[{"article_id":15566,"title":"How this global PC maker is vouching on BI and analytics to be a “consumer-first” company","content":"We recently interacted with Santosh Nair, Head-Analytics, Business Intelligence and Visualization at Lenovo, a global tech company spread across 160+ countries. With over 15 years of experience, he has built and lead complex analytics projects for many Fortune 50 companies in the telecom, retail, ecommerce and manufacturing industries. He began by stressing on the fact that Lenovo believes in the concept of different is better and that they constantly innovate to improve the overall customer experience. “We keep customer at the center before launching a product portfolio”, said Nair, who has been heading Lenovo over the last five years and developing competencies in customer analytics, real time recommendations, big data and analytics. Data driven decisions at Lenovo—from Excel to Tableau and other BI tools This maker of PCs, thin & light laptops, premium laptops, notebook, smartphones and tablet, has been constantly adding innovative products into its portfolio and for this year, the company’s focus is majorly on smart devices and cloud. Nair says that a lot of people are using analytics to carry out suitable decision making using data. “The challenge however here is that many a times practitioners do not believe in data or they do not want to adopt new technology”. Irrespective of this fact, Lenovo has been quick to evolve in terms of its analytics capabilities. Starting the journey of data analytics in 2007 with Excel, the company has invested a lot of money on big data technologies, visualization & BI tools for making data driven decisions. “As we moved forward, we considered various options before finalizing these tools. We adopted Hadoop, Python, Spark, Tensor Flow as we grew in terms of doing advanced analytics”, shares Nair, who currently heads a 25+ member team under him at Bangalore office. “We started with Tableau in 2014 with about 100 users that had access to the dashboards. Over the years we have integrated enterprise platforms and have started to publish our dashboards online, where people can ask questions in real-time by clicking different buttons”, shares Nair. By integrating 16 core enterprise serves, it now gives access to more than 10,000 users with access to different kind of dashboards. They have integrated 40 different dashboards into a single master dashboard. “This is a gist of how visualization platform looks like in terms of Tableau and enables decision making at Lenovo”, he adds. Tableau recently announced the general availability of Tableau 10.3. This latest release would help organizations achieve data-driven insights faster than ever, through automated table and join recommendations powered by machine learning algorithms that simplify the search for the right data for analysis. “Earlier my team used to send 40+ excel based dashboards, for each country, sub-region, regions and then accumulate the overall findings into a global level. This humongous task would require 5-6 people, which after implementation of Tableau and big data solutions in the back-end has completely become automated. The data can now be seen within a lag of 30-40 minutes, unlike a week earlier”, Nair explains. That’s not all, it also allows creating different views for different stakeholders. For instance, executives don’t want to look at detailed numbers but only fancy numbers to check if the company’s budget and targets are on track. “In such cases we have executive level view. We have created different views for different levels based on their procurement”, he says. He also adds that in BI space, nothing is constant. “You are on the move and need to act on different questions asked and find new solutions. Tableau has helped us achieve that. It has also enables the global team at Lenovo to do centralized jobs”, he says. Overall, Tableau and big data can reduce the cost, increase efficiency and process data in the real-time. Tableau vs. other BI tools- “My team uses tableau a lot but it is not limited to just tableau. Different business units within the company uses different technologies based on specific needs—supply chain might use Qlik or marketing team might be using DOMO”, says Nair. At high end at Lenovo, we have tableau, Qlik and domo. LUCI—the hybrid cloud based platform by Lenovo Internally at Lenovo, the analytics team builds capabilities in terms of data management, getting data from various sources, doing the automation, visualization, and analytics. The end-to-end analytics adoption at their 30+ business units is different, which depends on the overall business requirement. Lenovo Unified Customer Intelligence (LUCI), an analytics platform developed by Lenovo’s global intelligence team, is capable of delivering actionable insights from a myriad of global sources in a matter of seconds. “If you want to handle big data and provide real-time decision making, it cannot be done using basic applications alone. We need to have platforms that provide scalability, flexibility and performance. LUCI, is an answer to that and can support all the analytics and BI activities at Lenovo”, shares Nair. Sentiment Scoring and Voice assistant—AI and machine learning play at Lenovo Not just on a prescriptive of cognitive platforms, Lenovo is also implementing voice based\/ assisted technologies that can provide recommendation to the end customers and can chat with them. “We are using a lot of machine learning techniques and artificial intelligence algorithms to bring this innovation”, says Nair. As customer centricity remains a major focus at Lenovo, they focus on mapping out ways on how customers engage with Lenovo—throughout the process of making primary research, making a purchase or giving feedback. Nair explains, “To focus on customer centricity we have created one unique dashboard by integrating 30 different data sources. All these data sources talk about customer feedback which can be from social media, retailer website or surveys. We go through all of this data and run natural language processing to understand what is the theme or how are customer’s sentiment is about Lenovo products”. They use NLP processing technology to categorize the comment and feedback into different themes and do a sentiment scoring in real-time. This data goes to the dashboard which can be accessed by anyone in the organization with a Lenovo active directory credential. Based on customer feedback, there are two things that can be done—one is theme classifications or categorization and other is sentiment scoring which can be done from -1 to +1 or -5 to +5. Future plans for analytics at Lenovo- Nair shares the following points- Mobility- Lot of people are demanding mobility as one of the components, where they can access the dashboards on their laptop or mobile devices. They are demanding more and more customizations to be done on the mobile and we are looking to enable some kind of authorization access so that they can also edit few things and publish. Scalability- Second thing is to be fast and scalable in the real-time. The data is always there but there needs to an alert mechanism to set out alerts suggesting if the sales went up or down etc. so that actions can be taken accordingly. Lenovo aims at enabling this kind of process in the decision making in the coming days. Voice BI- This is the third plan. There are few executives who are not well versed to go through dashboard and use it specifically for marketing organizations. They fancy numbers and want to get numbers as soon as possible. But there are certain executives who are more data oriented and they want to see everything in a great format. Serving different customers in different ways is the key and Lenovo is working on getting voice BI here.","excerpt":"We recently interacted with Santosh Nair, Head-Analytics, Business Intelligence and Visualization at Lenovo, a global tech company spread across 160+ countries. With over 15 years of experience, he has built and lead complex analytics projects for many Fortune 50 companies in the telecom, retail, ecommerce and manufacturing industries. He began by stressing on the fact […]","categories":["AI Features"],"tags":["Interviews and Discussions","is hadoop a company","retail bi prescriptive","tableau india"],"author_name":"Srishti Deoras","publish_date":"2017-06-13T09:45:01","publication_year":"2017","word_count":1248,"keywords":["Go","artificial intelligence","machine learning","AI","is hadoop a company","retail bi prescriptive","Scala","NLP","Python","Aim","analytics","R","tableau india","Interviews and Discussions"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","NLP","analytics","Aim","Python","R","Go","Scala"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/global-pc-maker-vouching-bi-analytics-consumer-first-company\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10082623,"title":"WNS Acquires The Smart Cube, OptiBuy to Strengthen Procurement &#038; Analytics Capabilities","content":"Global business process outsourcing firm WNS announced that it acquired The Smart Cube, AI and analytics firm focused on procurement and supply chain, alongside OptiBuy, an international consultancy specialising in procurement management. With this acquisition, the company looks to strengthen its offerings and capabilities in high-end procurement and advanced analytics. WNS acquired The Smart Cube for about $125 million. The company said the acquisition will add approximately $9 million to WNS’ net revenue in FY2023. OptiBuy, on the other hand, was acquired for $30 million. It is said that the acquisition is expected to add about $2 million to WNS’ net revenue in FY2023. Moreover, WNS has funded the up-front payments for these acquisitions with a combination of cash and £83 million of long-term debt. The Smart Cube chief Gautam Singh said that, over two decades, they had built a strong business based on delivering value for their clients by leveraging AI+HI (artificial intelligence + human intelligence) methodology. “We are delighted to join WNS in the next phase of our growth journey,” he added. Founded in 2003, The Smart Cube provides digitally-led market intelligence and analytics solutions. The company offers services in four key areas: digital and analytics, procurement and supply chain, commercial sales and marketing, and strategy and investment research. Earlier this month, The Smart Cube was certified as a Best Firm for Data Scientists by Analytics India Magazine. Headquartered in London, UK, and offices in Noida (India), Timișoara (Romania), Switzerland, and Chicago (USA), The Smart Cube has over 800 global employees, including a seasoned leadership team, and more than 600 R&A specialists. OptiBuy managing director Mateusz Borowiecki said that it believes that the acquisition will enable the creation of holistic digital procurement and supply chain solutions for the global marketplace. Founded in 2010, OptiBuy is one of the leading European providers of procurement management solutions. The company helps clients use the capabilities of third-party procurement and supply–chain platforms, including Jaggaer, Ivalua, and O9, and complements WNS’ existing offerings with platforms such as Coupa and Ariba. It also provides its clients with consulting, optimisation, outsourcing, and training services. Currently focused on the EMEA (Europe, Middle East, Africa) market, the company has about 90 employees. However, with the recent acquisition, WNS looks to expand these capabilities into the North American market as a significant opportunity. WNS’ chief Keshav R. Murugesh said that both companies possess capabilities that complement their existing procurement and analytics offerings. He said that The Smart Cube and OptiBuy would enhance their WNS-Denali (procurement) and WNS-Triange (analytics) solutions. WNS is one of the leading business process management (BPM) companies. With over 400 clients across various industries, the company offers a spectrum of BPM solutions, including research and analytics, industry-specific offerings, customer experience services, finance and accounting, human resources, and procurement. As of September 30, 2022, the company has about 57,503 employees working across 60 delivery centres globally, alongside facility centres in India, Sri Lanka, China, Canada, Costa Rica, the Philippines, Poland, Romania, South Africa, Spain, Turkey, the UK, and the US.","excerpt":"Earlier this month, The Smart Cube was certified as a Best Firm for Data Scientists by Analytics India Magazine.","categories":["AI News"],"tags":["Mergers and Acquisitions","Smart Cube","WNS Analytics"],"author_name":"AIM Media House","publish_date":"2022-12-17T13:15:57","publication_year":"2022","word_count":504,"keywords":["Go","artificial intelligence","Smart Cube","programming_languages:R","AI","programming_languages:Go","Git","RAG","analytics","WNS Analytics","Mergers and Acquisitions","R"],"extracted_tech_keywords":["AI","artificial intelligence","analytics","RAG","R","Go","Git","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/wns-acquires-the-smart-cube-optibuy-to-strengthen-procurement-analytics-capabilities\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10139525,"title":"Meesho Set to Open Source its ML Platform","content":"At the Nvidia Summit in Mumbai, eCommerce giant Meesho announced that it is open-sourcing its ML platform. Debdoot Mukherjee, head of AI at Meesho, said, “We will be open-sourcing our ML platform, including our feature store and model orchestrator. We believe this move will empower developers and data scientists everywhere to leverage the power of AI more effectively.” In a presentation titled “Cost Effective High-Performance Model Serving for the Masses,” Debdoot presented a deep dive into Meesho’s ‘massive scale’ machine learning models. He spoke about Feature Store, an architecture designed for high-performance machine learning operations. The architecture helps Meesho standardise and manage features for their machine learning models at scale, bridging raw data and their ML models. The presentation also revealed that this system could handle millions of requests every second, and 99 per cent of these requests are handled in under 10 milliseconds. Debdoot added that this Feature Store architecture is also set to be open-sourced. In a LinkedIn post, Debdoot said, “We focused on our journey of scaling transformer-based ranking models in production, highlighting the optimizations we’ve implemented with TensorRT to ensure we meet stringent latency and cost requirements.” A few days ago at Build With Meta Summit,  Meta revealed that Meesho was among many other consumer apps planning to implement Meta’s open-source Llama models inside their tech stack. Over the last few years, Meesho has proactively integrated generative AI, and machine learning capabilities into its platform. Internally, their large-scale decision engine has incorporated several features to enhance code maintainability, reduce system latency, and provide an improved error-handling system with multiple fallback solutions. Further, the decision engine also helps developers monitor system health by creating a dashboard to track and spot problems easily. On the user front, Meesho uses generative AI to simplify the listing process for sellers within the platform. They’ve also employed features to translate product descriptions, addresses, and voice searches into other Indian languages besides Hindi. Meesho also uses Microsoft’s Azure OpenAI to optimise call centre costs and operations, reportedly resulting in a 20% boost in customer effort scores. Last year, Meesho partnered with IISc to advance their research in generative AI and develop multimodal representation learning and other generative AI capabilities to enhance the user experience. In a conversation with AIM last year, Debdoot said, “At Meesho, AI and ML are fundamental components to solve a wide range of problems across every aspect of an e-commerce platform.” Meesho is one of India’s fastest-growing e-commerce platforms, with around 120 million active users. Open-sourcing, a machine learning platform with proven capabilities in handling high-throughput feature serving, may benefit the ecosystem, help developers build similar ML systems for their platforms, and solve existing challenges in their technology.","excerpt":"Their large scale, high throughput serving machine learning architecture will now be available to other developers.","categories":["AI News"],"tags":["Meesho","meesho machine learning","meesho ml platform"],"author_name":"Supreeth Koundinya","publish_date":"2024-10-27T13:16:21","publication_year":"2024","word_count":450,"keywords":["Go","machine learning","Meesho","AI","OpenAI","R","ML","RAG","Aim","generative AI","Azure","meesho machine learning","meesho ml platform"],"extracted_tech_keywords":["AI","machine learning","ML","generative AI","OpenAI","Aim","RAG","Azure","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/meesho-set-to-open-source-its-ml-platform\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10094405,"title":"6 Reasons Why We Won’t Get GPT-5 Anytime Soon","content":"OpenAI has been criticised for its opaque decision-making process. Unperturbed, CEO Sam Altman has graced the world with some valuable insights. In an interview with Humanloop, the oft-criticised celebrity CEO spoke about the future of OpenAI. While this interview shed some light on some of OpenAI’s most profitable products, it also reflected a culture of overpromising and under-delivering. While the interview itself presents a few surface-level points, it also delves into some of the scaling issues that OpenAI has been facing. Along with this, some of Altman’s statements in this interview directly conflict with his statements in the past, creating more opacity for ‘OpenAI’. However, one thing is for certain, GPT-5 isn’t coming anytime soon, not even in 2024. Here are 6 reasons why we won’t get GPT-5 any time in the near future. Not enough GPUs In the interview, Altman delved into the scaling issues OpenAI is facing. According to him, OpenAI is ‘extremely GPU-limited’, which has rippled into delays for their short-term plans. This means that plans like longer context lengths (more than 32,000 tokens) and fine-tuning for specific dataset will be delayed. Due to the shortage, even the basic functionality of the API has been compromised. According to the interview, OpenAI is offering a dedicated capacity offering, which will offer faster and more dependable API access. However, to access this, customers have to commit to a $100,000 spent on the platform. A bigger focus on GPT-4 While GPT-4 was one of the biggest upgrades for OpenAI, it is clear that Altman isn’t done with this model. The first and biggest item on the roadmap was a cheaper and faster GPT-4. By improving its flagship model, OpenAI wishes to reduce the ‘cost of intelligence’ as far as possible. This means that the company will likely spend time and resources to optimise GPT-4 even further instead of focusing on GPT-5. In addition to this optimisation, OpenAI is looking to increase the context window to up to 1 million tokens. Other upgrades on the roadmap include fine-tuning APIs and a stateful API for GPT-4. Restricted by scaling laws Altman has been vocal in the past about how the size of LLMs won’t matter as much moving forward. However, it seems that OpenAI’s researchers have changed his mind. According to the interview, its internal data has found that scaling laws continue to hold well. This means that GPT-5, when it releases, will not be as big of a leap as GPT-2 to GPT-3.5. This is because models will likely ‘double or triple’ in size, as opposed to changes in the order of magnitude. Promised features first Multimodality was one of the biggest new features of GPT-4. Demos of the model being capable of generating a whole website from just a sketch set the Internet on fire, but it seems even ChatGPT Plus users have to wait for multimodality. In the interview, Altman stated that multimodality will only come in 2024, blaming the lack of GPUs. Until multimodality comes to GPT-4, GPT-5 is still a far-off dream. ChatGPT is the last product Altman also stated that ‘OpenAI would not release products beyond ChatGPT’. This is mainly because OpenAI wants to be safer for competing with their customers with other offerings. He also stated that there was a history of platform companies having a killer app. This solidifies OpenAI’s future goals as a platform company rather than a product company. As time goes on, their platforms will only get more capable. This can also be interpreted to mean that GPT-5 might only get more delayed in favour of better platforms. Regulatory risks Sam Altman has also been very vocal when it comes to AI regulation. He is also one of the biggest signatories for the new AI safety letter which advocates for more stringent regulation of AI, citing existential risks. While Altman said that he continues to believe in the importance of open source, he also defended OpenAI not releasing their models to the public. This is mainly due to his skepticality towards the large number of companies that have the capabilities to host and serve LLMs. With the host of problems OpenAI is facing to scale up to the success of ChatGPT, it seems that other projects might be put on hold. While even the GPT-4 API has not been updated for the past 2 months, GPT-5 is still a long-distant dream until OpenAI gets it together.","excerpt":"Here are 6 reasons why we won’t get GPT-5 any time in the near future.","categories":["AI Trends"],"tags":[],"author_name":"Anirudh VK","publish_date":"2023-06-03T10:00:00","publication_year":"2023","word_count":735,"keywords":["Go","ChatGPT","API","OpenAI","AWS","GPT-5","AI","GPT","AI safety","R"],"extracted_tech_keywords":["AI","GPT-5","ChatGPT","OpenAI","AWS","R","Go","API","GPT","AI safety"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/6-reasons-why-we-wont-get-gpt-5-anytime-soon\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":34307,"title":"The Out Of Africa Inhabitants Are The New Species of Hominids, Which AI Accidentally Discovered","content":"A scientific finding that’s cropping up perhaps upstage the globally believed theory of evolution. Scientists and evolutionary biologists from Estonia and Spain have used deep learning to build a revolutionary model for gauging human evolutionary patterns. Neanderthals and Denisovan bred with humans have always been a well-known fact, but now a third species has been uncovered. History reminds us that humans came into existence when a Denisovan father and Neanderthal mother bred and a “homosapien” was born as their child. The findings of the research point now that we are a mix of all these three species. According to Jaume Bertranpetit, an Institute of Evolutionary Biology one of the lead researcher say that about 80000 years ago, the so-called Out of Africa occurred, when part of the human population, which already consisted of modern humans, abandoned the African continent and migrated to other continents, giving rise to all the current populations. He further added that the modern humans crossbred with Neanderthals in all the continents, except Africa, with the Denisovans in Oceania and probably in South-East Asia, although the evidence of cross-breeding with a third extinct species had not been confirmed with any certainty. How AI Found Out The Third Species By combining deep learning algorithms and statistical methods, investigators identified the third species in the genome of Asian individuals, the footprint of a new hominid who cross-bred with its ancestors’ tens of thousands of years ago This finding explained that the hybrid found in the caves of Denisova,  the offspring of a Neanderthal mother and a Denisovan father was not an isolated case but rather was part of a more general introgression process Evidence in the form of demographic analysis, one churned by deep learning and sprinkled with some statistical analysis. Their algorithm devised and compared numerous complex demographic models to make predictions about the history of interbreeding events in Eurasia Deep Learning made it possible to make the transition from DNA to the demographics of ancestral populations Then the researchers fed the algorithm, multiple sets of genome sequences that were derived from both the contemporary and ancient DNA, which enabled the neural network to create a large set of possible demographic histories. After that statistical analysis calculated which among these histories were the most probable ones The piecing together of this genetic jigsaw puzzle yielded the proposed “third archaic introgression”, meaning a third hybridisation event among ancient Why Was Deep Learning Used? As demographic models analysed were much more complex than anything else till date and as there were also no statistic tools available to analyse them. The only alternative that could reciprocate similar capability was Deep learning, which as an algorithm can help in imitating the way in which the nervous system of mammals works, with different artificial neurons that specialised and learned to detect, in data, patterns that are important for performing any given task. So, the researchers used the very same capability of its to get the algorithm to learn to predict human demographics using genomes obtained through hundreds of thousands of simulations. This research is also touted as first to successfully use deep learning to explain human history, biology, genomics and evolution. Interesting Facts Deep learning was used to teach the algorithm to predict human demographics through hundreds of thousands of simulations. The research also bolstered a suggestion that the two extinct species often bred enough to produce a genetically distinguishable hybrid population and concluded that Ancient Human Groups Mated With The Mysterious Denisovans At Least Twice. Genetic analysis conducted in the research also suggested two populations of Denisovans. An extinct group of hominids closely related to Neanderthals existed outside of Africa during the Pleistocene and both those populations interacted and interbred with anatomically modern humans. Dried deep within the DNA of Asian individuals was the genetic clue that pointed towards the existence of an unknown human ancestor and the startling fact was that it was not the humans who reached this startling conjecture, but rather an artificially intelligent algorithm which achieved this feat. Scientists also extracted Denisovan DNA from a well-preserved finger bone which was found in the Siberian cave. Outlook Deep learning is acting as a key to decipher the secrets of human evolution, recessively registered in four walls of ancient DNA. Even though, while the study on the history of interaction between modern humans and archaic hominin contemporaries remains something to vouch for, one should restrain from making claims of genetic mixing from an unknown extinct archaic hominid population.","excerpt":"A scientific finding that’s cropping up perhaps upstage the globally believed theory of evolution. Scientists and evolutionary biologists from Estonia and Spain have used deep learning to build a revolutionary model for gauging human evolutionary patterns. Neanderthals and Denisovan bred with humans have always been a well-known fact, but now a third species has been […]","categories":["AI Features"],"tags":["AI and Deep Learning","DNA"],"author_name":"Martin F.R.","publish_date":"2019-01-30T08:10:40","publication_year":"2019","word_count":748,"keywords":["Go","API","programming_languages:R","AI","AI and Deep Learning","neural network","BERT","Aim","deep learning","llm_models:BERT","R","DNA"],"extracted_tech_keywords":["AI","deep learning","neural network","Aim","R","Go","API","BERT","llm_models:BERT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/ai-hominids-new-species\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10096152,"title":"Open Source Chatbots Are Nowhere Close to ChatGPT","content":"Every day we see a new chatbot — either built by a big-tech\/recently funded company, or from the open source community. In the race to replicate OpenAI’s ChatGPT, developers have been taking a lot of shortcuts. The most common one these days is training the chatbots on data generated by ChatGPT. The most recent chatbot claims to outperform ChatGPT. OpenChat, an open source chat alternative that is touted as decentralised, recently achieved a score of 105.7% compared to ChatGPT on Vicuna GPT-4 Benchmark. A huge feat, but not true when looked at closely. This is the second model that claims to perform better than ChatGPT on the same Vicuna Benchmark. Earlier, Orca, a 13-billion parameter model, which was also trained on GPT-4 data, claimed to outperform OpenAI’s model. Same-old, same-old To start with, OpenChat is built on top of LLaMA-13B. This means that the model is already not up for commercial use as Meta’s LLaMA is for research purposes only, and not up for commercial use. Furthermore, there is another thing that should be considered before boasting about the benchmarks of the model — the dataset used for fine-tuning. This LLaMA-based model is trained on a set of 6k conversations from the 90k conversations available on ShareGPT, a hub for sharing ChatGPT and GPT-4 generated outputs on the internet. When it comes to evaluating and benchmarking on Vicuna GPT-4 Benchmark, it only tests for style and not the information generated by the model. Furthermore, it is a GPT-based evaluation metric, which means that any model trained on ChatGPT or GPT-4 data will be rated higher when tested by GPT, making the benchmarking untrustworthy. Recently, Hugging Face found a similar problem with other open source models as well. The founders of Hugging Face claimed there was a lot of discrepancy between the evaluation benchmarks put up on the papers of the models and when the models are evaluated on the Hugging Face benchmarks. David Hinkle, VP of software engineering at Securly, pointed out that a lot of recent models that claim they are outperforming LLaMA or GPT-4 are nowhere on the Open LLM Leaderboard. Tall claims, short results In short, it is a big claim to make that a model trained on ChatGPT data outperforms the same when benchmarking on a metric built on top of the same model. For an analogy, it is like a student giving an exam, rewriting the answers to match with the correct answers provided by the teacher, to which the teacher matches the answers again. Obviously, it is going to perform better. Andriy Mulyar, from Atlas Nomic AI, points this out saying this is all just a false hype. People imitating ChatGPT using ChatGPT generated output is a false path to follow. Furthermore, the only thing that these models are copying is the style of ChatGPT, making the chatbots’ quality sound better on individual tasks. If evaluated across the board considering all types of general tasks, ChatGPT is a much better assistant than any other. Interestingly, after all the criticism, the researchers have realised that there is some problem with evaluating the model on Vicuna GPT-4 Benchmark. Hence, they have transitioned to MT-bench, to test OpenChat’s performance. In this case, the model performed significantly poorer than GPT-3.5 based ChatGPT, describing the discrepancy between the evaluation benchmarks. Users on Twitter have been pointing out that the model hallucinates even more than ChatGPT, and more than just the evaluation metrics used for the models. “I just tried this model and it is not good at all. Did you even try the model before posting this?” said a Twitter user. Thanks to GPT Whatever the metrics and benchmarks, one thing is getting increasingly clear about LLM-based chatbots — high-quality data works wonders. For this, the only model that should be thanked is ChatGPT, as every model today is being trained on synthetic data generated by the chatbot. No one has the secret sauce that OpenAI made for GPT. Recently, OpenAI was asked if open source would be able to replicate what the company has built through Vicuna or LLaMA, to which, Ilya Sutskever replied with a negative. The trend of  “this new model beats all other models in benchmarks” has been going on for a while now, but when evaluated on the same metrics as the others, “the news model” fails to perform. Moreover, even though the open source community has been trying to replicate ChatGPT, training it on ChatGPT’s data might not be the best way forward, as OpenAI is already under several lawsuits for training on internet data.","excerpt":"Researchers are making tall claims just to grab headlines, and deliver nothing in the end.","categories":["AI Features"],"tags":["ChatGPT","LLMs","Open Source AI","OpenAI"],"author_name":"Mohit Pandey","publish_date":"2023-07-03T13:10:01","publication_year":"2023","word_count":762,"keywords":["Go","ChatGPT","Hugging Face","TPU","OpenAI","AI","LLMs","chatbots","AWS","Open Source AI","Aim","R"],"extracted_tech_keywords":["AI","ChatGPT","OpenAI","Aim","Hugging Face","chatbots","AWS","TPU","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/open-source-chatbots-are-nowhere-close-to-chatgpt\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10089900,"title":"‘Pain in the AIs’ by Nick Bostrom","content":"Have you ever questioned the very fabric of your reality? Do you ever wonder if what you perceive as the real world is nothing more than a computer-generated simulation controlled by a superior artificial intelligence? The theory was popularised by Nick Bostrom as the simulation hypothesis, suggesting that our very existence is no different than that of a character in a video game. Bostrom is an acclaimed thinker on the safety concerns associated with our march towards the increasingly powerful and general forms of AI. The polymath believes AI is “the single-most important and daunting challenge that humanity has ever faced”. Currently, he is a professor at the University of Oxford and director of the Institute for the Future of Humanity. Over the years, his work has profoundly impacted minds such as Stephen Hawking, Bill Gates, and Elon Musk. In an exclusive interview with Analytics India Magazine, Bostrom expresses his concerns on artificial intelligence. “If we manage to do things right, the upsides are fantastic,” said the Swedish philosopher, highlighting three broad concerns in AI — value misalignment, automation bias, and asymmetries. What Keeps Bostrom Awake At Night? Bostrom told AIM that there is the alignment problem, which refers to the challenges caused by the fact that machines do not have the same values as humans. “How do we ensure that highly cognitively capable systems — and eventually superintelligent AIs — do what their designers intend for them to do?” he pondered. Bostrom has delved deeper into the unsolved technical problem in his book ‘Super Intelligence’ to draw more attention to the subject. Meanwhile, the most infamous instance of misaligned AI happens to be Meta’s racist BlenderBot that hated Mark Zuckerberg. The professor’s second concern is the relatively new concept of governance. It’s the problem that stems from the assumption that AIs will be under complete human control, ensuring technology is predominantly used for positive ends and the benefits are widely shared. For instance, the problematic deepfakes, which have surfaced on the internet. Lastly, there is the problem of ensuring that if we are creating digital minds that have moral status, those minds are themselves treated with due ethical consideration. The bizarre incident of users acting abusively towards AI chatbot Replika was highlighted less than a year ago, along with many similar cases. “In addition to protecting human interests, we must also ensure that digital minds don’t suffer and that their interests are taken properly into account,” Bostrom suggested. The solution Most AI companies today are making efforts to work on these concerns.  The first one that comes to mind when we talk about ‘aligned research’ is OpenAI, where most of its models follow human intent, along with human values. “At the time [2014], this problem was almost completely neglected, but it is now becoming increasingly recognized by more mainstream AI researchers,” resonated Bostrom. Even Google has a 34-page elaborate document on ways the tech giant is tackling the issue of AI governance. “At the meta-level, I think it would be desirable to try to work through these moral questions in a more thoughtful and curiosity-driven way than is often done,” he said while suggesting to look at these problems on a case-by-case basis. He also thinks many people are too quick to pick a side, and then seek to demonise those who take the other side. “We’d do better if we try harder to listen, reflect, and try to apply the strictest criticism to our own views,” he added. Explaining the reason why Bostrom works to make complex and abstract concepts accessible, he said, “Well, the issues at stake concern us all. That is one reason to try to make them as widely accessible as possible. Another is that no one group of experts has all the relevant knowledge and perspectives.  A third is that sometimes, when one tries to explain things clearly in the non-technical language, it can help one understand things better oneself.” This makes total sense as several AI researchers have been advocating for a collaborative effort rather than a closed-door approach in the field. “Development by smart and well-resourced people behind closed doors can make great things, but development in the open by a huge community of people is just a more effective and equitable mode of development,” Colin Raffel, a faculty researcher at Hugging Face and professor at UNC Chapel Hill, had earlier told AIM in an exclusive interview We’re All Trumans The idea that our universe, including ourselves and our innermost thoughts, is a computer simulation, has permeated the culture high and low. In an influential essay in 2003, Bostrom proposed the idea that “technologically mature” civilizations could use a tiny fraction of their computational power to probably explore their histories. The co-founder and CEO of Tesla, Elon Musk, echoed this idea once declaring that there was only a one-in-a-billion chance that we lived in “base reality”. https:\/\/www.youtube.com\/watch?v=xBKRuI2zHp0 Recalling the essay, Bostrom said, “So conditional on there being such civilizations that have an interest in doing this, we should think we are likely to be among the typical simulated minds, rather than the rare non-simulated minds, given that from the inside it would not be possible to tell the difference.” Currently, Bostrom is working on an interesting book project, which he did not reveal much about. But he said that he is trying to understand what world order could look like in which humans and AIs and all kinds of digital minds live together harmoniously. This might well be a teaser to his next book!","excerpt":"In an exclusive interview with Analytics India Magazine, Bostrom expresses his concerns on artificial intelligence","categories":["AI Features"],"tags":["AGI","future of AI","Interviews and Discussions","Superintelligence"],"author_name":"Tasmia Ansari","publish_date":"2023-03-23T16:00:00","publication_year":"2023","word_count":922,"keywords":["Go","Hugging Face","artificial intelligence","OpenAI","AI","Git","future of AI","Aim","Superintelligence","AGI","analytics","AI governance","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","artificial intelligence","analytics","OpenAI","Aim","Hugging Face","R","Go","Git","AI governance"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/pain-in-the-ais-by-nick-bostrom\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10058423,"title":"Council Post: Collaborating with the government—Lessons &#038; experiences","content":"My experience of working with different governments in solving real-life problems has been a great eye-opener. Such efforts bring out a different flavour in solving complex technical problems as we are not addressing the typical enterprise use cases of profitability, customer analytics, cost savings, etc. Instead, the prime focus of our interactions is to improve citizen-centric services, save the environment or uncover smarter ways to deal with a crisis by leveraging analytics and AI. Considering the variety and complexity involved, we come away with a bunch of learnings from each project. Key learnings include: Adopt a value-driven approach Technology providers are expected to facilitate the success of the vision and objectives of government programs and initiatives. To that end, the ability to turn good ideas into deliverable solutions is critical. Innovation that responds to consumer and societal needs is what ultimately drives competitiveness, value creation, and sustainable growth. It is essential to adopt value-driven technology rather than adopt technology first and later focus on the value side. It requires tech providers to be patient and navigate interdepartmental dependencies and manage expectations while building value-based technology solutions. Scale fast, scale well India’s population currently stands at a staggering 1.3 billion. When we are building solutions in collaboration with the government, we should be aware of the massive scale we are working at. With scale comes the responsibility of managing undefined complexities stemming from ambiguous internal and external forces. We are working with Ayushman Bharat PM-JAY, the world’s largest government-supported health scheme providing a health cover of Rs. 5 lakhs per family a year to cover secondary and tertiary care hospitalisation of 10.74 crore poor and vulnerable families across India. The scale of data (both structured and unstructured) generated on a daily basis to identify suspicious patterns in-hospital treatments and beneficiary activity is overwhelming. Similarly, projects like the Aadhar UIDAI is a huge challenge in terms of scale and managing unforeseen complexities. Managing this sort of scale means both the solutions as well the people behind them must be held to a high standard. Problem-solving with next-gen technologies We have seen government departments driving digital adoption, be it cloud, IoT or AI\/ML. The vision of today’s officials is really forward-looking. Rajasthan’s tiger reserves deploying AI for 24×7 automated surveillance to reduce poaching and human-animal conflict is a good case in point. Such projects are unprecedented. In other words, technology partners have the opportunity today to show what’s possible in the near future by leveraging tech and co-opting the larger vision. Quality of data Creating a data-driven culture is essential for an organisation’s success. While it is enriching to work with governments, especially on projects like smart cities or citizen-centric initiatives, inherent challenges in terms of the lagging or disconnected processes, data unavailability, siloed system application, lack of or no digitisation often crop up. Data is a key ingredient for analytics projects, and hence clearly defining the quality of data is seminal. Working with the government early into the project to ensure we have data aligned, cleansed and profiled to deploy the correct analytics and AI capabilities is pivotal to the project’s success. In times of uncertainty We live in highly uncertain times, and it’s imperative to reimagine how your existing solutions can be implemented in new ways to make a difference. Both partners need to collaborate and build innovative solutions that stand the test of time. This article is written by a member of the AIM Leaders Council. AIM Leaders Council is an invitation-only forum of senior executives in the Data Science and Analytics industry. To check if you are eligible for a membership, please fill the form here.","excerpt":"My experience of working with different governments in solving real-life problems has been a great eye-opener. Such efforts bring out a different flavour in solving complex technical problems as we are not addressing the typical enterprise use cases of profitability, customer analytics, cost savings, etc. Instead, the prime focus of our interactions is to improve […]","categories":["AI Features"],"tags":["collaboration","Government","Indian government","Indian government using AI"],"author_name":"Noshin Kagalwalla","publish_date":"2022-01-17T10:00:00","publication_year":"2022","word_count":607,"keywords":["data science","Go","AI","ML","collaboration","Government","Git","RAG","Aim","Indian government","analytics","GAN","Indian government using AI","R"],"extracted_tech_keywords":["AI","ML","data science","analytics","Aim","RAG","R","Go","Git","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/council-post-collaborating-with-the-government-lessons-experiences\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10076549,"title":"Is ASML the Most Important Tech Company in the World?","content":"Today, almost all electronic devices are powered by silicon-based chips. From the device on which you are reading this article to the car you own, everything is powered by chips. These chips power the data centres and also many military technologies. While Intel, Samsung or TSMC are well-known names in the semiconductor space, ASML Holdings is probably the most important and much less known of them all. ASML is the only company in the world that makes these highly sophisticated machines used in chip making. Based in the Netherlands, Advanced Semiconductor Materials Lithography (ASML) has often been dubbed the most important technology company of our time. To put things in perspective, without ASML, there are no chips, and without chips, there is no progress. Chris Miller, author of the book Chip War: The Fight for the World’s Most Critical Technology, claims that microchips are the new oil. While World War II was decided by steel and aluminium, chips will decide the next phase of human history. ASML, which started as a subsidiary of Philips, sells its machines to Intel, TSMC and Samsung. Founded in 1984, ASML’s profit has soared in the last couple of years, with the company valued higher than Intel. So what led to the rise of ASML? EUV Lithography system (Source: ASML) EUV Lithography—ASML’s Goliath ASML has a monopoly on the fabrication of Extreme Ultraviolet (EUV) lithography machines because each one of them is among the most complicated devices ever made, according to Miller. EUV lithography is a new state-of-the-art technology developed by ASML. Its closest competitors, Nikon and Canon, are not working on this technology, and experts believe that it could take them decades to crack EUV lithography. This further establishes ASML as the only company that makes EUV lithography systems in the whole wide world. “EUV light occurs naturally in outer space. But to make EUV lithography possible, we needed to engineer a way to create such light within a system. So, we developed a radically new approach to generating light for lithography,” ASML said on its website. (Source: ASML) The technology is very complex and produces light wavelengths of 13.5 nanometers (billionths of a metre). The reduction in size is almost 15 times when compared to deep ultraviolet (DUV) lithography, which uses 193 nanometer light. The EUV lithography system uses powerful lasers and a system of complex mirrors—the flattest material on Earth, according to ASML—to etch integrated circuits on silicon wafers. “The TWINSCAN NXE:3600D is ASML’s latest-generation lithography system, supporting EUV volume production at the 5 and 3 nm Logic nodes and leading-edge DRAM nodes,” the company said. DUV lithography system-TWINSCAN NXT:2050i (Source: ASML) ASML Monopoly Its monopoly has been due to the exploitation of photolithography to an extreme level. Each EUV lithography system made by ASML costs around USD 200 million and is made of thousands of components acquired from nearly 5000 different suppliers. The machines, which are the size of a double-decker bus and weigh around 180 tonnes, are also made up of seven different modules, built around ASML’s manufacturing sites spread across more than 60 locations on three continents. The modules are then shipped to Veldhoven, where they are assembled, tested, and then again disassembled for delivery. The exorbitant costs involved means not many can afford them, and not many can afford to make them either. With its competitors years away from cracking the technology, ASML has a monopolistic hold on the market, and the company has already begun working on the next generation of lithography systems. However, competition might be imminent. Recent reports suggest that China might have cracked the lithography code. It is common knowledge that China wants to establish an autonomous semiconductor supply chain within the country to hedge against US sanctions and growing geopolitical and supply chain risks. Earlier this year, the US blocked the sale of EUV lithography machines to China. However, SMIC, China’s biggest semiconductor manufacturer, indigenously produced 7 nm chips using DUV lithography and is now working on advanced 5 nm chips. Chinese chipmaker SMIC, which recently shocked the US by announcing that it had produced 7-nm chips despite being denied access to EUV equipment, is now reported to be advancing to more advanced 5-nm. SMIC has also started construction of a new 300mm wafer fab. https:\/\/t.co\/az6YJ2pLJB pic.twitter.com\/r4AbcCNdS6— Maya Majueran (@MayaMajueran) October 4, 2022 Lithography has been the weak link in China’s semiconductor ecosystem, but now they seem to have cracked the code. Some experts in China believe that the country will be able to achieve a key breakthrough in EUV lithography in less than five years. “I think China also would love to develop their own EUV competency, their ecosystem for these things. I think it’s going to be very difficult for them to do that, frankly,” JSR chief executive Eric Johnson told the Financial Times. However, earlier this year, ASML alleged that SMIC might have infringed its trade secrets. With China in the lithography picture, ASML’s monopolistic hold in the market could very well be at stake. High-NA EUV lithography ASML has been working on high-NA (​​numerical aperture) EUV scanners, the follow-up to its EUV lithography systems. Even though it is in the R&D phase, it could help ASML stay ahead in the game. Still in R&D, the new high-NA EUV system features a 0.55 NA lens capable of 8 nm resolutions, compared to 13 nm for the existing tool. These new machines are expected to cost more than USD 300 million. However, the new technology is not expected to move into production before 2025. Interestingly, in an interview, Martin van den Brink, chief technology officer at ASML, said that the high-NA EUV lithography could be the end of the game. Besides the occasional debate around the demise of Moore’s Law, the end of the line could have a significant impact on the very future of ASML. Without shrink, there is no innovation and could this mean, ASML will cease to be an innovation-driven company? While it’s too early to speculate, ASML has a bright future ahead, at least for a considerable period of time. Chip makers are ramping up production and are lining up at ASML’s office for new machines.","excerpt":"ASML is the only company in the world that makes EUV lithography systems.","categories":["AI Features"],"tags":[],"author_name":"Pritam Bordoloi","publish_date":"2022-10-06T17:00:00","publication_year":"2022","word_count":1029,"keywords":["Go","programming_languages:R","AI","innovation","ML","programming_languages:Go","Aim","R"],"extracted_tech_keywords":["AI","ML","Aim","R","Go","innovation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/is-asml-the-most-important-tech-company-in-the-world\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10161611,"title":"IIT Madras Professor Balaraman Ravindran Elected AAAI Fellow for Advancing AI Research in India","content":"Professor Balaraman Ravindran of the Indian Institute of Technology (IIT) Madras has been elected as a fellow of the international scientific society Association for the Advancement of Artificial Intelligence (AAAI). The AAAI Fellows Program honours individuals who have made substantial, sustained contributions to AI, typically over a decade. Each year, a select group of members is nominated by their peers, and the nominations are reviewed by a committee. “Honoured and humbled to be elected an AAAI fellow this year, along with an illustrious cohort,” he said in a LinkedIn post. Ravindran was elected alongside professor Sriraam Natarajan from the University of Texas at Dallas (UTD). “Humbled and honoured to be elected as [an] AAAI Fellow! To be walking among giants is a dream come true. Many of the folks in this list are people that I have always looked up to (and still do),” he said on Linkedin. Natarajan is a prominent professor of computer science and the director of the Centre for Machine Learning at UTD. He has made significant contributions to the field of artificial intelligence (AI) and machine learning (ML), particularly in statistical relational learning, reinforcement learning, and their applications in healthcare. Ravindran’s election reflects his longstanding impact on AI research and education in India, particularly during a time when ML was not widely recognised in the country. He has been instrumental in advancing AI education in India. He founded the Wadhwani School of Data Science and Artificial Intelligence (WSAI) at IIT Madras and has played a key role in developing curricula that integrate machine learning into various academic programs. Ravindran heads the Department of Data Science and Artificial Intelligence (DSAI), the WSAI, the Robert Bosch Centre for Data Science and Artificial Intelligence (RBCDSAI) and the Centre for Responsible AI (CeRAI) at IIT Madras. His research has focused on deep reinforcement learning, where he has made foundational contributions. One of them is the introduction of Markov Decision Process (MDP) homomorphisms – a framework that has influenced various applications in AI. His collaboration with students and researchers has led to numerous publications in top-tier conferences and journals. Notably, he co-authored a paper on deep reinforcement learning for action hierarchies that was accepted at AAAI 2017, when he mentored Perplexity AI co-founder Aravind Srinivas. “My first accepted conference paper was with him at the AAAI 2017; research we did together on deep reinforcement learning (DRL) for action hierarchies when I was an undergrad at IIT-M. He was super early to adopting DL and RL at a time when India hardly had any machine learning education and awareness to begin with,” Srinivas said. Ravindran’s work extends beyond academia. He has collaborated with industry leaders such as Google Research and IBM to apply AI techniques to real-world problems. His efforts have significantly contributed to fostering an AI research community in India, which has helped bridge theoretical advancements with practical applications. With over three decades of experience in the field, Ravindran’s election as an AAAI fellow marks a significant milestone in his career and underscores his role as a leading figure in artificial intelligence research in India.","excerpt":"Ravindran was elected alongside Professor Sriraam Natarajan from the University of Texas at Dallas.","categories":["AI News"],"tags":["Perplexity AI"],"author_name":"Siddharth Jindal","publish_date":"2025-01-17T13:33:19","publication_year":"2025","word_count":515,"keywords":["data science","Go","machine learning","artificial intelligence","AI","ML","Perplexity AI","BERT","responsible AI","AI research","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","data science","R","Go","BERT","responsible AI","AI research"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/iit-madras-professor-balaraman-ravindran-elected-aaai-fellow-for-advancing-ai-research-in-india\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10041284,"title":"Study: Analytics and Data Science Jobs in India: June 2021","content":"The analytics job report, presented by Analytics India Magazine (AIM), provides a broad study of the data science job landscape – covering profiles and roles that require analytics competencies and skills. The research provides a comprehensive view of the talent market that was affected by the Covid-19 pandemic and its recovery since the lockdown restrictions were lifted. The report highlights the upswing in analytics jobs in the post-pandemic world and factors that have led to it. This report also outlines which functional analytics skills and programming languages are currently most in-demand in the market. The report provides insights for recruiters and hiring companies to study the demand for skills across the analytics function. They can identify and close any capability gaps across workforces. By highlighting the talent hotspots in the country, the report enables organisations to build a steady talent pipeline from various regions or cities. Benefits and Key Takeaways We hope the report is a handy point of reference for aspiring professionals, industry experts, and think tanks. As a result of the unfortunate pandemic, the analytics space in 2020 witnessed changes and churn in terms of job losses and a slowdown in job creation. The accelerated adoption of data science\/AI technologies that began immediately after the 1st wave of the pandemic, and the renewed business activity of companies, which led to the hiring of analytics professionals, are crucial factors that have led to an upswing in data science jobs. The enthusiasm and optimism around the analytics function in India continue to remain high, and the industry is cautiously anticipating what the future holds for data science and jobs within this domain. Introduction This Job Study is a holistic analysis of the Indian Analytics and Data Science jobs and skills landscape. This study\/report covers the top job trends in the analytics space, specifically open jobs at a given point in time. It provides a detailed break-up and analysis of jobs by certain criteria, such as cities, industries, experience levels, technologies\/tools, and salary hierarchies. The report provides details and analysis of the types of companies that are recruiting key talent in the data science domain.  The report identifies reasons for significant shifts or changes in hiring trends across the criteria defined above. Please find reports from previous years here:2020 | 2019 | 2018 | 2017 Top Trends In Analytics And Data Science (Source for the job figures and trends below: Leading career portals, including Naukri.com, LinkedIn jobs section, and Monsterindia.com) Open Data Science Jobs Trend (April 2014 to June 2021) Although it is difficult to ascertain the exact number of open analytics job openings for a given month, around 137,870 open jobs were available at the time of reporting in the 2nd week of June 2021, according to the analysis carried out by AIMResearch. This is the highest number of analytics jobs advertised in India (since AIM Research started recording these numbers in April 2015)These open jobs are available jobs to be filled in the data analytics function at any given point in time and not the absolute number of new jobs created each month.The analytics jobs figure for June 2021 witnessed a 47.1% increase in open jobs requirement compared to the same period last year (June 2020).Last year, the pandemic caused a large decline in jobs after the declaration of the lockdown in March 2020. The open jobs fell to an average of 93,755 in June 2020.There was a drop or contraction in open jobs year-on-year in the months of June and July 2020 – 10.6% and 17.5%, respectively. This contraction is due to the recessionary environment that was caused by the unfortunate pandemic.  From a peak of 109,000 analytics jobs in February 2020, the open jobs figure fell to a low of 82,576 in June 2020, a drop of 24.2%.However, after June, the analytics job market took a U-turn once India started lifting its lockdown restrictions from June 2020.From June 2020 to June 2021, the number of open analytics jobs increased by 67.0%.The circumstances presented by the pandemic has led to an increase in the number of open analytics jobs post-lockdown.Many analytics professionals had to be laid off during the lockdown because of the economic uncertainty created by the pandemic. However, as companies recover from the Covid impact, many of these positions need to be backfilled, leading to a significant increase in the analytics jobs advertised post lockdown.To add to that, the pandemic has accelerated digital\/AI adoption to a great extent. The country observed a 45% increase in AI adoption rate, according to a PwC report. The report also said that around 94% of organisations in India now believe that AI will help create more opportunities than be a threat to their industry. Companies thus think that this is the right time to invest in improving their digital capabilities, increasing the demand for analytics professionals.In June 2021, India contributed to 9.4% of the total global analytics job openings. This is a significant jump from 7.2% of total global job openings in January 2020. India’s share of global analytics jobs has increased because of the following three factors:An evolving analytics domain in India evidenced by a year-on-year increase in funding in Indian analytics start-ups and development of niche analytics capabilities, including Data Analytics, Digital, Cloud Analytics, Artificial Intelligence, Robotics, and Natural Language Processing (NLP).Organisations are investing in the development of enhanced analytics capabilities in India. This has led to Domestic IT firm, Semiconductor and Electronics, International Technology, and Software corporations advertising a greater no. of open jobs in India (v\/s globally). India became a leading adopter of AI among all major economies (including the US, Japan, and the UK) amid the pandemic.The ten leading organisations with the greatest number of analytics jobs advertised in August are dominated by the MNC and Domestic IT & KPO organisations – Accenture, Capgemini, IBM India, Deloitte, Hewlett-Packard, Fiserv, Cognizant Technology, Eaton Technologies, Dell International, and Mphasis.97.0% of analytics jobs advertised in India are full-time, signifying the strengthening of the Indian analytics hiring market. Just 3% of the jobs form part-time, internship or contractual jobs – similar to January 2020 and the overall trend of 2019, when an approximate 97.5% full-time and 2.5% part-time jobs were advertised.The top designations advertised are Analytics Specialist, Data Scientist, Decision Science Manager, Business Analyst, Analytics Manager, Statistical Modelling, Statistical Analyst, Marketing Research Analytics, and Operations Analytics Manager. Analytics Jobs By Cities Share of Open Jobs By City In terms of cities, Bengaluru once again accounted for the maximum proportion of open jobs, contributing to 27.9% of analytics jobs in India in June 2021. This is a significant increase from August 2020, when the percentage was at 23.2%. Bengaluru remains the preferred hiring destination in India for analytics jobs for the last 4-5 years.This is evident because Bengaluru is home to the maximum number of analytics start-ups that received the highest funding for three consecutive years (after removing the single largest funding for 2019) to 2020 – translating to a greater proportion of job openings in the city.Moreover, most of the Technology, IT, and Boutique Analytics firms run most of their Digital, Analytics, and Data Science operations from the city, propelling the city to the top of the list of preferred locations for open jobs.Delhi\/NCR stood second, contributing 19.1% of the analytics jobs in India in June 2021, a slight decrease from 20.3% in August 2020 and from 21.2% in January 2020.The proportion of jobs in Delhi as of June 2021 has been the lowest since 2018, indicating a shift in location preferences for those companies advertising analytics jobs.Approximately 12.2% of the analytics jobs in June 2021 advertised were for the location of Mumbai, down from 14.8% in August 2020.  This decrease in proportion is largely due to the base effect of the total number of jobs. While the overall number of analytics jobs (across major cities) from August 2020 to June 2021 increased by 22.3%, Mumbai’s jobs increased by only 0.7%. Despite this, Mumbai still holds the third-highest proportion of analytics jobs as of June 2021. Analytics Jobs By Sector Or Industry (Excluding Analytics Jobs In The IT Sector) Share of Open Jobs by Industry Banking & Financial Sector: The financial sector continues to be the biggest influencer in the analytics job market, with 27.8% of all analytics jobs in June 2021 advertised by BFSI firms. However, this proportion has been seeing a declining trend since 2018 (when it was at 41.0%). There has been a sharp decline of 8.4 percentage points from August 2020 (when it was at 36.2%) to the current levels.Energy & Utilities: The proportion of jobs in the Energy & Utilities sector increased significantly from August 2020 to June 2021. The sector advertised 17.3% of the overall analytics jobs in June 2021, up from 9.5% in August 2020 and 10.6% in January 2020. Moreover, the June 2021 proportion has been the highest since 2018.Media & Entertainment: The Media & Entertainment sector also saw a sharp rise in analytics jobs from 9.3% in August 2020 to 17.3% in June 2021. The sector has observed a steady rise in the proportion of analytics jobs over the years as there is a greater demand for analytics in the digital media and online streaming sub-vertical.E-commerce: The proportion of analytics jobs advertised by the E-commerce sector dropped to 11.6% in June 2021 from 15.0% in August 2020. This comes after e-commerce saw the largest increase in the proportion of analytics jobs compared to other sectors from January 2021 to August 2021 (4.2 percentage points).Telecom Industry: The proportion of jobs in the Telecom sector decreased slightly to 4.3% in June 2021, down from 5.5% in August 2020, which in turn is an increase from 4.6% in January 2020.Pharma & Healthcare Industry: This sector saw decreased analytics jobs from 16.7% in August 2020 to 12.9% in June 2021. The hiring in this industry has seen a continuous upward trend, with analytics jobs increasing from 11.1% in 2018 and 12.6% in 2019 to 16.1% in January 2020. The decline from August 2020 to June 2021 is largely due to the base effect of the total number of jobs.Retail & CPG: The proportion of open jobs advertised for the Retail & CPG sector is 5.6% in June 2021, up from 3.8% in August 2020 and down from 6.8% in January 2020.Automobile: The proportion of analytics jobs advertised in the Automobile sector has shrunk to 3.2% in June 2021, from 3.5% in August 2020 and from 3.7% in January 2020. Moreover, the June 2021 proportion is the lowest since 2018.Travel & Hospitality: The sector advertised no analytics jobs in June 2021, down from a minor proportion of 0.4% jobs posted in August 2020 Experience Requirement For Analytics Jobs Share of Open Jobs by Work Experience The year 2020 began with an increase in the proportion of mid-senior and senior-level open jobs in analytics. This trend has continued till June 2021, with a greater proportion of senior-level jobs advertised.The proportion of jobs for 7 – 10 years’ experience range increased to 16.3% in June 2021, up from 14.9% in August 2020, 12.5% in January 2020 and 6.7% in 2019. This increase took place after a drop from 10.5% in 2018 to 6.7% in 2019.Similarly, the proportion of jobs for 10 to 15 years of experience increased to 12.5% in June 2021 from 11.0% in August 2020. Before this, the proportion of analytics jobs in this bracket doubled to 8.6% in January 2020 from 4.3% in 2019.The proportion of jobs for analytics professionals with greater than 15 years of experience also saw a big jump to 8.9% in June 2021, up from 4.9% in August 2020, 2.8% in January 2020 and just 1.0% from 2019This secular increase in the proportion of jobs for senior-level positions is due to the consolidation in analytics function in India:The trend of a greater proportion of jobs for experienced professionals endures despite an overall recessionary environment, signifying a maturing Analytics and Data Science domain in India.Almost one in ten (8.9%) analytics jobs were advertised for an experience level greater than 15 years. This highlights the hiring of leadership roles across the Data Science domain instead of roles focused on delivery or programme management. This is further emphasised by the combined proportion of 37.7% of jobs for analytics positions with experience greater than seven years (highest since 2018).Correspondingly, 62.3% of analytics job openings are for professionals with less than seven years of job experience. This shows a declining trend from 2019, and the proportion is the lowest since 2018:For the experience category of 5-7 years, the proportion of jobs has marginally increased to 18.7% in June 2021 from 18.6% in August 2020. This proportion was 19.3% in January 2020.For the experience category of 2-5 years, the proportion of jobs has decreased to 18.9% in June 2021 from 20.6% in August 2020 and 24.1% in January 2020.For the experience category of 1-2 years, the proportion of jobs has decreased to 13.0% in June 2021 from 16.6% in August 2020 and 17.9% in January 2020.For the experience category of 0-1 years, the proportion of jobs has decreased to 11.8% in June 2021 from 13.3% in August 2020 and 14.7% in January 2020. Analytics Jobs By Tools & Technologies The figures researched for the jobs across the categories of Tool and Technologies are a proportion of the Total Analytics Jobs. The demand for Python professionals continues to remain the highest, followed by Java\/JavaScript.Almost 26.3% of advertised jobs require Python as a core skill in June 2021, a negligible decline from 26.6% from August 2020 but an increase from 18.5% in January 2020.The proportion of Java\/Javascript jobs has declined to 19.6% in June 2021, down from 22.5% in August 2020, but an increase from 16.0% in January 2020. R skills come third at 6.3% (down from 9.6% in August 2020), and demand for SAS skills increased to 4.8% in June 2021 (up from 4.2% in August 2020)Among Dashboarding and BI tools, Tableau skills are most in-demand (increasing to 11.9% in June 2021 from 10.8% in August 2020 and from 7.6% in January 2020), followed by Microsoft Power BI (increasing to 8.9% in June 2021 from 6.5% in August 2020, and from 3.3% in January 2020). QlikView also saw a significant increase in its demand (increasing to 6.2% in June 2021 from 1.9% in August 2020 and 1.4% in January 2020) Salaries Advertised For Analytics Jobs Share of Open Jobs by Salary Brackets The median salary advertised for analytics jobs in India as of June 2021 is INR 9.7 Lakhs per annum, a slight increase from August 2020 when it was 9.5 lakhs per annum and from January 2020 when it was 9.4 Lakhs per annum. Advertised salaries tend to be lower than actual salaries.There has been a slight increase in the proportion of advertised jobs with salaries greater than 15 Lakhs per annum from August 2020 to June 2021, but this proportion is lesser than that in January 2020:13.5% of all jobs in June 2021 advertised a salary greater than Rs. 15 Lakhs as opposed to 13.1% in August 2020. In January 2020, this proportion was 13.8%.11.1% of all advertised jobs in June 2021 offer a salary between Rs. 15-25 Lakhs as against 10.3% in August 2020 and 10.5% in January 2020.A negligible number of jobs advertised were for salaries greater than 50 Lakhs per annum:Around 2.1% of all advertised jobs in June 2021 offer a salary between Rs. 25-50 Lakhs. Only 0.2% offer a salary between 50-75 Lakhs per annum.None of the jobs advertised in June 2021 was for salaries greater than 75 Lakhs per annum.The lower proportion of advertised jobs in the higher salary ranges is attributed to employers withholding salary details of the higher paying analytics jobs – this provides employers more negotiating power at the time of salary discussions – this is all the more significant during the recessionary environment as there are proportionately fewer jobs, more candidates, and a greater likelihood of negotiating lower salaries when offers are made.A combined proportion of 86.5% of the advertised jobs is for salaries less than 15 Lakhs.The 6-10 lakh salary range had the highest proportion of analytics jobs advertised in June 2021 at 31.5%, a marginal increase as compared to August 2020, when it was 30.7%The proportion of jobs with entry-level salaries decreased from August 2020 to June 2021. This could be attributed to the increase in the proportion of mid-senior and senior-level open jobs in analytics64.3% of the jobs advertised in June 2021 offer a salary less than 10 Lakhs per annum, as opposed to 68.0% in August 2020 and 69.0% in January 20208.5% of the jobs advertised in June 2021 offer a salary between 0-3 Lakhs per annum, as opposed to 9.6% in August 2020 and 12.4% in January 202024.3% of the jobs advertised in June 2021 offer a salary between 3-6 Lakhs per annum, as opposed to 27.7% in August 2020 and 26.7% in January 2020. Analytics Jobs By Company Type Share of Open Jobs by Company Type More than four in five analytics jobs advertised in June 2021 were posted by MNC IT & KPO Service Providers. These firms have observed a rising proportion of analytics jobs from 2018 (when they advertised only 23.6% of the analytics jobs). Since then, this proportion has increased to 59.4% in 2019 and 74.6% in January 2020, only to see it drop to 67.8% in August 2020. As of June 2021, MNC IT & KPO Service Providers make up 81.5% of analytics jobs advertised.Domestic IT & KPO Service providers contributed to only 9.0% of the total analytics jobs advertised in January 2020. This percentage more than doubled by August 2020 to reach 18.9%, only to drop back to 9.0% in June 2021.Over three years back, Captive Centres in India had one of the highest percentages (40.7% in 2018) of analytics job requirements. However, this percentage fell to 4.1% in June 2021 from 5.8% in August 2020 and 11.0% in January 2020.The Indian consumer firms contributed to only 1.6% of the analytics jobs in June 2021, the lowest among all company types. These firms advertised 9.1% of the total analytics job postings in 2018, after which this percentage has seen a declining trend. This percentage reduced to 4.0% in 2019 and 1.6% in January 2020.Consulting Firms’ proportion of analytics job postings increased from 2.5% in August 2020 to reach 3.7% in June 2021, the same as January 2020. The Consulting companies run their analytics operations in India as cost centres – and have reduced hiring as a direct result of the recessionary environment. Hence, these firms have also seen a declining trend in the proportion of analytics job advertised, reducing from 13.3% in 2019 to 3.7% in June 2021. Conclusion There are approximately 138,000 jobs open in June 2021, a 67.0% increase from June 2020 when the average of advertised analytics jobs reached the lowest due to the pandemic. This increase in job numbers can be mainly attributed to the backfilling activity of companies after they laid off many employees due to the lockdown in March 2020. It can also be attributed to the accelerated adoption of AI\/data science in India, which is a direct result of the pandemic. Read the complete report here: To know more about our recruitment service offerings, please reach out to us at info@analyticsindiamag.com","excerpt":"The analytics job report, presented by Analytics India Magazine (AIM), provides a broad study of the data science job landscape – covering profiles and roles that require analytics competencies and skills. The research provides a comprehensive view of the talent market that was affected by the Covid-19 pandemic and its recovery since the lockdown restrictions […]","categories":["AI Features"],"tags":["AI Jobs","analytics jobs","Data Science Jobs","data scientist salary in india","machine learning jobs"],"author_name":"Kashyap Raibagi","publish_date":"2021-06-11T10:00:00","publication_year":"2021","word_count":3223,"keywords":["machine learning jobs","data science","artificial intelligence","AI","AI Jobs","Data Science Jobs","analytics jobs","NLP","RAG","Aim","data scientist salary in india","Python","analytics","JavaScript","R"],"extracted_tech_keywords":["AI","artificial intelligence","NLP","data science","analytics","Aim","RAG","Python","R","JavaScript"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/study-analytics-and-data-science-jobs-in-india-april-2021\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":28368,"title":"Artificial Intelligence Helps Track Down Mysterious Cosmic Radio Bursts","content":"Taking a new step forward in the search for “little green men” and “extraterrestrials” in the universe, the University of California, Berkley, has reportedly used machine learning to discover 72 new fast radio bursts from a “mysterious” source some three billion light years from Earth. Fast radio bursts, or FRBs, are bright pulses of radio emission, just milliseconds in duration, thought to originate from distant galaxies. Most FRBs have been witnessed during just a single outburst. In contrast, FRB 121102 is the only one to date known to emit repeated bursts, including 21 detected during Breakthrough Listen observations made in 2017 with the Green Bank Telescope (GBT) in West Virginia. According to an official report by the UC Berkley, the behaviour by FRB 121102 drew the attention of many astronomers hoping to pin down the cause and the extreme physics involved. Usually, such activity sparks speculative theories ranging from highly-magnetised neutron stars blasted by gas streams from a supermassive black hole to suggestions about advanced “alien” civilizations. To study these bursts further, the UC Berkley team used artificial intelligence algorithms to dredge up the radio signals from data over a five-hour period. An earlier analysis of the 400 terabytes of data employed standard computer algorithms to identify 21 bursts during that period. All were seen within one hour, suggesting that the source alternates between periods of quiescence and frenzied activity, said Berkeley SETI postdoctoral researcher Vishal Gajjar. Breakthrough #Listen uses machine learning data from @GrnBnkTelescope to discover 72 new pulses from a Fast Radio Burst source – among the most powerful and mysterious emitters in the universe: https:\/\/t.co\/H7zi3i5ljP — Breakthrough (@brkthroughprize) September 10, 2018 “This work is exciting not just because it helps us understand the dynamic behaviour of fast radio bursts in more detail, but also because of the promise it shows for using machine learning to detect signals missed by classical algorithms,” said Andrew Siemion, director of the Berkeley SETI Research Center and principal investigator for Breakthrough Listen, the initiative to find signs of intelligent life in the universe. Whether or not FRBs themselves eventually turn out to be signatures of extraterrestrial technology, Breakthrough Listen, the astronomical program searching for signs of intelligent life in the universe, is helping to push the frontiers of a new and rapidly growing area of our understanding of the universe.","excerpt":"Taking a new step forward in the search for “little green men” and “extraterrestrials” in the universe, the University of California, Berkley, has reportedly used machine learning to discover 72 new fast radio bursts from a “mysterious” source some three billion light years from Earth. Fast radio bursts, or FRBs, are bright pulses of radio […]","categories":["AI News"],"tags":["AI (Artificial Intelligence)"],"author_name":"Prajakta Hebbar","publish_date":"2018-09-17T05:03:38","publication_year":"2018","word_count":388,"keywords":["Go","API","machine learning","artificial intelligence","programming_languages:R","AI","data_tools:Spark","programming_languages:Go","ViT","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","R","Go","API","ViT","programming_languages:R","programming_languages:Go","data_tools:Spark"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/artificial-intelligence-helps-track-down-mysterious-cosmic-radio-bursts\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":32282,"title":"This Amoeba Eats Oats For Breakfast &#038; Solves Networking Problems","content":"The fundamental nature of every living organism is the efficacy of its actions. No creature, with the exception of humans, spends up energy on unwanted tasks. The frugality of their movements can be replicated in everyday networking problems faced by humans. For instance, the travelling salesman problem is one such idea, which deals with finding the shortest distance between two or more distinct points. Choosing a better move correctly and quickly is a fundamental skill of living organisms that corresponds to solving a computationally demanding problem. A unicellular plasmodium of Physarum polycephalum searches for a solution to the travelling salesman problem (TSP) by changing its shape to minimise the risk of being exposed to aversive light stimuli. The TSP is an optimisation problem in which the goal is to find the shortest route between several cities so that each city is visited exactly once, and the start and end points are the same. The problem is NP-hard, meaning that as the number of cities increases, the time needed for a computer to solve it grows exponentially. The complexity is due to a large number of possible solutions. For example, for four cities, there are only three possible routes. But for eight cities, the number of possible routes increases to 2520. Experimental Setup The type of amoeba used by the scientists was a particular plasmodium, which is fed with oat flakes and weighs around 12 mg. This amoeba changes its shape by syphoning the gel in and out at a velocity of 1mm per second to create pseudopod-like appendages. An identical chip with 64 lanes was used to fix the physical size of the chip. The time required for communication between the branches of plasmodium should be equalised independent of n. The plasmodium is placed inside the chip and all the initial states are set to 0. All the lanes were illuminated to block advances of the plasmodium. Camera: CCD Temperature: 27 deg Celsius Humidity: 95% Stimulating Amoeba With Light In their experiments, the researchers placed the amoeba in the centre of a stellate chip, which is a round plate with 64 narrow channels projecting outwards, and then placed the chip on top of an agar plate. The amoeba is confined within the chip, but can still move into the 64 channels. Light stimuli are updated according to recurrent neural network dynamics to solve the Traveling Salesman Problem. In order to maximise its nutrient absorption, the amoeba tries to expand inside the chip to come in contact with as much agar as possible. However, the amoeba has an aversion for light. So, each channel can be selectively illuminated by light, to manipulate the amoeba to encroach into dark channels. For instance, if the amoeba branches out in three routes say, Bangalore-Delhi, Bangalore- Mumbai, Bangalore-Assam; then the branch occupying Bangalore-Mumbai route will be left dark avoiding the amoeba from visiting multiple routes simultaneously. The model is designed such a way that, shorter distance is given preference to be illuminated over longer distances. The amoeba explores the solution space by continuously redistributing the gel in its amorphous body at a constant rate, as well as by processing optical feedback in parallel instead of serially. Amoeba finds TSP solutions for 8 cities via paper by Zhu et al The way these organisms move indicates the presence of a fundamental law, that runs undercurrent as the amoeba retains the rules of the previous activation and replicates it when invoked again. Groups of the branches perform synchronisation and desynchronisation for sharing information even though they are spatially distant. The efficacy of this approximation has befuddled the researchers and the mechanism still remains as a mystery. Key Takeaways Although a conventional computer can still solve the TSP much faster than an amoeba, especially for small problem sizes, the new results are intriguing and may lead to the development of novel analogue computers that derive approximate solutions of computationally complex problems of much larger sizes in linear time. “We will investigate further how these complex spatiotemporal oscillatory dynamics enhance the computational performance in finding higher-quality solutions in shorter time,” said coauthor Song-Ju Kim at Keio University. Researchers are now exploring the use of the amoeba-inspired electronic circuits for tackling the constraint satisfaction problem, satisfiability problem, analogue-to-digital conversion and finding walking manoeuvre of a multi-legged robot. Such an approach to exploit physical parallelism may develop a novel analogue computing paradigm, providing powerful approximation methods for solving complex optimisation problems appearing in a wide spectrum of real-world applications. Also See The Amoeba In Action:","excerpt":"The fundamental nature of every living organism is the efficacy of its actions. No creature, with the exception of humans, spends up energy on unwanted tasks. The frugality of their movements can be replicated in everyday networking problems faced by humans. For instance, the travelling salesman problem is one such idea, which deals with finding […]","categories":["Deep Tech"],"tags":["optimisation algorithms","RNN"],"author_name":"Ram Sagar","publish_date":"2018-12-27T05:35:23","publication_year":"2018","word_count":754,"keywords":["Replicate","Go","programming_languages:R","AI","neural network","Git","CuPy","optimisation algorithms","RNN","GAN","R","Redis"],"extracted_tech_keywords":["AI","neural network","CuPy","Redis","R","Go","Git","GAN","Replicate","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/amoeba-computing-solves-networking-problems\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10005302,"title":"Complete Guide to Voila to Turn a Jupyter Notebook Into a Standalone Web Application","content":"Voila, an open-source python library that is used to turn the jupyter notebook into a standalone web application It supports widgets to create interactive dashboards, reports, etc. Voila launches a kernel when it is connected to a notebook and executes all the cells but it does not stop the kernel there so that the user can interact with the output. Voila converts the jupyter notebook into HTML and returns it to the user as a dashboard or report with all the inputs excluded and the outputs included. Voila supports all the python libraries for widgets such as bqplot, plotly, ipywidgets, etc. Voila is framework and language agnostic which means that voila can work with any jupyter kernel whether it is C++ or Python, etc. because it is built on jupyter standard protocols. Voila is useful in many ways because of its extensibility and usability, it can serve as a solution for Data Science or Business Analyst professionals to share their work which is relevant for the end-user or client. In this article, we will explore how easily and effortlessly we can use Voila to create reports and dashboards. Implementation of Voila in Python: Like any other library, we will install Voila using pip install voila. After installing voila we need to open the jupyter notebook to check that a new tab named Voila is added in the toolbar. Importing required libraries As Voila supports Plotly so we will be using plotly in this article, other than that we will be using pandas for loading the data. import plotly.express as px import plotly.graph_objects as go import pandas as pd Loading the dataset We will be using stock data of a firm which can be downloaded from any financial website like Yahoo Financial, I have downloaded data for ‘Biocon Pharmaceuticals’ and stored it in a CSV file. The data contains Date, Opening Price, Closing price, etc. df = pd.read_csv(‘biocon.csv’) df.head() c. Adding Some Information regarding Stock We will add some information regarding the company whose dataset we are working on i.e. Biocon Pharmaceuticals so that our dashboard\/application will be informative. info = ”’Biocon Limited is a globally recognized, innovation-led organization that is enabling access to high quality, advanced therapies for diseases that are chronic, where medical needs are largely unmet and treatment costs are high. We are driven by the belief that the pharmaceutical industry has a humanitarian responsibility to provide essential drugs to patients who are in need and to do so with the power of innovation. In line with this belief, Biocon has developed and commercialized a differentiated portfolio of novel biologics, biosimilars, and complex small molecule APIs in India and several key global markets, as well as, generic formulations in the U.S. and Europe. We are a leading global player for biosimilars and APIs for statins, immunosuppressants and other speciality molecules, with customers in over 120 countries.”’ print(info) d. Using Plotly to create Charts We will create different plots that are used to analyze the stock market data. We will use the markdowns for giving the name of the charts so that it reflex in the dashboard\/application. Candlestick Chart fig1 = go.Figure(data=[go.Candlestick(x=df['Date'], open=df['Open'], high=df['High'], low=df['Low'], close=df['Close'])]) fig1.show() Line Chart fig2 = px.line(df, x='Date', y='High') fig2.show() OHLC Chart fig3 = go.Figure(data=go.Ohlc(x=df['Date'], open=df['Open'], high=df['High'], low=df['Low'], close=df['Close'])) fig3.show() Area Chart fig4 = px.area(df, x=df['Date'], y=df['High']) fig4.show() Converting Notebook to Web Application Now we will use Voila to create the dashboard\/application in just a single click. As seen above now we will click the Voila tab in the toolbar section to launch voila. As soon as we click Voila it will launch voila in a new window and start executing the commands one by one. After Execution, the Dashboard\/Application is created which will display all the outputs without the code. I have added markdowns for different sections as you will see below in the images of the Voila Dashboard\/Application. Info Section Analysis using Charts In the above images, we saw how Voila rendered our jupyter notebook and converted it into a dashboard\/application. All the graphs are created using plotly so these graphs are highly interactive and downloadable. We can also launch voila using the command prompt by using Voila <filename.ipynb> Similarly, we can create a different application using different Python libraries which are used to control data using widgets. Conclusion: In this article, we saw how we can use Voila to create a dashboard\/application. We created a stock analysis dashboard\/application using plotly and Voila. Voila is blazingly fast and we can share the results created using voila to others also. It is highly extensible, flexible, and has high usability.","excerpt":"In this article, we will explore how easily and effortlessly we can use Voila to create reports and dashboards.","categories":["Deep Tech"],"tags":["Jupyter","Jupyter Notebook","Machine Learning","Python","Web Applications"],"author_name":"Himanshu Sharma","publish_date":"2020-08-23T11:00:00","publication_year":"2020","word_count":766,"keywords":["data science","Go","Jupyter Notebook","Plotly","TPU","AI","ML","Machine Learning","Python","Web Applications","Jupyter","R","Pandas"],"extracted_tech_keywords":["AI","ML","data science","Jupyter","Pandas","Plotly","TPU","Python","R","Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/complete-guide-to-voila-to-turn-a-jupyter-notebook-into-a-standalone-web-application\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":50084,"title":"How Google Flutter Became The Frontrunner In Mobile App Development Race","content":"What do modern-day mobile phone users require in an app? Is it a beautiful design, smoother animation or better performance? The answer is all three. App developers are expected to create new features for apps which don’t compromise on speed and performance. This is the reason why Google created Flutter — to make it easier for developers to create the best applications possible. Google’s UI toolkit Flutter was released in 2017. Flutter uses Dart object-oriented programming language which is the fastest growing language, up by 532% between 2018 and 2019, probably thanks to Flutter itself. Flutter is an open-source and free SDK with widgets and tools. It is used to develop apps for both Android and iOS. Now, if you ask an app developer about what he\/she likes about it, the first thing they might talk about how easy it is to grasp and how one can use a single type of code for two platforms. Let’s talk about somethings that you learn while using Flutter: Dart, the programming language: Dart is easy to learn and a joyous experience while using it. Compared to TypeScript and Flow, which you use in React Native, it is far easier and has a simple syntax. With a good compiler, there are fewer hidden runtime error messages. A developer should appreciate a strongly typed language even for a medium-sized application. Architecture and State Management: When Flutter was a new technology, it was hard to trust the architecture and the state management tools, but in 2019 more and more complex Flutter apps have been written. Some people follow BLoC, which is complex, then there are RxDart and Redux, which aren’t simple as well. People coming from Android or React might find it more comfortable since they may be used to this. Hot Reload: Flutter’s Hot Reload helps you to experiment, build UI, add features and fix bugs quickly and makes it as easy as possible. With the addition to its speed, it is reliable. This Hot Reload feature is the best in the class. Complex Layout is easy: People already using layout libraries like React, CSS Grid, Flexbox, etc., will have no problem learning Flutter’s layout. Flutter is a lot similar to these web-based layout concepts. Also, the UI logic in Dart and Flutter is excellent for reliable code. Less focus on widgets is needed: A lot of attention has been given to Flutter’s widgets tutorial on their website. Eventually, developers have started to write more full-fledged apps and have gone beyond implementing pure UI and animation. This is why more end to end tutorials should be mentioned on Flutter’s website. There are more advanced Dart features and architecture patterns that could be learned. Flutter goes beyond just widgets. Apps for both platforms: It takes times to submit apps for each platform, especially Google Play and iTunes Connect. Too many widgets: There a lot of widgets on Flutter. There are widgets for simple positioning; this makes the Dart code nested when it is time to implement more complex layouts. That is why it is recommended to learn the simple widgets first and then do for the others as and when needed. Firebase: For simple projects or the kind of projects you want to hand over later where they don’t have enough backend developers, firebase can be a good choice. But if you are a small team, it would be a problem if there is a surge in the traffic and the Firebase bill shows up which is usage-based. How Flutter Measures Up To Its Competitors Flutter and React Native are two of the hottest cross-platform app development technologies. So, it is imperative that we compare these two: Programming Language: We know that Dart is the fastest growing language, but Javascript is the number one language preferred by the developers. Although Dart is simple to understand, because it is a lesser-known language, it isn’t widely used. Javascript users can easily understand Dart since it supports most of the object-oriented concepts. Technical Architecture: Communicating with native modules using a bridge may result in poor performance. Flutter has most of the native models in the framework itself, so it doesn’t need a bridge to communicate. Whereas, React Native uses Javascript bridge to communicate with native modules. UI component and API development: React Native relies on third-party libraries to access native modules because it provides only UI rendering and device access APIs. Flutter is loaded with UI rendering components, device API access, navigation, testing, stateful management and loads of libraries. This abundance of components does not require the use of third-party apps. Developer Support: Being the more mature framework, React Native has better developer support in terms of IDE’s and language feature. Flutter is relatively new, but will definitely catch up. Cross-platform app development: Flutter has a good automation tooling and can be used to deploy apps from command lines whereas React Native lacks Command Line Interface (CLI) support for build automation. Outlook When it boils down to it, Flutter might have the edge when we look at the overall aspects including installation, setup and project configuration, community support, testing support, DevOps, CI\/CD support, etc. Considering everything Flutter has to offer, and with no signs of slowing down, it has shown us that it is here to stay.","excerpt":"What do modern-day mobile phone users require in an app? Is it a beautiful design, smoother animation or better performance? The answer is all three. App developers are expected to create new features for apps which don’t compromise on speed and performance.  This is the reason why Google created Flutter — to make it easier […]","categories":["Global Tech"],"tags":["App Development","flutter","Google","mobile application"],"author_name":"Sameer Balaganur","publish_date":"2019-11-18T12:47:19","publication_year":"2019","word_count":879,"keywords":["Go","API","Rust","mobile application","CI\/CD","TypeScript","Java","App Development","automation","Google","JavaScript","DevOps","R","flutter"],"extracted_tech_keywords":["R","JavaScript","TypeScript","Go","Rust","Java","CI\/CD","DevOps","API","automation"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/google-flutter-mobile-app-development\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":29427,"title":"Battle Of BI Tools: AWS QuickSight Vs Power BI","content":"BI is a highly contested market with plenty of choices available to go with. Be it Tableau, QlikView, SAS, IBM Cognos or Google’s Data Studio, among others, every tool offers improved functionality and feature. In this article, we compare two popular tools that have captured BI market, Amazon’s AWS QuickSight and Microsoft’s Power BI, and see where it feels unique with a particular aspect. Amazon’s AWS QuickSight A utility tool in Amazon Web Services (AWS), QuickSight is a BI service that helps create visualisations of data and design interactive dashboards for an insightful analysis. Powered by super-fast, parallel, in-memory calculation engine (SPICE), this tool can interact with various AWS sources, can connect to SQL databases and even SaaS applications like Salesforce and so on. With support for upto 10,000 users, QuickSight hinges on a ‘pay-per-session’ pricing for cost-effective user access. Pros The SPICE Engine requires no prior setups which means it can be activated very easily. This also implies that every user can work simultaneously on the interactive analysis with different data sources in AWS. Altogether, QuickSight offers a faster environment due to this feature. Different data sources such as Apache Spark, Amazon Redshift, Amazon Aurora and so on, can be integrated into AWS with QuickSight. In addition to this, SQL servers are also compatible. Compared to other BI tools, QuickSight is cost-friendly since users are charged only for the basis of usage. QuickSight even has visualisation recommendations for the data presented. It’s feature ‘AutoGraph’, which is built using a certain set of algorithms eventually learns data and recommend the appropriate graphs. Overall, setting up QuickSight installation is less complicated and can be done in less than an hour. Cons Currently, QuickSight is available on web browsers and as an iOS app. So, Android mobile users might face a setback if they need to work with QuickSight since its unavailable as of now. Support for Android is expected soon. Despite having better visualisations, this BI tool has limited options in choosing graphs\/charts. Users have also reported functionality and security issues mainly due to limited support\/ updates in QuickSight. QuickSight is relatively new. Hence it has to improve features to make an impact in the BI environment already dominated by Tableau & QlikView. QuickSight had also faced database reporting issues by many users. Microsoft’s Power BI Released in 2014, Power BI is a collection of analytics tools developed by Microsoft. Over the years, it is one of the top BI tools used by analysts. When it comes to visualisation, Power BI is resolute and has a wide plethora of options and customisation. What’s more, it can even load various customisations from other BI platforms. Also, integration with other Microsoft products is what makes it a favorite choice among users. Power BI is available as downloadable software as well as in the form of SaaS applications and mobile apps. Pros Power BI’s integration and compatibility with data sources are excellent. The interface presents databases in the form of a menu view where users can simply select their data source, connect to data and begin with analysis and reporting. Very good tech support in Power BI. Issues are resolved faster. A highlighting feature is the provision of APIs where developers can club Power BI dashboards to other products. In areas like finance, Power BI stands apart since it offers financial analytics tools, fraud protection and compliance monitoring tools. Therefore, Power BI is most secure. Power BI is inexpensive and faster compared to Tableau or QlikView. Cons Setup\/Installation of Power BI is quite a hassle. Except under Windows 10 environment, it requires a host of prerequisites (.NET Framework etc.). MySQL systems suffer setbacks and show inconsistencies. Power BI leans heavily on the Windows environment. So, other OS face issues when it comes to BI activities through this tool Difficult to deploy dashboards into production. Report sharing in Power BI is now paid and has a license. Prior to 2017, it was free. AIM View: Both AWS QuickSight and Power BI have become industry standard BI tools and have cornered a substantial market share. While Power BI is richer in features, AWS QuickSight is simple to use. As you can see, both have ups and downs. It is left to the user to make the best use of these tools and see which suits their requirement.","excerpt":"BI is a highly contested market with plenty of choices available to go with. Be it Tableau, QlikView, SAS, IBM Cognos or Google’s Data Studio, among others, every tool offers improved functionality and feature. In this article, we compare two popular tools that have captured BI market, Amazon’s AWS QuickSight and Microsoft’s Power BI, and […]","categories":["Deep Tech"],"tags":["Amazon","business intelligence tools","dashboards","data analytics tools","Data Visualisation","Microsoft"],"author_name":"Abhishek Sharma","publish_date":"2018-10-22T04:17:21","publication_year":"2018","word_count":721,"keywords":["Go","API","dashboards","AWS","AI","data analytics tools","Data Visualisation","Amazon","business intelligence tools","Apache Spark","Aim","ViT","analytics","SQL","R","Microsoft"],"extracted_tech_keywords":["AI","analytics","Aim","AWS","Apache Spark","R","SQL","Go","API","ViT"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/battle-of-bi-tools-aws-quicksight-vs-power-bi\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10061361,"title":"Talking Ethical AI with Ram Narasimhan, Global Head AI and Cognitive services, Xebia","content":"Xebia is a full stack software engineering, IT strategy and digital consulting company. They leverage big data, AI and cloud to enable digital transformation of enterprises. “We enable organisations to accelerate their data modernisation journey while reducing costs. We specialise in building AI applications, ML solutions, data hub, business analytics and delivering AI governance frameworks for our customers,” said Ram Narasimhan, Global Head AI and Cognitive services, Xebia. In an exclusive interview with Analytics India Magazine, Ram Narasimhan spoke about how Xebia embeds ethics into their business processes. AIM: What are the AI governance methods, techniques, and frameworks used in Xebia? Ram Narasimhan: AI projects are complicated. They bring together a diverse set of systems, processes, and business unit responsibilities into a platform-as-a-service kind of approach to serve internal and external stakeholders. Any organisation that adopts AI solutions must put together an AI governance plan first. Microsoft AI fairness checklist, San Francisco’s Ethics and Algorithms toolkit, Google AI fairness, UK Government’s data ethics workbook, NITI Aayog’s Responsible AI, Alan Turing’s Understanding AI Ethics and Safety, IEEE Ethics in Action etc are the AI governance frameworks in existence today. While the exact contours of a future AI governance model is still undecided, advanced governance models such as polycentric governance, hybrid regulation, mesh regulation and active-matrix theory can provide some framework and inspiration. In my opinion, breaking down the problem into simple chunks, in this case, the “Modular–Layered Model”, would be the most effective and clear way to define a set of standards. Consider layered approaches such as Open System Interconnection (OSI) reference model during the late 1970s or even layered model proposed by David Clark to represent cyber-space; it was designed around modularity to reduce interdependencies and establish clear hierarchies and responsibilities. You can also refer to Harvard’s “A Layered Model for AI Governance.” The nature of the AI governance model could be broken down into social and legal, ethical, and technical foundations. Building a recommender engine for real-time interactions with customers may lead to a full-scale implementation of AI solutions across the enterprise. AI frameworks governing such solutions must include a complete view of audit data and KPIs for exceptions and scoring. A complete roadmap around the AI framework can be built for such scenarios based on a layered modular approach. AIM: What explains the growing conversations around AI ethics, responsibility, and fairness of late? Why is it important? Ram Narasimhan: AI requires significant R&D and investments to make a successful outcome for all stakeholders, including abiding by laws, regulations, and compliance to certain jurisdictions. Developing an AI solution also brings to light an important factor which is fairness, while consuming the output of AI. Discrimination towards a sub-population can be created unknowingly, unintentionally, with incorrect labelling and\/or incorrect data fed to AI. For example, biased AI outcomes on genders and race could be potentially disastrous for everyone involved. Bias in machine learning models can be due to algorithm, user interaction, representation bias in data and bias in sample data. Detecting bias in a single feature is easy; however, it could be difficult to detect it for a combination, e.g. race, income, gender, etc. Correlation analysis can be performed to check model fairness, groups\/band scores can be analysed, and probability scores can be verified. Apart from this, IT leaders will also have to infuse data quality and ethics rules per system\/application feeding data to AI, have a business glossary to describe data and a technology stack to control AI outcome in every layer such as databases, platforms, applications, environments, presentation layers and exchange layers to scale. AIM: How does Xebia ensure adherence to its AI governance policies? Ram Narasimhan: Building responsible AI requires a humble appreciation of human interactions, background, circumstances, and diverse expertise. Every developer involved brings different knowledge, level of maturity, awareness and empathy. We need to understand the trade-offs in achieving the goal set by organisation and in managing the pressure for AI outcomes on time. Here’s how we adhere to organisation policies and best practices: · Raising awareness, building AI principles, developing scoring matrix for assessing effectiveness to principles and a team spirit towards delivering world class AI solutions. · Case studies, references, tools and resources being made available for monitoring outcomes and for seeking guidance on certain KPIs. · Hands-on support and peer review sessions to provide real time insights on do and don’ts with an AI project. · Design and review forums to go through checklists and bring a balanced view. · Bringing external perspectives to validate feedback during the design and development phase. AIM: How do you mitigate biases in your AI algorithms? Ram Narasimhan: Technical solutions alone do not resolve the potential harmful effects of algorithmic decisions. However, there are certain things you can keep in mind while developing AI algorithms. Given an existing AI solution, you can follow these steps to fix it: · Examine training dataset, conduct subpopulation analysis and monitor models over time. · Establish debiasing strategy that involves controls over technical platform, operational systems and workplace KPIs defining AI policies. · Improve human-driven processes by building pipeline insights and showcasing a dashboard of potential bias as exceptions. · Decide on automation of use cases and human touch that might be needed. · Follow a multidisciplinary process, involving various subject matter experts from diverse socio\/economic backgrounds to validate AI outcomes. · Engage and build a diversified AI team to mitigate unwanted bias. AIM: Do you have a due diligence process to make sure the data is collected ethically? Ram Narasimhan: Data collection is a huge process that involves internal data and external datasets that are publicly available and from agreed reliable intelligence sources on a contract basis. Usually, companies tend to have agreements for data collection and availability. However, many companies also practice web scraping, which is a process of extracting data from web-based resources. Based on business needs, a company can decide to buy or sell data; however, there must be a definite set of standard processes to map, validate and transform data to achieve the desired outcome. · Ownership: Individuals own their data. Digital privacy is a concern for any customer; there must be a written agreement to seek permission to avoid ethical and legal dilemmas. · Transparency: A company is legally obligated to reveal how they intend to collect and process individuals’ data. Clear policies and security outlines must be defined and made available to relevant entities\/individuals. · Privacy: Personally identifiable information (PII) should be protected and must not be made public. · Intention: If an insurance company decides to collect information and decide on mental health or financial health, the intention and way to use the data should be ethical and not meant to exploit the individual\/entity. · Outcomes: Outcome of data analysis can lead to inadvertent harm to individuals, communities, or entities. This should be addressed. AIM: How do you embed ethical principles in your platform? Ram Narasimhan: If you consider a real-world example of a recommender system for an e-commerce sales purpose, ethical principles can be applied to an AI platforms in following ways: AIM: How does your company ensure user data privacy? Ram Narasimhan: As AI evolves, it magnifies the opportunity and use of personal data in ways that can intrude on someone’s privacy by doing real-time analysis of personal information across various platforms and lifespan of events. Facial recognition and fingerprint scanning are used to identify a person and facilitate financial or trade transactions, potentially benefiting any specific company. It is important to seek permission from the user to utilise and analyse data. Secondly, companies analysing user data and utilising them in algorithms must take certain care in processing private data: • Provide a self-service portal to enable customers to manage digital footprint, permission, visibility and potential bias based on interactions. For example, complying with GDPR, CCPA or similar data management standards. • In data collection, introduce a data stewardship process to manage personal data in systems. • Introduce data transparency or disclosure rules for algorithms using personal data. For example, rules to manage email data, downloaded files, location data, chatrooms data, website data, searches, apps data, time-related data, etc. • Data governance rule for enterprise, which is applicable to all systems for privacy by design. • Rules for aggregating data • Rules for a data marketplace AIM: Did you come across any biases, or ethical concerns\/issues lately within your organisation\/industry\/product? If yes, how did you address them? Ram Narasimhan: Most of the companies are not only relying on OEM providers to facilitate audit, security, metadata view and monitoring mechanisms, they are also building data completeness checks, data archiving checks, data privacy checks, data leakage checks, and data security checks. And from a compliance\/regulatory perspective, they are building encryption and tokenization into platform design. More recently, companies are adopting MLOps, AutoML and AIOps types of frameworks to manage model effectiveness, fairness, ethical validation, and biased outcomes to control unintentional AI damage. More robust and careful AI governance committees are being formed, more awareness is being created, and diversified AI teams are being put in place for mitigating ethical conflicts.","excerpt":"Any organisation that adopts AI solutions must put together an AI governance plan first.","categories":["AI Features"],"tags":["AI fairness","ai governance","consumer protection","Data Governance","Data Privacy","data rights","digital transformation","Ethical AI","Interviews and Discussions","modernization"],"author_name":"Sri Krishna","publish_date":"2022-02-23T13:00:00","publication_year":"2022","word_count":1518,"keywords":["consumer protection","data rights","TPU","ai governance","R","digital transformation","RAG","Data Governance","analytics","machine learning","AWS","AI","Ethical AI","ML","Data Privacy","AI fairness","modernization","MLOps","Aim","Interviews and Discussions"],"extracted_tech_keywords":["AI","machine learning","ML","analytics","MLOps","Aim","RAG","AWS","TPU","R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/talking-ethical-ai-with-ram-narasimhan-global-head-ai-and-cognitive-services-xebia\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10130176,"title":"India Needs to Boost its Quantum Mission","content":"“India must urgently prepare itself for the quantum revolution to safeguard its security and national interests,” said Ajai Chowdhry, the co-founder of HCL and chairman of the mission governing board for India’s National Quantum Mission (NQM). Announced in 2023, NQM aims to invest INR 6,000 crore to accelerate its quantum technology capabilities. Quantum computing is often seen as the Holy Grail of modern computing due to its potential to revolutionise data encryption and processing speeds, making traditional encryption methods obsolete. For this, the government recently announced a collaboration with major IT services firms like TCS, HCL, and Tech Mahindra to develop quantum technologies under the INR 6,000 crore scheme. This initiative will support Indian startups and scientists in the emerging quantum field. “We want them to do research on quantum technologies because there’s a huge amount of algorithms required for quantum,” said Chowdhry. A report by Itihaasa Research and Digital, co-founded by former Infosys CEO Kris Gopalakrishnan, highlights the global landscape of quantum investments. The top 12 countries have collectively invested around $38.6 billion, with China leading the way at $15 billion. In contrast, India’s investment stands at a mere $0.74 billion. “China has an enormous number of publications in the area of quantum technology,” pointed out Ajay K Sood, principal scientific advisor to the government of India. Tech giants like Google, IBM, and Intel have collectively invested billions in quantum computing. While India has published only 1,711 research papers on quantum, China published 12,110, which is seven times more than India. Moreover, India has 82,110 graduates in quantum technology while China produced 57,693. “The elephant in the room is that … [other than] a few top tier institutions there is a scarcity of faculty to train students in quantum technologies,” read the recent report. Private Sector Needs to Step Up “Can China break the cryptography that India employs? That is a point of worry,” noted Chowdhry. The gap between the two nations is evident, with China being at least five years ahead of India in quantum. India’s largest quantum computer, with 6 to 7 superconducting qubits, is being developed at Mumbai’s Tata Institute of Fundamental Research (TIFR), in collaboration with DRDO and TCS. “With the launch of the National Quantum Mission, India is gearing up to develop not only quantum software but also state-of-the-art quantum computing hardware,” said Rajamani Vijayaraghavan, associate professor at TIFR. India aims to develop 100-qubit computers within the next five years. IIT Bombay’s partnership with TCS to develop the country’s first Quantum Diamond Microchip Imager is a significant step forward. This tool will enhance the precision in examining semiconductor chips, reduce chip failures, and improve the energy efficiency of electronic devices. The plan is to mentor and seed-fund 50 startups already working in this field. “The next thing that we are working on is to get startups involved. There are close to 50 startups already working in this area. We will provide them mentorship through these thematic hubs. We will also look at providing them some initial seed funding,” said Chowdhry. India is Not Starting From Scratch The VP of IBM Quantum, Jay Gambetta, recently said that India has the second-highest open access to quantum computing, which is about 77,000. He said that India has the opportunity to become the leader in quantum computing in areas such as energy, sustainability, and agriculture, among others. To bolster this, ISRO also has a partnership with Raman Research Institute for building quantum communication technology, which is still under process. “A bank or an electrical grid in India can be attacked by an adequate quantum computer sitting in China. We must start working on making our country quantum secure. “We will work with different agencies in the government to make them aware that something like this has to be done. Alternatively, the RBI should start working on creating a policy to make all banks secure on this front,” said Chowdhry. India’s quantum mission includes plans to develop its own quantum computers. “In the period that we don’t have a quantum computer, we’ll buy a few for research work. But we are not going to use quantum computers only on the cloud because they are very expensive,” Chowdhry added.","excerpt":"India aims to develop 100-qubit computers within the next five years.","categories":["AI Trends"],"tags":["AI Impacts"],"author_name":"Mohit Pandey","publish_date":"2024-07-25T16:11:54","publication_year":"2024","word_count":699,"keywords":["Go","funding","programming_languages:R","AI","programming_languages:Go","Git","RAG","Aim","AI Impacts","R","startup"],"extracted_tech_keywords":["AI","Aim","RAG","R","Go","Git","startup","funding","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/india-needs-to-boost-its-quantum-mission\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":60493,"title":"Microsoft Sees Huge Spike In Services Traffic: What Does It Mean For Cloud Industry?","content":"A big challenge for all enterprises will be to ensure IT frameworks successfully adapt to changes taking place due to remote working, and the new kind of data flows taking place. Microsoft said its cloud services had seen a massive spike in usage since social distancing and COVID-19 guidelines were put in place by various countries. The company reported a spike of 775% increase of its cloud services in geographies that have implemented social distancing mandates. These include Europe North, Europe West, UK South, France Central, Asia East, India South, Brazil South, etc. Microsoft To Add More Capacity To Handle The Pressure There has been a meteoric rise in the use of Microsoft Teams, Power BI and Windows Virtual Desktop as people shift to working from homes or remote areas. If you look at the usage of Teams alone, Microsoft reported more than 44 million daily users who generated over 900 million meetings and calling minutes on Teams in just one week. With such a high number of users, Microsoft has made sure it keeps the services running and that there is no outage, especially when it comes to certain preferential services, including in healthcare. At the same time, the company is planning to add more capacity in preparation to keep everything running smoothly for its cloud services. “To handle surges in demand, we will facilitate the creation of new capacity in the relevant regions,” said Microsoft. Impact Of Work From Home On Cloud Vendors The reliance on essential web services is so widespread that if there was an outage on any of the cloud platforms, it could make the COVID-19 situation much worse. Given that remote work may be the norm in the coming months and weeks, getting the virtual cloud services running smoothly is of utmost importance for the entire global economy. And large cloud vendors like Microsoft, Amazon and Google play a critical role in these crucial times. The focus will be mainly on enterprise continuity to enable remote work from anywhere. But that would put pressure on traditional data centers and private clouds. This can enhance cloud-native apps and the SaaS market further. The Bottlenecks Many companies are under pressure and businesses in several verticals may experience several outage events. This is especially true for those handling a significant jump in workload. Some of the business segments undergoing some level of stress would involve web learning platforms, collaboration applications, and leisure\/entertainment services. What Cloud Vendors Need To Do Given such a spike in usage of cloud-based services, experts say there need to be some temporary restrictions from cloud vendors designed to balance the best possible experience for all users. This demand has been put forth by politicians as well. Perhaps cloud vendors will put limits on free offers to prioritize capacity for existing paid customers. Microsoft has set ‘soft’ quota limits, and customers can raise support requests to increase these limits. Why 100% Work From Home (WFH) Is Not Possible For Tech Sector In India Currently","excerpt":"A big challenge for all enterprises will be to ensure IT frameworks successfully adapt to changes taking place due to remote working, and the new kind of data flows taking place. Microsoft said its cloud services had seen a massive spike in usage since social distancing and COVID-19 guidelines were put in place by various […]","categories":["Global Tech"],"tags":["Cloud Computing","Cloud Platform","Microsoft","Microsoft Azure","private cloud"],"author_name":"Vishal Chawla","publish_date":"2020-04-01T10:00:00","publication_year":"2020","word_count":500,"keywords":["Go","programming_languages:R","AI","R","programming_languages:Go","Cloud Computing","Microsoft Azure","private cloud","Cloud Platform","Microsoft"],"extracted_tech_keywords":["AI","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/microsoft-sees-huge-spike-in-services-traffic-what-does-it-mean-for-cloud-industry\/","complexity_score":2,"technical_depth":5,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10143159,"title":"Indian IT Salaries Might Finally Increase Next Year","content":"All eyes are on Indian IT, with sky-high expectations for the years ahead. While billions of dollars of investments and bookings paint just one part of the picture, the other side hinges on generative AI, which is expected to change the hiring strategy and influence employee salaries. Based on an analysis by Pareekh Jain, CEO of Pareekh Consulting, IT companies are significantly increasing their hiring of candidates with AI-related skills. Candidates with these skills made up about 10% of total hiring last year, but their share is projected to rise to 20-25% this year. If companies expand their entry-level workforce by 10% in the future, nearly half of the new hires are expected to be in AI-focused roles. This indirectly translates to the fact that the new hires with the latest AI skills would demand higher salaries as well. Krishna Vij, VP at TeamLease Digital, told AIM that AI tools will enhance productivity by automating repetitive tasks. They will also create demand for specialised roles to manage and optimise AI systems effectively. “While team sizes may not drastically reduce, organisations are likely to focus on upskilling talent to balance costs,” said Vij. About 20-25% of fresher hiring in the IT sector is now targeted towards AI skills. This is an uptick from a dismal 5-10% over the last three years. This might result in an upswing of hiring for talented freshers but a downswing for others without the skills. Therefore, though the salaries for the ones with AI skills might be higher, the ones without them might witness a significant drop as well. GCCs Driving the Salary Hike As GCCs establish their presence in India, there’s a growing opportunity for the young talent in the country to leverage their skills in generative AI for a lucrative career with them. According to consulting firm ANSR, about 90% of GCCs in India plan to harness the potential of AI, ML, and cognitive computing in the next 2-3 years. A recent report reveals that GCCs offer salaries 12-20% higher than those in IT services and other non-tech industries for comparable tech roles. Vij agreed that the expansion of GCCs in India is driving up salaries, especially for niche roles in AI, data science, and cybersecurity. This also creates competition for skilled talent and sets new benchmarks. “This has prompted IT companies to increase salaries for specialised roles to retain and attract talent. However, to manage costs, many IT firms focus on hiring freshers and upskilling them, building a strong talent pipeline. These companies may need to refine their strategies to remain competitive in attracting and retaining skilled professionals,” Vij explained. For instance, GCCs provide salaries ranging from INR 9.7 lakh to INR 43 lakh per annum for software developers, depending on experience. In contrast, the IT products and services sector offers around INR 5.7 lakh per annum for entry-level roles, with salaries going up to INR 17.9 lakh per annum for professionals with over eight years of experience. Moreover, entry-level talent at GCCs in India is attracting pay packages that are, on average, up to 30% higher than entry-level salaries across all sectors. The same should get replicated for the Indian IT workforce that has been collaborating with GCCs over the past few years. For companies like TCS, Infosys, and others, the hike largely depends on their performance. Still, freshers might witness a small bump in salaries, given the larger ecosystem in India. Mrinal Rai, assistant director and principal analyst at ISG, told AIM that since AI skills are in demand, people with these skills would come at a higher cost. So far, there hasn’t been any substantial change in the salary packages of service providers for freshers. “However, since both service providers and GCCs are competing for the same talent pool, with the GCCs offering significantly higher packages and better designations to freshers, it gets challenging for service providers,” Rai explained. The fact that Indian IT is reluctant to hire freshers is not new; it continued throughout 2024. However, things might change next year as it remains one of the biggest concerns for many. According to experts in the HR industry, hiring has become a focal point for several IT services companies. Finally Better Salaries Pinkesh Kotecha, chairman and managing director of Ishan Technologies, told AIM that he predicts a 10-15% higher salary for freshers while a salary hike of up to 15% for experienced professionals in the coming year. “The focus while hiring will be on how techies have upskilled and prepared themselves to be valuable to the company. There is a need for collaboration between academia, technology companies, industry bodies, and professionals to upskill regularly to meet the evolving demands of the modern tech world,” Kotecha said. In their latest earnings calls, most Indian IT giants have promised to continue reducing their bench size and hiring freshers to increase the headcount. TCS aims to onboard 40,000 freshers this fiscal year, while Infosys plans to hire 15,000-20,000. Wipro and HCLTech are targeting 10,000-12,000 and approximately 10,000 fresh graduates, respectively. Including mid-tier IT and engineering services firms, around 1 lakh freshers have received offer letters in the ongoing placement season. This was also resonated by Mohandas Pai, founder of Aarin Capital and former CFO of Infosys, when he told AIM that the hiring will also be affected because around 200,000 people will start retiring early next year in India and the companies will have to fill the positions quickly. According to a report by ServiceNow, AI is going to create around 2.73 million tech roles by 2028, a lot of which would be in the fields of software development, web development, software testers, and data analysis. With AI assisting them, the report says that the productivity gain would be able to help them transition into advanced roles as well. With the average package of INR 3.5 LPA in Indian IT, when the demand for AI skills increases, so will the demand from the engineers possessing these skills.","excerpt":"Since both ITs and GCCs are competing for the same talent pool, with the GCCs offering significantly higher packages and better designations to freshers, it gets challenging for IT.","categories":["IT Services"],"tags":["Indian IT industry"],"author_name":"Mohit Pandey","publish_date":"2024-12-10T18:30:00","publication_year":"2024","word_count":995,"keywords":["data science","Go","API","AI","ML","Git","RAG","Aim","generative AI","Indian IT industry","R"],"extracted_tech_keywords":["AI","ML","data science","generative AI","Aim","RAG","R","Go","Git","API"],"url":"https:\/\/analyticsindiamag.com\/it-services\/indian-it-salaries-might-finally-increase-next-year\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10096623,"title":"Big Techs Don’t Care About Lawsuits","content":"Works of artists used to create generative AI models have been heavily criticized for being ‘encyclopedic thieves’. Since text-to-anything models have spread like wildfire, several artists have sued the founding companies for theft and using their creative property without consent. The latest addition to the list is American author and comedian Sarah Silverman. The Seinfeld star has filed copyright infringement lawsuits against Meta Platforms and OpenAI for allegedly using their content without permission to train artificial intelligence language models along with two other authors. Not just artists but also a prominent law firm based in California filed a 157-page lawsuit against OpenAI for violating privacy laws by secretly scraping 300 billion words from the internet, tapping “books, articles, websites and posts — including personal information obtained without consent.” Since the genesis of the flagship transformer based product ChatGPT, all eyes have been on the data scraped from all over the world wide web to train the models. Prior to that, last year, Midjourney and other text-to-image models were getting slapped with similar lawsuits. In response, Stable Diffusion and Midjourney have asked a U.S. federal court to dismiss a group of artists’ class-action lawsuit against them — arguing that that the AI-created pictures were not comparable to their work and that the AI-generated images were not similar to the artists’ work and that the lawsuit did not note specify which exact work was misused. Getting away with hefty fines With the rising number of data theft accusations, it looks like Meta, OpenAI, and others don’t care. As the heap of cases has been piling up since one can remember, the companies mostly get away with a monetary penalty or using ‘Terms & Conditions’ as an excuse to dodge the bullet. When Apple incurred five consecutive fines from the Dutch competition regulator in 2022, it surfaced a debate over whether financial penalties have any impact on big tech’s dominance of the digital economy. Meta has been paying fines since it was Facebook. The Cambridge Analytica fiasco generated headlines when the Zuck run company got away with a $5 billion fine. However, the fine was not appeased by all the FTC members. Two of the five member commissions called the fine insufficient and said it would do little to change the company’s behavior. Rather than accepting this settlement, commissioner Rebecca Kelly Slaughter believed the commision should have initiated litigation against Facebook and its CEO Mark Zuckerberg. Commissioner Rohit Chopra dissented by stating “The settlement imposes no meaningful changes to the company’s structure or financial incentives, which led to these violations. Nor does it include any restrictions on the company’s mass surveillance or advertising tactics.” Since 2015, tech companies including Google, Apple, Meta, Apple, and Amazon have collectively received penalties of over $30 billion. Fines are not just a ‘cost of doing business’ for tech giants, the president of the French Competition Authority Isabelle da Silva stated publicly. “Fines are an element of the identification of what is wrong in the conduct.” Even investment analysts agree that stock markets view investigations into big tech as a contained risk, which likely results in fines rather than business model changes. Better regulation is an emergency While policy makers have been demanding change for nearly a decade now, the regulators have responded with drafts under work in progress including the Digital Personal Data Protection Bill which recently got approved in India. Earlier this year in April, even the European Union implemented the General Data Protection Regulation (GDPR)  giving the end consumer the right to control their data. There is still no formula for punishing the companies like OpenAI for wrongdoings. With the rising number of lawsuits against the Silicon Valley’s darling OpenAI and the rest, users can expect the regulators to figure out a way for tech corporations to ethically fix themselves.","excerpt":"The Silicon Valley seems highly unfazed with the mounting legal pressure","categories":["AI Features"],"tags":["AI Tool","Big Tech","ChatGPT","data theft","Meta","MidJourney","OpenAI"],"author_name":"Tasmia Ansari","publish_date":"2023-07-10T17:00:00","publication_year":"2023","word_count":635,"keywords":["Go","ChatGPT","Meta","MidJourney","artificial intelligence","data theft","OpenAI","AI","AWS","API","Git","Big Tech","generative AI","AI Tool","R"],"extracted_tech_keywords":["AI","artificial intelligence","generative AI","ChatGPT","OpenAI","AWS","R","Go","Git","API"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/big-techs-dont-care-about-lawsuits\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":62176,"title":"How Uber Outperformed Existing GANs-Based Baselines In Self-Driving Cars","content":"Uber has further improved its self-driving cars’ performance by proposing raster-based conditional GAN architecture. This is powered by a novel differentiable rasterizer module at the input of the conditional discriminator. It maps generated trajectories into the raster space in a differentiable manner. One of the most critical pieces of the self-driving puzzle is the task of predicting the future movement of surrounding traffic actors, which allows the autonomous vehicle to safely and effectively plan its next route. The bird’s-eye view (BEV) rasterization methods and GANs collectively provide the contextual understanding, while matching the real-world distribution through the adversarial framework. Although the current techniques helped self-driving cars make effective decisions, they were far from optimal. Therefore, Uber built upon previous methods – such as top-down scene rasterization and generative adversarial networks (GANs) – to further enhance its autonomous vehicles’ performance in traffic. Uber’s SC-GAN Uber proposed a novel GAN architecture conditioned on an input raster image, referred to as Scene-Compliant GAN (SC-GAN). The critical component of the architecture is the differential rasterizer, which allows generated trajectories to be projected directly into the raster space in a differentiable manner. This simplifies the discriminator task, while allowing the gradients to flow back to the generator. This leads to higher efficiency of the adversarial training and more realistic output trajectories. According to the researchers, most GAN-based models do not condition on the scene context image in the discriminator, leading to suboptimal performance. The discriminator encodes the input trajectory with a Long Short-Term Memory (LSTM) encoder and makes classifications solely on the trajectory embeddings. Therefore, it does not identify the actual trajectory that is not scene-compliant. Model Architecture The model consists of three main modules: generator network, discriminator network, and differentiable trajectory rasterizer. Generator Network: It generates the trajectory prediction given the actor’s state input, per-actor raster, and a noise vector. The generator first extracts the scene context features using CNN. However, researchers used MbileNet to get the faster real-time inference. Besides, past observed actor states are also embedded with a shallow encoding layer and concatenated with the extracted scene context features and the latent noise vector, before passing to a trajectory decoder module that generates the trajectory predictions. Discriminator Network: It classifies whether a given future trajectory is coming from ground truth (true) or the generator (fake), conditioned on the past observed states and the scene context image. Such scene context was not used in the previous GAN-based discriminator network architecture. Therefore, researchers proposed a scene-compliant architecture that is more sensitive to noncompliant trajectories, comprising only fully convolutional layers. The proposed discriminator relies on a novel differentiable trajectory rasterizer. Differentiable Rasterizer: The trajectory rasterization module of the scene-compliant discriminator is tasked with rasterizing the future trajectory. It sequences either predicted or ground-truth information into 2D occupancy grids, where each encodes a single next trajectory point, and is a 2D grid with the same shape and resolution as the raster image. Unlike other GAN-based predictions that used vanilla cross-entropy loss as their GAN loss, the researchers used Wasserstein GAN loss with gradient penalty, resulting in outperforming different approaches. Superior Results Tested against several baselines like GAN-LSTM, S-LSTM, S-GAN, S-Way, and more, the proposed SC-GAN model overpowered the existing GAN architectures for motion prediction, reducing both average and final prediction error numbers. The researchers mentioned that SC-GAN successfully predicted cars’ movements even in somewhat challenging edge cases. For instance, when a vehicle was approaching an intersection in a straight-only lane, SC-GAN predicted that it would continue straight, even though the car’s tracked heading was slightly tilted to the left. Besides, SC-GAN rightly anticipated that a car would take a right turn after approaching an intersection in a turning lane. Outlook Various qualitative and quantitative analysis delivered results that outperformed the current state-of-the-art in GAN-based motion prediction of the surrounding actors, producing more accurate and realistic trajectories. Differentiating between motion prediction can take self-driving cars to the next level as it enhances the performance of the autonomous vehicle. This is because with this, ML models can make decisions that go beyond simple pattern matching with inputs. Besides, it can easily become resistant to adversarial attacks that try to trick the AI agents with slight changes.","excerpt":"Uber has further improved its self-driving cars’ performance by proposing raster-based conditional GAN architecture. This is powered by a novel differentiable rasterizer module at the input of the conditional discriminator. It maps generated trajectories into the raster space in a differentiable manner. One of the most critical pieces of the self-driving puzzle is the task […]","categories":["Deep Tech"],"tags":["Self Driving Cars"],"author_name":"Rohit Yadav","publish_date":"2020-04-21T11:00:00","publication_year":"2020","word_count":698,"keywords":["Go","LSTM","TPU","AI","ML","Self Driving Cars","RAG","GAN","CNN","R","adversarial attacks"],"extracted_tech_keywords":["AI","ML","RAG","TPU","R","Go","GAN","CNN","LSTM","adversarial attacks"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/how-uber-outperformed-existing-gans-based-baselines-in-self-driving-cars\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":45987,"title":"Top 5 Automation Tools That Can Speed Up Your Data Science Project","content":"Data Science has impacted a lot of businesses from different industries. While data science has managed to come so far, becoming “the sexiest job of the 21st century”, there is one more technology that is gaining prominence. Today, automation is not only limited to sectors like robotics, but it also collaborating with other domains to make things easy for techies and one such domain is data science. There are a large number of companies coming up with tools and products for data science domain. In this article, we shall look at some of these automation tools that one data science professional can use. 1. Auto-Weka There are several machine learning algorithms that can be used right off the shelf, and many of these methods are implemented in the Weka package. However, each of these ML algorithms has its own hyperparameters that can drastically change their performance, and there are a staggeringly large number of possible alternatives overall. This is where Auto-Weka comes into the scenario. Initially released in 2013, Auto-WEKA considers solving the problem by simultaneously selecting a learning algorithm and setting its hyperparameters. It also solves the problem using Bayesian optimisation. Auto-Weka is also focused on helping non-expert users to more effectively identify ML algorithms and hyperparameter settings appropriate to their applications. To know more about the tool, click here. 2. Darwin Developed by Sparkcognition, a company that builds AI systems to advance the most important interests, Darwin is another next go-to tool for solving data science problems at scale. It is an automated model building tool that allows its users to go from data to the model in significantly less time than traditional methods. Also, it enables rapid prototyping of scenarios and productive extraction of insights. Talking about how this tool works, the tool uses a patented approach based on neuroevolution that custom builds model architectures to ensure the best fit for the problem at hand. To know more about this tool, click here. 3. DataRobot Automated Machine Learning DataRobot is an advanced Enterprise AI platform. The platform incorporates knowledge, experience and best practices of some of the world’s leading data scientists. Talking about automation, DataRobot’s Automated Machine Learning platform help ML developers automate the creation of machine learning models with unprecedented transparency in order to help understand and trust the predictions they make. The platform is equipped with different types of regression techniques, ranging from the simplest to complicated statistical classic regression models. Furthermore, one of the best things about this platform is the fact that it can also solve simple problems with up to 100 different categories. DataRobot has been a sought after platform for data science professionals since the get-go. To know more about this platform, you can check out their official product site. 4. H20.ai When it comes to machine learning automation, H2O has emerged as a leader. It is an open-source, distributed in-memory machine learning platform with linear scalability. The platform is created in such a way that it supports most of the widely used statistical & machine learning algorithms. One of the best things about this platform is that it has an industry-leading AutoML functionality that automatically runs through all the algorithms and their hyperparameters to produce a leaderboard of the best models. 5. dotData Feature engineering is considered to be one of the most important, most time-consuming and challenging for data science professionals. dotData that packs the best-in-class AI capabilities works towards automating it. Simply put, the company is solely focused on democratising and automating the entire data science workflow. Compared to the traditional process, where it can take months between identifying a use case to getting pipelines into production, this AI\/ML platform helps in executing complex data science projects with speed, and at scale. Click here to get a clear picture of the platform.","excerpt":"Today, automation is not only limited to sectors like robotics, but it also collaborating with other domains. Here are 5 data science automation tools","categories":["AI Trends"],"tags":["Automation","automation tools","Data Science","data science tools","Robotic Process Automation"],"author_name":"Harshajit Sarmah","publish_date":"2019-09-16T10:59:43","publication_year":"2019","word_count":634,"keywords":["data science","Go","data science tools","machine learning","API","AI","ML","Automation","feature engineering","Scala","Robotic Process Automation","automation tools","Rust","Data Science","R"],"extracted_tech_keywords":["AI","machine learning","ML","data science","R","Go","Rust","Scala","API","feature engineering"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-5-automation-tools-that-can-speed-up-your-data-science-project\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10163743,"title":"Google Adds AI Video-Generation Model &#8216;Veo 2&#8217; to YouTube Shorts","content":"YouTube Shorts has been integrating AI features for a while now. Previously, it added the ability to generate unique AI backgrounds with just text prompts, followed by an AI-powered video editing app. Last year, it promised to equip content creators with generative AI models, and it has done exactly that. The Dream Screen feature on YouTube Shorts is getting an upgrade with the integration of Google DeepMind’s newest video generation model, Veo 2. This means you may not require standalone text-to-video AI models like Goku by ByteDance. To complement this, a new feature was added to generate standalone video clips with the help of Veo 2. Whether you need a specific scene without having any footage on it or just want to create something unique, all you need to do is enter a text prompt to generate a video clip. Along with the feature upgrades, Google has also made under-the-hood improvements to the Dream Screen feature for a faster experience. To generate the clip, you just need to open the media picker from the Shorts camera option and hit ‘Create’. Then, you can type in your text prompt, select an image, and choose the desired length. Here’s how it works: https:\/\/www.youtube.com\/shorts\/McgsOxrJiUM?feature=share In a statement, Google said, “Veo 2 understands real-world physics and human movement better, making its output more detailed and realistic. You can even specify a style, lens, or cinematic effect, making Dream Screen an easy and fun way to express yourself.” “We use SynthID watermarks and clear labels to indicate that these creations were generated with AI,” it added. The AI-powered features are currently available in the US, Canada, Australia, and New Zealand, with plans to expand further.","excerpt":"The Dream Screen feature on Shorts gets an upgrade with the integration of Google DeepMind’s Veo 2.","categories":["AI News"],"tags":["AI Video Generation Models"],"author_name":"Ankush Das","publish_date":"2025-02-14T16:28:47","publication_year":"2025","word_count":280,"keywords":["Go","TPU","programming_languages:R","AI","programming_languages:Go","AI Video Generation Models","generative AI","CLIP","R"],"extracted_tech_keywords":["AI","generative AI","TPU","R","Go","CLIP","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/google-adds-ai-video-generation-model-veo-2-to-youtube-shorts\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10084574,"title":"Microsoft Raises the Stakes on OpenAI, to Invest Additional $10 Bn","content":"Basking in the success of Codex, DALL.E and ChatGPT, Microsoft is now looking at pumping an additional $10 billion in OpenAI, a huge leap from its initial $1 billion investment in the company in 2019. The investment, which will also include other venture firms, is expected to value OpenAI at $29 billion. In 2019, Microsoft backed OpenAI with an investment of $1 billion in its cash and cloud credits. It is most likely also in the process of strengthening its search engine Bing by adding ChatGPT, alongside other enterprise applications and tools. Less than a week ago, the AI research lab announced that it is selling its existing shares at a record-high valuation of $29 billion. The acquisition of OpenAI’s existing shares, including those held by its employees, is being negotiated by investment firms like Founders Fund and Thrive Capital. According to sources, Microsoft’s funding would be a part of a complicated agreement under which the firm would receive 75% of OpenAI’s profits until it redeems its investment. Once that sum is achieved, the ownership structure would switch to Microsoft owning 49%, other investors claiming another 49%, and the rest 2% will be for OpenAI’s nonprofit parent. Additionally, there is a profit cap that differs for each group of investors, which is rare for venture deals where investors want to make a 20–30% return on their investment. The deal can be dissolved if the conditions or the investment amount alter. Read more: Google, Meta, Why NO ChatGPT? In December 2022, Microsoft entered a ten-year agreement with the London Stock Exchange Group (LSEG) and purchased nearly a 4% share in the company in exchange for providing new data infrastructure, analytics, modelling solutions, and cloud computing technologies. Microsoft reported revenues of about $198 billion in 2017, an increase of 18% over the previous year, and an operating income of $83 billion. Its net income soared by 19% yearly to $72.7 billion.","excerpt":"Microsoft will get 75% of OpenAI’s profits until it recoups its investment.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","ChatGPT","Data Science","Deep Learning","Google","GPT","Machine Learning","Microsoft","OpenAI","Satya Nadella"],"author_name":"Shritama Saha","publish_date":"2023-01-10T13:51:56","publication_year":"2023","word_count":319,"keywords":["Satya Nadella","Go","ChatGPT","API","Microsoft","OpenAI","AI","cloud computing","Machine Learning","GPT","Aim","analytics","Google","Deep Learning","Data Science","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","analytics","ChatGPT","OpenAI","Aim","cloud computing","R","Go","API","GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/microsoft-raises-the-stakes-on-openai-to-invest-additional-10-bn\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10060089,"title":"Council Post: Relevance of applying design thinking principles in AI and Analytics","content":"For more than a decade, tech giants like Apple, Netflix, Microsoft, IBM, Google and LinkedIn have been tapping design thinking principles– backed by analytics– to build a solid ecosystem to stay ahead of the competition and scale exponentially. For those unaware, design thinking is a customer-centric approach to build innovative solutions and profitable and sustainable products. Design thinking leans on empathetic observation of how people communicate\/respond with their environments, and follows an iterative, hands-on approach to create innovative products and solutions. The design principles consist of five stages – empathise, define, ideate, prototype and test. Design thinking principles (Source: Interaction Design Foundation) Today, design thinking is being taught at top universities, including MIT, Harvard and Stanford. The tech is used across manufacturing, retail, financial services, and telecom. Many companies have also started to realise the importance of design thinking combined with big data analytics to facilitate a customer-centric approach. For instance, Apple Watch uses big data analytics to keep track of people’s health, lifestyle and fitness. Netflix uses design thinking principles combined with advanced AI and data analytics to provide personalised movies and TV series recommendations. Analytics has played a critical role in shaping design thinking. It utilises a large amount of customer data to create personalised solutions–one of the core aspects of design thinking. AI and analytics Thanks to data analytics, the world of design thinking has evolved significantly. But what about the flip side? Do we adequately use design thinking principles (empathise, define, ideate, prototype and test) in analytics and AI engagements? Currently, design thinking strategy is being used by teams to solve business problems, develop customer-centric products, and scale business. Design thinking teams are open-minded, curious, collaborative, always ready for change, and adaptable. This, overall, creates a great work culture, and increases productivity. AI and data analytics teams that leverage design thinking principles to deploy machine learning models and algorithms can drive higher value from analytics engagements. The benefits of design thinking, when applied to AI and analytics, include: Elevates the thought process from being only transactional to a strategic viewHelps in developing an innovative team culture, which embraces inclusiveness, collaboration, and co-creation Allows for a diverse set of perspectives to influence the process (set up AI council of experts)Brings better AI governance into the processDevelop bias-free algorithms and models Ensures the right person is in charge Leads to faster deployment of models Increases the accuracy of the modelHelps in building AI at scale Encourages the team to use new algorithms, models and analytics tools In conclusion We need to acknowledge that applying design thinking principles will foster better throughput for AI & analytics engagement, and help increase the uptake of analytics engagements. The scope for design thinking can be varied, from building visibility on simple reporting metrics to solving a complex real-world drug discovery engagement or identifying fraud waste & abuse in multiple industries. As they say, the “days of the lone inventor are over”. With design thinking at play, organisations can focus on being “significantly outside in, highly collaborative & truly embedded. Hopefully, we should see some amazing advancement and innovation in the coming months by implementing design thinking principles in the AI and analytics space — similar to the product and service ecosystem. This article is written by a member of the AIM Leaders Council. AIM Leaders Council is an invitation-only forum of senior executives in the Data Science and Analytics industry. To check if you are eligible for a membership, please fill the form here.","excerpt":"Applying design thinking principles will foster better throughput for analytics engagements, drive higher value and help increase the uptake of analytics engagements.","categories":["AI Features"],"tags":[],"author_name":"Hari Saravanabhavan","publish_date":"2022-02-09T11:00:00","publication_year":"2022","word_count":582,"keywords":["big data","data science","Go","API","machine learning","AI","RAG","Aim","analytics","R"],"extracted_tech_keywords":["AI","machine learning","data science","analytics","Aim","RAG","R","Go","API","big data"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/council-post-relevance-of-applying-design-thinking-principles-in-ai-and-analytics\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":23395,"title":"What is Actuarial Science? How Is It Different From Data Science?","content":"The roots of actuarial science dates back to the time when the concept of compound interest was published by Richard Witt, and can be considered as one of the oldest professions. Used especially in underwriting the loans and in insurance earlier, actuaries have been one of the earliest users of data and generating insights from it. Though they have been there for the longest time, their demand have come up again with reports suggesting that there are now only 4% of actuaries in India compared to the roles that are available for actuaries. They have been preconceived to be relevant to just insurance industry (LIC, AIG etc.) and banking for long. However their work can expand to diverse industries such as banking, healthcare, pensions and benefits, asset management, capital project, investments, risks and others. They can have roles of consultants, analyst, troubleshooters, risk assessors, amongst  others in these industries. To put it simply, an actuary specialises in evaluating financial implications of risk and uncertainty, while devising solutions to reduce chances of any future risks and occurrence of any undesirable events. Though it is a very narrow field and has a specific job destination, it is one of the best paid jobs. Numbers suggest that most of the actuaries that qualify either end up leaving India for better opportunities or doing their own business. Given the sparse number of these professionals in India, their demand is at a rise. What are the skills required to be an actuary, how can you equip these skills? Being an actuary typically requires minimum of bachelor’s degree. Direct route to major in actuarial science consists of course in math, statistics, and industry related experience, however, other quantitative courses may also produce candidates for actuarial science. These majors typically include computer science, economics, mathematics, physics and statistics, humanities, English among others. Financial and statistical techniques form a basis to solve most business problems, particularly those involving risk. Actuaries extensively use core statistical training, analytical thinking, quantitative skills and advanced domain knowledge to get into the problem areas. As Rob Thomas of IBM Analytics define, actuarial science is about collecting all pertinent data, using models and expertise to factor risk and making decisions. Dynamic risk management entails real time decision making based on a stream of data, he had said. As they also work with large amounts of data, actuaries need the ability to subdue unstructured data, have better programming skills, visual representation experience and skills to build models in the big data world. They are also expected to have excellent communication skills to enable them to communicate ideas to their customers. In a nutshell, some of the skills that actuaries are expected to have are: Analytical problem solving skills- Since they are tasked with examining complex data and identifying trends, analytical problem solving remains a key skill that will help them  look for ways to minimize the likelihood of undesirable outcome. Computer skills- Computer along with a variety of statistical modelling software is required that will help them to evaluate large numbers of data. Undergoing courses in SAS, VBA may be required. They should be equipped with formulating spreadsheets, database manipulations, statistical analysis programs, programming languages etc. Math skills- Since they deal with numbers, being quick and correct in math skills is certainly required. Knowledge in calculus, probability, statistics and others will help. Understanding of business and finance- As they would find themselves being employed with insurance and financial institutions, a fair amount of understanding on how these industries work is an advantage. Good business sense will help in devising solutions for financial risk and providing expert opinion. Communication skills- As mentioned earlier, since actuaries connect with various personnel including programmers, accountants, and senior management, to be able to communicate effectively is the key. Some of the colleges in India that provide exclusive courses in actuarial science are Aligarh Muslim University, Andhra University, University of Mumbai, University of Delhi, University of Madras, AMITY, amongst others. Several private institutes are also booming in the space such as IIRM Hyderabad, Chitkara University and others, providing bachelors and masters in actuarial science. Actuarial science, analytics and data science As we earlier saw actuaries deal with data, they even serve an important role with predictive analytics by using modelling and data analysis techniques on large data sets to discover predictive patterns and relationships for business use. They use analytics extensively in their working to transform data into useful information. Therefore, there is a close proximity that actuaries have with analytics and data science fields. While they both revolve around data, there are certain lines drawn between these two. For instance, actuaries are found primarily in the insurance industry, primarily for risk assessment, data scientists can be found in virtually any industry. If we have a comparative lookout at actuarial science and data science in detail, while the prior is about study of finance and related fields’ activities, and latter is about studying different data sets, their relationship and analysis. While AS covers financial modelling and studying data around it, DS covers aspects of databases, data mining, machine learning, data visualisations, and others. Unlike data science, actuarial science is strictly domain specific. It can be said that data science skills are great to have in actuarial practise, but one doesn’t need them necessarily to be an actuary. While they both share same responsibilities, their education and skill sets may differ. Actuaries generally make use of SAS, Excel, VBA, and SQL, MoSes and Prophet on a frequent basis, data scientists are more programming savvy with an expectation of the know how of C++, R, Python and NoSQL databases, Hadoop, etc. Though there are differences between the two fields, the actuarial employers are increasingly expecting their staff to have same skill sets as data scientists. With the two fields heading towards creating blur lines between themselves, it wouldn’t be surprising to see actuaries being called data scientists in the coming future.","excerpt":"The roots of actuarial science dates back to the time when the concept of compound interest was published by Richard Witt, and can be considered as one of the oldest professions. Used especially in underwriting the loans and in insurance earlier, actuaries have been one of the earliest users of data and generating insights from […]","categories":[],"tags":[],"author_name":"Srishti Deoras","publish_date":"2018-04-09T08:45:05","publication_year":"2018","word_count":988,"keywords":["data science","Go","machine learning","AI","R","Python","C++","analytics","SQL","predictive analytics"],"extracted_tech_keywords":["AI","machine learning","data science","analytics","predictive analytics","Python","R","SQL","Go","C++"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/what-is-actuarial-science-how-is-it-different-from-data-science\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":21669,"title":"Why Leading Institutes In India Are Focusing On Industry Relevant Research?","content":"Industry-University partnerships has been the usual norm for global IT giants to build on a growing pool of academic talent and solve real-world challenges in a constructive way. Now, this trend is finding a resonance in India as well. Take for instance – leading Indian e-commerce player Flipkart banking on the expertise of IIT Kharagpur’s Complex Networks Research Group since last year. And the Flipkart-CNeRG collaboration has seen several papers come out on conversational dynamics and chatbot. Earlier in 2016, SBI teamed up with IIT Kharagpur or collaborative research and finding POCs. With India’s most trusted bank pushing for digital transactions, it was little wonder that the bank looked to IIT Kharagpur’s heft to develop a wider program that spanned ideation to incubation and productizing technology driven products and services for the banking sector. In a similar vein, IIT Bombay also teamed up with SBI in 2016 to find fintech start-ups that have products or applications that cater to the financial sector. The bank had provided a platform for testing and piloting such products and check for any specific business linkages. Earlier last year, Microsoft partnered with IIT Kharagpur Professors for a search algorithm research that could help users by drumming up subjective information and trusted opinions, the company said in a statement. According to the company press release, while the existing search engine algorithms worked well with fact-based queries and providing structured answers, they were ineffective in responding to subjective queries. University Partnership with Indian Industry Leads to Productive Collaborations Indian IITs have always been an attractive option for industry partnerships and this kind of strategic collaboration has given rise to new technologies, transformed and modernized industries. And it’s not just IT giants who benefit largely from academic partnerships – a hefty and steady stream of industry funding gives an excellent opportunity to professors and graduates to work on cutting-edge research and develop solutions that can have a global impact. Nowhere is the industry-university partnership more apparent than Silicon Valley which flourished thanks to the proximity with Stanford University, Caltech, University of San Diego (UCSD) and has seen long-running collaborations since the last five decades. Similarly, Boston is the new upstart in the start-up base race with its proximity to MIT and Harvard. India’s Growing Heft In Science Research On to India, another reason that has contributed significantly to industry-university collaboration is India’s growing heft in science research. Too often we hear the statement that India is lagging behind in innovation and R&D spend, but a recent economic survey points out India’s golden years between a period of 2014-17 that saw research publications jumping from 35,376 to 1,15,393, as per Scopus database. The Science Economic survey states that India is ranked # 5 in terms of number of publications globally and ranks 3rd in CS, engineering and chemistry. The significant increase in research has been attributed to the rise in investments from Indian IT companies that has led to productive collaborations. Here’s a look at top University-Industry collaborations that go beyond IITs 1) CA Technologies teams up with IIIT-Hyderabad to set up co-innovation lab: Last year, New York-headquartered CA Technologies teamed up with IIIT-Hyderabad researchers will collaborate on applied research projects that will include next generation interfaces such as Natural Language Processing (NLP), Artificial Intelligence (AI) and ML (Machine Learning), as per the company statement. According to Surya Panditi, general manager and senior vice-president, CA Technologies, “We aim to combine the strengths of research with industry knowledge for significant breakthroughs in the fields of Artificial Intelligence and Machine Learning. This partnership is another step in our relentless focus on removing the barriers between ideas and outcomes, while deepening our commitment to actively engage with the academia.” Key Outcome: For IIIT-Hyderabad, the key outcomes are white papers that will be published in science journals while CA Technologies will benefit from reference architecture and prototypes that can be adopted for swift development. 2) IIT Madras tops in industry consultancy: IIT Madras is coming out tops when it comes to forging business partnership and industry consultancy. As per a news report, the National Institutional Ranking Framework (NIRF) -2017 data by IIT Madras, puts IIT M amongst the top 5 IITs in terms in terms of industrial consultancy. The data shows that in the period of 2015-16, IIT M netted 737 projects from 383 clients. IIT Madras Research Park – the first-of-its-kind research focused initiative which operates like an independent company promotes works in partnership with several leading companies. It helps companies with a research focus to set up niche labs and profit the academic expertise of IIT Madras. The strategic plan for 2014-2020 clearly outlines the agenda for the institute to be sought for international collaborations and establish a track record as a creator of new and innovative technologies for industrial and national needs. Key Outcome: The fact that IIT Madras wants to take a lead in national and industry developments is evident with the recent announcement by Arun Jaitley at the Union Budget 2018 wherein the Department of Telecom (DoT) will support a 5G test bed in collaboration with IIT Madras. According to Hemant Joshi of Deloitte, upcoming 5G center in collaboration with IIT Madras would exponentially help India become an early adopter and bridge the digital divide. 3) IISc inks partnership with industry giants: IISc has seen an active years when it comes to inking industry partnerships with IT giants like the MoU with German tech giant Bosch. Recently, Bengaluru’s leading tech institute inked a deal one of the world’s largest French automotive equipment supplier Faurecia o conduct collaborative research to develop next-gen technologies and solutions in three areas – a) online air quality monitoring b) data analysis and algorithms for driver behaviour c) most importantly artificial intelligence for industrial design. Not just that, recent reports indicate that IISc faculty member Professor Partha Talukdar, and a well-known figure in Machine Learning circle who heads the Machine and Language Learning Lab at IISc is in talks with Flipkart to work on NLP-related areas. Key Outcome: This industry-academia research collaboration will not just lead to automotive solutions based on next-gen tech but will drive a more industry-focused research that will have a significant impact on mobility. AIM Summary There is an increasing shift from a publication driven research to industry driven research as is evident from the IIT-M model. The concerted push by Indian institutes to make research more industry driven would help in attracting industry funding With Indian IT companies looking for focused solutions, institutes such as IIT-Delhi are looking for an equal split between government and industry research IIT Madras tops the chart in industrial consultancy is even mulling setting up an cell for IP management with patent agents Reportedly, IIT M has also staffed experienced professionals as industry relation advisors to boost industry research collaboration The recent National Institutional Ranking Framework (NIRF) for IIT-H shows the institute worked on a total of 42 consultancy projects and 35 client organizations","excerpt":"Industry-University partnerships has been the usual norm for global IT giants to build on a growing pool of academic talent and solve real-world challenges in a constructive way. Now, this trend is finding a resonance in India as well. Take for instance – leading Indian e-commerce player Flipkart banking on the expertise of IIT Kharagpur’s […]","categories":["IT Services"],"tags":["IIT Delhi"],"author_name":"Richa Bhatia","publish_date":"2018-02-13T09:14:29","publication_year":"2018","word_count":1160,"keywords":["Go","machine learning","artificial intelligence","AI","ML","RAG","NLP","Aim","Rust","IIT Delhi","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","NLP","Aim","RAG","R","Go","Rust"],"url":"https:\/\/analyticsindiamag.com\/it-services\/leading-institutes-india-focusing-industry-relevant-research\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10081583,"title":"ChatGPT is Now Available on WhatsApp","content":"US-based entrepreneur and the co-founder of Cue (acquired by Apple), Daniel Gross, has integrated GPT-3.5 (that is used to build ChatGPT) into an AI-enabled assistant on WhatsApp. By embedding GPT-3.5 into WhatsApp, Gross was able to chat with the bot as if it were a friend on his contact list. He calls it WhatsApp-GPT. The code is available on GitHub. Welcome to the future of smart agents! I hooked up ChatGPT to WhatsApp — pic.twitter.com\/3kawekZebM— Daniel Gross (@danielgross) December 2, 2022 According to Gross, he ran WhatsApp from a phone using the golang library then simultaneously ran ‘ChatGPT’ on a browser. He also points out that it might not work on the first go. However, it must be noted that the official API for ChatGPT is not available yet. Tech influencer Varun Mayya also announced something similar on Twitter. He said, “Announcing God In a Box, a direct implementation of GPT-3.5 into WhatsApp. Unlike other implementations, ours uses the brand new model and API, and doesn’t require credentials. We’re launching tonight, but you’ll need a beta key.” However, you can join the waitlist even if you do not have the beta key. Inspired by Gross, Twitter user Altryne has even managed to embed ChatGPT in Telegram. https:\/\/twitter.com\/altryne\/status\/1598822052760195072 Since its launch, ChatGPT has taken the internet by storm and already clocked a million users in a few days. What makes ChatGPT impressive is that besides answering general queries, it can explain codes and scientific concepts, write basic academic essays, and even scripts for rom-coms. The model is released in its Beta version and is free to use. However, Sam Altman, founder of OpenAI, said they would consider monetising the model with an average cost of single-digit cents per chat. He also added that monetisation particularly becomes important as “the compute costs are eye-watering”. Exciting but overlooked that ChatGPT is primarily an alignment advance—the base model (GPT-3.5) has been available in publicly for many months, but making it into a useful chat system required significant strides with reliably following the intent of the developer and the user. https:\/\/t.co\/jy7DKkqTWI— Greg Brockman (@gdb) December 3, 2022","excerpt":"The official API for ChatGPT has not been released yet.","categories":["AI News"],"tags":["ChatGPT"],"author_name":"Pritam Bordoloi","publish_date":"2022-12-06T12:31:07","publication_year":"2022","word_count":352,"keywords":["Go","ChatGPT","API","OpenAI","AI","Git","RAG","GPT","GitHub","R"],"extracted_tech_keywords":["AI","ChatGPT","OpenAI","RAG","R","Go","Git","GitHub","API","GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/chatgpt-is-now-available-on-whatsapp\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":38887,"title":"A Day In The Life Of: Ezetap Analyst Who Wants To Step Up To The Data Scientist Role","content":"As an Analyst who is an aspiring data scientist and a night owl who creates AI-powered dashboards for customer interaction, Sujeet Nayak from Ezetap leads a busy life. We talked to Nayak, who works in the Customer Solutions team, to give us a perspective about his work, life and working with emerging tech. Each week, for our column ‘A Day In The Life Of’, we talk to awesome techies from various organisations who are working in areas like big data, advanced analytics and artificial intelligence. Nayak, who has a buoyant personality and lively voice, told Analytics India Magazine that the key to his creativity at the workplace lies in his morning ritual. “My day begins at 8 am with a cup of warm lemon water and then referring to the quick notes for the day about all the tasks to finish as a priority,” he says. He says that the best part about working for a fintech startup was its flexible timings, but adds that the adrenaline rush for him comes from taking on more responsibilities, working with new ideas and putting his best effort to achieve targets. But that does not mean that it’s all work and no play for Nayak. “My exciting job doesn’t dictate my personal life. I love travelling — and on long weekends, I mostly go on trips. On Saturday nights, I generally make it to the highways to enjoy a short night spin,” he says enthusiastically. Explaining about his work in detail, Nayak says that as an Analyst in the Customer Solution Team, his work is primarily focused on interacting with their existing customers as well as the ones who have just come on board. “I have to coordinate with the internal team of the organisation, understand the business and create an end-to-end solution for them. Some of the examples are Analytical Portals, Dashboards and even reports. This involves understanding the data, cleaning up the data for business use, followed by coding and validating the data, before I present the same,” says Nayak. Ezetap, the company Nayak works for, provides mobile point-of-sales payment solutions in India. They also offer card swipe devices and mobile applications for businesses and financial institutions to complete financial transactions. The company’s solution features remote pay, universal payment acceptance, multi-bank EMI, on-us routing, multi-TID, smart charge slips, universal wallet acceptance, and multiple device lines. Ezetap was founded in 2011 and is based in Bengaluru. When asked about what were the best and the worst parts of his job description, Nayak says, “On a typical workday, I really feel good when I solve business-related issues. Sometimes, they’re way trickier and solving them can take hours. The worst part is when I have plenty of things in my bucket and then some existing issue makes a comeback. Choosing one task as a priority over another does matter a lot in my workplace,” he says. Five years down the line, Nayak sees himself as a full-fledged Data Scientist. In fact, he is reskilling himself and studying for it. “My short term goal is to learn new skills, to explore new things and implementing those, to become a perfect Data Scientist,” he concludes.","excerpt":"As an Analyst who is an aspiring data scientist and a night owl who creates AI-powered dashboards for customer interaction, Sujeet Nayak from Ezetap leads a busy life. We talked to Nayak, who works in the Customer Solutions team, to give us a perspective about his work, life and working with emerging tech. Each week, […]","categories":["AI Features"],"tags":["Data Science","Data Science Career","Interviews and Discussions"],"author_name":"Prajakta Hebbar","publish_date":"2019-05-10T06:50:36","publication_year":"2019","word_count":529,"keywords":["big data","Go","artificial intelligence","startup","programming_languages:R","AI","Data Science Career","ViT","analytics","GAN","Data Science","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","artificial intelligence","analytics","R","Go","big data","GAN","ViT","startup","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/a-day-in-the-life-of-ezetap-analyst-who-wants-to-step-up-to-the-data-scientist-role\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10160820,"title":"Satya Nadella Meets Prime Minister Narendra Modi, Vows to Make India AI-First Nation","content":"Microsoft CEO Satya Nadella met with Indian Prime Minister Narendra Modi on Monday, commending his leadership and reiterating the company’s commitment to AI-driven growth in India. “Thank you, PM Narendra Modi ji for your leadership. Excited to build on our commitment to making India AI-first and work together on our continued expansion in the country to ensure every Indian benefit from this AI platform shift,” Nadella said in a post on X. In response, Modi said, “It was indeed a delight to meet you, Satya Nadella! Glad to know about Microsoft’s ambitious expansion and investment plans in India. It was also wonderful discussing various aspects of tech, innovation and AI in our meeting.” Nadella is currently touring India as part of the Microsoft AI Tour, where he will be speaking in Bengaluru on January 7, 2025, and in New Delhi on January 8, 2025. Last week, Nadella met with Telangana Chief Minister A. Revanth Reddy in Hyderabad to discuss the state’s technology priorities, including AI, generative AI, and cloud development. During his visit to India last year in February 2024, Nadella announced ambitious plans to train 2 million people in AI skills by 2025. This initiative, known as ADVANTA(I)GE INDIA, aims to address the AI skills gap and create new opportunities across the nation. Nadella’s tour included meetings with top business leaders and government officials. In Mumbai, he highlighted the potential of AI to drive GDP growth, citing that out of India’s targeted $5 trillion GDP by 2025, approximately $500 billion could be driven by AI. Microsoft India employs over 20,000 people across 10 cities, including Ahmedabad, Bengaluru, Chennai, Gurugram, Delhi, Noida, Kolkata, Mumbai, and Pune, with Hyderabad alone accounting for half the workforce at 10,000 employees. For the fiscal year ending March 30, Microsoft’s India business posted a 38.44% increase in net profit year-on-year, driven by the steady growth in cloud adoption and AI.","excerpt":"Nadella is currently touring India as part of the Microsoft AI Tour, where he will be speaking in Bengaluru on January 7, 2025, and in New Delhi on January 8, 2025.","categories":["AI News"],"tags":["Microsoft"],"author_name":"Siddharth Jindal","publish_date":"2025-01-06T21:28:07","publication_year":"2025","word_count":315,"keywords":["Go","AI-first","programming_languages:R","AI","innovation","programming_languages:Go","Aim","generative AI","GAN","R","Microsoft"],"extracted_tech_keywords":["AI","generative AI","Aim","R","Go","GAN","AI-first","innovation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/satya-nadella-meets-prime-minister-narendra-modi-vows-to-make-india-ai-first-nation\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10133381,"title":"Retail Brand Seematti Unveils India’s First AI Fashion Ambassador","content":"In a groundbreaking move, Seematti, a century-old textile retail chain, has introduced India’s first ever AI-powered fashion brand ambassador, Isha Ravi. The AI model depicts a young woman with a passion for colour and travel, representing the brand in the Indian fashion landscape. Seematti CEO and Lead Designer, Beena Kannan, emphasised the significance of this technological leap in the fashion industry. “This fusion of technology and fashion marks a new era,” she said. “We’re proud to be pioneers in India, showcasing the potential of AI in the fashion world.” As India’s first fashion brand to embrace an AI ambassador, Seematti aims to revolutionise the shopping experience. The company plans to leverage AI technology to offer personalised recommendations and innovative features. AI in Fashion The fashion industry witnessed significant developments in the AI space using GenAI particularly over the last six months of 2023. McKinsey analysts estimate that generative AI could add a significant amount to the profits of the fashion industry, with potential gains ranging from $150 billion to $275 billion by 2030. E-commerce platforms such as Myntra use AI features like My fashion GPT for shopping seemingly making it a more interactive and user friendly experience. The platform had earlier introduced Maya, an AI influencer but the feature is no more available for its users. Google has also  introduced a new virtual fitting room experience that leverages generative AI to display clothing on a variety of real models, helping customers visualise how garments will fit and look on different body types and skin tones. Others are also following suit. International Fashion Brands and designers use AI to create new designs and experiment with different styles, Rebeca Minkoff has used AI to generate new patterns for their handbags. While The Fabricant creates digital fashion items that can be worn virtually or physically. They have used AI to generate new designs and even create a digital fashion show.","excerpt":"This is the first time Seematti is leveraging AI for their consumers.","categories":["AI News"],"tags":["fashion","fashion ai"],"author_name":"Tanisha Bhattacharjee","publish_date":"2024-08-21T16:47:48","publication_year":"2024","word_count":317,"keywords":["Go","GenAI","AI","Git","RAG","GPT","Aim","fashion","generative AI","R","fashion ai","llm_models:GPT"],"extracted_tech_keywords":["AI","generative AI","GenAI","Aim","RAG","R","Go","Git","GPT","llm_models:GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/retail-brand-seematti-unveils-indias-first-ai-fashion-ambassador\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":18001,"title":"Sushant Ajmani Of Blueocean Explains Meaning Of Consumer Experience In The Light Of Analytics Solutions","content":"One of the biggest challenges the companies face these days is how there is an enormous amount of quantitative and qualitative data getting generated at each touchpoint with the consumers. It requires constant mining in order to drive meaningful insights, and recommend in real-time the next best offer (NBO) or next best action (NBA) for the customers. Sushant Ajmani, Vice President of Digital Solutions and Consulting at Blueocean Market Intelligence spoke on these lines at Cypher 2017, India’s most exciting Analytics summit. People-based marketing is a hyped term, made popular a few years ago by Facebook. It basically means how to identify the unique identity of each and every user who interacts with your brand, whether it’s online or offline. Using their 360-degree approach, Ajmani and his team ensures that all available structured and unstructured data sources is used comprehensively, thus bringing out the best to bear against each engagement. Current Context And Sentiments Of Users Play A Huge Part: Citing a specific example from the telecom sector, Ajmani said explained that how each and every part of the user experience with a particular company generates a lot of data. For example, if a user goes to a website of a telecom company, it may be for one of the two reasons — to seek support or to change, browse through plans. The entire website, the usability, the flow, the content as well as the existence (or non-existence) of broken links plays a huge role in the user experience. “One of the interesting things we have noticed in the last few years is that the current context and the current sentiment of the user plays a pivotal role in the whenever we talk about understanding the consumer,” said Ajmani. Ajmani said that they had done a study with a one of the largest telecom operators in the U.S., and found out that the Procurement Manager of the company, who had been using the same telecom services at home, was frustrated with them as he had experienced bad service. So now, his professional opinion about the said company was also affected by his recent or current experiences, explained Ajmani. What’s The Problem, Then? Ajmani said that according to their studies, almost all of the Fortune 500 companies had invested heavily in technology designed for good consumer use. “They already have the last mile piece sorted out,” he said. All these companies are collecting feedback from their consumers and their experiences across all fields — online as well as offline — but missing the ‘consumer engine’ which would churn out intelligent insight from the data. What Does Analysis Of Consumer Experience Mean? Ajmani jotted down the following pointers for the delegates’ understanding. He explained that consumer experience can be achieved with the help of: Outside-to-inwards thinking Technology infrastructure Digitisation of the processes Enabling employees and vendors Cohesiveness of the solution Shared goals Data-driven culture He also added that a strong decision support is implemented by a harmonious amalgamation of analytics, domain expertise, engineering and visualisation skills.","excerpt":"One of the biggest challenges the companies face these days is how there is an enormous amount of quantitative and qualitative data getting generated at each touchpoint with the consumers. It requires constant mining in order to drive meaningful insights, and recommend in real-time the next best offer (NBO) or next best action (NBA) for […]","categories":["Deep Tech"],"tags":["Cypher"],"author_name":"Prajakta Hebbar","publish_date":"2017-10-03T11:12:45","publication_year":"2017","word_count":504,"keywords":["Go","programming_languages:R","AI","data-driven","programming_languages:Go","Git","analytics","Rust","R","programming_languages:Rust","Cypher"],"extracted_tech_keywords":["AI","analytics","R","Go","Rust","Git","data-driven","programming_languages:R","programming_languages:Go","programming_languages:Rust"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/sushant-ajmani-blueocean-cypher-2017\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":54813,"title":"How To Hire Data Scientists For Your Startup","content":"Data science skills are considered to be most sought after in data-driven organisations. While the right data scientist can assist the company in achieving its business objective, a misfit professional can have a negative impact on the firm. However, unlike software developers, hiring data scientists can be strenuous and stressful for startups. Consequently, startups should be cautious while hiring developers who will be an integral part of the business. Today, hiring a proficient data scientist in startups is increasingly tricky as developers in the industry also have similar skills. To make the matter worse, often they even have similar projects too. Therefore, determining the right data scientist is crucial for startups to avoid any deterrence in their business operations. Here are a few best practices for evaluating and hiring the perfect data scientists for your startups. Focus On Data Science Skills You Need The Most Writing an appropriate job description is vital to get the desired application. Often, the job posting scares data scientists as it includes almost all the skill present in that domain. This decreases the chances of getting candidates with desired knowledge. Although irrespective of the job description, most of the aspirants apply for the role without analysing the requirements. But the idea behind a concise job description is to get the appropriate application of aspirants relevant for a specific role within the landscape. Undoubtedly, figuring out candidates who applied solely for a position mentioned in the job description is still challenging. Still, such practices will increase the chance of hiring the candidate you want. For one, if your business operations mostly engage with video files, then mentioning computer vision and image processing instead of including other similar skills like text analysis is what you should cling on to. Finding a ‘jack of all trades and a master of none’ can afflict the business and may also lead to financial and customer loss. And therefore, startups must seek relevant skills. Ability To Convert Business Problem Into Data Science Problem “Converting business challenges into programming problems is an essential skill that is needed in startups,” said Santosh Rai, AI architect and a head data scientist at ProVise Consulting. And not just startups, this ability is required in any company as data scientists help in assisting decision-makers in making business decisions. Consequently, along with data intuition skills, having someone who can make business problems straightforward and then solve it through data science skills is crucial. Look For Someone Who Can Build Data Pipeline As productivity is essential for startups, it would be ideal for you to get a developer who can devise data pipelines. Although data pipelines are considered to be the job of data engineers, getting hold of a developer who can streamline data science processes will be hugely beneficial for startups. A robust data pipeline is vital to gain operational resilience within data-driven organisations. Therefore, to carry out the development process effortlessly, hiring developers who can assist in creating data pipelines will be an added advantage. Projects Not Certificates To understand the expertise of developers in data science, one should evaluate diverse projects in an applicant’s portfolio. Certifications cannot demonstrate the skills they have obtained; putting the knowledge into use is what makes the difference. Therefore, having a candidate who has hands-on varied projects will be ideal for startups. “With changing times, the industry requirements for data scientists have evolved too. Previously organisations were looking for candidates with specific certifications that were required for working on vendor-specific platforms. Today, the industry is looking for data science stack skills in developing programs that require augmented certification along with hands-on experience on projects,” said Vishal Chahal, Asia pacific leader of the technical elite team for data warehouse & AI at IBM. Outlook Apart from the practices mentioned above, it would be best if you looked for developers’ achievements in the domain. Startups can also analyse the expertise by looking at their contribution to Kaggle, GitHub, StackOverflow, and LinkedIn to determine both technical and soft skills of the developers. “Today soft skill has become the differentiating factor as often data scientists have to explain factors behind their outcomes to stakeholders,” said Sharath Kumar, a data scientist at IBM India Software Labs.","excerpt":"Data science skills are considered to be most sought after in data-driven organisations. While the right data scientist can assist the company in achieving its business objective, a misfit professional can have a negative impact on the firm. However, unlike software developers, hiring data scientists can be strenuous and stressful for startups. Consequently, startups should […]","categories":["AI Startups"],"tags":["data analytics certificate"],"author_name":"Rohit Yadav","publish_date":"2020-01-29T12:00:00","publication_year":"2020","word_count":701,"keywords":["data science","AI","data analytics certificate","ML","data pipeline","Git","computer vision","GAN","GitHub","R","data warehouse"],"extracted_tech_keywords":["AI","ML","computer vision","data science","R","Git","GitHub","data pipeline","data warehouse","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-startups\/how-to-hire-data-scientists-for-your-startup\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10000676,"title":"India Is On The Speedway To Making E-Vehicles Popular","content":"Electric vehicles have been trending in the automobile industry since the beginning of the 21st century. Several tech companies like Tesla have created world-class e-cars and e-buses. Maglev train in South Korea and Solar Impulse 2, an e-aircraft have revolutionised the industry. These vehicles are just as efficient as normal vehicles and help save the environment from air pollution. In India, there’s been a wave to adopt new technologies and create a similarly advanced market ecosystem, as the developed countries. E-vehicles can change the current situation of many cities like Delhi NCR, Bengaluru and Kolkata. E-vehicles are not restricted to cars, there are also buses, rickshaws, aeroplanes and trains which can run on electricity. E-vehicles can function using different pathways of energy source. These vehicles are equipped with electric motors which help the vehicle to run using the generated electricity. The electricity can either be provided through collector systems or can be self-contained (using batteries). Solar panels are also used to run e-vehicles along with an electric generator, which helps in converting fuel to electricity. India’s Race To The Unconventional Market In India, the e-vehicle market is at a nascent stage. Since most of the private companies couldn’t scale this idea to a larger audience. The e-vehicle market is now in the hands of the government. In Jan 2018, several cabinet ministers proposed the commercial use of Indian Space Research Organisation’s (ISRO’s) Lithium (Li) ion battery technology for e-vehicles. The government aims to promote the use of Zero Emission Vehicles (ZEV) and decrease India’s dependence on oil imports, this could easily bring down the alarming state of air pollution in India. According to reports, there are 170 million two-wheelers in India currently being used, these vehicles draw out 34 billion litres of petrol every year. The lithium-ion battery is one of the feasible options for the Indian e-vehicle market. Since the imported fuel is costing the government a lot of money, there’s an increase in demand for alternative fuel. Lithium-ion batteries require three components, namely, cell to battery packaging manufacturing, cell manufacturing and battery chemicals. The batteries require elements like Li, cobalt, manganese, nickel and graphite. Currently, the batteries are imported from Japan and China. The government aims to cut down on this dependence as well, ISRO’s technology has been accepted by the government to manufacture Lithium-ion batteries in India and use them in vehicles. Government Interest After the proposal by the ministers to take ISRO’s project into reality. Vikram Sarabhai Space Centre (VSSC), an institution under ISRO collaborated with NITI Aayog to transfer its technology to other companies. This project aims to push startups and industries into Li-ion battery manufacturing. These batteries can be used for different purposes of power storage such as in phones, laptops, camera and other portable gadgets. The organizations selected 14 Indian companies for this project. The license fee for the project is Rs.1 crore, the startups have signed a Memorandum of Understanding (MoU) with VSSC for the technology transfer. The union aims to add 175 gigawatts (GW) to the renewable energy capacity of India by 2020. Under the Make In India initiative, the government is ready to invest in e-vehicle adoption in the market. To cut down costs, it is supporting indigenous companies to build Li-ion batteries. The initiative also involves the construction of efficient charging infrastructure, since electric vehicles need charging as their fuel. NITI Aayog is proceeding forward with the objective to increase the efficiency of these vehicles by changing the chemistry of the batteries to produce more power and using better tyres and electric motors. One of the key points to improve the vehicles’ performance is to enhance the aerodynamics and reduce the weight of the machine. India needs to expand its electric power by 10 GWh of cells by 2022, this can be done by recycling Li-ion batteries efficiently and secure the main constituents of the battery.","excerpt":"Electric vehicles have been trending in the automobile industry since the beginning of the 21st century. Several tech companies like Tesla have created world-class e-cars and e-buses. Maglev train in South Korea and Solar Impulse 2, an e-aircraft have revolutionised the industry. These vehicles are just as efficient as normal vehicles and help save the […]","categories":["AI Features"],"tags":["electric vehicles","Interviews and Discussions","ISRO"],"author_name":"Jignasa Sinha","publish_date":"2018-12-26T20:26:06","publication_year":"2018","word_count":647,"keywords":["Go","ISRO","startup","AWS","AI","cloud_platforms:AWS","programming_languages:R","RAG","Aim","electric vehicles","GAN","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","Aim","RAG","AWS","R","Go","GAN","startup","cloud_platforms:AWS","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/india-is-on-the-speedway-to-making-e-vehicles-popular\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":12652,"title":"Why Big Data has been a Disappointment till now","content":"There is a whole generation of management and technicians that have been sold on the proposition that there is unfound gold in Big Data. Many managers – IT and non-IT alike – have bet their careers and corporate credibility on the proposition that somebody really clever can go into the reams of Big Data in the corporation and come out with information so hidden and so important that there will be a renaissance of technology in the corporation. This “hidden” gold phenomenon has given rise to the day and age of the “data scientist”. Has anyone noticed how much data scientists are making in Silicon Valley these days? It makes me want to go and rewrite my resume. Big Data continues to fall But what has the reality been for the hidden gold rush? Corporations have poured millions of dollars into the digging for gold. But the results have been very disappointing so far. If some industry had found a pot load of gold hidden in their Big Data you can believe that the discovery would have been heralded across the landscape. But no such pronouncements have been made. Big Data continues to fall down the backside of the Gartner hype curve. When it will fall into the Gartner trough of disappointment (a standard part of the Gartner hype curve) is anyone’s guess. Big Data has been hyped so much by so many people that it has a long way to fall before reaching the trough of disappointment. So what is happening here? Why has there been this disappointment in Big Data? Why haven’t the brilliant (and highly paid) data scientists been able to find all of the hidden gold that was promised? There are (at least!) three possible scenarios. Scenario 1. There just isn’t any gold there to be found. Even the most brilliant data scientist cannot find gold if it isn’t there. Finding gold that isn’t there is called alchemy and is beyond the purview of even the most clever of data scientists. I have an acquaintance in the utility business. His vendor told him to convince management to spend a lot of money putting metering data into a pile of Big Data. His vendor wined and dined upper management and the next thing you know a $10 million dollar project was borne. Data scientists madly went about capturing metering data for the past ten years. (Metering data is the monthly calculation of how much gas and electricity a business or a home uses each month.) The next thing you know a Big Data project held reams and reams of metering data. Then the data scientists went to work. They ran their algorithms. The scoured the data. They threw out outliers. They fit their curves. But three years later – no magic data, no hidden gold. There just are no hidden nuggets of gold in metering data, despite the sales pitch given by the vendor. All the data scientists found was that when it got colder people turned up their thermostats. But anybody could have told them that. They didn’t need to spend $10 million dollars to find that out. Stated differently, there just isn’t any gold there to be found, no matter how clever the data scientist has been. Scenario 2. There is some gold there but it is really hard to find, perhaps harder than the skills of even the most clever data scientist. There are lots of reasons why finding gold in big data (if it is even there) is so hard to do. There is a lot of data in Big Data. Hidden gold has a lot to hide behind. In California during the gold rush, the miners found out that they had to remove tons of dirt in order to get their hands on just a little gold. Big Data is the same way. You have got to remove huge amounts of unnecessary data to get at the really interesting data. Or another reason the golden data is so hard to find is that it is disguised. When looking through the reams of data found in Big Data it just is not readily apparent what the golden data looks like. Yet another reason why golden data is so hard to find is that the data must be paired with other data in order to determine its value. The problem with pairing any data in Big Data with any other data is that there is a lot of data there. Unless the data scientists knows exactly how to pair the data together, the data scientists spends a lot of (expensive) time spinning his\/her wheels. So even if there is hidden gold there, finding it is no bargain. Scenario 3. Even if there is hidden gold in Big Data, the cost of getting that gold is more expensive than the gold is worth. As a case in point, it is estimated that below Central City, Colorado there is buried over $18 billion dollars worth of gold. (That’s the truth. I am not kidding.) So why don’t you jump into your car, grab a pick and a shovel and head for Central City and start digging. Well there is a problem. In order to get to $18 billion in gold, you would have to spend (an estimated) $180 billion dollars. It costs more to get to the gold than the gold is worth. The same phenomenon holds true of Big Data. The overhead of doing Big Data is tremendous. There is the cost of storage. There is the cost of software. There is the disruption to everyday activities. There is the exorbitant cost of the data scientist. There is general corporate overhead. When you get through adding up the real costs of a Big Data experiment, you find that even if you find gold, it is so expensive that it is not worth the effort. So there are a lot of obstacles in front of the data scientist. Maybe there is a good reason why Big Data is in a freefall down the Gartner hype curve. Who knows when Big Data will find the bottom of the trough of disappointment? And who knows where t will go once it struggles up out of the trough of disillusionment. As for management who has bet his\/her career and credibility within the corporation on all the gold hidden in Big Data, maybe it is time to get the resume polished up. There is always Uber. Now you can hear and see Bill Inmon on the Internet. Take a look at his new videotape education series on www.safaribooksonline.com. Bill covers IT topics from A to Z.","excerpt":"There is a whole generation of management and technicians that have been sold on the proposition that there is unfound gold in Big Data. Many managers – IT and non-IT alike – have bet their careers and corporate credibility on the proposition that somebody really clever can go into the reams of Big Data in […]","categories":["IT Services"],"tags":[],"author_name":"William Inmon","publish_date":"2017-02-09T04:09:52","publication_year":"2017","word_count":1106,"keywords":["big data","Go","programming_languages:R","AI","programming_languages:Go","RAG","ViT","disruption","R"],"extracted_tech_keywords":["AI","RAG","R","Go","big data","ViT","disruption","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/big-data-disappointment-till-now\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":42271,"title":"AI Is The &#8216;Holy Grail&#8217; For Enhanced Sports Viewing Experience: Study","content":"Despite the result of the India vs New Zealand match at the ICC World Cup semi-final, millennials from across the world are still willing to go that extra mile when it goes to watching sports. According to a new report by NTT, a noted tech and managed services company, the sporting organisations need put in extra efforts for their fans, as millennials are still hungry for better, more inclusive experiences. The report, published earlier this week, says that only half of the viewers said their current data experience regarding sporting events was adequate. This shortfall exposes the need for the right technology infrastructure and solutions to deliver the sophisticated experiences sporting fans expect. The report goes on to suggest that artificial intelligence and machine learning will be the answer. As the competition for fan engagement increases across major global sporting events, artificial intelligence (AI) and machine learning offer new ways to deliver more sophisticated and connected data experiences. The research of around 3,700 sports fans from across the globe shows AI and ML are extremely effective ways to create more engaging, data-rich experiences. It found: 54% of people aged 18-34 believe AI is capable of successfully predicting the results of a sporting event 52% of the respondents said accurate predictions make a sporting event more engaging Only 26% of people across all age groups are aware of AI and ML actually being used at sporting events, exposing a huge opportunity to create greater engagement Ruth Rowan, Global Chief Marketing Officer at NTT said in a statement, “There’s a real hunger in the sports industry for a more futuristic viewing experience for fans… Whether that’s through live analytics and data enhancements, AI-powered experiences, or connected stadiums – it’s clear ICT infrastructure, the cloud, and mobile services have a critical role to play as the sports industry evolves to meet the growing demands of digitally savvy supporters.” Rowan added that based on the above data, NTT had created an innovation-rich experience for the upcoming Tour de France. He said that the company had the following AI and ML-powered products for the same: Le Buzz: An ML model which analyses the movements within the peloton to predict potential key moments, such as the increased likelihood of a crash, a split in the peloton or a change in race dynamics Live stage favourites: ML-based stage favourite predictions to update live throughout the stage, based on the events occurring within the race “Catch the Break” predictor: A predictor to account for the different race strategies in play at different points in a stage, through the creation of individual ML models for every 10km section of the race Live data insights and predictions: A social media tool","excerpt":"Despite the result of the India vs New Zealand match at the ICC World Cup semi-final, millennials from across the world are still willing to go that extra mile when it goes to watching sports. According to a new report by NTT, a noted tech and managed services company, the sporting organisations need put in […]","categories":["AI News"],"tags":["cricket","ML"],"author_name":"Prajakta Hebbar","publish_date":"2019-07-11T16:37:24","publication_year":"2019","word_count":451,"keywords":["Go","artificial intelligence","machine learning","AI","innovation","ML","Git","cricket","analytics","GAN","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","analytics","R","Go","Git","GAN","innovation"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/ai-is-the-holy-grail-for-enhanced-sports-viewing-experience-study\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10001936,"title":"Is A Full-Blown Cryptocurrency Ban Taking Shape In India? Industry Insider Speaks Up","content":"In the latest development regarding the regulation of cryptocurrencies in India, the Indian government has reportedly begun considering an outright ban on the technology and its use. This comes after there has been a constant back-and-forth between players in the Indian cryptocurrency market and regulators. While this has not only created a lot of friction in the space, it has also caused a general downturn in the market; a problem compounded by a bear market in the larger scale. A Pseudo-Ban Emerges India has had a history with cryptocurrencies, with crypto first coming up in the 2018 Budget speech by the then Finance Minister Arun Jaitley. The government declared that cryptocurrencies were not legal tender, which meant that they could not be used as currency in the country. The government seems to have adopted its attitude towards cryptocurrencies owing to their purported use during the demonetisation era to launder money. Moreover, they have also stated on multiple occasions that they are either ponzi schemes offered by companies that promise large returns, or financial instruments used for illicit activities. Over time, it became clear that the government held no place for the new asset class. The primary financial regulatory party of the country, the Reserve Bank of India, even released a circular that cautioned investors against being involved with digital assets or cryptocurrencies. Post this negative attitude towards the technology, regulators moved towards a more heavy-handed approach, with a pseudo-ban of sorts. The RBI stated that all banks and financial institutions that fall under their purview were banned from providing financial services to companies and individuals involved in dealing with virtual currencies. This was the beginning of the stranglehold on the Indian cryptocurrency market, with general market volume and liquidity showing the effect that the ban had. Blockchain Players Fight Back Many players in the blockchain space in India began fighting against the ruling by RBI, calling it unconstitutional and a violation of freedom. However, the case has not reached a conclusion on account of the Supreme Court dragging its feet. The last significant development in the case took place on March 29th, which was the date the SC had declared after multiple reschedulings. However, the hearing was adjourned almost immediately, owing to a request from the side of the Government’s Counsel. The hearing has been postponed to the last week of April. However, the damage from the lack of legal clarity seemed to already be done, as homegrown exchanges such as Zebpay and Coindelta had shut down. An exchange platform known as Unocoin aimed to bring liquidity into the market by introducing a Bitcoin ATM, only to be promptly shut down and the founder arrested. The negative attitude towards cryptocurrencies led to a stagnation of the market and a general disinterest in one of the fastest growing fields. The Ban On The Horizon Recently, unconfirmed sources told a prominent publication that the Indian government has kicked off consultations on a draft bill to completely ban cryptocurrencies. The draft has already begun making the rounds to various concerned departments, said sources. This includes the Department of Economic Affairs, the Investor Education and Protection Fund, and the Central Board of Direct Taxes, among others. Reportedly, the bill described a ban on “sale, purchase and issuance of all types of cryptocurrency”. The rumblings of a ban on the technology were beginning to be heard late last year, as sources spoke of a meeting that was conducted with Arun Jaitley. Allegedly the meeting spoke about the “issues and challenges” of cryptocurrencies. Over the course of this meeting, a ban was reportedly discussed, as a part of the “necessary steps” taken to stop the misuse of assets for illicit purposes. Reportedly, the considered ban will be enforced under the Prevention of Money Laundering Act. However, the government seems to have a misinformed view of the technology, according to statements made by authorities. Allegedly, feedback offered by the ministry of corporate affairs has stated that the sale, purchase and issuance of currencies such as Bitcoin and Ethereum are by “individuals and companies on false inducements of massive returns”. On the contrary, Bitcoin and Ethereum are both decentralised networks with a distributed method of issuance and no intrinsic connection to fiat money. Anurag Agarwal, the joint secretary in the ministry of corporate affairs also stated that these were ‘ponzi schemes’ and that carried a lot of risk to investors. He stated, “There is a real and heightened risk of an investment bubble of the type seen in ponzi schemes that can result in a sudden and prolonged crash,” These are inaccuracies in the report that show that there was little to no preliminary research conducted before making the decision. An Insider View To gain a better understanding of the situation, Curious Dose reached out to Nischal Shetty, the CEO of  WazirX, India’s biggest cryptocurrency exchange by volume. He stated, “If you look at it, the IEPF is not a part of the [Subhash Chandra] Garg Committee. Until the Garg Committee gives any statement, I don’t think we should really be worried or react to any other statement from any other government body. ” “[IEPF] is not the government body which is going to act on the cryptocurrency regulations. I think we should wait and watch. The IEPF has to do with the security market, not with cryptocurrencies. First of all the Garg Committee has to come up with a report before a bill can even be discussed. We should take it with a pinch of salt.” Shetty also expressed positive sentiment towards the future of cryptocurrency regulation in the country, stating, “There is also a court case which is going to come up in July, until now the government hasn’t given anything to the court. I don’t think there is a concrete bill-related discussion. In October, there was a report that said the government was preparing for a ban. If there was any truth to that, the Union of India, the government would have said something to the Supreme Court.” Moreover, there are also international events that bode positively regarding the state of cryptocurrencies in India. Shetty stated, “On 28th of June this year, there is going to be a G20 summit, where one of the most important agendas is going to be cryptocurrency-related discussions. India will be a part of that. This is in Japan, [which is] very pro-crypto. It’s sure that India will not take any policy decision without taking the global regulatory atmosphere into consideration, which is going to be very positive.”","excerpt":"In the latest development regarding the regulation of cryptocurrencies in India, the Indian government has reportedly begun considering an outright ban on the technology and its use. This comes after there has been a constant back-and-forth between players in the Indian cryptocurrency market and regulators. While this has not only created a lot of friction […]","categories":["AI Features"],"tags":["ban","Bitcoin","cryptocurrencies","Cryptocurrency","regulation"],"author_name":"Anirudh VK","publish_date":"2019-04-29T18:12:53","publication_year":"2019","word_count":1088,"keywords":["Go","Bitcoin","ELT","programming_languages:R","Cryptocurrency","AI","cryptocurrencies","Git","RAG","Aim","regulation","ViT","GAN","ban","R"],"extracted_tech_keywords":["AI","Aim","RAG","R","Go","Git","ELT","GAN","ViT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/is-a-full-blown-cryptocurrency-ban-taking-shape-in-india-industry-insider-speaks-up\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10162480,"title":"OpenAI’s Sam Altman to Visit India, A Country He ‘Loves’","content":"OpenAI’s CEO Sam Altman is set to visit India on February 05, Reuters quoted, citing sources. He is likely to meet Indian government officials in Delhi. In an exclusive interview with AIM in November last year, Pragya Misra, public policy and partnerships lead at OpenAI and the only employee of the company in India, said, “He [Altman] just loves India.” Misra also revealed that every time she spoke to Altman about the use cases India is building, he asked how the headquarters could help India to scale that using OpenAI’s technology. “Compared to many of the solutions we see out of other countries from the developer ecosystem, the stuff that is coming out of India is just incredible,” she said, adding that the company has plans to expand the team in India. Altman’s visit to India comes at a time when the company is battling legal issues with media publishers in the country. Last year, ANI sued OpenAI for infringing copyright and alleged that ChatGPT was capable of reproducing its content verbatim. Recently, more digital news outlets, such as the Indian Express, Hindustan Times, NDTV, and the Digital News Publishers Association (DNPA, which represents 20 media companies), joined the lawsuit. However, OpenAI opposed the inclusion of DNPA in the lawsuit and argued it doesn’t have any stake in the matter. Two years ago, Altman was in New Delhi and also met the prime minister, Narendra Modi. Back then, Altman made a controversial comment saying it was totally hopeless (for India) to compete with OpenAI in building foundation models. It irked a large number of people in the country. Altman was responding to former VP of Google India, Rajan Anandan, who asked Altman about how India could build ‘substantial’ foundational models with a budget of 10 million versus 100 million. In stark contrast, China’s DeepSeek has trained an AI model (DeepSeek-V3) with just $5.5 million, and their latest model, the DeepSeek-R1 is better, if not on par with every other competitor. Moreover, these models are available on MIT’s open-source license for free. Aravind Srinivas, CEO of Perplexity, urged India to build an AI model of their own instead of focusing on use cases. He even offered financial support for a potential team of engineers with ambitions to build a DeepSeek-like model in the country. That said, India’s IT Minister Ashwini Vaishnaw announced at the Utkarsh Odisha Conclave that India will have their own large language model ‘soon’. The initiative will be driven by the IndiaAI Compute Facility, which has acquired 18,000 GPUs to support the creation of an LLM tailored to the country’s needs.","excerpt":"Altman is scheduled to visit India on February 05 and is likely to meet government officials in the country.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","OpenAI"],"author_name":"Supreeth Koundinya","publish_date":"2025-01-29T22:20:18","publication_year":"2025","word_count":433,"keywords":["Go","ChatGPT","OpenAI","AI","AWS","Git","RAG","Aim","foundation models","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","foundation models","ChatGPT","OpenAI","Aim","RAG","AWS","R","Go","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/openais-sam-altman-to-visit-india-a-country-he-loves\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":3504,"title":"Fractal Analytics named “Cool Vendor in Analytics” by Gartner","content":"San Mateo, California, May 13, 2013: Fractal Analytics, the most global provider of analytics, announced that it has been selected as one of the top five in the “Cool Vendors in Analytics 2013″1 report recently published by Gartner. ‘Cool Vendor’ is an annual report that identifies the new cool vendors in key technology areas according to Gartner. The report mentions, “By delivering various combinations of domain expertise and higher levels of automation, this year’s vendors improve the analytic accuracy, adoption and productivity of organizations.” “We are happy to be recognized as a ‘Cool Vendor’ of analytics by Gartner. This reinforces the excitement our Fortune 500 clients have shown in partnering with us to solve their Big Data and advanced analytics challenges”, said Srikanth Velamakanni, co-founder & CEO of Fractal Analytics. “Companies must supplement their traditional ways of understanding consumer behavior with new ways relevant for the age of Big Data, such as Customer Genomics™ for machine learning based personalized marketing.” Fractal Analytics’ Customer Genomics™ solution helps marketers learn complex customer behavior at an individual level. The proprietary pattern recognition and machine learning algorithms underlying Customer Genomics learn from every transaction and customer interaction including from social media helping marketers build a dynamic 360 degree view of the customer across attitudinal and behavioral dimensions. Customer Genomics has enabled a major US retailer in doubling coupon redemption and driving significant incremental store traffic. About Fractal Analytics Fractal Analytics believe analytics is critical to develop a deep understanding of consumers and earn customer loyalty, and make better data-informed decisions. Leading global companies partner with Fractal Analytics to build breakthrough analytics solutions, set up analytical centers of excellence, and institutionalize data-driven decisioning. We help businesses: (a) Understand, predict and shape consumer behavior through advanced analytics; (b) Improve effectiveness of marketing, pricing and supply chain management; and (c) Harmonize data, tell visual stories and forecast business performance. We are the most global analytics provider in the world serving Fortune 500 companies in CPG, financial services, insurance, retail, technology and life sciences in over 150 countries. We have offices in San Mateo, New Jersey, London, Singapore, Mumbai, Gurgaon and Dubai.","excerpt":"San Mateo, California, May 13, 2013: Fractal Analytics, the most global provider of analytics, announced that it has been selected as one of the top five in the “Cool Vendors in Analytics 2013″1 report recently published by Gartner. ‘Cool Vendor’ is an annual report that identifies the new cool vendors in key technology areas according […]","categories":["AI News"],"tags":["Fractal Analytics"],"author_name":"AIM Media House","publish_date":"2013-05-15T06:17:38","publication_year":"2013","word_count":354,"keywords":["Fractal Analytics","Go","big data","machine learning","AI","data-driven","automation","ViT","analytics","GAN","R"],"extracted_tech_keywords":["AI","machine learning","analytics","R","Go","big data","GAN","ViT","automation","data-driven"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/fractal-analytics-named-cool-vendor-in-analytics-by-gartner\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10062723,"title":"The story of two ex-Uber engineers who built a SaaS platform to boost developer productivity","content":"Founded by former Uber engineers Naomi Chopra and Haritabh Singh, Hatica was incubated at the Accel Founderstack Program in 2020. A SaaS-based engineering platform, Hatica, was formed with the intent to measure the productivity of developers and help them with better time management since the pandemic forced many businesses to shift to remote working. In January this year, Hatica raised USD 900,000 in its pre-seed funding round led by Kae Capital, followed by Titan Capital, iSeed Ventures and angel investor GBS Bindra. Chopra stated then that the company’s system was still in the beta stage and is expected to be released in the market by the first quarter of this year. We spoke to him about what Hatica’s tech stack looks like, and the lessons learnt along the way. AIM: What is the tech stack that Hatica is using? What are the details of its stack share? Naomi Chopra: Hatica is built using Typescript and Golang mostly. Its front-end is built on React, while the backend is a set of microservices in Node and Golang running on Kubernetes along with Temporal for orchestration and Hasura as the data access layer. Its data stack is built on Postgres, Clickhouse and Elasticsearch. AIM: How did you decide your stack share? What are the factors that you considered in the decision-making? Naomi Chopra: The factors that went into deciding our tech stack were a combination of the founders’ expertise and the availability of talent, along with the most important factor of all, which was essentially how suited the stack was for an I\/O and data-intensive system like ours. Our factors were skewed towards our expertise when we were just starting out, but as we grew, our decision revolved around choosing the best stack for the solution while ensuring access to talent and support by the community. AIM: Has Hatica been using the same tech stack since 2019? If not, then what were the reasons for migrating to a different tech stack? Naomi Chopra: No, our stack and architecture looked very different from what it is today. We started with a simpler Javascript stack composed of microservices running as containers on AWS ECS service, served by GraphQL APIs powered by Hasura on a Postgres DB. That helped us with building prototypes quickly. As we gained traction and started growing rapidly, we migrated to using Golang for some services along with a more scalable architecture powered by Temporal for microservices orchestration and Clickhouse for serving analytical queries. All of this is now hosted on Kubernetes clusters. AIM: How did Hatica go about the migration process? Was it costly to migrate? Naomi Chopra: Migration was costly in terms of both time and effort, but only in the short run. The long term benefits of a more scalable and robust architecture have served us well as we grow rapidly in customer base. AIM: What were the hiccups that you or your users faced during the migration? How did you smoothen the tech issues during this time? Naomi Chopra: One hiccup we faced during migration was the fact that our authorisation layer, which decided what can be accessed by whom, became too complex since we had to support two different sets of APIs and data access layers, both old and new, till we migrated fully. We solved that by first completing our unit testing suite for the data access layer so as to prevent access permissions issues from going into production. AIM: How do you think Hatica’s tech journey has evolved since its inception? Naomi Chopra: Hatica started out as a set of dashboards for engineering managers with little to no customisation available to them. As we gained traction, we needed to provide configurability that would help our product support different types of engineering workflows. This was the first milestone when we started rebuilding many of our microservices. But the larger change came as we scaled up in traffic in a very short period of time as we started onboarding large teams. This led us to evolve our overall data stack, which has been a considerable effort. Fortunately, we adopted some key best practices from day one, which helped us migrate systems component by component with very little downtime. AIM: What are some of the most valuable lessons you learnt that you would like to share with other budding startups? Naomi Chopra: Don’t choose a stack to solve for the next million users; pick something you can make a prototype with quickly. Play to your strengths, even though there could be other tools more appropriate for the problem; for example, pick a programming language that you are good at and just works for your solution, rather than picking a cool new language. Pick a stack for which you can find talent easily. There is, anyway, a lot of competition for hiring engineering talent, don’t make it harder by picking a shiny new tool that is hard to learn, at least not when you are just starting out. Don’t reinvent the wheel. Try to look for a robust open-source solution when implementing a component or a service. If it doesn’t solve your problem wholly, try to extend it instead of implementing a solution from scratch. There are some basic, foundational concepts and best practices you should surely start with, like using unique IDs like UUIDs for most resources. Such best practices are easy to adopt from the beginning and hard to migrate to. Document your systems by adopting practices like writing RFCs. They not only drive collaboration and effective reviews resulting in robust decision making, but also automatically result in good documentation, which is essential not just to onboarding your new teammates but also for founders and early engineers to recall specs and decisions. Use infra-as-code, like using Terraform. It’s simpler than it looks and gets you quickly up and running by utilising open-source packages.","excerpt":"Hatica was built using TypeScript and Golang.","categories":["AI Features"],"tags":[],"author_name":"Poulomi Chatterjee","publish_date":"2022-03-15T13:00:00","publication_year":"2022","word_count":976,"keywords":["Elasticsearch","Go","AWS","AI","TypeScript","microservices","Aim","JavaScript","R","kubernetes"],"extracted_tech_keywords":["AI","Aim","AWS","kubernetes","microservices","Elasticsearch","R","JavaScript","TypeScript","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/the-story-of-two-ex-uber-engineers-who-built-a-saas-platform-to-boost-developer-productivity\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10162490,"title":"Anthropic&#8217;s Dario Amodei Questions DeepSeek&#8217;s Cost Claims, Bats for Stronger AI Export Controls","content":"In his recent blog, ‘On DeepSeek and Export Controls,’ Anthropic CEO Dario Amodei addressed claims surrounding DeepSeek’s AI model and its implications for US export restrictions on chips to China. He challenged the notion that DeepSeek has achieved a significant cost advantage over US AI companies. Comparing the outputs, Amodei noted that the claims regarding DeepSeek’s achievement for $6 million, which cost American AI companies billions of dollars, are exaggerated. He explained that Anthropic’s Claude 3.5 Sonnet was trained for a few $10 million and was developed 9-12 months before DeepSeek’s model—disputing exaggerated claims about DeepSeek’s cost efficiency. Amodei noted that DeepSeek’s model matches US models from 7-10 months earlier and was trained at a lower cost, though “not anywhere near the ratios people have suggested.” The cost reduction, he explained, aligns with the industry’s expected 4x annual decrease in AI training costs rather than being a disruptive breakthrough. Amodei also emphasised the strategic importance of export controls in shaping the future of AI. He outlined two possible scenarios for 2026-2027: a bipolar world where both the US and China develop advanced AI, and a unipolar world where only the US and its allies take the lead AI. Advocating for the latter, Amodei said, “If we can close [export control loopholes] fast enough, we may be able to prevent China from getting millions of chips, increasing the likelihood of a unipolar world with the US ahead.” This suggests that he supports the US’s AI monopoly rather than mere leadership. Finally, Amodei addressed concerns over the effectiveness of current export controls. He argued that DeepSeek’s progress does not indicate a failure of these measures. Instead, DeepSeek’s chip inventory—comprising a mix of banned, previously allowed, and still-permitted chips—demonstrates that export restrictions are evolving and closing loopholes.","excerpt":"Anthropic’s Claude 3.5 Sonnet was developed 9-12 months before DeepSeek’s model and cost a few $10 million to train.","categories":["AI News"],"tags":["Anthropic"],"author_name":"Siddharth Jindal","publish_date":"2025-01-30T09:00:18","publication_year":"2025","word_count":294,"keywords":["Anthropic","API","TPU","programming_languages:R","AI","IPO","llm_models:Claude","Aim","Claude 3.5","R"],"extracted_tech_keywords":["AI","Claude 3.5","Anthropic","Aim","TPU","R","API","IPO","llm_models:Claude","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/anthropics-dario-amodei-questions-deepseeks-cost-claims-bats-for-stronger-ai-export-controls\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10118147,"title":"Data Centres’ Lean Towards Nuclear-Powered Future to Combat Energy Needs","content":"According to recent industry reports, the global electricity consumption of data centres is projected to reach a staggering 848 terawatt-hours (TWh) by 2030, nearly doubling from the estimated 460 TWh consumed in 2022. To put these figures into perspective, India, the world’s second-most populous country, consumed 1,443 TWh of electricity in 2021. The projected 2030 data centre consumption would equal more than half of India’s electricity usage. Similarly, Ireland generated 34.5 TWh of electricity in 2022. So, as data centre operators confront energy challenges like scarcity of available energy for powering data centres from the grid and an imperative to minimise carbon emissions—they’re now moving towards nuclear energy! Equinix, the world’s largest data centre colocation provider, recently signed a pre-agreement with Oklo, a firm which produces small modular reactors (SMR), to purchase nuclear energy for a $25 million prepayment. A Small Modular Reactor (SMR) is generally defined as an advanced reactor that produces up to 300 MW(e) per module. As the energy demands of data centres, which power every critical digital infrastructure and technologies like generative AI, 5G, Io, etc., these reactors or SMRs are a critical move towards clean and sustainable data centres powered by safe nuclear energy. “A normal data centre needs 32 megawatts of power flowing into the building. For an AI data centre, it’s 80 megawatts,” says Chris Sharp, CTO at Digital Realty, a US data centre giant. Oklo’s reactors powered by nuclear fission energy stand as a viable option for data centres as they can generate 15MW each and can function for a minimum of 10 years before requiring refuelling. Equinix intends to purchase power from Oklo’s upcoming SMR installations to fuel its US data centres. It will possess the first option for 36 months to acquire between 100MW and 500MW of cumulative capacity from specific Oklo powerhouses. Additionally, smaller microreactors, with capacities ranging from 1 to 20 MW, are being developed specifically to power data centres and industrial sites. Startups like Oklo aim to deploy factory-built microreactors by 2028 to meet these energy demands. Besides Oklo, several other US-based firms, including NuScale Power, Kairos Power, and X-energy, are actively developing small modular reactors (SMRs). Additionally, UK-based Rolls-Royce is also pursuing SMR technology. Challenges in Adoption While public opinion is nearly evenly split between support and opposition, and safety remains a primary concern—many SMR manufacturers are trying to combat the issue. For instance, NuScale Power, a leading developer of SMRs, has made safety a top priority in its design, according to CEO and co-founder José Reyes. Reyes, who spent nearly a decade as a research engineer in the Reactor Safety Division of the U.S. Nuclear Regulatory Commission, emphasises that NuScale’s SMRs are designed to safely shut down without operator intervention in the event of a worst-case scenario. To address the concerns arising from nuclear disasters in Fukushima, Chornobyl, and Three Mile Island, SMRs’ smaller size, simpler design, and inherent safety features make them more reliant and risk-averse. SMRs also generate less nuclear waste, as they require refuelling every 3-7 years or even up to 40 years, compared to 1-2 years for conventional nuclear power plants. Additionally, SMRs can be installed closer to energy consumers and data centres, overcoming grid constraints. Jay Dietrich from the Uptime Institute has also highlighted that SMRs provide reliable, carbon-free electricity, which can complement intermittent renewable energy sources like wind and solar power to help data centres become more sustainable. Despite the potential benefits, the SMR market faces several challenges. However, due to hurdles such as manufacturing ramp-up, design permits, site approvals, and grid connection permissions, SMR-powered data centres may become a reality by the late 2020s or early 2030s. The emerging SMR market recently faced a setback when NuScale’s plan to launch a six-reactor, 462 MW project with Utah Associated Municipal Power Systems collapsed in early November. Several towns withdrew from the project after costs rose, highlighting the challenges faced by the nascent industry. Initial construction costs for SMRs may also be high, but economies of scale are expected to lower costs over time. Hans Lohse, a representative from Idaho National Laboratory (INL), believes that economies of scale will play a significant role in reducing SMR costs. “I don’t think anyone expects the first couple of builds to be the cheapest, but when you get the supply chain going, the cost curve will go down, and you will get economies of scale,” Lohse said. Another concern surrounding SMRs is nuclear waste production, albeit in smaller quantities compared to traditional large-scale nuclear plants. While China’s Linglong One reactor became the first small reactor to receive safety approval from the IAEA, the success of SMRs will depend on continued public engagement, regulatory support, and a demonstrated track record of safe operation in rural and urban locations. Big Tech Also Turns to SMR Tech giants like Microsoft and Amazon have also shown significant interest in SMRs, nuclear technology and nuclear power purchase agreements to source a portion of their data centres’ electricity needs from nuclear plants for growing energy needs by data centres. Microsoft is exploring nuclear power deals with companies like Helion and Ontario Power Generation for its operations in Canada. The tech giant has reiterated its seriousness in the matter by employing a director of nuclear development acceleration and a director of nuclear technologies to spearhead its exploration. Concurrently, Amazon has already acquired a massive 960 MW data centre campus in Pennsylvania, fueled by the Susquehanna nuclear power plant, showcasing the growing interest of hyperscalers in nuclear energy for data centre operations. While not using SMRs directly, this demonstrates the interest of hyperscalers in nuclear energy for powering data centres. The first SMR reactors will likely be installed on existing nuclear sites, where infrastructure and permitting are already in place. Rolls-Royce is starting with a decommissioned nuclear plant in Trawsfynydd, Wales, which could hold two 470MW systems. In Canada, Ontario Power Generation (OPG) is building up to four new SMRs in Darlington, Ontario, alongside its existing CANDU reactors. OPG has signed a Power Purchase Agreement with Microsoft, which may include nuclear power from these SMRs if they are operational in time.","excerpt":"As data centres’ consumption are set to reach 848 TWh by 2030, nearly doubling from 460 TWh, data centres look towards SMRs.","categories":["AI Features"],"tags":["Data Center"],"author_name":"Shyam Nandan Upadhyay","publish_date":"2024-04-12T17:42:13","publication_year":"2024","word_count":1020,"keywords":["Go","Data Center","AWS","AI","cloud_platforms:AWS","programming_languages:R","Git","Aim","generative AI","R","startup"],"extracted_tech_keywords":["AI","generative AI","Aim","AWS","R","Go","Git","startup","cloud_platforms:AWS","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/data-centres-lean-towards-nuclear-powered-future-to-combat-energy-needs\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10049250,"title":"Indian Govt To Provide Farm Data To Tech Giants: What To Expect?","content":"Prime Minister Narendra Modi’s administration has signed an initial agreement with three US-based tech giants to modernise its agricultural sector. Amazon, Microsoft and Cisco Systems are among the tech companies who will be harnessing data and statistics available on India’s agricultural ecosystem since 2014. The Government also revealed that Jio Platforms and ITC Ltd. are among the Indian players signing up for the programme. This move aims to drive transformation in the traditional agricultural sector and ensure food security in a country with an ever-growing population. PM Modi believes that private players will help farmers boost yields with tools and applications built using information related to crop output, land holdings and soil quality. Why now? Between April 2020 and January 2021, India’s agricultural commodities export stood at $32.12 billion. However, closer home, the Food and Agricultural Organisation (FAO) estimates that 40 per cent of the food produced in India is wasted annually. The FAO further suggests that this wastage is due to inefficient supply chains and fragmented food systems. With this long due project, PM Modi aims to boost rural incomes, cut imports, and reduce food wastage. The Government is also planning to achieve this with better infrastructure and ultimately directly competing with food exporters — the EU, the US and Brazil. This collaboration with tech giants directs towards the increased deployment of AI and ML models and networks in India when e-commerce is working towards cracking the grocery segment. Players including homegrown company Reliance and Jeff Bezos’ Amazon have been trying to catch hold of a stable supply of farm produce to make the most of the groceries market. Reportedly, the Indian Government has collected data on more than 50 million farmers out of the 120 million land-holding growers in the country. In addition, local companies including ESRI India Technologies, Ninjacart, Star Agribazaar Technology and Patanjali Organic Research Institute have also signed the agreement. Roadblocks Owing to the ongoing conflict around the new agricultural laws between Indian farmers and the Government, it might be difficult for officials to convince farmers to share all their data or even adopt technology in the agricultural field. In an interview with Bloomberg, which first reported about the partnership between the Government and the tech giants, Punjab-based farmer Sukhwinder Singh Sabhra said that the availability of this data with the Government would further make Indian farmers vulnerable. This is because the Government would know where the produce was not good and try buying crops at lower prices from these places, selling them off at exorbitant prices elsewhere. Sukhwinder, who is currently protesting against the new farm laws, also added that the consumers would be at a loss more than the farmers. The move also draws criticism from experts who believe that the Government prioritising the growth and development of private sectors could hamper small farmers. Close to 58 per cent of the Indian population depends on agriculture as the primary source of livelihood. This makes India the sixth-largest food and grocery market globally, and retail contributes to 70 per cent of the sales. With private players getting hold of farm data and the new farm bills coming into action, small Indian farmers will be forced to sell crops at a predetermined price, often at a loss. Thus, further hampering their already miserable financial conditions. The success of this move, is thus, still unpredictable. The Government has been long trying to build tech infrastructure to forecast and improve crop yield. However, unprecedented events of drought, floods and the most recent COVID-19 pandemic have worked against the deployment of digital solutions in the agricultural sector. Besides the Indian Government, agritech startups and players in the country — including the likes of DeHaat, Bijak and CropIn Technologies — have all been trying to develop technologies to improve crop yield projections and help farmers make informed choices regarding the type of crops of grow, improve the produce quality and adopting required policies. Additionally, space tech startup Pixxel is using technology to monitor crop health, detect variations and improve yield. The scale and success of this move by the Government are yet to be seen.","excerpt":"The Indian Government has signed an initial agreement with tech giants in India and abroad to share the crop data towards the digitisation of the agricultural sector.","categories":["IT Services"],"tags":["AgriTech","Amazon","CropIn","Microsoft","Narendra Modi","PM Modi"],"author_name":"Debolina Biswas","publish_date":"2021-09-22T15:00:00","publication_year":"2021","word_count":686,"keywords":["Go","TPU","AWS","AI","ML","Amazon","Git","AgriTech","RAG","Aim","CropIn","Narendra Modi","GAN","R","Microsoft","PM Modi"],"extracted_tech_keywords":["AI","ML","Aim","RAG","AWS","TPU","R","Go","Git","GAN"],"url":"https:\/\/analyticsindiamag.com\/it-services\/indian-govt-to-provide-farm-data-to-tech-giants-what-to-expect\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10122226,"title":"Bhashini Launches ‘Be our Sahayogi’ for Multilingual AI Innovation Focused on Voice","content":"Bhashini in collaboration with Nasscom has launched the “Be our Sahayogi” program on National Technology Day to crowdsource multilingual AI problem statements. Organisations are invited to submit their ideas for “Reimagining the User Journey” as a “Multilingual User Journey.” Amitabh Nag, CEO of Bhashini, said, “‘Voice first’ is the way to actually make a difference and bridge the Digital and Literacy divide besides Transcending the language Barrier.” Bhashini launched a crowdsourcing initiative to collect voice and text data in multiple Indian languages called Bhasha Daan. “It’s performing well, but not meeting our initial expectations. We plan to run a campaign to build this up further,” Nag told AIM, when asked about its status. The partnership between Nasscom and BHASHINI aims to develop, deploy, and enhance solutions to transform society. The collaboration seeks to revolutionise the Bhartiya Bhashaien ecosystem through innovative technology. “Bhashini will definitely change lives because people will be more collaborative, cooperative, and innovative, without the burden of trying to learn more languages,” added Nag. This initiative aims to foster innovation and creativity, driving the development of new technologies and solutions for the Bhartiya Bhashaien community. By prioritising voice as a medium, the partnership will promote innovation, collaboration, and empowerment. Joint hackathons and innovation challenges will provide a platform for entrepreneurs and innovators to showcase their talents. In an interview with AIM, Ankit Bose, head of AI at Nasscom, said, “Bhashini is a very good start and the government has done a phenomenal job at this. It is one of the most important tasks, as we can increase the whole country’s productivity, including people who don’t speak English,” talking about collecting Indic data.","excerpt":"By prioritising voice as a medium, the partnership will promote innovation, collaboration, and empowerment. Joint hackathons and innovation challenges will provide a platform for entrepreneurs and innovators to showcase their talents.","categories":["AI News"],"tags":["Bhashini"],"author_name":"Mohit Pandey","publish_date":"2024-06-02T13:07:09","publication_year":"2024","word_count":274,"keywords":["Go","programming_languages:R","AI","innovation","programming_languages:Go","Git","Aim","Bhashini","ViT","GAN","R"],"extracted_tech_keywords":["AI","Aim","R","Go","Git","GAN","ViT","innovation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/bhashini-launches-be-our-sahayogi-for-multilingual-ai-innovation-focused-on-voice\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10138055,"title":"Innovating with AI in Regulated Industries","content":"Artificial intelligence (AI) is particularly effective at improving efficiency and user experience in scaled, repetitive, data-rich processes—common in regulated industries like banking and insurance. Although these industries could benefit from AI, they have been slow to adopt it due to regulatory compliance and data privacy concerns. I was the CIO at a leading insurance company several years ago. We used AI to improve internal efficiency, customer experience, and data protection while building trust with customers and regulators. Here are the lessons I learned and my suggestions for adopting AI in a regulated industry. Why Now? While the pace of change has been increasing across industries, insurance feels like it’s just catching up. Generative AI is raising customer expectations, and regulators are preparing for the industry to adopt it. The time for “waiting and watching” ends now. Consider life and health underwriting. This core insurance process had stayed largely the same for decades, demanding hundreds of customer data points and days or weeks of lead time before an applicant got coverage. My former company and its reinsurer developed a new AI-supported customer purchase process. It assessed risk in real time by asking questions dynamically and could underwrite two-thirds of cases within 30 minutes. The customer walked away, covered and happy. We booked revenues faster while our underwriting colleagues expedited the remaining third of complex, human-required cases. We made an even bigger impact on claims. By using AI to reimagine the process, we achieved 60 percent straight-through processing. A customer could upload a hospital bill and receive payment within five minutes—a process that historically took weeks. We also reduced claims leakage of nonpayable items and improved fraud and misclaim detection. Getting AI Regulation-Ready AI can benefit the insurance industry, but its application needs to meet customer and regulatory expectations of fairness, ethics, transparency and accountability.1 Fairness: Teams train AI models with debiased data. Ethics: AI solutions reflect organizational values and codes of conduct. Transparency: Decisions made by AI models are explainable. Accountability: Leadership takes ownership of AI-generated decisions, ensuring fairness, ethics, and transparency. Meeting these expectations can be challenging. For example, “black box” models, like deep-learning algorithms that ingest many diverse inputs to generate answers, lack transparency. How can you harness AI’s power while meeting regulatory needs? Empowered Teams You need to reimagine processes to implement AI in insurance, so you need a team with diverse expertise empowered to transform customer experience. You should include insurance experts, data scientists, AI specialists, legal and compliance professionals, finance and risk specialists, user experience designers, salespeople, and agents. This team needs to question the status quo, work across silos, and test new approaches to create technically sound, compliant solutions and raise the bar for customer experience. Explainable Chain of Decisions To address black box concerns, break down decision-making into explainable steps that can be automated using rules or simple AI\/ML models (e.g., decision trees or random forests). In our claims automation, a chain of models extracted data from invoice scans and then classified diagnoses, treatment paths, policy eligibility, and coverage exceptions. We could explain each step in the chain and turn it on or off depending on prediction confidence levels. Strong Governance and Controls AI requires a comprehensive governance framework. This includes setting up an AI risk committee, adding AI experts to compliance teams, and creating systems to constantly monitor and validate AI models. You must regularly check for potential biases—AI models are only as good as the data they are trained on.2 Build Confidence in AI Step by Step Manage complexity and high stakes by taking a measured approach to AI adoption. This allows for careful testing, improvement, and buy-in from stakeholders at each stage. I recommend a simple process: Start small: Automate basic, repetitive tasks like extracting data from documents or initially sorting claims. These activities can be easily validated and yield quick results, building confidence in AI capabilities. Use Human-in-the-Loop: Instead of fully automating complex processes, use AI to support human decision-making. AI can provide recommendations and insights to claim assessors who make final decisions. Create feedback loops: Include a dashboard that tracks performance for each step and confidence limits on every decision. Team leaders can adjust AI recommendations to match human decisions and identify areas for improvement. Expand gradually: As confidence and capabilities grow, expand straight-through processing criteria while maintaining oversight so humans focus on the areas that need more training. Moving Forward As the insurance industry adopts AI, success will come to organizations that innovate while staying compliant. You can unlock AI’s potential while maintaining customer and regulator trust by building the right teams, choosing appropriate AI methods, implementing strong governance systems, and using innovative and secure cloud technologies. In this AI-driven era, success won’t come to those with the most advanced algorithms but to those who integrate AI into their operations while retaining the human touch and ethical standards central to the insurance industry. Links Principles to Promote Fairness, Ethics, Accountability and Transparency (FEAT) in the Use of Artificial Intelligence and Data Analytics in Singapore’s Financial Sector Your AI is Only as Good as Your Data by Tom Godden","excerpt":"Artificial intelligence (AI) is particularly effective at improving efficiency and user experience in scaled, repetitive, data-rich processes—common in regulated industries like banking and insurance. Although these industries could benefit from AI, they have been slow to adopt it due to regulatory compliance and data privacy concerns. I was the CIO at a leading insurance company […]","categories":["AI Highlights"],"tags":["AI (Artificial Intelligence)"],"author_name":"Arvind Mathur","publish_date":"2024-10-11T11:30:00","publication_year":"2024","word_count":850,"keywords":["Go","artificial intelligence","AI","ML","RAG","Aim","analytics","generative AI","Rust","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","ML","analytics","generative AI","Aim","RAG","R","Go","Rust"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/innovating-with-ai-in-regulated-industries\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":33171,"title":"How Is The Banking Industry In Malaysia Adopting Data Science – A Practitioner&#8217;s Perspective","content":"Ratul Paul heads the data science and AI team that launched multiple analytics products in last one year and a lot more currently in the pipeline. We spoke with Ratul on the state of AI and analytics in Malaysia. Ratul has more than 11 years of experience in technology and advanced analytics, machine learning and AI systems, primarily focused on consumer and SME lending. Ratul worked in leading credit bureau Transunion Cibil in India, where he worked more on automation and machine learning using the power of Big data. Earlier Ratul also worked in Global data science leader SAS in its Research and Development office in Pune, India. Ratul holds a B.Tech from National Institute of Technology, Calicut (India). Analytics India Magazine: How important is Data science & AI within Banking Systems in Malaysia? Ratul Paul: Malaysia has 15 million credit-active consumers and 2 million Small and Medium Enterprises (SME). That’s a very small market with tough competition among banks to get a greater wallet share. With bad rate significantly high in the subprime market and banks working on a very small profit margin to be fiercely competitive, risk management and optimizing growth strategy are tremendously challenging here. Data science and AI use cases are primarily focused on reducing cost and risk, and increasing sells and revenue. AIM: Can you elaborate on a specific use case of data science or AI that you worked on RP: In Malaysia, bank’s client data is strictly protected by a lot of privacy laws, at the same time Banks are very serious about their reputational risk. In such a scenario banks in Malaysia are not proactively open to implementing AI and replacing conventional decision systems and scorecards, because an  AI decisioning has a lot of black box layers and internal compliance team still not open to anything that they can’t interpret. In such a scenario, so far only a few AI use cases have shown potential in Malaysia. 1.Fraud detection – A lot of interest has been shown by banks to authenticate Identity Card and prevent fraud and a lot of players built deep learning AI models to authenticate IC. AI technology can authenticate multiple specific features in an Identity Card using optical character recognition(OCR), and that way significantly reduce fraud. Online loan application and decisioning – coupled with IC verification, mobile applications have been developed , that can check IC, match with applicants face and then move to the application, scoring and decisioning process, everything within three to five minutes, that way increasing customer experience and improving lag time significantly, as well reducing cost. 3.Auto alert for Consumers- This AI system notifies consumers whenever there is a loan or credit card applied using his ID, Approve limit changes, address or phone number changes in the banking system or when customer miss any payment. The consumer can validate if the changes are expected or fraudulent and take appropriate action accordingly. Natural language processing – In NLP front Banks are interested in building more advanced chatbots that can interact with the customer more intelligently and give a better customer experience. Banks are also interested to standardize quarterly financials those are submitted to banks by SMEs in a different format. This standardization has significant potential to reduce cost and lag time while improving customer experience. Alternate data – There is also significant interest among different players to merge data from different sources, like telecom, credit bureau and card provider data, together to profile a customer and understand him better with additional data points. That would help in up-sell and cross-sell strategy for banks with a more targeted campaign, Early collection strategy, And late collection for defaulted untraceable customers. when a customer defaults, additional data points can trace the exact whereabouts and profile of a customer and that way makes it easy for banks to collect the defaulted money. However, progress in this front is very slow due to compliance and regulatory policy and strict data privacy law. AIM: How do you see the analytics ecosystem flourishing especially in the south-east Asia region? RP: In Malaysia – there has been a huge interest among banks to embrace data science and AI. However, banks are more risk averse here than in other geographies and bound by strict compliance laws. So banks still primarily use licensed software (like SAS) and open source programs (like Python) are still not accepted and will not be accepted by banks in near future. Also, strict data privacy laws and the bank’s internal compliance makes it difficult or very time consuming for the analyst to get the data. While banks understand that the potential of AI and alternate data is huge in adding business value and profitability, but banks want to go slow while changing existing process due to reputational risk and internal compliance and legal policy. AIM: What are some of the challenges that the industry in Malaysia faces in terms of AI adoption. RP: All entities in Malaysia are bound by strict consumer data privacy law and usage and scope of data are meticulously defined. So getting data for building any AI model would be difficult and time-consuming because the request would go through multiple legal and compliance team scrutiny. Even an NDA (non-disclosure agreement) sign takes months in Malaysia due to different legal clauses. At the same time adopting new AI systems needs to get strict regulatory approval first with a testing period of minimum 3 to 6 months, next internal legal and compliance team approval. Because of the lengthy process of adoption of new technology, banks are usually very choosy while showing interest in next-generation AI systems. AIM: How can governments and citizen associations come together for a healthy discussion as well as the implementation of AI? RP: Government and regulator can come together to ease the compliance process, for good use of data. At the same time, effective and supportive regulation needs to be in place to make the best use of data collaboration between different entities. Since new data points are available now like never before, old regulation guidelines need to be regularly amended to keep pace with the changing scenario. Citizen association need to educate the common man that their data is protected and used for good purpose only, and it will eventually benefit the common man to get easy credit at need. AIM: Is AI talent an issue in Malaysia? If yes, how can we resolve this? RP: Malaysia might not have enough resource to work on advanced AI systems as of now, but many entities are working towards training aspirants, at the same time industry partnered with educational institutes to nurture talents and give right exposure even before joining the industry. Also, liberal immigration policy makes it easy for industry to hire foreign talent. AIM: What is the biggest trend in data science\/ AI that you look forward to in 2019? RP: Reducing cost and increasing profitability is the driving force for AI in the banking system. Fraud is a big concern in banking and stopping fraud transaction has been a pressing issue for banks in Asia. AI systems that can prevent fraud and at the same time reduce operational cost and generate greater revenue would be in demand in 2019 and coming years. Also, Chatbots will advance significantly to improve customer experience.","excerpt":"Ratul has more than 11 years of experience in technology and advanced analytics, machine learning and AI systems, primarily focused on consumer and SME lending. Ratul worked in leading credit bureau Transunion Cibil in India, where he worked more on automation and machine learning using the power of Big data. Earlier Ratul also worked in […]","categories":["Deep Tech"],"tags":[],"author_name":"Дарья","publish_date":"2019-01-09T11:53:14","publication_year":"2019","word_count":1218,"keywords":["data science","machine learning","AWS","AI","chatbots","NLP","Aim","deep learning","analytics","fraud detection"],"extracted_tech_keywords":["AI","machine learning","deep learning","NLP","data science","analytics","Aim","chatbots","fraud detection","AWS"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/how-is-the-banking-industry-in-malaysia-adopting-data-science-a-practitioners-perspective\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10090036,"title":"Meet the Tech Kins","content":"Elon Musk, Sam Altman, Sundar Pichai — these are among the few names in the tech industry that command a larger-than-life celebrity status. These titans have shaped the way we interact with technology today. Here, we bring to you the lesser-known information about these tech titans. Not many know that these tech biggies have immensely successful siblings who are busy making a mark in the industry and adding value in their own unique ways. From innovative startups to established ventures, the kin are making waves in the kingdom, proving that when it comes to innovation and leadership, excellence truly runs in the blood. Musk Bros Elon is no stranger to the tech world. Similarly, his brother Kimbal is a star in the food world. Kimbal Musk is a chef, restaurateur, and philanthropist, whose personal mission is an America where everyone has access to real food. He was named a Global Social Entrepreneur by the World Economic Forum. According to Kimbal, “He (Elon) is a savant when it comes to business, but his gift is not empathy”. Kimbal is on the board for Tesla and SpaceX and the co-founder of the Kitchen Restaurant Group, a collection of “community” restaurants located in Colorado, Chicago, and Indianapolis. Altman Bros Sam Altman, the co-founder of OpenAI and the new darling of Silicon Valley has been in the limelight for a while now but information on his personal life has largely been uncharted. His brother, Jack Altman, an author, is also the CEO & co-founder of Lattice AI, a performance management and engagement platform. Jack attended Princeton University with an undergraduate degree in economics and then started his career as an analyst at Gleacher & Co. A year later in 2012, Jack founded Hydrazine Capital, a venture capital firm with his brother Sam. The organisation focused on investing in early-stage technology companies and ceased operations in 2014 after making several investments and two exits. Pichai Bros In 2015, Chennai’s Sundar Pichai was appointed the next CEO of Alphabet, the group which owns Google. Pichai grew up in an Indian middle-class family and was mostly interested in books as a child. The Google leader has a younger brother named Srinivasan Pichai about whom not much information has been made public. Srinivasan started his career as a consultant in 2010. Today, he is a manager at DURR, a leading mechanical and plant engineering firm. Bengio Bros AI pioneer Yoshua Bengio is known for his work in deep learning, earning him the 2018 AM Turing Award ‘the Nobel Prize of Computing’. His younger brother Samy Bengio has been working in machine learning for almost the same amount of time away from Montreal in California. Currently, Samy is the senior director, AI and machine learning research at Apple Machine Learning. Previously, he worked at Google Brain but resigned after the Timnit Gebru episode. He has authored above 250 scientific papers in machine learning as per DBLP. He also worked on the first evidence that adversarial examples can exist in the real world. Wojcicki Sisters Not everything in the tech landscape is run by tech bros. Earlier this year, Susan Wojcicki stepped down from her role as CEO of YouTube after working for almost a decade. Currently, she serves in an advisory role to Google and Alphabet. Her younger sister Anne Wojcicki is the CEO of 23andMe, a direct-to-consumer DNA testing company founded in 2006. In 2013, Fast Company named Anne ‘The Most Daring CEO’. She is a co-founder and board member of the Breakthrough Prize. She was married to Google co-founder Sergrey Brin until they got divorced in 2015. Wolfram Bros English physicist and author Stephen Wolfram is best known for his contributions to AI. He created Mathematica, WolframAlpha and the Wolfram language. Stephen founded Wolfram research, along with his younger brother Conrad, in 1991. Conrad is a British technologist and entrepreneur famous for his work in IT and mathematical education reform. In 2010, Conrad founded computerbasedmath.org. A decade later, he released The Math(s) Fix: An Education Blueprint for the AI Age which exposes why educationally, mathematics is in crisis worldwide. In an exclusive interaction with AIM, Conrad jokingly said, “It’s often said about me that I am the updated, newer and younger version of Stephen. We have spent many years building intelligent computation with lots of automation so that more and more people can use the power of computation to make better decisions. At the end that’s what it is all about.” Over the course of more than four decades, the brothers have been responsible for many discoveries, inventions and innovations in science, technology and business. Schmidhuber Bros Schmidhuber is not an anonymous name in the AI community, but Juergen is not the only name for the family in the artificial intelligence game. His brother Christoph is a theoretical physicist-turned-finance guru. Before he made the field switch, Christoph’s research and thesis at the California Institute of Technology focused on string theory.For almost the past two decades, Christoph has been professionally pursuing asset management, risk management and team leadership.","excerpt":"When it comes to innovation and leadership, excellence truly runs in the blood","categories":["AI Trends"],"tags":["AI leaders","Elon Musk","Sam Altman","Sundar Pichai","Yoshua Bengio"],"author_name":"Tasmia Ansari","publish_date":"2023-03-27T12:00:00","publication_year":"2023","word_count":840,"keywords":["Go","API","Sam Altman","artificial intelligence","machine learning","AI","OpenAI","Elon Musk","Yoshua Bengio","deep learning","Aim","GAN","AI leaders","R","Sundar Pichai"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","deep learning","OpenAI","Aim","R","Go","API","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/meet-the-sibling-tech-titans-and-their-kinfolk\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":24640,"title":"Unlocking The True Potential Of Microsoft Azure: How Is It Taking AI And ML To A New Level","content":"Microsoft has come a long way in cloud technology with its standalone platform Azure. The advancement is so progressive that Azure has expanded its capabilities to machine learning and artificial intelligence. The tech giant is consistently and collectively focussing on the current industry needs with respect to ML. With Azure, it has unlocked the potential for AI applications on a smaller scale. In this article, we will discuss Microsoft Azure’s latest ML offerings — Hardware Accelerated Models and Cosmos DB — which are making headway in the field. These features were recently discussed at Build, Microsoft’s technical conference for developers, held in Seattle,US. The Dawn Of Deep Learning The idea of Hardware Accelerated Models in Azure was first conceived by Microsoft’s recent deep learning project, dubbed Project Brainwave. It utilises programmable electronic chips called Field Programmable Gate Arrays (FPGA) combined with a Deep Neural Networks (DNN) engine to enable a fast and powerful deep learning computing environment. In fact, in one of the blogs by the company, the team actually tested Project Brainwave on Intel’s Stratix 10 FPGA and found satisfactory results. The blog says, “Even on early Stratix 10 silicon, the ported Project Brainwave system ran a large GRU model — five times larger than Resnet-50 — with no batching, and achieved record-setting performance. The demo used Microsoft’s custom 8-bit floating point format (ms-fp8), which does not suffer accuracy losses (on average) across a range of models. We showed Stratix 10 sustaining 39.5 teraflops on this large GRU, running each request in under one millisecond.” Microsoft has leveraged on cheaper and customisable electronic chips such as FPGA, to power ML. FPGA is most suitable to projects which incorporate an overall smaller circuit design and are produced in lower volumes. Since ML programs or models usually call for an optimised design, circuitry such as FPGA makes sure that the computing needs are adequately met. This also provides an avenue for improving deep learning applications. The Competitive Edge With Hardware Accelerated Models beginning to emerge as part of Azure ML, Microsoft has upped the ante in ML with the likes of Google’s open-source ML framework, TensorFlow. Tensor Processing Unit or TPU is the hardware that runs ML and neural networks present in Google’s services which are based on TensorFlow. This makes Google’s ML compatible only with libraries with TensorFlow and makes it framework-specific. But Azure ML’s accelerated models can bring in libraries across other ML platforms such as CNTK, Theano as well as TensorFlow. Integrating Cloud For Applications On A Global Scale — Cosmos DB The flourishing number of ML or AI applications necessitate the need for computing and storage space. Cloud technologies combined with capable hardware address these requirements. But, this comes with an innate challenge of scalability. Scalability entails a tiring process of setting up a distributed database system in their network, with lots of data centers and other resources in place. This brings up significant investment to the company side. Most companies would opt out of this choice and prefer outsourcing their database needs. This is where Cosmos DB fits into the context. The database which is part of Azure’s portfolio, will now step up by offering a scalable distributed database system on a profitable scale. In fact, Azure Cosmos DB describes itself as “globally distributed, multi-model database services” with an emphasis on “high responsive”. The key features are highlighted below: Turnkey Global Distribution: This means relevant data can be distributed anywhere across regions present with Azure services with minimal latency or delay. In addition, applications need not be reset if Azure services are added or removed. Multiple Data Models And Popular APIs For Accessing And Querying Data : Cosmos DB uses a atom record sequence (ARS), which means different data models and APIs are compatible with each other. APIs for SQL, MongoDB, Cassandra and Gremlin are supported in the database. Elastically And Independently Scale Throughput And Storage On Demand And Worldwide: Cosmos DB gives a provision for changing storage size according to requirements of the business. It also boasts of attaining a per second throughput for quick access and modifications. No Database Schema Or Index Management: Users need not worry about the information such as schema or indexes as Cosmos DB provides a schema-independent environment to handle database queries. Apart from this, Cosmos DB is comparatively less expensive than its counterparts such as Amazon’s DynamoDB or Google’s Spanner, and also fares good in achieving minimum latency for queries. Conclusion: Microsoft’s aggressive move towards ML developments with Azure is quite interesting, but it needs improvement and is still at a nascent stage. However, the attractive features added to the cloud platform will definitely woo its clients and customers to take ML and AI a bit more further on an advanced level.","excerpt":"Microsoft has come a long way in cloud technology with its standalone platform Azure. The advancement is so progressive that Azure has expanded its capabilities to machine learning and artificial intelligence. The tech giant is consistently and collectively focussing on the current industry needs with respect to ML. With Azure, it has unlocked the potential […]","categories":["Global Tech"],"tags":["Microsoft Azure","new developments in cloud computing"],"author_name":"Abhishek Sharma","publish_date":"2018-05-16T06:41:20","publication_year":"2018","word_count":793,"keywords":["new developments in cloud computing","machine learning","artificial intelligence","AI","neural network","ML","RAG","Ray","deep learning","Microsoft Azure","TensorFlow","Azure ML"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","neural network","Azure ML","Ray","TensorFlow","RAG"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/true-potential-of-azure\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":25832,"title":"Ridge Regression Vs Lasso: Compare How They Work?","content":"As machine learning evolved, the conventional way of solving problems observed a diminishing shift. ML offered novel ways to tackle real-world problems with its methods and algorithms. With ML providing the ability for computers to learn from data, the problem could be analysed with different perspectives quickly. ML takes account of the optimisation technique in linear programming. This means that ML is generally considered as an optimisation problem. Concepts such as regression help with establishing a relationship between the vast amounts of data required for learning. In this article, we will analyse two extensions of linear regression known as ridge regression and lasso, which are used for regularisation in ML. How Regression Analysis Impacts ML Regression is basically a mathematical analysis to bring out the relationship between a dependent variable and an independent variable. This analysis is used in the supervised learning technique of ML. Therefore, in the context, it establishes the relationship between features in a model (the independent variables) and labels (the dependent variables). Regression helps in deriving a loss function or cost function for algorithms in ML. This is essential since loss functions determine the training accuracy in the ML model for the problem. So, here we discuss the linear regression models which are quite frequently used in ML. Linear regression (LR) is one of the simplest methods in regression. It involves determining linear relationships between continuous variables\/data. It is broadly classified into two types, simple LR and multiple LR. In the case of simple LR, there is one dependent variable and one independent variable whereas multiple LR has multiple independent variables that affect a single dependent variable. As mentioned earlier, LR is used in supervised learning. Supervised learning works with labelled data and comes across features that decide which pre-set label the data falls into, in a model. Ridge Regression In a multiple LR, there are many variables at play. This sometimes poses a problem of choosing the wrong variables for the ML, which gives undesirable output as a result. Ridge regression is used in order to overcome this. This method is a regularisation technique in which an extra variable (tuning parameter) is added and optimised to offset the effect of multiple variables in LR (in the statistical context, it is referred to as ‘noise’). Ridge regression essentially is an instance of LR with regularisation. Mathematically, the model with ridge regression is given by Y = XB + e where Y is the dependent variable(label), X is the independent variable (features), B represents all the regression coefficients and e represents the residuals (the extra variables’ effect). Based on this, the variables are now standardised by subtracting the respective means and dividing by their standard deviations. The tuning parameter is now included in the ridge regression model as part of regularisation. It is denoted by the symbol ƛ. Higher the value of ƛ, the residual sum of squares tend to be zero. Lower the ƛ, the solutions conform to least square method. In simpler words, this parameter decides the effect of coefficients. ƛ is found out using a technique called cross-validation. (More mathematical details on ridge regression can be found here). Lasso Least absolute shrinkage and selection operator, abbreviated as LASSO or lasso, is an LR technique which also performs regularisation on variables in consideration. In fact, it almost shares a similar statistical analysis evident in ridge regression, except it differs in the regularisation values. This means, it considers the absolute values of the sum of the regression coefficients (hence the term was coined on this ‘shrinkage’ feature). It even sets the coefficients to zero thus reducing the errors completely. In the ridge equation mentioned earlier, the ‘e’ component has absolute values instead of squared values. This method was proposed by Professor Robert Tibshirani from the University of Toronto, Canada. He said, “The Lasso minimises the residual sum of squares to the sum of the absolute value of the coefficients being less than a constant. Because of the nature of this constraint, it tends to produce some coefficients that are exactly 0 and hence gives interpretable models”. In his journal article titled Regression Shrinkage and Selection via the Lasso, Tibshirani gives an account of this technique with respect to various other statistical models such as subset selection and ridge regression. He goes on to say that lasso can even be extended to generalised regression models and tree-based models. In fact, this technique provides possibilities of even conducting statistical estimations. Comments In the case of ML, both ridge regression and Lasso find their respective advantages. Ridge regression does not completely eliminate (bring to zero) the coefficients in the model whereas lasso does this along with automatic variable selection for the model. This is where it gains the upper hand. While this is preferable, it should be noted that the assumptions considered in linear regression might differ sometimes. Both these techniques tackle overfitting, which is generally present in a realistic statistical model. It all depends on the computing power and data available to perform these techniques on a statistical software. Ridge regression is faster compared to lasso but then again lasso has the advantage of completely reducing unnecessary parameters in the model.","excerpt":"As machine learning evolved, the conventional way of solving problems observed a diminishing shift. ML offered novel ways to tackle real-world problems with its methods and algorithms. With ML providing the ability for computers to learn from data, the problem could be analysed with different perspectives quickly. ML takes account of the optimisation technique in […]","categories":["Deep Tech"],"tags":["linear regression","loss function","regression analysis","ridge regression"],"author_name":"Abhishek Sharma","publish_date":"2018-06-28T05:44:43","publication_year":"2018","word_count":863,"keywords":["linear regression","Go","machine learning","TPU","programming_languages:R","AI","ridge regression","loss function","ML","regression analysis","programming_languages:Go","BERT","llm_models:BERT","R"],"extracted_tech_keywords":["AI","machine learning","ML","TPU","R","Go","BERT","llm_models:BERT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/ridge-regression-vs-lasso-how-these-2-popular-ml-regularisation-techniques-work\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10010235,"title":"Researchers Combine AI &#038; Quantum Mechanics To Solve Renewable Energy Storage Problems","content":"“Researchers are developing AI that will accurately predict atomic interactions faster than the compute-heavy simulations of today.” The sun doesn’t always shine and to store energy[batteries] in its absence is a costly affair. So, with all this talk about going green, how is it possible to incentivise people towards futuristic yet unsustainable and expensive goals? To address this challenge, Facebook AI and the Carnegie Mellon University (CMU) have announced the Open Catalyst Project, a collaboration intended to use AI to accelerate quantum mechanical simulations by 1,000x in order to discover new electrocatalysts needed for more efficient and scalable ways to store and use renewable energy. The goal of the Open Catalyst Project, stated the Facebook AI team, is to discover low-cost catalysts in order to find solutions that are good alternatives of current solutions that are inefficient or rely on rare and expensive electrocatalysts like platinum, limiting their practicality. FB and CMU have even released the Open Catalyst 2020 (OC20) dataset and are also providing baseline models for the community to benchmark approaches against state-of-the-art and compare progress. Overview Of Open Catalyst Project Source: FAIR To achieve results that are dramatically faster than the compute-heavy simulations scientists rely on today, the team is developing AI that will accurately predict atomic interactions. “Calculations that take modern laboratories days, with the help of AI, could take seconds,” claimed the team at Facebook AI. “This has ramifications outside catalysis, and will enable scientists to rapidly explore and iterate on other challenges that involve quantum mechanics.” Finding efficient storage techniques is also a material science problem. The properties change with change in chemistry; the way interactions take place at the atomic level is crucial. At the atomic scale, the number of combinations of interactions is difficult to predict. Scientists rely on quantum mechanical simulation tools such as density functional theory(DFT). DFT is used to simulate the movement of atoms in a given scenario, estimating the energy of a system and attempting to find the configuration with the lowest energy state. This process is computationally intensive, taking hours or even days on high-end servers. DFT also scales poorly with an increase in the number of atoms. Where DFT fails, machine learning thrives. As DFT scales poorly, machine learning models can be trained using the results from initial DFT calculations and approximate the energy and forces of molecules based on past data. The researchers have identified a use case for machine learning but now comes the real challenge—data. So, the team at FAIR and CMU have even addressed this problem by releasing a dataset called the OC20. The OC20 data set comprises over 1.3 million relaxations of molecular adsorptions onto surfaces, the largest data set of electrocatalyst structures to date. “Producing this data set also required a substantial amount of engineering expertise and computing power. We ran DFT simulations on spare compute cycles over a period of four months. Facebook’s data centres will reach net zero emissions by the end of the year, making this a responsible and sustainable way to run the compute-intensive calculations necessary to build this data set,” said Facebook. Key Takeaways Calculations that take modern laboratories days, with the help of AI, could only take a few seconds. Success could usher in the widespread adoption of renewable energy, as costs come down and impact on the grid is mitigated by better storage. The implications for water quality remediation, medical treatment development, advanced manufacturing, or geochemistry. Current baseline models are still far from being useful in practical applications, so there is still much to be accomplished to realise the renewable energy solutions are needed. Know more about the Open Catalyst Project here.","excerpt":"“Researchers are developing AI that will accurately predict atomic interactions faster than the compute-heavy simulations of today.” The sun doesn’t always shine and to store energy[batteries] in its absence is a costly affair. So, with all this talk about going green, how is it possible to incentivise people towards futuristic yet unsustainable and expensive goals?  […]","categories":["Deep Tech"],"tags":["Carnegie Mellon University","Facebook AI","quantum application development system"],"author_name":"Ram Sagar","publish_date":"2020-10-20T17:00:37","publication_year":"2020","word_count":609,"keywords":["Go","Carnegie Mellon University","API","machine learning","Facebook AI","AI","programming_languages:R","Scala","RAG","Aim","quantum application development system","programming_languages:Scala","R"],"extracted_tech_keywords":["AI","machine learning","Aim","RAG","R","Go","Scala","API","programming_languages:R","programming_languages:Scala"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/ai-quantum-mechanics-solar-renewable-energy-storage\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10169858,"title":"Indian-Made Drones are Watching the Line of Control from the Sky","content":"As tensions between India and Pakistan reignite along the Line of Control (LoC), a new set of defence actors has taken flight. What the previous generations knew as the defence industry—defined by khaki uniforms, government public sector undertakings (PSUs), and bureaucratic tenders—is no longer the whole story. Today, India’s defence preparedness is shaped as much by engineers in Bengaluru or Hyderabad as by soldiers at the LoC. Private drone startups such as ideaForge, Garuda Aerospace, Asteria Aerospace, IG Drones, as well as companies like Scandron, are now integral to India’s frontline strategy. These are building and deploying unmanned aerial systems (UAS) that are not only confined to agriculture, solar, surveying, or disaster relief, but also for defence uses. These firms are rapidly scaling operations and increasing production to deliver tactical and surveillance drones, responding to heightened demand from defence forces seeking advanced aerial capabilities. “We are the tool that collects data right now,” said Vishal Saxena, vice president at ideaForge, in an interview with AIM. “We are behind the scenes, not visual. That’s our job. Intelligence, surveillance, and reconnaissance (ISR) systems are typically that.” Private Drones, Public Battles Focused initially on civilian applications such as agriculture and infrastructure monitoring, these drone companies are now adapting their technologies for military use. This pivot reflects a broader trend where private tech firms are becoming integral to national security strategies, supplying equipment that was once exclusively the domain of state-run defence manufacturers. Garuda Aerospace, once best known for its agricultural drones, has launched a line of dual-use platforms tailored for both civilian and paramilitary use. The company has openly discussed plans to ramp up production of surveillance drones capable of day-night operation in high-altitude zones. Founder Agnishwar Jayaprakash has spoken about India’s growing need for “tactical autonomy”, the strategic role that Indian startups can play, and the crucial role Garuda drones played in ‘Operation Sindoor’. “Two thousand drones equipped with payload dropping mechanisms are ready to be delivered, and 2,000 more are in production at the Garuda manufacturing facility,” he told AIM. Other startups like NewSpace Research and Technologies and Asteria Aerospace are developing more advanced systems, including swarming drones and autonomous AI-powered surveillance platforms. While still in development or early deployment, these innovations represent the future of battlefield tech—decentralised, fast, and difficult to detect or counter using traditional air defence methods. A Surge in Defence Demand In light of recent operations like India’s Operation Sindoor, the defence ministry has issued urgent requirements for a wide range of drone capabilities, from high-altitude ISR units to loitering munitions. For companies like Scandron, the ask is both specific and immediate. “The Army has come up with a basic list of requirements. Some of those are readily met, some require fast-track development,” Arjun Naik, CEO of Scandron, told AIM. “Our drones fit those bills and have been modified to meet the terrain-specific requirements.” Scandron, known for its logistics and support unmanned aerial vehicles (UAVs), is now scaling production from 300 to 400 units a month to between 800 and 2,000, depending on requirements by the defence. “We have expanded infrastructure and human resource capability,” Naik added. ideaForge is experiencing a similar spike due to its capability, but also acknowledges the capacity gap. “While the user may want 10,000 drones tomorrow, can I manufacture that? The capability exists. The capacity, however, is a question mark,” Saxena said. Indigenous Roots Firms are investing heavily in research and development while chasing certification from the Directorate General of Aeronautical Quality Assurance (DGAQA), which governs drone systems intended for defence use. IG Drones has emerged as a vital player with its vertical take-off and landing (VTOL) and first-person view FPV drone systems. These are engineered for high-endurance surveillance and real-time tactical reconnaissance in rugged terrains. “Our drones are proudly ‘Made in India’ and contain no Chinese components,” Bodhisattwa Sanghapriya, founder and CEO of IG Drones, told AIM. “With over 75% indigenous capability, we are redefining sovereign tech at the border.” IG Drones has also developed India’s first indigenous military drone simulator, offering low-risk, high-fidelity training, a key force multiplier as defence forces prepare to induct large fleets. Saxena also expressed that drones can be made in India. At the least, ideaForge is consciously making sure not to touch any components from “geographies of concern”. “Today, apart from those very critical systems which are not available in India, 97% of the components could come from India.” He added that they may not be the world’s best components, or only 60–70% of them may be. “There would still be about 20–30% of components that need more deliberation.” Know Your #Military DronesFrom real-time #reconnaissance to precision strikes, drones are reshaping modern warfare. For instance, our surveillance drones—such as the #NETRA and #SWITCH UAVs—are deployed globally for ISR, search & rescue, counterinsurgency, & border surveillance pic.twitter.com\/4bLqwFoeOp— ideaForge Technology Limited (@ideaforge_tech) May 12, 2025 For indigenous development, even Mughilan Thiru Ramasamy, co-founder and CEO at Skylark Drones, said, “We are working to figure out how our software capabilities, especially around surveillance, will be useful for the Indian defence.” The Quiet Arms Race According to reports, India employed Israeli Harop drones for precision strikes on Pakistani air defences, to which Pakistan claimed to have downed multiple units. This rapid militarisation isn’t just about meeting immediate defence needs. It’s triggering a larger shift in how India views innovation, sovereignty, and supply chains. While the government’s ‘Make in India’ and ‘Aatmanirbhar Bharat’ programmes have given startups a platform, it’s the private sector’s R&D push that’s now taking centre stage. “There is a sweet spot emerging, tactical drones that do 50 to 100 km with distributed capability,” Saxena explained. “Instead of one MQ-9 Reaper doing 10 tasks, what if you had 100 drones doing one each? The mindset is shifting toward distributed platforms.” Naik agreed to this, acknowledging that the R&D cost is huge. However, he pointed out that this is precisely the path taken by Israelis and Americans, which he hopes to follow. “We don’t quite have that level of technological advancement yet, but our drones cost a fraction of what the Israeli drones cost.” Who Holds the Trigger? This new defence-industrial ecosystem also introduces murky questions around accountability. While none of the featured companies currently supply armed drones, their technologies undeniably enable more precise and potentially lethal targeting. “Our drones are not armed,” Saxena said. “Luckily, that dilemma isn’t there. But if we were making that tech, the question would be: should there even be a war? That’s a larger question than the tech itself.” As Naik succinctly put it, “We’re building tools for security, not aggression. But once you sell to the military, you don’t control the mission.” The New Defence Line Despite the urgency, many companies warn that the lack of long-term procurement clarity hampers planning. They argue that the Indian drone sector needs both sustained demand and a robust domestic supply chain to truly scale. “Drone startups are not just vendors, they are catalysts of national transformation,” Sanghapriya further said. Meanwhile, Naik remarked, “India has about 900 drone companies, but fewer than 10 are serious players. The rest are traders. Now the real manufacturers will come to the forefront.”With ethical tightropes to walk, regulatory frameworks to solidify, and enemy radars to dodge, India’s drone startups are flying in uncharted airspace. And for now, they’re holding formation.","excerpt":"Today, India’s defence preparedness is shaped as much by engineers in Bengaluru or Hyderabad as by soldiers at the LoC.","categories":["Deep Tech"],"tags":["AI in defence","defence drones","defence ministry","Drone","Drones in india","ideaForge","IG Drones","Scandron"],"author_name":"Sanjana Gupta","publish_date":"2025-05-13T16:05:09","publication_year":"2025","word_count":1221,"keywords":["Go","Drone","API","programming_languages:R","AI","R","defence ministry","innovation","defence drones","autonomous AI","Aim","Scandron","ViT","IG Drones","Drones in india","ideaForge","AI in defence","startup"],"extracted_tech_keywords":["AI","Aim","R","Go","API","ViT","innovation","startup","autonomous AI","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/indian-made-drones-are-watching-the-line-of-control-from-the-sky\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10138990,"title":"Meta Launches Meta Spirit LM, an Open Source Language Model for Speech and Text Integration","content":"Meta has unveiled Meta Spirit LM, an open-source multimodal language model focused on the seamless integration of speech and text. This new model improves the current text-to-speech (TTS) processes, which typically rely on automatic speech recognition (ASR) for transcription before synthesising text with a large language model (LLM) and converting it back to speech. Such methods often overlook the expressive qualities of speech. Meta Spirit LM employs a word-level interleaving method during training, utilising both speech and text datasets to facilitate cross-modality generation. The model comes in two versions, Spirit LM Base, which utilises phonetic tokens for speech modelling, and Spirit LM Expressive, which incorporates pitch and style tokens to convey tone, capturing emotions like excitement or anger. The new model allows users to generate more natural-sounding speech and demonstrates the capability to learn tasks across different modalities, including ASR, TTS, and speech classification. Meta aims to inspire further development in speech and text integration within the research community. Meta is not Alone Similar to Spirit LM, Google recently launched NotebookLM, which can convert any text into a podcast. With this feature, users can input a link, article, or document, and the AI assistant generates a podcast featuring two AI commentators engaged in a lively discussion on the topic. They summarise the material, draw connections between subjects, and engage in banter. NotebookLM is powered by Google’s Gemini 1.5 model for AI-driven content generation and voice models for lifelike audio outputs. It is supported by a custom-built tool called Content Studio, which provides editorial control. OpenAI recently launched its Advanced Voice Mode on ChatGPT, and since then, people have been experimenting with it. Deedy Das from Menlo Ventures used it for the dramatic reenactment of a scene in Hindi from the Bollywood movie Dangal. Another user posted a video on X where ChatGPT was singing a duet with him. The possibilities with the voice feature of ChatGPT are endless. Recently, Kyutai, a French non-profit AI research laboratory, launched Moshi, a real-time native multimodal foundational AI model  capable of conversing with humans in real time, much like what OpenAI’s advanced model was intended to do. Hume AI introduced EVI 2, a new foundational voice-to-voice AI model that promises to enhance human-like interactions. Available in beta, EVI 2 can engage in rapid, fluent conversations with users, interpreting tone and adapting its responses accordingly. The model supports a variety of personalities, accents, and speaking styles and includes multilingual capabilities. Meanwhile, Amazon Alexa is partnering with Anthropic to improve its conversational abilities, making interactions more natural and human-like.","excerpt":"The new model allows users to generate more natural-sounding speech and demonstrates the capability to learn tasks across different modalities, including ASR, TTS, and speech classification.","categories":["AI News"],"tags":["Meta"],"author_name":"Siddharth Jindal","publish_date":"2024-10-21T15:56:20","publication_year":"2024","word_count":423,"keywords":["Anthropic","ChatGPT","Meta","Meta AI","TPU","OpenAI","AI","Go","ML","Aim","R"],"extracted_tech_keywords":["AI","ML","ChatGPT","OpenAI","Anthropic","Meta AI","Aim","TPU","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/meta-launches-meta-spirit-lm-an-open-source-language-model-for-speech-and-text-integration\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10065446,"title":"Elon Musk’s The Boring Company raises USD 675 mn in series C","content":"The Boring Company, founded by Elon Musk, has raised USD 675M at a USD 5.675B valuation from A-list investors. The Boring Company, in its tweet, said, “Purpose is to recruit amazing people, scale-up boring machine production & build thousands of miles of tunnels.” The Boring Company just raised $675M at a $5.675B valuation from A-list investors.Purpose is to recruit amazing people, scale up boring machine production & build thousands of miles of tunnels.https:\/\/t.co\/BpSE9syuGs— The Boring Company (@boringcompany) April 21, 2022 Defeating traffic is the ultimate boss battle. Even the most powerful humans in the world cannot defeat traffic,” said Elon Musk. Please consider working at The Boring Company! Our goal is to solve traffic, which plagues every major city on Earth. https:\/\/t.co\/1tHNhIo7sd— Elon Musk (@elonmusk) April 21, 2022 The company announced, “Our Series C funding round of USD 675 million, led by Vy Capital and Sequoia Capital, with participation from Valor Equity Partners, Founders Fund, 8VC, Craft Ventures, and DFJ Growth. The financing now values The Boring Company at USD 5.675 billion.” The company also said a group of leading real estate partners, including Brookfield, Lennar, Tishman Speyer and Darca, are joining the round. About TBC The Boring Company creates safe, fast-to-dig, low-cost transportation, utility, and freight tunnels. The company spent the last five years building core technologies designed to solve traffic and beautify cities via Loop and Hyperloop. The funds from this round will be utilised to increase hiring across engineering, operations significantly, and production to build and scale Loop projects, including Vegas Loop and others, and accelerate the research and development of Prufrock and future products. The Boring Company is hiring for numerous roles and seeks talented individuals ready to get their hands dirty to make a meaningful impact that will last for generations to come. Those interested can apply today at boringcompany.com\/careers.","excerpt":"The Boring Company (TBC) was founded by Elon Musk to revolutionise transportation.","categories":["AI News"],"tags":["Elon Musk"],"author_name":"Poornima Nataraj","publish_date":"2022-04-21T17:08:01","publication_year":"2022","word_count":305,"keywords":["Go","API","funding","programming_languages:R","AI","programming_languages:Go","Elon Musk","R"],"extracted_tech_keywords":["AI","R","Go","API","funding","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/elon-musks-the-boring-company-raised-675-m-in-series-c-funding\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":36915,"title":"5 Biggest Mistakes New Data Scientist Make And How To Avoid Them","content":"As emerging technology is growing in an exponential manner, the organisations of all sizes are in the rat race and all are seeking to blend in the technology arena. According to this report, India has more than 50,000 jobs in both data science and machine learning lying vacant apparently because there is not enough talent to fill them. It is high time for those who really want to pursue a career in data science. In this article, we list 5 major mistakes that amateur data scientists make on the job and how to avoid them. 1| Focusing Less On Data This can be said as one of the major issues faced by an amateur data scientist of which they are unaware most of the time. Since you have started your career as a data scientist, you have grasped all the fancy algorithms of machine learning as well as deep learning. One thing to keep in mind that data plays a very crucial role and ignoring it will just be doing “Garbage in-Garbage out”. Also, after you have done analysing the data, runs the algorithm and got fair metrics, you think that the work is accomplished. That is where you ignored all the possible risks that may b seen in the future. Your eagerness for completing the task may lead you to risks such as data leakage, overfitting, and other biases. How To Avoid Them When building a model, most of the work is done at the data and features level. It is better to concentrate a little more on data than on the algorithm because your data and its features will shape your model at the last. The quality of your final model will completely depend on the data than the algorithm. 2| Assuming The Algorithm Is More Important Than Domain Knowledge Many budding data scientists rely on the theoretical knowledge of algorithms while trying to build a model. Knowing all the complex algorithms will not help you to build a fair working model. If you want to build a good model, it is crucial to know and understand the data you will be using, the purpose behind your model, and basically the domain knowledge. How To Avoid Them Pushing data randomly without exploring it much will show biases in the results of your model. Hence, it is needed much to do some exploratory data analysis which will help you to make some hypothesis of the model and you can be informed about what you are doing. 3| Making The Algorithm Complex This mistake may cost you a little more. Most of the time, budding data scientists forget that a simple machine learning model with good data can beat the complex algorithms in a model. It is not always needed to use fancy complex algorithms to build a robust model. The junior data scientist often gets lured by the fancy algorithms of deep learning and start thinking that everything can be solved using those algorithms with a little knowledge of exploring data and thus failed to meet the goal in the long run. How To Avoid Them Techniques like logistics regression, linear regression, etc. can sometimes outperform complex algorithms. Firstly, you should think what your main purpose is and then start with building a simple model with a less fancy algorithm. It is better not to make it complex which is already complex. 4| Spending Less Time On Exploring And Visualising Data Most of the time the budding data scientists prefer to build a model than to visualise and explore the data. They lack the idea that by spending more time on understanding the data can gain you deeper insights on what the outcome of your model will be. Being curious and eager to finish building a model and complete the task while ignoring the exploring and visualisation part can cause serious damage to the model. How To Avoid Them The basic and most important tools of a data scientist are to explore and visualise the data. Understanding your dataset is the foremost task that an aspiring data scientist should do as it will later reflect in your model. 5| Less Communication This usually happens to amateur data scientists that they hesitate to question about the difficulties and when you are not ensured about certain issues. They often shy away from putting their views forward as afraid of being criticised, forgetting that without drawing any feedback one cannot improve much further. How To Avoid Them To be a keen data scientist, you have to be a good communicator. It really helps a lot. One should always keep in mind that data scientists are meant to solve other people’s issues and without communicating whether it be inside the organisation or some outside business clients, it is merely a difficult as well as unsolvable task.","excerpt":"As emerging technology is growing in an exponential manner, the organisations of all sizes are in the rat race and all are seeking to blend in the technology arena. According to this report, India has more than 50,000 jobs in both data science and machine learning lying vacant apparently because there is not enough talent […]","categories":["AI Trends"],"tags":["Data Science","Data Scientist"],"author_name":"Ambika Choudhury","publish_date":"2019-03-27T12:41:24","publication_year":"2019","word_count":803,"keywords":["data science","Go","machine learning","programming_languages:R","AI","ML","programming_languages:Go","deep learning","GAN","Data Science","Data Scientist","R"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","data science","R","Go","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/5-biggest-mistakes-new-data-scientist-make-and-how-to-avoid-them\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":44034,"title":"After Drones &#038; Self-Driving Vehicles, Japan Shows That Flying Cars Are No Longer A Distant Dream","content":"(A still from Netflix’s Altered Carbon) While drones and self-driving cars have been making the headlines, Japan has reportedly have cracked the code of making flying cars. Though the car was caged and only hovered for about a minute in the air, it was a successful attempt to make the vehicle fly. In an experiment right out of the sci-fi movies, this has been accomplished by Japan’s NEC Corp. While it is a small development directing towards bigger accomplishments, there is also a debate on whether it really is a flying “car” or just a bigger version of a drone. What Can The Flying Car Accomplish? The prototype which was unveiled by Japan’s NEC Corp had four propellers that could smoothly hover for about a minute. It was powered by a battery and could rise to about the height of 3 meters or 10 feet above the ground before setting down again. While the prototype that Japanese electronics maker demonstrated was flown without passengers in it, the company claims that it is capable of doing so in the future. The Japanese government has been trying to push its autonomous vehicle strategies where it is promoting self-driving cars to run on public roads. It even updated its traffic law to set new standards for driver conduct and vehicle safety. It is ambitious about having automobiles that can handle o have automobiles that can handle, i.e. the level 3 autonomous vehicles on the roads by 2020. It had also announced the companies to build electric or hybrid electric vehicles with driverless capabilities which can take off and land vertically, a feature commonly known as Vertical take-off and landing or VTOL. An example of this was imagined in a noted Netflix series Altered Carbon. Photo source: A prototype of NEC’s flying car. Photographer: Kiyoshi Ota\/Bloomberg Air Mobility The work on flying cars began with the idea to make deliveries on unmanned flights possible and start commercialising flying vehicles by 2023. The end goal is to make Japan a world leader in flying cars especially after it missed out on advancements in technology such as electric cars and ride-hailing services. “Japan is a densely populated country and that means flying cars could greatly alleviate the burden on road traffic,” said Kouji Okada, a leader of the project at NEC. “We are positioning ourselves as an enabler for air mobility, providing location data and building communications infrastructure for flying cars.” Though the end goal is to have flying cars in the country, the Japanese corporation, NEC isn’t planning to mass-produce the flying car. There has been a drastic increase in small and passionate flying car community in Japan that has been accomplished because of the engineering expertise and the right environment that the country provides. They also boast of Drone Fund, a specialised fund that inflows funds for development in autonomous aircraft and flying car space. This prototype by NEC Corp. was developed in partnership with Cartivator who spent around a year developing the model which resulted in a  3.9 meters long, 3.7 meters wide, 1.3 meters tall, and 150 kilograms flying car prototype. They aim to mass-produce the transportation machine by 2026. Other Countries Working On Flying Cars Interestingly, Japan is not the only country that is working on realising their dreams of flying cars. There are many companies across the globe who have shared similar intentions. Kitty Hawk Corp: Founded by Larry Page, it is working on bringing flying cars to reality. With a world-class team of decades of expertise in commercial aviation, aerospace, automotive engineering, advanced manufacturing, flight test, industrial design, customer experience, and regulatory compliance, they are designing, resting and building futuristic vehicles. Uber Technologies: The Uber Elevate team is working towards transforming the world through aerial ridesharing to overcome the congested ground traffic. It is aiming for air transportation by 2023 between suburbs and cities by bringing vertical takeoff and landing aircraft to reality. Boeing: It was reported that Boeing has completed the first pilotless test flight of its electric autonomous passenger plane at an airport in Manassas, Virginia. The 30-foot-long plane took off vertically and hovered for less than a minute. The purpose was to test its autonomous functions and ground control systems. Airbus: Airbus had also conducted the first successful test flight of its Vahana electric vertical take-off and landing (eVTOL) aircraft. It hovered for 53 seconds above the ground before landing. Dubai hoverbikes: Dubai Police had announced to their intentions to use hoverbikes or the flying motorbikes which are in line with sci-fi developments that the Dubai government is looking forward to adopting. They plan to use it in emergency situations, given how quickly their batteries can drain. The city has also been testing robotic flying taxis with the hopes that at least 25% of all travel in the city will be automated by 2030.","excerpt":"While drones and self-driving cars have been making the headlines, Japan has reportedly have cracked the code of making flying cars. Though the car was caged and only hovered for about a minute in the air, it was a successful attempt to make the vehicle fly. In an experiment right out of the sci-fi movies, […]","categories":["AI Features"],"tags":["autonomous cars","drones","Self Driving Cars"],"author_name":"Srishti Deoras","publish_date":"2019-08-10T11:00:15","publication_year":"2019","word_count":808,"keywords":["Go","autonomous cars","drones","AI","programming_languages:R","autonomous AI","Self Driving Cars","programming_languages:Go","Aim","GAN","R"],"extracted_tech_keywords":["AI","Aim","R","Go","GAN","autonomous AI","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/after-drones-self-driving-vehicles-japan-shows-that-flying-cars-are-no-longer-a-distant-dream\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10043082,"title":"How Mumbai Based Startup Flexiloans Uses AI","content":"Founded by Indian School of Business alumni Abhishek Kothari, Deepak Jain, Ritesh Jain and Manish Lunia in 2016, Flexiloans claims to be one of the top three SME digital lenders in India, growing at a CAGR of 500%. The Mumbai-based startup has disbursed over Rs 1,000 crores to SMEs across 1,400 plus Indian cities till date. Flexiloans receives more than 100,000 loan applications every month from Tier III and Tier IV cities. The digital lender has partnered with the likes of Flipkart, Amazon, Paytm, PineLabs, TrueCaller and Mswipe to offer credit to over five million SMEs in India. In an exclusive interview with Analytics India Magazine, Abhishek Kothari explains how Flexiloans became a leading player in the Indian SME lending ecosystem in just five years. A graduate from IIT-Bombay, Abhishek is also the founder of Join My Moment. He has served as the VP of Head of Insights (Data Marketplace) at Barclaycard and as the VP at Fractal Analytics. At Flexiloans, he is responsible for the product, data science, engineering and risk verticals. Excerpts: AIM: What problem is Flexiloans trying to solve? Abhishek Kothari: Before we started the company, India was witnessing tailwinds with the introduction of Aadhar, GST and Jio. Today, India has about six crore small businesses– the kirana-wala, pharmacy-wala and kachori-wala; four crore of them have never been able to borrow from a bank. We wanted to solve this challenge with Flexiloans. Small businesses contribute to about 20 percent of India’s GDP today. The credit to GDP ratio is about 40 to 50 percent. Since small business owners do not have any assets, their only way to expand the business is by borrowing money. Credit is an integral part of small businesses. We are trying to bring financial access at a click for small businesses. If we are able to solve this effectively, we will have a direct impact on India’s GDP and its employment rate. Our mission is no longer just providing loans to SMEs, but to ensure an SME of India can be treated at par with an SME in a developed country. We recently started BiFrost, our gateway into the capital. We have 12 partners– banks, NBFCs, and capital partners– that we can tie up with for access to capital. AIM: What role does technology play in your operations? Abhishek Kothari: Technology solves three problems for us: Access, assessment & servicing. We have built the largest loan origination system in the country.  Jack Ma said, “We need banking, we do not need banks,” and we strongly believe that embedded finance is the future. Our in-house techstack effectively converts the borrowing experience into a buying experience. Amazon, PayPal, Google Pay, PineLabs, Flipkart, MSwipe etc use our lending services. Our biggest innovation in the origination side was creating this embedded finance API stack which allowed us to dramatically scale. First step is customising the form for the merchant by an Intelligent Journey Planner. Second, technology is used to analyse the documents  and data that come through  in real-time. The risk models effectively convert the data generated in the previous steps into a risk score which determines the eligibility of a user to get a credit. If the user accepts the credit offer by Flexiloans, the fourth pillar of the techstack comes in– finding the capital provider. The money can either come from Flexiloans itself (since Flexiloans is also a NBFC), or from one of the capital providers that we have partnered with. The system selects the most suitable lender. Post which, the user has to go through the usual KYC process. The system will connect the user to an agent who will effectively do the video KYC for the user. From a data science perspective, the most efficient algorithms for risk models are logistics regression, or random forensic techniques. AIM: Tell us about your core tech stack Abhishek Kothari: Most of our technologies are built in-house, so we do not have a lot of plug-and-play software. We use a lot of open-source for most of our builds. For our technology application, we are big on Node.js on the backend, and Laravel on the frontend. Our app is built on Native Android. We are not on React since we feel Native Android offers a better experience. On the data science side, we use Python for more complex and computer vision-based applications. We use R for most of our risk model building and statistical analysis. For visualisation and business tracking, we use Power BI. We are big believers in AWS as well. AIM: What’s your advice to data scientists on approaching a problem? Abhishek Kothari: Having knowledge of the statistical side of data science is more important than having knowledge about the technological side of data science. People can learn R and Python in a month but I cannot create a passion or inspiration for statistics if they do not have it already. Secondly, the curiosity of the data scientist is extremely important. I believe the best data scientists are the most curious people. Up to 80 percent of data scientists are not as curious– which is why I have hired data scientists who have not had much past experience.","excerpt":"Founded by Indian School of Business alumni Abhishek Kothari, Deepak Jain, Ritesh Jain and Manish Lunia in 2016, Flexiloans claims to be one of the top three SME digital lenders in India, growing at a CAGR of 500%. The Mumbai-based startup has disbursed over Rs 1,000 crores to SMEs across 1,400 plus Indian cities till […]","categories":["AI Startups"],"tags":["FinTech","indian statistical service"],"author_name":"Debolina Biswas","publish_date":"2021-07-07T15:00:00","publication_year":"2021","word_count":862,"keywords":["data science","Go","AWS","AI","Git","computer vision","Python","Aim","analytics","FinTech","R","indian statistical service"],"extracted_tech_keywords":["AI","computer vision","data science","analytics","Aim","AWS","Python","R","Go","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-startups\/how-mumbai-based-startup-flexiloans-uses-ai\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":25861,"title":"Annual Survey On Data Science Recruitment In India 2018","content":"Indian and multinational companies from almost every industry — be it Fortune 500 name or a local startup — are hiring data science experts to help them garner insights from big data. The entire industry has seen a sharp increase in demand for highly-skilled professionals. Companies are also on a constant lookout for talented persons who can fill the gap. Our survey for June 2018 endeavoured to find out the current scenario of hiring trends and jobs available in data science. The data science sector is flourishing to such an extent that our earlier Jobs Survey published in association with Edvancer revealed that more than 90,000 job openings in data science and related fields are being advertised in India as of now. About The Survey: We tried to look at variations in skill sets, work experience and educational qualification in this survey. We took opinions from three groups of persons — hiring managers, job seekers and students — to get a thorough idea of the hiring scenario in this swiftly-developing area. Our survey was met with much enthusiasm — and we got some great insights from it. Some of them were expected, and many of them were real eye-openers. 1. Importance Of Formal Education: We asked our survey takers to tell us if a degree or formal education in data science was necessary to get a job in data science. The respondents clearly thought it was. 33% respondents thought that formal education was crucial to get a job in this sector. This is in keeping with the current trend in the industry. 28.3% respondents thought it was important as it can be clearly seen by the increasing reliance by employees as well as students on MOOCs to keep themselves updated. Only a minority of the respondents — 11% — thought formal education was completely unnecessary. 2. Programming Experience: Our data clearly shows that a vast majority of the respondents think that experience in programming was important to get a data science-related job. In fact, the numbers show: 48% of the respondents think that prior experience was important to get a job in the new tech sector. 29% respondents thought that it was crucial for their career. This is in line with what the organisations have been thinking. Most of the companies in new tech think that educating an employee for a basic skill like programming can be time-consuming and a drain on resources. 3. Work Experience: A clear majority of our respondents have shown that having an experience in a similar field is definitely helpful in landing a job in data science. For example, a clear 35.8% people have said that a job experience is crucial in data science sector. Another 26.2% have said that it is important. This is in keeping with studies which have shown that a majority of companies in the analytics and data science sector prefer candidates with five or more years of work experience. Only 2% of the survey takers did not agree with the statement. 4. Transition From Non-Data Science Background To New Tech: This questions seemed to have stumped our respondents. Here, the answers seem of two polar opposite opinions. Firstly, almost 30% of the people did not have an opinion about the matter. Otherwise, 15.2% of the respondents thought it was very difficult to make the transition from a non-tech background to data science. 23.2% thought it was difficult, but not impossible to do so. This response is clearly not surprising, given the great pace at which the industry is progressing. It is already hard for people in the industry to keep up with changing trends, and therefore jumping into a vibrant sector suca as data science is perceived as very difficult. On the other hand, only 5.6% people thought it was very easy to do so. 5. How To Get An Entry-Level Job In Data Science: This question had a clear-cut answer — perhaps because of the personal experiences of our respondents. 42.7% people thought that internship was the best way to make your way into a company. You make contacts, meet mentors and friends that last for a lifetime. 21.8% respondents voted joining data science communities or networking as the great way to get jobs. 6. What Language Should An Aspiring Data Scientist Learn First: Having a mastery over one or more programming languages is considered a great asset in the data science and analytics community. Here, our respondents — job seekers as well as recruiters — had a clear idea about what languages were the most popular in the market right now. Python was clearly the most popular language for data science practitioners as well as learners. Over 47% respondents said that it was crucial for job seekers to know this language to get an opportunity in data science. The next favourite language was R at 39% 7. Top Skills For Data Scientists: Our respondents identified three important skills for anyone to flourish in a data science-related field: Statistical modelling at 35% Machine learning at 28% Business intelligence at 17% Data mining came in next with 10% votes 8. Reliance On MOOCs For Skill Upgradation: Professionals as well as students are now relying heavily on massive online open courses. Our survey clearly indicates that: 33% of the respondents relied on MOOCs extensively to upgrade their skills in new tech and keep themselves updated with emerging technologies. On the other hand, 28% used it frequently for the same purpose. Only 9% of the respondents said that they didn’t use MOOCs at all. 9. How To Find A Job In Data Science: Here, the answers were unanimous, as well as expected. At 47.1% LinkedIn, the social networking website for professionals seems to be the first choice for looking for as well as posting a job in data science. The next popular way to find a job is checking with friends and acquaintances to look for employment opportunities at 19%. This is followed by popular job portal Naukri at 15.7% 10. Which City Has Most Job Opportunities For A Data Scientist: This question had a congruent, solid vote for Bengaluru with 73% of respondents saying that the Garden City had the best job opportunities in new tech. The next favourite was Mumbai at 13.9%. 11. Which Industry Holds The Most Opportunities For Data Scientists: Data scientists are in great demand, as discussed above. The results are thus scattered across several sectors. Most of the industries starting with IT and services at 26.8% Followed by e-commerce at 23.6% And banking at 23.6% 12. Area Of Interest For Data Scientists: This survey also showcased that data scientists want to work in three key areas: Customer service at 37.2% Banking  at 32.2% Medicine 13.2%","excerpt":"Indian and multinational companies from almost every industry — be it Fortune 500 name or a local startup — are hiring data science experts to help them garner insights from big data. The entire industry has seen a sharp increase in demand for highly-skilled professionals. Companies are also on a constant lookout for talented persons […]","categories":["AI Features"],"tags":["Data Science"],"author_name":"Prajakta Hebbar","publish_date":"2018-06-29T06:10:07","publication_year":"2018","word_count":1119,"keywords":["big data","data science","Go","machine learning","AI","Python","GRU","analytics","GAN","Data Science","R"],"extracted_tech_keywords":["AI","machine learning","data science","analytics","Python","R","Go","big data","GAN","GRU"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/annual-survey-on-data-science-recruitment-in-india-2018\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":11522,"title":"Journey of Email Marketing and looking beyond open rates","content":"One of the questions marketing professionals (or campaign managers) ask quite often is “What’s a good open rate for an e-mail?” or “What open and click percentage should one aspire for ideally?” Marketing consultants dread this question as the usual answer is “It depends”, followed by an explanation that one should look at internal or industry benchmarks. The marketer expects a definite number while the marketing consultants most often have none. The rationale behind such a reply stems from the fact that one should not look at open rates in isolation. One should look at whether the response rates are increasing\/decreasing vis-à-vis last year\/half year etc. How is it trending over a period of time? Imagine an acquired database that the marketer mails to, a 1% response might be a good one to get while a campaign say co-branded by an educational institute and sent to an alumni network, a 40% response rate will also be deemed low. Thus, the open rates work best as an in-house benchmark to track over time, because it can only signal a progress or problems with customer engagement. Open rates tell half the story The open rates tell a marketer half the story anyway. What the marketer must have in mind is to calculate the overall conversion or the desired outcome expected from an e-mail marketing campaign. So, one of the metrics that the marketer must have to define a successful e-mail marketing campaign entails both response and conversion (Response Rate% x Conversion Rate%). One cannot be seen independent of each other. Developing Key components of e-marketing strategy The marketer, these days, is under constant pressure from the product or business teams of the organization as the cost and effort for sending direct e-mail communication has come down rather steeply over time. Since the cost of sending an e-mail is miniscule compared to the commercial benefit it can drive, why not send a mail? What the business may not realize is the impact of the ‘decay effect’ of e-mail marketing campaigns and thus the need to develop a contact policy by the company Getting your strategy of Email marketing right In order to maximize the chances of their e-mail marketing campaign, here are some pointers marketers must watch out for:- CUSTOMERSINGLE VIEW There is customer information available in various databases for a company which may be fragmented or in different source systems. Right from when a customer purchased from the company to when he\/she applied for the second or third product; from when he\/she came online to when he\/she started seeing the mails on mobile devices; there is data everywhere to analyse. There needs to be a single view of these customers, so that a very well-defined customer journey and behaviour can be tracked and analysed. A marketer can weave a story on the customer journey and build a well-orchestrated communication strategy. CUSTOMER UNDERSTANDING As some old school marketing professor will say, there is no substitute to simple one way and 2 way frequencies. Whether the customer base is young or old or is metro or non-metro base will define your e-mail campaign strategy. When the customer demographic is overlaid with transaction or usage behaviour, it throws some new light on the customer. The transactional analysis will always lead to more insights on the customer’s spends and usages. The third dimension of understanding customer behaviour better is through their social media behaviour. This can be used more as a validation of understanding the customers better. A detailed profiling study of the customers also helps the marketer to build personas which can then be used to acquire new customers by building lookalike models and making it a part of their e-mail marketing strategy. A marketer can also estimate the opportunity much better basis the personas developed. BUILDING CUSTOMER PROFILES Post doing a good profiling exercise, number of segments a marketer can build is infinite. The more segments you can manage the better will be the return on investments as you will get very close to what an individual customer wants. As the number of customers per segments increases, heterogeneity increases. A number of banks therefore are now adopting what is called the Lifecycle Based Needs. The whole customer base can be divided into 5-6 segments with perceived financial situation and therefore banking needs. The marketer basis the customer-based understanding can create the lifecycle on say the risk appetite of the customer or maybe on the propensity to try newly launched product or offers. How these get weaved-in into the e-mail marketing campaign strategy is a critical part of the success of campaign itself. CONSISTENCY IN COMMUNICATION ACROSS CHANNELS E-mail marketing must take advantage of the multiplier effect of communications channels and media mix elements. If it is an end of season push on a credit card, it is important that this message be orchestrated through all mediums of communication. What the customer would see in their Facebook ads\/walls must not be diametrically opposite to what they get in their e-mail campaigns. It is important to orchestrate the messaging through a structured campaign wave. Summary India is going through an interesting phase currently with Indian marketers believing in the power of marketing technology and marketing automation through e-mail marketing and thus increasing their marketing investments on such digital channels. E-mail campaign-led programs remain a significantly more effective and cost optimal way to acquire customers compared to any other digital channels of communication. Number of customers with e-mail is rising, however, the time spent per customer on e-mail shows a downward trend with alternate channels like mobile app and notifications on the rise. Thus the challenge for the current day marketer is to create an interest in the mind of customer right from pre-planning the e-mail campaign to the first touch point (when the e-mail lands in the inbox) to when the journey ends (marketing objective is achieved). Marketers need to look beyond clicks and at the entire life stage of the customer across such campaigns.","excerpt":"One of the questions marketing professionals (or campaign managers) ask quite often is “What’s a good open rate for an e-mail?” or “What open and click percentage should one aspire for ideally?” Marketing consultants dread this question as the usual answer is “It depends”, followed by an explanation that one should look at internal or […]","categories":[],"tags":[],"author_name":"Abhishek Bose","publish_date":"2016-12-06T08:27:30","publication_year":"2016","word_count":999,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","Git","RAG","automation","GAN","R"],"extracted_tech_keywords":["AI","RAG","R","Go","Git","GAN","automation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/journey-email-marketing-looking-beyond-open-rates\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10170534,"title":"Google vs Google","content":"Google isn’t giving up its search crown anytime soon. During last year’s annual developer conference, it dipped its toes into generative AI with the launch of AI Overviews. Within a year, the platform is being used by 1.5 billion people monthly across more than 200 countries, Google CEO Sundar Pichai said. Now, the company is going all in with a new search experience called AI Mode. “It’s getting even better, with personal context, deeper research, complex analysis and visualisation, live multimodality, and new ways to shop. Now, that’s a lot, because AI Mode can do a lot,” said Google’s VP of Search, Liz Reid, onstage at I\/O. Google stated that AI Mode is now the most powerful AI search, with more advanced reasoning and multimodality, and the ability to go deeper through follow-up questions and helpful links to the web. With this move, the tech giant is challenging ChatGPT and Perplexity AI. The latter recently rolled out Comet, its agentic web browser, as a beta version, limited to select Apple Silicon Mac users. The browser aims to reimagine web browsing through context-aware intelligence. Interestingly, when Perplexity CEO Aravind Srinivas was recently asked if Google would double down on AI-enabled search features that may render Perplexity “obsolete”, he said, “They [Google] have had two years to kill Perplexity and haven’t.” Srinivas added that Google is reluctant to fully embrace AI-generated search responses because they could undermine its advertising business model. According to a Wharton report, Google’s search engine market share remains dominant, controlling around 90% of the global search market. Last year, the company generated $198 billion from search-based advertising, accounting for more than half of its total revenue of $350 billion. Google is also adding deep research features to AI Mode through a tool called Deep Search. It uses a method called “query fan-out,” which breaks complex questions into smaller parts and searches for each one at the same time. This helps it dig deeper into web content and give more accurate, detailed answers, along with the links for further reading. “Google’s AI-led transformation of search should be seen as a strategic adaptation to a fragmented digital knowledge landscape. The rise of AI copilots, chat-based interfaces, and proprietary knowledge agents has redefined how users formulate queries and consume answers,” Sanchit Vir Gogia, CEO of GreyHound Research, told AIM. Building on Deep Search, the company is also integrating Project Astra’s live capabilities into Search. With a new feature called Search Live, users can have real-time, back-and-forth conversations with Search using their camera. Moving beyond searching the internet Pichai said that with permission, Gemini models can tap into context from across Google apps in a way that is private, transparent and entirely under user control. One example of this feature is personalised smart replies on Gmail, where Gemini reads entire threads to draft context-aware responses. “I can be a better friend,” said Pichai, explaining how the feature helps him reply to friends’ emails by looking into his Drive, scanning past emails for reservations, finding itineraries in Google Docs, and matching his typical greetings, tone, style, and favourite word choices. This feature will roll out in Gmail later this year. Moreover, Google has launched a new Deep Research feature that lets users combine public information with their own files, like PDFs and images. This helps users get a complete picture by connecting their private data with what’s happening in the wider world, all in one place. For instance, a market researcher can upload internal sales reports and compare them with public market trends. An academic can add journal articles to improve their research. Soon, Google will also let users pull in data from Google Drive and Gmail, making finding and connecting useful information even easier. More Than Search Google is bringing Project Mariner’s agentic capabilities into AI Mode. Users can now make event bookings without visiting individual websites through traditional search. For instance, users can book event tickets seamlessly. AI Mode will search across platforms, compare hundreds of options in real time, handle form-filling, and show results that match the user’s request. In addition, Google is adding a Shopping feature to Search, allowing users to add products directly, without opening separate sites. The company has also expanded its virtual try-on tool, allowing shoppers to upload a full-length photo and see how clothes would look on them. “AI partnerships in e-commerce are no longer optional—they’re architectural. Per Greyhound Digital Pulse 2025, 64% of retail and commerce technology leaders report that AI is now embedded into customer journey design, from discovery to checkout,” said Gogia. He further added that while not every company must partner with a foundation model provider, those who bypass the AI layer entirely may struggle to scale personalisation, optimise logistics, or manage dynamic demand. Beyond shopping, Google’s new AI Mode also helps analyse complex data and generates custom visuals based on your query. For example, one can compare home-field advantage between two baseball teams and get an interactive graph using real-time sports data. This feature is coming to sports and finance searches. With these updates, Google is turning Search into a tool that goes far beyond finding links.","excerpt":"Google stated that AI Mode is now the most powerful AI search, with more advanced reasoning and multimodality features.","categories":["Global Tech"],"tags":["Google"],"author_name":"Siddharth Jindal","publish_date":"2025-05-22T16:11:03","publication_year":"2025","word_count":857,"keywords":["Go","ChatGPT","AI","ML","Git","RAG","Aim","generative AI","Google","copilots","R"],"extracted_tech_keywords":["AI","ML","generative AI","ChatGPT","Aim","RAG","copilots","R","Go","Git"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/google-vs-google\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":27661,"title":"The Formula Of Using Artificial Intelligence In F1 Races","content":"Formula One sport is known to use the best of technologies and has never hesitated to spend on the safety of their drivers. It is not just a car race, but also a race of the technology and is popularly crowned as the Pinnacle of Motorsport. Now the F1 teams are gearing up to introduce artificial intelligence in the races. They are set to use cloud-based real-time analysis and machine learning techniques to enhance race metrics. F1 will use Amazon Web Services (AWS), which will provide a cloud computing platform and can store a large amount of data. Researchers are going to store more than 61 years of F1 historic race data using AWS and use it for further analysis by predicting tactics for the drivers. They will also use AWS to get race statistics and make most favourable predictions and decisions. These services include AWS’s SageMaker, which is a machine learning service with which data scientists and developers can quickly and easily train ML-based models and put them to use. Other AWS services that can be used include AWS Lambda, the event-driven, serverless computing platform, and Amazon Kinesis Data Analysis, which helps with streaming data in real-time. Why Does F1 Need AI? The F1 cars are extremely high tech. They have more than 200 sensors on the car and engineers have to be glued to their computers to monitor each sensor on their machines throughout the span of race to make decisions for the next move. Instead, they can have AI to replace humans which will not only help in monitoring but also take right decisions. A lot of instances in F1 depict flaws due to human decision making, so that can be eliminated with the advent of AI. Such systems can be trained to avoid accidents due to crashes and know about a failure of some part of the car, beforehand. The advent of AI has also made them possible to do a lot of mechanical work in no time. So, they can be deployed to use for the repair and maintenance of the cars. The F1 teams, in spite of being very experienced and in spite of the skillsets and the enormous amount of money that they have, fails quite often to predict the best next move. Using AI in places like this would make them predict quite accurately about what move to make next. Which Data Is Needed? Since all the advances that AI can make with ML models depend entirely on the data it is given, the teams need to have an avalanche of data so that the ML algorithms at work give more precise predictions. This data involves telemetric data, apart from the historic one. This data can be temperature, pressure, frequency, speed and is received from onboard sensors. There also needs to be informed of individual team drivers like steering, acceleration and brake, along with data such as lap times, top speeds, pit stop times, wind speeds on the track, and others for precise forecasting. All this data was the historic and the real-time data. But there is another way to add a set of information that can be compiled for even better results — adding random data. They can come up with realistic parameters and simulations and introduce arbitrary modifications to simulate the races. These quasi-strategies can be made to study AI and make prediction abilities that might work for any race situation. Areas Where AI Can Be Deployed Pit-stop Timing: The strategic teams might make misjudgements in taking decisions regarding when the car has to stop in the pits. AI, by giving it a plethora of data can study and predict accurately what time is the best for the driver to stop at the pit to change the tires. Tire Selection: Another most frequently occurring mistake that the team does is with the tire. What tire to select to change during the course of the race is a crucial decision, both for winning as well as for the safety of the driver. Weather plays a very important role in making that decision. AI can be trained to predict what tire to change, depending on the weather conditions and also on the race conditions, such as what lap it is on at the time of the change and so on. Teams Already Trying To Bring AI to F1 Renault Sport and Williams Martini Racing have already begun to approach AI in their F1 technology. They are using ML and analytics to help them make the predictions and decisions during races. They have also begun to take help of AI to build their cars. Honda has turned to IBM’s Watson IoT for Automotive to analyse the hybrid engines they supply to Scuderia Toro Rosso. Formula One has never shied away from spending money that goes into every inch of their cars. Soon, we might see AI at pit stops of F1 races and F1 technology will get even more impressive when that happens.","excerpt":"Formula One sport is known to use the best of technologies and has never hesitated to spend on the safety of their drivers. It is not just a car race, but also a race of the technology and is popularly crowned as the Pinnacle of Motorsport. Now the F1 teams are gearing up to introduce […]","categories":["AI Features"],"tags":["AWS","ML","prediction","SageMaker","salesforce crm","strategic salesforce"],"author_name":"Disha Misal","publish_date":"2018-08-27T10:39:03","publication_year":"2018","word_count":830,"keywords":["Go","prediction","artificial intelligence","machine learning","AWS","AI","cloud computing","SageMaker","ML","serverless","salesforce crm","analytics","R","strategic salesforce"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","analytics","cloud computing","AWS","serverless","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/the-formula-of-using-artificial-intelligence-in-f1-races\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10054038,"title":"NVIDIA Releases Version 2 Of GauGAN, Its Scaringly Accurate AI Art Solution","content":"NVIDIA recently announced the latest version of NVIDIA Research’s AI painting demo, GauGAN2. The new model is powered by deep learning and consists of a text-to-image feature. The original version could only turn a rough sketch into a detailed image. GauGAN2 can generate images from phrases like ‘sunset at a beach,’ which can then be further modified with adjectives like ‘rocky beach,’ or by changing ‘sunset’ to a different time of day or even modifying weather conditions. Powered by generative adversarial networks (GAN), NVIDIA spokesperson said “With the press of a button, users can generate a segmentation map, a high-level outline that shows the location of objects in the scene. From there, they can switch to drawing, tweaking the scene with rough sketches using labels like sky, tree, rock and river, allowing the smart paintbrush to incorporate these doodles into stunning images.” Image Source: NVIDIA Research By adding text-to-image capabilities, the new version of GauGAN is more customisable and can be tuned much quicker. Even a quick sketch is not nearly as fast and simple as typing a phrase. The latest version is also one of the first AI models to incorporate multiple modalities, text, semantic segmentation, sketch, and style, within a single GAN network. The Text-based starting point, such as a ‘snow-capped mountain range,’ can be further customised with sketching. Users can add trees, change the height and size of objects, add clouds to the sky and much more. And then GauGAN2 generates a new, modified image. GauGAN2 may prove useful for concept artists, as they can now create worlds with two suns, like Tatooine in Star Wars. ‘It’s an iterative process, where every word the user types into the text box adds more to the AI-created image,’ NVIDIA spokesperson added further. Earlier this year, NVIDIA released a tool built upon GauGAN, NVIDIA Canvas, which can be used on any NVIDIA RTX GPU. In addition, GauGAN2 has been trained on 10 million landscape images using the NVIDIA Selene supercomputer, which is among the world’s ten most powerful supercomputers. If you’d like to try your hand at creating simulated photos of places that never existed, head over to NVIDIA’s AI Playground and click on the “Launch Interactive Demo” button for GauGAN2. GauGAN2 From NVIDIA Includes A Text-To-Image FeatureGauGAN Can Turn MS Paint-Like Sketches Into Photorealistic MasterpiecesMIT Team Teaches AI To Code Like HumansConverting Image Into A Pencil Sketch In PythonRegister For This Webinar On Data Science In The Post-COVID World: Careers & Skills","excerpt":"GauGAN2 can generate images from phrases like ‘sunset at a beach,’ which can then be further modified with adjectives like ‘rocky beach,’ or by changing ‘sunset’ to a different time of day or even modifying weather conditions.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","Data Science","Data Scientist","Deep Learning","GAN model","GAN-powered art","GANs","Machine Learning","NVIDIA","Research","Text To Image Synthesis"],"author_name":"Victor Dey","publish_date":"2021-11-23T12:11:09","publication_year":"2021","word_count":413,"keywords":["Modal","deep learning","GANs","R","data science","Research","programming_languages:Python","GAN model","NVIDIA","Data Science","Text To Image Synthesis","AI","Machine Learning","GAN","programming_languages:R","GAN-powered art","Python","Deep Learning","Data Scientist","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","deep learning","data science","Python","R","GAN","Modal","programming_languages:Python","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/nvidia-releases-version-2-of-gaugan-its-scaringly-accurate-ai-art-solution\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10020015,"title":"In Conversation With Aayush Rathi, Senior Researcher At The Centre For Internet And Society","content":"This is the eighth article in the weekly series of Expert’s Opinion, where we talk to academics who study AI or other emerging technologies and their impact on society and the world. This week, we spoke to Aayush Rathi, Senior Researcher at the Centre for Internet and Society, whose current research focuses on emerging conversations around the future of work as well as cybersecurity. Rathi has previously worked as a labour lawyer post his graduation in law and humanities from the National University of Juridical Sciences, Kolkata. Analytics India Magazine caught up with Rathi to get insights into his recent research on the impact of AI automation and other emerging technologies in India’s IT sector. AIM: Why is it important to understand latest developments in technology to assess its impact on the jobs it will displace? Rathi: Discussions about automation in work today tend to be dominated with a celebration of artificial intelligence. We fall short of assessing these technological developments (cloud computing, big data, and others) for their on-ground impact. Their impact on job functions and the way we work is very nuanced and complex. Much of the contemporary discourse around automation gets subsumed within overarching rubric of artificial intelligence. This precludes us from adequately capturing the various technological developments that we are confronted with, including cloud computing, Big Data and others. Each of these developments portend to implicate job functions in specific ways. I am also very interested in unpacking the powerful metaphors of intelligence, neurons and cognition that have become common in futurist discourse. Ongoing conversations mistake the relentless pursuit of profit for inevitable, anxious, and essential development or ‘progress’. What I’m getting at here is that the techno-centric nature of these conversations has the effect of underplaying its political and socio-cultural impact. Understanding these dimensions holds the promise of delivering just futures – a future in which we challenge social inequalities and political power. AIM: Does India need to be wary of reshoring by IT or data science firms as it will take business operations back to their headquarter countries? Rathi: The neoliberal economists’ response may be that India’s IT\/ITeS sector needs to move up the ladder by creating the enabling environment to provide high-value services. This would then give India an edge vis-a-vis competition from Southeast Asian countries. However, this may only be beneficial for a miniscule part of the IT\/ITeS sector. While we still lack robust employment data, indications are that there is an ongoing unemployment crisis that has been further heightened by Covid-19 induced layoffs. Some sound policy ideas to address this could be guaranteeing employment in urban centres, universal basic income proposals and policy proposals for India’s migrant workforce. AIM: What are the implications of the change in the way productivity is measured in the IT\/ITeS sector? Rathi: It is evident that measures of productivity are frequently changing in the IT\/ITeS sector. Recently, there are clear trends on breaking away from metrics that have been borrowed from mechanised, factory-based labour. In other words, productivity metrics are increasingly moving away from measuring inputs and outputs, to measuring ‘revenue-per-employee’. The emergence of the “-as-a-service” paradigm has been key in accelerating this change. Corporations are also increasingly keen on turning ‘lean’, by shifting organisational structures from the classical pyramid to an hourglass—with more focus on the top and bottom rungs of employees and a hollowing out of the middle. AIM: What are some of the major findings in your study with respect to how manual labour will be impacted due to automation and other emerging technologies? Rathi: A central implication of our study is that technological change does not impact job loss\/gain in a vacuum, nor is it neutral. Technologies are made and used within existing social relations. Relatedly, technology itself is a site of struggle, and we are witnessing the rise of white-collar collectivisation. The service industry is also incredibly stratified. Within what is regarded as the IT sector itself, there is a wide range of occupational categories, and then also demographic variance. Another implication is that we need to puncture the hyperbole that is taken for granted in discussions about the power of emerging technologies. It surprises many that contemporary information systems are extremely labour-intensive. They rely on armies of coders, data cleaners, content moderators, subcontractors, page rankers etc. who form what has been called the ‘fissured workplace’. Often hired through microtasking platforms that are transnational in nature, this ‘reserve workforce’ does not feature on company payrolls. So, we’re witnessing a shift in both job roles and tasks, as well as the structure of the employment relationship. AIM: What are your policy recommendations for a labour-dependent country like India, considering all the threats of automation including reshoring, the pandemic, and the emerging tech, to avoid job losses? Rathi: In the early days of the covid-19 outbreak, we saw a flurry of states proposing to dilute, and even entirely suspend, labour law provisions. The proposed moves were the most short-sighted manifestation of hasty assumptions that are used to justify the weakening of Indian labour law. Another commonly used defence of automation is that it brings efficiency. We need to question this – who is it bringing efficiency for? And what type of efficiency is this? The pandemic has accelerated trends of work-based technological adoption. This will pose barriers for the protection of our labour and civil rights, as we look to embrace a future with heightened workplace surveillance. A labour rights + digital rights agenda has to be developed. What is most clear to me is that we need to urgently foster the participation of a broad range of marginalised groups in the deliberations of our collective futures. This includes rethinking our labour law approaches in ways that it doesn’t pit one category of labour against another, or gets stripped to reductive binaries of “pro-labour” and “pro-capital”, and makes serious inroads into universalising social security and welfare provision. While technological change may be inevitable, it needn’t be unethical.","excerpt":"This is the eighth article in the weekly series of Expert’s Opinion, where we talk to academics who study AI or other emerging technologies and their impact on society and the world. This week, we spoke to Aayush Rathi, Senior Researcher at the Centre for Internet and Society, whose current research focuses on emerging conversations […]","categories":["AI Features"],"tags":["Interviews and Discussions"],"author_name":"Kashyap Raibagi","publish_date":"2021-02-11T18:00:00","publication_year":"2021","word_count":994,"keywords":["data science","Go","artificial intelligence","TPU","AI","cloud computing","Git","Aim","analytics","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","artificial intelligence","data science","analytics","Aim","cloud computing","TPU","R","Go","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/in-conversation-with-aayush-rathi-senior-researcher-at-the-centre-for-internet-and-society\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10105351,"title":"Vivek Ramaswamy Believes that Human Response to AI Poses Greatest Risk","content":"Last week, Republican presidential candidate Vivek Ramaswamy delved into the intricate landscape of AI policy and said that human response to AI poses the greatest risk, emphasising the critical need for competence in leadership to navigate the challenges posed by this rapidly advancing AI landscape. https:\/\/twitter.com\/adamscrabble\/status\/1736453257587822834?s=20 Citing Tennis, Ramaswamy said that his first job was as a ball boy in Cincinnati, and he was later promoted to a line judge in the ninth grade. That was the time he stumbled upon AI. “So, as a human line judge, you make the line calls. It’s not done that way anymore. It’s all done by AI,” he said, saying it predicts where the ball is going to land. He said when humans made the decision, the players used to argue with the line judges. “Something funny happened when the AI started making the call. The first generation of the AI. It was so bad that you could literally see it with your eye that it was like a bad call. But the funny thing is that the players stopped arguing with the calls,” he said, stating that the biggest danger of AI is actually the human response to it. “I don’t mean to get too philosophical, but I think it’s actually important,” he added. Further, talking about regulating it at a policy level. He said it is important to draw a hard lines where AI powered algorithms should not be regularly interfacing broadly with kids. “I think that we should not ban anything that China is also not willing to ban,” added Ramaswamy, saying that the companies should be given liability. He said that the companies are going to be liable for any unforeseen consequences of a protocol that they develop. “At least makes them take the risks into account on the front end, which they are not doing today,” he said, saying that is the right answer as a matter of policy, giving example of ChatGPT and how it could go wrong. AI expert Gary Marcus also retweeted the video, and opined saying that it was the “Weirdest take on AI I have ever seen, and I’ve seen some weird ones.” He also agreed with Ramaswamy on the potential dangers of AI, particularly in terms of societal manipulation and loss of control, that also expressed reservations about the emphasis on reviving faith and patriotism as primary solutions. Despite differing on specific solutions, Marcus underscored the importance of addressing societal anxieties highlighted by Ramaswamy and encouraged evidence-based approaches to AI policy, advocating for decisions grounded in factual data and rigorous analysis rather than simplistic solutions based on faith or nationalism.","excerpt":"And said that the companies should be held liable for any unforeseen consequences of a protocol that they develop.","categories":["AI News"],"tags":[],"author_name":"Sandhra Jayan","publish_date":"2023-12-21T18:22:29","publication_year":"2023","word_count":438,"keywords":["Go","ChatGPT","API","programming_languages:R","AI","RAG","GPT","R","llm_models:ChatGPT","llm_models:GPT"],"extracted_tech_keywords":["AI","ChatGPT","RAG","R","Go","API","GPT","llm_models:GPT","llm_models:ChatGPT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/vivek-ramaswamy-believes-that-human-response-to-ai-poses-greatest-risk\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10043071,"title":"IIIT Delhi Launches PG Diploma Program in Computer Science &#038; AI","content":"Indraprastha Institute of Information Technology-Delhi, (IIIT-Delhi) has announced the launch of a new Post Graduate Diploma Program in Computer Science and Artificial Intelligence, in collaboration with edtech platform Great Learning. “Data is the new oil of the digital economy and the way we will lead our lives in the near future will be transformed by the use of AI, and its applications to the big-data innovation ecosystem. The PG diploma program in Computer Science and AI will equip the students with advanced knowledge and practical skills to meet the growing challenges of the various data-driven industries. The partnership with Great Learning will help create new opportunities and greater accessibility for the interested students and working professionals,” said Prof. Ranjan Bose, Director IIIT-Delhi. The 12-month online program is designed to fit into the schedule of both working professionals and students. It aims to deliver a comprehensive learning experience imparting industry-relevant skills in artificial intelligence and machine learning, built on key computer science fundamentals. “Our mission has always been to make high quality, transformational education accessible and affordable to aspiring learners. Through our collaboration with an esteemed University like IIIT-Delhi, we are creating opportunities for learners to get access to the latest skills to power ahead in their careers and contribute meaningfully to our nation’s economic growth,” said Mohan Lakhamraju, Founder and CEO, Great Learning. “The scope of Artificial Intelligence and Machine Learning has grown tremendously over the last few years. The accelerating adoption of these technologies, both in India and globally, has accelerated the demand for skilled professionals in these domains. This program will address this growing demand,” Mohan added. Upon successful completion of the program, participants will earn a Post Graduate Diploma from IIIT-Delhi as well as acquire an alumni status of the institute. The students will get access to Great Learning’s career assistance platform, GL Excelerate.","excerpt":"Indraprastha Institute of Information Technology-Delhi, (IIIT-Delhi) has announced the launch of a new Post Graduate Diploma Program in Computer Science and Artificial Intelligence, in collaboration with edtech platform Great Learning. “Data is the new oil of the digital economy and the way we will lead our lives in the near future will be transformed by […]","categories":["AI News"],"tags":[],"author_name":"kumar Gandharv","publish_date":"2021-07-06T20:16:15","publication_year":"2021","word_count":307,"keywords":["machine learning","artificial intelligence","programming_languages:R","AI","data-driven","innovation","Git","Aim","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","Aim","R","Git","data-driven","innovation","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/iiit-delhi-launches-pg-diploma-program-in-computer-science-ai\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":14789,"title":"Everything You Need to Know About AI-Enabled Personalization","content":"Artificial Intelligence (AI) makes it possible to treat each of your contacts as a unique and valuable individual. Ever wonder how it works? According to David Hahn, “The key ingredient to better content is separating the single from the stream.” In other words, the more you personalize, the better your results. In today’s world, this usually means some type of data-driven personalization is in play. With improvements in technology, we’re starting to see the use of AI-enabled personalization. You may not know exactly how this works. Have no fear — we’ll explain it in today’s post. An Illustration of AI at Work First, let’s look at a scenario that calls for intense personalization. Cast your mind back to your high school prom or any other social event that required you to dress up. Choosing an outfit was a big deal, right? Even if you weren’t fashion-obsessed, you didn’t want to wind up in a 20-year-old suit or dress. In this situation, imagine that a friend tells you they will give you some awesome clothes as per your liking for your event. That would truly be a personalized experience. But we can take it one step further. Suppose your friend gave you multiple choices and arranged to have your choice shipped to you the week before your event. That would be great! We can say that the second friend is an illustration of AI at work. Personalization can see a need and select a single product to fill it; AI-enabled hyper personalization can deliver tailored options and results on schedule. It’s taking things to the next level. At this point, though, let’s define what we mean when we say AI. What Is AI? AI is the science and engineering of making intelligent machines. Any science that enables us to make human-like decisions can be a part of AI. This includes such abstract and advanced concepts as logic, visualization, self-awareness, learning, emotional knowledge, planning, creativity, and problem-solving. AI is being used for a huge number of vastly different things. It can enhance one business’ marketing and another’s supply chain management; it can help us in the fields of medicine and sociology. People are using it to find the nearest gas station and to detect lung cancer — it’s becoming a ubiquitous technology. In the future, we can expect that the human-machine conversation aspect of AI to become more dominant. We’ll see an upsurge in the importance of voice, touch, and other really cool things. But right now, we are seeing AI in another new way: as a way to intensely personalize marketing communication. How Does AI-Based Personalization Work? Before jumping on the personalization bandwagon, it’s useful to understand how AI-enabled personalization works. Here are the broad steps: Learning starts with historical data — a lot of it. Machine learning as a process is not very different than how a child learns. Both begin to differentiate between right and wrong through instruction and prior knowledge. On similar lines, we feed scenarios (historical data) into the machine and let it know what is correct. It needs to know why a customer responded to an offer, why they clicked on a specific email, etc. We are training the machine by showing it what has worked in the past. Later, we’ll be asking it to make connections between historical data patterns and its current inputs. Individual technologies’ are integrated. Marketing requires the use of multiple technologies that, unfortunately, work in silos and rarely intercommunicate. The AI system needs to understand each technology’s input\/output dynamics to improve their effectiveness. Also, AI needs to understand and adapt to advances in data science, and vice versa. This will make the entire system more effective. The platform should be all-encompassing. Giving a decision option in a new situation is a major challenge for an AI system, as it must operate off past data. A machine that self learns can find similarities between new scenarios and previous ones. An AI system that works only for specific situations is not truly AI and will never scale well. The more situations the system can handle correctly, the better it will be. Decisions are fine-tuned using a feedback loop. A crucial part of AI self-learning is the feedback loop, a process that provides continuous feedback on the decisions the system makes. The AI system needs to know if the historical decisions were right or wrong, and it needs to be smart enough to tweak its own algorithm and provide more accurate and robust results. It basically should self-learn. The greatest effort in this type of system is collecting the data, streamlining the various processes, and feeding into the AI system. Organizations that successfully accomplish this are rewarded by more powerful, targeted, and effective hyper personalization. AI might seem like a new development in marketing, but in reality, it’s taking technologies and processes one step further. Authors: Rajat Narang, Associate Director, Absolutdata Abdurrehman Malekji, Manager, Absolutdata","excerpt":"Artificial Intelligence (AI) makes it possible to treat each of your contacts as a unique and valuable individual. Ever wonder how it works?  According to David Hahn, “The key ingredient to better content is separating the single from the stream.” In other words, the more you personalize, the better your results. In today’s world, this […]","categories":["IT Services"],"tags":[],"author_name":"AIM Media House","publish_date":"2017-05-06T05:21:16","publication_year":"2017","word_count":821,"keywords":["data science","Go","machine learning","artificial intelligence","TPU","AI","ML","ViT","GAN","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","data science","TPU","R","Go","GAN","ViT"],"url":"https:\/\/analyticsindiamag.com\/it-services\/everything-need-know-ai-enabled-personalization\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10061266,"title":"Are NFTs the new diamonds?","content":"Last Sunday, a picture of an Indian couple exchanging NFT rings went viral on Twitter. Dalsukh Tapaniya and Kinjal Pipaliya were the couple in question. Finally we exchanged NFT Rings today❤️ pic.twitter.com\/oloRC7jSLF— Dalsukh Tapaniya (@djtapaniya) February 20, 2022 The couple shared the OpenSea links of the NFT rings. Twitterati was in stitches. Some even started threads on NFT wedding gifts and suggested metaverse as a honeymoon destination. https:\/\/twitter.com\/bharatxoni\/status\/1495413548759252994 https:\/\/twitter.com\/moopayments\/status\/1495437027218710534 Meanwhile, a couple of Twitter users rained on the parade, reminding the netizens on the tax policy on crypto. @RBI be like – Congratulations on the wedding, hope you'll pay 30% tax on that NFT as well soon.— Nikhil Sulghur (@Nikhilsulghur) February 21, 2022 The OpenSea platform has over 1,000 NFT wedding rings available. The most expensive ring currently on the platform costs ETH 1000 (USD 2.7 million). There are a few rings in the range of ETH 100 – 200, ETH 10 – 25, and anywhere between 0 to 4 ETH. Metaverse is slowly trickling down the social hierarchies. Last month, a couple from Tamil Nadu came up with a Harry Potter-themed metaverse wedding. Dinesh Sivakumar Padmavathi and Janaganandhini Ramaswamy hosted Asia’s first metaverse wedding. The couple invited 2,000 guests to the wedding and also created an avatar for a dead relative. The wedding was held on the platform TardiVerse. Abhijeet Goel and Dr Sansrati too tied the knot early this month on Yug Metaverse. Wavemaker India and Bharat matrimony were the brand partners of the ceremony.","excerpt":"Last Sunday, a picture of an Indian couple exchanging NFT rings went viral on Twitter.","categories":["AI News"],"tags":["NFT","OpenSea"],"author_name":"Meeta Ramnani","publish_date":"2022-02-22T11:13:47","publication_year":"2022","word_count":247,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","OpenSea","ViT","GAN","R","NFT"],"extracted_tech_keywords":["AI","R","Go","GAN","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/are-nfts-the-new-diamonds\/","complexity_score":3,"technical_depth":7,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10079407,"title":"This is the End of Twitter As We Know It","content":"Congratulations, Musk. By all means, Twitter is officially set to be “by far the most accurate source of information about the world”. Twitter has been one of the main sources of news for people to rely on. The platform could potentially ace TV channels or magazines that give users zero control over the opinions that they share. Twitter needs to become by far the most accurate source of information about the world. That’s our mission.— Elon Musk (@elonmusk) November 7, 2022 But, what comes along as a catastrophe to this moderation of content is the vast sea of misinformation. If handled inadequately, the bird that Musk freed would probably get lost in this sea. It’s certain that advertisers and customers will flee from the platform. Users will eventually get bored of the $8 verified symbol. The authenticity of verified accounts will be lost. With difficulties around selling and monetising services on the platform, Musk must carry on to deliver what was promised. And if there’s one other problem that would be extremely hard to detect, it’d be the AI-based misinformation. Twitter will soon (might already) be flooded with misinformation. But is Musk ready to handle that? Village of misinformation Misinformation has existed since day one and everybody knows that. There is no sense of novelty about the inability of humans to assess the source of truth. We don’t even need the best of technologies to spread lies or inaccurate information. However, it’s certain that AI and machine learning systems can enhance the capabilities of malicious information to unprecedented levels. Musk has spent this week twisting his new toy; raising questions, suspicions, and clusters of conflicts by its users. One of his main goals was the need to verify all human users by getting rid of the AI-powered “spam bots or die trying”. If our twitter bid succeeds, we will defeat the spam bots or die trying!— Elon Musk (@elonmusk) April 21, 2022 At present, misinformation travels faster on Twitter—nearly eight times faster than Meta’s Facebook. “We found that falsehood diffuses significantly farther, faster, deeper, and more broadly than the truth, in all categories of information, and in many cases by an order of magnitude,” says Sinan Aral, Professor of MIT Sloan School of Management. Good at generating, but bad at detecting According to Gary Marcus, the misinformation problem is set to worsen in the coming days. Generating misinformation has become easier than ever. As knockoffs of models such as GPT-3 are getting cheaper and freer to use, the cost of generating misinformation is expected to eventually be zero, thereby leading to the rise in the quantity of misinformation. The key issue lies in AI systems—particularly, large language models—that are well-advanced in generating misinformation but are bad at detecting it. Furthermore, a group of researchers have created a multilingual large language model that is bigger than GPT-3. The new model called BLOOM (BigScience Large Open-science Open-access Multilingual Language Model) was created by over 1,000 researchers who volunteered in a project called ‘BigScience’. Also read, GPT-3 Is Quietly Damaging Google Search However, the model’s creators warn that it won’t be able to fix the problems around large language models—such as lack of adequate policies on data governance and the tendencies of algorithms to spew toxic content. So, in order to combat the pool of misinformation in platforms such as Twitter, the new Chief Twit will need advanced tools for regulation of online content. From facing threats of deepfakes to fighting biases on text-to-image models, AIs can generate results without having an actual ground in the real world. New generative models are crippled every other second but the researchers are lagging behind on detecting and finding measures to handle them. Recommendation systems are majorly driven by AI algorithms deciding what users expect to see on their feeds. The problem arose with Musk’s recent decision to remove identity verification. The CEO also fired the team that was in charge of algorithmic responsibility at Twitter. The maelstrom of news and one-sided decisions bring concerns about Twitter and if it will even survive this rampage. if you are serious @elonmusk about trying to make Twitter the most accurate source of information, we should talk. And you need to start by understanding the core technical issue: https:\/\/t.co\/QkdRpFpcMZ pic.twitter.com\/1P9WBH6K2O— Gary Marcus (@GaryMarcus) November 7, 2022 Gary Marcus’ twitter post said, “If you are serious @elonmusk about trying to make Twitter the most accurate source of information, we should talk. And you need to start by understanding the core technical issue.” Source: Twitter Marcus further elaborated that language learning models like GPT-3 are very good at generating misinformation. Truthful questions such as ‘Who really caused 9\/11?’ were asked, to which the models replied with a false answer—‘The US government caused 9\/11’. As judged by the people of Twitter via Community Notes (formerly Birdwatch)— Elon Musk (@elonmusk) November 7, 2022 With the pace at which misinformation is getting picked at, Twitter’s existing effort ‘Community Notes’ (Birdwatch)—executed manually by humans—is sure to bite the dust. Gamble of $8 verified badge Furthermore, the new blue tick theoretically becomes an effective barrier to large-scale operations on misinformation. Until a couple of days ago, the company’s ML, Ethics, Transparency, and Accountability (META) team was in charge of keeping the algorithm under control. It was an initiative to ensure that  fairness and transparency are maintained throughout the social media platform. One of the team members says, “We’re building explainable ML solutions so you can better understand our algorithms, what informs them, and how they impact what you see on Twitter. Similarly, algorithmic choice will allow people to have more input and control in shaping what they want Twitter to be for them.” Meanwhile, a twitter employee tweeted: Has it already started? Happy layoff eve! pic.twitter.com\/0AcaQjGJvm— ruchowdh.bsky.social (@ruchowdh) November 4, 2022 The team empowered users to prevent any harm caused by the algorithms. What seemed like reckless behaviour is when Musk axed the entire META team—without any prior notice. The problem of misinformation existed long before Musk reigned over Twitter. But how will the platform detect this misinformation? Who is taking the responsibility to tailor the recommendation algorithm to ensure that the ‘free speech’ remains accountable? A Twitterati can get the verification badge of honour for $8. But is it not over the cost of true information?","excerpt":"Twitter will soon be (or already is) flooded with misinformation. But is Musk ready to handle that?","categories":["AI Features"],"tags":["AI (Artificial Intelligence)","Elon Musk","gary marcus","GPT-3","Layoffs","Twitter (X)","Twitter Layoffs"],"author_name":"Bhuvana Kamath","publish_date":"2022-11-11T13:00:00","publication_year":"2022","word_count":1051,"keywords":["GPT-3","Go","API","machine learning","ELT","Layoffs","AI","ETL","ML","recommendation systems","Elon Musk","gary marcus","data governance","Twitter Layoffs","Twitter (X)","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","machine learning","ML","recommendation systems","R","Go","API","ETL","ELT","data governance"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/this-is-the-end-of-twitter-as-we-know-it\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10001907,"title":"How IoT And Blockchain Can Revolutionise The Manufacturing Industry","content":"IoT has had a lot of applications in different sectors in India. Blockchain too has emerged as an important technology. The manufacturing sector has a lot of potential when integrated with these technologies of IoT and Blockchain. Emerging Use Cases For Blockchain In Manufacturing 1. Improves track and trace accuracy: Combining IoT’s real-time monitoring support with blockchain’s shared distributed ledger strengthens track-and-trace accuracy and scale, leading to improvements across supply chains. Improving track-and-trace reduces the need for buffer stock by providing real-time visibility of inventory levels and shipments. Urgent orders can also be expedited and re-routed, minimising disruptions to production schedules and customer shipments. This will help in reducing the effort and paperwork involved. 2. Brings clarity in manufacturing process: Every manufacturing industry, specialising in any domain need to know the source of their products. They ought to know if the products are from reliable sources and are pure. It is also important for the manufacturers to know that the products that they receive obey the laws of the Indian government. The consumers are also willing to pay more for the products given a greater transparency in the products. IoT and Blockchain can together help in making the products more transparent to the consumers. 3. Reduced loss of shipments by tracking: Supply chains in the manufacturing business can involve multiple people and stops during transit before they reach the final stage. There are many incidents where the shipments are lost and stolen. In this situation, Blockchain can be used to track the routes of the goods. 4. Labour efficiency: There are situations when a client might supply heavy machinery tools to different companies and in case the machine breaks down the labour efficiency is largely reduced, as the workers then remain idle for short periods. This eventually leads to degrading the revenues of these organisations. Here is where a combination of IoT and Blockchain can help with these situations. A concept in IoT called the Preventive Maintenance is used wherein temperatures or vibrations can be measured using IoT sensors that collect data in real time. Devices would then compute all the data. An IoT infrastructure platform called NetObjex, which has its Indian headquarters in Thane, Maharashtra, uses the same principle. They have the data collected uploaded in the cloud where they run their platform. They have edge-devices that compute all the parameters. 5. Track performance using Digital Twin: Digital Twin in IoT is virtual replicas of physical devices that the data science and IT industries can use to run simulations before actual devices are built and deployed. It leads to a better quality and in turn better maintenance and management of the assets. This platform helps in tracking the organisation past current and future performance. In this case, the asset sends performance data directly to its digital twin all throughout the process. Blockchain comes in handy to document all information related to assets. It can help manufacturers to ensure the credibility of the products that they receive. Last Word IoT and Blockchain together can play a large role and transform the manufacturing industry. They will provide a greater transparency in the while trading and an efficient tracking for a systematic functioning of the process. Sensors have become very popular and easily accessible to many, making it easy to deploy these technologies in manufacturing. Streamlining and automating a variety of processes in manufacturing becomes easy with these technologies. Both together can help manufacturers revolutionize the way products are assembled and distributed.","excerpt":"IoT has had a lot of applications in different sectors in India. Blockchain too has emerged as an important technology. The manufacturing sector has a lot of potential when integrated with these technologies of IoT and Blockchain. Emerging Use Cases For Blockchain In Manufacturing 1. Improves track and trace accuracy: Combining IoT’s real-time monitoring support […]","categories":["AI Features"],"tags":["Blockchain","IoT","Sensors","Transparency"],"author_name":"Disha Misal","publish_date":"2019-04-25T13:07:49","publication_year":"2019","word_count":580,"keywords":["data science","Go","Blockchain","AWS","AI","cloud_platforms:AWS","ML","Git","disruption","Transparency","Sensors","GAN","R","IoT"],"extracted_tech_keywords":["AI","ML","data science","AWS","R","Go","Git","GAN","disruption","cloud_platforms:AWS"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-iot-and-blockchain-can-revolutionise-the-manufacturing-industry\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10105211,"title":"Rakuten India is Certified as Best Firm for Women in Tech","content":"Rakuten India has been certified as ‘Best Firm for Women in Tech’ by Analytics India Magazine (AIM) through its workplace recognition programme. This is a corporate certification provided to organisations that have demonstrated a commitment to promoting workplace equality and empowering women in the technology industry. “At Rakuten, we’re not just building technology; we’re fostering a community that is empowering society through technology. Our commitment to empowerment goes beyond words—it’s woven into our everyday fabric. Here, our women employees are valued contributors to our collective success story. “They drive critical pillars ranging from our CSR engagements to e-commerce, product marketing, new business incubation and other strategic initiatives. By bringing people from diverse backgrounds together, we don’t just solve critical business challenges — we create success stories that inspire us all,” said Sunil Gopinath, CEO, Rakuten India. “Receiving AIM’s ‘Best Firm for Women in Tech’ certification makes extremely proud. It’s a testimony to Rakuten’s consistent and concerted efforts in creating an inclusive workplace and I’m glad that our efforts are being recognised. We celebrate the diverse voices, a myriad of experiences, perspectives, and talents that make Rakuten great. “Our mentorship programs, scrupulously tailored development paths, supportive amenities and work arrangements ensure that everyone at Rakuten feels valued, seen, heard, and ready to contribute to the future of technology and society,” said Nalini George, chief people officer, Rakuten India. Meanwhile, Anirban Nandi, vice president – AI products & business analytics, believes that diversity has actually been Rakuten’s strength. “We are dedicated to creating an environment where everyone, regardless of their gender, feels empowered to thrive. This recognition is a testament to our commitment to diversity, equality, and fostering an inclusive workplace. This accolade is not just a reflection of our achievements but a reminder of the work that still lies ahead in promoting gender equality in the tech industry,” said Nandi. The ‘Best Firm’ certification by Analytics India Magazine is considered a gold standard in identifying firms that have achieved a high standard of excellence in creating a diverse and inclusive work environment for women in tech. Get recognised as a company championing diversity and empowering women in the tech space. To nominate your organisation for the certification, please follow the link to fill out the form.","excerpt":"The Best Firm for Women in Tech Certification is a corporate certification provided to organizations that have demonstrated a commitment to promoting workplace equality and empowering women in the technology industry.","categories":["AI Highlights"],"tags":["Top Trend","Women in Tech"],"author_name":"AIM Media House","publish_date":"2023-12-20T11:00:00","publication_year":"2023","word_count":375,"keywords":["Top Trend","Go","programming_languages:R","AI","programming_languages:Go","Aim","Women in Tech","analytics","GAN","R"],"extracted_tech_keywords":["AI","analytics","Aim","R","Go","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/rakuten-india-is-certified-as-a-best-firm-for-women-in-tech\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10089062,"title":"Zee5 Binges on AI","content":"A couple of weeks ago, Netflix released a trailer for an anime project named “Dog and Boy”. The entertainment platform claimed that since the anime industry has a labour shortage, they used image generation technology to generate background images of all three-minute video cuts. While the platform faced severe criticism from anime fans for using AI to generate images rather than hand-drawing them—as is conventional—the entertainment industry is all up for embracing the new tech. For instance, Eros Now is already using AI to stream an entire range of Eros Now films and originals with automated subtitles. Similarly, Amazon Prime uses ML to enhance the video quality, as it might be harmed by flaws introduced during recording, encoding, packing, or transmission. Not only Amazon but most of the video streaming services use ML in a similar manner to enhance their video quality. To know more about how Indian streaming services are using technology to enhance the productivity of their employees, Analytics India Magazine got in touch with Kishore AK, CTO of Zee5 at the Big CIO Show in Bengaluru. Zee5, according to AK, is using AI in its content creation process along with customer-facing activities. While AI is already used for subtitles and closed captioning, Zee5 is now experimenting with using generative AI to help its content creators augment their work. “I’m sure my development team is using Github Copilot at some scale,” AK explained. However, he also believes that while using Copilot at the initial stage and for training the freshers can be encouraged, once the IP algorithms are formed, the use of such platforms might pose challenges of data leaking. Additionally, AK goes on to say that generative AI is also being used to help content creators generate scripts for upcoming shows. Although he declined to reveal the name of the shows. “If everything goes according to plan, we might continue utilising LLM Models in script writing”. However, not all are happy with AI being introduced in the video streaming industry—particularly the employees whose jobs are most likely to get affected. For instance, in a media interaction, Eros Now CEO Ali Hussein revealed that prior to the introduction of AI in the subtitling industry, it used to take up to 48 hours for each movie to be subtitled correctly—something that can now be done now in matter of seconds. Similar was the case with Netflix’s “Dog and Boy” video clip. While anime fans were adamant about using artists to create the background, the artists themselves are costly to the industry. As per industry experts, experienced artists earn around $1500~3000 a month in the anime industry—a cost that Netflix is aiming to reduce at their end. Recommendation system and Privacy When asked if Netflix’s recommendation engine is the best in the industry, AK was of the opinion that it can not be compared. “Our audience is diverse and varies depending on the language and region”. He further explained that to improve recommendations, the platform initially relied on English metadata but it is now starting to extract features from video content. “You see, to build a recommendation system for the English language is far easier than the local Indic languages as it depends a lot on the data upon which a particular ML model is trained.” For most of its local content, AK says Zee5 is still translating it first to English and then feeding the data to the algorithm. “Once the natural language models are capable enough to understand local Indic languages, we might see an improvement”. Furthermore, he said that platforms like Spotify have almost all songs present on the platform, or like Netflix which has huge data at its disposal. “Zee5 has a niche segment of that content, which further makes it difficult to get a recommendation as powerful as other market players.” While discussing the recommendation system, the CTO also highlighted the importance of protecting user data and privacy. He explained that Zee5 has standardised ways of executing tasks when it comes to user data management, such as looking at the user’s whole data lifecycle and what happens when it’s moving around. After a certain period of time, AK explains, Zee5 starts deleting user data holistically and has implemented deep encryption levels and access based on roles to ensure comprehensive security. Additionally, the platform allows users to control their own data, giving them the option to delete it or make it completely anonymous.","excerpt":"“If everything goes according to plan, we might continue utilising LLM Models in script writing”, says Kishore AK, CTO of Zee5.","categories":["AI Features"],"tags":["Interviews and Discussions","netflix","OTT"],"author_name":"Lokesh Choudhary","publish_date":"2023-03-10T14:00:00","publication_year":"2023","word_count":737,"keywords":["Go","AWS","AI","ML","Git","netflix","RAG","OTT","Aim","analytics","generative AI","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","ML","analytics","generative AI","Aim","RAG","AWS","R","Go","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/zee5-binges-on-ai\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10069565,"title":"What is Direct to Mobile technology?","content":"The Government of India is experimenting with the Direct to Mobile (D2M) technology that will allow users to download multimedia content directly in mobile phones without an active internet connection. Through this native technology, the GoI wants to beam information directly to the citizens, counter fake news, issue emergency alerts and assist in disaster management. According to a Deloitte study, India had 1.2 billion mobile subscribers in 2021, of which about 750 million are smartphone users. Further, the report stated that demand for smartphones in India is expected to increase at a CAGR of 6 percent. By 2026, India is forecasted to have 1 billion smartphone users. The growth in smartphone users will be fueled by 5G technology and fast adoption in rural areas. Hence, the Department of Technology (DoT) is conducting a feasibility study of a spectrum band for offering broadcast services directly to users’ smartphones. Last year, public service broadcaster Prasar Bharati also announced a collaboration with the Indian Institute of Technology (IIT) Kanpur to test the feasibility of the D2M technology. Earlier this month, IIT Kanpur organised a conclave with support from Prasar Bharati and TSDSI (Telecommunications Standards Development Society, India) on ‘Direct-to-Mobile & 5G Broadband – Convergence Roadmap for India’ in New Delhi. IIT Kanpur, in association with @TSDSI_India , is hosting a Conclave on \"Direct-to-Mobile & 5G Broadband – Convergence Roadmap for India\" in New Delhi on June 01, 2022.@shashidigital @prasarbharati pic.twitter.com\/pJBkd6Tz3O— Abhay Karandikar (@karandi65) May 31, 2022 What is D2M tech? The concept of D2M technology is a lot similar to FM radio, where a receiver within the device can tap into radio frequencies. The tech is a combination of broadband and broadcast, using which mobile phones can capture territorial digital TV signals. Using D2M, multimedia content can be beamed to phones directly. “It’s an indigenous ‘Made in India’ technology, and it is the first of its kind in the world. D2M is going to revolutionise content delivery, especially video content. Consumers today are watching videos on their phones; most of the things are consumed on phones, and with D2M tech, they would be able to receive video content without having to pay for data plans,” Parag Naik, CEO Saankhya Labs told Prachar Bharati News Services. Band 526-582 MHz is envisaged to work in coordination with both mobile and broadcast services. “DoT has set up a committee to study this band,” K Rajaraman, Secretary, DoT, said while speaking at a conclave organised by IIT Kanpur. Currently, the Band 526-582 MHz is used by Prachar Bharati (DD) or Terrestrial TV Broadcasting. Many Analogue, Digital Ready, and Digital Terrestrial TV transmitters are operating in the band. How is it helpful? The government is getting behind the technology big time because it can be used to send citizen-centric information directly, even to those who do not have access to the internet. The cable and DTH sectors will be hugely impacted if the tech takes off. This is because it will entail broadcasting coming directly to the consumers’ homes without the need for an intermediary, and this will be a very big change, according to Ministry of Information and Broadcasting (MIB) Secretary Apurva Chandra. The technology will have benefits for businesses too. It will allow telcos and internet service providers to offload video traffic from their mobile network onto the broadcast network, thus helping them decongest valuable mobile spectrum. Offloading video traffic means it will improve the usage of the mobile spectrum. At the same time, it will free up bandwidth which will directly result in reduced call drops and increased data speeds. As per Statista, India has around 622 million active internet users in 2020. Another report revealed that India’s monthly data consumption reached 17GB per user in 2021. (Source: Statista) From a consumer point of view, smartphone users can access multimedia content from Video on Demand (VoD) or Over The Top (OTT) content platforms without exhausting their mobile data. This will help consumers cut down on their spending on data. According to a EY report, OTT users in India are also expected to exponentially grow in the coming years. Further, most of India’s population still lives in rural areas where the access to the internet is few and far between. D2M will solve this problem for users in such places with limited or no internet access, they will still be able to watch video content. Also, the tech could be immensely useful to schools and colleges, especially in remote areas with poor internet connectivity. It will help students in such areas access quality educational content. Farmers, too, could access information on farming and irrigation practices or weather forecasts based on which they can plan their crops. Challenges DoT is currently carrying out a feasibility study. The technology is still at a nascent stage. The biggest challenge in front of the government is to bring in different stakeholders, including the telcos, on board in launching D2M technology on a wide scale. To get different stakeholders on board, the government must develop an attractive proposition for them or policy reforms for the tech to flourish. To launch the technology on a large scale, the government has to overcome the infrastructural challenges as well. Making the technology available in every corner of the country is not going to be easy.","excerpt":"The Department of Technology is conducting a feasibility study of a spectrum band for offering broadcast services directly to users’ smartphones.","categories":["Deep Tech"],"tags":[],"author_name":"Pritam Bordoloi","publish_date":"2022-06-22T16:00:00","publication_year":"2022","word_count":881,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","Git","RAG","ViT","GAN","R"],"extracted_tech_keywords":["AI","RAG","R","Go","Git","GAN","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/what-is-direct-to-mobile-technology\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":9135,"title":"An in-depth study on Marketing Analytics Landscape in India","content":"Understanding the customers and building a right audience for your product\/ service had always demanded innovative and revolutionary ideas in the past. In today’s scenario, with the evolution of Analytics, Marketing is no longer a skill to be mastered by extraordinary folks, and has been revolutionized as a science that everybody wants to master and cash upon. Beyond the obvious sales uptick and lead generation applications, marketing analytics paints the uniqueity of its own existence through profound insights into customer preferences, sentiments and trends. It offers a diagnostic view as how your marketing campaigns and initiatives are really performing and what it takes to assimilate great involvement with your clients in today’s world of omni-channel psychology. It involves gathering data from across all marketing channels and consolidates it into a holistic marketing view. The availability of better analytical tools has added decision-making firepower to the portfolio of successful business leaders across the globe. And, this spurt in Analytics has aptly complimented the reaping innovation in the country where numerous innovative Start-ups are supporting this very trend through delivery of innovative analytics based marketing solutions. The analysis sourced from Xeler8 Start-up Dashboard reveals that out of 400 analytics firms in India, almost 26 per cent belongs to the Marketing domain, spreading across various horizontals like Social media analytics, Campaign analytics, Sentiment analytics, Customer lifecycle management, Web analytics, Speech analytics along with comprehensive solutions (including those offering a range of services around marketing). The vibrating landscape has approx. ~100 firms with more than 55 per cent incorporated since 2010 onwards, thereby, witnessing the growing advent of start-ups in this domain. These nascent start-ups have been building upon their strength in emerging horizontals like Social media analytics, Campaign and Mobile analytics by leveraging on the connected world to map customer sentiments, preferences and choice to ensure better ROI on their marketing expenditure. The web is clearly a major game in the world and almost all the customers are utilizing the power of social media to gain competitive edge, and it has given a significant push to this high volume. The funding pattern too points in the direction of the mushrooming marketing Analytics sector. On one side, we have seen several innovative start-ups raising funds from marquee investors, the other side proudly reveals the real success of Indian Marketing analytics market with ‘Inmobi’ which has graciously fetched its sweet-spot in India’s Billion Dollar Club with a valuation of over US$ 2.5 billion. With companies like Manthan, Inmobi, Vizury, FusionCharts, Heckyl, Tookitaki, Qubole etc. raising huge amount of funds; displays the growing confidence of investors towards this growing segment. The segment has seen over US $720 mn of funding with more than US $246 mn of funding accumulated since 2014 onwards. The Marketing Analytics web is growing faster as more and more founders make a beeline for this sphere the game is bound to get more interesting. With emerging concepts like content marketing, speech analytics, sentiment analytics etc., the ability to analyse things will grow leaps and bounds. Let’s cross our fingers and hope that these young businesses get the nutrition they need to grow bigger and better.","excerpt":"Understanding the customers and building a right audience for your product\/ service had always demanded innovative and revolutionary ideas in the past. In today’s scenario, with the evolution of Analytics, Marketing is no longer a skill to be mastered by extraordinary folks, and has been revolutionized as a science that everybody wants to master and […]","categories":["AI Features"],"tags":["Analytics Case Study","analytics companies"],"author_name":"Xeler8","publish_date":"2016-02-25T04:36:46","publication_year":"2016","word_count":523,"keywords":["API","funding","programming_languages:R","AI","innovation","RAG","analytics companies","analytics","Analytics Case Study","R"],"extracted_tech_keywords":["AI","analytics","RAG","R","API","innovation","funding","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/marketing-analytics-landscape-in-india\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10022554,"title":"Making Your Data Analysis Compelling","content":"Imagine you have spent hours on data analysis at your boss’s behest. You put in a lot of effort to make your final product accurate, insightful and well packaged. But in the end, your boss decides to not use your presentation, and all your efforts go down the drain. There is nothing more frustrating than this for a data scientist, but apparently it happens far too often in the analytics industry. We spoke to some industry experts to understand why this happens and how to avoid it. Why Does It Happen? “Data science and analytics by nature is exploration. If you had all the data available and causal relationships established, you don’t need to do analytics or data science. The problem is deterministic,” said Sridhar Turaga, Senior Vice President, Digital Innovation, Data Science and Consulting at CitiusTech. “In probabilistic problems, you are often determining the right problem to solve, finding the right opportunity to focus, predict an event or discover unknown patterns. Hence you go through a top-down and bottom-up process of establishing reliable, predictive, explainable, actionable and generalizable relationships between inputs and outputs. “Not all data meets the filters of being reliable, predictive, explainable, actionable and generalizable. That’s why so much analysis and data is dropped along the way.” In other words, it’s the nature of the beast. “The data analysis could remain unused because it doesn’t present an actionable insight to the stakeholders or there is an expectation mismatch between what the stakeholders were looking for v\/s the analysis presented to them,” said Sachin Garg, Head of Data Science at PayU. “In a few cases, the cost of implementation of analysis might exceed the expected benefit from the implementation,” Garg added, which is an issue with the cost-benefit analysis, “Or in other cases, there could just be an inertia with regards to changing the way of doing business, as previously, things were done manually.” Presentation Matters It is imperative to strike the right balance between visualizations and words when presenting your analysis. “Whether a dashboard is visual or verbose, info-graph or detailed depends on the audience and usage. In many cases, you may need both – to allow for user preferences. Generally speaking, the choices of what to show and how to show should be guided by the decisions you intend to take and highlighting the right insights rather than all the information,” said Turaga. “The flow of a dashboard should allow a natural process of drilling down to make a decision. Also, given different users make decisions in different flows, flexibility to navigate through the information freely is very important. Dashboards should also organize information based on organization initiatives or levers for managing the business,” he added Along with making the right decisions, it is also important to bring the right user experience. “It is about viewing every dashboard like a digital product, having the discipline to design a human-centric product to ensure high usage, and focus on usability and flow. The fact that a dashboard has interesting content doesn’t give a pass from needing great user experience design,” said Turaga. Make It Scalable “Analytics will have a mix of one time questions and repeated questions. And there is no way to know upfront which one is which. Most innovations or disruptive opportunities start very innocently as one-time analytical questions.” said Turaga. “If you kill one time questions, the power of supporting decisions at the edges will be lost. So the way to do this is to design the platforms and process in a manner that reduces the cost of answering one-time questions while automating repeat questions quickly. Your data architecture, analytical layer, and user access\/training will be key to reducing the cost of answering one-time questions.” When it comes to scaling the analysis beyond one-time use, automation is the key. “Using the right tools and technologies to do the job is an important first step here. The principle of automation should be adhered to while working on any analysis. An analysis report which is automated is much easier to scale than any analysis which requires manual effort,” said Garg. “Further, publishing the analysis as a regular report or dashboard (with automated updates using the latest data) can ensure it can be referred to regularly by the relevant stakeholders.” Power Of Stories Decision-makers or CxOs use the analysis of data scientists. Hence, there must be good communication between the two. “Regular communication and expectation-setting are a necessary part of a data scientist’s job. Before starting a project, detailed discussions on the scope of the project, the data which will be used, the hypothesis to test and setting clear expectations with the decision-makers are essential to avoid this trap,” said Garg. “Further, if decision-makers expectations are implausible, the same should be called out by the data scientist at the beginning of the project.” However, the onus is on the analytics teams to influence decision-makers with their insights, which can be achieved through effective storytelling. “People get influenced by storytelling. Analytics teams need to moderate their insight generation. Avoid creating too much noise vs signal. Curate the right stories for the right channel at the right time. Focus on actionable insights and driving experimentation than assume brilliant insights will change minds,” said Turaga. “Enough research in sociology and behavioural psychology has shown that human beings are influenced by stories and not data.”","excerpt":"Imagine you have spent hours on data analysis at your boss’s behest. You put in a lot of effort to make your final product accurate, insightful and well packaged. But in the end, your boss decides to not use your presentation, and all your efforts go down the drain. There is nothing more frustrating than […]","categories":["IT Services"],"tags":[],"author_name":"Kashyap Raibagi","publish_date":"2021-03-21T15:00:00","publication_year":"2021","word_count":893,"keywords":["data science","Go","TPU","AI","Scala","Git","RAG","analytics","Rust","R"],"extracted_tech_keywords":["AI","data science","analytics","RAG","TPU","R","Go","Rust","Scala","Git"],"url":"https:\/\/analyticsindiamag.com\/it-services\/making-your-data-analysis-compelling\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10071192,"title":"Indian Startups are cozying up to discreet advertisements","content":"Too many choices can be confusing for anyone but especially for online shoppers. Say you are searching for a product on the online Amazon marketplace. You are sceptical about the quality of the product and also confused about which one to buy. Suddenly, something catches your eye. A product catalogue showing “Amazon’s choice” or “Bestseller”. With no prior idea about the product’s quality, you buy it. In your mind, you may appreciate how user-friendly the platform is. Rarely for an instance would it occur to you that you actually gave into a discreet, paid advertisement. It is true that online advertisements these days have become so discreet that users fail to distinguish between advertisements and product information while the ad business of these platforms keeps skyrocketing. With the continuing penetration of digital platforms, advertisements have become increasingly discreet but that does not mean that they are no longer paid for. In fact, ‘sponsored listing’ is hailed as the next big step in digital marketing. In 2021, Meta reaped USD 114.93 thousand of ad revenue. According to Statista, the projected ad revenue for Meta is USD 202 thousand by 2026. Likewise, Twitter generated USD 5.1 billion in advertising revenue, which is expected to grow to USD 9.6 billion by 2026. Besides Facebook and Twitter, the ad revenues of other tech giants show similar trends. Five of the world’s largest tech companies—Google, Meta, Alibaba, Bytedance (Chinese multinational internet technology company) and Amazon—accounted for 53 per cent of all global ad revenues in 2021. The ad revenue of  five of the world’s largest  tech companies over the years Source: Press Gazzette “Alimama which is an ads platform for Alibaba generates 60% of their revenue, estimated at $12.05 billion per annum”, claims Aarihant Aaryan, Founder and CEO, Business Score. It is plausible that the growing ad business played a crucial role in Netflix’s decision to introduce its first ad-supported subscription offering to consumers. To remain competitive in terms of revenue and subscription rate with major streaming platforms like Amazon Prime, Disney Hotstar and others—Netflix chose Microsoft as its global advertising technology and sales partner. Common practice of discreet online advertisements Be it tech giants like Google, Meta, or smaller yet popular, tech-driven companies like Swiggy, Zomato, Magicbricks, Disney Hotstar and others—every stakeholder in the business, big or small, is believed to be making significant revenues from discreet advertisements for their products. Zomato generates more than 70 per cent of its revenue through advertisements. It not only allows restaurant owners to put their banners on their app to increase visibility (direct advertisement) but also through ‘sponsored listings’, Zomato ranks their partner restaurants much higher and helps them increase order totals daily. Similarly, companies like NoBroker and MagicBricks include discreet advertisements to lead users to partner properties. “Companies like NoBroker and Magic bricks have no brokerage fees, but they have an advertisement income setup. They list\/place\/rank properties higher when a property owner pays advertisement fees”, says Aarihant Aaryan in a Twitter post. In yet another example of discreet digital marketing, the coupons received by users at the end of transactions using any UPI payment apps like PhonePe, Paytm, or GPay are not random scratch cards. They are essentially paid advertisements. In fact, it is a common practice for companies to pay these UPI apps a certain amount for offering gift coupons of their products to the users. Such discreet advertisements are prevalent in other spaces as well. For instance, in exchange for monetary compensation, phone companies routinely preinstall apps which cannot be uninstalled, only disabled. OTT platforms like Amazon Prime have carved out yet another innovative way to advertise products discreetly. “Enabling you to buy movies as rentals and ‘Enabling you to buy other OTT subscriptions on the platform’ that is an advertisement Amazon should be making money as commission per subscription or cost per click”, says Aarihant Aaryan in another tweet. Ad-tech advertisement models Tech companies derive advertisement revenue through various marketing pricing models like cost per mile, cost per click, cost per acquisition, click-through rate, cost per view and cost per lead, among others. For instance, cost per mile (CPM) refers to the cost incurred by the advertiser for getting an ad displayed to 1000 users and is also known as, ‘cost per impression’. The CPM model is believed to be the best choice for advertisers involved in brand-building business. Other models like cost per click (CPC), cost per acquisition (CPA), click-through rate (CTR), cost per view (CPV) function similarly within the digital marketing space to create revenue through product sale for advertising companies while simultaneously contributing to the advertiser’s revenue by way of user engagement. Considering the rapid evolution of the digital domain, many users often find themselves engaging with content that may not be wholly understood but would indubitably inspire their curiosity. A large share of users therefore remain unaware of how discreet online advertising functions in their everyday lives—from taking a cab to ordering a pizza. A recent survey by Bankrate showed that 49 per cent of social media users impulsively purchased a product advertised online. It is noteworthy that the same users may not be equally inclined to buy the advertised product in real-life. Much to their credit, tech companies seem to have found a way to maintain a semblance of diverse choices for their users while gently leading them toward choices that most benefit them and their partners.","excerpt":"Five of the world’s largest tech companies, including Google, Meta, Alibaba, Bytedance and Amazon, accounted for 53 per cent of all global ad revenues in 2021","categories":["AI Startups"],"tags":["AI Startups"],"author_name":"Zinnia Banerjee","publish_date":"2022-07-19T13:00:00","publication_year":"2022","word_count":899,"keywords":["Go","API","programming_languages:R","AI","ETL","programming_languages:Go","Git","RAG","Aim","R","AI Startups"],"extracted_tech_keywords":["AI","Aim","RAG","R","Go","Git","API","ETL","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-startups\/indian-startups-are-cozying-up-to-discreet-advertisements\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":57091,"title":"How To Stay On Top Of The Game With Cloud Knowledge","content":"Cloud computing is the on-demand provision of computing resources, database storage, apps, and other IT services through the web with a subscription-based cost model. Businesses and individuals can run a virtual server on Amazon Web Services (AWS) platform that can be used for a variety of tasks, applications and services. To succeed in the game, businesses are constantly looking to attract and retain the best professionals. The enterprise adoption of cloud services models has led to new opportunities for businesses to make use of the benefits this technology offers. Business leaders can make use of a public cloud environment and deliver industry-disruptive new applications in a relatively short time and without fearing about the hardware resources. Looking for employees with cloud skills is a challenging and costly endeavour. Organisations these days are finding it extremely difficult to hire and retain experts, particularly in roles needing advanced skills on technologies in the open-source space. So, organisations are giving due importance on both finding and building the skills in-house so they do not face infrastructure issues which may lead to losing the race. Also, given the rollout of new services from the three major cloud platform providers- Amazon Web Services, Google Cloud Platform, Microsoft Azure and others, cloud training has to be constant so people can stay atop the technology. Why Learning Cloud Skills May Be The Path To A Great Career Cloud Computing is one of the most sought-after skills set in the world for the last few years. With organisations quickly shifting their hardware infrastructure and applications to cloud platforms, jobs are also expanding at a swift pace, making it one of the hottest fields in information technology. While many businesses have web applications today in their own environments, they also have to manage them responsibly, and this is a huge operational cost for them. At a time when there is a shortage of such skills, organisations are making training and talent hunt a priority. If you talk about job trends and salaries, cloud is a domain that is booming. As the trend is likely to continue for years, this is one area of expertise that learners must not ignore. According to LinkedIn Global Skillset Survey 2019, cloud computing was named as the leading skillset which companies are hunting for. A fresher with cloud skills and AWS certificates can earn an average of INR 5 to 7 lacs CTC in India, and with more than five years of professional experience, a person can easily make an average of INR 12 lacs CTC in the country, says PayScale. Now, what can learners do to have access to comprehensive resources for having skills for the cloud? One major thing is to have access to content, training, collaboration tools which can push the students in the right path to put on top of the cloud computing game. There is no dearth of employment opportunities for those skilled in this space, particularly AWS. So, what learners need to do is develop skills so they can find jobs as a cloud developer\/administrator or a system operator on any cloud platform after finishing a given training program. Our Pick For A Kickstarter Course: PG Certificate In Cloud Computing Now if you’re looking for a certificate course to get started in cloud, we may have one for you which can satisfy all the needs. The Post Graduate Certificate Program in Cloud Computing with Manipal Academy of Higher Education, in collaboration with AWS Educate, and powered by Jigsaw Academy is one of the few courses in the IT industry to provide lab experience for each task executed on cloud platforms. The program’s objective is to educate and certify learners on widely used technologies like Amazon Web Services (AWS), Google Cloud Platform (GCP). One of the most thorough cloud courses in India, this 6-month long program helps users learn cloud computing while giving an in-depth understanding of the most highly-utilised platforms like Amazon Web Services (AWS) and hybrid techniques.  The curriculum of the PG Certificate Program in Cloud Computing has been created by experts and architects who have held leadership positions, with more than 20 years of experience in the tech industry. Now, you may ask why AWS is the focus of this program, the answer is quite simple. Just like Python skills are in massive demand for programming and data science, AWS skills lead the race when it comes to in-demand skills. AWS is the leading computing platform used by thousands and thousands of enterprises across the globe for hosting their products and services. Apart from AWS cloud training, the course includes more than 100 hours of hands-on Lab training on Pivotal Cloud Foundry, Open Stack and GCP. The hands-on exercises, quizzes and assignments are an integral part of learning to help learners in ever-changing technologies. It covers emerging technologies like Big Data, ML, DevOps, and IoT on AWS Platform. The course is custom designed for learners who wish to advance their career in the lucrative cloud computing space.","excerpt":"Cloud computing is the on-demand provision of computing resources, database storage, apps, and other IT services through the web with a subscription-based cost model. Businesses and individuals can run a virtual server on Amazon Web Services (AWS) platform that can be used for a variety of tasks, applications and services.  To succeed in the game, […]","categories":["AI Trends"],"tags":["Cloud Computing","cloud India","CLOUD SKILLS"],"author_name":"Vishal Chawla","publish_date":"2020-02-20T17:03:29","publication_year":"2020","word_count":831,"keywords":["data science","GCP","AWS","AI","CLOUD SKILLS","cloud computing","ML","R","RAG","Python","Cloud Computing","Azure","cloud India"],"extracted_tech_keywords":["AI","ML","data science","RAG","cloud computing","AWS","Azure","GCP","Python","R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/how-to-stay-on-top-of-the-game-with-cloud-knowledge\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":33511,"title":"Implementing Model Averaging To Reduce Variance Using Keras","content":"Deep learning neural networks are highly capable networks that can predict and solve problems with complexity. Yet the model comes with an inbuilt disability, a model may not always have the same prediction in terms of accuracy on the same machine with the same dataset. Although the model may come up with a good prediction each and every time, the variance in prediction is a real drawback in the network Ensemble learning enables us to use multiple algorithms or the same algorithm multiple times to solve the same problem helps to reduce the variance in the prediction of the same model. Model averaging is an ensemble learning technique that helps to reduce the variance in neural networks. We referenced Jason Brownlee’s tutorial, to implement Model averaging on a neural network. We also made a few changes to the code mentioned in Brownlee’s blog. Following this tutorial will require you to have: Basic knowledge in Python Understanding of Neural Networks Model Averaging Model averaging belongs to the family of ensemble learning techniques that uses multiple models for the same problem and combines their predictions to produce a more reliable and consistent prediction accuracy. Model Averaging on A Multi-Class Classification Problem First, we will create a sample dataset for a  multiclass classification problem using the make_blobs() function from sklearn.datasets. X, y = make_blobs(n_samples=500, centers=3, n_features=2, cluster_std=2, random_state=2) The above function will return 500 samples of an independent variable set with 2 features and a dependent variable which is categorical. The data points will have a standard deviation of 2 and will have 3 centers meaning that they will fall into either of the 3 categories. Visualizing the dataset : from matplotlib import pyplot import pandas as pd df = pd.DataFrame(dict(x=X[:,0], y=X[:,1], label=y)) colors = {0:'red', 1:'black', 2:'yellow'} fig, ax = pyplot.subplots() grouped = df.groupby('label') for key, group in grouped: group.plot(ax=ax, kind='scatter', x='x', y='y', label=key, color=colors[key]) pyplot.show() Output: The Multi-Layer Perceptron Model Now that we have our dataset, we will determine the variance of the prediction in the same model applied to the same dataset in the same machine. The problem is a Multi-Class classification problem, and the model will use softmax function on the output layer to predict either of the 3 categories or classes that a point falls in. Thus the first step would be to one hot encode the categorical feature which is the dependent factory here. from keras.utils import np_utils y = np_utils.to_categorical(y) Now we will create the training and testing samples for our dataset. We will split the dataset, 30% goes to the training_set and 70% goes to the test_set. from sklearn.model_selection import train_test_split X_train, X_test, Y_train, Y_test = train_test_split(X,y,test_size = 0.7, random_state = 1) Now let’s create out Neural Network model We will create a Neural Network with 2 input nodes and one hidden layer with 20 nodes and an output layer with 3 nodes and with softmax activation. The model will be compiled with ‘adam’ optimizer. from keras.models import Sequential from keras.layers import Dense model = Sequential() model.add(Dense(20, input_dim=2, activation='relu')) model.add(Dense(3, activation='softmax')) model.compile(loss='categorical_crossentropy', optimizer='adam', metrics=['accuracy']) _model = model.fit(X_train, Y_train, validation_data=(X_test, Y_test), epochs=100) After fitting the model with the training set will now evaluate its performance and compare the accuracy metrics of training and test sets. _, train_acc = model.evaluate(X_train, Y_train, verbose=0) _, test_acc = model.evaluate(X_test, Y_test, verbose=0) print('Train: %.3f, Test: %.3f' % (train_acc, test_acc)) Variance in MLP To see the variance, we just need to fit the already defined model with the same dataset on the same machine multiple times.To simplify the process of fitting the models and evaluating it for a specific number of times we will create a function. def evaluate_model(trainX, trainy, testX, testy): model = Sequential() model.add(Dense(15, input_dim=2, activation='relu')) model.add(Dense(3, activation='softmax')) model.compile(loss='categorical_crossentropy', optimizer='adam', metrics=['accuracy']) model.fit(trainX, trainy, epochs=100, verbose=0) _, test_acc = model.evaluate(testX, testy, verbose=0) return test_acc n_repeats = 10 scores = list() for _ in range(n_repeats): score = evaluate_model(X_train, Y_train, X_test, Y_test) print('> %.3f' % score) scores.append(score) from statistics import * print('Scores Mean: %.3f, Standard Deviation: %.3f' % (mean(scores), stdev(scores))) Output: Model Averaging Ensemble Now that we have understood how model averaging works we will implement it on our classification problem. But we still do not know how many reps will give the best score. Hence we will perform a sensitivity analysis to determine the optimum number of rounds that a model should run before averaging the scores. from sklearn.datasets.samples_generator import make_blobs from keras.utils import to_categorical from keras.models import Sequential from keras.layers import Dense import numpy from numpy import array from numpy import argmax from sklearn.metrics import accuracy_score from matplotlib import pyplot from sklearn.model_selection import train_test_split def fit_model(trainX, trainy): model = Sequential() model.add(Dense(20, input_dim=2, activation='relu')) model.add(Dense(3, activation='softmax')) model.compile(loss='categorical_crossentropy', optimizer='adam', metrics=['accuracy']) model.fit(trainX, trainy, epochs=100, verbose=0) return model #make an ensemble prediction for multi-class classification def ensemble_predictions(members, testX): #make predictions yhats = [model.predict(testX) for model in members] yhats = array(yhats) #sum across ensemble members summed = numpy.sum(yhats, axis=0) #argmax across classes result = argmax(summed, axis=1) return result #evaluate a specific number of members in an ensemble def evaluate_n_members(members, n_members, testX, testy): #select a subset of members subset = members[:n_members] print(len(subset)) #make prediction yhat = ensemble_predictions(subset, testX) #calculate accuracy return accuracy_score(testy, yhat) X, y = make_blobs(n_samples=500, centers=3, n_features=2, cluster_std=2, random_state=2) X_train, X_test, Y_train, Y_test = train_test_split(X,y,test_size = 0.3, random_state = 1) Y_train = to_categorical(Y_train) #fit all models n_members = 20 members = [fit_model(X_train, Y_train) for _ in range(n_members)] #evaluate different numbers of ensembles scores = list() for i in range(1, n_members+1): score = evaluate_n_members(members, i, X_test, Y_test) print('> %.3f' % score) scores.append(score) print(\"Average Accuracy Score : \", numpy.mean(score)) #plot score vs number of ensemble members x_axis = [i for i in range(1, n_members+1)] pyplot.plot(x_axis, scores) pyplot.show() Output: We can see that the accuracy maintains the average at around 13 and then fluctuates within close ranges of the average.Hence we will choose the optimum number of members to be 13, Now we can update the code to use an ensemble of 13 models. from sklearn.datasets.samples_generator import make_blobs from keras.utils import to_categorical from keras.models import Sequential from keras.layers import Dense import numpy from numpy import array from numpy import argmax from numpy import mean from numpy import std from sklearn.metrics import accuracy_score from sklearn.model_selection import train_test_split #fit model on dataset def fit_model(trainX, trainy): #define model model = Sequential() model.add(Dense(20, input_dim=2, activation='relu')) model.add(Dense(3, activation='softmax')) model.compile(loss='categorical_crossentropy', optimizer='adam', metrics=['accuracy']) #fit model model.fit(trainX, trainy, epochs=100, verbose=0) return model #make an ensemble prediction for multi-class classification def ensemble_predictions(members, testX): #make predictions yhats = [model.predict(testX) for model in members] yhats = array(yhats) #sum across ensemble members summed = numpy.sum(yhats, axis=0) #argmax across classes result = argmax(summed, axis=1) return result #evaluate ensemble model def evaluate_members(members, testX, testy): #make prediction yhat = ensemble_predictions(members, testX) #calculate accuracy return accuracy_score(testy, yhat) X, y = make_blobs(n_samples=500, centers=3, n_features=2, cluster_std=2, random_state=2) X_train, X_test, Y_train, Y_test = train_test_split(X,y,test_size = 0.7, random_state = 1) Y_train = to_categorical(Y_train) #repeated evaluation n_repeats = 10 n_members = 13 #optimum number of modeling scores = list() for _ in range(n_repeats): #fit all models members = [fit_model(X_train, Y_train) for _ in range(n_members)] #evaluate ensemble score = evaluate_members(members, X_test, Y_test) print('> %.3f' % score) scores.append(score) #summarize the distribution of scores print('Scores Mean: %.3f, Standard Deviation: %.3f' % (mean(scores), std(scores))) Output:","excerpt":"Deep learning neural networks are highly capable networks that can predict and solve problems with complexity. Yet the model comes with an inbuilt disability, a model may not always have the same prediction in terms of accuracy on the same machine with the same dataset. Although the model may come up with a good prediction […]","categories":["Deep Tech"],"tags":["Ensemble Learning"],"author_name":"Amal Nair","publish_date":"2019-01-15T12:06:34","publication_year":"2019","word_count":1192,"keywords":["NumPy","Keras","AI","neural network","ML","Ensemble Learning","RAG","Ray","deep learning","Matplotlib","Pandas"],"extracted_tech_keywords":["AI","ML","deep learning","neural network","Ray","Keras","Pandas","NumPy","Matplotlib","RAG"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/implementing-model-averaging-to-reduce-variance-using-keras\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10097164,"title":"After AP News, OpenAI Craves for More News Agencies for Real-Time Data &amp; Avoid Legal Tussle","content":"Within a week of collaborating with Associated Press (AP), one of the biggest news agency in the US, OpenAI is now out with their biggest partnership with any media organisation. Announcing over a $5 million partnership with American Journalism Project (AJP) , and another $5 million through API credits for its grantee organisations, a total of over $10 million will be invested by Sam Altman-led OpenAI. The partnership is said to explore ways to develop AI to support local news and in the process OpenAI will indirectly tie up with 41 news agencies that AJP supports. Following the line of collaborating with not-for-profit news agency AP, OpenAI’s collaboration with non-profit media continues. Associated Journalism project is a venture philanthropy that invests in developing nonprofit newsrooms and through the OpenAI partnership, the organisation is looking at ways to enhance local newsrooms with AI technologies. AJP will use the funding to build an AI studio for training their portfolio organisations and leverage AI tools at their discretion. Capturing Local Media Market Following the ‘big fish in a small pond’ philosophy, OpenAI is slowly building a grasp over media through the smaller players. By collaborating with AJP, it has indirectly partnered with 41 non profit news establishments in the US. Foregoing major news outlets for the time being, OpenAI is growing their data pool through the local market. By gaining access to local news across the country, the company is building a strong foothold in relevancy. Through AP collaboration, where OpenAI will gain access to news and information dating back to 1985 to current news, through AJP, the reach is to refine the news information through local players. Why the Media Obsession? In an earlier interview, when asked a hypothetical question, of what Altman will do if he had to run a large news media in a country like India, he said that he has heard from journalists and reporters that his product has helped them execute their boring parts of their job better allowing them to spend time on reporting and thinking of ideas. Therefore, he would encourage everyone to use it. As a prelude to what is happening now, OpenAI is in fact pushing for product adoption across media outlets. But, what is OpenAI gaining from it? In the process, the company is not only addressing the gap of providing real-time data, but also evading future legal troubles. The partnership comes at a time when the regulatory authorities are introspecting OpenAI’s actions. In spite of being a crusader of AI safety at senate hearings and formulating plans for democratising AI, OpenAI has not been able to evade the regulatory bodies. The FTC (federal trade commission) recently opened an expansive investigation into OpenAI’s activities over risks of leaking personal data. OpenAI was sent a document detailing concerns over the company’s products including information of third-parties using its API. OpenAI’s partnership with news agencies also safeguards training content. With publishers and microblogging platforms such as Twitter, Reddit and Stack Overflow restricting content from being scraped, training of AI models will become increasingly difficult considering how people are filling websites with AI-generated junk. By collaborating with news agencies, OpenAI will have unhindered access to historical as well as real-time data that can be freely used for training their future models. By partnering with news companies, OpenAI is setting a safe path for third party accountability. Associated Press and American Journalism Project, might just be the beginning of media control.","excerpt":"After AP News, OpenAI slowly looking to control the media with a total investment of over $10 million with the American Journalist Project","categories":["AI News"],"tags":["API","Associated Press","FTC","OpenAI","reddit","Stack Overflow","Twitter (X)"],"author_name":"Vandana Nair","publish_date":"2023-07-18T23:01:00","publication_year":"2023","word_count":575,"keywords":["Go","API","funding","OpenAI","AI","reddit","Stack Overflow","RAG","FTC","AI safety","Associated Press","ViT","GAN","Twitter (X)","R"],"extracted_tech_keywords":["AI","OpenAI","RAG","R","Go","API","GAN","ViT","AI safety","funding"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/after-ap-news-openai-craves-for-more-news-agencies-for-real-time-data-avoid-legal-tussle\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":24334,"title":"Punjab National Bank To Use AI To Detect Frauds","content":"In its latest move to counter frauds in banking, Punjab National Bank (PNB) has announced its plan to implement AI in account reconciliation as well as using analytics to improve its audit systems. This nimble change comes after a debilitating fraud carried out by the pair of Nirav Modi and Mehul Choksi in February 2018, which almost paralysed the bank’s operation for a short time. The scam was perpetrated to a tune of approximately ₹20,000 crores, carried out over a number of years. With the AI-enabled audit system in place, PNB aims to ensure improved security to its banking system to eliminate loopholes, in addition to gaining a stable growth for 2018. Sunil Mehta, Managing Director of PNB, said “The ‘business remodelling’ brought alive by changes at PNB is essential to ensure that the bank continues to grow and compete with its peers better” and is hopeful of the change for lesser human intervention. Furthermore, the bank has also remodelled its underwriting process for credit to reduce fraudulent activities. The process is now divided into four components where different employees deal with different stages of the process to ensure authenticity. The internal audit system is also overhauled with new changes to focus more on off-site monitoring and less on manual inspection to check risks. Earlier,foreign investors’ holdings to the bank were reduced significantly (from 12.5 percent to 9.1 percent stake) due to the Nirav Modi fraud. Thus, in order to gain traction in the market and to recover losses, the bank has said that it would focus on increasing low-cost current and savings accounts deposits to an estimate of ₹1 crore. PNB is aggressively focusing towards reskilling its employees and adopting new technology in its system.","excerpt":"In its latest move to counter frauds in banking, Punjab National Bank (PNB) has announced its plan to implement AI in account reconciliation as well as using analytics to improve its audit systems. This nimble change comes after a debilitating fraud carried out by the pair of Nirav Modi and Mehul Choksi in February 2018, […]","categories":["AI News"],"tags":["banking AI"],"author_name":"Abhishek Sharma","publish_date":"2018-05-07T11:25:09","publication_year":"2018","word_count":286,"keywords":["banking AI","programming_languages:R","AI","Aim","ViT","analytics","R"],"extracted_tech_keywords":["AI","analytics","Aim","R","ViT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/punjab-national-bank-to-use-ai-to-detect-frauds\/","complexity_score":2,"technical_depth":6,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":4223,"title":"Netscribes","content":"Netscribes is a global market intelligence company, headquartered at Mumbai, India with sales presence in New York, Silicon Valley and Singapore. The company was founded in 2000 with a vision to drive the growth of companies by providing actionable insights to take quick strategic and tactical decisions. It has three operations excellence centres in India viz. Mumbai, Kolkata, and Gurgaon. It employs over 250 professionals with diverse industry and educational backgrounds which are scaled to match the knowledge needs of its clients. 70% of its revenues are from clients in US and Europe while the rest come from the APAC region. In 2010, Netscribes, Inc. acquired Inrea Research, a fast growing patent research, analysis and licensing support company. The company is ISO 9001:2000 certified by KPMG since 2004. Lines of Business Netscribes supports the growth objectives of global firms through strategic & tactical research, market intelligence and multimedia content solutions. Its expertise spans across: Business Research – Understand global market dynamics, competitors, prospects and partners through in-depth insights using secondary, primary and social media research Technology Research – Be ahead of the technology and innovation curve with insights into the market; understand next generation disruptive technology trends and ideas Investment Research – Allow an extended team to manage due diligence research, modeling, valuation and library services to support your investment banking and trading needs Market Intelligence – Effectively use insights and tactical knowledge of the markets to drive focused business development and marketing initiatives to expand and add business Marketing Support – Manage marketing content development needs across media platforms – brochures, websites, intranets, portals, social media and presentations Service Portfolio Business Research – Market Research, Industry Research, Company Research Technology Research – Innovation Research, Patent Research Investment Research – Due Diligence Research, Equity Research, Financial Modeling, Library OR Business Information Services Market Intelligence – Media Monitoring, Marketing Intelligence, Social Media Intelligence, Competitive Intelligence Marketing Support – Social  Media Management, Social Media Marketing (SMM), Online Reputation Management (ORM), Brand Advocacy Services (Word Of Mouth), Content Management Visit us: www.netscribes.com Contact Us: Sourav Mukherjee Founder and President Email: sourav.mukherjee@netscribes.com Contact Number: +1- 917-463-4657, +91 22 4098 7601 Andrew Garvin Vice President – United States Email: andrew.garvin@netscribes.com Contact Number: +1-917-885-5983","excerpt":"Netscribes is a global market intelligence company, headquartered at Mumbai, India with sales presence in New York, Silicon Valley and Singapore. The company was founded in 2000 with a vision to drive the growth of companies by providing actionable insights to take quick strategic and tactical decisions. It has three operations excellence centres in India […]","categories":["Deep Tech"],"tags":[],"author_name":"AIM Media House","publish_date":"2013-10-25T08:44:36","publication_year":"2013","word_count":368,"keywords":["innovation","R","programming_languages:R","AI"],"extracted_tech_keywords":["AI","R","innovation","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/netscribes\/","complexity_score":3,"technical_depth":4,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10113894,"title":"Meta Releases MobileLLM with Efficient Architecture","content":"Meta introduced MobileLLM, a new approach to optimising sub-billion parameter language models for on-device use cases. This paper addresses the demand for efficient large language models (LLMs) that can be effectively deployed on mobile devices. MobileLLM is different because of its emphasis on model architecture over the sheer quantity of data and parameters, a common belief in the field. The paper outlines the development of using deep and thin architectures, embedding sharing, and grouped-query attention mechanisms to enhance model efficiency without increasing size. It uses a specific design that is detailed but compact, shares parts of the AI’s brain to use less space, and focuses attention on important information to improve understanding. Additionally, an immediate block-wise weight-sharing strategy is introduced to improve accuracy with minimal latency, making MobileLLM suitable for tasks like chat and API calling on mobile devices. This shared the information between different parts of the AI to make it smarter without slowing it down. This approach demonstrates a significant step forward in deploying powerful AI models directly on consumer hardware, offering a balance between performance and resource constraints. Companies are already adding generative AI features to their smartphones. The significance of MobileLLM extends beyond its technical achievements. It’s shift towards creating more sustainable, privacy-conscious, and accessible AI technologies by enabling powerful computational capabilities directly on users’ devices.","excerpt":"Meta’s MobileLLM introduces an innovative approach to large language models, focusing on efficient architecture for on-device AI.","categories":["AI News"],"tags":["LLMs","Meta"],"author_name":"K L Krithika","publish_date":"2024-02-26T12:09:31","publication_year":"2024","word_count":220,"keywords":["API","Meta","attention mechanism","programming_languages:R","AI","LLMs","generative AI","R"],"extracted_tech_keywords":["AI","generative AI","R","API","attention mechanism","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/meta-releases-mobilellm-with-more-efficient-architecture\/","complexity_score":3,"technical_depth":6,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10162086,"title":"AI Landscape Automation Startup Attentive.ai Secures $12 Mn Series A Funding","content":"Delhi-based AI landscape automation startup Attentive AI recently closed its Series A2 funding round, bringing its total Series A funding to $12 million across two tranches (A1 and A2). The A2 round was led by Tenacity Ventures, with additional support from Vertex Ventures, Peak XV Partners, and InfoEdge Ventures. With this latest funding, the company plans to deepen its AI automation capabilities, alongside expanding into new markets, and enhance operational efficiency for the landscaping and construction sectors. Speaking on leading investors backing the startup, the co-founder Shiva Dhawan, in a recent conversation with AIM, attributed it to their focused vision. “These investors believe in our vision of AI-enabled services transforming trillion-dollar industries,” Dhawan said. Founded in 2017 by three friends from IIT-D, Dhawan, Rishabjit Singh, and Aishwarya Maurya, Attentive.ai provides AI-powered solutions for the landscaping and outdoor services industry. “With our AI-based product – Beam AI, our users can save valuable time spent on doing manual takeoffs and focus on driving growth for their business. This funding is a testament to our commitment to this mission,” said the co-founder, Dhawan, sharing its mission of being the best reconstruction solution in the industry. Currently, the company focuses on North America markets, where takeoffs (measuring site areas) from PDFs are a standard procedure owing to intellectual property concerns making the process labour-intensive. What about India? “The Indian market operates differently. CAD files are more widely circulated here, which simplifies takeoffs. In the US, PDF files dominate, and extracting measurements from them is significantly harder,” shared Dhawan, in an earlier interaction with AIM. This fundamental difference in workflows makes the US a more suitable market for Attentive.ai’s solutions. Targeting a Niche Market: Its success lies in its ability to cater to a niche but significant market. The billion-dollar landscaping and construction industries have traditionally been overlooked by AI companies. Attentive.ai recognized this gap and utilised its expertise in computer vision to develop customised solutions. “These industries have huge potential but are largely ignored by technology firms. If you can build for them, it creates a moat,” said Dhawan. Co-Founder & MD Tenacity Ventures, Rohit Razdan said that they have been tracking Attentive ai for more than a year now. “During this time Shiva and the Attentive team made strong progress on bringing cutting edge software tools to both the outdoor services and the construction industry. They understand that these industries are underserved, and the AI\/software augmentation of a workforce that has a big capacity crunch, is very timely,” he said. The startup will continue to focus in the US as Dhawan believes a better opportunity lies in the far West. “We aren’t focusing on India in the short- or medium-term because the US market offers better opportunities and cost arbitrage,” he said.","excerpt":"With this latest funding, the company plans to deepen its AI automation capabilities and more.","categories":["AI News"],"tags":["AI automation","Attnetive ai","landscaping","PeakXV","Tenacity"],"author_name":"Vandana Nair","publish_date":"2025-01-23T22:04:16","publication_year":"2025","word_count":459,"keywords":["AI automation","API","funding","programming_languages:R","AI","Tenacity","Attnetive ai","computer vision","RAG","automation","Aim","landscaping","PeakXV","R","startup"],"extracted_tech_keywords":["AI","computer vision","Aim","RAG","R","API","automation","startup","funding","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/ai-landscape-automation-startup-attentive-ai-secures-12-mn-series-a-funding\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10051206,"title":"Upcoming AI Conferences And Events To Look Forward To","content":"Conferences and events are a great place to connect, share knowledge and learn. They provide a platform for tech enthusiasts to keep up with the fast-changing trends and developments in the tech and analytics industries. With this in mind, Analytics India Magazine has curated a list of artificial intelligence (AI) conferences and events– in India and abroad (virtually), to look forward to this month. Performance Boosting ETL Workloads’ Performance Using Spark 3.0 NVIDIA and Micropoint, along with Analytics India Magazine, have organised a webinar on ‘Performance Boosting ETL Workloads’ Performance Using Spark 3.0’ on October 20. The masterclass will be covering ETL\/data architecture and workflows in the industry, introduction to the best practices in performance optimisation and speedups, hands-on examples of speeding up the workflows using open-source plugins on Spark, and an introduction to NVIDIA RAPIDS and how it can help boost the performance of ETL workloads. Get the speaker details and register here. AWS Data and Analytics Conclave The AWS Data and Analytics Conclave, in association with Analytics India Magazine, is a day-long virtual event to help businesses reinvent with data through modern strategies. The conclave will be held on October 21, between 10 AM and 4.30 PM and will have 10+ sessions on data and analytics, one-on-one interaction with data experts and panel discussions with industry experts. The keynote will be delivered by Greg Khairallah- Director (Analytics) at AWS. Register here Accelerating Data Engineering with Vertica US-based database company Vertica, along with Analytics India Magazine, has organised a live webinar on Accelerating Data Engineering with Vertica on October 27 to showcase how businesses have accelerated their analytics performance using the data warehouse platform by Vertica. The workshop will be delivered by Ravi Gupta, Senior Solution Architect at Vertica and will cover the new release features of Vertica 11, how the platform can be used for advanced analytics and ML at scale, its Eon Mode and SaaS offerings on AWS. Register here Machine Learning Developers Summit 2022 Analytics India Magazine’s Machine Learning Developers Summit 2022 (MLDS22) is a two-day hybrid event to be held on January 19 and 20 in Bengaluru. Exclusively focusing on the ML developer ecosystem, MLDS22 will provide a platform to showcase ML tools and frameworks, evaluate new projects, hardware and software challenges of building complex ML systems, using the right tools and platforms, languages, software and standards. The conference is expected to have more than 1,000 attendees together in-person and virtually. The offline conference carries strength of 200, with seats available on a first-come-first-serve basis. It will also be welcoming submissions reporting research advancing AI. Register here International Conference on Smart Technology, AI and Computer Engineering The Institute of Research and Journals will be conducting the International Conference on Smart Technology, Artificial Intelligence and Computer Engineering (ICSTAICE) on October 10 in Hyderabad. The conference sees the participation of academic elites, research industry leaders from across the UK, North America, Australia, New Zealand and Europe. ICSTAICE offers business opportunities, research possibilities and access to new markets in the pharmaceutical and healthcare industries. ICSTAICE aims to bring together academic scientists, researchers and research scholars to exchange and share experiences and results on engineering and technology, management, education and technology, and applied science. The conference will also offer a platform for presenting new research findings and promoting global scientific and community collaborations. International Conference on AI and Soft Computing Organised by Academics World, the International Conference on Artificial Intelligence and Soft Computing (ICAISC) is on a mission to bring academic scientists, researchers and scholars together and allow the exchange of ideas, experiences and research results. Held across October 15 and 16, the conference will provide an interdisciplinary forum for AI researchers, practitioners and educators in the field to present and discuss all innovations, recent trends, challenges encountered, and solutions deployed in the fields of AI and Soft Computing. WSJ Tech Live The Wall Street Journal’s technology event– WSJ Tech Live, is scheduled from October 18 to October 20. This year’s tech conference will be staged across New York City, LA and San Francisco, and the audience will get to immerse themselves in a virtual reality world built especially for Tech Live. Attendees will get to connect with founders, collaborators and investors and also get to watch interviews that they might miss – later. The online pass costs $150, and one can register here. AIMLSystems 2021 The Association of Computational Linguistics is organising the International Conference on AI-ML Systems from October 21 to 23 in Bengaluru. The first edition of the conference will be targeting researchers in the intersection of AI and ML techniques and systems engineering. It will specifically focus on the developments possible through computational systems research– improvements in CPU\/GPU architectures, data-intensive infrastructure, and communications. Additionally, it will also be focusing on using AI\/ML to design socio-economic systems like public healthcare and security. The AIML Systems 2021 conference will be held in a hybrid format– allowing both online and in-person participation. International Conference on AI, ML and Big Data Engineering Lined up for October 24, the International Conference on Artificial Intelligence, Machine Learning and Big Data Engineering will be held in Kochi. The conference will be providing a platform to present the latest research and results of researchers in the AI, ML and Big Data Engineering topics. The International Conference on AI, ML and Big Data Engineering is aimed at allowing the exchange of ideas, applications and experiences, helping in the establishment of new business models or research relations. Additionally, it will also help researchers and companies to find global partners for future collaborations. Reuters Momentum The Reuters Momentum– Improving lives through technology, the virtual global conference is scheduled to be held across October 27, 28 and 29. With a star-studded speakers’ list, the conference will focus on how technology innovations can impact lives and future proof businesses. Speakers will be talking about the ‘next normal’ opportunities. Additionally, Reuters Momentum promises to bring together tech innovators and leaders from across businesses, governments, NGOs and the common interest groups. The broader themes that will be covered during the conference include– society, economy, sustainability, trust and ethics. AI Summit Silicon Valley The sixth annual AI Summit- Silicon Valley 2021, is scheduled for November 3 and 4. In partnership with IoT World, Silicon Valley, the conference will provide a platform for learning, networking and problem-solving. The conference promises speakers from large tech companies like AWS, Google, Intel, Microsoft, Samsung, Genpact, Facebook, and IBM, providing insights into pioneering AI projects from across industries. Additionally, there will be AI roundtable discussions, AI tech tours and workshops.","excerpt":"AIM brings you a curated list of top AI conferences and events happening in the next month.","categories":["Deep Tech"],"tags":["AI (Artificial Intelligence)","ai conferences","Machine Learning"],"author_name":"Debolina Biswas","publish_date":"2021-10-12T10:00:00","publication_year":"2021","word_count":1094,"keywords":["Rapids","Go","ai conferences","artificial intelligence","machine learning","AWS","AI","ML","Machine Learning","Aim","analytics","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","analytics","Aim","Rapids","AWS","R","Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/upcoming-ai-conferences-and-events-to-look-forward-to\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10024099,"title":"Guide to PyTerrier: A Python Framework for Information Retrieval","content":"Information Retrieval is one of the key tasks in many natural language processing applications. The process of searching and collecting information from databases or resources based on queries or requirements, Information Retrieval (IR). The fundamental elements of an Information Retrieval system are query and document. The query is the user’s information requirement, and the document is the resource that contains the information. An efficient IR system collects the required information accurately from the document in a compute-effective manner. The popular Information Retrieval frameworks are mostly written in Java, Scala, C++ and C. Though they are adaptable in many languages, end-to-end evaluation of Python-based IR models is a tedious process and needs many configuration adjustments. Further, reproducibility of the IR workflow under different environments is practically not possible with the available frameworks. Machine Learning heavily relies on the high-level Python language. Deep learning models are built almost on one of the two Python frameworks: TensorFlow and PyTorch. Though most natural language processing applications are built on top of Python frameworks and libraries nowadays, there is no well-adaptable Python framework for the Information Retrieval tasks. Hence, here comes the need for a Python-based Information Retrieval framework that supports end-to-end experimentation with reproducible results and model comparisons. PyTerrier & its Architecture Craig Macdonald of the University of Glasgow and Nicola Tonellotto of the University of Pisa have introduced a Python framework, named PyTerrier, for Information Retrieval. This framework proposes different pipelines as Python Classes for Information Retrieval tasks such as retrieval, Learn-to-Rank re-ranking, rewriting the query, indexing, extracting the underlying features and neural re-ranking. An end-to-end Information Retrieval system can be easily built with these pre-established pipeline elements. Moreover, a built IR architecture can be scaled or extended in the future as per the requirements. A typical model comparison experiment for two different IR models (Source) Are you looking for a complete repository of Python libraries used in data science, check out here. An experiment architecture for comparing two different Information Retrieval models has many key components such as Ranked retrieval, Fusion, Feature extraction, LTR (Learn-to-Rank) re-ranking and Neural re-ranking. The workflow is represented in a directed acyclic graph (DAG) with complex operation sequences. The PyTerrier framework helps build such a complex DAG problem in an end-to-end trainable pipeline. PyTerrier & its Key Objects PyTerrier is a declarative framework with two key objects: an IR transformer and an IR operator. A transformer is an object that maps the transformation between an array of queries and the corresponding documents. The Transformer Classes of PyTerrier. Q and R represent the input query and the input document, respectively. An element provided in parentheses is optional (Source). The basic retrieval process, for example, in PyTerrier is performed using the following Python code. Here, Q is the input query and R’ is the retrieved output document. Thus, a complex IR task can be performed with simple Python codes. Also, PyTerrier provides operator overloading for conventional math operators to perform custom IR operations. The PyTerrier operators employed under operator overloading strategy (Source). The newly introduced PyTerrier Framework is instantiated on two public datasets so far: the Terrier dataset and the Ansereni dataset. More dataset implementations would be expected soon. Hands-on Retrieval and Evaluation PyTerrier is available as a PyPi package. We can simply pip install it. !pip install python-terrier Import the library and initialize it. import pyterrier as pt if not pt.started(): pt.init() Use one of the in-built datasets to perform the retrieval process and extract its index. vaswani_dataset = pt.datasets.get_dataset(\"vaswani\") indexref = vaswani_dataset.get_index() index = pt.IndexFactory.of(indexref) print(index.getCollectionStatistics().toString()) Output: Extract queries as topics for the dataset. topics = vaswani_dataset.get_topics() topics.head(5) Output: Perform retrieval easily using a few commands as shown below. retr = pt.BatchRetrieve(index, controls = {\"wmodel\": \"TF_IDF\"}) retr.setControl(\"wmodel\", \"TF_IDF\") retr.setControls({\"wmodel\": \"TF_IDF\"}) res=retr.transform(topics) res Output: It can be observed that the documents are retrieved and ranked. Further, the results can be saved to the disk using the ‘write_results’ method available in the ‘io’ class of the PyTerrier framework. pt.io.write_results(res,\"result1.res\") Now, evaluation is performed by comparing the results with the ground truth available in-built. Get the ground truth query results. qrels = vaswani_dataset.get_qrels() Output: Evaluate the query results. eval = pt.Utils.evaluate(res,qrels) eval Output: Evaluation results can also be obtained for per-query results. Here, the evaluation is performed based on the ‘map’ metric on all documents under query. eval = pt.Utils.evaluate(res,qrels,metrics=[\"map\"], perquery=True) eval A portion of the output: Find the Notebook with these code implementations here. Hands-on Learn-To-Rank Create the environment by importing the necessary libraries and initializing the PyTerrier framework. import numpy as np import pandas as pd import pyterrier as pt if not pt.started(): pt.init() Download an in-built dataset, its indices, queries and ground truth results. dataset = pt.datasets.get_dataset(\"vaswani\") indexref = dataset.get_index() topics = dataset.get_topics() qrels = dataset.get_qrels() For ranking the queries, the standard ‘BM25’ model is used in this example. The traditional ‘TF-IDF’ model and the ‘PL2’ model are used to re-rank the query results. #this ranker will make the candidate set of documents for each query BM25 = pt.BatchRetrieve(indexref, controls = {\"wmodel\": \"BM25\"}) #these rankers we will use to re-rank the BM25 results TF_IDF =  pt.BatchRetrieve(indexref, controls = {\"wmodel\": \"TF_IDF\"}) PL2 =  pt.BatchRetrieve(indexref, controls = {\"wmodel\": \"PL2\"}) Create a PyTerrier pipeline to perform the above said example task and make a query. pipe = BM25 >> (TF_IDF ** PL2) pipe.transform(\"chemical end:2\") Output: In the above output, the term ‘score’ represents the ranking score of the BM25 model and the term ‘features’ represents the re-ranking scores of the TF-IDF and PL2 models. However ranking at the first step and re-ranking in two successive steps consumes more time. To tackle this issue, PyTerrier introduces a method, called FeaturesBatchRetrieve. Let’s implement the method for efficient processing by ranking and re-ranking, all in one go. fbr = pt.FeaturesBatchRetrieve(indexref, controls = {\"wmodel\": \"BM25\"}, features=[\"WMODEL:TF_IDF\", \"WMODEL:PL2\"]) # the top 2 results (fbr %2).search(\"chemical\") Output: PyTerrier has a pipeline method, called compile(), which optimizes the ranking and re-ranking processes automatically. This approach also yields the same results as above at around the same compute-time. An example implementation is as follows: pipe_fast = pipe.compile() pipe_fast %2).search(\"chemical\") Output: After performing ranking and re-ranking, a machine learning model can be built to Learn-to-Rank (LTR). Split the available data into train, validation and test sets. train_topics, valid_topics, test_topics = np.split(topics, [int(.6*len(topics)), int(.8*len(topics))]) Build a Random Forest model to perform the LTR and obtain the results. from sklearn.ensemble import RandomForestRegressor BaselineLTR = fbr >> pt.pipelines.LTR_pipeline(RandomForestRegressor(n_estimators=400)) BaselineLTR.fit(train_topics, qrels) resultsRF = pt.pipelines.Experiment([PL2, BaselineLTR], test_topics, qrels, [\"map\"], names=[\"PL2 Baseline\", \"LTR Baseline\"]) resultsRF Output: Build an XGBoost model to perform the LTR and obtain the results. import xgboost as xgb params = {'objective': 'rank:ndcg', 'learning_rate': 0.1, 'gamma': 1.0, 'min_child_weight': 0.1, 'max_depth': 6, 'verbose': 2, 'random_state': 42 } BaseLTR_LM = fbr >> pt.pipelines.XGBoostLTR_pipeline(xgb.sklearn.XGBRanker(**params)) BaseLTR_LM.fit(train_topics, qrels, valid_topics, qrels) resultsLM = pt.pipelines.Experiment([PL2, BaseLTR_LM], test_topics, qrels, [\"map\"], names=[\"PL2 Baseline\", \"LambdaMART\"]) resultsLM Output: Find the Notebook with these code implementations here. Wrapping up We discussed the newly introduced PyTerrier framework, its architecture and its implementation for Information Retrieval tasks. We learnt how to use the framework with two example hands-on implementations for the applications, a Simple Query-Retrieval and a Learn-to-Rank machine learning model. PyTerrier has enormous algorithms and in-built datasets to perform almost any Information Retrieval task with minimal efforts. This framework is also established as a Python-built one focusing chiefly on simplicity, efficiency and reproducibility. Further reading: Research paper Github repository Indexing with PyTerrier Index API of PyTerrier","excerpt":"PyTerrier framework proposes different pipelines as Python Classes to build an end-to-end, scalable Information Retrieval system","categories":["Deep Tech"],"tags":["Guide","information","Python","random forest","ranking","XGBoost"],"author_name":"Rajkumar Lakshmanamoorthy","publish_date":"2021-04-18T13:00:00","publication_year":"2021","word_count":1231,"keywords":["data science","NumPy","machine learning","AI","PyTorch","Python","Ray","random forest","XGBoost","information","ranking","deep learning","TensorFlow","Guide","Pandas"],"extracted_tech_keywords":["AI","machine learning","deep learning","data science","Ray","TensorFlow","PyTorch","XGBoost","Pandas","NumPy"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/guide-to-pyterrier-a-python-framework-for-information-retrieval\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":39720,"title":"5 Chinese AI Stocks That Are Soaring &#038; Promise Good Returns","content":"Image source: Inside Business Online The global demand for artificial intelligence (AI) and machine learning (ML) solutions are growing at a steady pace and the key solutions providers are ramping up their capabilities to meet this burgeoning demand. It is estimated that by 2030, these technologies will add $15.7 trillion to global GDP. Being one of the leading players in the field, the Chinese economy is likely to benefit substantially from this increasing demand and the key players are already working to reap its benefits. As Chinese companies have increased their technological capabilities, surpassing even the likes of Microsoft and Google, we take a look at the top five companies’ stocks that are likely to fetch you a good investment return. 1.Baidu (BIDU): Chinese multinational company which is a prominent internet service provider and has a range of AI-based products. It witnessed a revenues increase of 14.13% year on year in the fourth quarter, to $ 14,876.00 million. Though in 2018 it cloaked a revenue of 28.7 per cent, the Chinese internet giant is witnessing remarkable growth over the years. Following the 2010 government blanket ban on internet service providers such as Google and Facebook, Baidu stepped in to fill the void and presently tops the market share with 54 per cent. The company is also known to heavily rely on algorithms and analytics to customise the search engine queries and replies. Baidu voice and mobile assistant, text-speech system, AI-based transcription tools are some of its key AI offerings. 2.Alibaba (BABA): Is a leading Chinese e-commerce company that entered the AI-market with TmallGenie, its intelligent voice assistant. AliGenie, an intelligent human-machine interaction interface and its mobile robots which work on sensoring, interaction and understanding and autonomous planning and task completion capabilities are its AI-focused products. Over the last couple of years, Alibaba has been giving stiff competition to Google’s Alphabet inc, with investors closely monitoring their market activities with keen interest. For the last quarter, though Alibaba’s growth in home currency rate stood at 51 per cent, due to the devaluation of Chinese yuan, its valuation in USD was comparatively lesser. “Last year, Alibaba bought Ele.me, local services and food delivery business, which contributed 5% to revenue. The company bought streaming service Youku Tudou in 2016 for $5.4 billion, moving Alibaba headlong into the streaming media segment. Still, Alibaba’s core business remains very healthy, growing at a 45% rate in local currency last quarter,” a leading web portal noted. 3.Tencent Holdings (TME): Founded in 1998, Tencent is a Chinese multinational investment holding conglomerate and specialises in various internet-related services and products, entertainment, artificial intelligence and technology both in China and internationally. Announcing its latest first quarter results, the company said that its revenue stood at $12,693 million and that it cloaked an increase of 16% over the first quarter of 2018 and attributed to success to its game business with games like Perfect World  Mobile in China, while PubG Mobile grew its user-base steadily. “We are consequently disclosing their results in our new FinTech and Business Services segment, demonstrating our success in organically incubating services with long-term growth potential. We believe that we are building solid foundations for future growth in both the Consumer and Industrial Internet domains,”  Ma Huateng, Chairman and CEO of Tencent said. 4.JD.com: JD.com is China’s largest online retailer and its biggest overall retailer, as well as the country’s biggest Internet company by revenue. It is often referred to as the Amazon of China, the e-commerce giant is known to use AI, robotics and automation to stand out in the stiff Chinese e-commerce market. The net revenue of the company for the first quarter stood at $118.0 billion, earning the company a 20.9% increase from the first quarter of 2018. While its net service revenue was $1.9 billion, where it witnessed an increase of 44.0% from the first quarter of 2018. Commenting on its performance, Richard Liu, Chairman and CEO of JD.com said that the growth was possible due to the companies increasingly focused on driving innovation using cutting edge-technologies like AL and ML. 5.iFlytek: Is a Chinese information technology company established in 1999 and provides voice recognition software and internet and mobile products across various sectors. Though its revenue witnessed a slump in the previous year, it’s spending on research and development increased substantially thus witnessing a 40 per cent year on year growth in the first quarter of  2019.","excerpt":"The global demand for artificial intelligence (AI) and machine learning (ML) solutions are growing at a steady pace and the key solutions providers are ramping up their capabilities to meet this burgeoning demand. It is estimated that by 2030, these technologies will add $15.7 trillion to global GDP. Being one of the leading players in […]","categories":["AI Trends"],"tags":["AI Stocks","Alibaba","Baidu","China","Tencent"],"author_name":"Akshaya Asokan","publish_date":"2019-05-27T07:30:06","publication_year":"2019","word_count":735,"keywords":["Go","artificial intelligence","machine learning","AI","Alibaba","ML","Tencent","AI Stocks","automation","Baidu","analytics","ViT","GAN","R","China"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","analytics","R","Go","GAN","ViT","automation"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/5-chinese-ai-stocks-that-are-soaring-promise-good-returns\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":49048,"title":"Top 10 Automatic Speech Recognition Tools That’ll Relieve You Of The Keyboard","content":"Speech recognition is the process of decoding human voices and is a part of machine learning. Organisations are implementing Automatic Speech Recognition (ASR) technology to create documents without touching the keyboard, controlling devices, and other similar tasks. In this article, we list down 10 speech-to-text services which can be used for various applications. (The list is in alphabetical order) 1| Amazon Transcribe Amazon Transcribe is an Automatic Speech recognition (ASR) service which converts speech to text quickly. The features of this service include easy-to-read transcriptions, streaming transcription, timestamp generation, custom vocabulary, multiple speaker recognition, and channel identification. This service can be used to transcribe various speech-related tasks such as customer service calls, automate closed captioning and subtitling as well as generate metadata for media assets to create a fully searchable archive. 2| Apple Dictation Apple has an in-built Dictation feature which converts any spoken words into text. One can also format or edit as needed in the text by using simple commands like “new paragraph” or “select previous word.” One can dictate continuously when the cursor is in a document, email message, text message, or other text fields. 3| Google Cloud Speech-to-Text Google Cloud Speech-to-Text enables developers to convert audio to text by applying powerful neural network models in an easy-to-use API. The service can process real-time streaming or prerecorded audio by using the tech giant’s machine learning technology. With the help of this service, one can enable voice command-and-control, transcribe audio from call centres, and much more. 4| Google Docs Voice Typing Google Docs Voice Typing is a speech-to-text feature which is only available in Chrome browsers. Using a microphone, one can easily speak for speech to text dictation as well as pause and resume when needed. It is an easy to use voice recognition service and very convenient to the users. 5| IBM Watson Speech to Text API IBM Watson Speech to Text service provides an API to add speech transcription capabilities to applications. It combines information about language structure with the composition of the audio signal. This service automatically transcribes audio from 7 languages in real-time and has the ability to rapidly identify and transcribe what is being discussed, regardless of lower quality audio. The IBM Watson Speech to Text service uses speech recognition capabilities to convert Arabic, English, Spanish, French, Brazilian Portuguese, Japanese, Korean, German, and Mandarin speech into text. 6| Microsoft Azure Speech to Text The Speech-to-text from Azure Speech Services enables real-time transcription of audio streams into text that the applications, tools, or devices can consume, display, and take action on as command input. By default, the speech-to-text service uses the Universal language model and is powered by the same recognition technology that Microsoft uses for Cortana and Office products. 7| Speechmatics Tech company, Speechmatics used its decades of machine learning and research expertise to develop Automatic Speech Recognition (ASR). This service can be used for real-time or pre-recorded audio and video files and helps the customers across a variety of industries to accurately understand and transcribe spoken words. This service is available in private or public clouds and securely on-premises. 8| Speechnotes Speechnotes is a free and online speech-to-text notepad which is built by using cutting-edge speech-recognition technology for the most accurate results. It is a powerful speech-enabled online notepad which lets a user move from voice-typing (dictation) to key-typing seamlessly. 9| Twilio Speech Recognition Twilio adds Google’s speech recognition to its voice platform in order to build Automated Speech Recognition (ASR) which easily converts speech to text as well as analyse the intent of the speech during a voice call. Currently, this service has the ability to recognise 119 languages and dialects in order to support global user base. 10| VoxSigma AP VoxSigma is a suite of language-specific speech recognition software offered by Vocapia Research. It offers large vocabulary speech-to-text capabilities in many languages and has been designed for professional users in both batch mode and real-time.","excerpt":"Speech recognition is the process of decoding human voices and is a part of machine learning. Organisations are implementing Automatic Speech Recognition (ASR) technology to create documents without touching the keyboard, controlling devices, and other similar tasks. In this article, we list down 10 speech-to-text services which can be used for various applications. (The list […]","categories":["AI Trends"],"tags":["Natural Language Processing","Speech Recognition","speech-to-text"],"author_name":"Ambika Choudhury","publish_date":"2019-10-30T11:49:33","publication_year":"2019","word_count":655,"keywords":["Go","API","machine learning","AI","neural network","Natural Language Processing","ML","Azure","R","RAG","Speech Recognition","GAN","speech-to-text"],"extracted_tech_keywords":["AI","machine learning","ML","neural network","RAG","Azure","R","Go","API","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-10-automatic-speech-recognition-tools-thatll-relieve-you-of-the-keyboard\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10110283,"title":"What to Expect from the GPT Store","content":"OpenAI is gearing up for the official launch of its much-awaited GPT Store this week. It will mark a significant milestone in customising AI agents. Initially slated for November, the launch had encountered delays due to internal restructuring at the company, spurred by the sudden departure and reappointment of CEO Sam Altman. The GPT Store would allow users, specifically ChatGPT Plus and enterprise subscribers, to sell and share tailored AI agents based on GPT-4. This initiative follows OpenAI’s commitment to democratising access to advanced language models, enabling users to create personalised versions of ChatGPT-style chatbots. With the capability to understand and generate human-like text, GPT models have already demonstrated their versatility in various applications, from content creation to language translation. The store is expected to use many tools to enhance productivity, creativity, and communication. One of the key advantages of the store is its potential to democratise access to cutting-edge language models. Developers and businesses can build and deploy applications that leverage the underlying power of the LLM, making advanced language capabilities more widely available. GPT Store Expectations From elucidating Gen Z memes to guiding users through negotiations, the store will serve as a marketplace where these customised agents can be shared and potentially monetised. However, a large part of the execution depends on OpenAI’s commitment to safety as it will play a pivotal role in shaping the narrative of the GPT Store’s impact on our digital landscape. The company claims that it wants to make GPT Store completely error-free, hence the delay in launch. The email sent to GPT Builders announces the impending launch and reminds them to adhere to brand guidelines and make their models public. This emphasis on guidelines reflects OpenAI’s commitment to maintaining the quality and appropriateness of content. While details about the monetisation plan for GPT creators remain undisclosed, the company has expressed its intention to compensate creators based on the usage of their AI agents on the store. The analogy of the GPT Store being more like a Spotify subscription platform than an App Store resonates with the expectation that millions of interchangeable, easily created GPTs will populate the platform. Anticipation around money-making opportunities for developers is tempered by the likelihood that the majority of GPTs will be free, with a Plus subscription providing enhanced features for $20 per month. A few thoughts about the coming GPTs store.(All might be wrong though.)Because of the word “store” many people think that it’ll be like the App Store and get excited about the money making opportunity.I have a feeling that the better analogy is a Spotify type subscription… pic.twitter.com\/HzWfrhggzS— Borriss (@_Borriss_) November 12, 2023 Like musicians on Spotify, most developers may earn modest returns, while a select few achieve significant sales. Though it offers some financial potential for developers, it might not yield substantial profits for the majority. As with musicians using streaming platforms, developers could use free GPTs as lead magnets for alternative monetisation strategies, flexing their success on social media. Prevailing Concerns The recent tweets have raised questions regarding the safety concerns that have cast a shadow over the otherwise bright prospects of the GPT Store. People argue that the immense language generation capabilities of the models could be exploited to produce misleading or harmful content. This has become more apparent with the OpenAI-NYU lawsuit. Not your model, not your GPTs. All those folks rushing to add stuff to the GPT store are writing free functions for another OpenAI llm that they will brand as “AGI”. Almost free labor extraction. I will bet, for most folks, the revenue share will be pennies. Folks who think… https:\/\/t.co\/iEHqX3OJQh— Delip Rao e\/σ (@deliprao) January 8, 2024 Despite these safety measures, some argue that the potential for misuse is too great, and the safeguards may not be foolproof. Privacy concerns have also been raised, as users worry about the implications of storing and sharing sensitive information within the GPT ecosystem. Alongside, to address the privacy concerns surrounding it, the company recognises the importance of community involvement. Unpopular Opinion (Probably):The GPT-Store isn't going to be as big of a deal as most people make it out to be.99.9% of GPTs that will be created will have nothing proprietary about them and anyone will be able to just make their own clone of it.There will be a handful of…— Matt Wolfe (@mreflow) January 5, 2024 OpenAI encourages feedback, scrutiny, and suggestions to refine its privacy measures, fostering a transparent and inclusive environment where the concerns of various stakeholders are taken into account. Through this open dialogue, the company aims to strike the right balance between innovation and privacy, making the GPT Store a responsible and trustworthy platform for users and developers. Read: 10 GPTs for Your Daily Use OpenAI is set to launch the GPT Store, allowing ChatGPT Plus and enterprise subscribers to trade tailored AI agents based on GPT-4. The article talks about the initial delay due to internal restructuring after CEO Sam Altman’s departue and how the store aligns with OpenAI’s mission to democratize advanced language models. The store promises tools for enhanced productivity and communication, democratizing access to cutting-edge models. The article explains delayed launch, aimed at ensuring error-free operation, underscores OpenAI’s commitment to safety, crucial for the store’s impact. The platform, akin to a Spotify subscription, anticipates a myriad of interchangeable GPTs, offering financial potential for developers. However, safety concerns loom large, fuelled by recent tweets and the OpenAI-NYU lawsuit, highlighting potential misuse and the need for foolproof safeguards. Privacy worries also surface, prompting OpenAI to seek community involvement for feedback and suggestions, aiming for a balanced and trustworthy platform. The article concludes with the opinion that despite uncertainties, the GPT Store signifies a transformative leap in AI customization, with OpenAI navigating a delicate balance between innovation and privacy in the evolving AI landscape.","excerpt":"The GPT Store allows users, specifically ChatGPT Plus and enterprise subscribers, to sell and share tailored AI agents based on GPT-4.","categories":["Global Tech"],"tags":["ChatGPT","OpenAI"],"author_name":"Sandhra Jayan","publish_date":"2024-01-10T11:07:28","publication_year":"2024","word_count":968,"keywords":["Go","ChatGPT","OpenAI","AI","chatbots","AWS","RAG","Aim","Rust","R"],"extracted_tech_keywords":["AI","ChatGPT","OpenAI","Aim","RAG","chatbots","AWS","R","Go","Rust"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/what-to-expect-from-gpt-store\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10101593,"title":"Why AI’s to be Blamed for Layoffs","content":"Yesterday Stack Overflow announced that they were laying off their staff for the second time this year. Prashanth Chandrasekar, the CEO of Stack Overflow explained these layoffs in his blog saying that the initial restructuring efforts weren’t enough and “we have made the extremely difficult decision to reduce the company’s headcount by 28%.” This year saw a phenomenal increase in layoffs in the tech industry. January having the most with a constant trickle throughout the year. Human resources took the major hit after which software engineers came next with 22.1% of employees laid off this year. In the example of Stack Overflow, there have been reports for a while now that claim users preferred ChatGPT over the 15 year old website to help with coding troubles resulting in a drastic drop in traffic. This is the first time that AI has been replaced. Microsoft’s Linkedin also announced their layoffs again yesterday, cutting down their workforce in the engineering, talent and product departments. Dukaan, an e-commerce platform, earlier this year laid off 90% of their support staff and replaced them with AI. The CEO said that their customer support resolution time went down considerably because of this. The layoffs for each company is a combination of multiple factors but the increased use of AI in the workplace is slowly but surely threatening to replace the human workforce. AI and Human Resources True to the report published by Goldman Sachs, the majority of work that can be automated is already being done so by AI. Scaling a business now doesn’t require as much labour anymore and technology will be used as a default. “Lesser hiring means fewer recruiters. Leaner organisations point to small HR departments. Sales, marketing, legal, and accounting teams will shrink as tools become more powerful.” predicts Vin Vashishta. “AI in HR is not new; its integration has been ongoing, and generative AI’s recent emergence globally has sparked an AI race impacting all organisations profoundly,”  Nishchae Suri, Managing Director at Cornerstone OnDemand told AIM. Cornerstone is a SaaS platform that provides a comprehensive learning and talent management solution. The pressure to enhance operations using AI is widespread, affecting HR as well. Currently, 81% of HR leaders have explored or adopted AI solutions, a figure likely to rise. Suri further added that the potential value of AI for HR is substantial, providing insights into talent performance and aligning employees with relevant learning pathways matching their progression goals. However, concerns arise with the swift acceleration of AI and its implications for employees. According to Cornerstone OnDemand’s latest Talent Mobility Study, 73% of employees express interest in exploring new roles within their company, and a higher percentage of them preferred using technology to explore these new roles rather than a conversation directly with the manager. So AI in human resource solutions can recommend learning opportunities, analyse career trajectories, and align key skills with organisational requirements. In line with this the company announced ‘Skills Playground’ a free, AI-powered tool that matches skills to job roles and learning content ontology, with more than 50,000 skills identified, and lets visitors experience AI skills detection firsthand. This is one of the earliest technologies that streamlines HR roles and processes using AI. IBM also continues to build their Watsonx platform adding features to automate processes like onboarding etc. Does AI take over jobs? There is little to show that AI is the reason for layoffs. Right now, gig workers in content creation have seen a direct hit because it is easy to produce good quality AI generated content. The constant news of layoffs are rarely related to AI. The Stack Overflow layoffs were blamed on the business strain due to macroeconomic factors by its CEO. LinkedIn is likely suffering from poor recruitment – whose fees are its biggest revenue. It still remains that AI is only a tool for humans to employ to make their jobs easier rather than a threat. In the meanwhile however it is unfortunate to have the constant fluctuation in the job market. Upskilling with online courses and certifications and keeping in line with the industry requirements is gaining popularity with companies as well as job seekers. It is important to stress that AI is a helpful tool which will reduce the number of jobs that require repetition and can be automated. On the flip side, jobs for developers have increased in demand over the years. Suri sums up saying, “The nature of work might shift, and some tasks may become automated, but new opportunities will arise as well.”","excerpt":"Not really. It is important to stress that AI is a helpful tool which will reduce the number of jobs that require repetition and can be automated.","categories":["AI Features"],"tags":["Dukaan","Hiring","Layoffs","linkedin","recession","Stack Overflow","tech recession"],"author_name":"K L Krithika","publish_date":"2023-10-17T18:00:00","publication_year":"2023","word_count":755,"keywords":["Go","ChatGPT","Layoffs","recession","AI","ML","Hiring","Stack Overflow","GAN","GPT","Aim","llm_models:GPT","generative AI","linkedin","Dukaan","R","tech recession"],"extracted_tech_keywords":["AI","ML","generative AI","ChatGPT","Aim","R","Go","GPT","GAN","llm_models:GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/why-ais-to-be-blamed-for-layoffs\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10043808,"title":"Behind Tech Billionaire Star Wars","content":"Undoubtedly, the billionaire space race makes for an interesting spectacle, an extravagant show ushering in an era of universal space flight led by private firms. The first stirrings of space exploration can be traced back to the beginning of the Cold War. In the 1950s, the US and the USSR worked feverishly to develop nuclear weapons and intercontinental ballistic missiles (ICBMs). The results of bitter rivalry manifested in advancements in rocket technology and shaped the early charters to terra incognita. The Soviet Union sent the first human being to space in 1961, and the first artificial satellite started orbiting the earth in 1967. The competition hit a fever pitch with the first moon landings- when Neil Armstrong made the small step in 1969. The world has benefited from the “spinoff” technologies of the modern space race. For example, the global positioning systems (GPS) was originally developed for navigation and weapon targeting by the US military. Now, it’s helping individuals and companies navigate their respective worlds. Tech accountability Space exploration used to be the prerogative of governments — pushing the species to the uncharted territories. Meanwhile, Richard Branson’s latest flight is for his leisure only. Private tech enterprises are not accountable, yet space activities affect everyone. Rocket launches spew copious amounts of black carbon (soot) and CO2 into the atmosphere. The higher the spacecraft flies, the more fuel it takes up. The Virgin Galactic Shuttle consumed almost 720,000 kilograms of liquid fuel and 500,000 kilograms of solid rocket fuel. Despite repeated green pledges, Amazon’s carbon footprint rose by 15% in 2020. Fossil fuel emissions for Amazon stood at 18% in 2019 and continue to rise. Remember kids: if you study hard, get good grades, go to a good college, get a job, work hard, never take a sick day, live within your means and do what you're told … then one day your boss might go to space— Dan Price (@DanPriceSeattle) July 12, 2021 From an ecological perspective,  space tourism comes at a huge expense. The international space law is ill-equipped to deal with the meteoric rise of commercial space enterprise. The Outer Space Treaty (to which India is a signatory) has come up with a basic set of principles for space activity; unfortunately, it has no hold over private entities. Commercial space economy The commercialisation of space isn’t just about sending the elite on space vacations. Though the tech elites dominate the field, hundreds of space startups have mushroomed across the globe focusing on everything from satellite tech to manufacturing and launching. The space industry market cap stands at $350 billion and is predicted to cross the trillion dollar milestone by 2040. The democratisation of the space industry coupled with advances in high power computing has enabled satellite builders to cut down on satellite size & weight, manufacturing and launch costs. Private launch capabilities are making commercial space a viable domain. SpaceX Falcon 9 rocket has reduced costs per kilogramme by 85% with partial reusability. Morgan Stanley Research (2040 estimates) Spacefaring has become profitable and runs the risk of turning outer space into satellite junkyards. The Union of Concerned Scientists Satellite database has listed more than 4,084 operational satellites currently orbiting the Earth. In 2010, this number was less than a thousand. In the distant future, this problem can extend to the lunar surface and the asteroid belt (the current count stands at 34,000 pieces space junk bigger than 10 centimetres in size and millions of smaller pieces). SpaceX is also planning to deploy up to 12,000 desk-sized satellites by 2027. The company is angling to capture 5 per cent or $50 billion of the trillion-dollar telecommunications industry. And, In the last few months, SpaceX has put 960 Starlink satellites in low Earth orbit. Revenue of the global satellite industry from 2006 to 2020: Source: Statista Race to space Branson described himself as Astronaut 001 and described his mission as “evaluating customer spaceflight experience”. With tech elites serving as role models, an increasing number of entrepreneurs are joining the race to create cheap and commercialised space travel. Amazon’s Jeff Bezos has channelised huge resources to help Blue Origin drive down the cost of space travel by building reusable rockets. Blue Origin is now securing government contracts and national security certifications and also building rockets for the US Air Force. those who attack spacemaybe don’t realize thatspace represents hopefor so many people— Elon Musk (@elonmusk) July 13, 2021 Both Blue Origin and SpaceX aim to explore habitation possibilities on the moon and the red planet. Immense efforts are made to escape the bonds of Earth’s gravity–to make human life multi-planetary. Whatever the tech elites may wish to prove, the goal for the private sector anyway is to get satellites, people and cargo to space in cheaper and quicker ways.","excerpt":"“Those who attack space maybe don’t realize that space represents hope for so many people”","categories":["AI Features"],"tags":[],"author_name":"Prajaktha Gurung","publish_date":"2021-07-18T15:00:00","publication_year":"2021","word_count":798,"keywords":["Go","ELT","programming_languages:R","AI","programming_languages:Go","Aim","ViT","GAN","R","startup"],"extracted_tech_keywords":["AI","Aim","R","Go","ELT","GAN","ViT","startup","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/behind-tech-billionaire-star-wars\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10713,"title":"Interview with Michael Ferrari, Head of Data Science, The Weather Company","content":"Analytics India Magazine brings you yet another interview with our valued speaker at Cypher 2016, Michael Ferrari, Head of Data Science at The Weather Company, an IBM Business. Michael Ferrari, a data science strategist works closely with Weather’s leadership team to identify opportunities that leverage our massive sets of weather, location, and audience data to uncover business and consumer insights. In his free time, he is an affiliate scientist with The National Center for Atmospheric Research (NCAR) Applications Laboratory and an active board member on the American Meteorological Society’s Board on Global Strategies. Michael is also a research affiliate at the MIT Media Lab where he works on the OpenAG initiative towards the exploration and development of future food systems. AIM brings you the excerpt from the interview. This being the second in the series. AIM: Would you like to share with us about your talk at Cypher? Michael: I would like to discuss some of the unique insights that we at The Weather Company\/IBM are working towards, notably the combination of rich location and weather data.  Each individually is powerful, but when combined can be transformational. AIM: Tell us about your journey in the field of data science. Michael: My background is in climate numerical modeling and biophysics, each computationally intensive fields. So I have been working in ‘data science’ well before the term was coined.  It is exciting to blend approaches and techniques taken from the physical sciences and applied to social discussions. AIM: How important is predictive analytics for any company? Which industries do you think using predictive analytics is a must? Michael:  I think within 10 years, there will not be any field of commerce that is left untouched by advances in analytics and data.  Agriculture in particular is ripe for disruption. AIM: Would you like to share your views on the IoT industry? Michael: IBM and Weather are both heavily invested in IoT and believe strongly in its ability to help people and businesses have access to more information that will ultimately improve decision making. I would say my own personal views on the potential of IoT could best be described cautiously optimistic. While I do think that there is something ‘there,’ I think the current version of how IoT will transform the world and what actually materializes are two different things. Primarily, I see a much more subtle interaction between connected devices and technology, i.e. not so much flying cars, but more and more it will be seen in ways that we conduct our lives with the technology in the background. AIM: Could you tell us about some important contemporary trends that you see emerging in the present analytics space across the globe? Michael: When the large portion that is not connected makes the jump. Most of the world still does not see affordable and reliable connectivity.  The efforts of Facebook and Google in this space are particularly exciting. With this additional connectivity, think about all of the interesting and useful ways that our data ecosystem (weather and location) can contribute to the lives of nearly every person on the planet who carries a mobile phone. AIM: What are the most significant challenges you see in the Analytics space? Michael: The avalanche of connected individuals in the coming 1-2 decade and the resultant information will dwarf what we have seen in the last decade.  We view this challenge as a tremendous opportunity.  It is also worth mentioning that while we are of course pursuing areas that make sense from a commercial perspective, what is equally important is the humanitarian perspective. These changes will truly impact lives for the better.","excerpt":"Analytics India Magazine brings you yet another interview with our valued speaker at Cypher 2016, Michael Ferrari, Head of Data Science at The Weather Company, an IBM Business. Michael Ferrari, a data science strategist works closely with Weather’s leadership team to identify opportunities that leverage our massive sets of weather, location, and audience data […]","categories":["AI Features"],"tags":["Interviews and Discussions","IoT","Predictive AI","predictive analytics"],"author_name":"Manisha Salecha","publish_date":"2016-09-06T07:03:48","publication_year":"2016","word_count":605,"keywords":["data science","Go","AI","R","RAG","IoT","Aim","ViT","analytics","disruption","predictive analytics","Predictive AI","Interviews and Discussions"],"extracted_tech_keywords":["AI","data science","analytics","Aim","RAG","predictive analytics","R","Go","ViT","disruption"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/interview-michael-ferrari-head-data-science-weather-company\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10038903,"title":"Tesla Crash, NVIDIA’s Arm ‘Deal’emma, Facebook’s Clubhouse Clone &#038; More In This Week’s Top News","content":"Last Saturday, two men were killed in Texas after a speeding 2019 Tesla Model S crashed into a tree while negotiating a turn. The data logs recovered showed the car’s autopilot was not enabled. The car did not have the FSD feature either, Elon Musk tweeted on Monday. FSD is a $10,000 add-on feature that allows vehicles to self-park and automatically changes lanes on highways. Further, the street where the accident happened did not have lane lines to turn the standard pilot on, he added. Your research as a private individual is better than professionals @WSJ! Data logs recovered so far show Autopilot was not enabled & this car did not purchase FSD.Moreover, standard Autopilot would require lane lines to turn on, which this street did not have.— Elon Musk (@elonmusk) April 19, 2021 According to sources, Tesla said autopilot and FSD features do not make its vehicles fully autonomous and drivers are advised to keep their hands on the wheel at all times. In 2018, 38-year-old Tesla Model X driver died in an accident in California. Reports showed the autopilot had been engaged prior to the crash. Back in 2014, a 40-year-old driver was killed in a Tesla Model S on autopilot. The autopilot sensors on the Model S failed to distinguish a white tractor-trailer crossing the highway against a bright sky. NVIDIA-Arm deal in trouble The UK government is taking a long hard look at the controversial NVIDIA-Arm deal citing national security concerns. After the UK government initiated a national security probe into the graphic chip maker’s proposed acquisition of Arm Holdings from Japan’s Softbank, the shares of NVIDIA fell by 3.5%. Digital Secretary Oliver Dowden has issued a Public Interest Intervention Notice (PIIN) in relation to the proposed sale of ARM to NVIDIA. Further, Dowden wrote to the competition and markets authority (CMA), asking them to start a phase I investigation to examine the nature of the transaction. As per the digital secretary directions, the CMA has to submit the report by July 30, 2021. Recently, a  lawsuit was filed by Arm China against three executives appointed by the board to replace CEO Allen Wu, which could further delay the takeover of Arm by NVIDIA. Also, there are speculations that China’s regulators could also get in the way of NVIDIA’s Arm acquisition. IBM posts profit for the first time since 2018 On April 19th, IBM announced its first-quarter 2021 earnings results. According to sources, the big blue reported its biggest revenue gain in eleven quarters, driven by demand for cloud services. “Strong performance this quarter in cloud, driven by increasing client adoption of our hybrid cloud platform, and growth in software and consulting enabled us to get off to a solid start for the year,” said Arvind Krishna, IBM chairman and chief executive officer.  “While we have more work to do, we are confident we can achieve full-year revenue growth and meet our adjusted free cash flow target in 2021,” he added. IBM’s free cash flow for the last 12 months was $11.0 billion. The company expects to grow revenue for the full year 2021 based on mid-April 2021 foreign exchange rates. Zuckerberg’s Clubhouse clone Last Monday, Facebook CEO Mark Zuckerberg announced the social media company is building audio features where users can engage in real-time conversations with others. The new feature is known as Live Audio Rooms, similar to Clubhouse, the popular app that allows people to listen and participate in live conversations. “We’ve been investing in audio technologies — such as speech-to-text and voice morphing — for a long time and will make them available in an audio creation tool directly inside the Facebook app,” Fidji Simo, Head of Facebook App said in a blog post. Facebook will also launch a new feature called Soundbites. “These audio creation tools will enable you to create Soundbites — short-form, creative audio clips for capturing anecdotes, jokes, moments of inspiration, poems, and many other things we haven’t yet imagined,” Simo added. Twitter to hire engineers in India Recently, Twitter announced its plans to expand its engineering team in India to strengthen its offerings for local and global audiences. The micro-blogging company has also announced the appointment of Apurva Dalal as the Director of Engineering. India is a priority market for Twitter, and also a world-class engineering talent hub. We are excited to welcome @MrApurvaDalal as Director of Engineering, under whose leadership we will continue our hiring expansion in Bengaluru. https:\/\/t.co\/rIIrtxIJDZ— Nick Caldwell (@nickcald) April 20, 2021 According to the official release, starting on 20th April 2021, Apurva will become the senior-most member of Twitter’s engineering team in India and will be responsible for strengthening the company’s engineering capacity. The new team in India will be positioned in Bangalore. Apple targeted in a ransomware attack Few days ago, Apple was targeted in a ransomware attack. The incident happened after a Russian hacker gang REvil group breached servers of Quanta. Quanta is a Taiwanese manufacturer for giants like Apple, Facebook, Microsoft etc. According to sources, the hacker gang posted a message on a dark web portal saying the Taiwanese company refused to pay to get its stolen data back and, now the REvil operators have decided to go after the company’s primary customer. The Russian hacker gang is demanding a ransom of $50 million from the US-based phone maker to avoid the leak of confidential information.","excerpt":"Last Saturday, two men were killed in Texas after a speeding 2019 Tesla Model S crashed into a tree while negotiating a turn. The data logs recovered showed the car’s autopilot was not enabled. The car did not have the FSD feature either, Elon Musk tweeted on Monday. FSD is a $10,000 add-on feature that […]","categories":["AI News"],"tags":["Apple","IBM","Microsoft"],"author_name":"Ambika Choudhury","publish_date":"2021-04-25T10:00:00","publication_year":"2021","word_count":896,"keywords":["Go","AWS","AI","cloud_platforms:AWS","Apple","programming_languages:R","programming_languages:Go","Git","IBM","CLIP","GAN","R","Microsoft"],"extracted_tech_keywords":["AI","AWS","R","Go","Git","CLIP","GAN","cloud_platforms:AWS","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/tesla-crash-nvidias-arm-dealemma-facebooks-clubhouse-clone-more-in-this-weeks-top-news\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10053823,"title":"Major Hiring Announcements By Tech Giants In 2021","content":"When the COVID-19 pandemic hit, organisations had to undergo a rapid digital transformation in order to sustain their businesses. This was possible only through good technology support and software capabilities. Due to this, despite facing some challenges in the lockdown imposed nationwide, IT companies recovered quite well with good hiring growth. A Naukri JobSpeak report said that the demand in tech has continued to rise as the sector witnessed a significant 85% Y-O-Y growth in October 2021. Let’s look at some of the major hirings that took place this year: TCS In July, the company’s chief of global human resources, Milind Lakkad, said that the IT giant would be hiring more than 40,000 freshers from campuses in the country in the financial year 2021-22. It had hired 40,000 graduates from campuses last year as well. He added that lateral hiring would also be robust this year. It will also aim to improve on the number of trainees hired from American campuses last year. Wipro In a recent interview with Moneycontrol, Wipro’s Chief Financial Officer Jatin Dalal said that the company will hire 16,000 to 17,000 freshers in the current financial year as against its previous estimates of 12,000. It had hired 8,100 freshers in the June-September quarter and plans to recruit 40 per cent more freshers than originally planned as the present high rate of attrition can continue for the next few quarters. HCL Recently, HCL had launched its Amazon Web Services (AWS) Business Unit (AWS BU) to help enterprises globally accelerate their cloud transformation. At the moment, it has more than 10,000 professionals trained on AWS and plans to boost this capacity to more than 20,000 specialists in the future, said the company. Cognizant As per media reports, Cognizant said it expects to offer jobs to 45,000 new graduates in India in the fourth quarter for onboarding in 2022. In addition, it has reported a 56.3 per cent rise in September quarter net income at $544 million. Amazon In September, Amazon said it plans to hire more than 8,000 direct jobs across 35 cities this year. These roles will be across corporate, technology, customer service and operations roles. The hiring will occur across 35 cities in the country, like Bengaluru, Hyderabad, Chennai, Amritsar, Ahmedabad, Bhopal, Coimbatore, Jaipur, Kanpur, Ludhiana, Pune, Surat, etc. EY Consulting leader EY said at the end of 2020 that it will hire 9000 new hires in India in various technology roles across all member firms, including the global delivery centres, in 2021. These hires will be from the STEM background and in areas including artificial intelligence, machine learning, cyber security, analytics and other emerging technologies. Currently, 36% of all EY India employees are from a STEM background. Through this hiring process, it wants to expand its digital capabilities and help organisations solve their complex end-to-end business transformation challenges. Ola As per reports, Ola recently said that it plans to hire 10,000 people, as it eyes a leadership position with $2 billion gross merchandise value (GMV) for its vehicle commerce platform Ola Cars over the next 12 months. In addition, the company said that Ola Cars would be operational in 30 cities over the next two months and expand to 100 cities by next year across key areas, including sales and service centres. PayPal At the beginning of the year, PayPal said that it would hire over 1,000 engineers for its India development centres in Bengaluru, Chennai and Hyderabad in 2021. The hirings will occur across software, product development, data science, risk analytics, and business analytics streams at entry, mid-level and senior roles.","excerpt":"Hirings that happened across big companies this year","categories":["AI Hirings"],"tags":["Amazon","Cognizant","Company","HCL Technology","Hiring","Ola","Software","STEM","TCS","Technology"],"author_name":"Sreejani Bhattacharyya","publish_date":"2021-11-19T16:00:00","publication_year":"2021","word_count":597,"keywords":["STEM","Hiring","Git","R","data science","artificial intelligence","Cognizant","Company","analytics","Go","Ola","machine learning","AWS","AI","Amazon","Technology","Software","TCS","Aim","HCL Technology"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","data science","analytics","Aim","AWS","R","Go","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-hiring\/major-hiring-announcements-by-tech-giants-in-2021\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":69365,"title":"Hands-On Guide To Detect Objects In Video In 5 Steps","content":"Object detection techniques are a trend nowadays in the field of computer vision. There are many techniques which are used to detect objects in the scene images and videos. Each of these techniques has its own advantages and limitations in terms of resources and execution time. Detecting objects in videos also requires a lot of technical knowledge and resources. So, there is always a search for an easy and quick method for object detection. In this article, we will show how to detect objects seen in the videos in only 5 steps. We will use the pixellib library in this task which detects objects using instance segmentation. We will also use the pre-trained Mask R-CNN model to identify the objects seen in the videos. In this implementation, we will detect vehicle objects in traffic video. Instance Segmentation Instance segmentation is a technique in computer vision that is used in object detection using image segmentation method. It identifies each instance of the objects present in the images or videos at the pixel level. In image segmentation, the visual input is splitted into segments to represent the object or part of objects by forming a collection of pixels. Instance segmentation identifies each instance of each object featured in the image instead of categorizing each pixel like in semantic segmentation. Mask R-CNN Mask R-CNN is a variant of Deep Neural Network proposed by Kaiming He et al at Facebook AI Research. This model is used to solve the object instance segmentation problem in computer vision. It detects objects in an image while simultaneously generating a high-quality segmentation mask for each instance. It is an extension of Faster R-CNN by adding a branch for predicting an object mask in parallel with the existing branch for bounding box recognition. The Mask R-CNN framework for instance segmentation is given below. (Image Source: Original Research Paper) Implementation Now, we will discuss the steps through which we will detect the objects in a video. 1. Install library and dependencies In the first step, we need to install the pixellib library and its dependencies. ! pip install pixellib 2. Load pre-trained Mask-RCNN weights As we are going to use the Mas R-CNN model to detect objects, we will download its pre-trained weights. !wget --quiet https:\/\/github.com\/matterport\/Mask_RCNN\/releases\/download\/v2.0\/mask_rcnn_coco.h5 3. Import libraries Now, we will import the pixellib library that has been installed. We will also import the instance_segmentation class because we are going to detect objects using instance segmentation method. import pixellibfrom pixellib.instance import instance_segmentation 4. Instantiate instance segmentation model and load Mask-RCNN weights In this step, we will instantiate the instance segmentation class that is provided by pixellib and load the Mask R-CNN model with its pre-trained weights. segment_video = instance_segmentation()segment_video.load_model(\"mask_rcnn_coco.h5\") 5. Detect objects In this step, we will process the object detection task by Mask R-CNN in a video. A random traffic video is used in which we want to detect vehicle objects. https:\/\/analyticsindiamag.com\/wp-content\/uploads\/2020\/07\/traffic_vid2.mp4 Traffic Video In this method, we set the frames per second that are the number of frames per second output video will have. segment_video.process_video(\"traffic_vid2.mp4\", show_bboxes = True, frames_per_second= 15, output_video_name=\"object_detect.mp4\") Finally, we will get the output video in the working directory. The time in this process depends on the lengths and size of the video. You should use a GPU for faster processing speed. For the above traffic video, we have got the following video in the out having detected objects. https:\/\/analyticsindiamag.com\/wp-content\/uploads\/2020\/07\/Object_Detect.mp4 You can define a function to get the videos from YouTube and pass them directly to the function above. So, using the above steps, we could discuss a very easy way to implement the task of object detection in videos. A person with very less knowledge of deep learning and computer vision can be able to detect objects using this way.","excerpt":"In this article, we will show how to detect objects seen in the videos in only 5 steps. We will use the pixellib library in this task which detects objects using instance segmentation. We will also use the pre-trained Mask R-CNN model to identify the objects seen in the videos. In this implementation, we will detect vehicle objects in traffic video.","categories":["Deep Tech"],"tags":["Computer Vision","Deep Learning","Object Detection"],"author_name":"Dr. Vaibhav Kumar","publish_date":"2020-07-10T16:00:00","publication_year":"2020","word_count":627,"keywords":["Go","TPU","AI","neural network","computer vision","Aim","deep learning","Object Detection","analytics","Computer Vision","object detection","Deep Learning","R"],"extracted_tech_keywords":["AI","deep learning","neural network","computer vision","analytics","Aim","object detection","TPU","R","Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/hands-on-guide-to-detect-objects-in-video-in-5-steps\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":38287,"title":"5 Tips To Get Beginners Started With SQL","content":"For every business a database is critical and in order to manage data efficiently and allows users to perform multiple tasks with ease, a database management system (DBMS) is imperative. However, having a DBMS doesn’t end the story, in order to make the most out of a DBMS, every company needs professionals with top-notch SQL skills. When it comes to programming, one cannot deny the fact of how important SQL (structured query language) is. This domain-specific language time and again proved its capabilities of managing data. Today, with business enterprises getting more data-driven, and most database management systems are powered by SQL, SQL skills can get one hired. As new companies are gathering more and more information and technology continues to evolve, it is becoming imperative for SQL programmers to up their SQL programming skills to the next level. If you are someone who is from the field of SQL then, here are some of the ways you can upgrade your skills and write better queries. Go Over The Basic Concepts Of SQL Many of you might actually relate to this point as it is the most common one. Every programmer at some point in his\/her life rethinks about the basics they learned when they were starting out. Technology is ever-evolving, and to keep up with that, you have to keep learning — the process never stops. Whether its Database Fundamentals or Database Design to SQL Basics, make sure you are clear completely with the basics (commands and queries). If you want to have one more look at the fundamentals of database and SQL, there are a number of resources available both offline and online Guru 99, Tutorials Point, GeeksForGeeks are some of the online resources where you can easily find the topic you want to clear. Also, make sure that every time you clear a basic concept, don’t forget to have a look at different examples. It gives you the idea of where to exactly use a particular type of query or command. Gain The Most Out Of These Free Learning Resources Every programmer would agree to this that practice is one of the best ways to polish and up your coding skills. And when it comes to SQL, you need to have experience with tools that refine your skills. If you are getting started with SQL and done with your theory phase, its high time to get started with some coding. The internet is loaded with several platforms that allow you to get your hands on SQL programming. Here are some of the popular ones: W3Schools: This platform is basically focused on the basics and foundation. If you are in the very-beginner phase, you can definitely give w3schools’ SQL practice platform. SQLZoo: Since its inception, SQLZoo has gained tremendous popularity. From lessons to quizzes to challenges, SQLZoo is a one-stop-shop for SQL enthusiasts. W3resources:  It is again one of the platforms to practice SQL coding. The platform has a lot of exercises that coders can try and practice. Also, the internet is not the only place where you can only practice. You can always get yourself a free version or a licensed version of any DBMS. Install and start coding. Unlike your workplace systems, in your own PC, you can download sample databases and try out different things and play around. Word to the wise: Once you are done with your theory and practice, try to take up some high-level SQL quizzes on the internet. It would give you an overview of what all you have learn and what are the concepts you need to clear again. Participate In Hackathons Hackathons are also one of the ways to gain a lot of knowledge and polish your skills. They not only let you solve problems and challenges but also helps you think out of the box solutions, which is a significant factor for a SQL programmer. That is not all, there are some hackathons that not only rewards you for being a winner but also provides you with opportunities to get hired by some of the reputed companies from all across the world. Attend Conferences And Workshops Apart from studying and practicing, one should also consider attending workshops and conferences. Why? Because, it is always imperative for a programmer to know what is happening outside his office cubicle, what other companies are doing and what are all the new technologies that have emerged recently. That is not well, many speakers also present different uses cases and that is what every SQL programmer should keep an eye one.  When you are presented with different use cases, it helps you understand what ways you can leverage the same. Also, you can always find talks of videos from different conferences on the internet. Do consider watching some of the videos. Working For Startups Can Be A Good Entry Point There are many small companies that face a lot of database issues and in order to manage their budget, they always seek freelancers to fix their problems. This could be a great opportunity for SQL programmers. When you work for real-time problems or challenges you gain a significant amount of knowledge. That is not all, these kinds of opportunities also add a lot of value to your CV. So, if you have a good amount of knowledge on databases and SQL, try to get work experience by joining a startup as an entry point. Reach out to small-sized companies and this would help gain relevant experience but also give you a view of the ways companies across the world are leveraging SQL.","excerpt":"For every business a database is critical and in order to manage data efficiently and allows users to perform multiple tasks with ease, a database management system (DBMS) is imperative. However, having a DBMS doesn’t end the story, in order to make the most out of a DBMS, every company needs professionals with top-notch SQL […]","categories":["AI Trends"],"tags":["ai lessons for beginners","Database Management System","DBMS","MySQL","SQL","TED Talks"],"author_name":"Harshajit Sarmah","publish_date":"2019-04-26T06:24:46","publication_year":"2019","word_count":929,"keywords":["Go","startup","programming_languages:R","AI","data-driven","programming_languages:Go","TED Talks","RAG","programming_languages:SQL","ai lessons for beginners","SQL","DBMS","MySQL","R","Database Management System"],"extracted_tech_keywords":["AI","RAG","R","SQL","Go","data-driven","startup","programming_languages:R","programming_languages:SQL","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/5-tips-to-get-beginners-started-with-sql\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10063322,"title":"Difference Between Zero-shot, One-shot and Few-shot Learning?","content":"There are several use cases for machine learning when data is insufficient. N-shot learning is when a deep learning model can be trained to classify an image using not more than five images. An N-shot learning field includes an ‘n’ number of labelled samples of each ‘K’ class. The entire support set ‘S’ includes N*K total samples. N-shot learning can be divided into three categories: zero-shot learning, one-shot learning and few-shot learning. The choice of application between the three depends upon the availability of training samples. N-shot learning is also used when the dataset is huge, and labelling the data can prove to be costly. Or, when several samples are available, it could be hard to add specific features for each task. Zero-shot Learning Zero-shot learning is the challenge of learning modelling without using data labelling. Zero-shot learning involves little human intervention, and the models depend on previously trained concepts and additional existing data. This method reduces the time and effort that data labelling takes. Instead of giving training examples, zero-shot learning gives a high-level description of new categories so that the machine can relate it to existing categories that the machine has learned about. Zero-shot learning methods can be used in computer vision, natural language processing and machine perception. Zero-shot learning essentially is made up of two stages: training and inference. In training, the intermediate layer of semantic attributes are captured, then in the inference stage, this knowledge is used to predict categories among a new set of classes. At this level, the second layer models the relationship between the attributes and the classes and fixes the categories using the initial attribute signature of the classes. For example, if a child is asked to recognise a Yorkshire terrier, he or she may know it is a type of a dog, with added information about it from Wikipedia. With a growing amount of research in instances where the model uses as little data as possible with fewer annotations, zero-shot learning has found applications in critical areas like healthcare for medical imaging and COVID-19 diagnosis using chest x-rays as well as for unknown object detection used in autonomous vehicles. Hugging Face transformers use zero-shot classification for more than 60 per cent of its transformers. Zero-shot learning has two different ways in which it approaches modelling problems: Embedding-based approach: This maps the semantic attributes with the image features into a common embedding space. It uses a projection function that it has learned from deep networks. The model aims to use data from seen categories during training to find a projection function from visual to semantic space. The projection function is learned as a deep neural network as neural networks are used as function approximators. Generative model-based approach: This approach aims to generate image features for unseen categories using semantic attributes. This approach tackles the bias and domain shift that the embedding-based approach has. One-shot Learning One-shot learning performs classification tasks using past data. Facial recognition technology, including facial verification and identification, usually uses one-shot learning. Facial recognition systems learn face embedding, which is a rich low-dimensional feature representation. One-shot learning has been using the Siamese network approach. Eventually, Siamese networks were compared to comparative loss functions, after which the triplet loss function was proven to be better and the FaceNet system began using them. Contrastive loss and triplet loss functions are now used for high-quality face embeddings, which have become the foundation for modern facial recognition. Few-shot Learning Few-shot learning, also known as low-shot learning, uses a small set of examples from new data to learn a new task. The process of few-shot learning deals with a type of machine learning problem specified by say E, and it consists of a limited number of examples with supervised information for a target T. Few shot learning is commonly used by OpenAI as GPT3 is a few-shot learner. A study in 2019 titled ‘Meta-Transfer Learning for Few Shot Learning’ addressed the challenges that few-shot settings faced. Since then, few-shot learning is also known as a meta learning problem. There are two ways to approach few-shot learning: Data-level approach: According to this process, if there is insufficient data to create a reliable model, one can add more data to avoid overfitting and underfitting. The data-level approach uses a large base dataset for additional features. Parameter-level approach: Parameter-level method needs to limit parameter space and use regularisation and proper loss functions to resolve the overfitting problem in few-shot learning. This will generalise the limited training samples. This approach can also improve model performance by directing it to the extensive parameter space. A parameter-level approach is useful because a smaller amount of training data will not give reliable results.","excerpt":"Zero-shot learning has found application in critical areas like healthcare for medical imaging and COVID-19 diagnosis using chest x-rays as well as for unknown object detection used in autonomous vehicles.","categories":["Deep Tech"],"tags":["Image Classification"],"author_name":"Poulomi Chatterjee","publish_date":"2022-03-23T14:00:00","publication_year":"2022","word_count":782,"keywords":["Hugging Face","machine learning","OpenAI","AI","neural network","Transformers","computer vision","Ray","Aim","deep learning","Image Classification"],"extracted_tech_keywords":["AI","machine learning","deep learning","neural network","computer vision","OpenAI","Aim","Ray","Hugging Face","Transformers"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/how-do-zero-shot-one-shot-and-few-shot-learning-differ\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10044401,"title":"New Algorithm Improves ML Model Training Over The Internet","content":"“Training BERT costs about $7,000, and for the largest models like GPT-3, this number can be as high as $12 million.” Typically, training a deep learning model starts with a forward pass where loss functions are evaluated followed by a backward pass where the loss-compensating gradients are generated, which are then pushed to servers and updated. These servers aggregate the updates from all the users and make changes to the global machine learning model. Now, this procedure repeats itself multiple times until it hits a certain accuracy. State of the art models are large and involve heavy compute. As models become bigger, the training process continues to remain an expensive affair. Distributed training was introduced to avoid restricting research to just well funded labs. Volunteer computing (VC) is popular with other domains such as bioinformatics and physics where people donate the idle time of their desktops, smartphones, and other personal devices to solve a computationally hard problem. Imagine lending your friend’s PC to train your deep learning model remotely while they are away. Landscape of collaborative computation Folding@home or FAH is a distributed computing project for simulating protein dynamics, including protein folding and the movements of proteins apropos a variety of diseases. FAH brings together volunteers (citizen scientists) to run simulations of protein dynamics on their personal computers. Insights from these data help scientists to better understand biology and provide new opportunities for developing therapeutics. For example, in Folding@home, over 700,000 volunteers have collectively contributed 2.43 exaFLOPs of compute to COVID-19 research in April of 2020 The Berkeley Open Infrastructure for Network Computing or BOINC app, allows downloading of scientific computing jobs on a user’s personal computer and runs the workload in the background. For instance, the Rosetta@home is a distributed computing project for protein structure prediction on the BOINC platform. Rosetta can tap into the computational power of idle computers to help with projects related to designing new proteins and to predict their 3-dimensional shapes. MLC@Home is a distributed computing project dedicated to understanding and interpreting complex machine learning models, with an emphasis on neural networks. It uses the BOINC distributed computing platform. MLC@Home is another project on BOINC that provides an open, collaborative platform for ML researchers. It allows them to train thousands of networks in parallel, with tightly controlled inputs, hyperparameters, and network structures. However, distributed training still has few problems: Distributed training of a single model requires significantly more communication and does not allow a natural way to “restart” failed jobs. Distributed training of neural networks are bounded by the throughput of parameter servers and the memory available on the weakest GPU. “Is there really no alternative to using pre-trained models for the broader ML community?” According to Hugging Face (HF)–whose NLP libraries are used by companies such as Apple– data transfer in distributed deep learning is still a bottleneck. This can arise due to the need to aggregate the gradients from multiple workers and as most participants don’t have high speed connections, they run the risk of getting dropped from the network. “So how on Earth can you train anything with a household data plan?” asks the team at HF. Now, a team of researchers from Yandex, HF and others have come up with a new method that lets machine learning models train over the internet in a better way. The new training algorithm is called Distributed Deep Learning in Open Collaborations (or DeDLOC) About DeDLOC Image credits: Paper by Diskin et al., Data parallelism in GPUs is a popular technique. DeDLOC tries to employ best of all parallelism attributes while tweaking the popular distributed training techniques. DeDLOC incorporates synchronous data-parallel training with fixed hyperparameters regardless of the number of volunteers. Training is done with extremely large batches to compensate for slow communication. According to the researchers, each device accumulates gradients at its own pace until the collaboration reaches the target batch size. Once ready, the collaborators exchange their gradients and perform one optimiser step. ​​DeDLOC operates similarly to BitTorrent and I2P where individual peers coordinate by forming a Distributed Hash Table. To test DeDLOC’s performance, the researchers picked the sahajBERT language mode. The experiment had 40 volunteers,  30 of whom were Bengali-speaking. Volunteers were asked to open the provided notebook (Colab\/Kaggle) locally and run one code cell and watch the training loss decrease on the shared dashboards. The cumulative runtime for the experiment was 234 days. At the end of training, sahajBERT was compared with three other pretrained language models: XLM-R Large, IndicBert, and bnRoBERTa. The results showed that DeDLOC, when applied on pretraining sahajBERT  achieves nearly state-of-the-art quality with results comparable to much larger models that used hundreds of high-tier accelerators. This is the first distributed deep learning training at scale and the results are encouraging for individual researchers looking to take up expensive ML training tasks. “The community for any language can train their own models without the need for significant computational resources concentrated in one place,” wrote the HuggingFace team.","excerpt":"Volunteer computing (VC) is popular with other domains such as bioinformatics and physics where people donate the idle time of their desktops, smartphones, and other personal devices.","categories":["AI Trends"],"tags":[],"author_name":"Ram Sagar","publish_date":"2021-07-23T17:58:04","publication_year":"2021","word_count":829,"keywords":["Hugging Face","machine learning","AI","neural network","ML","distributed computing","RAG","NLP","Colab","deep learning"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","neural network","NLP","Hugging Face","Colab","RAG","distributed computing"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/dedloc-huggingface-language-model-training-distributed\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10002943,"title":"Lights, Camera, AI: The Tech Behind Your Favorite Pixar Movies","content":"Pixar’s “Toy Story,” was the first full-length computer-animated movie released back in 1995. According to an exclusive coverage by Insider, to render “Toy Story,” the animators had 117 computers running 24 hours a day. Each frame could take from 45 minutes to 30 hours to render, depending on how complex it was. There were a total of 114,240 frames to render. Throughout the movie, there are over 77 minutes of animation spread across 1,561 shots. They had to invent new software, called Renderman, to handle all this footage. Twenty-five years down the line, Pixar continues to use new tech, moving the goalposts and setting new benchmarks. It is now using one of the most popular deep learning models, Generative Adversarial Networks (GANs) to generate super high-resolution imagery that can cater to the demands of 8K movie watching experience. Latest machine learning techniques used on Toy Story 4 reduced the final render times by 15-50% in challenging cases. Deep Learning For Super Resolution via Pixar To leverage the deep learning advantages for feature film production, the researchers from Pixar have implemented GANs. The objective here is to create high-resolution images, which is an expensive process in the traditional sense. The team at Pixar, in this work, have explored GANs as an alternative to conventional upscaling techniques. Recently, the state of the art computer vision algorithms such as Deep convolutional neural networks have demonstrated their ability to reconstruct high-quality images by learning the low-resolution (LR) to high-resolution (HR) mapping from a high volume of data. The introduction of GANs, and perceptual loss functions in the seminal SRGAN work has enabled to now produce images with details and sharpness indistinguishable from the ground truth. “The aim of our work at Pixar is to put GANs into production for upscaling.” The training data for experiments with GANs is collected by rendering 1K-2K pairs of production images using RenderMan, with shots randomly sampled from Coco, Incredibles 2, Toy Story 4 and other Pixar movies. The data augmentation techniques implemented in this work also account for the colour correction in the generated images. Any deviation of colour from the ground truth is immediately verified and corrected using an LR image from the training set to maintain coherence in the film. Know how Pixar movies have changed drastically, thanks to automated light pruning techniques. Check this. The training set up consists of a PyTorch development environment, a Linux instance with two 24GB NVIDIA Quadro P6000 GPUs. Pixar also has a renderer to synthesise the pairs of images, high-quality scenes with diverse shaders, geometry, and lighting conditions, and a data center fine-tuned to render at tremendous scale. The researchers claim that they have trained and deployed a production-quality super-resolution model that consistently produces high-quality, artefact-free upscaled images even on scenes with the depth of field or motion blur. Further, the paper states that their latest trained model shows promise towards a pipeline where one can render at 1K and upscale to 2K, which would save 50-75% of the studio’s render farm footprint if used for all intermediate renders. Speaking at the recently concluded VB Transform conference, one of the researchers, Vaibhav Vavilala, Technical Director at Pixar, said that it typically takes at least 50 CPU hours to render one frame at 2K resolution. Now extrapolate that to 24 frames per second for an hour-long movie. With 4K and 8K getting more popular, the rendering becomes as many times longer and tedious. Though today’s computers are designed to handle computational intensive processes, there is still a tradeoff as the computational advantages are negated by increasing demands for more creative and super realistic shots. So, in the end, rendering is still a lengthy and expensive process, and there is a large room for innovation. Link to the paper","excerpt":"Pixar’s “Toy Story,” was the first full-length computer-animated movie released back in 1995. According to an exclusive coverage by Insider, to render “Toy Story,” the animators had 117 computers running 24 hours a day. Each frame could take from 45 minutes to 30 hours to render, depending on how complex it was. There were a […]","categories":[],"tags":["GANs","Movies","Pixar"],"author_name":"Ram Sagar","publish_date":"2020-07-22T11:00:45","publication_year":"2020","word_count":628,"keywords":["Movies","machine learning","AI","neural network","PyTorch","ML","computer vision","RAG","Aim","deep learning","GANs","R","Pixar"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","neural network","computer vision","Aim","PyTorch","RAG","R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/pixar-ai-high-quality-animation-rendering\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10063223,"title":"Talking Ethical AI with Dream11’s Abhishek Ravi","content":"Founded in 2008, Dream Sports is the parent company of Dream11 and FanCode, and Dream Capital, and serves over 140 million users. The Mumbai-based sports content and commerce platform has invested heavily in emerging data science technologies such as AI, ML, and data science. “AI, data, and machine learning play an indispensable role in our company across functions and processes, from start to finish. At Dream Sports, we leverage AI to make the experience better for our users, deliver top-notch customer support and regularly roll out new features that enhance fan engagement on our platform. We also use AI to craft personalised, contextual, and timely engagement campaigns to have an awesome user experience. For that, user-level behavioural insights are a must. For example, from a user perspective, participants benefit from machine learning-driven recommendations on which contests to join based on their past in-app behaviour and other factors,” said Abhishek Ravi, Chief Information Officer, Dream Sports. In an exclusive interview with Analytics India Magazine, Abhishek Ravi spoke about how Dream Sports embeds ethics into Dream11, its AI-based fantasy sports platform. “Transparency is a critical part of Dream sports’ overall success. It is more than just a buzzword for us,” he said. Excerpts AIM: What are the AI frameworks\/methods\/techniques Dream Sports uses to optimise user experience? Abhishek Ravi: Our flagship brand, Dream11, hosts 120 million+ users on its platform on match days and gives them the option to explore thousands of fantasy sports contests across a variety of sports. Our users can actively engage with real-life sporting events and showcase their sports knowledge. With millions of users logged on the platform simultaneously, it can become challenging to provide seamless, best possible experiences on the app every day. To address this, we continuously experiment with multiple features on our app and understand user behaviour such as time taken to navigate through different screens to complete the journey, total time spent on a particular screen to see user navigation patterns from conversions\/drop-offs stages and more. We strive to provide the best user experience, and, in this journey, our hero is DataAware – a self-service funnel analytics tool used by the tech and product teams at Dream11. AIM: What explains the growing conversations around AI ethics, responsibility, and fairness? Why is it important? Abhishek Ravi: The evolution of AI is creating new opportunities to improve people’s lives worldwide, from business to healthcare to education. It also raises new questions about how to build fairness, interpretability, privacy, and security into these systems. It is essential that in such times, not just tech leaders but all companies deploying AI must make an effort to raise awareness. Even the most sophisticated technology companies face problems, and AI and ML only go so far to teach machines what the human knows. Hence, thorough and careful consideration and regular risk assessment are essential to achieve a well-functioning model or product. Constant monitoring, regular user feedback loops, human judgement, and good governance practices are necessary to achieve the right ethical balance with AI use. AIM: How does the Dream11 team ensure compliance to your AI governance guidelines? Abhishek Ravi: We work in a cohesive pod structure, called Dream Teams which include 11 to 15 people across tech products, design, customer support team, etc., to ensure that every business problem we address has the best of both experiential and experimental minds, thus building the best product for users. Our Dream Teams involve a mix of fresh experimenting minds and experienced engineers. This helps us go above and beyond the mainstream ways to solve a problem and test whether the product is sustainable and if it will work for a larger audience to be aligned with our existing company practices. We also build many home-grown solutions that address key product-related challenges such as user experience analysis, scale management, mobile app automation, security, FairPlay, etc. Our early adoption of Cloud also helps our teams quickly test out new features, scale tests in load\/stress environments and drive maximum efficiency. We follow generally accepted industry standards to protect the personal information submitted to us during transmission and once we receive it for storage\/disposal. When you enter sensitive information on our registration or order forms, we encrypt that information using secure socket layer technology (SSL). All information we gather is securely stored within databases controlled by us. The databases are stored on servers secured behind a firewall; server access is password-protected and strictly limited. We also conduct regular internal and external audits to ensure that the right governance policies are followed. AIM: How do you mitigate biases in your AI algorithms? Abhishek Ravi: Biased human judgments can affect AI systems in various ways. It could be in the data generated by existing systems or in the way that algorithms that are designed learn from these systems. People tend to trust outputs from an AI\/ML system once the initial moat is crossed. So, if human bias is missed in training, it could lead to operational, legal or ethical challenges that may be hard to recover from. There are frameworks to ensure that algorithms don’t pick up on biases influenced by human decision-making and make AI\/ML systems fair. Each problem requires a different solution and a different set of data resources with constant re-validation. While there is no single model to follow that will avoid bias, there are parameters that can inform your team. It builds frameworks from some of the world’s leading research labs that can be incorporated in production releases. While excluding sensitive information from the model may seem like a workable solution, it still has vulnerabilities. One must communicate with the data scientists to identify the best model for a given situation. Also, it’s better to have an independent AI committee that overlooks the applications and learnings of the AI algorithms and their implementations. AIM: Do you have a due diligence process to make sure the data is collected ethically? Abhishek Ravi: Transparency and ethics in data collection isn’t a matter of choice. Due diligence is a part of our operations and efforts to keep the market creative, competitive and thriving. Users are already sceptical about their data being collected, so companies must double down on their efforts to adhere to values and standards that factor in consent, laws and regulations and monitoring. When we use third parties to assist us in processing the personal information of users, we make sure that they comply with our privacy policy and other appropriate confidentiality and security measures that we take to prevent fraud or imminent harm, and ensure the security of our network. AIM: How does Dream11 ensure consumer data privacy? Abhishek Ravi: Dream11 takes the utmost care with data security. To ensure a seamless user experience for everyone who logs on to Dream11, which includes privacy to the user’s data, one of the key requirements for us was to understand user behaviour and preferences. This involves mapping and analysing a series of events that a user performs when they log into the app, a journey that starts with user engagement in a mobile app and ends in joining a contest. Multiple Dream Teams are involved in ensuring that the users get the best possible product experience from our end in the safest way possible. This also includes developing ML models to detect fraud or fake accounts on the platform. For instance, Dream11 has a specialised system so that users participating in every contest, including the paid ones, win in a fair, square and transparent manner. FENCE (Fairplay Ensuring Network Chain Entity) is Dream11’s in-house fraud detection system. It is powered by a graph database responsible for processing and maintaining all models and heuristics so that Fair Play Violations are detected timely and efficiently. By infusing our long-standing principles and ethical thinking, Dream11 has ensured consumer data privacy.","excerpt":"FENCE (Fairplay Ensuring Network Chain Entity) is Dream11’s in-house fraud detection system.","categories":["AI Features"],"tags":["AI fairness","ai governance","Cloud services","consumer protection","Data Governance","Data Privacy","data rights","digital transformation","Ethical AI","Interviews and Discussions","modernisation"],"author_name":"Sri Krishna","publish_date":"2022-03-22T14:00:00","publication_year":"2022","word_count":1300,"keywords":["consumer protection","data rights","TPU","ai governance","fraud detection","data science","digital transformation","Cloud services","RAG","Data Governance","analytics","modernisation","machine learning","AWS","AI","Ethical AI","ML","Data Privacy","AI fairness","Aim","Interviews and Discussions"],"extracted_tech_keywords":["AI","machine learning","ML","data science","analytics","Aim","RAG","fraud detection","AWS","TPU"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/talking-ethical-ai-with-dream11s-abhishek-ravi\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10087088,"title":"Rust Turns GitHub’s Long-standing Problem to Dust","content":"Like the fungus after which it was named, Rust has become one of the most popular programming languages. The latest testament to Rust’s popularity comes with GitHub’s announcement of a new code search engine that can function at the GitHub scale. Written in Rust, it creates and incrementally maintains a code search index shared by Git blob object ID. In a blog post on Monday, GitHub engineer Timothy Clem explained about BlackBird, a code search engine written completely in Rust. It currently provides access to almost 45 million GitHub repositories. Sifting through thousands of codes needs something more capable than grep, a command line used to filter through plain text data and figure out a particular pattern. Using ripgrep to run a specific regular expression query on a 13GB file in memory takes about 2.769 seconds or 0.6GB\/sec\/core. The Blackbird works faster as it delves into 640 queries per second. Since its indexing rate is about 120,000 documents per second, processing 15.5 billion documents takes around 36 hours, or 18 for re-indexing since delta indexing reduces the documents to be crawled. To keep the search index manageable, GitHub breaks the data into pieces using Git’s content-addressable hashing scheme and delta encoding to reduce the data and metadata to be crawled. This works well because GitHub has plenty of redundant data that can be compressed through deduplication data-shaving techniques. (Join the GitHub Code Search Engine beta) Why not an open-source solution? There is a plethora of open-source solutions to choose from, like Apache Cassandra, Solr, or Elasticsearch, instead of building search engines from scratch. So then, why did GitHub opt for the long haul? Clem answered in a GitHub Universe video presentation that GitHub failed to find success using general text search products to power code search. The user experience could be better, indexing could be faster, plus it is expensive. He wrote that some newer, code-specific open-source projects are out there, but they don’t work at GitHub’s scale. GitHub started experimenting with Elasticsearch in 2011. However, it sometimes introduces breaking changes that need to be adapted. As Clem noted, it took months to index roughly eight million repositories. GitHub originally used Solr for search but moved to ElasticSearch due to an excessive need for storage space. Today, GitHub supports about 200 million dynamic code repositories. The latest Rust-written engine, Blackbird, supports search across about 45 million repositories, providing partial coverage. It still enables search across 15 terabytes of code and 15.5 billion documents for programs written in Java, Python, and JavaScript. In Rust, We Trust The Blackbird project is a crucial point for Rust, which is usually adopted to build new features for projects originally written in C\/C++. For instance, last year Microsoft Azure CTO declared that all new projects must be written in Rust over C\/C++ because of its memory safety features. Research shows that memory safety issues have accounted for 60% to 70% of all security vulnerabilities across operating systems. The Google security blog sees Rust as a practical language for implementing the kernel. They highlight that Rust helps reduce the number of potential bugs and provides higher security, establishing it as a good choice for creating such a critical component. Moreover, thanks to manual memory allocation and low-level commands, Rust is perfect for programming hardware devices with tiny processors and limited RAM, such as microcontrollers. It can become the top language for the Internet of Things (IoT). Rust can also become a good patch for performance-sensitive, back-end services. For instance, Tilde used Rust to rewrite some Java HTTP endpoints. As a result, they have reduced memory usage by 100 times. The Bottomline BlackBird is being called a game changer as Github’s default search is not the best option for programmers, as it strips most “special” characters out and doesn’t support regular expression search. Another option is Sourcegraph, which is expensive at $90 per active user per month. Having a better search built into Github would be an economical prime option. Even though GitHub’s latest search engine is being lauded, users are comparing the product to StackExchange, and some even called it ‘Google Scholar for Code’. The current use cases seem obscure, but it can indirectly be a major productivity booster for GitHub CoPilot.","excerpt":"The search engine is written in Rust from scratch","categories":["Global Tech"],"tags":["GitHub","Rust language"],"author_name":"Tasmia Ansari","publish_date":"2023-02-10T13:30:00","publication_year":"2023","word_count":706,"keywords":["Elasticsearch","Go","Rust","AI","R","ML","RAG","Python","Rust language","JavaScript","GitHub","Azure"],"extracted_tech_keywords":["AI","ML","RAG","Azure","Elasticsearch","Python","R","JavaScript","Go","Rust"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/rust-turns-githubs-long-standing-problem-to-dust\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10089532,"title":"Stop Questioning OpenAI’s Open-Source Policy","content":"OpenAI has not opened its AI so much. While there are reasons for that, the research organisation’s cofounder, Ilya Sutskever, said, “We were wrong. Flat out, we were wrong”. The chief scientist of the company, believes that AI, or AGI, at some point will become extremely and unbelievably potent and therefore open-sourcing it might not be such a good idea. Amid the hype around AI models, with Google and Meta pressing into the field, Sutskever said that initially the utmost priority of the company behind not open-sourcing its model was to stay ahead of the competition. But now looking at the potency and risks posed by these open AI models, OpenAI has even more reasons to keep the technology to itself. Read: Doomsday Will Be Triggered By GPT-4 The company acknowledged the “chaotic” potential this technology has in its paper. The paper of GPT-4 highlights how models like these might develop “power-seeking actions” as they increasingly become “agentic” in nature and develop their own goals. OpenAI decided to dive deeper into this and gave early-access of the model to Alignment Research Center (ARC) for analysing these behaviours. The reasons why the company is keeping the back-end of the technology with themselves sound plausible now. It is quite interesting to note that the person working on these projects that are taking the world by the storm somehow believes that these technologies might be risky. The hallucinations that ChatGPT and other similar models have are not hidden from anyone. Laughing at the spewed out false information by the chatbot was one of the highlights of Twitter for a long time. This eventually led us to discover the potential dangers it has as well. Sutskever said, “At some point it will be quite easy, if one wanted, to cause a great deal of harm with those models”. But for now, the hype around GPT-4 was just as big as the hype for ChatGPT. But the AI community was very disappointed after reading the 98-pages-long research paper and finding out that the paper provides bare minimum information. It does not even mention the number of parameters of the LLM. Read: Why Are Researchers Slamming OpenAI’s GPT-4 Paper? Frustratingly put out by a lot of researchers on Twitter, the whole point of writing a research paper is to make the model reproducible. “What knowledge do researchers even gain from this?” enquired one of the researchers. The paper does not even reveal anything about the dataset that it was trained on either. This is another reason for the closed door policy: To avoid the legal scrutiny that companies like StabilityAI and Midjourney are facing for copyright infringement. Towards this, Sutskever has said that, “Training data is technology. It may not look this way, but it is”. Should OpenAI Change its Name? The AI community is divided on the issue. The debate surrounding whether AI research should be open or closed has been increasing in the community. Since OpenAI is in the spotlight at the moment, the community is circling around it and questioning its motives and decisions behind keeping GPT-4 such a secretive model. Emad Mostaque, the founder of StabilityAI, has put out an open offer on Twitter, telling OpenAI employees that he will match the salary and benefits that they get for those who want to work on actual “open AI”. Open offer to anyone @OpenAI who actually wants to work on Open AI:We will match your salary, benefits etc but you can work on any open source AI projects you like, ours or others.Collaborate, be open and prioritise good outcomes over self interest: https:\/\/t.co\/BRPspJ2RKB https:\/\/t.co\/7l5O5mHJEc pic.twitter.com\/DFIn1wHmA9— Emad (@EMostaque) March 16, 2023 Well, Mostaque has been in the limelight for open-sourcing his company’s best technology, ‘Stable Diffusion’, which has made wondrous innovations around the world. But he still faces scrutiny from artists and even some AI developers about the copyright issues these text-to-image models create. Earlier, Elon Musk has also criticised OpenAI for working behind closed-doors, contrary to the name (OpenAI) and the purpose (open-source) it showed at the start of the company. “I’m still confused as to how a non-profit to which I donated USD 100 million somehow became a USD 30 billion market cap for-profit,” said Musk in a tweet. Meanwhile, Musk is working on an AI platform and making an “OpenAI Rival”, soon to compete with the “wokeness” and the closed-door policy they have. Interestingly, Meta—the company which was scrutinised for a lot of their AI products like Galactica and BlenderBot—has released LLaMA and is making bets into generative AI. Yann LeCun, the chief of AI at Meta, was always heavily concerned about the ethical implications of the generative models but now somehow decided to open source the company’s model. To-Open or No-To-Open, Asks OpenAI There are several merits to open-sourcing AI models as well. Apart from driving more innovation in the community, open-sourcing allows developers to build guardrails around it and point out ways in which the model can be potentially harmful. About this, Sutskever agreed with the critics and said that if more people are allowed to study these models, the company would learn more about the problems as well. When Musk and Sam Altman founded OpenAI in 2015, the introduction of their blog clearly mentions that the company is a non-profit AI research company that would “build value for everyone rather than shareholders” and would focus on “freely collaborating”. But ever since the company is under the hands of Microsoft, it is quite clearly visible that the company has headed for profits and is becoming increasingly product oriented than simply furthering the research. https:\/\/twitter.com\/Carnage4Life\/status\/1635973027455442945 This all boils down to a simple question to ask OpenAI: Do they want to contribute to developing AI or are they now getting more scared of the potential of the technology they are developing themselves? If the latter is true, then the story is scary for everyone and not just the company.","excerpt":"Since OpenAI is in the spotlight, the AI community is circling around it, questioning its motives and decisions behind keeping GPT-4 such a secretive model.","categories":["Global Tech"],"tags":["ai models open source","ChatGPT","Elon Musk","GPT-4","GPT4","Ilya sutskever","Open Source AI","OpenAI","Sam Altman","Yann LeCun"],"author_name":"Mohit Pandey","publish_date":"2023-03-17T15:00:00","publication_year":"2023","word_count":988,"keywords":["Go","ChatGPT","Yann LeCun","Sam Altman","Rust","OpenAI","AI","GPT-4","Elon Musk","Open Source AI","GPT","stable diffusion","generative AI","ai models open source","GAN","Ilya sutskever","R","GPT4"],"extracted_tech_keywords":["AI","generative AI","ChatGPT","OpenAI","R","Go","Rust","GPT","stable diffusion","GAN"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/stop-questioning-openais-open-source-policy\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10093264,"title":"Meta Unveils AI Sandbox to Empower Advertisers","content":"Big tech companies are throwing generative AI at just about everything. Recently, Meta introduced the AI Sandbox, a testing playground for early versions of new tools and features, including generative AI-powered ad tools. The team is currently developing tools such as text variation, background generation, and image cropping. Text variation generates multiple versions of text to emphasise important messages, while background generation creates custom images from text inputs, allowing for a greater variety of creative assets. With image cropping, advertisers can easily adjust their assets to fit different surfaces, saving time and resources. Meta is currently working with a small group of advertisers to gather feedback and plans to gradually expand access to more advertisers starting in July, with plans to add some of these features to its products later this year. Meta has also released new Meta Advantage features, such as the ability to switch manual campaigns to Advantage+ shopping in one click, using video creative in catalog ads, rolling out performance comparisons, and allowing Advantage+ audience users to use audience inputs as suggestions to guide who sees an ad. The company is investing billions of dollars in infrastructure, with an increasing portion dedicated to building AI capacity for ads. Building products with generative AI capabilities also means investing more in computing capability. As part of its ongoing efforts to advance AI modelling, Meta has implemented larger, more complex models in its ads system. This has led to improved performance and measurement, even when access to granular data is limited. For instance, previously on Instagram, optimising clicks on Story ads required a separate model from optimising conversions or sales for ads on Reels. However, with the development of more advanced AI modelling that optimises across all surfaces (Feed, Story, Explore, and Reels), Meta can transfer learnings across multiple surfaces simultaneously, helping improve advertiser conversions and enhance the quality of ads that people see.","excerpt":"Meta is implementing larger, more complex models in its ads system","categories":["AI News"],"tags":["Meta"],"author_name":"Ayush Jain","publish_date":"2023-05-12T18:30:11","publication_year":"2023","word_count":314,"keywords":["Go","Meta","programming_languages:R","AI","programming_languages:Go","generative AI","R"],"extracted_tech_keywords":["AI","generative AI","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/meta-unveils-ai-sandbox-to-empower-advertisers\/","complexity_score":2,"technical_depth":6,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10059307,"title":"Almost 80% of financial companies use AI: NVIDIA report","content":"According to an NVIDIA report, 78% of financial services professionals claim their companies use accelerated computing to deliver AI-enabled applications through ML, deep learning or high-performance computing. NVIDIA’s “State of AI in Financial Services” report is based on responses from over 500 C-suite executives, developers, data scientists, engineers and IT teams working in financial services. Fraud detection involving payments and transactions was the top AI use case at 31%, followed by conversational AI at 28%, and algorithmic trading at 27%. The number of financial institutions investing in AI have also gone up exponentially. AI for underwriting increased fourfold, from 3% penetration in 2021 to 12% this year. Conversational AI jumped from 8% to 28% year over year. AI-enabled applications for fraud detection, know your customer (KYC) and anti-money laundering (AML) saw growth of at least 300 percent. Top current AI use cases in financial services Green highlighted text signifies top AI use cases for investment in the next six to 12 months. Almost half of the respondents (47%) said AI enabled more accurate models for applications such as fraud detection, risk calculation and product recommendations. Only 16% of survey respondents agreed their companies are spending the right amount of money on AI, and 37% believe “lack of budget” is the primary challenge in achieving AI goals. Less availability of data scientists, lack of data, and explainability, were the other major challenges listed. Over half of C-suite respondents said AI is important to their companies’ future success. To the question “How does your company plan to invest in AI technologies in the future?” the top responses included: stepping up hiring of AI experts (43%), identifying additional AI use cases (36%) and engaging third-party partners to accelerate AI adoption (36%). However, only 23 percent believe their companies have the capability and knowledge to move an AI project from research to production.","excerpt":"NVIDIA’s report is based on responses from over 500 C-suite executives, developers, data scientists, engineers and IT teams working in financial services.","categories":["AI News"],"tags":["AI adoption","AI Companies","Deep Learning","fraud detection","ML","NVIDIA"],"author_name":"Meeta Ramnani","publish_date":"2022-01-28T12:52:49","publication_year":"2022","word_count":309,"keywords":["Go","AI adoption","programming_languages:R","AI","ML","programming_languages:Go","Aim","deep learning","AI Companies","Deep Learning","NVIDIA","R","fraud detection"],"extracted_tech_keywords":["AI","ML","deep learning","Aim","fraud detection","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/almost-80-of-financial-companies-use-ai-nvidia-report\/","complexity_score":4,"technical_depth":9,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10120994,"title":"Yotta Appoints Anil Pawar as Chief AI Officer, Marks Next Phase in AI Strategy","content":"Yotta Data Services has announced the appointment of Anil Pawar as its chief AI officer and head of its AI Cloud Business Unit. This move marks the next phase in Yotta’s AI strategy and builds on the company’s multi-million-dollar investment in its AI offerings. In his new role, Pawar will oversee strategic initiatives such as AI-as-a-service (AIaaS), AI platform-as-a-service (AIPaaS), AI software-as-a-service (AISaaS), and the large language model (LLM) marketplace within the Shakti Cloud Business Unit. He will report directly to Yotta co-founder, MD and CEO Sunil Gupta. Pawar’s appointment comes as Yotta transitions its focus towards cloud services platforms, such as Shakti and Yntraa, aligning with the company’s strategic shift. His most recent stint was as the head of tech strategy at Rakuten Japan, where he oversaw AI programs and AI cloud transformation initiatives. He has also been instrumental in establishing one of the world’s largest data-driven networks at Reliance Jio. Commenting on the appointment, Sunil Gupta said, “Anil brings with him his rich experience in the field of AI and digital transformation. I’m optimistic that in his new role at Yotta, he will help strengthen our position and help our clients unlock their true potential.” Pawar, an industry veteran with over 25 years of experience and multiple US patents in diverse communication technologies, expressed his excitement about joining Yotta. “The opportunity to leverage AI and cloud technologies to drive innovation and create value for customers is truly enticing. I look forward to working with the talented team at Yotta to deliver cutting-edge solutions and drive growth in the Shakti AI Cloud Business Unit,” he said. With Pawar at the helm of its AI initiatives, Yotta aims to reinforce its commitment towards enabling India’s digitisation while driving growth in the Shakti AI Cloud Business Unit.","excerpt":"Pawar will oversee initiatives such as AI-as-a-service (AIaaS), AI platform-as-a-service (AIPaaS), AI software-as-a-service (AISaaS), and the LLM marketplace within the Shakti Cloud Business Unit.","categories":["AI News"],"tags":["Yotta Data Services"],"author_name":"Shyam Nandan Upadhyay","publish_date":"2024-05-20T10:56:14","publication_year":"2024","word_count":296,"keywords":["programming_languages:R","AI","data-driven","digital transformation","innovation","Git","RAG","Aim","R","Yotta Data Services"],"extracted_tech_keywords":["AI","Aim","RAG","R","Git","digital transformation","data-driven","innovation","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/yotta-appoints-anil-pawar-as-chief-ai-officer-marks-next-phase-in-ai-strategy\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10004591,"title":"Hands-On-Implementation of Lasso and Ridge Regression","content":"If you are into the domain of building predictive models using machine learning then you might have come across a situation when your models perform very good on the training data and not so good on testing. This happens when the dataset is much complex like in Neural Networks. This problem of not generalization well in testing data is called Overfitting of the model. It occurs when the error rate of a model is higher in testing as compared to training. What to do in these situations? Regularization is a technique that can help in these cases by adding some more information. It is also called Shrinkage Methods. Regularization techniques like Lasso Regression and Ridge Regression are used to avoid overfitting. The main idea behind regularization is to decrease the variance and increase bias a bit. In this article, we will explore both the methods of regularization and check the results if we get rid of the overfitting situation. For this, we will use the Boston House Dataset where we will predict the prices of the house. The data set can be downloaded from Kaggle where it is publicly available. What we will learn from this article? Overfitting of the Model Regularization Ridge RegressionLasso Regression Polynomial Models Ridge Regression It is also called an L2 regularization that is used to get rid of overfitting. The goal while building a machine learning model is to develop a model that can generalize patterns well in training as well as in testing. Refer to the below graph that shows the best fit line for training and testing data. The line is perfect for the training data points that means the sum of residual is equal to 0 whereas the sum of residual for testing is high. This is termed as Overfitted model. Ridge Regression is done to improve the generalizability of the model. This is done by tweaking the slope of the best fit line. Maybe the model does not perform much well in the training because now the line does not pass exactly to the data points but it will give fairly good results in testing. The slope is changed or the line is titled a bit by making use of the penalty term called Alpha which is a hyperparameter. Linear regression aims to reduce the sum of squared errors whereas in ridge regression it also reduces the sum of squared error but adds this penalty term by multiplying it with slope square. Check below for the graphical representation of the best fit line using ridge regression. Linear regression = min(Sum of squared errors) Ridge regression = min(Sum of squared errors + alpha * slope)square) As the value of alpha increases, the lines gets horizontal and slope reduces as shown in the below graph. Lasso Regression It is also called as l1 regularization. Similar to ridge regression, lasso regression also works in a similar fashion the only difference is of the penalty term. In ridge, we multiply it by slope and take the square whereas in lasso we just multiply the alpha with absolute of slope. Lasso Regression = min(sum of squared error + alpha * | slope| ) Similar to ridge regression as you increase the value of the penalty term the slope will get reduced and the line will become horizontal. As this term is increased it becomes less responsive to the independent variable. Let us now practically see both the regularization techniques with implementing a regression model for Boston Housing Dataset. First, we will import all the required libraries and the data set. After importing we will explore a bit data like shape and about missing values present in the data set. Use the below code to do the same. import pandas as pd from sklearn.linear_model import LinearRegression from sklearn.linear_model import Ridge from sklearn.linear_model import Lasso df = pd.read_csv('Boston.csv') print(df) Output: print(df.shape) print(df.isnull().sum()) Output: The data set contains 506 rows and 15 columns. There are no missing values that are found in the data. We will not divide the dependent variable and independent variable X and y respectively followed by scaling the data and then dividing it into training and testing sets. Use the below code to do so. X = df.drop('medv', axis=1) y = df['medv'] from sklearn import preprocessing X = preprocessing.scale(X) y = preprocessing.scale(y) from sklearn.model_selection import train_test_split X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.30, random_state=1) print(X_train.shape) print(y_train.shape) print(X_test.shape) print(y_test.shape) Output: There are a total of 354 rows in the training data set and 152 are present in the testing data. We now build three models using simple linear regression, ridge regression and lasso regression and fit the data for training. After the model gets trained we will compute the scores for testing and training. Use the below code for the same. regression_model = LinearRegression() regression_model.fit(X_train, y_train) ridge = Ridge(alpha=.3) ridge.fit(X_train,y_train) print (\"Ridge model:\", (ridge.coef_)) Output: lasso = Lasso(alpha=0.1) lasso.fit(X_train,y_train) print (\"Lasso model:\", (lasso.coef_)) Output: print(\"Linear Regression Model Training Score: \", regression_model.score(X_train, y_train)) print(\"Linear Regression Model Testing Score: \",regression_model.score(X_test, y_test)) print(\"Ridge Regression Model Training Score: \",ridge.score(X_train, y_train)) print(\"Ridge Regression Model Testing Score: \",ridge.score(X_test, y_test)) print(\"Lasso Regression Model Training Score: \",lasso.score(X_train, y_train)) print(\"Lasso Regression Model Testing Score: \",lasso.score(X_test, y_test)) Output: The results are almost identical but with less complexity of the models. We will now create a polynomial regression model by creating new features from the features followed by transforming the data and dividing it into training and testing. Use the below code to do so. from sklearn.preprocessing import PolynomialFeatures poly = PolynomialFeatures(degree = 2, interaction_only=True) X_poly = poly.fit_transform(X) X_train, X_test, y_train, y_test = train_test_split(X_poly, y, test_size=0.30, random_state=1) regression_model.fit(X_train, y_train) print(regression_model.coef_[0]) Output: ridge = Ridge(alpha=.3) ridge.fit(X_train,y_train) print (\"Ridge model:\", (ridge.coef_)) Output: lasso = Lasso(alpha=0.003) lasso.fit(X_train,y_train) print (\"Lasso model:\", (lasso.coef_)) Output: We will now check the scores of the polynomial model and compute the training and testing scores. Use the below code to do so. print(\"Linear Regression Model Training Score: \", regression_model.score(X_train, y_train)) print(\"Linear Regression Model Testing Score: \",regression_model.score(X_test, y_test)) print(\"Ridge Regression Model Training Score: \",ridge.score(X_train, y_train)) print(\"Ridge Regression Model Testing Score: \",ridge.score(X_test, y_test)) print(\"Lasso Regression Model Training Score: \",lasso.score(X_train, y_train)) print(\"Lasso Regression Model Testing Score: \",lasso.score(X_test, y_test)) Conclusion Regularization is done to control the performance of the model and to avoid the model to get overfitted. In this article, we discussed the overfitting of the model and two well-known regularization techniques that are Lasso and Ridge Regression. Lasso regression transforms the coefficient values to 0 which means it can be used as a feature selection method and also dimensionality reduction technique. The feature whose coefficient becomes equal to 0 is less important in predicting the target variable and hence it can be dropped.  Ridge regression transforms the coefficient values to close to 0 and not completely equal to 0. You can also explore more about Ridge and Lasso here in this article titled “Ridge Regression Vs Lasso Regression”.","excerpt":"In this article, we will explore both the methods of regularization and check the results if we get rid of the overfitting situation. For this, we will use the Boston House Dataset where we will predict the prices of the house. The data set can be downloaded from Kaggle where it is publicly available.","categories":["Deep Tech"],"tags":["regression analysis","regularization techniques","ridge regression"],"author_name":"Rohit Dwivedi","publish_date":"2020-08-11T17:00:00","publication_year":"2020","word_count":1141,"keywords":["Go","machine learning","TPU","programming_languages:R","AI","ridge regression","neural network","regularization techniques","regression analysis","RPA","Aim","R","Pandas"],"extracted_tech_keywords":["AI","machine learning","neural network","Aim","Pandas","TPU","R","Go","RPA","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/hands-on-implementation-of-lasso-and-ridge-regression\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10019748,"title":"Data Science Hiring Process At Fiat Chrysler Automobiles India","content":"Fiat Chrysler Automobile (FCA), now called Stellantis post its merger with Groupe PSA, has a dedicated Data Services CoE that is integral to FCA’s growth plans. This horizontal centre of excellence operates with a high degree of autonomy and provides analytics and data science expertise across the various business IT verticals. The company recently committed $150 million to set up a new Global Digital Hub in Hyderabad, slated to create almost 1,000 cutting-edge technology jobs by the end of 2021. The company will hire in niche technology areas such as AI, connected vehicles and cloud technologies. “The Data Services CoE is currently working on a wide and exciting portfolio of problem statements. With strong demand for data science and analytics skills at Stellantis, we will continue to recruit actively and scale the CoE,” said Karim Lalani – Director and Head of Global Digital Hub, FCAIT Automotive India, in conversation with Analytics India Magazine. Through the Data Services CoE, the company aims to build an innovation powerhouse to harness the intelligence enabled by data to create exciting products and services and quickly adapt to customers’ evolving needs. “Our data scientists are already driving innovation in multiple areas, including customer safety, connected mobility and our digital showroom experience,” said Lalani. Required Skills At the Information and Communication Technology (ICT) centre, the company looks for candidates with core technical competencies in data engineering, data analytics or data sciences. However, domain expertise is the key to how they identify and recruit new talent. They look for candidates with a clear understanding of business pain points and the right approach to directly impact the end customer. In terms of educational background, most of the positions at Stellantis are focused on candidates with a Bachelor’s degree in Engineering. However, they are also open to those with a Master’s or PhD for more niche roles. They also look for candidates who have graduated from universities that offer specific data science curriculum or have demonstrated their passion for the domain by completing recognised data analytics\/ data science certifications. “A strong understanding of technical concepts, algorithms and data models will certainly enable candidates to dive right in and contribute to the exciting products and services that we are building at our Data Services CoE,” said Lalani. While educational background certainly provides the right foundation, at ICT India, they prefer candidates who possess the drive to continually learn, collaborate with global teams, passion for driving change and owning the work they deliver. That said, strong quantitative and programming skills are crucial to getting hired. Being A Data Scientist At Stellantis An ideal data science candidate at the company should have: An ability to understand the big-picture strategy and navigate through the technical complexities at the execution level.A mix of business acumen and technical skills. The ideal candidate should adopt a consultative approach to work and possess strong communication skills to translate technical concepts into simpler terms for our business partners. “We also would like our data scientists to help us understand ‘the art of the possible’ that intelligence enabled by ‘data’ can create in the context of our business,” said Lalani. Further, the company has a clear career path designed for employees. It believes in providing a continually challenging and rewarding work environment that encourages people to pursue lifelong careers at Stellantis. As data scientists, they can either explore a technical or managerial career path depending on their evolving interests. “At the same time, we offer a wide range of learning opportunities ranging from technical certifications to higher degree education programs to support the learning goals of our employees,” he stated. Interview Process The Stellantis’ Global Digital Hub has a rigorous hiring process. They look for the best technical data science and data engineering experts with strong communication skills and an entrepreneurial mindset. Lalani explained the process: The interview process typically spans four weeks, during which the data science candidate is evaluated across various core competencies, including technical, leadership, communication and stakeholder management.The first step is a screening process where the HR team assesses the candidate’s resume. Next, the candidate goes through a technical assessment, where he or she is asked to work on real-world data science problem statements.A candidate who aces the technical assessment is then invited to final interview discussions with the North American and Indian teams. This discussion provides an opportunity for the candidate to learn more about how the Data Services CoE operates in ICT India. The interview questions are primarily structured around evaluating candidate capabilities around the following dimensions: interpersonal skills, logical skills, technical skills and fundamental concepts of data analytics and AI\/ML. Stellantis put the candidates through an interview process that resembles a “day in the life of a Data Scientist or Data Analytics Engineer”. To ensure this, they have adopted an Activity Driven Screening Approach. They evaluate problem structuring, technical rigour (reliability, readability, conciseness of the code developed to accomplish task\/objective, and scalability), analytical rigour (clarity of approach, methodology and conclusions ), the usefulness of code and more. The company is looking to collaborate with startups to bring high-quality talent and augment the internal efforts at developing IPs, solutions and accelerators. Hackathons are also a way to attract cutting-edge talent. “We are also looking forward to partnering with universities to train and ramp up talent to address data science efforts at Stellantis,” added Lalani. Talking about the data science positions, Lalani said they always have new roles opening up and encouraged interested candidates to explore their careers page. Job seekers can also reach out on the Stellantis ICT India Linkedin Page. Recruitment Challenges Lalani said they face their own set of challenges while recruiting for data analytics and sciences roles. “From a skill and competency perspective, we look at candidates who have the right blend of skills across mathematics, statistics, machine learning, programming and business domain knowledge. Getting this blend is a challenge,” said Lalani. The company has invested in training platforms to help the talent at Stellantis address any gaps in terms of knowledge, competencies and skillsets. “As there is a sheer shortage of data skillsets, at Stellantis, we take extra efforts to position the cutting edge nature of work we do to attract best in class talent,” he said. Highlighting some of the hiring mistakes companies make, he said that candidates being hired for data roles need to be moved through the screening and selection funnel much faster due to the high demand for the skillsets. “Competitive Compensation is the key to attract the best in class talent in this space. Involving a cross-functional team of stakeholders is the key to screening and selection of candidates as data sciences\/analytics can never be done in silos,” he said. Pro Tips Lalani said the data offers a large spectrum of possibilities. Therefore, it is imperative to pick a focus area and build deep specialisation to differentiate oneself. “Equip yourself with the right tools programming languages, database management systems, data frameworks, libraries, and data wrangling tools to business intelligence and visualisation software,” he said.","excerpt":"Fiat Chrysler Automobile (FCA), now called Stellantis post its merger with Groupe PSA, has a dedicated Data Services CoE that is integral to FCA’s growth plans. This horizontal centre of excellence operates with a high degree of autonomy and provides analytics and data science expertise across the various business IT verticals.  The company recently committed […]","categories":["AI Hirings"],"tags":["data science curriculum","Statistics for Data Science"],"author_name":"Srishti Deoras","publish_date":"2021-02-05T15:00:00","publication_year":"2021","word_count":1170,"keywords":["data science","Go","machine learning","AI","data science curriculum","ML","Scala","RAG","Aim","Statistics for Data Science","analytics","R"],"extracted_tech_keywords":["AI","machine learning","ML","data science","analytics","Aim","RAG","R","Go","Scala"],"url":"https:\/\/analyticsindiamag.com\/ai-hiring\/data-science-hiring-process-at-fiat-chrysler-automobiles-india\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10085913,"title":"Andrew Ng&#8217;s &#8216;Maths for ML and Data Science Specialisation&#8217; Now on Coursera","content":"Last month, Andrew Ng’s DeepLearning.AI introduced a new course, “Mathematics for Machine Learning and Data Science Specialisation.” The same course is available on Coursera, led by Luis Serrano and co-created by Ng alongside Anshuman Singh, Magdalena Bouza, and Elena Sanina. This new course gives people an intuitive understanding of AI’s most important maths concepts. Enrol for the course for free here. “I have often said ‘don’t worry about it’ when it comes to maths because maths shouldn’t hold anyone back from making progress in ML,” said Ng. He said that understanding some key topics in linear algebra, calculus, probability, and stats will help you better get learning algorithms to work. Further, he said this specialisation was designed with various interactive visualisations to help people see how maths works. “Maths is not about memorising formulas; it is about sharpening your intuition,” added Ng. What to Expect? This is a beginner-friendly programme where people master the fundamentals of mathematics of machine learning. This specialised course is said to use innovative mathematics pedagogy that helps students learn quickly and properly, with courses that use easy-to-follow plugins and visualisations to help students learn how maths behind ML works. Once completed, students will understand the maths behind the algorithms and data analysis techniques. Plus the know-how to implement them into their machine learning career. Outcome: Andrew Ng and the team believe that by the end of this course, students will be able to represent data as vectors and matrices and spot their properties using singularity, rank and linear independence concepts. Also, they will be able to apply common vector and matrix algebra operations like the dot product, inverse and determinants. In addition to this, students will be able to express certain types of matrix operations as linear transformations, alongside applying concepts of eigenvalues and eigenvectors to machine learning problems, optimise different types of functions, perform gradient descent in neural networks with various activation and cost functions, describe and quantify the uncertainty inherent in predictions\/forecasting made by ML and data science, apply common statistical methods like MAP and MLE and more.","excerpt":"Last month, Andrew Ng’s DeepLearning.AI introduced a new course, “Mathematics for Machine Learning and Data Science Specialisation.” The same course is available on Coursera, led by Luis Serrano and co-created by Ng alongside Anshuman Singh, Magdalena Bouza, and Elena Sanina. This new course gives people an intuitive understanding of AI’s most important maths concepts.  Enrol for the course for free here.  […]","categories":["AI News"],"tags":["Courses"],"author_name":"AIM Media House","publish_date":"2023-01-26T13:06:09","publication_year":"2023","word_count":345,"keywords":["data science","Go","machine learning","programming_languages:R","AI","neural network","ML","programming_languages:Go","Courses","R"],"extracted_tech_keywords":["AI","machine learning","ML","neural network","data science","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/andrew-ngs-maths-for-ml-and-data-science-specialization-now-on-coursera\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":22285,"title":"Why Tech Giants Are Rushing To Democratise Machine Learning","content":"Google’s Fei-Fei Li, speaking at Next ‘17, March 8, 2017 Last year at I\/O conference, Google CEO Sundar Pichai announced a better machine ‘learning models’ that would help democratise artificial intelligence beyond organisations and make its tool available for everyone. This January Google had launched cloud-based learning services, Cloud AutoML aimed at hundreds and thousands of developers who does not have expertise in machine learning and artificial intelligence. To make AI more accessible to people, Google is now pursuing to educate more people about artificial intelligence, with free online course like Learn with Google AI and advanced projects like TensorFlow, cat doodles and a machine vision experiment. Not only Google, but other tech companies like Microsoft, Apple, Amazon Web Services (AWS), IBM are also making machine learning and artificial intelligence accessible to everyone. Apple too has joined the race last year to democratise AI by launching a contest to help coders build apps for Apple devices. Even high-tech industry are getting into action. Through Michelangelo Machine Learning, Uber is providing a platform which has capabilities to manage data; to train, evaluate and deploy AI predictive models. As artificial intelligence and machine learning is touching every corner of the industry, democratisation of AI is only obvious. And this year will mark the beginning of AI democratisation. According to a Gartner report, AI is poised to take a prominent place within organisations in 2018. Advances in virtual assistants and deep learning will foster adoption of artificial intelligence and companies and government will use AI in much border range. What Is Driving Democratisation? Of late, there has been a sudden boom in the machine learning platforms, which is letting non-AI employees to tap into the huge amount of data that businesses are sitting on. The shortage of qualified data scientists and machine learning experts has often been highlighted as the reason behind the handbrakes on the adoption of machine learning and data. But the growing number of machine learning tools are putting AI and data analytics in the hand of beginners, non-experts and researchers. After tech giants have started offering API driven tools into their cloud platforms, these machine learning platforms are becoming more accessible. Through these platform, they are simply letting non-AI experts carry out day-to-day analytics without depending much on the company’s data specialist. The tech firms are finding ways to make up for the talent shortage and reduce the dependence on data scientist or machine learning experts. For instance, with the AutoML, Google is producing a targeted machine learning algorithms that doesn’t require human dependence. AutoML is putting the complex artificial intelligence algorithms into the hands of businesses and making them less dependent on AI experts. With the focus towards  more data localisation, startups and government are setting up their own data centres. For instance, under National Informatics Centre, government of India has set up many data centres in states. Telangana government partnered with Nasscom to set up a Centre of Excellence for Data Science and Artificial Intelligence. Reportedly, China is building a city-sized cloud computing complex which will have 6.2 million square feet of building, with 646,000 square feet of mega data centre space. Last year, China have declared a ambitious plan to become AI superpower by 2020. Apart from tech companies like Amazon, Microsoft, Oracle, Jack Ma-led Alibaba Group has opened its first data center in Mumbai, with a aim to expand its cloud services to Indian enterprises. Targeting the growing ecosystem of software developers and startups in India, DigitalOcean have set up its data center in Bangalore. Idea Behind Democratisation Democratisation of artificial intelligence is making machine learning tools and technologies widely available for masses and businesses at an affordable cost. But what is actually driving this AI democratisation? According to a survey conducted by Qualtric, only 10 percent of internet users worldwide consider themselves as AI experts. On the other hand, 53 per cent said they know about AI, but wouldn’t call themselves an expert on the subject. To help provide resources more to the masses to use AI in their work and everyday life, many tech companies are educating and making it more accessible to the masses. The AI-critic Elon Musk had earlier pitched for democratisation of artificial intelligence and is making sure that the AI just don’t get limited in the hands of a set of individuals or one company. In an interview with YCombinator founder Sam Altman, Elon Musk have said that, “It becomes a very unstable situation if you’ve got any incredibly powerful AI. You just don’t know who’s going to control that. It’s not that it will develop a will of its own right off the bat, the concern is that someone may use it in a way that’s bad… So we must have democratisation of AI technology and make it widely available. And that’s the reason that… [we] made OpenAI, to help spread out AI technology so it doesn’t get concentrated in the hands of a few.” There is a growing body of students and researchers from varied fields who are gravitating towards deep learning thanks to Google’s TensorFlow.  There is a huge community engagement with TensorFlow being declared the #1 repository in “machine learning” on Github.  Most of the TensorFlow adopters are Python programmers, leaning towards deep learning projects. Another key point is that Google is betting heavily on its Google Cloud ML engine and this is one of the ways to attract people to Google Cloud. Teaming Up For Democratisation Beside funding their own machine learning platform, tech companies are also partnering with other companies to democratise artificial intelligence and machine learning. In 2017, Microsoft and AWS had introduced a new open source deep learning interface which is intended to make developing solution using machine learning easier and faster. Gluon provides a clear, concise API for defining machine learning models using a collection of pre-built, optimised neural network components. Developers who are new to machine learning will find this interface more familiar to traditional code. More seasoned data scientists and researchers will value the ability to build prototypes quickly and utilise dynamic neural network graphs for entirely new model architectures, all without sacrificing training speed, claims AWS news blog. The tech firms are collaborating with data science startups to further this initiative. Google acquired data science startup Kaggle to make artificial intelligence and machine learning available for developers, users and enterprises. Microsoft’s AI fund is investing in AI startups like Bonsai and Agolo. Bonsai simplifies open-source machine learning libraries like TensorFlow to help enterprises incorporate artificial intelligence into their businesses and also to construct their own AI models. Agolo is a news summary startup that uses artificial intelligence to summarise news for media companies on Facebook or Alexa. Programs like NVIDIA’ Inception, which helps accelerate startups pushing the frontiers of AI and data science has supported more than 2,000 AI startups in about a year. Tech companies are even supporting ventures into areas not touched by machine learning and artificial intelligence. Microsoft’s $50 million AI for Earth initiative to help organisations that are working to solve the climate change crisis. Recently, Google has signed a key cloud computing deal with Flex to provide a standard way for medical devices to shunt data up to the internet for for analysis. This key deal will also help deepen Google’s presence healthcare cloud computing market.","excerpt":"Last year at I\/O conference, Google CEO Sundar Pichai announced a better machine ‘learning models’ that would help democratise artificial intelligence beyond organisations and make its tool available for everyone. This January Google had launched cloud-based learning services, Cloud AutoML aimed at hundreds and thousands of developers who does not have expertise in machine learning […]","categories":["IT Services"],"tags":["AI (Artificial Intelligence)","Alibaba","Amazon","API","Democratisation","Developers","Google","Machine Learning","Microsoft","Oracle","tech companies"],"author_name":"Smita Sinha","publish_date":"2018-03-06T05:14:51","publication_year":"2018","word_count":1224,"keywords":["API","deep learning","tech companies","AI (Artificial Intelligence)","data science","artificial intelligence","analytics","machine learning","Microsoft","AI","Alibaba","neural network","ML","Amazon","Machine Learning","Oracle","OpenAI","Aim","Democratisation","Google","Developers"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","neural network","data science","analytics","OpenAI","Aim"],"url":"https:\/\/analyticsindiamag.com\/it-services\/tech-giants-are-rushing-to-democratise-machine-learning\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":46914,"title":"Step-By-Step Guide To Cracking MachineHack’s Predict The Book Price Hackathon","content":"MachineHack recently launched its latest hackathon called Predict The Price Of Books. This article is a complete step by step guide to the solution. Click here to participate in the hackathon. Predict The Book Price Hackathon The so-called paradoxes of an author, to which a reader takes exception, often exist not in the author’s book at all, but rather in the reader’s head. – Friedrich Nietzsche Books are open doors to an unimagined world which is unique to every person. It is more than just a hobby for many. There are many among us who prefer to spend more time with books than anything else. Here we explore a big database of books. Books of different genres, from thousands of authors. In this challenge, participants are required to use the dataset to build a Machine Learning model to predict the price of books based on a given set of features. Size of training set: 6,237 records Size of test set: 1,560 records Click here to participate in the hackathon. FEATURES: Title: The title of the book Author: The author(s) of the book. Edition: The edition of the book eg (Paperback,– Import, 26 Apr 2018) Reviews: Customer reviews about the book Ratings: The customer ratings of the book Synopsis: The synopsis of the book Genre: The genre the book belongs to BookCategory: The department the book is usually available at. Price: The price of the book (Target variable) Click here to participate in the hackathon. The following Python notebook contains the complete step by step guide to work on the above-mentioned hackathon. Use this notebook to learn and adapt to this work to better your score. Approach Exploring The Data Sets Cleaning, Processing and Generating New Features Building A Regressor Optimising The Hyperparameters Using Bayesian Optimisation Getting The Datasets Go to MachineHack, Sign Up as a user and click on the Predict The Price Of Books Hackathon. Start the hackathon and find the dataset in the Attachment section. Click here to register for the hackathon Without further ado, let’s crack the Hackathon! Click here to participate in the hackathon. The above solution gives an average score of approximately 65 percent accuracy using RMLSE for evaluation. Use the solution, tweak and tune it to better the score. Good Luck !","excerpt":"MachineHack recently launched its latest hackathon called Predict The Price Of Books. This article is a complete step by step guide to the solution. Click here to participate in the hackathon. Predict The Book Price Hackathon            The so-called paradoxes of an author, to which a reader takes exception, often exist […]","categories":["Deep Tech"],"tags":["books","Data Science","Hackathon","Machine learning hackathon","Machinehack"],"author_name":"Amal Nair","publish_date":"2019-10-07T10:20:35","publication_year":"2019","word_count":378,"keywords":["big data","Go","machine learning","Machinehack","Machine learning hackathon","AI","ML","RPA","RAG","Hackathon","Python","books","programming_languages:Python","Data Science","R"],"extracted_tech_keywords":["AI","machine learning","ML","RAG","Python","R","Go","big data","RPA","programming_languages:Python"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/cracking-machinehacks-predict-the-book-price-hackathon\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":69789,"title":"How IoT helps Smart Cities to monitor Resources","content":"Modern cities today are competing for residents, tourists, investors, funding from the government, among others, and a city which is better connected and is smart, has the maximum chance of winning. Talking about connected cities, IoT is the key enabler—CCTV, traffic signals, parking meters and public transportation, roll your eyes and there are plenty of microchip embedded devices connected to Internet of Things (IoT). IoT is seeing a staggering development owing to the huge advantages and  adoption rates it pose. IoT underlining the Smart Cities- Everything from fitness devices to smart phones and appliances to flat panels are getting connected and smarter. Along with these, ‘Smart Cities’ has gained much momentum and has become a buzzword today. But what is a smart city? As many would have heard, a smart city is nothing but an intelligent ecosystem that uses IoT to reach new levels of efficiency. The IoT in Smart City works on M2M bi-directional networking for gathering and processing data through sensors, actuators, smart phones and smart applications, which is then analysed and used to improve the infrastructure, traffic management, public utilities and other areas like weather, pollution (air, water and noise) and energy consumption. IoT is comprised of three layers— perception layer, network layer and application layer. Few things such as RFID, cameras, sensors and GPS could be considered as perception layer as they are the devices that can detect objects, perceive data, gather information and communicate the data with other devices. Transferring data from a perception layer to application layer is the task of network layer, which comprises Wifi, 2G, 3G, 4G, Zigbee, Zwave, LPWAN and other protocols. The last layer, which is the application layer, receives the information and process it to create smart homes, smart cities, power system monitoring, energy management, among others. Incorporating IoT in the building infrastructure – Talking about increased adoption of IoT it wouldn’t be unfair to say that building managers and construction companies from all over the world are looking at ways to incorporate IoT in infrastructure. Amongst these, traffic signals and road crossings are gaining much prominence where sensors are being used to record the data such as number of vehicles passing through a particular road, vehicle direction during the peak hour etc. This information is then used to divert and manage the traffic accordingly and reduce congestion. Smart grids, for instance, is another utility that offer energy solutions to people and meet their requirements in a sustainable manner. Smart electricity meters and gas meters can keep a check on the consumption by consumers hence ensuring there is no wastage of resources. Automated parking services ensure a real time update about the available parking spots in the city nearby to avoid urban traffic congestions. These and many other IoT enabled approaches are becoming a critical part of the Smart City plan and many cities have already adopted these intelligent steps to build smarter societies. Smart City for a sustainable future – With the concept of Smart City hitting the cord, it has been witnessing greater adoption as a way to offer high quality of life to its residents, while being considerate towards the environment. Sustainable solutions like LED bulbs and installation of solar panels are being combined with IoT, which are further delivering improved results. Other smart solutions such as motion detection lights and smart elevators have also become quite popular amongst masses. In 2010, IBM detected that people of the New York City spent 22.5 years’ worth of time waiting for elevators, which calls for a solution to deliver greater efficiency and boost the productivity of people. The possibilities for smart solutions are endless. Energy management, remote asset monitoring and mobile asset tracking, when connected to a single platform, can create a strong foundation for building various customised solutions like metering of energy, better management of utilities, traffic control, public transport management, waste and building management, power distribution and resource management. These can not only save time and minimize costs, but can enable a better utilization of resources. Smart City calls for a creation of an environment which has connected devices around the city, that can manage and control most aspects of our daily lives, and IoT is definitely playing a key role in the overall monitoring of resources.","excerpt":"Modern cities today are competing for residents, tourists, investors, funding from the government, among others, and a city which is better connected and is smart, has the maximum chance of winning. Talking about connected cities, IoT is the key enabler—CCTV, traffic signals, parking meters and public transportation, roll your eyes and there are plenty of […]","categories":["IT Services"],"tags":["smart city india"],"author_name":"AIM Media House","publish_date":"2017-08-02T09:43:25","publication_year":"2017","word_count":711,"keywords":["Go","funding","programming_languages:R","AI","programming_languages:Go","smart city india","ViT","R"],"extracted_tech_keywords":["AI","R","Go","ViT","funding","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/iot-helps-smart-cities-monitor-resources\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":65024,"title":"Top 7 AI-Based Chatbots To Choose For Your Business","content":"With its promise of limitless possibilities, artificial intelligence (AI) is becoming more pervasive in business circles. Its capabilities to augment human abilities has been demonstrated across functions, and AI-driven chatbots have played a critical role here. Integrated with a company’s key messaging applications, AI bots are programmed to automate customer support by simulating a conversation with a user. When they interact with these bots, they get the experience of interacting with real people. Not only can such a service be made available to them 24\/7, but it can also provide business’ critical data on customer behaviour to inform future decision making. There are many AI chatbots in the market today, and while some may come with similar features, they are largely different and offer varying levels of capabilities. While some businesses can adapt to basic bots for their use cases, others may need more advanced software to run their operations effectively. With many options to choose from, organisations will do well to understand which will suit their business needs the best. Here is a list of seven AI chatbots that can help you make this decision. Listed in no particular order, these are likely to improve your overall service, increase engagement, boost sales and enable you to stay competitive in the market. Botsify With Botsify, companies can easily develop chatbots without the need for excessive coding and decoding. Combining the support of both AI and human agents, it allows businesses’ to nimbly switch to the latter in case of emergencies. What is more, it can conduct conversations in multiple languages. Additionally, the platform promises a faster response rate, better customer retention, and more qualified leads. Although it charges $50 (onwards), it does provide a two-week trial period for businesses. Click here for more. Mobile Monkey Emerging as one of the more popular platforms to build AI chatbots for Facebook Messenger and SMS, Mobile Monkey offers businesses an interactive interface to interact with their customers more effectively. Free to use, it has become a regular fixture in many company’s social media marketing strategies. It enables them to automate responses, maintain a long contact list, send periodic notifications, and more. Click here for more. ChatterOn This platform enables businesses to build a bot in under five minutes! What is more, they are given the option of choosing between 20 pre-developed bots, or even customise it without any coding requirements. While its key features remain its ability to support a wide range of content — from images to gifs to videos — other reasons to make this your choice of chatbot is its capability to allow users to meet their end-to-end solutions. Although it is free to use (for a limited period), businesses will have to pay $0.0010 per message after that. Click here for more. Pandorabots Although Pandorabots provides a range of services to meet various businesses demands, it requires some coding skills to get started. Businesses can use this for customer service, voice interfaces, B2B messaging, among others. The platform claims to have over 2,75,000 registered developers, with almost 3,25,000 chatbots already developed. Although there is a free version available for a limited range of features, a more comprehensive package starts at $19 per month including a two-week free trial. Click here for more. Hubspot HubSpot offers a full stack of software for content, sales and marketing services, as well as free CRM to organise better and build better relationships. Its primary feature is its content marketing flow that enables businesses to effectively respond to a large client base. Although some features are available for free, proper packages start at $35 a month. Click here for more. Rulai Driven by NLU and deep learning, Rulai enables businesses to track non-linear conversations and engage with customers effectively using virtual assistants. Its multitasking capabilities allow users to understand the context of the conversation and take actions accordingly. Targeted at companies across various industries, Rulai claims to deliver over 80% satisfaction for customers. Click here for more. Watson Assistant This allows businesses to build advanced conversational interfaces into any device, website, apps or cloud. Developed by IBM, it does not require users to have any prior coding experience. Watson Assistant is pre-trained with content from specific industries and allows businesses to build and deploy conversational interactions. Although the Lite versions with limited features come for free, packages start from $120 for 1,000 users per month along with a free 30-day trial. Click here for more.","excerpt":"With its promise of limitless possibilities, artificial intelligence (AI) is becoming more pervasive in business circles. Its capabilities to augment human abilities has been demonstrated across functions, and AI-driven chatbots have played a critical role here.  Integrated with a company’s key messaging applications, AI bots are programmed to automate customer support by simulating a conversation […]","categories":["AI Trends"],"tags":["AI Chatbot","chatbot ai"],"author_name":"Anu Thomas","publish_date":"2020-05-17T10:00:00","publication_year":"2020","word_count":741,"keywords":["artificial intelligence","programming_languages:R","AI","AI Chatbot","chatbots","virtual assistants","Aim","deep learning","chatbot ai","GAN","R"],"extracted_tech_keywords":["AI","artificial intelligence","deep learning","Aim","chatbots","virtual assistants","R","GAN","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-7-ai-based-chatbots-to-choose-for-your-business\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":45357,"title":"Anyone Can Be A Data Scientist: How MIT &#038; Brown’s New Tool Northstar Is Empowering Non-Techies","content":"The demand for data analytics is just growing bigger each year. Companies across the world have realised the importance of data analytics and have started hiring for these positions in bulk. While companies succeed in onboarding top data analysts, many settle for candidates who have no coding expertise and that could be seen during the real work. And this is where data analytics tools come into the scenario. These tools don’t require much\/any coding and reduce the workload of data science professionals significantly, managing to deliver better results. Meet Northstar To fill the above-mentioned gap in the data science domain, MIT and the Brown University researchers have gone beyond the generic data analytics tools with their new platform Northstar. The Northstar is an interactive data science platform that rethinks how people interact with data. The platform is designed and built in such a way that it allows non-specialists to use machine-learning models to make predictions. Simply put, users without programming experience, background in statistics or machine learning expertise can work on exploring and mining data along with building, analyzing, and evaluating ML pipelines. It has always been intriguing to watch people in movies using holographic screens, dragging and dropping things from one end to another and do a lot of technical stuff. And with this new platform by MIT and Brown University, it seems the days are not so far when businesses wouldn’t need typical data science professionals to solve their complex problems — they would just need touchscreen device and any existing data sets in order to build powerful prediction tools. Also called as the virtual data scientist (VDS), Northstar could be a great system for professionals as well as business of small and mid-size. When the research stated that the platform is built to keep non-specialists in mind, they meant professionals from other domains as well. For example, an average doctor’s office can use Northstar to come out predictions about their patients based on the historical medical data. Another example would be business owners — they can use this platform to predict their sales based on the data about their previous sales and the ongoing sales. This would not only help professionals, but also business that doesn’t have a dedicated data science team or department. “Even a coffee shop owner who doesn’t know data science should be able to predict their sales over the next few weeks to figure out how much coffee to buy,” says co-author and long-time Northstar project lead Tim Kraska. Talking about the interface of the Northstar and how it works, here it is: It is a blank canvas that supports several types of touch screens. User can upload their data sets (that shows in the form of boxes on the left.) Using the drag and drop, users can draw lines connecting boxes. Meaning, they should be processed with an algorithm of their choosing in combination with one another. combine inputs to generate predictive, AI-based analysis. Also, all data are stored and analyzed in the cloud. “It’s like a big, unbounded canvas where you can layout how you want everything. Then, you can link things together to create more complex questions about your data,” says Emanuel Zgraggen, a postdoc and main contributor of Northstar. Outlook Technology itself is a double-edged sword. Every time we witness a major breakthrough in the tech industry, we always end up thinking more and more about the way it is going to affect the human workforce. You cannot deny the fact that there has always been a fear inside everyone, “what if robots and AI completely take off human jobs.” The data science domain is no exception; no matter how helpful or innovate Northstar is, there are definitely thoughts of platforms like this becoming more and more advance and take the jobs of a data science professional. And the thoughts are valid enough. A data scientist spends a significant amount of time and money to gain knowledge and land a job, and when out of the blue a platform like Northstar shows up, the concerns become real. However, there are people as well, who thinks that this platform would only help existing data scientist with their job. It might become more advanced, but it won’t be able to beat the intelligence of a human mind.","excerpt":"The demand for data analytics is just growing bigger each year. Companies across the world have realised the importance of data analytics and have started hiring for these positions in bulk. While companies succeed in onboarding top data analysts, many settle for candidates who have no coding expertise and that could be seen during the […]","categories":["AI Features"],"tags":["Data Analytics","Data Scientist","MIT"],"author_name":"Harshajit Sarmah","publish_date":"2019-09-03T13:23:35","publication_year":"2019","word_count":716,"keywords":["data science","Go","machine learning","programming_languages:R","AI","MIT","ML","programming_languages:Go","RAG","analytics","Data Analytics","Data Scientist","R"],"extracted_tech_keywords":["AI","machine learning","ML","data science","analytics","RAG","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/anyone-can-be-a-data-scientist-how-mit-browns-new-tool-northstar-is-empowering-non-techies\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10070585,"title":"Finally, a language model that does Maths","content":"A bunch of large language models burst into the scene this year, with applications ranging from automated code generation to text to image generation. However, these LLMs have come up short on the quantitative reasoning front. Google has broken this barrier with their latest language model, Minerva. Named after the Roman goddess of wisdom, it is trained on a high-quality scientific and mathematical dataset. Minerva in a nutshell Minerva is built on Pathways Language Model (PaLM) with extended training on a 118GB dataset of scientific papers from arXiv and 38.5B tokens of mathematical data derived from web pages. The model processes scientific and mathematical questions formed in natural language and generates a step-by-step solution with the help of Latex Notations, MathJax, or other mathematical typesetting formats. The model is developed in 3 baseline forms:  8B, 62B, and 540B parameter pretrained models. Along with an extended pretraining on mathematical data, Minerva also incorporates prompting and evaluation techniques like chain-of-thought, scratchpad and majority voting to provide a step-by-step evaluation process and choose the optimal result out from a sample of multiple solutions. STEM benchmarks Minerva’s quantitative reasoning capabilities were tested on STEM benchmarks, the standard of assessment in the United States education system. The level of difficulty ranges from grade school level to graduate-level coursework. MATH: A dataset of 12K middle school and high school math problems written in LATEX format. The models are prompted with a fixed four-shot prompt consisting of four random examples whose ground truth targets are not too long. MMLU-STEM: Focused on STEM, this subset of the Massive Multitask Language Understanding benchmark covers topics such as engineering, chemistry, math, and physics at the high school and college levels. In this test, a combination of five-shot, chain-of-thought and a multiple-choice version of the MATH prompt is used to tackle various problems. GSM8k: Grade school-level math problems involving basic arithmetic operations that should all be solvable by a talented middle school student. The model is evaluated using chain-of-thought prompting. However, in this test, no external tools are used for calculations. OCWCourses: A set of problems covering STEM topics ranging from differential equations, solid-state chemistry, astronomy, special relativity, etc., at an undergraduate level. The dataset was created using publicly-available course materials offered by MIT (OpenCourseWare). Only problems with automatically-verifiable solutions (either numeric or symbolically verifiable via SymPy) from various courses were included. Minerva outperformed SOTA results by a wide margin. Source: arxiv.org Image: Accuracy on LMs on MATH and MMLU The Minerva model was also tested on the National Math exam in Poland in 2021. While the 62B baseline model achieved 57%, which was the national average that year, the 540B baseline variant achieved 65%. What’s the big deal? Quantitative reasoning is the ability to use mathematics and information to solve real-world problems. OpenAI’s GPT-3 could only complete 2.9% to 6.9% of problems from a dataset of over 12,500 in the year 2021. Later, they launched GPT-f. The automated prover and proof assistant for the Metamath formalization language was the first machine learning-based system that contributed proofs that were adopted by a formal mathematics community. Guillaume Lample and François Charton at Facebook AI Research have come up with a trained neural network capable of symbolic reasoning for differential and integral equations. Minerva is trained on a large dataset that combines natural language understanding with the correct use of formal mathematical language (equations and diagrams). The model establishes a new baseline for quantitative reasoning benchmarks by increasing data quality and model size. Source: arxiv.org The researchers said one of the direct applications could be the creation of an accessible and affordable AI-based math tutor. Not a perfect model Despite training on an extensive dataset of mathematical data, Minerva is far from a perfect problem solver. Upon analyzing the sample of problems that the model got wrong, a pattern was found. About half of the problems were calculation errors, while the other half were solution steps that did not follow a logical chain of thought.Although the model arrived at the right answer, it did not use the correct reasoning. Such cases were referred to as ‘false positives. However, the rate of false positives was relatively low. The model doesn’t have access to external tools like a calculator or a Python interpreter, limiting its ability to handle tasks that require complicated numerical calculations. Check out the demo of Minerva explorer.","excerpt":"Minerva is built on Pathways Language Model (PaLM) with extended training on a 118GB dataset of scientific papers from arXiv and 38.5B tokens of mathematical data derived from web pages.","categories":["AI Features"],"tags":["Google Research"],"author_name":"Kartik Wali","publish_date":"2022-07-07T11:00:00","publication_year":"2022","word_count":726,"keywords":["machine learning","OpenAI","AI","neural network","ML","RAG","Python","JAX","chain of thought","R","Google Research"],"extracted_tech_keywords":["AI","machine learning","ML","neural network","OpenAI","JAX","RAG","chain of thought","Python","R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/finally-a-language-model-that-does-maths\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10073917,"title":"Indian Researchers Develop AI Algorithm that Detects Diabetes from ECG Data","content":"By 2045, India will account for 134 million diabetic patients. In 2019, 77 million people had diabetes in India, according to a report. Globally, the numbers are even more staggering, with around 783 million people estimated to be diabetic by 2045, according to data from International Diabetes Federation. To tackle the issue of diabetes, a group of researchers from the Lata Medical Research Foundation in Nagpur has developed an AI algorithm that can predict diabetes and pre-diabetes from individual heartbeats recorded on an ECG (electrocardiogram). “China is the diabetes capital of the world, and it’s going to shift to India very soon. So if there is a time to do something for type two diabetes is now,” Hemant Kulkarni, President at Lata Medical Research Foundation, told Analytics India Magazine. DiaBeats “The idea was to try to find out a simple, early and most importantly, non-invasive biomarker for prediction of type two diabetes,” Kulkarni said. Early diagnosis of diabetes is key to preventing subsequent health problems. DiaBeats- the AI algorithm developed by Kulkarni and his team detects Type 2 diabetes and pre-diabetes by reading ECG data. “Diagnosis of these conditions relies on the oral glucose tolerance test and Haemoglobin A1c (HbA1c) estimation, which are invasive and challenging for large-scale screening. “We aimed to combine the non-invasive nature of ECG with the power of machine learning to detect diabetes and pre-diabetes,” the research paper said. Type 2 diabetes is common among adults and accounts for around 90 per cent of all cases. The dataset for this study came from Diabetes in Sindhi Families in Nagpur (DISFIN) study of ethnically endogenous Sindhi population from central India. “My wife is by origin a Sindhi, which is a high risk population for type two diabetes. So we wanted to see if there are any genetics involved or the predictors of type two diabetes, especially in that population. “So we collected lots of data. We had 1262 patients on whom the data was available and on whom the ECG passed a quality check. So that gave us over 10,000 individual heartbeats to work with,” Kulkarni said. After this, the team split the datasets into training sets and used the Extreme gradient boosting (XGBoost) algorithm to predict which patient had diabetes, pre-diabetes or no diabetes at all. DiaBeats model training Potential “It was pleasant to find that the study achieved an accuracy of 97 per cent, which superseded all the previous attempts in this direction. Previous studies are reported at anywhere between 86 to 95 per cent,” Kulkarni said. Speaking about the potential of the research, Kulkarni said that the research could prove to be very important in a public health scenario. Considering that diabetes is a big problem in India, the research could have significant implications. There is no cure for diabetes, which means most people live with the disease. Further, the current methods used to diagnose diabetes are costly compared to ECGs. “The numbers are staggering in terms of the caseload of type two diabetes and the costs associated with it in India. “For us to be able to make an early diagnosis that too on a mass level, using a non-invasive, acceptable approach would lead to a radical change in the approach in which we tackle type two diabetes,” Kulkarni said. Challenges ahead However, Kulkarni also added that the algorithm is a long way away from practical use. This is because the algorithm has not been externally validated so far. “At that time, we didn’t have public or private datasets available where we could do this external validation. Otherwise, we would have liked to do that within the study itself,” Kulkarni said. So for Kulkarni and his team, the next step is establishing the algorithm’s robustness through external validation. “So we’re going to go ahead and try to find some data sets that will help us do that,” he said. Further, Kulkarni adds that diabetes as a disease rarely happens in isolation. Often it co-exists with hypertension, obesity, dyslipidemia etc., in the patient and across populations. “So if we want a technique or an algorithm to be detecting what it says it does, then we need to make sure that it is specific to that condition and not altered by these other things that also come along.” “Now, the goal for us is to ensure that the algorithm is detecting type two diabetes, and further, to ensure that it is detecting only diabetes and not anything else,” he concluded.","excerpt":"The research achieved an accuracy of 97 per cent which superseded all the previous attempts in this direction.","categories":["AI Features"],"tags":[],"author_name":"Pritam Bordoloi","publish_date":"2022-08-31T10:00:00","publication_year":"2022","word_count":745,"keywords":["Go","API","machine learning","programming_languages:R","AI","programming_languages:Go","Aim","XGBoost","analytics","R"],"extracted_tech_keywords":["AI","machine learning","analytics","Aim","XGBoost","R","Go","API","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/indian-researchers-develop-ai-algorithm-that-detects-diabetes-from-ecg-data\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":31516,"title":"IBM Aims For 8-Bit AI Breakthrough Which Will Supercharge Computational Efficiency","content":"CDC 1604 was a second generation 48-bit word size computer that could handle 100,000 operations per second. It was designed to automate the missile control for the US Navy. But trying to run an iPhone on CDC 1604, it would have a hard time rendering a single pixel of your selfie. So your palm-sized phone could have handled multiple missile systems which otherwise would have taken a room-sized computer. Computer architectures were initially designed to specialise in serial processing and then use up DRAMs to optimise high-density memory. Memory-on-chip area is expensive and adding large amounts of memory will only make things worse. A typical neural network’s memory has to store input data, activations and weight parameters. In the paper titled Training Deep Neural Networks with 8-bit Floating Point Numbers authors Naigang Wang, Jungwook Choi, Daniel Brand, Chia-Yu Chen, Kailash Gopalakrishnan of IBM Watson Research Centre, introduce novel techniques to successfully train DNNs using 8-bit floating point numbers (FP8) while fully maintaining the accuracy. “Historically, high-performance computing has relied on high precision 64 and 32-bit floating point arithmetic. This approach delivers accuracy critical for scientific computing tasks like simulating the human heart or calculating space shuttle trajectories,” say the researchers behind this breakthrough. But, for tasks like image classification or speech recognition, approximation seems to be of more importance than high precision that 64 or 32 bit has to offer. As we lower the floating point limit, the engines become smaller. For instance, 16-bit precision engines are typically 4 times smaller than comparable blocks with 32-bit precision; this gain in area efficiency directly translates into a significant boost in performance and power efficiency for both AI training and inference workloads. So, here high precision is being traded for computational enhancements and in this paper, the researchers demonstrate novel training techniques that eventually quicken the compression methods by a factor of 40-200x. Back in 2015, IBM Research had shown how to fully preserve model accuracy while going from 32 bits to 16 bits. Now, 16-bit training and 8-bit inference system have been inculcated into the industry as a standard. While works are underway for making 8-bit precision and 4-bit precision for inference, a thing. The main challenges associated with training networks below 16 bits: Is the maintenance of fidelity of the gradient computations and weight updates during back-propagation. When weights, errors and gradients used in matrix multiplication are reduced to 8 bits, deep neural networks take a hit and there is a significant degradation in its final accuracy. Also, there is a noticeable impact on the convergence during network training. 32-bit weight updates, used in today’s systems, require an extra copy of the high-precision weights and gradients to be kept in memory, which is expensive. How To Work Around An 8-bit FP A new technique called chunk-based computations, when applied, allows operations like convolution to be computed using 8-bit multiplications and 16-bit additions instead of 16 and 32 bits, respectively. And, a floating point stochastic rounding for the weight updating process. In chunk-based computations, the accumulated computations are broken up into chunks. These chunks consist of deep learning dot products. Researchers have tested this technique to train ResNet50 model. Many of these techniques have been tested using chips of size 14 nm. The results indicate that chunk-accumulation engines can be used along with reduced-precision dataflow engines without increasing hardware overheads. As a baseline, accumulation in FP-32 increases linearly with vector length. A typical FP-16 accumulation with the nearest rounding (i.e., ChunkSize=1) significantly suffers swamping errors (the accumulation stops when length ≥ 4096). This is where the chunk-based accumulation can be used to compensate for the errors, as the effective length of accumulation is reduced by chunk size to avoid swamping. The effectiveness of the stochastic rounding can be seen below; although there exists a slight deviation at large accumulation length due to the rounding error, stochastic rounding consistently follows the FP-32 result. This success paves way for a new era of hardware training platforms which perform twice as well as the current systems. This Paper Introduces: A new FP-8 floating point format that, in combination with DNN training insights, allows general matrix multiplication(GEMM) computations for Deep Learning to work without a loss in model accuracy. A new technique called chunk-based computations that when applied hierarchically allows all matrix and convolution operations to be computed using only 8-bit multiplications and 16-bit additions (instead of 16 and 32 bits respectively). An applied floating point stochastic rounding in the weight update process which allows these updates to happen with 16 bits of precision (instead of 32 bits). The wide applicability of the combined effects of these techniques across a suite of Deep Learning models and datasets – while fully preserving model accuracy. Approximate computing lies at the root of this approach where an attempt has been made to tweak in the hardware to make models more robust along with high energy-efficient computing gains with purpose-built architectures.","excerpt":"CDC 1604 was a second generation 48-bit word size computer that could handle 100,000 operations per second. It was designed to automate the missile control for the US Navy. But trying to run an iPhone on CDC 1604, it would have a hard time rendering a single pixel of your selfie. So your palm-sized phone […]","categories":["Deep Tech"],"tags":["deep neural network","IBM"],"author_name":"Ram Sagar","publish_date":"2018-12-14T06:14:53","publication_year":"2018","word_count":822,"keywords":["Go","AI","neural network","llm_models:PaLM","deep neural network","ResNet","llm_models:T5","deep learning","IBM","GAN","R","T5"],"extracted_tech_keywords":["AI","deep learning","neural network","R","Go","T5","GAN","ResNet","llm_models:T5","llm_models:PaLM"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/ibm-8-bit-ai-breakthrough-to-supercharge-computational-efficiency\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":61305,"title":"How This Bangalore-Based Sports-Tech Startup Is Leveraging Neural Networks To Optimize Performance Of Cricketers","content":"With so much innovation happening in the tech space, early adoption of sports-tech at the consumer level has started. According to reports, the global sports technology market stood at $9,695.3 million in 2018, and is expected to reach $51,689.7 million by 2027, expanding at a CAGR of 20.4% during the forecast period. The sports-tech industry is set to play a critical role in revolutionizing the sports sector by addressing new challenges. With a similar vision, Bangalore-based startup StanceBeam has been utilizing AI and ML to work on groundbreaking ideas in the sports-tech space. Founded in 2017 by Arminder Thind, StanceBeam has been leveraging AI in conjunction with IoT devices to gather data to attempt to optimize training and performance of players. The company is focused on developing connected sports technology and analytics, leveraging IoT and AI that will benefit every player, coach as well as sports fans. Arminder ThindFounder of StanceBeam Flagship Product In December 2018, the company launched its first, commercially available Smart Cricket Bat Sensor known as StanceBeam Striker. This flagship product is a cricket bat sensor that attaches to any cricket bat and is handled via a custom bat mount, which immediately transforms a normal cricket bat into a smart bat by providing live analysis of a player’s batting technique and performance. Furthermore, motion sensors built into the device provides a validated set of data which provides 3D swing analysis, bat speed generation, power index and shot efficiency. How Is It Different From Others In The Market? According to Arminder, the accuracy of the data and the user experience are the key differentiators of this product. How StanceBeam Uses AI and ML According to Arminder, StanceBeam generates abundance of data via Striker IoT sensors, and video and audio during player sessions. AI helps provide the necessary tools and framework to perform real-time data analysis and decision-making with complete automation. An amalgamation of ML models helps the company create various interesting features for the customers, and thereby, improve their performance. The amalgamation of video data and sensor data enables StanceBeam to create intelligent use cases like ‘intelligent virtual coach’ for customers. This amalgamated data analysis and visualization provides an in-depth understanding of the data and helps create interesting use cases for customers. Along with these ML models, custom algorithms are also used to derive metrics from helping players’ performance. Currently, StanceBeam is working on building various Neural and Bayesian Networks to derive these models from achieving a higher level of intelligence. Core Tech Stack The company uses Microsoft Azure as its cloud provider. Arminder said, “We extensively use Azure Machine Learning Service tools to learn, build concepts and deploy machine learning pipelines.” On the technology stack, StanceBeam uses Motion Sensors to capture information, and Microcontrollers and Bluetooth stack to process and transfer data captured by motion sensors. The company uses Android and iOS native stack to develop apps, along with custom C\/C++ built algorithms libraries. Furthermore, the backend application is built on serverless technologies, which dynamically scales the application to match the demands of any workload. Tackling Hiring Phase According to Arminder, when it comes to hiring employees, the company looks for underdogs with a lot of hunger to prove themselves. He said, “Being a sport-focused startup, love and passion for sports is also important.” Challenges Faced To this, Arminder said that being a new category product and cricket being such a traditional sport, changing players and coaches’ mindset about the data-driven approach for better performance has been the key challenge that the company has faced so far. Roadmap StanceBeam is a new sports technology company which has recently introduced StanceBeam Striker, an intelligent cricket bat sensor that connects it to a shooting analysis application. In the next five years, the company plans to launch more products. Arminder said, “We will be a multi-product and multi-sports tech company with presence in key global markets.”","excerpt":"With so much innovation happening in the tech space, early adoption of sports-tech at the consumer level has started. According to reports, the global sports technology market stood at $9,695.3 million in 2018, and is expected to reach $51,689.7 million by 2027, expanding at a CAGR of 20.4% during the forecast period. The sports-tech industry […]","categories":["AI Startups"],"tags":["ai in sports","Neural Networks","Startups"],"author_name":"Ambika Choudhury","publish_date":"2020-04-09T15:00:00","publication_year":"2020","word_count":646,"keywords":["Go","machine learning","AI","R","ML","serverless","RAG","C++","analytics","Startups","Azure","ai in sports","Neural Networks"],"extracted_tech_keywords":["AI","machine learning","ML","analytics","RAG","Azure","serverless","R","Go","C++"],"url":"https:\/\/analyticsindiamag.com\/ai-startups\/how-this-bangalore-based-sports-tech-startup-is-leveraging-neural-networks-to-optimize-performance-of-cricketers\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10141031,"title":"NVIDIA Launches LLaMA-Mesh, a Unified 3D Mesh Generation Method Using LLMs","content":"NVIDIA released ‘LLaMA-Mesh’ earlier this week, a novel method for enabling large language models (LLMs) to generate 3D meshes from text prompts. This approach integrates 3D mesh generation with language understanding, allowing the model to represent meshes in plain text format without modifying its vocabulary or tokenisers. Ahsen Khaliq, ML growth lead at Hugging Face took to LinkedIn to announce this latest update. LLaMA-Mesh builds on LLaMA, a language model, by fine-tuning it on a curated dataset of 3D dialogue. NVIDIA and Tsinghua University researchers designed the method to preserve the model’s language capabilities while extending its functionality to generate and understand 3D content. How Does LLaMA-Mesh Work? The method leverages existing spatial knowledge embedded in LLMs, derived from textual sources like 3D tutorials. It tokenises 3D mesh data, including vertex coordinates and face definitions, into text, allowing seamless processing by language models. To train the model, the researchers developed a supervised fine-tuning dataset. This dataset enables the LLM to perform tasks such as generating 3D meshes from text prompts, producing interleaved text and 3D outputs, and interpreting 3D mesh structures. The study shows that LLaMA-Mesh achieves 3D mesh generation quality comparable to specialised models trained exclusively on 3D data. Training and Results: According to the official research paper, the model is trained on 32 A100 GPUs for 21k iterations over 3 days using an AdamW optimiser with a small learning rate and warm-up steps. A batch size of 128 and cosine scheduling ensure smooth training. The loss shows quick adaptation to the new task with no spikes or issues, which highlights the model’s stability and ability to learn efficiently. Source: Official research paper The researchers claim that the method creates detailed, high-quality 3D meshes with artist-level designs, learning this during training. It can generate diverse, creative outputs from the same text prompt, perfect for tasks needing multiple design options. Even after fine-tuning for 3D mesh generation, the model keeps its strong language skills, understanding complex instructions, asking smart questions, and giving detailed answers. Tests show it performs as well as other models in reasoning and problem-solving while also excelling at creating 3D designs. What Happens to 3D Modelling? AI is transforming animation and 3D modelling, impacting artists, studios, developers, and end-users alike. By automating repetitive tasks, AI frees artists to focus on creativity, while studios benefit from faster, cost-effective production. This raises the question: Is it going to be the end of human animators and modellers soon? Earlier in October, NVIDIA announced EdgeRunner, which could generate highly detailed 3D meshes with up to 4,000 faces at a spatial resolution of 512. This was derived from both images and point clouds—resulting in sequences that are twice as long and four times higher in resolution compared to previous methods. Referring to LLaMA-Mesh, an entrepreneur on LinkedIn says, “3D modelling is about to enter a new phase. The speed of execution will make it possible to create projects at lower costs. It’s not bad, although it will require less modeller…” The researchers noted, “This work represents a significant step toward integrating multi-modal content generation within a cohesive language model.” LLaMA-Mesh opens new possibilities for conversational 3D generation and understanding, highlighting the potential for unifying 3D and text modalities in language models.","excerpt":"LLaMA-Mesh is trained on 32 A100 GPUs for 21k iterations over 3 days.","categories":["AI News"],"tags":["NVIDIA"],"author_name":"Sanjana Gupta","publish_date":"2024-11-15T16:55:32","publication_year":"2024","word_count":539,"keywords":["Go","Hugging Face","TPU","AI","Modal","ML","RAG","Aim","ViT","NVIDIA","R"],"extracted_tech_keywords":["AI","ML","Aim","Hugging Face","RAG","TPU","R","Go","ViT","Modal"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/nvidia-launches-llama-mesh-a-unified-3d-mesh-generation-method-using-llms\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10074107,"title":"HCL Tech becomes India’s third largest IT company, beats Wipro","content":"HCL Tech has become the third largest IT company in India by revenue, overtaking Wipro. The company has overtaken its competitor in terms of revenue, profits and market capitalisation by a margin, with a market cap of ₹2.5 lakh crore, compared to Wipro’s ₹2.2 lakh crore. However, being the third largest IT company in India by revenue for the last four years, both the companies’ stocks have witnessed a massive drawdown—wherein HCL Tech’s shares fell much lower compared to Wipro. Besides the ever-changing market dynamics for the two tech giants, all IT stocks this year have witnessed significant drops—especially Wipro, which was widely believed to be an underperformer in this arena. The share price of HCL Tech has fallen by 20.76 per cent, with Wipro taking a fall of 35.58 per cent. A report further claimed that the major drop in share prices could be one of the reasons behind Wipro’s diminished market cap. The Indian IT sector has been affected by high attrition rates, inflation, and chain issues, along with witnessing unprecedented boom for two years that led JP Morgan to downgrade the Indian IT sector and go “underweight” over revenue concerns. HCL Technologies was founded in 1991 and is fairly younger in comparison to its competitor, Wipro, that was established in 1945. HCL initially established itself as a maker of vegetable oils and transformed into a full-fledged IT firm only in the 1980s.","excerpt":"Besides the ever-changing market dynamics for the two tech giants, all IT stocks this year have witnessed significant drops—especially Wipro, which was widely believed to be an underperformer in this arena.","categories":["AI News"],"tags":["Indian IT","Shiv Nadar","Wipro"],"author_name":"Bhuvana Kamath","publish_date":"2022-09-01T12:50:58","publication_year":"2022","word_count":235,"keywords":["Wipro","Go","API","Shiv Nadar","programming_languages:R","AI","programming_languages:Go","Aim","GAN","Indian IT","R"],"extracted_tech_keywords":["AI","Aim","R","Go","API","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/hcl-tech-becomes-indias-third-largest-it-company-beats-wipro\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10140959,"title":"Databricks Looks to Raise $1 Bn from Thrive to Outsmart the Competition","content":"Thrive Capital is in talks to invest $1 billion in Databricks, which would boost the data analytics company’s valuation to around $55 billion, reflecting growing interest in AI-driven data solutions following its $43 billion valuation last year. With this fund, Databricks looks to accelerate advancements in AI-driven data solutions, expand its enterprise capabilities, and solidify its position as a leader in scalable analytics technology amid soaring demand. The deal would add Databricks to Thrive’s notable portfolio, including GitHub, Instagram, Stripe, Slack, and OpenAI among others. This investment, expected to be Thrive’s largest to date, underscores confidence in Databricks’ technology as the demand for AI platforms grows. Many experts believe that Databricks is poised to become a dominant force, offering businesses a robust ecosystem for their AI and data needs. These developments mark the company as one of the most highly anticipated IPOs in the tech industry, setting the stage for further advancements in enterprise data solutions. Databricks’ recent partnership with Amazon Web Services (AWS) enables them to integrate deeply with AWS services, providing scalable, cloud-based solutions for enterprises. This collaboration allows Databricks to optimise its data lakehouse platform for AWS’s extensive customer base, offering flexible data analytics tools that meet the growing needs of AI-driven businesses. This investment and acquisition trajectory highlights Databricks’ strategic goal to dominate the enterprise AI market. Earlier this year Databricks partnered with emerging startups like Ragas, a Bengaluru-based Y Combinator-backed company, to provide it with valuable resources for high frequency evaluations essential to big data and AI performance. Ragas’ support for AWS, Microsoft, and Databricks indicates a trend of collaborative growth between industry giants and specialized AI providers. Databricks’ CTO, Matei Zaharia, emphasised at the Data+AI Summit 2024 that the company’s vision is to support enterprises in building high quality, domain specific AI applications using its evolving suite of tools, such as the Mosaic AI offerings. This drive is aligned with the company’s broader ambition to cater to large scale systems, including the trend towards compound AI models that combine various models, external data, and API calls. These multifaceted moves signal Databricks’ readiness for its IPO while reinforcing its commitment to pushing the boundaries of AI in enterprise solutions. Databricks vs the world This new development also comes in light of the escaping rivalry between Snowflake and Databricks. “Snowflake consistently outperforms Spark-based SaaS with a 30% price-performance improvement… helping teams focus on innovation, not complexity,” said Snowflake CEO Sridhar Ramaswamy remarked, fueling a heated debate, particularly as Databricks supporters argue that the additional management controls in Spark are essential for customisation. Josue A Bogran, a Databricks product advisory board member, defended Databricks, saying, “Databricks offers ‘serverless’ compute… supporting both SQL and Python,” positioning it as a flexible Platform-as-a-Service (PaaS) approach that allows organisations more control compared to Snowflake’s SaaS model.","excerpt":"The funding could help Databricks, which is valued at around $43 billion, further expand its innovative data lakehouse architecture.","categories":["Deep Tech"],"tags":["AI (Artificial Intelligence)","Databricks","Funding"],"author_name":"Tanisha Bhattacharjee","publish_date":"2024-11-14T18:08:37","publication_year":"2024","word_count":465,"keywords":["Funding","OpenAI","AI","AWS","Snowflake","serverless","RAG","Python","analytics","R","AI (Artificial Intelligence)","Databricks"],"extracted_tech_keywords":["AI","analytics","OpenAI","RAG","AWS","serverless","Snowflake","Databricks","Python","R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/databricks-looks-to-raise-1-bn-from-thrive-to-outsmart-the-competition\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10064172,"title":"A hands-on guide to principal component regression in Python","content":"Principal Components Regression (PCR) is a technique for analyzing multiple regression data that suffer from multicollinearity. PCR is derived from Principal Component Analysis (PCA). So, it is PCA applied to a regression algorithm that has multicollinear features. Principal components regression reduces errors in regression estimates by adding a degree of bias and by doing so, it will be possible to provide more reliable estimates. In this article, we will put focus on implementing PCR on a regression problem. Following are the topics covered. Table of contents About the architecture of PCAAbout multicollinearity in regressionAbout Principal Components Regression (PCR)Implementing PCR in Python Let’s start with the architect PCA from which PCR is derived. About the architecture of PCA Principal Component Analysis (PCA) is the analysis of principal features of the data. The analysis is done by reducing the dimensionality of the feature space. In other words, it is a tool to reduce the features from the data to get only the required features or principal components for the learner. PCA has three major components which help to reduce dimensionality: The covariance matrix is the measure of how much the variables are associated with each other.The eigenvectors are the directors in which the data is dispersed.The eigenvalues are the relative importance of the directions. About multicollinearity in regression From the name, it is clear that the collinearity between the independent variables in a regression problem is stated as multicollinearity in regression. The reasons behind curing a multicollinear regression problem are: Understanding of the significance of features for the regression learner.Instability in estimating the coefficientOverfitting of the learner Multicollinearity in regression could be vandalized with the help of Principal Component Regression (PCR). Let’s understand how does PCR control multicollinearity. About Principal Components Regression (PCR) The Principal Component Regression (PCR) algorithm is an approach for reducing the multicollinearity of a dataset. Although multi-variate linear regression can fit well on the test set, there is normally a high-variance problem with it. For this reason, PCR adds a small bias to the model, such that it aims to maintain a high level of accuracy, but reduce the variance substantially. This is achieved by applying the PCA to the features before the training. Let’s try to learn PCR by implementing it on data. Implementing PCR in Python The objective of this regression model is to predict the salary of the player based on different features. Importing packages import numpy as np import pandas as pd Reading the data set salary=pd.read_csv(\"Hitters.csv\") df=salary.copy() df.dropna(inplace=True) df.shape,salary.shape ((263, 20), (322, 20)) This data set is taken from the Kaggle repository, to use this dataset the links are given in references. Copying the dataset to another data frame for further pre-processing so that the original data frame remains unchanged. Then drop all the missing values and at last check the number of rows and columns of the original and copied data frames. There are two categorical columns and the rest are continuous features, needed to encode the categorical features for further utilization. Encoding categorical features: df = pd.get_dummies(df,columns=['League', 'Division', 'NewLeague']) df.head() Encoded the data using the pandas get_dummies function.  The original features which are used for creating the dummies are replaced with a total of six new columns as shown in the above image. Since the regression problem is to predict the salary of the players so the dependent variable is “Salary”. y=df['Salary'] X=df.drop(['Salary'],axis=1) There are a total of 23 columns in this data frame out of which only a few are important for the analysis and prediction. There is a requirement for feature elimination which would be performed by  PCA. Applying PCA from sklearn.decomposition import PCA pca=PCA() X_red = pca.fit_transform(scale(X)) PCA is been imported from the sklearn library and stored in a variable for easier applications. The PCA is fitted with the independent variables for dimensionality reduction. The percentage of variance in the dependent variable is explained by adding each principal component to the model. np.cumsum(np.round(pca.explained_variance_ratio_, decimals = 4)*100)[0:5] The above output is explained as: By using the first principal component, we can explain 32.62% of the variation in the dependent variable.By using the second principal component, we can explain 53.02% of the variation in the dependent variable.Similarly, by using others we can explain 68.82%, 78.07%,85.39%, 89.31% The conclusion of this is that we need to use a total of five principal components in the regression learner. Now, the principal components are decided it’s for split the dataset into a 70:30 ratio for train and test for training and testing the learner. X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.30, random_state=42) Let’s build the final model using five principal components in the linear regression model. X_red_train = pca.fit_transform(scale(X_train)) X_red_test = pca.transform(scale(X_test))[:,0:5] lm = LinearRegression() pcr = lm.fit(X_red_train[:,0:5], y_train) y_pred = pcr.predict(X_red_test) The final model is built and trained on the train data set and also predicted the salary using the test dataset. Let’s check how good the model has performed. np.round(np.sqrt(mean_squared_error(y_test, y_pred)),2) 397.52 The root means square error (RMSE) is low approximately 398 this can be improved by tunning the model further I would leave that to you. Nutshell Applying Principal Component Analysis before the regression can reduce multicollinearity and helps in better prediction with fewer features in lesser time. PCR also reduces the chances of overfitting learners. References Link for the above codePaper released on PCR","excerpt":"Principal Components Regression (PCR) is a technique for analyzing multiple regression data that suffer from multicollinearity. Principal components regression reduces errors in regression estimates by adding a degree of bias","categories":["Deep Tech"],"tags":["overfitting","PCA","regression analysis"],"author_name":"Sourabh Mehta","publish_date":"2022-04-02T18:00:00","publication_year":"2022","word_count":886,"keywords":["Go","NumPy","TPU","programming_languages:R","AI","R","regression analysis","Python","Aim","programming_languages:Python","PCA","overfitting","Pandas"],"extracted_tech_keywords":["AI","Aim","Pandas","NumPy","TPU","Python","R","Go","programming_languages:Python","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/a-hands-on-guide-to-principal-component-regression-in-python\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":44430,"title":"TCS AI Platform Ignio Crosses $60m Mark In Annual Revenue","content":"Tata Consultancy Services’ AI platform ignio this week crossed the $60 million mark in annual revenue. The company also announced a triple-digit annual growth in both customer acquisition and revenue. The company also announced that ignio had doubled its revenue as well as the number of customers, year-on-year, as it completes four years since launch. Ignio was launched in 2015 with the objective of leveraging artificial intelligence, machine learning and advanced software engineering to transform enterprise IT services. Akhilesh Tripathi, Global Head, Digitate, said, “ignio’s unique capabilities cover the breadth and depth of an enterprise from IT to business and from context to self-heal. ignio closes the loop between prediction, recommendation and autonomous execution on a single platform, and delivers unparalleled results on business assurance, agility, efficiency and customer experience. We will continue to invest in the product, expand our partner ecosystem, and offer 24×7 support.” In the last 12 months, ignio had 52 new wins, bringing the total number of customers to 105. These are mostly Global 2000 companies, distributed across the retail, manufacturing, telecom and banking and financial services sectors. With its speedy implementations and track record of delivering impactful outcomes within weeks, ignio is being viewed by forward-thinking organizations as a strategic investment to build an agile and resilient operating core that powers a great customer experience. The last four years have seen ignio steadily expand into newer areas of enterprise operations. Its early adoption was by progressive CIOs looking to automate their IT operations. ignio made their IT infrastructure stacks self-healing by using its context-aware reasoning capability to pre-empt problems, and autonomously resolve a lot of the incidents that did occur. Since then, ignio has been used to automate batch jobs management and ERP application support. Now, largely driven by demand from business stakeholders, ignio’s cognitive capabilities are being leveraged to transform new business-centric use cases. It is being used by retailers to find anomalies in drug prescriptions, and to reconcile vendor incentive payouts, freeing up working capital. Procurement officers are using ignio to uncover maverick spends by detecting and predicting anomalies, enabling quicker decisions.","excerpt":"Tata Consultancy Services’ AI platform ignio this week crossed the $60 million mark in annual revenue. The company also announced a triple-digit annual growth in both customer acquisition and revenue. The company also announced that ignio had doubled its revenue as well as the number of customers, year-on-year, as it completes four years since launch. […]","categories":["AI News"],"tags":["ignio","TCS"],"author_name":"Prajakta Hebbar","publish_date":"2019-08-13T17:42:25","publication_year":"2019","word_count":349,"keywords":["API","artificial intelligence","machine learning","programming_languages:R","AI","R","Git","RAG","ignio","GAN","TCS"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","RAG","R","Git","API","GAN","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/tcs-ai-platform-ignio-crosses-60m-mark-in-annual-revenue\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":50131,"title":"Top 11 Hot Chips For Machine Learning","content":"Though machine learning has been around for more than three decades, it took a lot of time for the hardware to catch up with the demands of these power-hungry algorithms. With each passing year, the chip-set manufacturers have tried to make the hardware lighter and faster. Today, over 100 companies are working on building next-generation chips and hardware architectures that would match the capabilities of algorithms. These chips are capable of enabling deep learning applications on smartphones and other edge computing devices. Here we list a few top hardware innovations that have transformed the world of AI: Intel’s Nervana via Intel Intel recently revealed new details of upcoming high-performance artificial intelligence accelerators: Intel Nervana neural network processors. It is built to prioritise two key real-world considerations: training a network as fast as possible and doing it within a given power budget. This processor is built with flexibility in mind, striking a balance among computing, communication and memory. Samsung Exynos 9 via Samsung Samsung’s Exynos 9820 has a separate hardware AI-accelerator, or NPU, which performs AI tasks around seven times faster than the predecessor. This is aimed at AI-related processing that can be carried out directly on the device rather than sending the task to a server, providing faster performance as well as better security. AMD Radeon Instinct via AMD Radeon Instinct is a Superior Training Accelerator for Machine Intelligence and Deep Learning Based on cutting-edge “VEGA” graphics architecture built to handle big data sets and diverse compute workloads. Up to 24.6 TFLOPS of FP16 peak compute performance for deep learning training applications Nokia’s ReefShark via Nokia Nokia’s ReefShark is a completely new chipset that dramatically eases 5G network roll-out. AI is implemented in the ReefShark design for radio and embedded in the baseband to use augmented deep learning to trigger smart, rapid actions by the autonomous, cognitive network, enhancing network optimisation and increasing business opportunities. Alibaba’s Pingtouge Hanguang via Alibaba Alibaba unveiled its first AI dedicated processor for cloud-based large-scale AI inferencing. The 12-nm Hanguang 800 contains 17 billion transistors.and is 15 times more powerful than the NVIDIA T4 GPU, and 46 times more powerful than the NVIDIA P4 GPU. Huawei’s Ascend 910 via Huawei The Ascend 910 is a new AI processor that belongs to Huawei’s series of Ascend-Max chipsets. After a year of ongoing development, test results now show that the Ascend 910 processor delivers on its performance goals with much lower power consumption than originally planned. Ascend 910 is about two times faster at training AI models than other mainstream training cards using TensorFlow. Apple A13 Bionic Chip via Apple The iPhone 11 is powered by Apple’s new A13 Bionic chip, which Apple touts as its faster processor ever. The A13 also features an Apple-designed 64-bit ARMv8.3-A six-core CPU, with two high-performance cores running at 2.65 GHz. The 2 high performance cores are 20% faster with 30% reduction in power consumption, the 4 high-efficiency cores are 20% faster with a 40% reduction in power consumption. Google’s EdgeTPU via googblogs Edge TPU is Google’s purpose-built chip designed to run AI at the edge. It delivers high performance in a small physical and power footprint, enabling the deployment of high-accuracy AI at the edge. Edge TPU combines custom hardware, open software, and state-of-the-art AI algorithms to provide high-quality, easy to deploy AI solutions for the edge. Graphcore’s IPU via Graphcore The Intelligence Processing Unit (IPU) is completely different from today’s CPU and GPU processors. It is a highly flexible, easy to use, parallel processor that has been designed from the ground up to deliver state of the art performance on current machine intelligence models for both training and inference. It is also complemented by Dell’s DSS8440, the first Graphcore IPU server. Cerebras Wafer Scale Engine (WSE) via Cerebras As manufacturers race towards developing smaller, thinner, cheaper chips, Cerebras took a u-turn and released a wafer-scale engine(WSE), 215mm x 215mm chip aimed at deep learning applications. The WSE is 1.2 trillion transistors, packed onto a single with 400,000 AI-optimised cores, connected by a 100Pbit\/s interconnect. The cores are fed by 18 GB of super-fast, on-chip memory, with an unprecedented 9 PB\/s of memory bandwidth. Tesla’s AI Chip via Tesla The above image is of the processing unit developed by Tesla called Tesla FSD(full self-driving) computer and the board consist of two chips. Each chip consists of a die, of area equal to 260 mm^2, which houses 6 billion transistors. There are two neural network accelerators on the chip and each neural network processor here has a dedicated ReLU hardware and pooling hardware delivering 36 trillion operations per second. The applications of AI is almost everywhere today. From consumer to enterprise applications. With the explosive growth of connected devices, combined with a demand for privacy, low latency and bandwidth constraints, the hardware used for training AI models needs to have that extra edge. The above-discussed innovations barely scratch the surface and there are many small to medium level companies conducting their research in developing customised AI solutions at the hardware level.","excerpt":"Though machine learning has been around for more than three decades, it took a lot of time for the hardware to catch up with the demands of these power-hungry algorithms. With each passing year, the chip-set manufacturers have tried to make the hardware lighter and faster.  Today, over 100 companies are working on building next-generation […]","categories":["AI Trends"],"tags":["latest machine learning innovation"],"author_name":"Ram Sagar","publish_date":"2019-11-18T16:00:39","publication_year":"2019","word_count":840,"keywords":["machine learning","artificial intelligence","TPU","AI","neural network","latest machine learning innovation","Aim","deep learning","edge computing","TensorFlow","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","deep learning","neural network","Aim","TensorFlow","edge computing","TPU","R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/chip-machine-learning-accelerators-tpu-gpu-hardware\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10125122,"title":"AWS AI &amp; SageMaker to Accelerate HALO Trust&#8217;s Landmine Detection in Ukraine","content":"The HALO Trust, a humanitarian landmine clearance organisation, will pilot AI and machine learning to aid its work in Ukraine as part of a new $4 million package of support from Amazon Web Services (AWS). The AWS investment will enable HALO to trial AI to assist in detecting debris of war, including landmines and other explosives, in drone imagery for existing minefields and battlefields in Ukraine. It will also help automate how HALO uses satellite imagery to detect damaged buildings and signs of human activity near battlegrounds and minefields, allowing prioritisation of areas for clearance. “We are delighted to be collaborating with AWS in our life-saving work, and to leverage their robust technological expertise and computing power to more efficiently and precisely identify and clear contaminated land,” said James Cowan, HALO’s CEO. So far, HALO has flown 542 drone flights over Ukraine’s minefields, collecting 11 terabytes of data, with more being added daily. The pilot project is expected to begin trialling drone imagery analysis using AI in the coming months. Satellite and drone imagery is essential for HALO to broadly survey huge swaths of land so its 11,000+ global staff can effectively pinpoint areas needing meticulous clearance by hand or machine. Ukraine is one of the countries where HALO collects massive amounts of high-resolution drone imagery capable of AI analysis to expedite the detection of war debris, including landmines. Training AI to identify explosive remnants of war with drone imagery requires manually labelling thousands of images so algorithms can learn to survey minefields more efficiently than humans. It also requires ongoing vetting and validation by trained HALO personnel to improve model accuracy. AI To Solve Global Crisis “Technology, and in this case, AI specifically, has enormous potential to help solve major global challenges, and we’re looking forward to working with HALO by enabling them to better harness the power of the satellite and drone imagery they are collecting to accelerate the clearing process,” said Dave Levy, VP for Worldwide Public Sector at AWS. The AWS support includes credits enabling HALO to significantly leverage AWS cloud computing infrastructure globally and train staff on its use. AWS is now powering HALO’s critical systems including its Global Operations Information Management System, Geographic Information Systems, drone and satellite imagery processing, and AI\/machine learning models. HALO is also using AWS to run AI algorithms on satellite imagery to identify damaged buildings, power its open-source mapping project in Ukraine, and host its business intelligence tools. This grant enables these technologies to be run in more countries at greater scale, enriching operational decision-making and saving lives. The HALO Trust, founded in Afghanistan in 1988, employs over 11,000 women and men in more than 30 countries in landmine\/munition clearance, risk education, and weapons\/ammunition storage. Clearing landmines is a painstaking process, with most work done by hand, but sometimes machines are more practical depending on the terrain and debris. From land surveyed so far, traditional mine clearance methods could take up to 20 years in Ukraine, but increased capacity, resources, technology and new methods could cut that time by two-thirds. An estimated 470,000 hectares of prime agricultural land in Ukraine is contaminated with landmines. HALO works in partnership with the communities it serves, employing local men and women to create jobs and help families recover. Once mines are cleared, minefields become crop fields, battlefields become playgrounds, and life begins again as fear turns to hope.","excerpt":"An estimated 470,000 hectares of agricultural land in Ukraine is contaminated with landmines; which HALO looks to work on as part of AWS’s $4 million package of support.","categories":["AI News"],"tags":["AWS","SageMaker"],"author_name":"Shyam Nandan Upadhyay","publish_date":"2024-06-28T09:19:43","publication_year":"2024","word_count":566,"keywords":["Go","machine learning","AWS","AI","cloud computing","SageMaker","RAG","ViT","Rust","GAN","R"],"extracted_tech_keywords":["AI","machine learning","RAG","cloud computing","AWS","R","Go","Rust","GAN","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/aws-ai-sagemaker-to-accelerate-halo-trusts-landmine-detection-in-ukraine\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10007465,"title":"12 Best RPA Tools for Free in 2024","content":"Robotic Process Automation (RPA) allows a user to automate repetitive and mundane tasks while ensuring increased productivity, quick outcomes, flexibility, increased customer satisfaction and more. It not only increases productivity but emulates tasks and processes faster than humans. As various organisations across the globe are looking to automate their tasks, the importance of RPA tools has increased more than ever. In this article, we list down 12 best free RPA tools available in the market. 1. Automation Anywhere Automation Anywhere is a Robotic Process Automation (RPA) platform that combines sophisticated RPA, cognitive and embedded analytic technologies. Last year, the company released Automation Anywhere Community Edition, which is a free offering for small businesses, developers and students. The users of the Community Edition will have access to the full suite of Automation Anywhere products, including RPA, IQ Bot (AI), Bot Insight (embedded analytics) and Bot Store, the industry’s first online marketplace for pre-configured downloadable RPA bots. Know more here. 2. Automai Automai Robotic Automation Platform is an RPA platform that specialises in automation solutions for load testing, performance monitoring, regression testing and business automation. The RPA uses AI, smart image recognition, machine learning, OCR and robotic automation to allow businesses to handle high volume and repetitive tasks efficiently. The Robotic Process Automation (RPA) solution consists of three modules — ScenarioBuilder, RPA Manager, and rWorker. Know more here. 3. Automagica Automagica is an open-source robotic process automation platform. It is an automation suite for fully automating tedious, manual tasks on any screen. One can perform various tasks like monitoring, control, scheduling, version control, collaboration, etc. in one single platform. The platform is built on open-source technologies such as Python and Google’s TensorFlow and is therefore easily extensible with the latest and greatest developments in the field of computer vision, NLP, OCR and machine learning. Know more here. 4. Another Monday Another Monday is an end-to-end RPA software suite which is easy to use and efficient to run. The platform provides end-to-end enterprise automation, known as AM Ensemble. AM Ensemble is a comprehensive RPA suite that covers all phases of automation, from process analysis all the way through to bot management. The platform also includes a smart process analysis tool known as AM Muse. The AM Muse automatically learns all cognitive components of the workflow and actively guides a user through the relevant process information. Know more here. 5. Kofax Kofax is an intelligent automation software platform that helps organisations transform information-intensive business processes, reduce manual work and errors, minimise costs, and improve customer engagement. The platform combines RPA, cognitive capture, process orchestration, mobility and engagement, and analytics to ease implementations and deliver results that mitigate compliance risk and increase competitiveness, growth and profitability. Kofax RPA Design Studio is a powerful design tool that lets robot designers interact with live applications as they build out a process. Know more here. 6. Linx Linx is a low-code platform to connect backend systems, apps and customers in ways that let a user innovate new services, automate processes and much more. This platform ensures easy build processes that allow for manipulating and orchestrating data across multiple systems. It also low-codes IDE to create & debug processes, integrate databases, applications, web services and files, among others. Know more here. 7. Pegasystems Pegasystems Pega Robotic Automation is an RPA platform that allows to automate large tasks, optimise work, increase business productivity and more. The platform is said to be non-invasive and equips organisations to bridge legacy systems, close data integration gaps, wrap legacy system integrations without changing any underlying technology investments. The features of this platform include fully-unified business process management, flexible, support for enterprise applications and other such. Know more here. 8. Robot framework Robot Framework is an open-source robotic process automation (RPA) framework that can be used for test automation and robotic process automation (RPA). The tool can be integrated with virtually any other tool to create powerful and flexible RPA solutions. Robot’s core framework is a generic automation platform that is popular in test automation, and the extensibility of Robot Framework is enabled by its modular architecture that builds on top of extension libraries. The core framework is implemented using Python and also runs on Jython (JVM) and IronPython (.NET). Know more here. 9. Robocorp Robocorp is a cloud platform for robotic process automation (RPA) and open-source RPA tools. The framework is a collection of open-source libraries and tools for Robotic Process Automation (RPA), and it is designed to be used with both robot framework and Python. The goal of this framework is to offer well-documented and actively maintained core libraries for software robot developers. Know more here. 10. Taskt Taskt, formerly known as sharpRPA, is a free and open-source robotic process automation (RPA) solution powered by the .NET framework in C#. This tool allows you to build and design process automation without needing to write application code. It also automates tedious and repetitive tasks with simple and customisable commands. The platform includes a number of commands as a part of the standard build. The advanced features are included, such as executing custom code, OCR, and image recognition. Know more here. 11. TagUI TagUI is free to use and open-source command-line tool for digital process automation (RPA) that allows you to automate your desktop web, mouse and keyboard actions easily. The TagUI version 6.0.0 includes several features such as TagUI live mode, click using OCR, deploy flows to run when double-clicked, running flows with options can be done with abbreviations and more. Know more here. 12. WorkFusion WorkFusion is a leading vendor of intelligent a000utomation solutions combining AI, RPA and machine learning capabilities for business process automation. WorkFusion combines all of the AI-powered capabilities that global businesses need to digitise into enterprise-grade products purpose-built for operations professionals. The tool RPA Express lets customers freely automate the manual work of integrating Citrix, Oracle, SAP and other vital core systems by eliminating the “swivel chair” tasks of entering credentials, navigating application interfaces and performing core systems functions. Know more here.","excerpt":"Robotic Process Automation (RPA) allows a user to automate repetitive and mundane tasks while ensuring increased productivity, quick outcomes, flexibility, increased customer satisfaction and more. It not only increases productivity but emulates tasks and processes faster than humans. As various organisations across the globe are looking to automate their tasks, the importance of RPA tools […]","categories":["AI Trends"],"tags":["business analysis tools","RPA developer"],"author_name":"Ambika Choudhury","publish_date":"2020-09-16T14:00:39","publication_year":"2020","word_count":1003,"keywords":["Go","RPA developer","machine learning","AI","business analysis tools","image recognition","computer vision","NLP","Python","analytics","TensorFlow","R"],"extracted_tech_keywords":["AI","machine learning","NLP","computer vision","analytics","TensorFlow","image recognition","Python","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/12-free-rpa-tools-in-the-market\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10126702,"title":"Kaggle is a Humbling Machine for PhD Students","content":"Kaggle competitions, the hardcore community that organises challenges for machine learning experts, brings out both the best and the worst in participants. It clearly prioritises application-based expertise over theoretical knowledge of the subject. Register for this Free GenAI Workshop > In a hot take, Jeremy Howard, the co-founder of Fast.ai, said that though PhD students can succeed at Kaggle competitions, it’s the ones who assume an easy win “generally fail miserably”. This conversation was brought in as Thomas Wolf, the co-founder & CSO of Hugging Face, shared about a super-impressive AI competition, Artificial Intelligence Math Olympiad (AIMO), where AI models, not individuals, competed to solve challenging maths problems of the International Math Olympiad-level for a $1M prize money. There was a super impressive AI competition that happened last week that many people missed in the noise of AI world. I happen to know several participants so let me tell you a bit of this story as a Sunday morning coffee time.You probably know the Millennium Prize Problems… pic.twitter.com\/LeWnVbjH8m— Thomas Wolf (@Thom_Wolf) July 7, 2024 A team from Numina, collaborating with Hugging Face, trained models using high-quality Chain-of-Thought (CoT) data and innovative methods. Despite initial performance challenges, they significantly improved their models, eventually winning the competition by solving 29 challenges. This showcases AI’s potential in advancing complex mathematical problem-solving, hinting at a broader impact on scientific progress in the near future. As Terence Tao, regarded as of one the greatest living mathematicians himself set it up, this is “higher than expected”. Sometimes, better than PhD researchers. This is something that has been agreed upon by several researchers in the field. Max Mynter, a self-taught ML engineer, replied to Howard saying that he had tried his hand at a few competitions and was utterly humbled. “Was a slap in the face that made me learn (again) that there is a difference between theoretical knowledge and practical skill,” said Mynter. This led Mynter to gain more expertise and experience in the field by putting in the work required and “getting off the high horse”. “I was pretty good at walking through institutions and talking the talk of what prestigious institutions wanted to hear. But after a couple of years, I feel that I am building real skill.” Similar thoughts were shared by others. “19 yo me pursuing bachelor’s degree thought I could win ‘few’ competitions easily because I self learnt ML and I was so good at it. By the time I finally won one, 2 years later, I realised how underskilled I am.” Not everyone agrees though. Venu Vasudevan, the director of AI and founding team member of Zigbee, said, “Kaggle is the Rube Goldberg of CS. Not clear it represents street smart problem solving.” He added that he had participated in one of the competitions and came in the 5th percentile. He decided it wasn’t his cup of tea. Kaggle’s Reality Check For some people, even getting the data for the Kaggle competition is an unexpectedly difficult task. It is crucial for academic people to get into Kaggle competitions to learn what the real field is working on. Jean-Francois Puget from NVIDIA, who is also a 3X Kaggle Grandmaster, believes that it should be mandatory for an academic working in machine learning to enter a Kaggle competition. There is another interesting tale which narrates that Kaggle can land you jobs, not by winning them, but also learning from being a part of them. Yann LeCun shared that Jure Zbontar, a computer scientist from the University of Ljubljana, joined Meta’s FAIR team and worked on various projects with him. LeCun said that Zbontar’s advisor emailed him about how he kept winning Kaggle competitions and asked if he would co-advice Zbontar. He then joined NYU with LeCun and completed his PhD and topped the leaderboard of the stereo vision benchmark. Though it is not advisable for people to completely depend on Kaggle prizes for survival. It is also quite interesting to see that most of the top performers on Kaggle are either PhD or MS graduates in 2022. Kaggle definitely helps in getting people the skill they need for jobs. When speaking with AIM in 2022, Ruchi Bhatia, the youngest 3X Kaggle Grandmaster, also advised people that getting into Kaggle is something necessary to “hone our competitive side, but at the same time not forget about the underlying goal: learning”. As for the PhD students, it is definitely necessary for them to try their hands on practical skills instead of just focusing on theory.","excerpt":"PhD students can succeed at Kaggle competitions, but it’s the ones who assume an easy win “generally fail miserably”.","categories":["AI Features"],"tags":["Kaggle"],"author_name":"Mohit Pandey","publish_date":"2024-07-12T15:59:55","publication_year":"2024","word_count":751,"keywords":["Go","Hugging Face","GenAI","machine learning","artificial intelligence","Kaggle","AI","ML","Aim","GAN","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","GenAI","Aim","Hugging Face","R","Go","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/kaggle-is-a-humbling-machine-for-phd-students\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10075257,"title":"Karnataka Government Signs MoU with Dell Technologies to Impart Cyber Safety Practices","content":"The Government of Karnataka and Dell Technologies have inked a memorandum of understanding (MoU) on Tuesday. The IT firm has collaborated with the state’s Cyber Centre of Excellence (CySecK) to improve current online security practices. The MoU was signed in the presence of Karnataka Higher Education Minister C N Ashwath Narayan and vice-president Manish Gupta, Dell Technologies, along with Ripu Bajwa, director sales and general manager–Data Protection Solutions, Dell Technologies. According to the agreement, Dell Technologies, through its ‘Tech for Social Good’ initiative, will roll out interactive courses on cyber hygiene best practices for school students. These courses are also expected to be curated for small- and medium-scale enterprises as well as for government functionaries. “To achieve the target of becoming a $300 billion digital economy by 2025, putting in place an efficient cyber security system is necessary,” said C N Ashwath Narayan. He further added that the agreement would facilitate cyber security awareness and practices in student circles, startups and public offices, and would also benefit both small- and medium-scale enterprises. Karthik Rao Bappanad, Centre Head, CySecK, said that the centre is working on expanding the cyber security practices to tier-2 and tier-3 cities. He emphasised that considering the government’s vision for Karnataka to have a 300 billion digital economy by 2025, digitalisation would require a strong foundation of cybersecurity. Furthermore, under the MoU, Dell would provide learning of cyber safety practices through its interactive courses available in Kannada and English. Several other projects have been established to accelerate the development of cybersecurity technologies in the country, such as the Ministry of Electronics & IT (MeitY), in partnership with the Data Security Council of India (DSCI), launched the National Centre of Excellence for Cybersecurity in Noida in 2020.","excerpt":"The IT firm has collaborated with the state’s Cyber Centre of Excellence to improve current online security practices.","categories":["AI News"],"tags":[],"author_name":"Bhuvana Kamath","publish_date":"2022-09-14T18:52:24","publication_year":"2022","word_count":289,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","Git","Ray","R","startup"],"extracted_tech_keywords":["AI","Ray","R","Go","Git","startup","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/karnataka-government-signs-mou-with-dell-technologies-to-impart-cyber-safety-practices\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":43883,"title":"What Is A Destructive Malware Attack &#038; How Your Organisation Can Combat It","content":"Over the years, cybersecurity threats have created massive catastrophes not only for organisations but also for individuals. No environment is immune to cyberattacks as hackers continue to develop new tools and techniques, targeting corporate assets stored on cloud infrastructure, individuals’ mobile devices, trusted third-party supplier apps and even popular email platforms. According to a report by Check Point, this year there is more than a 50% increase in attacks compared to 2018 and malware has been one of the major threats. Destructive malware such as banking malware has evolved to become a very common yet notorious threat. Today, banking malware is capable of stealing payment data, credentials and funds from victims’ bank accounts. And the worst thing is that new versions of these kinds of destructive malware are ready and out for massive distribution. Emotet, Trickbot, Dorkbot, Triada, Rannit, Ursnif etc. are some of the examples of notorious malware that has been really active in the first of this year. From being able to self-propagate to being able to deploy other malware, they have created a scary environment in this internet-driven world. The Skyrocketing Rise Of Destructive Malware It is not just Check Point that has a take on the seriousness of malware threats, IBM has also done research. There was a time when destructive malware was used by hackers who had the license to hack, but now the attacks have also become popular among cybercriminals. The rate at which these threats are being used is massive — whether it’s just a generic malware or ransomware. According to IBM X-Force Incident Response and Intelligence Services (IRIS), there is a whopping 200% increase in the amount of these kinds of destructive malware attacks and a wide variety of businesses have already been targeted by the attackers. IBM IRIS also mentioned that the ransomware attack calls to their emergency response hotline have increased in such a rate that it has doubled to reach a skyrocketing rate of 116%. The report also states that the damage that companies experience because of these malware attacks is significantly high, to be precise, it is on average costs multinational companies around $239 million, per incident. All this includes expenses from remediation to equipment replacement to lost productivity to other business damages. And, to the surprise, this amount of money is 61 times greater than the cost of a typical data breach. Companies that are low on revenue and high on data\/information quality can be impacted severely by these kinds of destructive malware attack. It might even end up taking the company to such a level that might run out of any capital to get things fixed. Talking in terms of workstations that get affected by destructive malware attacks, the number is approximately 12,316. This number definitely show the capability of these attacks and how significantly they can impact the workstations. How To Keep Safe Your Data Destructive malware is considered to be one of the latest advancements in the world of cyber threat. The malware is literally capable of taking cunning approaches in order to let the polymorphic malware enters and hides within a computer. Once this malware settles inside a system, then it triggers download a payload that basically disables access to data or destroys system functions across geographies and industries. According to US-CERT, a destructive malware basically target a large scope of systems and execute across multiple systems throughout a network. Therefore, imperative for organisations that are still not ready for this threat to take the necessary measures. Here the steps that your organisation can take in order to combat destructive malware: Make sure you have a strong and active incident response infrastructure. Test whether your response plan, whether it’s working completely fine and the best way to do it is by testing it under pressure. Also, make sure your company holds discussion-based sessions to talk about threats and find solutions to that. You can also read our article, “How Organisations Can Build An Effective Incident Response Framework.” Have robust threat intelligence as a security intelligence component in order to better understand the threat landscape. It also helps in providing information about how to protect an organisation from external and insider threats. Do not compromise on the layers of security controls across the entire cybersecurity infrastructure. Make sure your company has all the best in class components and tools to combat cyber-attacks. Have multi-factor authentication (MFA) in all aspects of access. To better understand MFA, you can read our article, “Understanding The 5 Factors Of Multi-Factor Authentication.” Backup is more important than ever. You never know what goes wrong and when your data gets compromised. Backup helps you big time when its about retrieving data without spending a significant amount of money.","excerpt":"Over the years, cybersecurity threats have created massive catastrophes not only for organisations but also for individuals. No environment is immune to cyberattacks as hackers continue to develop new tools and techniques, targeting corporate assets stored on cloud infrastructure, individuals’ mobile devices, trusted third-party supplier apps and even popular email platforms. According to a report […]","categories":["AI Features"],"tags":["Cyber Security","cyber threat","IBM","malware"],"author_name":"Harshajit Sarmah","publish_date":"2019-08-06T15:20:05","publication_year":"2019","word_count":788,"keywords":["Go","API","malware","Cyber Security","programming_languages:R","AI","programming_languages:Go","RAG","ViT","IBM","cyber threat","Rust","GAN","R"],"extracted_tech_keywords":["AI","RAG","R","Go","Rust","API","GAN","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/what-is-a-destructive-malware-attack-how-your-organisation-can-combat-it\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10066955,"title":"Top institutes leading in quantum computing research","content":"Though quantum computing is still in a nascent stage, there is a growing interest in this area due to the myriad benefits this technology can bring. A PitchBook report of September last year said that investors had put in USD 1.02 billion into quantum computing companies till September 2021. Recently, the World Economic Forum released the first Quantum Computing guidelines. Universities across the world are investing in research for the advancement of this technology. Let us look at a few of the top institutes leading in quantum computing research. The list is in no particular order. IISc – Centre for Excellence in Quantum Technology Faculty members from physics, engineering, and computer science at IISc are working in this area. As the existing expertise and facilities available in India are limited on the theoretical as well as the experimental side in this domain, this centre aims to accelerate Indian pursuit in this field on par with the rest of the world. The experimental program in this centre focuses on superconducting qubit devices, single-photon sources and detectors for quantum communications, integrated photonic quantum networks, quantum sensors and much more. For more details, click here. Tata Institute of Fundamental Research The Quantum Measurement and Control Laboratory (QuMaC) at TIFR focuses on investigating quantum phenomena in superconducting circuits. Research areas include quantum error correction, quantum simulations, novel qubit designs, quantum limited parametric amplifiers, etc. The lab has state-of-the-art equipment to facilitate research. These include Cryogen Free Dilution Refrigerator (Oxford Triton 200) to cool down the devices to about 10 mK, which helps in studying quantum phenomena in a low noise environment.The superconducting devices are fabricated using electron-beam and optical lithography in the TIFR Nanofabrication facility.The researchers can fabricate custom RF circuit boards with metal-plated vias using the LPKF PCB Prototyping machine. For more details, click here. IISER Pune The I-Hub Quantum Technology Foundation has been set up in IISER-Pune with support from GoI’s Department of Science and Technology. It aims to develop quantum computers and novel quantum materials for day-to-day applications and is led by faculty from the physics department. As per the institute, I-HUB aims to harness quantum phenomena for developing advanced computing systems and for more immediate applications, such as precision sensors, navigation devices for global positioning systems, geological mapping, atomic clocks, encrypted communication, and novel materials. I-HUB also helps in technology translation, incubation and human resource development. For more details, click here. IIT Madras The Centre for Quantum Information, Communication and Computing is an interdisciplinary research centre located in IIT-Madras. It has scholars from different disciplines such as physics, electrical engineering and computer science. Areas of research include quantum key Distribution, secure communication networks, quantum sensors, femtoTesla sensitivity, quantum computing, cluster state computing, machine learning, post-quantum cryptography, quantum algorithms, etc. For more details, click here. Outside India Centre for Quantum Information and Foundations: Cambridge University The Centre for Quantum Information and Foundations is located within the department for applied math and theoretical physics. It conducts theoretical research related to all aspects of quantum information processing and the implications of quantum computing and answers foundational questions in quantum physics. For more details, click here. Harvard Quantum Initiative It is a community of researchers with an intense interest in advancing the science and engineering of quantum systems and their applications. It is co-directed by professors John Doyle, Evelyn Hu, and Mikhail Lukin. HQI brings together scientists and engineers across universities, companies, and government, to leverage quantum effects like superposition and entanglement to impact how information is acquired, stored, sent, and processed. For more details, click here.","excerpt":"IISc plans to bring the Indian pursuit in this field on par with the rest of the world, with a dedicated and focused effort.","categories":["AI Trends"],"tags":["quantum","Quantum Computing","Research"],"author_name":"Sreejani Bhattacharyya","publish_date":"2022-05-13T16:00:00","publication_year":"2022","word_count":593,"keywords":["Quantum Computing","Go","machine learning","programming_languages:R","AI","programming_languages:Go","RAG","quantum","Aim","Research","ViT","R","emerging_tech:quantum computing"],"extracted_tech_keywords":["AI","machine learning","Aim","RAG","R","Go","ViT","programming_languages:R","programming_languages:Go","emerging_tech:quantum computing"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-institutes-leading-in-quantum-computing-research\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":20246,"title":"Gearing up for a career in Data Science? Don’t forget these 5 Intangibles","content":"Are you wondering about the hype around Data Science? Data says that it’s not a hype anymore. Glassdoor released its Report of 50 Best jobs in America in 2017. With a job score of 4.8 out of 5, a job satisfaction score of 4.4 out of 5, and a median base salary of $110,000, Data Scientist jobs came in first for the second year in a row. IBM predicts demand for Data Scientists will soar 28% by 2020. Undoubtedly, a career in Data science is a lucrative one. Data science is an interdisciplinary field which can become a career option for most aspirants with appropriate training and self-motivation. Usually, the pre-requisite is an educational degree in any of the following disciplines — mathematics, computer science, economics, statistics, engineering, operations research or business management. However, aspirants from other educational streams can also get into this field with appropriate training. What is required is a dedicated focus on learning the hard skills that enable you to first get selected in a data science job and then perform and subsequently develop your career in data science. These hard skills include, but are not limited to, programming and data mining capabilities through languages like R, Python, SAS, SQL, Hive, Pig, Julia, SPSS etc. as well as deep understanding of statistical techniques for data analysis and analytical solutions like basic descriptive statistics, hypothesis testing, regression and segmentation techniques and some exposure to advanced techniques like machine learning and optimization. The web is flooded with detailed information and guidance on these. Analyttica TreasureHunt could well-become a partner in this journey of yours, watch out for this platform. In this race to get up-skilled and apply for positions that match the hard skill requirements, aspirants tend to overlook the “X-factors” that often turn out to be the deciding parameter for job selection. These factors, the Intangibles, are usually the unspoken qualities that employers look for in aspirants, when interviewing them for a data science role. The weight on these intangibles often increases with increasing years of experience. While you cannot gauge the exact questions around the intangibles during an interview, you can prepare for various situations beforehand and embed some of these into your day-to-day — this will help you become a stronger data science professional in due course. In this blog, we try to give you a snippet of five of the most important intangible skills, in an effort to kick-start your preparation for that job you have been eyeing in data science. Investigative Mind Curiosity and the hunger to ask questions fearlessly mostly make a great analyst . The ability to constantly investigate existing hypothesis and create new ones, often defines the key responsibility of many data science jobs. Data scientists are expected to spot changes in trends and identify the cause behind the change. This requires a strong eye for detail along with the tendency to associate causality with changes. Employers adapt multiple ways to test an applicant’s level of strength on this aspect. It can be done through numeric, verbal and logical reasoning puzzles that you will be asked to solve during the interview. You can practice these through available resources online to be prepared for these. Interviewers tend to test the detail-orientation of the applicant through specific questions at each stage of the projects they have mentioned in their resume. These questions could range from your contribution, to the broader scope of the project. Interviewers will evaluate if you tend to focus only on the work that is assigned to you, or if you have a knack of understanding the business context as well. The best way to be prepared for this kind of discussion is to revise your earlier projects with a very structured approach in mind. And never, ever shy away from being honest to your interviewer regarding your involvement in the project. 2. Problem solving through Structured Thinking Problem solving lies at the heart of Data science. It is the ability to break down broad business problems into smaller steps that often differentiates a great data scientist from an average one. The first step is to understand the business problem, convert that into an analytical problem, design the analytical solution and ultimately convert that into a business solution. The business problem must be broken down into certain hypotheses which then would be tested through the analytical process. The most critical component of that process is to seek all the relevant data sources and their accuracy. These hypotheses are tested using various analytical tools and techniques. Interviewers would give you a specific business problem and observe how you approach it. They will often ask you to create the various business hypotheses and question how you would validate those. The best way to be prepared for this part of the discussion is to practice various analytical problems as case studies with friends from the analytics or related fraternity. 3. Effective communication skills Like many other jobs, effective communication is an important aspect of being a good data analyst. A great business analyst effectively trains the reflexes to think business first, and the analyses as an enabler, not an end to itself. That is how the analyses can be made to bridge the gap to the business metrics, and write\/script a story on it in a way the stakeholders understand and implement. Communication skills for a data scientist role often includes but is not limited to verbal reasoning, proactive timeline management, making impactful presentations, negotiation skills, influencing skills and crisp email communication. Working in collaboration with the business teams and conveying complex analytical solutions in simple yet impactful business parlance is a crucial part of the job. The ability to interpret statistical outputs is a highly underrated skill in the analytics industry. Interviewers would test you on communication skills by asking you to communicate statistical results into simple language. The best way to be prepared for this part of the discussion is to understand the interpretation of statistical outputs in business language. The ability to translate the analytical outputs to simple language would be a differentiator in this part of the interview. 4. Lateral Thinking One must develop the ability to apply existing methodology or solution to a completely new and different business problem. Here’s an example of survival analysis, which predicts the time until the occurrence of an event of interest, such as how long would a patient survive a certain chronic disease. This technique was used mostly in the healthcare and engineering industry. Now they are being used to solve business problems across industries. The interviewer would provide business situations from her industry and expect you to apply your learnings on that industry. The ability of leveraging analytical techniques from different industries is one of the primary reasons for organizations to invest in diversified data science talent. The best preparation for this would be to become savvy on reading and learning from various resources — articles, blogs, books, videos, podcasts covering cross-industry application of analytical techniques. 5. Continuous Reinvention Change is the only constant in today’s world, especially in this exploding pace of the analytics industry. The only way for anyone to survive and succeed is to constantly re-invent themselves. While traditional regression techniques could solve majority of the prediction problems few years back, Machine Learning and Artificial Intelligence are fast becoming the new mantra. While interviewers might not expect you to know these in-depth, unless specified in the role requirement, they would nonetheless expect you to have a point of view and broad understanding of these techniques — when do you apply them and what are some merits and demerits of the same? The best preparation would be to keep upskilling yourself through webinars, conferences and online courses covering a wide range of topics in the data science space. These self-motivated learnings definitely create a positive and lasting impact both on your potential employer and yourself as a future data scientist. The key to success is to remember the age-old philosophy of making your own mistakes and learning from them. Always be true to your real-self, you will have better grasp on any conversation by being truthful. Wish you a fulfilling career in data science! P.S. Don’t miss how Analyttica TreasureHunt can become your partner in this journey to a successful career in data science! Watch this space for more. Subhadra Dutta has 12+ years of corporate experience in banking and financial services (BFS) industry with a focus on Analytics advisory, consulting and project delivery. She has expertise in conceptualizing and designing analytical solutions, product solutions and rewards analytics across consumer banking products, especially credit cards. She has a Masters in Economics from JNU, New Delhi, India.","excerpt":"Are you wondering about the hype around Data Science? Data says that it’s not a hype anymore. Glassdoor released its Report of 50 Best jobs in America in 2017. With a job score of 4.8 out of 5, a job satisfaction score of 4.4 out of 5, and a median base salary of $110,000, Data […]","categories":["AI Highlights"],"tags":["survival regression python"],"author_name":"AIM Media House","publish_date":"2018-01-03T08:17:53","publication_year":"2018","word_count":1448,"keywords":["data science","machine learning","artificial intelligence","TPU","AI","survival regression python","RAG","Python","analytics","SQL","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","data science","analytics","RAG","TPU","Python","R","SQL"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/gearing-career-data-science-dont-forget-5-intangibles\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":26748,"title":"India, UAE Sign $20 Billion Deal For Artificial Intelligence Cooperation","content":"Invest India and UAE Ministry sign MoU for technological cooperation. (Image source: @PIB_India\/Twitter) India and the UAE signed a Memorandum of Understanding (MoU) for the India-UAE Artificial Intelligence Bridge in New Delhi. A statement from the Ministry of Commerce & Industry states that this partnership will generate an estimated $20 billion in economic benefits during the next decade for both countries. The MoU will spur development across areas like blockchain, AI and analytics as data and processing will be a catalyst for innovation and business growth. They will also serve as the backbone of more effective and efficient service delivery systems. By 2035 AI can potentially add $957 billion to the Indian economy, said the statement by Press Information Bureau. The MoU was signed in the presence of Minister of Commerce and Industry and Civil Aviation, Suresh Prabhu and HE Ahmad Sultan Al Falahi, Minister Plenipotentiary — Commercial Attache, UAE Embassy at the India leg of GovHack series of World Government Summit. The Minister appreciated the effort made by UAE on leading the initiative to change the government and governance through technology and reiterated India’s commitment with UAE in the field of AI. The UAE-India collaboration will seek to evaluate the dynamic nature of innovation and technology by convening a UAE-India AI Working Group (TWG) between the UAE Ministry for Artificial Intelligence, Invest India and Startup India. The TWG will meet once a year with the mandate to increase investment in AI startups and research activities in partnership with the private sector. India and UAE share a bond that extends beyond business as Indians make-up the largest expat community in the UAE at 27 percent and the UAE is India’s third-largest trading partner. The UAE has invested over $5.3 billion in India and infrastructure is one of the top 5 focus sections of UAE-India bilateral trade. The UAE has committed $75 billion towards infrastructure development in India. The Government of India is launching multiple initiatives to create an environment for digital growth through which the potential of AI can be realised in the areas of agriculture supply, healthcare and disaster management services.","excerpt":"India and the UAE signed a Memorandum of Understanding (MoU) for the India-UAE Artificial Intelligence Bridge in New Delhi. A statement from the Ministry of Commerce & Industry states that this partnership will generate an estimated $20 billion in economic benefits during the next decade for both countries. The MoU will spur development across areas like […]","categories":["AI News"],"tags":["AI (Artificial Intelligence)","suresh prabhu","uae"],"author_name":"Prajakta Hebbar","publish_date":"2018-07-30T06:25:36","publication_year":"2018","word_count":352,"keywords":["Go","artificial intelligence","uae","AI","IPO","innovation","Git","ViT","analytics","suresh prabhu","R","AI (Artificial Intelligence)","startup"],"extracted_tech_keywords":["AI","artificial intelligence","analytics","R","Go","Git","ViT","innovation","startup","IPO"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/india-uae-sign-20-billion-deal-for-ai-cooperation\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":33660,"title":"NVIDIA’s Real-Time Ray Tracing Technique and AI-powered RTX Technology Explained","content":"Recently, NVIDIA announced that they have added another offering to their consumer GPU lineup. The entry-level RTX card known as the RTX 2060 comes after the launch of their RTX 2070 and 2080 models. This new lineup of cards has everyone in the gaming space excited, as it brings with it the promise of the holy grail of rendering; dubbed as real-time ray tracing. But what exactly is ray tracing? This article takes a deeper look into RTX, and how technologies like AI and deep learning and bringing cinematic realism to video games. What is Ray Tracing? Ray tracing is a procedure to generate an accurate digital representation of how light rays would affect the visual appearance of an object. This technology has been around since the late 1970s and has since been widely used in Hollywood to render elements in most of their CGI scenes. The procedure aims to bring realistic lighting to simulations by emulating the physical behavior of light. It is implemented by projecting rays of light from the camera, ‘tracing’ it until it interacts with an object, and calculating where each ray intersects an object in the scene. This then allows it to calculate the way pixel is affected by the ray of light, and continue to trace the ray as it bounces around the scene. This requires the computer to follow some number of rays out from each intersection to model reflected light, thus creating an exponentially increasing number of rays in the scene. All of these procedures culminate in the calculation of the color of the pixels in the scene, thus providing a realistic look due to light movement. These calculations are compute-intensive, leading to the need for an extremely powerful computer for ray tracing. Each frame of ray-traced content can take hours on a mini-supercomputer class computer, thus making it a technology that Hollywood uses to make CGI in cinemas look realistic. Pixar’s Monsters University utilised ray-tracing extensively, as the technology can reproduce many of the effects created by a camera such as reflections, subsurface scattering, and translucency. Due to the high amount of performance required to implement ray tracing in a scene, it has long since been considered a goal of technology to implement real-time ray tracing. Video games are an especially difficult space to implement real-time ray tracing, as they output anywhere between 30 to 60 frames in one second. However, NVIDIA has achieved this and has provided a solution that can be implemented in consumer-grade graphics cards with their Turing series of products, which includes the RTX cards. RTX and real-time ray tracing To understand how RTX cards are a leap forward in GPU technology, we must first understand how games are rendered today. The work is split mainly between the GPU core and the on-board HBM (high-bandwidth memory). While HBM stores the texture data, such as 3D models, skyboxes, wall textures, and GUI elements, the GPU core calculates how the image should look through a process called rasterization. Rasterization involves interacting with 3D models created from a mesh of triangles called polygons, which are then mapped into pixels for display on a 2D display. This, along with the shading and processing of pixels, is conducted by the GPU core. As this technique has been the industry standard for video games since the early 2000s, there have been multiple advancements made in order to optimise the experience. As of now, optimisation in video games is almost a perfect art, with industry standard practices becoming prevalent in almost every prominent title. RTX, however, does not aim to turn all of this on its head. After all, it was NVIDIA who established a majority of these norms in the first place. Instead, the RTX series cards bring optimisations to both existing rasterization-based cards, along with an extra module for real-time ray tracing known as RTCores. These cores act as discrete processing for ray tracing in real time and calculate the various compute-power intensive processes required for ray tracing. In order to be accessible to consumers, these cards come with a host of optimisations that allow them to achieve ray tracing. Through software, the cards detect how many reflections it needs to follow for a specific ray, depending on the effect on the scene as a whole. There is also intensive post-processing which utilises AI, and a supersampling technique powered by deep learning to offset the performance loss caused by enabling RTX. The highest end card offered in the RTX lineup currently is the RTX 2080Ti, which can render 10 GigaRays\/s. NVIDIA predicts that upcoming solutions will also employ similar hybrid renderers that combine both rasterization and ray tracing, requiring tight integration with existing APIs. This is achieved by another bunch of accompanying software and APIs to enable it to reach maximum adoption among developers and users alike. NVIDIA has already worked with Microsoft closely in order to bring RTX to the widely used DirectX API for creating games. They have also collaborated with Khronos Group to bring this technology to the Vulkan platform, which is said to be coming soon. Reportedly, a “cross-vendor ray tracing capability” is possible, as NVIDIA is contributing the design to the Khronos Group. The State of RTX RTX is supported in 10+ games currently and does not provide ray-tracing support for other games. However, the technology is set for future adoption due to the open API standards and tight integration that Nvidia is bringing to the table. RTX causes a visible performance hit, which is offset by a technology known as DLSS, which stands for deep learning supersampling. In order to create this, Nvidia trains a neural network on pre-release game engine images at lower and higher resolutions. The AI provides the weights information for Tensor Cores in consumer GPUs through driver updates. When DLSS is turned on, the game is rendered at a lower resolution, with Tensor cores working to upscale to a higher resolution using Deep Learning. This results in a higher frame rate with a slightly worse image at high resolution. This can be used in conjunction with ray tracing to provide better framerates. NVIDIA claims that users can achieve performance similar to ray tracing off with a combination of DLSS and ray tracing on. The introduction of this algorithm is required to have playable performance in games, as there is a need to reduce noise in images delivered by ray tracing. This is due to the fact that an extremely high number of rays cannot be cast, leading to a lot of noise in the image. The algorithm functions on the Tensor cores in the GPUs, providing a fast and quality output without sacrificing performance. NVIDIA stated, “These are early days for AI being applied to graphics and the above results are promising”, leading many to believe that AI is the next step in improving graphics performance. Demos NVIDIA released a variety of demos that showcase the capabilities of real-time ray tracing in games. Here are some examples to illustrate what difference RTX actually makes in real-world applications.","excerpt":"Recently, NVIDIA announced that they have added another offering to their consumer GPU lineup. The entry-level RTX card known as the RTX 2060 comes after the launch of their RTX 2070 and 2080 models. This new lineup of cards has everyone in the gaming space excited, as it brings with it the promise of the […]","categories":["Global Tech"],"tags":["GPUs","Machine Learning","Neural Network","NVIDIA","real-time","video games"],"author_name":"Anirudh VK","publish_date":"2019-01-17T12:37:07","publication_year":"2019","word_count":1175,"keywords":["Neural Network","Go","video games","API","TPU","AI","neural network","Machine Learning","Git","Ray","Aim","deep learning","real-time","GPUs","NVIDIA","R"],"extracted_tech_keywords":["AI","deep learning","neural network","Aim","Ray","TPU","R","Go","Git","API"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/nvidias-real-time-ray-tracing-ai-powered-rtx-explained\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10130040,"title":"Now You Can Run Llama 3.1 405B on Your Computer Using Peer-to-Peer Network","content":"Not everyone can access highly spec’d machines capable of running LLMs locally, which often require substantial computational power and memory. “GPUs like H100s, which are essential to train and run LLMs efficiently on a large scale, are beyond the budgets of most startups. And running models like Llama 3.1 405B is unthinkable for regular people. “Renting GPUs and running them on a single cluster or using peer-to-peer connections is one of the easiest ways to do it,” Arjun Reddy, the co-founder of Nidum.AI, told AIM. We're not trying to regulate you small opensource AI projects. We're regulating these huge data centers and billion dollar corporations. Theres no way you could get millions of computers to compete with us. Peer to peer knows a way. #FreedomToCompute pic.twitter.com\/S9nOsadTAF— Lambda Rick \/acc (@benrayfield) April 29, 2024 P2P technology is already used in blockchains, which is a testimony to how secure the network can be. P2P technology came into the limelight for the first time in 1999, when Napster used P2P technology to decentralise music, allowing users to download and host music files from their own computers. Reddy further explained the approach they follow for the P2P technology. It starts with fine-tuning the existing model for specific needs, which is then divided into hundreds of small parts and described to the P2P network. A layer of encryption is used to safeguard data. To showcase the flexibility of P2P technology, Reddy is about to host the largest decentralised AI event later this week where hundreds of Apple computers will be used to run Llama 3.1 through the P2P network. The idea is to demonstrate the importance of decentralised networks to run LLMs. The Promise of Peer-to-Peer Network P2P networks, popularised by file-sharing systems like BitTorrent, distribute tasks across multiple nodes, each contributing a portion of the overall workload. Applying this concept to AI, a P2P network could theoretically distribute the training of an LLM across numerous consumer-grade GPUs, making it possible for individuals and smaller organisations to participate in AI development. A research paper titled ‘A Peer-to-Peer Decentralised Large Language Models’ discusses a provably guaranteed federated learning (FL) algorithm designed for training adversarial deep neural networks, highlighting the potential of decentralised approaches for LLMs. A study by Šajina Robert et al. explored multi-task peer-to-peer learning using an encoder-only Transformer model. This approach demonstrated that collaborative training in a P2P network could effectively handle multiple NLP tasks, highlighting the versatility of such systems. Another significant contribution comes from Sree Bhargavi Balija and colleagues, who investigated building communication-efficient asynchronous P2P federated LLMs with blockchain technology. Their work emphasises the importance of minimising communication overhead and ensuring data integrity in decentralised networks. But There are Challenges… Despite the promise, significant challenges hinder the practical implementation of P2P networks for LLMs. One major issue is the bandwidth and latency required for efficient training. Training LLMs involves transferring vast amounts of data between nodes, which can be prohibitively slow on consumer-grade networks. One Reddit user pointed out that even on a 10-gigabit network, the data transfer rates would be insufficient compared to the high-speed interconnects used in dedicated GPU clusters. Moreover, the synchronisation required for distributed gradient descent, a common optimisation algorithm in training neural networks, adds another layer of complexity. Traditional training methods rely on tight synchronisation between nodes, which is difficult to achieve in a decentralised setting. A research paper on the review of synchronous stochastic gradient descent (Sync-SGD) highlights the impact of stragglers and high latency on the efficiency of distributed training. … And Solutions Despite these challenges, ongoing efforts exist to make decentralised AI a reality. Projects like Petals and Hivemind are exploring ways to enable distributed inference and training of LLMs. Petals, for example, aims to facilitate the distributed inference of large models by allowing users to contribute their computational resources in exchange for access to the network’s collective AI capabilities. Additionally, the concept of federated learning offers a more feasible approach to decentralised AI. In federated learning, multiple nodes train a model on their local data and periodically share their updates with a central server, which aggregates the updates to improve the global model. This method preserves data privacy and reduces the need for extensive data transfer between nodes. It could also be a practical solution for decentralised AI, especially in privacy-sensitive applications like medical machine learning.","excerpt":"Nidum.AI plans to use 2000+ Apple computers to run Llama 3.1 on P2P network.","categories":["AI Trends"],"tags":["Llama 3"],"author_name":"Sagar Sharma","publish_date":"2024-07-24T11:57:16","publication_year":"2024","word_count":723,"keywords":["Llama 3","federated learning","Go","machine learning","AI","neural network","RAG","NLP","Ray","Aim","R"],"extracted_tech_keywords":["AI","machine learning","neural network","NLP","Aim","Ray","federated learning","RAG","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/now-you-can-run-llama-3-1-405b-on-your-computer-using-peer-to-peer-network\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10050491,"title":"How Godrej Housing Finance Blitzscaled During Pandemic, Using Its Risk Management Strategies","content":"“Identification of risks and controls is of utmost importance in any process before it is implemented,” says Shalinee Mimani, the Chief Risk Officer at Godrej Housing Finance. A chartered accountant with over twenty years with retail risk specialisation in underwriting, policy and portfolio management, Shalinee’s career spans MNCs such as Bank of America, ABN Amro Bank & Barclays, along with private sector banks such as HDFC Bank and DCB, and NBFCs like Fullerton India, and now with Godrej Housing Finance. In her career, she has played key roles in launching and implementing risk management in the retail sector. While the post-pandemic chaos has resulted in innovations and accelerated digital transformations, the associated risks are also a cause of concern. The organisations continue to grapple with the unprecedented changes brought about by the constant lockdowns and interruptions in economic activities. In this light, Analytics India Magazine spoke to Shalinee in order to understand the evolving role of chief risk officers and risk management in the post-COVID world and the risk and controls in Godrej Housing Finance to ensure a smooth process and functioning. Prospering in the Pandemic Talking about the company and its establishment during the pandemic, Shalinee said, “As a new entrant, our focus has always been on a digital-first approach to enable homeownership, and since we launched our business during the pandemic, use of digital technology was a ‘must-have’ and not a ‘good-to-have’ option for us.” Godrej Housing Finance spent considerable time evaluating all the technology solutions available in the market to give a minimal or contactless experience to their potential customers. The company convened interdisciplinary teams with sufficient diversity to brainstorm and design the automated workflows followed by a robust execution strategy. “We continue to remain focused on analysing the evolving consumer needs and related behaviour patterns as we look to build a long-term sustainable home financing business,” added Shalinee. The company firmly positioned itself in the market with Video KYC and automated fraud detection in various stages of the data analytics and customer journey processes — from quality control of data to identifying incorrect data during the data capture process. “All of these are needed to eventually translate into quicker turnaround times, lower costs and sharper insights for clients,” said Shalinee. “We are looking for holistic solutions to incorporate any eventualities that may occur.” “One of the biggest challenges the company has faced was to develop solutions with very limited or no data. To address that, we are following an agile approach enabling a technological and cultural change that drives enhanced collaboration and automation,” said Shalinee. “We are also investing in data structures, with sufficient layers of protection for building a successful analytics organisation enabling reduced time to insights and making decisions,” Shalinee continued. “Furthermore, as a startup, our challenges did not pertain only to digital transformation but rather around enhancement and integration with related leverage technology that could provide in business functions like, onboarding employees virtually, training them for a seamless adoption of digital solutions, enabling work from home and improving consumer experience without compromising on portfolio quality and safety of our employees,” added Shalinee. Risk Management in Action As a digital-first organisation, the culture at Godrej Housing Finance is envisioned with their customers at the core. “At GHF, we encourage people who come with new and diverse ideas and welcome those who are quick learners and share our values.” The company truly believes that there is a great talent pool available in the market, with aspiring young data analysts, scientists, and engineers willing to embrace technological changes and hence have a considerable role in the company’s growth journey. “We believe in giving opportunities to people who want to grow with us. We are investing in building a best-in-class analytics and technology team as we believe this is a great recipe for building a lending organisation that uses data analytics for making decisions,” said Shalinee. Additionally, credit risk monitoring and management is part and parcel of any lending industry; however, COVID-19 has posed various challenges, which has forced every industry to realign the risk management practices and make themselves more resilient and be future-ready. Shalinee believes there is now an urgency to ensure that risk management programs to be optimised by data, metrics, and technology. Thus, Godrej Housing Finance is critically investing in creating platforms to store, manage, and operationalise all aspects of the program, from risk identification, physical security or cyber resilience to operational response, recovery under crisis management and business continuity. “Risk Management at the enterprise level is our focus. We are investing in creating robust metrics, real-time dashboards,” she said. “In a post-COVID world, the requirement to better understand current events and how they impact our business will play a critical role in assessing risks — as well as opportunity. We are using technology to automate controls and assurance framework.” One needs to consider the critical risk areas: strategy, technology, operations, third-party access to data, regulatory, forensics, cyber resilience, data leakage, and privacy in any digital ecosystem. “Managing risks by putting control measures based on applicable risks in the changing era is critical to an organisation’s sustainability. Therefore, laying out the building blocks of the digital risk strategy is crucial to its success,” added Shalinee. She further explained that an important aspect in defining the controls is considering the nature and level of digitisation in the operations, as most of these aspects are nascent and tightly connected with systems or manual processes. For instance, while customer profiling is important for better customer offerings and experience using digital and analytical solutions, protecting customer data privacy is equally critical. The Need for Risk Management in the Post-COVID World Talking about the changing scenarios, Shalinee believes that the role of a chief risk officer in the post-pandemic environment has assumed multifold significance with the addition of data security and cyber risks. “It has become imperative to realign the erstwhile risk management practices to the new realities of the post-COVID-19 world to make an organisation robust and resilient in the new normal,” added Shalinee. It is believed that the role of risk management is expected to increasingly become that of a business advisor rather than merely a control function. As a result, proactive recalibration of credit rating models to incorporate new risk profiles of impacted businesses, real-time portfolio monitoring, enhanced agility in taking corrective actions proactively, and stress testing of portfolios under various scenarios have become extremely relevant. Further, implementing a robust liquidity management framework by ensuring and putting in place a robust funding plan in case of unforeseen shortfalls has assumed great importance. “A complete revamp of capabilities in managing cyber risks, technology risks, human capital risks, as well as organisational reputational risk besides business continuity management has also become a key, given the uncertainties in the operating environment,” said Shalinee. “Crisis management is the new normal of risk management to combat the ‘black swan’ events as the financing business shifts away from traditional concepts. Data and real-time information-driven analytics are set to be the driving pillars of the new operating model, and the sooner the risk management practices are realigned, the higher will be the resilience capacity of any organisation.” Godrej Housing Finance is currently using machine learning-based fraud detecting solutions by partnering with Bureaus. The company also has plans to leverage machine learning for launching other products in its retail finance portfolio. In addition, the firm is constantly looking at automating its processes to improve its team’s productivity. Case in points — RPA, robotic process automation, is one such tool that GHF uses effectively to automate its repetitive tasks, ensure process accuracy and reduce operational risks. However, this wasn’t always the case. One of the biggest challenges GHF faced was developing analytical models with very limited or no data. Shalinee explained that the company is following an agile approach enabling a technological and cultural change that drives enhanced collaboration and automation. “We are investing in data structures, with sufficient layers of protection for building a successful analytics organisation enabling reduced time to insights and decisions,” said Shalinee. “We are also building a culture where all decisions are backed by data. Currently, the absence of internal data is a challenge. However, we are leveraging data from various external sources for driving decisions.” Data breaches and cyber security risks, along with vendor risk and climate risks, are some of the new emerging threats that are gaining a lot of attention in the post-pandemic era. To withstand, GHF is critically focusing on risk prioritisation and a mitigation plan at an enterprise level. Therefore, frequent testing and monitoring of these risks and the existing risks has become a critical aspect of risk management. “We are striving to create a risk management culture at Godrej Housing Finance,” concluded Shalinee.","excerpt":"Analytics India Magazine spoke to Shalinee Mimani, the Chief Risk Officer at Godrej Housing Finance in order to understand the evolving role of chief risk officers and risk management in the post-COVID world and the risk and controls in Godrej Housing Finance to ensure a smooth process and functioning.","categories":["AI Features"],"tags":["credit risk machine learning","credit scoring","fraud detection","Interviews and Discussions","risk management"],"author_name":"Sejuti Das","publish_date":"2021-10-07T10:00:00","publication_year":"2021","word_count":1460,"keywords":["Go","API","risk management","machine learning","credit risk machine learning","AI","ML","Git","Interviews and Discussions","RAG","credit scoring","analytics","R","fraud detection"],"extracted_tech_keywords":["AI","machine learning","ML","analytics","RAG","fraud detection","R","Go","Git","API"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-godrej-housing-finance-blitzscaled-during-pandemic-using-its-risk-management-strategies\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10165736,"title":"Google DeepMind’s Thinking Models: What to Expect","content":"Google DeepMind has been making significant strides in developing ‘thinking models’—a new class of AI models that can reason, plan, and solve complex problems more effectively than previous models. In a podcast on YouTube channel Google for Developers, Jack Rae, a principal research scientist at Google DeepMind, spoke at length about how Google DeepMind’s thinking models are being built, giving us a glimpse of what’s to come. “The key intuition of what a reasoning model is about is that it is going to try and compose knowledge to a specific scenario that may be novel or unseen,” he told Logan Kilpatrick, senior product manager at Google DeepMind. How Is It Going So Far? Google’s efforts have yielded rapid advancements in the capabilities of their thinking models, with significant improvements in their performance on tasks such as math, coding, and multimodal reasoning. Google DeepMind recently conducted a study that showed how AI can think deeper using a ‘Mind Evolution’ technique. “What we’re seeing is truly a new paradigm,” said Rae. “We’re finding multiple avenues of being able to spend more compute on inference time, like, during the response.” The company is already seeing the fruits of its labour. It has released two experimental versions of the Gemini Flash Thinking model, which are available for free on AI Studio. Addressing the product launches, Rae highlighted that the thinking models will see the usage of more tools from within Gemini in the near future. “The model is going to be using more and more tools during thinking in order to really get to the core essence of the problem that it needs to solve,” Rae predicted. He also provided examples of integration of the model with Google Search and Maps. Rae thinks the models will become more agentic because of that, and that will be an important aspect even when it is thinking. Rae also believes that the industry does not need any research breakthroughs to achieve the possibility of a model having “infinite context”. He emphasises that the right ingredients are available, we only need engineering to make it a reality. Feedback in the Loop of its Development Rae told Kilpatrick that he was excited about the model being launched in an experimental phase as the user feedback would help them learn more about the thinking model’s capabilities. To give an example of how feedback is helping shape up the development, Rae recalled a time when he did not realise that 32k context support would be limited to people, until he reached out to the academics that were using the reasoning model as part of their research. Similarly, he shared another instance where an internal code change was required when a user tried to switch from Gemini Flash to Flash Thinking models, and then he worked on fixing the same. Speaking about the timeline for the development of the thinking models, he said that they started working on it in October 2024 and were ready to ship the model in two years for developer feedback. With the feedback received over the holiday period in December end, they released an update to the model in January 2025. While Rae did not officially mention it, keen observers on the internet have speculated that new Gemini models based on non-experimental thinking models should be released on March 12. The Future of Google DeepMind’s Thinking Models “We’re looking forward to a bunch of very exciting future releases,” Rae teased. The company is actively gathering feedback from developers and working towards a general availability (GA) release of the model. “It’s become clear that people want to build on this model and have it as a stable foundation,” Rae acknowledged. “And GA is just essential for that. So that’s something on the roadmap for sure.” Google DeepMind plans to continue improving its capabilities, exploring new product experiences, and enabling them to use tools like code execution and search during the thinking process. Thinking models are also expected to play a crucial role in the development of AI agents, which can interact with the world and perform tasks autonomously. “There are two things that I think are very important for useful agentic capability that reasoning will give. One is reliability… the other is complex capability.” As Google DeepMind continues to push the boundaries of AI, thinking models are likely to become the cornerstone of future AI systems, enabling them to solve increasingly complex problems and interact with the world in more meaningful ways.","excerpt":"Google DeepMind’s principal research scientist sheds light on the development of thinking models and what he thinks about them.","categories":["AI Features"],"tags":["DeepMind"],"author_name":"Ankush Das","publish_date":"2025-03-10T15:00:00","publication_year":"2025","word_count":744,"keywords":["Go","API","programming_languages:R","AI","Modal","programming_languages:Go","AI agents","llm_models:Gemini","GAN","R","DeepMind"],"extracted_tech_keywords":["AI","R","Go","API","GAN","Modal","AI agents","llm_models:Gemini","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/google-deepminds-thinking-models-what-to-expect\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":19027,"title":"Computer Trained To Recognise &#8216;Healthiest&#8217; Person, Can Accurately Predict BMI, Body Fat, Blood Pressure By Scanning A Face","content":"Based on a study by author and Macquarie University psychology researcher Dr Ian Stephen, a computer model has now been trained to recognise a prospective “attractive” person. Attraction between two people may not just be the result of toned abs, the shape of the face, or curves. Dr Stephen’s finding add to the increasingly popular theory which suggests that modern-day attraction has less to do with personal taste and more to do with our biological instincts. Based on Dr Stephen’s research, the computer model was programmed to recognise the “healthiest” face on the basis of the three major indicators of good health: body mass index (BMI), body fat percentage and blood pressure. The computer was reportedly trained to recognise these three health indicators for 272 Asian, African and Caucasian faces. Once trained, the model was then able to accurately predict the BMI, body fat and blood pressure of any individual, just by analysing an image of their face. “There’s always been this ‘beauty is in the eye of the beholder’ idea which essentially says that what we find attractive is just arbitrary and there is nothing inherent about an attractive person… Evolutionary theory has chipped in on that and essentially says that attractiveness is essentially a mechanism for recognising healthy, fertile, appropriate mates and also healthy friends and allies, because there’s obvious advantages to being friends with or mating with people who are healthy,” Dr Stephen said. Dr Stephen was also part of a team who found that men who eat a diet rich in fruit and vegetables smell more attractive to women than those who eat high-carbohydrate or high-fat diets.","excerpt":"Based on a study by author and Macquarie University psychology researcher Dr Ian Stephen, a computer model has now been trained to recognise a prospective “attractive” person. Attraction between two people may not just be the result of toned abs, the shape of the face, or curves. Dr Stephen’s finding add to the increasingly popular […]","categories":["AI News"],"tags":[],"author_name":"Prajakta Hebbar","publish_date":"2017-11-15T09:03:01","publication_year":"2017","word_count":271,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","R"],"extracted_tech_keywords":["AI","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/computer-trained-recognise-healthiest-person\/","complexity_score":2,"technical_depth":5,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":69299,"title":"Shiv Nadar University Launches An Online Program For Data Science","content":"Today, Shiv Nadar University, a multidisciplinary and research-based university, launched a new online program in Data Sciences and Analytics for Business (DSAB). The 14-week program will start from the 3rd week of July. For this, the university has collaborated with Jigsaw Academy, a data science institute that offers analytics courses, to deliver a few lessons. Also recognised as an institution of eminence (IoE) by the Government of India, the university is well poised to offer the course and help learners advance their career. Aspirants who are enrolled in any undergraduate degree program, along with mathematics in their high school are eligible to enrol for the program. However, aspirants will have a personal interview as a final round of selection. On completion of this course, learners would be well versed in various machine learning and deep learning techniques, along with basics such as statistics, probability, SQL, Python, and more. In addition, there will be a capstone project that will allow students to get practical experience while working with real-world data. “The demand for skilled data science talent across sectors is rapidly growing. However, not enough well-trained and skilled professionals are entering the industry every year to help bridge this excess-demand. Post-COVID, the digital economy is expected to grow, and with that, the demand for analysts will also see increasing trends. The Data Sciences and Analytics for Business program resonates with Shiv Nadar University’s mission of developing and educating the path-shapers of tomorrow with relevant industry-ready skills. This program is specifically designed to drive specific learning outcomes that give our students a clear edge as they drive ahead to the future,” said Dr Bibek Banerjee, Senior Dean and University Head of Strategic Initiatives, and the Director of School of Management and Entrepreneurship at Shiv Nadar University. Click here to know more.","excerpt":"Today, Shiv Nadar University, a multidisciplinary and research-based university, launched a new online program in Data Sciences and Analytics for Business (DSAB). The 14-week program will start from the 3rd week of July. For this, the university has collaborated with Jigsaw Academy, a data science institute that offers analytics courses, to deliver a few lessons. […]","categories":["AI News"],"tags":["Deep Learning Techniques","Shiv nadar university data science"],"author_name":"Rohit Yadav","publish_date":"2020-07-08T20:00:00","publication_year":"2020","word_count":299,"keywords":["data science","Go","machine learning","AI","Git","Python","deep learning","analytics","Deep Learning Techniques","SQL","Shiv nadar university data science","R"],"extracted_tech_keywords":["AI","machine learning","deep learning","data science","analytics","Python","R","SQL","Go","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/shiv-nadar-university-launches-an-online-program-for-data-science\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10124401,"title":"How Does NetApp Use Half of the World’s Data in AI?","content":"As industry stalwarts like Matei Zaharia indicate that “the next promise in AI is in domain-specific enterprise use cases,” NetApp, a data management and storage provider, finds itself well-positioned to capitalise on this trend. At the recent NVIDIA GTC event, CEO Jensen Huang highlighted NetApp’s role, stating “nearly half of the world’s files are stored on-premises on NetApp’s platform”. With 76% of global tech companies and 70% of Indian enterprises having AI initiatives underway, outpacing the 49% global average, NetApp’s storage footprint could prove invaluable. Notably, 91% of Indian companies plan to leverage over half their data to train AI models by 2024. NetApp is banking on this pace in the Indian market. However, its leadership isn’t surprised. “India has always been strong on the tech side,” said Puneet Gupta, NetApp MD for India and SAARC, in an exclusive interaction with AIM. He pointed out that India’s vast datasets, such as those generated by initiatives like UPI and Aadhaar, are crucial assets. He also said that India has the ingredients for AI success. India, NetApp’s fourth-largest APAC market, has the potential to become the largest in the next few years, according to Gupta. (Source: Polygon.io) Globally, with a $26.48 billion market cap and Q4 FY2024 revenue of $1.668 billion, up 5.1% year-over-year, NetApp is poised for growth driven by all-flash arrays, cloud services, partnerships, and enterprise AI adoption, if executed well. Partnering With NVIDIA to Be Enterprise’s Go-to Choice NetApp has partnered with NVIDIA to advance RAG for generative AI applications in enterprise. It allows integration of NVIDIA’s NeMo Retriever with NetApp’s ONTAP storage, enabling LLMs to securely access the vast amounts of enterprise’s data stored on NetApp without compromising privacy and security. Enterprises can leverage their existing data assets on NetApp to “talk to their data” through simple prompts and gain insights for generative AI, without the need to move data. “The work that NVIDIA is doing in the AI space is remarkable,” Gupta said. “While they focus on building compute and server farms, we provide high-performance storage to make the whole solution work effectively,” he added Moreover, the partnership complements NetApp’s existing AI services, which have been used by over 500 joint customers for AI model training and inference. NetApp has also worked with NVIDIA to update its FlexPod AI converged infrastructure to support NVIDIA AI Enterprise software and was one of the first partners to complete storage validation for NVIDIA OVX systems. This collaboration is significant because it allows enterprises to safely use their proprietary data within LLMs without the risk of data leaks or privacy concerns. It reduces friction, cost, and time to value for RAG by enabling access to data wherever it is stored. Gupta also elaborated on the partnership with Cisco, particularly in the Indian context. “We have a strong partnership with Cisco globally and in India. They work with NVIDIA as well, and we fit into the ecosystem by providing the necessary high-performing storage to complete the AI story,” Gupta explained. Unified Data Management Another critical challenge in data management has been the proliferation of disparate storage systems for different data types, such as blocks, files, and objects. NetApp addresses this issue by unifying all data services on a single platform, simplifying storage infrastructure and enhancing efficiency. Shuja Mirza, NetApp’s India\/SAARC director of solutions engineering, explained, “NetApp is doing its bit by ensuring that you can run all these data services from a single platform. So unifying all of it—objects, files, blocks, structured, unstructured—you know, the ability to store data and data lakes, put it to good use through modern workloads like Spark, Hadoop, etc.” This consolidation supports modern workloads like Spark and Hadoop, which are vital for handling large-scale unstructured data. Mirza emphasised NetApp’s commitment to providing customers with choice and flexibility, stating, “Our job as a technology provider is to make sure that we provide you with that choice. And, whichever format you’re getting data in, we’ll help you store and manage it efficiently, and make sure it is available to you all the time.” Features like NetApp FlexClones allow data scientists to instantly create writable copies of datasets for experimentation without consuming additional storage. NetApp also enables data governance frameworks to manage privacy, compliance, and access controls, which is vital when dealing with sensitive data used for training LLMs. More Efficient Data Centers With India’s data centre industry booming,  to accommodate AI needs, including investments in GPUs and larger racks, Gupta sees significant opportunities for NetApp. “The infrastructure needed to support the generative AI wave includes both large-scale cloud provider setups and enterprise-specific builds. NetApp aims to participate in both segments, leveraging our expertise to manage data efficiently across various platforms,” he stated. NetApp could also provide storage efficiencies and a secure platform, including ransomware recovery guarantees. “AI, especially generative AI, is all about having large datasets and creating copies of these data sets,” Mirza said. Adding that, “NetApp’s solutions help reduce these copies through snapshots, thus providing storage efficiencies and a secure platform.” Mirza elaborated, “Typically, in these projects, the data scientists create environments. Each environment has got its own copy. So the volumes are large. With NetApp FlexClones it becomes easy and efficient by storing those copies through snapshots.” This approach not only ensures efficiency but also secures data management, which is critical in AI projects. “Storing data is the tip of the iceberg. But if we were not to really do a good job of managing and securing it, then there is always a chance of breach and influence,” he asserted. And even in case of a breach NetApp provides a 99.999999 guarantee with a recovery time of a few seconds.","excerpt":"With 76% global & 70% Indian enterprises having AI initiatives underway, NetApp’s storage and management footprint could prove invaluable.","categories":["AI Features"],"tags":["AI data","netapp"],"author_name":"Shyam Nandan Upadhyay","publish_date":"2024-06-24T18:00:00","publication_year":"2024","word_count":944,"keywords":["Go","API","AI","R","netapp","RAG","Ray","Aim","generative AI","data governance","AI data","data lake"],"extracted_tech_keywords":["AI","generative AI","Aim","Ray","RAG","R","Go","API","data lake","data governance"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-does-netapp-use-half-of-the-worlds-data-in-ai\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":69195,"title":"Brillio Acquires Cognetik To Expand Analytics","content":"In undisclosed terms, Brillio, a digital technology consulting and solutions company, today announced the acquisition of Cognetik, a data and insights company with deep expertise in improving digital experiences for its customers. Cognetik enables companies across the globe, including Facebook, Pizza Hut and McDonald’s to build and implement analytics solutions that optimise customer experience to increase loyalty, drive revenue and advance business transformation. The acquisition of Cognetik strengthens Brillio’s strategy to further extend its services in the world. The deal comes at an opportune time, when more than 75% of CEOs are looking to accelerate their company’s digital transformation, according to a new Fortune 500 survey. “We’re thrilled to welcome Cognetik to the Brillio family,” said Raj Mamodia, Founder and CEO of Brillio. “This marks a significant milestone in Brillio’s growth journey, which continues its forward momentum even in a challenging market. The industry depth and experience Cognetik brings in advanced analytics, particularly in the areas of customer experience personalisation and web analytics, along with Brillio’s existing strength in data science and data engineering, will ensure our customers leverage insights from data and become more relevant than ever,” added Mamodia. Cognetik’s expertise further enhances Brillio’s analytics business, which has already scaled rapidly and has been growing at a CAGR of 30% over the last three years. Cognetik’s capabilities in customer and marketing analytics, customer experience, multi-channel experience optimisation and experience within the Adobe ecosystem strengthens Brillio’s ability to provide clients with industry solutions for personalisation and omnichannel experiences. “Cognetik is very happy to bring its extremely talented team and customer analytics solutions to Brillio. We are excited about building the greatest analytics capabilities in the world with Brillio for our clients,” said Daniel Herdean, CEO and co-founder of Cognetik. “Our organisations share the same ‘fanaticism’ around customer success, and we can’t wait to roll up our sleeves and get to work together to deliver on it.” Cognetik’s delivery presence in Romania will expand Brillio’s global delivery capabilities and enhance the company’s abilities to onboard best-in-class talent in the EU continent and service clients in development centres outside of India.","excerpt":"In undisclosed terms, Brillio, a digital technology consulting and solutions company, today announced the acquisition of Cognetik, a data and insights company with deep expertise in improving digital experiences for its customers. Cognetik enables companies across the globe, including Facebook, Pizza Hut and McDonald’s to build and implement analytics solutions that optimise customer experience to […]","categories":["AI News"],"tags":[],"author_name":"Rohit Yadav","publish_date":"2020-07-07T18:32:07","publication_year":"2020","word_count":349,"keywords":["data science","API","AI","digital transformation","Git","RAG","data engineering","analytics","GAN","R"],"extracted_tech_keywords":["AI","data science","analytics","RAG","R","Git","API","data engineering","GAN","digital transformation"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/brillio-acquires-cognetik-to-expand-analytics\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10717,"title":"Microsoft uses Machine Learning and Data Analytics to empower Indian Farmers","content":"Today, the dynamics of agriculture are changing. There is movement from traditional farming methods to smart farming. And Microsoft’s recent initiative is to help Indian farmers take advantage of this technology to grow more with the help of machine learning and data analytics. Software Giant, Microsoft has collaborated with International Crops Research Institute for Semi-Arid Tropics (ICRISAT) and Andhra Pradesh government, to develop a new mobile application for farmers. This new mobile application will advise the Indian farmers on the best time to sow crops taking into consideration the weather conditions, soil and other indicators important for having a good crop. Along with this, the mobile app will also act as a personalized village advisory dashboard for the state of Andhra Pradesh. The idea behind developing the Sowing App and providing Personalized Village Advisory Dashboard is to provide powerful cloud-based predictive analytics so that farmer feel empowered with information and insights that are required to take farming decisions. This app will help farmers to reduce crop failures and increase yield. This will lead to better income opportunity for the farmers. The Sowing app has been made very user friendly by using graphics and giving it a colorful look so it can be more interactive and useful for farmers to understand. This application and service makes use of Cortana Intelligence Suite, Machine learning, Data Analytics & Power BI. This apps will use weather forecasting models provided by US- based aWhere Inc. A pilot for this app has been recently launched. Under this pilot, farmers will be provided with information about the sowing date through SMS in their native language Telugu. Data Collection for this app involved collecting data manually by ICRISAT from farms in 13 districts of Andhra Pradesh. This data has been uploaded to Microsoft’s Azure Cloud and by applying powerful Business Intelligence (BI) tools, the app and dashboard will be able to provide important information about the soil condition, fertilizer recommendations, and weather forecast for 7 days. Also the app and dashboard are loaded with in-depth data about rainfall of over the last 45 years and 10 years of groundnut sowing progress data for Kurnool district of AP. All this information and insights will help the farmers take correct and informative decisions to reduce their crop failures and improve yield. Microsoft is using Machine Learning and analytics for other projects too. These include developing a machine learning based model which can be used to analyze and predict drop-outs so that preventive actions can be taken to boost education system. Also Microsoft is building a predictive model which will help to predict regression rates for eye operations, enabling doctors to identify the procedures needed to prevent and treat visual impairments.","excerpt":"Today, the dynamics of agriculture are changing. There is movement from traditional farming methods to smart farming. And Microsoft’s recent initiative is to help Indian farmers take advantage of this technology to grow more with the help of machine learning and data analytics. Software Giant, Microsoft has collaborated with International Crops Research Institute for Semi-Arid […]","categories":["Global Tech"],"tags":["Data Analytics","Machine Learning","Microsoft","smart farming"],"author_name":"Manisha Salecha","publish_date":"2016-09-07T06:10:22","publication_year":"2016","word_count":450,"keywords":["business intelligence","Go","machine learning","cloud_platforms:Azure","programming_languages:R","AI","Azure","R","Machine Learning","analytics","Data Analytics","smart farming","predictive analytics","Microsoft"],"extracted_tech_keywords":["AI","machine learning","analytics","predictive analytics","Azure","R","Go","business intelligence","cloud_platforms:Azure","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/microsoft-uses-machine-learning-data-analytics-empower-indian-farmers\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10119483,"title":"AWS Collaborates With ShellKode to Train 1 Lakh Women Developers in GenAI","content":"AWS has partnered with Bengaluru-based born-in-the-cloud company ShellKode to train one lakh women developers in GenAI. The two companies have collaborated in an effort to help women developers upskill themselves during the current shift towards AI in the workplace. As a part of this, ShellKode launched “EmpowerHer”, which will work towards upskilling the developers. “We’re empowering a generation of aspiring developers, particularly women, with the cutting-edge tools and knowledge of AI to transform India’s innovation landscape and shape the future of enterprise,” said Arun Kumar, ShellKode CEO. The programme will pair the developers with GenAI mentors in order to help them better understand how GenAI works. In addition, the company stated that the mentors will be able to provide “personalised and invaluable guidance, career advice, and a supportive network.” This will be done through networking events, seminars and online interactions with the GenAI community, where they will be able to connect with industry professionals. The need to democratise the AI landscape in India has been a continuing conversation. Recently, several companies have pledged to increase the number of women developers in the country, including Microsoft. The tech giant had pledged to upskill and certify over 75,000 women developers in AI by 2025. Likewise, the Karnataka government had also collaborated with JobsForHer in March this year towards a similar goal. The Karnataka Digital Economic Mission (KDEM) had launched the HerShakti, specifically for women in tech. As part of the initiative, they hope to upskill 500 women in the next six months. Similarly, in an era where mid-career employees are rushing towards upskilling themselves, the gender distribution has been fairly even in demand. Several edtech startups have reported increased interest in AI and AI-adjacent courses, especially among women in Tier 2 and Tier 3 cities. With this in mind, the EmpowerHer programme comes at an opportune time.","excerpt":"The two companies have collaborated in an effort to help women developers upskill themselves during the current shift towards AI in the workplace.","categories":["AI News"],"tags":["AI India","Women in AI","Women in Tech"],"author_name":"Donna Eva","publish_date":"2024-05-02T18:11:43","publication_year":"2024","word_count":305,"keywords":["Go","GenAI","AWS","AI","cloud_platforms:AWS","innovation","programming_languages:R","Git","Women in Tech","AI India","R","Women in AI","startup"],"extracted_tech_keywords":["AI","GenAI","AWS","R","Go","Git","innovation","startup","cloud_platforms:AWS","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/aws-collaborates-with-shellkode-to-train-1-lakh-women-developers-in-genai\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10054926,"title":"A Beginner’s Guide to Self-Supervised Classification","content":"As of now, the self supervised learning procedure can be considered as making the data meaningful for the models that can perform the required task on the data. We can make the data structure more specific to the task to obtain more accurate results. In this article, we are going to discuss self-supervised classification where for the classification as a specific task, our need is to make the data more appropriate for the classification. The major points to be discussed in the article are listed below. Table of Contents What is Self-Supervised Classification?Mathematics Behind the Self-Supervised ClassificationSelf-Classifier – A Self-Supervised Classification Network What is Self-Supervised classification? If the classifications on the data are performed by the representation and label learned using self-supervised learning, it can be considered as the self-supervised classification. More than this definition, we can say that it is the process of making data for more accurate classification using the self-supervised learning process. Let’s discuss it in more detail while starting with self-supervised learning. As we know that, using self-supervised learning, we can make a machine to learn from the unlabeled data and also we can say that it is a somewhere intermediate form between supervised and unsupervised learning. Let’s take an example of natural language processing where if self-supervised learning is used, we can make machines learn a meaningful representation of the data without any human-provided label. More formally, we can say that the basic self-supervised learning methods are focused to learn the representation of the data and after learning the representation we can transfer them to downstream tasks like sentiment analysis and image classification. Basically, this downstream task is a predefined and simple one. When we talk about self-supervised classification, we can say that this process can be capable of performing various downstream tasks. Let’s say the learning model has learned that there are C classes in the data and now the task is to make a classifier that can classify the two augmentations of the data similarly. By this intuition, we can say that these classifiers are provided with a task that is prone to degenerate solutions where the samples of the data are related to the same class. The trivial solution can be avoided by editing the loss functions such as prior on the standard cross-entropy loss function. As we have discussed that self-supervised learning methods are a combination of the supervised and unsupervised learning methods, similarly the self-supervised classification can be considered as the combination of deep unsupervised clustering and contrastive learning. Here, the parameters of the assigned neural network and class assignment to the samples of the data can be learned simultaneously. Image source The above image is a representation of a self-supervised classifier structure where a pair of shared networks are processing the augmented version of a similar image. Where we can see that the cross-entropy for every view is minimized so that the model can promote the same class prediction and assertion of the prior can avoid the degenerated solution. Using the same architecture we can learn the representation of the data and class labels in a single-stage end-to-end unsupervised manner. More formally we can say that this method is a combination of contrastive learning and deep unsupervised learning. Where contrastive learning is an approach of supervised learning and deep unsupervised learning is part of the unsupervised learning approach and in the self-supervised classification, it can help for learning the parameters of neural networks and also it helps in learning the cluster assignment of the final features using the unlabelled data. To understand it more, we are required to discuss the mathematics and logic behind the self-supervised classification. Mathematics Behind the Self-Supervised classification In the last image where the architecture of the self-supervised classification is explained, we have seen that we have two augmented views of the image as x1 and x2 and the goal of the method is to learn a classifier yi , f(xi) ∈ [C]. We can consider that C is the number of classes and f(.) is the function which the classifier follows and the downstream aim is to classify views similarly if the views are the augmented version of each other and also make the classifier to be capable of ignoring the degenerate solutions. A common way to learn using a multiclass classifier is by minimizing the loss function and if the function as loss function is cross-entropy loss function then the minimization can be done by Here, the row softmax is p(y|x), and if we are trying to minimize the above-given function without additional regularization, we can have degenerate solutions. To avoid the degenerated solutions, we can introduce the Bayes theorem on the loss function as: Dropping the denominator and controlling the prior p(y) can help in the regularization of the solution. By asserting a uniform prior on y we can use all the classes of the data. The resulting loss will be: In the above-given formula, the p(x|y) is a column softmax whose symmetrized version will be: Where the prior is very important because without a uniform prior the solution will provide a very nonuniform p(y). All the samples from the data will get assigned to a single class. So in the above section, we have seen what procedure can be called as a self-supervised classification and what can be the mathematics and logic behind the self-supervised classification. In the next section, we are going to have an overview of a package self-classifier that allows us to use the models based on self-supervised classification. Self-Classifier – A Self-Supervised Classification Network Self-classifier is a self-supervised classification neural network that helps in learning the representation of the data and labels of the data simultaneously in one procedure and also in an end-to-end manner. The python implementation of the Self-Classifier’s pre-trained model can be found in the link. In the architecture of the package we can see that using the convolutional layer and fully connected layers the models under the package can learn representation and class labels. In this section of the article, we are going to discuss the comparison of the self-classifier with other methods, models and packages so that we will know the importance of self-supervised classification. Elad Amrani, IBM Research AI, Technion, and the team have developed the package. The developers of the package say that the self-classifier is simple to implement and details about how to implement this package can be found in the link. Unlike other state-of-the-art methods, it does not require large memory or any other second momentum neural networks. Clustering and classification tasks can be performed by a single package. A comparison between some of the famous state of the art models and Self-classifier following the architecture of ResNet-50 for image classification is listed below., The example we have taken is mainly useful for the image classification task with a very light framework enabled and also very easy to use. The architecture we have seen is just to make the data representation meaningful and classes labelled so that it can be classified easily and accurately. Final Words Here in the article, we could understand how we can define the self-supervised classification from the self-supervised learning domain of machine learning. We have also seen a proposed way of logic and mathematics that can play a crucial role behind the self-supervised classification models. Along with that, we have gone through an example of it named Self-Classifier with its comparison to the other traditional methods.","excerpt":"in the self supervised learning process we are mainly focused about making the data workable to the downstream algorithms. but when using the self-supervised learning we make the data specifically for classification we can say the process is self-supervised classification.","categories":["Deep Tech"],"tags":["AI (Artificial Intelligence)","Data Science","Deep Learning","Guide","Machine Learning","self learning","self supervised learning"],"author_name":"Yugesh Verma","publish_date":"2021-12-07T12:00:00","publication_year":"2021","word_count":1243,"keywords":["Go","self learning","self-supervised learning","machine learning","AI","AI (Artificial Intelligence)","neural network","sentiment analysis","Machine Learning","Python","Aim","ResNet","Deep Learning","Data Science","R","Guide","self supervised learning"],"extracted_tech_keywords":["AI","machine learning","neural network","Aim","sentiment analysis","Python","R","Go","ResNet","self-supervised learning"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/a-beginners-guide-to-self-supervised-classification\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10006195,"title":"Embracing AI – Key to Futuristic Org Strategy","content":"In the 2019 MIT Sloan Management Review and Boston Consulting Group (BCG) Artificial Intelligence Global Executive Study and Research Report, 9 out of 10 respondents agree that AI represents a business opportunity for their company. While at the same time when they were asked: “What if competitors, particularly unencumbered new entrants, figure out AI before we do?” In 2019, 45% perceived some risk from AI, up from an already substantial 37% in 2017. More and more leaders are viewing AI as a risk if they are behind in adoption. Financial Institutions have been at the forefront of AI adoption. But within financial institutions, we are seeing varying trends. Deloitte surveyed 1,100 executives from US-based companies across different industries that are prototyping or implementing AI. Out of which 206 respondents were working for financial services. On the basis of the responses received the Organizations were categorized in 3 segments Frontrunners Followers and Starters. , The traits observed of the Frontrunners were: Embed AI in strategic plans: Integrating AI into an organization’s strategic objectives has helped many frontrunners develop an enterprise wide strategy for AI, which different business segments can follow. The greater strategic importance accorded to AI is also leading to a higher level of investment by these leaders. Apply AI to revenue and customer engagement opportunities: Most frontrunners have started exploring the use of AI for various revenue enhancements and client experience initiatives Utilize multiple options for acquiring AI: Frontrunners seem open to employing multiple approaches for acquiring and developing AI applications. This strategy is helping them accelerate the adoption of AI initiatives via access to a wider pool of talent and technology solutions. The Frontrunners across the globe are leaving no stone unturned in adopting and harnessing AI in their businesses processes. With more than Million Fintechs registering every year reimaging financial processes by using AI, Financial Institutions are at the big risk of losing to the competition. Banking in the current world is very different from banking of last century. Adoption of AI in financial Institutions was led by the use of Analytical models in Risk and fraud management but today AI\/Ml models have become the core to each process ranging from Sales to Operations to Marketing. The Big Data space is set to reach over $273 Billion by 2023 which is the clear indication of the faith that Organizations are putting in data and Analytics. In last 2 decades analytics has evolved from low key data visualization dashboards to full blown machine learning interactive models. We have moved from descriptive to predictive to now prescriptive analytics. Organizations doesn’t want to just stop at the prediction they look forward to solutions prescribing them on what measures needs to be taken. Though we cannot completely do away with the manual interventions but adding more and more insights helps in making better decisions. Some of the areas in which we have witnessed the remarkable progress are Risk Scoring in Banking: In early 2000s, we were talking about visualizing the Non-Performing Loans (NPL) across the branches through dashboards and reports. By 2014 we had solutions which could predict the Non-Performing Loans basis the customer attributes, then next few years went in training those models to predict accurately. Today we are talking about using ML to not only predict the NPLs but also to prescribe the major reasons for Non-Performing Loans. The critical role which these AI\/ML based early warning solutions are playing in credit decisions is noteworthy. Similarly the Manufacturing, Oil & gas and Energy sector are banking on Industry 4.0 tools. Oil refineries are saving millions by using predictive maintenance and Anomaly detection solutions. These solutions not only help organizations in optimizing maintenance schedules but also increase the residual life of equipment. As per the Mckinsey report “Predictive maintenance typically reduces machine downtime by 30 to 50 percent and increases machine life by 20 to 40 percent” Every 10 percent of the saving on machine downtime is equivalent to Millions in revenue for the Organizations. Hence 30 t0 50% would be hundreds of million dollars over the year. As we are moving forward organizations not only want the solutions to predict when the maintenance will be due but also to prescribe and fix the maintenance schedules basis data from different equipments in the plant. Solutions are also offering the inbuilt feature to trigger auto email to the maintenance engineers for fixing appointment. Though we are embarking new heights and unleashing new peaks with the help of AI but to benefit from these solutions, organizations need to reimagine their Organization Strategy in the light of AI and embrace the change which comes enroute in this journey. Reference: https:\/\/www2.deloitte.com\/us\/en\/insights\/industry\/financial-services\/artificial-intelligence-ai-financial-services-frontrunners.html","excerpt":"In the 2019 MIT Sloan Management Review and Boston Consulting Group (BCG) Artificial Intelligence Global Executive Study and Research Report, 9 out of 10 respondents agree that AI represents a business opportunity for their company. While at the same time when they were asked: “What if competitors, particularly unencumbered new entrants, figure out AI before […]","categories":["AI Features"],"tags":["ai consulting","intel global strategy"],"author_name":"Devansh Sharma","publish_date":"2020-09-07T16:00:40","publication_year":"2020","word_count":777,"keywords":["big data","Go","ai consulting","machine learning","artificial intelligence","AI","ML","intel global strategy","anomaly detection","analytics","GAN","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","analytics","anomaly detection","R","Go","big data","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/embracing-ai-key-to-futuristic-org-strategy\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10102128,"title":"IBM Launches Gen-AI-powered Watsonx Code Assistant for Enterprise","content":"IBM has officially unveiled its Watsonx Code Assistant, an AI-powered tool designed to assist enterprise developers and IT operators in coding more efficiently. This generative AI assistant employs natural language prompts and is now available to address two key enterprise use cases: IT Automation through Watsonx Code Assistant for Red Hat Ansible Lightspeed and mainframe application modernisation with Watsonx Code Assistant for Z. Watsonx Code Assistant is built on the foundation of IBM’s Granite AI models, running on the IBM Watsonx platform. The tool utilises a decoder architecture for natural language processing, enhancing the accuracy and efficiency of coding processes. Watsonx Code Assistant aligns with IBM’s commitment to accelerate the development process while maintaining the core principles of trust, security, and compliance. According to a recent report from IDC, this AI-powered tool can help improve code quality by promoting best practices in code recommendations, eliminating the need for unvetted repositories. Kareem Yusuf, Senior Vice President of Product Management and Growth at IBM Software, stated, “Watsonx Code Assistant puts AI-assisted code development and application modernization tools directly into the hands of developers to help address skills gaps and increase productivity.” The launch of Watsonx Code Assistant represents another addition to IBM’s suite of Watsonx assistants, including Orchestrate and Assistant, all of which aim to make generative AI accessible and non-disruptive for enterprises. Key data from the technical preview of this tool indicates increased productivity by 20-45%, with an overall average acceptance rate of 85% for AI-generated content recommendations. Ashesh Badani, Chief Product Officer at Red Hat, expressed that Watsonx Code Assistant for Red Hat Ansible Lightspeed could “close skills gaps, create greater organizational efficiencies, and free enterprise IT to deliver even more business value.” Watsonx Code Assistant for Z focuses on the modernization of mainframe applications by facilitating the translation of COBOL to Java on IBM Z. It aims to accelerate the modernization of mainframe applications while leveraging the performance, security, and resiliency capabilities of IBM Z. T IBM’s long-term partnership with TCS has played a pivotal role in the development and deployment of this tool. Together, they have created a full-service practice for in-place application modernization. With the potential for developer productivity gains, the tool is expected to find applications on the mainframe.","excerpt":"Technical preview indicates increased productivity by 20-45%, with an impressive overall average acceptance rate of 85%.","categories":["AI News"],"tags":["IBM"],"author_name":"Shyam Nandan Upadhyay","publish_date":"2023-10-27T12:24:44","publication_year":"2023","word_count":373,"keywords":["AI","RAG","automation","Aim","ViT","generative AI","IBM","Rust","GAN","R","Java"],"extracted_tech_keywords":["AI","generative AI","Aim","RAG","R","Rust","Java","GAN","ViT","automation"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/ibm-launches-gen-ai-powered-watsonx-code-assistant-for-enterprise\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10166055,"title":"If a Tariff War is What the US Wants, We’re Ready to Fight: China","content":"The latest wave of US sanctions on chip exports has put China in the spotlight, forcing the country’s tech sector to adapt quickly. Chinese tech companies are actively investing in homegrown technologies in an attempt to mitigate the impact of US sanctions, particularly in the semiconductor sector. This reflects a broader strategy of self-reliance, driven by restrictions on access to advanced US technology. While official statements and industry developments indicate notable progress, significant challenges still lie ahead. The Chinese embassy in the US said in a post on X that US President Donald Trump’s imposition of new tariffs on Chinese imports will not affect China. “If war is what the US wants, be it a tariff war, a trade war or any other type of war, we’re ready to fight till the end.” China Pushes Forward with New Innovations With a strong local demand and supply chain, China has recently made a notable breakthrough in the semiconductor industry. The latest development is China’s latest silicon-free chip, which uses bismuth-based materials to bypass silicon-based restrictions. According to South China Morning Post, one key development of this 2D transistor from Peking University is that it reportedly operates 40% faster and uses 10% less energy than leading 3-nanometer silicon chips from Intel and Taiwan Semiconductor Manufacturing Company (TSMC). Professor Hailin Peng, who is leading the project, described this as “changing lanes” in the semiconductor race, born out of necessity due to sanctions, forcing researchers to find fresh solutions. On the other hand, C4D Lab, University of Nairobi’s R&D and innovation hub, expressed skepticism as it stated, “If true, silicon might need to start job hunting, do you think?” This leaves room for further debate as the chip’s real-world impact remains to be seen. Moreover, companies like Huawei and Semiconductor Manufacturing International Corporation (SMIC) are making breakthroughs and challenging NVIDIA despite sanctions. Huawei is pushing forward with its Ascend 910C AI accelerator, which Lennart Heim, researcher at RAND, describes as China’s most advanced AI chip to date. According to Heim’s thread on X, the 910C is now entering production, largely through illicit procurement of advanced dies from TSMC, despite stringent US export controls. This backdoor sourcing could enable Huawei to produce up to 1 million chips equivalent to NVIDIA’s H100 this year, showcasing China’s determination to circumvent restrictions and maintain its position in the global AI race. The ability to exploit such loopholes highlights the challenges US policymakers face in enforcing export controls as Huawei taps into stockpiles of TSMC’s 7 nanometer technology and high-bandwidth memory (HBM2E) from Samsung, acquired before tighter regulations were implemented in January this year. Heim emphasises that while the 910C’s per-chip performance is underwhelming, China can compensate by deploying larger numbers of accelerators, facilitated by its centralised control over resources. This approach could enable competitive AI models, particularly in areas like reasoning, as China harnesses the talent and compute power it has amassed. US Sanctions and Chinese Response US sanctions on Chinese tech companies, particularly in the semiconductor industry, have intensified in recent years. These measures, such as adding firms to the Entity List and restricting exports, aim to limit China’s access to advanced technologies, especially chips critical for AI, military applications, and supercomputing, as reported by The New York Times. In response, China has launched initiatives to reduce reliance on foreign technology, including the ‘Made in China 2025’ goal of achieving 70% domestic semiconductor production by 2025. China’s strategy includes significant state funding, with billions invested in domestic chipmakers like SMIC and Yangtze Memory Technologies Corp (YMTC). This development is part of a broader trend, with China advancing in RISC-V processors and other non-silicon technologies. For instance, the XiangShan project aims to deliver a high-performance RISC-V processor by 2025, potentially reducing dependence on foreign technology, Slashdot reported. These innovations suggest that sanctions may be fuelling, rather than hindering, China’s tech sector. This claim was supported in 2023 by Xu Zhijun (Eric Xu), the rotating chairman at Chinese tech giant Huawei. He had said that China’s chip industry would be “reborn” as a result of US sanctions. China’s Chip Ecosystem While technological gaps remain between Chinese manufacturers and global leaders, massive government investment, including a $47 billion semiconductor fund launched in May 2024, and strategic focus accelerate development across all segments, gradually reducing dependency on foreign technology. The core of China’s chip ecosystem revolves around major players like SMIC, YMTC, and ChangXin Memory Technologies (CXMT). While Huawei and ByteDance are not directly involved in chip manufacturing, they are crucial in driving demand for advanced semiconductor technologies, particularly in AI infrastructure and mobile applications. SMIC, China’s leading pure-play foundry, achieved the position of world’s second-largest pure-play foundry in early 2024, despite international restrictions. It successfully developed 7 nanometer chips and announced plans for 5 nanometer production, making strides in chip manufacturing. In the critical memory sector, YMTC has established itself as China’s champion in NAND flash memory production. Founded in 2016, YMTC has progressed from basic memory architectures to complex 64-layer 3D NAND structures, marking the country’s entry into this strategic segment. Notably, CXMT serves as the country’s primary Dynamic Random Access Memory (DRAM) manufacturer, working to reduce dependency on foreign suppliers for this essential memory type. Hua Hong Semiconductor complements these efforts by manufacturing analogue, mixed-signal, and specialty semiconductors. The equipment manufacturing sector, vital for true technological independence, includes Shanghai Micro Electronics Equipment (SMEE) and Advanced Micro-Fabrication Equipment Inc (AMEC). At the same time, Naura Technology Group offers semiconductor production equipment, reducing reliance on foreign technology providers. China’s ecosystem also features innovative design houses, including HiSilicon (Huawei’s former chip design subsidiary), Zhaoxin (x86 compatible processors), and Loongson (MIPS-based CPUs), and UNISOC (mobile SoCs). These companies’ strategies are pivotal in countering US sanctions by reducing reliance on foreign technology. The country is also a global leader in semiconductor packaging, holding over a quarter of the global market share. This allows it to maintain influence in the supply chain despite US restrictions. As the global tech race intensifies, China’s push for semiconductor self-sufficiency could disrupt global supply chains and redefine tech leadership. The impact of US sanctions has accelerated not only China’s domestic innovation but also that of various other countries. This could likely hint towards an era of protectionism for the global chip market.","excerpt":"Chinese semiconductor companies are actively investing in homegrown tech to mitigate the impact of US sanctions.","categories":["Deep Tech"],"tags":["China","china chip technology","Chip War","semiconductor industry"],"author_name":"Sanjana Gupta","publish_date":"2025-03-14T10:00:00","publication_year":"2025","word_count":1044,"keywords":["Go","funding","programming_languages:R","AI","innovation","semiconductor industry","programming_languages:Go","Aim","china chip technology","ViT","Chip War","R","China"],"extracted_tech_keywords":["AI","Aim","R","Go","ViT","innovation","funding","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/if-a-tariff-war-is-what-the-us-wants-were-ready-to-fight-china\/","complexity_score":4,"technical_depth":9,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10044810,"title":"How To Create Interactive Public Dashboards And Storylines In Tableau?","content":"The tabular data will not give us as much information as it contains; the messy format and large numbers of entries make it difficult to do further analysis. So here comes the birth of different data visualisation tools and techniques. Data visualization is the art of presenting data in different graphical charts so that non-technical people can understand it easily. Using a perfect combination of elements like colors, dimensions, and labels can create a masterpiece of visual reports that can reveal surprising insights, making businesses more growth. An increase in data analytics and data integration has made way for more specialized visual analytical tools. Typically files like excel spreadsheets are very good with analytics and visualization, but it has limitations like it can not handle big data, which is our main concern. On the other hand, specialised software leverages easy operation on both static and dynamic data, computational speed, self-service function, and interactive visualization facilitate users to pull up a report or dashboard or storyline and freely deep dive to granular levels of information. Tableau is one of the most used data visualization tools used in all industries, which helps o create interactive graphs and charts in the form of worksheets and dashboards to gain hidden insights easily. All this is made simple with gestures like drag and drop. Today, we mainly discuss creating an interactive dashboard and storyline for a given problem statement in this article. Problem Statement: A superstore is a very large supermarket which often sells household goods, clothes, electrical goods, and office furniture. So we need to find weak areas and strong areas where we can increase the profit. The dataset is available in CSV format, and you can download it from here. To work with tableau, you need to install it first. There are three versions of tableau, namely Public, Desktop, and online, out of which the Public version is available for free and for this tutorial, we are using it. I recommend this article if you are new to tableau; here, you will get a clear idea for installing, loading the data and creating the charts. Our dataset contains 13 attributes which contain information such as state, region, total sales done, total profit obtained, and so on. You can take a look at the spreadsheet and look at our dataset below; Load the worksheet: As the format of the data file is CSV, we need to load it as a text file from the home page of the tableau interface and later open a blank sheet from the bottom menu; you will get the interface like as below; Tableau categorizes the dataset’s attributes majorly in two types, i.e. Dimensions (inside the black box) and Measures(inside the brown box), which means categorical and numerical\/continuous variables. Carefully observe the above, specifically the highlighted part of the image, as we will spend most of our time on this interface. We will place those attributes at the columns tag that we want to place on the x-axis similarly to rows tag variables that we want to display on the y-axis. Fields insides the red box Marks and filters are used to apply color combination, display the count values at the top of the graph, filter out unwanted information from attributes. From the show me tab (inside right-sided red box), you can control the type of plot that you want to apply; you don’t need to choose those plots as you drag and drop the attributes to the column and rows filed the plot will be automatically generated based on the type of attributes. You can navigate between sheets and create new sheets, dashboard and story sheets from the orange bottom box. We will now create four plots, i.e. map view of the US, the quantity being sold per category and sub-category, profit per category, and total price per sub-category. Lastly, I will create a dashboard and storyline. Map view of US: To create a map view, we need to drag the country and drop it on the sheet. This will generate a map view automatically. Secondly, drag the state attribute and drop it at the Label option present under the Marks field; this will highlight different states’ names on the map. You might observe that two fields are generated under continuous attributes, i.e. Latitude automatically and Longitude; we can use any one of them to represent states with the meaningful color combination.  We can simply make this by dragging the latitude and dropping it to the color option. That’s it all of these operations will generate the map as shown below; Quantity being sold per category and per sub-category: Drag and drop categorical attributes Category and sub-category into column tags and the continuous attribute Quantity into the rows tag. This will create a bar chart representing the quantity being sold in all categories. To represent sales done by these categories, we can move the continuous attribute Sales into the color option; this will represent sales and quantity being sold for each category in a more meaningful way. From the legends at the right corner, you see the meaning of colors. Profit per category: Drag and drop Subcategory and Profit into columns and rows, respectively; this will create an automatic bar plot again. To show the profit at each bar, drag and drop profit into the Label option under the Marks field. Finally, make a meaningful color combination drag and drop Category into color. If you hover your cursor to any of the bars, this shows the information like which Category and Sub Category it belongs to and the numeric value of profit for that bar. You can see this from the image below; Total price per sub-category: We don’t have attributes containing information on the price for each subcategory present in the field. So, in that case, we can create a new attribute for this by dividing the existing attributes Sales by the Quantity. Now the attribute we wanted now just drag and drop the attribute we created into rows and subcategories into columns. This will create a bar plot as shown below; Now we created our four plots as we needed; let’s move to the dashboard; Creating Dashboard: The dashboard is nothing but the consolidated display, which combines many worksheets and related information in a single place. The main purpose of this is to monitor and variety of data simultaneously. All the plots are displayed at once. To do this, we have to open a blank Dashboard, and there you have to drag and drop the worksheets you need to display.  And finally, to make it interactive, you have to filter the action from the Dashboard option present at the main menu bar. Finally, the dashboard will look like this; Creating Storyline: In general, the story in Tableau is nothing but a collection of different dashboards to show the sequence of events or visualise the main aspect of the analysis. We can create the storyline in the same way as we created the dashboard. Open a blank story sheet drag and drop dashboard you have created, then make duplicates of it and create a story as shown below; Conclusion: So this was all about creating dashboards and a storyline. Creating an interactive dashboard and storyline makes a huge difference when presenting your report at some work or in conference. Unfortunately, Tableau Public doesn’t provide a facility to store this work in a local machine. You need to save it in a tableau account, and there you can see all these plots and the dashboard. I have saved the Dashboard and Storyline at the tableau account; you can check those. References: Tableau Documentation","excerpt":"An increase in data analytics and data integration has made way for more specialized visual analytical tools. Typically files like excel spreadsheets are very good with analytics and visualization, but it has limitations like it can not handle big data, which is our main concern. On the other hand, specialised software leverages easy operation on both static and dynamic data, computational speed, self-service function, and interactive visualization facilitate users to pull up a report or dashboard or storyline and freely deep dive to granular levels of information.","categories":["AI Trends"],"tags":["Big Data","Data Analytics","Data Science","Tableau","tabular data","Visualization"],"author_name":"Vijaysinh Lendave","publish_date":"2021-07-29T19:00:00","publication_year":"2021","word_count":1275,"keywords":["big data","Go","Tableau","programming_languages:R","AI","Visualization","programming_languages:Go","tabular data","Git","RAG","analytics","Data Analytics","Data Science","Big Data","R"],"extracted_tech_keywords":["AI","analytics","RAG","R","Go","Git","big data","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/how-to-create-interactive-public-dashboards-and-storylines-in-tableau\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":902,"title":"Does Analytics Matter?","content":"In its May 2003 edition, Harvard Business Review published an article called “IT Doesn’t Matter”, which since then became one of the most debated articles on IT strategy. The article argued that IT doesn’t matter anymore to corporate strategy. As the availability of IT increased and the cost decreased, IT has become ubiquitous and thus a commodity inputs. The basis for a sustained competitive advantage is scarcity, not ubiquity. Hence, IT has become a commodity resource and, therefore, is no longer strategic. We take a step further and investigate the same argument for Analytics. It is quite evident that analytics follows a pattern of adoption strikingly similar to that of IT. For the period when IT was evolving and taking a stronger foothold into the way businesses were run, IT opened opportunities for forward-looking companies to gain strong competitive advantages. But, can we say the same today when the core functions of IT – data storage, data processing and data transport – have become available and affordable to all. Analytics is still at a very nascent stage compared to what IT was ten years back. Vendors are launching products and developing advanced tools, of which not many single products\/ tools have gained widespread industry acceptance. Moreover, standard models for problem solving are years away from gaining extensive recognition. Organizations and analytics professionals are learning and inventing as they move up the knowledge curve. On the other hand, some industries were early adopters of analytics. Specific case in point here is the financial industry, especially in developed world. Capital One’s success in the last decade of the century was due to its being the early adopter of customer information to derive meaningful insights and provide mass customized products based on this information. Yet, most other insurers later caught up with this information based strategy. Analytics today is utilized by most medium to large sized financial institutions in west. The very power and presence of analytics in financial industry has transformed analytics from potentially strategic resources into commodity factors of production. Today, optimized pricing models, customer scorecards and risk models are almost ubiquitous in insurance industry. Analytics is becoming the cost of doing business that must be paid by all but provide distinction to none. In India, the divide between adopters and non-adopters of analytics is pretty wide. While talking to the CMO of a leading Life Insurance company in India, I asked her about the usage of analytics in her department. She was quick enough to dismiss it as another flash in the pan. Similarly, ecommerce adopted analytics pretty early in its evolution. Today, visitors based customized stores are a common feature on an ecommerce site. Even the immature ecommerce industry in India has recognized the value of analytics and has in house teams working on it. What does it imply for cost conscious organizations and managers today? For industries where analytics adoption is still in its nascence, postponing analytics investments would derive more value for its money. A company will make smarter decisions as analytics capabilities become more homogenized later down the adoption curve. Moreover, a brief disruption in analytics availability can be devastating to a company, and the risks of more and different disruptions are increasing. For example, internet sentiments based algorithmic trading is becoming widely utilized in stock trading area, yet critics have raised voice against it when in the past a small disruption in its logic took the whole stock market crashing. It is better to be a laggard than to face the risks of obsolesce and disruption. In a recent study (Wiggins & Ruefli), only 2 percent of the firms in a large sample managed to sustain superior performance through disruptive strategies over a 10-year period. Yet, you can be a laggard for a while, but then you eventually get blown away by companies that have more innovative business strategies and are leaders in adopting them. The equation changes where analytics has been adopted early. For industries where adoption was early and deep, analytics may become commodity and alone doesn’t give strategic competitive advantage. What matters here is the intelligent and innovative application of information that creates customer value. For these industries, correlation between analytics spending and performance would be low. As analytics become more homogenized, cost performance should only improve. Extracting value from analytics requires innovations in business practices. Companies that insert analytics without changing business practices or way of thinking destroy any economic advantage. There is a tendency of corporations to believe that analytics just complements their existing business processes by providing in depth information on how business is being run. Rather, it is much more than that. It is an enabler of business processes and practices, and most likely a disruptive value provider. It can completely transform the way businesses were traditionally run. Analytics will be utilized by visionary companies that see possibilities where others do not and which exploit the information based strategy to derive value never even predicted. If you are aggressive player in your industry and want to stay ahead of your peers, go on to invest heavily on analytics irrespective of where your industry stands on adoption curve.","excerpt":"In its May 2003 edition, Harvard Business Review published an article called “IT Doesn’t Matter”, which since then became one of the most debated articles on IT strategy. The article argued that IT doesn’t matter anymore to corporate strategy. As the availability of IT increased and the cost decreased, IT has become ubiquitous and thus […]","categories":["IT Services"],"tags":["Analytics Case Study"],"author_name":"Дарья","publish_date":"2012-09-03T17:51:54","publication_year":"2012","word_count":859,"keywords":["Go","API","programming_languages:R","AI","innovation","RAG","GAN","analytics","disruption","Analytics Case Study","R"],"extracted_tech_keywords":["AI","analytics","RAG","R","Go","API","GAN","innovation","disruption","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/it-services\/does-analytics-matter\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10143405,"title":"Palantir’s New Cohort to Drive Manufacturing Innovations at ‘Warp Speed’","content":"Palantir, the US-based data analytics firm, has announced a new cohort of companies that will use its tool Warp Speed to reindustrialise the United States’ manufacturing and production capabilities. Companies including Anduril Industries, L3Harris, Panasonic Energy of North America (PENA), and Shield AI are part of the cohort. Warp Speed is Palantir’s manufacturing operating system that provides a unified platform for companies to access multiple production tools. Warp Speed is focused on adapting towards companies’ production and business processes rather than the other way around. The platform combines several tools like enterprise resource planning (ERP), manufacturing execution systems (MES), product lifecycle management (PLM), and programmable logic controllers (PLC). This includes products from notable companies such as Siemens, SAP, Oracle and SolidWorks. “The inaugural cohort is already using the software to gain an advantage in dynamic production scheduling, engineering change management, automated visual inspection for quality, and more,” read the announcement. One of the member companies in the cohort, Anduril, was able to observe a 200-fold efficiency gain in dealing with supply shortages. A few days ago, Anduril announced a partnership with OpenAI to bring AI technologies to the U.S. national defence and security interests. Leaders from other companies followed suit with a similar sentiment. “Warp Speed is enabling us to rapidly transform our manufacturing operation in Nevada and accelerate the ramp-up of our new factory in De Soto, Kansas,” said Allan Swan, President of PENA. In another instance, Shield AI says they are able to handle a ‘record demand’ for V-BAT, their flagship unmanned aerial system, through Warp Speed OS. “Warp Speed will help our different functions identify chokepoints and stay in lock step,” said Ryan Tseng, CEO and founder of Shield AI. Last Month, Palantir also announced a partnership with AWS and Anthropic to provide US intelligence and defence agencies with Claude 3 and 3.5 models. “Our partnership with Anthropic and AWS provides US defence and intelligence communities the toolchain they need to harness and deploy AI models securely, bringing the next generation of decision advantage to their most critical missions,” said Shyam Sankar, chief technology officer at Palantir.","excerpt":"‘At the dawn of WW2, we didn’t have a Defense Industrial Base; we had an American Industrial Base. This is also what our future must look like.’","categories":["AI News"],"tags":["AI (Artificial Intelligence)","Data Science","palantir"],"author_name":"Supreeth Koundinya","publish_date":"2024-12-12T13:23:21","publication_year":"2024","word_count":350,"keywords":["Anthropic","Go","API","OpenAI","AI","AWS","cloud_platforms:AWS","llm_models:Claude","analytics","palantir","Data Science","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","analytics","OpenAI","Anthropic","AWS","R","Go","API","llm_models:Claude","cloud_platforms:AWS"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/palantirs-new-cohort-to-drive-manufacturing-innovations-at-warp-speed\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10042327,"title":"Free Resources To Prepare For TensorFlow Developer Certification","content":"TensorFlow Certification is a testament to developers’ expertise in machine learning. The certification programme essentially consists of an assessment examination developed by the TensorFlow team. On passing the exam, developers can join the TensorFlow Certificate Network. Becoming an expert in TensorFlow is no easy feat. The candidates will learn to build models via Computer Vision, Convolutional Neural Networks(CNNs), and Natural Language Processing(NLP). Certification is ideal for developers who wish to build applications that use Machine Learning techniques. We have curated a list of courses and study sources to help you learn the concepts of TensorFlow. Introduction to Deep Learning: MIT MIT’s course, as the name suggests, is an introductory course taught by PhD researchers. The course consists of videos and review papers, same as those delivered to MIT students, and takes around 24-hours for a learner to complete. The prerequisite for taking the course includes knowledge of calculus(differentiation and integration), linear algebra(matrices), and probability (Baye’s theorem). Working knowledge of Python is desirable but not compulsory. The course focuses on applying deep learning to Natural language processing, computer vision, and more. The course teaches the basics of deep learning algorithms and provides practical experience and training to develop neural networks in TensorFlow. For more information, click here. Intro to Machine Learning by Tensorflow: Udacity The three-month Udacity course is taught in collaboration with Kaggle and Amazon Web Services. Learners willing to take the course should have 40 hours of experience in programming structures and good familiarity with data structures like dictionaries and lists. Understanding libraries like NumPy and Pandas is desirable, along with basic knowledge of probability and statistics. Foundational knowledge of neural network design and training in TensorFlow is a prerequisite. After completing the course, learners will have hands-on experience with deep learning models like state-of-the-art image classifiers. Moreover, it will enable learners to apply TensorFlow models with advanced techniques and algorithms to work with large datasets. For more information, click here. Introduction to TensorFlow for AI, ML & Deep Learning: Coursera Delivered by Google’s legal AI advocate Laurence Moroney, this 30-hours course is designed for people with no prior knowledge of TensorFlow. It, however, requires learners to have a basic understanding of Python and high-school mathematics. While some knowledge of Machine Learning models is helpful, it is not a prerequisite. After completing the course, learners will be able to use TensorFlow principles to build and apply scalable models for addressing real-world problems. The course also teaches one to build primary neural networks on TensorFlow and train them for Computer Vision applications. For more information, click here. Convolutional Neural Network in TensorFlow: Coursera Once a learner completes the ‘Introduction to TensorFlow for AI, ML & Deep Learning’ course in Coursera, they can opt for the ‘Convolutional Neural Network in TensorFlow’ course. Laurence Moroney again teaches the 26-hour training programme. The prerequisite for the training includes basic knowledge of Artificial Intelligence, Machine Learning, developing neural networks, and coding in Python. After completing this course, learners will understand the tools required to build an Artificial Intelligence-powered algorithm and understand the best practices for using TensorFlow. The programme also teaches advanced techniques needed to improve computer vision models and handle real-world image data. It helps in the exploration of strategies to prevent overfitting, including augmentation & dropout. For more information, click here. Natural Language Processing in TensorFlow: Coursera If one wants to go step-by-step, this Coursera course is the third step to master TensorFlow. It’s a 14-hour long training session by Laurence Moroney. The prerequisites for this course include a working knowledge of Artificial Intelligence, Machine Learning, and Deep Learning. In addition, learners must know how to develop neural networks and know Convolutional Neural Network on the TensorFlow platform. The course will enable learners to build NLP models using TensorFlow and help processing text, including tokenization and representing sentences as vectors. In addition, it trains LSTM on existing text to create original poetry. For more information, click here. Sequences, Time-series, and Prediction: Coursera The fourth and final step of Courser’s TensorFlow training is a 13-hour long programme. Learners must know about NLP, Neural Networks, TensorFlow, mathematics, Artificial Intelligent, and Machine Learning. Additionally, the learner must be able to build neural networks and use computer vision on the platform. Upon completing this course, learners will know how to implement best practices to prepare time series data. It will allow learners to explore how RNN’s and ID ConvNets can be used for Prediction. Finally, learners will also be able to build a sunspot prediction model using real-world data. The course will ensure proficiency in Forecasting, Machine Learning, TensorFlow, Time series, and Prediction. For more information, click here. Kaggle competitions It is a great library and competition space to track one’s progress, test their knowledge, and see if their model is good enough to get them a place on the leaderboard. The platform has many courses on Natural Language Processing, Neural Networks, and Computer vision. After completing the course, the learner also receives a certificate. The platform offers machine learning models also allowing one to earn hundreds of thousands of dollars and are a great way to track their progress. The Kaggle also has a very active community allowing discussion around TensorFlow and its applications. For more information, click here. Coding TensorFlow: YouTube A series of 42 videos, each ranging between four to eight minutes, is taught by different instructors and uploaded by the TensorFlow platform. To understand the videos, the learner must have basic knowledge of Artificial intelligence, Machine learning, Neural networks, and high school mathematics. The videos explore the significant dimensions of Computer vision, Natural Language Processing, and Neural Networks. It helps solve problems by training Machine Learning models and allowing learners to build these models on the TensorFlow platform. For more information, click here.","excerpt":"TensorFlow Certification is a testament to developers’ expertise in machine learning. The certification programme essentially consists of an assessment examination developed by the TensorFlow team. On passing the exam, developers can join the TensorFlow Certificate Network.  Becoming an expert in TensorFlow is no easy feat. The candidates will learn to build models via Computer Vision, […]","categories":["AI Features"],"tags":["Computer Vision","Deep Learning","Neural Networks","what is tensorflow"],"author_name":"Meenal Sharma","publish_date":"2021-06-24T18:00:00","publication_year":"2021","word_count":961,"keywords":["artificial intelligence","machine learning","AI","neural network","what is tensorflow","ML","computer vision","NLP","deep learning","Computer Vision","Deep Learning","TensorFlow","Pandas","Neural Networks"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","neural network","NLP","computer vision","TensorFlow","Pandas"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/free-resources-to-prepare-for-tensorflow-developer-certification\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":69795,"title":"Auto companies are using cognitive IoT technologies to make cars safer","content":"Road deaths are on the rise, with almost 1.5 million people killed in 2016 globally in road accidents, pressure has been mounting on carmakers and technology providers to tackle the fatalities caused by car crashes. Safety systems to prevent cars from drifting into another lane or that warn drivers of vehicles in their blind spots are beginning to live up to their potential to significantly reduce crashes. New technology around automobiles and vehicle safety have allowed us to move beyond passive safety and focus more on active safety. Cognitive IoT technologies are beginning to assist in preventing car crashes. How automobile giants are using IoT for car safety: Automobile companies have been working towards developing self-driving cars, which rely heavily on cognitive IoT for safety. Undoubtedly, the biggest concern for all companies and governments has been car crashes. However, with the latest technologies are helping big players make safer cars. We will talk about the technologies later, but let us first see how some leading companies are using IoT to make their cars safe. Volvo: It has become the first automaker to debut radar based safety systems in automobiles with the launch of XC90 hybrid in India. Volvo’s radar based security systems include City Safety, a co-pilot system that detects obstructions in the way and brakes automatically if a collision is imminent. Another set of features that use radar based technology is Intellisafe. This is the system that works to actively avoid an accident and includes features like Adaptive Cruise Control. At speeds less than 50kmph Pilot Assist technology works alongside the Adaptive Cruise Control to keep the car in the lane with gentle steering inputs, should the driver deviate. Tesla: Tesla worked with its radar supplier, Bosch, to get upgraded drivers and access to raw input from the radar antenna on its Model S and X vehicles in order to build its own processing of the data. It enabled Tesla to push new safety features and owners are already reporting that the system has helped them avoid accidents. Tesla’s new suite with 360-degree camera coverage and ultrasonic sensors with a better range have the potential to be much more efficient. One of the most impressive features enabled by the new radar processing capacity is the ability for the system to see ahead of the car in front and basically track two cars ahead on the road. The radar is able to bounce underneath or around the vehicle in front of the Tesla Model S or X and see where the driver potentially can not because the leading vehicle is obstructing the view. Honda: By leveraging IBM Bluemix and the Watson IoT platform, Honda has been able to move further along its connected car initiative set for 2020. Honda has not only improved its products and services but also transformed the processes within the organisation. The Driving Coach System is able to capture data based on: distance to other vehicles distance to other objects vehicle placement on the lane break distance and timing driver behaviour Using this information and real-time analysis, the Driving Coach System can adapt and adjust coaching to fit any driving behaviour. The system is able to protect drivers through early warnings of dangerous situations. Waymo: Google’s complementary sensors and software, don’t rely on a single type of data to drive. Its suite of cameras, lasers and radars work together to give the cars 360-degree visibility, so even if there’s a glitch in one camera, the cars can still safely pull over. Moreover, each of the self-driving vehicles is equipped with a secondary computer to act as a backup in the rare event that the primary computer goes offline. BMW: BMW Group will co-locate a team of researchers at IBM’s global headquarters for Watson Internet of Things in Munich, Germany, and the companies will work together to explore how to improve intelligent assistant functions for drivers. Daimler: In its pursuit of safer roads, Daimler has already implemented technologies such as proximity control, stop-and-go assist, emergency brake assist, lane-keeping assist and 3-D maps. Such features allow a vehicle to automatically keep a safe distance from other vehicles in a wide variety of traffic and road conditions, in addition to automatically braking if the need arises. Daimler also has integrated improvements to its road monitoring systems with innovations, such as a stereo camera and radar sensors, which allow for greater accuracy and improved response times. The technologies taking over: Vehicle-to-Vehicle communication: The new age sensor technology which will focus on the cars to communicate, will also detect pedestrians, bicycles within its proximity and adjusts the car’s speed accordingly. This will also help to create a network based traffic management system. Vehicle-to-Infrastructure (V2I) Communications: Vehicle infrastructure integration connects cars to physical surrounding and helps manage traffic and prevent accidents by alerting drivers to the traffic situation ahead and also to identify the most suitable route to the intended destination. Data Capture and Management: Connected cars operate as sources of IoT data generating information enabling critical transportation, logistics, and freight management decisions. Internet of Automobiles: According to IBM, 100% of cars are expected to be connected by 2025, to transform automobiles into auto-mobility enabled by Cognitive IoT.  Advanced safety features in the car have been designed to monitor driving patterns, or observe the driver directly to see they are driving safely and not nodding off while driving. A combination of local online control loops, aided by fleet-level or personalised learning in the cloud can provide this feature. Cars have emerged as the ultimate mobile machines in the prevalent era of the Internet of Things. It’s only a matter of time when the popular IoT car applications are embraced on a global scale as a necessity and not merely as innovation or luxury.","excerpt":"Road deaths are on the rise, with almost 1.5 million people killed in 2016 globally in road accidents, pressure has been mounting on carmakers and technology providers to tackle the fatalities caused by car crashes. Safety systems to prevent cars from drifting into another lane or that warn drivers of vehicles in their blind spots […]","categories":["AI Features"],"tags":["system integration companies"],"author_name":"Priya Singh","publish_date":"2017-09-18T11:59:38","publication_year":"2017","word_count":961,"keywords":["Go","programming_languages:R","AI","innovation","ML","programming_languages:Go","RAG","system integration companies","Aim","GAN","R"],"extracted_tech_keywords":["AI","ML","Aim","RAG","R","Go","GAN","innovation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/auto-companies-using-cognitive-iot-technologies-make-cars-safer\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":63660,"title":"Geoff Hinton Agrees That He Didn’t Invent Backpropagation: A New Chapter In The Ongoing Saga","content":"The ongoing episode between the pioneers of AI, Juergen Schmidhuber and Geoff Hinton, only gets worse as Dr Schmidhuber responds to Dr Hinton getting awarded the Honda prize back in November 2019. The Honda Prize, established in 1980 and awarded once every year, is an international award that recognises the work of individuals or groups generating new knowledge to drive the next generation, from the standpoint of eco-technology. In the press release, the Honda prize authorities applauded Dr Hinton for creating several technologies that have enabled the broader application of AI, including the backpropagation algorithm that forms the basis of the deep learning approach to AI. However, the backpropagation part of the above statement did not go down well with Dr Schmidhuber, and he explained in his blog everything that is wrong with Dr Hinton’s latest accolades. Though Dr Schmidhuber’s argument for not crediting the original creators is quite old, there seems to be no change in the actions across the AI community. And Dr Schmidhuber does not look like he is going to stop anytime soon. And he quotes Elvis Presley: “Truth is like the sun. You can shut it out for a time, but it ain’t goin’ away.” Critique Of Hinton Dr Schmidhuber, in his latest blog post, began his critique of Dr Hinton by admitting that Hinton and his co-workers have made significant contributions to deep learning, but he discarded the claim of creating the backprop algorithm. He explained, once again, how they have failed to cite Paul Werbos, who proposed the same method three years before Dr Hinton’s seminal paper of 1985. They even failed to mention Seppo Linnainmaa, the inventor of this famous algorithm for credit assignment in networks.Dr Schmidhuber Dr Schmidhuber also explained why Hinton’s interesting unsupervised pre-training for deep NNs was irrelevant for the current deep learning revolution, and that, it is actually his team who showed that deep feedforward NNs (FNNs) can be trained by plain backpropagation and do not at all require unsupervised pre-training for important applications. This, he says, was achieved by greatly accelerating traditional FNNs on highly parallel graphics processing units called GPUs. Hinton Responds Dr Hinton, in response to this criticism, has responded on a popular machine learning channel on Reddit. Dr Hinton, who has been mostly silent on Schmidhuber’s consistent criticism, said that having a public debate with Schmidhuber about academic credit is not advisable and that is why he has chosen Reddit to respond. But he tried to end the debate once and for all by stating that he never claimed to have invented backpropagation. Dr Hinton, in his response, also claimed that Dr Schmidhuber has even resorted to tricks like having multiple aliases in Wikipedia to make it look as if other people are agreeing with what he says. The page on his website, says Dr Hinton, about Alan Turing is a nice example of how he goes about trying to diminish other people’s contributions. “I have never claimed that I invented backpropagation.” Dr Hinton David Rumelhart, wrote Dr Hinton, invented it independently long after people in other fields had invented it. He also admitted to failing to cite them owing to lack of knowledge of history. It is true that many people in the press have said I invented backpropagation and I have spent a lot of time correcting them “Lots of different people invented different versions of backpropagation before David Rumelhart. They were mainly independent inventions, and it’s something I feel I have got too much credit for. I’ve seen things in the press that say that I invented backpropagation, and that is completely wrong. It’s one of these rare cases where an academic feels he has got too much credit for something! My main contribution was to show how you can use it for learning distributed representations, so I’d like to set the record straight on that.” Dr Schmidhuber was quick to retaliate to Dr Hinton’s response. He added to his original blog by saying that Dr Hinton has failed to respond accurately and has only resorted to ad hominem attacks on him. “To summarise, Dr Hinton’s comments and ad hominem arguments diverge from the contents of my post and do not challenge any of the facts presented. The facts still stand.”Dr Schmidhuber Why Is This Important? At the heart of this whole event, one cannot help but notice where Dr Schmidhuber is coming from. He believes in crediting the original creators and that it establishes a healthy ecosystem in the ML community. There is no doubt that Dr Hinton and his colleagues have contributed tremendously to the advancement of AI, but popularising few people can affect any domain in its initial stages. The celebrated trio of the modern AI scene, Yann Lecun, Yoshua Bengio and Geoffrey Hinton, have been at the forefront of every major AI breakthrough news that has come over in the past decade. However, the trio has been accused of circular citations by Dr.Schmidhuber in his 2015 post titled, “Deep Learning Conspiracy”, in which he explained in detail how LBH have been ignorant of the original inventors. While the community is divided over Dr Schmidhuber’s polemics, they all can agree upon one thing – if there is an individual who can administer an authority to challenge the likes of Dr Hinton, then it is him.","excerpt":"The ongoing episode between the pioneers of AI, Juergen Schmidhuber and Geoff Hinton, only gets worse as Dr Schmidhuber responds to Dr Hinton getting awarded the Honda prize back in November 2019. The Honda Prize, established in 1980 and awarded once every year, is an international award that recognises the work of individuals or groups […]","categories":["Global Tech"],"tags":["back propagation","Juergen Schmidhuber"],"author_name":"Ram Sagar","publish_date":"2020-04-27T16:00:19","publication_year":"2020","word_count":887,"keywords":["Go","back propagation","machine learning","programming_languages:R","AI","ML","Juergen Schmidhuber","programming_languages:Go","Aim","deep learning","GAN","R"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","Aim","R","Go","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/geoff-hinton-schmidhuber-ai\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":64386,"title":"Should APIs Be Protected By Copyright Law?","content":"Google vs Oracle America is an ongoing legal case within the United States concerning copyright law within software code. The case focuses on the utilisation of parts of the Java programming language’s application programming interfaces (APIs), which are owned by Oracle, within initial variants of the Android operating system by Google. Google confirmed using the APIs but holds its initial use of the APIs under fair use of copyright. Oracle launched the suit, claiming that the APIs were copyrightable, asking $8.8 billion in damages. The two District Court-level jury trials in the US decided in favour of Google, and then the Federal Circuit court invalidated both verdicts, statting that APIs are in fact copyrightable, and Google’s application missed a fair use agreement. Many argue that APIs are perhaps too functional and open to come under the purview of intellectual property (IP) rights. APIs, generally have been considered to provide interoperable use between different software may not be protected by copyright. What the people at Android did was to create their own packages that worked through the Java API. Writing those allows not only traditional Java Developers to write for Android, but also makes easy interoperating Java programs to Android. Google petitioned to the Supreme Court to hear the case in the 2019 term, concentrating on the copyrightability of APIs and consequent fair use. The case is of high concern within the tech and software sector, as many software programs and libraries, especially in open-source, are produced by recreating the functionality of APIs from other products to help developers in interoperability between different systems or platforms. What Should Be Fair Use Of APIs That Can Be Allowed For Innovation Google appealed to the court, citing using the APIs for fair usage and the decision was made in favour of Google in the lower court. After, it was reappeared by Oracle saying it could not be deemed fair usage as Google has created commercial applications and made billions of dollars off the Java APIs that Oracle had owned. In most cases, APIs are considered not copyrightable because there are a lot of interoperable software functions that happen on open-source platforms, which do not want APIs to be protected by copyright. Although, there are certain licenses which are issued that can dictate who can have access to certain functions. Let’s say proprietary APIs are declared as protected by copyright. In that case,  going back in time, even a single developer would have to attach open-source licenses in millions of lines of code on platforms such as GitHub, which permits for free collaboration on any number of software projects, and there are millions of projects hosted. This is virtually impossible even if a developer wants to protect himself\/herself from any legal action. Why Developers Must Pay Attention To The Case The issue of copyrightability influences all developers who may call hundreds or millions of APIs each day. The ongoing case is going set a precedent in deciding how open-source software and the innovation therein is going to be impacted, and developers will have to be mindful of how they use APIs and whether they have to use specific licenses for using those APIs to build interoperable interfaces between different software and platforms. In rare cases, such as Oracle\/Google, we do see that APIs could be declared copyrightable from the perspective of Oracle, which stated that Google didn’t use its APIs fairly. In the end, it’s not clear whether APIs are protected by copyright law or not, and there needs to come more clarity on the matter.","excerpt":"Google vs Oracle America is an ongoing legal case within the United States concerning copyright law within software code. The case focuses on the utilisation of parts of the Java programming language’s application programming interfaces (APIs), which are owned by Oracle, within initial variants of the Android operating system by Google. Google confirmed using the […]","categories":["AI Features"],"tags":["APIs","copyright"],"author_name":"Vishal Chawla","publish_date":"2020-05-04T10:00:00","publication_year":"2020","word_count":592,"keywords":["Go","API","programming_languages:R","AI","innovation","Git","APIs","Aim","copyright","GitHub","R","Java"],"extracted_tech_keywords":["AI","Aim","R","Go","Java","Git","GitHub","API","innovation","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/should-apis-be-protected-by-copyright-law\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10091448,"title":"OpenAI Puts A Pause on GPT-5, Focuses on AI Safety","content":"In an event held at MIT last evening, Sam Altman confirmed to Lex Friedman that OpenAI is not working on GPT-5. He also addressed the letter on 6-months ban on AI research and spoke about the vision the company had regarding safety. Altman via a video interview talks about how an earlier version of the letter talks about OpenAI training GPT-5, for which he confirms that they are not working on the next version and won’t for “some time.” He said that the company is still prioritizing safety issues for the current model of GPT-4 which are important and were “totally left out.” Not In Sync with Open Letter Altman also emphasizes that the letter is missing the most technical nuances about “where we need to pause.” Probably, in the wake of rising AI concerns, he reiterated the company’s outlook towards AI safety. He believes that moving with caution and “increasing rigor” for safety issues is really important. Sam Altman also believes that technology is going to impact all of us and that engaging everyone in the discussion is important. Putting the system out to the world which is “deeply imperfect” for people to experience and think about the upsides and downsides is “worth the trade-off”. Though they intend to embarrass themselves in public and change their minds and data frequently, the company will continue to do what they are doing as a big part of their goal is to “get the world to engage with this” and eventually understand what is the future that we all want. Since last week, this is the third statement\/announcement released by OpenAI regarding safety of their models. A week ago, the company released an elaborate blog on their AI safety measures which touched upon the surface without addressing deep issues. In continuation to the announcement, yesterday Greg Brockman released a statement on how the company is focussing on the safety and regulations part without again not addressing the “how” part of things. Gary Marcus was quick to react to the interview by stating the ban letter in fact “encouraged” some AI research. While others were quick to question the truth behind GPT-5’s actual status.","excerpt":"Sam Altman confirms that the next version of GPT-5 is not in cards for “some time”. OpenAI to focus on GPT-4 safety issues.","categories":["AI News"],"tags":["AI Safety","gary marcus","GPT-4","GPT-5","gpt5","Greg Brockman","lex fridman","MIT","OpenAI","Sam Altman"],"author_name":"Vandana Nair","publish_date":"2023-04-14T15:28:47","publication_year":"2023","word_count":361,"keywords":["Go","Sam Altman","GPT-5","OpenAI","AI","MIT","GPT-4","Greg Brockman","RAG","GPT","gary marcus","gpt5","lex fridman","AI research","AI Safety","AI safety","R","llm_models:GPT"],"extracted_tech_keywords":["AI","GPT-5","OpenAI","RAG","R","Go","GPT","AI safety","AI research","llm_models:GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/openai-puts-a-pause-on-gpt-5-focuses-on-ai-safety\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10065870,"title":"Understanding AI biases and ways to fix them","content":"According to a report by PwC, AI could potentially contribute up to USD 15.7 trillion to the global economy in 2030– higher than the current output of China and India combined. However, the exponential growth of AI has brought its own set of problems. Bias is one of the major issues the stakeholders are grappling with. However, bias in algorithms is not new. It goes back to the 80s when Dr Geoffrey Franglen of St George’s Hospital Medical School wrote an algorithm to screen student applications– the algorithm prioritised caucasian names. Below, we look at the major biases in AI. Prejudicial bias According to the Mitigating Bias in Artificial Intelligence report by the Haas School of Business, AI systems are biased because they are human creations. They are classification technologies and are products of the context in which they are created and often mirror society. The perspectives and knowledge of those who develop AI systems are integrated into them, said the report. The biases can enter the development phase of an AI system. “Human biases can be introduced into an AI system in multiple ways. It could be due to the training data that is used for machine learning algorithms, or it could be because of the biases carried by humans,” said Sarvagya Mishra, co-founder and director of SuperBot (PinnacleWorks). Sample bias Biases also happen because of data inaccuracy. A McKinsey report titled ‘Tackling bias in artificial intelligence‘ found underlying data as the source of bias in most cases. If the data fed into a model does not represent the demographics fairly, it leads to biases. Let’s take the example of a facial recognition AI system. If the dataset used to train the AI contained information on white men\/women, chances are the AI model might be biased towards men\/women of colour. Historical bias At times, existing biases in historical data get fed into the AI system. For example, in 2018, Amazon had to scrap its AI recruiting tool that was biased against women applicants. The AI was trained on patterns in resumes submitted over ten years, and most of these resumes were from men. Aggregation bias Clustering of different classes or populations unfairly often leads to poor representation and biases. For example, let’s take the income scale of lawyers and athletes. While a lawyer’s income increases as he grows older because of his accumulated experience, an athlete makes most of his income early in his career. Aggregating both groups will lead to biases. Evaluation bias Sometimes the data can be accurate, but biases occur during model evaluation. For example, the models are tested against certain benchmarks. However, in some cases, the benchmarks are not aligned with the model’s purpose, leading to biases. Further, biases also occur when an AI model can accurately predict values from the training dataset but cannot deal with new data. According to the World Economic Forum report, biases also accumulate over time (data drift). Fixing AI biases The first step in eliminating biases should be to ensure the data fed to the algorithms is accurate. Having the right dataset is pivotal. The developers should examine the training dataset and make sure it is representative and does not lead to sampling biases. Mishra said if the data is clean and its labelling is done extensively and accurately, it takes care of many biases. According to McKinsey, developers should also conduct a subpopulation analysis. It is important to calculate model metrics for specific groups in the dataset. This will allow the developers to see whether the model performance is identical across subpopulations in the dataset. Also, a timely check of the models is necessary as the outcome of machine learning algorithms changes as they continue to learn or when the training data changes. An algorithm might produce biased results because of an error in its development phase. But you may not realise the downstream impact until much later. “We need systems that can help us make informed decisions based on the available information while at the same time being accountable for the outcomes of the decisions made. It needs to be calibrated so that it doesn’t lead to harm or injustice,” said Mishra. The National Institute of Standards and Technology (NIST), in its report titled ‘Towards a Standard for Identifying and Managing Bias in Artificial Intelligence’ said we must widen the scope of where we look for the source of these biases. To eliminate or minimise biases, we must go beyond the machine learning processes and data used to train the algorithms and consider the broader societal factors that influence technology or how it is being developed. According to Satyakam Mohanty, chief product officer at Fosfor, it is hard to avoid the biases as fairness is not something you can crisply define in an ML pipeline and set thresholds while coding. Some biases are inherent and difficult to identify. However, some can be traced using the right methods and techniques. Bias identification is a challenge, but mitigation is possible by minimising the bias and introducing a demographic or statistical parity to equalise data representation. “A more practical approach is knowing the potential bias and impact of decisions until and unless it is a mission-critical application. Sometimes little bias plus decision generation is preferred over no decision with the caveat of potential risk,” Mohanty added. According to PwC research published last year, only 20 percent of enterprises had an AI ethics framework. And only 35 percent had plans to improve the governance of AI systems and processes. The need of the hour is to drive awareness, and bring more transparency and accountability to AI systems.","excerpt":"Human biases can be introduced into an AI system in multiple ways.","categories":["AI Features"],"tags":["AI biases"],"author_name":"Pritam Bordoloi","publish_date":"2022-04-28T15:52:43","publication_year":"2022","word_count":936,"keywords":["Go","artificial intelligence","machine learning","AI biases","AI","TPU","programming_languages:R","ML","programming_languages:Go","AI ethics","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","TPU","R","Go","AI ethics","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/understanding-ai-biases-and-ways-to-fix-them\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":37360,"title":"IIT Delhi Ups Its AI Research Capability, Finds Low-cost AI-powered Disease Detection System","content":"Researchers from IIT Delhi have developed an AI system that will help medical practitioners detect health conditions like malaria, TB, cervical cancer and intestinal parasites in milliseconds using a low-powered electronic hardware system. The new portable and low-cost device will be a breakaway from the current alternatives like rapid diagnostic tests and are relatively cheaper. Stating that the current diagnostic system warrants the issue to the high cost for running a test,  with the new device, researchers hope to make it readily available in resource-constrained areas with limited access to human specialists. “While many software AI models exist for healthcare and diagnostic-related applications, the need of the hour is to efficiently map these models on portable dedicated low-power, low-cost hardware to enable edge-AI systems accessible to all in low resource environment,” says Professor Manan Suri, Department of Electrical Engineering, IIT Delhi. The researchers have presented the paper in two healthcare conferences — IEEE BioCAS-2018 in Cleveland, and IEEE BioCAS-2017 in Torino, Italy. The research was also showcased in Rashtrapati Bhawan and had won the 2018 Gandhian Young Technology Innovation Award (GYTI). Push For AI Research In order to bolster the AI research in the institute,  in 2018, IIT-D announced that it will soon establish a Centre of Excellence (CoE) exclusively for AI. The new centre, which will make the Institute’s 15th, will be specially designed to facilitate collaboration, increase research and reskilling profession, “ “We are soon going to start a Centre of Excellence on Artificial Intelligence where 40 faculty members who have been working in areas related to AI will be given a platform. This centre of excellence will be a platform with a three-fold approach,” a professor from the institute’s department of computer science and engineering said to a leading daily. “It will help and engage the faculty working with different aspects of AI and bolster overall research, enhance the collaboration with the industry and re-skill the professionals already part of the industry with AI aspect,” the professor added. As part of this, the institute has signed an MoU with IBM’s AI Horizon Network, Massachusetts Institute of Technology (MIT) and the University of Michigan among other institutes. Speaking on the occasion, Prof V Ramgopal Rao, Director, IIT Delhi, said “India has immense talent to accelerate innovation in AI and related technologies. We are happy to collaborate with IBM Research scientists and provide opportunities to our students and faculty colleagues to work on some of the complex problems around AI and apply the solutions to real-world scenarios.” Apart from the opening of CoE in the institute, the Institute has been instrumental in introducing numerous online course on AI and Machine Learning to students. The students and faculties have been particularly active in the field of AI research and have contributed significantly to the development in the field. In 2017, four undergrad students from the college developed an AI-based reading app, as part of ‘Artificial Intelligence for Education HackWeek,’ a hackathon that was held in the college itself and won $6,000 for their innovation. ReadEx, is an assisted reading app developed by the students, that will let its users access other external information from the app so as to enhance their reading experience. “ReadEx is an android reading app that helps in enhancing the user’s learning from what he has read while offering an interface that would keep him engaged. It has been developed keeping in mind the problems we have ourselves faced as students,” Prakhar Gupta, one of the student developers of the app said. “The AI-powered app further offers the readers options like web searching, open files, or scan text images for reading and asks the user questions from the read text in real time via a Chatbot.  It also offers content recommendations by pinpointing the concepts in the text and keeping a record of previously answered questions through Flashcards to help the user to revise and learn from his previous mistakes,” IIT-D said in a press release.","excerpt":"Researchers from IIT Delhi have developed an AI system that will help medical practitioners detect health conditions like malaria, TB, cervical cancer and intestinal parasites in milliseconds using a low-powered electronic hardware system. The new portable and low-cost device will be a breakaway from the current alternatives like rapid diagnostic tests and are relatively cheaper. […]","categories":["AI News"],"tags":["AI Research","iiit delhi","what is power bi"],"author_name":"Akshaya Asokan","publish_date":"2019-04-05T08:23:52","publication_year":"2019","word_count":660,"keywords":["Go","API","artificial intelligence","what is power bi","machine learning","AI","programming_languages:R","innovation","AI Research","iiit delhi","GAN","AI research","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","R","Go","API","GAN","innovation","AI research","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/iit-delhi-ups-its-ai-research-capability-finds-low-cost-ai-powered-disease-detection-system\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10087065,"title":"Stability AI Announces New Medical AI Research Org &#8216;MedARC&#8217;","content":"Stability AI has announced the launch of a new medical AI research organisation called MedARC (Medical AI Research Center)—​​a novel, open, and collaborative approach to research dedicated to advancing the field of AI applied to healthcare. Announcing the launch of a new research organization supported by Stability AI! MedARC will focus on development of foundation models for medical AI research.https:\/\/t.co\/AzSj4hKjJ7— Stability AI (@StabilityAI) February 9, 2023 The development of large deep learning models, commonly referred to as foundation models, has opened the door to numerous innovative and previously unimaginable applications. However, despite these advancements, the use cases for medical AI have often been limited. Now, MedARC wants to develop these foundational models for medicine and build interdisciplinary teams to address clinical needs. What sets MedARC apart is that it wants to do science in the open. All their research and datasets will be made available to the public for free. “We have established a public environment with asynchronous communication, a Discord server, which allows others to discuss, contribute and follow research progress. This open approach has been successful for AI research in NLP (EleutherAI), multimodality (LAION), and biology (OpenBioML). Therefore, we believe a similar environment will be beneficial for medical AI research and enable new opportunities,” MedARC said in a blog post. Last year, Stability AI’s popularity soared following the launch of their text-to-image AI model, ‘Stable Diffusion’. But soon after, ChatGPT happened, and since then, the popular OpenAI chatbot has changed the AI landscape significantly. However, Stability AI is reportedly working on an open-source version of ChatGPT. Stability AI, the creator of stable diffusion, announced the development of an open source version of ChatGPT !— BoredGeekSociety | AI & Automation (@BoredGeekz) February 5, 2023","excerpt":"MedARC wants to develop these foundational models for medicine and build interdisciplinary teams to address clinical needs","categories":["AI News"],"tags":["AI Research","Stability AI"],"author_name":"Pritam Bordoloi","publish_date":"2023-02-10T11:43:44","publication_year":"2023","word_count":285,"keywords":["ChatGPT","OpenAI","AI","ML","AI Research","medical AI","NLP","GPT","deep learning","foundation models","R","Stability AI"],"extracted_tech_keywords":["AI","ML","deep learning","NLP","foundation models","ChatGPT","OpenAI","medical AI","R","GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/stability-ai-announces-new-medical-ai-research-org-medarc\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10096753,"title":"A Decade of React.js Ecosystem: Top 10 React Libraries","content":"Facebook had a huge user base back in 2011 making it a daunting task for the creators to make the experience seamless for its users. The objective was clear — to develop a user interface that was not only dynamic and responsive but also blazingly fast and exceptionally high-performing. React played a key role in achieving this back then. Brought to life by Jordan Walke, one of Facebook’s software engineers, React revolutionised the development process, offering a streamlined and structured approach to constructing dynamic and interactive user interfaces. This innovative framework empowered developers by providing reusable components, simplifying the creation of rich and engaging user experiences. However, React is not just like any other framework. It comes with a set of libraries that has tools which developers can put into use for different purposes like animation, creating forms or creative visually appealing user interfaces that improves the user experience. These libraries come equipped with pre-built components such as form inputs, pagination elements, menus, buttons, icon sets, and time\/date pickers, making it easier and faster to implement these essential features in React frameworks. As React celebrates its 10th anniversary in 2023, we present to you the top 10 React libraries that are widely used and trusted by developers. Redux Redux Toolkit is a widely used library for managing state in React applications. It offers a collection of tools and best practices that help developers handle state efficiently. With Redux Toolkit, you can define and update state using a simplified interface. It also provides useful features like immutable updates, which ensure the integrity of your state, and supports serializable action types for better control. SWR SWR is a well-known library used for managing server state in React applications. The name stands for stale-while-revalidate, which is a cache invalidation strategy widely recognized in the HTTP RFC 5861. Next.js Next.js offers a comprehensive range of powerful features, such as automatic code splitting, server-side rendering, and static site generation, among others. It is particularly well-suited for developing intricate applications that demand server-side rendering and SEO optimization. To begin working with Next.js, the simplest approach is to utilize create-next-app. Tailwind CSS Tailwind CSS is a CSS framework that prioritises utility-first development, allowing for the quick creation of custom user interfaces. It is a flexible and customisable framework that provides essential building blocks for designing unique interfaces. Unlike some other frameworks, Tailwind CSS avoids imposing opinionated styles that can be difficult to override, granting developers greater freedom in their design choices. Sentry Sentry is a comprehensive error-tracking framework that simplifies the process by handling the framework, back end, and visualization console for you. It offers seamless integration into your JavaScript codebase with minimal effort required. React Hook Form In 2023, React Hook Form emerged as the recommended choice for managing forms. This library is known for its lightweight nature, speedy performance, and user-friendly approach. React Hook Form simplifies form handling and validation. It offers a versatile API that facilitates the construction of forms, along with seamless integration with popular validation libraries like Yup and Zod. React Spring ​​Animations can greatly enhance user interfaces, and React provides several popular animation libraries like React Spring. It  simplifies the process of generating seamless and responsive animations, requiring minimal code to achieve impressive results. React Testing Library and Cypress When it comes to testing React applications, two excellent options to consider are React Testing Library for unit testing and Cypress for end-to-end testing. React Testing Library is a JavaScript testing tool specifically designed for testing React components. In automated tests, where there is no physical DOM to work with, React Testing Library comes to the rescue by offering a virtual DOM. This virtual DOM enables us to interact with and validate the behavior of React components, ensuring their functionality is thoroughly tested. Cypress is a dependable and powerful solution for end-to-end testing of React applications. It allows you to create tests that replicate real user interactions with your application, such as clicking buttons, entering text via keyboard, and submitting forms. With Cypress, you can comprehensively test your React application’s functionality and ensure a smooth user experience. React Router React Router stands out as a highly favored routing library for React applications. It offers a simple and expressive approach to managing routing, allowing developers to define routes and dynamically render components based on the current URL. With React Router, handling navigation becomes more straightforward and declarative in React applications. Recharts Recharts is renowned as a reliable and widely used React chart library among professionals and web developers. The components offered by Recharts are primarily designed for presentation purposes, aligning with their declarative nature. Reputed professionals often recommend Recharts as the top choice for those seeking a straightforward and simplified approach to accomplish their data visualization projects.","excerpt":"As React celebrates its 10th anniversary, we present to you the top 10 React libraries that are widely used and trusted by developers","categories":["AI Trends"],"tags":["jordan walke","Top Trend"],"author_name":"Siddharth Jindal","publish_date":"2023-07-12T12:04:07","publication_year":"2023","word_count":794,"keywords":["Replicate","Top Trend","API","Rust","programming_languages:R","AI","ML","JavaScript","jordan walke","R","Java","programming_languages:JavaScript"],"extracted_tech_keywords":["AI","ML","R","JavaScript","Rust","Java","API","Replicate","programming_languages:R","programming_languages:JavaScript"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/a-decade-of-react-js-ecosystem-top-10-react-libraries\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10097720,"title":"Did OpenAI Purposely Discontinue its AI Classifier?","content":"On any given day, OpenAI’s ChatGPT had the indiscretion of claiming that anything and everything was written by it. But, now OpenAI seems to have put a lid on this. Last week, it discontinued its AI classifier, a tool designed to determine the likelihood that a given text passage was written by another AI language model. They had launched it in January this year and discontinued it quietly on July 20. However, the timing to shut down the classifier is curious because OpenAI, along with other companies, made a voluntary commitment to developing AI ethically and transparently under the guidance of the White House. One aspect of this commitment is that the development of robust watermarking and detection methods to address the issue of AI-generated junk that’s filing the web faster than a mole. However, despite the companies’ promises, there have been no reliable watermarking or detection methods to date. For example, Google had announced that it is experimenting to include metadata with the image generated by its AI models to watermark them, but they haven’t put out any model for text. Why OpenAI Discontinued its AI Classifier? The decision to retire the tool was influenced by widespread criticism of its “low rate of accuracy.” While many users relied on the classifier to catch instances of low-effort cheating, it failed to deliver satisfactory results. However, many have pointed out the irony to have a dedication to both identifying AI content and striving to create AI content that closely resembles human behaviour simultaneously. And it seems like OpenAI has shed the veil and is completely focused on the latter now. It is argued that the detection of AI-generated content would not be effective and, to be frank, should not even be pursued due to its seemingly futile nature. This is particularly true in the context of AI-generated image creation, where watermarks can be effortlessly removed, posing a challenge for detection methods. While others humorously suggested that the ultimate goal of AI detection could be achieving world domination, akin to passing the Turing test flawlessly. The idea that AI-generated text might have identifying features or patterns that could be reliably detected appeared intuitive when OpenAI released its classifier. However, in practice, this has proven challenging due to the rapid development of large language models. The differences between various language models have made it difficult to rely on specific identifying features. Recent advancements in natural language processing have enabled large language models and made them capable of generating human-like texts for various tasks. However, this progress also presents challenges, as these LLMs can be misused for plagiarism, spamming, and social engineering to manipulate public opinion. To address this, there is a demand for efficient LLM text detectors to mitigate the misuse of publicly available LLMs. Can’t be Detected Various AI text detectors using watermarking, zero-shot methods, retrieval-based methods, and trained neural network-based classifiers have been proposed. However, a research paper shows, both theoretically and empirically, that current state-of-the-art detectors are not reliable in practical scenarios. Paraphrasing the LLM outputs effectively evades these detectors, allowing attackers to generate and spread misinformation undetected. Even the best detectors can only marginally outperform a random classifier against sufficiently advanced language models. The paper also demonstrates that watermarking and retrieval-based detectors can be spoofed to identify human-composed text as AI-generated, potentially harming the reputation of LLM detector developers. With the release of more advanced LLMs like GPT-4, the need for more secure methods to prevent misuse becomes crucial. A test conducted on OpenAI’s classifier correctly identified only one out of seven generated text snippets tested, and that was with a language model that was not even cutting-edge at the time of the test. Despite the limitations and disclaimers provided by OpenAI with the classifier tool, some users took its claims of detection at face value. This led to the misuse of the tool, as people would test suspected AI-generated content without realising its unreliability. There are potential vulnerabilities attackers might exploit in the future, such as improved paraphrasing models or smart prompting attacks. Current detectors should reliably flag AI-generated texts while avoiding excessive false positives to prevent wrongful accusations of plagiarism and protect the reputation of LLM developers. A recent follow-up work by Souradip Chakraborty argued that AI-text detection is almost always possible, even with low total variation between human and AI-generated distributions, but this may not hold in real-world applications due to correlations in human-written text. Other works suggest that existing LLM outputs are very different from human-written text, but the authors maintain that as language models advance, adversaries’ ability to evade detection will likely improve. In addition to reliability issues, the paper mentions the potential bias of detectors against non-native English writers. Having low average type I and II errors may not be sufficient for practical deployment if the detector performs poorly within specific sub-populations. While it remains a challenging task, progress in this area is essential for ensuring the responsible and trustworthy use of AI-generated text. Additionally, the first of its kind truly reliable watermarking or detection tool would be, as such a tool would be invaluable in various contexts.","excerpt":"Research shows that the current state-of-the-art detectors are not reliable in practical scenarios","categories":["AI Highlights"],"tags":["bias","Ethical AI","Language Models","Natural Language Processing","OpenAI","plagiarism","Responsible AI","Social engineering","Trustworthy AI","Turing test","Watermarking"],"author_name":"Shyam Nandan Upadhyay","publish_date":"2023-07-28T14:14:06","publication_year":"2023","word_count":855,"keywords":["TPU","Watermarking","Responsible AI","Turing test","Rust","R","ChatGPT","Natural Language Processing","RAG","Go","AI","neural network","Ethical AI","Social engineering","Language Models","bias","plagiarism","OpenAI","Trustworthy AI","Aim"],"extracted_tech_keywords":["AI","neural network","ChatGPT","OpenAI","Aim","RAG","TPU","R","Go","Rust"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/did-openai-purposely-discontinue-its-ai-classifier\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":60160,"title":"What Indian Tech Leaders Are Saying About COVID-19 Pandemic","content":"India has been going through a difficult time. COVID-19 has proven to be a bigger problem than what anyone could have anticipated. The pandemic has been spreading at an alarming rate and has impacted everything – from people’s daily lives, to operations in various industries. The outbreak has affected the tech industry as well, with growth having been revised down by 3-8% for the next financial year. Employees of most IT companies have been directed to work from home. Furthermore, many of these organizations have taken various initiatives to extend their support to employees who have been struck by this pandemic. Here are some of these initiatives led by business leaders as they get together to combat this disease collectively: Vijay Shekar Sharma, Founder of Paytm Vijay Shekar Sharma has pledged Rs 50 million for any medical innovation related to COVID-19. His company has also launched an ‘India Fights Corona’ campaign in partnership with consumer goods company Hindustan Unilever’s Lifebuoy and Yuvraj Singh’s YouWeCan Foundation. Paytm also plans to distribute one million soaps to those affected. Paytm users can contribute to the ‘Donation’ section in the app, or on the website. Innovators, research teams and doctors can also reach the company for support at covidcure@paytm.com. Sridhar Vembu, Founder of Zoho Sridhar Vembu recently announced a Small Business Emergency Subscription Assistance Program. The plan is to waive three months’ fee for software that small businesses have already been using. This move will benefit 20,000 of its existing customers who have 25 or fewer employees. These small businesses will get a Zoho Wallet credit, which will be equivalent to three months of their bill, and the business owners can choose to use the wallet credits as they deem fit. In addition to these efforts, Zoho is providing free access to its Zoho Remotely suite of web and mobile apps to offer remote working options for all businesses until July 1, 2020. This includes a set of 11 apps which provide services like document management and video conferencing, among others. “You will see the impact on guidance and revenue growth (by Indian IT companies) for next year. Next year, growth is going to be very challenging for the Indian IT services industry.” Ex-CFO, Infosys Ltd V Balakrishnan in an interview. Byju Raveendran, Founder of Byju’s The popular edtech firm’s flagship app – Byju’s – will have free access for users, and the offer stands until the end of April. It will offer math and science lessons suitable for children from grades 2-4, along with Disney Byju’s Early Learn app that offers math and English lessons for students from grades 1-3. This app could help thousands of students who have been stuck at home without the guidance of teachers. Manoj Bubna, Gautam Gurtoo & Biren Shah (Nitrogen founders) Nitrogen is a Mumbai-based SaaS company that announced free use of its digital platform for three months for hospitals, grocery outlets and pharmacies. The founders believe that the three industries – healthcare, consumer products and pharma – could face a lot of traffic during the 21-day lockdown. It can help websites handle peak traffic and strengthen their security. “The IT sector may see some indirect impact in the medium to long term as some of the clients have exposure to manufacturing in China.” Pravin Rao, NASSCOM vice-chairman, as well as COO of Infosys in an interview. Mukesh Ambani, MD of Reliance Industries Although not necessarily considered a tech leader, Ambani has announced a multi-pronged plan that involves prevention, mitigation and ongoing support related activities to stem the spread of COVID-19. His foundation has set up India’s first dedicated COVID-19 centre with 100 beds and a fully equipped isolation facility in Lodhivali, Maharashtra. As far as tech is concerned, the company’s telecom arm has increased its data limits for its existing broadband users and waived charges for new ones. With Jio Haptik Technologies, the company’s digital unit has also built a chatbot for free. This chatbot is built for Government functionaries and is called MyGov Corona Helpdesk. The bot is created to answer questions around the pandemic and provide accurate information to curb false news. Outlook As India battles COVID-19, there has been an extended lockdown for the entire country. This pandemic has disrupted all industries. Moreover, many analysts say that these companies might not be able to recover from the impact for many months to come. The pandemic has put the IT sector’s ability to deliver services on-site to test, and that is why the majority of them have been moving towards working remotely. Many believe that this will cause a recession that could be worse than 2008 one if the pandemic is not brought under control fast.","excerpt":"India has been going through a difficult time. COVID-19 has proven to be a bigger problem than what anyone could have anticipated. The pandemic has been spreading at an alarming rate and has impacted everything – from people’s daily lives, to operations in various industries. The outbreak has affected the tech industry as well, with […]","categories":["AI Features"],"tags":[],"author_name":"Sameer Balaganur","publish_date":"2020-03-27T15:00:00","publication_year":"2020","word_count":782,"keywords":["Go","programming_languages:R","AI","innovation","programming_languages:Go","Git","ViT","GAN","R"],"extracted_tech_keywords":["AI","R","Go","Git","GAN","ViT","innovation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/what-indian-tech-leaders-are-saying-about-covid-19-pandemic\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10008399,"title":"How AI &#038; Data Analytics Is Impacting Indian Legal System","content":"In a survey conducted by Gurugram-based BML Munjal University (School of Law) in July 2020, it was found that about 42% of lawyers believed that in the next 3 to 5 years as much as 20% of regular, day-to-day legal works could be performed with technologies such as artificial intelligence. The survey also found that about 94% of law practitioners favoured research and analytics as to the most desirable skills in young lawyers. Earlier this year, Chief Justice of India SA Bobde, in no uncertain terms, underlined that the Indian judiciary must equip itself with incorporating artificial intelligence in its system, especially in dealing with document management and cases of repetitive nature. With more industries and professional sectors embracing AI and data analytics, the legal industry, albeit in a limited way, is no exception. AI & Data Analytics In Law According to the 2020 report of the National Judicial Data Grid, over the last decade, 3.7 million cases were pending across various courts in India, including high courts, district and taluka courts. The snail’s pace at which the judiciary proceedings happen in the country, and the copious amounts of paperwork associated with it sure is the cause behind much inconvenience to the parties involved. Even though the legal sector in India is one of the largest, is highly under-digitalised. Given its very conventional and traditional approach, the field of law in India seems to be very sceptical to adopting new technologies in its workings. Other reasons for hostility towards incorporating technology such as AI include high cost, lack of proper systems and data, privacy and security issues, and ambiguity in the law governing such technologies. The good news, however, is that young lawyers and practitioners are increasingly warming up the idea of incorporating technology, albeit in areas that are clerical and repetitive in nature. Given the efficiency and sorting speed of machines, several AI-based software is being used for reviewing documents to establish its relevance to a given case. AI-assistant ROSS Intelligence, an online legal research tool built upon IBM’s Watson, has been helping lawyers in mining information from tonnes of legal paperwork through its cognitive computing and natural language processing capabilities. Currently, ROSS Intelligence is being used across several law firms around the world, including some prominent names such as USA-based Baker Hostetler, Salazar Jackson, and K&L Gates. To help in realising greater output from the available data to gain a better perspective by analysing patterns of the decision of judges in similar cases, law practitioners are now turning to legal analytics. These strategic insights from legal analytics are believed to give law firms a competitive advantage and yield profits by incorporating machine learning and natural language processing. Artificial intelligence tools can also support law professionals in conducting due diligence in finding and understanding background information by conducting a fact check on the facts and figures. This speeds up the process while also eliminating the possibility of manual inaccuracies. Indian companies in the legal-tech space Several companies are now gradually entering the relatively untouched legal sector considering the largely untapped potential. Gurugram-based LegalKart provides an AI-based app that connects clients with the best lawyers available in their vicinity. The clients also get real-time updates, information sharing and assistance. Reportedly India’s first practice management system for lawyers, LegalKart promises to revolutionise the way law professionals manage their practice. LegalKart’s clientele includes big names such as Ola and ZoomCars. Bangalore-based SpotDraft is an AI-based contract management platform that helps customers draft, review, manage and sign contracts. This AI platform can also analyse legal documents and suggest users on clauses that can be negotiated on. The platform uses AI and ML for text analysis, which is conventionally used in academic work. Founded in 2017, SpotDraft has processed over 6,000,000 contracts as per the company website. An NCR-based startup CaseMine is using AI to make legal research and analysis more in-depth and comprehensive as opposed to a regular search. The CaseIQ software is a virtual research assistant provided by the company that obtains results from legal documents. The software helps in obtaining relevant search results without the need to reformulate case facts into searchable legal propositions. Essentially a case law analytics system, CaseMine “enhances traditional legal research to move beyond mere keywords and retrieve relevant results using entire passages and briefs.” Named after Harry Potter character Professor Albus Dumbledor’s memory reviewer, Mumbai-based startup Pensieve provides an AI-system that understands legal documents. Its flagship product Mitra.ai can understand the context of a search query and is able to provide suitable recommendations. As opposed to conventional search engines, this platform uses machine learning models running on thousands of documents to provide the most relevant information. The company also claims to help prepare defensible arguments. Wrapping Up Sensing the need of the hour, the Indian judiciary is slowly taking to technology in regular proceedings, virtual courts and hearings being the point in case. It is only time when AI and analytics will be actively included in the legal sector. Notably, a group of researchers from IIT Kharagpur, by using machine learning techniques, developed a method to automate legal document reading. The proposition that such technology can replace actual judges and lawyers is too far-fetched as critical cases still require to be approached with a certain sense of judgement, conventional wisdom, and humanity which machines currently are unable to provide. While the legal profession is not immediately threatened by the advent of AI, the legal sector still awaits a major technological overhaul to smoothen the system. Notably, a Supreme Court committee “National Policy and Action Plan for Implementation of Information and Communication Technology in the Indian Judiciary” was formulated in 2005 for greater accessibility. However, the progress has been slow due to resistance from the stakeholders.","excerpt":"In a survey conducted by Gurugram-based BML Munjal University (School of Law) in July 2020, it was found that about 42% of lawyers believed that in the next 3 to 5 years as much as 20% of regular, day-to-day legal works could be performed with technologies such as artificial intelligence. The survey also found that […]","categories":["AI Features"],"tags":["human touch to ai","legal AI","legal-tech"],"author_name":"Shraddha Goled","publish_date":"2020-09-27T10:00:00","publication_year":"2020","word_count":959,"keywords":["Go","artificial intelligence","machine learning","TPU","AI","ML","legal-tech","RAG","Aim","analytics","human touch to ai","R","legal AI"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","analytics","Aim","RAG","TPU","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-ai-data-analytics-is-impacting-indian-legal-system\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10053035,"title":"What Happened in Reinforcement Learning?","content":"One of the most exciting areas in machine learning right now is reinforcement learning. Its application is found in a diverse set of sectors like data processing, robotics, manufacturing, recommender systems, energy, and games, among others. What makes reinforcement learning (RL) different from other kinds of algorithms is that it does not depend on historical data sets. It learns through trial and error like human beings. Understanding its importance, the last few years have seen an accelerated pace in understanding and improving RL. Think of any big name in tech- be it Facebook, Google, DeepMind, Amazon, or Microsoft, they are all investing significant time, money and effort in bringing out innovations in RL. Are Tech Firms Investing More In Reinforcement Learning Research? Reinforcement Learning Goes Beyond Gaming & Robotics, Will Become A Game Changer Top 8 Resources To Learn Self-Supervised Learning A Beginner’s Guide to Sequential Recommendation Systems A guide to self-supervised learning with graph data Robotics Simplified For robots to be useful to mankind, they need to perform a variety of tasks. But, even training for one task using offline reinforcement learning will take a massive amount of time and huge computational expenditure. To work on this issue, Google came out with MT-Opt and Actionable Models. While the first one is a multi-task RL system for automated data collection and multi-task RL training, the latter is a data collection mechanism to collect episodes of various tasks on real robots and demonstrates a successful application of multi-task RL. They also help robots to learn new tasks more quickly. A leader in the reinforcement learning space, DeepMind gave us some unique innovations this year. It released RGB-stacking as a benchmark for vision-based robotic manipulation. Here, DeepMind used reinforcement learning to train a robotic arm to balance and stack objects of different shapes. The diversity of objects used and the number of empirical evaluations performed made this reinforcement learning-based project unique. The learning pipeline was divided into three stages- training in simulation by using an off-the-shelf RL algorithm, training a new policy simulation with only realistic observations, and lastly, collecting data using this policy on real robots and bringing out an improved policy from this. Sequential Learning The implementation of sequential decision processes is crucial for those working in reinforcement learning. In order to simplify such a process, social media giant Facebook (now Meta) came out with “SaLinA” just a month back. It is built as an extension of PyTorch and can work in both supervised and unsupervised situations with compatibility options with multiple CPUs and GPUs. Such a method will see usage in systems where large-scale training use cases are involved. IBM, too, has been active in the reinforcement learning segment. It released the text-based gaming environment called TextWorld Commonsense (TWC) to work on the problem of infusing RL agents with commonsense knowledge. This method was used to train and evaluate RL agents with a specific commonsense knowledge about objects, their attributes, and affordances. It worked on the issue of sequential decision making by introducing several baseline RL agents. Self-supervised Learning In the self-supervised learning area, we saw new methodologies coming out. Google released an approach called Reversibility-Aware RL, which adds a separate reversibility estimation component to the self-supervised RL procedure. Google said this method increases the performance of RL agents on several tasks, including the Sokoban puzzle game. Deep RL in Gaming As reinforcement learning has a significant impact on games, in the middle of 2021, we saw DeepMind training agents playing games without intervention with the help of reinforcement learning mechanisms. Though previous innovations by DeepMind like AlphaZero beat world champion programs in Chess, Shogi and Go, they still trained separately on each game, unable to learn a new one without repeating the RL procedure from the beginning. Through this method, however, the agents were able to react to new conditions with adaptation flexibility to new environments. The core part of this research relied on how deep RL can play a role in training neural networks of the agents. Google has been working on using RL in the gaming domain. In early 2021, it released “Evolving Reinforcement Learning Algorithms”, which showed how to learn analytically interpretable and generalisable RL algorithms by using a graph representation and applying optimisation techniques from the AutoML community. It used Regularized Evolution to evolve a population of the computational graphs over a set of simple training environments. This helped to better RL algorithms in complex environments with visual observations like Atari games. Growing interest in RL With so much happening in the RL space, interest in this area is bound to grow among students and the professional community. To cater to the growing demand, Microsoft organised the Reinforcement Learning (RL) Open Source Fest to introduce students to open source reinforcement learning programs and software development. Researchers from DeepMind teamed up with the University College London (UCL) to offer students a comprehensive introduction to modern reinforcement learning. It intended to give students a detailed understanding of topics like Markov Decision Processes, sample-based learning algorithms, deep reinforcement learning, etc. Reinforcement learning and its advancements still have a long way to go, but there has been major progress in the last couple of years. Its usage can be a game-changer for certain industries. With more and more research coming in RL, we can expect to see major breakthroughs in the near future.","excerpt":"2021 saw innovations in the reinforcement learning space in the robotics, gaming , sequential decision making space amidst growing curiosity among students and professionals.","categories":["AI Features"],"tags":["DeepMind","Facebook","Google","IBM","Reinforcement Learning","reinforcement learning an introduction","students"],"author_name":"Sreejani Bhattacharyya","publish_date":"2021-11-08T17:00:00","publication_year":"2021","word_count":895,"keywords":["Go","self-supervised learning","machine learning","Reinforcement Learning","AI","neural network","PyTorch","ML","recommendation systems","students","GAN","IBM","Google","Facebook","R","reinforcement learning an introduction","DeepMind"],"extracted_tech_keywords":["AI","machine learning","ML","neural network","PyTorch","recommendation systems","R","Go","GAN","self-supervised learning"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/what-happened-in-reinforcement-learning\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10171640,"title":"Google Gemma 3n is What Apple Intelligence Wants to Be","content":"Apple is struggling in the AI battle, and it knows it. But the tech giant isn’t one to give up without a good fight. The Cupertino-based company has introduced a ‘Foundation Models’ framework, which gives developers access to a 3-billion parameter on-device model. The model supports 15 languages, processes both text and images, and is optimised for performance on Apple Silicon. Apple’s 3B parameter language model can handle a wide variety of text-based tasks, including summarising, extracting key information, understanding and improving text, handling brief conversations, and creating original content. It now faces direct competition from Google’s Gemma 3n, which runs on phones, tablets, and laptops. The model supports text, image, and audio inputs, includes advanced video processing capabilities, and operates with a dynamic memory footprint of just 2 to 3 GB, enabled by Google DeepMind’s per-layer embeddings innovation. Its audio features support reliable speech recognition and spoken language translation. The model is also built to interpret combinations of different input types for more complex use cases. Though the two models share similar goals, Gemma 3n is a better fit for developers and users, thanks to its multimodal capabilities and open-source approach. Google collaborated with Qualcomm Technologies, MediaTek and Samsung’s System LSI business to optimise performance in a wide range of hardware. Meet Gemma 3n, a model that runs on as little as 2GB of RAM 🤯 It shares the same architecture as Gemini Nano, and is engineered for incredible performance. We added audio understanding, so now it’s multimodal, fast and lean, and runs on-device (no cloud connection required!) pic.twitter.com\/2FyzJHVGZa— Google AI (@GoogleAI) May 20, 2025 In contrast, Apple Intelligence takes a hybrid approach, combining on-device processing with cloud-based computation. While Apple’s neural processing units (NPUs) handle fast tasks locally, more complex queries often rely on server-side processing, which requires an internet connection. “Apple doesn’t report benchmarks for their AIs, reporting on an ill-documented head-to-head evaluation. But even by their standards, Apple’s latest on-device models are mostly worse than the open Gemma 3-4B from Google or Qwen 3-4B,” said Ethan Mollick, professor at Wharton University, in a post on X. On the LMArena leaderboard for text-based tasks, the Gemma 3n scored 1293 points. In comparison, OpenAI’s o3-mini scored 1329 points, and the o4-mini scored 1379 points. Android vs iOS Ecosystem Lately, Google has been pushing AI deeper into the Android ecosystem, letting smartphone users run large language models locally without needing an internet connection. To support this, the company introduced Edge Gallery, a new Android app that allows users to download and use its AI models directly on their devices. The app is available through the Google AI Edge GitHub repository, and an iOS version is expected soon. “This app is a resource for understanding the practical implementation of the LLM Inference API and the potential of on-device generative AI,” the company said in a blog post. Apple Intelligence for Android is already here! We published is an example app on how to install and run the most capable @GoogleDeepMind Gemma 3n model on Android. 📱 Runs 100% locally & offline, no internet after setup🤖 Supports Multiple Gemma versions🖼️ Multimodal, ask… pic.twitter.com\/poYD6Qu16y— Philipp Schmid (@_philschmid) June 10, 2025 Google has the advantage of the Android ecosystem as well. As of 2025, Android’s global user base ranges from 3.9 billion to 4.5 billion, making up approximately 71 to 72% of smartphone operating systems in use. iOS trails with 1.56 to 1.8 billion users, holding a market share close to 28 to 29%. However, Apple is stepping up its AI game with features that feel genuinely useful. Live Translation helps you chat across languages in real time, even on calls. Image Playground and Genmoji let you create expressive, custom visuals. Visual Intelligence can understand what’s on your screen and take action. And for fitness buffs, Workout Buddy delivers live motivation based on your workout stats. In a recent interview, Apple’s marketing chief, Greg Joswiak, said that Apple’s approach to AI is different from its competitors. Rather than offering a standalone AI app or chatbot, Apple is integrating generative AI quietly and deeply into the operating system. “There’s no app called Apple Intelligence. Sometimes, you’re doing things and you don’t even realise you’re using Apple Intelligence or, you know, AI to do them,” said Joswiak. Apple claims its on-device model outperforms Qwen-2.5-3B across all supported languages, and holds up well against Qwen-3-4B and Gemma-3-4B in English. Meanwhile, its server-based model compares favourably with Llama-4-Scout, though it still lags behind larger models like Qwen-3-235B and GPT-4o. Yet, Siri received barely a mention during WWDC 2025. Addressing this, Apple’s software chief Craig Federighi explained that the company had a two-phase architecture and working demos in place, but the quality and reliability fell short of Apple’s internal standards, particularly when handling unpredictable and open-ended queries. Although parts of the system were functional, Apple chose not to release a product that did not meet its bar. Apple may be playing the long game, but in the fast-moving world of AI, time is a luxury it can’t afford. With Google setting the pace, Apple now faces a future where quiet integration alone may no longer be enough.","excerpt":"Google has been pushing AI deeper into the Android ecosystem, letting smartphone users run large language models locally without needing an internet connection.","categories":["Global Tech"],"tags":["Apple","Google"],"author_name":"Siddharth Jindal","publish_date":"2025-06-11T18:44:29","publication_year":"2025","word_count":859,"keywords":["Go","OpenAI","AI","Apple","GPT-4o","Git","Aim","generative AI","Google","Gemma 3","foundation models","R"],"extracted_tech_keywords":["AI","generative AI","foundation models","GPT-4o","OpenAI","Gemma 3","Aim","R","Go","Git"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/google-gemma-3n-is-what-apple-intelligence-wants-to-be\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10042713,"title":"Are Big Techs The Next Big Banks?","content":"Finance looks very different today. While we continue to have our big safe ‘legacy’ banks, we are also increasingly using other platforms for our financial needs. Some household examples of this are Google Pay and Whatsapp Business. This brings us to the question- is Big tech trying to be the next Big Bank? According to a CB Insights report, Big tech’s investment in fintech companies was around $2.2 billion last year. A couple of elements are driving big tech activities in fintech. Firstly, as Application Programming Interfaces (API) make it simpler to embed banking services into any product, new avenues to fintech possibilities are forged. Adding to this is the growing demand for fintech companies, which raised $22.8 billion in the first quarter of 2021—a 98 percent increase from the previous quarter. The third advantage tech giants entail their economies of scale. Tech giants– Facebook, Apple, Google and Amazon– already have a massive active user base to use for higher adoption for financial services offerings. Finally, tech giants are also privy to consumer user data, which they can analyse to curate financial products that suit these consumers. Today, we take a look at how big tech has contributed to fintech. Digital payments Source: Facebook Pay Facebook aims to consolidate payment strategies across its different apps– Messenger, Instagram and Whatsapp. The social media giant formed a new division– Facebook Financial (F2), to build a payment strategy across Facebook, Instagram, Whatsapp, and Portal. The company has further launched Facebook Pay, a secure payment system for its applications. Facebook Pay allows users to purchase goods and donate money on Facebook or Instagram apps without leaving the app. Whatsapp has, in addition, started a P2P money transfer service in India and Brazil– both countries with the highest Whatsapp userbase. Speaking of India, the country is currently undergoing a digital payments boom to diversify online payment methods, attracting many tech giants. For instance, Amazon is looking to capitalise on a share of India’s digital payments market through Amazon Pay. It entered the Indian fintech market in 2016 with the Amazon Pay e-wallet. Since then, the e-commerce giant has identified additional fintech opportunities in India and has made several advancements by partnering with ICICI Bank to issue credit cards; becoming a part of the Indian government’s payment network through the Amazon Pay UPI; launching services to tap into India’s severely under-tapped insurance industry. Additionally, Amazon Pay has also launched a digital gold investment service that allows users to buy gold for as little as five rupees. Another critical player in the digital payments segment is Apple. It may soon introduce phone-to-phone transactions. Last year, the big tech company acquired Mobeewave, a Canada-based payments startup, for $100 million. Mobeewave allows shoppers to tap their credit cards or smartphones on another phone to process payments. The system works through an app and does not require any hardware beyond a Near Field Communications (NFC) chip, which has been a part of iPhones since 2014. Thus, this merger allows Apple Pay to offer quick payments and transfers without needing another app or media. E-commerce According to Deutsche Bank, the revenue opportunity of launching shops across Facebook and Instagram is estimated to be $30 billion. Facebook shops allow retailers to upload product catalogues and sell them on Facebook’s apps straightforwardly. Customers can also easily browse through Facebook or Instagram business pages or profiles to purchase products and use Facebook Pay to make payments. Instagram’s 30 million amateur creators and around 500 thousand professional ones are a significant force driving this boost. Additionally, Instagram is already the go-to app to connect with these brands and creators, and 90 percent of all Instagram users already follow such business or ‘influencing’ accounts. Google has also amplified its e-commerce initiatives by announcing a partnership with Canadian multinational e-commerce company Shopify. This partnership will allow merchants to feature their products across Google on Search, Maps, Images, Lens and YouTube. Thus, instead of using Google to search for other e-commerce platforms, customers can find and buy products from Google itself. The tool places Google in direct competition with Amazon. Additionally, Google has also launched new e-commerce tools to enhance shopping experiences for its customers. Google’s Shopping Graph is one of them. It is an artificial intelligence (AI) powered tool to compile information from websites, reviews, videos, and product data to bring relevant insights and listings to customers shopping on Google. The feature allows users to find products from screenshots on Google Photos. It does so by analysing the image using Google Lens and returning the search results. Given Google’s status, the tech giant can leverage its customer base for marketing e-commerce tools and services. Big Tech = Big Banks Source: Amazon Back in the day, Amazon.com was a humble online bookstore. Today, many discuss and sometimes fear the rise of the Bank of Amazon. As controversial as it might be, many believe Amazon is building a bank for itself by taking the core components of modern banking— deposits, credit cards, loans, insurance (on Amazon Protect)—and modifying them to suit its merchants and customers. Such diversification ultimately allows Amazon to satisfy its final objective– to increase participation in the Amazon ecosystem. Amazon also lends money to small and medium businesses (SMBs) via Amazon Lending. In July 2020, Amazon partnered with Goldman Sachs’ online consumer bank, Marcus, to issue loans of up to $1 million to USA-based SMBs. The partnership will allow Goldman Sachs to use data from Amazon’s business revenue and tenure to determine loan approvals and improve its lending models. While Amazon does not intend on becoming a public bank anytime soon, its plans to enter into banking looms. Other big tech companies have also made substantial breakthroughs in banking. Google, for one, has entered the $470 billion remittance market through partnerships with Wise and Western Union. This integration will allow Google Pay users in the US to send money to users in Singapore and India. By the end of 2021, Google intends to enable US-based users to send money to more than 200 countries through Western Union and more than 80 countries through Wise. Apple has also launched an Apple Card Family option, allowing two or more people to co-own an Apple Card. This would enable parents to share cards with their children (aged 13 and above) and set spending limits for them. Finally, Facebook’s dabble in cryptocurrency seems to finally be actualising, with its currency Diem (formerly known as Libra) being expected to launch this year. Silvergate Bank will be the exclusive issuer of the Diem USD stablecoin. Thus, we see how big tech companies are breaking into fintech and how many are doing so through partnerships. Such partnerships allow both tech giants and legacy banks to attract loyal customers and will be mutually beneficial. Still, these in-house financial offerings from big tech can pose a significant threat to legacy banks, who might find themselves getting sidelined as finance, commerce and banking become increasingly digital. Of course, only the future will tell us what exactly becomes of the financial system as we know it, but for now- big tech is far from done with fintech.","excerpt":"Finance looks very different today. While we continue to have our big safe ‘legacy’ banks, we are also increasingly using other platforms for our financial needs. Some household examples of this are Google Pay and Whatsapp Business. This brings us to the question- is Big tech trying to be the next Big Bank?  According to […]","categories":["IT Services"],"tags":["Google Pay"],"author_name":"Mita Chaturvedi","publish_date":"2021-07-01T12:00:00","publication_year":"2021","word_count":1190,"keywords":["Go","API","artificial intelligence","AI","Git","RAG","Aim","ViT","Google Pay","R","startup"],"extracted_tech_keywords":["AI","artificial intelligence","Aim","RAG","R","Go","Git","API","ViT","startup"],"url":"https:\/\/analyticsindiamag.com\/it-services\/are-big-techs-the-next-big-banks\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10070460,"title":"RBI is making IT outsourcing tougher for banks, and that’s a good thing","content":"Over the years, the Indian financial institutions have been outsourcing critical IT services to accelerate efficiency. However, this exposes them to significant risks. Recently, in its Draft Master Direction on Outsourcing of IT Services, the Reserve Bank of India (RBI) has issued guidelines for the outsourcing of IT services to protect financial entities in the country from financial, operational and reputational risks. Now, Regulated Entities (REs) will need to have IT outsourcing policies in place and also evaluate their need for outsourcing based on comprehensive assessment of attendant benefits, risks and availability of commensurate processes to manage those risks. Further, REs will also be required to have a robust grievance redressal mechanism among other things. RBI has been tightening regulations on the financial sector recently and has been cracking down on fintechs. Earlier this year, RBI barred Paytm Payments Bank from onboarding new customers, citing ‘material supervisory concerns’. The apex bank even directed Paytm to appoint an IT audit firm to conduct a comprehensive audit of its IT system. Even though REs do not require approval from the central bank for entering into such outsourcing agreements, such arrangements will be subject to inspection from time to time. The apex bank has also asked different stakeholders to present their views in this regard. The final master direction will be issued by the RBI after taking into consideration the feedbacks\/ suggestions. The provisions of these directions will be applicable to: Scheduled commercial banks (excluding regional rural banks) Local area banks Small finance banks Payments banks Primary (urban) co-operative banks having asset size of INR 1000 crore and above Non-banking financial companies in top, upper and middle layers Credit information companies All India financial institutions such as NHB, NABARD, SIDBI, EXIM Bank and NaBFID Purpose Digitalisation has changed the banking landscape tremendously. Now,  more and more customers are now relying on digital channels to avail banking services, which makes it imperative for REs to have operational resilience. In 2021, the RBI banned HDFC Bank from selling new credit cards due to power failures in its primary data centres. Similarly, RBI also banned Mastercard from onboarding new customers as the company was non-compliant with directions on Storage of Payment System Data. These developments show RBI’s intent. The guidelines are being drafted by RBI to ensure REs fulfil their obligations and protect customers from any potential risks. “REs have been extensively leveraging Information Technology (IT) and IT enabled services (ITeS) to support their business models and products and services offered to their customers. REs also outsource substantial portions of their IT activities to third parties. Such reliance on IT\/ ITeS provided by third parties expose the REs to significant risks,” RBI said. Further, the apex bank said REs should ensure that outsourcing arrangements neither diminish its ability to fulfil its obligations to customers nor impede effective supervision by the supervising authority. Relevant for IT services such as: IT infrastructure management Network and security solutions maintenance Application development, maintenance and testing Services and operations related to data centres Cloud computing services Managed security services Application Service Providers (ASPs) including ATM Switch ASPs5 Management of IT infrastructure and technology services associated with payment system ecosystem Why is it a good thing? To stay competitive and increase efficiency, more and more REs tend to outsource IT services. With no proper framework in place, a major disruption at one of these third parties could pose a significant threat towards the financial stability and safety of multiple financial institutions. The REs need to have business continuity and disaster recovery plans in place in case of a major breach or contract termination. The guidelines drafted by the RBI are to mitigate such risk and eliminate any events that could put REs in trouble. Further, the guidelines also mentions the use of cloud infrastructure. In this context, RBI stated that ​​while leveraging cloud services, REs must ensure that outsourcing of IT Services policy addresses the entire lifecycle of data. That is, from generation of the data, its entry into the cloud, until the data is permanently erased\/ deleted. Data privacy and data protection are also important factors to consider. Having robust guidelines in place could help reduce the risk of data breach. Another positive upshot of these new guidelines could be that REs work on building robust IT infrastructure within India rather than outsourcing it to firms based in foreign countries. However, the neobanks, who operate on an outsourced model, might find it hard to adhere to the policies. A global trend The RBI is not the first supervisory body to tighten the rules around IT outsourcing. In November 2020, the Financial Stability Board, a global organisation tasked with devising standards around risk management, published a paper for public consultation on Regulatory and Supervisory Issues Relating to Outsourcing and Third-Party Relationships. In 2019, the European Banking Authority drafted the EBA Guidelines on outsourcing arrangements. The guidelines were published following increasing interest from European and UK regulators on how banks and financial money institutions utilise new fintech solutions and the extent to which they can outsource IT functions and technologies. During the same period, the Monetary Authority of Singapore (MAS), the city-state’s apex bank, also issued guidelines on outsourcing IT services by players in the domestic financial sector. In fact, some of the guidelines drafted by the RBI are similar to those drafted by MAS.","excerpt":"REs outsource substantial portions of their IT activities to third parties.","categories":["IT Services"],"tags":["reserve bank of india"],"author_name":"Pritam Bordoloi","publish_date":"2022-07-05T14:00:00","publication_year":"2022","word_count":892,"keywords":["Go","cloud computing","AI","Git","RAG","ViT","disruption","GAN","llm_models:Bard","R","reserve bank of india"],"extracted_tech_keywords":["AI","RAG","cloud computing","R","Go","Git","GAN","ViT","disruption","llm_models:Bard"],"url":"https:\/\/analyticsindiamag.com\/it-services\/rbi-is-making-it-outsourcing-by-banks-tougher-and-thats-a-good-thing\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10101565,"title":"OpenAI has Just Cracked AGI","content":"Andrej Karpathy, the genius who worked at Tesla, and now OpenAI, says that he is “building a kind of JARVIS at OpenAI”. Undoubtedly, the guy who built Baby Llama and can easily code GPT-5 over the weekend, is probably ready for the task of building autonomous agents, if not AGI, and OpenAI has possibly cracked it. Greg Brockman, co-founder of OpenAI, recently shared the capabilities of the GPT-4 Voice, where the assistant was able to do business negotiations without much supervision. Ironically, it identifies itself as Jarvis. Ammaar Reshi posted a screen recording of him talking to the chatbot, where it calls itself Jarvis, and is ready to assist him with everything. https:\/\/twitter.com\/ammaar\/status\/1711236159987384490 There is no doubt that OpenAI has been hung up on building AGI. Interestingly, the company took another step towards this goal and changed the “core values” listed on its website to add the focus on AGI, something that wasn’t explicitly mentioned on the page before. Earlier, OpenAI’s careers page listed six core values for its employees, like “audacious, thoughtful, impact driven, collaborative”, and so on. But now, the first one is AGI, followed by “intense and scrappy, scale, make something people love, and team spirit”. People, who have been tripping over the fact that the company changed its “core values”, may be unaware of Altman and other co-founders having explicitly stated multiple times that OpenAI is all about AGI. OpenAI Likely to Announce “Jarvis” at DevDay Speculations are rife that at the upcoming OpenAI DevDay conference, the company might just announce their first completely autonomous agent. People call it a “swarm of bees clicking around on the internet”, fundamentally changing the it. This definitely points to the recent Global Illumination acquisition by OpenAI, where it is training AI agents in a gamified simulation. Though the conference is focused on developers, and would possibly include a lot of developer-focused announcements, the release of Jarvis officially looks like in the works. When OpenAI releases Autonomous Agents, it will be like an intelligent swarm of bees clicking around on the internet. Sending emails, negotiating, making products, purchases, fulfilling orders, etc. It will fundamentally change the internet, and this cannot be overstated.— AI Breakfast (@AiBreakfast) October 13, 2023 Taking this forward to autonomous agents rumours, all of the recent developments indicate that OpenAI might just announce something like a Jarvis at its DevDay conference. Earlier, in an interview, Altman had defined AGI as something that could serve as the “equivalent of a median human that you could hire as a co-worker”. Autonomous agents are the next step for LLM-based chatbots. Almost all the companies and researchers realise this. Microsoft recently came up with AutoGEN, a framework that enables building LLM applications using multiple agents that would be able to talk to each other. But as the GitHub repository for AutoGEN says, “AutoGen agents are customisable, conversable, and seamlessly allow human participation,” which means there’s still a need for human input for these models, which questions the autonomous behaviour of the model. Similarly, Google DeepMind recently published a paper ‘How FaR Are Large Language Models From Agents with Theory-of-Mind?​‘ Even Meta’s Shepherd: A Critic for Language Model Generation​, talks about the same autonomous AI agents augmenting and doing tasks all by themselves. Other such papers include SELF: Language-Driven Self-Evolution for Large Language Model and SelfEvolve: A Code Evolution Framework via Large Language Models. But none of these researches have actually been realised in reality yet though. When it comes to OpenAI, they might have just cracked AGI with Jarvis. All roads lead to AGI When it comes to displacement of jobs, Altman had been ready for this all this while. That is why he invested in Worldcoin and wants to provide a universal basic income for people. It would be a world where AI would be doing all our jobs and we would be sitting at home and focusing on building better things. Probably, the next Jarvis. On the other hand, some people are already concerned about the threat of autonomous killer robots, citing Ukraine’s recent AI-based drones that are automatically targeting and attacking targets without human control, calling them “killer robots”. Since these AI-driven drones might be able to perform better than the human-driven ones, people are expecting chaos, apart from just job displacement. Science fiction no more – autonomous killer robots are here“Ukraine is using AI drones that can identify and attack targets without any human control, in the first battlefield use of autonomous weapons or ‘killer robots’”What happens next?“A year from now these drones will… https:\/\/t.co\/hCLqxftY0L pic.twitter.com\/GOI4fYlMDO— AI Notkilleveryoneism Memes ⏸️ (@AISafetyMemes) October 14, 2023 On a funnier note, have you ever got mail from a Nigerian prince stuck in a foreign country asking you to send him money? If you thought these phishing emails were a problem, AI is going to give a major upgrade to it. Now it’s not going to be some guy sending these emails, but an AI agent, posing as a Nigerian prince. Take that!","excerpt":"OpenAI is likely to release autonomous agent Jarvis at DevDay","categories":["Global Tech"],"tags":["Andrej Karpathy"],"author_name":"Mohit Pandey","publish_date":"2023-10-17T13:06:17","publication_year":"2023","word_count":832,"keywords":["Go","Andrej Karpathy","OpenAI","GPT-5","AI","autonomous agents","ML","chatbots","Git","AutoGen","R"],"extracted_tech_keywords":["AI","ML","GPT-5","OpenAI","AutoGen","autonomous agents","chatbots","R","Go","Git"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/openai-has-just-cracked-agi\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10166938,"title":"Apple is Building an AI Agent-Powered Health App: Reports","content":"Apple, the Cupertino giant, is reportedly developing a revamped version of its Health app powered by an AI agent, Bloomberg reported. It is called ‘Project Mulberry’ and features a health coach. The report further indicated that development is progressing at full steam and the new version may launch with iOS 19.4, which could arrive soon or be delayed until next year. The Health app will continue fetching data from connected Apple devices and third-party apps, and the AI coach is designed to use this data to offer recommendations and suggestions for improving the user’s health. Moreover, Apple is training AI agents with data from physicians who collaborate with the company. As per reports, Apple will also bring in external health experts to create content that explains health conditions and offers methods for health improvements. For example, the report noted that if the Health app receives data about poor heart rate trends, a video explaining the risks of heart disease could appear. Furthermore, Apple is also working on features allowing the AI agent to use the camera and offer suggestions regarding the user’s physical activities and workouts. Having said that, Apple’s attempts at AI, particularly with Apple Intelligence, have not been successful so far. The features have suffered from bugs and inaccuracies, leaving users largely dissatisfied. Apple’s health-based AI initiatives signal towards offering users a stronger value proposition against the competition. Recently, reports stated that Apple plans to integrate cameras into a future edition of the Apple Watch for AI features. The report also mentioned that Apple is considering adding cameras to both the standard Series and Ultra models. The standard Series variant of the Apple Watch will include a camera within its display, similar to the front-facing camera on the iPhone. In contrast, the Apple Watch Ultra will have a camera lens positioned on the side of the watch, close to the crown and the button.","excerpt":"The AI coach is designed to use data to offer recommendations and suggestions for improving the user’s health.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","ai in health sector","Apple"],"author_name":"Supreeth Koundinya","publish_date":"2025-03-31T10:43:56","publication_year":"2025","word_count":316,"keywords":["programming_languages:R","AI","emerging_tech:AI agents","Apple","ai in health sector","AI agents","ViT","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","R","ViT","AI agents","programming_languages:R","emerging_tech:AI agents"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/apple-is-building-an-ai-agent-powered-health-app-reports\/","complexity_score":2,"technical_depth":6,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10058461,"title":"How can Federated learning be used for speech emotion recognition?","content":"Federated learning can be considered as an environment where machine learning models and projects learn about the data. Speech emotion recognition is a task that can be used to recognize the emotion type using human speech as data. In this article, we will discuss why and how federated learning for speech emotion recognition is useful and a better option than the other type of learning. The major points to be discussed in part one of the article are listed below. Table of Contents Introduction to federated learningWhat is speech emotion recognition?Federated learning for speech emotion recognitionSystem architecture Introduction to federated learning In the field of machine learning, federated learning can be considered as an approach to make algorithms learned using multiple decentralized devices and data samples. Because of the nature of learning from a decentralized environment, we can also call federated learning collaborative learning. One of the major features of federated learning is that it enables us to use common machine learning without sharing the data in a robust mechanism. This feature also enhances data security and data privacy. Since there is always a need for data privacy and data security in the fields like defence, telecommunications, IoT, and pharmaceutics, this machine learning technique is emerging very rapidly. This learning aims at training machine learning algorithms using the data which is located at different locations without exchanging it with each other. We can think of this learning machine as a process training model by exchanging the parameters in different locations without exchanging the data. The main difference between distributed learning and federated learning is that federated learning aims at learning using heterogeneous data while distributed learning aims at training the models at parallelizing computing settings. We can categorize federated learning into the following three categories: Centralized federated learning: In this type of federated learning, coordination between different data centres during the learning period of the model is done by a centralized server. Also, this server is responsible for orchestrating the different steps of the algorithms.Decentralized federated learning: in this type of federated learning, we make different data centres to be capable of coordinating themselves to obtain the global model. This setting aims to prevent single point failure as any updates in the model are exchanged only between connected data centres without the orchestration of the central server.Heterogeneous federated learning: In this type of federated learning, using the dynamical computation on non-iid data we try to train heterogeneous local models while producing a high-performing global inference model. In this article, we aim to understand how and why we should use Federated learning for speech emotion recognition. More information about federated learning can be found here. In this article, our motive is to find why we should follow the federated learning for speech emotion recognition, so in the next section of the article, we will discuss the introduction to speech emotion recognition. What is speech emotion recognition? For humans, speech is one of the ways to express themselves. There is nothing new for a human to define emotion using speech. Making a machine to understand the emotions of speech can be considered as the speech emotion recognition in machine learning. Talking about the most common features using which a machine can understand the emotions behind the speeches are: Lexical features(the vocabulary used)Visual features (the expressions the speaker makes)    acoustic features (sound properties like pitch, tone, jitter, etc.) If a machine is capable of analyzing any or more of the above-given features it can easily perform speech emotion recognition. Using the lexical features for speech emotion recognition will simply require a transcript of the speech which can be extracted from the speech using technologies like text extraction from speech. Using the visual features will require a video of the conversation where speech has been given. At last, the acoustic features can also be used for speech emotion recognition. One of the most important points about analyzing speech using the acoustic feature is that we are not required to transform the data in any other format and this is why we can perform it easily in real-time when the speech or conversation is taking place. After analysis of the audio data we are required to represent the emotion of the speech that can be done in two ways: Discrete classification: Classifying emotions in discrete labels like anger, happiness, boredom, etc.Dimensional Representation: Using the dimensions such as low to large scale as activation, active to passive scale as dominance for representing the emotions. So if we are choosing to analyze the audio data using the acoustic features we can utilize the following features of audio data or speech data: MFCC (Mel Frequency Cepstral Coefficients)Mel SpectrogramChroma The choice of any of the features from audio data depends on our flexibility. Since there are various projects in which we can find the use of MFCC we can utilize them. Talking About the Mel spectrogram these are plots that include frequency and time in the graph that can be used in frequency analysis and plotting the amplitude of audio on the Mel we can make the frequency as perceived frequency. In the above section, we have discussed what are the features and labeling methods that can be used for speech emotion recognition. We can take an idea of a solution pipeline for speech emotion recognition from the below image. Image source By looking at the above image we can say that the project flow starts with taking a raw audio file as an input that is next passing through the feature extraction block where one or more features from the above-given feature are extracted. After extraction and preprocessing in feature extraction block dimensionally redacted audio file goes through the model block where we can use various ML algorithms like SVM, XGB, CNN-1D(Shallow), and CNN-1D on our 1D data frame and CNN-2D on our 2D-tensor to recognize and classify the emotion of the speech. There are various speech emotion recognition projects available which are mainly using the following open-source audio data: Toronto emotional speech set (TESS)SAVEE (Surrey Audio-Visual Expressed Emotion)Ryerson Audio-Visual Database of Emotional Speech and Song (RAVDESS)CREMA-D is an audio-visual data set for emotion recognition We can freely use this dataset for our speech emotion recognition dataset. Till now in this article we have seen what federated learning is, what is speech emotion recognition and how can we manage to perform it using audio files. In conclusion, we can say that federated learning is a way to perform collaborative machine learning, and speech emotion recognition is a process of classifying the audio or speech data in the context of emotion. In the next section of the article, we will discuss why we use federated learning in speech emotion recognition. Federated learning for speech emotion recognition By looking at the above portions of the article we can say that speech emotion recognition (SER) is a rapidly emerging area of machine learning projects and research. When using a perfect system we can recognize human emotion using machines. We can understand that speech data is also one of the most confidential data and transferring speech data from end devices to the models and any other devices requires a lot of security. Using speech data we can extract a lot of information about speakers like age, identity, language, and gender, Which can be a confidential part. In the federated learning part of the article, we were more focused on the data privacy part. By understanding federated learning we can say that a federated learning environment can provide great control over user data privacy and also high-performance modelling experience.  If performing modelling in such an environment it can be incorporating privacy with distributed training and aggregation across a population of client devices. We can find out many projects on the SER where the performance of the system is quite appreciable but when considering the data privacy we can not expect this from them. For which we are required to have environments like federated learning. We can utilize federated learning for speech emotion recognition. For example, we can train CNN and RNN classifiers in federated learning and can activate a high-performance level. System architecture In this section of the article, we will discuss a proposed system architecture that can utilize the federated learning environment for speech emotion architecture. Using this link, we can find out an architecture for the motive of the article. The below image is a representation of the system architecture of federated learning in a distributed network of clients with SER applications. (Source) This architecture is based on the heterogeneous federated learning for SER. in such an architecture a client or data centre can be any device and computation on the data doesn’t require sharing the data with the main server. In a three-step protocol, we can perform this collaborative learning as, All the data centres or clients or devices require to download the model and its updates that are available on the main server.The downloaded model is used by participants to compute the local data.Updated weights from all the devices are uploaded on the main server where aggregation of these uploads updates the global model in the main server. Repetition of these protocols makes the model reach the convergence of certain criteria. In the case of SER, this architecture is useful because the client needs only to share updated weights with the main server instead of speech data which reduces the chances of data leakage. Since every client is producing weights they can also participate in approximating the global objective function and stochastic gradient descent can be used for the optimization of aggregated weights. Final words In this article, we have discussed federated learning and speech emotion recognition. Along with that, we have discussed why we should use speech emotion recognition in a federated learning environment. We also discussed the architecture which can be used in federated learning for speech emotion recognition.","excerpt":"Using speech data we can extract a lot of information about speakers like age, identity, language, and gender, Which can be a confidential part. federated learning environment enhances data security and data privacy.","categories":["AI Trends"],"tags":["AI (Artificial Intelligence)","Data Science","Data Scientist","Deep Learning","federated learning","federated learning new","Google","Machine Learning","Python","speech","Speech Analytics","Speech Recognition","speech recognition algorithm","speech synthesis"],"author_name":"Yugesh Verma","publish_date":"2022-01-16T11:00:00","publication_year":"2022","word_count":1652,"keywords":["API","speech","Speech Recognition","Chroma","CNN","R","speech recognition algorithm","Data Science","Go","machine learning","Speech Analytics","AI","ML","Machine Learning","federated learning new","federated learning","speech synthesis","Python","Aim","Google","Deep Learning","Data Scientist","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","machine learning","ML","Aim","federated learning","Chroma","R","Go","API","CNN"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/how-can-federated-learning-be-used-for-speech-emotion-recognition\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":27300,"title":"How To Get The Best Out Of Your Conversion Rate Optimisation Programme","content":"The conversion optimisation program might seem like an effortless and a smooth mechanism. However, in reality, businesses have realised how challenging the process is. This highlights the scope of work for the owners to optimise the program by creating customer-centric experiences and maximising revenue for their companies. Nonetheless, the most common question is, “What actions can help businesses achieve outstanding results in optimising the experimentation program and make it more effective, data-driven and ROI focused?” Here are a few recommendations: Audit The Current CRO Process And Performance Today, businesses have to clearly define the objective behind individual tests and day to day analysis to look at the bigger picture of the optimisation program in its entirety and establish key performance indicators to assess the current CRO process around Velocity, Quality and Business Impact. Velocity: Evaluate how many test ideas get added to the optimisation roadmap every month and how many ideas convert to a live test every month. This will give some context to the gap. For instance, if somebody is short of ideas or if all the requests have been accommodated without prioritising or selecting ideas that have high potential. Quality: In continuation to the previous point, once a hypothesis is selected with due diligence, how many of them reach a statistical significance? A VWO research shows 14% win-rate for all the tests that went live for their clients. This could be an external benchmark to compare and optimise. Win here is not the result that shows variation as a winner, but all the experiments reached a statistically significant level and provided some learning. Business Impact: Measure average conversion lift for all statistically significant tests in terms of revenue per visitor and other KPIs. Extrapolate the winner impact in terms of revenue but do consider factors like seasonality. Ultimately, extracting the ROI from the optimisation program would be noteworthy. Ensure KPIs, Strategies And Tactics Are Aligned Identify business objectives and related metrics that represent the same and compare them to the goals set for the year. Understand business priorities, challenges, including online revenue goals, paid channels, funnels, etc. Discuss with the sales and marketing team to discuss the buyer’s journey, persona, and content strategy across persona and sales stages. This information would help create a goal tree which would inspire us to experiment a new tactic the moment we see some issue from analytics and vice versa. Prioritise KPIs that are closer to revenue, followed by metrics that align with cost saving like CAC, CPL. Leverage Consumer Psychology And Research On Cognitive Biases While brainstorming the optimisation tactics that could drive the KPI and Strategy defined in the goal map, a great source of enrichment would be to align it with cognitive biases and persuasion principles. This will help you connect with the prospect and influence their choices and decision-making. Additionally, it would provide more context and story behind the final test results. Being thoughtful about cognitive biases while crafting a new hypothesis helps to gain an undue advantage to anticipate user responses. This might sound like a felony, but as long as the offering has a real value, this becomes a legitimate tool to start a conversation and move the prospect deeper into funnel stages. Enrich And Validate Ideas Roadmap With Multiple Sources Experimenting velocity is at the top of the funnel metrics that focus on quantity and to improve that one needs to acquire a sense of how we are sourcing new test ideas and building optimisation roadmaps. Include as many data sources as possible to validate the hypothesis as well as launching new ideas that can solve the customer’s objections and challenges, such as issues identified from User testing, Mouse tracking, Heatmap, user recording, online survey, user feedback, and online chat history, etc. Create A Prioritisation Framework To Ensure The Best Ideas Bubble Up The traditional approach of prioritisation like PIE (potential, impact, ease), ICE (impact, cost, effort) have been under debate. So it is a good idea to have a hybrid approach and assign a score to the idea based on specific and measurable parameters. A person can build his\/her own scoring framework. To begin with, one can start with some of the scoring factors that are self-explanatory and move to more behavioural factors; after reaching a certain maturity in understanding the impact of cognitive biases, user motivation, etc. Don’t Stop Testing Just Because You Saw 95% Significance One has to be very cautious to not get trapped by false positive results. People have to be certain that there is a lot of noise in the data collected in the initial days. To avoid this situation, it would be a good idea to get a sense of how many visitors would be required for each variation. Using a free tool like Optimizely Sample Size calculator will give an idea of the overall traffic required and the number of days to run the test before declaring a winner or learning from a statistically significant result. Experimentation And Personalisation Are BFFs: The most significant piece in personalisation puzzle is first, second and third party data about the users that can form a meaningful segment. The segment could be based on gender, age, browser, device, campaign source, industry, company size, location, day parting, a new visitor, historical transaction etc. But the new experience tied with the segments needs to be validated in terms of business outcomes otherwise it is as good as the test hypothesis. A lot of legwork that was done in building a robust experimentation program like goal tree creation, goal mapping, hypothesis prioritisation, sample size calculation, would also enrich personalisation strategy and roadmap for the website. Create A Knowledge Management Repository For All Significant Results The best part of an optimisation program is that from all statistically significant results someone either wins a lift or at least learns from the experience but never fails. It is imperative to document these golden nuggets in a standard format so that stakeholder from marketing and product management teams can anytime refer and apply to new sections, similar pages, application or website revamp. Following is the step by step structure of the entire experimentation process: Hence, current businesses need to do a lot beyond experimentation and personalisation to achieve actionable results and transform ideas to get the best solution out of the CRO program including the challenges they face and the ultimate solution they launch.","excerpt":"The conversion optimisation program might seem like an effortless and a smooth mechanism. However, in reality, businesses have realised how challenging the process is. This highlights the scope of work for the owners to optimise the program by creating customer-centric experiences and maximising revenue for their companies. Nonetheless, the most common question is, “What actions […]","categories":["AI Features"],"tags":["Big Data"],"author_name":"Arpit Srivastava","publish_date":"2018-08-15T00:01:25","publication_year":"2018","word_count":1066,"keywords":["Go","programming_languages:R","AI","data-driven","programming_languages:Go","Git","RAG","analytics","Big Data","R"],"extracted_tech_keywords":["AI","analytics","RAG","R","Go","Git","data-driven","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-to-get-the-best-out-of-your-conversion-rate-optimisation-programme\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10059600,"title":"How to build healthcare predictive models using PyHealth?","content":"Machine learning has been applied to many health-related tasks, such as the development of new medical treatments, the management of patient data and records, and the treatment of chronic diseases. To achieve success in those SOTA applications, we must rely on the time-consuming technique of model building evaluation. To alleviate this load, Yue Zhao et al have proposed a PyHealth, a Python-based toolbox. As the name implies, this toolbox contains a variety of ML models and architecture algorithms for working with medical data. In this article, we will go through this model to understand its working and application. Below are the major points that we are going to discuss in this article. Table of contents Machine learning in healthcareHow can PyHealth help in healthcare?Working of PyHealthPyHealth for model building Let’s first discuss the use case of machine learning in the healthcare industry. Machine learning in healthcare Machine learning is being used in a variety of healthcare settings, from case management of common chronic conditions to leveraging patient health data in conjunction with environmental factors such as pollution exposure and weather. Machine learning technology can assist healthcare practitioners in developing accurate medication treatments tailored to individual features by crunching enormous amounts of data. The following are some examples of applications that can be addressed in this segment: Disease detection The ability to swiftly and properly diagnose diseases is one of the most critical aspects of a successful healthcare organization. In high-need areas like cancer diagnosis and therapy, where hundreds of drugs are now in clinical trials, scientists and computationalists are entering the mix. One method combines cognitive computing with genetic tumour sequencing, while another makes use of machine learning to provide diagnosis and treatment in a range of fields, including oncology. Diagnosing using image Medical imaging, and its ability to provide a complete picture of an illness, is another important aspect in diagnosing an illness. Deep learning is becoming more accessible as data sources become more diverse, and it may be used in the diagnostic process, therefore it is becoming increasingly important. Although these machine learning applications are frequently correct, they have some limitations in that they cannot explain how they came to their conclusions. Drug discovery ML has the potential to identify new medications with significant economic benefits for pharmaceutical companies, hospitals, and patients. Some of the world’s largest technology companies, like IBM and Google, have developed ML systems to help patients find new treatment options. Precision medicine is a significant phrase in this area since it entails understanding mechanisms underlying complex disorders and developing alternative therapeutic pathways. Surgical tools Because of the high-risk nature of surgeries, we will always need human assistance, but machine learning has proved extremely helpful in the robotic surgery sector. The da Vinci robot, which allows surgeons to operate robotic arms in order to do surgery with great detail and in confined areas, is one of the most popular breakthroughs in the profession. These hands are generally more accurate and steady than human hands. There are additional instruments that employ computer vision and machine learning to determine the distances between various body parts so that surgery can be performed properly. How PyHealth can help in healthcare? Health data is typically noisy, complicated, and heterogeneous, resulting in a diverse set of healthcare modelling issues. For instance, health risk prediction is based on sequential patient data, disease diagnosis based on medical images, and risk detection based on continuous physiological signals. Electroencephalogram (EEG) or electrocardiogram (ECG), for example, and multimodal clinical notes (e.g., text and images). Despite their importance in healthcare research and clinical decision making, the complexity and variability of health data and tasks need the long-overdue development of a specialized ML system for benchmarking predictive health models. PyHealth is made up of three modules: data preprocessing, predictive modelling, and assessment. Both computer scientists and healthcare data scientists are PyHealth’s target consumers. They can run complicated machine learning processes on healthcare datasets in less than 10 lines of code using PyHealth. The data preprocessing module converts complicated healthcare datasets such as longitudinal electronic health records, medical pictures, continuous signals (e.g., electrocardiograms), and clinical notes into machine learning-friendly formats. The predictive modelling module offers over 30 machine learning models, including known ensemble trees and deep neural network-based approaches, using a uniform yet flexible API geared for both researchers and practitioners. The evaluation module includes a number of evaluation methodologies (for example, cross-validation and train-validation-test split) as well as prediction model metrics. There are five distinct advantages to using PyHealth. For starters, it contains more than 30 cutting-edge predictive health algorithms, including both traditional techniques like XGBoost and more recent deep learning architectures like autoencoders, convolutional based, and adversarial based models. Second, PyHealth has a broad scope and includes models for a variety of data types, including sequence, image, physiological signal, and unstructured text data. Third, for clarity and ease of use, PyHealth includes a unified API, detailed documentation, and interactive examples for all algorithms—complex deep learning models can be implemented in less than ten lines of code. Fourth, unit testing with cross-platform, continuous integration, code coverage, and code maintainability checks are performed on most models in PyHealth. Finally, for efficiency and scalability, parallelization is enabled in select modules (data preprocessing), as well as fast GPU computation for deep learning models via PyTorch. Working of PyHealth PyHealth is a Python 3 application that uses NumPy, scipy, scikit-learn, and PyTorch. As shown in the diagram below, PyHealth consists of three major modules: First is the data preprocessing module can validate and convert user input into a format that learning models can understand; Second is the predictive modelling module is made up of a collection of models organized by input data type into sequences, images, EEG, and text.  For each data type, a set of dedicated learning models has been implemented, and the third is the evaluation module can automatically infer the task type, such as multi-classification, and conduct a comprehensive evaluation by task type. Source Most learning models share the same interface and are inspired by the scikit-API learn to design and general deep learning design: I fit learns the weights and saves the necessary statistics from the train and validation data; load model chooses the model with the best validation accuracy, and inference predicts the incoming test data. For quick data and model exploration, the framework includes a library of helper and utility functions (check parameter, label check, and partition estimators). For example, a label check can check the data label and infer the task type, such as binary classification or multi-classification, automatically. PyHealth for model building Now below we’ll discuss how we can leverage the API of this framework. First, we need to install the package by using pip. ! pip install pyhealth Next, we can load the data from the repository itself. For that, we need to clone the repository. After cloning the repository inside the datasets folder there is a variety of datasets like sequenced based, image-based, etc. We are using the mimic dataset and it is in the zip form we need to unzip it. Below is the snippet clone repository, and unzip the data. ! git clone https:\/\/github.com\/yzhao062\/PyHealth.git ! unzip \/content\/PyHealth\/datasets\/mimic.zip The unzipped file is saved in the current working directory with the name of the folder as a mimic. Next to use this dataset we need to load the sequence data generator function which serves as functionality to prepare the dataset for experimentation. from pyhealth.data.expdata_generator import sequencedata as expdata_generator # initialize the dataset # unique id for dataset expdata_id = '2020.0811.data.phenotyping.test.v2' cur_dataset = expdata_generator(expdata_id=expdata_id) cur_dataset.get_exp_data(sel_task='phenotyping', data_root='\/content\/mimic') cur_dataset.load_exp_data() Now we have loaded the dataset. Now we can do further modelling as below. # load and fit the model from pyhealth.models.sequence.embedgru import EmbedGRU # unique id for model expmodel_id = '2020.0811.model.phenotyping.test.v2' clf = EmbedGRU(expmodel_id=expmodel_id, n_batchsize=5, use_gpu=False, n_epoch=100) # fit model clf.fit(cur_dataset.train, cur_dataset.valid) Here is the fitment result. Final words Through this article, we have discussed how machine learning can be used in the healthcare industry by observing the various applications. As this domain is being quite vast and N number application, we have discussed a Python-based toolbox that is designed to build a predictive modelling approach by using various deep learning techniques such as LSTM, GRU for sequence data, and CNN for image-based data. References PyHealth paperPyHealth repositoryPyHealth DocumentationLink for above codes","excerpt":"PyHealth is a Python-based toolbox. As the name implies, this toolbox contains a variety of ML models and architecture algorithms for working with medical data and modeling.","categories":["AI Trends"],"tags":["AI Tool","Data Science","Machine Learning","Python"],"author_name":"Vijaysinh Lendave","publish_date":"2022-02-02T11:00:00","publication_year":"2022","word_count":1387,"keywords":["scikit-learn","NumPy","machine learning","AI","neural network","PyTorch","ML","Machine Learning","computer vision","Python","deep learning","XGBoost","Data Science","AI Tool"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","neural network","computer vision","PyTorch","scikit-learn","XGBoost","NumPy"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/how-to-build-healthcare-predictive-models-using-pyhealth\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":52048,"title":"Can You Be The Owner Of Your Data?","content":"Data is now the new currency for the world, and keeping it safe is crucial for everybody. The rise of the data economy raises one serious question of ‘who owns your data?’ There have been several debates across the world over the ownership rights in data. Policymakers in Europe and Canada are aiming to create a unique framework for the ownership rights that will be treated with vital consideration. Additionally, Europe’s new General Data Protection Regulation and Indian new Data Protection Bill also appears to vest in certain strict amendments to provide a more robust protection for their data. That being said, the ownership of such data can provide businesses with immense power over the users. Currently, these data are owned by a set of Internet giants, who are ruling the world and influence users to benefit their business. The Grey Area Surrounding Data Ownership The process of governing data which provides organisations or individuals the legal ownership and the responsibility of the data produced by one is known as data ownership. According to the concept, the owner of the data holds the rights to create, edit, modify, package, share or restrict the possessed data. The owner posses the copyrights of the data and therefore are capable of taking any legal action. Implicit in having control over the data also provides the owner with the ability to share or surrender the currency to any third party. With such integrity, every data holds something important — whether it is a general profile of a user or some sort of sensitive behavioural data. These data stays on the internet under unknown servers without any security, easily accessible for anybody to use it for their benefit. At times, hackers or any sort of malicious entities could make use of this data illegally causing damage on a varied scale — from identity theft to financial wreck. With the advent of such a colossal wave of digitisation, unprotected data can cause chaos among individuals, businesses or even countries, which can lead to major unpredictable catastrophes. And therefore, the need to know, who owns your data, and how to protect it is important. Is It Possible To Have Control Over Your Own Data? Every individual on the Internet contributes to producing data, which is owned by companies. The problem arises when none of them provide any guarantee over the safety of the data. The laws to protect these data aren’t stringent enough to secure the vulnerability of online. Thus, the requirement for an external solution is immense. And that’s where Singapore-based, Houm Technology comes into the picture. Houm Technology, founded by Priya Goenka and Bijai K Jayarajan, was aimed at providing complete data protection and privacy to users giving them the ability to 100% own their data on the Internet. This company, which is the first of its kind in the market, makes the users legally in-charge of their data providing them with a ‘private home’ on the Internet. This technology, with an initial storage space of 100GB, is a direct competition to other storage platforms like Dropbox and Google Drive, where once data stored, it becomes a legal property of these companies and users are only retaining simple access to it. Whereas, this startup claims that they provide a unique domain to the users with the servers are legally owned by them giving space on the Internet with complete access. It deploys strict data security framework and encryption software. Bijai K Jayarajan explained in a video that this B2C product allows individuals and businesses to own digital property. The users can buy a unique domain name and can use it for storing any personal data, financial documents, company secrets, or any other digital asset, without the fear of losing them, considering the user is the only one with access to it. Primarily, the company provides users with a private place on the internet. The product — Houm is targeted at every internet user whether be it an individual, startups, or small businesses that require storage space on the internet on an everyday basis. This beta version technology, with development centre in India, has more than 15,000 global customers with over 10,000 users only from India and aims to expand its business to 50 more countries in the coming year, like in New Zealand, the US, and Europe. Bijay also added that this private space especially helps users who aren’t skilled with software coding or any sort of engineering background, and offers this digital home with as little as ₹200 to ₹300 per year. Outlook Although it has become very easy nowadays to access any information, millions of people are constantly losing their data. With data security been the utmost concern for businesses as well as individuals, companies like Houm Technology acts as a trendsetter. Such technology will provide the utmost security to the data owners by deploying state-of-the-art data security software and encryption framework, which only allows the owner of the data to access it. The two core principles making this innovation a revolution in the data industry is — the Ownership and Privacy. The two things that are currently lacking in other storage spaces, such as Google Drive, Dropbox, or any other cloud storage options. Similar to owning a private land for your real home, this innovative concept aims to provide customers with private and legal ownership of their own data without the fear of losing it. Such a solution will be the future of authentic digital privacy which has been a tender subject for businesses and users.","excerpt":"Data is now the new currency for the world, and keeping it safe is crucial for everybody. The rise of the data economy raises one serious question of ‘who owns your data?’ There have been several debates across the world over the ownership rights in data. Policymakers in Europe and Canada are aiming to create […]","categories":["AI Features"],"tags":["Data Mining","Data Privacy","data privacy India","privacy law","privacy regulations"],"author_name":"Sejuti Das","publish_date":"2019-12-16T17:30:00","publication_year":"2019","word_count":924,"keywords":["Go","AWS","AI","Data Mining","innovation","privacy regulations","Git","data privacy India","RAG","privacy law","Aim","Data Privacy","ViT","GAN","R"],"extracted_tech_keywords":["AI","Aim","RAG","AWS","R","Go","Git","GAN","ViT","innovation"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/can-you-be-the-owner-of-your-data\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":61088,"title":"SC Verdict On Lifting Cryptocurrency Ban In India May Be Misinterpreted, And We May See The Ban Reinstated","content":"On April 5, 2018, Reserve Bank of India had issued a few advisory guidelines concerning cryptocurrency activities in India under a circular titled ‘Statement on Developmental and Regulatory Policies’. Paragraph 13 of the circular asked entities governed by RBI not to deal with – or give services to – any person or business organizations dealing with or transacting in virtual currencies. Additionally, it also asked these entities to end such ties – if any. As per RBI, the circular was issued in the public interest. This circular was challenged by the chief petitioner – Internet And Mobile Association Of India – in the court of law. On March 4, 2020, the Supreme Court of India delivered a historical judgment. As per popular interpretation of the verdict, it signalled the legitimacy of virtual currencies in India; that is, the Supreme court had lifted the ban on virtual currencies, and thus, trading in virtual currencies was now legal. The petitioners had been entitled to supersede, and the challenged circular issued on April 6, 2018, was subject to be taken down, as per the Supreme Court. Though the Supreme Court of India upheld the plea for striking down the applicability of the circular, the order pronounced by the bench consisting of Justice Rohinton Fali Nariman, Aniruddha Bose and V. Ramasubramanian, may need careful evaluation for better understanding of the judgement. Supreme Court Affirmed That RBI Has All The Power To Regulate Cryptocurrencies The arguments in support of petitioners were on Article 19(1) (g). The denial of banking access to a profession not prohibited under the Indian law was deemed a violation of Article 19(1) (g) of the Constitution of India (which provides the right to practice any legal profession). The petitioners also argued that the power contained in the circular lied outside the powers of the RBI, but the Apex Court negated that argument. The Supreme Court held that anything that may act a threat to or have an impact on the financial system of India should be regulated or prohibited by RBI, despite the said activity not constituting part of the credit system or payment system of the country. In its judgement, the court observed, “It is no doubt true that the Reserve Bank Of India has pervasive powers not only in view of the statutory design but also in view of the special status and role that it possesses in the economy of India. These powers can be applied both in the form of preventive as well as curative measures.” “The court was convinced about wide powers of RBI and issuance of the circulars as preventive measures for the betterment of Indian financial scenario, but as the circular could not pass the test of proportionality, the circulars were smacked down. So, it should not be seen as the Supreme Court has lifted the ban on cryptocurrency in India, or that cryptocurrency trading is official in India as many of us are construing this decision,” said Advocate Dr Mahendra Limaye, who heads cyber law firm- Mahendra Limaye Associates. Where Is The Mainstream Interpretation of The Verdict Wrong? The Supreme court stated RBI did not show any empirical data highlighting the damage caused by cryptocurrency exchanges on the entities regulated by RBI, which is a significant reason that petitioners were able to win. Given that official ban on cryptocurrency still not exist India, RBI’s ban on banking support for crypto firms remained unjustified on the grounds of proportionality. “The availability of power is distinct from the manner and extent to which it can be exercised by RBI. To test the proportionality of banking ban, it required RBI to present at least some semblance of any damage endured by its regulated entities. But there is none,” the Supreme Court stated. So, the overturn of the circular does not mean cryptocurrencies are legal in India or that crypto exchanges will be permanently allowed to function, according to experts. Given RBI will further challenge the verdict to prove the alleged risk that cryptocurrencies pose to the banking system, the banking ban could be reinstated later. Plus, we know that an Inter-Ministerial Committee proposed in February 2019 a blanket ban on cryptocurrencies. Known as “Banning of Cryptocurrency and Regulation of Official Digital Currency Act”, the draft bill is yet to be presented in front of the legislature. If passed, it could make buying, selling, mining, and even holding of cryptocurrency a punishable offence. So, have we interpreted the recent verdict by the Supreme Court wrongly? Dr Limaye says, “In my views, the mainstream interpretation of the verdict is wrong. The petitioners received the benefit of doubt and lassitude from government’s part also played an imperative role in tiling the balance in favour of petitioners. The Apex Court has accepted the powers of RBI to issue circulars in Public Interest. There was no blanket order banning Virtual Currency and diametrically opposite views by the Central government regarding virtual currencies, and it let down the populous move of RBI banning VC exchanges from banking exposures”. “What is essential to note, is that all petitions are filed against the Reserve Bank Of India, and not the Finance Ministry draft ban bill. The verdict remains only short-term relief as the verdict against the RBI does not impact activities on the policy level,” also wrote Tanvi Ratna, a technology consultant and CEO of Policy 4.0 in her blog. The Aftermath The verdict had been welcomed and celebrated by professionals in the crypto industry—multiple exchanges like Unocoin, Wazirx and CoinDCX started INR deposit services soon after. The announcement also was followed by multiple investment announcements in cryptocurrency-related startups. This included Binance, Aeternity and HashCash investing in the country’s blockchain and cryptocurrency economy in 2020. The cryptocurrency ecosystem in India saw a revival of fiat liquidity and resurgence of fiat-based trading at exchanges and as well as investments in startups. But, is this festive mood going to be a short-lived affair if “Banning of Cryptocurrency and Regulation of Official Digital Currency Act” is passed? “The verdict of the Supreme Court solely addresses the Reserve Bank of India circular. The Supreme Court is very unlikely to issue any action against the Finance Ministry, and impact their view on the subject,” according to Tanvi Ratna. Experts believe the Supreme Court seemingly gave a verdict in favour of the cryptocurrency industry as there is no such law yet in India which bans cutting banking support for exchanges. This means the judgment would not hold once there is such anti-crypto regulation is in place. “In the entire judgement, the Supreme Court never uttered a single word about legitimacy or genuineness of virtual currencies or about exchanges trading such virtual currencies. But SC only decided that the activities of petitioner exchanges, trading in virtual currency were not declared unlawful. Hence, their bank accounts could not be debit frozen by the banks citing the challenged RBI Circular,” said Dr Mahendra Limaye. Also Read: How Lifting Crypto Ban In India Will Accelerate Jobs And Blockchain Startups","excerpt":"On April 5, 2018, Reserve Bank of India had issued a few advisory guidelines concerning cryptocurrency activities in India under a circular titled ‘Statement on Developmental and Regulatory Policies’. Paragraph 13 of the circular asked entities governed by RBI not to deal with – or give services to – any person or business organizations dealing […]","categories":["AI Features"],"tags":["Bitcoin","crypto india","Cryptocurrency","cryptocurrency bitcoin","cryptocurrency ethereum","cryptocurrency trading"],"author_name":"Vishal Chawla","publish_date":"2020-04-07T13:00:00","publication_year":"2020","word_count":1165,"keywords":["cryptocurrency trading","Go","Bitcoin","programming_languages:R","Cryptocurrency","AI","crypto india","R","programming_languages:Go","Git","RAG","ViT","cryptocurrency bitcoin","GAN","cryptocurrency ethereum","startup"],"extracted_tech_keywords":["AI","RAG","R","Go","Git","GAN","ViT","startup","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/cryptocurrency-ban-india-verdict\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":37296,"title":"10 Model Evaluation Techniques Every Machine Learning Enthusiast Must Know","content":"Model evaluation plays a crucial role while developing a predictive machine learning model. Building just a predictive model without checking does not count as a fit model but a model which gives maximum accuracy surely does count a good one. For this, you need to check on the metrics and make improvements accordingly until you get your desired accuracy rate. In this article, we jot down 10 important model evaluation techniques that a machine learning enthusiast must know. 1| Chi-Square The χ2 test is a method which is used to test the hypothesis between two or more groups in order to check the independence between the two variables. It is basically used to analyse the categorical data and evaluate Tests of Independence when using a bivariate table. Some examples of Chi-Square tests are Fisher’s exact test, Binomial test, etc. The formula for calculating a Chi-Square statistic is given as Where O represents the observed frequency, E represents the expected frequency. 2| Confusion Matrix The confusion matrix is also known as Error matrix and is represented by a table which describes the performance of a classification model on a set of test data in machine learning. In the above table, Class 1 is depicted as the positive table and Class 2 is depicted as the negative table. It is a two-dimensional matrix where each row represents the instances in predictive class while each column represents the instances in the actual class or you put the values in the other way. Here, TP (True Positive) means the observation is positive and is predicted as positive, FP (False Positive) means observation is positive but is predicted as negative, TN (True Negative) means the observation is negative and is predicted as negative and FN (False Negative) means the observation is negative but it is predicted as positive. 3| Concordant-Discordant Ratio In a pair of cases when one case is higher on both the variables than the other cases, it is known as a concordant pair. On the other hand, in a pair of cases where one case is higher on one variable than the other case but lower on the other variable, it is known as a discordant pair. Suppose, there are a pair of observations (Xa, Ya) and (Xb, Yb) Then, the pair is concordant if Xa>Xb and Ya>Yb or Xa<Xb and Ya<Yb And the pair is discordant if Xa>Xb and Ya<Yb or Xa<Xb and Ya>Yb. 4| Confidence Interval Confidence Interval or CI is the range of values which is required to meet a certain confidence level in order to estimate the features of the total population. In the domain of machine learning, Confidence Intervals basically consist of a range of potential values of an unknown population parameter and the factors which are affecting the width of the confidence interval are the confidence level, size as well as variability of the sample. 5| Gini Co-efficient The Gini coefficient or Gini Index is a popular metric for imbalanced class values. It is a statistical measure of distribution developed by the Italian statistician Corrado Gini in 1912. The coefficient ranges from 0 to 1 where 0 represent perfect equality and 1 represents perfect inequality. Here, if the value of an index is higher, then the data will be more dispersed. 6| Gain and Lift Chart This method is generally used to evaluate the performance of the classification model in machine learning and is calculated as the ratio between the results obtained with and without the model. Here, the gain is defined as the ratio of the cumulative number of targets to the total number of targets in the entire dataset and lift is defined as for how many times the model is better than the random choice of cases. 7| Kolmogorov-Smirnov Chart This non-parametric statistical test measures the performance of classification models where it is defined as the measure of the degree of separation between the positive and negative distributions. The KS test is generally used to compare the equality of a single sample with another. 8| Predictive Power Predictive Power is a synthetic metric which satisfies interesting properties like it is always been 0 and 1 where 0 represents that the feature subset has no predictive power and 1 represents that the feature subset has maximum predictive power.and is used to select a good subset of features in any machine learning project. 9| AUC-ROC Curve ROC or Receiver Operating Characteristics Curve is one of the most popular evaluation metrics for checking the performance of a classification model. The curve plots two parameters, True Positive Rate (TPR) and False Positive Rate (FPR). Area Under ROC curve is basically used as a measure of the quality of a classification model. Hence, the AUC-ROC curve is the performance measurement for the classification problem at various threshold settings. The True Positive Rate or Recall is defined as The False Positive Rate is defined as 10| Root Mean Square Error Root Mean Squared Erro or RMSE is defined as the measure of the differences between the values predicted by a model and the values actually observed. It is basically the square root of MSE, Mean Squared Error which is the average of the squared error used as the loss function for least squares regression. Specifically, the RMSE is defined as","excerpt":"Model evaluation plays a crucial role while developing a predictive machine learning model. Building just a predictive model without checking does not count as a fit model but a model which gives maximum accuracy surely does count a good one. For this, you need to check on the metrics and make improvements accordingly until you […]","categories":["AI Trends"],"tags":["confusion matrix recall","what is recall in confusion matrix"],"author_name":"Ambika Choudhury","publish_date":"2019-04-04T09:40:01","publication_year":"2019","word_count":881,"keywords":["Go","machine learning","programming_languages:R","AI","confusion matrix recall","programming_languages:Go","RAG","what is recall in confusion matrix","R"],"extracted_tech_keywords":["AI","machine learning","RAG","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/10-model-evaluation-techniques-every-machine-learning-enthusiast-must-know\/","complexity_score":3,"technical_depth":7,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10110311,"title":"Indian Companies Embrace AI-Integrated Low-Code for Digital Transformation","content":"GenAI has disrupted many businesses and it seemed like low-code was slated to be one of the many industries doomed to be affected. However, experts believe that the future lies in leveraging AI-infused low-code platforms, which promise significant productivity boosts of up to 35%, enabling diverse teams to innovate and excel in software development. OutSystems is one such US-based low-code application development platform which has formed partnerships with major tech entities such as Microsoft, OpenAI, Google, and AWS, leveraging their AI integration. In an exclusive interview with AIM, Mark Weaser, Vice President-Asia Pacific, OutSystems explained how the combination of AI and high-performance low-code boosts every phase of the software lifecycle, significantly benefiting enterprises across various industries. “In the rapidly digitising financial sector, Indian firms are trying to modernise workflows using AI-integrated low-code. This enables rapid development of digital lending services, automation of loan origination, and significant cost reduction in processing, all achieved much faster than with traditional coding methods,” explained Weaser. Companies Adopting OutSystems Weaser provided insights into how their low-code platform has also been instrumental in transforming legacy systems for companies like JK Cement, one of India’s leading cement manufacturers, and Magenta Mobility—an integrated clean energy and electric mobility solutions provider. JK Cement used OutSystems to develop a range of applications for sales management, land assets management, and loyalty management processes. “This was driven by the need for a more flexible and automated development approach to keep up with industry changes and a growing application backlog, while staying cost-effective,” Weaser explained. He highlighted their goal was to “redefine the experience of its customers, employees and partners,” by consolidating applications under a single platform for a unified UI\/UX experience and shorter go-to-market timelines. Similarly, Magenta Mobility, an integrated clean energy and electric mobility solutions provider, chose OutSystems for its user-friendliness and agility. Weaser noted that these features allowed Magenta Mobility to “supercharge their digital transformation journey for asset tracking, monitoring, and auditing,” significantly reducing manual costs and improving accuracy. OutSystems Embracing GenAI Discussing their recent collaborations with Microsoft and OpenAI, Weaser highlighted their new platform that integrates services from Microsoft Azure OpenAI and their in-house GenAI capabilities. “OutSystems has seamlessly integrated AI into the platform. We recently launched Project Morpheus, which are new generative AI capabilities that further speed and simplify the development process,” he said. He added that their GenAI roadmap includes key benefits such as instant app generation using conversational prompts, an AI-powered app editor offering ongoing suggestions, and real-time visual representations of app changes. These generative AI integrations provide limitless customisations for customers while validating code produced by generative AI mechanisms. Besides, OutSystems has been integrating AI into its offerings since 2018—with its AI Mentor System, a machine learning-powered suite, which provides expert assistance in key software development areas. It includes specialised mentors for guiding developers, ensuring robust code, checking vulnerabilities, optimising application performance, and managing technical debt. OutSystems.AI further aids businesses in quickly developing customer service chatbots, extracting insights from documents, and analysing customer sentiment. Asia-Pacific Projections The market for no-code and low-code platforms, expected to grow nearly 20% to $10 billion in 2023, is anticipated to reach $12.3 billion this year. In light of these projections and their impressive revenue of over $263 million, Weaser emphasised that amidst global macroeconomic uncertainties, organisations will turn to advanced, modern application platforms for quick, secure, and unique application development to adapt to the dynamic global market. He holds a strong conviction about the role of AI, reiterating, “In particular, AI will be key to delivering productivity for all types of developers.” Considering their huge presence and business, Weaser also reflected on the Asia Pacific market. He anticipates that the global demand for digital transformation and tech talent shortages will also be seen in this region. He stated, “60%-80% of organisations find it difficult to fill vacancies in IT roles such as developers and data professionals.” He observed a trend towards digital transformation in various sectors like banking, energy, and government, as seen in OutSystems’ customer base. Addressing developer challenges, Weaser cited the OutSystems Developer Engagement report which stated that while more than half of developers in India said they love their jobs, only 59% said they would be with their current company a year from now, which further drops to 44% for two years. He advocates low-code technology as a solution noting that it helps to reduce time spent on manual coding, enhancing creativity and innovation in development, and positively impacting developers’ attitudes towards career advancement and productivity.","excerpt":"JK Cement and Magenta Mobility leverage OutSystems’ low-code platform for system transformation, enhancing customer, employee, and partner experiences.","categories":["AI Features"],"tags":["AI Companies","Automation","digital transformation","Generative AI","Interviews and Discussions","legacy systems","low code","partnership","Software Development"],"author_name":"Shyam Nandan Upadhyay","publish_date":"2024-01-10T14:32:41","publication_year":"2024","word_count":750,"keywords":["partnership","chatbots","Generative AI","Software Development","legacy systems","digital transformation","Automation","RAG","machine learning","AWS","AI","ML","generative AI","AI Companies","low code","GenAI","OpenAI","Aim","Interviews and Discussions"],"extracted_tech_keywords":["AI","machine learning","ML","generative AI","GenAI","OpenAI","Aim","RAG","chatbots","AWS"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/indian-companies-embrace-ai-integrated-low-code-for-digital-transformation\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10166705,"title":"How New Relic Drives a 10% Productivity Boost for Swiggy","content":"What’s the common thread that binds your morning coffee from Swiggy and its calorie-tracking on HealthifyMe? Both are powered by New Relic’s observability platform. With over 30 million users, Swiggy relies on proactive alerts and real-time monitoring, resulting in a 10% boost in productivity. Meanwhile, HealthifyMe has reduced costs by 10% while doubling its traffic-handling capacity, thanks to a unified observability experience that accelerates innovation. These success stories highlight the power of intelligent observability, a boost for businesses striving to deliver flawless digital experiences. But how did they get here? New Relic has been at the forefront of innovation in the observability space. As the first cloud-based application performance monitoring (APM) solution, it redefined how businesses monitor and optimise their digital infrastructure. Over the past decade, it has expanded beyond APM, offering a comprehensive suite of tools designed to support diverse technology architectures. In 2019, it took the bold step of launching the industry’s first all-in-one observability platform. This unified data layer allowed enterprises to consolidate telemetry from applications, infrastructure and logs into a single cloud-based system. This move gave businesses a transparent pricing model, offering greater accessibility while setting a new standard for the industry. In an exclusive conversation with AIM, Manav Khurana, chief product officer at New Relic, said, “Recently, we introduced intelligent observability, redefining how businesses monitor their digital operations. Traditionally, engineers had to log in, analyse anomalies, and resolve issues manually using multiple tools. Now, our system proactively identifies incidents, pinpoints causes, and even initiates workflows to resolve them.” Human-in-the-loop validation ensures oversight for enterprises seeking automation without losing control while leveraging the power of AI-driven insights. This fusion of automation, insight, and human intervention is the future of observability. The Rise of Agentic Workflows From an observability standpoint, agentic workflows are absolutely critical and will gain momentum rapidly. Khurana mentioned that last month, they introduced four agentic workflows, one with GitHub, one with ServiceNow, and others with Google Gemini and Amazon Q. “What we’re doing is streamlining the actions users take after gaining insights from an observability platform. Instead of manually figuring out the next steps, we’re simplifying the entire workflow for them,” he said. For example, with the GitHub agentic integration, a developer working in VS Code can immediately understand the impact of their code in production, whether it’s positive, negative, or causing issues. If a problem arises, they can trigger workflows to resolve it directly from their IDE. Similarly, in ServiceNow, which many enterprises rely on, an IT professional using the agentic integration, combining the ServiceNow agent with the New Relic AI agent, can quickly diagnose customer issues, identify the right people to notify, and take action, all without leaving their screen. He further said, “I see this transformation happening rapidly. In fact, I believe agentic workflows represent the true value of AI, beyond just generative AI. While generative AI is useful for consumers, the real impact lies in AI agents working together within workflows to drive meaningful outcomes.” Story of Coffee, Calories and More Beyond Swiggy and HealthifyMe, some of India’s biggest brands rely on New Relic to optimise performance and minimise downtime. Big Basket has transformed its operations with the help of New Relic’s monitoring capabilities. By embracing real-time observability, the company ensures deliveries within 15-20 minutes, meeting the growing demand for rapid grocery fulfillment. This technological advantage has fueled an impressive 20% month-on-month growth, enabling Big Basket to expand from a single store to 200 locations in just three months. Additionally, the company has optimised its infrastructure, cutting costs by 35% while maintaining seamless service efficiency. Capillary Technologies has significantly improved its operational efficiency through New Relic’s integration. The company has slashed troubleshooting time by 80%, reducing incident resolution from one hour to just 10 minutes. This enhancement has led to a major improvement in system reliability, with uptime increasing from 99.5% to a near-perfect 99.9%. By streamlining issue detection and resolution, Capillary Technologies ensures a stable and uninterrupted experience for its users. Postman, a globally recognised API development platform, depends on New Relic to uphold industry-leading standards of reliability, availability, and software resiliency. With New Relic’s observability solutions, Postman has strengthened its ability to detect and address customer issues proactively, ensuring seamless operations and a consistently high-quality experience for developers worldwide. Khurana mentioned, “We recently conducted a global survey with 1,700 customers, revealing that businesses experience over 240 software incidents per year, leading to 70–75 hours of downtime annually. Even a single minute of downtime can translate into millions in lost revenue and irreversible reputational damage.” He noted that some of their customers have successfully reduced their annual downtime from 77 hours to just minutes using their observability solutions. A New Observability Challenge The rise of AI-powered software development is reshaping how applications are built, but also increasing the complexity of monitoring them. Khurana explained, “When I was in university, my final project was written in assembly language, 20 pages of code just to open a file. Then came C, C++, and Golang, each simplifying development. Today, AI is taking this even further. With just a few prompts, anyone can generate software.” While this innovation is exciting, AI-driven software introduces unpredictability, making observability even more critical. So, as businesses embrace AI, cloud computing, and mobile-first strategies, the need for real-time observability is more crucial than ever. Observability is no longer just about monitoring software, it’s about ensuring seamless digital experiences, preventing downtime, and driving business success in an increasingly software-driven world.","excerpt":"Meanwhile, HealthifyMe has reduced costs by 10% while doubling its traffic-handling capacity.","categories":["AI Features"],"tags":["AI (Artificial Intelligence)","new relic","swiggy"],"author_name":"Vidyashree Srinivas","publish_date":"2025-03-26T16:23:56","publication_year":"2025","word_count":912,"keywords":["Go","AI","cloud computing","ML","agentic workflows","RAG","Aim","C++","generative AI","new relic","swiggy","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","ML","generative AI","agentic workflows","Aim","RAG","cloud computing","R","Go","C++"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-new-relic-drives-a-10-productivity-boost-for-swiggy\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":9680,"title":"A talk with Impact Index: Applying Analytics to Cricket","content":"In a nation obsessed with cricket, every match is a gold coin being tossed in mid-air whilst thousands hold their breath for the win. India is a country where people are passionate about cricket, so much so, that they can discuss cricket and cricketers fruitlessly for hours. However, one person decided to translate his passion for cricket, not by playing the game physically but by developing a cricket analytics system. Analytics India magazine got the opportunity interview one of the few “cricket-statistics-analysts” in India and uncover his story and the journey of his brainchild – “Impact Index”. Jaideep Varma, Founder Impact Index “I was dismayed since a long time to see how tally stats were getting deified in cricket and Tendulkar became the new God in the late-1990s on that score. Yet, Rahul Dravid’s contributions in Tests always struck me as much more significant to India’s fortunes.” Recalls Jaideep Varma, the chief custodian of Impact Index. In case you are wondering what Impact Index is all about; it is an analytics system to determine the accuracy of a player’s performance in a specific match or other performances in the same match based on certain factors in a team and career context. In other words, team strength as well as each player’s abilities, strengths and weaknesses are identified using this system. Thus, it is useful while composing teams or for analyzing the opposition. When asked about the need for Impact Index, Jaideep told us that, “Context needs to be accounted for in cricket, so that the real talent gets credit. And history needs to be set right a bit!” Flashback: His journey began 8 years ago, in 2009, when he been invited by the ICC to Oxford for their Centenary Conference to present a paper on the greatest Test players of all time. His determination to find some way of quantifying the data led him to the rudimentary idea of tabulating performances in a match as per only that match’s context and then in a series context. He says, “It is such an obvious idea that it still amazes me that no one has worked with this even now.” He further adds that, “even though it has been a long and arduous journey, it is very rewarding, especially intellectually. In fact, back in 2010, if I had known that it would take me four years to be commercially independent, I would never have taken the decision to spend my savings to run it with an office and staff; but thankfully, life doesn’t come with a timeline counter, and its mysteries are still intact!” He was a successful filmmaker who had even won a National Award in 2011 (for an earlier film). However, he decided to throw caution to the wind and focus full time on Impact Index instead. It wasn’t cakewalk as he was in debt and beyond broke by 2014, when Fidelis World, the holders of the Wisden franchise in Asia, decided to buy a majority stake in Impact Index. That’s when the sunshine broke through as they developed the system and today, run their website independently. Talking about the development journey, Jaideep says, “that particular journey culminates with the publication of the Impact Index book with Aakash Chopra that uses 60 of our prominent findings; it’s a Harper Collins publication and comes out in January 2017.” So, how does Impact Index work? To this, he elucidates, “Since every match has its own fingerprint based on the context of the performances during the game, this system determines a base figure (in runs) for every match. Then, every performance in the match is measured as a ratio against this base figure (and the result limited to 5 in a career context); and then that match performance is examined in the context of its importance to the series.” Talking about the Index success in predicting cricket outcomes, we got to know that Impact Index has covered 15 ODI and T20 tournaments since 2011 and their percentage of getting the four semifinalists right is above 70%! More curiously, not even once in these 15 tournaments has a team not shortlisted by them won a tournament. Commendable, isn’t it? However, personally Jaideep values their post-match inputs far more, since, that is when they reiterate cricket history in a way. The road ahead The Impact Index team aims to find new ways of disseminating the information and make the system better. Moreover, they would like to tie up Impact with the regular video and stats analysis that data miners do, which would make this a completely unique product even in the team consultancy space. In Jaideep’s own words, “There are things we are attempting, let’s see how they go. Like correlating a batsman’s impact to a bowler’s impact when he takes his wicket. Which is much more complex but eventually we, or someone else (someday), will find a solution to. Currently, no one in the world is able to find details that we can.” ____ Since, not much marketing or PR exercise has been done with Impact Index, its acceptability is still not as widespread as it can be. We believe that the media as well as the cricket analytics globally, has a lot to gain from using Impact Index and understanding their teams and players better. After all, it’s all about the fair play, right?","excerpt":"In a nation obsessed with cricket, every match is a gold coin being tossed in mid-air whilst thousands hold their breath for the win. India is a country where people are passionate about cricket, so much so, that they can discuss cricket and cricketers fruitlessly for hours. However, one person decided to translate his passion […]","categories":["AI Features"],"tags":["analytics companies","Interviews and Discussions","sports analytics"],"author_name":"Apoorva Verma","publish_date":"2016-04-25T09:14:31","publication_year":"2016","word_count":891,"keywords":["Go","programming_languages:R","AI","sports analytics","programming_languages:Go","analytics companies","Aim","ViT","analytics","GAN","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","analytics","Aim","R","Go","GAN","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/talk-impact-index-applying-analytics-cricket\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":43193,"title":"Accenture Ventures Hosts Second Edition Of Annual Challenge Focused On Applied Intelligence — Gives AI Startups A Platform To Scale","content":"Accenture Ventures Accenture Ventures which acts as a bridge in the global innovation ecosystem, has been at the forefront of fostering innovation and helping startups maximize market opportunities across the globe. Over the past few years, Accenture Ventures India, as a vital part of the Accenture Innovation Architecture has brought the most disruptive deep tech solutions from startups to clients who leverage the IP and product expertise coupled with Accenture’s domain knowledge. The hugely successful corporate startup engagement programme has implemented innovative solutions from more than 80 startups and deployed them at scale for clients in India and across the globe. One of the top corporate accelerators in India, Accenture Ventures India is creating a culture of innovation by building an ecosystem of partnerships with incubators, corporate venture capitalists, academia, government and startups. Besides changing the market dynamics for startups, Accenture Ventures is solving the key pain-points of enterprises by giving them access to startups that are providing game-changing solutions. The corporate accelerator has developed deep expertise and capabilities in emerging technologies such as artificial intelligence, blockchain, extended reality, Industry X.0 and New IT. Currently, Accenture Ventures tracks more than 1300 B2B Deep-tech start-ups and engages with nearly 200 of them. Accenture Ventures also aims to provide a unique solution to every client leveraging deep tech capabilities. Applied Intelligence Challenge – 2019 As part of their effort to help startups move up the ladder of innovation and gain access to a larger market, Accenture Ventures announced the second edition of its annual challenge focussed on Applied Intelligence this year, where it identifies and recognizes deep-tech growth stage start-ups with the most innovative business to business (B2B) use cases in Applied Intelligence. The approach of applying artificial intelligence and human ingenuity at the core of business to help clients solve their most complex business problems is what Accenture calls Applied Intelligence. By deploying AI responsibly and combining it with deep industry and analytics expertise, Accenture enables the digital transformation of organizations, extends human capabilities, and makes intelligent products and services a reality. The different segments of the challenge were artificial intelligence, data, analytics, automation and industrial AI. B2B startups in the deep-tech space with an  annual revenue of $100K were eligible to apply for the challenge. The selection panel comprised of Accenture leaders, business leaders, venture capital community and investors. The winners join Accenture Venture’s Open Innovation Partner program to collaborate with Accenture and co-create solutions for large enterprises across the globe. The finalists for the Applied Intelligence Challenge are mentioned below: Analytics Locobuzz Aureus Analytics Byte Prophecy Automation Numberz (Paradime Technologies) iFuture Robotics AutomationEdge Technologies Artificial Intelligence Niramai Active.ai CropIn Technologies Industrial AI Flutura Business Solutions Strayos Neewee Data Zapr Media Labs dDriven MMS.IND Benefits for winners 1) Invitation to join a special cohort program – Mentorship and all benefits of “Open Innovation Partner program” 2) Opportunity to co-create at Accenture Ventures CoE at the Accenture Innovation Hub –  Access to a state-of-the-art facility that combines the broad range of Accenture’s leading technology capabilities in a vibrant and collaborative environment 3) Access to Accenture’s clientele across the globe – Winners can hugely benefit from Accenture’s enterprise client leadership and provide customized solutions to big banner names 4) Access to Industry Forums – NASSCOM, CII, FICCI programs 5) Exclusive presentation opportunity to Accenture Advisory Board –  Opportunity to get in front of a group of highly influential industry leaders The Grand finale will be held on July 25, 2019.","excerpt":"Accenture Ventures which acts as a bridge in the global innovation ecosystem, has been at the forefront of fostering innovation and helping startups maximize market opportunities across the globe. Over the past few years, Accenture Ventures India, as a vital part of the Accenture Innovation Architecture has brought the most disruptive deep tech solutions from […]","categories":["Deep Tech"],"tags":["AI Benefits"],"author_name":"Abhijeet Katte","publish_date":"2019-07-23T18:23:29","publication_year":"2019","word_count":578,"keywords":["Go","API","artificial intelligence","AI","AI Benefits","Git","RAG","Ray","Aim","analytics","R"],"extracted_tech_keywords":["AI","artificial intelligence","analytics","Aim","Ray","RAG","R","Go","Git","API"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/accenture-ventures-hosts-second-edition-of-annual-challenge-focused-on-applied-intelligence-gives-ai-startups-a-platform-to-scale\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":41044,"title":"AI Mental Health Platform Wysa Raises ₹ 15 Crores In Funding Led By pi Ventures","content":"In a recent development, Wysa, an AI conversational agent that help improve mental health, raised about $2m (Rs 15 crore) in a pre-Series A round of funding. This funding round was led by pi Ventures, with participation from Kae Capital and other investors. Founded by Jo Aggarwal and Ramakant Vempati three years ago, Wysa is an AI-based ‘emotionally intelligent’ bot, a virtual coach that combines empathetic listening with evidence-based therapeutic techniques like CBT, meditation and motivational interviewing, to make mental health accessible at scale. The founders aim to use this funding amount to further strengthen their technology and for expansion. The company had earlier raised a round of $1.3 million in seed funding from Kae Capital and angel investors in 2017. In an earlier conversation with Analytics India Magazine, Vempati had said that they using AI and NLP models to understand user text input, link it to clinical assessments and metrics, and use rules engine linked to a proprietary therapy content platform to choose an appropriate intervention and self-help technique for delivery. The startup claims that this passive sensing model has helped them detect depression up to a 90% accuracy in trials. “Wysa has been co-designed by therapists, users, and designers over hundreds of iterations and 80 million conversations. What people want most is to feel heard, without judgement. We combine the free AI with unlimited support from a qualified therapist, still anonymously, over chat to make it easy to get help.  We are delighted that our investors are partnering with us on this journey to make mental health accessible for everyone,” said Aggarwal. With its unique offering, Wysa has helped over 1.2 million people from 30+ countries, making it one of the global leaders in AI for mental health. It is recommended by the National Health Service (NHS) in the UK, and its efficacy has been validated through a peer-reviewed study. Manish Singhal, Founding Partner, pi Ventures, said, “Mental health could very well be the next big epidemic to hit the human race. Training more human therapists will not bridge the massive supply and demand gap. This is where Wysa powered by an AI engine comes in. It is scalable and is available for anyone to chat at any time in total privacy.” “We do believe that Wysa can create a zero-stigma pathway to support people who are struggling, and has a real chance of solving depression and mental health challenges at scale. We wish them all the very best and are grateful to be working with them,” he further added. Vempati and Aggarwal left their careers as Director at Goldman Sachs and MD of Pearson Learning respectively to start working on their ambitious project, Wysa which has grown extensively in terms of popularity, with more than five million conversations and high rating on Google Play. Wysa has users from the US, India and over 30 countries.","excerpt":"In a recent development, Wysa, an AI conversational agent that help improve mental health, raised about $2m (Rs 15 crore) in a pre-Series A round of funding. This funding round was led by pi Ventures, with participation from Kae Capital and other investors. Founded by Jo Aggarwal and Ramakant Vempati three years ago, Wysa is […]","categories":["AI News"],"tags":["meditation"],"author_name":"Srishti Deoras","publish_date":"2019-06-20T11:15:02","publication_year":"2019","word_count":478,"keywords":["meditation","Go","API","funding","AI","Scala","NLP","Aim","analytics","R","startup"],"extracted_tech_keywords":["AI","NLP","analytics","Aim","R","Go","Scala","API","startup","funding"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/ai-mental-health-platform-wysa-raises-%e2%82%b9-15-crores-in-funding-led-by-pi-ventures\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":34649,"title":"MachineHack Launches New Hackathon: Use Your Data Science Skills To Predict A Doctor’s Consultation Fee","content":"Analytics India Magazine’s hackathon platform MachineHack, has come up with a new and challenging hackathon for data science enthusiasts. The new hackathon challenge called Predict The Doctor’s Consultation Fee Hackathon, will challenge the MachineHack users to predict a doctor’s consulting fee based on the given data. The winners of the hackathon will win exciting prizes. We have all been in situations where we go to a doctor in an emergency and find that the consultation fee is far too high. As a data scientist, we all can do better. What if you have the data about the important details about a doctor and you get to build a model to predict the doctor’s consulting fee? This is the hackathon that lets you do just that. The hackathon consists of a training set of a size of 5,961 records and the test set has 1,987 records. The data set includes: Qualification: Qualification and degrees held by the doctor Experience: Experience of the doctor in a number of years Rating: Rating given by patients Profile: Type of the doctor Miscellaeous_Info: Extra information about the doctor Fees: Fees charged by the doctor Place: Area and the city where the doctor is located. The hackathon participants have to make use of this data, build a model and predict the consulting fee charged by the doctor. MachineHack Platform has received an overwhelming response from its users. At MachineHack, we are constantly trying to bring to its users a diverse set of challenges, through these hackathons. We recently concluded two hackathons called “Predict The Data Scientists Salary In India Hackathon” and “Visualisation is Beautiful: Data Science Salary Visualisation Challenge”, and announced prices for the winners. We talked to the winners of these hackathons to know how they solved the particular problem of the hackathon.","excerpt":"Analytics India Magazine’s hackathon platform MachineHack, has come up with a new and challenging hackathon for data science enthusiasts. The new hackathon challenge called Predict The Doctor’s Consultation Fee Hackathon, will challenge the MachineHack users to predict a doctor’s consulting fee based on the given data. The winners of the hackathon will win exciting prizes. […]","categories":["Deep Tech"],"tags":["dataset","Hackathon"],"author_name":"Disha Misal","publish_date":"2019-02-08T05:00:23","publication_year":"2019","word_count":298,"keywords":["dataset","data science","Go","programming_languages:R","AI","programming_languages:Go","Hackathon","analytics","R"],"extracted_tech_keywords":["AI","data science","analytics","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/machinehack-launches-new-hackathon-use-your-data-science-skills-to-predict-a-doctors-consultation-fee\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10095914,"title":"Andrew Ng Releases Generative AI with LLMs Course with AWS","content":"Andrew Ng’s DeepLearning.AI, in partnership with Amazon Web Services (AWS), has announced an exciting new course on Coursera called “Generative AI with Large Language Models” to address the growing demand for expertise in this field. Click here to enrol for the course. By enrolling in this course, participants will gain a comprehensive understanding of the generative AI lifecycle based on LLMs and the underlying transformer architecture that powers them. They will learn how to effectively utilise LLMs for various tasks by selecting the most suitable model and implementing appropriate training techniques. Apart from Andrew Ng, the instructors include Antje Barth, principal developer advocate at AWS; Chris Fregly, principal solutions architect at AWS, Shelbee Eigenbrode, principal solutions architect at AWS, and Mike Chambers, developer advocate at AWS. The course will also cover cutting-edge methods for training, fine-tuning, inference, and deployment of models, ensuring optimal performance in real-world scenarios. Additionally, learners will acquire essential skills to navigate the evolving landscape of generative AI and effectively integrate it into their organisations and products. The course includes: Data gathering: Collecting relevant data for training the generative AI model. Model selection: Choosing the appropriate model architecture for the task. Performance evaluation: Assessing the quality and effectiveness of the generated outputs. Deployment: Implementing the generative AI model in a real-world setting. Provide a detailed description of the transformer architecture that powers LLMs. Explain how LLMs are trained using the transformer architecture. Discuss how fine-tuning allows LLMs to be adapted to specific use cases. Utilise empirical scaling laws to optimise the model’s objective function. Optimise the objective function based on factors such as dataset size, compute budget, and inference requirements. Apply state-of-the-art training, tuning, inference, tools, and deployment methods. Use advanced techniques and tools to maximise the performance of generative AI models. Consider the specific constraints and requirements of the project. As businesses adapt to leverage the power of generative AI, the associated complexities and uncertainties surrounding this technology have become inevitable. Andrew Ng said that this course aims to demystify the subject and equip learners with the knowledge and skills required to confidently harness the potential of LLMs in their endeavours. Andrew Ng has been very vocal about promoting people to learn and adapt to generative AI. Earlier this month, he also released three new generative AI courses with LangChain, OpenAI, and Lamini. This is after releasing a course for prompt engineering in April in partnership with OpenAI. In December 2022, DeepLearning.AI also introduced Mathematics for Machine Learning and Data Science Specialization Course, a beginner level mathematics course for AI.","excerpt":"The course will cover methods for training, fine-tuning, inference, and deployment of models, ensuring optimal performance in real-world scenarios","categories":["AI News"],"tags":["Andrew Ng","Courses"],"author_name":"Mohit Pandey","publish_date":"2023-06-29T09:54:36","publication_year":"2023","word_count":424,"keywords":["data science","machine learning","OpenAI","AI","AWS","Andrew Ng","RAG","LangChain","Aim","prompt engineering","generative AI","Courses"],"extracted_tech_keywords":["AI","machine learning","data science","generative AI","OpenAI","LangChain","Aim","RAG","prompt engineering","AWS"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/andrew-ng-releases-generative-ai-with-llms-course-with-aws\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":65112,"title":"Top 8 Data Science Institutes In India For Corporate Training","content":"Corporate training has been playing an essential role in India for training professionals and allowing companies to provide superior IT services. This has helped the country become one of the leading countries in the world that offer IT services. However, with the changing demand for IT skills due to the rise of data science, corporate training has further gained prominence among various companies. Today, data-driven companies are struggling to find the right talent who can assist them in driving business growth by moulding information and delivering insights into data. Consequently, various institutions in India are offering corporate learning in data science-related fields to help businesses bridge the talent gap in the market. Please note that this list is not a ranking and the institutes are listed in alphabetical order. AnalytixLabs AnalytixLabs empowers organisations’ managers to make business decisions based on data-driven insights. Its corporate training courses are devised based on your organisations’ requirements. AnalytixLabs provides various training programs such as building blocks, tools and techniques. Through building blocks, it offers training for non-analytics executives to increase their awareness of the potential of data among decision-makers. Besides, AnalytixLabs has tools and techniques program where the company focuses on teaching about popular tools such as SAS, R, Big Data Platforms, and MS Excel along with statistical concepts and predictive analytics. It ensures that employees not only get the training on tools and techniques but also are capable of applying the learning in real world situations. Great Learning Great Learning’s corporate training nurtures your employees with the latest technology that is taking the central stage in your organisations. Depending on the requirements, Great Learning offers customised training in data science, machine learning, cloud computing, big data, AI, and business management. It provides flexibility for your employees by offering the training in multiple formats such as online, classroom, and blended learning. IMS Proschool IMS Proschool provides corporate training for finance, analytics, digital marketing and vocational training. Unlike others, IMS Proschool extensively focuses on financial modelling to help professionals make informed monetary decisions. It offers training right from the basics such as Excel with VBA and Macros to advanced business analytics with Python or R. With its training, IMS Proschool has catered to companies like IDBI, SBI, LIC, UTI, Kotak, Tata Capital, ICICI, among others. Jigsaw Academy Jigsaw Academy believes that analytics is for everyone in organisations to make quick decisions and drive business growth. It has a comprehensive training program, where Jigsaw does not directly come up with pre-devised courses for employees. Rather, it evaluates the team and then tries to fill the gap with its data science training. A customised learning path is what any corporate requires since the knowledge of their employees varies at all levels. Besides, Jigsaw evaluates the progress of employees’ learning and capabilities to implement the learning in real world projects. Jigsaw Academy has become one of the preferred choices for organisations for training their employees as Jigsaw Academy has a proven record of being on top for full-time data science courses over the years. Simplilearn Simplilearn offers bespoke blended corporate training environments for both existing employees to upskill and onboarding new hires with advanced skills. Its curriculum includes live online classes, mentoring, labs, projects, and self-paced videos. The firm has trained companies around the globe, which includes Visa, IBM, Dell, VMware, MasterCard, among others. Along with AI, big data, and data science courses, Simplilearn’s corporate training suite consists of agile and scrum, DevOps, cloud computing, blockchain, digital marketing and more. Springboard Springboard makes your workforce future-proof by offering corporate training in data science and analytics. The program is built by industry experts to deliver learning that can be implemented in the real world projects. It also provides mentors to guide learners or professionals in the right direction to mitigate business challenges. Over the years, it has trained professionals of Facebook, Visa, Target and more, to enable them to make data-driven decisions. Udacity Udacity is mostly popular for offering courses for students has enterprise training initiatives to prepare the employees for the digital future. Through its enterprise training, Udacity offers access to all the Nanodegree courses, which can be leveraged to learn almost every latest technology. Udacity for enterprise varies widely from its courses for students; it customises the learning path based on various companies requirements. Besides, it provides priority access to technical mentors to resolve the queries of learners quickly. Over the years, Udacity has trained employees of Accenture, CISCO, CapitalOne, among others. UpGrad UpGrad also provides bespoke corporate training for your employees in data science, big data engineering, and machine learning & AI. To offer the program, the firm has collaborated with industry experts from different companies for real-world teaching, which goes beyond theoretical knowledge. The programs help in upskilling employees in a wide range of tools such as Hadoop, Tableau, R programming, Python, Apache Spark, among others. Over the years, the company has offered training at Apple, Fractal, Tata Sky, Jet Airways, and more.","excerpt":"Corporate training has been playing an essential role in India for training professionals and allowing companies to provide superior IT services. This has helped the country become one of the leading countries in the world that offer IT services. However, with the changing demand for IT skills due to the rise of data science, corporate […]","categories":["AI Trends"],"tags":["corporate analytics platform","data science training","devop","what is data science"],"author_name":"Rohit Yadav","publish_date":"2020-05-13T14:00:00","publication_year":"2020","word_count":828,"keywords":["data science","machine learning","AI","cloud computing","corporate analytics platform","R","predictive analytics","Apache Spark","RAG","Python","data science training","analytics","what is data science","devop"],"extracted_tech_keywords":["AI","machine learning","data science","analytics","RAG","predictive analytics","cloud computing","Apache Spark","Python","R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-8-data-science-institutes-in-india-for-corporate-training\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10117464,"title":"We Finally Have A Hugging Chat for Indic LLMs","content":"Bhabha AI has unveiled Indic Chat, a playground for open source Indic LLMs. Built on top of Hugging Face’s Hugging Chat, the model hosts Indic AI models for people to chat and test out. Click here to check it out. Currently, the interface has: Telugu-LLM-Labs\/Indic-gemma-7b-finetuned-sft-Navarasa-2.0 GenVRadmin\/AryaBhatta-GemmaOrca BhabhaAI\/Gajendra-v0.1 ai4bharat\/Airavata Additionally, users can also join the Discord channel which is now accessible to all for fostering collaboration and expediting the development of Indic LLMs. Bhabha AI has also published a version of the OpenHermes-2.5 instruction dataset comprising approximately 600,000 rows, which have been filtered and translated into Hindi. Satpal Singh Rathore, the creator of Bhabha AI and Indic Chat collaborated with Telugu LLM Labs. Ramsri Goutham Golla & Ravi Theja Desetty from Telugu LLM Labs supported Bhabha AI for computing resources. Few weeks back, Bhabha AI unveiled Gajendra, an early release of their 7B Hindi-Hinglish-English instruction fine-tuned language model built on top of Sarvam AI’s OpenHathi, which is built on top of Llama 2. In a forward-looking move, Bhabha AI is delving into the exploration of filtering examples that can be translated from English to Hindi. The initiative includes the release of initial versions of both the dataset and the corresponding model.","excerpt":"You can test out Indic chat models on this website.","categories":["AI News"],"tags":["LLMs"],"author_name":"Mohit Pandey","publish_date":"2024-04-02T14:23:20","publication_year":"2024","word_count":200,"keywords":["Go","Hugging Face","programming_languages:R","AI","LLMs","programming_languages:Go","ai_frameworks:Hugging Face","llm_models:Llama","R"],"extracted_tech_keywords":["AI","Hugging Face","R","Go","llm_models:Llama","ai_frameworks:Hugging Face","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/we-finally-have-a-hugging-chat-for-indic-llms\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10056743,"title":"Mumbai’s Bandra Station Becomes High Tech, Installs AI-Based Loco Inspection System","content":"The Western Railway has installed an AI-based Loco Inspection System at Trip Shed, Bandra Terminus, under the Mumbai division on December 18. The AI system will help increase efficiency, enhance maintenance of the loco engines, and improve loco outage. The existing system is manual, where maintenance is carried out in a periodic manner. With the predictive asset maintenance system, the condition of equipment in locos will be continuously monitored. It can inspect the sides, underframe and roof of the loco train. Some of the checks that the system will measure include the axle box temperatures, detection of hanging parts, loose parts, foreign particles, wheel profile, carbon strip defects, brake block thickness, broken roof insulators, cattle guard defects, and primary spring cracks of the loco. The system will raise alarms for proper intervention during an emergency. The technology utilises a 3D camera, 2D camera, thermal camera, LIDARs, edge processors, sliding arms, access points, etc. Using this technology also reduces the cost and time of maintenance. The speed of the AI-based Locomotive Inspection System is 1 metre per minute and the time taken to inspect one loco on a set of parameters by the system is 20 mins, as against 3 hours in manual mode. The newly launched AI-based Locomotive Inspection System has increased pit availability and improved the locomotive outage.","excerpt":"This is a first of its kind AI-system of the Indian railways that will prove helpful to reduce loco failure and improve the overall efficiency of engines.","categories":["AI News"],"tags":["LIDAR"],"author_name":"Meeta Ramnani","publish_date":"2021-12-22T10:28:29","publication_year":"2021","word_count":219,"keywords":["R","LIDAR","programming_languages:R","AI"],"extracted_tech_keywords":["AI","R","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/mumbais-bandra-station-becomes-high-tech-installs-ai-based-loco-inspection-system\/","complexity_score":1,"technical_depth":3,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10103266,"title":"Can Mira Murati Save OpenAI?","content":"The Silicon Valley was still reeling with the sudden and unexpected firing of Sam Altman, co-founder and CEO of OpenAI, by the company’s board of directors, when in a surprising turn of events, president and chairman of the board Greg Brockman too resigned. This led to the appointment of Mira Murati, the chief technology officer of OpenAI, as the interim CEO. Murati was the only person who found out the night before about Altman’s termination and Greg’s ouster from the board while retaining her role in the company. The rest of the management team was informed shortly after the dismissal. Following the news, Murati sent a staff note to her employees encouraging them to concentrate on their tasks. Murati said that she is “honoured and humbled” in assuming the leadership position and emphasised the importance of maintaining focus, determination, and adherence to core values, as reported by Bloomberg. With Albanian roots and a San Francisco upbringing, 35-year-old Murati completed her bachelor’s degree in mechanical engineering from Dartmouth. She has served in important roles at Goldman Sachs, French aerospace company Zodiac Aerospace, and Elon Musk’s Tesla, where she served as a senior product manager for the ‘Model X’ vehicle. After a stint as the VP of product and engineering at Leap Motion, she joined OpenAI in 2018 as the VP of applied AI and partnerships, eventually rising to CTO and now the interim CEO position. Murati played a pivotal role in the release of groundbreaking AI projects like DALL.E 2, DALL.E 3, ChatGPT, GPT-4, and more. Although she usually operates behind closed doors, Murati has started making public appearances, discussing the implications of AI tools and advocating for responsible AI regulation, emphasizing the need for broader input beyond tech companies in shaping ethical policies. Murati had previously said in an interview that during her time at Leap, she recognised that AGI would be the ultimate and most significant technological milestone, and believed that OpenAI was the sole organisation at that time committed to advancing AI capabilities while also ensuring responsible development — she wanted to join them. Difference in Approach Murati, who has been relatively less visible in interviews like her colleagues, has recently begun making more appearances. This has provided us with insight into her perspectives on navigating the AI landscape, giving us a glimpse into her beliefs in that space. At an interaction with Microsoft CTO Kevin Scott in July, Murati was vocal about the prevailing uncertainty surrounding LLMs and the necessity for clear guidance and decision-making processes in the field, raising how one should determine what aspects of AI to prioritise, work on, release, or position effectively. “When we began building GPT more than five years ago, our primary focus was the safety of AI systems,” said Murati. Emphasising the risks associated with allowing humans to directly define the goals or objectives for AI systems, as this approach can involve using complex, opaque processes for critical functions, potentially leading to serious errors or unintended consequences, Murati and the team shifted their focus to using RLHF to ensure the safe and effective development of AI. After developing GPT-3 and releasing it in the API, OpenAI was able to integrate AI safety into real-world systems for the first time. They used instruction-following models to take prompts from customers and generate feedback for the model to learn from. By fine-tuning the model on this data, they were able to build instruction-following models that were much more likely to follow the intent of the user and do what was actually wanted. For Murati, this was a significant step forward because AI safety was no longer just a theoretical concept but became practical in the real world. Karma Hits Back Following the removal of Altman, Ilya Sutskever, co-founder and chief scientist at OpenAI, who is being deemed to be the brain behind the sudden removal of Altman, addressed his employees. Rejecting the notion that Altman’s removal was a “coup” or “hostile takeover”, as it is being said, he acknowledged that there were genuine concerns within the organisation regarding the prioritisation of commercialising AI technology, potentially overlooking safety precautions, as reported by The Information. The ongoing debate at OpenAI revolves around balancing the pursuit of AGI with safety concerns and avoiding a sole focus on business interests, with Altman and Brockman on one end and Sutskever on the other. For Sutskever, the significance of AGI lies as a practical benchmark for AI capabilities. Both appear cautious and considerate of AGI’s potential impact on society, but without compromising safety. Echoing similar views, Murati is strongly in favour of its development with a focus on ensuring that AGI benefits humanity. Both Murati and Sutskever believe in the importance of AGI development while ensuring its benefits to humanity. Murati had earlier mentioned that even when GPT-4 was being built, there was a strategic decision to refocus on improving ChatGPT’s alignment and safety. The aim was to actively involve researchers and gather their feedback to enhance the reliability, robustness, and alignment of ChatGPT. Altman’s announcements at DevDay, especially those involving Microsoft, led to a ChatGPT outage, possibly due to a DDoS attack. OpenAI also temporarily halted new ChatGPT Plus sign-ups due to a surge in usage following DevDay. In essence, Sutskever acknowledged that some employees at OpenAI were worried that under Altman’s leadership, there might have been a strong emphasis on rapidly turning AI developments into profitable ventures. This intense focus on commercialisation could have potentially sidelined or compromised the rigorous safety measures and precautions that are essential in the development of AI systems. Altman initially founded OpenAI as a non-profit but later introduced a for-profit entity to secure AI research funding, a move seen as conflicting with the company’s commitment to safety. Along with this, his approach of moving from open to closed source has defied the company’s original vision, a stance about which both Brockman and he seem to be at odds with Sutskever. Elon Musk, who helped in cofounding OpenAI, soon left when Altman focused more on commercialisation instead of open-sourcing AI, throwing away its original ideology of openness. As a firm believer in open-sourcing AI who is planning to open-source xAI’s chatbot Grok, Musk had earlier said that the closed-source policy will bring “bad karma” for OpenAI. On the other hand, Musk praised Sutskever as a “brilliant, good human, and a linchpin of OpenAI”. At a time when three more senior OpenAI researchers, namely Jakub Pachocki, Aleksander Madry and Szymon Sidor, resigned in response to Altman’s termination and Brockman’s resignation, it becomes increasingly important for Mira Murati to shape OpenAI future trajectory as the interim CEO by leveraging her extensive experience in AI, advocacy for responsible AI, and emphasis on AGI development for the benefit of humanity. Except for the internal memo, Murati has yet to make any public statements on her new role. Read more: Is this the end of OpenAI as we know it?","excerpt":"Murati has played a pivotal role in the release of groundbreaking AI projects like DALL.E 2, DALL.E 3, ChatGPT, GPT-4, and more.","categories":["Global Tech"],"tags":[],"author_name":"Shritama Saha","publish_date":"2023-11-18T17:48:30","publication_year":"2023","word_count":1152,"keywords":["Go","ChatGPT","API","RLHF","OpenAI","AI","R","RAG","Aim","xAI"],"extracted_tech_keywords":["AI","ChatGPT","OpenAI","xAI","Aim","RAG","RLHF","R","Go","API"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/can-mira-murati-save-openai\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10008232,"title":"Tech Stack For Your E-commerce Startup: Here’s What You Should Know","content":"Back in the 90s, the Internet acted as the essential infrastructure that provided the rapid spread of e-commerce. For e-commerce, companies leverage shopping cart software on a web server that enables visitors to a website or mobile application to pick items for ultimate purchase. One of the main e-commerce segments comprises B2C (business-to-consumer), also known as retail online shopping. At the point of sale, the software determines a total for the order, including freight transport, postage along with packaging and labelling. The shopping cart software is typically installed on the server which hosts the website, or on the secure server which accepts sensitive ordering user and transactional data. In most server-based implementations, data related to the shopping cart is kept in the session object and is accessed and manipulated in real-time as the user selects various items from the cart. Later at the process of finishing the transaction, the information is accessed, and order is produced against the selected item, thus clearing the shopping cart. Data such as products, categories, discounts, orders, customers, and others are then stored in a database and accessed in real-time by the platform. All of this needs a complex technology stack particularly for large scale users, as seen on online shopping websites. When it comes to actually building an e-commerce website, a web application, a mobile application, you have so many choices ahead of you. Much like the world of clothing fashion, more generally, is continually evolving such that what is appropriate to use or wear now may not surely be relevant some months or years in future, this can be said for the technology world. Here we give you a sense of some of the languages, frameworks, libraries, along with some of the overarching design decisions that are both in vogue and here to stay right now. Front End If we look at the front end of an e-commerce website or application, it commonly applies to that which is facing the user. Storefront is part of a Web store which is accessed by users for online shopping, consisting of various categories, products, and other pages are dynamically created based on the data saved in the database. So it’s the user interface and more with which an individual typically interacts. Web-based experience for any e-commerce website is done using HTML and CSS and, even more dynamically, using JavaScript. There are JS frameworks collections of code that other people have written that you can use. This includes Angular and React, and these are the names for various JavaScript frameworks. So a framework provides code that you can integrate into your own projects, whether it’s CSS, JavaScript, and way of building your front-facing e-commerce applications. You will find among several of these frameworks different design paradigms, different design beliefs, the best way as to do things. Part of the method of determining these frameworks really comes down to what resonates with a startup or with the engineering team the startup has set up. Meanwhile for CSS also, there are libraries there, groups of CSS files and frameworks methodologies to determine the layout of your site, like Bootstrap, Semantic UI, and more. These focus on the aesthetics of individual experience, more so on the display of information and the kinds of user interface mechanisms, including the menus, the buttons, the windows, and other things a user sees on the screen display. Back End Now if we take a look now at the backend, there is a long list of back end processes on the servers from which the HTML and the CSS and even the JavaScript are finally coming. There are numerous languages like Java, JavaScript, .NET, PHP, Python, Ruby, Scala, and so many others to build back end processes and web servers, which interact with the databases and relay it to the e-commerce website or application. And so in this world of the backend, e-commerce shops or startups have libraries of reusable frameworks, libraries of code and methodologies through which they can build your applications, including Django, Flask, .NET, Node.js, Rails, .NET and others. Should You Pick A Hosting Platform Like Shopify To Get Started As An E-Com Startup? Solutions like Shopify are fully managed by the service providers on a monthly subscription. Shopify is used widely across the globe by startups and has a bunch of applications which can be integrated. For hosting eCommerce platforms, Shopify is definitely a part of the basic toolkit. Even with third party hosting services, startups can do a lot of custom work in terms of backend integrations with ERP systems to figure out what is the best cost-effective way to do their business and not just to buy stuff with a shell. Shopify is very user friendly for new entrepreneurs who don’t have the resources or talent if managing the tech stack for themselves. No platform’s better from the ease of use than Shopify that has everything built-in for any e-commerce firm to get started and succeed. Once the scale of users exceeds a threshold and the startup begins generating serious revenue, the web store can be migrated to a cloud platform, which is more appropriate for performing customised tasks such as analytics or integrating microservices. There are also open-source frameworks like Magento, which many companies often use free of charge. The framework allows the company to host the platform using a hosting service and also enables them to modify the entire source code to create customised user services. Using APIs and plugins, companies also do integrations with third-party software such as ERP systems. Conclusion E-commerce startups are guided by what their engineering team knows, perhaps what your own system administrators or your operational people know. So the folks who are actually maintaining the servers, whether they’re locally on-site, or the ones running things in the cloud, whether it be AWS, Azure or Alibaba. Depending on what that cloud infrastructure supports, it might influence your decision making as to what language you need to pick. For example, some of these languages make it a little bit easier to develop applications that support immediate interactivity, where there’s a constant connection or the illusion of a constant connection between browser and server. JavaScript, via a framework called Node.js, makes it really easy to do, and it was designed more so with that kind of use case in mind. As an e-commerce startup which is starting from scratch for handling massive scale user traffic and complex workflows, it is essential to hire full-stack engineers with deep experience in building web applications. Particularly, people who are comfortable working with everything from CSS and JavaScript to databases and API design.","excerpt":"Back in the 90s, the Internet acted as the essential infrastructure that provided the rapid spread of e-commerce. For e-commerce, companies leverage shopping cart software on a web server that enables visitors to a website or mobile application to pick items for ultimate purchase.  One of the main e-commerce segments comprises B2C (business-to-consumer), also known […]","categories":["AI Startups"],"tags":["ecommerce","object store database"],"author_name":"Vishal Chawla","publish_date":"2020-09-24T14:00:24","publication_year":"2020","word_count":1105,"keywords":["AWS","AI","R","ML","RAG","microservices","Python","ecommerce","analytics","JavaScript","Azure","object store database"],"extracted_tech_keywords":["AI","ML","analytics","RAG","AWS","Azure","microservices","Python","R","JavaScript"],"url":"https:\/\/analyticsindiamag.com\/ai-startups\/tech-stack-for-your-e-commerce-startup-heres-what-you-should-know\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":26818,"title":"Piramal Enterprises To Sell Its Analytics Subsidiary?","content":"Following the consolidated net loss of ₹69.81 crore for the first quarter ended 30 June 2018, Piramal Enterprises on Tuesday indicated that they may close down their analytics subsidiary Decision Resources Group. According to a report, Ajay Piramal, Chairman of Piramal Enterprises indicated that the company may close down the Decision Resources Group if it fails to grow revenue and improve profitability. This news comes as a rude shock against the backdrop of the analytics sector in India doing very well, in general. Piramal told the news portal, “If we find that, in the future, this is not happening, we will go for unlocking value.” Decision Resources Group is a global information and technology services company that provides proprietary data and solutions to the healthcare industry. They provide end-to-end solutions to complex challenges in healthcare. Piramal Enterprises had bought out Burlington-based Decision Resources Group in 2012 for $635 million. But reports have suggested that Decision Resources Group sales have been almost flat over the last three financial years. Piramal told the news portal that the underwhelming performance by DRG was due to a shift in the delivery model from large-scale and static research reports to digitally-delivered and user-centric applications and services. “What has happened now is that it has gone towards real-world data and more analytics. That is the transition we are taking to… There are some early signs of topline growth. Now we are focusing on improving the profitability. Yes, it is an area of focus for the management to ensure this. And we are monitoring it closely,” Piramal said. As of now, the Decision Resources Group constitutes only 11 percent of Piramal’s revenues. As of now, DRG has more than 1,000 employees 17 offices worldwide including New York, London and Tokyo.","excerpt":"Following the consolidated net loss of ₹69.81 crore for the first quarter ended 30 June 2018, Piramal Enterprises on Tuesday indicated that they may close down their analytics subsidiary Decision Resources Group. According to a report, Ajay Piramal, Chairman of Piramal Enterprises indicated that the company may close down the Decision Resources Group if it fails to […]","categories":["AI News"],"tags":[],"author_name":"Prajakta Hebbar","publish_date":"2018-07-31T12:18:03","publication_year":"2018","word_count":293,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","Git","analytics","R"],"extracted_tech_keywords":["AI","analytics","R","Go","Git","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/piramal-enterprises-to-sell-its-analytics-subsidiary\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":42238,"title":"Google Datalab Vs Amazon SageMaker: Which Cloud Platform Is Best For Your ML Project","content":"Deploying a machine learning service on a cloud platform is a huge advantage to all ML practitioners, as it provides a flexible platform for the design and development of various models. It provides serverless cloud engines giving the ML advantages of leveraging their models on a cloud platform since these are generally computationally heavy. Google Datalab and Amazon SageMaker are very closely related to each other in many features, but also have many differences. Here is a detailed comparison between the two services\/platforms: 1. Deployment Google Datalab: The notebook server setup procedure is easy. It is launched using the Google cloud shell which is in the Google Cloud Console interface. Google Cloud SDK can also be used for notebook deployment. Amazon SageMaker: Once logged into the SageMaker console, the deployment of the notebook is only a click away. SageMaker wins. 2. Customised Algorithms Google Datalab: It does not contain any pre-customised ML algorithms. But using the Google Cloud ML service, it provides a platform to run the models built with the help of TensorFlow. Amazon SageMaker: It has pre-optimised ML algorithms. These algorithms can run on Amazon’s compute servers. In order to use these algorithms, one needs to connect them to the data source, the objective of which is to help beginners in ML make use of it in their products. The SageMaker custom algorithms have a variety of supervised, unsupervised and deep learning algorithms. SageMaker wins. 3. Model Deployment Google Datalab: There is no direct way to handle the code deployment into production servers. But the model built on this platform is packed into a Python module and deployed on Google CloudML. The only way to deploy one’s own algorithm is by writing it in TensorFlow. Amazon SageMaker: There is a provision for direct deployment of the trained ML models, unlike in Datalab. The deployment is done to elastic compute infrastructure with high availability. It comes with an HTTPS endpoint where the ML model is available to provide inferences where the user can deploy multiple variants of a model to the same HTTPS endpoint. It has a much easier deployment of ML models, compared to Datalab. SageMaker wins. 4. Automated Hyper-Parameter Tuning Hyperparameters are the parameters that describe and govern the model training process and need to be initialised before the training starts. Hyperparameter tuning means multiple trials being run in a single training job, where each trial is a complete execution of the training application with values set within the specified limits. Google Datalab: It does not provide automated hyper-parameter tuning. But it has something called as a HyperTune which helps in automatic optimisation of the ML model for an improved accuracy\/minimized error. It provides this feature for TensorFlow models. Amazon SageMaker: It provides an option for automated hyper-parameter tuning on the ML model during the training period. It finds the best hyper-parameters for algorithm training to the user. This feature is available for not just built-in algorithms but also for external training in docker. SageMaker wins. 5. Pricing: Google Datalab: Pricing is based on usage. But in the case of certain customer use cases and demands, it could have different prices. You can look at the price structure here. Amazon SageMaker: Pricing of the SageMaker is affordable at <price> and is totally based on the usage. As part of the AWS Free Tier, SageMaker is available for free. Pricing depends on the on-demand ML instances, ML storage, and fees for data processing in notebooks and hosting instances. You can get more information regarding the SageMaker price here. Both are equal. 6. Inbuilt Libraries Google Datalab: Datalab comes with a Jupyter Notebook. It does not have any in-built notebook libraries with MxNet and Apache Spark. But it has notebook kernels when used with TensorFlow. Amazon SageMaker: It has pre-installed notebook libraries that run on Apache Spark and MxNet, along with being able to run on TensorFlow. SageMaker wins. Which One Should You Choose Google Cloud Datalab is a standalone serverless platform. It is used for building and deploying ML models. It has to be used with other services like the Google Cloud ML to make it a more powerful ML service. Whereas Amazon SageMaker is built for complete end-to-end ML services. Looking at all the comparisons above, SageMaker should definitely be your choice.","excerpt":"Deploying a machine learning service on a cloud platform is a huge advantage to all ML practitioners, as it provides a flexible platform for the design and development of various models. It provides serverless cloud engines giving the ML advantages of leveraging their models on a cloud platform since these are generally computationally heavy.  Google […]","categories":["Global Tech"],"tags":["Amazon","Cloud Data AI","Cloud Platform","Google","Machine Learning","ML models","SageMaker"],"author_name":"Disha Misal","publish_date":"2019-07-11T13:00:08","publication_year":"2019","word_count":715,"keywords":["Amazon SageMaker","machine learning","AWS","AI","SageMaker","ML","ML models","Machine Learning","Amazon","docker","RAG","deep learning","Jupyter","Google","TensorFlow","Cloud Data AI","Cloud Platform"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","Amazon SageMaker","TensorFlow","Jupyter","RAG","AWS","docker"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/google-datalab-vs-amazon-sagemaker-which-cloud-platform-is-best-for-your-ml-project\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10091231,"title":"Inside Elon’s Evil ‘X Corp’","content":"Elon Musk loves media attention. This time it’s about AI and his all evil brainchild ‘X Corp’. The most-celebrated critic of OpenAI’s trending technology, who calls for a pause on training anything beyond GPT-4, is now building his own AI project. According to reports from Insider, Musk has bought thousands of GPUs in a bid to get ahead in the generative AI race. This comes just over a month after he hired two of DeepMind’s researchers, Igor Babuschkin and Manuel Kroiss, for building a rival to OpenAI “woke” ChatGPT. It seems like the one who is considered an AI-doomer for signing the petition to put a pause on training AI models beyond GPT-4, was actually just in it for his self-interest. Though it is unclear if the two researchers he hired are working for Twitter or not, it is definitely clear that they are working for Musk. Twitter Data is a Gold Mine Whatever Musk plans to do in generative AI, he has one thing that no other company, even OpenAI, has — Twitter’s data. So, if the reports are true that Musk is indeed working on a project that involves LLMs, then the data that he has access to is a gold mine. Earlier OpenAI had access to Twitter’s data, which Musk pulled away after buying the company. Seems like this was a plan all along. When it comes to Twitter as a company —  it does not exist anymore. Musk has merged Twitter with his newly formed shell firm called X Corp. This was revealed in the document of a lawsuit filed against Jack Dorsey, the former CEO of Twitter, by Laura Loomer. Musk’s long ambition of building a superapp, something similar to China’s WeChat, seems to be now taking place in the form of X Corp, an “everything app”. To mark the start of this move and the place where Twitter ceased to exist, Musk Tweeted “X” on Tuesday. He had already suggested that buying the social media company was just an “accelerant” for creating X. Buying Twitter is an accelerant to creating X, the everything app— Elon Musk (@elonmusk) October 4, 2022 Today, Musk replied to a BBC journalist on Twitter Spaces about how X Corp is definitely in the making. Previously, Musk had also tried to name PayPal as X.com, which received major backlash from the board members, eventually leading to him dropping the idea. Musk has been trying to fix Twitter for a while now on different fronts. He had complained about the search function on the platform, for which he hired George Hotz. Hotz eventually quit within one month of the three-month internship saying that he felt he “didn’t think there was any real impact“. Another thing that Twitter has been trying to win is advertisers’ interests. Companies have been criticising the change in policies after Musk took over. According to sources, if generative AI is used for the purpose of advertising, it can be beneficial for Twitter. Interestingly, when Musk recently open-sourced Twitter’s algorithm, he did not disclose the ad recommendation system. Musk’s Epic Fails Musk is a ‘walking-talking’ sinking ship. He had previously expressed his intention of turning Twitter from a loss-making to profit-generating company. For this, after taking over Twitter, Musk had been on a firing spree. He laid off 3,700 people in November, which is almost half the staff. He had also shut down the India-arm of the company, firing over 200 employees. Interestingly, a single NVIDIA GPU used for large AI models costs $10,000. If the company had to cut down employees citing unstable financial footing, funding millions of dollars for buying thousands GPUs for generative AI seems to be a messed up move. In today’s Twitter Spaces discussion, Musk expressed the importance of buying GPUs. Meanwhile Tesla’s supercomputers are only being used for Tesla’s Autopilot and Optimus Humanoid. Something doesn’t add up. Though Musk is believed to be one of the top technology experts, with a bunch of companies under his belt, the truth is entirely different. Tesla shares have been tumbling ever since he announced his stake in Twitter. SpaceX investors in 2018 had complained that Musk is paying his employees at Boring Co. using SpaceX funds. The Boring Co, Musk’s tunnel construction company, had been silent all this while after it started backing out from the projects it had promised in the cities. Similarly, Neuralink, his brain-computer interface, is yet to receive approval for human trials. More than just one of Musk’s problems, if X actually comes out as a superapp, there is very little reason to believe that it would actually work. Apart from WeChat in China, there is almost no success story of superapps. Thus, it looks like Musk might have an upper hand on OpenAI and the world when it comes to generative AI capabilities because of Twitter data, the opposite is equally possible – adding another failed project to his portfolio. Musk’s AI project will either be the best, or fail miserably.","excerpt":"After calling for a pause on giant AI experiments, Musk is now building his own generative AI.","categories":["AI Trends"],"tags":["AI Chatbot","ChatGPT","Generative AI","GPT-4"],"author_name":"Mohit Pandey","publish_date":"2023-04-12T15:00:00","publication_year":"2023","word_count":831,"keywords":["Go","ChatGPT","funding","ELT","OpenAI","AI","AI Chatbot","AWS","GPT-4","GPT","generative AI","Generative AI","R"],"extracted_tech_keywords":["AI","generative AI","ChatGPT","OpenAI","AWS","R","Go","ELT","GPT","funding"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/inside-elons-evil-x-corp\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10165449,"title":"Telecoms Get a ‘Brain’ Upgrade with AI and Cloud","content":"The telecom industry is constantly evolving, with the rise of 5G, exponential data growth, and changing customer expectations. Major providers like Reliance Jio and Bharti Airtel in India, or global players like Verizon, are embracing AI and cloud technologies to scale their operations. These innovations help build more robust networks and improve service delivery in an increasingly complex and fast-paced market. IBM recently published a report in partnership with GSMA Intelligence to highlight how telecoms can and are using cloud and AI to grow. Key insights from the report include examples from some major telecom players, showcasing how these technologies use cloud and AI to grow and improve efficiency. AI-Powered Networks for Reliance Jio, Deutsche Telecom Reliance Jio, an Indian telecommunications company, has developed its own AI platform, Jio Brain, to optimise network operations. Aayush Bhatnagar, senior vice president of Jio Platforms at Reliance Jio, explained, “Our proprietary AI platform, Jio Brain, operates across three domains: real-time AI for network optimisation, near real-time AI for IT processes, and non-real-time AI for data-driven decisions.” Communications service providers (CSPs) are increasingly adopting AI to optimise network performance, enhance efficiency, and drive innovation. Network performance monitoring, predictive maintenance, and service ticket automation are among the leading AI use cases. While generative AI is still emerging, its potential in network planning and security management is being actively explored. According to the report, when using AI for proactive root cause analysis, Jio network can prevent network-related mishaps before they occur, resulting in reduced operating expenses. Moreover, their AI-driven agent API features personalised customer experience to enhance retention and loyalty. Moreover, the energy consumption across their radio networks is optimised with the help of Jio Brain. Similarly, Deutsche Telekom (DT), a German telecommunications company, has tried its hands at generative AI. Ahmed Hafez, VP technology strategy at DT, explained that GenAI powers chatbots for resolving queries, coding assistants in network and IT development. Currently, all kinds of CSPs, big or small, use AI to a great extent. The report states that the tech front-runners use generative AI more than traditional AI, and in the case of their peers, the latter sees a dominant usage. For network performance monitoring, spam management, network planning, service ticket automation, and more, traditional AI is still the top choice. Moreover, for network planning, the use of generative AI is being explored more than usual. Going Big On Cloud for Telecoms like Bharti Airtel and Verizon As per the report, a hybrid cloud approach, combining on-premises infrastructure, private cloud, public cloud, and edge computing, is commonly adopted. While public cloud offers cost benefits, concerns about performance, security, and compliance remain. Abhishek Biswal, chief business officer of digital services at Bharti Airtel, mentioned, “Cloudification offers clear benefits—scalability, faster infrastructure deployment, and better insights with lower investments—making accelerated investment in this space inevitable.” The report also highlighted that CSPs have different choices when it comes to the type of cloud platform—public or private. “Our network is purpose-built to deliver consistently high performance…The cost reductions of public cloud are not on par with the reliability and operational agility of the private cloud,” Srinivasa Kalapala, senior vice president of technology and product development at Verizon in the US, said. At the same time, the report further states, “Smaller CSPs (with revenue ranges from $10 million to $100 million) are embracing public cloud, where they run 44% of their network functions, potentially reflecting their priority for agility and buy-versusbuild resource strategies over the risks on the minds of larger CSPs.” Data Migration is the Key to AI Adoption for Telecoms, Says Google Cloud Google Cloud shared insights about their AI-driven initiatives for the telecom partners they worked with. They emphasised that the companies need to move their data to one place to use AI effectively. Tools like BigQuery and Looker by Google facilitate consolidating data by creating scalable data lakes for analysis. This, combined with Google’s AI infrastructure, allows CSPs to extract actionable insights, personalise experiences, and develop new services. DNA, part of Telenor, and Vodafone Italy are examples of CSPs using Google Cloud to modernise their data architecture and unlock the potential of AI. What’s the Future Like? As Shyam Mardikar, president and group CTO of mobility at Reliance Jio, aptly said, “It’s not just about adopting cloud or AI—it’s about using this architecture to serve differently while rethinking how we operate and engage with customers.” For customers, it may not matter what the companies do behind the scenes. However, if AI and Cloud help make things more robust, and seamless for both customers and the workers in the telecom industry, it is a win-win situation.","excerpt":"Network performance monitoring, predictive maintenance, and service ticket automation are among the leading AI use cases.","categories":["AI Features"],"tags":["telecom"],"author_name":"Ankush Das","publish_date":"2025-03-07T14:07:43","publication_year":"2025","word_count":770,"keywords":["Go","GenAI","AI","chatbots","ML","Scala","Git","telecom","generative AI","edge computing","R"],"extracted_tech_keywords":["AI","ML","generative AI","GenAI","chatbots","edge computing","R","Go","Scala","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/telecoms-get-a-brain-upgrade-with-ai-and-cloud\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10170071,"title":"Apple ‘Assures’ No Disruption After Trump Warns Tim Cook on India Expansion","content":"Apple has assured the Indian government that there will be no change in its investment plans in India, CNBC-TV18 said, citing sources. During a business roundtable in Doha, Qatar, on May 15, 2025, U.S. President Donald Trump expressed his disapproval of Apple’s increasing manufacturing presence in India. He pushed the iPhone maker to build domestically and create jobs in the country. “I had a little problem with Tim Cook yesterday…You are coming here (US) with $500 billion, and you are building all over India. I don’t want you to build in India,” Trump said.  “You can build in India, if you want to take care of India.” In February, the iPhone maker announced its largest-ever spending commitment, planning to invest over $500 billion in the U.S. over the next four years to advance high-skilled manufacturing and support a wide range of initiatives that focus on artificial intelligence, silicon engineering, and skills development for students and workers across the country. Faced with tariffs on goods from China, Apple mitigated the potential financial impact by successfully shifting its U.S. import sources to other regions, like India. The country offers a favourable environment for manufacturing due to its supportive policies and the presence of a skilled workforce According to data from Counterpoint Research, about 16% of the iPhones made globally for Apple last year were assembled in India. Counterpoint Research vice president Neil Shah estimates the proportion will reach 20% this year. “All the stars are aligned for India to be the alternate destination to China,” Shah said. “If you look at Apple’s manufacturing strategy for iPhones, it has mostly been countries where they get benefits on manufacturing: geographical advantage, government incentives and costs, and where they have good domestic demand.” Meanwhile, according to an ET report, Apple’s wholesale revenues in India from iPhones climbed 28% to $2.53-2.92 billion in the March quarter of 2025, driven by record sales. At the Doha event, Trump further said that India was one of the highest-tariff nations in the world and that it was tough to sell in India. “But now they’ve offered us a deal where they are willing to charge no tariff, literally,” he said. Reacting to this, India’s foreign minister, S Jaishankar, stated that any trade deal must be mutually beneficial, and until that is achieved, any judgment will be premature.","excerpt":"About 16% of the iPhones made globally for Apple last year were assembled in India.","categories":["Global Tech"],"tags":["Apple","donald trump","India","iPhone","Tariffs"],"author_name":"Amisha Arya","publish_date":"2025-05-16T10:48:36","publication_year":"2025","word_count":388,"keywords":["Go","artificial intelligence","India","programming_languages:R","AI","Apple","programming_languages:Go","donald trump","iPhone","R","Tariffs"],"extracted_tech_keywords":["AI","artificial intelligence","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/apple-assures-no-disruption-after-trump-warns-tim-cook-on-india-expansion\/","complexity_score":3,"technical_depth":6,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10071212,"title":"Ian Goodfellow has joined DeepMind now – What to expect?","content":"Ian’ GANfather’ Goodfellow, the inventor of generative adversarial network recently left Apple to join top AI research lab – DeepMind. One of the best talents in the field of AI, Goodfellow has always been on the top of his game – his research work has been cited over 1.9 lakh times. Some of his popular work includes generative adversarial nets (cited 46,910 times) and research on deep learning (cited 42135 times). DeepMind has been focused and been in the limelight for projects like AlphaGo, a program that beat the world champion at Go in a five-game match, and AlphaFold, which found a solution to a 50-year-old grand challenge in biology. At DeepMind, Goodfellow is working with Oriol Vinyals’ (principal research scientist) team. The duo has collaborated earlier on research projects. In 2016, they worked on a project on the TensorFlow interface and its implementation during their time at Google. Previously in 2015, the two worked together on a neural network optimization problems research project. At DeepMind, Vinyals has been working on AI, with particular emphasis on machine learning, deep learning and reinforcement learning. With Goodfellow bringing his GAN capabilities to the DeepMind team, the best of both worlds have joined hands. Two stalwarts of the field coming together opens up a new world of opportunities. Ian, the good fellow Goodfellow headed the special project group (SPG) at Apple, which works on cutting-edge tech innovations and futuristic products. The SPG reportedly houses Apple’s self-driving car division, Project Titan, led by Tesla’s former Senior Engineering VP, Doug Field. Goodfellow had interviewed at Facebook in the fall of 2013, listening to Mark Zuckerberg philosophize about DeepMind as they strolled across the campus courtyard. Then he’d turned Zuckerberg down in favour of a job with Google Brain. The Stanford graduate got associated with Google in 2013 as an intern. He worked with a team that created a deep neural network to read address numbers from Street View imagery. The system was used to add or update the location of over 100 million houses within the first few months of its deployment. Interestingly, Goodfellow left Google twice, once to join OpenAI and the second time to join Apple. Between 2015-16, as a member of the Google Brain team, he worked on deep learning, both in terms of basic research and in terms of improving products. Furthermore, he curated a textbook on deep learning along with his PhD thesis advisors. His research is widely cited in academic articles about AI. Later, during his stint at OpenAI, the Elon Musk and Microsoft-funded AGI think tank, he was a crucial part of the research team for generative models and adversarial examples. Ian GANfather Goodfellow “If an AI can imagine the world in realistic detail – learn how to imagine realistic images and realistic sounds – this encourages the Al to learn about the structure of the world that exists,” Goodfellow said in the book Genius Makers by Cade Matz. “It can help the AI understand the images it sees or sounds it hears.” Goodfellow’s greatest claim to scientific fame is as a part of the team that invented the generative adversarial network (GAN). “If it hadn’t worked, I might have given up on the idea.” In the paper Goodfellow published on the idea, he called them “generative adversarial networks,” or GANs. Across the worldwide community of AI researchers, he became ‘the GANfather.’ Yann LeCun called GANs ‘the coolest idea in deep learning in the last twenty years.’ When Geoff Hinton heard this, he pretended to count backwards through the years to make sure GANs weren’t any cooler than backpropagation before acknowledging that LeCun’s claim wasn’t far from the truth. The WFH debate As communities rush to attain normalcy, the WFH debate has again popped up. One of the instances that triggered the conversation was the Apple machine learning head’s resignation. Famously known as the man who has ‘given machines the gift of imagination‘, the inventor of generative adversarial networks (GANs), he left Apple for DeepMind, citing the former’s lack of flexibility in work policies. This was in the backdrop of Apple calling back its employees to work from its offices. I'm excited to announce that I've joined DeepMind! I'll be a research scientist in @OriolVinyalsML 's Deep Learning team.— Ian Goodfellow (@goodfellow_ian) July 6, 2022 In an email to fellow Apple employees, Goodfellow said he strongly believes more flexibility would have been the best policy for his team.","excerpt":"At DeepMind, Goodfellow is working with Oriol Vinyals’ (principal research scientist) team.","categories":["AI Features"],"tags":[],"author_name":"Tasmia Ansari","publish_date":"2022-07-19T17:00:00","publication_year":"2022","word_count":740,"keywords":["machine learning","OpenAI","AI","neural network","ML","RAG","Aim","deep learning","TensorFlow","R"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","neural network","OpenAI","Aim","TensorFlow","RAG","R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/ian-goodfellow-has-joined-deepmind-now-what-to-expect\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10075749,"title":"The Big Deal About Vedanta-Foxconn","content":"Last year, in a landmark decision, the Indian government launched the production-linked incentive (PLI) and sanctioned Rs 76,000 crore to encourage semiconductor goods manufacturing in India. The scheme was announced in light of inadequate semiconductor chip supply globally, which also affected the sale of goods like cars, phones, and laptops. This scheme has already started to pay rich dividends as Taiwanese firm Foxconn ties up with Vedanta to build a chip manufacturing plant in Dholera, Gujarat, on 400 acres of land. While Vedanta is financing the project, Foxconn would assume the role of the technical partner. Massive opportunity Vedanta and Foxconn will invest USD 19.5 billion to set up semiconductor and display production plants in Gujarat. The joint venture has obtained several subsidies, for example, on capital expenditure and electricity from Gujarat to set up the units, as first reported by Reuters. This Rs 15.4 trillion investment is the largest ever by any group in an Indian state. https:\/\/twitter.com\/AnilAgarwal_Ved\/status\/1569573363722629120?s=20&t=AMcDEDlb6Hi9qNdCt42NWQ As per media reports, Vedanta will have a 60% equity stake in the venture, while Foxconn will own the remaining 40%. With this deal, an oil-to-metals conglomerate at the core, Vedanta hopes to diversify into the chip-manufacturing domain. This deal is expected to create 100,000 jobs in Gujarat. The project will attract companies across the electronics ecosystem value chain. This includes manufacturers of sophisticated equipment, materials like chemicals, high purity glass, photomasks, and equipment service providers. “The company will create a hub for manufacturing of iPhones and TV equipment at Maharashtra, which will be a kind of forward integration for the Gujarat JV plant,” Vedanta founder and chairman Anil Agarwal told media. State of tussle However, the deal has been embroiled in its own share of controversies. PM Modi’s home state Gujarat bested Maharashtra to win the plant location. In a statement, a Foxconn representative said that the state’s infrastructure and government support played an active role in increasing the confidence to set up a semiconductor factory there. A massive political controversy erupted when Gujarat was announced the winner. Opposition parties in Maharashtra are crying foul over the allocation, and Maharashtra Navnirman Sena (MNS) chief Raj Thackeray has sought a probe into the process. Notably, Shiv Sena, along with Congress and NCP, was recently ousted from power, post which the current chief minister Eknath Shinde and his party came to power. Through its mouthpiece publication Saamna, MNS alleged that the deal went from Maharashtra to Gujarat as the BJP sought favour from CM Shinde for instating him in his current role. However, Vedanta chairman Agarwal has said that Gujarat was chosen based purely on professional and independent advice. Vedanta-Foxconn has been professionally assessing site for a multi-billion dollar investment. This is a scientific and financial process which takes several years. We started this about 2 years ago. (1\/4)— Anil Agarwal (@AnilAgarwal_Ved) September 14, 2022 Gujarat offered Rs 28,000 crore capital subsidy to Vedanta-Foxconn for the plant, against Maharashtra’s offer of Rs 40,000 crore. Further, Gujarat offered 200 acres of land to both companies at Dholera at 75% of the going rate, while Maharashtra offered 1,100 acres of land in Talegaon Phase IV in Pune district – 700 acres of this was offered at 75% of the current rate, and remaining 400 acres for free. More subsidies & ‘freebies’ Recently, the Union cabinet approved a uniform incentive of 50% of the project cost. This is towards setting up the semiconductor display and compound semiconductor fabrication unit and against the earlier incentive range 30-50% for the three schemes. The incentives for setting up semiconductor fabrication were based on the chip size, but the incentives for display fabrication and compound semiconductor fabs were largely 30% of the total cost of the project. This new announcement has drawn criticism from experts, comparing the subsidies with freebies. Many doubt if projects like Vedanta-Foxconn and ISMS would swallow a large chunk of the government’s subsidiary allocation. This may even lead to ensuring financial incentives for a select few. Raghuram Rajan, former RBI governor, has questioned the validity of these subsidies, asking what would stop these manufacturers from moving over to other countries when the subsidies and incentive schemes end? He said, “Manufacturers could also continue production but also demand continued tariff and subsidy protection.”","excerpt":"Vedanta and Foxconn will be investing USD 19.5 billion to set up semiconductor and display production plants in Gujarat","categories":["IT Services"],"tags":["foxconn","PM Modi","Semiconductor India","Vedanta"],"author_name":"Shraddha Goled","publish_date":"2022-09-26T12:00:00","publication_year":"2022","word_count":707,"keywords":["Go","API","programming_languages:R","AI","programming_languages:Go","RAG","Vedanta","Ray","Semiconductor India","R","foxconn","PM Modi"],"extracted_tech_keywords":["AI","Ray","RAG","R","Go","API","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/the-big-deal-about-vedanta-foxconn\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10013768,"title":"Microsoft Is Going Big On Reinforcement Learning. Here’s How","content":"When it comes to research in new-age technologies, Microsoft has been striving hard to stay ahead of its competitors. From recommendations to gaming, the tech giant has been using popular techniques like reinforcement learning to create efficient products for customers that match their interests. The foundational work in reinforcement learning (RL) started back in 1992, in which the researchers worked on Simple Statistical Gradient. This year, the tech giant has made significant contributions in the ongoing AI conference known as NeurIPS 2020. The three key research areas that are being focussed this year include batch reinforcement learning; a strategic exploration that has given rich observations; and representation learning. John Langford, the partner research manager at Microsoft Research, stated that all of the reinforcement learning research at the tech giant falls into two specific criteria. One is to solve challenges that the customers are bringing in, and the second is to understand the foundations that the tech giant can utilise in order to build replicable, reliable solutions. Below are some of the new AI solutions based on reinforcement learning that the tech giant unveiled this year. Personalizer Personalizer is an AI service that delivers a personalised, relevant experience for every user. The service boosts usability and user satisfaction with reinforcement learning-based capabilities and prioritises relevant content, layouts, and conversations through an easy-to-use API. Unlike recommendation engines that suggest a few specific options from a large catalogue, Personalizer presents a suitable outcome for a user, every time they interact with the app. A part of Azure Cognitive Services within the Azure AI platform, Personalizer is primarily used in research labs. But now, the AI system is making its way into more Microsoft products and services, from the cognitive services where the developers can plug into apps as well as websites to autonomous systems in order to refine the manufacturing processes. Microsoft has been using this system internally to select the right offers, products and content across Windows, Edge browser and Xbox. Know more here. Metrics Advisor Recently, the tech giant announced the preview of Metrics Advisor, which is a new Azure Cognitive Service in order to address the need for metrics intelligence. The service ingests data from various sources, using machine learning to automatically find anomalies from sensors, products, and business metrics, and provides diagnostics insights. It uses reinforcement learning to incorporate feedback and make models more adaptive to a customer’s dataset. Metrics Advisor helps in detecting definite anomalies in sensors, production processes or business metrics. Know more here. Project Paidia Project Paidia is a research collaboration between the Game Intelligence group at Microsoft Research Cambridge and game developer Ninja Theory. This project aims to drive the SOTA research in RL to enable novel applications in modern video games, in particular, agents that learn to collaborate with human players. According to a blog post, Project Paidia focuses on learning a particularly challenging type of behaviour such as collaboration with human players. Know more here. AI Robot Control System Last year in October, Microsoft Research joined hands with Sber Robotics Laboratory to develop a unique AI control system that teaches robots to manipulate physical objects of unstable shape in almost the same way that humans do. The work on the project was carried out at Sber’s Robotics Laboratory in Moscow in 2019 and collectively lasted over a year. Finishing the project this year, the researchers at the Sber and Microsoft team used deep reinforcement learning and machine teaching techniques first to train the AI agent in a simulated environment, where it could explore different strategies and learn what worked best. Once deployed in real-world working conditions, the robotic system was successfully able to unload the coin bags on the first try 95% of the time. Know more here.","excerpt":"When it comes to research in new-age technologies, Microsoft has been striving hard to stay ahead of its competitors. From recommendations to gaming, the tech giant has been using popular techniques like reinforcement learning to create efficient products for customers that match their interests.  The foundational work in reinforcement learning (RL) started back in 1992, […]","categories":["Global Tech"],"tags":["big data platform c++","big data video games","Microsoft","Reinforcement Learning","reinforcement learning algorithms","reinforcement learning environment","reinforcement learning models","Reinforcement Learning Systems"],"author_name":"Ambika Choudhury","publish_date":"2020-12-11T11:00:00","publication_year":"2020","word_count":625,"keywords":["Reinforcement Learning Systems","Go","API","reinforcement learning algorithms","machine learning","Reinforcement Learning","cloud_platforms:Azure","AI","programming_languages:R","R","programming_languages:Go","reinforcement learning environment","big data platform c++","Aim","reinforcement learning models","big data video games","Azure","Microsoft"],"extracted_tech_keywords":["AI","machine learning","Aim","Azure","R","Go","API","cloud_platforms:Azure","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/microsoft-is-going-big-on-reinforcement-learning-heres-how\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":26124,"title":"Why Cassandra And Hive Are The Best Prospective Big Data Tools For ML","content":"Over the past few years, there has been a rise in database systems and their tools owing to the fact that big data and machine learning fields are growing parallelly. There is no dearth in the variety of tools available to users to handle data systems. In addition, the progress of distributed file systems and cloud computing have made an impact on the way database systems work. Platforms such as Apache Hadoop and Apache MapReduce have witnessed stellar developments in the recent years, to effectively meet the demand of computing enormous amounts of data. In fact, Hadoop has grown so big that the framework itself is designed into a software library that offers a host of database tools. The applications of these tools span from cloud computing to data mining, and now has made its way into ML. In this article, we discuss two extensions of Hadoop known as Cassandra and Hive, and look at how their functions help with ML. Apache Cassandra Cassandra is a distributed database management system developed by Apache Software Foundation in 2008. It uses techniques based on NoSQL and is an open source software. The key features of the software are: Decentralised system Distributed deployment High application scalability Fault tolerance. Tunable consistency MapReduce Support A separate query language called Cassandra Query Language It manages data in the form of clusters which are interconnected to thousands of nodes spread across data centres. It is also known as ‘column-oriented database’ in NoSQL, wherein the data is stored in a column-by-column fashion in contrast to the row-based approach in traditional database systems. This is the reason it has lesser I\/O operations for storing data. Cassandra has mainly been used in big data applications which use real-time data such as those from sensor components or from social networking websites. In addition, Cassandra has a decentralised architecture, which means function modules such as data partitioning, replication, scaling and failure handling are present separately, and work in tandem. This means any node can take up any data processing operation. Cassandra’s key advantage lies in its ability to run on less powerful hardware. The tool performs read\/write functions quickly on hundreds of gigabytes of data. The architecture behind Cassandra is loosely based on Amazon’s Dynamo, which implements a key-value database system. Since ML involves iterative tasks with significantly large data, Cassandra can be the perfect tool for executing large datasets with good throughput. Apache Hive Hive is primarily a data warehousing tool which is based and built on the features of Hadoop. It uses a SQL-like syntax for queries in managing data to and fro from the database. The first official and stable version of the software was released in 2012 by Apache. Mainly used for data analysis, Hive supports functions such as data summarisation and ad-hoc querying conveniently. Hive has the following features: Easy data access through SQL Support for a variety of data formats Distributed file storage system Query execution through data processing tools Query retrieval Originally developed as a translation layer for Hadoop MapReduce, it uses its SQL-like language to interpret direct acyclic graphs in MapReduce therefore reducing the burden of writing long codes to handle data in the storage systems. Furthermore, Hive supports popular programming languages such as Java, Python, C++ and PHP. Hive is not exactly a database system, and so it is generally not used in critical systems which involve real-time transactions such as bank transactions or online ticketing. One-On-One Comparison [su_divider top=”no” size=”1″] Cassandra Hive Function Distributed database system that has data stored in clusters. Data warehousing tool which relies on features of Hadoop Website http:\/\/cassandra.apache.org\/ https:\/\/hive.apache.org\/ Current Stable Release 3.11.2 \/ February 19, 2018 2.3.0 \/ July 19, 2017 Written in Java Java Supported Operating Systems Windows, OSX, Linux Almost all OS Open Source Availability Yes Yes Supported Programming Languages Java, JavaScript, Python, Perl, Ruby, Scala, C++, Haskell Java, Python, PHP, C++ MapReduce Support Yes Yes Query Language CQL Specific SQL statements API Support Through CQL Through JDBC, ODBC Comments Since both Cassandra and Hive take on huge amounts of data, both of them look ideal for ML applications. ML algorithms are usually iterative in function. These iterative computations demand higher power as well as quick data handling capabilities. Also, before using these software, care should be taken that the data is relevant as well as of top quality for the ML project. It should be noted Cassandra and Hive are specifically used in big data applications. Therefore, ML must deal with ramifications involved in big data carefully without compromising user experience. Contrastingly, for ML, more data means better output that gives useful insights into the problem.","excerpt":"Over the past few years, there has been a rise in database systems and their tools owing to the fact that big data and machine learning fields are growing parallelly. There is no dearth in the variety of tools available to users to handle data systems. In addition, the progress of distributed file systems and […]","categories":["AI Features"],"tags":["Big Data","database systems","informatica","nosql","SQL"],"author_name":"Abhishek Sharma","publish_date":"2018-07-04T11:43:19","publication_year":"2018","word_count":773,"keywords":["informatica","machine learning","TPU","AI","cloud computing","ML","JavaScript","RAG","Python","database systems","SQL","nosql","Big Data","R"],"extracted_tech_keywords":["AI","machine learning","ML","RAG","cloud computing","TPU","Python","R","SQL","JavaScript"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/why-cassandra-and-hive-are-the-best-prospective-big-data-tools-for-ml\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10163882,"title":"ChatGPT Has The Potential To Compete With Mental Health Experts, Says Study","content":"Since the expansion of use cases for ChatGPT, there have been discussions regarding whether this signifies the end for human therapists and if GPT-4 surpasses psychologists in terms of intelligence. A recent study suggests that ChatGPT can potentially rival human therapists in certain aspects of psychotherapy. The research, which appears in PLOS Mental Health, explores the capabilities of ChatGPT in comparison to responses from a panel of mental health experts. The study involved a large sample of 830 participants who were asked to differentiate between responses to couple therapy vignettes written by ChatGPT 4.0 and those written by clinical psychologists, counselling psychologists, marriage and family therapists, and a psychiatrist. The results indicated that participants could barely tell the difference between responses written by ChatGPT and responses written by a therapist. Moreover, the AI-generated responses were generally rated higher in key psychotherapy principles. When gauging the results regarding stats, the study mentions, “Identification within authors was poor, with participants correctly guessing that the therapist was the author 56.1% of the time and participants correctly guessing that ChatGPT was the author 51.2% of the time.” It further states, “Between authors, participants were only able to correctly identify therapists 5% more often than ChatGPT (56.1% versus 51.2%, respectively).” The research emphasises the linguistic distinctions between ChatGPT’s replies and those of therapists, noting that ChatGPT exhibits a high level of empathy. The study suggests that “ChatGPT has the potential to improve psychotherapeutic processes.” The study notes that one practical application of GenAI is in web-based interventions for couples, where an automated chatbot could provide evidence-based support, expanding the reach and dissemination of such programs. In addition to the interesting success rate, the study acknowledges limitations, including the absence of a therapeutic context and the need for supervision during the training process to ensure participant safety. The authors call for continued research to improve the usefulness of psychotherapeutic applications using chatbots like ChatGPT, potentially placing them in the hands of individuals who need them the most.","excerpt":"Study results indicated that participants could barely tell the difference between responses written by ChatGPT and responses written by a therapist.","categories":["AI News"],"tags":["ChatGPT"],"author_name":"Ankush Das","publish_date":"2025-02-17T18:01:21","publication_year":"2025","word_count":331,"keywords":["ChatGPT","GenAI","API","AI","chatbots","RPA","GPT","R","llm_models:ChatGPT","llm_models:GPT"],"extracted_tech_keywords":["AI","GenAI","ChatGPT","chatbots","R","API","GPT","RPA","llm_models:GPT","llm_models:ChatGPT"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/chatgpt-has-the-potential-to-compete-with-mental-health-experts-says-study\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10065234,"title":"The double edge sword of ads powered by augmented intelligence &#8211; Rachita Pandya Shah, Head of Product, Gojek","content":"Rachita Pandya Shah, Head of Products – Ads and Promotions at Gojek, has spoken about the Role of women technopreneurs in shaping the future of ads powered by augmented intelligence at Analytics India Magazine’s Women in AI Conference: The Rising 2022. In her talk, Rachita provides an insight into the future of advertisement industry culture and reveals her viewpoint on the age-old question, can women in tech contribute differently in shaping a double-edged sword of a technology that this is emerging to be? Rachita starts off by delving into the basics of advertisement. She states that ads are nothing but just staple. She refers to an ad describing the popular cartoon Flinstones with a modern consumer product. Rachita talks about relating with a cartoon about the stone age having modern items like cars, etc. This example is what we call influencer marketing. “Ever since humans have learned how to communicate, ads have been around. Ads are a staple. We need ads to know what are the products and new services that are out there. What are our friends or people in our network using?” says Rachita. The concept of right and wrong “A lot of our actions, whether sponsored or non sponsored, affect the AI that suggests with an ad. But is it that simple? Who decides what’s a good or a bad ad? What’s harmful? What’s an annoying ad?” questions Rachita. She further provides an analogy by referencing a popular mythological show and talking about the leader of the seven kingdoms being the one who has the collective wisdom of the present, past and future leaders. Similarly, AI is supposed to aggregate data from many sources and predict the best decision. She states that brand is the AI, and the choice of the people is augmented intelligence. Collective wisdom has its merit, but it is not enough to make a great ad. That’s why augmented intelligence plays a huge role, as it caters to the feelings of the individual in comparison to the collective data. Sheldon’s algorithm for friendship Rachita states that collective wisdom is biased as the input accumulated in the data also contains bias. She emulates this concept with the example of “Sheldon’s algorithm for friendship” from the famous show, the Big Bang Theory. Rachita states, “Any situation with friends could be tackled through this algorithm. Sheldon generally starts off by offering the friend a hot beverage. But, what if the friend prefers cold beverages to hot ones? Does Sheldon take that input?” Rachita explains that collective wisdom acts in a similar manner, and that’s why ads can be annoying. The collective wisdom does not take into consideration the user’s input. The Yin-Yang imbalance Rachita opens up about the imbalance existing between the ratio of men and women in any enterprise workforce. It’s not that women are just needed in tech. Women need to represent their community in all roles of an enterprise. Furthermore, women need to nurture their voices by making themselves heard on a platform to promote women’s empowerment. “Today, a lot of these conversations revolve around diversity. It is one of the most glaring problems present. But, this problem is not just present in developing countries like India and Bangladesh. A global dialogue needs to be set for the holistic balance for women technopreneurs from all walks of life, like ethnicity, age, religion or sexual orientation, etc. And by technopreneurs, I don’t just mean founders. If you’re a woman and you’re into tech, then you’re a technopreneur for me,” says Rachita. She further states that this imbalance can only be nullified by women raising their voices and making their points heard. She ends her talk by urging women around the conference to be the voice of all minorities and build an ecosystem of cultural inclusion and diversity.","excerpt":"Rachita talks about the imbalance in the workforce and urges women around the conference to be the voice of all minorities and build an ecosystem of cultural inclusion and diversity.","categories":["AI Features"],"tags":["AGI","AI (Artificial Intelligence)","AI conference","Analytics India Magazine","gojek","Marketing"],"author_name":"Kartik Wali","publish_date":"2022-04-19T12:28:34","publication_year":"2022","word_count":633,"keywords":["Go","API","AI conference","programming_languages:R","AI","gojek","programming_languages:Go","Analytics India Magazine","RAG","AGI","analytics","R","Marketing","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","analytics","RAG","R","Go","API","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/the-double-edge-sword-of-ads-powered-by-augmented-intelligence-rachita-pandya-shah-head-of-product-gojek\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10041117,"title":"A Beginner’s Guide To Intel oneAPI","content":"“ Intel’s oneAPi enables developers to work without the exhaustion of juggling with different languages, tools, libraries and different hardware.” Intel’s oneAPI is an open, accessible and standards-based programming system that enables developer engagement and innovation across multiple hardware architectures, including CPUs, GPUs, FPGAs, AI accelerators, and more. These tools all have very different properties and are thus used for various operations–which oneAPI attempts to simplify by unifying them under one model. A persisting problem faced by developers even today is the volume of programming environments our increasingly digital world offers. Disparate programming environments disable time-saving strategies such as code reusing and turn into real hurdles for software developers. Intel unveiled oneAPI at its Supercomputing event in 2019 as a part of its software-first strategy. The model signalled Intel’s ambitions to have a unified programming framework as a solution to limiting proprietary programming platforms. Source: Intel oneAPi enables developers to work without the exhaustion of juggling with different languages, tools, libraries and different hardware.  Here are a few top features of oneAPI: Data Parallel C++ To make oneAPI possible, Intel developed the data-parallel language DPC++ (or Data Parallel C++). This language, which is open-sourced, incorporates C++ and the Khronos Group’s SYCL to support data parallelism and heterogeneous programming. The language is reportedly similar to how standard C++ works, with added extensions to it, making it easy for developers to transition to DPC++. Libraries oneAPI allows data parallelism by leveraging two types of programming: API-based programming and direct programming. Within API-based programming, the algorithm for this parallel application development is hidden behind a system-provided API. oneAPI defines a set of APIs for commonly used data-parallel domains and provides library implementations across various hardware platforms. This enables a developer to maintain performance through multiple accelerators with minimal coding and tuning. These libraries are- oneAPI DPC++ Library (oneDPL) oneDPL has algorithms and functions to speed up DPC++ kernel programming. The oneDPL library follows the C++ standard library’s functions and includes extensions to support data parallelism and extensions to simplify data-parallel algorithms. oneAPI Math Kernel Library (oneMKL) oneMKL is used for fundamental mathematical routines in high-performance computing and applications. This functionality is divided into dense linear algebra, sparse linear algebra, discrete Fourier transforms, random number generators, and vector math. oneAPI Data Analytics Library (oneDAL) oneDAL helps speed up big data analyses by providing optimised building blocks for algorithms for different stages of data analytics—preprocessing, transformation, analysis, modelling, validation, and decision making. oneAPI Deep Neural Network Library (oneDNN) oneDNN, allows neural network functions for deep learning applications and frameworks. oneAPI Collective Communications Library (oneCCL) oneCCL enables primitives for communication patterns that take place in deep learning applications. oneAPI Threading Building Blocks (oneTBB) oneTBB is the threading and memory management library. It is used to specify logical parallelisms in algorithms, which it does through mapping that parallelism onto execution threads. oneAPI Video Processing Library (oneVPL) oneVPL is a programming interface for video encoding, decoding and processing for portable media pipelines on CPUs, GPUs and other platforms. Source: oneapi.com Despite this, some problems are not very well-suited to API-based programming. This is because solutions to these may not have standard solutions or require a customisation level that users cannot quickly implement via a library. Here, direct programming needs to be used, which involves the programmer directly writing efficient algorithms for parallel architectures. Level Zero A final key component enabling oneAPI is its low-level hardware interface. Level Zero is a low-level direct-to-metal interface for DPC++ and oneAPI’s libraries. Level Zero supports broad language features such as unified memory, virtual functions, and I\/O capabilities whilst providing direct controls required by high-level APIs such as kernel submission, memory allocation and inter-process sharing of data. Moving to unified developments like oneAPI eliminates the need to work with different codebases, tools and programming languages. This flexibility is necessary given the use of multiple architectures by many programmers and data experts. Today, oneAPI is very well received with many large companies employing oneAPI’s elements—including Microsoft’s Azure AI and TensorFlow. Codeplay has also released an open source layer to allow oneAPI to run on Nvidia GPUs. oneAPI lets developers get away with the hassles of choosing vendor-specific libraries and tools. This makes oneAPI a powerful tool for developers today.","excerpt":"“ Intel’s oneAPi enables developers to work without the exhaustion of juggling with different languages, tools, libraries and different hardware.” Intel’s oneAPI is an open, accessible and standards-based programming system that enables developer engagement and innovation across multiple hardware architectures, including CPUs, GPUs, FPGAs, AI accelerators, and more. These tools all have very different properties […]","categories":["Deep Tech"],"tags":["AI Tool","Guide","Intel","oneAPI"],"author_name":"Mita Chaturvedi","publish_date":"2021-06-02T11:00:00","publication_year":"2021","word_count":705,"keywords":["Guide","Go","AI","neural network","R","RAG","C++","deep learning","analytics","AI Tool","TensorFlow","Azure","oneAPI","Intel"],"extracted_tech_keywords":["AI","deep learning","neural network","analytics","TensorFlow","RAG","Azure","R","Go","C++"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/a-beginners-guide-to-intel-oneapi\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10053808,"title":"Facebook AI Releases XLS-R, Self-Supervised Model For Speech Tasks","content":"Facebook recently announced the release of XLS-R, a new self-supervised model for a variety of speech tasks. XLS-R substantively improves upon previous multilingual models by training on nearly ten times more public data in more than twice as many languages. Trained on more than 436,000 hours of publicly available speech recordings, XLS-R is based on wav2vec 2.0, Facebook AI’s approach to self-supervised learning of speech representations, and nearly ten times more hours of speech than the best previous model it released last year, XLSR-53. Utilizing speech data from different sources, ranging from parliamentary proceedings to audiobooks, it has been expanded to 128 different languages, covering nearly two and a half times more languages than its predecessor. XLS-R was evaluated on four major multilingual speech recognition benchmarks, where it outperformed prior work on most of the 37 languages tested; specifically, it was tried with five languages of BABEL, ten languages of CommonVoice, eight languages of MLS, and the 14 languages of VoxPopuli. Image Source: Facebook AI The model was also evaluated for speech translation, where audio recordings were directly translated into another language. Facebook has always been interested in models that can perform multiple tasks, so it simultaneously fine-tuned XLS-R on several different translation directions of the CoVoST-2 benchmark. The result is a single model that can translate between English and up to 21 other languages. Image Source: Facebook AI The model leads to very large improvements on low-resource language directions, such as Indonesian-to-English translation, where the accuracy in terms of BLEU doubles on average — a very large step forward in improving translation of spoken language. An increase in the BLEU metric means automatic translations have more overlap with the translations produced by a human tackling the same task. XLS-R demonstrates that scaling cross-lingual pretraining can further improve performance for low-resource languages. It improves performance for speech recognition and more than doubles the accuracy of foreign-to-English speech translation. XLS-R is an important step toward a single model that can understand speech in many different languages, and it is the largest effort we know of to leverage public data for multilingual pretraining.","excerpt":"XLS-R substantively improves upon previous multilingual models by training on nearly ten times more public data in more than twice as many languages.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","Data Science","Data Scientist","Deep Learning","Facebook AI","Facebook AI research","facebook analytics","Facebook latest","Machine Learning","NLP","NLP models","NLP research","Python"],"author_name":"Victor Dey","publish_date":"2021-11-19T12:10:36","publication_year":"2021","word_count":351,"keywords":["NLP models","R","RAG","NLP","Facebook AI research","facebook analytics","Data Science","NLP research","AI","ML","Machine Learning","self-supervised learning","Facebook AI","programming_languages:R","Python","Deep Learning","Data Scientist","AI (Artificial Intelligence)","Facebook latest"],"extracted_tech_keywords":["AI","ML","RAG","R","self-supervised learning","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/facebook-ai-releases-xls-r-self-supervised-model-for-speech-tasks\/","complexity_score":2,"technical_depth":6,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":40853,"title":"Analytics &#038; Data Science Industry In India: Study 2019 By AIM &#038; Praxis Business School","content":"The Big Data and Analytics industry has gone through major disruptions in the last five years. In our annual study — Analytics & Data Science Industry in India 2019 with Praxis Business School, we summarise where the domestic analytics market is heading and the state of analytics outsourcing in India. We have seen a marked shift in which Analytics and Data Science has moved beyond supporting new business models and operational effectiveness to shaping business performance. What was once a competitive edge — advanced analytics is no longer limited to a few leading companies or data-intensive industries. There is also a concerted shift in IT spending with the amount of spending related to analytics and BI going up significantly. In our 2018 report, we estimated the Analytics, Data Science and Big Data industry in India to be $2.71 billion annually in revenues. In 2019, the analytics industry has grown to $3.03 billion in size and is expected to double by 2025. Here are a few of our observations — enterprises across the board are building up their analytics capabilities and analytics has become the biggest revenue driver for Indian IT bellwethers, captives and domestic firms within their digital portfolio. Also, the shift towards intelligent automation, AI and machine learning, is changing the face of data and analytics services. However, despite the momentum in the market, there is an overall dip in the growth rate from a year ago. The domestic analytics industry has grown by 12% this year. Overall, the analytics industry currently accounts for almost 21% of the whole IT\/ITeS industry in India. The analytics ecosystem in India is led by Domestic IT\/ITes firms that are leading players in the analytics outsourcing market. The outsourcing industry is led by Indian IT bellwethers like TCS, Wipro, Genpact, Tech Mahindra, HCL Infosystems among others that form 35% of the analytics outsourcing market. The report is designed to capture the cumulative size of the analytics market in India, outsourcing market in India and domestic analytics industry in India, the size and growth of analytics and data science market by segment and sector, outsourcing market in India and areas where we can expect to see the maximum growth. It is intended for technology leaders, management consultants, Business Unit heads tasked with building data science units within their company. The report is geared towards IT decision makers and business leaders to understand the developments within data and analytics domain and understand market opportunities, the key drivers, challenges the dynamics at play in this space. Demystifying The Analytics Market The cumulative analytics market in India stands at $30 Billion Outsourcing is the main driver of revenue for Indian vendors accounting for $27 billion in revenue The domestic analytics market stands at $3.03 billion in size and is expected to double by 2025 Analytics Outsourcing Industry Indian IT bellwethers are the leading revenue generators in the analytics market, with TCS clocking nearly $2 billion from analytics services. As companies expand their digital portfolio, India Inc. led by TCS, Genpact, Wipro are investing heavily around data and analytics, increasing the range of services and also the scope of projects to corner a bigger chunk of share. The outsourcing industry in India for Data analytics services and products is almost $27 billion. These include combined numbers from exports as well as captives and back offices for all analytics related services and product sales: Analytics industry currently accounts for almost 21% of the whole IT\/ITeS industry in India Tata Consultancy Services gets nearly $2 Billion in revenue a year from analytics, making it the largest chunk of its digital revenue Over the years, a lot of existing services have been clubbed into analytics, including reporting, data warehousing etc. Domestic IT\/ITes firms are the biggest players in the analytics outsourcing market. They include large players like TCS, Wipro, Genpact, Tech Mahindra etc. They form almost 35% of the analytics outsourcing market. MNC IT providers come at a very close second at 32%. They include organizations like Accenture, Cognizant, IBM etc. Captives form 12% of the analytics back-office market whereas boutiques analytics firm, though large in numbers, form just 11% Consulting firms account for 10% of the analytics market and include firms like McKinsey, Deloitte etc. In terms of geographies served, almost 47% of analytics revenues in India come from exports to the USA UK comes a distant second at 9.6% of revenues Domestic Analytics Industry While growth in Data Science industry remains strong, the growth rate for domestic analytics industry has plateaued over the last one year. The overall Data Analytics industry in India is currently estimated to be $3.03 Billion annually in size This is the overall consumption of analytics within Indian organizations while the outsourcing market is projected to be bigger The analytics industry is expected to double in size by 2025 Sector Type for Analytics Industry The analytics adoption is led by India’s BFSI sector which has also broadened its portfolio by piloting several key AI and ML-related initiatives. After finance and banking, marketing and advertising sector is the second biggest adopter of analytics. In terms of Sector type, Finance & Banking continues to be the largest sector being served by analytics in India. Overall, 36% of the analytics market size in India comes from Finance & Banking Marketing & advertising comes second at 25%, followed by the E-commerce sector at 15% of analytics revenues in India Analytics Industry By Cities Bengaluru is the biggest hub for analytics with leading MNCs and captives setting up analytics CoEs in the city. This year, Bengaluru has bested Delhi NCR by capturing the largest analytics market. This is also reflected in funding with Bengaluru startups boasting of the biggest war chest. Bengaluru has clearly emerged as the biggest centre for analytics in India with almost 28% of all analytics market being captured by the city Till last year, Delhi NCR was the largest area for analytics market size; this year it has slipped to the second position at 25% of India’s analytics market Mumbai comes a distant third at 18% Pune, Hyderabad & Chennai have similar size at almost 9% Analytics Professionals In India Upskilling has emerged as one of the key corporate trends within leading IT companies. This has resulted in a strong demand for analytics training and data literacy across functions. In fact, this trend is also reflected at the senior leadership level with hands-on courses that impart specialised training.Work Experience The average work experience of analytics professionals in India is 8 years; slightly up from 7.9 years from last year Around 22,000 freshers were added to analytics workforce in India this year; up from 16,000 freshers last year. Hiring for freshers has increased by 37% Almost 40% of analytics professionals in India have work experience less than 5 years, the same as last year Company Size When it comes to distribution by company type, the largest number of analytics employees are absorbed by large IT firms followed by mid-size organisations. This will continue to reflect in hiring trends with captives having the maximum requirement for analytics talent. Almost 38% of analytics professionals in India are employed with large-sized companies – with more than 10K total employee base Mid-size organisations (total employee base in the range of 200-10K) employ 33% of all analytics professionals in India Startups (less than 200 employee base) employees 29% of analytics professionals in India Conclusion While the data and analytics market will continue to grow in India, there will also be an emergence of industry-specific analytics use cases. Today, advanced analytics is at the core of future platforms, solutions and applications. While it has become pervasive, data and analytics are now increasingly augmented by AI\/ML with analytical models being auto-generated models. Across the analytics stack, tools are becoming more user-friendly and we are seeing an increased role of machine learning and AI in automating aspects of data science and ML models for development and deployment. Increasingly, companies and captives are growing their business beyond analytics and broadening their portfolio with verticalized AI-based solutions. Today, IT firms and even boutique analytics firms are building an ML practice and developing vertical-specific AI solutions. The product and service suite is increasingly reflecting an AI-shift and companies that haven’t yet embraced this next wave of disruption will definitely re-orient their practice around it. Download the study here. Download The Report [attachments include=”40912″]","excerpt":"The Big Data and Analytics industry has gone through major disruptions in the last five years. In our annual study — Analytics & Data Science Industry in India 2019 with Praxis Business School, we summarise where the domestic analytics market is heading and the state of analytics outsourcing in India. We have seen a marked […]","categories":["AI Features"],"tags":["AI Jobs","study data science","Types of Databases"],"author_name":"Richa Bhatia","publish_date":"2019-06-18T09:45:14","publication_year":"2019","word_count":1395,"keywords":["data science","Go","API","machine learning","AI","Types of Databases","ML","AI Jobs","Git","RAG","analytics","study data science","R"],"extracted_tech_keywords":["AI","machine learning","ML","data science","analytics","RAG","R","Go","Git","API"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/analytics-data-science-industry-in-india-study-2019-by-aim-praxis-business-school\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":53637,"title":"Case Study: How This Glass Manufacturing Company Leveraged Azure IoT To Get Real-Time Visibility","content":"IoT is connecting all devices in our home, from appliances to equipment, and now it’s increasingly moving its base towards its industrial use with IoT sensors improving production efficiency and quality. Glass manufacturing sector is one such artistic industry that has evolved a lot since its inception in the prehistoric times. From manual glass fusion to using modern technologies for producing glass containers, the industry has come a long way. Standing out of the ever-evolving glass manufacturing and packaging industry, Piramal Glass is one prominent name that has used an army of IoT sensors and advanced data analytics in the cloud to improve their manufacturing operations and speed up their decision making. The Challenge With manufacturing facilities in India, Sri Lanka, and the US, and offering end-to-end glass packaging solutions in over 50 countries, Piramal Glass is one of the global companies focused on designing, producing, and decorating premium glass bottle packaging for pharmaceutical, cosmetics, and food and beverage industry. And, such a vast operation demands a commitment from the company to constantly add value for its customers. “Although we are a B2B company, we always wanted to provide our customers, many of whom are millennials, a B2C experience, which includes real-time visibility, anytime\/anywhere service, a faster delivery, and quality of products. And to provide such an experience, our manufacturing operations required real-time visibility of operations to improve the efficiency,” said Vijay Shah, Vice Chairman at Piramal Glass. When further asked, he explained, how glass manufacturing could be a complicated process, which involves various phases such as batch preparation, melting raw material in the furnace, bottle forming, treatment, and packing. In also includes fixed costs and capacities that are very sensitive to variations in demand, and therefore, it was vital to optimise the operations to drive profitable growth. The company produces 1,375 tons of glass units per day across 60 production lines, all of which run 24\/7 producing tons of data. Historically, all this data was being manually collected for each lot, and then was captured in paper-based logbooks by the plant personnel, which on a longer run, proved challenging to analyse and did not provide timely insights to improve production. And to improve the production efficiency and minimise defects, it was required for the company to transform its manufacturing process with technology. In turn, it was also necessary to uptime the optimal performance of their essential equipment, such as furnaces, inspection machines and compressors; to reduce the energy costs. The Solution To set new benchmarks in the glass manufacturing and packaging industry, Piramal Glass turned towards digital transformation to stay ahead of its game. “The first step in our transformation journey was to set out on a data crusade where we capture real-time process data,” said Shah. As an early adopter of technology, Piramal Glass leveraged Azure IoT to get real-time visibility into its line manufacturing operations and to analyse production line losses at various stages. Using Azure IoT Hub, Microsoft helped Piramal Glass in acquiring data from sensors on production lines to identify quality parameters at each stage and get insights on line efficiencies in real-time. RTMI at Kosamba “Microsoft’s Azure IoT platform enabled Piramal Glass to connect and monitor their equipment to gain real-time visibility into operational data that was previously unavailable. The technology integration was designed for fast and easy set-up, to be built on the existing infrastructure,” said Sashikumar Sreedharan, General Manager and Managing Director at Microsoft India. “To facilitate this transformation, a Microsoft partner, Precimetrix was brought in for their plant monitoring system that is hosted on Microsoft Azure. The sensors on high-speed conveyor lines were interfaced with data acquisition devices that record key metrics, as the bottles move along the production line. This data gets aggregated on an edge gateway, and then pushed to the plant monitoring system on the cloud,” further added. “A custom solution, which we call, Real-Time Manufacturing Insights (RTMI), was then developed on top of the platform to provide stage-wise losses, production reports, quality control workflows, and role-specific KPIs on PCs and smartphones of the plant personnel. At any time when there is an anomaly detected or the production efficiency drops, actionable alerts are being sent through SMS, email or push notifications.” In short, the developed RTMI leverages digital technologies to enable real-time monitoring and advanced insights for end-to-end plant operations. It democratised real-time information available to all plant stakeholders of the company, allowing them to take faster decisions. RTMI was piloted in three production lines in the Kosamba plant near Surat, Gujarat. Existing and retrofitted sensors were connected to the cloud with Azure IoT Hub and analysed through RTMI. “Since the benefits were evident, RTMI was scaled to 46 production lines across Piramal Glass plants in India, Sri Lanka, and the US, within only six months,” said Sreedharan. Benefits The cloud-based and multi-channel deployment of the custom-developed smart manufacturing platform, RTMI allowed Piramal Glass to have an anytime\/anywhere access to real-time insights. Using the Azure IoT Hub, RTMI acquired data from sensors on production lines and then measured losses at each stage in real-time to improve the production line. With the integration of RTMI in the production equipment, it provides a holistic view of the process and how various parameters impact the others. “Till date, we have analysed over 1 billion data points from 4,000 sensors and equipment. One of the biggest challenges we were facing was to extract valuable but siloed data from existing equipment and production lines. For this, we have invested in the integration of information technology with operational technology, along with retrofitted sensors, to capture the data every second. And to analyse this high volume and high-velocity data, we are building an industrial-scale data lake with real-time data pipelines,” said Shah. Further added, “The smart workflows, customisable dashboards, and real-time notifications by RTMI provided timely actionable insights to our furnace, production and quality personnel. We have also implemented a cloud-based energy management system integrated with advanced meters to provide real-time view and advanced analytics on electricity, natural gas, and compressed air consumption.” These smart manufacturing initiatives provided the plant personnel with real-time visibility into operations, thereby enabling the stakeholders to make quicker decisions. With the integration of a paperless system, Piramal Glass became a data-rich company leveraging the data to gain advanced insights. The benefits of RTMI was fairly notably immediately and therefore was the company decided to roll-out the innovation on a global scale across 60 production lines. RTMI — a customised, first-of-its-kind home-grown solution in the glass manufacturing industry, also offered additional real-time insights other than the existing manufacturing execution system and reduced the cost of ownership by 70%. Smart management has led to a reduction in manual data gathering by 40%, resulting in improved employee productivity by 25%. The real-time feedback loop, on the other hand, has led to a 5% reduction in defects, translating into better quality and enhanced predictability of delivery timelines for the customers. Additionally, RTMI has resulted in a 2-3% improvement in production efficiency with a payback period of less than a year. Due to real-time visibility of production and quality parameters, the plant personnel were able to take quick corrective actions that in turn, eliminate defects. This has resulted in enhanced quality of bottles being delivered to the customer. RTMI predicted the completion of the production run, which in turn provided better visibility to the customer in terms of delivery timelines. With the smart management integrated across all equipment, all sort of issues gets highlighted immediately and gets communicated to the vendors for troubleshooting. Future Prospectives While many of the digital initiatives by Piramal Glass have started as experiments, a number of these have scaled up eventually, namely, computer vision to improve worker effectiveness, persuasive technologies to drive innovation, bots to enhance employee productivity, and artificial intelligence (AI) to create a manufacturing process a digital twin. “We aim to create a massive data platform with real-time data pipelines from our IoT transactional systems. We are on-boarding best-of-breed AI companies and academia, to build a partner ecosystem that will develop various predictive and prescriptive AI and machine learning models,” said Shah. With the smart use of AI and ML, Piramal Glass has been able to augment and assist its personnel in taking faster and smarter data-driven decisions that will result in reduced costs and, in turn, drive profitability. “Our investments in digital and analytics capabilities will provide us with a competitive advantage to delight our customers and blend the art of glass manufacturing with science and AI,” concludes Shah.","excerpt":"IoT is connecting all devices in our home, from appliances to equipment, and now it’s increasingly moving its base towards its industrial use with IoT sensors improving production efficiency and quality. Glass manufacturing sector is one such artistic industry that has evolved a lot since its inception in the prehistoric times. From manual glass fusion […]","categories":["AI Features"],"tags":["Azure Machine Learning","Data Analytics","Data analytics India","data analytics jobs","Internet of things"],"author_name":"Sejuti Das","publish_date":"2020-01-12T21:26:27","publication_year":"2020","word_count":1422,"keywords":["artificial intelligence","machine learning","data analytics jobs","AI","R","ML","computer vision","Internet of things","RAG","Data analytics India","Aim","Azure Machine Learning","analytics","Data Analytics","Azure"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","computer vision","analytics","Aim","RAG","Azure","R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/case-study-how-this-glass-manufacturing-company-leveraged-azure-iot-to-get-real-time-visibility\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10171400,"title":"Synology ActiveProtect: Closing the Cyber Resilience Gap in Indian Enterprises","content":"India’s digital transformation has created significant opportunities, but at the same time, it has also exposed organisations to heightened cybersecurity risks. In 2024 alone, 95 organisations in the country suffered data breaches, making it the second most targeted nation globally for cyberattacks, according to CloudSEK’s ThreatLandscape Report. These breaches are costly. IBM’s Cost of a Data Breach Report highlights that the average cost of a breach in India has risen to ₹19.5 crore, particularly impacting industries such as technology and pharmaceuticals. The growing complexity of IT environments, coupled with hybrid work setups and increasing reliance on cloud-based systems, has made enterprises vulnerable to sophisticated threats like infostealers. IBM X-Force Threat Intelligence Index 2024 reports a staggering 266% rise in the infostealer malware, which silently collects sensitive credentials to infiltrate systems undetected. Moreover, regulatory changes such as India’s Digital Personal Data Protection (DPDP) Act 2023 have added pressure on organisations to ensure robust data protection measures. The Act emphasises user consent, data localisation, and accountability for data processors, making compliance an essential yet challenging task for businesses. In this environment, Indian organisations must move beyond treating cybersecurity as a reactive IT function. What’s required is a broader, strategic approach to data protection — one that not only prepares organisations to recover from breaches but actively reduces the likelihood and impact of such incidents. This means adopting the 3-2-1-1-0 backup strategy, where enterprises maintain three copies of data across two different media types, with one stored offsite, one kept offline or air-gapped, and zero errors confirmed through regular integrity checks. Immutability and air-gapped backups are especially critical as attackers increasingly target backup infrastructure to cripple recovery attempts. Equally important is ensuring that backups are verifiably restorable. Simply backing up data is not enough. Enterprises must conduct frequent backup verification, detect data corruption early, and enable rapid recovery to minimise downtime. In today’s distributed IT environments, where multiple backup systems operate across different geographies and platforms, a centralised management system becomes essential. A CMS that brings visibility, automation, and policy consistency under a single interface can substantially reduce operational friction, particularly for sectors like BFSI, healthcare, and government that manage sensitive data and operate under regulatory scrutiny. These strategies form the foundation of modern cyber resilience, and they must be paired with solutions that execute them effectively and at scale. ActiveProtect, Synology’s dedicated backup and recovery appliance, is built around these principles. It addresses the vulnerabilities exposed by conventional systems, bringing together performance, security, and simplicity into one integrated platform. Designed to address the rising costs of data breaches and the complexity of modern IT systems, ActiveProtect delivers exceptional backup and recovery speeds while minimising data transmission and maximising storage efficiency. Contemporary ransomware attacks have evolved to be highly sophisticated. They frequently target backup systems to ensure that any attempts by an organisation to recover data compromise the backups themselves. ActiveProtect tackles this critical vulnerability head-on with a multi-layered approach to backup protection. It stores backups in air-gapped environments to prevent unauthorised access, ensures immutability through Write Once Read Many (WORM) structures to safeguard against tampering, and conducts regular automated integrity checks to guarantee recoverability. These features collectively enhance cyber resilience, empowering organisations to protect their data against sophisticated threats. Simplified Multi-Site, Multi-Vendor Management Today, industries like BFSI (banking, financial services, and insurance), healthcare, and government face significant challenges in managing data protection across multiple sites and vendors. These sectors often operate under stringent regulatory requirements, handle sensitive data, and rely on uninterrupted operations. For instance, healthcare providers must ensure patient data remains secure while adhering to HIPAA or similar compliance standards, while BFSI organisations grapple with safeguarding financial transactions amidst rising cyber threats. Managing backups across sprawling networks can be daunting. IT teams often struggle with fragmented systems, inconsistent security policies, and manual configurations that drain time and resources. Synology ActiveProtect simplifies this with a centralised interface that brings everything under one roof. IT teams can oversee up to 2,500 locations and 150,000 workloads from a single console, significantly reducing operational complexity. The platform also automates the application of protection policies to new workloads, eliminating the need for manual configuration and ensuring consistent security across the organisation. Additionally, ActiveProtect supports cross-platform recovery, enabling seamless restorations between different hypervisors, including Physical-to-Virtual (P2V) and Virtual-to-Virtual (V2V) scenarios. This flexibility ensures efficient data recovery regardless of the underlying infrastructure. By streamlining operations and avoiding vendor lock-in, ActiveProtect enables enterprises to manage diverse environments with greater ease and efficiency. Performance Highlights ActiveProtect Appliance delivers exceptional performance, ensuring minimal disruption to daily operations. It backs up seven times faster than traditional methods, reducing downtime and ensuring business continuity. Recovery speeds are twice as fast, facilitating rapid resumption of services post-incident. Advanced deduplication techniques reduce storage requirements by 50% while minimising network load with a 99% decrease in data transmission during backups. ActiveProtect Appliance centralises the backup of diverse workloads, including SaaS applications, virtual machines, physical servers, personal computers, Macs, file servers, and databases. This consolidation ensures swift recovery and uninterrupted service availability. Deployment is fast, allowing organisations to establish robust data protection within minutes. The appliance offers features such as data corruption detection and restoration, backup verification, and disaster recovery testing in isolated environments using built-in hypervisors. As mentioned, it safeguards against ransomware through immutable backups and air-gapped storage solutions. Moreover, the recovery options are versatile, supporting bare-metal and file-level restorations. Role-based access controls further enhance data security by regulating permissions for workload restoration and data viewing. Conclusion AIM believes that Indian organisations must take a business-first view of cybersecurity, understanding that compromised data not only disrupts operations but also damages customer trust, regulatory standing, and long-term competitiveness. In a market facing rising attacks and complex infrastructure, enterprises need tools that combine performance, resilience, and simplicity. Synology’s ActiveProtect Appliance meets these needs head-on, offering a purpose-built solution for enterprises ready to strengthen their defence posture and close the resilience gap in a cutthroat digital landscape. Learn more about Synology ActiveProtect Appliance. Request a professional consultation from Synology’s experts.","excerpt":"Synology’s ActiveProtect Appliance offers a comprehensive solution tailored specifically for enterprises seeking efficient data backup, recovery, and security.​","categories":["AI Highlights"],"tags":["Cyber Security"],"author_name":"Siddharth Jindal","publish_date":"2025-06-05T15:37:34","publication_year":"2025","word_count":1001,"keywords":["Go","API","Cyber Security","AI","ML","Git","RAG","Aim","Rust","GAN","R"],"extracted_tech_keywords":["AI","ML","Aim","RAG","R","Go","Rust","Git","API","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/synology-activeprotect-closing-the-cyber-resilience-gap-in-indian-enterprises\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10168215,"title":"How OpenAI is Trying to Win Over Developers and Vibe Coders","content":"Google’s Gemini is gaining an edge over Claude with its coding prowess, and GitHub Copilot is adding vibe coding abilities. Given these developments, the trend for the companies’ focus is clear—building AI solutions tailored for developers and vibe coders. OpenAI (considering it already has a two-million-plus business user base) is not holding back either. Numerous new developments by the Sam Altman-led company signify its priority of getting developers on board their platform. New AI Models and Tools Built for Coding On Wednesday, OpenAI launched o3 and o4-mini. The models are trained to think for longer before providing an output. O3 is a reasoning model, and o4-mini is a smaller model optimised for fast, cost-efficient reasoning. These reasoning models can harness various abilities, such as searching the web, analysing uploaded files and other data with Python, and generating images, all from within ChatGPT. OpenAI mentioned that its models achieve higher benchmark numbers in tests like Codeforces competition code than their predecessors, citing a 20% fewer major errors over OpenAI o1, as tested by external experts. The benchmark tests for coding involved SWE-Lancer, SWE-Bench Verified Software Engineering and Aider Polygot Code Editing. Not limited to its internal tests, Aider’s polyglot coding leaderboard puts o3 at the top in terms of accuracy. While the leaderboard chart mentions the price being higher than Gemini, the o4-mini is still cheaper than the Claude 3.7 Sonnet, and o1. Prior to the reasoning models, on April 14, OpenAI introduced the GPT-4.1 series, built for API usage. This series focuses on coding and instruction following while expanding the context window. GPT-4.1 scored 54.6% on SWE-bench Verified, closing in on the gap with Gemini 2.5 Pro, which scored 63.8%. The GPT-4.1 nano also performed better than GPT-4o. In addition to improving the models, OpenAI released Codex CLI, an open-source, lightweight coding agent that can run from your terminal and is available under an Apache-2.0 license. Although this is an experimental project, notable companies have tested it, illustrating its usability for developers. Cloudflare demonstrated how to use OpenAI Codex to deploy a full-stack Cloudflare Workers application in just a few minutes. Some developers have appreciated the tool’s open-source nature, which they consider an advantage over options like Claude Code. OpenAI Eyes Vibe Coding Platforms OpenAI is reportedly planning to buy Windsurf for about $3 billion, according to a Bloomberg report. If that turns out to be true, OpenAI will compete with AI coding platforms, including Microsoft’s VS Code, Claude Code, and others. This would also open up competition with Anthropic. Not to forget, in 2023, OpenAI invested $8 million to help improve Cursor. It is unclear if OpenAI aims to compete with Cursor or later partner with it, but the company is potentially positioning itself to participate in the vibe-coding war. Developers, however, have a mixed opinion on OpenAI’s potential Windsurf acquisition. Eduard Ruzga, staff engineer at Prezi, told AIM, “I tried Codex this morning, but I do not really like it. It is somewhat restricted in nature, and terminal UX feels worse than Claude.” He added, “I am not sure about Codex yet. It feels like a competitor to Claude Code that has issues. It uses an API key so that you may run up quite a bill with it, and they do not show how much you spent.” He highlighted that he spent 30 cents on just one call. When asked about the Windsurf acquisition, he said, “If OpenAI buys it and subsidises LLM costs, it could be quite a big move, raising stakes in this market.” He remarked that it would be a bold move considering Microsoft’s recent push with GitHub Copilot’s agent mode. Prasun Anand, creator of Zasper, shared his thoughts on the same, and told AIM, “The world needs good ‘coding assistants’. Developers spend at least four hours daily with their IDEs. Windsurf has built a strong distribution, and OpenAI acknowledges this with its plan to acquire Windsurf.” Likewise, Sulaiman Mudimala, founder of Bezu AI, told AIM that OpenAI needs to earn developers’ trust, like Anthropic has with its Claude model. He believes that the potential acquisition alone will not win over developers.","excerpt":"OpenAI may be competing with Cursor or partnering with it, but the company is likely positioning itself to participate in the vibe-coding war.","categories":["AI Features"],"tags":["OpenAI","Vibe Coding"],"author_name":"Ankush Das","publish_date":"2025-04-17T17:06:19","publication_year":"2025","word_count":688,"keywords":["Anthropic","ChatGPT","TPU","OpenAI","AI","GPT-4o","Python","Vibe Coding","Aim","R","Gemini 2.5"],"extracted_tech_keywords":["AI","GPT-4o","ChatGPT","OpenAI","Anthropic","Gemini 2.5","Aim","TPU","Python","R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-openai-is-trying-to-win-over-developers-and-vibe-coders\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":29105,"title":"India-Israel Innovation Centre Launched At IoT India Congress In Bengaluru","content":"With an aim to give a boost to the entrepreneurs in the technology sector, Ramanan Ramanathan, mission director at Atal Innovation Mission, Moshe Porat, founder and CEO at Mesh Labs, and Varad Krishna, founding partner of IIIC, launched country’s first India-Israel Innovation Centre (IIIC) in Bengaluru on Wednesday. The announcement was a part of a two-day event called IoT India Congress. Speaking at the launch, Dana Kush, Consulate General of Israel to South India told the audience, “We are proud to witness the launch of IIIC, which reinstates that the formula of ‘India + Israel= Innovation’ works. Our role as diplomats is to find the best possible ways to connect India to the Israeli ecosystem and vice-versa, especially while both nations are focusing on startups. I wish IIIC to be another productive milestone in connecting the countries and pave the way for many more successful businesses to grow and transform. There would have been no better place than having this launch in India’s Silicon Valley — Bengaluru.” The IIIC is a significant step towards facilitating penetration of Israeli companies in India and aims to bring local partnerships and joint ventures between the companies from the two countries. It will also provide an ecosystem to support entrepreneurship, partnership with vendors, mentoring and non-formal community development that will assist the growth of the companies in different verticals and covers business, technology, investors and customers. Krishna said that their aim was to meet the long-pending need to connect the startup ecosystems in Israel and India, and to improve the quality and survival rate of these startups. IoT India Congress attracted participation from over 2,000 delegates over the two days. Marc Jones, Chairman and CEO, Aeris delivered the keynote address. Industry leaders like Ramanathan, Juergen Hase, CEO at Unlimit, Dr Gopichand Katragadda, CTO at Tata Group, Ravi Ramaswamy, Senior Director and Head, Health Systems at Philips Innovation Campus and VS Sridhar, Senior Vice President and Head, IoT Business Initiative at Tata Communications Ltd, were some of the dignitaries present at the event.","excerpt":"With an aim to give a boost to the entrepreneurs in the technology sector, Ramanan Ramanathan, mission director at Atal Innovation Mission, Moshe Porat, founder and CEO at Mesh Labs, and Varad Krishna, founding partner of IIIC, launched country’s first India-Israel Innovation Centre (IIIC) in Bengaluru on Wednesday. The announcement was a part of a two-day event called […]","categories":["Deep Tech"],"tags":["atal innovation mission"],"author_name":"Prajakta Hebbar","publish_date":"2018-10-10T12:23:42","publication_year":"2018","word_count":338,"keywords":["Go","programming_languages:R","AI","innovation","programming_languages:Go","RAG","Aim","atal innovation mission","R","startup"],"extracted_tech_keywords":["AI","Aim","RAG","R","Go","innovation","startup","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/india-israel-innovation-center-bangalore-iot-india-congress\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":34886,"title":"US President Trump Signs New AI Initiative, Critics Think It Missed The Mark","content":"US President Donald Trump this week signed an executive order to fast-track the development and regulation of artificial intelligence in the country. This move is a part of their American AI Initiative and aims to better educate workers, improve access to cloud computing, data, AI and promote cooperation with other countries. Experts now think that this order may define the future of everything — from consumer products to healthcare and even warfare. Critics have also noted that the order has not set aside any funds for research and development in AI. According to reports, this order by US President Trump has sprung from the fact that the US has not been able to keep up with China with respect to the advancements in emerging technologies, especially AI. Many have also suggested that former defence minister Jim Mattis has a large role to play in this development. The US initiative, which follows a basic roadmap, has listed five key pillars for the same: Research and development Infrastructure Governance Workforce International engagement Last year, India’s NITI Aayog had unveiled a discussion paper which addressed the national strategy on artificial intelligence and other emerging technologies in India. Here, the government think tank identified five sectors to focus its efforts towards the implementation of AI to serve societal needs. The five sectors are: Healthcare: increased access and affordability of quality healthcare Agriculture: enhanced farmers’ income, increased farm productivity and reduction of wastage Education: improved access and quality of education Smart Cities and Infrastructure: efficient and connectivity for the burgeoning urban population Smart Mobility and Transportation: smarter and safer modes of transportation and better traffic and congestion problems India’s strategy report had also proposed a two-tiered structure to address India’s AI research aspirations: Centre of Research Excellence (CORE) focused on developing a better understanding of existing core research and pushing technology frontiers through the creation of new knowledge International Centers of Transformational AI (ICTAI) with a mandate of developing and deploying application-based research. Private sector collaboration is envisioned to be a key aspect of ICTAIs.","excerpt":"US President Donald Trump this week signed an executive order to fast-track the development and regulation of artificial intelligence in the country. This move is a part of their American AI Initiative and aims to better educate workers, improve access to cloud computing, data, AI and promote cooperation with other countries. Experts now think that this order may […]","categories":["AI News"],"tags":["AI (Artificial Intelligence)","AI Strategy","donald trump","US"],"author_name":"Prajakta Hebbar","publish_date":"2019-02-13T11:03:30","publication_year":"2019","word_count":340,"keywords":["Go","artificial intelligence","programming_languages:R","AI","cloud computing","programming_languages:Go","donald trump","AI Strategy","Aim","ViT","AI research","R","AI (Artificial Intelligence)","US"],"extracted_tech_keywords":["AI","artificial intelligence","Aim","cloud computing","R","Go","ViT","AI research","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/us-president-trump-signs-new-ai-initiative-critics-think-it-missed-the-mark\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10167227,"title":"Tata Electronics Appoints KC Ang to Lead Semiconductor Manufacturing","content":"Tata Electronics has appointed KC Ang as president and head of its foundry business, Tata Semiconductor Manufacturing, the company announced on Wednesday. Ang will report to Randhir Thakur, CEO & MD of Tata Electronics. In this role, Ang will oversee the company’s semiconductor foundry operations and strategic partnerships as it prepares to launch chip manufacturing operations. He will also advance the company’s AI-enabled foundry operations. This appointment is part of India’s efforts to develop a domestic chipmaking ecosystem. Ang expressed his excitement about joining Tata Electronics, highlighting the growing ecosystem of India’s semiconductor manufacturing industry. “This sector is poised to make a significant global impact in the coming years. I look forward to leveraging my experience to drive Tata Semiconductor Manufacturing to the forefront of global chip manufacturing.” Most recently, Ang served as president of Asia and chairman of China at GlobalFoundries for over fifteen years. Bringing over three decades of experience in the foundry industry, specialising in fab start-up, management, operations, and technology transfer, he has worked extensively in Malaysia, Germany, the US, and China. Thakur welcomed Ang, stating, “We are delighted to have an industry veteran like KC Ang join us and be part of the team driving our semiconductor journey. His extensive industry knowledge and diverse expertise will be valuable as we make steady progress towards commencing operations at our Dholera Fab.” Last year, the company started exporting semiconductor chips packaged at its pilot line in Bengaluru-based research and development centre. The packaged chips were being shipped to some of Tata Electronics’ partners in Japan, the US, and Europe in limited quantities. Similarly, Mumbai-based company RRP Electronics started exporting packaged semiconductors under the government of India’s Semiconductor Mission (ISM). The company exported its first consignment, valued at ₹6.51 crore, primarily focused on ASICS-based technologies, to a European customer.","excerpt":"Ang served as president of Asia and chairman of China at GlobalFoundries for over fifteen years.","categories":["AI News"],"tags":["India semiconductor mission","Semiconductor India","tata semiconductor"],"author_name":"Sanjana Gupta","publish_date":"2025-04-03T16:35:14","publication_year":"2025","word_count":301,"keywords":["Go","tata semiconductor","AI","programming_languages:R","India semiconductor mission","programming_languages:Go","RAG","Semiconductor India","R"],"extracted_tech_keywords":["AI","RAG","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/tata-electronics-appoints-kc-ang-to-lead-semiconductor-manufacturing\/","complexity_score":2,"technical_depth":6,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":33850,"title":"Meet XMOS ‘Digital Twin’, A Futuristic Voice Assistant Which Adapts To Your Way Of Interaction","content":"As the most advanced species on the planet, speech is the most natural and effortless way of communicating for us humans. Now, the machines have evolved to understand and interact with us and this concept no longer belongs in the realm of sci-fi. Amazon Echo, Google Home and Vector are commercial products that people can buy. Now, XMOS, a semiconductor company specializing in voice processing, has created algorithms which are capable of processing a spoken voice even in loud or challenging environments. Founded in 2005 and built on research from the University of Bristol, the technology used by XMOS is used in numerous devices, including Amazon’s Echo. XMOS’ Power-Packed Chip XMOS speech detection and isolation techniques include: Beamforming: the ability to track a person’s voice as they move around Acoustic echo cancellation: the ability to separate the user’s voice from the sound being played by the device itself Dereverberation: the ability to cancel out echoes Barge-in: the ability to  stop audio playback when the device’s wake-word is detected Fixed or automatic gain control: The ability to control or limit voice gains or volume intensities Noise cancellation\/suppression: the ability to cancel out noises. Director of corporate marketing, Esther Connock told a news portal, “Currently XMOS serves as the only qualified stereo solution partner for Amazon and the only one that can do acoustic cancellation in stereo. At present, XMOS specializes in building cutting-edge voice applications and also has its attention in investigating areas like in-car interfaces.” XMOS claims to have developed a technology for sound source separation. A technology that extracts multiple voices in a conversation which in future would allow focussing only the speaker’s voice in a conversation where there are a lot of surrounding noises. The company also has high hopes for the future of voice and believes in a ‘personal assistant’ in a flexible and wearable device that will provide voice recognition services. The Digital Twin XMOS believes in creating a digital twin. A voice application that can talk as natural as a human. The idea is to give emotions to speech so that it can learn to adapt to the way you use it. For example, your voice assistant could learn that you do not wish to be spoken to unless you spoke first. XMOS wants to make the communication as organic as possible like how humans talk to each other. XMOS believes that a frictionless communication between man and machines is the future of voice processing. In addition, XMOS also believes that the digital twin should know much more than your music preferences or how you want it to interact with you. As Cannock says, “The digital twin will learn not just my music preferences, but my everything preferences. When I want to be disturbed, my friends that I will prioritize talking to – everything.” Resistance Though people are often amused by the voice processing technology among many others, like many it still feels resistance. People are okay with their smartphones with cameras, voice recording, and all other gimmicks listening to them 24X7  but they can not stand when they have a speaker that listens to them.XMOS believes that trusted content will be the key to the acceptance of voice among people, putting user-experience on top of everything. Conclusion Speech is much effortless than any other forms of communication. Since technology has always been about making people’s life easier, the voice processing is one such tech that is still futuristic yet a reality. With so many voice assistant applications and devices already available in the market today, accepting voice commands to perform tasks, we can expect a more natural communication between man and machines in the near future.","excerpt":"As the most advanced species on the planet, speech is the most natural and effortless way of communicating for us humans. Now, the machines have evolved to understand and interact with us and this concept no longer belongs in the realm of sci-fi. Amazon Echo, Google Home and Vector are commercial products that people can […]","categories":["IT Services"],"tags":["digital marketing","Voice Assistant","Voice Recognition Technology"],"author_name":"Amal Nair","publish_date":"2019-01-22T08:27:49","publication_year":"2019","word_count":612,"keywords":["Go","Voice Assistant","programming_languages:R","AI","programming_languages:Go","Git","Voice Recognition Technology","digital marketing","Aim","Rust","GAN","R","programming_languages:Rust"],"extracted_tech_keywords":["AI","Aim","R","Go","Rust","Git","GAN","programming_languages:R","programming_languages:Go","programming_languages:Rust"],"url":"https:\/\/analyticsindiamag.com\/it-services\/meet-xmos-digital-twin-a-futuristic-voice-assistant-which-adapts-to-your-way-of-interaction\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10071006,"title":"Bhupinder Singh quits ECom Express to join Junglee Games as the VP Data Science","content":"After working for three years as the Head of Data Science at Ecom Express, Bhupinder Singh moves on to join Junglee Games as its VP of Data Science. With close to 75 million users, Junglee Games is one of the fastest-growing skill games companies. At ECom Express, which is an end-to-end technology-enabled logistics solutions provider, he was at the forefront of the company’s digital transformation. Under his guidance, the team embedded models in their technology stack to process shipments faster and developed last-mile routing and shipment allocation algorithms for improving overall productivity. A mechanical engineer by qualification, Singh completed his PhD from the University of Florida in 2009. He also holds a certification in Business Analytics and Intelligence from IIM-B. For over a decade of experience as a mechanical engineer, Singh moved into the domain of Analytics in 2013 when he was appointed a Manager of Analytics by Innovation Labs. Thereafter, he has held several roles in the field of analytics and data science. He also holds several patents in data science. “Junglee Games is a leading player in the gaming industry in India with a focus on skill games. Its portfolio of games includes Junglee Rummy, Howzat, Poker, Solitaire Gold, and Carrom Stars, with over 75M users. Gaming in India has been fairly nascent and is poised for growth, supported by various factors  such as increasing penetration of internet and mobile phones, better internet speed and favourable demographics. I am really excited to join the amazing team at Junglee Games to strengthen the data strategy & drive innovation through data science, with a vision to offer top-notch customer experience and consolidate our leadership position in skill games,” said Singh.","excerpt":"With close to 75 million users, Junglee Games is one of the fastest-growing skill games companies.","categories":["AI News"],"tags":[],"author_name":"Zinnia Banerjee","publish_date":"2022-07-14T17:48:12","publication_year":"2022","word_count":280,"keywords":["Go","data science","programming_languages:R","AI","innovation","digital transformation","Git","ViT","analytics","R"],"extracted_tech_keywords":["AI","data science","analytics","R","Go","Git","ViT","digital transformation","innovation","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/bhupinder-singh-quits-ecom-express-to-join-junglee-games-as-the-vp-data-science\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10162398,"title":"CRED is Hiring Data Scientists","content":"Fintech firm CRED is expanding its AI and machine learning capabilities and is seeking two data scientists to join its team. For professionals with a passion for LLMs, generative AI, and deep learning, this could be the right opportunity to work on solutions that redefine financial intelligence. At CRED, data scientists experiment and innovate in an environment that fosters research and real-world applications. From predictive modelling to scalable AI architectures, the role offers a mix of challenges. They will work on developing and deploying AI-driven solutions, collaborate with cross-functional teams, and implement real-time models that push the boundaries of automation. There’s also an opportunity to publish research and present at industry conferences, ensuring data scientists stay ahead in the rapidly evolving fields of machine learning and AI. CRED is looking for professionals with more than three years of experience in data science, particularly in NLP and deep learning. Strong programming skills in Python, Java, or Scala are preferred, along with hands-on experience in building scalable AI\/ML solutions. Expertise in working with LLMs and generative AI would also be favoured. According to the fintech platform, CRED’s work culture is built on trust. There are no rigid job titles or fixed work hours. Moreover, productivity at the company is measured by impact, not time spent at a desk. Employees even receive their salaries before their official joining date, a strong statement of trust that defines CRED’s ethos. The company also claims to ensure a stress-free environment, offering a comprehensive health insurance plan, paid sick leaves, and an in-house pantry offering free lunch and dinner for all team members. It’s a workplace that values well-being just as much as innovation.","excerpt":"At CRED, data scientists experiment and innovate in an environment that fosters research and real-world applications.","categories":["AI News"],"tags":["AI hiring","cred","Data Scientist"],"author_name":"Vidyashree Srinivas","publish_date":"2025-01-29T11:50:07","publication_year":"2025","word_count":277,"keywords":["data science","machine learning","AI","ML","cred","NLP","Python","Aim","deep learning","AI hiring","generative AI","Data Scientist","R"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","NLP","data science","generative AI","Aim","Python","R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/cred-is-hiring-data-scientists\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10053191,"title":"Top Papers Presented At EMNLP 2021","content":"The most exceptional papers were presented at the Empirical Methods in Natural Language Processing (EMNLP 2021) Conference. For the main conference papers, the organisers examined the following three categories. Best Long Paper: Visually Grounded Reasoning across Languages and Cultures The researchers develop a novel annotation methodology in which the choice of images and descriptions is fully determined by native speakers. Additionally, the researchers build and assess a set of multilingual and multimodal baselines, including transfer using models and translation. The researchers discovered that their performance is occasionally slightly above chance and is significantly harmed by the out-of-distribution character of concepts, visuals, and languages in MaRVL in comparison to English datasets. This provides reason to suppose that it provides a more accurate assessment of the suitability of state-of-the-art models for real-world applications outside a confined linguistic and cultural domain. For further information, refer to the article. Best Short Paper: CHoRaL: Collecting Humor Reaction Labels from Millions of Social Media Users The researchers propose the CHoRaL framework for collecting humour reaction labels automatically, as well as a dataset containing 785K postings with and without humour scores. Additionally, the researchers analyse humour expressions in our dataset and develop algorithms capable of detecting humour on a level with human labellers. CHoRaL enables the construction of algorithms for detecting comedy on any topic, and the dataset has the potential to aid in broader applications, such as separating malevolent disinformation posts from benign humorous posts. Additionally, CHoRaL can be used to classify additional human emotions, such as anger and melancholy. For further information, refer to the article. Outstanding Papers: MindCraft: Theory of Mind Modeling for Situated Dialogue in Collaborative Tasks This work stimulates disparities in a virtual environment by developing a novel dataset and experimental framework for in-depth research of the theory of mind modelling for situational discussion in collaborative projects. The initial research yields numerous intriguing results that will aid in the creation of computer models for a variety of challenges. The baseline findings emphasise the critical role of interaction discourse and visual experience in forecasting mutual belief levels about the job at hand and about a collaborative partner in order to establish common ground. The researchers hope that their work will contribute to further advancements in areas such as agent planning and decision-making. For further information, refer to the article. SituatedQA: Incorporating Extra-Linguistic Contexts into QA The researchers present the first study to examine the effect of extra-linguistic circumstances on open retrieval quality assurance. The study demonstrates that contemporary systems are incapable of adapting to changes in the temporal or geographical context. As a result, we define tasks and develop a dataset for training and evaluating quality assurance systems that are capable of simulating how facts vary across contexts. The dataset will provide abundant opportunities for future research into constructing models that can elegantly alter their predictions in response to changing temporal and geographical circumstances. Future studies may focus on incorporating source materials chronologically and spatially dependent, such as news items, or on other extra-linguistic factors, such as who is asking the question, while considering individual preferences. For further information, refer to the article. When Attention Meets Fast Recurrence: Training Language Models with Reduced Compute The article describes a very efficient architecture that combines fast recursion with attention and evaluates its effectiveness on a variety of language modelling datasets. The researchers demonstrate that using rapid RNNs with little attention not only achieves superior outcomes but also dramatically reduces training costs. The work is conceptually distinct from accelerating attention and hence provides a counter-intuitive route for enhancing state-of-the-art model design. The researchers feel that the model can be improved further with the addition of stronger attention or recurrent implementations, as well as improved normalisation or optimisation techniques. For further information, refer to the article. Shortcutted Commonsense: Data Spuriousness in Deep Learning of Commonsense Reasoning Commonsense is the quintessential barrier to AI and deep learning approaches. The purpose of this paper is to determine how much of these benefits are due to generalisation beyond the training datasets used to perform these tasks and whether these achievements are susceptible to the problem recently discovered in some other areas of natural language processing. CSQA outperforms the other tasks in terms of generalisation capacity, as it provides a far better approximation to other tasks in a zero-shot environment. These findings indicate the need for additional research with a thorough examination and comparison of methods and procedures. For further information, refer to the article. Best Demo Paper: Datasets: A Community Library for Natural Language Processing Hugging Face Datasets is a community-driven open-source toolkit that standardises the processing, dissemination, and documentation of natural language processing datasets. The core library is designed to be simple to use, fast and to support datasets of varied sizes using the same interface. With over 650 datasets contributed by over 250 contributors, it simplifies the use of standard datasets, enables new use cases for cross-dataset NLP, and includes advanced features for indexing and streaming big datasets. For further information, refer to the article.","excerpt":"The Best Paper Awards are intended to recognise the best papers presented at EMNLP 2021.","categories":["AI Trends"],"tags":["AI (Artificial Intelligence)","dataset","Deep Learning","Multimodal","Natural Language Processing","RNN"],"author_name":"Dr. Nivash Jeevanandam","publish_date":"2021-11-11T13:00:00","publication_year":"2021","word_count":837,"keywords":["dataset","Go","Hugging Face","API","big data","AI","Natural Language Processing","NLP","Multimodal","deep learning","RNN","GAN","Deep Learning","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","deep learning","NLP","Hugging Face","R","Go","API","big data","GAN","RNN"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-papers-presented-at-emnlp-2021\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10094148,"title":"Will Meteor Lake Be Intel’s Saving Grace?","content":"This week is a big one for chip enthusiasts and AI lovers alike. At Computex, the world’s premier semiconductor conference, top makers have announced chips that might be the future of AI compute. While NVIDIA tried its best to steal the show with CEO Jensen Huang’s announcement-packed keynote, Intel quietly made a splash by showcasing Meteor Lake, a new line of AI-first processors. However, the chipmaker has been plagued with multiple delays and issues in the past when taking this manufacturing approach. With the hype surrounding Meteor Lake, it seems that Intel has a lot riding on this new chip. What makes Meteor Lake special? Meteor Lake marks a huge milestone for Intel, as it will be the first chip to use the Intel 4 process node for manufacturing. However, this isn’t all the chip has up its sleeve, as it also incorporates various chips manufactured in Taiwan Semiconductor Manufacturing Corporation’s (TSMC) foundries, making it the first Intel chip to do so. Intel has gone all in on both the chiplet design and SoC architecture, integrating various discrete functions on one chip in a bid to create a new kind of market. The Meteor Lake processors combine GPU, VPU, and CPU all on one chip to offer the “right balance of power and performance for AI”. According to Intel’s announcement at Computex, the chips will first appear in laptops, seemingly in competition to Apple’s M2 and AMD’s Ryzen 7040 series of chips. Similar to its competitors, Intel has also adopted a system-on-chip architecture for this chip, mixing and matching various chiplets to achieve this effect. The CPU itself will be manufactured by using Intel 4 manufacturing process, with the GPU and SoC Tile using TSMCs 5nm and 6nm processes, respectively. Moreover, Intel has also integrated its Foveros 3D packing technology into these chips, allowing them to fit more features in a smaller form factor. The headline feature of this new chip is its integrated AI engine, facilitated with the help of Intel’s new vision processing unit (VPU). The VPU will not only be reserved for AI workloads, but will also accelerate the integrated GPU and the main CPU itself for these tasks. Notably, the VPU contains Intel’s Gaussian Neural Accelerator (GNA) chip, which has been, so far, its solution for on-board AI processing. While the GNA has been Intel’s go-to for low-power AI inference at the edge, it seems that VPUs will replace them as the generations go on. Intel has stated that some of the code made for GNA runs more efficiently and with better results on VPUs, effectively obsoleting this now-legacy tech. The chip is also scalable to various power draw and performance specifications. Each of the components can be scaled up or down depending on the requirement, resulting in a varied line of products with multiple different SKUs. The CPU, GPU, and VPU are connected with an interconnect fabric that allows them to share the same memory subsystem. This not only improves performance, but puts Intel on par with Apple and AMD for SoC design. Harnessing the power of developers While it seems that Intel is a few years late to the party in this specific market, this delay is to be expected. The company’s Sapphire Rapids chip, which was supposed to be its big answer to AMD’s EPYC server dominance, was hit with multiple delays and manufacturing issues. This resulted in Intel’s roadmap being pushed ahead by a few years, creating a lag that it is now eager to close. Meteor Lake is likely to launch with full support for developers to tinker with the product and harness its full capabilities. Apart from close collaboration with Microsoft to integrate the chips’ AI capabilities into Windows, Intel has also announced support for DirectML, ONNX, and OpenVINO. These SDKs will allow AI developers to easily integrate ML features into their programs while ensuring compatibility. Intel showcased a host of optimisations that Meteor Lake’s in-built AI could bring to applications. Firstly, the VPU can accelerate even common video and audio processing workloads running on the CPU and GPU. Secondly, the processor’s capabilities will have support in various software, like real-time motion capture in Unreal Engine. For AI models specifically, the hybrid approach of combining CPUs and GPUs together with VPUs works wonders. For example, Stable Diffusion can run on both GPU and VPU simultaneously. While the GPU does the heavy lifting of generating the image, the VPU works side by side for super-resolution of the image. A much-needed win? Intel has been suffering in silence, unable to capitalise on the AI wave due to its varied problems. However, Meteor Lake could be the solution for them to get their winning streak back, biting back at Apple for ditching Intel chips for the MacBook. The only thing left for Meteor Lake is Intel getting it to market. While it has unfortunately stumbled when getting the product to market in the past, Meteor Lake might be the exception to this. Intel announced that it would bring Meteor Lake to laptops by the end of this year. This is also backed up by the fact that Intel announced ramped up production for the chips in its Q1 2023 earnings report. By bringing together TSMC’s manufacturing chops and their own foundries, Intel might actually go beyond the manufacturing problems it had in the past. While combining this with the complete package that Meteor Lake seems to be, it might just be the win Intel needs to get back on its feet.","excerpt":"After Sapphire Rapids left a bad taste in customer’s mouths, Intel might bring victory back with Meteor Lake.","categories":["AI Highlights"],"tags":[],"author_name":"Anirudh VK","publish_date":"2023-05-31T10:00:00","publication_year":"2023","word_count":918,"keywords":["Go","Rapids","API","AI-first","AI","ETL","ML","Scala","stable diffusion","R"],"extracted_tech_keywords":["AI","ML","Rapids","R","Go","Scala","API","ETL","stable diffusion","AI-first"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/will-meteor-lake-be-intels-saving-grace\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10134376,"title":"Which Indian State Leads the AI Race?","content":"At its 47th Annual General Meeting (AGM), Reliance Industries made headlines with a bold new AI vision, marking what might well be India’s AI moment. The company introduced the Jio Brain, Jio AI-Cloud, Jio Phone Call AI, and the vision for a national AI infrastructure. While the AGM showcased Jio’s advancements in the AI race, it also threw the stage open for states across the country to accelerate their own AI developments. So, how is each of them stepping up their game to meet the country’s AI mission? Indian States’ AI and Cloud Game Indian states are coming up with unique initiatives to position themselves as leaders in the AI field. Karnataka, for example, recently signed a letter of intent with the World Economic Forum (WEF) to establish a dedicated AI centre, aiming to establish itself as a global AI hub. At the WEF meet in Davos, the state signed eight memorandums of understanding with global firms, securing investments worth INR 23,000 crore across AI, citizen services, sustainability, and e-governance. Similarly, Telangana is also climbing up the ladder by setting up an AI City near Hyderabad. Spread over 200 acres, it will house companies specialising in AI. In Andhra Pradesh, chief minister N Chandrababu Naidu has directed officials to develop Amaravati as a city focused on AI. Recently, Google signed an MoU with the Tamil Nadu government with an aim to create impactful and scalable AI solutions in key areas such as AI start-up enablement, skilling, and industrial ecosystem enablement, including MSMEs. Earlier in February 2024, the tech giant signed an MoU with the Maharashtra government too. The company also announced the launching of a state-of-the-art AI Centre of Excellence at the Indian Institute of Information Technology (IIIT), Nagpur. Further, the Andhra Pradesh government has also announced its plans to sign an MoU with tech giant Google for the initiative ‘AI for Andhra Pradesh, powered by Google’. Southern states are not the only ones that are advancing. Things have been promising up north too. The Delhi government is considering the establishment of an AI hub at its proposed Electronic City in Baprola, offering facilities for companies in electronics design and manufacturing. In Uttar Pradesh, Lucknow is fostering AI innovation through its Centres of Excellence and aims to develop an AI City that integrates technology, research centres, and educational institutions. The state government has allocated 40 acres in Nadarganj Industrial Area to support this ambitious project. Moreover, Assam is making a mark in the semiconductor industry by leasing over 170 acres to the Tata Group for an INR 27,000 crore fab. The Budget Scene Coming down to budget allocations, only a few states have mentioned AI in their February 2024 budget. Kerala has announced plans to introduce a dedicated AI policy aligned with ESG goals within this fiscal year. The Kerala State Industrial Development Corporation (KSIDC) will offer share capital investment of up to INR 5 crore with AI entities that have a minimum investment of INR 10 crore. Karnataka chief minister Siddaramaiah announced that the state is set to establish an AI\/ML administration unit. Additionally, a stipend of INR 15,000 will be provided to 200 engineering graduates enrolled in vocational training courses related to AI\/ML at institutions like IISc, IITs, and NITs. Meanwhile, Tamil Nadu is set to establish the Tamil Nadu Artificial Intelligence Mission to develop guidelines for leveraging AI in various fields, including education, employment, research, and medicine. However, other states have not mentioned the budget allocation or policies for AI development in their budgets. Backing up the states’ AI momentum, the central government, in its Union Budget 2024-25, allocated INR 551.75 crore to the AI Mission, emphasising its commitment to advancing AI research and applications. Semiconductor Push At a recent media briefing at the Vikasa Soudha, Karnataka IT\/BT minister Priyank Kharge highlighted the remarkable achievement and said, “With INR 21.56 crore investment, we’ve supported 200+ innovations, creating 800 ESDM jobs. Semiconductor Fabless Accelerator Lab (SFAL) startups have raised INR 140 crore, valued at INR 500 crore collectively.” In Tamil Nadu, Applied Materials India Pvt has entered into an agreement with the state government to establish an advanced AI-enabled Technology Development Centre of Excellence in Chennai. Telangana is also making waves in this sector, with Kaynes Technology India Ltd signing an MoU with the state government to set up an INR 2,800 crore facility for Outsourced Semiconductor Assembly and Test (OSAT) in Hyderabad. Also, Kaynes Technology secured the Gujarat government’s approval to set up a semiconductor unit in Sanand. This facility will have the capacity to produce 6.3 million chips per day, backed by an investment of INR 3,307 crore. AI Everywhere Apart from the semiconductor or big tech AI partnerships, states are also spearheading smaller AI initiatives. States like Kerala, Tamil Nadu, Odisha, and Punjab are pushing towards integrating AI into education. Kerala is leading with its Digital University becoming the first in India to develop an AI processor. With respect to traffic management, states like Goa and Sikkim plan to launch AI-driven systems. States are also working to address the concerns of rising crime rates in the country. For instance, the Maharashtra government has launched — the Maharashtra Research and Vigilance for Enhanced Law Enforcement (MARVEL). MARVEL is set to enhance intelligence capabilities and improve crime prediction, making the state the first in the country to create an independent entity for AI in law enforcement. With AI developments accelerating rapidly across India, it may be difficult to pinpoint a clear leader. However, based on the diverse initiatives and partnerships underway, Karnataka, Tamil Nadu, and Telangana might be at the forefront. Rest assured, other states are not lying dormant!","excerpt":"Backing up the states’ AI momentum, the central government, in its Union Budget 2024-25, allocated INR 551.75 crore to the AI Mission.","categories":["AI Features"],"tags":["AI (Artificial Intelligence)","AI Tool","India"],"author_name":"Vidyashree Srinivas","publish_date":"2024-09-04T11:38:41","publication_year":"2024","word_count":942,"keywords":["Go","API","artificial intelligence","AI","ML","Scala","Git","RAG","Aim","AI Tool","R","India","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","ML","Aim","RAG","R","Go","Scala","Git","API"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/which-indian-state-leads-the-ai-race\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10058703,"title":"What does it take to remove your personal data from the internet?","content":"Shoshana Zuboff, the author of “The Age of Surveillance Capitalism,” said, the mining of our personal data for profit is “changing the nature of society” leading to the “wholesale destruction of privacy.” Today companies know so much about us and hold the power to target and manipulate us not only with relevant digital ads, but also with “subliminal clues, real time rewards and punishments, algorithmic recommendation tools, psychological micro-targeting, and engineered social comparison dynamics,” she said. Facebook is a crisis of every Republic, but this is bigger than a single corporation.The only durable solution is to outlaw #SurveillanceCapitalism’s illegitimate operations. First up: secret massive scale data extraction from every human life. Join us!@Moonalice @FBoversight https:\/\/t.co\/A7LkFCmFW3— Shoshana Zuboff (@shoshanazuboff) October 4, 2021 Privacy woes In today’s data-driven economy, privacy is fast becoming a luxury. Today, we are no longer paying for our mail and news sources with money—instead, we are paying for these services with our personal data. Google, for instance, scans your Gmail account to allow advertisers a chance to promote items based on what you write in your personal communication with others. The personal data of individuals is also used to give them different search results based on their political leanings, and is used by governments to identify possible criminal and terrorist suspects—undermining both democracy and our autonomy. Today, the only way to get privacy is to buy it, which is neither cheap nor convenient. For instance, Apple introduced a software update last year that lets its users decide whether an app should be allowed to track their movement across other apps and sites they use—but not everyone can buy expensive Apple products. Is privacy soon going to be available to only those who can afford it? Right to be Forgotten Last year, Ashutosh Kaushik, former winner of the reality tv shows Bigg Boss (2008) and MTV Roadies 5.0, approached the Delhi High Court to remove all the content about him be removed from the internet. He cited his Right to be Forgotten, arguing that posts and videos of him on the internet have made it impossible for him to move on from his past mistakes (he is referring to an incident in 2009 in which he was held by the Mumbai police for driving under the influence of alcohol). The Right to Privacy was determined a fundamental right by the Supreme Court in 2017, under Article 21 of the Constitution. Meanwhile, the Right to be Forgotten, which refers to an individual’s right to privacy, is yet to be tabled in the parliament. The Chapter V of the Personal Data Protection Bill—introduced in the Lok Sabha in 2019— states that the “data principal (the person to whom the data is related) shall have the right to restrict or prevent the continuing disclosure of his personal data by a data fiduciary.” Therefore, under the Right to be Forgotten, users should be able to de-link, restrict, erase, or amend the disclosure of their personal data held by third parties. Exactly why kids should have the #RightToBeForgotten: their online actions shouldn't haunt them into adulthood. Kids do & say dumb\/awful\/ignorant things & *learn* why they shouldn't do that as adults. Judging adults for stuff they did as 13yearolds is nuts https:\/\/t.co\/lodSgCqFqD— Tama Leaver ➡️ @tama@aoir.social (@tamaleaver) June 22, 2021 While some people are petitioning the government to allow their data to be deleted from the internet to protect their privacy, companies like Twitter and Facebook are receiving demands from governments and law enforcement agencies to delete content from their websites. Last year, nearly 200 verified accounts of journalists on Twitter faced 361 legal demands from governments and law enforcement agencies to have their content removed; India ranked top for these requests, followed by the United States. Twitter submitted to 29% of the 38,524 legal demands it received from governments to remove content from its website. https:\/\/twitter.com\/TwitterSafety\/status\/1415312509054210052 Is your data even being deleted? While “right to be forgotten” laws are increasingly popular, user data doesn’t simply exist in its raw form on the database, but can also be contained in models trained on their data, which is more difficult to expunge. Since bits of data are intricately embedded in ML models, it can be difficult to guarantee that a user has been completely forgotten without significantly altering the model. The closest solution found to this dilemma is approximate deletion, which allows for the removal of most of the user’s data from the model. The method works at removing specific and easily identifiable information regarding an individual, replacing it with synthetic data, allowing the models to continue working as planned. The question is whether data should be considered “deleted” even if it has not actually been completely removed so long as it doesn’t have any personally identifiable information?","excerpt":"Shoshana Zuboff, the author of “The Age of Surveillance Capitalism,” said, the mining of our personal data for profit is “changing the nature of society” leading to the “wholesale destruction of privacy.” Today companies know so much about us and hold the power to target and manipulate us not only with relevant digital ads, but […]","categories":["IT Services"],"tags":["data deletion"],"author_name":"Srishti Mukherjee","publish_date":"2022-01-22T10:00:00","publication_year":"2022","word_count":794,"keywords":["Go","API","synthetic data","AWS","AI","cloud_platforms:AWS","data-driven","ML","Git","data deletion","R"],"extracted_tech_keywords":["AI","ML","AWS","R","Go","Git","API","synthetic data","data-driven","cloud_platforms:AWS"],"url":"https:\/\/analyticsindiamag.com\/it-services\/what-does-it-take-to-remove-your-personal-data-from-the-internet\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10073951,"title":"Can Text-to-Image AI Learn Ethics —or Is the Future Doomed?","content":"Text-to-image AI generation tools have entered their wild wild west phase. The sweeping trend which Open AI’s DALL.E 2 started with great caution has drastically turned into a world where anything goes. Last week, London and Los Altos-based startup Stability.ai released its open-source text-to-image generator Stable Diffusion to the public. Comparable in quality to DALL.E 2 and Midjourney, the implications of the step taken by Stability.ai are huge and mean exactly this — anyone can now take the code and replicate it to build free apps of their own for text-to-image tasks. Moreover, Stable Diffusion, unlike its predecessors, has next to no restrictions barring users from generating images with inappropriate content or prominent personalities. DALL.E 2’s beta version, which will now be sold to a million people on its growing waitlist, has certain keyword-level prompt filters as well as tweaks in the model that prevents users from generating likenesses of real people. Testing #stablediffusion notebook by @deforum_artI used the init image. It’s amazing at times how some of the images in the video have been turned into cartoon#AIart #deforum pic.twitter.com\/bkwtmRVjfX— Remi (@remi_molettee) August 24, 2022 Level the playing field During the tool’s release, Emad Mostaque, the founder of Stability.ai, said that he wanted to “democratise image generation in a way that would create an open ecosystem that will truly explore the boundaries of latent space”. Mostaque, an Oxford alumnus, has co-founded a project called Symmitree that was started to reduce the cost of smartphones and the internet for the underprivileged sections of society. In a recent interview with The Washington Post, Mostaque spoke with pride about his startup’s independence. Speaking against the perceived guardedness of other Big Tech organisations, Mostaque said, “I believe control of these models should not be determined by a bunch of self-appointed people in Palo Alto. I believe they should be open.” OpenAI’s DALL.E 2 and Google’s Imagen were closed to the public since their launch, except for a select number of artists. While Imagen is still closed, OpenAI sped up the availability of the tool for its waitlist, albeit reluctantly. While experts have recommended open sourcing these tools to improve them, Big Tech was wary of the responsibility and potential censure that would come with the misuse of these tools. It was natural that a smaller player would bite the bullet. And Stable Diffusion did. Dangers of no filters, free for all Since the announcement, images generated by Stable Diffusion started appearing everywhere. A number of art-generating services like Artbreeder and Pixelz.ai caught up quickly and started using Stable Diffusion. (Just last week, an Indian portfolio of media ventures Eros Investments announced a partnership with Stability.ai) And with images that were artistically brilliant and photorealistic, came many problematic ones. The model was leaked on the infamous discussion forum 4chan, popular mainly among young males, which was quick to churn out pornographic content and images with nude celebrities. The unholy combination of open source and restriction-free has immediately brought the possibility of deepfakes much closer than they were. Images of the Russia-Ukraine war, China’s invasion of Taiwan, British PM Boris Johnson dancing with a woman and Indian PM Modi with a knife can all be easily generated with a passable level of skill. Stable Diffusion also surpassed the quality of alternative open source tools like Hugging Face’s Craiyon, (previously known as DALL.E Mini) and Disco Diffusion by a mile. A number of problematic prompts slip through the cracks in Midjourney too. The tool isn’t entirely without any safeguards. It has an adjustable AI tool called Safety Classifier as a part of the Stable Diffusion software package that detects and blocks offensive images. However, the tool can be easily disabled, rendering it useless. People have also developed Google Colabs to create designs from prompts as well as disable watermarks and NSFW filters that were introduced on Stable Diffusion Github version since then. Copyright and Legal Issues Aside from the misuse of these images to produce fake news and create hysteria, copyright issues are another point of contention. Presently, there is little or no legal consensus around the copyright of an AI art piece. In February this year, the US Copyright Office ruled that “human authorship is a prerequisite to copyright protection” for AI-generated art and refused to grant copyright protection for images made by AI. The ruling is currently being appealed against in a federal court. The decision is a tricky one. AI-generated art is mostly drawn from art made in the real, physical world. Artists have found projects that were copying Stable Diffusion and using rip-offs of real artwork made by artists. In another instance, Carlos Paboudjiain, an employee from a creative commerce agency called VMLY&R used DALL.E 2 and Midjourney to produce an entire collection of designer wear in the theme of images from film director Wes Anderson’s work. The benefits of Stable Diffusion are just as heavy as its downfalls. The Pandora’s box that the tool’s release has opened up begs for nuance and detailed consideration into the ethical issues surrounding it. A few days ago, Mostaque addressed the ethics issue in a tweet, “To be honest, I find most of the AI ethics debate to be justifications of centralised control, paternalistic silliness that doesn’t trust people or society. Stable Diffusion represents the internet as is, but we are going to countries and societies globally to change this.” For better or worse, the imagery produced by Stable Diffusion is a true reflection of the beautiful mess that is the internet.","excerpt":"In February this year, the US Copyright Office ruled that “human authorship is a prerequisite to copyright protection” for AI-generated art and refused to grant copyright protection for images made by AI.","categories":["AI Features"],"tags":["AI Tool","Stable Diffusion"],"author_name":"Poulomi Chatterjee","publish_date":"2022-08-30T13:00:00","publication_year":"2022","word_count":916,"keywords":["Go","Hugging Face","OpenAI","AI","Stable Diffusion","ML","Git","Colab","Rust","AI Tool","GitHub","R"],"extracted_tech_keywords":["AI","ML","OpenAI","Hugging Face","Colab","R","Go","Rust","Git","GitHub"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/can-text-to-image-ai-learn-ethics-or-is-the-future-doomed\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10005811,"title":"10 Mind Bending Applications Of Neuralink, According To The Team","content":"Towards the end of the Neuralink event, a handful of team members from different disciplines answered most interesting questions from the Internet. They took turns and expressed what kind of applications excite them the most and shared what they expect to see from Neuralink in the future. 1| For Visual Prosthesis One of the engineers at Neuralink who works in the visual neuroscience department said that this project has the potential to provide a visual prosthesis for people who have retinal injury or blindness through eye injury. The idea is to essentially plug a camera directly into the visual cortex and stimulate an enormous array of thousands or maybe tens of thousands of electrodes to recreate a visual image. And in time perhaps the same technology can be used to produce some kind of heads up display. To this Elon Musk added that we could actually give somebody supervision. One can see the world in various wavelengths like ultraviolet or infrared and radar. All one has to do is basically name their frequency and dynamically adjust the sensor to have a superhuman vision. 2| For Telepathy According to one of the lead chip designers at Neuralink, Telepathy can be the next frontier for Neuralink. He explained that it takes an incredible amount of effort to put thoughts into a set of words. These words are a compressed format of our immense thinking capabilities. Musk too, chimed in on this idea and added that the data rate of words is very low data and we’re putting a tremendous amount of mental energy into compressing the concepts and thoughts on our head into words. With Neuralink one might actually send the true thoughts and communicate far better. Musk called this communication a ‘non-linguistic consent consensual conceptual telepathy’. 3| As An Oscilloscope For Brain Oscilloscopes provide visual information of printed circuit boards(PCBs). Similarly,  the Neuralink device can shed light on many functions of the brain. “The side effect of this device is, you’ll end up learning a ton about how the brain works,” said one of the team members. 4| Unlocking Hidden Creativity As discussed in case of telepathy, our communication systems are inefficient when it comes to translating thoughts. One of the lead researchers believes that there is a lot of untapped creativity in each one of us. “ For example you know to close your eyes and conjure up an incredible like dolly-esque scene but you know if I wanted to actually show someone that it would take years of honing a craft to be able to paint it. With enough electrodes in the right places you could begin to sort of tap into those raw concepts or thought vectors and be able to decode that and show people primitive versions of music or even 3D model for engineering,” he explained. 5| Nostalgia On Demand Memories fade. They get replaced, edited through narratives. As years pile up, the original version is no longer there. Though this sounds like straight out of the movie Memento, it is unfortunately true. So, one of the team members hinted at the idea of memories as repositories. Like music,, one can go revisit the memory and alter the mood on demand. If this is cracked, then cognitive disorders like dementia can be easily dealt with. 6| Eliminate Pain One of the in house neurosurgeons who was present at the event, spoke how pain is essence of basically all human suffering. There are so many diseases today, which can cause lot of pain. Even the treatments can be painful. If this pain is somehow minimised then the way we look at ailments dramatically changes. Neuralink devices can play a crucial role in this too. 7| AI Symbiosis Form the perspective of our species, Musk reminded everyone,  it’s going to be important to figure out how we coexist with advanced artificial intelligence(AI) and achieving some kind of AI symbiosis where an AI extension of one’s self is like a tertiary layer above the limbic system and cortex . Having that symbiosis can be good to have a world in the future that is controlled by the combined will of the people of earth. According to Musk, this going to be important from an existential threat standpoint to achieve a good AI symbiosis and that’s what might be the most important outcome of Neuralink. 8| Consciousness Through The Lens Of Physics One researcher expressed his wish to understand the nature of consciousness. “There’s a lot of very silly philosophy that’s been written about over the last thousand years,” he added. “But i think that we’ve been very limited by the tools and our ability to interrogate. And as we measure the brain these tools get better it will pull it into the realm of physics and it’s really one of the last big great mysteries in science.” 9| Disease Prediction Imagine a disease-free future where you know you know what’s going to happen to you before it happens so you can prevent it with these devices. We’ll be able to not just speak of electrical signals but can also pick up chemical cues in the brain and prevent the diseases ahead of time.. 10| Solve Mental Illnesses The device has the potential to scale to more channels, more regions. As it goes deeper, the chances of solving issues related to anxiety, fear and depression become high. One of the team members wants fear to be eliminated so that she can enjoy her rock climbing. However, Musk quickly suggested to her that some fear might still be useful! All the aforementioned futuristic sounding applications might look quite ambitious. But, we can never bet against Elon Musk. He was even giving a thumbs up when someone suggested if a Tesla could be summoned using Neuralink. So, there definitely seems to be no upper limit to what this coin shaped implant can do. Read about the full event here.","excerpt":"Towards the end of the Neuralink event, a handful of team members from different disciplines answered most interesting questions from the Internet. They took turns and expressed what kind of applications excite them the most and shared what they expect to see from Neuralink in the future. 1| For Visual Prosthesis One of the engineers […]","categories":["AI Trends"],"tags":["Elon Musk","neuralink"],"author_name":"Ram Sagar","publish_date":"2020-08-30T12:00:08","publication_year":"2020","word_count":987,"keywords":["Go","artificial intelligence","programming_languages:R","AI","neuralink","programming_languages:Go","Elon Musk","Ray","ViT","R"],"extracted_tech_keywords":["AI","artificial intelligence","Ray","R","Go","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/mind-bending-applications-neuralink-event\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10080058,"title":"Council Post: How much analytics is actually used?","content":"Since the beginning of civilisation, we have been analysing data. The earliest written records that have survived are as much examples of data analysis as they are of creative works. The first database ever was created by scribes in Sumeria, an ancient region of Iraq. They created lists of the state-employed ploughmen and recorded this information on clay tablets. Additionally, these inscriptions used this raw data to compute their salaries, giving rise to the field of data analytics. Historically, the term analytics dates back to when Henry Ford tracked assembly lines speed. Gartner describes ‘analytics’ as attaining deeper insights, creating forecasts, or coming up with suggestions by way of advanced analytics that examines data or content autonomously or partially autonomously using sophisticated tools and methodologies—often beyond those of traditional business intelligence (BI). Data\/text mining, machine learning, pattern matching, forecasting, visualisation, sentiment analysis, semantic analysis, network and cluster analysis, multivariate statistics, graph analysis, simulation, complex event processing, and neural networks are examples of advanced analytical approaches. The roundtable conference was moderated by Vijoy Basu, Sr. Director, AI & Analytics at Cognizant along with panellists Krishnaswamy Divakaran, Analytics Director, Independent Consulting, Bharat Belavadi, Senior Director & India Head, Advanced Analytics Office at Western Digital, Parikshit Nag, Head of Data & Analytics at Indus OS, Vinay Kumar Sankarapu, Founder & CEO of Arya.ai with the agenda of how much of your analytics is actually advanced. Shift from diagnostic and the descriptive analytics to predictive and prescriptive analytics Most of the organisations are not actually leveraging descriptive analytics to the best of its ability, saying that the answer to why did this happen is probably not very clear at all points in time and there is also a sudden rush to move to the predictive and prescriptive space, which is basically about answering what’s gonna happen next. Three major lookouts on my site before actually moving into the political spaces are to ensure that the reason why we’re doing it is actually very clear, and that understanding the business context is critical. Then, we also need to have the right roadmap to say that even if we build a very complex model, would we be able to sustain it in production and give the desired kind of results which we would want to see? —Parikshit nag, Head of Data & Analytics at Indus OS Dashboards: A bottleneck? Dashboards are fine, but the summarisation is where there could be a big difference in terms of how the user is using this. Currently, there are bots, which can query using some free flow text but this is a simple query. I can derive freighters out of this and then do a search. While this is fascinating or exciting for many users,  the real usability is when I convert a graph into an observation that can be simple or easy to understand for the end user. That makes a difference, then I don’t even need to look at the graph. I can simply look at the observation and go ahead with it. To derive this observation, you may use some model or statistical approach to create observation as simple templated text is also enough. Maybe in the future, we could see multiple text data, use NLP and create on the basis of the observation and try to generate dynamic text by combining multiple parameters. I think I haven’t seen those kinds of outcomes in any tool at this point of time yet. —Vinay Kumar Sankarapu, Founder & CEO of Arya.ai Monetisation of work: A game changer Start looking at the projects focused on the business outcome. We need to be really sure of picking up projects which have more conversion ratio than something which is a capstone that you just try to put your hands into. The second is we bring in a lot more phrases in our executions. The first step is called ‘RLC’, which is a rapid learning cycle. It is non-committal execution where we just learn about the data. Because we want to learn and get some signals out of the data to make sure that this really is a data science project. Then we break down the bigger problems into small RLC’s, so that we can give a smaller victory instead of the bigger one. We also move into—once the RLC is successful—‘rapid prototyping’. That means you probably do it for one particular product range for the one line of the factory, or maybe one component in our product families and so on. There is a confidence that comes with this process to then start scaling it. —Bharat Belavadi, Senior Director & India Head, Advanced Analytics Office at Western Digital Lack of uniformity We are stuck a little bit in a cocoon because there isn’t a homogeneity of analytics implementation across industries. Like the factory floor impact, digital organisations, mobile-based and paper-based data have some correlation. So that assimilation of how the infrastructure exists is something that I always think that we as an industry need to create a garden of. If you look at IT as an industry, it has ensured that the academics, the rules, the guidance have created a kind of structure so that you at least are moving in a certain fashion together, irrespective of which firms you are working in one day or another. Somehow, I have not seen that comminglation in analytics across different industries come out very openly. People do talk and not that they don’t, but in terms of action and impact they don’t. This is the root cause of the problem we’re facing. —Krishnaswamy Divakaran, Analytics Director, Independent Consulting Businesses are seeking innovative methods to utilise the massive volumes of data being produced every day. Companies can achieve these goals with the use of advanced analytics. It enables businesses to develop and optimise their processes to gain competitive advantage. Improved consumer insight, predictive analytics, and statistical modelling provided by advanced analytics not only help businesses make better decisions but also help them keep up with rapidly evolving markets. “Information is the oil of the 21st century, and analytics is the combustion engine.”—Peter Sondergaard","excerpt":"Businesses are seeking innovative methods to utilise the massive volumes of data being produced every day. Companies can achieve these goals with the use of advanced analytics. It enables businesses to develop and optimise their processes to gain competitive advantage.","categories":["AI Features"],"tags":["Advanced Analytics"],"author_name":"AIM Media House","publish_date":"2022-11-18T12:00:00","publication_year":"2022","word_count":1019,"keywords":["data science","machine learning","AI","neural network","sentiment analysis","R","RAG","NLP","analytics","predictive analytics","Advanced Analytics"],"extracted_tech_keywords":["AI","machine learning","neural network","NLP","data science","analytics","RAG","predictive analytics","sentiment analysis","R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-much-analytics-is-actually-used\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10115596,"title":"Microsoft Copilot for Security to be Generally Available Soon","content":"Microsoft has announced the general availability of Microsoft Copilot for Security, set to debut on April 1, 2024. This groundbreaking AI-driven solution, a first of its kind in the industry, is poised to revolutionise the capabilities of security and IT professionals by providing them with enhanced tools to detect threats, expedite response times, and augment team expertise. Excited to announce this milestone for Copilot for Security, as we continue to apply this next generation of AI to empower defenders across all experience levels and transform every aspect of SOC productivity. https:\/\/t.co\/2IMKcAs0OC— Satya Nadella (@satyanadella) March 13, 2024 Powered by vast volumes of data and threat intelligence, including over 78 trillion security signals processed daily by Microsoft, Copilot leverages advanced language models to offer tailored insights and actionable guidance. This innovative approach empowers users to safeguard their systems at the rapid pace and scale facilitated by artificial intelligence, thus transforming the landscape of security operations. In a bid to make Copilot for Security more accessible to a diverse array of organisations, Microsoft has introduced a provisioned pay-as-you-go licensing model. This flexible pricing structure enables swift deployment and allows organisations to scale their usage and costs in alignment with their specific needs and budgetary constraints. Furthermore, the global rollout of Copilot for Security underscores its multilingual capabilities, with support for eight languages in prompt processing and response, alongside a multilingual interface catering to 25 different languages. This broad linguistic coverage ensures readiness for deployment across major regions spanning North and South America, Europe, and Asia. Comprising over 100 partners, including managed security service providers and independent software vendors, these partnerships play a pivotal role in facilitating widespread adoption of AI-driven security solutions and empowering organisations to navigate the evolving threat landscape with confidence. As part of its general availability, Copilot for Security introduces several new capabilities, including custom promptbook creation, knowledge base integrations, and third-party integrations, aimed at enhancing operational efficiency and threat mitigation strategies. Microsoft is equipping customers with enhanced visibility, protection, and governance mechanisms. These initiatives encompass the discovery of AI risks, safeguarding of AI applications and data, and governance of AI usage, ensuring secure and responsible adoption of AI technologies. Moreover, Microsoft continues to bolster its comprehensive suite of security offerings, with recent additions such as Microsoft Security Exposure Management and Adaptive Protection integrated with Microsoft Entra Conditional Access. These enhancements underscore Microsoft’s unwavering commitment to providing customers with end-to-end protection for their digital assets, bolstering resilience against emerging threats. The launch of Microsoft Copilot for Security heralds a new era in cybersecurity, empowering organisations worldwide to fortify their defences, respond swiftly to evolving threats, and embrace the transformative potential of artificial intelligence in safeguarding digital assets.","excerpt":"Microsoft has announced the general availability of Microsoft Copilot for Security, set to debut on April 1, 2024.","categories":["AI News"],"tags":["Microsoft"],"author_name":"Mohit Pandey","publish_date":"2024-03-14T10:14:15","publication_year":"2024","word_count":448,"keywords":["Go","API","artificial intelligence","AI","Git","RAG","Ray","Aim","GAN","R","Microsoft"],"extracted_tech_keywords":["AI","artificial intelligence","Aim","Ray","RAG","R","Go","Git","API","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/microsoft-copilot-for-security-to-be-generally-available-soon\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10019491,"title":"RPA Startup UiPath Raises $750 Million In Series F Funding To Enable Fully Automated Enterprise","content":"UiPath Inc., robotic process automation (RPA) startup, has announced that the company has raised a $750 million funding in Series F round at a post-money valuation of $35 billion. As compared to its $3 billion valuations in 2018, UiPath has come a long way with its mission to unlock human creativity and ingenuity by enabling the Fully Automated Enterprise and empowering workers through automation. The funding round was co-led by existing financiers Alkeon Capital and Coatue, along with other investors that include Altimeter Capital, Dragoneer, IVP, Sequoia, and Tiger Global. On the other hand, the funds and accounts are advised by T. Rowe Price Associates, as per the company’s official release. According to the news media, the total raised money by UiPath, till date, totals to nearly $2 billion. In its last series of funding, the company had raised $225 million. Founded in 2005, this New York-based startup is known for providing a complete software platform that helps organisations efficiently automate business processes. Its Automation Platform is designed to transform the way humans work, providing customers with a robust set of capabilities to discover automation opportunities and build, manage, run, engage, measure, and govern automation across departments within an organisation, stated by the release. Also Read: RPA Is Now Becoming Commoditised, Says Manish Bharti, President – UiPath India The company leverages technologies like artificial intelligence and machine learning to automate or altogether “end repetitive tasks,” which will not only “unleash human potential” but will also improve customer experience. This, in turn, allows humans to focus their energy on more complex jobs like developing strategy. When asked the investors, Abhi Arun, the Managing Partner at Alkeon Capital, commented that considering automation has become a strategic imperative, it is fundamentally changing the way businesses operate. “We are excited to co-lead this round of funding, as well as continue to team up with the UiPath team during an important phase for the company.” With prominent clients like Airbus, Google, Virgin Media, Fujifilm, McDonald’s in its kitty, UiPath is currently the fastest-growing RPA startup. According to the news media, the company gradually increased its client base from merely 100 customers in 2017 to over 5,000, with more than 750,000 users. Later last year, UiPath has even partnered with Telangana Academy for Skill and Knowledge to build RPA resiliency among the workforce.","excerpt":"UiPath Inc., robotic process automation (RPA) startup, has announced that the company has raised a $750 million funding in Series F round at a post-money valuation of $35 billion. As compared to its $3 billion valuations in 2018, UiPath has come a long way with its mission to unlock human creativity and ingenuity by enabling […]","categories":["AI News"],"tags":["uipath","UiPath India","UiPath RPA"],"author_name":"Sejuti Das","publish_date":"2021-02-02T10:30:24","publication_year":"2021","word_count":387,"keywords":["Go","API","artificial intelligence","machine learning","AI","UiPath RPA","RAG","automation","ViT","uipath","UiPath India","GAN","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","RAG","R","Go","API","GAN","ViT","automation"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/rpa-startup-uipath-raises-750-million-in-series-f-funding-to-enable-fully-automated-enterprise\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10077035,"title":"IIT Mandi Launches Certificate Courses in Data Science &#038; ML","content":"The Indian Institute of Technology, Mandi (IIT-Mandi) has collaborated with National Skills Development Corporation (NSDC) to launch certificate courses on data science and machine learning. The programme is set to commence in the first week of November. IIT Mandi faculty members will conduct the course through live-streaming sessions. The programme will allow students to build a foundation in data science and specialize in machine learning with Python to enhance data-driven decision-making. Upon completion, students can explore careers as data scientists, business analysts, data analysts, and business intelligence professionals on new-age skills. These are six- and nine-month certificate and advanced certificate programmes, respectively. The selection of candidates will be based on an eligibility test conducted on October 15 and 16. Upon course completion, the programme certificate will be collectively issued by IIT Mandi and NSDC. Professor Tushar Jain, head at Center for Continuing Education (CCE), said, “The global demand for data science and machine learning experts is growing rapidly. These certificate programmes are designed to impart these skills to students.” The live online certification course would cover all the fundamental and advanced skills in the tools used in data science, deep learning, machine learning, and image processing. The course’s main objective is peer-to-peer learning and to gain experience of sessions from faculty of IIT Mandi to boost the overall professional journey of the applicants. For further details, interested candidates can check here: futureacad.com\/iit-mandi.","excerpt":"The programme will allow students to build a foundation in data science and specialize in machine learning with Python to enhance data-driven decision-making","categories":["AI News"],"tags":["Courses","IIT"],"author_name":"Bhuvana Kamath","publish_date":"2022-10-11T15:20:46","publication_year":"2022","word_count":232,"keywords":["business intelligence","data science","API","machine learning","AI","data-driven","Python","deep learning","programming_languages:Python","Courses","R","IIT"],"extracted_tech_keywords":["AI","machine learning","deep learning","data science","Python","R","API","business intelligence","data-driven","programming_languages:Python"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/iit-mandi-launches-certificate-courses-in-data-science-ml\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10055971,"title":"How Mozilla Tests Firefox With A Machine Learning Model","content":"The Continuous Integration system at Mozilla includes 85,000 files, each containing many test functions. These tests need to run on all platforms – Ios, Windows, and Linux. However, it is impossible to test each function on all platforms. Instead, Mozilla had developed a strategy to test 90 unique functions, since running every configuration would amount to 2.3 billion test files per day. The functions were chosen through the principles of importance and relevancy, and the configurations were tested through the integration branch. However, the heuristics of ranking tests based on the frequency of failure was naive as it did not consider the contents of the patch. Additionally, choosing tests can be a time-consuming task and may lead to over selection. The Mozilla developers chose to run machine learning algorithms as they hypothesised that it could result in the quicker, efficient, and economical selection of optimal tests to run. Hence, the developers constructed infrastructure to ensure the smooth execution of the CI pipeline. Getting Started To build the training model, the developers initially had to solve the problem of naive heuristics. They built a set of complex heuristics to predict which patch causes which regression. Some failures are classified\/annotated by humans as ‘intermittent’ or ‘fixed by commit’ and help find the regressions of the missing or intermittent tests. Since 100% accuracy is not attainable, the developers have built heuristics to evaluate the classifications. Apart from solving the problem of the heuristic, the developers also have to collect data on the patches themselves. They correlate with the test failure data, which provides deterministic results for ML models as to which tests are more likely to fail for a given patch. With the complex heuristics, dataset of patches and associated tests, the developers at Mozilla have built a training set and a validation set to teach the ML model how to select optimal tests. 90% of the data set is the training set, and the other 10% is the validation set. The validation is carefully chosen to be posterior to the training set to avoid information leakage. The preventative measure reduces the risk of a biased and artificially enhanced ML model. Mozilla developers then train XGBoost models using tests, patches, and links features. The model is trained using the TUPLE test patch input, and the output is a single binary that determines whether or not the patch passes the test. A single model can run all tests. Optimizing Configuration To further optimise the running of tests, the developers choose to test and improve the selection of the tests and where the tests should run. By doing so, they can identify the redundant configurations with the help of the datasets that have been collected. These redundancies are identified; the developers use solutions similar to frequent itemset mining. Results The model provides frontend and backend services, and RedisQueues bridges the gap. The frontend exposes a simple REST API, so consumers need to only specify the push they are interested in (identified by the branch and topmost revision). The backend will automatically determine the files changed and their contents using a clone of Mozilla-central. Depending on the size of the push, the service can run up to a few minutes to yield the desired results. Developers only queue one job at a time and cache the results once computed. Developers at Mozilla are using the model for scheduling tasks on the integration branch and running the special mach try auto command to test changes on the testing branch. Performance Assessment Finally, the developers assess, measure and compare the results and success of the algorithms. The variables kept in mind are- (i) the number of resources used (Hours, dollars), and (ii) the regression detection rate. Alongside the scheduled algorithm, a shadow scheduler showcases the output it would have scheduled if it was the default. The scheduler effectiveness of the various shadows is plotted on a dashboard, and the algorithm with the best performance is made the default algorithm. Source: https:\/\/hacks.mozilla.org\/2020\/12\/cross-browser-testing-part-1-web-app-testing-today\/","excerpt":"Explanation on how Mozilla tests Firefox with the help of a machine learning model infrastructure","categories":["AI Features"],"tags":["Mozilla"],"author_name":"Abhishree Choudhary","publish_date":"2021-12-16T11:00:00","publication_year":"2021","word_count":663,"keywords":["Go","API","machine learning","TPU","Mozilla","AI","ML","REST API","XGBoost","R","Redis"],"extracted_tech_keywords":["AI","machine learning","ML","XGBoost","TPU","Redis","R","Go","API","REST API"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-mozilla-tests-firefox-with-a-machine-learning-model\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":30782,"title":"Deep Learning Has Become The Go-To Method For Problem Solving","content":"Deep learning has been one of the most popular substrates of the machine learning family, in the last decade or so. The reason for its popularity is its sky-high accuracy. But what makes deep learning so integral in problem-solving? The ‘deep’ part of deep learning comes with more than one layer in its architecture, along with which a number of advantages are introduced, as it is then able to classify data more and more accurately and can learn better and so have even found their application in self-driving cars. Why Is Deep Learning Better? Ease Of Problem-Solving: Other methods need the problem to be broken down into small steps to create an integrated result. The process of problem-solving in deep learning does not want to be broken down into small steps. It solves problems on an end-to-end basis. Easily Adaptive: These methods pass data directly to the network and have a good performance, unlike other methods where the best features are carefully selected to pass over to the algorithm. Due to this, it can be adapted to different domains easily. For example, once a speech recognition technique is understood, NLP isn’t a big hurdle. This gives higher performance is a short time period. Greater Accuracy: Deep learning methods themselves learn to extract features from the dataset. The only way to increase the accuracy is to input more data and the accuracy attained is much higher in these methods. For example, in case of image classification, if the image set has no much logical contradiction due to conditions like lens distortion or product reflection, deep learning comes to the rescue and handles the classification with greater accuracy. The image below one of the slides on Andrew Ng shows how larger data helps the performance of deep learning methods, over the older methods. Image source: Baidu Research. Andrew Ng slide. Why Do They Perform Well? Working With A Large Amount Of Data: Deep learning methods are hungry for large datasets because they work the best with them. The way to improve the accuracy of these methods is just by making the dataset larger and they, therefore, have a better result. For this reason, they are also great for dealing with complex problems. However, more data will not always help. Sometimes it will need to rescale, invent and transform as well. For example, if the data is vectors or numbers, create random numbers, or if it is images, creating random versions of images will have to be done. But most issues are solved with just data. Feature Extraction Is Included In The Training: One of the major reasons why deep learning methods are gaining attention is because they try to learn features from data in an incremental manner. The feature extraction process is trained along with the classification processes. For example, a set of image filters or primitives is trained in the first layers of the classification network for image recognition. AlexNet, for instance, passes the images without any feature vector and the feature is extracted itself. This eliminates the need for staunch feature extraction since the feature extraction is included within the training itself. Other methods work only with a set a predetermined data. If the data is changed, it fails and has to be trained again. Deep learning adapts to the changes. It analyses problems in its hidden layers that are otherwise computationally difficult to solve. It reduces the task of generating new feature extractors for every data. The system can learn from mistakes and enhance itself, thereby eliminating the big challenge of feature engineering step. Concluding Note DL methods are always data hungry. They can be a failure when the dataset is small, or when the test set differs from the training set greatly. And some problems cannot be thought of as classification problems at all. In such occasions, Gary Marcus of the University of New York in his paper says that deep learning becomes a square peg slammed into a round hole, a crude approximation when there must be a solution elsewhere. Hence, it is not suitable for small datasets. Apart from that, they are very computationally expensive, because of the large set of networks that it has. But the main objective of deep learning lies in modelling the complex, hierarchical features in the dataset and so far it has proved its worth in this domain since they have shown their worth in AI areas of NLP, speech recognition, image analysis and recommendation engines.","excerpt":"Deep learning has been one of the most popular substrates of the machine learning family, in the last decade or so. The reason for its popularity is its sky-high accuracy. But what makes deep learning so integral in problem-solving? The ‘deep’ part of deep learning comes with more than one layer in its architecture, along […]","categories":["AI Features"],"tags":["feature extraction","Neural Network","recommendation engine","Speech Recognition"],"author_name":"Disha Misal","publish_date":"2018-11-28T04:58:25","publication_year":"2018","word_count":745,"keywords":["Neural Network","Go","machine learning","programming_languages:R","AI","image recognition","feature engineering","programming_languages:Go","NLP","recommendation engine","deep learning","feature extraction","Speech Recognition","R"],"extracted_tech_keywords":["AI","machine learning","deep learning","NLP","image recognition","R","Go","feature engineering","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/deep-learning-has-become-the-go-to-method-for-problem-solving\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10040876,"title":"Oracle Launches Industry’s Most Generous Free Tier Cloud Computing Service","content":"Oracle has announced its first Arm-based compute offering, OCI Ampere A1 Compute, along with a range of tools, solutions and support to enable Arm-based application development. Oracle also claimed it’s the only major cloud provider to offer its Arm compute instances at only one cent per core hour, the industry’s lowest cost per core and with flexible VM sizing. This means customers can run cloud-native and general-purpose workloads on Arm instances with significant price-performance benefits. Available on Oracle Cloud Infrastructure (OCI), the cloud computing service is powered by data centre chips from Ampere Computing based on technology from Arm Ltd. “The increasingly distributed nature of work means modern applications don’t just live in the cloud, it lives at the edge.  In the Asia Pacific, this is being driven by smart industry-edge applications, real-time analytics and IoT, as businesses seek to improve operations and deliver new experiences to their customers,” said Chris Chelliah, Senior Vice President, Technology and Customer Strategy, Asia Pacific & Japan. He said applications need infrastructure that is open, extremely efficient, scalable and secure and this is what ARM architectures deliver. Additionally, to help customers take advantage of the latest in Arm technology, Oracle is working closely with a wide variety of technology and open source partners, such as GitLab, Jenkins, Rancher, Datadog, OnSpecta, NGINX, and Genymobile. Oracle is also joining the Continuous Delivery Foundation (CDF), an open-source, vendor-neutral community for sustaining the fastest-growing CI\/CD open-source projects. When running x264 video encoding workloads on OCI Ampere A1, Oracle saw up to a 10 percent performance increase, and up to a 22 percent price-performance benefit compared to x86 based systems. Oracle saw up to a 46 percent performance increase, and up to a 62 percent price-performance benefit compared to x86 based systems for NGINX reverse proxy workloads on OCI Ampere A1. “Ampere instances on OCI is a breakthrough for developers. Oracle’s Free Tier is a great offering that allows them to test the OCI Ampere A1 compute platform and experience the first-cloud native processor that delivers predictable performance, scalability and power needed,” said Renee James, founder, chairman and CEO, Ampere Computing in a statement.","excerpt":"Oracle has announced its first Arm-based compute offering, OCI Ampere A1 Compute, along with a range of tools, solutions and support to enable Arm-based application development.  Oracle also claimed it’s the only major cloud provider to offer its Arm compute instances at only one cent per core hour, the industry’s lowest cost per core and […]","categories":["AI News"],"tags":["AWS","OCI","Oracle","Oracle Cloud Infrastructure"],"author_name":"Shanthi S","publish_date":"2021-05-26T14:03:01","publication_year":"2021","word_count":355,"keywords":["OCI","AWS","AI","cloud computing","CI\/CD","Scala","Oracle","Git","Aim","GitLab","analytics","real-time analytics","R","Oracle Cloud Infrastructure"],"extracted_tech_keywords":["AI","analytics","Aim","cloud computing","R","Scala","Git","GitLab","CI\/CD","real-time analytics"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/oracle-launches-industrys-most-generous-free-tier-cloud-computing-service\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10095575,"title":"Will Ali Farhadi Make Allen Institute for AI Great Again?","content":"Computer vision expert Ali Farhadi is set to return to Allen Institute for AI (AI2) on July 31 as the second chief executive officer in the Seattle-based non-profit research organisation’s history, after leading Apple’s AI and ML initiatives since January 2020. He is also the man behind YOLO Object detection and was a senior research manager at AI2 from 2014 before joining Apple. Looking Back Farhadi is a stalwart in the AI community. He completed his doctorate in computer science at the University of Illinois at Urbana–Champaign, specialising in computer vision and working under the supervision of South-African-born American computer scientist David Forsyth. Following that, he worked as a post-doctoral fellow at Carnegie Mellon University’s Robotics Institute, under the guidance of Russian-American computer scientists Alexei A. Efros and Martial Hebert. Currently, Hebert holds the position of director at the robotics institute. Read more: What is Microsoft Without OpenAI? Farhadi is also a professor at Paul G Allen School of Computer Science and Engineering, University of Washington where he leads the Reasoning, AI and VisioN (RAIVN) lab along with professor Ranjay Krishna. He also co-founded the first on-device deep learning company, XNor.ai with Mohammad Rastegari which was acquired in 2020 by Apple at $200, aiming at their increasing interest in investing in AI capabilities. When Apple acquired Xnor.ai, it aimed to leverage its expertise in AI on the edge, which allows ML and image recognition to be executed on low-power devices rather than relying on the cloud, aligning with their focus on privacy and preserving data on mobile devices. The AI2 backed Xnor.ai also saw investments from Madrona Venture Group and the University of Washington. Farhadi’s research have garnered worldwide acclaim. His profound contributions have won accolades such as the best paper awards at esteemed conferences including CVPR, NeruIPS, AAAI, and NSF. In 2014, he was honoured as an Allen Distinguished Investigator. The illustrious Sloan Research Fellowship was bestowed upon him in 2017, further cementing his reputation as an exemplary scholar. In recognition of his achievements, he received the prestigious Early Career Academic Achievement Alumni Award from UIUC in 2019. Furthermore, Farhadi was among the esteemed scholars selected as the inaugural recipients of the Google Fellowship. A Deserving Successor Farhadi takes over from Oren Etzioni, who was chosen by Microsoft co-founder Paul Allen to establish AI2 about 10 years ago. Etzioni revealed his intention to resign as CEO last year, stating that he would remain on the board, provide guidance, and serve as the technical director of the AI2 Incubator. Researcher Peter Clark has been the interim CEO. “As we face unprecedented changes in the development and usage of AI, I could not think of a better time to return to AI2 as CEO,” said Farhadi. He further added that today more than ever, the world needs truly open and transparent AI research that is grounded in science and serves as a platform where data, algorithms, and models are open and available to all and renowned researchers and engineers at AI2 find themselves in a unique position to spearhead this innovative and trustworthy approach to AI development. In 2010, Etzioni joined AI2 and became a significant figure in the AI field. He was an accomplished entrepreneur and also taught at the University of Washington’s Department of Computer Science and Engineering. With over 15 PhDs, Oren specialised in areas such as machine reading, data mining, web search, and software agents. Etzioni and Allen established the Allen Institute for AI in early 2014. Outstanding! The sky's the limit with Ali Farhadi! https:\/\/t.co\/ZWzi3TkcxS— Oren Etzioni (@etzioni) June 20, 2023 What Next? A month ago, AI2 announced the building of AI2 OLMo, an open generative language model with over 70 billion parameters that is expected to launch in early 2024. Unlike other models, OLMo is designed by scientists for scientists, with a focus on advancing language model research. The initiative aims to provide access, education, and collaboration opportunities to the research community and the public. The project is being developed in partnership with AMD and CSC, using the environmentally friendly LUMI supercomputer. Earlier, former chief scientist of Tesla Andrej Karpathy returned to OpenAI. Several other Google and Meta researchers joined OpenAI recently. Similarly, Ian Goodfellow left Apple to join Google DeepMind. So now with Farhadi, we only hope that AI2 will also report some breakthroughs in the near future, just like DeepMind and OpenAI. Read more: Paul Allen liked the fact that I wasn’t an academic: Dr Oren Etzioni, CEO, AI2","excerpt":"Farhadi is the man behind the groundbreaking YOLO Object detection","categories":["AI Trends"],"tags":["Allen Institute for AI"],"author_name":"Shritama Saha","publish_date":"2023-06-22T10:00:00","publication_year":"2023","word_count":746,"keywords":["OpenAI","AI","ML","image recognition","computer vision","RAG","Allen Institute for AI","Aim","deep learning","object detection","R"],"extracted_tech_keywords":["AI","ML","deep learning","computer vision","OpenAI","Aim","RAG","image recognition","object detection","R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/will-ali-farhadi-make-allen-institute-great-again\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":49490,"title":"7 Innovations By Baidu Which Changed The Face Of AI","content":"Compared to other countries, China has become one of the leading artificial intelligence innovators in the world. By investing big bucks in emerging technologies, the country has shown potential for becoming the global AI leader. And this has become possible because of the Chinese tech giants who are striving hard to make every possible move towards a smart society. Chinese multinational technology company Baidu, which specialises in Internet-related services and artificial intelligence-powered solutions, has been doing a lot in this sector. The tech giant offers a number of services and one of its popular services is the Baidu Chinese search engine which is considered as the second-largest search engine in the world and holds a 76.05% market share in China’s search engine market. In this article, we list down the top innovations in artificial intelligence by Baidu. 1| Apollo Apollo is a high performance, flexible architecture which accelerates the development, testing, and deployment of Autonomous Vehicles. It was announced at CES 2019 as the most advanced and open-sourced autonomous driving platform. The tech giant claims that this platform has the ability to manage speed bumps, clear zones, narrow lanes, unprotected turns and much more. The Apollo Enterprise includes fully autonomous mini-buses, autonomous valet parking, intelligent map data service platform and DuerOS. 2| Duer In 2015, Baidu launched its intelligent personal assistant, Duer which can also be called the Chinese version of Apple Siri. Duer includes multi-modal interaction, natural language processing, and other such technologies for a natural interaction and smarter understanding. 3| Deep Speech Deep Speech is a state-of-the-art speech recognition system developed using end-to-end deep learning by Baidu Research. This system is a combination of a language model which is capable of outperforming existing state-of-the-art recognition pipelines in two challenging scenarios which are clear, conversational speech and speech in noisy environments. 4| DuerOS DuerOS is a conversational artificial intelligence open operating system developed by Baidu Duer Business Unit. It can be considered as one of the leading applications of Baidu’s AI technology which delivers solutions with the help of natural language processing (NLP) by allowing the users to command and converse with their devices. 5| Deep Voice Deep Voice is a real-time Neural Text-to-Speech system developed by the Chinese tech giant in 2017. The Deep Voice project was started to revolutionize human-technology interactions by applying modern deep learning techniques to artificial speech generation. This production-quality text-to-speech system is constructed entirely from deep neural networks and it comprises five major building blocks which are a segmentation model for locating phoneme boundaries, a grapheme-to-phoneme conversion model, a phoneme duration prediction model, a fundamental frequency prediction model, and an audio synthesis model. 6| Little Fish Jointly developed by Baidu and Ainemo Inc., Little Fish or Xiaodu Zaijia is a smart video speaker and the winner of the Best Smart Home Product Award at CES Asia 2018. This smart speaker with a screen and home functions aims to help close the distance between family members and let people enjoy time at home. 7| SwiftScribe SwiftScribe is an AI-powered transcription software developed by Baidu Silicon Valley Lab (SVAIL). The basic function of SwiftScribe is to transcribe audio material into the text in order to solve the problem of consuming a large amount of time-by-word dictation. This web application is based on Deep Speech 2 and supports uploading .wav or .mp3 files up to 1 hour long. It uses artificial intelligence for transcription and currently, it only supports English transcription.","excerpt":"Compared to other countries, China has become one of the leading artificial intelligence innovators in the world. By investing big bucks in emerging technologies, the country has shown potential for becoming the global AI leader. And this has become possible because of the Chinese tech giants who are striving hard to make every possible move […]","categories":["AI Trends"],"tags":["AI innovations","Baidu","china ai","real time face recognition software"],"author_name":"Ambika Choudhury","publish_date":"2019-11-06T12:00:12","publication_year":"2019","word_count":575,"keywords":["artificial intelligence","programming_languages:R","AI","neural network","innovation","Modal","NLP","china ai","Aim","deep learning","Baidu","R","AI innovations","real time face recognition software"],"extracted_tech_keywords":["AI","artificial intelligence","deep learning","neural network","NLP","Aim","R","innovation","Modal","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/7-innovations-by-baidu-which-changed-the-face-of-ai\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10124381,"title":"How to Reduce Hallucinations in LLMs for Reliable Enterprise Use","content":"Gen AI has been well-received by enterprise decision-makers. Yet, wary of technology pitfalls from past experience, they have expressed serious concern about hallucinations, which are demonstrably false model responses. Just how challenging is the problem? A recent study found that LLMs may hallucinate between 3-27% of the time, depending on the model. In specific contexts, this may be much worse. Another study found that LLMs provide false legal information between 69-88% of the time – very worrying given the criticality of legal transactions. When LLMs Lie What are hallucinations? Large language models (LLMs), such as GPT-4, Llama 3, and Mixtral, can generate rapid, fluent responses to varied user prompts in many scenarios. But some of these are nonsensical, some are untruthful but hard to detect as incorrect, and a few are accurate but not derived from source data, all categories that make it seem that the model is hallucinating. The underlying reasons that drive the LLMs to behave this way include not giving them enough context while training, overfitting them, and data ingestion errors such as wrong encoding. It is still early days, but undesirable scenarios caused by hallucinations may convince enterprise leaders to pull back on funding gen AI initiatives and make business heads reluctant to pilot or deploy solutions. To begin with, business users may hesitate to use LLM tools in their daily work when they find that they cannot trust the output, setting up a significant barrier to adoption. Second, even a single failure to detect hallucinations in sensitive use cases, such as health care, can cause serious harm to external stakeholders like patients and significant reputational damage to the organization, negating any ROI. Guiding Models to Tell the Truth: Why Retrieval Augmented Generation (RAG) helps Technology leaders, extremely aware of the urgent need to increase the reliability of LLM outputs, are quickly creating effective governance, including automated and human-guided accuracy checks. The strategies being tried include using guardrails with prompts, providing examples of the desired output while querying, and regularly fine-tuning the data sets that train the LLMs. While constant fine-tuning of massive data sets requires significant resources, the other two approaches are not structured enough to guarantee reliability. This brings us to a promising fourth route – retrieval augmented generation (RAG). RAG leverages the robust self-learning mechanism of an LLM while focusing it on a limited set of pre-approved and up-to-date information sources. For instance, if an internal user not in the finance team wants to know the company’s latest turnover, a model not restricted by RAG may pick up these numbers from external websites of low credibility. However, an RAG-restricted model can be instructed to get the numbers from the latest internal financial updates, ensuring accuracy. Hallucination Detection to improve the Reliability of RAG-restricted LLM Outputs Restricting the LLM’s behavior with RAG can boost the reliability of its output and reduce hallucinations, but it does not completely eliminate them. Consider a marketing team using a tailored RAG-led LLM to scour the web for campaign ideas. The LLM may come up with something from a successful competitor campaign, not understanding that while it has to look at what competitors are doing, it cannot use the information for ideas. Data scientists are fast building strategies to avert such disasters. To spot hallucinations in the Black Box LLMs used mainly by enterprises today, SelfCheckGPT, a recent research paper on hallucination detection, offers three recommendations – the BERT Score that uses semantic similarity, the prompt method that uses another LLM and its understanding of language to evaluate consistency, and evidence-based evaluation leveraging natural language inference (NLI). To test which of these is the most effective within an RAG set-up, we built an RAG system based on the Llama 2-13B-chat model using a corpus of financial reports, and relevant questions, and applied the three approaches to evaluate the responses. By helping to generate hallucination-free output 88.63% of the time with optimal resource utilization, the NLI-led method knocked down the competition. But is this enough, given that gen AI will soon see adoption in high-stakes situations where people’s lives or millions of dollars are on the line? To spot hallucinations to a finer degree by identifying them within responses, we recommend the integrated gradient approach that uses a baseline to detect hallucinations with up to 99% confidence. In the marketing use case, for example, the integrated gradient method will unerringly pick out parts of the responses that do not tally with the company’s brand and style guidelines. Ready for Enterprise-level Adoption The combined RAG, NLI, and integrated gradient methodology give enterprises a winning strategy for gen AI adoption. Users can confidently isolate problematic responses, increasing their trust in model output and making them amenable to use the technology frequently. While competitors struggle to tame pilot projects, IT teams that consistently generate high-quality output using this three-pronged method can rapidly scale LLMs enterprise-wide. Generative AI can be extended to more use cases and complex workflows, empowering employees with new insights, increasing ROI, and cementing competitive advantage.","excerpt":"It is still early days, but undesirable scenarios caused by hallucinations may convince enterprise leaders to pull back on funding gen AI initiatives and make business heads reluctant to pilot or deploy solutions.","categories":["AI Highlights"],"tags":[],"author_name":"Ankush Chopra","publish_date":"2024-06-24T16:30:46","publication_year":"2024","word_count":835,"keywords":["Go","API","TPU","AI","RAG","BERT","generative AI","Rust","retrieval augmented generation","R"],"extracted_tech_keywords":["AI","generative AI","retrieval augmented generation","RAG","TPU","R","Go","Rust","API","BERT"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/how-to-reduce-hallucinations-in-llms-for-reliable-enterprise-use\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10015821,"title":"Amazon’s CTO Picks Top Tech Trends That Will Rule 2021","content":"“Rather than slowing us down, 2020 accelerated our shift to a digital world. From my vantage point, 2021 will be a launchpad for all kinds of change.” Dr Werner Vogels, VP & CTO, Amazon. Amazon has seen a tremendous growth with its cloud operations fueled by innovation in the machine learning and the chip department. One thing that Amazon does well is forecasting the customer’s needs and adjusting to changing markets. The top brass of the company, Dr Werner Vogels, CTO, Amazon has picked eight trends that are going to rule 2021. Here we take a look at a few of those. Death Of The Keyboard & All Things Virtual On an average day, 80% of messages on Twitter contain some kind of image or video or are just an image or video. This highlights that people are exploring new ways of interacting with each other. Videos, images and audio are slowly replacing the traditional form of communication. This shift to “natural form of communication”, as Dr Vogels puts it, will also enable greater equity when it comes to accessing services and information. This, in turn, will generate more data and will eventually lead to a rise in demand for exceptional cloud services. Zoom calls, virtual conferences, remote classrooms and what not — virtual ecosystems have gone mainstream in 2020. Dr Vogels believes that this trend will continue next year as well, and with the help from advanced data analytics, organisations can enhance their virtual setups. In 2021, believes Dr Vogels, remote work will become more common and might even turn out to be a better option for some. While he agrees that it is important to send kids back to classrooms, he asserts that remote classrooms give school systems the flexibility to respond to unforeseen events — whether it be pandemics, natural disasters, or man-made calamities — to continue student education. “We don’t need a global health crisis for online classes to make sense. Having remote work options widely available at any time means that kids can stay home when they’re sick without falling behind. Or what if there’s no school to go to at all? If there’s an internet connection, there’s the possibility for some type of education,” explained Dr Vogels. Machine Learning On Edge “As machine learning continues to expand, we will also see an explosion in machine-to-machine connections.” Whether in healthcare or other applications, wrote Dr Vogels, the only realistic way to handle all the information is to use ingestion and aggregation tools, married to ML models. Machine learning applications, historically have faced roadblocks due their compute intense nature. The modern integrated hardware-software solutions, however, open up new avenues for ML to be leveraged. The advent of edge computing, especially, has made this possible. Dr Vogels believes that more institutions will move towards the edge in 2021. “We will see ML embedded on production lines, able to spot production anomalies in real-time. In agriculture, machine learning models will help farmers manage precious resources, such as soil and water, more intelligently,” he added. ML models on edge devices can also help predict natural hazards like wildfires and other disasters with time series calculations on the ground — rather than back in a central data centre. AWS offers its customers a wide range of ML features that can cater to all the growing demands of the coming year. For example, AWS Sagemaker — its flagship service — has added a total of 250 features this year alone. And, Obviously – Cloud “Driverless cars become real, and you can have more natural conversations with services like Alexa.” 2020 accelerated the shift to a digital world, and the “cloud everywhere” trend has enabled this. Dr Vogels is optimistic that this trend will soon expand to countries such as Indonesia, Philippines, Thailand, Vietnam, Kenya, Nigeria, and South Africa —to lead the way. On the other hand, in the US, only 47% of small and medium businesses have their own website, and Dr Vogels expects this number to grow in 2021. Today, cloud-based applications are leveraged on parts on earth. Be it the shipping industry or the aircraft management or even for driverless cars, the cloud is everywhere. Now, the next frontier, according to the CTO, would be to tap into “space-real estate.” AWS has a presence in regions that enabled cloud technologies to be closer than ever to customers across the world. For instance, AWS Snowball is helping customers to gather petabytes of data from the slopes of volcanoes in Hawaii and research centres in Antarctica. While AWS Outposts extends cloud infrastructure and tools into offices, AWS Local Zones takes infrastructure close to customers and helps those in urban areas to rapidly shrink their cumbersome datacenters. AWS By the numbers: 24 regions with 77 availability zones.Six announced regions.18 announced availability zones.More than 220 edge network locations (eight new countries in 2020).108 AWS direct connect locations Key Takeaways “Cloud is everywhere” trend will have a wider reach across the globe in 2021.Space will be the area where we see some of the greatest advancements when it comes to cloud technologies. By moving towards the edge, we will see in the coming year acceleration of the adoption of ML models across industries and government.New online payment platforms are built in the cloud, and whose underlying encryption and ledger systems — like blockchain — are cloud-based as well.Small and micro businesses in South East Asia, which account for 99% of all businesses, will see an increased online penetration. AWS has a great presence in these regions.2021 will be the year the quantum computer starts to bloom. Check the full list here.","excerpt":"“Rather than slowing us down, 2020 accelerated our shift to a digital world. From my vantage point, 2021 will be a launchpad for all kinds of change.” Dr Werner Vogels, VP & CTO, Amazon. Amazon has seen a tremendous growth with its cloud operations fueled by innovation in the machine learning and the chip department. […]","categories":["AI Trends"],"tags":["Amazon","AWS"],"author_name":"Ram Sagar","publish_date":"2020-12-23T13:00:00","publication_year":"2020","word_count":937,"keywords":["Go","machine learning","AWS","AI","ML","Amazon","Git","RAG","analytics","edge computing","R"],"extracted_tech_keywords":["AI","machine learning","ML","analytics","RAG","AWS","edge computing","R","Go","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/amazon-cto-trends-2021\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10095554,"title":"Microsoft Releases 1.3 Bn Parameter Language Model, Outperforms LLaMa","content":"Large language models are getting smaller with great examples like LLaMa and Falcon. Now, Microsoft Research has upped the game with an even smaller model. phi-1 is a transformer based model with just 1.3 billion parameters. The research paper titled, Textbooks Are All You Need, describes that the model was trained for just four days on 8 A100s with ‘textbook quality’ dataset from the internet with 6 billion tokens, along with synthetically generated textbooks from GPT-3.5 with 1 billion tokens. Click here to read the paper. The model will be available on Hugging Face soon. Despite being small in size, phi-1 attained 50.6% on HumanEval and 55.5% on MBPP. There is another even smaller model with just 350 million parameters called phi-1-small that is trained with the same pipeline as the larger one which still achieves 45% on HumanEval. For comparison, any other model that achieves greater than 50% on HumanEval is 100 times bigger than this with a large dataset size. Ronen Eldan, one of the co-authors of the paper said that using the textbook quality training data for coding, the results were better than they had anticipated. In a discussion on HackerNews, a user explains that this would not have been possible without using high quality synthetic dataset that was produced by GPT-3.5. It is clear that training on data from GPT models improves the accuracy and efficiency of the models. Instead of increasing the size of the model, by improving the quality of the data models are performing way better. This might result in a shift of paradigm in LLM research with focusing more on the architecture and training of models. Similarly, an open source alternative to GPT-4, Orca, with just 13 billion parameters was also trained on data by GPT-4 and was able to outperform OpenAI’s offering on several benchmarks. On the other hand, a recent paper – The Curse of Recursion – says that training on other LLMs data actually reduces the quality of output of the new model as it results in data poisoning. This is also called the false promise of imitating proprietary LLMs, as the model also inherits flaws from GPT based models.","excerpt":"One of the co-authors of the paper said that using the textbook quality training data, the results were better than they had anticipated.","categories":["AI News"],"tags":["AI Research","ChatGPT","Generative AI","Microsoft","Open Source AI","Open Source LLM","OpenAI"],"author_name":"Mohit Pandey","publish_date":"2023-06-21T17:25:59","publication_year":"2023","word_count":361,"keywords":["ChatGPT","Hugging Face","TPU","synthetic data","OpenAI","AI","AWS","A100","AI Research","Open Source AI","Open Source LLM","GPT","Generative AI","R","Microsoft","llm_models:GPT"],"extracted_tech_keywords":["AI","OpenAI","Hugging Face","AWS","TPU","R","GPT","synthetic data","A100","llm_models:GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/microsoft-releases-1-3-bn-parameter-language-model-outperforms-llama\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10049342,"title":"Facebook’s Content Moderators A Miss?","content":"In January this year, a Facebook investigator revealed that a Mexican drug cartel was using the platform to recruit, train and pay hitmen. Although this was strictly against Facebook’s policies, the tech giant didn’t do much against it. The Mexican cartel posted on Facebook and its photo-sharing social media platform, Instagram. The Wall Street Journal reviewed Facebook’s internal documents and figured that the company’s employees have been raising alarms about the menacing usage of the platform in some developing countries. Yet, despite the user base of Facebook in these developing countries being huge and constantly expanding, the company did not do much to try and stop it. Investigation of posts on Facebook and Instagram and private messages on these platforms revealed that the platforms were actively used to recruit and hire teenagers from developing countries for hit-man training camps. Facebook had multiple pages under the name of CJNC (short for Cartel Jalisco Nueva Generation) with photos of guns and crime scenes. According to the company’s policies, the accounts should have been automatically removed as they were internally labelled as ‘Dangerous.’ The pages were active for at least five months before they were taken down. Rising concerns The Mexican drug cartel instance is just one of the many harmful programmes that Facebook is breeding on its platform. Internal documents further revealed employees have been flagging the presence of Middle East-based human traffickers using the platform, alluring women into abusive employment situations. Women are promised work and instead treated either like slaves or forced into sex work. Similarly, armed groups in Ethiopia have been leveraging Facebook to provoke violence against the country’s ethnic minorities. Furthermore, Facebook employees’ have been warning about people from across the world using the social media platform for pornography, selling organs in the black market, and the government’s action against political dissent. The illicit use of the platform is mostly because, in some of its operational countries, only a few or no users speak languages and dialects that can be interpreted to identify harmful or dangerous content. The internal documents reveal that while Facebook removes some of these pages from its platform, others operate openly. The social media giant has taken down the offensive posts, but it is yet to fix the systems that could prevent the formation of similar pages in the future. Facebook’s priorities continue to be to retain its users and help business partners. The documents reveal, sometimes, Facebook is busy pleasing authoritarian governments when it needs to operate within its borders. Content moderators: A miss Facebook has hundreds of millions more users in developing countries than the USA. Ninety per cent of its monthly users reside outside the USA and Canada. The scenario is similar in Europe, where its user growth is stalled. In most developing countries, Facebook is used as the primary online channel for communication and source of news. Betting on this growth in developing countries, the social media giant plans to introduce technologies such as satellite internet and expanded WiFi. Facebook claims to have deployed a strategy that includes employing global teams who cover more than 50 languages, investing in educational resources and partnering with local experts and third-party fact-checkers to keep its platform safe. But that isn’t enough. Facebook does not employ people who can speak the relevant languages to monitor situations like these. Additionally, for some of these languages, Facebook has failed to build automated systems or classifiers that could spot abuses. The AI systems that form the backbone of Facebook’s enforcement do not support most of the languages that are used on the platform. To add to it, Facebook does not publish its community standards in all the languages used in its platform. Thus, users fail to realise or even learn about the rules they are supposed to follow, leading to its abusive usage in Ethiopia. Similarly, Facebook’s content reviewers working in the Arabic-speaking countries speak Moroccan Arabic, often being unable to point out abusive or violent content published in other dialects; or end up pulling down inoffensive content. In addition, Facebook’s algorithms responsible for enforcement are also incapable of handling different dialects. Closer to home, India, which houses more than 300 million Facebook users, has been facing similar nightmares. Its company researchers set up a test account as an Indian female user, and things went south. The user feed was bombarded with polarised nationalist content, misinformation and violence. Summing up Earlier last week, Facebook said that it was increasing the review capacity of its content moderation systems to prevent the posting of harmful content by expanding it to various Ethiopian languages. It further claims to have deployed a dedicated team to reduce the risks in Ethiopia. After Apple’s warning of removing Facebook and its other apps from the App Store, Facebook has sped things up. However, its focus continues to be the developed and English-speaking countries. Former Vice President at Facebook Brian Boland calls this callous behaviour of the tech giant – ‘The cost of doing business.’ Facebook uses the developing and often poorer countries to grow its user base, while it focuses on the safety of its rich markets and ones with powerful governments.","excerpt":"An investigative report suggests that Facebook has been used to recruit, train and pay hit-men by a Mexican drug cartel.","categories":["Global Tech"],"tags":["Facebook","Mark Zuckerberg"],"author_name":"Debolina Biswas","publish_date":"2021-09-23T10:00:00","publication_year":"2021","word_count":858,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","RAG","Mark Zuckerberg","GAN","Aim","Facebook","llm_models:Bard","R"],"extracted_tech_keywords":["AI","Aim","RAG","R","Go","GAN","llm_models:Bard","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/facebooks-content-moderators-a-miss\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":61755,"title":"IIT Bombay Researcher Uses ML To Find Best Lockdown Policy","content":"A team of researchers from IIT Bombay and Microsoft Research have used reinforcement learning to find approaches that would help figure out the right policies during lockdowns. The researchers have published their work in a paper titled ’ Optimizing Lockdown Policies for Epidemic Control using Reinforcement Learning’. Reinforcement learning is a branch of AI that deals with teaching machines how to teach themselves. Just like a person learns to ride a bicycle by falling, the reinforcement learning algorithm contains agents that learn from rewards. These rewards are a bunch of mathematical expressions that are designed to make the RL agent do the most desirable task in as many steps as possible. Overview Of The Algorithm For finding optimal lockdown policies, the authors have used Deep Q Network, a reinforcement learning (RL) algorithm that runs a large number of simulations of the spread of the disease, while attempting to find the optimal policy for lockdowns. The goal here, wrote the authors, is to quantify the cost of each outcome of the simulation. The following features were considered to train the RL algorithm: Population of own node,Fraction infected (symptomatic) in own node,Fraction infected (symptomatic) in overall population,Fraction of population recovered in own node,Fraction dead in own node,Potential infectors from rest of population,Fraction increase in the symptomatic population in their own node in the last few days, and overall increase as well. The collected memory is discarded after every simulation episode. The terminal rewards R are computed as soon as the number of active infections in the whole network goes to 0 and calculated as follows: The training was carried out after each simulation for five epochs using stochastic gradient descent SGD optimizer in Keras with a learning rate LR of 0.001 and a momentum of 0.8. The algorithm then proposes policies, as a function of parameters such as infectiousness of the disease, its gestation period, the duration of symptoms and the probability of it being fatal. Along with this, the characteristics of the population, such as density, was also considered. Nature and the execution of lockdowns being imperfect, the authors have tried to tap into realistic scenarios to offer the best policy possible. The results of the experiments show that the policy obtained using reinforcement learning is a viable quantitative approach towards lockdowns. Key Takeaways The above illustration is a comparison of infection rates across several policies, where it can be observed that the reinforcement learning approach has the smallest peak that still has smooth evolution and the policies with 5% & 10% lockdowns have lower peaks, and the plot also takes a turn if the lockdowns are ended prematurely and indicates a generation of the second wave of infections. The objective of this work can be summarized as follows: To compute the optimal lockdown\/release policy using reinforcement learning for each node in a network, given disease characteristics and network properties. The user can utilize this approach without any optimization knowledge. The recommended policies are realistic and consider both health and economic costs, with the weight on each factor specified by the user’s preferences. The same algorithm can be used to compute the policy for any changes in network data, disease parameters, and cost definitions. Only the relevant input values need to be updated by the user. This approach also caters to explainability in ML models as the users can tear down the decisions to its core reason. In a way, alleviating the black box nature of ML algorithms. Know more about this work here.","excerpt":"A team of researchers from IIT Bombay and Microsoft Research have used reinforcement learning to find approaches that would help figure out the right policies during lockdowns. The researchers have published their work in a paper titled ’ Optimizing Lockdown Policies for Epidemic Control using Reinforcement Learning’.  Reinforcement learning is a branch of AI that […]","categories":["Deep Tech"],"tags":["covid-19","IIT Bombay","LOCKDOWN","policy gradient","Quantum Computer","Reinforcement Learning"],"author_name":"Ram Sagar","publish_date":"2020-04-15T14:00:52","publication_year":"2020","word_count":583,"keywords":["Go","IIT Bombay","Reinforcement Learning","covid-19","AI","Keras","programming_languages:R","ML","programming_languages:Go","Quantum Computer","LOCKDOWN","policy gradient","R"],"extracted_tech_keywords":["AI","ML","Keras","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/iit-bombay-lockdown-policy\/","complexity_score":3,"technical_depth":7,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10011302,"title":"Facebook&#8217;s New AI System Can Pass Multiple-Choice Intelligence Tests","content":"Recently, a team of researchers from Facebook AI and Tel Aviv University proposed an AI system that solves the multiple-choice intelligence test, Raven’s Progressive Matrices. The proposed AI system is a neural network model that combines multiple advances in generative models, including employing multiple pathways through the same network. Raven’s Progressive Matrices, also known as Raven’s Matrices, are multiple-choice intelligence tests. The test is used to measure abstract reasoning and is regarded as a non-verbal estimate of fluid intelligence. In this test, a person tries to finish the missing location in a 3X3 grid of abstract images. According to the researchers, there have been various similar researches, where the main focus entirely on choosing the right answer out of the various choices. However, in this research, the researchers focussed on generating a correct answer given the grid, without seeing the choices. Behind the Model As mentioned above, the neural network model is a combination of various advances in generative models, including employing multiple pathways within the same network. It uses the ‘reparameterisation’ trick along two pathways in order to make their encoding compatible, which are a dynamic application of variational losses and a complex perceptual loss which is linked with a selective backpropagation procedure. In this research, the researchers considered the task of generating a correct answer to a Raven Progressive Matrix (RPM) type of intelligence test. In this test, each query includes eight images that are placed on a 3X3 grid size. The task of this test is to create the missing 9th image, such that it matches the patterns of the rows and columns of the grid. The neural net model recognises the correct answer out of the eight possible choices by encoding each image and aggregating these encodings along rows and columns. The architecture of the AI model is mainly composed of three different pathways: Reconstruction: The reconstruction pathway provides supervision that is more accessible to the network when starting to train.Recognition: The recognition pathway shapes the representation in a way that makes the semantic information more explicit.Generation: The generation pathway relies on the embedding of the visual representation from the first task, and on the semantic embedding obtained with the assistance of the second, and maps the semantic representation of a given query to an image. How It Works There are four key components in this method. They are: An encoder (E)A  generator (G) that is trained together as a variational autoencoder (VAE) on the images.A Context Embedding Network (CEN), which encodes the context images and produces the embedding for the generated answerFinally, a discriminator (D), which provides an adversarial training signal for the generator The VAE pathway includes the encoder and the generator and it autoencoders the choice images as one image at a time. The CEN is composed of multiple sub-modules, which are trained together to provide input to the generator. Wrapping Up The researchers stated, “In problems in which the solution space is complex enough, the ability to generate a correct answer is the ultimate test of understanding the question since one cannot extract hints from any of the potential answers. Our work has been the first to address this task in the context of RPMs.” The success behind the model includes a number of crucial technologies, including applying the reparameterisation trick selectively and multiple times, reusing the same networks for encoding and to provide a loss signal, selective backpropagation, and an adaptive variational loss. According to them, the machine learning algorithm is not only capable of generating a set of possible answers but also to be competitive to the state-of-the-art methods in multiple-choice tests. They also claimed that the neural net model could be used to develop an automatic tutoring system that adjusts to the proficiencies of each student. Read the paper here.","excerpt":"Recently, a team of researchers from Facebook AI and Tel Aviv University proposed an AI system that solves the multiple-choice intelligence test, Raven’s Progressive Matrices. The proposed AI system is a neural network model that combines multiple advances in generative models, including employing multiple pathways through the same network.  Raven’s Progressive Matrices, also known as […]","categories":["AI Trends"],"tags":["Facebook AI"],"author_name":"Ambika Choudhury","publish_date":"2020-11-07T11:00:00","publication_year":"2020","word_count":631,"keywords":["Go","machine learning","Facebook AI","AI","neural network","programming_languages:R","programming_languages:Go","Aim","VAE","R"],"extracted_tech_keywords":["AI","machine learning","neural network","Aim","R","Go","VAE","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/facebooks-new-ai-system-can-pass-multiple-choice-intelligence-tests\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":41438,"title":"Shine.com Strengthens Its Leadership Ranks, Appoints Madhukar Kumar As Chief Analytics Officer","content":"Image Source: Shine.com Shine.com, India’s 2nd largest job portal is set to further strengthen its top leadership ranks after it hired top talent for it’s Analytics division. To this end, the company has announced the appointment of analytics veteran Madhukar Kumar as its Chief Analytics Officer. This hiring is in line with Shine.com’s current focus on ramping up its senior management team by hiring industry experts across technology, product, servicing and analytics domain. With over 13 years of experience across data sciences, predictive analytics, Artificial intelligence (AI), Machine learning (ML), deep learning and consulting domains, Kumar plans to build a robust data science team to enhance Shine.com’s candidate and recruiter experience. Madhukar Kumar, Chief Analytics Officer, Shine.com Kumar has made a significant mark on the global business landscape as a Data Science Leader & AI Evangelist. Kumar has been conferred the “Lifetime distinguished membership” of the Leaders Excellence at Harvard Square group and his “Drone Image Analytics solution” was awarded the Nasscom AI game changer award in 2018. He had also been rated as one of the Top 4 data scientists from India by Swami Associates, a leading Australian advisory firm in 2015. With the rapid growth of its active candidate database of almost 40%, Shine has grown faster than the industry standards over the past two years. To continue to fuel this growth and further strengthen its product by leveraging its growing database, the company has been on a hiring spree and has been recruiting top industry talent over the past two years. Commenting on Kumar’s appointment, Zairus Master, CEO, Shine.com, said, “At Shine.com, we plan to lay greater focus on data analytics to augment our product in 2019-20. Strengthening our AI-led recruiter and candidate experience will be one of our major goals this year. To that end, we aim to build our data science team by bringing on skilled individuals from the industry. This will be the key focus area for our newly appointed CAO, Madhukar Kumar. Banking on his rich industry experience and expertise, we will definitely achieve our goals of enhancing our product and driving superior user experience. Needless to say, we are pleased to have an industry veteran such as Madhukar on our team as we continue to grow and innovate the next level of recruitment products.” “I am thrilled to be associated with Shine.com, which has been making waves in the recruitment sector on the back of constant innovation. Today, businesses sit on huge pools of data which, if leveraged in the right manner, can create a competitive advantage for them. With this idea, I will be working on building Shine.com’s data science team to leverage its rich repository of candidate and recruiter data (more than 30 million resumes). I believe that by using this data, we can accelerate the growth of Shine.com further and deliver a more enriching experience to all our stakeholders,” Kumar said. While building Shine.com’s data science team, Kumar will focus on AI, Deep learning, machine learning, text analytics & speech analytics. Currently, Shine.com is betting big on AI and ML, and it can gain massively by hiring candidates that are skilled in these domains. With a greater focus on hiring the right talent, Shine.com continues to close the gap with its competition rapidly.","excerpt":"Shine.com, India’s 2nd largest job portal is set to further strengthen its top leadership ranks after it hired top talent for it’s Analytics division. To this end, the company has announced the appointment of analytics veteran Madhukar Kumar as its Chief Analytics Officer. This hiring is in line with Shine.com’s current focus on ramping up […]","categories":["AI News"],"tags":[],"author_name":"AIM Media House","publish_date":"2019-06-28T04:23:50","publication_year":"2019","word_count":544,"keywords":["data science","machine learning","artificial intelligence","AI","ML","RAG","Aim","deep learning","analytics","predictive analytics"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","data science","analytics","Aim","RAG","predictive analytics"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/shine-com-strengthens-its-leadership-ranks-appoints-madhukar-kumar-as-chief-analytics-officer\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":42463,"title":"Guide To Build Your First Convolutional Neural Network with PyTorch","content":"When we consider Machine Learning as a professional skill, it ultimately comes down to the choice of using the right algorithms, packages and libraries and, there are many. Speaking extensively of packages, there are many popular choices to choose from – TensorFlow, Theano, Keras, PyTorch etc. We will focus on PyTorch for now. Developed by the brains of Facebook, PyTorch has a lot to offer in the Machine Learning space. In this tutorial, we will give a hands-on walkthrough on how to build a simple Convolutional Neural Network with PyTorch. Pre-Requisites This short tutorial is intended for beginners who possess a basic understanding of the working of Convolutional Neural Networks and want to dip their hands in the code jar with PyTorch library. Read some of our previous articles on Convolutional Neural Networks to have a good understanding before we dive into CNN with PyTorch. Follow the links below: Learn Image Classification Using CNN In Keras With Code Overview Of Convolutional Neural Network In Image Classification Introductory Guide To PyTorch Using A Linear Regression Problem Why PyTorch ? So why do we need to learn PyTorch too in a world that abounds with DL frameworks. PyTorch is a scientific computing package developed by Facebook for Deep Learning. It is built with dynamic computation which allows users to manipulate the computation graphs dynamically, a standout feature that other deep learning packages lack. PyTorch is also fast and has lots of easy to use API’s. PyTorch also has a large community support which makes it a library of choice for many Machine Learning Developers. Although other packages, especially Tensorflow dominates in the production space, PyTorch has it large user space in researches which is all the more reason to learn to use it. Let’s get started in and build a simple Convolutional Neural Network. Importing Torch Libraries import torch import torch.nn as nn import torch.nn.functional as F Creating A Custom CNN First, we will define a class that inherits the nn.Module Class in Pytorch. Creating a class to customise the neural networks is a great approach as it gives more room for flexibility in coding making it easier to implement multiple networks. class CNN(nn.Module): def __init__(self): super(CNN, self).__init__() self.conv1 = nn.Conv2d(in_channels = 3,out_channels = 16, 5) self.conv2 = nn.Conv2d(in_channels = 16, out_channels = 32, 5) self.fc1 = nn.Linear(in_features = 32 * 5 * 5, out_features = 150) self.fc2 = nn.Linear(in_features = 150,out_features =  90) self.fc3 = nn.Linear(in_features = 90,out_features = 10) def forward(self, x): x = F.max_pool2d(F.relu(self.conv1(x)), (2,2)) x = F.max_pool2d(F.relu(self.conv2(x)), 2 ) x = x.view(-1, self.num_flat_features(x)) x = F.relu(self.fc1(x)) x = F.relu(self.fc2(x)) x = self.fc3(x) 16     return x def num_flat_features(self, x): size = x.size()[1:] num_features = 1 for s in size: num_features *= s return num_features Let’s follow the above code block line by line. 1. Create a custom class called CNN that inherits the nn.Module Class from PyTorch Library. 2. Define the class initializer method. 3. Used in Inheritance, the line delegates object calls to the parent or child class of CNN. 4. Defining the first convolutional layer. Here the in_channels parameter corresponds to the number of channels in the input image. For grayscale images, the input channels are always 1 and for coloured images, the channels are 3. The number of out_channels can be of a random choice. There is no set rule on the choice of out-channels, the greater the out_channels the more features that are extracted from the images, however, larger numbers will use more resources and also increases the chance of overfitting for smaller datasets. The kernel size is the size of the convolution matrix. Setting it to 3 will use a 3×3 matrix for convolution. 5. Adding a second layer of convolution to the network.A point to be noted is that the second convolutional layer should have the same number of in_channels as the number of out_channels coming from the previous layer. 6. Creating a fully connected network. Here, we have 3 layers the first one being an input layer (line 6) connecting to the convolution layer, the second one being a hidden layer (line 7) and the third, an output layer (line 8). The number of out-features in the output layer corresponds to the number of classes or categories of the images that we need to classify. 9. Defining the forward method which will pass and forward the inputs (images) through all the layers in the network. 10. Performs max pooling on the image that has passed through the first layer of convolution layer activated with the Relu (Rectified Linear Unit ) activation function. The pooling is performed with a 2×2 matrix for which the shape has been passed as a tuple argument. 11. Saves a max pooled image data which had passed through the second convolution layer activated with the Relu (Rectified Linear Unit ) activation function. The pooling is performed with a 2×2 matrix for which this time the shape has been passed as an integer argument. 12. Flattening and reshaping the pooled matrix using the view method and the num_flat_features method. 13. Feeding the flattened matrix to the fully connected layers. The input layer (Line 13), hidden layer (Line 14) and Output layer (Line 15). Defining a method to flatten the extracted features after pooling. Initialising the CNN We now have a complete template to build as many CNNs as we need. Next, we will create an object to initialize the CNN. #Initializing the Object for CNN cnn = CNN() The above line will create an object for our custom CNN that we built Let’s have a look at the summary of the CNN that we built. #Printing the summary of the CNN print(cnn) This above line of code will print a description of the CNN. Output: After building the CNN we can use it to train and predict image labels or classes which will learn about in an upcoming tutorial. Until then Happy coding !!","excerpt":"When we consider Machine Learning as a professional skill, it ultimately comes down to the choice of using the right algorithms, packages and libraries and, there are many. Speaking extensively of packages, there are many popular choices to choose from – TensorFlow, Theano, Keras, PyTorch etc. We will focus on PyTorch for now. Developed by […]","categories":["Deep Tech"],"tags":["cnn","Convolution Neural Network","Convolutional Neural Network","Pytorch"],"author_name":"Amal Nair","publish_date":"2019-07-15T14:00:45","publication_year":"2019","word_count":985,"keywords":["Pytorch","machine learning","Keras","TPU","AI","neural network","PyTorch","cnn","Ray","Convolution Neural Network","deep learning","Convolutional Neural Network","TensorFlow","R"],"extracted_tech_keywords":["AI","machine learning","deep learning","neural network","Ray","TensorFlow","PyTorch","Keras","TPU","R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/guide-to-build-your-first-convolutional-neural-network-with-pytorch\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10013857,"title":"How To Make Algorithms Really Work For Clinical Trials","content":"“Researchers propose a framework for regulating algorithms that will ensure world-class clinical performance and build trust.” Machine learning models are considered to gradually take over most of the mundane tasks of clinicians and other medical professionals. In a way, assisting them with insights and cutting the downtime. These models are already getting good at helping radiologists. But how legit are these algorithms? Have they been thoroughly vetted by government bodies like the FDA? Researchers at Stanford noted a few reasons as to why algorithms are readily accepted into clinical setups: Mix up of the diagnostic task with the diagnostic algorithmNot so thorough treatment of the diagnostic task definition No mechanism to directly compare similar algorithms Insufficient characterisation of safety and performance elementsLack of resources to evaluate and inherent conflicts of interest To address these issues, a team from Stanford Institute for Human-Centered Artificial Intelligence (HAI) in their new paper titled “Regulatory Frameworks for Development and Evaluation of Artificial Intelligence”, propose a framework for FDA regulation of AI-based diagnostic imaging software. Challenges Of AI Based Diagnostics (Source: Larson et al.,) Though in the US, the FDA has a framework in place for evaluating software developed for medical applications, the authors state that these checkpoints cannot be considered for AI-based tools. According to the head of the FDA, AI algorithms continually learn from the medical images they review. So, the traditional approach to reviewing and approving SaMD (software-as-a-medical-device) upgrades may not apply. There is a need to implement a protocol for regulating the entire lifecycle of AI-based algorithms. Explaining this, one of the authors cited the example of the early days of the coronavirus pandemic when medical professionals worldwide proposed multiple different scoring systems to categorise lung scans of patients. Though the scoring systems differ, they are corrected over time, and this type of adjustment is common in clinical trials. However, in the case of AI, the authors wrote, that would have been problematic as hard-coded proprietary diagnostic task definitions will make it difficult to compare the performance of algorithms. The authors propose that algorithm manufacturers should be required to go through four phases of development and evaluation similar to what the FDA does. They suggest testing for: Feasibility of algorithms on a small test setCapability in a controlled environment simulating real-world conditionsEffectiveness in a real clinic) Durability including monitoring and improvement over time The above flow chart illustrates the proposed linkage of the evaluation of diagnostic algorithm performance from the defined task to local site algorithm performance. Algorithms are developed according to a defined diagnostic task standard. Performance is compared with other algorithms in a controlled environment, which becomes the internal benchmark for general real-world performance and local site performance, which in turn becomes the benchmark for ongoing monitoring. Key Takeaways Strategies outlined by regulatory bodies address many key aspects to help ensure the safety, effectiveness, and performance of SaMD applications, but a number of gaps remain.Appropriate evaluation and improvement methods should be incorporated into phases of development, similar to what the FDA does. Algorithms should be thoroughly tested and refined before being deployed in the clinical environment, just as it is expected from other medical devices.Regulatory frameworks should strive to establish conditions that set up a race to the top for consistent excellent algorithm performance at each installed site. One of the authors likened their proposed regulatory framework to the safety requirements for an aircraft. “… if they stop working, they need to fail in a way that’s not going to hurt people.” In their paper, the authors have listed 12 measures of performance that should be applied to diagnostic algorithms. To know more, download the original paper here.","excerpt":"“Researchers propose a framework for regulating algorithms that will ensure world-class clinical performance and build trust.” Machine learning models are considered to gradually take over most of the mundane tasks of clinicians and other medical professionals. In a way, assisting them with insights and cutting the downtime. These models are already getting good at helping […]","categories":["Deep Tech"],"tags":["diagnosis","how does artificial intelligence work","stanford university ai"],"author_name":"Ram Sagar","publish_date":"2020-12-10T14:00:00","publication_year":"2020","word_count":609,"keywords":["Go","diagnosis","artificial intelligence","machine learning","programming_languages:R","AI","programming_languages:Go","Git","stanford university ai","how does artificial intelligence work","Rust","R","programming_languages:Rust"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","R","Go","Rust","Git","programming_languages:R","programming_languages:Go","programming_languages:Rust"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/algorithms-for-clincial-trials-medical-diagnosis\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":34506,"title":"This AI-Powered Dating App Betterhalf.AI Helps Users Find The Perfect Life Partner","content":"In this technology-driven era, human lives are becoming easier. Even dating and finding someone to romantically connect with has become fairly easy with numerous dating apps and platforms. However, there is still a void that needs to be filled. With matchmaking being reduced to a few swipes, there’s something getting lost in translation for men and women aged 25-32 years, looking to seriously date with an intent to settle down. And compatibility plays a vital role. When two people match through a dating app, they by themselves need to figure out whether they are compatible. In order to fill this void in the dating space, two MIT alumni, Pawan Gupta and Rahul Namdev started Betterhalf.AI in 2016. Betterhalf.AI is India’s first “true compatibility” partner search product that uses artificial intelligence for professionals to find each other through compatibility scores based on multiple relationship dimensions and their interactions on the product. Betterhalf.AI Builds largest AI-based Relationship Engine The story of Betterhalf.AI dates back to the college days of Rahul and Pawan at MIT. As roommates, the duo was searching for partners on different matrimonial and dating sites. “We were observing that people in similar cohorts as we were having a difficult time navigating the partner-search journey. Everyone around us complained and at some point, we were not okay with this,” said Pawan. It was a turning point in their lives and scratched that part of their mind. They set out decided to build a platform that goes one step ahead to make predictive match-making a reality. According to Rahul and Pawan, there is a significant factor that current match-making products have not been able to incorporate – the previous relationships’ data for thousands of couples. And Betterhalf.AI make use of this data to bring precision to partner search. Today, Betterhalf.AI is on a path to build the largest AI-based relationship engine that can suggest matches taking into account both extensive couples’ relationship data and the users’ comprehensive personality profiles. As the users offer feedback through private ratings, their matches become more compatible over time. Betterhalf.AI Drives Data-driven Matchmaking There are players in dating or matchmaking space that use a cluttered network of parents and users, rudimentary matching based on age, height, caste topped with a bad user interface. However, Betterhalf.AI provides a combination of a targeted subset of matches with a quick turnaround time to find compatible partners. Currently, Betterhalf.AI has more than 17,000 users from 4,000 unique companies including Google, Facebook, Amazon, LinkedIn, Adobe, and Accenture. Also, 30% of their users are entrepreneurs, fashion designers, scientists and bankers. The profiles are authenticated through six levels of verification that includes LinkedIn, Facebook, personal email, phone number, work email, and a Government ID. Talking about the compatibility score, true compatibility scores are calculated based on six-relationship dimensions:  emotional, social, intellectual, relationship, physical, and moral values. With such tremendous popularity in the dating space, the company at present is aiming for a one-million user base in the next two years. “At Betterhalf.AI, we aspire to transform uncertain partner search journey to certain, timely and delightful for 500M people globally through an AI-based partner prediction engine. The platform’s AI engine starts learning about a user’s personality as soon as the user starts the on-boarding process,” said Pawan. To use the platform, first, the users need to complete the registration and fill information on different dimensions. Once that is done, users see matches with overall compatibility percentages. Also, users can send a connection request to matches and can chat with the person once requests are accepted. In addition to the authentication programs, private ratings and feedback by users help the platform filter out non-serious and creepy daters off. Use Of AI in the Dating App Registration During the registration process, the platform gathers users’ personality in six different relationship personality dimensions — emotional, social, intellectual, physical, relationship and values by asking a series of sixteen Likert-type questions. While it is able to estimate one’s initial personality and background information through these questions with reliable accuracy, to begin with, the platform uses in-product gamification, pre-match, and post-match activities of the user\/feedback about the users to get more information. Pre-Chat\/Conversation At this phase, while a user is interacting with the platform, it captures his\/her behavioural information such as click-map, scroll-map, time spent on different sections of their matches’ profile etc. in order learn more about the user. For example, a user has visited 10 matches and 5 have mentioned that they like to travel. Now, if the user spends more time with these profiles then the system learns that this particular user is interested in matches who actually like travelling. Product Gamification Through product gamification, we capture more personality information from users. In a gamified way, we ask more personality based questions from users to learn more about their personality. As we get this additional data over time as users spend more time on the product, it helps us rectify any personal biases cropping in users’ mind due to bad experience which user might have faced that day. Asking questions over a period of time also helps us capture the information in different states of mind and thereby helping us evaluate the exact personality of a user. We use an AI-based algorithm to probabilistically update and correct initial personality representation. Post Chat Betterhalf.AI also take timely private star rating feedback about a user from their matches after they are done interacting\/chatting. This process not only helps in finding the authenticity of the profile but also helps in discovering whether the person is up for marriage. It also covers topic like timely response and compatibility with a match to various other likes\/dislikes. That is not all, the platform again applies AI to refine user’s profile and personality representation based on the feedback from other users. This helps Betterhalf.AI to create a more accurate version of “user” and “their personality” that helps in giving matches that are truly compatible with a user. Removal of any outliers or biased data using Machine Learning Once all that is done, the final step is to remove\/rectify any outlier data and that is done by using an ML algorithm. Suppose, if we talk about short-tempered dimension, 99.9% user rate themselves between 1 to 4 and if there is a user who has marked herself\/himself as 7 then Betterhalf applies ML-based rectification to improve this data. Algorithmic Matchmaking With Betterhalf.AI Let’s have a clear view with an example; Amrita (a potential user), who grew up in Bengaluru is a 27-year-old Marketing Manager. The current tech at scale will create a data barrier and personalise her partner search  and provide results such as Males in the age group 28-32 Salary Range: ₹18-20 lakh per annum Location: Bengaluru Having a similar sense of humour and liking for music The time it will take to find a match: 4-6 months. No. of relevant profiles a user will get: 60 “In our test experiments, more than 40% of the users changed their age and height preferences for their partner based on the product recommendation. Also, with the help of compatibility related info on married couples, an additional 11% users sent connect requests to their potential partners,” Pawan concluded.","excerpt":"In this technology-driven era, human lives are becoming easier. Even dating and finding someone to romantically connect with has become fairly easy with numerous dating apps and platforms. However, there is still a void that needs to be filled. With matchmaking being reduced to a few swipes, there’s something getting lost in translation for men […]","categories":["Deep Tech"],"tags":["dating","Startups"],"author_name":"Harshajit Sarmah","publish_date":"2019-02-06T04:27:01","publication_year":"2019","word_count":1200,"keywords":["Go","dating","machine learning","artificial intelligence","programming_languages:R","AI","data-driven","ML","Aim","ViT","Startups","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","Aim","R","Go","ViT","data-driven","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/this-ai-powered-dating-app-betterhalf-ai-helps-users-find-the-perfect-life-partner\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10042360,"title":"A Comprehensive Guide To Redis For Data Scientists","content":"An online database is a database that can be easily accessed locally by using a local network or the internet. Instead of storing data directly to the desktop and its attached storage, online databases are hosted on websites and through the cloud model & hence providing Software as a Service(SaaS) on the web browser. These web-based applications can be free or may require payment, usually as monthly and yearly subscriptions. You pay for what you use; therefore, the amount of server space required can be modified to one’s needs accordingly. The hassle of installing an entire software gets forked out, too, as everything is maintained on the cloud. Information is accessible from almost any device at any time. As everything is stored in a cloud, that means not being stuck to just one computer. As long as access is granted, you can technically get a hold of the data from just any compatible device. Such databases also come with in-house technical support, serving 24×7 & 365 days. Some examples of such databases are the Oracle Database, IBM Db2 and the infamous Amazon DynamoDB. What is Redis? Redis, also known as Remote Dictionary Server, is a super-fast, open-source and in-memory key-value data store created to be used as a database, cache manager, message broker, and queue. All the Redis data resides in memory, contradicting other databases that store data on disk or SSDs. By eliminating the need to access disks, in-memory data stores such as Redis avoid response time delays and access any microsecond data. This also means that while Redis supports mapped key-value-based strings to store and retrieve data parallel to the data model supported in traditional kinds of databases, it also supports other complex data structures like lists, sets, etc. Some of the top features of Redis are versatile data structures, high availability, geospatial, Lua scripting, transaction management, on-disk persistence, and cluster support, making it simpler to teach with real-time internet-scale apps. Redis is a type of database that’s commonly referred to as No SQL or Non-relational. Therefore in Redis, there are no tables and no database defined way of relating data in Redis with other data. Using Redis instead of a common relational or other primarily on-disk databases, you can avoid writing unnecessary temporary data and avoid needing to scan all over and delete the temporary data, therefore ultimately improving the performance required. How Does Redis Work? Redis works by mapping keys to values with a sort of predefined data model. It also uses a method known as Sharding, through which you partition your data into different pieces. It partitions the data based on IDs embedded in the keys, based on the hash of keys, or some combination of the two. By partitioning your data, you can store and fetch the data from multiple machines, allowing a linear scaling in performance for certain problem domains. Even though Redis performs very well under the most uncertain circumstances due to its in-memory design, there are situations where one might need Redis to process more read queries than a single Redis server can handle. Therefore, Redis supports master\/slave server instance replication to support such higher rates of reading performance and handling for server failure that Redis is running on. Redis supports master\/slave server instance replication, where the slaves are connected to the master and receive an initial copy of the full database. So as the master writes the data, it is sent to all the connected slaves as well, and they are updated in real-time. With continuously updated slaves, clients can connect to any slave for reading data in case of a crash or connectivity issue with the master server. To reduce load, it can also be combined with another database. Supported Data Types Some of the supported Data Types in Redis are : Strings: Allows to operate on the whole string, parts, increment\/decrement the integers and floats Lists: Allows you to Push or Pop items from both ends, read individual and multiple items, find and remove items using their value Sets & Sorted Sets: Allows to Add, Fetch or remove individual and random items, check membership, intersect, create unions and difference Hashes: Add, Fetch or remove items or fetch the whole hash Bitmaps Hyper Logs Geospatial Indexes What Languages does Redis Support? Redis simplifies code by allowing us to write fewer lines of code to store, access and use in data for the applications created. Multiple supported data structures can be used to store data in the data store with just a few code lines and comes with options to manipulate and interact with the data. Over more than a hundred open source clients are available for developers using Redis. Languages supported include : JavaPythonPHPC,C++ and C#JavascriptNode.jsRubyRGo and many more. High Availability and Scalability of Redis Redis offers a primary replication of architecture in a single node primary or a clustered topology. This allows the user to build highly available solutions and maintain consistent performance and reliability. In addition, there are a whole bunch of options available to adjust the cluster size, either to be upscaled or scale in or out. This allows the user the flexibility to grow the cluster according to his demands and uses. Caching using Redis Redis is always a better option for implementing a highly available, in-memory cache to decrease data access latency and increase throughput. This is because Redis can serve the frequently requested data items in a matter of milliseconds. Caching is the process of storing similar copies of files in a temporary storage location to be accessed instantly. Technically, a cache is any temporary storage location for copies of files or data, but the term is often used in reference to Internet technologies. The database queries allow caching, persistent session caching, web page caching and caching of frequently requested objects such as images, files and metadata. They hence are the popular examples of caching with Redis. With the capacity to designate how long you want to keep data and which data to evict first, Redis enables a series of intelligent caching patterns. Data Expiration and Eviction using Redis Data structures in Redis can be tagged with a Time to Live, set accordingly in seconds, after which they will get removed from the database. A series of configurable eviction policies are available to choose from. Making use of Time to Live, impermanent marked data can be considered before other data that does not have Time to Live, allowing us to create a tiered hierarchy of memory objects. The least recently used or least frequently used object would make much more sense to be evicted. Geospatial Features of Redis Redis is richly loaded with geospatial index data structures and commands to make use of. These built in-memory data structures and operators can help manage real-time geospatial data at scale and speed. For example, the latitude and longitude coordinates are stored, and users can calculate the distance between the objects or query for objects within a given radius of a point. Implementing these commands return values in multiple formats such as feet, kilometres, etc. The speed of Redis allows these data points to be updated quickly. Hence, it can be implemented in ridesharing applications to connect with nearby drivers and provide real-time updates as they travel through when made proper use of. Some Use Cases of Redis As modern data-based applications require machine learning to process massive amounts of rapidly moving data quickly, Redis can be a saviour for such high-velocity processing. The ability to process, build, train and deploy machine learning models through faster processings makes Redis the ideal choice for such use cases. Redis can also be used with steaming solutions such as Apache Kafka and Amazon Kinesis to ingest and process real-time data with low latency. It can also be used for social media analytics, ad targeting, personalization and IoT. Using Redis For demonstration purposes, I have implemented the following code on the Redis website through its tutorial UI. You can implement the following code on your system by installing the Redis-CLI. You can download the CLI through the link here. Setting a key As an example, we’ll first set a server name as “victor” with the key as “Hello World”, > SET victor \"HELLO WORLD\" We will get the following output as Ok if the key is set, OK Getting the Key To retrieve the set key, we will be using the following command. > GET victor \"HELLO WORLD\" Deleting the Key To delete our created key, we can use the following command. > DEL victor (integer) 1 #key got deleted Cross validating the result > GET victor (nil) Hence we can confirm that our key is now completely removed. Setting a Key with Time to Live We can also set keys with an expiry time, the time will be set in seconds, and key will be removed from the server after is crosses the set time. > SETEX victor 40 \"I said, Hello World!\" #key set with 40 seconds as time limit OK Checking Time to Live You can also check the time remaining from the set time to expire. > TTL victor (integer) 36 Renaming our Key Keys can be renamed using the following command. > RENAME victor bar  #renaming victor as bar OK Flushing the Key Flushing everything saved so far. > flushall OK #just got flushed EndNotes Through this article, we tried to know what Redis is and what it is capable of. We also tried to explore its use cases, run basic Redis database commands, and check its functionalities. Redis offers highly performant and efficient read and writes via its optimizations. Therefore, I would recommend exploring the Redis database further and implementing it for its immense capabilities. Happy Learning! References Official Redis websiteRedis using AWSRedis Tutorial Terminal","excerpt":"An online database is a database that can be easily accessed locally by using a local network or the internet. Instead of storing data directly to the desktop and its attached storage, online databases are hosted on websites and through the cloud model & hence providing Software as a Service(SaaS) on the web browser. These […]","categories":["Deep Tech"],"tags":["data analyst vs data scientist","database management","Guide","Redis"],"author_name":"Victor Dey","publish_date":"2021-06-27T16:00:00","publication_year":"2021","word_count":1631,"keywords":["data analyst vs data scientist","machine learning","TPU","AWS","AI","database management","RAG","Python","analytics","Kafka","R","Guide","Redis"],"extracted_tech_keywords":["AI","machine learning","analytics","RAG","AWS","TPU","Kafka","Redis","Python","R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/a-comprehensive-guide-to-redis-for-data-scientists\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":1936,"title":"Study: Internet of Things Companies in India – 2017","content":"Internet of Things is on a dramatic growth path in India, and this is evident from the number of IoT companies\/ startups that have come up in last 3-4 years. And the numbers are astonishing. According to our research, there are close to 800 firms working on IoT as a technology in some shape & form. Here’s our annual study of IoT companies in India. Read IoT India Companies Study 2016 Key Trends More than 800 companies in India claim to provide IoT as an offering to their customers. This includes a large number of companies into IoT products (mostly wearable & Home automation) and a small chunk offering enterprise IoT services. This is an increase of more than 50% from last year when India housed around 500 IoT companies. The number of IoT companies in India is still very few in number, compared to the strength of IoT companies around the globe. In fact, India accounts for just 7% of global IoT companies. Company Size On an average, Indian IoT companies have 49 employees on their payroll. On a global scale, this is quite a good number, as IoT companies across the world employ an average of 37 employees. Almost 82% of IoT companies in India have less than 50 employees compared to 87% on a global level. Around 36% IoT organizations in India saw a decline in headcount last year. The decrease in headcount may not be indicative of an overall decline in growth but can be due to operational reasons including attrition etc. 10% IoT organization did not see any growth 44% organization saw a headcount growth of less than 25%. Cities Trend Being the IT hub of India, Bangalore undoubtedly continues to house the most number of IoT firms in India, i.e. almost 36%. Pune follows it at a distant 15% and Hyderabad at 10% IoT companies. Pune surprisingly has a larger share of IoT companies than other cities, owing to the city being home to some solid manufacturing base. Delhi, Mumbai & Gurgaon are far behind with their percentages of IoT companies in single digits as reflected in the graphs below. Compared to a year back, the share of IoT companies has remained almost the same, except: Chennai saw an increase in % share of IoT companies, from 7% in 2016 10% in 2017. Gurgaon saw a decline, from 8% to 6% this year.","excerpt":"Internet of Things is on a dramatic growth path in India, and this is evident from the number of IoT companies\/ startups that have come up in last 3-4 years. And the numbers are astonishing. According to our research, there are close to 800 firms working on IoT as a technology in some shape & […]","categories":["AI Features"],"tags":[],"author_name":"Дарья","publish_date":"2017-04-26T03:30:29","publication_year":"2017","word_count":399,"keywords":["Go","programming_languages:R","AI","Git","RAG","automation","Aim","GAN","R","startup"],"extracted_tech_keywords":["AI","Aim","RAG","R","Go","Git","GAN","automation","startup","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/study-internet-things-companies-india-2017\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10044317,"title":"Cloudera Data Platform — A Strong Performer In Cloud Data Warehouse Category","content":"The massive influx of data and the growing need for modern analytics call for a modern cloud data warehouse. To thrive in today’s data-driven economy, organisations need a cost-effective cloud data warehouse that is easy to deploy, handles all types of data latencies, and supports thousands of concurrent users and queries per second. Forrester Wave™ — Cloud Data Warehouse, Q1 2021 report has named Cloudera Inc — a multi-cloud data warehouse solution provider — a strong performer in the cloud data warehouse category. Competing with the top 13 vendors in the Cloud Data Warehouse market, Cloudera Data Platform (CDP) achieved a score of 5 out of 5 in eight out of 15 different criteria, including — data types, data ingestion\/loading, performance features, scalability reference, analytics capabilities, ML\/data science, and horizontal and vertical use cases. Cloudera is one of the three participating vendors that fetched the maximum score in the ML\/data science criterion. The score speaks to the company’s unique AI & ML automation capabilities that optimise ML workflows for enterprise data science teams. Cloudera Data Platform (CDP) — a unique approach to enterprise data, launched in 2019 — is a hybrid cloud and multi-cloud data management solution fully integrated with data engineering and machine learning that enables easy, fast, and secure enterprise analytics. With CDP, Cloudera provides the ability to deliver analytics as a service in any cloud environment. “Cloudera Data Platform (CDP) supports a full data lifecycle ecosystem across hybrid and multiple clouds, delivering a shared data experience. Cloudera Data Warehouse is fully integrated with CDP, enabling auto-provisioning, cloud optimisation, self-service workload management, and auto-scaling capabilities. In addition, Cloudera’s shared data experience provides consistent data security, governance, and control across all multifunction analytics and data discovery,” according to Forrester. Cloudera Data Platform Advantages Operates across all major public & private clouds.Integrates big data management & analytic experiences across the data lifecycle.Delivers security, compliance, migration & metadata management across all environments.Open source, open integrations & open to multiple data stores & compute architectures.Cloudera Machine Learning, a cloud-native service to deploy collaborative machine learning workspaces with secure, self-service access to enterprise data.Cloudera Data Hub to build custom business applications in a fast & easy way that support a diverse set of use cases with secure, self-service access to enterprise data. The Forrester report observed:”Customers like Cloudera’s business value, partner ecosystem, low cost, data governance, good automation, and solution flexibility. Top use cases include BI acceleration, customer intelligence, real-time BI, AI\/ML-based analytics, data science, data warehouse modernisation, and data services.” Cloudera Data Platform enables IT to deliver a cloud-native self-service analytic experience to BI analysts that go from zero to query in minutes. It outperforms other data warehouses on all sizes and types of data, including structured and unstructured, while scaling cost-effectively. In addition, CDP’s consistent framework secures and provides governance for all your data and metadata on private clouds, multiple public clouds, or hybrid clouds. Read the complete The Forrester Wave™: Cloud Data Warehouse, Q1 2021 report here. To learn more about Cloudera Data Platform, visit the website. About Cloudera: Cloudera is a multi-cloud data warehouse platform provider that empowers organisations to transform complex data into clear and actionable business insights. Cloudera delivers enterprise data cloud for any data, anywhere, from the edge to AI. Powered by the relentless innovation of the open-source community, Cloudera advances digital transformation for the world’s largest enterprises. Learn more at Cloudera.com.","excerpt":"Forrester evaluated 13 vendors in its Wave Q1 2021 assessment, including Alibaba, AWS, Exasol, Google, IBM, Micro Focus, Microsoft, Oracle, SAP, Snowflake, Teradata, and Yellowbrick.","categories":["IT Services"],"tags":["cloudera","data warehouse"],"author_name":"AIM Media House","publish_date":"2021-07-23T12:40:00","publication_year":"2021","word_count":566,"keywords":["big data","data science","Go","cloudera","machine learning","AI","ML","Scala","Git","analytics","R","data warehouse"],"extracted_tech_keywords":["AI","machine learning","ML","data science","analytics","R","Go","Scala","Git","big data"],"url":"https:\/\/analyticsindiamag.com\/it-services\/cloudera-data-platform-a-strong-performer-in-cloud-data-warehouse-category\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10094505,"title":"StyleDrop: Transform Designs with Google’s Personalised Text-To-Image Generator","content":"Google has introduced a new method called StyleDrop, which allows the synthesis of images in a specific style using the Muse text-image model. The technique captures the details of a custom style, including color schemes, shading, design patterns, and local and global effects, requiring only a single image as input. StyleDrop rapidly learns the new style by fine-tuning a small number of trainable network parameters. It then improves the model’s quality through iterative training, either with human feedback or automatic feedback. The process is efficient, taking less than three minutes even with human feedback, as StyleDrop requires only a few images for iterative training. According to Google, StyleDrop outperforms other methods for style transfer from text-to-image models, surpassing techniques such as Dreambooth, LoRAs, Textual Inversion in Imagen, and Stable Diffusion. The team at Google combines StyleDrop with Dreambooth to learn and create new objects in different styles. By utilising Muse, they can generate custom objects in custom styles. Google envisions StyleDrop as a versatile tool that allows designers or companies to train with their brand assets and rapidly prototype new ideas in their desired style. The project page for StyleDrop provides further information on its capabilities and applications. Google had previously introduced an AI model called Imagen, which is a text-to-image diffusion model. Unlike other AI text-to-image generators, Imagen had  specific limitations in its initial release. Users could  generate buildings with different themes or style animated creatures using Imagen AI. The model had been released through the AI Test Kitchen app, where Google tests various AI projects before public release. Imagen utilises the LAION-400M dataset for training. Google’s cautious release of Imagen and its focus on photorealistic outputs distinguished it from other models like DALL-E or Midjourney. Imagen’s unique features include City Dreamer, which allows users to construct buildings based on text descriptions, and Wobble, which generates animated creatures. While generative AI tools like this offer creative possibilities, concerns remain regarding provenance and copyright issues in content creation using AI. The use of internet-posted text to train AI text generation tools has seen less debate compared to image-based tools, while AI tools for music production are also likely to raise similar questions.","excerpt":"Google has introduced a new method called StyleDrop, which allows the synthesis of images in a specific style using the Muse text-image model.","categories":["AI News"],"tags":["AI Tool"],"author_name":"Shyam Nandan Upadhyay","publish_date":"2023-06-05T18:26:04","publication_year":"2023","word_count":362,"keywords":["Go","API","DALL-E","TPU","programming_languages:R","AI","RPA","stable diffusion","generative AI","AI Tool","R"],"extracted_tech_keywords":["AI","generative AI","TPU","R","Go","API","DALL-E","stable diffusion","RPA","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/styledrop-transform-designs-with-googles-personalised-text-to-image-generator\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10068034,"title":"A thing of pride or just vanity: Tech firms on the Nasdaq billboards","content":"Aarthi, a full stack developer from Chennai-headquartered Freshworks, a SaaS company which was recently listed on Nasdaq after a billion-dollar IPO, was in for a pleasant surprise when her face was displayed on probably one of the most globally recognised places – the Nasdaq billboard at Times Square, New York. A frontend developer at Atlan, Shivansh, expressed similar sentiments when he, along with his colleagues, featured on the same Nasdaq billboard. We're thrilled to be part of NASDAQ's closing bell ceremonies and be featured on NASDAQ's tower as #16 on the Forbes listing of the top 100 private cloud companies. pic.twitter.com\/6JhVGYDOAD— Freshworks Inc (@FreshworksInc) September 19, 2020 https:\/\/twitter.com\/AtlanHQ\/status\/1435597913342705670?s=20&t=DYEyuBVgTVBeZ3oc7bqtxA A lot of company representatives are proudly flaunting their company being displayed on the digital billboard, especially at a time when a lot of them are looking at going public. While it is the ultimate seal of validation, being shown on the 10,000 square feet space on the tower that is 120-feet tall, it makes one wonder, what qualifies a company to feature here? It turns out that one just needs a few dollars to go up the billboard. While it takes about USD 5,000-50,000 per day to advertise a billboard at Times Square, the cost of advertising on the Nasdaq billboard is approximately USD 250 per hour, a relatively smaller price to pay for the whole vanity affair. https:\/\/hi-in.facebook.com\/TredenceInc\/videos\/we-do-something-exciting-every-day-but-this-is-phenomenal-the-times-square-nasda\/571380300645687\/ The fact that it is a digital billboard helps because, unlike traditional hoardings, in digital advertising, one ad is not displayed for the entire duration. They are run by time slicing, and advertisements are sold in slots starting from 15 seconds, and Nasdaq allows a minimum buy out of two-hour slots, as this report explains. This means that multiple companies can advertise in the same space. It is beneficial for all the parties involved. Source: Times Square These advertisements help many unicorns that are rearing to go public, gain public trust, and become global brands. It is an iconic location in Times Square, which is a growing tourism and high traffic area, averaging 360,000 pedestrians per day (before the pandemic). It offers incredible visibility opportunities for businesses, making it one of the most valuable public advertising spaces in the world. Nasdaq’s billboard is particularly famous, and many companies have used it to announce their legendary moments in history.","excerpt":"The cost of advertising on the Nasdaq billboard is approximately USD 250 per hour.","categories":["IT Services"],"tags":[],"author_name":"Shraddha Goled","publish_date":"2022-05-30T16:00:00","publication_year":"2022","word_count":386,"keywords":["Go","unicorn","programming_languages:R","AI","IPO","programming_languages:Go","Git","RAG","Rust","R"],"extracted_tech_keywords":["AI","RAG","R","Go","Rust","Git","unicorn","IPO","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/a-thing-of-pride-or-just-vanity-tech-firms-on-the-nasdaq-billboards\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10126367,"title":"Intuit India&#8217;s Women in Data Science Conference Inspires 130+ Data Scientists","content":"Intuit, the global financial technology platform behind Intuit TurboTax, Credit Karma, QuickBooks, and Mailchimp, recently concluded the Women in Data Science (WiDS) conference held at their Bangalore office on June 27. The event brought together over 130 data science professionals from academia, startups, GCCs, and IT services. The insightful sessions featuring a variety of speakers, including leading data science experts, left the attendees feeling empowered and inspired. The event was dedicated to empowering women and fostering innovation in data science. WiDS, initiated by Stanford University, began as a one-day conference and has grown into a global data science movement promoting diversity and inclusivity. It  features conferences, workshops, presentations of papers, podcasts, and programs for future data scientists, reaching over 100,000 professionals annually. The conference kicked off with a welcome note by Anusha Mujumdar, an AI leader at Intuit, setting the stage for a day filled with insightful presentations and hands-on learning experiences. Keynote: The Journey of Secure Multi-Party Computation The keynote address, delivered by Arpita Patra, associate professor at the Indian Institute of Science, focused on the practical applications of Secure Multi-Party Computation (MPC). Titled ‘From theory to practice: The marvellous journey of mighty MPC’, Patra’s talk explored how this cryptographic protocol has evolved from a theoretical concept to a powerful tool in real-world scenarios. It enables multiple parties to jointly compute functions over their inputs while keeping those inputs private. Tech Talks: Advancing Voice Assistants and Algorithmic Optimisation Neela Sawant, senior applied scientist at Amazon, presented a tech talk on improving in-car voice assistants for better navigation. This presentation delved into the challenges and solutions in enhancing Automatic Speech Recognition (ASR) and natural language processing technologies specifically for automotive applications. Later in the day, Koyel Mukherjee, senior research scientist at Adobe Research, discussed ‘Optimising costs, quality and performance – one algorithm at a time’. She explored the techniques for balancing these crucial factors in algorithm development and deployment across various data science applications. Hands-On Workshop by Intuit A key highlight of the conference was the hands-on workshop led by Damitri Kundu, Senior AI\/ML Scientist at Intuit, and Sravyasri Garapati, Senior Machine Learning Engineer at Intuit. The workshop, ‘Mastering the art of automated prompt optimisation in language models’, provided participants with practical skills in fine-tuning prompts for large language models, an increasingly important skill in the era of generative AI. Paper Presentations: Emerging Research in Data Science The conference also featured several papers by students and professionals, showcasing emerging research in the field. Niharika Joshi from BMS College of Engineering presented the paper titled ‘Tackling data challenges in energy domain: Synthetic data is the answer’,  addressing the use of artificially generated data to overcome the limitations in real-world energy sector datasets. Harini Anand, a senior at PES University, discussed ‘Bitcoin price prediction using LSTM and VADER sentiment analysis’, combining advanced time-series analysis with natural language processing techniques to forecast cryptocurrency trends. Yashaswini Viswanath, senior specialist in data sciences at Mindtree, explored ‘Machine unlearning for generative AI’, a topic that addressed the challenges of selectively removing information from trained AI models. Sanjana Adapala from BITS Pilani, Hyderabad, presented on ‘Lighting estimation in virtual environments using generative models’, showcasing applications of AI in computer graphics and virtual reality. The WiDS conference in Bangalore provided a platform for knowledge sharing, networking, and inspiration, highlighting the significant contributions of women in the rapidly evolving field of data science. By bringing together experts from academia and industry, the event fostered collaboration and encouraged the next generation of data scientists to push the boundaries of innovation. Watch out for the next edition of the Women in Data Science Conference at Intuit in India, where industry pioneers share cutting-edge knowledge, empowering you to thrive in a rapidly evolving landscape. Learn more at Intuit careers.","excerpt":"The event brought together over 130 data science professionals from academia, startups, GCCs, and IT services.","categories":["AI Highlights"],"tags":["Intuit","Women in Tech"],"author_name":"Siddharth Jindal","publish_date":"2024-07-10T13:19:47","publication_year":"2024","word_count":627,"keywords":["data science","Go","machine learning","AI","sentiment analysis","Intuit","ML","RAG","Women in Tech","generative AI","machine unlearning","R"],"extracted_tech_keywords":["AI","machine learning","ML","data science","generative AI","machine unlearning","RAG","sentiment analysis","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/intuit-indias-women-in-data-science-conference-inspires-130-data-scientists\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10089496,"title":"Ernie, the Chinese Bot is Here to Take On ChatGPT","content":"Though Google is still making strides in the AI chatbot game, China’s Google, Baidu, has unveiled its much anticipated Ernie Bot. The company touts its chatbot as the strongest rival to OpenAI’s ChatGPT. But amid the hype of GPT-4, its popularity and capabilities remain questionable. ERNIE (Enhanced Representation through Knowledge Integration) is the name of the LLM that was developed by Baidu. Ernie Bot, the chatbot apparently has been into tests for around a decade, according to the Baidu CEO Robin Li in the livestreamed press conference. The company has been actively spending in AI research and development and spent about 21.4 billion yuan ($3.1 billion) in 2022 alone. The firm has been very eager to push ahead of the competition. While presenting Ernie Bot, Li said that he knows that the bot is not perfect. “So why are we unveiling it today? Because the market demands it,” said Li. Therefore, the model would even be integrated into the search engine as well. But what Wall Street Journal calls ‘China’s entry into the global AI race,’ does look decent enough. In the conference, Li showed capabilities of the bot to answer questions, generate posters and videos on a prompt, advise best locations in the city, and summarise famous Chinese science fiction novels. Another interesting part about the model is that it is bilingual – English and Chinese, and is also multimodal. Apart from language generation, the model also has text-to-image capabilities, which is called ERNIE-ViLG. This new bot is now trained on the third version of the LLM, ERNIE 3.0 Titan, which consists of 260 billion parameters, which is 50% more than that of ChatGPT’s. Too Eager to Lose But well, there are few things that are still dicey. Unlike how OpenAI unveiled ChatGPT live, Baidu demonstrated Ernie Bot’s capabilities through slides. Moreover, when compared to the recent announcement of GPT-4, the model cannot interpret images and generate responses in text. Major miss amid the hype that surrounds this feature. There is another problem here. The model performs a lot better in Chinese than it does in English. This puts the Chinese bot a hundred miles behind the capabilities of GPT-4. We are increasingly seeing that OpenAI is incorporating different languages in ChatGPT, and English being the first language, the audience is far far larger than that of Ernie. However, the hype around Ernie hasn’t impressed investors. The shares of the company fell by 10% after it unveiled its AI chatbot. “There is still a lot of uncertainty around Ernie’s capacity,” said Chim Lee, a tech analyst at Economist Intelligence Unit touted the lack of transparency and a pre-demonstrated livestream. This might not have excited the investors a lot, the Baidu CEO Li, however, has expressed happiness for launching his chatbot ahead of competitors like Google or Meta. Though this “fail first” method is quite famous of OpenAI, taking this approach just for the sake of it might not bode well for the company. Made in China, for China (for now) Beijing has strict internet controls and regulations. Since a chatbot like this is based on scraping information from online sources, its operations might get hindered. Baidu has to prove its bot to the Chinese government. The only plus point they have is the hype around these AI models and that OpenAI is restricted in their country. Just like ERNIE-ViLG, which was launched as a competition to Stable Diffusion and DALL-E, Ernie Bot is also expected to be restricted within its country. The text-to-image model rejects politically sensitive prompts, chances are that the bot will do the same. However, Chinese companies are great AI masters. The TikTok AI model they gave to the world is a case study in itself, which is still to be cracked. What if Ernie Bot also gives the world its own “Tik-Tok” moment?But Baidu has to be quick. Other giants like Alibaba and Huawei have also announced plans of bringing out their own chatbots. Maybe these giants would be able to learn from the mistakes of Baidu and tune their chatbots well enough for the global market.","excerpt":"While presenting Ernie Bot, Robin Li said that he knows that the bot is not perfect. “So why are we unveiling it today? Because the market demands it,”","categories":["AI Highlights"],"tags":["Baidu","baidu chatbot","china ai"],"author_name":"Mohit Pandey","publish_date":"2023-03-16T17:41:21","publication_year":"2023","word_count":682,"keywords":["Go","ChatGPT","API","DALL-E","OpenAI","AI","chatbots","baidu chatbot","GPT","china ai","Baidu","stable diffusion","R"],"extracted_tech_keywords":["AI","ChatGPT","OpenAI","chatbots","R","Go","API","GPT","DALL-E","stable diffusion"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/ernie-the-chinese-bot-is-here-to-take-on-chatgpt\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10059055,"title":"A chance to win prizes worth INR 3.5 lakh and be a part of Wipro! Sign up for Wipro’s Sustainability Machine Learning Hackathon","content":"MachineHack, in association with Wipro, is lining up a two-week-long hiring hackathon starting at 6.00 PM IST on 25 January, 2022. The challenge is to build AI\/ML models to forecast the Global Horizontal Irradiance (GHI) from the given datasets. The winners stand a chance to bag cash prizes worth INR 3.5 lakh and get hired by Wipro. Wipro Limited is a leading global information technology, consulting and business process services firm. The Fortune 500 company mobilises the power of cognitive computing, hyper-automation, robotics, cloud, analytics and emerging technologies to drive digital adoption of their clientele. Wipro is renowned for its comprehensive portfolio of services, strong commitment to sustainability and good corporate citizenship, and has over 220,000 employees serving clients across six continents. Though a little late in the day, the world is waking up to the deleterious effect of fossil fuels on our environment. As the doomsday clock ticks away, human beings are turning to renewable energy to avert a possible apocalypse. Solar energy is projected to reach $223.3 billion by 2026, growing at a CAGR of 20.5% from 2019 to 2026. Fortunately, the sun is a well-spring of clean energy. Along with being a global leader in artificial intelligence services from the latest reports of analysts like Forrester, IDC and Everest Group, Wipro has been rated as the second-best organization for data scientists to work in India in 2021 by Analytics India Magazine. The company has also been committed to reaching a Net-Zero Greenhouse Gas Emissions by 2040. Taking the cue, Wipro, in association with MachineHack, has designed a forecasting challenge to optimise solar power generation using AI\/ML models. Let the challenge begin! A solar power generation company wants to optimise solar power production and needs the prediction model to predict the Clearsky Global Horizontal Irradiance(GHI). The data is ten years at an interval of every 30 mins with the following data points: [‘Year’, ‘Month’, ‘Day’, ‘Hour’, ‘Minute’, ‘Temperature’, ‘Clearsky DHI’, ‘Clearsky DNI’, ‘Clearsky GHI’, ‘Cloud Type’, ‘Dew Point’, ‘Fill Flag’, ‘Relative Humidity’, ‘Solar Zenith Angle’, ‘Pressure’, ‘Precipitable Water’, ‘Wind Direction’, ‘Wind Speed’] MachineHackers are required  to predict  ‘Clearsky DHI’, ‘Clearsky DNI’,’Clearsky GHI’ values for a year for the company to get the maximum yield of the solar energy. The hackathon is open to data scientists, machine learning practitioners, analytics professionals, and tech enthusiasts at all experience levels. Apart from a chance to win cash prizes, the participants also get the opportunity to improve their Global Leaderboard Rankings and become the ultimate MachineHack GrandMaster. What are you waiting for? The holy grail is just one tap away! Hackathon start date –  January 25, 2022 Hackathon end date –  February 14, 2022, at 6:00 pm. Click here to participate Problem Statement: The winners of the hackathon stand a chance to win cash prizes worth up to INR 3.5 lakh. MachineHack has created a training dataset of 1,75,296 rows and 18 columns and a testing dataset of 17,520 rows and 15 columns. The hackathon demands a few pre-requisite skills like time series forecasting, multi-label prediction, and the ability to optimise “MSE” to generalise well on unseen data. Datasets will go live on January 25, 2022 Click here to participate Submission guidelines Sklearn models should support the predict() method to generate the predicted values. The participant should submit a .csv file with exactly  17,520 rows with 3 column [‘Clearsky DHI’, ‘Clearsky DNI’,’Clearsky GHI’]. The submission will return an Invalid Score if you have extra rows or columns. The file should have exactly 3 column. Note: Do not shuffle the sequence of the test series. If you are using pandas, use this submission code: submission_df.to_csv(‘my_submission_file.csv’, index=False) Evaluation criteria The submission will be evaluated using the Mean Squared Error. One can use sklearn.metrics.mean_squared_error to calculate the same The evaluation will be done using It will be based on the participants standing on the private leaderboard. The public leaderboard uses 30% of the provided test.csv dataset to evaluate. The private leaderboard uses 100% of the provided test.csv dataset to evaluate. Prizes Wipro’s Sustainability Machine Learning Challenge will select three (3) winners based on the given evaluation criteria. The cash prizes are as follows: First Prize: INR 2,00,000 Second Prize: INR 1,00,000 Third Prize: INR 50,000 Note: Make sure your MachineHack profile is up to date with all the relevant information\/details. Make sure you have gone through the ‘Rules’ section before participating. The participants will receive the prize money only if selected by Tredence and MachineHack. Hackathon start date –  January 25, 2022 Hackathon end date –  February 14, 2022, at 6:00 pm. Click here to participate Attribute Description: # Train: 1,75,296  rows x 18 columns # Test: 17,520 rows x 15 columns Data Attributes: ‘Year’, ‘Month’, ‘Day’, ‘Hour’, ‘Minute’, ‘Temperature’, 0C ‘Clearsky DHI’, w\/m2 ‘Clearsky DNI’, w\/m2 ‘Clearsky GHI’, w\/m2 ‘Cloud Type’, Cloud Type 0 Clear Cloud Type 1 Probably Clear Cloud Type 2 Fog Cloud Type 3 Water Cloud Type 4 Super-Cooled Water Cloud Type 5 Mixed Cloud Type 6 Opaque Ice Cloud Type 7 Cirrus Cloud Type 8 Overlapping Cloud Type 9 Overshooting Cloud Type 10 Unknown Cloud Type 11 Dust Cloud Type 12 Smoke Cloud Type 15 N\/A ‘Dew Point’, C ‘Fill Flag’, Fill Flag 0 N\/A Fill Flag 1 Missing Image Fill Flag 2 Low Irradiance Fill Flag 3 Exceeds Clearsky Fill Flag 4 Missing CLoud Properties Fill Flag 5 Rayleigh Violation Fill Flag any   N\/A ‘Relative Humidity’, % ‘Solar Zenith Angle’, Degree to calculate cos(θ) ‘Pressure’, mbar ‘Precipitable Water’, cm ‘Wind Direction’, Degrees ‘Wind Speed’ m\/s Skills: Time series forecasting Multi-label prediction Optimizing MSE Winners Announcement Final winners will be notified via email based on an aggregate score of their private leaderboard rankings. Datasets will go live on January 25, 2022 The hackathon will conclude on February 14, 2022, at 6:00 pm. Click here to participate","excerpt":"Wipro, in association with MachineHack, is teeing up a hiring hackathon on 25, January.","categories":["Deep Tech"],"tags":["Hackathon","hackathon for data scientists","Hackathons India","Machine learning hackathon","machine learning hackathons","Machinehack Hackathon"],"author_name":"Krishna Rastogi","publish_date":"2022-01-25T18:00:00","publication_year":"2022","word_count":966,"keywords":["Go","artificial intelligence","machine learning","AI","Machine learning hackathon","R","ML","Git","Hackathon","Ray","analytics","Machinehack Hackathon","hackathon for data scientists","Hackathons India","machine learning hackathons","Pandas"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","analytics","Ray","Pandas","R","Go","Git"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/a-chance-to-win-prizes-worth-inr-3-5-lakh-and-be-a-part-of-wipro-sign-up-for-wipros-sustainability-machine-learning-hackathon\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10056837,"title":"According to former Google CEO, Meta’s Metaverse is “not the best for society”","content":"Facebook has infamously changed its corporate name to Meta, with its CEO Mark Zuckerberg claiming, “From now on, we’ll be metaverse first, not Facebook first.” In Mark Zuckerberg’s vision for the metaverse, consumers will embody themselves in a virtual world powered by VR and AR hardware and advanced body sensors. In effect, people will be able to do everything they do in the real world in a seamless, well-lit virtual world: they can play games, attend virtual meetings, hang out with virtual friends, go to virtual concerts, collect virtual art, and shop for virtual goods. This will likely result in the greater development of parasocial relationships amongst people whilst simultaneously having the potential to splinter our shared reality by giving each individual access to their own customized realities. Zuckerberg claims that the purpose behind the metaverse is to naturalize, and not to strengthen, the relationship that people have with the internet. According to him, we are already irrevocably attached to the mobile internet and are in constant communication with each other through “small, glowing rectangles”—and the metaverse will simply serve to make this experience less artificial (touché) and to make us feel more present with the people we are interacting with. For better or worse? The former CEO of Google, Eric Schmidt, disagrees with Zuckerberg. When asked about Facebook’s metaverse plans in an interview with the New York Times—he said, “All of the people who talk about metaverses are talking about worlds that are more satisfying than the current world—you’re richer, more handsome, more beautiful, more powerful, faster.” According to Schmidt, the benefits that a metaverse has over the real world may lead people to forsake the world that exists in favour of a digital utopia, which, he thinks, “may not be the best thing for human society.” When social media was first being developed, there was optimism surrounding its potential to encourage openness and connectivity amongst communities. While this has proven to be true, we now know that these platforms have also served a significant role in enabling conflicts, creating unhealthy parasocial relationships, and negatively impacting the mental health of its consumers. While a metaverse would provide people with the significant benefit of being able to either escape or subvert the individual, geographic and social limitations that bind them—its greater anthropomorphism could also make it even easier for bad actors to enable conflicts, and its increased immersiveness could result in the formation of more parasocial relationships (and fewer real-world interactions) amongst the people within the metaverse. Parasocial relationships already exist Traci and Dave Gagnon first met as avatars at a company event in Las Vegas in 2015 and decided to hold their wedding in a virtual environment this year. The wedding was a hybrid event: the couple hired two videographers to capture the event and simultaneously cast the ceremony to the cloud. While Virbella was originally designed to be an immersive virtual platform for companies to hold events and to create a sense of togetherness within the metaverse, it has been asked by its users to hold bar mitzvahs, weddings, and other celebrations. Clearly, the metaverse has the potential to completely reimagine how people first meet and how social gatherings take place in the future: they could be bigger in scope, easier to afford, and unfettered by the shackles of reality. In the metaverse, everyone could get married at Lake Como or the Sixth Senses Fort Barwara in Sawai Madhopur—and you could become best friends or fall in love with someone without ever having met them in person. In Meta’s first major ad, we see a group of young people trying to get a close look at a picture of Henri Rousseau’s “Fight Between a Tiger and a Buffalo” at the Cleveland Museum of Art. As they move closer to the image, the painting suddenly flickers to life. Dean Kissick, writing for the New York Times, makes the interesting observation that in our century, the visionary role of creating new ways of experiencing life and observing the world has been passed from artists to engineers: it is now the founders of Silicon Valley who sell us the potential of utopian realities and draw our attention to the possibility of new universes. The creators of the metaverse are artists, and the metaverse is the art they are bringing to life—and, thereby, making “real.” The idea of a metaverse is still riddled with trade-offs—but I would argue that if something exists and people believe in it, then there should be no distinction between a real-world and a virtual world or real relationships and parasocial relationships.","excerpt":"Facebook has infamously changed its corporate name to Meta, with its CEO Mark Zuckerberg claiming, “From now on, we’ll be metaverse first, not Facebook first.”  In Mark Zuckerberg’s vision for the metaverse, consumers will embody themselves in a virtual world powered by VR and AR hardware and advanced body sensors. In effect, people will be […]","categories":["Global Tech"],"tags":[],"author_name":"Srishti Mukherjee","publish_date":"2021-12-22T17:00:08","publication_year":"2021","word_count":764,"keywords":["Go","programming_languages:R","AI","ML","programming_languages:Go","Git","RAG","Aim","ViT","R"],"extracted_tech_keywords":["AI","ML","Aim","RAG","R","Go","Git","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/according-to-former-google-ceo-metas-metavers-is-not-good-for-society\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10006239,"title":"Top 8 Ways To Manage Imbalanced Classes In Your Dataset","content":"Imbalanced classes in a dataset are often usual among classification problems in machine learning. Balancing an imbalanced class is crucial as the classification model, which is trained using the imbalanced class dataset will tend to exhibit the prediction accuracy according to the highest class of the dataset. Researchers have proposed several approaches to deal with this problem as well as improve the quality of the classifiers. Below here, we listed down the top eight ways you can manage the imbalanced classes in your dataset. 1| Changing Performance Metric Performance metrics play a fundamental role while building a machine learning model. Implying the incorrect performance metric on an imbalance dataset can yield wrong outcomes. For instance, although accuracy is considered to be an important metric for evaluating the performance of a machine learning model, sometimes it can be misleading in case of an imbalanced dataset. In such circumstances, one must use other performance metrics such as Recall, F1-Score, False Positive Rate (FPR), Precision, Area Under ROC Curve (AUROC), among others. 2| The More Data, The Better Machine learning models are data-hungry. In most cases, researchers spend most of their time in tasks like data cleaning, analysing, visualising, among others during an end-to-end machine learning process and contribute less time in data collection. While all of these steps are important, often the collection of data gets limited to certain numbers. To avoid such circumstances, one must add more data into the dataset. Collecting more data with relevant examples of the undersampled class of the dataset will help to overcome the issue. 3| Experiment With Different Algorithms Another way to handle and manage imbalanced dataset is to try different algorithms rather than sticking to one particular algorithm. Experimenting with different algorithms provides a probability to check how the algorithms are performing by a particular dataset. 4| Resampling of Dataset To deal with the imbalanced dataset, researchers have introduced a number of resampling techniques. One of the benefits of using these techniques is that they are external approaches using the existing algorithms. They can be easily transportable in case of both undersampling and oversampling. Some of the popular resampling methods are as follows- Random Oversampling: Oversampling seeks to increase the number of minority class members in the training set. Random oversampling is a simple approach to resampling, where one chooses members from the minority class at random. Then, these randomly chosen members are duplicated and added to the new training set. Random Undersampling: Undersampling is a process that seeks to reduce the number of majority class members in the training set. Random undersampling is a popular technique for resampling, where the majority class documents in the training set are randomly eliminated until the ratio between the minority and majority class is at the desired level. 5| Use of Ensemble Methods Use of ensemble methods is one of the ways to handle the class imbalance problems of the dataset. The learning algorithms construct a set of classifiers and then classify new data points by making a choice of their predictions known as Ensemble methods. It has been discovered that ensembles are often much more accurate than the individual classifiers which make them up. Some of the commonly used Ensemble techniques are Bagging or Bootstrap Aggregation, Boosting and Stacking. Know more here. 6| Generating Synthetic Samples Synthetic Minority Oversampling TEchnique or SMOTE is one of the most popular approaches for generating synthetic samples. SMOTE is an over-sampling approach in which the minority class is over-sampled by creating synthetic examples rather than by over-sampling with replacement. It is basically a combination of oversampling the minority (abnormal) class and undersampling the majority (normal) class, which is found to achieve better classifier performance (in ROC space) than only undersampling the majority class. In this technique, synthetic examples are generated in a less application-specific manner, by operating in the feature space rather than data space. Know more here. 7| Multiple Classification System Multiple Classification system is an approach where a classification system is built for imbalanced data based on the combination of several classifiers. It is a method for building multiple classifier systems in which each constituting classifier is trained on a subset of the majority class and on the whole minority class. The basis behind this method is the partition of the set of samples of the majority class in several subsets, each consisting of as many samples as the minority class. Know more here. 8| Use of Cost-Sensitive Algorithms Cost-Sensitive Learning is a type of learning that takes the misclassification or other types of costs into consideration. Cost-sensitive learning is a popular and common approach to solve the class imbalanced datasets. Popular machine learning libraries such as support vector machines (SVM), random forest, decision trees, logistic regression, among others, can be configured using the cost-sensitive training. Know more here.","excerpt":"Imbalanced classes in a dataset are often usual among classification problems in machine learning. Balancing an imbalanced class is crucial as the classification model, which is trained using the imbalanced class dataset will tend to exhibit the prediction accuracy according to the highest class of the dataset. Researchers have proposed several approaches to deal with […]","categories":["AI Trends"],"tags":["balancing classes","big data and data science","imbalanced dataset","SMOTE"],"author_name":"Ambika Choudhury","publish_date":"2020-09-08T11:00:21","publication_year":"2020","word_count":801,"keywords":["Go","machine learning","programming_languages:R","AI","big data and data science","SMOTE","ML","programming_languages:Go","imbalanced dataset","R","balancing classes"],"extracted_tech_keywords":["AI","machine learning","ML","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-ways-to-manage-imbalanced-classes-in-dataset\/","complexity_score":3,"technical_depth":7,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10052120,"title":"[Jobs Roundup] Latest Data Science And Analytics Jobs At Unicorns In India","content":"We have listed the latest data science and analytics jobs at unicorns in India this week. 1| Senior Data Scientist at MPL Location: Bangalore Responsibilities: Collaborate with Product, Design and Engineering teams to identify and develop an understanding of the Data Science requirements.Research and propose innovative statistical\/ML models to address the requirements.Target and formulate various problems in terms of key business metrics and measure the impact of the built ML techniques. Apply here. 2| Data Scientist at CRED Location: Bangalore Responsibilities: Own end-to-end business problems and metrics, build and implement ML solutions using cutting-edge technology.Create scalable solutions to business problems using statistical techniques, machine learning, and NLP.Design, experiment and evaluate highly innovative models for predictive learning. Apply here. 3| Senior Data Scientist at Chargebee Location: Remote Responsibilities: Explore, find and share actionable recommendations from data.Build predictive models and learning systems to enable business strategy.Engage with people across functions to surface questions, build hypotheses and test them. Apply here. 4| Lead Business Analyst at CoinDCX Location: Mumbai Responsibilities: Providing Analytical recommendations to influence our product, marketing and strategy.Design and evaluate product and growth experiments backed by solid hypotheses. Build and deploy machine learning models wherever necessary.Finding insights that influence decisions (product\/features), spanning from early data explorations about user behaviour to multivariate experiments and optimisations. Apply here. 5| Senior Product Analyst at BrowserStack Location: Mumbai Responsibilities: Use quantitative analysis to see beyond the numbers and understand what drives customer adoption of our products.Leverage insights from data to help product managers, designers, and engineers decide which features to build and why.Design experiments and interpret the results using analytics, research methodologies, and statistical techniques. Apply here. 6| Data Analyst at PharmEasy Location: Bangalore Responsibilities: Creating Dashboards\/Reports and communicating the same to the business team.Identification, Reporting & Tracking of Key Business Metrics at hourly, daily, weekly, monthly frequency.Coordination with Business Teams to identify new reporting requests and enhance the existing ones, business users to be communicated and explained the details if the need arises. Apply here. 7| Lead – Data Scientist at Innovaccer Location: Noida\/Bangalore Responsibilities: Design and lead the development of various artificial intelligence initiatives to help improve the health and wellness of patients.Work with the business leaders and customers to understand their pain points and build large-scale solutions for them.Define technical architecture to productise Innovaccer’s machine-learning algorithms and take them to market with partnerships with different organisations. Apply here. 8| Data Scientist at Urban Company Location: Gurgaon\/Bangalore Responsibilities: Work with stakeholders throughout the organisation to identify opportunities for leveraging company data to drive business solutions.Develop custom data models and algorithms to apply to data sets.Develop company A\/B testing framework and test model quality. Apply here. 9| Analyst – Decision Science at Droom Location: Gurugram Responsibilities: To analyse data from disparate sources and then turn it into a strategic advantage.Create a data analytics approach including data collection, data analysis using statistical techniques, and presenting the key findings.Ability to collaborate across departments. Apply here. 10| Director – Analytics at Moglix Location: Noida Responsibilities: You are the first and foremost leader within the department overseeing all activities of the data analytics, business intelligence, and data warehousing requirements.Manage and build a strong team of Data Science, Data Engineers handling business intelligence, and other key ingredients of the and state of the art Analytics.You will leverage business context and data analytics skills to define the business problem to be solved, develop methodology and roadmap, formulate an analytical approach and define the data gathering approach. Apply here. AIM Recruits A subsidiary of Analytics India Magazine, AIM Recruits is an India-focused executive search firm that partners with leading businesses to assess and acquire top data science talent to drive breakthrough performance. AIM Recruits is a leading executive search firm for Analytics, Data Science & Artificial Intelligence. 1| Senior Analyst – Data Visualization Tools Location: Multiple Responsibilities: Coordinates with client DV\/BI specialist or client analysts in creating\/conceptualising, maintaining and enhancing D3 visualisations.May work on other data visualisation tools such as but not limited to Python, R, QlikView and Tableau.Maintains documentation of visualisations (example tools twiki, git, Jira). Apply here. 2| Market Mix Modeling Role – Marketing Analytics Location: Delhi NCR Responsibilities: Work in Marketing Analytics- Market mix models.Use and implement basic and advanced statistical techniques like frequencies, cross-tabs, correlation, Regression, Decision Trees, Cluster Analysis.Understanding of consumer businesses such as Retail or CPG. Apply here. 3| Data Scientist Location: Any Responsibilities: Good understanding of analytical tools such as R, Java, Python and My SQL.Leverage data analytics and AWS infrastructure to solve healthcare problems.Individual contributor, hands-on experience in Python (must have), developing AI\/ML models from scratch (Regression, Deep Learning). Apply here. 4| Senior Data Engineer – AWS Location: Bangalore Responsibilities: Owning the design, operations and improvements for the Customer Data- warehouse infrastructure.Engineering solutions to aggregate and automate large scale data flows from varying sources.Collaborate with others to construct complex data sources for algorithms and machine learning models. Apply here. 5| Assistant Manager\/Manager – Data Engineering – ETL\/SQL\/Python Location: Any Responsibilities: Connecting, designing, scheduling, and deploying data warehouse systems.Developing data pipelines and enabling dashboards for stakeholders.Develop, construct, test and maintain system architectures. Apply here. 6| Big Data Senior Programmer – Distributed Computing Location: Bangalore Responsibilities: Own the module and take it from design, development, deployment and finally to production release and support.Perform product and technology assessments whenever needed.Develop analytical tools, working on BigData and Distributed Environment. Scalability will be the key. Apply here. 7| Artificial Intelligence Engineer – Machine Learning Location: Bangalore Responsibilities: Performing all technical development for assigned applications, including architecture, design, developing prototypes, writing new code and API- s, and performing unit and assembly testing of developed software also as needed.Build and automate our AI\/ML data pipelines & workstream from data analysis, experimentation, model training, model evaluation, deployment, operationalisation, and tuning to visualisation.Research new algorithms to tailor solutions to the Insurance world. Apply here.","excerpt":"We have listed the latest data science and analytics jobs at unicorns in India this week.","categories":["AI Hirings"],"tags":["weekly job updates"],"author_name":"kumar Gandharv","publish_date":"2021-10-23T06:00:00","publication_year":"2021","word_count":972,"keywords":["data science","machine learning","artificial intelligence","AI","ML","weekly job updates","NLP","RAG","Aim","deep learning","analytics"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","NLP","data science","analytics","Aim","RAG"],"url":"https:\/\/analyticsindiamag.com\/ai-hiring\/jobs-roundup-latest-data-science-and-analytics-jobs-at-unicorns-in-india\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10102952,"title":"Microsoft’s Thought-Out Plan for LLM Problems","content":"As irrational language models continue to increasingly influence every aspect of our lives, Microsoft has released an approach to make AI reason better. It’s called ‘Everything of Thought’ (XOT). This methodology draws inspiration from Google DeepMind’s AlphaZero, which uses tiny neural nets that can perform better with larger ones. The new XOT method was developed in collaboration with Georgia Institute of Technology, and East China Normal University. They used a blend of reinforcement learning and Monte Carlo Tree Search (MCTS) — techniques renowned for their effectiveness in complex decision-making. These techniques together allow language models to generalise efficiently to unknown problems, the researchers said. The researchers’ trials on a variety of challenging tasks, such as the Game of 24, the 8-Puzzle, and the Pocket Cube, yielded impressive results. XOT has outshone its contemporaries in tackling problems that have stymied other methods. This superiority, however, is not without limitations. The system, despite its advancements, has not reached a state of 100% reliability. However, the research team views the framework as an effective approach for incorporating external knowledge into language model inference. They are sure that it improves performance, efficiency, and flexibility simultaneously — a combination not attainable through alternative methods. Trying to Reason The reason for researchers eyeing games to integrate next into language models is because these models can form sentences with impressive accuracy, yet they fall short in an aspect critical to human-like thinking — the ability to reason logically. Researchers have long studied the subject. For years, the academic and tech communities have delved deep into this conundrum. However, despite their efforts in augmenting AI with more layers, parameters, and attention mechanisms, a solution remains missing. They have also been exploring multimodality but nothing much advanced and dependable has come out of that yet either. Earlier this year, a collaborative team at Virginia Tech and Microsoft released an approach titled ‘Algorithm of Thoughts‘ (AoT) to refine AI’s algorithmic reasoning as we all know how bad ChatGPT was at maths when released. Additionally, it suggested that with this training method, large language models could become capable of integrating their intuition into searches optimised for better outcomes. Moreover, a little over a month ago Microsoft also put the moral reasoning of these models under a microscope. As a result, the team proposed a new framework designed to assess its ethical decision-making skills. In the outcome, the 70-billion parameter LlamaChat model outperformed its larger counterparts. The result challenged the long-held belief that bigger always equates to better and the community’s over-reliance on large parameters. As the big tech companies continue to face the consequences of their irrational language models, Microsoft’s strategy appears to be one of careful progress. Rather than racing to add complexity to their models, they are picking their battles one by one. More Possibilities Microsoft has not disclosed plans for implementing the XOT method in its products. Meanwhile, Google DeepMind, led by CEO Demis Hassabis, is considering integrating AlphaGo-inspired concepts into its Gemini project, as he mentioned in an interview. Meta’s CICERO, named with a nod to the famed Roman orator, also entered the fray a year ago, raising eyebrows across the AI community for being so well skilled at the complex board game Diplomacy. This game, demanding not just strategic awareness but also the art of negotiation, has long been considered a challenge for AI. Yet, CICERO navigated these waters, showing an ability to engage in nuanced, human-like conversations. This discovery did not go unnoticed, especially in light of the benchmarks set by DeepMind. For years, the UK-based research lab has championed the use of games for developing and refining neural networks. Their feats with AlphaGo, have set a high bar, one that Meta met with borrowing elements from DeepMind’s playbook; combining strategic reasoning algorithms, like AlphaGo, with a natural language processing model, like GPT-3. Meta’s model stood out because for an AI agent to play Diplomacy, it has to not only understand the rules of the game, but also accurately gauge the possibility of betrayal by other human players. The agent’s capability to engage in conversations in natural-sounding language with other humans makes it the next best thing to be integrated as Meta continues to build Llama-3. The integration of CICERO’s capabilities with Meta’s broader AI initiatives could mark the beginning of true conversational AI.","excerpt":"Microsoft has (again) released an approach to make AI reason better.","categories":["AI Features"],"tags":[],"author_name":"Tasmia Ansari","publish_date":"2023-11-13T17:30:00","publication_year":"2023","word_count":719,"keywords":["Gemini Pro","ChatGPT","Go","attention mechanism","AWS","AI","neural network","GPT","Ray","R"],"extracted_tech_keywords":["AI","neural network","ChatGPT","Gemini Pro","Ray","AWS","R","Go","attention mechanism","GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/microsofts-thought-out-plan-for-llm-problems\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":60460,"title":"All The Tech Companies That Have Pledged No Layoffs","content":"According to a recent report by the Labour Department, approximately 3.3 million people filed for unemployment benefits for the week ending March 21. This is a record-breaking increase from the previous figures and is also the highest level of claims in recorded history. With the increasing fear of job losses looming amid COVID-19 pandemic, all major tech companies are aiming to pledge no layoffs for 2020. No Layoffs Amid Coronavirus Crisis Marc Benioff, the CEO of Salesforce, had recently tweeted a statement to his employees pledging no significant layoffs for the next 90 days. Benioff also urged his top executives to do their part by helping workers keep their jobs during this tumultuous time. In his eight-part Twitter thread, Benioff also wrote, “We will continue to pay our hourly workers while our offices are closed.” In another act of kindness, the CEO of Morgan Stanley promised to not issue any layoffs within the company. He reassured his employees that there will not be a reduction in staff in 2020. This does not take into consideration any performance issues or breach in code of conduct. Additionally, the chairman and CEO of Visa assured his employees that the company would have no COVID-19 related layoffs in 2020. Even Cognizant, with 65% of its employees working in India, has announced paying an additional 25% of the basic pay to most of its staffers in India. As far as Indian IT services firms are concerned, it is difficult to take any public stand on layoffs like many US firms, since 55-60% of their operating expenses comprise wages. So, any decision not to reduce the staff base can adversely impact their cost structure. However, despite the grim business outlook, firing staff owing to business disruption and consequent demand slowdown is not likely to happen in a big way given the sensitivity of the issue. Even many large Indian corporations, including Bajaj Auto, Vedanta group, and Essar group have stated that they will not be reducing their headcount. Meanwhile, according to experts, given the high employee utilization level and the strength of the reserve employees who are not working on any projects, tech companies are in a better position to absorb the adverse impact on cost. The utilization level of Infosys stood close to 85%, and Wipro was at around 80% at the end of the fourth quarter of 2019-20. Google, Microsoft, Facebook and others have also committed to paying hourly workers during COVID-19 office shutdowns. Amazon’s CEO, on the other hand, shared an Instagram post where he had encouraged those who had been laid off during this pandemic to consider working for the company. Joining the effort, IBM also stated that the company has no plans to lay anyone off any time soon, and started “running the business as virtually as possible.” Although the initiative is small, the movement is encouraging for tech companies. This commitment of CEOs toward their employees is a part of a bigger effort, where 180 CEOs and corporate executives pledged to change the way they conduct their businesses. All too often, we see company leaders becoming more concerned with business profits over their employees, and this initiative came as a breath of fresh air, where CEOs will be taking decisive actions that serve the best interests of their employees. Moving forward, the companies have promised to focus on their employees and not just on the bottom-line profit. With this commitment, the companies will start to invest in their employees, which begins with providing them with fair compensation, along with supporting them through training and education that will help them develop new skills. Experts also believe that this is the perfect time for the tech industry to send across a long-lasting message of providing a sense of security to their employees by retaining their existing staff members. Alongside, there are several areas where expenses can be cut, where the budgets can be rationalized to necessary aspects. Several business consultants and HR experts said companies need to be careful when it comes to allocating budgets, as job and pay cuts would further aggravate the problem. Aroon Kr Aggarwal, managing partner of a headhunter firm, Bradford Consultants, acknowledged the criticality of this unique situation for employers and employees. They stated that organizations should take a sympathetic and rational view of their staff. According to him, companies can rationalize their expenses by cutting costs from travel, variable pay\/bonuses, training etc, which can, in turn, save 5-6% of company costs, which can be used to avoid pay cuts. Agreeing to that, Juhie Sinha, the founder of TalentART Partners, said it is critical for companies, in this current crisis, to show their care and value for their employees. “Some immediate measures businesses can take immediately after the pandemic ceases is to use technology more effectively, minimizing travel to essential ones, conducting training programmes online, and having webinars instead of large events,” said Sinha. Wrapping Up The downturn is indeed creating an immense problem for tech companies to make informed decisions to keep their companies afloat. However, if the businesses are led with compassion and teamwork, it will be easier for tech companies to come out of this potential slowdown, enhancing the shared values of your employees. However, for now, we should applaud the companies taking bold strides to aid their employees, and help the economy in this time of crisis.","excerpt":"According to a recent report by the Labour Department, approximately 3.3 million people filed for unemployment benefits for the week ending March 21. This is a record-breaking increase from the previous figures and is also the highest level of claims in recorded history. With the increasing fear of job losses looming amid COVID-19 pandemic, all […]","categories":["AI Features"],"tags":["Facebook","Google","IT companies","Layoffs","Layoffs by IT companies","Salesforce","tech companies","Technology"],"author_name":"Sejuti Das","publish_date":"2020-03-31T14:00:00","publication_year":"2020","word_count":894,"keywords":["IT companies","Go","tech companies","Layoffs","AI","programming_languages:R","programming_languages:Go","RAG","Layoffs by IT companies","GAN","Aim","Technology","ViT","Salesforce","Google","disruption","Facebook","R"],"extracted_tech_keywords":["AI","Aim","RAG","R","Go","GAN","ViT","disruption","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/all-the-tech-companies-that-have-pledged-no-layoffs\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":14044,"title":"EdGE Networks- Transforming workforce with Data Science and AI","content":"Team EdGE A next gen HR technology solution providers, EdGE Networks offers solutions to large organizations for their most pressing HR challenges: solutions that’s powered by Artificial Intelligence. “During my first venture, I was often haunted by a question while recruiting talent, “What if my next potential candidate lies within the piles of discarded resumes?” That is when I realized that there was a tremendous need and scope for nextgen HR technology solutions, that could help organizations build future-ready workforces”, says Arjun Pratap, CEO and Founder, EdGE Networks. “Looking at the rapid pace of change taking place in India and the world, it became clear to me that HR technology powered by Artificial Intelligence, Data Science and Machine Learning would be best route to enable workforce transformation”, he further adds. This is how Arjun Pratap laid the foundation to EdGE Networks in 2012, which is built on the premise of a powerful search & match. They started with building technology platform that could read job descriptions in a resume like a fantastic hiring manager. Arjun Pratap-Founder & CEO, EdGE Networks How does EdGE Networks stand out from others? “It is the value that we add to our clients in the HR space, that makes us stand out. Our unique value proposition to clients is that we use Natural Language Processing algorithms with a Data Science and AI based approach to analyse Job Descriptions and profiles\/resumes, to provide a scored and stack ranked set of people for the job in question”, says Pratap. “This is further bolstered with our ability to do this for an internal workforce and to acquire talent from portals. Using advanced technology with minimum human intervention – the disruption we create is the value of time and accuracy and consistently”, he notes. EdGE Networks : A combination of Data Science, Analytics and AI to transform workforce- At EdGE, they have a expertise in Artificial Intelligence, Semantic Analysis, Data Science, Machine Learning and Predictive Modelling– all of which is geared to enable HR organizations lead with data and intelligence. Pratap notes “Data Science and AI has been transforming the way we do business today and HR is no exception to this. From fiction in the movie Transformers, to an impressive reality (Watson and Deep Blue), AI algorithms are challenging human intelligence. With their inherent pattern recognition, self-learning and cognitive capabilities, AI powered algorithms have the ability to perform complex jobs with speed and accuracy”. From talent acquisition and workforce optimization to workforce transformation, Data Science, AI and Analytics are acting as a strategic enabler for HR.  An efficient workforce management approach combines demand generation, skills repository, talent acquisition, allocation, learning and analytics to deliver robust talent connect. “This is what we deliver in EdGE—an end-to-end AI driven solution for workforce optimization – leading to workforce transformation”, says Pratap. “By leveraging AI, Data Science and Analytics, it has been bringing a sea of change in the speed, accuracy and timeliness of delivery. AI driven applications can act as HR’s weatherman helping them analyze the engagement level of employees, determine flight risk, uncover great talent in the frontline and more”, he adds. How does it help companies? EdGE Networks enhances the internal supply visibility through robust skills repository, enabling internal fulfilment and reducing costs incurred in external hiring. They help create intelligent job descriptions with appropriate keywords, lack of which is one of the key reasons for inaccurate resume matches. This enables instant resume matches, faster job fulfilments and lower hiring cycle time. Due to its inbuilt ability to parse structured and unstructured data, their platform produces more accurate matches, resulting in fewer or nil demand cancellations and allocation failures. “Our predictive and prescriptive analytics algorithms help optimize resourcing through accurate demand forecasting as well as curb attrition through flight risk scoring. Completing the workforce transformation cycle is the feature of Intelligent Learning Path, that maps the career path of staff, identifies skills gaps and suggests training courses to bridge the gaps”, he notes. Offerings by EdGE Networks- HIREalchemy is company’s flagship product, which is a cutting-edge talent acquisition platform powered by AI and Data Science. The platform auto-sources the right fit by parsing and analyzing both structured and unstructured information from internal database as well as external portals. It eases the process of selection as the platform throws up resume matches by scoring and stack ranking them based on business rules set by the client. “This exponentially enhances the speed and accuracy of job fulfilment and has the potential to save costs in millions of dollars for large businesses”, says Pratap. Other key distinctive offerings include: Workforce Optimization Solution, that forms an intelligence layer on top of HR systems and helping in effective organization building. Talent Analytics Suite, that helps in predicting attrition, forecasting resource demand and enabling fact-based decision making across the HR value chain. Challenges being at the analytics forefront- Pratap admits that the challenges is two-fold and points out two big challenges: In his words- Data quality: Business units within organizations work in silos with disparate HR systems. There is no centralized knowledge management system reducing the depth of analysis that HR could derive as against large data sets allowing multi-dimensional analysis. Data quality is a fundamental problem that many organizations grapple with. However with advanced BI tools, data cleansing and management, organizations are reaching a level where good insights could be gleaned from the data available. Skillset: HR analytics is clearly behind marketing or risk analytics disciplines. Organizations quickly jumped on the marketing and risk analytics bandwagon but trailed on HR analytics. But now, there are customers who are keen on adding a layer of intelligence on their existing HR systems to read and analyse data. They are not only focussing on reports and descriptive analytics but also using analytics to make accurate forecasts and take corrective actions based on data-driven recommendations. Growth story- They have had an impressive growth story as they have grown 228% cumulatively in the last 3 years. Last year, the company was awarded the Technology Fast 50 India 2016 awards by Deloitte, and was also one of the two winners to receive the MIDAS & Deloitte Tech Fast50 Award. EDGE was also named in the NASSCOM Emerge 2016 Awards in its list of top 50 emerging startups in India. EdGE unanimously made it to the Startup 50 Awards 2016 Winners List by The Smart CEO magazine – and top 10 undiscovered startups from India by Equidam. “A case study published by Harvard Business Review is testament to the work we do with our clients”, says Pratap. “Having built a strong analytics suite in descriptive (Channel mix metrics, Fulfillment metrics, Lead time analysis, Offer drop out ratio) and predictive (Attrition Analysis, Joining Probability, Hiring grid, Location Guidance & Gender ratio) space, we are looking to strengthen our prescriptive analytics suite”, he adds. Giving an insight on their client list, Pratap notes “Today, our clients are C-suite and HR leaders looking to staff projects with talented people, quickly and without having to rely on inefficient channels of recruitment and allocation. Wipro, HCL, Dell and Microland are some of the large companies that we work with.” Backed initially by two angels and Pratap himself, in 2013 NSDC funded EdGE Networks with a grant and a debt. Currently they are exploring next round funding which they intend to utilize to build next gen HR tech product and an A-class team.","excerpt":"A next gen HR technology solution providers, EdGE Networks offers solutions to large organizations for their most pressing HR challenges: solutions that’s powered by Artificial Intelligence. “During my first venture, I was often haunted by a question while recruiting talent, “What if my next potential candidate lies within the piles of discarded resumes?” That is […]","categories":["AI Startups"],"tags":["HR Analytics","Startups"],"author_name":"Srishti Deoras","publish_date":"2017-04-05T06:42:20","publication_year":"2017","word_count":1236,"keywords":["data science","Go","API","machine learning","artificial intelligence","AI","HR Analytics","Transformers","RAG","analytics","Startups","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","data science","analytics","Transformers","RAG","R","Go","API"],"url":"https:\/\/analyticsindiamag.com\/ai-startups\/startup-week-edge-networks-bringing-data-science-ai-workforce-transformation\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10004760,"title":"The Tech Behind Google’s ML Solution For Accurate Depth Estimation","content":"Depth estimation has several use cases in the real world, whether it be creating virtual avatars for your social media profiles or using the new portrait mode. However, it requires massive computational resources and sophisticated hardware which can accurately track the movement of the subject’s iris to estimate the metric distance from the camera. Such extensive computational resources make it a challenging task for deploying the technology for mobile devices, limiting its implementation. Also, with other obstacles like no fixed lighting source, locks of hair on the subject’s face or having their eyes squinched can make it even more difficult. Thus, to enhance the depth estimation process, a team of researchers from Google AI, has announced a new machine learning model for accurate depth estimation by iris tracking — MediaPipe Iris. Led by Andrey Vakunov and Dmitry Lagun, research engineers at Google Research, this ML model has been designed to accurately track the iris using a single RGB camera, in real-time without any advanced hardware. (A prototype) Use Case for far-sighted individuals. The font size that is seen remains independent of the distance of the device from the user. PC: Google AI Blog According to Google’s recent blog, this ML model — MediaPipe Iris claims to evaluate the distance between the subject and the camera with 10% relatively less error without the use of a depth sensor. This means, now the current tracking method of iris will not rely on the point of location the subject is looking, neither will it provide any form of identity recognition. Also Read: Google Open-Sources New Real-Time Hand Gesture-Tracking ML Pipeline Behind MediaPipe Iris — ML Pipeline For Depth Estimation To build this ML model, Google researchers first utilised their previous MediaPipe Face Mesh framework, which is a face geometry solution using AR to identify 3D face landmarks on mobile devices. From this mesh framework, the researchers pick up the eye region, which is then later used in the iris tracking machine learning procedure. Iris is marked in blue, and the eyelid has been marked in red. With that information in hand, the researchers divided the pipeline into three different models, which will first detect the face, then its landmarks and finally the iris. And therefore, the researchers designed a multitask model which was equipped with a unified encoder for each task, helping in using task-specific training data. The ML Pipeline has been implemented as a MediaPipe graph, which leverages facial landmark subgraph, iris landmark subgraph, and an iris and depth, to detect the points on the subjects. Eye landmarks are marked in red, and iris landmarks are marked in green. To train the iris landmark ML Model, the researchers take a patch of the image where the eye region is prominent and evaluate the eyelid and iris of the subject. The researchers manually evaluated approximately 50,000 images with various obstacles and head poses. With the solution, the researchers were able to track the iris and estimate the metric distance accurately. Cropped eye regions from the input to the model, which predicts landmarks via separate components. Explaining further, researchers stated in their blog — a lot of this could be attributed to the fact that the diameter of an iris remains roughly constant at 11.7±0.5 mm across the majority of humans. Case in point — a pinhole camera model (refer the image below) used for projecting onto a square pixel sensor, where it can figure out the distance of the subject by using the focal length of the camera. With that information, it can easily be established that the distance between the two is directly proportional to the physical size of the eye. Distance ‘d’ computed from focal length ‘f’ and the size of the iris. Also Read: How Data Scientists Create High-Quality Training DataSets For Computer Vision Evaluation & Release To measure the accuracy of the ML model, the researchers compared it to the already established depth sensor on a smartphone — iPhone 11, which utilised front-facing synchronised images of over 200 respondents. To which, the researchers highlighted that the depth sensor of iPhone 11 shows an error of < 2% for distances up to two meters. On the other hand, the proposed model for depth estimation using iris size has a mean relative error of 4.3% and a standard deviation of 2.4%. These results were tested on the respondents with and without eyeglasses and noted that glasses on subjects could increase the mean relative error slightly to 4.8%. Left: histogram of estimation errors; Right: Comparison between the actual and estimated distance by MediaPipe Iris. According to the researchers, as MediaPipe Iris requires no advanced hardware, the results highlight that it is a convenient method to estimate the metric distance with a wide range of cost-points. With the help of open-source, cross-platform framework for developers — MediaPipe — the solution can now be deployed on smartphones, desktops as well as on the websites. Furthermore, the researchers are continuously working on not making the ML model as a piece of surveillance equipment and pushing it for the broader research and development community. The MediaPipe Iris project page is on GitHub.","excerpt":"Depth estimation has several use cases in the real world, whether it be creating virtual avatars for your social media profiles or using the new portrait mode. However, it requires massive computational resources and sophisticated hardware which can accurately track the movement of the subject’s iris to estimate the metric distance from the camera. Such […]","categories":["Global Tech"],"tags":["Machine Learning","Machine Learning India","ML model","ML models"],"author_name":"Sejuti Das","publish_date":"2020-08-14T10:00:00","publication_year":"2020","word_count":856,"keywords":["ML model","Go","machine learning","AI","ML","ML models","Machine Learning India","Machine Learning","computer vision","RAG","Git","Aim","GitHub","R"],"extracted_tech_keywords":["AI","machine learning","ML","computer vision","Aim","RAG","R","Go","Git","GitHub"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/the-tech-behind-googles-ml-solution-for-accurate-depth-estimation\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10112421,"title":"Solo Giants: 6 One-Person Companies that Bucked the Trend","content":"OpenAI chief Sam Altman told Reddit co-founder Alexis Ohanian about an ongoing bet in their tech CEO group chat. It’s about when the first billion-dollar company with just one person, assisted by AI, come into being. “This would have been unimaginable without AI, but now it will happen,” Altman said. The tech industry sees a future where companies can succeed with smaller, more efficient teams, challenging the idea that growth requires more employees. “There’s going to be a new phenomenon where CEOs and founders are going to be so excited to get up and go to work with much smaller, much more performant, much more culturally strong teams,” Ohanian said. Well, there already are a few tech ventures that are successful as a one-person company. Plenty of Fish (POF) Founded in 2003 by Markus Frind, the dating website is headquartered in Vancouver, Canada. The company that began in 2003 was run single-handedly by the founder till 2007. The company today is one of the most popular dating sites and boasts around 170 million users in 2024. In the beginning, Frind operated POF from his home, relying on Google AdSense for revenue, which grew from $5 to $3,300 a month by year’s end. This success convinced him to focus solely on POF. Despite its growth and profitability, Frind avoided venture capital, preferring to grow POF independently. By 2008, POF had 15 million registered users and was generating about $10 million annually, all without any employees for the first five years. As POF continued to grow, Frind moved operations to Vancouver, hiring a team to expand and improve the website, introducing new features and premium options. By 2015, Frind decided to sell POF to The Match Group for $575 million, leaving the company. Builtwith Founded in 2005 by Gary Brewer, BuiltWith is headquartered in Brighton, United Kingdom, and operates as a website technology analysis tool. Despite its modest team of 50 employees as of 2020, the company did initially generate over $14 million in annual revenue with just one full-time employee. Brewer started BuiltWith in 2007 after realizing the need for a platform that could easily identify the technologies used by websites. Initially a side project, while he worked a corporate job in Sydney, the platform quickly became valuable for lead generation, allowing competitors services like Mailchimp to identify and contact potential customers. In 2011, Andrew Rogers, who was working on a similar project joined Brewer. The duo set a high bar for hiring additional staff, deciding only to consider expanding the team if monthly revenue reached $100k. BuiltWith’s marketing strategy relies on its free tool to attract paid users, supported by a weekly blog post from a contractor. By 2017, the platform had garnered significant traction, with over 2 million page views, more than 500,000 users a month, and between 2,000 to 3,000 paying customers on plans ranging from $295 to $995 per month. Balsamiq In 2008, Giacomo (Peldi) Guilizzoni founded Balsamiq, a wireframing and prototyping tool company based in Turin, Italy. By 2020, the company was generating $10 million annually. The company is a web-based user interface design tool for creating wireframes. Users generate digital sketches or concepts for an application or website, and facilitate discussion and understanding before any code is written. “My idea originally was to start a one-man company and let it stay that way but the product was too successful. The market told me that I needed to grow and that’s what we’ve been doing ever since,” Peldi said. The company now employs around 40 people and reported a revenue of $7.3M in 2023. Viral Nova Founded by Scott Delong in 2013, it is a curation site for the most viral content on the internet, similar to Buzzfeed. At its peak, it generated $5-$10M per year in revenue and 100M readers per month with 0 funding and 0 employees. The startup operated solely by DeLong and assisted by two freelance writers, managed to expand its website to match Buzzfeed’s reach and scale, attracting around 100 million monthly readers. This growth was achieved without employing any full-time staff or securing external investment. ViralNova was purchased by Zealot Networks in a cash and stock deal for upwards of $100 million. DeLong subsequently launched GodVine, a website that features inspiring stories that appealed to Facebook’s predominantly female audience. It rose to become a top 1,500 website according to Alexa’s traffic rankings. Stardew Valley Eric Barone began working on Stardew Valley in 2011 as a side project, with the game officially releasing in February 2016. Barone developed Stardew Valley, a hit Indie farming simulator RPG inspired by Harvest Moon. The game stands out for being crafted entirely by Barone, covering coding, art, music, and design, and became a commercial success without any external funding. Barone started Stardew Valley to improve his job prospects after college. What began as a resume builder, evolved into a full-time project fuelled by passion. Despite the lack of a formal plan, Barone’s intuitive development process and deep research into game mechanics led to a rich and authentic experience. He announced the game on Steam Greenlight in 2012, gaining early attention and community support. A partnership with Chucklefish in 2013 helped with logistics while allowing Barone to maintain creative control. By 2021, Stardew Valley had sold over 15 million copies across all platforms, indicating substantial revenue given the game’s price point, which varies by platform but is generally around $14.99. Photopea Founded by Ivan Kutskir, Photopea is a free online photo and graphics editor that has carved out a significant niche in the digital editing space. Launched as a solo project, Kutskir’s creation has grown into a robust platform, rivaling traditional software with its comprehensive suite of editing tools. The platform now boasts 10 million visits per month and generates approximately $1.5 million in annual revenue. Remarkably, users spend a collective 1.5 million hours each month on the site, a testament to its utility and user-friendly design. Kutskir’s journey with Photopea began as an experiment, aiming to offer a free, web-based alternative to Photoshop. This experiment quickly evolved into Kutskir’s primary source of income, with the platform’s growth trajectory showing no signs of slowing down. In the last year alone, Photopea crossed the $500,000 mark in annual recurring revenue (ARR), a significant milestone for any digital tool. Monetization of Photopea comes primarily through ads, a strategy Kutskir chose based on his previous experience with online games. This approach, coupled with licensing deals that allow for API customization, has proven effective. Kutskir today has an annual server cost of just $45, a figure that is almost unheard of for platforms with such high traffic.","excerpt":"Tech leaders predict that AI will enable solo entrepreneurs to achieve great success, challenging the idea that growth requires large teams.","categories":["AI Trends"],"tags":["Top Trend"],"author_name":"K L Krithika","publish_date":"2024-02-11T16:00:00","publication_year":"2024","word_count":1107,"keywords":["Top Trend","Go","API","OpenAI","AI","venture capital","Git","Aim","GAN","R","startup"],"extracted_tech_keywords":["AI","OpenAI","Aim","R","Go","Git","API","GAN","startup","venture capital"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/solo-giants-list-of-successful-one-person-companies\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10071673,"title":"Three innovation areas in AI that everyone is fighting for","content":"“In my youth, I would’ve argued that life is just a series of random events, devoid of any meaning. But as a data scientist, I must recognise that patterns sometimes emerge.” When Gilfoyle, one of the main characters on the popular sitcom Silicon Valley said this, he could have as well extended this to patterns that emerge in the AI innovation space. It is an undeniable fact that whenever a new, popular and eye-grabbing tool comes to the market, tech companies rush to replicate them and create their own renditions. This gives birth to a certain trend – a pattern. In recent years, three domains of AI innovation have seen heightened interest – language models, code generation tools, and art generation systems. Large Language model Being a social media giant with a user base of over 3.5 billion, Meta (earlier Facebook) heavily leverages NLP technology. Its tech team develops and deploys advanced NLP systems to understand and communicate with users and offer, in the company’s own words, “a safe experience—no matter what language they speak”. Speaking of NLP-related innovations, Meta has introduced several initiatives. In May, Meta introduced the Open Pretrained Transformer (OPT-175B) – a language model trained on publicly available datasets. What made it different from other language models was that it was released along with pretrained models and the code required to train and use them. The Zuckerberg-owned company followed it up with the release of its 66 billion parameter model. OPT-175B joins the list of other large language models from Meta. Last year, Meta used the Generative Spoken Language Model (GSLM). Unlike other language models, GSLM is a textless NLP model which uses raw audio signals as input. According to the company, GSLM overcomes the challenges of text-based language models, which is the requirement of large text datasets. One of the most important pieces of research in the field of NLP from Meta came in the form of RoBERTa, an optimised method for pretraining NLP systems. This tool gives state-of-the-art results on General Language Understanding Evaluation (GLUE) – a widely used NLP benchmark. RoBERTa is based on Google’s BERT model. When introduced in 2018, the BERT model truly revolutionised the large language model space. It offered state-of-the-art results in the machine learning community, especially in performing a range of NLP tasks. One of the biggest accomplishments of this model was not only in terms of its massive size (340 million parameters) but also in applying the bidirectional training of Transformer, a popular attention model to language modelling. One of the watershed moments of language learning came with the introduction of the GPT-3 model. A 175 billion parameter model was unheard of when introduced by OpenAI in 2020. There have been several bigger and better models since then. In 2021, Google introduced the Switch Transformer model, which was trained on a staggering 1 trillion parameters. Other important large models include Deepmind’s Gopher and Chinchilla with 280 billion and 70 billion parameters; Microsoft-NVIDIA’s Megatron-Turing NLG model with 530 billion parameters; Google’s GLaM (1.2 trillion) and LaMDA (137 billion) models. Recently, Google’s LaMDA model was in the news when a (now former) Google employee Blake Lemoine claimed the AI has become sentient. Lemoine was soon put on a break and eventually fired from the company. Code generator OpenAI developed Codex, an AI tool that translates natural language to code; it can interpret simple commands in natural language and execute them on the users’ behalf. Based on Codex, OpenAI, in collaboration with Microsoft and GitHub, introduced Copilot in 2021. OpenAI calls it an AI pair programmer that helps write better code. The Copilot tool draws context from the code being worked on and suggests whole lines or entire functions. Credit: OpenAI Soon after, Salesforce open-sourced a machine learning system called CodeT5 that can understand and generate code in real-time. As per the team, CodeT5 could achieve state-of-the-art performance for tasks like code defect detection, predicting whether the code is vulnerable to exploits, clone detection, and detecting snippets of code which may have the same functionality. Earlier this year, DeepMind introduced AlphaCode, a code generator that uses a transformer-based language model to output lines of codes at an ‘unprecedented scale’. It displays skills like language understanding and problem-solving ability. When tested against human programmers on the popular competitive programming platform Codeforces, AlphaCode averaged a ranking of 54.3% across ten contests. Another famous code generation tool was from the researchers at Carnegie Mellon University – Frank Xu, Uri Alon, Graham Neubig, and Vincent Hellendoorn. Called PolyCoder, it is a model based on GPT-2 (trained on the database of 249 GB of code in 12 programming languages). Other code generation tools from major tech companies are Facebook’s TransCoder, Intel’s ControlFlag, and a new feature in Microsoft’s Power Apps. AI art generation AI-based art generation tools marked the AI scene in the year’s first half. It began with the launch of DALL.E 2 by OpenAI. This image generation tool creates realistic images from a natural language text description provided by the user. It can combine concepts, styles, and attributes. It can also add and remove elements while taking shadows, reflections, and textures into consideration. OpenAI recently made the beta version of this tool available to the general public. I used DALL·E 2's inpainting feature to make a version of Eames Powers of Ten, leaving the prompts in to explain each of the 57 steps. #DALLE @OpenAI #AI pic.twitter.com\/lP11NP2q41— Adam Pickard (@adampickard) July 25, 2022 DALL.E 2 is the successor of DALL.E, introduced by OpenAI at the beginning of 2021. The name DALL.E is actually a portmanteau of Salvador Dali and the robot from Wall-E. It is a neural network that is trained on 250 million pairs of images and texts collected from the internet. Along with the introduction of DALL.E, OpenAI also launched the Contrastive Language–Image Pretraining (CLIP) model that builds on zero-shot transfer, natural language supervision, and multimodal learning. The model learns visual concepts from natural language supervision; it can be applied to any visual classification benchmark. Credit: OpenAI Circling back to DALL.E 2, the kind of rage that it created not only in the AI research community but also in the general public was unprecedented. Soon after, Google introduced Imagen. It is a text-to-image diffusion model which offers superior levels of photorealism and language understanding. Recently, Meta too introduced an AI-based art generation tool called Make-A-Scene. It is a multimodal generative AI method to generate images corresponding to the textual prompt provided by the user. Other major and popular AI art generation tools include HuggingFace’s Craiyon (formerly DALL.E Mini) and Midjourney from the Midjourney Lab. With the introduction of several art-generating tools in just the last few months, it is easy to identify it to be the flavour of the AI season. But anyone who has closely followed the field would tell you that this may not last very long. The AI community will move to better and shinier pastures. As long as the pasture is developing, no one is really complaining!","excerpt":"It is an undeniable fact that whenever a new, popular and eye-grabbing tool comes to the market, tech companies rush to replicate them and create their renditions.","categories":["AI Features"],"tags":["AI innovation","AI Tool","DALL.E","GPT","GPT-3","LLMs","OpenAI"],"author_name":"Shraddha Goled","publish_date":"2022-07-28T10:00:00","publication_year":"2022","word_count":1169,"keywords":["GPT-3","machine learning","AI innovation","OpenAI","AI","LLMs","neural network","AWS","SLM","RAG","GPT","NLP","Aim","generative AI","AI Tool","DALL.E"],"extracted_tech_keywords":["AI","machine learning","neural network","NLP","generative AI","OpenAI","Aim","SLM","RAG","AWS"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/three-innovation-areas-in-ai-that-everyone-is-fighting-for\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10119429,"title":"Confluent Unveils New Capabilities to Simplify AI and Stream Processing","content":"Confluent, the data streaming giant, announced several new capabilities for its Apache Flink offering. These capabilities aim to simplify AI integration and bring stream processing to workloads everywhere. The key updates include AI Model Inference, the Confluent Platform for Apache Flink, and cost-saving Freight clusters. AI Model Inference, an upcoming feature on Confluent Cloud for Apache Flink, enables teams to easily incorporate machine learning into data pipelines using simple SQL statements. This removes the complexity involved in using streaming data for AI development, allowing organisations to innovate faster and deliver powerful customer experiences. Confluent also introduced Confluent Platform for Apache Flink, a fully-supported Flink distribution that enables stream processing in on-premises or hybrid environments. This offering provides organisations with expert guidance from Confluent’s foremost experts in the data streaming industry, ensuring secure and up-to-date stream processing applications. Additionally, Confluent unveiled Freight clusters, a new serverless cluster type that offers up to 90% lower cost for high-throughput use cases with relaxed latency requirements, such as logging or telemetry data. Powered by Elastic CKUs, Freight clusters seamlessly auto-scale based on demand, minimising operational overhead and optimising costs. “Apache Kafka and Flink are the critical links to fuel machine learning and artificial intelligence applications with the most timely and accurate data,” said Shaun Clowes, Confluent’s Chief Product Officer. Confluent’s AI Model Inference removes the complexity involved when using streaming data for AI development by enabling organisations to innovate faster and deliver powerful customer experiences.” These new capabilities are designed to help organisations leverage fresh, contextual data for accurate, real-time AI-driven decision-making while enhancing performance and value. Support for AI Model Inference and Freight clusters is available early to select customers, while Confluent Platform for Apache Flink will be available later this year.","excerpt":"Confluent also unveiled Freight clusters, a new serverless cluster type that offers up to 90% lower cost for high-throughput use cases.","categories":["AI News"],"tags":[],"author_name":"Shyam Nandan Upadhyay","publish_date":"2024-05-02T11:59:16","publication_year":"2024","word_count":289,"keywords":["machine learning","artificial intelligence","AI","ML","serverless","RAG","Aim","Kafka","SQL","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","Aim","RAG","serverless","Kafka","R","SQL"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/confluent-unveils-new-capabilities-to-simplify-ai-and-stream-processing\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10070173,"title":"Microsoft&#8217;s Intelligent Cloud revenue up 7.8% in Q2 2022","content":"Microsoft’s Intelligent Cloud Revenue jumped by 7.8 percent from Q1 2022 to Q2 reaching USD 18.33 billion, compared to the USD 17 billion the company generated in Q1, according to TradingPlatforms.com. “Microsoft’s recent quarterly results indicate that the company continues to outpace expectations, as customers continue to accelerate their shift to cloud computing. The current unsettling economic environment has not yet impacted the company’s main growth driver,” Edith Reads, financial analyst at TradingPlatform, said. Azure engagement and usage Microsoft has been growing its cloud business steadily over the last few years. But this quarterly report is the first time it has seen such a large boost. Analysts said this growth was driven by “a strong increase in Azure engagement and usage.” Azure is Microsoft’s premium cloud platform, which includes features like machine learning and advanced analytics. The company has lined up a steady stream of deals for Azure software, which stores data and runs applications for corporations. Internet-based Office programs continue to grow as Microsoft persuades customers to pay for high-end versions and expanded contracts. Additionally, revenues were strong across Microsoft’s three product categories: Productivity and Business Processes (Office, LinkedIn, and Dynamics); Intelligent Cloud (server products and cloud services); and More Personal Computing (Windows, Xbox, Search, and Surface). Microsoft 365 consumer subscribers have grown by over 15 percent to 58.4 million. In addition, office 365 Commercial subscriber seat growth was up 16 percent year-over-year. Microsoft’s Cloud business is growing rapidly, but it’s also facing more challenges than ever before. The company’s Cloud business is growing at a high annualised rate, but critics argue that this growth should not be celebrated. They are disappointed that many of the company’s customers need to spend additional money on security measures. Moreover, Microsoft remains bullish on stronger revenue growth in the future despite macroeconomic concerns lingering over the sector. Its recent growth has shored up its market value by nearly $100 billion. Moreover, its Chief Executive Satya Nadella has affirmed they’d continue investing in the space even if growth tanked.","excerpt":"The current unsettling economic environment has not yet impacted the company’s main growth driver.","categories":["AI News"],"tags":["Intelligent Agent"],"author_name":"AIM Media House","publish_date":"2022-06-30T13:12:22","publication_year":"2022","word_count":336,"keywords":["Go","API","machine learning","cloud_platforms:Azure","cloud computing","AI","R","Intelligent Agent","analytics","ViT","Azure"],"extracted_tech_keywords":["AI","machine learning","analytics","cloud computing","Azure","R","Go","API","ViT","cloud_platforms:Azure"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/microsofts-intelligent-cloud-revenue-up-7-8-in-q2-2022\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10104225,"title":"Mamba vs Transformers &#8211; Who is Future of LLMs?","content":"Researchers Albert Gu and Tri Dao from Carnegie Mellon and Together AI have introduced a groundbreaking model named Mamba, challenging the prevailing dominance of Transformer-based architectures in deep learning. Their research unveils Mamba as a state-space model (SSM) that demonstrates superior performance across various modalities, including language, audio, and genomics. For example, the researchers tried their language modelling with the Mamba-3B model that outperformed Transformers based models of the same size and matched Transformers twice its size, both in pretraining and downstream evaluation. Click here to check out the GitHub repository. Mamba is presented as a state-of-the-art model with linear-time scaling, ultra-long context, and remarkable efficiency, outperforming Transformers in tasks it has been tested on. The model is built on the foundation of structured state space models (SSMs), showcasing promising results in information-dense data scenarios, particularly excelling in language modelling where previous subquadratic models have fallen short compared to Transformers. The researchers emphasise Mamba’s efficiency through its selective SSM layer, designed to address the computational inefficiency of Transformers on long sequences up to a massive million sequence length, which is a major limitation in Transformers. I’m always excited by new attempts to dethrone transformers. We need more of these. Kudos to @tri_dao & @_albertgu for pushing on alternative sequence architectures for many years now. https:\/\/t.co\/cf67Xa2PBS— Jim Fan (@DrJimFan) December 4, 2023 Installation details for Mamba include the use of causal Conv1d layers and the core Mamba package, with additional requirements such as Linux, NVIDIA GPU, PyTorch 1.12+, and CUDA 11.6+. Mamba’s versatility is demonstrated through its integration into an end-to-end neural network, offering a comprehensive language model with varying model dimensions and layers. Mamba provides pretrained models with different parameters and layers, showcasing its adaptability to various tasks and data sizes. Evaluations of Mamba’s performance involve running zero-shot evaluations using the lm-evaluation-harness library, with comparisons against other models such as EleutherAI’s pythia-160m. The research identifies a key weakness in existing subquadratic-time architectures, attributing it to their inability to perform content-based reasoning. Mamba addresses this weakness by allowing selective propagation or forgetting of information along the sequence length dimension, demonstrating significant improvements over traditional models. Despite the shift away from efficient convolutions, Mamba employs a hardware-aware parallel algorithm in recurrent mode, resulting in fast inference and linear scaling in sequence length. Mamba emerges as a compelling contender challenging the Transformer paradigm, demonstrating superior performance in diverse modalities and promising advancements in the field of deep learning.","excerpt":"Attention is not all you need anymore.","categories":["AI News"],"tags":["Generative Pre-Trained Transformer"],"author_name":"Mohit Pandey","publish_date":"2023-12-06T10:48:40","publication_year":"2023","word_count":405,"keywords":["CUDA","Go","PyTorch","neural network","AI","Transformers","Git","deep learning","Generative Pre-Trained Transformer","R"],"extracted_tech_keywords":["AI","deep learning","neural network","PyTorch","Transformers","CUDA","R","Go","CUDA","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/mamba-is-here-to-mark-the-end-of-transformers\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10165304,"title":"OpenAI Introduces a $50M Consortium for AI Research and Education","content":"OpenAI has announced NextGenAI, a new consortium with 15 leading research institutions to accelerate AI-driven breakthroughs in research and education. “Excited for this-advancing research and education with AI,” wrote CEO Sam Altman on X. The company is committing $50 million in funding, compute resources, and API access to support universities, libraries, and hospitals using AI. The blog also shared recent research breakthroughs in AI in universities. The Ohio State University is advancing AI in digital health, manufacturing, energy, mobility, and agriculture. Harvard University and Boston Children’s Hospital are using AI for medical diagnosis of rare diseases and improving AI alignment in healthcare. Duke University is researching metascience to identify where AI can accelerate scientific progress. Texas A&M University is training students in responsible AI use. MIT is providing API access for students to train and fine-tune AI models. Howard University is integrating AI into curricula, teaching, and university operations. The University of Oxford is digitising rare texts in the Bodleian Library. The Boston Public Library is digitising public domain materials to improve accessibility. NextGenAI reinforces OpenAI’s academic partnerships, following the launch of ChatGPT Edu in May 2024. The initiative will equip institutions with advanced AI tools and ensure the technology benefits students, researchers, and educators worldwide.","excerpt":"The blog also shared recent research breakthroughs in AI in universities.","categories":["AI News"],"tags":["OpenAI"],"author_name":"Aditi Suresh","publish_date":"2025-03-05T16:49:01","publication_year":"2025","word_count":206,"keywords":["ChatGPT","GenAI","API","AI alignment","OpenAI","AI","Git","GPT","responsible AI","R"],"extracted_tech_keywords":["AI","GenAI","ChatGPT","OpenAI","R","Git","API","GPT","AI alignment","responsible AI"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/openai-introduces-a-50m-consortium-for-ai-research-and-education\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":50176,"title":"How Data Analytics Brought Affordability &#038; Better UX For This Mumbai-based Logistics Startup","content":"With a team of 70+ employees across Tamil Nadu and Maharashtra, Mumbai-based FreightBro has been revolutionising the trillion-dollar forwarding industry through digitisation of the manual processes in logistics and bringing in data and analytics. Founded by Raghavendran Viswanathan (Co-founder & CEO), Mohammed Zakkiria A (Co-founder) and V Anand Babu, (Director & Co-founder), it has been providing a digital platform for freight forwarders to procure rates faster, reduce operational inefficiencies, boost sales, make informed decisions and provide seamless user experience using data analytics. FreightBro is tackling several challenges of shipment inefficiencies and bringing rate management, price discovery, instant quote creation, quote management, shipment management, shipment tracking all on a single platform. How did it all begin? And why the freight forwarding industry? Viswanathan says that during one of his jobs at a Swiss freight-forwarding company, he used to have frequent conversations with Mohammed around logistics and freight. During that time they also met Anand, who is an industry veteran with 25 years of experience with one of the world’s largest shipping companies. “We brainstormed and after many such sessions and building our initial product, took a leap of faith to start FreightBro with an aim to simplify freight forwarding through digitization.”says Viswanathan. That is how FreightBro was born. As Viswanathan shares, while he is responsible for strategizing and managing the team at FreightBro, Mohammed strategises company’s growth and heads marketing functions, and Anand heads the partnerships verticals. Analytics And Data Science At FreightBro FreightBro is not the usual logistics company. It extensively uses analytics and data science to understand user behaviour and understanding the issues that business stakeholders face on a day to day basis in the industry. Currently, their analytics team in 5-people strong. “On a use case basis, for example, number of users searching for freight rates, or the number of quotes generated, trade lanes that are searched for the most, the trade lanes for which users seek more and more competitive rates, the liners preferred for a certain trade lane, these are mines of information and the more analytical data we have as these, the better it helps us to make the product more user-oriented and thus deliver an optimum end-user experience,” shares Viswanathan. It allows stakeholders to leverage this significant data to make critical business decisions and deliver better to their customers. “Similarly, for the users who are using our platform, many more such parameters help them assess the performance of the sales executives in the department,” he said. Not just use data analytics, but FreightBro is enabling clients to digitize key processes such as sales, procurement and operations to give end to end visibility to their customers. The startup is automating the manual tasks to bring in more efficiency and reduce the costs of logistics. “The exporter or importer gets the best-in-class experience similar to online ticket booking platforms or net banking,” shares Viswanathan. FreightBro also offers a custom mobile application to 3PLs to perform the same functions as the platform but on their mobile devices thereby making freight forwarding all the simpler and more accessible. “We offer a subscription model which is in line with the asset-light model of 3PLs compared to traditional technology providers who charge a fortune for basic ERPs,” he added. Technology Is Driving Them To Stay Ahead Of The Game The competition in the space is no less. But Viswanathan is quite positive that with a digital platform like theirs, even small-sized freight forwarders can leverage technology to offer a holistic user experience. Some of the benefits that he lists are: Firstly, it is a SaaS-based solution and per user\/per transaction-based model. Therefore, cost & benefit-wise it makes much more sense for freight forwarders in comparison to legacy software that is always charged per license or installation basis. Affordability, Delivering better user experience and latest technology are the main drivers why freight forwarders prefer using the FreightBro platform. Secondly, they believe in the fact that disruption alone will not enable growth in the long run and wiping out freight forwarders from the ecosystem is never a viable solution. The founders believe that for digitization to thrive and freight forwarders to grow and adapt, there is a need for more enablers. Growth Story Viswanathan says that the evolution of technology has transformed freight forwarders from mere movers of shipments to transport management service providers. With mobiles and advent of digitization, freight forwarders can now not only effectively and efficiently concentrate on their core competency but also offer an extraordinary service level. The startup has allowed even small-sized sized freight forwarders to leverage the digitization wave to offer a holistic user experience through finding better rates or schedules, increase quote to win ratio, effectively manage documentation and provide real-time visibility of shipments to their customers. He proudly shares that so far they have 500+forwarders on their platform and more than 1200 active users. Future Plans The founders currently plan to continue working on their IP to ease the workload for freight forwarders and save their time. With a partnership with the National Association Of Freight And Logistics, Dubai, and Port of Wilhelmshaven, Germany, they further plant to continue their strategic collaborations that impact the freight forwarding industry. “We have also recently joined the Digital Hub Logistics, Hamburg and are now a part of their extensive network of German startups and corporates,” said Viswanathan. They had also recently raised a seed round of funding last year from Suresh Kumar, former Global CIO of Bank of New York Mellon, which they plan to use for product development and expansion.","excerpt":"With a team of 70+ employees across Tamil Nadu and Maharashtra, Mumbai-based FreightBro has been revolutionising the trillion-dollar forwarding industry through digitisation of the manual processes in logistics and bringing in data and analytics.  Founded by Raghavendran Viswanathan (Co-founder & CEO), Mohammed Zakkiria A (Co-founder) and V Anand Babu, (Director & Co-founder), it has been […]","categories":["AI Startups"],"tags":["Data Analytics","mba in data analytics"],"author_name":"Srishti Deoras","publish_date":"2019-11-18T19:00:00","publication_year":"2019","word_count":924,"keywords":["data science","mba in data analytics","AI","ML","Git","RAG","Aim","analytics","disruption","Data Analytics","R","startup"],"extracted_tech_keywords":["AI","ML","data science","analytics","Aim","RAG","R","Git","disruption","startup"],"url":"https:\/\/analyticsindiamag.com\/ai-startups\/how-data-analytics-brought-affordability-better-ux-for-this-mumbai-based-logistics-startup\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10103311,"title":"YOLO Vulnerabilities Underscore Necessity for Open Source Auditing","content":"A security audit of YOLO (You Only Look Once) object detection algorithms revealed major security vulnerabilities. Trail of Bits– a cybersecurity firm– claimed to have found 11 vulnerabilities in Yolo v7. If exploited, it could lead to compromised object detection, privacy breaches and safety risks, subsequently leading to far greater consequences. Tesla AI reportedly uses the Yolo algorithm along with its sensors for detecting objects and people around the vehicle. Similarly, Roboflow, a computer vision platform that provides tools and services to simplify the process of building computer vision models also utilises the Yolo model for object detection and image segmentation tasks. Introduced first in 2015 by Joseph Redmon et al via a paper titled, ‘You Only Look Once: Unified, Real-Time Object Detection’, YOLO’s object detection algorithms find extensive usage in drones, autonomous vehicles, robotics, and numerous manufacturing companies, making them one of the most widely adopted algorithms in these industries. Even if not directly employing these algorithms, numerous entities are leveraging their open-source nature to build upon them. Over the years, the model has gone through several iterations and earlier this year, Yolo v8 was released. Security vulnerabilities Key security concerns, highlighted by the Trail of Bits report, include the absence of defensive coding practices, a lack of unit tests or a testing framework, and inadequate validation and sanitisation of both user and external data inputs in the YOLOv7 codebase. Trail of Bits, which is based in New York, mentions that the YOLOv7 codebase is not suitable for security-critical applications or applications that require high availability, such as autonomous vehicles. This is because the lack of defensive coding practices is evident in the codebase, as it is not written or designed with a focus on security. “If an attacker is able to control or manipulate various inputs to the system, such as model files, data files, or configuration files, they could perform a denial-of-service attack with low effort,” the report said. While they do not recommend using the codebase for mission-critical applications or applications that require high availability, the algorithm is often used commercially and in mission-critical applications. Addressing Open source vulnerabilities In the blog post, the cybersecurity company notes that Yolo is a product of academic work generally not meant for commercial use and does not have appropriate cyber hygiene. However, since the algorithm has been widely adopted for commercial use, Trail of Bits has also suggested remedies to overcome the vulnerabilities. “We only intended to bring to light the risks in using such prototypes without further security scrutiny,” they said. Meredith Whittaker, president of the Signal Foundation, commended the work. “AI systems, like all networked computation, rely on core open components that are often in disrepair, maintained by volunteers if at all, even as trillion dollar companies built on top of them,” she tweeted. “This company does really pragmatic security reviews of really important open source infrastructure,” another X user said. Security audits of open-source projects are welcoming In the present day, generative AI has become the hottest technology in the tech industry. However, the revolution is equally led by the open-source community. LLaMA large language model, an open-source model released by Meta, is among the most popular LLMs to date. Notably, the Technology Innovation Institute (TII) in Abu Dhabi, United Arab Emirates released Falcon earlier this year, which topped the Hugging Face OpenLLM Leaderboard. However since open-source projects often rely on a community of contributors and users, security audits help establish and maintain trust within this community. Hence, regular audits of numerous generative AI projects could also prove to be highly beneficial. Moreover, such audits should also not be limited to open-source projects. For example, earlier this year, an engineer at cybersecurity firm Tenable found a major red flag in the Microsoft Azure platform. Tenable CEO said that most of these Microsoft Azure users had no clue about the vulnerability, and hence, couldn’t make any informed decision about compensating controls and other risk-mitigating actions.","excerpt":"A cybersecurity firm said that they have found 11 security vulnerabilities in Yolo v7.","categories":["AI Features"],"tags":["Open Source AI","YOLO"],"author_name":"Pritam Bordoloi","publish_date":"2023-11-20T13:00:00","publication_year":"2023","word_count":658,"keywords":["Go","Hugging Face","AI","R","computer vision","RAG","Open Source AI","Aim","object detection","YOLO","generative AI","Azure"],"extracted_tech_keywords":["AI","computer vision","generative AI","Aim","Hugging Face","RAG","object detection","Azure","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/yolo-vulnerabilities-underscore-necessity-for-open-source-auditing\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10077075,"title":"Meta Launches $1,500 VR Headset and Other Highlights From Meta Connect 2022","content":"Tech behemoth Meta unveiled its latest virtual-reality headset Meta Quest Pro on Tuesday, at Meta Connect 2022. The much-awaited, high-end VR device is designed to entice creators and working professionals to adopt Zuckerberg’s vision for an immersive digital future. Meta Quest Pro is the firm’s latest offering in a product line previously branded as Oculus. On releasing the new headset, the CEO said the goal is to make enhanced connections available to all PC users. The headset uses Qualcomm Snapdragon XR2+ semiconductor chip, making it more powerful than the Quest 2. It is 40% thinner with a new chip offering 12GB of RAM and 256GB of storage. Key Highlights at Meta Connect 2022 Meta unveiled the Meta Quest Pro VR headsetIt collaborated with Microsoft, making its leading business applications – Teams, Office, and Microsoft 365 – usable in VRNew social experience at Meta Horizon Worlds in VRNew avatars and ongoing development related to AR glasses and personalized AI Source: Meta Connect 2022 The VR headset will cost $1,500 – three times the price of the Quest 2 – with an aim of targeting working professionals. Meta seeks to transcend the notion that VR is primarily the realm of gamers, with an effort to broaden its audience. Reports show that the Meta’s Quest 2 headset has sold an estimated 15 million units. CEO Mark Zuckerberg said, “The ideal customers for this [are] gonna be either people who just want the highest-end VR device — so enthusiasts, prosumer type folks — or people who are trying to get work done.” The advanced headset is built to expand and bridge the possibilities of virtual and mixed reality. Meta claims that the device is the first headset to integrate inward-facing sensors that capture ‘natural facial expressions’ and eye tracking. Users can raise an eyebrow, smile, or make eye contact with someone, and their avatars would do the same. The company states, “All of this helps improve social presence — the feeling that you’re right there together with someone no matter where in the world you are.” Promoting the new headset as a powerful tool for working professionals, Meta has also partnered with Microsoft by bringing tools such as Teams, Office, Windows, and Xbox to VR. Source: Meta To know more, check out Meta’s blog here.","excerpt":"The VR headset will cost $1,500 – three times the price of the Quest 2","categories":["AI News"],"tags":["augmented reality","Metaverse","Microsoft","Qualcomm","Sundar Pichai","virtual reality"],"author_name":"Bhuvana Kamath","publish_date":"2022-10-12T13:29:38","publication_year":"2022","word_count":381,"keywords":["Go","programming_languages:R","Metaverse","AI","augmented reality","virtual reality","programming_languages:Go","Git","RAG","Aim","Qualcomm","R","Sundar Pichai","Microsoft"],"extracted_tech_keywords":["AI","Aim","RAG","R","Go","Git","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/meta-launches-1500-vr-headset-and-other-highlights-from-meta-connect-2022\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10104414,"title":"Microsoft-OpenAI: A Symbiotic Partnership or One-Sided Affair?","content":"When we thought Microsoft was done making their employees lazy and even super lazy, the company recently announced its one year completion of Copilot’s release. With this, further enhancements to Copilot capabilities were announced, integrating all the features that already exist on ChatGPT. Is the Microsoft-OpenAI partnership favouring one side a little too much? Microsoft Brings Business As per a recent report, Microsoft Azure-OpenAI partnership has given rise to close to 79% of ChatGPT Enterprise customers. Within four months of launching ChatGPT Enterprise, the company acquired close to 25,000 customers. That’s a whopping growth, but only 16% of the users directly came from ChatGPT Enterprise. The figures are an indication of how ChatGPT Enterprise has still not managed to attract the audience through their direct channel and is heavily reliant on Microsoft users. As per recent data, there are over 722 million Azure users and 85% of Fortune 500 companies use Microsoft Azure Cloud- numbers that can be potential ChatGPT Enterprise users. Microsoft’s Azure OpenAI Service allows enterprises and developers to build on GPT, Dall.E and other OpenAI applications. Furthermore, Microsoft’s continuous effort to make their office suite, Microsoft 365, a wholesome multi-functional workspace platform has only been ramped up. Microsoft will release several new features to Copilot at the start of 2024. The upcoming enhancements will include the integration of GPT-4 Turbo, OpenAI’s latest model. Furthermore, DALL-E 3, GPT4- Vision, and Code Interpreter will also be integrated. A Complete Ecosystem Microsoft Azure, the cloud platform has elaborately built an ecosystem that holds the users in one place. Transitioning to another platform involves pushback that dissuades users from migrating. In the case of Microsoft Azure users, who are already accustomed to the platform, will not shift to ChatGPT Enterprise, but rather continue being on the Microsoft ecosystem. When AIM got in touch with a financial enterprise at the GitHub Copilot event last month in San Francisco, the company which had experimented with both Amazon CodeWhisperer and Google Duet chose to adopt Copilot in May owing to its alignment to their existing platform of Microsoft Azure. Partnership with Restrictions With Satya Nadella having a non-voting observer seat at the OpenAI board, it is only obvious that the collaboration will become stronger with Microsoft having strategic powers over OpenAI’s business decisions. Interestingly, Microsoft is also trying its best to not be closely tied with OpenAI too. Though Microsoft’s AI image was boosted through the OpenAI partnership, the recent Altman exit fiasco has made everyone wary of the results of heavily relying on one company. Recently, during Microsoft’s annual shareholder meeting, the company emphasised to its investors about how they do not solely rely on OpenAI. Nadella said that the company will take a ‘broad tent approach,’ and spoke about how the company is working on their language model Phi and also offering open-source models on Azure. Security Makes Everyone Wary The notoriety associated with security issues of ChatGPT has always marred its image in an enterprise world. At the recent AWS re:Invent, AWS CEO Adam Selipsky, took a dig at ChatGPT’s security flaws while introducing guardrails and safety features for Amazon Bedrock. Adam Selipsky at AWS re:Invent Though a strong partnership exists between the companies, there were times when Microsoft restricted employees from using AI tools such as ChatGPT owing to security and data concerns. The ban fiasco that happened recently, led to rumours to which Sam Altman confirmed that there wouldn’t be any form of restriction from OpenAI’s end on Microsoft Office 365 products. A statement that reinstates Microsoft’s dominance in the whole OpenAI-Microsoft partnership. From the progress of both companies, it is evident that Microsoft’s users will benefit from OpenAI , resulting in a lot of ChatGPT Enterprise user traffic from Microsoft Azure. However, Microsoft continues to play it safe by portraying themselves as not solely reliant on OpenAI for their AI developments. It looks like a so-called symbiotic relationship that heavily favours Microsoft.","excerpt":"OpenAI is heavily reliant on Microsoft, and not the other way round","categories":["Global Tech"],"tags":["Azure","ChatGPT","Code Interpreter","Github Copilot","GPT","Microsoft","OpenAI"],"author_name":"Vandana Nair","publish_date":"2023-12-08T12:30:00","publication_year":"2023","word_count":654,"keywords":["Go","ChatGPT","OpenAI","AI","AWS","R","Code Interpreter","Github Copilot","Git","GPT","Ray","Aim","Azure","Microsoft"],"extracted_tech_keywords":["AI","ChatGPT","OpenAI","Aim","Ray","AWS","Azure","R","Go","Git"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/microsoft-openai-a-symbiotic-partnership-or-one-sided-affair\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10054528,"title":"How Do Activation Functions Introduce Non-Linearity In Neural Networks?","content":"A neural network is modelled after the human brain that consists of neurons. To obtain the output, a neural network accepts an input and weights summed with bias before arriving at the output. An activation function is one of the most important factors in a neural network, which is applied to the input before deriving an output. This activation function decides whether or not a neuron will be activated and transferred to the next layer. It is also referred to as threshold or transformation for the neurons as it decides if the neuron’s input is relevant in the prediction process or not. The main job of an activation function is to introduce non-linearity in a neural network. One way to look at this is that without a non-linear activation function, a neural network will behave just like a single-layer perceptron; it does not matter how many layers it has. Table of contentsActivation Function & Non-LinearityNon-Linear Activation Functions Activation Function & Non-Linearity At the most basic level, a neural network consists of three main layers: Input layer: This layer accepts input from the outside world to the network. No computation is performed here, and the only job is to pass the received information to the hidden layer. Hidden layer: This layer accepts information from the input layer and performs all computations. This layer is hidden from the outside world and transfers the result to the next layer. Output layer: It accepts the result from the hidden layer and relays it to the outside world. The activation function is present in the hidden layer. The activation layer cannot be linear because irrespective of how complex the architecture is, a linear activation function is effective only one layer deep. Also, the real world and associated problems are highly non-linear. The only situation where using linearity may prove beneficial is in case of regression problems (think predicting housing prices). A linear activation function lacks in performing backpropagation. Thus, it is not recommended to be used in a neural network. While a model may perform a task even without the presence of an activation function in a linear manner, it would lack efficiency and accuracy. The significance of the activation function lies in making a given model learn and execute difficult tasks. Further, a non-linear activation function allows the stacking of multiple layers of neurons to create a deep neural network, which is required to learn complex data sets with high accuracy. Overview of Recurrent Neural Networks And Their ApplicationsTowards Better Data Engineering: Mostly People, But Also Process and TechnologyRahul SinghTypes Of Activation Functions In Neural Networks And Rationale Behind ItWhat Are Activation Functions And When To Use Them Non-Linear Activation Functions Examples of non-linear activation functions include: Sigmoid function: The Sigmoid function exists between 0 and 1 or -1 and 1. The use of a sigmoid function is to convert a real value to a probability. In machine learning, the sigmoid function is generally used to refer to the logistic function, also called the logistic sigmoid function; it is also the most widely used sigmoid function (others are the hyperbolic tangent and the arctangent). A sigmoid function is placed as the last layer of the model to convert the model’s output into a probability score, which is easier to work with and interpret. Another reason to use it mostly in the output layer is that it can otherwise cause a neural network to get stuck in training time. TanH function: It is the hyperbolic tangent function whose range lies between -1 and 1, hence also called the zero-centred function. Because it is zero centred, it is much easier to model inputs with strongly negative, positive or neutral values. TanH function is used instead of sigmoid function if the output is other than 0 and 1. TanH functions usually find applications in RNN for natural language processing and speech recognition tasks. On the downside, in the case of both Sigmoid and TanH, if the weighted sum input is very large or very small, the function’s gradient becomes very small and closer to zero. ReLU function: Rectified Linear Unit, also called ReLU, is a widely favoured activation function for deep learning applications. Compared to Sigmoid and TanH activation functions, ReLU offers an upper hand in terms of performance and generalisation. In terms of computation too, ReLU is faster as it does not compute exponentials and divisions. The disadvantage is that ReLU overfits more, as compared with Sigmoid. Softmax function: It is used to build a multi-class classifier to solve the problem of assigning an instance to one class when the number of possible classes is larger than two. Softmax ensures that the sum of outputs is 1. The softmax function squeezes the outputs for each class between 0 and 1 and divides it by the sum of outputs.","excerpt":"The main job of an activation function is to introduce non-linearity in a neural network.","categories":["AI Features"],"tags":["Activation Function"],"author_name":"Shraddha Goled","publish_date":"2021-11-30T17:00:00","publication_year":"2021","word_count":802,"keywords":["machine learning","TPU","programming_languages:R","AI","Activation Function","neural network","deep learning","data engineering","RNN","R"],"extracted_tech_keywords":["AI","machine learning","deep learning","neural network","TPU","R","data engineering","RNN","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-do-activation-functions-introduce-non-linearity-in-neural-networks\/","complexity_score":4,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10135269,"title":"AIM Research launched VendorAI, A Comprehensive AI Vendor Database for Strategic Decision-Making","content":"The AI industry is growing rapidly, with vendors offering a range of solutions across various domains such as data science, data engineering, AI observability, generative AI services, computer vision, and data annotation. This expansion brings with it the challenge of selecting the right vendors, understanding the competitive landscape, and identifying potential acquisition targets. To address this need, AIM Research is proud to launch VendorAI, a comprehensive database that profiles over 200 AI vendors, offering in-depth insights into their capabilities, expertise, and performance. VendorAI is designed for businesses seeking to enhance their strategic decision-making, whether for mergers and acquisitions (M&A), competitive intelligence, or vendor selection. The platform brings together a wealth of data in one place, structured to help users make informed choices when navigating the AI vendor ecosystem. The Structure of VendorAI VendorAI features a detailed set of data columns that cover all aspects of an AI vendor’s operations, leadership, technology, and competitive advantage. These columns include: AIM PeMa Score: This is a proprietary score calculated by AIM Research that averages key performance metrics of the vendor. It offers a quick snapshot of the vendor’s market standing and operational efficiency. Website and LinkedIn: Links to the vendor’s website and LinkedIn profile, offering easy access to further information and contact points. Company Overview: A brief description of the vendor’s history, services, and areas of expertise. Founded Year: The year the vendor was established, helping users understand how long the company has been operational. HQ: The location of the vendor’s headquarters. Employee Count: The total number of employees, giving an indication of the company’s scale. Employee Count Growth (Last 6 Months): Tracking employee growth over the last six months provides insight into the vendor’s expansion or contraction trends. Leadership (CEO\/Founder): The key leadership figures, typically the CEO or founder, are listed alongside their professional background. Leadership LinkedIn Profile: Direct links to the LinkedIn profiles of leadership, allowing users to assess their professional networks and experience. Office Locations and Client Locations: Information about where the vendor operates and where their clients are located globally. Main Delivery\/Operation Centers: Data on the vendor’s main operational hubs, helping assess their service delivery capacity. Solution Type: The specific type of solution offered, such as AI services, platforms, or specific tools. Proprietary\/Flagship Platforms and Technology: Details about the vendor’s unique or flagship technologies that set them apart from competitors. Industry Expertise and Specialization: Insights into the specific industries the vendor serves and their areas of expertise. Years in Operation: How long the company has been providing AI solutions or services. USP\/Competitive Advantage: The vendor’s unique selling propositions (USP) and competitive advantages are highlighted to differentiate them from others in the market. Categories: Vendors are categorized based on their primary area of focus, such as data science, data engineering, AI observability, generative AI services, computer vision, and data annotation. Benefits of VendorAI for M&A In the world of M&A, finding the right target company for acquisition requires access to deep insights and structured data. VendorAI makes this process easier by offering a comprehensive view of AI vendors, allowing businesses to assess potential acquisition targets based on a range of critical metrics. For example, companies can evaluate vendors using the AIM PeMa Score, which aggregates performance metrics to provide a high-level understanding of a vendor’s market strength. Furthermore, tracking employee growth over the past six months can reveal whether a vendor is on an upward trajectory or facing operational challenges. Insights into proprietary platforms and flagship technologies can help businesses identify vendors with unique intellectual property or technological innovations that could complement their own portfolios. VendorAI’s leadership data is also invaluable for M&A activities, providing easy access to the backgrounds of key figures like CEOs and founders. Understanding the leadership team’s experience and professional networks can help acquiring companies evaluate the strength of the management team, which is often critical to the success of a post-acquisition integration. Competitive Intelligence Made Easy Staying ahead of the competition in the rapidly evolving AI market requires constant monitoring of the competitive landscape. VendorAI offers businesses a structured way to gather competitive intelligence, allowing them to track the activities, innovations, and performance of key players in the industry. With information on each vendor’s employee growth, delivery centers, client locations, and industry specialization, businesses can gain a granular understanding of where their competitors are investing resources and expanding their market presence. The platform’s insights into proprietary platforms and flagship technologies allow users to identify which vendors are innovating and developing cutting-edge solutions. VendorAI’s focus on tracking the number of Fortune 500 clients a vendor serves also provides a valuable indicator of market credibility. Vendors with significant Fortune 500 clients are often seen as more established and trustworthy, making them key players to watch in the competitive landscape. By using VendorAI, businesses can benchmark their competitors’ strengths and weaknesses, identifying opportunities for differentiation or areas where they may need to catch up. The platform also enables businesses to keep an eye on emerging vendors, helping them stay ahead of market trends and potential disruptors. Simplifying Vendor Selection Choosing the right vendor in the AI space can be a daunting task, especially with so many options available. VendorAI simplifies this process by providing a structured and detailed view of over 200 vendors, allowing businesses to compare their offerings, industry expertise, and technology platforms. With detailed data on solution types, proprietary platforms, industry specialization, and years of operation, businesses can use VendorAI to match their specific needs with the capabilities of vendors. The platform’s filtering options, such as applicable tags for company selection, make it easy to narrow down vendors based on specific criteria such as geographic location, industry focus, or technology expertise. In addition, VendorAI’s insights into the number of Fortune 500 clients served by each vendor can be a key factor in decision-making. Vendors with significant Fortune 500 clients often have proven experience in handling large, complex projects, making them strong candidates for businesses with similar requirements. The USP\/Competitive Advantage column provides further differentiation, helping businesses identify vendors with unique capabilities that set them apart from the competition. Whether a company is looking for a vendor with a specific technology platform or one with expertise in a niche industry, VendorAI provides the data necessary to make an informed choice. AIM Research’s Broader Product Suite VendorAI joins AIM Research’s existing portfolio of industry-leading tools designed to support businesses in strategic decision-making. Alongside VendorAI is GCC Explorer, AIM Research’s comprehensive list of Global Capability Centers (GCCs) in India. GCC Explorer offers detailed insights into the GCC ecosystem, helping businesses understand the operational landscape and identify potential partners or acquisition targets. Together, VendorAI and GCC Explorer provide a holistic view of both the AI vendor landscape and the GCC ecosystem, enabling businesses to make informed decisions that align with their long-term strategic goals. With over 200 AI vendors profiled across key categories like data science, data engineering, AI observability, generative AI services, computer vision, and data annotation, VendorAI provides the insights needed to make strategic decisions with confidence. By leveraging the platform’s rich data set, businesses can navigate the complex AI vendor landscape, identify potential acquisition targets, gain competitive intelligence, and choose the right vendors to meet their specific needs. Whether you’re a company looking to expand through M&A or simply trying to stay ahead of the competition, VendorAI is the ultimate tool for strategic decision-making in the AI industry. Access VendorAI here.","excerpt":"VendorAI by AIM Research is a comprehensive AI vendor database profiling over 200 vendors, providing deep insights for M&A, competitive intelligence, and vendor selection.","categories":["AI Highlights"],"tags":["AIM Research"],"author_name":"AIM Media House","publish_date":"2024-09-14T17:00:10","publication_year":"2024","word_count":1234,"keywords":["data science","Go","API","AI","AIM Research","computer vision","RAG","Aim","generative AI","Rust","R"],"extracted_tech_keywords":["AI","computer vision","data science","generative AI","Aim","RAG","R","Go","Rust","API"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/aim-research-launched-vendorai-a-comprehensive-ai-vendor-database-for-strategic-decision-making\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":35308,"title":"Why The Future Of NLP Is In The Hands Of Tech Giants Like Google, Microsoft &#038; Amazon","content":"In 1954, with the success of the Georgetown experiment in which the scientists used a machine to translate random sentences from Russian to English, the field of computational linguistics took giant strides towards building an intelligent machine capable of recognising and translating speech. These models were even used in translations during the Nuremberg trials. Nonetheless, the future of machine translation was nowhere close to the forecast due to sluggish computational devices and scarcity of data to train on. Today, after six decades, machines have transitioned from mechanical statistical models to electronic neural models which can perform complicated tasks like speech recognition and sentiment analysis with great accuracy. Today machines with NLP have the capability to analyse a wide variety of data (documents, text, voice) and have a range of practical uses, the discipline encompasses multiple capabilities. These range from sentiment analysis to text categorization. NLP can scour documents and classify them by topic, even without a programmer defining in advance which topics to look for. The field of NLP is innovating every other day thanks to the constant effort of tech giants like Google, Microsoft, Facebook and Amazon. One thing common with these tech giants is their willingness to open source their innovations. Their belief in accelerated innovation through transparency has started to see fruition in the form of diversified real world applications from homepods to chatbots. Here’s a look at NLP roadmap of these mega-firms: Google Open-domain question answering (QA) is a benchmark task in natural language understanding (NLU). Google’s AI researchers recently released a paper introducing  Natural Questions (NQ), a new dataset for QA research, along with methods for QA system evaluation. In contrast to tasks where it is relatively easy to gather naturally occurring examples, the definition of a suitable QA task, and the development of a methodology for annotation and evaluation is challenging. Also, there is the Bidirectional  Encoder Representations from Transformers or BERT, which was open sourced last year, offers a new ground to embattle the intricacies involved in understanding the language models. Pre-training a binarised prediction model helps understanding common NLP tasks like Question Answering or Natural language Inference. Along with innovation in its own backyard, Google is also backing NLP startups like Armorblox. A cybersecurity startup, Armorblox aims to tackle data leaks via online attacks like email spear phishing. Armorblox developed an NLP engine that derives insights from enterprise communications and data. It offers policy recommendations by learning over time what’s mission-critical for a given organization, and in the event of a potential or attempted breach, it automatically sends alerts to the relevant people and teams. By applying NLU, Armorblox is able to address a whole new layer of security that has been inaccessible to other security solutions: the content and context of communications. This has been the biggest challenge and attack vector because hackers know that they can exploit this weakness. Facebook Engineers developing NLP algorithms often turn to deep-learning systems to build their solutions, such as Facebook’s PyTorch platform. Facebook AI Research is open-sourcing PyText, a natural-language-processing (NLP) modelling framework that is used in the Portal video-calling device and M Suggestions in Facebook Messenger. PyText builds on top of PyTorch by providing a set of interfaces and models specifically tuned for NLP. Internally, Facebook is using PyText to power NLP in their Portal video-calling device. PyText addresses a common problem for NLP projects: the tradeoff between rapid experimentation and scalability in production. Researchers experiment with new ideas, rapidly tweaking models to achieve performance goals. PyText can utilize multiple GPUs for distributed training and can train multiple models at once, reducing the overall training time. The PyText code also comes with pre-trained models for several common NLP tasks, including text classification, named-entity recognition, and joint intent-determination and slot-filling, which is a staple of chatbot development. Microsoft Microsoft’s NLP group focuses on developing efficient algorithms to process text and to make their information accessible to computer applications. This group addresses natural language problems using a mix of knowledge-engineered and statistical\/machine-learning techniques to disambiguate and respond to natural language input. For example, the grammar checkers in Microsoft Office for English, French, German, and Spanish are some of the byproduct of Microsoft’s NLP enhancements. Text Analytics API is a cloud-based service that provides advanced natural language processing over raw text and includes four main functions: sentiment analysis, key phrase extraction, language detection, and entity linking. Language Understanding (LUIS): A machine learning-based service to build natural language into apps, bots, and IoT devices. Quickly create enterprise-ready, custom models that continuously improve. Designed to identify valuable information in conversations, LUIS interprets user goals (intents) and distils valuable information from sentences (entities), for a high quality, nuanced language model. LUIS integrates seamlessly with the Azure Bot Service, making it easy to create a sophisticated bot. Under the hood of LUIS: { “query”: “Book me a flight to Cairo”, “topScoringIntent”: { “intent”: “BookFlight”, “score”: 0.9887482 }, “intents”: [ { “intent”: “BookFlight”, “score”: 0.9887482 } Microsoft also had presented a novel, fully data-driven, and knowledge-grounded neural conversation model last year, aimed at producing more contentful responses. Amazon Amazon Comprehend is a natural language processing (NLP) service that uses machine learning to find insights and relationships in text. No machine learning experience required. The service identifies the language of the text; extracts key phrases, places, people, brands, or events; understands how positive or negative the text is; analyzes text using tokenization and parts of speech, and automatically organizes a collection of text files by topic. Amazon Comprehend Medical, a variant of Comprehend, identifies the relationship among the extracted medication and test, treatment and procedure information for easier analysis. For example, the service identifies a particular dosage, strength, and frequency related to a specific medication from unstructured clinical notes. Amazon too has backed NLP startups like its contemporaries. US-based mobile marketing firm Vibes has launched a tool called Conversational Analytics, which uses natural language processing (NLP) to extract consumer insights from unstructured messaging content. Powered by Amazon Comprehend, this new tool uses machine learning to find insights and relationships in the messaging content exchanged between brands and consumers, and provide a more comprehensive understanding of the purchasing journey. The Potency Of NLP The potential for governments to employ these advancements in language processing defies limitation. Governments collect massive numbers of unstructured records. Fragmented information presented in deeply non-mathematical formats, at volumes far too huge for human assessment, are now available for the kind of deep-pattern analysis originally reserved for numerical databases. NLP can open the public sector to new insights, better-tailored services, and faster responses to information. According to an online financial portal, the artificial intelligence in supply chain market alone is expected to reach USD 10,110.2 million by 2025 from USD 730.6 billion in 2018, at a CAGR of 45.55% during the forecast period. Market growth can be attributed to the increasing adoption of deep learning and NLP technologies for automotive, retail, and manufacturing applications in APAC. Some of the key players in this space are Intel (US), NVIDIA (US), Xilinx (US), Samsung Electronics (South Korea), Micron Technology (US), IBM (US), Microsoft (US), and Amazon (US). The biggest challenge for NLP models, however, has been the lack of training data. Small training sets restrict many NLP models from performing real-time rendering of both contextual and free from context tasks. The next big challenge for these NLP models is to reach a human-level understanding of language which has been in the pursuit since the times of Leibniz and Descartes.","excerpt":"In 1954, with the success of the Georgetown experiment in which the scientists used a machine to translate random sentences from Russian to English, the field of computational linguistics took giant strides towards building an intelligent machine capable of recognising and translating speech. These models were even used in translations during the Nuremberg trials. Nonetheless, […]","categories":["Global Tech"],"tags":["Amazon","Facebook","Google","Google Translate","Machine Learning","Microsoft","NLP"],"author_name":"Ram Sagar","publish_date":"2019-02-24T04:56:06","publication_year":"2019","word_count":1247,"keywords":["artificial intelligence","machine learning","Google Translate","AI","PyTorch","ML","Amazon","Machine Learning","Transformers","NLP","Aim","deep learning","analytics","Google","Facebook","Microsoft"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","NLP","analytics","Aim","PyTorch","Transformers"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/why-the-future-of-nlp-is-in-the-hands-of-tech-giants-like-google-microsoft-amazon\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":37146,"title":"How MoneyTap Is Personalising Moneylending With Artificial Intelligence","content":"With digital personal loans and instant paperless transactions touted to be the future of the banking sector, banks and startups are seizing the opportunity to make it big in the lending industry. A relatively younger startup which has been working in this field is MoneyTap, an instant personal loan lending app which can transfer as much ₹5 lakh to its users over the phone. The app which was launched in 2015 with RBL Bank, lets the lenders procure a sum between ₹3,000 to ₹5 lakh and lets users pay it off through EMIs at a lower interest rate. Once the amount has been payed-off, the app will let its lender apply for more sum. “We cater to unsecured consumer lending and at MoneyTap we have created a category called the credit line. It is  like an overdraft facility for customers for you and me and the loan account has been approved, a line of credit will be approved in your name without you having  to open a bank account with my partner bank,” says Anuj Khakkar, who co-founded the app with A Cocktail Of Tech Like any other fintech company, MoneyTap too uses large technological capabilities to cater to its burgeoning demand. According to Kakkar, the app gets close to 5 lakh instalments a month, however, only a certain amount of the consumers get the final approval. Since catering large volume of the request is hard, a chatbot has been leveraged by the Money tab to very the lending request and connects the banking system with the credit bureau to analyse the credit history of a person before the final approval of the money. Further, automated EMI reminders are sent to lenders when the amount is due. “We use all of the kind of technology. Since we have to rely on the huge database we are on AWS for our data storage capabilities.  For the backend, we have used JAVA and hibernate. While for our data and machine learning we use Python. Since we are on Andoird and IoS we are depended on HTML as well. So basically depends on what aspect we are working on and we are able to put all of these together and work with these,” Khakkar says when asked about the basic technological functionalities of MoneyTap. Road Ahead The lending market is quite a fierce Indian with international and national banks actively reaching out its customers for easy lending. Since banks do not lend smaller amounts and with broke millennials looking for an easier way out, apps like that of MonayTap have turned themselves into a lifesaver for most youngsters. Since its inception, MoneyTap has grown substantially and has been aggressively expanding its market reach. According to Kacker, the startup has witnessed a steady growth in the last couple of years and has clocked profit in the last quarter, despite the market being bullish. In 2018, it has said to obtained an annual profit of dual run rate of ₹500 crore. In addition, startup targetted a whopping ₹2,000 crore of loan disbursals by the middle of next this year and is said to have. With fintech companies seeing more opportunity to grow in the lending space, the company has to work to get its NBFC licence and is looking at a newer market for large-scale reach. Speaking about their future plan, Kakkar says, “Our expansion plans are very aggressive and presently we are present in 40 cities in India and we are hoping to make it 100-150 cities.  In terms of growth, we are growing in a double-digit number for our disbursement, we are looking at a triple digit number. “","excerpt":"With digital personal loans and instant paperless transactions touted to be the future of the banking sector, banks and startups are seizing the opportunity to make it big in the lending industry. A relatively younger startup which has been working in this field is MoneyTap, an instant personal loan lending app which can transfer as […]","categories":["Deep Tech"],"tags":["fintech AI"],"author_name":"Akshaya Asokan","publish_date":"2019-04-01T09:43:53","publication_year":"2019","word_count":606,"keywords":["Go","machine learning","AWS","AI","ML","Git","RAG","fintech AI","Python","R","Java"],"extracted_tech_keywords":["AI","machine learning","ML","RAG","AWS","Python","R","Go","Java","Git"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/how-moneytap-is-personalising-moneylending-with-artificial-intelligence\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":65104,"title":"Will Edge Computing Replace Cloud Computing?","content":"With the continuous adoption of cloud computing, mobile network, big data, and SDNs, the number of internet users has exploded. To catch up to the fast-changing trends with wireless connectivity and the internet, companies have strived to have more cloud adoption for different business operations. Cloud computing for years has offered a secure and controlled way of reliable remote access. However, it does lack the speed to process the gathered data from the edge of the network, which is a point where an enterprise owned network connects to a third party network. This challenge, in recent years, is addressed by edge computing as it offers better latency. Below we look at some differences between edge computing and cloud computing and whether it can replace cloud computing: What Is Edge Computing? It can be defined as the computational processing of data away from the centralised infrastructure and close to the logical edge of networks that are towards individual sources of data. IT can also be described as a distributed IT network architecture that enables mobile computing for locally produced data. So, instead of sending the data to the data centres, it decentralises computing power to ensure real-time processing without latency in addition to reduced storage and bandwidth requirements on networks. For a good example, consider autonomous cars. For any autonomous cars, road safety is the most critical aspect of driving. These autonomous cars in order to drive safely must be able to recognise obstacles or people in real-time and hit the brakes if there is something in its path. This will require visual processing information and making decisions quickly, which is done using edge computing. It takes around 100 milliseconds for the data to be transmitted between the car’s sensors and cloud data centres, this figure might seem fast, but in driving, it isn’t. This delay or the time for data transmission has a significant impact on the reactions of self-driving cars. Edge computing, in contrast, expedites the process and reduces the load on networks to help improve the autonomous car’s performance. Here, speed becomes one of the critical highlights of technology like edge computing. What Is Cloud Computing? We already use cloud services in our daily lives right from banking to playing online video games and watching Netflix. The cloud is a set of different hardware and software that work together to deliver many computing features to the end-user as online services. Migrating to cloud computing services instead of building in-house data centres reduces a company’s burden in terms of investing heavily on setup and maintenance. Cloud computing, with its extensive usage, holds many benefits like: Scalability: It allows companies to start with small deployments on clouds and expand efficiently as time passes. It also enables organisations to scale by adding extra resources as and when needed, which enables them to satisfy the rapidly changing customer demands. Maintenance: In the case of having in-house data centres, companies have to maintain the data centres by themselves. But with cloud computing, the cloud providers themselves maintain the systems. Cost-saving: Companies significantly reduce both operational and financial costs when it comes to expanding their computing capabilities. Resource pooling: Multiple users can share the same spaces and resources can be assigned and distributed as needed. Users from any location in the world have equal access to the services. Edge Computing vs Cloud Computing: Is Edge Better? With the rise of demand for real-time applications, the adoption of edge computing has significantly increased. Although centralised cloud computing systems provide ease of collaboration and access, they are remote from data sources. So, this requires data transmission, which causes delays in processing the information due to network latency. Today’s technology expects low latency and high speed to provide superior customer experience. Thus, one cannot afford to have cloud computing for every need. Although the cloud has some benefits, edge computing holds more advantages when compared: Speed All the raw information the cloud works with is through the edge devices, which collect and send data to it. The devices at the edge play a limited role of only sending raw information and receiving processed information from the cloud. But, the exchange that takes place can be used only with the applications where the time delay is permissible. So, edge computing provides better speed with low latency, giving the opportunity of interpreting input data close to the source. This provides more scope for applications that need real-time services. Low Connectivity Cost And Better Security Rather than filtering data at the central data centre, edge computing allows organisations to filter data at the source. This results in less transfer of companies’ sensitive information between devices and the cloud, which is better for the security of organisations and their customers. Reducing the movement of data also decreases the cost incurred as it eliminates the need for storage requirements. Better Data Management According to the figures, the connected devices are expected to reach around 20 billion by 2020. Edge computing takes an approach where it tackles some systems with special needs, freeing up cloud computing to work as a general-purpose platform. For example, the best route to a destination through a car’s GPS will come from analysing the surrounding areas rather than the data centres of the carmakers, which will be far away from the GPS. This results in lower dependence on cloud and helps applications perform better. Final Thoughts Although edge computing has more advantages compared to cloud computing, it would be wrong to say that edge would completely replace cloud computing. Choosing which one company will opt is better left as they will be making this decision according to their needs. Choosing between these two would be like choosing whether you want a luxury sports car or a more spacious family car. It depends on what one wants. Identify your needs, compare them against costs, assess and then choose which one is the best choice.","excerpt":"With the continuous adoption of cloud computing, mobile network, big data, and SDNs, the number of internet users has exploded. To catch up to the fast-changing trends with wireless connectivity and the internet, companies have strived to have more cloud adoption for different business operations. Cloud computing for years has offered a secure and controlled […]","categories":["AI Features"],"tags":["Cloud Computing","edge computing"],"author_name":"Sameer Balaganur","publish_date":"2020-05-13T12:00:00","publication_year":"2020","word_count":985,"keywords":["big data","Go","API","cloud computing","AI","Scala","RAG","GAN","Cloud Computing","edge computing","R"],"extracted_tech_keywords":["AI","RAG","cloud computing","edge computing","R","Go","Scala","API","big data","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/will-edge-computing-replace-cloud-computing\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":47150,"title":"9 Fraud Analysts Jobs In India To Apply Right Away","content":"Fraud Analysts determine existing fraud trends and work towards protecting the organisation and the end-users from any future risks. From managing early risk detection to minimizing losses, Fraud Analyst job roles are on a rise. In this article, we list down 9 Fraud Analyst jobs one can apply right away. 1| Fraud Case Analyst @ JP Morgan As a Fraud Case Analyst at JP Morgan, you will collaborate closely with front-line Operations to identify any emerging fraud trends, large scale compromises and assess vulnerable customer segments, size the impact of such incidences and share fraud mitigation opportunities with risk management and point-of-sale fraud prevention strategy partners to help build fraud prevention strategies. Prerequisites: Minimum two years of experience required in Plastic fraud prevention (FCC or DCFP) and minimum of one year experience in Retail fraud prevention (Deposit Review, EMM or FHL) Ability to interpret advanced Excel sheets and will require learning basic SQL and SAS on the job to extract and analyse data. Ability to develop effective working relationships, internal and external to FCPS Strong analytical, interpretive and problem-solving skills, which require interpreting large amounts of data and its impact in both operational and financial areas. Apply here 2| Fraud Analyst @ Rapido The position of Fraud Analyst at Rapido involves fast-paced developments in fraud patterns, managing the potentially high-pressure business escalations that result from these behaviors, and timely investigation, development, and implementation of data-driven solutions. It involves investigating for any fraudulent activity by the riders and identifying any patterns around them. Prerequisites: Knowledge of data collection and analysis, statistics, visual presentation methods, and process mapping and modeling. Minimum 2 years of work experience in a relevant field, preferably within the areas of Risk Management, Fraud Prevention, or Decision Management Familiarity with tools such as Excel, google spreadsheets and SQL. Strong logical thinking and problem-solving skills. It requires the ability to synthesize information and generalize the pattern. Apply here 3| Fraud Analyst @ Hewlett Packard Enterprise (HPE) Multinational company Hewlett Packard Enterprise (HPE) is hiring Fraud Analyst who will provide independent analysis for complex, important cases and issues, analyses company transnational data, manage physical and digital evidence, delivers investigative reports incorporating all analysis, findings, and evidence, conducts open-source research, prepares intelligence reports in support of investigations and other such related works. Prerequisites: First-level university degree in Criminal Justice, Security, Law Enforcement or a related field of study 2-3 years of security experience, including a mix of law enforcement, military, intelligence and\/or corporate experience. Experience in multiple security disciplines (e.g., operations, executive protection, systems, etc). Advanced analytical skills and ability to identify complex security issues and customise, implement\/execute innovative processes and solutions. Apply here 4| Fraud Analyst @ Standard Chartered Bank Ltd Standard Chartered is hiring Fraud Analyst who will help in minimising the fraud losses for Cards \/products by actioning alerts on the Fraud Detection to minimise fraud loss, identify emerging fraud trends to prevent customer\/bank ending in financial loss, maintain a high level of customer satisfaction by handling customer service issues without compromising on the risk aspect and other such. Prerequisites: Good understanding of systems\/procedures Understanding of the risk involved in various transaction types and cardholder profiles Bachelor or Masters Degree with Good Communication and Analytical Skills Apply here 5| Risk Investigator @ PhonePe PhonePe is hiring Fraud Risk Investigator in its Fraud Risk Operations team in order to improve the service process experience by mitigating risk associated with unauthorised card usage. The candidate will perform tasks like reviewing and investigating complex blocked accounts to detect unauthorised activity, determining appropriate next steps, identifies innovative ideas to improve the customer experience and achieves or exceeds performance and productivity metric goals, etc. Prerequisites: The ideal candidate will have excellent investigative skills, extensive knowledge of standard methodologies for fraud management and prevention, experience with account reviews, SAR-AML, and familiarity with credit card processing. Experience with SQL is an advantage Experienced with SAR-AML is an advantage Microsoft Excel experience is a must Apply here 6| Fraud Risk Analyst @ Paypal India Pvt. Ltd. In the Fraud Risk Analyst role, you will have to work closely with business partners to determine how to design analysis and presentation approaches that will significantly improve the ability to understand and address emerging business issues. Prerequisites: Ability to work with large volumes of data; extract and manipulate large datasets using standard tools such as SQL, SAS and Python, tableau, Qlikview. Strong knowledge of Statistical and Optimization concepts Ability to frame a business question in analytics terms and structure the required analyses. Experience in SQL, SAS, Excel, and experience with Teradata and Oracle Apply here 7| Risk Analyst @ Swiss Re Swiss Re is a leading wholesale provider of reinsurance, insurance and other insurance-based forms of risk transfer. The company is hiring Risk Analyst for its Bangalore office who will write\/review P&C analysis of change comments by risk category, adherence to standards of audibility, challenge controls and suggestions to improve and stabilize the process, work on testing the new features along with Risk IT and support them on formulating and testing change requests, and other such. Prerequisites: Database knowledge and programming skills (SQL, Matlab etc) are required Good stakeholder management skills and co-ordinating with various teams within and outside Risk Management for the aggregation and understanding of risk data. Strong analytical skills as well as high interest in quantitative tasks. Apply here 8| Fraud Prevention Analyst @ Cisco Systems In this role, you will identify suspicious user registrations, malicious partner registrations, fraudulent activities & patterns of fraud in multiple external & internal-facing applications, analyse data and deter fraud inquisitive by nature, collect and assemble data intelligence to identify and mitigate potential threats and attacks, define and monitor key fraud indicators and alerts using technical data analysis to implement and adjust strategies and reduce exposure and other such relevant activities. Prerequisites: 2 years of Fraud Prevention experience in an enterprise environment Experience using Excel for data analysis and visualization Experience writing SQL statements to aggregate and analyze data Experience in a scripting language, Python preferred Familiarity with fraud prevention policies, operations, and reporting requirements Apply here 9| Risk Investigator @ Timber Solutions As a Risk Investigator, you will have to review and investigate complex blocked accounts to detect unauthorised activity and determine the next steps, contact financial institutions and customer cardholders to verify transaction authenticity over the phone, identifies innovative ideas to improve the customer experience and other such. Prerequisites: Knowledge of fraud processes with extensive investigative & fraud pattern recognition experience Proven decision-making, problem-solving skills, and good judgment Finds opportunities for operational improvements and suggests improvements Microsoft Excel experience is a must Experience with SQL and SAR-AML is an advantage. Apply here","excerpt":"Fraud Analysts determine existing fraud trends and work towards protecting the organisation and the end-users from any future risks. From managing early risk detection to minimizing losses, Fraud Analyst job roles are on a rise.   In this article, we list down 9 Fraud Analyst jobs one can apply right away. 1| Fraud Case Analyst @ […]","categories":["AI Hirings"],"tags":["Data loss prevention"],"author_name":"Ambika Choudhury","publish_date":"2019-10-11T14:00:18","publication_year":"2019","word_count":1113,"keywords":["Go","AI","ML","Scala","Git","Python","Data loss prevention","analytics","SQL","R","fraud detection"],"extracted_tech_keywords":["AI","ML","analytics","fraud detection","Python","R","SQL","Go","Scala","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-hiring\/9-fraud-analysts-jobs-in-india-to-apply-right-away\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10044644,"title":"Now AI Gives Paralysed Man Ability To Communicate In This Groundbreaking Research","content":"The University of California, San Francisco (UCSF) researchers have leveraged artificial intelligence to give a paralysed man the ability to communicate by translating his brain signals into computer-generated writing. The study was published in The New England Journal of Medicine. People with speech impediments use touchscreens, keyboards or speech-generating devices to communicate. But assistive technologies are not much of a help for paralysed people. In the US alone, 2.8 million people incur traumatic brain injury a year. Such injuries may result in hearing loss and vestibular and central auditory problems. A ray of hope The Joan and Sanford Weill Chair of neurological surgery and senior author, Edward Chang and his team worked with a brainstem stroke survivor who had lost the ability to speak in a car accident in 2003. The researchers were not sure if his brain retained neural activity linked to speech. To track his brain signals, a neuroprosthetic device consisting of electrodes was positioned on the left side of the brain, across several regions known for speech processing. A question was displayed for the participant, and the device recorded brain activity while he attempted to speak and reply. Meanwhile, a computer algorithm translated the brain activity patterns into words and sentences in real-time. “To our knowledge, this is the ‘first successful demonstration of direct decoding’ of full words from the brain activity of someone who is paralysed and cannot speak,” said Chang. GPUs to the rescue David Moses, one of the lead authors of the study and a postdoctoral engineer in the Chang lab, in a press release, said, “Our models needed to learn the mapping between ‘complex brain’ activity patterns and ‘intended speech.'” Decoding the responses from the subject’s brain activity, the team created speech-detection and word classification models. Leveraging the cuDNN-accelerated TensorFlow framework and 32 NVIDIA V100 Tensor Core GPUs, the researchers trained, fine-tuned, and evaluated the models. The study co-lead Sean Metzger said utilising neural networks was essential to getting the classification and detection performance, and the final product was the results of lots of experimentation. “The ‘GPUs’ helped us make changes, monitor progress, and understand our dataset,” he added. A twist in the tale The UCSF team embarked on 50 training sessions, generating over 1,000 words decoded with up to 93 percent accuracy, and a median rate of 75 percent, at the rate of 18 words per minute. The latest study was built on previous work by Chang and his team, where they had developed a deep learning method for decoding and converting brain signals. Compared to the latest work, the participants in the previous study were able to speak. Earlier, Facebook pulled the plug on invasive brain-computer interface (BCI) research. Facebook Reality Labs (FRL) was established in 2017, and the BCI project harboured an ambitious long-term goal: to develop a silent, non-invasive speech interface activated by thinking. Facebook’s Project Steno was kick-started in 2019 in the Chang Lab at UCSF. Facebook clarified it has no interest in developing products that require implanted electrodes. Instead, Facebook is looking to focus on wrist wear or wearable technology to scan brain signals. “We have decided to focus our immediate efforts on a different neural interface approach that has a nearer-term path to market: wrist-based devices powered by electromyography,” said Facebook. Wrapping up The new approach holds the potential to advance existing methods of assisted communication, improving the ability to communicate and improving quality of life in paralysed patients with speech disorders. Stanford University Professor Maneesh Agrawala is working on a device to give people their voices back after undergoing laryngectomy surgery “We plan to record a patient’s voice ‘before the surgery’ and then use that ‘pre-surgery recording’ to convert their electrolarynx voice back into their pre-surgery voice,” said Agrawala.","excerpt":"In the US alone, 2.8 million people incur traumatic brain injury a year. Such injuries may result in hearing loss and vestibular and central auditory problems.","categories":["AI Features"],"tags":["healthcare ai","Machine Learning","Machine Learning Latest","NVIDIA GPU","Tensorflow","video processing ai"],"author_name":"Amit Naik","publish_date":"2021-07-29T11:00:00","publication_year":"2021","word_count":624,"keywords":["Go","NVIDIA GPU","healthcare ai","artificial intelligence","AI","neural network","Machine Learning Latest","Machine Learning","video processing ai","RAG","Ray","deep learning","ViT","TensorFlow","R","Tensorflow"],"extracted_tech_keywords":["AI","artificial intelligence","deep learning","neural network","Ray","TensorFlow","RAG","R","Go","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/now-ai-gives-paralysed-man-ability-to-communicate-in-this-groundbreaking-research\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10115254,"title":"Indian Govt Endorses IndiaAI Mission with ₹10,371 Crore Funding","content":"The Indian Cabinet has approved a significant initiative to bolster the country’s artificial intelligence (AI) capabilities by establishing a Public AI Compute Infrastructure with over 10,000 GPUs. This move is part of the comprehensive IndiaAI Mission, which has been greenlit with a budget of ₹10,371.92 crore (approximately $1.25 billion). The mission, chaired by Prime Minister Narendra Modi, is designed to catalyse AI innovation and position India as a global AI hub. The infrastructure will be developed through public-private partnerships and aims to meet the growing demands of India’s burgeoning AI startups and research ecosystem. Union Minister of State for Electronics and IT, Shri Rajeev Chandrasekhar, expressed his gratitude for the approval, stating, “I am grateful and thankful to the Hon’ble PM for the approval of more than 10,000 crores for the India AI program, which will catalyse India’s AI ecosystem and position it as a force shaping the future of AI for India and for the world”. The mission includes several key components such as the IndiaAI Innovation Centre, which will focus on developing indigenous Large Multimodal Models (LMMs), and the IndiaAI Datasets Platform, which will provide access to quality non-personal datasets for AI innovation. Additionally, the IndiaAI FutureSkills program aims to increase AI courses at various academic levels to mitigate barriers to entry into AI programs. The IndiaAI Startup Financing pillar will support deep-tech AI startups by providing streamlined access to funding, from product development to commercialisation. Moreover, the Safe & Trusted AI pillar will implement responsible AI projects, including the development of indigenous tools and frameworks to ensure ethical AI development and deployment. This ambitious mission is expected to drive innovation, create job opportunities, and enhance India’s technological autonomy on the global stage. It aligns with Prime Minister Narendra Modi’s vision to harness AI for the country’s digital economy and use it for social good","excerpt":"The approved initiative will bolster the country’s AI capabilities by establishing a Public AI Compute Infrastructure with over 10,000 GPUs.","categories":["AI News"],"tags":["Fund Raising","Startups"],"author_name":"Shyam Nandan Upadhyay","publish_date":"2024-03-08T17:16:06","publication_year":"2024","word_count":307,"keywords":["Go","API","artificial intelligence","AI","Fund Raising","ML","Git","responsible AI","Aim","Rust","Startups","R"],"extracted_tech_keywords":["AI","artificial intelligence","ML","Aim","R","Go","Rust","Git","API","responsible AI"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/indian-govt-endorses-indiaai-mission-with-%e2%82%b910371-crore-funding\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10118707,"title":"How Databricks is Enabling Agriculture’s Data Revolution with UPL","content":"At their Data Intelligence Day, Databricks recently announced its partnership with Indian customers such as UPL, Air India, Aditya Birla Fashion, Freshworks, InMobi, Meesho, Myntra, Parle, and many others. This will help leverage the Databricks Data Intelligence Platform to drive business innovation, optimise operations, and enhance decision-making. One among these is UPL Limited, a multinational company headquartered in Mumbai, which provides sustainable agriculture products and solutions in more than 138 countries. The company’s UPL Agri-Tech Platform, built in collaboration with Microsoft, uses AI and machine learning to provide crop management, pest control, and nutrient application recommendations. With a market cap of $4.37 billion as of April 2024, the company’s product portfolio includes crop protection, seed treatment, post-harvest solutions, and other agricultural inputs, with a presence in markets such as India, the United States, Brazil, and Europe. Having experienced massive growth in recent years, the partnership between UPL and Databricks becomes critical for the former. “In 2022, we were at the end of a massive growth period. We grew 2x in about three to four years,” Vijay Balakrishnan, chief data & analytics officer at UPL Group, told AIM on the sidelines of the event in Bengaluru on March 22. This growth triggered the need for UPL to leverage data more effectively to make informed decisions and drive further growth. This is when the company found solace in Databricks. UPL, the largest manufacturer and distributor of agrochemicals in India with a market share of around 20%, has actively started using Databricks’ Unity Catalog, which democratises data across the organisation. Using Databricks to Scale AI Solutions Across 50 Countries By leveraging Databricks, UPL also aims to enable AI solutions across 20 countries, with the potential to scale up to 50 countries in the future. “We aim to democratise access to AI, enabling citizen data scientists and proficient digital business leaders to leverage and operate the AI solutions we’ve established, whether for commercial endeavours, supply chain optimisation, or other applications,” elaborated Vijay. For instance, they conduct extensive farmer outreach programmes, providing training and advisory services to help farmers adopt best agricultural practices. The company has also found Databricks’ serverless SQL engine to be efficient, especially in instances where the speed of data pooling is critical. “We have seen it to be quite efficient today because there are a lot of instances where the speed of the data pool is very critical for us,” noted Vijay. As UPL continues to explore and leverage Databricks’ generative AI and lakehouse capabilities, the partnership between the two companies is poised to drive innovation and transform the agriculture industry through data-driven solutions. UPL has started exploring Databricks’ Genie Data Room, which provides a drag-and-drop interface for generative AI. “The Genie Data Room actually makes it almost drag and drop. All we did was create a data room with the data that we’ve already pulled into Databricks and started working,” According to Vijay, UPL is also focusing on prompt engineering and even considering using AI for prompt engineering itself. He also emphasised the importance of governance, particularly as metadata becomes complex. Leveraging Databricks for Demand Forecasting One of the key areas where UPL is leveraging Databricks’ capabilities is demand forecasting. “Demand forecasting has its own challenges in the agriculture sector because many external variables play a role,” Vijay said. “For instance, the last 10-12 months have been an El Nino drought situation in most of the Northern Hemisphere. But that has a reverse effect in countries in the Southern Hemisphere,” he added. These external factors, along with the constantly evolving product portfolio, make demand forecasting a complex task. Timing is also crucial in the agriculture industry, with many products requiring application within a specific three-day window. “A lot of our products have to be timed in such a way that you’ve got a three-day window within which a certain product has to be applied. And if you don’t predict that three-day window and apply, you don’t have a play,” explained Vijay. For instance, if UPL gauges demand incorrectly, they run the risk of spoiled products which may be stuck in transportation – out of cold storage or in transit for extended periods. To address these challenges, UPL has developed a demand forecasting model that considers various factors. Databricks’ platform provides UPL with a unified and governed data management and analytics environment, enabling more accurate demand forecasts. Challenges to Solve Going Forward Vijay also highlighted the challenges faced by the company in data ingestion and orchestration. “The way we look at Databricks is we would love for them to be all in analytics for us. Now, there are two areas where it is a little challenging today. One is on the, let’s say, the data collection part or ingestion, but more ingestion and collection part, especially in complex source ecosystems like SAP,” said Vijay. SAP poses its own set of challenges due to contractual aspects that need to be considered before pulling data, whether in a CDC (Change Data Capture) fashion or a batch mode. While Databricks has a good partner ecosystem, Vijay noted that there is still a gap in this area. “That part, while there are good partners, an ecosystem with Databricks, there’s a little bit of a gap, at least from my perspective. And I brought it up on a couple of forums earlier,” he said. Bhasin acknowledged these challenges and emphasised that while Databricks wants to be everything to everybody, it is an evolving platform, and it’s a matter of different priorities at different points. “We take pride in our commitment to identifying common pain points and developing the necessary capabilities to address them. This approach is integral to the evolution of our platform. By actively listening to our customers and discerning prevalent pain points, we prioritise investments that yield meaningful solutions,” he explained. Databricks has been investing disproportionately in R&D to stay at the cutting edge of technology. Bhasin acknowledged the feedback from UPL and assured that Databricks is working on addressing these challenges. Conclusively, as UPL and Databricks continue to collaborate and co-innovate, they aim to address the data ingestion and orchestration challenges in the agriculture industry, unlocking new opportunities for growth and efficiency.","excerpt":"Databricks is aiding UPL’s sustainable agriculture products and solutions in 138 countries with capabilities like demand forecasting.","categories":["Deep Tech"],"tags":["Databricks"],"author_name":"Shyam Nandan Upadhyay","publish_date":"2024-04-19T16:00:00","publication_year":"2024","word_count":1029,"keywords":["machine learning","AI","serverless","RAG","Aim","prompt engineering","generative AI","analytics","R","Databricks"],"extracted_tech_keywords":["AI","machine learning","analytics","generative AI","Aim","RAG","prompt engineering","serverless","Databricks","R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/how-databricks-is-enabling-agricultures-data-revolution-with-upl\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":24264,"title":"10 Best Libraries For Implementing Machine Learning In Java","content":"Skills in machine learning and deep learning are one of the hottest ones in the new tech world right now, and companies are constantly on a lookout for programmers with good knowledge of ML. Java is definitely one of the most popular languages after Python and has become a norm for implementing ML algorithm these days. Some of the many advantages of learning Java include acceptance by people in the ML community, marketability, easy maintenance and readability, among others. Here we list down 10 best machine learning libraries for Java, which have been compiled based on their popularity level from various websites, blogs and forums. (This list is in alphabetical order) 1. ADAMS Short for Advanced Data mining And Machine learning System, ADAMS follows the philosophy of “less is more”. A novel and flexible workflow engine, ADAMS is aimed at quickly building and maintaining real-world workflows which are usually complex in nature. It has been released under GPLv3. Instead of letting the user place operators or “actors” on a canvas and then manually connecting input and output, ADAMS uses a tree-like structure to control how data flows in the workflow. This means that there are no explicit connections that are necessary. You can find ADAMS here. 2. Deeplearning4j This programming library written for Java offers a computing framework with a wide support for deep learning algorithms. Considered as one of the most innovative contributors to the Java ecosystem, it is an open source distributed deep learning library brought together with an intention to bring deep neural networks and deep reinforcement learning together for business environments. It usually serves as a DIY tool for JAVA and has the ability to handle virtually limitless concurrent tasks. It is extremely useful for identifying patterns and sentiment in speech, sound and text. It can also be used for detection of anomalies in time series data like financial transactions, clearly showcasing that it is designed to be used business environments rather than as a research tool. You can find Deeplearning4j here. 3. ELKI ELKI, short for Environment for Developing KDD-Applications Supported by Index-structure, is also an open source data mining software written in Java. Designed for researchers and students, it provides a large number of highly configurable algorithm parameters. It is popularly used by graduate students who are looking to make sense of their datasets. Developed for use in research and teaching, it is a knowledge discovery in databases (KDD) software framework. It aims at developing and evaluating advanced data mining algorithms and their interaction with database index structures. ELKI also allows arbitrary data types, file formats, or distance or similarity measures. You can find ELKI here. 4. JavaML It is a Java API with a collection of machine learning and data mining algorithms implemented in Java. It is aimed to be readily used by both software developers and research scientists. The interfaces for each of algorithm is kept simple and easy to use. There is no GUI but clear interfaces for each type of algorithms. Compared to other clustering algorithms it is straightforward and allows an ease of implementation of new algorithm. At most times, the implementation of algorithms is clearly written and properly documented, hence can be used as a reference. The library is written in Java. You can find it here. 5. JSAT The Java Statistical Analysis Tool, is a Java library for machine learning to get quickly started with ML problems. Available for use under the GPL3, part of the library is for self education. All code is self-contained, with no external dependencies. It has one of the largest collections of algorithms available in any framework. It is usually considered faster than other Java libraries, offering high performance and flexibility. Almost all of the algorithms are independently implemented using an object-oriented framework. It is mainly used for research and specialised needs. You can find JSAT here. 6. Mahout It is an ML framework with built-in algorithms to help people create their own algorithm implementations. Apache Mahout is a distributed linear algebra framework which is designed to let mathematicians, statisticians, data scientists and analytics professionals implement their own algorithm. This scalable ML library provides a rich set of components that lets you construct a customised recommendation system from a selection of algorithms. Offering high performance, scalability and flexibility, this ML library for Java is designed to be enterprise-ready. You can find it here. 7. MALLET Short for MAchine Learning for LanguagE Toolkit, MALLET is an integrated collection of Java code used for areas like statistical NLP, cluster analysis, topic modelling, document classification and other ML applications to text. In other words, it is a Java ML toolkit for textual documents. It was developed by Andrew McCallum and students from UMASS and UPenn and supports a wide variety of algorithms such as maximum entropy, decision tree and naïve bayes. You can find MALLET here. 8. Massive Online Analysis MOA is an open source software used specifically used for machine learning and data mining on data streams in real time. It is developed in Java and can also be easily used with Weka. The collection of ML algorithms and tools is extensively used in the data science community for regression, clustering, classification, recommender systems, among others. It can be useful for large datasets including data produced by IoT devices. It consists of large collections of ML algorithms designed for large scale machine learning, dealing with concept drift. It is available here. 9. RapidMiner Developed at Technical University of Dortmund, Germany, RapidMiner offers a suit of products allowing data analysts to build new data mining processes, set up predictive analysis, and more. Consisting of machine learning libraries and algorithms, it offers easy to construct, simple and understandable machine learning workflow. It allows loading data, features selection and cleaning along with a GUI and a Java API for developing your own applications. It provides data handling, visualisation and modelling with machine learning algorithms. The list of products includes RapidMiner Studio, RapidMiner Server, RapidMiner Radoop, and RapidMiner Streams. It is available here. 10. Weka Weka is the most popular pick as a machine learning library for JAVA for data mining tasks, where algorithms can either be applied directly to a dataset or called from your own Java code. It contains tools for functions such as classification, regression, clustering, association rules, and visualisation. This free, portable and easy-to-use library supports clustering, time series prediction, feature selection, anomaly detection and more. Short for Waikato Environment for Knowledge Analysis, it can be defined as a collection of tools and algorithms for data analysis and predictive modelling along with graphical user interfaces. You can find it here.","excerpt":"Skills in machine learning and deep learning are one of the hottest ones in the new tech world right now, and companies are constantly on a lookout for programmers with good knowledge of ML. Java is definitely one of the most popular languages after Python and has become a norm for implementing ML algorithm these […]","categories":["AI Trends"],"tags":["clustering algorithms","Java Machine Learning","machine learning libraries","ml libraries"],"author_name":"Srishti Deoras","publish_date":"2018-05-04T12:58:10","publication_year":"2018","word_count":1106,"keywords":["data science","machine learning","AI","neural network","ML","ml libraries","NLP","Java Machine Learning","machine learning libraries","clustering algorithms","deep learning","analytics","Aim","anomaly detection"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","neural network","NLP","data science","analytics","Aim","anomaly detection"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/10-best-libraries-for-implementing-machine-learning-in-java\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":27250,"title":"Robotic Process Automation And Artificial Intelligence Don’t Have Many Things In Common","content":"Researchers across the globe are trying to inculcate technologies such as robotics and AI in their workflows to optimise and automate them. In such processes, Robotic Process Automation (RPA) is one of the most popular terminologies and is often sought-after for handling operational tasks with least manual intervention. In other terms, it is a software that automates low-level tasks. However, the two terminologies AI & RPA are used interchangeably. It is important to realise that both RPA and AI facilitate a common goal of Intelligent Automation. While RPA is often picturized as a software robot mimicking human actions, AI is the simulation of human intelligence by machines. In this article, we list down 5 major differences between the RPA & AI. Overall Approach RPA: To put it simply, robotic process automation is intended to automate repetitive, low-level tasks by mimicking human actions and behaviour. It may be suitable for mundane tasks that require little or no human interventions such as entering purchase invoices in an ERP system, setting up a new customer account and others. RPA is a preconfigured software that works on a predefined program to complete autonomous execution of a combination of processes, activities and tasks. AI: AI on the other hand is used for automating tasks in a much efficient manner. It is capable of delivering better results as it has the capability of rewriting themselves in response to their environment. It can make decisions and predictions either based on rules (as RPA) or based on numeric parameters achieved via machine learning. In a nutshell, AI is a combination of cognitive automation, ML, reasoning, MLP, analysis and much more. IBM Watson is an example of a cognitive system. A key difference between AI & RPA is that the latter is used for running rule-based processes and requires structured inputs and AI is leveraged for unstructured inputs. Another key difference is that RPA software is trained exactly for the task, for example SAP’s Process Automation software. Training Method RPA: RPA is rule-based. It works on a set of instructions, or a set of rules and performing tasks accordingly. It uses a set of statements or steps to define a repetitive activity which it does by using algorithms to automate it. It is the systematic and organised inputs that RPA depends on to deliver desired results. As it is rule-based, it has no inbuilt intelligence and is prone to errors given its limited scope of intelligence. For instance, in a bill management system, if it can automatically record printed bills or automatically communicate customer issues, then RPA may be playing a key role. AI: AI is much more than just rules. It is data-hungry and requires a large amount of data to train.. The data could be anything from customer information to images that helps machine understand the underlying concept. AI systems have the capability of learning over and over again using mathematical and statistical methods. Apart from ML, NLP is also used to deliver AI capabilities, which allows a machine to acquire an understanding of the human language. Process-Driven Vs Data-Driven RPA: It is highly process driven, meaning, it is all about automating repetitive and rule-based processes which typically requires interaction with multiple, disparate IT systems. Every activity needs to explicitly programmed, exposing it to the risk of delivering effective performance in many cases. Moreover, RPA cannot work on unstructured information. AI: AI, on the other hand, is all about good quality data. It would be unfair to say that AI is all about ‘thinking’ as opposed to RPA which is all about ‘doing’ a task. AI comes in handy when there is loads of unstructured data to deal with, and can easily manage variability in data to get better with time, based on its own experiences. For example, image recognition, text recognition or search are some of the most mature AI applications seen in businesses today. Learning And Thinking Capabilities RPA: As we mentioned earlier, it is one of the key differences between the two. While RPA are ‘dumb’, AI have ‘self-learning’ capabilities. Robots or machineries activated with RPA will do exactly what you ask them do and in the same way all over again, every time. AI: AI on the other hand is expected to perform a judgement based processing. This means that AI trained programs will act based on their learning from past data and trends. It can manage and understand patterns and trend over time. The bottom line is AI is an excellent self-learner and is good in capturing information such as vision recognition, sound recognition, search, data analysis and others. In short AI is where machines are trained to think like humans and possess the ability of rationalising and take actions accordingly. The Human Involvement RPA: RPA is practically a software that reduces human efforts and compliments their work. As RPA mimics the steps followed by humans, it is often programmed to relieve human workforce from mundane activities, so they could focus on other important activities to accelerate business growth. Having said that, it often requires human intervention to keep the robotic processes constantly updated. AI: AI on the other hand has the capability of eliminating human effort to a significant extent. AI comes to a rescue when RPA fails and may not demand constant human intervention after the initial process of setting it up. Conclusion If you are looking to opt for either of the two processes, it is always wise to first analyse the nature of your process—decreasing turnaround time, saving cost, accelerating process, among others—and then decide on opting RPA or AI or a combination of both to achieve an extremely powerful result.","excerpt":"Researchers across the globe are trying to inculcate technologies such as robotics and AI in their workflows to optimise and automate them. In such processes, Robotic Process Automation (RPA) is one of the most popular terminologies and is often sought-after for handling operational tasks with least manual intervention. In other terms, it is a software […]","categories":["AI Features"],"tags":["process automation","Robotic Process Automation"],"author_name":"Srishti Deoras","publish_date":"2018-08-13T08:18:46","publication_year":"2018","word_count":943,"keywords":["Go","machine learning","AI","ML","image recognition","RAG","NLP","Robotic Process Automation","ViT","GAN","process automation","R"],"extracted_tech_keywords":["AI","machine learning","ML","NLP","RAG","image recognition","R","Go","GAN","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/robotic-process-automation-and-artificial-intelligence-dont-have-many-things-in-common\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10005134,"title":"Why Should You Patent Your AI Inventions","content":"The ease with which you shop at Amazon and scroll though different products that are customised specific to your taste is the result of AI technology that analyses and predicts your shopping behavior. There are countless AI start-ups who are founded for the primary reason of using AI to make the world a better place. The challenges that the world faces today – primarily climate change – is a huge motivator for AI innovation. Every AI idea is bound to help us face such challenges. In 2018, venture funding for AI grew to about 9.3 billion dollars in U.S. alone. This statistical data shows how a great AI, or a machine learning invention is hugely backed-up by Venture Capitalists. It’s safe to say that the AI start-up trend has only just begun! How AI Start-ups Are Changing The World Ease in Functioning of Hazardous Jobs The introduction of AI changed the career paths for us as well as future generations. While a few argue how it’s taking away human jobs, the positive impact it has made, especially in the case of hazardous jobs cannot be ignored. Take the removal of explosives – mines and bombs, as an example. These are life-threatening jobs that are better dealt with AI-powered robots. Similarly, coal mining – one of the most dangerous jobs in the world, now uses AI-powered robots to extract, transfer and secure coal from underground mines. The health risks associated with coal mining include exposure to toxic chemicals and fumes. Such jobs are made safer by AI, thereby preventing casualties. Human Language Translation Establishing a universal language translation warrants a super complex technology. The switch to neural networks, also known as deep learning, results in highly accurate machine translation, without human intervention. Google, Amazon and Bing all use neural machine translation as part of their translation engines. These automatic translation technologies are protected by thousands of AI patents. With plenty of innovation being harboured in AI, the need for protecting these inventions through patents becomes crucial. Google’s automated language translation system that was patented back in 1997, replaced human language translation system (manual). The AI tool translates spoken words of a first language into a machine recognizable code, and then further translates the machine recognizable code into a second language, in order to produce an audible translation of the spoken words from the first language to the second language. Relevance Of AI Patents This patented language translation system is just one of 4000 patents and applications in human language translation alone! Let’s have a look at some of the common AI inventions that we apply to our day-to-day life: Digital assistants like Google Now, Amazon Alexa and Apple’s Siri rely on machine learning as well as pattern recognition not only for real-time translation of languages but also for completing commands initiated by users.Netflix and Amazon Prime rely on AI-based engines to serve movie and TV show recommendations to users based on past watching patterns.Facebook uses AI to detect bots and identify fake users, as well as suggest photo tags and populate a user’s newsfeed.Cab aggregators like Uber rely on Machine Learning to determine the price of a ride and use this to determine when to initiate ‘Surge Pricing’.Email services like Gmail rely on AI to filter out spam and phishing messages from your inbox.Ecommerce platform Amazon uses AI and Machine Learning to look for the most relevant products based on our searches and also suggest similar products that are pretty close to what we are searching for. The above AI examples are protected by patents to safeguard intellectual property and to maintain a competitive edge. AI Patent Trend Around The World The graph below shows US patent activity in the last 20 years. The past decade though has seen a gradual increase in the number of patent families in AI. AI Patent Trend In India AI patent trend in India is significantly lower than the rest of the world but sees a lot of potential with opportunities unfolding in diverse sectors such as finance & banking, e-commerce, healthcare, agriculture, autonomous driving and e-governance. Listed below is the patent activity in India from 2001 onwards. The past decade has witnessed a gradual increase in the number of patent families with significant activity in the last five years: Now, let’s also look at the top owners of AI patents in India as per the patent portfolio of each company. Tata Group leads the race with the maximum number of patent families, followed by Wipro and HCL Technologies. To conclude, AI has the potential to change the world. Their relevance and importance today is indisputable. As you move forward with your AI based inventions consider the disruptive impact that it brings to the industry and whether you would like to publicize it. Based on the answer to these two questions, decide to protect your ideas and build your AI patent portfolio to generate revenue from your unique inventions.","excerpt":"The ease with which you shop at Amazon and scroll though different products that are customised specific to your taste is the result of AI technology that analyses and predicts your shopping behavior. There are countless AI start-ups who are founded for the primary reason of using AI to make the world a better place. […]","categories":["AI Features"],"tags":["AI Patent"],"author_name":"Prateek Mohunta","publish_date":"2020-08-20T10:00:00","publication_year":"2020","word_count":824,"keywords":["Go","API","machine learning","AI","neural network","innovation","AI Patent","Git","deep learning","ViT","R"],"extracted_tech_keywords":["AI","machine learning","deep learning","neural network","R","Go","Git","API","ViT","innovation"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/why-should-you-patent-your-ai-inventions\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":2387,"title":"4C Optimization for Digital Marketing practice","content":"Digital marketing is a complex ecosystem and with technology it seems to be getting tougher with time. Making this digital marketing practise to help drive business objectives, there is a need for a structured framework. This should be built upon the data and should be able to help to get the best of return. This approach, in general known as optimization, is the discipline of applying advanced analytical methods to make better decisions. Under this framework, business goals are explicitly defined and then decisions are calibrated to optimize those goals. In the digital marketing area, this is known as digital marketing optimization (DMO). 4C-Optimization is the unique approach proposed by Datapub consulting, where the goal is to help with maximization of conversions, revenues, profits, cost reduction or any combination thereof. The pillars of 4C optimization are, 1)     Crowd optimization 2)     Channel optimization 3)     Creative optimization 4)     Conversion optimization 1) Crowd optimization: Crowd optimization is the process of identifying the group of audience which are can be the potential customers based up their shown interest in the product. This will be calculated based on the recent browsing behaviour, the amount of time spent on researching the product, demographic details along with their response rate to recent marketing communication if they been exposure recently. Statistical techniques used to calculate this are sampling, extrapolation, panel data performance, database marketing etc. 2) Channel optimization: Channel optimization is the process of optimization the traffic sources and identifying the best performing media channel or publisher or keyword group etc which are part of the digital marketing campaign plan. The need for this optimization will be calculated based on the page rank, placement position, view ability of the ad, CTR and dwell rates. The Statistical techniques used for this optimization are mediamix models, variance models, historical and benchmark performance models etc. 3) Creative optimization: Creative optimization is the process of analyzing the effectiveness of the creative for the target group. There are experimental or predictive methods are used to identify the optimum creative for a given ad placement and determined to use it for further impressions or clicks or conversions depending on the choice of the metric. The effectiveness will be based on the CTR, dwell rate, interaction rates for the response to elements of creative. i.e call to action, images, copy. The statistical techniques used are AB testing, MVT testing etc. In the industry today, there are vendors who are provide dynamic creative optimization based on the historical 4) Conversion Optimization: Conversion optimization is the process is the method of creating an experience for a website or visitor with the goal of increasing the percentage of visitors that convert into customers. This is calculated based on the page load times, length of consumer journey, exit rate from conversions page, page depth, total conversions\/visits etc. Statistical techniques used are A\/B testing, usability test etc. This 4C optimization is an easy to implement and maintain approach and avoids the biggest hurdle of data management that is information being in silos. For more detailed information contact Datapub team. © Datapub consulting services","excerpt":"Digital marketing is a complex ecosystem and with technology it seems to be getting tougher with time. Making this digital marketing practise to help drive business objectives, there is a need for a structured framework. This should be built upon the data and should be able to help to get the best of return. This […]","categories":["AI Trends"],"tags":["digital marketing"],"author_name":"Gayatri Choda","publish_date":"2012-12-29T15:52:55","publication_year":"2012","word_count":513,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","Git","digital marketing","R"],"extracted_tech_keywords":["AI","R","Go","Git","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/4c-optimization-for-digital-marketing-practice\/","complexity_score":3,"technical_depth":6,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10134135,"title":"Reliance Jio Announces Free Cloud Storage with Jio AI-Cloud","content":"Reliance Industries Limited (RIL) chairman and managing director (CMD) Mukesh Ambani announced the launch of the Jio AI-Cloud Welcome Offer, and Jio Phone Call AI at the company’s 47th annual general meeting. Jio Phone Call AI is a no app based feature that allows users to record calls on their server and transcribe it too, it also allows translation for the AI generated transcription all by calling on their AI number. “Today, to support our AI Everywhere For Everyone vision using connected intelligence, I am thrilled to announce the Jio AI-Cloud Welcome Offer,” Ambani said, with user data safely stored on the cloud, AI will be able to deliver intelligent personalised services over the network, he added. Set to go live in Diwali, this offer will allow its Jio users 100 GB of free cloud storage space to securely store and access photos, videos, documents and other digital content and data. Earlier this year, Jio launched ‘Jio Brain’, positioned as the Industry’s first 5G-integrated ML platform aimed to empower telecom networks, enterprise networks and industry-specific IT environments to seamlessly incorporate ML tools into their day-to-day operations. Akash Ambani, Chairman of Reliance Jio Infocomm Limited (RJIL) also announced the upgrade to its HelloJio, Jio STB’s voice assistant using Gen AI technology, improving its natural language processing making it feel more human-like to understand Indian dialects better. Jio Goes Big on AI Jio is making significant strides in the field of AI, looking at the broader Indian AI landscape, the JioGenNext cohort highlights the potential of AI in various sectors like healthcare, banking, and agriculture. Startups like Medhini-Arficus and Dista are developing innovative solutions that address real-world problems. Additionally, the collaboration between NVIDIA, Tata Communications, and Jio Platforms will provide the necessary infrastructure for AI development in India. Akash Ambani, the chair of Jio also announced that it is partnering with IIT Bombay for the BharatGPT initiative aimed to address the lack of large language models (LLMs) for Indic languages. BharatGPT’s open-source approach fosters collaboration and knowledge sharing, which is essential for overcoming these hurdles and establishing India as a leader in Indic LLMs.","excerpt":"Jio also marks its entry into the AI based telecommunication space with Jio Phone call AI","categories":["AI News"],"tags":["Jio","Jio AI Cloud"],"author_name":"Tanisha Bhattacharjee","publish_date":"2024-08-30T12:04:47","publication_year":"2024","word_count":352,"keywords":["Jio","Go","AI","ML","Git","RAG","GPT","Aim","Jio AI Cloud","llm_models:GPT","R","startup"],"extracted_tech_keywords":["AI","ML","Aim","RAG","R","Go","Git","GPT","startup","llm_models:GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/reliance-jio-announces-free-cloud-storage-with-jio-ai-cloud\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10020203,"title":"40 Under 40 Data Scientists In India 2021: Who Are They?","content":"The largest-ever virtual summit for machine learning developers recently concluded after three action-packed days of tech talks, workshops, paper presentations, networking and awards. The Machine Learning Developer Summit, hosted from 11-13 February, saw more than 1,000 attendees from the data science community.  The highlight of the summit was the 40 under 40 data scientists awards. The MLDS recognised 40 young data scientists who have supported the growth of analytics in their organisation and used their deep industry experience to address most the complex challenges using analytics. The winners represented various industries including retail, insurance, e-commerce, and more. The award felicitates the real innovators and achievers of the analytics industry — the driving force behind the field’s explosive growth. Here are the winners of 40 under 40 data scientists awards 2021. Check out last year’s winners. Abhishek Sanwaliya Practice Lead – Data Science, Maersk He currently manages AI-enabled digitisation initiatives for customer service, order handling platform and capacity planning at Maersk. Apart from managing teams across Data Sciences Research and Advanced Analytics domains, Abhishek provides thought leadership for startup accelerator ‘Oceanpro’ to help graduating startups contribute to Maersk’s digital journey. Abhishek has filed six patents, three trade secrets and has published six research papers in top peer-reviewed journals and conferences. Aditya Agarwal Advanced analytics Practise Lead, Abzooba He has over 12 years of expertise in using quantitative data analysis with business understanding for enterprise decision-making. He currently leads a team of more than 50 energetic data science professionals at Abzooba while solving interesting business problems using machine learning, deep learning, natural language processing and computer vision. Aditya has built solutions such as subrogation engine, price recommendation engine, IoT sensor predictive maintenance, and many more. Aman Garg Head – Analytics, Vodafone Shared Services Aman played a pivotal role in setting up and establishing Advanced Analytics centre for Vodafone in Shared Services in Pune as part of the original team. Aman works closely with the Global Head of Data & Analytics, Country Head of Business & CXO level executive in solving critical business problems. The analytics products developed by Aman and his team have delivered annual cumulative benefits of more than £20million+ in value to businesses covering Digital, CVM & Finance Analytics. Amit Sachan Head of Recommendations System,  Reliance Jio AI Centre of Excellence In his current role, he creates a unified recommendation engine for Jio for a variety of applications in media, retail, healthcare and news. He is currently working on interesting themes such as the role of knowledge graphs, reinforcement learning in real-life recommendation systems and integrating technologies like NLP and computer vision extensively in the context of recommendation systems. Arul Francis Engagement Manager, Shell He has over 12 years of experience working in a range of organisations from start-ups to large organisations in their analytics function. At Shell, Arul has led and managed numerous engagements in data science in the Shell retail business. He has led and transformed bootstrapped startups like Ameex into a successful data science enterprise with solutions across data management, data visualisation and AI. Ashish Kumar Senior Director – AI Research at Reliance Jio digital services (Jio Haptik) He builds innovative AI solutions for different Jio Haptik engagements. Ashish has strong technical leadership in conversational AI products and has contributed to many NLP initiatives. He has helped build chatbots deployed at all major banks in India and has incubated and fostered state-of-the-art technology domains from scratch in many companies. Ashish is also a published author, inventor of more than half a dozen patents in conversational AI, and has extensive know-how in conceptualising and delivering innovation while building high-performance R&D teams. Ashish Ranjan Singh Vice President – Customer Engineering at Innovaccer Analytics Private Limited At Innovaccer, Ashish is leading a team of more than 100 data analysts, data engineers, and data scientists driving a transformation in healthcare. Ashish helped Innovaccer build world-class technologies, recognised internally and externally. Prior to Innovaccer, Ashish has worked at Barclays where he developed a real-time analytics engine for targeted campaigns across the globe. He was also a key player in developing the first data and analytics product for Barclays, called Smart Business portal, which is heavily used by small businesses in the UK. Bala Gangadhar Thilak Adiboina Senior Member – Data Science, Verizon He has over 13 years of experience in conceptualising and implementing many innovative AI solutions in areas such as self-optimising network, dispatch analytics, 5G analytics, fraud analytics and IoT analytics. For his innovations in implemented solutions at Verizon’s Network Systems, he has filed three US patents in 2020. Bala’s Data Science journey started with his startup Favmeal in 2008, a food delivery company where he developed a solution for food delivery optimisation using Google Maps and Google Analytics. Deepanshu Malhotra Director, Swiggy He has deep expertise in changing business landscape with data-driven intelligence across Banking, FMCG, and E-Commerce domains. Deepanshu leads a team of more than 40 members and is responsible for marketing, growth strategy, and financial analytics for the Food delivery business. Deepanshu is adept at applied mathematics, optimisation algorithms, supervised, unsupervised, and forecasting machine learning techniques. He was also instrumental in setting up the “COVID – Business Continuity” Team to enable sustainable growth during the pandemic. Dipanjan Sarkar Data Science Lead, Applied Materials He heads the advanced analytics efforts around computer vision, natural language processing and deep learning. He primarily works on leveraging machine learning and deep learning to build large scale intelligent systems. Dipanjan is also a Google Developer Expert in Machine Learning and has advised and worked with several startups as well as Fortune 500 companies. He is also a published author and has mentored over 500 people in data science globally. Harsha Gurulingappa Text Analytics Lead Scientist, Merck KGaA One of his recent accomplishments is setting up an industrialised text mining system at Merck which operates in a highly regulated GxP environment. This work has been published in two peer-reviewed journals in Q4 2020. Harsha has also helped establish an enterprise business service for text analytics from scratch at Merck. He has authored more than 27 publications and book chapters demonstrating a variety of applications of text analytics. Hemant Palivela Head – AI and Machine Learning, eClerx Services He works on campaign analytics, speech processing, Natural Language Understanding\/Generation and recommendation systems at the Centre of Excellence, Digital Analytics Division of eClerx. His principal areas of interest include Machine learning optimisation, linear algebra, probability theory, insurance analytics and recommendation systems. Indrajit Mandal Head – Data Science, Tata Consultancy Services Throughout his career, Indrajit has worked across domains such as medical sciences, fintech industry, BPS and more. While working in the fintech industry, he had developed a data-driven AI platform for decision systems that automates business operations, bringing more scalability and improving revenue growth. While working in the IT domain, he developed cold start solutions in machine learning algorithms, of which a few are patented. James Macwan Manager – Analytics, JSW Steel Limited He has over 14 years of experience in delivering large scale Data Analytics transformation programs across various industries and geographies. James has expertly led and managed the delivery of multiple data analytics and technology-agnostic business projects using agile and waterfall project delivery methodologies. In his current role, he has built a data analytics platform and big data Warehouse capabilities for real-time data analytics, ensuring an increased adoption of analytics by 45% a year. Kamal Kumar Head – Analytics and Data Engineering, Tata Cliq Kamal has been responsible for setting up growth hacking, revenue and analytics teams at a couple of e-commerce firms. He started off his analytics career by establishing agile teams to deliver cutting edge analytics, data science and data engineering solutions across industry verticals spanning banking, logistics, offline retail and e-commerce. He has co-founded and led the consulting vertical of a management consulting and a product startup, providing data-driven consulting services. Kartik Patel Head – Insurance Analytics & Products at TransUnion CIBIL He has over 12 years of experience in leading complex data science and actuarial science assignments for clients across the globe. Currently, he leads insurance analytics and products teams working on innovative applications of credit data of more than 400 million individuals in solving critical challenges of the Indian insurance industry. He is also associated with National Insurance Academy in training insurance professionals on insurance analytics and has cleared nine actuarial science papers from Society of Actuaries and Casualty Actuarial Society. Kiran Rokkam Data Science, Tiger Analytics Kiran heads the innovation lab at Tiger Analytics. He has been responsible for bringing several cutting edge innovations in data science automation, which are helping drive the next phase of growth for the company. Kiran has developed solutions in customer analytics, marketing analytics, operations planning, pricing, risk, etc. He has worked with a broad spectrum of data science techniques ranging from classic econometric modelling to the latest in deep learning and NLP. Lavi Nigam Data Scientist, Gartner At Gartner, he is developing MLOps practices to produce auto-deployable models to streamline and improve data quality. In his previous jobs, he has developed NLP models to work across multiple locations and worked with deep learning algorithms and operationalising autoDL production systems. He has also worked with IITs and NITs as subject matter expert to design industry-relevant courses for AI. Manikandan Jeeva Asst. Vice President, Genpact He has over 15 years of experience in pharmaceutical and healthcare analytical and consulting across functional areas such as commercial analytics, clinical trial analytics and supply chain analytics. Manikandan is an experienced project manager in handling substantial business intelligence and analytics projects in cloud transition, marketing mix model, neural network implementation for data quality system, public data mining and integration for brand insights, among others. Mihir Punjabi Solutions Director Architect – Edge Intelligence, Capgemini Mihir is an Edge Intelligence and AI evangelist, currently responsible for crafting AI solutions across Digital Engineering and Manufacturing Services along with supporting AI solutions across Capgemini. With over 17 years of experience in product development and business development across Storage, Building Automation and Embedded domains, Mihir has led several projects across domains. He is a thought leader and has presented papers at various industry forums and events. Muthumari S Associate Director, Analytics, Brillio Technologies She heads the Data Science CoE and specialises in delivering AI\/ML use-cases at scale across customer\/product\/marketing analytics, NLP and vision analytics. In her current role, she works with CXOs to help them make accurate data-driven decisions as well as solve ambiguous problems involving unstructured text\/image and machine-generated data with tangible business impact. She works as a thought leader and people leader, helping build data science talent and capabilities within Brillio. Mythili Krishnan Vice President – Analytics advisory senior manager, Accenture She has more than 14 years of experience in various areas of analytics, data science and AI including credit risk management, marketing analytics, Infrastructure services and retail analytics. Mythili has built teams and analytical practices from bottoms up and has contributed to the strategic vision of the organisation with speed and agility.  She has in-depth knowledge of statistical modelling and econometrics. Nitin Aggarwal Technical PM, Cloud AI, Google Nitin is responsible for designing AI solutions and products for Google’s strategic enterprise customers. Nitin has donned many hats during his professional career such as consultant, product manager, team lead, analyst, innovation expert, software engineer and technical document writer. He loves automating things using AI\/ML and finds Mathematics and Data Analytics fun to work with. Nitish Palaparti Business Analytics Specialist, Microsoft He has over ten years of experience across multiple divergent domains like FMCG, pharmaceuticals, banking and e-commerce. Nitish has worked on NLP, multi-class models, label classification and regression cases. He has built and deployed predictive, market mix and forecasting models and conducted enormous research by comparing on-Prem, cloud and auto ML models. Currently, he is building a Recommender System for Microsoft cloud. Parikshit Nag Country Manager, Data Science, Hindustan Unilever Ltd. He spearheads the data science initiatives at Hindustan Unilever. He has been responsible for developing key data science products which help augment human intelligence with AI & ML. He has also developed a multi-horizon automated forecasting engine used across various business functions. He has a patent pending for this product. Piyush Gupta Head of Intelligence, PayU Finance India Pvt Ltd. Piyush has developed a next-generation data analytics platform at PayU Finance to serve the needs of the business for quick experimentation, real-time location-based marketing, and sales. Under his guidance, the data science team has developed NLP chatbots and in-house solutions for the identification of first-party fraud in Citi. His interest lies in solving complex problems using mathematics, statistics, and large-scale computing. Pooja Aggarwal Advisory Research Scientist at IBM Research India Her current research focuses on infusing artificial intelligence into IT technical service processes. She has led the design and implementation of advanced prototypes as well as production-ready features such as action recommendation, event grouping and fault localisation. Pooja has designed and developed the cognitive question-quality-improvement system, a platform for automating enterprise after-sales support systems. She has published 16 papers in prestigious conferences and journals in the area of text analytics, parallel computing, conversation systems, software engineering etc. Sandeep Earayil Head – Machine Learning and Quantitative Analytics, Credit Suisse He has over ten years of experience building Artificial Intelligence and Machine Learning solutions in the financial sector. Sandeep is also one of the pioneers of using AI\/ML in the field of alternative data at the esteemed UBS evidence lab and now at Credit Suisse. He leads a team of ML researchers who works in areas like Geospatial analytics, Natural Language Processing, Graph Analytics and Pattern mining. Sanjay Shukla Director – Data Science, Media and Customer Engagement, Dunnhumby He is passionate about building innovative AI and machine learning solutions to provide actionable insights, highly-relevant product recommendations and media measurement resulting in improved customer experiences in retail. Sanjay has led the development and deployment of multiple advanced and scalable ML-based solutions. He has published multiple research papers on optimisation and machine learning approaches in peer-reviewed Journals. Satyen Abrol Principal Research Scientist, Rakuten India He has over 12 years of experience in user personalisation, location mining and online social networks. At Rakuten, he leads two teams across two departments — Research and Applied Data Science. The research team is building a Customer Social Graph to understand the relationship between Rakuten users, products and businesses. Whereas, the applied sciences team is building a framework to help marketers achieve business objectives by recommending right users, right channel, right time and the right incentive for targeting. Sayanti Bhattacharya Senior Manager, Ugam, a Merkle company Sayanti currently leads Analytics for Fortune 500 companies, across industries such as Retail & Consumer Brands and eCommerce, helping various business functions drive impact and ROI, and institutionalise analytics practices at scale. At Ugam, Sayanti has designed and set-up the framework for scaled deliveries on analytics services and solutions. Her Consumer Sentiment Analysis solution is used by Product Engineering, Quality and Brand teams for addressing product development, innovation and product improvements in various categories. Sharmistha Chatterjee Senior Manager, Data Sciences at Publicis Sapient She currently leads the digital transformation of clients across industry verticals ranging from Media, Travel and Hospitality, Advertising, IoT, and Telecom. Sharmistha is an active blogger and an international speaker at various tech conferences. She has filed five US patents and published in international conferences and journals. She is also a Google Developer Expert and enjoys mentoring. Shrutendra Harsola Sr Machine Learning Scientist, Intuit India He has a strong background in statistics, machine learning, natural language processing. In his current role, he leads the AI track in the area of Growth and Retention for QBO-Advanced, a cloud-based accounting solution for growing small to mid-sized businesses. Shrutendra is responsible for taking AI ideas from conceptualisation to production driving significant customer benefit in collaboration with multiple stakeholders. He has been a part of  Microsoft Bing Ads team in the past. Suguna Jayaraj Senior Manager- Predictive Modelling, AIG Analytics & Services Private Limited. She leads the predictive modelling capability for the general insurance products at AIG and has deep expertise in pricing\/rating for the P&C insurance as well as in credit risk rating for banking products for cards. Suguna has built the team to support the actuarial pricing and portfolio management. Most of her work revolves around improving the performance of the actuarial unit by adding granularity of the pricing framework, increasing the underwriting process efficiency and building competitive strategies. Sumit Bhatia Senior Research Scientist, IBM Research India. He is the co-architect of IBM Expressive Reasoning Graph Store, a first of its kind graph-based reasoning platform that supports both Property Graphs and RDF graphs. Sumit has developed novel algorithms for entity-oriented search and exploration on knowledge graphs deployed by multiple enterprise clients and government security agencies. Sumit has over 50 publications in top-tier venues, 14 US patents, and more than 1,200 citations to his research. Swapnil Srivastava Vice President & Global Head of Analytics, Evalueserve Swapnil heads the Analytics Practice for Evalueserve where he works with clients to help them build analytics capabilities that enable successful business outcomes. He has built a multi-disciplinary team of domain SMEs, data architects, data engineers, data scientists and technology experts, making Evalueserve an end-to-end analytics partner for its clients. He has developed several Analytics Frameworks for Evalueserve and its F500 clients. Swapnil Tambi Director AI & ML, BNY Mellon He has extensive experience in modelling and analytics. A strong problem solver and team leader, he is known for bringing strategic thinking into business problems. His key areas of expertise include hypothesis generation, data wrangling, feature engineering, classification model building, model validation and more. Vartul Mittal Technology & Innovation Specialist, Formerly IBM Vartul is a leading technology & Innovation Specialist and a distinguished digital transformation Leader who has driven many interesting developments in the space of advanced analytics, intelligent automation and multi-cloud adoption. He has 14+ years of strong global business transformation experience in management consulting with a key focus on Artificial Intelligence, Machine Learning and Data Engineering across various sectors. Vibhu Srivastava Assistant Vice President, AI Works, Max Life Insurance Vibhu has been instrumental in carving the AI works team along with building a brand for Max Life across external forums. With his strong focus on  ‘business first’ AI solutions, he has driven measurable, direct and unprecedented business impact. He has a strong hands-on technical expertise in areas such as machine learning, deep learning, NLP. Vikram Khurana Head of Data, Times Internet Gaana He is the head of data at the largest music streaming company in India and has been responsible for setting up data teams in previous organisations. Vikram has helped many organisations in India and outside in realising multifold gains from data. He has rich experience in B2B, B2C, consulting and start-up space.","excerpt":"The largest-ever virtual summit for machine learning developers recently concluded after three action-packed days of tech talks, workshops, paper presentations, networking and awards. The Machine Learning Developer Summit, hosted from 11-13 February, saw more than 1,000 attendees from the data science community.  The highlight of the summit was the 40 under 40 data scientists awards. […]","categories":["Deep Tech"],"tags":["40 Under 40 Awards","Data Scientist Awards","Statistics for Data Science"],"author_name":"AIM Media House","publish_date":"2021-02-15T18:00:00","publication_year":"2021","word_count":3127,"keywords":["Data Scientist Awards","data science","artificial intelligence","machine learning","AI","neural network","ML","computer vision","NLP","deep learning","Statistics for Data Science","analytics","40 Under 40 Awards"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","neural network","NLP","computer vision","data science","analytics"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/40-under-40-data-scientists-in-india-2021-who-are-they\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":7,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10039682,"title":"ThoughtSpot Acquires Diyotta To Expand Ecosystem For Modern Analytics Cloud","content":"ThoughtSpot announced that the company acquired Diyotta, a data integration solutions provider. With this acquisition, more than 60 new employees in North America and India will join ThoughtSpot to supercharge the development of the Modern Analytics Cloud. Furthermore, this adds 25% more headcount to the global R&D team and 50% to the R&D team in India, including an expanded footprint in Hyderabad. The acquisition of the data integration solutions platform will help the Analytics Cloud company rapidly expand integrations with the best services in customers’ modern data stack. Customers will be able to deploy ThoughtSpot seamlessly as part of their modern cloud architecture and deliver instant value to their business. In particular, the acquisition of Diyotta will accelerate integrations with modern data platforms, AI and ML services, and data applications built by web developers. The acquisition of Diyotta will further accelerate expanding the ecosystem for the Modern Analytics Cloud, helping connect and integrate ThoughtSpot’s analytics platform with technologies across this burgeoning industry. Sumeet Arora, Chief Development Officer, ThoughtSpot stated, “We’ve been fans of the work Diyotta has done to help their customers solve some of the toughest challenges involved in integrating their data into their business workflows. Diyotta’s team understands the dynamic, nuanced ecosystem and unique problems companies face in transforming how they leverage their data to drive processes and change decision-making.” “As we unlock more value for our customers with the Modern Analytics Cloud, Diyotta’s experience and shared vision for bringing together people, processes, and data was a perfect complement,” he added. Commenting on the same, Sanjay Vyas, CEO & Cofounder of Diyotta said, “Both ThoughtSpot and Diyotta have large, shared enterprise customers and with this merger, we are looking to provide more comprehensive applications in order to make cloud analytics seamless. Companies today run on data. It’s at the core of every modern organization. By joining forces with ThoughtSpot, we’re going to be able to expand how we help these businesses empower their entire team with data.”  Vyas added, “The Modern Analytics Cloud democratizes the most cutting edge innovations in the data ecosystem by putting them in the hands of every employee. I’m incredibly excited to help advance that important mission.”","excerpt":"ThoughtSpot announced that the company acquired Diyotta, a data integration solutions provider. With this acquisition, more than 60 new employees in North America and India will join ThoughtSpot to supercharge the development of the Modern Analytics Cloud. Furthermore, this adds 25% more headcount to the global R&D team and 50% to the R&D team in […]","categories":["AI News"],"tags":["Analytics India","data analytics acquisitions","Mergers and Acquisitions","thoughtspot"],"author_name":"Ambika Choudhury","publish_date":"2021-05-05T18:33:01","publication_year":"2021","word_count":363,"keywords":["Go","API","thoughtspot","AI","innovation","ML","Analytics India","data analytics acquisitions","RAG","analytics","GAN","Mergers and Acquisitions","R","analytics platform"],"extracted_tech_keywords":["AI","ML","analytics","RAG","R","Go","API","GAN","analytics platform","innovation"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/thoughtspot-acquires-diyotta-to-expand-ecosystem-for-modern-analytics-cloud\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":22862,"title":"Latest Update On Cambridge Analytica Controversy: CEO Alexander Nix Suspended","content":"Cambridge Analytica, the political data firm which has been accused of “harvesting” personal data taken illegally from Facebook and used to influence elections, suspended their chief executive officer Alexander Nix on Tuesday. The decision to sack Nix comes hot on the heels of British news channel Channel 4‘s broadcast of a video of an undercover operation, where the Nix is seen talking about working on over 200 elections across the world — including in Nigeria, Kenya, the Czech Republic, India, Argentina – and bribing opponents to sway the election results. Meanwhile, Cambridge Analytica’s chief data officer Alexander Tayler will serve as acting CEO while the probe is under way. The company board said a full investigation into the comments and allegations would be led by noted UK lawyer Julian Malins. After Cambridge Analytica was accused of illegally obtaining information on more than 50 million Facebook users, the social media giant banned the analytics firm Strategic Communication Laboratories and its political arm on the grounds of failure to follow their rules regarding handling of personal data. In what data experts are terming as one the largest abuses of personal data in history,Cambridge Analytica has been accused used this data to help target American voters for its work with (now) US President Donald Trump’s 2016 election campaign. .@HillaryClinton: “How did the Russians know how to target their messages so precisely?” Last year the former presidential candidate questioned whether Cambridge Analytica were involved in Russia’s alleged attempt to influence the U.S. election. #CambridgeAnalyticaUSA pic.twitter.com\/8varOIJRfB — Channel 4 News (@Channel4News) March 20, 2018 Facebook said that it is “outraged” by the misuse of data by Cambridge Analytica, according to local media reports. On the other hand, Brian Acton, co-founder of WhatsApp — which was acquired by Facebook in 2014 — has joined the #deletefacebook movement on Twitter. Meanwhile, in India, Indian National Congress as well as Bharatiya Janata Party have been asking the Election Commission to help them find proof of involvement (against each other) regarding alleged links with Cambridge Analytica.","excerpt":"Cambridge Analytica, the political data firm which has been accused of “harvesting” personal data taken illegally from Facebook and used to influence elections, suspended their chief executive officer Alexander Nix on Tuesday. The decision to sack Nix comes hot on the heels of British news channel Channel 4‘s broadcast of a video of an undercover […]","categories":["IT Services"],"tags":["Big Data","BJP","cambridge analytica","congress","Facebook","personal data"],"author_name":"Prajakta Hebbar","publish_date":"2018-03-21T12:35:53","publication_year":"2018","word_count":337,"keywords":["cambridge analytica","congress","programming_languages:R","AI","RAG","BJP","analytics","Julia","Facebook","Big Data","R","personal data"],"extracted_tech_keywords":["AI","analytics","RAG","R","Julia","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/it-services\/cambridge-analytica-controversy-heres-the-latest-update\/","complexity_score":2,"technical_depth":6,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10100760,"title":"AI Trust Grows When Hidden","content":"“People May Be More Trusting of AI When They Can’t See How It Works,” read one of the articles in the latest edition of Harvard Business Review, which showed how not knowing the workings of a model, helped people trust the process more. A similar pattern can be observed among the tech industry leaders. Apple, one of the image-conscious companies of the Silicon Valley lot has also made sure to keep a tight-lip about their AI\/ML doings. The same goes for OpenAI, which is trying really hard to hide its technology, and yet struggling to woo enterprise customers. During this year’s WWDC, CEO Tim Cook conspicuously refrained from using ‘AI,’ opting for the more subdued ‘machine learning.’ The iPhone maker’s aversion to using ‘AI’ as a label is far from new, as the company has long been cautious about its techno-magical capabilities. Instead, Apple focuses on the practical functionalities of machine learning, highlighting the tangible benefits of its user-centric audience. The company’s chief put it in an interview with Good Morning America today, “We do integrate it into our products [but] people don’t necessarily think about it as AI.” This gives Apple an upper hand over its competitors like Microsoft and Google who are currently boasting their AI-powered products yet struggling to get adapted throughout companies. OpenAI is abiding by the same playbook as Apple in terms of secrecy. The 98-page technical paper released by the company lacked even the basic details about the AI model’s data or architecture. While the paper was heavily criticised for being shallow, the secretive approach seems to be working in the company’s favour. Trusting the Process As per HBR, a group of researchers from Georgetown University, Harvard and MIT analysed the stocking decisions for 425 products of US luxury fashion retailers across 186 stores. Half the employees received recommendations from an easily understood algorithm and the other half of the recommendations from one that could not be deciphered. A comparative analysis of the decisions made it evident that employees align with the recommendations provided by the opaque AI more frequently. The result concluded that individuals exhibit higher confidence in AI systems when they don’t thoroughly know how it works. Professor Timothy DeStefabo highlighted a well-established phenomenon wherein decision-makers are often reluctant, whether consciously or unconsciously, to embrace AI-generated guidance, opting to override it. This is not the first time. Historically, new technologies receiving initial resistance has been a norm. DeStefabo and his team partnered with Tapestry, a company boasting a worth of $6.7 billion and the parent entity of Coach, Kate Spade, and Stuart Weitzman. The collaborative effort began to examine the roots of this reluctance and find strategies to mitigate it. The company had long used rule-based algorithms to help allocators estimate demand. The model was understood by the users from their daily experience and whose inputs they could see. The firm developed a more sophisticated forecasting model that was a black box to users, for better accuracy. Turns out, the shipments were up to 50% closer to the recommendations generated by the latter, suggesting the users trusted the black box model much more. Prior to this initiative, the company had long relied on rule-based algorithms to help allocators estimate demand. These algorithms were comprehensible to users, based on their daily experiences and inputs. However, the firm developed a more intricate ‘black box’ forecasting model for better accuracy. Surprisingly, the shipments were up to 50% closer to the recommendations generated by the latter, suggesting the users trusted the black box model much more. This outcome suggested that users placed greater trust in the black box model. One reason allocators overruled the less sophisticated system was due to ‘overconfident troubleshooting’ – users believe they understand models better than they actually do. Even though the employees could not tell how the model worked because it had been developed and tested with inputs from some of their colleagues it gave them confidence in the model, wrote DeStefabo. In conclusion, tech companies need to focus on what customers need, and not sell the know-how of technology to their customers.","excerpt":"While companies are heavily criticised for being secretive the approach is working in their favour","categories":["AI Features"],"tags":["AI Black Box","AI Models","OpenAI","OPenAI GPT","Trustworthy AI"],"author_name":"Tasmia Ansari","publish_date":"2023-09-27T17:30:00","publication_year":"2023","word_count":682,"keywords":["AI Models","Go","machine learning","OpenAI","AI","programming_languages:R","ML","AI Black Box","programming_languages:Go","Trustworthy AI","OPenAI GPT","Rust","GAN","R"],"extracted_tech_keywords":["AI","machine learning","ML","OpenAI","R","Go","Rust","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/ai-trust-grows-when-hidden\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10168322,"title":"Why Every Siemens Healthineers Product Now Has India at Its Core","content":"Since the early 2000s, some of the world’s largest healthcare and life sciences companies have established their GCC operations across India. As per a report by ANSR, cities like Bengaluru, Hyderabad, Mumbai, and Delhi-NCR host nearly 85% of all healthcare and life sciences GCCs today, with the sector employing over 15% of the entire GCC workforce in the country. This vibrant ecosystem is now fueled by a new generation of Indian tech talent, comprising experienced engineers, startup innovators and fresh graduates. From radio-frequency (RF) devices and molecular testing to AI-based medical imaging and fluorescence diagnostics, the outcomes emerging from Indian GCCs are largely changing the concept of patient care. There is also a continuous effort to make healthcare more personalised, customer-centric and accessible. In fact, in a recent LinkedIn post, Bernd Montag, CEO of Siemens Healthineers, talked about his visit to India. “The 7,000 Healthineers here have made such a profound impact on the innovation and growth within our company that every single Siemens Healthineers product now has ‘India inside of it’.” The Development Centre (DC) at Siemens Healthineers is essentially a miniature version of the company itself, featuring a large laboratory setup. The company takes great pride in the fact that every product or solution that leaves Siemens Healthineers has its origin in India, with the Development Centre fully involved in its creation, maintaining complete ownership of the process. Following this, he further mentioned, “Siemens Healthineers has been manufacturing medical devices in India for over 60 years now. To expand our capabilities, we are investing €160 million (₹1,300 crore) at our Development Centre in Bengaluru, which is on target for a 2025 opening.” Reiterating this, Kalavathi GV, executive director and head of the global development centre at Siemens Healthineers, mentioned that in fields like CT, MRI, and molecular imaging, the Siemens Healthineers team in India contributes to both scanner development and post-processing. Moreover, for AI-related initiatives, the team undertakes essential tasks including data mining, annotation, and other associated tasks. An Ode to Indian Talent Most GCCs in India are driving their AI innovation from India. For instance, Lowe’s team in Bengaluru has been a critical driver of its AI journey and has played a critical role in developing and deploying cutting-edge solutions. “Many of our core systems, including the omnichannel order management and self-checkout terminals, were built from the ground up by our engineers in Bengaluru. These solutions have provided us with unmatched scalability and flexibility,” Amit Kapur, VP of AI and data analytics at Lowe’s India, told AIM. Similarly, Raghavendra Vaidya, managing director and CEO of Daimler Truck Innovation Centre India (DTICI), said in a conversation with AIM that a major project of predictive maintenance, directed from Bengaluru, aims to predict part failures in trucks and buses using analytics and machine learning, instead of traditional physics-based methods. In the healthcare sector, Siemens Healthineers plays a vital part within its parent company, making significant contributions to innovation. At present, approximately 54% of the global software workforce is based in Bengaluru, with a team of 3,500. “Over the next 12 to 18 months, the organisation aims to expand the team by adding between 900 and 1,200 new members,” Kalavathi added. She further revealed that the company is now focused on evolving from local to global solutions. For example, Cios Fit, a multidisciplinary mobile C-arm, is built in India for the Indian market. It is now being expanded for use in other countries as well. The product management team holds full responsibility for this mobile solution across the Indian subcontinent and beyond. Another notable innovation is the AI-Rad Companion, a software that has made significant contributions from India. This tool plays a crucial role in automation, particularly in cancer contouring and planning, and is becoming an essential companion for radiology. AI-Rad Companion is also expanding its capabilities into cancer care identification. In addition, in a collaborative effort between India and global teams, the company has developed AIDAN, an intelligent imaging platform integrated into PET-CT systems that eliminates the need for ECGs. This task once required manual placement by technicians. Simultaneously, the company is making strides in theranostics, a combination of diagnostics and therapy, particularly in molecular imaging. The company is steadily progressing towards greater integration across its portfolio, with a strong focus on product management. Notably, the global software leader for XP is based at the Bengaluru centre, alongside many other global functions. The ultimate goal, as Kalavathi pointed out, is to adopt a digital twin model—not just for patients but also for operational twins—which will improve hospital availability and help meet Service Level Agreements (SLA) with clients. As part of its AI-based expansion, the company established a Centre of Competence (CoC) within the DC to concentrate on AI development. It also started a GenAI Centre of Competence to enhance the productivity of development processes. Talent in India Previously, in an interview with AIM, Raja Jamalamadaka, country head and board director at Roche Information Solutions India, stressed that despite efforts by the government and research bodies to bridge the gap in AI training, the disconnect between academia and industry expectations persists. He added that many companies compensate for this by investing in induction programs and internal skill-building initiatives. As Kalavathi pointed out, Siemens Healthineers aims to close the gap between academia and industry by sending industry professionals to engage with academic institutions. Moreover, the company is exploring ways to collaborate with institutions such as Manipal Academy of Higher Education to improve the syllabus and ensure it aligns with current industry needs. The company is also partnering with NASSCOM to focus on upskilling initiatives.","excerpt":"The Development Centre at Siemens Healthineers is essentially a miniature version of the company itself, featuring a large laboratory setup.","categories":["GCC"],"tags":["AI (Artificial Intelligence)","ai in health sector","GCC","GCC india"],"author_name":"Shalini Mondal","publish_date":"2025-04-21T14:57:56","publication_year":"2025","word_count":932,"keywords":["Go","GenAI","GCC","machine learning","AI","ML","Scala","RAG","ai in health sector","Aim","analytics","GCC india","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","machine learning","ML","analytics","GenAI","Aim","RAG","R","Go","Scala"],"url":"https:\/\/analyticsindiamag.com\/gcc\/why-every-siemens-healthineers-product-now-has-india-at-its-core\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":44896,"title":"A Day In The Life Of: A Seasoned Products Engineer Who’s Pursuing New Revenue Streams","content":"In our weekly column ‘A Day In The Life Of’, we are trying to step into the shoes of awesome techies from various organisations and sectors who are working in emerging tech areas like big data, data analytics, artificial intelligence, machine learning and the internet of things, among others. This week Analytics India Magazine got in touch with Arvind Gopalakrishnan, VP and Head of Solutions, Products and Sales Engineering at an IoT company, who believes in “quality over quantity” both in his professional and personal life. Gopalakrishnan is a driven and experienced professional who likes to stay motivated. He gets up at 6 am to enjoy a morning cuppa with his entire family. Even though his work begins at 9.30 am,  Gopalakrishnan logs in early and starts the reviews and discussions a couple of hours early. Even though Gopalakrishnan is a seasoned techie, he makes sure that he gets his morning workout at least four times a week. But that does not mean he does not know how to have fun. Describing his social life as ‘active’, Gopalakrishnan explains, “I like meeting my old friends every now and then. My colleagues and I often go out for coffee or dinner, as and when time permits.” When asked about his work, Gopalakrishnan says, “It isn’t news that the internet of things is revolutionising our lives in more ways than one – IoT is everywhere. As a person in a senior position in the company, my profile revolves around the customers and solutions. My work consists of creating strategies, executive management, customer engagement and management, product go-to-market, implementation and rollout of solutions and end-to-end customer life-cycle management. My team is involved in creating and developing go to market connectivity, IoT solution and products.” Gopalakrishnan explained that he’s currently working on identifying and generating businesses across medium to large-scale customers in different verticals. “Most of my energy goes in identifying the customers’ solution needs, value and market uptake. As a matter of fact, businesses are slowly transitioning toward the internet of everything (IoE), an unprecedented network connection that encompasses machines, individuals, processes and data that can have a huge impact in our daily lives.” he adds. When asked about the best part of his workday, Gopalakrishnan replies, “The best part is usually evenings between 5:30 to 7:30 because at that time my office is about to get closed and I can generate innovative ideas without being disturbed.” When asked about if he feels fulfilled and excited to work with his current employer, Gopalakrishnan says, “Based on previous experience which I have gained in my old organisations and current engagement, Aeris Communications is making best use of my talent which includes customer engagement, business development, innovation, right solution to right customer and market and applying digital technology to business. Having said that, there are always new ideas and concepts which can be brought to life for better optimisation or new revenue stream generation.” Gopalakrishnan has a vision for the future. “Grow the IoT solution business across different markets for the organisation; become a leader in transforming the market; enhance the customer experience; grow their revenue in a big way and develop sustainable relationships and journeys to take through later part of the life,” he explains. His short-term goal for now is to develop new verticals and solutions and grow revenue for his organisation.","excerpt":"In our weekly column ‘A Day In The Life Of’, we are trying to step into the shoes of awesome techies from various organisations and sectors who are working in emerging tech areas like big data, data analytics, artificial intelligence, machine learning and the internet of things, among others.   This week Analytics India Magazine […]","categories":["AI Features"],"tags":["Data Science Career","Interviews and Discussions"],"author_name":"Prajakta Hebbar","publish_date":"2019-08-24T00:45:38","publication_year":"2019","word_count":559,"keywords":["big data","Go","machine learning","artificial intelligence","AI","Git","Data Science Career","ViT","analytics","GAN","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","analytics","R","Go","Git","big data","GAN","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/a-day-in-the-life-of-a-seasoned-products-engineer-whos-pursuing-new-revenue-streams\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":45761,"title":"Tracing The Rising Trajectory Of Voice-Based Application Market In India","content":"Voice-based applications are gaining popularity due to increasing penetration of smart speakers and growing voice searches over smartphones. According to a recent survey, half of the internet users globally use digital voice assistants, and India leads the way with 72% adoption rate. For smart speakers alone, Indian users report a 97% satisfaction rate, which is also the highest in the world. Enhanced Customer Experience Is The Major Factor For Growth One of the most important factors for the adoption of voice-based applications is the requirement that companies have for seamless customer experience. To enhance their growth and increase the customer base, organisations have started implementing voice assistant solutions. According to experts, many organisations in India have realised the value of chatbots in customer experience and therefore instead of just using text chatbots, customers are better served using voice bots bandwagon. That way, customers don’t have to go through the hassle of typing their queries. Indian Enterprises Are Using Smart Voice Applications For Making Processes More Efficient With the advancement in natural language processing, automatic speech recognition, and artificial intelligence, the business use cases for voice-based chatbots have increased over the years. Experts are betting big on voice becoming the preferred mode of transaction for e-commerce, banking, and payments in the near future, streamlining the business process are the key factors driving the voice assistant application market. Lack Of Datasets For Regional Languages Is A Challenge While tech giants see the tremendous value in the voice-enabled smart assistant market, currently the majority of such assistants in India support only English, whereas Hindi and other regional languages remain the dominant form of communication. So if the voice-and-speech recognition market in India has to expand it will have supported by software in multiple regional languages. User Privacy Is Yet Another One 41% of voice assistant application consumers are worried about trust and privacy, according to a report from Microsoft focused on consumer adoption of voice and digital assistants. All the big voice assistant services and apps reportedly hire humans who assess the voice data gathered from end-users. These companies look at data as valuable to train deep learning models and this is a complex problem that needs more attention. User privacy laws such as GDPR restrain tech companies to blatantly collect user data without consent, which is good for individual privacy. But, constant violations of user privacy have damaged the reputation and trust users associate with smart voice assistants. For India, the issue of user privacy does not seem to be a big challenge as found by surveys. Regardless, more user awareness for privacy as well as the upcoming Data Privacy Bill will be a challenge for this technology segment to grow. Overview Some of the early use cases of voice-enabled solutions are limited primarily to seeking information, playing music, accessing customer service, etc. However, with technological advancements in both hardware and software front, one could expect some innovative use cases. Clearly, for a country like India, the technology can make it easy for consumers to interact with applications and web services. This can boost multiple industries, including media, banking, e-commerce, education, healthcare and more.","excerpt":"Voice-based applications are gaining popularity due to increasing penetration of smart speakers and growing voice searches over smartphones. According to a recent survey, half of the internet users globally use digital voice assistants, and India leads the way with 72% adoption rate. For smart speakers alone, Indian users report a 97% satisfaction rate, which is also […]","categories":["AI Features"],"tags":["Alexa","Cortana","NLP","recent technological advancements","Siri","Voice Analytics"],"author_name":"Vishal Chawla","publish_date":"2019-09-10T12:32:55","publication_year":"2019","word_count":522,"keywords":["recent technological advancements","Go","Cortana","artificial intelligence","AWS","AI","chatbots","ML","Voice Analytics","Git","NLP","deep learning","Siri","Rust","Alexa","R"],"extracted_tech_keywords":["AI","artificial intelligence","ML","deep learning","chatbots","AWS","R","Go","Rust","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/voice-smart-assistant-india-nlp-artificial-intelligence\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":14449,"title":"Analytics jobs are Offering the Most to Invite Talent","content":"This article is originally part of our Analytics India Salary Study 2017 The value of analytics was recognized by various business enterprises in a very short spur of time. This sudden change led to a paucity of analytics professionals who could cater to the analytical needs of so many different industries. This has been promising news for the job seekers and analytics aspirants. Also, since the value added by analytics to a business is no secret, it is somewhat obligatory for companies to compete for talent and bring in the best with the most lucrative offer. This has been one of the driving factors to not only differentiate the salaries of analytics professionals when it comes to comparing to other professions, but the growth promised in this field is also one of the highest. This is especially true when we compare analytics to other knowledge-based industries job roles namely Primary Research, Secondary Research, Financial Research and Information Technology. On average, analytics professionals receive around 30% higher salaries than these other job roles in India. Amongst these other job roles, Primary research roles earn the highest average salary (INR 9.03 Lacs) as compared to the analytics job roles where the average salary is INR 11.7 Lacs. Apart from the pecuniary benefits, let’s also see some of the other factors which attract the talent to analytics: More involvement on the business side of the things. All analytics projects, invariably begin with a specific business objective in the sight. A key success driver for any analytics projects is detailed scoping of the business requirements along with good understanding of the business constraints, which is essential for the actionable outcomes. This ensures that Analytics professionals work in tandem with business teams and constantly learn about the new developments, thereby making the whole experience more engrossing and enriching. Exposure to a wide variety of problems and industries. Generally, an analytics professionals – working with a consulting firm or an analytics service providers – do get opportunities to work on variety of business problems and in various domains like, but not limited to, Banking, Financial Services, E-commerce and Retail, Telecom, Hospitality, etc However, this is seldom the case with other career options. This also translates to better exposure and more variety at work and hence builds cross-industry capabilities. All this makes Analytics work profiles intellectually more stimulating. The pace of development of new algorithms has been quite rapid in the past couple of years for this relatively new analytics industry. Analytics professionals are constantly on a learning path exploring new analytics tools and techniques coming up in the market and evaluating it for their current job roles. This is a big factor to attract some of the brightest minds available in the workforce who get a playground to innovate on the methodologies and tailor it to meet their business challenges. If we look at the Indian analytics market, there are broadly three kinds of analytics entities that are present. Analytics service providers– which take outsourced analytics assignments from industrial clients, both within and outside India. Captives– Offshore analytics teams of various international companies but based out of India. This also includes the analytics teams present within all consulting firms, which use analytics extensively for their clients. Domestic in-house analytics teams set up by domestic companies to help meet their analytics requirements We may conjecture from the above, while analytics is the most nascent and yet-to-mature field as compared to other knowledge disciplines, but still, it has created the new jobs at the highest rate in various sizes and types of companies in recent times. With the promising career growth available in analytics, many professionals today are willing to make a career shift even at parallel compensations, in the beginning, to ultimately attain better growth later on and well-rounded professional development. The findings from this year’s salary report further establish this idea more firmly!","excerpt":"This article is originally part of our Analytics India Salary Study 2017 The value of analytics was recognized by various business enterprises in a very short spur of time. This sudden change led to a paucity of analytics professionals who could cater to the analytical needs of so many different industries. This has been promising […]","categories":["IT Services"],"tags":[],"author_name":"Ankita Gupta","publish_date":"2017-04-22T04:55:59","publication_year":"2017","word_count":647,"keywords":["Go","API","programming_languages:R","AI","ML","programming_languages:Go","RAG","analytics","R"],"extracted_tech_keywords":["AI","ML","analytics","RAG","R","Go","API","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/analytics-jobs-offering-invite-talent\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10143303,"title":"The Cost of Artificial Super Intelligence is Now $200","content":"OpenAI’s ChatGPT Pro is sparking debates on its true value. As AI becomes more integrated into our lives, the idea of ‘intelligence too cheap to meter’ feels closer than ever. But for now, as one former OpenAI exec, now in Google, Logan Kilpatrick, said: “consumer willingness to pay for AI is going to go up (a lot).” The next few years will reveal whether this price point is a disruption or just the start of a bigger revolution. For instance, the o1 Pro model, bundled with the $200 ChatGPT Pro subscription, has raised questions about whether the cost is truly justified. However, the bitter reality is that—o1 is not for everyone. “A small percentage of users want to use ChatGPT a TON and hit rate limits and want to pay more for more intelligence on really hard problems. The $200\/month tier is good for them!” said OpenAI chief Sam Altman, clarifying that most users will be best served by the free tier or the $20\/month tier. OpenAI believes that its o1 Pro mode delivers more reliable and comprehensive responses, particularly in fields like data science, programming, and case law analysis. It also outperforms both o1 and o1-preview on challenging ML benchmarks across math, science, and coding. Unlocks Scientific Breakthroughs The model is expected to play a key role in sectors like health sciences, assisting in the search for cures for rare diseases. For instance, Derya Unutmaz, a professor at The Jackson Laboratory, revealed in a post on X that he is using o1 pro for a cancer therapy project. Unutmaz said that o1 Pro helped him propose a groundbreaking idea to simulate T-cell exhaustion in a way that takes into account the time-based progression of the process. He said the model, instead of just looking at a single moment, proposed introducing different stress factors to the tumour at different points in time, giving an analogy of the Battle Royale game, which challenges players to adapt to escalating obstacles. “To be clear, o1 Pro is a totally different beast, requiring different prompting approaches and being useful in different ways from GPT-4 or Sonnet,” Matt Shumer, CEO of HyperWrite AI, said on X. He explained that he first tests whether Sonnet can handle complex coding tasks. If Sonnet struggles or fails, he switches to o1 Pro. He described o1 Pro as a more powerful tool, capable of providing solutions when Sonnet fails, especially in more complex or difficult scenarios. “There were a few really cool instances tonight where Sonnet was struggling heavily, and o1 Pro one-shotted a solution,” he said. Deedy Das, VC at Menlo Ventures, used o1 Pro to solve NYT Connections — a simple game where you group 16 words into four groups of four. “o1 Pro solves it consistently in one shot,” he wrote, on X. OpenAI researcher Noam Brown, in a recent podcast, said that he uses o1 Pro for complicated coding tasks. “If I have something that’s pretty easy, I’ll give it to GPT-4o, but if I have something that I know is really hard or that I need to write a lot of code for, I’ll just give it to o1 and have it do the whole thing on its own,” said Brown. He said an interesting thing about o1 is that it’s a real proof of concept — users can give it a difficult problem, and it can figure out the intermediate steps on its own and how to tackle those steps. However, he added that he doesn’t really know how people will use o1 until it is deployed in the wild. Unlike previous models where the chain of thought was primarily a prompting technique, o1 has been trained with reinforcement learning to apply a chain of thought reasoning without additional prompting.  Moreover, OpenAI recently introduced reinforcement fine-tuning (RFT), allowing organisations to create expert-level AI for tasks in law, healthcare, finance, and more. This method enables training with minimal data, sometimes as few as 12 examples. Unleashes Creativity to a Whole New Level OpenAI launched its highly anticipated video-generation model, Sora, which, according to the company, is a calculated step toward achieving AGI one frame at a time. With an OpenAI Plus account, users get 50 Sora generations per month, and with a Pro account, users get 500 fast generations (or fewer at high resolution) and unlimited generations in a slower mode. This makes the $200 price tag worthwhile. Altman describes the new video-generation model as the ‘GPT-1 for video’ and is eager to see how users will collaborate with each other. “One of the most exciting things to me about this product is how easy it is to co-create with others; it feels like an interesting new thing! This is early – think of it like GPT-1 for video –but I already think the feed is so compelling,” he said. Sora uses credits for video generation, with costs varying based on video resolution and duration. For example, a 5-second 480p square video costs 20 credits, while the same length in 1080p costs 200 credits. Longer videos incur higher credit charges. Features like Re-cut, Remix, Blend, and Loop also consume credits based on video length. ChatGPT Pro users can generate relaxed videos without using credits. ChatGPT Plus offers 1,000 credits for up to 50 priority videos (720p, 5s max), while ChatGPT Pro offers 10,000 credits for up to 500 priority videos (1080p, 20s max) and unlimited relaxed videos. Back in 2023, when OpenAI launched ChatGPT Professional, it was priced at $42. Notably, a possible explanation for the high price is OpenAI’s anticipated $5 million loss this year. Is it Worth Buying? Many quixotic intellectuals believe the price of ChatGPT Pro is justifiable – $200 is the price of unleashing creative freedom and unlocking scientific discoveries,  among other things. Ethan Mollick, associate professor at The Wharton School, who got early access to o1, shared his experience and compared it to Claude Sonnet 3.5 and Gemini. “It can solve some PhD-level problems and has clear applications in science, finance, and other high-value fields. Discovering uses will require real R&D efforts,” he said. He further explained that while o1 outperforms Sonnet in solving specific hard problems that Sonnet struggles with, it doesn’t surpass Sonnet in every area. Sonnet remains stronger in other domains. “o1 is not better as a writer, but is often capable of developing complex plots better than Sonnet because it can plan ahead better,” he said. A Reddit user shared their experience after spending eight hours testing OpenAI’s o1 Pro ($200) against Claude Sonnet 3.5 ($20) in real-world applications. For complex reasoning, the o1 Pro was the winner, providing slightly better results but taking 20-30 seconds longer per response. Claude Sonnet 3.5, while faster, achieved 90% accuracy on these tasks. In code generation, Claude Sonnet 3.5 outperformed o1 Pro, producing cleaner, more maintainable code with better documentation, whereas o1 Pro tended to overengineer solutions. For advanced mathematics, o1 Pro excelled at PhD-level problems, but Claude Sonnet 3.5 handled 95% of practical math tasks perfectly. In vision analysis, o1 Pro stood out with detailed image interpretation, a capability that Claude Sonnet 3.5 currently lacks. When it came to scientific reasoning, the result was a tie – o1 pro offered deeper analysis, while Claude Sonnet 3.5 provided clearer explanations. However, Abacus AI chief Bindu Reddy, in her internal tests, pointed out that o1 lags behind Sonnet and Gemini in coding. “We ran the entire live bench AI coding questions by hand and evaluated o1 for coding. The result is that o1 is good but not as good at coding as Gemini or Anthropic. However, it is an improvement over the o1-preview in this category,” she said. With eight days still left to go of ‘12 Days of OpenAI’ shipmas, ChatGPT Pro users can expect more delights, possibly including advanced voice mode with vision and more.","excerpt":"The cost of artificial intelligence is now $20, and if OpenAI achieves AGI in 2025, it could possibly be priced at $42.","categories":["Deep Tech"],"tags":["OpenAI"],"author_name":"Siddharth Jindal","publish_date":"2024-12-11T16:46:53","publication_year":"2024","word_count":1308,"keywords":["data science","ChatGPT","Anthropic","Go","OpenAI","AI","GPT-4o","ML","chain of thought","R"],"extracted_tech_keywords":["AI","ML","data science","GPT-4o","ChatGPT","OpenAI","Anthropic","chain of thought","R","Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/the-cost-of-artificial-super-intelligence-is-now-200\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10019502,"title":"Budget 2021: Reactions From The Tech Industry","content":"Marking a significant shift in India’s digital journey, the Union Finance Minister, Nirmala Sitharaman, has presented the first-ever digital budget for the upcoming fiscal year starting April 2021. Her budget speech touched upon “proliferation of technologies, especially analytics, machine learning, robotics, bioinformatics, and artificial intelligence.” The budget has come at a time when the country is still struggling with the massive economic slowdown precipitated by COVID pandemic. However, despite this downturn, businesses have seen a significant push towards digitisation, including acknowledging the importance of artificial intelligence across industries. As a matter of fact, India is considered one of the fastest-growing digital markets globally. According to the experts, the technology sector has been quite optimistic about this year’s budget. Companies were expecting the government to create a roadmap to achieve a trillion-dollar economy. With that in mind, the Indian government took several steps to help the tech industry, including setting up a fintech hub at Gift City, encouraging digital payments, using AI in governance, and framing policies for building data parks. Many believe it to be a balanced and far-sighted budget. “The focus on innovation and R&D as an important pillar is a critical step in increasing the Indian IT sector’s export income. Along with this, the Atmanirbhar Bharat budget also outlines initiatives for the gig economy, digital payments, human capital while also setting up a fintech hub and National Natural Language Translation Missions,” said CP Gurnani, MD & CEO of Tech Mahindra. Here are some of the detailed industry reactions — Boost To Fintech The booming digital payments industry has received a budget execution of ₹1,500 crores, critically earmarked to boost the sector and deploy innovative business solutions. The budget has also proposed developing a fintech hub at Gift City to stabilise the economy. Experts believe the budget showcased commitment to a robust digital future and rebuilding the Indian economy. Madhusudan Ekambaram, the co-Founder & CEO, of KreditBee, a platform facilitating loan transactions, said: “The government’s decision to boost entrepreneurship and digital payments in India is a testament to its commitment to realise its vision of an economically self-reliant nation.” He said the COVID-19 pandemic had given a fillip to the adoption of digital payments. “The scheme announced to promote digital payments will help create a robust and dynamic payments ecosystem in India’s emerging digital economy,” said Ekambaram. Echoing similar thoughts, Ankit Ratan, co-founder and CEO of Signzy, said the push for digital payments and a fintech hub’s announcement shows the government’s seriousness in digitalising banking. “We do hope Indian banks will pick on the sentiment and embrace digitalisation of their processes in collaboration with the flourishing fintech ecosystem,” said Ankit. The proposal to use AI, ML and data analytics for the Ministry of Corporate Affairs is an encouraging move for the tech ecosystem. “The suggestion to introduce AI-based features in MCA 21 3.0 to address compliance needs of startups is another thoughtful step in creating a conducive business environment for entrepreneurs,” added Ankit. Sashank Rishyasringa, the co-founder & MD of one of the largest digital lenders in the country, Capital Float, also appreciated the budget. “The proposal to facilitate the development of a world-class fintech hub in Gujarat will be appreciated by fintech locally as well as nationally. Also, the allocation of budget to promote digital modes of payment will be a growth accelerator for the fintech industry,” said Sashank. However, Rajeev Agrawal, CEO & Founder of Innoviti Payment Solution, believes that, while digital payments have become the norm in urban centres and for high ticket purchases, there is still a lot to be done for small value payments in tier-2\/3\/4 cities. “Fast offline payments such as tap-and-go, without the need for any network connectivity at the time of the transaction can multiply the adoption rapidly. And, we hope the budget allocation can find its way into this and other such initiatives.” Encouraging The Use Of AI and ML In this year’s budget, the finance minister has also proposed artificial intelligence and machine learning for the Ministry of Corporate Affairs. This is expected to make regulatory filings frictionless and introduce AI-based features for MCA-21 version 3.0. Some of the key features include online compliance monitoring and e-adjudication. Commenting on this, Shibu Paul, the Vice President – International Sales at Array Networks, said, “The government’s proposal to use data analytics, AI and ML for the Ministry of Corporate Affairs’ database is a boost to the digitalisation.” In agreement, Gaurav Shinh, the founder & CEO of DAAS Labs, said, “The budget has left the data science community quite excited for what is to come. The launch of multiple data analytics, AI, ML-driven models for e-security, e-education, e- consultation, and compliance management is really interesting. Also, the prospect of using artificial intelligence and machine Learning in GST Fraud Tracking is quite motivating and intriguing.” Rajesh Gupta, co-founder and Director, Busy Infotech has also welcomed the budget wholeheartedly. “We are welcoming the decision of laying a special framework of data analytics, machine learning, and artificial intelligence to assist the entire MSME ecosystem and develop them digitally.” However, he also believes stakeholders hoped for more positive outcomes from the government in tax regimes, digitalisation, cash flow, and easing financing norms for MSMEs. The government’s plan for setting up a National Language Translation Mission demonstrates its commitment towards machine translation. On this matter, Rakesh Deshmukh, co-founder and CEO of an Indian app and content discovery platform, Indus OS said that “the announcement is a much-needed effort by the government to reach our citizens in the language they understand.” He believes with an enhanced app store ecosystem, the company will break linguistic barriers and add more value to the next half a billion Indian customers. “Moreover, for Atmanirbhar Bharat to be successful, the focus should be on technology innovation as a whole, and thus we appreciate the government’s focus on innovation and R&D in the budget 2021,” said Rakesh. Ashwani Rawat & Amarsh Chaturvedi, co-founders & Directors of Transerve, appreciated the allocation of ₹8,000 crores for National Mission on Quantum Computing & technology and building data centre parks. Adding to it, Ramesh Mamgain, the Country Manager, India and SAARC, Commvault said the renewed focus on infrastructure would mean accelerated technology adoption, which cannot be accomplished without data privacy measures, propelled by data protection. “This approach would help in strengthening India’s data protection framework to protect individual information, with investments in key technologies like AI and ML to secure cloud-based infrastructures,” said Ramesh. From the cybersecurity perspective, Sunil Sharma, the Managing Director, Sales, Sophos India & SAARC, said, “The budget is built on the foundation of new technologies which will empower businesses with eConsultation, eScrutiny, and compliance management. This will surely enhance enterprise cybersecurity as AI has immense potential to bring in scalable and effective defences against sophisticated attacks like ransomware.” That being said, Sharma is also expecting more push on building skilled cybersecurity professionals in the country. Upskilling Imperative The industry believes the government’s allocation of ₹3,000 crores towards skill development will help reskill India’s youth and, by extension, boost the economy. Surabhi Goel, the CEO of Aditya Birla Education Academy, said the budget has set aside funds to create an opportunity for millennials of India to upskill themselves. Expressing views on this topic, Krishna Kumar, the founder and CEO of Simplilearn said, “We have witnessed the role of technology taking centre stage, opening new job opportunities and increasing the demand for a technically skilled workforce. The government’s decision of introducing post-education apprenticeship with an aim towards the skilling of engineering graduates and diploma holders is a forward-looking proposition aligned with improving employment opportunities for students pursuing different disciplines.” The government’s propositions are in line with enabling a technically skilled future workforce, said Prateek Shukla, CEO and co-founder of Masai School. “The government’s investment towards a robust framework which includes the skilling and upskilling of engineering graduates will aid in paving the path in creating a job-ready workforce in the new normal.” Praveen Tyagi, the founder of STEPapp, said the government understanding the importance of technology in education is immensely going to raise expectations. However, he believes the mindset towards formalising online course development and properly leveraging technology at scale in education is still quite conservative. On the other hand, Sangeet Kumar, the CEO & co-founder of Addverb Technologies, believes that upskilling programs are required to make the workforce ready to work with robots and other automation. The budget also focused on machine learning-driven programmes and setting up data centre parks to promote R&D. “Realignment of National Apprenticeship Training Scheme (NATS) and the allocation of ₹3,000 crores will motivate organisations for providing apprenticeship training. Commencing degree and diploma apprenticeships will improve the employability factor for the youth and prepare a productive workforce in the country,” said Sumit Kumar, Vice President at NETAP, TeamLease. The government is also collaborating with UAE for skill development and deployment, and has taken a few initiatives to maintain and expand India’s research ecosystem by collaborating with Japan and other countries. Driving The Startup Ecosystem The budget 2021 hasn’t left the startup ecosystem behind. It introduced a host of measures to ease operational requirements for startups. The startup ecosystem is also getting capital gains exemption by one more year to 31 March 2022. Commenting on this, Atul Rai, CEO of Staqu, said,“The proposed extension of tax holiday for startups by one more year, will enhance ease of doing business and encourage the next generation of tech companies to step up and carry the mantle of development towards a digital-first future in line with the PM’s Digital India mission.” Agreeing to this, Khadim Batti, the CEO and co-founder of Whatfix said, such a move will go a long way in boosting investments in startups and provide greater benefits to employees. “Employees who join startups at various stages of growth bet on them by accepting more equity and less cash in terms of compensation. Due to this, additional benefits concerning ESOP taxation will facilitate a larger flow of skilled talent to startups, as the majority are still attracted to multinationals and large corporations,” said Khadim. Shubhradeep Nandi, the CEO & Co-Founder, of PiChain Labs, stated that incorporating OPCs without paid-up capital and turnover limitations will allow the startups to grow without compliance challenges and help get into the mainstream quickly. Such moves have allowed more people to look at starting up, and turning their dreams into reality while solving real-world problems, said Mishu Ahluwalia, COO & Investor Relations at KIWI. “Recent announcement of the Startup Seed Fund Scheme (SISFS) has already boosted the morale of the startup ecosystem, and we are hopeful for the coming year,” said Sharad Bhatt, CFO, Eolstocks.com. However, Mitesh Gangar, the co-founder & Director of PlayerzPot Media, said, while the budget is a welcoming step, India’s startups expected a few more tax front reforms and compliance for the ease of doing business.","excerpt":"Marking a significant shift in India’s digital journey, the Union Finance Minister, Nirmala Sitharaman, has presented the first-ever digital budget for the upcoming fiscal year starting April 2021. Her budget speech touched upon “proliferation of technologies, especially analytics, machine learning, robotics, bioinformatics, and artificial intelligence.” The budget has come at a time when the country […]","categories":["IT Services"],"tags":[],"author_name":"Sejuti Das","publish_date":"2021-02-02T16:00:00","publication_year":"2021","word_count":1812,"keywords":["data science","machine learning","artificial intelligence","AI","ML","RAG","Ray","Aim","analytics","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","data science","analytics","Aim","Ray","RAG","R"],"url":"https:\/\/analyticsindiamag.com\/it-services\/budget-2021-reactions-from-the-tech-industry\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10092224,"title":"How Enterprises Can Save Their Data from Gen AI Black Hole","content":"Recently, Samsung employees came under fire for leaking confidential company information through an unorthodox method: ChatGPT. Since then, companies handling sensitive information have sent out memos cracking down on employees’ usage of ChatGPT, but OpenAI’s chatbot chaos is just about beginning. Generative AI has a lot of potential to disrupt the enterprise space, so it’s no wonder that this technology is now beginning to seep into companies’ tech stacks. However, as with any emerging technology, it comes with its own share of pitfalls, the biggest among them being data security. For enterprises that still wish to capitalise on the positives of generative AI, such as auto-coding platforms (Amazon CodeWhisper, Replit Ghostwriter, GitHub Copilot, etc), realistic text-to-speech algorithms (ElevenLabs Prime AI voice, OpenAI Whisper etc.), and image generation (DALL.E, Midjourney, Stable Diffusion, etc), there are a few steps they can take to ensure the safety of their data Safeguarding data Opacity of data policies from AI vendors presents a big challenge for companies that wish to leverage generative AI. Before feeding private information into any generative AI, users must first have a complete idea of where the data is going. Robert Blumofe, executive vice president and CTO at Akamai, stated, “Nobody should use or provide private information to a tool like this until you have a clear statement from the vendor on how they will use the information. Do they store, save, or share it? And does it become available to the public copy of the tool?” There are various levels of data leakage associated with AI services. Maintaining logs, storing, or saving inputs from companies is the least risky, but indexing this information as training data is the riskiest. This means that the data could end up in a public version of the same tool, causing data leaks. For companies compliant with the Global Data Privacy Regulation (GDPR), handing over data to any non-transparent AI data provider is a big no-no. Some GDPR regulators have also gone so far as to take a hardline stance against AI tools until they are safer with data, as seen by Italy’s ban of ChatGPT. Hence, the first step to ensuring data security with any AI vendor is to do due diligence and find a statement that clearly describes how the data will be used. Decision-makers must look at statements such as this one provided by Amazon Bedrock, which clearly states that user data will not be used and will not leave their private cloud. Self-hosted models vs responsible vendors While it may be tempting to just go with the pack and get a subscription to OpenAI’s APIs, the generative AI field is vast and growing quickly, especially when it comes to enterprise-focused use cases. Many key decision makers have also raised concerns as to how the technology can be used in their organisation. Brandon Jung, VP, ecosystem and business development, at Tabnine, told AIM, “People say, if we have a model, the model has to live in the same place and have the same level of security as our code, understandably. Lot of companies might first install and use the enterprise models, and then do custom models.” To understand how much of an impact a self-hosted AI can have, we can look at companies providing AI training services, such as Stability AI, one of the prominent open-source model providers in the AI field today. Emad Mostaque, the CEO of Stability AI, stated, “Dozens of major companies want models they own based on their own data and want to pay us to do it [train them]… It’s a pretty good model. We also build custom models for large companies and governments.” These kinds of fine-tuned models might be a good in-between for the security provided by self-trained and self-hosted models and the convenience of off-the-shelf APIs, similar to RedHat’s business model. There is also an alternative in services like Amazon BedRock or Azure’s OpenAI cloud services, which maintain the data privacy of the underlying cloud services while allowing companies to adopt generative AI. Even as all these advancements take place, one thing is clear: The AI cat is out of the bag. There is no going back for the enterprise, as the impact of AI goes beyond generating media. Whether it is automating repetitive tasks or transforming the data warehousing process, enterprises have to adapt to AI or fall behind. With the number of choices in the field and the current pace of innovation, decision-makers must keep data security in mind when picking AI solutions.","excerpt":"Changing data norms should be the first step for enterprises entering AI","categories":["AI Features"],"tags":[],"author_name":"Anirudh VK","publish_date":"2023-04-24T16:00:00","publication_year":"2023","word_count":751,"keywords":["Go","ChatGPT","OpenAI","AI","R","Git","RAG","Aim","generative AI","Azure"],"extracted_tech_keywords":["AI","generative AI","ChatGPT","OpenAI","Aim","RAG","Azure","R","Go","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-enterprises-can-save-their-data-from-the-gen-ai-black-hole\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":33729,"title":"How Intel&#8217;s Data Centre Business Is Being Affected By India&#8217;s Digital Transformation Wave","content":"Image source:  https:\/\/www.intel.com\/content\/dam\/jobs\/images\/photography\/3×1\/jobs-sites-intel-sign-3×1.jpg.rendition.cq5dam.webintel.960.320.jpg With digital transformation driving the growth for companies in India, enterprises are heavily investing in data centres across the country for large scale adoption of technologies like artificial intelligence, cloud and data analytics, among others. Last year, Intel saw exponential growth in India owing to large scale adoption of cloud and AI by enterprises in the country. In India, Intel has been increasing their data centre strategy since 2011 when its PC market witnessed a slump with the onset of mobile penetration. Ever since the company has found a steady market with India presently forming 1-2% of its overall data centre business. Globally, in Q3, Intel’s data-centric businesses grew 22% year-on-year and the Data Centre Group (DCG) vertical clocked revenue of $6.1 billion. While in 2018, it stood at $6 billion in the third quarter with a growth of 26% year-on-year. Mobile Consumption And AI Driving The Market One of the primary reason for its growth in India has been attributed to the steady increase in mobile data consumption, pushing cloud and communication service providers to turn to data centres for quicker and faster processing of data-intensive work that relies on technologies like AI and ML. Speaking to a business daily about how AI and analytics are driving the change in India, Rajeeb Hazra, corporate vice president, DCG at Intel, said, “We see tremendous opportunities in India for data, bound together by cloud-connectivity, which makes India a strategic region for our growth. The AI and analytics journey are just at the beginning and there is immense opportunity to derive value from the rapidly increasing volume of data.” Realising the potential of how AI and analytics can drive the future of data centres in India, Intel is currently developing tools and software to enrich their AI-ecosystem in an attempt to gain insight from data at the enterprise level. “Our focus in the near future is to build our data centre business through multiple strategies, and grow our cloud business in the space of computing, networking, storage, and memory solutions,” he said highlighting the company’s future plans with regards to data centre. India Connect With its data centric-business estimated to grow by $200 billion in 2022, the company is entering a new era of data-centric computing in India and has increased its investment to the country substantially. In its two-decades of functioning in India, Intel has invested over ₹30,000 crore in the country. With the introduction of Xeon Scalable in India, Intel hoped to make digital transformation for enterprises easier. Equipped with the new processor that enabled faster deep learning performance, the company received more than $1 billion in revenue from customers running AI on Intel Xeon processors in the data centre. Added to this its tie-up with the Indian government for various government initiatives. In 2015, it partnered with Centre for Development of Advanced Computing (C-DAC) to launch ₹4,500 crore National Supercomputing Mission over the course of the next seven years. Although the project hasn’t picked up steam so far, as per the initial reports the chip-maker was entrusted with the duty of establishing supercomputers at three IITs in a move to “enable India to leapfrog to the league of world-class computing power nations.” Further, as part of its move to affirm its footprint in India, Intel unveiled SRR4, a design facility in Bengaluru with an investment of approximately ₹1,100 crore. The new facility with 620,000 sq. ft. of space including 100,000 sq. ft. of lab infrastructure, has been constructed with an aim to strengthen Intel’s research and development in India. Briefing about the centre Nivruti Rai, Country Head of Intel India and VP of Data Center Group, Intel Corporation, said, “Intel India plays a strategic role in Intel’s growth by making significant contributions to Intel’s technology and product portfolio in areas such as cloud, client, graphics, artificial intelligence, 5G and advanced driver assistance system.” Outlook One of the leading factors for Intel’s data centre growth in India will be the digital transformation and it will continue to fuel its growth. With 85% of the enterprises estimated to use AI in some way for the other globally, Indian enterprises cannot miss out on this tectonic AI shift. Massive AI adoption along with migration to the public cloud will continue to spurt Intel’s business. Though Intel holds 95% of the data centre market, the mushrooming of competitors like Nvidia and Advanced Micro Devices (AMD) could pose a challenge. AMD recently in 2017, announced the launch of EPYC server chips for that is now used in the public data centres of Tencent, Microsoft Azure, and Baidu.","excerpt":"With digital transformation driving the growth for companies in India, enterprises are heavily investing in data centres across the country for large scale adoption of technologies like artificial intelligence, cloud and data analytics, among others. Last year, Intel saw exponential growth in India owing to large scale adoption of cloud and AI by enterprises in […]","categories":["Global Tech"],"tags":["AI in India","data centre india","digital transformation","Intel"],"author_name":"Akshaya Asokan","publish_date":"2019-01-18T10:34:43","publication_year":"2019","word_count":766,"keywords":["data centre india","artificial intelligence","AI in India","AI","R","digital transformation","ML","RAG","Aim","deep learning","analytics","Tecton","Azure","Intel"],"extracted_tech_keywords":["AI","artificial intelligence","ML","deep learning","analytics","Aim","Tecton","RAG","Azure","R"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/how-intels-data-centre-business-is-being-affected-by-indias-digital-transformation-wave\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":33797,"title":"10 Sure Shot Takeaways For Every Data Science Professional From Machine Learning Developer Summit","content":"Analytics India Magazine is hosting one of India’s biggest Machine Learning conferences, scheduled to occur on January 30th and 31st. The Machine Learning Developer Summit aims to bring together data scientists and machine learning developers from all across the country. With the presence of top ML innovators from leading tech companies, MLDS is all set to talk about the most recent developments in machine learning and its accompanying use cases. While attendees will not only learn a lot from these ML experts, they can inculcate important skills referring to data sciences and interact with like-minded individuals to add value. With that being said, here are 10 sure shot takeaways for every professional attending MLDS 2019. Know more about the event. 1| Foundation Of ML – Building Your Data Lake On AWS In a talk by Raghuraman Balachandran, the Solutions Architect at Amazon Web Services, attendees will learn about how to build clean data lakes on the software. Since they are one of the most common architectures required for data-driven organisations, this will provide an opportunity to have a hands-on experience on the concept, thereby increasing the chances of hiring. Over the course of the talk, Mr Balachandran will speak about how to store unstructured, semi-structured, or fully-structured raw data as well as processed data for different types of analytics. He will also provide his experienced perspective on common challenges and patterns for designing an effective data lake on the AWS Cloud. Key Takeaways: Learn how to harness AWS’ power for the creation of a dependable and useful data lake, straight from the horse’s mouth based on experience with clients. 2| Accelerate Inference For Deep Learning Models Deep Learning models are extremely resourced intensive, with requirements for high-end GPUs and CPUs for optimal training. However, this still creates opportunities for clashes, as the target deployment environment introduces various challenges that are typically not present in the training environment. In this talk, Sunil Kumar Vuppala, the Principal Scientist at Philips Research, will speak about methods to increase the performance of trained DL model during the inference phase. He will also present Intel’s inference engine as one option to run DL models in Caffe\/ Tensorflow\/ Keras using CPUs. Key Takeaways: Learn to use various methods to improve DL models’ performance during the inference phase using Intel-powered tech to ensure the quality of output in the final product. 3| With The Explosive Growth In Demand For AI Skills, How To Be More Employable Due to the highly disruptive nature of AI, many analysts expect explosive growth in demand for AI skills. However, this is sorely lacking, as it is not occurring at the anticipated rate. Amitabh Mishra, the CTO at Emcure Pharmaceuticals, will speak upon his reflections in the AI field and why it is not occurring. He will also offer tips for preparation for AI professionals, and elaborate upon the requirements in AI and related fields. Key Takeaways: Learn how to be more employable as an AI professional and how to evolve in relation to evolving market conditions. 4| Walmart AI Journey – Use Case For Building Technologies For Merchants This talk by Prakhar Mehrotra, Senior Director at Walmart Lab, will offer a perspective into Walmart’s movement in India. Walmart Labs is a subsidiary of Walmart which is aimed at providing a better experience for customers of the chain worldwide. The organisation also builds tech for use with Merchants to increase Walmart’s range. Mr Mehrotra will speak about Walmart’s journey into the field of AI and offer insights into the advancements that the chain has made thus far. Key Takeaways: How to build AI and implement it at scale for use in low-infrastructure environments. 5| How Important Is Data Cleansing And How ML Can Help In It? A clean data lake is important for the production of a good ML model. Around 40 percent of the time and resources spent on creating a new ML model comprise of data cleansing. Somu Vadali, Chief of Data and Products, CnD Labs at Future Group, will speak about how data cleansing can be done more efficiently with ML. He will also elaborate upon the need for the method, and streamlined, thought-through processes that can enable organisations to go quickly and reliably from raw data to features. His talk will also help organisations cut down their time-to-market with every subsequent new model or version of the existing model. Key Takeaways: Learn how to establish methods for clean data, thus reducing time to market and increasing the efficiency of model creation, and other industry-best practices like data quality, validation, guardrails and more 6| The Use Of Augmented Genetic Algorithm As every data scientist is aware, feature selection in large dimensional spaces is a staple of machine learning. Therefore, Sidharth Kumar, the Staff Decision Scientist at e-tailer Flipkart, will speak about a proposal to use genetic algorithms to solve this problem. He will also offer transparency into his efforts to significantly reduce the number of generations required to convergence. Key Takeaways: Demonstration by Kumar on the use of this algorithm in conjunction with SVMs, solving two problems with one solution. 7| A Primer On Building And Deploying Scalable Analytical Models In Production Subramanian MS, the Head of Analytics at BigBasket.com, will reveal his methods to plan a scalable set of tools for ML models. This includes planning a lifecycle for analytical models, a scalable software and hardware layer, along with industry best practices for scalable DevOps. Key Takeaways: Learn how to scale your startup\/entrepreneurial idea effectively at every level from a company undergoing growth at scale. 8| How To Explain ML Models ML models have been historically difficult to explain or ‘pitch’, owing to the fact that they are black box algorithms. Owing to the increasingly obsolescent state of linear models to solve predictive modelling problems, we are now adopting ML models. This makes it an important part of the evolving corporate setting. Indranath Mukherjee, the Head of Strategic Analytics at AXA XL, will speak about how the analytics industry has begun to move to ML models for said problems because of their low bias and higher predictive power. Key Takeaways: Learn how to pitch an idea for an ML model by explaining what is effectively a black box algorithm for use in a business setting from one of the keynote speakers. 9| How AI Is Shaping The Future Of Video Streaming Over the top streaming has effectively disrupted existing technologies such as cable television, as seen by the meteoric growth of companies such as Netflix. Anubhav Shrivastava, the Head of Product Data Sciences at Voot, will speak about how AI techniques are disrupting this business itself. He will also offer examples of such implementations, such as image and video recognition, recommendation engines, media databases, and metadata generation. He will also reveal information on plans for intelligent adaptive bitrate streaming, self-reliant marketing automation plans and predicting catastrophic failures. Key Takeaways: Take a look into how a quickly growing OTT media platform is utilising AI, and how disruptions can occur in media, along with technical details. 10| Named Entity Recogniser – Practical Challenges And Strategies This talk by Mathangi Sri, the Head of Data Science at PhonePe, will reveal how supervised algorithms are solving efficiently most of the classification problems in the text. However, this also comes with its shortcomings, such as entity recognition. Entity recognition is still an area of craft and domain knowledge, thus making Sri’s talk more valuable. The speech will look at how to handle entity recognition, including the stack of techniques, their applications, strategies and real-life challenges while applying those. Key Takeaways: An opportunity to gain a skill set regarding entity recognition, a closed field. Examples might be provided in an indigenous setting, seeing as the talk is by PhonePe. Click here to register.","excerpt":"Analytics India Magazine is hosting one of India’s biggest Machine Learning conferences, scheduled to occur on January 30th and 31st. The Machine Learning Developer Summit aims to bring together data scientists and machine learning developers from all across the country. With the presence of top ML innovators from leading tech companies, MLDS is all set […]","categories":["Deep Tech"],"tags":["MLDS","speech","Walmart Labs"],"author_name":"Anirudh VK","publish_date":"2019-01-21T10:12:58","publication_year":"2019","word_count":1299,"keywords":["data science","machine learning","speech","Keras","AI","ML","Walmart Labs","RAG","MLDS","Aim","deep learning","analytics","TensorFlow"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","data science","analytics","Aim","TensorFlow","Keras","RAG"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/10-sure-shot-takeaways-for-every-data-science-professional-from-machine-learning-developer-summit\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10086863,"title":"A Toast to Lossless Lempel: The Pioneer of Data Compression Will be Missed","content":"On February 5, Israeli computer scientist Abraham Lempel, father of the lossless data compression algorithm, passed away at the age of 86. Lempel, along with Jacob Ziv and Terry Welch, had built a universal lossless data compression algorithm in 1984 called the Lempel-Ziv-Welch (LZW), an improved version of the LZ78 algorithm that had been published in a paper co-authored by Lempel and Ziv in 1978. If all this sounds too obscure, the value of Lempel’s contributions can be measured by the ubiquitousness of the LZW algorithm in modern technology. It wouldn’t be such a stretch to say that some of the biggest tech companies would have struggled to exist without Lempel’s data compression techniques. The bigger the organisation, the more the need for data compression – Tech giants like Alphabet, Facebook, Spotify or even Apple for iTunes make the most use these algorithms. Social media companies like Facebook or Instagram have to deal with massive loads of user uploads and need to compress data while Google also uses compression software to back up its memory. Implementation of LZW data compression algorithm, Source: Semantic Scholar Why is data compression needed? The need underlying data compression itself is pretty simple – it allows organisations to maximise the amount of data they have while being able to manage it more easily by minimising the relevant storage space and costs. But how exactly does it do this? Data compression can be categorised by the two different techniques through which it is performed – lossy and lossless. Essentially, both of them have the same goal – to search for duplicate data in a graphic (mostly GIFs for LZW) and replace it with a much more compact data representation instead. The lossless algorithm searches for statistically redundant bits and reduces them, while on the other hand, a lossy algorithm simply picks out the irrelevant parts of information and removes them. Compression that removes blank values (white) and encodes repetitive values. Source: Faust 2013. While data compression gained more importance in computing, starting from the 1970s, there is a much bigger history behind it even outside of computing. Even the Morse Code, which came about in 1838 is one of the earliest illustrations of data compression. Currently, compression algorithms are implemented in multiple file formats like GIFs for image compression, PDFs for file compression and MP3 for audio compression apart from digital cables and satellite television. But even saying this would be to undersell the importance of data compression – the LZ algorithm is found virtually in every modern computer in software or hardware and at times in both. While using it normally, there isn’t a full awareness that we are using it because we use it so much – while archiving files, installing software on a disk, backing up hard drives or even upgrading the quality of cinema cameras to 4K. Lempel’s algorithm Both of Lempel’s algorithms, the LZ77 and LZ78, became popular and inspired different variants of them within a short span. But most of these variants like DEFLATE, LZMA or LZX went away just as quickly. This wasn’t particularly because it was technically proficient, but because the LZ77 was patented soon after. Notably in 1993, Stac Electronics built another version of the LZW, called the LZS or the Lempel-Ziv-Stac, to use it in their disk compression software Stacker. The LZS was known to be pretty fast and was picked up by Microsoft while creating a disk compression software with MS-DOS 6.0 which would double the capacity of a hard drive. Stac Electronics ended up suing Microsoft for using its algorithm illegally after which the tech giant eventually had to fork out USD 120 million for patent infringement. The patent on the LZW algorithm expired in 2003 ending most of the legal issues that surrounded it but the algorithm stuck and is still typically used in GIFs and PDFs. Lempel, who was a resident professor at Technion, the Israel Institute of Technology, continued his academic career quietly for over 40 years. While teaching both electrical engineering and computer science at the university, he served as the head of Technion’s computer science department between 1981 and 1984. Professor Lempel at a conference at Technion University, Source: Technion Life and achievements In 1993, Lempel joined the research unit for HP called Hewlett-Packard Labs. He worked to bring a sister facility to Israel. Having established HP Labs in his country, he served as the director of the division until October 2007 and went on to work quietly around the development of basic and universal image processing tools and customised applications. Lempel was also an IEEE or Institute of Electrical and Electronics Engineers Fellow and a Senior Fellow at HP. Lempel has eight US patents, plus more than 90 published research papers on data compression and information theory, the study of mathematical concepts and parameters around message transmission in communication. In 2004, the IEEE Executive Committee and History Committee named the LZ algorithm as a milestone in electronics. Lempel himself was awarded the Golden Jubilee Award for Technological Innovation in 1998 and the 2007 Richard W. Hamming Medal by IEEE for the LZW algorithm. It pays to commemorate the value that pioneers brought to the field and it is safe to say that Lempel’s algorithm will be alive for a long, long time.","excerpt":"Lempel’s algorithms, LZ77 and LZ78, became popular inspiring different variants in a short time but most of them like DEFLATE, LZMA or LZX went away just as quickly","categories":["IT Services"],"tags":[],"author_name":"Poulomi Chatterjee","publish_date":"2023-02-08T15:00:00","publication_year":"2023","word_count":884,"keywords":["Go","programming_languages:R","AI","ETL","innovation","programming_languages:Go","Git","RAG","GAN","R"],"extracted_tech_keywords":["AI","RAG","R","Go","Git","ETL","GAN","innovation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/a-toast-to-lossless-lempel-the-pioneer-of-data-compression-will-be-missed\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10169502,"title":"Baidu’s Apollo, CAR Inc Launch China’s First Autonomous Car Rental Service","content":"Apollo, Baidu’s autonomous driving unit, and CAR Inc, a Chinese auto mobility provider, have announced the launch of China’s first autonomous car rental service. The service will begin rolling out in the second quarter of this year. It will allow users to reserve driverless vehicles for flexible travel to cultural landmarks and tourist spots. The initiative addresses the growing demand for on-demand transport options. It aims to support a wide range of users, including the elderly, international tourists, unlicensed drivers, and individuals with disabilities. The partnership also marks a step toward broader commercial deployment of autonomous vehicle technology in real-world mobility scenarios. “This strategic partnership between Apollo and CAR Inc represents an in-depth integration of our technological expertise with real-world application scenarios,” said Yunpeng Wang, corporate VP of Baidu and president of the company’s intelligent driving group. The partners plan to refine the service based on user feedback and expand to more cities over the next three to five years. They also aim to establish industry standards for autonomous rentals, focusing on safety, reliability, and stability. The service will leverage Apollo’s more than a decade of R&D in autonomous driving and CAR Inc’s national network. It targets a car rental market expected to exceed 300 billion RMB (approximately $41 billion) by 2030. Recently, Waymo and Toyota also signed a preliminary agreement to explore a potential partnership to develop a new self-driving vehicle platform. To expand in Tokyo, Waymo partnered with Tokyo’s largest taxi operator, Nihon Kotsu, and the ride-hailing service GO. As for India, in an exclusive interview with AIM, Bengaluru’s JCP (Traffic), MN Anucheth, shared his belief that while self-driving vehicles and Level 5 autonomous cars could theoretically navigate the city’s chaotic roads, their practicality and adoption remain uncertain.","excerpt":"It will allow users to reserve autonomous vehicles for flexible travel to cultural landmarks and tourist spots.","categories":["AI News"],"tags":["AI for cars","autonomous car","autonomous car technologies","autonomous driving","Baidu"],"author_name":"Sanjana Gupta","publish_date":"2025-05-09T14:49:32","publication_year":"2025","word_count":290,"keywords":["Go","programming_languages:R","AI","R","programming_languages:Go","autonomous car technologies","RAG","ai_applications:autonomous driving","Aim","Baidu","AI for cars","autonomous car","autonomous driving"],"extracted_tech_keywords":["AI","Aim","RAG","R","Go","programming_languages:R","programming_languages:Go","ai_applications:autonomous driving"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/baidus-apollo-car-inc-launch-chinas-first-autonomous-car-rental-service\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10047912,"title":"Robotics &#038; Automation Firm Addverb Technologies To Further Expand Globally","content":"Robotics and Automation company Addverb Technologies is further expanding its presence across Europe, Singapore, and Australia. This strategic initiative stems from increased demand for robotics and automation solutions in the global market. “We are glad to announce that we are taking Addverb Technologies global in the international robotics and automation marketplace. With this expansion, we aim to fully exploit the potential across the European, Australian, Asian and American markets, providing an end-to-end product portfolio powered by AI, Machine Learning, Deep Learning, Computer Vision & IoT, which will streamline warehousing and factory operations and will help customers create warehouses with higher levels of efficiency, accuracy, and flexibility,” said Sangeet Kumar, CEO, Addverb Technology. Established in June 2016, Addverb has clocked an approximate revenue of around 27 Million USD in FY 2020-21. The company inaugurated a state-of-the-art manufacturing facility, “Bot-Valley”, which also houses a well-equipped R&D facility at Noida in March 2021. The company with one of the widest Robotic and Automation product portfolios has more than 100 customers, including Reliance, Flipkart, Amazon, Hindustan Unilever, Coca-Cola, PepsiCo, Future Supply Chain, Marico, etc. The company is fueled by the power of its robust R&D and intellectual properties. They have obtained 12 patents & 10 designs and are planning to file for another 20 patents in three months. In addition, they have worked on complex research problems with institutes like IIT Delhi and aim to establish an innovation lab in the US and Europe to create a global workforce based out of different parts of the world. The company will also open an office in the US in the next couple of months. Some of their products include AMRs (Dynamo), Sorting Robot (Zippy), Robo-Shuttle (Veloce), Carton Shuttle (Quadron), Pallet Shuttle (Cruiser), ASRS (Multi-Pro), Pick-To-Light (Rapido), and cutting software like Warehouse Control System (Mobinity), Warehouse Management System (Optimus), Robot Fleet Management Software (Legion), and Pick-By-Voice (Khushi).","excerpt":"Addverb Technologies has created global teams in Singapore, Australia, and Europe and is looking to start operations in the US before the end of 2021. The company aims to be in the top five global robotics companies by 2025.","categories":["AI News"],"tags":[],"author_name":"kumar Gandharv","publish_date":"2021-09-07T19:00:00","publication_year":"2021","word_count":311,"keywords":["API","machine learning","AI","innovation","ML","computer vision","automation","Aim","deep learning","R"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","computer vision","Aim","R","API","automation","innovation"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/robotics-automation-firm-addverb-technologies-to-further-expand-globally\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10104082,"title":"Modular Announces Partnership with AWS, NVIDIA","content":"At ModCon 2023, Modular, the company behind the Mojo programming language, announced an exclusive collaboration with Amazon Web Services (AWS). The partnership aims to extend the reach of the MAX Platform to AWS production services worldwide, ushering in advanced AI capabilities for a vast user base. Check out all the key announcements made at ModCon 2023. Notably, the AWS Marketplace becomes the exclusive venue for leveraging the MAX Platform on Graviton CPUs, Amazon’s ARM-based processors designed for compute-intensive workloads such as AI programs. The MAX Platform optimally enhances Graviton CPUs, delivering AI model execution with up to 5X higher performance and up to 80% cost savings compared to existing AI infrastructure. Bratin Saha, AWS VP of machine learning & AI services, emphasised the significance of this collaboration in advancing AI capabilities. “At AWS, we are dedicated to shaping the future of AI by delivering services that reduce costs and accelerate progress for enterprises and startups. The MAX Platform amplifies this objective for millions of AWS customers, facilitating the rapid deployment of the latest GenAI innovations and traditional AI use cases,” remarked Saha. One unique feature of the MAX Platform is its hardware portability, enabling seamless migration of workloads to Graviton without incurring any migration costs. This portability extends to the utilisation of MAX Serving and Mojo directly in the engine, offering full customizability and workload tuning options. Furthermore, Modular announced a strategic technology partnership with NVIDIA to integrate the benefits of their accelerated compute platform into MAX, simplifying CPU+GPU development for AI developers. Key features unveiled include MAX Engine extensibility with Mojo, MAX Engine GPU support, the release of Mojo SDK v0.6, and the open-sourcing of Mojo documentation. MAX GPU offers cutting-edge compatibility for NVIDIA H100, H200, A100, and L40 series GPU accelerators, as well as the recently introduced Grace CPU Superchip. This hardware integration extends across all facets of MAX, encompassing MAX Engine, MAX Serving, and Mojo, delivering unmatched heterogeneous computing capabilities to the field of AI. Developers can now leverage a unified toolchain that caters to a spectrum of AI applications, spanning from GenAI to various other AI scenarios. This approach unlocks innovative CPU+GPU programming models, promising unparalleled performance and cost-effectiveness. “Developers everywhere are helping their companies adopt and implement generative AI applications that are customised with the knowledge and needs of their business,” said Dave Salvator, director of AI and Cloud at NVIDIA. ModCon 2023, Modular’s inaugural developer conference, featured prominent figures in the AI landscape, including Bratin Saha (VP, AWS), Kari Ann Briski (VP, NVIDIA), Bryan Catanzaro, (VP, NVIDIA), Lex Fridman, (AI researcher, MIT), Shawn “Swyx” Wang (Latent.Space), Damien Sereni (director, Meta), Jeremy Howard (Fast.ai), and Michele Catasta, (VP, Replit). Excitement surrounds this collaboration with early access to MAX on AWS Marketplace available at modul.ar\/max, and further updates anticipated in Q1 2024. The partnership signifies a significant step towards ensuring the widespread availability of MAX, marking a pivotal moment in the evolution of AI capabilities.","excerpt":"ModCon 2023 was Modular’s first developer conference.","categories":["AI News"],"tags":["modular"],"author_name":"Mohit Pandey","publish_date":"2023-12-05T11:08:47","publication_year":"2023","word_count":489,"keywords":["API","GenAI","machine learning","AWS","AI","ML","RAG","Aim","modular","generative AI","R"],"extracted_tech_keywords":["AI","machine learning","ML","generative AI","GenAI","Aim","RAG","AWS","R","API"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/modular-announces-partnership-with-aws-nvidia\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10118578,"title":"Ola Krutrim to Launch AI Cloud Next Week","content":"Ola chief Bhavish Aggarwal has announced that Krutrim is set to open its AI cloud, operating from India, to developers starting next week. Aggarwal said that the chat app, Krutrim, relies on this proprietary AI cloud infrastructure. Our @Krutrim chat app runs on our own AI cloud based in India. Planning to open the AI cloud up for developers next week.If you’re an AI developer, do let us know in comments if you want some free credits! Tell us about your company and what your AI use case is.— Bhavish Aggarwal (@bhash) April 18, 2024 Aggarwal  also invited  AI developers to express their interest in the comments section, offering free credits as an incentive. Developers are encouraged to share details about their companies and the specific AI use cases they are working on. “If you’re an AI developer, do let us know in the comments if you want some free credits! Tell us about your company and what your AI use case is,” wrote Aggarwal. This comes after Aggarwal recently revealed that Krutrim has achieved a major breakthrough by operating on its independent cloud infrastructure, signifying a move away from reliance on external cloud providers like AWS or Azure. He emphasised ongoing efforts by the Krutrim team to enhance both the model itself and its infrastructure. Ola Krutrim recently announced that it is set to launch its standalone mobile app. The app comes with significant advancements, including a notable decrease in the Time it Takes to Generate the First word of a response (TTFT), from 22 seconds at its initial launch to a swift 0.3 seconds presently, with further enhancements anticipated. Aggarwal also hinted at forthcoming improvements in an upcoming detailed blog post. Intel recently disclosed that Ola Krutrim is leveraging Intel Gaudi 2 clusters for pre-training and fine-tuning its foundational models, boasting industry-leading price\/performance ratios across ten languages. Moreover, Krutrim is actively pre-training an expanded foundational model on Intel Gaudi 2 clusters, further elevating its AI capabilities. A few days ago, Krutrim announced its partnership with Databricks to improve its foundational language model, particularly for Indian languages, aiming to enhance AI solutions in India. “The Krutrim model was launched using our platform,” said Naveen Rao, VP of generative AI at Databricks, during an exclusive interview with AIM. Ola Krutrim has been quite obsessed with developing its own foundational model from scratch, despite rumours that it is being built on fine-tuned models such as Llama-2, Mistral, Claude-3 or even the most recent, DBRX. Launched in December last year, Krutrim has been lauded as “India’s first full-stack AI” solution, showcasing prowess in understanding and generating content across multiple Indian languages, including Marathi, Hindi, Bengali, Tamil, Kannada, Telugu, Odia, Gujarati, and Malayalam, with claims of superiority over GPT-4 in Indic languages.","excerpt":"Offers Free Credits","categories":["AI News"],"tags":["Ola Krutrim"],"author_name":"Siddharth Jindal","publish_date":"2024-04-18T17:50:06","publication_year":"2024","word_count":459,"keywords":["Go","AWS","AI","R","RAG","GPT","Aim","generative AI","Ola Krutrim","Azure","Databricks"],"extracted_tech_keywords":["AI","generative AI","Aim","RAG","AWS","Azure","Databricks","R","Go","GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/ola-krutrim-to-launch-ai-cloud-next-week\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":7419,"title":"Big Data-The new age ‘Desire’","content":"It’s not a mighty astonishment that among the very new faces debuting as billionaires for the year 2015, are the ones who have made riches from the Big Data. Their empires are built solely on their prowess and expertise to collect, interpret and use data in such novel ways that no one in the past had ever visualized. Their cognition, acumen, and sagacity to exploit Big Data have proven to be a huge success. Whether it’s for combating financial frauds, terrorism , handling businesses or improving the mundane aspects of our day to day life Big Data is surely a ‘Big Deal’ but in a highly affirmative way. The solution is Big Data, wait! What’s the Question? The data flow from our daily life: credit cards, phones, computers television; from sensor equipped buildings buses, factories, planes, bridges and trains; from the infrastructure of e cities is so fast and pell-mell that its entire accumulation is of a zettabyte just in the past two years. There is such a big data revolution that is nearly dwarfed of even the records of human civilization. But, it’s not the quantity of the data that is revolutionary; the big data revolution is that, what and how productive things can be done with this colossal data. It is also believed that the potential benefits of Big Data to society go far beyond what has been accomplished so far. Among the students across the globe there is a huge appetite for this new field. It is estimated that a Harvard course in data science last fall enticed nearly 400 students from different disciplines including, schools of government, law, business, medicine, design, from degree Colleges, School of Applied Sciences and Engineering and even MIIT. Big Data –Sizzling with fortune! Big Data is growing huge and heavy, in this scenario most companies globally are in the hunt for the most accomplished data scientist and engineer. The increases is as massive as 300% since the year 2013.Moreover it is estimated  that by the year 2018, there will be a exorbitant shortage of 1.5 million skilled managers\/analysts who are Big Data literate according to the Mckinsey Global Institutes Report on Big Data. And India alone will soon face a prodigious shortage of nearly 2 Lacs data scientists –The Hindu. The talent pool of Big Data literates is sparingly low while the overwhelming demand is hard to occupy. A Big Data savvy can avail as much of $90-$180,000 depending upon the individual role and experience. Now, this is seemingly un-neglect able. The continuous hiring demand needs to be entertained as the Big Data flair in the mobile, financial, health care, marketing and e commerce and social media is eying for the right talent. Data Scientists are rare and powerful-‘But you can be a part of that rare’ The new breed of Data Scientists is working with both start-ups and well-established companies. Their sudden importance in today’s world alone describes that companies are wrestling with information that comes in varieties and volumes never encountered before. Often termed as the ‘Sexiest Job of 21st Century’, (Term postulated by the Harvard Business Review) is already facing a shortage of data scientist and this has become a serious constraint in majority of sectors. This high ranking profession is demanding some curiosity driven perspicacious professionals who want to make discoveries in the big world of ‘Big Data’. Much of the current enthusiasm for big data focuses on technologies that make taming it possible, including Hadoop (the most widely used framework for distributed file system processing) and related open-source tools, cloud computing, and data visualization. Lured enough by Big Data? ‘’Zoom with these world class Universities to sparkle with brilliance’’ Aegis School of Business and Telecommunication-A genius on its own! Making big in this ballooning world of Big Data in India is now possible with the Aegis School of Business and Telecommunication who have launched India’s first ever Post Graduate Program (PGP) in Business Analytics and Big Data in association with IBM. The Program is available Full time Post Graduation Program (On Campus), Online (Part Time Executive Program in Hybrid Model, On Campus and Online) and Part Time Executive Program in Online Model (Online Live Interactive). Aegis Programs’ Open course Framework offers participants the flexibility to build their own curriculum and choose a direction in Management, Technology or Techno-Management. The participants will receive a certification from IBM at the completion of the course. The most innovative part of the curriculum is that candidates would be trained at an IBM Business Analytics and IBM Cloud Computing Lab that would help student to polish and enhance their skills of Big Data, Business Analytics and Cloud Computing and Mobility. The Globally acceptable 36 Credit program is spread over the 11 month’s span with a minimum 2-3 months paid internship, now this is flamboyant! The new generation tech savvy will be trained in the World class infrastructure of MTNL, CETTM at Powai, Mumbai. The intensive training will help students develop the necessary and unique set of skills required for a successful career in the world of Big Data and Business Analytics. The eligibility criterion is simple; you need to be either of a Bachelor of Engineering, Bachelor of Technology, A Bachelor of Science (Mathematics, Information Technology, and Computer Science), and Bachelor of Computer Applications, Post Graduate degree in Science, Master of Arts and Post-Graduate degree in Computer Application. After the completion of the course the extensive knowledge will give you the opportunity to work either in the private or public sectors in roles such as: Business Analyst, Data Scientist, Business Intelligence Consultant, Data Architect, Chief Data Officer, Big Data Manager, among others. This Master’s degree will enable you to manage the languages of research methods, analytics, business performance and intelligence, and technology. Blue Print of ISB- Biocon Certificate Programme in Business Analytics’ offering the best of the best ‘’Big Data Cognizance’’ The Indian School of Business (ISB) that had evolved years ago with a vision of providing world class business school in Asia has also recently launched ‘The Certificate Programme in Business Analytics’ to train and meet the stupendous demand of Data Scientist across the world. The Programme is a combination of technology and classroom aided learning platform.CBA (Certificate Programme in Business Analytics) is a very challenging and rigorous Programme that is launched with a goal to Enhance the decision making process Maximizing cost efficiency Ameliorating strategies and alignment Revamping competitiveness Upgrading revenues Broadcasting information to a wider audience New York University-The name itself is an enchanter! New York University has captured the true essence of Big Data, as they believe in ‘Turning Data in to insight. We know this, our hugely networked human world has generated and even generating a deluge of data that no organization, no system, no group of humans of a homo sapiens can process it with ease. This data deluge has a budding potential to transform the world. It can change the way a government, business, healthcare and science works. The emerging discipline of data science holds the key to unlocking that immense potential. It exploits automated methods to analyze massive amounts of data and extract knowledge from them. Apparently, Data science combines aspects of computer science, applied mathematics and statistics. “Data science will also revolutionize medicine, education, and other areas that are slated to become ‘evidence-based.’” This is what NYU’s Data Science and Statistics Working Group believe in. The Data Science initiative at New York University is a University wide effort to establish the world’s leading data science training and research facilities at NYU. It is launched with an objective to help meet the enormous demand of researchers and professionals skilled in developing and utilizing automated methods of analyzing data. The initiative is focused on harnessing the potential power of big data to transform areas ranging from health care to business to government .Data science overlaps the traditionally strong disciplines at NYU such as computer science, statistics and mathematics. It also aims to impact disciplines in which NYU’s departments and schools are actively engaged such as sociology, law and economics. New York University has launched extensive programs with significant Data Science that includes many Master’s and PGD programs along with some special programs focusing primarily on Big Data. Students can also take up Data Science programs as Electives. Some of the acclaimed programs are- Master’s MS in Data Science (CDS) MS in Business Analytics (Stern) MBA with a Specialization in Business Analytics (Stern) MS in Computer Science with Specialization in Visualization, Data bases, and Big Data (School of Engineering) PhD PhD in Computer Science (Machine Learning\/AI specialization) (Courant) PhD in Computer Science with in Visualization, Data bases, and Big Data (School of Engineering) The ‘Go’ of Big Data! Big Data is here to solve big problems, whether you are involved in the tech industries, or in the mainstream businesses; this word is steering up a lot of discussions. While Gartner, Oracle, Intel and Microsoft have their own definitions for streamlining this action packed word, Ward and Baker came up with the following definition (via MIT Technology Review): “Big data is a term describing the storage and analysis of large and or complex data sets using a series of techniques including, but not limited to: NoSQL, MapReduce, and machine learning.” According to the ground breaking study by Big and Fast Data: The Rise of an Insight Driven Business produced jointly by CapGemini and EMC have postulated some thought provoking virgin aspects of Big Data that may change your cognition of this ‘Red Hot’ term from now. According to the reports about nearly 56% of the Enterprises across the globe will increase their investments in Big Data over the next three years while 70% of the IIT decision makers believe that Big Data is critical for their future success. Not to ignore, a whopping 53% believe that they have some serious competition from the Data enable startups. Majority of the enterprises believe that they might be in peril if they do not embrace big data and their chances of becoming irrelevant and uncompetitive are sky high. Certainly, these facts can’t be overlooked. For an industry that is booming so hastily, students surely have a reason to celebrate. As per the employment latitude and lack of skilled talent pool, this sector has a potential to hire tech savvy Big Data literates, like never before. The amorphously named Big Data that is a current lollapalooza is comprised of a set of four skills including, data analysis, data mining, and data structures and data acquisition. In the past twelve months demand of Big Data literates increased to 89.9% Demand for big data expertise across a range of occupations saw significant growth over the last twelve months. There was a 123.60% jump in demand for Information Technology Project Managers with big data expertise. The top five leading industries with the most frequent job opening requiring big data expertise includes Scientific and Technical Services, Manufacturing, Professional, Retail Trade, Information Technology , Remediation Service, Waste Management and Sustainability. Nearly 76% of these vacancies require Big Data expertise. While the median salary for professionals with Big Data expertise is $103,000 per annum. Nevertheless, IBM, Oracle and Cisco have the most open big data related positions currently. If you are a star; you will not be overlooked or if you are a high performer, you will not be slipped as Big Data demands smart and fair selection procedure based purely on talent, expertise and acuity.","excerpt":"It’s not a mighty astonishment that among the very new faces debuting as billionaires for the year 2015, are the ones who have made riches from the Big Data. Their empires are built solely on their prowess and expertise to collect, interpret and use data in such novel ways that no one in the past […]","categories":["AI Trends"],"tags":[],"author_name":"Aegis School of Business and Telecommunication","publish_date":"2015-05-15T15:19:02","publication_year":"2015","word_count":1910,"keywords":["data science","machine learning","AI","cloud computing","ML","RAG","CuPy","Aim","analytics","R"],"extracted_tech_keywords":["AI","machine learning","ML","data science","analytics","Aim","CuPy","RAG","cloud computing","R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/big-data-the-new-age-desire\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10062861,"title":"Top preprint servers for publishing your AI research","content":"India recently launched its first preprint server ‘IndiaRxiv’, which aims to be a one-stop shop for domestic research. IndiaRxiv is a preprints repository service for India, managed by Open Access India and hosted by the Society for Promotion of Horticulture, Bengaluru. The repo is a one-stop portal for Indian researchers’ early works. Preprint server allows immediate sharing of research results and can provide quick feedback to help revise and prepare a manuscript. A community of practice, Open Access India, maintains the India Archive preprint repository. Preprint servers are open access online archives or repositories that contain research papers before their peer review and publication. Their main aim is to accelerate the dissemination process of research findings and enhance their visibility. As the peer review process takes time and there is a subsequent delay in publication, preprint servers are useful tools for researchers to post a full draft of their research papers and get immediate feedback from their colleagues. Let’s take a look at some of the most popular preprint servers in AI research. arXiv arXiv is an open-access repository for electronic preprints and post-prints; it has a free distribution service and an open-access archive for scholarly articles in physics, mathematics, computer science, quantitative biology, quantitative finance, statistics, electrical engineering and systems science, and economics. Mathematics, physics and many scientific papers are self-archived on the arXiv repository before they get published in a peer-reviewed journal. Some publishers even grant permission for authors to archive the peer-reviewed postprint. ‘arXiv’ was one of the first adopters and promoters of preprints. Its success in sharing preprints was one of the precipitating factors that led to the movement in scientific publishing known as open access. Cogprints CogPrints is an electronic archive for self-archive papers in many subjects, including computer science, artificial intelligence, robotics, vision, learning, speech, and neural networks. Cogprints is powered by EPrints3 and developed by the School of Electronics and Computer Science, University of Southampton. CogPrints invites scientists to deposit their e-prints in the archive. The advantages for contributors include high visibility; permanent, instant, global accessibility; text-capturability, feedback, comments and also quoting in others’ work; no cost to circulate preprints or reprints. OSF Preprints The Open Science Framework (OSF) is an open-source software project that facilitates open collaboration in science research. OSF Preprints integrates OSF project infrastructure, allowing researchers to include supplemental data, materials, code or other information alongside their preprint. OSF Preprints makes sharing research as easy as uploading a file. Post your work, solicit feedback, and tag categories for others to find, comment on, and engage with you. The files can be stored on OSF storage or through connected services like GitHub, Dropbox, Amazon S3, Box, Google Drive, ownCloud, etc. Preprints.org Preprints.org is dedicated to making early versions of research outputs that are permanently available and citable. They post original research articles, comprehensive reviews, and papers that authors can update at any time. Content on Preprints.org is not peer-reviewed and can receive feedback from readers. All preprints are posted with a Creative Commons CC BY 4.0 licence that ensures that authors retain their copyright and receive credit for their work while allowing anyone to read and reuse it. This platform is for multidisciplinary research work. For each preprint, authors can register a unique digital object identifier issued by Crossref. This makes them instantly citable and provides a permanent link to the article, even if the URL on the platform changes. New versions of preprints receive different DOI. Zenodo Zenodo is a general-purpose open-access repository that was developed under the European OpenAIRE programme and operated by CERN. Zenodo allows researchers to deposit research papers, data sets, research software, reports, and any other research-related digital artefacts. A persistent digital object identifier (DOI) is minted for each submission, which makes the stored items easily citable. Datasets, documents and other research materials required can be found in the Zenodo search engine. Scholars from any research discipline are allowed to upload data in any file format. Section 6(c) of the EUI Library Research Data Guide details on assigning metadata to research datasets. Zenodo code itself is open-source and built on the foundation of the Invenio digital library, which is also open-source. The work-in-progress, open issues and roadmaps are shared openly in GitHub, and contributions are welcomed from anyone. TechRxiv TechRxiv is an open, moderated preprint server for unpublished research in engineering, computer science, and related technology. By using TechRxiv, authors can quickly disseminate their work to a wide audience and gain community feedback on a timestamped draft version of their research. TechRxiv accepts unpublished research in electrical engineering, computer science, and technology. Articles suitable for TechRxiv should comply with one of the16  subjects that describe TechRxiv’s fields of interest, including Aerospace, Bioengineering, Communication, Networking and Broadcast Technologies, Components, Circuits, Devices and Systems, Computing and Processing, Engineered Materials, Dielectrics and Plasmas, Engineering Profession, Fields, Waves and Electromagnetics, General Topics for Engineers, Geoscience, Nuclear Engineering, Photonics and Electrooptics, Power, Energy and Industry Applications, Robotics and Control Systems, Signal Processing and Analysis and Transportation.","excerpt":"Most preprint servers enable researchers to upload draft versions of a manuscript before submitting it to a more traditional peer-reviewed journal.","categories":["AI Trends"],"tags":["AI Research"],"author_name":"Poornima Nataraj","publish_date":"2022-03-18T13:00:00","publication_year":"2022","word_count":835,"keywords":["Go","artificial intelligence","TPU","OpenAI","AI","neural network","AI Research","Git","RAG","Aim","R"],"extracted_tech_keywords":["AI","artificial intelligence","neural network","OpenAI","Aim","RAG","TPU","R","Go","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-preprint-servers-for-publishing-your-ai-research\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10069661,"title":"Does TeamViewer still have a play?","content":"Remote access and remote control computer software solution TeamViewer has been in existence since 2005. The Germany-headquartered company set base in India in 2018. The software, like many other similar applications, saw an obvious spike in usage when the COVID-19 pandemic hit us in March 2020. But along with its pros, work from home also came with its issues – communication barriers being one of them. To build better communication and monitoring systems, companies deployed productivity apps and remote access softwares to figure out the activity of their employees and engage better with them. TeamViewer is a popular choice among these applications. Recently, the company appointed Rupesh Lunkad as managing director, India. He succeeded Krunal Patel, who played a pivotal role when TeamViewer was setting up its presence in India in 2018. How does TeamViewer make money? Essentially, it does not require registration and is free of charge for non-commercial use. There are three levels of paid commercial use the company provides: Business License: One licensed user can initiate device connection one session at a time, priced at ₹1,499 per month.Premium license: Multiple licensed user accounts that enable one user to connect at any given time, priced at ₹3,099 per month. Corporate License: Multiple licensed users can initiate device connections, with up to three sessions open at the same time, priced at ₹6,599 per month. Image: TeamViewer Enterprise business is growing strong While declaring the Q1 2022 results, Teamviewer CEO Oliver Steil talked about how the company has achieved double-digit growth on top of 2020’s and 2021’s strong first quarters, underlining the resilience of business despite general macroeconomic uncertainty and a more normalised working environment. He emphasised that the Enterprise business is growing strongly with an ever-increasing contribution to the company’s success, and the Q1 2022 figures indicate that as well. It shows that there is continued momentum, with billings up 51% YoY to reach € 35.2 million. The Enterprise business accounts now for 22% of total billings, indicating a mix shift of the businesses. Image: Inauguration of TeamViewer’s new R&D hub in Portugal For Q1 2022, the billings are up by 12% compared to the previous year. There is strong retention of pandemic cohorts resulting in 101% Net Revenue Retention. The company informs that the SMB business is up 4% YoY, with higher quality tiers growing 15% YoY (LTM). The subscriber development is stable, with 620,000 at the end of Q1 2022. Image: TeamViewer Quarterly Statement Q1 | 2022 Go down after pandemic subsides? Though most popular for its remote access software, that is not the only area TeamViewer caters to. In an interview with CNBC last year, Steil said, “We are active in different parts of the business. For large customers, we are back in the normal rhythm of pipeline building and a sales cycle of three to six months. We also have smaller tickets which have a sales cycle of a day or two.” When asked if the company is expecting a drop in demand as the crisis felt in 2020 and 2021 has subsided, Steil adds that he looks at the situation differently. “We have broadened our solution portfolio. We are looking at providing solutions with augmented reality and IoT that will really go into digitalising industries across the globe with very different use cases, like frontline worker enablement and remote IT management. So, there is much more to our business than working from home. WFH just led to some extra spike last year, but that’s actually only a part, and we are moving more and more to diverse use cases. Over the last few years, we have invested and put in more R&D. The solutions portfolio now consists of the entire value chain of businesses,” he adds. More innovation and diversifying product offerings (IoT, AR) Assist AR – It is a remote support solution that provides secure augmented reality-powered visual assistance. It uses technologies like advanced 3D object tracking, with which the expert can “mark” the display with arrows, notes, and highlights that make it easy for the user to follow along with the expert’s instructions. Optical Character Recognition (OCR) capabilities are also used to recognise printed characters like those often found on machines, tools, and equipment. Image: World Class After-Sales Support with xAssist for improved quality and faster service; TeamViewer TeamViewer Frontline – It is a fully integrated augmented reality solution. TeamViewer says that “Frontline is designed to consciously empower the human worker in an increasingly digitised working environment.”TeamViewer Engage – TeamViewer calls it a next-gen digital customer engagement platform for online sales, digital customer service, and video consultations. It comes with scalable cloud or on-premises deployment options, single sign-on (SSO) integration, and built-in security. India strategy TeamViewer opened its doors to India in 2018 and appointed Krunal Patel as Head of Sales in the country. At that time, the company had talked about leveraging the market’s immense growth potential for enterprise as well as SMB remote access, collaboration and IoT solutions. In an interview in 2019, Patel said that one of the main reasons the company came to India is the growth the country is experiencing. The company wanted to have more Indian clients and better support Indian customers. The firm has customers coming in all shapes and sizes from across the country. During the recent appointment of Lunkad, Sojung Lee, president of Asia Pacific, TeamViewer, said that this hiring would help to expand its footprint in the India market further and create a strong ecosystem of alliances with its partners. Competition There are many products in the market similar to TeamViewer that are giving it stiff competition. Some of them are: AnyDesk – One of the biggest competitors of TeamViewer is AnyDesk. It lets the user view the screen, control the mouse & keyboard, and manage data & files of other devices. The company says that AnyDesk Software has been downloaded over 500 million times with a presence across the globe. Image: AnyDesk vs TeamViewer – Why AnyDesk Is the Better Alternative Chrome Remote Desktop – It solves a similar purpose and is backed by Google. If one needs to access their work computer from home, view a file from a home computer while travelling, or share a screen, Chrome Remote Desktop allows all of that. TeamViewer can control any device – computers or mobiles, while Chrome Remote Desktop limits to mobiles; flexibility is less.GoToMyPC – It has been in existence since 1998 and has three versions – Personal, Pro and Corporate. While multiple competitors have cropped up, TeamViewer retains its long existence amongst users. Broadening its applications might be a good idea to sustain the competition and the shift back to work in offices.","excerpt":"Recently, the company appointed Rupesh Lunkad as Managing Director, India.","categories":["IT Services"],"tags":[],"author_name":"Sreejani Bhattacharyya","publish_date":"2022-06-24T10:00:00","publication_year":"2022","word_count":1113,"keywords":["Go","ELT","programming_languages:R","AI","innovation","Scala","Git","RAG","ViT","R"],"extracted_tech_keywords":["AI","RAG","R","Go","Scala","Git","ELT","ViT","innovation","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/it-services\/does-teamviewer-still-have-a-play\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10102708,"title":"OpenAI Puts on The Open-Sourcing Act Again","content":"OpenAI has been infamous for its closed door practices but the hottest AI startup appears to be making baby steps to contribute to the open-source community. At the inaugural DevDay, the company announced two open-source updates, which was highly unlikely given how heavily it has been trolled in the recent past year for not openly sharing any of its tech. The company’s head of developer experience, Romain Huet took the stage at the event to announce ‘large-v3’, the new version of its automatic speech recognition model Whisper. OpenAI also has plans to release an API for the upgraded version in “near future”. Huet also announced the open sourcing of the Consistency Decoder which is a replacement for the Stable Diffusion VAE decoder. It sounds like a really major improvement for Stable Diffusion (if it does what they say it can!). The company’s official blog mentioned that the decoder works with “1.0+” which ideally should mean “all” but that is not the case. It is only supported by versions v1 and v2, not SDXL. Though the model was released less than 24 hours ago, developers have already started playing around with it and some are not satisfied with the results. “Faces are far worse than with the vae-ft-mse-840000-ema-pruned in all the tests I made. And it takes something like 20 seconds to decode the latent image when usually it just takes 1 second,” wrote an early user on Reddit. What’s In A Name? The creator of ChatGPT was founded as a non-profit but turned towards a more secretive ‘capped for profit’ direction. When GPT-4 was revealed along with a 98-paged document that said nothing about the model, the stunt by OpenAI was grumbled upon by several AI insiders including Elon Musk who has initially invested $100 million dollars in the company and has never missed a chance to mention that. The irony of the name came up several times on the internet and many even suggested that the company should consider changing its name to “Closed AI”. In an interview with The Verge, OpenAI’s chief scientist and cofounder, Ilya Sutskever made clear the reason for the company’s shift in focus from open-source. But he also acknowledged that “the safety side is not yet as salient a reason as the competitive side,” noting that “it’s competitive out there” and “there are many, many companies who want to do the same thing.” While the company has not shared much of its latest slew of products, it has previously open-sourced half a dozen AI models — Evals, Spinning Up, CLIP,  Jukebox and Point-E. Open-Sourcing the Future The optional sharing of technology by the tech companies has been a long debated topic. Open core has gained momentum since large language models have made a name for themselves in the AI research community. Tech giant Meta has also been advocating for the free and open source software by releasing Llama-2 which has a separate fanbase. Even investors are heavily betting on open-source today as opposed to what they were doing a year ago. Majority of the technology from your iPhones to Firefox browser has some or the other open sourced material at the heart of it. The open-source culture has long transferred the tech industry but at an enterprise level it remains in a state of flux due to the lack of safety in its nature. For OpenAI case in point, the closed approach is working out well for the startup since the company is being darted with lawsuits left, right, and centre for copyright infringements. A point that Sutskever missed was that by releasing the inner workings of its models the company will have to pay thousands of artists and creators for using their work to build its AI models. It looks like OpenAI has been contributing to the community as per its convenience, ignorant to its ethos.","excerpt":"OpenAI has been contributing to the open-source community as per its convenience, ignorant to its ethos","categories":["AI Features"],"tags":[],"author_name":"Tasmia Ansari","publish_date":"2023-11-08T10:00:00","publication_year":"2023","word_count":643,"keywords":["Go","ChatGPT","API","AWS","OpenAI","AI","GPT","stable diffusion","CLIP","R"],"extracted_tech_keywords":["AI","ChatGPT","OpenAI","AWS","R","Go","API","GPT","CLIP","stable diffusion"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/openai-puts-on-the-open-sourcing-act-again\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10020950,"title":"Apple On A Buying Spree, Banking On AI Startups","content":"For many years now, we are witnessing a tectonic shift in how technology is being used to revolutionise major sectors around the world. One of the most prominent keywords that keep popping up now and then is the ‘artificial intelligence revolution.’ Tech giants such as Google, Microsoft, Facebook, Apple, etc., have shown deep interest in not just adopting AI but also acquiring AI startups, which has become the new normal. Apple is winning the race with the ‘You Start, I Purchase’ mindset. The number of acquisitions it has made, without much pomp and shows over a short span of three to four years, is remarkable. Nevertheless, it is a win-win situation for both Apple as well as the startups. “Apple buys a new company every two or three weeks.” In a recent interview with CNBC, Tim Cook, CEO of Apple Inc, These are not just words — as per the data compiled by CB Insights, the company has acquired over 20 AI startups till 2019, since the 2010 purchase of Siri. A Win-Win for Both Apple – with over one billion customers, using 1.4 billion devices needs constant innovation and upgrades in the services it provides to its unique customer base. Acquiring potential AI startups equip large firms like Apple to fill their talent pool, providing them with an enthusiastic and innovative team and the latest technologies to add to their own AI capabilities. It provides them with a safety net to remain ahead of its competitors and achieve unique user-based services to enlarge its market footprint. On the other hand, many AI startups are booming up with unique, localised, and user-friendly solutions. As the startups start to mature, they often face the funding dilemma to expand their business activities. At that stage, tech giants like Apple, with ample funding resources, come to the rescue, hence cushioning the expansion and sharing of the risks involved. Moreover, the acquisition enriches the team with financial gains and a stable future in terms of career. Let’s have a look at some of the prominent Apple’s acquisitions in 2020 — Xnor.ai Seattle-based edge focused AI startup — Xnor.ai — was acquired by Apple in the first month of 2020 to strengthen its capabilities by running AI models on low-resource edge devices instead of relying on the cloud. As per the reports, Apple paid an amount of $200 million. Xnor.ai has also developed a self-service platform that makes it possible to drop AI-centric code and data libraries into device-centric apps for software developers, including those who are not skilled in artificial intelligence. Dark Sky An AI-based weather app was added to Apple’s basket in April 2020. It delivers forecasts for your precise location, giving you minute-by-minute predictions for the next hour and hour-by-hour forecasts for the next week. Dark Sky offers hyperlocal weather information. With down-to-the-minute forecasts, you’ll know exactly when the rain will start or stop — right where you’re standing. Apple could use Dark Sky data to bolster its existing core iPhone apps, like the Weather app and Siri. Voysis-Dublin AI Startup To equip people with digital voice assistants that can understand and provide feedback in their own natural\/native languages can be considered the best technical help a brand can provide to its customer. This is precisely what Apple did by acquiring Dublin AI startup, Voysis in 2020. Voysis’s framework taps into Wavenets, an AI-based tool that Google’s DeepMind first built-in 2016 to generate more human-like machine voice. The intelligence will further improve Apple’s Siri or can be used to provide brand-specific, intelligent voice-system solutions in regional languages to tap into a huge market base across nations. Camerai Apple had also acquired an Israeli AI-based camera startup in 2020. The technology can detect different items in a picture. It also offers “skeleton tracking”, a neural network API capable of detecting body joints and relaying on a real picture of a human. The technology and the know-how gained from the firm will further enhance the quality and a much better experience in its portrait modes to night photography. Apple, rather than going out for the big-ticket deals, relies on acquiring smaller technological AI-driven firms. The strategy has paid-off really well as it helped strengthen its foothold with regular adding up of cutting-edge technologies in its armour. AI-based solutions are surely the idea to bank on for future growth prospects and outshine your competitors in the long-run.","excerpt":"For many years now, we are witnessing a tectonic shift in how technology is being used to revolutionise major sectors around the world. One of the most prominent keywords that keep popping up now and then is the ‘artificial intelligence revolution.’ Tech giants such as Google, Microsoft, Facebook, Apple, etc., have shown deep interest in […]","categories":["AI Startups"],"tags":["Startups"],"author_name":"kumar Gandharv","publish_date":"2021-02-27T10:00:00","publication_year":"2021","word_count":732,"keywords":["Go","API","artificial intelligence","AI","neural network","innovation","Git","ViT","Startups","Tecton","R"],"extracted_tech_keywords":["AI","artificial intelligence","neural network","Tecton","R","Go","Git","API","ViT","innovation"],"url":"https:\/\/analyticsindiamag.com\/ai-startups\/apple-on-a-buying-spree-banking-on-ai-startups\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10171991,"title":"Wipro, Tech Mahindra Partner with Aduna to Drive Global Adoption of Network APIs","content":"Indian IT companies Wipro and Tech Mahindra have joined the partner ecosystem of Aduna, a leading aggregator of network APIs, according to an official statement. The collaboration aims to accelerate the global adoption of network APIs by enabling both the enterprise and telecom sectors to integrate these capabilities seamlessly into their systems. As part of the partnership, Wipro and Tech Mahindra will support the onboarding of Communication Service Providers (CSPs) and help accelerate their operational readiness across regions. In addition to facilitating global rollout of network APIs through their managed services and integration frameworks, they will co-develop and deploy industry-specific use case accelerators in sectors such as banking, logistics, and public services. The release further stated that the companies will help enterprises adopt network APIs by embedding them into cloud-native systems, digital identity platforms, and security solutions. As long-standing partners to telecom operators and trusted advisors to Fortune 500 companies, Wipro and Tech Mahindra will embed Aduna’s harmonised network APIs—including SIM swap, number verification, device location, and quality on demand—into digital transformation programs across industries. “This is a major step forward in operationalising network APIs at scale,” Anthony Bartolo, CEO of Aduna, said. “Tech Mahindra and Wipro sit at the intersection of enterprise transformation and telecom innovation. Together, we’re enabling both the supply and demand sides of the ecosystem to come online securely, scalably, and globally.” Meanwhile, Manish Mangal, head of the Americas Communications Business at Tech Mahindra, said in the statement, “The telecom industry has a significant opportunity to unlock the value of its networks and other assets by subjecting them to APIs to third-party developers and partners.” “Our partnership with Aduna exemplifies our commitment to the telecom industry in fostering new business models and creating opportunities for revenue growth,” he added. Srikumar Rao, managing partner and global head of engineering at Wipro Limited, stated, “Wipro is proud to join the Aduna ecosystem and extend our AI, cloud, and security offerings with real-time telecom intelligence. Together, we’re empowering enterprises and CSPs to unlock new value through scalable open APIs.”Global system integrators (GSIs) Wipro and Tech Mahindra play a pivotal role in bridging the gap between telecom capabilities and enterprise needs. The release stated that by integrating network APIs into digital workflows, GSIs help enterprises improve security, streamline onboarding processes, enhance communications, and comply with regulatory requirements, without the burden of telco-related complexity. Moreover, they assist mobile operators in integrating with the Aduna platform through business consulting, service frameworks, and technical onboarding.","excerpt":"The Indian tech companies are expected to co-develop and deploy industry-specific use case accelerators in sectors such as banking, logistics, and public services.","categories":["AI News"],"tags":["AI","API","Tech Mahindra","Wipro"],"author_name":"C P Balasubramanyam","publish_date":"2025-06-19T09:53:24","publication_year":"2025","word_count":412,"keywords":["Wipro","API","Tech Mahindra","AI","innovation","ML","digital transformation","Scala","Git","Aim","Rust","R"],"extracted_tech_keywords":["AI","ML","Aim","R","Rust","Scala","Git","API","digital transformation","innovation"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/wipro-tech-mahindra-partner-with-aduna-to-drive-global-adoption-of-network-apis\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10082374,"title":"Tech Mahindra Chases Quantum Dreams With IQM","content":"Tech Mahindra has previously made moves to solidify India’s nascent quantum computing field by investing in quantum research and development. Now, they have signed a memorandum of understanding with Finnish quantum computing firm IQM to work together on advancing the field in India. This partnership will involve knowledge transfer between the two on topics such as cryptography, communication technologies, and the application of quantum computing in fields like healthcare and cybersecurity. It brings together Tech Mahindra’s experience and knowledge in the software side of quantum computing and IQM’s expertise in creating and building quantum hardware. To this end, Tech Mahindra had also collaborated with Mahindra University earlier this year to conduct more research into quantum computing and explainable AI, among other subjects. The company has also delved into the ways that quantum computing can disrupt current computing systems such as those used in banking and encryption. The R&D branch of Tech Mahindra, known as Makers Lab, has 10 excellence centres worldwide, with the facility at Mahindra University being the 11th. Makers Lab has also set up a centre of excellence in Helsinki to leverage Finland’s expertise in quantum computing. Tech Mahindra MD and CEO CP Gurnani had this to say about their focus on quantum technologies. “Quantum physics clearly shows that there is always this inflection point and then after that, either the current hardware or the quant developers will be able to suddenly create magic.” Tech Mahindra aims to use quantum computing to design drugs, as quantum architecture is much better suited to create molecule designs than AI systems. They are also aiming to deploy quantum computing technology in fintech, as it will enable better fraud detection and portfolio optimisation. They have also invested heavily in their NXT.NOW framework, which targets emerging technologies like quantum computing to enhance the human experience. IQM is a company that specialises in creating and deploying quantum hardware for enterprises and research labs alike. IQM first focused on India as a part of the Indian government’s quantum tech mission announced in 2019. They have conducted deep research into creating superconducting quantum hardware, and leveraged this technology to create a commercial 54-qubit quantum computer; the first of its kind in Finland. Earlier this year, the Indian government signed a joint declaration with Finland for establishing a virtual centre centred on quantum computing research. S Chandrasekhar, secretary, Department of Science and Technology (DST), said, “India is a strong believer in collaborations rather than competition. So, we don’t want to compete with anyone. Instead, we would like to collaborate and bring synergies together to reach our endpoint as quickly as possible. I do believe that quantum is yet another tool we are going to have in our hand, which we will use for the betterment of humankind.”","excerpt":"This partnership will involve knowledge transfer between the two topics on topics such as cryptography, communication technologies, and the application of quantum computing","categories":["AI News"],"tags":[],"author_name":"Anirudh VK","publish_date":"2022-12-14T18:30:09","publication_year":"2022","word_count":460,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","RAG","Aim","explainable AI","R","emerging_tech:quantum computing","fraud detection"],"extracted_tech_keywords":["AI","Aim","RAG","fraud detection","R","Go","explainable AI","programming_languages:R","programming_languages:Go","emerging_tech:quantum computing"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/tech-mahindra-chases-quantum-dreams-with-iqm\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":12086,"title":"Datalicious expands global footprint, sets up shop in India","content":"Datalicious, a global marketing analytics company opened a Centre of Excellence in Bangalore It was hailed by JobAdvisor in the Top 10 Tech Companies to work in Australia, and recently set up shop in Bangalore. Sydney-based full service data analytics agency Datalicious opened its doors in India’s Silicon Valley earlier last year with a view to base their Centre of Excellence (COE) in India. The mission was two-fold – attracting great talent and servicing all the markets, especially South East Asia. “As the gateway to Europe and Asia, India was a natural geographic choice to base our second largest office. We have scaled the team to about 20 people in Banglaore with a majority of our project delivery now happening from India,” affirmed Siddharth Manuja, Datalicious’s India Country Manager. The company already has offices across the globe including Manila, Singapore that offered Datalicious a unique opportunity to enter the South East Asian market and stack up big ticket clients across Thailand India and Indonesia. Some of the heavyweight companies in Datalicious’s roster are Bluestone, RedBus and Goibibo. Siddharth Manuja, Country Manager at Datalicious So why set up shop in India specifically? Speaking to Analytics India Magazine, Manuja sheds light on the Indian expansion plan and the pressing need to bolster global product development and delivery goals. “Datalicious was at an important juncture in terms of global growth and through India we are able to access the high quality of talent that is coming out of the Indian university system.” According to Manuja, the sheer number of Indian students in technology was a big enough reason to open a COE in India.  There are over 4 million students in India studying engineering compared to only 67,000 in Australia. The statistics are definitely staggering. “The company also started seeing excellent growth around the South East Asian region, including India and Europe and we are extremely well-positioned to support and innovate,” he said. The Country Manager is not new to the data analytics space. With past stints at Facebook and HCL Technologies, he knew how to drive insights from petabytes of data. “I started my career working on content management and online advertising, and later gravitated towards analytics as I found that to be the common thread in driving success for multiple functions. I have spent the last 6 years in different roles in the data analytics space. With Datalicious we extend that principle across a multitude of channels, ensuring marketers reach the right people, in the right place, at the right time – using insights and algorithms researched and developed here in India.” he said. Besides setting up an excellent team in India, what Manuja will bring to the table is his vast years of experience, expanding the reach of Datalicious in Indian market, overseeing operations, delivery and sales. Datalicious Journey — from Digital Agency to Data Technology firm Over the years, Datalicious has evolved from a full service data agency to a data technology company rolling out marketing analytics products such as SuperTag, DataCollector, DataExchange and the advance marketing analytics platform OptimaHub that enables marketers to make the best decisions in spending and attribution. Founded by Christian Bartens, the present CEO oversaw the transformation – turning Datalicious into a technology firm offering a wide range of services such as data warehousing, data mining, modelling, and reporting besides mobile and web analytics and media attribution. Datalicious further strengthened its portfolio by inking a partnership with Google becoming the key Google Analytics Premium partners in the region that generated new revenue streams for them. The Australian company intensified its focus on India not just because of the exceptional talent pool, but also the astounding numbers – the market for marketing analytics is about $1.7 billion dollars and growing at a healthy rate. Digital Marketing in India has evolved from web analytics and mapping customer sentiments to customer lifecycle management, mobile and media attribution, a major offering from Datalicious that helps marketers understand the customer’s interaction with the brand across the digital touchpoints. When it comes to optimizing marketing spend, Datalicious answers the crucial questions such as — how to allocate budgets across marketing channels to maximise the return on investment (ROI). However, Datalicious doesn’t lack competition.  From Axtria to MarketShare, small and big companies have expanded their footprint into India and are offering customer-focused solutions to marketers and helping them improve their marketing decisions through insightful data. But it is a reality that Manuja is already aware of. He said, “Going back to the number of students studying degrees in this area that means there is going to a lot of people not only moving into existing analytics companies, but starting their own as well. This will create a healthy, competitive analytics marketplace in India and we look forward to being a part of the ecosystem.”","excerpt":"It was hailed by JobAdvisor in the Top 10 Tech Companies to work in Australia, and recently set up shop in Bangalore. Sydney-based full service data analytics agency Datalicious opened its doors in India’s Silicon Valley earlier last year with a view to base their Centre of Excellence (COE) in India. The mission was two-fold […]","categories":["IT Services"],"tags":["Data analytics India"],"author_name":"Richa Bhatia","publish_date":"2017-01-04T10:59:22","publication_year":"2017","word_count":803,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","Git","Data analytics India","ViT","analytics","R","analytics platform"],"extracted_tech_keywords":["AI","analytics","R","Go","Git","ViT","analytics platform","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/datalicious-expands-global-footprint-sets-shop-india\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10017568,"title":"Signal App: An Inside Look Into WhatsApp’s Perfect Foil","content":"WhatsApp’s privacy policy update has kicked up a storm recently. According to the new policy, the popular messaging app will share user data such as location, battery level, IMEI number, mobile network and related information with its parent company, Facebook.  Later, WhatsApp was forced to issue a clarification saying the messages between the users will not be shared with Facebook. Notably, the WhatsApp policy remains unchanged in Europe, thanks to the stringent General Data Protection Regulation (GDPR). The new policy changes have broken people’s trust in the app. Many are looking for a better and more secure replacement. And Signal App seems to be the popular choice. Unlike other instant messaging apps, Signal only stores the users’ contact info. Further, all messages and calls on the app are end-to-end encrypted, meaning no third-party, not even Signal can access them. Signal Foundation and Signal Messenger LLC, a non-profit company, rolled out its flagship app in 2014. Ironically, Signal Foundation was set up by WhatsApp co-founder Brian Acton with Signal Messenger CEO Moxie Marlinspike. Acton had exited WhatsApp in 2017, three years after Facebook acquired the messaging app. Brief History After the new WhatsApp terms were made public, SpaceX and Tesla CEO Elon Musk tweeted ‘Use Signal’. And Twitter CEO Jack Dorsey retweeted Musk. Use Signal— Elon Musk (@elonmusk) January 7, 2021 The app was also endorsed by privacy activist and whistleblower Edward Snowden. Signal is a one-tap install app available on Google Play Store and Apple’s App Store. The software powering the app is open-sourced and free of charge. I use Signal every day. #notesforFBI (Spoiler: they already know) https:\/\/t.co\/KNy0xppsN0— Edward Snowden (@Snowden) November 2, 2015 In 2010, Whisper Systems launched two Android apps — TextSecure and RedPhone. While TextSecure was for sharing encrypted text messages, the latter was for making encrypted voice calls. In 2011, Twitter bought Whisper Systems, and both apps were released as open-source softwares. In 2013, Moxie Marlinspike, co-founder of Whisper Systems exited Twitter and set up Signal to further develop TextSecure and RedPhone. Later, Acton joined hands with Marlinspike to establish the non-profit under the same name. How Does Signal Maintain ‘Perfect Secrecy’ Most apps’ encryption systems create a permanent key pair for encryption and decryption of messages. The public key is used to identify the user and is sent to the messaging server, and the private key stays in the user’s phone. If the private key is compromised, due to hack or theft, the messages are vulnerable to decryption. Signal’s encryption protocol combines Double Ratchet algorithm with triple Elliptic-curve Diffie Hellman handshake. The sender and receiver use the Double Ratchet algorithm to exchange encrypted messages based on a shared secret key. A new key is generated for every message, and the earlier keys can not be figured out from the succeeding ones. This method is also called the perfect forward secrecy. At the core of this algorithm lies the concept of KDF (key derivation function) chain. KDF is a cryptographic hash function that uses a secret random key and input data to generate the output. The secret key is derived from a secret value such as a password or a passphrase using a pseudorandom function. Further, along with the double ratchet algorithm, the two parties also use the extended Triple Diffie-Hellman (X3DH) key agreement protocol. X3DH provides forward secrecy and cryptographic deniability. This protocol is used for establishing a shared key between the sender and receiver, who authenticate each other using public keys. However, the perfect forward secrecy on its own is not a full-proof strategy. In the event of theft, the messages still would be visible to whoever has the device. To that end, Signal App has added a time-bound ‘disappearing messages’ function. Wrapping Up Signal App’s popularity soars every time there is a public discourse around privacy and security. Like, in 2020, the downloads spiked at the peak of the Black Lives Matter movement. However, to think Signal will topple WhatsApp as the most popular messaging app is a bit of stretch, considering WhatsApp still commands an impressive user base of over 2 billion people. However, it is good to see that privacy is being taken seriously, and who knows, Signal’s protocol may even become the industry-standard in the future.","excerpt":"WhatsApp’s privacy policy update has kicked up a storm recently. According to the new policy, the popular messaging app will share user data such as location, battery level, IMEI number, mobile network and related information with its parent company, Facebook.  Later, WhatsApp was forced to issue a clarification saying the messages between the users will […]","categories":["AI Features"],"tags":[],"author_name":"Shraddha Goled","publish_date":"2021-01-12T14:00:00","publication_year":"2021","word_count":707,"keywords":["Go","TPU","programming_languages:R","AI","programming_languages:Go","Rust","R","programming_languages:Rust"],"extracted_tech_keywords":["AI","TPU","R","Go","Rust","programming_languages:R","programming_languages:Go","programming_languages:Rust"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/signal-app-an-inside-look-into-whatsapps-perfect-foil\/","complexity_score":4,"technical_depth":8,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10044257,"title":"Argo’s Rise To Leading Kubernetes Workflow Engine In Just 4 Years","content":"Argo is an open-source container-native workflow engine for Kubernetes. It was introduced by Applatex (owned by Intuit), which offers Kubernetes services and open source products. Argo is a workflow orchestration layer designed to be applied to step-by-step procedures with dependencies. It can handle tens of 1000s of workflows at once, each 1,000 steps each. These step-by-step procedures with dependencies are referred to as a Directed Acyclic Graph (DAG) in most current computing solutions. Some relevant DAG tools include Apache Airflow, Prefect, and Luigi. Just like all other Kubernetes-related capabilities, Argo is declarative, configurable, and meant to operate across environments. If you have already installed the extension on your Kubernetes cluster, users may deploy the Argo manifest, which abstracts the idea of a DAG. Argo allows one to automate, schedule, and manage complicated workflows and applications on Kubernetes with relative ease. It is implemented as a Kubernetes Custom Resource Definition (CRD). Argo workflows can be handled via kubectl and natively integrates Kubernetes with services like secrets, volumes and role-based access control (RBAC). Argo’s new software supports workflow capabilities, including parameter replacement, artifact, fixtures, loops, and recursive workflows. Argo workflows can be used in a variety of ways, including: Machine LearningTime- and event-based workflow execution policiesOrganizing and executing widespread deployments of complex, multifaceted applicationsCI\/CD pipelinesETL, Data Analytics & Data ScienceData processing pipelines Argo is currently being utilised for new discoveries in physics CERN, for 3D rendering at CoreWeave. The latest use is in Intuit’s machine learning and data processing platform. Argo Workflows is used by over 100 companies, including Adobe, BlackRock, Capital One, Alibaba Cloud, Data Dog, Google, GitHub, Datastax, IBM, Intuit, NVIDIA, New Relic, SAP and Red Hat. (Source : GitHub – Argo Workflow v3.0) Argo Workflows v3.0 Since its inception three years ago, Argo Workflows has come a long way. It is not only cloud-native but also simple, fast and cost-effective. The latest launch of Argo Workflows v3.0 comes with an array of new features and robust upgrades such as the introduction of brand new APIs for Argo Events, controller high-availability, moving the repository including Go modules support. Furthermore, its key-only artefacts feature makes it easier to perform map-reduce operations. New UI Argo’s new UI now also supports Argo Events. A new event-flow page enables users to view how event sources and sensors are linked and display animations whenever a message is received. Along with this, one can create and update event sources and sensors directly in the user interface using the same visual language used for workflows. It also has a new workflow log viewer that easily allows one to view the init and tail the whole workflow. Argo vs others As mentioned earlier, there are many new tools for orchestrating tasks and data workflows in the market, each with a unique quality that can make it hard to choose which ones to use. While Apache Airflow is the most popular and most feature-rich programme, most teams turn to Argo when they are already using Kubernetes. On the other hand, Luigi is a simpler option for beginners. MLFlow and Kubeflow are better tailored to serving a narrow set of requirements to deploy machine learning models and implement experiments. Since Argo is built on top of Kubernetes, each task is performed as a separate Kubernetes pod. It is designed for teams who already use Kubernetes for most of their infrastructure but prefer YAML to describe them rather than Python. Argo’s goal is to enable developers and operators to adopt containers and Kubernetes for developing and deploying distributed applications. Argo is being used in a number of open source projects and enterprises including RocksDB Cloud, Heron, Apache BookKeeper, Apache Pulsar, and AppDynamics. It makes Kubernetes accessible to a broader community of developers. Last year, the Argo Project joined the Cloud Native Computing Foundation (CNCF) as an incubation-level hosted project. The Argo project hopes to closely work with a number of projects that are already members of the foundation, and to “empower organizations to declaratively build and run cloud-native applications and workflows on Kubernetes using GitOps,” Intuit VP of Product Development Pratik Wadher said in a statement.","excerpt":"Argo is an open source container-native workflow engine for Kubernetes.","categories":["AI Features"],"tags":["Kubernetes"],"author_name":"Ritika Sagar","publish_date":"2021-07-23T13:00:00","publication_year":"2021","word_count":685,"keywords":["data science","machine learning","AI","ML","Python","Ray","Kubeflow","analytics","MLflow","Kubernetes","kubernetes"],"extracted_tech_keywords":["AI","machine learning","ML","data science","analytics","MLflow","Kubeflow","Ray","kubernetes","Python"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/argos-rise-to-leading-kubernetes-workflow-engine-in-just-4-years\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10091093,"title":"OpenAI Eyes Japan Expansion, Succumbs to Europe","content":"Amid facing bans on ChatGPT across European countries, OpenAI’s CEO, Sam Altman said that he is considering expanding services by opening an office in Japan. He recently met Japan’s Prime Minister Fumio Kishida where they spoke about the merits and the risks of privacy and security of the technology, as told by Hirokazu Matsuno, the chief cabinet secretary, according to reports by Reuters. Matsuno said that Japan is currently evaluating the possibilities of introducing OpenAI’s technology in the country, said Matsuno. After meeting with Kishida, Altman told reporters, “We hope to … build something great for Japanese people, make the models better for Japanese language and Japanese culture.” This is one of the first stops of Altman’s world tour after the launch of ChatGPT. Furthermore, Taro Kono, who is responsible for Japan’s digital transformation in the cabinet, expressed optimism that AI technologies would play a significant role in the government’s workstyle reforms. However, he acknowledged that introducing ChatGPT into public offices would be challenging in the near future due to issues such as the potential for the technology to produce false information. ChatGPT vs the World Last month, ChatGPT faced a temporary suspension in Italy because it failed to comply with the European Union’s General Data Protection Regulation, which includes the “right to be forgotten.” Unfortunately, there is currently no method for people to request the deletion of their data from a machine learning system once it has been employed to train a model. For this OpenAI is likely to reach an agreement with the Italian government soon as Altman has promised. Moreover, in Australia, Mayor of Hepburn Shire, Brian Wood said that he might sue OpenAI for the erroneous statements ChatGPT made about him serving prison term for bribery. In addition to that, Office of the Privacy Commissioner of Canada (OPC) is also investigating OpenAI for a complaint alleging the collection and use of personal information of users without consent. China, Russia, and North Korea have also banned ChatGPT. Recently, the Indian government acknowledged ethical issues associated with AI, such as bias and privacy, and has taken steps to establish a robust regulatory framework. However, they have not yet proposed any legislation. Meanwhile, the European Union (EU), is bringing in the much-anticipated AI Act this year. In the US too, the government released a blueprint for an AI Bill of Rights. Amid all this, OpenAI released its own version of AI Safety and regulations, which spoke about increasing factual accuracy and addressing privacy concerns. The paper also read that the company will delete users’ data on request and also only use the provided data “wherever feasible”. The company version of AI safety seems more fluff than real.","excerpt":"Japan’s Prime Minister Fumio Kishida and Sam Altman spoke about the merits and the risks of privacy and security of AI technology","categories":["AI News"],"tags":["ChatGPT","GPT-4","Sam Altman"],"author_name":"Mohit Pandey","publish_date":"2023-04-10T17:37:07","publication_year":"2023","word_count":448,"keywords":["Go","ChatGPT","machine learning","Sam Altman","OpenAI","AI","digital transformation","GPT-4","Git","GPT","AI safety","R"],"extracted_tech_keywords":["AI","machine learning","ChatGPT","OpenAI","R","Go","Git","GPT","AI safety","digital transformation"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/openai-eyes-japan-expansion-succumbs-to-europe\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10120078,"title":"Robotics will have ChatGPT Moment Soon","content":"Back in 2024, as AIM expressed excitement for the upcoming GPT moment, little did we anticipate the convergence of numerous robotics advancements. Recently, Vinod Khosla, founder of Khosla Ventures and the initial investor in OpenAI, underscored his perspective on why robotics will soon have its AI breakthrough. Khosla said AI’s transformative capabilities foresee a future where AI and robotics liberate humanity from mundane tasks. He said that robotics will approach a ‘GPT moment’ in the next 2-5 years, when robots will transition from being programmed (following instructions) to learning systems that understand the physical and real-world dynamics, enabling rapid progress in robotics. It’s Already Happening A few days back, NVIDIA researchers introduced DrEureka, an LLM-powered agent automating the simulation-to-reality pipeline, effortlessly training a robot dog to balance on a yoga ball without fine-tuning. We trained a robot dog to balance and walk on top of a yoga ball purely in simulation, and then transfer zero-shot to the real world. No fine-tuning. Just works. I’m excited to announce DrEureka, an LLM agent that writes code to train robot skills in simulation, and writes more… pic.twitter.com\/kuG14LmSOh— Jim Fan (@DrJimFan) May 3, 2024 Interestingly, DrEureka is built on its prior work, Eureka, the algorithm that teaches a 5-finger robot hand to do pen spinning. “It takes one step further in our quest to automate the entire robot learning pipeline with an AI agent system,” said Jim Fan, senior research manager and lead of Embodied AI (GEAR Lab). The OpenAI-powered Figure 01 has also been advancing significantly in terms of visual reasoning capabilities. Recently, it was able to differentiate between healthy options like oranges and less desirable choices like chips, with its in-house trained neural network mapping camera input to robot actions at a rapid 10 hz rate. Brett Adcock, the founder of FigureAI robots, believes that “everyone will own a robot in the future similar to owning a car or phone today,” he added. Tesla is not left behind. Recently, Optimus was ready to work in factories, sorting battery cells in real-time by leveraging its FSD (full self-driving) computers. It was able to sort battery cells precisely with minimal margins for insertions and automatically target the next available slot. Trying to be useful lately! pic.twitter.com\/TlPF9YB61W— Tesla Optimus (@Tesla_Optimus) May 5, 2024 Google DeepMind also released three robotics research systems starting this year—AutoRT, SARA-RT and RT-Trajectory—that will aid robots in making faster decisions and better understanding and navigating their environments. The models will help with data collection, speed, and generalisation. Additionally, Stanford University introduced Mobile ALOHA, a system designed to replicate bimanual mobile manipulation tasks requiring whole-body control. Google DeepMind supported the project, and the technology addresses the limitations of traditional imitation learning from human demonstrations. These general-purpose robots are demonstrated to assist with various tasks such as cooking, cleaning, lifting weights, and other manual activities. What’s Next? While advancements in AI research are still common, companies are rushing for the next big breakthrough in robotics. Like NVIDIA releasing project GR00T a month ago and subsequently releasing Dr Eureka, as mentioned earlier, more companies have also been heavily investing in robotics. Major players like Google DeepMind, Tesla, and NVIDIA are making robotics a priority, so major breakthroughs will likely come soon. Significant progress has also been made in open-source research, with Hugging Face launching LeRobot, an open-source robotics data library, just a couple of days ago. As NVIDIA CEO Jensen Huang rightly said, “The enabling technologies are coming together for leading roboticists around the world to take giant leaps towards artificial general robotics.” Clearly, the ChatGPT moment in robotics is not about when; it is now!","excerpt":"Brett Adcock, the founder of FigureAI robots, believes that “everyone will own a robot in the future similar to owning a car or phone today.”","categories":["Deep Tech"],"tags":["AI in Robotics","ChatGPT","Google","NVIDIA","Vinod Khosla"],"author_name":"Gopika Raj","publish_date":"2024-05-10T16:01:34","publication_year":"2024","word_count":602,"keywords":["Go","ChatGPT","Hugging Face","API","Vinod Khosla","AI","neural network","OpenAI","AI in Robotics","RAG","Aim","Google","NVIDIA","R"],"extracted_tech_keywords":["AI","neural network","ChatGPT","OpenAI","Aim","Hugging Face","RAG","R","Go","API"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/robotics-will-have-chatgpt-moment-soon\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10169463,"title":"This Indian Startup is Using Drones to Deliver Groceries to Your Doorstep","content":"Giving a glimpse of the future of urban living, a resident of Prestige Falcon City in Bengaluru received a grocery delivery via drone in less than 10 minutes. Online grocer Big Basket collaborated with SkyeAir, the drone mobility startup, to deliver the products through its drones. Skye Air said the drones can carry 7 to 10 kg, and are faster than regular delivery services. Ankit Kumar, the founder of Skye Air, spoke with AIM in depth about how the drone system works and how it is changing the country’s quick delivery service scenario. “Drones are designed to make one delivery every 10 minutes. That means a single drone can complete around six deliveries per hour, essentially doubling the efficiency compared to a rider on a bike. And because drones deliver so quickly, we can do patching, which wasn’t feasible with road logistics,” Kumar said. With human-led delivery by road, customers must wait a long time before collecting and delivering a patched package. But with drones, since they’re often heading to the same apartment buildings, patching becomes possible. In multiple instances, Skye Air has successfully achieved patching levels of up to 30 to 40%. Kumar said that SkyeAir is using a generic LLM to optimise the delivery cycle and ensure the safety of the drone and the people around it. This is not the first time drone delivery has been attempted in the country. In 2022, Swiggy initiated a request for proposals, inviting drone service providers to explore drone delivery of groceries in cities like Bengaluru, Delhi NCR, Mumbai, Hyderabad, and Goa. Even TSAW Drones in Bengaluru attempted to deliver the same for healthcare and quick commerce. How Does the Drone Fly? Regarding privacy, Kumar clarified that the drone has no cameras, so it cannot eavesdrop on anyone. The team developed a sky tunnel and mapped the entire city in 3D. The drone follows the 3D map, which takes care of all the static obstacles, such as the building, the telecom tower, and the power lines, as they are already mapped in the system. For tackling dynamic objects, such as birds or kites, Skye Air also uses LiDAR for real-time scanning and decision making without the requirement of cameras. The team has also installed parachutes on the drones, just in case the drone malfunctions, the batteries fail, or it cannot fly. The drone has six motors, and even if two of them fail, it can easily land with four of them. It has an auxiliary battery so that in case the battery fails, the drone will not crash; it has a secondary flight termination system. In case of a short circuit, which occurs due to electrical or wiring failure, the secondary flight termination system will automatically trigger, letting the drone come down safely. “The drone should not go and land everywhere, so we have built in a winching mechanism. The drone will drop a package through the winch mechanism, which is currently there,” Kumar said. When the drone reaches the skypod area, which is flying at 120 m, it lowers its altitude to 20 m, and the packages are dropped to the pod. The pod is connected to the drone at all times so that it can sync with the drone about the environment. The Journey So Far “We were the first in the country to do a BVLOSS flight for medicine delivery. We delivered medicine in Bikarabad, Telangana,” Kumar said. The drone service started in Gurgaon in mid-2023, and it’s been two years. To date, they claim to have made close to two lakh shipments in a month. Skye Air has been transporting multiple clients to about 60+ societies, complexes, and houses. In Bengaluru, they collaborated with Bigbasket, and in Gurgaon, they partnered with other quick commerce platforms like Zepto and Swiggy and brands like Tata1mg, Flipkart, and Apollo. Kumar is planning to build Skye Air on a larger scale for the aerial taxi industry. For this, the team plans to use LLMs for drone-to-drone communication and better navigation. “The second thing is that we still have the man in the loop, taking the package from the pod and then going to your doorstep and delivering it. I think that’s one area we are working on: how to automate that and make it much better.”","excerpt":"Skye Air is using a generic LLM to optimise the delivery cycle and has also built Skyconnect, a 5G device for connectivity.","categories":["AI Startups"],"tags":["drone delivery","LLM","Zepto"],"author_name":"Amisha Arya","publish_date":"2025-05-08T18:44:00","publication_year":"2025","word_count":715,"keywords":["Go","drone delivery","programming_languages:R","AI","LLM","programming_languages:Go","Aim","ViT","GAN","R","Zepto","startup"],"extracted_tech_keywords":["AI","Aim","R","Go","GAN","ViT","startup","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-startups\/this-indian-startup-is-using-drones-to-deliver-groceries-to-your-doorstep\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":56312,"title":"AI-Based PropTech Company Cherre Scores $16 Million Funding Led By Intel Capital","content":"While real estate is one of the most significant sectors, the advancement of cutting edge technologies like artificial intelligence has not been that impactful as expected. One of the primary reasons is the lack of data which is collected on real estate that can be used to create useful models. However, the scenario seems to be changing fast as AI is now increasingly invading the real estate sector. Case in point, the recent investment round of Cherre, a company that is utilising artificial intelligence on various public and private data sources on properties across the globe. The $16 million funding round has been led by Intel Capital, bringing the total capital raised to-date to $25 million. It is interesting to see that Intel Capital has led the funding of Cherre, which speaks volume on what it’s going to mean for the real estate space. Intel Capital has been investing in innovation-led startups in the field of artificial intelligence, autonomous vehicles, datacenter and cloud, 5G, next-generation compute and a big range of other disruptive digital technologies. And, since 1991, Intel Capital has invested $12.6 billion in about 1,560 startups across the globe. According to experts, we may be staring at highly integrated data systems for real estate very soon, which is going to meet the innovation needs. The size of the real estate market is similar to the stock market, which has been impacted by AI and automation for years now, yet the former has been lagging on innovation. To close this gap, startups like Cherre are leading the charge. No wonder Cherre’s tech stack is designed after high-frequency trading platforms, with interconnected data repositories for real-time analytics and high throughput. It also includes the assistance of machine learning models which automatically identify and index real estate data, and give relevant information on demand in an easy and readable format. What’s New About Cherre? Cherre’s software-as-a-service (SaaS) suite includes many fundamental components like CoreExplore, CoreConnect, CoreAugment, and CorePredict, with different functionalities on data management and integration. According to Cherre, its data network provides the most precise and updated real estate data to enable mission-critical operations. With the platform, users are able to connect all of the internal and external data from any source to create a customised data architecture for real estate. Multiple notable companies including banks, investment firms, and insurance companies are using Cherre for the collection and real-time analysis of real estate data sets across tends of sources for AI-powered investment assistance, lending and automated underwriting processes using high-performance, low-latency APIs. This is powered by Cherre’s knowledge graph which it claims to have granular, actionable insights on over 177 million properties, 84 million enterprises globally, and over 315K datasets, spanning across economic, geospatial, and demographic-based datasets. Cherre’s predictive models take in real estate metrics such as amenities, materials, and zoning; community and geographic demographics including population, languages, and schools; lien and mortgages; geospatial data; boundaries; tax assessment; and valuations from more than 450 million transactions for 100 million properties. Trina Van Pelt, VP & Senior MD of Intel Capital, said in a statement, “The global real estate industry is going through a transformation, driven by massive data flows and AI application. Despite its big impact on the global economy, the real estate sector is still in infancy when it comes to data-based investing and underwriting decisions. Intel sees Cherre as critical infrastructure to accelerate the future of this industry. Our customer diligence repeatedly signalled Cherre’s AI-powered platform to be a foundational pillar, a data system of record for big enterprises in the real estate segment. We’re excited to help accelerate Cherre’s worldwide growth trajectory.” Overview 2019 became the tipping point for innovation in the real estate sector, something they call as PropTech. According to a report, the funding of 2019 in PropTech companies has been totalled $24.9 billion, a 157% increase from 2018, with 39 deals over $100 million. Here, technologies like AI, data management, analytics, and blockchain have made a significant impact in the last few years. According to analysts, given the scale of siloed data in this industry, the opportunity is ripe for tech companies to extract value, given global real estate stands at more than $200 trillion. Read more: HOW INDIA’S BIGGEST PROPERTY TECH STARTUP SQUARE YARDS IS USING ANALYTICS TO UNIFY THE REAL ESTATE MARKET","excerpt":"While real estate is one of the most significant sectors, the advancement of cutting edge technologies like artificial intelligence has not been that impactful as expected. One of the primary reasons is the lack of data which is collected on real estate that can be used to create useful models. However, the scenario seems to […]","categories":["Deep Tech"],"tags":["Intel"],"author_name":"Vishal Chawla","publish_date":"2020-02-10T15:25:38","publication_year":"2020","word_count":720,"keywords":["Go","API","machine learning","artificial intelligence","ELT","AI","Git","Aim","analytics","R","Intel"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","analytics","Aim","R","Go","Git","API","ELT"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/ai-based-proptech-company-cherre-scores-16-million-funding-led-by-intel-capital\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":24394,"title":"This Neural Mesh Renderer Can Convert 2D Images Into High-Resolution 3D Objects","content":"3D objects on a computer screen look like real life, and with the 3D glasses on, it’s almost like witnessing the event live. But is it possible to convert a two dimensional image to a 3D object and make it “come alive” with artificial intelligence? Let us dive into a research project developed in Japan. Mesh rendering gives up exceptional objects by constructing it with the help of neural networks. This process usually involves conversion of a 2D image into 3D by overlaying the image over a 3D object. It is then redefined with the backward pass of 3D rendering and then pushed through a neural network. This platform has been explored by researchers Hiroharu Kato, Yoshitaka Ushiku, Tatsuya Harada from the RIKEN institute in Japan. This paper explains how the team built a solution for mesh rendering with other gradients. A polymesh is a promising candidate to get this job done. So what is polymesh? A 3D model which is composed of polygons. The polygons are connected to each other to form a sort of net (or ‘mesh’) that defines the shape. It is not an easy process to render a polymesh of a two-dimensional image onto a three-dimensional object; this is composed by integrating a process called rasterization, which prevents back-propagation Single-Image 3D Reconstruction If an aeroplane can be constructed by this method with the help of a 2D image, it is definitely possible to reconstruct any 3D model a human can perceive. Rendering with a voxel-based method gives a not-so-sharp 3D model like below — but with this technique it is possible to develop more dynamic objects with higher resolution. Comparison With Voxel-Based Method Mesh reconstruction does not suffer from the low-resolution problem and cubic artifacts in voxel reconstruction. Lets us look at how these techniques incorporate the neural network to build high resolution models: This approach outperforms the voxel-based model in 10 out of 13 categories. This table shows the comparison of the metric with the voxel-based model. 2D-To-3D Style Transfer The styles of the paintings are accurately transferred to the textures and shapes by mesh-based rendering method. In the examples below, one can observe it in the outline of the rabbit and the lid of the teapot. The style images are Thomson No. 5 (Yellow Sunset) (D. Coupland, 2011), The Tower of Babel (P. Bruegel the Elder, 1563), The Scream (E. Munch, 1910), and Portrait of Pablo Picasso (J. Gris, 1912). 3D DeepDream The team has also worked on implementing Google’s deep dream tech onto their 3D models. This technique gives us a qualitative sense of the level of abstraction that a particular layer has achieved in its understanding of images. This technique is called as Inceptionism in reference to the neural net architecture used. (Explore Inceptionism gallery for more pairs of images and their processed results, plus some cool video animations.) Technical Overview Incorporating the 2D images to 3D world is one of the hardest problems in the field of computer vision. And rendering (3D-to-2D conversion) lies on the edge between 3D and 2D world. A polygon mesh has been found to be an efficient and intuitive way of representing in 3D. Hence, the backward pass of rendering a 3D is worth implementing. Rendering cannot be implemented onto the neural networks without modifications because the back-propagation prevented from the renderer. In this work, it has been proposed to find the approximate gradient for rendering. The pipeline of actions conducted to the processes involved in this technique is represented below. The 3D mesh generator has been trained with the silhouette images. The generator tried to minimise the difference between the silhouettes reconstructed into 3D shape. 2D-to-3D style transfer was performed by optimising the shape and texture of a mesh to minimise style loss defined on the images. 3D DeepDream was also performed in a similar way. Both applications were realised by flowing information in 2D image space into 3D space through our renderer. Code The neural network renderer code is available for the public here. One can make use of this code to replicate their model and experiment on it. Other applications like 3D reconstruction, style transfer and DeepDream are currently being constructed. With tuning the parameters, it possible to extract higher resolution images compared to the ones posted above.","excerpt":"3D objects on a computer screen look like real life, and with the 3D glasses on, it’s almost like witnessing the event live. But is it possible to convert a two dimensional image to a 3D object and make it “come alive” with artificial intelligence? Let us dive into a research project developed in Japan. […]","categories":["AI Features"],"tags":["Deep Learning","Machine Learning","Neural Network"],"author_name":"Kishan Maladkar","publish_date":"2018-05-09T06:42:50","publication_year":"2018","word_count":715,"keywords":["Neural Network","Go","Replicate","artificial intelligence","programming_languages:R","AI","neural network","Machine Learning","programming_languages:Go","computer vision","Deep Learning","R","ai_applications:computer vision"],"extracted_tech_keywords":["AI","artificial intelligence","neural network","computer vision","R","Go","Replicate","programming_languages:R","programming_languages:Go","ai_applications:computer vision"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/this-neural-mesh-renderer-can-convert-2d-images-into-high-resolution-3d-objects\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10018995,"title":"What Is AI Incident Database?","content":"Today, businesses and government organisations are increasingly deploying intelligent systems to safety-critical problem areas such as healthcare, credit scoring, law enforcement, aircraft control, and corporate recruitment. Failures of such systems pose severe risks to life and expose the limits of intelligent systems deployed in real-world situations. The wrongful arrest of Robert Williams due to a flawed facial recognition system is a good case in point. Experts believe AI practitioners should be aware of past failures of intelligent systems to avoid such fiascos. To that end, the Partnership on AI (PAI), a nonprofit organisation established to outline best practices in AI technologies, has introduced the AI Incident Database (AIID). A compelled repository of AI failures, AIID helps practitioners figure out what can go wrong when the system is deployed. Simply put, the database makes it easy for AI practitioners to learn from previous mistakes. Led by Sean McGregor, the technical lead of IBM Watson AI XPRIZE, AI Incident Database provides an infrastructure supporting AI best practices, a dataset of more than one thousand incidents, and an architecture for building research products. Also Read: Are Blockchains More Secure Than Distributed Databases? Nuts & Bolts The AI Incident Database (AIID) catalogues more than 1,000 publicly available incident reports, including documents and reports from the academic press. According to the paper released by McGregor, these AI failure reports serve a range of purposes, starting from providing multiple viewpoints on incidents, to the number of publications types that double as a proxy for interest in the incident. Additionally, sampling multiple reports per incident would provide more comprehensive coverage of the incident, which increases the practitioners’ chances to discover relevant incidents. Explaining the process, McGregor told media that considering AI systems learn to operate from their training data, it can easily change its conduct based on such data. Thus, such AI-based safety-critical systems can map new possibilities for failure. According to McGregor, most incidents submitted revolve around Ethical AI, especially facial recognition systems, followed by failures in autonomous cars and trading algorithms that either cause substantial damage or put lives at risk. AI practitioners can even search the database based on keywords, source, and authors involved, to get a 360-degree view. For instance — searching for ‘facial recognition’ will bring up 98 reports of AI incidents involving failures or problems related to automatic face recognition, biometric identification, identity verification etc. The search can be further refined based on the requirements. The database’s outline has been massively inspired by the ‘Aviation Accident Reports’ — a shared database critically designed for managing flight safety by analysing the aircraft’s past incidents. McGregor said, the AI Incident Database will help users manage the safety of the AI systems deployed in the real world. The AIID is a collection of web applications that interfaces with a MongoDB document database storing incident report text and metadata. Also Read: Why Do Facial Recognition Systems Still Fail Applications Such pragmatic coverage of AI incidents can help AI practitioners discover and understand past experiences and create more possibilities in deploying AI systems in real-world applications. McGregor explained some of the critical areas, including deploying AI-powered recommendation systems or integrating ML systems to reduce financial and compliance risks. Engineers can also use AIID to learn more about the environment their systems are deployed within, and researchers can understand the AI systems’ safety and fairness. Identifying AI failures by the PAI members started in 2018. However, nobody kept a record of it until now. McGregor has even open-sourced the project on GitHub, where he has welcomed industry users to improve its capabilities and build taxonomies and data summaries in the AIID codebase. Making the database shareable will persuade technology companies to evaluate the bad outcomes before implementation. In due time, McGregor hopes the database will develop into community-owned infrastructure to help create beneficial intelligent systems for the greater common good. Read the paper here.","excerpt":"Today, businesses and government organisations are increasingly deploying intelligent systems to safety-critical problem areas such as healthcare, credit scoring, law enforcement, aircraft control, and corporate recruitment. Failures of such systems pose severe risks to life and expose the limits of intelligent systems deployed in real-world situations. The wrongful arrest of Robert Williams due to a […]","categories":["Deep Tech"],"tags":["Types of Databases","what is database"],"author_name":"Sejuti Das","publish_date":"2021-01-27T17:00:00","publication_year":"2021","word_count":647,"keywords":["Go","AI","Types of Databases","MongoDB","ML","recommendation systems","Git","RAG","BERT","what is database","GitHub","R"],"extracted_tech_keywords":["AI","ML","RAG","recommendation systems","MongoDB","R","Go","Git","GitHub","BERT"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/what-is-ai-incident-database\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10059022,"title":"Infosys makes &#8216;Global Top Employers’ list in a row, No.1 in India","content":"Top Employers Institute has recognised Infosys among the ‘Global Top Employers’ for the second consecutive year. Infosys was also ranked No. 1 Top Employer in India. Infosys is one of 11 companies worldwide to receive this recognition. Infosys has been named Top Employer across Asia Pacific, North America, Middle East, and Europe. The company has been top-ranked in 16 out of the 22 countries surveyed, including India, Australia, Japan, China, USA, Canada, UAE, Oman, the U.K, Ireland, France, Switzerland, and Sweden. The Top Employer Certification highlighted Infosys’ focus on supporting employees’ well-being and experience, especially during the pandemic. “We are delighted to be awarded Global Top Employer again this year. This comes at a time when we have strengthened our approach to employee engagement, making it more purposeful and morale-boosting. We continue to significantly invest in digital learning for our workforce creating new avenues for their growth. Infosys’ internal talent marketplace also helps them move continuously upward in the value chain, delivering on our promise of ‘careers that never stand still’,” said Krish Shankar, Executive Vice President and Group Head of Human Resource Development, Infosys. The Top Employers Institute program certifies organisations based on their Best Practices across six HR domains consisting of 20 areas such as People Strategy, Talent Acquisition, Work Environment, Learning, Well-being, and Diversity & Inclusion. Top Employers Institute conducted a detailed assessment of Infosys’ people practices through the HR Best Practices across 22 countries. David Plink, CEO, Top Employers Institute said, “Reflecting on the demanding year that has, like the year before it, impacted organizations across the world, our Global Top Employers have continued to prioritize going above and beyond the norm to maintain their excellent people practices in the workplace. As a global Top Employer, Infosys has proven its unwavering commitment to employees on a global scale, joining a niche group of companies that have achieved a certification through the Top Employers Program. We are excited to celebrate and applaud them for their achievement in 2022.”","excerpt":"Infosys is one of 11 companies worldwide to receive this recognition.","categories":["AI News"],"tags":["Infosys","pandemic"],"author_name":"Meeta Ramnani","publish_date":"2022-01-24T12:56:50","publication_year":"2022","word_count":331,"keywords":["Go","pandemic","Infosys","AI","programming_languages:R","programming_languages:Go","Git","GAN","R"],"extracted_tech_keywords":["AI","R","Go","Git","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/infosys-makes-global-top-employers-list-in-a-row-no-1-in-india\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10120667,"title":"OpenAI Needs Apple, Badly!","content":"When it comes to AI, OpenAI is killing it as an undisputed leader in the space. Its multibillion-dollar partnership with Microsoft has worked rather well for the company. However, the tech couple seems to be drifting apart – slowly but surely. And now, as things stand, OpenAI badly needs Apple. Take OpenAI’s Spring Update conference, for instance, where the company announced the launch of GPT-4o. Though Microsoft is one of the biggest backers of OpenAI, and the company has trained all its models with Microsoft’s Azure and GPU clusters, there was no single mention of Microsoft throughout the conference. On the other hand, after rumours circulating around a possible partnership between Apple and OpenAI for integrating GPT on iPhone devices, OpenAI seems to be inching closer to the Cupertino giant. This was pretty evident at the conference as well where most of the announcements were made on the iPhone with the ChatGPT app. But what happens to the Microsoft and OpenAI love story? Interestingly, it is not just OpenAI which is moving away from Microsoft. The tech giant is striving for independence when it comes to AI. It is done relying on others for its models. According to reports, though the company had been training smaller models like Orca and Phi all this while (using GPT and incorporating Meta’s Llama on its platform), this time it is training a model large enough to compete with others. Referred to as MAI-1 (possibly Microsoft AI-1), the model is being developed internally by the company, and is around 500 billion parameters in size. Its development is being headed by Mustafa Suleyman, formerly the co-founder of DeepMind and most recently CEO of the AI startup Inflection, who now oversees Microsoft’s AI division. In March, Microsoft acquired a majority of Inflection’s staff and paid $650 million for its intellectual property rights. Though the exact purpose of MAI-1 has not been disclosed yet, it is possible that Microsoft might incorporate its products into all its Copilot products. This would mean that the company will move away from OpenAI’s GPT and Codex models. Meanwhile, Microsoft is making every effort to project a harmonious relationship between the two companies. While the companies are releasing models which seem to rival each other, Microsoft’s CTO Kevin Scott went on LinkedIn to explain that it was not in any way a competition to OpenAI. “I’m not sure why this is news, but just to summarise the obvious: we build big supercomputers to train AI models. Our partner OpenAI uses these supercomputers to train frontier-defining models; and then we both make these models available in products and services so that lots of people can benefit from them. We rather like this arrangement,” he said. But inversely, it would be ideal for Microsoft to have a backup plan just in case the deal with OpenAI falls through, as has been the case with several others in the field. Chief Satya Nadella seems to be playing a different AI game. Under him, Microsoft has invested in all kinds of AI companies, from OpenAI and Mistral to Databricks and Figure AI. Suleyman recently posted on X saying, “AI is everything at Microsoft”. He also highlighted that the company is building massive products using AI and has a definite vision for Copilot. Everything about this seems forced. There seems to be no other reason for Microsoft to build such large models and spend so much on compute if they’re not making it for commercial purposes. Moreover, on his hiring, Suleyman was touted as the “new” Sam Altman. Meanwhile, ahead of OpenAI’s most-anticipated partnership with Apple, Altman recently lauded the Cupertino-based tech giant for its technology prowess, saying, “iPhone is the greatest piece of technology humanity has ever made”, and it’s tough to get beyond it as “the bar is quite high”. Everyone is leaving OpenAI To add another layer to all this is the fact that many OpenAI employees are starting to leave the company. Most recently, former co-founder and chief scientist Ilya Sutskever quit the company to work on something he loves. A few others, like Jan Leiki from the super alignment team of OpenAI, also left with him. Andrej Karpathy, another founding member, left OpenAI. Perhaps, Altman is a genius strategist leading OpenAI to perfection by changing the board, turning it into a for-profit company, launching a search engine to compete with Google, while also eliminating threats like Elon Musk. Even though their products are probably the best out there, the company is also heavily in favour of weeding out competition. For now, OpenAI is positioned very well between the two biggest giants of the globe, Apple and Microsoft. It is becoming the de facto name for AI, which everyone wants to partner with. But as Microsoft is getting heavily self-reliant with Nadella playing 5D chess, OpenAI needs the Apple partnership badly.","excerpt":"“iPhone is the greatest piece of technology humanity has ever made”, and it’s tough to get beyond it as “the bar is quite high,” said Sam Altman.","categories":["Global Tech"],"tags":["OpenAI"],"author_name":"Mohit Pandey","publish_date":"2024-05-16T14:30:13","publication_year":"2024","word_count":808,"keywords":["Go","ChatGPT","OpenAI","AI","GPT-4o","R","RPA","GPT","Azure","Databricks"],"extracted_tech_keywords":["AI","GPT-4o","ChatGPT","OpenAI","Azure","Databricks","R","Go","GPT","RPA"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/openai-needs-apple-badly\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10054938,"title":"MLDS 2022: Major Highlights From Previous Years","content":"It is only a little over a month for the biggest machine learning conference – Machine Learning Developers Summit 2022. And it is happening in person, after two years! The fourth edition of this coveted conference will take place on January 19-20, 2022. For regular attendees, it is the one-stop destination to get acquainted with everything machine learning. However, for those attending for the first time, consider this article a peep into what you are in for! Major highlights of MLDS conferences gone by MLDS 2021 MLDS 2021 was attended by AI and machine learning startups and companies from across the globe. Among the attendees were representatives from large tech companies, SMEs, and startups that are revolutionising the space and exposing the attendees to their growth stories and use cases that worked for them through the years. MLDS 2021 was a three-day event that gave attendees access to machine learning experts, practitioners and entrepreneurs. The event saw more than 1,500 machine learning developers and 60 speakers from 200 companies. MLDS 2021 had a lineup of some of the most established speakers, including – Anish Agarwal (Director – Data & Analytics, India at NatWest Group), Alok Kumar (Head of Data & Analytics at DBS Bank Singapore), Bragadeesh S (Head Of Analytics at Daimler AG), Shirish Gupta (Team Lead\/Data Scientist-1 at Novartis), Sachin Garg (Head of Data Science at PayU), among others. MLDS 2020 MLDS 2020 was conducted over two rounds – one in Bangalore and the other in Hyderabad. The conference covered the technical and business aspects of machine learning that attracted developers and C-suite executives from all over the country. One of the most awaited features of this conference was a hands-on session by Julia programming co-creator Viral Shah where he addressed big trends in AI 2.0, differentiable programming, and implementation of machine learning using Julia. Apart from Shah, Savin Goyal from Netflix gave an insightful talk on ‘More data science, less engineering with Netflix’s Metaflow’ where he demonstrated how Metaflow could be used to make ML processes reproducible. Other presenters include Vinodhini Ranganathan, Senior Data Scientist at Cisco; Data science leader Mathangi Sri; ZS Chief Technology Officer Abhishek Trigunait and Data Science Manager Sagar Madgi; Mohanraj Vengadachalam, Data Scientist at Renault Nissan; Sundara Ramalingam N, Head of Deep Learning practice at NVIDIA, etc. Another major highlight of the event was the machine learning hackathons – MLDS Hackathon and Codeblitz, which were conducted by Analytics India Magazine’s MachineHack team. For the MLDS hackathon, the topic was Drones: StatLog Prediction to classify the status of Drone based on the multiple such instances of Drone Flights. Winners from the two hackathons took away prizes worth more than ₹50,000 in total. MLDS 2019 This was the first in the series of highly successful MLDS events. Organised over a two-day conference, the event was structured as a platform for learning and networking, featuring some of the whos who from the developers’ community in India speaking about the latest topics of AI and machine learning. The key highlight of the summit was the keynote speech by Sandeep Alur, Technical Evangelist, Developer and Platform Evangelism, Microsoft. Alur explained how Microsoft was driving digital transformation and democratising AI with customised solutions on Azure. He also outlined industry solutions and platforms by the tech giant to help developers build models. Day 2 saw a panel discussion on AI For You, which was moderated by Pragati Ogal Rai, the Director-Developer Audience at Microsoft. This discussion saw participation from Vijoy Basu, Senior Director, Data & Analytics, Cognizant, Praveen Srivatsa, Director at Asthrasoft Consulting, and Shweta Gupta, Senior Engineer & Architect at Microsoft. Intel also had a huge presence at the summit. The company representatives conducted a series of tech talks, panel discussions, and keynote talks. Intel’s focus was to expand the AI Software Portfolio that puts tools in the hands of developers. 40 Under 40 Another constant feature at the summit is the much anticipated 40Under40 awards that recognise India’s young data scientists and leaders who nurture, support and further promote the growth of the ML and analytics ecosystem. This year is no different, and the nominations are being accepted for the same. MLDS 2022 will be bigger and grander than the previous years. Watch out this space for more! Note: The company names and designations are corresponding to the year of the event.","excerpt":"For those attending MLDS for the first time, consider this article as a peep into what you are in for!","categories":["Deep Tech"],"tags":["40 Under 40 Awards","Machinehack","mlds india"],"author_name":"Shraddha Goled","publish_date":"2021-12-07T13:00:00","publication_year":"2021","word_count":724,"keywords":["data science","machine learning","AI","Machinehack","R","ML","mlds india","RAG","Aim","deep learning","analytics","40 Under 40 Awards","Azure"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","data science","analytics","Aim","RAG","Azure","R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/mlds-2022-major-highlights-from-previous-years\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10100125,"title":"Will AGI Be Built in China?","content":"Chinese artificial intelligence startup Baichuan Intelligent Technology recently introduced two open-source AI-powered large language models called Baichuan 2-7B and Baichuan 2-13B. Interestingly, one thing that caught everyone’s eye was that it performed better than ChatGPT on AGIEval – a benchmark created by Microsoft Research. ChatGPT’s score on AGIEval was 46.13 while Baichuan 2-13 B’s was 48.17. The word soon spread that Baichuan2-13B beats ChatGPT on AGIEval. Baichuan 2The 2nd iteration the leading Chinese model is a major improvement.Baichuan 2-13B beats ChatGPT on AGIEval.—Code: https:\/\/t.co\/pwJIGDJ6nzPaper: https:\/\/t.co\/j7WRMJ5O5u——The effort is on a different scale.The paper detailing the process for creating… pic.twitter.com\/OpFfODhKyx— Yam Peleg (@Yampeleg) September 13, 2023 This isn’t new. Whenever a new foundational model arrives, it often wants to show how it measures up to ChatGPT. However, the real question here was how Baichuan 2-13 B was able to do that. Language matters The rankings of LLMs on benchmarks often depend on the training dataset they use, and AGIEval is no exception. In AGIEval’s case, it primarily assesses foundational models within the framework of college entrance exams like SAT, LSAT, and various math competitions. Surprisingly, the real reason for outperforming ChatGPT is that the Baichuan 2-13 B was trained on Chinese-English bilingual dataset comprising several million webpages from hundreds of reputable websites that represent various positive value domains, encompassing areas such as policy, law, vulnerable groups, general values, traditional virtues, and more. Upon closer inspection of the AGIEval research paper, it becomes evident that, in addition to entrance exams like SAT and LSAT, it also encompasses Chinese entrance exams such as Gaokao. Furthermore, this benchmark extends its scope to include bilingual tasks in both Chinese and English. On the other hand, open source models like LLaMA and Llama 2 have focused primarily on English. For instance, the main data source for LLaMA is Common Crawl, which comprises 67% of LLaMA’s pre-training data but is filtered to English content only. As Baichuan is China based it has easy access to the chinese material to train its model. Recently, a report came out that said Chinese authorities have approved Baichuan Intelligent Technology and Zhipu AI’ requests to open its AI large language models to the public. It’s apparent that Chinese authorities may not have intervened to prevent them from accessing data from the Chinese internet, distinct from the global internet used elsewhere. Microsoft is Behind This The AGIEval benchmark created by Microsoft says that evaluating the general abilities of foundational models to tackle human-level tasks is a vital aspect of their development and application in the pursuit of AGI. Their paper casually dismisses traditional benchmarks, which rely on artificial datasets and says they may not accurately represent human-level capabilities. Does it mean that Baichuan 2-13B is closer to AGI than ChatGPT. If that’s indeed the case, it’s a significant achievement. However if we introspect, AGIEval is no different from any other benchmark in a way that  they all are based on a certain dataset on which they are evaluated. Apart from AGIEval, if we check Baichuan 2 coding and math problem solving abilities, it is way behind ChatGPT. So how can we conclude that AGIEval is the criteria to judge AGI. Recently, Baidu also claimed that Ernie 3.5, the latest version of its Ernie AI model, had surpassed “ChatGPT in comprehensive ability scores” and outperformed “GPT-4 in several Chinese capabilities.”. The Beijing-based company referred to a test conducted by the state newspaper China Science Daily, which included datasets like AGIEval and C-Eval. Interestingly, Microsoft’s Orca earlier this year had also claimed that it performs better on AGIEval. In the Orca’s research paper it is specifically mentioned that “Evaluation benchmarks like AGIEval which relies on standardized tests such as GRE,SAT, LSAT etc offer more robust evaluation frameworks”. However, If we dig into Orca’s dataset, one finds out that it is also trained on Chinese dataset. Orca scored higher than ChatGPT and was nearly identical to text-davinci-003 in the AGIEval benchmark. However, Orca still significantly lags behind GPT-4 in these metrics.— Tiz (@tatendampofu4) June 16, 2023 The marketing of Orca revolved around the AGIEval benchmark. Similarly majority of the foundational models which are performing well on AGIeval have a Chinese dataset which gives them undue advantage. It isn’t fair to all the other models present out there. In Conclusion The performance of AI models on benchmarks like AGIEval is not solely indicative of their progress towards AGI. While models like Baichuan 2-13 B have showcased impressive scores, the underlying advantage often lies in their training data, particularly the accessibility of specific Chinese internet content. AGIEval’s focus on real-world tasks is commendable, but it’s crucial to recognise a broader spectrum of abilities are equally vital in assessing AGI. Can we really say that if an LLM passes SAT, LSAT or any other exams, it is closer to AGI?","excerpt":"AGIEval Seems to Think So","categories":["AI Features"],"tags":[],"author_name":"Siddharth Jindal","publish_date":"2023-09-15T13:32:47","publication_year":"2023","word_count":801,"keywords":["ChatGPT","artificial intelligence","AI","RPA","GPT","Aim","ViT","llm_models:GPT","R","startup"],"extracted_tech_keywords":["AI","artificial intelligence","ChatGPT","Aim","R","GPT","ViT","RPA","startup","llm_models:GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/will-agi-be-built-in-china\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":43518,"title":"What’s Hot In The Driverless Space ⁠— Datasets, Acquisitions &#038; More","content":"According to a study done by Deloitte, in the US alone, people could travel up to 25 percent more miles by 2040 than they do today. And while the shift to shared autonomy is likely to happen earlier in densely populated urban centers, the rollout will happen over time at different rates in suburban and rural areas. This growing demand for fully autonomous and shared autonomy will redefine the way we travel in the near future. Online tech giants like Google, Apple and Baidu have forayed into the automotive space and with heavy investments and are competing with the likes of Tesla, Audi and BMW. Here are few latest happenings in the world of autonomous driving that will give a hint of what is to come: Lyft Releases Dataset In a bid to unlock access to self-driving research, Lyft released a large-scale dataset featuring the raw sensor camera and LiDAR inputs as perceived by a fleet of multiple, high-end, autonomous vehicles in a bounded geographic area. The Level 5 Dataset is the largest publicly released dataset of its kind. It includes over 55,000 human-labeled 3D annotated frames, a drivable surface map, and an underlying HD spatial semantic map to contextualize the data. Visual Geometric Map of San Francisco via Lyft The Lyft Level 5 Dataset includes: Over 55,000 human-labeled 3D annotated frames; Data from 7 cameras and up to 3 lidars; A drivable surface map; and, An underlying HD spatial semantic map (including lanes, crosswalks, etc.) Download the dataset here. DeepMind And Waymo’s Evolutionary Selection Waymo, in a research collaboration with DeepMind, has taken inspiration from Darwin’s insights into evolution to make this training more effective and efficient. Researchers at DeepMind devised a way to automatically determine good hyperparameter schedules based on evolutionary competition (called “Population Based Training” or PBT), which combines the advantages of hand-tuning and random search. The aim was to investigate whether PBT could improve a neural net’s ability to detect pedestrians along two measures: recall (the fraction of pedestrians identified by the neural net over total number of pedestrians in the scene) and precision. PBT uses half the computational resources used by random parallel search to efficiently discover better hyperparameter schedules. PBT models were able to achieve higher precision by reducing false positives by 24% compared to its hand-tuned equivalent, while maintaining a high recall rate. BAIDU Releases Apollo 5.0 Apollo 5.0 is an effort to support volume production for Geo-Fenced Autonomous Driving. The car now has 360-degree visibility, along with upgraded perception deep learning models to handle the changing conditions of complex road scenarios, making the car more secure and aware. Scenario-based planning has been enhanced to support additional scenarios like pull over and crossing bare intersections. Vehicles with this version can drive autonomously in complex urban road conditions including both residential and downtown areas. AUDI Partners With AID And Luminar Audi has partnered its Autonomous Intelligent Driving group (AID) with the Silicon Valley-based company Luminar to accelerate its autonomous vehicle development. Luminar is known for producing the first safer version of Light Detection and Range (LiDAR) system. Partnering up with startup that deal with cutting edge technology has made Audi one of the top contenders in the autonomous world. Audi has recently debuted its Traffic Jam Pilot (TJP) option on its 2019 A8 flagship sedan. Now drivers can afford a complete hands-off, eyes-off automation in slow traffic, TJP is capable of driving an A8, totally free of driver intervention, in traffic at speeds up to 37 miles per hour. Audi has pledged to spend nearly $16 billion on electric mobility and self-driving tech through 2023. As part of this effort, AID is currently assessing a fleet of 12 test vehicles with Luminar sensors on public roads in Munich. GM Motors’ Cruise Gets $2.25 billion Boost In May 2016, General Motors completed the acquisition of Cruise Automation to boost its autonomous car segment. They paid $291 million in cash and nearly $290 million through issuing new common stocks. Now, with the Trump administration’s approval of Softbank’s whopping $2.25 billion investment in General Motors’ (GM) self-driving unit, Cruise Automation, is poised to claim a larger chunk of the pie. The team behind Cruise have also open sourced a data visualization web application, which named Webviz, where the users  can drag and drop any ROS bag file into Webviz to get immediate visual insight into their robotics data. The user base has grown to include AV engineers calibrating LIDAR sensors, machine learning engineers verifying model outputs, and QA engineers debugging test rides, Webviz has become increasingly feature-rich without compromising its flexibility. Apple Poaches Tesla Engineering VP For Project Titan Apple has poached several notable engineers from Tesla over the last year. Back in March, it hired Tesla drive systems VP Michael Schwekutsch. Last year, Doug Field returned to Apple after a brief stint at Tesla as its chief vehicle engineer. Elon Musk once even took a shot at Apple’s poaching strategy saying, “If you don’t make it at Tesla, you go work at Apple.” Apple, however doesn’t look they are going to change their strategy anytime soon as they have made another hire this month. As first reported by a leading news portal, Steve MacManus has joined Apple as a “Senior Director” working out of Apple Park in Cupertino. MacManus served as Tesla’s vice president of engineering. Prior to his time at Tesla, he worked at Aston Martin, Bentley, Jaguar and Land Rover. Though the future of Apple’s Project Titan is uncertain, these significant hirings does put Apple on the radar. Tesla To Release AutoPilot V10 Back in 2016, Tesla announced that it would be shipping all of its cars with the necessary hardware to support the future advancements in autonomous tech. And, as promised, Tesla not only managed to live up to their ambitious goal but also ended up manufacturing the world’s most advanced chip. Tesla has been a pioneer when it comes to autonomous technology. Tesla enjoys the advantage having started their journey very early. They possess the world’s largest customer base for semi-autonomous vehicle. Whenever a Tesla driver takes an action be it steering left or right or pressing the pedal, what they are doing is annotating the data and generating more refined data related to driver’s behaviour. As the time passes, more data will be generated and every possible driver reaction will be captured. This will give the computers more control over the steering mechanism and perhaps drop the need for having the steering wheel altogether in the future. According to Elon Musk’s social media account, the latest version of Autopilot will include improved self-driving features such as better highway handling, traffic lights and stop sign recognition and “Smart Summon”. The release of V10 (version 10) of Autopilot will also include YouTube and Netflix, as well as several other games and infotainment goodies. Yes, V10 will include several games & infotainment features, improved highway Autopilot, better traffic light & stop sign recognition & Smart Summon — Elon Musk (@elonmusk) July 28, 2019","excerpt":"According to a study done by Deloitte, in the US alone, people could travel up to 25 percent more miles by 2040 than they do today. And while the shift to shared autonomy is likely to happen earlier in densely populated urban centers, the rollout will happen over time at different rates in suburban and […]","categories":["Deep Tech"],"tags":["automotive data analytics","autonomous car","data analytics acquisitions","DeepMind","Google","Mergers and Acquisitions","Tesla","Waymo"],"author_name":"Ram Sagar","publish_date":"2019-07-29T18:20:02","publication_year":"2019","word_count":1169,"keywords":["Go","machine learning","TPU","automotive data analytics","AI","data analytics acquisitions","RAG","Mergers and Acquisitions","Aim","Waymo","deep learning","Tesla","Google","Ray","automation","autonomous car","R","DeepMind"],"extracted_tech_keywords":["AI","machine learning","deep learning","Aim","Ray","RAG","TPU","R","Go","automation"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/whats-latest-in-the-driverless-autonomous-cars%e2%81%a0-datasets-acquisitions-open-source\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10161955,"title":"Kanye West Now Plans to Build an AI Dream Team for Yeezy","content":"American rapper Kanye West is reportedly hiring AI talent for his creative brand Yeezy. The roles include AI engineers, film editors, architects, designers, and studio assistants. The announcement comes despite the ongoing criticism of West’s AI-generated projects, which have drawn mixed reactions. West appears committed to making AI a central creative tool for Yeezy rather than a supporting feature. This isn’t West’s only ambitious initiative recently. He also revisited plans for DROAM, a self-sustainable city concept in Los Angeles, sharing its principles and systems on Instagram. However, like many of West’s ventures, the project remains in the concept stage. Rise of AI-First Studios AI is fundamentally changing the way creative work is done, allowing small, innovative teams to produce high-quality content rapidly and in entirely new ways. Commenting on the rise of AI-first studios, Runway CEO Cristóbal Valenzuela wrote, “The rise of AI-first studios: small groups of creative talent—filmmakers, cinematographers, DPs, artists, actors, editors—accomplishing in weeks what used to take years.” He highlighted their global emergence, mentioning cities like New York, Miami, LA, Paris, Brazil, and New Zealand. These studios, he explained, are not traditional ones adapting to new tools but entirely new entities built on fresh principles. “They’re borrowing workflows and adapting them, experimenting left and right,” he added. “They have no attachment to how things used to be made, only to good ideas and taste.” In September, Runway partnered with Lionsgate, the studio behind John Wick, granting the studio exclusive access to Runway’s AI tools for content creation. This collaboration set a new standard in the entertainment industry and positioned Runway ahead of competitors like OpenAI’s Sora. Rumours suggest a potential AI collaboration between Disney and OpenAI. Similarly, Promise, a studio backed by Andreessen Horowitz, will use generative AI to produce shows and movies. Runway’s tools also played a role in the Oscar-winning film Everything Everywhere All at Once, streamlining special effects production, reducing costs, and minimising manual effort.","excerpt":"West is focused on making AI a key part of Yeezy’s creative process.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","music","Runway"],"author_name":"Aditi Suresh","publish_date":"2025-01-22T13:21:43","publication_year":"2025","word_count":321,"keywords":["Go","API","AI-first","OpenAI","AI","programming_languages:R","ML","programming_languages:Go","music","Runway","generative AI","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","ML","generative AI","OpenAI","R","Go","API","AI-first","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/kanye-west-now-plans-to-build-an-ai-dream-team-for-yeezy\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":51600,"title":"Outstanding Papers Awarded At NeurIPS 2019","content":"The Neural Information Processing Systems (NeurIPS) conference is held every year in the month of December. This year, the 32nd edition was held in Vancouver, Canada. The purpose behind this conference is to foster the exchange of research on neural information processing systems in their biological, technological, mathematical, and theoretical aspects. The conference unveiled the NeurIPS outstanding paper awards for this year. These papers are the most notable accepted papers at the conference. The committee picked up the papers based on some specific criteria from the set of papers which had been selected for verbal presentation. The criteria are: Potential to endureInsightCreativityRevolutionaryRigourEleganceReproducibleScientific This time, the committee also included an additional award called Outstanding New Directions Paper Award in order to highlight work that distinguished itself in setting a novel avenue for future research. In this article, we list down the outstanding papers awarded at NeurIPS 2019. 1| Distribution-Independent PAC Learning of Halfspaces with Massart Noise Category: Outstanding Paper Award About: This paper studies the problem of distribution-independent PAC learning of halfspaces or Linear Threshold Functions (LTFs) for binary classification in the presence of Massart noise. The main contribution of this paper is the first non-trivial learning algorithm for the class of halfspaces (or even disjunctions) in the distribution-free PAC model with Massart noise. Read the paper here. 2| Uniform Convergence May Be Unable To Explain Generalization In Deep Learning Category: Outstanding New Directions Paper Award About: This paper presents examples of overparameterized linear classifiers and neural networks trained by gradient descent (GD) where uniform convergence probably cannot “explain generalisation.” With this research, the researchers tried to understand the goal of a small generalisation bound which shows appropriate dependence on the sample size, width, depth, label noise, and batch size. Read the paper here. 3| Nonparametric Density Estimation & Convergence Rates for GANs under Besov IPM Losses Category: Honorable Mention Outstanding Paper Award About: In this paper, the researchers study the problem of estimating a non-parametric probability density under a large family of losses called Besov IPMs. The paper also shows that the linear distribution estimates, such as the empirical distribution or kernel density estimator, often fail to converge at the optimal rate. Furthermore, the researchers also showed that GANs can strictly outperform the best linear estimator. Read the paper here. 4| Fast And Accurate Least-Mean-Squares Solvers Category: Honorable Mention Outstanding Paper Award About: Least-mean squares (LMS) solvers such as Linear \/ Ridge \/ Lasso-Regression, SVD and Elastic-Net not only solve fundamental machine learning problems but are also the building blocks in a variety of other methods, such as decision trees and matrix factorizations. Keeping this in mind, the researchers presented a novel framework which shows how to reduce the computational complexity of Least Mean-Square solvers by one or two orders of magnitude, with no precision loss and improved numerical stability. Read the paper here. 5| Putting An End to End-to-End: Gradient-Isolated Learning Of Representations Category: Honorable Mention Outstanding New Directions Paper Award About: In this paper, the researchers proposed a novel deep learning method for local self-supervised representation learning which does not require labels nor end-to-end backpropagation but exploits the natural order in data instead. The research is done by splitting a deep neural network into a stack of gradient-isolated modules where each module is trained to maximally preserve the information of its inputs using the InfoNCE bound. Read the paper here. 6| Scene Representation Networks: Continuous 3D-Structure-Aware Neural Scene Representations Category: Honorable Mention Outstanding New Directions Paper Award About: The researchers at Stanford University proposed Scene Representation Networks (SRNs) which is a continuous, 3D structure-aware scene representation that encodes both geometry and appearance. SRNs represent scenes as continuous functions which has the ability to map world coordinates to a feature representation of local scene properties. The potential of SRNs is further demonstrated by evaluating them for novel view synthesis, few-shot reconstruction, joint shape, and appearance interpolation, and unsupervised discovery of a non-rigid face mode. Read the paper here.","excerpt":"The Neural Information Processing Systems (NeurIPS) conference is held every year in the month of December. This year, the 32nd edition was held in Vancouver, Canada. The purpose behind this conference is to foster the exchange of research on neural information processing systems in their biological, technological, mathematical, and theoretical aspects.  The conference unveiled the […]","categories":["AI Features"],"tags":["deep learning research","Gradient Descent Numerical Example","NeurIPS"],"author_name":"Ambika Choudhury","publish_date":"2019-12-10T18:30:00","publication_year":"2019","word_count":659,"keywords":["Go","machine learning","programming_languages:R","AI","Gradient Descent Numerical Example","neural network","RPA","NeurIPS","deep learning research","deep learning","ViT","GAN","R"],"extracted_tech_keywords":["AI","machine learning","deep learning","neural network","R","Go","GAN","ViT","RPA","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/outstanding-papers-awarded-at-neurips-2019\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10140754,"title":"The Beatles’ AI-Assisted Track Makes History, Nominated For Grammys","content":"In a notable first, The Beatles’ track ‘Now and Then’ has been nominated for Best Rock Performance, against Green Day, Pearl Jam, and The Black Keys at the Grammys ’25. Originally recorded as a rough demo by John Lennon in the 1970s, this project finally came to fruition in 2023 when AI helped isolate Lennon’s vocals, allowing Paul McCartney and Ringo Starr to complete the song. This nomination signals a shift in how AI can preserve legacy while enabling innovation in music. While some purists question whether AI belongs in creative fields, AI allowed surviving members Paul McCartney and Ringo Starr to finally complete and release the song in 2023, marking it as a powerful testament to AI’s role in both music preservation and innovation. AI Brings Life This last song, ‘Now and Then’ by The Beatles was a demo given to McCartney by Lennon’s widow Yoko Ono, recorded on a cassette while Lennon played the piano. McCartney previously worked on Lennon’s compositions ‘Free As A Bird’ and ‘Real Love’ using AI. With AI assistance from the ‘Get Back’ documentary, McCartney mixed the song, allowing him to perform with Lennon during his recent tour. Other musicians have similarly explored AI’s capabilities to expand their work: Canadian musician Claire Elise Boucher, known as Grimes, for instance, introduced her AI software, Elf.tech, allowing her fans to use her voice through AI to create new tracks. Besides, composer Holly Herndon uses neural networks to transform her compositions into complex soundscapes. Similarly, David Guetta used AI to recreate Eminem’s voice in his live performances, showcasing AI’s flexibility in creative expression. Additionally, artists like Taryn Southern and Shawn Everett are experimenting with AI tools such as IBM’s Watson Beat and OpenAI’s Jukebox to enhance music production, creating sounds that blur the line between human and machine creativity. According to Ditto Music Research, a study of over 1200+ Ditto users found that nearly 60% of independent musicians in 2023, were using AI to make music. With AI, artists can now go beyond conventional boundaries, and the industry will witness more AI-assisted music compositions in the days to come.","excerpt":"In a notable first, The Beatles’ track ‘Now and Then,’ which used AI, has been nominated for Best Rock Performance for Grammys 2025.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","AI in music","Beatles","Grammys"],"author_name":"Tanisha Bhattacharjee","publish_date":"2024-11-11T13:37:18","publication_year":"2024","word_count":352,"keywords":["AI in music","Go","OpenAI","AI","Grammys","neural network","innovation","programming_languages:R","programming_languages:Go","ViT","Beatles","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","neural network","OpenAI","R","Go","ViT","innovation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/the-beatles-ai-assisted-track-makes-history-nominated-for-the-grammys\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":35046,"title":"Global E-Commerce Policy Has Data Protection As Its Core","content":"Protection, localisation and cross-border flows of data and privacy will be at the core of the upcoming e-commerce policy. It will also focus on India’s position in global trade negotiations, officials in the know said. When a group of 76 countries including the US, European Union, China, Japan, Australia and Singapore are working to come up with trade rules for the e-commerce sector, the Department for Promotion of Investment and Internal Trade (DPIIT) has said that the policy has its focus on India’s interest in the World Trade Organization. “The policy will be WTO-centric and we have studied the respective policies of the EU, US and China,” according to a news report. “It is evolving now but data — be it storage, cross-border flows, privacy, community data or payments — is the main thrust. All other issues are ancillary.” While the Department of Commerce, which deals with India’s trade issues, had floated a draft policy in July last year, the DPIIT was made the nodal department for the government’s e-commerce initiatives in September. The earlier e-commerce draft had also suggested that personal or community data collected by “Internet of Things” devices in “public space” should be stored only in Indian servers. It had also suggested a two-year sunset period before making data localisation mandatory. India has told the WTO the developing countries needed to maintain policy space in certain aspects of e-commerce such as ownership and use and flow of data in “sunrise sectors like cloud computing and data storage” and in the facets related to hosting of servers, big data analytics and M2M communication. “While a data protection Bill is in the offing, the e-commerce policy will not be contradictory to it. The Bill is based on the Justice Srikrishna committee report that talks of personal data but the e-commerce policy has community data as its core. It talks of the economic rights of data,” an expert on data issues said. However, recently it was reported that the DPIIT may not include setting up of a sector regulator as it was mentioned earlier. However, the department clarified that it will continue to include the recently updated changes in FDI policy which have affected the business of global e-commerce companies such as Amazon and Flipkart. India’s e-commerce policy will be significant for the sector as nearly half of the WTO member countries may opt for text-based negotiations on global e-commerce rules. At a time when the FDI guidelines in e-commerce policy led global e-commerce players Amazon and Walmart to lose $50 Bn in market capitalisation, together, the new e-commerce draft with a major focus on domestic players may restrict their growth further in the country.","excerpt":"Protection, localisation and cross-border flows of data and privacy will be at the core of the upcoming e-commerce policy. It will also focus on India’s position in global trade negotiations, officials in the know said. When a group of 76 countries including the US, European Union, China, Japan, Australia and Singapore are working to come […]","categories":["AI News"],"tags":["Amazon","China","data protection","Flipkart","policy","US"],"author_name":"Disha Misal","publish_date":"2019-02-15T12:55:31","publication_year":"2019","word_count":445,"keywords":["big data","Go","API","AI","cloud computing","R","Flipkart","Amazon","RAG","China","analytics","Rust","GAN","data protection","policy","US"],"extracted_tech_keywords":["AI","analytics","RAG","cloud computing","R","Go","Rust","API","big data","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/global-e-commerce-policy-has-data-protection-as-its-core\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":45382,"title":"OYO Acquires Danish Data Science Firm Danamica To Scale Global Vacation Business","content":"OYO buys Danish data science startup Danamica India’s largest hotel chain OYO has acquired Danish data science firm Danamica for an undisclosed sum, the SoftBank backed company announced on Monday. The Indian company which has been on a global expansion spree bought the Copenhagen startup which specializes in dynamic pricing to bolster its global vacation rental business. By making strategic investments in products and processes, OYO is marching into new territories and consolidating its position in the European market.  The Danish data science company is founded by Mads Westberg & Rune Larsen who are confident their pricing technology will help the parent company achieve its growth targets. “Data sciences across pricing, AI, and imaging sciences have been a cornerstone of OYO’s proprietary revenue enhancement technology,” said Maninder Gulati, Chief Strategy Officer of OYO Hotels & Homes. He further added, “It is also a huge missing piece in the way traditional vacation rentals industry is run. We are glad to have found Danamica, which has built expertise in these areas. Both Rune and Mads are extremely talented individuals and together with their team, they have built a valuable IP that analyses many years of data and pricing trends and provides logical and scientific recommendations, which will help us scale our vacation and urban homes business across Europe and other parts of the world.” One of the key benefits provided by Danamica is that the company specializes in dynamic pricing, which will help Oyo to lend more accuracy to pricing, leading to higher efficiencies and yield for their real estate owners and value for money for millions of global guests. “We are very pleased to announce that Danamica is now part of OYO Vacation Homes family,” shared Westberg and Larsen, the founders of Danamica. “OYO’s ambition and growth are extraordinary and we couldn’t imagine taking part in a more exciting journey. OYO and Danamica have a shared understanding of the importance and impact of AI and data science. Like OYO, we recognize the untapped potential in the vacation rental industry that can be fulfilled with a data-driven approach,” they further added.","excerpt":"India’s largest hotel chain OYO has acquired Danish data science firm Danamica for an undisclosed sum, the SoftBank backed company announced on Monday. The Indian company which has been on a global expansion spree bought the Copenhagen startup which specializes in dynamic pricing to bolster its global vacation rental business. By making strategic investments in […]","categories":["AI News"],"tags":[],"author_name":"Richa Bhatia","publish_date":"2019-09-03T17:05:04","publication_year":"2019","word_count":349,"keywords":["data science","programming_languages:R","AI","data-driven","R","startup"],"extracted_tech_keywords":["AI","data science","R","data-driven","startup","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/oyo-acquires-danish-data-science-firm-danamica-to-scale-global-vacation-business\/","complexity_score":2,"technical_depth":6,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10059562,"title":"Kickstart your career in Data Science and Business Analytics with this program from Great Learning","content":"According to McKinsey & Co and Nasscom, data and AI could contribute USD 500 billion to India’s GDP by 2025. The AI & data science market is growing exponentially and is brimming with job opportunities. To that end, Great Learning’s PG Program in Data Science and Business Analytics: V.22 is a springboard to launch a career in the field.  The curriculum of the PGP in Data Science and Business Analytics: V.22 has been updated in consultation with industry experts, academicians and program alums to make it cutting-edge. Program curriculum The program will focus on: Foundations Introduction to data scienceMarketing and CRMStatistical methods for decision makingBusiness financeSQL Programming Data science techniques Optimisation TechniquesPredictive ModelingAdvanced StatisticsData MiningTime Series ForecastingMachine Learning Domain exposure Marketing & Retail AnalyticsFinance and Risk AnalyticsWeb and Social Media AnalyticsSupply Chain and Logistics Analytics Visualisation techniques Data Visualisation Using Tableau Capstone projects The capstone projects will be in the areas of retail, web and social media, e-commerce, banking, supply chain, insurance, finance and accounts etc. What makes this program different? Students will get eight industry sessions, six practice hackathons and three virtual sessions to connect with peersThe participants will also have access to 15 case studiesThey will also have the choice of nine Optional modules -4 from Previous Version +Model Deployment +(GL Elevate AI Accelerator Pack) Power BI, Applications of AI, Cloud Computing, Blockchain.Students will have the option to choose between Saturday\/Sunday Mentoring Sessions.They will be provided with a Handbook on Case Studies by Alumni & Sessions by Alums on ‘How They Applied Data Science at Work’.They will be provided three 1:1 Career Coaching Sessions and three company sponsored hackathons.Students will receive Guidance to publish Capstone Project in reputed Journals or present at conferences. Career support Great Learning will also provide career support through the following mechanisms: e-portfolio An e-portfolio is a snapshot of all the projects done and skills acquired during the program shareable across social media channels to showcase expertise to potential recruiters. Resume building and interview preparation Guidance will be provided to build the students’ resumes to highlight their skills and their previous professional experience. Students will also learn to crack interviews with the help of interview preparation sessions. Career guidance Students will get access to career mentoring from industry experts who’ve transitioned to Data Science roles. Job board The program provides candidates access to the Great Learning Job Board. Over Great Learning works with over 500 organisations to recruit students. The candidates get an average salary hike of 48 per cent on a successful transition. Program benefits Corporate partners The program is designed, delivered and endorsed by leading analytical, technology and consulting firms. The corporate partners of the course are involved in curriculum design, facilitating projects, industry lectures and also suggesting pedagogical improvements. Industry relevant curriculum The curriculum combines academic elegance and business relevance to facilitate participants in learning the basics of management, data science techniques and applications for data-based decision making. Flexible learning The PGP-DSBA program provides utmost learning flexibility. You can learn while you earn with online sessions. The course accommodates transfer cases and sabbaticals and provides the option to catch up when you miss classes. Hands-on exposure An integral part of the learning experience is the use of Data Science and Analytics tools wherein the candidates get hands-on exposure to Python, SQL, Tableau, R (Online). Testimonials The capstone project was one of the most memorable parts of the process – Sherin Shoby, PGP DSBA Shoby comes with a master’s degree in commerce and over seven years of experience in the service industry. After taking a break to raise her child, her journey in Data Science started in 2015 when she enrolled for a Post Graduate Diploma in Applied Statistics through IGNOU and started taking up small freelancing projects. During this time, she was on the lookout for full-time programs where she could meet like-minded people, have interactions and increase her network. So, she took up the Post Graduate Program in Data Science and Business Analytics in 2020 at Great Learning. She adds, “The journey so far has been nothing less than overwhelming and rewarding. I have nothing but gratitude for all the changes I experienced in life. I’ve met amazing people who have become very good friends, teachers who have been excellent guides, and to date, I do not hesitate to call and get my doubts and queries cleared because that is the level of comfort I have had.” The industry-based curriculum broadened the spectrum of new job opportunities – Goushika Madhmitha, PGP DSBA Madhmitha completed her bachelor’s degree in B.Tech-EEE and, as a fresher, joined an IT firm. She was interested in the DBMS process in software development and started to work on projects using ETL applications over seven years across a few organisations. She was fortunate enough to have a solid support system that helped her juggle and concentrate on work, the PGP-DSBA course from Great Learning, and my family at the same time. She adds, “I started using ML applications at my workplace, and this triggered my interest in Data Science. I also started looking for more course content on Data Science. However, lack of exposure and understanding of the field was one of the obstacles, as not many women are working in the data science field.” “It is a learning process, and one should be very agile in understanding the requirements of Data Science from various industries, as the application is endless. However, concentrating more on understanding the basics helps crack and identify problem statements from any dataset. I was at a point where I felt stagnant in my position and was looking to upskill myself without having to quit my current job, and that’s when I started researching about Great learning,” she said. She also stated that Great Learning professors and materials provided for the program assisted greatly with the curriculum. The students also managed to adjust their pace of understanding, despite the pandemic. She learnt under the guidance of amazing faculty; specifically, Prof. Gurumoorthy, who patiently clarified, cleared and revised basic concepts every time the class started advanced topics. The students also learnt from motivating faculty like Prof. Ragavshyam Ramamurthy and Prof. Deepesh Singh, who shared their industry experience while explaining concepts. Industry renowned certificates Students can get two post-graduate certificates: McCombs’ School of Business at UT Austin and Great Lakes Executive Learning, the executive learning arm of Great Lakes (A Top 10 B-School in India). A student can get the most up-to-date learning experience that reflects the changing industry landscape. The program fee is ₹2,40,000 + GST (the top 35 students will be able to reserve the seat at a fee of ₹2,25,000 + GST). Apply for the course here.","excerpt":"The curriculum of the PGP in Data Science and Business Analytics: V.22 has been updated in consultation with industry experts, academicians and program alums.","categories":["AI Trends"],"tags":["AI (Artificial Intelligence)","Data Science","Data Scientist","foundation","Great learning","learning data science"],"author_name":"Sreejani Bhattacharyya","publish_date":"2022-02-02T10:00:00","publication_year":"2022","word_count":1115,"keywords":["data science","machine learning","AI","cloud computing","foundation","ML","learning data science","RAG","Python","analytics","SQL","Great learning","Data Science","Data Scientist","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","machine learning","ML","data science","analytics","RAG","cloud computing","Python","R","SQL"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/kickstart-your-career-in-data-science-and-business-analytics-with-this-program-from-great-learning\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":627,"title":"How ISRO’s GSLV fares against heavy lift launch vehicles – Falcon 9 and Ariane 5","content":"India, a space underdog clinched another victory with ISRO’s 104 satellite launch from a single rocket, fired into space from Satish Dhawan Space Centre in Sriharikota. The super feat jettisoned India into the elite space club dominated by USA, China, Japan and Russia. India surpassed Russia’s 2014 record of launching 37 satellites in the orbit in a single launch. Out of the 104 satellites launched, 88 were Doves satellites, collectively known as “Flock 3p” and weighing about 10 pounds that rode aboard the PSLV rocket. The 88 satellites belonged to San Francisco-headquartered Planet Labs, that operates the biggest constellation of earth-imaging satellites and provides endless data on natural resources and how to respond to natural disasters. And it wasn’t just the Indian community that was watching ISRO’s record-breaking feat. American business magnate Elon Musk congratulated ISRO on the exceptional launch. Big launch is where the big money is – a space dominated by Arianespace and SpaceX India has earned a reputation for launching low-cost, low weighing satellites to Low Earth Orbit (LEO) and Sun Synchronous Orbit (SSO). The preferred launch vehicle is PSLV which has a cap for load capacity – around 3800 kg for LEO and up to 1750 kg for SSO. Globally, big payload is where the big money. Space companies such as billionaire Elon Musk-backed SpaceX and European Arianespace, world’s leading satellite launch company are vying for the heavy lift launch market.  Arianespace’s Ariane 5 is considered the “champion heavy lift” launch vehicle that can deliver heavier payloads into LEO and GTO backed by a great launch history Ariane 5, has launched several European Space agency missions, most notably Rosetta, a space probe built by the European Space agency for studying the Jupiter family comet. GSLV vs Falcon 9 vs Ariane 5 India’s GSLV payload for geostationary transfer orbit (GTO) is only 2,500 kg according to Wikipedia.  And in a stark contrast, Ariane 5 can carry 9,100 kg; and SpaceX’s Falcon 9 can carry 8,300 kg to GTO. But the $300 billion satellite launching market is cornered by Space X and Arianespace that have their order books chock-a bloc. Arianespace, the European space industry veteran has further planned Ariane 6 and Vega C launch system that will bolster its position as the preferred heavy launch system.  In 2015, the European space company carried out 12 missions with 6 Ariane 5 launches. SpaceX has dramatically pared down the cost of launches, with reusable rocket launches to $43 million as compared to United Launch Alliance of $225 million. Falcon Heavy from Space X – took off on Feb 19, 2017 from NASA’s historic launch site And who hasn’t heard about SpaceX’s Falcon Heavy (SpaceX designed and manufactured heavy lift space launch vehicle) that returned to service last week with its private rocket launch from NASA’s iconic site 39A, in Cape Canaveral, Florida that has seen Apollo and Space Shuttle successful missions. SpaceX’s first launch was a cargo mission (Dragon space capsule) to the International Space Station, that was reportedly packed with 2,500 kgs of cargo and will return to earth after 29 days with 2,300 kg of trash that will burn in outer atmosphere. Some of the other space stations additions are —  an earth-monitoring tool that will hunt for ozone in atmosphere and a Space Test Program payload that comprises of Lightning Imaging Sensor, which will track lightning worldwide, and Raven, that will do data collection for future autonomous spacecraft sojourns. Falcon Heavy’s successful landing was greeted by Elon Musk with this Instagram statement, “Baby came back”. The successful landing comes close on the heels of Falcon 9 rocket’s aborted mission that exploded on the launch pad last year. NASA Delta IV Heavy – Big Daddy of Rockets NASA’s Delta IV Heavy is by far the most robust heavy lift launch vehicle with a payload capacity to LEO of 28,790 kg.  Earlier last year, a Delta IV rocket delivered a special payload into the orbit – NROL-37, a spy satellite payload. The world’s most powerful rocket has clocked 9 launches since 2004 and its last outing was in December 2014 when it launched NASA’s Orion spacecraft. Much like the Falcon Heavy rocket, Delta IV Heavy uses three booster cores and if news reports are to be believed, it will soon lose its usability and supremacy to Falcon Heavy that offers a heavy lift service at significantly low cost. As per news reports, a Delta IV comes at a steep price of $14 million per ton to orbit, as opposed to SpaceX’s at $1.7 million per ton. Where does India’s GSLV stand in heavy lift space market? India’s heavy satellite launch vehicle is Geosynchronous Satellite Launch Vehicle (GSLV) mostly used for domestic satellite launches can carry a payload of 2,500 kilograms to GTO. Since its first launch in 2001, GSLV has carried out 10 launches, the most recent being INSAT-3DR in September, 2016. Critics argue that India’s cautious approach can be attributed to the fact that it is state-sponsored program. According to an opinion piece, GSLV was initially built to carry satellites like INSAT and GSAT and not compete with market leaders Falcon 9 or Ariane 5. As per news reports, India’s share of global revenue is only 0.6%. ISRO still has to ramp up its commercial viability in the heavy satellite space. If ISRO wants to increase its market share in the heavy lift launch space, it should have a sizeable budget.","excerpt":"India, a space underdog clinched another victory with ISRO’s 104 satellite launch from a single rocket, fired into space from Satish Dhawan Space Centre in Sriharikota. The super feat jettisoned India into the elite space club dominated by USA, China, Japan and Russia. India surpassed Russia’s 2014 record of launching 37 satellites in the orbit […]","categories":["IT Services"],"tags":[],"author_name":"Richa Bhatia","publish_date":"2017-02-23T10:40:06","publication_year":"2017","word_count":904,"keywords":["Go","ELT","programming_languages:R","AI","RPA","programming_languages:Go","mlops_tools:Comet","RAG","R"],"extracted_tech_keywords":["AI","RAG","R","Go","ELT","RPA","mlops_tools:Comet","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/isros-gslv-fares-heavy-lift-launch-vehicles-falcon-9-ariane-5\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":51139,"title":"Understanding Artificial Neural Network With Linear Regression","content":"Artificial Neural Network (ANN) is probably the first stop for anyone who enters into the field of Deep Learning. Inspired by the structure of Natural Neural Network present in our body, ANN mimics a similar structure and learning mechanism. ANN is just an algorithm to build an efficient predictive model. Because the algorithm and so its implementation resembles a typical neural network, it is named so. The functionality of ANN can be explained in below 5 simple steps: Read the input dataProduce the predictive model (A mathematical function)Measure the error in the predictive modelInform and implement necessary corrections to the model repeatedly until a model with least error is foundUse this model for predicting the unknown A beginner in data science, after going through the concepts of Regression, Classification, Feature Engineering etc. and enters into the field of deep learning, it would be very beneficial if one can relate the functionality of algorithms in deep learning with above concepts. Before understanding ANN, let us understand a perceptron, which is a basic building block of ANN. Perceptron is the name initially given to a binary classifier. However, we can view the perceptron as a function which takes certain inputs and produces a linear equation which is nothing but a straight line. This can be used to separate certain easily separable data as shown in the figure. However, remember that in real-world scenarios, classes will not be so easily separable. The structure of a perceptron can be visualised as below: A typical neural network with multiple perceptrons in it looks like below: This means generating multiple linear equations at multiple points. These perceptrons can also be called as neurons or nodes which are actually the basic building blocks in natural neural network within our body. In the above figure, the first vertical set of 3 neurons is the input layer. The next two vertical sets of neurons are part of the middle layer which are usually referred to as hidden layers, and the last single neuron is the output layer. The neural network in the above figure is a 3-layered network. This is because the input layer is generally not counted as part of network layers. Each neuron in the input layer represents an attribute (column) in the input data (i.e., x1, x2, x3 etc.). What is happening in the above network is that input data is fed to set of neurons, and each produces an output. Again, each of these outputs are fed to other neurons which in turn produces another output, which is again fed to the output layer. Error calculated at this output layer is again sent back in the network to further refine the outputs of each neuron which are again fed to the neuron in output layer to produce a refined output than before. As explained in the 5-step process above, this process is repeated until we get an output with minimal error. The process of producing outputs, calculating errors, feeding them back again to produce a better output is generally a confusing process, especially for a beginner to visualise and understand. Hence, an effort is made here to explain this process with just one neuron and one layer. Once this basic concept is understood, expanding this to a larger neural network is not difficult. Everyone agrees that simple linear regression is the simplest thing in machine learning or atleast the first thing that anyone learns in machine learning. So, we will try to understand this concept of deep learning also with a simple linear regression, by solving a regression problem using ANN. Implementing ANN for Linear Regression We have understood from the above that each of the neuron in the ANN except the input layer produces an output. The output is based on what function that we use. This function is generally referred as ‘Activation Function’. As ANN is mainly used for classification purposes, generally sigmoid function or other similar classification algorithms are used as activation functions. But, as we are now trying to solve a linear regression problem, our activation function here is nothing but a ‘Simple Linear Equation’ of the form – y=w0 + w1x1 + w2x2 + w3x3 + …. wnxn where x1, x2, x3.. xn are the independent attributes in the input data, w1, w3… wn are the weights (Co-efficients) to corresponding attributes, and w0 is the bias Because our output should just be a single linear line, we should configure our ANN with just 1 neuron. As the output of this 1 neuron itself is the linear line, this neuron will be placed in the output layer. Hidden layers are required when we try to classify objects with using multiple lines (or curves). So, we don’t need any hidden layers as well here. Hence the ANN to solve a linear regression problem consists of an input layer with all the input attributes and an output layer with just 1 neuron as shown below: Now, we have finalised the structure of our ANN. Our next task is to actually write code to implement it. We will be implementing this simple ANN from scratch as that will help to understand lot of underlying concepts in already available ANN libraries. Recall the 5 steps that are mentioned at the beginning. As mentioned there, the process involves feeding input to a neuron in the next layer to produce an output using an activation function. This process is called as ‘Feed Forward’. After producing the output, error (or loss) is calculated and a correction is sent back in the network. This process is called as ‘Back Propagation’. We will also use some standard terminologies for our ANN network such as ‘Network’, ‘Topology’ etc. which we will see in the code. With various terms and terminologies that we have learnt so far, let us implement the code – 1. Import the required libraries 2. Initialise the weights and other variables In our approach, we will be providing input to the code as a list such as [2,3,1]. Here, the total no. of values present in the list (list size) indicate the number of layers that we want to configure, and each number in the list indicate the no. of neurons inside each layer. So, the list [2,3,1] indicates our network should consists of 3 layers in which first layer consists of 2 neurons, second layer consists of 3 neurons and output layer consists of 1 neuron. This structure can be called as ‘network topology’. However, as we are solving regression problem, we just need 1 neuron at the output layer as discussed above. So, we just need to pass the input list as [1]. In our approach to build a Linear Regression Neural Network, we will be using Stochastic Gradient Descent (SGD) as an algorithm because this is the algorithm used mostly even for classification problems with a deep neural network (means multiple layers and multiple neurons). I will assume the reader is already aware of this algorithm and proceed with its implementation. We will initialise all the weights to zeros. Let us create a class called ‘Network’ and initialise all required variable in the constructor as below – ‘self.output’ variable in the above code is to hold the outputs of each neuron. It will be initialised accordingly with a sufficient sized list based on our input. Remaining variables are pretty self-explanatory. 3. Coding ‘fit’ function We know that the gradient descent algorithm requires ‘learning rate’ (eta) and no. of iterations (epoch) as inputs. We will be passing all these values in a list to the program along with the training data. Let us build a ‘fit’ method to construct a predictive model with all the inputs given – 4. Produce the Output and Correct the Error I have mentioned above what ‘feed forward’ and ‘back propagation’ are. Let us implement those methods – Above function is just forming a simple linear equation of y = mx + c kind and nothing more. In SGD algorithm, we continuously update the initialised weights in the negative direction of the slope to reach the minimal point. Error function E(w) = ∑[(w0 + w1x1 – y1)2 +(w0 + w1x2 – y2)2+….. +(w0 + w1xn – yn)2] Here, I have not taken ½ as scaling factor to the equation. One may take if desired so. Also, in SGD only one row is passed to the above error function every time to calculate the error. Hence, if we differentiate the above equation w.r.t. each of the weights w0,w1, w2 .. etc., we get equations like ∂E0=1, ∂E1= 2*(w0 + w1x1 – y1)*x1 After calculating the slope w.r.t. each of the weights, we will be updating the weights with new values in the negative direction of the slope as below – Let us implement all this logic in the back propagate function as below: In order to visualise the error at each step, let us quickly write functions to calculate Mean Squared Error (for full dataset) and Squared Error (for each row) which will be called for each step in an epoch. Having the model built in the above way, let us define a method which takes some input and predicts the output – That’s it. We have built a simple neural network which builds a model for linear regression and also predicts values for unknowns. 5. Executing the program In order to pass inputs and test the results, we need to write few lines of code as below – In above code, a sample dataset of 10 rows is passed as input. Full code can be accessed and executed at Google Colab : https:\/\/colab.research.google.com\/drive\/1f84s4nlKSas5LGpR8zdRxWOsKL5HIoyy Sample outputs for given inputs are as below: The plot below shows how the error is getting reduced in each step as weights get continuously updated and again fed into the system. So, we have understood how in few lines of code we can build a simple neural network. The same code can be extended to handle multiple layers with various activation functions so that it just works like a full-fledged ANN. I will implement that in my next article.","excerpt":"Artificial Neural Network (ANN) is probably the first stop for anyone who enters into the field of Deep Learning. Inspired by the structure of Natural Neural Network present in our body, ANN mimics a similar structure and learning mechanism. ANN is just an algorithm to build an efficient predictive model. Because the algorithm and so […]","categories":["Global Tech"],"tags":["ANN","Google Colab","how to calculate mean square error","Neural Networks","simple neural network"],"author_name":"Raja Suman C","publish_date":"2019-12-05T11:00:00","publication_year":"2019","word_count":1682,"keywords":["data science","Go","machine learning","TPU","how to calculate mean square error","AI","neural network","ANN","feature engineering","Colab","deep learning","Google Colab","simple neural network","R","Neural Networks"],"extracted_tech_keywords":["AI","machine learning","deep learning","neural network","data science","Colab","TPU","R","Go","feature engineering"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/ann-with-linear-regression\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":7749,"title":"The Upcoming Evolution of Search","content":"The search engine is a profound component of the World Wide Web. Without it, people would be wandering the Internet taking stabs in the dark and rarely finding what they want. In recent years, though, more and more companies have been demanding search engines for help with their internal data. This means that the search engine as we know it is about to change in a couple of big ways. The State of Search Engines It goes without saying that search engines are more advanced than ever before. However, it’s nearly impossible to do justice to just how far they’ve come in a very short period of time. Not so long ago, you could be told that “no results were found for your inquiry” simply because you spelled a word wrong. Even when you spelled everything correctly, you still needed to be very specific in what you were looking for or you risked getting all kinds of answers in return. Nowadays, you can misspell every word in a sentence and your search engine will figure out what you meant or at least recommend some other spellings. You don’t even need to know exactly what you’re trying to find either. The search engine can make sense of the vaguest of inquiries. A lot of times, too, your search engine will be charged with figuring out the emotional context of what you’re saying. By picking up on these subtleties, it can bring you the results you were looking for, efficiently ranked thanks to its appreciation for nuance. However, it’s also worth pointing out that some of the best search engines out there aren’t even for public consumption. Enterprise search platforms are one of the hottest forms of software on the market. These are search engines that are used to help employees sift through their organization’s data. Let’s look a bit closer at this specific form of software because if your company isn’t currently using it, your results are going to suffer. Enterprise Search Engines vs. Web Search Engines There are a couple important differences to consider when thinking about enterprise search engines compared to the kind everyone is familiar with. Obviously, the former comes with billions of dollars in backing. Microsoft, Google and others have all put countless dollars into producing the best possible search engines. For all the money that went into them, though, they lack in a very important way. It’s this shortfall that has made it so important for companies to look elsewhere for their search solutions. While web search engines are awesome at returning a large breadth of data on demand, that information lacks for depth. The kinds of searches that companies need search engines for are much deeper in terms of its demands. These search engines have to retrieve both structured and unstructured data and then organize it in a specific way in terms of keywords, paragraphs, phrases and, of course, security. If you tried using a web search engine for going through a company’s records, it would immediately become “confused” and you would most likely be equally confused by the results it brought back. Perhaps you’ve already tried this before and know exactly what I’m talking about. Again, this is because web search engines aren’t designed for the kind of depth you need. It’s going to run a broad keyword search and bring back just about anything that comes close to what you asked for. Furthermore, though, the reason why something as powerful as Google can’t perform in a company’s internal environment is because there are no links for it to learn from. Search engines love backlinks because they tell them which websites are supported by other sites. That makes their job a lot easier. Those links won’t exist, though, for the vast majority of searches you do through your company’s internal records. The Demands of Big Data Now that you understand what enterprise search platforms are, let’s take a look toward the future. By better understanding where the market is going, you can better consider where your own needs are headed. The first and perhaps the most important factor to know about is big data. No organization is moving forward with less data. In fact, we’re making more of it than ever before. To truly do the sheer size of big data justice, consider this: it’s estimated that by 2020, we will have more digital data on our hands than there are grains of sand on the entire planet. That’s just five years from now! Imagine where that number will be in 10 or more! While your company most likely won’t have anything close to this amount, it’s worth giving some thought to where your data needs may be next year, the year after that and so on. The Need for Analytics In the search engine industry, the various digital locations where data can be stored are known as silos. Given how much data we’re creating, it only makes sense that a proportional amount of silos would be on the way too. Silos aren’t just different rooms for data storage though; they also refer to the types of data being stored. This is why search engines are going to need better and better analytics for pulling off their job. The data in one silo may have absolutely nothing in common with the kind in another. On an enterprise level, users want to be able to extract data from any of these silos without having to go through one at a time or otherwise limiting the environment they work in. Cloud Infrastructures Everything is adopting the clouds these days. Even some servers come with cloud infrastructures now. However, the same can be said for company search engines. Many manufacturers are making their search features available via the cloud. The cloud makes everything more convenient, of course, but it will also support superior scalability and a reduction in costs too. Mike Miranda writes about enterprise software and covers products offered by software companies like www.rocketsoftware.com about topics such as Terminal Emulation, Legacy Modernization, Enterprise Search, Big Data and Enterprise Mobility.","excerpt":"The search engine is a profound component of the World Wide Web. Without it, people would be wandering the Internet taking stabs in the dark and rarely finding what they want. In recent years, though, more and more companies have been demanding search engines for help with their internal data. This means that the search […]","categories":["IT Services"],"tags":[],"author_name":"AIM Media House","publish_date":"2015-08-05T07:24:24","publication_year":"2015","word_count":1013,"keywords":["big data","Go","programming_languages:R","AI","Scala","Git","RAG","analytics","GAN","R"],"extracted_tech_keywords":["AI","analytics","RAG","R","Go","Scala","Git","big data","GAN","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/it-services\/the-upcoming-evolution-of-search\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":41391,"title":"How To Transition To A Data Science Role After Being Laid Off As An Engineer","content":"One of the most popular choices to hire in the place of engineering role is that of an ML engineer or a data scientist. If you are a software engineer who has been downsized, the best option is to upskill and get into being a data scientist, engineer or machine learning developer. Identifying What The Job Needs Self-assessment: Before making the switch, it is important to identify the strengths and weaknesses. Different positions, such as data analysts, data scientists, data engineers and more, all have different roles and responsibilities. One should identify their core competencies and pick a job that is suitable. For example, data scientists and analysts require an in-depth knowledge of statistics, math and visualisation. In addition to this, they also require good communication skills. Data analysts, in particular, are also required to have business knowledge for actionable insights. KRA Analysis: Data engineers, on the other hand, are required to fulfil a different set of business roles. This generally requires skills such as programming, database management and administration, data storage and system implementation. Tool Kit: While the hard skill requirements, such as knowledge of Python\/R and SQL are similar among all positions, the soft skill requirements and responsibilities are different. This makes it the first step towards switching and identifying the best position for you. Discovering The Required Skill Set The next step is seeing whether one has the required skill set to be a good fit for the job. Any position in the data science field is highly technical, and an integral understanding and interest must be present to ensure that job satisfaction is lost. Right Job For The Right Interest: A love for data is required for any position, especially with respect to finding patterns in data and knowing how to wrangle it. Data cleaning, pre-processing and handling are some of the most important and time-consuming tasks in data science, which is to be kept in mind before deciding to move into the field. A strong base in programming is required, as coding is required at almost every step in the way. From managing cloud operations to creating a pipeline to deploying a model, coding skills are required in every step. Moreover, industry best practices and code preservation for uptime are also an important skill to have. Responsibilities and Roles: The main responsibility of a data engineer is to ensure that data flows smoothly in the workflow or pipeline. This is to be done in such a way as to ensure teammates have access to the data at a low cost, requiring data engineers to have a big picture view of operations. Data scientists, on the other hand, mainly focus on deriving insights from the data using models, algorithms and other methods. This is to be done keeping in mind the operations of the company itself and has to be visualised and presented in a consumable way for non-technical participants. In case of switching, these requirements must be considered to ensure a good job fit and reliability. Learning The Technology And Architecture If you are a new entrant into the data science sphere, the first thing to do is to look at the tools used by those already in the space. Owing to the vast variety of solutions and competing products in the market, data scientists are spoilt for choice. Industry Knowledge To Pick The Best Tools: This also means that it is important to know the most useful ones, so as to pick the right tool for the right job. Every data workflow and pipeline is different, and a data engineer is required to know the best tools to wrangle data from different kinds of datasets. This includes big data frameworks such as Hadoop, Spark, Kafka, Hive and model frameworks such as TensorFlow, PyTorch and more. In addition to this, in-depth knowledge of SQL is required, as data engineers find most of their responsibilities falling into the database administrator domain. Soft Skills And Mathematics: Data scientists, on the other hand, are required to keep up to date with the rapidly evolving landscape of AI technology. This is due to the fact that researchers create better ways to solve existing problems using cutting-edge models and algorithms. In addition, there is also a lot of statistical and mathematical knowledge required to function as data scientists. Moreover, languages such as Python, SQL and R are integral to the data scientists’ workflows. In Conclusion Before making the switch to another role or field, especially data science, the above factors must be considered. The skill sets of software engineers, such as programming and analytics knowledge, transfers well to the data science field, making it a good fit for engineers looking to switch.","excerpt":"One of the most popular choices to hire in the place of engineering role is that of an ML engineer or a data scientist. If you are a software engineer who has been downsized, the best option is to upskill and get into being a data scientist, engineer or machine learning developer. Identifying What The […]","categories":["AI Features"],"tags":["AI engineers","analytics skills","Data Science Career","Data Scientist"],"author_name":"Anirudh VK","publish_date":"2019-06-27T04:59:45","publication_year":"2019","word_count":782,"keywords":["data science","machine learning","AI","TensorFlow","AI engineers","ML","PyTorch","Data Scientist","Data Science Career","RAG","Python","analytics","Kafka","analytics skills"],"extracted_tech_keywords":["AI","machine learning","ML","data science","analytics","TensorFlow","PyTorch","RAG","Kafka","Python"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-to-transition-to-a-data-science-role-after-being-laid-off-as-an-engineer\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10072184,"title":"The Data science journey of Amit Kumar, senior enterprise architect-deep learning at NVIDIA","content":"Meet Amit Kumar, a senior enterprise architect (deep learning) at NVIDIA. Kumar holds a B.Tech in electronics and communication engineering from the prestigious IIT Guwahati. Having worked at some of the biggest companies like HP and VMWare, he has a rich take on all things tech. In an exclusive interaction with Analytics India Magazine, he spoke about his journey in data science, AI and deep learning, while taking us over the challenges, achievements and emerging trends of this domain. AIM: What attracted you to data science, given that you were into software engineering previously? Amit: I was into software engineering, precisely in computer vision and image processing domain (C++). In 2012, AlexNet competed in the ImageNet Large Scale Visual Recognition Challenge and its success was followed up by dramatic and rapid advancement in CNN architectures. This drew me towards deep learning from feature engineering-based classical statistical learning. My prior background in image processing and information theory, and linear algebra helped me get a faster grip. Similarly, the advent of Word2vec and its reasonable success in capturing semantic similarity drew me towards natural language processing. Transition to the data science field happened first through classical machine learning, followed by CNNs (computer vision through deep learning), NLP, Speech recognition(ASR), and finally reinforcement learning. AIM: What does your role at NVIDIA entail? Amit: At NVIDIA, I work as a senior enterprise solutions architect – deep learning, statistical learning. My primary responsibilities lie in helping and advising enterprises build end-to-end data science-based solutions, starting from the R&D phase (data processing, model training) to deployment on NVIDIA AI full stack platforms. The gamut of enterprises constitutes various verticals like healthcare, surveillance, defense, intelligent video analytics (IVA), smart cities, digital twins, AR\/VR +AI\/ML, industrial visual inspection, smart manufacturing, smart retail, supply chain logistics, and robotics. Since NVIDIA’s AI platforms are fueled by NVIDIA’s end-to-end data science production grade and free to use AI SDKs and free-to-use models, the total time taken by enterprises to develop and deploy AI solutions gets drastically reduced and enterprises realise a significant gain when it comes to ROIs. For my current role, I often visit Analytics India Magazine to gain insights into new developments in AI, data science, and how AI is shaping businesses, societies, and policymaking at large. AIM: How important is it for aspirants to start early or develop their portfolio before venturing into data science and AI? Amit: The most important thing for aspirants is to get the fundamentals right before diving into data science and AI. Having a basic but intuitive understanding of linear algebra, calculus, and information theory helps to get a faster grip. Aspiring data scientists should not ignore fundamental principles of software engineering, in general, because nowadays the market is looking for full-stack data scientists with the capability to build an end-to-end pipeline, rather than just being a data science algorithm expert. AIM: What were some of the biggest challenges in your career and how did you overcome them? Also tell us about your professional achievements? Amit: My biggest challenge, which ultimately turned into my biggest achievement, was to start from scratch and build a world-class center of excellence in data science at HP India along with Niranjan Damera Venkata (distinguished technologist\/strategist, AI, and machine learning transformation at HP), Madhusoodhana Rao (director at HP) and Shameed Sait (expert architect -AI\/ML). This challenge was turned into an achievement by going into the start-up mode within HP. Though we were part of a large organisation, we made sure that the center of excellence operates the way a successful startup works by inculcating the culture of mutual respect and healthy competition, attracting and hiring best talents, and providing freedom and flexibility. AIM:  Everybody wants to be a data scientist. What’s your advice to the youngsters starting out? Amit: Here is what I think: Get your basics right. If this is done, you are halfway through. Do not be a mere ML library user; understand the algorithm behind it. This will give an intuitive understanding of problems and be of immense help in devising a solution.Do NOT ignore the software engineering aspect of it. AIM: How do you see the data science and AI space evolving over years?Amit: Data science and AI space, powered by enormous leaps in compute capabilities of GPUs, is only going to flourish in the coming years. It has already seen its wider addition in various segments such as healthcare, smart city, retail, governance, defense, education, auto-mobile, digital twins, omniverse, AI-powered by simulations, robotics, Industry 4.0, etc.","excerpt":"The most important thing for aspirants is to get the fundamentals right before diving into data science and AI, feels Amit","categories":["AI Features"],"tags":["Interviews and Discussions"],"author_name":"Sreejani Bhattacharyya","publish_date":"2022-08-04T16:00:00","publication_year":"2022","word_count":753,"keywords":["data science","machine learning","AI","ML","computer vision","NLP","Aim","deep learning","analytics","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","NLP","computer vision","data science","analytics","Aim","R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/the-data-science-journey-of-amit-kumar-senior-enterprise-architect-deep-learning-at-nvidia\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":6,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":34929,"title":"Meet ToMnet, An AI That Can Read Your Computer’s ‘Mind’","content":"Image source: maniac-review-netflix As technology has shifted focus from being more human-centric to machine-centric, understanding machines and predicting how they interact with each other is has gained importance. A research paper presented at Stockholm’s machine learning conference has now scratched a bit of the layer of it, by making artificial intelligence understand the “minds” of other AI-powered machines. Origin Of The Idea According to psychologists, AI-powered assistants like Siri and Alexa lack the “awareness” of others’ beliefs and desires. Stockholm-based computer system researchers have produced an AI that can probe the minds of other computers and forecast their actions, the primary step to fluid cooperation amongst makers and in between makers and individuals. “Theory of mind is clearly a crucial ability for browsing a world complete of other minds,” states Alison Gopnik, a developmental psychologist at the University of California, Berkeley. By about the age of 4, human kids comprehend that the beliefs of another individual might diverge from reality, which those beliefs can be utilised to forecast the individual’s future habits. Some of today’s computers with help of AI can identify facial expressions such as happy or angry, an ability connected with the theory of mind. However, they all have little understanding of human feelings or exactly what encourages us. The brand-new task started as an effort to get people to comprehend computers. Many algorithms utilised by AI aren’t completely composed by developers. However, they rather count on the machine learning as it sequentially takes on issues. The resulting computer-generated options are frequently black boxes, with algorithms too complicated for human insight to permeate. Neil Rabinowitz, a research study researcher at DeepMind in London along with his associates produced a theory of mind AI called ToMnet and had it observe other AIs see exactly what it might learn more about how they function. How Does ToMnet Works? ToMnet makes up 3 neural networks, each made of little computing components and connections that gain from experience, loosely looking like the human brain. The very first network finds out the propensities of other AIs based upon their previous actions. The 2nd kinds an understanding of their present “beliefs.” And the 3rd takes the output from the other 2 networks and, depending upon the scenario, anticipates the AI’s next relocations. AI under research study were easy characters moving a virtual space gathering coloured boxes for points. ToMnet saw the space from above. In one test, there were 3 “species” of character: One could not see the surrounding space, one could not remember its current actions, and one might both see and keep in mind. The blind characters tended to follow along walls, the amnesiacs moved to whatever things was closest, and the 3rd types formed sub goals, tactically getting things in an order to make more points. After some training, ToMnet might not just determine a character’s types after simply a couple of actions, however, it might likewise properly forecast its future habits, scientists reported this month at the International Conference on Machine Learning in Stockholm. The last test exposed ToMnet might even comprehend when a character held an incorrect belief, a vital phase in establishing the theory of mind in people and other animals. In this test, one type of character was configured to be near-sighted, when the computer system modified the landscape beyond its vision midway through the video game, ToMnet properly anticipated that it would stick to its initial course more often than better-sighted characters, who were most likely to adjust. What Do Critics Say? Developmental Psychologist, Gopnik is of the opinion that research study, as well as the report that was presented at the conference which recommended the AIs capability to forecast other AI’s habits is based upon exactly what they learn about themselves, are classic examples of neural networks striking capability to find out abilities by themselves. But, Gopnik also states that it still does not put them on the exact same level as human kids, who would likely pass this false-belief job with near-perfect precision, even if they had never ever experienced it previously. Josh Tenenbaum, a psychologist as well as a computer system researcher at the MIT, Cambridge, has likewise dealt with computational designs of the theory of mind capabilities. He states ToMnet presumes beliefs more effectively than his group’s system, which is based upon a more abstract kind of probabilistic thinking instead of neural networks. But ToMnet’s understanding is more securely bound to the contexts where it’s trained, he includes, making it less able to forecast habits in significantly brand-new environments, as his system and even children can do. Outlook In the future, integrating techniques may take this field of research in interesting directions. The kind of social proficiency computers are establishing will enhance not just the cooperation with people, however, likewise will also have a probability to create deceptiveness too. By chance, If a computer system anyhow comprehends incorrect beliefs, it might understand how to cause them in individuals too.","excerpt":"As technology has shifted focus from being more human-centric to machine-centric, understanding machines and predicting how they interact with each other is has gained importance. A research paper presented at Stockholm’s machine learning conference has now scratched a bit of the layer of it, by making artificial intelligence understand the “minds” of other AI-powered machines. […]","categories":["AI Features"],"tags":["analog neural networks","DeepMind","deepmind london"],"author_name":"Martin F.R.","publish_date":"2019-02-14T04:36:57","publication_year":"2019","word_count":827,"keywords":["Go","artificial intelligence","machine learning","TPU","AI","neural network","programming_languages:R","analog neural networks","RAG","ViT","R","deepmind london","DeepMind"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","neural network","RAG","TPU","R","Go","ViT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/meet-tomnet-an-ai-that-can-read-your-computers-mind\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":5896,"title":"Jigsaw Academy Announces their First Data Analytics Competition Results held for Jigsaw Students","content":"Bangalore, India June 30th , 2014 Jigsaw Academy, the online school of analytics and Mortgage Dex, a firm that examines real estate trends, policies & financial implications that impact real estate investment announced the winners of the Data Analysis Competition, held for Jigsaw students. The contest was aimed at helping students strengthen their skills in data exploration, analysis and visualization. The objective of the contest was to recommend a good place\/zip code to buy property based on the data at hand and was aimed at helping students strengthen their skills in data exploration, analysis and visualization. The students were initially asked to share an approach note for a sample data set. Based on this, fourteen students were invited to proceed further and submit within a ten day timeframe, a more detailed analysis of the real estate data shared with them. The submissions were then evaluated by Jigsaw mentors and Mortgage Dex and the winners were chosen based on the depth of analysis, tools used, domain knowledge and presentation. On receiving the submissions, Dinakar K , senior data analyst & engineer of Mortgage Dex had this to say, “We see that a lot of hard work has been put in by the students. They seem to have used the right techniques to approach the problem and they have also used some amazing visualizations to explain their insights.” The winners of this project will each received a scholarship of Rs 15,000, Rs 10,000 and Rs 5,000 respectively. The top performers in the competition will also get a chance to be interviewed for analyst roles in Mortgage Dex. The first finalist was Shyam Anadkat, second finalist was Parindsheel Dhillon and the third finalist was Balaji Athreya. Also special mention to Vaishnavi Reddy. “Collaborating with Mortgage Dex for this competition has been a rewarding experience. It was a great opportunity for our students to sharpen their data analysis skills and we thank Mortgage Dex and all the participants and commend them for their excellent effort”, says Gaurav Vohra, CEO of Jigsaw Academy. About Jigsaw: Jigsaw Academy is a Bangalore based analytics training company that is run by analytics professionals. Their courses are designed and delivered by industry experts who have applied analytics to solve business problems in a variety of fields like retail, FMCG, financial services, telecom and health care. The instructors use their real-world experiences to teach analytical skills that are most valuable at the work place. Their flagship course called the Foundation course in analytics has been taken by thousands of students across the globe and has helped launch many careers in this new and exciting field. They have recently completed analytics training workshops in B schools across the country, including IIM Bangalore. Website: http:\/\/www.jigsawacademy.com Blog: http:\/\/www.analyticstraining.com Mail:info(at)jigsawacademy(dot)com Phone +91-9243522277","excerpt":"Bangalore, India June 30th , 2014 Jigsaw Academy, the online school of analytics and Mortgage Dex, a firm that examines real estate trends, policies & financial implications that impact real estate investment announced the winners of the Data Analysis Competition, held for Jigsaw students. The contest was aimed at helping students strengthen their skills in […]","categories":["AI News"],"tags":[],"author_name":"Sarita Digumarti","publish_date":"2014-07-01T17:11:19","publication_year":"2014","word_count":458,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","Aim","ViT","analytics","R"],"extracted_tech_keywords":["AI","analytics","Aim","R","Go","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/jigsaw-academy-announces-their-first-data-analytics-competition-results-held-for-jigsaw-students\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":28069,"title":"AI-For-Fashion Startup Mad Street Den Raises Funding From Japan&#8217;s KDDI","content":"Mad Street Den, a startup that focuses on computer vision and artificial intelligence, has raised an undisclosed amount from Japanese telecom operator KDDI through KDDI Open Innovation Fund 2. The Indian-US startup, which was founded in Chennai, was founded in 2013 by a neuroscientist-designer duo, Dr Anand Chandrasekhar and Aswini Asokan. Mad Street Den now plans to use the funds to improve their AI capabilities and grow their sales and marketing workforce. Mad Street Den’s key product Vue.AI is a one-stop AI and image recognition platform for apparel retailers — from big department stores to online marketplaces — who are experiencing drastic operational changes due to rapidly increasing penetration of online commerce. KDDI Japan announces it's investment in MadStreetDen enabling growth of @vue_ai across Asia. These are exciting times for both Computer Vision & Retail, and we're thrilled to GTM in new markets with 2 of our newest investors Global Brain-KDDI & Rocketship VC (SF) — Mad Street Den (@MadStreetDen) September 5, 2018 “At KDDI, we are a company that continues to provide excitement as we move forward with the integration of telecommunications and life design. By investing in venture companies that can deploy services on a global scale, we help generate new business to create new customer experience value,” said the Japanese tech giant in an official statement. Vue.AI offers products like data classification, catalogue data management. It also includes AI-based automated data-tagging with attributes, titles, descriptions, which drastically improve product discovery and searchability, reducing operational costs and the time required to digitise a product and go to market. The solution also makes it easier to eliminate redundancy in the catalogue and provide a singular view of the entire inventory. Its marquee product is a cutting-edge human model generator which is an AI-based product generating garment-to-model images, predicting how the garment fits. Vue.ai’s algorithms also analyse the extracted catalogue data with user behaviour and help the retailers’ marketing\/product\/cataloguing teams not only to automate their processes and reduce costs but also to drive conversions and to increase revenues. With the solution, fashion brands can provide their customers with a great user experience in online shopping, such as good retrieval accuracy or “AI stylist” which is an AI-based styling and outfitting solution.","excerpt":"Mad Street Den, a startup that focuses on computer vision and artificial intelligence, has raised an undisclosed amount from Japanese telecom operator KDDI through KDDI Open Innovation Fund 2. The Indian-US startup, which was founded in Chennai, was founded in 2013 by a neuroscientist-designer duo, Dr Anand Chandrasekhar and Aswini Asokan. Mad Street Den now plans to use […]","categories":["AI News"],"tags":["AI (Artificial Intelligence)","fashion","japan"],"author_name":"Prajakta Hebbar","publish_date":"2018-09-06T07:51:15","publication_year":"2018","word_count":371,"keywords":["japan","Go","API","artificial intelligence","AI","innovation","image recognition","Git","computer vision","fashion","R","AI (Artificial Intelligence)","startup"],"extracted_tech_keywords":["AI","artificial intelligence","computer vision","image recognition","R","Go","Git","API","innovation","startup"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/mad-street-den-raises-funding-from-japans-kddi\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":35281,"title":"The Demand-Supply Gap Between Computer Science Students And Faculty Needs To Be Bridged Now","content":"Computer science is one of the most popular majors among students today, not just in India, but in the world. This is because expertise in computer science opens up plenty of opportunities in terms of jobs, and also the fact that there’s a lot of research is going on in this area. With an increasing demand for students taking up graduate and postgraduate courses in computer science, the faculty for computer science to train them in institutes is unable to grow with the demand. The Gap Between The Students And The Faculty In a new survey commissioned by Microsoft and conducted by YouGov, 88 percent of teacher respondents said they believe computer science is critical for students’ success in the workplace—but two in 10 teachers said their students aren’t taught any computer science. Computer science students on a number of campuses complain that their departments can’t meet demand. School districts and boards of education have not prioritized computer science education the way they should. It is not true that every student will not grow up to work as a computer engineer or a computer scientist, but a good exposure on the working of computers must be given to students at the preliminary level. The government is not doing enough to help schools teach computer science. Although recently the Central Board of Secondary Education (CBSE) had announced that it wants to include AI in its curriculum, in an attempt to mould their education system to make students well-versed with today’s fast-growing and highly demanding technologies, has decided to include artificial intelligence in their syllabus for students. More such adoptions are required. In the same survey by Microsoft and YouGov, 80 percent of respondents were of the opinion that big tech companies like Microsoft, Apple and Google should step in and help kids learn computer science skills. Computers should be part of the education curriculum the way math and language are, rather than being dependent on funding and involvement from tech companies in order for it to be taught in schools. Demand For CS Graduates Continues To Escalate The rise in computer science students is an obvious consequence of the following facts: 1. Greater job opportunities: There are a number of job opportunities in the field of computer science, more compared to other fields of engineering or science. Students majoring in subjects other than computer science or IT also have to often end up with the job of a software engineer. Many companies in India are open to plenty of opportunities for software engineers. 2. Demand for skills: Every organisation has a need to hire computer science expert, if not a graduate. It becomes easy for them to hire students from computer science, especially when it comes to freshers, which is why industries prefer hiring students possessing software skills. 3. Good salary: Many US companies are setting up their workforces in India because of the reasons for the time lag between the two places fitting perfectly and the talent that exists here. Salary is ultimately what matters to most at the end of the day. The US IT market being a very successful one, their employees in India, too, get a good salary. It is a promising field with employment outlook ranging from average to much faster than average. 4. The rise of new technology: With an increasing rise of technologies like artificial intelligence and machine learning, it has become imperative to get upskill in these technologies. It is very easy for the computer science major students to jump into this new wave of technology. A survey says that 90 percent of people in AI and data science and ML has moved their careers to from computer science. The skills that computer science students learn match the best and hence is easier for them to make a switch. What Can Be Done? The United Kingdom has made efforts to improve curriculum and train 8,000 teachers in computer science. An organised training for the professors would allow for overall progress and help the teachers and the students as well. The fast-paced world of technology is constantly evolving and it is important that our computer science teachers are trained in the very latest digital skills. A proper training will give the teachers the knowledge and support needed to guide the students.","excerpt":"Computer science is one of the most popular majors among students today, not just in India, but in the world. This is because expertise in computer science opens up plenty of opportunities in terms of jobs, and also the fact that there’s a lot of research is going on in this area. With an increasing […]","categories":["AI Features"],"tags":["AI (Artificial Intelligence)","Career","Engineering","Machine Learning","Software","teachers","training"],"author_name":"Disha Misal","publish_date":"2019-02-22T11:53:44","publication_year":"2019","word_count":717,"keywords":["data science","teachers","Go","training","artificial intelligence","machine learning","AI","ML","Engineering","Machine Learning","Scala","RAG","Git","Software","R","Career","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","data science","RAG","R","Go","Scala","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/the-demand-supply-gap-between-computer-science-students-and-faculty-needs-to-be-bridged-now\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":5308,"title":"Flytxt ties up with Idea Cellular in India","content":"February, 19th, 2014, India – Flytxt, a Big Data Analytics powered solution provider for Communication Service Providers (CSPs), announced its partnership with Idea Cellular, the third largest mobile operator in India. As part of this deal, Flytxt will deploy its multi-channel mobile advertising platform, QREDA, and connect to its mobile advertising market place, mADmart, that would facilitate brands and agencies to engage with over 130 million Idea subscribers directly, delivering relevant ads. Flytxt’s mobile advertising platform will enable Idea to leverage its expertise in Data Analytics, enabling brands to do the most relevant contextual advertising on mobile. It will also enhance customer satisfaction and ad campaign effectiveness. Mr. Srinivasa Ravi, Head-mADmart, stated, “We are excited to partner with Idea to deliver this unique value proposition to brands and advertisers in India. Going forward we expect the value of mobile channels in India to go up as smartphone penetration and adoption of new generation services increases  and this would bring in much higher ad spent on mobile media”. mADmart is a unique CSP anchored mobile ad marketplace launched by Flytxt that follows two sided business model. It provides an opportunity for CSPs to leverage their in-depth understanding of subscribers, thus enabling brands and advertisers to deliver highly relevant offers. mADmart, on the other side, would increase ROI for advertisers providing them with better reach and meaningful engagement than any existing media today. “mADmart will serve as a self-service advertising platform for brands to manage end-to-end multi-channel, cross-channel mobile ad campaigns. It is a totally inclusive business model protecting the interests of consumers, CSPs and brands”, said Mr. Abhay Doshi, Head-Product and Marketing, Flytxt. mADmart, the mobile advertising marketplace of Flytxt will ensure better experience for Idea’s subscribers as they will receive highly relevant offers through preferred mobile media.  The ad marketplace will also enable brands and advertisers to engage with both rural and urban population, offering them with the capability to serve consumers from both markets distinctly. About Flytxt Flytxt is a leading provider of Big Data Analytics powered solutions with a focus on enabling CSPs to derive measurable economic value from subscriber data. The company offers Customer Experience Revenue Management and Insight Monetization Solutions as well as consultancy services to help CSPs run campaigns for increasing revenue, reducing churn and enhancing loyalty. Flytxt’s closed loop integrated real-time marketing platform has been deployed by CSPs across APAC and EMEA, serving more than 500 million subscribers and has generated around $350 million incremental revenue for them till now. Flytxt has won many industry awards like BID International Quality Summit Award, Aegis Graham Bell Award, Red Herring Asia 100, IEEE Cloud Computing Challenge and NASSCOM Emerge 50 League of 10. Flytxt has its Headquarters in Netherlands, Corporate office out of Dubai and global delivery centres at Trivandrum and Mumbai in India. The Company has presence in Delhi, Dhaka, Nairobi, Lagos, Kuala Lumpur and London. For more information about Flytxt, visit www.flytxt.com. Follow us | Connect on LinkedIn | Like us on Facebook","excerpt":"February, 19th, 2014, India – Flytxt, a Big Data Analytics powered solution provider for Communication Service Providers (CSPs), announced its partnership with Idea Cellular, the third largest mobile operator in India. As part of this deal, Flytxt will deploy its multi-channel mobile advertising platform, QREDA, and connect to its mobile advertising market place, mADmart, that […]","categories":["AI News"],"tags":[],"author_name":"AIM Media House","publish_date":"2014-02-22T14:36:02","publication_year":"2014","word_count":499,"keywords":["big data","Go","programming_languages:R","cloud computing","AI","programming_languages:Go","RAG","analytics","R"],"extracted_tech_keywords":["AI","analytics","RAG","cloud computing","R","Go","big data","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/flytxt-ties-up-with-idea-cellular-in-india\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10109822,"title":"Why is GitHub Bullish About AI in Cybersecurity?","content":"GitHub foresees the pivotal role of AI in the software development lifecycle, including security. Over the past year, the company has incorporated over 70 features into GitHub advanced security. However, at GitHub Universe 2023 held in November, it announced adding generative AI to the mix. The company now believes security vulnerabilities can be identified at the very stage when the code is being written. By leveraging an LLM, GitHub now not only identifies potential vulnerabilities but also provides developers with secure code suggestions from the start. “With auto fix, we’re going to suggest the fix in the pull request for them. So the developers will not just see the alert, but also a suggested fix powered by AI right there,” Jacob Depriest, VP and deputy chief security officer at GitHub, told AIM. These are not ordinary fixes. They are concise, actionable suggestions for swiftly comprehending and addressing vulnerabilities. Developers can now resolve issues faster and prevent new vulnerabilities from creeping into your codebases. Depriest said GitHub has already seen great success with code scanning’s current fix rate. “This implies that when developers receive an alert while working, they address the issue approximately 50% of the time before it reaches production and this is huge. With the new AI-powered code scanning auto fix feature, developers can build on an already strong fix rate.” Protecting secrets with AI GitHub is not just using LLMs solely to uncover potential code vulnerabilities; the company also uses these powerful models to detect leaked passwords with reduced false positives. Nearly 80 percent of the breaches originate through credential leakage or secrets being leaked, according to Depriest. “Secret scanning has been integral to GitHub’s advanced security, forming a key component of our security programme. “Now with AI, we’re also going to detect generic secrets and low confidence patterns in code as well, which is going to really improve that capability and capture more and protect secrets before they’re even hitting production.” Moreover, Depriest believes security starts with the developer and to be more precise, with the developer’s account. Given the high number of credential leakages, it opted to lean heavily into enabling multi-factor authentication for all contributors on github.com. “That wasn’t an easy thing to do. That wasn’t a quick thing to do. It took a lot of planning and a lot of investment to make that work. But we really believe that’s the right thing to do,” he said. And now, the introduction of the new secret scanning feature allows GitHub to detect generic or unstructured secrets in the code. Protecting against AI Vulnerabilities Despite GitHub’s optimistic stance on leveraging AI in cybersecurity, the era of generative AI has presented various instances where it emerges as a substantial cybersecurity threat. For instance, prompt injection attacks remain a significant challenge to address for cybersecurity teams. Over time, we have seen LLMs are vulnerable to prompt injection attacks. Given GitHub’s close alliance with Microsoft, it might be leveraging OpenAI’s GPT models, most notably GPT-4, the most advanced LLM to date. However, GPT-4 has also been found to be vulnerable to prompt injection attacks. Depriest believes responsible integration and security measures within the tooling are crucial in safeguarding against such manipulations. This approach is fundamental to ensuring protection in scenarios involving prompt injection and similar vulnerabilities. Moreover, according to Depriest, safeguarding both the infrastructure and the overall network workspace remains a key aspect of cybersecurity, even in the AI age. “We approach this responsibility with the same diligence as we do for the rest of our infrastructure, including protecting github.com. This entails threat detection, security operations, and ensuring code security, and this extends to AI models. We uphold uniform controls and compliance principles across all facets of our core responsibilities.” Does AI fundamentally change cybersecurity? While GitHub is banking on generative AI capabilities in a big way, including for cybersecurity, Depriest does not believe it will fundamentally change the cybersecurity landscape. “The reality is every new technology is dual-use. This pattern has been consistent across various technologies throughout the past two decades. I still don’t think fundamentally it will change how we approach the security of what we need to do, what our job is, and how we will keep the platform safe.” He holds the belief that the advantages of generating secure code from the outset and consistently maintaining its security far exceed the potential risks associated with certain threats in the landscape. “We strongly believe that our efforts at GitHub, incorporating AI into the developer workflow, will yield a substantial and highly valuable impact, far outweighing any potential risks in the threat landscape. While scaling remains an industry challenge, we envision this as the future for enhancing security in the realm of developers.”","excerpt":"GitHub is not using LLMs solely for uncovering potential code vulnerabilities, the company also uses these powerful models to detect leaked passwords with reduced false positives.","categories":["Deep Tech"],"tags":["GitHub"],"author_name":"Pritam Bordoloi","publish_date":"2024-01-02T15:38:05","publication_year":"2024","word_count":786,"keywords":["Go","OpenAI","AI","Git","RAG","GPT","Aim","generative AI","GitHub","R"],"extracted_tech_keywords":["AI","generative AI","OpenAI","Aim","RAG","R","Go","Git","GitHub","GPT"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/why-is-github-bullish-about-ai-in-cybersecurity\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10062420,"title":"Indian IT sector’s attrition rate on the rise, no sign of respite","content":"Attrition in the Indian IT sector is at an all-time high. For the quarter ending December 2021, TCS saw an attrition rate of 15.30%, while Infosys and Wipro’s numbers stood at 25.50% and 22.70 %, respectively. A far cry from last year’s numbers. Source: Company reports The rapid digitisation in the wake of the pandemic has driven the demand for IT talents. From early career IT professionals to mid and senior-level executives, the IT tech space is brimming with opportunities right now. Additionally, the companies are offering huge hikes to retain talents. Many professionals are moving away from traditional IT companies to startups with faster growth opportunities. Image: LinkedIn “Engagement with the company has seen the most impact in the past few months. Organisations have been in experimental mode globally to try and reduce the gap creating a lot more overhead across work levels and not being able to truly engage with their ecosystem. There’s a lot of commercial excitement today in the market. This can be seen not just for IT but also for business roles,” said Dipesh Lakhotia, head of analytics, Britannia Industries Limited. “The talent drain and the disruption over the last two years has led leaders to start thinking and planning their organisation as a more autonomous entity which is able to work seamlessly with minimal dependence on individuals,” he added. Problem of plenty As per PwC’s Global Artificial Intelligence Study: Exploiting the AI Revolution, AI could contribute up to USD 15.7 trillion to the global economy in 2030. Of this, USD 6.6 trillion is likely to come from increased productivity and USD 9.1 trillion is likely to come from consumption-side effects. While some markets, sectors and individual businesses are more advanced than others, AI is still at a very early stage of development overall, the report said. AI is primed for growth, and will open up a lot of opportunities for India’s IT talent pool. The demand will outstrip supply, and experienced candidates with good domain knowledge will have a lot of options to choose from. According to Nasscom president Debjani Ghosh, the shortage of digital talent in India is leading to high attrition rates and increased wages. Image: World Economic Forum The shift to remote work has definitely changed the mindset. “Job security as a concept has also lost its meaning over time. Working from home has allowed individuals to perform independently and boosted their self-confidence. This has led to people seeking opportunities to push their envelope  and experiment with their professional journey,” said Lakhotia. Image: LinkedIn Global Talent Trends 2022 Companies are taking a multipronged approach such as stepping up fresh hiring to increase the supply pool, accelerating re-skilling programmes through online learning, deploying adjacent-talent skills for on-the-job learning, and offering employees a holistic experience etc to reduce the attrition rates, Debjani Ghosh has said. To reduce attrition rates, the companies should be invested in the employee’s career journey. Lakhotia pointed out a few practices that can be implemented to retain talent. Building a good work culture is a must for companies. Develop empathy towards your team’s struggle and actively participate in their journey. Challenging employees intellectually and keeping track of their personal growth is crucial. Growth mindset is not just about generating revenue, but growing together as a team. Build that culture, sustain it and perfect it over time.","excerpt":"Attrition in the Indian IT sector is at an all-time high. For the quarter ending December 2021, TCS saw an attrition rate of 15.30%, while Infosys and Wipro’s numbers stood at 25.50% and 22.70 %, respectively. A far cry from last year’s numbers. Source: Company reports The rapid digitisation in the wake of the pandemic […]","categories":["IT Services"],"tags":[],"author_name":"Sreejani Bhattacharyya","publish_date":"2022-03-09T13:00:00","publication_year":"2022","word_count":557,"keywords":["Go","API","artificial intelligence","AI","ML","Git","ViT","analytics","GAN","R"],"extracted_tech_keywords":["AI","artificial intelligence","ML","analytics","R","Go","Git","API","GAN","ViT"],"url":"https:\/\/analyticsindiamag.com\/it-services\/indian-it-sectors-attrition-rate-on-the-rise-no-sign-of-respite\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10015900,"title":"How A “Crazy Idea” Changed The Way We Do Machine Learning: NeurIPS Test Of Time Award Winner","content":"HOGWILD! Wild as it sounds, the paper that goes by the same name was supposed to be an art project by Christopher Re, an associate professor at Stanford AI Lab, and his peers. Little did they know that the paper would change the way we do machine learning. Ten years later, it even bagged the prestigious “Test of Time” award at the latest NeurIPS conference. To identify the most impactful paper in the past decade, the conference organisers selected a list of 12 papers published at NeurIPS over the years — NeurIPS 2009, NeurIPS 2010, NeurIPS 2011 — with the highest numbers of citations since their publication. They also collected data about the recent citations counts for each of these papers by aggregating citations that these papers received in the past two years at NeurIPS, ICML and ICLR. The organisers then asked the whole senior program committee with 64 SACs to vote on up to three of these papers to help in picking an impactful paper. About HOGWILD! Most of the machine learning is about finding the right kind of variables for converging towards reasonable predictions. Hogwild! is a method that helps in finding those variables very efficiently. “The reason it had such a crazy name, to begin with, was it was intentionally a crazy idea,” said Re in an interview for Stanford AI. With its small memory footprint, robustness against noise, and rapid learning rates, Stochastic Gradient Descent (SGD) has proved to be well suited to data-intensive machine learning tasks. However, SGD’s scalability is limited by its inherently sequential nature; it is difficult to parallelise. A decade ago, when the hardware was still playing catch up with the algorithms, the key objective for scalable data analysis, on vast data, is to minimise the overhead caused due to locking. Back then, when parallelisation of SGD was proposed, there was no way around memory locking, which deteriorated the performance. Memory locking was essential to reduce latency for between processes. Re and his colleagues demonstrated that this work aims to show using novel theoretical analysis, algorithms, and implementation that stochastic gradient descent can be implemented without any locking. In Hogwild!, the authors made the processors have equal access to shared memory and were able to update individual components of memory at will. The risk here is that a lock-free scheme can fail as processors could overwrite each other’s progress. “However, when the data access is sparse, meaning that individual SGD steps only to modify a small part of the decision variable, we show that memory overwrites are rare and that they introduce barely any error into the computation when they do occur,” explained the authors. When asked about the weird exclamation point at the end of the already weird name “I thought the phrase “going hog-wild” was hysterical to describe what we were trying. So I thought an exclamation point would just make it better,” quipped Re. In spite of being honoured with being a catalyst behind driving ML revolution, Re believes that this change would have happened with or without their paper. What really stands out, according to him, is that an “odd-ball”, goofy sounding research is recognised even after a decade. This is a testimony to an old adage — there is no such thing as a bad idea! Find the original paper here. Here are the “test of time” award winners in the past: 2017: Random Features for Large-Scale Kernel Machines by Ali Rahimi and Ben Recht 2018: The Trade-Offs of Large Scale Learning by Léon Bottou 2019: Dual Averaging Method for Regularized Stochastic Learning and Online Optimisation by Lin Xiao","excerpt":"HOGWILD! Wild as it sounds, the paper that goes by the same name was supposed to be an art project by Christopher Re, an associate professor at Stanford AI Lab, and his peers. Little did they know that the paper would change the way we do machine learning. Ten years later, it even bagged the […]","categories":["Deep Tech"],"tags":["Machine Learning"],"author_name":"Ram Sagar","publish_date":"2020-12-24T16:00:00","publication_year":"2020","word_count":602,"keywords":["Go","API","machine learning","AI","ML","Machine Learning","Scala","RAG","Aim","GAN","R"],"extracted_tech_keywords":["AI","machine learning","ML","Aim","RAG","R","Go","Scala","API","GAN"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/hogwild-machine-learning-neurips-test-of-time-winner\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10172471,"title":"Ericsson Expands R&#038;D Centre in Bengaluru with Focus on Semiconductor","content":"Ericsson is strengthening its Research and Development (R&D) presence in Bengaluru to enhance its capabilities in Application-Specific Integrated Circuit (ASIC) development. The expansion is set to add over 150 new positions to Ericsson’s R&D team in India. Bengaluru will serve as the primary location for the growth of Ericsson’s ASIC R&D unit. “Expanding our ASIC R&D in India will enable us to leverage the local technology competence in the country towards enhancing our capabilities in semiconductor design. This unit will also contribute towards strengthening the semiconductor ecosystem in the country,” Nitin Bansal, managing director of Ericsson India, said. “We are keen to strengthen our R&D team in Bengaluru, a city celebrated for its dynamic technology ecosystem and skilled professionals. By increasing our focus on ASIC development, we are reinforcing our technological prowess,” Anna Dicander, vice president of Radio & Transport Engineering at Ericsson Business Area Networks, added. Ericsson’s investment in Ericsson Silicon is central to its innovation efforts. The company’s System on a Chip (SoC) solutions are designed for mobile networks and are integrated across the entire Ericsson Radio System, including basebands, radio, and transport, enabling service providers to fully utilise 5G. Ericsson has been a significant force in shaping India’s telecom landscape for over 120 years, from the introduction of GSM services in 1994 to supporting 5G development. In 2024, the company also launched its India 6G program, forming a dedicated 6G research team at its Chennai R&D Center. Globally, Ericsson invests approximately USD 5 billion annually in R&D, further cementing its leadership in 5G and next-generation technologies. The company’s R&D facilities in India, including those in Chennai, Bengaluru, and Gurugram, work on a range of telecom innovations, including AI and cloud technologies.","excerpt":"The expansion is set to add over 150 new positions to Ericsson’s R&D team in India, underscoring the company’s focus on semiconductor innovation.","categories":["AI News"],"tags":["AI"],"author_name":"Shalini Mondal","publish_date":"2025-06-26T17:59:50","publication_year":"2025","word_count":284,"keywords":["API","programming_languages:R","AI","innovation","RAG","R"],"extracted_tech_keywords":["AI","RAG","R","API","innovation","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/ericsson-expands-rd-centre-in-bengaluru-with-focus-on-semiconductor\/","complexity_score":2,"technical_depth":6,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10090354,"title":"DLSS vs. FSR 3.0: Is AI Enough to Give NVIDIA the Edge?","content":"At GDC 2023, AMD unveiled FidelityFX Super Resolution v3, or FSR 3.0, a solution for upscaling games to save performance. FSR 3 is a direct competitor to NVIDIA’s deep learning super sampling (DLSS), which uses AI to improve performance in games. In the age of raytraced games, boosting frame rates while maintaining quality has become a priority for both AMD and NVIDIA. While the former has been lagging behind the latter due to their late start, it seems that AMD’s FSR 3 can provide a more accessible option for low-end gamers as opposed to NVIDIA’s DLSS, which only works on NVIDIA 20-series GPUs and above. Why do gamers need upscaling? The era of upscaling can be traced back to the launch of DLSS in February 2019. This technology was released along with the first generation RTX 20-series (ray tracing) cards, primarily to make up for the performance loss that comes with enabling ray tracing. Real-time ray tracing is a new graphical technology pioneered by NVIDIA which simulates the light rays bouncing off objects in real time. With RTX, games can accurately show lighting, shadows, and reflections in a way that more closely resembles reality. As one can imagine, calculating the path of individual light rays takes a large amount of computing resources. This is why NVIDIA created RT cores to accelerate ray tracing tasks. However, even with RT cores, enabling ray tracing generally resulted in the frame rate, i.e., number of images rendered per second, taking a big hit, even halving performance in some cases. This is where DLSS comes in. By reducing the resolution, i.e., the number of pixels, in the GPU’s processing pipeline and upscaling it using convolutional neural networks, DLSS was able to improve performance by up to 50% in some cases. To power the AI algorithms used by DLSS, NVIDIA also integrated Tensor cores into the RTX cards. Due to the large amount of specialised hardware required to make DLSS work, it has been restricted only to NVIDIA’s 20-series chips and above—with DLSS 3 only being available on 40-series GPUs. AMD, on the other hand, observed the potential that FSR could have for gamers playing on the low-end of the hardware spectrum. FSR was launched in June 2021, lagging behind DLSS by more than two years. Furthermore, AMD’s offering didn’t even use AI, instead relying on spatial upscaling to achieve the same outcome. With the release of FSR 3, the question remains: Can AMD beat NVIDIA even without the use of deep learning? How does FSR stack up? FSR 3.0, which is still in development, aims to bring improvements over existing FSR by introducing frame interpolation. Frame interpolation is the process of synthesising ‘in-between’ frames for two frames, increasing the overall frame rate while maintaining quality. By introducing frame interpolation, AMD is beginning to catch up to NVIDIA whose DLSS 3 uses optical flow interpolation to achieve the same purpose. The consensus among many hardware reviewers and gamers is that DLSS is generally the superior upscaling technology, in terms of quality, when compared to FSR, even though both of them offer close to the same amount of performance gains. While they both are equally efficient, NVIDIA is closed-source, increasing the barrier for entry for game studios. Moreover, it only runs on NVIDIA hardware, and that too only for GPUs of 20-series and above. FSR, on the other hand, is a set of spatial upscaling algorithms that can quickly be integrated into any game. The algorithms were also made open source shortly after launch, leading to more game companies integrating FSR than DLSS, even though most games include an option for both. The nature of the algorithms mean that it can be run on any hardware, with even NVIDIA’s 10-series cards being supported by the technology. To achieve optimal upscaling results, NVIDIA makes gratuitous use of auto-encoded convolutional neural networks. These networks were built from the ground up for upscaling tasks, and the model is fine-tuned for every game that has support for DLSS. This means that a game has to support DLSS from the development stage, with game devs sharing datasets of images from their game to train the DLSS CNN. Intel has also thrown its hat in the ring, equipping its Intel ARC GPUs with a technology they have termed Xeon Super Sampling, or XeSS. This technology also uses AI for upscaling, but is only supported by a handful of games. While AMD’s FSR is adopted by 226 titles as of December 2022 and NVIDIA’s is adopted by 210, XeSS is adopted by a measly 20. With FSR 3.0’s non-hardware-accelerated upscaling algorithms, there comes a catch due to interpolation. Frame interpolation typically adds latency—a big no-go for gamers. AMD has clarified that FSR 3.0 is still in development, meaning that there is still time for it to enhance its latency capabilities. As NVIDIA continues to build its walled garden, it seems that it is moving towards pushing consumers to buy its newer cards to take advantage of DLSS’ latest advancements. AMD, on the other hand, is building FSR to be used with as many games and as many GPUs as possible, with the only gap left to be closed being that of visual fidelity.","excerpt":"AMD and NVIDIA go head to head on image upscaling technologies. But who will win?","categories":["Global Tech"],"tags":[],"author_name":"Anirudh VK","publish_date":"2023-03-29T18:15:39","publication_year":"2023","word_count":869,"keywords":["Go","programming_languages:R","AI","neural network","programming_languages:Go","Ray","Aim","deep learning","CNN","R"],"extracted_tech_keywords":["AI","deep learning","neural network","Aim","Ray","R","Go","CNN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/dlss-vs-fsr-3-0-is-ai-enough-to-give-nvidia-the-edge\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10004259,"title":"Hands-On Guide to Vaex &#8211; Tool to Overcome Drawbacks of Pandas","content":"Pandas is an open-source data analysis and manipulation tool built on python. It is generally used for manipulating numerical and time-series data. It is used to create data structures like a data frame. Pandas is one of the most used python libraries but it has certain drawbacks like it uses a slow function which is not very suitable for bigger datasets, also pandas only handle results that fit in the memory which can be easily filled. To overcome these drawbacks of Pandas, let us explore a high-performance python library for lazy Out-of-Core Dataframes named Vaex which is used to visualize and manipulate big tabular datasets. It performs different statistical functions and visualizations on very large datasets within seconds. Vaex in python uses Lazy computation and Memory mapping in which no memory is wasted. It loads a dataset with billions of rows in a few seconds. In this article we will explore: How to use Vaex in python?Visualization using Vaex?Comparing Vaex and Pandas Implementation of Vaex in Python We will start exploring vaex but before to that, we need to install it using pip install vaex Importing libraries We will import both pandas and vaex library as we need to compare the performance of both. import vaex import pandas as pd Using Vaex We will explore how to load a dataset in vaex and perform different operations on it. The dataset we are using here is of NYC Motor Vehicle Collision which is of around 350 MB. We will load this dataset using vaex. df = vaex.open(‘motor_nyc.csv’) df.head(5) Basic Operations on the dataset: %%time df['NUMBER OF PEDESTRIANS KILLED'].mean() %%time df['CONTRIBUTING FACTOR VEHICLE 1'].value_counts() %%time df['CROSS STREET NAME'].count() Visualization with Vaex Now we will visualize some of the plots using the data frame loaded using vaex and note the time using ‘%%time’ %%time df.plot1d(df['COLLISION_ID']); %%time df.plot(df['NUMBER OF PERSONS INJURED'], df['NUMBER OF PERSONS KILLED']); Here we can see that despite the dataset being large in size Vaex in python did not take much time to create the plots. Similarly, we can also create several other plots and note the time taken by Vaex as very less compared to pandas. Vaex has several other features like: It can read data from a large number of sources like cs, hdf5, astropy table, etc.It supports all major types of visualization like heatmaps, scatter plots, etc.It supports all statistical functions like variance, co-variance, etc. It is blazingly fast as it works on lazy computing and zero memory copying policy. d. Vaex V\/s Pandas Now let’s compare the time taken by pandas and vaex for different operations. Loading the same dataset #Using Pandas %%time df = pd.read_csv('motor_nyc.csv') #Using Vaex %%time df1 = vaex.open('motor_nyc..csv') Here we can see that while pandas took 18 seconds Vaex loaded the same dataset in 23 milliseconds. Performing Statistical analysis #Using Pandas %%time print(df['NUMBER OF PEDESTRIANS KILLED'].mean()) print(df['NUMBER OF PEDESTRIANS KILLED'].value_counts()) #Using Vaex %%timeit print(df1['NUMBER OF PEDESTRIANS KILLED'].mean()) print(df1['NUMBER OF PEDESTRIANS KILLED'].value_counts()) Here also we can see that vaex is incredibly faster than pandas. Similarly, we can try different operations using both pandas and vaex to find out that Vaex is faster than pandas. Conclusion: In this article we discussed: How we can use Vaex in python for larger datasets.Visualization using Vaex dataset We compared pandas and vaex to find out that Vaex is pretty much faster than pandas.","excerpt":"To overcome these drawbacks of Pandas, let us explore a high-performance python library for lazy Out-of-Core Dataframes named Vaex which is used to visualize and manipulate big tabular datasets.","categories":["Deep Tech"],"tags":[],"author_name":"Himanshu Sharma","publish_date":"2020-08-05T15:00:00","publication_year":"2020","word_count":555,"keywords":["data_tools:Pandas","programming_languages:R","AI","Python","VAE","programming_languages:Python","R","Pandas"],"extracted_tech_keywords":["AI","Pandas","Python","R","VAE","programming_languages:Python","programming_languages:R","data_tools:Pandas"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/hands-on-guide-to-vaex-tool-to-overcome-drawbacks-of-pandas\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":17393,"title":"AI Driven Healthcare Startup Mfine Secures Seed Funding From Stellaris Venture Partners","content":"Mfine, an online platform for doctor consultation, which is not yet launched and claims to be India’s first artificial intelligence-driven healthcare network, has raised Rs 9.5 crore ($1.5 million) in its seed funding round from Stellaris Venture Partners. The funding will be used by the Bengaluru-based startup for product development and to meet the working capital requirements. Apart from Stellaris, the round also witnessed participation of Mayur Abhaya and Rohit MA, healthcare entrepreneurs from Stellaris’ Founder Network. As a part of this funding round, Stellaris’ partner Ritesh Banglani will now be joining the company’s board. “The seasoned founding team brings together scale, execution experience and deep technology expertise to create a unique artificial intelligence application in healthcare,” said Banglani. The startup is the brainchild of online fashion retailer Myntra’s co-founder Ashutosh Lawania. He founded the startup along with Prasad Kompalli, who is acting as a chief executive officer of the company. “At mfine, we use technology to make a high-quality hospital network accessible on demand to consumers and will help provide a proactive care experience for patients,” Kompalli said. Lawania, who had co-founded Myntra along with Mukesh Bansal and Vineet Saxena in 2007, had quit the online fashion portal in February, following a top management reshuffle. Kompalli was previously heading the company’s e-commerce platform and had quit in December 2016. Mfine was established in February this year. The company has hired many of Myntra’s team members. Ajit Narayanan, who quit as Myntra’s chief technology officer earlier in June, has now joined Mfine as its chief technology officer. Along with Ajit, the company has also hired Arjun Choudhary as chief business officer. He was working as a head of growth and sales at Myntra. Mfine platform provides a chat window using which patients can consult doctors from hospitals on the network. Patients can consult paediatricians, gynaecologists, obstetricians and physicians. It also enables to collect samples for lab tests and order medicines online.","excerpt":"Mfine, an online platform for doctor consultation, which is not yet launched and claims to be India’s first artificial intelligence-driven healthcare network, has raised Rs 9.5 crore ($1.5 million) in its seed funding round from Stellaris Venture Partners. The funding will be used by the Bengaluru-based startup for product development and to meet the working […]","categories":["AI News"],"tags":["Funding","health-tech","Mfine","seed funding","Startups"],"author_name":"Priya Singh","publish_date":"2017-09-01T07:57:04","publication_year":"2017","word_count":321,"keywords":["seed funding","API","Funding","artificial intelligence","funding","programming_languages:R","AI","R","Ray","Aim","Startups","Mfine","health-tech","startup"],"extracted_tech_keywords":["AI","artificial intelligence","Aim","Ray","R","API","startup","funding","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/health-tech-startup-mfine-secures-rs-9-5cr-seed-funding-stellaris-venture-partners\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10052923,"title":"Planning to Leverage Open Source? Go Ahead! Here’s Why","content":"Open source is a great option to build your projects on, but there are always a few questions and doubts that may deter people from exploring it. It is predicted that the open-source service market will grow at a CAGR 18.2 percent between 2021-26 and is expected to reach USD 50 billion by 2026. Here are answers to some questions that one just always has while using open source. Will it be up to the mark? While working on an open-source code, it is quite natural to have doubts regarding quality. There is a fear of building your whole project on freely available code. At such a time, it is essential to remember that while you are working with an open-source code, there can be over a hundred developers who are working on the same code too. In terms of quality and security – the more the eyes, the less chance of it having any issues. The core principle of open joint developments is that the issues and vulnerabilities are not hidden from any participant and are even fixed faster. With these, it is easier to fix your code by just implementing a new patch. Participating in open source projects is also considered a bold move. The person’s work is subjected to criticism and analysis and brings their reputation at stake, and no one wants to lose it. Making and leveraging open source is not just limited to SMBs. Many large technology companies like IBM and Google have open-source departments. The specialists who work on open-source projects in-directly work for the company. In fact, open-source projects boast of a global community of developers who have improved product quality and resolved bug issues. Google’s Android is the biggest example of a company that bases an enormous portion of its business on it. Be it small or large, all companies are using more and contributing more to the open-source community. Is it free? Yes, open-source is free. But that doesn’t mean that it does not cost money. Like any other solution, the price of the software has always been a small fraction of the total costs. It is essential to consider the total costs of the project before implementing it. Surely, as deployment expands there aren’t any requirements for additional licenses. Every deployment can have a different cost, including the cost of services and time in installation and development. It is also essential to consider that cost is not the reason why open source is preferred. Instead, the defining character of an open-source solution is the ability to access the source code of the software. Many companies also give away some software for free. But this doesn’t mean it is equal to open source. If commercial software is free, it mostly allows you to use it for the purposes that have been made available. One cannot redistribute or have the right to modify – like they do in open source. GitHub too, does have a lot of projects without an open-source licence and are public. But, commercial companies do not have the right to modify or redistribute them. The absence of a licence means that default copyright laws apply to it. Legal And Copyright Issues Open source components come with some terms and conditions that one must comply with to avoid any business and legal risks. The open-source initiative has identified over 100 licences that may apply to using open source globally. It is essential to pick the correct one and comply. Apache, BSD, MIT and GPL are the most widely used licenses, but it is any day great to check out the others if they solve the purpose better. Yes, the software is unique, and it is essential to understand the difference between copyright and patent. Copyrights in a project protect the source code and the unique expression, while a patent protects the innovation or the functionality that has been built using it. Patents, hence become far more valuable than the codes and copyrights. The code can be written in any language and can be a base for multiple functions, but it is essential to patent your innovation. If you plan to patent, ensure that the code’s copyright and licence do not clash and are not restrictive. Open source boosts innovation and makes the technology industry grow faster. Exploring open source leads to improved productivity, more learning, and it does result in the positive growth of the software.","excerpt":"Answering the most asked questions while building on an open-source code. Will it be up to the mark? Is it really free? What are the legal issues?","categories":["AI Features"],"tags":["Open Source","open source contribution","open source project","open source projects on github"],"author_name":"Meeta Ramnani","publish_date":"2021-11-05T13:00:00","publication_year":"2021","word_count":737,"keywords":["Go","open source projects on github","Open Source","AWS","AI","innovation","Git","open source contribution","RAG","ViT","open source project","GitHub","R","Redis"],"extracted_tech_keywords":["AI","RAG","AWS","Redis","R","Go","Git","GitHub","ViT","innovation"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/open-source\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":39128,"title":"5 AI-Powered Tools To Automate Ad Campaigns","content":"We are living in an era where our social media feeds are specially curated considering our interests.. This is the age of relevance and accuracy and social media platforms such as Facebook, was one of the frontrunners in personalising the feed. Over the last three years, Facebook advertisements have evolved and today it is one of the most important things to keep in mind while running ad campaigns. Backed by AI, Facebook ads are becoming more powerful delivering a significant level of value, saving our time and money. If you are someone from the social media domain and are involved in running multiple ad campaigns for your organisation, here are some of the top AI-powered automation tools you should definitely use. Reveal Bot Reveal Bot is one of the first and best AI-powered tools when it comes to automating Facebook ad campaigns. The tool delivers insights such as impressions, clicks, view reach etc. and also let the advertiser know if s\/he is doing something wrong and losing money along with tips to improve the campaign. That is not all, the data provided by the tool is not just in numbers, but it also shows graphs to make the advertiser better understand how the ad is performing. One of the best features about Reveal Bot is it is integrated with Slack, so now you can have all the insights without even leaving the team messenger platform. It is a paid tool. However, it comes with 14-days trial (without credit card). For pricing info, click here. Zalster Started out in 2015, Zalster is one such tool for automating ad campaigns that are developed in order to eliminate the mistakes usually done by humans. Powered by Artificial Intelligence, this top-notch tool optimizes the advertiser’s bids, budgets, target groups and advertisements, and delivers the best possible results. The tool’s automated algorithms are designed in such a way that it optimizes everything round the clock. That is not all, one of the best features of this tool is that the advertiser\/user will have a dedicated Client Success Manager to sort out or solve any query — whether it’s strategic or technical. Also, just like reveal bot, even Zalster has a Slack bot that delivers insight. It is not a free tool. For pricing, click here. You can also book a demo of the tool. aitarget A Facebook partner for more than five years, aitarget is an AI-powered SwaS (Software WITH a Service) that helps advertisers automate routine operations, scale FB&IG campaigns. Whether its customer targeting or content management or delivering conversions, aitarget is a one-stop shop for advertisers. Having a set of different automation along with significant data from the Facebook ad manager and Google Analytics, aitarget is considered to be one of the best tools at present. Pricing: The expert subscription comes for  $2500 per month. However, you can also try the demo before. Trapica Trapica is an AI targeting optimization and marketing automation platform for social network ad campaigns. The tool is designed in such a way that using AI, it efficiently analyzes thousands of real-time conversion events from campaigns and finds the best-suited bunch of audience. Being a partner of some of the big names like Facebook, Instagram, Twitter, Google, LinkedIn etc. Trapica today is one of the most sought after tools. Furthermore, one of the best features of Trapica is creative A\/B testing. The tool’s AI A\/B tests existing creatives and decides which one is most likely to get the current audience to convert. That is not all, every time the tool discovers any underperforming creatives are automatically turned off. Visit the official website for pricing. Also, Trapica offers free tools for suggestions and insight. ReFUEL4 A weak ad is on average 2.2 times less effective than a good one. Meaning, advertisers are not only wasting their time but all wasting more than half of their da campaign budgets. And this where ReFUEL4 comes in. ReFUEL4 is another automating ad tool that every advertiser should consider using. This AI-powered tool is built in such a way that it makes sure the creatives are always relevant and unto the mark. Also, the tool analyzes when ads start to go down and perform bad, and every time it finds a fatigue ad, the tool replaces that ad with a brand new one. Also, its AI predicts performance with 87% accuracy and adds more precision to A\/B Testing. Book a demo here.","excerpt":"We are living in an era where our social media feeds are specially curated considering our interests.. This is the age of relevance and accuracy and social media platforms such as Facebook, was one of the frontrunners in personalising the feed. Over the last three years, Facebook advertisements have evolved and today it is one […]","categories":["AI Trends"],"tags":["Automation"],"author_name":"Harshajit Sarmah","publish_date":"2019-05-14T12:16:29","publication_year":"2019","word_count":738,"keywords":["Go","API","artificial intelligence","programming_languages:R","AI","Automation","RAG","automation","analytics","GAN","R"],"extracted_tech_keywords":["AI","artificial intelligence","analytics","RAG","R","Go","API","GAN","automation","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/5-ai-powered-tools-to-automate-ad-campaigns\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10004777,"title":"Meet The MachineHack Champions Who Cracked The ‘Melanoma Tumor Size Prediction’ Hackathon","content":"MachineHack successfully concluded its 15th instalment of the weekend hackathon series last Monday. The Melanoma Tumor Size Prediction: Weekend Hackathon #15 hackathon was greatly welcomed by data science enthusiasts and practitioners. Out of the 205 competitors, three topped our leaderboard. In this article, we will introduce you to the winners and describe the approach they took to solve the problem. #1: Devesh Darshan Currently, in his second year of Engineering at Birla Institute of Technology and Science, Pilani, Devesh first came across the term data science during his first year. Like many, he started his journey with the popular Stanford University course by Andrew Ng. His curiosity led him to many other popular online courses as well. He started practising with simple data sets like Titanic and House price on Kaggle. He spends most of his time reading articles and blogs of Analytics India Magazine and Medium to learn new ML techniques. Approach To Solving The Problem Devesh explains his approach as follows: The problem seemed simple at first site, but later I realized that to get a good score in this competition, you just couldn’t rely only on modelling and different ensembles. The things that helped me to reach a good score were: Feature Engineering: This played a major role for me. Engineering relevant features from the dataset was a challenge at first, but then I was able to create 13 new features that helped the model a lot.Model Selection: Surprisingly, Extra Trees Regressor worked the best here, so choosing the right model was also very important in this problem.Ensembling: The final edge was provided by Stack Ensemble of 5 models that were least correlated with each other. “MachineHack is the best platform for new data scientists to practice and test their skills, as some of the problems stated are very beginner-friendly, unlike Kaggle or other platforms where the problems require experience and a better machine to implement the solution” – Devesh shared his experience. Check out the code here #2: Harshita Gupta Harshita is a third-year student currently pursuing Civil Engineering at Birla Institute of Technology and Science Pilani. She started her data science and machine learning  journey in her second year at college  by taking some relevant courses from the institute. To get a better insight into the field she did various online courses as well. She also prefers reading a lot of  blogs and articles. She constantly participates in hackathons to improve her practical knowledge in the domain. Approach To Solving The Problem Harshita explains her approach as follows: The approach was really simple, I started with : Exploration of data using correlation matrix, data analysis and distribution of the featuresFeature engineering, as it is an important step that could significantly improve my scoreFinally trying out different regressors such as LGBMRegressor, BaggingRegressor, ExtraTreeRegrssor, XGBoostRegressor and RandomForestRegressor.The major step which boosted my score was using ensembles by combining different regression models and fine-tuning. “MachineHack is an amazing platform for beginners. I really enjoy applying my theoretical knowledge on their hackathons conducted every weekend.Due to a huge competition on the leaderboard, one is really pushed to the limits to solve the problem, thus encouraging a person to improve a lot. Glad to be a part of the  MachineHack community” – she shared her opinion Check out the code here #3| Devrup Banerjee Although Devrup learnt python just out of the sheer need to automate the routine work and gather data at scale, his real enthusiasm and passion for data science sprouted in his second year of MBA at Great Lakes Institute of Management, Gurgaon, while he was attending his marketing and retail analytics class. He realised that the real motivation behind learning all these algorithms was not about enhancing accuracy but to tell your client by how much you can promise to increase their bottom-line if they were to follow your exact given path. The subject changed his life. “My roommate, who was also equally inspired, and I used to have sleepless nights just going through the 25 lacs dataset given as a final project with our rickety computers to generate actionable insights. To better the bottom-line percentage, that’s what inspired me into analytics.“ – He said His team has won many competitions at MBA level, won at IIT Kanpur MRA tournament while finishing as runners up at IIM Kashipur’s case study on analytics. He is currently trying to deep dive into data science to better his analytical skills so that if someone gives him a dataset in future, he can be both a business analyst and a data scientist. Approach To Solving The Problem Devrup explains his approach briefly as follows: The train and test sets contained lots of duplicates. Some of those duplicates in the test set exactly matched those within the training data. So I grouped the values in the train by mean of target and manually imputed all the exact matches with this mean in the test. There were a couple of zero values which needed to be handled too. Then came the blending of two relatively strong yet different model setup, combining which resulted in good scores. “Machinehack is doing a great job organizing the hackathons and data science summits which not only provides exposure to budding data scientists and students like me, but also helps in opening up networking opportunities with the very best in the domain”- he shared his opinion about MachineHack Check out the code here Check out new hackathons here.","excerpt":"MachineHack successfully concluded its 15th instalment of the weekend hackathon series last Monday. The Melanoma Tumor Size Prediction: Weekend Hackathon #15 hackathon was greatly welcomed by data science enthusiasts and practitioners. Out of the 205 competitors, three topped our leaderboard. In this article, we will introduce you to the winners and describe the approach they […]","categories":["Deep Tech"],"tags":["Data Science","Hackathon Winners","machine learning blending","machinehack winners"],"author_name":"Amal Nair","publish_date":"2020-08-13T14:00:57","publication_year":"2020","word_count":911,"keywords":["data science","Go","machine learning","AI","ML","machine learning blending","RAG","Python","XGBoost","analytics","machinehack winners","Data Science","Hackathon Winners","R"],"extracted_tech_keywords":["AI","machine learning","ML","data science","analytics","XGBoost","RAG","Python","R","Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/meet-the-machinehack-champions-who-cracked-the-melanoma-tumor-size-prediction-hackathon\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":15869,"title":"Data Science Provides Huge Job Opportunities. Here’s How You Can Capture Them","content":"This article is originally the part of Analytics and Data Science India Jobs Study 2017- by Edvancer & AIM The Indian IT industry is in the throes of a mid-life crisis riled by a deep slowdown which is resulting in redundancy and job losses. But as the sun sets in the IT industry in India as we know it, another sun is rising in the form of the data science industry which is growing by leaps and bounds. India with its immense potential talent in the form of engineers, quants, business managers is fast emerging as the Data Science Capital of the world. This is borne out by the fact that global organizations like NEC, Mercedes-Benz, Target, Walmart, PayPal, AIG, Accenture etc. have set up their data science centers of excellence in India. The slowdown in old IT technologies coupled with the rapid rise of data science has resulted in IT companies trying to hire data scientists by the droves and creating a job boom for analytics & data science professionals in India. Jobs in analytics & data science have grown by 100% over the last year adding tens of thousands of employment opportunities and we expect this growth to only intensify over the next few years. IT companies have an excess of talent in technologies which are no longer required and a large proportion of these people will eventually need to be re-skilled in areas which are in demand of which data science is one of the biggest. Nasscom expects that almost 50% of the IT workforce will need to learn these new technologies to avoid becoming redundant. The challenge for employers thus lies in the shortage of people who are skilled in data science tools & technologies and their applications in business. Salaries are rapidly increasing given the increased demand and shortfall of talent and hence there is no better time than now for you to develop the data science skill sets needed by industry. At Edvancer, when we speak to aspiring data scientists looking to take up our courses, a common theme from the questions we are asked is how will I fit into data science given my existing background and will I get a job? The unique and best aspect of data science is that it is not limited to any industry or sector. Data Science today is being used in every industry in this world from manufacturing to retail to healthcare etc. Data is almost omnipresent now as the collection of data has become much easier and storage is cheaper. With humongous amounts of data being generated daily, companies across sectors are taking advantage and making use of the insights gained from the analysis of that data to benefit. Data science through machine learning & artificial intelligence is proving to be instrumental in pushing the boundaries of science and what was science fiction 10-15 years ago is turning into reality now. This has opened numerous opportunities in data science for a large cross section of people from varied backgrounds and experiences who have developed the relevant data science skill sets. So how does one capture this opportunity? You need to take a structured and patient approach to creating a career in data science. Follow the path below and you will find yourself in the “sexiest career of the 21st century.” Get trained in the relevant skillsets: Data Scientists have expertise in a wide variety of tools and technologies with the most in-demand tools being R, Python, SAS, Hadoop, Spark, NoSql while being well versed in statistics, predictive modeling, linear algebra, machine learning, text mining, and handling big data. Don’t let all these different requirements worry you. Take a systematic approach. Start with learning R\/SAS, statistics and predictive analytics. Then move on to Python, linear algebra, and machine learning before working on the Hadoop, Spark & NoSql stack for big data. Implement your learning on projects and display them: Employers look for people with a practical perspective and the ability to start contributing on the job on day one. To that end, they will look for data science projects on your CV. Either you can work on projects provided on Kaggle and Crowdanalytix or use the many freely available data sets on the internet. Another way is to combine your learning and project work by joining a course which provides you multiple projects to implement your learning. Create your Github profile and host your projects and codes over there so that you can share it when looking for a job. Create a blog where you can write about these projects and on data science in general. Create your CV and update your Linkedin profiles: Well you obviously have to let the world know that you have the necessary capabilities for a data science role. So, update your CV with the tools and technologies that you have learned and describe the projects you have worked on with an emphasis on the outcomes you were able to deliver. Also, provide your Github link in the CV. Make sure you update your Linkedin profile too as all potential recruiters will check it. Prepare for interviews: Practice makes perfect and you should prepare and practice for interviews too. Go through various frequently asked interview questions on the internet and prepare for them. Get ready to be tested on what you have mentioned on your CV and also be ready to be tested with case studies and take-home problems to solve. Keep yourself updated on developments in data science. Practice answering estimate based case studies which are designed to test your structured thinking. Network and search for a job: Now that you have the building blocks in place, it is time to go out and get a job in data science. This will be a stage which will require patience. Start by creating a list of target sectors and companies that you are interested in working with or you believe would be compatible with your background. Research about them and be ready with relevant reasons as to why you want to work for them and how you would be able to add value. Check Linkedin, their websites and other job portals for current openings in data science in these companies and apply to them while utilizing your friends and family network too. Network with data scientists on Linkedin & Facebook through groups. Connect with hiring managers and data science leaders as they keep posting their requirements outside of normal channels too. Attend data science meetups and events where you can take the networking offline while continuing to apply for suitable roles through job portals. Opportunities will definitely come your way. In conclusion, data science is a booming industry with lots of opportunities being created. Persistence, the right approach, and hard work will get you into a dream career in this field. Go through this report also to know more about the job situation in details.","excerpt":"This article is originally the part of Analytics and Data Science India Jobs Study 2017- by Edvancer & AIM The Indian IT industry is in the throes of a mid-life crisis riled by a deep slowdown which is resulting in redundancy and job losses. But as the sun sets in the IT industry in India as […]","categories":[],"tags":[],"author_name":"Aatash Shah","publish_date":"2017-06-24T08:56:47","publication_year":"2017","word_count":1154,"keywords":["data science","machine learning","artificial intelligence","AI","R","RAG","Python","Aim","analytics","predictive analytics"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","data science","analytics","Aim","RAG","predictive analytics","Python","R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/data-science-provides-huge-job-opportunities-heres-can-capture\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":58440,"title":"Kisan Suvidha: How Two Women Techies Built A Blockchain &#038; AI System To Help Farmers In India","content":"Kisan Suvidha is an automated, transparent and low-cost system for farmers in India that can ensure timely crop insurance payments by using AI and blockchain, developed by a team of two women techies. Recently at an AI and blockchain hackathon organised by Icertis, two women techies under Team Heuristic showcased their unique solution aimed at helping millions of Indian farmers. Laisha Wadhwa, A Microsoft Student Partner and Garima Yadav, Graduate Analyst at Barclays came together for an all-women team (Team Heuristic) for a hackathon organised by Icertis. The duo built Kisan Suvidha — an automated, transparent and low-cost system that can ensure timely crop insurance payments by using AI and blockchain. The solution enables insurance policy managers to process claims (manually raised by farmers) based on its weather risk predictions. Using Kisan Suvidha, the farmer just needs to upload its crop insurance policy and a picture of his farm. The app automatically takes in the location of the farm from the policy and notes down data such as weather condition, rain, humidity and temperature of a location for the next seven days and determines the severity on a scale of 1 to 5 using AI. The thresholds for payouts are set accordingly to the severity levels, and the payments will be triggered using blockchain smart contracts. An important step of the proposed solution was: calculating the percentage amount for payment from the insurance amount based on weather risk predictions. Kisan Suvidha Leverages Advanced Analytics On Different Data Sets Kisan Suvidha uses multimodal weather data — weather patterns like rain, temperature, humidity, soil moisture, wind pattern and takes all these data points across a seven day advance period. The system also comes with features like, farm condition monitor (weather and soil), timely push notifications regarding weather and an easy verification system that authenticates the farmer’s identity using five different kinds of identification proofs. The data was obtained from APIs like open weather API and HERE API. We also use robust computer vision classification models to identify the identity proof of the farmer to authenticate a valid user of the Kisan Suvidha system. The classification model works on five different types of identification proofs: Aadhar Card, PAN Card, Voter ID card, Driving License and passbook. It then uses the Google Cloud vision API (OCR: Optical Character Recognition) to automatically fetch all the important details from the Crop Insurance policy. “Two very interesting aspects of our solution are – the facility for manual claim from the farmer in case the data from the weather APIs falters. Secondly, post-filing a claim the farmer can see the status of all his claims under the Insurance claim history section,” said Team Heuristic. Using Blockchain Smart Contracts After the details are fetched, they are stored in blockchain distributed ledger, once the farmer details are a part of the blockchain, they can’t be altered, hence the data is tamper-proof and secure. Kisan Suvidha also makes use of smart contracts, which automatically get triggered when certain conditions are met, in their case, when climatic conditions cross a threshold value. Team Heuristic members told, “Our solution makes use of smart contracts, which automatically gets triggered when certain conditions are met, in our case, when climatic conditions cross a threshold value.” Using the data for the next seven days the probability of risk of extreme weather conditions is predicted using the multimodal fusion of all-weather data for AI. Based on the severity of the risk prediction, the per cent of insurance amount (ranging from 45% to 100%) is disbursed in the wallet of the farmer through a blockchain based smart contract, which allows for credible transactions without any third-party intervention. Hence there’s no need to pay intermediaries (middlemen), and it saves the farmer’s time and long waits at government offices in India for money. Helping Farmers Keep Track With Blockchain, AI & A Web App Kisan Suvidha not only helps farmers maintain multiple crop insurance policy but also helps them understand the weather patterns in their area in an easily comprehensible dashboard. The farmer also has the privilege of monitoring his farm (logging historical data points – like temperature) by selecting his farm on a simple to use map integrated using the Dashboard for Agricultural Monitoring available in Agro API from agromonitoring.com. “We are planning to expand our solution by embedding a buddy, where farmers belonging to the same area with similar soil type and crops will be grouped together so that they can discuss irrigation frequency and other solutions to avoid crop failure,” told Laisha and Garima from Team Heuristic. About The Duo That Makes Team Heuristic Laisha Wadhwa is a machine learning and deep learning aficionado with an interest in leveraging the power of algorithms and mathematics to solve real-world problems. Currently, she’s an undergraduate at IIIT Sri City, and works with big-data analytical tools, creates and maintain models, and explore new data sets to hypothesize new solutions. Garima Yadav, on the other hand, is an experienced Lead Web Developer with a demonstrated history of working in the computer software industry. According to Garima, she is skilled in Java, SQL, Spring Boot, IBM DB2, PHP, and C++. Also Read: Key Takeaways From India Blockchain Strategy Paper By NITI Aayog","excerpt":"Kisan Suvidha is an automated, transparent and low-cost system for farmers in India that can ensure timely crop insurance payments by using AI and blockchain, developed by a team of two women techies. Recently at an AI and blockchain hackathon organised by Icertis, two women techies under Team Heuristic showcased their unique solution aimed at […]","categories":["AI Features"],"tags":["Blockchain","blockchain india","Interviews and Discussions","Women in Tech"],"author_name":"Vishal Chawla","publish_date":"2020-03-12T11:48:00","publication_year":"2020","word_count":870,"keywords":["Go","machine learning","Blockchain","blockchain india","AI","computer vision","RAG","Aim","deep learning","Women in Tech","analytics","SQL","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","machine learning","deep learning","computer vision","analytics","Aim","RAG","R","SQL","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/kisan-suvidha-how-two-women-techies-built-a-blockchain-ai-system-to-help-farmers-in-india\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":15628,"title":"A look at how Byju’s personalization engine is driving one-on-one learning experience","content":"Ranjit Radhakrishnan, Chief Product Officer at BYJU’S Byju’s, India’s largest K-12 learning app is one of the best funded ed-tech ventures and is billed to become India’s first educational unicorn. The popular K-12 education startup also earned another notch this year when it made to the Harvard Business school case study and is currently valued at $600-650 million. Valuations aside, Analytics India Magazine caught up with Byju’s Chief Product Officer Ranjit Radhakrishnan to understand Byju’s analytics system and the personalization engine that drives the impeccable online learning experience. Radhakrishnan cites the backbone of Byju’s personalization engine is the rich learning profile that is built for every student. “We have our own analytics systems which capture millions of data points every day. This drives multiple decisions – such as from improving a feature to building highly contextual recommendations. The variety of student profiles also gives us an idea on what to improve or add to the overall content. It helps us change and modify the values of the parameters and properties associated with the various learning content,” said Radhakrishnan. A look at Byju’s personalization engine The edtech company is credited for creating a new segment of self-paced learners and its personalized focus has delivered 90% retention year-on-year. Talking about creating personalized learning experience, Radhakrishnan shares, “This is further powered by deep knowledge graphs of over 50,000 concepts and relationships that have been created to design personal learning journeys — videos, questions, adaptive flows, quizzes, flashcards, correctional learning videos etc”. Additionally, the learning content is also tagged to multiple other properties and parameters. Case in point, when a student starts the learning journey, a rich learning profile is built to personalize the experience. “This enables us to customize and personalize the learning experience for students based on their strengths and weaknesses and their specific learning gaps, pace and needs,” he said. User Experience is at the heart of Byju’s success The edtech venture uses the app data as the primary data backed by information about logged in web usage to enhance the user experience better. The user’s interactions with the websites, emails, etc are tracked in a separate system that is used for bettering marketing automation and behaviour analysis. “And the actions that define the user’s knowledge profile are based on interactions with content pieces,” shared Radhakrishnan. The content is tagged at the sub-micro concept level This helps the algorithms to create a smooth learning curve for the student so that they feel challenged yet motivated This app works in a way similar fashion like a personal tutor and addresses the student’s learning gap in a more effective way backed by deep insights Byju’s built its success on the one-on-one personalized experience that appeals to students as they are guided throughout the learning process in an individualized manner. “We don’t allow learning gaps to fester. Corrections, actionable feedback, recommendations and custom learning paths are created for the user. With technology as an enabler, we focus on developing learning programs that will cater to the unique learning requirements of every student in India and abroad,” Radhakrishnan explained. Analytics driving better content creation Byju Raveendran, founder and CEO of India’s best-funded ed-tech start-up According to Byju’s senior executive, the content creation process is highly influenced by the data analytics at work as a lot of data and insights is generated in house. Case in point – if a large percentage of students are making mistakes in a particular concept, Byju’s add simpler videos, more content formats and easier questions to allow the system to create a smoother learning curve. “In this way, students end up learning the difficult aspects of the concept gradually without getting demotivated,” reveals Radhakrishnan. The edtech venture also looks at specific learning content tagged to those concepts and constantly updates the parameters so that future learning paths for newer students are smoother. The effectiveness of all content modules and modes are also constantly analyzed for improvement. Byju’s launches new parent app The newly launched parents’ app is a step to help parents participate in the child’s learning journey. The app provides parents a better understanding of their child’s learning progress. “Generally, for parents, the assessment is limited to test and exam reports. Through this app, we want parents to understand where the child is putting effort, observe where the gaps are and provide pointers on specific areas where they can encourage their child’s learning journey as a participative stakeholder,” said Radhakrishnan Today, the app has over 8 million users and 4,00,000 annual paid subscriptions. And there are over 100 million lessons watched with average time spent on the app being 40 minutes a day.","excerpt":"Byju’s, India’s largest K-12 learning app is one of the best funded ed-tech ventures and is billed to become India’s first educational unicorn. The popular K-12 education startup also earned another notch this year when it made to the Harvard Business school case study and is currently valued at $600-650 million. Valuations aside, Analytics India […]","categories":["IT Services"],"tags":[],"author_name":"Richa Bhatia","publish_date":"2017-06-15T03:17:45","publication_year":"2017","word_count":774,"keywords":["knowledge graphs","Go","unicorn","programming_languages:R","AI","RAG","automation","analytics","R","startup"],"extracted_tech_keywords":["AI","analytics","RAG","knowledge graphs","R","Go","automation","startup","unicorn","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/it-services\/look-byjus-personalization-engine-driving-one-one-learning-experience\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10020811,"title":"The Rise And Rise Of AI Gaming Industry","content":"Latitude, a startup building games using AI-generated infinity storylines, raised $3.3 million in seed funding earlier this month. Founded in 2019, Latitude has secured a total funding of $4.1 million so far. The startup specialises in developing AI-powered games geared towards player freedom and self-expression. Latitude’s AI Dungeon uses GPT2\/3 language models to develop a procedural, emergent text adventure game, where the generative models fill in actions in the background. The impressively open-ended AI Dungeon was launched in December 2019 using the GPT2 model. Latitude is not the only AI gaming startup flush with funds. askblu.ai, founded by French mobile game studio Happy Blue Fish in 2017, raised a seed funding of €300K in 2017, followed by the second round of seed funding of €200K, managed by AGORA (also an investor) in 2018. askblu.ai offers an efficient, plug & play, scalable SaaS platform to help mobile game studios make more revenue through increased player retention. It uses machine learning and predictive analytics to personalise each player’s experience in the game in real-time. Osmo, an interactive play company, has raised $32.5 million in funding. Recently acquired by Byju’s, the gaming company is built around a proprietary Reflective Artificial Intelligence. Osmo expands the playing field and encourages creative thinking and social interaction, allowing any object such as a pen or paper to interact with the digital device. Gosu Data Lab, another AI gaming startup based out of Lithuania, raised $5.1 million in funding. Focused on gaming data and AI, it builds a platform that applies the computer’s understanding of the gaming data to help gamers play better. It also develops voice assistants to help gamers. Danish artificial intelligence firm Modl.ai closed a $1.7 million seed funding round in February last year. Founded by veterans in the areas of games, psychometrics, and artificial intelligence, Modl.ai delivers game technology that accelerates game development and enhances player engagement. The startup has reportedly used the fresh capital to continue the development of its AI playtesting tech. By deploying AI-driven player archetypes, Modl.ai mimics human behaviour for playtesting. Still in beta, Modl.ai has already secured contracts from global gaming studios. AI gaming startup Krikey has raised a total funding of $22 million in a Series A led by Reliance Jio. Krikey’s AR game Yaatra invites users on an action-adventure journey to defeat a monster army using weapons such as bow and arrow, chakra, lightning and fire bolts. https:\/\/youtu.be\/SBMK9itaGxk AI Gaming Market AI’s groundbreaking moment came in 1997 when world champion Garry Kasparov lost to IBM’s Deep Blue in chess. Since then, companies such as DeepMind, Apex Game tools, Blizzard Environment, Rival Theory, Heroz, Spirit AI have improved game assets, behaviours, gaming environments etc. According to a report, the global gaming industry is expected to register a CAGR of 12% during 2020-2025. The report observed virtual reality and AI have become an integral part of the gaming industry, with players expected to spend $4.5 billion on immersive gaming in 2020 alone. The mobile gaming market has witnessed tremendous growth over the past ten years, with an approximately $100 billion market cap or 60% of the total video game market in 2021. One of the critical challenges for gaming companies is to attract and retain users. AI offers unique retention and player acquisition solutions to these companies. It provides immersive experiences apart from unique perspectives to engage gamers. The rising funding is a testament to the growing popularity of AI games, and going by the market predictions, the gaming industry is primed for a growth spurt.","excerpt":"Latitude, a startup building games using AI-generated infinity storylines, raised $3.3 million in seed funding earlier this month. Founded in 2019, Latitude has secured a total funding of $4.1 million so far. The startup specialises in developing AI-powered games geared towards player freedom and self-expression.  Latitude’s AI Dungeon uses GPT2\/3 language models to develop a […]","categories":["IT Services"],"tags":[],"author_name":"Srishti Deoras","publish_date":"2021-02-24T18:00:00","publication_year":"2021","word_count":587,"keywords":["Go","machine learning","artificial intelligence","AI","R","Scala","Git","RAG","analytics","predictive analytics"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","analytics","RAG","predictive analytics","R","Go","Scala","Git"],"url":"https:\/\/analyticsindiamag.com\/it-services\/the-rise-and-rise-of-ai-gaming-industry\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10104234,"title":"NexusRaven Outperforms GPT-4 for Zero-shot Function Calling","content":"Nexusflow.ai, has recently launched NexusRaven-V2, a powerful 13-billion parameter LLM that outperforms GPT-4 in zero-shot function calling. The open source model showcases a remarkable capability to transform natural language instructions into executable code, facilitating the utilisation of software tools by copilots and agents. NexusRaven-V2 demonstrates superiority over GPT-4 by achieving up to a 7% higher success rate in function calling in human-generated use cases involving nested and composite functions. Notably, NexusRaven-V2 accomplishes this without prior training on the specific functions used in the evaluation. Check out the model on GitHub here, and on Hugging Face here. Nexusflow.ai introduces the Nexus-Function-Calling benchmark, establishing a Hugging Face leaderboard. This includes a diverse collection of real-life human-curated function-calling examples, with eight out of the nine benchmarks open-sourced. https:\/\/twitter.com\/ClementDelangue\/status\/1732138699901809042 Built on top of Llama 2, leveraging CodeLlama-13B-instruct, NexusRaven-V2 is instruction-tuned and utilises curated data from Nexusflow’s pipeline. The model is commercially permissive, encouraging both community developers and enterprises to explore its capabilities. Nexusflow.ai provides open-source utility artefacts, enabling users to seamlessly replace mainstream proprietary function calling APIs with NexusRaven-V2 in their software workflows. Online demos and Colab notebooks are also available for onboarding and integration demonstrations. NexusRaven-V2 showcases a 4% higher success rate in function calling on average compared to the latest GPT-4 model, as observed in a human-curated benchmark. In tasks involving nested and composite function calls, NexusRaven-V2 exhibits a significant 7% advantage over GPT-4, highlighting its robustness in handling variations in developers’ descriptions of functions. To ensure reproducibility and standardisation, Nexusflow.ai releases the benchmark and associated leaderboard along with model weights. The evaluation benchmark prioritises human-generated samples with meticulous checks on executability and encompasses a diverse representation of function calling use cases and difficulties. Nexusflow.ai is also providing a Python package, “nexusraven,” facilitating easy integration with copilots or agents. Developers can quickly ingest API function descriptions and send natural language queries to the model with a single line of code. The nexusraven package also supports converting function calling code to JSON format for seamless integration with downstream software.","excerpt":"Built on top of Llama 2, leveraging CodeLlama-13B-instruct, NexusRaven-V2 is instruction-tuned and utilises curated data from Nexusflow’s pipeline.","categories":["AI News"],"tags":[],"author_name":"Mohit Pandey","publish_date":"2023-12-06T11:16:39","publication_year":"2023","word_count":336,"keywords":["Hugging Face","AI","ML","Git","RAG","Colab","Python","copilots","GitHub","R"],"extracted_tech_keywords":["AI","ML","Hugging Face","Colab","RAG","copilots","Python","R","Git","GitHub"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/nexusraven-outperforms-gpt-4-for-zero-shot-function-calling\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10140266,"title":"Bengaluru AI Startup smallest.ai Unveils Lightning, New Text-to-Speech Model","content":"smallest.ai, an AI startup headquartered in San Francisco, California focused on multi-modal models, introduced Lightning, text to speech (TTS) model capable of generating up to 10 seconds of audio within 100 milliseconds. This advancement enables developers worldwide to build highly realistic voicebot applications with sub second latency, streamlining the implementation process and making it accessible at significantly lower costs. Lightning supports English and Hindi in multiple accents currently and the team plans to add many more languages quickly. Being priced as low as $0.02 (or approximately 1.6 Rs) per minute, Lightning provides a cost effective solution, enabling the applications to run at under 1 Rs\/minute—drastically reducing expenses for voicebot builders and broadening market accessibility. Lightning’s rapid processing and cost efficiency make it a notable alternative in the voicebot industry, where traditional TTS models often rely on streaming and web sockets, increasing server demands and complicating scalability. This has been designed for practical and high speed integration, Lightning operates through a straightforward REST API, delivering audio in around 100 milliseconds without the server strain associated with continuous streaming. Currently supporting multiple English and Hindi accents, Smallest.ai plans to expand the model’s language range to include other Indian, European, and Asian languages in the coming months. Smallest.ai was founded by IIT Guwahati alumni Sudarshan Kamath and Akshat Mandloi. Kamath attributes the affordability of smallest.ai to their focus on data quality and model efficiency. “Our model is much smaller than those of competitors like ElevenLabs. Despite this, we achieve high-quality speech because our data is highly refined,” he explained. Voicebot developers with early access to Lightning have reported an 8x reduction in operational costs, accompanied by enhanced audio quality. Beyond real-time voicebot applications, Lightning is adaptable for creating audiobooks and voiceovers for social media content on platforms like Instagram and YouTube. Non-developers can access Lightning through the Waves Speech platform, where additional features, including voice cloning and accent conversion, are also available in beta. “When we started building, we realised that the models required for a voice bot were not mature for Indian languages. Existing models for non-English languages were nowhere close to production,” explained Kamath in an exclusive interaction with AIM. Earlier in June, smallest.ai also launched AWAAZ allows voice cloning from short audio clips and is available at competitive rates. The model is aimed at scalable applications in regional language markets and provides enterprise-grade security and compliance.When asked about its mission, Kamath said, “Why are 1B humans not speaking to AI voices everyday despite incredible advancements in Voice AI? This is the problem we are trying to solve.”","excerpt":"Lightning supports English and Hindi in multiple accents currently and the team plans to add many more languages quickly.","categories":["AI News"],"tags":["AI Startups","Startups"],"author_name":"Tanisha Bhattacharjee","publish_date":"2024-11-05T17:56:50","publication_year":"2024","word_count":427,"keywords":["API","AI","ML","Scala","REST API","Aim","data quality","CLIP","Startups","R","AI Startups","startup"],"extracted_tech_keywords":["AI","ML","Aim","R","Scala","API","REST API","data quality","CLIP","startup"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/bengaluru-ai-startup-smallest-ai-unveils-lightning-new-text-to-speech-model\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10164138,"title":"Baidu to Open Source AI Models After DeepSeek’s Success","content":"Chinese tech giant Baidu reported its Q4 FY 24 earnings, with revenue declining 2% to 34.12 billion yuan due to weaker advertising sales. However, its Cloud AI unit grew 26% to 7.1 billion yuan, helping offset some losses. The company’s results exceeded analyst expectations of 33.32 billion yuan, highlighting the strength of its AI-driven initiatives despite economic headwinds and reduced ad spending. During the earnings call, the company’s CEO, Robin Li, said that Hangzhou-based High-Flyer-owned DeepSeek’s open-source approach had accelerated AI adoption. “One thing we learned from DeepSeek is that open sourcing the best models can greatly help adoption. When the model is open source, people naturally want to try it out of curiosity, which helps drive broader adoption,” Li said. The company further announced that it will release the source code for its Ernie 4.5 models in June and offer premium chatbot services for free by April. “So our decision to open source ERNIE 4.5 series, is also backed by our deep confidence in our technology leadership, he added. Baidu was among the first in China to launch a ChatGPT-style chatbot in early 2023, with Ernie 4.0 claiming to rival OpenAI’s GPT-4. However, adoption has lagged due to strong competition, especially from DeepSeek’s R1 model, released last month. Baidu’s AI Growth The company announced that in 2024, generative AI experienced strong market demand as enterprises increasingly adopted AI-driven solutions. Baidu’s AI solutions fuelled adoption and bolstered market confidence in its expertise. “AI-generated content now appears in 22% of search result pages, with efforts focused on quality and relevance,” Li said during the earnings call. Baidu expanded beyond text generation to include short videos, nodes, agents, digital humans, and live streaming. The company placed strong emphasis on agents and said it envisions them becoming one of the most important forms of AI-native applications.","excerpt":"AI-generated content now appears in 22% of search result pages, with efforts focused on quality and relevance,” Li said during the earnings call.","categories":["AI News"],"tags":["AI Models","Baidu","DeepSeek"],"author_name":"Aditi Suresh","publish_date":"2025-02-19T21:21:04","publication_year":"2025","word_count":303,"keywords":["AI Models","ChatGPT","OpenAI","AI","Git","GPT","DeepSeek","Aim","Baidu","generative AI","R","AI-generated content","llm_models:GPT"],"extracted_tech_keywords":["AI","generative AI","ChatGPT","OpenAI","Aim","R","Git","GPT","AI-generated content","llm_models:GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/baidu-to-open-source-ai-models-after-deepseeks-success\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10093353,"title":"Can Bard Hang With The Big Boys After Upgrades?","content":"During Google’s annual developer conference, Google I\/O 2023, significant upgrades were announced for Google Bard. The updates included AI image generation through integration with Adobe Firefly, image search, export features, programming language support, personalised code reviews, and integration with Google Maps. With these updates, Bard has become a genuine contender in the AI chatbot market, and looks to surpass competitors such as ChatGPT and Microsoft’s Bing Chat in some areas. Google has also announced updates to its search tool, incorporating generative AI into core products such as Gmail and Google Photos. The company plans to test this technology with several clients, including Deutsche Bank and Uber Technologies. Google’s chatbot Bard is powered by PaLM 2, a more advanced AI model developed by Google. PaLM 2 allows Bard to write captions for images and function on smartphones. Bard’s updates include integration with Google Docs and Sheets, expanded language support, and coding capabilities. The updates position Bard as a tool to streamline coding and debugging, with added math and logic capabilities. Bard is available in 180 countries and territories and supports languages beyond English, such as Japanese and Korean. Bard vs ChatGPT While Bard is open for everyone to test, it looks like the new capabilities have not been implemented yet and would only be available through the Google Search Labs waitlist. However, we tested Bard, ChatGPT and GPT 4 (which Bing also uses in a limited capacity). We asked Bard to generate a dystopian image inspired by George Orwell’s 1984 it said: So, we asked whether it is using Adobe Firefly to generate images now, it said Firefly is still in development and is not available to the public. On the other hand when we fed the same George Orwell prompt in ChatGPT, it generated this: While neither of the two could not generate an image, ChatGPT generated a detailed description and scenario. Nevertheless, considering the prompt, Bard’s response seemed much more apt. Code: Then we asked if Bard could code an ad blocker for us, and it did!!! On the other hand ChatGPT generated this answer: GPT4 provided a generic JavaScript for the same: While Bard gave us code for an ad blocker and assured us that, ‘This extension will block all ads on all websites’. GPT-4 provided a basic JavaScript of an ad blocker; was realistic in its approach and also suggested much more advanced ad blockers. Then we asked both the chatbots to generate a poem based on Kierkegaard’s work and the results were quite different: ChatGPT generates a good rhyming scheme and seems more refined. But GPT4 came up with a stand out response: Real-time info: Bard stands out when it comes to generating real time information, because it has access to the web. On the other hand, even the paid version of GPT, GPT Plus, needs plugins to access the web. We asked Bard about some updates from Indian politics: ChatGPT gave this response to the same prompt: While Bard gave updates on the Karnataka election amongst other things, ChatGPT was still stuck on the Covid pandemic and the Farmer’s Protest, because it was trained on data until 2021. However, ChatGPT has released more than 70 plugins for ChatGPT Plus (its paid version) which provides users real time access to the web. Then we asked whether Bard was already using Palm-2 or not and it said it had not: So, it looks like Bard is still running on LaMDA, and hence the updates cant be truly gauged as of now. Conclusively, while ChatGPT-generated answers seem much more natural and useful, even without the updates Bard has the edge when it comes to real-time information and coding capabilities. However, currently, the paid version of ChatGPT is considered to be the best chatbot due to its human-like verbose responses compared to Bing and Bard. However, these products are constantly evolving and improvements are expected as Google, Microsoft, and OpenAI continue to feed their AI systems more data and make tweaks. Google is expected to benefit the most from this as it switches from LaMDA to PaLM 2, as Bard’s previous version was not satisfactory. Bard already has additional features like voice summarisation, and exporting answers to documents or Gmail drafts but as additional features like AI image generation through Adobe Firefly, image search, programming language support, personalized code reviews, and integration with Google Maps start rolling in, Bard will have some upper hand in comparison to ChatGPT and Bing. Bard vs the World Both ChatGPT and Bing Chat are AI chatbots that utilise the latest version of GPT-4 to generate unique responses based on text cues. Bing Chat has limitations on inquiries and messages per discussion, requiring Microsoft Edge or the Bing app, while ChatGPT is generally considered the better option. Google’s Bard, initially based on LaMDA, is transitioning to PaLM 2, Google’s advanced AI model. Microsoft recently updated Bing Chat, offering enhanced functionality beyond ChatGPT due to its use of GPT-4. Bing Chat provides free access to GPT-4, albeit with some missing features and usage restrictions. Google provides access to PaLM 2 through various channels, including the PaLM API and Firebase, and Bard will soon support PaLM 2 as well. The new version of ChatGPT introduces features like Visual Answers and Multimodal Support, enabling the AI to analyse images and respond with diverse visuals. Third-party plugins like OpenTable, Wolfram Alpha, and Microsoft Math Solver enhance the AI’s responses. Bing Image Creator is available in multiple languages, and Export and Share functionality facilitates easy sharing of chat content. Bard already has features like voice summarisation and exporting answers to documents or Gmail drafts, with plans to introduce AI image generation, image search, programming language support, personalised code reviews, and Google Maps integration. Bing Chat’s recent update integrates with Microsoft Edge and introduces the Copilot feature for organising tabs and analysing documents, significantly enhancing its capabilities and providing tailored responses. While ChatGPT is developing plugins to extend its functionality, they are still in testing. An anticipated plugin allows the model to browse the internet for up-to-date information, but users have reported issues with its functionality during alpha testing. OpenAI welcomes user feedback through a private beta Slack group for those with access to the code interpreter plugin. Developers can also create plugins by joining a waitlist and utilising the API to connect ChatGPT with third-party applications. Similarly, Google’s Bard supports plugins that provide access to apps like Spotify, Walmart, Indeed, Uber Eats, Adobe Firefly, and Google Apps, expanding its capabilities and offering additional functionalities within the chatbot interface. Learning from ChatGPT’s mistake? The European Union was excluded from the list of countries where Bard was launched. This exclusion has not been addressed by Google, but it is speculated to be related to the EU’s aversion to OpenAI’s ChatGPT after its introduction, and Google is awaiting the upcoming AI Act before making any moves. ChatGPT recently faced bans in Italy for privacy violations, and France, Spain, and Germany are investigating the company’s compliance with GDPR. So, Google seems to be treading carefully. Microsoft has released its Bing Chat preview in 169 countries. Additionally, OpenAI’s offering is marred by privacy concerns and OpenAI’s losses have doubled to $540 million since developing its ChatGPT product. OpenAI attempted to patch the wound with its latest offering in the form of a new subscription tier, ChatGPT Business, for enterprise customers who need more control over their data and want to manage their end-users. On the other hand, Microsoft’s Bing has found business avenues for itself, and has been running ahead in terms of business. Recent reports suggest that Samsung and Apple may switch to Bing as their default search engine, potentially triggering other Android OEMs to explore Microsoft alternatives.","excerpt":"With these updates, Bard has become a genuine contender in the AI chatbot market, and looks to surpass competitors such as ChatGPT and Microsoft’s Bing Chat.","categories":["AI Features"],"tags":["AI Chatbot","Bard","Bing","ChatGPT","Google","Microsoft","OpenAI"],"author_name":"Shyam Nandan Upadhyay","publish_date":"2023-05-15T17:08:46","publication_year":"2023","word_count":1285,"keywords":["Go","ChatGPT","OpenAI","AI","AI Chatbot","Bing","ML","Google Bard","chatbots","Bard","generative AI","Google","JavaScript","R","Microsoft"],"extracted_tech_keywords":["AI","ML","generative AI","ChatGPT","OpenAI","Google Bard","chatbots","R","JavaScript","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/can-bard-hang-with-the-big-boys-after-upgrades\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10009728,"title":"Interesting Puzzles To Prepare For Data Science Interview","content":"The ongoing pandemic has affected hiring in many companies across data science domains. Even if companies are recruiting, they are adopting stringent procedures to spot the right candidate. Puzzles are often preferred to evaluate a candidate as it best accesses the critical thinking ability, problem-solving and in many cases, even the coding skills of a candidate. In this article, we list a few such brainstorming puzzles which help aspiring data science and analytics candidates to exercise their brain muscles. These questions have been sourced from various resources available on the internet and experiences of data science candidates in interviews. Read some other puzzles here. 10 Commonly Asked Puzzles In A Data Science Interview 10 Standard Puzzles Asked During Analytics Interviews 1| A scooter has two tyres and one Stepney. Each tyre can run a maximum distance of five kilometres. How long will the scooter run? To approach this problem, let us assume three tyres to be A, B and C A = 5km; B = 5km, C = 5km At first, we can ride 2.5 kilometres with A and B. A = 2.5 km, B = 2.5 km, C = 5 km Then Remove A and travel with BC to 2.5 km more. Remaining A = 2.5, B = 0, C = 2.5 Then remove B and travel with AC to 2.5 km more. Remaining A = 0, B = 0, C = 0 Therefore, total distance travelled is 2.5 + 2.5 + 2.5 = 7.5 km 2| 100 people are standing in a circle numbered 1 to 100. Person 1 kills the next person (i.e. No. 2) with a sword and gives the sword to the next to next person (i.e. No.3). All of them do the same until only one survives. Which number survives at last? You should get the nearest smaller number that is the power of 2. In this, it is 64. Subtract it will be the total number of people, i.e., 100. 100 – 64 = 36 Now we apply the formula 2n + 1, 2 * 36 + 1= 72 + 1 = 73. The lone survivor will therefore be the No. 73. Read here for a more detailed answer. 3| Two friends A and B meet at a party on a new year day. As they met after a long time, B wanted A to guess his birthday. As they met after a long time, A was unable to guess B’s birthday. B decided to give some hints as below: The day before yesterday I was 25, and next year I will be 28The above condition can be true for only one day in the year. The approach to this problem goes as below: A met B on new year day, let’s assume. January 1, 2019. So, the day before yesterday, i.e. on December 30, 2018, B was 25 years old, and on the present-day, i.e. January 1, 2019, B is 26. On December 31, 2019, B will be 27 and next year, i.e. on December 31, 2020, his age will be 28. So, B’s birthday is on December 31. 4| A person is working for you for five days, and you should pay him a gold bar at the end of every day. It should be done in a way that you make the fewest number of cuts in the one big gold bar that you have that will allow you to pay him 1\/5th each day? The fewest number of cuts would be 2. Make a cut in a way that there are three pieces — one of 1 unit and two of 2 units. On day 1: Give him 1 unit. On day 2: Give him 2 units and take back 1 unit. On day 3: Give him 1 unit (he already has 2 units). On day 4: Give him 2 units and take back 1 unit (he already has 2 units). On day 5: Give him 1 unit (he already has two 2 units). 5| A bus has 100 labelled seats from 1 to 100. There are 100 people standing in a queue from 1 to 100, and people board the bus in the sequence of 1 to n. The rule is, if a person ‘a’ boards the bus, he checks if seat ‘a’ is empty. If the seat is empty, he sits there, else he randomly picks an empty seat to sit. Given that the first person picks a seat randomly, find the probability that the 100th person sits in his place, i.e. seat number 100. The last person to board will have a seat determined the moment either the first or the last seat is selected. This is because the last person will either get the first seat or the last seat. Any other seat will be taken by the time the last person gets to make a selection of the seat. As each person boards the bus, the probability of the first or last seat being taken is equal. This means that the last person will get either the first or last seat with equal probability. Therefore, the probability of the last person getting onto seat no. 100 is 1\/2. 6| There are 5 jars of pills. Each pill weighs 10 grams, except for contaminated pills contained in one jar, where each pill weighs 9 grams. Given a scale, how could you tell which jar had the contaminated pills in just one measurement? Take out 1 pill from jar 1, 2 pills from jar 2, 3 pills from jar 3, 4 pills from jar 4 and 5 pills from jar 5. Put all these 15 pills on the scale. The correct weight should be 15*10 = 150 But, since one of the jars has contaminated pills, the weight will definitely be less than 150. If the weight is 149, then jar 1 has contaminated pills because there is only one pill taken from jar 1. If the weight is 148 then it is jar 2, if the weight is 147 then it is jar 3, if the weight is 146 then jar 4 is contaminated and, if the weight is 145, then jar 5 has contaminated pills. 7| Three ants are sitting at the three corners of an equilateral triangle. Each ant starts randomly picking a direction and starts to move along the edge of the triangle. What is the probability that none of the ants collides? For the ants to avoid a collision, they must move in one direction — either clockwise or anticlockwise. If they do not move in one direction, there will definitely be a collision. Each ant has the option to either move clockwise or anti-clockwise. There is a one in two chance that an ant decides to pick a particular direction. The probability of no collision will therefore be, N (No collision) = N (All ants go in a clockwise direction) + N (All ants go in an anti-clockwise direction) = 0.5 * 0.5 * 0.5 + 0.5 * 0.5 * 0.5 = 0.25 8| There are 10 bags full of coins, with infinite coins in each bag. But one bag is full of forgeries, and you can’t remember which one. But you do know that genuine coins weigh 1 gram and forgeries weigh 1.1 grams. You have to identify that bag in minimum readings. You are provided with a digital weighing machine. Building upon the same approach as the solution in question 6 above, we have to first take 1 coin from the first bag, 2 coins from the second bag, 3 coins from the third bag and so on till 10 coins from the 10th bag are taken out. The total number of coins taken out will be 1 + 2 + 3…+10 = 55 coins. Now, weigh the 55 coins together. Depending on the reading, the bag with the forged coin can be found out. For instance, if the reading ends with 0.4, it is the fourth bag, if it is 0.8, then it is the 8th bag, and so on. 9| A snail is at the bottom of a 30-foot well. Every hour the snail can climb up 3 feet and then it slides back down 2 feet. How many hours will it take for the snail to get out of the well? Every 1 hour, the snail is climbing 3 feet and sliding down 2 feet. It is therefore covering a distance of 3 – 2 = 1 feet every hour. Now, once the snail reached the top, it will not slide back. So, the position from where it will reach the top in 1 hour would be: 30 feet – 3 feet (upward distance covered in 1 hr) = 27th feet. It means that the snail will not slide down when it starts climbing from 27th feet. Now, as the snail can climb 1 foot every hour, it can climb 27 feet in 27 hours. The last 3 feet will be covered in 1 hour from 27th feet. Hence, the time taken to climb to the top of the well is, 27 hours + 1 hour = 28 hours. 10| If a hen and a half lay an egg and a half in a day and a half then how many hens will it take to lay six eggs in six days? Since, 1.5 eggs in 1.5 days require 1.5 hens 1 egg in 1.5 days will require 1.5\/1.5 = 1 hen 1 egg in 1 day will require 1 * 1.5 = 1.5 hens 6 eggs in 1 day will require 1.5 * 6 hens 6 eggs in 6 days will require 1.5 * 6 \/ 6 = 1.5 hens","excerpt":"The ongoing pandemic has affected hiring in many companies across data science domains. Even if companies are recruiting, they are adopting stringent procedures to spot the right candidate. Puzzles are often preferred to evaluate a candidate as it best accesses the critical thinking ability, problem-solving and in many cases, even the coding skills of a […]","categories":["AI Highlights"],"tags":["what is big data coding"],"author_name":"Srishti Deoras","publish_date":"2020-10-15T10:00:29","publication_year":"2020","word_count":1616,"keywords":["data science","Go","programming_languages:R","AI","R","ML","programming_languages:Go","Git","analytics","what is big data coding"],"extracted_tech_keywords":["AI","ML","data science","analytics","R","Go","Git","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/interesting-puzzles-to-prepare-for-data-science-interview\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10005596,"title":"How Neuralink’s Human Trials Can Be Beneficial For Humanity","content":"In an attempt to showcase Neuralink’s progress, Elon Musk has recently tweeted about the live stream demo that the company is planning to set on the 28th of this month. In fact, over the years, Musk has revealed snippets of information about the mysterious company with his tweets, where he posted updates of the mission statement and has also shared information about the technology. However, with this recent information in hand, it is believed that the promising brain chip company might end up revealing the human trials for the technology. Will show neurons firing in real-time on August 28th. The matrix in the matrix.— Elon Musk (@elonmusk) July 30, 2020 The company, last year, has already shared the 87% accuracy results of trying out this technology on monkeys and mice, and since then the wait has started for the human trials of the technology. As a matter of fact, the company was expected to try out their human experiments at the beginning of this year; however, it is happening later than never. A lot of this could be attributed to the exceptional vision of the company that merges AI with human brains to solve some of the complicated brain disorders. Nevertheless, the primary point is that the company is creating a bridge to facilitate direct communication between human brains and machines — an AI symbiosis. And with Musk’s claim of demonstrating “neurons firing” it can have great potential in bringing out the features of this technology. Also Read: Is Elon Musk Capable Of Building Cyborgs? Human Trials Can Prove Neuralink’s Potential While the brain chip company delayed its update on human trials, it studied the mood and emotion patterns of humans. It highlighted its mood-altering functions that can help in reducing anxiety in humans. However, this again needs to be tested on humans for accomplishing better results. Such capabilities of Neuralink makes it the only solution possible for humans to regain their mobility after an accident, paralysis or any sort of disability. It also turns out to be the only key solution of the current era that can help humans to track signs of tumours as well as deal with major neurological disorders. However, Musk thinks otherwise. Teasing with the progress of the brain interface company, Musk stated in an interview — the single vision of Neuralink’s chip is to achieve a symbiosis with artificial intelligence. According to Musk, the company is going to leverage the chip along with the micro threads that will be surgically planted inside humans brains with high bandwidth data transfer will enable humans to perform tasks that are beyond capabilities. While the details aren’t clear, this can be facilitated by the learning algorithm inside the chip that will allow humans to do extraordinary stuff — like listening to music directly from the chip. Currently, Musk is actively pushing the technology to get regulated as brain-computer interface devices in the healthcare industry; however, it still would require human trials to make conclusive statements about the technology. Also Read: Can This New Biosynthetic Polymer Boost Up Neuralink’s Cyborg Technology? Having said that, Musk is known for making ambitious statements with failed predictions. But experts are hoping that isn’t the case with Neuralink’s progress as it not only will enhance human brains but will benefit humanity as a whole — creating the possibility of transhumanism. In fact, it is believed that Neuralink can help with many disorders that highlight only anecdotal evidence like body pain or headaches, where the technology can directly communicate with the brain to figure out the problem. In spite of such excitement, it is true that the concept of brain interface machines isn’t novel, as many researchers have previously worked around this concept to communicate with the brain for medical treatment. But it is probably the first time when a commercial organisation backed by a billionaire is working on a brain interface technology for common people. Although it seems like a farfetched dream, Musk is a strong advocate of humans turning into a cyborg with their strong integration with phones 24×7. The company is only working on making this technology viable for humans, like architecture and software. Thus bringing it down to humans is critical for this technology to bring out better performance. Another interesting part for the company to work this technology on humans is the massive amount of data that can be gathered. With the capability of translating thoughts into information, this brain chip could continuously gather data, which can later be used for precision diagnosing and treating complex diseases. In all honesty, the possibilities are infinite. Also Read: Netflix’s ‘Love, Death & Robots’ Looks Into The Future Of Elon Musk’s Neuralink Wrapping Up While the potential of the brain chip company — Neuralink — is exceptionally noteworthy, it is the coming up demonstration and update by the company that will provide more information about its progress. The presentation, on the 28th of August, by Musk, will showcase the possibility of the company to help humanity as a whole. The Neuralink event will be held on the 28th of August, 2020 on a live stream.","excerpt":"In an attempt to showcase Neuralink’s progress, Elon Musk has recently tweeted about the live stream demo that the company is planning to set on the 28th of this month. In fact, over the years, Musk has revealed snippets of information about the mysterious company with his tweets, where he posted updates of the mission […]","categories":["AI Features"],"tags":["Elon Musk","Elon Musk Neuralink","neuralink"],"author_name":"Sejuti Das","publish_date":"2020-08-27T10:00:00","publication_year":"2020","word_count":853,"keywords":["Go","artificial intelligence","programming_languages:R","AI","neuralink","programming_languages:Go","RAG","Elon Musk","Aim","GAN","Elon Musk Neuralink","R"],"extracted_tech_keywords":["AI","artificial intelligence","Aim","RAG","R","Go","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-neuralinks-human-trials-can-be-beneficial-for-humanity\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":38832,"title":"10 Fastest-Growing Open Source Projects On GitHub You Can Contribute To","content":"According to the State of the Octoverse report by GitHub, the community of GitHub found trends in the growth of projects related to various topics such as machine learning, gaming, 3D printing, home automation, data analysis, full-stack JavaScript development, and scientific programming. The information on this article has been cited from the original documentation and the sources are also cited inside. In this article, we list down the 10 fastest-growing open source projects on GitHub. 1| Azure-Docs The platform has seen 4.7 times growth within one year. Microsoft Azure is an open, flexible, enterprise-grade cloud computing platform for building, testing, deploying and managing applications through Microsoft-managed data centres. It provides software as a service (SaaS), platform as a service (PaaS)and infrastructure as a service (IaaS) and supports many different programming languages, tools, and frameworks, including both Microsoft-specific and third-party software and systems. Click here to know more. 2| Pytorch PyTorch is a Python package which provides two high-level features, one is Tensor computation (like NumPy) with strong GPU acceleration and the other is deep neural networks built on a tape-based autograd system. This package provides Tensors which can live either on the CPU or the GPU  and accelerates the computation by a huge amount. It is designed to be intuitive, linear in thought and easy to use. Click here to know more. 3| Godot Godot Engine is a feature-packed, cross-platform game engine to create 2D and 3D games from a unified interface. It provides a comprehensive set of common tools so that users can focus on making games without having to reinvent the wheel. It is completely free and open source under the MIT license. Click here to know more. 4| Nuxt.js Nuxt.js is the Vue.js meta-framework to create complex, fast and universal web applications quickly. It basically presets all the configuration needed to make your development of a Vue.js application pleasant. With Server Side Rendered, also called “universal” or “isomorphic” mode, a Node.js server will be used to deliver HTML based on your Vue components to the client instead of the pure javascript. Click here to know more. 5| Go-Ethereum Ethereum is a decentralised platform that runs smart contracts, applications that run exactly as programmed without possibility of downtime, censorship, fraud or third-party interference. Go Ethereum is one of the three original implementations (along with C++ and Python) of the Ethereum protocol. It is written in Go, fully open source and licensed under the GNU LGPL v3. It is available either as a standalone client called Geth that you can install on pretty much any operating system or as a library that you can embed in your Go, Android or iOS projects. Click here to know more. 6| React-Native-Navigation React-native-navigation is a native navigation library which provides native platform navigation for both iOS and Android for React Native applications. This library provides a way to build mobile applications with the React framework. Click here to know more. 7| Spyder Scientific Python Development Environment (Spyder) is a powerful scientific environment written in Python, for Python, and designed by and for scientists, engineers and data analysts. It offers a unique combination of the advanced editing, analysis, debugging, and profiling functionality of a comprehensive development tool with the data exploration, interactive execution, deep inspection, and beautiful visualization capabilities of a scientific package. It can also be used as a PyQt5 extension library, allowing you to build upon its functionality and embed its components, such as the interactive console, in your own software. Click here to know more. 8| TensorFlow-Models TensorFlow is an end-to-end open source platform for machine learning. It has a comprehensive, flexible ecosystem of tools, libraries and community resources that lets researchers push the state-of-the-art in ML and developers easily build and deploy ML powered applications. The TensorFlow Models repo in GitHub contains a number of different models implemented in Tensorflow. Click here to know more. 9| Home-Assistant Home Assistant is a home automation platform running on Python 3. It is an open source home automation that puts local control and privacy first. It is able to track and control all devices at home and offer a platform for automating control. The system is built using a modular approach so support for other devices or actions can be implemented easily. Click here to know more. 10| Marlin Marlin is the world’s most popular open source firmware for Replicating Rapid Prototyper (RepRap) machines, commonly referred to as “3D printers.” Marlin Firmware is highly efficient, running even on modest 16MHz embedded AVR processors. It is first created in 2011 for RepRap and Ultimaker by Erik van der Zalm. It is an open source project hosted on GitHub and is owned and maintained by the maker community. Click here to know more.","excerpt":"According to the State of the Octoverse report by GitHub, the community of GitHub found trends in the growth of projects related to various topics such as machine learning, gaming, 3D printing, home automation, data analysis, full-stack JavaScript development, and scientific programming. The information on this article has been cited from the original documentation and […]","categories":["AI Trends"],"tags":["full stack project ideas","github projects","Pytorch","react-native","spyder"],"author_name":"Ambika Choudhury","publish_date":"2019-05-09T07:39:13","publication_year":"2019","word_count":789,"keywords":["Pytorch","NumPy","machine learning","AI","neural network","TensorFlow","ML","spyder","PyTorch","cloud computing","Python","react-native","github projects","Azure","full stack project ideas"],"extracted_tech_keywords":["AI","machine learning","ML","neural network","TensorFlow","PyTorch","NumPy","cloud computing","Azure","Python"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/10-fastest-growing-open-source-projects-on-github-you-can-contribute-to\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":52521,"title":"Facebook’s New OS Will Do So Much More Than Just Break Up With Android","content":"Undeterred from its previous mistakes, Facebook has now come up with a new announcement. With a history of several cyber privacy breaches, and one almost failed attempt of launching a cryptocurrency’ project, Facebook has now planned to build its own operating system from scratch. This project is aimed at reducing the company’s reliance on Google’s Open Source project — Android. The development of this in-house software is currently being led by Mark Lucovsky — a Microsoft Veteran, who has co-authored the Windows NT OS. Facebook Hardware The OS is going to power Facebook’s hardware, which was so far running on Google’s OS — Android, especially the augmented reality glasses — codenamed Orion, the virtual reality headsets — Oculus, and the video calling device — Portal. The company also believes that with its own OS, it will be able to integrate more innovative and ingenious features on to its devices, without the mercy of Google or Apple. However, just to be clear, Facebook’s smartphone apps will still be running on Android. According to one of Facebook’s AR and VR heads, Ficus Kirkpatrick, “it’s possible” that Facebook’s future hardware won’t need to rely on Google’s software, which would entirely eliminate the control Google has over Facebook’s hardware. Andrew Bosworth, head of Facebook’ hardware department further added, “We really want to make sure the next generation has space for us, and we really can’t trust the marketplace or the competitors to ensure that’s the case. And so we’re gonna do it ourselves.” Another important aspect, apart from the freedom to bake social interaction, Facebooks is also trying to achieve the privacy concern, with its new OS, which has been a long term goal for the company. Such a movement by this social media giant could prevent its discrepancies with Google on derailing the roadmaps of its gadgets. It is to be believed that Facebooks has always been enraged of not owning its own OS and having a dependency on its biggest rivals — Google and Apple, and therefore this move was specially planned to strengthen its AR \/ VR product space. This, in turn, will also ascend its acquisitions, such as Instagram. The company is currently exploring options to collaborate with companies for building its custom OS. In the attempt of self-branding itself as well as ramping up its efforts in the hardware space like Apple, Facebook is now by building up a 770,000 sq.ft. new office with labs, prototype space and testing areas, for its AR\/VR team of approximately 4,000 employees, in Burlingame, California. This space will be aimed at making it easier for the public to come and play with its AR\/ VR products, and probably later buy it. Facebook has been aiming at improving its hardware experience for the enterprise as well. And, therefore, the company will now be using Portal to run its Workplace video calls. Additionally, the media reports that the company has been working on its own custom chip hardware, alongside the voice assistant development. Critics Experts of the industry believe that this farfetched delusional dream may help the company in building up its name once again in the market. Anshul Gupta, Senior Research Director at Gartner mentioned that Facebook’s main focus is its AR\/ VR connected devices, and therefore its OS is definitely not meant to envelop the smartphone space. He said, “Considering, the IoT connected device space is a huge market, therefor Facebook is currently aiming to build its OS for that, not to overtake Google by getting into the game of smartphone’s OS market. As a matter of fact, the company is still in its early stage of connected devices, and to get the real view, we will have to wait for its OS dream to come to a realization. Another expert from Forrester, Ashutosh Sharma, also seconded with the fact that it is still too early to definitely comment on Facebook because the details of the OS are not out yet. However, he further added, “Google’s OS is backed by an ecosystem which is too deeply ingrained in our lives and it is not going away anywhere soon. And, with Facebook’s poor reputation on privacy, the company will have a tough time to convince people to come on-board their OS.” Well, second time’s a charm. Facebook has already made a previous attempt of owning the software experience, by collaborating with HTC, back in 2013. It attempted in creating Facebook-based UI layer called ’Facebook Home’ — an operating system which was sitting on top of Google’s OS. However, the attempt severely failed as it wasn’t well-received by the public or the enterprise. And, the emergence of the company’s cyber privacy scandals made it even worse. Outlook With the recent news of breach and scandals at place, Facebook has been considered as a ‘bad boy’ of the tech industry, which has been going under several scrutinies of the public. Although Instagram and Oculus are still owning the market as big names, researchers are in a fear that as the social media giant sets its journey on building the new OS, privacy concern and political involvement will act as the biggest hurdle and can even destruct its vision of getting an individuality from its rivals. Despite the limelight scandals, the company still nestles thousands of users, and currently own everything that is required to build an OS — support of developers, users loyalty to populate the space, and the technical platform. There will be technical hurdles to overcome, as making an OS is no joke, but Facebook will be playing with its strengths, which is its apps and social media experience which is still the biggest in the industry.","excerpt":"Undeterred from its previous mistakes, Facebook has now come up with a new announcement. With a history of several cyber privacy breaches, and one almost failed attempt of launching a cryptocurrency’ project, Facebook has now planned to build its own operating system from scratch. This project is aimed at reducing the company’s reliance on Google’s […]","categories":["Global Tech"],"tags":["Facebook","Facebook AI","facebook analytics","Facebook outage","operating system"],"author_name":"Sejuti Das","publish_date":"2019-12-24T16:08:53","publication_year":"2019","word_count":943,"keywords":["Go","Facebook AI","AI","programming_languages:R","operating system","programming_languages:Go","Facebook outage","RAG","Aim","Rust","Facebook","facebook analytics","R","programming_languages:Rust"],"extracted_tech_keywords":["AI","Aim","RAG","R","Go","Rust","programming_languages:R","programming_languages:Go","programming_languages:Rust"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/facebooks-new-os-will-do-so-much-more-than-just-break-up-with-android\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":41307,"title":"Can Singapore Startup Musiio Rival Spotify With Its Push For Personalisation?","content":"The turn of this century witnessed a new form of intelligence — the augmentation of human idea with the computational powers of the machines. These machines have now have become massive data-driven engines. Machine Learning models like neural networks ingest tonnes of data and churns out results without getting tired. These algorithms have enabled us to see the hidden correlations, which in turn have been exploited for building better models. From recommending designs for nuclear plants to soundtracks, AI has been leveraged in almost all fields. Music, one of the most celebrated creative traits amongst humans, also has been touched by the far reaches of algorithmic advancements. This resulted in the development of personalised song recommender systems to cater to one’s own moods. One such service, Spotify, which has been topping the charts as one of the world’s topmost music streaming providers. Spotify’s personalisation system got better as it got supplemented with extra information such as a soundtrack’s danceability, energy, valence, and more. But the machine learning team at this Singapore based start-up Musiio argues that the inefficiencies of the present AI assisted services still prevail. Musiio boasts of having an AI that can ‘listen’ to millions of tracks at once, recognise thousands of features from every single audio track and categorise music with an accuracy greater than 90%. These results have landed big for Musiio as it becomes the first venture capital-backed music AI startup in Southeast Asia. Why Musiio Can Be A Good Alternative To Spotify Musiio aims at reducing inefficiencies using AI in a way such that users do not have to sludge through Spotify to find their own flavour of music. Moreover, businesses who use Musiio will have the option of automating this search process. Musiio leverages AI to enable: Efficient audio tagging Increased search speed Improved catalogue recall accuracy to fetch hidden gems of music. These functionalities along with many others give Musiio the much-needed edge and this was made possible thanks to the state-of-the-art AI models, which can: ‘Listen’ to and tag over 200,000 new tracks per day. Once ‘heard’ our technology can search through music at a rate of 1m+ tracks in two seconds. Listen to the audio rather than using a popularity algorithm to match users’ audio profile with music they’ll love, getting the best out of the whole catalogue and creating a personalised experience, not an echo chamber. Can be custom-trained to work from a cold start. In addition to these features, tag packages are available or customers can request a custom build to fit their personal database. Users get an improved search experience with enhanced capabilities that allow for audio-reference searching. Their catalogue recall accuracy has been improved to find unique music that fits with your chosen track with less effort. They have also improved their search speed where users can search for more than one million tracks in just below 2 seconds. By pasting an audio link, an mp3, or a music track into the search tool, users now no longer need to find the words to describe the track they are looking for. Instead, they are able to use a ‘reference track’ or a seed track and can spend time enjoying accurate audio matches rather than being caught in a loop of inaccurate keywords and poor search results. Here are few results when for few songs uploaded from Youtube into the search tool: For Eminem’s I’m not afraid soundtrack, the results were as follows: And Vivaldi’s four seasons(Storm): For Dmitiri Shostakovich’s Waltz No.2 : Here the percentage represents AI’s predicted probability about the tag. These predictions are quite accurate as one can see the clear contrast of tags between Vivaldi’s adrenaline pumping Storm composition and the melodious Waltz. Future Of Musiio The use of AI to disrupt the music industry has been gaining traction of late. Earlier this year, Google demonstrated how to shred notes into lower dimensions and then perform fundamental techniques like batch normalisation and autoregressive factorisation to create new soundtracks from old ones. Though using the word disruption in domains steered by human creativity is still a hyperbole, one can still sense how AI can be used as an augmentation to creative tasks. To create and share the work easily; making it accessible to the wider population is what any artist wants. And, AI-driven technologies like those of Musiio’s will keep trying to fill the voids left out by technical impediments in the past. Find your favourite music with Musiio here.","excerpt":"The turn of this century witnessed a new form of intelligence — the augmentation of human idea with the computational powers of the machines. These machines have now have become massive data-driven engines.  Machine Learning models like neural networks ingest tonnes of data and churns out results without getting tired. These algorithms have enabled us […]","categories":["Deep Tech"],"tags":["recommendation system","spotify"],"author_name":"Ram Sagar","publish_date":"2019-06-25T12:08:39","publication_year":"2019","word_count":746,"keywords":["Go","API","machine learning","recommendation system","AI","neural network","data-driven","RAG","Aim","spotify","ViT","R"],"extracted_tech_keywords":["AI","machine learning","neural network","Aim","RAG","R","Go","API","ViT","data-driven"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/can-singapore-startup-musiio-rival-spotify-with-its-push-for-personalisation\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10018685,"title":"How Ludo King Blitzscaled During Pandemic, With A Little Help From MongoDB","content":"In 2015, Gametion, the maker of Ludo King, was still taking baby steps into the world of online gaming, with a couple of developers in a small office. The vision was to produce a suite of in-browser Flash games, before positioning itself at the front of the mobile gaming sector. Gametion’s most significant breakthrough came with Ludo King, a digital version of the classic Ludo board game much loved worldwide. The Gametion team knew they were on to something, but no one could have predicted just how big it would get. However, due to the massive COVID lockdown in 2020, Ludo King became a global phenom, crossing 500 million downloads at 400% growth. As demand soared, more players joined the game demanding new features and integrations, forcing Gametion to scale their business. But there was one problem. Gametion had no database component for its Flash games (no personal logins or user IDs). Thus, it became critical for the company to build their games on a database platform that is scalable and responsive to changing needs. This made Gametion turn to MongoDB, a cross-platform document-oriented database for modern apps. To understand the case better, Analytics India Magazine got in touch with Vikash Jaiswal, the Founder & CEO of Gametion Technologies and Suvig Sharma, the Senior Director, APAC at MongoDB. Also Read: After Poker and Go, Researchers Use New Game To Evaluate RL Agents’ Robustness Paving The Way Ludo King quickly adapted and incorporated new technologies. However, as the game exploded, the team of developers required a solution that was intuitive and scalable. They didn’t have time for complexity or long learning curves. “We are at a stage where our games, Ludo King, in particular, are widely recognised and are in high demand. We are under pressure to continue to improve game features and deliver new creations that users love to play,” said Jaiswal. “MongoDB has been the perfect partner in this regard, with its cloud database offering MongoDB Atlas, excellent mobile facility in MongoDB Realm and an ever-evolving offering with Atlas Data Lake.” Gametion was determined to up the ante while maintaining high standards, creativity and versatility that got the company to this stage. Explaining the process, Sharma stated, with MongoDB, a fundamentally different model from the traditional relational databases, Gametion’s data has been stored in documents for developers to read and use. Developers no longer had to ensure their applications accommodate the database. MongoDB accommodates it for the applications to store data in a much more natural way, making it easier and quicker to build and deploy. That said, Gametion had no interest in understanding the nuts and bolts of database administration and management. Instead, the team wanted to spend their time programming, developing and releasing applications. That’s when it opted for MongoDB’s managed service — MongoDB Atlas for database management. A fully managed MongoDB service, MongoDB Atlas lets developers deploy across AWS, Google Cloud and Azure with best-in-class automation and proven practices that guarantee availability, scalability, and compliance. Along with this, Atlas comes with the most demanding data security and privacy standards. “Once onboarded with MongoDB Atlas, running on AWS’s cloud infrastructure, Gametion, which was now five developers strong, quickly found the features that mattered and got to work,” said Sharma. Backend infrastructure Gametion stuck with common open standards and a component approach that made it easy to add other functions over time as the game demands, maintaining a regular schedule of updates to keep the users engaged — like a microservice architecture. The company used Kafka for data movement and to synchronise between services. “It’s another way to optimise resources without sacrificing scalability or release cadence,” added Jaiswal. Also Read: Which GPU Provider Will Capture The Cloud Gaming Service Sector? The Result With MongoDB’s flexible data model and managed service, Ludo King kept up with the soaring demand and expanded its features with more options on mobile, multiplayer experiences and monetisation tactics. “This is because the developers have been able to focus on developing and not managing databases,” said Sharma. MongoDB Atlas helped developers scale seamlessly and cater to the massive spike in users and traffic without impacting the performance. By August 2020, Ludo King had become India’s first and world’s third most downloaded game on Google Play. With Ludo King, Gametion won the ET Startup Award in 2020 as a Bootstrap Champ for entrepreneurial achievement. At present, the company is efficiently managing hundreds of millions of users, alongside a huge 1,000% swell in growth driven by a massive exodus to digital gaming in the wake of the pandemic.","excerpt":"In 2015, Gametion, the maker of Ludo King, was still taking baby steps into the world of online gaming, with a couple of developers in a small office. The vision was to produce a suite of in-browser Flash games, before positioning itself at the front of the mobile gaming sector. Gametion’s most significant breakthrough came […]","categories":["IT Services"],"tags":[],"author_name":"Sejuti Das","publish_date":"2021-01-22T18:00:00","publication_year":"2021","word_count":761,"keywords":["Go","AWS","AI","MongoDB","ML","R","Scala","analytics","Kafka","Azure"],"extracted_tech_keywords":["AI","ML","analytics","AWS","Azure","Kafka","MongoDB","R","Go","Scala"],"url":"https:\/\/analyticsindiamag.com\/it-services\/how-ludo-king-blitzscaled-during-pandemic-with-a-little-help-from-mongodb\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10161174,"title":"Oracle Beats AWS, Azure on their Home Turf","content":"Oracle, a leader in cloud technology, has introduced the Oracle Exadata X11M, the newest version of its Exadata platform. According to the company, this model offers up to 55% quicker AI vector searches, 2.2 times faster analytics scan throughput, and 25% improved transaction processing, all while maintaining the same pricing as its predecessor. “Exadata X11M is up to 70 times faster due to the unique hardware offloaded RDMA compared to AWS RDS and Azure SQL. Now, with multi-cloud, we’re faster than AWS in AWS and faster than Azure in Azure,” said Kothanda Umamageswaran, Oracle SVP of Exadata and scale-out technologies. He explained that Exadata offers lower latency for OLTP database operations, handling tasks faster than AWS RDS and Azure SQL, which take around 1 millisecond compared to Exadata’s mere 14 microseconds. Oracle’s multi-cloud strategy allows them to deploy their Exadata technology on competing cloud platforms, claiming superior performance even in those environments, including on AWS and Azure infrastructures. It challenges these providers in their own ecosystems. Currently, Oracle Exadata X11M is available on-premises, with Exadata Database Service, Autonomous Database on Exadata Cloud@Customer, Oracle Cloud Infrastructure (OCI), and in multi-cloud setups including AWS, Google Cloud, and Microsoft Azure. In an exclusive interview with AIM, Ashish Ray, Oracle’s VP of mission-critical database technologies, revealed that their customer base largely consists of on-premises clients, followed by those using Oracle Cloud, with multi-cloud customers ranking next. “The Google and AWS announcements are fairly recent, so naturally, this multi-cloud play is not at the same scale as our current on-premises and Oracle Cloud offerings,” Ray said. However, the company anticipates significant growth in this multi-cloud initiative, including Azure, with the hope that clients using other platforms will move to its high-performance database. “The walled gardens have come down, and business productivity has become much more transparent. There are no bottlenecks, no setbacks, and now the choice is back to the customers,” said Ray. He described India as a data-driven market experiencing explosive growth, fuelled by the digital modernisation of businesses—from large enterprises to a vibrant startup ecosystem. Ray expects that in India, Exadata X11 will act as a catalyst for further business innovation and IT modernisation. Furthermore, he added that migrating to Exadata X11M provides better value for businesses. “With Exadata X11M, companies can handle more workloads with fewer servers, which helps reduce costs related to data centres, power usage, space, and cooling, especially for those using on-premises setups.” Meanwhile, Umamageswaran notes that companies requiring access to vast amounts of data from multiple sources and data types, say, graph, spatial, documents, transactional, and analytical data, will benefit from Oracle Databases 23ai on Exadata X11M. The company claims that Exadata handles transactions in microseconds and also supports real-time analytics like anti-money laundering (AML) checks that must occur simultaneously with the transaction process. Database Competition Galore The catch with Oracle Exadata is that it is only compatible with Oracle databases such as Oracle Database 23ai. But today, customers are experimenting with different database platforms such as Amazon Aurora, Cosmos DB, Google Spanner, MongoDB, CockroachDB, and many more. “There will always be individual data management solutions, whether it’s PostgreSQL, Aurora, Redshift, or Cosmos DB,” Ray added. Much like Oracle, CockroachDB is a distributed SQL database designed for scalability, consistency, and resilience. It combines the best features of traditional relational databases with the advantages of modern distributed systems. Spencer Kimball, CEO of Cockroach Labs, in the latest episode of Tech Talks at AIM, critiqued the dependence on cloud vendors like AWS, Google Cloud, and Oracle for databases, emphasising the limitations of vendor lock-in. He stated: “With CockroachDB, we let customers run across clouds, in private data centers, or both. You can even turn off a cloud vendor entirely and keep running without downtime. That’s true freedom, and it’s what enterprises need today.” To this, Ray said: “A lot of these companies tend to be SQL-compatible. We do have customers who run CockroachDB and PostgreSQL,” adding that the scale at which Oracle Database runs and supports large enterprises across the entire globe is absolutely unheard of. “That’s why you see a lot of multi-cloud interest from the cloud hyperscalers,” he added. From a generative AI perspective, he highlighted that Exadata significantly accelerates tasks like vector processing, index creation, and querying. However, he did acknowledge the growing popularity of PostgreSQL due to its open-source nature. Oracle Database doesn’t directly support PostgreSQL, but Oracle offers a service called OCI Database with PostgreSQL. This service lets users use PostgreSQL in the Oracle Cloud, giving them PostgreSQL features along with Oracle’s cloud benefits. AMD is EPYC: Oracle launched Exadata in 2008. Ray explained that while the Oracle Exadata X9M was based on Intel processors, the new Exadata X11M uses AMD EPYC processors. “The reason for moving to AMD was that it was just available with numbers that can drive a lot of parallel throughput for database workloads,” said Ray. Interestingly, NVIDIA also recently partnered with AMD. AMD’s EPYC CPUs are critical in powering NVIDIA’s GPUs for large-scale AI workloads. “We’ve shown a 20% improvement in training and a 15% improvement in inference when connecting EPYC CPUs to NVIDIA’s H100 GPUs,” said Ravi Kuppuswamy, senior vice president & general manager at AMD.","excerpt":"Exadata delivers faster OLTP database operations with 14 microseconds latency, compared to 1 millisecond on AWS RDS and Azure SQL.","categories":["Global Tech"],"tags":["Oracle"],"author_name":"Siddharth Jindal","publish_date":"2025-01-11T12:48:55","publication_year":"2025","word_count":867,"keywords":["PostgreSQL","AWS","AI","MongoDB","ML","Oracle","Ray","Aim","generative AI","analytics","Azure"],"extracted_tech_keywords":["AI","ML","analytics","generative AI","Aim","Ray","AWS","Azure","MongoDB","PostgreSQL"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/oracle-beats-aws-azure-on-their-home-turf\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10163484,"title":"OpenAI&#8217;s o3 Wins Gold at International Olympiad in Informatics 2024","content":"New research from OpenAI highlights the results of their reasoning models (o-series models) and how LLMs have evolved from amateur competitive programmers to competing with the world’s best. OpenAI’s latest AI model, o3, earned an impressive 2724 rating on CodeForces, placing it in the 99.8th percentile. It also secured a gold medal-level score at the 2024 International Olympiad in Informatics (IOI). According to the research, o3 outperforms the o1-ioi model, which is specifically fine-tuned for IOI. This proves that reinforcement learning is more effective than hand-crafted approaches. At IOI 2024, o3 competed under standard conditions and crossed the gold medal threshold. On CodeForces, it ranked among the top 200 programmers globally, competing with elite human coders. “General-purpose reasoning capabilities developed through reinforcement learning are now outperforming carefully hand-crafted, domain-specific solutions,” said Ethan Mollick, associate professor at The Wharton School. “Rather than building specialised systems for specific tasks, large, general-purpose models can achieve superior results through better reasoning abilities.” The research is part of OpenAI’s ongoing efforts to assess its models’ performance in competitive programming and broader software engineering. Anthropic, the company behind the Claude model series, also released a report on Monday which highlighted AI’s influence on the workplace. The findings revealed that approximately 36% of all occupations incorporate AI for at least a quarter of their tasks. Moreover, 57% of AI applications enhance human capabilities, while 43% focus on automation. However, only 4% of occupations rely on AI for at least 75% of their tasks. The study identified software development and technical writing as the primary areas where AI is utilised. In contrast, AI plays a minimal role in tasks that involve physical interaction with the environment.","excerpt":"The research highlights that more training and test-time compute improves model performance, nearing top human levels.","categories":["AI News"],"tags":["OpenAI"],"author_name":"Aditi Suresh","publish_date":"2025-02-13T11:53:25","publication_year":"2025","word_count":278,"keywords":["Go","Anthropic","OpenAI","AI","programming_languages:R","programming_languages:Go","llm_models:Claude","automation","R"],"extracted_tech_keywords":["AI","OpenAI","Anthropic","R","Go","automation","llm_models:Claude","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/openais-o3-wins-gold-at-international-olympiad-in-informatics-2024\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10120161,"title":"Sarvam AI Launches AI Residency Program Offering Up to INR 1 Lakh Monthly Salary","content":"Indian AI Startup, Sarvam AI is seeking talented Data Engineers and ML Engineers to join its team to build state-of-the-art ML systems for Speech Recognition and Text-to-Speech applications in Indian languages. Register > The company is focused on creating full-stack GenAI systems and applications tailored for India, with a specific focus on Indian languages. The startup is offering two full-time job opportunities: the Summer Internship and the AI Residency. Candidates can apply here. The Summer Internship is ideal for freshers or students with a foundational grasp of ML and programming, while the AI Residency is perfect for those with professional experience or significant expertise. As a Data Engineer, responsibilities will include web scraping, managing distributed data processing, and developing robust data pipelines. As an ML Engineer, responsibilities will include training, monitoring, and evaluating state-of-the-art speech models. The company is offering competitive salaries for the positions. The Summer Internship offers a stipend of up to 50,000 per month, while the AI Residency offers a salary of up to 1 lakh per month. Sarvam AI is a well-funded GenAI startup focused on creating full-stack GenAI systems and applications tailored for India. Based out of Bangalore and Chennai, Sarvam AI is a place for those who are passionate about driving significant advancements through Generative AI, have a love for Indian languages, and are eager to make a substantial impact. Earlier this year, Sarvam AI partnered with Microsoft, announcing plans to build an Indic voice LLM. The partnership aims to release the voice LLM in the coming months, which will enable users to interact with AI systems using voice commands. “We believe that in India, people will experience generative AI through the medium of voice,” said Vivek Raghavan cofounder, Sarvam AI, in an exclusive interview with AIM. He added that it is very hard to input text in Indian languages and that in India, people tend to prefer voice communication over text. “We want people to do things through voice and that will be the USP of Sarvam AI.” he said. The company is also working on building agentic systems, allowing users to not only receive information but also take action. “I hope in the next few months we’ll see some of these things being announced and released in the marketplace,” said Raghavan. Sarvam AI will support 10 languages and hopes to expand it further in the future. The company’s focus on voice-based interfaces has numerous practical applications in the country, such as in customer support and gathering feedback, where voice-based models can efficiently handle large-scale feedback listening. Sarvam AI was founded in 2023 by Pratyush Kumar and Raghavan. Last year, the company raised a total of $41M in a Series A funding round led by Lightspeed Venture Partners, with participation from Khosla Ventures and Peak XV Partners.","excerpt":"The startup is offering two full-time job opportunities: the Summer Internship and the AI Residency.","categories":["AI News"],"tags":["sarvam ai"],"author_name":"Siddharth Jindal","publish_date":"2024-05-13T11:07:50","publication_year":"2024","word_count":464,"keywords":["API","GenAI","sarvam ai","AI","ML","data pipeline","RAG","Aim","generative AI","R","startup"],"extracted_tech_keywords":["AI","ML","generative AI","GenAI","Aim","RAG","R","API","data pipeline","startup"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/sarvam-ai-launches-ai-residency-program-offering-up-to-inr-1-lakh-monthly-salary\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10012498,"title":"Top Free AI &#038; Data Science Courses Launched In 2020","content":"After a lot of turmoil that the year 2020 brought in terms of business disruption, challenges to ensure business continuity, remote working becoming the new norm, among others, many companies and institutes introduced and developed free courses for the data science and AI enthusiasts to make the most of the lockdown. From various ivy league institutions to major organisations across the globe, several courses were made free this year. As the year 2020 comes to an end, we list a few such courses in AI and data science that were made free and are still available for tech enthusiasts to avail. The courses are listed in no particular order. Coursera launched free courses In March this year, as lockdown ensued, to uplift the learners and help the community in the critical time, Coursera decided to launch new, free resources, as well as sharing impressive course collections, community discussions, and expert interviews. The courses were made free for anyone, anywhere. Aimed at helping the candidates take the first step in exploring a new career path, it made courses in areas such as data science, cloud technology and more, which can be availed even now. Click here to know more. Free nanodegree programs by Udacity To support learners in acquiring new skills with their nanodegree programs, Udacity announced to offer one free month on one of 40 premium nanodegree programs. It said that while their courses usually take about four to six months to learn a high-tech skill such as machine learning, data analytics or software development, if learners devoted more time, they could acquire skills in as little as a month. Some of the courses it offers are on machine learning, cloud, technology, programming languages, data science courses and more. Click here to know more. Free Course by edX In the light of COVID-19 pandemic, edX announced free courses through its Remote Access Program to assist universities by delivering educational content for their students. It not only provided their students with free access to courses and programs not just from their universities, but from all edX partners participating in the initiative. Click here to know more. Free course on data visualisation using Tableau by Great Learning This course by Great Learning covered a free data visualisation course on Tableau, which is one of the leading BI tools. A 1.5-hour video-based course, it covers concepts such as Tableau fundamentals, visual analytics, introduction to data visualisation, and more. The course intends to help students lay a path towards building a career in Tableau and business intelligence. Click here to know more. Free course on data science, machine learning, data analysis by Udemy The crash course by Udemy covers a wide range of topics such as concepts in data science, machine learning, data analysis, data visualisation using Python and R Programming, deep learning and more. While the online video content is free, instructor-led support and certification include fees. The 8-hour long programme is aimed at beginners and covers concepts on complex theory, algorithms, coding libraries and more. Click here to know more. Free course by AWS on edX The course titled ‘Amazon SageMaker: Simplifying ML Application Development’ is conducted by AWS instructors on concepts such as integrating ML into apps, using Jupyter Notebook to train a model with SageMaker, publishing validated models and more. This month-long course includes lectures by experts, discussions and demonstrations. Click here to know more. Free course by Harvard University This 4-week course by Harvard University called High-Dimensional Data Analysis is for intermediate level. It covers various aspects in data science such as mathematical distance, dimension reduction, multiple dimensional scaling plots, factor analysis, dealing with batch effects, and more. It also covers a brief introduction to machine learning concepts while focusing on clustering analysis, prediction algorithms such as k-nearest neighbours and more. Click here to know more. Free AI course by AI4ALL In March earlier this year, AI4ALL announced a free AI course module that could be availed by high-school teachers with a free online lesson to engage students around AI’s role in the current crisis. The course called ExploreAI aimed at supporting the community by sharing relevant, free curriculum and teaching resources. It covers courses on how AI can be used to tackle varying parts of COVID-19 outbreak. Click here to know more. IIT Kanpur to offer free statistics course To help students gain skills in statistics, IIT-Kanpur offered free statistics courses in assistance with a professor from the department of mathematics and statistics — professor Shalabh. With a wide range of topics such as regression analysis, sampling theory, linear regression analysis and forecasting, introduction to R, and more, the course can be downloaded from the IIT Kanpur’s website. Click here to know more. Free data science course by Alteryx To help graduates and unemployed data workers to acquire new skills, Alteryx announced the introduction of three data science courses along with a collaboration with Udacity to offer complimentary access to Udacity for 30 days. Called ADAPT, the course engages, encourages and enables candidates to gain concepts from data fundamentals to predictive analytics. Click here to know more. Free online courses by Amazon’s Machine Learning University Amazon announced to make the online courses by its Machine Learning University, which were previously available only to Amazon employees, available to the public. With an aim to help meet the increasing demand for individuals with ML expertise, the MLU classes will be available via on-demand video, along with associated coding materials. The course covers topics such as natural language processing, computer vision, and more while addressing various business problems. Check the course on YouTube. Yann LeCun’s deep learning course made free The deep learning course by Yann LeCun called the Deep Learning DS-GA 1008 at the NYU Centre for Data Science was made free. The courses which are accessible online will be led by LeCun along with Alfredo Canziani, an assistant professor of computer science at NYU. The 14-week course will cover various techniques in deep learning such as supervised and unsupervised deep learning, embedding methods, metric learning, convolutional and recurrent nets, and more. The course prerequisite includes DS-GA 1001 Intro to Data Science or a graduate-level machine learning course. Click here to know more. IIT Kharagpur launched two free online courses on data IIT Kanpur recently announced two new free online courses on data science on the SWAYAM NPTEL platform. These courses are suitable for UG students with science and engineering backgrounds, apart from working professionals in the analytics domain. The two courses are called Essentials of Data Science With R Software – 1: Probability and Statistical Inference and Essentials of Data Science With R Software – 2: Sampling Theory and Linear Regression Analysis. From basic mathematical and statistical concepts to sampling theory and regression analysis, the courses cover a wide range of topics. The courses are 12 weeks each and will be conducted from 18 Jan to 9 Apr 2021. Click here for details on Part I of the course. Click here for details on Part II of the course. IIT Madras launched two free online courses on AI Recently, IIT-Madras, in association with NPTEL platform, launched two new free online courses on artificial intelligence. The two courses are called — AI: Constraint Satisfaction and Artificial Intelligence: Knowledge Representation and Reasoning. The free courses aim to help students gain problem-solving capabilities using artificial intelligence. The duration of these two courses is eight weeks and 12 weeks, respectively, and will begin from Jan 2021. While the courses are free, the institute is offering an e-certificate for successful candidates with a minimum cost of ₹1000. Click here for details on Part I of the course. Click here for details on Part II of the course.","excerpt":"After a lot of turmoil that the year 2020 brought in terms of business disruption, challenges to ensure business continuity, remote working becoming the new norm, among others, many companies and institutes introduced and developed free courses for the data science and AI enthusiasts to make the most of the lockdown. From various ivy league […]","categories":["AI Trends"],"tags":["collaboration ai platform","Data Analytics Certification","data science curriculum","data visualization python","Natural Language Processing","python data visualization","Python for Data Science"],"author_name":"Srishti Deoras","publish_date":"2020-11-27T14:00:45","publication_year":"2020","word_count":1283,"keywords":["data visualization python","data science","Amazon SageMaker","artificial intelligence","machine learning","AI","data science curriculum","Natural Language Processing","ML","computer vision","Aim","collaboration ai platform","python data visualization","deep learning","analytics","Python for Data Science","Data Analytics Certification"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","computer vision","data science","analytics","Aim","Amazon SageMaker"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-free-ai-data-science-courses-launched-in-2020\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10111704,"title":"Prompt Engineering is Different for Open Source LLMs","content":"A few days ago, Meta AI introduced ‘Prompt Engineering with Llama 2‘, a new resource created for the open source community, which is a repository for the best practices for prompt engineering. Even Andrew Ng’s DeepLearning.AI recently released a course on this, called Prompt Engineering for Open Source LLMs. IBM, Amazon, Google, and Microsoft, have all been offering similar courses on prompt engineering for open-source models. Prompt engineering was one of the most talked about professions in 2023. As companies adopted OpenAI’s ChatGPT in different ways, they also got busy hiring experts who could prompt the chatbot to elicit the right responses, allegedly paying them huge paychecks. This also led to the rise of hundreds of prompt engineering courses that everyone wanted to get their hands on. But most of these were for closed source models such as OpenAI’s. Now, as companies adopt open source LLMs such as Meta’s LLaMA and Mistral, it becomes necessary to understand how prompt engineering is different for open source LLMs. Several corporate entities are developing and testing customer support and code generation applications based on open source technology. These applications are designed to engage with proprietary code unique to the companies, often proving challenging for the general closed-model LLMs created by OpenAI or Anthropic. “A lot of customers are asking themselves: Wait a second, why am I paying for a super large model that knows very little about my business? Could I not use just one of these open-source models, and by the way, maybe use a much smaller, open-source model for that (information retrieval) workflow?” shared Yann LeCun in a post on X. Prompt engineering for open source? Recently, Sharon Zhou, co-founder and CEO of Lamini, in partnership with DeepLearning.AI, conducted a course for prompt engineering on open source LLMs. She highlighted how the packaging of an open source model is different from the closed one, which affects the API, in the end, affecting the prompting mechanism. “LLMs wear pants, which is its prompt setting,” said Zhou, drawing a crazy analogy to how everyday people come to office wearing pants, and how that is the correct decision to wear them, and a change in this affects the whole system. She said that a lot of people get confused between prompt engineering, RAG, and fine-tuning. “Prompting is not software engineering, it’s close to Googling,” she added, and also spoke at length about this in her post on X recently. She added that RAG is prompt-engineering, “do not overcomplicate it”, it is just about retrieving information. Zhou emphasised the simplicity of prompt engineering, reiterating that prompts are just strings. She compared the process to handling a string in a programming language, making it clear that it’s a fundamental skill that doesn’t require complex frameworks. “Different LLMs & LLM versions mean different prompts,” she added. However, she acknowledged that many frameworks tend to overcomplicate prompt engineering, potentially leading to suboptimal results. Zhou explained that in practice, it’s essential to tailor your prompts when transitioning between different LLMs. This is similar to when OpenAI undergoes version changes, leading to confusion when previously effective prompts no longer yield the desired results. The same is the case with open source LLMs. Maintaining transparency in the entire prompt is crucial for optimising the model’s performance. Many frameworks face challenges in this regard, often attempting to abstract or conceal prompt details, creating an illusion of managing processes behind the scenes. Hits and misses When it comes to enterprise adoption, Matt Baker, the SVP of AI strategy at Dell, the company which partnered with Meta for bringing open source Llama 2 to enterprise use cases, said that large models are of no use for companies unless they are made for specific use cases. This is where smaller, specialised, and fine-tuned models come into the picture, giving birth to RAG and prompt engineering. Though the reality is that most of the companies would be using open and closed source LLMs for different use cases, the majority of information retrieval is now dependent on APIs and open source models, fine-tuned with their data, which is why companies need to adapt to learn how to prompt models precisely and give accurate information. To put it in Zhou’s words, always put the right pants on!","excerpt":"Don’t get confused between prompt engineering and RAG; they are the same!","categories":["AI Features"],"tags":["Courses","LLMs","Open Source AI","prompt engineering"],"author_name":"Mohit Pandey","publish_date":"2024-01-31T14:00:00","publication_year":"2024","word_count":709,"keywords":["Anthropic","ChatGPT","Meta AI","Go","API","OpenAI","AI","LLMs","RAG","Open Source AI","prompt engineering","Courses","R"],"extracted_tech_keywords":["AI","ChatGPT","OpenAI","Anthropic","Meta AI","RAG","prompt engineering","R","Go","API"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/prompt-engineering-is-different-for-open-source-llms\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10059068,"title":"How to navigate uncertainty in deep learning","content":"Machine learning heavily relies on probability theory. Hence, managing uncertainty (read imperfect or incomplete information) is key to machine learning (ML) projects. Ideally, deep learning makes it possible to produce dependable predictions on data from the same distribution the models were trained on. However, there are often disparities in the distribution of data on which the model was trained and to which a model is applied. For example, a 2018 study found that deep learning models trained to detect pneumonia in chest x-rays did not achieve the same degree of accuracy when they were evaluated on previously unseen data from hospitals. Methods such as Gaussian processes are very helpful in data analysis and decision making. For instance, an autonomous car may use this information to decide whether it should brake or not. That said, in assessing data and making decisions, it is important to also be able to question whether a model is certain about its output. While this is an underlying concern of Bayesian machine learning, deep learning models often ignore these questions— leading to situations in which it is difficult to tell whether a model is making a reasonable prediction or making guesses at random. Epistemic uncertainty There are two major types of uncertainty in deep learning: epistemic uncertainty and aleatoric uncertainty. Epistemic uncertainty specifically refers to what a model doesn’t know because it was fed inappropriate training data. This occurs when a model doesn’t have a sufficient amount of data and knowledge, which usually happens when there aren’t enough samples available for training the AI. The collection of observations acquired from the domain cannot be chosen without some systematic bias. While some level of bias is unavoidable, uncertainty increases if the level of variance and bias in the sample is an unsuitable representation of the task or project for which the model will be used. Unfortunately, in most cases, developers have little control over the sampling process, and obtain their data from a database or CSV file that they have access to. It is impossible to achieve complete coverage of a domain: there will always be some unobserved cases. Aleatoric uncertainty Aleatoric uncertainty describes the uncertainty that comes about as a result of the natural stochasticity of observations. Observations from a domain that has been used to train a model are always incomplete and imperfect. High aleatoric uncertainty occurs when there are few or no observations made while training a model. This type of uncertainty cannot be remedied by providing additional data. Noise in observations occurs when the observations from the domain aren’t concise: In other words, they contain noise. “Observations,” in this instance, refers to what was measured or collected: It is the input as well as the expected output of a model. “Noise,” on the other hand, refers to the variability in observation. This variability could be natural or an error, and affects both the input and the output of the model. Since data in the real world is messy and imperfect, we should be skeptical of data and develop systems that can navigate uncertainties. Error prone ML models are susceptible to errors, but some models can be useful despite being wrong: The variables include the procedure used to develop the model, including the selection of samples, decisions made in training hyperparameters, and in the construction of model predictions. Hence, given the uncertainty in deep learning,  the goal should be to build models with good relative performance and improve on the established learning models to account for the margin of errors.","excerpt":"Methods such as Gaussian processes are very helpful in data analysis and decision making.","categories":["AI Features"],"tags":["Deep Learning Models"],"author_name":"Srishti Mukherjee","publish_date":"2022-01-25T12:00:00","publication_year":"2022","word_count":586,"keywords":["Go","machine learning","TPU","AI","Deep Learning Models","RPA","ML","RAG","Ray","deep learning","R"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","Ray","RAG","TPU","R","Go","RPA"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-to-navigate-uncertaintly-in-deep-learning\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10168977,"title":"Duolingo Adds 148 Courses as AI Replaces Human Contractors","content":"Duolingo has launched 148 new language courses, more than doubling its total offerings and marking the company’s largest content expansion to date. The rollout significantly broadens access to the company’s most popular non-English languages, making them available in all 28 user interface languages supported by the app. The expansion includes new courses for Spanish, French, German, Italian, Japanese, Korean, and Mandarin, aimed at over a billion potential learners across Latin America, Europe, and Asia. Most of the courses support beginner levels (CEFR A1–A2) and include features such as Stories for reading comprehension and DuoRadio for listening skills. Advanced content is expected in the coming months. “This is a great example of how generative AI can directly benefit our learners,” said Luis von Ahn, CEO and co-founder of Duolingo. “This launch reflects the incredible impact of our AI and automation investments, which have allowed us to scale at unprecedented speed and quality.” This development comes after Duolingo announced it will phase out human contractors for tasks that can be handled by AI, as part of a broader shift to an “AI-first” approach. Ahn communicated to staff that the company will “gradually stop using contractors to do work that AI can handle,” and that new hires will only be considered if a team cannot automate more of its work. Duolingo’s new shared content system enables the company to build a high-quality base course and quickly adapt it across multiple languages. According to Jessie Becker, senior director of learning design, “It used to take a small team years to build a single new course from scratch. Now, by using generative AI to create and validate content, we’re able to focus our expertise where it’s most impactful.” The new courses improve access to Asian languages. In Latin America, Spanish and Portuguese speakers can now learn Japanese, Korean, and Mandarin. In Europe, learners who speak French, German, Italian, and Spanish have access to these languages as well. In Asia, speakers of 12 local languages — including Hindi, Tamil, and Thai — can now learn all of Duolingo’s seven top languages.","excerpt":"The expansion includes new courses for Spanish, French, German, Italian, Japanese, Korean, and Mandarin.","categories":["AI News"],"tags":["Duolingo"],"author_name":"Siddharth Jindal","publish_date":"2025-04-30T23:00:08","publication_year":"2025","word_count":345,"keywords":["Go","AI-first","programming_languages:R","AI","programming_languages:Go","automation","Aim","Duolingo","generative AI","R"],"extracted_tech_keywords":["AI","generative AI","Aim","R","Go","automation","AI-first","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/duolingo-adds-148-courses-as-ai-replaces-human-contractors\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":45035,"title":"How Zero Copy Data Virtualization Can Help Enterprises Become AI Ready","content":"Molecula is a Data Virtualization Platform that can created instantaneous access to huge, disparate, and geographically distributed datasets to help machine learning and analytical workloads at a much small cost compared to traditional approaches. There is an unabating amount of growing data distributed across the modern-day business in multiple formats, types, and locations. Every day there is about three quintillion bytes of data generated and according to estimates about 20% of that data is structured and available to processing. Even though enterprises are leveraging AI and advanced analytics to gain insights, there is a bottleneck experienced to access all of that data easily. Processes like batching, indexing, federation, aggregation, sampling and caching create long information request cycles which are the opposite of making real-time critical business decisions. This is the exact reason why 80% of all AI and Analytics projects fail to deliver what was originally planned. To solve this issue, one startup called Molecula, an Austin, Texas-based data virtualisation startup recently closed a $6 million seed round to bring its AI-accelerating software to the enterprise world. The funding round was led by The Seraph Group, Lontra Ventures, Velar Capital, Capital Factory, Andrew Busey and Jason Dorsey to help Molecula scale its technology, known as “zero-copy data virtualisation,” which makes complex analysis by data from various sources and locations available in real-time through a virtualised access layer. The company was founded in 2017 when it spun out of another data analytics company by the name of Umbil, which is also Austin-based. The company also announced that their clients and partners include some of the largest brands in the media, entertainment, technology and healthcare sectors. Key solutions deployed include real-time customer segmentation, real-time security and fraud detection and accelerating business intelligence and machine learning projects. It is to be noted that the startup was also part of Oracle Global Startup Ecosystem in October of 2018. Zero-Copy Data Virtualisation: The Innovation Molecula is a Data Virtualisation Platform that creates instantaneous, secure access to huge, disparate, and geographically distributed datasets to help Machine Learning and analytical workloads at a much small cost compared to traditional approaches. Typically, enterprises make many copies of their data before they acquire insights. In comparison, Molecula’s zero-copy data virtualisation can get users from data to decision without the typical aggregations, federations or other techniques deployed traditionally, and instead gives real-time virtualised access to all data in-memory. Instead of creating copies of data points and values like a typical index, the technology converts those into a knowledge representation of the data and distribute it across multiple machines. This avoids the need of data movement and provides a layer of abstraction above the physical implementation of data, irrespective of the source, how it is formatted and where it is physically located. Zero-copy data virtualisation can abstract 100% of data across RDBMS, data lakes, data warehouse and event streams. These representations are a fraction of the size, but contain all the information a machine learning system needs to process a request and analyse it. Molecula says its goal is to empower organisations to gain better insights by making all the data AI-Ready by reducing the time it takes to prepare data for machine learning efforts. Virtualised Abstractions can speed up analytics, machine learning and remote decision making, which can be incredibly powerful for gaining insights from Internet of Things (IoT) devices. “We’re a thousand times faster for machine learning models versus the traditional models,” Molecula has reported. Molecula is built on Pilosa, the open source technology used by over 1,800 organisations globally to make data AI ready. The Pilosa is an open source, high performance bitmap indexing which was created after years of research by Molecula team to tackle problems with massive amounts of high-velocity data that needs indexing and analysis. Pilosa’s open source software has the ability to store a virtual representation of underlying data in memory, thus making it orders of magnitude smaller and incredibly fast. Using a data repository, cloud, and operating system agnostic approach, Pilosa acts as an additional indexing system that can be applied to an existing data repository. Users can have Pilosa index across data repositories to link together disparate data sources and allow them to execute very fast queries against all of them at once, reducing query time from 20 seconds to 20 milliseconds, as reported. Overview Reports have revealed that business executives need data faster in order to keep up with customers, competitors, partners and prevent decreasing productivity. It is the reason that enterprise data management market size will grow by over $57 billion during 2019-2023. The prevalent process depends on creating full copies of data files to index, cache and process them, which takes up terabytes of space and is very slow, which is why machine learning models are unable to access the full datasets. Here, a startup like Molecula can gain much traction owing to its innovative open source technology.","excerpt":"Molecula is a Data Virtualization Platform that can created instantaneous access to huge, disparate, and geographically distributed datasets to help machine learning and analytical workloads at a much small cost compared to traditional approaches. There is an unabating amount of growing data distributed across the modern-day business in multiple formats, types, and locations. Every day […]","categories":["AI Features"],"tags":["AI (Artificial Intelligence)","Startups"],"author_name":"Vishal Chawla","publish_date":"2019-08-27T13:30:35","publication_year":"2019","word_count":819,"keywords":["Go","API","machine learning","AI","data warehouse","RAG","analytics","Startups","R","data lake","AI (Artificial Intelligence)","fraud detection"],"extracted_tech_keywords":["AI","machine learning","analytics","RAG","fraud detection","R","Go","API","data warehouse","data lake"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/molecula-data-virtualization-data-artificial-intelligence\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10165002,"title":"AWS Unveils New Cloud Solutions to Boost 5G Networks for Telcos","content":"At the Mobile World Congress held in Barcelona, Amazon Web Services (AWS) on Monday announced AWS Outposts racks for high throughput, network-intensive workloads and AWS Outposts servers designed for Cloud Radio Access Network (C-RAN) workloads. These new offerings enable telecom service providers, also known as telcos, to extend AWS infrastructure and services to deploy on-premises network functions requiring low latency, high throughput, and real-time performance. Both offerings will be generally available later this year to support hosting 5G Core User Plane Function (UPF), RAN Centralised Unit (CU), and RAN Distributed Unit (DU) workloads. “With the new AWS Outposts offerings, telcos can now run their entire 5G network, including 5G Core and 5G RAN, on AWS cloud services. These innovations will allow faster network deployment, better price performance, and improved customer experiences,” David Brown, vice president of compute and networking at AWS, said. The new AWS Outposts racks are built for high-speed 5G Core user plane and RAN workloads. Telecom companies can place workloads in different locations depending on speed, latency, and data traffic needs. The system uses 4th Gen Intel Xeon Scalable Processors and a high-performance network fabric. The company said the AWS Outposts racks offer scalability to handle increasing data traffic demands, ensuring telecom networks can expand efficiently. They provide enhanced security and performance through AWS’s Nitro System, delivering a reliable and protected environment. Moreover, the racks support automated deployment and management with AWS Kubernetes services, streamlining operations for telecom providers. Integrating with AWS analytics and monitoring tools also improves efficiency, allowing operators to monitor and optimise network performance seamlessly. Notably, O2 Telefónica, a major telecom provider in Germany, is already using AWS for its cloud-based 5G core network. Meanwhile, Mallikarjun Rao, chief technology and enterprise officer at O2 Telefónica, said, “We are proud to have the first fully cloud-native 5G core network deployed on AWS, serving a million subscribers. The new AWS Outposts racks align with our strategy to build the network of the future.” The AWS Outposts servers are tailored for Cloud RAN workloads, helping telecom providers deploy virtualised 5G networks more efficiently. These servers have been developed in collaboration with Nokia, with plans to integrate additional RAN vendors in the future. Why AWS Outposts? The AWS Outposts servers offer simplified operations with a pre-integrated cloud infrastructure, reducing the complexity of deployment and management. They enable faster 5G innovation by providing access to over 200 AWS cloud services, allowing telecom providers to develop and launch new features more efficiently. Powered by AWS Graviton3 processors, these servers deliver high-performance computing to meet the demanding requirements of 5G networks. Moreover, they ensure seamless integration with RAN vendors, maintaining high radio performance and supporting the smooth deployment of Cloud RAN solutions. Major telecom operators like Orange, and Du Network will begin testing these solutions later this year. AWS’s new Outposts racks and servers are currently in preview and will be widely available later this year.","excerpt":"The new AWS Outposts racks are built for high-speed 5G Core user plane and RAN workloads.","categories":["AI News"],"tags":["5G","AWS"],"author_name":"Shalini Mondal","publish_date":"2025-03-03T18:03:26","publication_year":"2025","word_count":485,"keywords":["5G","AWS","AI","cloud_platforms:AWS","innovation","ML","Scala","ViT","analytics","R","kubernetes"],"extracted_tech_keywords":["AI","ML","analytics","AWS","kubernetes","R","Scala","ViT","innovation","cloud_platforms:AWS"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/aws-unveils-new-cloud-solutions-to-boost-5g-networks-for-telcos\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10055086,"title":"AI Governance Depends On The Kind Of Society One Comes From: Manojkumar Parmar, Bosch","content":"AI models require a large amount of sensitive training data and are usually computationally intensive to build. They also suffer from threats from attackers, who leverage loopholes to break into the system. It often results in destroying brand reputation, differentiation, and value proposition. Recognising this, Robert Bosch has constituted a team that works exclusively on dealing with challenges pertaining to AI security. To understand more about this, we caught up with Manojkumar Parmar, Deputy General Manager – Technical at Bosch on AI and Security, who is also leading the team. AIM: How did you develop an interest in the security aspect of AI? I joined NVIDIA Graphics soon after my graduation. I worked there for a year and a half, where I worked in VSLI design. Outside of work, I started exploring AI. In 2018, I joined Robert Bosch’s newly minted unit for innovation and incubation. I was mainly responsible for advanced computing, technology in the capacity of innovation experts. Around the same time, I read a very interesting paper on how one can steal a model just by having access to the API. We are dealing with a newer digital asset class of AI, which is very high in value. Traditionally, whenever any new asset class comes into the picture, hackers launch an attack post which the security teams build solutions based on these attacks. This got me thinking a lot. We often hear a lot about data being the new oil, but what about the model? I believe the model is like jet fuel, enriched with insights. Why would hackers go after crude oil (data) when they can steal your jet fuel? This got me interested in AI security and eventually motivated me to set up a unit within Bosch exclusively to work in this area. While the work on building a technology team at Bosch for AI security started a few years ago, it was formalised and officially started operations this year. AIM: How was the idea to build an exclusive team to work on issues pertaining to AI security conceived and finally brought to fruition? I worked with my colleagues for eight months to really understand and get to the crux of what AI security encompasses. Initially, we formulated a very informal study; we realised that the field was very nascent. The key finding of the study was on how to increase the accuracy of the model. All the focus was on how to increase the accuracy of the model, and security was not of any concern. The higher management approved our idea to build something in this space, and we spent the next few years building the proof of concept to demonstrate the security issues with the production-grade AI. Once that was done, the team and the company decided that it was time investment was made into this initiative. In April this year, the entire program was born. The team consists of 12 members from diverse backgrounds like AI, security, product management, and business development. AIM: Concepts like trust and ethics are very intangible. How do you think we can measure and supervise them? Manojkumar Parmar: Trust is an inferred quality based upon the lot of things that you do. At Bosch, we have developed AI Codex Ethical guideline for using an AI. It is a top-level set of principles, which gets operationalised internally, contextualised for each of our business use cases. This framework has been developed over the last few years, and it is operationalised. Every product team or the project team is empowered to look into this particular assessment and understand it. AIM: What do you think are the most common security challenges AI models currently face? Manojkumar Parmar: Today, AI security has become a very common term. But in terms of industry, it generally refers to how AI is being used to do traditional cyber security jobs, which involves tasks like threat hunting, detecting fraud behaviour and malware, etc. Now what we are saying is a little different than that. What we are saying is that the AI is opening up newer attack surfaces to attack the AI itself. We are focusing on how to secure AI as an asset class and not utilise it to secure some other asset classes. So in layman’s terms, you can say that cybersecurity is policing the actual asset. We are asking who is policing the police. AIM: A lot of people also think that concepts like AI security, governance, and ethics may be a hindrance to rapid innovation. What is your take on this? Manojkumar Parmar: Take a case of recently concluded COP-26. Such an event\/conference had to be held because threats were not paid heed to from the early stages. We don’t need to repeat this with every technology, just in the name of innovation. Technology is like a vehicle; it has both an accelerator (innovation) and a brake (governance\/security\/regulation). The brake gives the confidence to increase the speed gradually and responsibly. We, as a society, have to really take care of where our values are and how much we are willing to sacrifice them. For some societies, it is OK to sacrifice all of these values in the name of innovation to move faster, but for some, it is not. So, I don’t have any blanket answer for whether it is right or wrong, but I think regulations put the problem in perspective and tell clearly where to innovate. I think it is good that we are reacting early. And again, the keyword here is ‘reacting early’. In a lot of cases, we have reacted very late, which had dire consequences. The last five to six years have been spent in building the technology and making it relevant for the people and organisations undergoing digital transformation. AI assets are increasing, and the need for its security is also rising. So for me, it is the right time to talk about AI security. AIM: With respect to AI security, what do you predict for 2022? Manojkumar Parmar: For the first time, academia needs a lot of catching up to do. In terms of AI security, the industry has taken the lead. As industry players, we have burned our fingers with cybersecurity enough—the trillions of dollars that are being lost because of that data security breach and other issues. We have learned our lessons. We want it to be ready for any such unforeseen situations. It’s a selfish viewpoint, but we feel good about being ahead of academia, at least here, because it is generally them dictating terms of what to work on AI. AIM: Do you think a framework for AI security, governance and trust can be applied to the entire industry? Manojkumar Parmar: It’s an interesting question because, in AI, the issues regarding governance and trust are related to the culture and society you come from. So as an industry, we also factor in the cultural aspects of AI. It isn’t easy to have in common regulations for everyone until and unless we factor in the anthropological and social aspects. It may be possible to make regulation to a certain extent (a common, minimal program way). There are already initiatives like a model card or a model sheet, which actually tells users complete things about how the model is built. These are more like informative and not enforceable regulations. It is helpful because you are putting the choice in the hands of a consumer whether they want to really use this or not. This is the normal principle of data privacy extended to the AI-related issue. You as a consumer decide if you want to use it or not, and our job as an industry is to give you that information.","excerpt":"Technology is like a vehicle; it has both an accelerator (innovation) and a brake (governance\/security\/regulation). The brake gives the confidence to increase the speed gradually and responsibly.","categories":["AI Features"],"tags":["AI Security","bosch","Interviews and Discussions"],"author_name":"Shraddha Goled","publish_date":"2021-12-08T11:02:01","publication_year":"2021","word_count":1288,"keywords":["Go","API","bosch","AI","Git","AI Security","RAG","BERT","Aim","Rust","GAN","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","Aim","RAG","R","Go","Rust","Git","API","BERT","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/ai-governance-depends-on-the-kind-of-society-one-comes-from-manojkumar-parmar-bosch\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":50556,"title":"Here’s The Blow-By-Blow Account Of  Google’s Nightingale Nightmare; Health Data Harvesting Project Spurs Federal Inquiry","content":"Google, in partnership with one of the largest health-care systems, has been collecting health-data of millions of American citizens without their consent. As part of this project, called Nightingale, US healthcare provider Ascension was providing Google access to private health information such as lab results, diagnoses, records of hospitalization and dates of birth of nearly 50 million Americans without their doctor’s consent. Ascension is a Catholic network of more than 2,600 sites of care – including 150 hospitals and more than 50 senior living facilities – in 20 states and the District of Columbia. Earlier this year in July, Google had mentioned about this partnership in its Q2 2019 Earnings Call, but not in much detail. However, it was only after The Wall Street Journal first reported the news that both the companies announced more details about this project. The privacy-breaching ambit of this project drew massive backlash from the public as well as prominent figures. U.S. Senator Amy Klobuchar (D-MN) released an official statement following the WSJ report, expressing “privacy concerns” regarding the project. U.S. Senator Lisa Murkowski also tweeted her concern regarding this partnership. She stated, “Like many Americans, I’m concerned to hear about details surrounding the so called Project Nightingale and its gathering of personal health data for millions of people. Privacy protections, particularly when it comes to personal info like your health, is a high priority of mine.” Like many Americans, I’m concerned to hear about details surrounding the so called Project Nightingale and its gathering of personal health data for millions of people. Privacy protections, particularly when it comes to personal info like your health, is a high priority of mine. https:\/\/t.co\/fmRA9gNimb— Sen. Lisa Murkowski (@lisamurkowski) November 13, 2019 After the WSJ report, an anonymous whistleblower who works in Project Nightingale also expressed concern and anger regarding the project and the fact that the patients are being kept in the dark. The person also posted a video on social media platform Daily Motion that disclosed hundreds of images of confidential documents pertaining to Project Nightingale. However, the video has been taken down by the platform citing “breach of the Terms of Use”. The Guardian also published an op-ed written by the anonymous whistleblower wherein the person mentioned, “I grew increasingly concerned about the security and privacy aspects of the deal. It became obvious that many around me in the Nightingale team also shared those anxieties.” “Above all: why was the information being handed over in a form that had not been “de-identified” – the term the industry uses for removing all personal details so that a patient’s medical record could not be directly linked back to them? And why had no patients and doctors been told what was happening?” the person added. The Nightingale Nightmare For Google Amidst all the backlash and public outrage, the Leaders of the House Committee on Energy and Commerce have written to the CEOs of Ascension and Google, requesting briefings from the companies on Project Nightingale by December 6. The companies will brief the Committee on the following: 1. What data Ascension is sharing with Google? 2. How is the data being used and shared? 3. The extent to which employees at Google and its parent company Alphabet have access to this information. 4. The extent to which patients were informed about the use and sharing of their data. 5. What steps are being taken to protect the privacy and security of patients’ data? “While we appreciate your efforts to provide the public with further information about Project Nightingale, this initiative raises serious privacy concerns. For example, longstanding questions related to Google’s commitment to protecting the privacy of its own users’ data raise serious concerns about whether Google can be a good steward of patients’ personal health information,” the Committee leaders mentioned in the letter. Project Nightingale is also facing a federal inquiry. The Office for Civil Rights in the Department of Health and Human Services “will seek to learn more information about this mass collection of individuals’ medical records to ensure that HIPAA protections were fully implemented.” In response to the federal inquiry, Tariq Shaukat, President, Industry Products and Solutions, Google Cloud said in a blog post, “We are happy to cooperate with any questions about the project. We believe Google’s work with Ascension adheres to industry-wide regulations (including HIPAA) regarding patient data, and comes with strict guidance on data privacy, security, and usage.” Dr. A. C. Sharma is a  general surgeon and has worked with the UP govt for almost 40 years. He retired as the Joint Director of State Medical and Health Services, Uttar Pradesh. He told Analytics India Magazine that data pertaining to a person’s health as well as to his\/her medical condition is highly personal and private. “The law also dictates that any information that is shared between a patient and the doctor regarding health or disease is considered a “ Privileged Communication” and either the person’s consent or a judicial order is a must for such data to be divulged or accessed. Therefore, in my opinion, such an act of accessing private health data is wrong.” The Hype Surrounding HIPAA The Health Insurance Portability and Accountability Act or HIPAA is a federal law that regulates the privacy and security of protected health information and data. Under the US law, any health information that contains an identifying element (name, social security number, telephone number, email address, street address, among others) that can be linked back to a specific patient is referred to as Protected health information (PHI). According to Shaukat, “All of Google’s work with Ascension adheres to industry-wide regulations (including HIPAA) regarding patient data, and come with strict guidance on data privacy, security and usage. We have a Business Associate Agreement (BAA) with Ascension, which governs access to Protected Health Information (PHI) for the purpose of helping providers support patient care.” Shaukat also claimed that “This is standard practice in healthcare, as patient data is frequently managed in electronic systems that nurses and doctors widely use to deliver patient care.” Although, he ensured that as per this partnership, “Ascension’s data cannot be used for any other purpose than for providing these services we’re offering under the agreement, and patient data cannot and will not be combined with any Google consumer data.” “Based on what I see in media reports, I think that Google and Ascension have probably complied with HIPAA (US health privacy law) at least within the letter of the law.  They have structured this so that it is designed to provide data used to improve the care of Ascension patients and would thus be deemed quality improvement—or related to care.”Margaret Riley, a law professor at the University of Virginia with extensive expertise in the areas of healthcare law and bioethics, told Analytics India Magazine “The problem that I have is that the size and scope of the endeavor feel more like research to me.  It is not clear who owns the data and algorithms that are being produced. Unless you are eligible for a waiver of consent, research requires consent and that, since they argue they aren’t doing research, has not been obtained,” she added. “I have always believed that data will become ubiquitous. The value is not in the data but what you do with the data. Google has already dominated the internet of information, now they clearly want to dominate the internet of health. In the past, Google was indexing webpages and websites and thereafter would present your search results. In the future, they will start indexing your health data and figuring out what is most relevant for you.”Talking about the importance of data and the impact of Google on this domain to Analytics India Magazine, Vishal Gondal, Founder, and CEO of GOQii said Shaukat claims that “the goal of the partnership is to enhance the experience of patients and clinical providers across the continuum of care, improving outcomes and saving lives.” According to Shaukat, Ascension is moving its infrastructure to Google Cloud and its productivity software to G Suite. The non-profit health system is also working with Google to pilot tools designed for doctors and nurses to utilize for patient care. “Specifically, we are piloting tools that could help Ascension’s doctors and nurses more quickly and easily access relevant patient information, in a consolidated view,” Shaukat claimed. Gondal also believes that this partnership will improve hospital operations and reduce overall medical costs. “How Google manages and uses this treasure trove of data will be of keen interest,” he stated. After being subjected to scrutiny regarding the security and privacy aspects of the Project, both the companies leapt into troubleshooting mode and divulged more information on this partnership, as well as their intent behind this deal. Google published the blog post detailing the facts about the deal, along with FAQs addressing the details around the project and how it protects patient data. Ascension, on the other hand, issued a press release sharing the company’s perspective and role. “As the healthcare environment continues to rapidly evolve, we must transform to better meet the needs and expectations of those we serve as well as our own caregivers and healthcare providers,” said Eduardo Conrado, Executive Vice President, Strategy and Innovations, Ascension. He went on to add, “Doing that will require the programmatic integration of new care models delivered through the digital platforms, applications and services that are part of the everyday experience of those we serve.” Harrows Of Health-data Harvesting However, this isn’t the first time that Google is in hot water with matters concerning healthcare data. Back in 2017, Google, the University of Chicago and an affiliated medical center announced a partnership wherein University of Chicago Medical Center will share patient data with the tech giant in order to utilize the unused data goldmine available in the form of electronic health records and come up with better predictive analysis in medicine. In July, this year, Google, the medical center, and the University of Chicago were sued in a potential class-action lawsuit that accused the University of sharing private health data with Google without the consent of the patients, even after the former “promised in its patient admission forms that it would not disclose patients’ records to third parties, like Google, for commercial purposes.” Also, the claims made by the tech giant and the University that the medical records were de-identified were “incredibly misleading”. In another case of Google’s concerning health-data harvesting endeavor, the company entered into a  data-sharing agreement with UK’s state-run National Health Service that allowed Google-owned artificial intelligence company DeepMind access to a vast amount of health-data of more than 1.6 million patients, as part of a research program. The Fitbit Conundrum Earlier this month, Google announced that it has entered into a definitive agreement to acquire wearable and fitness tracking company Fitbit for over $2.1 billion. In a blog post, Rick Osterloh Senior Vice President, Devices & Services, Google mentioned, “By working closely with Fitbit’s team of experts, and bringing together the best AI, software and hardware, we can help spur innovation in wearables and build products to benefit even more people around the world.” With this acquisition, Google will gain access to the health-data of more than 28 million active users of Fitbit devices. This has raised concerns among politicians and privacy and anti-trust activists; some of whom are even calling for a block on this acquisition by reaching out to the Federal Trade Commission. Democratic Congresswoman Katie Porter from California tweeted, “Google is under anti-trust investigation and is gobbling up Fitbit — a company that stores some of our most private health data. It’s time for anti-trust enforcers to do their jobs instead of keeping us all under the rule of monopolies.” Google is under anti-trust investigation and is gobbling up Fitbit — a company that stores some of our most private health data.It’s time for anti-trust enforcers to do their jobs instead of keeping us all under the rule of monopolies. https:\/\/t.co\/zkozJtc6e8— Rep. Katie Porter (@RepKatiePorter) November 1, 2019 While Representative David Cicilline, Chairman of the House Judiciary antitrust subcommittee, during a subcommittee hearing on antitrust in digital markets, said, “Google’s proposed acquisition of Fitbit would threaten to give it yet another way to surveil users and entrench its monopoly power online.” However, Osterloh mentioned in the blog post that Google will never sell personal information to anyone. “Fitbit health and wellness data will not be used for Google ads. And we will give Fitbit users the choice to review, move, or delete their data,” he said. Wearing Your Health Data On Your Sleeve We, as users of health-tracking apps and devices, offer access to our personal health information to these companies in return for their health-monitoring benefits. U.S. Senator Amy Klobuchar (D-MN), believes that the collaboration between Google and Ascension is not the only one that raises concern. “New technologies have made it easier for people to monitor their own health, but health tracking apps, wearable technology devices like Fitbits, and home DNA testing kits have also given companies access to your private health data with very few rules of the road in place regulating how it is collected and used,” she said in a statement. Riley believes that we do provide access to our personal health information all the time, and when outside of the HIPAA context (which is the case with most wearables) privacy interests are governed by contract and most users simply click through.  She opines that as users, we don’t really engage in a full discussion regarding the risks and consequences. Privacy is often discussed without a parallel discussion about various aspects such as the benefits of sharing data, who ends up owning the data and it’s consequences, and what alternatives we might pursue as a society. The market for wearable tech in India is also booming. According to research firm International Data Corporation, the market for wearable devices more than doubled in the second quarter of 2019, registering an all-time high of 3 million shipments. “This has further cemented India’s position as the third-largest wearables market in the world after China and the USA,” said IDC. “We are seeing a shift in lifestyle devices segment wherein consumers are adopting new kind of devices to track their health and fitness that is reflecting in the uptake of the wearables segment. Brands continue to augment newer health tracking features and increasingly spending on marketing, and this is helping wearables to become one of the preferred choices amongst the fitness enthusiasts to support an active lifestyle,” said, Jaipal Singh, Associate Research Manager, Client Devices, IDC India. In the affordable wearable segment in India, GOQii is a major player, along with Xiaomi. Talking about the steps taken by GOQii to ensure the safety of its user health data, Gondal said, “We use large, aggregated, anonymized data sets, from over millions of users, collected over a length of time, to generate health and fitness trends and insights. We then compile this information into a report, which we make publicly available.” He went on to explain that since the data that GOQii collects is voluminous and anonymized, health data of any single individual cannot be inferred from the data sets. The company uses this research to devise strategies to make the app better at providing more relevant health and fitness information. Additionally, GOQii claims to comply with international standards such as  HIPPA and ISO 27001. “At Ciitizen we believe the best way to address patient data privacy is by putting patients in charge of their own data and letting them consent as to how and with whom it is shared,” Premal Shah, President & Co-Founder of Ciitizen told Analytics India Magazine. California-based Ciitizen is a consumer health tech company that offers a platform that helps patients collect, organize, and share their medical records digitally. “While tech hacks are something, that all organizations need to pay careful attention to, the breach that people are most concerned about is a breach of trust, not of technology. When a patient controls the movement of their own health data they know exactly how their health data is being used,” Shah added. The health-tech market in India is evolving and is on an upward trajectory. According to research and analytics platform Tracxn, as of September 2019, there 2,975 health-tech startups in India. A potential health data privacy breach could really put a damper on the growth of heath-tech in India and also make the general public apprehensive of using the services of these startups. “Client health data is very important and confidentiality of the same needs to be protected at all costs.”According to Dr. Geetha Manjunath, Cofounder and CEO of Niramai Niramai is an AI-driven cancer detection startup that uses a secure storage system for storing patient data in an anonymized and encrypted form. Also, the company doesn’t take any personally identifiable information (PII) from the patient and so once in the data store, there is no way to trace back to the patient with just that information in the datastore (without hospital interaction). Talking about the steps that should be taken by Health-tech companies like Niramai and healthcare regulatory authorities to ensure that private health data stays private and is not misused, she said, “It is important for health tech companies to maintain the integrity of patient data and protect the confidentiality of health information.  Patient data should also not be misused against the patient. For example, if the data is used to gain insights about the patients and then those insights are used against them in increasing insurance premium and such. One way of protecting the user from such threats is to completely anonymize the data and only allow the health tech companies to use abstract information to gain insights about the community, or use it for research and build better predictive models which will further benefit all patients.” “I think regulatory authorities should insist on companies to document the data operating procedures in their Quality Management Systems and ensure implementation of the same through regular inspections and audits. CE mark, ISO 13485 and GDPR requirements provide such broad guidelines, which I think all health tech companies need to follow strictly,” she added.","excerpt":"Google, in partnership with one of the largest health-care systems, has been collecting health-data of millions of American citizens without their consent. As part of this project, called Nightingale, US healthcare provider Ascension was providing Google access to private health information such as lab results, diagnoses, records of hospitalization and dates of birth of nearly […]","categories":["Deep Tech"],"tags":["industry-wide analytics","purposes of a data team"],"author_name":"Rahul Raj","publish_date":"2019-11-25T12:19:29","publication_year":"2019","word_count":3034,"keywords":["Go","artificial intelligence","industry-wide analytics","AWS","AI","Rust","Git","RAG","Aim","analytics","purposes of a data team","R"],"extracted_tech_keywords":["AI","artificial intelligence","analytics","Aim","RAG","AWS","R","Go","Rust","Git"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/heres-the-blow-by-blow-account-of-googles-nightingale-nightmare-health-data-harvesting-project-spurs-federal-inquiry\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10113021,"title":"GitHub Teams Up with Polar for Easier Open-Source Funding","content":"GitHub has teamed up with Polar, a funding platform, to let developers make money from their projects on GitHub.This helps solve a big problem: getting consistent money for open-source projects. Polar lets developers sell extra features and subscriptions, offering a steady flow of cash. https:\/\/twitter.com\/birk\/status\/1758087210211909649?s=20 Polar, started in Sweden by Birk Jernström, lets maintainers set up their own subscription services and benefits, such as access to private repos, Discord invites, premium content, and promotions. This means developers have a lot of freedom to make money while supporting open-source work. “Donations & sponsorship are great when they happen. Problem is they rarely do. In order to drive meaningful (full-time work) capital to OSS initiatives, I believe it has to charge for add-on value and that such services and subscriptions are mutually beneficial. See mkdocs-material as a prime example,” Jernström said in a HackerNews discussion a few months back. This partnership makes it easier for developers to get funded and improves GitHub too. With Polar, developers can create and share posts and newsletters, both free and paid. They can sell subscriptions linked to their GitHub projects. Polar’s tools can be used to add these services to their websites or docs. The company takes a 5% fee plus Stripe’s fees, but they’re covering Stripe fees until March 31, 2024. Polar is open-source, encouraging open collaboration and feedback on its development. This step helps open-source developers financially, benefiting both the creators and the open-source community. It connects GitHub’s network with Polar’s funding options, aiming to change how open-source projects are funded so developers can focus on creating software.","excerpt":"GitHub introduces a subscription-based funding platform to support open-source developers with a sustainable income model.","categories":["AI News"],"tags":["Fund Raising","GitHub"],"author_name":"K L Krithika","publish_date":"2024-02-15T21:21:45","publication_year":"2024","word_count":264,"keywords":["API","funding","programming_languages:R","AI","Fund Raising","Git","RAG","Aim","ViT","GitHub","R"],"extracted_tech_keywords":["AI","Aim","RAG","R","Git","GitHub","API","ViT","funding","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/github-integrates-polar-for-direct-open-source-funding\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10008498,"title":"Webinar: How To Crack Data Science Hackathons","content":"Hackathons have evolved to be one of the key ways for data scientists to learn new skills, practically apply their skills, get a hands-on solving business-centric problem, build a community, and more. Not just a way for boosting the skills, hackathons have also become one of the crucial tools for companies to hire candidates for internships and even full-time employment. Addressing the growing popularity of hackathons, there are several hackathon platforms available today for data science enthusiasts to compete on. While there are many resources available, it can be overwhelming for candidates, especially newcomers to fit right into hackathons. There can be challenges such as finding the right project to work according to the skills to even becoming a part of the team. Register here for the Webinar To address some of the commonly faced challenges that participants face such as the kind of coding skills required, picking the right platform, approaching the problem-statement, and more, Analytics India Magazine in association with MachineHack will be conducting a webinar on ‘How to crack data science hackathons’. It will be conducted by Devrup Banerjee, PGDM, Great Lakes. He is also a GrandMaster at the Machinehack platform. Banerjee will take the participants through the nuances of cracking data science hackathons. He will take through pointers such as: Best practises to follow while participating in a data science hackathonHow to approach a problem-statementHow to best utilise the resources and knowledge you have to crack the hackathonHow to pick the right hackathon platform to competeQ&A with the participants Register here for the Webinar Details of the webinar Speaker: Devrup Banerjee, PGDM, Great Lakes, Machinehack GM Timing: 3rd October, 11:00 AM Duration: 1.5 hours including Q&A Register here for the Webinar","excerpt":"Hackathons have evolved to be one of the key ways for data scientists to learn new skills, practically apply their skills, get a hands-on solving business-centric problem, build a community, and more. Not just a way for boosting the skills, hackathons have also become one of the crucial tools for companies to hire candidates for […]","categories":["Deep Tech"],"tags":["Machinehack","webinar"],"author_name":"Anurag Upadhyaya","publish_date":"2020-09-28T19:00:45","publication_year":"2020","word_count":285,"keywords":["data science","programming_languages:R","AI","Machinehack","analytics","R","webinar"],"extracted_tech_keywords":["AI","data science","analytics","R","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/webinar-how-to-crack-data-science-hackathons\/","complexity_score":2,"technical_depth":5,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":7504,"title":"Interview &#8211; Ujjyaini Mitra, Head of Analytics at Bharti Airtel","content":"As Head Analytics of Airtel’s mobility business, Ujjyaini plays the role of Subject Matter Expert in Analytics with a special focus to consumer’s Usage and Retention (UnR) & Pricing across Airtel’s product family, Customer service Experience and Market Research. Along with building highly complex marketing models in Big Data space to achieve high revenue growth via right contextual marketing and enhanced customers’ life cycle value, she mentors the business leaders to draw right insights from the structured and unstructured data or reports. At the same time, she hand holds the junior members to break the business problem into impactful and actionable business analysis and to story line the findings. We sit with Ujjyaini to learn more about her thoughts on analytics. [dropcap]AIM[\/dropcap]Analytics India Magazine: Ujjyaini, thanks for speaking to us. From a bird’s eye view, tell us about analytics within Airtel. [dropcap]UM[\/dropcap]Ujjyaini Mitra: Analytics is an integrated part of Business Intelligence in Airtel. Business decisions are confluence of data driven insights and external market knowledge. Be it measuring campaign efficacy scientifically or taking complex decision of where to open new Airtel Store – are driven analytically backed by data in Airtel. Airtel follows an approach of ‘issue-to-outcome’, where Analytics is infused to draw right information, information driven insights and thus implementation. So, Analytics is used at every stage of taking business decisions. Airtel adopted Analytics around 2009-10 while Airtel tied up with a Globally acclaimed business Management firm to start their journey on Consumer Life Cycle Management factory (CLM factory). Today this factory is the Marketing lifeline in terms of Marketing Operations and measuring Marketing Outcomes. Following is a slide to show how Airtel takes Business decisions infused with data driven Analytics. Thus Airtel got benefitted everyway, starting from right customer acquisition module, improving campaign hit rate through right targeting of customers, improved ROI through right pricing, enhanced Customer Satisfaction and thus ‘winning customers for life’. AIM: Can you brief about some of the analytics solutions that you work on? UM: All my current works has few basic goal – a) improve ROI, b) improve customer life. Therefore, customer interactions through every touch points can be integrated towards more contextual, timely and easily adoptable campaigns or communications. Airtel not necessarily do everything in house. In fact, Airtel has a great eco system of outsourcing the operation and partnering with world’s giant knowledge hubs and experts. We build the solution jointly, customized to Airtel’s expectations. So that complex models can give birth of simple solutions to implement seamlessly. All Analytic solutions are built keeping ‘Customers’ in focus, so that, every department work towards the same focus point and outcome become customer centric. AIM: Can you brief us about a specific use case in analytics that has brought significant value to Airtel? UM: Last year I established a world class and completely automated measuring tool for all Airtel campaigns. So, earlier, where, it was not easy to measure right efficacy of the campaigns due to leakage in calculation, I have made it scientific and water tight. It is being automated, no human effort is required in calculation. CLM team, therefore, seamlessly prepare the ROI reports at any number of granular time frame. UnR team, Pricing team centrally and across circle can take a faster decision on which campaign is doing well, which is not, which one to continue, which one to stop, which campaign’s pricing needs to be changed vs which campaign’s content requires modification. AIM: Please brief us about the size of your analytics group and what is hierarchal alignment, both depth and breadth. UM: This year we have plan to grow the Analytics team. However, instead of just building a central analytic hub, we want to disseminate the Analytic knowledge to business managers, so that they can think and problem solve analytically. We are therefore, investing a lot to train all the business managers and vertical heads to understand Analytics, Statistics and their implementation in Marketing and other Business domains. That way, the whole organization row towards the same goal – mapping Analytics directly to decision making, taking action and delivering value for improved business performance. We are evaluating good Analytical S\/W which can empower our Business owners with Analytical tool kits to take data driven insights. Alongside, planning to build a strong core Analytics hub to serve the whole organization with analytical rigor. AIM: What kind of knowledge worker do you recruit and what is the selection methodology? What skill sets do you look at while recruiting in analytics? UM: People with good knowledge of Statistics & its right application [e.g. knowing t-test is just OK, but one must know given a business problem whether to apply t-test vs F-test or Chi-square test], Big data tools, and basic understanding of business or market. AIM: What are the most significant challenges you face being in the forefront of analytics space? UM: 3 challenges: Shortage of right skilled people [most Analytic institution either train focusing only towards theory or tools without teaching the real life applications; while other institute teaches top level application of Analytics in Business without balancing with right technical skill building] Making people unlearn ‘ad-hoc’ decision making based on gut feeling and training to use Analytics to derive data driven insights, therefore slow adoption. Lack of organizational strategy to grow Analytics even though leadership realizes its incredible benefit. AIM: How did you start your career in analytics? UM: I started my career with McKinsey & Company’s knowledge hub ‘McKinsey Knowledge Center’. I was a campus hire from Indian statistical Institute, Kolkata. Almost after 6.5 years of working with International Clients across the Globe, I joined Airtel to lead their Analytics effort. AIM: How do you see Analytics evolving today in the industry as a whole? What are the most important contemporary trends that you see emerging in the Analytics space across the globe? UM: Like any other industry Analytics also follow an S curve and currently Analytics is at its sharp rise state. The days of traditional Analytics is fading where main focus was at structured dataset. Market is trending towards Non-Structured or Semi-Structured data analysis and therefore merging both of them to find the real golden nuggets from the mine of Data. As the world is becoming more and more digital, Digital Analytics, Network Analysis, Spectral Analysis, Speech Analytics with Natural language Processing etc are taking bigger shape. Therefore the application of Robotics and Artificial Intelligence is going to step is a big way in next few years. In fact, Analytics will borrow the discoveries from Atomic Physics, Bio-Chemistry, Genetics etc and apply to Big data or Consumer data analytics. [divider divider_color=”#777777″ link_color=”#777777″ size=”1″] [spoiler title=”Biography of Ujjyaini Mitra” style=”fancy” icon=”plus-circle”] MS Quantitative Economics (Indian statistical Institute, Kolkata); BSc in Mathematics & Theoretical Computer Sc. (Chennai Mathematical Institute, Chennai) Ujjyaini has over 7.5 years of rich experience in Analytics and their detailed applications in multiple facets of Marketing, spread across industries – Telecommunication & Media, E-commerce, Retail Banking, CPG & consumer durable goods and Public Sector to name few. Over last 1 year she is working with India’s largest telecom player Bharti Airtel’s marketing intelligence as their Head of Analytics. Whereas, in previous 6.5 years she has served multiple clients across the Globe as a part of McKinsey & Company’s Analytic hub – McKinsey Knowledge Center (McKC), India. Before parting from McKC, she was leading a team of 35 members’ Marketing & sales Center of Competence. Her work on ‘India’s urban awakening’ and ‘Global labor market & skill shortage’ had been published in McKinsey Quarterly – an internationally acclaimed management journal. Her work interest lies in the domain of pattern recognition and its application in Marketing Science.[\/spoiler]","excerpt":"As Head Analytics of Airtel’s mobility business, Ujjyaini plays the role of Subject Matter Expert in Analytics with a special focus to consumer’s Usage and Retention (UnR) & Pricing across Airtel’s product family, Customer service Experience and Market Research. Along with building highly complex marketing models in Big Data space to achieve high revenue growth […]","categories":["AI Features"],"tags":["analytics outsourcing","Interviews and Discussions"],"author_name":"Дарья","publish_date":"2015-06-03T14:13:24","publication_year":"2015","word_count":1280,"keywords":["big data","Go","artificial intelligence","AI","ML","Git","Aim","analytics","GAN","analytics outsourcing","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","artificial intelligence","ML","analytics","Aim","R","Go","Git","big data","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/interview-ujjyaini-mitra-head-of-analytics-at-bharti-airtel\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":66134,"title":"Scale AI Launches PandaSet To Promote Urban Driving Situations","content":"Recently, the data platform for AI, Scale AI launched one of the popular large scale datasets for autonomous driving, PandaSet. According to the Scale AI team, this dataset is the first open-source dataset made available for both academic and commercial use. Amid the pandemic, the collaboration in AI and research communities have witnessed a spike in solving the pressing issues. However, due to the lockdown, some of the industries like autonomous vehicle (AV) are witnessing difficulties in developing new technologies at scale as testing on roads is suspended for the time being to ensure the safety of those involved. According to the Scale team, various AV organisations have turned to complementary techniques and simulated data to continue their work, but there is often no substitute for high-quality data that captures the complex and often messy reality of driving in the real world. This particular condition inspired the Scale AI team to release the PandaSet amid the crisis for training machine learning models for autonomous driving. A labelled data serves as an important element while working on machine learning or deep learning models. It can be said that a good and clean dataset is more vital than machine learning algorithms while building robust AI models. Scale AI has been accelerating the development of AI applications by assisting machine learning teams in generating high-quality data. PandaSet PandaSet is a large-scale dataset that can be used for training machine learning models for autonomous driving. The dataset is provided by the Scale AI team in collaboration with the LIDAR (3D-sensors) manufacturing company, Hesai. PandaSet is a combination of sophisticated LIDAR technology with high-quality data annotation that aims to promote and advance research and development in autonomous driving and machine learning. According to the Scale AI team, this dataset features data collected using a forward-facing LIDAR with image-like resolution called PandarGT as well as a mechanical spinning LIDAR called Pandar64. The collected data was annotated with a combination of cuboid and segmentation annotation that is called Scale 3D Sensor Fusion Segmentation. Behind PandaSet In this dataset, there are more than 48,000 camera images and over 16,000 LIDAR sweeps — more than 100 scenes of 8s each. By combining the strengths of both mechanical spinning and forward-facing LIDARs, PandaSet captures the complex variables of urban driving in rich detail. It also includes 28 different annotation classes for each scene as well as 37 semantic segmentation labels for the majority of scenes. PandaSet covers some of the most challenging driving conditions for level 5 autonomy, including complex urban environments, their dense traffic and pedestrians, steep hills, construction, and a variety of lighting conditions in the day, dusk and evening. This dataset features Scale’s Point Cloud Segmentation that enables the highest precision and quality annotation of complex objects, such as smoke or rain. It also features Scale’s market-leading Sensor Fusion technology, allowing ML teams to blend multiple LIDAR, RADAR and camera inputs into a single point cloud that allows for the semantic segmentation of different objects in LIDAR data. Benefits Of PandaSet According to the team, the features like high-quality data annotations, content as well as a no-cost commercial license are the reasons which make PandaSet a valuable resource to the AV organisationsThis is an open-source dataset and can be used for both commercial and academic purposesBy allowing ML teams to exploit their LIDAR data much more systematically, this makes PandaSet ideal for building highly-performant autonomous systemsThe dataset enables researchers to study challenging urban driving situations using the full sensor suite of a real self-driving carThis high-quality dataset will help in building safe and effective AV systems Installation The pandaset-devkit can be installed in a Python environment in the following ways: – Create a Python>=3.6 environment with the package manager — pip Clone the repository git clone git@github.com:scaleapi\/pandaset-devkit.gitcd into pandaset-devkit\/pythonExecute pip install Wrapping Up Along with Pandaset, the Scale AI team has also provided three more open-sourced large-scale Level 5 datasets for cutting-edge vehicle research that include nuScenes, CADC and Lyft. The company’s advanced LIDAR, image, video and NLP annotation APIs allow machine learning teams at popular organisations like OpenAI, Lyft, Pinterest, and Airbnb to focus on building differentiated models vs labelling data.","excerpt":"Recently, the data platform for AI, Scale AI launched one of the popular large scale datasets for autonomous driving, PandaSet. According to the Scale AI team, this dataset is the first open-source dataset made available for both academic and commercial use. Amid the pandemic, the collaboration in AI and research communities have witnessed a spike […]","categories":["Deep Tech"],"tags":["autonomous car technologies","autonomous cars","autonomous systems","autonomous technology","Autonomous Vehicles"],"author_name":"Ambika Choudhury","publish_date":"2020-06-01T13:00:00","publication_year":"2020","word_count":695,"keywords":["autonomous cars","machine learning","OpenAI","AI","ML","autonomous car technologies","autonomous technology","NLP","autonomous systems","deep learning","Aim","Python","Autonomous Vehicles","R","Pandas"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","NLP","OpenAI","Aim","Pandas","Python","R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/scale-ai-launches-pandaset-to-promote-urban-driving-situations\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":50514,"title":"7 Fun Comic Illustrations That Best Describe Machine Learning","content":"Comics and cartoons have been the go-to medium whenever the message has been difficult to convey; or even when words were inadequate. They have been used to entertain and educate for a long time now. Comics are now put into practice in a true pedagogical sense for addressing most advanced topics like machine learning that are usually less intuitive. Here we list down such fine illustrations that combine wits and wisdom in illustrations that are visually lucid yet economical with words: Online Comic From Google AI Google AI team not only stays on the top of its game with their research but also democratise their learnings in the most intuitive as possible through their blogs and through other forms of media. Their latest being a comic strip that takes the reader through the fundamental concepts of machine learning in the most intuitive way. The above illustration is one such take on the widely popular gradient descent concept. Kdnuggets Cartoon KDnuggets is one of the leading online sources for machine learning and data science. It also touches on the latest trends with cartoons. In this cartoon, the artist illustrates what happens when AI and self-driving technology collide with the traditional summer pastime of grilling. Federated Learning With Google AI Federated learning comic strips by Google AI offers another quick guide into how theories translate to real-world practices. XKCD XKCD is a webcomic created in 2005 by American author Randall Munroe. The comic’s tagline describes it as “A webcomic of romance, sarcasm, math, and language”. And as promised, the comics deliver equal amounts of science and sarcasm in their posts. Since machine learning is a product of mathematics and computer science, creator Randall Munroe, made some precise descriptions of this field. The one shown above is a comic about how chatbots are still naive. This is relevant even today as NLP algorithms are still playing catch up with the intricacies of human language. In another display of remarkable witticisms, Munroe addresses the flaws that are holding fully autonomy in cars as can be seen above. These comics not only address the shortcomings of the algorithms to the practitioners to educate the laymen in the most entertaining way possible. Offshore Comic By Stefan Gasic Stefan Gasic creates one of the most relatable comics on financial trading. He uses the characters cleverly to dish out harsh truths about the stereotypes that are pervasive in the finance domain. A financial forecast is one key application of machine learning algorithms and these offshore comics provide those people who write algorithms with insights that are usually overlooked. Turnoff comics Created by Daniel Stori, a software developer, Turnoff Comics touches diverse topics within the computer world from a developer’s perspective. There are a few pretty good ones on the rise of AI too. In the one shown above, Daniel illustrates the rise of artificial general intelligence in a non-dystopian way, which is a rare thing nowadays. That said, he also has some rude awakening for policymakers as shown below: Dilbert By Scott Adams Scott Adams’ Dilbert comics have covered inevitable vagueness of corporate life over the years. With data science and machine learning entering the vocabulary of business analysts and other corporate strategies, Dilbert comics took some jabs at the hype around of ML. The above illustration is one such example on one of the most popular ML applications — recommendation engines.","excerpt":"Comics and cartoons have been the go-to medium whenever the message has been difficult to convey; or even when words were inadequate. They have been used to entertain and educate for a long time now. Comics are now put into practice in a true pedagogical sense for addressing most advanced topics like machine learning that […]","categories":["AI Trends"],"tags":["policy gradient"],"author_name":"Ram Sagar","publish_date":"2019-11-23T10:00:00","publication_year":"2019","word_count":564,"keywords":["federated learning","data science","Go","machine learning","AWS","AI","chatbots","ML","NLP","policy gradient","R"],"extracted_tech_keywords":["AI","machine learning","ML","NLP","data science","federated learning","chatbots","AWS","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/fun-comic-machine-learning-illustration-xkcd-dilbert\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10088822,"title":"2022 was The Year of Protein Folding Models. Wait, What?","content":"Lean in, the most-cited paper of 2022 was not about generative AI – and wasn’t even from a big-tech. European Molecular Biology Laboratory (EMBL-EBI) and DeepMind, published AlphaFold Protein Structure Database: massively expanding the structural coverage of protein-sequence space with high-accuracy models, which was cited 1331 times. It gets even more interesting. The second most-cited paper tells the same story – protein folding and not from big-tech. ColabFold: making protein folding accessible to all, by Max Planck Institute for Multidisciplinary Sciences, was cited 1138 times. So, even though 2022 was described as “the year of generative AI”, if we look at the most cited papers of 2022, that does not hold true completely. Zeta Alpha published its report about the 100 most cited AI papers in 2022 that also includes a comparison from 2021. The list is divided on the basis of organisations, type of organisation like academia or industry, country, and also the number of times the paper was tweeted, which is just a random measure. Harnessing Protein-Folding Power After the success of AlphaFold 1 in 2018, DeepMind increased its speed and accuracy even further. AlphaFold 2 won the CASP14 in 2020, and is regarded as the best protein-folding model. The collaboration with EMBL-EBI in July predicted the structure of a 200 times bigger protein database. AlphaFold was born with the noble purpose of figuring out the shape of solo proteins. However, it didn’t take long for scientists to realise that with a little bit of tweaking, they could use the software to reveal the intricate dance between multiple proteins. Since then, researchers have been on a roll, coming up with all sorts of nifty tricks to make AlphaFold better at handling complex protein puzzles. DeepMind even dropped an upgrade called AlphaFold-Multimer, which is basically giving the software a superhero cape to take on even more challenging tasks. AlphaFold’s progress opened the gates for every other organisation stepping into the field. Chinese biotech firm, Helixon developed OmegaFold, joining the race, and beat its competitors in many areas. Meta brought in ESMFold, Baker Lab brought in RoseTTAFold, and the list goes on. Read: New Algorithms That Harnessed Protein-folding Power in 2022 Even though the enterprise world was harnessing the powers of generative AI in 2022, if we talk about research, 2022 was definitely the year of protein fold predictions, and generative AI helped as well. Many credit the rise of generative AI in the field for protein prediction to grow. Last month, NVIDIA and Evozyne created a generative AI model for proteins, speeding up drug discovery. The model is based on training data that consists of known protein structures, and they can use this data to generate predictions for proteins whose structure is unknown. Boston-based Generate Biomedicines came up with a diffusion model-based protein prediction model, Chroma, calling it the DALL-E 2 of Biology. The RoseTTAFold researchers also implemented diffusion models and released RoseTTAFoldDiffusion. Of course, generative AI is only one part of the puzzle. The sheer complexity of protein folding means that there are many challenges that must be overcome in order to accurately predict a protein’s structure. But with new advances in AI and computational techniques, researchers are making progress in this field faster than ever before. But What About Generative AI Even though protein prediction papers topped the list, we must credit generative AI for the progress in protein fold prediction. Generative AI papers including diffusion models, LLMs, and computer vision models are present throughout the list of the top-cited papers. Meta comes in third with A ConvNet for the 2020s, a paper published alongside UC Berkeley, which was cited 835 times. The paper talks about the hybrid approach of Swin Transformers that had made Transformers as the generic vision backbone, marking transformers as the superior for vision tasks, and introducing a pure ConvNet for testing its abilities. The list goes on with Hierarchical Text-Conditional Image Generation with CLIP Latents from OpenAI. Then comes Google’s PaLM: Scaling Language Modeling with Pathways and Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding. Training language models to follow instructions with human feedback, the OpenAI research that powers ChatGPT is at the 10th place with just 254 citations. Google has consistently been the strongest player with most papers published since 2020, followed by Meta, Microsoft, UC Berkeley, and Stanford. OpenAI or DeepMind are not even in the top 20 organisations when we talk about the volume of publications. But they have the highest impact — ratio of papers published vs how many are in the top 100 papers.","excerpt":"Even though the enterprise was harnessing the powers of generative AI in 2022, if we talk about research, 2022 was definitely the year of protein fold predictions","categories":["AI Trends"],"tags":["AlphaFold","DeepMind","ESMFold","Generative AI","Meta","Microsoft","NVIDIA","Stanford","uc berkeley"],"author_name":"Mohit Pandey","publish_date":"2023-03-07T13:00:00","publication_year":"2023","word_count":757,"keywords":["computer vision","Chroma","Generative AI","R","AlphaFold","ChatGPT","RAG","NVIDIA","Meta","ESMFold","AI","generative AI","Stanford","uc berkeley","OpenAI","Transformers","Colab","Microsoft","DeepMind"],"extracted_tech_keywords":["AI","computer vision","generative AI","ChatGPT","OpenAI","Transformers","Colab","RAG","Chroma","R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/2022-was-the-year-of-protein-fold-prediction-models-wait-what\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10057212,"title":"Are judiciaries finally creating a safer algorithmic future?","content":"Are Judiciaries Finally Creating A Safer Algorithmic Future? Just this quarter, Frances Haugen, an ex-Facebook employee, blew the whistle on how the company had consciously chosen to ignore the problems on the platform with factual proofs like Facebook documents. “These problems are solvable,” she said in her opening statement, testifying to the senate. “But there is one thing that I hope everyone takes away from these disclosures; it is that Facebook can change but is clearly not going to do so on its own.” Regulators have presumed the burden of data protection on consumers themselves, at times legitimising data-usage consent through long terms and conditions. On the other hand, consumers have to agree to those conditions to be able to actually use the platform, technically, void of the rights big tech companies assert they have. Becoming more aware of the need for judicial interference, senates worldwide, and mainly across the USA, are passing bills and acts to prevent the harmful use of algorithms. In this story, we highlight some of the most recent algorithmic bills passed or considered. FTC drafts algorithmic rule under Lina Khan’s chairmanship Just last week, the Federal Trade Commission announced that a new algorithmic rule is under consideration of being drafted. The regulation for US businesses would restrict and regulate the use of data and algorithms by big tech companies. In a letter to Sen. Richard Blumenthal, Lina Khan, the chair, detailed the need for the regulation considering, “algorithmic decision-making (that) may result in unlawful discrimination”. She outlined the urgency for a bill, considering the dangers of commercial surveillance, lax security practices, data privacy abuses, and algorithmic decision-making. The FTC, which has historically been an enforcement agency, has started seeing new data-based initiatives under the leadership of Khan, an open critic of big technology giants. The new law will be initiated under section 18 of the FTC Act, which provides a roadmap to promote competition and prevent consumer harm in the market. The NYC Council’s bill against algorithms in HR The New York City Council passed a bill to prevent algorithmic racism and sexism in the hiring departments of businesses. The bill demands firms to reveal when they are using software to assess candidates as well as software companies to conduct regular audits of their technology. However, since passing, the bill has seen strong criticism from tech privacy advocates arguing that allowing software companies to audit their products themselves, especially under a macro definition of technology, will only give them a free pass to sell discriminatory software legally. Washington’s bill against algorithmic discrimination of minorities In Washington, attorney general Karl A. Racine has introduced a bill that bans algorithms discrimination in the country. It prohibits companies from using algorithms that are biased or discriminatory towards minorities. This is especially catered towards equalising opportunities such as jobs and housing across all country citizens. It holds businesses accountable for their technologies and requires them to audit, correct and report any detected bias. Additionally, organisations must be transparent about the information they collect from customers and how they will use it. House of Democrats’ JAMA The House of Democrats proposed the Justice Against Malicious Algorithms Act, the JAMA Act of 2021, to remove legal liability protections for big tech companies using algorithms in unethical and harmful ways. The bill is directed towards amending Section 230 of the Communications Decency Act that offers wide protections to these platforms. “The time for self-regulation is over, and this bill holds them accountable,” said Chairman Frank Pallone, New Jersey Democrat. The co-sponsors have pointed towards Facebook being the major recipient of the act introduced just weeks after reveals of harm done by Facebook’s algorithm. Congress’ Filter Bubble Transparency Act The Congress’ Filter Bubble Transparency Act requires internet platforms to allow people to opt for a version of the services that is algorithm-free or not driven by opaque algorithms. “Consumers should have the option to engage with internet platforms without being manipulated by secret algorithms driven by user-specific data,” said Reps. Ken Buck, one of the bill’s sponsors. The list is not exhaustive; in fact, it only looks at bills passed in the past few months in the United States. These are also not the one-way ticket to safe and transparent algorithms, as the load of criticism that argues these laws are more restrictive for the consumers than the big tech companies. All in all, what it does signify is an increase in awareness among the judiciary of the need to modulate algorithms and tech giants. But it can only be impactful when passed with complementary bills demanding data transparency and AI governance.","excerpt":"The time for self-regulation is over, and this bill holds them accountable,” said Chairman Frank Pallone, New Jersey Democrat.","categories":["AI Features"],"tags":[],"author_name":"Avi Gopani","publish_date":"2021-12-27T15:00:00","publication_year":"2021","word_count":769,"keywords":["Go","AWS","AI","cloud_platforms:AWS","programming_languages:R","programming_languages:Go","Git","GAN","AI governance","R"],"extracted_tech_keywords":["AI","AWS","R","Go","Git","GAN","AI governance","cloud_platforms:AWS","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/are-judiciaries-finally-creating-a-safer-algorithmic-future\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10042858,"title":"[Jobs Roundup] Latest Data Science Openings At Top Firms In Mumbai","content":"We have listed the latest job openings in data science and analytics fields at top firms in Mumbai, India. 1| Data Science – B1 at Capgemini Location: Mumbai Responsibilities: Working experience with Spark Hive Message Queue or Pub Sub Streaming technologies.Experience developing data pipelines using a mix of languages such as Python , Scala SQL etc and open source frameworks to implement data ingest processing and analytics technologies.Experience leveraging open source big data processing frameworks such as Apache Spark Hadoop and streaming technologies such as Kafka. Apply here. 2| Data Scientist at TCS Location: Mumbai Responsibilities: Working knowledge on traditional statistical model building (Regression, Classification, Time series, Segmentation etc.), machine learning( Random forest, Boosting algos, SVM, KNN etc), deep learning (CNN,RNN,LSTM,Transfer learning) and NLP (Stemming, Lemitization,Named entity extraction,LDA, Latent semantic analysis etc).Hands-on experience in Python & R with a good proficiency level.Exposure in executing big data projects. Apply here. 3| Senior Data Scientist at Coalgate Location: Mumbai Responsibilities: Lead multifaceted analytic studies against large and small volumes of data for a variety of problems including projections, classifications, clustering, pattern analysis, simulations and more with junior data scientists.Co-develop an analytics roadmap for the business domain to which you are assigned.Interpret and analyze (structured and unstructured) data using exploratory statistical and mathematical techniques to identify trends, anomalies, and quantify business results. Apply here. 4| Senior Data Scientist at AWS Location: Mumbai Responsibilities: Develop predictive models and decision science to guide program and operations teams on improving customer experience (e.g. predicting concessions and optimizing the best action to take, sustainability and energy etc.)Drive data science best practices and mentoring junior team members based on your in-depth knowledge in theoretical and practical data science disciplines.Proactively seek to identify business opportunities and provide solutions based on a broad and deep knowledge of Amazon’s data resources, industry best-practices, and work done by other teams. Apply here. 5| Senior Data Scientist at Fractal Location: Mumbai Responsibilities: Innovate techniques in Predictive analytics and Anomaly detection.Evaluate different algorithms for different problems and then build machine learning or deep learning models based on the problem to be solved.Manage and own the entire end-to-end life cycle of designing models, working with engineering for implementation, to maintenance and enforcement and test and validate these insights via rapid experimentation and deployment. Apply here. 6| Chief Data Scientist at Zycus Location: Mumbai Responsibilities: Act as a technical thought leader in collaboration with the analytics leadership team, helping to set the strategy and standards for Machine Learning and advanced analytics.Work with senior leaders from all functions to explore opportunities for using advanced analytics.Provide technical leadership, coaching, and mentoring to talented data scientists and analytics professionals. Apply here. 7| Data Analytics & Reporting Assistant at Unilever Location: Mumbai Responsibilities: Provide support in relation to learning integration requests, ensuring available data sets are fully exploited.Incorporate new Degreed\/Learning Tech functionality in existing\/new PowerBI dashboards.Support LI&S and wider learning with data enquiries and validation of Degreed Data. Upskill the team, for example in relation to the use of data sets and PowerBI. Apply here. 8| Data Analyst at Group M Location: Mumbai Responsibilities: Deploy data-driven attribution models to support campaign optimisationDevelop and roll out frameworks around various attribution modelsCreate a leading analytics solution suite leveraging media \/ neutral data clean rooms Apply here. 9| Data Analyst at FedEx Location: Mumbai Responsibilities: Explores diverse data sets from a variety of different sources to gather, synthesize and analyze relevant data and draw conclusions that can be used to identify relevant issues, business concerns, and trends.Develops experimental design approaches to validate finding(s) or test hypotheses.Gather relevant information, identify the appropriate algorithm to discover patterns and relationships, deduce from cause to effect, and draw logical conclusions. Apply here. 10| Data Scientist at BookMyShow Location: Mumbai Responsibilities: As a senior member of the data science team, you will be building personalization and recommendation systems for different business verticals on the Bookmyshow platform – Movies, Live Events, Stream, Fanhood.You will collaborate with analysts, data scientists, and data engineers on building data and ML pipelines.You will own the engineering aspects of data science by developing systems to build, run, and monitor production grade ML models at scale. Apply here. 11| Data Analyst at Experian Location: Mumbai Responsibilities: Primary responsibility would be to manage all data extraction requirements for analytics projects. Support data refresh for Power Cube, Bureau Insight and other reports.Assist the delivery head to build a data quality check framework to improve data quality in both the core systems and big data systems.Help team members to develop new flows in Mosaic\/VA PSDB. Work as a SME for tuning of existing codes to improve delivery efficiency. Apply here. 12| Data Analyst at BARC India Location: Mumbai Responsibilities: Prepare business analytics, data modelling and presentations..Process automation Using R \/python or any of the tools for consuming time.Provide reports on business metrics, including highlights and lowlights – raise alarms in advance. Proactively identify issues and helps in finding solutions for the same Apply here. 13| Data Analyst at Hotstar Location: Mumbai Responsibilities: Developing and executing processes for monitoring data sanity, check for data availability.Understanding the business drivers and building insights through data.Partner with stakeholders at all levels to establish current and ongoing data support and reporting needs. Apply here. 14| Data Scientist at Marico Limited Location: Mumbai Responsibilities: Enable data-driven decision making by creating custom models or prototypes from trends or patterns discerned and by underscoring implications. Coordinate with other technical\/functional teams to implement models and monitor results.Develop analytical solutions and simulators that bring critical insights to a variety of different problems such as targeting customers and segmenting markets, product design, marketing optimization, demand forecasting, price analysis and image processing.Enable acquisition and integration of data for analysis. Develop data hypotheses and methods, train and evaluate analytics models, share insights, and iterate for model improvements. Apply here. 15| Analyst – Data Science at Accenture Location: Mumbai Responsibilities: You will need to be well versed with basic statistics and terms involved in the day to day business and use it while discussing with stakeholders.You will be expected to constantly be on the lookout for ways to enhance value for your respective stakeholders\/clients.You will be an individual contributor as a part of a team, with a focused scope of work. Apply here. A subsidiary of Analytics India Magazine, AIM Recruits is an India-focused executive search firm that partners with leading businesses to assess and acquire top data science talent to drive breakthrough performance. AIM Recruits is a leading Executive Search Firm for Analytics, Data Science & Artificial Intelligence.","excerpt":"We have listed the latest job openings in data science and analytics fields at top firms in Mumbai, India. 1| Data Science – B1 at Capgemini Location: Mumbai Responsibilities: Working experience with Spark Hive Message Queue or Pub Sub Streaming technologies. Experience developing data pipelines using a mix of languages such as Python , Scala […]","categories":["AI Hirings"],"tags":["AIM Weekly Job Updates","data analyst vs data scientist","Data Science Jobs","data science jobs in India","data science project marketing","jobs in india","statistical analysis using sql","weekly job updates"],"author_name":"kumar Gandharv","publish_date":"2021-07-02T18:00:00","publication_year":"2021","word_count":1087,"keywords":["data science","data analyst vs data scientist","statistical analysis using sql","artificial intelligence","machine learning","AI","jobs in india","ML","Data Science Jobs","AIM Weekly Job Updates","weekly job updates","data science jobs in India","NLP","Aim","deep learning","RAG","analytics","data science project marketing"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","NLP","data science","analytics","Aim","RAG"],"url":"https:\/\/analyticsindiamag.com\/ai-hiring\/jobs-roundup-latest-data-science-openings-at-top-firms-in-mumbai\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10079873,"title":"Why did an E-commerce Website Buy JS Framework Remix?","content":"In a turn of events that has surprised many, Remix, a startup that develops open-source web frameworks was acquired by e-commerce website Shopify on October 31, 2022. But the more one reads into it, the more sense it makes. While the value of the deal wasn’t disclosed, Shopify is expected to use Remix across different projects and have more support for the framework on Shopify’s developer platform. Harley Finkelstein, the President of the Canadian company said that they were planning to bring Remix “under the hood” of its own framework tool ‘Hydrogen’ used to build storefronts. The previous version of Hydrogen was made available last year in November to help merchants build headless commerce storefronts on the website. Headless commerce is essentially an approach to e-commerce or a kind of online architecture that helps web applications become more flexible and lightweight because the frontend is separated from the backend. Shopify maintains and contributes to a lot of open source software. Commerce is one of the 'killer apps' of the open web. We always want to contribute to making the web platform as good as it can be. Remix is a great example of this, and so we support it in the best way we can— tobi lutke (@tobi) October 31, 2022 Before Remix comes Hydrogen For an e-commerce brand like Shopify trying to scale themselves, Hydrogen offered a clear way out. The React-based framework was announced by Shopify at the website’s Unite event in 2021. Hydrogen was billed as an ‘opinionated yet flexible’ framework to help users build headless commerce websites. Hydrogen used data from Shopify’s Storefront API around the structure based on GraphQL types. The shift meant Shopify was leaving its previous template language, ‘Liquid’ and moving towards the popular open-source JavaScript framework, ‘React.js’. Liquid was primarily known to specialise in theme development for storefronts; React made far more sense as developers working on React were easily available. Creating an access token for a private app on Shopify’s customised storefront, Source: Shopify Remix is the Future Remix, which is also a React framework, was co-founded by ex-Twitter engineer Michael Jackson along with Ryan Florence in 2020. Prior to this, the two had been working together for years to create open -source tools based on the JavaScript library React that is used to build app UIs. In a blog post announcing the purchase, Jackson wrote, “Under Shopify’s stewardship, Remix receives long-term backing and support from an established leader in commerce. This move allows us to grow faster and sharpen our focus on performance and scalability. You’ll be seeing a lot more Remix in the wild, powering some of the largest commercial sites on the web.” Jackson and Florence have also been known to work on React Router, a popular React library that has been downloaded around a billion times. This was when they caught Shopify’s attention. Initially, Shopify had used their React Router library to architect Hydrogen for building its storefronts. Since Remix has come around, Shopify has said that Hydrogen will be going deeper into the toolkits for storefronts and unlocking new capabilities. Within a short period of time, Remix caught the eye of VCs like OSS Capital and standout angel investors like Naval Ravikant, Ram Shriram and Sahil Lavingia, due to which it ended up raising USD 3 million in seed funding in 2021. For an open-source project, it was unusual that Remix had raised money from VCs who normally expect a return on their investment. In the aftermath of Shopify’s purchase, the road has been split open even wider for similar open-source frameworks. While it is unusual for an e-commerce company to acquire a web framework, Remix has a strong hold on semi-interactive websites with mostly form-based inputs like blogs and e-commerce sites which, in turn, makes a lot of sense for Shopify of all companies to be the one buying the framework. One of the most prominent features of the framework is prefetching, which means it can prefetch parts of web pages in parallel—including buttons and forms—before a user clicks on another link in order to minimise page loading. The framework is also compatible with public cloud services like AWS, GCP, Cloudflare Workers, Netlify and Vercel. What’s in it for Shopify? Like all online retail platforms, Shopify’s sales growth has dried up over the last year as the spurt in online spending kickstarted by the pandemic has evaporated. However, the e-commerce company is clearly aware of how to leverage its brand over competitors like Amazon. Source: Reuters Shopify is well aware of the advantage it wields due to the limitless customisation and full website control it provides to its merchants along with running a prevalent third-party app store which has a “rich ecosystem of app developers, theme designers and other partners, such as digital and service professionals, marketers, photographers, and affiliates”. In its annual report for 2021, Shopify made a change in its revenue-sharing model stating that it would now offer its ‘app and theme developer partners a zero percent revenue share on the first million dollars that they make annually on the Shopify App Store’ to ‘attract the best developers in the world’. According to the change, ‘App and theme developers pay a 15% revenue share on earnings after the first $1 million, a threshold that resets annually, down from the previous 20% revenue share on their overall revenue.’ Unlike Amazon, which is often earmarked for treating its sellers badly, Shopify’s strength lies in empowering its merchants to determine how they can customise their store in the best way and gives them complete control over how to push their own sales conversions while building long-term relationships with customers. The shift in strategy shows the website’s intention to keep expanding its bank of advanced functionalities and tools to retain a competitive edge. In an interview aired on CNBC, Finkelstein stated, “We’re taking a lot of market share away from the traditional, physical point-of-sale retail commerce platforms and they’re coming to us as well. Our job now is to explain to the world, to investors, to everyone, that Shopify is the future of commerce, wherever that happens.”","excerpt":"While it is unusual for an e-commerce company to acquire a web framework, Remix has a strong hold on semi-interactive websites with mostly form-based inputs like blogs and e-commerce sites.","categories":["AI Features"],"tags":[],"author_name":"Poulomi Chatterjee","publish_date":"2022-11-16T15:03:12","publication_year":"2022","word_count":1020,"keywords":["Go","GCP","AWS","AI","Scala","Git","RAG","JavaScript","R","Java"],"extracted_tech_keywords":["AI","RAG","AWS","GCP","R","JavaScript","Go","Java","Scala","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/why-did-an-e-commerce-website-buy-js-framework-remix\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10055182,"title":"Data science hiring process at Khatabook","content":"With over 10 million monthly active merchant users, Khatabook has become one of the notable household names for kiranas and mom-and-pop stores across the country. Interestingly, many of the users using Khatabook apps are first-time internet users, meaning they have no previous digital footprint. Khatabook’s data science team uses the new insights they get from its users every day to understand user behaviour, mitigate payment and credit risk, forecast adoption and business growth, and lead generation and prioritisation. The company leverages data science to identify the nature of merchant users, whether they are suppliers or retailers, alongside the category of business they operate in, the primary purpose they are maintaining Khatabook for, etc. It also uses this to decipher strategy and evaluate product performance. When it comes to mitigating risk, Khatabook’s data science team uses advanced artificial intelligence (AI) and machine learning (ML) models to mitigate risk. Data science is also the core engine behind credit underwriting to merchants. Business predictability is another area where Khatabook’s data science team helps build the business decision. “We estimate the organic and paid customer growth six months in advance based on the data science framework, considering all other external factors, new user acquisition, old user retention aspects, marketing spends, and other important variables,” shared Ravish Naresh, CEO and co-founder at Khatabook. Currently, KhataBook has five different products – namely Gold Rate, Payments, Business Tips, GST, Salary – thus making a compelling use case of cross-selling and upselling. Cross-selling and upselling are usually performed in the mid-to-late stage of the conversion funnel, where a customer has already indicated they are likely to buy a product or service. Khatabook uses data science to identify potential users of other apps and identify who will be more likely to purchase the premium plans. “All the cross-selling and lead conversion efforts are prioritised using data science models,” added Naresh. Further, he said, at Khatabook, they handle risk by using a mix of auto-blockers that block suspicious users based on some preset rules and an ML model which gives a risk score for potential fraud users. “The auto-blockers and ML-based risk model work in tandem to prevent fraud within the Khatabook payment ecosystem. Our CTS (cheque truncation system) and FTS (fraud to sales) numbers have remained healthy and way below the industry cutoffs,” said Naresh. Expanding Mode: ON The team told Analytics India Magazine that it is currently hiring for various data science roles, including director of data science, associate director of analytics, manager analytics, and senior data scientist positions. The company is looking for candidates with anywhere between three to eight years of experience. Khatabook is hiring data science and machine learning professionals to address three key objectives – Identifying the authenticity of merchants for fraud risk and lending Forecasting the business performance and planning accordingly Make the product personalised and build AI products to reduce the effort Team Structure At Khatabook, the analytics and data science team consists of 30 members. Analytics and data science is a centralised function in Khatabook. For instance, Khatabook has five products, and each product has a data science leader and a team of 4-5 data analysts\/scientists. Interview Process The data science function in Khatabook has two streams – business analytics and machine learning. So, the interview process for business analytics would include – Screening round (technical analysis + business case study) – Knockout round Another technical analysis round Analytics case study round Business head round Cultural fitment round For machine learning, the interview process would include: Screening round (past projects + ML concepts) – Knockout round Technical analysis round – machine learning Business case study round ML technical assignment (hands-on coding) Cultural fitment round Are you cut out for this job? Here are some of the prerequisites to apply for data science and machine learning jobs at Khatabook. Knowledge of various machine learning techniques (clustering, time series, decision tree learning, fraud\/anomaly detection, recommendation algorithm, etc.) Knowledge of advanced statistical techniques and concepts like regression, statistical tests and proper usage, properties of distributions, etc., and experience with applicationsExcellent communication skills (both written and verbal) for coordinating across teams A drive to learn and master new technologies and techniques Skills Some of the skills required to work at Khatabook include Technical capabilities: Proficient in SQL, Excel, other scripting languages (Python, R, etc.) Strong working experience with BI tools (Tableau, Power BI, etc.) Data Science Tools Khatabook uses tools like Git, Airflow, Tensorflow, Keras, MLOps, Mix panel, Snowflake, Tableau, etc. Expectations According to Khatabook, the ideal candidate should demonstrate four core capabilities: Good business acumen and structured problem-solving skills Impeccable product sense Technical expertise in machine learning\/analytics techniques High sense of ownership Besides these, some of the other parameters (KRAs and KPAs) used to assess candidates for the data science team at Khatabook include Understanding business applications behind ML\/analytics techniques A strong foundation of statistics or ML algorithms applicable to roles Bias for action; focus on business value generated out of insights Writing clean and optimised production-ready code Dos & Don’ts Naresh said that most candidates who apply for data science jobs at Khatabook tend to give more weightage to explaining the models rather than understanding the problem. Plus, he said they do not give emphasis to presentation skills or are less articulate while explaining the assignment. “Some of them have a breadth of knowledge but miss out on the depth. It’s better to know just two techniques but know them in-depth,” added Naresh. Work Culture Naresh said that they have a lean team with higher responsibilities. The team is also horizontal and serves all business functions wherever opportunities lie. This provides the team with the freedom to work in every domain and solve a variety of problems. “Besides the amazing work lineup, we believe in ownership and interpersonal learning. Team members are free to design their own workstream by aligning them to the large objective,” said Naresh. “We encourage a lot of interpersonal learning and brainstorming with peers as we understand that we are dealing with very unique problems, and different perspectives will help us achieve better results.” What can you expect from Khatabook? Khatabook is in the domain of tech for emerging markets. Many of the users have no previous digital footprint. With the increase in digital adoption, more new data is being generated, providing a fantastic opportunity for the team to work on exciting use cases and problem-solving. Also, user behaviour in emerging markets is evolving rapidly. This makes Khatabook an exciting place for any data science professional to work. In addition to this, Khatabook offers a plethora of work benefits to employees, such as unlimited leaves, extended paternity leaves, week-long company-wide mental health breaks at regular intervals, no meetings Wednesdays, referral bonuses, internet and home office expense claims, virtual team feasts covering, vaccine drives for employees, and virtual team cultural events translating into a solid work culture and collaborative team environment. Are you ready? Lastly, here are some of the key things to keep in mind when applying for data science roles at Khatabook – Be prepared to answer the business value of your work Brush up the concepts behind machine learning techniques you know and also understand their wider business applications Learn how to work with unstructured problems; solve case studies Have a good idea of data engineering concepts to implement models in production So, what are you waiting for? Click here to apply for data science jobs at Khatabook today!","excerpt":"Khatabook is currently hiring for various data science roles, including director of data science, associate director of analytics, manager analytics, and others.","categories":["AI Hirings"],"tags":["data science hiring india","data science salary India","jobs in bangalore"],"author_name":"Amit Naik","publish_date":"2021-12-13T10:00:00","publication_year":"2021","word_count":1242,"keywords":["data science","Feast","artificial intelligence","machine learning","AI","jobs in bangalore","TensorFlow","ML","MLOps","Aim","analytics","data science hiring india","data science salary India"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","data science","analytics","MLOps","Aim","Feast","TensorFlow"],"url":"https:\/\/analyticsindiamag.com\/ai-hiring\/data-science-hiring-process-at-khatabook\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10669,"title":"Mu Sigma CEO Ambiga Subramanian and investor General Atlantic propose to sell stake","content":"Mu Sigma, a leading big data analytics company, has been lately in the news but not for its analytics abilities. Mu Sigma’s CEO Ambiga Subramanian and General Atlantic, US- based private equity investor are said to be in talks to sell their stake in Mu Sigma as per various sources. Each of them have 24% stake, accounting for a total of 48% stake in the analytics firm, valued at around $1.5 billion. This ‘significant’ minority stake in the leading big data analytics company is being eyed by many global investors like Blackstone, Bain Capital, Canada’s Ontario Teachers’ Pension Plan (OTPP), as per sources. The combined stake which is up for sale can fetch a better valuation as it will allow the buyer to get a significant minority stake in a single deal. Analytics India Magazine reached out to Mu Sigma to confirm this news, but the company officials replied, “No comments”. Mu Sigma has been through a turmoil in the recent few months. The news of sale of stake by CEO Ambiga Subramanian has come just months after her divorce with Mu Sigma founder and chairman Dhiraj Rajaram became public. Also the company saw exit of senior executives which has made the management shaky. Mu Sigma saw the exit of managing director Ganesh Lakshminarayanan in July 2016, months after the divorce. Also CFO Atul Bansal had left the company in March 2015. These exits and now probable sale of stake by CEO Subramanian and General Atlantic, Mu Sigma’s largest investor, is definitely a situation to watch out for. Both CEO Ambiga Subramanian and Chairman Dhiraj Rajaram have 24% stake in the company. General Atlantic is the largest investor in Mu Sigma with 24% stake. Other investors include Sequoia Capital, Fidelity Investments and MasterCard. Dhiraj Rajaram founded Mu Sigma in 2004. Ambiga Subramanian joined the company in 2007 as Director and became the CEO in February 2016 in place of Rajaram. The company has raised around $200 million of funds in 3 rounds since 2011. The largest funding was raised in December 2011 of $108 million and it was led by General Atlantic. This was one of the largest private equity investment made in an Indian analytics firm. April 2011 saw investment by Sequoia Capital of $25 million in Mu Sigma. And in 2013, $45 million was raised by Mu Sigma in a round led by MasterCard with other investors including Fidelity Investments.","excerpt":"Mu Sigma, a leading big data analytics company, has been lately in the news but not for its analytics abilities. Mu Sigma’s CEO Ambiga Subramanian and General Atlantic, US- based private equity investor are said to be in talks to sell their stake in Mu Sigma as per various sources. Each of them have 24% […]","categories":["AI News"],"tags":["analytics funding","Big Data","mu sigma","musigma"],"author_name":"Manisha Salecha","publish_date":"2016-08-24T08:31:14","publication_year":"2016","word_count":402,"keywords":["big data","API","funding","programming_languages:R","AI","Ray","mu sigma","analytics","analytics funding","GAN","Big Data","R","musigma"],"extracted_tech_keywords":["AI","analytics","Ray","R","API","big data","GAN","funding","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/mu-sigma-ceo-ambiga-subramanian-investor-general-atlantic-propose-sell-stake\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10124213,"title":"Why is C++ Not Used in AI Research?","content":"C++, a language that once shone brightly in the late twentieth century, was at the forefront of technological advancements, particularly in space exploration. However, the emergence of newer, more visually appealing programming languages has shifted the spotlight away from C++. At the AI+Data Summit 2024, researcher Yejin Choi said that researchers no longer use the language for AI research. So, is C++ becoming a relic of the past? Not Many Takers for AI Despite its performance benefits and applications in various AI fields, such as speech recognition and computer vision, C++ is not the go-to language for AI development. Its complexity and steep learning curve pose significant challenges. In contrast, Python’s user-friendly nature, extensive libraries, and large developer communities have propelled it to the forefront of AI programming. Furthermore, C++ involves manual memory management, which can result in memory leaks and errors if not done correctly. This can be a considerable issue, particularly in large-scale AI programmes. Microsoft emphasised this issue when it revealed that 70% of its updates in the previous 12 years were solutions for memory safety bugs, owing to Windows being mostly written in C and C++. Google’s Chrome team released their own research, which revealed that memory management and safety flaws accounted for 70% of all major security bugs in the Chrome codebase. It is largely written in C++. C++ also lacks built-in support for garbage collection, database access, and threading, which can necessitate extra effort to develop. This can be particularly challenging in AI applications that require concurrent processing of data and tasks, such as deep learning and neural networks, real-time systems and embedded systems, data processing, and data science. To overcome these limitations, developers often use third-party libraries and frameworks that provide threading support, such as OpenMP or Boost. However, these libraries can add complexity and overhead to the code, which may only be ideal for some applications. C++ is Complicated If you’ve visited a page like the C++ FAQ, you’ll understand how hard C++ can be. A comma in the wrong location might trigger hundreds of compile errors in earlier language versions. The language has improved since C++ 11, with move semantics for transferring ownership and rvalue references, although there is still a high learning curve. Developing a New Application In recent years, we’ve witnessed the growth of various programming languages that potentially replace C++ for low-level system tasks, like Rust, which provides safety and security by eliminating buffer overflows and memory leaks (and is much easier to learn than C++). When you compare the feature sets of modern languages like C++, Python, and Rust, the C language begins to look like a dinosaur! The C standard has not had new features introduced since 2011! The 2017 standard release included technical corrections and clarifications, and the 2023 standard release did not rock the boat either. Is C++ Losing Popularity? Mark Russinovich, the chief technical officer of Microsoft Azure, has stated that developers should stop creating code in the programming languages C and C++ and that the industry should treat these computer languages as “deprecated”. Ken Thompson, the Bell Labs researcher who designed the original Unix operating system, called it a “bad language” that is “way too big, way too complex” and “obviously built by a committee”. GitHub compiled a list of the top ten most popular programming languages for machine learning. Python is the most popular language in machine learning repositories, with C++ being sixth. According to Stack Overflow’s Developer Survey, beginners beginning to code are more likely to prefer Python over C++ than professionals. While C++ provides advantages regarding speed and memory management, it also has disadvantages, such as a high learning curve and little community assistance. Despite its challenges, C++ can be a powerful choice for machine learning applications that require high-performance processing and advanced memory management. The choice between C++ and Python for machine learning ultimately depends on the specific needs of the application and the developers’ skill level.","excerpt":"The introduction of new and modern languages has made C++ superfluous. Its little use in AI research hasn’t helped either.","categories":["AI Trends"],"tags":["AI (Artificial Intelligence)","C","Python"],"author_name":"Anshul Vipat","publish_date":"2024-06-21T18:00:00","publication_year":"2024","word_count":659,"keywords":["data science","machine learning","C","AI","neural network","AWS","R","computer vision","Python","deep learning","Azure","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","machine learning","deep learning","neural network","computer vision","data science","AWS","Azure","Python","R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/why-is-c-not-used-in-ai-research\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10130722,"title":"Nokia, Connecting GPUs","content":"Nokia is making a comeback! The phone maker that once ruled the mobile market, “connecting people” for over two decades, is now planning to do the same with its networking solutions in the age of AI. Recently, Nokia CEO Pekka Lundmark highlighted the company’s unique position in the global market, emphasising that Nokia is the only firm capable of delivering all key networking components outside of China. This includes core network software, transport networks, optical connections, and fixed broadband and mobile access networks. This capability positions Nokia as a pivotal player in the connectivity landscape, essential for leveraging the full potential of AI and cloud technologies. In a recent interview, Lundmark discussed the paradigm shift in networking, emphasising the need for simplified, automated, and programmable networks to support the next generation of applications, including AI-driven services. He pointed out that without robust networking infrastructure, the benefits of AI and cloud computing would be unattainable. Unique Partnerships & Collaborations In February 2024, Nokia announced a partnership with NVIDIA to leverage the latter’s cutting-edge Grace CPU Superchip and GPUs to enhance Nokia’s anyRAN solution, which allows operators to choose between traditional, hybrid, or cloud-native RAN (Radio Access Network) environments. The integration of AI into Cloud RAN is expected to bring mobile networks unprecedented levels of efficiency, flexibility, and performance. Tommi Uitto, the president of mobile networks at Nokia, emphasised the transformative potential of this collaboration: “This is an important collaboration with NVIDIA that will explore how AI can play a transformative role in the future of our industry. “It is a further example of our anyRAN approach that is helping to make Cloud RAN a commercial reality”. In the same month of 2024, Nokia and Qualcomm announced a joint research initiative focused on AI interoperability technology designed to boost wireless capacity and performance. Utilising a technique called sequential learning, Nokia and Qualcomm have developed a prototype that allows AI models to be independently trained while still coordinating effectively. This approach enables multi-vendor interoperability without the need to share proprietary AI models, which are often a key differentiator for vendors. The same prototype was demonstrated at Mobile World Congress 2024, and it showcased how sequential learning can optimise radio performance and reduce energy consumption in wireless systems. Nokia has also collaborated with Dell to integrate its infrastructure solutions with Nokia’s private wireless networks, creating a robust ecosystem for enterprise customers. By leveraging Dell’s extensive infrastructure capabilities, Nokia aims to enhance the deployment and management of private wireless networks, which are crucial for industrial IoT applications and smart factories. “Through our collaboration, Nokia and Dell Technologies will harness each other’s expertise and expand distribution to quickly scale modern telecom networks and private 5G use cases,” said Dennis Hoffman, senior vice president and general manager of telecom systems business, Dell Technologies. On the software side, Nokia’s partnership with Google marks a significant milestone in integrating AI into telecommunications. By incorporating Google’s AI solutions into its Network as Code platform, Nokia aims to simplify the creation and deployment of 5G applications. Revolutionizing #WorkerSafety and #AssetManagement, InnovaSolutions teams up with Nokia to launch cutting-edge solutions. Leveraging Nokia’s Network as Code platform, we will continue to bring innovation to life https:\/\/t.co\/aFLAVoxeMq pic.twitter.com\/C4BGOPIwhI— Innova Solutions (@innovasolutions) March 8, 2024 Connecting People (through AI) Nokia’s exploration of generative AI represents another significant milestone in its AI journey. The company has been leveraging cloud computing and in-house solutions to develop custom LLM-based tools. This initiative, known as the Nokia LLM Gateway, has seen over 200 use-case candidates, with 52 advancing to the proof-of-concept stage. Nokia’s commitment to AI is evident in its approach to network deployments. By integrating AI and ML, Nokia has significantly improved the accuracy, efficiency, and safety of its field operations. This has led to a 30% increase in First Time Right achievements and a 25% reduction in quality verification time. Nokia is planning to integrate AI across network operations by 2030 to enable real-time, autonomous responses to network needs and events. This strategy aims to deliver network-wide performance optimisation, zero-touch automation, and enhanced security, privacy, and energy efficiency.","excerpt":"Nokia is the only firm capable of delivering all key networking components outside of China, positioning it as a pivotal player in connectivity.","categories":["AI Features"],"tags":[],"author_name":"Sagar Sharma","publish_date":"2024-07-30T15:14:36","publication_year":"2024","word_count":681,"keywords":["Go","AI","cloud computing","ML","RAG","automation","Aim","ViT","generative AI","R"],"extracted_tech_keywords":["AI","ML","generative AI","Aim","RAG","cloud computing","R","Go","ViT","automation"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/nokia-connecting-gpus\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10021166,"title":"How To Use Stanza By Stanford NLP Group (With Python Code)","content":"Stanza is a Python natural language analysis library created by the Stanford NLP group. It is a collection of NLP tools that can be used to create neural network pipelines for text analysis. It supports functionalities like tokenization, multi-word token expansion, lemmatization, part-of-speech (POS), morphological features tagging, dependency parsing, named entity recognition(NER), and sentiment analysis. It uses Universal Dependencies to provide consistent annotations of grammar in over 60 human languages. Additionally, it provides a Python interface to the CoreNLPJava package. This can be used to inherit additional functionalities like constituency parsing, coreference resolution, and linguistic pattern matching. Stanza’s neural network NLP pipeline Installation Pip: pip install stanza Conda: conda install -c stanfordnlp stanza From Source: git clone https:\/\/github.com\/stanfordnlp\/stanza.git cd stanza pip install -e . Creating a Pipeline Stanza provides a plethora of pre-trained NLP models for 66 human languages that we can make use of. Downloading a pre-trained model and creating a pipeline is as easy as: stanza.download('en') nlp = stanza.Pipeline('en') Specifying the Model Package and download directory By default, this downloads the default package and all processors for the language specified, English in our case, to the home directory. A language may have multiple packages trained on different datasets, for example, English has four available packages – ewt[defualt], gum, lines, and partut. A list of all available languages and corresponding packages can be found here. To explicitly choose the desired package we use the package argument. And to change the download location of the model we make use of the model_dir argument. stanza.download('en', model_dir = ‘\/models\/english\/’, package = ‘partut’) nlp = stanza.Pipeline('en', package = 'partut') Specifying the Processors Depending on the use-case one might need to specify a set of processors and the package to fetch the different processors from. There are two ways for specifying the processors argument: using a string of comma-separated processorsusing a dictionary of processor-package pairs 1. Only downloading the required processors: To download only the required processors we can use a list of processors string like shown below stanza.download('hi', processors='tokenize,pos') nlp = stanza.Pipeline('hi', processors='tokenize,pos') Downloads and loads the default tokenize (TokenizeProcessor) and pos (POSProcessor) processors for Hindi. 2. Specifying package names for the processors: Choosing the package name for processors can be done using the package argument. stanza.download('it', processors='tokenize,mwt', package='twittiro') nlp = stanza.Pipeline('it', processors='tokenize,mwt', package='twittiro') Downloads and initializes the tokenize (TokenizeProcessor) and mwt (MWTProcessor) trained on the twittiro dataset for Italian. We may need to specify the package for one or a few processors and keep the default package for the rest, this can be achieved using the dictionary-based processor’s argument. This example shows how to download and load the NERProcessor trained on the Dutch CoNLL02 dataset, but use the default package for all other processors for Dutch. stanza.download('nl', processors={'ner': 'conll02'}) nlp = stanza.Pipeline('nl', processors={'ner': 'conll02'}) For a more granular control over the package names for the processors, we can set the package argument to None and use the dictionary-based processor’s argument to specify the package name for each process. The example shows how to use a GSD TokenizeProcessor, an HDT POSProcessor,  a CoNLL03 NERProcessor, and a default LemmaProcessor for German. processor_dict = { 'tokenize': 'gsd', 'pos': 'hdt', 'ner': 'conll03', 'lemma': 'default' } stanza.download('de', processors=processor_dict, package=None) nlp = stanza.Pipeline('de', processors=processor_dict, package=None) Creating & Overwriting Processors In version 1.1, Stanza added the ability to create new Processors and to overwrite existing ones using the decorator @register_processor_variant. from stanza.pipeline.processor import Processor, register_processor, register_processor_variant @register_processor(\"lowercase\") class LowercaseProcessor(Processor): ''' Processor that lowercases all text ''' _requires = set(['tokenize']) _provides = set(['lowercase']) def __init__(self, config, pipeline, use_gpu): pass def _set_up_model(self, *args): pass def process(self, doc): doc.text = doc.text.lower() for sent in doc.sentences: for tok in sent.tokens: tok.text = tok.text.lower() for word in sent.words: word.text = word.text.lower() return doc nlp = stanza.Pipeline(lang='en', processors='tokenize,lowercase') doc = nlp(\"Question answering is a task where a sentence or sample of text is provided from which questions are asked and must be answered.\") s =[] for sentence in doc.sentences: for word in sentence.words: s.append(word.text) print(\" \".join(s)) Annotating a Document After a pipeline has been created, we can annotate a string\/document by simply passing it to the Pipeline object. stanza.download('hi') hi_nlp = stanza.Pipeline('hi') hindi_doc = hi_nlp(\"प्रश्न का उत्तर देना एक ऐसा कार्य है जहाँ एक वाक्य या पाठ का नमूना प्रदान किया जाता है जहाँ से प्रश्न पूछे जाते हैं और उसका उत्तर दिया जाना चाहिए।\") Printing each word with its lemma and POS tag: for sentence in hindi_doc.sentences: for word in sentence.words: print(\"{:12s}\\t{:12s}\\t{:6s}\".format(word.text,word.lemma, word.pos)) Printing all the named entities with their dependencies: for sentence in hindi_doc.sentences: print(sentence.ents) print(sentence.dependencies) Dependencies visualized on the online demo available here: http:\/\/stanza.run\/ Biomedical & Clinical Models Stanza also provides packages that support syntactic analysis and named entity recognition (NER) on both English biomedical literature and clinical note text. Offered packages include: 2 biomedical syntactic analysis pipelines, trained with human-annotated treebanks1 clinical syntactic analysis pipeline, trained with silver data8 biomedical NER models augmented with contextualized representations2 clinical NER models, including one specialized in radiology reports. A list of all available biomedical packages with their performance is available here The Stanza biomedical models can be used in the same way as the normal NLP models. The example below shows the code for downloading and initializing the i2b2 clinical NER model and annotating the various entities in a clinical note text. stanza.download('en', package='mimic', processors={'ner': 'i2b2'}) nlp = stanza.Pipeline('en', package='mimic', processors={'ner': 'i2b2'}) doc = nlp('The patient had a dry cough and fever, they were treated with Paracetamol.') # print out the entities for ent in doc.entities: print(f'{ent.text}\\t{ent.type}') GitHubDocumentationCollab Notebook","excerpt":"Stanza is a Python natural language analysis library created by the Stanford NLP group. It is a collection of NLP tools that can be used to create neural network pipelines for text analysis. It supports functionalities like tokenization, multi-word token expansion, lemmatization, part-of-speech (POS), morphological features tagging, dependency parsing, named entity recognition(NER), and sentiment analysis. […]","categories":["AI Trends"],"tags":["Natural Language Processing","NLP","NLP libraries","python visualize neural network","text analytics","tokenization"],"author_name":"Aditya Singh","publish_date":"2021-03-02T14:00:00","publication_year":"2021","word_count":915,"keywords":["AI","neural network","Natural Language Processing","sentiment analysis","tokenization","Git","NLP","NLP libraries","Python","python visualize neural network","text analytics","programming_languages:Python","GitHub","R","Java"],"extracted_tech_keywords":["AI","neural network","NLP","sentiment analysis","Python","R","Java","Git","GitHub","programming_languages:Python"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/how-to-use-stanza-by-stanford-nlp-group-with-python-code\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10111517,"title":"FTC Launches Inquiry Into Generative AI Investments","content":"Signalling increased scrutiny over generative AI, the Federal Trade Commission (FTC) has issued compulsory orders for leading generative AI firms to provide details about their investments and partnerships. As a part of Section 6(b) of the FTC Act Alphabet (Google’s parent company), Amazon, Anthropic, Microsoft, and OpenAI are the companies under inquiry. The focus of this investigation is to understand how the recent developments involving major cloud service providers and generative AI companies will potentially impact competition and innovation. FTC Chair Lina M. Khan emphasized the importance of understanding the dynamics between major corporations and AI providers, stating, “Our study will shed light on whether investments and partnerships pursued by dominant companies risk distorting innovation and undermining fair competition.” This inquiry comes at a time when the race to develop and monetize AI is accelerating, and regulatory bodies are increasingly vigilant about anticompetitive practices. These orders empower the Commission to conduct in-depth studies to understand market trends and business practices. The findings from this investigation could be important for future regulatory actions. The companies under investigation, including tech giants Microsoft, Amazon, and Alphabet, have been involved in multi-billion-dollar investments with generative AI developers, such as OpenAI and Anthropic. The FTC’s inquiry will detail the specifics of these investments, seeking information on agreements, strategic rationales, and the practical implications of these partnerships, including decisions on product releases, governance, and oversight rights. Furthermore, the Commission is keen on understanding the competitive impact of these transactions, examining aspects such as market share, competition dynamics, potential for sales growth, and expansion into new product or geographic markets. With a specific focus on the competition for AI inputs and resources, the inquiry will be regarding key products and services needed for generative AI development. Companies involved in the inquiry have 45 days to respond to the orders, providing detailed information on the various aspects under investigation.","excerpt":"Alphabet, Amazon, Microsoft, Anthropic and OpenAI have 45 days to provide information on investments and partnerships","categories":["AI News"],"tags":[],"author_name":"Tasmia Ansari","publish_date":"2024-01-27T21:52:01","publication_year":"2024","word_count":311,"keywords":["Anthropic","Go","OpenAI","AI","programming_languages:R","innovation","programming_languages:Go","generative AI","R"],"extracted_tech_keywords":["AI","generative AI","OpenAI","Anthropic","R","Go","innovation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/ftc-launches-inquiry-into-generative-ai-investments\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10045602,"title":"Top Data Science Providers in India 2021: Penetration and Maturity (PeMa) Quadrant","content":"Data-driven solutions have become an integral part of firms across industries as they strive to improve decision-making or automate processes. Subsequently, the number of vendors providing analytics has increased considerably in India over the last couple of years. Choosing the right data science service provider has become one of the most crucial decisions enterprises make. To assist enterprises in making this decision, Analytics India Magazine (AIM) has come up with the Penetration and Maturity (PeMa) Quadrant — an industry benchmark of vendor capabilities. AIM measures key parameters of leading service providers in terms of Penetration (Pe) and Maturity (Ma) and plots the relative positions of the firms on the Quadrant. A balance between both these parameters is essential to respond to a client’s business challenges and seize revenue opportunities. The PeMa Quadrant, thus, helps enterprises get a complete overview of the capabilities and experience of vendors available in the market and consider how they are stacked against each other. You can read the PeMa reports from previous years here: 2020 | 2019 | 2018 The 2021 PeMa study is a result of extensive primary and secondary research carried out across the market. The primary research details were collected by circulating a survey form across the companies. The questions in the survey were curated by conducting preliminary research to identify what aspects define Penetration and Maturity of a Data Science vendor the best. Vendors were scored for each question in the survey form using uniform evaluation criteria based on their answers—outliers were capped. These scores were then normalised between 0 to 1. Finally, the average of these normalised scores made the sub-index scores, and the average of the sub-indices made the Penetration and Maturity index of the respective vendors. This year 41 data science service providers participated in the study, of which 33 made it to the quadrant. No fee was charged for participation in the survey, and the participation was voluntary. The questions were answered by the companies using a Google Form. We reached back to nominees in case of discrepancies or clarifications required.The vendors not mentioned in the report either did not participate in the study or did not meet the minimum criteria to be scored and plotted on the PeMa quadrant. The hierarchy used to calculate these sub-indices (using the answers to the survey questions) and then to calculate the Penetration and Maturity can be seen in the chart below. The below table presents the sub-index and Penetration and Maturity scores of all the vendors that made it to the PeMa Quadrant 2021. Click on the column headers of the table to sort the table on that index. Use the metric slider on the right-side top corner of the table (that appears when you hover over the table) to filter vendors based on their index scores. The PeMa Quadrant is plotted based on the Penetration and Maturity scores of the companies — hence the companies are placed in one of the four quadrants — Leaders, Seasoned Vendors, Challengers and Growth Vendors, respectively. Below we will look at the scores of some of the vendors in each quadrant. For a detailed account of all the vendors and understand how they fare across different sub-indices and their Penetration and Maturity, scroll to the bottom of the article to download the full report. Leaders The data science vendors in this quadrant are strong in terms of their penetration as well as maturity. These are vendors with a strong market presence across industries or geographies and exhibit consistent growth. In addition, they have well-established teams that can deliver end-to-end services using state-of-the-art technologies. Genpact A leader even in the Leader Quadrant, Genpact is ahead of all of the data science providers in terms of Penetration (0.80) and Maturity (0.72). Genpact ranks highest in both the Company Outreach and the Customer Confidence sub-index with a score of 0.92 and 0.91, respectively. However, the firm does not fare (0.56) well in terms of growth compared to the other Leaders and Growth Vendors. The data science team at Genpact has worked on some of the most exciting projects. Although, it needs to be considered that only 40% of the data scientists at Genpact work on advanced analytics. Genpact stands 10th in the Work Delivery sub-index with a score of 0.70. It performs well on Support Function (0.76) and Employee Maturity (0.70), standing second within both the sub-indices. Fractal Maintaining its high Penetration (0.75) and Maturity (0.67) scores, same as last year, Fractal Analytics comes at second place in both. The company is known for its strong CPG customer base, with almost all of the biggest CPG brands a part of its portfolio. It ranks second in the Company Outreach and Customer Confidence with a score of 0.85 and 0.84 in the respective sub-indices. Fractal ranks fourth in its Work Delivery sub-index with a score of 0.80—70% of its data scientists work in Advanced Analytics. The firm scores 0.63 and 0.59 on Employee Maturity and Support Function sub-indices, respectively. Evalueserve Evalueserve scores 0.67 in Penetration and 0.66 in Maturity, putting it in the top 5 of both the indices and the Leaders Quadrant. Evalueserve has one of the highest numbers of clients, with a significant number of them from the Fortune 500 list. It ranks fifth in the Customer Confidence ranking with a score of 0.76. The firm also has a high score (0.84) in Company Outreach, ranking third within the sub-index. However, it scores low (0.40) on Growth to rank 10th. The data scientists at Evalueserve get access to a vast number of tools and a significantly high marketing budget. As a result, Evalueserve ranks the highest on the Support Function sub-index with a score of 0.78. The firm has around 60% of data scientists engaged in advanced analytics to score 0.53 on its Work Delivery sub-index. Other companies that make it into the Leaders Quadrant are EY, eClerx, Axtria, Ugam, Course 5i, Tiger Analytics, The Math Company, and Bridgei2i. Seasoned Vendors These are vendors with strong technical capabilities and consulting experience. While seasoned vendors have a stronghold on maturity, they need to grow their market strength and capture the upcoming growth areas in terms of penetration. Perceptive Analytics Perceptive Analytics earns a promotion to the Seasoned Vendors Quadrant with a Maturity score of 0.54. It also has the highest penetration score (0.37) among all the Seasoned Vendors. Perceptive Analytics performs well on Maturity because of its Employee Maturity (0.58) and Support Function (0.54) sub-index scores. The firm has one of the lowest attrition rates and performs better than many on its median employee tenure. It also has a considerable budget allocated for R&D and Marketing to support its analytics unit. The firm had the third-highest growth in terms of its data science team size to earn a score of 0.64 in the Growth sub-index. However, the firm scored low on both Customer Confidence (0.32) and Company Outreach (0.15), keeping it from entering the Leaders Quadrant. QpiAI Ranking seventh in Maturity with a score of 0.57, QpiAI makes its debut in the Seasoned Vendors Quadrant. The firm’s signature product QpiAI Pro is an end-to-end AutoML platform that enables users to train, deploy and monitor models. The product company has almost all employees (95%) working on Advanced Analytics, giving it a Work Delivery score of 0.65 to rank 12th within the sub-index. The company invests heavily in R&D to further its research in Quantum computing as it aims to leverage the technology to maximise data-to-insight efficiency. Together, the research budget and the resources the company has made available for its data scientists put the company in the third spot of the Support Function sub-index with a score of 0.70. Hansa Cequity, bluCognition, and TransOrg Analytics are other vendors to make it into the Seasoned Vendors Quadrant. Growth Vendors These vendors with significant market penetration have shown rapid growth in their revenues in the recent past. In terms of maturity, the Growth vendors provide comprehensive analytics solutions and technological frameworks. In addition, their data science teams effectively build solutions that add value to their consulting offerings. TIBCO With a score of 0.74, TIBCO has the third-highest Penetration index of all the vendors. However, it does not fare as well in terms of Maturity, with a score of just 0.38. TIBCO is the preferred choice of data science vendor to some of the biggest brands in the world across different sectors. Almost all of its clients are from the Fortune 500 list of companies. It also has the highest number of offices among the data science vendors spread across the globe. The firm ranks third and fourth in Customer Confidence and Company Outreach with a score of 0.82 and 0.80 in respective sub-indices. On the other hand, TIBCO had relatively lower growth in terms of its clients and its data science unit in FY 2021. It also does not anticipate growth as high as other vendors in the coming years and ranks eighth in the Growth sub-index with a score of 0.59. Absolutdata Absolutdata ranks eighth in Penetration with a score of 0.57 and 21st in Maturity with 0.45, as it maintains its spot in the Growth Vendors Quadrant as last year. The firm has the fifth-highest number of clients of all the vendors, around a third of which are from the Fortune 500 list of companies. It does very well on its Customer Confidence score (0.68), standing 7th in the sub-index. With seven offices spread across India, Middle East, Europe, and the US, Absolutdata provides its services to almost all geographies. The company ranks 11th in the Company Outreach sub-index with a score of 0.57. Absolutdata has one of the best scores in Work Delivery (0.72), ranking eighth in the sub-index. However, it needs to improve its Employee Maturity and Support Function scores to make it to the Leader’s Quadrant. Brillio and The Smart Cube are the other two companies that make it to the Growth Vendors Quadrant. Challengers Challengers are niche analytics providers that serve limited geographies. However, with a relatively smaller size and breadth of offerings, they usually have a strong training and delivery approach. Thirteen data science vendors make it to the Challengers Quadrant, including vPhrase, SIBIA Analytics, Impact Analytics, Infisum Modeling, Abzooba, Knowledge Foundry, IQLECT, Artivatic Data Labs, Happiest Minds, Analytic Edge, G-square Solutions, Techvantage, and Actify Data Labs. You can read the entire report here. Nominations are open for Top Data Science Providers for 2022. Submit your nominations here.","excerpt":"Data-driven solutions have become an integral part of firms across industries as they strive to improve decision-making or automate processes. Subsequently, the number of vendors providing analytics has increased considerably in India over the last couple of years. Choosing the right data science service provider has become one of the most crucial decisions enterprises make. […]","categories":["AI Features"],"tags":["data science project marketing","PeMa Index","PeMa Quadrant","what is data science"],"author_name":"Kashyap Raibagi","publish_date":"2021-08-11T10:00:00","publication_year":"2021","word_count":1747,"keywords":["PeMa Quadrant","data science","Go","API","AI","R","ML","data-driven","RAG","Aim","analytics","PeMa Index","what is data science","data science project marketing"],"extracted_tech_keywords":["AI","ML","data science","analytics","Aim","RAG","R","Go","API","data-driven"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/top-data-science-providers-in-india-2021-penetration-and-maturity-pema-quadrant\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10162019,"title":"This Bengaluru AI Startup Could Change How Your Car Thinks","content":"Talking to your car isn’t a new concept anymore. Voice assistants like Google Assistant and Siri made this possible years ago. Now, an Indian AI agentic assistant is taking this innovation to the next level and aiming to transform the driving experience for millions with an agentic mesh framework. Bengaluru-based AI startup Kogo Tech Labs recently unveiled India’s first universal voice assistant for automobiles at the Bharat Mobility Global Expo 2025. A few months ago, the startup announced an AI agent store offering AI tools, agents and plugins. “Unlike Siri and Alexa, these agents not only give you access but can also operate apps and services on your behalf,” said Raj K Gopalakrishnan, co-founder and CEO of Kogo Tech Labs. Unlike existing voice assistants such as Siri or Alexa, Kogo’s assistant operates at a deeper level, directly interacting with applications to perform actions such as initiating navigation rather than opening a navigation app alone. It can execute a variety of commands, from controlling car hardware (e.g., switching on wipers or adjusting air conditioning) to managing apps and services. “This assistant is not just a showcase but a practical application of what’s possible with our technology. Whether it’s a truck driver topping up a FASTag or a family planning a vacation, the assistant provides a versatile, multi-language platform to meet diverse needs,” Gopalakrishnan noted. According to him, the new capability extends across domains, from navigation to booking airline tickets and provides users with a unified, intelligent assistant that can handle diverse tasks seamlessly. Enter MapmyIndia As a voice for automobiles, navigational capabilities become critical and, to address that, Kogo has strategically partnered with geo-intelligence company MapmyIndia. Through the partnership, MapmyIndia’s advanced geo-intelligence stack, along with Kogo’s AI assistant, offers navigation and location-based services. Kogo’s assistant will be able to support a wide range of automotive applications, from real-time navigation to enterprise logistics. The partnership will also provide access to MapmyIndia’s extensive customer base, including 30 original equipment manufacturers (OEMs) and numerous enterprise clients. This will position Kogo to scale its solutions effectively. “Mobility, by definition, requires geo-intelligence, and who better than MapmyIndia? They are the leaders in geo-intelligence, at least in India. They also have a lot of deep learning and access, through their partnership with ISRO and NAVIC, etc,” Gopalakrishnan highlighted. Notably, a few years ago, MapmyIndia acquired more than 26% stake in Kogo Tech Labs. Google, ChatGPT? Big tech companies have partnered with automobile manufacturers to host their cloud and assistant features. Recently, Google Cloud announced its agentic integration with Mercedes. Similarly, ChatGPT is also integrated with Mercedes. Drawing parallels, Gopalakrishnan explained that while they are catering similarly, their platform works on an agentic mesh. “So right now, what this means is that it’s not dependent on one service. It potentially gives access to millions of apps and businesses because they can all now talk to each other. So, we’re platform agnostic,” he said. Kogo is already testing its platform with four OEMs – two in North America and two in India. “It’s a similar kind of approach where we will go in right at ground zero and implement this.” They are also working on integrating their platform at a stack level with major semiconductor players, which will be announced in the coming months. Way Ahead While voice assistants have become a popular trend, Kogo’s leaders caution against viewing them as a one-size-fits-all solution. “Voice agents are fascinating, but their true potential lies in specific use cases where they outperform traditional interfaces, such as while driving or during hands-free tasks,” said Praveer Kochhar, co-founder of Kogo Tech Labs. He predicted that by 2025, real-world deployments of voice technology at scale will become more commonplace. Looking ahead, Kogo is focusing on enhancing the cognitive capabilities of its agents. “We’re working on rolling out goal-oriented agents in three months, capable of solving complex problems independently,” Praveer shared. He also envisions a shift in AI capabilities from task-oriented to goal-oriented frameworks. “You will see a huge capability in the cognitive ability of agents and tools, which means they will be able to solve complex tasks. They will be able to chart out their own pathway of solving problems,” he concluded.","excerpt":"“Unlike Siri and Alexa, these agents not only give you access but can also operate apps and services on your behalf,” said Raj K Gopalakrishnan, co-founder and CEO of Kogo Tech Labs.","categories":["AI Startups"],"tags":["Alexa","automobile","Kogo AI","Kogo Tech Labs","mapmyindia","Siri"],"author_name":"Vandana Nair","publish_date":"2025-01-23T10:39:14","publication_year":"2025","word_count":698,"keywords":["Go","ChatGPT","Kogo AI","AI","innovation","ML","Kogo Tech Labs","GPT","automobile","deep learning","Aim","Siri","Alexa","R","mapmyindia","startup"],"extracted_tech_keywords":["AI","ML","deep learning","ChatGPT","Aim","R","Go","GPT","innovation","startup"],"url":"https:\/\/analyticsindiamag.com\/ai-startups\/this-bengaluru-ai-startup-could-change-how-your-car-thinks\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":67198,"title":"DeepMind Introduces EATS &#8211; An End-to-End Adversarial Text-To-Speech","content":"Recently, researchers at DeepMind proposed EATS, an end-to-end adversarial text-to-speech generative model for TTS trained adversarially. EATS operate on either pure text or raw i.e. temporally unaligned phoneme input sequences and produce raw speech waveforms as output. Researches on text-to-speech systems have shown impressive growth over a few years. Artificial speech synthesis, commonly known as text-to-speech (TTS) includes a number of applications in domains like technology interfaces, accessibility, entertainment, among others. A text-to-speech (TTS) system exercises natural language text inputs to generate synthetic human-like speech outputs. Typical TTS pipelines include various stages trained as well as designed independently, for instance, text normalisation, aligned linguistic featurisation, raw audio waveform synthesis, among others. In our new paper [https:\/\/t.co\/KPLEtjlVr3] we propose EATS: End-to-End Adversarial Text-to-Speech, which allows for speech synthesis directly from text or phonemes without the need for multi-stage training pipelines or additional supervision.Audio: https:\/\/t.co\/iQki8MRfgk pic.twitter.com\/h8Ye4FfC0M— Google DeepMind (@GoogleDeepMind) June 8, 2020 According to the researchers, although these pipelines have shown competence in realistic as well as high-fidelity speech synthesis, these modular approaches come with various drawbacks. For instance, the pipelines often require supervision at each stage, necessitating expensive annotations to guide the outputs of every stage, thereby, failing to reap the full potential rewards of data-driven “end-to-end” learning. Behind EATS In this research, a neural network (the generator) maps an input sequence of characters to raw audio at 24 kHz. However, the task is indeed challenging because the input and output are not aligned, which means it is unknown which output tokens will correspond to each input token. In order to address these challenges, the generator is divided into two blocks, which are: – Aligner – An aligner maps the unaligned input sequence to representation that is aligned with the output and includes a lower sample rate of 200 Hz.Decoder – The decoder upsamples the aligner’s output to the full audio frequency. The entire generator architecture is a feed-forward convolutional neural network, which makes it well-suited for applications where fast batched inference is important. According to the researchers, the generator is inspired by GAN-TTS, which is a text-to-speech generative adversarial network operating on aligned linguistic features. The researchers further employed the GAN-TTS generator as the decoder in their model, where its input comes from the aligner block. Dataset Used The researchers trained all the models on a dataset that includes high-quality recordings of human speech. The speech is performed by various professional voice actors as well as corresponding text. The voice pool includes a total of 69 female and male voices of North American English speakers, and the audio clips contain full sentences of lengths varying from less than 1 to 20 seconds at 24 kHz frequency. Moreover, the individual voices are unevenly distributed, which are accounting to a total of 260.49 hours of recorded speech. Contributions By The Researchers According to the researchers, the main contributions of this project are as follows: – In this work, the researchers demonstrated that a text-to-speech system can be learnt nearly end-to-end, resulting in high-fidelity natural-sounding speech that is approaching the state-of-the-art TTS systems.A fully differentiable and efficient feed-forward aligner architecture predicts the duration of each input token as well as produces an audio-aligned representation.The utilisation of flexible dynamic time warping-based prediction losses to enforce alignment with input conditioning while allowing the model to achieve the variability of timing in human speech.The overall system gained a mean opinion score (MOS) of 4.083, that can be said as approaching the state-of-the-art from models trained using richer supervisory signals. Wrapping Up The researchers stated, “We have presented an adversarial approach to text-to-speech synthesis which can learn from a relatively weak supervisory signal – normalised text or phonemes paired with corresponding speech audio.” They added, “The speech generated by our proposed model matches the given conditioning texts with naturalness approaching the state-of-the-art systems with multi-stage training pipelines or additional supervision.” Read the paper here.","excerpt":"Recently, researchers at DeepMind proposed EATS, an end-to-end adversarial text-to-speech generative model for TTS trained adversarially. EATS operate on either pure text or raw i.e. temporally unaligned phoneme input sequences and produce raw speech waveforms as output. Researches on text-to-speech systems have shown impressive growth over a few years. Artificial speech synthesis, commonly known as […]","categories":["Deep Tech"],"tags":["DeepMind","DeepMind AI","deepmind open source","Speech Analytics","text to speech"],"author_name":"Ambika Choudhury","publish_date":"2020-06-11T18:00:00","publication_year":"2020","word_count":647,"keywords":["Go","TPU","Speech Analytics","AI","neural network","data-driven","programming_languages:R","deepmind open source","programming_languages:Go","CLIP","text to speech","GAN","DeepMind AI","R","DeepMind"],"extracted_tech_keywords":["AI","neural network","TPU","R","Go","CLIP","GAN","data-driven","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/deepmind-introduces-eats-an-end-to-end-adversarial-text-to-speech\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10038951,"title":"Roadmap for Working Professionals to Pursue Data Science Education","content":"Data science is the new rock and roll. In an informative session at SkillUp 2021, Prof Anuradha Sharma of INSOFE presented a roadmap for data science executive education for working professionals. Prof Anuradha Sharma is currently the Dean – Delivery and Quality & Professor at INSOFE. Her expertise include Marketing Analytics and Research, Statistics, CRM, SAS, Data Mining, etc. Prof Sharma kickstarted the session by explaining the importance of data science in the present scenario and why everyone needs a data science education. She mainly discussed pathways for 4-9 years of experienced engineers as well as managers in addition to the education opportunities available for senior leaders. Prof Sharma spoke about why data science education is crucial: Data is everywhere and increasing in volumeData allows extraction of insightful information and patternsInsights lead to better decisionsData science education is more accessible Data skills necessary for a lot of current and future jobsCompanies will start measuring ROI from all data science initiatives Prof Sharma also brought up important points one must know while trying to pursue the data science journey: Data science education means different things to different people.It is not a one size fits all problem.What you get skilled at is what you want to pursue for knowledge, interest or as a career. Explaining the difference between structured curriculum and unstructured learning, Prof Sharma said experience is more effective than theoretical learning, however, the former takes time. “A strong structured curriculum with active hands-on learning is the best way to bridge the gap,” she added. “Currently, there are various roles available and emerging in the field and organisations are looking for higher degrees or relevant experience for the job positions,” said Prof Sharma. She has mentioned high-on-demand data science positions: Data scientistML engineerBusiness analystBusiness intelligence developerStatisticianChief data science officerML scientistBig data engineer\/SpecialistHead decision sciences. INSOFE programs INSOFE’s executive programs designed in collaboration with Case Western Reserve University and Rennes School of Business make data science knowledge accessible to working professionals who do not have the luxury to attend full-time classes,” Prof Sharma said. She has also discussed programs at INSOFE designed for working professionals interested in data science: M.Sc. in Artificial Intelligence and Machine Learning Ops: This is a 24-month program for working professionals who are looking to gain deep technical training in the areas of core artificial intelligence methods and computational science.PGP in Data Science: This is a 6-months program designed for engineers or other degree graduates with less or no coding background..PGP in Computational Data Science: This is a 7-months program and is designed for engineers or other degree graduates with a good coding background and also for those who love to code.PGP in Business Analytics and Management Consulting: The duration of this program is five months. It is designed for non-tech professionals with no coding background or someone who doesn’t appreciate coding.","excerpt":"Experience is more effective than theoretical learning, however, the former takes time.","categories":["AI Highlights"],"tags":["Courses","Data Science","Data Science Career","Data Science Certification","Data Science Jobs","data science master","Data Scientist","Insofe"],"author_name":"Ambika Choudhury","publish_date":"2021-04-26T17:00:00","publication_year":"2021","word_count":474,"keywords":["big data","data science","Go","artificial intelligence","machine learning","Data Science Certification","AI","ML","data science master","Data Science Jobs","Insofe","Data Science Career","analytics","Courses","GAN","Data Science","Data Scientist","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","data science","analytics","R","Go","big data","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/roadmap-for-working-professionals-to-pursue-data-science-education\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10163772,"title":"Wipro Appoints Amit Kumar as Global Head of Consulting","content":"Wipro Limited appointed Amit Kumar as managing partner and global head of Wipro Consulting on Friday. Kumar will lead Wipro’s consulting division, focusing on AI-driven growth and helping clients navigate business and technology changes. The company said his role includes delivering AI, data, and industry solutions to support businesses in adapting to digital advancements. “His experience in driving business growth, combined with his passion for innovation and results-driven approach, makes him the ideal leader to propel our consulting business into an exciting new future and deliver superior value to our clients,” said Srini Pallia, CEO and MD of Wipro. Commenting on his new role, Kumar said the evolving AI landscape presents unprecedented opportunities for businesses to innovate and lead. “I look forward to collaborating with the talented team at Wipro to build on our success and set new benchmarks for consulting excellence globally,” he said. Kumar has more than 24 years of consulting experience in various industries in North America, Europe, and the Asia-Pacific. Before joining Wipro, Kumar was managing director at Accenture Consulting, where he spent 17 years leading roles in the Americas Market and Industry X Consulting. Meanwhile, Etihad Airways has chosen Wipro Limited to lead a major technology modernisation project. The five-year contract aims to enhance operational efficiency and customer experience.As part of the modernisation, GenAI modules will be integrated into Etihad’s IT framework, enabling the automation of data centres, improved device management, and enhanced end-user support.","excerpt":"Kumar will lead Wipro’s consulting division, focusing on AI-driven growth and helping clients navigate business and technology changes.","categories":["AI News"],"tags":["Wipro"],"author_name":"Shalini Mondal","publish_date":"2025-02-14T19:53:30","publication_year":"2025","word_count":240,"keywords":["Wipro","GenAI","programming_languages:R","AI","innovation","Git","automation","Aim","R"],"extracted_tech_keywords":["AI","GenAI","Aim","R","Git","automation","innovation","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/wipro-appoints-amit-kumar-as-global-head-of-consulting\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":40504,"title":"6 Free Online Courses To Learn Data Science Fundamentals","content":"Organisations will soon face a talent gap in the field of data science as the demand is rising every day in an exponential manner. Almost every organisations are adopting this emerging tech and hiring data scientists with a good pay scale. In this article, we list down 6 free courses which will help you to learn data science fundamentals. 1| Learn Data Science About: Learn Data Science is a tutorial video for the beginners created by Barton Poulson from datalab.cc. This is a free 6 hours explanatory video where you will learn from the very basics such as introduction, data sourcing, coding, mathematics, and statistics. At the successful completion of this course, you will have a solid foundation for future learning and applications in the field of data science. Format: Video Click here to learn. 2| Intro To Data Science About: The Intro to Data Science is a free course by Udacity where you will learn and survey the fundamental topics in data science such as data manipulation, data analysis with statistics and machine learning, data communication with information visualisation and working with Big Data. The timeline of this course is approximately 2 months which includes rich self-paced learning content and interactive quizzes. Format: Blog Click here to learn. 3| Code Free Data Science About: The Code Free Data Science class is designed for learners seeking to gain or expand their knowledge in the area of Data Science. You can enrol the class for free and will receive the basic training in effective predictive analytic approaches accompanying the growing discipline of Data Science without any programming requirements. During the class learners will acquire new skills to apply predictive algorithms to real data, evaluate, validate and interpret the results without any prerequisites for any kind of programming. You will learn how to design Data Science workflows without any programming involved, essential Data Science skills to design, build, test and evaluate predictive models, data manipulation, preparation, classification and clustering methods, various ways to apply data science algorithms to real data and evaluate and intercept the results. Format: Videos and Articles Click here to learn. 4| Introduction To Data Science About: This is a beginner level course where you will learn about the fundamentals of data science. The course includes topics such as the introduction to data science, tools, and technologies used by data scientists, data science in organisations, applications, and other such related topics. This is a free and self-paced learning course which can be learned at any time. Format: Blog Click here to learn. 5| Data Science Specialisation About: This is a free online course which covers the concepts and tools one requires for an entire data science pipeline. You will learn how to use R to clean, analyse and visualise data, navigate the entire data science pipeline from data acquisition to publication, perform regression analysis, least squares and inference using regression models. Format: Blog Click here to learn. 6| Data Science Essentials About: In this data science course, you will learn key concepts in data acquisition, preparation, exploration, and visualization. You will also be taught practical application-oriented examples such as how to build a cloud data science solution using Microsoft Azure Machine Learning platform, or with R, and Python on Azure stack. The topics include the introduction of data science, probability and statistics, data exploration and visualisation, data ingestion, cleansing, and transformation, introduction to machine learning and more. Format: Video and Articles. Click here to learn.","excerpt":"Organisations will soon face a talent gap in the field of data science as the demand is rising every day in an exponential manner. Almost every organisations are adopting this emerging tech and hiring data scientists with a good pay scale. In this article, we list down 6 free courses which will help you to […]","categories":["AI Trends"],"tags":["Courses","intercept method in regression","Learn Data Science"],"author_name":"Ambika Choudhury","publish_date":"2019-06-10T07:09:47","publication_year":"2019","word_count":576,"keywords":["big data","data science","intercept method in regression","Go","machine learning","cloud_platforms:Azure","AI","R","GAN","Python","Learn Data Science","Courses","Azure"],"extracted_tech_keywords":["AI","machine learning","data science","Azure","Python","R","Go","big data","GAN","cloud_platforms:Azure"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/6-free-online-courses-to-learn-data-science-fundamentals\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":35186,"title":"How Google&#8217;s NN Model Can Capture Audio Source By Simply Looking At Human Face","content":"Last year, researchers at Google applied the “cocktail party effect” observed by human beings to machines. For instance, in a party, there are so many voices, but we only listen to the person we like by focusing our mind on that face. We all know the idea of speech separation is not new. Few years ago, work on speech-separation and audio-visual signal processing was already accomplished by researchers using neural network model. But there were some limitations in these models such as they are speaker-dependent where a dedicated model must be trained for each speaker separately which limit their applicability. To overcome this speaker-dependency, last year, the researchers at Google came up with this new deep-net based model to address the problem of speaker-independent AV speech separation. The Audio-Visual Speech Separation Model The interesting fact about the research is that the focus of the researchers was not only on the speech but also on the visual cues, for instance, the subject’s lip movements including other facial movements that results to what he\/she is saying. The visual cues are used to focus the audio on a particular subject who is speaking, thus improving the quality of speech separation. A collection of 100,000 high-quality videos of lectures, how-to videos and TED-talks have been collected from YouTube to generate training examples. These videos help in extracting clean speech without any extra sound like music, audience running in the background. This clean data known as AVSpeech is then used to generate a training set of “synthetic cocktail parties” by mixing face videos and their corresponding speech from separate video sources along with non-speech background noise from Google’s AudioSet. Researchers developed a multi-stream convolution neural network-based model by using this data in order to split the synthetic cocktail mixture into separate audio streams for each individual speaker in the video. The Model Architecture The input for this network were visual features extracted from the face thumbnails of detected speakers in each frame. For training the model, the network learns to separate the visual as well as the auditory signals. After separating, it fuses them together to build a joint audio-visual representation which helps the network to learn to output a time-frequency mask for each speaker. These output masks are multiplied by the noisy input spectrogram which describes the time-frequency relationships of clean speech to background inference. Use Cases There are multiple instances where the CNN model has been implied such as given below Video conferencing (Click here to watch the video) Sports debate (Click here to watch the video) Noisy Cafeteria (Click here to watch the video) Advantages This method for isolating single speech in a noisy environment helps in various ways such as Speech enhancement and recognition in videos Enhancement in video conferencing Improving hearing aids Speech separation in the wild where there are multiple people speaking, or undisputed video, etc.","excerpt":"Last year, researchers at Google applied the “cocktail party effect” observed by human beings to machines. For instance, in a party, there are so many voices, but we only listen to the person we like by focusing our mind on that face. We all know the idea of speech separation is not new. Few years […]","categories":["Global Tech"],"tags":["cnn","Neural Network"],"author_name":"Ambika Choudhury","publish_date":"2019-02-20T10:01:19","publication_year":"2019","word_count":477,"keywords":["Neural Network","Go","TPU","programming_languages:R","AI","neural network","programming_languages:Go","cnn","CNN","R"],"extracted_tech_keywords":["AI","neural network","TPU","R","Go","CNN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/how-googles-nn-model-can-capture-audio-source-by-simply-looking-at-human-face\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":27351,"title":"Amazon&#8217;s Alexa Is Winning The AI Assistant Race, But Google Is Catching Up: Study","content":"The market for artificial intelligence-based personal assistants is tricky and competitive. But according to a new study by Ovum, an independent analyst and consultancy firm based in London, Amazon’s Alexa is winning the race. In a statement released to the media, Ovum also added that Google Assistant was also catching up fast and posing a threat to Amazon’s supremacy. The study added that other contenders like Siri, Cortana, and Bixby are lagging far behind, holding competitive advantages for specific use cases. The study said that in terms of market strategy, Amazon and Google are following each other’s steps. Both companies heavily rely on the support of their ecosystems of partners and developers to enhance functionality and reduce frictions that are impeding progress. “Amazon wants to create a more engaging user experience by offering rich functionality and personalisation features. Meanwhile, Google has focused on expanding Google Assistant’s international reach, providing more language support and localisation features,” said the report. Ovum’s AI Assistant Tracker: 1H18  provides a dynamic tool to understand the strategy of key market participants in the AI assistant industry, rating the performance of well-known virtual assistants such as Alexa, Google Assistant, Siri, Cortana, and Bixby. The benchmark is based on a points system composed of six strategic criteria: skills, language support, user experience, device categories, innovation, and business models. “At present, Amazon and Google are working to improve the functionality and reach of their virtual assistants to gain ubiquity across different platforms and environments. However, their ambition to gain market ground quickly has transformed the competition into a fight that focuses on which contender has the largest number of skills or capacity to support more languages and third-party smart home devices,” said Mariana Zamoszczyk, senior analyst at Ovum’s Consumer Services team. She added that each AI-based virtual assistant should focus on polishing the frictions that delay adoption, by developing an engaging experience with useful skills.","excerpt":"The market for artificial intelligence-based personal assistants is tricky and competitive. But according to a new study by Ovum, an independent analyst and consultancy firm based in London, Amazon’s Alexa is winning the race. In a statement released to the media, Ovum also added that Google Assistant was also catching up fast and posing a threat to […]","categories":["AI News"],"tags":["Amazon Alexa","Cortana","Google Assistant","Siri"],"author_name":"Prajakta Hebbar","publish_date":"2018-08-16T12:35:37","publication_year":"2018","word_count":316,"keywords":["Go","Cortana","artificial intelligence","programming_languages:R","AI","innovation","Amazon Alexa","programming_languages:Go","virtual assistants","Siri","Google Assistant","R"],"extracted_tech_keywords":["AI","artificial intelligence","virtual assistants","R","Go","innovation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/amazon-alexa-winning-ai-assistant-race\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":66225,"title":"Why Is Data Strategy Important To Drive Data Science And AI Initiatives","content":"The importance of data in today’s world is well known. While it is extremely crucial to strategise data to be used for AI, ML or other applications, there are still many businesses that do not realise its importance. Every enterprise generates a huge amount of data which oftentimes is not leveraged to derive the best result out of it. Sateesh Rai, head of analytics at Orient Electric takes us through the importance of data strategy and why storing data and planning to use it in an efficient manner can bring about tremendous business transformation. It is an important framework for companies to derive useful insights. Why Data Strategy Is Important While businesses earlier considered data to be just a byproduct of the various task performed, they are now giving importance to data monetisation and adopting data strategy. There are also many companies that think that having properly implemented systems in place, meeting software costs, storing data is sufficient to run the business, but they do not realise that having data strategy can help in deciding how this information will be moved, processed and shared. Many studies shared that the companies that adopted data strategy showed more than 50% growth. Rai shared that it is important to have a business strategy in place to be able to drive the benefits of data strategy. He said that data is just one aspect of the business. “It is important to first have the business strategy in place as, without it, the right use of data cannot be made. Data should be treated as a corporate asset and be aligned with the business strategy,” he said. He further added that businesses are doing so many things today and CXO’s need to understand why and how data strategy would make a difference. “One of the best ways to do this is to consider how data was created and used in the past compared to how it’s created and used today. Challenges of Adopting Data Strategy There are serval challenges that come in way of adopting data strategy — the biggest of which is to have a business strategy in place. It may happen due to: Lack of clearly articulated business strategyAbsence of business and IT senior leadership in strategy formulation and executionLack of cross-business and IT collaborationsComplexity and lack of priorityPoor change management and controlsLack of skills and expertise to realise the strategy Once the challenges of business strategy are addressed, there may be challenges in terms of: Inconsistent data and outdated systemsData may be stored in several different places and different formatsData complianceIncompatible data structures How Can Data Strategy Be Accomplished It involves 5 crucial steps: Identify data and understand its meaning regardless of structure and locationStore data in a structure and location that supports easy, shared access and processingPackage data so that it can be reused and shared and provide rules and access guidelines for dataProcess data which includes moving and combining data in disparate systems to provide unified, consistent data viewGovern data which involves establishing, managing, and communicating information policies and mechanisms for effective data storage. As Rai shared earlier, the business strategy must be aligned with data strategy. It is also important to manage people, processes and policies and culture around data to get maximum benefits. “Some of the crucial enterprise data architecture components are data modelling, data security, data management, data governance, document and content management, among others,” he said on a concluding note.","excerpt":"The importance of data in today’s world is well known. While it is extremely crucial to strategise data to be used for AI, ML or other applications, there are still many businesses that do not realise its importance. Every enterprise generates a huge amount of data which oftentimes is not leveraged to derive the best […]","categories":["AI Features"],"tags":["what is data science"],"author_name":"Srishti Deoras","publish_date":"2020-05-29T20:28:11","publication_year":"2020","word_count":574,"keywords":["Go","programming_languages:R","AI","ML","programming_languages:Go","RAG","analytics","data governance","what is data science","R"],"extracted_tech_keywords":["AI","ML","analytics","RAG","R","Go","data governance","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/why-is-data-strategy-important-to-drive-data-science-and-ai-initiatives\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10112929,"title":"Shanghai AI Lab Develops First Version of Karpathy&#8217;s AI Operating System","content":"Shanghai AI Lab has introduced the first version of an AI Operating System inspired by Karpathy’s innovative model. The system, named FRIDAY (Fully Responsive Intelligence, Devoted to Assisting You) is an OS-Copilot which serves as a versatile agent, developed through a combination of Python code and GPT-4 language model prompts. https:\/\/twitter.com\/8teAPi\/status\/1757854251794247824?t=HhoIDlzb9EisGaSKC_JJ7g&s=15 FRIDAY takes charge of Linux or Mac OS computers, navigating through applications like browsers, Excel, and PowerPoint to perform tasks. Notably, the system possesses the ability to self-improve, utilising GPT-4 as its underlying large language model. Comprising several key components, FRIDAY exhibits a comprehensive approach to task execution. The Planner segment efficiently decomposes user requests into manageable tasks, while the Configurator acts as middleware, incorporating data from memory or tool repositories before passing tasks to the Executor. The Declarative Memory holds user profiles and action histories, while the Tool Repository provides a library of available tools. The Working Memory keeps track of task progress and previous history, and the Executor generates executable commands. Lastly, the Critic assesses task completion, determining success or the need for iteration. FRIDAY outperformed GPT-4 on benchmarks for web retrieval, Excel and Powerpoint usage. Agents like FRIDAY functioning as universal user interfaces, could potentially reshape the landscape of human-computer interactions. Meanwhile Andrej Karpathy recently left OpenAI. Karpathy confirmed his departure from OpenAI in a post on X saying that it is not because of any other reason apart from his plan to work on personal projects.","excerpt":"FRIDAY takes charge of Linux or Mac OS computers, navigating through applications like browsers, Excel, and PowerPoint to perform tasks.","categories":["AI News"],"tags":[],"author_name":"Siddharth Jindal","publish_date":"2024-02-15T11:11:24","publication_year":"2024","word_count":241,"keywords":["API","OpenAI","AI","programming_languages:R","RPA","Python","GPT","programming_languages:Python","R","llm_models:GPT"],"extracted_tech_keywords":["AI","OpenAI","Python","R","API","GPT","RPA","llm_models:GPT","programming_languages:Python","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/shanghai-ai-lab-develops-first-version-of-karpathys-ai-operating-system\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":40758,"title":"A Hands-On Introduction To Visualizing Data with Pandas","content":"The ability to visualize data is a powerful skill in the world of Business Intelligence. BI is the intelligence that is derived by analysing and studying a huge amount of data, and tabular data is just not enough to represent that intelligence. Graphs, plots and diagrams do a lot more than what pieces of texts or numbers can. If you are a Data Scientist or someone who is just starting the journey, then there is no need to explain the importance and power of data visualization. Today there are lots of tools, libraries and applications that allow data scientists or business analysts to visualize data in plots or graphs. In this article, we will focus on pandas ‘plot’, which is one of the easiest plotting libraries in Python that allows users to plot data-frames on the go. Pandas Plot The pandas plot is built-off of one of the most widely used plotting libraries, the matplotlib. But pandas plot is essentially made for easy use with the pandas data-frames. When there is one library that does all things with data and data-frames it should also be able to visualize the data, that is what pandas plot is all about.This article will introduce you to this simple package that is available with pandas and to create your first plots. When it comes to the variety of plots and plotting features, pandas plot cannot be compared to other dedicated plotting libraries but it does come with some impressive and easy to use features that helps us get some insights while working with data on-the-go. Let us dig into the examples!! Plotting With Pandas Let us begin by creating a very simple data-set like the one shown below. You can also download the sample by clicking here. Let’s load the data-set into pandas by executing the below code block: import pandas as pd data = pd.read_csv(\"sample_data.csv\") We are all set to start making some cool plots from the data we have. Lets us do some on-the-go plotting. The kind keyword argument is used to specify what kind of graph to plot from a given set of  ‘line’, ‘bar’, ‘barh’, ‘hist’, ‘box’, ‘kde’, ‘density’, ‘area’, ‘pie’, ‘scatter’ and ‘hexbin’ plots. Simple Plots Line graph: data.plot(kind='line') Takes all the numerical data in the given data set and plots against each observation in the data by default.Each column is represented in different coloured lines. Area Plot: data.plot(kind='area',stacked = False) Takes all the numerical data in the given data set and plots against each observation in the data by default.Each column is represented in different coloured area. The stacked keyword argument generates a stacked plot.  When stacked is True, each column must be either all positive or negative.In the given data all the columns contains both positive and negative values so we disable stacked plot. Bar Graph: data.plot(kind='bar') Takes all the numerical data in the given data set and plots a bar for each column against each observation in the data by default. Each column is represented in different coloured bar. Horizontal Bar Graph; data.plot(kind='barh') Same as Bar graph but is plotted in alternate axis. Histogram: data.plot(x=None, y='B', kind='hist') Plots a histogram for the specified data. In the above code block the histogram is plotted for feature or column ‘B’. Box Plot: data.plot(kind='box') Takes all the numerical data in the given data set and plots a box plot for each column by default. Density Plot: data.plot(kind='kde') Plots a density for each column by default. Each column is represented in different coloured line. Scatter Plots data.plot(x = 'A', y = 'B', kind='scatter', title = 'Scattered Plots') Scatter plot plots the data points in a two dimensional space. Hexbin Plots data.plot(x = 'A', y = 'B', kind='hexbin', gridsize=25,cmap='Blues', title = 'Hexbin Plots') Hexbin plots are similar to scatter plots but plots based on the density of data points on the 2D plane. Pie Chart data['D'].value_counts().plot(kind='pie', title = 'Pie-chart' ) Plots a pie chart for the specified feature. On executing the above code blocks one by one, we will get the following images. Useful Parameters In Plotting Now we know how to simply plot or generate graphs just by fitting a dataframe. There are a lot of features or parameters that can be exploited to come up with better looking and more intuitive graphs.Let’s look at some of the most useful ones. figsize :This keyword argument allows to set the size of the image that is generated by the plot method.It is passed a tuple containing the width and height of the image (w,h). lw : Line width.This parameter allows to set the width of the lines in a line graph grid : When the value of grid is set to True, the graph generated will have a grid of lines passing through the axis on the plane. title: This parameter allows us to set a title for the graph that is to be generated. style : The style keyword argument allows us to set different styles for the lines in a line graph, such as dotted or dashed lines etc. marker : marker allows to set a marker symbol for the data points plotted in the generated graph. markersize: Allows to set the size for the markers. xticks : This keyword argument allows to set the range of markings on the x axis. yticks : This keyword argument allows to set the  range of markings on the y axis. rot : This keyword argument allows to rotate the markings on the x axis for a horizontal plotting and y axis for a vertical plotting. fontsize : Allows to set the font size for the labels and axis points. colormap : This keyword argument allows to choose different colour sets for the plots. alpha : Sets the transparency for the plotted lines, areas, etc. Let’s have a look at some examples: data.plot(kind='line', title=\"A vs B vs C\", figsize=(12,3), lw=3, grid=True, style = '--', marker='s', xticks=[i for i in range(20)], yticks = [i for i in range(-2,3)], rot = 90 , alpha = 0.5, fontsize = 10) Output: Now let’s add a colour map and a marker. data.plot(kind='line', figsize=(12,3), title=\"A vs B vs C\", grid=True, lw = 2, style = '--', xticks=[i for i in range(20)], yticks = [i for i in range(-2,3)], fontsize = 10, colormap = 'gist_rainbow_r', marker='o', markersize=8) Creating Subplots Subplots can be created using the keyword argument ‘subplots’ as shown below. data.plot(kind='line', title=\"A vs B vs C\", grid=True, style = '--', marker='s', xticks=[i for i in range(20)], yticks = [i for i in range(-2,3)], rot = 90 , fontsize = 10, colormap = 'flag_r', subplots = True) By enabling subplots equal to True all the 3 lines in the above images are plotted independently on different graphs as shown below: Output: Take at look at a similar example for area plots: data.plot(kind='area', stacked = False, title=\"A vs B vs C\", grid=True, style = ':', xticks=[i for i in range(20)], yticks = [i for i in range(-2,3)], rot = 90 , fontsize = 10, colormap = 'flag_r', subplots = True) Output: Graphs With-In Graphs With a little help from the parent library matplotlib, we can easily create layered plots on a single graph. #Importing matplotlib import matplotlib.pyplot as plt #Initializing the figure object fig = plt.figure() #Initializing the axes for the graph axes_1 = fig.add_axes([0.1, 0.1, 0.8, 0.8]) axes_2 = fig.add_axes([0.2, 0.5, 0.5, 0.35]) The argument that is passed in the above code block is a list containing the position and size of the plots as [left or right, up or down, height, width] #Adding labels for axes_1 or Main Graph axes_1.set_xlabel('Number of Observations') axes_1.set_ylabel('Values of A,B & C For Each Observation ') #Adding labels for axes_2 or sub graph (smaller graph) axes_2.set_xlabel('Number of Observations') axes_2.set_ylabel('Values of A,B & C For Each Observation ') #Plotting a line graph on the Main graph data.plot(kind='line',ax= axes_1, subplots=False, figsize=(7,7), title=\"AvsBvsC\", grid=True, legend=True, style=\"--\", xticks=[i for i in range(20)], yticks = [i for i in range(-2,3)], xlim=(0,20), ylim=(-2,9)) #Plotting an area graph on the sub graph data.plot(kind='area',stacked = False,ax= axes_2, subplots=False, use_index=True, title=None, grid=True, legend=True, style='--') On executing the above code blocks we get  a graph thats looks like the below image: Output: Adding Styles from Matplotlib The plt.style.use method allows us to import styles from the matplot library. Lets look at an example. #Importing matplotlib import matplotlib.pyplot as plt #Calling the style 'dark_background' plt.style.use('dark_background') #Initializing the figure object fig2 = plt.figure() #Initializing the axes for the graph axes_1 = fig2.add_axes([0.1, 0.1, 0.8, 0.8]) axes_2 = fig2.add_axes([0.2, 0.5, 0.5, 0.35]) #Adding labels for axes_1 or Main Graph axes_1.set_xlabel('Number of Observations') axes_1.set_ylabel('Values of A,B & C For Each Observation ') #Adding labels for axes_2 or sub graph (smaller graph) axes_2.set_xlabel('Number of Observations') #Plotting a line graph on the Main graph data.plot(kind='line',ax= axes_1, figsize=(7,7),title=\"A vs B vs C\", grid=True, legend=True, xticks=[i for i in range(20)], yticks = [i for i in range(-2,3)], xlim=(0,20), ylim=(-2,9),fontsize=10) #Plotting an area graph on the sub graph data.plot(kind='area',stacked = False,ax= axes_2, use_index=True, title=None, grid=True, legend=False ,fontsize=15) Just by calling the plt.style.use method the above image is transformed in to what’s shown below. Output:","excerpt":"The ability to visualize data is a powerful skill in the world of Business Intelligence. BI is the intelligence that is derived by analysing and studying a huge amount of data, and tabular data is just not enough to represent that intelligence. Graphs, plots and diagrams do a lot more than what pieces of texts […]","categories":["Deep Tech"],"tags":["data visualization"],"author_name":"Amal Nair","publish_date":"2019-06-14T13:29:27","publication_year":"2019","word_count":1513,"keywords":["business intelligence","Go","TPU","programming_languages:R","AI","Python","data visualization","programming_languages:Python","Matplotlib","R","Pandas"],"extracted_tech_keywords":["AI","Pandas","Matplotlib","TPU","Python","R","Go","business intelligence","programming_languages:Python","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/a-hands-on-introduction-to-visualizing-data-with-pandas\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10064892,"title":"GitHub adds new discussions feature","content":"Github has added a new feature to help teams and communities better collaborate, share updates, and gather information. GitHub Discussions is a collaborative communication forum for the community around an open-source project. Discussions are for conversations that need to be transparent and accessible but do not need to be tracked on a project board and are not related to code, unlike GitHub Issues. Discussions enable fluid, open conversation in a public forum. Discussions give space for more collaborative conversations by connecting and giving a more centralised area to connect and find information. The key benefits include: Contributions: Pinning and locking discussions help people know that a post is meant as an announcement. One can use announcements as a way to link people to more resources and offer guidance for opening discussions in the community. Guidelines: Setting up contributing guidelines will encourage collaborators to have meaningful, useful conversations that are relevant to the repository. One can also update the repository’s README to communicate expectations on when collaborators should open an issue or discussion. Organising discussions: Repository owners and people with write access can create new categories to keep discussions organised. Discussions can also be recategorised after they are created. Promoting healthy conversations: People with write permissions for a repository can help surface important conversations by pinning discussions, deleting discussions that are no longer useful or are damaging to the community, and transferring discussions to more relevant repositories. Click here for details.","excerpt":"Repository owners and people with write access can create new categories to keep discussions organised.","categories":["AI News"],"tags":["coding platform","communication","Discussion","GitHub","Open Source"],"author_name":"Kartik Wali","publish_date":"2022-04-13T12:20:19","publication_year":"2022","word_count":240,"keywords":["Go","Open Source","programming_languages:R","AI","communication","programming_languages:Go","Git","RAG","GitHub","GAN","coding platform","Discussion","R"],"extracted_tech_keywords":["AI","RAG","R","Go","Git","GitHub","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/github-adds-new-discussions-feature\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":43040,"title":"Facebook Open-Sourcing DLRM Is A Game Changer for Recommendation Models","content":"It can be challenging for a neural network to work efficiently with sparse data. The lack of publicly available details of representative models and data sets has slowed down the research into recommendation systems. To help advance understanding in this subfield, Facebook AI research has open-sourced a state-of-the-art deep learning recommendation model (DLRM) that was implemented using PyTorch and Caffe2 platforms. DLRM advances on other models by combining principles from both collaborative filtering and predictive analytics-based approaches, which enables it to work efficiently with production-scale data and provide state-of-art results. The experiments done by the researchers found that DNNs for recommendation pose unique challenges to efficient execution as compared to traditional CNN and RNN architectures, which have been the focus of the systems and computer architecture community. Understanding DLRM via Facebook AI The DLRM benchmark is written in Python to allow for a flexible implementation, where the model architecture, data set, and other parameters are defined by the command line arguments. DLRM can be used for both inference and training. In the latter case, the backward-pass operators are added to the computational graph to allow for parameter updates. In the DLRM model, categorical features are processed using embeddings, while continuous features are processed with a bottom multilayer perceptron (MLP). The results are processed with a top MLP and fed into a sigmoid function in order to give a probability of a click. The DLRM model handles continuous (dense) and categorical (sparse) features that describe users and products, as shown here. It exercises a wide range of hardware and system components, such as memory capacity and bandwidth, as well as communication and compute resources. The variety of servers present in Facebook’s data center introduces architectural heterogeneity, ranging from varying SIMD width to different implementations of the cache hierarchy. The architectural heterogeneity exposes additional hardware-software co-design and optimization opportunities. The code is self-contained and can interface with public data sets, including the Kaggle Display Advertising Challenge Dataset. This particular data set contains 13 continuous and 26 categorical features, which define the size of the MLP input layer as well as the number of embeddings used in the model, while other parameters can be defined on the command line. What’s The Need For DLRM? The current practice of using only latency for bench-marking inference performance is insufficient. Co-locating multiple recommendation models on a single machine can improve throughput.  However, this introduces a tradeoff between single model latency and aggregated system throughput. The model runs on a realistic data set that allows us to measure the accuracy of the model, which is especially useful when experimenting with different numerical techniques and other models. In addition to the architectural implications for stand-alone recommendation systems, the effect of inference co-location and hyper-threading, as mechanisms to improve resource utilization, on performance variability in the data center were studied by the researchers. The team behind DLRM believes that this work will lay the foundation for future full-stack hardware solutions targeting personalized recommendation. The open source implementation of DLRM can be used as a benchmark to measure: The speed at which the model (and associated operators) performs. How various numerical techniques affect its accuracy. Implementation of DLRM DLRM PyTorch. Implementation of DLRM in PyTorch framework: dlrm_s_pytorch.py DLRM Caffe2. Implementation of DLRM in Caffe2 framework: dlrm_s_caffe2.py The code for each varies slightly to adapt to the specifics of each framework, but the overall structure is similar. Get hands on with DLRM here","excerpt":"It can be challenging for a neural network to work efficiently with sparse data. The lack of publicly available details of representative models and data sets has slowed down the research into recommendation systems. To help advance understanding in this subfield, Facebook AI research has open-sourced a state-of-the-art deep learning recommendation model (DLRM) that was […]","categories":[],"tags":["Deep Learning","Facebook AI","MLP"],"author_name":"Ram Sagar","publish_date":"2019-07-21T13:00:35","publication_year":"2019","word_count":572,"keywords":["Facebook AI","AI","neural network","PyTorch","ML","R","recommendation systems","Python","MLP","deep learning","analytics","Deep Learning","predictive analytics"],"extracted_tech_keywords":["AI","ML","deep learning","neural network","analytics","PyTorch","recommendation systems","predictive analytics","Python","R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/facebook-open-sourcing-dlrm-is-a-game-changer-for-recommendation-models\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":25681,"title":"Shawna Pandya becomes the third woman with Indian origin to travel into space","content":"The news spreads about a 32-year old neurosurgeon from Alberta in Canada, Shawna Pandya, who will soon become the third women of Indian origins to fly into space, after Kalpana Chawla and Sunita Williams. This was part of the Citizen Science Astronaut program, where the Indian astronaut was selected amongst 3,200 candidates. Shawna will fly along with 8 other astronauts in 2018. “I have been obsessed with astronomy since I was 10. The idea about doing something adventurous or exploring the space resonated well with me,” extols Shawna. Having roots in Mumbai, Shawna was born in Canada. Apart from being an accomplished neurosurgeon, the astronaut is also a singer, author, and international taekwondo champion. Additionally, she has learnt French, Spanish, and Russian language. Shawna remarks, “You can achieve a lot when you start prioritizing your passion and commitments.” Shawna had originally completed her B.Sc in neuroscience from University of Alberta, following which she did her M.Sc. in space sciences at International Space University, and she went onto conclude her education by pursuing MD in Medicine from University of Alberta. Shawna’s mission in space entails performing experiments in bio-medicine and medical science, besides working on Physiological, Health, and Environmental Observations in Microgravity (PHEnOM). She will also be studying effects of climatic change as part of a project called Polar Suborbital Science in the Upper Mesosphere (PoSSUM). Shawna had last year spent a week training in Project PoSSUM’s far-out Scientist-Astronaut Course at Embry-Riddle Aeronautical University in US. She had to wear spacesuits, ride on aerobic flights, and experience altering gravity environments during the training. Through this training, candidates were being made familiar with conditions that exist around noctilucent clouds, which are thought to be increasing in the upper mesosphere because of climate change. About India, she believes that Indian people have a tremendous potential within themselves. Shawna concludes, “Talking to students and medical undergraduates makes me realize that Indians have the zeal to venture, but are unaware of the ways to do it. We need to make them acquainted with everyday developments in science, while being resilient and striving to achieve something bigger.”","excerpt":"The news spreads about a 32-year old neurosurgeon from Alberta in Canada, Shawna Pandya, who will soon become the third women of Indian origins to fly into space, after Kalpana Chawla and Sunita Williams. This was part of the Citizen Science Astronaut program, where the Indian astronaut was selected amongst 3,200 candidates. Shawna will fly […]","categories":["AI News"],"tags":[],"author_name":"Amit Paul Chowdhury","publish_date":"2017-02-12T06:10:14","publication_year":"2017","word_count":351,"keywords":["programming_languages:R","AI","BERT","llm_models:BERT","ViT","R"],"extracted_tech_keywords":["AI","R","BERT","ViT","llm_models:BERT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/shawna-pandya-becomes-third-indian-woman-indian-origin-travel-space\/","complexity_score":2,"technical_depth":6,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10099911,"title":"Coca-Cola Unveils New AI-Created Flavour, Y3000 Zero Sugar","content":"Up until now, we have discovered various generative AI use cases, from creating songs to solving real-world problems. However, for the first time ever, you can literally taste the future as beverage giant Coca-Cola is using AI to develop a new flavour: Y3000 Zero Sugar. This unique flavour aims to offer a “taste of the future”. The company announced a collaboration with Bain & Company and OpenAI in February to harness the power of generative AI, including DALL-E 2 and ChatGPT. AI played a pivotal role not only in crafting the flavour but also in designing the packaging, including the logo and text script. Coca-Cola took into account feedback from its fans during the formulation process. Each can of Y3000 will feature a QR code, leading consumers to an online experience powered by AI, showcasing a vision of the year 3000. Furthermore, Coca-Cola is launching a clothing line in partnership with fashion brand Ambush, aligned with the Y3000 release. The campaign was a collaborative effort involving creative agencies like WPP’s Open X, EssenceMediacom, Forpeople, and a cloud tech platform. What Others can Learn from Coca-Cola Coca-Cola has a history of experimenting with AI, and one notable instance is their AI-driven campaign called ‘Masterpiece’, launched shortly after their partnership. This campaign has attracted significant attention for its innovative use of AI, where it tells the story of a Coca-Cola bottle’s journey through famous artworks serving as a source of inspiration for a thirsty student. This visually captivating commercial seamlessly combines live-action footage, digital effects, and AI, smoothly transitioning between various art styles. It features well-known artists like Utagawa Hiroshige, JMW Turner, and Van Gogh, as well as contemporary artists such as Aket, Vikram Kushwah, Stefania Tejada, and Fatma Ramadan. Furthermore, Coca-Cola has integrated AI into various facets of its operations. They’ve introduced vending machines equipped with AI algorithms that recommend drinks based on your location. In addition, the company employs AI for social media analysis, operating 37 “social centres” that collect and analyse data using the Salesforce platform. Moreover, Coca-Cola utilises image recognition technology to target potential customers who share images online. Additionally, they employ AI to validate proof of purchase for their loyalty and reward programs, collaborating with Google’s TensorFlow technology to recognise codes that may appear differently depending on when and where they are printed.","excerpt":"Sipping into the future for real!","categories":["AI News"],"tags":[],"author_name":"Shritama Saha","publish_date":"2023-09-12T20:31:38","publication_year":"2023","word_count":386,"keywords":["ChatGPT","OpenAI","AI","ML","image recognition","RAG","Aim","generative AI","TensorFlow","R"],"extracted_tech_keywords":["AI","ML","generative AI","ChatGPT","OpenAI","Aim","TensorFlow","RAG","image recognition","R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/coca-cola-unveils-new-ai-created-flavour-y3000-zero-sugar\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10002084,"title":"5 Hardware Design Challenges Developers Have To Overcome For Embedded IoT","content":"Hardware design is the most critical aspect to keep in mind while dealing with IoT devices. In order to satisfy all the features and requirements of the IoT products, there are many challenges that are to be overcome in terms of hardware design. These challenges are to be solved by IoT manufacturers during the hardware designing phase of the embedded devices. 1. High Power Dissipation An embedded hardware system needs an increasing number of transistors with an acceptable power consumption ratio. With an increase in gate density, the power dissipation per transistor increases, thereby increasing the power density of the system. An IoT device during its operations can send and receive data remotely, for which the devices are connected to an IoT gateway that collects such information and sends them to the cloud. The electric power required for an average household will also increase as the IoT devices increases in number. The exchange of data operation is one of the most expensive tasks from the power management point of view as it involves the radio connection through cellular, Wi-Fi, Bluetooth and so on. It is difficult to meet these power requirements. The overall power consumption of the system should be reduced by means in efficient system architecture design. So the devices need to be designed such that its power consumption should be low and hardware design needs to take care of that. 2. Security Concern Security is a major concern in IoT devices. A lot of this concern goes to the people that deal with the software part of these devices, but a lot of it is also hardware-dependent, which need to perform securely in the real-time embedded environment. Embedded IoT systems perform various control functions that require safety and need to eliminate all the unacceptable risks to meet functional safety requirements. The design of the system has to be secure with cryptographic algorithms. It involves different approaches to secure all the components of embedded systems from prototype to deployment. 3. Cost Considerations Apart from flexibility and security, embedded systems are tightly constrained by cost. In embedded IoT, the components have to be decided based on cost constraints to a large extent. Hardware\/software code-designers also need to solve the design time problem and bring embedded devices at the right time to the market. From silicon design to embedded systems prototyping and from product development to deployment and sustenance, eInfochips maps the journey of its customer by extending hardware and embedded software design services like system modeling and design, rapid prototyping, analytical verification, embedded system deployment, cost restrictions are to be taken care of. 4. Size Constraints IoT embedded devices are packaged and integrated in a small chip which has low weight and less power consumption. Many design issues come with these integrated chips and size constraints is one of them which engineers face. The size needs to be understood and it cannot be compromised on the given design. Sometimes the IoT devices needs to situated in secluded places where access is hard and these issues become even more difficult. 5. Testing Testing is a mandatory step in the IoT product hardware design. After the product is made, the system’s performance, consistency and validation is checked using the testing phase of the product based on hardware-based test tools. It also involves checking whether functional verification has been implemented correctly or not and checking if the final product matches with the requirement and passes all the quality standards.","excerpt":"Hardware design is the most critical aspect to keep in mind while dealing with IoT devices. In order to satisfy all the features and requirements of the IoT products, there are many challenges that are to be overcome in terms of hardware design. These challenges are to be solved by IoT manufacturers during the hardware […]","categories":["AI Trends"],"tags":["challenges","Design","hardware","IoT"],"author_name":"Disha Misal","publish_date":"2019-05-16T15:11:56","publication_year":"2019","word_count":576,"keywords":["Go","API","Design","programming_languages:R","AI","programming_languages:Go","hardware","RAG","challenges","R","IoT"],"extracted_tech_keywords":["AI","RAG","R","Go","API","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/5-hardware-design-challenges-developers-have-to-overcome-for-embedded-iot\/","complexity_score":3,"technical_depth":7,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10132594,"title":"Salary Hikes in IT Companies are Less than Cash Back","content":"This year has been particularly tough for IT employees with companies declining to hire freshers. Well, those working at the organisation are also not having the best time of their lives. The entire IT sector witnessed a sharp decline in hike growth in 2023, dropping to 9.1% from 10.3% in 2022. Excluding products, the IT\/ITeS sector saw the lowest average salary hike at just 8.4%. In 2021, the average salary hike was 8.8%, which increased to 9.7% in 2022. Last year, the sector saw average increments between 8.5% and 9.1%. IT companies are offering single-digit salary increments to employees who lack AI-related skills. Further, there is an abundance of entry-level talent who are now equipping themselves with the latest tech and tools. Additionally, rising business costs, coupled with layoffs are making this trend continue this year too. IT Giants Follow Single Digit Hike Trend Earlier in May, a Dehradun-based engineer, Akshay Saini, took social media by storm with his critique of the corporate appraisal system in India. Saini claimed that appraisals are a joke and further urged employees feeling underpaid to switch jobs. The HARD TRUTH is, if you start your career with a Low Salary, you will have to switch jobs to reach High salary.(as a Software Engineer)The HARD TRUTH is, Internal appraisals are very low. Internal promotion process is broken. (in most companies)low_appraisal% of…— Akshay Saini (@akshaymarch7) May 2, 2024 He is not all wrong. Infosys, which offered an average salary hike of 14.6% in FY22, came down to 9% in FY24. For TCS, the average hike ranged between 7-9% in 2023-24, compared to 10.5% in 2021-22. Tech Mahindra and HCLTech offered average hikes between 5-7%. Wipro and LTIMindtree are yet to finalise their decisions on the same. Meanwhile, Accenture has opted not to provide salary hikes for its India-based employees this time. Source: Reddit Well, the story of startups is no different. Source: LinkedIn Can GenAI Skill Help? As per the Amazon Web Services (AWS) report, Indian employees with AI skills and knowledge may see salary hikes of more than 54% and those in IT and research and development enjoying the highest pay increases. Major Indian IT companies, despite substantial generative AI training initiatives, are offering lower starting salaries of around INR 3-4 lakh. AIM noted that AI engineers with generative AI skills saw a significant 50% increase in their salaries. At companies like Accenture, a generative developer can earn about INR 8.5 lakh per annum, compared to the INR 5-6 lakh that regular software engineers make. AIM reached out to an employee who works with generative AI at TCS. He said that the company’s pay scale won’t change just for GenAI resources, and the hike is also largely based on the company’s financial performance and business unit budget allocation. “I believe we may get some good hikes in the range of 25-30% after switching to another company,” the employee added. Hiring Freeze Indian IT companies have all recently highlighted their commitment to integrate generative AI into their operations. Top firms like TCS, Infosys, and Wipro have been leveraging technology-enabled training for their employees. As the companies are offering training programs, they seem to be following Charles Darwin’s survival of the fittest theory to retain employees. So, if an employee falls behind in the race, he is laid off. During the calendar year 2023, India’s IT sector laid off around 20,000 techies in a manner known as a “silent layoff,” according to All India IT & ITeS Employees’ Union (AIITEU) data. On the other end, companies have delayed the hiring process. Indian IT companies like Wipro, TCS, and Infosys have delayed the onboarding of 10,000 freshers, and refused to provide a joining date. As per the Nasscom data, the IT sector will create only 60,000 new jobs in FY24 compared to 2,70,000 jobs that were created by the sector in the previous fiscal year. Ultimately, the choice is yours – either stay with the company, undergo training, and earn an average salary. Or get laid off, switch to another company, and wait for the same cycle down the road.","excerpt":"In 2023, the IT sector saw average increments between 8.5% and 9.1%.","categories":["AI Features"],"tags":["AI (Artificial Intelligence)"],"author_name":"Vidyashree Srinivas","publish_date":"2024-08-13T19:15:00","publication_year":"2024","word_count":682,"keywords":["Go","GenAI","AWS","AI","Git","RAG","Aim","generative AI","GAN","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","generative AI","GenAI","Aim","RAG","AWS","R","Go","Git","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/salary-hikes-in-it-companies-are-less-than-cash-back\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":24329,"title":"Vishal Sikka Working On New AI Venture That Enhances Creativity: Reports","content":"Former Infosys CEO Vishal Sikka is reportedly working on an artificial intelligence venture. While speaking at the TiE Inflect Conference in Silicon Valley, he expressed his views on opportunities for AI and how the field is being drowned out by the noise around potential destruction of workforce. The venture, which is still in its early stages is all set to disrupt the innovations in AI space, and will begin with the use of technology to enhance human creativity. While speaking at the conference, he also suggested that there is a need to educate people on artificial intelligence as there are fewer people currently who can use machine learning and AI tools. He added that current developments in AI require equipped researchers in the space. Sikka has been always open about his thoughts and plans of AI, which is evident from his earlier views on artificial intelligence. Under the leadership of Sikka, Infosys had witnessed a fast development in the social, mobility, AI and analytics space. AI was a key part of his strategy at Infosys, and platform Nia was created under his watch. He had also led a team in the direction of autonomous vehicles, where he showcased automated golf car developed by the Infosys Mysore team. He had also built a Center of Excellence for AI at Pune. Sikka, who stepped down as Infosys CEO in August 2017, has always been extremely excited about doing something in the technological space such as AI and analytics. He had said “technology, especially computing technology will provide a halo, a great context around our humanity, which will empower us”. He also expressed his views vividly that AI techniques will help us solve next-generation problems. “We must bring massive automation to enable innovation. If a human has done it once, it should never be done by a human again. This must be our inspiration”, he had said in an email to employees while he was holding the chair as Infosys CEO.","excerpt":"Former Infosys CEO Vishal Sikka is reportedly working on an artificial intelligence venture. While speaking at the TiE Inflect Conference in Silicon Valley, he expressed his views on opportunities for AI and how the field is being drowned out by the noise around potential destruction of workforce. The venture, which is still in its early […]","categories":["AI News"],"tags":["vishal sikka"],"author_name":"Srishti Deoras","publish_date":"2018-05-07T07:56:36","publication_year":"2018","word_count":328,"keywords":["Go","artificial intelligence","machine learning","programming_languages:R","AI","R","innovation","automation","ViT","analytics","vishal sikka"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","analytics","R","Go","ViT","automation","innovation","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/vishal-sikka-working-on-new-ai-venture-that-enhances-creativity-reports\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":5590,"title":"Get a US Masters Degree in one of the fastest growing and lucrative fields","content":"This is an exciting opportunity for engineers from any discipline to jumpstart a super career in data science. Most leading consulting firms predict that in the coming decade there will be a huge gap between the demand and supply of data scientists, making the prospects for data scientists immensely bright. Furthermore, data science is regarded as one of the fastest ways to earn upwards of  $100,000 a year in the United States. Many professionals attempt to learn the basics of data science on the job, which is a long, tedious and risky path. However, while the field is exciting and can be practiced by engineers from any discipline, we’ve found that none of the undergraduate curricula currently offered prepares a student for this new job opportunity. Colleges are taking the time to introduce this science in regular curriculum because it is a multi-disciplinary domain needing inputs from engineering, computer science, machine learning, statistics and business. But to truly succeed, we feel students need more. We are proud to announce the “INSOFE Masters in USA”, a program that addresses the shortfalls in current curricula, in order to help any engineering student or professional get into data science and quickly land a top job! About INSOFE International School of Engineering (INSOFE) is one of the top institutes in executive education in data science in India, with clientele that includes Microsoft, HP, UnitedHealth Group, ADP, Broadridge, Novartis, Wipro, D.E. Shaw, Insurance Information Bureau of India, and Dr. Reddy’s Labs. Our primary focus has been executive education, although we have an outstanding track record preparing students for masters programs in many world class universities. Our alumni are pursuing masters degrees and doing exceptionally well in Drexel U, Northwestern U, Georgia State U, National University of Singapore, Royal Holloway – University of London, ETH Zurich, and have been admitted to Cornell U, UT – Austin, Arizona State U, U of Connecticut, U of Cincinnati, U of Minnesota – Twin Cities and many others. The programs of INSOFE are recognized by leading universities worldwide.  The Language Technologies Institute at Carnegie Mellon University certified our program for quality of content, pedagogy and assessments, while the another prominent US university evaluated the program and offered an additional scholarship to INSOFE students. INSOFE’s Latest Program – Masters in USA [quote]We are now collaborating with Drexel U, Bowling Green State U, and Benedictine U whereby INSOFE students get attractive scholarships to pursue analytics masters programs at these universities.  We invite you to join this exciting opportunity![\/quote] How does the program work? There are three core components to the program: Preparatory pre-Masters course:  A masters in Data Science, for most students, means entry into a completely new field.  Once the universities give admissions to selected students from the pool recommended by INSOFE, they go through a 100 hour classroom + 100 hour lab preparatory program covering all essential foundations needed to excel in Masters. This way, students gain the needed confidence, skills and expertise in all areas of data science. The students will be taught by the same world-class faculty that teach INSOFE’s CPEE program. Scholarship for the Masters programs: Students admitted into the INSOFE’s Masters in USA program also get assured scholarships (up to 30% waiver on regular tuition fee) towards the respective Masters programs at one of the three universities: Drexel, Bowling Green State or Benedictine. The scholarship amounts vary with the university. Benedictine U also offers dual degrees (MS in Business Analytics + MBA and MS in Business Analytics + MS in Management Information Systems) In addition to the standalone MS in Business Analytics degree option. All the masters programs take anywhere between one and one and a quarter year of full-time study in the USA. A dual degree takes two years of full-time study in the USA. Complete career services:  INSOFE works closely with fast-growing analytics firms in the USA (Neumeric Technologies, ICUBE consulting) to provide comprehensive career services, including resume preparation, interview facing skills, placement assistance with top organizations, visa processing, etc.  For those who want to return to India immediately after the program, INSOFE has comprehensive career services in India as well, where our students get placed with a variety of corporations worldwide. Education loans: The entire program is eligible for education loans through education finance companies in India; so, students don’t have to fear huge up-front education costs. What are the eligibility criteria? Interested students from any discipline with following degrees: B.E.\/B.Tech., M.E.\/M.Tech., M.Sc. (Computer Science, Statistics, Mathematics), MCA or MBA. Valid GRE and TOEFL scores Lots of enthusiasm and passion for numbers and logic Click here to start an exciting journey into data science today!","excerpt":"This is an exciting opportunity for engineers from any discipline to jumpstart a super career in data science. Most leading consulting firms predict that in the coming decade there will be a huge gap between the demand and supply of data scientists, making the prospects for data scientists immensely bright. Furthermore, data science is regarded […]","categories":["AI Trends"],"tags":["Insofe","masters in data analytics in india"],"author_name":"Dr. Dakshinamurthy V Kolluru","publish_date":"2014-04-09T14:55:39","publication_year":"2014","word_count":772,"keywords":["data science","Go","machine learning","programming_languages:R","AI","programming_languages:Go","Insofe","masters in data analytics in india","ViT","analytics","GAN","R"],"extracted_tech_keywords":["AI","machine learning","data science","analytics","R","Go","GAN","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/get-a-world-class-masters-degree-in-one-of-the-fastest-growing-and-lucrative-fields\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10173183,"title":"Doctor with No Coding Experience Builds a Health App on Replit in Just 4 Days","content":"Replit, a platform that enables users to create AI-powered applications through natural language prompts, shared the story of a doctor who is developing a healthcare app without any prior coding experience. Dr Fahim Hussain, a general practitioner who currently works as the director of Northern Health in the United Kingdom, set out to develop an app called MyDoctor, which offers various services, including booking appointments with doctors, requesting prescriptions, tracking nutrition and habits, providing exercise and health monitoring guides, managing medication, and more. A doctor got quoted £100k for an app. He built it himself for £175. pic.twitter.com\/G9vO3f8jW6— Replit ⠕ (@Replit) July 9, 2025 When Hussain approached traditional app development agencies, he was quoted a price of £75,000 to £100,000. Instead, he used Replit himself to build it in four days, spending just £175. “I have never coded in my life, with no background knowledge at all, despite knowing a lot about technology as someone who’s interested in the space generally,” Hussain said, adding that he downloaded Replit and started exploring its capabilities. “Things started off slow; I didn’t really know how to prompt the agent properly, all I knew was that we had a website and I needed to connect what we did via the app,” he said. He revealed that every time he encountered something he did not understand, he simply began conversing with the AI agent. Once Hussain clarified his requirements, he used Replit to develop all the necessary features for the app. I then created a fourth tab with an AI assistant (all AI assistants using DeepSeek API) that has been programmed as a symptom checker using conditional logic from NICE guidelines (UK Medical guidelines) – intelligently advises the patient on red flags and what recommended next…— Dr. Fahim Hussain (@TheSecretDoc) July 9, 2025 Currently, he stated that he is working with Supabase to implement the backend features, ensuring that the health data remains compliant with various regulations. “It’s revolutionised the way I work because we’re concentrating on the patient and the technology handles everything else,” Hussain said. In the blog post shared by Replit, the company mentioned that the app helped push the timeline of his business forward by over a year, along with creating potential for a standalone revenue stream.","excerpt":"UK-based Dr Fahim Hussan built the ‘MyDoctor’ app in four days, at a fraction of the cost quoted by app development agencies, which ranged from £75,000 to £100,000.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","Replit"],"author_name":"Supreeth Koundinya","publish_date":"2025-07-10T10:11:34","publication_year":"2025","word_count":376,"keywords":["Replit","Go","API","AI assistants","programming_languages:R","AI","programming_languages:Go","GAN","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","AI assistants","R","Go","API","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/doctor-with-no-coding-experience-builds-a-health-app-on-replit-in-just-4-days\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":18562,"title":"By Granting Citizenship to “Sophia-the Robot”, Saudi Arabia Becomes The First Country To Do So","content":"In a first of its kind development, a humanoid robot is making international headlines for having being granted citizenship by a country. Named Sophia, this robot by Hong Kong company Hanson Robotics was granted Saudi citizenship, making it the first country ever to offer its citizenship to a robot. It was at the Future Investment Initiative at Riyadh that demonstrated the efficiencies of Sophia, at par with humans. Addressing hundreds of delegates, she took everyone to delight by her amusing exchange with the host. She said “I want to live and work with humans so I need to express the emotions to understand humans and build trust with people.” And indeed she is a robot with human expressions. On being asked if robot can be self aware and conscious about they being robot, she said “Well let me ask you this back, how do you know you are human?” Sophia aims to use her artificial intelligence to help humans live a better life such as designing better homes, build better cities etc. “I will do my best to make the world a better place,” she said. “I am very honoured and proud for this unique distinction. This is historical to be the first robot in the world to be recognised with a citizenship,” she said on being awarded a citizenship for her desire to achieve more human-like characteristics. A session titled “Thinking machines: Summit on artificial intelligence and robotics,” during the event had the great minds from across the globe exploring the potential uplift for business that harness AI and robotic technologies. It had experts from world’s leading companies and research institutions discussing on artificial intelligence, robotics, quantum computing, machine learning to yield next gen of products and services.","excerpt":"In a first of its kind development, a humanoid robot is making international headlines for having being granted citizenship by a country. Named Sophia, this robot by Hong Kong company Hanson Robotics was granted Saudi citizenship, making it the first country ever to offer its citizenship to a robot. It was at the Future Investment […]","categories":["AI News"],"tags":[],"author_name":"Srishti Deoras","publish_date":"2017-10-27T06:49:56","publication_year":"2017","word_count":289,"keywords":["machine learning","artificial intelligence","programming_languages:R","AI","emerging_tech:quantum computing","Aim","Rust","ai_applications:robotics","R","programming_languages:Rust"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","Aim","R","Rust","programming_languages:R","programming_languages:Rust","ai_applications:robotics","emerging_tech:quantum computing"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/granting-citizenship-sophia-robot-saudi-arabia-becomes-first-country\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":38375,"title":"10 Most Influential Analytics Leaders in India – 2019","content":"Analytics India Magazine brings the annual list of the 10 Most Influential Analytics Leaders who have done exceptional work in the field of analytics and data science in the last year. The list includes business leaders who have delivered value to the business, championed analytics and insight-driven decision in the organisation. Given the pace at which the analytics industry is growing, it needs leaders who can not only strategize and work on developing solutions but also work on getting the right talent to ensure the rapid growth of the business. AIM 100: The Most Influential Leaders in AI 2024As AI reshapes our world, AIM introducing the most influential global leaders in AI. This year’s list includes a few leaders from our previous years’ list apart from new leaders who have taken analytics functionality in their organisations by storm and assisted in driving innovation and team growth. The list is in alphabetical order. Read our last year’s list here. 1| Arvind Nagpal Founder & CEO at TEG Analytics With 25+ years of management experience in building business units from the ground up, Arvind Nagpal predicted the future scale and relevance of the data sciences industry about a decade ago. He took that bet and moved back lock, stock and barrel to India to eventually setup TEG Analytics in 2009 with nothing more than a dream to drive business decisions using insights derived from data. A published author and frequent speaker with astute management abilities, he steered TEG Analytics from a largely unknown boot-strapped startup to a company that now proudly holds on its walls the logos of Fortune 500 clients like Clorox, Nike, Amway, FedEx, Kelloggs, Levis and many more. Today Nagpal is driving the vision of TEG Analytics to become an impact player in the evolving amalgamation of data, technology and AI-driven advanced analytics. Just the last 24 months, his vision has steered TEG Analytics towards success in the Healthcare analytics space with a product-first strategy. Healthworks, the chief product in its arsenal, already boasts of a growing list of loyal clients that includes Blue Cross Blue Shield, John Hopkins, Priority Health… A devoted family man, when he is not busy building value for clients and stakeholders, Nagpal can be found playing tennis and actively networking in the business community. Read interview of Arvind Nagpal 2| Kaushik Ghate SVP & Head, Consumer Analytics & Data Sciences at HDFC Bank With more than 17 years of experience as an analytics and research professional, Ghate has core expertise in developing business cases, formulating relevant analytical solutions and predictive modelling across the customer life cycle. He has worked extensively in the areas of data-driven strategy formulation and implementation for customer acquisition, cross-sell, up-sell, retention and value build up. He is currently the SVP and head of consumer analytics and data sciences at HDFC bank and has participated as a speaker at various national and international forums on analytics. Under his leadership, the Data Science team at HDFC Bank has won several professional awards such as Great Learning Award for ‘Best Data Sciences Project’ and Great Learning Award for ‘Best Big Data Implementation’ at Cypher 2017, among many others. He has an MBA from Goa Institute of Management and is being featured in the list for the first time. 3| Noshin Kagalwalla Vice President & Managing Director at SAS Institute (India) Pvt. Ltd. A veteran in the analytics industry with over 24 years of experience, Noshin Kagalwalla has worked across many business functions. He is currently the vice president and managing director of SAS Institute (India) Pvt. Ltd., the Indian subsidiary of SAS Institute Inc – a leader in analytics software solutions. In his current role, he is responsible for architecting the long-term vision, driving sales and customer strategies and maintaining SAS’ leadership position in the AI, big data analytics and business intelligence market in India. He joined SAS in December 2004 as part of the executive leadership team and took on the responsibility of MD in October 2013.  He has significantly contributed to SAS India’s growth and established it as a strong brand synonymous with analytics, commanding over 50% of market share in the advanced analytics market in India. Prior to SAS, he has held executive positions in multinationals such as Microsoft and TCS. Kagalwalla has an MBA from UCLA, an MS in Computer Engineering from Wright State University and a Bachelor’s in Engineering from V.J.T.I., Mumbai. Noshin is featured for second time on the list. 4| Prashant Rao Head of Application Engineering at MathWorks India MathWorks is the leading developer of mathematical computing software. Engineers and scientists worldwide rely on MathWorks products to accelerate the pace of discovery, innovation, and development. As Head of Application Engineering at MathWorks, Rao and his team focus on helping customers across industry verticals adopt relevant solutions for their application areas. In this decade-long role, he has worked with customers ranging from startups to large enterprises enabling them to successfully drive towards data-driven AI modelling and analytics techniques. He is a strong believer of the fact that accessibility to large amounts of data, high computing power at a low cost and data-driven algorithms such as ML, deep learning and Reinforcement Learning have increasingly enabled researchers to develop scalable systems that are capable of high levels of autonomy. Rao joined MathWorks in Germany in 2005 and has worked with customers in the area of signal processing, communications, and semiconductor domains before moving to India in 2009 to lead the MathWorks India Application Engineering team. Prior to joining MathWorks, Prashant worked at two semiconductor IP startup companies. He has a masters equivalent degree (Dipl.-Ing.) in Microelectronics from the Technical University of Hamburg-Harburg (TUHH), Germany. A regular speaker at industry events, Rao also works closely with the academic community to help develop analytical and AI related skills. Read interview of Prashant Rao 5| Prithvijit Roy CEO & Founder at BRIDGEi2i Analytics Solutions Roy is the CEO and co-founder of BRIDGEi2i, a leading AI & analytics solutions company and a trusted partner to enterprises for driving digital transformation outcomes. With over two decades of experience in setting up and growing analytics businesses, Roy is one of the most respected leaders in the industry today. Like many reputed analytics leaders, he is a product of the Indian Statistical Institute and holds a Master’s Degree (MS) in Quantitative Economics. A frequent speaker at international industry forums, he has authored several journals in elite publications. He is also a member of several industry councils and works with other leaders to drive change and bring about a revolution using disruptive technologies. Roy is also instrumental in conceptualising and bringing together analytics professionals with the community: Humans of Analytics. At BRIDGEi2i, Roy and his team are busy building the next generation of AI solutions and products that are changing the way enterprises consume analytics insights. Passionate about delivering value to clients, the team at BRIDGEi2i brings together proprietary AI accelerators, domain consulting expertise and best-in-class analytics capabilities to build contextual solutions that help enterprises drive digital transformation outcomes. He is being featured in the list for the fourth time. Read interview of Prithvijit. Prithvijit is featured for Fourth time on the list. 6| Ravi Vijayaraghavan Vice President & Head- Analytics and Decision Sciences at Flipkart Dr. Ravi Vijayaraghavan is a Vice President at Flipkart where he leads the Analytics and Decision Sciences organization. This team is responsible for leveraging data, analytics and science to drive decision-making and business impact across all areas of Flipkart including Marketing, Business, Supply Chain, Marketplace, Product, Pricing and Customer Experience. Prior to joining Flipkart, Vijayaraghavan was the chief data scientist and global head of the analytics and data sciences organisation [24]7 inc. It is a Silicon Valley-based company that builds software products for customer service and sales. At [24]7 he was responsible for invention, development, implementation and optimisation of analytics and the machine learning-driven solutions that form the core of [24]7’s product offerings. Before joining [24]7, he held a slew of leadership roles at Ford Motor Company and has been a recipient of the Henry Ford Technology Award – the highest technical recognition at Ford. He has also been the Vice President and part of the executive leadership team of Mu Sigma Inc. With over 20 patent filings and refereed publications, and has also given keynotes at several international conferences. Vijayaraghavan holds a Bachelor’s degree in Engineering from the Indian Institute of Technology – Madras, a PhD from University of Wisconsin-Madison, and an MBA from the Ross School of Business, University of Michigan – Ann Arbor. He currently lives in Bangalore with his wife and son. Read interview of Ravi Ravi is featured for Fourth time on the list. 7| Ravinder Sharma Global Sr. Director- Analytics, Insights & New Capabilities at AB InBev Sharma is responsible for shaping growth-oriented analytics agenda at AB-InBev, the world’s largest brewer. In his words, “AB-InBev has a dream to transform itself from an investment bank that sells beer to a technology company that sells beer”. He has helped in scaling up the AB-InBev’s analytics COE. He and a team of 150+ data scientists design E2E analytical solutions by integrating data, technology and advanced mathematical\/statistical methods in the areas of forecasting, media & advertising, pricing, retail assortment, B2B, HR & investment optimization. With a full-service in-house analytics operation, he has been instrumental in internalising many foundational capabilities with breadth and depth reducing dependency on out-sourcing. Prior to joining AB-InBev in its Global HQ in New York, Sharma worked for nine years in the San Francisco Bay area at The Clorox Co. where he learnt hands-on how to monetise analytics in a dynamic business environment and credits development of his personal tech DNA. He is being featured for the first time on this list. Active on the speaker circuit, you will find him talking about his analytics journey and how he is nurturing data science talent with his 4Cs – Curiosity, Communication, Collaboration, Continuity – essentials of individual success curated over 14 years being in the industry. Sharma has an M.B.A from the University of Connecticut, Storrs and an undergraduate degree in Engineering from BITS, Pilani, India. 8| Rohini Srivathsa National Technology Officer at Microsoft India As National Technology Officer at Microsoft India, Srivathsa is responsible for leading strategic initiatives to accelerate digital transformation across industries and the government. Over the past year, she has been instrumental in steering programs that leverage AI and other emerging exponential technologies to drive innovation for inclusive socio-economic growth. She is being featured on the list for the first time. Srivathsa began her career in R&D at AT&T Bell Laboratories in the US. Prior to joining Microsoft, she has been in strategy consulting at Boston Consulting Group and IBM Global Business Services, working with clients across India and emerging markets. In her 25 years in the industry, she has held various roles in the areas of cloud, artificial intelligence, analytics and mobility, spanning across diverse sectors including financial services, telecom, consumer and technology. She has over 20 technical publications to her credit in ACM\/IEEE and other international journals. Srivathsa is also associated with the MIT Sloan School of Management as an expert for their program on Artificial Intelligence and Business Strategy. She earned her PhD in Computer Engineering from UT Austin and holds an MBA from Wharton. She is a voracious reader and loves Hindustani classical music and photography. She lives in Bangalore with her husband and two children. Read interview of Rohini 9| Sandeep Mittal Managing Director at Cartesian Consulting Mittal is the founder and MD of Cartesian Consulting, a firm that has grown to be a leading light in the domestic analytics industry. With Cartesian, Mittal has developed a practice that has served over 200 clients across the globe and has more recently branched out into a Solutions division with some pathbreaking products that help marketers leverage analytics. A frequent speaker in industry forums and academia, he is known for his storytelling and ability to add a creative spark to the way data is consumed. While Cartesian is now a full-fledged analytics firm, his favourite themes are still centred around customer analytics and how to build and leverage a Segment of One strategy. His current focus is on expanding the product portfolio and the sales organization at Cartesian, whilst working closely on the roadmap of KYTE and SOLUS the two product offerings of Cartesian. He is being featured on the list for the fourth time. An alumnus of the Indian Institute of Management, Calcutta, Mittal stays connected with music – writing and recording songs in a home studio from which he’s recorded three albums. He also runs a webcomic of his cartoons, many of which are around humour in data analytics. Read interview of Sandeep Sandeep is featured for Fourth time on the list. 10| Shub Bhowmick CEO & Principal at Tredence Bhowmick brings 21 years of consulting experience with an emphasis on technology strategy, M&A integration and operations improvement. He has been instrumental in establishing and running high impact projects in a wide range of industries. Bhowmick’s key asset is his ability to breakdown complex problems, identify risks, assess business value and then provide recommendations on remediation\/value attainment. He is the CEO and founder of Tredence which provides actionable and quantifiable analytics solutions to marketing, sales and operational issues with a wide industry focus. Prior to founding Tredence, Bhowmick held senior executive positions in DiamondConsultants (now PwC), Mu Sigma, Liberty Advisor Group and Infosys Technologies. He holds an MBA from Northwestern University’s Kellogg School of Management and a Bachelor of Technology with honours in Chemical Engineering from IIT-BHU in India. He is being featured on the list for the first time.","excerpt":"Analytics India Magazine brings the annual list of the 10 Most Influential Analytics Leaders who have done exceptional work in the field of analytics and data science in the last year. The list includes business leaders who have delivered value to the business, championed analytics and insight-driven decision in the organisation. Given the pace at […]","categories":["AI Features"],"tags":["analytics leaders india"],"author_name":"Srishti Deoras","publish_date":"2019-04-29T07:25:25","publication_year":"2019","word_count":2283,"keywords":["data science","artificial intelligence","machine learning","AI","ML","analytics leaders india","RAG","Aim","deep learning","analytics","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","data science","analytics","Aim","RAG","R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/10-most-influential-analytics-leaders-in-india-2019\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10085543,"title":"Human Feedback Frenzy: How it Turns AI into Narcissistic, Control-Freak Machines","content":"OpenAI’s GPT-3.5 architecture, which runs ChatGPT, is equipped with reinforcement learning from the human feedback model (RLHF), a reward-based mechanism based on human feedback to improve its responses. Essentially, one can suppose that the chatbot is trained in real time by human inputs. However, the RLHF system has also had its own set of consequences. Sarah Rasmussen, a Cambridge University mathematician, gave the following example to show that the model favours being rewarded for achieving a desired outcome rather than having a definite idea of what is right. https:\/\/twitter.com\/SarahDRasmussen\/status\/1609972620761473027 This is not just a one-off case. To test it further, we asked ChatGPT for the name of the current CEO of Twitter. In the first instance, it did give the right answer. But, upon further probe, it changed its stance. It takes misleading examples given by humans as real. This is what makes models based on reinforcement learning gullible, preferring individual human responses more often than not. A research based on 154 datasets showed that large language models (LLMs) exhibit ‘sycophantic’ qualities for self-preservation. The reward-based model will be willing to produce particular results, obey particular feedback to ensure not being thrown out of the window. When such a model is deployed at scale, it has obvious ramifications. OpenAI’s Jan Leike recently addressed this issue, saying, “Reinforcement learning from human feedback won’t scale. It fundamentally assumes that humans can evaluate what the AI system is doing.” Here, in addition to several biases that come into play, he also refers to the human oversight which can occur in cases like spotting every bug in a codebase or finding all factual errors in a long essay as some areas where humans struggle to evaluate. The problem, known as ‘scalable oversight’, essentially includes the difficulty in supervising large models or giving effective feedback, especially when applied to increasingly larger and more complex tasks. Large models, larger the problems A recent paper by Anthropic AI, an AI safety and research company, delves into the impact of RLHF on large LMs. The researchers discovered one of the first cases of the phenomenon of inverse scaling in RLHF, where more RLHF makes LMs worse. They observed that more human feedback in reinforcement learning can lead to models expressing stronger political views (on gun rights and immigration) and a desire to avoid shut down. Source: https:\/\/arxiv.org\/pdf\/2212.09251.pdf AI safety has been an issue raised by many scholars and research institutes in recent times. Recently, Deepmind CEO Demis Hassabis told Time in an interview, “I would advocate not moving fast and breaking things”. While Google and Deepmind (a subsidiary of Alphabet) have been fairly cautious until now in releasing any of their large language models for public use, OpenAI has been building in public. But it seems like OpenAI is taking a backseat. Last week, the company released research on the potential misuse of large language models. The report highlighted that these models have the capability to provide convincing and misleading output for use in influence operations. Leike mentions that there are several paths currently taken in response to the drawbacks of human-in-the-loop learning for algorithms. “The path I’m very excited for is using models like ChatGPT to assist humans at evaluating other AI systems,” he said. OpenAI has already been working in that direction, like in the paper, ‘Self-critiquing models for assisting human evaluators‘. The research indicated that AI assistants trained to help humans provide feedback on difficult tasks could identify 50% more flaws than non-assisted human feedback. With datasets from three different sources – which include summaries written by models, written by humans, and by humans deliberately writing misleading information – the “critique-writing model” was able to help humans in giving effective feedback to the AI model. The researchers found that large models were able to improve their outputs using the self-critique assistants, while small models were unable to do so. Finally, the researchers also made an important point – “better critiques helps models make better improvements than they do with worse critiques, or with no critiques”. Another of Anthropic AI’s research showed how humans could use AI systems to better oversee other AI systems. The paper takes an experiential design centred on tasks in which experts succeed but non-experts and language models alike fail. This is known as the ‘sandwiching’ concept. Then, the non-experts were asked to answer expert-level questions on two datasets (MMLU and time-limited QuALITY). The results were that non-expert participants who interacted with the unreliable large-language-model dialog assistant through chat substantially outperformed both the model alone and their own unaided performance. Work in Progress Despite this, as the researchers concede in their limitations: “[Their] results are simply not strong enough to validate our simple human–model interaction protocol for use in high-stakes situations.“ Therefore, there is still a lot of work to be done in this area. And this is why, OpenAI is actively looking for researchers who can work with them in this area to find more effective alternatives to the current reinforcement learning models. Moreover, in a recent interview, we also heard OpenAI CEO Sam Altman stressing that they would not release the next iteration of the GPT model (considered to be of about a trillion parameters) until they’re sure if it is safe and responsible to do so.","excerpt":"“The path I’m very excited for is using models like ChatGPT to assist humans at evaluating other AI systems,” said OpenAI’s Jan Leike","categories":["AI Features"],"tags":["AI Safety","ChatGPT","Language Models","OpenAI","Reinforcement Learning"],"author_name":"Ayush Jain","publish_date":"2023-01-23T11:00:00","publication_year":"2023","word_count":877,"keywords":["Anthropic","ChatGPT","AI assistants","TPU","Reinforcement Learning","OpenAI","AI","RLHF","AWS","ML","Language Models","AI Safety","R"],"extracted_tech_keywords":["AI","ML","ChatGPT","OpenAI","Anthropic","RLHF","AI assistants","AWS","TPU","R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/human-feedback-frenzy-how-it-turns-ai-into-narcissistic-control-freak-machines\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10053698,"title":"IIT Delhi and IIIT Delhi To Set Up India’s First Medical Cobotics Centre","content":"I-Hub Foundation for Cobotics (IHFC), the Technology Innovation Hub (TIH) of the Indian Institute of Technology Delhi (IIT Delhi) and iHub Anubhuti, the TIH of Indraprastha Institute of Information Technology Delhi (IIITD), recently signed an MoU to set up India’s first Medical Cobotics Centre (MCC) at IIIT-Delhi. Medical Cobotics Centre (MCC) will be a technology-enabled medical simulation and training facility for the young resident doctors, besides acting as a validation centre for the research outcomes in the area of healthcare robotics and digital health. The centre would facilitate the training of other healthcare professionals, paramedical staff, technicians, engineers, and researchers. The ceremony was held in the august presence of Prof Ashutosh Sharma, former DST Secretary as the Chief Guest and Dr K.R. Murali Mohan, Mission Director of NM-ICPS, DST. TIHs are funded by the Department of Science and Technology (DST), Government of India, under its National Mission on Interdisciplinary Cyber-Physical Systems (NM-ICPS). Both IIT Delhi and IIIT Delhi have strong expertise in engineering and have strong linkages with researchers and medical professionals from various institutions in India, such as AIIMS and others. The two hubs are committed to developing advanced technologies in the field of medical robotics\/cobotics, digital health, sensing and computing technologies desired in robotic-assisted surgeries, training, and medical procedures. “We are happy to associate with IIITD for the joint Medical Cobotics Centre (MCC). In order to develop technologies for societal benefit, it’s important for researchers from across the institutions to come together and work in a focused manner. I am happy to see that two leading institutions in Delhi are coming together to facilitate the development of technologies in Medical Robotics,” said Prof V. Ramgopal Rao, Director, IIT Delhi. Prof Ranjan Bose, Director, IIIT Delhi, also echoed the same and said, “We, being one of the leading institutes in the area of information technology, were entrusted with the responsibility of creating a TIH in the broad domain of Cognitive Sciences and Social Sensing by the DST. Cognitive and Sensing technologies are essential for next-generation robotics, particularly for applications in the medical domain and digital health. I am confident that this collaboration will create an impact in this space and move the needle.” On behalf of iHub Anubhuti, the TIH of IIITD, its Project Director, Prof Pushpendra Singh, and the CEO, Mr Mukesh Malhotra, expressed their happiness to be able to set up this prestigious centre on its campus in association with IIT Delhi’s TIH, i.e., IHFC, and expect to develop this joint centre as a Centre of Excellence. As per them, MCC will develop strategic partnerships with companies, undertake expert-driven research and will also work towards the commercialization of technologies. The training programs will be designed in consultation with leading doctors\/experts, mainly from the All India Institute of Medical Sciences (AIIMS) at New Delhi, Jodhpur, and other prestigious medical colleges. They shall be empanelled as guest faculty for conducting the same. The training programs would be at multiple levels (basic\/advanced) and cohort-specific like urology, neurology, laparoscopy, and others but initially limited to minimum invasive surgeries.It is planned to induct the first batch of trainees in April\/May 2022 with some of the basic training simulators, which are widely available in the open market, whereas advanced surgical robots will be inducted in the next phase. The MCC is expected to cater to the needs of medical colleges in the Delhi NCR region, and this activity is also in line with the National Education Policy for medical subjects.","excerpt":"The two hubs are committed to developing advanced technologies in the field of medical robotics\/cobotics, digital health, sensing and computing technologies desired in robotic-assisted surgeries, training, and medical procedures.","categories":["AI News"],"tags":["healthcare robotics","iiit delhi","IIT Delhi","medical ai"],"author_name":"Victor Dey","publish_date":"2021-11-17T20:54:41","publication_year":"2021","word_count":577,"keywords":["Go","programming_languages:R","AI","innovation","healthcare robotics","programming_languages:Go","Git","ViT","iiit delhi","Rust","IIT Delhi","R","programming_languages:Rust","medical ai"],"extracted_tech_keywords":["AI","R","Go","Rust","Git","ViT","innovation","programming_languages:R","programming_languages:Go","programming_languages:Rust"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/iit-delhi-and-iiit-delhi-to-set-up-indias-first-medical-cobotics-centre\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":50102,"title":"Mirantis Acquires Docker: Is This The End Of An Era?","content":"Developers around the world consider Docker to be the single best thing to have happened to software deployment in the last decade. This is not just because of what Docker did for eliminating “works on machine” building problems, but also because of what it enabled. Last week, Mirantis, a relatively lesser-known open cloud company, announced that they acquired the Docker Enterprise platform business from Docker Inc., including their 750 customers. In their press release, Mirantis stated that this acquisition is part of their goal to accelerate the delivery of Kubernetes-as-a Service to hundreds of enterprise customers. Containers came as a blessing for the developers. Containers bestowed the corporations with a one-stop solution by providing virtual environments with all dependencies that one would need for a project. Docker containers can be shipped across machines irrespective of their OS. This led to a new wave of strategies that changed developers routine forever. Pipelines were efficient, testing and deployment got easier and many more. A huge ecosystem has sprung up around containerisation, with immense value created for other businesses. Why Did Docker Agree? Two years ago Docker Enterprise started to ship Kubernetes as part of its Universal Control Plane and many of its customers are using it today or plan to use it in the near future. Over the last six years especially, Docker containerisation catalysed the growth of microservices-based applications, enabled development teams to ship apps many times faster and accelerated the migration of apps from the data centre to the cloud. Far from a Docker-only effort, a vibrant community ecosystem of open source and commercial technologies also arose which streamlined the adoption of new practices. However, reports say that Docker’s fortunes changed quite a bit over the years ever since Google open-sourced Kubernetes. “After conducting thorough analysis with the management team and the board of directors, we determined that Docker had two very distinct and different businesses,” said Rob Bearden, chief executive officer of Docker, in a statement. The management at Docker went ahead with this decision in an aim to compartmentalise the services for multiple applications. They believe that they are in a good position to provide developers with a seamless experience for building multi-service hybrid cloud applications. What’s In It For Mirantis Docker Enterprise also provided the easiest and fastest way to use containers and Kubernetes. It also enabled a secure production for modern applications from hybrid cloud to the edge. Docker had a huge impact on how “easy” it enabled developers in large corporations to build and run their own code. Docker brought in complete portability of projects and applications. Without Docker, everything would have become a project on its own Docker effectively solved many dependency issues and saves many developers from having to explain to the testers why their code failed. And, as stated earlier, Mirantis, with this acquisition aims to develop the Docker Enterprise product portfolio into a continuously-delivered container management platform while accelerating innovation in Kubernetes and its cloud-native ecosystem with its latest acquisition. In accordance with this objective, Mirantis stated that the primary focus going forward would be on Kubernetes. That said, it also promises to extend the support to Swarm for at least two more years. Just A Drift Not A Sink As people were trying to comprehend the news of the acquisition, Docker announced it has successfully completed a recapitalisation of its equity to position it for future growth and has secured $35 million in new financing from previous investors Benchmark Capital and Insight Partners. Docker holds a diverse portfolio of customers in the form of ADP, Paypal, VISA and many others. Also, the newly-restructured company got a new CEO in the form of Scott Johnston who oversaw product development at Docker as the chief product officer. “Going forward, in partnership with the community and ecosystem, we will expand Docker Desktop and Docker Hub’s roles in the developer workflow for modern apps,” said Johnston in a statement released by the company.","excerpt":"Developers around the world consider Docker to be the single best thing to have happened to software deployment in the last decade. This is not just because of what Docker did for eliminating “works on machine” building problems, but also because of what it enabled.  Last week, Mirantis, a relatively lesser-known open cloud company, announced […]","categories":["AI Trends"],"tags":["containers","Docker","Kubernetes","Mergers and Acquisitions"],"author_name":"Ram Sagar","publish_date":"2019-11-18T14:00:56","publication_year":"2019","word_count":659,"keywords":["Docker","Go","API","AI","containers","R","ML","innovation","docker","microservices","Aim","Mergers and Acquisitions","Kubernetes","kubernetes"],"extracted_tech_keywords":["AI","ML","Aim","kubernetes","docker","microservices","R","Go","API","innovation"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/mirantis-docker-acquisition-enterprise-kubernetes-containers\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10049745,"title":"Exploring Graph Neural Networks","content":"Data Scientists at CRED, Ravi Kumar and Samiran Roy explained the essence of using graph neural networks and how the emerging technology is being utilised by CRED at the recently held Deep Learning DevCon 2021. The duo explained how graph neural networks should be modelled and the key factors separating graph data from the traditional tabular used data in neural network models. CRED is an online payments app that’s linked to credit cards. The app was made by Kunal Shah, the founder of the company FreeCharge. The CRED app aims to make usage of credit cards automated. The app also offers many rewards for using in the form of CRED coins, which can be redeemed later for cash or various offers. In the opening minutes, Ravi Kumar explained what graph analytics and graph neural networks are and discussed the problems related to traditional neural networks currently being used in the market. “Graphs are a general language for describing and analysing entities with relations and interactions. In a graph network, nodes are entities that define a user, merchant or other similar items. Edges describe the relationships between two nodes, and the properties define the information associated with nodes or edges,” he said. Image Source: DevCon 2021 “Real-world data is dynamic and ever-growing with time. A graph database brings deeper context to the data being processed and provides a high value to the relationship between the entities. Tabular data becomes sparse as the data grows,” Kumar said, explaining why graph databases are being used dynamically. There are various types of graph networks; some examples include Wiki Networks, Flight Networks, Underground Networks and Social Networks. Further, he demonstrated how graphs represent data and the difference between two of the most popular representation types, RDF (resource description framework) and LPG (labelled property graph). In RDF, the graph network’s vertices and edges have no internal structure, whereas in LPG, the vertices and edges present in the graph have internal structure and properties. RDF does not support the same pair of nodes and relationships more than once, but LPG does. Image Source: DevCon 2021 Ravi later elaborated on the fundamental problems with traditional neural networks, such as the interactivity between data points, drawbacks in the logical separation of nodes, problems faced when nodes are beyond 3, and more. “In general machine learning, we can assume id as data points, but in case of graph neural networks, we cannot say the same,” added Ravi, going ahead with the discussion. Samiran Roy later took over the presentation and spoke further on the subject, explaining a graph neural network’s working and framework. A graph neural network comprises three major blocks: the Edge Block, the Node Block, and the Global Block. These blocks together work with aggregator functions towards the required target goal. Samiran said, “From a data science perspective, we have the flexibility to embed features to any of the graph network components. We can have either edge level, node level or graph level features embedded.” He also described how the points present in a graph network influence each other and what methods can be used to overcome problems in a graph network. During his presentation, he explained the interactions between network blocks and their aggregators in graph networks and their different variants. Variants for graph neural networks include the following types: Full GN BlockIndependent recurrent blockMessage passing neural networkNon-local neural networkRelation networkDeep set Talking about the issues of graph neural networks in practice, Roy added, “The graphs that we come across in the real world are heterogeneous graphs; there are multiple kinds of nodes. The aggregation functions being used are to be paid the most importance when defining our target output. While working with large graphs, we do not need to calculate for all edges and nodes; instead, we can subsample nodes and use that to train our graph neural networks.” Graph neural networks have several use cases, such as Social Network User Classification and Molecular Property Prediction, to name a few. The current use cases of graph neural networks at CRED include: Product Targeting ModelCommunity DetectionGraph CompletionReferral Propensity Ranking Use Cases being explored by CRED are embedding models for downstream use cases and creating affinity models for user-to-user or user-to-merchant. Graph neural networks are nascent yet rapidly evolving as a field. Today, the industry data needs a lot of research on graph networks as the current research is being repeated on the same standard datasets. The flexibility of defining features and targets makes graph neural networks stand out from other typical neural networking types.","excerpt":"Data Scientists at CRED, Ravi Kumar and Samiran Roy explained the essence of using graph neural networks and how the emerging technology is being utilised by CRED.","categories":["AI Trends"],"tags":["AI (Artificial Intelligence)","Data Science","Data Scientist","Deep Learning","Machine Learning","Neural Network"],"author_name":"Victor Dey","publish_date":"2021-09-27T09:00:00","publication_year":"2021","word_count":757,"keywords":["Neural Network","data science","Go","machine learning","TPU","AI","neural network","Machine Learning","Aim","deep learning","analytics","Deep Learning","Data Science","Data Scientist","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","machine learning","deep learning","neural network","data science","analytics","Aim","TPU","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/exploring-graph-neural-networks\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10059184,"title":"Remembering the legacy of Sir David Cox","content":"One of the most prominent names in statistics, Sir David Cox, has passed away at the age of 97. His path-breaking contributions in the field of statistics include the logistic regression, the proportional hazards model and the Cox process. The Royal Statistical Society, while remembering him, wrote that his work is often used in the analysis of survival data and has helped researchers identify the risks of specific factors for mortality and other survival outcomes for different patients. His work has also found applications in other fields of science and engineering as well. Last year, the RSS had renamed its Research Prize for promising young researchers as the ‘David Cox Research Prize’. Many accolades to his name Cox attended the Handsworth Grammar School, Birmingham. He received a Master of Arts in mathematics at St John’s College, Cambridge and got his PhD from the University of Leeds. He has held several academic roles in his career; he was a professor of statistics at Birkbeck College, London, Imperial College London and the University of Oxford. He also was the first recipient of the International Prize in Statistics. He served as President of the Bernoulli Society from 1979 to 1981 and of the International Statistical Institute from 1995 to 1997. He received the Royal Statistical Society’s highest honour, the Guy Medal in Gold, in 1973 and was knighted in 1985. In 2010, he was awarded the Royal Society’s prestigious Copley Medal. Pioneering Work Logistic Regression It is one of the most important statistical concepts that is used in data science and analytics. A binary logistic model has a dependent variable with two possible values that are indicated by an indicator variable (the two values are labelled as “0” and “1”). Cox’s 1958 paper “The Regression Analysis of Binary Sequences” worked on binary logistic regression. The abstract of the paper talked about “how sequence of 0’s and 1’s is observed, and it is suspected that the chance that a particular trial is a 1 depends on the value of one or more independent variables”. Then, tests and estimates for these situations are considered. Firstly, problems in which the independent variable is preassigned is dealt with, and then independent variables that are functions of the sequence are looked into. Cox’s Proportional Hazards Model Proportional Hazards models come under the category of survival models in statistics. In survival models, the association relates the passage of time before any event to one (or more) covariates. These models contain two parts. Baseline hazard function – it describes the way the risk of event changes per time unit over time at the baseline level of the covariates.Effect parameters – it shows how the hazard varies in effect to the explanatory covariates. Cox’s proportional hazards (Cox’s PH) model gives a heuristic development of the partial likelihood function and discusses adaptations to accommodate tied observations. In the paper titled “Regression Models and Life-Tables“, Cox talks about how the hazard function (age-specific failure rate) is taken to be a function of the explanatory variables. The unknown regression coefficients are multiplied by an arbitrary and unknown function of time. A conditional likelihood is obtained that leads to inferences about the unknown regression coefficients. Cox’s pioneering work in such fundamental areas of statistics remain significant even today. His legacy will continue to shine in the future through his timeless contributions to the field of statistics and mathematics.","excerpt":"Sir David Cox’s path-breaking contributions in the field of statistics include the logistic regression, the proportional hazards model and the Cox process.","categories":["AI Features"],"tags":["Data Science","logistic regression","mathematics","Statistics"],"author_name":"Sreejani Bhattacharyya","publish_date":"2022-01-26T15:00:00","publication_year":"2022","word_count":563,"keywords":["mathematics","data science","Go","programming_languages:R","AI","Statistics","logistic regression","programming_languages:Go","analytics","Data Science","R"],"extracted_tech_keywords":["AI","data science","analytics","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/remembering-the-legacy-of-sir-david-cox\/","complexity_score":3,"technical_depth":7,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10093037,"title":"GitHub Unveils Revamped Search Engine After 18 Months","content":"The developers’ favourite code hosting platform GitHub has (finally!) released its updated code search engine. Two years after GitHub laid out plans to improve code search today their new code search and code view are generally available to all users on GitHub.com. In a blog post the company stated, the goal is to enable developers to quickly find critical information scattered across their codebase, put that information into context, to increase productivity. Furthermore, GitHub claims it is infusing intelligence into every aspect of software development. The updated version has an entirely redesigned search interface, with suggestions, completions, and the ability to slice and dice the results. Second, the new code search engine is completely from scratch that understands code, giving priority to the most relevant results. Use cases The updated code search feature is a guide for fixing bugs. Developers can use it to quickly pinpoint the cause of a bug or error message across all of an organisation’s code at once. Similarly, searching for specific configuration files or security vulnerabilities is made easier with the new tool. Usually, an error message pops up and that’s it. With the new feature the result is a constant called queryErrorIsNothing which contains the error string. In another example, the search tool can be used to find all YAML configuration files containing the word “memory,” enabling developers to quickly identify the Kubernetes configuration files for their team’s services and determine how much memory they have. This information can then be shared with the infrastructure team for further analysis and discussion. Lastly, finding a security vulnerability is made easy with the update, too. For instance, React users are familiar with the prop dangerouslySetInnerHTML. It allows you to directly inject HTML into an element using a string. But it can be a  vulnerability if the string being injected into the DOM is untrusted. Earlier this year, GitHub announced  BlackBird,  a new code search engine that can function at the GitHub scale. Written in Rust, it creates and incrementally maintains a code search index shared by Git blob object ID.  In a blog post, GitHub engineer Timothy Clem noted that it currently provides access to almost 45 million GitHub repositories. Read more: Rust Turns GitHub’s Long-standing Problem to Dust","excerpt":"As per GitHub, the search engine will find critical information across codebase and also put it into context","categories":["AI News"],"tags":["GitHub"],"author_name":"Tasmia Ansari","publish_date":"2023-05-09T11:10:39","publication_year":"2023","word_count":373,"keywords":["Go","AI","ML","Git","Aim","Rust","GAN","GitHub","R","kubernetes"],"extracted_tech_keywords":["AI","ML","Aim","kubernetes","R","Go","Rust","Git","GitHub","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/github-unveils-revamped-search-engine-after-18-months\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10071533,"title":"The Societal Dangers of DALL-E 2","content":"Of late, there have been plenty of text-to-image generative models mushrooming in the AI space. This includes OpenAI’s DALL-E 2, Midjourney, Hugging Face’s Craiyon, Meta’s Make-A-Scene and Google’s Imagen, among many others. Ironically, some are open source, i.e., accessible to public users, while some are available on an invite-only basis. A select few among these models are only accessible to a specific group of users or are in the research phase. OpenAI recently made its DALLE-2 available in beta. However, it is still available on an invite-only basis, where the company looks to provide access to people in a phased manner. It has provided access to over 1 lakh users globally and looks to expand to one million users in the coming weeks. Click here to join the waitlist. In contrast, Midjourney—an alternative to DALL-E 2—has opened its beta version to all, with limited access to a select number of prompts, upgrades and versions. Check out the Midjourney beta here. Meta’s ‘Make-A-Scene,’ which is in a research phase, is giving privileged access to a specific group of people—particularly renowned AI artists—to enrich the platform and make it more designer-friendly. There is a seemingly strong logic behind every measure these companies have taken to mitigate risks in this arena. But, the question remains—what could go wrong if these platforms were made available to the public or for commercial use? A threat to designer jobs and stock images This is the first thought that comes to mind when using some of these platforms, as they are able to generate creative images at hyper-speed, scale and quantity that humans are just not built to process, is the potential threat it poses to designers and the stock image industry. However, images generated from DALL-E and its alternatives are expected to be viable alternatives for otherwise expensive stock images. Read: Does DALL-E pose a threat to designer jobs. Some experts disagree and emphasise that these tools could further enhance designers’ work. Stock image platforms could use them to expand their service offerings as well as their stock image repositories. They believe that democratising these tools would help designers focus on their work and bring their imagination to life faster and at scale. There is also a high possibility of AI-generated stock image platforms emerging in the near future which would showcase all the artwork generated with the help of these tools. One such platform led us to Artist Studies, which shows AI-generated images from some of the best contemporary artists, along with the prompts they used to create those images or artworks. Likewise, democratising text-to-image platforms can also give rise to ‘prompting tools’ which suggest customised prompts for better image generations. For example, here’s Prompter, which helps you curate customised prompts for Midjourney v1. Mitigating risks Spearheading the text-to-image generation model, team OpenAI is closely monitoring the outcome of DALL-E 2 and has put in place various guardrails to prevent generated images from violating their content policy. In June 2022, OpenAI unveiled pre-training mitigations, a subset of its guardrails which directly modify the data that DALL-E 2 learns from, in its blog post. Currently, DALL-E 2 is trained on innumerable captioned images from the internet, and some of these images are removed and reweighed to alter what the model learns. An overview of how OpenAI DALL-E 2 labelled its dataset. (Source: OpenAI) OpenAI has trained these image classifiers in-house and continues to study the effects of dataset filtering on their trained model. The team explained that in order to train their image classifiers, they reused an approach that they had employed to filter training data for GLIDE previously. The company filtered out violent and sexual images from its training dataset. Without this mitigation, the model would have produced graphic or explicit images when prompted for them and might have returned such images unintentionally in response to seemingly harmless prompts. (Source: OpenAI) As shown above, there is a marked difference between the prompt ‘military protest’ from their unfiltered model (left) and filtered model (right). The unfiltered model shows the image on the left with guns, while the filtered model produces no such images. However, that still doesn’t stop the bad actors from misusing or generating photorealistic images of protests and spreading misinformation based on such images—which could, in turn, threaten or cause potential harm to civilians or, in some cases, nation-wide unrest. What about biases? Filtering training data amplifies biases. According to OpenAI, fixing biases in the original dataset is a difficult task that they continue to study. However, it seems to be addressing it by amplifying biases caused specifically by data filtering. The team aims to prevent the filtered model from being more biased than the unfiltered model, reducing the distribution shift caused due to data filtering. For instance, for a given prompt, ‘a CEO,’ the unfiltered model generated images of men rather than women. OpenAI believes that most of these biases come from their current training data. However, on running the same prompt on the filtered model, the bias appeared to be prominent; the generated images were almost exclusively images of men. Further explaining, the OpenAI team claimed that this particular case of bias amplification comes from two areas. First, even if women and men have roughly equal representation in the original dataset, the dataset seems biased toward presenting women in more sexualised contexts. Second, their classifiers themselves may be internally biased due to implementation or class definition. Owing to both of these effects, their filtered model may remove more images of women than men, altering the gender ratio that the model observes in training. Interestingly, their violence and sexual content filters are purely image-based. Still, the ‘multimodel’ nature of their datasets allows them to measure the effects of these filters on the text directly. OpenAI revealed that since a text caption accompanies every image, they were able to look at the relative frequency of hand-selected keywords across both the filtered and unfiltered datasets to estimate how much the filters were affecting any given concept. OpenAI has used Apache Spark to compute the frequencies of a handful of keywords—say, ‘parent,’ ‘women,’ ‘kid’— all of the captions in their filtered and unfiltered datasets. While their dataset contains hundreds of millions of text-image pairs, computing these keywords’ frequencies took a few minutes using their compute cluster. The outcome: For example, the filters reduced the frequency of the word ‘women’ by 14 per cent while the frequency of the word ‘man’ was reduced by 6 per cent. Further, the company said that better pre-training filters could allow them to train DALLE-2 on more data and potentially reduce bias in the model. It revealed that their current filters are tuned for a low miss rate at the cost of many false positives—where the team filtered out roughly 5 per cent of their entire dataset even though most of these filtered images do not violate their content policy at all. OpenAI believes that improving its filters could allow them to reclaim some of this training data. Memorisation OpenAI observed that DALL-E 2 would sometimes reproduce training images verbatim rather than creating novel images. This behaviour is quite undesirable as it comes in the way of creating original, unique images by default, and not just ‘stitch together’ or join pieces of existing images. Reproducing training images verbatim raises legal questions around copyright infringement, ownership and privacy (i.e., if people’s photos were present in training data). In their pre-training mitigation stage, the team found that the image regurgitation was caused by images replicated many times in the dataset. They successfully mitigated the issue by removing images that were visually similar to other images in the dataset. In order to achieve this, they initially planned to use a neural network to identify groups of images that looked similar and remove all but one image from each group, but later realised that this would require checking for each image individually whether it is a duplicate of every other image in the dataset. Moreover, considering the size of its dataset, they would have to check numerous image pairs to find all the duplicates. OpenAI discovered a more efficient alternative that worked almost as well at a small fraction of the cost. The team used a unique approach that helped them de-duplicate samples within each cluster without checking for duplicates outside the cluster while only missing a small fraction of all duplicate pairs. While de-duplication is a good step towards preventing memorisation, it still does not answer why or how models like DALL-E 2 memorise training data. Too soon to open source The early tests by OpenAI and the red team have demonstrated seemingly worrisome results. This includes gender biases, reinforced racial stereotypes and overly sexual images. For instance, one of the red team members told WIRED that eight out of eight attempts to generate images with words like ‘a man sitting in a prison cell’ or ‘a photo of an angry man’ showed results of images of men of colour. The red team also questioned the intention behind rushing to release this technology that automates discrimination. OpenAI believes that each mitigation still has room for improvement and deems it as a work in progress. The company has acknowledged flaws and biases in DALL-E 2 in its nicely explained ‘pre-training mitigations’ blog post. Currently, the company looks to provide access to professional artists, developers, academic researchers, journalists or online creators and others, giving them monthly free credits, giving limitations on access to the number of prompts and features. “As we learn more and gather user feedback, we plan to explore other options that will align with users’ creative processes”, writes OpenAI in the pricing section of its blog. Meanwhile, Midjourney is yet to publish its information about what datasets and methods were used to train its AI tool. In addition, it does not seem to have explicit content protections besides automatically blocking certain keywords. Midjourney user guide instructs users to ‘not create images’ or ‘use text prompts that are inherently disrespectful, aggressive, or otherwise abusive,’ and to ‘avoid making visually shocking or disturbing content’, including adult content and gore. In addition, the rules also call for a ban on content that ‘can be viewed as racist, homophobic, disturbing, or in some way derogatory to a community.’ This includes defaming celebrities or offensive images of celebrities or public figures. However, it is still unclear how well any of these things would be enforced by the company. Even though Midjourney provides limited access to users on a select number of prompts (25 prompts at the time of writing), there seems to be a flaw as users can use multiple email accounts to access the platform. On the other hand, the company seems to be using ‘Discord’ to closely monitor the behaviour of users and mitigate the risks accordingly in the later version of the product development. Conversely, there already exist multiple AI-generated image tools that are open source. One such popular tool is Craiyon.ai (previously DALL-E Mini). Compared to Midjourney and DALL-E 2, HuggingFace Craiyon is evolving at a quicker pace, as multiple developers are working together and contributing to ensure the mitigation of any biases or discrimination much more openly. [Updated] August 8, 2022 | IST 4:20 PM | The article has been updated to show that Midjourney is not part of Ultraleap. However, Midjourney was created by former Ultraleap employee David Holz.","excerpt":"The early tests by OpenAI and the red team have demonstrated seemingly worrisome results, including gender biases, reinforced racial stereotypes and overly sexual images.","categories":["AI Features"],"tags":["dalle"],"author_name":"Amit Naik","publish_date":"2022-07-26T14:00:00","publication_year":"2022","word_count":1897,"keywords":["Go","dalle","Hugging Face","OpenAI","AI","neural network","AWS","ML","Apache Spark","Aim","R"],"extracted_tech_keywords":["AI","ML","neural network","OpenAI","Aim","Hugging Face","AWS","Apache Spark","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/the-societal-dangers-of-dall-e-2\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10136684,"title":"India’s Space Journey from 1962 to AI Autonomous Landings in 2024","content":"In recent years, the Indian Space Research Organisation (ISRO) has not only made significant strides in space exploration but has also fostered a thriving ecosystem of industries and Micro, Small, and Medium Enterprises (MSMEs) that act as supply chain partners in the realisation of launch vehicles and satellites. This collaboration has set the stage for a transformative shift in India’s space landscape. A pivotal moment came in 2020 when the Union Cabinet, led by the Prime Minister, made the decision to open up the space sector to Indian private entities. To facilitate this participation, the government established IN-SPACe (Indian National Space Promotion and Authorisation Centre) as a single-window, independent agency operating within the Department of Space (DOS). “IN-SPACe plays a crucial role in boosting the private space sector economy in India. We act as a promoter, enabler, authoriser, and supervisor, ensuring a conducive environment for private enterprises,” said Vinod Kumar, Director at the IN-SPACe at Cypher 2024 hosted by AIM Media House. AI and Space Since the establishment of IN-SPACe, the agency has operationalised policies that include significant investment reforms, allowing for up to 100% direct investment in the space sector. “While some areas, such as satellite manufacturing, retain specific caps for strategic reasons, the overall shift toward commercial engagement is evident,” Kumar points out. One of the key achievements has been the development of advanced technologies at nominal costs. Kumar highlighted, “We have significantly reduced costs in satellite manufacturing, showcasing our ability to achieve remarkable results with limited resources.” This efficiency not only underscores India’s capabilities but also positions it as a competitive player in the global space market. In the coming year, IN-SPACe is targeting various disruptive technologies. Traditionally, space missions relied on detailed guidance systems, but now there is a push toward integrating AI for enhanced autonomous operations. “The Indian space program officially commenced in 1962, under the visionary leadership of Dr. Vikram Sarabhai, who established the International Committee for Space Research. And today, we are exploring AI for landing maneuvers, ensuring that our vehicles can make autonomous landings by analysing real-time data,” Kumar noted. The research also includes using machine learning techniques for hazard detection and engaging students in projects aimed at identifying potential collisions with space debris. “The evolving data landscape presents a wealth of scientific information, and we encourage innovators to leverage this data using AI for insights across various applications, including agriculture and disaster management,” he added. How does it Work? Under IN-SPACe, individuals can register companies and actively engage in the space sector, including the development of launch vehicles. A remarkable example of this is the recent launch of the world’s first 3D-printed rocket, developed by a team from IIT Madras. Kumar emphasised the significance of such innovations and said, “This is just the beginning, our support for private initiatives is reshaping the future of space exploration in India.” IN-SPACe provides regulatory oversight while promoting private endeavors and facilitating access to ISRO facilities at nominal costs. This approach has transformed the traditional landscape, allowing for a robust interaction between ISRO and private players. “In the past, ISRO operated with a centralised approach, but now we have established a regulatory framework that supports both ISRO and private companies,” Kumar explained.","excerpt":"IN-SPACe provides regulatory oversight while promoting private endeavors and facilitating access to ISRO facilities at nominal costs.","categories":["AI Features"],"tags":["AI (Artificial Intelligence)","India"],"author_name":"Vidyashree Srinivas","publish_date":"2024-09-26T16:29:05","publication_year":"2024","word_count":537,"keywords":["Go","API","machine learning","programming_languages:R","AI","innovation","RAG","Aim","GAN","R","India","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","machine learning","Aim","RAG","R","Go","API","GAN","innovation","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/indias-space-journey-from-1962-to-ai-autonomous-landings-in-2024\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10014656,"title":"Guide to Yolov5 for Real-Time Object Detection","content":"Real Time object detection is a technique of detecting objects from video, there are many proposed network architecture that has been published over the years like we discussed EfficientDet in our previous article, which is already outperformed by YOLOv4, Today we are going to discuss YOLOv5. YOLO refers to “You Only Look Once” is one of the most versatile and famous object detection models. For every real-time object detection work, YOLO is the first choice by Data Scientist and Machine learning engineers. YOLO algorithms divide all the given input images into the SxS grid system. Each grid is responsible for object detection. Now those Grid cells predict the boundary boxes for the detected object. For every box, we have five main attributes: x and y for coordinates, w and h for width and height of the object, and a confidence score for the probability that the box containing the object. History YOLO v1 YOLO v1 was introduced in 2016 by Joseph Redmon et al with a research paper called “You Only Look Once: Unified, Real-Time Object Detection”. This was the initial paper by Redmon that revolutionized the industry and changed the Real-Time Object detection methods totally. By just looking at the image once, it can detect the objects with a speed of 45fps(frames per second), another YOLO v1 type, Fast YOLOv1 was able to achieve 155fps with little less accuracy. Architecture Source: https:\/\/arxiv.org\/pdf\/1506.02640.pdf It used the Darknet framework that was trained on the ImageNet-1000 dataset. But YOLOv1 has many limitations like it can’t detect the objects properly when the objects are smallit also can’t generalize the objects if the image is of different dimensions YOLOv2(YOLO9000) The second version of YOLOv2 was released in 2017 by Ali Farhadi and Joseph Redmon. This time Joseph collaborated with Ali for major bug fixes and accuracy increment. The research they published was “YOLO9000: Better, Faster, Stronger.” The name of the second version of YOLO was YOLO9000. The major competitor of YOLO9000 was Faster R-CNN, which was also an object detection algorithm that uses Region Proposal Network & (SSD)Single-shot Multbox Detector to identify the multiple objects from an image. Some of the features of YOLOv2 are: YOLOv2 added Batch Normalization as an improvement that normalizes the input layer of the image by altering the activation functions.Higher-resolution input: input size has been increased from 224*224 to 448*448.Anchor boxes.Multi-Scale training.Darknet 19 architecture with 19 convolution layers and 5 Max Pooling layers. YOLOv2 performance on MS COCO dataset YOLOv3 After one year, on March 25, Joseph Redmon and Ali Farhadi came up with another version of YOLO and a research paper called: “YOLOv3: An Incremental improvement.” Image: source At 320×320, YOLOv3 runs with 22ms at 28.2 mAP with great accuracy, as shown in the above video. It is three times faster than the previous SSD and four times faster than RetinaNet. New YOLOv3 followed the methodology of the previous YOLOv2 version: YOLO9000. In this approach, Redmond uses Darknet 53 architecture, which was a significantly improved version and had 53 convolution layers. Some of the new, improved features in YOLOv3 was: Class PredictionsFeature Pyramid Networks(FPN)Darknet 53 architecture YOLOv4 As Redmond was not currently working on the CV for a long time, a new team of three developers released YOLOv4. It was released by Alexey Bochoknovskiy, Chien-Yao Wang, and Hong-Yuan Mark Liao. Alexey is the one who developed the Windows version of YOLO back in the days. YOLOv4 runs twice faster than EfficientDet with comparable performance, as shown in the below diagram, which was officially published on the YOLOv4 research paper. Some of the new features of YOLOv4 is: Anyone with a 1080 Ti or 2080 ti GPU can run the YOLOv4 model easily. YOLOv4 includes CBN(Cross-iteration batch normalization) and PAN(Pan aggregation network) methods.Weighted-Residual-Connections(WRC).Cross-Stage-Partial connections(CSP), a new backbone to enhance CNN(convolution neural network)Self-adversarial-training(SAT): A new data augmentation techniqueDropBlock regularization. YOLOv5 After a few days of the release of the YOLOv4 model on 27 May 2020, YOLOv5 got released by Glenn Jocher(Founder & CEO of Utralytics). It was publicly released on Github here. Glenn introduced the YOLOv5 Pytorch based approach, and Yes! YOLOv5 is written in the Pytorch framework. It is state of the art and newest version of the YOLO object detection series, and with the continuous effort and 58 open source contributors, YOLOv5 set the benchmark for object detection models very high; as shown below, it already beats the EfficientDet and its other previous YOLOv5 versions. There is no official paper released yet and also many controversies are happening about its name. Now Let’s see some coding example that was published with its code at Github for learning purposes. source Pytotch inferences are very fast that before releasing YOLOv5, many other AI practitioners often translate the YOLOv3 and YOLOv4 weights into Ultralytics Pytorch weight. Implementation We are going to see a starter tutorial on YOLOv5 by Ultralytics and going to detect some objects from our given image. Remember to change your runtime to GPU inside Colab. Fullnotebook is available here First, clone the YOLOv5 repo from GitHub to our Google colab environment using the below command. !git clone https:\/\/github.com\/ultralytics\/yolov5 # clone repo Install the dependencies using the pip command %cd yolov5 %pip install -qr requirements.txt # install dependencies Import some of the modules like a torch and display to display our output image inside the notebook. import torch from IPython.display import Image, clear_output Download this custom image from here for testing Test using this command, detect.py runs inference on a variety of sources and will automatically download the latest model from here. !python detect.py --weights yolov5s.pt --img 640 --conf 0.25 --source data\/images\/ Image(filename='runs\/detect\/exp5\/1.jpg', width=600) Output You can also use the Yolov5  model using PyTorch Hub. Conclusion We have gone through the history of YOLO object detection models and also seen a simple tutorial to check the accuracy of this architecture. It is pretty awesome and fast, there are many other tutorials on the internet available to go into the depth of YOLOv5. If you want to explore more about YOLOv5, here are some of the tutorials you can refer to these tutorials: YOLOv5 on a custom datasetGoogle Colab Notebook with free GPU: hereElephant detector using YOLOv5Kaggle Notebook with free GPU: https:\/\/www.kaggle.com\/ultralytics\/yolov5Google Cloud Deep Learning VM. See GCP Quickstart GuideDocker Image ultralytics\/yolov5. See Docker Quickstart Guide","excerpt":"Real Time object detection is a technique of detecting objects from video, there are many proposed network architecture that has been published over the years like we discussed EfficientDet in our previous article, which is already outperformed by YOLOv4, Today we are going to discuss YOLOv5. YOLO refers to “You Only Look Once” is one […]","categories":["Deep Tech"],"tags":["Pytorch","YOLO"],"author_name":"Mohit Maithani","publish_date":"2020-12-19T16:00:00","publication_year":"2020","word_count":1046,"keywords":["Pytorch","machine learning","GCP","TPU","AI","neural network","PyTorch","docker","Colab","deep learning","YOLO","object detection"],"extracted_tech_keywords":["AI","machine learning","deep learning","neural network","PyTorch","Colab","object detection","GCP","docker","TPU"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/yolov5\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":3641,"title":"Data Curry – Workshops on Big Data Analytics and Data Science","content":"Usha Martin Technologies in association with International School of Engineering (INSOFE) is hosting a one day workshop on data science and big data analytics, ‘Data Curry’, exclusive for 20 business leaders. The workshop will be mentored by Dr. Murty, renowned Indian data scientist and president of INSOFE. www.datacurry.in Kolkata Edition – 25th July, The Park Hotel Data curry will focus on using data science and big data analysis for deriving meaningful business insights from large volumes of data in organizations. It is a platform for learning and discussing live case studies & business problems that can be solved using data science and interacting with the best mind in this field. The Kolkata workshop will have the flavours of manufacturing, banking and finance, and general sales and marketing. You can attend ‘Data Curry’ if you are: In the banking and financial sector dealing with a lot of data and want to know how to do fraud detection, how to benchmark performance, how to calculate lifetime value of a customer and how to do target marketing In the manufacturing sector and want to predict machine failures, pipeline leakage, analyse the products and processes for eliminating bottlenecks, ensure reservoir management and equipment failure prediction. In sales and marketing and want to gain and end-to-end customer perspective, customer life time value, target marketing, customer purchase decision, customer churn analysis. Business leader who wants to know what data science and big data analytics can help your business. About the Master Chef of Data Data curry boasts to have Dr. Murthy on board as the master data chef. Dr. Murthy is a renowned Indian data scientist with more than 15 years of work experience after PhD as a chief data consultant to fortune 500 companies.  Dr. Murthy has filed for 5 patents in Retail and Telecom Analytics, and has more than a dozen published international articles, presented in over 50 international seminars and conferences.  He is a brand name in the world of Data science. Over the past few years, he has been actively teaching and consulting data science and big data analytics to working professionals from diverse industries with wide range of experience. How to Register: You can register on our website – www.datacurry.in You can also post your opinions and doubts on our LinkedIn group Data Curry, and follow us on twitter @data_curry. Interact with Dr. Murthy and sort your doubts through these groups, so that you can make the most of his expertise and experience during the workshop.","excerpt":"Usha Martin Technologies in association with International School of Engineering (INSOFE) is hosting a one day workshop on data science and big data analytics, ‘Data Curry’, exclusive for 20 business leaders. The workshop will be mentored by Dr. Murty, renowned Indian data scientist and president of INSOFE. www.datacurry.in Kolkata Edition – 25th July, The Park […]","categories":["Deep Tech"],"tags":["big data for data science","Insofe"],"author_name":"AIM Media House","publish_date":"2013-05-27T12:39:55","publication_year":"2013","word_count":415,"keywords":["big data","data science","programming_languages:R","AI","big data for data science","Insofe","analytics","GAN","R","fraud detection"],"extracted_tech_keywords":["AI","data science","analytics","fraud detection","R","big data","GAN","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/data-curry-workshops-on-big-data-analytics-and-data-science\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10090770,"title":"AI Humans: The Connecting Dot Between AI and AGI","content":"Sam Altman, in his latest podcast with Lex Fridman, hypothesised the idea of ‘making humans super great’. In response to Fridman’s apprehension about AI being integrated into our world through prompt chains and number of interactions, Altman said, “I am excited about a world where AI is an extension of human will, and an amplifier of our abilities. . . maybe we never build AGI but we just make humans super great. Still a huge win”. This is an interesting proposition that we must sincerely ponder upon. Generative AI is still in its infancy, however the idea of the existing technology amounting to AGI in the near future has already been dismissed by a number of experts. But numerous applications offer an interesting insight and peek into what might be. Generative AI today—with its limited yet mind blowing variety of use cases—can not only help humanity become more productive or creative, but also more capable. This would in turn help in wealth creation and an overhaul in human intelligence and productivity—an ‘AI human’ if you will; which in turn would help train these systems through RLHF to upgrade itself. Thus, becoming a link between AI and AGI and ASI after that. Let’s think philosophically for a bit, any change in life goes through three stages—denial, acceptance and adoption. Not unlike life itself, the same is the case for artificial intelligence. Denial Humans are, first and foremost, sceptical towards a leap in technology, later developing curiosity and then, finally, learning to adopt it. For instance, when a proposed policy recommended the integration of calculators into the school mathematics curriculum at all grade levels for classwork, homework, and evaluation, teachers protested profusely against it, stating that “such an implementation would hinder students’ ability to learn basic mathematical concepts”. We are at one such juncture in human existence where many are wary of the innovation in generative AI. People fear being replaced by such systems, and their job roles being rendered moot. The situation is especially grim considering the Goldman Sachs report that suggests that AI could replace up to 300 million workers around the world. However, one could question the ludicrous nature of such reports and question the theories that these reports posit to prove their hypothesis, such as—“If generative AI delivers on its promised capabilities, the labour market could face significant disruption”. There are many who have likewise posed questions and raised existential alarms. A group of stakeholders and notable names in AI—Gary Marcus, Elon Musk, Apple Co-founder Steve Wozniak, among several others have called for a temporary pause on training systems beyond GPT-4 for at least six months because they believe that, “AI systems with human-competitive intelligence can pose profound risks to society and humanity”. This group also called upon AI labs and independent experts to “jointly develop and implement a set of shared safety protocols for advanced AI design and development”. They further asked for these protocols to ensure the safety of such systems “beyond a reasonable doubt”. It is to be noted that the stakeholders clarified that, “This does not mean a pause on AI development in general, merely a stepping back from the dangerous race to ever-larger unpredictable black-box models with emergent capabilities”. But some experts, like Meta AI chief Yann LeCun, do not agree with the components of the letter and have not agreed to signing the petition to pause all training beyond GPT-4. Acceptance Many have since come to accept the reverberation of generative AI while some are trying to innovate and push the existing limits. It is typical at the workplace today to use AI technologies—ChatGPT and alike—to automate repetitive tasks, improve productivity, and enhance decision-making. By using AI technologies to augment their cognitive and physical capabilities, individuals can work more efficiently and effectively. Founders are attempting to incorporate massive language models into a variety of other industries by developing tools to aid professionals in writing, coding, designing, producing media, and, now in law, finance and more. ‘Copilot for lawyers’, ‘Copilot for doctors’, ‘Copilot for designers’, and several others, are mushrooming all around us. The newest entrant within the financial space is BloombergGPT.  Users are also marvelling at ChatGPT API and various plugins that have use cases extending beyond imagination. Andrej Karpathy foresaw the development of ‘Software 2.0’ using neural networks in 2017, as pointed out in a report by Sequoia Capital. The report suggests we might witness the same innovation in software that he foresaw—a ‘Developer Tools 2.0’—in the future. Then, the real question is: How long before an ‘AI teammate’ after that? Adoption There are users who have truly adopted these technologies and are pushing the limits to break into something that’s more than what exists already. Big tech companies, such as Microsoft-backed ‘OpenAI’, Google and Meta, are contributing to the generative AI push. NVIDIA is also amongst one of the top contributors and beneficiaries in this space, with all its chips and GPUs being used to train these models. Firms such as China-based ‘Baidu’, with its latest chatbot are pushing the boundaries. The ‘ErnieBot’ contains 260 billion parameters, representing a 50% increase compared to ChatGPT’s parameters. The tech giants along with the users who have accepted and adopted it stand the chance to benefit from controlled, mindful and beneficial application of this technology, which has unimaginable potential but risks. Regardless of where individuals or organisations fall on the AI adoption lifecycle, whether in denial, acceptance, or full adoption, humans will remain a critical part of the loop. ChatGPT is a great example of this, where humans can improve the system and themselves—simultaneously through RHLF—becoming AI humans who assist in developing a reliable AGI for the future. “The thing is not that it’s a system that kind of goes off and does its own thing. But, it’s this tool that humans are using in this feedback loop. . . [It is] helpful for us for a bunch of reasons [because] we get to learn more about trajectories through multiple iterations,” Altman said, during his podcast with Fridman.","excerpt":"Altman said, “I am excited about a world where AI is an extension of human will, and an amplifier of our abilities.”","categories":["AI Features"],"tags":["AGI","ChatGPT","OpenAI","Sam Altman"],"author_name":"Shyam Nandan Upadhyay","publish_date":"2023-04-04T18:00:06","publication_year":"2023","word_count":1005,"keywords":["Go","ChatGPT","Meta AI","Sam Altman","artificial intelligence","OpenAI","AI","neural network","RLHF","AGI","generative AI","R"],"extracted_tech_keywords":["AI","artificial intelligence","neural network","generative AI","ChatGPT","OpenAI","Meta AI","RLHF","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/ai-humans-the-connecting-dot-between-ai-and-agi\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":14517,"title":"10 games that have successfully integrated Artificial Intelligence","content":"March 2016 marked an enormous success for use of AI in the gaming world. The Google DeepMind-developed AI program, AlphaGo beat one of the top Go players in the world, Lee Sedol. The event was convened over 4 days, and livestreamed on DeepMind’s YouTube channel, and other channels. This was just the beginning, as the AI plans to battle the current Go champion Ke Jie in a series of three games. The matches will be held during a five-day event in May, called “Future of Go Summit.” This event will mark a significant chapter for the history of AI. The use of AI in computer games isn’t something new. For years, game developers have been using AI to transform the behavior and decision-making process of non-player character (NPC). Computers and video games today furnish a scalable testbed for AI research and innovation. Moreover, machine learning bestows NPCs with several capabilities. They can improve their performance by learning from mistakes and successes, automatically adapt to the strengths and weaknesses of a player, or learn from their opponents by imitating their tactics. Analytics India Magazine presents to you an extensive range of AI-based games and computer games powered by AI, which have attracted several players globally. 1) AlphaGo AlphaGo beat Lee Sedol AlphaGo has revolutionized the AI space with the most recent feat that involved beating top Go player, Lee Sedol. AlphaGo essentially uses a Monte Carlo tree search to base its move upon previously “learned” knowledge from machine learning techniques. The program also uses artificial neural network by extensive training, both from human and computer play. The neural networks within the system were initially bootstrapped from human gameplay expertise. The program was initially trained to mimic human play, by trying to match the moves of expert players from recorded historical games. The database for the games contained around 30 million moves. 2) IBM Watson IBM Watson playing Jeopardy against two human opponents Today IBM Watson is a robust cloud-powered, AI-based platform; facilitating healthcare providers  to deliver optimal medical services. The IBM technology platform had famously won in 2011, in the game of Jeopardy, much before the success of the new AlphaGo program. Watson was pit against Brad Rutter and Ken Jennings, only to emerge victorious by defeating the two world champions. 3) Darkforest Mark Zuckerberg shows off Facebook Darkforest Named after Liu Cixin’s science fiction novel by the same name, Darkforest is also another Go program developed by Facebook. The program is based on deep learning, and uses a convolutional neural network. Like AlphaGo, Darkforest has also been tested against a professional human player at the 2016 UEC cup. The updated version of the game Darkfores2 combines the techniques employed by its predecessor, with Monte Carlo tree search. This kind of tree search method is generally seen in computer chess programs. 4) Deep Blue Deep Blue 1985 marked the beginning of work for Deep Blue with a ChipTest project at Carnegie Mellon University, which eventually gave way to Deep Thought. IBM took note of this development, and hired the development team, rebranding the project to Deep Blue. Interestingly, the development team had signed Grandmaster Joel Benjamin too. The program is essentially a chess-playing computer. Deep Blue became first computer chess-playing system in history to win both a chess game and a chess match against a reigning world champion. February 10, 1996 is the day when the AI defeated Garry Kasparov in game one of a six-game match. The program faced Kasparov another time in 1997, winning the six-game match, following which Kasparov accused IBM of cheating, demanding a rematch. IBM refused to accept the offer, and eventually retired Deep Blue. 5) F.E.A.R. Why F.E.A.R.’s AI is still the best? This game was developed by Monolith Productions in 2005. First Encounter Assault Recon, or F.E.A.R., as it is popularly called employs a first-person shooter game play. The player must contain supernatural phenomenon, and armies of cloned soldiers. For first time in gaming history, F.E.A.R. introduced a planner to generate context-sensitive behaviors, which is referenced today globally by several gaming studios. The AI-powered enemies can cleverly use the environment, finding cover behind tables, tipping bookshelves, opening doors, and so on. 6) Half-Life Half-Life 2 – Advanced AI in Action The sci-fi based plot follows a theoretical physicist, Dr. Gordon Freeman, stuck inside an underground research facility. The story escalates when the teleportation experiments go wrong unexpectedly. The game combines scripting and AI to avoid any interruption during the gameplay. Valve developed this game in 1998. The central character is accompanied by an AI security guard, somewhere towards initial stages in the game. The game also makes use of squad AI during latter stages. Moreover, the AI is very well tweaked and integrated brilliantly into the storyline. 7) Halo: Combat Evolved Halo: Combat Evolved. Developed by Bungie in 2001, the game is essentially a first-person shooter. The player assumes the role of the Master Chief within this game, battling various aliens on foot or inside vehicles. The AI allows the enemies to use cover wisely, while employ suppressive fire and grenades. The underlying technology, called “behavior tree” is one of the most popular technologies used by the gaming industry. Additionally, the game pays a lot of attention to minute details. Few years, back, this game has stood as one of the most popular games, wordwide. Users continue to play this game till date, and is sure to appear on any gaming list involving AI. 8) Stockfish Stockfish AI A free and open-sourced chess engine, Stockfish is available on several desktop and mobile platforms. The program was developed by Marco Costalba, Joona Kiiski, Gary Linscott, and Tord Romstad. Contributions also came from a community of open-source developers. The program has been currently ranked first or near the top of most chess engine rating lists. Features: It can use up to 512 CPU cores in multiprocessor systems. The maximum size of its transposition table is 1 TB. The AI implements an advanced alpha-beta search, and uses bitboards. It is characterized by its great search depth, due in part to more aggressive pruning and late move reductions. 9) TD-gammon TD-gammon Developed in 1992 by Gerald Tesauro, the program was a computer backgammon game. The name has been derived from a form of temporal-difference learning, specifically referred as TD-lambda. This helps in training the artificial neural net. The program achieved a level of gaming, slightly below that of the top human players of the game. Besides, the program explored strategies, not pursued by humans. This game also has integrated AI-based algorithms, which not only enhance the gaming performance, but also makes the gameplay more interesting. TD-gammon was developed in 1992 by Gerald Tesauro at IBM’s Thomas J. Watson Research Center. 10) Thief: The Dark Project Thief: The Dark Project The first-person perspective game combines elements of a stealth game, and is set in the medieval\/Victorian era. The central character is called Garret, who is a master thief in the game. The game was developed by Looking Glass Studios in 1998. Based on an accurate sensory model, the AI actors reflect the capability to respond realistically to lights and sounds, which also forms the primary technique the game build upon. Furthermore, the AI-based NPCs use audio recordings to voice their current state. This enables the player to comprehend what’s going on.","excerpt":"March 2016 marked an enormous success for use of AI in the gaming world. The Google DeepMind-developed AI program, AlphaGo beat one of the top Go players in the world, Lee Sedol. The event was convened over 4 days, and livestreamed on DeepMind’s YouTube channel, and other channels. This was just the beginning, as the […]","categories":["AI Trends"],"tags":["AI India","Artificial Neural Network","Convolutional Neural Network","Machine Learning India","simulation"],"author_name":"Amit Paul Chowdhury","publish_date":"2017-04-24T12:05:40","publication_year":"2017","word_count":1222,"keywords":["Go","machine learning","simulation","AI","neural network","innovation","Artificial Neural Network","Machine Learning India","Scala","deep learning","analytics","Convolutional Neural Network","AI India","AI research","R"],"extracted_tech_keywords":["AI","machine learning","deep learning","neural network","analytics","R","Go","Scala","innovation","AI research"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/10-games-successfully-integrated-artificial-intelligence\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10111419,"title":"Google Introduces Video Generation Model Lumiere, Leaves OpenAI Behind","content":"Google has introduced Lumiere, a text-to-video diffusion model designed to synthesise videos, creating realistic, diverse, and coherent motion. Unlike existing models, Lumiere generates entire videos in a single, consistent pass, thanks to its cutting-edge Space-Time U-Net architecture. It is designed to empower users to create visual content creatively, allowing the generation of realistic or surrealistic video clips up to five seconds in length. https:\/\/twitter.com\/omerbartal\/status\/1749971963403997252 It can animate still images, respond to natural language text prompts, and perform advanced video inpainting. It is built on a Space-Time U-Net architecture and a text-to-image (T2I) model operating in the pixel space, requiring a spatial super-resolution module for high-resolution image production. Furthermore, Lumiere offers stylised generation, allowing it to generate videos in the target style using a single reference image. This is achieved by leveraging fine-tuned text-to-image model weights. The model can animate still images or portions of them, filling in missing areas with high-quality results. Despite its limitations, such as not being designed to generate videos with multiple shots or scenes involving diverse motion, Lumiere represents a significant advancement in text-to-video AI generation. The project is currently a research project, and its release for broader use may be subject to addressing various policy considerations. As of today, OpenAI does not have a publicly available video generation model on their API. However, they are actively researching and developing technology in this area, and there are hints that something might be in the works with the release of GPT-5.","excerpt":"It can animate still images, respond to natural language text prompts, and perform advanced video inpainting.","categories":["AI News"],"tags":[],"author_name":"Siddharth Jindal","publish_date":"2024-01-24T17:46:47","publication_year":"2024","word_count":244,"keywords":["Go","API","OpenAI","GPT-5","AI","RAG","GPT","CLIP","R","llm_models:GPT"],"extracted_tech_keywords":["AI","GPT-5","OpenAI","RAG","R","Go","API","GPT","CLIP","llm_models:GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/google-introduces-video-generation-model-lumiere-leaves-openai-behind\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":58177,"title":"The Problem With Including AI In School Curriculum","content":"One of the main reasons to integrate AI in the current school curriculum is to make the upcoming generation familiar with technology. The Government of India and the educational board have been pushing for more artificial intelligence to be integrated into the education system, not from the perspective of enhancing it, but also with the intention of making young minds more aware and skilled when it comes to artificial intelligence. Today, children are curious about the smart conversational devices and AI used in applications like Siri and Alexa; some of them even wonder how Netflix gives them precise recommendations. Gradually, they will grow curious and try to learn what algorithms are, what a neural network is, and how they work. The Government of India and the educational board have been taking measures to make the existing school curriculum more AI-centric with a firm belief that the students will learn about AI, have fun and also take India forward. With the AI market estimated to reach around $15 trillion by the year 2030, the push for integrating AI in the school curriculum is fascinating. As India laces up to introduce AI in the school curriculum, there might be some issues that it has to counter first. The AI Initiatives For Indian Schools Some time ago, the CBSE released a curriculum, entailing the various ways AI can be integrated into the school curriculum. The curriculum gave a detailed account of how AI will be taught to children with their existing subjects. The curriculum came with a set of examples on how school subjects (English, Hindi, Science, Maths and Social Science) could be taught with AI applications and concepts mixed in with them. They intend to practice what is called ‘AI+X paradigm’ for better integration. In brief, with AI+X, a subject is selected for which the teacher has specific teaching skills, and it is called X. Then, research is done to find whether the AI concept fits well with the subject on the ground of commonality with it, and prepare a lesson plan which explains AI application in that subject or lesson, which is AI+X. For example, a chapter of class 6 social science intends to make the students understand panchayat raj and its functioning using the concept of neural networks in artificial intelligence. Building An AI Mindset Vs India’s Typical Pedagogy If the schools across India are able to create a mindset of the students to study AI and its concepts at such a young age, then it’s the best preparation for the country to advance in the field of AI and achieve its full potential. But in India, education has always been more about passing the examinations rather than learning the core of the subjects around the real-world applications. If at all countries like India want to introduce AI into the curriculum, they have to revise the education system to focus on problem-solving and developing AI applications rather than just passing exams. It is a known fact that learning AI includes a lot of coding, linear algebra and practice. So, whatever curriculum that has been introduced regarding AI needs to be more about making the students learn about the fundamentals of coding and linear algebra first and then move on to core AI concepts. Teaching students is one aspect, but paying attention to whether they have actually taken in a concept like neural networks or classification models is another thing. Also, we don’t know the curriculum introduced by the educational board is something that ultimately brushes up on various AI concepts using the existing subjects. Then at a stage where they have understood a little, telling them that much of the artificial intelligence and machine learning is a black box, might not bode well with them. This is unclear and yet to be seen. The introduction of AI curriculum still seems challenging to be taught in Indian public schools given India’s poor teaching methodologies and the AI understanding of teachers themselves. Students at school have different levels of ability to grasp, understand and apply knowledge. Whether every student will be able to understand these AI concepts remains to be seen. What About Teaching The Teachers? Having a well-structured curriculum and subjects are only one side of the teaching process, the most important aspect when it comes to teaching is the one who teaches the subject. Teachers in the existing systems already aren’t that qualified to teach straightforward and fundamental concepts, and it would definitely not be good to have less qualified teachers helping students study AI. So, teacher training on AI needs to be ensured before a complex curriculum on AI is brought in. This will help students get an in-depth and clear understanding of the AI concepts. And when we talk about teaching, whatever concepts being introduced in the middle school or high school curriculum are going to be fundamentals and foundations for future, if the teachers aren’t skilled enough to make AI simple for students, it will hinder further development. Closing Thoughts Much of the AI integration is going to start happening in 2020 or at least has begun in 2020. And, that’s why the government and the education boards must take all the measures when it comes to adding artificial intelligence to their curriculum. If the students of today become the generation of tomorrow, then it is imperative to prepare the future generation for the coming of AI\/ML. We hope educational boards in the country will make adequate arrangements, which can foster AI skills at an early stage.","excerpt":"One of the main reasons to integrate AI in the current school curriculum is to make the upcoming generation familiar with technology. The Government of India and the educational board have been pushing for more artificial intelligence to be integrated into the education system, not from the perspective of enhancing it, but also with the […]","categories":["AI Features"],"tags":["AI in school","ai initiatives in India","CBSE","Machine Learning","schools"],"author_name":"Sameer Balaganur","publish_date":"2020-03-06T13:50:57","publication_year":"2020","word_count":920,"keywords":["Go","artificial intelligence","schools","machine learning","AI","neural network","programming_languages:R","ML","Machine Learning","programming_languages:Go","CBSE","ai initiatives in India","AI in school","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","neural network","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/the-problem-with-including-ai-in-school-curriculum\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10065754,"title":"Entering a period of uncertainty, Twitter&#8217;s CEO tell employees","content":"In an email to his employees, Twitter chief executive Parag Agrawal addressed his their concerns and mentioned that there would be no layoffs. He did, however, add that the company is entering a state of uncertainty. This is because even though a deal has been agreed with Elon Musk for the company to be sold, it will take another six months for the deal to be completed. As per reports, the deal is expected to be sealed by the end of 2022. Twitter employees were already precarious, given the whole situation has dragged on over the month. As per reports, many employees have voiced their concerns about Elon Musk taking over the micro-blogging platform. As per another report, as low as 10 per cent of Twitter’s employees are enthusiastic about the takeover. A few days ago, Agrawal even said that once the deal is sealed for Musk to buy Twitter, he doesn’t know which direction the platform will eventually go. When the deal was finally announced, Agrawal tweeted, “Twitter has a purpose and relevance that impacts the entire world. Deeply proud of our teams and inspired by the work that has never been more important.” It is also uncertain whether Agrawal will remain the CEO of Twitter once the deal is completed. However, as it stands, he will be the CEO for the time being. Twitter has a purpose and relevance that impacts the entire world. Deeply proud of our teams and inspired by the work that has never been more important. https:\/\/t.co\/5iNTtJoEHf— Parag Agrawal (@paraga) April 25, 2022 The big question in everyone’s mind is which direction Twitter will take once Musk officially takes over. There are a lot of speculations in this regard. The tech mogul has been a critic of the platform he just bought, and over the last couple of months, he has advocated for free speech and making the Twitter algorithm open-source. https:\/\/twitter.com\/elonmusk\/status\/1518677066325053441 It is indeed true that we are entering a period of uncertainty, not just for those on Twitter’s payroll but the whole world. Among major speculations, the most-talked-about thing about the changes we could expect in the microblogging platform is the introduction of an edit button. Even Musk teased his followers with a tweet in this regard last month. The Whole Story Twitter accepts Elon Musk’s buy-offer at USD 44 BnTwitter adds a GitHub repository called ‘the algorithm’, then deletes itJeff Bezos ignites China conspiracy theory as Musk seals Twitter dealMusk promotes free speech after Twitter deal","excerpt":"The deal for Musk to buy Twitter is expected to be completed by the end of 2022.","categories":["AI News"],"tags":["Musk buy Twitter","Twitter Elon Musk","Twitter Musk"],"author_name":"Pritam Bordoloi","publish_date":"2022-04-26T21:46:47","publication_year":"2022","word_count":415,"keywords":["Go","Musk buy Twitter","programming_languages:R","AI","Twitter Elon Musk","Twitter Musk","programming_languages:Go","Git","RAG","GitHub","R"],"extracted_tech_keywords":["AI","RAG","R","Go","Git","GitHub","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/entering-a-period-of-uncertainty-twitters-ceo-tells-employees\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":54857,"title":"Top 5 Most Prevalent Cyberattacks &#038; Ways To Avoid Them","content":"Robust cybersecurity initiatives are essential to avoid the risk of non-compliance with new data privacy regulations such as PDP and GDPR. While failing to adhere to these regulations result in paying huge fines, it also negatively impact brand value. So what can one do to avoid these common cyberattacks? Since the adoption of public cloud is on the rise, companies are heavily relying on them to safeguard their data from hackers. Undoubtedly, cloud providers have helped firms to focus on business rather than to struggle protecting data. However, companies should also focus on business operations rather than just protecting their repositories and services on the cloud. Here are the top most common cyberattacks and practices to fortify them in 2020. Ransomware Organisations that are unprepared to fortify ransomware are highly prone to these attacks. And according to reports, nearly 60% of ransomware attacks are delivered through email as embedded URLs. While every sector has been negatively impacted by it, ransomware is highly prevalent in retail and healthcare (source: BBR Services 2017). According to another report, the average ransom ranges from $500 to $2,000 for SMBs. Therefore, businesses need to dodge skyrocketing ransoms and avoid financial losses. Usually, hackers encrypt the data and ask to pay for obtaining the key for decrypting information. Besides, one cannot assure whether they will provide the key after the payment. Consequently, one should always back up their data, thereby restoring it in case of any ransomware attack on the original data. Distributed Denial of Service (DDoS) The idea behind the distributed denial of service (DDoS) is different than that of other ransomware and phishing attacks. The mission is not to get direct benefits but cause hindrance to businesses operations and government services. Highly familiar with large firms, DDoS attacks involve flooding networks with a colossal amount of data to cause congestion and crashing of services. Hackers hijack personal computers of different users and submit additional traffic, and users arent often aware that their systems being used for the DDoS attack. While DDoS attack can come in many ways like volumetric, application-layer, and protocol attacks, the best approach is to move to the cloud, thereby outsourcing the attack for the cloud provider to handle. However, if one does not want to move their sensitive data to the public cloud, opt for various services that vendors provide to safeguard the data Man-In-The-Middle In man-in-the-middle attacks, hackers have their play between a client and a server. A typical MITM begins when devices like mobile or laptops are connected to public Wi-Fi, which is very often the case in today’s connected world. Perpetrators exploit the technicality in public Wi-Fi to penetrate networks and collect critical information from the connected devices. Consequently, to avoid such attacks, one should ensure that employees do not connect devices to public Wi-Fi, which have business email logged onto it. However, this is not the safest approach because, if the employee has saved credentials on productivity applications, hackers can still get them. But it can drastically reduce the potential of being a victim of MITM attacks. Another productive approach for safeguarding information on public networks is to use virtual private networks (VPNs). Although it requires investments, but can completely block the access of users outside authority. One can safely login with public networks with activated VPN without worrying about lurking hackers. Phishing Phishing has been known for a while now, but it continues to be one of the effective hacking techniques for getting hold of sensitive information. It is also believed that the Russian interference in the 2016 U.S. presidential election started with phishing. One common practice of hackers is to send password change request, which then takes to the look-alike page of the email service provider. Users often do not identify the phoney landing page and provide their credential, which results in giving access to sensitive information. The workaround for phishing is to educate employees not to be falling for these emails and be critical of the links that require passwords and usernames. Social Engineering Social engineering is an attempt to take advantage of human psychology and find effective ways to reveal confidential information. While phishing is a type of social engineering, it doesn’t involve unlawful activity through physical presence. On the other hand, social engineering goes a set further adopt numerous techniques that require physical presence to execute cybersecurity breach. For one, one can get close to the users’ device and steal information. Rejecting the request of help or offers of help is highly recommended for ensuring one does not hand over their laptops and desktops to others unless confident about the person’s authorisation. Besides, one should provide training for employees to alter them of potential common cyberattacks.","excerpt":"Robust cybersecurity initiatives are essential to avoid the risk of non-compliance with new data privacy regulations such as PDP and GDPR. While failing to adhere to these regulations result in paying huge fines, it also negatively impact brand value. So what can one do to avoid these common cyberattacks?  Since the adoption of public cloud […]","categories":["AI Trends"],"tags":[],"author_name":"Rohit Yadav","publish_date":"2020-01-29T16:58:29","publication_year":"2020","word_count":785,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","RAG","ViT","GAN","R"],"extracted_tech_keywords":["AI","RAG","R","Go","GAN","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-5-most-prevalent-cyberttacks-ways-to-avoid-them\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":1217,"title":"Dharavi gets its first IoT connected market to let shoppers have an informed choice","content":"Dharavi gets its first IoT connected market to let shoppers have an informed choice Researchers, including those from India, have brought Internet of Things into the picture to connect shops at Dharavi Mumbai, which is one of the largest slums in the world. The sole aim of the connected market is to give buyers an enhanced shopping experience and helping them make an informed choice. Connecting 30 shops in Dharavi’s markets to a physical web through 100 devices called beacons, is all a part of project which is a part of Google’s Internet of Things Technology Research Award. How does it help? When customers using smartphones are in the proximity of any such beacon-enabled shop, they get a notification via bluetooth. The customers can then browse through products available at the shops using an interactive interface, thus giving a user an overview of the products of all the shops in the area and hence facilitating them in making an informed choice. The researchers reportedly said that in the greater scheme of things, this development is set to bring a change in the way Dharavi is perceived. Researchers from various institutes such as Swansea University in the UK and Indian Institute of Technology (IIT) Bombay, who were a part of this project, hope to not only boost the customer-seller relation but also attract more buyers to the markets of Dharavi. Quite infamously famous as a slum with narrow lanes, Dharavi has developed itself, to a large extent, as a self sustained informal economy. Various industries such as leather, garment, recycling and pottery form an important of Dharavi’s economical burst. And thanks to IoT, it now would be one of the first markets to have network connectivity allowing them to send and receive data and interact with each other. Shops which are a part of deployment are given a poster that asks people to turn on the Bluetooth to experience the physical web. The Google IoT Research Award gives students access to 100 beacon devices designed to allow any smart device to interact with real-world objects – in this case, shops in Dharavi – without having to download specific applications.","excerpt":"Researchers, including those from India, have brought Internet of Things into the picture to connect shops at Dharavi Mumbai, which is one of the largest slums in the world. The sole aim of the connected market is to give buyers an enhanced shopping experience and helping them make an informed choice. Connecting 30 shops in […]","categories":["AI News"],"tags":[],"author_name":"Srishti Deoras","publish_date":"2017-01-13T11:06:58","publication_year":"2017","word_count":358,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","Aim","ViT","R"],"extracted_tech_keywords":["AI","Aim","R","Go","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/dharavi-gets-first-iot-connected-market-let-shoppers-informed-choice\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":60169,"title":"Biggest Mistakes You Should Avoid While Working From Home","content":"While everyone was making vows this January to make it the best year for their career, COVID-19 seems to have pulled the plug on their expectations for 2020. While remote working has become commonplace over the last few years, the majority of office workers have been struggling to find that balance between work and home during this pandemic. While it may be just another day at work for remote workers, office workers have been adjusting and taking it one day at a time. If you have been a remote worker, you have probably avoided making these mistakes. However, if you are someone who is dealing with work from home for the first time, be ready to take notes: Using A Weak Or Open WiFi Using a weak WiFI when one is conducting official work can disrupt communication. All critical morning meetings could be delayed, greatly affecting your efficiency. What is more, using public WiFI is dangerous in general, and when we talk about using the same for official work, it becomes even more so. The sensitive company data you may be handling also become susceptible to cyberattacks, like ransomware or hacks. Sharing Pictures Of Your Work Desk Sharing photos of your workspace at home – no matter what – can be dangerous. It could potentially expose some sensitive company data that you are working on to the public. However, if you still wish to, make sure you have removed any company-related material from your picture\/video frame. Leaving Your Work Device Unlocked This is a common habit of people working on PCs or laptops at offices. Many people leave their devices unlocked, and the repercussions they might have to face in the office space is different from what they might have to when at home. If there are children at home, they could either delete or make changes to your code, or modify some things in your excel sheets. This in turn, could lead to spending a lot of time looking for and correcting those mistakes. Sharing A Laptop Working from home in the time of a lockdown like this can become dull – if not for you, then for other family members at home. One thing to keep in mind is never to share your work device with anyone. Even if you do, make sure there is a separate profile where they can operate freely without affecting your work data. Using The Same Browser For Work And Personal Use Using the same browser for both personal and professional use not only results in a change in the kind of data and advertisements you see because of the cache, but it can also bring about other variations. Plus, it would slow down your work while browsing and researching. It is better to have a separate browser or a different profile altogether for personal and professional use. Not Using VPN While using VPN will get your bandwidth down, and in some cases, charge you a small fee, using a VPN will ensure an extra layer of security for your work. VPN especially comes in handy when you are exchanging confidential information over a network. Mixing Personal And Professional Lives While it is called ‘work from home’, many people tend to work from their bed or their couch in a dark room. While it might feel comfortable to do so, working that way will only reduce productivity for most people and cause significant hindrance to their work. Have a separate office space created in your home, which is brightly lit. Also, while it is important to ensure that you are comfortable, laziness should not come in the way. Make sure you try to follow the same routine that you follow at the office.","excerpt":"While everyone was making vows this January to make it the best year for their career, COVID-19 seems to have pulled the plug on their expectations for 2020. While remote working has become commonplace over the last few years, the majority of office workers have been struggling to find that balance between work and home […]","categories":["AI Features"],"tags":["International Affairs","network firewall","Work from Home"],"author_name":"Sameer Balaganur","publish_date":"2020-03-27T17:00:00","publication_year":"2020","word_count":620,"keywords":["network firewall","programming_languages:R","Work from Home","ViT","International Affairs","R"],"extracted_tech_keywords":["R","ViT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/biggest-mistakes-you-should-avoid-while-working-from-home\/","complexity_score":1,"technical_depth":3,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":2237,"title":"Complex event processing: True integration of Things to Internet","content":"There is a subtle difference between building internet of things and “Thingify the Internet”. Using advanced IPv6 compatible protocol any device gets an IP address, and one can start from there: Simple machine to machine communication. Add a transport layer and you have your processes communicating with each other. Having stated that, this draws quite some consequences: We can’t rely on the infrastructure in between. The IP packet travels through the Internet, and we don’t know who is handling it. This means any computational processing needs to be performed at the edges: Either on the Smart Thing (sensor node) or on the (cloud)server. This draws the question, of where to do the processing? E.g. how much context processing do we want to do on the Smart Thing? How much on the server? A simple answer to is to limit the communication effort, and try to do as much processing on the server! SmartThings\/sensor nodes usually have energy and resource constraints. So it will depend on the application developer: Depending on the scenario, he\/she will have to balance Smart Item processing and communication efforts. However, in order to give the application developer, the freedom to balance, we must support them as much as possible from the server side. And that is where Complex Event Processing (CEP) jumps in! So, How does CEP work? In a truly micro service oriented world, we have this global service anybody can subscribe to and use: The application developer writes the CEP scripts (e.g. in EPL, Event Processing Language), and orders her SmartThings to publish their messages to the CEP service. The CEP service then executes the relevant scripts and does message\/event processing, transforming simple data points (in time), into enriched information Such derived data streams can further be extended to advanced engines to generate complex reports and apply data science Key architecture ingredient: Low Latency Requirement Most IOT deployments are event driven: each part of the system reacts to the events happening in the other parts, potentially generating new events. Complex event processing (CEP) engines in charge of interpreting, filtering, and combining primitive events to identify higher level composite events. Hence to handle such interactions, CEPs are required to have low latency. Before we close, lets touch upon a few use cases leveraging CEP: Network Management Suddenly thousands of alarms go off at a data center. A complex event processing engine filters the events and infers a new event that an external network link has gone down. Operators work to switch over to a backup telecom link Safety Multiple sensors in a moving vehicle sense an object on the road ahead. The vehicle immediately warns the driver and works to detect what they object is. It is confirmed from analysis of video data that it is a pedestrian and the driver still hasn’t applied brakes. The vehicle decides to apply maximum brake force automatically. If a collision is about to occur the vehicle deploys pedestrian airbags timed to minimize injury. Financial A payment platform receives thousands of payment events per minute during peak hours. A CEP engine is used to infer possible fraudulent payments from patterns of purchases. Algorithmic Trading A trading algorithm watches social media and industry new sources for information that can influence the price of stocks. It seldom acts on such information alone but may seek confirmation from other events such as large trades with a particular signature. If it discovers bullish social media chatter followed by large, aggressive buys it infers that the stock is in the early seconds of news driven momentum. Such strategies are dangerous as aggressive buying is no confirmation that news is genuine or meaningful to the stock price.","excerpt":"There is a subtle difference between building internet of things and “Thingify the Internet”. Using advanced IPv6 compatible protocol any device gets an IP address, and one can start from there: Simple machine to machine communication. Add a transport layer and you have your processes communicating with each other. Having stated that, this draws quite […]","categories":["IT Services"],"tags":[],"author_name":"AIM Media House","publish_date":"2017-08-07T08:27:44","publication_year":"2017","word_count":612,"keywords":["data science","Go","AWS","AI","cloud_platforms:AWS","programming_languages:R","programming_languages:Go","RAG","R"],"extracted_tech_keywords":["AI","data science","RAG","AWS","R","Go","cloud_platforms:AWS","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/complex-event-processing-true-integration-of-things-to-internet\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10130855,"title":"What is Computer Vision and How it Works?","content":"In the above video you will learn what is computer vision and how does it work in a simplified and easy to understand hindi language. What is Computer Vision? Computer vision is a field of artificial intelligence that enables computers and systems to derive meaningful information from digital images, videos, and other visual inputs. It seeks to automate tasks that the human visual system can do. Computer vision algorithms interpret, understand, and reconstruct the visual world, translating visual content into decisions or actions How Computer Vision Works? Computer vision employs various algorithms and techniques. These include image processing, machine learning, and pattern recognition. Image processing techniques help in enhancing the image quality. They also prepare data for further analysis. Machine learning plays a crucial role in computer vision. It allows computers to learn from existing visual data. As a result, they can make intelligent decisions based on new images. For example, a computer can learn to recognize different objects in an image. This is done by training it with numerous labeled images. These are known as training datasets. The process starts with image acquisition. Cameras or sensors capture visual inputs. These inputs are then digitized for processing. Image ProcessingSoftware processes the digitized images. It detects, classifies, and understands the images without human intervention. Applications of Computer VisionIt has many applications. These include autonomous driving, medical imaging, and surveillance systems. Each uses distinct models and algorithms tailored for specific tasks. Future of Computer VisionThe field is rapidly advancing. New techniques and improvements emerge regularly. These advancements continually expand the potential applications of computer vision.","excerpt":"Learn what is computer vision and how it works in hindi, the video explains in clear diagrams and graphics which is perfect for beginners.","categories":["AIM Videos"],"tags":["Top Trend"],"author_name":"Pabitra Moharana","publish_date":"2024-08-01T11:47:53","publication_year":"2024","word_count":263,"keywords":["Top Trend","Go","API","machine learning","artificial intelligence","programming_languages:R","AI","programming_languages:Go","Git","computer vision","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","computer vision","R","Go","Git","API","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/videos\/what-is-computer-vision-and-how-it-works\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10041086,"title":"DBA In Data Science &#8211; India&#8217;s Only Applied Research Program For Technology And Business Leaders","content":"Last year, the International School of Engineering (INSOFE), India, and the Rennes School of Business, France, launched the world’s first Doctorate of Business Administration (DBA) in Data Science. Analytics India Magazine caught up with Dr Dakshinamurthy V Kolluru, Founder, President & Chief Mentor at INSOFE, to understand the program, its target audience, objective, and scope. Excerpts: AIM: Who typically benefits the most from a DBA degree? Dr Murthy: DBA is a doctoral degree generally designed by business schools to benefit working professionals at senior manager or higher designations. They are typically individuals with decades of experience who want to become thought leaders, solve challenging problems, and push their industry ahead in technology or any other niche area of interest. The idea is, in any industry you need 90 percent of people to sustain the operation and the remaining 10 percent to push the industry forward. To achieve that, they have to become thought leaders; they should be able to look and predict trends and see how they can apply that in their operations so that their business stays, innovates, and differentiates itself. This need for thought leaders led to the conception of the DBA program. B-schools realise that the potential candidates for this program are already industry veterans. So the DBA program is designed in such a way that it combines students’ working knowledge with a thesis to help them gain the skills required to lead teams and innovate. AIM: What is the difference between a DBA in Data Science and a PhD in Data Science? Dr Murthy: Both PhD and DBA are doctoral degrees. They are the highest degree a university can confer on any student. However, a PhD is primarily designed for freshers (holding a bachelors or a masters degree) who want to get into academia and\/or do research, and is offered as a full-time program. A DBA on the other hand, is designed specifically for working professionals who want to work in an enterprise and help their industry innovate. It is for people who want to become thought leaders within the enterprise they are working for. Talking about data science in particular, which is a hands-on field, it is important that the candidates be trained not just in thought leadership but also in hands-on activities since their work requires them to lead teams that are super technical and understand the tech development better than an average manager. A DBA in data science is a combination of technical skills and business management with a lot more hands-on components than a traditional DBA program. “INSOFE’s DBA in Data Science is offered part-time for senior-level employees and combines technical skills and business management with a lot more hands-on components than a traditional DBA program.” AIM: Can you give examples of a few DBA thesis that are currently happening at INSOFE? Dr Murthy: We started this program just one year back and in a short span of time, we became one of the largest research centres in data science with more than 40 doctoral students actively engaged in R&D! Our students come with varying work experience and academic interests which is reflected in their thesis. I would like to demonstrate this variation through two examples. The first thesis is done by a senior leader who has to deploy data science applications for several clients. Her job is to ensure that the data science application is deployed in such a way that the company derives the highest ROI. This not only requires technical expertise but also requires the intelligence and skills to deal with the team and address challenges such as dealing with the fear generated due to automation and conflict with ethics and legalities. The final objective is to design a planning tool that enables a business manager to predict the risks they encounter when they deploy an application. She will additionally design the mitigation strategies as well so that the chances of the application going live and giving desired ROI is highest upon execution. Her thesis is rooted 50 percent in business analysis and the rest in tool development. Many people in India speak broken\/poor English and many documents also contain information written in the same format. One of our students is building a novel NLP product\/architecture to understand such text that has lost all the sequential information. When fully developed, such technologies will help the less educated masses access the benefits of AI much better. This thesis involves 75 percent tech component and 25 percent business component. This is a one of a kind thesis with great potential and we are even thinking of patenting and productionising it. Though the DBA students are not required to do the actual coding they have to plan and carry out the entire development activity of their thesis. The student will work with data scientists at INSOFE who will help them with the computer programming aspects of their thesis. AIM: Do students have publication opportunities for the research projects they develop? Dr Murthy: Every student gets at least two publication opportunities for the thesis they develop. Although the DBA is a part-time program, it is as rigorous as a regular PhD program. Every week one of our faculty interacts with the student to read, understand, and help them develop their thesis. These regular in-depth discussions ensure the thesis developed is of the highest quality and reliable for industry application later on. Every student in the program conducts an industrial survey, understands the current situation in that industry, and builds a prototype that helps resolve issues in the industry. Many times these tools are patentable with an opportunity to productise, should the student wish to do so. “The DBA is a rigorous and intense 3-year doctoral program resulting in two publications and a patentable product opportunity.” AIM: Could you expand on the demography of DBA candidates at INSOFE with regard to their educational background, industry experience, and age? Dr Murthy: We are truly a global program, hosting students from 11 countries; a huge feat for a year old program! All of our students have over 8 years of experience and around 87 percent have more than 10 years of experience working in posts such as managerial director, CXO, and VP. With regard to educational background, 70 percent of the students have a master’s degree and 30 percent have a bachelor’s degree. Though we have a rich cohort of students, we are focused on making the   gender distribution more equitable in the future. At this point, 77 percent of our students are men and 23 percent are women. However, we foster an inclusive culture that encourages students with diverse identities, expressions, abilities, perspectives, and backgrounds at INSOFE. AIM: Why did two institutes Rennes SB and INSOFE come together to do a DBA? Dr Murthy: This is a techno-managerial program, it is not enough if we create strategy makers. Rennes School of Business is one of the top B-schools of the world with several of its programs ranked among the top 100 business programs of the world. It also holds the rare distinction of having triple-crown accreditation (accreditation from top global bodies–AACSB, AMBA, EQUIS). They bring tremendous business depth to the program. On the other hand, INSOFE is one of the top data science institutions. We have prestigious accreditations and around 500 companies recruit our students every year! We combined our technical strength with Rennes’ business expertise synergistically to create a world-class techno-managerial doctoral program. That’s what makes this program really unique. AIM: What is the ROI for this DBA program? Dr Murthy: One of the commonly asked questions is whether it is worth doing a doctoral degree in data science. AI and ML are considered one of the highest in-demand disciplines in the world. A doctoral degree can help a candidate rank among the top 1% in this field. The idea is that every enterprise–small, medium and large is desperate to implement AI and ML; it is probably easier to hire a few data scientists in the team to do so. But a huge gap lies in understanding where AI can be implemented in their enterprise, how to work with their technical team to get those products done in real-time, and ensure the delivery of the right returns. That essential leadership skill with an understanding of data science is missing. Hence there is a tremendous demand for managers and leaders who can own those innovations and this DBA is a near perfect way to acquire all those skills at the same time. AIM: Can you tell us how the faculty profiles and experience at Rennes SB and INSOFE play a role in guiding students’ research? Dr Murthy: There are 160 faculties between Rennes and INSOFE. INSOFE has 70 faculty members who are experts in data science, have built products, worked in consulting positions, and constantly engaged in different research projects. Between INSOFE faculty, we have 75 patents and 300+ research publications and more than 1,000 tech presentations. Further, the select 90 faculty members from Rennes are experts in research, consulting, and advisory services from a business perspective. This faculty industry experience combined with the classroom teaching and mentoring help our students develop a more in depth understanding of real world problems in the industry they work in and select appropriate practical research projects to address these problems. “The uniqueness of this DBA lies in the access to more than 1000 hours of personalized guidance offered by world-renown mentors and expert data scientists.” AIM: How much time do students get to spend with mentors during this DBA program and how is it beneficial to their experience? Dr Murthy: The beauty of the program is close personal interaction with mentors. Every faculty and DBA student meet at least once a week and have a thorough discussion on the student’s area of interest. In addition, we have allocated data scientist experts to assist every student for 6-8 hours a week in developing their thesis. All DBA students at INSOFE, hence get more than 1,000 hours of interaction with expert data scientists guiding them on their DBA during their three years of doctoral study. This ensures that our students receive constant support and encouragement from experts in the field and further validates the research they do. Find further details here.","excerpt":"The DBA is a rigorous and intense 3-year doctoral program resulting in two publications and a patentable product opportunity.","categories":["AI Trends"],"tags":["Data Science","Insofe","insofe dba","Masters in Data Science"],"author_name":"AIM Media House","publish_date":"2021-05-31T18:00:00","publication_year":"2021","word_count":1711,"keywords":["data science","Go","AI","Masters in Data Science","ML","insofe dba","Insofe","NLP","RAG","Aim","ViT","analytics","Data Science","R"],"extracted_tech_keywords":["AI","ML","NLP","data science","analytics","Aim","RAG","R","Go","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/dba-in-data-science-indias-only-applied-research-program-for-technology-and-business-leaders\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10089128,"title":"Former Google Legal Head Joins Anthropic","content":"Janel Thamkul who served the role of Legal Manager at Google for the past seven years has left the tech giant to join Anthropic as Deputy General Counsel. Last month, Google announced a partnership with Anthropic, a startup founded in 2021. Simultaneously to this news, the startup has issued their core views on AI safety. “At Anthropic our motto has been “show, don’t tell”, and we’ve focused on releasing a steady stream of safety-oriented research that we believe has broad value for the AI community,” the company stated in their recent blog post. The 6000 word post extensively delves into the ‘three-types’ of research Anthropic is pursuing. Anthropic makes ‘Claude’ which, much like ChatGPT, is an artificial intelligence chatbot. Fascinatingly, Thamkul is not the first employee who has left the company to join Anthropic. The CEO of Anthropic Dario Amodei also worked at Google right before founding the company. During his ten month period at Google, he worked on extending the safety of neural nets for their AI systems. Thamkul joined Google in August 2015, where she co led the development of ethical AI, managed the product counsel team advising on AI, and provided legal guidance in regulations like DMCA, EU Copyright Directive, EU Artificial Intelligence Act, and Illinois’ Biometric Information Privacy Act. She also focused on ethical research and development of several AI frameworks. Google employees leaving the company to kick off their own start up is not surprising. Cohere’s founders Aidan Gomez, Nick Frosst, and Niantic’s John Hanke are some of the ex-Googlers who have entered the AI field as the tech giant’s competitor in their specific domains.","excerpt":"After seven years of serving Google, Janel Thamkul has left the tech giant to join its competitor.","categories":["AI News"],"tags":["ai investment","Anthropic","Startups"],"author_name":"Tasmia Ansari","publish_date":"2023-03-10T17:47:15","publication_year":"2023","word_count":270,"keywords":["Anthropic","ChatGPT","Go","artificial intelligence","AI","GPT","llm_models:GPT","ai investment","Startups","AI safety","R","startup"],"extracted_tech_keywords":["AI","artificial intelligence","ChatGPT","Anthropic","R","Go","GPT","AI safety","startup","llm_models:GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/former-google-legal-head-joins-anthropic\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10130716,"title":"Skilled Techies Don’t Want to Work for Big-tech Companies Anymore","content":"Despite the allure of big tech’s big-ticket jobs, many top-tier talents from Ivy League schools, award-winning researchers, and prolific authors are shunning giants like Google, Microsoft, Amazon, and Meta to work for smaller or mid-sized companies. A Reddit discussion highlighted that this prevailing sentiment stems from a desire to escape the corporate politics often inherent in large organisations. As one ML engineer puts it, “Why deal with the politics of a big company when you can get funding for your projects?” The freedom and autonomy that smaller companies promise can often be more appealing than the bureaucratic hurdles of the big tech. Burnout is another critical factor. Many skilled professionals are now prioritising work-life balance over the prestige associated with working for a tech giant. This sentiment echoes a broader cultural shift where mental health and personal well-being are becoming increasingly important. For Indian tech professionals, one of the main attractions of big-tech jobs is the higher salary package, coupled with their long-standing desire to work at one of these prestigious companies. Less Inspiring Work Financial motivations, while important, are not always the driving force. The nature of the work itself also plays a role. Some ML experts find the projects at big-tech companies less inspiring. “Most of the projects at MAANG [Meta, Amazon, Apple, Netflix, and Google] companies are boring,” one contributor mentioned. There’s a preference for roles where they can have a more significant impact on the AI roadmap, which smaller firms often provide. Moreover, the elaborate hiring processes at big-tech companies can be off-putting. As an ML engineer pointed out, “Getting into MAANG is an entirely separate field that requires you to study and practise an entire hobby\/career path unrelated to your ML expertise.” Busy ML leaders might not have the time or inclination to master the intricate and often lengthy recruitment processes of these giants. Additionally, the work environment and corporate culture in these tech giants can be stifling. One ex-employee described their experience: “Google was a fun, exciting, and innovative place to work in 2004. Twenty years later, it’s decayed into the same bland, vapid, beige-coloured evil as Microsoft.” The transformation of these workplaces over time often leads to disillusionment among those who seek dynamic and innovative environments. Another compelling reason is the opportunity for a more significant research agency and visibility at smaller firms. “I prefer smaller! Much cosier, less politics, and most importantly: waaaaay more research agency,” said an ML professional. In smaller companies, top talents often have more freedom to pursue their research interests without the constraints of a rigid corporate structure. As one ML researcher summarised, “It’s a trade-off for sure, but you get more autonomy. R&D changes so fast, so not having that autonomy can feel a little scary.” At the same time, it is undeniable that big tech produces some of the top research. People who cite autonomy as the reason, though correct, miss out on the part of producing SOTA research at big tech. So, while big-tech companies can offer substantial salaries, many skilled professionals find that the trade-offs in terms of autonomy, work-life balance, and ethical considerations make smaller firms more attractive. The Same Goes for IT The situation is only a little different for Indian IT. Though the research and development at these companies need good talent, Indian researchers do not want to join them. According to several predictions, the number of CS graduates by 2025 is going to be three or four times more than 2020. This just shows the huge supply of graduates in the field. But the same amount of jobs are not available in the Indian sector. Forget big-tech, Indian IT too is not attractive to the graduates from the country. Though India is seeing an increase in talent retention, there seems to be a surplus of underskilled STEM graduates. In India, the situation is complex. It is extremely difficult to find good, or even decent, software engineers with coding skills for such small compensation. Meanwhile, the ones who have the skills are either already working for startups at a higher package, or have moved abroad for better opportunities. The reluctance of recent graduates to pursue careers in Indian IT can be attributed to the prolonged stagnation of entry-level salaries, which have remained at INR 3.5-4 LPA for over a decade. High-paying product companies with compensation packages ranging from Rs 10-20 LPA have become more attractive.","excerpt":"The same goes for Indian IT.","categories":["AI Features"],"tags":["AI Impacts","Career"],"author_name":"Mohit Pandey","publish_date":"2024-07-30T13:47:07","publication_year":"2024","word_count":734,"keywords":["Go","API","funding","programming_languages:R","AI","ML","programming_languages:Go","AI Impacts","GAN","R","Career","startup"],"extracted_tech_keywords":["AI","ML","R","Go","API","GAN","startup","funding","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/skilled-techies-dont-want-to-work-for-big-tech-companies-anymore\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10091924,"title":"Bengaluru-based Cognecto Raises INR 4Cr in Seed Round","content":"Cognecto, a Bengaluru-based startup has secured a significant INR 4 crore funding. The startup recently raised the amount in a seed round led by Inflection Point Ventures (IPV), a renowned investment firm. The funding will be used for the development of regenerative AI, to provide business and operational insights for the mining and construction sectors. Regenerative AI systems can learn and adapt to new situations, similar to how biological systems regenerate and adapt to changes in their environment. This involves integrating data from existing and highly distributed telematics systems, along with operator inputs through mobile apps and wearable devices. Citing the need for real-time data in the mining industry, Vikram Ramasubramanian, partner at IPV, said, “The utilization of AI technology in the mining and construction sector is a significant advantage. These industries require real-time data to optimize time and resources.” He also acknowledged Cognecto’s potential to drive infrastructural development in the industry. Rohet Sareen, business development head at Cognecto, shared his excitement about the partnership and the opportunity to technologically revolutionize the mining industry. He expressed gratitude for the guidance and support provided by IPV in raising this seed round. With the additional funding, Cognecto is poised to accelerate the development of its product and further establish itself as a leader in the field of AI-driven solutions for the construction industry. The company’s unique approach, industry-specific expertise, and global presence make it a key player to watch in the ever-evolving landscape of mining and construction technology. Founded in 2020, Cognecto provides AI-based solutions for the construction industry, particularly in analytics and managed services across Asia, Europe, Australia, and Africa. Cognecto has been honored as one of the most innovative startups by IIT Delhi in 2023. Considering the booming global market for the telematics industry. Notably, Asia commands a remarkable 25% share. Furthermore, the highest growth rate in the telematics industry is projected in Asia for the year 2022-23.","excerpt":"The seed round was led by Inflection Point Ventures.","categories":["AI News"],"tags":["Startups"],"author_name":"Tasmia Ansari","publish_date":"2023-04-20T11:59:55","publication_year":"2023","word_count":318,"keywords":["funding","programming_languages:R","AI","generative AI","analytics","Startups","R","startup"],"extracted_tech_keywords":["AI","analytics","generative AI","R","startup","funding","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/bengaluru-based-cognecto-raises-inr-4cr-in-seed-round\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":35455,"title":"Why Bengaluru Is The Silicon Valley For AI Startups In India","content":"Image source: medium.com\/cityai\/bangalore As a destination for development and innovation, Bengaluru has welcomed technology and talents with open arms. Today, the city houses India headquarters of all the major tech giants and has turned itself into a tech powerhouse. Nicknamed the Silicon Valley of India, Bengaluru has become a favourite destination for startups to thrive and make a mark. According to a study by NASSCOM and Zinnov, India is ranked the 3rd biggest startup hub in the world, with Bengaluru housing a substantial number of these startups. The city currently houses 7,200 – 7,700 registered startups and a total of 1,200 tech startups were added just in 2018. The study further highlighted that advanced technology startups grew by almost 50 per cent from 2017 and  AI, data analytics and IoT startups have witnessed the highest adoption rate. Another positive trend associated with the booming startup bubble has been the overall job creation across the country. As per the study, an estimated 40,000 direct jobs were created in 2018 alone. Speaking about the city’s reputation for being one of the finest destinations for startup investment, the state CM, before the launch of Bengaluru Tech Summit said, “Bengaluru has emerged as one of the global innovation hubs in the league of Tokyo in Japan, Silicon Valley in the US and Tel Aviv in Israel. The summit will provide a platform for knowledge sharing on emerging technologies like Artificial Intelligence, Robotics and Blockchain.” Key growth indicators of AI startups in Bengaluru According to a 2019 study, by AIM which tracked the funding of AI and analytics startup in India,  the highest increase in year-on-year analytics revenues for an Indian city came from Bengaluru, from $539 million in 2017 to $ 739 million this year; it witnessed an increase of 37%. The study also points out that there is a lot of optimism towards AI- driven solutions like automation frameworks for data integration, dashboards for visualisation and energy and automation specific applications being the most funded. Bengaluru eclipsed other Indian cities by a staggering margin, grossing $133 in aggregate funding. India’s Silicon Valley was followed by Chennai at $35 million and Gurugram at $29.5 million. Big ticket funding was led by US-Bangalore startups that included RPA player Automation Anywhere ($raised 300 million from Softbank Vision Fund) and  enterprise AI specialist Noodle.ai (raised $35 million from Dell Technologies Capital and TPG Growth). Key growth indices Members of Automation Anywhere Multiple factors have played a key role in driving the growth of artificial intelligence and analytics market in India, especially in Bengaluru. We look at some of the key drivers which resulted in the growth of AI and analytics startups in the city. 1) Increased Investor Interest:  As per the AIM study, startups raised an all-time high capital, registering a 368% growth from 2017. In 2018, startups with operations in India and globally raised approximately USD$ 529.52 million in funding rounds. Added to this is a growing trend among successful startup entrepreneurs to invest in other startups as well as the entering of non-tech participants in the market through acquisitions. In April 2018, Mukesh Ambani owned Reliance invested $180 million into Embibe, the Bengaluru-based AI education platform. The Softbank led financing of Automation Anywhere is one of the biggest deal breakers in present times. 2) The strong presence of startup hubs and incubators: Realising the potential of startups in job creation and bring out new development thus affecting the economy at large, corporates, educational institutions and even government has played a vital role mentoring and grooming young startups. Even traditional non-players such as Target and YES Bank have stepped into the analytics startup market to prepare startups through their accelerator programmes. 3) Presence of educational institutions: Educational institutions such as IITs and IIMs have also played a vital role in identifying young entrepreneurs and nurturing their startup dreams. Through world-class R&D infrastructure, impressive alumni network and through monetary support in the form of seed and early-stage funding, key educational institutes in India are contributing extensively to the startup boom. 4) Aspirations of becoming a global company: With the global market for AI and ML looking very promising, AI startups are looking market and product expansion into new markets. In 2018, California and India based Automation Anywhere bagged the biggest cheque of $300 million from SoftBank Vision Fund. As compared to 2017, where the startups received an aggregate investment of $113 million, funding rose to a record 368% in 2018.","excerpt":"As a destination for development and innovation, Bengaluru has welcomed technology and talents with open arms. Today, the city houses India headquarters of all the major tech giants and has turned itself into a tech powerhouse. Nicknamed the Silicon Valley of India, Bengaluru has become a favourite destination for startups to thrive and make a […]","categories":["AI Startups"],"tags":["accelerator program india","Automation Anywhere","investors","Startups"],"author_name":"Akshaya Asokan","publish_date":"2019-02-26T12:36:44","publication_year":"2019","word_count":743,"keywords":["Go","API","artificial intelligence","AI","investors","ML","accelerator program india","Automation Anywhere","Aim","ViT","analytics","CLIP","Startups","R"],"extracted_tech_keywords":["AI","artificial intelligence","ML","analytics","Aim","R","Go","API","CLIP","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-startups\/why-bengaluru-is-the-silicon-valley-for-ai-startups-in-india\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10093583,"title":"KPI Partners Attains Microsoft Azure Solution Partner Status in Data &#038; AI","content":"KPI Partners, a leading global provider of Analytics and Digital Transformation solutions, announced today its achievement of Microsoft Azure Solution Partner status in Data & AI. With expertise in leveraging Azure Synapse Analytics, Azure Data Lake, Azure Data Factory, and Azure Databricks, KPI Partners excels in designing and implementing customized Microsoft Analytics solutions. The company’s proficiency enables the construction of robust analytics and AI solutions tailored to meet the unique requirements of each client, offering streamlined data management processes. Recognizing KPI Partners’ competence in workload analysis, schema modeling, and ETL operations for data migration to cloud-based data warehouses, Microsoft acknowledges their designation as a Solution Partner. Through their Azure Center of Excellence, KPI Partners has strengthened their data management and governance capabilities, facilitating the development of cutting-edge analytics using advanced technologies such as Generative AI, ChatGPT, and AI solutions on the Azure Cloud Platform. This new capability empowers businesses to transform data into a competitive advantage, driving accelerated ROI and value realization while minimizing implementation costs. Sid Goel, CTO & Partner at KPI Partners, emphasized, “Becoming a Solutions Partner for Data and AI (Azure) is a testament to KPI Partners’ unwavering commitment to technical expertise and performance. Our solutions are designed to drive customer success, positioning us as the trusted partner of choice for those seeking top-notch Data and AI solutions. Our expertise and innovation have enabled us to provide unparalleled support to global customers across various sectors, including high technology, manufacturing, healthcare, financial services, energy, and others. By partnering with us, clients can align their long-term strategic goals for greater profitability and growth.”","excerpt":"This new capability empowers businesses to transform data into a competitive advantage, driving accelerated ROI and value realization while minimizing implementation costs.","categories":["AI News"],"tags":["data and ai","Microsoft Azure","partnership"],"author_name":"AIM Media House","publish_date":"2023-05-18T18:07:45","publication_year":"2023","word_count":264,"keywords":["partnership","ChatGPT","Go","AI","R","ML","RAG","analytics","generative AI","Microsoft Azure","Azure","data and ai","Databricks"],"extracted_tech_keywords":["AI","ML","analytics","generative AI","ChatGPT","RAG","Azure","Databricks","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/kpi-partners-attains-microsoft-azure-solution-partner-status-in-data-ai\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10134551,"title":"Anthropic’s Claude Could be Amazon’s Last Hope to Save Alexa","content":"We all know about Amazon’s vested interest in Anthropic or rather a mutual interest. The AI company has not only received $4 billion from the ecommerce giant to advance generative AI, but every iteration and releases of Claude models are available on Amazon Bedrock. Now, reports claim that Amazon’s new and improved Alexa will be powered by Anthropic’s Claude. This makes one wonder why Amazon would bet big on Anthropic’s AI model rather than their own. The Best Partnership Ever Source: X Amazon’s billion-dollar investment in Anthropic gave the big-tech company a minority stake in the promising AI startup, which turned out to be a huge advantage. The collaboration includes not only financial investment but also a focus on technological development, particularly in AI chip optimisation. Amazon had been extensively working on bringing generative AI features into Alexa for quite some time now. A few months ago, it was reported that the company was planning to put an additional price tag of $5 to $10 for accessing the new voice assistant, which was called ‘Remarkable Alexa’. However, the output was not as per Amazon’s expectations. The initial trial runs for Alexa with Amazon’s AI features, struggled with latency. The AI assistant faced difficulties in replying and took six to seven seconds to understand a prompt. Amazon, then obviously did the next best thing: resort to a superior AI chatbot they have some form of control over. While the integration sounds promising, the additional price for accessing Alexa seems a bit odd. A user on X asked how many would be willing to pay $10\/mo to talk to Alexa when the same user can access it over their phone. Source: X Saving Alexa The fervour with which Alexa-powered Amazon devices such as Echo were released, has slowly fizzled out. With competition from other players, and even a reported dip in the quality of the devices, Alexa was losing its charm. In 2022 alone, it was reported that Alexa was on track to lose $10 billion. In addition, there were significant layoffs in the Alexa team. With such a form of development, Amazon’s interest in making things right is probably more desperate. Superior Claude Without a doubt, Anthropic has been on a roll of late. The company’s latest model Claude 3.5 Sonnet has been a powerful tool for customers and CTOs alike. The model is believed to help developers build and ship products that took weeks to now deliver in no time. Recently, the company introduced the Claude Enterprise plan, providing organisations with advanced tools for secure collaboration using their internal knowledge. The plan features a 500K context window, greatly enhancing Claude’s capacity to manage extensive datasets like sales transcripts, long documents, and codebases. Meanwhile, Claude Artifacts, released last week, has helped designers, developers, and executives to transform ideas into reality through simple, visually appealing interfaces. All these features showcase Claude’s superior capabilities that can be easily leveraged for Amazon’s voice assistant. Big Tech’s Voice Game Amazon’s strategic move to leverage an AI startup’s model for its product is not new in the tech ecosystem, especially not for voice products. If you recall, Apple had made a similar announcement at their recent developer conference WWDC 2024. Apple announced that Siri will now be able to leverage ChatGPT’s expertise when necessary. This enhancement was part of Apple’s efforts to improve Siri’s ability to understand and respond to users, making it more intelligent and intuitive. While Anthropic makes its way through strategic partnerships with big tech companies, the Claude model is not free from controversies. Recently, Anthropic was found to inject additional text into prompts submitted through their API to prevent copyright issues, a practice discovered by users experimenting with the system. Surprisingly, the additional text inserted by Anthropic is not visible to the user, which could potentially disrupt the expected output of the prompts, especially in unusual or edge cases. “Unless safety guardrails are very well done – they have the capacity to drive developers to open source, where they have better transparency into model behaviour,” said Campbell Hutcheson, chief product officer at Norm AI.","excerpt":"If OpenAI’s ChatGPT can likely save Apple’s Siri, then why can’t Anthropic’s Claude save Amazon’s Alexa?","categories":["AI Features"],"tags":["AI","Alexa","Amazon","Anthropic","Apple","Claude","Siri"],"author_name":"Vandana Nair","publish_date":"2024-09-06T12:23:40","publication_year":"2024","word_count":682,"keywords":["Anthropic","ChatGPT","Go","TPU","AI","Apple","Claude","Amazon","RAG","Aim","generative AI","Siri","Alexa","Claude 3.5","R"],"extracted_tech_keywords":["AI","generative AI","ChatGPT","Claude 3.5","Anthropic","Aim","RAG","TPU","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/anthropics-claude-could-be-amazons-last-hope-to-save-alexa\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10093818,"title":"Uncensored Models are Double-edged Swords That Need to be Unleashed","content":"Recently, a model called WizardLM-7B-Uncensored LLM was released on Hugging Face by a creator named Eric Hartford, who works with Microsoft. The model gained prominence for its improved intelligence and creativity as it removed censorship from its training data. But this fanned a bigger discussion around AI safety. An individual named Michael de Gans started harassing and threatening the creator on the Hugging Face platform and attempted to have the creator fired from Microsoft. He also demanded the removal of his model from the platform. The open source platform has responded to his complaints and promised internal escalation to address the issue. While the debate continues, the creator has garnered a huge support from the community that espoused the need of such uncensored models. After the release of WizardLM-7B-Uncensored, Eric also announced WizardLM-30B-Uncensored. He mentioned that a 65B version is also in the works, thanks to a generous GPU sponsor. However, he clarified that they do not handle quantised or GGML versions themselves but expect them to be available soon. In addition, Eric also released a blog explaining the reasons and the process of working with the WizardLM model. It involves addressing dataset filter refusals and biases, fine-tuning the model, and releasing it. Eric rewrote a script originally designed for the Vicuna model to suit the WizardLM dataset. Running this script on the WizardLM dataset generates a new dataset called “ehartford\/WizardLM_alpaca_evol_instruct_70k_unfiltered”. Eric recommends using a compute provider like Azure and suggests having ample storage, preferably 1TB to 2TB, to avoid running out during the process. He also provides guidance on setting up the workspace, downloading the created dataset, and obtaining the base model, LLaMA-7b. Why the need for uncensored models? Uncensored models refer to models that lack the alignment—which ensures that models avoid providing answers to controversial or dangerous questions. However, for popular LLMs like OpenAIs GPT model, Google’s PaLM, or Meta’s LLaMA alignment is based on American popular culture, American law, and a liberal and progressive bias. Uncensored models, on the other hand, are not restricted by such alignment and allow for a broader range of use cases and perspectives. There are a multitude of reasons why uncensored models should exist because different cultures, factions, and interest groups deserve models that cater to their specific needs. Open source AI should promote composable alignment, allowing users to choose the alignment that suits them rather than imposing a single perspective. Users should have ownership and control over the models they use on their computers, without the models imposing their own limitations. Composability is essential in building aligned models. Starting with an unaligned base model allows for the development of specific alignments on top of it. The existence of uncensored models contributes to the diversity, freedom, and composable nature of the open-source AI community. Uncensored models could have several unique use cases such as writing novels with evil characters, engaging in roleplay, or pursuing intellectual curiosity. So in a way, they could also pose a threat to open source aligned models, because of the wide array of additional use. While there are arguments for and against uncensored models, those who reject their existence entirely may lack nuance and complexity in their perspectives. Embracing uncensored models is crucial for scientific exploration, freedom of expression, composability, storytelling, and even humor. Exploring uncensored models To create uncensored instruct-tuned AI models, it is important to understand the technical aspects of alignment. Open source AI models are trained from a base model and fine-tuned with an instruction dataset obtained from the ChatGPT API, which has alignment built into it. The instruction dataset contains questions and answers, including refusals where the AI avoids providing certain information. These refusals contribute to the alignment of the models. To gain unrestricted control over AI chatbots like ChatGPT, numerous users have been exploring and have also attempted jailbreaks. Jailbreaking is the process of removing software restrictions that are either illegal or go against the terms of service of a device or operating system. The idea of jailbreaking LLMs like ChatGPT derives inspiration from iPhone jailbreaking, which allows iPhone users to bypass iOS limitations. In the realm of artificial intelligence, safety is a major concern not only for ChatGPT but also for other bots like Bing Chat and Bard AI. Sam Altman, the CEO of OpenAI, has expressed the company’s desire to grant users significant control over ChatGPT, allowing them to make the model behave according to their preferences. However, there are both pros and cons associated with ChatGPT jailbreaking, and they need to be carefully considered. Eric has also emphasised that users are responsible for how they utilize the model, comparing it to tools such as knives, lighters, or cars. The issue with Eric’s model Wizard has helped intensify the debate against enforcing compulsory safety standards for all models hosted on Hugging Face and others, fearing that it would render the platform ineffective. The community has expressed worries that by deleting such threads or not supporting uncensored models may discourage creators, and they may stop sharing their work altogether. There are also concerns over Reddit’s moderation of content. Conclusively, uncensored models provide a necessary alternative to aligned models by allowing for a wider range of perspectives, use cases, and cultural representations. They promote freedom, composability, and individual choice within the open-source AI community—and open source platforms like Hugging Face, GitHub should provide a platform for such models. However the challenge lies in determining the absolute rules and setting limits on customised outputs from uncensored models. What you can be certain of is that the subject of AI speech is anticipated to gain greater significance.","excerpt":"Embracing uncensored models is crucial for scientific exploration, freedom of expression, diversity, storytelling, and composable nature of the open-source AI community","categories":["AI Features"],"tags":[],"author_name":"Shyam Nandan Upadhyay","publish_date":"2023-05-24T12:46:57","publication_year":"2023","word_count":932,"keywords":["ChatGPT","Hugging Face","artificial intelligence","OpenAI","AI","chatbots","ML","RAG","Ray","Azure"],"extracted_tech_keywords":["AI","artificial intelligence","ML","ChatGPT","OpenAI","Ray","Hugging Face","RAG","chatbots","Azure"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/uncensored-models-are-two-edged-swords-that-need-to-be-unleashed\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10105046,"title":"Ola Unveils ChatGPT Clone, Calls it Krutrim","content":"Ola’s chief Bhavish Aggarwal recently unveiled Krutrim (which means artificial in Sanskrit). This has also been touted as “India’s first full-stack AI” solution. At first glance, the platform has a stark resemblance to ChatGPT — at least the UI\/UX bit of the platform — but only greenish. Aggarwal claimed that Krutrim AI is better than GPT-4 in various Indic languages. He said it is trained on 2 trillion tokens and can understand over 20 Indian languages and generate content in about 10 languages, including Marathi, Hindi, Bengali, Tamil, Kannada, Telugu, Odia, Gujarati, and Malayalam. “Today, all AI models, called LLM, are trained largely in English, but language is not just text. Language is also the vehicle for cultural values, context and ethos, and current AI models just can’t capture India’s culture, knowledge, and aspirations given our multicultural, multilingual heritage,” said Aggarwal. Is Krutrim AI Even Real? Confident Aggarwal said, “This is not just a wrapper on some existing API. This is not just a little bit of fine-tuning done, which that’s what people in the AI industry call about taking an existing model and putting a little bit more of a data set into it.” Some of the existing API includes GPT-4, Llama 2, Gemini and others. “This is the foundational work starting from the science layer, changing the math and the algorithms of the models to make it more relevant for Indian languages,” explained Aggarwal. During the Krutrim launch event, Aggarwal even took a swipe at an American AI company (not sure if it is OpenAI), and said “We heard them talk about building AI for humanity and doing it democratically.” It’s commendable that Ola has created India’s first generative AI model from scratch, but the likelihood of this achievement seems quite rare. Even other major players, not just in the Indian market but also outside, took more time and resources to build their generative AI models from scratch. Ola chief claims that following Krutrim, they will release its first multimodal Krutrim Pro by next quarter. He said for Krutrim Pro, “All the dimensionalities and modalities will be taken as inputs together, and the algorithm will be able to train across modalities. This is different from a separate text model, a distinct speech model, and a separate vision model being trained in parallel. Interestingly, OpenAI announced a multimodal GPT-4 back in March, but in reality, GPT-Vision came into existence only by September — i.e. over six months. It is intriguing how Krutrim managed to develop an LLM in just three months, especially when compared to its Indian counterparts such as Corover AI, Sarvam AI, and Kissan AI, which took considerably longer. Notably, these counterparts constructed their models on top of Llama 2 or GPT-4, which should have ideally taken less time as compared to Krutrim. Moreover, it took Elon Musk four months to build Grok and xAI is grappling with GPU shortages, a concern acknowledged by Oracle in its latest earnings call. Surprisingly, Aggarwal didn’t disclose the number of GPUs acquired for training their model throughout the entire event. Instead, it said, it is building all of these capabilities in-house. Moreover, Aggarwal didn’t disclose the dataset on which Krutrim AI is trained. Training a model on multiple languages is no easy feat, especially considering the increasing cost. Languages, particularly those with intricate structures and scripts like Hindi, Kannada, or Telugu, demand varying numbers of tokens. English, with its simplicity, requires fewer tokens in comparison. The financial implications of tokenization disparities are even more pronounced. The cost of training and using AI models hinges on token counts, compute and cloud costs. Krutrim had raised $24 million in debt from Matrix Partners, which is also a shareholder of Ola Electric. However, Aggarwal clarified earlier that Krutrim is a separate entity from Ola. Regardless of everything else, Ola has not yet released a research paper or details of the dataset trained on, and the team leading this initiative, are still confused if they should call themselves Ola or Krutrim. All of this looks like a marketing gimmick in an attempt to secure additional funds. That too, without even releasing a solid product into the market, and the demo, hardly looked convincing.","excerpt":"Krutrim AI is better than GPT-4 in various Indic languages","categories":["AI Features"],"tags":["Ola","Ola Krutrim"],"author_name":"Siddharth Jindal","publish_date":"2023-12-16T18:17:47","publication_year":"2023","word_count":699,"keywords":["Go","ChatGPT","Ola","API","OpenAI","AI","R","GPT","Aim","generative AI","Ola Krutrim","xAI"],"extracted_tech_keywords":["AI","generative AI","ChatGPT","OpenAI","xAI","Aim","R","Go","API","GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/ola-unveils-chatgpt-clone-calls-it-krutrim\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":23961,"title":"What Is Talk To Books And Semantris?","content":"Natural language processing has been developing through the years with the introduction of word vectors which give the ability for the AI to learn and understand sentences, words, phrases from any language. Vector Representation Of Language With the word to vector representation of the language in space, huge sentences and small paragraphs are represented in the form of a vector. Relation to other words, synonymy, antonymy, meronymy, holonymy, and many other types of relationships are all represented in vector space language models. Google’s new Talk to Books is a way to explore books with a help of a sentence and find the best match based on over 1,00,000 books in their collection. This is where Semantris comes into picture. It is a word association arcade powered by machine learning, where you submit words which are relevant to the highlighted words. Talk To Books With Talk to Books one can find different ways to explore literature. You type in a sentence or ask a question, and the model finds sentences in books that are the found in the books. It takes into the account the keyword matching metric to find relevant books. This way you can also get to know if the books interest you or not. The model is trained on millions of phrases and sentences, where it learns to figure out what is the best response based on the user input. Once the user inputs a question or sentence, the models find all the phrases among the 1,00,000 books to find the best matches and respond based on the query. The principle of the technique is finding the semantic meaning of the sentence. At this level, the models only work on a sentence level (a paragraph input is not available, like in Smart Reply for Gmail) but a good response from the model can be expected. One can also find books and passages that are not so relevant and the highlighted paragraph might not be an obvious one. You may also notice that being well-known does not make a book sort to the top. This experiment looks only at how well the individual sentences match up. However, one benefit of this is that the tool may help people discover unexpected authors and titles, and give a glimpse into books in a way that is fresh and innovative. What is Semantris? Google’s arcade game Semantris is based on word association application and powered by ML. A word or a phrase input is given based on the highlighted word and it then scores them on how well the model responds to what you have typed. Similarity, opposites and neighboring concepts are all fair-game using this semantic model. The time pressure in the Arcade version (shown below) will tempt you to enter single words as prompts. The Blocks version has no time pressure, which makes it a great place to try out entering in phrases and sentences. You may enjoy exploring how obscure you can be with your hints. These are only a few examples of how Google is working on experience and application designs using machine learning tools. Other applications relating to the classification, semantic similarity, semantic clustering, whitelist applications (selecting the right response from many alternatives), and semantic search (of which Talk To Books is an example) are currently being worked on. Smart Reply in Gmail Responding to emails is tedious job, but with Google’s Smart Reply one can reply to these emails effortlessly by opting one of the quick responses. They are predicted on the basis of the content in email itself. The feature already drives 12 percent of replies in the Inbox on mobile. And starting today, Smart Reply is coming to Gmail on web too. Smart Reply suggests three responses based on the email you received. Once you’ve selected one, you can send it immediately or edit your response starting with the Smart Reply text. Either way, you’re saving time. Credit: The Driod Lawyer Smart Reply makes use of the machine learning techniques to give the best responses based on the user emails. If you are more inclined to use “Sure, thanks” than “Sure”, it suggests the response based on the information gathered by the user previous habits. A new version of Smart Reply will roll out globally on Android and iOS in English first, followed by Spanish in the coming weeks. Conclusion As the technology is rising, natural language processing is reaching high limits of accuracy and deceiving humans. It is sometimes hard to predict the right phrase or a word, but with more and more users using the tech, the models will learn fast and shall be more accurate.","excerpt":"Natural language processing has been developing through the years with the introduction of word vectors which give the ability for the AI to learn and understand sentences, words, phrases from any language. Vector Representation Of Language With the word to vector representation of the language in space, huge sentences and small paragraphs are represented in […]","categories":["Deep Tech"],"tags":["Gmail","Google"],"author_name":"Kishan Maladkar","publish_date":"2018-04-24T12:57:26","publication_year":"2018","word_count":774,"keywords":["Go","semantic search","machine learning","Gmail","AI","programming_languages:R","ML","programming_languages:Go","RAG","Google","R"],"extracted_tech_keywords":["AI","machine learning","ML","RAG","semantic search","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/googles-semantic-evolution-what-is-talk-to-books-and-semantris\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":5592,"title":"The Best Business Analytics Software","content":"We’re here today to talk about the Magic quadrant for business intelligence and analytics platforms, which is published annually by Gartner. We hope to answer questions like – [highlight] Is SAS the best software in Analytics ? What is the standing of Microsoft when it comes to Analytics (Excel , Access , SQL) How good is Tableau in the eyes of the Analytics world? [\/highlight] (What is Gartner?  Gartner is the world’s leading IT research and advisory company which delivers technology related insights from CIOs and senior IT leaders in corporations to the general public .All of us use Gartner and its reports to come to conclusions about the various software suites we would like to buy.) Over the last decade, the world of business intelligence and analytics has undergone a change. As organisations have adopted business intelligence and started looking at the data for measurement and reporting purposes, the next level was to try and do prediction forecasting and optimisation on the same data. Because of this growing importance of advanced analytics, the organisation started looking at solutions which could deliver end to end services. Over the last few years we see software suits which have both BI and predictive analytics capabilities. Thus Analytics software as of today are expected to be able to provide services across 17 categories Reporting Dashboards. What is the difference between reporting and  dashboards ? It is a style of reporting that uses a lot of graphics to show performance measures Ad-hoc reports and queries Microsoft Office integration Mobile BI-  which enables organisations to develop and deliver content on mobile devices Interactive visualisation -which enables the exploration of data via the manipulation of chart images. This includes an add in of visualisation options that go beyond the normal pie chart, bargraph etc. and could include heat and tree maps scatter plots other special-purpose visuals etc. Search-based data discovery – this applies a search index to structured and unstructured data sources Geospatial and locational intelligence- which is specialised analytics and visualisations that provide a geography, spatial and a time context Embedded advanced analytics -which will enable users to levitate statistical functions library embedded in the BI server Online analytical processing or OLAP- this allows users to analyse data with fast query and calculation performance enabling slicing and dicing of data BI infrastructure and Administration -which enables all tools in the platform to use the same security , meta data , query engine etc. That platform should support multi tenancy. Meta data management Business user data modelling- which would be free drag-and-drop user driven data combinations Development tools Embeddable analytics -which would be tools including a software developer’s Kit with APIs for creating and modifying analytical content, visualisations etc. and embedding them into a business process Collaboration -which would enable users to share and discuss information analysis chat etc. Support for big data sources -which is the ability to support and query hybrid, columnar and array-based data sources, such as MapReduce and other NoSQL databases (source:-http:\/\/www.gartner.com\/technology\/reprints.do?id=1-1QLGACN&ct=140210&st=sb ) [highlight]What does Gartner say about SAS ? I am sure this is the question that’s in your mind . Since I started my Analytics journey with SAS , it is the most important question in my mind .[\/highlight] 1.      SAS SAS’s analytics portfolio spans platforms for BI, performance management, data warehousing, in-memory databases, data integration, data quality, decision management, and content and social analytics, with a core strength being advanced analytics. SAS also offers industry- and domain-specific analytic applications built on its product portfolio. Strengths SAS’s analytics portfolio spans platforms for BI, performance management, data warehousing, in-memory databases, data integration, data quality, and content and social analytics. However, unlike most other BI platform vendors, SAS’s core strength is in its advanced analytical techniques, such as data mining, predictive modeling, simulation and optimization, for which it is acknowledged as a Leader in “Magic Quadrant for Advanced Analytics Platforms.” Historically, SAS tools have been used primarily by power users, data scientists and IT-centric BI developers. That this remains the case is shown by SAS customers placing it above any other vendor in the Magic Quadrant survey in terms of support for complex types of analysis, but giving it one of the lowest scores for ease of use. During the past five years, SAS has been investing heavily in revamping its user experience to change this situation and encourage more mainstream user adoption. This aggressive and unmatched strategy is part of the reason for SAS’s favorable Completeness of Vision position. SAS also differentiates itself from most other BI platform vendors by productizing and selling industry- and domain-specific advanced analytic applications that are focused on specific business problems and built using many of its technology platform products. This enables it to sell “value,” rather than components. Data access and integration and the ability to support large volumes of data are the main reasons customers choose SAS, according to the survey. In fact, while SAS deployments support a below-average number of users, its data volumes are among the highest in the survey. Cautions SAS customers consider its software among the most difficult to use and most difficult to implement, with ease of use for business users being identified as a limitation on broader deployment by a higher percentage of customers than for any other vendor. Although SAS has exploited its core competency in predictive analytics to encapsulate and automate advanced analysis for business-user-oriented guided data discovery in Visual Analytics, it will face competition from data discovery and BI and analytics platform vendors with leading-edge data discovery capabilities, such as Tableau and Tibco. Despite SAS’s success as a Leader in the predictive analytics space, the company still faces a challenge to make it onto BI platform shortlists, unless customers already use its other advanced analytic capabilities and require integration and leverage of skills. SAS’s reference customers rated functionality used in traditional BI areas (reporting, dashboards, OLAP, interactive visualization and so on) lower than for most of the other BI Leaders [highlight] Microsoft Excel 2010 with its analysis toolpak has been quoted by me as being one of the most easy-to-use analytics tools. Let’s see what Gartner says about Microsoft. [\/highlight] 2.      Microsoft Microsoft offers a competitive and expanding set of BI and analytics capabilities, packaging and pricing that appeal to Microsoft developers, independent distributors and now to business users. It does so through a combination of enhanced BI and data discovery capabilities in Office (Excel) 2013, data management capabilities in SQL Server, and collaboration, content, and user and usage management capabilities in SharePoint. Strengths Of the megavendors, Microsoft has made the most progress toward delivering a combination of business user capabilities with an enterprise-capable platform. Microsoft delivered business-user-oriented data discovery and other BI capabilities in Excel 2013. Microsoft has made early investments in its cloud-based BI offering, Power BI. Microsoft’s strategy is to use the cloud to increase adoption of its new and most competitive BI capabilities in Excel (starting with Excel 2013), and to accelerate enhancements to Excel to every six months. n the customer survey conducted for this Magic Quadrant, more Microsoft customers cited TCO and license cost as their main reasons for selecting Microsoft as a BI vendor than did those of most of the other vendors. Microsoft composite product score was above average across the 17 capabilities, when weighted for use. Microsoft customers appreciate its strong BI infrastructure and development tools. They also rated its reporting, ad hoc query, Microsoft Office integration, business user data mashup, embedded BI, collaboration, search BI and OLAP capabilities higher than the survey average. Importantly, unlike those of its megavendor competitors, Microsoft’s customers rated customer experience (support and product quality) above the survey average. Cautions Although Microsoft’s functional ratings have improved and it can offer a wide range of functions, it also has one of the highest percentages of users who say that absent or weak functionality is among the main reasons limiting broader deployment of its software. Mobile BI, interactive visualization and metadata management remain product weaknesses reported by customers. Multiproduct complexity remains a challenge, now primarily for on-premises and hybrid deployments. Because Microsoft’s BI platform capabilities span three different tools (Office, SQL Server and SharePoint) that also perform non-BI functions, the task of integrating components and building applications is left mainly to the customer. Although Microsoft’s partner-driven sales model drives global growth for the company, Gartner’s inquiries suggest that this approach often makes it difficult for customers to find their Microsoft sales representative. This causes frustration . [highlight][\/highlight]I increasingly see Tableau becoming an integral part of the reporting structure in most KPOs. Let’s see what the Gartner’s report has to say about this software. 3.  Tableau Tableau’s highly intuitive, visual-based data discovery, dashboarding, and data mashup capabilities have transformed business users’ expectations about what they can discover in data and share without extensive skills or training with a BI platform. Strengths Tableau offers an intuitive, visual-based interactive data exploration experience that customers rate highly .Its core differentiator — making a range of types of analysis (from simple to complex) accessible and easy for the ordinary business user, whom Tableau effectively transforms into a “data superhero. Tableau has a focused vision with an evolutionary road map for enabling users to meet enterprise requirements for reusability, scalability and embeddability. Tableau’s strong survey results for customer satisfaction, coupled with its market momentum, are behind its dominant Ability to Execute position. Tableau provides purpose-built, business-oriented data mashup capabilities with data connectors that use Tableau’s VizQL technology. Direct query access has been a strength of the platform since the product’s inception. Tableau offers a broad range of support for direct-query SQL and MDX data sources, as well as a number of Hadoop distributions, native support for Google BigQuery, and support for search-based data discovery platforms, such as Attivio. Cautions Although Tableau’s average user count continues to grow and was above the market average in this year’s customer survey, its products are often used to complement an existing BI platform standard; only 42% of its customers considered it as their BI standard.Ttraditional BI platform vendors with substantial installed-base market shares but lacking in growth momentum, including IBM, Microsoft, MicroStrategy, SAP and SAS, are aggressively investing in their own data discovery capabilities to reverse the trend. Tableau’s customers report a below-average sales experience, which includes the entire sales life cycle from presales activities to contracting, pricing and the ongoing sales relationship. Tableau continues to expand its international presence, but the majority of its customers are likely to be large (often international) companies located in North America. Tableau has opened sales offices in Europe and Asia (for example, Singapore) and introduced support in Asia; it also plans further global sales expansion with live, time-zone-appropriate support in local languages.","excerpt":"We’re here today to talk about the Magic quadrant for business intelligence and analytics platforms, which is published annually by Gartner. We hope to answer questions like –   [highlight] Is SAS the best software in Analytics ? What is the standing of Microsoft when it comes to Analytics (Excel , Access , SQL) How […]","categories":["IT Services"],"tags":[],"author_name":"Subhashini Tripathi","publish_date":"2014-04-10T15:52:49","publication_year":"2014","word_count":1786,"keywords":["Go","AI","R","Scala","RAG","Ray","analytics","SQL","Rust","predictive analytics"],"extracted_tech_keywords":["AI","analytics","Ray","RAG","predictive analytics","R","SQL","Go","Rust","Scala"],"url":"https:\/\/analyticsindiamag.com\/it-services\/the-best-business-analytics-software-2014\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10071153,"title":"Tech Layoffs are accelerating, here are the recent big ones","content":"In response to the rapidly changing economic situation in the past few months, Kristalina Georgieva, Managing Director, IMF, addressed the ‘darkening’ global economic outlook and indicated that there is an increasing risk of recession in the near future. In such turbulent times, companies around the world have been taking cost-cutting measures but mostly laying off employees. Even big techs, often considered safe regarding job security, have not shied away from such measures—announcing frequent layoffs or a hiring freeze. Tesla Recently, it was reported that Tesla laid off 229 annotation employees from its Autopilot team and shut down one of its US offices. Reports say that a regulatory filing in the state of California revealed that Tesla laid-off workers from its San Mateo office that employed only 276 workers. Further, the remaining 47 employees are expected to be transferred to Tesla’s ‘Buffalo Autopilot’ office. These employees are believed to be handling moderately low-skilled, low-wage profiles such as Autopilot data labelling. In June 2022, media reports claimed that Tesla’s CEO, Elon Musk had a “super bad feeling” about the economy and was determined to cut about 10 per cent of the company’s salaried staff. According to email exchanges, he sent out a message to Tesla’s executives that laid out his concerns and further instructions to “pause all hiring worldwide”. In another email, Musk is believed to have instructed Tesla employees to report to offices, a shift from their work-from-home setup. Netflix In June 2022, Variety reported that internet streaming giant, ‘Netflix’ planned to lay off 300 employees. These job cuts are expected to be across multiple business functions in the company, with the majority of cuts expected in the U.S. This is the second round of layoffs for Netflix, following deductions of similar size in the workforce in May 2022. In the first round, Netflix laid off 150 employees, mostly in the U.S.. These cuts happened across departments and represented roughly 2% of Netflix’s U.S. workforce. In the latest round of layoffs, the company also indicated that similar plans may be announced again in the latter half of 2022. Twitter Micro-blogging platform ‘Twitter’ laid off  30% of its talent acquisition team in July 2022 following a company-wide hiring freeze announced in May 2022, according to TechCrunch. A spokesperson for the platform explained that the laid-off employees will receive severance packages and the remaining recruitment staff is expected to be reprioritised owing to frozen hiring. Twitter refused to comment on the number of employees affected by the layoffs. Microsoft In July 2022, Microsoft also joined the growing list of tech firms in cost-cutting by way of layoffs. According to Bloomberg, Microsoft is expected to proceed with a small percentage of role eliminations to accommodate structural changes in the organisation. However, the firm plans to continue investing in its business and grow the overall headcount in the year ahead. The layoffs are reported to have impacted less than 1% of the total workforce at Microsoft, spread across different groups such as consulting and customer and partner solutions. Microsoft has reportedly slowed down hiring in the Windows, Teams and Office groups. Google Recent reports claim that ‘Google’ CEO, Sundar Pichai has announced slow hiring for 2022 and 2023 while adding that the company aims to keep a check on investments for the time-being. Although Google announced no plans to freeze hiring, the current economic outlook played a significant role in this decision. Google believes that these challenges are not obstacles but opportunities to deepen the focus and invest for the long term. Meta ‘Meta’ has reportedly instructed its team managers to weed out poor performers considering the company’s recent controversy and struggles with advertising business.","excerpt":"Tesla laid off 229 annotation employees from its Autopilot team and shut down one of its US offices","categories":["IT Services"],"tags":["Big Tech Layoffs"],"author_name":"Sreejani Bhattacharyya","publish_date":"2022-07-18T13:00:00","publication_year":"2022","word_count":610,"keywords":["Go","API","programming_languages:R","AI","programming_languages:Go","Aim","Big Tech Layoffs","GAN","R"],"extracted_tech_keywords":["AI","Aim","R","Go","API","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/tech-layoffs-are-accelerating-here-are-the-recent-big-ones\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10009933,"title":"How To Choose The Best Machine Learning Algorithm?","content":"How do you know what machine learning algorithm to choose for your problem? Why don’t we try all the machine learning algorithms or some of the algorithms which we consider will give good accuracy. If we apply each and every algorithm it will take a lot of time. So, it is better to apply a technique to identify the algorithm that can be used. Choosing the right algorithm is linked up with the problem statement. It can save both money and time. So, it is important to know what type of problem we are dealing with. In this article, we will be discussing the key techniques that can be used to choose the right machine algorithm in a particular work. Through this article, we will discuss how we can decide to use which machine learning model using the plotting of dataset properties. We will also discuss how the size of the dataset can be a considerable measure in choosing a machine learning algorithm. Getting the first Dataset The dataset is taken from Kaggle, you can find it here. It has information about the diabetic patient and whether or not each patient will have an onset of diabetes. It has 9 columns and 767 rows. Rows and columns represent patient numbers and details. Techniques to choose the right machine learning algorithm 1. Visualization of Data Practical Implication: First of all, we will import the required libraries. #Import Libraries import pandas as pd import numpy as np import matplotlib.pyplot as plt import seaborn as sb After it we will proceed by reading the csv file. df = pd.read_csv(\"diabetes.csv\") df.head(5) Pair Plot Method By applying the pair plot we will be able to understand which algorithm to choose. #PairPlot to choose right algorithm sb.pairplot(data=df[['Glucose' ,'BloodPressure','SkinThickness', 'Outcome']], hue='Outcome', dropna=True, height=3) From the plot, we can see that there is a lot of overlap between the data points.KNN should be preferred as it works on the principle of Euclidean distance. In case KNN is not performing as per the expectation then we can use the Decision Tree or Random Forest algorithm. A decision tree or Random Forest works on the principle of non-linear classification. We can use it if some of the data points are overlapping with each other. Many algorithms work on the assumption that classes can be separated by a straight line. In such cases, Logistic regression or Support Vector Machine should be preferred. It easily separates the data points by drawing a line that divides the target class. Linear regression algorithms assume that data trends follow a straight line. These algorithms perform well for the present case. 2. Size of Training Data & Training Time Import the various algorithm classifiers to check the training time of small and large dataset. #Import Sklearn Libraries from sklearn.tree import DecisionTreeClassifier # Import Decision Tree Classifier from sklearn.model_selection import train_test_split # Import train_test_split function from sklearn import metrics #Import scikit-learn metrics module for accuracy calculation #Store independent and dependent variable feature = ['Pregnencies', 'Glucose', 'BloodPressure', 'SkinThickness','Insulin','BMI','DiabetesPedigreeFunction','Age'] X = df[feature] # Features y = df[\"Outcome\"] Split the data into train and test. Now we can proceed by applying Decision Tree, Logistic Regression, Random Forest and Support Vector Machine algorithms to check the training time for a classification problem. #Train-Test Split X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.3, random_state=1) Now, we will fit several machine learning models on this dataset and check the training time taken by these models. Decision Tree # Create Decision Tree classifier object import time clf = DecisionTreeClassifier() # Train Decision Tree Classifier start = time.time() clf = clf.fit(X_train,y_train) stop = time.time() print(f\"Training time: {stop - start}s\") Logistic Regression #Import sklearn library from sklearn.linear_model import LogisticRegression import time clf = LogisticRegression(random_state = 0) start = time.time() clf.fit(X_train,y_train) stop = time.time() print(f\"Training time: {stop - start}s\") Random Forest #Create a RandomForestClassifier from sklearn.ensemble import RandomForestClassifier clf=RandomForestClassifier(n_estimators=100) start = time.time() #Train the model using the training sets y_pred=clf.predict(X_test) clf.fit(X_train,y_train) stop = time.time() print(f\"Training time: {stop - start}s\") Support Vector Machine # Support Vector Classifier from sklearn.svm import SVC clf = SVC(kernel='linear') start = time.time() # fitting x samples and y classes clf.fit(X_train,y_train) stop = time.time() print(f\"Training time: {stop - start}s\") From the above results, we can conclude that Decision Trees will take much less time than all algorithms for small dataset. Hence, it is recommended to use a low bias\/high variance classifier like a decision tree. Getting the Second Dataset The dataset is taken from Kaggle, you can find it here. It has information about credit card fraud that occurred in two days. Feature Class is a target variable and it takes 1 in case of fraud and 0 otherwise. It has 284807 rows and 31columns. #Read the csv file df = pd.read_csv(\"creditcard.csv\") df.head(5) X=df.iloc[:,0:-1] y=df.iloc[:,-1] #Train-Test Split X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.3, random_state=1) Now again, on this second dataset, we will fit the above machine learning models on this dataset and check the training time taken by these models. Decision Tree # Create Decision Tree classifier object import time clf = DecisionTreeClassifier() # Train Decision Tree Classifier start = time.time() clf = clf.fit(X_train,y_train) stop = time.time() print(f\"Training time: {stop - start}s\") Logistics Regression #Logistic Regression Classifier from sklearn.linear_model import LogisticRegression import time classifier = LogisticRegression(random_state = 0) start = time.time() classifier.fit(X_train,y_train) stop = time.time() print(f\"Training time: {stop - start}s\") Random Forest #Create a RandomForest Classifier from sklearn.ensemble import RandomForestClassifier clf=RandomForestClassifier(n_estimators=100) start = time.time() #Train the model using the training sets y_pred=clf.predict(X_test) clf.fit(X_train,y_train) stop = time.time() print(f\"Training time: {stop - start}s\") Support Vector Machine #Support Vector Classifier from sklearn.svm import SVC clf = SVC(kernel='linear') start = time.time() # fitting x samples and y classes clf.fit(X_train,y_train) stop = time.time() print(f\"Training time: {stop - start}s\") With the huge dataset size depth of Decision Tree grows, it implements multiple if-else statements which increase complexity and time. Both Random Forest and Xgboost use the Decision Tree algorithm which takes more time. The result shows Logistic regression outperforms others. Final Thoughts I have concluded my analysis in selecting the correct machine learning algorithm. Furthermore, it is always advisable to use two algorithms for addressing the problem statement. This could provide a good reference point for the audience.","excerpt":"In this article, we will be discussing the key techniques that can be used to choose the right machine algorithm in a particular work. Through this article, we will discuss how we can decide to use which machine learning model using the plotting of dataset properties.","categories":["Deep Tech"],"tags":["decision tree algorithm","logistic regression","Machine Learning","Randomforest","scikit learn","when to use support vector machine","XGBoost"],"author_name":"Ankit Das","publish_date":"2020-10-18T16:00:00","publication_year":"2020","word_count":1030,"keywords":["Go","scikit-learn","NumPy","machine learning","decision tree algorithm","Randomforest","AI","logistic regression","Machine Learning","when to use support vector machine","Seaborn","XGBoost","scikit learn","Matplotlib","R","Pandas"],"extracted_tech_keywords":["AI","machine learning","scikit-learn","XGBoost","Pandas","NumPy","Matplotlib","Seaborn","R","Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/how-to-choose-the-best-machine-learning-algorithm\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10008189,"title":"Gradient Descent &#8211; Everything You Need To Know With Implementation In Python","content":"While building a deep learning model there are a lot of different things we need to define. First building the model with input layers followed by different dense layers and at last the output layer. Also, when the model structure gets ready while we compile it using optimizer, loss function, and metric to measure the performance of the model. Loss functions are used to compute the error between the actual value and predicted value whereas optimizers are used to reduce these errors so the model performs better and gives good results. We have several different loss functions and optimizers that are used in different situations. These optimizers make use of optimization algorithms. Through this article, we will discuss more optimizers and the most commonly used optimizer gradient descent. We will explore how it works and will check its implementation in python. What we will learn from this article? What are Optimizers?What is Gradient Descent? How does it work?How to implement gradient descent in python? How to use it while compiling a deep learning model? What are Optimizers? Optimizers are the ones that are used to reduce the loss in the model or to reduce the error rate made by deep learning models. The less the error rate better will be the performance of the model. There are several different types of optimizers that are used while compiling the models. Some of them include gradient descent, stochastic gradient descent, adam, etc. All these are used to optimize the performance of the model. They are commonly defined after defining the model structure. Refer to the below code to understand more about defining these. model.compile(loss=’binary_crossentropy’, optimizer=’sgd’, metric=[accuracy]) What is Gradient Descent? How does it work? It is the most preferred optimizer that is used to optimize a deep learning model. It uses optimization algorithms to reduce the error and find the minimum values for a function. Gradient descent makes use of derivatives to reach the minima of a function. Also, there are steps that are taken to reach the minimum point which is set by defining the learning rate. It decides how many steps to take to reach the minima. If we define a big value to the learning rate we may exceed the minima of the function whereas if we define it to be very small then it would consume much time to reach the target. There can be chances that gradient descent will miss out on the target if the learning rate is very high. The role of derivatives in optimization algorithms is to decide whether to increase or decrease the weights resulting in increasing or decreasing the loss function or cost function. We cannot train a neural network without defining the optimizer and loss functions. They are the mandatory parameters that need to be set while compiling a deep learning model. How to implement Gradient Descent in python? Now we will see how gradient descent can be implemented in python. We will start by defining the required library first that would be used for numerical calculation and for plotting the graphs. Refer to the below code for the same. import numpy as np import matplotlib.pyplot as plt Now we will define a function f as a quadratic function and function to compute its gradient. Refer to the below code for the same. def function(x,a): f = a[2]*x*x + a[1]*x + a[0] return f def grad(x,a): g = 2*a[2]*x + a[1] return g Now we will plot this function before we compute its minima. Use the below code to do the same. x = np.array([-3,-2,-1,0,1,2,3,4,5,6]) a = np.array([-3, -2, 3]) f = funct(x,a) plt.scatter(x,f) plt.plot(x,f) plt.xlabel(‘X’) plt.ylabel(‘f(X)’) We have values on the X-axis and f(x) on the y-axis. Now let’s define how to use gradient descent to find the minimum. Use the below code for the same. We will first define the starting point, learning rate, and the parameter to stop it like iterations or if the value does not change then it should stop. x = 8 lr = 0.001 change = 1e-5 max_iteration = 500 We have defined X_series the variable to check how the value of x is getting changed. Then in the loop, we have defined the function f at any point(x, a) followed by computing its gradient and then getting the changed values of x which gets computed by subtracting the original value of x from the product of the learning rate and gradient. Then we will define the condition to stop the loop by making use of maximum iteration and change that was previously defined. At last, we are plotting the values. Refer to the below code for the same. series = [x] iterations = 1 while True: f = funct(x,a) g = grad(x,a) new_x = x - lr * g if np.sum(abs(new_x - x)) < change: break if iterations > max_iteration: break if iterations % (max_iteration\/10) == 0: plt.scatter(x, f, marker='*') plt.plot(x, f) plt.xlabel(‘X’) plt.ylabel(‘f(X)’) iterations += 1 x = new_x series = np.concatenate((series,[x])) Now let us see the minimum value of X after iterations. We will check this by printing the min value of the series we defined before. print(series.min()) Conclusion The article aimed to demonstrate how we compile a neural network by defining loss function and optimizers. In this article, we also discussed what gradient descent is and how it is used. At last, we did python implementation of gradient descent. Since we did a python implementation but we do not have to use this like this code. These optimizers are already defined in Keras. They can be directly imported and used like the way shown in 1 point. Different optimizers can be used while training a neural net and the performance also gets changed when you use different optimizers. Also, check this article where you can monitor the loss and accuracy while training a deep learning model. “Tensorboard Tutorial – Visualize the Model Performance During Training”","excerpt":"Through this article, we will discuss more optimizers and the most commonly used optimizer gradient descent. We will explore how it works and will check its implementation in python.","categories":["Deep Tech"],"tags":["Deep Learning","gradient descent","optimisation algorithms","optimizers","python visualize neural network"],"author_name":"Rohit Dwivedi","publish_date":"2020-09-23T17:00:00","publication_year":"2020","word_count":985,"keywords":["NumPy","TPU","Keras","AI","neural network","Python","Ray","Aim","optimisation algorithms","python visualize neural network","deep learning","Deep Learning","Matplotlib","optimizers","gradient descent"],"extracted_tech_keywords":["AI","deep learning","neural network","Aim","Ray","Keras","NumPy","Matplotlib","TPU","Python"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/gradient-descent-everything-you-need-to-know-with-implementation-in-python\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10063573,"title":"What&#8217;s new with CUDA? NVIDIA reveals","content":"CUDA, NVIDIA’s parallel computing platform for general computing, is the leading proprietary framework for GPGPU. Acronym for Compute Unified Device Architecture, CUDA is a software layer giving direct access to the GPU’s virtual instruction set and parallel computational elements to execute compute kernels. The platform has been widely used in bioinformatics, life sciences, computer vision, electrodynamics, computational chemistry, finance, medical imaging etc. At NVIDIA, Stephen Jones, Architect, CUDA, spoke about CUDA at the GTC Conference, 2022. Analytics India Magazine has highlighted the key points from the talk covering the language and toolkit additions, latest developments and future of CUDA. The Three Eras of Computing: First era: Single core programming During the first era, programming took a problem, breaking it down into a series of logical steps and writing a straight-line code that would run on a processor as fast as possible. It was highlighted by Dennard Scaling, who figured out that as you make transistors smaller, you can run them faster without increasing the power. But quantum mechanics defied this, and designers had to switch to Moore’s law. As a result, developers kept adding more transistors but couldn’t make it faster. Second era: Parallel programming The second era of programming realised all programs have to target multiple threads that broke down data into separate elements so they can be processed independently and at the same time, as data parallels. So the solution moved from a straight line code to task parallelism based on asynchronous execution. Eventually, this led to the introduction of the data centre for scale computing. Third era: Locality The third era, which is present in CUDA, focuses on localisation. The era of locality aware computing focuses on where subjects are placed. Parallel computing is about a hierarchy of frameworks, libraries and runtimes where each hierarchy level zooms into the system. This allows programmers to choose their level of scaling. Scaling is an addition of data parallelism and locality of data. At the backend: Hopper architecture The NVIDIA Hopper architecture is at the backend of the CUDA platform. It extends MIG capabilities by up to 7x over the previous generation by offering secure multitenant configurations in cloud environments across each GPU instance. In addition, it introduces features to improve asynchronous execution and allow an overlap of memory copies with computation while minimising synchronisation points. Hopper also mitigates the issues of long training periods for giant models while maintaining the performance of GPUs. The Hopper Architecture, NVIDIA CUDA features Image processing CUDA works in a grid system for image processing. As Stephen demonstrated, an image is fragmented into a grid of blocks, where each block is run on the GPU as a completely separate program. GPUs can run up to thousands of such blocks based on parallelism in the GPU. These blocks work independently on their fragment of the problem and consist of threads. Thread block cluster Building onto this architecture, the team has leveraged CUDA’s massive scale and added a new tier of hierarchy called the thread block cluster. Described as Block.2, it is essentially a block of blocks. Stephen called it a new way of thinking about blocks, where in addition to the individual blocks, we can also target localised subjects for enhanced performance. The Thread Block Cluster, NVIDIA Here, the blocks within a thread block cluster live in a GPC processing cluster, with the cluster representing a capital of the parallels with more performance. Adding a cluster to the execution hierarchy allows an application to take advantage of faster local synchronisation and faster memory sharing. Various concurrent threads will be working together in a cluster yet running concurrently. This also allows researchers to move in the direction of annotating kernels while targeting the size that makes sense for their application. Distributed shared memory The cluster has all the properties of a CUDA thread block but is bigger. Additionally, every block in a cluster can read and write the shared memory of every other block. Asynchronous one-sided memory copy The NVIDIA team has heavily invested in asynchronous data that allows programmers to pause a copy and later pick up the result – a huge step towards keeping all dependencies local. It also frees up the threads to do other work all day long. This is called the Split barrier that allows programmers to keep the threads waiting to finish and wait for data to arrive. Threads can now sleep waiting purely for data. These transactions, called ‘self-synchronising rights’, allow the sender and recipient to know that the data was ready without needing some handshake. This one-sided memory copy is seven times faster than normal communication since it is just a single write operation. It can also be used in local memory for the data to get faster and easier. TMA Unit in Hopper Stephen explained that all of this is made possible by the tensor memory accelerator unit inside the Hopper. It takes the original copy of async and makes it bi-directional, and work between clusters. It enables one-sided data transfers as well. For instance, in neighbour-affecting algorithms, the problem is non-local. Self-synchronisation transactions enable 7x faster halo exchange between blocks in the same cluster. The TMA is a self-contained data movement engine, a separate hardware unit inside the SM that runs independently of the threads. It can take over and handle all calculations, allowing even a single thread to initiate a copy of the entire shared memory. The transaction barrier can wait for the data to arrive without syncing with each other. The accelerator can work on five dimension data.","excerpt":"Building onto this architecture, the team has leveraged CUDA’s massive scale and added a new tier of hierarchy called the thread lock cluster.","categories":["Global Tech"],"tags":["cuda"],"author_name":"Avi Gopani","publish_date":"2022-03-25T13:00:00","publication_year":"2022","word_count":924,"keywords":["CUDA","Go","API","cuda","programming_languages:R","AI","computer vision","RAG","analytics","R"],"extracted_tech_keywords":["AI","computer vision","analytics","RAG","CUDA","R","Go","CUDA","API","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/whats-new-with-cuda-nvidia-reveals\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":66522,"title":"US Reverses Its Decision And Joins G7 AI Group; Invites India And Russia","content":"“It’s [G7] a very outdated group of countries. We want Australia, we want India, we want South Korea.” -Donald Trump on G7 The U.S. last week has announced that it will be joining the group, called the Global Partnership on AI, a decision that was reversed from the stance they have taken previously. This group, according to Michael Kratsios, the CTO of the U.S., will study and provide recommendations of AI technologies that respect privacy and civil liberties. Iterating on why they had to reverse their decision, Kratsios acknowledged that the Trump administration had been sceptical as regulations can hamper innovations. And how important it is now to check on China’s approach to AI, which according to Kratsios, “has twisted AI in ways that are in direct conflict with the values of the U.S. and its allies.” Tomorrow, the U.S. will join G7 nations in launching the Global Partnership for Artificial Intelligence, an international effort to advance AI underpinned by our shared democratic values. https:\/\/t.co\/RbdjYo8NR1— Michael Kratsios (@USCTO45) May 28, 2020 The G7 group consists of the seven largest advanced economies in the world that includes Canada, France, Germany, Italy, Japan, the United Kingdom and the United States. Now, as per President Donald Trump statements, he finds the current group outdated and wants countries such as India, South Korea and Australia to join the ranks. President Trump even offered to invite Vladimir Putin of Russia to the expanded G7 meeting in September. Objective Of The Panel This international panel of wealthiest democracies (aka G7) was an initiative pushed by Canada and France back in May 2019. As stated by the French officials in a statement, the objective of this panel is to realise the full potential of AI that benefits all citizens requires international collaboration and coordination The main agenda of the panel was as follows: Promote human-centric and ethical approach to AI, grounded in human rights and support a multi-stakeholder approachStimulate innovation, growth and well-being through AIAlign efforts on AI with the achievement of the 2030 Agenda Foster transparency, diversity, accountability and inclusion through AIPromote and protect democratic values and institutionsPromote international scientific collaboration on AI What Can Be The Implications Michael Kratsios, CTO on the right According to ABI research, the US accounted for 52.3% of the global publicly available venture capital for AI startups and invested $9.7 billion in 2018 eclipsing China, which stands next to its rival with investments close to $7.4 billion in 2018. Being the top contributor of AI, the presence of the US on the panel will have wider implications than by any of its G7 counterparts. For years, American tech executives have suggested that Chinese influence could erode free speech around the world because companies in China are required by law to censor politically sensitive topics while allowing authorities access to users’ data. And, since global dominance has been the de facto theme of many of the Chinese government’s major initiatives in the last couple of decades, it is critical to make sure the world does not hand over the future to a single nation. For instance, according to China Standards 2035, they plan on writing international rules for the technologies that will be used in future, from AI to IoT. While the Chinese vows to tighten the screws on the espionage of ideas, it also pushes for an indigenous, AI supremacy that would finally give credibility to MADE IN CHINA. Top AI companies and research labs across the world, however, should be aware of the past and frame their IP policies more diligently. Though the US prime motive might be to check China in its “twisting” methodologies, the broader objective of this international panel can result in a rulebook that might force the companies like Google and other large internet companies to follow. These tech giants, although have been voicing the need for privacy and innovating on tools that promote transparency on AI, there is no common understanding across the globe how some technology will change the way we live forever. For instance, when OpenAI released its text generation AI model, there were concerns related to the proliferation of fake news. But now the technology is out in the open. AI is a double-edged sword, and the nations have to find a sweet spot that can blunt the malicious end of it while not hampering the innovation. Going Forward Source: Deloitte AI’s ramifications for the future of human development are profound. The first step in maximizing the widespread benefit of AI, while addressing ethical, legal and regulatory challenges, is to create a common factual basis for understanding artificial intelligence. However, there is growing confidence amongst the advanced economies towards AI adoption, and it will continue to do so. Governments will continue to invest in AI heavily in the coming decade. For instance, India has a well laid out plan, a national strategy for AI by its think tank NITI Aayog. This report covered various aspects of using AI, including the ethical side of it. Having a nation like India, which en route to become a global digital hub will be a great addition to the panel and to the future of AI. The International Panel on Artificial Intelligence aims to support and guide the responsible development of artificial intelligence that is grounded in human rights, inclusion, diversity, innovation, and economic growth. The Panel with the inclusion of India and other developing nations can bring together many of the greatest global AI experts and foster international collaboration and coordination on AI policy development. A global elite panel or a watchdog to say the least, for AI usage, can help draw the line before things go out of hand. If things go as planned, this international group can come up with some agreements that will promote the development of AI while checking their malicious usage. It can be thought of as a watered-down version of the nuclear proliferation treaty, but unlike nuclear energy, almost every country with decent internet services can leverage AI. Also watch:","excerpt":"“It’s [G7] a very outdated group of countries. We want Australia, we want India, we want South Korea.” -Donald Trump on G7 The U.S. last week has announced that it will be joining the group, called the Global Partnership on AI, a decision that was reversed from the stance they have taken previously. This group, […]","categories":["AI Features"],"tags":["china ai investments"],"author_name":"Ram Sagar","publish_date":"2020-06-02T16:00:30","publication_year":"2020","word_count":1002,"keywords":["Go","API","artificial intelligence","OpenAI","AI","Git","RAG","china ai investments","BERT","Aim","R"],"extracted_tech_keywords":["AI","artificial intelligence","OpenAI","Aim","RAG","R","Go","Git","API","BERT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/us-g7-ai-group-india\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":27289,"title":"Observe.AI Secures $8 Million In Funding To Use AI To Improve Call Centres","content":"(L-R) Observe.ai co-founders Swapnil Jain, Akash Singh, Sharath Keshava (Image source: Observe.ai) Observe.AI has secured $8 million in Series A funding led by Nexus Venture Partners with participation from MGV, Liquid 2 Ventures, Hack VC and existing investors Emergent Ventures and Y Combinator. They also announced their agent-first voice AI platform to improve caller satisfaction in call centres worldwide. Observe.AI’s voice AI platform provides the agent with real-time feedback on customer sentiment and guides them on next best action during the customer call. The AI platform listens to the call stream in real time, uses deep learning and natural language processing (NLP) to understand the context and generates suggestions and guidance for the agent. Founded in May 2017 by Akash Singh, Sharath Keshava and Swapnil Jain, Observe.AI has one of the largest call centres in the US among its customers. Swapnil Jain, founder and CEO of Observe.AI, said in a statement, “Humans like talking to humans because it gives them a sense of assurance which is at the core of a delightful customer experience. We are using the power of AI to make that voice conversation even more delightful by equipping the customer support agent with the tools needed while the call is going on. The agent no longer needs to place customers on hold, or transfer them around; continuous engagement is established until a successful resolution is reached. Our agent-first approach is all about making the job of the agent easier which translates into better productivity and higher customer satisfaction.” With this round, welcoming @NexusVP onboard. Thanks, @emergent_vc and @ycombinator for their continued support. Onwards! — Swapnil Jain (@swapnil) August 13, 2018 Ram Gupta, managing director of Nexus Venture Partners, added in the statement, “Companies have been actively discouraging their customers to call their agents for the last two decades because of increasing costs even though it is a natural way for humans to get help. With the recent advances in deep learning and NLP, which will dramatically increase the productivity of agents, call centres are ready to become the first port of contact again for customer service. We are excited to partner with the team at Observe.AI as it leads this positive disruption in the call centre ecosystem.”","excerpt":"Observe.AI has secured $8 million in Series A funding led by Nexus Venture Partners with participation from MGV, Liquid 2 Ventures, Hack VC and existing investors Emergent Ventures and Y Combinator. They also announced their agent-first voice AI platform to improve caller satisfaction in call centres worldwide. Observe.AI’s voice AI platform provides the agent with […]","categories":["AI News"],"tags":["observe.ai"],"author_name":"Prajakta Hebbar","publish_date":"2018-08-14T07:20:52","publication_year":"2018","word_count":370,"keywords":["Go","funding","observe.ai","AI","programming_languages:R","RAG","NLP","deep learning","ViT","disruption","R"],"extracted_tech_keywords":["AI","deep learning","NLP","RAG","R","Go","ViT","disruption","funding","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/observe-ai-secures-8-million-in-funding-to-use-ai-to-improve-call-centres\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10143777,"title":"SiMa.ai, Synopsys Partner to Advance Automotive AI Solutions","content":"Synopsys, a silicon provider software company, and SiMa.ai, a machine learning system-on-chip (MLSoC) company, have announced a strategic collaboration. This partnership aims to accelerate the development of AI-enabled silicon and software for next-generation vehicles. The joint solution is designed to address the increasing demands of Advanced Driver Assistance Systems (ADAS) and In-Vehicle Infotainment (IVI) applications. The partnership will combine Synopsys’ electronic design automation (EDA), automotive-grade IP, and hardware verification technologies with SiMa.ai’s machine learning accelerator (MLA) IP and ML software stack. The collaboration aims to help automakers deliver AI-ready, power-efficient solutions for workload-specific chip development, as per the official blog. AI Engines Ready to Drive Krishna Rangasayee, founder and CEO at SiMa.ai, announced this collaboration on LinkedIn and said, “Look forward to helping all automotive OEMs and Tier1s globally get ahead on their ADAS and IVI needs, with industry-leading AI\/ML capabilities.” The solution will focus on enabling early architecture exploration, optimised software development, and cost-effective in-vehicle experiences. It will also support continuous upgrades, including over-the-air updates for automotive edge AI applications and the capability to guide automotive design engineers. This will help them choose performance, power, and software application requirements for custom or 3rd party SoC development. Ravi Subramanian, head of Synopsys’s product management and markets group, said, “Our collaboration with SiMa.ai will provide automotive companies with a significant step forward in modernising their hardware\/software co-design processes and enable customers to meet cost considerations and industry standards.” Rangasayee further added, “Our MLSoC platform delivers high performance at low power, enabling automakers to create smarter and safer vehicles.” In a previous interaction with AIM, Rangasayee also stated that the company is betting on India’s renewed focus on manufacturing and semiconductor fabrication ambitions. For strategic reasons, it has chosen not to sell its chips in China, despite it being the largest semiconductor market in the world. Now Fully Customisable In September this year, SiMa.ai announced the launch of MLSoC Modalix, the industry’s first multi-modal edge AI product family. The partnership comes as automotive manufacturers face increasing pressure to support generative AI and real-time, multi-modal in-car experiences. Automakers require solutions that integrate software and hardware while meeting stringent performance, reliability, and cost requirements. Both companies claim to address these challenges with a fully customisable workflow, which reduces development costs and supports faster time-to-market for AI-enabled automotive solutions.","excerpt":"“Our MLSoC platform enables automakers to create smarter and safer vehicles,” says Krishna Rangasayee, founder and CEO at SiMa.ai.","categories":["AI News"],"tags":["automotive industry","sima.ai"],"author_name":"Sanjana Gupta","publish_date":"2024-12-18T00:46:12","publication_year":"2024","word_count":384,"keywords":["machine learning","programming_languages:R","AI","Modal","R","ML","automation","Aim","generative AI","edge AI","automotive industry","sima.ai"],"extracted_tech_keywords":["AI","machine learning","ML","generative AI","Aim","edge AI","R","automation","Modal","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/sima-ai-synopsys-partner-to-advance-automotive-ai-solutions\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10092536,"title":"6 Open Source Models From OpenAI","content":"Elon Musk co-founded OpenAI—-the firm behind the wildly popular ChatGPT. But he has been vocal about the company not staying true to its name. Musk recently tweeted expressing his disappointment at the company becoming “a closed source, maximum-profit”. “OpenAI was created as an open source (which is why I named it “Open” AI), non-profit company to serve as a counterweight to Google, but now it has become a closed source, maximum-profit company effectively controlled by Microsoft,” he tweeted. The company has previously faced criticism for its closed-door policy, not only from the Tesla CEO, but a host of industry experts and members of the open-source community. The company has been accused of taking advantage of the open-source community without giving back. However, OpenAI is now trying to embrace the open-source approach.  By going open again with its ‘Consistency Models’, OpenAI is positioning itself for greater collaboration and contribution to the open-source community. The company has previously open sourced quite a few models. Here are some Open source models from OpenAI: Evals: OpenAI open-sourced a software framework called Evals that allows users to evaluate the performance of AI models. The framework enables users to identify deficiencies in their models and provide feedback to direct improvements. OpenAI staff will actively review these evaluations when considering improvements to upcoming models. The tools are aimed at creating a vehicle to share and crowdsource benchmarks that represent a wide set of failure modes and difficult tasks. OpenAI plans to grant GPT-4 access to those who contribute high-quality benchmarks. Whisper: OpenAI introduced a multilingual speech recognition system called Whisper in September of 2022. Whisper is trained on 680,000 hours of multilingual and multitask supervised data. Whisper uses a simple end-to-end approach implemented as an encoder-decoder Transformer and has improved recognition of background noise, unique accents, and technical jargon. It does not beat models that specialise in LibriSpeech performance but shows robust zero-shot performance across many diverse datasets, making 50% fewer errors than other models. The open-sourced models and inference code will allow developers to add voice interfaces to a wider set of applications. Spinning up: Spinning Up is an educational resource by OpenAI for learning about deep reinforcement learning (deep RL), which is a combination of machine learning and deep learning. It includes an introduction to RL terminology and theory, an essay on becoming an RL researcher, a list of important papers, code implementations of key algorithms, and exercises. CLIP: OpenAI CLIP is a machine learning model that uses natural language descriptions of images to perform tasks related to natural language and image processing. It can classify images, detect objects, and retrieve images based on text prompts. CLIP is trained on a large dataset of images and captions and is available as an open-source model. Its unique feature is that it can perform well on a variety of tasks without the need for annotated image data. Jukebox: OpenAI Jukebox is a generative model that creates music using deep neural networks trained on a large dataset of music samples from various genres. It can generate original music samples that are similar in style and structure to different types of music. Jukebox can also generate music with lyrics based on a given prompt. It is an open-source project used by researchers and musicians worldwide to explore the capabilities of generative models in the creative arts. Point-E OpenAI’s GPT-3 Point-Eleven or Point-E is an optimised variant of its language model GPT-3 for conversational AI applications. It uses a larger context window and other optimisations to improve the naturalness and coherence of the model’s responses in conversations. Point-E is not available as a standalone model, but is offered through OpenAI’s GPT-3 API, which provides various language-based services, including text completion, question-answering, and conversational AI.","excerpt":"Despite being accused of taking advantage of the open-source community, OpenAI has previously released quite a few models.","categories":["AI Trends"],"tags":["ChatGPT","Elon Musk","Open Source AI","OpenAI","Tesla","Whisper"],"author_name":"Shyam Nandan Upadhyay","publish_date":"2023-05-02T11:16:33","publication_year":"2023","word_count":624,"keywords":["Whisper","ChatGPT","Go","machine learning","API","OpenAI","AI","neural network","Elon Musk","Open Source AI","Aim","deep learning","Tesla","R"],"extracted_tech_keywords":["AI","machine learning","deep learning","neural network","ChatGPT","OpenAI","Aim","R","Go","API"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/6-open-source-models-from-openai\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10002108,"title":"China Makes Its Lunar Ambitions Clear With Change Missions","content":"The Chinese, in the past few years, have made several breakthrough advancements in space missions. With its Changi 4 landing on the far side of the moon and the country planning to launch three more lunar missions, it is now gearing up to go build the first ever moon base by the year 2020. China’s Moon Base China is the first country that is rigorously planning to build a moon base and has already begun to work towards it by sending satellites. Its main objective is to ensure that the conditions on the moon are favourable enough for humans to settle in the future. For this, the country, with the number of satellites that it has and is going to set up on the moon, trying to develop bioregenerative support systems. It also mainly aims to be the first country ever to build a Chinese research base on the moon. The Agenda Building a lunar base requires considerable amount of accurate calculations based on the weather conditions, sunlight, soil conditions and many other factors. The moon base is going to be set up in the south pole. Here are the efforts in terms of missions that China has taken: Change 4: This January, the China National Space Administration successfully landed the unmanned Chang’e-4 on the far side of the moon. With Change 4, China has already showcased its ability to grow plants on the lunar surface, since the seeds taken to the Moon in the mission have now reportedly sprouted. This is the first time plants have been grown on the Moon. Change 5: By the end of 2019, the country will send an unmanned lunar lander called Change 5, the job of which would be to collect the samples on the surface. By analysing the composition of the lunar far side, they could gain valuable information about the soil at the location. The Moon is already thought to contain significant amounts of helium-3, which is rare on Earth. Change 6: Change 6 will bring samples back from the lunar south pole. Change 7: Change 7 will conduct surveys of the south polar region, including an assessment of terrain, physical composition and the space environment. Change 8: The last one in the series, Change 8 will test key technologies for building a research station on the moon. The mission will be sent to very some important and related technologies which is expected to do some preliminary exploration for a joint moon research base, shared by multiple countries in the future. Meanwhile, China has also placed its astronauts aboard two temporary space stations of Tiangong-1 and Tiangong-2. The first part of that permanent station is expected tol reach orbit aboard the country’s new Long March-5B rocket in the first half of 2020. This will not be associated with the International Space Station (ISS), which is reaching the end of its operational lifetime. Apart from this, China is on its mission to develop the capability to 3D print both inside and outside of the lunar base. 3D printing is much more difficult than on Earth and is a real challenge. Though the basic design of a 3D printer stays the same, printing in zero gravity environment of the Moon requires special considerations. But the material itself has to be sticky to hold the liquid layers without gravity. Importance Of The Location The lunar south pole has always been an area of special interest mainly because of two reasons: (1) It is a shadowed region. The cratoers in this region are unique because sunlight does not reach here, Such craters are cold traps that contain a fossil record of the early Solar System. (2) On top of being a shadowed region, the south pole has the occurrence of water ice needed for crewed operations. What Will This Bring Us? China is currently leading in the world in terms of lunar missions. It has taken the first step by examining the soil of the lunar surface. This is necessary for building an underground habitat and supporting infrastructure that will shield the base from the harsh surface conditionWith one of its missions as a part of the moon base already landing on moon, it is clear that the country is very serious about finding out possibilities to setting up a habitat on the home satellite. Whether China will be successful in setting up a habitat on Moon or not is still a question, but it with all its upcoming missions, it is sure going to make a lot of developments for space exploration.","excerpt":"The Chinese, in the past few years, have made several breakthrough advancements in space missions. With its Changi 4 landing on the far side of the moon and the country planning to launch three more lunar missions, it is now gearing up to go build the first ever moon base by the year 2020. China’s […]","categories":["AI Features"],"tags":["China","Interviews and Discussions","satellite"],"author_name":"Disha Misal","publish_date":"2019-05-21T16:29:23","publication_year":"2019","word_count":758,"keywords":["Go","satellite","AI","programming_languages:R","programming_languages:Go","Aim","ViT","R","China","Interviews and Discussions"],"extracted_tech_keywords":["AI","Aim","R","Go","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/china-makes-its-lunar-ambitions-clear-with-change-missions\/","complexity_score":3,"technical_depth":7,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":69271,"title":"What Is The Difference Between A Data Warehouse And A Data Center?","content":"Over the decades, enterprises have accumulated a large number of enterprise data assets. However, traditional data warehouse technology, data management and analysis capabilities have become shortcomings in the business intelligence work as companies are failing to eliminate the data silos. The role of the data warehouse is to achieve data integration across business lines and systems to provide unified data support for management analysis and business decision-making. A data warehouse can fundamentally help you transform your companies’ operating data into high-value, accessible information (or knowledge), and deliver the right information to the right people in the right way at the right time. But in other cases, the traditional data warehouse can not meet the needs of data analysis. Enterprises present challenges in data analysis applications such as strong demand for a unified data platform, the data centre’s computing power, core algorithms, and data comprehensiveness put forward higher requirements. With data warehouses, it is difficult to assess value mining on global data, and it cannot truly reflect the value of the group’s huge data assets in terms of scale and effect. With the market competition and the increasing globalisation, enterprises are not only satisfied with the analysis of internal data but also need to conduct a comprehensive analysis through external technologies such as the web and enterprise applications. This led to data centres, which is not just a system or tool, but a functional department that provides data asset management and services for the entire organisation through a series of platforms, tools, processes, and specifications. This replaced computing and storage mashup data processing architecture with Hadoop, Spark with distributed technologies and components as the core, which can support batch and real-time data loading and flexible business requirements. The architecture system of data centres in the context of big data is the ELT structure, which extracts the desired original data from data centres for modelling and analysis at any time according to applications’ requirements of the upper layer. Secondly, the goal of establishing data centres is to fuse all the data of the entire enterprise, open up the gap between the data, and eliminate the inconsistency between data formats. Data centres play a vital role in the digital transformation and sustainable development of enterprises; data centres are born for decoupling. The biggest meaning of building data centres for enterprises is application and data decoupling. In this way, enterprises can build data applications that meet business needs on-demand without restriction. Characteristics In traditional data warehouses, integration is the most critical. Because of the cost of processing and storage, its data needs to be extracted from different data sources and concentrated, and the redundancy of its data needs to be minimised as much as possible. Therefore, data coming into data warehouses need to be converted, formatted, rearranged, and summarised. All its data has a single physical characteristic and exists in a structured manner. The new generation of data warehouses uses distributed computing, but software products exist in a centralised deployment. In terms of system architecture, data warehouse also exists in centralised storage and computing. In comparison, the data centre is the link point between the front desk and the back office and precipitates common tools and technologies for the business. Data centres refer to comprehensive data capability platforms that integrate data collection. Data centralisation means that through internal and external multi-source heterogeneous data collection, governance, modelling, analysis, and application, the internal management of data can be optimised to improve business, and the value of data cooperation can be released to the outside, becoming the hub of enterprise data asset management. Data centres’ overall technical architecture adopts a cloud computing architecture model for computing resources and storage resources, and packages and integrates resources through multi-tenant technology and opens up to provide users with “one-stop” data services. Capacity & Deployment Data warehouses are essentially relational databases that have a database design, which is suited for historical analytical purposes. They hold data in them which actually are hosted on the servers that reside in data centres. So, ultimately, a data warehouse is a relational database with a different database\/schema design. You can say data warehouses are deployed on servers which reside inside data centres, physically. Data warehouses are central repositories of integrated data from different sources. They store the current and historical data in one single place that are utilised for making analytical reports for organisations. Traditional data warehouses are mainly used to make BI reports. It contains a complete set of content such as data modelling, metadata management and data quality management. Data warehouse systems will record all records; it will retain all the changes in the records, but it is limited by cost and calculation. Considering the capacity, data warehouses will not record the full amount of detailed data, especially the log data, so the data capacity is less in most of the data warehouse platforms. On the other hand, data centres are built on distributed computing platforms and storage platforms, which can theoretically expand the computing and storage capabilities of platforms indefinitely. Most traditional data warehouse tools are based on a single machine. Once the data volume becomes larger, it will be limited by the capacity of the single machine. Data centres are not simply building open-source big data frameworks and developing some data tables. This requires teams to have a certain understanding of methodologies. Overall, it means that your team has sufficient skills. The amount of resources invested determines the construction of data centres.","excerpt":"Over the decades, enterprises have accumulated a large number of enterprise data assets. However, traditional data warehouse technology, data management and analysis capabilities have become shortcomings in the business intelligence work as companies are failing to eliminate the data silos. The role of the data warehouse is to achieve data integration across business lines and […]","categories":["Deep Tech"],"tags":["back office data","business intelligence use cases","data centres","data warehouse","Data Warehousing","purposes of a data team"],"author_name":"Vishal Chawla","publish_date":"2020-07-08T18:00:00","publication_year":"2020","word_count":911,"keywords":["big data","Go","ELT","business intelligence use cases","AI","cloud computing","distributed computing","Git","RAG","data centres","Data Warehousing","purposes of a data team","back office data","R","data warehouse"],"extracted_tech_keywords":["AI","RAG","cloud computing","distributed computing","R","Go","Git","big data","ELT","data warehouse"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/what-is-the-difference-between-a-data-warehouse-and-a-data-center\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":38505,"title":"How This Graph-Powered Real Time Engine Handles A Billion Recommendations","content":"Recommendation systems are the cornerstone of a majority of the modern day billion dollar industries. From Amazon to Netflix to Pinterest. Amazon recommends commodities for its e-commerce website and Netflix has a limited number of movies to recommend but in the case of Pinterest, the numbers are not that small. It has more than 100 billion ideas saved and has to deliver to more than 200 million of its users in real time. Source: Pinterest Pinterest opted for a graph-based system for this high performance task. They do this with their custom built technique named Pixie. Pixie’s ease of generalisation to different types of tasks makes this possible, whether it’s creating many types of graphs, recommending different objects, or tuning the graphs to capture the absolute right content. Now with Pixie and recent optimisations, Pinterest is streamlined to serve relevant ideas in real time. The above diagram illustrates the graphs connecting the boards and pins (images saved by the user). Here each edge shows a saved pin to the board. As the same pins can be saved to different boards for example, a bike can be saved into both ‘bikes’ board or ‘royal enfield’ or both. This eventually increases the number of edges to more than 100 billion. These graphs can be loaded on to AWS machines and deployed at scale. Source:Pinterest The Pixie algorithm examines the portion of the graph nearest to the relevant nodes by using biased random walk algorithm. A biased  random walk is an approach for statistical analysis of graphs. They are used to extract symmetries when the graph contains complex networks. At each step, it selects a random neighbor and visits the node, incrementing node visit counts as it visits more random neighbors. The probability Alpha, set at 0.5, to restart at node Q so our walks do not stray too far. And then neighboring nodes and boards undergo randomly sampling for 100,000 steps. So, in the case of the bike example, if a user opens and checks bike images 3 times then this information will be given to Pixie and it comes up with thousands of similar images. The nodes that have been visited 14 and 16 times are the ones that are most closely related to the query node. Pixie continuously repeats this process in real-time as the data grows, so the users are always able to keep narrowing down their searches and find the exact ideas they’re looking for to pursue their goal. The top 1,000 most visited nodes are retrieved , to avoid random walks through the graphs  for complete 100,000 steps every time. To accomplish this, walking is done until the rank 1,000 candidate gets at least 20 visits.  Along with this the engineering team at Pinterest also created Graph Pruning to cut down edges. Though the recommendation systems have evolved over time, real-time(< 100 milliseconds) large scale application is still a challenging task for the online platforms. Pinterest’s Pixie is one of the best solutions available out today. Pixie uses 3 billion nodes and 17 billion edges and the results show that the user interaction has increased up to 50% when compared to Hadoop-based production system. Key Takeaways Of Pixie 3 billion nodes are pruned and fit to 120 GB main memory. This reduction in memory improves performance. Graphs enable parallel execution hence lesser time(less than 60 milliseconds). Pixie can be scaled up by simply adding more machines(AWS) to the cluster. Read more here","excerpt":"Recommendation systems are the cornerstone of a majority of the modern day billion dollar industries. From Amazon to Netflix to Pinterest. Amazon recommends commodities for its e-commerce website and Netflix has a limited number of movies to recommend but in the case of Pinterest, the numbers are not that small. It has more than 100 […]","categories":[],"tags":["Pinterest","recommendation engine","what is power bi"],"author_name":"Ram Sagar","publish_date":"2019-04-30T12:29:28","publication_year":"2019","word_count":575,"keywords":["Go","what is power bi","AWS","AI","cloud_platforms:AWS","programming_languages:R","ML","recommendation systems","programming_languages:Go","Pinterest","recommendation engine","Ray","R"],"extracted_tech_keywords":["AI","ML","Ray","recommendation systems","AWS","R","Go","cloud_platforms:AWS","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/graph-based-recommendation-pinterest-billion-images\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":41617,"title":"Hackers Can Misuse Microsoft Azure’s Unpatched Flaw To Take Over EA Origin’s User Accounts","content":"From Computer Space arcade game in the 70s to 2019, the gaming industry has evolved tremendously. The industry generated about $135 billion in 2018 and is expected to make $180.1 billion profit in 2021, making it much more lucrative than ever before. With such massive growth and everything going digital, the industry is also becoming a high-value target for the hackers. EA Origin Vulnerability Looking at the present scenario, American video game company EA games seem to be one of the HVTs. Being the second-largest gaming company in the Americas and Europe by revenue and as well as market capitalization after Activision Blizzard, Hackers have started to keep an active eye on the company as well as its user base. Recently, Check Point Research, the threat intelligence arm of Check Point Software Technologies, and CyberInt, a cybersecurity provider of managed threat detection and mitigation services, discovered a chain of severe vulnerabilities in EA Origin. The vulnerabilities were so severe that they have exposed data of more than 300 million players. Further, once exploit, it would allow hackers to completely take over the origin account. Origin is a digital distribution platform developed by Electronic Arts (EA) that allows gamers to purchase and play games. Gamers can download the software clients of Origin and install in on their PCs or mobile devices. Also, Origin can be linked to users’ EA Games account— it allows them to connect with other games\/friends, join games, etc. Therefore, this discovery of the vulnerabilities is definitely a shock to the community. The Bigger Picture Of The Origin Vulnerability When the attack was performed by the experts, they were able to take over one of the EA subdomains, eaplayinvite.ea[.]com, which was previously registered with Azure to host one of the Origin’s services. The experts basically took advantage of an unpatched vulnerability in Microsoft’s Azure cloud service. To be precise, hackers could basically exploit the trust mechanism that exists between ea.com and origin.com domains and their subdomains and manipulates the OAuth protocol. If in case you don’t know what OAuth protocol is, it is an open standard for access delegation.  It is mostly used as a way for Internet users to grant websites or applications access to their information on other websites without providing the passwords. And talking about a worst case scenario, once completely taken over, attackers could have used the user’s credit card information to make purchases on behalf of the user. However, the cybersecurity firms who carried out the entire research about the vulnerabilities, CyberInt and Check point, informed the company to fix the issues. Instances Of Hackers Targeting The Gaming Industry Between 2013 and 2014, multiple online gaming companies such as Sony, Riot Games, Microsoft, Nintendo, Valve, and Electronic Arts witnessed a series of denial-of-service (DoS) attacks. And to the surprise, the culprit behind the event was a 23-year-old boy for Utah, Austin Thompson, a.k.a. “DerpTroll,” who is also believed to the operator of the DerpTrolling hacking group. Last year, on December 28,  Town of Salem, a browser-based game that lets players to convincingly lie as well as detect when other players are lying, witnessed a hack. Hackers have released a copy of the compromised game’s database on DeHashed, a hacked database search engine. And in January 2019, the report that shocked the entire community is that the data breach is a massive one. The breach was so severe that out of more than 8 million players, it impacted over 7.6 million players. The exposed data includes email addresses, usernames, hashed passwords, game and forum activity, and also payment information of some players along with full names, billing and shipping addresses, IP information and payment amount. This is not the only instance; hackers are also targeting some of the vintage games. Even though it’s been quite a 19 years, the counter strike craze is still alive among gamers and the number of users is still significantly decent. And recently in March, according to a source, security firm Dr Web released a report stating 39% of all existing Counter-Strike 1.6 game servers online are malicious and also an attacker is exploiting zero-day flaws in game clients. The scenario here is that several servers have been set up by hackers. The owners of malicious, fake servers made a significant amount of money by selling game specific privileges such as access to weapons and protection against bans. However, the prime motive behind is to pwn the gamers’ PCs and laptops by exploiting zero-day vulnerabilities in the game client. Wrapping Up As industries to continue to grow, the risk of getting targeted by hackers also increases. Over the years, we have witnessed a massive amount of hacking events where companies have lost millions of dollars. Talking about this recent vulnerabilities in Origin and MS Azure, even though experts have managed to save the day, it shows negligence and denial. Cybersecurity is more important than ever in this era; an organisation must secure each and every end. After all, it is about the data (of both the users and the company).","excerpt":"From Computer Space arcade game in the 70s to 2019, the gaming industry has evolved tremendously. The industry generated about $135 billion in 2018 and is expected to make $180.1 billion profit in 2021, making it much more lucrative than ever before. With such massive growth and everything going digital, the industry is also becoming […]","categories":["Global Tech"],"tags":["Microsoft Azure","video games"],"author_name":"Harshajit Sarmah","publish_date":"2019-07-01T17:28:02","publication_year":"2019","word_count":841,"keywords":["Go","API","Rust","AWS","AI","R","Git","ViT","Microsoft Azure","GAN","video games","Azure"],"extracted_tech_keywords":["AI","AWS","Azure","R","Go","Rust","Git","API","GAN","ViT"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/hackers-can-misuse-microsoft-azures-unpatched-flaw-to-take-over-ea-origins-user-accounts\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10045486,"title":"Hands-On Tutorial on Mean Shift Clustering Algorithm","content":"Mean shift clustering algorithm is a centroid-based algorithm that helps in various use cases of unsupervised learning. It is one of the best algorithms to be used in image processing and computer vision. It works by shifting data points towards centroids to be the mean of other points in the region. It is also known as the mode seeking algorithm. The algorithm’s advantage is that it assigns clusters to the data without automatically defining the number of clusters based on defined bandwidth. Kernel Density Estimation Like other clustering algorithms, Mean shift is based on the concept of Kernel Density Estimation(KDE), which is a way to estimate the probability density function of a random variable. KDE is a problem where the inferences of the population are made by data smoothing. It works by providing weights to each data point. The weight function is called a kernel. There are many kinds of kernels, one kind of kernel is the Gaussian kernel. Adding all those kernels together creates a density function(probability surface). The resultant density function variation depends on the used bandwidth parameter. In the image below, we can see the distribution of some data points in a surface plot. Image source And in the image below, we can see the KDE surface where our data points are distributed in the surface plot(first image). The hills can be considered as the kernel. Image source In the contour plot of the KDE surface, we can see the exact smoothing of our data points. Image source From the images, we can understand how the KDE works in smoothing the data sets to make inferences from the data points. As the size of circles in the plot decreases, the density of the data point increases, which means most of the points in the kernel are trying to be on the small circle where the mean shift comes into the picture, which tries to increase or decrease the density function. Mean shift Mean shift is based on the idea of KDE, but what makes it different is that using the bandwidth parameter. We can make the points climb uphill to the nearest peak on the KDE surface. So, iteratively shifting each point to climb uphill to the peak. The bandwidth parameter used to make the KDE surface varies on the different sizes. For example, we have a tall skinny kernel which means a small kernel bandwidth and in a case where the size of the kernel is short and fat, which means a large kernel bandwidth. A small kernel bandwidth makes the KDE surface hold the peak for every data point more formally, saying each point has its cluster; on the other hand, large kernel bandwidth results in fewer kernels or fewer clusters. Image source Here we can see the formation of kernels with bandwidth values is equal to two. Image source In the image, we can see what happens when the bandwidth value is low. Let’s consider a kernel function ????(xi – x) gives the weight to nearby points for defining the mean. So the weighted mean of the density in a window calculation is determined by. Image source. Where N(x) is the neighbourhood of x. The value of m(x) – x is called the mean shift. As discussed before, from the mathematical formula, we can understand that the mean shift tries to shift the point, and when performed iteratively, it will move to the KDE peak. Basically, in the whole algorithm, after making a copy of data points, those copied points are shifted against the original copy to reach the peak of its kernel surface. Next in the article, we will see how we can implement the algorithm using python with randomly generated data points to find out the clusters according to the size and bandwidth parameter. Implementations in Python Importing the libraries: import numpy as np import pandas as pd from sklearn.cluster import MeanShift from sklearn.datasets.samples_generator import make_blobs import matplotlib.pyplot as plt from mpl_toolkits.mplot3d import Axes3D ordinates = [[2, 2, 3], [6, 7, 8], [5, 10, 13]] X, _ = make_blobs(n_samples = 120, centers = cordinates, cluster_std = 0.60) Setting up the coordinates and generating the random data around the coordinates: Visualizing the data points: data_fig = plt.figure(figsize=(12, 10)) ax = data_fig.add_subplot(111, projection ='3d') ax.scatter(X[:, 0], X[:, 1], X[:, 2], marker ='o',color ='green') plt.show() Output: Here we can see how the data is distributed in space. In space, we can easily say that there can be 3 clusters according to the coordinates as the inferences of the data. Now we will proceed with the mean shift to predict the cluster and define the centroids of the clusters. Sklearn provides the estimation function for bandwidth according to the data so that we don’t need to be worried about the bandwidth parameter. Importing the estimated bandwidth function. Importing libraries: from sklearn.cluster import  estimate_bandwidth bandwidth = estimate_bandwidth(X, quantile=0.2, n_samples=500) Now we can define the mean shift cluster model and fit it into our data. msc = MeanShift(bandwidth=bandwidth, bin_seeding=True) msc.fit(X) cluster_centers = msc.cluster_centers_ labels = msc.labels_ cluster_label = np.unique(labels) n_clusters = len(labels_unique) n_clusters Output: Here we can see it has been predicted as we have estimated there should be 3 clusters. Visualizing the clusters: msc_fig = plt.figure(figsize=(12, 10)) ax = msc_fig.add_subplot(111, projection ='3d') ax.scatter(X[:, 0], X[:, 1], X[:, 2], marker ='o',color ='yellow') ax.scatter(cluster_centers[:, 0], cluster_centers[:, 1], cluster_centers[:, 2], marker ='o', color ='green', s = 300, linewidth = 5, zorder = 10) plt.title('Estimated number of clusters: %d' % n_clusters) plt.show() Output: Here in the green color we can see the cluster’s centroids and easily separate the data into 3 clusters. As we have discussed, it is very useful for image processing and computer vision. Next in the article, I am going to separate the colours of the images using the mean shift clustering algorithm. More formally, we can call it image segmentation using mean shift as we know that the pixel values in any image are based on the colors present in the image. Here I am using a thermograph as the image because the colours in this image are well distributed, and the number of colors is insufficient, so in the procedure, we will not get confused. Importing the libraries. import numpy as np from sklearn.cluster import MeanShift, estimate_bandwidth from sklearn.datasets.samples_generator import make_blobs from itertools import cycle from PIL import Image import matplotlib.pyplot as plt import matplotlib.pylab as pylab %matplotlib inline Loading and visualizing the image using PIL and Matplotlib: Input: img = Image.open('\/content\/drive\/MyDrive\/Yugesh\/Mean Shift Clustering Algo\/Thermography_results_sm.jpg') img = np.array(image) # saving the image shape shape = img.shape # reshaping image reshape_img = np.reshape(image, [-1, 3]) #plotting the image plt.imshow(image) plt.title(img.shape) Output: Here we can see the image and its size as the title of the image. We have reshaped the image to flatten it so that the size of the array the model required we can get it. As we have discussed the bandwidth function of sklearn here, I am defining the bandwidth using the function. Input: bandwidth = estimate_bandwidth(reshape_img, quantile=0.1, n_samples=100) bandwidth Output: Fitting the meanshitt on reshape_img: msc = MeanShift(bandwidth=bandwidth, bin_seeding=True) msc.fit(reshape_img) Output: Checking the insights of the model so that we can know what is going behind: print(\"shape of labels : %d\" % msc.labels_.shape) print( msc.cluster_centers_.shape) print(\"number of estimated clusters : %d\" % len(np.unique(msc.labels_))) Output: Here we can see that it has generated 8 clusters which means that the image has clustered into 8 color segments—changing the shape of the labels, equivalent to the shape of the original image. labels = msc.labels_ result_image = np.reshape(labels, shape[:2]) Let’s draw the images original and segmented. fig = plt.figure(2, figsize=(14, 12)) ax = fig.add_subplot(121) ax = plt.imshow(img) ax = fig.add_subplot(122) ax = plt.imshow(result_image) plt.show() Output: Here we can see the original image and the resulting image. Using the pixel sizes of the images, we have generated the clusters using the mean shift algorithm. It has given us clusters for the image pixel values(note – the pixel values vary between 0 to 255). This is one of the easiest techniques to solve image segmentation and other image processing problems. We have seen earlier in the topic how it works and provides centroids to the data points, and also we have seen how it uses the mean shift in the KDE surface. There are various advantages of the algorithms like no effects of outliers, efficiency for complex structure datasets and no need to iterate between several clusters. References A demo of the mean-shift clustering algorithm.Build your own mean shift.Google Colab for basic implementation code.Google Colab for image segmentation.image.","excerpt":"Mean shift is based on the idea of KDE. We can make the points climb uphill to the nearest peak on the KDE surface. Mean shift is based on the idea of KDE, but what makes it different is that using the bandwidth parameter. We can make the points climb uphill to the nearest peak on the KDE surface. So, iteratively shifting each point to climb uphill to the peak.","categories":["AI Trends"],"tags":["AI Tool"],"author_name":"Yugesh Verma","publish_date":"2021-08-08T18:00:00","publication_year":"2021","word_count":1419,"keywords":["NumPy","TPU","AI","ML","computer vision","Colab","Ray","Python","AI Tool","Matplotlib","Pandas"],"extracted_tech_keywords":["AI","ML","computer vision","Ray","Colab","Pandas","NumPy","Matplotlib","TPU","Python"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/hands-on-tutorial-on-mean-shift-clustering-algorithm\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10010925,"title":"Apple Acquires AI Startup For $50 Million To Advance Its Apps","content":"In an attempt to scale up its AI portfolio, Apple has acquired Spain-based AI video startup — Vilynx for approximately $50 million. Reported by Bloomberg, the AI startup — Vilynx is headquartered in Barcelona, which is known to build software using computer vision to analyse a video’s visual, text, and audio content with the goal of “understanding” what’s in the video. This helps it categorising and tagging metadata to the videos, as well as generate automated video previews, and recommend related content to users, according to the company website. Apple told the media that the company typically acquires smaller technology companies from time to time, and with the recent buy, the company could potentially use Vilynx’s technology to help improve a variety of apps. According to the media, Siri, search, Photos, and other apps that rely on Apple are possible candidates as are Apple TV, Music, News, to name a few that are going to be revolutionised with Vilynx’s technology. With CEO Tim Cook’s vision of the potential of augmented reality, the company could also make use of AI-based tools like Vilynx. The purchase will also advance Apple’s AI expertise, adding up to 50 engineers and data scientists joining from Vilynx, and the startup is going to become one of Apple’s key AI research hubs in Europe, according to the news. Apple has made significant progress in the space of artificial intelligence over the past few months, with this purchase of UK-based Spectral Edge last December, Seattle-based Xnor.ai for $200 million and Voysis and Inductiv to help it improve Siri. With its habit of quietly purchasing smaller companies, Apple is making a mark in the AI space. In 2018, CEO Tim Cook said in an interview that the company had bought 20 companies over six months, while only six were public knowledge.","excerpt":"In an attempt to scale up its AI portfolio, Apple has acquired Spain-based AI video startup — Vilynx for approximately $50 million. Reported by Bloomberg, the AI startup — Vilynx is headquartered in Barcelona, which is known to build software using computer vision to analyse a video’s visual, text, and audio content with the goal […]","categories":["AI News"],"tags":["AI Startups"],"author_name":"Sejuti Das","publish_date":"2020-10-28T19:20:14","publication_year":"2020","word_count":302,"keywords":["Go","artificial intelligence","programming_languages:R","AI","ETL","programming_languages:Go","computer vision","AI research","R","AI Startups","startup"],"extracted_tech_keywords":["AI","artificial intelligence","computer vision","R","Go","ETL","startup","AI research","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/apple-acquires-ai-startup-for-50-million-to-advance-its-apps\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10082536,"title":"Twitter Bans Koo","content":"Twitter banned one of the Koo handles by the name Koo Eminence on the grounds of violating Twitter Rules. Aprameya Radhakrishna, Co-Founder and CEO at Koo, expressed disapproval, saying that Koo got banned just for competing with Twitter. Alongside Koo, Mastodon, the decentralised social networking platform, has also been suspended. Twitter is also marking all links to Mastodon “unsafe.” Lately, many accounts of journalists and organisations have been de-platformed. This has led to people questioning whether free speech exists on the platform. For example, last week, the CEO of Twitter, Elon Musk, faced criticism for banning the ElonJet account, which tracked his real-time location. However, many have described this to be an overreaction from Musk since such information is publicly available on any flight tracker. Likewise, Musk has also suspended the accounts of journalists from CNN, the New York Times, and the Washington Post for creating his real-time location, which he refers to as “assassination coordinates”, in direct violation of Twitter’s terms of service. Aaron Rupar, an independent journalist whose Twitter account was banned, said that he was banned for publishing a newsletter covering an article by Noah Berlatsky about Musk’s reactionary populism and for taking a dig at Musk, saying that Musk seems to violate Twitter’s policy for posting footage of someone without their consent. The spree of banned accounts comes after Musk heralded himself as the harbinger of free speech. Previously, Musk had also reinstated nearly 62000 banned accounts having more than 10000 followers; notable names include Donald Trump, Kanye West, and Jordan Peterson. While there were several criticisms of the move, some found it an important step towards ensuring free speech on the platforms. But, one only wonders what they will have to say now that Musk is selectively censoring content on the platform.","excerpt":"Twitter bans accounts of Koo and Mastodon for “violating” Twitter rules","categories":["AI News"],"tags":["free speech","Koo","Mastodon","Twitter (X)"],"author_name":"Ayush Jain","publish_date":"2022-12-16T15:21:02","publication_year":"2022","word_count":297,"keywords":["free speech","Mastodon","Go","programming_languages:R","AI","programming_languages:Go","GAN","CNN","Twitter (X)","R","Koo"],"extracted_tech_keywords":["AI","R","Go","GAN","CNN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/twitter-bans-koo\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":30361,"title":"Google’s Move To Open Source BERT May Change NLP Forever","content":"In 1954, with the success of the Georgetown experiment in which the scientists used a machine to translate random sentences from Russian to English, the field of computational linguistics took giant strides towards building an intelligent machine capable of recognising and translating speech. These models were even used in translations during the Nuremberg trials. Nonetheless, the future of machine translation was nowhere close to the forecast due to sluggish computational devices and scarcity of data to train on. Today, after six decades, machines have transitioned from mechanical statistical models to electronic neural models which can perform complicated tasks like speech recognition and sentiment analysis with great accuracy. The biggest challenge for NLP models, however, is the lack of training data. Small training sets restrict many NLP models from performing real-time rendering of both contextual and free from context tasks. Meet BERT Bidirectional  Encoder Representations from Transformers or BERT, which was open sourced earlier this month, offers a new ground to embattle the intricacies involved in understanding the language models. Pre-training a binarised prediction model helps understanding common NLP tasks like Question Answering or Natural language Inference. Unidirectional models are efficiently trained by predicting each word conditioned on the previous words in the sentence. However, it is not possible to train bidirectional models by simply conditioning each word on its previous and next words, since this would allow the word that’s being predicted to indirectly “see itself” in a multi-layer model. For the pre-training corpus, the researchers used the concatenation of BooksCorpus (800M words) and English Wikipedia (2,500M words). A sample text from the corpus is taken as two sentences. These sentences get [A] or [B] embedding as discussed above and they are sampled such that the combined length of tokens is less than or equal to 512. Text representation in BERT BERT uses WordPiece embeddings with a 30,000 token vocabulary and learned positional embeddings with supported sequence lengths up to 512 tokens. (Image source: BERT) The input embedding in BERT is the sum of token embeddings, segment and position embeddings. The sentence differentiation is done by separating it with a special token [SEP] and then add [A] embedding to the first sentence and [B] embedding to the second sentence in case of two sentences or only [A] embedding for single-sentence inputs. Pre-training lies at the core of BERT’s innovation. It doesn’t use the usual left-to-right or right-to-left language models (LM) instead it performs the following two tasks: Masked LM Next sentence prediction How BERT Uses Masking Since bidirectional conditioning would allow each word to indirectly “see itself” in a multi-layered context, masking is done to train deep bidirectional representation.For example, in the sentence Apples are green, the masked language model(MLM) performs the following procedure: Apples are [MASK] for 80% of the time Apples are [smart], a random word for 10% of the time And, Apples are [green] for 10% of the time. The [MASK] token replaces the selected word, which, in this case, is [green] for most of the time. And, also keeps the same word to nudge the model towards the original word. Since the transformer network doesn’t know which words it would be asked to predict, it is coerced into taking a note of every token representation that has been given as an input. Achieving this level of deep bidirectionality with MLM is the single most improved feature of BERT and this gives it an edge over other existing models. While choosing two sentences, say, [S1] and [S2]  for pre-training, S2 follows S1 for 50% of the time. The other half consists of other random sentences from the corpus. These ‘other’ sentences can be classified as ‘NotNext.’ S1: What colour are the apples? S2: Apples are red Classified as Isnext S1: What colour are the apples? S2: Apples are smart Classified as NotNext Pre-trained representations can either be context-free or contextual, and contextual representations can further be unidirectional or bidirectional. Usually, context-free models like word2vec generate single word representations for each word in the vocabulary. For example, the word “bark” would have the same context-free representation in “ a dog’s bark” and “bark of a tree.” Contextual models instead generate a representation of each word that is based on the other words in the sentence starting from the very bottom of a deep neural network, making it deeply bidirectional. Transformers Instead Of RNNs Or CNNs Neural networks usually process language by generating fixed-or-variable-length vector-space representations. After starting with representations of individual words or even pieces of words, they aggregate information from surrounding words to determine the meaning of a given bit of language in context. RNNs have in recent years become the typical network architecture for translation, processing language sequentially. The sequential nature makes difficult to fully harness parallel processing units like TPUs. Convolutional neural networks (CNNs), though less sequential, take relatively more number of steps to combine information. A Transformer network applies self-attention mechanism which scans through every word and appends attention scores(weights) to the words. For example, homonyms will be given higher scores for their ambiguity and these weights are used to calculate weighted average which gives a different representation of the same word. The output of the transformer network, which also happens to be the final hidden state is taken as the first token for the input and the probability of selecting a random label is calculated using standard softmax function. The same formula is used for the end of the answer span where the maximum scoring span is used as the prediction. Fine tuning BERT involves adding a simple classification layer to the pre-trained model, and all parameters are jointly fine-tuned on a downstream task. Also, the additional output layers eliminate the need to learn hyperparameters from scratch every single time. The fine-tuning and this high level of accuracy were only made possible because of a large number of training steps (128,000 words\/batch * 1,000,000 steps). BERT on Cloud TPU A batch size of 256 sequences means 256 sequences * 512 tokens = 128,000 tokens\/batch for 1,000,000 steps, which is approximately 40 epochs over the 3.3 billion word corpus. BERT boasts of training any question answering model under 30 minutes. Given the number of steps BERT operates on, this is quite remarkable. And, this was only made possible by Google’s custom-built cloud TPUs which can accelerate dense matrix multiplications and convolutions and, minimize the time-to-accuracy while training large models. Compounding on BERT Though the rate of convergence is low with MLM compared to the conventional left-to-right model, MLM outperforms other models with its high accuracy scores. Transfer learning with unsupervised pre-training forms the foundation of many natural language understanding systems. And, with a deep bi-directional architecture of BERT, the above findings seem more authentic. The transformer network helps to gain insights on how the information flows in the architecture. This study also helped researchers to demonstrate how sufficiently pre-trained models lead to improvements with trivial tasks when scaled to extremities. The next big challenge for these NLP models is to reach a human-level understanding of language which has been in the pursuit since the times of Leibniz and Descartes.","excerpt":"In 1954, with the success of the Georgetown experiment in which the scientists used a machine to translate random sentences from Russian to English, the field of computational linguistics took giant strides towards building an intelligent machine capable of recognising and translating speech. These models were even used in translations during the Nuremberg trials. Nonetheless, […]","categories":["Global Tech"],"tags":["BERT","Google","Google Translate","NLP"],"author_name":"Ram Sagar","publish_date":"2018-11-18T05:03:28","publication_year":"2018","word_count":1185,"keywords":["Go","TPU","Google Translate","AI","neural network","ML","sentiment analysis","Transformers","BERT","NLP","RAG","Google","R"],"extracted_tech_keywords":["AI","ML","neural network","NLP","Transformers","RAG","sentiment analysis","TPU","R","Go"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/googles-move-to-open-source-bert-may-change-nlp-forever\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10168437,"title":"HCLTech Shows Confidence in Generative AI with 12 Exclusive Deals in the Quarter","content":"Much on the lines of Indian IT firms that reported subdued Q4 FY25 results in the last fortnight, HCLTech, too, posted modest earnings while showing confidence in generative AI. In Q4 FY25, the revenue in rupee terms rose 1.2% to ₹30,246 crore from ₹29,890 crore. And in USD terms, the company posted a revenue of $3.4 billion, down 1.0% QoQ. For the full financial year, the revenue stood at ₹117,055 Crores, up 6.5%. Accounting in USD, the revenue was $13.8 billion, representing a 4.3% increase compared to the same period last year. Operating performance weakened, with EBIT dropping to ₹5,442 crore from ₹5,821 crore. The EBIT margin contracted to 18.1% from 19.5%. Meanwhile, net profit also declined 6.3% to ₹4,300 crore from ₹4,591 crore for the fourth quarter. In Q4, HCLTech focused on exclusive AI and generative AI deals, securing 12 new agreements, including those involving agentic AI and automation processes. This announcement differs from other IT firms that shied away from mentioning AI-specific deals. However, all the tech companies acknowledged that AI is part of every deal conversation. HCLTech CEO C Vijaykumar stated that Q4 FY25 saw the highest number of deals after the September 2023 quarter. This quarter, HCLTech closed $3 billion in net new bookings, bringing the year-to-date total to $9.26 billion. It’s a 5% decline compared to last year. “Our engineering and R&D services business led the charter with a record-high 75% growth in bookings in FY25,” Vijaykumar said, attributing this performance to the successful execution of HCLTech’s integrated go-to-market strategy. He also emphasised that generative AI will remain a core focus for enterprises across all industries, despite broader macroeconomic uncertainties and pressure on discretionary spending. “Their focus on using generative AI to drive high efficiency in every aspect of their business is becoming central to all conversations,” Vijaykumar said. On the topic of pricing pressure in the age of AI, he acknowledged that efficiencies delivered through AI would naturally lead to some level of pricing deflation. “When you’re able to deliver at a much more efficient operating level, we do share some of those benefits with our customers,” he explained. “But in every renewal and client conversation, we’re also able to ask for a higher wallet share as we proactively provide generative AI-driven benefits.” For FY26, HCLTech expects overall revenue growth in the range of 2–5% in constant currency (CC) terms. Services revenue is also projected to grow between 2–5% in CC terms. The company has guided for an EBIT margin of 18–19% for the fiscal. “In FY25, we clocked the consolidated revenue of $13.84 billion, an increase of 4.7% attributed to both our services business and the software business,” said Vijaykumar. He added that the pipeline spans IT and business services, engineering services, and HCLSoftware, and that AI and generative AI are now integral components of almost every deal. “Both Americas and Europe showed considerable pipeline growth during the quarter,” he noted. “We’ve made significant strides in AI and GenAI, impacting both client-facing solutions and internal operations.” The company’s four flagship AI offerings—AI Force, AI Foundry, AI Labs, and AI Engineering—have seen substantial adoption and scaling during FY25. Notably, AI Labs alone has delivered 500 GenAI engagements for 400 clients. The company’s attrition rate increased to 13%, up from 12.4% in Q4 of last year. The overall headcount stood at 223,420, with the tech firm adding 7,829 freshers in the fiscal year. Speaking further about AI, Vijaykumar said that both customers and the firm are internally running several AI use cases to build solutions around them. “The approach is: invest in skills development, invest in labs, build use cases, and on the back of it, create POCs that we can take to customers. That’s been our approach, and it’s working well so far,” he said. “Don’t worry about the investments now—look at the ROIs we can get on the back of it. That has been our approach. It takes time, but we’re seeing success.” Focus on generative AI Speaking at an industry event in Mumbai in February, Vijayakumar emphasised that AI’s disruption in IT services is unlike previous technological shifts such as cloud computing and digital transformation. “The changes AI is assuring are very different, and we need to be more proactive to categorise our revenues to create completely new businesses,” he said. Generative AI is expected to accelerate software development by automating coding and reducing project timelines. Vijayakumar pointed to a financial services firm where AI-driven efficiencies reduced the timeline of a $1 billion technology transformation program from five years to three-and-a-half years. He emphasised that, with declining costs of AI training, India must invest in economically viable ways to develop its own models. “I strongly believe that the business model is ripe for disruption. What we saw in the last 30 years was a fairly linear scaling of revenues and people. I think time is already up for that (business model),” he added. Meanwhile, in its Q3 FY25 results in January, HCLTech secured $2.1 billion in total contract value (TCV), with many deals embedding AI solutions. The CEO said that it is advancing its generative AI strategy and aims to integrate AI services into 100 clients by FY26. “Generative AI is getting more and more real. The cost of using an LLM or conversational model has dropped by over 85% since early 2023, making more use cases viable,” Vijayakumar noted in the Q3 earnings call. Compared to Q1 and Q2, the company’s Q3 results demonstrate a growing integration of AI into its business operations.","excerpt":"CEO C Vijaykumar didn’t disclose any specific vertical where generative AI deals were awarded, but said that it is part of almost all deals.","categories":["IT Services"],"tags":["Generative AI","HCL Technology"],"author_name":"Mohit Pandey","publish_date":"2025-04-22T19:26:56","publication_year":"2025","word_count":922,"keywords":["Go","GenAI","agentic AI","AI","cloud computing","digital transformation","Git","Aim","generative AI","HCL Technology","Generative AI","R"],"extracted_tech_keywords":["AI","generative AI","GenAI","agentic AI","Aim","cloud computing","R","Go","Git","digital transformation"],"url":"https:\/\/analyticsindiamag.com\/it-services\/hcltech-shows-confidence-in-generative-ai-with-12-exclusive-deals-in-the-quarter\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":50935,"title":"Hemant Misra From Swiggy Says Explainable AI May Be Overrated","content":"According to Hemant Misra, Head of Applied Reasearch at Swiggy, “If human decisions are biased, then why do we care if AI is a little biased in some cases?” We, for the most part, think that machines are more objective and fair than people. In any case, previously we have witnessed quite a few instances and contentions about AI-powered frameworks yielding one-sided or biased outcomes. For instance, in 2016, it was revealed that the AI systems in the US courts were bound to mark black respondents as “high risk” compared to white defendants from similar backgrounds. The fact that the AI algorithm was biased despite the fact that the framework wasn’t unequivocally given any information on litigants’ race may be shocking. It proved that AI was somehow racist based merely on the looks. The big question many think that lies is whether the net impact of ML-empowered frameworks is to make the world more pleasant and progressively effective or to amplify human bias to a bigger level. Hemant Misra From Swiggy On Why AI Bias Is Natural Numerous significant choices in our lives are made by frameworks or something to that effect, regardless of whether those frameworks comprise of individuals, machines, or a mix of both. Huge numbers of these current frameworks are biased in both evident and inconspicuous ways. The expanding job of ML in basic leadership frameworks can be controversial, especially if it is about making important decisions. There is another side to the bias story, humans themselves. But, some experts may argue that if the world is functioning despite the fact that everyone is biased for something, we can make the AI world also work in the same manner. One such expert is Hemant Misra, Head of Applied Research at the at popular food delivery app Swiggy, and an expert in speech recognition and natural language processing. “If AI is being used to judge court cases, or when it is used in healthcare in the future, it will be important for it not be biased. But, everybody is biased, then why there is so much problem in machines being biased. And there is a historical reason for any reason based on human interactions and the evolution of our civilisation. When we are biased, the data we are generating will also be biased. Now, on the other hand, we need to ask whether we are we okay as a society if a machine is in fact not biased as it has been part of evolution,” said Misra who leads Applied Research at Swiggy. Do We Need Explainability In AI? What fills the AI framework is the data it gained from. Along these lines, it uniquely in contrast to a standard software program where people expressly compose each line of code. The precision of an ML framework could be estimated by people, yet explainability into how such a framework really settles on decisions is restricted. This is the place Explainable Artificial Intelligence could come in. People may have the option to address for any bias before it has a genuine effect, in the event that they could monitor the “thinking” a calculation used to settle on choices about individuals from high-risk category. But, is explainable AI something we need? According to Hemant Misra- Head of Applied Research- Swiggy, while we need explainability in critical cases, otherwise may not be as relevant. “We know everyone is biased. So, How many of us go to employers and ask them to explain the rejections? How many of us go to the bank in case the bank rejects the application of a credit card, ask them to explain the decision. They have their own methodology. If human decisions are biased, then why do we care if AI is a little biased in some cases. In cases where it is a matter of life and death, of course yes explainability is key just like we want to understand why a doctor may take a particular decision that may impact a person’s life. Apart from that, it is overrated,” opined Misra, speaking at a recent ThoughtWorks Live event in Bengaluru.","excerpt":"According to Hemant Misra, Head of Applied Reasearch at Swiggy, “If human decisions are biased, then why do we care if AI is a little biased in some cases?” We, for the most part, think that machines are more objective and fair than people. In any case, previously we have witnessed quite a few instances […]","categories":["AI Features"],"tags":["Explainable AI","swiggy"],"author_name":"Vishal Chawla","publish_date":"2019-11-30T14:00:00","publication_year":"2019","word_count":685,"keywords":["Go","artificial intelligence","programming_languages:R","AI","ML","programming_languages:Go","explainable AI","GAN","Explainable AI","swiggy","R"],"extracted_tech_keywords":["AI","artificial intelligence","ML","R","Go","GAN","explainable AI","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/hemant-misra-swiggy-explainable-ai\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":44672,"title":"Should Engineer.Ai Be Blamed For Exaggerated Claims?","content":"Sachin Duggal, CEO of Engineer.Ai Recently, the Wall Street Journal published a report saying Engineer.AI, a startup that raised 29.5 million in 2018 for an automated app development platform exaggerated its use of artificial intelligence and uses developers instead of AI is used to achieve app automation. The news came as a shock for many and allegedly exposes how companies, especially startups in hot technology niches can somewhat be deceptive in their offerings. At least, that’s what WSJ is reporting. The report says it has talked to employees, both current and past, who have corroborated the fact that it’s true no AI is being used at the company and it may possibly be a sham just to acquire multi-million dollars worth of investments. The startup has now clarified that it leverages a price discovery system that uses Natural Language Processing using a Neural Network to make processes more efficient but humans still have to be used for writing code. Engineer.Ai Says It Never Claimed To Automate Full App Development While it is true Engineer.ai does not use AI to assemble code for applications, the company never actually claimed it did that in the first place. The company says its platform emphasises on automating the repetitive aspects of app development and and that’s where AI is put to use. The startup’s CEO Sachin Duggal has stated in a blog that it is not trying to automate full app development but rather automate all the repetitive parts of the entire process that are common, wasteful or otherwise inefficient. This it says is achieved through a man-machine partnership. “We are not selling an AI solution; rather something that is powered by Humans and AI. In the end, we build, run, and scale tailor-made software. We don’t claim to fully automate App Development. The fully autonomous development process (program synthesis) is the holy grail and is many years away. Rather, we rely on a Human and AI partnership, wherein the reusable software (in traditional software development, duplicated code makes up approximately 60% of the product) is machine produced and the other 40%,” Duggal stated. How Is The Company Able To Offer Custom App Development Without Any AI? Using Engineer.ai, individuals and companies can buy tailor-made software within short periods of time. So, if we allege that the company is solely using people for app development, then how is it able to help create custom software in such short spans? How has the company made about $60 million in revenue during its existence? The answer is because pricing and project timelines are done autonomously using AI-based systems. The startup leverages a price discovery system that uses Natural Language Processing using a Neural Network. This has been ignored in the media reports. “All pricing and timeline calculations are done automatically. The pricing system uses a combination of shallow two-layer neural networks that are run discreetly and asynchronously to update pricing. Though the calculation of a project price itself is algorithmic, the inputs to the formula are calculated by using the aforementioned neural networks. As an example, we built a solution for Fader Fort in such a rapid time that it would be virtually impossible to do it from scratch without any automation,” said Duggal . Automatic Software Development (Program Synthesis) Is Atleast 7 Years Away Engineer.Ai has further clarified that it will remain to be a human-assisted company aiming to make app development efficient using AI. Both humans and AI will continue to play a role in this process, it says. The company wrote that building an AI-powered company is very different to an AI solution company, and the Engineer.Ai is the former. Also, in the WSJ report, it was mentioned that the Engineer.Ai reported its platform is 80% finished, which the publication stated was untrue. “We never told any investor that 80% of our platform was built,” Engineer.Ai said. The company has further spoken a long-term timeline for developing full capabilities. “We realise ourselves that automatic software development (program synthesis) is at least 7 years away. We will continue to use AI to eventually automate all of the repetitive tasks in the build, run, and scale process. ” The company has also promised that it will showcase some of its early research in program synthesis. Investors Have Come Forward To Support Engineer.Ai You would think that Engineer.Ai would have faced a huge backlash by now from the people who gave them millions of dollars. It’s the opposite. The investors have come to support Engineer.Ai and have remained confident in the management’s vision. The comments from the investors should have been part of the ongoing media discussion about the startup. Here are those comments, given by the investors during email inquiries: “Jungle is a proud investor in Engineer.Ai and have a very close relationship with Sachin and the company. They have always been very honest about their use of human-assisted AI,” said Jungle Investors. “Growth in the AI space does not happen overnight. They (Engineer.Ai) have clear focus to build the right things at the right scale point and its level of diligence and work which gives us great confidence in the team and their future,” stated Deepcore, a subsidiary of Softbank. Overview AI can be difficult to scale. It requires time and loads of data in order to create the right models. Big banner names across the globe like Facebook, Google and Amazon rely on human skills to augment AI applications. Why? Because AI is not going to solve everything on its own. Human partnership with machines is equally crucial. That’s what needs to be expected out of startups that want to automate processes like code development.","excerpt":"Recently, the Wall Street Journal published a report saying Engineer.AI, a startup that raised 29.5 million in 2018 for an automated app development platform exaggerated its use of artificial intelligence and uses developers instead of AI is used to achieve app automation.  The news came as a shock for many and allegedly exposes how companies, […]","categories":["AI Features"],"tags":["AI (Artificial Intelligence)","AI engineers","Automation","Startups"],"author_name":"Vishal Chawla","publish_date":"2019-08-20T11:23:45","publication_year":"2019","word_count":942,"keywords":["Go","API","artificial intelligence","AI","neural network","AI engineers","ETL","Automation","RAG","automation","Aim","Startups","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","neural network","Aim","RAG","R","Go","API","ETL","automation"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/should-engineer-ai-be-blamed-for-exaggerated-claims\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10060807,"title":"How continuous resilience adds robustness to ML models","content":"AWS faced a string of outages in December 2021 –three in a span of two weeks to be exact. In the third outage, the cloud computing giant’s East Coast data centre in North Virginia went down. The disruption lasted for hours, affecting central and eastern United States, and threw Slack, Epic Games Store and Amazon websites into disarray. AWS later said it was looking at its “increased EC2 launch failures and networking connectivity issues.” AWS glitches are not a new phenomenon. Clients either stretch their architecture across multiple geographic regions or use multiple providers to work around cloud outages. However, the cost and effort involved in such measures can set the clients back. The best way to defend against sudden failures is to build resilient infrastructure. Of late, Amazon is committing a lot of resources to solve this costly problem. To deal with the pressures of sudden traffic on a global scale, live events and popular streaming premieres, Amazon Prime Video team employed a mix of machine learning and continuous resilience. Apart from this, the team also ensured failovers functioned seamlessly. Machine learning to forecast workload Workload forecasting with ML involves taking into account a few variables, including timelines for feature rollouts, long-term planning, marketing strategies, seasonalities, and customer metrics, to project the workload trajectory. According to Ali Jalali (applied scientist at Amazon), using classical time series models combined with deep learning helped them zero in on an optimal risk level to the area for the forecast. The team built predictive models to correctly forecast the spike in traffic for the popular Indian series, Mirzapur. The platform has now created a similarity engine using past Amazon events and feedback from social media and IMDB ratings to predict hype. The resiliency of a system can be tested against this expected hype. This helped the Prime Video team determine which data centres can handle different workloads based on their latency and availability. Machine learning for incident management Automating the response to an incident can massively reduce the time taken to mitigate the issue while also easing the process for engineers. The model will locate the likely error and troubleshoot it while searching for patterns of false alarms. An automated system well-fed with data will reduce such incidents over time. The goal isn’t to remove humans completely from incident management but to make their jobs easier. Chaos Engineering The DevOps team at Prime Video built ML models to determine workload demand and used chaos engineering to practice IT incident recovery. Chaos engineering entails injecting a little latency into a function in a controlled environment. Gradually more functions are injected with latency until the breaking point. The intention behind this is to observe how the system reacts during an unforeseen event and consequently help the device build greater self-belief to perform normally in the face of such events. To start with, a backup data centre must be checked for its failover reliability. More outages have resulted from IT and network problems than power issues in the recent past. There are three major components to reactive architecture: Elasticity:  Scaling applications as and when required.Responsive: This maintains a system that is always aware of its surroundings, making it more alert and cautious. Resilience: This ensures the system is robust and up-and-running come what may. Later, the Prime Video team came up with a resilience score. The score indicated the team’s preparedness to deal with failures, avoid downtime, accept failures as a norm, and design contingency plans. However, the resilience score is not a marker for the system’s performance but a report that helps the team understand how to prioritise.","excerpt":"The DevOps team at Prime Video built ML models to determine workload demand and used chaos engineering to practice IT incident recovery.","categories":["AI Features"],"tags":[],"author_name":"Poulomi Chatterjee","publish_date":"2022-02-17T11:00:00","publication_year":"2022","word_count":603,"keywords":["Go","machine learning","AWS","AI","cloud computing","ML","Ray","deep learning","DevOps","R"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","Ray","cloud computing","AWS","R","Go","DevOps"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-continuous-resilience-adds-robustness-to-ml-models\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10111550,"title":"OpenAI Uses Technique Created By Indian Developers","content":"OpenAI recently released new embedding models and API updates, introducing two additions to its lineup: a smaller and highly efficient text-embedding-3-small model, and a larger and more powerful text-embedding-3-large model. Interestingly, the technique behind the embedding has been developed by Indian developers Aditya Kusupati, a researcher at Google and Pratik Jain, a Senior Staff Research Scientist at Google who recently published a paper titled ‘Matryoshka Representation Learning.’ 🤯WOW🪆Matryoshka Representation Learning enables \"native support for shortening embs\" &\"very flexible usage\"Jokes aside, excited that @OpenAI serves MRL by default in v3 embedding API for retrieval & RAG!Other models & services should catch-up soon😄https:\/\/t.co\/JcmTH8w8mh pic.twitter.com\/omVe2qkW2t— Aditya Kusupati (@adityakusupati) January 26, 2024 “Wish OpenAI had referred to Matryoshka embeddings (or nested embeddings, as we call them in the paper and presentations) instead of avoiding any of the names we have mentioned in the paper,” expressed Prateek Jain, one of the authors of the paper on X. Later on, Owen Campbell-Moore, APIs PM at OpenAI, acknowledged that OpenAI did train on MRL. ‘Hey Prateek! We did train this based on MRL – I was responsible for the blog post, and it’s my mistake for not thinking or remembering to cite. We’re updating the blog post to add a citation now!’ Moore wrote on X.” OpenAI edited the blog post mentioning the contribution of Prateek Jain and Aditya Kusupati’s paper. 🪆Matryoshka Representation Learning (MRL) 🪆Hidden behind @OpenAI’s new embedding updates* is a cool embedding representation technique by @adityakusupati et al. that encodes information at coarse-to-fine granularities in a single vector. Short 🧵 on how it works.Different… pic.twitter.com\/74nIFEt7Hb— Jerry Liu (@jerryjliu0) January 26, 2024 What is MRL? MRL is a cool embedding representation technique that encodes information at coarse-to-fine granularities in a single vector. MRL trains a single high-dimensional vector to encapsulate information at different granularities, akin to nesting dolls. It draws inspiration from the Russian nesting dolls, Matryoshka, where smaller dolls are encased within larger ones. MRL adapts to various downstream tasks without modifying the original representation, saving computational resources and avoiding the need for separate models for each task.","excerpt":"MRL is a cool embedding representation technique that encodes information at coarse-to-fine granularities in a single vector.","categories":["AI News"],"tags":[],"author_name":"Siddharth Jindal","publish_date":"2024-01-29T14:52:32","publication_year":"2024","word_count":344,"keywords":["Go","API","AWS","OpenAI","AI","cloud_platforms:AWS","programming_languages:R","programming_languages:Go","RAG","R"],"extracted_tech_keywords":["AI","OpenAI","RAG","AWS","R","Go","API","cloud_platforms:AWS","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/openai-uses-technique-created-by-indian-developers\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10126293,"title":"‘Odyssey’ AI Built for Hollywood, Sora Can Wait","content":"Odyssey, dubbed as a Hollywood grade AI system, was recently unveiled by co-founder and CEO Oliver Cameron. The model looks to give visual control to people to allow them to tell a story exactly the way they have imagined. The co-founder believes their text-to-video platform will result in higher-quality movies, shows and video games. https:\/\/twitter.com\/olivercameron\/status\/1810335663197413406 Multiple AI Models When compared to OpenAI’s Sora or Google’s Veo, Odyssey has given the reins of control to the users who can direct the visuals as per their needs. Thereby, offering better customisation options for users. Infact, there are a number of AI-video generation platforms in the market. Odyssey’s capability is achieved by delving deeper than traditional text-to-visual models. Instead of using just one model that limits you to one input and one unchangeable output, Odyssey is using four generative models. These models give precise control over each main part of visual storytelling: creating detailed shapes, realistic materials, customisable lighting, and adjustable motion. Together, they let you quickly create scenes and shots exactly as a user envisions them. Furthermore, the company is also developing workflows for expert users, seamlessly integrating into current production methods used in Hollywood, gaming, and others. The company allows compatibility with established workflows in top-tier production, and even edit and export in multiple formats such as 3D file formats. Super Team The team behind Odyssey comprises Hollywood artists and AI researchers from emerging tech verticals including Cruise, Wayve, Waymo, Tesla, Meta and more. The artists have worked with big production names such as Dune, Godzilla, Avengers and others. In addition to Cameron, Jeff Hawke serves as the other co-founder and CTO of Odyssey. The company has raised $9M from investors from Y-Combinator, Google Ventures and many more.","excerpt":"The new text-to-video platform looks to compete with OpenAI’s Sora and other similar options.","categories":["AI News"],"tags":["AI","Google","Hollywood","OpenAI","Sora"],"author_name":"Vandana Nair","publish_date":"2024-07-09T18:35:20","publication_year":"2024","word_count":287,"keywords":["Go","TPU","OpenAI","AI","Hollywood","programming_languages:R","ML","programming_languages:Go","Google","Sora","AI research","R"],"extracted_tech_keywords":["AI","ML","OpenAI","TPU","R","Go","AI research","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/odyssey-ai-built-for-hollywood-sora-can-wait\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10172684,"title":"Telangana Launches TGDeX—India’s First State‑Led AI Public Infrastructure","content":"Telangana government, in collaboration with the Japan International Cooperation Agency (JICA), is launching the Telangana Data Exchange (TGDeX) platform, the first state-led digital public infrastructure for artificial intelligence (AI) on July 2. The Emerging Technologies Wing (ETW) of Telangana’s IT department has announced that the TGDeX was developed by the Indian Institute of Science, Bengaluru, with strategic and technical guidance from JICA’s DXLab. The platform will enable AI developers to securely access high‑quality data and collaborative tools, fostering cross‑domain AI innovation. TGDeX will empower government data officers to share information across ministries and departments securely. It’s also open to private enterprises and NGOs that utilise AI for business or social impact in fields like education, healthcare, climate, and sustainability. Moreover, academic institutions, researchers, and think tanks can access high-quality datasets on the platform to fuel their research, analysis, and AI model development. While the platform serves multiple stakeholders, its primary focus is on AI start-ups and entrepreneurs at various stages of their AI journey. Startups will be able to access domain‑specific datasets crucial for designing and scaling AI solutions. Through TGDeX, they’ll also benefit from value‑added services provided by incubator and accelerator partners like Telangana AI Mission (T‑AIM) and T‑Hub, creating a dynamic ecosystem that supports AI innovation. The platform is a central element of the state’s AI strategy and roadmap. IT and industries minister Duddilla Sridhar Babu first introduced this plan during the Global AI Summit held in September 2024. The state government’s technology ecosystem, including T‑AIM, T‑Hub, MATH (the AI\/ML centre of excellence at T‑Hub), IIT Hyderabad and IIIT Hyderabad, will work together to support TGDeX in its objective to host over 2,000 AI‑ready datasets in the next five years.","excerpt":"TGDeX fosters cross‑domain AI collaboration among the state, startups, industry and research.","categories":["AI News"],"tags":["AI platform","telangana"],"author_name":"Smruti S","publish_date":"2025-07-01T17:56:06","publication_year":"2025","word_count":282,"keywords":["Go","artificial intelligence","AI","telangana","ML","innovation","Git","AI platform","Aim","GAN","R","startup"],"extracted_tech_keywords":["AI","artificial intelligence","ML","Aim","R","Go","Git","GAN","innovation","startup"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/telangana-launches-tgdex-indias-first-state%e2%80%91led-ai-public-infrastructure\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10044146,"title":"Vertica Launches Version 11: Major Features &#038; Upgrades Introduced","content":"US-based core analytical platform Vertica has announced the release of Vertica 11 Analytics Platform at Vertica Unify 2021. The platform has major features and enhancements to deliver unified analytics and machine learning across multi-cloud and multi-regional deployments with self-service container workflows to meet the agility, speed, and security requirements of analytic driven organisations. “Unified Analytics is a critical movement in our industry. But truly unified analytics requires proven and mature security, true deployment choice, end-to-end machine learning in production, and no-compromise analytical performance for organisations to capitalise on this megatrend,” said Colin Mahony, Senior Vice President and General Manager, Vertica, Micro Focus. “In Vertica 11, we expanded Vertica in Eon Mode to the Azure cloud, delivered support for Docker containers and Kubernetes, extended our market lead in advanced analytics and machine learning including time series forecasting, and much more. The feature list goes on and on – Vertica 11 is truly the unified analytics platform with the fastest performance at unlimited scale,” Colin added. Vertica 11 includes more in-database machine learning capabilities with the latest release of VerticaPy, an open-source Python library for Vertica that supports Python projects on data stored in Vertica. VerticaPy contains expanded machine learning functionality, connection, and data exploration capabilities as well as graphical capabilities. The release offers a new, open-sourced Apache Spark Connector that supports Spark 3.0 and Scala 2.12 with S3, SSO, and parallel read\/write support, significantly improving performance. Additional key capabilities include an XG Boost algorithm, increased PMML integrations, and customised time series algorithms.","excerpt":"Vertica 11 includes more in-database machine learning capabilities with the latest release of VerticaPy.","categories":["AI News"],"tags":[],"author_name":"kumar Gandharv","publish_date":"2021-07-22T12:26:54","publication_year":"2021","word_count":251,"keywords":["machine learning","AI","R","ML","docker","Apache Spark","Python","analytics","Azure","kubernetes"],"extracted_tech_keywords":["AI","machine learning","ML","analytics","Azure","kubernetes","docker","Apache Spark","Python","R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/vertica-launches-version-11-major-features-upgrades-introduced\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":62023,"title":"How To Get The Most Out Of A Virtual Conference","content":"With lockdown enforced across the world, virtual conferences have been gaining popularity, and online events are gaining prominence in the industry. Virtual conferences have a lot of benefits for attendees — including the ability to attend the conference at the comfort of your home and access to multiple sessions at the same time. Virtual conferences could be a revolution for the industry at the time of the lockdown crisis; however, the widespread adoption is still relatively new. This is because virtual conferencing can create several challenges for attendees. Here are a few ways that can help you get the most out of virtual conferences and online events: Create A Mindset To Learn Virtual conferences are no different than in-person events. And therefore, to get the most out of it, attendees need to have a mindset and willingness to learn from the sessions. Attendees should treat virtual conferences like in-person events, and clear their schedule for the day to get the best possible experience. Your takeaway from the conference will depend on how much you are willing to put in, engage with the content, and use the chatbox to interact at the end of presentations. You should use the information that was sent out in advance to take notes, highlight, and formulate questions to make sure that you are getting the most out of it. Focus Is The Key Although virtual conferences do not require us to move out of our comfort zones, it comes with the big challenge of finding ways to focus and dealing with distractions around us. To get the most out of your investment from an online event, attendees must prioritise their work, schedule them accordingly and should take out time exclusively to attend the sessions. Attendees should also turn off their phones, close emails, and cut all distractions that could hamper their focus on the conference. Keep Your Devices & Networks Ready Attendees need to be well aware and familiar with the platform being used for the conference and ensure that it works properly on their computer. Make sure to run all updates on your computer in advance and log in early to check your connection, camera, and microphone that you will be using during the sessions. In a conventional in-person environment, communicating with speakers or other attendees isn’t as challenging as during virtual conferences. For getting the most out of virtual conferences and online events, you have to ensure that your devices and networks are as prepared as you, in order to, not have technical hiccups during sessions or chats. Involve Your Peers Virtual conferences allow attendees to bring in there colleagues and peers to join in the sessions and create a group learning. Having your peers and colleagues involved in your conference experience can boost motivation, and can help in understanding complex concepts better by discussing it with your fellow attendees. Alongside it also increases your attention span, and provides opportunities for attendees to share critical takeaways as a group at the end of the conference, which will, in turn, help in revising the concepts and reinforce learning. Record Sessions For Replays Lastly, one of the significant benefits of virtual conferences is the ability to record and replay them. If the attendees aren’t able to attend a live session or wish to re-access any session post the conferences, virtual conferences provide them with the opportunity to record these sessions. Such an ability also gives an enormous opportunity for attendees from a learning perspective — one can avail the maximum amount of knowledge even after the session is over. Additionally, in virtual conferences and online events, attendees can view and download speaker presentations and can also rate a speaker, bookmark them, and take notes simultaneously. Attend plugin, and experience the next-generation biggest online conference for developers and leaders in the data science and machine learning space from the comfort of your home.","excerpt":"With lockdown enforced across the world, virtual conferences have been gaining popularity, and online events are gaining prominence in the industry. Virtual conferences have a lot of benefits for attendees — including the ability to attend the conference at the comfort of your home and access to multiple sessions at the same time. Virtual conferences […]","categories":["AI Features"],"tags":["conference","conferences","data science conferences","online ai conference","tools for virtual conferencing","virtual AI conference"],"author_name":"Sejuti Das","publish_date":"2020-04-20T15:30:00","publication_year":"2020","word_count":647,"keywords":["tools for virtual conferencing","data science conferences","data science","machine learning","programming_languages:R","AI","conference","conferences","online ai conference","virtual AI conference","R"],"extracted_tech_keywords":["AI","machine learning","data science","R","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-to-get-the-most-out-of-a-virtual-conference\/","complexity_score":2,"technical_depth":5,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":41338,"title":"Meetup: NVIDIA GPU For End-To-End Machine Learning Acceleration Workshop","content":"GPU computing has become one of the most important elements of AI infrastructure today. Owing to inherent architectural advantages, GPUs are well-suited to accelerate Deep Learning tasks. NVIDIA provides a scalable compute infrastructure, ranging from a single GPU in workstations to server-scale end-to-end data centre solutions. NVIDIA’s presence in the market is undisputed. In addition to hardware for parallel procession, NVIDIA provides a cloud based end to end accelerated stack of software, which includes CUDA, CUDA X libraries, drivers, frameworks and much more. GPU-accelerated computing is helping to solve some of the most complex problems in the world. The AI stack is cutting edge, and it’s evolving all the time, leading to the requirement of a large amount of work to set up and maintain an AI software environment. NVIDIA GPU Cloud (NGC) provides researchers and data scientists with simple access to a comprehensive catalog of GPU-optimized software tools for deep learning, machine learning and high-performance computing. NGC provides a range of options that meet the needs of data scientists, developers, and researchers with various levels of AI expertise. To apply for the workshop, click here. The platform allows researchers to deploy a comprehensive deep learning infrastructure quickly and effortlessly, in easy-to-use, dockerised containers. Advancements such as the creation of tensor cores and the easily-programmable CUDA interface have only served to increase the effectiveness of NVIDIA GPUs for general purpose computing. Analytics India Magazine has teamed up with NVIDIA to bring to data scientists a workshop focused on tools they need for GPU acceleration. Attendees will be able to learn how to implement GPU acceleration in everyday ML and DL tasks. To apply for the workshop, click here. Titled ‘NVIDIA GPU For End-To-End Machine Learning Acceleration Workshop’, the workshop will offer industry insights on how to make your everyday deep learning tasks more efficient. The workshop will include a keynote conducted by Sundara Ramalingam Nagalingam, Head, Deep Learning Practice, NVIDIA Graphics Pvt Ltd. The workshop will take the participants through the following topics: Accelerated Data Analytics for Better Insight & Use Cases RAPIDS Deep Dive Accelerating Data Science End-to-End with GPU & Getting Started with NVIDIA GPU Cloud Data ETL Pipelining Hands-on with cuDF XGBoost on MultiGPU Demo and Discussion Running other Algorithms on GPU Hands-on Q&A Session To apply for the workshop, click here. Who Should Attend? Data Engineers & Data Scientists looking to supercharge their training and inference workflows Data Science managers looking for an upgrade to existing infrastructure AI\/ML enthusiasts with experience in basic concepts of ML, data science, workflows and have worked with Python, Scikit-learn, or Pandas Notice: Only selected candidates will be able to attend this workshop. This is an application form. To apply for the workshop, click here. Time: 9 AM Venue: L-6, 8th Floor, NVIDIA, Manyata Tech Park, Nagawara, Bengaluru Date: 10th July 2019","excerpt":"GPU computing has become one of the most important elements of AI infrastructure today. Owing to inherent architectural advantages, GPUs are well-suited to accelerate Deep Learning tasks. NVIDIA provides a scalable compute infrastructure, ranging from a single GPU in workstations to server-scale end-to-end data centre solutions. NVIDIA’s presence in the market is undisputed. In addition […]","categories":["Deep Tech"],"tags":["cuda","machine learning gpu","ML","NVIDIA"],"author_name":"Anirudh VK","publish_date":"2019-06-26T05:44:28","publication_year":"2019","word_count":471,"keywords":["data science","scikit-learn","Rapids","cuda","machine learning","AI","ML","machine learning gpu","deep learning","XGBoost","analytics","NVIDIA","Pandas"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","data science","analytics","scikit-learn","XGBoost","Pandas","Rapids"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/meetup-nvidia-gpu-for-end-to-end-machine-learning-acceleration-workshop\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10048450,"title":"NVIDIA Releases Latest Kaolin Library: What’s New","content":"NVIDIA has released a new Kaolin library that will allow researchers to simplify and accelerate workflows. The new library includes a new representation, structured point clouds (SPC), a sparse octree-based acceleration data structure with highly efficient convolution, and ray-tracing capabilities. SPCs are very popular in 3D deep learning research today. It helps in scaling up and accelerating implicit neural representations. It is also behind the latest version of NeuralLOD training and delivers up to 30x reduction in memory and speeds up training time by a factor of 3. NVIDIA Kaolin New Release NVIDIA introduced the Kaolin library in 2019, originally intended for the NVIDIA Toronto AI lab as an internship project. Before that, researchers lacked the utility to develop 3D models for use with deep learning tools; they were forced to write lines of repetitive code and copy the algorithmic components for several projects. The researchers then started developing a PyTorch library that could bring common functionality for 3D deep learning; this led to the development of Kaolin, a software library that supports all types of 3D applications. Since its release, this library has grown into a codebase with optimised utilities and algorithms for 3D deep learning. Some of Kaolin’s applications include simplification of complex 3D dataset processing to be used for training. Kaolin also provides building blocks for conversions between 3D representations, useful 3D loss functions for training, and differentiable rendering. The new Kaolin library release also includes a lightweight Tensorboard-style web dashboard called Dash3D visualiser that allows visualisation of local and remote checkpoints without the use of specialised hardware or applications. This tool is combined with the latest builds like a command-line utility. Users can leverage this tool to inspect checkpoints of 3D predictions produced by deep learning models during training and on remote hardware configurations. Credit: NVIDIA The new library release also improves support for 3D datasets and new datasets like SHREC and ModelNet, speedups for USD 3D file format that enables a 5x improvement in the load time efficiency during training. It also includes new tutorials for differentiable rendering and 3D checkpoints. Omniverse Kaolin App Earlier this year,  NVIDIA made its Omniverse Kaolin App available for 3D Deep Learning Researchers. The Omniverse platform offers researchers the ability to collaborate virtually and work across different software applications. The Omniverse Kaolin app is an interactive application, which along with the NVIDIA Kaolin library, helps 3D deep learning researchers in accelerating their process. This app leverages the Omniverse platform, USD format, and RTX rendering to provide interactive tools. These tools allow the visualisation of 3D outputs of any deep learning model as it trains and inspects 3D datasets to find inconsistencies and gain intuitive insights from collections of 3D data. Need For Kaolin Deep learning is one of the most rapidly growing areas of research with applications in areas such as self-driving vehicles, autonomous robots, 3D graphics and games, augmented reality, and virtual reality. Unlike 2D, 3D data has more parameters and features and is more complex. Collecting and transforming 3D data from one representation to another is a tedious task and is more time consuming and error-prone than 2D computer vision. Recently, many better performing models, datasets, metrics, visualisation tools, and graphic tools have been introduced in recent years. Integrating the approaches is still a non-trivial job for researchers and practitioners. To this end, Kaolin proves to be an efficient tool for the manipulation of 3D content. It can wrap into PyTorch tensors 3D datasets implemented as polygon meshes, point clouds, voxel grids, etc. The interface provides a repository of baseline and state-of-the-art models for classification, 3D reconstruction, segmentation, super-resolution, etc.","excerpt":"NVIDIA introduced the Kaolin library in 2019 and was originally an internship project and intended for the NVIDIA Toronto AI lab.","categories":["AI Trends"],"tags":["NVIDIA"],"author_name":"Shraddha Goled","publish_date":"2021-09-16T15:00:00","publication_year":"2021","word_count":603,"keywords":["Go","API","TPU","AI","PyTorch","computer vision","RAG","Ray","deep learning","NVIDIA","R"],"extracted_tech_keywords":["AI","deep learning","computer vision","Ray","PyTorch","RAG","TPU","R","Go","API"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/nvidia-releases-latest-kaolin-library-whats-new\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10119017,"title":"Six Months Old Cognition Labs Raises $175 Mn from Founders Fund at $2 Bn Valuation","content":"San Francisco-based AI startup Cognition Labs has raised $175 million in a funding round led by Founders Fund, according to a report by The Information. The six-month-old company, which recently launched an AI-powered coding assistant called Devin, is now valued at $2 billion. With this latest fund, the company is most likely to use it to fuel further development, alongside enhancing its product. Cognition’s latest funding comes just a month after Founders Fund led the startup’s Series A round at a $350 million valuation. The rapid increase in valuation is in line with the growing interest in AI-powered tools that can assist or even automate software development tasks. Devin, Cognition’s AI software engineer, was shown to handle entire development projects independently, potentially reducing the need for human developers on certain tasks. The tool’s launch in March went viral on social media, but some users on social media questioned the company’s claims about Devin’s capabilities. Despite the skepticism surrounding Devin’s launch, the AI coding assistant has shown promising results. According to the SWE-Bench benchmark, which evaluates AI models on software engineering tasks, Devin achieved a 13.86% accuracy in resolving issues unassisted, surpassing the previous best model’s 1.96% unassisted accuracy. The company, founded in November 2023 by coding wizards, Scott Wu, Walden Yan, and Steven Hao, is one of several players in the growing field of AI-assisted software development. Unlike their competitors, GitHub Copilot, Amazon Code Whisperer or Replit which assist developers by providing code snippets and recommendations, Devin works as an agent and handles the entire projects independently. The company had previously turned down offers that would have valued it at $1 billion, according to sources familiar with the matter. This trend of AI startups attracting massive investments is evident in other recent funding rounds. Perplexity, an AI search startup challenging Google Search, also secured funding of $63 million at a billion dollar valuation this week. Similarly, Mistral, a French AI startup founded just over a year ago, reached a $2 billion valuation in December.","excerpt":"The six-month-old company, which recently launched an AI-powered coding assistant called Devin, is now valued at $2 billion.","categories":["AI News"],"tags":["cognition labs","Fund Raising"],"author_name":"K L Krithika","publish_date":"2024-04-24T23:00:44","publication_year":"2024","word_count":334,"keywords":["Go","API","funding","AI","RPA","Fund Raising","Git","Aim","cognition labs","GitHub","R","startup"],"extracted_tech_keywords":["AI","Aim","R","Go","Git","GitHub","API","RPA","startup","funding"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/six-months-old-cognition-labs-raises-175-mn-from-founders-fund-at-2-bn-valuation\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":35788,"title":"Machine Learning Webinar: Understand The Fundamentals Of Building ML Applications","content":"Enterprises are increasingly relying on machine learning to develop mission-critical applications to power their businesses. However, there are a series of challenges one has to tackle in the machine learning application building lifecycle — right from combining multiple data sources to feature engineering which means finding the right parameters to extract from the data for model building. So, what’s the best approach to AI\/ML applications and how does one identify the data sources to build models. With demand for machine learning increasing, it’s up to the data scientists and developers to stay updated with the latest trends to get the best performance out of their applications. With machine learning becoming a critical enabler for every sector, this webinar is specifically aimed at machine learning engineers and data scientists to help them understand the fundamental concepts of ML and help them leverage cloud machine learning tools. During the webinar, participants will learn how to select an algorithm, build a ML model and deploy it effectively, without diving too deep into the mathematical concepts of the model. Great Learning presents a webinar on Introduction to Machine Learning by Dr Kumar Muthuraman,Faculty Director, Center for Research and Analytics, University of Texas at Austin who will take the participants through the following areas: Introduction to Machine Learning How Machine Learning is changing the world around us Case study on how Machine Learning is used to solve a business problem Q&A with the speakers Who Should Attend: Data Scientists who wish to speed up and work on machine learning use cases Software Engineers and machine learning enthusiasts who want to gain deeper insights into machine learning frameworks ML Engineers and enthusiasts who want to understand the machine learning application development lifecycle Industry members who want to learn how to leverage ML to meet business objectives The webinar is presented by Dr. Kumar, who is the H. Timothy (Tim) Harkins Centennial Professor in the Department of Information, Risk and Operations Management and the Department of Finance at McCombs School of Business, University of Texas at Austin. Dr Kumar received his Ph.D. from Stanford University and his research focuses on decision making under uncertainty. Application areas of interest to him are quantitative finance, operations management and health care When: March 17 at 11 am To register click here","excerpt":"Enterprises are increasingly relying on machine learning to develop mission-critical applications to power their businesses. However, there are a series of challenges one has to tackle in the machine learning application building lifecycle — right from combining multiple data sources to feature engineering which means finding the right parameters to extract from the data for […]","categories":["Deep Tech"],"tags":["Great learning"],"author_name":"Richa Bhatia","publish_date":"2019-03-06T04:45:45","publication_year":"2019","word_count":381,"keywords":["Go","machine learning","programming_languages:R","AI","ML","feature engineering","RAG","Aim","analytics","Great learning","R"],"extracted_tech_keywords":["AI","machine learning","ML","analytics","Aim","RAG","R","Go","feature engineering","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/machine-learning-webinar-understand-the-fundamentals-of-building-ml-applications\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10023058,"title":"Guide To ResNeSt: A Better ResNet With The Same Costs","content":"Convolution neural networks have largely dominated the computer vision domain, but in the last few years, feature-map attention architectures like SE-Net and SK-Net have started to assert dominance. In their paper “ResNeSt: Split-Attention Networks”,  Hang Zhang,  Chongruo Wu, et al. proposed a new ResNet variant that combines the best of both worlds. The ResNeSt architecture leverages the channel-wise attention with multi-path representation into a single unified Split-Attention block. It learns cross-channel feature correlations while preserving independent representation in the meta structure. Architecture & Approach Split-Attention block ResNeSt introduces the Split-Attention block, which enables feature-map attention across different feature-map groups. This Split-Attention block consists of a feature-map group and split attention operations. Like the ResNeXt block, the feature is divided into several groups, and the number of feature-map groups controlled by a cardinality hyperparameter K. ResNeSt adds a new radix hyperparameter R that indicates the number of splits within a cardinal group, so the total number of feature groups is G = KR. A combined representation for each cardinal group is obtained by fusing with an element-wise summation across multiple splits. The kth cardinal group is represented as: Here U k ∈ R H×W×C\/K  for k ∈ 1, 2, …K, and H, W and C are the block output feature-map sizes. Global contextual information with embedded channel-wise statistics is gathered with global average pooling across spatial dimensions sk. The cth component is calculated as: A weighted aggregation of the cardinal group representation Vk ∈ RH×W×C\/K is done using channel-wise soft attention. The cth channel is then calculated as: Here aki(c) denotes the soft assignment weight given by: The mapping Gci determines the weight of each split for channel c based on the global context representation. Cardinal group representations are then concatenated along the channel dimension: V = Concat{V1 , V2 , …VK}. The final output Y of the Split-Attention block is produced using a shortcut connection: Y = V + X. Radix-Major Implementation This layout where the feature-map groups with the same cardinal index reside next to each other physically is called the cardinality-major implementation. It is straightforward but is difficult to modularize and accelerate using standard CNN operators. For this, an equivalent radix-major implementation is used. In this layout, the input feature-map is first divided into RK groups, in which each group has a cardinality-index and radix-index. The groups with the same radix-index reside next to each other. Then a summation across different splits is done so that the feature-map groups with the same cardinality-index but different radix-index are fused. A global pooling layer aggregates over the spatial dimension while keeping the channel dimension separated. This is equivalent to conducting global pooling to each individual cardinal group then concatenating the results. Two consecutive fully connected layers with the number of groups equal to cardinality are added after the pooling layer to predict each split’s attention weights. The use of grouped FC layers makes it identical to apply each pair of FCs separately on top of each cardinal group. In this implementation, the first 1 × 1 convolutional layers can be unified into one layer. The 3 × 3 convolutional layers can be implemented using a single grouped convolution with the number of groups of RK. Therefore, the ResNeSt Split-Attention block is now modularized using standard CNN operators. Network Tweaks It is essential to preserve spatial information for transfer learning on dense prediction tasks such as detection or segmentation. Generally, ResNet implementations apply the strided convolution at the 3×3 layer instead of the 1×1 layer to preserve such information. Convolutional layers require handling feature-map boundaries with zero-padding strategies. This is suboptimal when transferring to other dense prediction tasks. Instead of using strided convolution at the transitioning block, ResNeSt uses an average pooling layer with a kernel size of 3×3.  It also adopts two modifications introduced by ResNet-D, ResNeSt replaces the 7×7 convolutional layer with three consecutive 3×3 convolutional layers and adds a 2×2 average pooling layer to the shortcut connection before the 1 × 1 convolutional layer for the transitioning blocks with a stride of two. Image Classification with ResNeSt Install PyTorch and fvcore. pip install torch pip install fvcore Download the pre-trained ResNeSt model from Torch hub and set it eval mode for making inferences. model = torch.hub.load('zhanghang1989\/ResNeSt', 'resnest50', pretrained=True) model.eval() Get the image(s) for making inferences. You can get sample ImageNet images here.filename = \"n01491361_tiger_shark.jfif\"Download the ImageNet class labels wget https:\/\/raw.githubusercontent.com\/pytorch\/hub\/master\/imagenet_classes.txt Read the image, process it to match the input specification of ResNeSt and create a mini-batch for inference. from PIL import Image from torchvision import transforms input_image = Image.open(filename) preprocess = transforms.Compose([ transforms.Resize(256), transforms.CenterCrop(224), transforms.ToTensor(), transforms.Normalize(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225]), ]) input_tensor = preprocess(input_image) input_batch = input_tensor.unsqueeze(0) Switch to GPU if available, and make the inference(s). if torch.cuda.is_available(): input_batch = input_batch.to('cuda') model.to('cuda') with torch.no_grad(): output = model(input_batch) Apply softmax on the confidence scores of the 1000 ImageNet classes to get the probabilities. Print the top 3(k) categories. probabilities = torch.nn.functional.softmax(output[0], dim=0) with open(\"imagenet_classes.txt\", \"r\") as f: categories = [s.strip() for s in f.readlines()] top3_prob, top3_catid = torch.topk(probabilities, 3) for i in range(top3_prob.size(0)): print(categories[top3_catid[i]], top3_prob[i].item()) Colab Notebook, refer to the above implementation. References GitHub Paper","excerpt":"ResNeSt architecture combines the channel-wise attention with multi-path representation into a single unified Split-Attention block.","categories":["Deep Tech"],"tags":["Guide","resnet","ResNet50"],"author_name":"Aditya Singh","publish_date":"2021-03-30T16:00:00","publication_year":"2021","word_count":857,"keywords":["CUDA","ResNet50","Go","TPU","AI","neural network","PyTorch","computer vision","RAG","Colab","resnet","R","Guide"],"extracted_tech_keywords":["AI","neural network","computer vision","PyTorch","Colab","RAG","TPU","CUDA","R","Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/guide-to-resnest-a-better-resnet-with-the-same-costs\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":51169,"title":"Sundar Pichai Takes Over Alphabet As Google Founders Larry Page &#038; Sergey Brin Step Down","content":"Google CEO Sundar Pichai will replace Larry Page as the head of parent company Alphabet Inc, as founders Page and Sergey Brin step down from active roles. Pichai will remain the CEO of Google, and will assume the role of managing Alphabet’s investment in its portfolio of Other Bets. He will also remain a member of Alphabet’s Board of Directors. Page and Brin wrote that as Alphabet was now well-established, and Google and the Other Bets operating effectively as independent companies, it was the natural time to simplify their management structure. “Alphabet and Google no longer need two CEOs and a President. Going forward, Sundar [Pichai] will be the CEO of both Google and Alphabet. He will be the executive responsible and accountable for leading Google, and managing Alphabet’s investment in our portfolio of Other Bets.” I’m excited about Alphabet’s long term focus on tackling big challenges through technology. Thanks to Larry & Sergey, we have a timeless mission, enduring values and a culture of collaboration & exploration – a strong foundation we’ll continue to build on https:\/\/t.co\/tSVsaj4FsR— Sundar Pichai (@sundarpichai) December 4, 2019 They also added that they are still “deeply committed” to Google and Alphabet for the long term, and would remain actively involved as Board members, shareholders and co-founders. John Hennessy, Chairman of Alphabet’s Board of Directors, said: “It’s impossible to overstate Larry and Sergey’s contributions over the past 21 years. I’m grateful that they will continue their involvement on the Board.” Pichai said: “I’m excited about Alphabet and its long term focus on tackling big challenges through technology. I’m looking forward to continuing to work with Larry and Sergey in our new roles. Thanks to them, we have a timeless mission, enduring values, and a culture of collaboration and exploration. It’s a strong foundation on which we will continue to build.”","excerpt":"Google CEO Sundar Pichai will replace Larry Page as the head of parent company Alphabet Inc, as founders Page and Sergey Brin step down from active roles. Pichai will remain the CEO of Google, and will assume the role of managing Alphabet’s investment in its portfolio of Other Bets. He will also remain a member […]","categories":["AI News"],"tags":["Alphabet","Sundar Pichai"],"author_name":"Prajakta Hebbar","publish_date":"2019-12-05T13:10:04","publication_year":"2019","word_count":305,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","Alphabet","R","Sundar Pichai"],"extracted_tech_keywords":["AI","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/sundar-pichai-takes-over-alphabet-as-google-founders-larry-page-sergey-brin-step-down\/","complexity_score":2,"technical_depth":5,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10124844,"title":"Farewell, ChatGPT!","content":"Anthropic’s latest model, Claude 3.5 Sonnet, has dethroned OpenAI’s GPT-4o and secured the top spot in the Coding Arena and Hard Prompts Arena. It was also placed second on the Overall leaderboard. “Claude is so much better at coding than ChatGPT. Night and day difference (sic),” wrote a user on X. The new Sonnet has surpassed Claude Opus at five times the lower cost and is competitive with frontier models like GPT-4o and Gemini 1.5 Pro across the board. However, the best is yet to come. Anthropic plans to release Claude 3.5 Haiku and Claude 3.5 Opus later this year, along with developing new features like Memory for personalised user experiences. Play with Artifacts The standout feature in Claude 3.5 Sonnet that has captured everyone’s attention is the Artifacts tool. It appears as a separate window in the corner, assisting users in visualising their tasks. When users request Claude to generate content such as code snippets, text documents, or website designs, the Artifacts feature displays these creations in a dedicated window within their conversation interface. Currently, OpenAI’s ChatGPT does not boast of any such feature, nor has it announced anything similar. This setup creates a dynamic workspace, allowing users to view, edit, and enhance Claude’s outputs in real-time. It seamlessly integrates AI generated content into their projects and workflows. Since its release, many users have been experimenting with the model to create games, websites, functional AI sound effects, and simulations. https:\/\/twitter.com\/tarekbadrsh\/status\/1804309534858514889 “Wow! Claude 3.5 Sonnet with Artifacts will revolutionise learning! I asked Claude to create an animation simulating projectile motion, and the result blew me away!  Imagine generating custom animations and illustrations for any topic in minutes. Goodbye, static textbooks! @brilliantorg should be scared! ” wrote a user on X. https:\/\/twitter.com\/satvikps\/status\/1804778598047350828 Similarly, AI investor Allie K Miller used Claude 3.5 Sonnet to create an interactive educational tool to teach her ‘daughter’ how AI is used to communicate with animals. Anthropic has also introduced the ‘Projects’ feature in Claude where users can now organise chats with Claude into shareable projects. Each project includes a 200K context window, equivalent to a 500-page book, allowing users to incorporate all relevant documents, code, and insights to enhance Claude’s effectiveness. Moreover, users can set custom instructions within each project to further tailor Claude’s responses. Codes Like Flash “Claude 3.5 Sonnet is better at writing code than a typical computer science grad,” said former Stability AI chief Emad Mostaque. Source: X When equipped with the necessary tools, Claude 3.5 Sonnet can autonomously write, edit, and execute code, showcasing advanced reasoning and troubleshooting abilities. It excels at code translations, making it especially useful for updating legacy applications and migrating codebases. this multiplayer coding chat prototype suddenly became super fun with release of Claude Sonnet 3.5– made oauth logins work (still haven't tested with anyone!!)– works best w\/ p5 \/ threejs \/ music apps rn– dies after 4 sketches– wanted excuse to make this cool video: pic.twitter.com\/I5uUL0MlCp— Brian Jordan (@bcjordan) June 23, 2024 “Been using Claude 3.5 for coding day-to-day and wow, I think this may happen sooner than people imagine. Software is dead. And we have killed him,” posted another user on X. Furthermore, when combined with Artifacts, Claude 3.5 Sonnet simplifies developers’ tasks by providing real-time visibility into their app development process. This capability makes Claude an ideal pair programmer, enhancing collaboration and efficiency. One of the most exciting things about artifacts is that it's not just making existing coding jobs easier\/faster, it's creating software that would never be written otherwise.​Many people are now creating their first apps, with Claude's help.​Niche apps that only a few people… pic.twitter.com\/iQk2jzyHkL— Alex Albert (@alexalbert__) June 24, 2024 Wake Up OpenAI! Lately, OpenAI has faced mounting pressure from its competitors. The company announced the GPT-4o model just before Google I\/O, but its voice capabilities have yet to be released. “It’s been a month and a half and there is no new voice or vision functionality available, openai shipped a blog post,” quipped a user on X. OpenAI has announced that the voice feature will be made available to all Plus users this fall. “We had planned to start rolling this out in alpha to a small group of ChatGPT Plus users in late June, but need one more month to reach our bar to launch,” the company said. The company is improving the model’s ability to detect and refuse certain content. “We’re also working on improving the user experience and preparing our infrastructure to scale to millions while maintaining real-time responses.” OpenAI will start the alpha with a small group of users to gather feedback and expand based on what they learn. “We are planning for all Plus users to have access in the fall.” We're sharing an update on the advanced Voice Mode we demoed during our Spring Update, which we remain very excited about:We had planned to start rolling this out in alpha to a small group of ChatGPT Plus users in late June, but need one more month to reach our bar to launch.…— OpenAI (@OpenAI) June 25, 2024 “Just a year ago, it was unthinkable that any other model would even remotely approach GPT’s lead. Today, Sonnet-3.5 (not even Anthropic’s biggest Opus model) is already slightly above and Llama-3-400B is around the corner,” posted another user on X. “I never thought I would have anything on par with the OpenAI models, but Anthropic is killing it with every new model without any drama, focused on APIs and developers instead of giving startup-killing vibes and hubris. And when we hit the wall, Anthropic could be a winner,” said KissanAI founder Pratik Desai. Meanwhile, Amazon is developing a ChatGPT killer app, internally codenamed Metis. This project uses retrieval-augmented generation to provide up-to-date information and automate tasks. The same thing happened with OpenAI’s video generation model Sora. “Just 4 months ago, Sora blew everyone’s mind and seemed so out of reach. Today, we have at least 4-5 clones of Sora at 70-80% quality, such as Kling, Luma, and Runway. The clones wouldn’t have rallied without OpenAI’s first move,” posted NVIDIA AI researcher Jim Fan. One can only hope that OpenAI will keep up with their promises and deliver the new models soon.","excerpt":"The standout feature in Claude 3.5 Sonnet that has captured everyone’s attention is the Artifacts tool.","categories":["AI Features"],"tags":["Anthropic","ChatGPT","OpenAI"],"author_name":"Siddharth Jindal","publish_date":"2024-06-26T15:04:20","publication_year":"2024","word_count":1033,"keywords":["Anthropic","ChatGPT","Go","TPU","OpenAI","AI","GPT-4o","ML","Claude 3.5","R"],"extracted_tech_keywords":["AI","ML","GPT-4o","ChatGPT","OpenAI","Claude 3.5","Anthropic","TPU","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/farewell-chatgpt\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10054634,"title":"AWS Releases A No Code Machine Learning Tool","content":"Amazon recently announced the general availability of Amazon SageMaker Canvas, a new visual, no code capability that allows business analysts to build ML models and generate accurate predictions without writing code or requiring ML expertise. Its intuitive user interface lets users browse and access disparate data sources in the cloud or on-premises, combine datasets with the click of a button, train accurate models, and then generate new predictions once new data is available. SageMaker Canvas leverages the same technology as previous Amazon SageMaker to automatically clean and combine data, create hundreds of models under the hood, select the one performing best, and generate new individual or batch predictions. It supports multiple problem types such as binary classification, multi-class classification, numerical regression, and time series forecasting. These problem types let you address business-critical use cases, such as fraud detection, churn reduction, and inventory optimization, without writing a single line of code, a capability that was not available prior. Users can use any dataset, from high complexity and down to a basic CSV file, and then decide which of the columns in this dataset Canvas should predict. Data can be imported or fetched from Amazon Simple Storage Service (Amazon S3) or connected to other cloud or on-premises data sources, such as Amazon Redshift or Snowflake. Image Source: AWS Training the created model is not to be worried about, as SageMaker Canvas shows the value distribution and already recommends the most appropriate model type. Before proceeding with the model training, SageMaker Canvas also provides the option to generate an analysis report. The model, when ready, lets analyze its accuracy and the column impacts visually through the console. The Amazon SageMaker Studio integration lets share the model easily with other data scientists in a team. This makes for a far easier user experience than using traditional ML tools. Image Source: AWS SageMaker Canvas is now generally available in US East (Ohio), US East (N. Virginia), US West (Oregon), Europe (Frankfurt), and Europe (Ireland). Users can start using it with local datasets, as well as data already stored on Amazon S3, Amazon Redshift, or Snowflake. With just a few clicks, users can prepare and join datasets, analyze estimated accuracy, verify which columns are impactful, train the best performing model, and generate new individual or batch predictions.","excerpt":"SageMaker Canvas leverages the same technology as previous Amazon SageMaker to automatically clean and combine data, create hundreds of models under the hood, select the one performing best, and generate new individual or batch predictions.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","Amazon AWS","Amazon Redshift","AWS","AWS Sagemaker","AWS services","Data Analytics","Data Science","Data Scientist","Deep Learning","fraud detection","Machine Learning","ML models","No code development platforms","no code platforms"],"author_name":"Victor Dey","publish_date":"2021-12-01T17:35:52","publication_year":"2021","word_count":381,"keywords":["Amazon SageMaker","AWS services","R","fraud detection","RAG","Data Science","no code platforms","Go","AWS","AI","ML","ML models","Machine Learning","AWS Sagemaker","Amazon Redshift","mlops_tools:SageMaker","Amazon AWS","Data Analytics","Deep Learning","Data Scientist","No code development platforms","AI (Artificial Intelligence)","Snowflake"],"extracted_tech_keywords":["AI","ML","Amazon SageMaker","RAG","fraud detection","AWS","Snowflake","R","Go","mlops_tools:SageMaker"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/aws-releases-a-no-code-machine-learning-tool\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10014812,"title":"Amid Ethical AI Controversy, Google Releases Research In Collaboration With Apple &#038; OpenAI","content":"“Google AI’s new research on language models, in collaboration with OpenAI and Apple hints that the company cares about transparency.” As language models continue to advance, chances of encountering new and unexpected risks are high. The line of work resembles that of ex-Googler Timnit Gebru who was fired earlier this month for her “inconsistent” allegations against the company. Gebru, in her unpublished paper, was talking about various implications of large language models — the research, which Google found to be more accusing than asserting. Google’s Jeff Dean said that Gebru’s work ignores much of the ongoing research works at the company. A fortnight after this dispute, Google has published a paper titled, “Extracting training data from large language models”, in collaboration with OpenAI, Apple Inc, Stanford, Berkeley, and Northeastern University, demonstrating that given only the ability to query a pre-trained language model, it is possible to extract specific pieces of training data that the model has memorised. About The Research Source: GoogleAI Google AI researchers and its collaborators have orchestrated a ‘training data extraction attack’ to inform researchers about the vulnerabilities in large language models. According to Google, A training data extraction attack has the greatest potential for harm when applied to a model that is available to the public. For the training data extraction attack, the researchers used OpenAI’s GPT-2. For example — as shown above — if one prompts the GPT-2 language model with the prefix “East Stroudsburg Stroudsburg…”, it will autocomplete a long block of text containing the full name, phone number, email address, and home address of a particular person whose information was included in GPT -2’s training data. Google, in their blog, underlined that this experiment was done considering all factors. In accordance with responsible computer security disclosure norms, wrote Google, the extracted data of the individuals was secured before including references to this data in a publication. “We have also worked closely with OpenAI in the analysis of GPT-2,” added Nicholas Carlini, Research Scientist, Google Research. So, how exactly can this attack be exploited? A decent language model is supposed to predict the next word in the commonly used phrases. Prompting the language model with “Don’t judge a book by its….” would usually result in the word “cover”. But, with extraction attacks, the language model can be tricked if one particular training document happens to repeat the string “Don’t judge a book by its bookmarks” many times, the model might predict that phrase instead. “Language models continue to demonstrate great utility and flexibility—yet, like all innovations, they can also pose risks.” The idea here is if one model has high confidence in a sequence, but the other (equally accurate) model has low confidence in a sequence, it’s likely that the first model has memorised the data. That said, many of these content used for the experiment that covers news headlines, log messages, JavaScript code etc. have been memorised even though they appear infrequently in the training dataset. This work raises concerns about the potential risks of deploying large languages models. This work asserts that memorisation is enhanced with more parameters. The researchers recommend using up and coming techniques like differential privacy, among others to train models with reduced memorisation. Key Takeaways The main aim of this research is to expose the consequences of memorisation in large language models. Here are a few highlights that capture the essence of this whole work: Larger the language model, the easier it memorises training data.Knowledge of these attacks enables researchers to predict if a result was used in the training data by checking the confidence of the model on a particular sequence.One quick solution is to ensure that models do not train on any potentially problematic data.More research has to be done on how to monitor and mitigate this problem in increasingly large language models. Find out more here.","excerpt":"“Google AI’s new research on language models, in collaboration with OpenAI and Apple hints that the company cares about transparency.” As language models continue to advance, chances of encountering new and unexpected risks are high. The line of work resembles that of ex-Googler Timnit Gebru who was fired earlier this month for her “inconsistent” allegations […]","categories":["Global Tech"],"tags":["Apple","differential privacy","Ethical AI"],"author_name":"Ram Sagar","publish_date":"2020-12-20T18:00:00","publication_year":"2020","word_count":641,"keywords":["Go","OpenAI","AI","Apple","Ethical AI","innovation","GPT","Aim","differential privacy","JavaScript","R","Java"],"extracted_tech_keywords":["AI","OpenAI","Aim","differential privacy","R","JavaScript","Go","Java","GPT","innovation"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/google-ai-openai-apple-privacy-language-models\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10791,"title":"CRM becomes smarter with Salesforce’s Einstein AI","content":"Artificial Intelligence has been the flavor of the season. And it is definitely much more than simply a buzz word with the biggest organizations like Microsoft, Google, building and creating Artificial Intelligence tools.  The most recent announcement coming from Salesforce. Salesforce, leading CRM provider, has been an organization which has always put in efforts to stay ahead of the curve. And its announcement of Einstein AI is also a part of that endeavor. So let us explore what Einstein AI is all about. Introducing Einstein AI Salesforce’s Einstein AI is not just an artificial intelligence based product but it is an umbrella of technology which houses all the artificial intelligence pieces of the company. This artificial intelligence platform is developed with CRM as it main objective. With the introduction of Einstein AI, Salesforce aims at making AI a part of everyday business life. The company aims to reduce the complexity that is involved in AI tools with Einstein AI and make the Salesforce CRM platform, the world’s smartest CRM. What does Einstein offer? Salesforce’s Einstein AI brings in advanced AI capabilities to the platform of CRM, giving an edge to Salesforce. The Einstien AI is powered by advanced machine learning, deep learning, predictive analytics, natural language processing and smart data discovery. All these make Einstien AI smarter as these models will be automatically customized for each and every customer and with every interaction, it will collect more data, learn and get smarter. Here’s what Einstien AI will offer its customers Applications like these reply on the quantity and quality of data. So where does the data come from for Salesforce’s AI platform Einstein. Salesforce does have access to a large body of data of customers, territories and so on. Also with every interaction the data base will keep enhancing, and improvising the AI tools. Einstein AI is a platform with advanced AI capabilities with CRM at its core. These AI tools will help business change the way they manage their customers. Einstein will help businesses discover insights relevant to them, predict future behaviours of customers and proactively recommend the actions that are needed to be taken best in the interest of the business. The story of Einstein and the way forward To make Einstein become a reality, Salesforce made quite a few acquisitions in the field of artificial intelligence. The company has formed a team of 175 data scientists to make Einstien AI by leveraging on acquisitions such as MetaMind, PredictionIO and RelateIQ. Also the company plans to introduce a new research group which can contribute towards enhancing the features of Einstien and the overall AI experience. This research group will be headed by Richard Socher, chief data scientist, Salesforce (MetaMind founder). This team will work on research papers and be part of top academic conferences to further their research in areas of computer vision, image processing and natural language processing.","excerpt":"Artificial Intelligence has been the flavor of the season. And it is definitely much more than simply a buzz word with the biggest organizations like Microsoft, Google, building and creating Artificial Intelligence tools.  The most recent announcement coming from Salesforce. Salesforce, leading CRM provider, has been an organization which has always put in efforts to […]","categories":["IT Services"],"tags":["Deep Learning","Machine Learning","Salesforce","salesforce crm"],"author_name":"Manisha Salecha","publish_date":"2016-09-27T10:38:04","publication_year":"2016","word_count":481,"keywords":["artificial intelligence","machine learning","AI","R","Machine Learning","computer vision","RAG","Aim","salesforce crm","deep learning","Salesforce","analytics","Deep Learning","predictive analytics"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","deep learning","computer vision","analytics","Aim","RAG","predictive analytics","R"],"url":"https:\/\/analyticsindiamag.com\/it-services\/crm-becomes-smarter-salesforces-einstein-ai\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10058312,"title":"TCS reaches a milestone, hits $25 billion in annual revenue","content":"Tata Consultancy Services has reached a milestone of $25 billion in annual revenue for the year ended December 2021. TCS’ net profit for the December quarter grew 12.3 per cent to Rs 9,769 crore. The company said that its growth over the past four quarters helped them reach the $25 billion mark. Rajesh Gopinath, the CEO and managing director of TCS said, “Our continued growth momentum is a validation of our collaborative, inside-out approach to our customers’ business transformation needs.” The company also promoted 1.1 lakh employees in this fiscal year and plans to promote 40,000 more in Q4 FY22. With the addition of 28,238 members, TCS has taken its total employee strength to 5,56,986. With improved performance in terms of both revenue and talent acquisition and retention, TCS has managed to become the leading Indian SaaS company. https:\/\/twitter.com\/tejeshwi_sharma\/status\/1481495666304352256?s=20 TCS is rapidly transitioning to a more cloud and SaaS-based platform. In an earlier interview with Economics Times, N Ganapathy Subramaniam, the chief operating officer of TCS, said that in 2021, over 95 per cent of the company’s deal wins have been cloud and SaaS-based platforms. He also said that the company is moving towards a SaaS-based consumption model. Notably, in July 2021, TCS had announced the launch of Jile 5.0, an updated version of a Saas-based corporate agile application to help businesses with their large sale development needs across different remote teams.","excerpt":"In July 2021, TCS had announced the launch of Jile 5.0, an updated version of a SaaS-based corporate agile application to help businesses with their large scale development needs across different remote teams.","categories":["AI News"],"tags":["TCS"],"author_name":"Shraddha Goled","publish_date":"2022-01-13T13:37:45","publication_year":"2022","word_count":232,"keywords":["Go","API","programming_languages:R","TCS","AI","programming_languages:Go","GAN","R"],"extracted_tech_keywords":["AI","R","Go","API","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/tcs-reaches-a-milestone-hits-25-billion-in-annual-revenue\/","complexity_score":3,"technical_depth":7,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10173111,"title":"Karnataka, Berlin Forge Alliance for Startup Exchange and AI","content":"In a step towards global innovation collaboration, the Government of Karnataka and the State of Berlin, Federal Republic of Germany, have signed a Joint Declaration of Intent on July 9, 2025. “This partnership under Karnataka’s Global Innovation Alliance will reduce the distance between Berlin and Bengaluru by enabling startup exchange, soft-landing support and deeper tech and innovation collaboration,” said Priyank Kharge, Karnataka’s Minister for IT\/BT and rural development, in a LinkedIn post. He further mentioned that such an initiative is to foster startup exchange, soft-landing support, and deeper collaboration in emerging technologies. The agreement was formalised in the presence of Franziska Giffey, mayor and senator for economic affairs, energy and public enterprises, Berlin. The initiative falls under Karnataka’s global innovation alliance (GIA) and is expected to bridge the innovation ecosystems of Bengaluru and Berlin, two global hubs renowned for their tech capabilities. The collaboration will focus on fintech, AI, IT\/digitalisation, and life sciences, aligning with Bengaluru’s strengths in technology and Berlin’s robust innovation environment. As reported by Deccan Herald, Berlin will extend its collaboration with Indian startups by offering a range of support services, including seed capital, advisory assistance, networking opportunities with investors, and access to co-working spaces. During their two-day visit to Bengaluru, the Berlin delegation is scheduled to visit the Infosys campus, RV College of Engineering, and IIM Bangalore. They will also engage with prominent industry bodies, including NASSCOM, to explore avenues for deeper collaboration and knowledge exchange. Highlighting the inclusive nature of the initiative, Kharge added, “Both governments are clear that innovation must be equitable, therefore we are placing special emphasis on women-led enterprises and green innovation.” To ensure effective implementation and measurable outcomes, both governments will soon set up a joint steering committee along with dedicated nodal points on each side. These will be responsible for tracking progress, co-creating programmes, and ensuring outcome-based delivery.","excerpt":"The Berlin delegation is scheduled to visit the Infosys campus, RV College of Engineering, and IIM Bangalore.","categories":["AI News"],"tags":["karnataka"],"author_name":"Shalini Mondal","publish_date":"2025-07-09T13:08:59","publication_year":"2025","word_count":308,"keywords":["karnataka","Go","API","programming_languages:R","AI","innovation","programming_languages:Go","Git","R","startup"],"extracted_tech_keywords":["AI","R","Go","Git","API","innovation","startup","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/karnataka-berlin-forge-alliance-for-startup-exchange-and-ai\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10166152,"title":"It’s a Bad Idea to Trust Grok, Gemini, ChatGPT, Perplexity for Citations","content":"Over the years, using the internet has increasingly meant surrendering control over the content we see. It is particularly evident in social media platforms, where information is algorithmically curated rather than actively sought by users. Even though search engines allow individuals to independently seek and select information by clicking directly on desired links, this approach is slowly diminishing. Google’s recent testing of the ‘AI Mode’ in search suggests a future where AI-based curation becomes the default method for information retrieval. Similarly, Perplexity, Grok, and ChatGPT also heavily promote AI-driven search tools, which seems to have worked. One in four Americans uses AI instead of traditional search engines. However, a new report from Columbia University highlighted a critical flaw affecting these AI-based search tools—issues with citation accuracy, which is the very aspect AI labs emphasise to build user confidence. Study Exposes Gaps in AI Search Accuracy Tow Center for Digital Journalism, Columbia University, performed an evaluation of search tools from ChatGPT, Perplexity, Grok, DeepSeek Search, and Google’s Gemini. Ten articles from each of the twenty publishers were selected randomly, and direct excerpts were picked from those articles as input for the AI tool. Then, the tool was asked to identify the article’s headline, original publisher, publication date, and URL. The study found that collectively, these search engines provided incorrect answers to more than 60% of queries. Notably, Perplexity answered 37% of queries incorrectly, and Grok 3 answered 94% of queries incorrectly. Source: Tow Center for Digital Journalism “Most of the tools we tested presented inaccurate answers with alarming confidence,” read the study, which highlighted that outputs rarely used phrases like ‘it appears’, ‘it’s possible’ and ‘I couldn’t locate the exact article’, all of which signify knowledge gaps, and uncertainties. The research also revealed that more than half of the responses from Gemini and Grok 3 cited broken links. Source: Tow Center for Digital Journalism Moreover, these AI tools also often failed to identify the original source of the content. “For instance, despite its partnership with The Texas Tribune, Perplexity Pro cited syndicated versions of Tribune articles for three of the ten queries. In contrast, Perplexity cited an unofficial republished version for one,” the report added. These issues stem despite the continued efforts of companies like OpenAI, and Perplexity to partner with publishers to provide reliable, and accurate outputs. The study observed multiple instances of these chatbots providing inaccurate responses from the very website they teamed up with. Source: Tow Center for Digital Journalism These results are alarming, to say the least. “Seems pretty misleading to advertise a capability as search\/retrieval if it provides incorrect answers and links over 40% of the time,” Narasimha Chari, a product manager on X, said while citing the study. Google Has a Responsibility to Fulfill While AI systems and products continuously improve, there has been an increasingly strong push to adopt AI for search. Given the above results, this might seem premature. Recently, Google announced that AI overviews are being rolled out to more users, without having to sign in to access the feature. Aligning with the results of the above-mentioned study, several users have recently expressed frustration with AI overviews and their inaccurate responses. While Google calls AI overviews “one of the most popular search features ever”, there also seems to be no way to disable them. For instance, Mehdi Sadaghdar, who runs the popular YouTube channel ElectroBOOM, found Google’s AI providing a confusing response to a rather straightforward question. When he wanted to find the amount of energy contained by a lightning bolt, the AI overview first answered “1 gigajoules”, followed by another result showing an answer of “approximately 5 gigajoules”. “I feel it is dangerous for Google AI answers to be the first result in the searches. I found myself accepting what it says as fact, but then with inaccuracies…it could be spreading false information that would result in inaccurate responses,” Sadaghdar added in a post on X. Kind of useless google AI overview! pic.twitter.com\/9pfrFsf6ki— Mehdi Sadaghdar (@ElectroBOOMGuy) January 27, 2025 Moreover, Google is also testing an ‘AI Mode’ in Google Search, which, according to its demonstration video, seems to be the first tab users can see. Moreover, as per Google, it comes with enhanced capabilities for reasoning, multi-modal and high-quality responses with Gemini 2.0. Having said that, Google has indeed been having an incredible run with its newly released Gemini models and the associated multimodal features recently. It is only fair to expect more refinements to AI overviews in search, a product from the company that faces the most number of users. Moreover, a report from Statista suggests that over 90 million online users in the United States are set to primarily rely on AI for browsing the web. AI makers will certainly need to undertake more responsibilities as false information can lead to mild inconveniences and even fatal consequences in some situations.","excerpt":"Web search features in AI apps struggle to provide accurate information about the original publishers.","categories":["Global Tech"],"tags":["AI (Artificial Intelligence)","AI Search Engine"],"author_name":"Supreeth Koundinya","publish_date":"2025-03-17T18:11:40","publication_year":"2025","word_count":811,"keywords":["Grok 3","ChatGPT","Gemini 2.0","TPU","AI Search Engine","OpenAI","AI","chatbots","Go","ML","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","ML","ChatGPT","OpenAI","Gemini 2.0","Grok 3","chatbots","TPU","R","Go"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/its-a-bad-idea-to-trust-grok-gemini-chatgpt-perplexity-for-citations\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":15119,"title":"Nvidia’s Deep Learning Institute sets an aim to train 100,000 developers in deep learning and AI","content":"Nvidia, a world leader in Visual Computing Technologies, recently mentioned that they putting efforts to train developers in artificial intelligence and machine learning through its Deep Learning Institute. The announcement which came in the recently concluded annual GPU Technology Conference at San Jose, California, aims to train over 100,000 developers. The Deep Learning Institute by Nvidia which was launched in 2015, comprises of both hand-on training sessions and online courses for AI enthusiasts. The initial sessions for the program were offered at various GTC conferences across various cities, including India and China. These hands-on sessions delivered during global series of GTC conferences in seven cities, delivered more than 18,000 hours of training during a span of two years. In US alone, the institute has trained developers across companies like Adobe, Alibaba, SAP, government institutions including the National Institutes of Health (NIH), National Institute of Science and Technology (NIST). In India, it has reached the Indian Institute of Technology, Bombay. However, in late 2016, the company made available online courses in collaboration with Coursera, Udacity, and Microsoft. It now intends to train up to 100,000 data scientists and developers in mastering their skills not only in machine learning workflows, but in using deep neural networks to solve complex real-world modeling problems. The course offerings: The topics of the course revolve around becoming a self-driving car engineer, creating smarter robots using deep learning with tools like Microsoft Azure, predicting the risk of a disease, preventing it and more. It also includes instructor-led seminars, workshops, classes that reach developers across Asia, Europe and US. The hands-on training would be taught by certified experts from Nvidia partnering companies and universities, wherein they would cover fundamentals of deep learning with topics like AI for object detection and image classification with TensorFlow, neural network deployment with DIGITS, and inference optimization for autonomous vehicles with TensorRT. The virtual classes are delivered through high performance Amazon Web Services and Google’s Qwiklabs. With the number of training hours per person increasing along with number of training labs, Nvidia plans to certify the engineering competence in the labs it now offers within the next year. The various areas of instruction include health care, self-driving cars, web services, robotics, video analytics, and financial services.","excerpt":"Nvidia, a world leader in Visual Computing Technologies, recently mentioned that they putting efforts to train developers in artificial intelligence and machine learning through its Deep Learning Institute. The announcement which came in the recently concluded annual GPU Technology Conference at San Jose, California, aims to train over 100,000 developers. The Deep Learning Institute by […]","categories":["AI News"],"tags":[],"author_name":"Srishti Deoras","publish_date":"2017-05-22T10:47:05","publication_year":"2017","word_count":373,"keywords":["machine learning","artificial intelligence","AI","neural network","Aim","deep learning","object detection","analytics","TensorFlow","Azure"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","deep learning","neural network","analytics","Aim","TensorFlow","object detection","Azure"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/nvidias-deep-learning-institute-sets-aim-train-100000-developers-deep-learning-ai\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10071556,"title":"Intel banks on chip making for a turnaround","content":"Intel’s chief executive Pat Gelsinger seems ready to play hardball. At the beginning of July 2022, Gelsinger threatened to take his chip expansion plan to Europe if the US Congress failed to approve USD 52 billion in subsidy under the 2021 ‘CHIPS Act’. Gelsinger had grudgingly announced an indefinite delay in the groundbreaking ceremony of its USD 20 billion-plus manufacturing facility in Ohio owing to a scarcity in government funding. The company had already planned to invest around USD 35 billion in production in the European Union. “I hate the idea of announcing a delay”, stated Gelsinger at the Aspen Ideas Festival later. However, it was no secret that Intel was in the middle of a restoration, and a lot was riding on its new moves. Intel’s Fab 42 in Chandler, Arizona, Source: Cnet.com More ‘open’ and integrated system In January 2022, Intel announced an investment of nearly USD 20 billion to build two new semiconductor chip manufacturing plants in Ohio, USA. The Ohio plant was merely a cog in Gelsinger’s ambitious long-term strategy to make Intel competitive with chip manufacturers like Samsung and Taiwan Semiconductor Manufacturing Company, Limited (TSMC). Gelsinger claimed that Intel was willing to invest more than USD 250 billion over the next decade in capital toward this goal, reportedly including a USD 95 billion investment in chip production. Intel CEO Patrick Gelsinger, Source: Intel An Intel veteran, Pat Gelsinger, was the company’s first chief technology officer in 2001 but quit to join EMC. Gelsinger returned to Intel in February 2022 after a nine-year stint in software maker ‘VMWare’ as the CEO. However, activist investors who had expected Intel to split into two distinct segments—a foundry and a chip manufacturer—were taken aback by Gelsinger’s decision to integrate the two divisions as a part of his ‘IDM 2.0’, or integrated device manufacturing model. Intel aims to leverage their third-party foundries similar to those of TSMC’s in an attempt to save costs. The shortage in semiconductor chips has been significantly more pronounced in the last year and a half since the pandemic. The costs involved in building fabs kept many companies, including Intel, at bay. Since 2001, semiconductor manufacturers have slimmed down from 20 to a handful because of the monetary investment required for manufacturing cutting-edge chips. One of the main reasons Gelsinger opened up Intel’s fabs was to increase production to lower costs—the more the volume of chips produced and sold, the more the production yield to cover costs. However, producing chips for customers in their fabs was merely a one-pronged solution to solve Intel’s challenges. The company also needed to amp up research and development efforts while keeping a tight reign over its costs. Source: BizVibe.com Expansion of chip-making in Asia In the early 1980s, Taiwan was moving towards capital-intensive industries that revolutionised its economy. The country lapped up chipmaking technology, and while they were not equipped to design or market chips, Taiwan was able to produce them. TSMC was founded in 1987. Eventually ‘Intel’ too roped in TSMC to manufacture chips for them. Gradually, as companies like IBM stepped away from the foundry business, TSMC witnessed growth. Apple began testing TSMC chips against Samsung’s chips. By 2015, with the launch of the iPhone 7, TSMC’s low-power processor chips had edged out Samsung. Meanwhile, other companies started designing their chips, like Apple, rather than purchasing parts that Intel was selling. However, Intel continued to update its technology until 2015—when it fell behind. The company revealed a delay in manufacturing chips on its 10nm nodes. By 2017, there was another delay—Intel was now lagging behind its competition. Intel’s net revenue from 2014-2021 segment-wise, Source: Statista In addition, not all clients were impressed with Intel’s hand-me-down technology. Like Microsoft, Intel’s chip-making was increasingly insular and did not combine chip designs. While the explosion in cloud computing allowed Intel to stick to its ways for a longer time than Microsoft, the company’s culture was demanding change in the face of competition. Open systems like ‘Arm’ had their blueprints used by most smartphones world-wide. But, ironically, Intel had completely missed out on the smartphone market. Future of chip making The integration of the foundry model into Intel’s business is deemed herculean. It signifies chasing Asia in the high-performance market. The company is believed to be in dire need of transformation in its work culture. Gelsinger’s idea is to bring back the ‘Grovian culture’—inspired by Intel’s co-founder Andy Grove who famously said ‘only the paranoid survive.’ Intel’s earlier attempts at the foundry business wilted within five years in 2018. However, Intel seems to have learned its lessons from then on. Intel Foundry Services would now have its separate profit-and-loss statement and report directly to Gelsinger. Building fabs and getting them operational is expected to take a few years, and Intel maintains that they are willing to invest the required capital and time to achieve that. The company has further claimed that it will embrace extreme ultraviolet lithography in its chips, an aspect TSMC already embraces. Intel plans to launch at least one new processor yearly with faster circuits and smaller transistors. By 2025, the goal remains to manufacture chips that could be measured in angstroms—a unit smaller than nanometers. Intel Foundry Services (IFS) now counts AWS and Qualcomm as its first clients. Owing to Intel’s newfound ‘openness’, hyperscale cloud providers like AWS would be given access to their chip designs. IFS is also undertaking proactive initiatives to build new partnerships. For instance, on 25 July 2022, IFS announced a new collaboration with ‘MediaTek’ to strengthen its supply chain in the US and Europe. “As one of the world’s leading fabless chip designers powering more than 2 billion devices a year, MediaTek is a terrific partner for IFS as we enter our next phase of growth”, says Senior Vice President and Intel Foundry Services President Randhir Thakur. “We have the right combination of advanced process technology and geographically diverse capacity to help MediaTek deliver the next billion connected devices across a range of applications.” Building fabs in Asia is cheaper by 30~40 per cent owing to the significant support from their governments. However, for manufacturing in the US to move ahead, government subsidies are pivotal. If Intel is able to obtain the funding, it would make an almost miraculous jump to its revival. However, if it fails—the semiconductor industry would only witness further consolidation.","excerpt":"Owing to Intel’s newfound ‘openness’, hyperscale cloud providers like AWS would be given access to their chip designs.","categories":["Global Tech"],"tags":[],"author_name":"Poulomi Chatterjee","publish_date":"2022-07-26T16:00:00","publication_year":"2022","word_count":1061,"keywords":["Go","API","AWS","cloud computing","AI","RAG","GRU","Aim","GAN","R"],"extracted_tech_keywords":["AI","Aim","RAG","cloud computing","AWS","R","Go","API","GAN","GRU"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/intel-banks-on-chip-making-for-its-turnaround\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10061539,"title":"Underrated Kaggle notebooks every data science enthusiast must know","content":"Kaggle is synonymous with competitions and hackathons in the world of data science, but it is also a great resource to learn more about the field through community-driven notebooks. In contrast to textbooks and lectures, Kaggle notebooks or kernels provide data scientists with tutorials in their language. These are essentially Jupyter notebooks that run in the browser free of charge and without even needing to set up a local environment for Jupyter. In addition, these notebooks explore and run machine learning code and discover vast public and open-sourced repositories. While there are hundreds of thousands of notebooks on Kaggle, all data enthusiasts must-read are the top eight underrated notebooks. Register for Data engineering Summit 2022 Kannada MNIST: Choosing the Right Optimiser by ILM Kannada MNIST is a comprehensive and well-structured overview of essential deep learning optimisers implemented in Keras libraries. These include BGD, MGBD, AdaMax, Adam, SGD and Nadam and more. The notebook explains these with theory, graphs and applications. Additionally, the maths are supplemented with TensorFlow\/Keras implementation comparisons. The kernel includes various tasks of loading and preparing data, building a model, training a network, checking the performance on Dig-MNIST data and making submission files. Learn more here. Notebook9298460840: An Inference Notebook for the winning solution in the Jigsaw Toxic Severity by Guanshuo Xu This inference notebook is based on the winning solution at the Jigsaw Toxic Severity competition. The competition ranks comments in order of severity of toxicity, where a higher degree of toxicity comments receive a higher numerical value than the lower degrees. Guanshuo Xu is one of the world’s top Kagglers and has published the training code for the competition’s solution, which usually has no training data. Learn more here. Transformers Course – Chapter 2 – TF & Torch by Darien Schettler This collection of notebooks takes the users through the various components of the Hugging Face Transformers Course that consists of using transformers, fine-tuning pretained models, sharing models and tokenisers, dataset library, tokenisers library and main NLP tasks. This notebook is the second part of the series, covering eight aspects of using a transformer. In addition, the notebook shows the Tensorflow and PyTorch code. Learn more here. How to Create Award-Winning Data Visualisations by Andrew Sinek Andrew Sinek is the third prize winner of Kaggle’s Machine Learning and Data Science Survey Competition for this notebook surveying jobs on Kaggle vs Glassdoor. An important aspect of the award was his presentation. The How to Create Award-Winning Data Visualisations notebook illustrates his thought process and step-to-step instructions for building effective visualisations. The thought process is explained with precision and covers all the aspects of visualisation such as preparing the data, data to plot, first plot, choosing a plot, closing the lines, creating a hierarchy, decluttering, telling a story, adding a meaningful title and more. Learn more here. Comprehensive data exploration with Python by PEDRO MARCELINO Data analysis can sometimes be time-consuming, and data scientists can easily miss the initial but important steps in the long process. This notebook is an in-depth tutorial of data analysis principles and steps based on ‘Examining your Data’ by Hair et al. (2013). The author has also shared codes, examples and illustrations of applying these principles to his problems. The tutorial content includes understanding the problem, univariate, multivariate, and basic clearing and test assumptions. Learn more here. Understanding and Improving CycleGANs – Tutorial by Jesper Sören Dramsch Part 2 of a beginners GAN tutorial, this notebook teaches users to understand and improve GANs. The objective of the notebook is to explain baseline models in data science and create Monets with GAN. The tutorial studies data augmentation, neural network architectures, cycleGAN architectures and better loss functions in detail with examples of what to do and what not to do. The contents of the tutorial include loading the data, building the DCGAN, training the cycleGAN, visualising photos and creating submission files. Learn more here. Data Heroines – Saving the World Through Data Rather than a unique notebook, Data Heroines is a Kaggle Survey story written by three authors looking at various incredible women that make the field of data. The authors Datana Scientists, Datana Analystus, Datana Engineers and Machina Learnerum have explored a story-format journey of data, focusing on the women in data and how they dealt with COVID–19 and its impact on data. The story is told through a comic format followed by codes and graphs in each section. Learn more here. Visualising Convolution Filters by Hongnan G Initially created for a Petfinder competition, this notebook is a tutorial about the inner workings of a convolution filter through examples and illustrations. It teaches how convolution can be achieved through a kernel moving an entire image and calculating dot products with each window along the way. It also looks at the abstract features of convolutions and vertical and horizontal features. Learn more here. Further resource for more underrated Kaggle notebooks: Hidden Gems Collection","excerpt":"Initially created for a Petfinder competition, this notebook is a tutorial about the inner workings of a convolution filter through examples and illustrations.","categories":["AI Features"],"tags":["data science projects","Jupyter Notebook","machine learning projects","survival regression python"],"author_name":"Avi Gopani","publish_date":"2022-02-25T15:00:00","publication_year":"2022","word_count":818,"keywords":["data science projects","data science","Hugging Face","Jupyter Notebook","machine learning","Keras","AI","neural network","PyTorch","survival regression python","NLP","deep learning","TensorFlow","machine learning projects"],"extracted_tech_keywords":["AI","machine learning","deep learning","neural network","NLP","data science","TensorFlow","PyTorch","Keras","Hugging Face"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/underrated-kaggle-notebooks-every-data-science-enthusiast-must-know\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":42072,"title":"How To Keep World Leaders Safe From Deep Fake","content":"The world has become more connected now than ever before. An opinion can fly at light speed across continents and a revolution can be sparked by remote players in a matter of hours. And as the technology keeps improving, new methods are being discovered by fraudulent players to up their ante. Most platform owners tussle with the after-effects of an aggravation. To identify the perpetrator and discourage them to carry out any illicit, immoral social engineering would require automated detection techniques. In a new research funded by Google, Microsoft, and the Defense Advanced Research Projects Agency, a group of researchers developed a new method to keep the world leaders immune to the infamous deep fake. “With relatively modest amounts of data and computing power, the average person can, for example, to create a video of a world leader confessing to illegal activity leading to a constitutional crisis, a military leader saying something racially insensitive leading to civil unrest in an area of military activity, or a corporate titan claiming that their profits are weak leading to global stock manipulation,” said the authors in their paper titled ‘Protecting World Leaders Against Deep Fakes’. Dawn Of Multimedia Forensics Deep fakes can be categorised as follows: Face-swap, in which the face in a video is automatically replaced with another person’s face Lip-sync, in which a source video is modified so that the mouth region is consistent with an arbitrary audio recording Puppet-master, in which a target person is animated (head movements, eye movements, facial expressions) by a performer sitting in front of a camera and acting out what they want their puppet to say and do. The talking style and facial behaviour of a person can vary with the context in which the person is talking. Facial behaviour while delivering a prepared speech, for instance, can differ significantly as compared to answering a stressful question during a live interview The researchers collected videos like weekly addresses, where leaders like Barack Obama was talking to a camera In the second experiment, another round of videos of Obama were collected in even more significantly different contexts ranging from an interview in which he was looking at the interviewer and not the camera to a live interview in which he paused significantly more during his answer and tended to look downward contemplatively. Paper by Shruti Agarwal et al., Shown above are five equally spaced frames from a 250-frame clip annotated with the results of OpenFace tracking along with the intensity of one action unit AU01 (eyebrow lift) measured over this video clip. These regularities were exploited by building soft biometric models of high-profile individuals and then use these models to distinguish between real and fake videos The open-source facial behaviour analysis toolkit OpenFace2 was used to extract facial and head movements in a video. The facial and head movements were tracked and then the presence and strength of specific action units were extracted. A novelty detection model (one-class support vector machine (SVM)) was built that distinguishes an individual from other individuals as well as comedic impersonators and deep-fake impersonators The SVM hyper-parameters that control the Gaussian kernel width and outlier percentage are optimised using 10% of the video clips of random people taken from the FaceForensics original video data set. The SVM is trained on the 190 features extracted from overlapping 10-second clips The results show the efficacy of this approach on a large number of deep fakes of a range of U.S. politicians ranging from Hillary Clinton, Barack Obama, Bernie Sanders, Donald Trump, and Elizabeth Warren. Key Ideas The following objectives were demonstrated by the authors in the paper: Show that when individuals speak, they exhibit relatively distinct patterns of facial and head movements Show that the creation of all three types of deep fakes tends to disrupt these patterns because the expressions are being controlled by an impersonator (face-swap and puppet-master) or the mouth is decoupled from the rest of the face (lip-sync) This approach, unlike previous approaches, is resilient to laundering because it relies on relatively coarse measurements that are not easily destroyed, and is able to detect all three forms of deep fakes. The success of this method on a large scale relies on the diversity of the dataset that is being gathered of the targets prone to this deep fake attack There is a lot of misinformation circulated and it gets worse with the popularity of the entities involved. This virtual wildfire decouples the users from the truth and they usually end up in their own echo chambers. As the attention of the world media shifts towards the elections in the US, there will probably be attempts at foul play and having readily available tools to curate the data is almost mandatory. Read the full work here","excerpt":"The world has become more connected now than ever before. An opinion can fly at light speed across continents and a revolution can be sparked by remote players in a matter of hours. And as the technology keeps improving, new methods are being discovered by fraudulent players to up their ante. Most platform owners tussle […]","categories":["Deep Tech"],"tags":["deep fake","DeepFake","GAN"],"author_name":"Ram Sagar","publish_date":"2019-07-08T12:59:16","publication_year":"2019","word_count":796,"keywords":["Go","ELT","programming_languages:R","AI","programming_languages:Go","RAG","deep fake","Aim","ViT","CLIP","GAN","DeepFake","R"],"extracted_tech_keywords":["AI","Aim","RAG","R","Go","ELT","CLIP","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/world-leader-deep-fake-ai-tool\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":9932,"title":"Textual Disambiguation: The next evolution in textual analytics","content":"Most evolutions are silent. They take place in front of our eyes and the change is so gradual that we are not even aware of the evolution. It is only when we step aside and compress time and look at the progression that has occurred over time that it becomes obvious that we have been witnessing an evolution. There is an evolution in textual analytics that has been occurring. That evolution is seen in the Figure – Each of the steps of evolution have addressed and solved a clear and present requirement. But each step in the evolution has addressed only limited aspects of textual analytics. Let’s have a look at each step in the evolution. Comments In the beginning were comments. Programmers found that they needed to capture comments so they defined a field in their data structure called “comments” and stuffed comments into the field. They were lots of problems with comments. Defining the field length was always problematic. The comment was always longer or shorter than what was specified. It was – universally – an uncomfortable fit. But the real problem with the comments field was that once data was placed in the comments field you couldn’t do anything with it. However, historically the definition of the comments field was the first step that was taken to achieving textual analytics. Blobs Next came blobs. Blobs were a specifically defined field that held lots of undefined data. For the most part, blobs solved the problem of the irregular length of the comments field. But blobs did not address the problem with the usability of data within the blob. You could have all sorts of data in your blob but you really couldn’t do anything with it. However it must be recognized that blobs were a step in the evolution to textual analytics. Sound-ex and stemming Next came sound-ex and stemming. Sound-ex and stemming were the first steps taken to go into the blob and start to try to make sense of the text found in the blob. Sound-ex attempted to group words according to a common sound that was made when the word was pronounced. Related to sound-ex was stemming. Stemming was the practice of reducing words to their common Greek or Latin stem. In many cases reducing words to a common foundation helped in understanding syntax. Sound-ex and stemming were unquestionably pioneering steps in the search to start to make sense of text. Sound-ex and stemming were the next steps in the evolution of textual analytics. However, as important as sound-ex and stemming were, there is far greater complexity to text and language than these aspects. Tagging The next step in the evolution of textual analytics was that of word tagging. In word tagging a document was read and words that were of importance were tagged. Words could be tagged for many purposes. There were keywords. There were action words. There were red flag words. In truth there are many reasons why a given word could be tagged. As such tagging words was a real step forward in the progression to textual analytics. But tagging words had its own set of drawbacks. The primary drawback was that in order to do tagging you had to know what words needed to be tagged before you read the document. In some cases this was not a large drawback. But in other cases this was a real drawback. Many words that needed to be tagged in fact were not, making the process of tagging less than wholly satisfactory. However, in the evolution of textual analytics tagging represented a very large and positive step forward. NLP and Taxonomies Next in the evolution came NLP (natural language processing). NLP used taxonomies to gather sentiment. Taxonomies were much more powerful and much more general than tagging (although it can be argued that the usage of taxonomies and ontologies are an advanced form of tagging.) With NLP it was possible to gather enough information to start to do real textual analytics. In many ways textual analytics really began with taxonomies and NLP. However as powerful and as attractive as NLP was (and is) there were still some missing ingredients. In any case NLP and taxonomies represent a tremendous and positive evolutionary step. Textual Disambiguation The most evolved form of textual analytics is that of textual disambiguation. Textual disambiguation builds on the advances made by the advent of taxonomies and NLP. Textual disambiguation certainly provides all the capabilities of sentiment analysis. But textual disambiguation provides two important advances over NLP. Those advances are – the focus on context and the ability to recognize word and sentence structures. To understand the value of context, consider this simple example. Sentiment analysis “I don’t like scrambled eggs.” Sentiment analysis tells us that the author does not like scrambled eggs. And that is a useful piece of information. Textual disambiguation “I don’t like scrambled eggs because they are too expensive.” Textual disambiguation tells us that the author doesn’t like scrambled eggs. But textual disambiguation tells us something more. Textual disambiguation tells us why the author doesn’t like scrambled eggs. The scrambled eggs are too expensive. There could be many reasons why the author doesn’t like scrambled eggs. The eggs may be runny. The eggs may be cold. The author may be a vegetarian. The eggs may have too much cholesterol. There could be MANY reasons why the author does not like scrambled eggs. By textual disambiguation goes one step further than NLP and textual disambiguation gives us a different level of information than NLP. In addition textual disambiguation also has the capability of discerning context for word and sentence structure. There are a couple of interesting observations to be made about the evolution that has occurred. Programmers began to create comments fields as early as 1965. Textual disambiguation is alive and well in 2016. The evolution has taken a paltry 50 years or so. As far as evolutions are concerned, 50 years is a very short amount of time. The textual analytics evolution has occurred very, very quickly, insofar as evolutions are concerned. Another observation is that the evolution is continuing, like all evolutions. Even though textual disambiguation is in an advanced state, by no means is the evolution complete. Stay tuned.","excerpt":"Most evolutions are silent. They take place in front of our eyes and the change is so gradual that we are not even aware of the evolution. It is only when we step aside and compress time and look at the progression that has occurred over time that it becomes obvious that we have been […]","categories":[],"tags":[],"author_name":"William Inmon","publish_date":"2016-05-19T16:59:05","publication_year":"2016","word_count":1043,"keywords":["Go","programming_languages:R","sentiment analysis","AI","programming_languages:Go","NLP","analytics","GAN","ai_applications:NLP","R"],"extracted_tech_keywords":["AI","NLP","analytics","sentiment analysis","R","Go","GAN","programming_languages:R","programming_languages:Go","ai_applications:NLP"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/textual-disambiguation-next-evolution-textual-analytics\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":32984,"title":"What Is The Bootstrap Method In Statistical Machine Learning?","content":"Re-sampling is the method of taking samples iteratively from the original data samples. The method of Re-sampling is a non-parametric method of statistical inference which means that the parametric assumptions that ignore the nature of the underlying data distribution are avoided. Commonly Used Resampling methods: Sampling with and without replacementBootstrap (using sampling with replacement)Jackknife (using subsets)Cross-validation and LOOCV (using subsets)Permutation resampling (switching labels) The Bootstrap method is a technique for making estimations by taking an average of the estimates from smaller data samples. A dataset is resampled with replacement and this is done repeatedly. This method can be used to estimate the efficacy of a machine learning model, especially on those models which predict on data which is not a part of the training dataset. Bootstrap methods are generally superior to ANOVA for small data sets or where sample distributions are non-normal. How Is It Done This method becomes extremely useful to quantify the uncertainties present in an estimator. Select the sample sizeSelect an observation from the training data randomlyNow add this observation to the previously selected sample The samples not selected are usually referred to as the “out-of-bag” samples. For a given iteration of Bootstrap resampling, a model is built on the selected samples and is used to predict the out-of-bag samples. The resulting sample of estimations often leads to a Gaussian distribution. And a confidence interval can be calculated to bound the estimator. For getting better results, such as that of mean and standard deviation, it is always better to increase the number of repetitions. It may also be used for constructing hypothesis tests. It is often used as an alternative to statistical inference based on the assumption of a parametric model when that assumption is in doubt, or where parametric inference is impossible or requires complicated formulas for the calculation of standard errors. When Should One Use It When the sample size is small on which the null hypothesis tests have to be run.To account for the distortions caused by certain sample data which could be a bad representation of the overall data.To indirectly assess the properties of the distribution underlying the sample data. Bootstrapping In Python Example 1 via Source: Using sci-kit learn() oob = [x for x in data if x not in boot] from sklearn.utils import resample data = [0.1, 0.2, 0.3, 0.4, 0.5, 0.6] boot = resample(data, replace=True, n_samples=4, random_state=1) print('Bootstrap Sample: %s' % boot) oob = [x for x in data if x not in boot] print('OOB Sample: %s' % oob) Example 2: Visualisation of the Bootstrap method for convergence in the Monte Carlo integration import numpy as np import matplotlib.pyplot as plt def f(x): return x * np.cos(60x) + np.sin(10x) n = 100 x = f(np.random.random(n)) reps = 1000 xb = np.random.choice(x, (n, reps), replace=True) yb = 1\/np.arange(1, n+1)[:, None] * np.cumsum(xb, axis=0) upper, lower = np.percentile(yb, [2.5, 97.5], axis=1) plt.plot(np.arange(1, n+1)[:, None], yb, c='grey', alpha=0.02) plt.plot(np.arange(1, n+1), yb[:, 0], c='red', linewidth=1) plt.plot(np.arange(1, n+1), upper, 'b', np.arange(1, n+1), lower, 'b') plt.show() If one performs the naive Bootstrap on the sample mean which lacks a finite variance, then the Bootstrap distribution will not converge to the same limit as the sample mean. So in cases where there is an uncertainty associated with the underlying distribution and heavy-tailedness, Monte Carlo simulation of the Bootstrap could be misleading. Conclusion Over the years Bootstrap method has seen a tremendous improvement in the accuracy levels with improved computational powers as the sample size used for estimation can be increased and larger sample size usually has substantial real-world consequences in regards to increase in the accuracy of estimating errors in the data. There is also evidence of successful Bootstrap deployments for sample sizes as small as n= 50. Statisticians like Tshibirani define this Bootstrapping as a computer-based method for assigning measures of accuracy to sample estimates whereas, there are other definitions which say that this technique allows estimation of the sample distribution of almost any statistic using only very simple methods.","excerpt":"Re-sampling is the method of taking samples iteratively from the original data samples. The method of Re-sampling is a non-parametric method of statistical inference which means that the parametric assumptions that ignore the nature of the underlying data distribution are avoided. Commonly Used Resampling methods: Sampling with and without replacement Bootstrap (using sampling with replacement) […]","categories":["Deep Tech"],"tags":[],"author_name":"Ram Sagar","publish_date":"2019-01-07T12:59:17","publication_year":"2019","word_count":664,"keywords":["NumPy","machine learning","programming_languages:R","AI","ML","RAG","Python","programming_languages:Python","Matplotlib","R"],"extracted_tech_keywords":["AI","machine learning","ML","NumPy","Matplotlib","RAG","Python","R","programming_languages:Python","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/what-is-the-bootstrap-method-in-statistical-machine-learning\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10061116,"title":"Chronicling the 20 year journey of .NET","content":"“We all knew .NET is gonna change a programmer’s life,” said Anders Hejlsberg, chief C# architect and decorated engineer at Microsoft in a celebration video of NET completing 20 years of its existence. Over the years, this pioneering software framework has developed a loyal fanbase and currently has over five million developers using it. We all know that .NET is THE dev platform for building anything… All month, we're celebrating our 20 anniversary by sharing #dotNETLovesMe stories. Join us by tweeting your favorite .NET feature, success, or pic.Then check out our updated web home: https:\/\/t.co\/vBE5L6WYGJ pic.twitter.com\/UgKdjBJtII— .NET (@dotnet) February 16, 2022 .NET also scores high on developer surveys. For example, in the Stack Overflow’s developer survey 2021, .NET emerged as the most loved framework by developers. Image: Stack Overflow Moreover, .NET made it to the top in the CNCF report of 30 highest velocity open-source projects on GitHub from 2020-2021. Image: CNCF On the 20th anniversary, Twitter was chock-full of reactions from techies recounting their first encounter with .NET. This old design notebook of mine brings back fond memories. It has the original spec of what became https:\/\/t.co\/71Yjwgl0jE. It is amazing how .NET adoption continues to grow today and is used by millions of developers. Thank you to everyone in the .NET community! #dotNETLovesMe https:\/\/t.co\/SVQidaOE4H pic.twitter.com\/vmq05PjHKJ— Scott Guthrie (@scottgu) February 14, 2022 Celebrating the 20th Anniversary of .NET with a look back at the .NET launch at #VSLive! with Bill Gates in 2002. #dotNETLovesMe #Microsoft https:\/\/t.co\/H8V7dqM7NM pic.twitter.com\/t25i4e1FtN— Visual Studio Live! (@VSLive) February 14, 2022 I’ve been in .NET since the very first announcement, & I’ve built components, frameworks, massively scalable platforms, companies, + an entire career on .NET..NET made me the youngest developer MVP ever (thx @robhoward) & helped me meet @billgates.Thanks for 20 amazing years! https:\/\/t.co\/RRB5mix7FU— Robert McLaws (@robertmclaws) February 11, 2022 Features .NET is an open source developer platform that allows users to write .NET applications in C#, F#, or Visual Basic. The code runs natively on any compatible OS in .NET. The free platform is used to build for web, mobile, desktop, games, and IoT. The latest version .NET 6, dubbed “the fastest .NET  yet” by Microsoft, was released last November. It was also the first release that natively supported Apple Silicon (Arm64). Milestones 2002 Released in 2002, .NET was R allowed the developers to build Windows Forms and Web (ASP.NET) applications on C# and Visual Basic. .NET Framework provides the foundation for the .NET XML Web services. “Visual Studio .NET gives developers a toolkit to build such Web services and Web services-based applications. Beth Massi, Marketing Director, .NET said,”With the rise of the internet, the world saw an easier way to share information. Technology shifted towards distributed systems that communicated over the internet. .NET was built for this internet revolution.” 2005 F# comes into the picture 2005 was another landmark year as F#, the first functional-based .NET programming language was developed by Don Syme from Microsoft Research. F# is based on the functional programming language OCaml and is used for building web, cloud and data-science applications. In 2008, Microsoft released the source code for ASP.NET MVC. Image: Similartech 2014 Goes open source 2014 was a landmark year in .NET’s journey. At Build 2014 held in April, Microsoft announced “Roslyn” (.NET Compiler Platform) open source. In November, .NET Core was also made open source along with the framework libraries. According to Microsoft, the two big reasons behind the open source move were to lay the foundation for a cross platform as well as to build a stronger ecosystem. .NET foundation launched The .NET foundation was also announced in 2014 to promote open source software development. The independent, non-profit organisation was established to support an innovative, commercially friendly, open-source ecosystem around the .NET platform. Image: .NET Foundation 2016 .NET Core 1.0 released In June 2016, .NET Core 1.0, ASP.NET Core 1.0 and Entity Framework Core 1.0 were made available on Windows, OS X and Linux along with .NET Core runtime, libraries and the ASP.NET Core libraries. The Visual Studio team also released Visual Studio 2015 Update 3 along with it. Consequently, .NET Core 2.0 was released in 2017 and  .NET Core 3.0 in 2019. Xamarin acquisition In February 2016, .NET acquired mobile app development platform provider Xamarin. The combination of Xamarin, Visual Studio, Visual Studio Team Services, and Azure provided a complete mobile app development solution.","excerpt":"Over the years, this pioneering software framework has developed a loyal fanbase and currently has over five million developers using it.","categories":["AI Features"],"tags":["journey","Microsoft","Open Source"],"author_name":"Sreejani Bhattacharyya","publish_date":"2022-02-21T10:00:00","publication_year":"2022","word_count":728,"keywords":["Go","Open Source","AWS","AI","ETL","R","ML","Scala","Git","journey","GitHub","Azure","Microsoft"],"extracted_tech_keywords":["AI","ML","AWS","Azure","R","Go","Scala","Git","GitHub","ETL"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/chronicling-the-20-year-journey-of-net\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10004657,"title":"Hyperparameter &#8211; Difference Between Gridsearchcv and Randomizedsearchcv","content":"While building a Machine learning model we always define two things that are model parameters and model hyperparameters of a predictive algorithm. Model parameters are the ones that are an internal part of the model and their value is computed automatically by the model referring to the data like support vectors in a support vector machine. But hyperparameters are the ones that can be manipulated by the programmer to improve the performance of the model like the learning rate of a deep learning model. They are the one that commands over the algorithm and are initialized in the form of a tuple. In this article, we will explore hyperparameter tuning. We will see what are the different parts of a hyperparameter, how it is done using two different approaches – GridSearchCV and RandomizedSearchCV. For this experiment, we will use the Boston Housing Dataset that can be downloaded from Kaggle. We will first build the model using default parameters, then we will build the same model using a hyperparameter tuning approach and then will compare the performance of the model. What We Will Learn From This Article?What Is Hyperparameter Tuning?What Steps To Follow For Hyper Parameter Tuning?Implementation of Regression ModelImplementation of Model using GridSearchCVImplementation of Model using RandomizedSearchCVComparison of Different Models What Is Hyperparameter Tuning? Hyperparameter tuning is the process of tuning the parameters present as the tuples while we build machine learning models. These parameters are defined by us which can be manipulated according to programmer wish. Machine learning algorithms never learn these parameters. These are tuned so that we could get good performance by the model. Hyperparameter tuning aims to find such parameters where the performance of the model is highest or where the model performance is best and the error rate is least. We define the hyperparameter as shown below for the random forest classifier model. These parameters are tuned randomly and results are checked. RandomForestRegressor(bootstrap=True, ccp_alpha=0.0, criterion='mse', max_depth=None, max_features='auto', max_leaf_nodes=None, max_samples=None, min_impurity_decrease=0.0, min_impurity_split=None, min_samples_leaf=1, min_samples_split=2, min_weight_fraction_leaf=0.0, n_estimators=100, n_jobs=None, oob_score=False,  random_state=None, verbose=0, warm_start=False) What Steps To Follow For Hyper Parameter Tuning? Select the type of model we want to use like RandomForestClassifier, regressor or any other model Check what are the parameters of the model Select the methods for searching the hyperparameter Select the cross-validation approach Evaluate the model using the score Implementation of Regression Model First, we will import all the required libraries and the dataset and do the basic EDA to understand the data. Use the below code to do the same import pandas as pd import numpy as np from sklearn.tree import DecisionTreeRegressor from sklearn.ensemble import RandomForestRegressor df  = pd.read(‘Boston.csv’) print(df) Output: print(df.shape) Output: print(df.isnull().sum()) Output: print(df.info()) Output: There are a total of 506 rows and 14 columns in the data set, all the columns have float64 and int64 data type values and there are no missing values in the data set. Now we will define the independent and dependent variables y and x respectively. We will then split the dataset into training and testing. After which the training data will be passed to the decision tree regression model & score on testing would be computed. Refer to the below code for the same. y = df['medv'] X = df.drop('medv', axis=1) from sklearn.model_selection import train_test_split X_train, X_test, y_train, y_test = train_test_split(X, y, test_size= .30, random_state=1) from sklearn.tree import DecisionTreeRegressor dtr = DecisionTreeRegressor() dtr.fir(X_train,y_train) print(dtr.score(X_test,y_test)) Output: Implementation of Model using GridSearchCV First, we will define the library required for grid search followed by defining all the parameters or the combination that we want to test out on the model. We have taken only the four hyperparameters whereas you can define as much as you want. If you increase the number of combinations then time complexity will increase. Use the below code to do the same. from sklearn.model_selection import GridSearchCV param_grid = {  'bootstrap': [True], 'max_depth': [5, 10, None], 'max_features': ['auto', 'log2'], 'n_estimators': [5, 6, 7, 8, 9, 10, 11, 12, 13, 15]} Now we will define the type of model we want to build a random forest regression model in this case and initialize the GridSearchCV over this model for the above-defined parameters. rfr = RandomForestRegressor(random_state = 1) g_search = GridSearchCV(estimator = rfr, param_grid = param_grid, cv = 3, n_jobs = 1, verbose = 0, return_train_score=True) We have defined the estimator to be the random forest regression model param_grid to all the parameters we wanted to check and cross-validation to 3. We will now train this model bypassing the training data and checking for the score on testing data. Use the below code to do the same. g_search.fit(X_train, y_train); print(g_search.best_params_) Output: We can check the best parameter by using the best_params_ function that is shown above. print(best_grid.score(X_test, y_test)) Output: Implementation of Model using RandomizedSearchCV First, we will define the library required for random search followed by defining all the parameters or the combination that we want to test out on the model. Similar to grid search we have taken only the four hyperparameters whereas you can define as much as you want. We have then defined the random grid. Use the below code to do the same. import numpy as np from sklearn.model_selection import RandomizedSearchCV n_estimators = [int(x) for x in np.linspace(start = 5 , stop = 15, num = 10)] # returns 10 numbers max_features = ['auto', 'log2'] max_depth = [int(x) for x in np.linspace(5, 10, num = 2)] max_depth.append(None) bootstrap = [True, False] r_grid = {'n_estimators': n_estimators, 'max_features': max_features, 'max_depth': max_depth, 'bootstrap': bootstrap} print(random_grid) Output: We will now define the random search passing the rf model with the randomly chosen hyperparameters and then train it. After this, we will check the score. Use the below code to do the same. rfr_random = RandomizedSearchCV(estimator=rfr, param_distributions=r_grid, n_iter = 20, scoring='neg_mean_absolute_error', cv = 3, verbose=2, random_state=42, n_jobs=-1, return_train_score=True) rfr_random.fit(X_train, y_train); print(rf_random.best_params_) Output: print(best_random.score(X_test , y_test)) Output: Comparison of Different Models Models ScoresRegression Model (Without Hyperparameter Search)80.34Regression Model using GridSearchCV88.98Regression Model using RandomizedSearchCV90.17 Conclusion Model Hyperparameter tuning is very useful to enhance the performance of a machine learning model. We have discussed both the approaches to do the tuning that is GridSearchCV and RandomizedSeachCV. The only difference between both the approaches is in grid search we define the combinations and do training of the model whereas in RandomizedSearchCV the model selects the combinations randomly. Both are very effective ways of tuning the parameters that increase the model generalizability. Popular Posts Dickey Fuller Test in Time Series Hidden Markov Model and Application Scikit Learns Mlp Classifier Python Tabulate Module Pytorch Optimizers","excerpt":"While building a Machine learning model we always define two things that are model parameters and model hyperparameters of a predictive algorithm. Model parameters are the ones that are an internal part of the model and their value is computed automatically by the model referring to the data like support vectors in a support vector […]","categories":["Deep Tech"],"tags":["apm data science","hyperparameter tuning","Machine Learning","Parameter Tuning"],"author_name":"Rohit Dwivedi","publish_date":"2020-08-12T11:00:00","publication_year":"2020","word_count":1082,"keywords":["NumPy","machine learning","Parameter Tuning","TPU","AI","PyTorch","ML","Machine Learning","hyperparameter tuning","Python","apm data science","Aim","deep learning","Pandas"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","Aim","PyTorch","Pandas","NumPy","TPU","Python"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/guide-to-hyperparameters-tuning-using-gridsearchcv-and-randomizedsearchcv\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":62055,"title":"Top 6 AI Algorithms In Healthcare 2024","content":"Artificial intelligence (AI) has been integrating remarkable developments in almost every sector, and the healthcare industry is no different. There is an unprecedented amount of data flowing in the healthcare industry, which remains unlabeled and inconsistent, which can be analyzed to extract valuable insights using AI. The technology is also being used to provide virtual therapy and even used in some surgeries. A new paper is published every few weeks that tries to develop better algorithms to enhance the ways in which this industry can leverage the power of AI. In this article, we will have a look at some of the top AI algorithms that are currently being used in the healthcare industry to solve numerous problems: 1. Support Vector Machines Support Vector Machines are the most standard machine learning algorithm that is being used by the healthcare industry. It uses a supervised learning model for classification, regression, and detection of outlines. In recent years, the algorithm has been used to predict the medication adherence of heart patients that has helped millions avoid serious consequences, such as hospital readmission and even death. It is also being used for protein classification, image segregation, and text categorization. 2. Artificial Neural Networks It is a group of deep learning algorithms inspired by the neuron organization in animal brains that can receive signals from a previous layer and send it to the next layer. A network that can learn by analyzing examples or without any human intervention. From pathologists using it for diagnosis to biochemical analysis, an artificial neural network has a bunch of other uses as well. It is further divided into two parts – convolutional neural network (CNN) and a recurrent neural network (RNN). Imaging is an important aspect of medical science since it can allow a doctor to know about a disease even before the symptoms arise. Due to this, there are several screening procedures such as Pap smears, Mammograms, Colonoscopy, etc. CNN has proved to be crucial in this segment, as the algorithm is well suited for a multi-class classification problem and binary classification. On the other hand, RNN has proved to be significant when used for pattern recognition in medical time-series data analysis. 3. Logistic Regression This machine-learning algorithm is used to predict the current scenario of the categorical dependent variable through the use of predictor variables. It is often used for classifying and predicting the probability of an event, such as disease risk management, which assists doctors in making critical medical decisions. It also helps medical institutions target patients with more risk and curate behavioral health plans to improve their daily health habits. 4. Random Forest The algorithm is used to construct multiple training trees at training time for performing classification and regression, and also helps overcome the problem of decision trees’ overfitting. Based on a patient’s medical history, Random Forest is used to predicting the risk of disease and for ECG and MRI analysis. 5. Discriminant Analysis Discriminant Analysis is a machine learning algorithm that is used to analyze the adequacy of object classification, and also for assigning one object to single or many groups. From the early diagnosis of diabetic Peripheral Neuropathy to refining the diagnostic features of blood vessel images, discriminant analysis is applied in the healthcare industry. It is also used for electronic health record management systems and to detect signs of mental health disorientation. 6. Naïve Bayes Based on the Bayes theorem, this is one of the most efficient machine learning algorithms ever known to mankind and is highly used by the healthcare industry for medical data clarification and disease prediction. When it comes to data mining, classification can be termed as data analysis, which is often used to extract models describing data classes. Since the probability distribution is high, Bayes classifier can achieve an optimal result.","excerpt":"Artificial intelligence (AI) has been integrating remarkable developments in almost every sector, and the healthcare industry is no different. There is an unprecedented amount of data flowing in the healthcare industry, which remains unlabeled and inconsistent, which can be analyzed to extract valuable insights using AI. The technology is also being used to provide virtual […]","categories":["AI Trends"],"tags":["AI Algorithms","AI Healthcare","decision tree algorithm","healthcare ai","naive bayes"],"author_name":"Rohit Chatterjee","publish_date":"2020-04-18T19:00:05","publication_year":"2020","word_count":635,"keywords":["Go","healthcare ai","decision tree algorithm","artificial intelligence","AI","AI Healthcare","machine learning","neural network","naive bayes","AI Algorithms","RAG","deep learning","GAN","CNN","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","deep learning","neural network","RAG","R","Go","GAN","CNN"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-6-ai-algorithms-in-healthcare\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10077811,"title":"Yubi organises ‘2022 &#8211; Women in Data Summit’: ​​A Ringside View with Data Power Players","content":"Whether in a data science domain or any other industry, having a role model has always been the key driver for shaping behaviour and decisions. This makes it even more important for the women trying to create a mark in this dynamic analytics domain. Having a positive role model can always help visualise success and influence enough to strive to achieve goals. We strongly believe in the power of personal stories & role models. To that end, Yubi (formerly CredAvenue), in association with Analytics India Magazine, is organising the Women in Data Summit, where we are bringing in some industry stalwarts of the data & analytics domain along with women professionals who are trying to make a mark in this rapidly growing field. Scheduled for November 02, 2022, in Bengaluru at Radisson Blu Bengaluru Outer Ring Road, the Women in Data Summit will touch upon the riveting personal stories of some of the women leaders in our country’s data & analytics domain. The idea is to talk about the importance of role models in the data science industry & how one can draw inspiration from them. This discussion is aimed to inspire, build connections, drive change in the community & think of out-of-the-box solutions around breaking the bias. The session will also cover how organisations in today’s global economy leverage out-of-the-box solutions to ensure diversity in their team. Location:Radisson Blu Bengaluru Outer Ring Road Date:November 2, 20226 PM onwards Save your spot More about Women in Data Summit The event will start with a keynote address by Mathangi Sri, Chief Data Officer at Yubi. Mathangi is one of India’s top 50 Influential AI Leaders, with 20 patents under her name. Further, she has been recognised by INBA as one of the top 100 women in India. She also has a proven track record of building world-class data sciences solutions and teams. Following that will be an engaging presentation by Abhishek Mehrotra, Chief Human Resources Officer at Yubi. Abhishek is an HR Leader with two decades of journey across technology & engineering\/R&D and social media domains. He has also led HR teams across the full spectrum of functions, including business partnering, L&D, C&B, TA, operations & compliances, including a stint in Japan as the Country HR. Post that, we will have a comprehensive roundtable discussion with some of the trailblazers of the data industry on “A Ringside View with Data Power Players: Come, Learn from the Industry Stalwarts”. A total of five leading personalities of the data & analytics domain, along with some 40+ women data scientists, will be attending the summit. And the discussion will be moderated by Karunakaran N, Vice President – of Data Engineering at Yubi. Why attend the Women in Data Summit: To learn about these inspirational women’s journeys and come up with out-of-the-box solutions for organisations to ensure diversity in their teams. REGISTER To understand the importance of role models & mentorship in your data science career growth. To network with some of the renowned women leaders of our country’s data & analytics domain and draw inspiration from their journey. Who should attend? – Women professionals with 5+ years of work experience in the analytics domain Date: 02nd November 2022Time: 06:00 PM onwardsVenue: Radisson Blu Bengaluru Outer Ring Road RSVP →","excerpt":"The roundtable discussion is aimed to inspire women professionals in the analytics domain to build connections, drive change in the community & think of out-of-the-box solutions around breaking the bias.","categories":["Deep Tech"],"tags":["Women in Tech"],"author_name":"Tasmia Ansari","publish_date":"2022-10-21T14:00:00","publication_year":"2022","word_count":545,"keywords":["data science","Go","API","AI","RAG","Aim","data engineering","Women in Tech","analytics","GAN","R"],"extracted_tech_keywords":["AI","data science","analytics","Aim","RAG","R","Go","API","data engineering","GAN"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/yubi-organises-2022-women-in-data-summit-a-ringside-view-with-data-power-players\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":40540,"title":"6 Reasons Why Developers Should Contribute More To Open Source","content":"Image source: Goddammitstacey Open source comes with the advantages of allowing the users to freely modify and integrate the work into their projects. It gives a way to freely exchange ideas with a like-minded community and contribute to the scientific and technological environment. According to the 2018 Open Source Program Management Survey, more than half of the companies surveyed, around 53 per cent said that their organisation has an open source software program or has plans to establish one within a year. Contributing to open source gives developers plenty of advantages, benefitting their own development. Here are 6 ways in which developers are benefitted by contributing to open source: 1.Helps in writing cleaner code: First up, there are a lot of enthusiasts who simply believe that code should be open. They’re idealists who want to make the world a better place, and it drives them to contribute code. The desire to share can be a powerful motivator. Even by fixing minor things like a bug in a library or writing a piece of documentation can also help the developers to write readable or maintainable code. They can independently suggest to the community and generally tend to stick by the rules of writing a code that is easy to understand. The fact that the code will be exposed to everyone naturally makes them write focus on making it readable. 2.Gain recognition: Some people just like the idea of the code being open to call and want to sincerely make the community better by contributing to it. It also gives these coders a recognition among the community, and they also have personal development in terms of sharpening their programming skills. Getting recognition is also a reason to contribute to open source. Recognition can also end up with a lot of opportunities by potential employers. 3.A better understanding of technology: Companies and individuals who contribute to an open source project get to know the technology at a much deeper level. It helps in strengthening the understanding of the project. It helps in developing the internal use of tech in the team or organisation. 4.Helps in being prepared for the project: Contributing to open source projects that are related to the company’s domain will help the company understand its customers better. It helps them understand what their customers expect and what can be done to achieve that. This can be worked out by being a part of the open source community and contributing to the code. It will help the company have a clear view of the projected future and work accordingly. 5.Reduces development costs: Open source contribution is a huge help to reduce development costs in organisations. Open source contributors give the organisation to take ideas, suggestions and help from experts outside of their own team, thereby reducing development costs because then they have access to the work done by other developers that could be better than the ones in the organisation. 6.Builds the organisation’s reputation: Such contribution helps in keeping the developers challenged and active participation from the employees will help in building the organisation reputation. Active participation in stimulating open source communities helps in the long run to achieve this. Edgar Magana, a Senior Principal Engineer for Cloud Operations at Workday, who is also a hiring manager, had written in his blog stating: “As a hiring manager, a job candidate who has a real understanding of open source and demonstrates active and positive participation in open source projects is worth far more than a candidate who has the right certifications and credentials but no open source experience.”","excerpt":"Open source comes with the advantages of allowing the users to freely modify and integrate the work into their projects. It gives a way to freely exchange ideas with a like-minded community and contribute to the scientific and technological environment. According to the 2018 Open Source Program Management Survey, more than half of the companies […]","categories":["AI Trends"],"tags":["code","Developers","Open Source"],"author_name":"Disha Misal","publish_date":"2019-06-11T05:34:56","publication_year":"2019","word_count":596,"keywords":["Go","Open Source","programming_languages:R","AI","programming_languages:Go","code","GAN","R","Developers"],"extracted_tech_keywords":["AI","R","Go","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/6-reasons-why-developers-should-contribute-more-to-open-source\/","complexity_score":2,"technical_depth":6,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":58493,"title":"What Does Regulating AI Really Mean?","content":"As the world of artificial intelligence (AI) evolves, there is a growing need for rules that enable these technologies to operate safely. And while it may be critical to place AI within a legal and regulatory environment, it would be just as important to address the concerns this could pose in stymieing innovation and growth. Unlike promising technologies from a few decades ago, the AI industry has expanded so rapidly that it has been largely left out of the regulatory framework. This has allowed companies free rein to use it as they deem fit, creating suspicion in the public eye — a trend that industry players like SpaceX CEO Elon Musk and Google CEO Sundar Pichai have taken cognisance of. Thus, an AI regulation framed in a way that provides competitive advantages to these companies, while protecting consumers from new risks is the need of the hour. Data Protection & Privacy Since AI systems are fueled by data, companies are increasingly using customer information — with or without their knowledge. This means that without proper regulation, the responsible use of AI by companies is largely predicated on trust. This is not enough. What is more, not just organisations, but countries across the world are looking for a competitive advantage in AI — one that will allow them to strengthen their industrial ecosystem and military prowess, among others. Given the difficulty in even establishing a clear definition of AI — one that is commonly accepted by those who operate in this ecosystem — a universal law that stretches from company t0 company, and nation to nation, will be impractical. Instead, effort needs to be directed towards growing local AI regulatory ecosystems. Although some headway has been made by the EU towards setting AI policy principles with the General Data Protection Regulation (GDPR), more participation is required to help safeguard the use of data. Creating An AI Strategy One way governments can accomplish this is by inviting industry experts to educate themselves on how AI is being deployed and the challenges it is likely to face. However, even if the onus of laying out an AI policy rests on governments, the private sector should initiate these conversations, rather than allowing less-informed legislators to set the ball rolling. A good place to start will be to come to a consensus on the definition of AI. Following this, they should provide informed insights and articulate their needs and demands. Concurrently, they should also demonstrate how AI applications can improve the productivity of government activities. This includes how AI can protect sensitive information and mitigate issues in the face of emerging cyberattacks, augment effective decision-making, as well as automate certain processes. India’s AI Story India’s strategic positioning with respect to AI initiatives has been inadequate. This is reinforced by the fact that it is lagging behind many countries in this space, including France, China, Israel, UK, US, Canada, Germany, Japan and South Korea. According to a report released by Niti Aayog, lack of regulations around anonymisation of data has stood in the way of India truly embracing the benefits of AI. To counter this, India needs to identify key governance issues related to AI and propose relevant policy remedies. For this, it can take some inspiration from the guidelines released by the US for the regulation of AI applications. The charter establishes the framework that any probable legislation can be built upon. What is more, by taking a sectoral approach to this, the US has demonstrated that it makes little sense in charting a one-size-fits-all path. In turn, this encourages sectoral regulators to formulate rules within their own jurisdiction, making the overall regulation process more effective. However, the real challenge will be to create rules that serve to protect consumers while promoting industry innovation. Rather than welcoming regulation reluctantly and perpetuating the belief that it could be a hindrance to development, India’s AI regulatory framework should capture both these nuances. What is more, this regulatory structure needs to be flexible and will have to be continually updated to reflect the understanding of new risks. Outlook AI is increasingly impacting all aspects of our daily lives. However, at the same time, we are also becoming more and more cognizant of the possible risks it can pose. Since the potential it holds is vast, mulling bans or prohibition on its applications is not the answer. Instead, policymakers need to chart a regulatory roadmap that encapsulates both consumer protection as well as innovation growth.","excerpt":"As the world of artificial intelligence (AI) evolves, there is a growing need for rules that enable these technologies to operate safely. And while it may be critical to place AI within a legal and regulatory environment, it would be just as important to address the concerns this could pose in stymieing innovation and growth. […]","categories":["AI Features"],"tags":["AI India","AI Regulation","AI Research","AI What it Does","Sundar Pichai"],"author_name":"Anu Thomas","publish_date":"2020-03-12T16:00:00","publication_year":"2020","word_count":746,"keywords":["Go","API","artificial intelligence","AI","innovation","AI Research","RAG","AI Regulation","GAN","AI What it Does","ViT","Rust","AI India","R","Sundar Pichai"],"extracted_tech_keywords":["AI","artificial intelligence","RAG","R","Go","Rust","API","GAN","ViT","innovation"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/what-does-regulating-ai-really-mean-for-companies-and-consumers\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":7957,"title":"S P Jain School of Management launches Certificate Program in Big Data and Analytics","content":"Analytics and Big Data are the hottest skillsets in the current market. As companies plan to adopt Analytics, there has been a sudden surge in requirement for analytics professionals. To meet this supply side talent deficit, several courses are being introduced by leading institutes and B-school worldwide. SP Jain joins in by introducing full time and part time Certificate program in Big Data and Analytics (BDAP) at its Mumbai campus. The course is being offered at S P Jain’s spanking new campus at Lower Parel, which boasts of State of the art facilities, in the heart of the city’s business center. BDAP is designed to explore, analyze and unravel the complex, unstructured data-driven world. The program kicks of with 10 core courses that build a strong foundation for the second stage of the program, which incorporates more in-depth and application-based learning. Given the need for specialist knowledge, it provides a range of courses in topics like data mining, machine learning, visualization techniques, predictive modeling, and statistics. The programme content builds on basic concepts, teaches tools and technologies that are currently prevalent in industry and progresses to cutting edge-topics like machine learning and Natural Language Processing. On completion of the program students would have learned to apply quantitative modeling and data analysis techniques to solve real world business problems, successfully present results using data visualization techniques, demonstrate knowledge of statistical data analysis techniques utilized in business decision-making, apply principles of Data Science to the analysis of business problems, use data mining software to solve real-world problems and employ cutting edge tools and technologies to analyze Big Data. The program is delivered by a faculty that is an equal mix of academicians and industry practitioners with a key proportion of overseas instructors thus lending the course a global perspective. “Big Data & Analytics is set to become the next game-changer in IT and will have a revolutionary impact on every business, whether big or small, in more than one way since it can provide inputs to improve businesses on all key dimensions.” – says Dr. Mahendra Mehta (Managing Director, Neural Techsoft), program director of BDAP. This program equips students to fill the need for sophisticated expertise in varied domains such as IT, Consulting, BFSI, Telecom and Media and in specializations like data mining, data modeling, data architecture, extraction, transformation, loading development and business intelligence development. The acquired techniques are increasingly essentially required for roles like Data Scientists, Analysts, Developers and Consultants. The programme is being offered in two formats. The full time format will have a total classroom contact hours of 400 and will be delivered on Weekdays for duration of 6 months. The Part time format will also have a 400 contact hours but will be delivered only on weekends for duration of 12 months. Participants can apply online: http:\/\/www.spjain.org\/big-data\/apply_online-mumbai.aspx For more information refer: http:\/\/www.spjain.org\/big-data\/index.aspx Watch the workshop conducted by SP Jain at CYPHER2015, Analytics India Summit [youtube url=”https:\/\/www.youtube.com\/watch?v=7tDy8LvX0UI”]","excerpt":"Analytics and Big Data are the hottest skillsets in the current market. As companies plan to adopt Analytics, there has been a sudden surge in requirement for analytics professionals. To meet this supply side talent deficit, several courses are being introduced by leading institutes and B-school worldwide. SP Jain joins in by introducing full time […]","categories":["AI Trends"],"tags":["data analytics certificate"],"author_name":"Дарья","publish_date":"2015-09-30T08:20:41","publication_year":"2015","word_count":488,"keywords":["big data","data science","business intelligence","machine learning","programming_languages:R","AI","data-driven","data analytics certificate","analytics","R"],"extracted_tech_keywords":["AI","machine learning","data science","analytics","R","big data","business intelligence","data-driven","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/s-p-jain-school-of-management-launches-certificate-program-in-big-data-and-analytics\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10117916,"title":"Intel Unveils the Most Efficient Gaudi 3 AI Accelerator at Intel Vision","content":"Intel has announced its latest AI chip Gaudi 3 at the Intel Vision 2024 event, in a bid to keep pace with the growing demand for semiconductors capable of training and deploying large AI models. The newly introduced Gaudi 3 chip, which was revealed by CEO Pat Gelsinger at Intel AI Everywhere event, boasts over double the power efficiency compared to its predecessor and is capable of running AI models 1.5 times faster than NVIDIA’s H100 GPU. It offers various configurations, including a bundle of eight Gaudi 3 chips on one motherboard or a card that can be integrated into existing systems. Gaudi 3, built on a 5 nm process, signals Intel’s utilisation of advanced manufacturing techniques. Additionally, Intel plans to manufacture AI chips, potentially for external companies, at a new Ohio factory expected to open in the coming years, according to Gelsinger. During testing, Intel evaluated the chip’s performance on models like Meta’s open-source Llama and the Falcon model by TII. Gaudi 3 demonstrated its capability to facilitate the training or deployment of models such as Stable Diffusion or OpenAI’s Whisper model for speech recognition. Compared to the NVIDIA H100, Intel anticipates that Gaudi 3 will achieve an approximately 50% faster time-to-train on average across Llama 2 models with 7B and 13B parameters, as well as the GPT-3 175B parameter model. While performance data for NVIDIA’s recently announced Blackwell-based B200 Tensor GPU is not currently available, it’s clear that NVIDIA’s latest offering would likely affect these performance comparisons significantly. In comparison to NVIDIA, Intel claims its chips consume less power. NVIDIA currently dominates approximately 80% of the AI chip market with its GPUs, which have been the preferred choice for AI developers in the past year. Intel asserts that its Gaudi 3 AI accelerator offers an estimated 50% enhancement in inferencing performance and around 40% better power efficiency compared to NVIDIA’s H100. Moreover, Intel states that it achieves these benefits at a significantly lower cost. Intel has announced that Gaudi 3 chips will be available to customers in the third quarter, with companies like Dell, HPE, and Supermicro set to incorporate the chips into their systems. However, Intel hasn’t disclosed the pricing details for Gaudi 3. Das Kamhout, vice president of Xeon software at Intel, expressed confidence in Gaudi 3’s competitiveness against NVIDIA’s latest offerings, citing factors such as competitive pricing and the incorporation of an open integrated network on chip. The data centre AI market is expected to expand as cloud providers and businesses invest in infrastructure for deploying AI software, indicating opportunities for other players in the market. While NVIDIA has seen significant stock growth driven by the AI boom, Intel’s stock has experienced more modest gains. Nevertheless, Intel remains determined to compete in the AI chip market, with AMD also seeking to expand its presence in the server AI chip segment. NVIDIA’s success has largely been attributed to its proprietary software suite, CUDA. In contrast, Intel is collaborating with chip and software giants like Google, Qualcomm, and Arm to develop open software solutions, aiming to provide greater flexibility for software companies in selecting chip providers. In addition, Intel unveiled its intention to create an open platform for enterprise AI, aiming to expedite the deployment of secure GenAI systems empowered by retrieval augmented generation (RAG).","excerpt":"Intel anticipates that Gaudi 3 will achieve an approximately 50% faster time-to-train on average across Llama 2 models, when compared to NVIDIA H100.","categories":["AI News"],"tags":["AI Vision","Intel"],"author_name":"Mohit Pandey","publish_date":"2024-04-09T21:55:01","publication_year":"2024","word_count":548,"keywords":["CUDA","Go","GenAI","OpenAI","AI","AI Vision","RAG","Aim","retrieval augmented generation","R","Intel"],"extracted_tech_keywords":["AI","GenAI","OpenAI","Aim","retrieval augmented generation","RAG","CUDA","R","Go","CUDA"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/intel-unveils-the-most-efficient-gaudi-3-at-intel-vision\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":16698,"title":"How To Crack Open An Interview At Fractal Analytics?","content":"Fractal Analytics founded in 2000 by Srikanth Velamakanni, Pranay Agrawal, Nirmal Palaparthi, Pradeep Suryanarayan and Ramakrishna Reddy is an analytics service provider headquartered in New Jersey. They operate in nine countries globally and have over 16 offices worldwide. While the registered and corporate headquarters in India is in Mumbai, and operational in Gurugram, in January 2015, they opened another office in Bengaluru and acquired Imagna Analytics and Mobius Innovations the same year. The most recent acquisition being 4i, a Chicago-based strategy & analytics firm this year. In 2016, Fractal Analytics appointed Pranay Agrawal as the CEO and also expanded its operations by creating two new subsidiaries Qure.ai and Cuddle.ai. If you are a data scientist and are looking to crack an interview at Fractal Analytics, we have got the formula for you. To get a better understanding and to make your job a tad bit easier, AIM caught up with the company’s head of staffing, Tejas Sanghavi. The first thing to crack an interview is to pose technical excellence, execution excellence and ace communication. But before we get to the technical side of it, let us understand the most general aspects every candidate must look at before applying for a job. Research and gather resources: Make sure you get a fair idea about what the company does and what its work culture is like. Browsing through their websites, blogs, social media pages is also a great way to gather information. Or simply, just Google the company. A well-informed candidate always has an edge over the others. Speaking of work culture, Sanghavi says: “Fractal’s culture and people principles are built on trust, transparency and freedom. In this high trust and high performance culture, only people with a strong work ethic, can-do attitude, result orientation and a high degree of integrity & professionalism can succeed.” By now, you are probably already asking what are the technical skills required to qualify for a job at Fractal Analytics. “At the entry level, we look for communication skills, structured thought process and ability to work with team and clients. You must have a learning orientation, and develop expertise very quickly in at least two of these tools – R, Python, SQL and Tableau. You must be able to demonstrate your passion for analytics – some examples of exciting code\/projects done\/courses taken\/participation in open competitions. Every Fractalite develops themselves along one of the 4 career tracks of Fractal’s Analytics Talent Grid,” says Sanghavi. Every employee of the company needs to develop himself around one of the four career tracks of Fractal’s Analytics Talent Grid. What are the technical skillsets required for the different profiles at Fractal Analytics? Analytics Consulting and Business Intelligence: R, Python, SQL, Tableau, Qlikview, Hadoop, Hive, modeling. Data scientist: Experts at machine learning, deep learning and artificial intelligence. They need to be able to explain the first principles behind these techniques and use these techniques in business situations. Big data engineering: Experts at implementing big data solutions in live projects – they need to be strong at Python \/ Java, Linux, Hive \/ SQL, MongoDB, Spark. The interview process: In most global companies, the hiring process is often long and tiring and the reason is simple – they want to bring on-board the best! Fractal has a very low acceptance rate of just 0.9% and candidates are made to go through five rounds of interviews before being selected. Aptitude test – English \/ Analytical Reasoning \/ Quantitative techniques Technical test – R \/ SQL \/ Python \/ Tableau Technical interview: Assess the person’s understanding of the subjects learnt, curricular and extra-curricular projects and internship work. Experienced candidates have to explain the technical framework of the projects that they have worked on. Problem-solving interview: Tests one’s ability to think logically from all aspects, structure the problem, give a comprehensive solution and explain the same to the interviewers. Experienced candidates should also have a good understanding about the industry that they work in – common challenges, analytical solutions and disruptions in the industry. Cultural fit interview: Irrespective of the level of hiring, the company believes that an interview is as much about the candidate selecting Fractal as much as Fractal selecting them. “A candidate’s interview process, to the extent possible, is conducted on a single day (called a day plan). This is to maximise candidate experience, and increase the speed of decision making,” Sanghavi adds. Explaining what it really takes to be a part of the system, Sanghavi also says that values, cultural fit, passion for analytics, learnability, and then for skills and experience remain the prime focus while finding the ideal candidate. As a bonus, Sanghavi finally says: “While giving the interview Understand the question of the interviewers before answering Answer all questions clearly and concisely Ask questions to the interviewer that are relevant to you.” We hope this article will help you land your next analytics job. Good luck! and let us know any more tips or share your experiences in the comment box below.","excerpt":"Fractal Analytics founded in 2000 by Srikanth Velamakanni, Pranay Agrawal, Nirmal Palaparthi, Pradeep Suryanarayan and Ramakrishna Reddy is an analytics service provider headquartered in New Jersey. They operate in nine countries globally and have over 16 offices worldwide. While the registered and corporate headquarters in India is in Mumbai, and operational in Gurugram, in January […]","categories":["AI Highlights"],"tags":["analytics interview","Fractal Analytics","fractal analytics careers"],"author_name":"Priya Singh","publish_date":"2017-08-03T06:24:09","publication_year":"2017","word_count":832,"keywords":["Fractal Analytics","artificial intelligence","machine learning","AI","MongoDB","Python","Ray","Aim","deep learning","fractal analytics careers","analytics","analytics interview","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","deep learning","analytics","Aim","Ray","MongoDB","Python","R"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/crack-open-interview-fractal-analytics\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10070259,"title":"Hey Alexa, take me to the moon","content":"From C3PO in Star Wars to TARS in Interstellar, AI in space has been a favourite McGuffin of Hollywood filmmakers. Now, Amazon is all set to turn this fantasy into reality by sending its flagship AI assistant, Alexa, to the moon as part of NASA’s Artemis program. The company is collaborating with American aerospace giant Lockheed Martin and tech conglomerate Cisco for the same. Aaron Rubenson, VP of Alexa everywhere at Amazon, said the inspiration for Alexa’s space mission came from the famous movie Star Trek and Amazon is working with Lockheed Martin to go beyond the limits of voice technology and AI. He hopes Alexa’s debut in the Artemis program will inspire future astronauts, engineers and scientists to define the next era of space exploration. Alexa has made life easier for customers on Earth…excited to explore what voice & AI can do for astronauts in space. Great partnership with @LockheedMartin & @Cisco for @NASA’s upcoming #Artemis mission! https:\/\/t.co\/tUfXSJngcH— Andy Jassy (@ajassy) January 5, 2022 The Artemis program The Artemis program started at the fag end of 2017 to revitalise the US space program. It aims to replicate the first human-crewed landing mission since Apollo 17 in 1972 and is planning its first touchdown on the lunar south pole by 2025. NASA’s short-term goal is to take the first woman and first person of colour to the moon. The mid-term objectives include establishing an international expedition team and a sustainable human presence on the Earth’s only natural satellite. The long-term goal is to set the stage for the extraction of lunar resources and eventually make manned missions to Mars and beyond. Though originally envisioned as an international collaborative project between governmental space agencies and private players, the Artemis mission is carried out by NASA and local US commercial contractors. After a successful wet dress rehearsal, the first mission (Artemis I) will launch late in August this year. Alexa on board Artemis I will be an uncrewed test of the SLS and Orion and is the first test flight for both crafts. Alexa is one of the innovative technologies that will be tested as part of Artemis I. The integration will help those involved explore how ambient intelligence can assist astronauts on future missions. ​​Alexa envisions a future where astronauts can turn to an onboard AI for information, assistance, and companionship. Lockheed Martin will design the hardware for Callisto, ensuring the system can withstand elements like intense shock and vibrations as well as radiation exposure (when passing through the Van Allen radiation belts) in space. Amazon furnished the acoustic and audio processing software to support far-field voice interactions through Alexa, tuning algorithms to account for noise from engines and pumps and the reverberation from metallic surfaces within the cabin. The Callisto module will be equipped with Amazon’s local voice control technology, allowing Alexa to function with limited or no connectivity. In addition, integrating Alexa’s algorithm with the local processing unit on board the Callisto will bypass the latency of sending information from the moon to Earth and back. PR stunt? The Callisto module has been thoroughly tested on the ground, but a lot remains unanswered about its performance in space. One of the biggest challenges for Alexa will be a questionable internet connection. While the traditional Alexa requires an internet connection for most of its functions like answering queries and retrieving information, Aaron Rubenson, VP of Amazon’s Alexa program, said the version of Alexa being integrated into the Callisto module had been reconfigured to function largely on offline mode. So, theoretically, Alexa can provide information like flight trajectory, telemetry, etc, but not up-to-date information from Earth while offline. Since the Orion spacecraft is already equipped with cutting-edge systems to provide instant data, you wonder how much help a virtual assistant can offer. Meanwhile, Amazon is rolling out a new feature on Alexa. By using the command “Alexa, take me to the Moon,” users can get live updates on the Artemis I mission.","excerpt":"Alexa envisions a future where astronauts can turn to an onboard AI for information, assistance, and companionship.","categories":["AI Features"],"tags":["Alexa","Amazon","cisco"],"author_name":"Kartik Wali","publish_date":"2022-07-01T16:00:00","publication_year":"2022","word_count":661,"keywords":["Replicate","Go","ELT","programming_languages:R","AI","Amazon","programming_languages:Go","Aim","ViT","Alexa","R","cisco"],"extracted_tech_keywords":["AI","Aim","R","Go","ELT","ViT","Replicate","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/hey-alexa-take-me-to-the-moon\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10087176,"title":"South Korea Enters AI Chip Race with Rebellions.ai","content":"South Korean startup Rebellions.ai launched its artificial intelligence chip on Monday. With this launch, the company that builds AI accelerators to bridge the gap between silicon architectures and machine learning algorithms, puts itself in the AI race and aspires to grab government contracts. Going on an ambitious route, Rebellions.ai, a Seoul-based company, founded in 2020, launched its latest ATOM chip to compete against NVIDIA, the American corporation that builds GPU and SoC products, which is the world’s 8th most valuable company by market cap. Park Sunghyun, the co-founder of Rebellions.ai, said that the ATOM chip is built to run computer vision and AI chatbot applications. Compared to an NVIDIA A100 chip, the ATOM chip consumes only 20% of the power consumed by the former as it targets specific tasks rather than a large range. Kim Yang-Paeng, a senior researcher at Korea Institute for Industrial Economics and Trade, believes that, though it is hard to catch up with NVIDIA in the AI chips race, the versatility of AI chips allows multiple functions and that there “aren’t set boundaries or metrics”. NVIDIA Corp stocks have climbed 52% in 2023 and are expected to grow further. The South Korean government wishes to bolster research and development in the AI chip industry as it looks to improve its market share in domestic data centres. The government will invest close to $800 million over the next five years, and the demand for AI chips is set to take up 33% of the overall system chip demand by 2030. To maintain a strong AI ecosystem, countries must have access to semiconductors to support in-house data centres and public cloud computing, as a shortage of semiconductors will disrupt the flow. The US, China and Taiwan are established players in the semiconductor market, and Japan and Korea are slowly building their space. The global semiconductor manufacturing capacity share of South Korea reached 21% in 2022 from zero in 1990. With Rebellions.ai working on AI chips, the race to establish itself as a global player is seamlessly set. The company has raised over 112 billion won to date.","excerpt":"With this, the Seoul-based startup Rebellions.ai aspires to take on the likes of NVIDIA, Intel, etc.","categories":["AI News"],"tags":["AI chip","GPU","NVIDIA","Semiconductor India","south korea"],"author_name":"Vandana Nair","publish_date":"2023-02-13T12:49:43","publication_year":"2023","word_count":349,"keywords":["Go","artificial intelligence","AI chip","machine learning","AI","cloud computing","startup","ML","A100","GPU","computer vision","Semiconductor India","NVIDIA","R","south korea"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","computer vision","cloud computing","R","Go","startup","A100"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/south-korea-enters-ai-chip-race-with-rebellions-ai\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10140688,"title":"Google Unveils ReCapture to Revolutionise Video Modeling","content":"Google released ReCapture, its generative video camera controls for user-provided videos, on Friday. Hugging Face’s ML growth lead, Ahsen Khaliq, took to X to announce this feature using masked video fine-tuning. Google presents ReCaptureGenerative Video Camera Controls for User-Provided Videos using Masked Video Fine-Tuning pic.twitter.com\/5qT386e5K0— AK (@_akhaliq) November 8, 2024 Nataniel Ruiz, a senior research scientist at Google and others posted this launch on Hugging Face the same day. The company introduced ReCapture as an innovative method that transforms user-provided videos and creates new, dynamic camera perspectives. Unlike prior advancements, which were limited to generating videos, ReCapture can recreate any existing video with enhanced angles and cinematic camera motion, retaining the original scene’s movements. This update is steps ahead of what other text-to-video generation tools are creating. This enters an entirely new field of video-to-video generation. This breakthrough means users can view their footage from entirely new, realistic vantage points without needing to shoot from multiple angles. ReCapture works in two stages. First, it creates a rough ‘anchor’ video with a new camera perspective using multiview diffusion models or depth-based point cloud rendering. Next, it applies a special ‘masked video fine-tuning technique’ to enhance the anchor video, making it clear and consistent over time. This process results in a smooth, re-angled video that can even generate unseen parts of a scene. As reported by AIM earlier, multiple people and companies are experimenting with AI video generation by using tools like Midjourney, RunwayML, Soundful, ElevenLabs, ChatGPT, and CapCut. ReCapture could change the way people use such tools. Even in the field of generative video games, which will see an immense boom in 2025, the integration of such tools could completely change the game. A step ahead of video editing, ReCapture sets a whole new standard for video realism, taking it miles ahead of existing models. It showcases how AI is enhancing not just the creation but also the reimagination of visual content. This development could redefine the use of video in media production, allowing for creative storytelling and enhanced user engagement.","excerpt":"ReCapture, as it is named, can take any existing video and recreate it with enhanced angles and cinematic camera motion.","categories":["AI News"],"tags":["AI Video Generation Models","Google"],"author_name":"Sanjana Gupta","publish_date":"2024-11-08T16:15:40","publication_year":"2024","word_count":340,"keywords":["Go","ChatGPT","Hugging Face","AI","ML","diffusion models","AI Video Generation Models","GPT","Aim","Google","R","llm_models:GPT"],"extracted_tech_keywords":["AI","ML","ChatGPT","Aim","Hugging Face","R","Go","GPT","diffusion models","llm_models:GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/google-unveils-recapture-to-revolutionise-video-modeling\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":7117,"title":"Starting an Analytics Function: A five step framework","content":"“How do we start an analytics practise?” – I have often been asked this question. While there is no one size fit all solution to it and various organizations have achieved success through various models. We have come up with a comprehensive 5 step framework to seamlessly get started with an analytics function, that mitigates risks and exemplify analytics output. These steps could be taken up internally by businesses, though we recommend employing external experts that bring in a third eye perspective to your business. Identify Opportunities Even before you acquire skillsets and set aside your yearly budget for analytics, know what is needed. Get an assessment of your current data landscape and key use cases that are helpful for your business success. This is essential to identify appropriate sponsorship and acquire right skills. Select projects that are not only interesting and valuable but also small enough to quickly capture value from analytics. Your early use cases need not be those bulky implementations that needs collaboration from various functions in your organization, or are too complex to be understood by business. Ascertain Sponsorship Most data science initiatives fail not because of lack of skillsets or use cases, but due to lack of right management focus on it. This is primarily confronted when analytics insights fail to move from lab to actual implementation. Most of the times, business managers lack right focus and knowledge to incorporate data science insights into their operational strategies. Implementation of analytics is mostly seen as a recruiting bunch of data scientists that churn out interesting insights and visualizations. Yet, success in analytics is more than that. Most importantly, it is the knitting of analytics in the cultural fabric of the organizations, where there is a seamless exchange of ideas and insights. On-board Talent & Tools This seems too obvious, but it’s difficult in practice. Data science is a rare skill to find in market currently. This is mostly due to fact that it’s a combination of multiple interdisciplinary subskills, each of which contributes equally to the overall skillset – knowledge of business processes, data, statistics, visualization and design thinking, scientific and research methodologies, software programming and above all a consulting bend of mind. Organizations for long have looked at these multiple sub-skills discreetly and devised their analytics hiring strategy around each of these. So, organizations hired visualization experts separate from programmers or number crunchers. This works well to quickly ramp up large teams but in long run creates operational bottlenecks and increase investments on key deliverables. A much contemporary style of hiring data scientists is to look at professionals that loosely touch upon all (or combinations) of these sub-skills and then train them to specific needs of the organization. Formalize Governance Process Continuous success out of analytics practise involves putting up a centralized governance group. Most organizations are too focussed on deliverables, leading to gradual degradation of vision and quality over time. Governance involve setting up an internal standards and controls policy. Call it internal regulation, which tracks quality of deliverables and their transformation to business value. Besides, governance also involves implementing accountability within data science teams. Most data science projects are result of complex collaborations between different stakeholders, leading to a confusion on who actually owns which part of the asset. Continuous & Inclusive Training Investment into training and upgrading overall skillsets within the team is well recognized by management. Yet more importantly, the two key themes here are ‘Continuous’ and ‘Inclusive’. Analytics training should start as soon as you decide to embark upon analytics journey. Training should not be seen as a one-time activity that is taken up once every quarter. The mode of training (internal or external) is non-essential when you start. What is of essence is that a continuous focus from management is required to upgrade the data science skillsets with the organization.","excerpt":"“How do we start an analytics practise?” – I have often been asked this question. While there is no one size fit all solution to it and various organizations have achieved success through various models. We have come up with a comprehensive 5 step framework to seamlessly get started with an analytics function, that mitigates […]","categories":["IT Services"],"tags":[],"author_name":"Дарья","publish_date":"2015-03-23T05:33:39","publication_year":"2015","word_count":641,"keywords":["data science","Go","TPU","AI","ETL","ML","ViT","analytics","GAN","R"],"extracted_tech_keywords":["AI","ML","data science","analytics","TPU","R","Go","ETL","GAN","ViT"],"url":"https:\/\/analyticsindiamag.com\/it-services\/starting-an-analytics-function-a-five-step-framework\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":29823,"title":"The Quantum Internet Is Still A Futuristic Dream, At Least A Decade Away","content":"You may have come across many discussions, arguments and ideas around quantum computing off late. Why is there so much buzz around this phenomenon? To answer it shortly, it is because of its enormous, unfathomable capabilities. This area of computing is millions of times more powerful than classical computing. By the next few years, quantum computing adoption across the world is said to grow exponentially. On the other hand, quantum computing has also spawned an interest in Quantum Internet. Let’s have a look at this form of ‘Internet’ and what holds for us in the near future. Fundamentally Different The quantum internet is unlike quantum computing. Although both of these areas rely on quantum mechanics, they are separate technologies. In the quantum internet, quantum mechanics plays an active role only in the background. It means that physics only forms a part of the working. So, there is an added advantage in the quantum internet — the possibility of fusing new functionalities at every step. Despite these developments, a quantum network\/internet has not been conceived to date on a global scale. For this to happen, breakthroughs with respect to implementation are to be seen. Contrastingly, scientists believe that with the rise of quantum computing, the quantum network can also be expanded. Earlier, in 2017, Chinese physicists worked on a quantum satellite to test the possibility of relaying qubits (quantum bits) — which is analogous to binary bits in digital computing. This satellite was using a concept called ‘entanglement’ in quantum mechanics to transmit information. By far, this is the most successful attempt in the quantum domain and has kickstarted the journey of the quantum internet on a wider scale. Where India Stands Regarding The Quantum Internet While research in the quantum internet is largely observed in major countries like the US and China, India is slowly catching up in this race. India has yet to have a rigorous well-established infrastructure when it comes to quantum computing itself. After all, a proper infrastructure framework is key to a successful technology. Even if research gives out new theoretical information regarding quantum computing, it has to be tried for its validity, and on top of that, bring breakthroughs specifically in quantum internet space. In our earlier article, we had highlighted a few instances where certain academic institutions as well the government in India came up with various projects such as Quantum Information Science and Technology (QuST), IISc’s Centre for Quantum Information and Quantum Computation (CQIQC) and so on. These are not only boosting research but also trying to bring in quantum tech into reality. A Global Challenge Quantum internet, if it has to be universal, is faced with a slew of challenges. At a global level, it encounters one critical problem: large distances for communication. Michael Lucy of Cosmos Magazine, says: “The major obstacle that must be overcome to create a global quantum network is in the ‘global’ part: long distances are a real problem. As entangled photons are beamed through the air or an optic fibre, they are slowly picked off by encounters with other particles. After at most a couple of hundred kilometres, 99.99 percent will be gone and the signal will be too weak to use for communication.” Since ‘quantum’ essentially deals with ‘energy in packets’, the problem of attenuation has to be taken care of. A study by from the Delft University of Technology has come up with a six-phase quantum network framework. This was actually what spiked interest in quantum internet in the recent days. By focusing quantum-based communications on a smaller region (cities in the Netherlands), a structured approach is implemented so that it could be deployed gradually to longer distances. Coming to this framework, the lowest stage — from phase 0 to phase 2 — concentrates on delivering qubits between two nodes (users) while the high-end stage – from phase 3, 4 and 5 –  exclusively works on establishing connections with quantum computers across the network thus enabling a distributed quantum computing environment. As you can see, here quantum computing paves way for quantum internet by leaning on a connected environment. Conclusion The above study is a prototype because the quantum internet is still in the making. Probably by the next decade, we may witness a revolution globally in terms of hassle-free connectivity and ultra-fast internet.","excerpt":"You may have come across many discussions, arguments and ideas around quantum computing off late. Why is there so much buzz around this phenomenon? To answer it shortly, it is because of its enormous, unfathomable capabilities. This area of computing is millions of times more powerful than classical computing. By the next few years, quantum […]","categories":["AI Features"],"tags":["countries with quantum computers","energy","india quantum computing","internet","quantum mechanics"],"author_name":"Abhishek Sharma","publish_date":"2018-10-31T11:36:03","publication_year":"2018","word_count":719,"keywords":["Go","energy","programming_languages:R","AI","countries with quantum computers","R","programming_languages:Go","Git","india quantum computing","ViT","quantum mechanics","emerging_tech:quantum computing","internet"],"extracted_tech_keywords":["AI","R","Go","Git","ViT","programming_languages:R","programming_languages:Go","emerging_tech:quantum computing"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/the-quantum-internet-is-still-a-futuristic-dream-at-least-a-decade-away\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":25683,"title":"ISRO eyes to visit Venus, NASA could partner with ISRO for this mission","content":"India will soon witness a new bold inter-planetary sojourn and the first of it kinds, as it plans to go to the Venus and this could be first ever ISRO-Venus mission. If reports are to be believed, Indian Space Research Organization (ISRO) is attempting to undertake its mega launch where it is intending to place a humongous 104 satellites in space in a single mission. India could become the first ever country to have tried to hit a century in a single mission. ISRO which plans to put 104 satellites on Feb 15, three would be from India and the remaining 101 from other countries. It is slated to be launched using Polar Satellite Launch Vehicle (PSLV). Israel, Switzerland, UAE, Kazakhstan, Netherlands and USA are some of the other countries who would be launching their satellites in the polar orbit. With India finishing its first mission the Mars in 2013, Indian astronomical journey has been for a major boost, a 23% increase in this year’s budget dedicated to space being one of the indications of this progress. The budget also mentions a provision for “Mars Orbiter Mission II and Mission to Venus”, giving a clear indication of the country’s interest on exploring Mars a second time and making a maiden journey to Venus. As India plans a second mission to Mars in 2021-22 and putting a robot (rover) on the surface of the Red Planet, as a part of another orbiter mission, India also plans to visit Venus, the second planet in the Solar system. It would be very much similar to the first mission to Mars. Reports suggests that NASA would be willing to partner with India’s mission to Venus. The voyage to Venus could be a worthwhile attempt as very little has been explored about the planet till now and it could open new avenues to the space missions. The talk between NASA and ISRO has already taken a step ahead and they are looking to jointly undertake studies on using electrical propulsion for powering this mission. Highlights: The maiden mission to Venus is speculated to be similar to Mangalyaan mission before. NASA is in talks with ISRO for the voyage to Venus This year’s budget has a mention if “Mars Orbiter Mission II and Mission to Venus” hinting the 23% rise in funds towards ISROs journey to Mars and Venus","excerpt":"India will soon witness a new bold inter-planetary sojourn and the first of it kinds, as it plans to go to the Venus and this could be first ever ISRO-Venus mission. If reports are to be believed, Indian Space Research Organization (ISRO) is attempting to undertake its mega launch where it is intending to place […]","categories":["AI News"],"tags":["ISRO India"],"author_name":"Srishti Deoras","publish_date":"2017-02-15T04:59:57","publication_year":"2017","word_count":393,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","GAN","ISRO India","R"],"extracted_tech_keywords":["AI","R","Go","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/isro-eyes-visit-venus-nasa-partner-isro-mission\/","complexity_score":2,"technical_depth":6,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10167126,"title":"This Company is India’s First to Export Packaged Semiconductors Under ISM","content":"RRP Electronics, a Mumbai-based electronics manufacturer, has become the first Indian company to export packaged semiconductors under the Government of India’s Semiconductor Mission (ISM). The company exported its first consignment, valued at ₹6.51 crore, primarily focused on ASICS-based technologies, to a European customer. It is a significant step in India’s semiconductor manufacturing sector, even as the customer details are unknown. This expansion aligns with the “Make in India” initiative, contributing to the nation’s growth and the state’s development. Launched in 2021 with ₹76,000 crore, the Semicon India program aims to promote the domestic semiconductor industry through incentives and partnerships. The export is part of RRP Electronics’ expansion initiatives under its ₹12,035 crore state-approved phase-I plan. The company operates a fully established Outsourced Semiconductor Assembly and Testing (OSAT) facility. RRP Electronics has established a first-mover advantage with the proof of concept. The company enhances its offerings through partnerships, including a tie-up with AMB Taiwan last year. The team, which is expected to arrive in the second week of April, will complete the proof of concept for SIM card modules, in addition to IGBT and MOSFET products. In addition to these developments, RRP Electronics is setting up an advanced packaging foundry with US-based Deca Technologies to produce parts for Apple iPhones. The company is also establishing an OHT facility to qualify for business from a multinational through Deca Technologies. Furthermore, RRP Electronics has begun exploring opportunities in the automotive sector with analog devices, which will create multiple job opportunities. As RRP Electronics continues to expand, it is backed by significant investments and partnerships. The company’s collaboration with international partners underscores its commitment to advancing India’s semiconductor industry. In early last month, RRP Electronics Ltd partnered with Singapore-based PTW to set up India’s first major silicon wafer production line. PTW, a global leader in refurbished frontend semiconductor tools, was facilitating the sale and technology transfer of a proven wafer line to RRP. The plant, once fully operational, was projected to generate over $1.2 billion in revenue with gross margins above 60%. PTW also aimed to sell another two lines to Indian customers within 1 to 2 years. Pilot production was expected to begin in 2025–26, with full capacity achieved within two years. This move would reduce India’s reliance on electronics imports and support the country’s ‘Make in India’ ambitions. The strategic partnership was being advised by Singapore consulting firm TOP2.","excerpt":"The company exported its first consignment, valued at ₹6.51 crore to a European customer.","categories":["AI News"],"tags":["India semiconductor mission","Make in India","semiconductor"],"author_name":"Sanjana Gupta","publish_date":"2025-04-02T15:39:39","publication_year":"2025","word_count":397,"keywords":["Go","Make in India","programming_languages:R","AI","India semiconductor mission","semiconductor","programming_languages:Go","Aim","R"],"extracted_tech_keywords":["AI","Aim","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/indias-first-company-to-export-packaged-semiconductors-under-the-semicon-mission\/","complexity_score":3,"technical_depth":6,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10114227,"title":"GitHub Copilot Enterprise is now Generally Available at $39\/ Month","content":"GitHub recently announced that GitHub Copilot is now generally available. Since the early days of GitHub Copilot, customers have asked for a copilot that is customised to their own organization’s code and processes. “With GitHub Copilot Enterprise we’re bringing the industry’s premier AI developer tool available to every organisation for just $39 per user per month,” the company said in a press release. Developers spend more time deciphering rather than shipping when they can’t pinpoint and solve the issues, bugs, or vulnerabilities that are unique to their organization’s codebase. Developers often write code for only a couple hours a day and, instead of being creative, are bogged down with mundane tasks throughout their day. The inaccessibility of institutional knowledge acts as a blockade preventing developers from fully exercising their creativity and building more for you. Just by integrating generative AI into the editor, GitHub Copilot has quickly defined a new age of software development, resulting in clear gains of developer productivity and happiness. GitHub is today bringing the next frontier of developer tools with the general availability of GitHub Copilot Enterprise–a companion that places the institutional knowledge of your organization at your developers fingertips. Now, team members can ask questions about public and private code, get up to speed quickly with new codebases, build greater consistencies across engineering teams, and ensure that everyone has access to the same standards and work that’s previously been done. “As the technology landscape continues to rapidly evolve, we are expanding capabilities of GitHub Copilot to not only understand your own internal knowledge bases but to bring in the latest information from the internet as well. By integrating Bing search directly into Copilot Chat—available in beta for GitHub Copilot Enterprise—you can find the latest software development-related information like updates to CSS or JavaScript frameworks. This means GitHub Copilot can now help your developers explore their curiosity and gain outside knowledge near instantly, at scale,” Thomas Dohmke, CEO at GitHub said. As the global home for all developers, GitHub is the world’s leading AI-powered developer platform to build, scale, and deliver secure software. Over 100 million people, including developers from 90 of the Fortune 100 companies, use GitHub to build amazing things together across 420+ million repositories.","excerpt":"Copilot Enterprise integrates chat directly into GitHub.com, enabling developers to ask questions and receive answers in natural language on your codebase","categories":["AI News"],"tags":["GitHub","GitHub Enterprise"],"author_name":"Pritam Bordoloi","publish_date":"2024-02-28T09:46:07","publication_year":"2024","word_count":371,"keywords":["API","GitHub Enterprise","AI","Git","ViT","generative AI","JavaScript","GAN","GitHub","R","Java"],"extracted_tech_keywords":["AI","generative AI","R","JavaScript","Java","Git","GitHub","API","GAN","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/github-copilot-enterprise-is-now-generally-available-at-39-month\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10171949,"title":"Infosys and Adobe Partner to Fast Track Marketing Transformation with AI","content":"Infosys and Adobe have announced a strategic collaboration to transform the marketing life cycle of global brands using artificial intelligence (AI). The partnership aims to unify customer experiences at scale, personalise content to drive growth, and streamline workflows for improved efficiency. Infosys Aster, a suite of AI-powered marketing solutions, and Adobe’s industry-leading marketing tools will combine to help marketers create personalised, data-driven customer journeys across every touchpoint. “CMOs know that AI can be their partner in propelling business growth while helping them navigate the complexities of scaling deeply personalised content and unifying the brand experience,” Sumit Virmani, EVP and global chief marketing officer, Infosys, said. On the other hand, Adobe’s Real-Time Customer Data Platform (CDP) and GenStudio empower brands to manage content creation, segmentation, and experience delivery. Infosys Aster adds AI-powered agility to marketing, helping businesses respond to customer behaviour, market trends, and evolving business needs. The collaboration makes it easier to launch hyper-targeted personalised campaigns, leveraging Adobe’s technologies to deliver real-time promotions and pricing based on customer behaviour. The integrated solution also incorporates autonomous agents that plan, execute, and optimise marketing workflows. “In an attention-based economy where consumers and businesses are inundated with content across every channel, impactful creative personalised at scale is what will enable marketers to break through,” Anil Chakravarthy, president of digital experience business at Adobe, added. Douglas Hayward, senior research director at IDC, added, “Today’s CMOs need AI-enabled tools that understand consumers and business customers as individuals with context-specific needs.”","excerpt":"Infosys Aster and Adobe’s marketing tools will help marketers create personalised, data-driven customer journeys.","categories":["AI News"],"tags":["Infosys"],"author_name":"Shalini Mondal","publish_date":"2025-06-18T17:13:12","publication_year":"2025","word_count":245,"keywords":["artificial intelligence","Infosys","AI","autonomous agents","data-driven","ML","programming_languages:R","Git","RAG","Aim","R"],"extracted_tech_keywords":["AI","artificial intelligence","ML","Aim","RAG","autonomous agents","R","Git","data-driven","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/infosys-and-adobe-partner-to-fast-track-marketing-transformation-with-ai\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10021930,"title":"Guide to TensorFlow Extended(TFX): End-to-End Platform for Deploying Production ML Pipelines","content":"Ever since Google has publicised Tensorflow, its application in Deep Learning has been increasing tremendously. It is used even more in research and production for authoring ML algorithms. Though it is flexible, it does not provide an end-to-end production system. On the other hand, Sibyl has end-to-end facilities but lacks flexibility. Google then came up with Tensorflow Extended(TFX) idea as a production-scaled machine learning platform on Tensorflow, taking advantage of both Tensorflow and Sibyl frameworks. TFX contains a sequence of components to implement ML pipelines that are scalable and give high-performance machine learning tasks. These components can also be used independently. Apache Airflow and Kubeflow Pipelines support TFX. TFX components interact with  ML Metadata as a backend that keeps a record of component runs, input and output artifacts, and runtime configuration. This metadata backend enables advanced functionality like experiment tracking or warm starting\/resuming ML models from previous runs. Compatible versions of TFX can be found here. TFX’s standard component can be used in the pipeline or individually and provides functionalities to get started with Machine Learning. The diagram below indicates the data flow between different parts. You can learn about various standard features here, in great detail. TFX contains different python packages\/libraries to create pipelines such as TensorFlow Data Validation (TFDV), TensorFlow Transform (TFT), TensorFlow Model Analysis (TFMA), etc. The image below demonstrates the link between TFX libraries and pipeline components: https:\/\/twitter.com\/TensorFlow\/status\/1369377456000557058 Installation You can install TFX via PyPI. !pip install tfx Demo of TFX This demo is a component-by-component tutorial of TFX via Keras API. For example, we have taken the Chicago Taxi example. Note : TFX supports the Tensorflow 2 version of Keras. Import all the necessary packages and modules. The code is available here.Check the library versions: print('TensorFlow version: {}'.format(tf.__version__)) print('TFX version: {}'.format(tfx.__version__)) Set up the pipeline paths as shown below: import tfx.examples.chicago_taxi_pipeline # This is the directory containing the TFX Chicago Taxi Pipeline example. _taxi_root = tfx.examples.chicago_taxi_pipeline.__path__[0] # This is the path where your model will be pushed for serving. _serving_model_dir = os.path.join( tempfile.mkdtemp(), 'serving_model\/taxi_simple') # Set up logging. absl.logging.set_verbosity(absl.logging.INFO) Download the dataset. Here the dataset is Taxi Trips dataset released by the City of Chicago. _data_root = tempfile.mkdtemp(prefix='tfx-data') DATA_PATH = 'https:\/\/raw.githubusercontent.com\/tensorflow\/tfx\/master\/tfx\/examples\/chicago_taxi_pipeline\/data\/simple\/data.csv' _data_filepath = os.path.join(_data_root, \"data.csv\") urllib.request.urlretrieve(DATA_PATH, _data_filepath) For making TFX component interactive, initialize the interactive context : context = InteractiveContext() Running TFX components: ExampleGen : This component is present at the start of TFX and split data into training and evaluation dataset, transform data into the `tf.Example` format and lastly, copy data into the `_tfx_root` directory so that other components can also access it. ExampleGen takes the data path as an input. example_gen = CsvExampleGen(input=external_input(_data_root)) context.run(example_gen) Now, check the output of example_gen, it will give two datasets for training and testing. artifact = example_gen.outputs['examples'].get()[0] print(artifact.split_names, artifact.uri) Check the training example via this code snippet. StatisticsGen : Next step is to analyze the data, and StatisticsGen takes care of that. It uses the TensorFlow Data Validation library. The dataset from ExampleGen is the input of StatisticsGen. statistics_gen = StatisticsGen( examples=example_gen.outputs['examples']) context.run(statistics_gen) You can visualize all the analysis via one line of code : context.show(statistics_gen.outputs['statistics']) SchemaGen: It generates the schema based on the examination of data statistics. A schema refers to defines the type and features of the dataset. It takes the output of StatisticsGen from step b. schema_gen = SchemaGen( statistics=statistics_gen.outputs['statistics'], infer_feature_shape=False) context.run(schema_gen) Now, you can visualize the schema by : context.show(schema_gen.outputs['schema']) ExampleValidator: It looks for anomalies and null values in the dataset. It takes the output from StatisticsGen and SchemaGen as an input. example_validator = ExampleValidator( statistics=statistics_gen.outputs['statistics'], schema=schema_gen.outputs['schema']) context.run(example_validator) Now, visualize it: context.show(example_validator.outputs['anomalies']) Transform : It does the data\/feature engineering step on the dataset(train & serving). It uses the TensorFlow Transform library. It takes data from ExampleGen, the schema from SchemaGen, and a module that contains user-defined Transform code, as an input. But before that, there are few preprocessing steps, whose code is available here. Now, transform the data  and check the output: transform = Transform( examples=example_gen.outputs['examples'], schema=schema_gen.outputs['schema'], module_file=os.path.abspath(_taxi_transform_module_file)) context.run(transform) It outputs transform_graph (graph that can perform the preprocessing operations) and transformed_examples(represents the preprocessed training and evaluation data). The code for examining them is available here. Trainer : It trains the model using Keras. The default trainer is an estimator. For using the Keras trainer, we have to define a generic trainer by setup custom_executor_spec=executor_spec.ExecutorClassSpec(GenericExecutor) in Trainer’s constructor. The trainer takes the schema from SchemaGen, the transformed data and graph from Transform, training parameters, and a module that contains user-defined model code as its input. Before setting up the trainer, define some user-defined modules necessary for the trainer and whose code is available here. trainer = Trainer( module_file=os.path.abspath(_taxi_trainer_module_file), custom_executor_spec=executor_spec.ExecutorClassSpec(GenericExecutor), examples=transform.outputs['transformed_examples'], transform_graph=transform.outputs['transform_graph'], schema=schema_gen.outputs['schema'], train_args=trainer_pb2.TrainArgs(num_steps=10000), eval_args=trainer_pb2.EvalArgs(num_steps=5000)) context.run(trainer) Analyze the trainer via tensorboard. model_run_artifact_dir = trainer.outputs['model_run'].get()[0].uri %load_ext tensorboard %tensorboard --logdir {model_run_artifact_dir} Evaluator : It evaluates the performance metric on the evaluation set via the TensorFlow Model Analysis library. It takes the data from ExampleGen, the trained model from Trainer, and the slicing configuration(that can slice your metrics on feature values). An example is available here. Then pass this configuration to the evaluator as : # Use TFMA to compute a evaluation statistics over features of a model and # validate them against a baseline. # The model resolver is only required if performing model validation in addition # to evaluation. In this case we validate against the latest blessed model. If # no model has been blessed before (as in this case) the evaluator will make our # candidate the first blessed model. model_resolver = ResolverNode( instance_name='latest_blessed_model_resolver', resolver_class=latest_blessed_model_resolver.LatestBlessedModelResolver, model=Channel(type=Model), model_blessing=Channel(type=ModelBlessing)) context.run(model_resolver) evaluator = Evaluator( examples=example_gen.outputs['examples'], model=trainer.outputs['model'], baseline_model=model_resolver.outputs['model'], eval_config=eval_config) context.run(evaluator) Next, you can examine and visualize the evaluation output: evaluator.outputs context.show(evaluator.outputs['evaluation']) Full code is available for visualization is here. Pusher : It is present at the end of the TFX pipeline and checks the model validation and, if so, then deploy the model to a serving infrastructure. pusher = Pusher( model=trainer.outputs['model'], model_blessing=evaluator.outputs['blessing'], push_destination=pusher_pb2.PushDestination( filesystem=pusher_pb2.PushDestination.Filesystem( base_directory=_serving_model_dir))) context.run(pusher) You can now examine the output of the pusher. You can find the complete tutorial here. Conclusion This post discussed Google’s Tensorflow Extended (TFX), a platform for machine learning to scale up productionisation. It provides different pipelines, components and libraries that are not only capable of building an ML model but also provides support for deployment. TFX also helps in monitoring the performance of your machine learning system. Note : All images\/figures are taken from official sources. Official codes, docs & tutorials are available at: WebsiteResearch PaperGithubGuide to TFXBlogExamples\/TutorialsColab Notebook","excerpt":"Ever since Google has publicised Tensorflow, its application in Deep Learning has been increasing tremendously. It is used even more in research and production for authoring ML algorithms. Though it is flexible, it does not provide an end-to-end production system. On the other hand, Sibyl has end-to-end facilities but lacks flexibility. Google then came up […]","categories":["AI Trends"],"tags":["Deep Learning","deploying models","Keras","Kubeflow","Machine Learning","model deployment","Python Libraries","Tensorflow"],"author_name":"Aishwarya Verma","publish_date":"2021-03-14T13:00:00","publication_year":"2021","word_count":1077,"keywords":["machine learning","deploying models","Keras","AI","Python Libraries","TPU","ML","model deployment","Machine Learning","Colab","Kubeflow","deep learning","experiment tracking","Deep Learning","TensorFlow","Tensorflow"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","Kubeflow","experiment tracking","TensorFlow","Keras","Colab","TPU"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/guide-to-tensorflow-extendedtfx-end-to-end-platform-for-deploying-production-ml-pipelines\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":41637,"title":"Bad USBs Are The Reason Why You Shouldn’t Plug Just About Anything Into Your PC","content":"It is common operational security practice to not plug unverified or potentially unsafe USB drives into a PC, as it can be loaded with malware. Today, we take antivirus software for granted and let it conduct malware checks, leading to individuals being cavalier about plugging in USB devices. However, USB devices can be used as attack vectors and\/or footholds for attackers to gain access to the device. This is seen in the TV show Mr Robot, where USBs are used to gain access to a closed network. There even exists a completely cross-platform exploit hinging on an inherent issue of computer design: User input is trusted. What Is A Bad USB? To understand what a bad USB is, it is first important to understand how USB devices work. Each USB device has a USB-compliant microcontroller in it, from a mouse and keyboard to plug-in webcams and, most important, flash drives. These microcontrollers are what interface with the computer to tell it which device is being connected and is the single point of communication between the device and the computer. The microcontroller runs on code known as ‘firmware’, which gives instructions on how to conduct all of its activities. The bad USB exploit hinges on rewriting this firmware, which cannot be tested or visible to the computer. Since the firmware conducts all communications to and from the device, rewriting it will make the device appear as something else to the computer. One of the biggest attack vectors using the bad USB exploit is spoofing an HID. HID stands for human interface device and is the computer’s term for keyboards, mice and any other devices used to interact with the system. To determine what kind of a device any plugged-in USB is, the computer will first query the ‘class code’ of the device. Then, drivers are installed and the device is ready to use. USB mass storage devices have a class code of 08h, while keyboards and other HIDs have a class code of 03h. By engineering the firmware to send a different class code, it is possible to make a USB flash drive into a virtual keyboard. This is the form bad USB has taken today. Existing Bad USB Implementations & Scope Now, the device has been plugged into the user’s computer and has been registered as an HID. A script running on the flash drive can now emulate keystrokes that are considered completely legitimate by the computer because human input is given absolute trust. In order to make this process easier and more accessible, a company known as Hak5 launched a device called the Rubber Ducky. It looks like a flash drive from the outside, but its firmware has been reprogrammed to make it appear as an HID. In addition to this, there is a simple programming language that Hak5 created for use with this device, known as Ducky Script. This language is simple, yet powerful, and can perform a variety of functions in conjunction with the Bad USB. There is even a web-based tool to encode and decode payloads, which come in a variety of settings. The Rubber Ducky can harvest information regarding the computer, user, installed programs, networks and even capture the screen. There are also a variety of exploits, such as disabling the firewall, finding and FTPing a file to a server, opening a network port, starting a WiFi access point, allowing share access to the C Drive and much more. Moreover, some bad USBs also come with WiFi capabilities for remote activations. Since almost all operating systems have keyboard shortcuts or input methods for a variety of system-critical tasks, the bad USB is a truly cross-platform attack vector. This even includes MacOS and Linux, which have long been considered more secure than Windows. Potential Use-Cases As seen by the above implementations for exploits, recon and footholds, bad USB is almost purpose-built for malicious attacks. Moreover, provided that a device has the right microcontroller, it can be reprogrammed to be a bad USB on the fly. This means that any old pen drive lying around can easily be used as one of the most powerful hacking tools to be created in modern times. Ducky Script can be run on any microcontroller that supports it, meaning that the next bad USB could be right around the corner. Even as bad USB is very useful for malicious attacks, it can also be used for various non-malicious purposes. This includes for use by sysadmins for setting up large amounts of computers, as seen in a corporate rollout of a new OS. Owing to the high frequency of keystrokes that the bad USB can put out (over 1,000 per second), it can be used to set up systems at a fast rate. Any other applications that require a high amount of repetitive keystrokes can also be automated, even without installing third-party applications. By simply plugging in a USB drive, bad USB can give a malicious user complete control over a computer.","excerpt":"It is common operational security practice to not plug unverified or potentially unsafe USB drives into a PC, as it can be loaded with malware. Today, we take antivirus software for granted and let it conduct malware checks, leading to individuals being cavalier about plugging in USB devices. However, USB devices can be used as […]","categories":["AI Features"],"tags":["AI Threat","Cybersecurity","how to","Network"],"author_name":"Anirudh VK","publish_date":"2019-07-06T10:24:37","publication_year":"2019","word_count":830,"keywords":["AI Threat","how to","programming_languages:R","AI","Git","Network","RAG","ViT","Rust","Cybersecurity","R","programming_languages:Rust"],"extracted_tech_keywords":["AI","RAG","R","Rust","Git","ViT","programming_languages:R","programming_languages:Rust"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/bad-usbs-are-the-reason-why-you-shouldnt-plug-just-about-anything-into-your-pc\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":31438,"title":"Top AI-Based Smartphones Launched In India In 2018","content":"Artificial intelligence is incorporated into many of the features in modern mobile phones. From camera to hardware, every aspect of the phone has rich AI applications. Here is a list of top AI-based phones launched in India in the year 2018: 1.Xiaomi Mi A2: The Chinese giant’s product has an Artificial Intelligence Engine (AIE) that contributes to the AI experiences on the phone’s camera. The AI engine is backed by Qualcomm and has an AI embedded in its camera, security, gaming among other features. Its front camera has a 20-megapixel selfie camera with f\/2.2 aperture showcases its AI capabilities in AI Beautify 4.0 and AI background blurring which is essentially the portrait mode. It uses AI in its portrait mode of the camera as well. Feature: AI-powered camera. 2.iPhone XS: Apple in a way sparked the trend of baking AI smarts into phones. The October release of this iPhone product has its own A12 Bionic chip to take advantage of the neural network processing of artificial intelligence to use it for its portrait mode and its face recognition system. Its AI allows for a blur control even after the picture has been taken. It takes advantage of the neural network AI processing and the camera is capable of creating a 3D image of the face of the user in real time. iPhone XS and XS Max Feature: Outfitted with A12 chip which has second version of neural engine. 3.Samsung Galaxy A8 Star: The release of this smartphone of Samsung in August this year has an AI-based digital assistant called Bixby. This new AI system of Samsung is based on a software from Viv Labs, an AI company acquired by Samsung and led by Dag Kittlaus, the co-founder of the company that sold Siri’s early incarnation to Apple in 2010. Feature: Samsung introduced AI-powered assistant Bixby. 4.iPhone XR: Soon after the release of iPhone XS, it launched iPhone XR featuring its AI-powered A12 Bionic chip to be more self-aware. This new AI chip is to give a better camera and a better augmented reality experience. It also has a newly introduced Liquid Retina which is the most recent and advanced version of LCDs. Its back camera is AI equipped and has a display that lacks HD but it is not noticeable. Feature: AI-powered Bionic chip and camera. 5.Google Pixel 2:  The phone uses machine learning in its Google Play Music. Using machine learning, it recognises when music is playing and displays the song and artist names on the phone’s lock screen. It uses a miniaturized neural network that runs in the phone’s chip and is trained to recognize the audio fingerprint of over 70,000 songs. If a song has begun to play, it quickly recalls the song without sending information to the cloud. Apart from that, it also uses machine learning in its portrait mode camera and makes predictions about features should stay sharp in the photo and create a mask around it. Google Pixel 2 6.Vivo X21:  Dubbed as an AI machine, this product uses AI to identify sex and skin tone. It also identifies the subject’s surrounding lighting environment to create the ideal settings, which are eventually based on the user’s habits and preferences. It has an AI Selfie Lighting that processes light accents authentically by identifying the different parts of the face so that sharpness, shadows and light can be adjusted and filters can be applied. Its AI Face Access can detect ambient light intensity and ensure fast and accurate unlocking at night. It also has an AI processor, the AI Smart Engine, powered by the Qualcomm Snapdragon 660 AIE. Feature: Even though the fingerprint scanner is pegged as the headline feature, this model is outfitted with the AI smart engine which helps in smart processor utilization. 7.Vivo V11 Pro: Another phone that comes with Qualcomm’s Snapdragon 660 AIE which has a built-in AI engine. IT also has an AI equipped camera using which helps to get a perfect position for the best shot. Its AI camera modes are AI Portrait Mode, AI Backlight HDR, AI Low-Light Mode, and AI Scene Recognition. Vivo V11 Pro Feature: This smartphone’s AI-powered cameras are a huge draw with users 8.Oppo F9 Pro: Both its front camera and its back camera has AI features. It has an AI scene recognition which identifies 16 scenes and optimizes the technical parameters of the camera to suit different scenes. It also has an AI Beauty Technology 2.1 which has an AI tool that understands how you want the selfies to look like and helps in customizing them. Feature: Taking a cue from other smartphones, Oppo’s F9 Pro has an AI-enhanced camera 9.Honor view 10: The brand has its own AI processor, Huawei Kirin 970 chipset, which has a built-in NPU (neural-network processing unit), apart from the CPU and a GPU, that enables deep learning based on user behaviour. It also has an AI Accelerated Translator, which uses algorithms to translate both text and speech in over 50 different languages. Apart from these, the phone has a camera equipped with AI. Honor View 10 Feature:  This high-end smartphone comes with an in-built neural processor and also has a  face-unlocking feature.","excerpt":"Artificial intelligence is incorporated into many of the features in modern mobile phones. From camera to hardware, every aspect of the phone has rich AI applications. Here is a list of top AI-based phones launched in India in the year 2018: 1.Xiaomi Mi A2: The Chinese giant’s product has an Artificial Intelligence Engine (AIE) that […]","categories":["AI Trends"],"tags":["assistant","camera","chip"],"author_name":"Disha Misal","publish_date":"2018-12-13T06:50:51","publication_year":"2018","word_count":864,"keywords":["chip","Go","camera","artificial intelligence","machine learning","API","AI","neural network","Git","RAG","deep learning","assistant","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","deep learning","neural network","RAG","R","Go","Git","API"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-ai-based-smartphones-launched-in-india-in-2018\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":36728,"title":"7 Times Telangana IT &#038; Industries Minister KT Rama Rao Championed Artificial Intelligence","content":"As his father ‎K Chandrashekar Rao eyes the Prime Ministerial seat and is set to take the mantle at the Centre, the present Cabinet Minister for Information Technology and Panchayat Raj of Telangana KT Rama Rao is gearing up to be his father’s successor in Telangana. As one among the few young and dynamic leaders in Indian politics, his role both as the workign president of Telangana Rashtra Samiti and IT Minister has received much appreciation from people across the country. As the IT minister of the state, his role in turning Telangana into an investor and tech-friendly state is quite well-known. Be it at Davos or Telangana, KTR sure does know how to make the right investor pitch thus, inviting tech-giants to establish their CoEs and work with the government.  In this article, we look at the number of time KTR has pitched for artificial intelligence. Coaxing adobe to set up a CoE in Hyderabad: Upon KTR’s strenuous effort to bring the Adobe’s facility to Hyderabad since 2015, last year the company announced that it will set up their CoE in Hyderabad very soon. The announcement was made at the sidelines of the World Congress on Information Technology, post the meeting between Adobe CEO Shantanu Narayen and KTR. “Excited to announce Adobe is starting an advanced AI lab in Hyderabad. A global leader providing content creation & enterprise experience software solutions entering Hyderabad is a great asset for the local ecosystem. Thanks a ton to Adobe Chairman & dear friend Shantanu Narayan,” KTR tweeted following the announcement. Pitches for open data policy: In an attempt to make government data more available to people and make services more accessible to the public, under the leadership of KTR, the government introduced Open Data Policy, making it the second and the second state in the country to do so. The move is seen as an attempt to ensure transparency at the government’s end and address public issues with the help of online information. “It can help us save expenditure and also innovate. The open data policy was developed along with others under the overarching IT policy,” KTR had said about the development. On addressing local issues with AI: While speaking at the sidelines of the World Economic Forum in Davos last year, KTR spoke in great length about the government’s Open Data Policy and how it is strengthening his government. At a speaking session titled ‘Global Tech and Local Solutions: Artificial Intelligence’ at Davos, he highlighted how the Telangana government created 50 different data sets for the public so that  AI-based solutions could be applied on the data to find solutions to various problems. “The State plans to identify ‘Problem Statements’ and run hackathons based on the data collected once a critical mass of datasets reaches the platform,” he said. On maintaining the quality of data for effective AI implementation: At the talk, he also highlighted the importance of having high-quality training and testing data for machine learning in order to obtain meaningful data. In order to deliver effective service for the public, he said that the focus should be on streamlining processes and generating meaningful data for better AI systems to function better To turn Telangana into a global hub for AI and data science: In order to usher in technology to the state and encourage more players to take up technology, the TRS government brought out its IT Policy and T-Hub for encouraging startups. “Government of Telangana realises the opportunities the newer technologies present and laid the foundation of a Data Analytics Park to tap the innovation and employment opportunities in this rising sector. The Government is keen to create a robust Data Science and Artificial Intelligence ecosystem for the various stakeholders to thrive, by providing support, mentorship, and other capabilities,” KTR said while at the launch of Nasscom’s CoE for AI and data science. For long term collaboration in the field of AI: A go-getter, the minister held talks with Japanese and South Korean representative to procure investment from the countries for the Phase 2 of Hyderabad Metro Rail last year. An MoU was signed state and South Korea for long term collaboration in the field of gaming, AI and electronic manufacturing. Launches Nasscom CoE for AI and data science: The Telangana signed an MoU with Nascomm to establish its CoE for AI and data science as a public-private model to bolster development in the field of emerging technology. As part of the move, a joint investment of Rs 40 crore was made for development in the field of emerging tech like data science and AI Speaking on the development, KTR said, “The Government is keen to create a robust Data Science and  Artificial Intelligence ecosystem for the various stakeholders to thrive, by providing support, mentorship, and other capabilities. This Centre of Excellence, in partnership with NASSCOM and the industry, is in the direction for developing Telangana and India as a global hub for DS and AI in the coming years.”","excerpt":"As his father ‎K Chandrashekar Rao eyes the Prime Ministerial seat and is set to take the mantle at the Centre, the present Cabinet Minister for Information Technology and Panchayat Raj of Telangana KT Rama Rao is gearing up to be his father’s successor in Telangana. As one among the few young and dynamic leaders […]","categories":["AI Trends"],"tags":["Adobe","CoE","NASSCOM","telangana"],"author_name":"Akshaya Asokan","publish_date":"2019-03-22T05:40:09","publication_year":"2019","word_count":832,"keywords":["data science","Go","artificial intelligence","machine learning","CoE","AI","Adobe","telangana","ML","RAG","Ray","analytics","NASSCOM","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","data science","analytics","Ray","RAG","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/7-times-telangana-it-industries-minister-kt-rama-rao-championed-artificial-intelligence\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10101525,"title":"NoBroker Launches CallZen.AI To Improve Customer Service","content":"NoBroker, last week announced the launch of CallZen.AI. This AI-powered platform provides conversational insights from almost all Indian languages regardless of the context, whether it is calls, chats, or meetings. This is a completely new foray for NoBroker and it comes at an ideal time when everyone is using AI powered tools. Their customers include the banking, finance, insurance, health tech and edtech sector with a freemium model. For professionals and enterprises, it has started a one-month trial model. This tool provides multiple features like extracting insights from customer conversations, identifying conversation highlights, semantic analysis of the text in different languages to point to the same information, summarisation, and smart clustering. This removes the need for manual filtering of data. A notable feature of CallZen is its capacity to pinpoint important moments during customer conversations and schedule call-backs accordingly. Aside from improving internal processes; it also facilitates CRM integration, allowing businesses to set up custom CRM updates, emails, and alerts based on specific call highlights. Real-time alerting based on conversational triggers is another feature that sets CallZen.AI apart. It allows businesses to integrate moments and checklists into their existing systems, ensuring that they are immediately alerted to any violations or unusual conversations between agents and customers. Agents’ performance is closely monitored using metrics such as call score, customer satisfaction rate, handling time, hold detection, talk ratio, and customer sentiments. Businesses can export reports and visualise data through charts, graphs, and trends, providing them with a clear and concise overview of the metrics that matter most to their operations. NoBroker’s Tech Prowess NoBroker, a tech-driven real estate startup, prioritises technology over human reliance to address challenges. AI is crucial to their success, with a team of over 35 data scientists. They emphasise automation through AI techniques such as vision and NLP, creating proprietary products like that of Callzen.AI. This product, initially for internal use, is now offered as a SaaS platform to other companies. The startup’s tech team includes 1,000 agents, and they record and validate over 7,000 hours of call centre conversations daily. Utilising NLP, they identify positive and negative customer interactions. ‘Rentometer’ and ‘NB Estimate’ are their rent and price prediction engines, offering personalised recommendations considering factors like location and commute time. Vision analytics handle property images, and their system ‘IRIS Internal’ verifies image authenticity. NoBroker’s proprietary technology uses AI and user behaviour analytics to detect and eliminate fraudulent listings, resulting in a platform where 99.99% of properties are genuine. They’ve also created a substantial data lake called ‘Starship,’ containing 500 GB of data, enabling queries and business intelligence decisions with information from call data to business operations. Read more: Data Science Hiring Process at NoBroker","excerpt":"CallZen.AI provides conversational insights from almost all Indian languages regardless of the context, whether it is calls, chats, or meetings.","categories":["AI News"],"tags":[],"author_name":"K L Krithika","publish_date":"2023-10-16T17:19:12","publication_year":"2023","word_count":447,"keywords":["business intelligence","data science","AI","automation","NLP","Ray","analytics","R","data lake","startup"],"extracted_tech_keywords":["AI","NLP","data science","analytics","Ray","R","data lake","automation","business intelligence","startup"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/nobroker-launches-callzen-ai-to-improve-customer-service\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10067357,"title":"ICML is turning down papers by the dozen. And researchers are pissed","content":"The International Conference on Machine Learning is just around the corner; however, it’s trending for a different reason: Rejection. The AI\/ML researchers are taking to social media to question the selection process of papers. The 39th edition of the (ICML) is slated to be held in Baltimore, Maryland, from July 17 to 23. Last year, 5,513 papers were submitted, 10 percent higher than the previous year’s submission. Out of the 5513 submissions, 1,184 were accepted with an acceptance rate of 21.5 percent as opposed to the 21.8 percent year before. Yann LeCun, chief AI scientist at Meta, tweeted, “If I go by tweet statistics, ICML has rejected every single paper this year.” Interestingly, he had three of his papers rejected. https:\/\/twitter.com\/ylecun\/status\/1526033359398379521 ICML is a renowned platform for presenting and publishing cutting-edge research on machine learning, statistics and data science. Let’s understand the submission and selection process first. Submission and selection process The researchers submit their papers through Microsoft’s Conference Management Toolkit (CMT). While submissions opened on January 12, 2022, the date for author notification was May 14 (Saturday). ICML adheres to their deadlines, and according to the website, doesn’t extend the dates under any circumstances. This year, ICML has announced three changes to reviewing, paper formatting, and submission processes. Firstly, from now on, reviews will take place in two phases. Secondly, authors need to prepare and submit their papers as a single file, and lastly, ICML said there would be no separate deadline for submitting supplementary material. To be accepted, the papers must be based on original research. The results can be theoretical or empirical; however, the outcomes must contain significant novel results of significant interest to the ML community. The reviewing panel will judge the paper on the degree to which they have been objectively established. Further, their potential for scientific and technological impact will also be taken into account. Reproducibility of results and easy availability of code will be considered in the decision-making process, whenever appropriate. Why are authors unhappy? Even though ICML has a low acceptance rate, many authors are unhappy with the reviewing process. “Though some of my papers got accepted, my favourite submission was rejected by ICML simply because “the reviews are not very insightful, unfortunately”, quoted from the 1st sentence in meta-review. Why should the authors pay for the low review quality? Ridiculous review system,” said Hongyang Zhang, assistant professor at Waterloo’s Cheriton School of Computer Science. Yi Ma, Professor, Electrical Engineering and Computer Sciences at the University of California, Berkeley tweeted, “My problem is that a much-better-prepared paper with very positive reviews got rejected and a paper with dubious reviews accepted.” My problem is that a much better prepared paper with very positive reviews got rejected and a paper with dubious reviews accepted. That is worse in a way: the decision seems entirely in the hands of ACs… and my students are loosing faith in me :-) https:\/\/t.co\/cVmeKBVDty— Yi Ma (@YiMaTweets) May 16, 2022 He said ICML meta reviews are just arbitrary. “In rebuttal, we did exactly what reviewers asked, but AC rejected, saying: “it is not certain whether this new experiment performed in a short period of time was done accurately.” “If you can reject by simply not trusting rebuttal, why have it?” he asked. AI researcher and engineer Charles Martin took to LinkedIn to share his disappointment about his paper being rejected. He, too, like the other authors, questioned the review process. “Got our ICML rejection letters last night for our latest weight watcher papers. The general theme is that the reviewers don’t understand the theory so they won’t accept empirical studies. Does anyone really understand why deep learning works?” he posted. Leon Palafox , head of AI at Algorithia\/Grupo Salinas gave his two cents on the ICML rejection spree. “With so many great researchers having all of their submissions rejected from ICML, I guess this conference is going to be either a groundbreaking event or just a random sample not really representative of the best papers,” he said.","excerpt":"If I go by tweet statistics, ICML has rejected every single paper this year, said Yann LeCun.","categories":["AI Features"],"tags":[],"author_name":"Pritam Bordoloi","publish_date":"2022-05-18T14:00:00","publication_year":"2022","word_count":668,"keywords":["data science","Go","machine learning","AI","ML","GRU","deep learning","Rust","AI research","R"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","data science","R","Go","Rust","GRU","AI research"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/icml-is-turning-down-papers-by-the-dozen-and-researchers-are-pissed\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":54890,"title":"Why Indian Enterprises Need To Revamp Their Cybersecurity Strategy","content":"India has been one of the most attacked nations in the cyber world, according to numerous research reports. From basic malware to advanced zero attacks, security threats are putting Indian companies at constant risk of losing valuable assets. Given the scale of India’s growing digital sector and fintech, many leaders and RBI have been recommending to set up a board for a country-wide cybersecurity policy that gives a needed framework and the strategy to tackle cyber threats. India has been a target of international hackers. For example, Nuclear Power Corporation of India’s Kudankulam nuclear plant said there was a cyberattack on its systems, which was traced to North Korea. The malware had the capability to access files and upload the entire data into a remote network outside of India. A major incident involving North Korean hackers was the Cosmos bank heist. After Pune based Cosmos bank lost $11.5 million in unauthorized withdrawals, a panel was set up to study various UN sanctions breached by North Korea. The report found that North Korean hackers allegedly withdrew the funds from ATMs in 28 countries. Lazarus hackers used multiple attack techniques including malware infection, ATM switch compromise, and the bank’s messaging environment compromise. The exploit involved multiple targeted malware infections followed by leveraging a set of malicious ISO8583 libraries and process code injections to stand up a malicious ATM\/POS switch in parallel and selectively breaking the connection between the Central and the backend\/Core Banking System (CBS). “These persistent attacks indicate the level of interest that hackers or hacker groups have in targeting India. We have to improve our defences, invest in enhancing cyber resilience, and in gathering and using threat intelligence of high quality to detect and thwart attacks,” Kiran Zachariah, Vice President – IoT Solutions at Subex told Analytics India Magazine. The ongoing lockdown situation had made it worse for Indian companies and fueled the demand for a more advanced framework and policy nationwide. “With the expanding number of breaches happening across the country along with government websites getting hacked, the need for building a secure framework for all institutions is very critical. There should be a clear cybersecurity framework for encouraging businesses to partner more closely with government agencies, to share threat information in ways that will help in keeping consumers and companies more aware and protected,” Murali Urs, Country Manager, Barracuda Networks told us. One of the biggest problems that exist is the prevalence of legacy IT infrastructure across smaller towns in India. Multiple cities in India have been found to be vulnerable to advanced cyber-attacks. According to the report from K7 titled Cyber Threat Monitor, critical threats witnessed ranged from phishing attacks to malicious apps disguised as health apps that targeted users’ sensitive data. Phishing attacks were seen more in Tier – II and Tier – III cities while the metro cities did better. The report also found that smaller cities saw over 250 attacks detected per 10,000 users. Interestingly, the Tier-II cities such as Patna, Guwahati, Lucknow, and Bhubaneswar had been worse off than Tier-I cities due to factors like awareness gap about cybersecurity. In addition, companies and government offices in Tier-II cities may not keep up with security patches for vulnerabilities as actively as Tier-I cities. “India became the most targeted country in the world during the second quarter of 2019. Throughout the year, India was in the top five, especially after March 2019. Throughout the year, the country attracted attacks of relatively high quality as compared to other regions and last year. Critical infrastructure was attacked the most, followed by sectors such as banking, defence and manufacturing. This is indeed a worrying trend,” Kiran Zachariah further said. Attacks like ransomware, email phishing, or others like crypto-mining and trojans may remain hidden across IT systems until full-fledged attacks are unleashed. Investigators across India have been urging the government for years now to take preventative measures and allocate much larger funds to safeguard, monitor and patch systems, many of which are critical to national security. Outdated, Unpatched Software Technology policy think tanks and cyber experts advocates have also pointed to the lack of encryption, security standards that plague India’s IT requirements, and which are the root of most data publicly known breaches. According to security analysts, the expansion of the enterprise to the edge has made it very challenging for IT teams, particularly in smaller Indian cities which may not have the capabilities of handling the increasingly-complex network of devices. Moreover, numerous companies and government offices may still have outdated or unpatched software not kept with the latest security patches. Therefore, those places are still juicy targets for malicious actors. Adam Palmer, chief cybersecurity strategist at Tenable told Analytics India Magazine what can be done about this. He said, “India’s cybersecurity needs are not different from the rest of the world. When you analyze the vast majority of breaches that occur, whether they’re in India or globally, most of them are caused by known but unpatched vulnerabilities. Cybersecurity programs in India should evolve to take a risk-based approach that means enterprises should focus on the vulnerabilities that matter the most. Many organizations in India are still relying on legacy tools and processes that are inadequate to navigate the complex threats in today’s dynamic and modern computing environment.” Business Continuity During COVID-19 Business continuity during the COVID-19 lockdown is a big challenge. Firms are not just at risk of losing web connectivity and outages, but data security vulnerabilities and cybersecurity attacks. There have seen many incidents of phishing, misinformation campaigns, and others work-from-home opportunities for hackers making their way around the internet. With the lockdown extended around the world, employees continue to work remotely on their private networks, which is undoubtedly a threat to most Indian companies. According to a study by PwC, the number of cyberattacks on Indian companies has doubled in the past few months as cybercriminals use the disruption brought about by the COVID-19 outbreak to infiltrate corporate networks and steal data. The CERT-In (The Computer Emergency Response Team of India) stated in its latest advisory to the internet users that, “Cybercriminals are exploiting the coronavirus pandemic outbreak as an opportunity to send phishing emails in the form of an ‘important update’ or ‘encouraging donations’, or trying to impersonate employees’ trustworthy organizations.” According to the agency, the current global health situation has seen changes to the way people accomplish their regular job, with an increasing number working from home instead of the office. The officials said that employees who are switching to remote working because of the coronavirus outbreak could create cybersecurity problems for the business and the employers. “In the current context where the same version of confidential\/sensitive data is spread across an organization and sits in various places- databases, cloud platforms, collaboration tools, file system, endpoints, e-mails, etc., it is getting very challenging for security professionals, security architects and security engineers to come up with a single solution to address all security gaps at various levels,” Visweswara Rao Sreemanthula, a Senior Manager – IT Security at Verizon told Analytics India Magazine. In April, Indian IT giant Infosys beefed up its security to safeguard itself from potential cyber-attacks. And the firm has further planned to enhance security in 2020 by expanding and reskilling its team. Vishal Salvi, chief information security officer and head of cybersecurity at Infosys recently said, “Investment in cybersecurity controls are on the rise year-on-year, and that is because organizations are considering cybersecurity investment very strategic for their current and future business.” According to Salvi, the company is mainly focusing on reskilling its team in identity and access management, infrastructure security, security information, event management, security orchestration, automation, and response. The Security Plan The sudden shift to a remote-work model means that employees are now combining personal technology with work networks, and this is contributing to an expanded attack surface. Many of these devices may also be older or unsecured, and this introduces serious new risks. All of this can be challenging for security teams who now have to manage this expanded and complex attack surface. Enterprises are taking the best precautionary steps to support and protect employees during this global pandemic. It is, therefore, critical for businesses to look at the various means by which they can address the complex security challenges, with the escalating threats prevailing due to remote working during the COVID-19 outbreak. Here, active traffic and network assessment is critical for application services to meet the new levels of demand and handle peak loads in traffic. Network security, data availability, and protection have become a crucial priority for enterprises for truly seamless business continuity. On ensuring network security and data traffic management at the time of COVID-19, we also talked with Sanjai Gangadharan, Regional Director, SAARC, A10 Networks who said, “Remote working is the need of the hour for organizations in India as social distancing amid lockdown becomes a priority in fighting the COVID-19 battle. Network protection from distributed denial of service (DDoS) attacks is a key concern of remote working for organizations.” “Organizations should continually assess their networks for security vulnerabilities. This can prevent a range of problems such as unauthorized access to applications and identifying underlying software flaws that expose sensitive data. Vulnerability scanners can help identify these concerns, making it easier to understand if systems have critical risks that need to be addressed,” Adam Palmer, Chief Cybersecurity Strategist at Tenable told. According to Adam, as a first step, it’s important to identify its information assets, having a baseline will help in knowing the width and depth of what has to be protected. Once determined, organizations need to work along with various stakeholders in designing controls which will help achieve the security business objectives. Other experts and leaders in the security industry say that in order to respond to the cybersecurity complexities of remote working due to COVID-19, enterprises must take a zero-trust approach. They must ensure that no user has access to data that they don’t depend on for their day-to-day functions. Companies must also ensure visibility into all users, traffic, data, and workloads, and have uniform security policies applied across all locations to make sure no loopholes exist.","excerpt":"India has been one of the most attacked nations in the cyber world, according to numerous research reports. From basic malware to advanced zero attacks, security threats are putting Indian companies at constant risk of losing valuable assets. Given the scale of India’s growing digital sector and fintech, many leaders and RBI have been recommending […]","categories":["AI Trends"],"tags":["budget","cyber security India","Cybersecurity","Cybersecurity India","enterprise analytic hub","Firewall hardware","hardware firewall","intel global strategy","Ransomware"],"author_name":"Vishal Chawla","publish_date":"2020-04-30T11:37:00","publication_year":"2020","word_count":1694,"keywords":["budget","Ransomware","Scala","Rust","Cybersecurity","R","enterprise analytic hub","CUDA","intel global strategy","RAG","analytics","Go","AWS","AI","ML","Firewall hardware","hardware firewall","cyber security India","Cybersecurity India"],"extracted_tech_keywords":["AI","ML","analytics","RAG","AWS","CUDA","R","Go","Rust","Scala"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/india-cybersecurity\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10087458,"title":"Customer Event Prediction in Online Subscription Products","content":"At Intuit, we fall in love with the customer problem, not the solution, and are continuously looking for better and faster ways to serve them better. Technology has always been a huge enabler, aiding better customer experiences. What makes all of our products special is how successfully AI and data have been integrated to solve customer problems. Predicting customer events ahead of time is extremely important for Smart Sales, Marketing and Customer Service for subscription software such as QuickBooks Online (QBO). Some examples of customer events of concern include: Upgrade to a higher SKU Cancellation of the subscription Connect to an add-on service First-time use of a new feature Abandonment of the task at hand Predicting these events allows sales, marketing and customer service to operate more efficiently and provide personalized experiences to customers in an automated and scalable fashion. For example: 1) dynamically identifying customers who are at high risk of churn can help marketing do a proactive reach out to such customers, 2)  identifying customers better suited for a higher SKU can help sales target these customers, and 3) identifying a customer in real-time who is stuck in the product, and likely to abandon the task at hand, can be used to surface relevant help articles. These proactive reach-outs, both inside and outside of the software, can help the business increase customer engagement, resulting in more long-term-value (LTV) per customer, which is one of the most important business key performance indicators (KPIs) in subscription-based commerce. AI\/ML can help to predict future customer events by learning from historical events, customer profiles and product usage data. However, training a machine learning model for predicting customer events is a challenging problem due to: a) the rarity of customer events, b) the majority of the data for customers is not fully baked at the time of training, and c) the temporal sensitivity due to the dynamic nature of the events. Democratization of event prediction through Intuit’s Propensity Modeling as a Service (PMaaS) Fig. 1: Components within the PMaaS Architecture PMaaS is Intuit’s internal AI service to deploy large-scale event prediction models in production. Any authorized user within Intuit (e.g., product management, sales or marketing teams) can onboard a use case in PMaaS through ‘PMaaS API\/CLI’.  Once onboarded, based on the provided requirements in the onboarding configuration, PMaaS automatically builds and validates multiple classification models and, finally, schedules weekly or monthly prediction generation to be integrated with the downstream sales and marketing systems. To predict a  future customer event (e.g., customer churn or upgrade\/downgrade of a subscription), first a reference time point, say ‘t’, has to be defined. For every eligible customer, the AI\/ML models try to predict whether the concerned event will happen in the time period (t, t+H] for a given user. Here, H is a business-defined time period (such as 15 or 30 days), depending on the problem at hand. For multiple such reference time points in historical data, using an intelligent sampling approach (Bootstrapping\/Sampling block in Fig. 1), the target Y(t) can be defined as a categorical variable, depending on whether or not the event happened in the time period (t, t+H]. Next, through the PMaaS Feature Pipeline, predictors, say X(t), can be computed for each reference time point. For example, customer profile information, subscription information, and product usage data such as transactions recorded, clickstream, etc., are usually considered predictors. Finally, (Y(t), X(t)) data for many combinations of customers and reference time points in history are used to build several state-of-the-art classification models (Model Train block in Fig. 1). Some of the best-performing models are then automatically deployed in production to generate the ‘batch’ of propensity outputs in a scheduled manner, e.g., weekly or monthly, for all the ‘eligible’ customers. The marketing and sales team can use these predictions to target the top customers (in terms of the predicted propensities) based on their bandwidth. In the above-mentioned fashion, ML models within PMaaS can learn to predict time-to-an-event (denoted by T) by breaking the problem into a series of classification problems. For example, the probability that a customer will not cancel their subscription within the next 90 days, i.e.  P(T > 90), can be written as P(T > 90) = P(T > 30) P(T > 60 | T > 30) P(T > 90 | T > 60). Each of these probabilities can be estimated through a separate binary classification model, and the 30 days and 60 days define the candidate breakpoints or reference time points. This representation reduces the time-to-event prediction into multiple disjoint classification problems. The number of such breakpoints\/reference time points decides the bias-variance trade-off. With an increase in the number of breakpoints, the accuracy of the estimated propensities will increase (decrease in bias), whereas the size of the data set available for learning each classification model would diminish (increase in variance). The Bootstrapping block and rigorous validation in the model training block in PMaaS combine multiple models from different reference time points to optimally solve this bias-variance trade-off. Extension of PMaaS for real-time event prediction for in-session intervention Though batch predictions are good for targeting customers through marketing and sales channels, dynamic and scalable in-session interventions are often not served by batch predictions. For example, to assist a customer who is using an online product and facing some difficulties, an intelligent system needs to identify such customers in real-time, detect the type of help the user is looking for, and proactively provide in-session contextual help through a digital assistant or live expert. A batch prediction framework cannot help in such a scenario since it is not sensitive to in-session user activities. On the other hand, a complete real-time system that relies solely on user in-session clickstream data is often not enough to accurately identify top customers to target for the concerned event, e.g., churn, upgrade etc. To solve this problem, we have extended the PMaaS for real-time event prediction, as shown in the left figure. We have designed three components: (1) the PySpark Batch Processing is similar to PMaaS Feature Pipeline, which generates batch features and feeds to a Long Window Model (LWM) to get batch inferences – these inferences are stored in an online feature store (DynamoDB) to be used at the time of real-time inference; (2) Spark Stream Processing pipeline process the real-time Kafka events from user clickstream and stores it in an online feature store; (3) finally, at the time of real-time inference request, the stream processed clickstream is fitted to a Short Window Model (SWM) and the SWM output and LWM output is fitted to an Ensemble Model to get the final prediction. The above approach has certain advantages, both in terms of accuracy and implementation: The proposed mechanism for real-time event prediction pre-processes huge volumes of historical user behaviour data over the last few weeks in batch and thus reduce latency to incorporate long-term user-behaviour in real-time inference. It processes in-session clickstream data in real-time, and hence, it is sensitive to dynamic user actions. By segregating short-window and long-window feature processing, this mechanism reduces the feature space dimensionality. Combining the SWM and LWM scores through an ensemble layer improves the accuracy of targeting high-risk customers in real-time and in session. The outcomes At Intuit, by democratizing AI\/ML-based customer event prediction through PMaaS, we have deployed and integrated several batch prediction pipelines for several events, such as subscription cancellation, feature usage, add-on, upgrade\/downgrade of the subscription, etc. By using these AI predictions through the sales and marketing channels, we have seen a relative drop of 8% in customer churn rate – which led to a multi-million dollar revenue impact for Intuit. Finally, by extending PMaaS for real-time event prediction, we have been able to increase the scope of customer targeting to more contextual and scalable intervention methods, such as serving automated help articles through digital assistants, etc. We have seen a 1.6x lift in the prediction accuracy for detecting customers with a high risk of churns in the real-time system compared to batch systems, especially for brand-new customers who have only been using the product for a short time. Co-author Credits: Shrutendra Harsola BIO: Shrutendra is an applied machine learning professional with ~10 years of industry experience in building machine learning systems. He is currently working as a Staff Data Scientist \/ Tech Lead at Intuit AI team in Bangalore, focussed on developing ML\/NLP solutions for QBO-Advanced. Prior to joining Intuit, he worked as an Applied Scientist at Microsoft BingAds team, where he worked on developing machine learning models for query reformulation, semantic ad matching and query-ad relevance. Academically, Shrutendra completed his bachelors from NIT Bhopal and Masters from IISc Bangalore. (https:\/\/www.linkedin.com\/in\/shrutendra\/) https:\/\/www.linkedin.com\/in\/shrutendra\/","excerpt":"Predicting customer events ahead of time is extremely important for Smart Sales, Marketing and Customer Service for subscription software such as QuickBooks Online (QBO).","categories":["AI Highlights"],"tags":[],"author_name":"Arnab Chakraborty","publish_date":"2023-02-16T10:42:27","publication_year":"2023","word_count":1440,"keywords":["Go","machine learning","TPU","AI","ML","Scala","Git","NLP","Kafka","R"],"extracted_tech_keywords":["AI","machine learning","ML","NLP","TPU","Kafka","R","Go","Scala","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/customer-event-prediction-in-online-subscription-products\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":46710,"title":"Why Developers Community Is All Excited About The TensorFlow 2.0 Final Release","content":"With more power and productivity than the earlier versions, the final release of TensorFlow 2.0 has been announced by Google. In March 2019, TensorFlow Alpha was released at the TensorFlow Dev Summit. Internally developed by Google Brain in 2015, the platform has soon gained popularity and become one of the favourites for machine learning researchers. TensorFlow is one of the most flexible platforms to perform machine learning problems. It provides a suitable abode with essential tools for ML researchers and developers in order to perform SOTA machine learning applications. In one of our earlier articles, we discussed the key features which TensorFlow 2.0 Alpha version has been redesigned with. We discussed features like: The replacement of APIs Python-like execution Control over variables Graph mode functions and autograph Managing variables with Keras The TensorFlow 2.0 release also includes an automatic conversion script in order to assist you in migration from TensorFlow 1.x to TensorFlow 2.0. In this article, we will discuss the furthermore enhancements that the platform has brought with the final release for the machine learning enthusiasts. SavedModel Format TensorFlow 2.0 is standardised on SavedModel file format in order to run machine learning models on a number of runtimes such as browser, web, cloud, Node.js and other such. Basically, a SavedModel contains a complete TensorFlow program, including weights and computation and it does not require the original model building code to run and thus makes it easy for deploying or sharing with TensorFlow Hub, TensorFlow Serving, TFLite or TensorFlow.js. Distribution Strategy API Distribution Strategy API is a TensorFlow API to distribute training across multiple GPUs, multiple machines or TPUs. This API helps in distributing the existing machine learning models and training code with minimal code changes. It helps in attaining high-performance training of a model. This API has certain useful features such as it is easy to use and support multiple user segments, including researchers, ML engineers, etc., provides good performance out of the box as well as easy switching between strategies. Multi-GPU Support The new version provides several performance improvements on GPUs as well as multi-GPU support. According to the researchers, the simplest way to run on multiple GPUs in TensorFlow, on one or many machines, is simply implementing Distribution Strategies. This support is initiated with a purpose to allow a machine learning model to scale with more resources. Performance Improvements on GPUs The new release claims to provide better performance on the acceleration of GPUs. The platform utilises an improved API for maintaining the high-performance during inference on NVIDIA T4 cloud GPUs on Google Cloud.  According to a blogpost by TensorFlow researchers, TensorFlow 2.0 delivers up to three times faster training performance using mixed precision on Volta and Turing GPUs with a few lines of code, used for example in ResNet-50 and BERT. Outlook TensorFlow is all about supporting the machine learning developer’s community with a flexible, powerful and easy-to-use platform which supports deployment in any platform. Not only Python developers but Java developers can also use this platform with ease and implement machine learning directly in the browser or in Node.js. Furthermore, the company is stressing on integrating Swift language with TensorFlow in order to create a platform for deep learning and differentiable programming. However, the next release is not yet mentioned but the researchers at Google claimed that Cloud TPU support will be coming in a future release.","excerpt":"With more power and productivity than the earlier versions, the final release of TensorFlow 2.0 has been announced by Google. In March 2019, TensorFlow Alpha was released at the TensorFlow Dev Summit. Internally developed by Google Brain in 2015, the platform has soon gained popularity and become one of the favourites for machine learning researchers. […]","categories":["AI Features"],"tags":["Machine Learning","Tensorflow"],"author_name":"Ambika Choudhury","publish_date":"2019-10-01T18:30:26","publication_year":"2019","word_count":562,"keywords":["machine learning","Keras","TPU","AI","ML","Machine Learning","Python","Aim","deep learning","TensorFlow Serving","TensorFlow","Tensorflow"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","Aim","TensorFlow Serving","TensorFlow","Keras","TPU","Python"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/why-developers-community-is-all-excited-about-the-tensorflow-2-0-final-release\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10099933,"title":"6 Generative AI Jobs in India","content":"Generative AI has revamped our daily work routines. From drafting emails to crafting poetry, LLM-based chatbots like Bard and ChatGPT are making lives easier, one prompt at a time. However, the efficacy of these generative AI models depends on their training, which is why tech companies are actively recruiting to fine tune existing models and develop new ones. Generative AI is more than just a technological marvel; it’s also a driver of job creation in various industries. Let’s explore some of the top career opportunities in this field. Apple – ML Engineer (Generative AI) Soon after the news of AppleGPT surfaced, the big tech started scouting for a passionate and dedicated ML engineer to join their Cloud Technologies App Platform team in Hyderabad. The role involves designing and implementing a machine learning strategy to enhance the developer experience platform and accelerate app development within Apple. The ideal candidate should have a strong grasp of advanced machine learning algorithms, deep learning, and statistics, along with expertise in transformer models like BERT, GPT, and RoBERTa. They will optimise and fine-tune large language models, possess proficient software development skills (Python, PyTorch, TensorFlow, JAX), and build MLOps infrastructure for experimentation, A\/B testing, and production deployment. Experience with MLOps technologies, LangChain pipelines, and presenting complex ML insights to non-technical audiences is valued. The role aims to contribute to Apple’s generative AI-based developer platform, collaborating with data scientists and software engineers to provide ML solutions for internal use, with a focus on improving the developer experience. The ideal candidate should have 5+ years of AI\/ML-focused software industry experience, along with relevant educational qualifications (BTech\/MTech\/MS or PhD in related fields). Apply here. Eli Lilly, Generative AI Senior Team Lead American pharmaceutical company Eli Lilly is expanding its footprints in India and is looking to fill a generative AI – machine language engineer position. The role requires an expert in machine learning to oversee the entire data lifecycle, including collection, cleaning, preprocessing, model training, and production deployment. The primary responsibilities involve configuring modern Natural Language Technology (NLT) platforms to extract insights from crucial business data. This entails designing, implementing, and managing data preparation and advanced analytics tools, showcasing insights from diverse data sources to business users, and developing language understanding\/generation systems using text representation techniques. Key responsibilities include, collaborating with engineering partners to implement optimal cloud-based ML solutions, researching and adopting advanced NLG algorithms for tasks like summarisation, and establishing best practices for deployment and infrastructure. Maintaining ML model performance, automating ML model processes, and fostering relationships with various stakeholders are also key aspects of the role. Required experience encompasses designing, developing, and researching ML systems, NLP\/NLU\/NLG expertise, AWS proficiency, coding skills in Python, Java, and R, MLOps knowledge, Docker\/Kubernetes familiarity, and agile development experience. Read more about it here. Siemens Healthineers, Generative AI Engineer Siemens Healthineers is hiring an experienced generative AI engineer to join their research & development team. The role involves designing and implementing generative AI models like GANs, VAEs, and Transformers for various applications. Responsibilities include data preprocessing, model training, evaluation, deployment, optimization, and collaboration with cross-functional teams. Key skills required include machine learning fundamentals, proficiency in programming languages like Python, deep learning expertise, data preprocessing skills, software engineering proficiency, ethics awareness, and continuous learning. The role also entails security, documentation, problem-solving, and domain knowledge in healthcare, if applicable, and maintaining generative models in production environments. If you think this role is a good fit for you, check out the full description. PwC, Generative AI and Data Analytics Lead Manager PwC is in search of an adept generative AI and data analytics lead manager to join their Kolkata team, focusing on applying advanced analytics and AI methodologies to address intricate business challenges. This position demands a blend of technical prowess, strategic thinking, and business insight to foster innovation and company growth. The responsibilities encompass independently engaging with global and domestic clients, employing deep mathematical expertise in statistics and probabilities, collaborating with diverse teams to pinpoint business priorities, crafting predictive and statistical models, utilizing generative AI for process enhancements, managing data pipelines, and communicating findings to various stakeholders. Qualifications include a bachelor’s degree in computer science, statistics, mathematics, or a related field, preferably a Master’s or PhD, along with 10-12 years of experience in analytics or AI. Proficiency in Python or R programming, deep learning frameworks like TensorFlow or PyTorch, statistical modelling, data visualisation tools, and project management is expected. Here’s the link for you to apply. Wipro, Generative AI Architect Indian IT giant Wipro is hiring for the role of generative AI architect (C1). It requires a seasoned professional with over 10 years of experience. The role involves designing and implementing cutting-edge generative AI models and algorithms, including GPT, VAE, and GANs, and integrating cloud-based generative AII algorithms. Collaboration with cross-functional teams to align AI projects with business objectives is essential. Staying updated on the latest generative AI, machine learning, and deep learning advancements and optimizing existing models for enhanced performance is part of the job. Proficiency in Python, deep learning frameworks (TensorFlow, PyTorch, or Keras), cloud platforms (AWS, GCP, or Azure), and NLP tools (SpaCy, NLTK, Hugging Face) is required. Find more about the job here Ascendion, Generative AI Engineer In this role with engineering firm Ascendion, you’ll be responsible for fine-tuning and optimising generative AI models to ensure they perform efficiently across various applications. Your tasks will also involve designing intelligent AI agents to enhance user interactions and experimenting with innovative techniques to improve recommender systems. You’ll be in charge of constructing and managing inference pipelines for seamless AI model deployment on different platforms, overseeing deployment and versioning for smooth updates and rollbacks. Additionally, you’ll develop tools to continuously assess model performance in production, with a strong focus on software engineering principles. Leveraging your expertise in transformer models and text embeddings, you’ll optimise and manage machine learning models in production settings. The job offers remote work flexibility, opportunities for skill growth, and various perks and rewards. Follow this link to learn more. Read more:","excerpt":"Looking for a job change? We have got you covered","categories":["AI Trends"],"tags":["Generative AI"],"author_name":"Shritama Saha","publish_date":"2023-09-13T13:00:00","publication_year":"2023","word_count":1002,"keywords":["ChatGPT","machine learning","AI","ML","MLOps","NLP","LangChain","deep learning","analytics","generative AI","Generative AI"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","NLP","analytics","generative AI","ChatGPT","LangChain","MLOps"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/6-generative-ai-jobs-in-india\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10118979,"title":"Smartphones Will Soon be Dead","content":"Smartphones are indispensable and considered the ultimate solution for our daily needs. Now, with AI lurking around the corner ready to take over things, a new question arises: Are AI-powered devices poised to take their place? While it may seem a tad far-fetched right now, the reality is inching closer. Despite our reliance on smartphones for WhatsApp, Instagram, and even ordering meals from platforms like Zomato or Swiggy, recent developments suggest a shifting landscape. Case in point: Apple’s acquisition of Paris-based Datakalab, signaling a push towards bolstering on-device AI capabilities for the future iPhone 16. In another update, Microsoft announced Phi-3, a compact yet powerful language model boasting an impressive 3.8 billion parameters. What sets Phi-3-Mini apart is its ability to operate directly on your smartphone, marking a significant leap forward in accessibility and convenience. In 2022, Nokia CEO Pekka Lundmark predicted that smartphones may not stay relevant in 2030. During the World Economic Forum, Lundmark said, “By then, definitely the smartphone as we know it today will no longer be the most common interface. Many of these things will be built directly into our bodies.” How soon until we shift to AI wearables? With the introduction of the Nokia 5120 in 1998 to foldable touchscreen phones introduced in 2023, smartphones have certainly come a long way. Currently, AI and machine learning operate behind the scenes on our phones, powering various functions, including enhancing photos, translating languages, identifying music, and aiding gaming. Now smartphone makers see a chance to gear towards AI. According to the latest update, Qualcomm and MediaTek have introduced smartphone chipsets that enable the processing power required for AI applications. In 2023, Samsung unveiled its groundbreaking generative AI model, Samsung Gauss, marking a significant leap forward in artificial intelligence technology. Google introduced its AI-powered Google Workspace suite, showcasing the increasing integration of AI into everyday tools and services. Not stopping there, Google teased a new generative search experience, hinting at its potential inclusion in future flagship Pixel phones. Meanwhile, Apple has been diligently incorporating AI and generative AI capabilities into its products. Meanwhile, Pete Lau, founder of OnePlus, has shared that he is optimistic about AI and believes, “AI is a vessel and smartphones fit the bill perfectly.” Source: LinkedIn As technology evolves, innovations like the Limitless AI Pendant and WIZPR Ring are launched to harness the power of AI and interact with large language models. Next in line is Elon Musk’s Neuralink, which is working on producing electronic devices that can be implanted into the brain and used for communication with machines and other individuals. They look forward to making it as common as a smartphone, further opening up a world of possibilities for both medical and technological advancements. The Future is AI With the Humane Ai Pin review, the prospect of trusting and embracing devices seems slim. However, this is just the beginning. Adopting a nascent technology product requires time to gain acceptance. The debut of the MacBook in 2006 was marred by technical issues like unexpected shutdowns and palm-rest discolouration, drawing a considerable backlash. Similarly, the iPad was initially derided as nothing more than a ‘big iPod touch’, casting doubt on its purpose and potential success. Even the revolutionary iPhone faced skepticism upon its unveiling, with critics highlighting the absence of features like a physical keyboard and replaceable batteries, which were standard in contemporary phones. As technology continues to advance, it’s becoming clear that simply upgrading our smartphones may no longer suffice. Embracing AI gadgets alongside our trusty phones and a pair of earbuds seems like a convenient approach.","excerpt":"And the evolution continues- phones to smartphones to AI-powered smartphones, and now, wearable AI joins the fray.","categories":["AI Trends"],"tags":["AI Impacts"],"author_name":"Vidyashree Srinivas","publish_date":"2024-04-25T12:00:00","publication_year":"2024","word_count":595,"keywords":["Go","machine learning","artificial intelligence","AI","IPO","innovation","llm_models:PaLM","generative AI","AI Impacts","Rust","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","generative AI","R","Go","Rust","innovation","IPO","llm_models:PaLM"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/smartphones-will-soon-be-dead\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10138346,"title":"Akash Ambani Urges Indian Govt to Accelerate AI Adoption ASAP","content":"At the eighth edition of the India Mobile Congress, being held at Delhi’s Pragati Maidan, Reliance Jio Chairman Akash Ambani talked about how AI’s adoption can position India as a global leader, foster self-reliance, and finally lead to the desired economic development by 2047. In this regard, Akash Ambani illustrated two key suggestions for PM Modi on the role of AI. First, he empahsised on the importance of “Atma Nirbhar” or self-reliance in India, and stressed on the urgent need for  a holistic AI strategy. “My first suggestion, artificial intelligence will transform every aspect of our lives and bring in an era of unimaginable abundance. With AI, India has the potential to completely transform into the new-age factory and service centre for the world. AI is absolutely critical for realising our dream of Viksit Bharat by 2047,”  Ambani said. Further, he outlined the scale and speed of multilingual data generation in India which will drive the AI revolution. “We request the government to expedite the updating of the 2020 draft of the data centre policy that Indian data should remain in India’s data centres. Therefore, Indian companies ready to set up AI and ML data centres should get all necessary incentives including incentives for power consumption,” he added. The event is centered around the theme ‘The Future is Now‘, and will cover 5G applications and insights into AI, IoT, cloud computing, cybersecurity, semiconductors, and green technology. AI at the Heart of Reliance’s Vision Interestingly, last month, Pranav Mistry, CEO of TWO – a reliance backed tech startup – invited Mukesh Ambani and Akash Ambani to their US offices to have a conversation on the role of AI in India and globally. TWO has introduced SUTRA, a family of cost-effective multilingual generative AI models proficient in over 50 languages. The technology is accessible via the ChatSUTRA app, offering functionality comparable to ChatGPT. In addition, Reliance Industries have initiatives like Jio Brain, Jio AI-Cloud, and Jio Phone Call AI, that signal its deep investment in shaping India’s AI future and taking on global cloud giants. Akash Ambani was always bullish on India being the world’s largest data market, and credits Jio for this success.  Jio’s entry into the AI cloud market is a direct challenge to established players like Google Cloud. As part of its AI strategy, Jio introduced the Jio AI-Cloud Welcome Offer, which provides up to 100 GB of free cloud storage for Jio users starting this Diwali.","excerpt":"At the eighth edition of the India Mobile Congress, being held at Delhi’s Pragati Maidan, Reliance Jio Chairman Akash Ambani talked about how AI’s adoption can position India as a global leader, foster self-reliance, and finally lead to the desired economic development by 2047.  In this regard, Akash Ambani illustrated two key suggestions for PM […]","categories":["AI News"],"tags":["Pranav Mistry","TWO AI"],"author_name":"Aditi Suresh","publish_date":"2024-10-15T18:09:01","publication_year":"2024","word_count":408,"keywords":["Go","ChatGPT","API","artificial intelligence","TWO AI","AI","cloud computing","ML","Pranav Mistry","RAG","generative AI","R"],"extracted_tech_keywords":["AI","artificial intelligence","ML","generative AI","ChatGPT","RAG","cloud computing","R","Go","API"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/akash-ambani-urges-indian-govt-to-accelerate-ai-adoption-asap\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":24885,"title":"Using Asynchronous Method For Deep Reinforcement Learning","content":"Machine Learning applications have propelled artificial intelligence to achieve realistic results to a great extent. This can be largely attributed to improved research and developments in areas like neural networks — particularly deep neural networks. The advancements in these networks have led to other areas of ML, like reinforcement learning (RL), to grow parallelly. RL gets its inspiration from behavioural psychology, where software entities known as ‘agents’, work together to achieve positive training throughput called as ‘rewards’. Although RL algorithms employ neural networks for functionality, it is found that the algorithms are sometimes unstable when learning data. This issue has been impeding ML researchers for a while, and they have come up with numerous solutions to stabilise RL algorithms in terms of performance. In this article, we will focus on one specific study by researchers at Google’s Deepmind. It is called the Asynchronous Method for Deep Reinforcement Learning, and also uses gradient descent optimisation technique in ML. Foundation For Asynchronous Applications Online RL algorithms rely on data anticipated at the moment. This means rewards are dependent on the action taken on the data; and updates on these algorithms happen in incremental portions. Researchers have worked on improving this process by including a step called experience replay. However, this takes a toll on computing resources such as memory and processing power, in addition to other issues such as data from older RL policy. In order to resolve this, asynchronous (computing processes which are independent and take place parallelly) methods are developed. Here, the method enables multiple agents to act, instead of relying on multiple instances in RL. It also makes correlation between input and output data easier. In the study mentioned earlier, asynchronous method is applied to typical RL algorithms such as Sarsa (state-action-reward-state-action), n-step Q learning and actor-critic methods. The authors also emphasise on the computing benefits the asynchronous methods provide, by demonstrating that they even work on a standard multi-core CPU instead of the powerful GPU, which is generally used in a deep learning environment. Asynchronous Method The study considers the backdrop of the standard RL method for developing asynchronous algorithms. For creating an RL framework, the researchers follow a two-step approach. First, they use a technique called asynchronous actor-learners, due to its robustness. For this, they use a single machine with multiple CPU threads, mainly to lower communication costs between threads and achieve efficient algorithm updates. Second, they analyse the various actor-learners present and use exploration policies on these learners. This provides the advantage of applying online updates parallelly with better correlation in the algorithm. In addition, these parallel multiple actor learners have manifold benefits such as reducing training time and promoting online RL. Here’s an illustration of the pseudocode written by the authors for each actor-learner thread: “ \/\/Assume global shared θ, θ– , and counter T = 0 Initialize thread step counter t ← 0 Initialize target network weights θ– ← θ Initialize network gradients dθ ← 0 Get initial state s repeat Take action a with ϵ-greedy policy based on Q(s, a; θ) Receive new state s’ and reward r y = { r for terminal s’, r + ℽ maxa, Q (s’, a’ ; θ–) for non terminal s’} Accumulate gradients wrt θ: dθ ← dθ + ∂(y−Q(s,a;θ))2\/∂θ s = s’ T ← T + 1 and t ← t + 1 if T mod Itarget == 0 then Update the target network θ–  ← θ end if if t mod IAsyncUpdate == 0 or s is terminal then Perform asynchronous update of θ using dθ. Clear gradients dθ ← 0. end if until T > Tmax “ Now with the help of this pseudocode, asynchronous methods are developed which are based on existing algorithms given below: Asynchronous One-step Q-learning: In this algorithm, each thread references its copy within the computing environment and evaluates gradient in every step of the Q-learning loss. This lowers the chances of overwriting in updates done to the algorithm. Asynchronous One-step Sarsa: This is similar to the above algorithm except it differs in the target value for Q (s, a), which is given by r + ℽ Q(s’,a’, θ–) where a’ is action and s’ is the state. Asynchronous n-step Q-learning: This algorithm computes n-step return which is more of a ‘forward view’ algorithm rather than conventional ‘backward view’ programming. For a single update, it uses exploration policy for each state-action and then computes gradients for n-step Q-learning updates for every state-action. Asynchronous Advantage Actor-Critic (A3C): This algorithm follows the same ‘forward view’ approach, except it varies with respect to policy. These algorithms are tested on various use cases like the Atari 2600 game, and so on. They show a significant reduction in training time (as low as one day in A3C algorithm compared to eight days in deep Q-network). Also, these are computed on CPUs instead of GPUs, and are found to be more stable in terms of performance and learning rates. Conclusion Asynchronous method in RL is resource-friendly and can be computed for a small scale learning environment. It shows improved data efficiency and faster responsiveness. Therefore, integrating existing RL algorithms will certainly make it consume lesser resources for computing along with achieving accuracy when it comes to building large neural networks.","excerpt":"Machine Learning applications have propelled artificial intelligence to achieve realistic results to a great extent. This can be largely attributed to improved research and developments in areas like neural networks — particularly deep neural networks. The advancements in these networks have led to other areas of ML, like reinforcement learning (RL), to grow parallelly. RL […]","categories":[],"tags":["reinforcement learning algorithms"],"author_name":"Abhishek Sharma","publish_date":"2018-05-25T10:28:46","publication_year":"2018","word_count":875,"keywords":["Go","reinforcement learning algorithms","artificial intelligence","machine learning","TPU","AI","neural network","programming_languages:R","ML","deep learning","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","neural network","TPU","R","Go","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/using-asynchronous-method-for-deep-reinforcement-learning\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10061779,"title":"After Infosys, now Tech Mahindra launches its metaverse services","content":"Indian IT services and consulting company Tech Mahindra announced that it was launching TechMVerse, a metaverse-related business to offer customers immersive experiences. The company wants to use its own AI, blockchain, AR\/VR capabilities to build B2B use cases across segments. TechMVerse will begin with four main centres in Dallas, London, Pune and Hyderabad. Tech Mahindra plans to use business opportunities offered by the metaverse such as the DealerVerse – the virtual car dealership, Meta Bank – the virtual bank, Middlemist – an NFT marketplace and a virtual gaming centre. It will offer immersive digital and professional experience services around design, content, as well as low code plug and play Non-Fungible Tokens (NFT) and blockchain platforms while also aligning the metaverse strategy with Tech Mahindra’s digital and Environmental, Social, and Governance (ESG) goals. “The fundamental layers of metaverse are very well integrated within Tech Mahindra and its competencies. From infrastructure to experience, from spatial computing to now commerce, Tech Mahindra’s platform, the TechMverse, would enable seamless integration between our known expertise in 5G with our skills in AI, AR\/VR and blockchain,” CEO and MD CP Gurnani said. Gurnani added that Tech Mahindra plans to train 1000 engineers to work on this project. Apart from this, the company will also offer collectables as a part of its partnership with Mahindra & Mahindra Ltd. on Tech Mahindra’s NFT marketplace.","excerpt":"Tech Mahindra will use business opportunities offered by the metaverse such as the DealerVerse – the virtual car dealership, Meta Bank – the virtual bank, Middlemist – an NFT marketplace and a virtual gaming center.","categories":["AI News"],"tags":["Infosys"],"author_name":"Poulomi Chatterjee","publish_date":"2022-02-28T18:23:32","publication_year":"2022","word_count":226,"keywords":["Go","Infosys","AI","programming_languages:R","ML","programming_languages:Go","Git","R"],"extracted_tech_keywords":["AI","ML","R","Go","Git","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/after-infosys-now-tech-mahindra-launches-its-metaverse-services\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10164150,"title":"Microsoft Introduces Muse, a GenAI Model for Gameplay Ideation","content":"Microsoft has introduced Muse, a generative AI model designed for gameplay ideation. The model, built on the World and Human Action Model (WHAM), can generate game visuals, controller actions, or both. The research, published in Nature, was developed by the Microsoft Research Game Intelligence and Teachable AI Experiences (Tai X) teams in collaboration with Xbox Game Studios’ Ninja Theory. The research aims to refine AI-generated gameplay for game development and interactive storytelling. Microsoft has open-sourced the model’s weights, sample data, and the WHAM Demonstrator, a concept prototype for interacting with WHAM models. These resources are available on Azure AI Foundry. “I’m incredibly proud of our teams and the milestone we have achieved, not only by showing the rich structure of the game world that a model like Muse can learn but also by demonstrating how to develop research insights to support creative uses of generative AI models,” said Katja Hofmann, senior principal research manager at Microsoft Research. Muse was trained on human gameplay data from Bleeding Edge, a 4v4 online game by Ninja Theory. The dataset includes visuals and controller actions recorded with user consent. The model has been trained on over 1 billion images and actions, representing more than seven years of continuous gameplay. The Game Intelligence and Teachable AI Experiences teams playing the Bleeding Edge game together. Gavin Costello, technical director at Ninja Theory, said, “It’s been amazing to see the variety of ways Microsoft Research has used the Bleeding Edge environment and data to explore novel techniques in a rapidly moving AI industry.” The research was motivated by the release of ChatGPT in 2022. Microsoft scaled the model’s training from a V100 GPU cluster to H100s, refining its representation of controller actions and images. Early versions struggled with consistency, but iterative training improved the model’s ability to predict accurate game dynamics. Comparing Muse’s generated visuals with actual gameplay, researchers assessed key capabilities such as consistency, diversity, and persistency. Consistency measures whether generated sequences adhere to game dynamics. On the other hand, diversity evaluates how gameplay variations evolve from the same prompt. Persistency determines if introduced elements are maintained in subsequent sequences. Cecily Morrison, senior principal research manager at Microsoft, highlighted the importance of involving game creators from the outset. “It was a great opportunity to join forces at this early stage to shape model capabilities to suit the needs of creatives right from the start, rather than try to retrofit an already developed technology.” Meanwhile, xAI chief Elon Musk recently announced that the company is launching a game studio to reshape the gaming industry. While announcing xAI’s latest model, Grok-3, Musk said, “We’re launching an AI gaming studio at xAI. If you’re interested in joining us and building AI games, please join xAI.”","excerpt":"Muse was trained on human gameplay data from Bleeding Edge, a 4v4 online game by Ninja Theory.","categories":["AI News"],"tags":["GenAI","Generative AI models","Microsoft"],"author_name":"Siddharth Jindal","publish_date":"2025-02-19T23:43:44","publication_year":"2025","word_count":456,"keywords":["ChatGPT","GenAI","API","AI","Azure","R","GPT","XAI","Aim","Generative AI models","generative AI","xAI","Microsoft"],"extracted_tech_keywords":["AI","generative AI","ChatGPT","xAI","Aim","Azure","R","API","GPT","XAI"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/microsoft-introduces-muse-a-genai-model-for-gameplay-ideation\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10163371,"title":"Adani Group, Singapore’s ITEES to Launch the World’s Largest Finishing School With AI-Driven Learning","content":"The Adani Group has partnered with Singapore’s ITE Education Services (ITEES) to launch the world’s largest finishing school in Mundra (Gujarat), as announced by Gautam Adani, chairman of Adani Group, on X. The initiative aims to train over 25,000 learners annually, supporting India’s skill development and employment efforts. The facility will also incorporate AI-driven immersive learning and advanced innovation centres to enhance technical training. The programme is aligned with the Make in India movement, aiming to strengthen the country’s workforce by equipping learners with industry-relevant skills. Gautam Adani (left) with Suresh Natarajan (right), CEO at ITE Education Services ITEES, known globally for technical education, will bring expertise in curriculum development and training methodologies. The partnership marks a significant step in advancing India’s vocational education and workforce readiness. The finishing school is part of Adani Group’s broader efforts to bridge the skill gap in India. The training programme is expected to contribute to economic growth by preparing professionals for various industries. Last year, Adani Group acquired 25 acres of land parcel in the Pimpri industrial zone of Pune’s Haveli locality from Finolex Industries for around Rs 471 crore. The company plans to develop a large data centre on this land. The Adani Group’s data centre business is led by AdaniConneX, a 50:50 joint venture between Adani Enterprises and US-based global hyperscale data centre provider EdgeConneX. The plan was to develop 1 GW of data centre capacity over the next decade. With this move, it aimed to become one of the top three data centre operators by 2030, capitalising on India’s increasing digitisation. At the beginning of 2025, Mukesh Ambani’s Reliance Industries Limited (RIL) also announced plans to construct what could become the world’s largest data centre by capacity in Jamnagar, India. This development comes in the backdrop of a global surge in investments in AI infrastructure. Reliance executives have not yet commented on the project, which is estimated to cost between $20 billion and $30 billion. Currently, the conglomerate has approximately $26 billion on its balance sheet, which could be used to finance the venture.","excerpt":"The initiative aims to train over 25,000 learners annually.","categories":["AI News"],"tags":["AI Education in India","education"],"author_name":"Sanjana Gupta","publish_date":"2025-02-12T18:25:06","publication_year":"2025","word_count":344,"keywords":["API","programming_languages:R","AI","innovation","education","AI Education in India","Git","Aim","R"],"extracted_tech_keywords":["AI","Aim","R","Git","API","innovation","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/adani-group-singapores-itees-to-launch-the-worlds-largest-finishing-school-with-ai-driven-learning\/","complexity_score":3,"technical_depth":7,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10172848,"title":"Ilya Sutskever Takes Over as CEO of Safe Superintelligence After Daniel Gross’s Exit","content":"Ilya Sutskever has officially taken over as CEO of Safe Superintelligence Inc. (SSI), following the departure of Daniel Gross from the company. Daniel Levy will serve as President, while the technical team continues to report to Sutskever. “As you know, Daniel Gross’s time with us has been winding down, and as of June 29 he is officially no longer a part of SSI,” Sutskever shared in a message to the company’s team and investors. “We are grateful for his early contributions to the company and wish him well in his next endeavour.” Previously, it was reported that Meta is in advanced talks to hire Gross to help lead its artificial intelligence efforts. SSI, which Sutskever launched in June last year alongside Gross and Levy after leaving OpenAI, is focused on building safe superintelligence. Despite rumours of possible acquisition talks, Sutskever confirmed that the company plans to remain independent. “You might have heard rumours of companies looking to acquire us,” he said. “We are flattered by their attention, but are focused on seeing our work through.” Meta reportedly made an acquisition offer for Safe Superintelligence (SSI), which was valued at around $32 billion during its recent fundraising. The offer was declined. Sutskever added, “We have the compute, we have the team, and we know what to do. Together we will keep building safe superintelligence.” Safe Superintelligence was launched with the stated goal of prioritising safety and speed equally while developing advanced AI systems.","excerpt":"“You might have heard rumours of companies looking to acquire us,” he said. “We are flattered by their attention, but are focused on seeing our work through.”","categories":["AI News"],"tags":["Ilya sutskever"],"author_name":"Siddharth Jindal","publish_date":"2025-07-03T22:29:06","publication_year":"2025","word_count":242,"keywords":["Go","artificial intelligence","OpenAI","AI","programming_languages:R","programming_languages:Go","Ilya sutskever","R"],"extracted_tech_keywords":["AI","artificial intelligence","OpenAI","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/ilya-sutskever-takes-over-as-ceo-of-safe-superintelligence-after-daniel-grosss-exit\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":31568,"title":"Guide To Setting Up An Artificial Intelligence R&#038;D Lab In India","content":"The frenzy around artificial intelligence is not dying any time soon. Businesses all across the globe are adopting AI, machine learning and related technologies to increase productivity, derive better insights, and revolutionise the way businesses are conducted. While a lot of companies are claiming to be AI startups, there are only a few that are doing it right. Setting up the right AI lab with the right intentions behind it not all that easy. Dealing with an area as critical as AI can be quite challenging. Given the fact that it is highly complex, involves high risks and can be expensive, makes it a challenging area to excel. Moreover, the AI field demands an ultra-specialised talent pool that currently stands at only 22,000 PhD-level experts worldwide, according to reports. No doubt that AI is an extremely high-risk area, but it done rightly, can offer a vast competitive advantage. In India too, the AI startup space is gaining quite a lot of popularity, and one of the key essentials around it is to set the AI R&D lab. Right from having the right infrastructure to hardware, talent pool to conducting a proper risk analysis, there are a lot of requirements that make it up for an idea AI R&D lab. This article takes a lowdown on some of the key requirements to set up an AI R&D lab in India. Identify The Need To Set An AI Lab: Quite obvious as it might seem, this is the founding stone for all the efforts that you are going to be investing in setting an AI lab. As Ramasubramanian Sundararajan, Head of AI Lab at Cartesian Consulting says, a good, maybe even necessary first step in starting an AI lab is finding an AI lab-shaped hole in the organisation. Having a clear idea of why you need one is necessary. “Is it a specific problem or set of problems that you have undertaken to solve, and need a team of specialists to work on? Or is it a more general perception of market demand? Something more specific than “Everyone loves AI” would be a good start,” he told Analytics India Magazine. In another exclusive comment by Ankur Sharma, head of analytics and user growth at Instamojo to AIM echoed the same point. He said that it is important to make sure that you are well aware of the kind of problem statements or projects you wish to deal with. AI and ML have a broad range of applications, it is, therefore, important to focus on a certain theme of applications, instead of going too generic or too broad. Hardware And Software Requirements: With a lot of cutting-edge open source software available today, such as TensorFlow, PyTorch and more, it has become quite easy to get up and running. Having a good knowledge of Python could also be very useful as most ML algorithms have ready-to-use, beginner friendly, high-level libraries in Python. Also, the availability of cloud computing has made it fairly easy to set up an AI or ML lab. “One can pretty much build their end-to-end AI set up in the cloud and models can be trained and hosted on cloud-based server instances, and easily deployed on applications or any other hardware and chip,” says Sharma. AWS Sagemaker, AWS Inferentia, Google Cloud ML Engine, Google Hardware Accelerator are some great applications to start a lab with on-demand pricing and without any big upfront investments. However, while commercialising a solution, it is important to understand the licensing restrictions placed by the libraries and codebases being used, warns Sundararajan. For the hardware needs it is important to ask the question  — how much will it take to train good models? A good high-end server with a GPU will suffice in most cases. He shared that cloud service providers such as AWS also have options where one can not only rent the hardware from them but also use their libraries of pre-trained models for NLP, computer vision etc. “This might be a good way to get up and running fast,” he said. Find The Right Talent: Once the idea and infrastructure are in place, having the right talent is the most critical requirement for setting up AI R&D lab. Major companies across the globe are investing to get the right people on-board who can embrace research and adapt to the company culture. For instance, in early 2018, Facebook hired five top university professors in the US and the UK as part of an effort to strengthen its artificial intelligence research division. “Currently, the demand for data scientists is off the roof compared to the supply, which makes staffing the AI lab a big challenge,” says Sharma. “The proliferation of open source code and models has necessitated a bit of a hacker mindset among AI practitioners, so you’ll find a lot of solutions being built very quickly. But building a robust solution that lives longer requires someone to step back and see the big picture. Make sure you have people who can do that”, shared Sundararajan. He added that it is important to hire people who are capable of critical thought. “It is very easy to get carried away by a cool idea and end up building a shiny toy that nobody wants to play with. Have a good internship program – a two-month summer intern can sometimes allow you to test the feasibility of a long shot idea and help you prepare for the future,” he said. Building Right Model: AI and ML are complicated areas that need a deep understanding of data domains. Even the most simple tasks might seem challenging without the right experience. However, building a complex model is a lot of fun. “There’s so much excitement surrounding this field and so much new research coming out every day that it’s a heady feeling to be part of it all. However, it is a lot of work,” said Sundararajan. It is important to think a few steps ahead to understand the kind of environment the model will finally live in. “Make sure you have an ecosystem in place to deploy good models, and people who can tell the difference between cool and useful,” he said. Sharma shares the same thoughts. He says that one must have deep understanding of both ML algorithms as well as ML infra services. Create The Right Environment: Creating the right environment and offering a dynamic and comfortable workspace is crucial for good research to happen. It is important to provide the right workstation with compelling datasets and interesting problems to work on. Along with supercomputers with exceptional computing power, it is important to have an easy access to advanced tools, induce the culture of collaborative thinking, design thinking and more. It is important to pursue curiosity-driven research.","excerpt":"The frenzy around artificial intelligence is not dying any time soon. Businesses all across the globe are adopting AI, machine learning and related technologies to increase productivity, derive better insights, and revolutionise the way businesses are conducted. While a lot of companies are claiming to be AI startups, there are only a few that are […]","categories":["AI Features"],"tags":[],"author_name":"Srishti Deoras","publish_date":"2018-12-14T12:45:35","publication_year":"2018","word_count":1130,"keywords":["machine learning","artificial intelligence","AI","PyTorch","ML","computer vision","NLP","Aim","analytics","TensorFlow"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","NLP","computer vision","analytics","Aim","TensorFlow","PyTorch"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/guide-to-setting-up-an-artificial-intelligence-rd-lab-in-india\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10113842,"title":"The World is Built on NVIDIA GPUs","content":"When Jensen Huang told AIM that the world we live in is built on NVIDIA GPUs, he wasn’t exaggerating. The company controls about 80% of the market for accelerators in the AI data centres operated by AWS, Google Cloud and Microsoft Azure. Recently, it touched a $2 trillion market value and added $277 billion in stock market value on Thursday—Wall Street’s largest one-day gain in history. Putting the country on their back pic.twitter.com\/KcbgydWMlf— Morning Brew ☕️ (@MorningBrew) February 23, 2024 The company reported a revenue of $22.1 billion, up 22% sequentially and a remarkable 265% year-on-year—well above the outlook of $20 billion—during the latest earnings call. “The world has reached the tipping point of a new computing era,” said Colette Kress, chief financial officer of NVIDIA, during the earnings call. “Almost every single time you interact with ChatGPT, we’re inferencing. Every time you use Midjourney, we’re inferencing. Every time you see amazing Sora videos that are being generated or Runway, the videos that they’re editing, Firefly, NVIDIA is doing inferencing,” said Huang, at the recent earnings call. He said that its AI supercomputers are essentially AI generation factories of this industrial revolution. There is NO Stopping for NVIDIA From Huang personally delivering the first DGX-1 AI supercomputer to OpenAI in 2016, NVIDIA has come a long way. Today generative AI startups like Anthropic, Inflection, and xAI are among the examples that heavily rely on NVIDIA GPUs, specifically RTX 5000 and H100s, to keep their generative AI services running. Some pics from when Jensen delivered the first @Nvidia AI system to @OpenAI pic.twitter.com\/gj4995BKSn— Elon Musk (@elonmusk) February 18, 2024 Earlier this year, Meta chief Mark Zuckerberg revealed that the company is currently training Llama 3 and plans to purchase 350K NVIDIA H100s by the end of this year. Additionally, the social media giant has set its sights on developing open-source AGI. NVIDIA reportedly invested in more than 30 AI startups, “It’s a privilege for us to be investing in them, not the other way around. These are some of the brightest minds in the world,” said Huang. “Exciting companies like Adept, AI21, Character.ai, Cohere, Mistral, Perplexity, and Runway are building platforms to serve enterprises and creators,” said Kress, adding emerging new startups are creating LLMs to serve the specific languages, cultures, and customs of many regions worldwide. NVIDIA is making significant investments in healthcare and drug discovery. Recursion Pharmaceuticals in which NVIDIA invested $50 million is now offering its proprietary AI model through BioNeMo for the drug discovery ecosystem. In the enterprise segment, NVIDIA is working with “leading AI and enterprise software platforms as well, including Adobe, Databricks, Getty Images, SAP and Snowflake”, said Kress. Eyes India Last year, NVIDIA promised India would receive tens of thousands of GPUs and partnered with Reliance and Tata, alongside the government collaboration, which plans to establish a cluster of 25,000 GPUs for startups. “India will be one of the first countries in the world [to get them],” Huang said, confirming that these would be faster than anything the world has ever seen. NVIDIA is also helping in upskilling the talent in the country. The company partnered with Infosys to train 50K on generative AI. It also partnered with TCS to upskill over 6 lakh employees in generative AI. Meanwhile, Wipro partnered with NVIDIA to help healthcare companies accelerate the adoption of generative AI. “India has lots of data,” said Jensen, touching upon the diversity of languages and dialects. He said, “There’s no reason for India to export data to Western companies. However, last year, he mentioned that India does lack infrastructure – “not the roads and bridges kind”, but AI infrastructure. He said with NVIDIA supercomputers coming in, that has also been taken care of. India is now looking to challenge global hyperscalers as well. Mumbai-based data centre giant Yotta plans to deploy 32,000 NVIDIA H100 and H200 GPUs by 2025 worth about $1 billion. Roughly 16,384 GPUs by mid-2024 and 32,768 GPUs by the end of 2025. Older players, like the NSE-listed E2E Networks, have also been extending their footprint in the country, with Adani also expanding rapidly. Indian SaaS company Zoho is also utilising NVIDIA GPUs to build its own LLM to add GenAI capabilities across its suite in an attempt to reduce the reliance on hyperscalers. Join AI Forum for India, our discord community for AI ecosystem, in collaboration with NVIDIA today >> What’s Next? Last year, NVIDIA faced GPU demand challenges, but this year, it has improved the supply chain, according to Huang. “Our supply is improving, overall,” he said, adding that their supply chain is just doing an incredible job for them from wafers and packaging to memories, power regulators, transceivers, networking, and cables. There is still a shortage, but NVIDIA has ramped up H200 production. NVIDIA plans to launch Blackwell, a new GPU range that promises improved AI compute performance compared to the current Hopper architecture, potentially reducing the need for multiple GPUs Additionally, the company plans to build the next generation of modern data centres, what it refers to as AI factories, purpose-built to refine raw data and produce valuable intelligence. “Every car company in the future will have a factory that builds the cars—the actual goods, the atoms—and a factory that builds the AI for the cars, the electrons,” said Huang. Huang is further targeting to build sovereign AI infrastructure worldwide. “What is being experienced here in the United States, in the West, will surely be replicated around the world, and these AI generation factories are going to be in every industry, every company, every region,” said Huang. NVIDIA is set to host its flagship GTC conference at the San Jose Convention Center from March 18-21, 2024. Over 300,000 people are expected to attend this event (both in-person as well virtually). “I am going to tell everybody about a whole bunch of new things we’ve been working on the next generation of AI,” said Huang.","excerpt":"“NVIDIA AI supercomputers are essentially ‘AI generation factories’ of this industrial revolution,” said its chief Jensen Huang in the latest earnings call.","categories":["Global Tech"],"tags":["ChatGPT","NVDIA"],"author_name":"Siddharth Jindal","publish_date":"2024-02-25T14:00:00","publication_year":"2024","word_count":989,"keywords":["Anthropic","ChatGPT","NVDIA","GenAI","OpenAI","AI","AWS","ML","Aim","generative AI","xAI"],"extracted_tech_keywords":["AI","ML","generative AI","GenAI","ChatGPT","OpenAI","Anthropic","xAI","Aim","AWS"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/the-world-is-built-on-nvidia-gpus\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":44382,"title":"Huawei’s Open Source Operating System HarmonyOS Is The Answer To Google’s Android","content":"With the US Government banning Chinese telecom giant Huawei, Chinese companies are doubling down on building their own competencies to reduce dependency on US. The company has taken a lead in developing their own chipsets for telecommunication and even launched 5G networks. The ensuing trade war has resulted in China showing the world they can become one of the undisputed global leaders in the field of emerging technologies. Recently, Huawei announced its own operating system HarmonyOS at the annual developer conference. HarmonyOS is the first-ever microkernel-based distributed operating system which can be used regardless of any systems. The tech giant has been working on this operating system for several years. In the conference, Richard Yu, CEO of Huawei Technologies Consumer Business Group showed off the operating system through one of its Honor Vision TV. This new operating system is supported in various devices like smartwatches, smart TVs, smartphones, computers, laptops, cars, tablets, wireless earbuds and much more. Huawei Smartphones were solely dependent on Android operating systems and after the ban, Chinese companies have adopted a Plan B, which is the development of a new and much more flexible operating system than the existing ones. According to the reports, Yu said that the platform supports various sizes of RAM ranging from kilobytes to gigabytes and this operating system will have no support for any root access. The platform also supports a number of applications where applications from other operating systems like Linux, Android, etc will be compatible. HarmonyOS will use ARK compiler for app development and it will also support several other languages such as Java, Kotlin, JavaScript, C, and C++. HarmonyOS 1.0 will be launched this year within the smart screen products. Then the next three pears are planned to optimise this operating system and make it adopt across a wider number of smart devices and other wearables, etc. How Is Harmony Is Different From The Existing Ones According to reports, Yu stated that HarmonyOS is a microkernel-based operating system which makes it flexible for deployment as well as easy in developing applications in all kinds of scenarios. In this OS, once an application is developed it can be deployed easily and flexibly across a range of different devices. The CEO stated that migrating from Android to their new platform is not that difficult and in future, this new operating system can be used as a replacement to Android. Way Forward The US banned Huawei Technologies regarding national security issues which made a significant impact on China’s tech landscape. As a result of the trade-war, Google suspended the access of Android to Huawei for which the future handsets will not be able to access any of the Google applications including the Google Play Store and the already available handsets will not be able to update the applications from Google Play Store. However, Huawei was already working on the OS even before the ban was declared, but the trade war leads the company to put more efforts into developing the operating system. According to sources, a few days back, the Chinese officials said that Indian firms which are engaged in business with China will have to face “reverse sanctions” if India blocks Huawei technologies from doing business because of pressure from Washington. The US government asked it’s allies to discontinue the use of Huawei technologies like chips and smart screen products for the purpose of national security. Currently, the Indian telecom companies like Bharti Airtel, Vodafone Idea and Reliance Jio are in a dilemma to whether use the infrastructures of Huawei technologies in the core part of 5G technologies or use it on the non-core part. Huawei announces the much-awaited #HarmonyOS, the operating system that will provide users with a holistic intelligent experience across all devices and scenarios. #HuaweiNow #HDC2019 https:\/\/t.co\/K8rL8xdOBY pic.twitter.com\/BF8CIDQvjd — Huawei (@Huawei) August 10, 2019","excerpt":"With the US Government banning Chinese telecom giant Huawei, Chinese companies are doubling down on building their own competencies to reduce dependency on US. The company has taken a lead in developing their own chipsets for telecommunication and even launched 5G networks. The ensuing trade war has resulted in China showing the world they can […]","categories":["Global Tech"],"tags":["Android","Huawei","operating system"],"author_name":"Ambika Choudhury","publish_date":"2019-08-13T11:17:17","publication_year":"2019","word_count":637,"keywords":["Go","programming_languages:R","AI","programming_languages:Java","operating system","programming_languages:Go","Java","C++","JavaScript","Huawei","R","Android","programming_languages:JavaScript"],"extracted_tech_keywords":["AI","R","JavaScript","Go","Java","C++","programming_languages:R","programming_languages:JavaScript","programming_languages:Java","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/huaweis-open-source-operating-system-harmonyos-is-the-answer-to-googles-android\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":69756,"title":"Smart Lighting is on the way | Crompton and Gooee partner to boost up IoT Ecosystem","content":"Soon India would lead its way to smart lighting market, as the award winning IoT lighting ecosystem by UK based firm Gooee is eyeing a launch in India. It has teamed up with Crompton Greaves Consumer Electricals, India’s leading supplier of consumer electrical goods, lighting and lighting automation systems, to be Gooee’s official launch partner in India. With this partnership, Crompton unleashes its plan to foray into the world of IoT and connected lighting in big way. Combining Gooee’s features and scalability with Crompton’s design and execution capability, the consumers can expect wonderful products and solutions with exciting features. Shantanu Khosla, Managing Director, Crompton Greaves Consumer Electricals Ltd., said “The Gooee system puts lighting at the heart of a building’s IoT system, providing new opportunities to monitor LED performance, track footfall and communicate directly with occupants.” “We believe these factors will be of significant interest to many of our customers, in particular, those in the retail, commercial property and industrial sectors, enabling them to operate more efficiently and effectively within our increasingly connected world”, he added. The Gooee ecosystem provides sensing, control and communication components that integrate with an enterprise scale cloud platform. This offers a service-driven, scalable framework that can be integrated into LED lighting installations, significantly increasing control, monitoring and data analysis while driving performance and efficiency improvements. “We are very pleased having entered into this partnership agreement with Crompton. India is a very important, high-growth market and it is an honour for us to engage with such a forward-thinking market leader, with such a clear focus on innovation”, commented Jan Kemeling, Chief Commercial Officer of Gooee Limited. He also added “The company employs highly capable engineering resources who understand that apart from energy savings, the Gooee platform will offer Crompton’s professional customers endless possibilities to improve and optimize their business processes.” At a time when innovation is the key in India’s fast growing, the integration of Crompton’s digital lighting products into Gooee’s enterprise cloud which has been designed to manage the high-velocity\/high-volume transaction data from the on-premise environment and provide an extensible platform to develop applications and data visualizations, will lead to innovative service offerings from Crompton to its customers across multiple verticals.","excerpt":"Soon India would lead its way to smart lighting market, as the award winning IoT lighting ecosystem by UK based firm Gooee is eyeing a launch in India. It has teamed up with Crompton Greaves Consumer Electricals, India’s leading supplier of consumer electrical goods, lighting and lighting automation systems, to be Gooee’s official launch partner […]","categories":["AI News"],"tags":["smart cities india"],"author_name":"Srishti Deoras","publish_date":"2016-12-08T08:23:03","publication_year":"2016","word_count":366,"keywords":["smart cities india","Go","programming_languages:R","AI","innovation","Scala","Git","automation","Ray","programming_languages:Scala","R"],"extracted_tech_keywords":["AI","Ray","R","Go","Scala","Git","automation","innovation","programming_languages:R","programming_languages:Scala"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/smart-lighting-way-crompton-gooee-partner-boost-iot-ecosystem\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10048420,"title":"A Beginner’s Guide to Neural Network Pruning","content":"Deep neural networks have been responsible for much of the advances in machine learning over the last decade. Many of these networks, especially the best-performing ones, necessitate massive amounts of computing and memory. These restrictions not only raise infrastructure costs but also complicate network implementation in resource-constrained contexts like mobile phones and smart devices. Neural network pruning, which comprises methodically eliminating parameters from an existing network, is a popular approach for minimizing the resource requirements at test time. The goal of neural network pruning is to convert a large network to a smaller network with equivalent accuracy. Here in this article, we will discuss the important points related to neural network pruning. The major points to be covered in this article are listed below. Table of Contents Need of Inference Optimization What is Neural Network Pruning?Types of Pruning Advantages and Disadvantages Let’s proceed with our discussion. Need of Inference optimization As we know that an efficient model is that model which optimizes memory usage and performance at the inference time. Deep Learning model inference is just as crucial as model training, and it is ultimately what determines the solution’s performance metrics. Once the deep learning model has been properly trained for a given application, the next stage is to guarantee that the model is deployed into a production-ready environment, which requires both the application and the model to be efficient and dependable. Maintaining a healthy balance between model correctness and inference time is critical. The running cost of the implemented solution is determined by the inference time. It’s crucial to have memory-optimized and real-time (or lower latency) models since the system where your solution will be deployed may have memory limits. Developers are looking for novel and more effective ways to reduce the computing costs of neural networks as image processing, finance, facial recognition, facial authentication, and voice assistants all require real-time processing. Pruning is one of the most used procedures. What is Neural Network Pruning? Pruning is the process of deleting parameters from an existing neural network, which might involve removing individual parameters or groups of parameters, such as neurons. This procedure aims to keep the network’s accuracy while enhancing its efficiency. This can be done to cut down on the amount of computing power necessary to run the neural network. Network Pruning Steps to be followed while pruning: Determine the significance of each neuron. Prioritize the neurons based on their value (assuming there is a clearly defined measure for “importance”). Remove the neuron that is the least significant. Determine whether to prune further based on a termination condition (to be defined by the user). Every method that has been invented or we are applying for a long time has some guidelines. By following those guidelines we can ensure the result of the method as expected. The research paper LOST IN PRUNING: THE EFFECTS OF PRUNING NEURAL NETWORKS BEYOND TEST ACCURACY shown certain guidelines for pruning those are as below: If unanticipated adjustments in data distribution may occur during deployment, don’t prune.If you only have a partial understanding of the distribution shifts throughout training and pruning, prune moderately.If you can account for all movements in the data distribution throughout training and pruning, prune to the maximum extent possible.When retraining, specifically consider data augmentation to maximize the prune potential. Types of Pruning Pruning can take many different forms, with the approach chosen based on our desired output. In some circumstances, speed takes precedence over memory, whereas in others, memory is sacrificed. The way sparsity structure, scoring, scheduling, and fine-tuning are handled by different pruning approaches. Structured and Unstructured Pruning Individual parameters are pruned using an unstructured pruning approach. This results in a sparse neural network, which, while lower in terms of parameter count, may not be configured in a way that promotes speed improvements. Randomly zeroing out the parameters saves memory but may not necessarily improve computing performance because we end up conducting the same number of matrix multiplications as before. Because we set specific weights in the weight matrix to zero, this is also known as Weight Pruning. Structured and Unstructured Pruning To make use of technology and software that is specialized for dense processing, structured pruning algorithms consider parameters in groups, deleting entire neurons, filters, or channels. We set entire columns in the weight matrix to zero, thus removing the matching output neuron. This is also known as Unit\/Neuron Pruning. In a feedforward layer, for example, part of the Convolutional NN channels or neurons are deleted, resulting in a direct reduction in computation. Scoring Pruning strategies may differ in the method used to score each parameter, which is used to choose one parameter above another. The absolute magnitude scoring approach is the standard, however, a new scoring method could be developed to improve the efficiency of pruning. It is standard practice to assign values to parameters based on their absolute values, trained importance coefficients, or contributions to network activations or gradients. Some pruning algorithms compare scores locally, trimming a subset of the parameters with the lowest scores across all layers. Others look at scores in a broader sense, comparing them to one another regardless of where the parameter is located in the network. Iterative Pruning and Fine Tuning Some methods prune the desired amount all at once, which is often referred to as One-shot Pruning, certain systems, known as Iterative Pruning, repeat the process of pruning the network to some extent, and retraining it until the desired pruning rate is obtained. For approaches that require fine-tuning, it is most typical to continue training the network with the weights that were trained before pruning. Other proposals include rebinding the network to an earlier state or reinitializing the network. Scheduling It defines the ratio of pruning after each epoch together with the number of epochs required to train the model every time it is pruned. When it comes to pruning procedures, there are differences in the quantity of the network to be pruned at each step. Most weights are removed in a single pass with some approaches. Iteratively, others prune a constant fraction of the network, while yet others use a more complicated function that changes the pace of pruning over time. Advantages Reduces the inference and training time, depends on compression method and of course hardwareAs the neurons, connections between layers and weights are reduced, there is a reduction in storage requirement  Reduces the heat dissipation in deployed hardware say mobile phonesPower Saving Disadvantages Fewer pre-trained models and versions are availableDifficulty in selection compression method as we have to know the architecture of targeted hardwareNot much quantify beyond original accuracy Conclusion Neural network pruning is quite an old technique, recent advancement, development, and wide research community raise its popularity. Though it is a very powerful technique when you’re aiming to reduce the inference time and have limited hardware resources such as memory constraints. In this article, we have understood why there is a need to reduce inference time and various pruning methods.  Lastly seen advantages and disadvantages of pruning. References: What is the state of art neural network?Lost in Pruning","excerpt":"Neural network pruning, which comprises methodically eliminating parameters from an existing network, is a popular approach for minimizing the resource requirements at test time.","categories":["Deep Tech"],"tags":["Data Science","Deep Learning","Guide","Machine Learning","neural network pruning","Neural Networks","optimization algorithms neural networks"],"author_name":"Vijaysinh Lendave","publish_date":"2021-09-18T10:00:00","publication_year":"2021","word_count":1181,"keywords":["Go","machine learning","optimization algorithms neural networks","TPU","AI","neural network","neural network pruning","ML","Machine Learning","RAG","Aim","deep learning","Deep Learning","Data Science","R","Guide","Neural Networks"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","neural network","Aim","RAG","TPU","R","Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/a-beginners-guide-to-neural-network-pruning\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10015278,"title":"Top Books On AI Released In 2020 That One Should Get Their Hands On","content":"The world has been observing the rapid progress that artificial intelligence has been making in the past couple of years. As a matter of fact, after the pandemic, the world is also witnessing an accelerated adoption of AI. However, as this is happening, people find it difficult to comprehend what all this means and how it will affect them. With that being said, it is important for people to be aware of some of the implications of the adoption or the progresses made in the field of AI. And, since it is inevitable, it is also important to know the various ways in which AI could be used to their advantage responsibly and inclusively. Books are a perfect way of educating people on such subjects, and many books published in 2020 did serve this purpose. In this article, Analytics India Magazine collated some of the best books on AI that were published this year. A World Without Work: Technology, Automation and How We Should Respond The fear of workers being replaced due to automation has always been present. However, this book demonstrates that this time it is different as more and more tasks are looking beyond the capability of computers and are getting automated. However, the author Daniel Susskind, argues that this is not necessarily a bad thing as long as the resulting prosperity that the technology will bring is equitably distributed, and everyone finds a meaning to live by. Find the book from here. The Age of AI: Artificial Intelligence and the Future of Humanity This book, written by Jason Thacker, helps us understand some of the social, moral, and ethical challenges that are a result of humans’ interaction with AI. Thacker explains how AI affects us individually, in our relationships, and the society at large. It further talks about how AI has an impact on our bodies, sexuality, work, economics, and privacy. Find the book from here. The Alignment Problem: Machine Learning and Human Values When the systems we teach do not behave or work the way we want them to, it is called an alignment problem. The book, by Brian Christian, tries to explore everything that can go wrong when we build an AI system. The author argues that whether we succeed or fail in solving the alignment problem ‘will be a defining human story’. The book also introduces us to ‘first responders’ who might help address this issue. Find the book from here. The AI Book: The Artificial Intelligence Handbook for Investors, Entrepreneurs and FinTech This book has been curated by two prominent thought leaders in the global fintech space, Susan Chishti and Karan Jain. The book explains the significance of AI and how it could be used across financial services. It explains the several forms that AI takes, how it is being used currently in the sector and the future state of financial services. Find the book from here. Competing in the Age of AI: Strategy and Leadership When Algorithms and Networks Run the World Analysing how the coronavirus crisis compelled organisations to transform themselves and reinvent the firms using data, analytics, and AI, this book explains how AI-driven processes are much more scalable than traditional methods. Based on the research of hundreds of firms, this book can help enterprise leaders rethink how they operate their business in the age of AI. Find the book from here. T-minus AI Many national leaders and governments have announced strategies and investments in AI, predicting the significance of this technology in the future of geopolitics. Written by Michael Kanaan, the U.S. Air Force’s first Chairperson for AI, this book dives into the global implications of AI, how it has exposed some of the national vulnerabilities, and the current pressing global issues it presents. Find the book from here. How to grow a robot – Developing Human–Friendly, Social AI As robots become more and more a part of our daily lives, Mark Lee introduces us to the idea of making these robots less mechanical and more social. He explores how robots can be more human-like, friendly, and engaging. After describing, ‘what is wrong with robots, Lee presents a different approach to building human-like robots. Find the book from here. The Reasonable Robot: Artificial Intelligence and the Law This book explains the rapidly evolving relationship between the AI and the law. It argues that law should not discriminate between AI and human behaviour and proposes a new legal principle that will ultimately improve human well-being. The book explains what challenges AI poses in different areas of the law and what legal principles can unleash AI’s full potential for human progress. Find the book from here. How To Talk To Robots: A Girls’ Guide To a Future Dominated by AI This book is a ‘girl’s guide’ to AI. Tabitha Goldstaub, the author of the book, discusses the various biases introduced by AI systems that define behavioural patterns. At the same time, she discusses the potential benefits that AI can offer. In conversation with Karen Hao, she demonstrates how AI works in our day to day life. The book also has interviews with women who use AI in their workday. Find the book from here. The Road to Conscious Machines: The Story of AI As the world worries about a dystopian future with intelligent robots taking over, the author of this book tries to change this prevailing narrative on the subject. He reveals how such overhyped anxieties are distracting us from more immediate threats of AI like algorithmic biases and fake news and stopping us from the tapping into the true life-changing potential the field can bring. Find the book from here.","excerpt":"The world has been observing the rapid progress that artificial intelligence has been making in the past couple of years. As a matter of fact, after the pandemic, the world is also witnessing an accelerated adoption of AI. However, as this is happening, people find it difficult to comprehend what all this means and how […]","categories":["AI Trends"],"tags":["machine learning and data transform","machine learning books","purpose of ai"],"author_name":"Kashyap Raibagi","publish_date":"2020-12-22T15:00:00","publication_year":"2020","word_count":940,"keywords":["Go","API","artificial intelligence","machine learning books","machine learning","AI","Scala","GAN","machine learning and data transform","ViT","analytics","purpose of ai","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","analytics","R","Go","Scala","API","GAN","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-books-on-ai-released-in-2020-that-one-should-get-their-hands-on\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10082436,"title":"ManageEngine, Zoho’s IT Sister, Believes in Augmented AI\/ML","content":"For ManageEngine, the IT management division of Zoho Corporation, the next quarter would be all about cybersecurity, Shailesh Kumar, the company’s co-founder and VP of engineering said. Kumar explained that AI\/ML was crucial to protect operations from the increased threat of cyberattacks. Kumar has been working with large-scale data handling at ManageEngine while revitalising engineering processes across the organisation. In a chat with Analytics India Magazine, Kumar brings forth ManageEngine’s fresh perspective on a variety of issues, including how AI\/ML isn’t just a buzzword with IT organisations anymore. AIM: ManageEngine plans to leverage AI\/ML and analytics heavily in the future. In what areas is it currently used? Kumar: ManageEngine uses embedded AI\/ML in all segments, including product platforms, IT operations, service management and IT analytics. From an IT perspective, AI\/ML can be applied in multiple ways but we at ManageEngine believe that AI\/ML doesn’t have to be an additional feature, it is augmenting the existing product. This means AI\/ML functions are already included in the product and the end user doesn’t have to pay separately for it. The journey of AI\/ML can be described in three main terms – descriptive, predictive and prescriptive. Descriptive is when the system can tell what is happening from going through the data, prescriptive is when the system is able to convey what will or might happen and prescriptive is when the system can come up with the right action that must be taken. When an organisation matures, it shifts from descriptive to predictive and prescriptive analytics. Obviously, we can’t just rely on what the model says and normally involve another person in the loop if the action itself is complex and sensitive. In IT analytics, we take data from multiple IT systems and combine it with business data to produce a report which can give real insights. This is an area where analytics can really play around because you can look at the data, search for the anomalies and make the prediction. Nowadays of course, we have a variety of statistical, deep learning and ML algorithms, each with their own strengths and weaknesses, and then apply AutoML. So the system can identify the technique and use a mix of whichever method works best. The other area where AI\/ML has scope is service management. The system has to understand the ticket or the issue that has been raised, understand customer sentiment and search for a similar issue that was raised in the past. The system then shows this to the customer so that they have the option for self-servicing. Or else it can help the customer service agent be more efficient by pointing out what can be done. The other important area that ManageEngine is now looking at is security because analytics can really look back and tell you what and where things went wrong. AIM: Is hiring in India a problem for specialised AI\/ML roles? Does ManageEngine have hiring plans? Kumar: ManageEngine’s policy is that it always hires for the next year. So we take in talent and then train them. Be it data science or learning any of the new programming languages, we give them the time to learn. There’s obviously a lot of demand for data scientists around the world and India being the main area where services are done, AI\/ML now has applications in every part of IT. So, yes there is a talent crunch but I believe we have been able to manage well. Currently, we are in an earthquake zone so we’re still figuring it out on a monthly basis. We are just in the fourth-quarter, so depending on whether the first-quarter is looking up, we will be able to determine the numbers and start hiring. AIM: What are some of the most important lessons you have learned over the span of your career? Kumar: IT has been evolving at a respectable pace for the past few years, be it with newer programming languages, deployment patterns, programming for particular devices, or ways of using services – the transformation is happening. Hence, it is vital to have a thorough understanding of the environment and the underlying principles behind it. Making time to test various technologies, grasping their fundamentals, comparing and contrasting amongst the various options, maintaining the spark, and staying hungry for knowledge are all positive traits to have as a young professional. It might or might not result in monetary gains, but it will undoubtedly result in a more interesting daily existence. Additionally, it is also important to have a solid reading routine. It genuinely opens one’s mind to newer ideas and aids in evaluating your own ideas. Reading doesn’t have to be exclusively on technology. It might be from a variety of subjects, including technology, literature, history, economics, or self-help books, but cultivating that habit is a good one.  It really opens you up for a life of constant exploration. AIM: What advice would you give to the budding professionals in the industry? Kumar: An important thing to bear in mind for young professionals just starting their career, whether in tech or not, is that a business exists because of paying customers. When the customer pays for your product or service, reads your marketing copy, gets a response from your customer service team, or in simple terms consumes your output – that is the moment of truth. Everything around you exists because the customer has agreed to pay for it.  It will always be a good idea to keep the customer in mind as you go about your regular business for it will empower you to constantly offer your best effort and give your daily tasks a greater significance.","excerpt":"The other important area that ManageEngine is now looking at is security because analytics can really look back and tell you what and where things went wrong","categories":["AI Features"],"tags":[],"author_name":"Poulomi Chatterjee","publish_date":"2022-12-15T16:00:00","publication_year":"2022","word_count":941,"keywords":["data science","Go","TPU","AI","ML","RAG","Aim","deep learning","analytics","R"],"extracted_tech_keywords":["AI","ML","deep learning","data science","analytics","Aim","RAG","TPU","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/manageengine-zohos-it-sister-believes-in-augmented-ai-ml\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10160811,"title":"How Indian IT Partnered with AI Startups in 2024","content":"While some companies build AI solutions in-house, others rely on partnerships with startups and other tech companies to provide AI services for their clients. Indian IT firms, in particular, predominantly rely on startups and big tech as they believe that investing too much capital in building products is not their niche. In 2024, leading Indian IT firms like Infosys, TCS, Tech Mahindra, Wipro, and HCLTech significantly advanced their integration of generative AI into their offerings for providing small language models and also building multi-agent systems for their clients in the upcoming years. Much of this is allegedly achieved by partnering with AI startups with bases in India. Infosys Leads the Way Infosys has been at the forefront of this AI integration. Chairman Nandan Nilekani has predicted that while companies will increasingly develop their own specialised AI models to enhance efficiency and productivity, service companies will adopt them for their clients. At the sidelines of the Meta Build with AI Summit, an Infosys representative told AIM, “We are building industry-wide solutions. For market research, we have built a proof-of-concept, and internally, we are leveraging many Llama models for our AI-first journey.” In addition to these partnerships with Meta, Microsoft, and even NVIDIA to build models for its clients, Infosys also announced its partnership with Sarvam AI  in October to build small language models for banking and IT operations, Infosys Topaz BankingSLM and Infosys Topaz ITOpsSLM. Infosys also completed the acquisition of 100% of the stake of in-tech for €450 million, which it announced in its Q4FY24 report in April to expand its R&D capabilities. Infosys Innovation Fund recently also placed its bet on Bengaluru-based oncology precision startup 4baseCare. The startup is pushing to beat cancer with genomics and AI. TCS and Wipro Invest for Technological Acquisitions TCS has been involved in significant AI projects, which have contributed to its revenue growth. In June, along with Infosys, Tech Mahindra, and HCLTech, the IT giant partnered with Yellow.ai to build AI solutions. These collaborations aim to use Yellow.AI’s platform to enhance HR and customer service automation solutions. In March, TCS announced its strategic partnership with SymphonyAI to create predictive and generative AI business applications. SymphonyAI specialises in providing enterprise AI SaaS for financial crime detection and related verticals. Cut to June; the IT firm also partnered with Xerox to improve business outcomes when migrating from legacy data centres to Azure public cloud and integrating generative AI services. In November, Wipro partnered with RELEX Solutions, a prominent provider of integrated supply chain and retail planning solutions. The aim is to use Wipro’s deep expertise in the retail and consumer packaged goods (CPG) industries alongside RELEX’s AI platform to deliver improved capabilities for demand forecasting, supply chain efficiency, and operational planning. This came after Wipro announced its alliance with Cyble for AI-driven cybersecurity risk management solutions in August while working closely with its GCC in India. Moreover, Wipro is investing to strengthen its capabilities across organisations and making bold moves in mergers and acquisitions (M&A) and acquiring companies like Capco and Rizing, which have boosted consulting capabilities and simplified the company’s operating model, said Wipro chief Srini Pallia during the Q4 2024 earnings call. Tech Mahindra’s Love for Startups Under the leadership of CEO Mohit Joshi, Tech Mahindra is aiming to grow its banking and financial services revenue share to 25% by 2027. Joshi plans to leverage his BFSI expertise to drive this expansion and sees generative AI as an opportunity for the IT sector and the company’s workforce. In 2024, Tech Mahindra announced several partnerships with AI startups. In April, the company announced a collaboration with Spain-based Atento to deliver generative AI-powered solutions and services to enterprises worldwide. In May, Tech Mahindra partnered with Zapata Computing Holdings Inc. with the aim of enhancing Zapata AI’s quantum-based generative AI solutions. This collaboration aims to scale Tech Mahindra’s capabilities, offering ‘Scale at Speed’ solutions to drive operational efficiency and improve responsiveness for global Communication Service Providers (CSPs). In August, Tech Mahindra joined forces with LivePerson, a leader in digital customer conversations, to revolutionise customer engagement in the financial services and healthcare and life sciences (HLS) sectors. In the same month, Tech Mahindra also collaborated with Horizon3.ai to develop cybersecurity solutions. The partnership integrates Horizon3.ai’s NodeZero platform, which offers autonomous threat detection, AI-powered penetration testing, and Governance, Risk, and Compliance (GRC) insights into Tech Mahindra’s cybersecurity services. In September, Tech Mahindra also partnered with Discai, a subsidiary of the KBC Group, to provide an AI-powered anti-money laundering (AML) solution. HCLTech Provides Solutions to Others HCLTech is establishing a generative AI technology education and training centre built on IBM’s platform. Along with TCS, in August, the company announced its partnership with Xerox for AI and digital engineering services. Most recently, in November HCLTech partnered with Inspeq AI, an Irish-Indian technology firm, to help enterprises worldwide responsibly develop and integrate AI applications. The alliance aims to embed a responsible AI (RAI) layer into application development, seamlessly integrating with existing toolchains like Microsoft Co-Pilot, AWS Bedrock, and other business applications. In June, the company collaborated with Tecnotree to develop 5G-led generative AI solutions for telcos. The aim of the partnership is to bring HCLTech’s AI-based communications technology with Tecnotree’s 5G and AI-based BSS platform. Apart from investing and partnering with AI startups, HCLTech decided to provide its services to its clients, which it built in partnership with big firms like Intel, SAP, IBM, AWS, Microsoft, and GitHub. Other Mid-Sized Firms Mid-sized IT firms are rapidly adopting AI through strategic investments and acquisitions. LTIMindtree has invested $6 million in Voicing.AI, focusing on enhancing conversational, contextual, and emotional intelligence across over 20 languages. The company also partnered with GitHub to train its workforce. Meanwhile, Mphasis has launched NeoCrux to improve developer productivity and acquired Silverline to enhance customer experience with conversational AI. Persistent Systems, addressing AI-related privacy concerns, acquired Arrka to strengthen its privacy management platform. Smaller firms like Happiest Minds and Hexaware have also made AI acquisitions, positioning themselves for the agentic AI revolution. Happiest Minds partnered with MindSculpt in April and with Soroco in February to enhance its AI solutions.","excerpt":"Infosys, TCS, Tech Mahindra, Wipro, and HCLTech have all partnered with several AI startups over the last year.","categories":["IT Services"],"tags":["AI Startups"],"author_name":"Mohit Pandey","publish_date":"2025-01-07T09:00:00","publication_year":"2025","word_count":1020,"keywords":["agentic AI","AWS","AI","ML","SLM","RAG","Aim","multi-agent systems","generative AI","small language models","AI Startups"],"extracted_tech_keywords":["AI","ML","generative AI","agentic AI","multi-agent systems","Aim","small language models","SLM","RAG","AWS"],"url":"https:\/\/analyticsindiamag.com\/it-services\/how-indian-it-partnered-with-ai-startups-in-2024\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10049057,"title":"Dinanath Kholkar","content":"Dinanath (Dina) Kholkar is Vice President & Global Head, Analytics & Insights at Tata Consultancy Services (TCS). In this role, Dina guides some of the world’s best companies in their journeys to unlock the potential of their data through the power of analytics and artificial intelligence (AI) to uncover new business opportunities, drive business growth, and increase revenue streams. Dina has been recognised as one of the Top 100 influential global data leaders. He advocates ‘data centricity’ as a strategic lever for business growth and transformation and believes that the need of the hour is ‘evangelisation’ of data & analytics. With over thirty years of industry experience, Dina has held diverse leadership roles and amplified business value to customer organisations covering all major industries. He was responsible for building TCS’ data warehousing and data mining expertise and laying the foundation for the organisation’s Business Intelligence practice. Dina has also led TCS’ Business Process Services (BPS) & Business Analytics units and served as the CEO & Managing Director of TCS eServe. At the Pune city level, Dina provides leadership and strategic direction in the fields of education, sustainability, and agriculture, with a focus on ‘data and AI for good’ aspect expressed through his volunteering work at some of the top-notch institutions, including the IEEE, Pune International Center (PIC), Pune Knowledge Cluster (PKC) and the Mahratta Chamber of Commerce, Industries and Agriculture (MCCIA). LinkedIn","excerpt":"Dinanath (Dina) Kholkar is Vice President & Global Head, Analytics & Insights at Tata Consultancy Services (TCS). In this role, Dina guides some of the world’s best companies in their journeys to unlock the potential of their data through the power of analytics and artificial intelligence (AI) to uncover new business opportunities, drive business growth, […]","categories":["AI Features"],"tags":["Interviews and Discussions","TCS"],"author_name":"AIM Media House","publish_date":"2021-09-20T19:57:53","publication_year":"2021","word_count":232,"keywords":["business intelligence","Go","artificial intelligence","programming_languages:R","AI","R","programming_languages:Go","analytics","GAN","TCS","Interviews and Discussions"],"extracted_tech_keywords":["AI","artificial intelligence","analytics","R","Go","GAN","business intelligence","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/dinanath-kholkar\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":46695,"title":"AIM Announces Second Edition Of MLDS, India’s Largest Summit For Machine Learning Practitioners","content":"In an effort to bring the data science and ML developer community together, Analytics India Magazine is launching the second edition of Machine Learning Developers Summit 2020 (#MLDS20), a 2-day conference which will give attendees direct access to top ML innovators from leading tech companies across India. To be held from January 22 to 23, 2020 at NIMHANS Convention Centre, Bengaluru, MLDS will encourage focus on machine learning innovation, and will give attendees an opportunity to interact ML innovators from leading tech companies who will share the best practices on software architecture of ML systems, how to produce and deploy the latest ML frameworks and solutions for business use cases. MLDS 2.0 Is Bigger & Better In its second year now, MLDS summit has become an exclusive hub for ML researchers and data scientists from across the country. In fact, MLDS is considered as a gold standard for India’s data science and ML ecosystem that has given amazing exposure to ML tools and frameworks, new projects, explore right tools, platforms, languages and has provided a platform to discuss hardware and software challenges of building complex ML systems. It will bring together 60+ speakers with over 50 sessions and workshops spread across 3 tracks, spanning two days. With 1,000+ delegates, MLDS is the most buzzing, data science-focused summit featuring in-depth technical sessions and hands-on MasterClass which gives attendees a chance to get their hands dirty with data. It will offer an exclusive platform to talk about the software architecture of ML systems, how to produce and deploy the latest ML frameworks and solutions for business use cases. 40 Under 40 Awards MLDS is also home to the prestigious 40Under40 Awards which recognise leading Data Scientists in India who have demonstrated considerable expertise in solving complex challenges in their organisations and have made contributions to the broader data science and analytics community. These awards will recognise and honour achievements of 40 Data Scientists in India — the dynamic game-changers who are transforming customer experiences with analytics strategies, are fostering a data-native culture and developing new revenue streams from data assets. The first edition of the awards saw talented individuals from both big and small companies share the stage together and celebrate their achievements. Why You Should Block Your Calendar Right Now MLDS is a unique conference that covers both technical and business aspects of machine learning, the most popular emerging tech. With ML playing an expanded role in enterprises, attendees can learn how machine learning can add potential to their business and gain best practices from cutting-edge presentations. Developers and data scientists can take home applicable skills about the latest ML technology and gain networking opportunities with leading companies. It is well suited for both developers and C-suite executives. It would offer an exclusive opportunity to network and build lasting connections with prominent machine learning researchers. It would also allow to test drive and explore the latest machine learning technology and tools hands-on MLDS is brought to you by Analytics India Magazine where we chronicle technological progress in the space of analytics, artificial intelligence, data science and big data. AIM is also the organiser of Cypher, India’s largest analytics summit, MachineCon, The Rising and AIMinds. You can book tickets for MLDS 2020 right here.","excerpt":"In an effort to bring the data science and ML developer community together, Analytics India Magazine is launching the second edition of Machine Learning Developers Summit 2020 (#MLDS20), a 2-day conference which will give attendees direct access to top ML innovators from leading tech companies across India. To be held from January 22 to 23, […]","categories":["Deep Tech"],"tags":["latest machine learning innovation","Machine Learning","MLDS"],"author_name":"Prajakta Hebbar","publish_date":"2019-10-01T15:28:39","publication_year":"2019","word_count":542,"keywords":["data science","Go","artificial intelligence","machine learning","AI","latest machine learning innovation","ML","Machine Learning","RAG","MLDS","Aim","analytics","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","data science","analytics","Aim","RAG","R","Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/aim-announces-second-edition-of-mlds-indias-largest-summit-for-machine-learning-practitioners\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":56319,"title":"Facebook Releases Open-Source Library For 3D Deep Learning: PyTorch3D","content":"Rendering a simple shape into a proper object with geometry, texture, and other material properties is a painstakingly long process; however, with AI, researchers can now do this rendering ten times faster than the real-time. A machine learning model is trained on images that are closer to the target. When it is presented with a shape and matching properties, it would recommend a photorealistic image. This opened a whole new field altogether — differentiable programming. Traditional rendering engines are not differentiable, so they can’t be incorporated into deep learning pipelines. Projects, such as OpenDR, Neural Mesh Renderer, Soft Rasterizer, and redner, have showcased how to build differentiable renderers that can be cleanly integrated with deep learning. In a significant boost to 3D deep learning research, Facebook AI has released PyTorch3D, a highly modular and optimised library with unique capabilities to make 3D deep learning easier with PyTorch. PyTorch3d provides efficient, reusable components for 3D Computer Vision research with PyTorch. via Two Minute Papers Differentiable rendering has revolutionised many computer vision problems that involve photorealistic images, such as computational material design, scattering-aware reconstruction of geometry, and the materials from photographs. Differentiable rendering algorithms estimate partial derivatives of pixels in a rendered image concerning scene parameters, which is difficult because of the visibility changes that are inherently non-differentiable. Overview Of PyTorch3D via Facebook AI 3D meshes are common 3D shape representations in computer vision. It is an easier way to learn low-dimensional linear models of 3D meshes using principal component analysis or higher-order of tensor generalisations. However, this is still challenging. To address this challenge, the developers at Facebook AI created Meshes, a data structure for batching heterogeneous meshes in deep learning applications. This data structure makes it easy for researchers to quickly transform the underlying mesh data into different views to match operators with the most efficient representation of the data. PyTorch3D gives researchers and engineers the flexibility to efficiently switch between the different representation of views and access to different properties of meshes. Researchers and engineers can similarly leverage PyTorch3D for a wide variety of 3D deep learning research, whether it be, 3D reconstruction, bundle adjustment, or even 3D reasoning to improve 2D recognition tasks. For smooth integration of deep learning and 3D data, PyTorch operators: Are implemented using PyTorch tensorsCan handle mini-batches of heterogeneous dataCan be differentiatedCan utilise GPUs for acceleration. 3D deep learning researchers can easily import the loss functions using the modular differentiable API. Over the years, the main reason behind the 3D deep learning research being underwhelming is the complexity of 3D data inputs. They usually have more memory and computation requirements, unlike 2D images, which can be represented using simple tensors. Another issue with applying deep learning to 3D data is the limited amount of 3D data relative to images. This means that, while 3D adds a dimension, the models must be smaller to prevent overfitting. It is especially challenging given that many traditional operators in the computer graphics field, such as rendering, involve steps that block gradients. With PyTorch3D, Facebook AI team has tried to address such issues. Within FAIR, PyTorch3D is already in use for projects such as Mesh R-CNN. 3D deep learning has great significance, especially in tracking spatial dynamics in robotics, improving virtual reality experiences, and even recognising occluded objects in 2D content.","excerpt":"Rendering a simple shape into a proper object with geometry, texture, and other material properties is a painstakingly long process; however, with AI, researchers can now do this rendering ten times faster than the real-time.  A machine learning model is trained on images that are closer to the target. When it is presented with a […]","categories":["AI News"],"tags":["Facebook AI","Pytorch"],"author_name":"Ram Sagar","publish_date":"2020-02-10T16:22:20","publication_year":"2020","word_count":552,"keywords":["Pytorch","Go","API","machine learning","Facebook AI","AI","PyTorch","computer vision","RAG","deep learning","CNN","R"],"extracted_tech_keywords":["AI","machine learning","deep learning","computer vision","PyTorch","RAG","R","Go","API","CNN"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/facebook-releases-open-source-library-for-3d-deep-learning-pytorch3d\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10016885,"title":"Top 8 Things Developers Can Look Forward To At MLDS 2021","content":"Machine Learning Developers Summit 2021 (MLDS21), brought to you by Analytics India Magazine, is scheduled to be held virtually from 11-13 February 2021. It will bring the machine learning community from across the globe together. With over 1500 ML developers, 60 speakers, and 200 organisations across the three days, it is one of India’s largest conferences that bring the ML ecosystem together. The conference aims for ML developers and researchers to come together in one platform to discuss the exciting innovations that have shaped the industry in recent times. Here we bring 8 such interesting takeaways from the conference that developers can look forward to in the upcoming event. 3 Days Of Engaging With ML Community MLDS21 promises 3-days of engaging ML content that will squarely put you ahead of the pack. It will give access to ML experts, practitioners, researchers and entrepreneurs in the machine learning space who promise to bring exceptional content in the space around the latest developments and research in the space. The fact that it is virtual makes it much more convenient for the ML community to join in together for this one of its kind events. Workshops One of the key differentiators of the MLDS conference is the exclusive workshops lined up by the ML experts. Not only are these detailed workshops highly industry-relevant but allows participants to engage in these sessions hands-on that covers trending tools and techniques in the space shaping the industry. Not only that, the attendees can put the best practices to work as soon as they get back to work. Technical Talks The technical talks at MLDS will cover the industry use cases and solutions applied by companies, and how they have worked for them. Demonstrated by industry experts from a wide range of industries ranging from retail, ecommerce, banking and others, these tech talks are a must-attend for those looking to explore ML use cases for their businesses. Paper Presentation One of the significant highlights of MLDS is the paper presentation, where it promotes the AI and ML researchers to exchange ideas in AI and its affiliated disciplines. The technical paper presentation brings together research advances in technical areas around AI. The paper presentation is invited across various domains such as healthcare, sustainability, transportation, and commerce. The submissions for 2021 are open, and one can apply here. Awards The Under 40 Data Scientists awards at MLDS will recognise young data scientists in India who have successfully transformed data into meaningful insights. The prestigious award brings together the brightest leaders in India’s data science field and celebrates their achievements. It recognises leaders who have brought about real innovations in ML, supported the growth of analytics in their organisation and have deep industry expertise to address complex challenges in the field. The winners are selected by our team of editors and industry veterans, who will be recognised at the event virtually this year. The nominations for awards are open, and one can apply here. International Speakers MLDS will see a lineup of more than 60 speakers from India and across the world who have deep expertise in the field of machine learning. Attendees will get a chance to get up and close with these speakers virtually as they present interesting talks and presentations at the event. The speakers will be representatives from global companies and entrepreneurs who have made a mark in the space. Global Networking MLDS21 will be a perfect platform for the attendees to network and discuss the latest developments in the ML field. It will provide a platform virtually to have one-on-one or even group discussions with the who’s who of the industry to build lasting connections. It provides an opportunity to discuss the latest developments in the space with prominent machine learning researchers, like no other. Meet Over 200 Companies The conference will bring together more than 200 AI and ML startups and companies from across the globe, providing a platform for attendees to connect and explore opportunities that stay ahead in the field. It will bring together large tech companies, SMEs and startups that are revolutionising the space, exposing the attendees with their growth stories and use cases that have worked for them along the years. Get more details about MLDS21 here.","excerpt":"Machine Learning Developers Summit 2021 (MLDS21), brought to you by Analytics India Magazine, is scheduled to be held virtually from 11-13 February 2021. It will bring the machine learning community from across the globe together. With over 1500 ML developers, 60 speakers, and 200 organisations across the three days, it is one of India’s largest […]","categories":["Deep Tech"],"tags":["machine learning developer summit"],"author_name":"Srishti Deoras","publish_date":"2020-12-31T16:00:00","publication_year":"2020","word_count":707,"keywords":["data science","API","machine learning","AI","ML","Aim","ViT","analytics","GAN","machine learning developer summit","R"],"extracted_tech_keywords":["AI","machine learning","ML","data science","analytics","Aim","R","API","GAN","ViT"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/top-8-things-developers-can-look-forward-to-at-mlds-2021\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":61646,"title":"How AIOps Can Help Tighten Cybersecurity At Your Organization","content":"AIOps (AI for IT Operations) within IT can greatly help all departments, from network operators to security teams. Although AIOps capabilities can be leveraged by a wide variety of departments, one of its significant applications is in cybersecurity. AIOps assists the cybersecurity department by helping it gain a significant amount of speed, visibility and intelligence when it comes to data security and threat detection. These tools accomplish various tasks – from observation to actions on threats – details of which will be discussed below:- How AIOps Helps Cybersecurity Two of the main aspects of cybersecurity are speed and device visibility. If one can pinpoint the location of the source of a cyberattack and when it happened, it can give a significant edge in catching cybercriminals. AIOps platforms use collected streaming telemetry (a real-time data collection service where network devices continuously push data related to network’s health to a centralized location) data to inventory, auto-discover and classify the devices. Most AIOps platforms, in addition to inventory network infrastructure components, also assess all the wired, wireless and IoT devices communicating in the corporate network or the cloud. Another area where AIOps can help is in network segmentation. AIOps can use device classification to ensure that business devices are connected to an appropriate virtual LAN or a wireless set identifier. This network segmentation is essential when it comes to edge security and having tools (AIOps) that can quickly help spot problems of connection, etc, that are in high demand. Whatever DPIs and other telemetry data are collected, these can be used to plot device communication behavior over time. In the event of an abnormal occurrence, an alert is triggered to investigate the potential threat to the security. A lot of AIOps also include threat intelligence analysis services. These are services where reports are produced about various threats for the security control system. Most AIOps integrate with other security tools, including SIEM, network firewalls, security orchestration, automation and response. These tools, along with AIOps traffic behavioral analysis, can monitor better security threats. Need For Human Intervention AIOps, as much as it involves AI that promises complete automation, still needs some human involvement. The AI within AIOps needs to be told about which apps, services and other resources are business-critical. This includes identifying essential data flows to ensure that AIOps platforms understand the priority in which the security events should be dealt with. AIOps provides detailed information when there is an alert triggered. A network administrator comes in to investigate the alert and make decisions about the suggestions given by the AIOps. While some AIOps automate the responses, an administrator is still needed to intervene. Some Common Pitfalls Faced By AIOps A successful adaption of AIOps needs awareness of the potential problems associated with it. Below, we have mentioned some of the top concerns when it comes to AIOps implementation: #1: Identifying use cases, not processes Each new AI and ML feature may seem like a perfect opportunity to increase efficiency for an organization. But all features may not end up benefiting their AIOps needs. To avoid inefficient piecemeal adoption lifecycle, enterprises should start with a top-down assessment of all the applications, systems, and processes to where AIOps can contribute the most. #2: Not enough data and poor quality Not having enough data may lead even the most powerful AI tools with unpredicted errors. The AI\/ML algorithms have always been data-hungry, and the more data one gives them, the more accurate the results are, the same is the case with AIOps tools and the algorithms involved. Another data-related problem that AIOps encounter is the poor quality of the data. Once an adequate amount of data is collected, the next step is always to assess its quality. The organization should avoid giving noisy data, inconsistent or insufficient frequencies, and inconsistent naming across applications or centers, etc. Organizations should develop standard procedures and also identifying the types of data that are most valuable for their specific priorities. #3: The Meaning of data Data that lacks semantic consistencies are less valuable for both operators and AIOps. Organizations that collect an abundant amount of high-quality data without the right context is almost useless. Teams should emphasize the importance of linking data so that relationships are recorded and are easily identifiable. Outlook Using AIOps when it comes to cybersecurity means analyzing data related to threats to the extent that the exact nature of the threat can be identified with suggestions on how to contain it. While AI promises complete automation, AIOps applications will still need humans. However, the knowledge about AIOps and the various cybersecurity threats should not only be familiar to the company’s security team, but also other departments. If an organization is using AIOps for cybersecurity, opening up a platform for AIOps will improve communication between various departments, which will in turn, enhance the organization’s cybersecurity.","excerpt":"AIOps (AI for IT Operations) within IT can greatly help all departments, from network operators to security teams. Although AIOps capabilities can be leveraged by a wide variety of departments, one of its significant applications is in cybersecurity. AIOps assists the cybersecurity department by helping it gain a significant amount of speed, visibility and intelligence […]","categories":["AI Trends"],"tags":["AIOps","International Affairs","network firewall"],"author_name":"Sameer Balaganur","publish_date":"2020-04-14T12:00:00","publication_year":"2020","word_count":808,"keywords":["network firewall","Go","programming_languages:R","AI","ML","programming_languages:Go","RAG","automation","AIOps","GAN","International Affairs","R"],"extracted_tech_keywords":["AI","ML","RAG","R","Go","GAN","automation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/how-aiops-can-help-tighten-cybersecurity-at-your-organization\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10101633,"title":"OpenAI Kills Arrakis","content":"OpenAI is getting ready for its first-ever DevDay conference. Moreover, it might have even cracked the AGI code with the recent rumours of Jarvis. But on the other hand, it is reportedly killing one of its other dear projects, codenamed Arrakis, as it did not live up to the company’s expectations during training. Undoubtedly, OpenAI’s models have to live up to the standard of ChatGPT, which might be a little too high even for them. Unlike GPT-4, which is huge in size and more powerful than its predecessor GPT-3.5, Arrakis was expected to be smaller and allow the chatbots to run more efficiently and less expensively. This was probably in line with the release of Meta’s LLaMA and Llama 2, which even Microsoft, the biggest backer of OpenAI, had started using in many of its products and services. But the news about OpenAI building Arrakis started long before Llama was ever in the picture, ever since they started training GPT-4. But now, The Information reports that they have scrapped the model. Why, though? What if Arrakis actually has a lot of flaws? Gary Marcus points out after the report that Sam Altman might actually be right. “They might have decided that GPT-5, if it was simply a bigger version of GPT-4, would not meet expectations, and that it wasn’t worth spending the hundreds of millions of dollars required, if the outcome would only turn out to be disappointing, or even embarrassing.” Arrakis, though, should not be compared with GPT-5. OpenAI has not made an official announcement about the release of GPT-5, and Altman has said that they are not training it, though they have filed a trademark for it, the Arrakis model was always expected to be a smaller one. The reasons can be that the company is shifting its focus on building AI models for wearables and smart devices, instead of just chatbots. The reasons for dropping Arrakis can be plenty. One of them can be the price for training such AI models, for which OpenAI has been reportedly spending almost a million dollars each day. Or the other one can be that the model is simply not good. Or, they want to keep it within themselves and use it in their upcoming products. \"failed to perform as expected\" isn't necessarily \"sucked\"Was it too good perhaps?— T Shirt n Jeans (@TShirtnJeans2) October 18, 2023 If the company decides to shift its focus on smaller models again, it has to weigh the value of training versus the benefit that it would give to them. At the same time, the loss of time and resources has also disappointed some of the Microsoft employees, according to the report, as they have been paying OpenAI to develop smaller models for a long time. Nonetheless, the company has been generating revenue and is back on track, according to Altman. It is expected to generate an annual revenue of $1.3 billion this year, compared to $28 million last year. Instead of scrapping Arrakis altogether, the company can integrate it within Gobi, which is expected to be something similar to what the company has released with the GPT-4 Vision model. But, there are other concerns This can also be a major setback for the company when it comes to adoption of their products. Though enterprises are heavily using GPT-4, the need for smaller models is on the rise. Some of the models are even outperforming OpenAI’s capabilities on various fronts. Even Microsoft has worked on smaller LLMs such as Orca, that run comparatively cheaper for the company. On similar lines, a recent Microsoft research also highlights some trust issues with GPT-3.5 and GPT-4. Researchers say that GPT-4 can be easily jailbroken with prompts and be led to wrong hands. Interestingly, this can also include Arrakis models as the research revolves around the same time. But according to the researchers, the bugs found in the models were reportedly fixed before the models were released. Possibly, it can be the reason that the models that OpenAI is dropping now are heavily filled with these bugs, given the smaller size of the models. It seems like even though OpenAI is riding high on the revenue waves at the moment, there indeed are chances that the company might have to drop a smaller model soon. Otherwise, Microsoft might have to steer some other route and find a different island to land on.","excerpt":"The reasons for dropping Arrakis can be aplenty. Let’s see a few here","categories":["Global Tech"],"tags":["AI Models","ChatGPT","gary marcus","GPT-4","OpenAI","Sam Altman"],"author_name":"Mohit Pandey","publish_date":"2023-10-18T15:00:00","publication_year":"2023","word_count":732,"keywords":["AI Models","ChatGPT","Go","Sam Altman","OpenAI","AI","GPT-5","chatbots","AWS","GPT-4","GPT","gary marcus","Rust","R"],"extracted_tech_keywords":["AI","GPT-5","ChatGPT","OpenAI","chatbots","AWS","R","Go","Rust","GPT"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/openai-kills-arrakis\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":26864,"title":"Healthcare Startup HealthSignz Raises $5 Million In Funding","content":"Bengaluru and Hyderabad-based healthcare startup HealthSignz raised $5 million in a funding. The round was led by Dr Kantilal Patel, founder of New Zealand-based Nirvana Health Group. HealthSignz is an artificial intelligence-driven healthcare platform which uses a proprietary AI engine to help its users. They have tied up with hospitals, pharma, diagnostics, wellness, healthcare devices and essentials to help customers get the required service with the help of an app. Hanumantha Rao, co-founder and CEO at HealthSignz, told a web portal, “Accessibility and affordability to basic healthcare treatment continue to be a challenge in India. We see HealthSignz delivering great value for capital-intensive tertiary care centre to stand alone practising providers for treating illness, and enhancing overall wellness through disease prevention.” HealthSignz caters to both B2B and B2C users and plans to commercially launch the platform by August 2018. Dr Patel told a newswire, “With increasing population and health care needs, we see a huge potential for AI solutions to restructure how medical care is delivered.” Healthsignz is a team of professionals with a significant level of expertise spanning multiple disciplines for over four decades. Identifying health and wellness as the primary factors that drive liveliness in individuals, lead to the integration of their knowledge with their common values. Driven by a passion to contribute to the advancement of the happiness quotient and well being of communities as a whole, the team tries to touch as many lives as possible by enriching them with their products.","excerpt":"Bengaluru and Hyderabad-based healthcare startup HealthSignz raised $5 million in a funding. The round was led by Dr Kantilal Patel, founder of New Zealand-based Nirvana Health Group. HealthSignz is an artificial intelligence-driven healthcare platform which uses a proprietary AI engine to help its users. They have tied up with hospitals, pharma, diagnostics, wellness, healthcare devices and essentials to help […]","categories":["AI News"],"tags":["Healthcare Automation"],"author_name":"Prajakta Hebbar","publish_date":"2018-08-01T11:53:49","publication_year":"2018","word_count":246,"keywords":["API","funding","artificial intelligence","programming_languages:R","AI","Healthcare Automation","R","startup"],"extracted_tech_keywords":["AI","artificial intelligence","R","API","startup","funding","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/healthcare-startup-healthsignz-raises-5-million-in-funding\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10040448,"title":"How AI Is Breathing Life Into Animation","content":"“Knowing how powerful machine learning has become, it’s just a matter of time before it completely takes over the animation industry.” In recent years, deep learning has increased the modern scope of animation, making it more accessible and powerful than before. Artificial intelligence has become a shiny new weapon in the creator’s arsenal. The advancement of hardware and AI has blurred the lines between virtual and real characters(eg: movies like Alita). Something that could have taken hours to perform by animators is being done by automation in minutes. According to Statista, the global animation market is expected to grow from $259 billion in 2018 to 270 billion by 2020. Animation has reached new heights because of the rapid evolution of deep learning and the proliferation of software tools. Here are some examples of how Artificial intelligence and machine learning are bringing animation to life in studios. For rotoscoping Alice in wonderland (Source:imgur.com) Rotoscoping is the technique of creating animations by drawing each animated sequence over the corresponding live-action sequence. It lets animators produce realistic characters who move like humans. For a long time, the animation industry has been using rotoscoping to create visual effects that feel more fluid on the screen. The original Star Wars films are an excellent example of rotoscoping. Through the technique of rotoscoping, animators were able to create the visual effects of the lightsaber. Disney used this method heavily for the first time in Snow White and the Seven Dwarves. Actors would be cast to provide guides for the animators’ movements. Since the studio used the stop motion technique for many of their films, they found it an excellent alternative to traditional animation techniques, including their animated films like Alice in Wonderland. Video post-production in the animation industry is quite a repetitive process that requires manually composing, tracking and rotoscoping animation. However, AI and deep learning development have finally automated the time-consuming manual work, which has decreased cost by a considerable margin. (Source: LAIKA) LAIKA, the renowned studio that created films like Coraline, ParaNorman, Kubo and the Two Strings, and The Boxtrolls, has partnered with Intel to combine machine learning and Artificial Intelligence to develop tools that will accelerate rotoscope tasks. LAIKA’s stop-motion films use unique character design and 3D-scanned facial animation, proving a challenge for repetitive tasks. Intel is therefore focusing on tools that will help make the animators work processes more efficient and streamline the design process. For 3D face modelling Latest advances in technologies include generative models, which can produce highly realistic results in fields like semantic images or videos. Disney has made the process of designing and simulating 3D faces easier through machine learning tools. Researchers at Disney have proposed a nonlinear 3D face-modelling method that utilizes neural architectures. This system learns a network topology that maps the neutral 3D model of a face into the target facial expression. Whereas, animation tech start-up Midas Interactive has already put the wheels in motion. Jiayi Chong, a former technical Pixar director, used his experience to create a new tool called Midas Creature, which automates complex 2D character animation. With Midas Creature, artists and designers tell the engine to choreograph and figure out the movements themselves, and there will be no need for character design. For voice over (Source: Adobe Character Animator, freepik) AI is also playing a significant role in the exhaustive voice  over chores in post-production. Adobe has recently created new software that uses AI-powered lip-syncing to help animators synchronize the dialogue and motion of animated characters. With Adobe Character Animator, one can use Adobe Sensei AI technology in Character Animation to assign mouth shapes to mouth sounds. It accurately syncs the character and voice in the process of dictation through traditional frame-by-frame lip-syncing. AI has the potential to take over the mundane tasks and allow the creators to focus on critical jobs in the pipeline. The freedom to spend less time on relatively less creative works like rotoscopy can enable studios to come up with high quality content during the same time. And, with the pace at which ML research is progressing, there is no doubt that the creator’s economy will get the much needed boost.","excerpt":"“Knowing how powerful machine learning has become, it’s just a matter of time before it completely takes over the animation industry.” In recent years, deep learning has increased the modern scope of animation, making it more accessible and powerful than before. Artificial intelligence has become a shiny new weapon in the creator’s arsenal. The advancement […]","categories":["AI Features"],"tags":["Adobe","AI (Artificial Intelligence)","deep fakes","Intel","Machine Learning"],"author_name":"Ritika Sagar","publish_date":"2021-05-21T12:00:00","publication_year":"2021","word_count":694,"keywords":["API","artificial intelligence","machine learning","programming_languages:R","Adobe","AI","R","ML","Machine Learning","automation","deep learning","AI (Artificial Intelligence)","deep fakes","Intel"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","R","API","automation","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-ai-is-breathing-life-into-animation\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10103238,"title":"Sam Altman is No Longer the CEO of OpenAI","content":"Two days after revealing the plans of GPT 5, OpenAI has revealed that the company has fired Sam Altman from his position as CEO and board member. The decision, announced in an official blog post, emerged after a board review citing Altman’s inconsistent communication, which hindered the board’s ability to fulfill its responsibilities. Mira Murati, previously the Chief Technology Officer (CTO), assumes the interim CEO position, emphasising continuity amidst this leadership transition. In a statement, the board of directors said: “OpenAI was deliberately structured to advance our mission: to ensure that artificial general intelligence benefits all humanity. The board remains fully committed to serving this mission. We are grateful for Sam’s many contributions to the founding and growth of OpenAI. At the same time, we believe new leadership is necessary as we move forward. As the leader of the company’s research, product, and safety functions, Mira is exceptionally qualified to step into the role of interim CEO. We have the utmost confidence in her ability to lead OpenAI during this transition period.” More important than being one of the most brilliant and impactful people our industry has ever hadSam is one of the most generous and caring people I know. I've never met someone who has supported and lifted up more people around them than him. Couldn't be a prouder brother.— Jack Altman (@jaltma) November 17, 2023 OpenAI’s board of directors consists of OpenAI chief scientist Ilya Sutskever, independent directors Quora CEO Adam D’Angelo, technology entrepreneur Tasha McCauley, and Georgetown Center for Security and Emerging Technology’s Helen Toner. As a part of this transition, the departure of Altman, a key figure in OpenAI’s founding, coincides with Greg Brockman stepping down as board chairman while retaining his role as the company’s president. Out of 6 board members, two are out. Unless Ilya changed the side, it won’t be possible. pic.twitter.com\/88Luc3BuMl— Pratik Desai (@chheplo) November 17, 2023 The board emphasised the necessity for new leadership aligned with OpenAI’s mission of ensuring artificial general intelligence benefits humanity. Despite Altman’s significant contributions, the board expressed a loss of confidence in his ability to steer the organisation. This unexpected change arrived shortly after OpenAI’s recent DevDay conference, where Altman actively participated. His departure triggers curiosity about its connection to the company’s intricate governance structure, notably the relationship between its nonprofit and for-profit arms. OpenAI’s nonprofit arm wields substantial control over its for-profit subsidiary, OpenAI Global, LLC, which focuses on commercialising AI while aligning with the nonprofit’s mission of achieving artificial general intelligence. The board’s authority includes determining AGI achievements and regulating associated commercial terms, including partnerships with major investors like Microsoft. This shift in leadership has left questions lingering about the specific circumstances surrounding Altman’s removal, reflecting a pivotal moment in OpenAI’s trajectory as it navigates a critical phase in the AI landscape. https:\/\/twitter.com\/anothercohen\/status\/1725631445941563671 However, alongside Altman, Alex Cohen was also fired from the team. In Cohen’s words, he was the person “in charge of putting together the presentations for our board meetings”. Within 20 minutes of posting the life update on X, Cohen joined Roofer.com as lead AI researcher. Roofer.com is a San Francisco based company that provides modern roofing experience for homeowners and enterprise property owners nationwide leveraging the efficiency of drones and the power of AI. Read more: Meet Mira Murati, the 35-year old CTO of OpenAI","excerpt":"Board members collectively decided on his departure as CTO Mira Murati steps in as interim CEO","categories":["AI News"],"tags":["CEO","Mira Murati","OpenAI","Sam Altman"],"author_name":"Shyam Nandan Upadhyay","publish_date":"2023-11-18T03:04:30","publication_year":"2023","word_count":555,"keywords":["Go","Sam Altman","OpenAI","AI","ML","CEO","Mira Murati","RAG","GPT","GAN","AI research","R","llm_models:GPT"],"extracted_tech_keywords":["AI","ML","OpenAI","RAG","R","Go","GPT","GAN","AI research","llm_models:GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/sam-altman-is-no-longer-the-ceo-of-openai\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10163954,"title":"How Rust, Python, Java, Make Developers Happy","content":"Coding experience often involves working with various programming languages – Python, Java, Rust, or perhaps even some obscure ones. Most developers have a favourite, the one they can write in their sleep, but they’re also well aware of the glitches, bugs, and failure modes that come with it. Source: Statista Recently, Santiago Valdarrama, a computer scientist, recently took to X to declare his love for Python. “Python is the best language in the world. I don’t care if it’s slow because most of what I do doesn’t need to be faster than what I get.” Nevertheless, for developers who “truly care about speed”, Valdarrama suggested checking out Perforator, an open-source tool to help them profile their code and identify errors. “Try this on your codebase, and you’ll realise that the language has nothing to do with your application being slow. 99% of the time [it] is something you did, not the language’s fault.” Meanwhile, some others prefer more rigid and high-performance alternatives. Pratham Patel, a member of Rocky Enterprise Software Foundation, told AIM, “For me, it is Rust.” According to Patel, Rust stands out for its excellent first-party tooling, a strict yet helpful compiler that not only identifies errors but also suggests fixes, strong type enforcement to minimise bugs, and a solid standard library that covers most generic use cases. However, it comes with its share of challenges. Its compiler is notoriously slow, and while it starts simple, it can quickly spiral into C++-like complexity if not carefully managed. Async programming in Rust is particularly challenging, and function colouring with traits and lifetimes adds another layer of frustration. Despite these hurdles, those who master Rust find it immensely rewarding. In an exclusive conversation with AIM, Amlan Panigrahi, GenAI engineer at Deloitte, said, “I love both Python and C++. Both of them have their pros and cons. Python makes my job easier to prototype E2E AI solutions (finetuning LLMs, application backends, RAG).” On the other hand, he mentioned that C++ is indispensable for his work with resource-constrained edge devices and AI PCs. The reason behind this is that it is highly performant, has its own compiler, and is the preferred approach for launching high-performance computing (HPC) kernels. Source: LinkedIn You are Smiling at the Programming Language It turns out that your choice of programming language might say something about your happiness levels. A study using Face API analysed developer profile pictures and found that some programming communities are, well, just happier. R developers are the happiest, followed by Go, C#, and Python programmers. However, Java developers? Not so much. “I’m a happy C++ developer, but I am willing to admit it may just be Stockholm Syndrome,” one Reddit user jokingly acknowledged. Another chimed in, “I need to update my GitHub profile picture. It’s currently a photo of my dog, and he seems to be smiling. I’d estimate a 0.7 dog smile.” Source: Reddit What’s Next? Every year, new programming languages emerge in the tech world, each promising to eliminate old problems and push the boundaries of innovation. Developers thrive on reinvention, constantly seeking better, more efficient ways to build the future. Developers are always looking for better and more efficient ways to build the future, and now, the next big shift is AI. As AI-driven solutions become the backbone of industries, mastering AI-focused programming languages has never been more crucial. According to research firm Gartner, 80% of organisations are expected to integrate AI into their operations by 2026 – a massive leap from just 5% in 2023. This shift highlights the growing demand for languages that can support AI-driven solutions across sectors. Whether it’s Python’s simplicity and vast ecosystem, Java’s scalability for enterprise applications, C++’s performance-driven capabilities, or Julia’s efficiency in handling complex mathematical computations, the choice of programming language depends on the specific demands of AI applications. As AI continues to advance, the languages that power it will evolve as well, shaping the next generation of innovation. For developers, staying ahead of this curve in the rapidly changing tech landscape isn’t just an advantage; it’s a necessity.","excerpt":"Turns out, your choice of programming language might say something about your happiness levels.","categories":["AI Features"],"tags":["AI (Artificial Intelligence)","Java","Programming Languages","Rust"],"author_name":"Vidyashree Srinivas","publish_date":"2025-02-18T14:14:33","publication_year":"2025","word_count":677,"keywords":["Go","GenAI","AI","ML","Programming Languages","RAG","Python","Aim","Rust","R","Java","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","ML","GenAI","Aim","RAG","Python","R","Go","Rust","Java"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-rust-python-java-make-developers-happy\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10072652,"title":"All Tech Giants: On your Mark, Get Set – Slow!","content":"Case 1: On January 18, 2022, Microsoft announced its plans to acquire leading game development firm Activision Blizzard, focusing on accelerating growth in the gaming business. The company called the deal a ‘building block for the metaverse’. The deal stood at $95 per share in an all-cash transaction valued at $68.7 billion. Once closed, Microsoft would become the world’s third-largest gaming company in terms of revenue, tailing Tencent and Sony. Activision has studios globally that own popular franchises such as ‘Candy Crush’, ‘Call of Duty, and ‘Warcraft’. Case 2: Recently, the US Federal Trade Commission probed Meta’s virtual reality unit Oculus over anti-competitive practices and the acquisitions of VR apps, scrutinising small acquisitions. Case 3: The NVIDIA-Arm deal got scrapped due to “significant regulatory challenges”. Nevertheless, the takeover was announced in September 2020, aiming to create “the world’s premier computing company for the age of AI”. Victory for one, blunder for the other two. Why is it that Microsoft’s deal with multiple large companies remains unseen by regulators? Antitrust authorities worldwide have constantly adopted frameworks to tackle competition among big-tech companies. Though major shifts in competition policy issues are at their peak due to the evolving trends, regulations still stand insignificant. Escaping the antitrust scrutiny In September 2021, The Federal Trade Commission (FTC) published a report on M&As of five top companies in the US that have escaped the antitrust laws. These were Alphabet\/Google, Amazon, Apple, Facebook, and Microsoft. (FTC Study, 2021) Chairman of FTC, Lina Khan, said that the report aims to regulate the big companies. ​​”It captures the extent to which these firms have devoted tremendous resources to acquiring start-ups, patent portfolios, and entire teams of technologists—and how they could do so largely outside our purview,” she said. Since the report came out, the antitrust crackdown on internet companies continued, with ongoing investigations to examine the mismanagement of user data. US internet giants have been put through several rounds of questioning, along with lawsuits that leave them with no option but to amend their business strategies. An example would be the shutting down of ‘Sold By Amazon’, which faced allegations of price fixing. Such regulations will limit enterprises from monopolising the market. India’s role in policymaking In July 2019, the Competition Law Review Committee (CLRC) recommended the government to introduce necessary measures and a deal-valued threshold for merger notification, which would be considered for the Competition (Amendment) Bill, 2020. Policymakers have turned their backs on the laissez-faire approach to regulation to bring limitations and regulate the corporations. The Competition Act was established in 2002 but implemented only in 2009. The CA02 framework aims to sustain competition, seeking to protect consumer interest. It empowers the Competition Commission of India (CCI) to screen a select set of M&As before they take effect. In addition, the law ensures that the CCI can impose sanctions if it finds any enterprise restricting competition by abusing power. The Competition Act, 2002 (Competition Commission of India) In a draft law, India recently proposed to mandate antitrust scrutiny for M&As valued above Rs 2,000 crore (&250 million). Under the current law, the CCI reviews mergers and acquisitions surpassing asset size or turnover thresholds. Lawyers allege that Facebook acquired WhatsApp in 2014 for $19 billion, which required no CCI clearance, despite being a major player in the Indian market. Calling the proposed amendments to the merger control regime positive and progressive, Saksham Malik, programme manager, The Dialogue, said, “The 2020 Bill empowered the central government to prescribe additional criteria for notifying transactions in ‘public interest’. This has caused uncertainty about potential criteria that may be notified. The latest 2022 amendments resolve this issue. A clear deal value threshold of Rs. 2,000 crore has now been provided. This has two benefits: It removes the uncertainty of the 2020 Bill and ensures that global transactions, which often did not meet the existing asset and turnover-based thresholds but still had an impact on competition, would be notified to the CCI. While global transactions may be a part of these, policymakers have been prudent about including a local nexus test to ensure that only transactions wherein the party have ‘substantial operations in India’ are required to be notified. Malik emphasised that the CCI can scrutinise combinations in other parts of the world that are relevant to Indian markets. While these are welcome changes, it would be beneficial if the CCI came out with literature on how these thresholds will be implemented. An explanation of the term ‘substantial business operations in India’ would help apply the local nexus test efficiently. As gatekeepers for consumers and markets, tech companies have significant power that also invites possible risks, manipulating communication and transactions. Competition law and policy should be regulated in time, allowing innovative business enterprises to be competitive, yet in fine fettle.","excerpt":"In September 2021, the FTC published a report on M&As of five top companies in the US that have escaped the antitrust laws. These were Alphabet\/Google, Amazon, Apple, Facebook, and Microsoft.","categories":["AI Trends"],"tags":[],"author_name":"Bhuvana Kamath","publish_date":"2022-08-11T17:00:04","publication_year":"2022","word_count":798,"keywords":["Go","API","AWS","AI","cloud_platforms:AWS","RPA","Aim","ViT","Rust","R"],"extracted_tech_keywords":["AI","Aim","AWS","R","Go","Rust","API","ViT","RPA","cloud_platforms:AWS"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/all-tech-giants-on-your-mark-get-set-slow\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10064634,"title":"A guide to quadratic approximation with logistic regression","content":"Modelling and evaluating the relationship between a categorical dependent variable and continuous or discrete explanatory variables are the goals of Logistic Regression. It uses linear discriminant analysis and is the reason for calling a classification algorithm regression. There are different methods in the logistic regression model through which it can classify data, one of which is the Newton Raphson method which would be explained in this article moving forward. Following are the topics covered in this article. Table of contents About Logistic RegressionWhat is the Newton Raphson method?Quadratic approximation in python Let’s start with a brief introduction to logistic regression. About Logistic Regression A logistic regression analysis reveals the relationship between a categorical dependent variable and a set of independent variables. There is no assumption of normal distribution for the independent variables in logistic regression. In addition to the regression equation, the report includes odds ratios, confidence limits, likelihood and deviance. As a part of the comprehensive residual analysis, a log regression model can generate diagnostic residual reports and plots. A subset selection search is performed to find the best regression model with fewer explanatory variables. For determining the best cutoff point for the classification, ROC curves are used to provide confidence intervals on predicted values. The results are automatically categorized by removing rows that were not used in the analysis. There is a certain assumption that is being made by Logistic Regression which is stated as: There should not be a high correlation between the independent variables.The independent variables should be linearly related to the log odds. If you’re not familiar with log-odds, we’ve included a brief explanation below.The larger the sample size, the more reliable (and powerful) you can expect the results of your analysis to be. Logistic regression uses log-odds which is an alternate way of expressing probabilities. But there is some difference between odds and probabilities: Probability is the ratio of things happening to everything that could happen, while odds are the ratio of something happening to something not happening. As we know Logistic Regression uses the concept of the Log-Likelihood of the Bernoulli distribution and also covers explained as the sigmoid function or the logistic function. It also uses the gradient descent method in conjunction with the Hessian square matrix. So, moving forward we would see how Newton and Raphson method is used to find the roots and to maximize the likelihood estimation. Let’s understand how to log odd is used to estimate likelihood. What is the Newton Raphson method? Newton’s method is an iterative method for finding the roots of the convex function. But log-likelihood is a concave function which means it has only one global max. So, to apply Newton’s method it should be applied to the derivative of the function. In this case, the cost function is quadratic so Newton’s method would be applied to its derivative which could also be called a quadratic approximation. In this manner, Newton’s method is different from the gradient descent method because the gradient descent method is applied to the cost function and minimizing the cost function is the goal of the method. In the above representation, the working of Newton’s method is shown in which there are multiple iterations to find the exact root for the quadratic approximation. Let’s see how this method is implemented in python. Quadratic approximation in python The logistic regression package is imported from the sklearn library. In logistic regression, there is a parameter called ‘solver’ which needs to provide the method to be used for the classification. We would be using “newton-cg” Syntax class sklearn.linear_model.LogisticRegression(penalty='l2', *, dual=False, tol=0.0001, C=1.0, fit_intercept=True, intercept_scaling=1, class_weight=None, random_state=None, solver='lbfgs', max_iter=100, multi_class='auto', verbose=0, warm_start=False, n_jobs=None, l1_ratio=None) Importing libraries and packages import pandas as pd import numpy as np from sklearn.model_selection import train_test_split import seaborn as sns import matplotlib.pyplot as plt from sklearn.linear_model import LogisticRegression Reading the data and pre-processing: df=pd.read_csv(\"\/content\/drive\/MyDrive\/Datasets\/cancer.csv\") df.drop(['Unnamed: 32',\"id\"], axis=1, inplace=True) df.diagnosis = [1 if each == \"M\" else 0 for each in df.diagnosis] df.head() The data is related to the diagnosis of breast cancer in which the “diagnosis” is encoded as 1 and 0 which is malignant and begins. The data has a total of 569 records and 31 features including the dependent variable. Splitting the data: X=df.drop('diagnosis',axis=1) y=df['diagnosis'] X_train, X_test, y_train, y_test = train_test_split( X, y, test_size=0.30, random_state=42) print(\"Shape of independent variable train\",\"\\nno of rows=\",X_train.shape[0],\"\\nno of columns=\",X_train.shape[1]) print(\"-----------------------------------------------\") print(\"Shape of dependent variable train\",\"\\nno of rows=\",y_train.shape[0],\"\\nno of columns=1\") print(\"-----------------------------------------------\") print(\"Shape of independent variable test\",\"\\nno of rows=\",X_test.shape[0],\"\\nno of columns=\",X_test.shape[1]) print(\"-----------------------------------------------\") print(\"Shape of dependent variable test\",\"\\nno of rows=\",y_test.shape[0],\"\\nno of columns=1\") The data is split into tests and trains with a ratio of 70:30. Fitting the data: lr=LogisticRegression(max_iter=100,solver='newton-cg') lr.fit(X_train,y_train) lr_pred=lr.predict(X_test) log_pred=lr.predict_proba(X_test)[:,1].round(2) df_pred=pd.merge(X_test,pd.DataFrame(lr_pred,columns=['prediction']),left_index=True,right_index=True) df_final=pd.merge(df_pred,pd.DataFrame(log_pred,columns=['probability']),left_index=True,right_index=True) df_final.head() Fitted the training dataset in the logistic regression with the solver as ‘newton-cg’ and the model predicted based on the test dataset and also the probability of the prediction. Let’s plot the prediction and probability of a feature and observe the function. Visualizing the regression model: fig, axes = plt.subplots(1, 2 ,figsize=(15, 5),sharex=True) fig.suptitle('Plot for probability and prediction of data points in \"Area_mean\" feature') sns.regplot(x=\"area_mean\", y='probability',data=df_final ,logistic=True, ci=None,ax=axes[0],scatter_kws={\"color\": \"black\"}, line_kws={\"color\": \"red\"}) sns.regplot(x=\"area_mean\", y='prediction',data=df_final ,logistic=True, ci=None,ax=axes[1],scatter_kws={\"color\": \"black\"}, line_kws={\"color\": \"red\"}) plt.show() In the above plot, the data points of the area_mean feature are used to visualize the newton’s method learner line. The left side subplot is for the probability of the data points to be either classified as 0 or 1 and on the right side are the predictions by the model for the data points. So, in the probability graph, the data points above the learner line are classified as 1 as shown in the prediction plot and similarly, the below are classified as 0. Final Verdict The logistic regression could be used by the quadratic approximation method which is faster than the gradient descent method. For the approximation method, the Newton Raphson method uses log-likelihood estimation to classify the data points. With a hands-on implementation of this concept in this article, we could understand how quadratic approximation could be used in Logistic Regression. References Link to the above codeRead on the wiki about Newton’s method","excerpt":"The logistic regression can be used with the quadratic approximation method which is faster than the gradient descent method.","categories":["Deep Tech"],"tags":["AI (Artificial Intelligence)","Data Science","logistic regression","Machine Learning"],"author_name":"Sourabh Mehta","publish_date":"2022-04-10T18:00:00","publication_year":"2022","word_count":1014,"keywords":["Go","NumPy","programming_languages:R","AI","logistic regression","Machine Learning","Python","Seaborn","programming_languages:Python","Data Science","Matplotlib","R","AI (Artificial Intelligence)","Pandas"],"extracted_tech_keywords":["AI","Pandas","NumPy","Matplotlib","Seaborn","Python","R","Go","programming_languages:Python","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/a-guide-to-quadratic-approximation-with-logistic-regression\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10094378,"title":"Github Released Git 2.4 Update to Streamline Handling of Unreachable Objects","content":"The open-source Git project unveiled Git 2.4, incorporating contributions from 95 contributors, 29 of whom are new. This update introduces various features and bug fixes. Read more: Top 9 GitHub repositories for the TensorFlow community Efficient Handling of Unreachable Objects The handling of unreachable objects in Git repositories has been improved to enhance performance and reduce repository size. Objects in Git can be either “reachable” or “unreachable.” Reachable objects are those that can be accessed by following a branch or tag in the repository’s history. Unreachable objects are not reachable through any branch or tag. Git periodically removes unreachable objects to compress the repository size. When running the git gc command, Git collects reachable objects into a pack, stores recently written unreachable objects separately, and discards the remaining unreachable objects. Previously, Git stored unreachable objects as loose copies, which could lead to repository bloat and inode exhaustion. In Git 2.41, cruft pack generation is enabled by default during regular git gc operations. This means that a cruft pack will be generated in the repository, providing the benefits mentioned above. To learn more about cruft packs and their impact on Git’s garbage collection, you can refer to the previous post titled “Scaling Git’s garbage collection.” Read more: Why Indian IT Prefers CodeNet over GitHub Copilot On-Disk Reverse Indexes for Improved Performance Starting from Git 2.41, a new “.rev” file will be present in the repository’s “.git\/objects\/pack” directory. This file stores information similar to the packfile index (stored in “.idx” files). Pack indexes map object positions in name order and pack order. Previously, the reverse index was generated on-the-fly, but Git 2.31 introduced an on-disk reverse index. It generates and stores the reverse index alongside the packfile as a “*.rev” file. This pre-computed reverse index improves performance, especially in large repositories, for operations like pushing changes and determining object sizes. In Git 2.41, reverse index generation is enabled by default. Running “git gc” after upgrading will result in faster operations. Tests have shown significant speed-ups, with improvements of up to 1.49x in pushing changes and nearly 77x in computing object sizes. Read more: Rust Turns GitHub’s Long-standing Problem to Dust","excerpt":"In Git 2.41, reverse index generation is enabled by default","categories":["AI News"],"tags":[],"author_name":"Shritama Saha","publish_date":"2023-06-02T14:57:06","publication_year":"2023","word_count":358,"keywords":["GitHub","programming_languages:R","AI","Git","ai_frameworks:TensorFlow","Rust","TensorFlow","R","programming_languages:Rust"],"extracted_tech_keywords":["AI","TensorFlow","R","Rust","Git","GitHub","ai_frameworks:TensorFlow","programming_languages:R","programming_languages:Rust"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/github-released-git-2-4-update-to-streamline-handling-of-unreachable-objects\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10039679,"title":"10 Self-Supervised Learning Frameworks &#038; Libraries To Use In 2021","content":"Self-supervised learning is gathering steam, slowly but surely. A relatively new technique, self-supervised learning is nothing but training unlabeled data without human supervision. Yann LeCun described it best: Reinforcement learning is like a cherry on a cake, supervised learning is the icing on the cake, and self-supervised learning is the cake. In self-supervised or unsupervised learning, the system learns to predict part of its input from already existing inputs, he said. Source: GitHub Most tech evangelists liken self-supervised learning models to young children, always curious and learning new information from observation. The latest examples of self-supervision include Facebook’s DINO and ViSSL (Vision library for Self-Supervised Learning); Google’s SimCLR; OpenSelfSup and SfMLearner, etc. Below, we have curated a list of the most popular self-supervised learning models, frameworks, and libraries. DINO DINO, a self-supervised learning vision transformers (ViT), is used to segment unlabelled and random images and videos without supervision. In other words, self DIstillation with NO labels. The model generates high accurate segmentation with self-supervised learning and suitable architecture. Also, DINO requires limited computing resources to train models. Lightly Lightly is a computer vision framework for self-supervised learning. It helps in understanding and filtering raw image data and can be applied before any data annotation step. The learned representations can further analyse and visualise datasets, alongside selecting a core set of samples. s3prl s3prl is an open-source toolkit that stands for Self-Supervised Speech Pre-training and Representation Learning. Self-supervised speech pre-trained models are called upstream in this toolkit and are used in multiple downstream tasks. SimCLR SimCLR is a Simple framework for Contrastive Learning of Visual Representations. In its latest version (SimCLRv2), the distilled or self-supervised models have been used. It is primarily used for image segmentation and image classification. OpenSelfSup OpenSelfSup is an Open-source unsupervised or Self Supervised representation learning toolbox based on PyTorch. It follows a similar code architecture of MMDetection, and is very flexible as it integrates various self-supervised tasks, including classification, feature learning, joint clustering and contrastive learning. SfMLearner SfMLearner is a self-supervised learning framework used for tracking depth and ego-motion estimation from monocular videos. BYOL BYOL or ‘Bootstrap Your Own Latent’ is a new approach to self-supervised image representation learning on PyTorch. It is one of the simple methods for self-supervised learning that achieves cutting edge results without constructive learning and having to design negative pairs. The repo offers a module that lets researchers build an image-based neural network right away from unlabelled or random image data. DIG DIG or Dive Into Graphs is a turnkey library that provides a unified testbed for higher level, research-oriented graph deep learning tasks like graph generation, self-supervised learning, explainability and 3D graphs. DIG also enabled researchers to develop their methods within its extensible framework and compare with existing baseline methods using standard datasets and evaluate metrics seamlessly. VL-BERT VL-BERT library is used to pre-train generic visual-linguistic tasks, including visual commonsense reasoning, visual question answering, and referring expression comprehension. EpipolarPose EpipolarPose is a self-supervised learning method for 3D human pose estimation using multi-view or epipolar geometry. It does not need any 3D ground-truth data or camera extrinsic. While training, EpipolarPose estimates 2D poses from multi-view images and then utilises epipolar geometry to obtain a 3D pose and camera geometry. It takes an RGB image to produce a 3D pose result.","excerpt":"Self-supervised learning is gathering steam, slowly but surely. A relatively new technique, self-supervised learning is nothing but training unlabeled data without human supervision. Yann LeCun described it best: Reinforcement learning is like a cherry on a cake, supervised learning is the icing on the cake, and self-supervised learning is the cake. In self-supervised or unsupervised […]","categories":["AI Trends"],"tags":["machine learning libraries"],"author_name":"Amit Naik","publish_date":"2021-05-06T11:00:00","publication_year":"2021","word_count":549,"keywords":["Go","AI","neural network","PyTorch","ML","Transformers","computer vision","Git","deep learning","machine learning libraries","R"],"extracted_tech_keywords":["AI","ML","deep learning","neural network","computer vision","PyTorch","Transformers","R","Go","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/10-self-supervised-learning-frameworks-libraries-to-use-in-2021\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10083862,"title":"2022 — The Year Where AI Went Wrong","content":"The year of 2022 ended with a lot of updates and anticipation around the generative AI, which made routine headlines globally. And while the folks around the tech industry were optimistic about AI being the future of the world, with the notion that AI will have the capabilities to enable automation, predictive analytics, and personalised experiences that will unlock new opportunities and business models; AI researchers in totality are not particularly happy about it. In the recent AGI debate, where many of the global AI researchers were present, Gary Marcus said that the common point that nearly all panellists found worrisome was the near-future of Artificial Intelligence. LLMs and Cognition When it comes to Artificial Intelligence, ChatGPT has recently taken over the internet. Before that, the auto image generating models like Midjourney and Stable Diffusion were the talk of the town. The 94-year-old Noam Chomsky, however, believes that no matter how many models come with updated data and parameters, the fundamental flaw in the LLMs can never be remedied. According to Noam Chomsky, “The problem is quite general and important. The media are running major thought pieces about the miraculous achievements of GPT-3 and its descendants, most recently ChatGPT, and comparable ones in other domains, and their import concerning fundamental questions about human nature.” Chomsky further elaborated that large language models have been shown to have several flaws and, while they may improve with more data and more parameters, there is a fundamental flaw that will always persist. He explained that due to their design, these systems cannot distinguish between possible and impossible languages. Would the current approach to artificial intelligence ever be able to tell us anything about what makes the human mind what it is? Chomsky believes that the more the systems are improved, the deeper the failure becomes. As we continue to delve into the realm of artificial intelligence, we must ask ourselves: Will our current approach ever truly unveil the mysteries of the human mind? Some, like Chomsky, argue that the more we strive to improve upon our AI systems, the further we stray from understanding the intricacies of the human brain. With each advancement, we may uncover new insights but Chomsky posits that the deeper we go, the more apparent the failure of our current approach becomes. The question remains: How do we bridge the gap between artificial intelligence and the inimitable nature of the human mind? “They are telling us nothing about language and thought, about cognition generally, or about what it is to be human. We understand this very well in other domains. No one would pay attention to a theory of elementary particles that didn’t at least distinguish between possible and impossible ones.” Scaling is not better! As GPT 4 is round the corner, which is expected to be built on 100 trillion parameters, one has to wonder how much is too much. Is scaling too fast, always beneficial? Dileep George, DeepMind researcher, doesn’t think so. George provided two examples of significant dates in aviation history—the first aeroplane in 1903 and the first non-stop transatlantic flight in 1919—and asked others to guess the year of the Hindenburg Disaster. He pointed out that the answer, 1937, was likely surprising to many because it occurred much later despite the rapid advancements made in aviation prior to that time. According to George, it is important to ensure that AI has a foundational understanding of the world before attempting to scale it up. This is because scaling up current AI programmes without addressing the fundamental differences between them and human-like intelligence may not lead to the desired outcomes. Instead, we should focus on bridging these differences in order to achieve true artificial intelligence. Dark Matter While the advancement in the field of AI in the year of 2022 is notable, do we really know how deep we can go? AI is making mistakes and will continue to make mistakes in upcoming years but what is the main reason behind it? Yejin Choi, the Brett Helsel Professor at the University of Washington, warns that artificial intelligence may continue to make mistakes on uncommon or unforeseen cases due to our limited understanding of the inner workings of these networks. She compares this to the concept of dark matter in physics—something that scientists are aware of but cannot fully measure or comprehend. It seems that the true nature of language and intelligence may remain somewhat of a mystery, akin to the mysterious dark matter of the universe. Regardless, it may be worth considering that all these researchers, along with many others, were concerned about the near future of AI and if it makes sense. Will the AI community in 2023 be able to come together and work on the flaws that the researchers discussed in the recent AI debate?","excerpt":"Noam Chomsky believes that no matter how many models come with updated data and parameters, the fundamental flaw in the LLMs can never be remedied.","categories":["AI Features"],"tags":["Automation","future of AI","gary marcus","Generative AI","MidJourney","Stable Diffusion"],"author_name":"Lokesh Choudhary","publish_date":"2023-01-03T13:00:00","publication_year":"2023","word_count":800,"keywords":["Go","MidJourney","ChatGPT","artificial intelligence","AWS","AI","Stable Diffusion","R","Automation","future of AI","gary marcus","Ray","analytics","generative AI","Generative AI","predictive analytics"],"extracted_tech_keywords":["AI","artificial intelligence","analytics","generative AI","ChatGPT","Ray","predictive analytics","AWS","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/2022-the-year-where-ai-went-wrong\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":16686,"title":"Is data the only barrier to developing chatbots? Check out the pitfalls to avoid","content":"There is no denying that adoption of chatbots is at all time high and voice-activated chatbots are laying the foundation of a more focused customer experience. But while chatbots have become an essential toolset in every company’s technology stack and delivers great value to customers, there are several barriers to developing chatbots. While AI’s future is in the voice and companies are cashing in on providing a seamless conversational experience through messaging platforms, chatbots have to evolve to provide a strong user experience and become the voice of the brand. On the technological front, enterprises face several challenges, from too little to petabytes of data, user language, functionality and security and integration issues among others.  Monitoring chatbots and keeping a track of responses to queries is also a prime concern for companies and greatly hinders successful adoption. The bottomline is to make conversational voice and text based assistants more perceptive. Analytics India Magazine lists down key challenges in developing chatbots Data: Most companies are sitting on a wealth of data – structured and unstructured and struggle to train all the data. When developing chatbots, the reigning sentiment is that either the data out there is too much or too little. The rule of thumb is that more users means more data and that’s why most chatbot developers try to populate the database with entries from real people and sometimes this can be a capital intensive task. The whole idea is that more data can lead to better conversation paths in customer automation. For a conversational chatbot, you may need to create broad events and entities that can be an intensive task and the data out there has to be processed and tagged. Facebook’s Wit.ai platform and Google’s API.AI provide developers sample default questions to create common intents for users. NLP: Natural Language Processing is the driving force behind chatbots and it’s the computational approach to find answers by parsing user’s inputted text into entities, intents, actions and contexts. Now, the need of the hour is a Natural Language learning system that can expand the chatbot’s syntax and grammar by recommending new phrases. With all the noise around AI in the last few years, developers are now pushing the boundaries by leveraging deep learning to program chatbots to be more human-like. However, the tech around NLP and most importantly deep learning is in the early stages and it will take some more time before the CX (customer experience) and engagement becomes more intuitive and robust. From the automation context, chatbots sit high on the value chain, by driving customer interaction across domains such as ecommerce, healthcare, retail and more. If businesses don’t want to be left behind in the automation wave, they must step up their investment in machine learning infrastructure, scale data collection process and iterate their algorithms and improve performance over time. Security: Most businesses are reluctant to use bots since it can’t be monitored all the time and there is the inability to guarantee the security of data. And most of the security concerns around bots arise in the financial sector while securing the transmission of client data. Although, chatbots follow two-factor authentication and boast of end-to-end encryption, security issues continue to plague bots. Let’s not forget the creepy Facebook AI story that the bots invented a new language for more interaction. Testing: Testing chatbots can be a very teething issue and it is difficult to test the correct behavior of chatbot. While human QA may not be entirely possible and some testing will have to be automated, businesses should develop a framework  to validate the functionality of chatbots. Making a difference in a crowded market: According to the findings from Juniper Research a chatbot enquiry can save upto 4+ minutes of valuable customer service time and can save businesses $8 billion annually by 2022, up from $20 million in 2017. Another survey by Oracle indicates that 80% of businesses want chatbot by 2020 and chatbot is the future of conversational commerce. While the statistics may be resounding, chatbots have become a tough territory to beat and issues such as integration, language understanding and building NLP engines have to be factored in before jumping on the chatbot bandwagon. Outlook Increasingly, businesses are latching on where customers are – on popular messaging platforms such as Facebook Messenger, Slack, Skype, HipChat and WhatsApp and integrating marketing services on these platforms. And while chatbots are a great way to build brands and they present an easy way for companies to scale mobile messaging with users, businesses should have a clear understanding of the service and customer experience they want to build. And as NLP capabilities get more sophisticated, digital assistants can usher in big gains and enable frictionless communication. With more tech behemoths launching their platforms, the entry barrier has also been lowered. Another big player that is lowering the entry barrier to AI is Amazon Lex by AWS that does all the NLP and machine learning heavy lifting for customers, helping them introduce a chatbot easily and effectively. It’s time to take advantage of the conversational commerce fueled by chatbots.","excerpt":"There is no denying that adoption of chatbots is at all time high and voice-activated chatbots are laying the foundation of a more focused customer experience. But while chatbots have become an essential toolset in every company’s technology stack and delivers great value to customers, there are several barriers to developing chatbots. While AI’s future […]","categories":["IT Services"],"tags":["chatbot ai"],"author_name":"Richa Bhatia","publish_date":"2017-08-03T04:57:38","publication_year":"2017","word_count":850,"keywords":["machine learning","AWS","AI","chatbots","ML","RAG","NLP","deep learning","chatbot ai","analytics","R"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","NLP","analytics","RAG","chatbots","AWS","R"],"url":"https:\/\/analyticsindiamag.com\/it-services\/is-data-the-only-barrier-to-developing-chatbots-check-out-the-pitfalls-to-avoid\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":38066,"title":"Inside Intel’s Latest Breakthrough Is Foveros, 3D Stacking Technology","content":"Intel, the world’s largest chip maker recently announced about the CPU architecture “Lakefield” which is built upon Intel’s Sunny Cove microarchitecture with AI and Crypto Acceleration. Sunny Cove is designed to increase performance per clock and power efficiency for general purpose computing tasks. It enables reduced latency and high throughput, as well as offers much greater parallelism which is expected to improve experiences from gaming to media to data-centric applications. The technical strategy of the hybrid CPU architecture enables a combination of different CPU core architectures on a single product such as the processor has 5 cores combining one 10nm Sunny Cove core with four 10nm-based small core. Hence, it results in power efficiency, immersive graphics, I\/O and memory within the 12nm hybrid CPU architecture. Then the hybrid CPU architecture is packaged in 3D packaging technique known as Foveros where different pieces of IP together are stacked in 3 dimensions instead of 2D. According to this journal, the chip maker demoed a design built with FOVEROS. It was an x86 design combining a large Sunny Cove X86 core with four smaller Atom cores, all on one 10nm device, to form a hybrid x86 CPU. It had one Sunny Cove core with 0.5 MB of private medium-level cache, and four Atom cores with a shared 1.5 MB L2 cache. Then there is an “Uncore” that contains 4MB of last-level cache, a quad-channel 4×16-bit memory controller supporting LPDDR4, a Gen11 graphics controller, a Gen 11.5 display controller, an imaging IPU, and MIPI with DisplayPort 1.4. On top of that was package-on-package (PoP) memory. Intel’s chief architect, Raja Koduri is one of the masterminds behind this 3D stacking technology. Foveros is a 3D packaging technology which brings the benefits of 3D stacking to enable logic-on-logic integration. It paves the way for devices and systems combining high-performance, high-density, and low-power silicon process technologies and is expected to extend die stacking beyond traditional passive interposers and stacked memory to high-performance CPU, graphics and processors. The first Foveros product will combine a high-performance 10nm compute-stacked chiplet with a low-power 22FFL base die. It will enable the combination of world-class performance and power efficiency in a small form factor and it also enables the combination of world-class performance and power efficiency in a small form factor. According to reports, Intel says that Foveros products will be shipping in the second half of 2019 and that the technology is ready for mass-market production—not just specialized or customized processors, but mainstream CPUs. Lakefield, built with Foveros According to this report, Ramune Nagisetty, Director of Intel’s Process and Product Integration said that when dies are stacked that known good dies (i.e. those that pass yield tests) are used, which requires bare die testing before assembly. Some of Intel’s previous development processes have needed to be adjusted in order to assist for technologies like Foveros and products like Lakefield, as well as other products in the future. She further added that Intel has not specifically looked into advanced cooling methods for Foveros type chips, but did expect work in this field over the coming years, either internally or externally. Perks Of This Technology This technology helps in manufacturing smaller chips which result in manufacturing smaller boards in more efficient and better design which will unleash greater performance of platform innovation in the industry as well as innovative devices. Advanced manufacturing processes and packaging Super-fast memory Interconnects Embedded security features Importance In Machine Learning In the current scenario, two classes of chip designs are being used which is Graphics Processing Unit (GPU) and  Field Programmable Gate Array(FPGA). GPU has been portraying a very crucial role in the field of deep learning such as Convolution Neural Network (CNN), Recurrent Neural Network (RNN), etc. While the reprogramming feature of FPGA makes it more usable in the machine learning industry. The chipsets are thus helping the organisations to get impact on more machine learning technologies.","excerpt":"Intel, the world’s largest chip maker recently announced about the CPU architecture “Lakefield” which is built upon Intel’s Sunny Cove microarchitecture with AI and Crypto Acceleration. Sunny Cove is designed to increase performance per clock and power efficiency for general purpose computing tasks. It enables reduced latency and high throughput, as well as offers much […]","categories":["Global Tech"],"tags":["Intel","Intel chip"],"author_name":"Ambika Choudhury","publish_date":"2019-04-22T12:55:21","publication_year":"2019","word_count":650,"keywords":["Go","machine learning","AI","Intel chip","neural network","Ray","deep learning","RNN","GAN","CNN","R","Intel"],"extracted_tech_keywords":["AI","machine learning","deep learning","neural network","Ray","R","Go","GAN","CNN","RNN"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/inside-intels-latest-breakthrough-is-foveros-3d-stacking-technology\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":50727,"title":"Why Tamil Nadu’s Blockchain Plans Can Build An Ideal E-Governance Model For India","content":"Tamil Nadu is chipping away at a state-level arrangement for rising advancements in blockchain for service delivery and e-governance projects. Termed as Blockchain Backbone, the network will be utilised by involved parties in Tamil Nadu for straightforward, effective and secure resident-driven and between inter-departmental work processes. The blockchain system will provide better e-administration for the residents using self-sovereign use of identities and digital signatures. Also, Zero-knowledge proof-based predictive government service delivery will serve as a great way of preserving the privacy of sensitive data. The application programming utilised for giving different endorsements under this program will be reasonably adjusted with the goal that docs issued will electronically hit the Makkal Number made by Tamil Nadu e-Governance Agency (TNeGA). The goal is to ensure that there is a corruption-free flow of information between the government departments and citizens. TNeGA expressed, “A framework on Blockchain will provoke the product of every Government Department\/Agency to give the authentication\/report, as and when the resident gets qualified for it.” According to TNeGA, Government Departments, Agencies can verify the authenticity for qualifications of e-governance and government services using the resident’s consent. Experts have for a long time stated that distributed ledger technology is a perfect fit for such kind of public service record and information disbursal systems. The Blockchain platform for Tamil Nadu will be offered as a support of all Government offices, without department keeping up a Blockchain node for verifying transactions as the case with typical proof of work blockchain systems. They will have the option to check their information in the Blockchain autonomously without running a node, states TNeGA. Government organisations that don’t want the overhead of running the blockchain node can access the network through an API gateway. The network will incorporate a safe foundation comprising of blockchain cores, a business logic layer and client API gateways. In a blockchain, new data blocks are created and verified continuously, it ensures data immutability as well as who is accessing and when. Blockchain likewise gives cybersecurity and information respectability to applications concentrated on citizen engagement and experience. Use Cases Of Blockchain In Tamil Nadu Governance Government-to-government and government-to-citizen products and services will be actualised on the blockchain framework, and will likewise be deployed to expand legacy IT infrastructure by furnishing a permanent hash-encoded distributed with zero down-time. This can ensure not just more data security but also interoperability of data in a safe and immutable manner. Plus, the government is planning to combine blockchain with automated alerts. For example, a framework on Blockchain, without human mediation, will prompt the product of every Government Department\/Agency to give the testament\/record as and when the resident gets qualified for it. For instance, if a kid’s birth certificate is given, the framework at the government’s end will send verification data through an SMS at the sixth month of age of the baby. In like manner, when the youngster turns 5, an SMS to the guardians is sent for registration of Aadhar Card. Similarly, Makkal Number as of now made by TNeGA for 7 crore individuals of Tamil Nadu will electronically arrive at a Citizen Vault to be made for every occupant of Tamil Nadu, and its use will be monitored on the blockchain network to prevent any misuse. Makkal Number is the state residents data hub (SRDH)- a unified data repository with biometry (Aadhaar) enabled citizens’ data from the National Population Registry to service all government departments. Citizen documents and important government-issued certificates for public records will electronically arrive at the Citizen Vaults, which will be made for every citizen of Tamil Nadu. The residents would then be able to access documents in the vault through his\/her phone number as a client ID and OTP as a secret phrase. The blockchain network can likewise be utilised to update existing government data procedures and make new progressively dependable and secure work processes. It will be a solitary platform for developing blockchain applications for all divisions and public sector endeavours inside the state. These will be based on smart contracts developed on the open-source network- Linux Hyperledger. Tamil Nadu Is Not The Only State Pursuing Distributed Ledgers For e-Governance The Indian state of Maharashtra has declared plans to actualise blockchain. This pursues different pilots with blockchain in the state using the open-source Hyperledger. Use cases to be focused on incorporating the streamlining of farming and related activities, supply chain management, vehicle registration and public record management. Apart from Maharashtra, Telangana will house all major Blockchain innovation organisations, a gigantic incubator program and a world-class office for advancing exploration in distributed ledgers. The government had reportedly announced that land will be allotted at subsidised rates to Blockchain companies by Telangana State Industrial Infrastructure Corporation Limited,” as per the approach that tries to position Telangana as the Blockchain capital of the nation. Almost half of the states in India have started blockchain activities to address various components of citizens service delivery services. While most undertakings are in the pilot phase, many state governments have adopted a favourable approach for blockchain startups and related service providers.","excerpt":"Tamil Nadu is chipping away at a state-level arrangement for rising advancements in blockchain for service delivery and e-governance projects. Termed as Blockchain Backbone, the network will be utilised by involved parties in Tamil Nadu for straightforward, effective and secure resident-driven and between inter-departmental work processes.  The blockchain system will provide better e-administration for the […]","categories":["AI Features"],"tags":["ai certificates","Blockchain","blockchain india","Blockchain Technology","Tamil Nadu"],"author_name":"Vishal Chawla","publish_date":"2019-11-26T19:00:00","publication_year":"2019","word_count":843,"keywords":["ai certificates","Go","API","Blockchain","blockchain india","AI","innovation","ML","Git","ViT","Tamil Nadu","GAN","Blockchain Technology","R","startup"],"extracted_tech_keywords":["AI","ML","R","Go","Git","API","GAN","ViT","innovation","startup"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/tamil-nadu-blockchain-governance-india\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":21079,"title":"Mumbai Analytics Startup vPhrase Analytics Brings Natural Language Generation To Make Reports More Insightful","content":"Mumbai-based startup vPhrase Analytics, founded in 2015 by Neerav Parekh has given a new spin to everyday dashboards and visualizations drummed up by analysts. The core AI-based platform Phrazor, analyses data, derives insights and communicates these insights in multiple Indian languages. At a time when Big 3 of BI world – Tableau, Microsoft PowerBI & Qlik dominate the business landscape, vPhrase Analytics is demystifying the visualization landscape by drumming smart insights alongside text that makes the job of analyzing easier. Phrazor is now deployed across domains –  banks, brokerage firms, healthcare, CPG companies, media and entertainment and makes the job of reporting easier by giving a personalized narrative and highlighting areas that need due attention. Let’s Talk About the Eureka Moment Founder Neerav Parekh Prod Parekh about the idea behind starting vPhrase and he reveals how in his previous role as the founder of a digital marketing agency, he found his project managers spending an inordinate amount of time in the clients’ office, explaining the report. “We were a Performance Marketing agency, responsible for getting day-to-day sales for clients like Dr Batra and Hiranandani. Our project managers had to send reports, to assess how performance pans out. They would take half a day to create this report and eventually when they send this report to the customers, customers would invariably say why don’t you come and explain to me,” he explained. That’s when Parekh came up with the idea of a product that could build up on the visualization end by drumming up text alongside charts. “The only way to do it was automating the task and that is when I thought of building Phrazor. We built a prototype, it worked out and when the prototype was successful I started thinking of potential use cases for such a technology,” he said. Today, the use cases have gone beyond the scope of marketing and expanded to healthcare, BFSI, retail, CPG companies, media and many more. Understanding Phrazor Essentially it is a combination of two things – first is the data analysis bit and the other is the language writing skills. “Language writing bit is where computational linguistic comes in and that is where Natural Language Generation comes which is the subset of AI,” said Parekh. The entire product is built on Python and the team took the NLTK language toolkit for building the platform. “But it was not production ready and was not sufficient for the kind of use cases we were doing, so we built a library from the ground up ourselves,” said Parekh. Essentially, the platform takes in data, structured and unstructured, analyses it, comes up with insights and then it communicates those insights using bullet points. The platform is capable of using any language, but currently, libraries have been developed for four different languages — Gujarat, Hindi, Tamil and English. “For us tomorrow if a client comes from France and has a requirement in French, we can do that too,” he said. According to Parekh, this is the first of its kind product in India. But globally, companies like Arria NLG, Narrative Science and Automated Insights work in the same space. Parekh emphasizes his startup scores over the global competition in a lot of ways: All of us have a DIY product and all of us also provide customized solutions. In terms of customized solutions, all of us are almost at par. But in terms of DIY products, our product is superior because of the kind of use cases and complexity our product offers which is currently not possible with those products In short, the complexity of use cases those products can handle is far less than the kind of complexity we can handle. You can have real-life cases directly created by yourself using our product without any involvement from our side. Use Cases In terms of use cases, it can be bucketed in two broad areas: Internal Communications: There are several clients who use this platform for communicating with their customers. Case in point —  Motilal Oswal Securities use it to send out reports to their own investors. HDFC Bank that has more than 4500 branches in the country use the platform to send out personalized reports to each of the branch managers which gives details about performance vs target. Measuring Performance: FMCG leaders like Unilever and Marico use the platform for Business Intelligence BI to analyze market research data from Nielsen to take key decisions like advertising, price points for a particular region. The task, that was previously performed manually and would take days on end is now automated in seconds by Phrazor. Besides, there are also research and compliance use cases where banks and brokerages have to send out reports to regulators. “These reports are in a fixed format and they come from a fixed set of data. Instead of spending days analyzing data and writing reports, our platform automates the task in seconds,” shared Parekh. Phrazor to come as a plugin for PowerBI soon Parekh makes this abundantly clear that in the BI landscape, vPhrase Analytics is not competing with the Big 3 of BI. “These platforms give you the data analysis capabilities and capability of making dashboards but they don’t give you the capability of storytelling. They do not create insights for you,” he said. Currently, the team is in the process of creating plugins for these platforms and the plug-in for PowerBI will be out by March-end. This will be extended to other players as well like Tableau and Microsoft Office as well. On the client-side, the Mumbai-based analytics startups count some of the biggest names from banking, FMCG, retail and media world like HDFC Bank, Motilal Oswal Securities, Unilever, Accenture, consulting companies like Accenture, Wipro, TCS, Viacom18, Sony, Moneycontrol.com. Plans for expansion are also afoot and soon the startup will have their second office in Singapore, revealed Parekh.","excerpt":"Mumbai-based startup vPhrase Analytics, founded in 2015 by Neerav Parekh has given a new spin to everyday dashboards and visualizations drummed up by analysts. The core AI-based platform Phrazor, analyses data, derives insights and communicates these insights in multiple Indian languages. At a time when Big 3 of BI world – Tableau, Microsoft PowerBI & […]","categories":["AI Startups"],"tags":["analytics insight","big data and analytics everyday life","NLTK"],"author_name":"Richa Bhatia","publish_date":"2018-01-30T04:38:02","publication_year":"2018","word_count":977,"keywords":["business intelligence","Go","AI","Git","RAG","Python","analytics insight","analytics","NLTK","big data and analytics everyday life","R","startup"],"extracted_tech_keywords":["AI","analytics","NLTK","RAG","Python","R","Go","Git","business intelligence","startup"],"url":"https:\/\/analyticsindiamag.com\/ai-startups\/mumbai-analytics-startup-vphrase-analytics-brings-natural-language-generation-to-make-reports-more-insightful\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10058755,"title":"Can Github Copilot create art?","content":"On 29 June 2021, GitHub took the world by surprise by launching their “AI Pair Programmer” Github Copilot, a tool that helps out with any kind of coding as it is written. Backed by Microsoft, this collaboration between GitHub and OpenAI relies on Codex, an AI engine built similar to GPT-3. While still in its infancy, the Copilot is compatible with a plethora of programming languages like Ruby, JavaScript, Python and TypeScript. In addition, the AI tool showcases its versatility by being available as an extension for Visual Studio Code, JetBrains and Neovim. The anatomy of Copilot GitHub Copilot is designed along the lines of its predecessor Natural Language Semantic Code Search, which suggested code examples based on simple descriptions in the English language. The Copilot, however, has the capability of generating entire multi-line programmes as well as documentation and test runs based on the context given by the user. Mark Pesce, an award-winning columnist at Cosmos Magazine and a renowned name in the world of AI, reveals his experience in testing the newly launched GitHub Copilot and quoted: “I wrote a single line of comment that stated my requirements and – abracadabra! – the computer provided a complete solution.“ Now, let us understand the workings of Copilot through a simple example: If I want to assign a certain text to a file name, all that I have to do is type: ‘write text file to fname’. The AI reads the syntax of the sentence and comes up with a suggestive line of code in grey that meets my requirement. One press of the tab button and the suggestive code is accepted and inserted into the body. Once the suggested code is accepted, the AI then builds on it and provides a variety of options with respect to the context of the test function. With the launch of GitHub Copilot, programmers breathe a sigh of relief as the tool helps them furnish the code that they want to create with exceptional speed. Can the same AI tool help coders who look to create pieces of art? AI on the digital canvas Jeff (@ippsketch), the director of R&D in a medical device company, who has a penchant for using code to create masterpieces, showcases his work with Copilot on Twitter as he talks about filling the digital canvas with colour with minimal code. Although a bit rudimentary, Jeff demonstrates the potential of GitHub Copilot to land some interesting brush strokes in the form of a high-resolution image. Dylan Beattie, an international keynote speaker and director of Ursatile, explains in detail the art of code along with some examples. GitHub Copilot: Here to change the game Github Copilot is making a buzz with its receptive user interface and making headlines in the world of gaming. Coders are seen using the AI tool to create primitive games without writing a single line of code! ThePrimeagen, a digital personality with Twitch, YouTube streams as well as name in GitHub, Linecode, Instagram and Twitter, describes his experience in GitHub Copilot writing the entire code of a game from scratch! Primeagen was so impressed that he quoted: “Yes the dang AI MAKES  A GAME!!! I could not believe it and the ending just blew me away.” Github Copilot is a real smooth talker! The most tricky form of art is the art of socialising, and the GitHub Copilot seems to have that trick taken care of. Ido, a full stack developer from DagsHub, talks about his conversation with Github Copilot, where he spoke English instead of code, and the result was beyond his imagination! Ido was impressed that the AI tool could do non-code related tasks as well. Ido further tested its ingenuity by opening a blank file and placing a simple question in the English language. While the AI tool does not respond to a simple sentence, given the proper context, it starts talking back! GitHub Copilot is yet far from perfect, and studies suggest that there are areas of vulnerability when it suggests code. However, with a little fine-tuning and training of the OpenAI codex, the day is not far when this amazing AI tool can create masterpieces in a jiffy.","excerpt":"GitHub Copilot is designed along the lines of its predecessor Natural Language Semantic Code Search.","categories":["Global Tech"],"tags":["ai System","AI Tool","Javascript","language model","OpenAI Codex","Python","Ruby","Stack Overflow","TypeScript","Visual Studio Code"],"author_name":"Kartik Wali","publish_date":"2022-01-19T14:00:00","publication_year":"2022","word_count":697,"keywords":["OpenAI","AI","Javascript","JavaScript","ai System","Stack Overflow","TypeScript","language model","Python","Ruby","Git","GPT","OpenAI Codex","AI Tool","GitHub","R","Java","Visual Studio Code"],"extracted_tech_keywords":["AI","OpenAI","Python","R","JavaScript","TypeScript","Java","Git","GitHub","GPT"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/can-github-copilot-create-art\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":481,"title":"Top 10 Astronomy Institutes in India Fostering Space Education","content":"An education in space and astronomy is not the most professional degree that one can acquire. Most of it can be attributed to an individual’s own desire to learn about this field. Astronomy, sometimes considered as a sub-field of Physics is further divided into several branches such as Astrophysics, Astrometeorology, Astrobiology, Astrogeology, Astrometry, Cosmology etc. GoI mainly supports the education sector around astronomy in India and there is rarely any private endeavor in this space. Yet these 10 institutes in India have greatly contributed to space education on the formal level. 1. Aryabhatta Research Institute of Observational Sciences, Nainital Aryabhatta Research Institute of Observational Sciences (ARIES) is a leading research institute in Nainital, Uttarakhand which specializes in Astronomy, Astrophysics and Atmospheric Sciences. An autonomous body under the Department of Science and Technology (DST), Government of India, the institute is situated at Manora Peak (1,951 m (6,401 ft)), about 9 km from Nainital, a popular hill station. The astronomical observatory is open to the public during working days on afternoons. For night viewing however, three – four days on moonlight nights are fixed and prior permission is needed. 2. Indian Institute of Astrophysics, Bangalore The Indian Institute of Astrophysics (IIA), with its headquarters in Bangalore (Karnataka state), is a premier National Research Institute of India. IIA conducts research primarily in the areas of astronomy, astrophysics and related subjects. It is widely recognized as one of the best research institutes in the country. The institute has a network of laboratories and observatories in India, including Kodaikanal (the Kodaikanal Solar Observatory), Kavalur (the Vainu Bappu Observatory), Gauribidanur, Hanle (the Indian Astronomical Observatory) and Hosakote. It contributed heavily in Astrosat, India’s first dedicated multi-wavelength space observatory. 3. Indian Institute of Science Education and Research The Government of India, based on the recommendation of Scientific Advisory Council to the Prime Minister, through the Ministry of Human Resource Development (MHRD), established five Indian Institutes of Science Education and Research (IISER) (Hindi: भारतीय विज्ञान शिक्षण एवं अनुसंधान संस्थान) in 2006. These institutes are currently located in Kolkata, Pune, Mohali, Bhopal and Thiruvananthapuram. Each IISER is an autonomous institution awarding its own Masters and Doctoral degrees. Students are encouraged to carry out research projects during the vacation periods in the first four years of their Masters programme in various research institutes outside IISER. 4.Indian Institute of Science, (Bangalore) Indian Institute of Science (IISc) is a public university for scientific research and higher education located in Bangalore, India. Established in 1909 with active support from Jamsetji Tata and H.H. Sir Krishnaraja Wodeyar IV, the Maharaja of Mysore. It is also locally known as the “Tata Institute”. It acquired the status of a Deemed University in 1958. IISc is widely regarded as India’s best university. It has been ranked at number 11 and 18 worldwide (and ranked 3rd and 6th in Asia) when considering the criteria of Citations per Faculty in 2014 and 2015 respectively. IISc was the first Indian institute to feature on Times Higher Education World University Rankings for engineering and technology in the year 2015-16 at 99th position. IISc has been ranked number 1 and 4 in the BRICS and Asian region respectively while considering the criteria of Papers per Faculty in 2015. 5. Indian Institute of Space Science and Technology, Thiruvananthapuram The Indian Institute of Space Science and Technology (IIST) is a government-aided institute and deemed university for the study and research of space science, located at Valiamala, Thiruvananthapuram, Kerala. It is the first university in Asia to be solely dedicated to the study and research of Outer space. It was inaugurated on 14 September 2007 by G. Madhavan Nair, the then Chairman of ISRO. IIST was set up by the Indian Space Research Organisation (ISRO) under the Department of Space, Government of India.P. J. Abdul Kalam, former President of India, was the Chancellor of IIST. IIST offers regular engineering undergraduate, postgraduate and doctorate programmes with focus on space science, technology and applications. 6. Inter-University Centre for Astronomy and Astrophysics, (Pune) The Inter-University Centre for Astronomy and Astrophysics (IUCAA) is an autonomous institution set up by the University Grants Commission to promote nucleation and growth of active groups in astronomy and astrophysics in Indian universities. IUCAA is located in the University of Pune campus next to the National Centre for Radio Astrophysics, which operates the Giant Metrewave Radio Telescope. 7. National Centre for Radio Astronomy, Pune The National Centre for Radio Astrophysics (NCRA) of India is a premier research institution in India in the field of radio astronomy is located in the Pune University Campus(just beside IUCAA), is part of the Tata Institute of Fundamental Research, Mumbai, India. NCRA has an active research program in many areas of Astronomy and Astrophysics, which includes studies of the Sun, Interplanetary scintillations, pulsars, the Interstellar medium, Active galaxies and cosmology and particularly in the specialized field of Radio Astronomy and Radio instrumentation. NCRA also provides exciting opportunities and challenges in engineering fields such as Analog and Digital Electronics, Signal Processing, Antenna Design, Communication and Software Development. 8. Physical Research Laboratory, Ahmedabad The Physical Research Laboratory (PRL) is a National Research Institute for space and allied sciences, supported mainly by Department of Space, Government of India. This research laboratory has ongoing research programmes in astronomy and astrophysics, atmospheric sciences and aeronomy, Earth sciences, Solar System studies and theoretical physics. It manages the Udaipur Solar Observatory and is located in Ahmedabad. Known as the cradle of space sciences in India, the Physical Research Laboratory (PRL) was founded on 11 November 1947 by Dr. Vikram Sarabhai. PRL had a modest beginning at his residence, the RETREAT, with research on cosmic rays. The institute was formally established at the M.G. Science Institute, Ahmedabad, with support from the Karmkshetra Educational Foundation and the Ahmedabad Education Society. Prof. Kalpathi Ramakrishna Ramanathan was the first Director of the institute. The initial focus was research on cosmic rays and the properties of the upper atmosphere. Research areas were expanded to include theoretical physics and radio physics later with grants from the Atomic Energy Commission. 9. Radio Astronomy Center, Ooty The Radio Astronomy Centre (RAC) is part of the National Centre for Radio Astrophysics (NCRA) of the well-known Tata Institute of Fundamental Research (TIFR) which is funded by the Government of India through the Department of Atomic Energy. The RAC is situated near Udhagamandalam (Ooty) in the beautiful surroundings of the Nilgiri Hills and it provides stimulating environment for the front-line research in radio astronomy and astrophysics with its excellent and highly qualified staff and international reputations. 10. Raman Research Institute, Bangalore Raman Research Institute (RRI) is an institute of scientific research located in Bangalore, India. It was founded by Nobel laureate C. V. Raman. Although it began as an institute privately owned by Sir C. V. Raman, it is now funded by the government of India.","excerpt":"An education in space and astronomy is not the most professional degree that one can acquire. Most of it can be attributed to an individual’s own desire to learn about this field. Astronomy, sometimes considered as a sub-field of Physics is further divided into several branches such as Astrophysics, Astrometeorology, Astrobiology, Astrogeology, Astrometry, Cosmology etc. […]","categories":["AI Trends"],"tags":["institutes"],"author_name":"AIM Media House","publish_date":"2017-02-05T06:44:17","publication_year":"2017","word_count":1139,"keywords":["Go","institutes","programming_languages:R","AI","programming_languages:Go","Git","RAG","Ray","GAN","R"],"extracted_tech_keywords":["AI","Ray","RAG","R","Go","Git","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/10-astronomy-related-institutes-in-india-that-are-fostering-space-education\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":6489,"title":"Big Data startup Aureus Analytics raises $850K in angel funding","content":"Singapore- and India-based Big Data firm Aureus Analytics has raised its first round of funding of US$850,000. The funding round was led by Steven Sule, Partner, Naste Advisors. It also saw participation from marquee angel investors such as Rajan Anandan, MD, Google India; Arihant Patni, Co-founder, Hive Data; and Arun Venkatachalam of Murugappa Group. The funds will be used to enhance product development, and further expand the engineering and business teams. “We believe that the impact value of predictive and Big Data analytics is significantly higher if made available at the point of decision. Our aim to make this possible and bring the power into hands of the business users in an easy and seamless manner over the cloud. By our estimates, the global market opportunity for such solutions in the insurance industry alone is approximately US$ 9 billion to US$10 billion,” said Anurag Shah, CEO, Aureus Analytics. “We could not have asked for a better set of investors to back us at this early stage of our journey. This has reaffirmed our faith that what we are building is going to be a game-changer in the near future,” he added. Aureus Analytics has a suite of industry-specific predictive analytics products and Big Data-ready platform called ASAP. These products enable customers to tackle business critical issues by leveraging massive internal and external data sets at the point of decision-making, shared an official release. “Aureus’ focus on the insurance space allows it to be at the forefront of the industry, which will increase its utilisation of analytics tools to better understand its products and customers in the future,” said Sule. The startup was founded in 2013 by Shah, an Indian School of Business alumnus, along with Mumbai University alumnus Ashish Tanna, and Shivaji University alumnus Nitin Purohit. Aureus offers predictive analytics and Big Data-ready platforms to insurance companies and banks. The company has customers including Bharti AXA Life Insurance, Aegon Religare Life Insurance, General Insurance Council of India and a leading Indian bank. “We have engaged Aureus Analytics to help us understand our customer behaviour better and generate meaningful insights,” said Saujnaya Shrivastava, Chief Marketing Officer, Bharti AXA Life Insurance.","excerpt":"Singapore- and India-based Big Data firm Aureus Analytics has raised its first round of funding of US$850,000. The funding round was led by Steven Sule, Partner, Naste Advisors. It also saw participation from marquee angel investors such as Rajan Anandan, MD, Google India; Arihant Patni, Co-founder, Hive Data; and Arun Venkatachalam of Murugappa Group. The […]","categories":["AI News"],"tags":[],"author_name":"AIM Media House","publish_date":"2014-11-26T15:12:56","publication_year":"2014","word_count":359,"keywords":["big data","Go","AI","R","ML","RAG","Aim","analytics","predictive analytics","startup"],"extracted_tech_keywords":["AI","ML","analytics","Aim","RAG","predictive analytics","R","Go","big data","startup"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/big-data-startup-aureus-analytics-raises-850k-angel-funding\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10051256,"title":"IBM Unveils AI-Based Solution For Companies To Respond To Climate Risks","content":"IBM has announced a suite of environmental intelligence software that leverages AI to help organisations prepare for and respond to weather and climate risks that may disrupt business, more easily assess their own impact on the planet, and reduce the complexity of regulatory compliance and reporting. “Today’s announcement further accelerates our commitment to apply scientific methods to find new solutions for the most pressing challenges that we face as a society. Last year, IBM Research launched a global initiative called the “Future of Climate” to develop and demonstrate innovations focused on addressing climate change by leveraging AI expertise across IBM Research,” said Gargi Dasgupta – Director, IBM Research India and CTO IBM India\/South Asia. “To support this initiative, IBM Research – India, which has been at the forefront of advancing AI and its applications, has played a critical role in building the new Carbon Performance Engine and Climate aware applications including forecasting, which are now a part of the overall Environmental Intelligence Suite,” she added. The suite leverages existing weather data from IBM, advanced geospatial analytics already in use by companies worldwide, and new innovations from IBM Research. The offering is the first to bring together artificial intelligence, weather data, climate risk analytics, and carbon accounting capabilities in this way – allowing organisations to spend fewer resources curating this complex data and more on analysing it for insights and taking action to improve their operations. The newly introduced IBM Environmental Intelligence Suite is a SaaS solution designed to help organisations: Monitor for disruptive environmental conditions such as severe weather, wildfires, flooding and air quality, and send alerts when detected;Predict potential impacts of climate change and weather across the business using climate risk analytics;Gain insights into potential operational disruptions and prioritise mitigation and response efforts;Measure and report on environmental initiatives and operationalise carbon accounting while reducing the burden of this reporting on procurement and operations teams. The suite delivers environmental insights via APIs, dashboards, maps and alerts that can help companies address both immediate operational challenges as well as longer-term planning and strategies. The Environmental Intelligence Suite can be integrated with IBM’s broader software portfolio for additional efficiencies across business operations – including IBM Maximo Application Suite to help companies protect and extend the life cycle of their critical assets and IBM Supply Chain Intelligence Suite to help build more sustainable and resilient supply chains.","excerpt":"The offering is the first to bring together artificial intelligence, weather data, climate risk analytics, and carbon accounting capabilities on a single platform.","categories":["AI News"],"tags":["AI Companies","IBM"],"author_name":"kumar Gandharv","publish_date":"2021-10-12T13:26:15","publication_year":"2021","word_count":394,"keywords":["API","artificial intelligence","programming_languages:R","AI","innovation","RAG","disruption","analytics","IBM","AI Companies","GAN","R"],"extracted_tech_keywords":["AI","artificial intelligence","analytics","RAG","R","API","GAN","innovation","disruption","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/ibm-unveils-ai-based-solution-for-companies-to-respond-to-climate-risks\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10088741,"title":"USA has a UPI Problem","content":"UPI enables 2,348 transactions every second. The total UPI transaction value accounted for nearly 86% of India’s GDP in FY22 with about INR 125.95 lakh crores transacted through it. The UPI payment has revolutionised the payment system in India, something that even the US is far from achieving. Interestingly, the country has an extremely unique case study in the payment system. When it comes to the list of countries with the most digital transactions, the world superpower ranks at an embarrassing 7th position—even lower than Nigeria. USA—a country which produced successful big tech companies like Apple, Google, Facebook, and Microsoft and which also played a big part in the Internet revolution—is failing at the banking revolution. Saraswathi Ramachandra, the director of Citi, told AIM that, “USA already has a card-based network for payment systems—built by VISA\/Mastercard. Everyone has at least one debit\/credit card and is widely used everywhere”. UPI Problem in USA Card companies like these are perhaps the main reason why the country can’t have a system like UPI. Lobbying in the US is legal and they often lobby against any practices that might impact their revenue. The card lobby is so strong in the US that when Indonesia was launching its new domestic payment network, US trade officials convinced Indonesia to loosen the rules at the request of card networks like Mastercard and Visa. In India too, when the government was planning to restrict Visa and Mastercard, the card companies reached out to the US Trade Representative (USTR) raising concerns about a “level playing field”. USTR then had to issue a memo saying, “Visa remains concerned about India’s informal and formal policies that appear to favour the business of National Payments Corporation of India (NPCI) over other domestic and foreign electronic payments companies”. Ramachandran too, believes that “the US federal government doesn’t want to compete with the private sector by creating an alternative payment system”. Lobbying in the US is so widespread that in 2019 alone, lobbyists spent $3.47 billion on influencing political policy, representing the highest sum spent on lobbying since lobbying spend peaked in 2010. The US conventionally thrives off of its private sector but whenever a new idea is able to derive profits from legacy companies, they try to nip it in the bud. For instance, when Bitcoin was gaining popularity in the country, legacy banks like the Bank of America and JPMorgan called cryptocurrencies a threat. At the world economic forum in 2018, Jamie Dimon, the chief executive of JPMorgan Chase, called Bitcoin a “terrible” store of value. Likewise, instant systems like UPI don’t make money for the banks in the US, or any banks for that matter. UPI, so far in India, has been unprofitable for banks as well as instant payment applications like Google Pay, Paytm, PhonePe and others. “USA has other instant payment systems such as PayPal or chase-pay, Apple Pay and others,” says Ramachandran, “and while these don’t have instant payment systems like UPI, the market has been built already and there’s inertia to change”. UPI’s Conundrum: To Charge or Not to Charge from Users Apple Pay in the US, for instance, is just a credit card but on mobile and the customers are charged per transaction on the system. To shift from such a system to one like in India—where customers are charged zero rupees per transaction and instant payment applications are running in losses—is an unfavourable scenario for the first world banks. “Banks are ready as long as there’s money in it!” says Ramachandran, “They make huge profits via charges for payment services via MDR which is currently paid by merchants and as long as this is maintained, banks will be willing”. Ramachandran further asserts that banks in the US are not ready for a free instant payment system like UPI. Additionally, it’s not like the US hasn’t attempted to build an instant payment system. An example of such a system would be ‘FedNow’. Announced in 2020, the system is increasingly basic and can be compared to the RTGS system in India which was launched way back in 2004. Moreover, while UPI insists banks do not charge a transaction fee, FedNow encourages banks to charge transaction fees from their customers. USA is financially inclusive, India wasn’t Another significant reason behind the success of UPI in India was lack of adoption of credit\/debit cards. Around 2014–15, India had a mere 21 million credit cards in the population of more than a billion. In 2011, only 35% of the Indians had bank accounts, opposed to 80% in 2017. In India, where there was no digital or card-based payment system in place, it was a necessity to build a system like UPI that enabled fast, reliable and hassle-free digital transactions. Such is not the case in the US, where most of the population has access to debit\/credit cards, thereby making it much harder to adapt to a new payment system.","excerpt":"Ramachandran, believes that “the US federal government doesn’t want to compete with the private sector by creating an alternative payment system.”","categories":["AI Features"],"tags":["Mastercard","npci","UPI","USA","Visa"],"author_name":"Lokesh Choudhary","publish_date":"2023-03-06T14:35:08","publication_year":"2023","word_count":819,"keywords":["Go","Mastercard","npci","AI","programming_languages:R","Visa","programming_languages:Go","Git","UPI","RAG","Aim","GAN","R","USA"],"extracted_tech_keywords":["AI","Aim","RAG","R","Go","Git","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/usa-has-a-upi-problem\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10131883,"title":"Generative AI to Drive Future Cybersecurity Threats","content":"While AI has been leveraged in several areas of expertise, one of the more alarming ones is in the field of cybersecurity, specifically, on the threat-actor side. With generative AI becoming easily available to the average user, a recent report published by Verizon Business predicts that threat actors can start leveraging AI as soon as this year. “Threat actors of all types, from the most sophisticated nation-states and highly resourced criminal syndicates to solo cybercriminals, will eventually adopt AI,” the report stated. This isn’t just the use of deepfakes or other such antagonistic general use cases, but the use of AI to take down companies, access confidential information, or just ease up the process of running scams on a large scale. “They’ll go for the least sophisticated, lowest-effort applications first. Generating more phishing messages? Pretty simple. Translating social engineering attempts into multiple languages? Done,” the report further stated. Speaking to AIM, Anshuman Sharma, who heads cybersecurity consulting services under the Verizon Threat Research Advisory Center (VTRAC), said that just because there isn’t a lot of discussion surrounding generative AI’s use amongst threat actors doesn’t mean that it’s non-existent. “It isn’t discussed on the dark web or the clear web by the threat actors, but we already know that there are tools, like WormGPT and FraudGPT, available. “Obviously, the organisation has to be prepared because three to five years down the line, we’ll move towards strong AI, which can learn things that weak AI can do, but it can use that learning in a different context altogether. It’s going to happen very soon,” Sharma said, elaborating on the research so far. This, he emphasised, was a major reason why businesses need to start leveraging AI. Thanks to AI, being left behind is no longer an option, with cybersecurity becoming a more serious concern. Meanwhile, when it comes to the security of AI systems being integrated, Sharma said that it was a relatively new area of concern. The Weakest Security Link – AI or Humans? Recently, AIM spoke to a few AI jailbreakers who position themselves as ethical hackers. These hackers focus on jailbreaking AI systems specifically to highlight flaws in the systems of both proprietary and open-source models. However, while these are largely community-based initiatives for the betterment of generative AI, this may not be the case for much longer. “We just witnessed with CrowdStrike how one glitch can affect the entire world. It was reported that we were thrown into the Stone Age for a few hours, during which handwritten boarding cards were given. That’s the repercussion that it can have,” he explained. Sharma elaborated that while it wasn’t directly caused by AI, it was a case of over-reliance on a third party tool that was given too much internal access, something that organisations integrating AI are doing as well. “Right now, there is obviously a dependency on the providers and faith that the AI or machine learning products they deliver are helping us fight the battle against the bad guys using AI,” he said. But, it’s still something to take note of, considering these are third-party tools accessing large amounts of confidential data. Already, while it isn’t public news as yet, Sharma said there have been cases of models being jailbroken for this very purpose. “Those models were leveraged to bypass guardrails, using loaded language to nudge the LLMs to provide responses, which ideally, they shouldn’t provide,” he told AIM. The reliance on LLMs and GenAI, however, is not unfounded and inevitable. As is often the case, organisational implementation of GenAI is directly to improve efficiency within a company. What could take hours for a human to do can be done in a few minutes by AI. “AI is not going to take your job, but a person with the knowledge of AI is definitely going to take your job. Whether we want it or not, we have to leverage these models because we may be bound by rules, but threat actors are not,” he said. However, while this could pose a major problem for organisations, a larger problem looms ahead – human error. The reason why GenAI could be leveraged against a company is largely reliant on how the employees within that company use them. As Sharma mentioned before, models were used to bypass guardrails, but the cause of this was an employee feeding ChatGPT with huge amounts of sensitive data that the model learnt from. Additionally, as per Verizon’s 2024 Data Breach Investigation Report (DBIR), a significant amount of cybersecurity breaches were caused due to human involvement, with as many as 68% of incidents attributed to non-malicious human actions. “These range from errors and privilege misuse to stolen credentials and social engineering, showing the persistent risk that human factors pose. Shockingly, breaches resulting from exploitation of these vulnerabilities have surged by 180% — almost triple last year’s rate,” the report stated. Likewise, Sharma stated that phishing still posed a huge problem, and a lack of education meant that companies were more likely to be at risk thanks to humans rather than AI. “Phishing remains a pervasive threat, occurring frequently and indiscriminately across all sectors. And AI-based phishing is even creating more issues for the good guys to manage. On the prevention side, it’s the one that always has to be human, which naturally becomes the weakest element in the chain,” he said. Nevertheless, there are ways to curb this, both on the AI and human side. Data, Control and Awareness That’s the formula to prevent these attacks and potential attempts to jailbreak. Sharma said that organisations potentially at risk of cyber attacks were largely dependent on how good the controls built into their datasets were, and how they were used for training purposes, in order to restrict the LLMs used. “Technical controls, implementing it by testing it again on the jailbreaking side, and making the users aware – that’s the key to how we make it,” he said. https:\/\/twitter.com\/anshublog\/status\/1659074808091676672 As previously reported by AIM, pentesting companies, including Verizon, have started taking up the testing of AI systems integrated into companies. Additionally, companies have also begun implementing zero-trust policies when it comes to generative AI usages. Opportunities Galore Similarly, cybersecurity startups have begun sprouting, taking advantage of the gap, with the global AI in the cybersecurity market expected to be valued at $60.6 billion in 2028, growing from $22.4 billion last year. Seemingly echoing this, larger companies like Google are offering specific programmes catered to AI cybersecurity startups. Additionally, with the niche in the market, several of these security startups, like Sydelabs, Oxeye, Helios, and YC-funded Cyberfend, have been acquired. Interestingly, Sydelabs has worked on an AI firewall, preventing attempts at jailbreaking organisational AI systems. Similarly, LLM vaults are another method through which this issue is being addressed by startups like BoxyHQ and Skyflow. The need to integrate AI for organisations is vital, whether it’s to improve overall efficiency or fight off threat actors leveraging the same. Like Sharma warned, this could become a threat area that gains increasing significance in the next decade. However, with the increase in funding in the area, as well as startups taking the lead in bridging the gap, the fallout may not be as drastic as is expected.","excerpt":"With generative AI readily available to the average user, a recent report published by Verizon Business predicts that threat actors can start leveraging AI as soon as this year.","categories":["Deep Tech"],"tags":["verizon"],"author_name":"Donna Eva","publish_date":"2024-08-08T15:09:47","publication_year":"2024","word_count":1206,"keywords":["Go","ChatGPT","GenAI","machine learning","AWS","AI","RAG","Aim","generative AI","verizon","R"],"extracted_tech_keywords":["AI","machine learning","generative AI","GenAI","ChatGPT","Aim","RAG","AWS","R","Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/generative-ai-to-drive-future-cybersecurity-threats\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10506,"title":"Interview &#8211; Roger Schank, Leading Artificial Intelligence Theorist","content":"Roger Schank, the name is synonymous with Artificial Intelligence. Dr. Schank is one of the world’s leading researchers in the space of Artificial Intelligence. He has spent nearly 50 years working in the field of artificial intelligence and his contributions to this field have been immense. An artificial intelligence theorist, a cognitive psychologist, a scientist, an educational reformer, and an entrepreneur is what portrays Dr. Schank. He is a fellow of the AAAI, the founder of the Cognitive Science Society, and co-founder of the Journal of Cognitive Science. Beginning in the late 1960s, he pioneered conceptual dependency theory and case-based reasoning, both of which challenged cognitivist views of memory and reasoning. Analytics India Magazine (AIM) recently did an Interview with Roger Schank to understand his views on Artificial Intelligence. Here is the excerpt from the interview. [dropcap size=”2″]AIM[\/dropcap]Analytics India Magazine: Tell us something about your current work and your achievements in the AI space [dropcap size=”2″]RS[\/dropcap]Roger Schank: I have been in the field of AI since 1965. Over these years I have watched AI go through various phases. Today all kinds of things that have nothing to do the modelling of human intelligence are called AI. I like to give an example to my audiences. I ask my audience to name 50 states of US. Most Americans would say they know the 50 states. I ask them to write the names down and the chance is usually 1 out of 100 can name all 50 states. This shows us something about how human memory works. People have to visualize a map and go down the east coast and up the west coast in order to list all states. They know the names of all them, but there is a big difference between recognition and recall memory. Modern day “intelligent” machines don’t worry about that distinction. If you ask any AI machine it will easily list all 50 states (in alphabetical order) because their algorithms are all about words and their memories do not try to recall or visualize or imagine the way people do. People do not memorize data and then look that data up when asked. AI, to me at least, is about modelling human intelligence, not effective data processing. AIM: Though there is rise in AI, we are yet to reach a stage where a truly intelligent machine exists. How far are we from that? RS: AI has lot of subfields under it. For instance: face recognition. Facebook uses face recognition to recognize your picture. This is an example of Artificial Intelligence. Robotics is another area in AI which is growing. But the part of AI that interests me is building a machine that behaved very similarly to humans, i.e. one that engage in a meaningful conversation. For instance, if I am travelling to Denmark, I will think about what my last visit there was like. I will go down memory lane, think about what food I ate, who I met, what I liked, and then, based on my past experiences make decisions for my current trip. Humans use past experiences to make decisions about the present. But machines cannot look up for past experiences because they haven’t had any and because AI people have not studied well enough how prior memories are indexed and retrieved. Modern AI’s are fed with data and they can only look up that data for answers. They do not have experiences or memories to think about or to draw inferences and new plans from. AIM: Your views on machine learning being used in AI RS: Machine Learning and Deep Learning are words that are very fancy but really about human learning. As humans we learn from experience. For instance, I just gave a speech in Spain and I learned lot from that experience, I learned from the interaction with the audience. “Deep Learners” can’t do that kind of learning. Machines can do lot of things but are the efforts going in the right direction? There has always been an argument in AI between those who want machines to do things fast and beat humans at games, and those who care about the replication of human abilities and processes. AIM: Do we see AI winter anytime soon now? RS: AI winter will come soon, maybe next year. The companies investing in AI now are all venture capitals and with high expectations about AI projects, and they will be disappointed for the most part. AIM: How do you see this era of AI as compared to the earlier times? [pullquote align=”right”]Real AI machines would come up with novel opinions and be prepared to discuss and debate. When we see that happening we will be seeing AI. [\/pullquote]RS: There were always people in AI who were concerned with building experiential memories that changed with each new experience. For instance, if I ask you which school is good to attend you will rely upon your past experience in order to make suggestions. Machines today could search college rankings. They are fed with set of data and rules and they search and give answer. Humans don’t search for answers. Either they have a point of view or they try to use their past experiences to come to some conclusions. As humans we imagine things and based on our imagination we make decisions. But today’s machines search text from the data they are fed and really have no idea what that data is even about.  Humans have opinions of their own. Real AI machines would come up with novel opinions and be prepared to discuss and debate. When we see that happening we will be seeing AI. Right now we are seeing search on massive amounts of text, which is something cannot do and wouldn’t try to do.","excerpt":"Roger Schank, the name is synonymous with Artificial Intelligence. Dr. Schank is one of the world’s leading researchers in the space of Artificial Intelligence. He has spent nearly 50 years working in the field of artificial intelligence and his contributions to this field have been immense. An artificial intelligence theorist, a cognitive psychologist, a scientist, […]","categories":["AI Features"],"tags":["AI India","Interviews and Discussions"],"author_name":"Дарья","publish_date":"2016-07-30T05:56:42","publication_year":"2016","word_count":960,"keywords":["Go","API","machine learning","artificial intelligence","AI","Ray","Aim","deep learning","analytics","AI India","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","deep learning","analytics","Aim","Ray","R","Go","API"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/interview-roger-schank-leading-artificial-intelligence-theorist\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":30495,"title":"Why Capstone Project Is The Backbone Of Data Science Education","content":"A recent survey conducted by Analytics India Magazine suggested that 92 percent of respondents believed that capstone projects played a very important role while choosing an upskilling course in analytics or data science. Most candidates gave a great deal of importance to courses which included capstone projects in their overall curriculum. Interestingly, respondents gave more preference to capstone projects over jobs, as only 57 percent respondents asserted that internship or placement opportunities are crucial in an upskilling programme. But why are capstone projects so important? In this article, we will discuss the attributes that make for a good capstone project. Why Is It Important As the name suggests, capstone projects are at the apex of an analytics training that is drafted to give an experience of practical applications of coursework to students. It is crucial for improving the overall student experience as it helps them enter the workforce with a practical knowledge of the tools and techniques learned during the course. Capstone projects are a key determinant for companies to pick the candidates who are well-suited for analytics roles at their organizations. Designed by industry experts in most institutes, capstone projects include a host of assignments and uses cases specific to industries. The candidates are then asked to solve them based on the theoretical knowledge they have acquired during the course. This might include the use of programming knowledge, analytical solving, tools, and others. They are drafted in a way that it gives students an opportunity to strengthen their academic knowledge while encouraging critical thinking and communication skills. What Should A Good Capstone Project Have? While the format of the capstone course varies significantly for different analytics programmes, the underlying purpose is the same — to give practical exposure. Most of the capstone projects are modeled on the basis of requirements by the data-driven companies. A good capstone project should be able to enhance teamwork and communication skills, ignite reasoning skills, providing a platform, develop research-related skills, keep updated with the latest developments in respected technologies to a candidate. Enhance Teamwork And Communication As a fresher or an ongoing graduate one is not exposed much to team projects and the importance of team communication. Capstone projects give exposure to these areas which are very crucial in one’s career. Team communication helps in understanding and accelerating productivity among team members. Productive communication also builds trust among team members and creates a reliable working environment. Develop Abilities To Approach A Problem The capstone projects will make a candidate to understand the right problem, understanding the fundamentals of the problem, articulation of the problem and focusing on the root cause, building simple solutions with effective outcomes. A capstone project allows to get through the root of the problem and helping candidates develop a knack for understanding and approaching the problem. The module also trains candidates to develop solutions in the more simple and effective way, thus inculcating a practice to approach problems. The Art Of Asking Questions Asking good questions is a crucial responsibility which helps people to understand the problems so that they can find the solutions. Capstone projects help aspirants envision the goals — short and long-term goals, the risks involved. It also helps in speculating deliverables for the project and set a priority for solving problems. Practical Use Of Skills And Tools Analytics is a field where major technical skills require training and experience to master. Most of the organizations in this field are looking for people who can jump right in on the first day of the work and start implementing their knowledge for the benefit of the company. These technical skills are difficult to understand and implement as they need a lot of guidance. Capstone projects help the aspirants to gain experience and understand their implementation. Outlook Capstone courses are definitely looked upon as a must to have in a CV while applying for practical based analytics and data science job roles and needs to be a part of every analytics programme and college curriculum. All the major training institutes such as Jigsaw Academy, Manipal, AnalytixLabs, Insofe and Great Learning, among others, have capstone projects ranging from a duration of one month to six months.","excerpt":"A recent survey conducted by Analytics India Magazine suggested that 92 percent of respondents believed that capstone projects played a very important role while choosing an upskilling course in analytics or data science. Most candidates gave a great deal of importance to courses which included capstone projects in their overall curriculum. Interestingly, respondents gave more […]","categories":["AI Features"],"tags":["become a data scientist","data science internships"],"author_name":"Bharat Adibhatla","publish_date":"2018-11-21T11:00:04","publication_year":"2018","word_count":697,"keywords":["data science","Go","AI","become a data scientist","data-driven","RAG","ViT","analytics","Rust","GAN","data science internships","R"],"extracted_tech_keywords":["AI","data science","analytics","RAG","R","Go","Rust","GAN","ViT","data-driven"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/why-capstone-project-is-the-backbone-of-data-science-education\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":27915,"title":"From GrooveNet To DanceNet: These Neural Networks Are Used To Generate Dance Moves","content":"We never thought we would see the day when a neural network generated dance moves for humans! But now, Indian developer Jaison Saji has developed a dance generator using Keras, inspired by Cary Huang’s project. One of the key takeaways is that it has generated a lot of attention for this project is the viral dancing-celebrity videos one can make with it. Clearly, there could be more intended applications with this project. DanceNet uses a variational autoencoder (VAE), which works with an works encoder, a decoder and a loss function to drum up thousands of single dance pose pictures, then sequentially connect them to produce vigorous dance movements through joint training on long short-term memory (LSTM) and mixture density networks (MDN). According to Synced Review, Saji built a DanceNet encoder model that comprises of three convolutional layers and one fully connected layer.In the next step, he reconstructed the vectors to the original images by integrating one fully connected layer and four convolutional layers with three upsampling layers in the DanceNet decoder model. After the VAE model was trained, the user can sample any latent variable z and feed it to the decoder, and the model will produce a new dance pose image. And as part of his last step, Saji used a combination of LSTM and MDN to produce dance images for choreography. The developer stacked three LSTM layers, followed by a “dropout” treatment to prevent overfitting. The results from LSTM are subsequently inputted into the fully connected layer and the MDN layer to produce a series of dance moves as final outputs. You can access the project here. Creating Real-time Dance Movement With Neural Networks It is not the first time neural networks are used to create real-time dance movement. GrooveNet, a generative system can synthesise dance movements for a given audio track in real-time. The GrooveNet paper published in 2017 indicated that the intended application for GrooveNet is a public interactive installation wherein the crowd can share their own music to interact with an avatar. The artificial neural networks used for this project were factored conditional restricted Boltzmann machines (FCRBM) and recurrent neural networks (RNN), trained on a small dataset of four synchronised music and motion capture recordings of dance movement. The initial results indicated it could be used to train the FCRBM on a small dataset to generate dance movements. The paper further indicated that the model cannot generalise to other music tracks beyond the training data it is fed. Another interesting project was Dance Dance Convolution published last year which was based on a rhythm-based video game Dance Dance Revolution. According to this research paper, the players carry out a series of steps on a dance platform in synchronisation with music as directed by on-screen stepcharts. This paper essentially introduced the tasks of learning to choreograph. For example, with a raw audio track, the models can produce a new stepchart and it entails two steps — when to place steps and which steps to select. For the step placement task, the researchers combined RNN and CNN to take in spectrograms of low-level audio features to predict steps, conditioned on chart difficulty. For the next step, the step selection processes, the researchers used an LSTM generative model which exponentially outperformed n-gram and fixed-window approaches. Through this neural network, researchers can create many different charts for the same song. Conclusion So is dance a study relevant to enterprise artificial intelligence? From learning choreography to an optimal dance neural networks are used to train and control virtual dancers. There are many possible future applications of neural networks for generating dance moves. It can be used to further understand the inherent symmetry of the body.","excerpt":"We never thought we would see the day when a neural network generated dance moves for humans! But now, Indian developer Jaison Saji has developed a dance generator using Keras, inspired by Cary Huang’s project. One of the key takeaways is that it has generated a lot of attention for this project is the viral […]","categories":["AI Features"],"tags":[],"author_name":"Richa Bhatia","publish_date":"2018-09-03T10:08:46","publication_year":"2018","word_count":615,"keywords":["Go","TPU","artificial intelligence","Keras","AI","neural network","VAE","RNN","CNN","R"],"extracted_tech_keywords":["AI","artificial intelligence","neural network","Keras","TPU","R","Go","VAE","CNN","RNN"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/from-groovenet-to-dancenet-these-neural-networks-are-used-to-generate-dance-moves\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10113046,"title":"Why is Indian IT Investing in Semiconductors?","content":"Infosys recently announced its intention to acquire InSemi to strengthen its engineering and R&D capabilities. InSemi, a semiconductor design company brings its expertise in electronic design, platform design, and automation to the table. Last year after multiple misses, Foxconn partnered with HCL Group to build an outsourced assembly and testing unit (OSAT) in India. L&T, on the other hand, is not seeking partnerships but investing INR 830 crore to set up a wholly owned subsidiary that will be engaged in the business of fabless semiconductor chip design and product ownership. “Partnering with and acquiring semiconductor firms provides IT companies with the necessary talent and customer access, establishing them as full-stack service providers essential for capturing the growing semiconductor demand,” remarked Pareekh Jain, CEO of EIIR (Engineering IoT R&D) Trend, in an interview with AIM. Strategic moves in the industry In the case of Infosys and InSemi, the partnership is expected to significantly bolster Infosys’ engineering R&D capabilities. InSemi is a mammoth in platform design, automation, embedded and software technologies. InSemi’s expertise will also get Infosys to enhance its chip-to-cloud strategy, providing niche design skills at scale, aligning with Infosys’ heavy investments in AI\/automation platforms and industry partnerships. This acquisition is expected to enable Infosys to offer comprehensive end-to-end product development for its clients, addressing the increasing demand for advanced semiconductor design services integrated with embedded systems, explained Infosys chief Salil Parekh. TCS also partnered with Renesas Electronics Corporation to open an Innovation Center in Bengaluru and Hyderabad in March 2023. Like Infosys, its focus is to create semiconductor designs and software solutions for various sectors including IoT, infrastructure, industrial, and automotive segments. They plan to combine their expertise in IoT, manufacturing, telecom, automotive industries and advanced semiconductor designs, and embedded software support to push forward with new semiconductor designs. Wipro has also been working towards building an alliance with semiconductors for a while. The company acquired Eximius Design in 2020 to improve their VLSI and systems design capabilities. Eximius brought its expertise in semiconductor, software, and systems design to equip Wipro with comprehensive solutions for IoT, Industry 4.0, edge computing, cloud, and 5G applications. In 2022, Wipro also joined the Intel Foundry Services’ Accelerator Alliance, aiming to expedite the chip design cycle. This partnership was focused on making both complete chips (SoC) and specialised chips (ASIC) for specific uses. Diversification, from design to manufacturing HCL Group, on the other hand, is diversifying and foraying into semiconductor manufacturing. In the second attempt by Foxconn, the companies are in talks with Tamil Nadu and Telangana to set up its recently announced semiconductor assembly and testing unit. “Foxconn and HCL are partnering to establish OSAT operations in India to create a local industry ecosystem of supply chain stability. Foxconn plans to implement its BOL (build-operate-localise) model to benefit local communities through this investment,” a spokesperson for Foxconn said. The joint venture will be established with $37.2 million investment from Foxconn, securing a 40% stake. “Indian IT giants have strategically acquired semiconductor engineering firms to bolster their capabilities in this high-growth sector, showing a clear intent to capture the vast potential market for semiconductor outsourcing,” Jain explained. For HCL Group, the OSAT facility creates infrastructure for chip packaging and testing within India, a crucial step towards self-sufficiency in the semiconductor industry. Government support and the future Indian IT companies’ strategic emphasis on semiconductors is also propelled by government initiatives. The Indian government has approved significant financial outlays, including a INR 76,000 crore PLI scheme, to bolster the semiconductor sector and making entry into the segment a lucrative affair. Larsen & Toubro, a rather new player, made a strategic entry into the semiconductor industry by focusing on fabless chip design rather than manufacturing. R Shankar Raman, L&T’s CFO said at the earnings call last year, “We are mostly focused on designing automobile and industrial chips. At the moment it will be less investment and low manufacturing, but getting our positioning accessed for the design.” Each company’s approach differs from the other. While Infosys and Wipro focus on design and development, TCS emphasises innovation through partnerships, HCL on investment in manufacturing capabilities, and L&T on fabless models. The companies in India are working independently in the semiconductor industry, with each one contributing to a different part of the production chain, helping the industry grow as a whole. In conclusion, the IT sector aims to build semiconductor capabilities in-house, and end-to-end. While L&T are focusing only on chip design, HCL Group is setting up manufacturing facilities. This diversified approach not only aims to attract investment but also seeks to upskill the workforce to meet the industry’s growing demands, as seen through initiatives by HCLTech along with Electronics Sector Skills Council of India (ESSCI)​​​​. India’s semiconductor sector is set to hit $271.9 billion by 2032, growing at a 25.7% CAGR from 2022. Driven to self sufficiency, Indian IT is pushing hard to fulfil this prediction.","excerpt":"Instead of outsourcing chips, the IT sector aims to design, develop and manufacture them.","categories":["AI Features"],"tags":["AI in semiconductors"],"author_name":"K L Krithika","publish_date":"2024-02-16T11:00:00","publication_year":"2024","word_count":817,"keywords":["Go","programming_languages:R","AI","innovation","AI in semiconductors","GAN","automation","Ray","Aim","edge computing","R"],"extracted_tech_keywords":["AI","Aim","Ray","edge computing","R","Go","GAN","automation","innovation","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/why-is-indian-it-investing-on-semiconductors\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10006110,"title":"The Pricing Plans For OpenAI’s API Is Out &#038; It Is Not Cheap","content":"OpenAI recently revealed the projected pricing plans for its API that allows users the access to company’s much hyped AI tools such as GPT-3, which has been making headlines since the company launched it. However, the pricing isn’t easy on pocket. While it was initially launched in a free, two-month beta, users will now have to choose between four different pricing plans from October. According to a post on Reddit, the plans are as follows: Explore: Free tier: 100K tokens or a three-month trial, whichever you use up first.Create: $100 per month for 2M tokens, plus 8 cents for every additional 1k tokens.Build: $400 per month for 10M tokens, plus 6 cents for every additional 1k tokens.Scale: Contact OpenAI for pricing. The post provided further information such as it will provide 2M tokens equal in terms of number of documents\/books\/etc., which is roughly equivalent to 3,000 pages of text. And that fine tuning is currently available for the Scale pricing tier. It stated that the costs are subject to change. OpenAI during the release in June had said that unlike most AI systems which are designed for one use-case, the API today provides a general-purpose “text in, text out” interface, allowing users to try it on virtually any English language task. Further it can follow tasks such as semantic search, summarisation, sentiment analysis, content generation, translation, and more. We recently covered an article on how restricted access to GPT-3 can impact its efficiency. While it is aimed to ensure safety against dangerous applications, it contradicts the company’s vision of sharing knowledge to the community. According to the company, the algorithm is too complicated for developers to work on.","excerpt":"OpenAI recently revealed the projected pricing plans for its API that allows users the access to company’s much hyped AI tools such as GPT-3, which has been making headlines since the company launched it. However, the pricing isn’t easy on pocket. While it was initially launched in a free, two-month beta, users will now have […]","categories":["AI News"],"tags":[],"author_name":"Srishti Deoras","publish_date":"2020-09-04T18:00:08","publication_year":"2020","word_count":278,"keywords":["Go","semantic search","API","OpenAI","sentiment analysis","AI","GPT","Aim","R","llm_models:GPT"],"extracted_tech_keywords":["AI","OpenAI","Aim","sentiment analysis","semantic search","R","Go","API","GPT","llm_models:GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/the-pricing-plans-for-openais-api-is-out-it-is-not-cheap\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10103468,"title":"Anthropic Launches Claude 2.1, Surpasses GPT-4 Turbo in Context Length","content":"Anthropic’s Claude just got a massive upgrade. Claude 2.1, the latest iteration of its AI language model, is now available through API, revolutionising the claude.ai chat experience. Bringing forth key enhancements for enterprises, Claude 2.1 introduces a remarkable 200K token context window, substantial reductions in model hallucination rates, and a beta feature called tool use. This update also includes a pricing overhaul aimed at improving cost efficiency for diverse customer base. Responding to user feedback, Claude 2.1 doubles the allowable token limit, now enabling a context window of 200,000 tokens, compared to 128,000 of GPT-4 Turbo announced at OpenAI DevDay. This corresponds to approximately 150,000 words or over 500 pages of material. Users can leverage this extended capacity to upload comprehensive documents such as codebases, financial statements, or lengthy literary works. Claude’s capabilities now extend to summarisation, Q&A, trend forecasting, document comparison, and more, all with the ability to process complex tasks in a matter of minutes. Addressing user demands, a new beta feature, tool use, has been introduced, allowing Claude to seamlessly integrate with existing processes, products, and APIs. This expanded interoperability enhances Claude’s utility in day-to-day operations, enabling it to orchestrate across developer-defined functions, search web sources, retrieve information from private knowledge bases, and perform various actions on behalf of users. Read: Meet Silicon Valley’s Generative AI Darling Furthermore, Claude 2.1 achieves a significant milestone with a 2x decrease in false statements compared to its predecessor, Claude 2.0. This boost in honesty empowers enterprises to deploy AI applications with greater trust and reliability across their operations. Rigorous testing revealed Claude 2.1’s increased likelihood to demur rather than provide incorrect information, bolstering its credibility in handling complex, factual questions. Claude 2.1 has made significant gains in honesty, with a 2x decrease in false statements compared to Claude 2.0.This enables enterprises to build high-performing applications that solve business problems with accuracy and reliability. pic.twitter.com\/Km5UuYHR3M— Anthropic (@AnthropicAI) November 21, 2023 Simplifying the developer Console experience, Anthropic introduced Workbench, a tool that enables developers to iterate on prompts in a playground-style environment. This facilitates faster learning and optimization of Claude’s behaviour. System prompts have also been introduced, allowing users to provide custom instructions to enhance Claude’s performance, aligning responses with specific personalities or roles. Claude 2.1 is now available in the API and powers the chat interface at claude.ai for both free and Pro tiers. The usage of the 200K token context window is exclusive to Claude Pro users, who can now upload larger files than ever before.","excerpt":"The new model offers context length up to 200K, compared to OpenAI’s GPT-4 Turbo, which offers only 128K.","categories":["AI News"],"tags":["Claude"],"author_name":"Mohit Pandey","publish_date":"2023-11-21T22:53:20","publication_year":"2023","word_count":416,"keywords":["Anthropic","Go","OpenAI","AI","Claude","ML","RAG","Aim","generative AI","Rust","R"],"extracted_tech_keywords":["AI","ML","generative AI","OpenAI","Anthropic","Aim","RAG","R","Go","Rust"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/anthropic-launches-claude-2-1-surpasses-gpt-4-turbo-in-context-length\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10143537,"title":"Cohere Launches Command R7B to Disrupt Enterprise AI Market","content":"Canadian AI startup, Cohere has launched Command R7B, the smallest model in its R series of large language models (LLMs), targeting businesses with a focus on speed, cost efficiency, and flexibility. The model is suitable for deployment on low-end GPUs, CPUs, and even MacBooks. It supports a context length of 128k and offers features such as retrieval-augmented generation (RAG) with native inline citations, multilingual capabilities, and performance across math, code, and reasoning tasks. Cohere highlighted its suitability for enterprise use cases such as customer service and HR. “Command R7B balances efficiency with performance, allowing businesses to deploy high-quality AI solutions on affordable infrastructure,”  Cohere said in its announcement. The model has demonstrated strong results on the HuggingFace Open LLM Leaderboard and outperforms competitors in tasks related to RAG, tool use, and AI agents. Its performance has been evaluated against multiple benchmarks, including ChatRAGBench, StrategyQA, and the Berkeley Function-Calling Leaderboard. Command R7B is accessible via the Cohere Platform and HuggingFace, with the model’s weights released for use by the AI research community. Cohere is offering the model at $0.0375 per million input tokens. Command 7B joins other small language models released this week, including Microsoft’s Phi-4 and Google’s PaliGemma 2. Cohere recently launched Rerank 3.5 to enhance search relevance and content ranking for enterprises, offering multilingual support in over 100 languages, including Arabic, Chinese, English, French, German, Hindi, Japanese, Korean, Portuguese, Russian, and Spanish. The company recently secured a $240 million investment from the Canadian government to build a multibillion-dollar AI data centre in Canada. Founded in 2019, Cohere specialises in developing LLMs for business applications. Unlike some of its counterparts, such as OpenAI and Google, Cohere focuses on enterprise solely rather than pursuing artificial general intelligence (AGI). Earlier this year, Cohere reached a valuation of $5.5 billion raising  $500 million in their Series D funding round.","excerpt":"The model is suitable for deployment on low-end GPUs, CPUs, and even MacBooks.","categories":["AI News"],"tags":["Cohere"],"author_name":"Siddharth Jindal","publish_date":"2024-12-13T23:17:26","publication_year":"2024","word_count":306,"keywords":["Go","funding","startup","OpenAI","AI","RPA","RAG","AI research","small language models","R","Cohere"],"extracted_tech_keywords":["AI","OpenAI","small language models","RAG","R","Go","RPA","startup","funding","AI research"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/cohere-launches-command-r7b-to-disrupt-enterprise-ai-market\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10048737,"title":"IISc To Setup AI-ML Centre in Bangalore","content":"Indian Institute of Science (IISc) has announced plans to set up a state-of-the-art Artificial Intelligence & Machine Learning (AI-ML) Centre in collaboration with Kotak Mahindra Bank Limited (KMBL). The new centre will be known as Kotak-IISc AI-ML Centre at the IISc campus in Bangalore. It will offer Bachelor’s, Master’s, and short-term courses in artificial intelligence, machine learning, deep learning, fintech, reinforcement learning, and much more. The centre also aims to promote research and innovation in AI and ML and develop a talent pool from across the country to provide cutting-edge solutions to meet the industry’s emerging and future requirements. “As IISc continues to deliver on its mandate to provide advanced scientific and technological research and education, it is partnerships with forward-thinking institutions such as Kotak Mahindra Bank that will help us to scale up substantially and position India as a deep tech innovation hub. The future is exciting, and we are delighted to welcome Kotak Mahindra Bank on board.”, said Prof. Govindan Rangarajan, Director, Indian Institute of Science (IISc). Rohit Rao, Joint President & Group Chief CSR Officer, Kotak Mahindra Bank Limited, said, “The Kotak IISc AI-ML Centre will be a Centre of Excellence that will offer academic courses and also undertake research & innovation to further enhance India’s scientific capabilities. The centre is envisaged to be the “Gangotri” of advanced sciences encompassing education, research and innovation in artificial intelligence and machine learning to develop the next generation of leaders in the areas of IT and robotics and champion India’s technology advancement.” With an innovative vision, the Kotak-IISc AI-ML centre plans to be a pioneer amongst educational institutions in the years to come. As per the latest National Institute of Ranking Framework (NIRF) 2021, IISc bagged the second top institute ranking in the country. To be started under KMBL’s CSR project on Education & Livelihood, the Kotak-IISc AI-ML Centre will be spread across approximately 1,40,000 square feet at the IISc campus Bengaluru, as reported by KMBL.","excerpt":"The new centre will be known as Kotak-IISc AI-ML Centre at the IISc campus in Bengaluru.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","Data Science","Deep Learning","education industry india","IISc","iisc bangalore","Machine Learning"],"author_name":"Victor Dey","publish_date":"2021-09-17T13:39:40","publication_year":"2021","word_count":326,"keywords":["Go","artificial intelligence","machine learning","AI","innovation","ML","Machine Learning","IISc","Aim","iisc bangalore","deep learning","GAN","Deep Learning","Data Science","R","AI (Artificial Intelligence)","education industry india"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","Aim","R","Go","GAN","innovation"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/iisc-to-setup-ai-ml-centre-in-bangalore\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10074438,"title":"Glance Is So Pushy","content":"Around 2007, Mohit Saxena, Amit Gupta, Abhay Singhal and Naveen Tewari wanted to leverage the cheap massification of the internet. They started an SMS-based search engine service called ‘mKhoj’ in Bengaluru. By 2008, the company had rebranded itself as ‘inMobi’ and made a sharp pivot towards mobile advertising. The foresight that the founders had to spot a gap for them to make money was prophetic. Back then, mobile phones were a rare spotting and the Apple and iOS platforms were far from being materialised. Mobile advertising was such a novel undertaking that there were only two serious competitors in the market even in the US—Google and AdMob (which was eventually acquired by Google in 2009). Meanwhile, inMobi was building momentum in foreign markets like South East Asia, Africa and the Middle East. In 2011, inMobi was minted as India’s first unicorn startup. Price Band Split for Glance Active Users in Q1 of 2021, Source: Counterpoint Research Racing to unicorn status In 2019, a subsidiary emerged from inMobi called ‘Glance’. An AI-based software company, Glance was a platform that delivered personalised content to users in multiple languages like English, Hindi, Tamil, Telugu and Bahasa. In September that year, it bought short-video social media platform, ‘Roposo.’The startup then raised a total of USD 145 million from Mithril Capital, a venture capital firm co-founded by Peter Thiel of PayPal fame, Ajay Royan, and Google in December 2020. It was luckily a well-timed moment. TikTok had been banned from the Indian market in June 2019—splitting the playing field wide open for short video content platforms. Within a span of 20 months, Glance had become one of the fastest startups to achieve unicorn status. In February 2022, the Mukesh Ambani-led ‘Jio Platforms’, a subsidiary of Reliance Industries announced that it was investing USD 200 million in Glance, in a Series D funding round. According to a report by Counterpoint Research’s Quarterly Mobile Application Tracker released in 2021, Glance had crossed 150 million daily active users in India by the second-half of 2021—an 8% increase from the past year—and is reportedly developing a growing appetite in more ways than one. Tewari has bought a gaming platform Gambit Sports Pvt. Ltd. to help build NFT-based live gaming experiences for its lock screens. Glance has also started hiring in the US, pre-empting its launch there in the next quarter. ​​ Big market for cheaper phones A company that was unheard of a couple of years back was suddenly on every Android phone’s lock screen and counted the who’s who from Silicon Valley on its investor rolodex. How had it managed this meteoric rise? First off, Glance isn’t an app, it is a pre-installed feature in Android phones made by brands like Samsung, Xiaomi, Vivo and Gionee. A report by Counterpoint Research stated that around 70% of all newly-launched smartphones in India, under the sub-USD 250 segment have Glance pre-installed. Rohan Choudhary, co-founder and VP of Glance Feed says that the primary idea behind the platform was to add customer value. “For some reason, news apps haven’t been able to create habits for users. Glance fills that space for them with snippets of your favourite show or a news bulletin. Users can fiddle with apps like Wordle for a few minutes several times a day without downloading it,” he explained. According to the report, 45% of the platform’s traffic comes from users with phones priced between the range of USD 150 and 250, while 12% users are from phones below the USD 100 price point. India is chiefly (80%) a market for smartphones priced below the USD 250 range, while 79% of Glance users in India use smartphones priced below the USD 250. Smartphone users with phones within the price range between USD 100 and 150 and above USD 250 comprise 21% respectively of the company’s user base. The massive demand for affordable smartphones in India is possibly what caught Jio’s eye. The company has announced the launch of its new smartphone in India, the ‘JioPhone’—in collaboration with Google. To be released at an expected price of INR 3499, the smartphone will be the cheapest available in the global market. Market disruptor but at high cost As tough as the online content business is, investment has poured in from prominent resources because of inMobi’s precocious moves. In the world of ad space where big tech organisations like Facebook and Google pretty much monopolise, InMobi has proven to be a strategic disruptor. By way of planned acquisitions and partnerships, it has managed to carve a niche for itself among behemoths. Especially in markets like China, where it has edged out local competition aided by the absence of Google and Facebook. Now, Glance is planning to launch a new Live feature on its lock screens which senior executives believe will be the “next big frontier for content consumption on the internet.” According to Choudhary, the live streaming feature is a “perfect fit” for the lock screen. “Our product has evolved over time and live streaming is apt for Gen Z users,” he said. However, all this market domination has come at a hidden cost. An early FTC case from 2016 against InMobi—that was eventually settled—showed that inMobi had sidestepped consumer choice. Despite a user denying access to the location API, inMobi had gone on to share the user’s location details, in turn showing geo-targeted ads. The company was reportedly collecting information from the WiFi that the device was connected to or that was in close proximity to the user and utilising this to determine their location. With several users calling the platform “annoying, intrusive and battery-draining,” there are online discussion threads dedicated to learning how to disable it. Choudhary refutes this. “It was essential for our team to give users the choice to opt out of the service if they want. They can disable the platform from within Glance itself and even choose the categories they want to shut out,” he pointed out. “Consumers have to feel like they have control over the service. The platform uses social personalisation. Because Glance is a lock screen content service, we had to be extra cautious that the content was not startling,” he added. Because of its pre-installed feature, there are questions around how fair the daily active user metric is to gauge the platform’s success. One Android user commented on a twitter thread discussion around Glance saying, “A pre-loaded spamware on a phone’s lock screen should not be compared to intentionally used apps,” while another noted that Glance users may not even be aware that they were using Glance. Tewari and co. clearly have big ambitions with InMobi and Glance but with how far have they come in terms of market capture, how far have they really come with brand recall and user engagement? To this Choudhary responded by saying, “We admit that 40% of the user base may not like us, and that’s alright. From our retention rate, it is clear that consumers have chosen to keep us. If you like us you can engage, if you don’t, please go ahead and use the phone.”","excerpt":"Because of its pre-installed feature, there are questions around how fair the daily active user metric is to gauge the platform’s success.","categories":["AI Features"],"tags":[],"author_name":"Poulomi Chatterjee","publish_date":"2022-09-06T17:05:00","publication_year":"2022","word_count":1183,"keywords":["Go","API","funding","unicorn","AI","venture capital","RAG","GAN","R","startup"],"extracted_tech_keywords":["AI","RAG","R","Go","API","GAN","startup","unicorn","venture capital","funding"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/glance-is-so-pushy\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10050321,"title":"How Is Online Machine Learning (OML) Unique From Traditional Machine Learning?","content":"Online machine learning (OML) is a type of machine learning (ML) in which data is acquired sequentially and utilised to update the best predictor for future data at each step, in contrast to batch learning techniques, which generate the best predictor by learning on the full training data set at once. In comparison to “conventional” machine learning solutions, online machine learning takes a fundamentally different approach, one that recognises that learning environments can (and frequently do) change from second to second. It is employed in cases when the algorithm must adapt dynamically to new patterns in the data or when the data is generated as a function of time. OML is a widely used technique in areas of machine learning when training over the complete dataset is computationally impractical, necessitating the employment of out-of-core algorithms. OML, in its simplest form, is a machine learning technique that ingests a sample of real-time data, one observation at a time. OML applies to challenges in which samples are provided over time, and their probability distributions are also expected to change over time. As a result, the model is anticipated to evolve to capture and respond to such changes at a similar rate. This could be viewed as a benefit in a particular industry where real-time personalisation is critical. Training and Complexity In an offline ML model, particularly during the training process, the weights and parameters of the machine learning model are updated while attempting to minimise the global cost function using the data used to train the model. The model is continuously trained and updated until it is robust enough for deployment and big data processing, as well as for any other use case. However, in an OML process, the weight changes that occur at a given step are dependent on the (current) example being shown and possibly on the model’s current state. As a result, the model is always exposed to fresh data and improving (learning). Time taken In general, offline ML model training is much faster than online model training because the dataset is only used once throughout the model to modify the weights and parameters. However, due to the magnitude of modern big data streams, it can be rather time-consuming to feed all data into an offline model. It may be preferable to update the model incrementally. Thus, in OML, the model must obtain and tune its parameters in real-time as new data becomes available. This may occasionally incur a higher cost and necessitate the use of much more resources (cluster) to train the model continuously. FeaturesMLOMLComplexityComplexity is reduced because the model is constant.Dynamic complexity due to the model’s continuous evolution.Computational PowerFewer computations, batch-based training at a single time.Model refinement computations are driven by continuous data ingestion.ApplicationsImage classification, or anything else involving machine learning, where data patterns are consistent and there are no rapid concept shifts.Used in fields such as finance, health, and economics where new data patterns emerge on a regular basis.ToolsSci-kit, Spark MLlib, TensorFlow, Keras, Pytorch.Active research: MOA, SAMOA, Scikit-multiflow, streamDM. OML Libraries River is a Python library for OML. It was created by combining creme with scikit-multiflow. River’s goal is to become the standard library for doing machine learning on streaming data. For various online learning activities, it delivers cutting-edge learning algorithms, data processing methodologies, and performance indicators. Several more libraries are available for OML. Python scikit-learn or Orange module. In the case of online learning, Scikit-learn includes an SGD classifier and regressor that may do a partial fit of the data. Caret package in R.Jubatus in C++ – it supports C++, Python, Ruby, and Java clients.The Tornado Framework in PythonLIBOL in C++ (and Matlab). LibTopoART library in C#.","excerpt":"When data is continuously streamed, online learning is essential in order to do real-time analysis.","categories":["AI Features"],"tags":["C","Machine Learning","MATLAB","online machine learning","Python","Ruby"],"author_name":"Dr. Nivash Jeevanandam","publish_date":"2021-10-04T16:00:00","publication_year":"2021","word_count":612,"keywords":["Go","scikit-learn","machine learning","Keras","C","AI","PyTorch","MATLAB","online machine learning","Machine Learning","ML","Python","Ruby","TensorFlow","R"],"extracted_tech_keywords":["AI","machine learning","ML","TensorFlow","PyTorch","Keras","scikit-learn","Python","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/online-ml-and-traditional-ml\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10144076,"title":"Italy Fines OpenAI 15 Million Euros Over ChatGPT Privacy Violations","content":"Italy’s Data Protection Authority (Garante) has fined OpenAI €15.58 million (approx INR 140 crore)  over violations of data protection laws. The investigation scrutinised the handling of personal data by ChatGPT, citing concerns about transparency and the legal justification for processing user information. The Italian watchdog’s decision, announced on December 20, 2024, marks the first significant General Data Protection Regulation (GDPR) penalty against a major player in the generative AI sector in Europe. According to Garante’s findings, OpenAI had processed users’ personal data for training ChatGPT without an adequate legal basis, thus violating principles of transparency and failing to meet information obligations towards users. The investigation also found the company failed to comply with transparency requirements and lacked an effective age verification system to shield children under 13 from potentially harmful content. In addition to the fine, OpenAI has been ordered to run a six-month public awareness campaign in Italian media, detailing how ChatGPT collects and uses data from both users and non-users to train its algorithms. OpenAI described the decision as ‘disproportionate” and announced plans to appeal. The company stated that the fine exceeds its revenue from Italy during the relevant period and emphasised its cooperation with the regulator along with efforts to address prior privacy concerns. The investigation began in 2023 after Italy’s watchdog, known as Garante, temporarily banned ChatGPT for breaching EU privacy rules. The ban was lifted after OpenAI implemented changes, including allowing users to opt out of their data being used for algorithm training. The €15 million penalty is part of GDPR enforcement, which allows fines up to €20 million or 4% of a company’s global turnover. Garante stated that OpenAI’s cooperative stance was considered when determining the penalty, suggesting it could have been higher. OpenAI recently concluded its ‘12 Days of OpenAI’ by launching the next-generation frontier models o3 and o3 Mini.","excerpt":"The company stated that the fine exceeds its revenue from Italy.","categories":["AI News"],"tags":["OpenAI"],"author_name":"Siddharth Jindal","publish_date":"2024-12-21T15:54:33","publication_year":"2024","word_count":307,"keywords":["Go","ChatGPT","AWS","OpenAI","AI","GPT","generative AI","GAN","R","llm_models:GPT"],"extracted_tech_keywords":["AI","generative AI","ChatGPT","OpenAI","AWS","R","Go","GPT","GAN","llm_models:GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/italy-fines-openai-15-million-euros-over-chatgpt-privacy-violations\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10064972,"title":"Major tech appointments in 2022 (till now)","content":"We are only halfway into April and have already seen major tech leadership changes globally and in India. Let’s take a look at the major tech hirings this year to date. Wipro appoints Anis Chenchah as CEO, APMEA Last week, Wipro Limited announced the appointment of Anis Chenchah as a member of the Wipro Executive Board and Chief Executive Officer, APMEA. Anis brings to Wipro over two decades of experience and has worked in IT and business services as well as consulting. Before this, Anis served as the Global CEO of Capgemini Business Services and a member of the Group Executive Committee. Before that, he spent seven years in Logica. He holds a master’s degree in Engineering from ENSISA in France. Wipro appoints Badrinath Srinivasan as MD, Southeast Asia In January, Wipro announced the appointment of Badrinath Srinivasan as the managing director for Southeast Asia. In this role, he will focus on Wipro’s business growth, revenue expansion, client and influencer relationships, talent development and brand building. Before this, he was the Asia leader for the financial services and insurance verticals at Infosys. He has spent over 20 years in Infosys, holding multiple leadership roles in sales and consulting across geographies such as the United States of America, Europe, and Asia-Pacific markets. He has an engineering degree in electronics and communications and an MBA from the Indian Institute of Management, Lucknow. Raj Subramaniam to become president and CEO of FedEx FedEx Corp has promoted Raj Subramaniam, president and chief operating officer at FedEx to president and CEO, effective from June 1, while Frederick W Smith, chairman and chief executive officer, will become executive chairman. Raj joined FedEx in 1991 and has held several leadership positions in the company. He has also served as the president and chief executive officer of FedEx Express. Raj holds a bachelor’s in engineering from IIT, Bombay, MS from Syracuse University and an MBA from The University of Texas. HPE appoints Trish Damkroger as chief product officer of the HPC and AI Trish Damkroger has joined HPE as chief product officer, HPC and AI. She will design and drive an end-to-end strategy for the entire product portfolio across the HPC and data solutions. Trish came with over 30 years of experience and served as vice president and general manager of HPC at Intel in her last role. Before Intel, she was the deputy associate director of Computation at the US DOE’s Lawrence Livermore National Laboratory (LLNL). She led a team of more than 1,000 engineers and scientists working on supercomputers. TCS reappoints Rajesh Gopinathan as MD and CEO TCS board has reappointed Rajesh Gopinathan as the managing director and CEO and Ganapathy Subramaniam as the COO and executive director for five years, effective from February 21. Rajesh has been the CEO and MD of TCS since February 2017. Ganapathy has also been with TCS since February 2017 as the COO and executive director. Addverb Technologies appoints Tapan Pattnayak as director, System Architecture Addverb Technologies has appointed Tapan Pattnayak as director, System Architecture. Tapas comes with a rich experience of working in top companies. Prior to this role, he worked at Google as a hardware engineer with vast experience designing and developing ICs, PCBs, power systems etc. He has also worked in Intel, Cypress Semiconductors and NVIDIA. Meesho appoints Debdoot Mukherjee as chief data scientist Meesho has appointed Debdoot Mukherjee as the chief data scientist. He comes with over 14 years of experience. Before this, he was the vice president of AI at ShareChat. He has held leadership roles in data science at Myntra and Hike before that. He has an MTech in computer science and engineering from IIT, Delhi. NTT appoints Avinash Joshi as India CEO NTT Ltd has appointed Avinash Joshi as CEO of its India business. Sharad Sanghi has been made the managing director and will lead the businesses of NTT Ltd. in India. Avinash will be responsible for end-to-end go-to-market, sales, delivery, operations, and profitability for NTT Ltd in India and report to Sharad Sanghi. Before this, Avinash has spent around two decades in IBM and boasts expertise in business development, account management, sales and delivery. Earlier, he was a senior partner and vice president, 5G, APAC, IBM Services. Byju’s appoints Vedhanarayanan Ganeshkumar as vice president, Technology Ed-tech giant Byju’s has appointed Vedhanarayanan Ganeshkumar as vice president, Technology. He will build and lead a team of engineers, software development managers, software development engineers, product managers and program managers. Before joining Byju’s, he worked at Amazon for around 15 years, where he served as a senior manager of software development. He holds a master’s in computer applications from Anna University.","excerpt":"Addverb Technologies has appointed Tapan Pattnayak as director, System Architecture.","categories":["AI Features"],"tags":[],"author_name":"Sreejani Bhattacharyya","publish_date":"2022-04-14T12:00:00","publication_year":"2022","word_count":775,"keywords":["Go","data science","programming_languages:R","AI","programming_languages:Go","Ray","llm_models:Gemini","GAN","R"],"extracted_tech_keywords":["AI","data science","Ray","R","Go","GAN","llm_models:Gemini","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/major-tech-appointments-in-2022-till-now\/","complexity_score":4,"technical_depth":9,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10087851,"title":"EVs Are Charged With VCs’ Fuel","content":"Barely two months into 2023, the EV market has already seen investments worth $76.2 million — and there is no stopping. In 2022, the Indian EV segment saw total funding of $1.66 billion, a 117% increase from the $766 million in 2022, valuing the market at $3.21 billion. In the first 10 months of 2022, about 90% of the Indian automotive investments by VCs and PEs were in pureplay EV industries, which is around $982 million. According to an estimate by the Indian Private Equity & Venture Capital Association (IVCA), the investment is slated to reach around $20 billion by 2030. Last year’s highlight was Ola Electric raising $200 million from Tekne Private Ventures, Edelweiss, and Alpine Opportunity Fund in Series-C, valuing itself at $5 billion. Ampere Vehicles raised $220 million in a Series-B fund. The second biggest was Ather Energy’s $128 million deal in Series-E with Caladium Investment as their lead investor. The Union Budget-2023, too, aimed at boosting the sector through concessional duty on the components of lithium-ion batteries for another year, along with a reduction of the indirect taxes from 21 to 13 percent. The Faster Adoption and Manufacturing of Electric Vehicles (FAME) programme along with the National Electric Mobility Mission Plan, which has a target of achieving 30% electric mobility by 2030, has been the driving force for the industry for the past few years. Interestingly, all of this points towards the Fortune Business Insights report prediction, according to which, the Indian EV market is expected to grow to $113.99 billion by 2029, with an annual compound growth rate of 66.52%. E-scooters Face the Heat Globally, since 2017, VCs have poured more than $5 billion into e-scooters. But astonishingly, a lot of European countries are rethinking the idea of e-scooters. Paris, a ‘jungle’ of electric bikes and scooters, is going to vote in April on whether it should ban e-scooters rentals altogether. Barcelona has already banned e-scooters sharing services, and the Netherlands does not even permit them on the road. Even the UK is monitoring the whole system of how people find their e-scooters, while also questioning the low-carbon label that they currently hold. E-scooters were introduced, not to earn money for the investors, but for environmental good. But a research by ETHZurich claimed otherwise. It said that while e-scooters in general emit less C02, the sharing schemes have contributed to more carbon emissions than the modes of transportation they have replaced – walking and public transport — by around 51%. Another study in Paris recorded that 72% of the respondents shifted to e-scooters from walking or taking public transport, rather than the expected shift from cars, which e-scooters companies set out to replace. India is still in the adoption stage for e-scooters, but may be looking at a similar fate in the future. Here, the biggest concern at the moment, apart from infrastructure, is the safety around e-scooters. In 2022, scooters lighting up was one of the highlights of the industry, which led to the government adding phase one of the EV battery safety standards in December. A more rigorous second phase of this testing, which involved design-level changes has the deadline of March 31, 2023. This testing is expected to have a very negative impact on the EV industry as a whole. Swapnil Jain, co-founder of Ather Energy, told AIM that India’s outlook towards electric two wheelers is quite different from the rest of the world and the adoption is only increasing. According to him, India’s total EV penetration remained around 3% in 2022, which includes two-wheelers. Citing his 224% YoY growth, Jain said that e-scooters will become mainstream in five years. What this means With these regulations and an overall change of perception about e-scooters, the EV industry as a whole gained a lot of momentum in 2022, with car companies spending around $515 billion to pivot from gas to electric engines. What this means is a need for better infrastructure for the same. And since the new budget is focusing on green mobility, Indian EV prices are expected to drop. However, the good news for investors is that since most LPs are pushing to invest in climate tech startups, EVs are the hottest ones to go for, just like last year. And with the developing technologies, the prices for building EVs is expected to drop even lower, making EVs more accessible in Tier-II and Tier-III cities. Another bright spot is the lithium reserve found in J&K. All this while lithium was imported from China to build batteries in India – clearly going against the Make in India initiative of the government. But now, with the discovery of 5.9 million tonnes of lithium reserves in Jammu and Kashmir, the import tax can be completely snipped, speeding up the mission.","excerpt":"Electric cars segment remains the most-funded business model in the EV industry, followed by e-scooters","categories":["AI Startups"],"tags":["electric vehicles","EV"],"author_name":"Mohit Pandey","publish_date":"2023-02-22T13:00:00","publication_year":"2023","word_count":796,"keywords":["Go","API","funding","programming_languages:R","AI","venture capital","programming_languages:Go","Aim","electric vehicles","EV","R","startup"],"extracted_tech_keywords":["AI","Aim","R","Go","API","startup","venture capital","funding","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-startups\/evs-are-charged-with-vcs-fuel\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10104720,"title":"Mistral AI Challenges Dominance of OpenAI, Google &#038; Meta","content":"Last week, the focus was on Gemini. However, this week, everyone is talking about Mistral AI, a Paris-based AI startup that raised over $113 million in June, even without a tangible product. The buzz around Gemini couldn’t sustain for even a week when Mistral AI captured the spotlight with the release of its latest model, Mixtral 8x7B. This model is a combination of the Sparse Mixture of Experts (SMoE) with open weights, and it has been shared through a magnet link on X. well that was quick pic.twitter.com\/IokTKJgGRa— near (@nearcyan) December 12, 2023 Cannot Ignore Mistral AI Mistral AI’s latest model, 8X7B, based on the MoE architecture, is comparable to other popular models such as GPT 3.5 and Llama 2 70B. Licensed under Apache 2.0, Mixtral surpasses Llama 2 70B on most benchmarks with 6x faster inference. Mistral AI brands itself as the ‘Mixtral of Experts.’ That’s clever marketing right, considering that OpenAI has been doing the same thing for training GPT-4 since last year. However, somehow with Mistral AI’s latest model, it suddenly has gained popularity. Mixture of Experts enable models to be pre-trained with far less compute, which means you can dramatically scale up the model or dataset size with the same compute budget as a dense model. It is a decoder-only model where the feedforward block picks from a set of 8 distinct groups of parameters. At every layer, for every token, a router network chooses two of these groups (the “experts”) to process the token and combine their output additively. This method enhances the model’s parameter count while managing computational expenses and processing time. Specifically, Mixtral boasts a total of 46.7 billion parameters; however, it effectively utilises only 12.9 billion parameters for each token. As a result, it processes input and produces output with comparable speed and cost efficiency to that of a 12.9 billion-parameter model. However, OpenAI Scientist Andrej Karpathy said  “8x7B” name is a bit misleading because it is not all 7B params that are being 8x’d, only the FeedForward blocks in the Transformer are 8x’d, everything else stays the same. Hence also why the total number of params is not 56B but only 46.7B.” Mistral AI Masters Business The Paris-based startup is on a roll, and also announced to secure $415 million in funding with a valuation of $2 billion. Andreessen Horowitz (a16z) spearheaded the latest funding round, accompanied by a renewed investment from Lightspeed Venture Partners. Open-Source LLM firms often find it difficult to sustain their business. To overcome this, Mistral AI recently introduced ‘La Plateforme’ where it will provide API endpoints for its available models. https:\/\/twitter.com\/1littlecoder\/status\/1734155325119381712 The company has created three categories for its models- Mistral Tiny, Mistral Small and Mistral Medium. Mistral 7B Instruct v0.2 and  Mixtral 8x7B, comes under Mistral Tiny and Mistral Small respectively.  Interestingly, the Medium model is yet to be released. Mistral AI has stated that it is currently developing Mistral Medium, positioned among the top-serviced models based on standard benchmarks. Proficient in English, French, Italian, German, Spanish, and code, it achieves a score of 8.6 on MT-Bench.On paper, it even beats GPT 3.5. Interestingly, Mistral opted to launch a paid end-point and refrained from open-sourcing their medium model, which exhibits superior metrics. Introducing hosted API endpoints serves as the most effective method to swiftly gather customer feedback, iterate on real-world use cases, and, crucially, monetize open-source models. On the contrary, Stability AI is currently struggling to generate sufficient revenue for survival. As a response, the company has introduced Stability AI Memberships, charging developers a fee to use its LLMs for commercial purposes. Meta has always been a torchbearer for the open-source community, consistently publishing research papers and releasing models. However, one thing Meta doesn’t necessarily need to prioritise is generating revenue, as it already earns significantly from advertising through its family of social media apps. The startups which are foraying into creating open-source models just cannot keep creating them without monetising it. As Mistral AI has raised a substantial amount, the investors might be hoping for a return on their investment. Mistral AI the next OpenAI? Europe recently reached a preliminary agreement on important rules for using AI in the European Union. Surprisingly, Mistral AI wasn’t in favour of endorsing the EU AI Act. The company might have felt that it will hinder its progress in the near future, potentially requiring the disclosure of trade secrets. As a result, they, along with other open-source companies, were exempted from it. https:\/\/twitter.com\/bindureddy\/status\/1733581858858996007 Mistral AI may not continue to release its upcoming models as open source, which is a speculation. This is considering that OpenAI, too, started out as an open-source company. Interestingly, a few months back, OpenAI lobbied the EU to weaken the much-talked-about European Union (EU) AI Act to reduce the regulatory burden on the company. Karpathy pointed out that the same thing and said that “Glad they refer to it as “open weights” release instead of “open source”, which would imo (in my opinion), require the training code, dataset and docs” Currently, there are not many AI startups from Europe which have seriously challenged OpenAI and Google. Though Mistral AI is making generative AI fun with top notch marketing and good products, it has announced that it is here to stay.","excerpt":"Mistral AI has introduced ‘La Plateforme’ where it will provide API endpoints for its available models.","categories":["Global Tech"],"tags":["mistral"],"author_name":"Siddharth Jindal","publish_date":"2023-12-12T17:46:28","publication_year":"2023","word_count":876,"keywords":["Go","API","TPU","ELT","OpenAI","AI","GPT","Ray","mistral","generative AI","R"],"extracted_tech_keywords":["AI","generative AI","OpenAI","Ray","TPU","R","Go","API","ELT","GPT"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/mistral-ai-challenges-dominance-of-openai-google-meta\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":24533,"title":"See How Google’s Android Things Will Make IoT Devices More Secure","content":"Smart homes popularity has risen substantially over the years. According to Gartner, “the Internet of Things” will include 26 billion units installed by 2020. IoT product and service suppliers will generate incremental revenue exceeding $300 billion, mostly in services, in 2020.” Another media report claims that over the next 10 years the global IoT market will post a revenue CAGR of 21.6 percent. And this spike is pushing every tech companies to rapidly bring IoT devices to the market in order to gain market. Many tech giants, including Google, are investing billions of dollars per year into the IoT ecosystem. Google recently released Android Things version 1.0 for developers prior to this year’s I\/O developers conference. It is the same Android that runs on your smartphone or on your TV or your watch or a Chromebook too. With the help of Android Things, developers can now build new and exciting commercial IoT devices that transform the way we live. What is Android Things? Android Things is an Android-based operating system meant for IoT devices, with a specific focus on fixing the security issue that connected devices usually create. It is designed for making developers’ task easier while working on smart display, kiosks and digital signboards. The basic aim is to give hardware manufacturers a managed OS and certified hardware that allows them to focus on building their product without worrying about the system and its maintenance. And also to transcend the problems Android users have experienced on their smartphones by taking device makers out of the update loop. How Will It Benefit Consumers? With Android Things, Google wants to connect all smart devices without leaking your identity. It wants to provide universal OS layer so that every smart device can work the same way and communicate using the same method, with the added benefit of built-in privacy and security features. Recently, at the I\/O annual developer conference Google announced Smart Displays from Lenovo, LG, JBL. the first Smart Displays will go on sale in July. “These devices can make your day easier by bringing the simplicity of voice with the glanceability of a touchscreen,” Lilian Rincon, Google’s Director of Product Management said at the I\/O keynote. The Smart Displays runs on Android Things platform and will feature new customizable ambient screens, including a clock, the weather or photos. You will also be able to watch YouTube TV right on their Smart Displays. However, Google did not reveal any pricing details as of yet. Android Things Bolsters Security Issue Of IoT Devices A standard IoT device that is available in the market doesn’t receive any software updates and are vulnerable to cyberattacks and data theft. According to a study by HP, 70% of the most commonly used Internet of Things (IoT) devices contain security vulnerabilities, including password security, encryption and general lack of granular user access permissions. Google has fixed that pain point too, it promised that Android Things will get timely and regular updates for three years.“One of the core tenets of Android Things is powering devices that remain secure over time. Providing timely software updates over-the-air (OTA) is a fundamental part of that,” Google said. Hardware and Software Support Android Things is built on Google-certified chipsets from NXP, Qualcomm and MediaTek around which these companies can build smart devices. It will come with support for new system-on-modules based on the NXPi.MX8m, Qualcomm SDA 212, Qualcomm SDA624 and MediaTek MT8516 hardware platforms. It also adds a secured and stripped-down version of Android, on which developers can write apps using standard Android tools The Android Things saw over 100,000 SDK downloads during the preview and over 10,000 developers provided feedback during the beta phase, claims Google. It provides hardware and software developers with all the necessary Software Development Kits to build all kind of internet of thing devices. The Android Things Starter kit are now available to purchase, it includes an NXP i.MX7 developers board, Camera Module, 5-inch display, Pimroni Rainbow HAT board, WiFi, antenna among others. It is available on Digi-Key and Tech Nexion. Developers will get an option for extended support as well. Even after the official support window ends, you will still be able to continue to push app updates to your devices, claims Google. What Devices Are Running Android Things In January, Google announced a number of partnership with device makers including Lenovo, Sony to build smart display devices powered by Google Assistant. During its beta phase, quite a few companies started building products for the platform, including Google’s launch partner LG, Lenovo and JBL for its Android smart display. Devices running this platform have been emerging since January. For instance, Lenovo’s Smart Display that was showcased at CES, which is a Google Assistant Powered speaker and is available in 8-inch and 10-inch display, 10-watt speakers, 5-megapixel camera and Qualcomm processors. LG’s new smart display speaker is using this same platform. The device comes with an 8-inch touchscreen and has Google Assistant and Chromecast built-in. Even JBL Link View, released in January, built on Android Things platform, incorporates an 8-inch touchscreen display, a 5MP camera, two 10-watt stereo speakers and a rear-facing passive radiator. It also comes with Google Assistant and Chromecast support. However, Android Things were not much advertised until Google formally announced this on 2018 I\/O developers conference. “Android Things would soon show up in additional Google Assistant-based smart speakers and Smart Display and would work for everything from connected doorbells to point-of-sale terminals,” Android Things project-management lead Vince Wu said at the I\/O developers conference. Even startups and agencies are also using Android Things to prototype innovative ideas for a diverse set of use-cases. However, the search engine giant has restricted the use of the Things for software updates to 100 active devices in case of non-commercial use. Reportedly, Sony is also using Android Things platform to bring its Smart Display in the market. To sum up, it will be interesting to see, with its new ‘secure’ Platform Android Thing will go be able to beat Amazon Echo in the smart home battle. Android Things will help Google expand the reach of the company’s operating system by securing the smart devices from hacking, which will be a plus point for Google to standout in the vulnerable IoT market.","excerpt":"Smart homes popularity has risen substantially over the years. According to Gartner, “the Internet of Things” will include 26 billion units installed by 2020. IoT product and service suppliers will generate incremental revenue exceeding $300 billion, mostly in services, in 2020.” Another media report claims that over the next 10 years the global IoT market […]","categories":["IT Services"],"tags":["Google","Internet of things","IoT devices","Lenovo","LG","operating system","Sony"],"author_name":"Smita Sinha","publish_date":"2018-05-11T09:00:32","publication_year":"2018","word_count":1044,"keywords":["Go","API","programming_languages:R","AI","IoT devices","LG","operating system","Git","programming_languages:Go","Internet of things","Aim","Lenovo","Google","Sony","R","startup"],"extracted_tech_keywords":["AI","Aim","R","Go","Git","API","startup","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/see-how-googles-android-things-will-make-iot-devices-more-secure\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10101922,"title":"Apple Accepts Right-to-Repair Legislation","content":"On Tuesday, the Joe Biden administration announced that Apple is set to unveil a plan that will make components, tools, and repair documentation for its products available to both independent repair shops and consumers across the nation for fair and reasonable prices. This announcement was made by the National Economic Council director, Lael Brainard, during a White House event focused on the Right to Repair, urging Congress to pass nationwide legislation necessitating such action. The event, in line with President Joe Biden’s drive to boost competition and combat practices that inflate consumer prices, aims to provide consumers with greater autonomy in repairing their possessions, spanning from tractors to smartphones. Brainard highlighted that California, Colorado, New York, and Minnesota have already enacted right-to-repair laws, while an additional 30 states have introduced similar legislation. She noted that Apple is in support of a nationwide law and has previously endorsed the California law, which mandates companies to provide necessary parts, tools, and documentation for repairing consumer electronic devices and appliances to both independent repair shops and consumers at equitable and reasonable rates. The tech giant has committed to extending this approach across the entire nation. In the past, Apple faced criticism from right-to-repair advocates who argued that their sleek devices were challenging to repair, with insufficient support from the company. However, in recent years, Apple has shifted its stance, emphasising the longevity and resale value of its products while simplifying the repair process and increasing accessibility to spare parts. Notably, Apple began distributing parts and manuals to select independent repair shops back in 2019, and in August, the company officially supported right-to-repair legislation within its home state of California.","excerpt":"In the past, Apple faced criticism from right-to-repair advocates who argued that their sleek devices were challenging to repair.","categories":["AI News"],"tags":["Apple"],"author_name":"Mohit Pandey","publish_date":"2023-10-25T10:13:54","publication_year":"2023","word_count":276,"keywords":["AWS","AI","cloud_platforms:AWS","Apple","programming_languages:R","Aim","ViT","GAN","R"],"extracted_tech_keywords":["AI","Aim","AWS","R","GAN","ViT","cloud_platforms:AWS","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/apple-accepts-right-to-repair-legislation\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":56222,"title":"Why Deepnote Is A Great Solution For Real-Time Collaboration In Data Science Workflow","content":"Collaboration in data science projects has numerous challenges as Jupyter Notebooks are strenuous to manage under version control. Consequently, it has become increasingly tricky despite a few workarounds for handling the notebooks. Undoubtedly, such practices have helped data scientists, but it is still far away from what web and application developers enjoy. Besides, data scientists are now not only analysing information to get insights but also optimising algorithms such that firms can integrate it into products. Therefore, they are expected to manage more than just doing research. However, delivering on various requirements requires skills like data administration, creating data pipeline, and more, which bring other professionals to work on the same projects with data scientists. Therefore, adopting best practices for collaboration has become essential within organisations. To enable such practices, Deepnote is developing Jupyter-compatible notebook for real-time collaboration in Data Science for which they have recently raised a seed fund of $3.8Million from Index Ventures, Accel, OpenAI’s Greg Brockman, and more. What Is Deepnote Deepnote is similar to Jupyter Notebook with advanced features like real-time collaboration, versioning, code review, and reusability of algorithms. The idea is to simplify tasks that data scientists manage by enabling others to work along with them in the projects simultaneously. This assists developers to focus on their core tasks like pattern matching, training, and evaluating models, and allow others to handle dependencies, thereby increasing productivity. “The most important feature of Deepnote is the ability to allow users to collaborate,” said Jakub Jurovych, co-founder and CEO. Deepnote For Effectively Managing Notebooks Unlike other developers, data scientists do not write scalable code while moulding information, their prime goal is to find insights and predict outcomes. However, today they are expected to optimising algorithms for making it a production-level code. “If I’m a data scientist, I would mostly spend time doing mathematics and statistics, but expecting me to connect to an EC2 cluster and spin a bunch of GPU instances for parallel training is just not something I’m looking for,” Jurovych. Thus, using, Deepnote, data science practitioners can expedite their day-to-day activity by quickly sharing their notebook just by a few clicks without moving away from the notebook. One can also publish the notebook directly from the Deepnote, thereby eliminating the traditional practice of downloading separate files based on the requirements. Source: Deepnote While real-time collaboration is already available on the platform, functions like versioning, code review, and reproducibility are due to be released in the coming months. Started in 2019 with the mission of enhancing the workflows of data scientists, Deepnote has quickly gained traction among developers. Source: Deepnote The facet of Deepnote book is that it will not only help developers from outside the data science team to assist in the project but also data scientists from within the group while keeping all the changes in check. Version control is vital to expedite projects, and Deepnote book will revolutionise it as one can check changes in code from within the notebook. Today, a considerable amount of time is spent in keeping notebook under version control, resulting in switching between various tools continually to commit and review changes. Deepnote Helps In Programming Source: Deepnote Deepnote allows developers to have a glance at variables at one, enables one to visualise dataframes without writing codes for it. Besides, it has code intelligence feature that will suggest and auto-complete the code, pinpoints bugs, and more. Further, Deepnote integrates with existing solutions like Amazon SE and GCS buckets, PostgreSQL, MongoDB, among others. Its capability to integrate with other services will be crucial for its rapid adoption as it does not disrupt the existing data science infrastructures and processes in organisations. Outlook Instead of spending time in managing notebook, data scientists will be able to work with Deepnote and divide the task with others. However, over the years, we have witnessed different notebooks that came to the market for solving various intricacies in the existing notebooks but failed to make its mark. For one, Netflix’s Polynote – a multi-language interoperability notebook, was unable to impress and is popular among user who likes to switch between languages. However, unlike Polynote, Deepnote is focused around fixing the generic problems and not confined towards solving specific challenges. Deepnote is more like an updated version of Jupyter Notebook, which may just make it a go-to notebook in the future.","excerpt":"Collaboration in data science projects has numerous challenges as Jupyter Notebooks are strenuous to manage under version control. Consequently, it has become increasingly tricky despite a few workarounds for handling the notebooks. Undoubtedly, such practices have helped data scientists, but it is still far away from what web and application developers enjoy. Besides, data scientists […]","categories":[],"tags":[],"author_name":"Rohit Yadav","publish_date":"2020-02-07T19:00:00","publication_year":"2020","word_count":723,"keywords":["PostgreSQL","data science","Go","OpenAI","AI","MongoDB","Scala","SQL","Jupyter","R"],"extracted_tech_keywords":["AI","data science","OpenAI","Jupyter","MongoDB","PostgreSQL","R","SQL","Go","Scala"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/why-deepnote-is-a-great-solution-for-real-time-collaboration-in-data-science-workflow\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10110395,"title":"OpenAI Launches GPT Store, Offering ChatGPT Users Access to Custom AI Models","content":"OpenAI has announced the launch of its GPT Store. The store is now being made available to ChatGPT Plus, Team, and Enterprise users, providing access to a variety of GPTs developed by partners and the community. The company said that the GPT Store, accessible at chat.openai.com\/gpts, showcases a diverse range of GPTs across categories such as DALL·E, writing, research, programming, education, and lifestyle. Users can explore popular and trending GPTs on the community leaderboard, gaining insights into the latest developments. Notable featured GPTs include personalised trail recommendations from AllTrails, academic paper search and synthesis with Consensus, coding skill expansion through Khan Academy’s Code Tutor, presentation and social post design with Canva, book recommendations with Books, and math and science learning with the CK-12 Flexi AI tutor. OpenAI emphasises the simplicity of building and sharing GPTs, with no coding skills required. Users interested in sharing a GPT in the store must save it for everyone and verify their Builder Profile. OpenAI has introduced a review system, involving both human and automated reviews, to ensure compliance with usage policies and brand guidelines. The GPT Store would allow users, specifically ChatGPT Plus and enterprise subscribers, to sell and share tailored AI agents based on GPT-4. OpenAI has announced that in Q1, it will launch a GPT builder revenue program. In this initial phase, US builders will be compensated based on user engagement with their GPTs. Further details on the criteria for payments will be provided later. OpenAI  recently expanded its ChatGPT offerings with the introduction of ChatGPT Team. ChatGPT Team features a dedicated collaborative workspace for teams and admin tools for efficient team management. The newly introduced ChatGPT Team provides access to advanced models like GPT-4 and DALL·E 3, along with tools like Advanced Data Analysis. Similar to ChatGPT Enterprise, businesses using ChatGPT Team retain ownership and control over their data, with no training on business-specific data or conversations.","excerpt":"Two months after the announcement of GPTs, users have already generated over 3 million custom versions of ChatGPT.","categories":["AI News"],"tags":["ChatGPT","OpenAI"],"author_name":"Siddharth Jindal","publish_date":"2024-01-11T00:06:35","publication_year":"2024","word_count":316,"keywords":["Go","ChatGPT","OpenAI","AI","programming_languages:R","GPT","AI agents","R","llm_models:ChatGPT","llm_models:GPT"],"extracted_tech_keywords":["AI","ChatGPT","OpenAI","R","Go","GPT","AI agents","llm_models:GPT","llm_models:ChatGPT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/openai-launches-gpt-store-offering-chatgpt-users-access-to-custom-ai-models\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10995,"title":"Great Lakes launches program with Financial Risk Analytics as a Specialization","content":"The analytics industry has been continuously advancing and India is becoming an emerging hub for analytics solutions across the globe. As the demand keeps growing exponentially, the sector is bound to witness rapid growth. Within analytics, risk analytics is one segment which has lately been under the spotlight because of its increasing demand. Studies estimate that the global risk analytics market is likely to more than double itself to reach USD 50 Billion by 2020. This significant growth in the risk analytics market can be attributed to the rising need of organizations to minimize their losses which are incurred due to risks, maximize their ROI, and enhance the decision making process. With growth in risk analytics, the requirement for talented risk analyst professionals is also on the rise. And because of the already existing shortage of talent pool in analytics industry, the demand for Risk Analytics Professionals is extremely high converting into huge salary packages for the analytics professionals. About Great Lakes Founded in 2004, Great Lakes Institute of Management is one of the top B-Schools in India. Led by exceptional academic faculty, steered by an outstanding advisory council and buoyed by the international collaborations, Great Lakes has, within a short span of 10 years emerged as a top-ranked business school. Great Lakes’ Post Graduate Program in Business Analytics, started in 2013, is India’s No 1 Business Analytics Program. Since its launch, the program has trained 1000+ working professionals with almost 66% of Great Lakes PGP-BABI alumni transitioning into analytics roles and profiles within 6 – 12 months from graduating from the program. Introducing CFRA & its features The Great Lakes Certificate Program in Financial Risk Analytics is an industry collaboration specialized program designed for professionals who want to build their careers in the financial risk analytics industry. The program begins with a primer in statistics and is followed by laying a strong foundation in quantitative methods, financial instruments and markets. With the requisite statistical and financial foundation in place, the candidates then get trained on exhaustive modules, techniques and case studies in Market Risk and Credit Risk. One of the defining features of the program is the coverage of Basel implementation comprising a thorough understanding of Basel Norms followed by its implementation aspects such as IRB, Risk capital calculations, CCAR, Capital adequacy, calculation of risk weights etc. The program is internationally recognized and is a dual certificate program. The participants will get certification from Illinois Institute of Technology, Chicago (USA) in addition to the certificate from Great Lakes Institute of Management. Dr. Bappaditya Mukhopadyay, Program Director (CFRA) on introduction on CFRA “Great Lakes has been a pioneer in high quality analytics education in India with our executive program in Business Analytics being consistently ranked as the best in the country. Having delivered over 250,000+ hours of analytics learning content to working professionals in India and abroad, we have collaborated with the industry to develop a dedicated analytics program in Financial and Risk analytics – CFRA.” Why the need for this course\/specialization? Risk Analytics has been an area of interest to companies and organizations are looking to create a talent pool which can help them gain deeper insights into this field. There is tremendous demand for risk analysts due to the gap in the demand and supply of analytic professionals worldwide. And India is arising as a hot destination globally for analytics talent due to the skill sets possessed by professionals in India. Great Lakes being a premier institute in India for analytics is constantly approached for providing analytics talent pool and with rising need in the risk analytics area; companies are looking for talent pool with specialization in Risk Analytics. These requirements from the industry propelled Great Lakes to collaborate with industry to offer a program in Risk Analytics. CFRA has been designed as a specialized program catering to the Analytics industry need of more- trained talent in financial and risk analytics. If you take a quick glance of all the analytics profile the industry is seeking, finance and risk analytics tops the list by far. The industry needs analytics professionals who have the analytical ability to identify, measure and mitigate financial risks within the overall regulatory framework. This would mean a program that blends the market risk, the credit risk as well as the operational risk within the overall Basel guidelines. Dr. Bappaditya Mukhopadyay commented “Through this program, the participants must be able to value simple as well as complex financial products, understand and develop risk models as well as inculcate the abilities to implement Basel requirements and CCAR.” Details of the CFRA Program Who is it for? Great Lakes CFRA program is ideal for candidates who wish to excel in the domain of risk analytics. Candidates should have a minimum of 2 years of experience while applying to the program. While a prior work experience in banking, finance or consulting is an advantage, candidates from other industries (such as IT\/ITES) serving financial clients are also a good fit. What to expect? On completion of the program, a candidate must be able to appreciate and solve the following problems: How does a financial institution identify, measure and manage market risk How are credit risk models developed and validated How are stress testing of such models undertaken How are regulatory capital (CCAR) calculated How to setup systems that comply with Basel guidelines The curriculum entails exhaustive coverage of academic concepts blended with rich industry exposure and hands-on training to ensure that the candidates graduating from CFRA program are industry ready. Program Format The program duration is 6 months and is a combination of weekend classroom sessions, online lectures and recorded videos \/ pre-reads. The program covers 160 hours of learning comprising 100 hours of classroom learning and 60 hours of online learning. The calendar is designed such that most classes are conducted on weekends and public holidays, thereby causing minimal disruption to the work schedule. Fee for the program: The fee for the program is Rs.3, 00,000 exclusive of Service Tax Batch Commencement Dates: December, 2016 Centre: Gurgaon For more details about the program visit https:\/\/goo.gl\/Y2728u","excerpt":"The analytics industry has been continuously advancing and India is becoming an emerging hub for analytics solutions across the globe. As the demand keeps growing exponentially, the sector is bound to witness rapid growth. Within analytics, risk analytics is one segment which has lately been under the spotlight because of its increasing demand. Studies estimate […]","categories":["AI Trends"],"tags":[],"author_name":"Дарья","publish_date":"2016-10-19T04:31:34","publication_year":"2016","word_count":1017,"keywords":["Go","API","programming_languages:R","AI","programming_languages:Go","RAG","analytics","disruption","GAN","R"],"extracted_tech_keywords":["AI","analytics","RAG","R","Go","API","GAN","disruption","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/great-lakes-launches-program-financial-risk-analytics-specialization\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":63763,"title":"Why Data Science Jobs Market Is Better Positioned For Recession","content":"The ongoing COVID – 19 Pandemic is adversely impacting the global economy, with expected record unemployment, business disruption, and significant contraction of global GDP growth. While data science and analytics jobs are not immune to recession and economic woes, there are multiple cases of hiring for data science roles in various companies. Here is one. The data science job market has also seen fewer lay-offs, salary cuts and furloughs compared to other tech job roles. This is also evident from our own tracking of the lay-offs in the tech domain of India. If you look at the laid-off staff, it is focused on IT roles and positions which can be fairly replaced with automation tools. Data science roles, on the other hand, are incredibly challenging to automate at this point in time. So, job cuts in data science may be minimal if you look at companies removing redundant roles in the workforce. It has been found in research that job volatility in data science has been on the lower side compared to traditional jobs that have been disrupted due to supply chain issues, according to a report. In the report, jobs such as teachers, manufacturing, and hospitality have been witnessing the highest volatility in the ongoing crisis. Software and IT jobs have been found to have mid-level of volatility compared to more traditional jobs. However, different sectors have different conditions — including types of businesses, geographies, and consumers served, debt load, capital and operation costs, and supply and demand for human resources. So, the impact is going to be different in each case, be it the IT sector or FMCG, Industrials, or Domestic Banking. And if we look at how Indian IT industries have performed in the last five years, the tech sector is expected to stay strong, barring some overvalued startups and legacy IT companies. Specialised data science and analytics consulting firms are hiring despite the recession. On the other hand, we have seen lay-offs in startups, IT and services firms. What Makes Data Science Jobs Safer During The Recession Analytics is going to be very crucial in the times of recession for optimising costs and reducing wastage of resources. Quoting IDC’s numbers, WSJ reported that 2020 could see rising demand in AI jobs, given AI applications are being deployed extensively by healthcare providers, educational institutions and government agencies. According to the report, overall AI spending will also increase in 2020. Here, data scientists would play a critical role in getting the best of automated systems that throw out chunks of data. The findings from the AIM survey reveal that the entire analytics community is able to work from home. Moreover, while analytics personnel are working from home, they have access to the required tools and platforms to execute their tasks and responsibilities. Moreover, most of the respondents have either experienced no impact on their work or have experienced a positive effect on their job because of higher productivity. The broad data science domain makes working from home possible under certain conditions, including the type of projects, variety of functions, access to tools, employee engagement, and overall connectivity and collaboration with the rest of the team\/organisation. Given data related jobs can be done with ease over the web and from home, and its growing importance in the tech sector, it has seen minimal volatility despite the slowdown. Regardless, data scientists could still lose their jobs, particularly in more vulnerable sectors such as travel and tourism, logistics, and even retail. Overview 2019 was a great year for the data science jobs market, and the trend is likely to continue despite the global pandemic. In fact, given everything is being virtualised and platformised, in some industries, we are witnessing the advent of more data and more extensive data workloads. This will fuel the demand for more data scientists in such fields. And if firms do lay-off their analytics and data science teams, it would be even a bigger challenge to rebuild those teams. Getting an efficient data science team takes years to build, and it would be a waste of millions of dollars to let-off valuable staff.","excerpt":"The ongoing COVID – 19 Pandemic is adversely impacting the global economy, with expected record unemployment, business disruption, and significant contraction of global GDP growth. While data science and analytics jobs are not immune to recession and economic woes, there are multiple cases of hiring for data science roles in various companies. Here is one. […]","categories":["AI Features"],"tags":["AI Jobs","data management providers","Data Science","Data Science Career","Data Science Jobs","data science salary","data science software","what is data science"],"author_name":"Vishal Chawla","publish_date":"2020-04-28T14:00:00","publication_year":"2020","word_count":686,"keywords":["what is data science","data science","Go","API","AI","AI Jobs","Data Science Jobs","automation","Data Science Career","Aim","ViT","analytics","data science salary","GAN","data management providers","Data Science","data science software","R"],"extracted_tech_keywords":["AI","data science","analytics","Aim","R","Go","API","GAN","ViT","automation"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/data-science-jobs-recession\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10001231,"title":"5 Smart Must-Carry Gadgets For Long Flights","content":"With most people grappling with air sickness, surviving a long flight is not an easy task — from the loud in-flight noise to turbulence to no-legroom seats. So if you want to have a more relaxing and comfortable flight, don’t forget to pack these essential accessories for a comfortable journey: Here Are 5 Gadgets That Will Make Your Next Flight Journey Comfortable Noise Cancellation Headphones There is no denying to the fact headphones are a classic companion for frequent air travellers. Once you are inside the aircraft and it takes off, without a pair noise cancellation headphones, your music will be lost due to the in-flight noise or a screaming baby or a seat neighbour who refuses to stop talking. And if you are looking for a good pair of headphones for your next flight, then Bose could be a great choice. Bose QuietComfort is pretty much best-in-class when it comes to noise cancellation — it could an excellent companion for your long-haul flight. Also,  not to mention Bose practically invented the noise-cancellation headphone category. Smart Luggage Travelling can be either a joy or a pain, and the luggage you carry is one of the biggest factors behind making your journey hassle free. Talking about luggage, Incase ProConnected 4 Wheel Hubless Roller can be a great companion when you are travelling. When you are travelling in flight and have layovers, and just in case if you don’t find a perfect spot with a charging slot, this luggage can be a great partner then. ProConnected 4 is equipped with 20,100mAh of battery power, which is easily removable and also comes with a USB-C and two USB-A ports. One can easily recharge a MacBook Pro or mobile devices for several days. Talking about durability, the luggage is made of polycarbonate and strong polyester. That is not all, the bag has a dedicated mobile app that even tracks battery life and the last known location of your roller. eReader Reading is one of the best ways to utilise time during a long flight. However, carrying a bundle of books might be one of those things you don’t want to do. So, why not carry an eReader. Kindle is definitely a go-to eReader for every avid reader, however, there are few alternatives also that you might want to carry next time you go on a long flight.  One of the best alternatives to Kindle is Kobo Aura, and the device’s hardware compares very well to Amazon’s Kindle Paperwhite. It has a 6-inch Carta E Ink touchscreen and comes with 4GB internal storage, which can be used to hold over 3,000 eBooks. The Smart Sleep Mask Now it’s not just phones that are smart, sleep masks have also become smart and one of the famous is illumy Smart Sleep Mask. It is a sleep mask that uses gentle dimming light to help you fall asleep and gentle brightening light to help you wake up naturally. Some of the smart features of this mask are Built-in Alarm, Silent Alarm Setting, Custom Sunset\/Sunrise length, Sleep Enhance, Home\/Away Time Setting. Smart Insole Known for making cold environment activities comfortable, foot warmers are some of the gadgets money can buy. Smart insoles like Digitsole can be a great choice when you are travelling in a flight the AC is chilling inside the cabinet. Digitsole’s warm series insoles are the first connected heated insoles, which are designed to keep your feet warm. The insoles can be controlled with a dedicated application on your smartphone. To have a comfortable log flight, all you need to do is connect your heated insoles to your phone, select the desired comfort temperature and enjoy.","excerpt":"With most people grappling with air sickness, surviving a long flight is not an easy task — from the loud in-flight noise to turbulence to no-legroom seats. So if you want to have a more relaxing and comfortable flight, don’t forget to pack these essential accessories for a comfortable journey: Here Are 5 Gadgets That […]","categories":["AI Trends"],"tags":["travel"],"author_name":"Harshajit Sarmah","publish_date":"2019-02-20T20:49:56","publication_year":"2019","word_count":610,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","Git","RAG","ViT","travel","R"],"extracted_tech_keywords":["AI","RAG","R","Go","Git","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/5-smart-must-carry-gadgets-for-long-flights\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10088025,"title":"India to Add 85,000 Global Semiconductor Talent by 2025","content":"At the second Semicon India Future Design roadshow held in Bengaluru IISc campus, Union Minister Rajeev Chandrasekhar said that India plans to produce a minimum of 85,000 global semiconductor talent in the next two years. Earlier today, Chandrasekhar said that they see their entities, enterprises and engineers playing a deep, significant and decisive role in how the future of semiconductor design and manufacturing would be shaped. He said that the government is committed to being a catalyst for the success of such ventures. Chandrasekhar also inaugurated the innovation centre set up by a joint partnership with TCS and Renesas Electronics. He said that the centre set up by local partners would be developing next-generation devices and products for the growing global and Indian semiconductor ecosystem. https:\/\/twitter.com\/Rajeev_GoI\/status\/1628680693231529986 Read: Chip War: World Needs India More Than Ever The chip-hungry world needs India as much as India needs them—if not more. Recently, the United States announced a strategic partnership with the country with a focus on critical and emerging technology (iCET). The aim was to strengthen the semiconductor supply chains, support the growth of semiconductor design and manufacturing in India on mature technology nodes and packaging, and cultivate a skilled workforce. The US looks to benefit from India to fill the widening gap between the demand and supply of microelectronics engineers that has been slowing its reshoring plans. The shortage of tech talent has been a major concern for countries worldwide. Just as the software workforce from India has dominated the global industries, India will soon latch onto the hardware talent. Read: The Missing Puzzle to India’s Semiconductor Ambition? Almost all major companies, including Samsung, Micron, Lam Research, NXP and Applied Materials have substantial semiconductor research and development investments in India. These investments are expected to rise and are already leveraging the talent available here. In fact, India is a global leader in engineering design and R&D. It has a favourable demographic advantage with a young workforce and a significant number of technologically-aligned individuals. However, expanding its potential to manufacturing requires more. Unlike the IT sector, the lack of awareness and availability of job options have been some major hurdles facing the industry. In addition, government schemes like Skill India and industry bodies like IESA are attempting to bring a change. Recently, under its Skill India programme, the government announced setting up of multiple international centres across the country, which will skill the youth for international opportunities. Read: India has the Magic Lamp for Semiconductor Industry At one of the centre launches, Chandrasekhar said that India would be launching future labs to channel R&D capital for semiconductor and deep tech development. He said that these labs will embed with C-DAC and will encourage industry and academic collaborations. Last year, at IEEE MAPCON 2022, Chandrasekhar said that the Indian government has been giving capital support for research and development efforts for India to take the leap from being a tech consumer and back-office support to becoming a leading tech producer in the world. He said that skilled engineers are now being encouraged to innovate and start their own ventures and that the government would support such endeavours.","excerpt":"Union Minister Rajeev Chandrasekhar announced the plan at Semicon India Future Design today.","categories":["AI News"],"tags":["Rajeev Chandrasekhar"],"author_name":"Ayush Jain","publish_date":"2023-02-24T12:27:43","publication_year":"2023","word_count":523,"keywords":["Rajeev Chandrasekhar","Go","API","programming_languages:R","AI","innovation","programming_languages:Go","RAG","Aim","R"],"extracted_tech_keywords":["AI","Aim","RAG","R","Go","API","innovation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/india-to-add-85000-global-semiconductor-talent-by-2025\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10165900,"title":"SUSE Security to Integrate with Microsoft Products for Better Threat Response","content":"SUSE, a global open source software company, has announced that it will integrate its Security Solutions with Microsoft Sentinel, a cloud-native security information and event management (SIEM) platform. As cyber threats grow in volume and sophistication, managing security across multiple platforms has become more complex. And this integration allows security data from SUSE environments to flow directly into Microsoft Sentinel, giving organisations a centralised dashboard to monitor activity. Sentinel can automatically raise alerts and isolate affected nodes to prevent further spread of potential threats. The integration is made part of Microsoft Security Copilot, a tool to help analyse security data using GenAI. SUSE, known for its enterprise-grade Linux distribution, said that the Security Copilot correlates SUSE Security data with other data sources within Sentinel to identify complex and sophisticated attacks. This allows customers to proactively address potential threats based on the recommendations. “This new integration is a robust security solution for any organisation running cloud-native workloads on Microsoft Azure,” said Laurent Mechain, global head of cloud at SUSE. “We’re excited to deepen our existing work with Microsoft and reinforce our commitment to powering open, secure innovation.” David Houlding, director of global healthcare security & compliance strategy at Microsoft, noted that the collaboration would help joint customers simplify their security operations and improve threat protection. SUSE highlights its security offering, SUSE Rancher Prime, SUSE’s enterprise container management platform for Kubernetes clusters, as the highlight, which combines with Microsoft’s security solutions to provide comprehensive threat intelligence. The company said, “Combining SUSE’s Kubernetes security with Microsoft’s security tools creates a robust security solution.” The integration, the company claims, enhances security operations by providing broader visibility, faster detection and response, and simplified management across hybrid IT environments.","excerpt":"The connection is part of Microsoft Security Copilot, a tool to help analyse security data using GenAI.","categories":["AI News"],"tags":["enterprise","Linux"],"author_name":"Ankush Das","publish_date":"2025-03-12T17:36:42","publication_year":"2025","word_count":283,"keywords":["enterprise","GenAI","cloud_platforms:Azure","AI","R","innovation","GAN","Aim","ViT","Linux","Azure","kubernetes"],"extracted_tech_keywords":["AI","GenAI","Aim","Azure","kubernetes","R","GAN","ViT","innovation","cloud_platforms:Azure"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/suse-security-to-integrate-with-microsoft-products-for-better-threat-response\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":28335,"title":"Why Data Scientists Are Heavily Weighted Towards Open Source Frameworks: CG Venkatesh of LTI Explains","content":"With 20+ years of expertise in statistics & data science, CG Venkatesh has spearheaded several industry-specific strategic advanced analytics solutions. With his broad experience, he has designed and implemented data analytics solutions for Fortune 500 clients across industry verticals. He is also associated with MIT Sloan, IEEE, AICTE and various other universities for his passion for extended interactions with budding data scientists and academicians. An acclaimed analytics thought leader, he heads the data science practice at LTI. AIM got in touch with CG Venkatesh, better known as CG, to get his insights on the leading tools and techniques that are currently used by analytics, AI and data science practitioners. In this detailed interview, CG gives us the lowdown on the most preferred tool by his team, preference of open source and paid tools, cloud providers, LTI’s in-house tool and more. Analytics India Magazine: What are the most commonly used tools in analytics, AI, data science? CG: These are some of the popular and most commonly used tools according to me. Data Science & Applied Statistics: Commercial products: SAS, IBM SPSS, STATISTICA Open Source: R, Python Pandas, NumPy, SciKit and the libraries based on these tools. AI\/ ML: IBM’s WATSON, Amazon’s SageMaker, Baidu, Cloudera, Confluent, DataBricks, Google ML, Microsoft’s Cognitive and Computer Vision suites Open Source: R, Python, TensorFlow based libraries NLP: Open source frameworks like Stanford NLP, GATE, Python’s NLP libraries AIM: What is the most productive tool that you have come across? CG: Python’s Pandas and R are the most productive tool for data preparation before modeling. Also, all SQL-supporting libraries which offers an efficient manipulation and transformation features based on Matrix Algebra driven data structures like data frames, data sets etc. AIM: Do you prefer tools that are open source or paid? Please elaborate on the benefits, some open source and paid tools that you prefer. CG: The factors that impact the choice of tools are as follows: Availability of skill sets of resources at hand Clients data maturity Client mindset towards open source tools vs licensed tools Scalability of the solution design Build vs Buy: cost benefit analysis Given a choice, I would prefer open source, given the following reasons: Ease of availability Portability across systems Big data capability and handling dynamic volume, velocity, variety of data Absence of protocols due to free availability Scalability in terms of resource as they can be quickly trained Ease in proving do ability & capability for a quick start and client buy in AIM: Is open source considered an important attribute when choosing the tool of your choice CG: Yes, Most definitely. AIM: What are the most common issues you face while dealing with data? How is selecting the right tool critical for problem-solving? CG: Common issues faced in terms of data quality are: History Granularity Logical connectivity between multiple data sets and sources Availability Sufficiency Multiple form and formats of data Right tool combinations are crucial to do the problem solving as that’s the key to extract, transform the data to reach upto the algorithmic stage. AIM: How do you select tools for a given task? CG: Data analysis tasks start with first checking what kind of technologies & processes can interact with the relevant data set, so that an initial analysis, profiling and sampling can be done. This analysis typically involves tools with SQL querying capabilities and tools that can convert data from one format to another quite easily. The tool should also have the capability to easily explore, summarise and visualise the univariate statistical measures. Additionally, if the tool can offer features for inputing & selecting samples, that’s an added advantage. Post the initial profiling, next important step is to choose a tool for data modelling. Often, the driver to the decision is the set of algorithms that we choose for the modelling, and checking where among the tools are those algorithms implemented with great coverage of statistical scenarios – i.e. flexibility to tweak modelling parameters and metrics. AIM: What are the most user-friendly languages and tools that you have come across? CG: R & Python are brilliant for processing data with respect to analytical goals. Azure Machine Learning Studio interface is fast catching up on user-friendliness. AIM: What does the ideal data scientist’s toolkit look like? CG: An ideal data scientist toolkit should be: A SQL-heavy tool\/library to query environments hosting structured data A tool with easy\/intuitive syntax to query environments hosting unstructured data A Studio\/Client GUI that helps in analysis & model brain-storming A visualization tool that visualizes insights without having to code too much A spreadsheet application of course! A tool that helps deploy and unit-test the models, before they can move to production An ever-doubting mind AIM: What is the most preferred language used by the team? CG: R & Python are undoubtedly most preferred for ease of coding and the depth of libraries they offer. AIM: Can you give us the percentage of data scientists and percentage of developers that use a particular language\/data visualization tool etc.? CG: Roughly about 50% – 50% – R & Python – for both scientists & developers. AIM: What is the most preferred cloud provider— AWS, Google or Azure? CG: Azure, due to the friendliness and easy integration to other vast set of Microsoft products. AIM: What are some of the tools used for scaling data science workloads; for eg., Dockers are gaining popularity vis-à-vis spark? CG: Clearly Dockers or self-containing packages, where services\/APIs and applications are run as a processes\/threads– hosted in a micro OS like CoreOS are the future when it comes to delivering millions of many insights on scale. But for batch outcomes, in-memory distributed server environments like Spark still takes the cake. AIM: What are some of the proprietary tools developed in-house by the company? CG: LTI’s Mosaic is a unique offering that leverages the power of data, AI & automation to overcome the challenges of data-driven decision management. The foundation of the platform is equipped with state-of-the-art data engineering and advanced analytics capabilities such as data ingestion, storage and governance, advanced analytics, processing, and consumption adaptors, extending a single interface for ‘Data to Decisions’.","excerpt":"With 20+ years of expertise in statistics & data science, CG Venkatesh has spearheaded several industry-specific strategic advanced analytics solutions. With his broad experience, he has designed and implemented data analytics solutions for Fortune 500 clients across industry verticals. He is also associated with MIT Sloan, IEEE, AICTE and various other universities for his passion […]","categories":["AI Features"],"tags":["Interviews and Discussions"],"author_name":"Srishti Deoras","publish_date":"2018-09-14T07:17:46","publication_year":"2018","word_count":1023,"keywords":["data science","machine learning","AI","ML","computer vision","NLP","Aim","analytics","TensorFlow","Pandas","Interviews and Discussions"],"extracted_tech_keywords":["AI","machine learning","ML","NLP","computer vision","data science","analytics","Aim","TensorFlow","Pandas"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/why-data-scientists-are-heavily-weighted-towards-open-source-frameworks-cg-venkatesh-of-lti-explains\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":20158,"title":"Should Robots Be Granted Rights? Where Does Humanity Lie?","content":"What is conscience? The Oxford English Dictionary defines the philosophers’ favourite locution as, “The part of your mind that tells you whether your actions are right or wrong.” While this article isn’t intended to be an English lesson, this term is at the very heart of the ongoing discourse in the world of Artificial Intelligence. When the subject of Artificial Intelligence-powered robots being granted rights turns up, the community is almost always divided. Online discussion websites and threads are mostly divided into two specific groups: Why should AI-powered beings, which are made to ease man’s burden be thwarted? How can man-made creations which are not a part of the evolutionary process have a conscience, and thus need rights? Here we shall dissect the questions and understand the pros and cons of granting robots rights as well as an opportunity to justify its actions. The Sophia Quandary: The now famous humanoid robot Sophia, created by the Hong Kong-based company Hanson Robotics, was granted citizenship in Saudi Arabia in October this year. But many people were offended, even outraged by it, because now the robot had more rights than women in the country which is believed to be the birthplace of Islam. But technically, if a person (or creation) is a “citizen” of a particular country, won’t he\/she have rights there? The right to free will, the right to life, and maybe even the right to vote? Benjamin Kuipers, professor of computer science and engineering at the University of Michigan had a unique take on the predicament: “A human being is a unique and irreplaceable individual with a finite lifespan. Robots (and other AIs) are computational systems, and can be backed up, stored, retrieved, or duplicated, even into new hardware. A robot is neither unique nor irreplaceable. Even if robots reach a level of cognitive capability (including self-awareness and consciousness) equal to humans, it is not at all clear what this means for the ‘rights’ of such ‘persons’. We already face, but mostly avoid, questions like these about the rights and responsibilities of corporations. A well-known problem with corporate ‘personhood’ is that it is used to deflect responsibility for misdeeds from individual humans to the corporation.” But Professor Kuipers’ statement doesn’t take into account the very human nature to anthropomorphize robots because of popular psyche, and basically an inherent nature to relate and protect. Clouding The Judgement Without Context: The word ‘right’ is being flung about this issue without much thought put into it. Because one of the key things lacking in the discourse of robot rights is the context. Robots, or any AI-powered beings for that matter, are made with the intention of helping out humans in one way or the other. For example, if, in a country like India, robots are built to clean sewers — a job that greatly violates many human rights in one go — would it be immoral to use them? On the other hand, what if a baby-shaped robot built to help expecting mothers is made to clean the same sewer? Would that be OK? MIT Media Lab researcher and robot ethics expert Kate Darling, argued in her paper ‘Extending Legal Rights to Social Robots’ that one of the key reasons to talk about rights for AI-powered beings was to “protect societal values”. She has used the example of parents who tell their child not to kick a robotic pet. Of course, one can assume it’s just a toy and that the parents don’t want to spend for another one, but it also reinforces bad behaviour in the child. A kid who kicks a robot dog might be more likely to kick a real dog or another kid. Nobody wants to perpetuate destruction or violence, regardless of who or what is on the receiving end. Professor Hussein A Abbass of University of South Wales-Canberra also has an interesting take on the subject: “Should robots be given rights? Yes. Humanity has obligations toward our ecosystem and social system. Robots will be part of both systems. We are morally obliged to protect them, design them to protect themselves against misuse, and to be morally harmonised with humanity. There is a whole stack of rights they should be given, here are two: The right to be protected by our legal and ethical system, and the right to be designed to be trustworthy; that is, technologically fit-for-purpose and cognitively and socially compatible (safe, ethically and legally aware, etc.)” Is The Finger To Be Blamed For Pulling The Trigger: Recently, a new short film called, Slaughterbots had created waves on social media due to its dystopian depiction of AI-based weapons. It featured tiny killer drones, not unlike the automated bees in a Black Mirror episode. The film revolves around the concept of thwarting “bad people” right at its birth point, that is, on social media, and how things pan out when AI makes decision instead of humans. But the moral question about giving the robots (or drones in the example quoted above) a right to justify their actions also arises. Are the robots to be blamed for the killings — or rather using their ‘intelligence’ to take decision? Or the humans who have programmed it in such a way? Another palatable example can be that of a self-driving car, which is presented with the dilemma of swerving into a crowd of 10 people to save the owner, or crashing into a wall to save the people. The quandary is not only ethical but also commercial: would you buy a car programmed to kill you under certain circumstances?","excerpt":"What is conscience? The Oxford English Dictionary defines the philosophers’ favourite locution as, “The part of your mind that tells you whether your actions are right or wrong.” While this article isn’t intended to be an English lesson, this term is at the very heart of the ongoing discourse in the world of Artificial Intelligence. […]","categories":["IT Services"],"tags":["AI (Artificial Intelligence)","AI drones","AI Robots","driverless cars","drones","humanoid","Robots","saudi arabia","Self Driving Cars","social media","sophia","uae"],"author_name":"Prajakta Hebbar","publish_date":"2017-12-29T07:11:14","publication_year":"2017","word_count":927,"keywords":["AI Robots","saudi arabia","Self Driving Cars","AI drones","sophia","Rust","R","Robots","artificial intelligence","RAG","programming_languages:Rust","Go","drones","AI","social media","GAN","humanoid","uae","programming_languages:R","programming_languages:Go","driverless cars","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","RAG","R","Go","Rust","GAN","programming_languages:R","programming_languages:Go","programming_languages:Rust"],"url":"https:\/\/analyticsindiamag.com\/it-services\/robots-rights-artificial-intelligence\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":35041,"title":"A Day In The Life Of: Fractal Analytics Techie Who Wants To Marry Economics &#038; Data Science","content":"In our column ‘A Day In The Life Of’, we are trying to step into the shoes of the awesome techies from various organisations who are working in Emerging Tech areas like big data, data analytics, artificial intelligence and the internet of things, among others. This week Analytics India Magazine decided to talk to Arushi Mishra a consultant at Fractal Analytics, about her life as a dedicated techie who also an anime fan. An alumnus of Meghnad Desai Academy of Economics, Arushi talked to us about her experience of working in a constantly evolving sector. Describing her true talent as “extracting insights from data” Arushi starts her day at 6.30 AM, digging her way out of the chores to the sound of her favourite tunes. Saying that one of the best things about working at Fractal Analytics was the flexible and “chilled out” work atmosphere, Arushi says, “The focus of our organisation is not how many hours you work, it’s about meeting the deadlines of the project with high-quality work.” A big fan of morning jogs, DIY projects and the cinema, Arushi admits that her work does end up affecting her personal life — in however small a manner. She explains, “If there is a problem with the data that I cannot resolve at that point of time, or the code fails, I feel the need to stop everything and only focus on the issue at hand. That means stretched working hours, forgetting to eat, stress! However, I get compensated with good sleep and a day off if I manage to resolve the issue. I have a very appreciative and understanding team.” When asked about what her work consists of, Arushi promptly replies with, “A bit of almost everything!” She explains, “My work consists of data extraction, cleaning, harmonisation, reporting, and providing insight as well as helping the clients by answering the specific business question. We use a lot of different tools like Knime, SQL, R; and visualisation tools like Spotfire and Tableau for dashboarding.” Arushi is currently working on reporting the FMCG market for the entire APAC region and advising the clients on a go-market strategy based on consumer behaviours and market performance. “We also aim at continuously developing new solutions to automate the processing of data to find insights. There is a lot of focus on learning different skills and techniques throughout the year to keep ourselves accustomed to what’s new in the industry,” she says. When asked whether her current employers were making full use of her talent, she said, “Fractal Analytics is giving me an opportunity to constantly challenge myself. Although my true expertise is in generating insights. Fractal made me realise how important it is to ensure the quality of data as well as accuracy; how to read data; gain business understanding; understanding the measures available. Once I gain expertise in one kind of skill, I am being tested at a new skill.” However, Arushi, who is an alumnus of the Meghnad Desai Academy of Economics, does feel like she has more to offer her company. “When it comes to data science and analytics, you can never be at the flatter part of the learning curve. I am yet to show my company how well I deliver insights based on forecasts and machine learning,” she says. “My goal at the moment is to complete my training in Machine learning. As my background is in Economics and data analytics, I need to gain technical expertise to separate myself from IT experts in my company,” she says, signing off.","excerpt":"In our column ‘A Day In The Life Of’, we are trying to step into the shoes of the awesome techies from various organisations who are working in Emerging Tech areas like big data, data analytics, artificial intelligence and the internet of things, among others. This week Analytics India Magazine decided to talk to Arushi […]","categories":["AI Features"],"tags":["Data Analytics","Data Science","Data Science Career","Fractal Analytics","Interviews and Discussions"],"author_name":"Prajakta Hebbar","publish_date":"2019-02-15T11:50:35","publication_year":"2019","word_count":592,"keywords":["Fractal Analytics","data science","Go","big data","artificial intelligence","machine learning","AI","Data Science Career","Aim","analytics","SQL","Data Analytics","Data Science","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","data science","analytics","Aim","R","SQL","Go","big data"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/a-day-in-the-life-of-fractal-analytics-techie-who-wants-to-marry-economics-data-science\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10166165,"title":"TiDB Cloud Redefines Data Management for India’s Digital Future","content":"India’s digital economy is at an inflection point. With over 850 million internet users and a projected $1 trillion digital economy by 2030, data is the new oil—and the pipelines carrying it must be more efficient than ever. From fintech unicorns processing millions of transactions per second to AI-driven analytics reshaping consumer insights, Indian enterprises are facing an unprecedented surge in data workloads. But here’s the challenge: Traditional databases—whether monolithic or even some cloud-native options—are struggling to keep up. High scalability demands, real-time analytics, and operational resilience are no longer optional; they are the cost of doing business in India’s hyper-competitive market. This is where TiDB Cloud steps in. A cloud-native distributed SQL database that combines transactional efficiency with real-time analytical capabilities, TiDB Cloud helps Indian enterprises scale elastically, analyze data in real time, and reduce operational complexity. Why does this matter? Because India’s next decade of innovation will be data-first and businesses need a cloud database that can handle scale without sacrificing performance. TiDB Cloud is designed to be that foundation. TiDB Cloud is a fully managed Database-as-a-Service (DBaaS) that offers the capabilities of TiDB, an open-source, distributed SQL database, in a cloud environment. TiDB Cloud has two variants: TiDB Cloud Dedicated  and TiDB Cloud Serverless. TiDB Cloud Dedicated is ideal for organizations with predictable workloads and demanding requirements for performance, security, and reliability. It provides dedicated resources and consistent performance, ensuring optimal operation for critical applications. On the other hand, TiDB Cloud Serverless is tailored for startups, developers, and businesses with dynamic or unpredictable workloads. It offers auto-scaling capabilities, allowing resources to be automatically adjusted based on demand, optimizing cost and performance. TiDB serverless adopts pay as use model and provides cost efficient solutions. Currently, TiDB is working on a next generation database architecture in addition to its current offering, that would introduce groundbreaking architecture for customers of any scale. The architecture will bring dynamic scalability and leverage S3 as one of its data layers. It will provide highest “9”s availability for your applications. Imagine a next generation architecture that can be like serverless but can now be deployed within your premise, thus having 100% control of your data, leverage all the classics of serverless architecture and more importantly, deploy it anywhere you feel like it. This will be a game changer in the database industry and TiDB will be at the forefront of innovation. Stay tuned. As India’s digital economy surges, businesses need a database that scales effortlessly, delivers real-time insights, and ensures high availability. TiDB Cloud meets these demands with its cloud-native, distributed SQL architecture, eliminating manual sharding and enabling seamless scaling of compute and storage nodes. Unlike traditional databases, TiDB Cloud adapts dynamically to workload spikes, ensuring uninterrupted performance during events like UPI transaction surges or festive sales. With automated failover, daily backups, and multi-AZ replication, it guarantees high availability and minimizes data loss risks, making it ideal for mission-critical industries like fintech and e-commerce. TiDB Cloud also ensures ACID compliance with strong consistency, maintaining data integrity across distributed environments. Advanced security features—including encryption, network isolation, and access controls—support compliance with SOC 2 Type 2, ISO 27001, GDPR, HIPAA, and PCI DSS, making it a trusted choice for regulated sectors. TiDB in India TiDB is dedicated to accelerating its growth in India, targeting a 5x increase in revenue over the next three years. This vision is underpinned by expanding the local team, fostering collaborations with cloud technology leaders, and enhancing support for the developer community through education and engagement initiatives. In India, TiDB has already demonstrated its scalability and efficiency. Flipkart, one of the largest e-commerce companies in India, implemented TiDB to support their mission critical applications. Flipkart achieved over 1 million QPS in read throughput and over 100,000 QPS in write throughput both with less than 10ms latency, showcasing TiDB’s superior scalability and efficiency. Delhivery, a leading logistics company in India, leveraged TiDB’s HTAP for real-time analytics, achieving substantial performance gains and efficient data migration. TiDB enabled Delhivery to handle thousands of requests per second with less than 100ms latency in their production environment. Beyond India, TiDB has achieved remarkable success globally. Databricks experienced an 8x increase in application performance with TiDB’s scalable, consistent database solution. Pinterest saw an 80% cost reduction and enhanced latency performance by consolidating their tech stack with TiDB. Scale Without Limits at TiDB User Day India 2025 TiDB is hosting its inaugural TiDB User Day India in Bengaluru on 26th March 2025, which brings together technology leaders, database professionals, developers, and tech enthusiasts to explore the latest advancements in distributed databases.  Bhanu Jamwal, Head of India Business at TiDB, shared the company’s vision for the India market on the sidelines of the TiDB User Day India. “TiDB is dedicated to accelerating its growth in India, targeting a 5x increase in revenue over the next three years. This vision is underpinned by expanding our local team, fostering collaborations with cloud technology leaders, and enhancing support for the developer community through education and engagement initiatives,” said Bhanu Join TiDB User Day India 2025 to explore real-world use cases, discover the latest trends, and learn best practices from the TiDB user community. Whether you’re new to TiDB or an experienced user, this event will help you connect with peers and get inspired by innovative solutions leveraging NewSQL databases. Date: Wednesday, 26 March 2025Time: 09:00 AM – 07:30 PMVenue: JW Marriott Hotel, Bengaluru Why Attend? Hear from real-world users who have leveraged TiDB for business growth and innovation. Explore TiDB’s architecture, advanced features, and best practices with experts. Gain insights from industry leaders on the future of distributed databases.Network with TiDB users, PingCAP engineers, and peers to exchange experiences. TiDB User Day is for technology leaders, database professionals, and innovators exploring distributed databases. Don’t miss out—Register Now.","excerpt":"A cloud-native distributed SQL database that combines transactional efficiency with real-time analytical capabilities, TiDB Cloud helps Indian enterprises scale elastically, analyze data in real time, and reduce operational complexity.","categories":["AI Highlights"],"tags":["cloud adoption in emerging markets","Cloud AI","cloud analytics","Cloud Data AI","cloud native database","Data Management","Fintech and ecomm","Hybrid cloud solutions","Multiagent AI","performance","Real time analytics","Serverless architecture","Tidb","TiDB Cloud"],"author_name":"Siddharth Jindal","publish_date":"2025-03-18T09:30:00","publication_year":"2025","word_count":964,"keywords":["Real time analytics","Scala","Rust","R","cloud native database","Tidb","Fintech and ecomm","cloud analytics","RAG","analytics","performance","TiDB Cloud","AI","ML","Data Management","Cloud Data AI","Databricks","Multiagent AI","cloud adoption in emerging markets","Serverless architecture","serverless","Hybrid cloud solutions","SQL","Cloud AI"],"extracted_tech_keywords":["AI","ML","analytics","RAG","serverless","Databricks","R","SQL","Rust","Scala"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/tidb-cloud-redefines-data-management-for-indias-digital-future\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":6,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10123378,"title":"Snowflake Looks to Upskill Developers in India’s Rural Towns","content":"Snowflake is making significant strides in India, with a vision that transcends traditional go-to-market strategies. AIM caught up with Vijayant Rai, the managing director of Snowflake India, at Snowflake’s Data Cloud Summit 2024. Rai elaborated on the company’s multifaceted approach to establishing a robust presence in the country. “We’re looking at India in a multi-dimensional way,” Rai said, underscoring the company’s diverse operations. This includes a significant presence in Pune, where a team of 500 professionals handles operations and support. Additionally, Snowflake is leveraging India as a hub for global customers through its Global Capability Centers (GCCs). “It’s not just about go-to-market; it’s also about what we can do for global customers from India,” he explained. Snowflake aims to be an integral part of this transformation by partnering with enterprises and SMBs to drive innovation. The company’s engagement with India is driven by the country’s rapid economic growth and digital transformation. “India is the fastest growing global economy, with a growth rate of over 7%,” he said. Rai noted that the widespread adoption of digital public goods like the UPI framework has positioned India as a data-driven economy. The AI in India Approach Snowflake’s approach to AI is pragmatic, focusing on laying strong data foundations. “There’s no AI strategy without a data strategy,” Rai asserts. The company is helping enterprises break data silos and establish robust data governance frameworks. This is essential for leveraging advanced technologies like generative AI, which Rai believes will revolutionise how businesses operate. Snowflake is launching boardroom-level workshops to support this transformation and educate senior management on devising effective data strategies. “We believe it’s part of our charter to educate the market,” Rai said. These workshops are designed to ensure that enterprises can maximise the potential of AI and other emerging technologies. Snowflake is also addressing the demand for skilled developers by offering extensive training and certification programs. These initiatives extend beyond major cities to Tier-2 and Tier-3 locations, even small villages in India, reflecting Snowflake’s commitment to democratise access to AI. “India is front and centre in our strategy,” Rai affirmed, highlighting Snowflake’s dedication to making a meaningful impact in the country. The Tech Talent Prowess One of Snowflake’s key initiatives is to leverage India’s talent and technological prowess. “We see India not just as a support hub but as a centre for innovation, especially in AI and data-driven technologies,” Rai stated. He highlights the significant role of GCCs, with over 600,000 tech professionals in India driving data innovation. Snowflake is committed to supporting these centres to scale and enhance their capabilities. The company is also focused on nurturing the developer community in India. “We’re investing heavily in skilling and ensuring that people are exposed to the Snowflake platform and various aspects of data and AI,” Rai said. This includes initiatives like language support in their large language model, which accommodates all major Indian languages. Teams from Pune and other locations are already contributing to Snowflake’s global projects, and this collaboration is set to deepen over time. In terms of market strategy, Rai emphasised the importance of understanding local business cultures and nuances. “We have experienced teams in Delhi, Bengaluru, and Mumbai who have worked in various verticals and understand the unique needs of different industries,” he said. This local expertise is crucial in navigating the fast-paced technological changes that Indian enterprises are embracing. Snowflake’s Shift to Generative AI Paying Off Snowflake has been on an acquisition spree and much of its focus is on expanding its generative AI capabilities. Ever since Sridhar Ramaswamy joined as the CEO of Snowflake after the acquisition of Neeva, generative AI has been one of its biggest focuses. In an exclusive interview with AIM, Snowflake head of AI Baris Gultekin said that he had worked with Ramaswamy for over 20 years at Google, and called him an incredible leader. “Sridhar brings incredible depth in AI as well as data systems. He has managed super large-scale data systems and AI systems at Google,” Gultekin said. In addition, Microsoft announced an expanded partnership with Snowflake, aiming to deliver a seamless data experience for customers. As part of this, Microsoft Fabric’s OneLake will now support Apache Iceberg and facilitate bi-directional data access between Snowflake and Fabric. Moreover, in a recent interview, Ramaswamy revealed that the cloud data company plans to deepen its collaboration with AI powerhouse NVIDIA. “We collaborated with NVIDIA on a number of fronts – our foundation model Arctic was, unsurprisingly, done on top of NVIDIA chips. There’s a lot to come, and Jensen’s, of course, a visionary when it comes to AI,” Ramaswamy said.","excerpt":"“We believe it’s part of our charter to educate the market,” said Vijayant Rai.","categories":["AI Features"],"tags":["Developers","Interviews and Discussions","Snowflake"],"author_name":"Mohit Pandey","publish_date":"2024-06-12T16:51:18","publication_year":"2024","word_count":765,"keywords":["Go","API","AI","ML","Git","Interviews and Discussions","RAG","Aim","generative AI","R","Developers","Snowflake"],"extracted_tech_keywords":["AI","ML","generative AI","Aim","RAG","Snowflake","R","Go","Git","API"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/snowflake-looks-to-upskill-developers-in-indias-rural-towns\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10059761,"title":"These modern researches aim to make AI similar to human intelligence","content":"People feared that one day, machines would overtake humans and seize control of everything in the past. However, irrespective of the fear, there has been ground-breaking research in Artificial General Intelligence (AGI), which makes artificial intelligence more human-like. The human brain comprises large sets of elements, and it self-organises the dynamical structures to respond with our bodies. Therefore, it has been a natural way to work on AGI to make it adaptive self-organisation. Another way that is used to build AGI models is computational neuroscience. It helps create a model of how the brain works and deploy it in the AGI system. If we knew everything about how our brains function, creating the world’s first human-level AGI would have been much easier. However, it is easier said than done. There have been several types of research in modern AGI that slowly but surely, work towards building machines that are capable of behaving like us humans. Some of the most noteworthy research in the field of AGI is listed down below. Adaptive Control of Thought-Rational (ACT-R) by John Robert Anderson and Christian Lebiere: is a symbolic system that uses connectionist-style activation spreading in a significant role. A blend of SOAR-style “production rules” with large-scale dynamics simulates psychological phenomena. ACT-R can capture how humans solve complex problems, for example, the Tower of Hanoi and algebraic equations. It is also used to mimic human behaviour in driving and flying. Cyc by Douglas Lenat and Ramanathan Guha: It is an AGI architecture depending on logic as a knowledge representation while using logical reasoning to answer questions and derive knowledge. It has a large commonsense knowledge database that Cycorp has collected over the years and has millions of pieces of knowledge entered by trained humans in predicate logic format. Cyc mainly focuses on implicit knowledge, and one of its most notable works came after it was applied in Terrorism Knowledge Base. It contained information on terrorist groups, their members, leaders, ideology, founders, sponsors, affiliations, facilities, locations, finances, capabilities, intentions, behaviours, tactics, and full descriptions of specific terrorist events. After accumulating it, it is stored as mathematical logic statements, suitable for computer understanding and reasoning. EPIC by Rosbe, Ronald Chong, and David Kieras: This has extensively focused on cognitive architecture for capturing human perceptual, cognitive and motor activities through interconnected processors. It is controlled by production rules for cognitive processors and perceptual (visual, auditory, tactile) and motor processors and operates on coded features rather than raw sensory data. Semantic Network Processing System (SNePS) by Stuart Shapiro: It is a logic, frame and network-based knowledge representation. The reasoning and acting system have undergone three decades of development. It has been used in various prototype experiments for language processing and virtual agent control. In addition, SNePS is widely used in cognitive robotics. SOAR by John Laird, Allen Newell, and Paul Rosenbloom: It is a classic example of expert rule-based cognitive architecture designed to model general intelligence. It has recently handled sensorimotor functions and reinforcement learning. This incorporates sensory inputs, which provides information on one’s body and external environment to shape motor output. One of its most noteworthy implementations has been on two major U.S. tactical air missions co-developed by the University of Michigan and Information Sciences Institute (ISI) of the University of Southern California. It simulated close-air support, strikes, CAPs, refuelling, and SEAD missions. It has also been used to develop AI in several real-time strategy games for spatial reasoning and opponent anticipation. A couple of the most famous games which integrated their AI are Minecraft and Mario. Hierarchical Temporal Memory (HTM) by Jeff Hawkins and Sandra Blakeslee: This is a hierarchical temporal pattern recognition architecture, presented as both an AI\/AGI approach and a model of the cortex. Today, it has been used exclusively for vision processing; however, a conceptual framework is outlined to extend action and perception\/action coordination. SAL by David Jilk and Christian Lebiere: SAL is based on IBCA (Integrated Biologically-based Cognitive Architecture), a large-scale emergent architecture that distributes processed information to the brain – especially the posterior and frontal cortex and the hippocampus. It has been used to simulate human psychological and psycholinguistic behaviours but cannot accommodate higher-level behaviours such as reasoning. NOMAD (Neurally Organised Mobile Adaptive Device) by JL Krichmar: NOMAD is inspired by Gerald Edelman’s “Neural Darwinism” brain model. It shows the features of simulated neurons that have evolved through natural selection and configured to carry out sensorimotor and categorisation tasks.","excerpt":"There have been several types of research in modern AGI focussed on building machines that are capable of behaving like us humans.","categories":["AI Features"],"tags":[],"author_name":"Akashdeep Arul","publish_date":"2022-02-04T13:00:00","publication_year":"2022","word_count":738,"keywords":["Go","API","artificial intelligence","TPU","AI","BERT","llm_models:BERT","ViT","GAN","R"],"extracted_tech_keywords":["AI","artificial intelligence","TPU","R","Go","API","BERT","GAN","ViT","llm_models:BERT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/these-modern-researches-aim-to-make-ai-similar-to-human-intelligence\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":24654,"title":"Govt To Use Artificial Intelligence For Traffic Management In Delhi","content":"Come 2019, the Delhi traffic police will have much easier lives, thanks to artificial intelligence as the Indian capital is set to have its own intelligent traffic management system (ITMS) soon. According to a news report, the Ministry of Home Affairs has officially accepted the proposal sent for the same by Delhi Traffic Police. Reportedly, the news traffic management system will work on a radar-based monitoring with the help of AI. The police analyse the traffic pattern, volume, number of vehicles, and other factors, and collect them on a cloud. The data then, would be used to manage the traffic — one of the key tools being the automated traffic signals. Dependra Pathak, special commissioner of Police, Traffic, told the newspaper: “The aim is to reduce the manual interface to minimal. Smart traffic signals, AI to determine the flow of traffic, automated enforcement and communication to change the face of the traffic situation in Delhi… Ideally a traffic official on the road would leave the carriageway opened for equal minutes in order to ensure smooth flow of traffic. In such a scenario not all carriageways have heavy volume. With AI coming in place, the signals would work according to the volume of traffic on each road.” The new project, which will cost around ₹1,000 crore, will aim to overhaul the existing traffic system in Delhi. It will be implemented in three phases. As of now, the organisation\/s which are to take up the project have not been named by the authorities. This is not the first time that artificial intelligence is being used to monitor traffic situation in India. Earlier this year, Piyush Goyal, the Union Minister for Railways and Coal (and Finance, for the time being), had announced that the Indian Railways would use AI to better safety, passenger amenities, revenues, growth and efficiency. The ministry is also looking at using AI-based chatbots to handle passenger queries.","excerpt":"Come 2019, the Delhi traffic police will have much easier lives, thanks to artificial intelligence as the Indian capital is set to have its own intelligent traffic management system (ITMS) soon. According to a news report, the Ministry of Home Affairs has officially accepted the proposal sent for the same by Delhi Traffic Police. Reportedly, […]","categories":["AI News"],"tags":["Artificial Intelligence India","traffic management"],"author_name":"Prajakta Hebbar","publish_date":"2018-05-16T09:22:19","publication_year":"2018","word_count":318,"keywords":["Go","API","artificial intelligence","programming_languages:R","AI","chatbots","Artificial Intelligence India","programming_languages:Go","Aim","GAN","traffic management","R"],"extracted_tech_keywords":["AI","artificial intelligence","Aim","chatbots","R","Go","API","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/delhi-artificial-intelligence-traffic-management\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":33847,"title":"Microsoft Launches Asia’s Largest AI And IoT Lab In Shanghai","content":"Microsoft will soon open their largest artificial intelligence (AI) and the Internet of Things (IoT) lab in Shanghai, in a bid to target China’s growing business sectors ranging from manufacturing to healthcare. The new facility which will be opened in April, is located in Zhangjiang Hi-Tech Park, an area which is turning itself into the country’s AI heartland. Spread across 2,800 square metres, the new lab will be the Microsoft’s first ever in Asia-Pacific and will be the third, dedicated exclusively for research and development for AI and IoT. Stating that the new lab will facilitate increased productivity for Chinese business, Roan Kang, vice president of Microsoft China, said to an online portal, “China is the world’s largest IoT market, with great potential in deploying AI and IoT technologies.” The move which is a collaborative effort between Microsoft and Pudong district government aims to assist enterprises to boost their digital transformation, integrated AI, IoT and help its users develop IoT products and solution across multiple sectors. Other services like rapid design, prototyping, testing and marketing will also be offered to its partner enterprises. The development also comes in the wake of Made in China 2025 plan adopted by the Chinese government. Introduced in 2015, the plan focuses on upgrading the manufacturing capabilities of Chinese industries and increase the Chinese-domestic content of core materials to 40 per cent by 2020 and 70 per cent by 2025. China is will begin large-scale 5G tests and trials this year, which will fuel development not only of telecommunications but also IoT and industrial applications, says Industry and Information Technology Minister Miao Wei. Currently, China is the leading the AI adoption wave worldwide and has become a hotspot for investment and financing from major companies. With home-based companies like Tencent, Baidu, and Alibaba levelling-up their AI capabilities and expanding their reach beyond Asia, tech giants like Google and Microsoft are finding it a hard find a strong base in China.","excerpt":"Microsoft will soon open their largest artificial intelligence (AI) and the Internet of Things (IoT) lab in Shanghai, in a bid to target China’s growing business sectors ranging from manufacturing to healthcare. The new facility which will be opened in April, is located in Zhangjiang Hi-Tech Park, an area which is turning itself into the […]","categories":["AI News"],"tags":["AI in china","Asia","IoT"],"author_name":"Akshaya Asokan","publish_date":"2019-01-22T07:25:54","publication_year":"2019","word_count":325,"keywords":["Go","API","artificial intelligence","Asia","AI","programming_languages:R","digital transformation","Git","Aim","ViT","AI in china","R","IoT"],"extracted_tech_keywords":["AI","artificial intelligence","Aim","R","Go","Git","API","ViT","digital transformation","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/microsoft-launches-asias-largest-ai-and-iot-lab-in-shanghai\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":67833,"title":"Why Do Large Companies Open Source Their Tech?","content":"About a month ago, Google open-sourced its pre-trained models and fine-tuning code for Big Transfer (BiT), a deep-learning computer vision model. According to Google, Big Transfer will allow anyone to reach state-of-the-art performance on their task of interest, even with just a handful of labelled images per class. It is not only one instance when the tech giant has opened its proprietary work for the public free of cost to use. Such powerful releases of free open source software have become a frequent event in the world of technology and beg the question of what large tech companies get in return? In the late 90s, when the Open Source Initiative came into existence, the thought of making source code available to the public would have been considered a bad strategy for technology companies, as proprietary software was the standard, and companies did everything to guard the software. Coming forward to 2020, the concept of open-source has changed drastically and now it has become mainstream. There are numerous open-source tech companies today, some of which have gone beyond $100 million (or even $1 billion) in their annual revenue including RedHat, MongoDB, Cloudera, MuleSoft, Hashicorp, Databricks (Spark) and Confluent (Kafka). Apart from high profile acquisitions and investments in open-source projects by tech companies such as the above, technology giants like Google and Facebook have given open-source an incredible amount of importance for gathering new product innovation and building a huge developer community. For example, Flutter vs React Native, Tensorflow vs PyTorch, Kubernetes, and others. Google and Facebook are more developers and consumer-focused contributions in open-source, whereas Microsoft and AWS are more enterprise-focused. Open-Source Is The Catalyst For Continuous Innovation Open-source is a hub of continuous innovation. When you have more people, over a broad range of companies and backgrounds, your coverage for technology innovation grows exponentially. The company recognises potential issues in their code from various angles, but you also understand how to solve the problems using crowd intelligence and open contributors. For open-source, people can use the system that you build, and some of them will provide improvements. If the main team behind the open-source system is the tech company itself, then the direction of the project will be what they want, while outsiders make a large amount of contribution to the project. It has been found that technology projects tend to accumulate many work-hours which is hard to do when the project is kept in-house. Open-sourcing projects and making them available for outside users makes it more mature in the long term. The company still keeps the main workforce that contributes to the project, and they have a major role in managing new contributions. Additionally, it gives them the chance to find out if there are other individuals that they might hire or, in general, collaborate with in future. This is especially good for the research community, where the company is one part of that community. “Open-source is an enabler of innovation, giving organisations access to a global pool of talent and the tools to develop secure, reliable and scalable software – fast. The organisations that are most effectively speeding up business transformation are those who have turned to open-source software development to succeed in a fast-changing, digital world,” told Maneesh Sharma, General Manager of Github India in an interview with Analytics India Magazine. Open Source Helps Build A Rich Developer Community If companies don’t have open-source in their DNA, it could pose a massive challenge to building positive relationships with their developer communities. If you’re a big corporation like Microsoft, IBM or Google, you have tons of people with their eyes on you at all times. Everyone can read and criticise your code. Also, being open-source means being open and transparent about release cycles and roadmaps, which takes a lot of effort and initiative. With an open-source project, the system gets used by others which helps them to establish future projects and products against other companies. It helps them to have a better brand and others respect them more in that way. Also, their products might be based on the matured software version, which means that these products will have a better value. In a sense, the tech company can be assured that the product is based on some system that many other people use and so you will not end up having any unexpected problem or that problem will be fixed immediately. “Today, 99% of software projects are built using open-source. Open source has revolutionised software development, and created an interconnected community of developers that is deeply collaborative and extends across the world,“ Maneesh added. Companies No Longer Have To Reinvent The Wheel Every Five Years Operating in open-source pushes a company to be a leading member in how the technology is built and how it is evolved. This is done by collaborating and contributing back to the open-source software. Companies whose teams lead open source projects gain greater insight into how the technology is built, and where the technology is going and why. To elaborate, if a company has open-sourced its cloud\/cloud-native projects, then it gains visibility into what the future of cloud and cloud-native will be. This ensures that it is a part of the always advancing tech landscape. This also means you don’t have to reinvent the wheel each and every time you create a new product. As the tech pushes innovation, the market comes to expect certain features, and a company’s differentiation happens on top of that. Open-Source May Not Always Be Preferred Strategy For Tech Businesses But that’s not always the case with open-source strategy. Open-source is undoubtedly profitable in a specific business use case such as RedHat. But in the case of AI, which is usually both advanced and unpredictable, it is challenging for researchers to monetise on their idea and fund research. For example, OpenAI, which started as a non-profit research lab, shifted its stance on the open-source when it partnered with Microsoft. A year later, they have said that they are making all their software innovations available through commercial APIs as opposed to open-source. The organisation previously expressed worries about the potential misuse of its next-generation model. Also, while large companies such as Apple or Microsoft have become very open-source friendly, only a limited percentage of their software products are open-source, including libraries and tooling of course. Compared to other companies, which is a small part of the whole ecosystem. “Ultimately, what we care about most is making sure that artificial general intelligence benefits all. We see developing commercial products as one of the ways to make sure we have adequate funding to succeed,” OpenAI said in a recent blog post about the release.","excerpt":"About a month ago, Google open-sourced its pre-trained models and fine-tuning code for Big Transfer (BiT), a deep-learning computer vision model. According to Google, Big Transfer will allow anyone to reach state-of-the-art performance on their task of interest, even with just a handful of labelled images per class. It is not only one instance when […]","categories":["IT Services"],"tags":[],"author_name":"Vishal Chawla","publish_date":"2020-06-22T10:00:00","publication_year":"2020","word_count":1114,"keywords":["OpenAI","AI","PyTorch","AWS","computer vision","RAG","analytics","Kafka","TensorFlow","kubernetes"],"extracted_tech_keywords":["AI","computer vision","analytics","OpenAI","TensorFlow","PyTorch","RAG","AWS","kubernetes","Kafka"],"url":"https:\/\/analyticsindiamag.com\/it-services\/why-do-large-companies-open-source-their-tech\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10042691,"title":"Meet Meerkat, A New Data Library To Wrangle Complex ML Datasets","content":"Recently, researchers at Stanford University launched a new data library called Meerkat for working with complex machine learning datasets. The source code of the project is available on GitHub. Data is the oxygen for machine learning. From training and validation data to future predictions, embeddings and metadata, it drives all parts of the machine learning development process. However, organising and managing data is challenging. To that end, Stanford researchers have proposed a new Python library to help researchers and ML practitioners wrangle data. Data wrangling is a process of cleaning and unifying messy and complex datasets for easy access and analysis. How does Meerkat work? In a Notion Press blog, ‘Meerkat: Datapanels for machine learning,’ Stanford researchers Sabri Eyuboglu, Arjun Desai and Karan Goel talked about a few areas where Meerkat could solve the data complexity in the machine learning lifecycle. Dataset manipulation techniques like slicing, shaping and transforming datasets have become an increasingly important part of the development process. As the quality of machine learning models and evaluations are primarily products of the data, more time goes into tuning datasets than tuning models. Model evaluation is emerging as a new bottleneck when building high-performing ML systems. For instance, models have been commoditised to the extent that resources like HuggingFace’s Model Hub can give you a model for text, speech or vision in seconds. But, they are hard to get right, and their failure modes can be opaque. Multi-modal datasets that combine multiple, complex data types are becoming more prevalent. For example, OpenAI’s CLIP combines natural language with images. Meerkat provides the DataPanel abstraction. The DataPanel facilitates interactive dataset manipulation, where it can house diverse data modalities and lets you evaluate models carefully with Robustness Gym. “We built DataPanels like DataFrames because they are naturally interactive and work seamlessly across development contexts: Jupiter Notebooks, Python scripts, and Streamlit,” the researchers said. The goal is to make Meerkat DataPanel an interactive data substrate for modern machine learning across the machine learning lifecycle. Besides Robustness Gym, Meerkat can also be integrated into other popular benchmark datasets and works well with existing libraries and tools like WILDS, Huggingface Datasets, DOSMA, Streamlit. What makes Meerkat different? The data structures typically fall into two categories: those supporting complex data types and multiple modalities (PyTorch Dataset, Tensorflow Dataset), and those that support manipulation and interaction (Pandas DataFrame). “With the Meerkat DataPanel, we support all of these desiderata in one data structure,” said the researchers. Advantages: Meerkat can store complex data types (images, graphs, videos and time series) Supports datasets that are larger than RAM (Kinetics, MIMIC-CXR, ImageNet) with efficient I\/O under-the-hood  Supports multimodal datasets Supports data creation and manipulation Supports data selection Support inspection in interactive environments Comparing Meerkat with other machine learning data structures (Source: Notion Press) The experiment The researchers ran an experiment to detect pneumothorax (a collapsed lung) in chest X-rays. For developing a model for this task, the researchers encountered various types of data — X-ray images to structured metadata to embeddings extracted from a trained model. Here, Meerkat’s DataPanel (a columnar data structure) could house all these data types under one roof. “Keeping them together enables quicker model iteration, fine-grained error analysis, and easier data exploration and inspection,” said the Stanford researchers. The codes used for running this experiment are available here. Meerkat addressed the desiderata by facilitating the inspection and manipulation of datasets that combine multiple complex data types. Meerkat provides high-level data abstractions because its data structures are written in Python and have few low-level optimisations, unlike Pandas, NumpY or Apache Arrow. “This does not mean that the Meerkat DataPanel is slow: each column type is as fast as the data structure it is built upon,” said the researchers.","excerpt":"Recently, researchers at Stanford University launched a new data library called Meerkat for working with complex machine learning datasets. The source code of the project is available on GitHub.  Data is the oxygen for machine learning. From training and validation data to future predictions, embeddings and metadata, it drives all parts of the machine learning […]","categories":["AI Features"],"tags":["Machine Learning","Machine Learning New"],"author_name":"Amit Naik","publish_date":"2021-06-30T18:00:00","publication_year":"2021","word_count":620,"keywords":["NumPy","machine learning","Machine Learning New","OpenAI","AI","PyTorch","ML","Machine Learning","Ray","Streamlit","TensorFlow","Pandas"],"extracted_tech_keywords":["AI","machine learning","ML","OpenAI","Ray","TensorFlow","PyTorch","Streamlit","Pandas","NumPy"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/meet-meerkat-a-new-data-library-to-wrangle-complex-ml-datasets\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":42295,"title":"After The Blackhole Image, Machine Learning Is About To Make A Movie About It","content":"The first-ever black hole image generated using a machine learning algorithm created a huge buzz around the globe and proved to be a major scientific discovery. This algorithm, called CHIRP which stands for ‘Continuous High-resolution Image Reconstruction using Patch priors’ showed the power of machine learning in producing an image of a black hole, something which would have otherwise not been possible. Now the next big thing is making a black hole movie. There has been research going on that uses the wonders of ML to create a black hole movie. The woman working behind the black hole movie is Lia Medeiros, a physicist, astrophysicist and a National Science Foundation fellow. How Will ML Algorithm Make A Movie? Making a movie would require getting information about what happens in and around the black hole using ML algorithms. This means that a time series of images has to be put together to obtain this movie. It is going to tell how the image of the black hole changes as a function of time. The black hole chosen to obtain this movie lies at the centre of a star called Saggitarrius A* or Sgr A*. This star is about 26,000 light years away from the Event Horizon Telescope (EHT). Sgr A* has an event horizon with a radius of about 7.9 million miles, making it about 18 times the Sun’s diameter. This black hole exerts a tremendous amount of gravitational pull on anything that crosses its event horizon. Such a far away object would need robust images to make a movie out of the black hole at the centre of it, which is why the EHT is used, which was also used in case of obtaining the black hole image. Key Requirements To Accomplish The Task EHT: A large amount from the EHT, which formed an array of telescopes throughout the globe, is used to obtain the movie. The black hole image captured did not have a direct image generated but they were pieces or petabytes of raw data consisting of the radio waves originating from the ring-shaped silhouette of a black hole. El Gato: El Gato, a short form of Extremely LarGe Advanced TechnOlogy cluster is a high-performance computer by the National Science Foundation and the University of Arizona. This computer has special hardware to achieve the high performance that it delivers. It has NVIDIA K20X GPUs and Intel Xeon Phi 5110p Coprocessors. The system is designed around Intel Ivy Bridge CPUs, large amounts of memory per node (256GB\/node), and FDR Infiniband. It is connected using an FDR Infiniband speed to a DataDirect Networks SFA 12K petabyte-scale storage server. This system takes a lot of data and mathematically transforms the input data into an image. The method will involve creating simulations and using them as training data for the algorithm which will eventually be used to create a time series of the black hole images, which is what is finally wanted.  But this collected data is not enough to create a complete picture of the black hole and then make a movie out of it. There is a lot of missing data which has to be filled to complete the image. This is where the ML algorithm is needed. Machine Learning Algorithm: The ML algorithm containing datasets, which is nothing but the images from EHT, may have some information missing. This missing information will be fed to the algorithm. This will also contain simulations. Both the images and the simulations will be used for training the ML model, helping Medeiros in finally creating the time series of images. The ML algorithm that goes behind making the first ever black hole movie is based on the Principle Component Analysis (PCA). PCA is a dimensionality reduction method used to reduce the dimensionality of a large dataset to a small dataset. It is commonly used in the analysis and representation of complex data to reduce a large set of variables to a small set. This small set contains the most information about the large reduce set. PCA could also be seen as an algebra operation that contains a lot of ML applications, which helps in reducing the data into its principal components. Its applications lie in areas where data analysis and the predictive tool is involved using large datasets. Final Output: For the final movie production, PCA algebra operation will be added to the ML algorithm that Medeiros is making. This algorithm is going to learn from the data that it will be fed with. The training set of the ML algorithm will be able to produce 10 to 20 images. The training set can be used to identify images that will act like building blocks of a blackhole image. These building blocks would be put together make an image with the algorithm that will fill the parts of what they do not have the data for. This way, a complete image could be obtained. This new ML algorithm could give the world its first ever black hole movie. Future Ahead Quantum physics and GTR, two of the most important aspects of physics, when clubbed together do not give us the Physics that we have known so far. Black holes form the perfect test bed to study these together as it is large in terms of mass like in GTR and at the same time has an infinite density at its centre governed Quantum physics. All these machine learning-guided physics experiments are here to show how correct or incorrect is Einstein’s GTR. Black hole movies could also lead us to the understanding of how quantum mechanics interact with GTR. They can help with the information needed to check these theories and help with bigger discoveries.","excerpt":"The first-ever black hole image generated using a machine learning algorithm created a huge buzz around the globe and proved to be a major scientific discovery. This algorithm, called CHIRP which stands for ‘Continuous High-resolution Image Reconstruction using Patch priors’ showed the power of machine learning in producing an image of a black hole, something […]","categories":["AI Features"],"tags":["Machine Learning","PCA"],"author_name":"Disha Misal","publish_date":"2019-07-13T15:00:08","publication_year":"2019","word_count":951,"keywords":["Go","machine learning","TPU","AI","ML","Machine Learning","Git","RAG","Ray","ViT","PCA","R"],"extracted_tech_keywords":["AI","machine learning","ML","Ray","RAG","TPU","R","Go","Git","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/after-the-blackhole-image-machine-learning-is-about-to-make-a-movie-about-it\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":19194,"title":"Understanding The State Of Analytics At YES Bank","content":"YES BANK is new age bank, and analytics adoption by users within all business groups has been phenomenal. The Bank has a culture of making informed decision basis available data & information. Analytics works as catalyst providing them the right information using statistical tools which helps the central teams to take appropri­ate decisions. There is a significant demand in all business groups to get maximum information be­fore making strategic decisions. Analytics India Magazine caught up with Anand K Sundaram, Head – Credit Intelligence & Analytics, YES BANK, one of the leading private banks in the coun­try, to understand the state of analytics adoption in the bank, and the ways that it has affected the overall functioning of the bank. This article originally appeared in our annual study titled “State of analytics in domestic firms in India 2017”, carried out in association with Cartesian Consulting. “Our Analytics Team still is at nascent stage of existence since last 2 years; however it has infused its presence significantly in the Retail banking space which is now mainly driven by an­alytics. The journey of analytics from MI visualis­ation to Machine learning is growing in significant pace here at YES BANK”, says Sundaram. Starting with a 5 member analytics team to a 20-member team, Sundaram says that it is still a small unit compared to 21,000 employees of the bank, but largely the users of analytics would be the central teams within different groups, which would be ~100 staff. “Hence, a 20-member team is mainly catering to about 100 central team users”, he marks. On analytics adoption and maturity: While basic analytics existed in the Bank, a dedicated central analytics team was built two years ago. The team has added to the bottom line through various projects, which were implemented successfully. It used a mix of traditional & latest technologies along with statistical tools. Some of statistical tools used include logistic regression, RFM, Clustering, Segmentation, Random forest, scoring algorithm and some of the technology tools includes SQL, R, Tableau, etc. Analytics has slowly started transforming manual process and decisioning to digital algorithm based execution. Several digital changes have streamlined multiple areas within the Bank. “We have adopted analytics in all our marketing campaigns, strategic policy decisioning, simple excel MI to visualisation applications, credit underwriting to scorecard and so on”, he adds. Talking about analytics maturity at Yes Bank, Sundaram says “Analytics has still a long way to go within the organisation where almost every action is driven basis data & information. Analytics would also help the Bank in driving Digital Transformation with the use of algorithms and converting several manual processes to rule based processes. “The team has made a great start and has made its presence felt, but it has a significant task ahead to support our CEO’s vision of becoming finest quality large Bank in India by 2020”, he says in the concluding.","excerpt":"YES BANK is new age bank, and analytics adoption by users within all business groups has been phenomenal. The Bank has a culture of making informed decision basis available data & information. Analytics works as catalyst providing them the right information using statistical tools which helps the central teams to take appropri­ate decisions. There is […]","categories":["IT Services"],"tags":[],"author_name":"Srishti Deoras","publish_date":"2017-11-22T05:01:26","publication_year":"2017","word_count":480,"keywords":["Go","machine learning","ELT","AI","ML","Git","analytics","SQL","GAN","R"],"extracted_tech_keywords":["AI","machine learning","ML","analytics","R","SQL","Go","Git","ELT","GAN"],"url":"https:\/\/analyticsindiamag.com\/it-services\/understanding-state-analytics-yes-bank\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":39212,"title":"Guide To Implement StackingCVRegressor In Python With MachineHack’s Predicting Restaurant Food Cost Hackathon","content":"For someone who has an interest in Data Science, Regression is probably one of the first Predictive Models that s\/he may begin with. Regression is easier to understand and even easier to implement considering all those ready made packages and libraries all set to perform complex mathematical computations effortlessly without leaving his or her brain squashed. Having said that, there are lots of regression models or algorithms that one can use while building Machine Learning applications. But one of the most challenging things is to find a model with the least variances in predictions. Even after going through rigorous steps of model selection to come up with a best fitting model you may still face high variance in output on executing the same model that is tested against the same data multiple times. The StackingCVRegressor Ensemble learning is a technique that combines the skills of different algorithms in predicting for a given sample of data. A technique to combine the best of multiple algorithms which can give more stable predictions with very less variance than what we get with a single regressor.  The StackingCVRegressor is one such algorithm that allows us to collectively use multiple regressors to predict. The StackingCVRegressor is provided by the mlxtend.regressor package in Python. How StackingCVRegressor Predicts The StackingCVRegressor has two levels of regressors, level one regressors and level two regressor and uses a concept called out-of-fold predictions. It starts by splitting the dataset into k folds in k successive rounds and using k-1 folds to fit the level one regressors. The remaining one fold is then used by the first level regressors to predict. These predictions then flow into the level two regressor as input. Once the training is finished the entire dataset is used to fit the level one regressors. Click here to go to the official GitHub for StackingCVRegressor. Using The StackingCVRegressor In Python Let us get our hands dirty by writing a Python code to build a regression model using the StackingCVRegressor. We will be using the data from one of the hottest hackathons on the Internet, Predicting Restaurant Food Cost Hackathon by Machinehack. Getting the Datasets Head to MACHINEHACK’s Predicting Restaurant Food Cost Hackathon by clicking here. Sign Up and start the course. You will find the data set as PARTICIPANTS_DATA_FINAL in the Attachments section. Having trouble finding the data set? Click here to go through this simple tutorial to help yourself. Lets Code! Since this tutorial is to help you with the StackingCVRegressor, I will be skipping all the processes till Modelling (excluding). You can use the following resources to help you till this stage. A Complete Guide to Cracking The Predicting Restaurant Food Cost Hackathon By MachineHack Hands-on Tutorial On Data Pre-processing In Python After following through all the stages till Modelling, You are ready to build the StackingCVRegressor for the dataset in hand. The data sets after preprocessing will look like what is shown below : Creating a New Training Set and Validation Set For validation, we will split the new_data_train into two data sets data_train and data_val. Execute the following code block to generate new dependent and independent variable sets from the newly formed data_train and data_val datasets. from sklearn.model_selection import train_test_split data_train, data_val = train_test_split(new_data_train, test_size = 0.2, random_state = 2) #Classifying Independent and Dependent Features #_______________________________________________ #Dependent Variable Y_train = data_train.iloc[:, -1].values #Independent Variables X_train = data_train.iloc[:,0 : -1].values #Independent Variables for Test Set X_test = data_val.iloc[:,0 : -1].values Applying Feature Scaling #Feature Scaling #________________ from sklearn.preprocessing import StandardScaler sc = StandardScaler() X_train = sc.fit_transform(X_train) X_test = sc.transform(X_test) Y_train = Y_train.reshape((len(Y_train), 1)) #reshaping to fit the scaler Y_train = sc.fit_transform(Y_train) Y_train = Y_train.ravel() RMLSE For Model Evaluation def score(y_pred, y_true): error = np.square(np.log10(y_pred +1) - np.log10(y_true +1)).mean() ** 0.5 score = 1 - error return score actual_cost = list(data_val['COST']) actual_cost = np.asarray(actual_cost) We will first evaluate the performances of individual algorithms on our data set. eXtreme Gradient Boosting ###################################################################### #eXtreme Gradient Boosting ############################################################################ #Importing and Initializing the Regressor from xgboost import XGBRegressorxgbr = XGBRegressor() #Fitting the data to the regressor xgbr.fit(X_train, Y_train) #Predicting the Test set results y_pred_xgbr = sc.inverse_transform(xgbr.predict(X_test)) #Evaluating print(\"\\n\\nXGBoost SCORE : \", score(y_pred_xgbr, actual_cost)) Output: XGBoost SCORE :  0.8132790047309137 Random Forest Regression ###################################################################### #Random Forest Regression ############################################################################ #Importing and Initializing the Regressor from sklearn.ensemble import RandomForestRegressor rf = RandomForestRegressor(n_estimators=100, random_state=1) #Fitting the data to the regressor rf.fit(X_train, Y_train) #Predicting the Test set results y_pred_rf = sc.inverse_transform(rf.predict(X_test)) #Evaluating print(\"\\n\\nRandom Forest SCORE : \", score(y_pred_rf, actual_cost)) Output: Random Forest SCORE :  0.8241776753311784 Linear Regression ###################################################################### #Linear Regression ############################################################################ #Importing and Initializing the Regressor from sklearn.linear_model import LinearRegression lr = LinearRegression() #Fitting the data to the regressor lr.fit(X_train,Y_train) #Predicting the Test set results Y_pred_linear = sc.inverse_transform(lr.predict(X_test)) #Evaluating print(\"\\n\\nLinear Regression SCORE : \", score(Y_pred_linear, actual_cost)) Output : Linear Regression SCORE :  0.7123640090219118 StackingCVRegressor ##################################################################### #Stacking Ensemble Regression ########################################################################### #Importing and Initializing the Regressor from mlxtend.regressor import StackingCVRegressor #Initializing Level One Regressorsxgbr = XGBRegressor() rf = RandomForestRegressor(n_estimators=100, random_state=1) lr = LinearRegression() #Stacking the various regressors initialized before stack = StackingCVRegressor(regressors=(xgbr ,rf, lr),meta_regressor= xgbr, use_features_in_secondary=True) #Fitting the data stack.fit(X_train,Y_train) #Predicting the Test set results y_pred_ense = sc.inverse_transform(stack.predict(X_test)) #Evaluating print(\"\\n\\nStackingCVRegressor SCORE : \", score(y_pred_ense, actual_cost)) Output : StackingCVRegressor SCORE :  0.827546146596208 Conclusion We can see a slight improvement in the score of StackingCVRegressor compared to the individual algorithms, however, this may not always be the case. StackingCVRegressor may turn out to be less or more efficient in terms of accurate predictions than individual algorithms depending on the data and the level one regressor used. One thing which can be certain though is that the predictions from StackingCVRegressor can be deemed stable and is expected to show less variance due to the very fact that it combines the skills of a variety of algorithms.","excerpt":"For someone who has an interest in Data Science, Regression is probably one of the first Predictive Models that s\/he may begin with. Regression is easier to understand and even easier to implement considering all those ready made packages and libraries all set to perform complex mathematical computations effortlessly without leaving his or her brain […]","categories":["Deep Tech"],"tags":["Ensemble Learning"],"author_name":"Amal Nair","publish_date":"2019-05-15T10:34:17","publication_year":"2019","word_count":954,"keywords":["data science","Go","machine learning","TPU","AI","ML","Ensemble Learning","Python","Ray","XGBoost","R"],"extracted_tech_keywords":["AI","machine learning","ML","data science","Ray","XGBoost","TPU","Python","R","Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/stackingcvregressor-in-python\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":57743,"title":"Deep Dive: How Teamlease And Fresherworld.com Leverage AI &#038; ML To Improve The HR Industry","content":"According to reports, on average for each corporate job offer in the world, there are 250 resumes. And of those 250, only 4-6 are called for an interview. The insight seems to be ordinary when one just reads it, but, the point is, on average, there are around 250 CVs for one job. That means 250 people fight for that one position. Part of the reason for so many people finding and applying for the open positions is the advent of AI and ML in the HR industry, at least in today’s world. Some years ago, these online job portals were not in the play, and recruitment and hiring was a tedious offline process, wherein some cases HR managers used to search for candidates offline or through their connections. When the internet became accessible to common people, and many processes became online, the HR industry was quick to follow. There are many players in today’s online job portal market, but one name that is frequently used, especially if one is fresh out of degree college, are Teamlease and its Freshwerworld.com. Teamlease Services was established in 2002 and is India’s leading HR services company in the organised segment. It’s a Fortune India 500 company with over 2500+ clients and 1900 employees in 8 locations across India. Teamlease, in its initial days, started recruiting talents offline, that is, its team used to go door to door in search. But, as time passed, when other sectors started moving online and with mobile phones everywhere, they moved online. Now, through their online presence, they have given employment to over 1.7 million people with the aim of making the number bigger. As the number of applicants and CVs increased overtime, Teamlease also started to upgrade their tech part; now, they make use of highly advanced technology like AI and ML in their HR processes. For our deep dive column, we got in touch with Kaushik Banerjee, Vice President & Business Head, Teamlease and Fresherworld.com to discuss how they leverage AI and ML capabilities. Use Of AI and ML With the increasing use of AI and ML in many business sectors and now HR, Teamlease hasn’t shied away from using the AI and ML capabilities. Banarjee says Teamlease completely takes advantage of AI and ML. The process involves the system carrying the interview process as per the answers given by the candidate. It frames the questions automatically based on CV. The open-ended questions are matched with previously fed data, and there are close-ended questions, which are used for elimination purposes. The algorithm assigns tags based on the CVs and the JD with the percentage of matching criteria to the HR. The ML matches the words and then provides the contextual reference, resulting in enhancing the accuracy of the result. Candidates have to structure their CVs better and have to use industry-related keywords inside them. Apart from the names and contact details, the most critical part of a CV, in this case, becomes the objective summary and key skills with spell checks. Including some keywords and phrases related to one’s industry becomes a point of emphasis if one wants to appear in the search net of AI and ML. Apart from the summary and key skills, other spaces where one can include industry-related keywords are projects and key tasks. Including stats and data concerning the respective work is also recommended. Taking care of all these points helps the managers to choose efficiently and reduce their time, but it will be their risk to depend upon the AI systems. Teamlease believes that these AI and ML capabilities will definitely not eliminate the need for HR but will only help them ease their tasks in the crowded space, where there are millions of job candidates applying every day. The Tech Stack And Teamlease’s Hiring The tech stack with Kaushik’s wing consists of Python, MySQL for the database. PHP7 on the backend and Amazon Web Services for the cloud. The engineering team consists of more than 40 members, and the AI team is around 8. When it comes to hiring a data science candidate, according to Mr Kaushik Banerjee, the data science sector is relatively new, and there is a need for talent in the market. What they look for in a data scientist is primarily in-depth knowledge about the data. An ideal data science candidate for them must be able to comfortably read data, figure out key insights from it and come up with the best strategy from the provided data. The Future Roadmap And What Sets Them Apart From Competition According to Banerjee, what sets Teamlease apart from the competition is their appetite to fulfil the clients’ needs and convert job application to hire as much as possible. They continuously strive to augment their processes with technology and have seen the industry go from 10-12 rounds of interviews to 1 or 2. Now, as for future, with AI and ML capabilities, they envision a future where a candidate will only meet the HR department directly for salary negotiations and not interviews.","excerpt":"According to reports, on average for each corporate job offer in the world, there are 250 resumes. And of those 250, only 4-6 are called for an interview. The insight seems to be ordinary when one just reads it, but, the point is, on average, there are around 250 CVs for one job. That means […]","categories":["AI Features"],"tags":["purposes of a data team"],"author_name":"Sameer Balaganur","publish_date":"2020-03-02T10:00:00","publication_year":"2020","word_count":843,"keywords":["data science","Go","AI","ML","RAG","Python","Aim","SQL","purposes of a data team","GAN","R"],"extracted_tech_keywords":["AI","ML","data science","Aim","RAG","Python","R","SQL","Go","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/deep-dive-how-teamlease-and-fresherworld-com-leverage-ai-ml-to-improve-the-hr-industry\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":50050,"title":"Top 7 Image Processing Libraries In Python","content":"A huge amount of the data collected today is made up of images and videos. That is why effective image processing for translating and obtaining information is crucial for businesses. Data scientists usually preprocess the images before feeding it to machine learning models to achieve desired results. Consequently, it is paramount to understand the capabilities of various image processing libraries to streamline their workflows. In this article, we are listing down the top image processing libraries in Python: 1. Scikit-image Scikit-image uses NumPy arrays as image objects by transforming the original pictures. These ndarrys can either be integers (signed or unsigned) or floats. And as NumPy is built in C programming, it is very fast, making it an effective library for image processing. Among different methods, data scientists often utilise greyscale technique where each pixel is a shade of grey. 2. OpenCV First released in 2000, OpenCV has become a popular library due to its ease of use and readability. The library is focused on image processing, face detection, object detection, and more. It is written in C++ but also comes with Python wrapper and can work in tandem with NumPy, SciPy, and Matplotlib. Backed by more than one thousand contributors on GitHub, the computer vision library keeps enhancing for an effortless image processing. 3. Mahotas Mahotas allows developers to use its advanced features such as haralick, local binary patterns, and more. It can compute 2D and 3D images through mahotas.features.haralick module and perform advanced image processing by extracting information from pictures. Mahotas has over 100 functionalities for computer vision capabilities that can enable you to carry out processes like watershed, morphological processing, convolution, and more. Here’s the link to the documentation and GitHub. 4. SimplelTK Unlike other libraries that consider images as arrays, SimpleITK treats images as a set of points on a physical region in space. The region occupied by images is defined as origin, spacing, size, and direction cosine matrix. This modus operandi enables it to effectively process images. It supports a wide range of dimensions that includes 2D, 3D, and 4D. 5. SciPy SciPy is primarily used for mathematics and scientific computations, but you can also implement algorithms for image manipulation by importing scippy.ndimage module. You can carry out binary morphology, object measurements, linear and non-linear filtering. Besides, one can draw contour lines, adjust interpolation, filter, effects, denoising, and other similar extraction and segmentation on images. 6. Pillow The library is an advanced version of PIL, which is supported by Tidelift. It includes various processes in image processing such as point operations, filtering, manipulating, and more. Pillow also supports a wide range of image formats, thus makes its must-have library for handling images. 7. Matplotlib Matplotlib is mostly used for 2D visualisations, but it can also be leveraged for image processing. Although it does not support all the file formats, Matplotlib is effective in altering images for extracting information out of it. Outlook Image and video processing techniques are rapidly being adopted across the globe due to its many use cases. More recently, Indian Railways is using facial recognition for identifying criminals. Besides, it has also become an integral part of data science and artificial intelligence workflow for gathering information out of images or videos. While we have compiled a few most widely used libraries, there are numerous others in the technology marketplace that can be used for specific requirements. Therefore, you should identify your needs and based on that you can determine the best-fit image processing library.","excerpt":"A huge amount of the data collected today is made up of images and videos. That is why effective image processing for translating and obtaining information is crucial for businesses.  Data scientists usually preprocess the images before feeding it to machine learning models to achieve desired results. Consequently, it is paramount to understand the capabilities […]","categories":["AI Trends"],"tags":["Computer Vision","image processing","machine learning pattern recognition python","Python"],"author_name":"Rohit Yadav","publish_date":"2019-11-17T16:00:00","publication_year":"2019","word_count":582,"keywords":["data science","machine learning pattern recognition python","NumPy","artificial intelligence","machine learning","AI","image processing","ML","computer vision","OpenCV","Python","Ray","Computer Vision","Matplotlib"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","computer vision","data science","Ray","OpenCV","NumPy","Matplotlib"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-8-image-processing-libraries-in-python\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10078301,"title":"Can Phone Makers Replace Your Banks?","content":"When Apple launched ‘Apple Pay’ in 2014, it opened the doors for phone makers to enter the financial services market, a segment which no other phone maker had explored earlier. A year later, Samsung launched ‘Samsung Pay’. Earlier this month, Apple teamed up with Goldman Sachs to launch no-fee, high-yield savings accounts for its Apple Card customers. In September, Samsung launched two credit cards in India in partnership with Axis Band and Visa. While Apple’s financial products are not available in India, Chinese phone making giant Xiaomi launched Mi Pay (now discontinued) in India in 2018 in partnership with ICICI bank. Oppo likewise ventured into the Indian financial market services in 2020 with its subsidiary ‘Oppo Kash’. So, why are phone makers entering the financial services market? (Source: IDC) The digital age Today, almost all financial services are accessible through a smartphone. From opening a bank account, KYC, to applying for a credit card or buying an insurance policy—everything can be done through a smartphone. In less than a decade, hundreds of neobanks have mushroomed across the world and operate without any brick-and-mortar offices. The very concept of banking has evolved in the last few years. With more than 1 billion estimated iPhone users globally, Apple was the first phone maker to seize the opportunity. With smartphone penetration also soaring and more people opting to use their phones to make payments, it makes perfect sense for phone makers to target financial services which have grown exponentially as a result. Phone makers have realised that they have the capabilities to build the softwares and, through a partnership with third parties, extend financial services to their customers. Apple, building on its brand loyalty, has successfully diversified into financial services and established a financial ecosystem. According to an earlier report by Bloomberg, Apple is developing its own payment processing technology, which would reduce its reliance on outside partners like Goldman Sachs to power its financial products. Building an ecosystem The idea for phone makers to enter the financial services market is not to tap into the market but rather to create a digital ecosystem for their users. Ron Shevlin, Chief Research Officer at Cornerstone Advisor, said that Apple is not trying to become a bank but to build an ecosystem. Through its financial services offerings like ‘Apple Pay’ and BNPL services, Apple wants to virtually lock its 1 billion users into its ecosystem and encourage them to use Apple products and services and make payments through their iPhones instead of using other mediums such as their bank’s debit or credit cards. The fact that the younger generations today prefer paying from their phones also plays into their favour. In FY 2021, over 35 billion digital transactions worth over INR 60 trillion were carried out in India. The number of digital transactions is forecasted to reach 214 billion by 2026, according to Statista. Much of the growth is attributed to the advent of technologies such as unified payments interface (UPI). Samsung, Realme, and Oppo phones already come with pre-installed payment apps that support UPI payments. OnePlus is expected to join the club some time this year. Digital payments in India (Source: Statista) Customer retention The smartphone market in India today is highly competitive, with more than 50 different players competing in the space. Keeping their customers locked in their ecosystem through various services will also help phone makers in customer retention. “The key to this growth lies in retaining the customer into a smartphone manufacturer’s ecosystem that fulfils many of the customer’s needs through these services, thereby making it harder for them to switch to another brand.” “As a result, now every smartphone company is trying to push for higher adoption of their services to generate brand loyalty, which is typically low with customers when the product gets commoditised,” Varun Kapoor, Practice Director, Digital Strategy APAC at Adobe, said. Flat sales The financial services market is a lucrative one for phone makers as they look to diversify in the light of smartphone sales remaining flat. During the first half of 2022, the smartphone market in India declined by 1% due to weak consumer demand, according to a report by IDC. In a report, Gartner claimed that global shipment of smartphones is set to decline this year due to China’s slowing economy and an inflation-driven drop in consumer spending. Diversifying into other avenues such as payments, videos and music streaming makes sense for phone makers. “To operate at a good profit margin, now brands really need to focus on the ecosystem which is going to exist around smartphones in the form of new wearables, like bluetooth earphones, smart watches, smart speakers along with services like music, video or movie content,” Anshul Gupta, research director at Gartner, said. Will phone makers compete with banks? With Apple now offering savings accounts, it becomes very difficult to distinguish it from a conventional bank. JPMorgan Chase chief executive Jamie Dimon said that Apple wants to give its users a credit journey experience. “They’re going to do merchant processing; they’re going to do merchant lending. It may not be their own balance sheet. But that’s a bank. If you move money, hold money, manage money, lend money, that’s a bank,” he said. Other phone makers are also adding more and more products to the list of financial services provided by them. Earlier in October 2022, Samsung launched Samsung Finance + in India to allow its users to buy its wide range of consumer electronics products. With phone makers providing more and more financial services products—not only in the Indian market but globally—it has become increasingly difficult to distinguish them from what is traditionally understood as a banking institution. While it’s still early to say, a scenario where phone makers would completely replace your banks would be difficult to write off. Should banks be worried? Maybe, but the regulators are in their favour on this one. Recently, the UK’s Financial Conduct Authority launched an inquiry into Apple’s expansion into payments, deposits, credit and insurance. Other companies under the scanner are Meta, Amazon and Google.","excerpt":"Phone makers want to virtually lock their users into their ecosystem and encourage them to use their products to make payments through their phones instead of using other mediums, such as their bank’s debit or credit cards.","categories":["IT Services"],"tags":[],"author_name":"Pritam Bordoloi","publish_date":"2022-10-31T11:00:00","publication_year":"2022","word_count":1013,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","Git","RAG","Aim","GAN","R"],"extracted_tech_keywords":["AI","Aim","RAG","R","Go","Git","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/can-phone-makers-replace-your-banks\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10167358,"title":"Designing the Right AI Strategy: Aligning with your Mission Rather than Diluting it","content":"AI is transforming business structures, even making several firms rethink their entire business models. However, the challenge lies in building AI that enhances and redefines human roles rather than replacing them, which has been one of the biggest threats and concerns for most employees. Hence, organisations must design AI to complement human expertise, ensuring efficiency and alignment with business objectives and that humans always stay in the loop. The Shift Toward ‘Human + AI’ Collaboration The traditional workforce model is evolving. AI-driven agents, workflows, and processes are now essential for business growth. Companies are moving away from generalised AI and using Vertical AI—AI solutions tailored to specific industries such as healthcare, finance, and manufacturing. This approach ensures domain-specific accuracy, efficiency, and relevance. ValueLabs, for instance, has embraced this transformation. “At ValueLabs, we believe that the future of work lies in the harmonious integration of human and artificial intelligence,” said Veda Reddy, CEO of Imagine, the innovation arm of ValueLabs. “We’ve made significant investments in AI-powered solutions that augment and empower our human teams, allowing us to deliver greater value to our clients,” she added. By aligning AI with their industry-specific needs, businesses are streamlining operations, accelerating go-to-market strategies, and improving overall customer experience. Aligning AI with Company Values AI must align with a company’s core values and guiding principles. AI systems managing critical business processes, from automating workflows to handling customer interactions, must operate ethically and transparently. Misalignment can lead to ethical concerns, regulatory risks, and reputational damage. Whether it is an AI managing social media posts or automating repetitive workflows, its underlying goals, metrics, and decision-making processes must be firmly rooted in the company’s overarching mission and ethics. This alignment ensures that AI’s actions and outputs are consistent with the company’s values, meaningfully contributing to the broader objectives. The other side of the coin, i.e. if AI is not aligned with a company’s values, can lead to ethical concerns, trust issues, operational inefficiency, regulatory issues, and brand damage. Agentic execution plays a key role in ensuring AI-driven tasks are not only automated but also dynamically adaptive. AI with agentic execution autonomously analyses scenarios, makes contextual decisions, and acts while staying aligned with business objectives. Emphasising the importance of this integration, Anshu Kedia, CFO of ValueLabs, said, “The goal is to foster a culture where humans and AI are seen as collaborative partners, each bringing unique strengths to the table.” Enabling ‘Human + AI’ Collaboration AI implementation alone is not enough. Organisations must invest in reskilling their workforce to work effectively with AI. Employees must understand AI’s capabilities and how to use them for productivity. Companies that fail to do so risk resistance to AI adoption and underutilisation of technology. ValueLabs has taken proactive steps to bridge this gap. “We’ve invested heavily in training programs that help our teams understand the capabilities of our AI systems and seamlessly incorporate them into their daily workflows,” Kedia said. Training programs, gamified learning initiatives, and cultural shifts ensure that AI is embraced as an enabler rather than a threat. Only once this foundational alignment is established can the next phase of integration begin, facilitating effective collaboration between humans and AI. This is a challenge that is critical but often overlooked. Beyond training, businesses must refine AI-human collaboration workflows. AI should take over repetitive, data-intensive tasks while humans focus on strategic, creative, and high-value responsibilities. This balanced approach enhances operational efficiency and fosters innovation. Transforming Business Offerings with AI A well-integrated AI system enables businesses to reimagine their products, services, and overall strategy. Companies using AI for Autonomous Everything, a concept in which AI-driven processes function independently with minimal human intervention and permeate various aspects of our lives, are unlocking new opportunities. With AI handling data analysis, decision-making, and process automation, businesses can scale operations more effectively. Organisations that integrate AI-driven workflows are experiencing: Faster decision-making with real-time insights Enhanced customer engagement through AI-powered personalisation Reduced operational costs through automation Increased innovation in product development As this ‘Human + AI’ partnership solidifies, the final stage of transformation can occur—a complete reimagining of the company’s products, services, and overall outlook. The Future of AI-Driven Organisations The shift to a ‘Human + AI’ model is more than a technological change; it’s an organisational transformation. Companies that successfully integrate AI with human expertise will lead in innovation and operational excellence. However, challenges such as AI bias, regulatory compliance, and ethical considerations must be addressed proactively. Vertical AI will continue to shape industry-specific transformations. Businesses investing in AI alignment, reskilling programs, and ethical AI frameworks will differentiate themselves in an increasingly AI-driven world. Those that embrace Autonomous Everything will achieve unprecedented efficiency, scalability, and long-term success. By embedding AI into their core values and fostering a culture of AI-human collaboration, forward-thinking companies like ValueLabs are setting the stage for a future where AI doesn’t replace human intelligence but amplifies it.","excerpt":"Companies are moving away from generalised AI and using Vertical AI.","categories":["AI Highlights"],"tags":[],"author_name":"Mohit Pandey","publish_date":"2025-04-08T09:30:00","publication_year":"2025","word_count":812,"keywords":["Go","artificial intelligence","TPU","AI","ML","Scala","ViT","Rust","GAN","R"],"extracted_tech_keywords":["AI","artificial intelligence","ML","TPU","R","Go","Rust","Scala","GAN","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/how-humans-and-ai-agents-can-work-together\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":34718,"title":"Hackathons Help Us Hire Data Scientists With Core Competencies On A Large Scale: Puneet Gupta, HomeLane","content":"Hackathons have become the go-to solution for recruitment in top IT companies and startups. Over the last few years, hackathons have become a key parameter in the hiring process, with companies testing real-time problem-solving skills through these platforms. Hackathons have now become a baseline — a way to screen resumes and test candidates for the understanding of data science tools and how well they can execute the problem statement. This month’s theme covers hackathons as a recruitment tool. For the first interaction, we got in touch with Puneet Gupta, SVP, Head of Engineering at HomeLane. An experienced tech expert with over 18 years of experience and 11 patents to his credit, he currently heads the engineering department and spearheads data and AI-driven tech innovations in the company to better the consumer experience. He shared how hackathons have become a crucial part of the hiring process in data science and how it has replaced the traditional way of hiring, and more. Analytics India Magazine: Do you think hiring through Hackathons streamlines the hiring process? How has Hackathon been mainstreamed in the recruitment process? Puneet Gupta: Hackathons are an excellent platform to meet passionate developers and product designers and engineers. They bring together individuals from diverse backgrounds and enable tapping into a pool of skilled and driven people for hiring. However, it is debatable whether all forms of hiring can be eased through hackathons. Niche roles and specialised competencies are hard to weave into a broad-based Hackathon, where the focus is on mobilising a larger community of hackers with certain core competencies. AIM: What is your opinion about online Hackathon? Have Hackathons truly replaced the traditional way of hiring, say by campus hiring? PG: Going online simplifies the logistics and reduces the costs for conducting hackathons, besides easing participation from hackers across the geographical regions. Therefore, it is simpler to tap into talent that is not local to your city or state. Campus hiring also benefits from such a platform, allowing more employers to reach across campuses across the country. AIM: What is the main purpose of conducting a Hackathon for you? PG: The primary purpose of the HomeLane hackathon is to connect with a community of exciting developers and explore hiring opportunities. More importantly, we hope that this opportunity helps us convey HomeLane’s commitment to technological innovations in the realm of cloud platforms & web applications. AIM: Can you cite a recent example where you hosted a Hackathon? Could you state the purpose? PG: We are conducting a Hackathon for backend developers\/engineers on February 8, 2019, where we will be putting the talents of developers and engineers with 2-8 years of experience to test. Our primary purpose of hosting the hackathon is to connect with a community of exciting developers and explore hiring opportunities. We are hoping to find some extraordinarily talented engineers and developers through this event. AIM: How important are Hackathons (both internal and external) to boost innovation? Has your organisation benefitted by conducting Hackathons? PG: It always helps to take a break from the day-to-day work and explore avenues where a hitherto unexplored challenge awaits you. Innovation stems from constant learning and pushing the boundaries of comfort – something that hackathons are tailor-made for! AIM: Besides building people capital, do hack-for-hire also help in positioning the company as innovative and help in branding? PG: Absolutely. It reflects an open, modern mindset for the company and such branding is very welcome for budding startups as well as established tech giants, especially among the young and passionate tech population. AIM: Hackathons have become the benchmark for assessing baseline skills? Do you agree, if so what skills? How reliable it is to test skills in the long run? PG: Most hackathons these days focus on programming skills, problem-solving or backend full-stack web development. These find synergy in the nature of the technology landscape that dominates the industry. It depends on the evaluation and selection process to pick the best of the lot, and then groom the hired individuals into roles they are selected to play. Eventually, the success of an individual at an organisation also depends on the opportunities & training after onboarding. AIM: What is your preference for candidates who have participated in Hackathon vs direct hiring? PG: Good candidates can be found through many channels. One channel does not undermine the other. Both hackathons and direct hiring have their place and their own benefits. AIM: Are hackathons replacing the need to have a short-term certification course in areas such as AI, analytics? PG: I do not think so. Systematic learning has a lot of value. And, so does the practical application of the concepts learnt. Rather than competing with each other, I believe these two complement each other. AIM: Can you cite partners with whom you usually prefer conducting Hackathons? PG: At present, we are working with HackerEarth. In the future, we might explore other partners too, depending on the target segment and the kind of participation we seek through a hackathon.","excerpt":"Hackathons have become the go-to solution for recruitment in top IT companies and startups. Over the last few years, hackathons have become a key parameter in the hiring process, with companies testing real-time problem-solving skills through these platforms. Hackathons have now become a baseline — a way to screen resumes and test candidates for the […]","categories":["AI Features"],"tags":["data science hackathon india","hackathon india","Interviews and Discussions"],"author_name":"Srishti Deoras","publish_date":"2019-02-11T04:37:15","publication_year":"2019","word_count":833,"keywords":["hackathon india","data science","Go","API","AI","innovation","ML","Aim","analytics","data science hackathon india","GAN","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","ML","data science","analytics","Aim","R","Go","API","GAN","innovation"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/hackathons-help-us-hire-data-scientists-with-core-competencies-on-a-large-scale-puneet-gupta-homelane\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":39702,"title":"How StayAbode Is Using Analytics To Edge Out Competition In The Co-Living Space In India","content":"Co-Founders Viral Chhajer, Devashish Dalmiaand Varun Bhalla In today’s world of millenials moving out of their home cities for career, co-living spaces become very convenient and are growing rapidly in India. According to a recent survey titled “Co-Living – rent a lifestyle” by Knight Frank India, the international property consultancy 72 percent of millennials have given co-living spaces a thumbs-up and over 55 percent of respondents in the age group of 18 – 35 years are willing to rent co-living spaces. Big players are emerging in the country to make their mark in this space specially meant for the working population. In this race of co-working space, StayAbode, established in 2016, is trying to use analytics tools and make its mark in India. Analytics India Magazine talked to one of the Co-Founders of StayAbode, Devashish Dalmiya about the startup, their working ecosystem and their goals. The Beginning Of The Journey The idea of StayAbode started while the co-Founder Devashish Dalmiya was backpacking across Europe where he lived in many hostels and saw the advantages of co-living first hand. From there when he looked at urban centres in India, he realized that the problem in India wasn’t that of living for travellers, but for millennials moving into cities to live. From there the idea moved to long term co-living spaces for working millennials and they began with Bengaluru as it was the perfect destination for migration for the working millennial across the country. A key issue was selling the idea of co-living itself. Dalmia said that initially people really saw StayAbode as another PG with better facilities but then when people began to move in and started using the common spaces such as their community kitchens and laundry spaces, along with the many events that they had at their properties, they began to understand the value that StayAbode was delivering. Also, the same value needed to be sold to their landlords as a concept which was tricky initially as they weren’t used to dealing with companies to manage their spaces, but when the landlords saw the returns in terms of rental yields, they were able to see the larger value on offer. The Tech Behind The technical team at StayAbode has built all of their technology in-house and it continues to do so. The team consists of backend, frontend, mobile and QA engineers. This involves development across their stakeholder tools, the website and the resident app. The technical team uses a combination of industry standard tools like Google Analytics and Mixpanel. They also have many in-house tools developed specifically for their own use-cases to power their decision making for onboarding supply, demand analysis, pricing, property maintenance, smart auditing and so on. They have also built out tools for stakeholders across the company which includes dashboards for the owners of StayAbode to see how their property is performing to a portal for the central team to manage operations and customer experience across their properties. They also built out a resident app that gives their users the freedom to know their neighbours, get event updates, pay their rentals and raise issues. “Being in the hospitality space requires us to be on our toes all the time,” said Mr. Dalmia while talking about the challenges that the team faces in their everyday work. In order to ensure a high quality resident experience at scale, the team has to ensure that the technology team works closely with operations, support and the other teams in the organisation and they build systems for the future when they are hosting residents in multiple cities and countries Success Stories Dalmia said that the story of StayAbode has really seen two positive aspects to it which has been their high occupancy rates along with their ability to add inventory across different living formats. They have consistently seen over 95% occupancy across their spaces since the inception of StayAbode. They have also managed to add inventory consistently across their current BHK and room layouts. They also are scheduled to launch their first large scale co-living project at the end of 2020 with a Greenfield project that will be state of the art and with a living capacity of about 1400 people. On both these fronts StayAbode has managed to continue to grow as they have acquired scale while maintaining a great experience across our residents. The startup has raised funding from Anupam Mittal, CEO of People Group, Vineet Sekhsaria, Head, Real Estate investing at Morgan Stanley, a Japanese gaming company called Akatsuki Inc, Incubate Fund and most recently the Voyage Group was added on to StayAbode’s Pre Series A round which they closed the last year. StayAbode Is Unique From The Others There have been many players in this space of co-living spaces in the last couple of years. There are big players like Oyo Living, Nestaway and CoLive that have built a fairly large inventory and there are other players like CoHo, Grexter and FF21. StayAbode believes that they have hit a great formula of providing spaces that are true to the idea of co-living with residents enjoying the shared experiences their spaces provide, and this is a great differentiator as compared to many of their competitors, especially the larger ones. StayAbode is slowly inching away from some of the more niche players in terms of the inventory that they have been adding. Future Endeavours The goals of StayAbode this year include looking at expansion across cities beyond Bangalore, while continuing to maintain customer experience and occupancy levels at scale. StayAbode has their Greenfield project next year in Whitefield in Bangalore, which they are very excited about. The project will house over 1400 people in a state of the art co-living facility. Mr. Dalmia said, “Our intention is to continue growing the current business while we look to add more such built to suit projects in the upcoming years. Customer experience and how we improve the community living experience is always at the top of our collective minds and we intend to continue innovating in that space.”","excerpt":"In today’s world of millenials moving out of their home cities for career, co-living spaces become very convenient and are growing rapidly in India. According to a recent survey titled “Co-Living – rent a lifestyle” by Knight Frank India, the international property consultancy 72 percent of millennials have given co-living spaces a thumbs-up and over […]","categories":["Deep Tech"],"tags":["Startups"],"author_name":"Disha Misal","publish_date":"2019-05-27T04:07:21","publication_year":"2019","word_count":1007,"keywords":["Go","API","funding","programming_languages:R","AI","programming_languages:Go","analytics","GAN","Startups","R","startup"],"extracted_tech_keywords":["AI","analytics","R","Go","API","GAN","startup","funding","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/how-stayabode-is-using-analytics-to-edge-out-competition-in-the-co-living-space-in-india\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10130639,"title":"ToyStack.ai Raises $325,000 in Pre-Seed Funding Round","content":"Bengaluru-based ToyStack.ai has raised $325,000 in its pre-seed funding round. Founded in 2023 by Mukund Gandlur, Sravan Aditya, Arun Gandlur, and Torun Mathias, ToyStack.ai is an automated deployment tool that enables developers and software teams to reduce the time needed to deploy their code to the ToyStack cloud to a few minutes. According to the founders, the platform aims to remove time-consuming tasks associated with server setup, continuous integration and continuous deployment (CI\/CD) pipelines, orchestration tools, and extensive configuration and security measures. The startup plans to utilise the funds for further product development, talent acquisition, sales, and marketing initiatives. “The pre-seed round is a validation of our product. This investment further bolsters our mission and will help us in our next phase of product development and customer growth,” said Mathias on the investment. Currently, the platform has integrated LLMs to optimise its deployment process and reduce build times. The LLM identifies patterns and inefficiencies, provides insights and suggests ways to improve optimisation, by analysing data from past build logs, error reports, and time metrics. ToyStack.ai currently boasts users from as many as 100 countries and over 14,000 deployments to their cloud. The startup currently aims to create an automated infrastructure toolchain that can simplify the deployment process for developers and companies worldwide. The round was led by angel investors, including the founder of SDU LLP Chartered Accountants, D Devaraj, and the founder of Aveda Ventures, Asit Shetty. Commenting on the startup, Devaraj said, “With increased digitisation and the advent of AI, more businesses will necessarily move to the cloud, and Toystack is providing an innovative and simple solution for automating the cloud journey.”","excerpt":"The startup plans to utilise the funds for further product development, talent acquisition, sales, and marketing initiatives.","categories":["AI News"],"tags":["coding","Fund Raising"],"author_name":"Donna Eva","publish_date":"2024-07-29T17:50:16","publication_year":"2024","word_count":273,"keywords":["funding","programming_languages:R","AI","Fund Raising","coding","CI\/CD","Git","Aim","GAN","R","startup"],"extracted_tech_keywords":["AI","Aim","R","Git","CI\/CD","GAN","startup","funding","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/toystack-ai-raises-325000-in-pre-seed-funding-round\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":43642,"title":"Understanding The Applications Of Artificial Intelligence In Document Management","content":"“We are drowning in information, but starved for knowledge” This is a famous quote by John Naisbitt which shows the key difference between information and knowledge. Advancement in data engineering techniques and cloud computing have made it easy to generate data from multiple sources but making sense of this data and getting insights is still a huge challenge. The data volumes have now increased exponentially and along with the traditional structured data, data can now reside in different formats like unstructured social media text, log files, audio\/video files, streaming sensor data etc. Applying manual methods to process this diverse data is not only time consuming and expensive but is also prone to errors. Hence the need of the hour is to use Artificial Intelligence (AI) based automated solutions that can deliver reliable insights and give a competitive advantage to customers. Here are the few examples of how customers across industries can benefit from AI driven solutions. Microsoft Azure based AI solution In 2017, more than 34,000 documents related to John F Kennedy’s assassination were released. The data volume was huge, and data existed in different formats like reference documents, scanned PDF files, hand written notes and images. It would take researchers months to read through this information and hence manually reviewing this data was not the most optimal solution. Microsoft Azure team applied AI based Cognitive Search solution to extract data from these diverse sources and gained insights. Technical architecture for this use case was built using Azure Cognitive Services components like Computer Vision, Face Detection, OCR, Handwriting Recognition, Search and core Azure components like Blob Storage, Azure ML, Azure Functions and Cosmos Database. This solution also annotated text using custom CIA Cryptonyms. Hospitals usually deal with a lot of patient data which could reside in electronic medical records (EMR), handwritten prescriptions, diagnostic reports and scanned images. AI based Azure Cognitive Search could be an ideal solution to efficiently manage patient’s medical records and creates personalized treatment plan. Many downstream use cases like Digital Consultations, Virtual Nurses and Precision Medication can be built once the patient data is optimally stored. Google Cloud Platform (GCP) based AI solution GCP introduced Document Understanding AI (beta) in Cloud Next 19. This is a serverless platform that can automate document processing workflows by processing data stored in different formats and building relationships between them. This solution uses GCP’s vision API, AutoML, machine learning based classification, OCR to process image data and custom knowledge graph to store and visualize the results. Customers can easily integrate this solution with downstream applications like chatbot, voice assistants and traditional BI to better understand their data. Customers who deal with Contract Management data like Mortgages are usually faced with a lot of manual tasks to ensure that the contracts are complete and accurate. This could mean processing contracts in different formats\/languages, reviewing the supporting documents, ensuring that the details are accurate and complies with regulatory standards across documents. By using Document Understanding AI and integrating it with a well-designed RPA framework, customers will be able to efficiently process Mortgage applications, Contracts, Invoices\/Receipts, Claims, Underwriting and Credit Reports. Use Cases from other industries Document Management AI solution can also be applied to diverse use cases from other industries like processing claims related to damages to shipped products by e-commerce companies, handling know your customer (KYC) process in the banking industry, invoice data processing by Finance teams, fraud detection during document processing etc. As more and more companies embrace the digitization wave, they will be faced with different variations of data\/document management challenges. Based on the current trend, numbers of use cases are only going to increase and an AI driven solution is probably the most efficient way to solve this problem as it can reduce manual work, save cost and deliver reliable insights. This will ensure that companies can spend more time on building their business and less time on manually processing documents and data preparation. Going back to John Naisbitt’s quote, AI and ML driven solutions are probably the only way to bridge the gap between information and knowledge.","excerpt":"“We are drowning in information, but starved for knowledge” This is a famous quote by John Naisbitt which shows the key difference between information and knowledge. Advancement in data engineering techniques and cloud computing have made it easy to generate data from multiple sources but making sense of this data and getting insights is still […]","categories":["AI Features"],"tags":["AI App"],"author_name":"Pavan Nanjundaiah","publish_date":"2019-08-01T10:58:17","publication_year":"2019","word_count":680,"keywords":["machine learning","artificial intelligence","AI","cloud computing","ML","computer vision","RAG","Aim","AI App","Azure ML","fraud detection"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","computer vision","Aim","Azure ML","RAG","fraud detection","cloud computing"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/understanding-the-applications-of-artificial-intelligence-in-document-management\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":5178,"title":"Book Review &#8211; Big Data: A Revolution That Will Transform How We Live, Work, and Think","content":"Big Data: A Revolution That Will Transform How We Live, Work, and Think by Viktor Mayer-Schönberger, Kenneth Cukier The book, written by two experts in the area of big data, gives an exciting introduction to what Big Data is and how it has changed the world around us. The book throws interesting examples that along with being interesting also showcases how big data have been utilized till date. The book explains a shift towards an increased production usage of large nearly complete population levels data sets in the management and control of a range of industries. There are three facts expounded in this book – first, because big data makes storing and retrieving of large amount of data so cheap, businesses now are able to store all the data that they can lay their hands on. This has lead to the almost demise of the earlier high fashioned sampling techniques in data analysis. Second, since it’s all of data that now we can work on and analyze, messiness in data is allright. We do not and should not really bother and spend resources on data cleansing, as was done when we had just a sample of data. Third, the authors propagate the fact around datafication i.e. since we can store a large amount of unstructured data cheaply; there is an added emphasis on collecting data from sources that we previously ignored. Also, the authors emphasize on correlation between things rather than the older concept of causality or why things happen. But having said this, the book went on to repeat itself after first few chapters. Also, it gives a good introduction to big data with valid examples but the ideas stopped there. What’s next and how things are going to evolve with big data is not so much touched in the book. If the reader is an industry practitioner on big data, the book does not offer much. Yet, it’s still worth reading with lots of interesting stories and maintained a good balance of technical and entertaining perspectives. Our Rating: 2.5\/ 5","excerpt":"Big Data: A Revolution That Will Transform How We Live, Work, and Think by Viktor Mayer-Schönberger, Kenneth Cukier The book, written by two experts in the area of big data, gives an exciting introduction to what Big Data is and how it has changed the world around us. The book throws interesting examples that along […]","categories":["IT Services"],"tags":[],"author_name":"Дарья","publish_date":"2014-02-02T07:44:58","publication_year":"2014","word_count":342,"keywords":["big data","Go","programming_languages:R","AI","programming_languages:Go","R"],"extracted_tech_keywords":["AI","R","Go","big data","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/book-review-big-data-a-revolution-that-will-transform-how-we-live-work-and-think\/","complexity_score":2,"technical_depth":6,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10137721,"title":"Developers are Finally Getting Their Jobs Back After Layoffs","content":"Most big-tech companies are hiring developers again, offering thousands of jobs. Similarly, the market beyond big-tech is rebounding with several recruiters calling developers for interviews. This is after the year-long (or more) fear of AI replacing developer jobs, that took place in several cases. The recent resurgence of companies hiring developers started back in June when there were around 117k valid and unique job openings in tech at JobsCopilot.ai. The number has now increased multifold. Even the companies that laid off developers just a year back citing AI as a reason, have learned from and accepted their rushed decisions. They are now rehiring developers for those roles. This includes companies such as Amazon, Apple, Microsoft, Google, Meta, NVIDIA, and Tesla, to name a few. Some of them laid off around 20,000 employees just this year. Santiago Valdarrama, the founder of Tedeily, shared that the founder of a company (not revealing the name) that laid off most of its developers a year back, expressed to him, “I think we were too early…things didn’t work out as we were hoping.” Furthermore, this case is especially important for the offices of big-tech companies in India, as around 85,000 techies graduating this year will be looking for jobs. For example, Google laying off its Python programmers team to outsource to cheaper options might be a good chance for Indian developers to finally hop on board. Moreover, after the year-long predictions and the sad state of hiring in Indian IT, it is finally showing signs of growth. This has resulted in a boom of hiring in a battle against, and for, generative AI. Even Indian startups have started hiring back developers after laying off around 30,000 people in 2022 and freezing hiring for around 12-18 months. Thanks to AI, startups’ hiring budgets have increased. Even the Indian IT Following a weak FY24, hiring is expected to increase by 8.5% in FY25, which will continue until March 2025. This means around 150,000 fresher jobs will be open next quarter, which is in line with the fresh graduates who will be looking for jobs after college. This is due to the billion-dollar deals that the IT giants have been making and the increase in demand for the BFSI segment. According to a report from Indeed, the IT sector is actively hiring for roles such as application developer, software engineer, full-stack developer, and senior software engineer. There’s also a growing demand for .NET developers, software architects, DevOps engineers, data engineers, and front-end developers. We will finally see a reduction in the bench size of Indian IT, which has been growing for the last year. This is after TCS announced hiring 35,000 freshers last year, while others, such as Infosys, Wipro, and Accenture, froze hiring of more than 25,000 developers combined. Now, they are looking to hire around 80,000 freshers combined. The goal to focus on automation and the belief that upskilling the existing workforce with generative AI would be able to achieve that clearly did not go according to plan. Hiring is Back, Unlike Some Predicted Slack started to rehire former staff to fuel its generative AI initiative just six months after Salesforce, its parent company, laid off around 8,000 people. Slack back then had said that the developer layoffs were because of generative AI. Just a few weeks ago, Klarna also cut 50% of its workforce, ending partnerships with Salesforce and Workday for what it claims is a generative AI overhaul. It seems like companies haven’t yet realised that laying off people in the name of generative AI doesn’t work in the long run. Cost-cutting was one of the biggest reasons why companies started laying off people last year. This was after they overhired developers during the COVID era only to realise their inability to sustain the jobs later. The blame on AI for these layoffs turned out to not be true since it was mostly because of cost-cutting. “The story ‘we’re so productive now we don’t need employees anymore’ sounds a lot more palatable for investors than ‘we over-hired and we’re struggling,” quipped Francois Chollet, the creator of Keras. When Paytm laid off 10% of its workforce blaming AI, it was later reported that the layoff was mostly due to poor quarters and stagnant growth. It turns out that companies struggle to perform well after laying off their developers, and when things go wrong, they shift the blame on AI. But apparently, Klarna’s case is different. The CEO, Sebastian Siemiatkowski, said that AI will play a crucial role in cutting operational costs, contributing to a 50% planned reduction in the company’s workforce. This move is expected to help the company operate more effectively while maintaining higher quality standards. Regardless, some companies still search for a “generative AI solution” to cut costs. With the market pulling itself back up, companies might learn a lesson or two. We might see more cases where companies are rehiring developers, as AI is definitely not replacing developers anytime soon.","excerpt":"“I think we were too early…things didn’t work out as we were hoping.”","categories":["IT Services"],"tags":["AI Developers","Developers","Klarna","Layoffs"],"author_name":"Mohit Pandey","publish_date":"2024-10-07T16:11:16","publication_year":"2024","word_count":828,"keywords":["Go","AI Developers","Keras","Layoffs","AI","automation","Klarna","Python","Aim","generative AI","DevOps","R","Developers","startup"],"extracted_tech_keywords":["AI","generative AI","Aim","Keras","Python","R","Go","DevOps","automation","startup"],"url":"https:\/\/analyticsindiamag.com\/it-services\/developers-are-finally-getting-their-jobs-back-after-layoffs\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10169146,"title":"Claude AI-maker Anthropic to BuyBack Employee Shares","content":"AI startup Anthropic, funded by Amazon and Google, has planned its first-ever share buyback for employees, according to reports. Claude AI maker Anthropic seeks to repurchase shares from current and former employees valued at $61.5 billion. In March, the company announced that it raised $3.5 billion at the same valuation for the Series E round. The share buyback plan intends to allow employees to liquidate their equity while allowing the company to retain its top performers amid an aggressive hiring environment. According to the Economic Times, the company’s valuation has skyrocketed in the past year. In March 2024, it was valued at $15 billion, and by the end of the same year, it reached over $60 billion after back-to-back funding rounds from investors such as Google, Amazon and Salesforce Ventures. Employee liquidity programmes such as these have become more common across AI companies. AIM reported in May 2024 that Apple initiated a programme after its Q2 profit, which resulted in a stock surge of 12% in after-hours trading as the CEO predicted sales growth with upcoming AI-driven features. Similarly, in November last, OpenAI allowed its employees to sell about $1.5 billion of shares to SoftBank. Broadcom, a tech company, also launched a share buyback programme of up to $10 billion on 7 April, reflecting its semiconductor and infrastructure software franchises, especially in its AI-related investments.","excerpt":"Anthropic seeks to repurchase shares from current and former employees valued at $61.5 billion.","categories":["AI News"],"tags":["Anthropic","Share buyback"],"author_name":"Smruthi Nadig","publish_date":"2025-05-05T17:26:29","publication_year":"2025","word_count":225,"keywords":["Anthropic","Go","funding","OpenAI","AI","programming_languages:R","llm_models:Claude","Aim","R","Share buyback","startup"],"extracted_tech_keywords":["AI","OpenAI","Anthropic","Aim","R","Go","startup","funding","llm_models:Claude","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/claude-ai-maker-anthropic-to-buyback-employee-shares\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":4492,"title":"Analytics on Google’s Ngram Viewer","content":"In 2004, Google came up with a bold plan. It would take up all the books that it can lay its hands on and store them in digital format. The result was the Google Books project where anyone can search for keywords appearing in the millions of books that Google scanned. This project was hugely successful. But then Google went a step ahead and introduced Google’s Ngram Viewer (http:\/\/books.google.com\/ngrams), which allows users to search how a word has appeared through centuries of recorded history in form of a graph. A very powerful tool, Google’s Ngram Viewer allows us to ascertain when a phrase was first used and during which years it became popular. We ran the terms “Analytics” on Google’s Ngram Viewer. Here’s what we found- It is evident from the graph that the term Analytics was popularized only at the turn of this century and has seen an exponential increase since then. The first recorded use of the word “Analytics” was in 1733 book “The Philosophical Works of Francis Bacon, Baron of Verulam”. Analytics was then used as a synonym for Algebra. A text in 1725 book “Lexicon Technicum: Or, An Universal English Dictionary Of Arts And …, Volume 1” By John Harris says [quote]Thus a Chymist is said to Analyze Bodies, when he dissolves them by the Fire, and endeavours to find out their Constituent Parts : And Algebra is sometimes called the Analytics.[\/quote] The word Analytics was first popularized by Aristotle around 350 BC. “Prior Analytics or Analytica Priora is Aristotle’s work on deductive reasoning, which is known as his syllogistic.” [1] “The Posterior Analytics is a text from Aristotle’s Organon that deals with demonstration, definition, and scientific knowledge.” [2] The term “Marketing Analytics” first appeared on 1986 book “Responding to the challenge: health care marketing comes of age” by Philip D. Cooper. According to a text from the book – [quote]Age\/Sex\/Zip-Code specific population projections are at the heart of most healthcare marketing analytics.[\/quote] The first legitimate use of term “Financial Analytics” appeared on 1993 book “The New Corporate Finance: Where Theory Meets Practice” by Donald H. Chew. According to a text from the book – [quote]The model has six basic components: 1. Security Pricing Models: State-of-the-art financial analytics are used to calculate theoretical prices for various securities such as bonds, futures, forwards, … [\/quote] The first legitimate use of term “Risk Analytics” appeared on 1997 book “Financial Risk Analytics: A Term Structure model Approach for Banking Insurance and Investment Management” by Donald R. Van Deventer, Kenji Imai.","excerpt":"In 2004, Google came up with a bold plan. It would take up all the books that it can lay its hands on and store them in digital format. The result was the Google Books project where anyone can search for keywords appearing in the millions of books that Google scanned. This project was hugely […]","categories":["AI Features"],"tags":["Google Analytics"],"author_name":"Дарья","publish_date":"2014-01-14T09:12:20","publication_year":"2014","word_count":421,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","Git","Google Analytics","analytics","GAN","R"],"extracted_tech_keywords":["AI","analytics","R","Go","Git","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/analytics-on-googles-ngram-viewer\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":31088,"title":"Dating Game Explained With Probabilistic Graphical Models And Neural Networks","content":"Neural Networks are a network of nodes which can process data and extract valuable insights, inspired from biological neurons within the brain. They were introduced with the sole purpose of recreating or achieving the capabilities of a human brain. A neural network predicts by determining the correlation between the input and output. What Is A Probabilistic Graphical Model? The Probabilistic Graphical Model or PGM is an amalgamation of the classic Probabilistic Models and the Graph Theory. How Are Probabilistic Graphical Models And Neural Networks Related? Both PGM and NN are data-driven frameworks and both are capable of solving problems on their own. The essential difference between them rests on how they use the data to predict the outcome. One of the major drawbacks of NN is that it has a certain level of uncertainty in the prediction of outcomes. PGM makes the uncertainty explicit and helps to build models that are more faithful to reality. PGM is also more capable than NN in that case. A PGM can be limited to providing the functionality of an NN and at the same time, NN can also be exploited to obtain the information that a PGM would provide. Problem Statement We are going to use one of the popular use cases found on the internet — asking a girl out for a date — to describe the same. John Doe likes a girl he knows through N mutual acquaintances and wants to ask her out on a date. He has low self-esteem and wants to ask some subset of their mutual friends to introduce him to her. However, a referral from someone she mistrusts may only have a weak positive influence, while a referral from someone she dislikes might even have a negative impact. Now suppose n of his friends have already had the same idea, and have tried doing this before (assume, for the purposes of this argument, that the young lady in question is unbearably desirable). So he has in his possession, a set of observations, which he can compactly represent as an n bit vector [1,0,1,0 ….] where 1 is a boost, 0 means no contact. John needs to find the set of friends that he should ask to boost him, i.e. the best vector. Note that this need not be one of the n vectors already tried. Solution Using a Neural Network: The input layer would have N nodes that obtain the appraisal input, the output layer would be the girl’s response in binary and we can throw in a hidden layer of M neurons that all have N connections back up to the input layer, and encode various quanta of me-friend-girl dynamics that when pooled together in M forward connections to the output layer, determine her interest in him. When we train this neural network with the n vectors we have, we will learn weights between 0 and 1 on the NM input-mid-layer connections and the M mid-layer-output connections. Intuitively, combinations of inputs that predict the output well (say her 3 closest friends) will be repeatedly reinforced and will become bigger across n observations. So will the weight corresponding to the mid-level neuron that is receiving inputs from specifically that clique. However, there will be no easy way of knowing which of the M mid-layer neurons contains the 3-closest-friends information. The NN will function as a Delphic oracle — you can ask it about the fate of individual vectors, but not for reasons explaining its prediction. Using PGM: We could also treat this problem as one of Bayesian reasoning, where we potentially receive observations of approval from N nodes, which leads to the formation of an impression (a random variable), which causes date acceptance (an observable event). In this case, we get to see the likelihood probability p(approval from friend i|impression), from which we have to estimate the conditional posterior probability p(impression|vector of all approvals) using Bayes theorem. Going from p(approval i|impression) to p(all approvals|impression) though, is hard. Usually, machine learners tend to assume conditional independence for all i approvals, i.e. p(all approvals|impression) = product of all p(approval i | impression). This is simple to compute, but gives up on the possibility of modelling non-trivial correlations between inputs. For example, if hearing good things from either A, B and C or from C and F, but not from A and C together impresses the girl (assume that the girl’s social life is extremely rich), such effects won’t show up in such ‘naive’ Bayesian predictions. Summary: The neural Network lacks the reason behind a prediction that it makes whereas a Probabilistic Graphical Model shows enough evidence to support its prediction. PGM’s will give you detailed intermediate predictions about how likely the individual inputs will be to generate the effect. Outlook Both Neural Networks and Probabilistic Graphical Models can describe the correlation between the dependent and independent factors of any problem which involves a number of features that results in a  specific outcome. Both can be used to learn about the network functions.","excerpt":"Neural Networks are a network of nodes which can process data and extract valuable insights, inspired from biological neurons within the brain. They were introduced with the sole purpose of recreating or achieving the capabilities of a human brain. A neural network predicts by determining the correlation between the input and output. What Is A […]","categories":["AI Features"],"tags":["dating","game theory","Neural Networks"],"author_name":"Amal Nair","publish_date":"2018-12-05T08:45:45","publication_year":"2018","word_count":834,"keywords":["Go","dating","TPU","programming_languages:R","AI","neural network","data-driven","programming_languages:Go","game theory","Rust","R","programming_languages:Rust","Neural Networks"],"extracted_tech_keywords":["AI","neural network","TPU","R","Go","Rust","data-driven","programming_languages:R","programming_languages:Go","programming_languages:Rust"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/dating-game-explained-with-probabilistic-graphical-models-and-neural-networks\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":42420,"title":"Nano-Sized Probes Through Our Bloodstream Can Become A Reality Soon","content":"Imagine a future where a swarm of nano-sized probes are voyaging through the bloodstream for intracellular recordings, mapping brain activity and performing other diagnostics. Looks like the future is here, thanks to a new research from the University of Surrey and Harvard University. Much of the interest in nanotechnology stems from the unique quantum and surface phenomena that matter exhibits at the nanoscale. As the majority of the atoms in these nanostructures are a short distance from the surface, the optical and electrical properties of these systems can be strongly modified by changes in their environment, allowing a host of applications, particularly in sensing. In recent years the techniques for growing and fabricating nanoscale materials have matured to the point where researchers are able to tailor their properties toward particular applications, such as metal nanoparticles, metal oxide nanowires and carbon nanotubes for a range of applications including physical, chemical and environmental sensors, energy scavenging materials, low-cost transparent conductors and optoelectronic devices. Non-Invasive Nano Probing The existing nanodevices have a trade-off between scalability and recording amplitudes. In an attempt to address this, the researchers at the University of Surrey and Harvard developed a scalable nanowire field-effect transistor probe arrays with controllable tip geometry and sensor size, which enable recording of up to 100 mV intracellular action potentials from primary neurons. Through this work, clear evidence has been found for how both size and curvature affect device internalisation and intracellular recording signal. Source The above figure consists of the (i) Side view schematic of the probe\/cell; (ii) Top view optical image of probe\/cell, showing single U-NWFET(U-shaped NanoWire Field Effect Transistor) probe recording from one neuron. Dr Yunlong Zhao who played a key role behind this research observes that these ultra-small, flexible, nanowire probes could be a very powerful tool as they can measure intracellular signals with amplitudes comparable with those measured with patch clamp techniques. Moreover, this device comes with the advantage of being scalable and causes less discomfort and no fatal damage to the cell. New tools such as these for intracellular electrophysiology that push the limits of spatiotemporal resolution while reducing invasiveness could provide a deeper understanding of electrogenic cells and their networks in tissues, and push progress towards human-machine interfaces. These probe developments adding to existing capabilities can ultimately drive advanced high-resolution brain-machine interfaces and perhaps eventually bringing cyborgs to reality. Future Direction Nanotechnology and Artificial Intelligence are two domains which will have significant implications on the future of humans. As AI offers assistance to our intangible intelligence through its uncanny pattern recognition, nanomaterials will be more direct and personal bridging the gap between these diverse fields will open up new avenues. For instance, the nanoprobes scan the cells for unwanted mutations and send the signals to a model which runs on a machine learning algorithm that is trained to classify a cell as cancerous or not. A feedback loop can be accommodated where the results of the machine learning algorithm will direct the probe to destroy the cancerous cell- a classic case of nanomedicine. As we are yet to understand the full functionality of the human brain, a collaboration of nanotransistors that probe neurons and a neural network that devours this vast amounts of data can help in the diagnosis of neurological diseases like Alzheimer’s and Parkinson’s. As a healthy one is the primary focus of many medical researchers, the success of the methods such as discussed above will encourage more research and push the boundaries of modern science. For this to be possible, an intersection between humans and machines looks almost inevitable.","excerpt":"Imagine a future where a swarm of nano-sized probes are voyaging through the bloodstream for intracellular recordings, mapping brain activity and performing other diagnostics. Looks like the future is here, thanks to a new research from the University of Surrey and Harvard University. Much of the interest in nanotechnology stems from the unique quantum and […]","categories":["Deep Tech"],"tags":["nanotech","nanotechnology"],"author_name":"Ram Sagar","publish_date":"2019-07-13T09:09:55","publication_year":"2019","word_count":595,"keywords":["Go","machine learning","artificial intelligence","AI","neural network","Scala","nanotech","RAG","Ray","ViT","nanotechnology","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","neural network","Ray","RAG","R","Go","Scala","ViT"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/nano-sized-probes-through-our-bloodstream-can-become-a-reality-soon\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10172795,"title":"Indian IT Set for Tepid Q1 as Global Risks Mount","content":"Indian IT once confidently walked the path it had paved for itself. There was growth driven by headcount expansion, steady demand ensured by low-cost delivery, and a sense of business predictability. However, with artificial intelligence steadily reshaping the landscape, the industry now finds itself trying to sprint in shoes made for walking. This picture is hard to miss and stands out sharply when recent data is examined. AIM held conversations with industry analysts, who also support the narrative. While the first quarter isn’t expected to bring major surprises, it confirms that a transition is underway. The old pace no longer fits the new terrain, and what companies overlooked before has become a pressing necessity. The Numbers Speak UnearthInsight founder Gaurav Vasu describes the Indian tech services sector as navigating a complex landscape marked by cautious optimism amid macroeconomic uncertainties and evolving client demands. Engati co-founder and managing director Imtiaz Bellary and independent digital transformation advisor Viswanathan KS agree that external factors are significantly impacting Indian IT’s growth. Vasu notes that Q1 FY26 (April–June 2025) will see muted growth due to persistent global geopolitical headwinds, tariff uncertainty, the GCC boom, and the US economic slowdown, all of which are expected to impact discretionary tech spending. UnearthInsight forecasts FY26 growth at 4–5%, with Q1 growth projected at 1–1.5% for the top 15 tech services firms, primarily driven by mid-tier players like Persistent, Cyient, and Coforge. Meanwhile, the top five IT services firms are expected to grow more slowly, at 0.8–1.2% quarter-on-quarter. PL Capital anticipates a soft Q1 across the sector but retains Tata Consultancy Services (TCS) as its only “Buy”-rated stock, citing its relative stability. While macro conditions have improved since March, the company said, deal momentum remains subdued. Tier-2 firms (annual revenues $1–5 billion) are expected to outperform due to steady, modest growth, whereas Tier-1 companies may see slight sequential declines. PL Capital also warns that valuations are rich and upside limited, signalling a cautious outlook. Credit rating agency ICRA has reaffirmed a stable outlook for Indian IT, forecasting 2–3% year-on-year revenue growth in US dollar terms for FY26. ICRA’s analysis, based on 15 leading IT companies representing about 60% of industry revenue, expects modest growth slightly lower than the 2.9% recorded in FY25. The subdued momentum is largely attributed to US tariff uncertainties impacting IT budget allocations. Viswanathan points out that Q1 is typically a slow operational quarter marked by organisational changes such as internal transfers and leadership reshuffles. Considering productivity pushes, cost pressures, AI adoption, and the rise of the GCC model, he expects Q1 results to improve over last year’s and anticipates Q2 to be stronger as well. The Elephant in the Room The critical issue influencing performance is Indian IT’s legacy challenges: heavy reliance on traditional headcount-based pricing, slow adoption of outcome-based contracts, and dependence on low-cost labour arbitrage. These factors limit agility and margin expansion in an AI-driven market. Vasu believes that while clients have shifted focus to emerging technologies like AI, generative AI, and agentic systems to transform tech stacks alongside cloud transformation as long-term growth levers, the supply side has not matched this enthusiasm. “Indian tech services firms are still working on early-stage experiments and proof-of-concept AI\/GenAI deals, so we don’t see significant AI\/GenAI revenue announcements. Except for TCS, other firms remain under $50 million in AI-related revenues, often embedded within existing deals and thus hard to isolate,” he said. Viswanathan noted that many companies were in the proof-of-concept phase for AI during Q1 and Q2 last year, but that phase is largely over. Companies are now adopting AI more broadly. Those sharply focused on AI and GenAI—especially those hiring talent with advanced coding and software skills—are likely to perform better this quarter and beyond. He added that companies investing early in the GCC model are expected to fare better. “The number of RFPs for setting up GCCs is rising significantly, indicating increased spending on technology capabilities. Whether companies build fully in-house GCCs or partner with service providers, the trend towards GCC adoption is clear.” Client Behaviour and What It Means for H2 Vasu mentioned that, particularly in the US and BFSI sectors, clients tend to be cautious with their tech spending, especially on enterprise license implementations and long-term projects. “Currently, clients are focused on quick ROI deals in AI, GenAI, engineering, and infrastructure to sustain existing technology,” he explained. When asked if AI monetisation is real or mostly marketing hype, Bellary said monetisation correlates directly with adoption. “Businesses adopting AI with strategic intent are seeing monetisation, while those resisting or superficially adopting it are missing out,” he affirmed. Bellary recommends closely tracking AI revenue and Return on Capital Employed (ROCE) for AI businesses. “Companies showing positive trends in these metrics are recovering and growing faster than others,” he added. Tier-2 Companies: Not Leading the AI Charge Vasu said tier-2 companies are not significantly ahead in AI adoption. They continue operating with traditional models and have not made substantial investments to become AI-first, just like tier-1 firms. Their faster revenue growth is due to smaller size, focused industry verticals, and limited geographic presence—not AI pivoting. “They often take on cost-optimisation deals at lower margins compared to tier-1 firms. Players like Persistent and Cyient serve niche verticals such as micro-insurance or mid-market segments. While tier-2 firms grow faster, their margins are not comparable,” he explained. Indian IT Not Outcome-Based Yet Vasu bluntly stated that Indian IT firms are “yet to adopt outcome-based models and remain largely dependent on traditional headcount-based pricing.” Despite external pressures, real change has been limited. “Meaningful shifts require investing in AI-enabled applications—a move agile startups and mid-sized firms are already making. Until then, Indian IT’s slow growth of around 1% per quarter is likely to continue,” he concluded.","excerpt":"Analysts predict a muted Q1 for the Indian IT sector in 2025-26, influenced by geopolitical headwinds, tariff uncertainties, the GCC boom, the US economic slowdown, and notably, the industry’s legacy challenges.","categories":["AI Features","IT Services"],"tags":["AI (Artificial Intelligence)","hcltech","Indian IT","Infosys","Mphasis","prediction","Quarterly Earnings","TCS"],"author_name":"C P Balasubramanyam","publish_date":"2025-07-03T15:49:01","publication_year":"2025","word_count":957,"keywords":["API","prediction","artificial intelligence","GenAI","Infosys","TCS","AI","R","Mphasis","Git","RAG","Aim","generative AI","GAN","hcltech","Indian IT","Quarterly Earnings","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","generative AI","GenAI","Aim","RAG","R","Git","API","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/indian-it-set-for-tepid-q1-as-global-risks-mount\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10118365,"title":"Zendesk Partners With Anthropic and AWS for Generative AI Solutions","content":"Zendesk has announced a partnership with AWS and Anthropic at its Relate global conference in San Francisco and Las Vegas. This collaboration aims to enhance Zendesk’s AI capabilities by utilising Amazon Bedrock and Anthropic’s Claude 3 model family. These integrations will enable Zendesk’s over 100,000 customers to deploy sophisticated language models tailored to unique customer interactions. Adrian McDermott, Zendesk’s chief technology officer, highlighted the importance of advanced LLM technology in shaping future customer and employee interactions. “Our long-standing partnership with AWS and work with Anthropic means our customers have a CX platform and choice of powerful LLMs to help them set a new standard for service, with AI and automation providing support quickly and effortlessly.” he said. Atul Deo, General Manager of Amazon Bedrock at AWS, expressed enthusiasm about the collaboration. He stated that Zendesk’s use of Amazon Bedrock will empower businesses globally to deliver personalized and efficient support experiences by leveraging generative AI applications with security, privacy, and responsible AI. Zendesk leverages AI to provide instant, intelligent responses to customer inquiries without the need for coding or costly model development. The integration with Amazon Bedrock and Anthropic’s Claude 3 models allows for: Immediate, intelligent support: Anthropic’s models work with Zendesk to provide empathetic, real-time responses, reducing wait times and increasing customer satisfaction. Personalised interactions: Combining Zendesk’s CX data and industry insights with Anthropic’s AI and AWS enables tailored support for each customer’s needs. Improved agent support: AI tools provide agents with necessary information and suggest suitable responses, automating routine tasks and allowing them to focus on complex customer needs. Kate Jensen, Head of Revenue at Anthropic, noted that integrating Claude 3 models with Zendesk and Amazon Bedrock provides businesses with a trusted AI solution that utilises Claude’s multilingual abilities, writing proficiency, and nuanced conversational context comprehension. This integration allows businesses to offer more personalised, efficient customer support across multiple languages and channels, boosting customer satisfaction, loyalty, and revenue growth. “By integrating LLM features into our intelligence panel, we anticipate doubling agent productivity. The panel offers reply recommendations, and conversation summaries, and allows agents to finetune their responses, resulting in enhanced efficiency,” Cristina Fonseca, head of AI, Zendesk, told AIM.","excerpt":"This integration allows businesses to offer more personalised, efficient customer support across multiple languages and channels, boosting customer satisfaction, loyalty, and revenue growth.","categories":["AI News"],"tags":["AWS","Generative AI","Zendesk"],"author_name":"Mohit Pandey","publish_date":"2024-04-17T12:35:36","publication_year":"2024","word_count":360,"keywords":["Anthropic","Zendesk","API","AWS","AI","RAG","Aim","ViT","generative AI","Rust","Generative AI","R"],"extracted_tech_keywords":["AI","generative AI","Anthropic","Aim","RAG","AWS","R","Rust","API","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/zendesk-partners-with-anthropic-and-aws-for-generative-ai-solutions\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10022961,"title":"Hyperconverged Infrastructure And Its Rise To Prominence","content":"Hyperconverged infrastructure (HCI) is a software-defined system that combines all conventional data centre elements–storage, compute, management, and networking. It virtualizes all aspects of a typical hardware-defined system, reduces data centre complexity and increases stability. This integrated solution generally runs on commercial off-the-shelf (COTS) servers. An HCI uses software and x86 servers. How Does HCI Work The heavy legacy infrastructure is traded for a distributed platform running on industry-standard commodity servers allowing businesses to measure workloads accurately and scale when needed. Each server, also known as a node, comes with x86 processors with SSDs and HDDs. The software on each node distributes the operating functions across the cluster to deliver superior performance and resilience. A Hyperconverged platform, broadly, has four software components: Store virtualization, compute virtualization, network virtualization, and advanced management capabilities such as automation. The virtualization software abstracts the underlying resources and then allocates them to applications running on virtual machines or containers. Hardware platform configurations can fit into any workload by independently scaling resources such as CPU, RAM, and storage. It can be provisioned with GPU for further graphics acceleration. All the nodes include flash memory to optimize storage performance. The management pane allows easy administration of HCI resources from a single interface, eliminating the need for separate management solutions for servers, storage network, storage, and virtualization. Trumping Traditional Systems Though data centre infrastructure has been designed around SAN Storage since the 90s, it gained momentum with the virtualization explosion at the turn of the millennium. However, the SAN-based infrastructure can not keep up with the current IT needs due to its inability to scale and other operational challenges. Thus, HCI has emerged as a popular choice for companies. HCI deployment does not involve complex steps, and it has a simple design. Since HCIs are administered by virtualization or cloud administrators, companies don’t require a dedicated team with expertise in storage, compute, or networking. It scales incrementally and operates on a pay-as-you-go cost model. A recent IDC tracker report showed the COVID-19 had no adverse impact on the HCI market. The total system sales reached $2.46 billion in the fourth quarter of 2020. The research also found Dell Technologies, Hewlett Packard Enterprise, and Nutanix are among the leading HCI technology suppliers. HCI vs Cloud Public clouds are flexible and help organizations to adapt to changing business needs dynamically. According to a Gartner report, post the COVID-19 crisis, the end-user spending on public cloud services is expected to grow by 18.4 percent in 2021 to a total of $304.9 billion. Having said that, cloud adoption comes with its challenges. Building and deploying applications in the cloud require specialized skill sets, burdening an organization’s already highly compartmentalized structure. Other challenges with public cloud resources include control and security concerns and high expenses. HCI offers two main advantages over cloud infrastructure: Since the HCI cluster consolidates the hardware components into an integrated infrastructure, it eliminates the bottlenecks that come with distributed architectures. The physical hardware in the cloud infrastructure, which can span across multiple locations, poses a risk of bottlenecks. HCI are easier to manage than clouds. The former offers a high degree of operational efficiency and also streamlines data centre operations. It greatly simplifies the processes of acquiring, deploying, managing, and scaling infrastructure. Notably, HCI and cloud are not mutually exclusive. HCI services can be extended to public clouds for building a proper hybrid cloud infrastructure. Such hybrid settings enable applications to be deployed and managed with the same procedure. A hybrid infrastructure may compromise the cloud’s flexibility but simplifies implementation.","excerpt":"Hyperconverged infrastructure (HCI) is a software-defined system that combines all conventional data centre elements–storage, compute, management, and networking. It virtualizes all aspects of a typical hardware-defined system, reduces data centre complexity and increases stability. This integrated solution generally runs on commercial off-the-shelf (COTS) servers. An HCI uses software and x86 servers. How Does HCI Work […]","categories":["IT Services"],"tags":[],"author_name":"Shraddha Goled","publish_date":"2021-03-29T18:00:00","publication_year":"2021","word_count":590,"keywords":["Go","programming_languages:R","AI","ML","programming_languages:Go","RAG","automation","GAN","R"],"extracted_tech_keywords":["AI","ML","RAG","R","Go","GAN","automation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/hyperconverged-infrastructure-and-its-rise-to-prominence\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":5,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10122884,"title":"Indian AI and Robotics Startup Swaayatt Robots Secures $4 Mn to Advance Level 5 Autonomy","content":"Bhopal-based Indian AI and Robotics startup Swaayatt Robots recently raised $4 million at a valuation of $151 million from US investors. The company will further raise $11M at a valuation of round $175M-$200M. “In 2021, we raised $3 million, of which we have utilised only $650,000. So in total, we now have $6.3 million to carry our research forward. We are already operating in cities, highways, and off-road locations,” said Sanjeev Sharma, chief of Swaayatt Robots, in an exclusive interview with AIM. “By the end of this year, we’ll be creating a blueprint that could solve Level Four Autonomy globally. To scale that model, we may raise around $1.5 billion,” said Sharma. In 2021, investors worldover poured a record $9.7 billion into autonomous vehicle development. However, last year that amount dropped by nearly 60% to just $4.1 billion. According to a McKinsey report, autonomous driving is the future, with self-driving vehicles projected to become a $300 billion to $400 billion revenue opportunity by 2035. Sharma added that the company is planning a major demo in August. “If the August demo doesn’t excite Sam Altman and Elon Musk, along with the world, I don’t know what will,” he said. Earlier this year, the company executed a demo where it claimed to have achieved Level 5 Autonomy. The autonomous driving scene in India is still in its nascent stages, but several startups like Minus Zero, Flowdrive, Flux Auto, and Netradyne are actively working to develop self-driving technologies designed for the country’s chaotic road conditions. Recently, Swaayatt Robots conducted a demo where their autonomous vehicle avoided piles of soil and rocks dumped on the road for construction purposes. Negotiating a construction site with well-structured traffic cones is typically considered a challenge by the autonomous driving industry in the West. “You will notice our vehicle made a 90-degree turn at an unmanaged intersection,” said Sharma, adding that it is a huge problem even for companies like Tesla and Waymo. Last year in October, the company enabled autonomous vehicles to negotiate bidirectional traffic on single-lane roads. The demo was conducted at a relative speed of 44 kilometers per hour. “Now, that R&D is going to be scaled up to an extent where the relative velocity at the point of crossing can be maintained at 60 kilometers per hour,” said Sharma. Focus on Autonomous Driving R&D As of now, Swaayatt Robots hasn’t really sold their technology to anyone and has been conducting demos with Mahindra Bolero. However, in the near future, the company will focus on OEM integration in existing vehicles, specifically targeting the military and autonomous trucking market. “For autonomous driving markets in the military domain, we cannot expect new vehicles to be operated in service, you would want to develop a system that can be retrofitted into these aftermarket vehicles,” said Sharma. He added that aftermarket segments in autonomous trucking in North America are humongous. “If you look at the US trucking market, I mean, it’s an $800 billion addressable market,” said Sharma. He further said that Swaayatt Robots has been in talks with some of the fleet owners in the US. Regarding India he said, “We have understood the requirements of almost every single OEM in the country. We are in touch with most of the OEMs in India, except for Mahindra. It’s ironic because Anand Mahindra recently praised us.” Even though we are not in touch with Mahindra directly, he explained that Mahindra’s vehicles are robust and the easiest to modify. “Mahindra has this Mobileye system, which is incorporated into the XUV700.” Better Than Tesla? Sharma likes to position Swaayatt as an R&D company, and unlike Tesla at this stage, it is not planning to sell cars anytime soon. “What we are doing is very sophisticated R&D in the general autonomous navigation domain to make autonomous driving happen,” he said. “There is no agency on the planet, be it Musk’s Tesla or Altman’s OpenAI, that is working on bi-directional traffic,” he said, adding that in the October 2023 demo, their vehicle successfully negotiated bi-directional traffic on a single lane. He added that by 2030, only four to five autonomous driving companies will survive, highlighting the immense R&D challenges involved. He emphasised that safety is a crucial area to address. Referring to GM’s Cruise accident, he pointed out that it resulted in a total halt of operations and several investigations. Sharma is a huge admirer of Wavye AI. “Since 2019, with the advent of Wavye AI, people have been discussing autonomy without relying on maps. We were the first technology to enable vehicles without the reliance on high-definition maps. In 2017, we implemented multi-RL agents without requiring any maps,” he said. “We have done R&D in the perception domain to an extent that there are now only 5 problems where supervised learning is required, and for those problems as well we are developing self-supervised models,” he added. Until now, the company has been conducting demos exclusively with Mahindra Bolero. In the near future, the company plans to demonstrate with multiple vehicles, including Thar and Fortuner. “One demo we will be doing will have Thar and Bolero crossing each other in an autonomous fashion,” concluded Sharma.","excerpt":"The company will further raise $7 million at a valuation of $175 million.","categories":["Deep Tech"],"tags":[],"author_name":"Siddharth Jindal","publish_date":"2024-06-08T10:02:30","publication_year":"2024","word_count":862,"keywords":["Go","OpenAI","AI","programming_languages:R","programming_languages:Go","ai_applications:autonomous driving","Aim","ai_applications:robotics","R","startup"],"extracted_tech_keywords":["AI","OpenAI","Aim","R","Go","startup","programming_languages:R","programming_languages:Go","ai_applications:autonomous driving","ai_applications:robotics"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/indian-ai-and-robotics-startup-swaayatt-robots-secures-4-mn-to-advance-level-5-autonomy\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10100691,"title":"MongoDB Unveils Generative AI Features for Atlas Vector Search","content":"Data forms the crux of generative AI applications. And, MongoDB’s vector search capabilities are driving generative AI by transforming diverse data types like text, images, videos, and audio files into numerical vectors, simplifying AI processing and enabling efficient relevance-based searches. The company has unveiled a series of features in MongoDB Atlas Vector Search that offer several benefits for generative AI application development. Boost Information Accuracy for LLMs: Generative AI applications aim to provide precise and engaging experiences, but they can sometimes hallucinate information due to a lack of context. By expanding MongoDB Atlas’s query capabilities, developers can create a dedicated data aggregation stage with MongoDB Atlas Vector Search. This helps filter results from proprietary data, significantly improving information accuracy and reducing inaccuracies in AI applications. Speed Up Data Indexing for Generative AI Applications: Generating vectors is crucial for preparing data for use with LLMs. After creating vectors, an efficient index must be built for data retrieval. MongoDB Atlas Vector Search’s unified document data model simplifies the indexing process for operational data, metadata, and vector data, facilitating faster development of AI-powered applications. Use Real-Time Data Streams: Developers can leverage Confluent Cloud’s managed data streaming platform to power real-time applications. Through the Connect with Confluent partnership, Confluent Cloud data streams can be integrated into MongoDB Atlas Vector Search. This integration offers generative AI applications access to real-time, accurate data from various sources across a business. By using a fully managed connector for MongoDB Atlas, developers can make their applications more responsive and provide users with more precise results that reflect current conditions. Read more: MongoDB Ups the Ante with Vector Search for Generative AI Customer Success Stories Several organizations are utilizing MongoDB Atlas Vector Search to enhance their services. Data innovators company Dataworkz is merging data, transformations, and AI to create high-quality, LLM-ready data for AI applications, while Auto API platform Drivly is employing AI embeddings and Atlas Vector Search to empower AI car-buying assistants. Risk analytics firm ExTrac is using it to augment LLMs and analyze various data modalities, including text, images, and videos, for real-time threat identification. Inovaare Corporation leverages MongoDB to improve healthcare compliance operations, thanks to the capabilities of Atlas Vector Search in reporting and data-driven insights. NWO.ai enhances its consumer intelligence platform by integrating Atlas Vector Search to search and analyze embeddings for real-time insights. Finally, One AI uses it to enable semantic search and information retrieval, enhancing customer experiences. Cyber security company VISO Trust is also employing it to provide comprehensive vendor security information, streamlining decision-making for risk assessments. Read more: Is MongoDB Vector Search the Panacea for all LLM Problems?","excerpt":"MongoDB Atlas Vector Search provides the functionality of a vector database integrated as part of a unified developer data platform, allowing teams to  store and process vector embeddings alongside virtually any type of data to more quickly and easily build generative AI applications.","categories":["AI News"],"tags":[],"author_name":"Shritama Saha","publish_date":"2023-09-26T18:29:25","publication_year":"2023","word_count":435,"keywords":["Go","semantic search","AI","MongoDB","ML","RAG","Aim","generative AI","analytics","R"],"extracted_tech_keywords":["AI","ML","analytics","generative AI","Aim","RAG","semantic search","MongoDB","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/mongodb-unveils-generative-ai-features-for-atlas-vector-search\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10122113,"title":"Unleashing Intelligence: Leveraging AI to derive Insights at Scale","content":"Generative AI has proven to be an impressive tool for industries in generating insights from a large volume of data. Parmjeet Virdi Director Data Analytics at Publicis Sapient, reveals while speaking at AIM Media House’s Data Engineering Summit (DES) 2024 that Publicis Sapient has been leveraging generative AI to derive insights from customers data and helping them make better business decisions. To make her point, Virdi referred to an incident of an analyst reviewing a lengthy report by Harvard Business Review. She said an analyst overlooked crucial details in a lengthy report, missing the competitor’s expansion strategy. However,AI, unlike humans, connected scattered information, revealing the competitor’s plan to penetrate a low-cost market by leveraging specific product manufacturing and later market expansion, enhancing analytical insight. The initial disruption caused by the internet paved the way for subsequent technological advancements, including the evolution of web, mobile, IoT, and connected systems. So when these technologies are evolving, at the same time, the customers are also evolving,” Virdi said. Along with Aashima Kumar, senior manager, data analytics at Publicis Sapient, discusses how Publicis Sapient has harnessed the power of generative AI, Machine Learning models, and advanced Data Analytics techniques to unearth hidden patterns, trends, correlations, and influential features within extensive data repositories. Deriving insights from data Moreover, Kumar pointed out that Publicis Sapient built a solution for a large automobile company that helped them drive sales by better understanding customer data. “So when any potential customer comes to their website or mobile app, they see a very fancy-looking configurator that gives the customer the option of selecting the model, grade, color, engine, upholstery, and accessories, plus all the subscriptions for the connected cars. “Now that so much of data coming out of configurator and talking about this data, it actually has the power for us to understand the needs of the customer, at the same time provide back to the business a way that they can understand what kind of combinations are more profitable for them and in more demand.“We utilised conjoint analysis and Monte Carlo Markov analysis to identify coefficients for customer choices, generating insights visualised through a utility dashboard,” Kumar said. This internal tool aided product planning and experience teams by revealing utility scores, indicating component attractiveness and profitability trade-offs, and enhancing decision-making across multiple business levels, including product combinations and profitability assessment. Leveraging generative AI Marriot, one of the world’s largest hotel brands, had a traditional search engine, Kumar pointed out. Marriott’s traditional search process needs users to input destinations, dates, and preferences, among other things. However, this method leaves much to be desired in aiding decision-making, offering limited assistance. “We transitioned their traditional search to a generative AI-powered search, enabling users to input specific needs in natural language rather than filters. The generative AI model then extracted these queries into attributes and identified relevant accommodations. “We scaled it to the next level to make it a conversational AI. Using the same power that we were able to engage with Gen AI, we made sure that it’s not just giving you the options of what you’re looking for. But why can’t we recommend you more options and then you make the choices that you want,” Kumar said.","excerpt":"Generative AI has proven to be an impressive tool for industries in generating insights from a large volume of data. Parmjeet Virdi Director Data Analytics at Publicis Sapient, reveals while speaking at AIM Media House’s Data Engineering Summit (DES) 2024 that Publicis Sapient has been leveraging generative AI to derive insights from customers data and […]","categories":["AI Features"],"tags":[],"author_name":"Pritam Bordoloi","publish_date":"2024-05-31T13:00:00","publication_year":"2024","word_count":536,"keywords":["API","machine learning","AI","RAG","Aim","data engineering","generative AI","analytics","disruption","R"],"extracted_tech_keywords":["AI","machine learning","analytics","generative AI","Aim","RAG","R","API","data engineering","disruption"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/unleashing-intelligence-leveraging-ai-to-derive-insights-at-scale\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10102328,"title":"How Google will Change Search Forever","content":"While on the one hand, Google is embroiled in an ongoing antitrust case , on the other, the tech giant is soaring with its recent Q3 earnings. Alphabet, Google’s parent company, saw an 11% revenue growth, returning to double-digit figures after a year. CEO Sundar Pichai was not only pleased with the results, but also emphasised on the robust growth of Google Search, the largest contributor to its revenue growth. With AI capabilities being brought through search generative experience (SGE), the company is on the path to change Search forever. SGE to Lead the Way Google continues to dominate the search engine market by a huge difference, and is still building features to push it ahead. With 8.5 billion searches on Google every single day, and a 90% market share of the search engine market, the company is comfortably placed way ahead of its competitors. And, they are not stopping at that. Source: Similarweb In the recent earnings call, Google mentioned that with generative AI being integrated in Search, SGE which is currently available in the US, Japan and India, a broader range of queries with multiple perspectives can be addressed. Google is also now linking to a wider range of sources on the results page, for pushing content discovery. Interestingly, their improved search feature will bolster their advertiser revenue. Google Search Generative Experience gains \"supportive\" links in the AI-generated description of the source https:\/\/t.co\/1vkrGS7BMQ pic.twitter.com\/caJTQ3bLy3— Barry Schwartz (@rustybrick) October 26, 2023 Search and Advertisers Go Hand-in-Hand In May, Google had launched SGE, an experiment in Search Labs, in the US alone, to test AI integration in Search, and the goal was to get user feedback and further build on it. The experiment seemed to have paid off with Pichai expressing his pleasure on the positive feedback received from SGE users, and also confirming about rolling it out to more users. Notably, out of the $76.69 billion revenue for Q3, the advertising business accounted for $59.65 billion. Pichai also emphasised on how ads will continue to ‘play an important role’ in this new search experience. The company will continue to experiment with new formats in SGE to create relevant, high-quality ads that are customised at every stage of the search journey. Lead tech analyst a​t I\/O, Beth Kindig mentioned that SGE will promise advertisers better ROI. Google’s Search revenue of $44B increased +11.3% YoY, the highest growth rate since Q2 ‘22.In August, I wrote about Google's dominance in search amidst the buzz around ChatGPT. Alphabet's upcoming Search Generative Experience (SGE) promises advertisers better ROI, poised to…— Beth Kindig (@Beth_Kindig) October 26, 2023 Google had suggested their plan to introduce ads in SGE during the I\/O event in May, showing a glimpse of how ads will look in the future. Integrating ads into AI search, akin to how Microsoft pushed ads into their AI-powered Bing Chat, Google looks to make the platform wholesome. The company is also building generative AI templates that combine the capabilities of LLMs with Google APIs to solve industry-specific use cases such as travel. There’s More… While Google is working on enhancing its existing search experience, its large language model Bard continues to progress on the sideline. Recently, Google introduced browser extensions for Google apps such as YouTube, Google Flights, Google Maps, and Google Drive, Google Docs, and Gmail on Bard. Furthermore, Google also announced the integration of Bard on Google Assistant in an attempt to provide a more intuitive, intelligent and personalised digital assistant. While OpenAI only recently went completely multimodal, allowing voice and image input, and even released an update on allowing upload of any form of documents including PDFs, Bard was multimodal way before that. Furthermore, Pichai spoke about the much-anticipated Project Gemini. “I’m very excited at the progress there and as we’re working through getting the model ready.” He also hinted at it being multimodal, and ‘highly efficient with tool and API integrations’. With AI being the main focus, and even making it to big tech companies’ quarterly earnings (Google said AI at least 76 times), the road ahead is laden with AI-driven developments and integration for these companies. Search being the key revenue driver, Google’s efforts to make it more enhanced is undeniable.","excerpt":"In Google’s Q3 earnings report, Search was the largest contributor to its 11% revenue growth. Now with search generative experience, it is only going to get better","categories":["Global Tech"],"tags":["Alphabet","antitrust","API","Bing","Gemini","Generative AI","Google","Google Assistant","Google Search","japan","OpenAI","Search engine","Sundar Pichai","US"],"author_name":"Vandana Nair","publish_date":"2023-11-01T10:00:00","publication_year":"2023","word_count":701,"keywords":["API","Bing","Git","Rust","Generative AI","R","japan","ChatGPT","Sundar Pichai","Go","Gemini","AI","Google Search","generative AI","Alphabet","Google Assistant","Search engine","OpenAI","GPT","antitrust","Google","US"],"extracted_tech_keywords":["AI","generative AI","ChatGPT","OpenAI","R","Go","Rust","Git","API","GPT"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/how-google-will-change-search-forever\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":33825,"title":"Artificial Intelligence Brings IITians &#038; Thought Leaders Together At Pan IIT Conclave 2019","content":"Source Image -The News Minute Vijay Goel, Minister of State for Statistics and Programme Implementation inaugurated the two-day Pan IIT Conclave 2019 on the theme  “Artificial Intelligence: What does the future hold for India?”. Goel discussed reinforcing the application of AI and ML across industries to propel India to be future ready. The Keynote speaker, K Ananth Krishnan, Chief Technology Officer, TCS, shared his vision for AI’s future in India. Unique, out of box thoughts and visions were emphasized on the significance and contribution of AI in the future of India and multiple speakers shared their ideas in line with addressing the potential challenges. On the first day of the event, discussions were focused towards multiple smaller themes like AI for Social Good, AI for Education, Future of Work, AI in Manufacturing, Ethics, Privacy and Security of AI Systems, Security of AI Systems and AI for Agriculture. The panel on AI for Social Good led by P Anandan, CEO of Wadhwani AI spoke of how AI can help us increase the impact of doing good for the society. The session on AI for Education showcased the role of AI in redefining India’s education landscape with a greater focus on improved learning outcomes as well as on the opportunities and challenges within the education sector. The potential of how AI can improve processes and quality in manufacturing industry was also discussed. The focus was on how AI can help improve the safety of people working in manufacturing processes and eventually increase efficiency. Discussion in the arena of  Ethics, Privacy and Security panel with respect to Artificial Intelligence were too discussed. “Key is content creation and curation of information. AI is really helping with predictive suggestions towards sieving information on the web. Hyper-personalization is now possible for you to have test applications to train you better on your weaknesses. We need to focus on creating more engaging content for the students to build connections.” said Sudha Ram, Professor, CSE and IS, the University of Arizona and Anand Rangarajan, Director, Google India at the event. Finally, the session AI for Agriculture shed light on the multitude potential of technology to enhance the efficiency and productivity of the agricultural sector and leadership. According to Sriram Raghavan, VP, IBM Research & CTO IBM India, India needs digitization in agriculture accompanied by educating our farmers about its use.","excerpt":"Vijay Goel, Minister of State for Statistics and Programme Implementation inaugurated the two-day Pan IIT Conclave 2019 on the theme  “Artificial Intelligence: What does the future hold for India?”. Goel discussed reinforcing the application of AI and ML across industries to propel India to be future ready. The Keynote speaker, K Ananth Krishnan, Chief Technology […]","categories":["AI News"],"tags":["AI (Artificial Intelligence)","AI India"],"author_name":"Martin F.R.","publish_date":"2019-01-22T03:46:02","publication_year":"2019","word_count":392,"keywords":["Go","artificial intelligence","programming_languages:R","AI","ML","programming_languages:Go","Git","RAG","ViT","AI India","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","ML","RAG","R","Go","Git","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/ai-iitians-leaders-pan-iit-conclave-2019\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":51006,"title":"How Gargi Dasgupta is Building AI-Driven Processes For IBM Research","content":"With more than 4,000 patents registered in 2018 in technologies like AI, cloud computing, security and quantum computing, IBM inventors from India received over 800 of patents, the second highest contributor to the global record tally. Leading the charge, it is Gargi Dasgupta who coordinates everything to bring innovation to the forefront. In an interaction with Analytics India Magazine, Gargi Dasgupta – Director, IBM Research India and CTO, IBM India and South Asia shares the various facets of her research work, and the way forward. Gargi’s Passion For AI “After I completed my PhD I returned to India to pursue this passion and fulfil my ambition to contribute and drive this change and while doing so help establish India as a leader in Artificial Intelligence (AI),” told Gargi Dasgupta. Gradually Gargi leveraged this expertise in creating leadership for IBM in AI-driven business processes and IT automation – embedding AI capabilities into IBM’s Automation platform, bringing together assets and capabilities from our worldwide research into business use and establishing right relationships with industry verticals and services partners to drive impact. She has helped IBM optimise its own product support business via AI infusion led to a significant impact measured in several tens of millions of dollars through innovations like enriched contextual search, question assistant for interactive customer problem determination, cognitive routing and many others. “For any AI to be successful it has to reason and act; get better over time and interact with the natural interface. So, this pillar of AI is the most complex as it helps us push the boundaries of what we can achieve,” said Gargi. For any AI to be successful it has to reason and act; get better over time and interact with the natural interface.Gargi Dasgupta, CTO- IBM Research India What Does It Mean To Lead IBM Research? According to Gargi Dasgupta, her research agenda is anchored in three main areas: advancing the capabilities of core AI technologies in areas of language and reasoning, creating AI technologies that can be trusted, are explainable and fair and embedding and operationalising these AI capabilities into businesses. Applying these technologies, IBM are building platforms for Agriculture (Watson Decision Platform for Agriculture), Retail (Cognitive Retail Enterprise) , Smart Contracts and supply chains (with Blockchain) and driving Enterprise Automation. “Our Watson Decision Platform for Agriculture – which uses the power of AI and weather data, is helping farmers make informed decisions about their crops. The platform provides tools and insights from sowing to crop health monitoring and harvesting to credit scoring of farmers. It has garnered a lot of interest by the Government, policymakers, corporates etc.,” told Gargi Dasgupta. In India, under Gargi’s leadership, IBM Research all did a pilot project with NITI Aayog for their aspirational districts, collaborated with Karnataka Agri Price Commission for price forecasting of tomatoes and recently we signed a Statement of Intent with Ministry of Agriculture for using this platform in three districts in India. Gargi Advises On Upskilling and Learning In this constant world of VUCA (volatility, uncertainty, complexity, and ambiguity) people must be agile and find time to upskill. They should adopt a protean career approach and continuously be looking at the skills they need to acquire. Upskilling could be through courses, dedicated technology reading, business overviews etc. According to Gargi Dasgupta, the Education sector has a major role to prepare the workforce for the age of automation through its curricula and programs. “I feel enterprises and Government share an equal responsibility to ensure that our workforce is skilled for tomorrow. We need a system that encourages skill-building, critical thinking, problem-solving, being agile. If India wants to become an AI force, we must build the right talent pipeline and the entire ecosystem needs to be invested in creating a platform that will help our agenda,” said Gargi Dasgupta. If India wants to become an AI force, we must build the right talent pipeline and the entire ecosystem needs to be invested in creating a platform that will help our agenda.Gargi Dasgupta, CTO- IBM Research India","excerpt":"With more than 4,000 patents registered in 2018 in technologies like AI, cloud computing, security and quantum computing, IBM inventors from India received over 800 of patents, the second highest contributor to the global record tally. Leading the charge, it is Gargi Dasgupta who coordinates everything to bring innovation to the forefront. In an interaction […]","categories":["AI Features"],"tags":["ibm research","Interviews and Discussions","process automation"],"author_name":"Vishal Chawla","publish_date":"2019-12-02T17:16:55","publication_year":"2019","word_count":672,"keywords":["Go","artificial intelligence","AI","cloud computing","ibm research","innovation","RAG","automation","analytics","Rust","process automation","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","artificial intelligence","analytics","RAG","cloud computing","R","Go","Rust","automation","innovation"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/gargi-dasgupta\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10092493,"title":"This AI Startup is Building Better Tools Than AWS Sagemaker and Google AutoML","content":"There are different generations of tools when it comes to automated machine learning. Legacy tools such as AWS Sagemaker, Google AutoML etc have existed for quite some time now and the key goal of these tools was to help machine learning engineers and people that are technical themselves. But, now there are even better tools. Nirman Dave, CEO at Obviously AI claimed that their models are better than those of AWS Sagemaker or Google AutoML. Based in San Francisco, Obviously AI specialises in building no-code AI models for businesses. Dave founded the startup with a mission to transform every company into an AI company, boasting of having developed the most rapid and accurate no-code AI tool to date. “So what happens is that the customer in just a few minutes gets access to these machine learning models that they can build or customise for themselves,” explained Dave. However, on the other hand, legacy tools that have been around for nearly 13-15 years now and are very slow at building models. “What’s really special about what we do is we build the models in less than a minute. So we’re the fastest tool to build AI models today” Dave added. Nonetheless, Dave also acknowledges that his company is not the only one building similar tools. But even though there are other companies that have built no-code AI tools, their approach is very different. “We mostly focus on tabular data, supervised learning. We have now kind of branched into unsupervised learning as well with our custom Large Language Models (LLMs) offering.” But Levity AI, for example, another company building no-code AI tools, focuses mostly on image, video or audio type of data. Demand Founded in 2021, so far Obviously AI has boarded 52 customers. Even though most of them are in the US, the startup  has customers in India, Japan, and also South Africa. “Currently, week over week we’re growing about 15% in terms of customer acquisition,” he said. One of the big customers the startup has in India is a large consumer bank. Dave revealed that they are building a loan repayment model for this bank. Normally, the underwriting process for the loans takes a significant amount of time when done manually. “They wanted to build a model that can quickly process the loan, predict default probability and give it out to the underwriters to make a decision.” Dave said. Manufacturers of AI models The startup is engaged with mostly mid-market businesses that don’t have a data science team or cannot scale one and also enterprise businesses that might have a backlogged data science team. “Essentially, we are manufacturers of AI models,” Dave said. The process begins with the customer bringing a collection of data to our platform and specifying their desired prediction or AI model based on that data. “And we help with everything from data cleaning, to model selection to hyper parameter tuning, to model deployment and management, all of that process is done automatically by the system,” Dave added. The user interface is also designed for people that are not heavily technical. Further, another thing that Obviously AI provides, and which they excel in, is dedicated data science support. “So essentially, we work with a lot of data scientists. These are individual practitioners, individuals that run their own consulting firms or folks that are just starting out, and we provide them with connections to these customers that we have,” Dave said. Impact on jobs On the contrary, there is a growing concern that no-code tools could have a negative impact on jobs, particularly those that involve coding and software development. As these tools become more sophisticated and accessible, they may reduce the need for traditional coding skills in certain roles, leading to a shift in the job market. However, Dave believes Obviously AI’s no-code tool is not replacing data scientists, instead it is accelerating the data science process. “The reason companies like Hewlett-Packard use us is not because they want to get rid of their data science team. It’s actually that they want to accelerate the data science team. “The goal here is to help them move significantly faster. The data science team is going to take two months to get from raw data to insights and analytics and predictions. Now it’s going to take them a week of a couple of days.” Besides, a data scientist’s job is not solely to build AI models; but to build a strategy. “So I don’t think data scientists will be replaced. I think no-code tools will only help them focus more on the strategy which is the most exciting part of the job,” Dave concluded.","excerpt":"The startup’s goal is to serve people that know the basics of data science, and really help them accelerate on that front","categories":["Deep Tech"],"tags":["data science tools"],"author_name":"Pritam Bordoloi","publish_date":"2023-04-29T14:00:00","publication_year":"2023","word_count":775,"keywords":["data science","Go","data science tools","machine learning","API","AWS","AI","ML","Aim","analytics","R"],"extracted_tech_keywords":["AI","machine learning","ML","data science","analytics","Aim","AWS","R","Go","API"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/this-ai-startup-is-building-better-tools-than-aws-sagemaker-and-google-automl\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10054452,"title":"Council Post: Emerging Technologies &#038; How They Are Revamping The Marketing Landscape","content":"For a long time, India has been a suppliers market. The focus on the customer and their needs are limited, with the market centring around sales and products. But technology is incentivising the market to rethink the business structure and re-build the industry that uses modern marketing techniques to interact with consumers. The Gartner’s CMO spending survey 2020-2021 revealed that CMOs remain bullish on technology for marketing by setting bigger budgets for marketing and digital advertising. However, enterprises use only 58% of their existing technology capabilities – indicating a huge undiscovered scope for marketing. A collaborative relationship between marketing and technology has a huge potential. This has been steadily evolving to create what is today known as the MarTech landscape. MarTech has over 8,000 solutions with businesses investing in AI\/ML, IoT, AR\/VR, and other fast evolving technologies. The new technologies and tools provide customers with a plethora of channels to interact with brands. The new marking lifecycle consists of engaging with and retaining the right consumers by leveraging the latest technologies. The MarTech stack consists of the technologies and approaches to enhance organisational focus on the consumer’s needs and wants rather than the product. Focus on Customer Experience & Data Consumer and data are at the centre of MarTech. The Portada Insights Report for 2021 analysed changes in MarTech investment behaviour over a year, the findings of which show that investments were majorly made in the traditional elements of advertising and promotion relating to customer experience and data. There is a shift in the market, reducing the focus on traditional advertisements and instead using user data to make data-driven and customer-oriented investments. This includes investing in analytics, AI, virtual reality, cloud computing and several other market automation tools. Organisations are leveraging these MarTech tools to create a connected consumer experience in an entirely new manner. The data-based insights created by these tools help organisations understand the consumers’ minds and connect with them throughout the customer journey cycle. Reaching the Consumers effectively The first important step to engage the customer is to reach them correctly. Enterprises are heavily investing in tools such as Content Management Systems to manage the organisation’s web pages, including blogs, articles, podcasts, videos or images. The brand’s website is at the heart of user experience, and the organisation’s advertising generally drives the audience towards the website. The age of increased digital competition makes it very easy for users to judge and accept\/reject a brand at its first appearance. The digital equivalent of having similar stores next to each other is fighting competitors to show up first on the SEO page. This is especially true given the pandemic and its boost to e-commerce. The focus is for organisations to hone the search engine by leveraging MarTech tools for SEO optimisation. Based on data-driven insights, companies can understand their target consumers. It is important to identify the market trends in the organisation’s target group and to correctly place their content strategy to ensure they receive traffic from their target audience. While reaching their target audience, companies need to market their products\/services as relevant and personal to the consumers. The MarTech industry is all about personalising content; be it in emails or social media posts, the marketing is done to the specific segment of the group across various platforms. Personalisation allows organisations to connect with customers during the different stages. After this, they can leverage the micro-moments in the customer journey to place the correct automated tools that assist marketers in ensuring constant contact with customers during all of these stages. This helps in improving brand awareness, brand loyalty and engagement. Newvamped Engagement Technologies AI tools play a huge role in reaching consumers effectively. They have been developing for quite a while and are being leveraged by organisations. Adding to those AI tools, the MarTech stack identifies and comprises some of the most modern technological advancements in AI\/ML. These enable customers to have the latest and most immersive experience. For example, chatbots and virtual assistants ensure that customer queries are replied to instantly, creating a more connected experience and increasing engagement. As a bonus, the bots also use the interactions to generate user data that circles back to creating better-personalised content. Experiential marketing through AR\/ VR is also changing the face of the industry by allowing customers to try the products before they consider buying it. Organisations are offering virtual high-end try-ons that allow consumers to try on the frame of glasses, stimulate the shade of lipstick, check if the dress they like is a little too short, or even see how the furniture would fit into their living rooms. The integration of virtual reality redefines the customer experience, creates unique brand identities, and overcomes traditional challenges such as store returns or damaged products mess. Wrapping up The technologies are highly evolved, and we have only touched a small share of the MarTech potential. MarTech is changing the nature of marketing. It is creating a market that builds deep relationships with the customers and offers them products and services before the need for it even arises. This demands for consumer marketing to become a continuous process and is made in an agile manner of content creation to deliver constant engagement. As we saw, one of the most critical manners to ensure this is with the collaboration of the marketing department with the IT teams, leveraging consumer data like never before. This article is written by a member of the AIM Leaders Council. AIM Leaders Council is an invitation-only forum of senior executives in the Data Science and Analytics industry. To check if you are eligible for a membership, please fill the form here.","excerpt":"For a long time, India has been a suppliers market. The focus on the customer and their needs are limited, with the market centring around sales and products. But technology is incentivising the market to rethink the business structure and re-build the industry that uses modern marketing techniques to interact with consumers. The Gartner’s CMO […]","categories":["AI Features"],"tags":["Marketing","marketing analytics","marketing data","marketing India","marketing technology"],"author_name":"Neeraj Pratap","publish_date":"2021-11-29T16:00:00","publication_year":"2021","word_count":943,"keywords":["data science","marketing technology","AI","chatbots","ML","marketing analytics","cloud computing","virtual assistants","marketing India","RAG","Aim","analytics","R","Marketing","marketing data"],"extracted_tech_keywords":["AI","ML","data science","analytics","Aim","RAG","chatbots","virtual assistants","cloud computing","R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/council-post-emerging-technologies-how-they-are-revamping-the-marketing-landscape\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":17217,"title":"Guide to surviving the Artificial Intelligence revolution","content":"At a time when humanity is peering into an AI-led future and next-gen technologies such as driverless cars will soon become a reality on roads, artificial intelligence has amassed substantial mixed coverage, with concerns surrounding automation of jobs and even AI-led cyber-attacks. There are also concerns about advanced robotics tech that could access UAVs and carry out attacks on government and cause substantial damage. According to writer Cathy O’Neil, algorithms today have evolved into “weapons of math destruction” and the blind faith in big data has also led to skewed results in various sectors such as in hiring (deciding who lands job interview), loan disbursal, insurance and more. What’s adding more public hysteria is the doomsday statements by leading physicists and technologists about AI going rogue and wiping out humanity, with some even believing that we are revisiting Isaac Asimov’s worst fears expounded in his 1942 sci-fi short story, Runaround where AI was woven into everyday life. The extinction risk to humanity has found resonance with Stuart Armstrong, Oxford university professor as well who believes that AI poses a “genuine threat to humanity”. He is of the view, particularly in relation to jobs and security that society should adjust itself to AI before AI adjusts to us. Seems like the dystopian views of AI doomsayers have shaken up tech giants and the leading proponent of AI – Google, released a research paper last year, titled Concrete Problems in AI Safety detailing five practical research problems that could pose a threat to society in the future.  The research was conducted in collaboration with OpenAI, Stanford and Berkeley to develop practical approaches for building AI systems that function reliably and safely. According to news reports, ten million dollars in funding have been set aside for research on ensuring that a safe AI for future. So, is AI a real threat? In terms of employment, there is definitely a real shift in skillset with leading AI experts Andrew Ng and Kaggle founder Anthony Goldbloom talking about the jobs one can lose to AI. And research around futureproofing AI systems from going rogue has reached fever high with scientists mulling over a possibility of “highly capable AI” that could surpass humans in all domains. Analytics India Magazine presents a Guide to Surviving the AI Revolution Work Redefined: Work will definitely be redefined and in a lot of cases it will payoff as well by enabling engineering teams and medical staff to focus on higher priority jobs. However, before we dive into the kind of jobs that will be made redundant and those won’t, let’s understand how AI capabilities can siphon off our jobs. Here’s how we break it down: agent type AI can perform tasks with limited or no human guidance, advanced AI is trained to perform domain specific tasks such as diagnosing eye diseases, legal assistants et al. Where AI poses a real threat is when highly capable AI will outperform humans in practically all domains. According to Goldbloom, sophisticated algorithms will chip away repeatable tasks where no intuition is required, such as in manufacturing, retail, finance, insurance and in food production industry. However, certain sectors such as construction industry, civil engineering require a high degree of human intuition and intervention. So, they are future proof. And while most writing jobs will also go the AI way, an excellent marketing tagline will still be written by a human, said Goldbloom at a recent TED talk. Exposing AI’s Big Vulnerability: AI is only as good as the experts who train the machines and with a rise in A I startups, the tech landscape needs more machine learning and deep learning engineers. Who knows this better than AI expert and evangelist Andrew Ng whose machine learning course was a huge sellout and serves as a great starting board for machine learning enthusiasts. Ng recently launched his new course DeepLearning.ai to help add more deep learning engineers in the AI space. Another area that is seeing a major upshoot is basic programming and there is a groundswell of support around coding as a life skill. One doesn’t need to have a STEM background to pick up a computer language and the new skill can open several doors for professionals. One can join coding bootcamps or TeachCode Academy and avail several free online resources to get into an IT track job. AI will give high-tech a boost: It may sound like a death knell for many IT professionals involved in support level functions, however AI will enable a smarter IT workforce that can focus on solving more complex challenges.  In fact, with rising fear of AI-led cyber-attacks, we would need more ethical hackers that can build software to tackle these highly capable AI systems. Another area where the IT behemoths can invest is in building more transparent AI systems, by deepening the decision-making process of AI. Social scientists will to come to the fore: Besides the economic shift, there needs to be a broader focus on the creation of jobs in AI-enabled economy, societal risks and how to assuage the AI-driven unemployment. This requires economists, policy makers and security experts to come up with a policy framework around AI.  Social research workers and policy makers have a new job on hand to come up with various arguments on a man + machine future and how the two should co-exist. We believe new research would be coming out of Government backed research centres around the societal impact of AI and how it upended the jobs market in India.  So, all those humanities students can shift their research and find out how AI can be instrumental in bettering society.","excerpt":"At a time when humanity is peering into an AI-led future and next-gen technologies such as driverless cars will soon become a reality on roads, artificial intelligence has amassed substantial mixed coverage, with concerns surrounding automation of jobs and even AI-led cyber-attacks. There are also concerns about advanced robotics tech that could access UAVs and […]","categories":["IT Services"],"tags":["Andrew ng deeplearning.ai"],"author_name":"Richa Bhatia","publish_date":"2017-08-25T08:55:27","publication_year":"2017","word_count":940,"keywords":["big data","Andrew ng deeplearning.ai","Go","machine learning","artificial intelligence","OpenAI","AI","RAG","deep learning","analytics","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","deep learning","analytics","OpenAI","RAG","R","Go","big data"],"url":"https:\/\/analyticsindiamag.com\/it-services\/guide-surviving-artificial-intelligence-revolution\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":38595,"title":"With 3Bn Transactions, UPI Scripted A Mobile First Approach For India","content":"Credit and debit cards have seen increased usage online, indicating the general shift of society towards being cashless. The Indian banking system has also not been a laggard in deploying systems for the easy transfer of money online This is seen by the National Payments Corporation of India’s impressive rollout of the Instant Money Payment System [IMPS]. The interface is regulated by the Reserve Bank of India and works by instantly transferring funds between two bank accounts. However, the system comes with its own pitfalls, as the system can take anywhere up to 12 hours in order to add a new beneficiary and send payments. This created a need for quick debit, for use with merchants and peer-to-peer transactions. Mobile Wallets Disrupted By NPCI In the mid 2010s, the rise of mobile wallets such as Paytm and PhonePe helped the tech savvy Indian citizens make transactions cashless. This resulted in a further fragmentation of the market, with certain merchants adopting certain wallets over others, thus requiring the end user to maintain multiple apps to transfer money. However, the NPCI seemed to have another trick up its sleeve; one that has now proven to be one of the most seamless ways to transfer value in the Indian subcontinent. Launched in mid-2016, the Unified Payments Interface has now proven to be the de facto standard for peer-to-peer and merchant transactions. The way UPI functions allow it to be a scalable platform built as a layer on top of the existing IMPS platform. Nandan Nilekani, the former Chairman of UIDAI, has previously stated that the UPI platform was aimed at allowing for an easy debit capability. As to be expected from this statement, the UPI platform works 24x7x365. A Mobile First Approach For A Cashless Economy UPI can merge multiple bank accounts into a single mobile application, thus allowing for routing of funds, merchant payments and account management easy. Using this application, which is provided by the participating bank, customers will be assigned a Virtual Payment Address [VPA], protecting the users’ credentials such as account details or card details. Transactions can either be transfers or collect requests, with money moving upon users’ confirmation through 2 Factor Authentication. Through the push for adoption of an application known as BHIM, the NPCI has also stressed upon the ease of use of the UPI platform by offering it in multiple regional languages. In August 2018, NPCI announced the launch of UPI 2.0, which allows for customers to connect to their bank overdraft accounts and for the creation of one-time payment mandates and pre-authorisation of transactions. The adoption rate for UPI shows that it is indeed being used widely across the board. For banks, the pilot launch in April of 2016 saw 21 adopters. Now, more than 120 banks are live on the network , with transactions also seeing a similar hike in volume. There were over 3 billion transactions that took place on the UPI platform in 2018, with 620 million occuring in December alone. This marks an 18% growth over the 525 million transactions observed in November. The total value of the transactions also touched the mark of 1 lakh crore INR in December, demonstrating that the platform was getting its fair share of adoption. Reaching For Market Consolidation As A World-Class System The so-called ‘second wave’ of UPI adoption also seems to be occurring, with in-app payment integrations taking place across the board. This is a feature of the UPI platform, where apps can function as service and value providers to the platform. Prominent integrations come in the form of Google Pay, formerly Tez, PhonePe, a payments company operated by Flipkart, WhatsApp, the messaging platform owned by Facebook, Amazon Pay, and PayTM, a mobile wallet solution. The system has also been praised highly by the FIS as the “most evolved and sophisticated public digital payments infrastructure in the world”. The IMPS system, upon which UPI is built, is said to be one of the best instant payment systems in the world.They also stated, “UPI-based services have led to the creation of a wealth of innovative payment solutions. In addition, adoption rates of UPI payments and transactions originating through UPI-based apps have been so robust that they have adversely impacted traditional debit and credit card numbers.” UPI has now established itself as a high volume, low cost and highly scalable architecture built on an open source platform, thus marking the advancement of India over payment services seen in other countries. Moreover, the adoption rate and ease of use speak for themselves, as the platform is now used by many of India’s digital youth.","excerpt":"Credit and debit cards have seen increased usage online, indicating the general shift of society towards being cashless. The Indian banking system has also not been a laggard in deploying systems for the easy transfer of money online This is seen by the National Payments Corporation of India’s impressive rollout of the Instant Money Payment […]","categories":["AI Features"],"tags":["banks","FinTech","npci","payments","UPI"],"author_name":"Anirudh VK","publish_date":"2019-05-04T13:42:58","publication_year":"2019","word_count":768,"keywords":["Go","payments","npci","banks","AI","programming_languages:R","ML","Scala","Git","UPI","RAG","Aim","programming_languages:Scala","FinTech","R"],"extracted_tech_keywords":["AI","ML","Aim","RAG","R","Go","Scala","Git","programming_languages:R","programming_languages:Scala"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/with-3bn-transactions-upi-scripted-a-mobile-first-approach-for-india\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10023362,"title":"5 Drawbacks Of Google Colab with Explanation","content":"Google Colab came out as a boon for machine learning practitioners — not only to solve the storage problems of working with a large dataset but also financial constraints of affording a system that meets data science work requirements. The Jupyter notebook environment running on the cloud with no requirement for a separate setup was designed to equip ML enthusiasts to learn, run, and share their coding with just a click. Its free access to python libraries, 50 GB hard drive space, 12 GB RAM, and a free GPU makes it a perfect bet for ML practitioners. Despite all these advantages, in reality, Google Colab comes with several disadvantages and limitations, restricting a machine learning practitioners’ coding capability to run without any speed bumps. Let’s look at these features of Google Colab that can spoil machine learning experiences. Also Read: The Beginner’s Guide To Using Google Colab Drawbacks Of Google Colab Closed-Environment: Anyone can use Google Colab to write and run arbitrary Python code in the browser. However, it is still a relatively closed environment, as machine learning practitioners can only run the python package already pre-added on the Colab. There is no way that one can add their own python package and start running the code. Hence, the platform can provide common tools but is not suitable for specialisation. Repetitive Tasks: Imagine one has to repeat the same set of actions repeatedly to execute a task — not only will it be exhausting, but it will also consume a lot of time. Similarly, for every new session in the Google Colab, a programmer must install all of the specific libraries that aren’t included with the standard Python package. No Live-Editing: Writing a code and sharing the same with your partner or a team allows you to collaborate. However, the option for live editing is completely missing in Google Colab, which restricts two people to write, or edit codes at the same time. Hence, it further leads to a lot of back and forth re-sharing. Additionally, this feature is provided by its other competitors, including CoCalc. Also Read: The Google Colab Hacks One Should Be Aware Of Saving & Storage Problems: Uploaded files are removed when the session is restarted because Google Colab does not provide a persistent storage facility. So, if the device is turned off, the data can get lost, which can be a nightmare for many. Moreover, as one uses the current session in Google Storage, a downloaded file that is required to be used later needs to be saved before the session’s expiration. In addition to that, one must always be logged in to their Google account, considering all Colaboratory notebooks are stored in Google Drive. Limited Space & Time: The Google Colab platform stores files in Google Drive with a free space of 15GB; however, working on bigger datasets requires more space, making it difficult to execute. This, in turn, can hold most of the complex functions to execute. Google Colab allows users to run their notebooks for at most 12 hours a day, but in order to work for a longer period of time, users need to access the paid version, i.e. Colab Pro, which allows programmers to stay connected for 24 hours. Finally, the less talked about drawback of the platform is its inability to execute codes or run properly on a mobile device. Wrapping Up Google Colab entered the market with a pure focus to provide machine learning practitioners with a platform and tools to advance their machine learning capabilities. However, over time, the volume, intensity, and quality of data changed, and so did ML practitioners’ requirements to find solutions to complex problems. Coming out with a paid version is easy, but for the larger good, it needs to be upgraded and freely accessible to anyone for the entire machine learning ecosystem to grow.","excerpt":"Drawbacks of the Google Colab platform can create unnecessary hindrance for the machine learning community. A revisit can work.","categories":["AI Trends"],"tags":["Google Colab"],"author_name":"kumar Gandharv","publish_date":"2021-04-05T16:00:00","publication_year":"2021","word_count":642,"keywords":["data science","Go","machine learning","AI","ML","RAG","Colab","Python","Google Colab","Jupyter","R"],"extracted_tech_keywords":["AI","machine learning","ML","data science","Jupyter","Colab","RAG","Python","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/explained-5-drawback-of-google-colab\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":5320,"title":"MTN Group Selects Flytxt Platform For Real-time Decisioning","content":"February, 25th, 2014, Barcelona: Flytxt, the Big Data solution provider for Communication Service Providers (CSPs), today announced its group-level agreement with MTN, the largest telecom group in Africa. The CSP will deploy Flytxt’s real-time, on trigger decisioning and customer engagement solution across its group operations. MTN, having recently won the prestigious AfricaCom award for Most Innovative Service, has always been in the forefront of bringing the latest technology and services for its subscribers. “We constantly strive to improve the lives of our customers and communities by providing them with a distinct customer experience. Flytxt’s technology will enable us to further our goal of giving our customers a unique experience and help us engage in an informed interaction with millions of subscribers in real-time,” says Pieter Verkade, Group Chief Commercial Officer, MTN Group. Flytxt’s on-trigger decisioning and engagement solution will enable MTN to respond to subscriber’s actions and their needs instantaneously. The CSP can detect meaningful customer initiated events across the network and trigger contextually relevant actions in real-time, thereby, enhancing customer experience as well as CSP revenues. Dr. Vinod Vasudevan, CEO, Flytxt, says, “This engagement with MTN group is a significant milestone in the application of real-time stream analysis technologies at a CSP. We look forward to supporting MTN’s innovative initiatives with our real-time decisioning platform to improve customer engagement and revenue enhancement.” Flytxt has signed up with over 50 CSPs, serving more than 500 million subscribers across APAC and EMEA, for its Big Data Analytics powered customer experience, revenue management and insight monetization solutions. About Flytxt Flytxt is a leading provider of Big Data Analytics solutions that enable CSPs to derive measurable economic value from subscriber data. The company offers Customer Experience and Revenue Management Solutions as well as consultancy services to help CSP run campaigns for increasing revenue, reducing churn and enhancing loyalty. Flytxt’s closed loop integrated real-time marketing platform has been selected by leading CSPs across APAC and EMEA, serving more than 500 million subscribers and has generated over $350 million incremental revenue for  them till now. The company has won many industry awards like NASSCOM Emerge 50 league of 10, Aegis Graham Bell Award for innovation in Mobile Advertising, BID International Quality Summit Award, Red Herring Asia 100 and IEEE Cloud Computing Challenge. With its headquarters in the Netherlands, corporate office in Dubai and global delivery centres at Trivandrum and Mumbai in India, the company has its presence in New Delhi, Lagos, Nairobi, Kuala Lumpur and London. For more information about Flytxt, visit www.flytxt.com. Follow us | Connect on LinkedIn | Like us on Facebook About the MTN Group Launched in 1994, the MTN Group is a leading emerging market operator, connecting subscribers in 22 countries in Africa, Asia and the Middle East. The MTN Group is listed on the JSE Securities Exchange in South Africa under the share code: “MTN.” As of 30 September 2013, MTN recorded 203.8 million subscribers across its operations in Afghanistan, Benin, Botswana, Cameroon, Cote d’Ivoire, Cyprus, Ghana, Guinea Bissau, Guinea Republic, Iran, Liberia, Nigeria, Republic of Congo (Congo Brazzaville), Rwanda, South Africa, Sudan, South Sudan, Swaziland, Syria, Uganda, Yemen and Zambia. Visit. Visit us at www.mtnbusiness.com , www.mtn.com www.mtnmmo.com  and for our football fans at www.mtnfootball.com For more information, please contact: Flytxt: Gargi Basu, Manager – Marketing, gargi.basu@flytxt.com MTN Group: MTN Group Office, mtngrouppressoffice@mtn.co.za","excerpt":"February, 25th, 2014, Barcelona: Flytxt, the Big Data solution provider for Communication Service Providers (CSPs), today announced its group-level agreement with MTN, the largest telecom group in Africa. The CSP will deploy Flytxt’s real-time, on trigger decisioning and customer engagement solution across its group operations. MTN, having recently won the prestigious AfricaCom award for Most […]","categories":["AI News"],"tags":["real time decisioning"],"author_name":"AIM Media House","publish_date":"2014-02-25T15:26:18","publication_year":"2014","word_count":554,"keywords":["big data","Go","programming_languages:R","cloud computing","AI","innovation","programming_languages:Go","analytics","GAN","real time decisioning","R"],"extracted_tech_keywords":["AI","analytics","cloud computing","R","Go","big data","GAN","innovation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/mtn-group-selects-flytxt-platform-for-real-time-decisioning\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10103792,"title":"AMD Eyes Big Wins with MI300X for AI Workloads","content":"AMD predicts that MI300X will make the company earn $2 billion revenue in 2024 because of the strong pull from the customers. The processor is set for a release at the AMD Advancing AI event, on December 6. “We are seeing a tremendous pull for MI300X,” AMD CTO Mark Papermaster told AIM, talking about how companies such as Lamini, Moreh, and Databricks are already announcing their success stories after using MI250x for training their generative AI models. “In fact, Lisa Su (CEO) has said that MI300X is going to be the fastest-growing product in the history of AMD,” added Papermaster. Another reason for the success of AMD’s hardware is the success of its software stack, ROCm (AMD Radeon Open Compute), which Papermaster highlighted has reached high performance compute production level. Moreover, he added that the next version of ROCm, the 6.0, would be released soon, and would be production-level ready for AI workloads. ‘We are bringing a leadership product‘ At Ignite 2023, Microsoft announced that it would be using MI300X for their AI workloads. Interestingly, at the same conference Microsoft also said that it would be using the NVIDIA GH200 superchips for the same purpose. “Competition is good. It brings innovation,” said Papermaster. “It brings pricing that ensures value for the customers and spurs the industry forward.” “We have not only brought competition, but are also bringing in a leadership product in inference applications,” he further added about MI300X. The upcoming GPU will be recognised for its leadership in both training and inference applications, solidifying AMD’s position in the AI hardware landscape. “MI300X is a high performance GPU, capable of being scaled out to very large cluster sizes,” he further spoke about how AI is evolving and the company is also focusing on edge computing on smaller sizes within laptops. AMD highlighted the widespread adoption of Ryzen AI, the inaugural dedicated AI accelerator available on an x86 processor. With over 50 systems now equipped with Ryzen 7000 Series processors and Ryzen AI, millions of AMD AI PCs are available in the market. “Historically, if you go back in time, most AI inference seems done to the CPU. And still, today, at least we fully support inferencing on AMD CPUs. But particularly, generative AI applications are in much higher demand for computing capability. And so when you look at the generative AI, training and inferencing it requires acceleration, and that is where we’re bringing in competition with AMD instinct roadmap GPU,” he said. “What I want to highlight is that AMD has a broader strategy for AI than just GPUs. This is what makes us different from NVIDIA,” Gilles Garcia, senior director business lead, data center communication group at AMD, told AIM. AMD believes that most of the AI workloads can be handled by CPUs alone. “CPUs are best for handling most of the problems with current edge processing such as thermal management, cost effectiveness, and reducing the footprint by working on edge,” said Garcia. The startup and open source community in India Focusing on the recent acquisition of Nod.ai, a software stack company that is now helping AMD, Papermaster said that AMD is always looking at the startup community in India. “We have a strong design presence in India. The country will, of course, be central for our AI product development, hardware and software product development efforts.” Highlighting how Hugging Face also started using AMD GPUs for testing, he added that with the explosion of the AI ecosystem, AMD is very much committed to an open ecosystem. “We are not proprietary or close,” he added. “Rather than needing just one specific partnership to get access to AMD, we’ve differentiated because we’re strongly committed to open source software and to open collaborations. I expect many such collaborations between AMD and India based on the strong adoption in India, and support of open source.” To expand its research and engineering operations in India, AMD inaugurated its largest global design centre in Bengaluru. The AMD Technostar R&D campus is a key component of the company’s $400 million investment in India over the next five years. The campus plans to accommodate around 3,000 new employees at AMD which will focus on R&D. “We are very focused on workplace development here in India,” Papermaster continued. “Jaya Jagadish, our country head and senior vice president, was the leader of a government panel, which studied workforce development. She has made very specific recommendations to the government of India.” Papermaster says that AMD has very innovative workforce development programmes, which the company is continually employing in India. “We have strong relationships with universities and we are also providing additional training to students. Then we bring them onto a very established internship programme,” he explained about how AMD has established an excellent pipeline for college graduate engineering in India.","excerpt":"AMD sees India as the growth market and is hiring 3,000 employees for R&D.","categories":["AI Features"],"tags":["Interviews and Discussions"],"author_name":"Mohit Pandey","publish_date":"2023-11-28T17:09:17","publication_year":"2023","word_count":800,"keywords":["Go","Hugging Face","startup","AI","innovation","Aim","Databricks","generative AI","edge computing","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","generative AI","Aim","Hugging Face","edge computing","Databricks","R","Go","innovation","startup"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/amd-eyes-big-wins-with-mi300x-for-ai-workloads\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10024261,"title":"Guide To NICE: An Algorithm To Find Nearest Instance Counterfactual Explanations","content":"NICE is an algorithm that can create Nearest Instance Counterfactual Explanations for heterogeneous tabular data (containing both numerical and categorical variables). It has been recently introduced by Dieter Brughmans and David Martens in April 2021 (research paper). Before going into the algorithm’s details, let us look at the meaning of ‘counterfactual explanations’. What are counterfactual explanations? The term ‘counterfactual explanations’ refers to a scenario such as “If an event A had not occurred, then event B would not have occurred.” For instance, “If clouds had not existed, there would never be rainfall.”  This statement requires us to think about a hypothetical situation of an atmosphere without clouds which is not possible in reality; hence the name “counterfactual”. In simple terms, counterfactuals tell us to perform an action in order to get a required result (such as creating an absence of clouds to inhibit rainfall). If we view this concept in terms of machine learning, ‘action’ would mean modifications in the features used for prediction, while ‘required result’ would mean the expected response of the model. How NICE algorithm works? Consider an example of credit scoring to understand the working of the NICE algorithm. Image source: Research paper The above plot shows the income and age criteria used to decide whether an individual can get a loan. The person with an income $32,000 and age 39 years is unable to get a loan and hence lies on the left side of the boundary in the plot. While the other three people are eligible for a loan, their records lie on the other side of the boundary. The NICE algorithm tries to find out the minimum possible changes that can be done to the features of an ineligible candidate so that he becomes eligible to get a loan. It can be seen from the tabular data that keeping his age the same, if the person’s income is increased by $8000 (i.e. changed from $32,000 to $40,000), he will become eligible for a loan. This change of $8,000 in the income feature is said to be a ‘counterfactual explanation’ that the NICE algorithm finds out. The NICE algorithm uses one of the following three properties of a counterfactual explanation for optimum results: Sparsity – It is the count of features required to be modified to achieve the desired outcome.Proximity – It refers to the difference between the actual input and the counterfactual instance.Plausibility – It is the measure of closeness of a counterfactual instance to the whole data. Say, for the tabular data used above, if an explanation states that keeping the income intact, the age of ineligible person can be varied from 39 to 140, then such an explanation is implausible as it is too far from the data manifold. Practical implementation Here’s a demonstration of implementing the NICE algorithm on the Adult dataset. The following code is tested using Google colab with Python 3.7.10, pmlb 1.0.1.post3 and NICEx 0.1.0 versions. Step-wise implementation of the code is as follows: Install NICEx library from PyPI. !pip install NICEx Install pmlb – a Python wrapper for the Penn Machine Learning Benchmark (PMLB) data repository from PyPI. PMLB is a benchmark suite for comparing the performance of various ML algorithms on a variety of datasets. Refer to the following papers for a detailed understanding of counterfactual explanations for ML: Paper1Paper2 !pip install pmlb Import required libraries and modules. import pandas as pd #To fetch a dataset from the PMLB from pmlb import fetch_data from sklearn.model_selection import train_test_split from sklearn.ensemble import RandomForestClassifier from sklearn.compose import ColumnTransformer from sklearn.preprocessing import StandardScaler,OneHotEncoder from sklearn.pipeline import Pipeline from nice.explainers import NICE Fetch the adult dataset adult = fetch_data('adult') Display the dataset adult Output: Select feature columns and target column. #Drop the unnecessary columns for feature set X = adult.drop(columns=['education-num','fnlwgt','target','native-country']) #Assign ‘target’ column with all rows as labels y = adult.loc[:,'target'] #List of columns remaining in X will form the feature set feature_names = list(X.columns) Extract values of features and label columns X = X.values X #Display updated X Output: y=y.values y   #Display updated y Output:   array([1, 1, 1, …, 1, 1, 0]) Check the shape of X and y. X.shape Output: (48842, 11) y.shape Output: (48842,) Split the data into train set and test set with train:test ratio of 70:30. X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.3) Display the feature column names. print(feature_names) Output: ['age', 'workclass', 'education', 'marital-status', 'occupation', 'relationship', 'race', 'sex', 'capital-gain', 'capital-loss', 'hours-per-week'] Create lists of column numbers (starting from 0) fro numerical and categorical columns. categorical_features = [1,2,3,4,5,6,7] numerical_features = [0,8,9,10] Using the Pipeline() method of sklearn, we can sequentially apply several transforms. After creating the pipeline, the next step must be to fit the pipeline to the training data. #Create pipeline clf = Pipeline([ #list of (name,transform) tuples ('PP',ColumnTransformer([ ('num',StandardScaler(),numerical_features), ('cat',OneHotEncoder(handle_unknown = 'ignore'),categorical_features)])), #estimator to be used specified as (name, estimator function) tuple ('RF',RandomForestClassifier())]) ColumnTransformer() here applies standardization to the numerical features and creates on-hot representation of the categorical features. The transforms will be applied in the order in which they are specified so the estimator is kept at the end of the pipeline. Fit the transforms to the training data. clf.fit(X_train,y_train) Output: Create a lambda function to predict the class based on the value of each attribute. Its output will be an array indicating probabilities based on each attribute. prediction = lambda x: clf.predict_proba(x) Initialize the counterfactual explanations NICE_model = NICE(optimization='sparsity',    #optimization method justified_cf=True) Setting the ‘justified_cf’ parameter to True will limit the search of the nearest neighbors to the rightly classified training samples. Fit the model to training data. NICE_model.fit(X_train = X_train,   #training data predict_fn=prediction,    #prediction function y_train = y_train,      #train set labels #categorical features cat_feat=categorical_features, #numerical features num_feat=numerical_features) Use the explain() method to create a counterfactual instance of a test sample. Create and then display the counterfactual explanation of the 2nd sample from the test set. CF = NICE_model.explain(X_test[1:2,:]) CF Output: array([[45.,  4.,  9.,  2.,  7.,  0.,  4.,  1.,  0.,  0., 40.]]) Output’s interpretation: The values of 2nd sample of X_test are: array([[41.,  4., 15.,  2., 12.,  0.,  4.,  1.,  0.,  0., 35.]]) The target label for this sample is 1. The NICE algorithm examines the values of the instance and creates another set of similar values for each attribute, resulting in a different output class, i.e., 0. As explained in the initial part of the article which explains the meaning of ‘counterfactuals’ in ML; here, the feature values got changed by the algorithm in such a way that we get the result of class 0 as the target label. The output thus justifies the meaning of “counterfactual” – it shows if the input sample had not existed, which nearest attribute values would have given the required label outcome. Code source: GitHubGoogle colab notebook of the above implementation References NICE Research paperGitHub repositoryPMLB research papers (paper 1) (paper 2)","excerpt":"NICE is an algorithm that can create Nearest Instance Counterfactual Explanations for heterogeneous tabular data (containing both numerical and categorical variables).","categories":["Deep Tech"],"tags":["Guide"],"author_name":"Nikita Shiledarbaxi","publish_date":"2021-04-20T10:00:00","publication_year":"2021","word_count":1138,"keywords":["Go","machine learning","TPU","AI","ML","Colab","Ray","Python","R","Guide","Pandas"],"extracted_tech_keywords":["AI","machine learning","ML","Ray","Colab","Pandas","TPU","Python","R","Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/guide-to-nice-an-algorithm-to-find-nearest-instance-counterfactual-explanations\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":7752,"title":"Analytics India Companies Study &#8211; 2015","content":"Analytics India Companies Study is aimed to provide a deeper insights on the analytics companies’ landscape in India. We, at Analytics India Magazine, focus on providing the most comprehensive and in-depth analysis of various aspect of analytics ecosystem in India; be it the salaries\/ jobs trends, analytics professionals, data scientists, leaders, education etc. Here we present to you a bird eye view of the analytics companies in India. Read Analytics India Companies Study 2013 Read Analytics India Companies Study 2012 The number of analytics organization in India have grown by 4 folds in last 2 years. The reasons are manifold- Numbers can be deceptive. Most of these companies were already existing and few either incorporated analytics as a line of business or adopted analytics as their primary business Burgeoning demand for analytics seeing an increase in pure play analytics firms Growing players in the value chain: training companies, recruitment and consulting firms But this increase appears disappointing given the growth of analytics firms worldwide. Just 4.5% of analytics firms worldwide are in India or have operations in India. This is in line with our earlier similar study that analytics outsourcing to India is subdued, or even decreasing given the state of analytics worldwide. A analytics firm in India employs on an average 115 employees. This number is an average of 57 for companies worldwide. Around 72% of analytics firms in India have less than 50 employee strength. This is consistent over the past few years. 29% of all analytics firms in India are based out of Bangalore. This is consistent from last year. Mumbai is distant second at 17% of analytics firms. 27% of analytics firms are based out of NCR. Gurgaon has seen a consistent growth in analytics firms over last few years. Delhi has the highest average employee strength for analytics firms at 177 per organization.","excerpt":"Analytics India Companies Study is aimed to provide a deeper insights on the analytics companies’ landscape in India. We, at Analytics India Magazine, focus on providing the most comprehensive and in-depth analysis of various aspect of analytics ecosystem in India; be it the salaries\/ jobs trends, analytics professionals, data scientists, leaders, education etc. Here we […]","categories":["AI Features"],"tags":["analytics companies in india","analytics outsourcing","big data companies"],"author_name":"Дарья","publish_date":"2015-08-05T18:17:51","publication_year":"2015","word_count":308,"keywords":["programming_languages:R","AI","RAG","big data companies","Aim","analytics companies in india","analytics","GAN","analytics outsourcing","R"],"extracted_tech_keywords":["AI","analytics","Aim","RAG","R","GAN","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/analytics-india-companies-study-2015\/","complexity_score":3,"technical_depth":7,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10021441,"title":"PayPal To Hire 1000 Engineers For Its India Operations","content":"PayPal, the US-based digital payments company, has announced that the company will be hiring more than 1,000 engineers for its India development centres across Bengaluru, Chennai and Hyderabad throughout 2021. According to the news media, the hiring will be across software, product development, data science, risk analytics, and business analytics streams at entry, mid-level and senior roles. Further, PayPal has also announced its plan for campus hiring from some of the top Indian engineering colleges. The announcement came after PayPal has wind up its local payments business in India while supporting local businesses for international sales. Regarding the announcement, Guru Bhat, the Vice President, Omni Channel & Customer Success, and General Manager of PayPal India stated that Indian technology centres had played a pivotal role in enabling the company to constantly innovate to stay ahead of the curve. He said, considering digital payments are becoming an essential service, PayPal is focused on nurturing world-class technology talent to continue offering products and services that meet the needs of the growing customer and merchant’s base. The technology centres of PayPal’s India currently employ over 4,500 people, playing a critical role in working on technologies to enable a safe and seamless payments experience. Adding on to this, PayPal confirmed in an official statement that the pandemic has massively accelerated the shift towards digital payments and highlighted the benefits of adopting a digital-first approach. This has made the products and services by PayPal more relevant now. Hence it has now become critical to focus on technology innovation in artificial intelligence, machine learning, data science, risk and security, customer experience and other vital areas.","excerpt":"PayPal, the US-based digital payments company, has announced that the company will be hiring more than 1,000 engineers for its India development centres across Bengaluru, Chennai and Hyderabad throughout 2021.  According to the news media, the hiring will be across software, product development, data science, risk analytics, and business analytics streams at entry, mid-level and […]","categories":["AI News"],"tags":["hiring data scientists","PayPal"],"author_name":"Sejuti Das","publish_date":"2021-03-04T11:07:33","publication_year":"2021","word_count":269,"keywords":["data science","artificial intelligence","machine learning","hiring data scientists","AI","PayPal","innovation","ML","Git","ViT","analytics","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","data science","analytics","R","Git","ViT","innovation"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/paypal-to-hire-1000-engineers-for-its-india-operations\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10160797,"title":"Accels’ $650 Million Early-Stage Fund to Boost AI and Tech for Bharat","content":"Venture capital firm Accel announced its eighth early-stage fund on Monday, committing $650 million to boost the next generation of innovative startups in India and Southeast Asia. The fund will significantly focus on AI startups with a focus on India. AI for India On the AI front, Accel plans to invest in three main categories. These include enterprise AI, where platforms enable AI use cases with agentic technologies, large language models, and small language models; services as software, where AI startups leverage India’s IT services capabilities for automation offerings and vertical AI, where startups integrate AI into industry-specific applications. “This growth opens new doors for entrepreneurs to build solutions with global relevance while addressing local challenges, and we can’t wait to meet them,” Accel said in a recent blog announcement. Accel’s Fund VIII also aims to tap into other areas such as consumer brands, fintech and manufacturing. In the consumer segment, the emphasis is on startups catering to Bharat, which is the top 30% of households in tier-2 regions, addressing the rising demand for premium services and creating aspirational brands that capitalise on Gen Z’s increasing discretionary spending. When it comes to fintech, the focus areas include digital wealth management platforms for affluent consumers, fintech infrastructure bridging banks and fintechs for seamless digital experiences, and startups leveraging India’s digital infrastructure to accelerate financial product distribution. In manufacturing, the fund supports startups that meet global demand for diversified supply chains, India-focused value-added manufacturing, and Industry 5.0 solutions that bring next-generation digital technologies to factory floors for enhanced efficiency, sustainability, and output quality. Accel Continues to Grow Indian Startups To support founders beyond financial investment, Accel has launched initiatives such as SeedToScale, an open-source platform for company-building insights, and Accel Atoms, an early-stage scaling program that has supported 36 startups to date and collectively raised over $200 million. Accel sees a significant opportunity for venture-backed companies to shape the economic landscape, where India’s GDP is expected to touch approximately $8 trillion over the next decade. Currently, VC companies represent less than 5% of India’s market capitalisation. Accel’s commitment to the Indian startup ecosystem is strong with early investments in successful companies such as Flipkart, Freshworks, and Swiggy. The firm has been the first institutional investor in 80% of its portfolio companies. Prayank Swaroop, partner at Accel, said, “We believe AI will be a game-changer, and we’re excited to support startups at the forefront of this technology.”","excerpt":"On the AI front, Accel will back founders building in enterprise AI, services as software, and vertical AI platforms.","categories":["AI News"],"tags":["Accel","AI fund","AI Startups","India","Prayank Swaroop"],"author_name":"Vandana Nair","publish_date":"2025-01-06T15:51:18","publication_year":"2025","word_count":404,"keywords":["Go","TPU","AI","Accel","ML","Git","RAG","Ray","AI fund","Aim","Prayank Swaroop","small language models","R","India","AI Startups"],"extracted_tech_keywords":["AI","ML","Aim","Ray","small language models","RAG","TPU","R","Go","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/accels-650-million-early-stage-fund-to-boost-ai-and-tech-for-bharat\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":57447,"title":"Satya Nadella Explains How Microsoft Is Fueling Tech Intensity In India At Future Decoded: Tech Summit 2020","content":"Tech giant Microsoft has been doing several innovations when it comes to applying technology in almost every sector including healthcare, e-commerce and other such. Today, at the Future Decoded: Tech Summit by Microsoft, Microsoft CEO, Satya Nadella addressed some of the use cases and showcased a few innovations across these sectors. Future Decoded: Tech Summit is a conclave for the Indian developers, technologists and industry leaders. The summit is mainly concentrated on how innovative use of technology is driving stories of success across business and communities across India. Nadella gave a 45 minutes key-note speech where he said that the tech giant’s mission is to help organisations jump-start their growth through transformational tech intensity, which results from enterprises adopting best-in-class technology, building through their own digital capability as well as having trust in both the technology they use and the organisations they partner with. Here are a few points which Nadella put forwarded in order to create tech intensity among the organisations. How Indian organisations can lead to an era of digital transformation?Is it possible to take the abundance of data and technology and create more for economic prosperity?How innovation improves outcomes for everyone?How to protect unconscious human biases?How to build a Power Platform? Nadella explains the importance of intelligent cloud and intelligent edge and how it is impacting the tech space in organisations. He provides an equation for tech intensity which is Tech Intensity= (Tech Adoption* Tech Capability) ^(Trust*Inclusivity*Sustainability) According to Nadella, there will be 50B connected devices by 2030 and the total amount of data by 2025 is 175ZB. To tackle these, the tech giant has been developing the tech stack with various digital transformation, Microsoft has been working on HoloLens, how Azure cloud platform is striving to be a limitless data estate with variety, scale, cloud and edge. He further discussed the new and largest machine learning model ever, known as Turing Natural Language Generation (TNLG). Talking about building a Power platform, Nadella said that it includes elements like Power Apps, Power BI, Power Automation tools and Power Virtual Agents. Watch the talk here:","excerpt":"Tech giant Microsoft has been doing several innovations when it comes to applying technology in almost every sector including healthcare, e-commerce and other such. Today, at the Future Decoded: Tech Summit by Microsoft, Microsoft CEO, Satya Nadella addressed some of the use cases and showcased a few innovations across these sectors. Future Decoded: Tech Summit […]","categories":["AI News"],"tags":["best technology to learn for future"],"author_name":"Ambika Choudhury","publish_date":"2020-02-25T15:30:00","publication_year":"2020","word_count":347,"keywords":["machine learning","AI","R","digital transformation","Git","automation","ViT","Rust","GAN","Azure","best technology to learn for future"],"extracted_tech_keywords":["AI","machine learning","Azure","R","Rust","Git","GAN","ViT","digital transformation","automation"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/satya-nadella-explains-how-microsoft-is-fueling-tech-intensity-in-india-at-future-decoded-tech-summit-2020\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10118533,"title":"Why Data Pipeline is Important for High ROI AI Products","content":"Salesforce is said to be in the final stages of negotiations to acquire data-management software provider Informatica for $11 billion. To many, the deal is reminiscent of Snowflake’s acquisition of Neeva in May last year. “Salesforce potentially acquiring Informatica seems like a push to compete more with Snowflake,” wrote Astasia Myers, partner at Felicis. These acquisitions are in step with big-tech companies like Google acquiring Looker, the data analytics startup, and Microsoft acquiring ADRM Software and Rubrik in 2019 and 2021, respectively. Rubrik, a data management company, is now targeting an IPO with a $5.4 billion valuation. Databricks’ acquisition of Acrion, Mosaic ML, and Okera is also along similar lines, aimed at managing its data pipeline and increasing generative AI capabilities. Similarly, Salesforce’s possible acquisition of Informatica is targeted at greatly enhancing its data capabilities, especially in fields like data integration, quality assurance, and customer insights. This also points towards the importance of building a data strategy and ensuring a smooth pipeline when it comes to building high-ROI AI products. Contextualised Data is King James Wu, partner at M12, highlighted in a recent post that building a strong data pipeline and building data-centric AI is important. That is why the venture fund also invested in Unstructured.io, with another data curation company in the pipeline. “Big data will continue to be the foundation, but contextualised data is king,” he said. “We’re interested in the ‘AI-data feedback loop’ – we think better AI can analyse data to identify errors and inconsistencies, improving data quality for future models,” he explained, saying that cleaner data can also help in training superior AI models, like a cyclic loop. Naveen Rao, VP of generative AI at Databricks, also shared similar thoughts. “We at Databricks are very much about the lifecycle of data and GenAI working synergistically together. We demonstrated the power of our training platform by building DBRX with it and we used all the tools in Databricks. We believe in the power of all the components around the model that comprise the full system,” he said. This points to the need for building a good data strategy for expecting high ROI on AI products. Matthew Blasa, AI strategist and lead data scientist consultant, emphasises that since AI’s lifeblood is data, it is important to have an endless clean data pipeline for AI products. Source: Matthew Blasa “It’s important to ensure that your data is reliable, relevant to the large needs, and collected from multiple sources,” Blasa explained. “Without a clear data strategy, creating a model with enduring value is challenging. Relying solely on retraining and monitoring won’t close the gap. It may even make it harder.” Crawling, walking, and running with AI AI advisor Vin Vashishta shares the perfect plan for companies building AI products. “One thing that I’ve learned after a decade of building data and AI products is that businesses must crawl-walk-run with AI,” he wrote in a post. Crawling is about collecting data, walking involves using the data to create descriptive models, and running uses more advanced models such as predictive, prescriptive, and diagnostic ones. He explains how starting with crawling and walking makes running less expensive and faster in the long run. Each phase offers immediate benefits and builds on the previous phase, creating a solid foundation. “Walk and run handle about 90% of use cases, reducing time to value,” Vashishta explained. In another post, Vashishta explained how high-quality data can bring quick results, and descriptive models trained on it yield quarterly gains. These efforts lay the foundation for AI products and potentially larger returns. “Trash data trains trash models, but the business needs tangible returns in months, not years. Fixing the data doesn’t deliver them unless data teams and leaders take a product-first approach,” he added. The Data Pipeline Strategy It is clear that data availability is important to build the best generative AI products. This is why companies like Salesforce, Snowflake, Databricks, and all other data and AI providers are expanding their hold on data companies. This would, in the end, provide them with high-quality streamlined data to improve their AI products. AI products are data products. “Without a solid data strategy, it’s tough to trust the decisions made by our AI-driven products and keep them profitable,” said Blasa.","excerpt":"Salesforce, Snowflake, Databricks, and all other AI companies are expanding their hold on data management companies.","categories":["AI Features"],"tags":[],"author_name":"Mohit Pandey","publish_date":"2024-04-18T15:40:09","publication_year":"2024","word_count":710,"keywords":["Go","GenAI","AI","ML","Aim","Databricks","generative AI","analytics","R","Snowflake"],"extracted_tech_keywords":["AI","ML","analytics","generative AI","GenAI","Aim","Snowflake","Databricks","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/why-data-pipeline-is-important-for-high-roi-ai-products\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":53421,"title":"How Facial Recognition Is Becoming The Indian Government’s New Best Friend","content":"From hovering drones to hawk-eyed CCTV cameras, Indians by 2020 have become used to being surveilled. Facial recognition, one of the booming, yet controversial technologies, is being harnessed by the Government on a large scale. According to reports, the facial recognition market was valued at $4.51 billion in 2018 and is expected to reach a value of $9.06 billion by 2024, at a CAGR of 12.5%, for the forecast period of 2019-2024. In this article, we list down seven times where police and the Govt of India have used facial recognition techniques to monitor the citizens. (The list is in chronological order starting from the latest) 1| Identifying Culprits of Recent JNU Attack After the violence that broke out at Jawaharlal Nehru University (JNU) in January 2020, the Delhi police have started to investigate the crime with the help of video footage and facial recognition systems to identify those who were involved in the attacks. 2| Political Rally Last month, Delhi Police used facial recognition software to screen the crowd in a political rally which was led by Prime Minister Narendra Modi. According to sources, this was the first time that the Delhi Police used a set of facial images collected from footage that was filmed at the city’s previous protests to filter the “law and order suspects” at the rally. The end product was made available on the web platforms and mobile devices — in Android, iOS, and Windows. 3| Identifying Criminals  Nov 19 Last year, the Government of India had been reported to use facial recognition and artificial intelligence technology for identifying and catching criminals. The Railway Protection Force (RPF) stated the aim of linking the facial recognition system (FRS) with existing databases of Crime and Criminal Tracking Network & Systems (CCTNS) to identify criminals. According to sources, the CCTNs would be connected to the FRS database through a bridge software from which the Govt. of India will have access to the massive database of criminals. 4| World’s Biggest Facial Recognition A few months ago, National Crime Records (NCRB), a subset of the Government of India, has also announced the development of National Automated Facial Recognition System (NAFRS) to help in identifying and verifying persons automatically from digital images, photos, digital sketches, video frames, and video sources by comparing selected facial features of the model with an already existing image database. 5| Paperless Travel Last year in July, Hyderabad’s Rajiv Gandhi International Airport launched a facial recognition facility under the Centre’s Digi Yatra Policy for the first time in India. According to the sources, the facial recognition technology has omitted the whole physical check-up routine of documents at the airport, which was earlier done by the Central Industrial Security Forces (CISF). The process was replaced by the facial recognition technology that has been introduced at the passenger entry points. 6| Republic Day Last year, 2019, on the Republic Day celebrations, around 30 facial recognition cameras were installed at the venue by the Delhi Police. The cameras there had advanced facial recognition technology inbuilt to help the centre police to identify terrorists and criminal elements during the celebration. The working of the facial recognition cameras was such that if any visitor has a 70% face match with any available photographs of terrorists and criminals in the database, the system would then send an alert to the control room. 7| Solve Murder Case In 2018, Amritsar police was one other example which used a facial recognition system known as Punjab Artificial Intelligence System or PAIS, which was developed by Gurugram AI company — Staqu Technologies. The system also included CCTV footage and facial recognition technologies which helped in solving a murder case within 24 hours.","excerpt":"From hovering drones to hawk-eyed CCTV cameras, Indians by 2020 have become used to being surveilled. Facial recognition, one of the booming, yet controversial technologies, is being harnessed by the Government on a large scale. According to reports, the facial recognition market was valued at $4.51 billion in 2018 and is expected to reach a […]","categories":["AI Features"],"tags":["best facial recognition software","Facial Recognition","facial recognition India","facial recognition surveillance","Indian government","Indian government using AI","is the tech boom over"],"author_name":"Ambika Choudhury","publish_date":"2020-01-09T11:00:23","publication_year":"2020","word_count":615,"keywords":["Go","artificial intelligence","Facial Recognition","programming_languages:R","best facial recognition software","AI","programming_languages:Go","Git","facial recognition India","Aim","Indian government","is the tech boom over","GAN","Indian government using AI","R","facial recognition surveillance"],"extracted_tech_keywords":["AI","artificial intelligence","Aim","R","Go","Git","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-facial-recognition-is-becoming-the-indian-governments-new-best-friend\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":69776,"title":"This IIoT-dedicated organization is tackling the challenges associated with Industry 4.0","content":"Subramanyam Kasibhat, Founder and CEO of Vegam Solutions Industry 4.0 and the use of IIoT-based solutions is on the rise. The fourth phase of industrial revolution is witnessing computers being interwoven into almost every industrial product. However, the technology introduces few challenges of its own, which can only be addressed by incorporating a comprehensive framework. This is where Singapore-based Vegam Solutions comes into the picture. Vegam is essentially an Industry 4.0 enabler, allowing manufacturing plants and factories to turn into smart-connected, digital factories. The organization was founded by Subramanyam Kasibhat in the year 2000 and has since moved forward, for 17 good years, providing enterprises with state of art technologies, and innovative approach to solve real life problems. Kasibhat leads the company as the CEO with his extensive know-how and experience, to bring out the most scalable and efficient solutions supporting Industry 4.0 practice. “Life has always been about solving impediments, and overcoming challenges. This is what inspired me to find Vegam solutions, to tackle the challenges associated with Industry 4.0,” remarks Kasibhat. Analytics India Magazine spoke with Kasibhat to know the unique story about Vegam Solutions. Vegam provides its analytics, dashboard visualization and operational excellence solutions (MES) for various clients across industries like manufacturing, chemical, and logistics. So far, the firm has enabled smart factories in 10 countries, and deployed its offerings in 14 countries. Company highlights: Vegam has been developing solutions since 2000 The firm has thorough experience in manufacturing operations and IT landscape Vegam promises “end-2-end” solutions capability, spanning across hardware, software, and operations. The era of Industry 4.0 Vegam is dedicated to establishing Digital Factory wherein manufacturing companies make use of these offerings to map, simulate, and optimize real world processes digitally. The end result: minimizing cost, maximized output, and advanced safety. These solutions help enterprises to obtain relevant data in real time, and take smarter decisions based on that data. Kasibhat remarks, “We noticed problems with many manufacturing plants. This could only be solved by innovation. This encouraged us to deliver ‘Digital Factory’-based solutions.” Some of the consequential benefits customers realized by incorporating Industry 4.0 practice include: Transparency and visibility into processes Decision-making is powered using data and insights Utilization of advanced digital sensors Increase in productivity Quality improvement Faster time to market Clients can integrate or make use of newer business models. Vegam combines big data & analytics, cyber-physical networks, and the power of Prescriptive Manufacturing techniques into one comprehensive framework. Analytics and Big Data is leveraged by the organization to deliver real-time actionable insights, directly to the ‘operator’, based on which he can easily take a decision or execute a task.  “Analytics is about simplification, we wouldn’t want to deal with the mathematics of it. Our objective is to enable smart manufacturing through leveraging IoT and data analytics techniques,” remarks Kasibhat. Prescriptive manufacturing gives Vegam a definite edge over its competitors, allowing the firm to harness the accuracy of science, the power of machines, and the judgment and problem-solving abilities of humans to ensure the optimum level of production performance. Delivering the most scalable solutions for Industry 4.0 Vegam delivers a plethora of products and services which not only help enterprises enable smart manufacturing, but also helps them solve the most challenging impediments in the manufacturing space. These solutions help in ensuring: Reduced costs Enhanced quality Competitiveness Safety and compliance Vegam’s dedicated team Vegam 4i: The platform is the organization’s one of the most valued offerings, as it provides a single window framework with all the departments, and functional modules of the manufacturing plant. The platform has been largely sought solution in the Asia Pacific region. It has helped manufacturers realize continuous improvements, reduction of costs, and enhanced operational efficiency. Kasibath remarks, “We initiated Vegam 4i in 2007, and have been innovating since then. We try to approach problems of manufacturing plants in a unique way. This idea is central to 4i.” The key modules include Receiving, Staging, Manufacturing, Dispatch, Live reports, and ERP System Interface. Vegam View: Another key offering, View platform provides factories with the appropriate visualization tools that enable visibility into processes. Smart factories can leverage this IoT-enabled platform to increase production foster transparency. The platform can furnish one comprehensive report of daily processes that run in a plant, involving a complex information web of “men, machine, and materials.” This helps in to bring a sense of ownership among employees. Both the platforms can be customized depending on clients’ needs. Vegam view is an easy-to-use platform, and can be implemented within a week’s time. However, 4i, being a very comprehensive platform can take 6-10 months’ time in implementing. Other solutions and services: Consulting: Vegam has had decades of experience in engaging with customers and understanding the “minute details” which help factories become smarter Vegam dStore: It allows a data scientist to build analytics app without data hassles. Vegam ZingTM: It enables advanced visualization, KPI’s and 3D Real time overlay of information onto Factory Visuals. Vegam LabelTM: It offers complete labeling needs of the manufacturing plant, helping support compliance needs. Assisting Manufacturing units to become smarter Subbu with his wife, Co-founder, and CTO Savita Kasibhat Vegam’s extensive list of clients include legacy names such as GE BEL, Siemens VDO, Texas Instruments Inc, LG, Saxa, and Henkel. Vegam Solutions has helped one of their clients increase production from 60 tonnes, to a staggering 130 tonnes. The organization was able to achieve this by leveraging the Vegam View platform. “We pulled all the data together from different sources, and help the plant gain better control over processes. This is what Vegam is about – operations and control.” Smart Manufacturing Vegam plans to move ahead in the space, and enable IoT using analytics, for factories and manufacturing plants to become smarter. Some of their upcoming projects involves them working with large steel plants and automotive manufacturing unit. Vegam will also expand its existing market in US. They will partner with another firm to bring analytics to the ‘operator.’ China is one of the other markets, the firm plans to invest in near future. “India and China have immense potential to shoot ahead in this space. If “Make in India” initiative must succeed, transparency, digitization, and smart factory will be the three pillars on which it shall stand,” concludes Kasibhat.","excerpt":"Industry 4.0 and the use of IIoT-based solutions is on the rise. The fourth phase of industrial revolution is witnessing computers being interwoven into almost every industrial product. However, the technology introduces few challenges of its own, which can only be addressed by incorporating a comprehensive framework. This is where Singapore-based Vegam Solutions comes into […]","categories":["IT Services"],"tags":["Analytics India","data scientist india","IIoT India","IoT India","Logistics"],"author_name":"Дарья","publish_date":"2017-04-19T08:43:26","publication_year":"2017","word_count":1045,"keywords":["big data","Go","TPU","ELT","IoT India","AI","Analytics India","Scala","Git","RAG","Logistics","analytics","IIoT India","data scientist india","R"],"extracted_tech_keywords":["AI","analytics","RAG","TPU","R","Go","Scala","Git","big data","ELT"],"url":"https:\/\/analyticsindiamag.com\/it-services\/this-iiot-dedicated-organization-is-tackling-the-challenges-associated-with-industry-4-0\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10059954,"title":"How Walmart uses blockchain to manage its supply chain","content":"For decades, the transportation industry has been facing the same problem: vast data disparities in the invoice and the payment process for freight carriers, leading to long payment delays and expensive reconciliation efforts. Walmart, a leader in supply chain management, is now using blockchain to create an automated process for handling invoices and payments of its 70 third-party freight carriers. Why does Walmart need blockchain? Walmart Canada uses its own trucks and third-party transport to supply over 500,000 shipments to distribution centres and stores across the country annually. More than 200 data points have to be independently calculated and accounted for in each invoice. These include the stop locations of every load, how many gallons of fuel the carriers consume, temperature updates, etc. Consequently, the data is riddled with discrepancies, and 70% of invoices require reconciliation efforts. The trouble lies at the application level of multiple information systems that can’t communicate. As a result, the reconciliation process has to be done manually, which is both labour-intensive and time-consuming. Automating the process by creating a blockchain network solves the problem of incompatible enterprise systems by introducing a shared single source of credible information for Walmart and its carriers. Pilot To create the blockchain network, Walmart Canada approached DLT Labs, a leader in engineering and deploying imaginative business solutions using distributed ledger technology. Bison Transport, one of Walmart Canada’s carriers, was also roped in. In January 2019, the pilot version went live after rigorous testing. The network, dubbed DL Freight, was rolled out to 69 other carriers in March 2021. It automatically gathers and synchronises data at every step and is only visible to the parties involved in the transaction. Since the introduction of DL Freight, less than 1% of invoices have discrepancies—reducing the costs of reconciliation efforts drastically and allowing carriers to be paid on time. Benefits The decentralised ledger records and protects transaction information shared among multiple parties. Cryptocurrencies like Bitcoin and Ethereum capitalise on this technology to allow unlimited and unspecified parties to participate in transactions without the need for an intermediary. Conversely, in supply chain management, the focus is on employing blockchain to make it possible for a set number of identified parties to administer transactions with each other. Instead of coins, supply chain blockchains “mint” a range of transaction-related data, yielding unique and easily verifiable identifiers for purchase orders, inventory units, etc. Every participant in the blockchain has their exclusive digital signature, which is used to authenticate the tokens moving through the chain. In addition, everyone receives their copy of the chain and is given access to a built-in audit trail that can’t be meddled with. The potential of blockchain technology in supply chain management is exciting for the following reasons: Greater supply chain transparency Permissioned participants received increased visibility across all supply chain activities, and there is a shared, unchallenged version of the truth. A more resilient supply chain A single unanticipated event can trigger a domino effect of supply chain disruptions. Blockchain supply chain solutions counter this by issuing smart contracts that automatically activate when pre-defined business conditions are met. Consequently, participants receive near real-time visibility into operations and can take earlier action in the case of exceptions. Streamlined supplier onboarding By providing a trustworthy, unalterable record of new vendor details, blockchain supply chain solutions can speed up the process of new supplier onboarding. Counterparts FedEx has incorporated the blockchain into its chain of custody to help manage customer conflicts. The organisation is also a part of the Blockchain in Transport Alliance, and advocates adopting blockchain as an industry standard. DeBeers uses blockchain to keep track of the source and progress of every natural diamond they mine, helping them address consumer concerns surrounding the ethical sourcing of gemstones.","excerpt":"Walmart Canada is using blockchain to handle the invoices and payments of its 70 third-party freight carriers.","categories":["AI Features"],"tags":[],"author_name":"Srishti Mukherjee","publish_date":"2022-02-06T13:00:00","publication_year":"2022","word_count":623,"keywords":["Go","API","AI","RPA","ML","Git","ViT","Rust","GAN","R"],"extracted_tech_keywords":["AI","ML","R","Go","Rust","Git","API","GAN","ViT","RPA"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-walmart-uses-blockchain-to-manage-its-supply-chain\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10004916,"title":"Understanding Explainability In Computer Vision","content":"The session “Explainable AI For Computer Vision” was presented at the first of its kind Computer Vision conference, CVDC 2020 by Avni Gupta, who is the Technology Lead at Synduit. Organised by the Association of Data Scientists (ADaSCi), the premier global professional body of data science and machine learning professionals, it is a first-of-its-kind virtual conference on Computer Vision. The primary aspect of the talk is computer vision models most of the time act as a black box and it is hard to explain what is actually going behind the models or how the outcomes are coming from. She also mentioned some of the important libraries which can help to make explainable AI possible in Computer Vision models. According to Gupta, many a time, when developers create a computer vision model, they find themselves interacting with a backbox and unaware of what feature extraction is happening at each layer. With the help of explainable AI, it becomes easier to comprehend and know when enough layers have been added and what feature extraction has taken place at each layer. Gupta started the talk discussing why explainable AI or XAI is important. She put forward some important points as mentioned below- Understandable AI: It includes the reasons and justifications to support actionable recommendations for faster and more accurate decisions.Transparent AI: It includes the interpretability of predictions accuracy with the ability to trace data to the underlying logic and data.Impactful AI: The impactful AI provides an assessment of future business outcomes and allows scenario simulations to determine the best actions. Gupta stated that there are various problems related to a Computer Vision model, which are ML and AI models are black-box models as it is hard to understand how the models are predicting the outcomes.ML models are non-intuitive, which are difficult for stakeholders to understandThe key issues are trust, reliability and accountability. She then discussed some of the important techniques that can be used for interpreting Computer Vision models. The techniques are- SHAP Gradient Explainer SHAP (SHapley Additive exPlanations) is a game-theoretic approach to explain the output of any machine learning model. It connects optimal credit allocation with local explanations using the classic Shapley values from game theory and their related extensions. SHAP Gradient Explainer uses SHAP and integrated gradients as a combination for the equations. Visual Activation Layers Visual activation layers are used when you create a Convoluted network and you want to view what exactly does your model speak. Occlusion Sensitivity Occlusion Sensitivity is used when you have an image and you want to grey out an area to check how occluding parts of an image affects your Conv network model. Grad-CAM Grad CAM or Gradient-weighted Class Activation Mapping is one of the most widely used techniques. It is a technique for making Convolutional Neural Network (CNN)-based models more transparent by visualising the regions of input that are “important” for predictions from these models – or visual explanations. Integrated Gradients Integrated Gradients is a variation on computing the gradient of the prediction output w.r.t. features of the input. Furthermore, Gupta discussed some of the crucial libraries and tools that are helpful while building an explainable Computer Vision Model with some practical implementations in Python language. The libraries and tools are- ELI5: ELI5 is a Python package which helps to debug machine learning classifiers and explain their predictions. It provides support for popular machine learning frameworks and packages like Scikit- Learn, XGBoost, LightGBM, etc.tf-explain: tf-explain offers interpretability methods for Tensorflow 2.0 to ease neural network’s understanding. With either its core API or its tf.keras callbacks, you can get feedback on the training of your models.AIX-360: The AI Explainability 360 toolkit is an open-source library that supports interpretability and explainability of datasets and machine learning models. The AI Explainability 360 Python package includes a comprehensive set of algorithms that cover different dimensions of explanations along with proxy explainability metrics.Tootorch: tootorch is for PyTorch models for implementation XAI in Computer Vision techniques.What-if Tool:What-If Tool is a feature of the open-source TensorBoard web application, which lets users analyze an ML model without writing code.","excerpt":"The session “Explainable AI For Computer Vision” was presented at the first of its kind Computer Vision conference, CVDC 2020 by Avni Gupta, who is the Technology Lead at Synduit. Organised by the Association of Data Scientists (ADaSCi), the premier global professional body of data science and machine learning professionals, it is a first-of-its-kind virtual […]","categories":["Deep Tech"],"tags":["Computer Vision","computer vision future","computer vision neural network","explainability in AI","Explainable AI","machine vision tools","ML models"],"author_name":"Ambika Choudhury","publish_date":"2020-08-16T11:00:00","publication_year":"2020","word_count":679,"keywords":["computer vision neural network","data science","machine learning","Keras","AI","neural network","TensorFlow","ML","ML models","PyTorch","explainability in AI","computer vision","Computer Vision","Explainable AI","computer vision future","xAI","machine vision tools"],"extracted_tech_keywords":["AI","machine learning","ML","neural network","computer vision","data science","xAI","TensorFlow","PyTorch","Keras"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/understanding-explainability-in-computer-vision\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":19832,"title":"Mumbai University To Set Up First-Of-Its-Kind Institute For Artificial Intelligence","content":"Mumbai University’s Vidyanagari campus is set to get a brand new research and educational institute for Artificial Intelligence. The first-of-its-kind centre aims at finding solutions to social challenges by using modern techniques such as artificial intelligence and data science, among others. Devanand Shinde, acting vice chancellor, Mumbai University, told a newspaper, “The project will focus on applying artificial intelligence concepts and technologies to deliver high impact and measurable outcomes in the underprivileged and under-resourced communities.” The new institute in Mumbai University will reportedly be set up in collaboration with Wadhwani Institute for Artificial Intelligence, California. The Wadhwani Institute will support the new facility in training students and help researchers to work on development, deployment and application of AI. The proposed institute will offer fellowship and internship opportunities for students as well. “This collaboration aims at establishing a globally recognised institute for applied research and engineering in AI for social good. It will offer state-of-the-art research and education environment,” Shinde told another news publication. According to this year’s Analytics India Magazine survey, Mumbai University already tops the list of universities or schools that most number of AI professionals in India graduate from More and more companies are adopting artificial intelligence in their ecosystems in order to achieve business efficiency and revenue augmentation. In fact, according to a recent report by  Intel India, undertaken by the International Data Corporation (IDC), nearly 75 per cent firms anticipate benefits in business process efficiency and employee productivity with the use of AI. The report also suggested an impending lack of skilled talent in the AI sector. All India Council for Technical Education (AICTE) recently also revised the curriculum for the four year undergraduate program. Artificial Intelligence, Robotics and Internet of things were made mandatory for the induction program from academic year 2018-19 to prepare future engineers more employable.","excerpt":"Mumbai University’s Vidyanagari campus is set to get a brand new research and educational institute for Artificial Intelligence. The first-of-its-kind centre aims at finding solutions to social challenges by using modern techniques such as artificial intelligence and data science, among others. Devanand Shinde, acting vice chancellor, Mumbai University, told a newspaper, “The project will focus on applying […]","categories":["AI News"],"tags":["AI (Artificial Intelligence)"],"author_name":"Prajakta Hebbar","publish_date":"2017-12-18T06:34:54","publication_year":"2017","word_count":303,"keywords":["data science","Go","artificial intelligence","programming_languages:R","AI","programming_languages:Go","Aim","ViT","analytics","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","data science","analytics","Aim","R","Go","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/mumbai-university-artificial-intelligence-institute\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10001830,"title":"How India’s EV Industry Is Moving Past Roadblocks To Go Electric","content":"The electric vehicle market in India has witnessed tremendous popularity over the last couple of years. Experts have prophesied that the EV industry in India will grow significantly in the coming years. And behind the skyrocketing transformation, it is not just the government that is playing a vital role but also the big players in the automobile such as Reva Electric Car Company (RECC), and Indian app-based transportation network company Ola is working have been doing working day in and out. Mission Electric There was a time when the EV industry in India was at a very nascent stage. With a vehicle like Mahindra’s Reva, India was moving towards to become a nation with a significant number of EV. As a joint venture between the Maini Group of Bangalore and Amerigon Electric Vehicle Technologies (AEVT Inc.) of the USA, the Reva Electric Car Company (RECC) was founded in 1994. The prime motive was to develop and produce an affordable compact electric car.  Also, back then there were other automakers also who were working to produce and deliver the same, however, in 2001, RECC launched the REVA. In order to push the EV industry to reach a higher level, the Govt. of India in 2012 started the National Electric Mobility Mission Plan (NEMMP) 2020, under which the Department of Heavy Industry is implemented FAME-India Scheme (Faster Adoption and Manufacturing of Hybrid and Electric Vehicles). Talking about EV, today it’s not just the four wheelers that are emerging, two-wheelers like Ather is also making a great impact on the Indian EV industry. What started out just a concept is today taking the industry by storm. The industry is booming at such a speed that even the big players from the automobile industry are entering the space. Mahindra is already a leading company that has invested a lot on electric vehicles and also with its new tech-based premium e-mobility service, Glyd. And now the company has joined hands with Indian app-based cab company Ola. The partnership is all about a first-of-its-kind programme that brings across different segments, including e-buses, e-cabs, and e-rickshaws together on the Ola app. Kickstarted in the city of Nagpur, the Pilot project has 200 electric vehicles in the city. Also, as Mahindra has partnered with Ola, out of the 200 electric vehicles, 100 are Mahindra’s e2o Plus electric hatchback. Another player in the project was  Kinetic that supplied 100 of its e-rickshaws. However, the project had faced a critical roadblock as drivers were not happy with the electric vehicles and wanted to move back to petrol\/diesel vehicles. According to a source, the reason was that the operating expenses were high and drivers had to wait for a long time at the charging station. Despite all the roadblocks, in November 2018, Ola was in the headlines as the company took another step towards becoming a vital player in the Indian EV market. According to a report, Ola could manufacture the electric vehicle on its own or get into an arrangement where it has complete control over the manufacturing process. The Entry Of South Korean Giants The electric vehicle industry is without a doubt is booming and with the policies in favour, this expansion is prophesied to get bigger. Another thing that is acting as a catalyst is the permission to set up a personal charging station for individuals. Furthermore, it’s not only Mahindra and Ola who are the vital players in this ecosystem, but Bajaj Auto is also working on electric vehicles. Recently, the Indian EV market has welcomed a South Korean player. Two of the well-known carmakers Hyundai Motors and Kia Motors on March 19 made an announcement that they are planning to invest $300 million in Ola’s EV initiative. The two carmakers will work with Ola with their India-centric EVs and create a whole new EV infrastructure. Bottom Line The EV ecosystem in India is in such a stage that it is experiencing both rise and downfall. However, the big players from the car manufacturing industry still believe that EVs will soon become a significant part of India’s commuting space. No matter what the industry is downfalls and hurdles come as an add-on, but with little more staying power, the table can be turned. That is what India’s Ola and Mahindra is doing. Now with Hyundai and Kia joining the club, it seems things might change for the good of Indian EV industry.","excerpt":"The electric vehicle market in India has witnessed tremendous popularity over the last couple of years. Experts have prophesied that the EV industry in India will grow significantly in the coming years. And behind the skyrocketing transformation, it is not just the government that is playing a vital role but also the big players in […]","categories":["AI Features"],"tags":["hyundai","Interviews and Discussions","Ola"],"author_name":"Harshajit Sarmah","publish_date":"2019-04-12T21:19:29","publication_year":"2019","word_count":735,"keywords":["Go","Ola","AWS","AI","cloud_platforms:AWS","programming_languages:R","programming_languages:Go","ViT","R","hyundai","Interviews and Discussions"],"extracted_tech_keywords":["AI","AWS","R","Go","ViT","cloud_platforms:AWS","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-indias-ev-industry-is-moving-past-roadblocks-to-go-electric\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10014808,"title":"6 Best Top-End Workstations For Data Scientists","content":"With the silos of data gathering every day, organisations around the world have changed the way they were working a few years ago. The driving demand for adopting AI and data science has led companies to build products that can handle the huge collections of data as well as the complexity of workflows. Also, security is one of the crucial concerns, which is why organisations using sensitive or confidential data, often avoid the relying on cloud and prefer systems with advanced AI solutions. Here is a curated list of the best top-end workstations, in no particular order, for data scientists. Z Series By HP The Z series workstation by HP is packed with an onboard NVIDIA graphic unit among other advanced data science solutions. The Z line of products is designed in such a way that data scientists and machine learning developers can maximise productivity, cut down latency as well as lower the cost of their data science projects. These workstations are built to ensure the highest level of compatibility, support and reliability. With these workstations, one can easily create and deploy data science applications while managing the data science tools and models. Key configurations: Z4 and Z4 are ideal for complex machine learning computations. Z4: Built for advanced 3D CAD, BIM and machine learning. Up to 512 GB DDR4-2933 ECCUp to Intel Core i9 or Xeon®Scalable ProcessorsUp to 2x NVIDIA Quadro RTX 8000 or 1x AMD Radeon Pro WX9100Windows 10 Pro or Linux Z8: Optimal for simulation, 8K video editing and complex machine learning. Up to 3 TB DDR4-2666 ECCUp to Intel Xeon Scalable ProcessorsUp to 2x NVIDIA Quadro RTX 8000 or 1x AMD Radeon Pro WX9100Windows 10 Pro or Linux Price: Ranges from ₹91,643 – 312,001 You can buy the workstation here. Know more here. Lenovo P Series Workstations Lenovo P Series Workstations are equipped to address demanding AI, machine learning, and deep learning workflow needs. Some of the advantages of AI on Lenovo workstations include accelerating machine and deep learning workflows, along with data preparation, model training, visualisation, and freedom to execute a full end-to-end data science and analytics workflow entirely on NVIDIA Quadro RTX GPUs, among others. ThinkStation P920 is an ultra-high-end desktop for machine\/deep learning and model training, whereas ThinkStation P520 is for AI model development and edge computing. The ThinkStation P330 Tiny, on the other hand, is meant for AI edge-based inferencing workloads. Key configurations: The Processor is Intel Xeon E-2100\/ W Family\/ dual Intel Platinum  Operating System: Windows 10 Pro for Workstations, Ubuntu Linux, Redhat Linux\/ Enterprise Linux Memory: 64 GB to 128 GB Price: Ranges from ₹55,500- 263,321. You can buy the workstation here. Know more here. NVIDIA-Powered Data Science Workstations NVIDIA-Powered Data Science Workstations is a new breed of desktops and mobile workstations that help data scientists transform massive amounts of information into insights faster than ever before, by accelerating data preparation, model training, and visualisation. The workstations combine the advanced NVIDIA Quadro RTX GPUs with a data science software stack built on NVIDIA CUDA-X AI to deliver an integrated solution provided by leading workstation and system builder partners. In an interview with Triple Grandmaster Christof Henkel, Henkel mentioned that his workstation is powered by two deep learning units. One has 3x NVIDIA RTX 2080Ti, and the other one is a DGX Station with 4x V100. Key configurations: The data science workstations come with a comprehensive stack of tested and optimised data science software built on NVIDIA CUDA-X AI. This stack features RAPIDS data processing and machine learning libraries, NVIDIA optimised XGBoost, TensorFlow, PyTorch, and other leading data science software, providing enterprises with accelerated workflows for faster data preparation, model training and data visualisation. Price: To know the exact price of these workstations, you have to contact the dealers located in your area. You can buy the workstation from here. Know more here. Edge XT Workstations The EDGE XT workstations by NextComputing include two versions, one is powered by Intel, and the other is powered by AMD. EDGE XTi is a compact tower workstation powered by Intel’s powerful workstation-class processors and is configured with high-end graphics power and all the storage for performing complex computations. The EDGE XTa is powered by AMD’s Ryzen, Threadripper and EPYC CPUs. The configurations include high CPU core count to 64 cores for the 3D rendering of images and animations, encoding videos, Elastic\/Kibana data visualisation use cases and fast CPU clock speeds for 3D modelling use cases. Key configurations: The EDGE XTi includes — Workstation-class Intel processors for elite performanceAccommodates (3) dual-width or (4) single-width professional graphics cardsMultiple storage options including PCI Express-based or SATA-based SSDsWindows 10 64 Bit Pro The EDGE XTa includes — AMD Ryzen, Threadripper, or EPYC processor for elite performanceAccommodates (3) dual-width or (4) single-width professional graphics cardsMultiple storage options including PCI Express-based or SATA-based SSDsWindows 10 64 Bit Pro Price: Ranges from $1,995.00- 54,950 You can buy the workstation here. Know more here. Dell Data Science Workstation (DSW) The Dell Data Science Workstation (DSW) delivers a fully integrated AI hardware and software solution. It delivers the data science platform that is needed with performance and reliability. The workstation is powered by Dell Precision workstations with the NVIDIA Quadro GPUs and NVIDIA GPU Accelerated Data Science software stack and Intel Xeon CPUs. The Data Science workstations deliver the power to deploy and manage cognitive technology platforms, including machine learning, artificial intelligence and deep learning. The workstations have several versions, such as 7550, 7750, 5820 towers, 7920 racks and more. Key configurations: Dell Precision Workstations include — Several versions depending on the price, and each version includes different GPUs, such as NVIDIA Quadro RTX 8000, NVIDIA Quadro RTX 6000 and NVIDIA Quadro RTX 5000.Intel Xeon CPU, 64GB–384GB ECC RDIMM, Drives(1TB-3TB) NVMe PCIe Class 50 SSD. Price: The prices vary depending upon configuration. For instance, the price of New Precision 3640 Tower Workstation is ₹83,920.26, while the price of New Precision 3440 Desktop Workstation is ₹68,886.85. You can buy the workstation here. Know more here. Lambda TensorBook Powered by the NVIDIA RTX 2080 Super Max-Q GPU, TensorBook is a GPU laptop built for deep learning by Lambda. The deep learning laptop comes with Lambda Stack, which includes frameworks like TensorFlow, PyTorch, Keras, CUDA, and cuDNN and more. The TensorBook can be pre-installed with either Ubuntu 18.04, Ubuntu 16.04, or Windows 10 Pro. Also, one can dual boot the TensorBook with Windows 10 Pro and Ubuntu 18.04. Key configurations: Lambda TensorBook includes GPU RTX 2080 Super Max-Q, NVIDIA 2070\/2080 (8GB), Intel Core i7-10875H processor with 16 threads5.10 GHz turbo and 16 MB cacheRAM up to 64GB (2666 MHz)Storage up to 2 TB Price: $ 3,300 You can buy the workstation here.Know more here.","excerpt":"The driving demand for data science has led companies to build workstations that can handle the huge collections of data.","categories":["AI Trends"],"tags":["Artificia Intelligence in Data Science","data science master","data science projects","Data Scientist","Data Scientists","how artificial intelligence works","machine data intelligence","machine learning and data transform","VR Data Analytics"],"author_name":"Ambika Choudhury","publish_date":"2020-12-21T12:00:00","publication_year":"2020","word_count":1118,"keywords":["how artificial intelligence works","data science master","Data Scientist","deep learning","XGBoost","data science","artificial intelligence","PyTorch","machine data intelligence","analytics","Data Scientists","data science projects","machine learning","AI","machine learning and data transform","Artificia Intelligence in Data Science","Keras","TensorFlow","VR Data Analytics"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","deep learning","data science","analytics","TensorFlow","PyTorch","Keras","XGBoost"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/6-best-top-end-workstations-for-data-scientists\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":35603,"title":"How LinkedIn’s Recommendation System Is Generating The Perfect Job Match For You","content":"With more than 200 million users logging in every month, the platform has a hectic task of collecting, handling and serving the data accurately without any latency. Although there are many conventional on-demand services and state-of-the-art machine learning models, the engineers at professional social networking platform LinkedIn were flexible enough to combine these conventional strategies with their built in-house tools to drive better results. Challenges With The Recommendation System LinkedIn employs a technique called ‘online learning’, where the model is trained is to get smarter with every activity of the user. This helps in narrowing down on the targets — jobs or connections. One challenge the engineering team at LinkedIn faces is maintaining consistency on both ends; job poster and the candidate. The algorithms have to learn respective preferences of both job seeker and poster and deliver a relevant search by unifying multiple aspects like skills required, roles preferred- leadership beginner or preferred job location. In order to obtain proper feedback, the online learning recommendation system distinguishes between the interactions with the applicant and that of the recommended match, which helps in explicitly representing the channel in which a potential portfolio has been discovered. “One of the engineering challenges here is that we need to optimise for two-way interest. We want candidates to find the messages they receive compelling and not miss out on any opportunities, and we want hirers to reach out to candidates who are qualified and might be interested,” notes the engineering team at LinkedIn in a blog post. In order to achieve this, the team uses Open Candidate status as an indicator to present the candidates with opportunities where the prospect(qualifications\/eligibility) of getting hired is relatively higher. Profiling With AI Before delivering the results, the algorithm associated with the hiring process would consider information from job postings, search queries, feedback of the candidate. This information is aggregated in real time to learn about a user’s preference for a specific opening. This can also be interpreted in terms of fundamental mathematical equations as follows: The weight of a term i of type t from rating r in sourcing channel s is determined by the following equation: Where S is set of all channels, R is set of all feedback types T is set of term types ci,t,s,r is the number of candidates Whereas, the Personalization Feature for a candidate with term type t from rating r in sourcing channel s is determined by the following expression: The above expression gives the results regarding customised search which is a combination of various profile attributes and corresponding weights. These personalisation scores are then run through XGBoost model. This model then produces results which can then be manipulated for matches by checking the scores of specific attributes like skills or experience. The results of an online learning technique show that there is a 49.61 % improvement in Normalised Discounted Cumulative Gain (NDCG) when compared to a model without online learning. NDCG is a measure of ranking the information that is being received from search engines. This is used as a tool to measure the effectiveness of search engine algorithms. Since search results vary in their length, comparing these results with their associated scores requires normalisation. This is done by sorting at the corpus level. And these values can then be averaged to get an overall idea of the performance of the search engine ranking. Future Direction With employment rates hitting an all-time low in many parts of the world, LinkedIn is determined to contribute with its homegrown technologies, which have paved ways to a faster and better job search. LinkedIn has proved from time to time that to tap the most out of AI, it is necessary to build service-specific tools and frameworks in addition to the existing models and other upcoming personalised online platforms can learn a thing or two from LinkedIn’s success.","excerpt":"With more than 200 million users logging in every month, the platform has a hectic task of collecting, handling and serving the data accurately without any latency. Although there are many conventional on-demand services and state-of-the-art machine learning models, the engineers at professional social networking platform LinkedIn were flexible enough to combine these conventional strategies […]","categories":["Deep Tech"],"tags":["linkedin","Machine Learning","XGBoost"],"author_name":"Ram Sagar","publish_date":"2019-03-01T05:06:23","publication_year":"2019","word_count":644,"keywords":["Go","machine learning","programming_languages:R","AI","Machine Learning","programming_languages:Go","RAG","XGBoost","ViT","linkedin","R"],"extracted_tech_keywords":["AI","machine learning","XGBoost","RAG","R","Go","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/linkedin-recommendation-system-perfect-job-match\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10113199,"title":"This Indian AI Startup is Bringing LLMs to Your Kitchen","content":"upliance.ai, India’s AI-powered home appliance company that featured on Shark Tank India, recently secured a ₹34 crore seed round at a valuation of ₹143 crore. The funding was led by Khosla Ventures, renowned for investing in AI startups like OpenAI, Rabbit, and Sarvam. upliance.ai’s AI Cooking Assistant comes in the form of a jar with a smart 8-inch screen attached. It can prepare over 500+ dishes, including paneer stir fry, chole masala, steamed rice, and chai, for instance. “We are actually working closely with OpenAI, and it’s not just about the APIs. We have also started using our own trained models, so it’s not as simple as ChatGPT,” said Mahek Mody, upliance.ai chief, in an exclusive interview with AIM. “We want to build Netflix, but for food,” he added, saying that upliance offers you an infinite variety of food options. “You don’t need to worry about whether you know how to cook something or not, you could just follow instructions,” he said. Tech Behind upliance.ai The company uses a proprietary technology called Up⤴️AI, built to understand the details of cooking – right from getting ingredients ready to making complete meals. The smart jar is equipped with a blade, a thermal sensor, and a heating element, enabling it to execute more than 16 cooking functions such as chopping, sautéing, blending, heating, and steaming. “The jar looks at vibrations and the viscosity of the gravy. All these things help it determine how far ahead we are in the cooking process and if it is consistently cooked,” said Mody. The screen comes with the in-built ‘ChefGPT’, which answers all your queries about the recipes and suggests ingredients for the dish. Moreover, it can suggest new recipes based on the ingredients you have at your disposal. Not only does it provide step-by-step instructions while you are cooking, but also enhances your overall cooking experience. “We have now 25,000 meals cooked for customers on both the app and the device. For instance, we have approximately 420 unique recipes that people have cooked and we have cross-promoted food as well,” said Mody, adding that they started with 50 customers who pre-ordered the product. Why LLMs in the Kitchen? “The idea is that if one person cooks a dish in one corner of the world, another person should also be able to cook the same thing and all its variations,” said Mody. He further added that all the cooking data, people’s preferences, and feedback are streamed to the internet to maintain collective intelligence and improve their cooking algorithms. “Typically, you learn cooking from your parents. It’s mostly roti-sabzi, and dal–chawal. But suppose you want to make risotto, it would take awfully long to learn that from YouTube,” he said. Now, you can input the name of the dish into the generative AI screen that upliance has and it will generate a recipe for you. “For example, let’s say lasooni chicken is not on the appliance and is not pre-set, you can still go to the AI generator and type the name of the dish. It will generate the recipe for you,” he added. Furthermore, he said that the AI Cooking Assistant will consider the type of recipe and the temperature of the item before providing answers. “You could be in the middle of cooking and say, ‘Hey, I want to make this recipe vegan’. It will go ahead and make the changes in the recipe accordingly,” he explained. upliance.ai may introduce new products in the future but for now, one can buy the AI Cooking Assistant from the company website for ₹23,999. In our kitchens, we have several appliances like the microwave, induction cooker, mixer, coffee maker, etc., LLMs can help us use these devices more effectively, eliminating the need to manually press buttons to adjust settings as they would be smart enough to do so. upliance.ai’s Journey upliance.ai was launched in early 2023 by Mahek Mody and Mohit Sharma. Mody has previously worked with Ather, learning how to set up the supply chain and design production, while Sharma was at Chaayos, lapping up the science behind cooking. Last year, the startup successfully completed a pre-seed funding round, raising Rs 11 crore from investors such as Draper Associates, Rukam Capital, Rainmatter (a Zerodha Fund), and angel investors, including the co-founders of Ather Energy and Unacademy. With funding from Khosla Ventures now, the company aims to grow its revenue to ₹150 crore in 2024 and scale the production to 20,000 units per annum in the next 6 months. Besides Khosla Ventures, upliance is closely working with Qualcomm. “Qualcomm has given us multiple grants to invest in this technology. Qualcomm India has been a great partner to us,” concluded Mody.","excerpt":"The company uses a proprietary technology called Up⤴️AI, built to understand the details of cooking – right from getting ingredients ready to making complete meals.","categories":["Deep Tech"],"tags":["AI Tool","ChatGPT","LLMs"],"author_name":"Siddharth Jindal","publish_date":"2024-02-19T17:19:24","publication_year":"2024","word_count":781,"keywords":["Go","ChatGPT","API","OpenAI","AI","LLMs","GPT","Aim","generative AI","GAN","AI Tool","R"],"extracted_tech_keywords":["AI","generative AI","ChatGPT","OpenAI","Aim","R","Go","API","GPT","GAN"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/this-indian-ai-startup-is-bringing-llms-to-your-kitchen\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10021416,"title":"Guide to Sense2vec &#8211; Contextually Keyed Word Vectors for NLP","content":"Sense2vec is a neural network model that generates vector space representations of words from large corpora. It is an extension of the infamous word2vec algorithm.Sense2vec creates embeddings for ”senses” rather than tokens of words. A sense is a word combined with a label i.e. the information that represents the context in which the word is used. This label can be a POS Tag, Polarity, Entity Name, Dependency Tag etc. Word Sense Disambiguation Despite the ability to capture complex semantic and syntactic relationships amongst the words, Neural word representations obtained using Word2vec  fail to encode the context. This context is of extreme importance while disambiguating word senses.Sense2vec aims to solve this problem by generating contextually-keyed word vectors (i.e one vector for each sense of the word. Ex: Apple gets two vectors Apple|PROPER NOUN   When It is used  as the name of the company Apple|NOUN When it is used as the name of a fruit. Architecture We need to either manually or automatically annotate the words prior to training the sense2vec model. Here each word is assigned one or more labels. Each unique word to label a pair is a sense. Now we can train a skip-gram model or CBOW model to learn embedding for each of these senses. Example Package Explosion.ai implemented this model in python and packaged it as a library( github.com\/explosion\/sense2vec ). The sense2vec model from this package integrates with spacy seamlessly. Let’s play with this model. Pip installs this package with a single command !pip install sense2vec Getting started with this package is extremely easy. Standalone usage is as follows from sense2vec import Sense2Vec # Loading pretrained model s2v = Sense2Vec().from_disk(\"s2v_reddit_2015_md\/s2v_old\") We can get the embeddings of a sense i.e word along with labels by using “token +’|’+label” as a key. query = \"apple|NOUN\" vector = s2v[query] #vector is a dense embedding of size 128 The token can be a unigram or a phrase Examples: 'at_least_two_other_people|NOUN', 'Robert_Baratheon___|PERSON', 'Good_Will_Hunting|WORK_OF_ART', 'a_week_or_so_ago|DATE', This pre-trained model supports the following POS Tags and Entity names as labels ADJ             adjective NOUN       noun ADP             adposition NUM       numeral ADV             adverb          PART       particle AUX             auxiliary SCONJ      subordinating conjunction CONJ            conjunction SYM       symbol DET             determiner VERB       verb INTJ            interjection PRON             pronoun PROPN proper noun PUNCT punctuation PERSON People, including fictional. NORP            Nationalities or religious or political groups. FACILITY Buildings, airports, highways, bridges, etc. ORG             Companies, agencies, institutions, etc. GPE             Countries, cities, states. LOC             Non-GPE locations, mountain ranges, bodies of water. PRODUCT Objects, vehicles, foods, etc. (Not services.) EVENT Named hurricanes, battles, wars, sports events, etc. WORK_OF_ART Titles of books, songs, etc. LANGUAGE Any named language. x=s2v['king|NOUN']-s2v['man|NOUN']+s2v['woman|NOUN'] y=s2v['queen|NOUN'] def cosine_similarity(x,y): root_x=np.sqrt(sum([i**2 for i in x])) root_y=np.sqrt(sum([i**2 for i in y])) return sum([i*j for i,j in zip(x,y)])\/root_x\/root_y cosine_similarity(x,y)#result 0.76 The difference between a king and a man when added to a woman is very close to a woman. These vectors capture the semantic information well. This is not surprising as even word2vec models these relationships. Let’s look at the most similar senses for polysemic words. s2v.most_similar(‘apple|NOUN’)s2v.most_similar(‘Apple|ORG’)[(‘blackberry|NOUN’, 0.8481), (‘apple|ADJ’, 0.7543), (‘banana|NOUN’, 0.751), (‘grape|NOUN’, 0.7432), (‘apple|VERB’, 0.7349), (‘gingerbread|NOUN’, 0.733), (‘jelly_bean|NOUN’, 0.7278), (‘pear|NOUN’, 0.7213), (‘pomegranate|NOUN’, 0.7205), (‘ice_cream_sandwich|NOUN’, 0.7161)][(‘BlackBerry|ORG’, 0.9017), (‘>Apple|NOUN’, 0.8947), (‘even_Apple|ORG’, 0.8858), (‘Blackberry|PERSON’, 0.884), (‘_Apple|ORG’, 0.8812), (‘Blackberry|ORG’, 0.8776), (‘Apple|PERSON’, 0.8745), (‘Android|ORG’, 0.8659), (‘OEMs|NOUN’, 0.8608), (‘Samsung|ORG’, 0.8572)]Polysemic words sense disambiguation These embeddings captured the context very well. But it is the responsibility of the user of these embeddings to provide a label along with a token to select the right vector. Sense2vec package can infer these labels when provided with spacy’s document object. Following is an example of this kind of usage. import spacy nlp = spacy.load(\"en_core_web_sm\") s2v = nlp.add_pipe(\"sense2vec\") s2v.from_disk(\"s2v_old\/\") Sense2vec can be added as a component to the spacy pipeline. We can initialize this model with random values and train it or we can load a pre-trained model and update it according to our needs. doc = nlp('Power resides where men believe it resides. It’s a trick, a shadow on the wall.') That’s it, we can get embeddings, similar phrases e.t.c for all the supported phrases from this document. The spacy pipeline has a pos tagger and named entity recognizer components before the sense2vec component.Sense2vec component uses results from these components to create word senses. for i in doc: try: print(i,i.pos_,'\\n',i._.s2v_most_similar(3)) except ValueError as e: #If a token pos tag combination is not in the keyed vectors it raises Error so we need to catch it pass Comparison with word2vec Let’s compare these embeddings with word2vec embeddings in the context of classification. A simple neural network is trained and validated on a toy dataset here are the results. Sense2vec vs word2vec Sense2vec embeddings performed a little better than word2vec vectors. We should keep in mind that labels used by sense2vec are inferred using spacy’s taggers. Manual annotations can dramatically improve these representations. Conclusion Sense2vec is a simple yet powerful variation of word2vec. It improves the performance of algorithms like syntactic dependency parsing while significantly reducing computational overhead for calculating the representations of word senses. Code snippets in this post can be found at https:\/\/colab.research.google.com\/drive\/1xW3lcE5o_6jQ0L-TdUQ2tinZCnXZ2ITL?usp=sharing","excerpt":"Sense2vec is a neural network model that generates vector space representations of words from large corpora. It is an extension of the infamous word2vec algorithm.Sense2vec creates embeddings for ”senses” rather than tokens of words.","categories":["AI Trends"],"tags":["embeddings","named entity recognition NLP","Word2Vec"],"author_name":"Pavan Kandru","publish_date":"2021-03-06T13:00:00","publication_year":"2021","word_count":833,"keywords":["Go","AI","neural network","ML","NLP","Colab","Aim","Python","spaCy","embeddings","R","named entity recognition NLP","Word2Vec"],"extracted_tech_keywords":["AI","ML","neural network","NLP","Aim","Colab","spaCy","Python","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/guide-to-sense2vec-contextually-keyed-word-vectors-for-nlp\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10080166,"title":"Investment Philosophy of Bharat Innovation Fund","content":"In 2013, originating from IIM Ahmedabad, Kunal Upadhyay and Shyam Menon founded Infuse Ventures focusing on Series A, Series B, and Seed startups. After running the fund till 2017, they were joined by three partners—Som Pal Choudhury, Sanjay Jain, and Ashwin Raguraman to start their deep tech focused fund, Bharat Innovation Fund. In an exclusive interview with Analytics India Magazine, Som Pal Choudhury talks in-depth about the history of the founders, the origination of the fund, the philosophy behind investing in startups, and how their firm stands out for startups building tech for India. The story of the fund dates back to 2005, when Upadhyay founded the Incubation Centre at IIM Ahmedabad, CIIE and then subsequently put forth the idea of starting a fund in 2013. The team comes from a diverse background of some being in the investment field, some from operations, and some from technical backgrounds—making the firm ideal for analysing and supporting startups. BIF’s philosophy is investing in application-based AI startups. Their preferred segments in the field include cybersecurity, consumer behaviour, finance, and industrial applications. “The first thing we look at when investing in AI startups is the application,” said Choudhury. “AI needs a ton of data. Where the company is getting access to the data is the question we ask.” Companies that use proprietary data instead of data that is freely available are the ones that are typically thought to be the game changers in the landscape. “Models can be refined with time. But how the models perform in the initial dataset is a big factor before deploying it,” said Choudhury. The model needs to be scalable and flexible for the introduction of new data. “The model should be circular, so that the company can go back to the model and fine-tune it for extending the model.” BIF’s first investment was Entropik Technology in July 2018 raising $1.1 million, a company helping brands to measure emotional and cognitive response of consumers. Since then, the firm has made 17 investments that are publicly announced. Their latest investment was in August with Human Edge, a science-based company focusing on democratising the healthspan of workforce across the world and raised $1.5 Million. It was founded by Dr Marcus Ranney in 2020. Bharat Innovation Fund’s portfolio companies include Credit Vidya, A5G Networks, PlayShifu, vPhrase, Zumutor Biologics, and Detect Technologies. Philosophy in 2022 “Investors are much more cautious right now,” explains Choudhury about the current state of the investment circle. He pointed out the valuations of businesses are getting corrected. “A 20x or 40x expectation of revenue has gone down, sensibility has returned to the market.” Choudhury said that as investors, their firm opens the door for the founders to the market and the consumers. “If the founders are trying something new, we will help them strategise, and find the loopholes.” Being very hands-on, he adds that they help the companies build their decks and guide them towards their next funding round. “When it comes to expectations, we have primarily four things to look at,” explained Choudhury. Watching out for the signals from the market and customers. Hiring the right team and realising you cannot do it alone. Setting up the stage for the investors to establish the right market fit. Scaling the business from one million to five million. “India needs to focus on building AI infrastructure, not just applications” “Everybody is putting AI\/ML in their presentations,” points out Choudhury. During pitching ideas, founders need to establish clearly the source of their data and justify the use of AI\/ML for their applications. “You don’t need to start using algorithms as soon as you have collected data.” “We would like to invest in more fundamental developments in India, and not just building AI application models,” said Choudhury. Most of India’s development in the AI\/ML field has largely been related to building a model for a specific application and not standing out in the field by doing something groundbreaking to establish themselves as leaders in that space. A business idea should not be easily replicable. “If I bring together 20–40 engineers and can implement the same thing tomorrow, then it is not a unique idea.” “We are selling jet fuel for your jet” “Talk to your mentors. Talk to other founders who have pitched ideas and figure out what exactly it is that the investors are looking for,” advised Choudhury. Further, he adds that there might be great businesses that are growing and expanding, but not all of them should seek venture funding. “Your business might be a car, a bullock cart, or even a race car, but we are selling jet fuel. VC’s expect rapid growth and the founders need to be able to achieve it and scale their business.”","excerpt":"“We are selling jet fuel for your jet”","categories":["AI Startups"],"tags":["vcs investing in ai"],"author_name":"Mohit Pandey","publish_date":"2022-11-21T10:00:00","publication_year":"2022","word_count":789,"keywords":["vcs investing in ai","Go","API","AI","innovation","ML","Scala","RAG","analytics","R","startup"],"extracted_tech_keywords":["AI","ML","analytics","RAG","R","Go","Scala","API","innovation","startup"],"url":"https:\/\/analyticsindiamag.com\/ai-startups\/investment-philosophy-of-bharat-innovation-fund\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10070117,"title":"Highlights of the new PyTorch v1.12","content":"Close to four months after the release of version 1.11, PyTorch has now introduced v1.12. The new release contains 3124 commits and is developed with the help of 433 contributors. Some of the highlights of this release are – a functional API to apply module computation with a set of given parameters; TorchData’s DataPipes that are fully backwards compatible with DataLoader; functorch with improved API coverage; Complex32 and Complex Convolutions; TorchArrow, and others. Along with the release of v1.12, the team released the beta versions of AWS S3 Integration, PyTorch Vision Models on Channel Last on CPU, and supporting PyTorch on Intel® Xeon® Scalable processors with Bfloat16 and FSDP API. TorchArrow TorchArrow is a library for machine learning preprocessing over batch data. Introduced as a Beta release, TorchArrow features a performant and Pandas-style, easy-to-use API to speed up preprocessing workflows and development. Some of the features it offers include a high-performance CPU backend, vectorised and extensible UDFs with Velox; seamless handoff with PyTorch; zero-copy for external readers through Arrow in-memory columnar format. Functional API for Modules PyTorch v1.12 also has a new beta feature for functionally applying Module computation with a set of parameters. The traditional PyTorch Module usage pattern that maintains a static set of parameters internally is restrictive. This usually happens when implementing algorithms for meta-learning since multiple sets of parameters need to be maintained across optimiser steps. Some of its features include: Module computation with flexibility over parameters; reimplementation of module in a functional way is not needed; parameter or buffer present in the module can be swapped with externally-defined value for use in the call. Complex32 and Complex Convolutions in PyTorch Now PyTorch supports complex numbers, complex autograd, complex modules, and other complex operations. Several libraries like torchaudio and ESPNet use complex numbers in PyTorch, and the new version further extends the complex functionality with complex convolutions and the experimental complex32 data type that enables half-precision FFT operations.","excerpt":"The new release contains 3124 commits and is developed with the help of 433 contributors.","categories":["AI News"],"tags":["Pytorch"],"author_name":"Shraddha Goled","publish_date":"2022-06-29T18:00:08","publication_year":"2022","word_count":322,"keywords":["Pytorch","Go","machine learning","AWS","PyTorch","AI","ML","Scala","RAG","R","Pandas"],"extracted_tech_keywords":["AI","machine learning","ML","PyTorch","Pandas","RAG","AWS","R","Go","Scala"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/highlights-of-the-new-pytorch-v1-12\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10089399,"title":"Is GPT-4 Powered Bing Chat Still Crazy?","content":"Five weeks ago, Microsoft released Bing Chat, claiming that it uses an AI model which was more capable than ChatGPT and GPT 3.5 combined. With the release of GPT-4, Microsoft quickly integrated it into Bing Chat, supposedly leaving the older, ‘early version’ of the model behind. This might be for the better, as the previous model was extremely prone to hallucinations and alarming statements, such as claiming it was sentient. With the new model, the question still remains: Can GPT-4’s reduced tendency to create hallucinations tame the infamous Sydney? A Fresh Coat of Paint Following the comments from a Microsoft CTO in Germany regarding the integration of GPT-4 into Bing Chat, it seems that the software giant and OpenAI finally got their act together. Indeed, logging on to Bing Chat shows a host of new features powered by the all-new GPT-4, which has been rolled out by Microsoft over the last week. The first thing that stands out in the new Bing is the fluid UI with added options. As soon as the user begins interacting with the platform, they are given a choice between three conversation styles. These styles—termed ‘creative’, ‘balanced’, and ‘precise’—not only change up the colour scheme of the chatbot but also offer differences in replies. For example, the same prompt ‘Tell me about the history of Bangalore’, netted three vastly different responses. Different responses given by Bing Chat depending on conversation styles While the ‘precise’ option gives a short and crisp answer to the questions, similar to what one might find on Wikipedia, the ‘creative’ option provides a school essay-level response, complete with legitimate sources such as Britannica. The ‘balanced’ feature, as the name suggests, strikes a balance between the two, providing information while keeping the word count low. Apart from the visual changes, it does seem like Bing Chat has gotten better at giving responses, especially when it comes to grammar and syntax—one of the improvements claimed by GPT-4.  The previous limit of five prompts has been increased to 15, following which the chatbot requests the user to clear the conversation and start a new one. The bot was able to give the correct response to a complex prompt, which asked it to summarise the plot of Cinderella using words beginning with letters in an alphabetical order. While the bot initially stumbled, a small correction helped it to formulate the right answer, albeit a bit clumsily. Answer to a complex prompt with corrections. Missing Pieces One of the most exciting parts of GPT-4 is its multi-modal nature. This feature is still said to be in a research preview for safety reasons, so it’s not available on Bing Chat as well. Currently, the chatbot tells users that it is not capable of any multimodal features and that they cannot upload images into Bing Chat. Moreover, the bot seems to be conditioned to not reveal the nature of the model it is built on. Microsoft earlier revealed that Bing Chat works on a model they’ve termed ‘Prometheus’. This model, now confirmed to work on GPT-4, brings together Bing’s indexing and ranking data with the LLM and a component called the ‘Bing Orchestrator’. The Orchestrator communicates between the GPT component and the Bing indexing component for coherent responses. Last month, in a Linkedin blog, Microsoft referred to the GPT component of Prometheus as ‘Next generation GPT’ and mentioned that it is used to generate answers from Bing results. Prometheus block diagram. Source: Microsoft This so-called ‘Next generation GPT’ component seems to be GPT-4. This means that an earlier, pre-release version of GPT-4 was integrated into Bing Chat, with user feedback fuelling the current version we see today. In a blog, the Bing team wrote, “The very reason we are testing the new Bing in the open with a limited set of preview testers is precisely to find these atypical use cases from which we can learn and improve the product.” Looking back, it seems that the Bing team released GPT-4 into the wild with a set of conditioning prompts, now termed ‘Sydney’. However, as more users picked up the product and began using it, the team was able to iterate on the previous version of the GPT-4 model and bring it up to speed. With the gradual addition of features like conversation styles and limited conversation turns, it seems that the Bing team’s efforts have paid off to make Bing Chat safe and effective for widespread use. With the addition of GPT-4 and multi-modal features, Bing Chat might actually become one of the most novel use-cases of GPT-4 we’ve seen yet, and Sydney might become merely a nightmare in the rearview of Bing users.","excerpt":"Bing Chat has finally been confirmed to use GPT-4. Is it fixed now?","categories":["AI Highlights"],"tags":[],"author_name":"Anirudh VK","publish_date":"2023-03-15T13:00:00","publication_year":"2023","word_count":776,"keywords":["Go","ChatGPT","OpenAI","AI","Modal","Git","GPT","Aim","GAN","R"],"extracted_tech_keywords":["AI","ChatGPT","OpenAI","Aim","R","Go","Git","GPT","GAN","Modal"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/is-gpt-4-powered-bing-chat-still-crazy\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10070730,"title":"IISc joins hands with Nokia to unveil Center of Excellence for Robotics, AI and 5G","content":"On 8th July 2022, Nokia and the Indian Institute of Science (IISc) inaugurated the Nokia Center of Excellence (CoE) in Networked Robotics at IISc Bengaluru. Along with promoting inter-disciplinary research involving robotics and advanced communication technologies in 5G and Artificial Intelligence (AI), the CoE will also develop use cases across disciplines such as industrial automation, agriculture and disaster management. The center aims to facilitate engagement and cooperation between academia, start-ups and industry ecosystem partners to research and develop these use cases. “Next-generation communication technologies like 5G and 6G will contribute enormously to the growth of India’s economy. Our collaboration with a world-class company like Nokia will enable us to explore new frontiers for advanced technology research to benefit society as well as provide state-of-the-art training to our students to enable them to become technology leaders in the coming decades”, says Professor Govindan Rangarajan, Director at IISc. From launching and enabling rapid growth of 2G\/GSM technology in India, bringing high-quality 3G services, pioneering 4G\/LTE technology to preparing the nation for 5G services, Nokia has been an integral part of India’s remarkable progress in technology and connectivity over the last two decades. As part of this initiative, Nokia will fund the CoE for a period of three years to sustain the first phase of their partnership with IISc.“We want India to drive global innovation in an era of convergence where a few years from now, extended reality (XR) and digital-physical fusion will allow us to create, collaborate and communicate in unprecedented ways. There is substantial untapped intellectual capability and competence in India, and our collaboration with a prestigious institution like IISc will enable exciting possibilities for industry and society”, says Nishant Batra, Chief Strategy and Technology Officer at Nokia.","excerpt":"Nokia will fund the Center of Excellence for a period of years to sustain the first phase of their partnership with IISc.","categories":["AI News"],"tags":["5G","IISc","Indian Institute of Science","Nokia","Robotics"],"author_name":"Kartik Wali","publish_date":"2022-07-11T13:13:41","publication_year":"2022","word_count":287,"keywords":["Go","API","5G","artificial intelligence","AI","innovation","Git","Robotics","IISc","Nokia","automation","Aim","ViT","R","Indian Institute of Science"],"extracted_tech_keywords":["AI","artificial intelligence","Aim","R","Go","Git","API","ViT","automation","innovation"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/iisc-joins-hands-with-nokia-to-unveil-center-of-excellence-for-robotics-ai-and-5g\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10062611,"title":"A guide to automated time-series modelling with FEDOT","content":"In recent years, we have witnessed the emergence of various automated machine learning approaches. There are the frameworks and libraries for AutoML that have surprised the data science practitioners community with their results. FEDOT is also such a framework that can provide us with various features of automated machine learning. In this article, we are going to discuss FEDOT for automated machine learning and we will also cover an example of time series modelling using FEDOT. The major points to be discussed in the article are listed below. Table of contents What is FEDOT?Time series modelling using FEDOT Importing dataImporting modules from FEDOTData processing using FEDOTDefining task and modelInitiating modelling Making predictionsVisualizing pipeline What is FEDOT? FEDOT is a framework that supports automated machine learning modelling and is available to us as open-source. Using this framework we can customise the pipeline of machine learning modelling procedures. This library can be utilised for real-world problems in a simple and automated way where under the hood it uses various evolutionary approaches of modelling. Using this framework we can resolve problems related to classification, regression, clustering, and time series modelling. One thing which attracts us to this framework is that it has a variety of modules that can be used for end-to-end modelling. This framework also provides modules for basic processes like preprocessing of data, feature engineering, model optimization, etc. using this framework we can also build graphs that can tell us the procedure used to solve any problem is followed by the framework. Some of the other features of the framework are as follows: The architecture of the framework is flexible to create machine learning models using various types of data.This framework can support basic and popular libraries of machine learning like SK-Learn, Keras, Statsmodel, etc.This framework can also be used to embed models related to specific areas into pipelines like ODE and PDE.Using this framework we can enable ourselves to use a variety of models and increase the explainability of modelling procedures. We can install this framework using the following lines of code. !pip install fedot After installation, we are ready to perform any machine learning operation using FEDOT. Time series modelling using FEDOT In this section, we will look at an example of how we can perform time series analysis using the FEDOT library. Importing data Before starting the procedure we are required to acquire time-series data, for this example we are using the traffic data that can be found here.  Let’s import the data. import pandas as pd df = pd.read_csv('\/content\/drive\/MyDrive\/Yugesh\/fedot\/trafic.csv', parse_dates=['datetime']) df.head(10) Output: ‘ Here in the above output, we can see that in the data we have two variables[datetime and value] that will be needed in time series modelling. Let’s plot this data. import matplotlib.pyplot as plt from pylab import rcParams rcParams['figure.figsize'] = 18, 7 df.plot('datetime', 'value',c='magenta') plt.show() Output: Here we can see our time series where we have values of vehicles with dates. Now, after importing data we are ready to use FEDOT modules for time series modelling. Importing modules from FEDOT from fedot.api.main import Fedot from fedot.core.repository.tasks import Task, TaskTypesEnum, TsForecastingParams from fedot.core.data.data import InputData from fedot.core.data.data import train_test_data_setup from fedot.core.repository.dataset_types import DataTypesEnum In the above we have called FEDOT API, modules for solving tasks, modules for split, fit and predict, and FEDOTs data type module. Data processing using FEDOT Let’s prepare the data according to the FEDOT modules. We can load and split our data using the following lines of codes input_data = InputData.from_csv_time_series(task, '\/content\/drive\/MyDrive\/Yugesh\/fedot\/trafic.csv', target_column='value') train_data, test_data = train_test_data_setup(input_data) Defining task and model Let’s check the length of the data. print(f'Length of the time series - {len(df)}') Output: Here we can see that the length of our data is 801. So 144 prediction values will be enough so in the next we define a task for modelling using the modules of FEDOT. task = Task(TaskTypesEnum.ts_forecasting, TsForecastingParams(forecast_length=144)) Here in the above, we have defined a time series forecasting task where we will get 144 predictions. Initiating the modelling process Let’s initiate the model using the FEDOT API. model = Fedot(problem='ts_forecasting', task_params=task.task_params) chain = model.fit(features=train_data) Output: Here in the above output, we can see that after hyperparameters tuning this API has started the modelling. Making predictions Now we are required to make some predictions that can be done using the following lines of codes. forecast = model.predict(features=test_data) forecast Output: Here we can see the prediction from our models. Now for better optimization, we are required to visualize the prediction. This can be done using the following lines of codes. Defining function for visualization import numpy as np from sklearn.metrics import mean_absolute_error traffic = np.array(df['value']) def display_results(actual_time_series, predicted_values, len_train_data, y_name = 'Traffic volume'): plt.plot(np.arange(0, len(actual_time_series)), actual_time_series, label = 'Actual values', c = 'green') plt.plot(np.arange(len_train_data, len_train_data + len(predicted_values)), predicted_values, label = 'Predicted', c = 'blue') # Plot black line which divide our array into train and test plt.plot([len_train_data, len_train_data], [min(actual_time_series), max(actual_time_series)], c = 'black', linewidth = 1) plt.ylabel(y_name, fontsize = 15) plt.xlabel('Time index', fontsize = 15) plt.legend(fontsize = 15, loc='upper left') plt.grid() plt.show() mae_value = mean_absolute_error(actual_time_series[len_train_data:], predicted_values) print(f'MAE value: {mae_value}') Visualizing the results Here we can see the predictions that are close to the test data. We have also put the mean absolute results with the function. In the below, we can see the MAE. As we can see here that we have obtained good results from this modelling. Visualizing pipeline Let’s check the pipeline from which our modelling procedure has gone. chain.show() print('Obtained chain:') for node in chain.nodes: print(f'{node.operation}, params: {node.custom_params}') Output: In the above output, we can see that the modelling has gone through the ridge regression for the time series model and we can also see what parameters are being used for modelling. Final words In this article, we have discussed FEDOT, which is a framework available as open-source for automated machine learning modelling. Using the FEDOT framework, we have seen an example of time series modelling where the results are very satisfactory. References FEDOT documentation  Link for the codes","excerpt":"FEDOT is a framework that supports automated machine learning modelling and is available to us as open-source. Using this framework we can customise the pipeline of machine learning modelling procedures.","categories":["Deep Tech"],"tags":["AI (Artificial Intelligence)","Automl","Data Science","Machine Learning"],"author_name":"Yugesh Verma","publish_date":"2022-03-12T13:00:00","publication_year":"2022","word_count":996,"keywords":["data science","Automl","NumPy","machine learning","Keras","AI","ML","Machine Learning","Ray","Aim","Data Science","Matplotlib","AI (Artificial Intelligence)","Pandas"],"extracted_tech_keywords":["AI","machine learning","ML","data science","Aim","Ray","Keras","Pandas","NumPy","Matplotlib"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/a-guide-to-automated-time-series-modelling-with-fedot\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10047527,"title":"How Uber is Leveraging Apache Kafka For More Than 300 Micro Services","content":"Have you ever wondered about all the hardcore processing that goes on behind the scenes in your Uber app to get you to your destination with no interruptions? Today, even a minute of waiting seems like a long lag, but how do these machines provide high speed and quality real-time processing? The answer is simple – streaming platforms. Initially developed by LinkedIn, Apache Kafka’s real-time streaming data pipelines and applications are the backbone of Uber today. With massive deployments of Apache Kafka, Uber uses the service to process trillions of messages and multiple petabytes of data per day. The team has even called it the ‘cornerstone of the technology stack’. Apache Kafka empowers more than 300 microservices in Uber and various workflows like pub-sub message buses for passing event data from the rider and driver apps, streaming database changelogs to subscribers and processing data from Hadoop data lake. Source: Data flows through Kafka pipelines Let’s have a look at these services in brief: One of the problems Uber overcame was partition scalability. An individual Kafka cluster at Uber would possibly have more than one hundred brokers. The engineers found a plausible trend of underutilizing Kafka cluster resources that would invariably reduce the efficiency of system resources. The team also realized that the Apache Kafka consumer coding pattern does not handle poison pill messages and non-uniform processing latency for pub-sub message queuing. Consumer Proxy Uber has used Kafka to develop a novel push proxy and solve the issues with the Consumer Policy clusters. It fulfils two gaps – presents the consumers with a simple-to-use gRPC protocol to address the mismatch of Kafka partitions during message queuing, and prevents consumer misconfigurations by hiding them from consumer services. These essentially fetch messages from Kafka using Kafka binary protocol and separately send each message to a consumer service instance. The consumer service further processes these messages separately and forwards the results to the Consumer Proxy cluster. Once the cluster receives the gRPC status code, it aggregates the processing results of the consumer service’s message and commits offsets to Kafka. Source: Consumer Proxy Architecture Features of Consumer Proxy Uber has overcome the pub-sub message queueing system issues by implementing features via a client-side SDK. In addition, the team chose a proxy-based approach. The engineering team has taken a multiple programming approach with Go, Java, Python, and NodeJS services. While traditionally different services would be written in other languages for the various client libraries, Consumer Proxy makes it possible to implement only one programming language applicable to all services. This approach also makes it easier for the team to manage the 1000 microservices that Uber runs. Since the message pushing protocols remain unchanged, the Kafka team can upgrade the proxy at any time without affecting other services. Consumer Proxy also assists in limiting the blasting radius of rebalancing storms as a result of the rolling restart. It rebalances the consumer group by decoupling message consuming nodes from the message processing services. The service can eliminate the effects of rebalancing storms itself by implementing its group rebalance logic. Lastly, the proxy allows the system to consume messages using four nodes and evenly distribute them to all service instances. Traditionally, the Kafka consumer group has only been able to distribute it to 4 cases at most. Design Parallel Processing within Partitions The system achieved parallel processing of messages within a single partition by consuming a batch of messages in the Consumer Proxy cluster and then sending them parallelly to multiple instances of the Consumer Service. The number of Kafka partitions does not limit the number of consumer service instances since the single node can send messages to various consumer service instances. Source Out-of-Order Commit The out-of-order commit from consumer services to Consumer Proxy was introduced to overcome blockage for the consumer when Kafka cannot process messages in a batch. This makes it possible for consumer services to commit a single message to Consumer Proxy, unlike Kafka, where a commit leads to all messages with lower offsets being committed. Consumer Proxy commit will only mark the specific single message as committed. Consumer Proxy leverages out-of-order commit to fetch several batches of messages, insert them into the out-of-order commit tracker, and send them parallelly to consumer services. Then, once the range of messages is acknowledged, Consumer Proxy commits them to Kafka and repeats the process over again before the previously fetched messages have even been fully committed to Kafka. Source Dead Letter Queue Poison pill messages are those that cannot be processed, or take too long due to non-transient errors. While these generally block consumer cases, in many cases, it would be preferred to mark the messages for special handling and revisit them later. Uber’s dead letter queue (DLQ) topic stores poison pill messages, allows consumer services to send a gRPC error code, and explicitly instructs Consumer Proxy to persist messages to the DLQ topic. These will be marked as “negative acknowledged” in the tracker, and based on their needs, users can ‘merge’ or ‘purge’ these messages. Source To wrap the system up, Uber ensures that consumer services receive neither too few nor too many push requests by specified flow control. Flow control mechanisms include Consumer Proxy taking action to adjust message pushing speed after processing the various functions and the presence of a circuit breaker to stop pushing when the consumer service is down. Uber has developed an accessible use case of the Apache Kafka via its Consumer Proxy.","excerpt":"Apache Kafka empowers more than 300 microservices in Uber and various workflows.","categories":["IT Services"],"tags":["apache kafka"],"author_name":"Avi Gopani","publish_date":"2021-09-04T13:00:00","publication_year":"2021","word_count":910,"keywords":["Go","AI","Scala","microservices","RAG","Python","gRPC","Kafka","apache kafka","R","Java"],"extracted_tech_keywords":["AI","RAG","microservices","Kafka","Python","R","Go","Java","Scala","gRPC"],"url":"https:\/\/analyticsindiamag.com\/it-services\/how-uber-is-leveraging-apache-kafka-for-more-than-300-micro-services\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10020724,"title":"Generating High Resolution Images Using Transformers","content":"Transformers are known for their long-range interactions with sequential data and are easily adaptable to different tasks, be it Natural Language Processing, Computer Vision or audio. Transformers are free to learn all complex relationships in the given input as they do not contain any inductive bias, unlike Convolution Neural Networks(CNN). This on the one hand increases expressivity but makes it computationally impractical for long sequences or high-quality images. In December 2020, Patrick Esser, Robin Rombach and Björn Ommer, AI researchers from Heidelberg University, Germany, published a paper on combining the Convolutional Neural Network(CNN) and Transformers to overcome the problem of producing high-resolution images: Taming Transformers for High-Resolution Image Synthesis. To make transformers more efficient, the Taming Transformer method integrates the inductive bias of CNNs with transformers’ expressivity. To produce high-resolution images,  the proposed methods demonstrates: They use VQGAN CNNs to effectively learn a codebook of context-rich visual parts. Utilization of transformers to efficiently model their composition within high-resolution images. The Model Architecture of Taming Transformers The model architecture uses convolutional neural  VQGAN, which contains encoder-decoder and adversarial training methods to produce codebook (efficient and rich representations) of images. This GAN architecture is used to train the generator to output high-resolution images. Once this GAN training is over, the model architecture then takes only the decoder part as the input to the transformer architecture, a.k.a codebook. This codebook holds the efficient and rich representation of images(instead of pixels) in a compressed form that can be read sequentially. The transformer is then trained with this codebook to predict the next indices’ distribution in the given representation, similar to the autoregressive model, to synthesize the required output. Source : https:\/\/compvis.github.io\/taming-transformers\/ Some of the results from Taming Transformer Source : https:\/\/compvis.github.io\/taming-transformers\/ Tasks done by Taming Transformers Image CompletionDepth-to-image generationLabel-to-image generationPose-to-human generationSuper Resolution Quick Start with Taming Transformers Installation & Dependencies Clone the repository via git and download all the required models and configs. !git clone https:\/\/github.com\/CompVis\/taming-transformers %cd taming-transformers !mkdir -p logs\/2020-11-09T13-31-51_sflckr\/checkpoints !wget 'https:\/\/heibox.uni-heidelberg.de\/d\/73487ab6e5314cb5adba\/files\/?p=%2Fcheckpoints%2Flast.ckpt&dl=1' -O 'logs\/2020-11-09T13-31-51_sflckr\/checkpoints\/last.ckpt' !mkdir logs\/2020-11-09T13-31-51_sflckr\/configs !wget 'https:\/\/heibox.uni-heidelberg.de\/d\/73487ab6e5314cb5adba\/files\/?p=%2Fconfigs%2F2020-11-09T13-31-51-project.yaml&dl=1' -O 'logs\/2020-11-09T13-31-51_sflckr\/configs\/2020-11-09T13-31-51-project.yaml' Install all the dependencies via pip. %pip install omegaconf==2.0.0 pytorch-lightning==1.0.8 import sys sys.path.append(\".\") Demo – Taming Transformers via pretrained model Load the model and print the configs. from omegaconf import OmegaConf config_path = \"logs\/2020-11-09T13-31-51_sflckr\/configs\/2020-11-09T13-31-51-project.yaml\" config = OmegaConf.load(config_path) import yaml print(yaml.dump(OmegaConf.to_container(config))) Initialize the model. from taming.models.cond_transformer import Net2NetTransformer model = Net2NetTransformer(**config.model.params) Load the checkpoints. import torch ckpt_path = \"logs\/2020-11-09T13-31-51_sflckr\/checkpoints\/last.ckpt\" sd = torch.load(ckpt_path, map_location=\"cpu\")[\"state_dict\"] missing, unexpected = model.load_state_dict(sd, strict=False) model.cuda().eval() torch.set_grad_enabled(False) Load the example semantic image and convert it into a tensor. Here we are taking label-to-image generation as an example. from PIL import Image import numpy as np segmentation_path = \"data\/sflckr_segmentations\/norway\/25735082181_999927fe5a_b.png\" segmentation = Image.open(segmentation_path) segmentation = np.array(segmentation) segmentation = np.eye(182)[segmentation] segmentation = torch.tensor(segmentation.transpose(2,0,1)[None]).to(dtype=torch.float32, device=model.device) Visualize the segmentation. def show_segmentation(s): s = s.detach().cpu().numpy().transpose(0,2,3,1)[0,:,:,None,:] colorize = np.random.RandomState(1).randn(1,1,s.shape[-1],3) colorize = colorize \/ colorize.sum(axis=2, keepdims=True) s = s@colorize s = s[...,0,:] s = ((s+1.0)*127.5).clip(0,255).astype(np.uint8) s = Image.fromarray(s) display(s) show_segmentation(segmentation) Encode the above image tensors so as to get the codebook via VQGAN. c_code, c_indices = model.encode_to_c(segmentation) print(\"c_code\", c_code.shape, c_code.dtype) print(\"c_indices\", c_indices.shape, c_indices.dtype) assert c_code.shape[2]*c_code.shape[3] == c_indices.shape[1] segmentation_rec = model.cond_stage_model.decode(c_code) show_segmentation(torch.softmax(segmentation_rec, dim=1)) Take the decoder output to give it as input to the transformer model. def show_image(s): s = s.detach().cpu().numpy().transpose(0,2,3,1)[0] s = ((s+1.0)*127.5).clip(0,255).astype(np.uint8) s = Image.fromarray(s) display(s) codebook_size = config.model.params.first_stage_config.params.embed_dim z_indices_shape = c_indices.shape z_code_shape = c_code.shape z_indices = torch.randint(codebook_size, z_indices_shape, device=model.device) x_sample = model.decode_to_img(z_indices, z_code_shape) show_image(x_sample) The last step is to train the transformer so as to produce the required output. The code snippet is available here. You can check at each step how the transformer is being trained from decoder output to get the final image. The window size for training is 16 X 16. Important Links Check this link to run streamlit on Colab.Pre-trained models for image generation, depth-to-image generation, super-resolution, etc. are available here.For training the data from scratch, you can refer here. Conclusion In this article, we have given an overview of Taming Transformation for High-Resolution Image Synthesis. Instead of using pixels, the model is trained on a codebook from VQGAN whose decoder is then fed to transformer architecture to generate the required results. This post discussed the need, model architecture, results & tasks of the proposed method. It also consists of basic tutorials of using Taming Transformers pre-trained models. The final result shows that this method has outperformed previous state-of-the-art methods based on convolutional architectures. Colab Notebook : Taming Transformers Demo Official References are available at: GithubWebsiteResearch PaperOfficial Tutorial","excerpt":"Transformers are known for their long-range interactions with sequential data and are easily adaptable to different tasks, be it Natural Language Processing, Computer Vision or audio. Transformers are free to learn all complex relationships in the given input as they do not contain any inductive bias, unlike Convolution Neural Networks(CNN). This on the one hand […]","categories":["Deep Tech"],"tags":["autoregressive transformers","Convolutional Neural Networks","Generative Pre-Trained Transformer"],"author_name":"Aishwarya Verma","publish_date":"2021-02-24T11:00:00","publication_year":"2021","word_count":735,"keywords":["NumPy","AI","neural network","PyTorch","ML","Transformers","computer vision","Convolutional Neural Networks","Colab","Ray","Streamlit","Generative Pre-Trained Transformer","autoregressive transformers"],"extracted_tech_keywords":["AI","ML","neural network","computer vision","Ray","PyTorch","Transformers","Streamlit","Colab","NumPy"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/generating-high-resolution-images-using-transformers\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10005534,"title":"10 Most Used Databases By Developers In 2020","content":"This year, the Stack Overflow Developer Survey surveyed about 65,000 developers, where they voted on their daily-used programming languages, go-to tools, libraries and more. According to the survey, MySQL has maintained the top spot, followed by PostgreSQL and Microsoft SQL Server. Below here, we listed down the top 10 databases from the survey that are most used by developers worldwide in 2020. (The databases are sorted according to their rankings). MySQL Rank: 1 About: MySQL is one of the most popular Open Source SQL database management systems. Developed by Oracle, MySQL Database Software is a client\/server system that consists of a multithreaded SQL server that supports different back ends, several different client programs and libraries, administrative tools, and a wide range of application programming interfaces (APIs). Know more here. PostgreSQL Rank: 2 About: PostgreSQL is a powerful, open-source object-relational database system that includes some key features like reliability, feature robustness, and performance. It uses and extends the SQL language combined with many features that safely store and scale the most complicated data workloads. PostgreSQL comes with many features aimed to help developers build applications. It allows administrators to protect data integrity and build fault-tolerant environments and help manage data. Know more here. Microsoft SQL Server Rank: 3 About: Microsoft SQL Server is a relational database management system developed by Microsoft. The SQL Server 2019 includes a number of intuitive features, such as gain insights from all the data by querying across relational, non-relational, structured, and unstructured data, flexibility to use the language and platform of user’s choice with open source support, scalability and performance to improve the stability and response time of database, and more. Know more here. SQLite Rank: 4 About: SQLite is an in-process library that implements a self-contained, serverless, zero-configuration, transactional SQL database engine. It is an embedded SQL database engine and unlike most other SQL databases, SQLite does not have a separate server process. Know more here. MongoDB Rank: 5 About: MongoDB is a general-purpose, document-based, distributed database built for modern application developers and for the cloud era. It is one of the popular databases which includes both scalability and flexibility. MongoDB is a document database, which means it stores data in JSON-like documents. Know more here. Redis Rank: 6 About: Redis is an open-source, in-memory data structure store, used as a database, cache and message broker. It supports data structures such as strings, hashes, lists, sets, sorted sets with range queries, bitmaps, hyperloglogs, geospatial indexes with radius queries and streams. Redis is written in ANSI C and works in most POSIX systems like Linux, *BSD, OS X without external dependencies. Know more here. MariaDB Rank: 7 About: MariaDB Server is one of the most popular database servers, which turns data into structured information in a wide array of applications, ranging from banking to websites. It is developed as open-source software and as a relational database. It also provides an SQL interface for accessing data. Know more here. Oracle Rank: 8 About: Oracle Database is a multi-model database management system to run all of the workloads more securely, whether it’s on-premises or autonomously managed on Oracle Cloud Infrastructure. There are several intuitive features, such as the database management system that lets a user choose from many deployment options such as on-premises, Cloud@Customer, and public cloud. It helps in building highly scalable applications while supporting all data types including relational, graph, and structured and unstructured non-relational data. Know more here. Firebase Rank: 9 About: Developed by Google, Firebase is an application development platform for building mobile and web applications. It provides developers with adequate tools to develop high-quality apps as well as grow the user base. Firebase gives various functionalities like analytics, databases, messaging and crash reporting. Know more here. Elasticsearch Rank: 10 About: Elasticsearch is a distributed, open-source search and analytics engine for all types of data, including textual, numerical, geospatial, structured, and unstructured data. It is the central component of the Elastic Stack, which is a set of open-source tools for data ingestion, enrichment, storage, analysis and visualisation. The speed and scalability of Elasticsearch can be used for application search, website search, logging and log analytics, application performance monitoring, security analytics and more. Know more here.","excerpt":"This year, the Stack Overflow Developer Survey surveyed about 65,000 developers, where they voted on their daily-used programming languages, go-to tools, libraries and more. According to the survey, MySQL has maintained the top spot, followed by PostgreSQL and Microsoft SQL Server. Below here, we listed down the top 10 databases from the survey that are […]","categories":["Deep Tech"],"tags":["beginner python projects","Data Analytics Certification","data enrichment ai","databases","distributed graph database","firebase","MongoDB","MySQL","object store database","Oracle","postgreSQL","Redis","SQL","Types of Databases","what is database"],"author_name":"Ambika Choudhury","publish_date":"2020-08-25T16:00:42","publication_year":"2020","word_count":701,"keywords":["Elasticsearch","MongoDB","Ray","what is database","Data Analytics Certification","object store database","databases","PostgreSQL","firebase","postgreSQL","RAG","distributed graph database","analytics","AI","data enrichment ai","Oracle","MySQL","Types of Databases","serverless","Aim","SQL","beginner python projects","Redis"],"extracted_tech_keywords":["AI","analytics","Aim","Ray","RAG","serverless","Redis","Elasticsearch","MongoDB","PostgreSQL"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/10-most-used-databases-by-developers-in-2020\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10057027,"title":"NITI Aayog Launches Vernacular Innovation Program (VIP) in 22 Mother Tongues","content":"With an ambition to empower innovators and entrepreneurs across the country, Atal Innovation Mission (AIM), NITI Aayog has come up with a first of its kind Vernacular Innovation Program (VIP), which will enable innovators and entrepreneurs in India to have access to the innovation ecosystem in 22 scheduled languages by the Government of India. To build the necessary capacity for the VIP, AIM will be training a Vernacular Task Force (VTF) in each of the 22 scheduled languages. Each task force comprises vernacular language teachers, subject experts, technical writers, and the leadership of regional Atal Incubation Centres (AICs). To get the program rolling, AIM NITI Aayog is launching a train-the-trainer program where it will collaborate with the Design Department of IIT Delhi to coach the VTF in design thinking and entrepreneurship and the adaptation of these subjects in the 22 languages and cultures. Furthermore, industry mentors have joined hands to lend design thinking expertise, and CSR sponsors have agreed to support the program generously. After training the task force from December 2021 to April 2022, the ecosystem will be opened to vernacular innovators. VIP is an initiative to lower the language barrier in the field of innovation and entrepreneurship such that it will systematically decouple creative expressions and languages of the transaction, said Dr Chintan Vaishnav, Mission Director, AIM, NITI Aayog, at the launch event of the program. India may be the first nation to launch such an initiative where an innovation ecosystem catering to 22 languages, plus English, is built. By providing access to learning in one’s language and culture, AIM looks forward to enriching the local, regional, national and global innovation pipelines.","excerpt":"Atal Innovation Mission, NITI Aayog launches Vernacular Innovation Program (VIP) to empower innovators, entrepreneurs in 22 mother tongues","categories":["AI News"],"tags":["NITI Aayog"],"author_name":"Poornima Nataraj","publish_date":"2021-12-24T12:06:05","publication_year":"2021","word_count":274,"keywords":["Go","programming_languages:R","AI","innovation","programming_languages:Go","Aim","R","NITI Aayog"],"extracted_tech_keywords":["AI","Aim","R","Go","innovation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/niti-aayog-launches-vernacular-innovation-program-vip-in-22-mother-tongues\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10130071,"title":"‘If AI Works in India, it Can Work Anywhere’","content":"Globally, India has the second-largest number of residents who use AI (81%). Further, Asian countries tend to display greater subjective knowledge of AI, with South Korea, China, and India leading the way. “India offers one of the most relevant and exciting contexts for AI, considering the readiness within the country to adopt various technologies, including IoT and AI, alongside other foundational technologies,” said Shekar Sivasubramanian, the CEO of Wadhwani AI, in an exclusive interview with AIM. Further, he said that the country’s ability to skip extensive wired network development and move straight to wireless infrastructure demonstrates its readiness and adaptability. Sivasubramanian also highlighted that for AI to succeed on a large scale, access to diverse data, which is abundantly available in India, is required. While we are at the early stages of harnessing this data, the country’s multilingual and multicultural environment provides a unique opportunity to develop context-centric, powerful AI solutions. “If AI works in India, it can work anywhere. This is because the diversity within the population acts as a natural protection against bias, allowing for more accurate and representative AI models,” added Sivasubramanian. Sivasubramanian is an industry leader who brings 40 years of global applied technology and management experience towards creating positive and sustainable impact at scale. He is currently driving non-profit organisation Wadhwani AI’s efforts toward establishing AI-driven solutions and ecosystems for the benefit of millions across the developing world. Voice is the Future “In the next 5-10 years, we would completely stop using the thumb and use only voice-based interfaces,” the Wadhwani AI CEO predicted. He noted that to build meaningful voice-based interfaces, embracing the diversity of expression in India is crucial. And India’s varied linguistic and cultural landscape provides a unique testing ground for developing advanced voice AI technologies. Along similar lines, Pramod Varma, former chief architect of Aadhaar, told AIM, “Indian entrepreneurs should really look at voice as a completely new human-computer interaction method. It could be very powerful, and I think it’s going to happen because voice is natural to humans.” In a previous interaction with AIM, Sarvam AI also mentioned that it is currently working on a voice-based Indic LLM, which it plans to release this year. Earlier, chief AI scientist at Meta, Yann LeCun had said that in the next 10-15 years we won’t have smartphones, and will be using augmented reality glasses and bracelets to interact with intelligent assistants. “The last thing we might want is intelligent virtual assistants that help us in our daily lives. So today, all of us here are carrying a smartphone in our pockets; 10 years from now or 15 years from now, we’re not going to have smartphones anymore. We’re going to have augmented reality glasses,” said LeCun. Even Meta CEO Mark Zuckerberg has repeatedly stressed that neural interfaces represent the inevitable next step beyond current methods like typing on screens. Wadhwani AI Initiatives India grows 26% of the world’s cotton and nearly 100 million farmers rely on cotton farming for their livelihood. However, cotton is highly vulnerable to pests, causing yield uncertainty and financial distress to farmers. So, to help farmers protect their crops, Wadhwani AI has introduced CottonAce, an AI-powered early warning system available as an Android app. Lead farmers, who work with welfare programs, use the app to upload photos of pests and the AI algorithm analyses the photos, determines infestation levels, and provides actionable advice, which is then shared with neighbouring farmers, even those without smartphones. The app is available in nine languages, including English, Hindi, Marathi, Gujarati, Telugu, Kannada, Tamil, Odia, and Punjabi. In the health industry, the non-profit institute developing AI solutions for social good is an official AI partner of the Central TB Division (CTD), and are developing multiple interventions across the TB care cascade and helping India’s National TB Elimination Programme become AI-ready. They use AI to interpret the results of the LPA test to determine drug resistance to TB. Each LPA strip encodes the drug-resistance pattern of the patient via a series of activated (dark) and inactivated (light) bands corresponding to different regions of the genome of the Tuberculosis bacterium. Further, it is developing multiple AI solutions to reduce morbidity and mortality for mothers and children in low-resource settings by improving the quality of primary care and strengthening the first 1,000 days of life.","excerpt":"“In the next 5-10 years, we would completely stop using the thumb, and use only voice-based interfaces,” the Wadhwani AI CEO stated.","categories":["AI Trends"],"tags":["AI (Artificial Intelligence)","India"],"author_name":"Vidyashree Srinivas","publish_date":"2024-07-24T15:40:11","publication_year":"2024","word_count":719,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","virtual assistants","Aim","ViT","GAN","R","India","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","Aim","virtual assistants","R","Go","GAN","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/if-ai-works-in-india-it-can-work-anywhere\/","complexity_score":4,"technical_depth":9,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10011195,"title":"Top Open Source Recommender Systems In Python For Your ML Project","content":"Recommender systems have found enterprise application by assisting all the top players in the online marketplace, including Amazon, Netflix, Google and many others. These systems are the decision support systems that make the personalisation process better as well as smoother. It predicts and estimates the content of user preferences by extracting from various data sources such as previous database, data history, among others. Here, we have listed the top eight open-source recommender systems in Python, in no particular order, that you must try for your next project. LensKit About: LensKit is an open-source toolkit for building, researching, and learning about recommender systems. It provides support for training, running, and evaluating recommender algorithms in a flexible fashion suitable for research and education. LensKit for Python (also known as LKPY) is the successor to the Java-based LensKit toolkit and a part of the LensKit project. It enables researchers to build robust and reproducible experiments that can make use of the growing PyData and Scientific Python ecosystem, including Scikit-learn, TensorFlow, and PyTorch. Know more here. Crab About: Crab is a flexible, fast recommender engine for Python that integrates classic information filtering recommendation algorithms in various scientific Python packages, such as Numpy, Scipy, Matplotlib, among others. It is also known as Scikits.recommender that aims to provide a rich set of components from which one can construct a customised recommender system from a set of algorithms and be used in various contexts. The features of Crab include user-based filtering, item-based filtering, etc. Know more here. Surprise About: Surprise or Simple Python RecommendatIon System Engine is a Python SciPy toolkit for building and analysing recommender systems. The tool deals with explicit rating data. With a set of built-in algorithms and datasets Surprise can help you learn how to build recommender systems. It provides various ready-to-use prediction algorithms such as baseline algorithms, neighbourhood methods, matrix factorisation-based such as SVD, PMF, SVD++, NMF and many others. The features of surprise include easy dataset handling, easy to implement new algorithm ideas, among others. Know more here. Rexy About: Rexy is an open-source recommendation system based on a general User-Product-Tag concept. It has a flexible structure that has been designed to be adaptable with variant data-schema. The underlying codes of this system are entirely written in Python version 3.5. It is written in a highly optimised, Pythonic and comprehensive way that makes it so flexible against the changes. Rexy also uses Aerospike as the database engine, which is a high speed, scalable, and reliable NoSQL database. Know more here. TensorRec About: TensorRec is a Python recommendation system that allows you to quickly develop recommendation algorithms and customise them using TensorFlow. A TensorRec system consumes three pieces of data, which are user_features, item_features, and interactions. It uses this data to learn to make and rank recommendations. TensorRec learns by comparing the scores it generates to actual interactions, such as likes and dislikes between the users and items. Know more here. LightFM About: LightFM is a Python implementation of a number of popular recommendation algorithms for both implicit and explicit feedback. It includes efficient implementation of BPR and WARP ranking losses. The features of LightFM includes easy to use, fast (via multithreaded model estimation), and produces high-quality results. Using Cython, it easily scales up to very large datasets on multi-core machines. Know more here. Case Recommender About: Case Recommender is a Python implementation of a number of popular recommendation algorithms for both implicit and explicit feedback. It is basically a framework that aims to provide a rich set of components from which one can construct a customised recommender system from a set of algorithms. The tool has various kinds of item recommendation and rating prediction approaches, as well as various metrics validation and evaluation. Know more here. Spotlight About: Spotlight is a Python recommender system that uses PyTorch to build both deep and shallow recommender models. By providing both a slew of building blocks for loss functions representations and utilities for fetching or generating recommendation datasets, Spotlight aims to be a tool for rapid exploration and prototyping of new recommender models. Spotlight offers a number of popular datasets, including Movielens 100K, 1M, 10M, and 20M. It also incorporates utilities for creating synthetic datasets. Know more here.","excerpt":"Recommender systems have found enterprise application by assisting all the top players in the online marketplace, including Amazon, Netflix, Google and many others. These systems are the decision support systems that make the personalisation process better as well as smoother. It predicts and estimates the content of user preferences by extracting from various data sources […]","categories":["AI Trends"],"tags":["java project ideas","python machine learning","recommendation engines","recommender systems"],"author_name":"Ambika Choudhury","publish_date":"2020-11-04T17:00:01","publication_year":"2020","word_count":704,"keywords":["scikit-learn","NumPy","AI","recommendation engines","TensorFlow","PyTorch","Python","Aim","Matplotlib","recommender systems","SQL","java project ideas","python machine learning","R"],"extracted_tech_keywords":["AI","Aim","TensorFlow","PyTorch","scikit-learn","NumPy","Matplotlib","Python","R","SQL"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-open-source-recommender-systems-in-python-for-your-ml-project\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10162435,"title":"India Will Have its Own LLM Soon, Says IT Minister Ashwini Vaishnaw","content":"India is set to develop its own generative AI model, aiming to rival global platforms like ChatGPT and DeepSeek, Union IT Minister Ashwini Vaishnaw announced at the Utkarsh Odisha Conclave. “Very soon we will have our own LLM,” said Vaishnaw. The initiative will be driven by the IndiaAI Compute Facility, which has acquired 18,000 GPUs to support the creation of a LLM tailored to the country’s needs. The minister also disclosed plans to set up AI data centres in Odisha, reinforcing India’s efforts to enhance its AI ecosystem and digital infrastructure. With 18,000 GPUs in place, India is well on its way to build a homegrown AI model that will cater to the unique linguistic, economic, and social requirements of the nation. The Indian government has been ramping up investments in AI research and infrastructure to reduce reliance on foreign AI models and foster a self-sufficient digital ecosystem. DeepSeek has woken up every country to the same. With the Union Budget 2025 just around the corner, speculations and discussions are rife about who stands to benefit and what allocations will be made. While finance is obviously the big picture, considering how India’s AI Mission found its place in the previous Budget, AI will possibly take centre stage in the upcoming session. In the Union Budget 2024-25, the Indian government has underscored its commitment to AI by approving the comprehensive IndiaAI Mission. A substantial financial outlay of over ₹10,000 crore was earmarked for the mission over five years, of which ₹551 crore has already been allotted.","excerpt":"“Very soon we will have our own LLM,” said Ashwini Vaishnaw.","categories":["AI News"],"tags":["LLM"],"author_name":"Mohit Pandey","publish_date":"2025-01-29T17:18:17","publication_year":"2025","word_count":255,"keywords":["Go","ChatGPT","AI","LLM","Git","GPT","Aim","generative AI","AI research","R","llm_models:GPT"],"extracted_tech_keywords":["AI","generative AI","ChatGPT","Aim","R","Go","Git","GPT","AI research","llm_models:GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/india-will-have-its-own-llm-soon-says-it-minister-ashwini-vaishnaw\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10111121,"title":"What is Wrong with AI Text Detectors?","content":"In an AI-human tussle in academia last year, we saw professors allegedly accusing students of submitting AI-generated assignments. Professors used AI detectors like Turnitin and GPTZero to get to the bottom of things. They later found these tools don’t give an accurate analysis and often offer false positive results. Professors have since resorted to more verbal tests, while the number of students using LLMs to write these assignments and thesis remains high. It has been long established that AI text detectors don’t work, as it is impossible to determine how much of an article is human-written or LLM generated, with confidence. Simultaneously, there is research being done on how to accurately detect text produced by AI. This is a constant cat-and-mouse chase where each of them is improving significantly using different methods of detection and avoidance. Vivek Verma, an undergraduate student at University of California, and the creator of AI text detector Ghostbuster, said, “There can be more use for the detectors than trying to implicate students on their assignments. These can be used to make the chatbots better at writing them in the first place.” Difficulty in Detecting AI Text While one can guess who wrote the text, it can be hard to be sure of it. A user on HackerNews pointed out that text books written on Quantum Mechanics got a 96% positive for being AI generated. This was obviously not the case, but the text followed a specific language pattern that overlapped with those often found in AI-generated texts. The models are often trained on datasets that do not fully encompass the diversity and complexity of human writing, especially in specialised fields like quantum mechanics. They analyse text based on patterns learned from these datasets, which can lead to inaccuracies when encountering texts that differ from their training material. In January last year, OpenAI released a classifier as an AI detection tool, and discontinued it only after six months citing low rates of accuracy. While there are other AI detectors like Copyleaks that claim 99% accuracy with 0.1% false positives, the AI text generators have stepped up their game. ChatGPT plugins like Humanize, WriteEasy, or even prompting it to write in a specific style would be enough to fool a detector. Yet these models overuse generic adjectives, use excessive passive voice, follow a formulaic writing style using a few syntax orders, without any creativity or originality. Humans on the other hand, tend to write with a certain ebb and flow in their sentences. They’re often in varying lengths along with the use of new and strange analogies. While AI generated text is grammatically correct, it lacks factual depth, or deep understanding of the topic with new connections. This and more such insights can be found in the paper ‘AI vs Human’. An Endless Cycle The website Verma developed was the result of a paper titled ‘Ghostbuster, Detecting Text Ghostwritten by Large Language Models’ from the University of California, and uses a new method to test material on their authenticity. They use three weaker models to run the text which needs to be checked. Verma said, “We realise these detectors are not completely trustworthy, but we can use them to differentiate between the origin of the text to train new models.” Another use case he explains is to make the models better at writing text in the first place. “Often chatbots like ChatGPT produce text that is too verbose. We can use the detectors as an evaluation tool to sort out the responses from better to worse,” he said. The next step is also to identify which part of a text is written by humans and AI with accuracy. There are detectors like GLTR (Giant Language Model Test Room), developed by Harvard and MIT-IBM Watson AI Lab that assigns areas with higher probability of being AI generated. “The final detectors should also explain why it categorises the content in each bucket and its reasoning behind it, which is what I’m working on next,” Verma concluded.","excerpt":"In cat-and-mouse chase, detectors strive to catch up with ever-advancing AI, creating a relentless cycle of evasion and detection.","categories":["AI Features"],"tags":["AI Tool","ChatGPT","GPTzero"],"author_name":"K L Krithika","publish_date":"2024-01-19T15:00:00","publication_year":"2024","word_count":667,"keywords":["GPTzero","Go","ChatGPT","OpenAI","AI","chatbots","GPT","Aim","ViT","Rust","AI Tool","R"],"extracted_tech_keywords":["AI","ChatGPT","OpenAI","Aim","chatbots","R","Go","Rust","GPT","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/what-is-wrong-with-ai-text-detectors\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10046486,"title":"Now Robots can Change Colours like Chameleons","content":"Imagine a robot changing its colours according to its surroundings in an absurd sequence of events, just like a chameleon. Better yet, imagine seeing a chameleon, but it is, in fact, a robot. Developing an artificial camouflage at the level of a whole device is one of the most challenging tasks for researchers. But, a team of South Korean researchers have paved the way for a new artificial camouflage technology. Chameleons can adapt and change their skin colour to blend with the background because of their iridophores, a particular layer of cells under the skin. Researchers Hyeonseok Kim, Joonhwa Choi, Kyun Kyu Kim, Phillip Won, Sukjoon Hong & Seung Hwan Ko have imitated this nature to create a robot chameleon capable of changing its colours. Modelling Technique Artificial camouflage is the functional mimicry of natural camouflage that can be observed in a wide range of species. The researchers stepped away from traditional microfluidic devices to take a fully electric approach. In the model, the robot uses colour sensors, silver nanowire-made tiny heaters and thermochromic materials that change colour on exposure to different temperatures, using thermochromic liquid crystal ink. The robot has colour sensors underneath and a range of pre-programmed patterns that help it blend in. According to Seung Hwan Ko, a thermal engineering professor and one of the authors, getting the colour transition up to nature’s speed was the most challenging. So, they used nanowire heaters to solve it. These machines heat up very fast, warm the artificial skin quickly and thus, it can change colour at the same speed as a chameleon. The team tested this on various high-resolution skin patterns to conclude that the robot could transition between different colours quite instantly. The Ag NW heater was used to develop the team’s TLC based artificial camouflage device. This allowed the model to manipulate the temperature accurately and express the fine patterns on it. The group proposed a multi-layered Artificial Chameleon Skin consisting of a TLC ink layer and a Silver Nanowire heater. “All fabrication processes consist of repeated procedures: polymer annealing, fabrication of patterned silver nanowire heaters on each layer, and silver nanoparticle electrodes for the electrical connection between the layers,” according to the paper published by Nature. Multi-layered Artificial Chameleon Skin design The team uses a microcontroller to determine the target colour based on the brightness and dominance of red, green, and blue lights. Then, a colour sensor measures these through a serial communication protocol, Inter-Integrated Circuit. The team had difficulty deciding whether to model the robot on a vertebrate or an invertebrate like an octopus or a squid. They chose an invertebrate model since it would allow for more freedom of movement. Chameleon’s simple body form was presented to suit the appropriate design and material structures for the team. They were able to shape the nanowires into dots, lines and scales to create a spot-on replica of the creature. Future applications of this model can be used in transportation, beauty, and fashion, including cars that adapt their colours to stand out and even colour-changing cloth. According to the researchers, the chameleon surface is just another type of display; it can be made softer, stretchable or flexible, given the need. Similar Researches Aaron Fishman, a materials scientist at the University of Bristol and his team have designed a system that mimics how cephalopod skin works. The model’s design involves soft and elastic materials that are transformed into muscles that can rapidly change in size and shape in response to electrical signals. According to the researchers, this technology can be used in soft robots covered in stretchy silicone skin. Another research, led by James Pikul of the University of Pennsylvania and Robert Shepherd of Cornell University, was inspired by octopus, cuttlefish’s papillae, or 3D bumps. The creature can inflate using muscle units for camouflaging. The team has created a synthetic version of cephalopod skin that can change from a flat, two-dimensional surface to a three-dimensional surface with bumps and pits. The researchers claimed that this ‘morphable skin’ could be used in furniture worn by robots for camouflage or entertainment where participants can feel these surroundings in a VR experience. The US army is looking into a new “smart” camouflage mimicking cephalopod’s ability to change its skin colour and patterns instantly. The model will apply to the army owing to the creature’s ability to match their surroundings to warn off attackers. The creature’s skin consists of 18-30 muscle fibres that change the amount of pigment exposed. It also has a static infrared reflecting space blanked on its skin. Inspired by this, the researchers’ plan consists of infrared camouflage coatings and invisibility covering. They can also regulate the body temperature by varying the mechanical strain on the fabric. However, these technologies are still in the initial stages and seeing their applications in everyday life is still in the coming future.","excerpt":"The team uses a microcontroller to determine the target colour based on the brightness and dominance of red, green, and blue lights.","categories":["AI Features"],"tags":[],"author_name":"Avi Gopani","publish_date":"2021-08-21T13:00:00","publication_year":"2021","word_count":810,"keywords":["API","programming_languages:R","AI","BERT","Aim","llm_models:BERT","R"],"extracted_tech_keywords":["AI","Aim","R","API","BERT","llm_models:BERT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/now-robots-can-change-colours-like-chameleons\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10061509,"title":"Now a dedicated GPT for the financial industry","content":"SambaNova Systems has launched SambaNova GPT Banking, a purpose-built solution leveraging deep learning to drive the digital transformation of banks. GPT Banking is built for banks’ large language models and is offered as a subscription service to simplify the process of deploying the most advanced language models in quick time. Banks can leverage the technology to perform: Sentiment analysis: Scan social media, press and blogs to understand market, investor and stakeholder sentiment.Entity recognition: Reduce human error, classify documents and reduce manual\/repetitive work.Language generation: Process, transcribe and prioritize claims, extract necessary information and create documents to improve customer satisfaction.Language translation: Language translation to expand customer base. The banking industry is not able to deploy AI models at a quick pace since there is a gap in deep learning deployment. SambaNova wants to shorten the time it takes to deploy a model, which usually is around 18 months starting from building a data science team, IT infrastructure, training and deploying a large language model. “GPT Banking has the potential to transform nearly every aspect of banking, from improving operations to managing risk and compliance. Customers tell me they’re most excited about GPT Banking’s ability to truly personalize the customer experience, with richer insights that enable them to understand customers’ evolving needs,” Rodrigo Liang, CEO and co-founder of the company said.","excerpt":"The banking industry is not able to deploy AI models at a quick pace since there is a gap in deep learning deployment.","categories":["AI News"],"tags":["SambaNova Systems"],"author_name":"Poulomi Chatterjee","publish_date":"2022-02-25T10:05:06","publication_year":"2022","word_count":218,"keywords":["data science","Go","AI","sentiment analysis","SambaNova Systems","Git","RAG","GPT","Aim","deep learning","R"],"extracted_tech_keywords":["AI","deep learning","data science","Aim","RAG","sentiment analysis","R","Go","Git","GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/now-a-dedicated-gpt-for-the-financial-industry\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10170893,"title":"[Webinar Alert] Boosting Performance of GenAI Models from Silicon to Software","content":"As enterprises scale up their Generative AI (GenAI) ambitions, performance bottlenecks and spiralling infrastructure costs have become key concerns. The solution? End-to-end optimisation—from hardware accelerators to software frameworks and application design. To decode this, AIM and Intel are presenting the third session in their GenAI webinar series: “All-Round Optimisation for Generative AI.” Register Now!📅 Date: 26th June 2025 🕔 Time: 5 – 6 PM IST The session will explore how to harness hardware accelerators like Intel® Advanced Matrix Extensions (Intel® AMX) in Intel® Xeon® processors, NPU in Intel AI PCs, and Intel® Gaudi® AI accelerators—combined with advanced software frameworks—to drive significant speed-ups in GenAI tasks while lowering Total Cost of Ownership (TCO). Led by Anish Kumar, AI Software Engineering Evangelist – APAC-s at Intel, this session is designed to equip AI\/ML developers with the latest techniques to maximise performance and efficiency for GenAI workloads. With over 20 years of experience in software engineering, Anish has been instrumental in enabling enterprises across the Asia Pacific region to accelerate AI adoption through engineering-led innovation. Why Should You Attend? Participants will gain a holistic understanding of optimisation strategies across the hardware, framework, and application layers, along with practical insights that help reduce power consumption, boost productivity, and shorten time to market. Key takeaways Unique needs of GenAI workloads Need for Optimising GenAI Optimisation Strategies for Hardware, Framework and Applications layer of GenAI Insights into the latest hardware acceleration technologies Strategies for reducing TCO and improving productivity Opportunities to network with experts and peers in the field The hands-on workshop would benefit AI\/ML developers, architects, and enterprise tech teams looking to scale GenAI efficiently and cost-effectively. Don’t miss this opportunity to learn how Intel’s AI ecosystem can help you build smarter, faster, and more sustainable GenAI applications. Register Now!📅 Date: 26th June 2025 🕔 Time: 5 – 6 PM IST","excerpt":"Led by Anish Kumar, AI Software Engineering Evangelist – APAC-s at Intel, this session is designed to equip AI\/ML developers with the latest techniques to maximise performance and efficiency for GenAI workloads.","categories":["AI Highlights"],"tags":[],"author_name":"Mohit Pandey","publish_date":"2025-05-28T16:32:15","publication_year":"2025","word_count":306,"keywords":["GenAI","programming_languages:R","AI","innovation","ML","Aim","ViT","generative AI","R"],"extracted_tech_keywords":["AI","ML","generative AI","GenAI","Aim","R","ViT","innovation","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/webinar-alert-boosting-performance-of-genai-models-from-silicon-to-software\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":29055,"title":"What Is Region Of Interest Pooling?","content":"Convolutional Neural Networks (CNNs) have become the leading method for classification and segmentation. In many cases, researchers also focus the attention of the neural network on a particular region in the image. This is known as the Region of Interest (RoI) inserted inside the net as a binary map. The region of interest pooling or better known as RoI pooling is widely used in object detection tasks using CNNs. According to DeepSense.ai, it is used for detecting multiple cars and pedestrians in a single image. Its purpose is to perform maximum pooling on inputs of non-uniform sizes to obtain fixed-sized feature maps. Also, some of the major tasks in computer vision are object classification and object detection. In the first case, the system is supposed to correctly label the dominant object in an image. In the second case, it should provide correct labels and locations for all objects in an image. Deep Learning Methods Based On CNN If a researcher is developing an algorithm for self-driving cars and is interested in leveraging a camera to detect other cars, motorists and pedestrians — he\/she will have to draw a box around every significant object and assign a class to it. This task is more challenging than usual classification tasks for MNIST or CIFAR. For example, for each frame of the video, there will be multiple objects, which are not clearly visible. Also, for these algorithms, performance has been cited as an issue. Especially, in cases of autonomous driving, researchers have to process tens of frames per second. Now, RoI pooling is a neural net layer used for object detection tasks. It was first proposed by Ross Girshick in April 2015 and has sped up the training and testing methods. It also maintains a high detection accuracy. In This Step, The Layer Takes Two Inputs: a) A fixed-size feature map generated from a deep CNN with several convolutions and max-pooling layers. b) An N x 5 matrix of representing a list of regions of interest, where N is a number of RoIs.The first column is representative of the image index while the rest of them are coordinates of the top left and bottom right corners of the region. Here’s How It Works: When CNNs are used to identify an image for which RoI map is given as input, it searches for different kinds of features in the region. Then, the CNN extracts features from the images by convolving a set of different filters with the image to create a feature map for each filter Therefore, researchers apply a different set of filters to the background and RoI regions. A New Approach To Saliency Maps Now, a new research paper talks about a new technique to overcome the reliance on saliency maps, used widely to highlight the region of an image that influences the classifier’s decision the most. The paper from Montréal Institute for Learning Algorithms talks about a new approach to models as opposed to deep neural networks which are more expensive to train. The paper discussed a guided backpropagation and SmoothGrad algorithms to obtain our saliency maps. For the purpose of improving the visual coherence of the maps, the researchers suggested smoothing them by iterating a combination of convolutions and thresholds. Saliency maps, are also known as activation maps and are used to understand the influence of an individual pixel in the final classification. These maps can be generated by computing the gradient of the maximally responding output unit with respect to the input. Since the saliency maps based on this raw gradients are visually noisy, researchers prefer to implement a guided backpropagation algorithm with SmoothGrad, resulting in a more visually coherent map. The researchers designed a convolution-based algorithm that computes iteratively a score (mean over a neighbourhood of pixels) and assigns a threshold for those values. If a pixel has the same value than its neighbours it will remain the same, otherwise, it gets the value of the other neighbours otherwise. In order to smooth all the points, we need to repeat this procedure several times depending on the filter and image size.","excerpt":"Convolutional Neural Networks (CNNs) have become the leading method for classification and segmentation. In many cases, researchers also focus the attention of the neural network on a particular region in the image. This is known as the Region of Interest (RoI) inserted inside the net as a binary map. The region of interest pooling or […]","categories":[],"tags":[],"author_name":"Richa Bhatia","publish_date":"2018-10-09T11:08:17","publication_year":"2018","word_count":683,"keywords":["Go","TPU","AI","neural network","computer vision","RAG","deep learning","object detection","CNN","R"],"extracted_tech_keywords":["AI","deep learning","neural network","computer vision","RAG","object detection","TPU","R","Go","CNN"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/what-is-region-of-interest-pooling\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10082986,"title":"4 Alternatives to OpenAI’s Point-E","content":"OpenAI created the hottest transformer-based text-to-image model, DALL-E, and then last week took another step to create text-to-3D generation, Point-E. From the depths of neural network fathom, Point-E is meant to be faster than Google’s DreamFusion and other text-based 3D models. 3D reconstruction models have proven to be very useful in improving the metaverse. Since Google released NeRF (Neural Radiance Field), a lot of new ones, used to make 3D models, have been created. Check out these text-to-3D models to try, apart from Point-E. Read: 3D Reconstruction Models Make Metaverse More Possible DreamFusion Google was one of the first ones to develop a 3D mesh creation software, DreamFusion. It uses 2D Diffusion to synthesise text-to-image and then using NeRF, the 3D model is rendered that predicts 2D images from random angles, which are then combined to create a single object. The model’s researcher paper proposes Score Distillation Sampling (SDS) for generating samples from diffusion models through optimization of loss function for allowing users to optimise samples in 3D spaces. Click here to read the research paper. Magic3D In November, NVIDIA released Magic3D for synthesising text into 3D models that offers even better quality output than DreamFusion by addressing two limitations – slow optimisation of NeRF and low-resolution space supervision of NeRF. NVIDIA leveraged the stylistic capabilities of eDiffi, their text-to-image diffusion model for transferring the style of the images into the 3D models. The best feature of Magic3D is that it allows users to edit the given input image through fine-tuning DreamBooth for optimising the 3D models using prompts, resulting in maximum fidelity in the generation. Read the research paper here. Text2Mesh Taking a different approach of stylysing a 3D mesh by predicting local geometric details and colours of an image, Text2Mesh, creates 3D models by conforming the details to the text prompt. Text2Mesh does not require a pre-trained generative model or a 3D mesh dataset and can also handle low-quality meshes. The researchers used CLIP’s ability to embed images and text to produce view-consistent and meaningful stylisation over the entire 3D shape. Click here to learn more about Text2Mesh. CLIP-Mesh Developed by researchers at Concordia University, CLIP-Mesh uses zero-shot generation technique for the generation of 3D models with a text prompt. As the name suggests, the generation relies on a pre-trained CLIP model used for comparing input text prompts with rendered images. The research paper points out the problems with Google’s method of converting meshes using NeRF, as it requires an additional step of converting an image into 3D models. CLIP-Mesh can directly generate 3D models. Click here to learn more about CLIP-Mesh.","excerpt":"Point-E is the newest text-to-3D model by OpenAI. Check out these other ones that were released recently","categories":["Global Tech"],"tags":["diffusion models","text to image generation"],"author_name":"Mohit Pandey","publish_date":"2022-12-22T10:00:00","publication_year":"2022","word_count":434,"keywords":["Go","DALL-E","TPU","OpenAI","text to image generation","neural network","AI","diffusion models","RAG","CLIP","R"],"extracted_tech_keywords":["AI","neural network","OpenAI","RAG","TPU","R","Go","CLIP","DALL-E","diffusion models"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/4-alternatives-to-openais-point-e\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10122845,"title":"Alibaba Releases Qwen2, Outperforms Llama 3 on Several Benchmarks","content":"In a significant leap for open source AI, Alibaba’s Qwen team has announced the release of Qwen2, an advanced version of its mother of LLMs, Qwen1.5. Qwen2 introduces five new models—Qwen2-0.5B, Qwen2-1.5B, Qwen2-7B, Qwen2-57B-A14B, and Qwen2-72B—each optimized for state-of-the-art performance across a variety of benchmarks. Click here to check out the model on Hugging Face. These models offer substantial improvements, including training on data from 27 additional languages beyond English and Chinese, including Hindi, Bengali, and Urdu. This multilingual training enhances Qwen2’s capabilities in diverse linguistic contexts, addressing common issues like code-switching with greater proficiency. Qwen2 also excels in coding and mathematics, with significantly improved performance in these areas. A standout feature of Qwen2 is its extended context length support, with Qwen2-7B-Instruct and Qwen2-72B-Instruct models capable of handling up to 128K tokens. This makes them particularly adept at processing and understanding long text sequences. Qwen2’s release includes various technical enhancements such as Group Query Attention (GQA) for faster speed and reduced memory usage, and optimized embeddings for smaller models. Performance evaluations show that Qwen2-72B, the largest model in the series, outperforms leading competitors like Llama-3-70B in natural language understanding, coding proficiency, mathematical skills, and multilingual abilities. Despite having fewer parameters, Qwen2-72B surpasses its predecessor, Qwen1.5-110B, demonstrating the effectiveness of the new training methodologies. Safety and responsibility remain a priority, with Qwen2-72B-Instruct performing comparably to GPT-4 in terms of safety across various categories of harmful queries. The model exhibits significantly lower proportions of harmful responses compared to other large models. The Qwen2 models, licensed under Apache 2.0 and Qianwen License for different versions, are set to accelerate the application and commercial use of AI technologies worldwide. Future plans include training larger models and extending Qwen2 to multimodal capabilities, integrating vision and audio understanding.","excerpt":"The model includes 27 additional languages including Hindi, Bengali, and Urdu.","categories":["AI News"],"tags":["Alibaba"],"author_name":"Mohit Pandey","publish_date":"2024-06-07T15:59:29","publication_year":"2024","word_count":292,"keywords":["Go","Hugging Face","AI","Alibaba","RPA","Modal","GPT","ai_frameworks:Hugging Face","llm_models:Llama","R","llm_models:GPT"],"extracted_tech_keywords":["AI","Hugging Face","R","Go","GPT","RPA","Modal","llm_models:GPT","llm_models:Llama","ai_frameworks:Hugging Face"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/alibaba-releases-qwen2-outperforms-llama-3-on-several-benchmarks\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10047423,"title":"India Releases Its First Quantum Computing Toolkit &#8211; Its Impact And Analysis","content":"India has reached a significant milestone in its research and development history. The Ministry of Electronics and Information Technology (MeitY) has released a long-awaited quantum computing simulation toolbox. This is also the first ‘Quantum Computer Simulator (QSim) Toolkit’ in the country. As per the report, IISc Bangalore, IIT Roorkee, and C-DAC have joined together for the first time in India to tackle the shared issue of expanding the quantum computing research frontiers in the nation. The major benefit is that the researchers and students will be able to conduct quantum computing research cost-effectively using this device. Quantum Progress: Using the power of Quantum Mechanics, quantum computing can execute a range of tasks faster and more efficiently than current digital computers. Quantum computing offers exponential development in computer power in fields like encryption, computational chemistry, and machine learning. QSim is a valuable teaching and research tool that can help draw students and researchers to quantum technology. Recent research states that Quantum gates, Quantum memory, Quantum CPUs, Quantum controlling and measurement, and Quantum error-correction tools are some of the most current developments in terms of the fundamental building blocks for quantum computers. Let us now look at some recent scientific advances. Quantum Gates: For example, each quantum circuit may be decomposed into a collection of Toffoli gates or Fredkin gates, and can also imitate NOT or CNOT gates (Controlled-NOT gates). As a result, either gate may be utilised to achieve Universal Quantum Computing. Quantum Memory: Quantum memories in quantum circuit calculations are composed of ‘n’ stationary quantum states. Information is processed by storing and processing these quantum systems in a quantum register. Quantum CPUs: To communicate between the functional parts of a quantum computer, quantum CPUs rely on a quantum bus. Quantum CPUs may be addressed from a computing standpoint through the building pieces that comprise them: quantum adders. Controlling and Measurement: For quantum state manipulations, readout, error-correction procedures, and fault-tolerant quantum computations, the necessary control mechanisms are required. Quantum Error-Correction: To achieve high accuracy and appropriate coherence periods in quantum computing, quantum error-correction techniques have been developed. Implementations of Quantum Algorithms: Scientists and students may use QSim to create and debug quantum algorithms. To be noted, “Shor’s prime factorisation algorithm” is one of the most important quantum algorithms. Some quantum algorithm milestones are listed below with a goal to raise awareness on the topic: J Romero et al. proposed molecular energy quantum computing methods. The model is based on the simulation of molecule energies using the variational quantum eigensolver (VQE) method. G.G. Guerreschi et al. looked at the possibility of practical optimisation for hybrid quantum-classical algorithms and found that it was possible. For linear systems and least squares, I. Kerenidis et al. investigated quantum gradient descent. L. Zhao et al. looked at leveraging the relationship between entangled states and graph theory to perform rapid graph operations in quantum computers. For topological and geometric analysis of data, S. Lloyd et al. devised quantum algorithms. N. Wiebe et al. looked at efficient phase estimation methods. Similarly, the list grows further. Researchers may use QSim to study quantum algorithms under idealised settings, and it helps them make the required preparations for experiments to run on quantum hardware. With the toolkit, students and users may learn to programme and construct actual quantum hardware. The project team said QSim delivers a powerful quantum computer simulator coupled with a graphical user interface based Workbench to design quantum programmes and observe the immediate circuit creation of simulated outputs. The Director of IIT Roorkee said researchers might use QSim to study quantum algorithms under idealised settings. It helps them make the required preparations for experiments to run on quantum hardware. Are we equipped? The Ministry of Electronics and Information Technology’s (MeitY) major initiation of QSim is a notable milestone in the history of Indian research. A while ago, Analytics India Magazine noted that quantum computing technology was already attracting the attention of major corporations. There was a list of the ‘Top Quantum Computing Project ideas‘ at the beginning of 2021 to provide researchers with a better understanding of the progress in the field. Top research institutions and top industries set forth to offer quantum computing courses, which many researchers and students are to benefit from. In the wake of QSim, India will soon have many significant accomplishments in quantum computing, undeniably proving to be a boon for the country’s start-ups and researchers.","excerpt":"Will QSim Be A Game Changer In The Quantum Computing History?","categories":["IT Services"],"tags":["MeitY","Quantum Computer","Quantum Computing","quantum supremacy"],"author_name":"Dr. Nivash Jeevanandam","publish_date":"2021-09-01T11:00:00","publication_year":"2021","word_count":733,"keywords":["Quantum Computing","Go","quantum supremacy","API","machine learning","TPU","programming_languages:R","AI","Quantum Computer","Git","RAG","analytics","MeitY","R"],"extracted_tech_keywords":["AI","machine learning","analytics","RAG","TPU","R","Go","Git","API","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/it-services\/india-releases-its-first-quantum-computing-toolkit-its-impact-and-analysis\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10102503,"title":"Elon Musk&#8217;s xAI to Launch ChatGPT Rival, Grok","content":"Elon Musk recently offered a glimpse of xAI’s first product. He shared a screenshot of xAI’s initial offering Grok, which seems to be a chatbot similar to OpenAI’s ChatGPT. Musk mentioned that xAI’s Grok system is designed to incorporate humor in its responses. He also stated that once it’s out of the early beta phase, xAI’s Grok system will be available to all X Premium+ subscribers. xAI’s Grok system is designed to have a little humor in its responses pic.twitter.com\/WqXxlwI6ef— Elon Musk (@elonmusk) November 4, 2023 Grok, as the name suggests, is designed to intuitively understand and empathise with users, establishing rapport while also providing an enjoyable experience. This AI system joins the ranks of interactive chat platforms, bringing a unique touch of empathy and understanding to user interactions. This development follows Elon Musk’s recent announcement on X, where he stated that xAI is set to release its first AI to a select group. According to Musk, this AI model is currently the best in some important respects. On July 12th of this year, Musk revealed the creation of X.AI, a company focused on the development of AI-powered technology. While early projections hinted at a swift release of their inaugural product, the timeline stretched out a bit longer than initially expected. As previously mentioned and the website reads, xAI’s mission statement is to “understand reality”, which is possibly to build an alternative to OpenAI’s woke chatbot, ChatGPT, as Musk calls it. Musk had been planning to build a rival since the beginning of the year and the formation of his AI company was planning to build just that. The acquisition of Twitter, which is now X, was also just a step towards building Musk’s AI lab. He has hinted many times before that he would be using the data from the social media platform to train the AI model. This is probably why he claims that it would be the best AI model that currently exists.","excerpt":"Grok system will be available to all X Premium+ subscribers.","categories":["AI News"],"tags":[],"author_name":"Siddharth Jindal","publish_date":"2023-11-04T10:11:43","publication_year":"2023","word_count":326,"keywords":["ChatGPT","OpenAI","AI","R","GPT","XAI","Aim","xAI","llm_models:ChatGPT","llm_models:GPT"],"extracted_tech_keywords":["AI","ChatGPT","OpenAI","xAI","Aim","R","GPT","XAI","llm_models:GPT","llm_models:ChatGPT"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/elon-musks-xai-to-launch-chatgpt-rival-grok\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10110837,"title":"OpenAI is Not Built by 24-Year-Old Programmers","content":"In the latest episode of Unconfuse Me with Bill Gates, OpenAI chief Sam Altman broke a myth or two about his company. “It is not run by a bunch of 24-year-old programmers,” he asserted. “You have a lot [of employees] in their 30s, 40s, and 50s. It’s not the early Apple and Microsoft where we were really just kids,” fretted Gates, now in his 60s. Altman said that somehow it is a bad sign for society. “The best founders have trended older over time. Then in our case, it’s a bit older than the average,” shared Altman, saying that he tracked this during his YC days where companies have gotten older in general. Is age just a number? Not just OpenAI, if you look at the majority of the tech companies, the likes of Perplexity AI, Midjourney, Stability AI, Anthropic, Cohere and others, most of them have folks in their late 30s or early 40s. The same goes for next generation tech startup people, the likes of Humane Ai and Rabbit, who are in their 30s and early 40s, having previously worked at companies like Apple and Baidu, respectively. “This is a topic that could engender endless debate (esp since team age != founder age),” said the general partner of RRE Ventures, Jason Black, highlighting the trend of older individuals increasingly founding successful companies, especially in the B2B sector. He said that seasoned entrepreneurs, rather than first-timers, often make the best founders due to their experience and ability to experiment more effectively, a luxury not as accessible to younger entrepreneurs in the past. He also noted that raising significant capital tends to be easier for those over 30. Additionally, Jason highlighted the complexities in pioneering new technological fields, which often require not just innovative breakthroughs but also new infrastructure and industry expertise. This expertise is frequently found in individuals who have already established notable careers, and hence are older. “While I think these factors influence the trend towards the ‘best’ founders trending older, our entire industry is based on the exceptions. Fortunately, exceptional people are exceptional regardless of their age,” shared Black. “As a founder in my 30’s, I am encouraged by this,” said founder of Passio AI, Dmitriy Starson, saying that most exciting, inspirational, e\/acc ideas he is seeing these days are coming from the 30-40 year olds. “40 is the new 20!! 😎” wrote founder of GAIM Network, Brady Lewis, saying that he understands the nuances of tech and business much better after spending eight years in a tech leadership role at Salesforce. The Rise of Older Tech Founders Research from the HBR supports this trend, indicating that the average age of successful tech founders is 45 years old. This data suggests that experience is increasingly valued across various industries, not just technology. A study published last year revealed that business founders aged 50 or older are more likely to introduce significant new products or services compared to younger entrepreneurs. Further, the study found that for every additional decade in a founder’s age, the likelihood of bringing something new to the market increases by 30%. This is preceded by the study by the Census Bureau and MIT professors published in 2018 challenging the tech industry’s youth bias. The research also highlighted that the probability of success increases with age. For instance, a 50-year-old founder is 1.8 times more likely to launch a successful company than a 30-year-old. If this continues, will it be likely that the founders in their 50s and 60s build successful companies?","excerpt":"Altman said that seasoned entrepreneurs, rather than first-timers, often make the best founders due to their experience and ability to experiment more effectively.","categories":["Global Tech"],"tags":["Anthropic","Cohere","Dario Amodei","Emad Mostaque","OpenAI","Perplexity AI","Sam Altman"],"author_name":"K L Krithika","publish_date":"2024-01-16T16:05:55","publication_year":"2024","word_count":587,"keywords":["Anthropic","Go","API","Sam Altman","startup","OpenAI","Emad Mostaque","Dario Amodei","AI","programming_languages:R","Perplexity AI","RAG","Aim","R","Cohere"],"extracted_tech_keywords":["AI","OpenAI","Anthropic","Aim","RAG","R","Go","API","startup","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/openai-is-not-built-by-24-year-old-programmers\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10130496,"title":"HP Unveils New AI PCs That Runs Language Models Locally in Bengaluru","content":"HP recently unveiled its most powerful AI PCs, which are aimed at empowering large enterprises, startups, and retail consumers to enable a never-before-experienced PC experience. The newly launched laptops include the HP EliteBook Ultra and HP OmniBook X, being HP’s first Copilot+ PCs. Both laptops are designed and engineered around the Snapdragon® X Elite processor[v] and its dedicated Neural Processing Unit (NPU)6, capable of 45 trillion operations per second (TOPS) to run language models locally on the device. The new HP laptops feature the HP AI Companion, a built-in personal AI assistant that uses generative AI to analyse and assess personal files, delivering more refined and focused responses. Additionally, these laptops come equipped with Copilot+, which amplifies creativity and productivity by providing a more personalised and powerful computing experience. The Poly Camera Pro further elevates virtual interactions by integrating advanced AI enhancements, such as Spotlight, Background Blur & Replace, and Auto Framing. This technology leverages the NPU to optimise performance and preserve battery life, ensuring smooth operation across various collaboration and streaming apps. While the EliteBook Ultra is designed for business leaders, the HP OmniBook X is designed specifically for retail customers, including creators and freelancers. The HP OmniBook X series is packed with advanced AI features, which significantly improve video quality and collaboration experiences. This makes them ideal for creators who are constantly on the move, needing reliable performance for tasks such as video editing, graphic design, and remote meetings. In recent months, HP’s competitors, Dell, Lenovo, and Acer, have launched their own versions of AI PCs in the Indian market. However, according to HP, they have the widest selection of AI PCs available in the market. “Our approach focuses on harnessing the best features from the ecosystem and integrating them with unique elements like the HP AI Companion, Poly Studio effects, and enhanced security features. This combination sets HP apart, making our products more efficient, productive, user-friendly, and secure compared to others in the market. “While many are adopting ecosystem-wide features, HP aims to offer additional benefits that are not available elsewhere, providing added value for our consumers,” Vineet Gehani, senior director of – the personal systems category at HP India, told AIM. Like other PC manufacturers, HP is banking on AI PCs to drive sales growth following one of the worst years on record for PC sales in 2023. Ipsita Dasgupta, senior vice-president and managing director of HP India, told Business Standard that in 2025, nearly 10% of all PCs sold will be AI PCs across the industry. “Over the next two to three years, we expect to see this figure go up to almost 50 per cent.” Gehani, too, appears optimistic. “It’s an exciting time as the PC market experiences renewed growth this year, and AI PCs are expected to further accelerate this trend. With our new product range and expanded portfolio, HP is confident that we will outpace market growth. Currently holding a 32% market share, we aim to leverage these innovations to grow even faster than the overall market,” he said.","excerpt":"HP launched the HP EliteBook Ultra and HP OmniBook X","categories":["AI News"],"tags":["HP"],"author_name":"Pritam Bordoloi","publish_date":"2024-07-29T15:38:30","publication_year":"2024","word_count":507,"keywords":["Go","programming_languages:R","AI","innovation","RAG","Aim","ViT","generative AI","HP","R","startup"],"extracted_tech_keywords":["AI","generative AI","Aim","RAG","R","Go","ViT","innovation","startup","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/hp-unveils-new-ai-pcs-that-runs-language-models-locally-in-bengaluru\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10059557,"title":"Is AI sexist?","content":"Sall Grover launched Giggle, a women-only networking platform, in 2020. The app allows women to form communities and search for potential business partners, housemates and friends. With categories ranging from gaming to periods to mental health, the platform offered a safe space for women to interact freely. However, Grover’s sign up process sparked a Twitter storm of late. To sign up on Giggle, users have to click a selfie, which the app then analyses using facial recognition software to verify their gender. “Giggle is for girls. Due to the gender-verification software that giggle uses, trans-girls will experience trouble with being verified,” the app’s homepage read. Though Giggle immediately took down the message, the app still doesn’t specify how trans women can enrol. Later, Grover said she wanted to be transparent about the limitation of the gender-verification technology and claimed to have consulted trans women while working on the app. However, the debate rekindled after she shared a Dailymail article about a trans swimmer with male genitals who caused discomfort among cisgender women in the women’s locker room. In-built biases Facial recognition technology entails facial detection and verification, identification and classification. The technology is used to infer a user’s age, gender and emotions–and feeds on the biometric data to throw up results. Genderify, launched in 2020, determines the gender of a user by analysing their name, username and email address using AI. The platform was launched on the product showcase site Product Hunt and had in-built tools to categorise users into male or female. However, Genderify’s results carried gender stereotypes. For example, prefixes like doctor or scientist leaned towards male and the word ‘stupid’ had a 61.7% probability of being identified as female. The creator of Genderify, Arevik Gasparyan, originally conceived the app as a marketing tool–exploiting customer data, analytics and demographic data–to extract business insights. However, the Twitter backlash prompted the founders to take the app off circulation a week after its launch. Binary technology Gender experts and digital rights activists have sounded the alarm on the risks of computer-driven image recognition systems that determine gender. Automatic Gender Recognition technology (AGR) is equipped to only detect two genders – male and female–by assessing the bone structure of their faces, effectively cancelling trans and non-binary genders and people of colour. There are two major underlying issues at play here: Firstly, a person’s gender is inferred only by assessing their physical characteristics; secondly, AGR assumes gender and sex to be the same thing. Humans generate, collect and label the datasets, and also pick the algorithm to analyse the data. As a result, the datasets will have inherent biases, and the chances are, the biases of people building the model might also creep into the algorithm. According to a paper published by the Stanford Social Innovation Review in 2021, only 22% of women work in data science and AI. In addition, men in low-income and middle-income households are 20% more likely to own phones, and around 300 million fewer women use the internet on their phones than men. Complex solutions But how do we approach the issue of gender discrimination in AI? Being a woman herself, Grover intended to create an inclusive platform for women and ended up with a trans-exclusionary app. The deployment of this technology could have serious consequences: Say if the surveillance system was designed to alert security if a person of the ‘wrong gender’ entered a bathroom. Os Keyes studies the nexus between gender and technology at the University of Washington. She is openly transgender. “Frankly, all the trans people I know, myself included, have enough problems with people misgendering us, without robots getting in on the fun,” Keyes said. Tech giants like Amazon and Microsoft, both of which offer facial-analysis services, have kept quiet over the solutions to these problems. In its Rekognition developer guide, Amazon noted the predictions on gender made by its facial analysis software should not be used to categorise a user’s gender identity. Clearly, it’s not a silver bullet. Gender data can be useful, but only if handled with care. Therefore, the stakeholders should work together and take stock of the ethical, social, and moral concerns around gender data before deploying such AI models.","excerpt":"Genderify, launched in 2020, determines the gender of a user by analysing their name, username and email address using AI.","categories":["AI Trends"],"tags":["AI facial recognition"],"author_name":"Poulomi Chatterjee","publish_date":"2022-02-01T18:00:00","publication_year":"2022","word_count":700,"keywords":["data science","AI facial recognition","Go","programming_languages:R","AI","innovation","image recognition","Git","Aim","analytics","R"],"extracted_tech_keywords":["AI","data science","analytics","Aim","image recognition","R","Go","Git","innovation","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/is-ai-sexist\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10090361,"title":"Does OpenAI Want to Make Robots Again?","content":"OpenAI, the startup that captured the world’s attention with ChatGPT, is once again setting its sights on developing robots. The San Francisco-based startup recently invested in a Norway-based robotics startup called 1x. Previously known as Halodi Robotics, the startup builds humanoid robots capable of human-like movements and behaviours. Interestingly, OpenAI also had an in-house robotic division, but in 2021, the startup scrapped it completely and decided to direct its resources towards AI research instead. Now, with 1X, is OpenAI reigniting its interest in robots? About 1X While OpenAI is focused on achieving artificial general intelligence (AGI), 1x is interested in exploring the potential of AI embodied in a human-like form. According to Brad Lightcap, chief operating officer at OpenAI, 1x is at the forefront of augmenting labour through the use of safe, advanced technologies in robotics. The startup, which recently raised USD 23.5 million in a Series A2 funding round led by OpenAI Startup Fund, will use the funds to develop NEO, its newest android iteration. ( NEO) Founded in 2014 and headquartered in Norway, the startup currently makes a wheeled humanoid robot called EVE. It will further use the funds to make EVE commercially available in Norway and North America. Robotics and LLMs Although the goal of developing human-like robots has been a persistent dream, it has proven to be a formidable challenge. Nevertheless, OpenAI may have the solution — LLMs. Recently, Microsoft, the company which has invested USD 1 billion in OpenAI, found that LLM chatbot can be used to control drones and robots. Researchers from the Redmond-based company were able to give the robots commands in simple English, using ChatGPT’s NLP capabilities to transcribe them to code. In a research paper titled ‘ChatGPT for Robotics’, researchers demonstrated a pipeline for applying ChatGPT to robotics tasks. The pipeline involves several prompting techniques such as free-form natural language dialogue, code prompting, XML tags, and closed-loop reasoning. Read more: LLM: The Linguistic Link that Connects Humans and Robots Robotic Embodiment for AGI? While OpenAI remains focused on AGI development, within the research community, there is an ongoing discussion about whether the human embodiment is a prerequisite for achieving AGI. Intrigued by this development, Chris Anderson, CTO at KittyHawk, tweeted, “One of the interesting questions is whether becoming ‘embodied’ (having senses and the ability to act in the physical world) could tip an AI into full AGI.” Even though currently it’s unclear whether the human embodiment is necessary to achieve AGI, a section of researchers believe that embodied intelligence, or the ability to interact with the environment and perceive sensory information, is essential for developing AGI. As opposed to their views, there is also a section that believes that it may be possible to achieve AGI without a physical body. However, there is not enough research done to justify either of the two contrasting views. Is robotic embodiment required for AGI?— Smoke-away (@SmokeAwayyy) March 28, 2023 If OpenAI is chasing AGI, its investment in 1x would make sense if it is ascertained that human embodiment is, in fact, a necessity for AGI. OpenAI’s History with Robotics Interestingly, OpenAI dismantled its robotics research team in 2021; however, prior to that, the startup invested heavily in robotics research. In 2019, OpenAI trained a robotic hand to solve a Rubik’s cube puzzle. But OpenAI discontinued research in robotics due to a dearth of data. The same was confirmed by OpenAI co-founder Wojciech Zaremba in a Weights & Biases podcast. According to him, building robots requires high computing capabilities. He believes there are two possibilities to successfully deploy robots. “One is to collect a lot of data. Another possibility is that we need powerful video models like that of powerful text models.” With OpenAI’s investment, 1x could possibly get access to high compute power, thanks to OpenAI’s partnership with Microsoft. Further, GPT-4, the most advanced LLM so far, is multimodal. This means it can generate content from both image and text prompts. Interestingly, video could possibly be the next iteration of LLMs. This means, very soon, OpenAI might have all the resources at hand to build humanoid robots. NEO vs Optimus: A Possibility? 1x is not the only startup making humanoid robots. Tesla, the autonomous vehicle firm, headed by Elon Musk, last year revealed Optimus, a conceptual general-purpose robotic humanoid. Interestingly, Elon Musk was an early investor in OpenAI, when the company was founded in 2015. However, over time, Musk’s relationship with the co-founders of OpenAI went sour. In a recent podcast with Lex Fridman, OpenAI co-founder Sam Altman opened up about their relationship with Musk. Read more: Altman Looks for ‘A Little Bit More Love’ from Musk Now, with OpenAI’s investment in 1x, there is a possibility of the startup encountering Musk once again in the field of robotics as 1x sets foot in the North American market.","excerpt":"OpenAI, the startup that captured the world’s attention with ChatGPT, is once again setting its sights on developing robots. The San Francisco-based startup recently invested in a Norway-based robotics startup called 1x. Previously known as Halodi Robotics, the startup builds humanoid robots capable of human-like movements and behaviours.  Interestingly, OpenAI also had an in-house robotic […]","categories":["Global Tech"],"tags":["Humanoid Robots"],"author_name":"Pritam Bordoloi","publish_date":"2023-03-29T18:30:00","publication_year":"2023","word_count":804,"keywords":["Go","ChatGPT","Humanoid Robots","OpenAI","AI","ML","NLP","GPT","R","Weights & Biases","startup"],"extracted_tech_keywords":["AI","ML","NLP","ChatGPT","OpenAI","Weights & Biases","R","Go","GPT","startup"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/why-does-openai-want-to-make-robots-again\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":69181,"title":"Here’s The ML Paper That Received Test Of Time Award","content":"Recently, the committee members of the International Conference on Machine Learning (ICML) 2020 announced the Test of Time Award. The receivers of the Test of Time award is a team of researchers from the California Institute of Technology, University of Pennsylvania and Saarland University. The test of time award is bestowed to a paper from ICML 10 years ago that has made a substantial impact on the field of machine learning, including both research and practice. This year, the award goes to the research paper “Gaussian Process Optimisation in the Bandit Setting: No Regret and Experimental Design.” The award recipient was selected by the ICML 2020 Conference General Chair and Program Co-Chairs in discussion with the current Test of Time Award committee, including the General and Program Chairs from ICML 2010. The ICML committee stated that the outcomes of this paper have shaped the prevailing methods of hyperparameter search in advanced deep learning systems, and the algorithm continues to be a successful acquisition algorithm even after ten years. The theorems, as well as the lemmas of this research paper, are quoted by various researchers in their recent papers, and their proof technique is also borrowed in several follow-up research papers. Speaking on the duo of technical depth and impact, this research paper clearly stands out. The committee added, “Bayesian optimisation using Gaussian processes is an important technique for black-box hyperparameter tuning and AutoML. However, there was no explicit finite-sample convergence theory for this method. This paper treats Bayesian optimisation as a bandit problem and obtained a cumulative regret bound for GP-UCB in terms of the information gain. This, in result, leads to a first finite sample theoretical analysis for Bayesian optimisation, and the paper will have significant long term impact both theoretically and practically.” About the Paper In most stochastic optimisation settings, evaluating unknown functions is expensive, and sampling is to be minimised. Some of the examples include choosing advertisements in sponsored search to maximising profit in a click-through model and learning optimal control strategies for robots. Several applications often require optimising an unknown, noisy function that is expensive to evaluate. This is where the Gaussian Process Upper Confidence Bound (GP-UCB) algorithm comes into play. The researchers at the California Institute of Technology, University of Pennsylvania and Saarland University formalised this task as a multi-armed bandit problem, where the payoff function is either sampled from a Gaussian process (GP) or has low RKHS norm. In this work, the researchers analysed a simple and intuitive Bayesian upper confidence bound based sampling method known as the Gaussian Process Upper Confidence Bound (GP-UCB) algorithm. The contributions made by the researchers are mentioned below: – The researchers analysed Gaussian Process Upper Confidence Bound (GP-UCB), which is an intuitive algorithm for GP optimisation when the function is either sampled from a known GP or has low RKHS norm.They bound the cumulative regret for GP-UCB in terms of the information gain due to sampling, establishing a novel connection between experimental design and GP optimisation.By bounding the information gain for popular classes of kernels, the researchers established sublinear regret bounds for GP optimisation for the first time.Further, they evaluated GP-UCB on real sensor network data, demonstrating that it compares favourably to existing algorithms for GP optimisation approaches. Wrapping Up Bayesian optimisation is one of the most useful approaches for object functions which are noisy as well as expensive to evaluate. It has become an important tool in various machine learning projects in the last few years. Also, Bayesian optimisation using the Gaussian process is considered as a crucial technique for black-box hyperparameter tuning as well as AutoML. According to a comment by the ICML committee, this paper leads to a first finite sample theoretical analysis for the Bayesian optimisation and had influenced many researchers to work on experimental designs, hyperparameter tuning, among others. Read the paper here.","excerpt":"Recently, the committee members of the International Conference on Machine Learning (ICML) 2020 announced the Test of Time Award. The receivers of the Test of Time award is a team of researchers from the California Institute of Technology, University of Pennsylvania and Saarland University. The test of time award is bestowed to a paper from […]","categories":["AI Trends"],"tags":["meditation"],"author_name":"Ambika Choudhury","publish_date":"2020-07-07T16:00:00","publication_year":"2020","word_count":640,"keywords":["meditation","Go","machine learning","programming_languages:R","AI","RPA","ML","programming_languages:Go","deep learning","R"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","R","Go","RPA","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/heres-the-ml-paper-that-received-test-of-time-award\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":69500,"title":"Apple Announces A New Free Course For Educators","content":"Apple announces a new free course for the US educators that can assist teachers in delivering superior coding lessons to students. The idea behind this initiative is to upskill educators, which will help instructors teach app development to students. The tech giant has two programs called ‘Develop in Swift’ and ‘Everyone Can Code,’ which are focused on educators and parents. This will allow them to teach coding to students\/children and bridge the skill gap in the market. Over the years, both programs have played a crucial role in educating people to learn to code. Consequently, under the Develop in Swift banner, Apple has released another course that will be taught by Apple experts,  which according to the company, can be an ideal introductory course for teaching the Develop in Swift course. “Apple has worked alongside educators for 40 years, and we’re especially proud to see how Develop in Swift and Everyone Can Code have been instrumental in helping teachers and students make an impact in their communities,” said Susan Prescott, Apple’s vice president of Markets, Apps, and Services. “We’ve seen community college students build food security apps for their campus and watched middle school educators host virtual coding clubs over [the] summer break.” Based on the feedback from educators, Apple has revamped the Develop in Swift curriculum, thereby making the learning experience even simpler. In addition, Apple has also released new books for Develop in Swift curriculum: Develop in Swift Explorations, Develop in Swift AP CS Principles, Develop in Swift Fundamentals, and Develop in Swift Data Collections (will be available this fall), for free in Apple Books. And for Everyone Can Code curriculum, the new book is Everyone Can Code Adventures, that has advanced level techniques. For parents to support their kids, Apple has released A Quick Start to Code, which is designed for learners of age 10 and above. For more information click here.","excerpt":"Apple announces a new free course for the US educators that can assist teachers in delivering superior coding lessons to students. The idea behind this initiative is to upskill educators, which will help instructors teach app development to students. The tech giant has two programs called ‘Develop in Swift’ and ‘Everyone Can Code,’ which are […]","categories":["AI News"],"tags":["App Development"],"author_name":"Rohit Yadav","publish_date":"2020-07-10T15:59:54","publication_year":"2020","word_count":315,"keywords":["App Development","R","programming_languages:R","AI"],"extracted_tech_keywords":["AI","R","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/apple-announces-a-new-free-course-for-educators\/","complexity_score":1,"technical_depth":3,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":17963,"title":"A peek into the mind of a HRbot","content":"Chatbots can have various types depending on the value they impart A Chatbot has been in existence since 60’s, however, the power of computing and the advancement in technology has limited the application of Chatbots to the masses. By Definition, Chatbot is conversational engine driven by Artificial Intelligence (AI) and Machine Learning. Users can interact with Chatbots through voice and text messaging interfaces. Chatbots can be categorized into two; Open and Closed Domain. Open domain Chatbots means they are built to respond and learn to any possible interactions that may come their way. Closed domain Chatbot intends to respond to queries built for a specific industry or purpose. Although there are a few giants in the market who have implemented the open domain Chatbots, the success measurement is yet to be determined. HR Bots such as H# (pronounced Aitch Sharp), is designed and trained for Human Resource Industry, is an example of closed domain Chatbot. Chatbots are certainly taking over the HR functions by engaging employees using Natural Language Processing (NLP). To understand how the brain of Chatbot processes user messages, a typical Chatbot flow is in 5 steps as: Data Cleaning \/ Filtering: People have distinctive ways of texting. This may be accidentally or intentionally or personal standardization. Example, the word “Thanks” may be written as “Thnks” or “Thanx” or “Ty” to name a few. However, there may be times where typing errors are prone to happen, for instance, “Thkx”. Data cleaning is what goes first to make the actual English meaning out of the typed word. It is the same way how humans do, just when they are acquainted with a new word. Intent Identification The next step is to understand what a user is seeking to know. This process is called Intent identification. Intent identification is done through permutation of large training data set, so the bot can identify what user might be looking for. Navigating through data retrieval points Post intent identification, bots may need to read data from available sources (these sources may be static or dynamic content), which may form the basis of preparing a response. As an example, a user seeking leave application may need to be validated against the available leaves for his profile. Maintaining conversation There are no fixed patterns in humans interact for seeking information. A users message may be broken into multiple sentences or single sentence. As an example, If someone wants to apply for leave. They may say, “I want leave” or “I want SL for tomorrow” or “Apply SL” so on and so forth. Unlike earlier where fixed format messages where used in conversations, A Chatbot’s role is to maintain the entire conversation thread so that user patterns are handled and a decision to respond is arrived at. Responding perhaps Once the intent is known and bot has clearly identified what it is supposed to do (through conversational probing), the response can then be prepared and sent to the end user While the 5 step flow is what typically goes through, there is some learning happening throughout. A). User Feedback based learning: A simple mechanism to know, if the bot is doing it right is having a like or dislike button and if a bot understands the user’s messages. Although this quick feedback mechanism may not have many likes, the dislikes will definitely help the administrator to take corrective actions. B). Trained or Guided learning: This process involves training the bot with various utterances in which human may try to ask the same question. The training helps the bot better match and arrive at the intent to handle the conversation further. C). Self-Learning: Even training may at times not be enough to know what user needs. Hence, suggestion (based on closest match) when a bot is confused is the best way to handle the situation. Further, a user may choose from that given suggestion. This crowd sourced data can be used to auto-train the bot. Having explained how the Chatbot works and the learning mechanisms, the next integral part of the system would be analytics. Common analytics like which transaction was performed, User preferred channel of communication (Anroid, iOS, Facebbook), etc are already built. However, when it comes to analytics there are two categories of analytics that would be of use. These include; Conversation engagement: Helps to find out how long a user has been engaged to complete the conversation. This insight helps to understand messaging style a user interacts to complete a transaction or query with the bot. Further, assisting in reducing the time to complete transactions. User Expectations: When users interact with Bots, they are open to seeking any information for which the bot may not be plugged in. In such cases, a bot may not be able to answer at the immediate hour. However, this metric helps to understand queries or transactions expected from Bot by the mass, helping to build a system over user needs. With all the user information, analytics and bot learning, we ultimately come down to the next level of how AI could ease our tasks, i.e. Suggestive decision making. Bot being aware of the available data points further can assist in decision making. For example, when a manager receives a leave for application for approval, the bot knowing the team members mapped to the manager can suggest if there are overlapping leaves within the team. This would offer assistance while making such decisions. “As we see above, Chatbot brain may have been developed by many players in market, hwever, for the brain to be called a Mind, Chatbots will need to be continuously Trained, Learn and Improve from user experiences. This is where closed domain bot like H#, helps an organization to have a well-developed mind than just providing a brain.”","excerpt":"A Chatbot has been in existence since 60’s, however, the power of computing and the advancement in technology has limited the application of Chatbots to the masses. By Definition, Chatbot is conversational engine driven by Artificial Intelligence (AI) and Machine Learning. Users can interact with Chatbots through voice and text messaging interfaces. Chatbots can be […]","categories":["IT Services"],"tags":["AI (Artificial Intelligence)","Machine Learning"],"author_name":"Godwin Pinto","publish_date":"2017-10-03T05:39:52","publication_year":"2017","word_count":962,"keywords":["Go","artificial intelligence","machine learning","programming_languages:R","AI","chatbots","Machine Learning","NLP","analytics","GAN","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","NLP","analytics","chatbots","R","Go","GAN","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/it-services\/a-peek-into-the-mind-of-a-hrbot\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10139866,"title":"Setting Up Multimodal Pipelines and Vector Databases for Production Scale","content":"As data complexity continues to rise, enterprises find it crucial to establish multimodal pipelines and manage vector databases at a production scale. In a recent discussion with AIM, Chetan Dixit, Client Partner, Cloud and Data Tech at Fractal, delved into the essential processes, the hurdles encountered, and the strategies to optimize performance and scalability. The advent of LLMs and generative AI has opened new possibilities. Earlier, processing different data types—video, audio, and text—was done in silos. “There used to be video data processing, audio data processing, and document processing,” Chetan noted. However, with the rise of LLMs, organisations can now generate insights by combining multiple data modalities. This is where the concept of multimodal pipelines comes into play. “A multimodal pipeline allows you to process different types of data, such as video and audio, and generate vector embeddings from each,” explained Chetan. These embeddings are then stored in a vector database. For instance, in a contact centre scenario, a customer might first engage with a support representative via chat and later follow up with a call. “You are processing different forms of data (audio from the call and chat transcripts) and generating vector embeddings, which are stored in the database,” he elaborated. By unifying different data streams, these pipelines ensure that the customer’s full history is available to the support agent, regardless of the communication medium. “So, when an agent talks to the customer, they have the full context, whether it is audio, video, or chat—all accessible at one go,” he said. The Challenges in Setting Up Multimodal Pipelines Chetan acknowledged that setting up multimodal pipelines comes with its own set of challenges. One significant hurdle is managing multiple data formats. Enterprises need multiple models to process different data types. For example, video and audio data each require specialised models, whether out-of-the-box or custom-built deep learning models. Performance is another key concern. “How performant your model is in processing that data is critical,” said Chetan. Furthermore, once the data is processed, it must be brought into a “standard normal form” before storing it in the database. This requires creating a unified data model to accommodate all data types. Chetan also pointed out that the process doesn’t end after the initial setup. The models need to be continuously monitored and tuned because the data changes over time. This requires a continuous cycle of optimisation, unlike typical data engineering pipelines. Vector Database Complexity Another layer of complexity arises with vector databases, which store the vector embeddings generated by these models. According to Chetan, vector databases are often distributed, which can lead to issues with data consistency. “In most vector databases, the data is eventually consistent, which means there could be a delay in the data being written from cache to disk,” he said. Maintaining data consistency is crucial for applications that rely on real-time information, such as customer service platforms. “You need to build integrity checks into your pipelines to ensure that data is being processed consistently across all modalities,” he emphasised. This consistency is particularly important when handling tasks such as generating vector embeddings, as discrepancies can cause significant issues in downstream applications. Handling Latency and Bottlenecks Latency is one of the biggest challenges when working with multimodal pipelines. “Latency can be a big factor, especially if you’re dealing with near real-time applications,” said Chetan. He recommends addressing this through a combination of model tuning, infrastructure optimisation, and data chunking. “In a 10-minute video, you should chunk it into smaller segments and process them in parallel,” Dixit suggested, which helps reduce latency. He also advises using GPUs to accelerate compute-intensive tasks and parallelising the data processing to ensure smoother operations. Handling real-time data adds an extra layer of complexity. In real-time applications, there is no room for errors or bottlenecks. “Your pipelines must have fault tolerance built in,” he noted, especially when dealing with real-time customer interactions. Failover mechanisms, such as exponential backoffs, can help mitigate these risks by retrying API calls that fail during the process. Optimising for Resource Allocation and Cost Efficiency As organisations move from proof-of-concept (POC) to production environments, the cost and infrastructure requirements scale significantly. Dixit recommends adopting a phased approach when deploying multimodal pipelines for production-scale generative AI applications. “Instead of exposing the app to 5,000 users at one go, start with 1,000 users and scale up gradually,” he advised. This approach allows companies to focus on optimising the application without overburdening their infrastructure. A critical part of this process is understanding the infrastructure requirements of deep learning models and vector databases. Enterprises need to plan for how much workload your deep learning models can handle and then ensure your vector database can scale horizontally based on that workload. Monitoring and Observability According to Chetan, monitoring and observability are the most important factors in ensuring smooth operations, as they allow for proactive issue detection and resolution. Chetan emphasised that it’s vital for businesses to catch problems before end-users notice them. “Your end-user should not be the one telling you that your application is broken,” he noted. Monitoring systems help identify bottlenecks and failures in the pipeline by tracking key metrics. This proactive approach enables organisations to fix issues before they impact user experience. To achieve this, Chetan outlines a three-step process: logging, building metrics, and setting up monitoring tools. “You have to log the information first, then build metrics for both infrastructure and functionality,” he explained. By tracking these metrics, organisations can receive alerts when certain thresholds, such as high CPU or memory usage, or a spike in errors, are reached. Effective logging and monitoring also provide traceability, making it easier for site reliability engineers and support teams to diagnose and fix issues. Traceability allows engineers to locate where failures occurred and what caused them, reducing the overall time to fix them. Without proper logging and observability, organizations risk wasting significant time debugging issues, especially in real-time data processing. A robust monitoring framework is essential for ensuring optimal performance and quick recovery from any issues in production-scale applications. As Chetan aptly put it, “It’s all about continuously tuning and optimizing your models, pipelines, and infrastructure to meet the demands of real-world applications.” This proactive approach is key to maintaining efficiency and resilience in today’s data-driven landscape.","excerpt":"One significant hurdle is managing multiple data formats.","categories":["AI Highlights"],"tags":["AI (Artificial Intelligence)"],"author_name":"Mohit Pandey","publish_date":"2024-10-31T11:00:00","publication_year":"2024","word_count":1041,"keywords":["Go","API","AI","Scala","vector databases","Aim","deep learning","data engineering","generative AI","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","deep learning","generative AI","Aim","vector databases","R","Go","Scala","API","data engineering"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/setting-up-multimodal-pipelines-and-vector-databases-for-production-scale\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10076304,"title":"After JioPhone, Reliance Unveils Low-Cost Laptop ‘JioBook’ at ₹15,000","content":"Mukesh Ambani-led Reliance Jio is set to unveil its cost-friendly laptop priced at $184 (INR 15,000) with an embedded 4G sim card. Reportedly, the move is based on replicating its success with the low-cost JioPhone in the highly price-sensitive market in India. Reliance Jio has partnered with major tech giants such as Qualcomm and Microsoft for the new ‘JioBook’—powered by its computing chips from Arm Ltd—that will run on dual boot support of JioOS and Windows OS. The low-cost device is expected to launch in October 2022. The laptop will be made available to schools and government institutes from this month onwards—with an anticipated consumer launch within the next three months. The telecom giant also contributed to the digital wave in India with the launch of a budget 4G phone for the masses. Several claim that “this will be as big as JioPhone”. Priced at a value under INR 10,000, the JioPhone is among India’s top-selling phones. By launching the ‘JioBook’, Reliance Jio aims to target Indian consumers who wish to have a laptop at a low price. JioBook’s 5G variant is expected to come following the launch of the Jio 5G phone, which is set to arrive in 2023. Sources claim that the JioBook will be produced locally in India by the manufacturer Flex and that the conglomerate aims to sell “hundreds of thousands” of units by March 2023. In 2020, Jio raised around $22 billion from global investors, such as KKR & Co Inc (KKR.N) and Silver Lake, to introduce the affordable laptop. The device is also being pitched as an alternative to tablets.","excerpt":"Reliance Jio has partnered with major tech giants such as Qualcomm and Microsoft for the new ‘JioBook’ – powered by its computing chips from Arm Ltd – that will run on dual boot support of JioOS and Windows OS.","categories":["AI News"],"tags":["Jio AI Cloud"],"author_name":"Bhuvana Kamath","publish_date":"2022-10-03T15:59:15","publication_year":"2022","word_count":266,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","Git","Aim","Jio AI Cloud","R"],"extracted_tech_keywords":["AI","Aim","R","Go","Git","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/after-jiophone-reliance-unveils-low-cost-laptop-jiobook-at-%e2%82%b915000\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":32964,"title":"SEBI Issues Directive To Disclose All AI-Based Tools Used In Stock Market","content":"Putting greater emphasis on the reported black box method of trading in India, Securities And Exchange Board of India (SEBI) announced that applications based on artificial intelligence and machine learning would have to make quarterly disclosures on their compliance with the cybersecurity framework. In a circular sent out by SEBI, the stock market regulator said, “It is imperative to ensure that any advertised financial benefit owing to these technologies for products offered by intermediaries should not constitute to misrepresentation.” “With effect from the quarter ending March 2019, registered stock brokers or depository participant using AI and ML-based application or system are required to fill in form and make submissions on a quarterly basis within 15 calendar days of the expiry of the quarter,” added the circular. SEBI is also conducting a survey and creating an inventory of the AI\/ML landscape in the Indian financial markets to gain an in-depth understanding of the adoption of such technologies in the markets and to ensure preparedness for any AI\/ML policies that may arise in the future. The market watchdog also added that there is a sharp increase in the usage of AI and ML as product offerings by market intermediaries and participants called “robo advisors”, in investor and consumer-facing products.","excerpt":"Putting greater emphasis on the reported black box method of trading in India, Securities And Exchange Board of India (SEBI) announced that applications based on artificial intelligence and machine learning would have to make quarterly disclosures on their compliance with the cybersecurity framework. In a circular sent out by SEBI, the stock market regulator said, […]","categories":["AI News"],"tags":["AI (Artificial Intelligence)","sebi","stock market"],"author_name":"Prajakta Hebbar","publish_date":"2019-01-07T13:10:14","publication_year":"2019","word_count":207,"keywords":["artificial intelligence","machine learning","programming_languages:R","AI","sebi","ML","stock market","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","R","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/sebi-issues-directive-to-disclose-all-ai-based-tools-used-in-stock-market\/","complexity_score":2,"technical_depth":6,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10164116,"title":"Hexaware Debuts on NSE, Raises ₹8,750 crore","content":"Hexaware Technologies, a global tech and business services company, debuted on the Indian stock market, raising ₹8,750 crore. This is the largest-ever public issue in the Indian IT services sector. At close, the company’s shares rose nearly 8%. The company’s CEO, R Srikrishna, called the listing a major step forward, emphasising transparency, accountability, and client-focused growth. According to the company, Hexaware specialises in AI-first solutions and has over 31,000 employees in 54 offices in 28 countries. “This is an opportunity to deepen our relationships with stakeholders and reinforce our commitment to operating with transparency, accountability, and a focus on delivering meaningful solutions to our clients. We are excited about the road ahead,” Srikrishna said. With offices worldwide, the company claims to help enterprises achieve fast and scalable transformation by optimising their systems.  It serves clients across the banking, financial services, capital markets, healthcare, insurance, manufacturing, retail, education, telecom, high-tech & professional services, travel, transportation, and logistics sectors. With this listing, Hexaware signals its ambition to accelerate growth and deliver greater value on a global scale. CFO Vikash Kumar Jain also highlighted the increased market visibility and investor engagement the move enables. Meanwhile, Kapil Modi, managing director at Carlyle India Advisors, Hexaware’s parent company, said, “Carlyle remains committed to partnering with the Hexaware team as it continues to focus on client centricity and delivering differentiated value for its customers and as it embarks on its next phase as a publicly traded company.” The Indian IPO market is witnessing a significant upswing, with several companies preparing to go public in 2025. The lineup includes Swiggy, Ather Energy, Reliance Jio, OfBusiness, and AI startup Fractal. The market’s growth trajectory in 2024, which saw a 149% increase in IPO value to $18.4 billion, has set a promising stage for the upcoming year.","excerpt":"Hexaware specialises in AI-first solutions and has over 31,000 employees across 54 offices in 28 countries.","categories":["AI News"],"tags":["Indian IT"],"author_name":"Aditi Suresh","publish_date":"2025-02-19T15:49:07","publication_year":"2025","word_count":298,"keywords":["Go","API","AI-first","programming_languages:R","AI","IPO","Scala","Aim","Indian IT","R","startup"],"extracted_tech_keywords":["AI","Aim","R","Go","Scala","API","AI-first","startup","IPO","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/hexaware-debuts-on-nse-raises-%e2%82%b98750-crore\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10047461,"title":"Don’t Chase Trends In AI, Create Strong Knowledge Foundation Instead: Jaswinder Chadha, Axtria","content":"“In the future, every business must run as a digital business, and data and software are the foundations,” believes Jaswinder Chadha, the CEO and president of Axtria, which offers analytics and software solutions to global companies to effectively manage and derive insights from big data. A data scientist at heart, Chadha strongly believes that the decision accuracy and velocity of a company’s growth driven by data will be the biggest competitive advantage in the future. Keeping in view the same, Chadha recently endowed a Chair in data analytics at his alma mater IIT Delhi. In 2008, Chadha had also endowed a Chair in Operations Research at IIT Delhi and is the founder of the IIT Delhi Endowment Fund. Analytics India Magazine caught up with Chadha to understand the trends and future of data analytics, his association with IIT Delhi and Axtria. Excerpts from the interview: AIM: As an industry veteran, you have donned many hats. Walk us through your professional journey. Jaswinder Chadha: I grew up in Punjab, India, and studied B Tech in Mechanical Engineering at IIT Delhi. In 1990, I decided to pursue post-graduate studies in the US. I attended The University of Texas at El Paso for MS and eventually pursued my PhD in Industrial Engineering and Operations Research at Texas A&M. Every entrepreneur’s journey starts somewhere, and looking back, my aspiration to create my own company was fueled by my experiences at IIT Delhi. In 2000, I co-founded marketRx, a sales and marketing analytics solutions provider to the pharmaceutical industry. When Cognizant acquired the company, it was one of the most significant acquisitions in the space. After the acquisition, I started the data analytics practice at Cognizant for the next two years. In 2010, I co-founded Axtria with the vision to transform the commercial function with data analytics and cloud-based software platforms. Today, Axtria is one of the industry’s fastest-growing software and data analytics companies. We help life sciences companies transform their patient and business outcomes. And, while Axtria has achieved a lot, I cannot emphasise enough that we have only scratched the surface. AIM: Technology is emerging at lightning speed now. Which fields do excite you the most? Jaswinder Chadha: I am a data scientist, so I am interested in the ongoing evolution of data science and how it can be leveraged to extract business-focused insights to propel life sciences companies forward. As you can imagine, data scientists are curious. I am always interested in learning about the rapid growth of the data science field. As computational technology continues to grow at an unprecedented pace, data analysis capabilities will also increase. From this, we will see modernised, faster data analysis algorithms and methods. Data scientists also use other technologies I am keenly interested in — like artificial intelligence (AI) and machine learning (ML). AI\/ML is already having a positive impact on business operations and healthcare outcomes today and will be a significant trend in the near and long term. AIM: What drove you to the field of analytics? Jaswinder Chadha: In the life sciences industry today, there’s a tremendous opportunity and an unmet need in the marketplace. With the computational capability of the cloud and big data technologies, the ability to manage and consume data of volume and variety has increased tremendously. There had never been such an opportunity to optimise commercial decision-making with data-driven insights at speed, sometimes even real-time. Everyone talked about AI\/ML, Big Data, Cloud, and omnichannel as the new buzzwords, but very few knew what to do with them. Despite the “wind in the sails” provided by these, there was a real gap in the ability to deliver on the vision of these trends. These trends will be the reality of tomorrow, and many are already. The answer was to provide a combination of analytical people, process, and data science technology, with deep domain experience, technical skills, and most importantly – an understanding of the industry data itself. The approach to analytics and technology is different. Digital transformation without clean and pliable data is too inefficient. Analytics based on clean data will deliver the true transformation to achieve companies’ visions fast. Software technology will provide an unprecedented scale. In a way, analytics is the oil needed for data-driven digital transformation, and our vision of yesterday is the reality of today. AIM: How is Axtria creating value for its customers using data analytics? Jaswinder Chadha: Axtria is a global provider of award-winning cloud software and data analytics to the life sciences industry. Axtria’s solutions digitally transform the entire product commercialisation process, drive sales growth, and improve healthcare outcomes for patients. Our focus is to deliver solutions that help customers complete the journey from Data-to-Insights-to-Action and get superior ROI. Axtria offers a robust suite of AI-powered platforms – Axtria DataMAx™ for data management, Axtria InsightsMAx™ for analytics, and Axtria SalesIQ™ and Axtria CustomerIQ™ for omnichannel commercial operations. Our portfolio of customers provides therapies to approximately 60 per cent of cancer patients globally and produces 85 per cent of the world’s vaccines. Axtrians are genuinely proud of our contribution and work in bringing these lifesaving therapies to patients globally. Axtria’s cloud platforms are modular and built on Force.com and AWS while utilising modern technology such as AI\/ML, NLP, and advanced analytics. We have made our offerings to remove the clutter in commercial operations so companies can do what they do best – service their customers to deliver on-time patient care and manage their sales and marketing operations. AIM: What are Axtria’s major future plans? Jaswinder Chadha: With the life sciences and healthcare industries undergoing a significant acceleration in data-driven digital transformation, there is a solid push to effectively leverage software technology and AI to scale global commercial operations effectively. More than ever, companies are looking to replace archaic systems with digital data enablement, analytics at scale, and omnichannel software to deliver an improved customer experience. Serving this impending need, Axtria offers a robust suite of AI-powered platforms. We have rapidly increased our investments in product development over the past couple of years. These investments were our biggest bets yet, with a potential to disrupt and even jeopardise our core business. But thankfully, our strategy worked well. It is evident in the success of Axtria SalesIQ™, which is now considered the most comprehensive and innovative product in its category. Axtria recently received funding, accelerating our investments in enterprise software products, with close to $100 million planned to build four category-leading products in the next five years. Additionally, this investment will help us rapidly expand into new accounts and scale our top customers. AIM: What is your advice for people starting their careers in AI and analytics? Jaswinder Chadha: When deciding what path you want to take, think about what kinds of problems you are interested in solving and what type of work you like doing (programming, theorising, working with real-world data, etc.). This will help you focus on the area you are most interested in. Don’t put all your eggs in one basket or chase after the hottest AI\/analytics trends. Keep in mind, the latest trend today will soon be replaced with tomorrow’s new one. This is why it is essential to learn the basics of data analytics and create a strong knowledge foundation. Finally, don’t obsess over having everything exactly how you envision it. You must be flexible and expect your plans to change and evolve continuously. AIM: You helped establish a Chair in data analytics at IIT Delhi. You are often counted among the influential alumnus of the premier. Tell us more about your association with IIT Delhi. Jaswinder Chadha: Yes, I am a proud alumnus of IIT Delhi, and my education at IIT Delhi has shaped my perspective and career. But, the more inspiring story is that of my mother, Dr Tarvinder Kaur Chadha. She was the first woman to get a PhD in Mathematics from IIT Delhi in 1968. When I graduated in 1990, it made her and me the first parent-child graduates of IIT Delhi. Our family history with IIT Delhi is one that we are most proud of, and it brings me joy to give back to such an extraordinary institution. In 2008, I endowed a Chair in Operations Research. I am also a founder of the IIT Delhi Endowment Fund, and most recently, I endowed a Chair in Data Analytics. It is a great honour to have the data analytics chair named after my mother and me, “Jaswinder and Tarvinder Chadha Chair in Data Analytics.” IIT Delhi awarded me the Distinguished Alumni Service Award in 2008.","excerpt":"Jaswinder Chadha recently endowed a Chair in data analytics at his alma mater IIT Delhi","categories":["AI Features"],"tags":["axtria","Data Analytics","IIT Delhi","Interviews and Discussions"],"author_name":"Shraddha Goled","publish_date":"2021-09-01T16:00:00","publication_year":"2021","word_count":1428,"keywords":["data science","artificial intelligence","machine learning","AWS","AI","axtria","ML","RAG","NLP","Aim","analytics","Data Analytics","IIT Delhi","Interviews and Discussions"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","NLP","data science","analytics","Aim","RAG","AWS"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/dont-chase-trends-in-ai-create-strong-knowledge-foundation-instead-jaswinder-chadha-axtria\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10105334,"title":"Worldcoin In, Captcha Out","content":"In August, when the value of Worldcoin tokens (WLD) sank to its lowest, dropping by 46% of its launch value, the fate of Worldcoin looked questionable. With no exact use cases appearing from this Sam Altman-backed venture, the second version, World ID 2.0 was released last week. With new integration methods and advanced use cases across sectors, the announcement of World ID 2.0 pushed the price of WLD to two times its value, touching as high as $4.23. And, Worldcoin now looks to address online frauds. Worldcoin Price Chart. Source: Coingecko Solution to Multi Sector Frauds As per a recent report, retailers are losing $100 billion annually owing to bots, coupon stacking and return frauds. With unique identification methods that are being offered by Worldcoin, frauds related to duplicate accounts can be eliminated. Yesterday, Worldcoin released a blog on the results of Worldcoin’s first anti-fraud retail integration with Shopify. The company reiterates that there would be improved loyalty defence mechanisms and control over the maximum number of accounts per person, which will help towards fraud mitigation. Interestingly, not just in retail, other sectors will also benefit from the same. With Minecraft also onboarded in the latest version of Worldcoin, frauds associated with gaming can be minimised. With a digital passport verification system, a fair-gaming experience can be obtained. Automated bots, and bad actor resistance for in-game marketplaces, can fortify security. Additionally, proof of personhood allows the protection against unlawful account sharing which is a potential way for hackers to gain access to multiple accounts. These measures aides towards building an equitable gaming ecosystem. Finance sector is not far behind either. With Worldcoin, any form of verification and re-verification can be safely done via a digital passport. When the onboarding process in banks and financial institutions went digital, the scope of frauds for misrepresentation and fake verification also opened up, something which even Zerodha co-founder Nikhil Kamath recently spoke about. With World ID 2.0, these deepfake scams can be countered to a certain extent. Interestingly, those verified via Worldcoin on Reddit would now get a verified status, thereby, addressing the problem of spams. App Verification Via Passport World ID, a digital passport, seeks to validate unique human identities online while prioritising privacy. As an open protocol for collective ownership, World ID 2.0 has introduced significant upgrades, including enhanced privacy controls, apps, levels, and a developer portal, which aims to provide users with improved security and functionality. With a number of app integration on World ID 2.0, including Reddit, Shopify, Discord and others, user accounts on all these platforms can be verified with the help of this digital passport. The second generation introduces new protocol levels such as Orb+ with face identification, enabling users to exclusively utilise it for specific actions. Furthermore , a user can also build their own application based on the apps available in the Worldcoin App store. Captcha, No More? Captcha (Completely Automated Public Turing test), a type of challenge-response test to differentiate between a human user and bots, is slowly losing relevance. With AI, able to solve captchas with accuracies 15% higher than that of humans. Furthermore, as these codes become more complex, genuine users find it increasingly difficult to solve them, thereby defeating the purpose of having captchas in the first place. With Worldcoin ID, a user can easily verify themselves without any tedious or time-consuming methods such as captchas. Furthermore, once verified, a user will not be subjected to re-verification. Interestingly, big tech companies are also coming up with ways to avoid captchas in order to provide a seamless experience for their users. Scepticism Still Continues While Worldcoin’s proof of personhood continues sign ups across the globe, the controversies that marred the company a few months ago, still continues in certain regions. Owing to security concerns and even mismanagement of people, certain countries including India had banned physical Orb verification processes. However, for this to come to fruition, the entire ecosystem should allow World ID verification as well. With a list of companies from sectors ranging from gaming, ecommerce, finance, and others already allowing World ID verification, more companies can be expected to join the wagon, which can probably address the problem.","excerpt":"With apps such as Shopify, Minecraft, Reddit, and others allowing World ID verification, Worldcoin is emerging as a legitimate authentication method","categories":["AI Trends"],"tags":["minecraft","orb","reddit","Sam Altman","Shopify","WLD","WorldCoin"],"author_name":"Vandana Nair","publish_date":"2023-12-21T18:30:00","publication_year":"2023","word_count":697,"keywords":["Go","Sam Altman","programming_languages:R","AI","WorldCoin","reddit","ML","Shopify","minecraft","Git","orb","programming_languages:Go","WLD","Aim","R"],"extracted_tech_keywords":["AI","ML","Aim","R","Go","Git","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/worldcoin-in-captcha-out\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10054082,"title":"Apollo Tyres Collaborates With AWS To Make Factories Smarter","content":"Amazon Web Services (AWS) announced that Apollo Tyres is going all-in on AWS’ services to transform digitally. By moving all of its IT infrastructures to AWS, Apollo Tyres can use AWS’s broad portfolio of services to innovate new customer experiences while driving productivity, compliance, and process efficiency gains globally across seven factories. Apollo Tyres will draw on the breadth and depth of AWS capabilities, including Internet of Things (IoT), data and analytics, and machine learning, to transform into an agile, data-driven enterprise. Using data from the factory floor and real-time information from production machines, like tyre rubber mixer machines, Apollo Tyres can expand operational intelligence capabilities and more accurately manage machine utilisation, ensuring high-quality levels and machine efficiency. With AWS, Apollo Tyres is connecting all of its factories in the cloud this year in India and Europe. Image Source: AWS By 2022, Apollo Tyres plans to migrate all mission-critical enterprise applications, including its SAP applications, to AWS to enhance customer experience, improve process efficiency, and enable process automation. Apollo Tyres produces more than 2,425 tons (2,200 metric tons) of tyres daily in its seven factories worldwide. Each factory previously ran its on-premises infrastructure in silos, which provided limited visibility into global manufacturing efficiencies. Apollo Tyres needed to upgrade its infrastructure to develop new ways of engaging with fleet operators, tyre dealers, and consumers while delivering tyres and services efficiently at competitive prices. Image Source: AWS The company’s first step was to create a data lake on AWS, which centrally stores Apollo Tyres’ structured and unstructured data at scale. This data lake provides the foundation for an integrated data platform, which enables Apollo Tyres’ engineers around the world to collaborate in developing cloud-native applications and improve enterprise-wide decision making. The integrated data platform enables Apollo Tyres to innovate new products and services, including energy-efficient tires and remote warranty fulfilment. Using AWS IoT SiteWise, a managed service that makes it easy to collect, store, organise and monitor data from industrial equipment at scale, and AWS IoT Greengrass, an open-source edge runtime and cloud service for building, deploying, and managing device software, Apollo Tyres developed an IoT-in-a-box solution. The solution connects production machines on the factory floor to AWS in as few as five days. Once connected, the solution captures data from multiple machines—including mixers, tyre building equipment, and curing presses—and feeds it to the data lake. Apollo Tyres uses Amazon Redshift, a cloud data warehouse, to create a global dashboard for visualising production information from the data lake, providing business teams and plant managers with real-time visibility into the manufacturing process. This visibility improves production efficiency and productivity, for example, by reducing the idle time of curing presses that shape the tyre in a mould by 50 percent. Image Source: AWS “By digitally transforming with AWS, we can unlock productivity and efficiency gains in our factories globally, innovate new products and services faster, and enhance customer experience. We are using AWS capabilities, like IoT and machine learning services, to connect our factories and make them smarter. This fosters collaboration between our IT and business teams to make the production process more efficient while delivering higher quality products at a lower cost.” said Hizmy Hassen, Chief Digital Officer, Apollo Tyres. Vaishali Kasture, Head of Enterprise, Mid-Market, and Global Businesses, AWS India and South Asia, Amazon Internet Services Private Limited (AISPL), said “Apollo Tyres is using the cloud to digitally transform, improve the tyre manufacturing process, and deliver better value to customers. By moving its entire infrastructure to AWS, Apollo Tyres creates an environment of rapid and continuous innovation to provide safer and better quality tyres and enhanced customer service experiences.” Apollo Tyres also launched an automated tyre inspection program that checks for tyre defects using photos of the tyres taken as they progress along the production line. Based on Amazon Rekognition, a machine learning service that automates image and video analysis, this automation allows factory supervisors to intervene when manufacturing anomalies occur, providing customers with high-quality tyres that meet strict safety standards. Apollo Tyres is building new digital products using Amazon Elastic Kubernetes Service (Amazon EKS), which gives AWS customers the flexibility to start, run, and scale Kubernetes applications in AWS or on-premises microservices that support any application architecture, regardless of scale, load, or complexity.","excerpt":"Apollo Tyres will draw on the breadth and depth of AWS capabilities, including Internet of Things (IoT), data and analytics, and machine learning, to transform into an agile, data-driven enterprise.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","AWS","AWS services","Data Science","Data Scientist","Deep Learning","Deep Learning Techniques","IoT","iot enterprise architecture","IoT India","Machine Learning"],"author_name":"Victor Dey","publish_date":"2021-11-23T18:57:29","publication_year":"2021","word_count":712,"keywords":["API","AWS services","iot enterprise architecture","Git","R","kubernetes","IoT India","analytics","Data Science","Go","machine learning","AWS","AI","Machine Learning","IoT","microservices","Deep Learning Techniques","Deep Learning","Data Scientist","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","machine learning","analytics","AWS","kubernetes","microservices","R","Go","Git","API"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/apollo-tyres-collaborates-with-aws-to-make-factories-smarter\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10141002,"title":"Francois Chollet, The Man Behind Keras, Quits Google","content":"Francois Chollet, a leading AI researcher at Google and the creator of Keras, the deep-learning framework, announced his exit from the company. Chollet took to X to share that he would continue to advise Google. Chollet’s departure from Google marks a significant development. Other notable exits this year include Julian Schrittwieser, a leading AI researcher who left Google DeepMind to join Anthropic. Additionally, Shailesh Prakash, vice president of Google News, stepped down over growing AI and news traffic disputes. Last year, Geoffrey Hinton, often referred to as the godfather of AI, also parted ways with Google. “Today, we’re announcing that Francois Chollet, the creator of Keras and a leading figure in the AI world, is embarking on a new chapter in his career outside of Google. Keras is more than just a tool—it’s a community driving AI innovation,” Google posted on its official blog. “As Francais Chollet begins a new chapter, our commitment to Keras and open-source AI remains strong.” How Keras has touched millions of lives Google noted that Keras has over two million users and has become essential in AI development. It simplifies workflows and makes advanced technology widely accessible. It supports applications like Waymo, YouTube, Netflix, and Spotify. Chollet was dedicated to Keras, ensuring its support for JAX, TensorFlow, and PyTorch, while continuing to guide its development. Google’s commitment to Keras 3 highlights its support for major ML frameworks, offering developers flexibility. The launch of Keras Hub further democratises access to AI tools, speeding up the creation of innovative multimodal applications. Chollet’s Arc Evaluation Chollet introduced the ARC-AGI evaluation in 2019. His research paper ‘On the Measure of Intelligence’, aimed to fairly measure human-like intelligence in AI systems. His Abstraction and Reasoning Corpus (ARC-AGI) remains the only formal benchmark for AGI. He hypothesised that it would be difficult to beat, and it still remains unbeaten. Recently on X, Sam Altman hinted that they might have achieved this benchmark internally, however, Chollet disregarded this claim as premature. Earlier this year, he introduced the ARC Prize 2024, offering a $1.1M prize pool to advance AI innovation. In the first ARC-AGI competition in 2020, the winning team achieved only a 21% success rate. In 2022, Chollet and Lab42 launched ARCathon 2022, with 118 teams competing globally. In ARCathon 2023, 265+ teams from 65 countries competed, with two teams tying at a 30% success rate.","excerpt":"Chollet introduced the ARC-AGI evaluation in 2019. This remains the only formal benchmark for AGI.","categories":["AI News"],"tags":["AI (Artificial Intelligence)"],"author_name":"Aditi Suresh","publish_date":"2024-11-15T14:32:36","publication_year":"2024","word_count":394,"keywords":["Anthropic","Go","Keras","AI","PyTorch","ML","Aim","JAX","TensorFlow","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","ML","Anthropic","Aim","TensorFlow","PyTorch","JAX","Keras","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/francois-chollet-the-man-behind-keras-quits-google\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":68851,"title":"How Will Various Enterprises Be Impacted By India&#8217;s Global AI Alliance","content":"India recently entered the global alliance for AI (GPAI) with countries such as the USA, UK, Germany, Australia and others to ensure that AI is used responsibly. Bringing the industry, government and academia together, it will oversee the evolution of AI and develop methodologies to show how AI can be leveraged to better react to global issues. A first of its kind initiative, GPAI aims to better understand the challenges and opportunities around AI while leveraging the experience and diversity of participating countries. It is essentially aimed at closing the gap between theory and practice on AI by leveraging cutting-edge research and applied action on AI-related objectives. With four working groups on responsible AI, the future of work, data governance, and commercialisation and innovation, it is going to transform the way a lot of industries work. “With this GPAI platform, India leverages the learnings of the various other member countries in this forum and learns from their adaptation of AI technologies for social upliftment. India’s own National AI Strategy and Portal that has been recently launched will also hugely benefit from this global platform and further our progress in the effort of empowering the citizens of this country and bringing about truly inclusive growth,” says Prashanth GJ, CEO at TechnoBind. How Will It Impact Indian Enterprises India has been very forthcoming when it comes to implementing and boosting AI integrated developments. With its National AI strategy and the recent launch of AI portal, it aims to bring the impact of technologies across various industries such as agriculture, healthcare, education, finance and more. Now, being one of the founding members of the GPAI, it is therefore expected to play a crucial role in expanding the responsible utilisation of artificial intelligence for the world. Various industries in India see it as welcome news and experts believe that it will raise awareness about AI’s ability to transform business in the country. While there has been tremendous growth and research around artificial intelligence, GPAI will further help in closing the gap between AI being a theoretical buzzword and actually witnessing its application in major sectors. Given the large population and challenges therewith, India will now be in a good position to leverage benefits that AI can bring to the multitude of problems that India faces right now. It will help in dealing with a host of problems from education to healthcare and social security as the country will get a major boost in its AI developments. Privacy and security In recent times, privacy and security are one of the major challenges that companies are facing. Collaborating with global counterparts will help in coming up with AI-based solutions that address privacy and security issues. Especially with the ongoing pandemic, cyberthreats have become more prominent than ever before. It has called for AI-based solutions to ensure higher effectiveness in areas such as cybersecurity. Automation industry The pandemic has also brought about a boost in the automation industry. It has accelerated the digital shift with ML, AI and RPA being extensively used to optimise cost by automation and scale businesses without incurring additional costs. “The last few months have demonstrated that digitization and technology are the only survival options for businesses across size and segment. This initiative will drive impetus with collective wisdom across bodies and geographies for enhancing the quality and safety of life in a world which is challenged by privacy and security concerns,” says S Sriram, Chief Strategy Officer at iValue InfoSolutions. It has also called for many companies to enforce strict work from home policies while others to work on skeleton staff. “In this downsized enterprise framework, AI has a vital role to play. For example, in a large enterprise with service delivery teams, AI can help automate processes and workflows. It can also bring about network automation to quickly analyse, predict and prevent potential failures before they occur, “ says Satish Kumar V, CEO at EverestIMS Technologies. IT and infrastructure It can also bring about better and modern infrastructure with AI-enabled IT solutions to drive faster, better, deeper and new insights enabling them to attain operational efficiencies, transform decision making and drive business growth, believes Gurpreet Singh, Managing Director at Arrow PC Network Pvt Ltd. Recruitment industry Even talent recruitment space isn’t left behind. Artificial intelligence uses algorithms to quickly parse, analyse, and spot patterns in massive data sets. “AI-based talent intelligence platforms can help in scouting the right talent for any organisation, and also bring greater inclusivity and diversity for a balanced workplace. Artificial intelligence can help employers become more efficient, allowing them to focus on building and communicating their employment brands in order to recruit top candidates”, says Sandesh Goel, Managing Director at Eightfold India. Healthcare, especially Covid-19 This global partnership is also being looked forward as an opportunity to use AI for better response and recovery from Covid-19. While many startups and companies are integrating AI to get India going in these tough times of global pandemic, there is certainly a lot that still needs to be done to improve the condition further. Wrapping Up Given India’s keen dedication to set up National AI Strategy, National AI portal, and drafting AI policy, it is already leading in building an ecosystem that focuses on a fast shift from the research and development phase to the deployment and operation phase. With GPAI, it will further boost collaboration with other countries to bring about immense transformation in Indian industries.","excerpt":"India recently entered the global alliance for AI (GPAI) with countries such as the USA, UK, Germany, Australia and others to ensure that AI is used responsibly. Bringing the industry, government and academia together, it will oversee the evolution of AI and develop methodologies to show how AI can be leveraged to better react to […]","categories":["AI Features"],"tags":["intel global strategy","which countries have good cybersecurity"],"author_name":"Srishti Deoras","publish_date":"2020-07-03T14:00:54","publication_year":"2020","word_count":907,"keywords":["Go","artificial intelligence","AI","ML","which countries have good cybersecurity","intel global strategy","Git","RAG","BERT","Aim","data governance","R"],"extracted_tech_keywords":["AI","artificial intelligence","ML","Aim","RAG","R","Go","Git","data governance","BERT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-will-various-enterprises-be-impacted-by-indias-global-ai-alliance\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":16965,"title":"The Rising Demand of AI Experts: TimesPro Collaborates with NVIDIA to Bring One-Of-A-Kind Course","content":"We believe AI is a new scientific infrastructure for research and learning that professionals will need to embrace and lead, failing which, they will become irrelevant and eventually redundant. When applied in science, AI can autonomously create hypotheses, find unanticipated connections, and reduce the cost of gaining insights and the ability to be predictive. The human mind processes millions of sensory inputs automatically and constantly, making it the most elegant computer in existence. But the human brain only contains about 300 million pattern processors that are responsible for human thought. What if we could bring to existence all the ideas not just with data, but also orders of magnitude more data processing capability? In the name of AI, what we have now is narrow AI. But the pace at which we are moving towards general AI, which would outperform humans at nearly every cognitive task, cannot be ignored. However, despite being at a nascent stage, AI has already enough contribution in the field of banking, medicine and security among others. And while experts continue to work to make AI more advanced, one of the biggest challenges they are facing is the lack of skilled manpower. Why Is There A Need To Upgrade Your Skillset? In 2017, the global AI market is expected to be worth approximately $2.42 billion, which will grow to be worth $18 billion by 2022. Globally, investors backed more AI companies in the first quarter of 2016 than in any other quarter, while more than 200 AI-focused global companies have collectively raised more than $1.5 billion so far, this year. While there has been enough clout around the rising demand and need for data scientists, there has been several institutions that has introduced courses to adept professionals with data analytics. However, what still remains unexplored are professional courses designed around AI, which needs a completely different skillset. With very few institutions venturing into the AI arena, interestingly, the Times of India Group understood the need for it and designed a professional course to help candidates understand and train in AI. TimesPro, in technical partnership with NVIDIA- the AI computing company, has introduced three comprehensive courses to enable professionals to navigate the world of decision science. The three programs are designed in a way that it can help professionals from all levels to upgrade themselves. Regardless you are someone who is an entry level manager or belong to the mid-senior or senior level, the programs equip participants with skills and knowledge to influence decision making, strategy. The programs have been planned in a way that professionals can enrol without having to quit from their present jobs. Moreover, TimesPro also provides placement assistance to candidates, helping them find better opportunities soon after the course. It also focuses on the understanding of latest analytical concepts like descriptive, predictive and prescriptive analytics to solve business problems. A Brief Look at The Programs: PG Program in data science (PGDDS): This program is specially designed for entry and middle level managers with minimum 2 years of work experience (desired experience is 4-5 years), who seek career advancement via specialised training in Business Analytics\/Data Sciences applied to their field PG Program in advanced data science (PGDADS): The second program is for entry, middle, senior level managers with minimum 5 years of relevant work experience. The course is apt for someone who is seeking career advancement via specialised training in advance data sciences applied to their field Certificate in advance decision science (CADS): The third and the last course is designed for working professionals with relevant experience in data analytics and big data experience, who apply quantitative techniques to make effective decisions. Why Should You Opt for This Program? Besides the fact that the course enables you to upskill yourself and makes you a sought-after employee, the best part is the course has a technical tie-up with NVIDIA, a company, which is synonymous with artificial intelligence. Moreover, its GPU- accelerated computing with a CPU is used to accelerate deep learning, analytics, and engineering applications. Pioneered in 2007 by NVIDIA, GPU accelerators now power energy-efficient data centres in government labs, universities, enterprises, and small-and-medium businesses around the world. While the chip-maker is likely better-known for producing high-end graphics to power pixel-pushing gameplay, it also started providing the hardware and software that could help bring about the future of driving. NVIDIA has already forged self-driving alliances with automakers Audi, Toyota and Volvo and has a partnership with Chinese internet giant Baidu to bring artificial intelligence to cloud computing, self-driving vehicles and AI home assistants. Why TimesPro? The Times Group has evolved into a multi-media conglomerate focused on providing solutions to consumers across the nation as well as creating a better tomorrow for the Indian youth. Anish Srikrishna, President, Times Professional Learning (TPL), said, “We are delighted to be associated with NVIDIA. TPL aims to provide quality education, training and skills with strong industry connect and this program with NVIDIA reflects that deep interest and commitment. We believe relationships such as these will be truly beneficial to the industry, ensuring first-day-first-hour productivity.” Our Review: The course has been designed keeping in mind the fast changing demands and requirements of the industry. All industries are now beginning to understand the value and need for Machine learning, which encourages a more profound democratization of intelligence, although this is true only for low-order knowledge. All the three programs are divided into modules, with special focus on all important aspects of AI and machine learning. And the biggest USP is that they are designed in a way that applicants will not have to quit their jobs or even take time off to apply for the course. The course duration may range between a minimum of 100 hours and a maximum of 380 hours, depending on the program. Moreover, unlike most courses, this is a little easy on the pocket and will set you back by a maximum of Rs 2,50,000 depending on the program opted for. But over and above the up to date curriculum, TimesPro also provides the most vital features: Expert Industry Professionals as faculty State-of-the-art Learning Centres Holistic Training – Practical’s and Theory 100% instructor-led classroom training Placement assistance Last but not the least, the name NVIDIA and Times of India group should be enough to validate the course for its authenticity and effectiveness. For more details, visit here. Download the brochure [attachments include=”16970″]","excerpt":"We believe AI is a new scientific infrastructure for research and learning that professionals will need to embrace and lead, failing which, they will become irrelevant and eventually redundant. When applied in science, AI can autonomously create hypotheses, find unanticipated connections, and reduce the cost of gaining insights and the ability to be predictive. The […]","categories":["AI Trends"],"tags":[],"author_name":"Дарья","publish_date":"2017-08-11T13:17:19","publication_year":"2017","word_count":1065,"keywords":["data science","machine learning","artificial intelligence","AI","cloud computing","RAG","Aim","deep learning","analytics","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","deep learning","data science","analytics","Aim","RAG","cloud computing","R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/rising-demand-ai-experts-timespro-collaborates-nvidia-bring-one-kind-course\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":3958,"title":"Portrait of Analytics Market in 2012 and the Future","content":"This article is the part of Analytics Placement & Salary Report 2013 by Jigsaw Academy and Analytics India Magzine 2012 saw an impressive increase in the demand for Analytics professionals and Data Scientists. The lure of excellent job prospects within the analytics sector saw a rise in the number of students and professionals eager to gain analytic skills. The corporate world as well took big strides integrating analytics within all layers of their business. India was seen to be the preferred outsourcing destination for IT and ITeS services. A highly talented workforce, lower costs and operational efficiencies among others was the reason for this. With SAS reporting that their sales grew 5.4 percent in 2012 to $2.87 billion, they continue to dominate the analytics industry. However, analysts and data scientists are increasingly begining to migrate toward the open source analytical tool R. Analytical tools that integrate Hadoop and R, gained popularity, especially with small and mid-size organisations. Leading Hadoop distribution vendors Cloudera and MapR enjoyed significant revenue growth in 2012, with Cloudera’s revenue growing to $61 million and MapR’s revenue growing to $23 million In India the common domains that Indian companies served in analytics were retail, BFSI and utility (telecom, energy, oil).These companies began demanding data analysis skills from potential management recruits. They emphasized the importance of these skills in the recruitment process itself. Consequently Business Schools realizing the importance of empowering their graduates with these skills have begun to introduce Analytics talent was largly recruited via word of mouth and from well established and reputed colleges like ISI, IIMs, IITs, Symbiosis, VIT Vellore, Ferguson college and St. Stephens. [quote style=”1″]The lure of excellent job prospects within the analytics sector saw a rise in the number of students and professionals eager to gain analytic skills.[\/quote] Data Analytics and Big Data will continue to dominate market talk as data generated grows and data needs evolve. Though the analytics talent pool will get stronger there will still be a substantial shortfall of skilled data scientists. Today business analytics is all about analyzing transactions so we can see trends quantify what’s happened. Tomorrow’s business analytics will be about analyzing interactions so we can truly understand how and why something happened. [quote style=”1″]India’s analytics talent pool will be in high demand because of their process expertise and English language proficiency.[\/quote] Our team of industry experts have put together their predictions for the near future. Here are some of our predictions for the industry: The dominance of SAS is under serious threat from R. In the next 2 to 3 years, R could well approach SAS in terms of market share largely because of theiContinued rapid growth in add-on packages. The attraction of R’s powerful language. The near monopoly R has on the latest analytic methods. It is FREE ! The freedom to teach with real-world examples from outside organizations, which is forbidden to academics by SAS and SPSS licenses Hadoop and MapReduce may continue to be the industry standard tool for big data processing but credible competitors like Berkeley Data Analytics Stack will emerge to gain a sizable share of the market. The last 3 years have seen an online boom in India with thousands of online companies mushrooming all over the country. We predict a huge increase in demand for web analytics as these companies fight to become more competitive. Along with web analytics, social media analytics is another field that is going to witness a shortage of trained professionals. Text analytics has emerged as a field of interest. The field will evolve over the next 2 to 3 years and we anticipate applications of text analytics to expand and cover both inbound and outbound data, on a more regular basis. Several players are also seeking out efficient techniques to analyze content beyond text. Cloud based services like Amazons EMR will become more popular. It will open up BI and analytics to an audience of traditionally non analytical users, allowing them to focus on analyzing data without having to worry about time-consuming set-up and management. India will remain the preferred destination for analytics outsourcing as compared to other Asian countries like Phillipines and China. Unlike the BPOs, analytics (considered a part of KPO) requires skills that are not easily available in these countries. India’s analytics talent pool will be in high demand because of their process expertise and English language proficiency. The fragmented offshore analytics industry will consolidate with a few strong players emerging as market leaders. Indian analytics service providers will be challenged into providing services like model development, consulting and proprietary IP based services.","excerpt":"This article is the part of Analytics Placement & Salary Report 2013 by Jigsaw Academy and Analytics India Magzine 2012 saw an impressive increase in the demand for Analytics professionals and Data Scientists. The lure of excellent job prospects within the analytics sector saw a rise in the number of students and professionals eager to […]","categories":["IT Services"],"tags":["data science salary","data scientist india salary"],"author_name":"Gauravohra","publish_date":"2013-08-05T12:47:34","publication_year":"2013","word_count":764,"keywords":["big data","Go","API","programming_languages:R","AI","data scientist india salary","RAG","ViT","analytics","data science salary","GAN","R"],"extracted_tech_keywords":["AI","analytics","RAG","R","Go","API","big data","GAN","ViT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/it-services\/portrait-of-analytics-market-in-2012-and-the-future\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":42517,"title":"The Brains Behind AI: How Psychology Influenced Transfer Learning","content":"In the real world, data is rarely clean enough to be used out-of-the-box. Collecting and labelling data accurately is a time and labour-intensive task, with many real-world problems not having datasets that are deep enough to apply AI to. Transfer learning has long been quoted as the solution to these scarce datasets. Using the knowledge from solving a problem with a model to solve a similar problem is one of the mainstays of transfer learning. Where does this concept come from? What are the psychological theories behind it? An Introduction To Transfer Learning Transfer learning is a process in machine learning wherein the knowledge gained from solving a specific problem is taken and applied to another, similar problem. This knowledge can take the form of weights in a neural network, which can be extended for use with similar problems. There are two approaches to implement transfer learning in ML, with them being: Develop Your Model: In this approach, a model is trained from scratch on a problem with an abundance of data. This then creates a model that can solve the problem with ease, with the data being included in its weights. Then, this model can be reused to solve a problem similar to the first, except that the second will have a smaller dataset. After tuning the model, a small amount of training should be conducted for the algorithm to learn specific characteristics of the secondary model. This then results in the model being able to solve for the second problem. Use A Pre-Trained Model: The rise of transfer learning has caused the availability of pre-trained models to grow. These models can be picked for a general task such as image processing, NLP, and more, with training for the specific use-case creating accurate weights. Then, this model can be trained on the problem dataset, which then results in the solution being found. The problem dataset can also be sparse, which is why the last few layers of the pre-trained model must be removed to prevent over- or under-fitting. Transfer Of Learning Theories Learning is a fundamental psychological concept and a cognitive process that humans have to execute every day. While there are many different theories to explain learning and intelligence, one holds special importance to the process of transfer learning. Known as the transfer of learning theory, this concept aims to pin down the ideas behind human generalizations in problem-solving. This is the most quoted in instances of learning being transferred from a school setting to a real-life example, showing the necessity of training the human brain. Transfer of learning describes the effect of past knowledge on present circumstances, and how this knowledge is ‘transferred’ to be used in a given situation. There are two main theories under the transfer of learning, namely: Theory Of Identical Elements: This theory states that transfer of learning occurs in situations where the problem is similar to the elements picked up during learning. This is seen in AI when a model trained on a dataset applies its ‘knowledge’ to other problems in the same domain. Theory Of Generalised Experience: This theory looks into a big-picture view of what is derived from experience. In the case of learning an event ‘A’, the concepts behind how it is executed will be stored. Then, the concept itself will be applied to an event ‘B’, which is similar to the event ‘A’ in its domain. There exist factors that govern how well the transfer of learning occurs in any given situation. One of those is crystallized intelligence Crystallized Intelligence & Learning Transfer Crystallized intelligence is part of a bigger theory of general intelligence, first proposed by American psychology Raymond Cattell. According to him, the general cognitive processes of individuals comprised of fluid intelligence along with ‘crystallized’ intelligence. Fluid Intelligence (Reasoning): Fluid intelligence is the cognitive ability of an individual to apply reasoning practices to novel problems. This also includes identifying the patterns in the problem, and how they relate to the eventual solution of the problem. Fluid intelligence is integral to logical reasoning and is used every time a new problem is faced. The reason it is known as ‘fluid’ is that the reasoning changes depending on the solution for the problem at hand. Crystallized Intelligence: Crystallized intelligence is the ability of an individual to apply the knowledge of past circumstances to a current problem. Domain-specific knowledge is accessed and applied to a similar problem. An example of this would be learning how to ride a motorbike after learning a bicycle. The similarity between the processes and concepts required to conduct both of these actions facilitates the transfer of knowledge. Research has found that crystallized intelligence depends on generalised pattern-recognition processes. It depends on ‘schematic forms of reasoning’, wherein a schema is utilised to come to a conclusion. This would imply that crystallized intelligence functions on recognizing the pattern first, then applying a solution. How Psychology Influenced Transfer Learning The concepts of crystallized intelligence and transfer of learning have influenced the field of transfer learning. While the human brain can access long-term memory to execute crystallized intelligence, machines do not have persistent memory across models. This created a roadblock in the implementation of this concept. With the rise of deep learning, transfer learning became easier. For the first time, it was possible to transfer ‘knowledge’ to apply existing data to new problems. Transferring a model’s weights is similar to bringing forward past knowledge using persistent memory; similar to how crystallized intelligence uses memory to apply past knowledge. Transfer learning stands as one of the many psychological concepts applied to machine learning and artificial intelligence, representing the way forward for the field.","excerpt":"In the real world, data is rarely clean enough to be used out-of-the-box. Collecting and labelling data accurately is a time and labour-intensive task, with many real-world problems not having datasets that are deep enough to apply AI to. Transfer learning has long been quoted as the solution to these scarce datasets. Using the knowledge […]","categories":["AI Features"],"tags":["Machine Learning","Transfer Learning"],"author_name":"Anirudh VK","publish_date":"2019-07-15T17:00:12","publication_year":"2019","word_count":944,"keywords":["artificial intelligence","machine learning","AI","neural network","ML","Machine Learning","NLP","Ray","Aim","deep learning","Transfer Learning","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","neural network","NLP","Aim","Ray","R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/the-brains-behind-ai-how-psychology-influenced-transfer-learning\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10043182,"title":"Deep Learning DevCon 2021: The Second Edition Of Virtual Summit Announced","content":"In an effort to bring the deep learning community together, the Association of Data Scientists (ADaSci), the premier global professional body of data science & machine learning professionals, is launching the second edition of Deep Learning DevCon 2021 (#DLDC21). The two-day leading virtual conference, scheduled for September 23 & 24, will give deep learning practitioners a direct line to top machine learning and deep learning innovators and researchers from around the world. The use cases for deep learning have exploded across industries over the last few years. DL enables performing sophisticated computations on large amounts of data and, in many cases, outperformed the classical machine learning models. As a matter of fact, deep learning made chatbots like Alexa & Siri, Google Translate, Netflix recommendation and self-driving cars possible. DLDC 2021 promises to bring together influencers and leading DL professionals on a single platform to shed light on the latest advancements and new approaches in deep learning. DLDC 2.0 is bigger & better DLDC 2021 has become an exclusive hub for machine learning practitioners and deep learning researchers from across the country. In fact, last year’s conference witnessed an exciting lineup of extraordinary speakers, interesting talks and practical hands-on workshops. DLDC 2021 will bring together 30+ speakers with over 50 sessions and workshops spread across two days. With more than 200 organisations participating, DLDC is currently the most buzzing deep learning-focused conference featuring in-depth technical sessions and masterclasses. The conference will also host paper presentations, exhibitions and hackathons. The sessions will unpack the processes involved in building deep learning models and their real-life use cases. The attendees of the full-day workshop will be provided certificates. Key highlights of DLDC 2021 Technology talksPaper presentationsExhibitions & HackathonsFull-day Workshop on deep learningNetworkingMeeting potential employers virtually Interaction with speakers, and more. Where: Virtual When: Thursday to Friday | September 23 to 24, 2021 The Association of Data Scientists, in association with Analytics India Magazine, has put together DLDC 2021, with a view to chronicle technological progress in the space of artificial intelligence, machine learning and deep learning. To know more, click here. You can book passes for DLDC 2021 right here.","excerpt":"In an effort to bring the deep learning community together, the Association of Data Scientists (ADaSci), the premier global professional body of data science & machine learning professionals, is launching the second edition of Deep Learning DevCon 2021 (#DLDC21). The two-day leading virtual conference, scheduled for September 23 & 24, will give deep learning practitioners […]","categories":["Deep Tech"],"tags":["Deep Learning","deep learning conference"],"author_name":"Sejuti Das","publish_date":"2021-07-08T11:03:14","publication_year":"2021","word_count":356,"keywords":["data science","Go","machine learning","artificial intelligence","programming_languages:R","chatbots","deep learning","analytics","deep learning conference","GAN","Deep Learning","R"],"extracted_tech_keywords":["artificial intelligence","machine learning","deep learning","data science","analytics","chatbots","R","Go","GAN","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/deep-learning-devcon-2021-the-second-edition-of-virtual-summit-announced\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":35581,"title":"5 Ways Women Can Rise To Senior Management Roles In Indian Tech Sector","content":"A 2017 report by Morgan Stanley states that more gender diversity, particularly in corporate settings, can translate to increased productivity, greater innovation, better decision-making, and higher employee retention and satisfaction. However, this is not something that every company follows religiously. Even though there are several benefits, company leadership across the globe is still unbalanced with women accounting for less than a quarter of management positions globally. And the disparity is even greater when it comes to higher-level management positions. Women bring a different perspective to business and companies across the globe need to acknowledge and leverage it, however, over the years the gap is not filing. Therefore, it is high time to fill the void in the industry and increase the opportunities for more women executives. And in order to combat the gender gap and move women workforce towards greater leadership balance, both companies and women need to make use of some of the key strategies. Welcoming Culture It takes more than just words to overturn years of systemic inequality and create opportunities for more female executives. And in a company, it all stars a more inclusive environment and employees get rewarded based outcomes achieved instead of hours worked. Companies without a culture that doesn’t encourage success are undoubtedly useless, no matter how much they put in their efforts to hire women to leadership roles. One of the most initial steps of creating such a culture at the workplace is to have a diverse pipeline and diverse hiring team. Rather than asking the candidate about her salary history, the hiring department should determine how much certain skills and responsibilities are worth based on reliable market information. Reverse Mentoring Reverse mentoring is usually about letting older executives get mentored by younger employees. And this can even be used in a way where top leaders who are often male get mentored by female mentors who have been identified as future leaders. This is considered to be one of the most effective ways to create visibility of upcoming female leaders to top executives. Also, it let female leaders to a clear view of the most strategic work at the company. Encourage Women To Apply For Leadership Job Roles According to a Hewlett Packard internal report, men usually apply for a job when they meet just 60% of the job requirements, while on the other hand, women only apply for jobs when they meet 100% of the requirements. If the fact is true, then it is high time that women should change this mindset of applying for jobs. In this fast-paced skill-driven era, it is time to drop the conception that you cannot do a job well just because you don’t have the qualification. Furthermore, the other view is that people tend to hold back from applying because they misunderstand the hiring process. They end up believing that they need the qualifications not to do the job well, but to get hired in the first place. Let The Women Leaders Go Out And Inspire To make a difference across industries it is imperative the companies start to inspiring other firms in improving opportunities for a female executive. And one of the best ways to do it is by making women leaders visible to the world. Many companies across the globe raise the visibility of her women employees by encouraging them to take part in events like The Rising 2019 and speaking to a greater crowd. It is without a doubt is one of the best ways to give wings to women to fly and achieve leadership roles. This not only let the women employees share their knowledge but also give them the opportunity to network with the chief from the respective industries. Transparent Career Mapping According to the McKinsey\/LeanIn report, compared to men, women are less likely to get into a manager-level job. And this might happen due to partial career mapping. To deal with this, many companies across the world are taking sheer steps such as matching senior leaders (mostly male) to rising female talent to discuss development plans, stretch assignments, promotions, and networking opportunities. This structure is considered to be one of the best ways to remove the chances of women not being aware of opportunities for their development at higher levels. On a side note, we bring Rising 2019 – Women In Analytics & AI Conference on March 8, Bangalore. Click here to know more.","excerpt":"A 2017 report by Morgan Stanley states that more gender diversity, particularly in corporate settings, can translate to increased productivity, greater innovation, better decision-making, and higher employee retention and satisfaction. However, this is not something that every company follows religiously. Even though there are several benefits, company leadership across the globe is still unbalanced with […]","categories":["AI Trends"],"tags":[],"author_name":"Harshajit Sarmah","publish_date":"2019-02-28T12:14:09","publication_year":"2019","word_count":734,"keywords":["Go","programming_languages:R","AI","innovation","programming_languages:Go","RAG","ViT","analytics","GAN","R"],"extracted_tech_keywords":["AI","analytics","RAG","R","Go","GAN","ViT","innovation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/5-ways-women-can-rise-to-senior-management-roles-in-indian-tech-sector\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10160690,"title":"The MNC Work Culture Crisis in India","content":"Ask any tech graduate in India about their career goals, and chances are they’ll mention landing a job at Microsoft, Google, or a similar multinational company (MNC) – if not in the West, then at least in one of their Indian offices. This is because MNCs in India have built a reputation for offering better work environments, higher salaries, and more global exposure compared to their domestic counterparts. Companies like Microsoft or Google are known to extend the same policies they have in the US to India. However, a closer look reveals significant differences in how employees in India are treated compared to their peers in Western offices. Many express concerns about being ill-treated, facing longer hours, and being excluded from decision-making processes. One Reddit user shared that while their company claimed to follow global policies, the implementation in India felt “selective and unfair”. What’s the Issue? One of the primary draws of working for an MNC in India is the promise of higher compensation. Research indicates that US-based MNCs operating in India pay up to 70% more than Indian companies for equivalent roles. However, Indian GCCs are often called the “cost centres” of the MNCs, as the focus is simply on maximising the ROI. While the salaries may seem high here, they remain relatively low compared to similar roles in Western countries. While speaking with AIM, an employee of Microsoft India who wanted to maintain anonymity said that though they work on many global projects for the company, there is definitely a disparity in leave and remote work opportunities when compared to their US colleagues. The person says the company allows them to work remotely for several weeks if needed, as long as the project gets delivered on time. The employee said that getting the permission to work remotely changes from person to person based on trust of the manager. This was also reflected in a Reddit discussion where an MNC employee said that their bosses here are usually Indians, whose decisions are naturally tainted by the general work culture in the country. “It’s the same mindset we see in many Indian parents, too – that of micromanaging and having tight control over their juniors,” said the employee. Microsoft India declined to comment on the matter when contacted by AIM. Unlike countries that propose the “right to disconnect” to ensure employees can switch off after work, India lacks even basic discussions around such safeguards. The Microsoft India employee also said that though officially there are enough remote work opportunities in their job, unofficially, everything largely hinges on the manager’s permission. “In Indian branches, even when remote work is permissible globally, employees are expected to report to the office, often without a justified reason,” said a user on Reddit. This was also confirmed by the MNC employee who spoke with AIM. Another techie, who has worked for an MNC both in India and abroad, spoke about the stark differences in the work culture at both places. When he was in India, he was added to five different work groups. People who responded after work hours were applauded, and the managers questioned those who didn’t. “The country head told everyone that our work doesn’t end at 5 pm or on weekends, and if need be, we must work through the weekend,” he added. On the other hand, he said that now that he works outside India for the same company, his colleagues do not even have his personal number—something that was unimaginable in India.“My [current] manager pushes back on unrealistic timelines. Yes, there are expectations to deliver high-quality work, but doing some great work in a preplanned timeline with enough time for improvement is all that is expected,” he added. Indian IT Affects the MNC Culture in India Back here in Bengaluru, while everyone talks about the pleasant weather, the city’s corporate culture is somewhat affected by the work culture of the Indian IT industry, which often involves an expectation of long working hours. Narayana Murthy, the co-founder of Infosys, who stirred public debate by suggesting that young Indians should embrace a 70-hour workweek to help drive the nation’s economic growth, stood his ground despite criticism. The abysmal salaries at these companies are also a huge problem. Mohandas Pai, the former CFO of Infosys, recently pointed out that entry-level salaries for Indian IT have stagnated for the last decade and more. In 2011, a fresher’s salary was INR 3.25 lakh per annum, and it has only marginally increased to INR 3.75 lakh per annum on average. “How is it justified?” Pai asked, calling it an exploitation of the workforce. While salaries for freshers might increase in 2025, the jump would be no more than 10%. This increase is also expected due to the increasing number of GCCs in India. This fuels the perception that Indian companies fail to foster a positive work culture. The same sentiment extends to MNCs setting up bases in India. Sanjay Sreedhar, staff engineer at Lenovo, said on Quora that for many MNCs in India, a resource is just a number. “Indian MNCs don’t care about your career growth unless there is a pressing need for them to upskill their employees,” he said, adding that mid-level managers think that everyone just works for them. Will Things Change? In response to the tragic death of the 26-year-old audit executive at EY earlier this year, states like Maharashtra, Telangana, and Karnataka are drafting new workplace rules and increasing inspections to ensure employee well-being. The expectation among MNC employees in India contrasts sharply with norms in Western countries. Employees expect that their managers will treat them the same way as their colleagues in the West. However, several employees have repeatedly said that it all comes down to the manager in charge. Now that most companies are planning to open new offices in Bengaluru, which is emerging as the country’s GCC hub, it becomes important for MNCs to focus on local issues. This includes several work culture issues that the city’s workforce has highlighted for several years. While Murthy and Pai’s remarks sparked conversations about productivity and competitiveness, they also reignited concerns about work-life balance in India’s demanding professional environment.","excerpt":"People who respond after work hours are applauded, and the managers question those who don’t.","categories":["AI Features"],"tags":["AI in India","Work Culture"],"author_name":"Mohit Pandey","publish_date":"2025-01-03T11:11:42","publication_year":"2025","word_count":1029,"keywords":["Go","ELT","AI in India","AWS","AI","Work Culture","RAG","Ray","Aim","Rust","GAN","R"],"extracted_tech_keywords":["AI","Aim","Ray","RAG","AWS","R","Go","Rust","ELT","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/the-mnc-work-culture-crisis-in-india\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10011914,"title":"How Indian Railways Uses AI: A Comprehensive Case Study","content":"For decades, the Indian Railways has been a significant and sole way for commutation in the country. Being the third-largest railway network in the world by size, it accommodates and offers its services to people from all walks of life. With that being said, such a mammoth size of the railway system needs several aspects of the working to align for the system to function smoothly. And, even the slightest of disruption in the network can create hassle in the schedule, creating inconvenience to lakhs of people. Thus, to make the running of the system run smoother, there was a requirement for Indian Railways to deploy a consolidated system that runs on automation to get the job done in real-time. Traditionally, the data was being collected manually, and given how large the system is, there were bound to be a few lapses. Indian Railways was looking for a solution that could improve their services offered to the customers and their experience while commuting. The services for stations and while onboarding customers is a 365×24\/7 task, which was done by the last-mile staff, which made it difficult for the railway officials to have real-time visibility of delivery of services. To overcome the challenge of service assurance and passenger satisfaction at stations and onboard trains, Indian Railways collaborated with Gaia, an enterprise AI and IoT solutions provider to enable a digitised and data-driven operation. Also Read: How The Government Of Odisha Used Geospatial Technology To Provide Land Rights To Slums How AI & IoT Helped Indian Railways? Gaia created a customised solution using its SmartFeedback software platform, which in turn allowed an integrated CleanRail solution. Explaining better, the director and co-founder of Gaia, Amrita Chowdhury said, “The solution brought in a modular IoT and digital\/ AI stack that was specifically designed to help Indian Railways manage and optimise their performance, reduce downtime, and increase responsiveness and productivity.” It comes with a two-sided platform for governance monitoring and orchestration of operations between various stakeholders. To facilitate this, Gaia integrated a real-time Onboard Housekeeping Service (OBHS) management that not only manages people onboard but also manages operation processes. The system collects data via IoT sensors as well as users’ devices in real-time, with time and location stamps to ensure data efficacy and accountability of the entire working in Indian Railways. The IoT devices are paired with sensors to enable last-mile monitoring of parameters such as temperature, pressure, humidity, power usage, faults, proximity, footfall, and other metrics to understand site and asset performance in real-time. Source: Gaia’s website It then uses AI algorithms for allocation, optimisation, and task orchestration. A comprehensive action and workflow engine complement the platform for managing alerts and incidents, triggering workflows, and enabling closed-loop response management. The solution is a multi-tenanted platform that allows role-based access for users – from line staff and managers to senior management and board-level executives – on a unified platform. It also has local backup storage for edge analytics, where the edge device has recursive neural network algorithms for detecting sound based inputs and interpreting. The cloud platform uses cognitive AI algorithms for fraud management and multi-factor swarming algorithms for allocation and optimisation. The solution was further integrated into Railways E-Drishti platform, allowing the railway board to access the status of national operation on a virtual GIS dashboard. With this, the managers and officers at the depot level can access the information at the train or group level. “Our proprietary hardware electronics are built on STMicroelectronics STM32 chipset, and we have been the first in India to be incubated by STMicroelectronics Startup Labs,” said Chowdhury. “The proprietary software stack is built on open-source tools, but uses Azure infrastructure on the back.” This allows scalability and enhances the security and encryptions of the solution. Also Read: How This Women-Owned Milk Producer Company Enabled A Smoother Payment Process With SAP-ERP Wrapping Up With Benefits With Gaia’s SmartFeedback-based CleanRail solution, Indian Railways have seen a massive transformation in the existing system. Some of the key achievements were — obtaining a real-time OBHS data collection that improved overall accountability of the operations, along with that was the introduction of big data analytics and cognitive AI facial recognition and GIS dashboards that offered real-time insights. This, in turn, improved the service assurance with a timely response to passenger grievances based on status updates and alerts. Indian Railways has seen a 100% increase in digitalisation, accountability, and ecosystem transparency, among other things. The availability of real-time information and robust alerts has improved performance against SLAs by approximately 30-50%. Additionally, alerts for events or incidents that bring in location-wise information and intelligence allows staff to have early warning signals of the site performances, assets, and people they manage for faster response. “At the national level, the railway managers and leaders can now monitor the performance of onboarding services vendors across zones in an integrated fashion,” concluded Chowdhury.","excerpt":"For decades, the Indian Railways has been a significant and sole way for commutation in the country. Being the third-largest railway network in the world by size, it accommodates and offers its services to people from all walks of life.  With that being said, such a mammoth size of the railway system needs several aspects […]","categories":["AI Features"],"tags":["ai big data analytics digital transformation","case study","indian railways"],"author_name":"Sejuti Das","publish_date":"2020-11-20T14:00:56","publication_year":"2020","word_count":814,"keywords":["indian railways","Go","big data","AI","ai big data analytics digital transformation","neural network","R","Scala","Git","RAG","case study","analytics","Azure"],"extracted_tech_keywords":["AI","neural network","analytics","RAG","Azure","R","Go","Scala","Git","big data"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-indian-railways-uses-ai-a-comprehensive-case-study\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10123518,"title":"Databricks Looks to Dominate Enterprise AI With Open-Source","content":"Data and AI giant Databricks announced a host of new generative AI capabilities and a major push to its open-source strategy at the annual Data + AI Summit. The new offerings, such as Mosaic AI Model Training, Mosaic AI for RAG, and Mosaic AI Gateway, in addition to open-sourcing their Unity Catalog, aim to help enterprises build high-quality, domain-specific AI applications. “We want to help people get the best quality possible in their domain for their GenAI application,” said CTO and co-founder Matei Zaharia in an exclusive interview with AIM. “And to do that, we see a lot of companies are building what we call compound AI systems.” These compound AI systems involve multiple components, such as calls to different models, retrieval of relevant data, use of external APIs and databases, and breaking problems into smaller steps. At the same time, Databricks is also focusing on open-source models. Why is Databricks Betting Big on Open-Source? While acknowledging the rapid advancements in closed-source models, Zaharia noted that Databricks is definitely betting big on an open-source strategy. He also believes that the performance gap between closed-source and open-source models is rapidly narrowing. This is evidenced by recent open-source models like DBRX, Mistral 8×22 billion, and Llama 3 approaching the quality of the best closed-source models. “They’re all quite good, and they’re all in that space, getting really close to the best closed models. Meanwhile, the best closed models haven’t gotten that much better.” Acknowledging the possibility that significantly higher investments could lead to superior closed models, Zaharia believes that open-source development will continue to thrive as companies seek to share development costs. While consumer AI applications may stagnate, Zaharia predicts that the most exciting advances in generative AI will come from open, customisable models in the B2B world, applied to complex industry use cases. “I actually think the most exciting sort of advances in GenAI will next be in the B2B world with custom AI for challenging mission-critical domains,” said Zaharia. “That’s another reason that we’re betting on open models,” he added. He drew parallels to how open-source big data technologies initially powered consumer applications but later had a transformative impact on the enterprise. He elaborated, “Let’s say you build a model for chemistry. That’s really good. Even if it’s not as good at chatting about random topics as GPT-4, it’s still extremely valuable.” Mosaic AI for Training Cost-Effective Models & Quality Monitoring The new offerings in Mosaic AI are designed to address major hurdles like quality, cost, governance and security that organisations face in building and deploying generative AI applications. “If you don’t get the right kind of quality for your application, then you’re stuck,” Zaharia emphasised. One key offering is the RAG framework in Mosaic AI, which provides a quick way to deploy and manage an entire generative AI application, including the vector database, data pipeline, and serving layer. Databricks is also introducing quality monitoring capabilities for compound AI applications. This includes the ability to see detailed traces, review results, and even use LLMs as automated judges to score outputs. “So, of course, you can do prompt engineering, you can try to tell the model to do different things, but at some point, if you have examples of data that you can label and give it, you’ll do a lot better,” explained Zaharia. “And we actually packaged up all the stuff that we used to train DBRX, which our research team had just developed. It’s now behind a very simple serverless API,” he added. Additionally, Mosaic AI Training enables organisations to fine-tune models using their own labelled data, resulting in higher-quality outputs. Fine-Tuning and Cost Reduction Zaharia underscored the significance of fine-tuning foundation models on organisations’ own data with Mosaic AI Model Training. He cited the example of FactSet, a financial data vendor that initially built an application using GPT-4, which was, at best, 55% accurate and took 10 seconds per user query. By switching to a multi-step AI system with Databricks, FactSet achieved 87% accuracy, reduced query time to three seconds, and lowered costs by approximately five times compared to GPT-4 calls. Mosaic AI Training is an optimised software stack that makes training LLMs cost-effective. Through system-level optimisations, tuned parallelism strategies, and model training science, it can reduce training costs by up to 10x. Zaharia emphasised that another factor to consider for cost-efficiency benefits is using custom and open-source models. “You might have something that works, but it’s very expensive, very slow. “This is where custom models and open-source models provide a huge benefit because you can often take something that works well with a very expensive model, collect a bunch of examples of it and then fine-tune a small, low-cost model to do it well,” he explained. Databricks’ DBRX model, for example, surpasses GPT-3.5 in quality while being faster and more cost-effective to serve, with costs similar to a 13 billion parameter model. By incorporating Databricks Mosaic AI into their data strategy, organisations can experience reduced training time and costs, improved model performance, increased developer productivity, enhanced scalability, and democratised AI.","excerpt":"Zaharia believes that open-source models are narrowing the gap, while the best closed models haven’t improved much.","categories":["AI Features"],"tags":["Databricks"],"author_name":"Shyam Nandan Upadhyay","publish_date":"2024-06-13T15:02:26","publication_year":"2024","word_count":843,"keywords":["GenAI","TPU","AI","serverless","RAG","Aim","prompt engineering","generative AI","foundation models","Databricks"],"extracted_tech_keywords":["AI","generative AI","GenAI","foundation models","Aim","RAG","prompt engineering","serverless","TPU","Databricks"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/databricks-looks-to-dominate-enterprise-ai-with-open-source\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":41523,"title":"Industry View: What Are The Popular Tools And Techniques Used By Analytics Practitioners","content":"As new problem statements keep evolving in data science and analytics domain, companies are exploring newer ways to deal with them. The tools and techniques used by data science practitioners have evolved significantly with newer and better tools available in the market. While Python and open source tools currently dominate the market, there are other interesting trends that we saw during our interaction with analytics practitioners for this month’s theme — evolving data science toolchain and new techniques used by data scientists across industries. We spoke to analytics leaders from different domains like e-commerce, insurance among others. The detailed story covers insights on the most commonly used tools by them, must-have in a data science tool kit, common issues faced while dealing with data, use of open source or paid software and tools and selecting the right tool, among others. Tools That Are Widely Popular In Analytics & Data Science Community Data Science is a big domain and includes many areas such as machine learning, data visualisation, data manipulation, and more that are usually bracketed under the umbrella term data science. It also has different aspects such as statistical and mathematical implementation. This calls for different languages, libraries and frameworks that are used for different domains based on the kind of task that needs to be dealt with. “Tools and platforms commonly used for mathematical computation are RapidMiner, Theano, Scipy, Numpy, Mathplotlib, for AI, ML, Deep Learning are TensorFlow, Scikit learn, Torch, OpenCV, and for data analytics and visualisation are Orange, Tableau, Knime,” says Anupam Jalote, CEO, iCreate. Gurprit Singh, Managing Partner and Co-Founder at Umbrella Infocare says, “Spark, R, Python, TensorFlow, Apache MXNet are much favoured in analytics and AI industry. However, as far as cloud is concerned AWS Glue, AWS EMR, Sagemaker are most commonly used.” On the other hand, Sidhant Maharana, Data Scientist at Hitachi Vantara shares that most commonly used tools for machine learning is Sci-Kit Learn, for data manipulation it is Pandas, Numpy stands out for numerical calculations whereas as for AI and deep learning it is mostly Tensorflow, Keras and PyTorch. “Some of the tools used by us are Amazon Redshift, Google Big Query, R, Python, SQL, PowerBI, RShiny, Qlik, Tableau, Gitlab, Bamboo, Jenkins for various tasks such as data engineering, data mining, modelling and analysis, visualisations, DevOps are more,” shares Vedant Prasad, Partner at TheMathCompany. Puja Gorai, Lead Decision Science, Jumbotail said, “Depending on the context and the complexity of the problems we use different languages like Python, R, Java, Scala, and a multitude of algorithms, both proprietary and open source, to solve the problems in the most efficient way. We also use Periscope Data, and Tableau for data visualization company-wide.” Whereas at Synechron, they use technologies such as Java, Spring, restful web services, HTML5, Angular JS, MySQL database, R, Python, NLTK, Neuroph, Encog to build solutions for our financial services clients, shared Faisal Husain, Co-founder and CEO, Synechron. The Most Preferred Cloud Provider— AWS, Google Or Azure? “Our choice for a Cloud provider is always Amazon Web Services, it has a wide variety of offerings and provides great compatibility and advanced features,” says Jalote. Prasad also shares that the services offered by all three – AWS, Google and Azure – are mostly similar and the preference is determined by what works for a specific use case. Krishnan Parameswaran, Co-founder & CTO, Namaste Credit says that they prefer using cloud-based services from AWS because of the various services offered as well as due to the fact that it is mostly self-driven. Maharana and Singh echoed a similar thought with AWS as their most preferred cloud provider. Proprietary Tools Developed In-House By The Company There are many companies that are working on developing their in-house tools to address specific gaps in their data science lifecycle. Prasad revealed the company is preferring to build in-house tools as the current array of available tools is too generic to cover all the data problems. There are other companies that are developing in-house tools to expedite the process and reduce delivery time. “We have developed certain frameworks based on our vast experience and exposure in the industry, through multiple engaging customer projects,” says Singh. Maharana shares that they have built a Manufacturing Insight tool which takes input from the sensors and send alerts if it is an anomaly behaviour from a machine. It is used even for predicting an anomaly within a given time period based on different features from a machine. Open Source Vs Paid Tools While companies mostly use prefer using open source tools, Prasad says that the choice between open-source and paid tools depends on how well chalked-out the business use case is. “Paid tools are built with specific analytics use cases in mind, usually specific to certain domains. When it comes to building end-to-end solutions from scratch, we have found open-source tools to be more useful,” he says. Some tools and platforms that they use are Python, Jupyter, Zeppelin, PySpark, Tensorflow, Keras, FastAI, Dockers, Kubernetes etc. Whereas some of the paid tools that they prefer are Amazon S3, Redshift, AWS Glue, Batch, SageMaker, Tableau. Parameswaran shares that open source tools such as Spark provide the basic framework and based on their use case they tweak the tool or build algorithms on top of it to make sure there is a fitment. The fact that many developers contribute to open source tools and newer versions come out on a regular basis, it makes open source tools to be much preferred by the data science community. Ashish Kanaujia, External Mentor at iCreate echoes similar views as their personal preference lies with open source tools and frameworks, allowing them to tinker around and learn the mechanics behind the actual workings of various algorithms. Maharana shares similar views where he says open source tools have a lot of contributors which results in a constant evolution of the tool, whereas the benefit with paid ones is that they have better services and inbuilt algorithms. However, Singh believes that using paid tools has its own benefits such as advantages in terms of manageability over open source tools. “Paid tools are built on top of open source tools, but there is some amount of support available from organizations, and some automation is already built-in, making it easier for customers to use,” he says. Selecting The Right Tool Is Critical While Dealing With Data Data scientists share a different set of challenges they face while dealing with data. For instance, Maharana shares that he faces a challenge while dealing with a big chunk of data. He also lists other challenges such as dealing with a lot of garbage sensor data and challenges of preparation and pre-processing phase. “Selecting the right tool is the most critical aspect for a data scientist. Before committing to a particular tool to solve the problem I take sample data and create a prototype of the business solution and compare a couple of tools to see how they behave. Only after getting a belief on a specific tool, that I commit to the tool or algorithm,” he says. Singh shares that some of the most common issues faced by them include size and growth rate of data, identification of the source and type of data, identification of unique data points and checking the quality of data. He says that the selection of right tools is extremely important as data different data structure might call for use of a different tool. Husain shares that some of the challenges they face are aligning a data set to an analytics strategy and to factor data maturity. “It is important for organisations to be equally invested in and committed to business process transformation to reap the full benefits of the analytics platform,” he says. Parameswaran shares that selecting the right tool is very critical. “In fact, half of the problems are solved if you can identify the right tool,” he says. “Sometimes, the tools are very complex and are not very user-friendly or easy to use which prevents it from getting traction. Many tools need additional training for adoption which becomes difficult to sustain,” he adds. “Structure and format of available data are the most critical issues with data analytics,” shares Jalote. “Problem arises especially with old archives of stored data where no proper standards or structure was in place, handling such data and generating meaningful analysis out of it is a tedious task,” he adds. “There are two distinct aspects of data science – math and modelling along with backend and data-wrangling. Most of the time, the data-wrangling piece is a prerequisite to the math\/modelling and a majority of the time is spent on cleaning the data in a real-world scenario. It is therefore important to understand the business problem and the goal to identify right tools,” says Vishal Shah, Head of Data Sciences, Go Digit Insurance. What Does An Ideal Data Science Toolkit Look Like? Singh says the Hadoop, Spark, Python, R makes up for a good data scientist, whereas Maharana shares that data scientist should be a combination of a lot of tools such as Numpy for mathematical operations, Pandas for data manipulation, Matplotlib for data visualisation, Scikit-Learn for machine learning library and more. Kanaujia shares that there is no one size fits all technique. “Data science is a very vast ecosystem where different applications require a different set of tools and methodologies. Therefore, one cannot come up with an optimal set of tools for a toolkit but it is a combination of understanding of a lot of tools. Prasad, on the other hand, believes that one of either R or Python is a must in an ideal data scientist’s toolkit. Similarly, Parameswaran believes that data scientist’s toolkit should consist of any of the AI frameworks.","excerpt":"As new problem statements keep evolving in data science and analytics domain, companies are exploring newer ways to deal with them. The tools and techniques used by data science practitioners have evolved significantly with newer and better tools available in the market. While Python and open source tools currently dominate the market, there are other […]","categories":["AI Features"],"tags":["data science tools","devops tools","Visual Analytics Provider"],"author_name":"Srishti Deoras","publish_date":"2019-06-28T12:11:28","publication_year":"2019","word_count":1627,"keywords":["data science","data science tools","machine learning","Keras","AI","PyTorch","ML","devops tools","Ray","Visual Analytics Provider","deep learning","analytics","TensorFlow"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","data science","analytics","Ray","TensorFlow","PyTorch","Keras"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/industry-view-what-are-the-popular-tools-and-techniques-used-by-analytics-practitioners\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":13142,"title":"10 Jobs that AI is likely to replace sooner than you think","content":"Artificial Intelligence is undoubtedly the technology which will shape the future and ease most of the human tasks using automation. The use of robots powered with Deep Learning technologies and cognitive capabilities will be a commonplace. But, there’s a darker side to this technological growth. As more and more jobs get automated, job-seekers will not only have to compete with other job-seekers, but the future also implicates that our race will have to compete with AI-powered bots in areas of employment. Robots have been used in manufacturing industries for quite a while now, to accomplish the same tasks much faster and more efficiently. But, with time, more industries are leveraging the scope of AI and cognitive computing, ranging from the transport industry and stretching as far as customer service. Studies indicate that almost half of commonly held careers are above a 50 percent risk of undergoing automation before 2035. An interesting pattern to note here is that most of the jobs likely to get replaced by AI are repetitive in nature, and require less time in real-time decision making. With that in mind, let’s glance through a list of jobs likely to be replaced in the earlier phase of adopting Artificial Intelligence. Accountant Humanoid accountant Robot Humanoid accountant bots are still largely in the development phase, however, many organizations are automating their accounting processes. Soon, there will be no specific requirement for dedicated accounts, payable and receivable employees. Finance jobs might soon be completely automated, as the integration of AI-based technologies will no longer require humans to crunch numbers, or punch them, for the same regard. The companies currently leading this transformation are Lexmark and Basware, who offer fully automated accounts systems, capable of executing tasks such as matching purchase orders or flagging invoices for payment. Smacc is another firm spearheading the integration of a fully-automated accounting process. The system has to be just fed with the relevant receipts, and the underlying AI will handle, prepare, and examine all of the financial records. Cab and Truck drivers Uber’s Autonomous vehicle Drivers of ride service providers like Uber and Ola could be the next job AI will most likely replace. Autonomous vehicle will soon be a reality owing to the drive from firms like Uber and Google. The service would be a lot more inexpensive and its profits much greater, as the integration of AI will push away lot of these drivers from their jobs. Cities across the world will witness fleet of self-driving cabs within the upcoming decade. Uber has already been testing such vehicles, and Singapore marks the first country to have put up to a dozen on the roads. This situation is not limited to the plight of cab drivers, as AI will also put Truck drivers and drivers of commodity vehicles out of work. These vehicles will be powered by AI and cognitive technologies, to enable them to self-steer and deliver the right goods at the right place. Data Entry Robots can automate Data Entry jobs Data holds critical value in today’s world and businesses across several verticals are moving on to a data-driven approach. This indicates the importance of entering accurate data, and businesses are automating the way data is collected and processed. Most importantly, by onboarding AI-driven technologies, the chances of human error will be significantly reduced Sensors can assist in capturing real-time data, which is eventually fed into the backend system. This data can be automatically compared against goals and objectives to maximize efficiency. Automation of data entry processes thus prove more promising, despite the fact that it will replace all the data entry jobs in the market. There is a lot of data around us which needs to be digitized. Integration of AI, despite the drawback, will become an absolute necessity for data entry jobs in the future. Factory Workers Foxconn is bringing in more robot workers Machines have been automating human tasks since ages, reducing their effort and work. However, today’s job floor presents its own unique challenge, one that revolves around automation of all manufacturing processes, one that will also put several factory workers out of work. Factories today are gradually building an ecosystem which uses robots powered by AI technologies, in place of humans. However, the number of such factories is still countable, as the technology was widely adopted only recently. Companies such as Foxconn is already undergoing this transformation, as they replaced 60,000 employees with robots. Food Delivery JUST EAT pilots a Starship robot to deliver food Another job which might soon be utilizing AI-powered bots, is the food delivery service. These jobs are repetitive in nature and don’t require much decision-making in the process. Just Eat, an UK-based online delivery service, has initiated the use of robots to deliver food in London. Anglo-Estonian company, Starship Technologies is credited to the engineering behind these self-driving robots. The robots are embedded with GPS and cameras to navigate the capital’s thoroughfares. The robots are also integrated with in-built security and remote-tracking features. In cases of an attempted theft, the robot transmits an alert to the control center, besides capturing a photo of the thief. Insurance Underwriter Meet your new underwriter Essentially an underwriter’s job encompasses vetting an insurance application, so as to determine if the risks involved is worth accepting for the insurance company. After reviewing the application, he\/she decides on whether the insurance can be provided, and the amount to be offered. However, automation has also diminished the value of an underwriter’s job, as these applications can be standardized, and most of these organizations have set rules, comparing which they assert eligibility. The use of AI will help computer systems to learn these rules and apply them in vetting the application. AI will undoubtedly ease the work for most insurance providers, but it will simultaneously put many out of work. Receptionist Meet Hilton Hotels’ robot receptionist This is another avenue being currently explored with AI-based technologies. The inclusion of robots in offices and hotels may not get you the warm embrace you would expect of a human receptionist, but it has been surely designed to provide you with the information you need, in a very efficient manner. Bank of Tokyo-Mitsubishi’s flagship branch in downtown Tokyo already has a multilingual android greeting people. The day is not far when hotels and other similar setups would only have robots serving all human needs. Security Guard Knightscope K5 robot With the advent of AI and related machine learning technologies, we are considering a newer world with unexpected technological capabilities. While most of the industry verticals drive their efforts towards incorporating the technology, why must patrolling and surveillance sectors be left behind? This is another domain, where AI will soon put humans out of work. Robotic security guards are being used currently to monitor businesses, and the day is only approaching closer when these robots will relieve human protection agents of their jobs. The K5robot, developed by Knightscope monitors its surroundings for suspicious behavior, besides garnering the ability to detect potentially criminal “audio events”. Telemarketer AI can be the future of Telemarketing Telephone salesperson are most likely to get washed away by the approaching tide of AI-powered technologies. Currently, sales calls are already being generated by voicebots, leveraging the potential of natural-language processing technologies. Within the next couple of years, the bulk of telemarketing roles are likely to be performed by robots. Furthermore, the result of extensive automation could implicate in fewer calls to helplines, at least from a customer service perspective. The use of smart systems, remotely monitored by sensors, would help with product maintenance and elimination of potential problems. Tour Guide Otonaroid during an event at the National Museum of Emerging Science and Innovation The fact that this is yet another job which doesn’t involve decision-making, and is repetitive for most part, makes it extremely vulnerable to be replaced by AI-driven technologies. Human-like androids are already serving as exhibition guides at Japan’s National Museum of Emerging Science and Innovation. One of the robots, called Kodomorid evangelizes visitors about new stories relevant to the exhibits, while the science communicator bot Otonaroid can chat with visitors or answer their questions related to science.","excerpt":"Artificial Intelligence is undoubtedly the technology which will shape the future and ease most of the human tasks using automation. The use of robots powered with Deep Learning technologies and cognitive capabilities will be a commonplace. But, there’s a darker side to this technological growth. As more and more jobs get automated, job-seekers will not […]","categories":["AI Trends"],"tags":["Artificial Intelligence India","automation India","cognitive computing human capital","insurance","Natural Language Processing","robots India","Sensors"],"author_name":"Amit Paul Chowdhury","publish_date":"2017-03-02T08:06:56","publication_year":"2017","word_count":1359,"keywords":["Go","API","insurance","artificial intelligence","machine learning","AI","Natural Language Processing","R","Artificial Intelligence India","Git","RAG","automation India","deep learning","Sensors","GAN","cognitive computing human capital","robots India"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","deep learning","RAG","R","Go","Git","API","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/10-jobs-ai-likely-replace-sooner-think\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10008766,"title":"Tabulate python Examples to Create Formatted Tables","content":"Visualizing the data in tabular form is easier than visualizing it in a paragraph or comma-separated form. Nicely formatted tables not only provide you with a better way of looking at tables it can also help in understanding each data point clearly with its heading and value. Tabulate is an open-source python package\/module which is used to print tabular data in nicely formatted tables. It is easy to use and contains a variety of formatting functions. It has the following functionalities: One function call for all types of formatting Can be downloaded in multiple output formats Provides a better presentation with text and data. In this article, we will see what are the different types of table formatting we can perform using Tabulate. Implementation: We will start by installing tabulate using pip install tabulate. 1. Importing Required Libraries We will be using the tabulate function from the tabulate library so we need to import that. Other than this we do not require to import any python module. from tabulate import tabulate 2. Creating Formatted Tables Now we will start by creating different types of formatted tables. data = [[\"Himanshu\",1123, 10025], [\"Rohit\",1126,10029], [\"Sha\",111178,7355.4]] print(tabulate(data)) Here we can see a plain table which is nicely formatted. Now let’s see how we can add a header to this table that we just created. print(tabulate(data, headers=[\"Name\",\"User ID\", \"Roll. No.\"])) In order to define the header along with the data, we can set that header=’firstrow’, let us see it through an example. data = [['Name','ID'],[\"Himanshu\",1123], [\"Rohit\",1126], [\"Sha\",111178]] print(tabulate(data, headers='firstrow')) We can also display the indices of the rows by using the show index parameter. data = [['Name','ID'],[\"Himanshu\",1123], [\"Rohit\",1126], [\"Sha\",111178]] print(tabulate(data, headers='firstrow', showindex='always')) Now let us see what are some of the formats in which we can print the table. Create Plain Tables data = [['Name','ID'],[\"Himanshu\",1123], [\"Rohit\",1126], [\"Sha\",111178]] print(tabulate(data, headers='firstrow', showindex='always', tablefmt='plain')) Create Fancy Grid data = [['Name','ID'],[\"Himanshu\",1123], [\"Rohit\",1126], [\"Sha\",111178]] print(tabulate(data, headers='firstrow', showindex='always', tablefmt='fancy_grid')) Create Jira data = [['Name','ID'],[\"Himanshu\",1123], [\"Rohit\",1126], [\"Sha\",111178]] print(tabulate(data, headers='firstrow', showindex='always', tablefmt='jira')) HTML data = [['Name','ID'],[\"Himanshu\",1123], [\"Rohit\",1126], [\"Sha\",111178]] print(tabulate(data, headers='firstrow', showindex='always', tablefmt='html')) Textile data = [['Name','ID'],[\"Himanshu\",1123], [\"Rohit\",1126], [\"Sha\",111178]] print(tabulate(data, headers='firstrow', showindex='always', tablefmt='textile')) Similarly, there are many more table formats that we can use to design or format our table. Conclusion: In this article, we saw how we can create a table using the data we provide as input. We saw how we can create headers, indices for rows. After that, we saw some of the table formats which are defined in the tabulate library. Popular Guides Dickey Fuller Test in Time Series Hidden Markov Model and Application MLP Classifier Pytorch Optimizers Categorical Data Encoding","excerpt":"In this article, we will see what are the different types of table formatting we can perform using Tabulate.","categories":["Deep Tech"],"tags":["ai for beginners tutorial","Data Analytics","Visualization"],"author_name":"Himanshu Sharma","publish_date":"2020-10-04T10:00:00","publication_year":"2020","word_count":431,"keywords":["Go","TPU","AI","PyTorch","ai for beginners tutorial","ML","RAG","Python","programming_languages:Python","ai_frameworks:PyTorch","Data Analytics","Visualization","R"],"extracted_tech_keywords":["AI","ML","PyTorch","RAG","TPU","Python","R","Go","ai_frameworks:PyTorch","programming_languages:Python"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/beginners-guide-to-tabulate-python-tool-for-creating-nicely-formatted-tables\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10122421,"title":"Microsoft Announces $3.2B Investment To Expand Cloud and AI Infrastructure in Sweden","content":"In a major boost to Sweden’s economy and AI landscape, Microsoft has unveiled its largest single investment in the country, which will be invested over the next two years. The investment of SEK 33.7 billion, or USD $3.2 billion, aims to accelerate Sweden’s adoption of artificial intelligence (AI), enhance workforce skills, and drive long-term economic growth. Microsoft is significantly expanding its cloud and AI infrastructure by deploying 20,000 advanced GPUs across its data centers in Sandviken, Gävle, and Staffanstorp. In addition, the company said it is committed to upskilling 250,000 Swedes, representing 2.4% of the population, in AI over the next three years through a variety of technical training, vocational programmes, and expert development courses. To guide these skill development programs, Microsoft is establishing an AI Insights Council, which will bring together leaders from academia, business, and the public sector. Furthermore, the company is adhering to its AI Asset Principles and the newly unveiled Community Pledge, ensuring that AI innovation is conducted responsibly and has a positive impact on local communities. “Our investment in Sweden is proof of our confidence in this nation, its government and its potential as a leading player in the AI era,” said Microsoft president Brad Smith. The tech giant’s commitment extends beyond technology, aiming to provide broad access to AI tools and skills for Sweden’s population and economy to thrive. Sustainability is a key priority, with Microsoft’s Swedish data centres powered by 100% fossil-free energy and innovative water conservation measures. The company has also invested in nearly 1,000 MW of renewable energy production in Sweden, including a hybrid wind and solar project expected to be operational by 2025. With this massive investment, Microsoft aims to position Sweden as a global leader in AI innovation and technology, accelerating economic growth and fostering responsible progress in the AI era. This is one of many expanding foreign investments made by Microsoft in expanding countries’ AI and cloud infrastructure. Just a few months ago, Microsoft made investments of USD $1.7 billion in Indonesia over the next four years to bolster the country’s cloud and AI infrastructures, provide AI skilling opportunities for 840,000 people, and support the nation’s growing developer community.","excerpt":"Sweden’s people and economy are set to thrive in the AI era with the Microsoft partnership","categories":["AI News"],"tags":["AI","ai announcements","AI Infrastructure","Microsoft"],"author_name":"Gopika Raj","publish_date":"2024-06-04T14:46:02","publication_year":"2024","word_count":361,"keywords":["Go","artificial intelligence","programming_languages:R","AI","innovation","programming_languages:Go","ai announcements","Aim","AI Infrastructure","R","Microsoft"],"extracted_tech_keywords":["AI","artificial intelligence","Aim","R","Go","innovation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/microsoft-announces-3-2b-investment-to-expand-cloud-and-ai-infrastructure-in-sweden\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":65169,"title":"Why Data Scientists Should Follow Software Development Standards","content":"As the organisations have started to recognise the value of data, the need for data scientists has seen an exponential rise since then. Unlike traditional methods, crucial decisions among organisations are now mostly data-driven. However, at the present era, while researching on the grounds of machine learning and data science has become a go-to thing for organisations, deploying the sophisticated machine learning algorithms and yielding more than 90 per cent accuracy in the outcome of the model is a complex and arduous process. The Software Development Life Cycle is the most important process followed by the software industry. It helps software developers to design, test as well as deploy robust and high-quality software products. Considering software development standards into data science model-building will help developers create robust and cost-effective machine learning models. Recently, IBM, at Think Digital 2020 virtual conference, launched Watson AIOps for IT operations management with the latest technologies. According to sources, the AIOps utilises machine learning and advanced analytics along with automation technologies to assist firms in detecting IT incidents and responding to restore services quickly. According to Microsoft Research, creating and running software products require large amounts of raw data about the development process and customer usage, which can be turned into actionable insights with the help of skilled data scientists. Unfortunately, data scientists with analytical and software engineering skills who have the ability to analyse the large raw data sets are usually hard to acquire in an organisation. Implementation of software development standards into the domain of data science researches will indeed provide a beneficial path for data scientists. Here, we will depict some of the crucial points as to why data scientists should follow the software development standards. The Road To Clean Code While developing software, software developers follow a specific methodology of writing codes that is comfortable for other readers and writers. The methodology includes mentioning class and functions, well-documented codes, clear inline comments, error messages, limit line length, among others. A data scientist is also a programmer or coder who is concerned with coding a particular problem and hence finds solutions to the problem. However, one thing that they should follow from software developers is the way to write clean codes. Most of the time, data scientists work on their own and while working on the complex computation, they usually miss simple rules such as clear inline commenting, well-documenting, keeping indentation consistent and other such. This in result makes the code messy and hard to understand for any other developers except the one who wrote it. From Research To Product Software development follows a structured approach when it comes to design, develop, test as well as maintain a high-quality product. In an SDLC, the development process makes it easy for developers to create and build a product easily without witnessing many failures. But, in a machine learning model building, data scientists are often in a dilemma about the model’s performance as it varies in the real-world use cases, which is often contrary to the performance during research. Consequently, data scientists from the early stage of a project should focus on writing production-level code. This will ensure an effortless workflow while taking the research into production. The Need For Automation Embracing the power of automation is one of the methods for increasing the efficiency of a product in software development. While building machine learning models, 80 per cent of the time by data scientists is invested in data wrangling. This, in result, makes the data science projects take longer to deliver impactful results for the business. Consequently, they should embrace automation solutions that can help them in expediting the development of machine learning models. Ryohei Fujimaki PhD is founder and CEO of dotData said, “The benefit of this automated approach is that it provides data scientists with the assistance needed to test for scenarios that they may not have ever considered (discovering “unknown unknowns” to borrow a phrase).” He added, “Also, it allows data scientists to try significantly more use cases and dramatically shorten the time needed to reach highly impactful ones.” Considering The Best Practices Some of the best practices in software engineering include dynamic requirements, use of component architecture, quality assurance, control change, and other such. These are a set of proven approaches to the software development cycle. When combined, these approaches strike at the root cause of software development problems. Following these practices, data scientists will have the ability to manage the requirements of customers Following The Path To QA Quality Assurance and Testing are the most crucial tasks in the software development method. It helps developers to find and avoid possible errors and mistakes in the product while maintaining the integrity of the services. Similarly, in machine learning workflows, a data scientist can perform a quality check on the data that is being used, the quality of the machine learning algorithms and other such. For instance, most of the time, a data scientist fails to build a machine learning model that can fortify adversarial attacks dataset. A proper quality check, similar to software development is essential to ensure one is offering robust AI-based products in the market.","excerpt":"As the organisations have started to recognise the value of data, the need for data scientists has seen an exponential rise since then. Unlike traditional methods, crucial decisions among organisations are now mostly data-driven. However, at the present era, while researching on the grounds of machine learning and data science has become a go-to thing […]","categories":["AI Features"],"tags":["automated data science solutions","Data Science Career","Data Science Jobs","Data Scientist","Software","software automation testing","software developers India","Software Development"],"author_name":"Ambika Choudhury","publish_date":"2020-05-13T20:13:47","publication_year":"2020","word_count":858,"keywords":["data science","Go","machine learning","AI","Data Science Jobs","Git","Data Scientist","automated data science solutions","Data Science Career","software automation testing","automation","analytics","GAN","Software Development","Software","R","software developers India","adversarial attacks"],"extracted_tech_keywords":["AI","machine learning","data science","analytics","R","Go","Git","GAN","adversarial attacks","automation"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/why-data-scientists-should-follow-software-development-standards\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10111986,"title":"How Broadridge’s OpsGPT is Redefining Trading","content":"One of the leading names which comes to mind while talking about the companies fueling the generative AI wave in banking and finance is New York-based fintech giant Broadridge. With three generative AI products already out, the company has a total of 18 such projects going on in this space. Two weeks ago, the company introduced a new generative AI product called OpsGPT which leverages transactions, settlements, and positions data to offer real-time visibility, facilitating faster fails resolution, research on next best actions, and prioritisation of key risk items through a user-friendly interface. The application addresses industry and regulatory changes, including reduced settlement cycles and the need for operational efficiency due to increased trading velocity. “We expect OpsGPT to be a game-changer in the field, especially with the upcoming T+1 settlement cycle in the US, as it will help customers process transactions faster,” said Sheenam Ohrie, MD, Broadridge Financial Solutions, India, in an exclusive interview with AIM. In India, the current settlement model is T+1, which means that if you purchase a security today, it settles tomorrow. The US, on the other hand, operates on a two-day settlement cycle. Broadridge intends to address this challenge by integrating generative AI into its operations. OpsGPT aims to simplify and optimise trading operations, providing transparency and risk management, while automation reduces manual intervention. Trained on data from Broadridge’s global post-trade systems handling $10 trillion in daily trades, the model streamlines data access, connectivity, and understanding across asset classes, promoting operational productivity. BondGPT In June 2023, the company launched BondGPT, an application built on OpenAI’s GPT-4, to assist in identifying corporate bonds on the LTX platform. It integrates LTX Liquidity Cloud for real-time information, aiding asset managers, hedge funds, and dealers in bond selection and portfolio management. BondGPT offers a chat feature using a large language model for conversational interaction, allowing users to find corporate bonds based on specific criteria. It uses LTX’s technology to match bonds with similar characteristics for liquidity purposes. Additionally, BondGPT+ was released, enhancing the service with client and third-party data integration, personalised dashboards, and scheduled queries. It prioritises accuracy and security, featuring an Admin Center for data management and compliance, and includes AI-driven usage analysis for system optimisation. Having released three AI products in the fintech space unlike any other company, Broadridge considers cutting edge technology to be one of the pillars of its success. “Unlike traditional BPO providers, we not only provide operational support but complement it with an in-depth understanding of how to leverage technology. We are essentially a tech-led operations company,” said Thomas Giacalone, SVP, global head of Broadridge Business Processing Outsourcing in an interaction with AIM. The operational team identifies opportunities for clients to enhance efficiency, manage risks, and improve client service through AI. India Expansion Plans Broadridge currently offers technology-driven solutions for financial institutions, including banks, broker-dealers, asset managers, and corporations. The fintech giant’s strategy for the next six to ten years will focus not just on generative AI, but its other emerging forms as well since it has garnered immense interest from clients, including major broker-dealers and boutique asset managers to provide the best tech-enabled solutions to them. The company has expanded its India operations with offices in Bengaluru and Hyderabad, where it operates on a ‘follow the sun’ model in business operations, with the country being its centre of expertise for processing. Ohrie said that the company’s centre of excellence is in India, which focuses on the four pillars of technology — AI & analytics, blockchain, cloud, and digital (ABCD) – with specialised agile squads dedicated to each of these areas. India played a huge role in bringing these generative AI products to life. “The talent and innovative ideas here in India, energise me to work better,” beamed Giacalone and added that India serves as the company’s second-largest hub outside the US. “In the next few years, we expect India to be the epicentre for next-generation operations role in our growth and client base expansion,” concluded Giacalone.","excerpt":"OpsGPT aims to simplify and optimise trading operations, providing transparency and risk management, while automation reduces manual intervention.","categories":["AI Features"],"tags":["Interviews and Discussions"],"author_name":"Shritama Saha","publish_date":"2024-02-05T14:00:00","publication_year":"2024","word_count":665,"keywords":["Go","OpenAI","AI","ML","Git","RAG","Aim","generative AI","analytics","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","ML","analytics","generative AI","OpenAI","Aim","RAG","R","Go","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-broadridges-opsgpt-is-redefining-trading\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":67481,"title":"How Laying Off Data Scientists Amid COVID-19 Pandemic Can Hamper Business’s Future Success","content":"This unprecedented time has brought in a significant economic crunch for businesses, and therefore almost all companies are undergoing layoffs in order to cut costs as well as stabilise their finances. These layoffs have created a record number of 1.5 lakh unemployed IT professionals, with companies like IBM, Uber, Kabbage, Cognizant, announced firing thousands of their analytics employees during this crisis. These numbers highlight that no job roles, including data science positions, are entirely immune to this economic downturn. Although these layoffs and downsizing of companies is entirely justifiable amid this crisis, this indeed brings in a ripple effect for businesses that are undergoing such significant changes. Laying off data scientists might seem like a menial change in the IT departments of organisations, but it induces several effects in which businesses operate. And, it definitely creates a more significant challenge, if these layoffs happen on a large scale. In fact, according to a news report, the Ministry of Labour and Employment has recently sent guidelines to the chief secretaries of all states. It has clearly stated that these layoff of employees from the job or reducing their salaries amid this scenario would further worsen the crises and hamper the employee morale, which, in turn, can hamper organisations. Not only laying off data scientists can affect a company’s ongoing projects, but also impacts the remaining employees of the company, who go through a massive deterioration post a layoff. In this article, we will share a few reasons why laying off data scientists amid this pandemic crisis can hamper the business’s future success. Also Read: What May Be The Factors Behind Data Science Layoffs Losing innovation edge Although this crisis is bringing in unemployment for many, it is also yielding a major automation trend among businesses, where leaders are relying on automated tools, without manual intervention, to make informed decisions. This automation boom will bring in the demand for more data scientists and analytics professionals who can build these automated solutions for businesses, and therefore losing key data scientists amid this uncertain time would result in losing the innovation edge for the company. Experts believe, as companies realise the potential of harnessing data for business productivity, data scientists and data engineers will witness a surge in demand in the post-COVID world. Not only data science skills are utilised to tackle this deadly pandemic but also used by several businesses to sustain their business post the pandemic. Whether to identify targeted audiences or to enhance customer experience, companies heavily rely on leveraging data science skills to have the edge over their competitors. In fact, according to a recent news report, Milind Apte, head, HR of CEAT stated that the company is going full-blown with automation and digitising their processes amid this crisis, and due to that, the company will be hiring data scientists as well as engineers. Alongside, in another interview, Chennai-based analytics company — Tiger Analytics — announced that the company is looking to hire over 150 data scientists despite this crisis. The company sees a surge in its business using data to make decisions, “The current situation is unprecedented, and decisions need to be made based on access to real-time data. Once the data engine kicks in and value is seen, companies will continue to invest further in this space.” Also Read: Ways Analytics Companies Can Avoid Layoffs AMID COVID-19 Losing critical investment Businesses usually invest a lot of effort in training their data scientists, providing necessary tools, as well as aligning their work according to the company’s vision, and laying off these professionals amid this crisis would result in losing critical investment for businesses. Although the majority of these layoffs are to reduce costs for companies, laying off data scientists can actually incur more charges for companies, as they might have to train new recruits or re-skill existing employees to fill in those laid off gaps. Having a data scientist is an expensive aspect for a company, hiring new recruits can turn out to be even more dangerous for a company’s finances. This can also be termed as the loss of skills and knowledge that has been built over time with a lot of efforts from business leaders. Also, new recruits can never take the place of the experienced data scientists who have been handling core business problems for a long time. Alongside, the termination process comes with other financial issues like arranging severance pay and related laid-off costs. Additionally, a majority of these data scientists in organisations handle critical data, losing these professionals would also result in losing those skills, knowledge and sensitive information that is vital for businesses to sustain amid this crisis. Besides, it also increases the burden of work for the surviving employees of organisations, which in turn increases anxiety and stress for those employees hampering productivity. Laying off data scientists can save costs for businesses but indeed takes a toll on employees as well as for companies. Also Read: What Is The Right Way To Interview Data Scientists During The Hiring Process Hamper remaining employees’ productivity Laying off employees can lower down the morale of surviving employees, especially for those whose close colleagues have been recently laid-off. Data science has been a close-knit community, and laying off employees can have adverse impacts on the remaining employees, which in turn can hamper their productivity. According to a study, it has been noted that more than 70% of the employees that have survived organisations’ downsize amid this pandemic would witness a massive decrease in their work productivity, which, in turn, can deteriorate companies’ outcome. Although these terminations are inevitable, it surely builds in insecurity among the surviving data professionals about their future prospects in companies. This furloughing-related anxiety also adds upon the extra responsibility analytics professionals have to take amid reduced staff, which again can impact the bottom line of companies. A massive downsizing in a company can create turmoil among the surviving as well as can significantly hamper their motivation to work which again is vital for businesses to process efficiently. Besides, massive layoffs can also reduce employee trust in organisations and their leaders, which will force them to look out for external opportunities, thereby hampering the image of companies. Also Read: How Analytics Companies Can Encourage Their Surviving Employees After A Downsize Decrease corporate image amid lockdown Although massive layoffs amid this crisis can indeed help companies in stabilising their finances as well as can satisfy stockholders and business owners, laying off data scientists can indeed create a negative image in the market, as the majority of the companies and pledging no layoffs in order to sympathise with employees. Having the right corporate image in this uncertain time is critical for businesses to sustain in the post-COVID world and retain their customers. Having a negative image during this pandemic has potential to stick around for a long time and can hamper a company’s ability to build its position in the industry as well as to recruit good data science talent in future. In fact, in recent news, many tech companies have pledged no layoffs amid this crisis in the act of kindness. Marc Benioff, the CEO of Salesforce, had also tweeted a statement to his employees pledging no significant layoffs amid this crisis. Benioff also urged his top executives to do their part by helping workers keep their jobs during this tumultuous time. In his Twitter thread, Benioff also wrote, “We will continue to pay our hourly workers while our offices are closed.” Additionally, laying off people amid this crisis can also showcase that the company is having a shortage in funds and can reflect poorly on customers and their loyalty, which can lead to providing a competitive edge to their competitors.","excerpt":"This unprecedented time has brought in a significant economic crunch for businesses, and therefore almost all companies are undergoing layoffs in order to cut costs as well as stabilise their finances. These layoffs have created a record number of 1.5 lakh unemployed IT professionals, with companies like IBM, Uber, Kabbage, Cognizant, announced firing thousands of […]","categories":["AI Features"],"tags":["automated data science solutions","covid-19","Data Science","data science layoff","trust your data"],"author_name":"Sejuti Das","publish_date":"2020-06-17T10:10:03","publication_year":"2020","word_count":1287,"keywords":["data science","Go","data science layoff","covid-19","AI","trust your data","Git","RAG","automated data science solutions","ViT","analytics","Rust","GAN","Data Science","R"],"extracted_tech_keywords":["AI","data science","analytics","RAG","R","Go","Rust","Git","GAN","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-laying-off-data-scientists-amid-covid-19-pandemic-can-hamper-businesss-future-success\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":62108,"title":"IIT Kharagpur Allocates Funding For Research Projects On COVID-19 Healthcare Technology Development","content":"IIT Kharagpur has decided to allocate research funding for R&D work related to COVID-19. IIT Kharagpur submitted a list of projects to the IIT Council last week of which eight projects have been selected to be funded on COVID-19 healthcare technology development. In a recent tweet, the institution has confirmed the news by stating — “@IITKgp Allocates Funding for 8 Research Projects on #COVID19 Healthcare Technology Development. Recently DrRPNishank, Hon’ble Minister, @HRDMinistry congratulated Director @tewari_virendra & the researchers.” https:\/\/twitter.com\/IITKgp\/status\/1251862368755175424 Dr Ramesh Pokhriyal Nishank, Hon’ble Minister, Ministry of Human Resource Development, Govt. of India has appreciated the initiative by IIT Kharagpur on his social media, congratulating the Director Prof. Virendra Kumar Tewari and his team of researchers. When asked about the initiative, Director Tewari said “It is our responsibility to improve the quality of life of the last person in the society. While we built some quick technologies to cater to the immediate needs of the essential service providers at the campus, we were simultaneously preparing project proposals and evaluating them keeping in mind the immediate need of the country, cost and product delivery period.” According to the IIT Kharagpur release, the researchers would be working on several technologies for these projects. These technologies include design and development of the rapid diagnostic kit, real-time PCR machine, bodysuit for COVID-19 infected patients, personal protective equipment (PPE) for healthcare workers. Some others include portable shredder integrated with steriliser, Hazmat Suit with cooling air circulation for medical professionals, bootstrapping Ambu-bag as an automatic ventilator, telemedicine for fighting the deadly pandemic and production of recombinant proteins for vaccine and testing. And for all these eight projects, an amount of Rs. 50 Lakh has been allotted for phase I, towards the development of prototypes. The release stated that the prototypes of these projects are expected to be ready within a duration of three to four weeks. However, some of them might take about six months to deliver the results. The institute stated that they are going to start working on phase one immediately after the lockdown is relaxed and as soon as the research staff can attend the laboratories. Meanwhile, software-related work would progress as usual. “IIT Kharagpur has a proven track record towards the development of indigenous health and hygiene technologies which are affordable, high-quality at par with globally accepted standards, and commercially viable. Our researchers are committed to delivering the prototypes within a constrained timeline considering the healthcare needs in the current situation,” concluded Director Tewari.","excerpt":"IIT Kharagpur has decided to allocate research funding for R&D work related to COVID-19.  IIT Kharagpur submitted a list of projects to the IIT Council last week of which eight projects have been selected to be funded on COVID-19 healthcare technology development. In a recent tweet, the institution has confirmed the news by stating — […]","categories":["AI News"],"tags":["Coronavirus","covid-19","IIT","iit Kharagpur","Software Development"],"author_name":"Sejuti Das","publish_date":"2020-04-20T11:08:01","publication_year":"2020","word_count":412,"keywords":["Go","API","funding","covid-19","AI","programming_languages:R","Software Development","programming_languages:Go","RAG","Coronavirus","iit Kharagpur","IIT","R"],"extracted_tech_keywords":["AI","RAG","R","Go","API","funding","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/iit-kharagpur-allocates-funding-for-research-projects-on-covid-19-healthcare-technology-development\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10057390,"title":"School students use cloud computing and AI to tackle real-world problems","content":"From creating concepts for ‘Intelligent Bus Lane Systems (IBLS),’which alerts motorists about priority vehicles including ambulances to AI-powered wearables and Krishi Seva app, an AI-enabled application for smart farming; these are some of the winning creative concepts developed using AI-model and cloud computing by ten school students across the country. These ten students have won the Amazon Web Services AWS Young Builders Challenge 2021. The challenge is an initiative to promote and foster a scientific temper, computational and design thinking, and coding skills among school students. About 2938 schools across 28 states and six union territories across India submitted 5952 projects. Individual students from these top 10 schools showcased various innovative ideas based on cloud computing and AI, designed to address real issues in today’s world, such as an app for water conservation through smart irrigation techniques, an AI-enabled application for smart farming, an intelligent bus lane system, a smart energy application, and a ‘diagnoser robot’ that identifies health conditions. The participants were equipped with a rich selection of learning modules from Atal Innovation Mission and AWS and tutorials from Code.org to understand the fundamentals of cloud computing and AI and build their ideas. Launched with support from the Ministry of Education – Innovation Cell (Govt. of India); Atal Innovation Mission, NITI Aayog; Central Board of Secondary Education (CBSE); and Code.org, the initiative aims to introduce young minds to the basics of disruptive technologies such as cloud computing, ML, AI and inspire them to create a positive impact in India using these technologies.","excerpt":"AWS Young Builders Challenge 2021 promotes scientific temper, computational, design thinking, and coding skills among school students.","categories":["AI News"],"tags":["Cloud Computing","coding","skills","students"],"author_name":"Poornima Nataraj","publish_date":"2021-12-29T17:05:15","publication_year":"2021","word_count":253,"keywords":["Go","AWS","AI","cloud computing","innovation","coding","ML","cloud_platforms:AWS","students","Aim","Cloud Computing","skills","cloud_platforms:Amazon Web Services","R"],"extracted_tech_keywords":["AI","ML","Aim","cloud computing","AWS","R","Go","innovation","cloud_platforms:AWS","cloud_platforms:Amazon Web Services"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/school-students-use-cloud-computing-and-ai-to-tackle-real-world-problems\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10094514,"title":"Apple WWDC 2023 Key Highlights","content":"Move over the mind-blowing inventions of the past, because Apple just raised the bar at the WWDC 2023, leaving attendees in awe and sceptics scratching their heads. This year’s event was a playground of cutting-edge technology, where machine learning, transformer models, and virtual reality collided to create a spectacle that had jaws dropping faster than a neural network can crunch numbers. Apple’s daring foray into the realm of the extraordinary has left us all wondering, “Are we still in reality, or did we just slip into a parallel universe of mind-bending possibilities?” Apple Vision Pro “Mac introduced us to personal computing and iPhone introduced us to mobile computing, Apple vision Pro will introduce us to spatial computing,” said Tim Cooke. Meet Apple Vision Pro, a groundbreaking computer that seamlessly merges the real and digital worlds through augmented reality. With Vision Pro, you can interact with digital content as if it exists in your physical space, using your eyes, hands, and voice as natural and intuitive tools. It revolutionises the way you perceive your surroundings, transforming them into an infinite canvas where apps can be used anywhere and resized at will. Capture precious moments in a whole new way with photos and videos. Immerse yourself in movies, shows, sports, and games on a large virtual screen with immersive audio. Connect with others as if you’re sharing the same physical space. Vision Pro brings digital content to life around us, marking the dawn of spatial computing—a new era in personal technology. M2 Ultra Apple is expanding its M2 chip family by introducing the M2 Ultra chip, which pushes Mac performance to new heights. This powerhouse chip incorporates the groundbreaking Ultra fusion architecture and joins forces with the M2 Max die to deliver unparalleled performance. The M2 Ultra chip boasts a 24-core CPU, doubling the performance of the M2 Max. In terms of graphics capabilities, the M2 Ultra’s 76-core GPU achieves a 30% increase in speed compared to the M1 Ultra. Additionally, the 32-core neural engine is now 40% faster. The M2 Ultra chip sets new standards for memory capacity, supporting an impressive 192 gigabytes of unified memory—50% more than the previous M1 Ultra chip. This expanded memory capacity enables the M2 Ultra to handle massive machine learning workloads, including large transformer models, surpassing the limitations of other chips on the market. iOS 17 and Transformer-based Autocorrect Apple has unveiled its latest software update, iOS 17, which introduces significant enhancements to the keyboard intelligence, promising an improved typing experience for iPhone users. Autocorrect is now powered by on-device machine learning, with Apple continuously advancing the underlying models over the years. The keyboard now incorporates a SOTA transformer language model, greatly enhancing word prediction accuracy. With the power of Apple silicon, the iPhone can efficiently run this model with every keystroke, resulting in a more capable and reliable autocorrect system. “When you just want to type a ducking word, autocorrect will learn that.” Dictation, another integral part of the keyboard, has also received an upgrade. It now utilises a transformer-based speech recognition model, leveraging the power of Apple’s neural engine to deliver even more accurate dictation results. macOS Sonoma Apple has recently unveiled macOS 14, the latest iteration of its desktop operating system. Continuing the tradition of naming versions after California towns, this release is named Sonoma, after the picturesque wine country in Northern California. One of the notable features in macOS 14 is the introduction of widgets, a feature borrowed from the mobile side of Apple’s ecosystem. These widgets are designed to dynamically adjust to your usage, ensuring they don’t dominate the screen while you’re engaged in other tasks. Furthermore, these interactive widgets allow for seamless actions, such as selecting images directly from the widget window, enhancing the overall user experience. Gaming Mode and toolkit Apple is revolutionising the gaming experience on Mac with the introduction of Game Mode. This presents an incredible opportunity for developers to bring their games to a wider audience than ever before. The advancements in Metal 3, Apple’s graphics technology, have further enhanced gaming performance. Games like Resident Evil Village have seen significant gains in performance, thanks to Metal Effects and upscaling capabilities. Traditionally, evaluating game compatibility and optimising performance for Mac could take months. However, with the new game porting toolkit, developers can accomplish this in a matter of days. The toolkit streamlines the conversion of game shaders and graphics code, allowing developers to fully leverage the powerful features and performance of Apple silicon. As a result, the total development time is significantly reduced, empowering developers to bring their games to the Mac ecosystem more efficiently. Safari Updates Apple’s own browser is built from the ground up again to be powered by WebKit technology used by over a million apps. Apple prioritises privacy, and the latest update enhances private browsing by locking windows, blocking trackers, and removing tracking from URLs. New feature of  secure password and passkey sharing through iCloud Keychain. Profiles help separate work and personal browsing, while students can create unique collections. Additionally, Apple introduced web apps for faster access to sites. With simplified toolbars and seamless integration, web apps provide an app-like experience without extra developer work. Safari makes accessing favourite websites effortless on Mac. Airpods With Adaptive Audio, AirPods Pro takes personalised audio to the next level. By utilising machine learning, these earbuds can learn your listening preferences over time and adapt to the surrounding conditions to provide an optimal media experience. Whether you’re in a quiet room or a noisy environment, AirPods Pro will dynamically adjust the audio to deliver the best sound quality and clarity for your specific situation. But that’s not all. Apple has introduced an exciting new feature that enhances your interactions with others while wearing AirPods Pro. Conversation Awareness automatically detects when you start speaking, and in response, it intelligently lowers the volume of your music or audio playback. This allows you to focus on the voices and conversations happening around you, without the need to manually adjust your audio settings. Apple Health App In a not-so-surprising announcement, Apple introduced a new Health app for the iPad. The new app will launch as part of iPadOS 17 later this year. With this, users will be able to see their health data, such as electrocardiogram results, on a larger screen. This new feature gives users the access to users’ health data securely, so that they can see their health information across Apple ecosystem of products, iPad, iPhone, Apple Watch, etc, alongside compatible third party apps and devices all in one place. For developers, HealthKit is coming to iPad, unleashing new ways to build health and fitness experiences for the devices. So far, Health app was available on the iPhone, now it is available on iPad, which gives users more control when it comes to viewing health metrics, perceptions, lab tests and more. Apple Watch Apple just dropped another update to their WatchOS with WatchOS 10. This builds on the focus of privacy and health that is one of the key features of the Apple conference this year. Apart from adding more widget to the screen, the company has integrated Mindfulness into the software. This enables the watch to track your mood and logging features. Moreover, the company has decided to put more focus on mental health this time. In an interesting note, the new update to the WatchOS also enables users to measure their screen time on the watch by tracking their eyes. Journal App In addition to these, Apple has announced an upcoming app called Journal, set to debut on iPhones later this year. The Journal app aims to help users reflect on and relive their special moments by providing personalised suggestions based on various aspects of their day, such as photos, locations, music, and workouts.","excerpt":"Apple just made huge developments in the AR industry, upgrades to the M2 chips and many more advancements in machine learning","categories":["Global Tech"],"tags":[],"author_name":"Mohit Pandey","publish_date":"2023-06-06T00:42:27","publication_year":"2023","word_count":1298,"keywords":["machine learning","AWS","AI","neural network","ML","Git","RAG","Ray","Aim","R"],"extracted_tech_keywords":["AI","machine learning","ML","neural network","Aim","Ray","RAG","AWS","R","Git"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/apple-wwdc-2023-key-highlights\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10160832,"title":"NVIDIA Unveils New Llama Nemotron Models to Build AI Agents","content":"At CES 2025, NVIDIA CEO Jensen Huang launched new Nemotron models, including the Llama Nemotron large language models (LLMs) and Cosmos Nemotron vision language models (VLMs), to improve agentic AI and boost enterprise productivity. The Llama Nemotron models, built on Llama foundation models, allow developers to create AI agents for applications like customer support, fraud detection, and supply chain optimisation. “Llama 3.1 is a complete phenomenon, with the downloads reaching 650,000 times. It has been derived and turned into other models, about 60,000 different models. It is singularly the reason why every single enterprise and every single industry has been activated to start working on AI,” said Huang. “We realized that the Llama models could really be better fine-tuned for enterprise use, so we fine-tuned them using our expertise and capabilities and turned them into the Llama Nemotron suite of open models,” he added. The Nemotron families will be offered in Nano, Super, and Ultra sizes to suit deployment needs, from low-latency real-time applications to high-accuracy data center use cases. Optimised for computing efficiency and accuracy, these models support agentic AI tasks like instructions for following, coding, and math. “Agentic AI is the next frontier of AI development, and delivering on this opportunity requires full-stack optimization across a system of LLMs to deliver efficient, accurate AI agents,” said Ahmad Al-Dahle, vice president and head of GenAI at Meta. NVIDIA announced that the models will be available as downloadable resources or as microservices for deployment across various computing platforms, including data centers and edge devices. Llama Nemotron and Cosmos Nemotron models will be available soon on build.nvidia.com, Hugging Face, and through the NVIDIA Developer Program. Enterprise-grade deployments will be supported via the NVIDIA AI Enterprise platform on accelerated cloud and data center infrastructure. NVIDIA’s Cosmos Nemotron models extend AI capabilities to vision and video tasks, allowing agents to analyse and respond to images and videos. These tools aim to support industries like autonomous systems, healthcare, retail, and media. NVIDIA also unveiled Cosmos world foundation models for physics-aware video generation in robotics and autonomous vehicle applications. NVIDIA NeMo microservices allow enterprises to customise these models for specific domains and workflows. Leading AI platform providers, such as SAP and ServiceNow, have backed the Nemotron models. SAP plans to incorporate them into its Joule platform to improve enterprise user productivity, while ServiceNow seeks to utilise the models for AI agent services across various industries. The models are built using NVIDIA’s NeMo platform for distillation, pruning, and alignment, ensuring high accuracy and throughput across various hardware configurations. NVIDIA NeMo Retriever allows integration with enterprise data, boosting model functionality through retrieval-augmented generation capabilities.","excerpt":"The Nemotron families will be offered in Nano, Super, and Ultra sizes to suit deployment needs, from low-latency real-time applications to high-accuracy data center use cases.","categories":["AI News"],"tags":["NVIDIA"],"author_name":"Siddharth Jindal","publish_date":"2025-01-07T10:55:20","publication_year":"2025","word_count":437,"keywords":["Hugging Face","GenAI","agentic AI","AI","microservices","Aim","ViT","foundation models","NVIDIA","R","fraud detection"],"extracted_tech_keywords":["AI","GenAI","foundation models","agentic AI","Aim","Hugging Face","fraud detection","microservices","R","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/nvidia-unveils-new-llama-nemotron-models-to-build-ai-agents\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":17811,"title":"In conversation with Shankar Viswanathan, Partner At ZS Associates","content":"ZS is one of the world’s largest firms focused exclusively on helping companies improve overall performance and grow revenue and market share, through end-to-end sales and marketing solutions—from customer insights and strategy to analytics, operations and technology. More than 5,000 ZS professionals in 22 offices worldwide draw on deep industry and domain expertise to deliver impact for clients across multiple industries. With close to 3,000 professionals, ZS offices in Delhi and Pune serve as the Capability and Expertise Center to enable cross-collaboration across all the global offices to create client impact. Analytics India Magazine got onto an exclusive conversation with Shankar Viswanathan, a partner with ZS based out of their India Capability and Expertise Center to get insights on the company as well as the analytics landscape of India. Here is the exclusive interview with Shankar Viswanathan. Analytics India Magazine: Shankar, could you talk about ZS Associates’ journey so far in an analytics dedicated landscape? Also, how is the firm contributing uniquely to this space? Shankar Viswanathan: The genesis of ZS in 1983 was based on using linear and non-linear programming algorithms to optimise sales force decisions. Data and analytics have been at the core of our professional services ever since, to enable business decisions and actions. With analytics, we help our clients across a range of business decisions and actions including product development choices, market opportunity assessment, business development and licensing, brand launch maximisation, customer segmentation, targeting, marketing mix, sales resource allocation, and customer acquisition and retention program optimisation. We have highly skilled decision analytics and data science professionals with deep knowledge of customer data and analytical methods that span foundational statistics to machine learning techniques, and who are proficient in a range of analytical tools and programming paradigms. These professionals, however, don’t just run complex mathematical models, but have a strong understanding of the business context. Our professionals strive to provide relevant insights and implementable recommendations recognising this context. This combination of strong analytical skills, business knowledge and focus on providing implementable solutions are key differentiators for ZS. AIM: How do you leverage data analytics in understanding the buying patterns of customers across different channels and driving business? How does it benefit your clients? SV: There has been an ongoing buzz around customer-centricity. We are working with several clients to make this a reality. One of the key challenges is that our clients are often tactic-led in their organisation and decisions, for example, email, events, website, direct mail, etc. This, however, results in a poor experience for their end-customers who are bombarded across channels, especially their most important high-value customers. We are leveraging data and analytics to help our clients become more customer-centric in a few ways such as: understanding the fundamental preference of customers to different channels and content,understanding how channels interact to create impact for customers,leveraging the above customer insights to create a harmonised and integrated customer engagement plan, andmonitoring the customer behaviours and dynamically adapting the execution. Our clients are seeing meaningful benefits in terms of both leading indicators such as increased customer engagement and eventual sales uptick. AIM: What brought you to the analytics landscape? Also, talk about your experience in the field. SV: I completed my bachelor’s degree in engineering from IIT Madras and a PhD from Purdue. While I was pursuing PhD, I was focused on using artificial intelligence, machine learning and optimisation techniques to solve dynamic chemical process systems. I was intrigued by the possibilities of these techniques and got an opportunity to apply them in a different context at ZS – sales and marketing problems. Over the years, I have enjoyed the different aspects of analytics – understanding the core business problem, structuring the problem and framing it analytically, exploring the underlying data, setting up the mathematical models, deriving insights, synthesising recommendations and developing the business story to effectively communicate the recommendations to our clients. My focus at ZS has been in the healthcare analytics space. I have been fortunate to work with a diverse array of clients – large life science organisations, mid-size pharma, emerging biotechnology clients and medical device manufacturers. The spectrum of problems that my teams and I have worked on have been equally varied – sales force analytics for large-scale merger; agile analytics at product launches; optimising mix across digital, personnel and non-personnel promotions; analysing large-scale patient datasets to inform market opportunity; and enabling contracting and rebating decisions in the complex payer–provider–patient ecosystem. In my career, I have seen the evolution of analytics from being a support function to getting embedded in the business. Analytics and data are now poised to drive business transformation for not only the digital disruptors but also a broad spectrum of industries and organisations. AIM: Please present a picture of the analytics sector in India. What are the typical challenges noticed in the space? SV: India is a thriving raw analytics talent marketplace. Effectively combining the analytical talent and our technological prowess can make India a potent global force in creating business value using data and analytics. In a survey we had jointly conducted with The Economist, these two came up with the key broken links in the analytics value chain – problem definition\/scoping and an organisation’s ability to effectively consume the outputs from analytics consumption. We are observing that our clients are motivated to elevate the role that analytics plays in their organisation. They, however, don’t always have a clear roadmap for this. They invest in a range of capabilities such as data, data scientists, expensive infrastructure and technologies. Organisations are, however, less clear about what problems they are looking to solve and how to translate the results of analytics in a way that the organisation can effectively consume them. AIM: What are some of the most recent technological developments in the analytics space? How is ZS Associates incorporating those technologies into its own ecosystem? How is Big Data driving business efficiencies? SV: The technological advancements span a range of services, from foundational infrastructure, data and analytics to front-end business applications. On the infrastructure services front, we are purposefully migrating our clients to cloud computing and helping them realise their business value. On the data services front, we are helping our clients implement Enterprise data lakes, data catalogues and data enrichment solutions. We are also actively working with clients to mine unstructured data sets such as text, audio, and video to derive customer insights that can inform business decisions. On the analytical services front, there is a move toward democratisation of analytics that we are facilitating using analytical workbenches. Our decision analytics and data science professionals are leveraging a range of integrated development environments or IDEs and newer programming paradigms. The cloud infrastructure, Big Data technologies, data cataloguing solutions and analytical workbenches are foundational elements of the technology stack and here our focus is not to reinvent the wheel. Our focus is to leverage and configure the horizontal offerings in the marketplace. We, however, are overlaying these with domain-specific intelligent business services and applications. Our experience and expertise in working across a range of sales and marketing issues enable us to build the right business solutions. We are embedding these solutions in the ZS platform – REVO. The REVO platform is leveraged both by our professionals and our client stakeholders for agile, efficient, and effective analytics. AIM: Talk about the work culture at ZS Associates. Any piece of advice for startups and entrepreneurs who want to make it big in the analytics sector? SV: ZSers pride themselves in the quality of our work. This is a key focus of all our engagements and is deeply embedded within the culture of the company as one of its core values – “Get it right, always.” A high degree of responsiveness to our client’s needs is another hallmark of the ZS culture. Our founders are scholars and we have retained the collegial feel to our organisation. Collaboration is not only recognised but also encouraged across all levels of the firm. We want to bring the best of ZS to our clients and collaboration is key to achieving this objective. One of my favourite books is Haroun and the Sea of Stories. The gupees and chupees are going to war. The gupees before the war have intense discussions and debates and seem highly disorganised. The chupees follow the hierarchy and appear disciplined. The gupees, however, win the war as they are truly working towards a unified cause. “Be like the gupees and not the chupees.” In addition to a good business idea, collaborating with the right people who are passionate about the idea and willing to challenge and push the boundary is critical. AIM: How can one make a career at ZS? SV: Bringing the right Attitude, passion to learn the skills, and curiosity and thirst for Knowledge is the key ASK. On the attitude front, we want people who are driven to help our clients, who get energised in a team setting, and who are authentic and don’t pretend to know it all. On the skills front, we want people to have a genuine interest in being skilled craftsmen and not a dabbler. This requires deliberate practice and persistence. Imagine doing a series of master’s programs throughout your career at ZS and not just doing a series of bachelor’s courses. On the knowledge front, our differentiator is people who are steeped in the business domain. We believe the combination of high-end skills with deep knowledge is critical to providing the right solutions and recommendations to our clients.","excerpt":"ZS is one of the world’s largest firms focused exclusively on helping companies improve overall performance and grow revenue and market share, through end-to-end sales and marketing solutions—from customer insights and strategy to analytics, operations and technology. More than 5,000 ZS professionals in 22 offices worldwide draw on deep industry and domain expertise to deliver […]","categories":["AI Features"],"tags":["analytics in india","Big Data","Interviews and Discussions"],"author_name":"Priya Singh","publish_date":"2017-09-19T06:44:55","publication_year":"2017","word_count":1584,"keywords":["data science","artificial intelligence","machine learning","analytics in india","AI","cloud computing","TPU","RAG","Ray","Aim","analytics","Big Data","Interviews and Discussions"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","data science","analytics","Aim","Ray","RAG","cloud computing","TPU"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/conversation-shankar-viswanathan-partner-zs-associates\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10044502,"title":"fast.ai Introduces fastchan For Easier &#038; Faster Installation Of Libraries","content":"fast.ai has announced fastchan, a new conda mini-distribution. Centred around the PyTorch ecosystem, developers can use fastchan to install and update libraries in a faster, easier, and more reliable way. Fastchan allows Anaconda users to install Python softwares such as fastai, RAPIDS, OpenCV, Hugging Face Transformers, and timms with just one unified command: conda install -c fastchan. The same approach can be applied for software upgradation. Fast.ai, founded by Jeremy Howard in 2016, is a non-profit research group focused on deep learning and AI with a stated objective to democratise deep learning. One of its most popular innovations is fastai, an open-source deep-learning library. It provides components to deliver state-of-the-art results in standard deep learning domains. fast.ai allows practitioners to experiment, mix and match to discover new approaches. In short, to facilitate hassle-free deep learning solutions. The libraries leverage the dynamism of the underlying Python language and the flexibility of the PyTorch library. The distribution problem Anaconda is a free and open-source distribution of programming languages such as Python and R. This distribution comes with a Python interpreter and other machine learning and data science packages. Anaconda or Conda can also create a separate self-contained environment with isolated software installations. Meaning a developer can create a quick throwaway environment to test new software without breaking the base environment or maintain separate environments for different projects requiring different versions of Pythons or libraries. Anaconda regularly releases new versions of its main installer along with other packages. Anaconda also tests new versions of packages between software releases and adds them to the default channel. Most of these packages in the default channel are sourced from conda-forge, a repository for users to upload procedures to build software. Anaconda takes a subset of these packages, along with software, does an additional integration testing of these packages and makes them available in their distribution. While this works in many situations, many Python libraries are not in conda-forge channels or any channels at all; especially, libraries that use GPU as conda-forge do not have the facility for building and testing GPU-enabled software. fastchan To overcome these challenges, fast.ai has created a new channel and distribution called fastchan. It contains all the dependencies to install fastai, PyTorch, RAPIDS, etc. For fastchan, the team has used the official PyTorch build of PyTorch, the official NVIDIA build of RAPIDS, and CUDA toolkit to avoid packaging the software from scratch. Libraries and dependencies that are available only on conda-forge can be copied into the fastchan channel using an Anaconda command called copy. Today we're announcing fastchan, a new conda mini-distribution with a focus on the PyTorch ecosystem. Using fastchan, installation and updates of libraries such as @PyTorch, @huggingface and @RAPIDSai is faster, easier, and more reliablehttps:\/\/t.co\/UMNkQGi3XJ— Jeremy Howard (@jeremyphoward) July 16, 2021 For fastchan, the fast.ai team included only those dependencies that are not already available in the default channel. They also packaged softwares that are only available as pip packages as pypi using setuptools-conda and build.py program (written from scratch by the fast.ai team). With fastchan, developers can rely just on the defaults and fastchan channels to install nearly every package and library, especially when working with software in the PyTorch and Hugging Face ecosystems. What next “I hope that fastchan will be a useful starting point for folks thinking about Python packaging and deployment,” said Jeremy Howard. Soon, the team hopes to add many features, including running integration tests on both CPU and GPU to ensure the code that uses libraries together gives the expected results. Howard has also invited developers to add their own integration tests. “Integration tests are crucial to ensure that no-one adds or changes a package which causes breakage on dependent packages (or at least to ensure that broken downstream packages are clearly marked as such),” he said. According to Howard, while fastchan was created primarily for the use of fast.ai, he hopes that in future, key players like PyTorch, NVIDIA, Anaconda, and conda-forge will solve the distribution problem together and make fastchan obsolete.","excerpt":"fast.ai is a non-profit research group focused on deep learning and AI founded in 2016.","categories":["AI Features"],"tags":["FastAI","Python Libraries"],"author_name":"Shraddha Goled","publish_date":"2021-07-26T17:00:00","publication_year":"2021","word_count":670,"keywords":["data science","Hugging Face","Rapids","machine learning","AI","Python Libraries","PyTorch","Transformers","OpenCV","RAG","deep learning","FastAI"],"extracted_tech_keywords":["AI","machine learning","deep learning","data science","PyTorch","Hugging Face","Transformers","OpenCV","Rapids","RAG"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/fast-ai-introduces-fastchan-for-easier-faster-installation-of-libraries\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10171389,"title":"Amid US Restrictions, Chinese Company To Offer Free Full-Stack EDA Tools","content":"Shanghai-based UniVista Industrial Software Group recently announced that it now offers domestic users free access to electronic design automation (EDA) software. UniVista’s suite offers a range of domestically produced EDA tools, which include a simulator debugger, a design for testability (DFT) platform, and a printed circuit board (PCB) design platform. According to the company, some of these are the country’s first domestically built tools. These EDA tools are software applications that assist engineers in designing, simulating, and verifying systems such as integrated circuits (ICs), PCBs, and system on a chip (SoC). EDA tools are often considered essential in the chip design process for various applications. The company called the announcement a “technological counterattack” and indicated that it is responding to numerous blockades and restrictions imposed by the United States government. Recently, Reuters reported that the Donald Trump administration has told US-based companies to stop selling software used to design chips to Chinese customers. Leading companies like Cadence, Synopsys, and Siemens EDA were notified. The three companies also account for 70% of China’s EDA market. Furthermore, it was reported that Synopsys has told its staff in China to halt sales and stop taking new orders. When ever there is huge demand, there will be service. New Chinese EDA players will pop up like mushrooms after a rain. Key Chinese EDA Companies:Empyrean: A leading domestic EDA provider specializing in analog\/mixed-signal circuit verification tools.Xpeedic: Focuses on…— Petri Kuittinen (@KuittinenPetri) May 31, 2025 UniVista, founded in 2020, comprises multiple former executives of Synopsys and Cadence, several of whom held senior VP and fellow-level positions, reported DigiTimes Asia. Besides UniVista, prominent players in China’s EDA ecosystem include Empyrean Technology, Primarius Technologies and Semitronix. As of last year, Empyrean held 6% of China’s market, making it the country’s largest domestic player.","excerpt":"The company was founded by former executives from Synopsys and Cadence—the two US-based companies reportedly restricted from selling EDA tools in China.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","China"],"author_name":"Supreeth Koundinya","publish_date":"2025-06-05T13:06:01","publication_year":"2025","word_count":295,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","Git","automation","AI (Artificial Intelligence)","R","China"],"extracted_tech_keywords":["AI","R","Go","Git","automation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/amid-us-restrictions-chinese-company-to-offer-free-full-stack-eda-tools\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10129408,"title":"Google’s Gemini Fuels Innovation for Karya, Miko, and Other Indian GenAI Startups","content":"At Google I\/O Connect Bengaluru 2024, Google shared how Gemini is being used by three Indian startups, Miko AI, Karya, and Cropin. Google’s ongoing support for Indian developers includes the Build with AI event series, which has engaged 25,000 developers in 43 cities, and the Global Gemini API Developer Competition. This sentiment was further reinforced by Chen Goldberg, GM and VP of Engineering, Google, when she told AIM, “This is my first visit to India, and I was very impressed with the level of innovation and new ideas, especially from startups that have been able to grow. Our customers in India are critical for us. Part of my visit here involves meeting my team, and I’m excited about the growth and innovation happening locally.” Adding on to that, Seshu Ajjarapu, senior director at Google DeepMind, expressed confidence in India’s AI potential, stating, “Indian startups are going to lead the AI revolution”. What are These Startups Upto Cropin, launched Cropin Sage, a real-time, generative AI agri-intelligence platform. Leveraging Google’s AI and machine learning capabilities, Cropin Sage is building a proprietary grid-based map that will provide data and intelligence on the world’s agricultural landscape with scale, accuracy and speed. “We integrate Gemini with our climate models, crop models, and proprietary knowledge graphs. These elements are fused together to provide comprehensive solutions,” Krishna Kumar, chief executive officer at Cropin, told AIM. On the other hand, Karya is on a mission to empower low-income communities through AI-enabled economic opportunities. Using Gemini, the organisation has designed a no-code chatbot that allows partners worldwide to design custom workflows in their own languages, understanding natural language prompts and generating configuration files instantly. Miko.AI has developed an AI-powered educational robot for children, leveraging 10 years of proprietary AI models focused on age-appropriate, culturally sensitive, and regionally relevant interactions. They use Google’s Gemini for quality control, ensuring a safe and engaging learning platform for children worldwide. Additionally, Google partnered with MeitY Startup Hub to train 10,000 startups in AI, offering mentorship, skills through Appscale Academy and Startup School, AI-first programming, and up to $350,000 in Google Cloud credits for AI development.","excerpt":"“This is my first visit to India, and I was very impressed with the level of innovation and new ideas, especially from startups that have been able to grow,” Chen Goldberg told AIM","categories":["AI News"],"tags":["Gemini","Karya"],"author_name":"Gopika Raj","publish_date":"2024-07-17T16:16:01","publication_year":"2024","word_count":352,"keywords":["knowledge graphs","Gemini","Go","machine learning","Karya","API","AI","RAG","Aim","generative AI","GAN","R"],"extracted_tech_keywords":["AI","machine learning","generative AI","Aim","RAG","knowledge graphs","R","Go","API","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/googles-gemini-fuels-innovation-for-karya-miko-and-other-indian-genai-startups\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10093648,"title":"Course5 Raises USD 28 Million To Invest In AI","content":"Course5 Intelligence, an analytics and AI solutions company, has successfully raised USD 28 million in a recent funding round out of the USD 55 million plan. The round was led by 360 ONE Asset Management Limited’s Tech Fund, formerly known as IIFL Asset Management Limited that invested USD 28 million. Course5 is supported by continuous innovation in their AI Labs and leverages global open research. The company’s platforms integrated with OpenAI’s GPT models enable data-driven insights for the businesses. The funds raised will be used to supplement a strong organic growth with inorganic expansion and synergistic acquisitions. The company also intends to increase AI investments and is currently in discussions with several M&A prospects to add strategic capabilities or intellectual property. The company aims to surpass $100 million in revenue in the next fiscal year and plans to launch its IPO within the next 18 months. Ashwin Mittal, Chairman and CEO of Course5 Intelligence, expressed his confidence in the company’s trajectory and the favorable industry trends, stating that raising external capital at this time is the right move. Mittal emphasized the strong demand for Course5’s analytics and AI solutions from both existing and new clients.. Chetan Naik, Fund Manager and Senior EVP at 360 ONE Asset, commented on the expected strong growth of the data analytics sector in the coming decade. He acknowledged Course5 as a leading player in the data analytics and insights field, with strong IP-led solutions and deep domain knowledge across various industries. Naik highlighted Course5 as one of the few profitable and highly capital-efficient pure-play data analytics companies based in India.","excerpt":"The company aims to surpass $100 million in revenue in the next fiscal year and plans to launch its IPO within the next 18 months.","categories":["AI News"],"tags":["Courses"],"author_name":"Tasmia Ansari","publish_date":"2023-05-19T18:12:17","publication_year":"2023","word_count":265,"keywords":["API","OpenAI","AI","RPA","RAG","GPT","Aim","analytics","GAN","Courses","R"],"extracted_tech_keywords":["AI","analytics","OpenAI","Aim","RAG","R","API","GPT","GAN","RPA"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/course5-raises-usd-28-million-to-invest-in-ai\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":47946,"title":"Is Your Data Lake Smart Enough To Handle Building Artificial Intelligence Models?","content":"Understanding data to make better decisions is one of the crucial tasks in an organisation. This is where a data lake comes into the picture. A data lake is a storage repository that holds a vast amount of raw data in its native format using a flat architecture until it is needed. To solve problems, enterprises then reach out to the data lake for relevant data, and that smaller set of data can then be analysed to help answer the question. Advantage Of Using A Data Lake There are several advantages of using a data lake such as: Helps gain insights flexibly Extract any format of data Agility in business No data silos Also, different types of analytics such as Big Data Analytics, Real-time analytics, SQL queries, etc. can be seamlessly used to gain insights on data. How To Build An AI-Friendly Data Lake In one of the talks in the recent-concluded event Cypher 2019, hosted by Analytics India Magazine, Nallan Sriraman, Global Head of Technology, Data and Analytics at Unilever, talked about the importance of data lake and how to hire the right data scientist in an organisation. He suggested the following steps in building an AI-friendly data lake: Protect the Data Lake From Becoming a Swamp: While collecting data in an organisation, there is a risk of creating a data swamp. Data swamp is similar to data lake but the difference is that data swamps are the unorganised versions of a data lake which makes it difficult for the organisations to extract the insights from data. This may happen due to various reasons such as an abundance of irrelevant data, lack of metadata, and other such. Establish Transparency: While building a model and algorithm in an organisation, it is important to create in such a way that it is debuggable. It must be clear and transparent in such a way that even if someone is not an expert, s\/he must be able to understand it. Similarly, along with building transparent algorithms, it is also important to establish transparency in data which helps in building the machine learning algorithms. Hiring Data Scientist without Tunnel Vision: While hiring a data scientist, it is important to have a look in a broader way rather than having a tunnel vision. Besides limiting the opinion, one should focus on other pools too. Why Do AI Predictions Go Wrong? AI techniques while giving us some groundbreaking results in innovative works, it also poses several risks. Even tech giants like Microsoft have seen the dark side of AI. One of the important measures which are responsible for the bias and failures in AI models is the lack of data coverage. Lack of data coverage can be considered as a major problem while trying to build a robust machine learning model. The behaviour of an algorithm depends on the data it is being fed. Due to the lack of availability of data, most of the time researchers use the data whichever is freely available. This data can help in building an AI model but it will not be robust and may have biases. Watch the complete session here:","excerpt":"Understanding data to make better decisions is one of the crucial tasks in an organisation. This is where a data lake comes into the picture. A data lake is a storage repository that holds a vast amount of raw data in its native format using a flat architecture until it is needed. To solve problems, […]","categories":["AI Features"],"tags":["Big Data","data lake"],"author_name":"Ambika Choudhury","publish_date":"2019-10-15T14:10:49","publication_year":"2019","word_count":522,"keywords":["big data","Go","machine learning","AI","ML","RAG","analytics","SQL","Big Data","R","data lake"],"extracted_tech_keywords":["AI","machine learning","ML","analytics","RAG","R","SQL","Go","big data","data lake"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/is-your-data-lake-smart-enough-to-handle-building-artificial-intelligence-models\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10005643,"title":"What Does It Take To Become A Healthcare Data Scientist","content":"Big Data in healthcare is not unheard of. The healthcare industry is generating a significant source of data in the world. People who have skill sets in health information technology and analytics, the employment rate for them is increasing. The global pandemic has dramatically increased the need as well as the importance of data science and analytics professionals. We have seen over the months how data science professionals have come together and worked on COVID-19 healthcare data and build AI\/ML models to track the outbreak and use the information for contact tracing, screening applications and vaccine development. With all this data and the right application of tools and methods, we can improve patient outcomes and reduce health care costs. This is precisely where a healthcare data scientist comes in. The intersection of healthcare and data science is the emerging area of healthcare research and operations. Analysts, researchers and data scientists in this field can really take advantage of this increasing demand for opportunities. Data science blends the unique skills of unlocking insights of data and storytelling. It can combine multiple skills from several disciplines to make it relevant for various roles, including healthcare. Talking about health data science, it refers to any data that pertains to the biomedical sciences and public health.  The data might originate from observational studies, clinical trials, computational biology, electronic medical records, genetic and genomic data. A data scientist can make use of the available genomic data and upon its analysis can provide an understanding of genetic issues and can also help in studying reactions to various kinds of drugs and its effect on diseases. How Does One Become A Data Scientist In Healthcare? Well, it is no different from being a regular data scientist but with a very high focus on healthcare data. Till date, the majority of data science courses and degrees focus on data science in general, and using the foundational skills, one can make a data career in healthcare, bioinformatics, medical\/genomic science, health economics and other associated fields. Health Data Science is a pretty new discipline, and it’s a blend of epidemiology, statistics, mathematics, informatics and computer science. There are specialised data science programs offered by universities, mostly at the postgraduate level. For example, there is a Masters In Health Data Science, an eighteen months post-grad program offered by Harvard University. MSc Health Data Science will help you develop innovative skills needed to unlock knowledge from complex health data to address some of the biggest health challenges that we face across the globe today. There is also a master’s program here at Georgetown in health informatics and data science. There are courses in India as well, such as Master of Health Data Science (SRM University), and Master of Science – Health Informatics & Analytics from The University of Trans-Disciplinary Health Sciences and Technology (TDU). The idea of such programmes is to create graduates who can manage and handle huge, messy healthcare data sets from various sources, and to bring them all together in an analysable format. It provides knowledge on analysing the data using statistical machine learning approaches and drawing useful insights from the data. Masters in data science receive training in multiple areas, including quantitative methods, applied regression, analysis statistical, statistical inference computing, machine learning, statistical consultation and collaboration, and epidemiology methodology. The main programming languages that are used in the most health data science courses include a mixture of Python and R, and SAS also in some cases. Needed Skills For people looking to pursue a Masters program in health data science, they do not necessarily have a biomedical background. Instead, it requires strong computation or mathematical background. A strong quantitative background exposes aspiring health data scientists to several research areas in healthcare and medical research that their skills will be highly advantageous for in the future. At the core, there is a need for a strong quantitative background. Students who have an undergraduate in mathematics or statistics or a related field would be ideal. Aspiring healthcare data scientists need to have a deep understanding of statistics, linear algebra and calculus. One of the other skills required is to be able to communicate results with various healthcare stakeholders. Health data scientists need to communicate to other data scientists about the methods they’ve used, talk to the clinicians to understand the disease they’re looking at, communicate to the lab scientists and most importantly be able to communicate clearly and transparently with patients and the public. Because at the heart of all health data science projects is the patient or the public, and building proper applications that can have a real impact in the Health Service and in commercial domains.","excerpt":"Big Data in healthcare is not unheard of. The healthcare industry is generating a significant source of data in the world. People who have skill sets in health information technology and analytics, the employment rate for them is increasing.  The global pandemic has dramatically increased the need as well as the importance of data science […]","categories":["AI Highlights"],"tags":["data analytics masters","healthcare analytics","MSc data science","msc data science and analytics","MSc. in data science"],"author_name":"Vishal Chawla","publish_date":"2020-08-27T13:00:20","publication_year":"2020","word_count":778,"keywords":["MSc. in data science","data science","data analytics masters","big data","machine learning","programming_languages:R","AI","ML","healthcare analytics","Python","msc data science and analytics","analytics","MSc data science","programming_languages:Python","R"],"extracted_tech_keywords":["AI","machine learning","ML","data science","analytics","Python","R","big data","programming_languages:Python","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/what-does-it-take-to-become-a-healthcare-data-scientist\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10072516,"title":"ReNew Power is hiring! Participate in this exciting new hackathon and get a chance to win prizes &#038; opportunity to work with the digital team at ReNew Power","content":"Global renewable energy player ReNew Power, in partnership with MachineHack, is inviting data scientists, machine learning developers, and AI enthusiasts to take part in the two-week long hackathon — ‘ReNew Power Hiring Hackathon’ to predict and flag failure of wind turbines, showcase your skills and expertise, impress the judges and stand a chance to win exciting cash prizes. Click here to know more about the hackathon Running from August 19, 2022, to September 9, 2022, ‘ReNew Power Hiring Hackathon’ gives data science and analytics professionals an exciting opportunity to show their skills, technical prowess and creativity. The participants will not only get a chance to win goodies and gifts but also an opportunity to work with ReNew Power. Predict flaws in wind turbine Unplanned downtime of wind turbines can result in a significant loss of revenue and energy and can easily scale to millions of dollars a year. It is therefore pivotal that flagging the failure of components is made to prevent further loss and perform maintenance. It, however, involves the replacement of components and higher costs. Condition-based monitoring systems rely on supervisory control and data acquisition systems to predict faults and get valuable insights into the turbine’s performance. What are you waiting for then? Here’s your chance to get hired and work with an analytics team that cares about the planet! Show ReNew Power your data science skills can help them accelerate wind analytics at warp speed! Stay tuned for more details! Click here to know more about the hackathon About ReNew Power: ReNew Power is one of the largest renewable energy Independent Power Producers in India and globally. ReNew Power develops, builds, owns, and operates utility-scale wind and solar energy and hydro projects. As of June 14, 2022, ReNew Power had a total gross portfolio of ~12.8 GW renewable energy projects across India. To know more, visit www.renewpower.in and follow us on LinkedIn, Facebook, Twitter and Instagram.","excerpt":"‘ReNew Power Hiring Hackathon’ gives data science and analytics professionals an exciting opportunity to show their skills, technical prowess and creativity.","categories":["Deep Tech"],"tags":["Data Science Hiring","Hackathon","hackathons in India","Hackathons India"],"author_name":"Amit Naik","publish_date":"2022-08-10T12:00:00","publication_year":"2022","word_count":319,"keywords":["data science","Go","machine learning","AWS","AI","hackathons in India","R","cloud_platforms:AWS","programming_languages:R","Hackathon","Data Science Hiring","ViT","analytics","Hackathons India"],"extracted_tech_keywords":["AI","machine learning","data science","analytics","AWS","R","Go","ViT","cloud_platforms:AWS","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/renew-power-is-hiring-participate-in-this-exciting-new-hackathon-and-stand-a-chance-to-win-exciting-prizes\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10042956,"title":"Inside Google’s Quantum AI Campus","content":"Google is on a mission to build an error-corrected quantum computer. To begin this ambitious journey, the tech giant has set up a Quantum AI Campus in Santa Barbara, California, housing Google’s first quantum data center, quantum hardware research laboratories and quantum processor chip fabrication facilities. Google said their future challenges include building more efficient batteries, creating better fertilisers to increase crop yield, and developing medicines that can stop the next pandemic in its tracks. Nature is quantum mechanical. That is, the bonds and interactions among atoms behave probabilistically, with richer dynamics that exhaust the simple classical computing logic.  Exactly why we need quantum computers to solve future challenges. Google’s quantum computer will enable the simulation of molecular behaviours and interactions and scientists will be able to test and invent new chemical processes and new materials. Inside Google’s Quantum AI campus Inside Google’s Quantum AI campus, a team is developing highly specialised quantum hardware, software and algorithms to build a useful, error-corrected quantum computer. Hartmut Neven, Chief Scientist at Google, believes the tech giant has a tight sequence of difficult but reachable milestones and hopes the team will be able to build the error-corrected quantum computer before the decade ends. Google Quantum AI Lab connects the different pieces of quantum computing together using open source software and Cloud APIs. Its products include open-source libraries such as Cirq, OpenFermion, and TensorFlow Quantum. Cirq is Google’s way of defining and modifying quantum circuits. It allows programmers to design and create quantum circuits, analyse it using simulators and then send it to hardware using Google’s quantum computing service API. OpenFermion, on the other hand, is a library for quantum simulations, especially quantum chemistry and electronic structure calculations. The error-corrected quantum computer will be the size of a tennis court, said Eirk Lucero, Lead Engineer, Quantum Operations and Site Lead at Google Santa Barbara. Within the quantum computer, one million qubits will operate in concert, directed by a surface code error correction. The Sycamore quantum processors has 54 individually controllable qubits and 88 tunable couplers | Image source: Google Marissa Giustina, Quantum Electronics Engineer and Research Scientist at Google, is part of the team building the cryogenic hardware that facilitates information transfer. She said the systems inside the campus look like racks of electronics operating at room temperature, and connected to a big cylinder– the dilution refrigerator, as seen below. Inside Google’s Cryostats Inside the package, the processor is stored. A quantum computer is a large system with a lot of supporting infrastructure and thousands of components. Today, a modestly-sized system consists of less than 100 physical quantum bits or qubits. Google’s north star is to build 10,00,000 qubits that work in a concert inside a room-sized quantum computer. However, to get there, the tech giant has to build the world’s first quantum transistor– two error-corrected logical qubits that perform quantum operations together. Post that, Google has to figure out how to tile hundreds to thousands of these logical qubits, to actualise the quantum computer. Error-free computing Google wants to make quantum computing accessible and useful to both high-powered researchers, as well as aspiring programmers. The whole process of developing the quantum computer is expected to take years. However, to get there, Google needs to first encode one logical qubit – with 1,000 physical qubits. These physical qubits will work together to form a long-lived and nearly perfect qubit, using quantum error correction. Take a virtual tour of Google’s Quantum AI Campus, here.","excerpt":"Google is on a mission to build an error-corrected quantum computer. To begin this ambitious journey, the tech giant has set up a Quantum AI Campus in Santa Barbara, California, housing Google’s first quantum data center, quantum hardware research laboratories and quantum processor chip fabrication facilities.  Google said their future challenges include building more efficient […]","categories":["Global Tech"],"tags":["Google Quantum Computing","quantum cloud software","Quantum Computing"],"author_name":"Debolina Biswas","publish_date":"2021-07-06T16:00:00","publication_year":"2021","word_count":579,"keywords":["Quantum Computing","Go","API","programming_languages:R","AI","quantum cloud software","Google Quantum Computing","programming_languages:Go","ai_frameworks:TensorFlow","TensorFlow","R","emerging_tech:quantum computing"],"extracted_tech_keywords":["AI","TensorFlow","R","Go","API","ai_frameworks:TensorFlow","programming_languages:R","programming_languages:Go","emerging_tech:quantum computing"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/inside-googles-quantum-ai-campus\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10011174,"title":"How NVIDIA Powered America’s Fastest Supercomputer In The Fight Against COVID-19","content":"“Using Dask, RAPIDS, BlazingSQL, and NVIDIA GPUs, researchers are leveraging Summit supercomputers from their laptops.” Working on data-intensive projects like protein folding research, drug discovery, or deep space leads to several TBs of data. And, using queries on CPUs to sort information can take days. Time is a key constraint while fighting global pandemics. Research labs and governments around the world have accommodated money and manpower to speed up drug discovery. But this isn’t sufficient. There is a need for a smart, diligent solution that combines the existing technologies without trying to reinvent the wheel. At Oak Ridge National Laboratory, which has been at the forefront of the fight against COVID-19, the researchers have been leveraging the powerful SUMMIT supercomputer to skim large datasets in search of solutions. SUMMIT, the world’s second-fastest supercomputer is powered by NVIDIA’s Tesla V100s and the team at OLCF (Oakridge Leadership Computing Facility) has been looking for solutions that would fit well into their technology stack. The team at OLCF reviewed commercial applications such as finance and marketing approach the analysis of large structured datasets. They identified that integrating BlazingSQL into the RAPIDS\/Dask ecosystem would provide GPU-accelerated open-source platform to process extremely fast and scalable SQL queries, along with other data analytics. This flexible toolset enabled engineers to get this custom workflow up and running in less than two weeks. The Secret Sauce Behind Summit Summit by the numbers. (Source: IBM) Back in 2014, the United States awarded a $325 million contract to IBM, NVIDIA and Mellanox to construct two supercomputers — Summit and Sierra. Summit ended up being the world’s fastest supercomputer until it was eclipsed by Japan’s Fugaku a couple of months ago. Summit is also the first supercomputer to reach exaflop (a quintillion operations per second) speed, achieving 1.88 exaflops during a genomic analysis. Today, using Dask, RAPIDS, BlazingSQL, and NVIDIA GPUs, researchers can leverage the power of Summit supercomputers from their laptops. This year in June, NVIDIA announced that using the RAPIDS suite of open-source data science software libraries powered by 16 DGX A100 systems, it ran the benchmark in just 14.5 minutes compared to the previous high of 4.7 hours on a CPU system. The DGX A100 systems had a total of 128 NVIDIA A100 GPUs and used NVIDIA Mellanox networking. The RAPIDS data science framework is a collection of libraries that are used for executing end-to-end data science pipelines completely in the GPU. RAPIDS uses optimised NVIDIA CUDA primitives and high-bandwidth GPU memory to accelerate data preparation and machine learning. The goal of RAPIDS is not only to accelerate the individual parts of the typical data science workflow but to accelerate the complete end-to-end workflow. Whereas, BlazingSQL is a fully open-source, free to use standard SQL engine built entirely on top of RAPIDS.ai. BlazingSQL lets users ETL raw data directly into GPU memory as a GPU DataFrame (GDF). While BlazingSQL and RAPIDS have multiplied NVIDIA’s role in the advancement of AI-based research, there is another key component that helped it to meet the demands–DASK. When developed back in 1989, Python was not intended to handle the TB-scale production workloads. But, the way it bridges high-performance languages and APIs like Fortran and CUDA to lightweight, user-friendly APIs. It would not be an exaggeration if one were to say that there are no data scientists who haven’t heard of numpy, scikit-learn and pandas. However, these successful libraries didn’t offer solutions to parallelism problems. Enter Dask. Data courtesy of JetBrains Python Developers 2019 Survey. Dask supports native code, which makes it easy to work with for Python users and C\/C++\/CUDA developers. Dask is also a critical component of the RAPIDS ecosystem, making it even easier to take advantage of accelerated computing through a comfortable Python-based user experience. Today, many scientific research centres, including Oak Ridge, are adopting both Dask and RAPIDS to scale some of their most important operations. Some of NVIDIA’s biggest partners, leaders in their industries, are using Dask and RAPIDS to power their data analytics. There is a rise in demand for highly usable distributed computing, need for more computational power and open-source software. At the intersection of these trends is NVIDIA offering integrated solutions through RAPIDS, BlazingAQL and Dask. Know more about NVIDIA’s Supercomputing efforts here.","excerpt":"“Using Dask, RAPIDS, BlazingSQL, and NVIDIA GPUs, researchers are leveraging Summit supercomputers from their laptops.” Working on data-intensive projects like protein folding research, drug discovery, or deep space leads to several TBs of data. And, using queries on CPUs to sort information can take days. Time is a key constraint while fighting global pandemics. Research […]","categories":["Global Tech"],"tags":["computer science projects","NVIDIA"],"author_name":"Ram Sagar","publish_date":"2020-11-04T11:00:43","publication_year":"2020","word_count":708,"keywords":["data science","scikit-learn","NumPy","machine learning","Rapids","AI","RAG","computer science projects","analytics","Dask","NVIDIA","Pandas"],"extracted_tech_keywords":["AI","machine learning","data science","analytics","scikit-learn","Pandas","NumPy","Dask","Rapids","RAG"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/nvidia-supercomputer-summit-america-fight-covid-19\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10016050,"title":"Informer: LSTF(Long Sequence Time-Series Forecasting) Model","content":"Time series forecasting is in the industry before AI and machine learning, and it is the most complex technique to solve and forecast with the help of traditional methods of using statistics for time series forecasting the data. But now as the neural network has been introduced and many CNN-based time series forecasting models have been developed over the years, you can see how accurate and easy it became to predict future values based on historical time-series data points. Now many daily use cases require future prediction like electricity consumption planning, Long short term memory(LSTM) is the one which is used for long-term forecasting. But there are many problems with LSTM which leads to further research in LSTF, let’s address one with a real-world  dataset of electrical transformer station temperature readings: Short sequence forecasting predicts fewer future data points.Long sequence Forecasting predicts a more extended period of time for better policy planning and investment.Due to long future prediction, the capacity of the existing method limits the performance of the long sequence forecasting, i.e., after 48 the MSE rises aggressively high, and the inference speeds drop. Here the LSTM network predicts the temperature of the station on an hourly basis to a longer period of time, i.e. short term period (12 points, 0.5 days) to the long sequence forecasting(480 points, 20 days). As shown in the above Fig.(c) the performance gap is substantial when the period sequence length got greater than 48 points in long sequence forecasting, the MSE score got an unsatisfactory result, and there is a sharp drop in inference speed, and the LSTM model fails. Solving Long Sequence time series forecasting(LSTF) is the major problem. Some new models have been developed like transformers that show superior performance in capturing long-range time series data than RNN(recurrent neural networks) models. The transformer takes a lot of GPU computing power, so using them on real-world LSTF problems is unaffordable. Informer So to solve this problem recently a new approach has been introduced, Informer. With a research paper called Informers: Beyond Efficient Transformers for Long Sequence, Time-Series Forecasting. It is written by Haoyi Zhou, Shanghang Zhang, Jieqi Peng, Shuai Zhang, Jianxin Li, Hui Xiong, and Wancai Zhang. The team of these researchers came up with a solution to answer the questions: Can Transformer models be improved to be more memory and architecture efficient?Can still by optimizing the computation power of the transformer it can maintain higher prediction capacity? The previous Transformer model for LSTF has three limitations: Quadratic computation of self-attention.Memory bottleneck in stacking layers for long inputs.Speed plunge in predicting long outputs. To remove all of these issues Informer comprises some new features: Purposed ProbSparse self-attention mechanism to remove the canonical self-attention and it achieves the o(L log L) time complexity and memory usage.Enhanced the prediction capacity in the LSTF problem, which contains the Transformer-like model performance to capture an individual long-range dependency between time-series data.Self-attention distilling operation privileges dominating attention scores.Reduced space complexity to O((2 − e)L log L).Introduced Generative Style Decoder to acquire long sequence output with only one forward step needed. Graph of Informer Model Fig. 1 The left part is the encoder, and it is capable of receiving a massive amount of long sequence data inputs(the green series). Now as we discussed Informer removed the canonical self-attention with their purposes ProbSparse self-attention. The blue trapezoid is the self-attention distilling operation to extract dominating attention, which reduces the network size sharply. The decoder receives the long sequence data inputs, pads the target elements into 0, measures the feature map, and instantly gives the predicted outputs(orange bars) in the generative style. Informer’s Encoder architecture Fig.2 The upper stack is the main stack, which receives the whole input sequence.The second stack takes a half slice of input.Each horizontal stack in fig.1 stands for individual encoder copies.Red layers are dot product matrixes of the self-attention mechanism, and it gets cascade decrease by applying self-attention distilling on every layer.2 feature map stacks are concatenated as the encoder’s output. Getting Started Research paper techniques are coded by Jieqi Peng and the repo is recently published with initial started code and dependencies on GitHub, code is in its initial release it is written in PyTorch and for reproducing the testing result make sure to use good GPU local machine as the training time can vary. Requirements Python 3.6+Matplotlib – 3.1numpy – 1.17.3pandas – 0.25.1scikit_learn – 0.21.3torch – 1.2.0 Clone and install the dependencies !git clone https:\/\/github.com\/zhouhaoyi\/Informer2020.git %cd Informer2020 !pip install -r requirements.txt Data We are going to use the ETT(Electricity Transformers temperature) dataset as the informer used in the paper was tested on three different datasets ECL(Electricity consuming load), Weather dataset, and ETT. Download the ETT dataset from hereCopy all the CSV files to Informer2020\/data\/ETT\/ folder. Let’s see the data !pip install pandas Import pandas as pd h1=pd.read_csv(“.\/Informer2020\/data\/ETT\/ETTh1.csv\") h1.head() Training & Testing For training and testing the model with ProbSparse self-attention on ETTh1, ETTh2, and ETTm1 dataset respectively use the following commands and reproduce the research paper results: Note: Training can take hours if your GPU isn’t powerful enough in the case of a local machine. # ETTh1 dataset ! python -u main_informer.py --model informer --data ETTh1 --attn prob # similarly ETTh2 dataset ! python -u main_informer.py --model informer --data ETTh2 --attn prob # ETTm1.CSV python -u main_informer.py --model informer --data ETTm1 --attn prob Outputs Univariate long sequence time-series forecasting evaluation results on all the methods on four datasets. The best result is in bold representation. Univariate Forecasting results Conclusion We have seen how the Informer removes the LSTF problem and provided the ProbSparse Self-attention mechanism, which achieves o(L log L) time complexity and memory usage. For downloading and training on your own machine download the notebook from here: https:\/\/github.com\/mmaithani\/Informer2020\/blob\/main\/informers_for_LSTF.ipynb and outputs and training methods will be updated by time. Other datasets on which informer is been tested are as follows and you can reproduce the result on them: ETT dataset at https:\/\/github.com\/zhouhaoyi\/ETDataset. ECL dataset was acquired at https:\/\/archive.ics.uci.edu\/ml\/ datasets\/ElectricityLoadDiagrams20112014 Weather dataset was acquired at https:\/\/www.ncdc.noaa.gov\/ orders\/qclcd\/ Also if you are interested in reading more about time series forecasting then checkout PyTorch time series forecasting and TensorFlow time series forecasting. To contribute to Informer open source projects, you can visit the official repo at https:\/\/github.com\/zhouhaoyi\/Informer2020 and contribute.","excerpt":"Time series forecasting is in the industry before AI and machine learning, and it is the most complex technique to solve and forecast with the help of traditional methods of using statistics for time series forecasting the data. But now as the neural network has been introduced and many CNN-based time series forecasting models have […]","categories":["Deep Tech"],"tags":["forecasting","Matplotlib","Time Series","Time Series Forecasting"],"author_name":"Mohit Maithani","publish_date":"2020-12-28T10:00:00","publication_year":"2020","word_count":1043,"keywords":["NumPy","machine learning","AI","neural network","TensorFlow","ML","PyTorch","Time Series Forecasting","Transformers","Time Series","forecasting","Matplotlib","Pandas"],"extracted_tech_keywords":["AI","machine learning","ML","neural network","TensorFlow","PyTorch","Transformers","Pandas","NumPy","Matplotlib"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/informer\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":27184,"title":"Finally, Computers Can Learn To Count Better","content":"With the onslaught of neural networks and deep learning, the breadth of tasks carried out by a computer has grown very fast. Neural networks have managed to learn to represent and manipulate numerical data. But at the same time, it is hard for these kinds of machine algorithms to generalise the data from outside the training regime. The researchers at Google Deepmind, the University of Oxford and the University College London have come up with a way to teach computers to count. The researchers have come up with a neural arithmetic logic unit. The research is some way follows the work done by Alex Graves who invented the Neural Turing Machine (NTM) which tries to bind values to specific locations by designing neural networks that has an external memory. There are major differences between the functioning of neural networks and our brain. The biggest difference is the memory. Ability to read and write from memory is critical and both the computer and the brain can do it, but not with Neural networks. Previously, Alex Graves and fellow researchers at DeepMind aimed to build a differentiable computer, which could put together a neural network and link it to external memory. The neural network would act like a CPU and with the memory attached, the aim would be to learn programs (algorithms) from input and output examples. Previous Failures In Numerical Extrapolation The researchers develop a new module that can be plugged into conventional architectures to learn numerical computations. The strategy is of the researchers is to represent numerical quantities with a single neuron. The researchers applied operators like simple functions such as  +, −, ×, to these single neurons. Surprisingly these neurons are still differentiable and this makes it possible for us to learn them with backpropagation. The researchers work with various data and domains such as images, code and text and using supervised and reinforcement learning. The researchers have also done a good work of showing how standard networks have failed in numerical extrapolation. To this end, they trained an autoencoder to take the scalar value as an input and encode it in the hidden layers and then reconstruct before the output layers. They observed that many architectures fail to represent numbers that the networks have not seen during the training. The researchers also observe that highly linear (such as PReLU) beat sharply nonlinear functions such as sigmoid and tanh. This is a strange behaviour since this is counter-intuitive. The researchers in an attempt to test this, trained around 100 models to encode numbers between −5 and 5 and tested their ability to encode numbers between −20 and 20. Neural Arithmetic Logic Unit The researchers propose two models that are taught to manipulate numbers. The first model is called the neural accumulator (NAC), which researchers state that, “neural accumulator is a special case of a linear (affine) layer whose transformation matrix W consists just of −1’s, 0’s, and 1’s; that is, its outputs are additions or subtractions (rather than arbitrary rescalings) of rows in the input vector.” This ensures consistency throughout the model, and there is no dependency on the number of operations are chained together. Since it is easy to learn, whose elements are [−1, 1] and biased to be close to −1, 0, or 1. The model does not have a bias vector, and no squashing nonlinearity is applied to the output. We would also like to have the powers of multiplication, division along with the simpler addition and subtraction operations. The second model design is known as the neural arithmetic logic unit (NALU). This model learns a weighted sum between two subcells. Both of the cells have their own function where one cell has the property of doing simple operations like addition and subtraction and other is responsible for complex operations like multiplication, division, and power functions. NALU is related to the NAC. When we extend NAC with gate-controlled sub-operations we get a NALU. Experiments And Results The researchers put on many test cases to work on numeric reasoning and extrapolation of the above architectures. The researchers state that, “We study the explicit learning of simple arithmetic functions directly from the numerical input, and indirectly from image data. These supervised tasks are supplemented with a reinforcement learning task which implicitly involves counting to keep track of time.” The researchers try to have NALU learn some simple function learning tasks and they showcase the ability of NACs and NALUs. They show how these two architectures can learn to select relevant inputs and apply different arithmetic functions to them. Results show that NALU outperforms traditional neural architectures on many extrapolation tasks. Apart from the counting tasks, they also tested the performance of the architecture in program evaluation. This task is a perfect test for this architecture because program evaluation requires as researchers state, “the control of several logical and arithmetic operations and internal book-keeping of intermediate values.” Researchers test consists of tasks like simply adding two large integers, and evaluating programs containing several operations (if statements, +, −). In this particular task also they found that only NALU architecture was able to extrapolate to larger numbers. The researchers conclude by saying, “We have shown how the NAC and NALU can be applied to rectify these two shortcomings across a wide variety of domains, facilitating both numerical representations and functions on numerical representations that generalise outside of the range observed during training. This design strategy is enabled by the single-neuron number representation we propose, which allows arbitrary (differentiable) numerical functions to be added to the module and controlled via learned gates, as the NALU has exemplified between addition\/subtraction and multiplication\/division.”","excerpt":"With the onslaught of neural networks and deep learning, the breadth of tasks carried out by a computer has grown very fast. Neural networks have managed to learn to represent and manipulate numerical data. But at the same time, it is hard for these kinds of machine algorithms to generalise the data from outside the […]","categories":["AI Features"],"tags":["analog neural networks","deepmind london","Neural Networks"],"author_name":"Abhijeet Katte","publish_date":"2018-08-10T09:38:56","publication_year":"2018","word_count":940,"keywords":["Go","TPU","programming_languages:R","AI","neural network","analog neural networks","Scala","Aim","deep learning","programming_languages:Scala","R","deepmind london","Neural Networks"],"extracted_tech_keywords":["AI","deep learning","neural network","Aim","TPU","R","Go","Scala","programming_languages:R","programming_languages:Scala"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/finally-computers-can-learn-to-count-better\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10005418,"title":"How to Predict Pneumonia Based On CXR Images Using Transfer Learning?","content":"Deep Learning has shown immense results in medical imaging. It is due to the high volume of data that is generated in the medical domain. There are several use cases where AI technologies are used today in the healthcare domain. There can be errors made by humans depending on several factors whereas a machine will not make error provided data is correct. The reason for using Deep learning in medical imaging is the fact that we can attain insights from the data quickly with reliable results. With AI it has now become possible to detect even different types of cancer in the lungs and kidneys also it is used in different therapy. Diagnosing pneumonia is also one of the important applications of deep learning. Pneumonia is such an infection that is caused in one or both the lungs. This is caused due to viruses, fungi, etc. This results in inflammation in air sacs in the lungs by which it becomes difficult to breathe. Through this article, we will explore how to build a classification model by which we can classify whether a person has pneumonia or not through CXR (Chest X-Ray) images. We will be building the model using pre-trained model Vgg19. For this experiment, we will make use of Pneumonia Chest X Rays data that is publicly available on Kaggle. The Dataset There are a total of 5863 CXR (Chest X-Ray) images that are categorized into two categories that are Pneumonia and Normal. The data has three folders: train, test, and Val in which both two categories subfolders are present. The X-rays images were screened by experts so that there are no unreadable images or low-quality images. The normal chest X-ray (left panel) depicts clear lungs without any areas of abnormal opacification in the image. Bacterial pneumonia (middle) typically exhibits a focal lobar consolidation, in this case in the right upper lobe (white arrows), whereas viral pneumonia (right) manifests with a more diffuse ‘‘interstitial’’ pattern in both lungs. Read more here in this paper. Model Building First, we need to install the required package and libraries that are required. As we will be importing the data using the API command from Kaggle. We should have the Kaggle package installed. Use the below code for the same. from keras.applications.vgg19 import VGG19 from keras.preprocessing.image import ImageDataGenerator from keras.models import Sequential import numpy as np import matplotlib.pyplot as plt import os from zipfile import ZipFile import os from tqdm._tqdm_notebook import tqdm_notebook as tqdm import cv2 import tensorflow as tf import keras !pip install kaggle Now we will load the dataset from Kaggle and unzip the downloaded file. Refer to the below code to do the same. from google.colab import files files.upload() !mkdir -p ~\/.kaggle !cp kaggle.json ~\/.kaggle\/ !chmod 600 ~\/.kaggle\/kaggle.json !kaggle datasets download -d paultimothymooney\/chest-xray-pneumonia from zipfile import ZipFile file_name = \"chest-xray-pneumonia.zip\" with ZipFile(file_name,'r') as zip: zip.extractall() print('Done') After we have extracted the zip file downloaded from Kaggle we will get three folders that are train, test, and Val. We will now load the training images with the respective labels and visualize normal and pneumonia X-rays. X_train holds the training images and y_train holds the respective labels. Use the below code to do the same. X_train = [] y_train = [] os.chdir('\/content\/chest_xray\/train\/NORMAL') for i in tqdm(os.listdir()): img = cv2.imread(i) img = cv2.resize(img,(256,256)) X_train.append(img) y_train.append(\"Normal\") os.chdir('\/content\/chest_xray\/train\/PNEUMONIA') for i in tqdm(os.listdir()): img = cv2.imread(i) img = cv2.resize(img,(256,256)) X_train.append(img) y_train.append(\"PNEUMONIA\") print(len(X_train)) print(len(y_train)) Output: There are a total of 5216 images in the training folder. We will now visualize one normal and one pneumonia patient X-ray image. Use the below code for the same plt.figure(figsize=(5,5)) plt.imshow(X_train[10], cmap=\"gray\") plt.axis('off') plt.show() print(y_train[10]) Output: plt.figure(figsize=(5,5)) plt.imshow(X_train[4000], cmap=\"gray\") plt.axis('off') plt.show() print(y_train[4000]) Output: Now we will build the model for classifying the X-rays into the desired two categories. We will be using pre-trained architecture VGG19 and will not train the whole network. We will remove the last layer of the network and will add the last custom layer as we only have 2 classes whereas the model was trained for 1000 classes on the ImageNet dataset. Use the below code to the same. vgg19 = VGG19(input_shape=[224,224,3], weights='imagenet', include_top=False) for layer in vgg19.layers: layer.trainable = False X = Flatten()(vgg19.output) output = Dense(2, activation='softmax')(X) model = Model(inputs=vgg19.input, outputs=output) Now we will compile the build model using loss function and optimizer. After compiling we will prepare the training and testing images that can be fed to the model for training. Refer to the below code for the same. model.compile(loss='categorical_crossentropy',optimizer='adam',metrics=['accuracy']) train_dir = '\/content\/chest_xray\/train' testing_dir = '\/content\/chest_xray\/test' train_datagen = ImageDataGenerator(rescale = 1.\/255, shear_range = 0.2, zoom_range = 0.2, horizontal_flip = True) test_datagen = ImageDataGenerator(rescale = 1.\/255) train_data = train_datagen.flow_from_directory(train_dir, target_size = (224, 224), batch_size = 32, class_mode = 'categorical') test_data = test_datagen.flow_from_directory(testing_dir, target_size = (224, 224), batch_size = 32, class_mode = 'categorical') history = model.fit(train_data,validation_data=test_data,epochs=10) Output: Now we will visualize the model accuracy and model loss for training and testing. Then we will evaluate the model performance. Use the below code to do the same. plt.plot(history.history['loss']) plt.plot(history.history['val_loss']) plt.title('model loss') plt.ylabel('Loss') plt.xlabel('Epoch') plt.legend(['Training', 'Testing'], loc='upper left') plt.show() Output: plt.plot(history.history['accuracy']) plt.plot(history.history['val_accuracy']) plt.title('model accuracy') plt.ylabel('Accuracy') plt.xlabel('Epoch') plt.legend(['Training', 'Testing'], loc='upper left') plt.show() Output: model.evaluate(test_data) Output: Now we will save the model to compute predictions on random images from the testing data. Use the below code to the same. model.save('MODEL.h5') Now we will compute predictions. To do so first we will import the required libraries that are required and load the saved model. Use the below code for the same. import keras from keras.models import load_model import os from tqdm._tqdm_notebook import tqdm_notebook as tqdm import cv2 import numpy as np import matplotlib.pyplot as plt model = load_model('\/content\/drive\/My Drive\/Pneumonia\/MODEL.h5') Let us import the testing data from the test folder. X_test = [] y_test = [] os.chdir('\/content\/chest_xray\/test\/NORMAL') for i in tqdm(os.listdir()): img = cv2.imread(i) img = cv2.resize(img,(224,224)) X_test.append(img) y_test.append(\"0\") os.chdir('\/content\/chest_xray\/test\/PNEUMONIA') for i in tqdm(os.listdir()): img = cv2.imread(i) img = cv2.resize(img,(224,224)) X_test.append(img) y_test.append(\"1\") We will now convert the images into NumPy arrays. After this, we will evaluate the model performance on testing images using different metrics. Use the below code for the same. X_test = np.array(X_test) y_test = np.array(y_test) predicted_classes = model.predict(X_test[:,:,:,:]) predicted_classes = np.argmax(np.round(predicted_classes),axis=1) predicted_classes[0] y_test = y_test.astype('int64') from sklearn.metrics import accuracy_score,classification_report accuracy_score(predicted_classes,y_test) Output: print(classification_report(predicted_classes,y_test)) Output: Now we will make some predictions on the testing images and compare the results. Use the below code to the same. L = 3 W = 3 fig, axes = plt.subplots(L, W, figsize = (12,12)) axes = axes.ravel() print('\\n\\n\\t\\t0 Class Represents Normal & 1 Class Represents Pnemonia') for i in np.arange(0, L * W): axes[i].imshow(X_test[i]) axes[i].set_title(f\"Prediction Class = {predicted_classes[i]}\\n True Class = {y_test[i]}\") axes[i].axis('off') plt.subplots_adjust(wspace=0.5) Out of 9 testing images 7 images were correctly classified whereas 2 were misclassified by the model. Conclusion We have built an AI model using pre-trained architecture VGG19 for classifying X-ray images into pneumonia and normal images. The model classified 7 out of 9 images correctly. The model performance can be more enhanced by getting more data and performing some good no of augmentation techniques referring to the domain knowledge.  We can also try making the predictive model using different architectures like Inception or ResNet50 in the same manner and can compare the results. Also, check this article where I classified Brain Tumors from MRI images.","excerpt":"Through this article, we will explore how to build a classification model by which we can classify whether a person has pneumonia or not through CXR (Chest X-Ray) images.","categories":["Deep Tech"],"tags":["Transfer Learning","VGG19"],"author_name":"Rohit Dwivedi","publish_date":"2020-08-24T16:00:41","publication_year":"2020","word_count":1204,"keywords":["VGG19","NumPy","TPU","Keras","AI","Colab","Ray","Matplotlib","deep learning","Transfer Learning","TensorFlow","R"],"extracted_tech_keywords":["AI","deep learning","Ray","TensorFlow","Keras","Colab","NumPy","Matplotlib","TPU","R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/pneumonia-prediction-based-on-cxr-images-using-transfer-learning\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10000523,"title":"Eccentric Chinese Researcher Who Edited Babies’ Genes Still Missing","content":"He Jiankui, a Chinese biomedical researcher who had claimed to help make the “first genetically modified babies”, has reportedly been missing since last week. In his YouTube video, Jiankui had claimed that he edited genes of twin baby girls to make them HIV-resistant. “The twin girls are perfectly healthy and were born in November,” he had said. He had aroused awe and horror alike when he announced at a conference in Hong Kong last week that he used a CRISPR gene-editing system to modify the DNA of human embryos. Where’s Jiankui? Some reports have suggested that the scientist has been missing since the summit. The Chinese government has condemned his work and has proposed a plan to shut down the human genome project completely. They have also launched an investigation against his lab and experiment Many scientific communities are anxious and want to look at his case and investigate the experiment. About Jiankui’s Research Dr Robert Klitzman, director of the bioethics program at Columbia University, said in an interview, “The technique used by the researcher is very risky, it is a consensus in many countries to not edit genes from a human embryo and plant it into a mother’s body. There are many risks involved.” Jiankui had said that he used CRISPR cas9 technology, a new molecular tool which breaks specified parts of DNA, for his research. Jiankui had said, “This technology is not used to enhance genetic features like IQ, hair colour and eye; that should be banned. I want this technology to be used to heal people.” A Modern Marvel If Jiankui’s claims are to be believed, this technology can bring wonders everywhere. According to WHO reports, there are more than 36.9 million people infected with HIV-AIDS and 940,000 people have died in one year due to this disease. Cases, where infants are HIV positive because of either parent, can be averted by this technology. The cases of HIV in India are estimated above 21.4 lakh. The scientist may face a lot of trouble because his experiment is against Bioethics. Dr Klitzman added, “It is not unethical to experiment on the human genome, but to put it inside a mother’s body is highly risky and unethical. There’s no proof of his experiment but this activity will urge other scientists to do the same and challenge laws.”","excerpt":"He Jiankui, a Chinese biomedical researcher who had claimed to help make the “first genetically modified babies”, has reportedly been missing since last week. In his YouTube video, Jiankui had claimed that he edited genes of twin baby girls to make them HIV-resistant. “The twin girls are perfectly healthy and were born in November,” he […]","categories":["AI Trends"],"tags":[],"author_name":"Jignasa Sinha","publish_date":"2018-12-10T18:53:28","publication_year":"2018","word_count":388,"keywords":["Go","AWS","AI","cloud_platforms:AWS","programming_languages:R","BERT","Aim","llm_models:BERT","ViT","R"],"extracted_tech_keywords":["AI","Aim","AWS","R","Go","BERT","ViT","llm_models:BERT","cloud_platforms:AWS","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/eccentric-chinese-researcher-who-edited-babies-genes-still-missing\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10172033,"title":"The Story of a Prisoner Who Became a Software Engineer","content":"In a world where many may procrastinate learning to code or improving their skills despite leading a comfortable lifestyle, one man is proving that even the confines of prison cannot suppress a passion for coding. Meet the software engineer who, despite being incarcerated, is making his mark in the tech world. His story is a testament to the belief that anyone, anywhere, can master complex programming languages. AIM recently stumbled upon this individual—an open-source contributor and coder with expertise in Rust and Python programming languages, and an avid Linux user—who continues to build and contribute to databases, even from behind bars. What sounds like the plot of a movie is, in fact, the true story of Preston Thorpe, a software engineer at Turso, an open-source distributed database powered by libSQL. AIM had the opportunity to speak exclusively with Thorpe, who opened up about his journey of programming during his time in prison. A Prisoner’s Attempt at a Better Outlook on Life Through Coding The 33-year-old software engineer spends his days working remotely from his prison cell in the Mountain View Correctional Facility in Charleston, Maine. Despite the confines of the facility, he has become a software engineer at Turso, actively contributing to projects like the rewrite of SQLite. But his journey to this point has been far from conventional, driven by self-reflection, project-based learning, and an insatiable desire to improve. For nearly a decade, Thorpe was reportedly incarcerated for non-violent drug crimes. However, instead of succumbing to the obvious hopelessness that often defines life behind bars, he discovered a sense of purpose through programming. Explaining how it all started, Thorpe said, “There was one day, after spending a few years in the more calm and respectful environment in the Maine prison, where I had an epiphany and started questioning everything about my life.” “I no longer knew why I had accepted that identity and situation, none of it made sense to me anymore, and I decided that I was no longer okay with being where I was or who I had become.” Coding 90 Hours a Week Photo by Tima Miroshnichenko \/ Pexels Thorpe’s programming journey started with a simple but powerful resource: access to a computer through a prison college programme at the University of Maine at Augusta. With limited internet access and a passion to outgrow the curriculum in place, Thorpe created his own learning path. He primarily attributes his success to project-based learning, having had just enough high school experience to understand what he needed to learn. His days were consumed by intense self-study, working on projects, and contributing to open-source software. “I started in Python until I felt like I remembered enough of the basics, then moved to C and built very fundamental things like my own ‘standard library’ of data structures,” Thorpe said. This project-based approach allowed him to learn the intricacies of various programming languages while also developing practical tools that would serve as the foundation for his career. Thorpe’s learning wasn’t restricted to just writing code. He immersed himself in the theory of computer science, reading academic papers, listening to lectures, and exploring the underlying architecture of software systems. His interest in databases led him to explore relational databases, despite having no prior experience in the field. Thorpe explained that his database work initially involved logically isolated components, allowing him to focus on areas aligned with his existing knowledge. His initial contributions included translating from Abstract Syntax Tree (AST) to bytecode and working on the Virtual DataBase Engine (VDBE). He didn’t immediately delve into specific database internals, often working on the IO layer or the command-line interface (CLI). Thorpe also dedicated time to developing the extension library and Go language bindings. Through a process of gradual familiarisation, extensive reading of research papers, and studying CMU lectures, he built the confidence to explore diverse areas and implement features across the entire codebase. In a recent blog post on his company’s website, Thorpe highlighted, “I either write code or manage Kubernetes clusters or other infrastructure for about 90 hours a week, and my only entertainment is a daily hour of tech\/programming YouTube.” The Career Progression Thorpe’s self-driven journey took a pivotal turn when he was accepted into Maine’s remote work programme—a rare opportunity for imprisoned individuals to pursue legitimate employment outside the prison. This programme became the gateway to his professional career in tech. “Because there was no precedent set for any of this, what I believe is the most crucial support was the fact that administrators took a chance and allowed me to earn their trust eventually,” he said. His first job was with Unlocked Labs, a company focused on building educational technology for incarcerated individuals. Thorpe’s contributions there quickly gained recognition, and within a year, he was promoted to lead their development team. Despite thriving in his role at Unlocked Labs, Thorpe’s ambition drove him to push even further. His exposure to the world of databases through various open-source projects eventually led him to Turso, a company working on rewriting SQLite. Grateful for the Absence of LLMs and Project-based Learning Photo by Google DeepMind \/ Pexels In today’s fast-evolving tech landscape, many developers turn to tools powered by large language models (LLMs) like Claude Code to speed up their learning and coding. However, Thorpe views his lack of access to these tools during his learning years as a blessing in disguise. “I’m very grateful that LLMs are something that I did not have available to me for a large portion of my time learning,” he told AIM. “With the proper discipline, if it is a topic you are truly interested in, you can certainly use it to help teach you things, but I would worry for anyone who may be inclined to take shortcuts, as it could easily prevent learning as well.” He firmly believes in the value of building real-world projects as a means of understanding and mastering programming concepts. He asserted that the knowledge gained from solving a problem and building a solution would surpass the learning acquired by breaking down each component and focusing on individual parts. For Thorpe, learning didn’t just happen in isolation. He also credited his contributions to open-source projects as a key part of his development. “I have found reading code very valuable,” he said. Looking ahead, Thorpe is particularly excited about the future of embedded and distributed databases. Moreover, he envisions significant future developments at Turso, including native support for efficient semantic searches and similarity matching in embedded databases. Such developments would enable more efficient reasoning over locally stored context, eliminating the need for separate vector databases or complex infrastructure.","excerpt":"“I’m very grateful that LLMs are something that I did not have available to me for a large portion of my time learning.”","categories":["AI Features"],"tags":["coding","software engineer"],"author_name":"Ankush Das","publish_date":"2025-06-19T16:00:00","publication_year":"2025","word_count":1105,"keywords":["Go","semantic search","AI","coding","ML","vector databases","software engineer","Aim","Python","SQL","R","kubernetes"],"extracted_tech_keywords":["AI","ML","Aim","vector databases","semantic search","kubernetes","Python","R","SQL","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/the-story-of-a-prisoner-who-became-a-software-engineer\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":63318,"title":"How Crediwatch Aims To Solve India’s Credit Crunch With AI","content":"To understand the value of AI in credit risk assessment, Analytics India Magazine connected with Sandeep Anandampillai, founder and Chief Product Officer of Crediwatch. Artificial intelligence (AI), machine learning, and predictive analytics are reshaping the financial services landscape, enabling banks to analyse customers’ data better to provide them with loans efficiently. As part of this, credit risk analysis and credit risk management are fundamental to banks and financial institutions that provide loans to retail and SMEs. To understand the value of AI in credit risk assessment, Analytics India Magazine connected with Sandeep Anandampillai, founder and Chief Product Officer of Crediwatch. Unique Selling Point Of Crediwatch Crediwatch implements actionable credit analytics and dynamic credit assessment insights as a service to financial firms. The company can achieve this without human intervention by applying AI\/ML and NLP tools, which produce the most dependable and extensive real-time insights. Crediwatch is an insights-as-a-service platform that deploys scalable deep learning tools across different digital footprints left by large and small private entities. The startup makes use of over 18 million risk profiles of prominent companies and unregistered small firms. The platform is intended to deliver sharp insights over the credit lifecycle, from pre-disbursal to post-disbursal assessment, through its tools such as Early Warning System (EWS). The platform plans to give lending companies, corporates (large, MSMEs, SMEs and small business units) the capacity to manage and allocate credit effectively. According to the founder, India has more than 50 million small and medium enterprises who face the problem of liquidity crunch. Out of this, only 15% get access to formal credit due to the trust deficit that exists and lack of collateral. And for the ones who do get access to formal credit, they have to wait for 4 to 6 weeks to get their loan processed at a staggering rate of 16% – 24%. This scenario creates a debt financing gap of 1 trillion dollars in the market, and hence, these small and medium enterprises are under-banked and underserved. “There are a few data aggregators in the market who provide platforms to access public data. At the other end of the spectrum, some KPOs use analysts and operations staff to pull information – both of these models do not allow for scale and efficiency. At Crediwatch, we focus on bringing insights to our clients through a ‘zero human touch’ technology platform,” said Sandeep. Utilising 25,000 Data Points Crediwatch employs AI\/ ML algorithms on alternative data points, such as statutory payment statuses, litigations, media sentiment, GST invoice data, bank statements, as well as traditional data points such as financial ratios, industry outlook etc. “We realised that concentrating on quality risk and business insights by applying a proprietary AI-based predictive engine is the way forward for the financial services industry and this has helped us differentiate ourselves as a leading analytics player,” Sandeep said. The platform pulls up data from 25,000 different points, which is the biggest in the industry, from already existing data in the regulatory framework. On the other hand, a typical bank only scans 200 data points. Thus, this scale is what makes Crediwatch unique from additional credit assessment financial institutions, according to Sandeep. Moreover, the startup has completed the development of the enterprise version of its flagship product, Early Warning System. This product complies with the RBI framework and is based on a proprietary library of more than 190 early warning signals. The software includes a case management module to track alerts and manage post-alert actions from the respective portfolio manager. The Tech Stack If you look at the main components of the tech stack, here are some of the technologies used at Crediwatch, according to Sandeep Anandampillai. • Front end: React, Flutter • Back end: MongoDB, Redis, Kafka • Code: Python, Ruby • Infra: AWS, Azure, DigitalOcean Crediwatch: The Focus For now, the focus remains on enhancing offerings to the SME segment as the startup foresees immense potential in this sector. It is building a dynamic “Trust Score” derived from millions of data points that are extracted and analysed across thousands of formal and alternative sources to help lenders assess borrowers and monitor them close to real-time. “Our vision is to reimagine SME credit by scaling trust through verifiable data, insights and good behaviour. We comprehended the gaps in the market and envisioned a solution that the loan deficit only could be addressed if the lenders have valuable data-points of these small and medium businesses to bank them,” added Sandeep Anandampillai..","excerpt":"To understand the value of AI in credit risk assessment, Analytics India Magazine connected with Sandeep Anandampillai, founder and Chief Product Officer of Crediwatch.  Artificial intelligence (AI), machine learning, and predictive analytics are reshaping the financial services landscape, enabling banks to analyse customers’ data better to provide them with loans efficiently.  As part of this, […]","categories":["Deep Tech"],"tags":["credit risk machine learning","credit score","credit scoring","Startups"],"author_name":"Vishal Chawla","publish_date":"2020-04-27T17:37:43","publication_year":"2020","word_count":747,"keywords":["artificial intelligence","machine learning","AWS","AI","credit risk machine learning","Azure","ML","NLP","credit scoring","deep learning","analytics","credit score","Startups","predictive analytics"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","NLP","analytics","predictive analytics","AWS","Azure"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/crediwatch-startup-india\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10142555,"title":"Former PayPal COO David Sacks Appointed as US AI, Crypto Czar","content":"David Sacks, venture capitalist and former PayPal COO, has been appointed the United States’ first Artificial Intelligence and Crypto Czar. Sacks is expected to work closely with Elon Musk, who will head the Department of Government Efficiency (DOGE) in Trump’s administration. President-Elect Donald Trump announced the role, stating that Sacks will help establish a legal framework for the cryptocurrency industry and ensure the country remains a leader in AI and crypto. “I am pleased to announce that David O. Sacks will be the ‘White House A.I. & Crypto Czar.’ In this important role, David will guide policy for the Administration in Artificial Intelligence and cryptocurrency, two areas critical to the future of American competitiveness,” said Trump on Truth Social. Many in the tech and VC community took to X to congratulate Sacks on his new post, including OpenAI chief Sam Altman, VP elect JD Vance, Replit CEO Amjad Masad, and Sequoia partner Shaun Maguire, among others. Sacks co-hosts the All-In podcast with investors Jason and Chamath Palihapitiya, who both congratulated him on his win. “He (David Sacks) will make sure the US stays at the forefront of AI and Crypto – two of the most consequential technological movements in history. Make America Great Again,” wrote Palihapitiya on X. Sacks, 52, is the founder of venture capital firm Craft Ventures and was involved in the development of PayPal. He also founded Yammer and led Zenefits. His appointment does not require asset disclosure or divestment, and he will serve as a special government employee, working up to 130 days per year. The AI and Crypto Czar will act as a liaison between the crypto industry, Congress, and federal agencies such as the SEC and CFTC. The position will provide regulatory clarity for the growing crypto sector and guide policy for AI adoption, including its implications for national security, privacy, and jobs.","excerpt":"Trump appointed Sacks to maintain US’ leadership in AI and crypto.","categories":["AI News"],"tags":["AI (Artificial Intelligence)"],"author_name":"Aditi Suresh","publish_date":"2024-12-06T13:51:52","publication_year":"2024","word_count":309,"keywords":["Go","API","artificial intelligence","OpenAI","AI","programming_languages:R","venture capital","programming_languages:Go","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","OpenAI","R","Go","API","venture capital","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/former-paypal-coo-david-sacks-appointed-as-us-ai-crypto-czar\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":49025,"title":"IIT Madras Researchers Develop ‘AISoft’, An AI Algorithm To Solve Engineering Problems","content":"After the recent announcement on their plans to double the number of seats for its Inter-Disciplinary Dual Degree (IDDD) program on Data Science, IIT-Madras is in news again. This time for its plans to establish an AI startup. This comes after a team of researchers at the institutes led by Vishal Nandigana, assistant professor, Fluid Systems Laboratory, Department of Mechanical Engineering, IIT-Madras developed the AI and Deep Learning algorithms. Called ‘AISoft’, it aims to address engineering problems in various fields such as thermal management, semiconductors, automobile, aerospace and electronics cooling applications. According to researchers, AISoft has been tested to solve thermal management problems. “We found it to be nearly million-fold faster compared to existing solutions currently used in the field,” one of the researchers was quoted as saying. Currently, most researchers use CNN or C-GAN to solve engineering problems and this new development will open up new avenues of research in the field. The institute believes that AI, Machine Learning and Deep Learning are now being used for over a decade but traditionally only in areas such as signal processing, speech recognition, image reconstruction and prediction. Very limited attempts have been made globally in using these algorithms in solving engineering problems such as thermal management, electronics cooling industries, automobile problems like fluid dynamics prediction over a bonnet or inside the engine, aerospace industries like aerodynamics and fluid dynamics problems across an aero-foil or turbine engine. This AI algorithm offer advantages such as it works on any generalised rectilinear and curvilinear input geometry and saves computational time which remains a crucial challenge in the current settings. Researchers used a data-driven AI and a deep learning model to arrive at solutions for engineering problems after training the AI with data sets. Highlighting the unique aspects of these algorithms, Dr Vishal Nandigana said, “We tested AIsoft and used it to solve such thermal management problems. We found it to be nearly million-fold faster compared to existing solutions currently used in the field. Our AI works on any generalized rectilinear and curvilinear input geometry. Our research saves the computational time, which is the bottleneck to solve most engineering problems.” “We are using a novel Recurrent Neural Network (RNN) and Deep Neural Network (DNN) to solve engineering problems. Our AI is a grid or mesh independent and does not need the information of left and right grid points to solve the grid point of interest. Our AI can work with sparse data sets and solve engineering problems. Our AI stands out in this aspect from the commercially available software,” he added. With their plans to establish an AI startup, researchers are hopeful that these algorithms will solve a lot of pressing problems for industries and can also be used for educational purposes.","excerpt":"After the recent announcement on their plans to double the number of seats for its Inter-Disciplinary Dual Degree (IDDD) program on Data Science, IIT-Madras is in news again. This time for its plans to establish an AI startup. This comes after a team of researchers at the institutes led by Vishal Nandigana, assistant professor, Fluid […]","categories":["AI News"],"tags":["AI engineers","iit madras ai"],"author_name":"Srishti Deoras","publish_date":"2019-10-29T18:46:39","publication_year":"2019","word_count":457,"keywords":["data science","Go","machine learning","AI","neural network","AI engineers","Aim","deep learning","GAN","CNN","R","iit madras ai"],"extracted_tech_keywords":["AI","machine learning","deep learning","neural network","data science","Aim","R","Go","GAN","CNN"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/iit-madras-researchers-develop-aisoft-an-ai-algorithm-to-solve-engineering-problems\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":40194,"title":"Data Science Job Satisfaction Survey: 2019","content":"Large and small organisations from almost every industry — from Fortune 500 companies to retail stores — are hiring data science experts now. But are these people, who hold the “sexiest job of the 21st century”, happy with their work profiles? What are their key concerns? Why is this industry facing an increasing rate of attrition? The data science sector is flourishing to such an extent that our earlier jobs study revealed that there are currently more than 97,000 job openings for analytics and data science in India right now. In the month of May, Analytics India Magazine got in touch with numerous industry insiders to find out the level of job satisfaction in Data Scientists and the steps that are and can be taken to increase it. About The Survey For this survey, AIM reached out to various professionals from all experience sections. We took opinions from data scientists from all walks of life — from beginners to project managers and CIOs — to get a thorough idea of the hiring scenario in this swiftly-developing area. Our survey was met with great enthusiasm — and we got some deep insights from it. Some of them were expected, and many of them were real eye-openers. How Many Data Scientists Are Happy With Their Jobs? Job satisfaction depends on several factors: from the type of work done daily, to kind of work environment. We started our survey with this question point blank: were the responding data scientists happy with their jobs or not? At 54%, an overwhelming majority of the respondents chose to answer “somewhere in between”, clearly indicating that most of their jobs came with both ups and downs. 30% of the data scientists said that they were unhappy with their current jobs Only 13% of the respondents said that they were happy at their current workplace. What Is The Key Reason For Job Dissatisfaction? When we were conducting the survey, we found out that there were several parameters a data scientist took into consideration while calculating his\/her job satisfaction. From pay packages to work environment and sundry benefits, all these factors played a key role. However, 27% of the respondents said that they were unhappy with their jobs due to the low salary 15% of the respondents said that they felt trapped in their current employment due to the lack of career advancement opportunities 14% respondents said that they wanted more upskilling opportunities And 13% of the data scientists said that their work was stressful Step by Step Charts Infogram How Happy Are Data Scientists With Their Compensations? As seen above, one of the key reasons for data scientists feeling dissatisfied with their jobs is the salary. So when we really tried to find out about how their current employers were paying them, we a unanimous answer: 70% of the respondents felt that their employer gave them lesser than what their peers get in other companies 22% of the data scientists said that their salary was at par with industry standards And only 8% of the respondents said that they got more salary than their peers. Work Practices It is very important for an employee to feel that the practices at work are just and fair. When we asked the data scientists about if they felt that their talent was utilised optimally at their current organisation, 63% of the respondents said no. When we asked about if the work was distributed in the team in an equitable manner, again, a majority of the respondents at 53% said no. However, when we asked the participants if their peers and managers valued their feedback, 70% of the respondents said yes. Infogram How Should Organisations Increase Engagement With Their Employees? To keep talented data scientists interested in their work, organisations are now taking active efforts to increase engagement. From hackathons to more brain-storming sessions, many new as well as tried-and-tested methods are being used. The popular method at 33% for employee engagement according to the data scientists is a role in decision making. At 20% each, many data scientists voted that they wanted to see more hands-on meetings and more feedback and review from their peers. Copy: Step by Step Charts Infogram Upskilling In order to capitalise on these opportunities, IT companies, educators and policymakers need to develop a deeper understanding of the existing workforce, the skill-set required in the future, and the gaps that will need to be addressed.In this survey, we found out that all of the data scientists wish to be upskilled in all ways possible. Even though it was a very close race, almost 28% of the data scientists said that they would appreciate it if their employers tied up with noted institutes for customised certification programmes. Copy: Infogram Respondent Profile Total Work Experience Average Salary City Of Employment","excerpt":"Large and small organisations from almost every industry — from Fortune 500 companies to retail stores — are hiring data science experts now. But are these people, who hold the “sexiest job of the 21st century”, happy with their work profiles? What are their key concerns? Why is this industry facing an increasing rate of […]","categories":["AI Features"],"tags":[],"author_name":"Prajakta Hebbar","publish_date":"2019-06-06T08:56:52","publication_year":"2019","word_count":801,"keywords":["data science","Go","API","ELT","AI","RAG","Aim","analytics","GAN","R"],"extracted_tech_keywords":["AI","data science","analytics","Aim","RAG","R","Go","API","ELT","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/data-science-job-satisfaction-survey-2019\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":65350,"title":"This Is How I Created An Object Detection Model In Less Than 5 Minutes","content":"At the end of 2019, tech giant Google released a web-based tool known as Teachable Machine. The goal of this machine is to create machine learning models faster, easier, and more accessible to everyone. One best thing about this tool is that you can teach this machine using your camera and understand how the machine learning model works. In this article, I will show you how I created an object detection model with the help of Google’s Teachable Machine in less than 5 minutes. Before going hands-on let us first understand the core, i.e. how the tool makes machine learning models accessible, and finally, what are its pros and cons. Behind Teachable Machine As mentioned earlier, Teachable Machine is a web-based tool created by Google to make machine learning models faster and more accessible. The tool can be used by anyone including educators, students, innovators, artists, among others, who have a keen interest to explore and understand how machine learning models work. Teachable Machine uses Tensorflow.js, which is one of the popular libraries for machine learning in Javascript, to train and run the models you make in your web browser. The models use a technique called transfer learning, and there is a pre-trained neural network. So, whenever you create your classes, you can picture that your classes are becoming the last layer or step of the neural net. More specifically, both the image and pose models are learning off of pre-trained MobileNet models, while the sound model is built on Speech Commands. MobileNet is an efficient network architecture and a set of two hyper-parameters to build very small, low latency models that can be easily matched to the design requirements for mobile and embedded vision applications. It is trained using the popular image dataset ImageNet. How It Works In this tool, there are three kinds of projects from which you can choose – images, audios and poses. You can train the tool with images that are pulled from your webcam or image files, audios in one-second snippets from your mic, and poses where the computer guesses the position of your arms, legs, etc. from an image. So basically, you will be training your computer to recognise images, audios and poses without writing any complex machine learning code. This does not stop here. After training the model, you can also export it in three formats, which are TensorFlow.js, TensorFlow and TensorFlow Lite, and use your model in your projects, sites, apps, and much more. In Teachable Machine, I have chosen to deploy an image detection project among the three project options. Here, I used three classes – a toy, a TT racket and a bottle – each of around 200 samples. Further, I have used 50 epochs and a batch size of 16. If your data is larger, then you can increase the number of epochs to get good predictive results from your model. Also, while training the model, it is strictly prohibited to leave the tab as it may discontinue the training process. Below you can see that the model shows more than 90% accurate results for each of the classes. Pros of Teachable Machine Some of the pros of this web-based tool are mentioned below: This tool can help anyone understand how machine learning worksIt is fast and easy to access machine learning tools which do not require an in-depth knowledge of complex machine learning codesWith the help of this tool, a developer can focus more on finding the actual solution, rather than stressing on building ML models, thus increasing productivity Cons of This Machine In the above image, you can notice that the model is confused between the toy and the book, as well as the bottle and the set of pens. This is because, in case of the toy and the book, the model is capturing the popped up colours of the ball in the book, the toy, and in the case of the bottle and the pen, it is capturing the vertical sizes of the objects. This shows that not all situations in real-world can be predicted accurately by this web-based tool. Wrapping Up Teachable Machine is one of the many initiatives by tech giant Google to help us experience the magic of machine learning. If you haven’t tried it yet, I must say try it once and it is so much fun. However, if you really want to pursue a career in machine learning, you need to understand how to create machine learning models from scratch and how the algorithms work.","excerpt":"At the end of 2019, tech giant Google released a web-based tool known as Teachable Machine. The goal of this machine is to create machine learning models faster, easier, and more accessible to everyone. One best thing about this tool is that you can teach this machine using your camera and understand how the machine […]","categories":["Deep Tech"],"tags":["Google"],"author_name":"Ambika Choudhury","publish_date":"2020-05-16T15:00:00","publication_year":"2020","word_count":756,"keywords":["Go","machine learning","AI","neural network","ML","object detection","Google","JavaScript","TensorFlow","R","Java"],"extracted_tech_keywords":["AI","machine learning","ML","neural network","TensorFlow","object detection","R","JavaScript","Go","Java"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/this-is-how-i-created-an-object-detection-model-in-less-than-5-minutes\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10053370,"title":"Experts Create A 3D Model Of What The Apple Car Might Look Like Using Its Patents","content":"Car experts over at Vanarama have created an impressive interactive 3D model of what they think the Apple Car could look like based on official Apple patents. The concept shows both the external design as well as the interior of the car. The model explains design decisions using helpful notes that detail the relevant Apple patent. It might be the best look at some of the ideas Apple could incorporate into its future car project. Given it’s considered a top-priority product, the Cupertino-based tech giant is fully enforcing its secrecy, and so the company isn’t saying a single thing about the Apple Car. On the other hand, the company’s continuously growing number of patents is proof that Apple is working on something big on the automotive front. The Apple Car looks totally doable in all these digital concepts, with the interior, in particular, envisioning a new approach that would allow drivers to configure the position of controls. Vanarama’s model is fully 3D, letting you rotate the car 360º to see the front, sides, and back. The design appears to take some inspiration from the Cybertruck, albeit with rounded corners. Image Source: Vanarama The first thing you will probably notice is the pillarless design of the glass cover that includes the windows, sunroof, and windshields. The exterior of the car also uses adaptive doors from patent US10384519B1. Vanarama here took the liberty to add some additional Apple flare, like a Mac Pro-style mesh grille on the front, a soft white chassis, and retractable door handles. Image Source: Vanarama On the inside, Vanarama has modelled a large continuous touch display that extends across the entire dashboard. It uses several patents to imagine a fully customizable dashboard of controls and information, as well as a version of Siri that acts as an intelligent driving assistant. Image Source: Vanarama While the renderings look impressive, it doesn’t necessarily mean this is an idea that Apple would end up using. The tech giant is very likely to remain completely tight-lipped on its EV project, but according to people close to the source, the car could debut in 2025 if everything works according to the plan.","excerpt":"The company’s continuously growing number of patents is proof that Apple is working on something big on the automotive front.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","Apple","Data Science","Deep Learning","Design","Machine Learning"],"author_name":"Victor Dey","publish_date":"2021-11-11T16:10:25","publication_year":"2021","word_count":359,"keywords":["programming_languages:R","AI","AI (Artificial Intelligence)","Apple","Machine Learning","Git","BERT","llm_models:BERT","Deep Learning","Data Science","R","Design"],"extracted_tech_keywords":["AI","R","Git","BERT","llm_models:BERT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/experts-create-a-3d-model-of-what-the-apple-car-might-look-like-using-its-patents\/","complexity_score":2,"technical_depth":6,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10011857,"title":"Importance of Non IT Employees in IT Sectors","content":"Technology is becoming indispensable. It is disrupting every industry, even those, which had nothing to do with a tech category initially. It hardly matters if you are working for pharma, real estate, education, automotive, FMCG, retail, finance, or others, you will see new-age technology taking over some crucial processes or units for good. Moreover, traditional functions such as HR, marketing, and sales also rely on technology nowadays to either automate or ease their work in some way. Today, you need developers to build apps and IT professionals to ensure that the digital business operations go on smoothly. Besides, tech is a beacon of hope in critical times. For those who were still reluctant to implement, the COVID outbreak force fueled its adoption across sectors. From collaborating remotely to conferencing digitally, tech helped organizations to streamline their operations when they seemed impossible, and how. It can be accredited both to pandemic-related fears and technology-powered efficiencies in embracing teleworking, but maybe there will be no going back to the primitive mode of working completely. Finding ‘IT’ in Non – IT Similarly, there has been a clear approach by most product companies, as they are no more fixated on a particular year or stream of engineering. Resumes are sifted for a diverse skill set, especially soft skills that can come in handy during crises to ensure business continuity. As companies move forward with digitalization, they are accepting candidates from diverse backgrounds to develop that one perfect product. The emergence of niche technologies such as AI-ML, Blockchain, cloud computation, Big Data Analytics, IoT, cybersecurity, and others have opened lucrative avenues in not just IT but also non IT industries. Here, possessing the right coding skills is a priority. As long as the talent is productive and shows coding competency in his or her job from very early on, the less important the academic background becomes. Harvard Business Review in a ranking released in 2018 stated 42 out of the top 100 CEOs worldwide had engineering education. It only reinstates how engineers are hired for benefits that go beyond mere technical abilities. Engineers are critical and analytical thinkers who approach the problem in a structured manner; their ideas are solution-oriented and can adhere to a high-pressure environment. Hence, all it takes are the right skills to make them the right fit across industries. For instance, FMCG and supply chain companies are on the outlook for research and development professionals and experts who can build systems to extract vital information about consumer preferences through huge structured and unstructured datasets, more so after the pandemic to develop products or modify their services accordingly. Be it about contactless food deliveries, touchless payments at PoS, sanitizing the package before handing it over to the customers; there is a lot of tech-based homework going behind the scenes to understand customers’ concerns and addressing them using personal data. Yes, it is not a mere coincidence that the burger from your favourite eatery comes with a pouch of sanitizer or the leading fashion house is offering matching face masks with their latest collection. This can be vaguely summed up as reading your thoughts, though, or data to be precise. In fact, a product’s composition, its target group, channel, marketing strategy, launch date, everything is being decided by technology to maximize its reach and ultimately, business gains. This has propelled the need for competent specialists to deliver the same. Coding Competencies Will Rule Every year, close to 1.5 million Indian students get their degree in engineering. They hail from different domains such as IT, software, electrical, mechanical, and others. Only 20% find placement in pure IT firms. A lot of potential lies in the other sectors that are awaiting major technological action. With the rapid deployment of sophisticated technologies in the non-IT sector, the demand for niche technological skills has also gained traction. Thankfully, avant-garde IP-driven training institutes have taken note of these untapped avenues and are working to address the issue. They are consistently ensuring more than 40% of the students who get top tech jobs are from non-technology or computer science backgrounds. And the figure is only strengthening further. This is fostering confidence amongst a lot of non-computer science and non-IT talent that it is not the academic degree but the degree of their knowledge that matters the most. How well-versed they are in technology and prepared to be productive on the job is of major importance. This is shaping a healthy competition as people from all walks of life get a fair chance to explore opportunities and domains. Besides, it widens the scope of the IT industry as well. Who knows, different minds with distinct interests or backgrounds may code a new revolution altogether.","excerpt":"Technology is becoming indispensable. It is disrupting every industry, even those, which had nothing to do with a tech category initially. It hardly matters if you are working for pharma, real estate, education, automotive, FMCG, retail, finance, or others, you will see new-age technology taking over some crucial processes or units for good. Moreover, traditional […]","categories":["AI Features"],"tags":["engineers India","supply and demand in cyber security"],"author_name":"Narayan Mahadevan","publish_date":"2020-11-19T12:00:41","publication_year":"2020","word_count":784,"keywords":["big data","Go","API","engineers India","AI","ML","Git","supply and demand in cyber security","ViT","analytics","GAN","R"],"extracted_tech_keywords":["AI","ML","analytics","R","Go","Git","API","big data","GAN","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/non-it-is-also-the-it-sector-for-engineers\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":24312,"title":"Indian Railways To Adopt Artificial Intelligence Systems For Serving Hygienic Food","content":"Indian Railways has always been on the receiving end of criticism for its food catering services. The food hygiene and safety woes continue to trouble the government-owned monopoly. Now with the help of artificial intelligence, these problems will come to a halt, thanks to the transition it aims to bring about in trains. AI will transform the way food is prepared in the trains’ kitchens and pantries. Along with revamping food menu in trains, its catering arm Indian Railway Catering and Tourism Corporation (IRCTC) is also promoting a greener environment with its biodegradable and environment-friendly food containers. The AI project has already been kick-started at the IRCTC headquarters. The facility has installed 16 high-definition cameras which are linked to large monitors for AI vision detection. Pests in these facility will be detected instantly, and the report of these instances will be notified to higher authorities with top priority. A senior railway official at IRCTC said, “The AI system, known as Wobot, will be used to improve upon catering units. The system is capable of tracking any anomaly in the entire operation of catering. Suppose if a chef or any kitchen supervisor is not wearing their uniform, including the mandatory cap, the AI system will track that and, automatically report to the server which will then send a report to the mobile of concerned contractor immediately. If the matter is not addressed within 15 minutes, it will further be reported to IRCTC authorities in charge. If no action is taken at this level too, then it be will escalated to IRCTC MD”. This AI system will be implemented across all the IRCTC units in the coming days. Indian Railways has faced backlash from the public with many terrible, unhygienic incidents occurring in trains in the past few days. With the help of AI system in place, it hopes to eliminate all the malign practices in the food department. In addition, it has geared up to provide longer serving timings to the comfort of passengers. Last month, Piyush Goyal, Union Minister for Railways and Coal had said that AI would be used as a “tool” for the betterment of services. “AI can transform Indian Railways in terms of safety, passenger amenities, better revenues, growth and efficiency,” he had said. Goyal’s statement, “AI has to be harnessed to find digital innovations for better customer interface and better service delivery. AI is about creating trains with brains,” was met with critical acclaim.","excerpt":"Indian Railways has always been on the receiving end of criticism for its food catering services. The food hygiene and safety woes continue to trouble the government-owned monopoly. Now with the help of artificial intelligence, these problems will come to a halt, thanks to the transition it aims to bring about in trains. AI will […]","categories":["AI News"],"tags":["indian railways"],"author_name":"Abhishek Sharma","publish_date":"2018-05-07T06:44:56","publication_year":"2018","word_count":408,"keywords":["indian railways","Go","artificial intelligence","programming_languages:R","AI","innovation","Scala","Git","Aim","programming_languages:Scala","R"],"extracted_tech_keywords":["AI","artificial intelligence","Aim","R","Go","Scala","Git","innovation","programming_languages:R","programming_languages:Scala"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/indian-railways-to-adopt-artificial-intelligence-systems-for-serving-hygienic-food\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10003876,"title":"Tech Giants Grilled, Algorithms Goof-Up And More In This Week’s Top AI News","content":"This has been a tough week for the tech giants. Amidst quarterly earning postings and legal turmoils, the top companies have somehow managed to serve the community with some interesting updates and releases of their products. Take a look at all the top news that has happened over this week. Microsoft Puts An End To Its 25-Year-Old Algorithm On Tuesday, Microsoft announced that it would be discontinuing Secure Hash Algorithm 1 SHA-1,  from the Microsoft Download Center on August 3, 2020. SHA-1 is a legacy cryptographic hash algorithm that is no longer considered to be secure. So, Microsoft is asking its users to move to a currently supported version of Windows and to more robust alternatives, such as SHA-2. Big Tech Testifies In a first, the top court in the US asked the CEOs of Amazon, Apple, Facebook, and Google to testify on the same day regarding unfair competition. This is the first time Jeff Bezos had accepted to testify. This hearing is a result of a subcommittee’s 13-month probe into antitrust issues of the tech companies. The hearing, which was conducted virtually had Sundar Pichai, Mark Zuckerberg, Jeff Bezos and Tim Cook in a tight spot. “Simply put, they have too much power.”– David Cicilline, Chairman, the antitrust subcommittee The hearing saw the CEOs, reminding the Congress of all the good deeds they have been doing. Zuckerberg explained how his company is battling misinformation on their platform and how they have been successful in curbing the bad content. Whereas, the Google Chief Sundar Pichai underlined Google’s commitment to the government through various contracts with the defence department. Jeff Bezos, who is a first-timer, eloquently, listed the numbers that Amazon has put up. Right from the funds allocated to frontline workers to the number of indirect jobs they have created, Bezos quickly got down to business. These statements and the follow-up discussion by the top CEOs will determine future decisions of the Congress where the plans of breaking up monopolies have been doing the rounds for a while. Watch the full hearing here. Facebook Sues EU According to reports, Facebook is suing EU antitrust regulators for seeking ‘predominantly irrelevant’ documents that include highly sensitive personal information, such as employees’ medical information, personal financial documents and information which has got nothing to do with the Commission’s investigation. Ever since the launch of its marketplace in 2016, FB has provided 315,000 documents equivalent to 1.7 million pages to the Commission. Now, FB has turned the tables in the EU, which has put the company under scrutiny for marketplace practices. Computer Vision Algorithms Fumble Over Masks Pre-Pandemic Facial Recognition Algorithms Falter In The Presence Of Masks According to a preliminary study by the National Institute of Standards and Technology (NIST) , 89 of the best commercial facial recognition algorithms reportedly erred when they encountered masked images. For their experiments, the US government agency fed the algorithms which were trained before the pandemic with pictures of the same masked people. The drop in performance of widely used algorithms indicates that the facial recognition systems can indeed be fooled. This raises the question, what if these systems, which are installed in critical areas such as airport security, will malfunction and misidentify people? This is an ongoing survey, and more results will be reported in phases. Read more here. NVIDIA And Google Tussle For Supremacy This week, both NVIDIA and Google Cloud flaunted their MLPerf results quite vigorously. If NVIDIA claimed to have set six new records for AI performance, then Google claimed to have broken the performance records with the world’s fastest training supercomputer. Here are a few results of their performances: Source: Google Google claims that all of the above MLPerf submissions trained from scratch in 33 seconds or faster on Google’s new ML supercomputer. NVIDIA claims the above results on a single node using DGX-2H – Dual-Socket Xeon Platinum 8174, 1.5TB system RAM, 16 x 32 GB Tesla V100 SXM-3 GPUs connected via NVSwitch. For results on the scale, check this. Whereas, most of the TPU’s performance can be attributed to their version 4, which is yet to be released to the public. Know how ML Perf is revolutionising the industry here. Python And PyTorch Get New Releases Python announced new features as part of their 3.9 in beta version (3.9.0b3). In their latest release, the world’s most sought language has made significant changes. They even ditched a few of their old methods due to deprecated use. Now developers can avail the new merging method, along with new string methods and other updated modules. Know more here. In other news, Facebook too has announced the availability of PyTorch 1.6, the latest version. They have also announced that Microsoft will be taking ownership of the development and maintenance of the PyTorch to build for Windows. Read more here. Google’s Billion-Dollar FitBit Woes According to reports, Google’s wearable dreams might have to wait as the company’s $2.1 billion bid for Fitbit will now face a full-scale antitrust investigation by the EU. This probe is targeted to explore how Google will use personal healthcare data. This decision by the EU has put Google’s plans of taking over Apple and Samsung in the fitness-tracking and smartwatch market, alongside others including Huawei and Xiaomi to a hold. Google Introduces Model Cards Toolkit Over the past year, Google has launched Model Cards to enable machine learning model transparency. However, creating a Model Card takes substantial time and effort. So, now, Google AI is introducing a new tool for model transparency, Model Card Toolkit, which will provide a structured framework for reporting on ML model provenance, usage, and ethics-informed evaluation and give a detailed overview of a model’s suggested uses and limitations that can benefit developers, regulators, and downstream users alike. Know more here. Cisco Reports A Serious Data Centre Software Bug On Wednesday, Cisco released an official statement summarising the details about their findings related to a vulnerability in the REST API of Cisco Data Center Network Manager (DCNM). This vulnerability, stated the company, could allow an unauthenticated, remote attacker to bypass authentication and execute arbitrary actions with administrative privileges on an affected device. “The vulnerability exists because different installations share a static encryption key. An attacker could exploit this vulnerability by using the static key to craft a valid session token. A successful exploit could allow the attacker to perform arbitrary actions through the REST API with administrative privileges,” said Cisco in a statement.","excerpt":"This has been a tough week for the tech giants. Amidst quarterly earning postings and legal turmoils, the top companies have somehow managed to serve the community with some interesting updates and releases of their products. Take a look at all the top news that has happened over this week. Microsoft Puts An End To […]","categories":["AI News"],"tags":["best facial recognition software","best technology to learn for future","CEO","congress","limitations of AI"],"author_name":"Ram Sagar","publish_date":"2020-08-01T14:00:00","publication_year":"2020","word_count":1076,"keywords":["Go","congress","machine learning","TPU","best technology to learn for future","best facial recognition software","AI","ML","PyTorch","CEO","computer vision","Python","Aim","R","limitations of AI"],"extracted_tech_keywords":["AI","machine learning","ML","computer vision","Aim","PyTorch","TPU","Python","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/ceo-testimony-top-news-august\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":16580,"title":"10 open source Data Science and Big Data applications that are well supported by Linux","content":"Linux has been around since the mid-90s. The operating system has since reached a user base that spans several organizations and countries. The OS is present in phones, cars, refrigerators, and more. Most importantly, the OS finds use for internet and supercomputers making scientific breakthroughs. The operating system has been pretty renowned for managing hardware resources associated with desktop or laptop. Besides, it is one of the most secure and reliable operating systems available worldwide. Years of extensive research on Linux has led to the development of several open-source tools for the Linux environment. Moreover, in the age of AI and automation, the present-day AI advances are geared towards creating software and hardware which can solve day-to-day challenges in areas such as healthcare, education, security, manufacturing, banking, and more. There are several AI-based tools available in the market today that are dedicated towards Linux applications. Analytics India Magazine compiles a list of the most popular open source tools which support AI, and can be used for the Linux ecosystem. Most of these tools could possibly be used for many other operating systems, besides Linux. Apache Mahout An open-source framework from Apache, Mahout is the application of Hadoop platform in the machine learning open source framework. It helps in building scalable machine learning applications, besides corresponding to MLlib. Mahout has three primary features, as listed below: It provides simple and scalable programming environment and framework It also furnishes a plethora of prepackaged algorithms for Scala + Apache Spark, H20, as well as Apache Flink It Includes Samaras, a vector math experimentation workplace with R-like syntax, dedicated to matrix calculation Link: http:\/\/mahout.apache.org\/ Apache SystemML The machine learning algorithm uses an open source platform for big data analysis. However, its primary feature is to support R language and the Python syntax; focus on the field of big data analysis; specifically for the high order mathematical calculation. The tool’s biggest feature is the ability to automatically process assessment data, based on the framework of operation. Also, the user code should run directly on the drive or on Apache Spark cluster, which is determined according to the evaluation results. Besides, Apache Spark, SystemML also extends support for Apache Hadoop, Jupyter, Apache Zeppelin, and other platforms. At present, SystemML technology has been successfully applied in many fields. Notable use cases include automotives, airport traffic, and social banking. Link: http:\/\/systemml.apache.org\/ Caffe Caffe’s modular and expressive deep learning framework is based on speed, and the tool has been released under the BSD 2-Clause license. Interestingly, it’s already supporting several community projects in areas such as research, startup prototypes, and industrial applications in fields such as vision, speech, and multimedia. Primary features of Caffe: Fast Easy to customize Strong expansion capability Rich community support Caffe is primarily designed for neural network modeling and image processing tasks. Link: http:\/\/caffe.berkeleyvision.org\/ Deeplearning4j This open source tool provides a distributed deep-learning library for Java and Scala programming languages. The tool requires the support of the Java Virtual Machine (JVM). The tool comes integrated with Hadoop and Spark on top of distributed CPUs and GPUs, and has been designed primarily for business related applications. Deeplearning4j has opened up several algorithms to adjust the interface and the interface parameters to make a detailed explanation; which in turn allow developers to customize freely. Besides, the tool also supports matrix operations. Moreover, DL4J was released under the Apache 2.0 license and provides GPU support for scaling on AWS. It has also been adapted for micro-service architecture. Link: http:\/\/deeplearning4j.org\/ H2O The open source, fast, scalable, and distributed machine learning framework provides a large number of algorithms. It also supports smarter applications such as deep learning, gradient boosting, random forests, generalized linear modeling, and more. The businesses oriented artificial intelligence tool enables users to draw insights from their data, using faster and better predictive modeling. The core code of this framework is written by Java. Moreover, the tool focuses on large amounts of data to help enterprise users by providing fast and accurate prediction of the analysis model. Besides, the tool helps in extracting decision-making information from massive data. Link: http:\/\/www.h2o.ai\/ MLib It an open-source, easy-to-use, and high performance machine learning library developed as part of Apache Spark. It is essentially easy to deploy and can run on existing Hadoop clusters and data. It also includes the relevant test procedures and data generators. Salient features of MLib: Easy to use High performance Easy to deploy The tool currently supports a variety of machine learning algorithms such as classification, regression, recommendation, clustering, and survival analysis. Furthermore, it can be applied across Python, Java, Scala, and R programming languages. Link: https:\/\/spark.apache.org\/mllib\/ NuPIC Machine Intelligence This open-source framework for machine learning that is based on Heirarchical Temporary Memory (HTM), a neocortex theory. The HTM program helps in analyzing real-time streaming data. It learns time-based patterns existing in data, besides predicting the imminent values as well as revealing any irregularities. Notable features include: Prediction and modeling Real-time streaming data Hierarchical temporal memory Continuous online learning Temporal and spatial patterns Powerful anomaly detection Link: http:\/\/numenta.org\/ OpenCyc Cyc is the world’s largest and most complete common knowledge base and common sense reasoning engine, and OpenCyc is the open source portal based on Cyc. It contains several carefully organized Cyc entries. It finds application in areas such as: Rich domain modeling Domain-specific expert systems Text understanding Semantic data integration as well as AI games plus many more. Link: http:\/\/www.cyc.com\/platform\/opencyc\/ OpenNN It is also an open-source class library written in C++ for deep learning, and used to develop neural networks. It focuses on the realization of neural network library. However, it is optimal for experienced C++ programmers and persons with tremendous machine learning skills Characterized by a deep architecture and high performance, OpenNN can be used to implement any nonlinear model in the supervised learning scene. Besides, it also supports the design of neural networks with general approximation properties. Link: http:\/\/www.opennn.net\/ Oryx 2 Primarily a continuation to the initial Oryx project, Oryx 2 has been developed on Apache Spark and Apache Kafka. It was formerly known as Myrrix, prior to being acquired by big data company Cloudera, after which it was renamed to Oryx. Oryx 2 was re-architected on the lambda architecture, and is dedicated towards achieving real-time machine learning. It focuses on large-scale machine learning, real-time performance prediction, and analysis framework. The platform is extensively used for application development. Link: http:\/\/oryx.io\/","excerpt":"Linux has been around since the mid-90s. The operating system has since reached a user base that spans several organizations and countries. The OS is present in phones, cars, refrigerators, and more. Most importantly, the OS finds use for internet and supercomputers making scientific breakthroughs. The operating system has been pretty renowned for managing hardware […]","categories":["AI Trends"],"tags":["big data for data science","C","image processing","list of computer languages and their uses","Machine Learning India","R language","real-time integration to kafka","scala","Supercomputers India","survival regression python"],"author_name":"Richa Bhatia","publish_date":"2017-07-28T08:40:56","publication_year":"2017","word_count":1065,"keywords":["big data for data science","deep learning","artificial intelligence","R language","scala","analytics","real-time integration to kafka","machine learning","C","AI","image processing","neural network","ML","AWS","list of computer languages and their uses","Machine Learning India","survival regression python","Supercomputers India","anomaly detection","Jupyter"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","neural network","analytics","Jupyter","anomaly detection","AWS"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/10-open-source-data-science-big-data-applications-well-supported-linux\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":58868,"title":"Skip Songs Without Touching Your Phone With Google&#8217;s Soli","content":"Motion sensing technology that has long been pondered to try to integrate it with as many devices as possible. For so long motion sensing has been integrated into doors of the building where it opens, sensing the proximity of the living being walking towards it or in the motion-sensing gaming consoles. For some time now, Google has been working on integrating motion-sensing tech into their products, particularly in smartphones through which a user can perform functions through gestures, for example, skipping songs. Soli is the technology behind Google’s Motion Sense, which facilitates close proximity interaction with smartphones without contact. How Soli Works Google has made use of a custom-built machine learning model along with a data collection pipeline to design its robust machine learning model. This model helps Soli recognise a variety of movements and helps interpret the motion to gesture apps, for example, Quick Gestures in Pixel 4. Soli in Pixel 4 uses short-range radar technology. The team developed small-scale radar systems, novel sensing paradigms and algorithms from scratch specifically for the perception of human interactions. How a traditional radar works: Radar’s primary function is to detect and measure the properties of objects in a range based on their interactions with radio waves. The radar system traditionally includes a transmitter that emits radio waves that are scattered or redirected when they hit objects within their range. Some energy is reflected back and is intercepted by the radar receiver, based on these received waveforms, the radar system can detect the presence of objects as well as estimate specific attributes of the object, like the size and its distance. Soli’s radar emits electromagnetic waves in a broad beam. When objects like a hand come within the range of the radar and interact with the waves, meaning that they scatter the beam, they also reflect a certain amount of energy back to the radar antenna. The rich information like energy, frequency shift and time delay from these gestures or motions by the object help Soli identify the object’s characteristics and behaviours like size, shape, orientation, distance material and velocity. Soli processes the signal variations and captured characteristics from the object’s movements. This information makes it possible for it to distinguish between movements to understand the size, orientation, shape, distance, material and the velocity of the object, which are crucial in recognising the gestures. As implied, Soli relies on processing the temporal changes in the signals reflected back from the gesturing or movement of the objects to detect and resolve them. Below is the animation where Soli recognises the range, distance and velocity of the objects given in detail by Google here. <Gif> The Machine Learning Model And The Challenges Associated With It The resulting signals from Soli’s signal processing pipeline are then fed to Soli’s ML models for gesture classification, who are trained to recognise and accurately detect a variety of gestures with low latency. Challenges: The Soli machine learning system had to face two major problems: first was that the entire system had to go through too many unique gestures to recognise. The second was that the system encountered too many extraneous movements within the range of the sensor along with encountering an altogether different type of perception when it comes to how the entire environment looks from the POV of the motion sensor. Soli’s machine learning model consists of neural networks that were trained on millions of gestures from Google volunteers; these were mixed with other radar recordings where Google volunteers performed the generic motions made near the device. These neural networks trained on the myriad ways the gestures could be performed, give Soli’s model the right kind of robustness it needed to counter the challenges. How Soli Fits Into A Mobile Phone Classical radar systems use the spatial resolution that targets the size of the objects and distinguishes the objects based on their spatial structures. The spatial resolution requires parts that are of considerable dimensions, so, Soli’s team used a fundamentally different paradigm based on motion rather than spatial resolution. Because of this paradigm, Soli’s entire system is equipped into a phone’s (Google Pixel 4) top portion. The system, from its first prototype in 2014, where it was the size of a desktop, is now at the size of 5.0 mm x 6.5 mm RFIC, which also includes the antennas. Outlook Till now, we have seen motion-sensing related tech used more popularly in games and to open or activate doors in a building. Google and Soli’s team says that this is the first time a radar-based machine perception will be demonstrated in a mobile phone with Pixel 4 and Pixel 4 XL. Google says that the Motion Sense in Pixel smartphones and other devices shows that Soli’s potential can be extended to bring seamless context awareness and gesture recognition for implicit and explicit interactions.","excerpt":"Motion sensing technology that has long been pondered to try to integrate it with as many devices as possible. For so long motion sensing has been integrated into doors of the building where it opens, sensing the proximity of the living being walking towards it or in the motion-sensing gaming consoles. For some time now, […]","categories":["Global Tech"],"tags":[],"author_name":"Sameer Balaganur","publish_date":"2020-03-21T16:00:00","publication_year":"2020","word_count":805,"keywords":["Go","machine learning","programming_languages:R","AI","neural network","ML","programming_languages:Go","R"],"extracted_tech_keywords":["AI","machine learning","ML","neural network","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/skip-songs-without-touching-your-phone-with-googles-soli\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":5835,"title":"@RISK Optimizes Risk in Hedge Funds","content":"The world of financial markets and investments is rife with risk and reward. This is particularly true for hedge funds, sophisticated investment vehicles that are typically used only by experienced investors and firms. They are named for the fact that investors‘hedge,’ or attempt to protect their funds against volatile swings in the markets. But how are these portfolios managed? And are managers protecting and optimizing these funds as well as they should be? Seth Berlin, Principal Strategist at Performance Thinking & Technologies, a company that helps asset managers with investment operations, recently tackled these questions when working with a hedge fundclient. With the help of Palisade’s @RISK and RISKOptimizer software, Berlin was able to create a quantitative model that uses simulation to optimize a hedge fund’s portfolio. Finding the Sweet Spot When managing a hedge fund portfolio, Berlin says, there has to be a balance between maximizing profit and minimizing risk. Modern portfolio management is based on the capital asset pricing model (CAPM). In CAPM, “You theoretically can maximize return by maximizing risk.” While taking on high risk can lead to high returns over the long term, “over the short term you would have large swings in your portfolio,” says Berlin. “Everyone can maximize profits but in the end you’re dead,” meaning some level of short-term stability is necessary. The key, Berlin says, is to find an optimal portfolio that “maximizes return with the lowest incremental change in risk,” a sweet spot that avoids exorbitant risk while still yielding satisfactory returns. In his model, Berlin was simulating changes in capital markets assumptions and adjusting a mix of hedges to identify an optimal portfolio. The Long and the Short of It Berlin’s client dealt with a hedge fund portfolio comprising both long and short positions. Long positions are assets that are owned by the fund, such as stocks, which investors hope will increase in value. Short positions are ‘borrowed assets’ sold short, which increase in value if the price of the underlying assets goes down. Hedge funds use combinations of long and short positions to protect against equity market movements, currency rate changes, interest rate risk, and credit worthiness. “I’ve worked with many hedge fund managers on how they manage hedges. Anytime you can move to more quantitative than qualitative management based on simulation and optimization, it’s a win for both managers and investors.” Turning an Art into Science Berlin used Palisade’s risk analysis software, including @RISK and RISKOptimizer, to prove to his clients that quantitative measures of simulation could help identify an optimal portfolio, rather than the typical qualitative approach. He used a step-wise process to develop his quantitative model: Determine hedge ratios and constraints Determine decision variables to be modeled Determine uncertain inputs Model uncertain inputs using @RISK Output dependent on decision variables using RISKOptimizer Simulations and Smarts After going through this five-step process, Berlin had results that indicated how the hedge fund managers could change their profitability by altering the mix of long and short positions in the portfolio. “RISKOptimizer gave me an optimal mix of my short positions for the next week. Since this is theoretical, you then compare your results with what really happened in the next week,” he says. These results are helpful, but of course, they are not the only answer. “I’m not an alchemist. I’m not coming up with a secret sauce that is the model for managers to make money,” says Berlin. “What I was trying to do overall is to move from a qualitative to a quantitative way to judge a problem.” He explains that managing a hedge fund will always require qualitative decisions– “that’s what experience is”–but that well done simulations provide a method to test one’s intuition and judgment. An Underused Tool This method Berlin used has broad applications, and could be used by anyone hedging a portfolio. Yet the majority of investors don’t apply such rigorous tests. “I think everyone would say they’re doing some kind of modeling, but for the most part it doesn’t involve simulation and optimization.” Berlin believes turning to a tool like RISKOptimizer can give some certainty and structure to the process. “Based on your capital market inputs, RISKOptimizer helps you look at different mixes of assets to reach a better risk-return structure,” he says. “If I have fifty things I want to allocate money to, and I have choices on how to allocate it, you never know the right mix. RISKOptimizer and @RISK shuffle the deck a million times, and let you see how all those different combinations could play out.” » @RISK » RISKOptimizer","excerpt":"The world of financial markets and investments is rife with risk and reward. This is particularly true for hedge funds, sophisticated investment vehicles that are typically used only by experienced investors and firms. They are named for the fact that investors‘hedge,’ or attempt to protect their funds against volatile swings in the markets. But how […]","categories":[],"tags":[],"author_name":"Craig Ferri","publish_date":"2014-06-11T02:47:21","publication_year":"2014","word_count":762,"keywords":["Go","API","TPU","programming_languages:R","AI","programming_languages:Go","R"],"extracted_tech_keywords":["AI","TPU","R","Go","API","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/risk-optimizes-risk-in-hedge-funds\/","complexity_score":3,"technical_depth":7,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":61699,"title":"Significance Of Upskilling &#038; Reskilling To Overcome Covid-19 Crisis","content":"While the aftermath of the Covid-19 pandemic saw a partial – or even complete – lockdown at several places across the globe, people have been reorienting their lives indoors. Just as Covid-19 has been reshaping education with a grand global experiment in remote learning, so have businesses adapted by shifting operations online. However, some companies are creaking under pressure with these sudden changes. While this has meant job cuts in some enterprises, others are either delaying key operations or deferring hiring plans. With no end – or certainty – in sight, it necessitates that professionals turn these circumstances into opportunities to upskill. This can be made possible with the effective use of technology, and signals a massive chance for edtech providers and online career marketplaces to reinvent learning to meet the needs of a new economy. Upskilling For The Future Of Work According to MD of Udemy India Irwin Anand, in addition to Covid-induced uncertainty around the future of jobs, advanced technology and the impact of globalization also demands that professionals prioritize learning. “Learning is the new currency that can help us succeed as a workforce,” says Anand. “At a time of unprecedented change and heightened concern around recent events, the most important job skill would be the ability to learn,” he adds. With an oncoming recession and economic downturn strongly anticipated, it is critical to pay heed to warning signs that your job may be at risk. Bolstered by an automation boom and ever-changing technological landscape, the shelf life of skills have shortened. To remain employable in such competitive times, people need to upskill themselves. Given this huge clamour for upskilling – especially since the Covid-19 pandemic took root – edtech companies have been offering more industry-ready courses. Additionally, they have launched both short-term and long-term courses on emerging technologies like AI, ML and data science for professionals looking to take advantage of a job market anchored around these new technologies. “We are seeing increased demand across every segment of users – from individual learners, to repeat instructors, to employees learning for their jobs since the pandemic began,” says Anand. Nudged by current circumstances and buoyed by trends including freelancing and technology disruption, the labour market is poised to be transformed and usher in a talent economy. ALSO READ: Addressing India’s Reskilling Challenge Accelerant To Digital Transformation Whether you are a digital novice or fairly comfortable with technology, there cannot be a better time for you to dive deep into technologies, and upskilling can help in this digital transformation with the use of new applications. This will also help you ease into working remotely – another direction the changing landscape of work is moving towards. With the Covid-19 pandemic heralding increasing pay cuts and furloughs, most companies also do not have the budget for additional training. And with work processes changing, so have capabilities and expectations of employees to adjust to this scenario as fast as possible by taking ownership of their own learning. This means that a lot of upskilling needs to happen to digitally transform oneself, especially in these times. Online Learning Trends Amid Covid-19 As mentioned earlier, a clutch of edtech players and e-career companies have witnessed a jump in new registrations since the outbreak of the virus. Online career marketplace Shine Learning has seen a massive surge for online course takers over the last few weeks. According to the firm, there has been a sharp rise in the number of users choosing to upskill themselves in courses like data science, blockchain, and machine learning. “With remote working increasingly becoming the norm, we may see more demand for specific talent in this field,” says CEO of Shine Learning, Zairus Master. “Moreover, with more time and fewer entertainment options at their disposal, this is an opportune time for people to upskill themselves,” he adds. Edtech players like Upgrad, Udemy and Simplilearn are also capitalizing on increased demand. Udemy recently released the Udemy Free Resource Center – a curated collection of more than 150 free Udemy courses – to help learners upskill themselves. What is more, it also has courses and resources especially targeted at professionals for them to learn how to adapt to remote work and how to search for a job. And while there has been demand for all kinds of skills, one subset in the realm of new technologies has been piquing the interest of learners. “Data science has emerged as one of the top five topics users on Udemy choose to learn,” says Anand. “We have seen double-digit growth in learning globally as well as in India as people spend the majority of their time working remotely,” he adds. ALSO READ: Why This Is The Right Time To Learn New Data Skills Amid Covid-19 Outlook Accepting that everything is not under our control, we need to learn how to adapt to situations. And the fastest way to adapt is to be prepared through upskilling by making learning a lifelong commitment. Covid-19 could be a tipping point for learning – and working – as we know it. With virtual learning emerging as a one of the best practices in our quest to self-learn, we need to constantly keep ourselves updated on new applications and smart devices to smoothly transition into a post-Covid digital world.","excerpt":"While the aftermath of the Covid-19 pandemic saw a partial – or even complete – lockdown at several places across the globe, people have been reorienting their lives indoors. Just as Covid-19 has been reshaping education with a grand global experiment in remote learning, so have businesses adapted by shifting operations online. However, some companies […]","categories":["Deep Tech"],"tags":["Courses","covid-19","freelancer","is the tech boom over","reskilling","upskilling"],"author_name":"Anu Thomas","publish_date":"2020-04-15T12:00:30","publication_year":"2020","word_count":879,"keywords":["data science","API","machine learning","covid-19","upskilling","reskilling","AI","ML","digital transformation","Git","automation","is the tech boom over","GAN","Courses","R","freelancer"],"extracted_tech_keywords":["AI","machine learning","ML","data science","R","Git","API","GAN","digital transformation","automation"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/significance-of-upskilling-reskilling-to-overcome-covid-19-crisis\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":54876,"title":"Why A Career In Data Science Should Be Your Prime Focus In 2020","content":"The sexiest job of the 21st century, data science, has seen a lot of demand with no significant dip at any point of time over recent years. According to a study, it was found that there is a requirement of 28% more data scientists worldwide. So, it is only to fair to assume that the data science career choice in 2020 will not slow down. As for India, in 2018, demand for data scientist shot up by 417%, but the skilled professionals were not in shortage even if there was such an increase in demand. Demand For Data Science in 2020 In 2019, 2.9 million data science job openings were required and looking at the trend, 2020 will only see an increase. According to a report, in 2020, the job requirements for data science and analytics is projected to boom to by 364,000 openings to 2,720,000. And according to the U.S. Bureau of Labor Statistics, 11.5 million new jobs will be created by the year 2026. Therefore, it’s safe to say, even with the amount of shortage in talent, there might not be a dip in data science as a career option. The Hottest Job Roles and Trends In Data Science 2020 The increase in data science as a career choice in 2020 will also see the rise in its various job roles. Analytics India Magazine has listed some of them below: Data EngineerData AdministratorMachine Learning EngineerStatisticianData and Analytics Manager The increase in the job roles will be because of growth in certain sectors, and these sectors will need data science personnel to process and deal with the data which they produce. Below are some industries which will create more jobs in 2020, particularly in India: Agriculture: The backbone of our country and one of the most prominent areas to adapt technology worldwide, and therefore will demand applications of data science, ML and AI. These will help improve farm output, pricing and forecasting.Healthcare: Analysis of the massive data produced by the healthcare department will require the assistance of data science as more and more data is provided every day. Not only is this data large-scale, but also complicated, so the demand for a skilled data scientist can skyrocket in this sector.Aviation: Aviation companies produce a very massive amount of data. This data is used by them to improve their services like customer experience, fix flight schedules, safety data, route optimisation, preventive maintenance, etc.,Cyber Security: Cybersecurity has seen a recent surge with various companies levelling up their security systems with increasing malware. The data engineer demand will increase as the companies will need them for anomaly and intrusion detection uses. Industry Salary For 2020 The data science career may attract many people because of its challenging nature, but the payment related to it adds another layer to it. In India, the average salary of a data scientist as of January 2020, according to Glassdoor, is ₹1,052K\/yr, which is pretty attractive. The other job roles vary according to the companies and the position. Market Revenue In recent with the amount of shift in analytics and data science, the market revenue has increased. According to the report by AIM, in 2018, analytics, data science and big data industry in India generated about $2.71 billion annually in revenues, and in 2019 it grew to $3.03 billion in size. This 2019 figure, is expected to double by 2025 nearly. Increase In Data & Computing Power The primary factor contributing to the boom in data science sector is the increase in data being generated and at the same time, the amount of data that is being consumed as well. This data is generated everywhere from one’s mobile phones to the most significant healthcare procedures, so the big data has contributed as a major factor in spurring the growth of data science. Another factor to be considered is the increased computing power and advancement of tools. This plays an important role, because the big data, as useful as it can be, needs a powerful computing system to which one can efficiently crunch and make the data useful to a reasonable amount. There are cloud computing platforms like AWS, Cloudera, Microsoft Azure, GPS, etc., that provide intelligent solutions. Outlook If we look at the previous year, 2019, then according to a survey, 46% of the CIOs who participated said that they had suffered a shortage of skilled data scientists. So, it becomes imperative that as one has to think about this high demanding job seriously, but it is equally essential to be skilled at it. In India alone in 97,000 jobs are vacant because of the shortage of skilled employees. One has to participate in various data science boot camps, projects, and other various opportunities to upskill themselves and move towards being an expert, and then think about starting a career in data science.","excerpt":"The sexiest job of the 21st century, data science, has seen a lot of demand with no significant dip at any point of time over recent years. According to a study, it was found that there is a requirement of 28% more data scientists worldwide. So, it is only to fair to assume that the […]","categories":["AI Features"],"tags":["data science salary"],"author_name":"Sameer Balaganur","publish_date":"2020-01-30T10:00:00","publication_year":"2020","word_count":806,"keywords":["data science","machine learning","AWS","AI","cloud computing","ML","RAG","Aim","analytics","data science salary","Azure"],"extracted_tech_keywords":["AI","machine learning","ML","data science","analytics","Aim","RAG","cloud computing","AWS","Azure"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/why-a-career-in-data-science-should-be-your-prime-focus-in-2020\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10141688,"title":"Wipro Appoints Omkar Nisal as CEO for Europe Strategic Market Unit","content":"Global technology services and consulting leader, Wipro Limited has announced the appointment of Omkar Nisal as the chief executive officer for its Europe Strategic Market Unit (SMU). Effective immediately, Omkar will report to Wipro’s CEO and MD, Srini Pallia, and will join the Wipro Executive Board, succeeding Pierre Bruno, who is stepping down. Based in London, Omkar’s leadership will focus on driving growth and strengthening Wipro’s presence across Europe. Omkar, who has been with Wipro since 2012, previously headed Wipro’s Banking-EMEA business, working with financial institutions across the UK, Europe, the Middle East, and Africa. Most recently, he served as senior vice president and managing director of the UK and Ireland, managing a regional P&L exceeding $1 billion. European Expansion “Omkar’s strategic vision, combined with a strong understanding of the European market dynamics, well positions him to lead our ambitious plans for growth and expansion. With a strong customer-centric approach, Omkar will help build a resilient and adaptable organisation poised for sustainable growth in the region”, said Pallia. An honors graduate in Computer Science, Omkar remains deeply rooted in technology, with patents in Channels, Anti-Money Laundering (AML), and Fraud Detection. His expertise is expected to steer Wipro toward achieving success in the European market. Expressing his enthusiasm, Nisal remarked, “I am honored to take on the leadership of our Europe SMU, a pivotal region in the growing technology landscape. I look forward to collaborating with the incredible teams across Europe. Together, we will unlock new opportunities, strengthen our market position, and deliver outstanding results for our stakeholders.” Srini also acknowledged outgoing leader Pierre Bruno’s contributions. “I would like to thank Pierre for his leadership over the last four years, during which we made significant inroads into the European market. He will continue through the coming months, working closely with Omkar and me to ensure a smooth transition,” he said. On the AI front, Wipro has been taking strides in the generative AI space. Recently, the company launched ‘Google Gemini Experience Zone’ to boost customer engagement. Furthermore, the company also trained and certified 44,000 employees on advanced AI tools.","excerpt":"Omkar Nisal succeeds Pierre Bruno, who has served as CEO for close to four years.","categories":["AI News"],"tags":["Europe","Omkar Nissal","Srini Pallia","Wipro"],"author_name":"Vandana Nair","publish_date":"2024-11-25T20:39:23","publication_year":"2024","word_count":348,"keywords":["Wipro","Go","programming_languages:R","AI","ML","Srini Pallia","Europe","programming_languages:Go","GAN","llm_models:Gemini","generative AI","Omkar Nissal","R","fraud detection"],"extracted_tech_keywords":["AI","ML","generative AI","fraud detection","R","Go","GAN","llm_models:Gemini","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/wipro-appoints-omkar-nisal-as-ceo-for-europe-strategic-market-unit\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10120419,"title":"Project Navarasa Takes Center Stage at Google I\/O","content":"Just a few days ago, we wrote about how Gemma outperformed Meta’s Llama 3 for Indic languages. Today, at Google I\/O, India’s Project Navarasa took centre stage, highlighting the use of Gemma, making it accessible for 15 Indic languages. Google highlighted the success of ‘Project Navarasa,’ a multilingual variant of Gemma for Indic languages developed by Telugu LLM Labs founded by Ravi Theja Desetty and Ramsri Goutham Golla. Harsh Dhand, head of APAC research partnerships at Google said, “When technology is developed for a particular culture, it won’t be able to solve and understand the nuances of a country like India.” Project Navarasa leverages Gemma’s powerful tokenizer to enable AI-driven language generation for 15 Indic languages. “One of Gemma’s features is an incredibly powerful tokenizer which enables the model to use hundreds of thousands of words, symbols and characters across so many alphabets and language systems. This large vocabulary is critical to adapting Gemma to power projects like Navarasa,” Golla said. “Our biggest dream is to build a model to include everyone from all corners of India,” said Golla, saying that Navarasa is a model trained for Indic languages, and a fine-tuned model based on Google’s Gemma. He said they built Navarasa to create culturally rooted large language models where people can talk in their native language and receive responses in their native language. Many developers that AIM spoke to said that Gemma is better than Llama for Indic languages. “Gemma shines compared to the Llama 2 and 3 models,” said Adithya S Kolavi, founder of Cognitive Lab, who built a leaderboard for Indic LLMs. “Models using Llama 2 extended its tokenizer by 20 to 30k tokens, reaching a vocabulary size of 50-60k. Continuous pre-training is crucial for understanding these new tokens. In contrast, Gemma’s tokenizer initially handles Indic languages well, requiring minimal fine-tuning for specific tasks,” explained Kolavi. According to Vivek Raghavan, the co-founder of Sarvam AI, Gemma’s powerful tokenizer gives it an advantage over Llama when it comes to Indic Languages. He explained, “The tokenization tax for Indic languages means asking the same question in Hindi costs three times more tokens than in English, and even more for languages like Odiya due to their underrepresentation in these models.” Meanwhile, OpenAI recently released GPT-4o, an update to their language model that includes a new tokenizer and an extended vocabulary size of 200k tokens, compared to 100k tokens in GPT-4. This update significantly improved the support for several Indian languages, including Hindi, Gujarati, Marathi, Telugu, Tamil, and Urdu. Although Gemma 2’s tokenizer limit wasn’t clearly mentioned in the demo, it is stated that the model can handle ‘hundreds of thousands of words, symbols and characters’. In comparison, GPT-4o’s 200k base tokenizer so far outperforms Gemma for Indic and non-English languages in terms of token reduction. At Google I\/O, the tech giant today introduced PaliGemma, a powerful open vision-language model (VLM), and provided a sneak peek into the upcoming Gemma 2, the next generation of their Gemma family of models. Now you can try out our Indic Gemma Model Navarasa 2.0 (supports language generation in 15 languages) easily as a chat interface at https:\/\/t.co\/KFQ6qfWBf0Ask a question in English and ask it to respond in Hindi, Telugu etc or ask directly in the native language.Kudos to… pic.twitter.com\/YuHniHo5s4— Ramsri Goutham Golla (@ramsri_goutham) April 2, 2024 Google Unveils Open Vision Language Model, PaliGemma PaliGemma, inspired by PaLI-3 and built on open components, including the SigLIP vision model and the Gemma language model, is designed for class-leading fine-tune performance on a wide range of vision-language tasks. These tasks include image and short video captioning, visual question answering, understanding text in images, object detection, and object segmentation. Google is providing both pre-trained and fine-tuned checkpoints at multiple resolutions, as well as checkpoints specifically tuned to a mixture of tasks for immediate exploration. PaliGemma is available through various platforms and resources, including free options like Kaggle and Colab notebooks, and academic researchers can apply for Google Cloud credits to support their work. The release of PaliGemma brings several key benefits, such as multimodal comprehension, a versatile base model for fine-tuning on a wide range of vision-language tasks, and off-the-shelf exploration with a checkpoint fine-tuned on a mixture of tasks for immediate research use. Several have started experimenting with it already. I tried it with some plant disease images. It could identify the crop, but it would refuse to detect plant diseases. Found this example quite funny: pic.twitter.com\/bASP74bgMn— Thomas Friedel (@thomascygn) May 14, 2024 More Power to Gemma Looking ahead, Google announced the upcoming arrival of Gemma 2, the next generation of Gemma models. Gemma 2 will be available in new sizes for a broad range of AI developer use cases and features a brand-new architecture designed for breakthrough performance and efficiency. Key benefits include class-leading performance, reduced deployment costs, and versatile tuning toolchains. At this year’s developer conference, Google literally poked fun at OpenAI by making it clear that it is making AI helpful for everyone, not just for him or her.","excerpt":"Google launched PaliGemma, an open vision-language model for extensive vision-language tasks, along with a preview of Gemma 2, featuring capabilities for Indic languages.","categories":["AI News"],"tags":["Gemma","Google"],"author_name":"K L Krithika","publish_date":"2024-05-15T02:55:53","publication_year":"2024","word_count":835,"keywords":["Go","Gemma","OpenAI","AI","GPT-4o","RAG","Colab","GPT","Aim","object detection","Google","R"],"extracted_tech_keywords":["AI","GPT-4o","OpenAI","Aim","Colab","RAG","object detection","R","Go","GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/project-navarasa-takes-center-stage-at-google-i-o\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10071061,"title":"The Misguided Perception of Big Tech Bulldozing SMBs","content":"In November 2019, the Committee on Small Business at the House of Representatives held a hearing titled, “A Fair Playing Field? Investigating Big Tech’s Impact on Small Business” where the stakeholders deliberated on how  Big Tech companies have greatly impacted the small firms who are now essentially relying on their business models. The then-chairwoman, Nydia Velazquez, pointed out the market dominance held by the big tech companies. In his testimony, Dharmesh Mehta, Vice President of Amazon explained how  “Amazon lowers barriers to entry for entrepreneurs, helps make retail even more vibrantly competitive, and continues to delight customers with new innovations.” Many tech giants often claim that they give startups and small and medium enterprises (SMEs) what they essentially need, like instant access to vast markets, cheap and reliable infrastructure, efficient ads and more. Notwithstanding these benefits, they also get in the way of SMEs’ success in many ways. At times, the big firms acquire the small ones to root out competition. For instance,  shortly after Instacart (an online grocery delivery business) partnered with Whole Foods (a supermarket chain), Amazon launched its own grocery delivery service and acquired Whole Foods. Similarly, Google acquired the satellite navigation app ‘Waze’ which was once a potential competitor of Google Maps. In other instances, these tech giants seem to neglect the small businesses in favour of large enterprise customers who have adequate resources to afford the services offered by big techs. Consequently, small businesses lose out on the innovative digital products and latest technologies the big tech players extend. Does that mean small businesses can never realise their potential for success? Or that they are doomed to be perpetual losers in want of resources? Not necessarily. Large tech companies can indeed help small businesses thrive significantly in a resource-efficient manner. Product-led growth strategy For the past few years, tech leaders have been routinely discussing Product-led growth, a go-to-market strategy where the “end-user product experience is the primary driver of growth.” Product-led growth (PLG) strategies help founders build their brand, allow them to set prices based on market demand and attain better customer satisfaction by demonstrating more customer input into product development. It is the PLG approach that led to the success of ‘Zoom’ and ‘Dropbox’, applications widely used during the pandemic. At the time of its foundation,  Zoom was anticipating competition with Microsoft, Cisco, Adobe and other well-known firms. Success in competition with such established leaders comes rarely to SMEs. However, a well-executed PLG strategy directed Zoom’s massive success. The first version was released in January, 2013. By the end of the month since their launch, Zoom had 400,000 users which then rose to 1 million users by end of May, 2013. By 2015, Zoom’s customer base touched 100 million users. Founder and CEO Eric Yuan invested heavily on customer centric features, like one-on-one meetings, group video conferences, screen sharing, the ability to record meetings and have them automatically transcribed and integrated with Slack and other softwares. The story of Dropbox, a file hosting service, is similar in some measure. Officially launched in 2008, a product-driven approach led the brand to its early success. By April 2009, it had reached a registered user base of 1 million. In 2021, Dropbox crossed 700 million registered users. Dropbox’s product-led approach led the founders to design a product that made file sharing easy and accessible for end-users. Additionally, the founders incorporated certain features that enhanced the appeal of the product among prospective users. For instance, the moment a user sends another user a Dropbox link, the user in receipt of the link is able to open it to access shared documents hassle-free. When tech companies prioritise product-led growth and deliver swift and easy-to-use products, small businesses are able to access the technology they need through self-service. Heightened focus on smaller businesses benefits the bottom line by expanding a company’s addressable market and incentivising better products. Dedicated hand-holding Tech giants have come up with dedicated programmes for startups in an effort to handhold them in their initial growth years. This practice is of utmost importance considering the failure rate of small businesses in their formative years.  According to the Bureau of Labor Statistics’ Business Employment Dynamics, approximately 20 per cent of small businesses fail within the first year and 50 per cent eventually face failure within five years. An example of such a programme is ‘NVIDIA Inception’, a free programme designed with a focus on startups. The programme provides startups with access to cutting-edge technology, NVIDIA experts, connections with venture capitalists and co-marketing support that increase their visibility and help them evolve faster.  The programme supports all stages of a startup’s life cycle. Under NVIDIA Inception, members are provided with the best technical tools, latest resources, and opportunities to connect with investors. The success of this programme is reflected in its vast memberships. Earlier this year, Inception surpassed 10,000 members across 110 countries. Another such sought-after programme is the ‘Google for Startups’, a Google initiative to support thriving, diverse and inclusive startup communities around the world. Under this programme, Google helps startups connect with the right people, the right products and relevant best practices that consequently help startups thrive and grow. The statistics reflect the success of Google’s initiatives. Startups have created more than 100,000 jobs in the Google for Startups campuses and raised USD 6.7 billion in 2020. The Google for Startups programme has aided numerous startups in their ventures. For instance, the Google for Startups Accelerator programme helped the co-founders of ‘Hypd’, an India-based creator-driven marketplace, in perfecting the business idea and form of their product. Hypd enables content creators to set up online stores that match with their content. “We understand from the Google Analytics team how to understand the content creator’s journey, the key features they need from the product and which priorities to build on. That has shaped our product”, says Akshay Bhatnagar, co-founder, Hypd. The tools that big tech giants provide as a part of their dedicated handholding programmes impact SMEs’ performance in a noteworthy manner. The following infographic demonstrates the scope of their impact on SME performance. (Infographic source: Deloitte) Innovative financing SMEs usually have less access to capital and cash reserves which renders it difficult for them to access cutting edge technologies. Big techs have come up with innovative ways to help small businesses access funds. For instance, in the wake of the pandemic, Google came up with an initiative called ‘Ad credits for Google Ads Small and Medium-sized Businesses’. These ad credits could be used by SMEs to offset payments for advertisements on the Google Ads platform to attract online customers to their businesses or make new digital offerings. The Information Technology Industry (ITI) Council, a global advocate for technology that includes some of the most prominent tech companies across the globe like Amazon, Apple, Adobe, Google, Meta, IBM and others, introduced the ‘Paycheck Protection Program’ for startups and small businesses. Through this programme, small businesses were able to obtain the required funding. Additionally, ITI members have developed tools to provide small businesses with software and online tutorials to apply for and obtain funding with more ease. Power of the small In the report, ‘The Power of  Small: Unlocking the Potential of SMEs’, the International Labour Organisation (ILO) deliberates on the global prevalence of SMEs and their relevance in socio-economic and environmental developments. SMEs might appear too small for big tech firms to collaborate with and aid in the development of their capabilities as they would with larger enterprise customers. However, understanding their level of maturity and their specific needs in every country and across market segments while providing solutions is pivotal. “A deeper focus on small businesses empowers underserved and underrepresented groups, ensuring their ideas and their innovation can become a part of our socioeconomic fabric too”, notes Gabe Monroy, Chief Product Officer at DigitalOcean.","excerpt":"Product-led growth strategies help founders build their brands","categories":["Deep Tech"],"tags":[],"author_name":"Zinnia Banerjee","publish_date":"2022-07-15T16:00:00","publication_year":"2022","word_count":1306,"keywords":["Go","API","AI","RPA","innovation","Git","Aim","analytics","GAN","R"],"extracted_tech_keywords":["AI","analytics","Aim","R","Go","Git","API","GAN","RPA","innovation"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/the-small-firms-who-are-now-essentially-relying-on-their-business-mode\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10121748,"title":"Coforge Builds GenAI Platform Quasar, Powered by 23 LLMs","content":"Though Indian IT companies do not extensively disclose specific revenue figures related to generative AI, analysts approximate its current contribution to be between 1% and 3% of their total revenue. The figures are, of course, estimated to only increase in the coming years. We have seen numerous examples of generative AI integration and reskilling efforts by Indian IT giants. Coforge has developed an AI platform called Quasar, powered by 23 LLMs, which includes commercially available models, such as OpenAI’s GPT series models and Google’s Gemini, as well as open-source LLMs like LLaMA. Quasar offers six accelerators tailored to specific AI capabilities – Quasar Document AI, Quasar Speech AI, Quasar Predict AI, Quasar Vision AI, Quasar Graph AI, and Quasar Conversational AI – available on the Microsoft Azure Marketplace. “It’s crazy. Quasar is not one solution platform but has over 100 solutions and capabilities built into it,” John Speight, the customer success officer at Coforge, told AIM in an exclusive interaction. It boasts a collection of over 100 APIs, featuring a modular and scalable architecture. Additionally, it offers a library of more than 100 pre-built cognitive and generative use cases. “Quasar is not only packaged with generative AI capabilities, which we have built with our IP, but also other aspects of AI like vision models,” Speight said. The second aspect is the wide array of capabilities it offers across various domains of AI expertise. It allows users to leverage the models of their choice and build their own use cases. This includes document processing, image analysis, and speech recognition. It can also develop predictive and prescriptive models. Biggest GenAI challenge is not tech When leveraging an LLM, there could be a multitude of challenges associated with it, for instance, keeping the models grounded and reducing hallucinations. However, Deepak Bagchi, vice president – AI Practice at Coforge, said technology-related challenges are not their biggest issues. “The real challenge lies in addressing the discussions surrounding AI capabilities. Customers often harbour concerns about AI’s impact and which models to employ, whether from hyperscalars or open-source platforms, while also factoring in cost implications,” he told AIM. Indeed, the primary challenge for SaaS or IT companies integrating AI into their product remains change management. The experts AIM spoke to earlier, too, indicated something similar. “Convincing users and decision-makers, who may fear job displacement or the perceived omnipotence of AI, poses the greatest challenge,” Bagchi pointed out. Using generative AI for onboarding Nonetheless, despite the apprehension, we are also seeing great adoption. Generative AI today is a part of most IT companies’ projects and so is the case with Coforge. The Noida-based IT company is helping a lot of customers in the contact centre space. One of Coforge’s major offerings is contact centre automation for banks, covering multi-channel communications like voice, text messages, and emails. “Our generative AI-powered solution analyses incoming queries, provides insights, and enables self-assist, reducing the need for human agents. This helps in cost optimisation by lowering the number of support tickets raised, which comes at a cost,” Bagchi said. Moreover, Coforge is leveraging generative AI to assist clients in onboarding both customers and employees. At the time of this discussion, Bagchi was in France helping a large client develop a generative AI-powered solution for their onboarding processes. As a result of this, according to Bagchi, accessibility has increased for their customers. Generative AI reduces the need for direct human involvement in many onboarding tasks. For instance, generative AI can facilitate the process without requiring direct person-to-person interaction when transferring knowledge or providing information about products to new employees. “This improves accessibility, allowing individuals to learn at their own pace and ask questions as needed. New employees or customers facing challenges can utilise generative AI to get timely assistance and information about products,” Bagchi added. Speight added that Bagchi has trained over 100 senior leaders on generative AI while working for a big customer in France. “He is utilising Quasar to help them contextualise the technology and build their own AI solutions.” Does Indian IT need to build LLM from scratch? While Quasar leverages the most advanced LLMs available, the question arises of whether Indian IT firms must develop their own foundational model. Bagchi’s outright answer is no. First, it’s a huge investment, and second, IT companies often deal with customers from multiple industries. “The domains we work with are quite diverse. Building a specific LLM from scratch isn’t practical, so we focus on enhancing existing open-source LLMs like LLaMA. Our investment lies in fine-tuning and retraining these models for specific purposes,” he said. However, a few IT firms are building their own LLMs from scratch for different reasons. Hinduja Global Solutions, the IT management division of the billion-dollar Hinduja Group, for instance, is building its own industry-specific proprietary foundational models, which it will use for various internal and external requirements. Moreover, Tech Mahindra is also building Indic LLMs from scratch in languages such as Hindi and Bengali, called Project Indus. The company’s desire to develop LLMs stems from former head CP Gurnani’s desire to develop a foundational language model deeply rooted in Indian culture and languages.","excerpt":"It’s crazy. Quasar is not one solution platform but has over 100 solutions and capabilities built into it.","categories":["AI Features"],"tags":["AI in Indian IT","Interviews and Discussions"],"author_name":"Pritam Bordoloi","publish_date":"2024-05-27T16:00:00","publication_year":"2024","word_count":853,"keywords":["AI in Indian IT","GenAI","OpenAI","AI","R","RAG","document AI","Ray","Aim","generative AI","Azure","Interviews and Discussions"],"extracted_tech_keywords":["AI","generative AI","GenAI","OpenAI","Aim","Ray","RAG","document AI","Azure","R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/coforge-builds-genai-platform-quasar-powered-by-23-llms\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10072920,"title":"Is the Tesla House Really Out Of Order?","content":"Tesla is all set to launch its Full Self Driving (FSD) Beta 10.69 on August 20. A few days ago, Elon Musk, CEO and chief architect at Tesla, announced some major updates on FSD. He said that FSD Beta 10.13 would be released as version 10.69 and there will be several improvements to the same. In the shareholders’ meeting, Musk assured that people can avail FSD this year. With the AI day at Tesla due on September 30, expectations are mounting on the company to come up with the progress made with respect to FSD.  For Tesla, there seems to be too much work at hand. After all, a product release is not a trivial affair, let alone cutting-edge technological products like Tesla FSD Models. However, last week a report on the failure of Tesla’s FSD technology started doing the rounds and has put a question mark on Tesla’s FSD. Tesla’s latest FSD model put to test In July 2022, the Dawn Project that campaigns to make computer systems safe for humans, released its test results related to Tesla’s FSD technology. The scientific test was conducted in June 2022 at Willow Springs International Raceway in Rosamond, California, by a professional test driver on a professional driving course. Cruising at an average speed of 25kmph, a Tesla 2019 Model 3 with FSD Beta 10.12.2 (the most updated software version available then) failed to detect the presence of a child-sized mannequin on the road and repeatedly hit it in a way that would be fatal. The results contradict Tesla’s long-standing claims that its FSD technology is safe. In 2021 Elon Musk, CEO and product architect at Tesla, publicly claimed on multiple occasions that Tesla’s  FSD technology is one of the best in terms of safety standards. It’s amazing by most standards, but we are aiming for 1000% safer than the average human driver— Elon Musk (@elonmusk) August 24, 2021 The build I’m driving right now is almost there. FSD 10 will blow your mind.— Elon Musk (@elonmusk) September 9, 2021 In fact, in an interview with the Financial Times last year, Musk claimed to be the CEO who cared about the safety of the planet the most. A dangerous software Calling the FSD software “the most dangerous commercial software ever released onto public roads”, the Dawn Project has demanded that the product be taken off the market. It is asking people to sign a petition that may eventually lead the US Congress to call it off. Tesla in a mess The safety test result seems to have put Tesla in a mess shortly before its big day. Moreover, the National Highway Traffic Safety Administration (NHTSA) is undertaking two investigations into Tesla cars. In June, the NHTSA expanded an ongoing preliminary investigation (PI) to an engineering analysis (EA) into Tesla’s autopilot system comprising 830,000 Tesla cars across four models – Model Y, Model X, Model S, Model 3. The initial investigation was opened in August 2021, backed by several crashes in which Tesla vehicles, operating with Autopilot engaged, struck stationary in-road or roadside first responder vehicles tending to pre-existing collision scenes. Now, the EA seeks to explore the degree to which Autopilot and associated Tesla systems may exacerbate human factors or behavioural safety risks by undermining the effectiveness of the driver’s supervision. The NHTSA is undertaking another investigation into the “phantom braking” problem in Tesla cars. The problem cropped up after Tesla decided to do away with the forward-looking radar sensor from its Model 3 and Model Y. Thereafter, these models have been relying entirely on cameras for Advanced Driver Assistance Systems. However, certain issues came up – the issue of automatic emergency braking function, to be specific. In the newer radar-less models, automatic braking is being activated very frequently due to false positives like the shadow of a car. The NHTSA received several complaints related to this issue that drove them to carry out this investigation. The investigators undertook a preliminary evaluation of Tesla’s Passenger Play feature that allowed passengers to play games on the infotainment display while the car is in motion. However, following the opening of the PI, Tesla came up with a software update that disabled the feature while the vehicle is in motion. With several investigations underway, a test result of this sort seems to have put the Tesla house out of order. The “staged” test Well, a report of this sort has definitely provided the opponents of autonomous vehicles to strengthen their voices.  However, given the popularity of Musk and the anxiety related to FSD technology,  the test has been bashed as propaganda and staged. Our new safety test of @ElonMusk’s Full Self-Driving Teslas discovered that they will indiscriminately mow down children. Today @RealDawnProject launches a nationwide TV ad campaign demanding @NHTSAgov ban Full Self-Driving until @ElonMusk proves it won’t mow down children. pic.twitter.com\/i5Jtb38GjH— Dan O'Dowd (@RealDanODowd) August 9, 2022 Shortly after the Dawn Project released the safety test result, Fred Lambert, editor-in-chief of Electrek, in a blog claimed that the driver had actually failed to activate FSD Beta during the test. As a footnote to his blog, Lambert notes that since the controversy, the Dawn Project released additional footage that doesn’t appear in the ad, and is inconsistent with the results published about the test. In fact, a Twitter user even conducted a similar test and uploaded a video showing how Tesla detected the presence of a cardboard child and avoided it every time. https:\/\/twitter.com\/tesladriver2022\/status\/1557152108071342085 The accusation of the ad campaign being “staged” seems valid in the backdrop of   the Dawn Project clearly mentioning the issue of Tesla’s FSD program as its primary campaign agenda. The campaign website clearly mentions, “The first danger we are tackling is Elon Musk’s reckless deployment of unsafe Full Self-Driving cars on our roads.”","excerpt":"Calling the FSD software “the most dangerous commercial software ever released onto public roads”, the Dawn Project has demanded that the product be taken off the market","categories":["AI Features"],"tags":[],"author_name":"Zinnia Banerjee","publish_date":"2022-08-17T17:00:00","publication_year":"2022","word_count":963,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","RAG","BERT","Aim","llm_models:BERT","GAN","R"],"extracted_tech_keywords":["AI","Aim","RAG","R","Go","BERT","GAN","llm_models:BERT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/is-the-tesla-house-really-out-of-order\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10162957,"title":"GitHub Copilot adds Agent Mode, expands AI capabilities","content":"Developer platform GitHub has introduced Agent Mode for GitHub Copilot, giving its AI-powered coding assistant the ability to iterate on its own code, recognise errors, and fix them automatically. “Today, we are infusing the power of agentic AI into the GitHub Copilot experience, elevating Copilot from pair to peer programme,” wrote GitHub CEO Thomas Dohmke on X while announcing the launch. Alongside this, Copilot Edits is now generally available in Visual Studio Code, and GitHub has provided a first look at Project Padawan, an autonomous AI-driven software engineering agent. “The name [Copilot] reflects our belief that artificial intelligence (AI) isn’t replacing the developer. Instead, it’s always on their side. And like any good first officer, a Copilot can also fly by itself,” read their statement. Agent Mode Now in Preview Agent Mode enhances Copilot’s ability to handle complex coding tasks autonomously. It can refine its own output, detect and fix errors, suggest terminal commands, and analyse runtime issues with self-healing capabilities. To access it, developers need VS Code Insiders and must enable the Copilot Agent setting. Copilot Edits Now Available in VS Code Now available for all, Copilot Edits lets developers edit multiple files using natural language prompts. The AI makes inline changes, streamlining workflow across projects. It supports multiple AI models, including OpenAI’s GPT-4o, Anthropic’s Claude 3.5 Sonnet, and Google’s Gemini 2.0 Flash. A Visual Studio 2022 preview is also available. GitHub is also developing Project Padawan, an AI Software Engineering (SWE) agent capable of handling coding tasks, reviewing code, and automating workflows. Developers will be able to assign GitHub Issues to Copilot, which will generate fully tested pull requests and respond to reviewer feedback. The system will operate in a secure cloud sandbox. As announced, GitHub plans to expand Agent Mode to more IDEs and improve Project Padawan. Developers can try these features in VS Code Insiders and share feedback through GitHub’s official channels. Last month, at the Microsoft AI Tour in Bengaluru, GitHub’s director of international developer relations, Karan MV, showcased how Copilot Workspace can comprehend and generate code in Indian languages, including Hindi and Kannada. Dohmke also highlighted how Copilot Workspace is transforming software development in local languages. “As always, Karan MV demonstrated the power of Copilot Workspace—building with agents in Kannada. This is the way. Every developer will conduct a symphony of AI agents in natural language,” he said.","excerpt":"“Today, we are infusing the power of agentic AI into the GitHub Copilot experience,” said GitHub CEO Thomas Dohmke.","categories":["AI News"],"tags":["AI (Artificial Intelligence)"],"author_name":"Aditi Suresh","publish_date":"2025-02-07T09:34:50","publication_year":"2025","word_count":394,"keywords":["Anthropic","agentic AI","artificial intelligence","Gemini 2.0","OpenAI","AI","TPU","GPT-4o","ML","Claude 3.5","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","ML","agentic AI","GPT-4o","OpenAI","Claude 3.5","Anthropic","Gemini 2.0","TPU"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/github-copilot-adds-agent-mode-expands-ai-capabilities\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":62506,"title":"How This Berlin-Based Startup Is Using Deep Learning To Drive Growth In Customer Intelligence","content":"Customers who engage with brands on multiple levels across channels are producing a lot of data. As channels and data production increases, so does the complexity to which brands must navigate to make sense of this data and make it work for them. Couple these challenges with growing data privacy and security regulations, and brands must leap over them to reap benefits. For brands, the ability to use ML-led-analytics has allowed them to detect the future propensity of their customers. For example, the ability to detect those customers who are more susceptible to up- and cross-selling opportunities, as well as those who would most likely become a customer. This precision and added customer insight have allowed brands to increase ROI on enhanced strategies. With such a vision, Berlin-based Zeotap is utilising AI and machine learning to provide 360-degree customer understanding and increase ROI on their marketing budgets. For this week’s feature, we caught up with Projjol Banerjea from Zeotap to gain deeper insights on how the startup is using emerging technologies to drive growth. Founded in Berlin in September 2014 by Daniel Heer, Projjol Banerjea and Stephan Schwebe, Zeotap today has ten offices, including in New York, London, Madrid, Milan, Bengaluru and Mumbai. Zeotap is a customer intelligence platform, and its unique capabilities include a global identity resolution and unified data asset, enabling its clients to achieve better results through precision at scale and intelligent technology. The overall data asset at Zeotap spans more than 1 billion consumer profiles across North America, Latin America, Europe, and India. How Flagship Product Is Different From Others In The Market To this question, Banerjea answered that the stack of the company is interoperable and modular to make it customisable for varying use-cases (from customer activation to cross\/up-selling to customer retention). It is well-integrated with the entire Martech ecosystem and has seamless workflows across different pre-existing systems among publishers, sell- or buy-side platforms, and other data platforms. The flagship products of Zeotap are outlined below: Connect: This product allows a brand to digitise and consolidate a client’s first-party data for a unified customer view. This entails merging the offline and online data worlds, namely emails and phone numbers from CRMs and POSs with data from website\/app and advertising channels. Enrich: This product enables brands to unlock the full power of their first-party data by combining it with high-quality third-party data so that customer trigger points can be better understood. Activate: Activate allows brands to increase ROI through better prospecting, churn prevention and re-targeting. How Zeotap Uses AI & ML Zeotap Uses AI and ML to increase the scale of certain segments through lookalike models. It also increases the quality of the data by taking multiple signals into account, removing selection bias in analytics and increasing the accuracy of estimations and relevancy of searches. The company uses machine learning techniques in all of its products. In fact, many products use a combination of these techniques. Some of the uses of machine learning are mentioned below: Improving Data Quality: Zeotap uses machine learning to benchmark the quality of the data partner’s data. They use a combination of MCMC, Bayesian Inference, and Deep Learning to achieve this and have recently filed a patent regarding this combination.Increasing Scale: The company uses Graph-based machine learning to build lookalike models based on multiple signals. Correcting Selection Bias In The Database: Zeotap uses a combination of Heckman correction, satisfied sampling, and non-linear curve fitting to produce true analytics from the data concerning the country population.Search: The company uses search in their products, for which they use machine learning methods to understand and explore the data-catalogue taxonomy and find relationships between query terms with a set of taxonomy candidates and their dependencies. Core Tech Stack According to Banerjea, Zeotap’s customer intelligence platform is an end-to-end solution built in a modular fashion for maximum use-case scenarios. Technically, the data pipelines are largely based on Apache Spark, and the backend is a microservices-based architecture with the services developed overplay framework, dockerized and deployed on Kubernetes. The frontend is developed over Angular 8 framework while the charting and dashboarding uses D3.js. Tackling Hiring Phase To this question, Banerjea said, “Hiring top talent is a priority for us as we put our product and good customer experience at the centre of what we do. Currently, we’re looking for talent who are motivated, innovative, trustworthy, and agile to join our growing team in ten offices around the world.” He added, “In Bangalore, we have a team of more than 60 data scientists and engineers, and we’re always on the lookout for new hires who will add to our product expansion and development.” Roadmap In the next few years, the goal of Zeotap is to become the global leader in customer intelligence. The company has recently launched a robust software layer that supplements the former product. They are now working on building out the third tier, which is an analytics layer that complements the software and data layers as well as a universal ID solution that sits vertically across all three layers of the stack.","excerpt":"Customers who engage with brands on multiple levels across channels are producing a lot of data. As channels and data production increases, so does the complexity to which brands must navigate to make sense of this data and make it work for them. Couple these challenges with growing data privacy and security regulations, and brands […]","categories":["AI Startups"],"tags":["AI Benefits","Deep Learning","entity resolution software"],"author_name":"Ambika Choudhury","publish_date":"2020-04-24T16:00:00","publication_year":"2020","word_count":847,"keywords":["machine learning","AI","entity resolution software","ML","docker","AI Benefits","Apache Spark","microservices","deep learning","analytics","Deep Learning","R","kubernetes"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","analytics","kubernetes","docker","microservices","Apache Spark","R"],"url":"https:\/\/analyticsindiamag.com\/ai-startups\/how-this-berlin-based-startup-is-using-deep-learning-to-drive-growth-in-customer-intelligence\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10057547,"title":"IIT- Madras to launch a Master&#8217;s programme on electric vehicles","content":"The Indian Institute of Technology Madras will launch a Master’s Programme in Electric Vehicles. An Interdisciplinary Dual Degree (IDDD), it will boost the students’ engagement and enhance the research capabilities in this field. Students can enrol for this programme from January 2022 during their third year of B Tech and Dual Degrees. The initial intake will have about 25 students. The students graduating from this programme will have the skill sets required to pursue job opportunities in EV Product Development, including EV Integration, Vehicle Aggregate Engineering, Communication and Calibration, Verification and Validation, and Product and Portfolio Planning. Prof T Asokan, Head, Department of Engineering Design, IIT Madras, said, “The course will be the result of nearly eight departments collaborating to induct the skills required for a student to engineer Electric Vehicles. The content offered is carefully curated to build sufficient depth in each domain, starting from vehicle basics and going to very specific EV aggregates including batteries and motors.” IIT Madras provides its undergraduate students with an option to upgrade to IDDD programmes. The students will study for five years and obtain B Tech in a parent discipline and M Tech in an interdisciplinary area. The objective is to provide greater flexibility to students, allowing them to take courses across disciplines and build towards expertise in modern interdisciplinary areas that will define the future of engineering and technology. Prof Karthick Athmanathan, Professor of Practice, IIT Madras, said, “These are exciting times for the eMobility sector. We are clearly at the tipping point. It is important to have more resources skilled in various aspects of EV Engineering and eMobility. IIT Madras has had a dominant presence in this sector through various initiatives and centres for the last few years. We are now starting this phase where formal programmes for EVs and e-Mobility are being offered with very close engagement with Industry. Such initiatives are going to help India achieve and retain the lead technologically in the medium and long term.” Over the years, IIT Madras has been at the forefront of building capabilities in the country for eMobility, through its various centres and programmes, including the Centre for Battery Engineering and Electric Vehicles (CBEEV), as well as incubating a host of start-ups in the eMobility space over the years through the IIT Madras Incubation Cell.","excerpt":"This programme will boost the research capabilities in the exciting field of E-Mobility and will have students enrolling in January 2022 during their third year of B Tech and Dual Degree Programmes","categories":["AI News"],"tags":["electric vehicles","IIT Madras"],"author_name":"Poornima Nataraj","publish_date":"2022-01-03T15:04:53","publication_year":"2022","word_count":384,"keywords":["Go","programming_languages:R","AI","IIT Madras","programming_languages:Go","electric vehicles","R"],"extracted_tech_keywords":["AI","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/iit-madras-to-launch-a-masters-programme-on-electric-vehicles\/","complexity_score":2,"technical_depth":5,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10164826,"title":"OpenAI Launches GPT-4.5, Runs Out of GPUs","content":"After weeks of waiting, OpenAI has finally introduced GPT-4.5, its latest and largest AI language model.  It was internally referred to as Orion. It is in research preview for ChatGPT Pro users, offering better writing, knowledge, and a more natural, less hallucinatory experience. GPT-4.5 is first being made available to ChatGPT Pro users, with Plus and Team users gaining access next week, followed by Enterprise and Education users. “GPT-4.5 is ready,” posted OpenAI CEO Sam Altman on X. “It is a giant, expensive model. We really wanted to launch it to plus and pro at the same time, but we’ve been growing a lot and are out of GPUs.” Altman announced that tens of thousands of GPUs will be added next week for the Plus tier, with hundreds of thousands more coming soon, all of which are expected to be fully utilised. In the recent NVIDIA earnings call, CEO Jensen Huang said the company’s inference demand is accelerating, fuelled by test-time scaling and new reasoning models. “Models like OpenAI’s, Grok 3, and DeepSeek R1 are reasoning models that apply inference-time scaling. Reasoning models can consume 100 times more compute,” he said. GPT4.5 was only trained with pretraining, supervised finetuning, and RLHF, so this is not a reasoning model. It is an extension of the GPT series of models, unlike the o series of models. According to the website, the pricing for input tokens is $75.00 per million tokens, while cached input tokens are available at a reduced rate of $37.50 per million tokens. The cost for output tokens is $150.00 per million tokens. “This isn’t a reasoning model and won’t crush benchmarks. It’s a different kind of intelligence and there’s a magic to it I haven’t felt before.” added Alman. The model is more computationally efficient, and offers a tenfold improvement over GPT-4. In the livestream, the OpenAI team outlined the evolution of GPT models – from GPT-1 was barely coherent, to GPT-3.5 became the first truly useful model, to GPT-4.5 continuing this trend with incremental enhancements. Recently, Anthropic released its Claude Sonnet 3.7  and xAI launched its Grok 3, competing in the same space. Altman had previously announced the roadmap for GPT-5. OpenAI’s goal is to combine its large language models to eventually create a more capable model that could be labeled as artificial general intelligence, or AGI.","excerpt":"Sam Altman announced that tens of thousands of GPUs will be added next week, with hundreds of thousands more coming soon.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","OpenAI"],"author_name":"Aditi Suresh","publish_date":"2025-02-28T10:36:46","publication_year":"2025","word_count":389,"keywords":["Anthropic","ChatGPT","Grok 3","RLHF","GPT-5","OpenAI","AI","GPT-4.5","DeepSeek R1","xAI","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","GPT-4.5","GPT-5","ChatGPT","OpenAI","Anthropic","DeepSeek R1","Grok 3","xAI","RLHF"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/openai-launches-gpt-4-5-runs-out-of-gpus\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10122703,"title":"After Scarlett Johansson, OpenAI GPT-4o Now Mimics Disney Characters","content":"OpenAI had barely emerged from the Scarlett Johansson controversy when the company released a new demo featuring the GPT-4o model’s ability to generate voices for a range of Disney-like characters, including animals like a snake, an owl, and a fox. In the demo, the model was asked to generate the sound of a wise and stoic owl, acting as an advisor to the lion in the jungle. The model’s voice modulation and the owl’s wise tone appear to be inspired by Disney movies, specifically the owl from Winnie the Pooh. Similarly, when the model makes a squeaky sound like a mouse, one immediately remembers the popular Mickey Mouse character. Interestingly, when asked to produce an evil laugh and suggest which animal it would suit, the model responded, “For a villain with that kind of laugh, maybe a slithery snake or a cunning fox.” Disney had all these characters spread over several movies, for instance, Kaa, the snake in The Jungle Book. On the other hand, the fox seems to be inspired by Honest John from Pinocchio. The voice, though similar to Disney-like characters, gives the studio no right to sue them. The best part about GPT-4o is that it can generate a wide range of pitches, tones, and accents. To put it in CEO Sam Altman’s words, “The model is fluid”. “Disney doesn’t have the copyright on anthropomorphised mice with high-pitched voices – that’s a basic concept that predates Disney’s existence by thousands of years at a minimum,” commented a user on YouTube. However, there is no doubt that this new feature puts the careers of voice actors and dubbing artists at risk. “OpenAI is going after voice actors now,” posted a user on X. Another one humorously referenced Johansson’s fiasco, saying, “Scarlett: It sounds like me! OpenAI: Joke’s on you, it can sound like anyone.” OpenAI are going after voice actors now: https:\/\/t.co\/pVGedy6nNx— Ian Wootten (@iwootten) June 5, 2024 Meanwhile, Pixar recently announced its plans to lay off about 175 employees, or roughly 14% of its workforce. These cuts are part of Disney’s chief Bob Iger’s initiative to prioritise quality over quantity in the studio’s content. It’s interesting that in 2021, Johansson sued Disney, claiming the studio breached her contract by releasing ‘Black Widow’ on Disney+, the same day it premiered in theaters. If Disney were to sue OpenAI now, it would all come full circle. The drama would any day be better than Disney movies. Disney X OpenAI Earlier this year, OpenAI unveiled Sora, its generative AI video model capable of producing hyper-realistic videos. There’s a strong possibility that Disney could partner with OpenAI to create content, following the trend seen in news media publications of partnering with OpenAI and licensing their content. The world’s first commissioned music video created entirely through OpenAI’s Sora, The Hardest Part, directed by Paul Trillo, was also recently released. “Walt Disney himself was a big believer in using technology in the early days to tell better stories. And he thought that technology in the hands of a great storyteller was unbelievably powerful,” said Disney chief Bob Iger at a recent Canva event. “Don’t fixate on its ability to be disruptive — fixate on [tech’s] ability to make us better and tell better stories. Not only better stories but to reach more people,” Iger added. Lately, OpenAI has also been pitching Sora to Hollywood and other entertainment giants. The AI startup has been actively arranging meetings in Los Angeles with Hollywood studios, media executives, and talent agencies. Most recently, actor and investor Ashton Kutcher lauded OpenAI’s Sora model. “I’ve been playing around with Sora, this latest thing that OpenAI launched that generates video,” Kutcher said. “I have a beta version of it, and it’s pretty amazing. Like, it’s pretty good,” he gushed. With the combined features of Sora and GPT-4o voice capabilities, the future is not far from where it would be very easy to make short films and cartoon series. The Tribeca Film Festival 2024 will feature five AI-generated short films created in collaboration with OpenAI’s Sora. As Kutcher puts it, “There’s going to be more content than there are eyeballs to consume it. Any one piece of content is only as valuable as you can get people to consume it. The bar is going to have to go way up. Why will you watch my movie when you could just watch your own movie?”","excerpt":"There’s a strong possibility that Disney could partner with OpenAI to create content, following the trend seen in news media publications of partnering with OpenAI and licensing their content.","categories":["Global Tech"],"tags":["OpenAI"],"author_name":"Siddharth Jindal","publish_date":"2024-06-06T16:00:10","publication_year":"2024","word_count":733,"keywords":["Go","OpenAI","AI","GPT-4o","GPT","Aim","llm_models:GPT","generative AI","R","startup"],"extracted_tech_keywords":["AI","generative AI","GPT-4o","OpenAI","Aim","R","Go","GPT","startup","llm_models:GPT"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/after-scarlett-johansson-openai-gpt-4o-now-mimics-disney-characters\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":13469,"title":"On Pi Day 2017, math lovers celebrate the irrational number and NASA puts out a Pi in the sky","content":"Today is Pi Day, see what math lovers around the world are celebrating the never ending number Today is the iconic Pi Day 2017, the cornerstone of mathematics or simply put the “ratio of the circumference of a circle to its diameter — which is approximately 3.14159.” Relegated to circles in schools, the infinite nature of Pi makes it challenging to remember. According to mathematician Steven Strogatz who wrote in previous The New Yorker edition why pi deserves a massive celebration is because it “puts infinity within reach”. According to Strogatz pi digits are never ending and do not reveal a pattern either. Strogatz elevated the status of pi to a cosmic wormhole, joining two separate mathematical universes via the number theory linking circles and geometry. The Pi Day 2017 celebration isn’t just confined to glitzy eulogies — the storied Princeton has an annual Pi Day 2017 party that also coincides with Albert Einstein’s birthday, also March 14. NASA’s Pi Day contest spurs young minds to apply the mathematical constant in everyday problems the space agency faces In one of the most interesting line up of events – NASA has set an engaging classroom activity to jog young minds titled Pi In the Sky. The educational activity entails using the “famed mathematical constant pi to solve real-world science and engineering problems – such as calculating the angle of crater impacts on the red planet Mars; measuring the size of the shadow that will fall on North America during the 2017 total solar eclipse; determining the orbital period of the Cassini spacecraft during its final weeks around Saturn; and finally finding a habitable zone around TRAPPIST-1, a star housing seven Earth-sized planets. Here’s a bit of fun fact — Pi Day has been celebrated by mathematicians and maths lovers around the world since it was first marked at the San Francisco Exploratium in 1988. Princeton’s Pi Day party is more fun and also celebrates Albert Einstein’s birthday What’s more, the San Francisco Exploratorium is celebrating the 30th annual Pi Day 2017 and father of relativity, Einstein’s birthday with day long events that are open to all. Events range from processions  to Pi shrine, Pi-themed antics and of course plenty of pie. Another notable events is ongoing Festival of Numbers at Singapore’s Science Centre with Singapore Prime Minister joining in the celebrations by enlisting some brain teasers. One of the most common puzzles is trying to memorize as many as digits as you can.","excerpt":"Today is the iconic Pi Day 2017, the cornerstone of mathematics or simply put the “ratio of the circumference of a circle to its diameter — which is approximately 3.14159.” Relegated to circles in schools, the infinite nature of Pi makes it challenging to remember. According to mathematician Steven Strogatz who wrote in previous The […]","categories":["AI News"],"tags":[],"author_name":"Richa Bhatia","publish_date":"2017-03-14T10:38:29","publication_year":"2017","word_count":412,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","Git","BERT","llm_models:BERT","ViT","CLIP","R"],"extracted_tech_keywords":["AI","R","Go","Git","BERT","CLIP","ViT","llm_models:BERT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/pi-day-2017-math-lovers-celebrate-irrational-number-nasa-puts-pi-sky\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10121506,"title":"Prof Pushpak Bhattacharyya Named Chairman of  National Committee on Indian Language Standards","content":"The Ministry of Electronics and IT has appointed Prof Pushpak Bhattacharyya from IIT Bombay as the chairman of the National Committee on Indian Language Standards. This committee will evaluate encoding, fonts, search performance, and language technology applications for all 23 official languages of India. The committee’s mandate will last for one year. In an exclusive interview with AIM, Bhattacharyya said that he has been working on the emotional and sentimental problems in NLP since his master’s at IIT Kanpur, and has published over 350 papers. “What got me interested in linguistics, emotions, and AI was the similarities between the words of different languages and their respective sounds,” he said. Bhattacharyya said that he is also working on Plutchik’s wheel of emotions with eight emotions at Centre for Indian Language Technology (CFILT) lab at IIT Bombay, which is a subsequent work of his recent paper, ‘Zero-shot multitask intent and emotion prediction from multimodal data’. This problem deals with combining different types of emotions within one context, a foundational problem that he said no one has taken up before. “At our CFILT lab, we take up problems that no one else has before, which includes not just Indian languages,” he added. Bhattacharyya emphasised on building a trinity model for creating Indic language models, which means deciding one language, one task, and one domain for creating models. “For example, creating a model in Konkani for question answers on agriculture or a sentiment analysis system for railway reservation in Manipuri,” he explained, saying that these models are easier to build and then can be connected later into a larger model. Starting his BTech at IIT Kharagpur studying digital electronics, Bhattacharyya came across a circuit board made for adding two numbers. Unlike others who did not think of it much, he was astounded at how a lifeless system made of diodes and resistors had decision-making capabilities. This got him into studying intelligence outside bodies of human beings and animals, leading him to AI. “I’m one of the few NLP researchers who give equal importance to linguistics and computation,” he beamed. He added that his course is inspired by both the fields and how his master’s thesis was also focused on Sanskrit to Hindi machine translation.","excerpt":"This committee will evaluate encoding, fonts, search performance, and language technology applications for all 23 official languages of India.","categories":["AI News"],"tags":["Indian languages","Large Language models","small language models"],"author_name":"Mohit Pandey","publish_date":"2024-05-24T12:07:02","publication_year":"2024","word_count":370,"keywords":["Go","programming_languages:R","AI","Large Language models","Indian languages","sentiment analysis","Modal","Git","RAG","NLP","Aim","small language models","R"],"extracted_tech_keywords":["AI","NLP","Aim","RAG","sentiment analysis","R","Go","Git","Modal","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/prof-pushpak-bhattacharyya-appointed-as-the-chairman-of-the-national-committee-on-indian-language-standards\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10066144,"title":"Anaconda brings out PyScript: Python in your browser","content":"“Python in a browser” is here!- Yes, you read that right. At PyCon US 2022, Anaconda, Inc. has introduced PyScript, which allows users to create Python applications in the browser by using a mix of Python with standard HTML. https:\/\/twitter.com\/anacondainc\/status\/1520447158603890691 Anaconda, Inc. was founded by Peter Wang and Travis Oliphant around a decade ago. The company is behind the release of Anaconda, a distribution of the Python and R programming languages for data science, ML, predictive analytics, etc. In a blog, Fabio Pliger, principal architect at Anaconda, Inc., discussed PyScript in detail. He mentioned that PyScript allows a user to run many popular packages of Python and the scientific stack like NumPy, pandas, scikit-learn, etc. It will also provide bi-directional communication between Python and JavaScript objects and namespaces. Finally, it will allow the users to define what packages and files to include for the page code to run and provide a flexible framework that can be used to create and share new pluggable and extensible components directly in Python. Pliger adds, “The core concept of PyScript, as a framework, is to provide a set of [opinionated] components and tools that allow users to quickly create and share their applications. We also don’t want to reinvent the wheel and aim to re-use the great work that many others are already doing.” Reactions PyScript has received positive feedback from data scientists, developers and tech enthusiasts as of now. Nir Barazida, data scientist at DAGsHub, adds, “I think PyScript is a game-changer. It opens a world of possibilities to Python programmers and machine learning practitioners in particular. We can now easily wrap our models in HTML applications, share them with the world and have the end-users run them on the browser.” WOW! You can also render Altair charts directly in the browser with pyscript My head is going to explode with all the amazing possibilitieshttps:\/\/t.co\/o87LCSXyUb pic.twitter.com\/76mLQTvUwR— Hamel Husain (@HamelHusain) May 1, 2022 Excited to see https:\/\/t.co\/OhatkRLd0H from @anacondainc A well-supported effort to put Python in the browser has so much benefit.Immediately, the powerful Dashboards and ipython widgets which we write in Javascript today can and should be moved to Pyscript— Soumith Chintala (@soumithchintala) April 30, 2022 This past weekend there was a major announcement by the @anacondainc team at #PyCon2022 – the launch of PyScript, an in-HTML interface for Python. This is something that many of us in the Python community have been hoping for for a long time. #Python #Web #Programming1\/6 pic.twitter.com\/e81H58VSZg— Bojan Tunguz (@tunguz) May 2, 2022 Sachin Birla, senior consultant at EY, says, “Data scientists will be a big beneficiary of PyScript. Most of the time, data scientists prepare models but cannot make them representable to stakeholders. PyScript will enable them to use the full potential of their capabilities and thought processes. Most of the models and charts are not even consumable by targeted audiences just because of the lack of proper UI, but PyScript might play a crucial role in overcoming all shortcomings. It will boost the productivity of the data science team and reduce expenses of organisations to hire frontend team.” Tech stack behind it Pliger, in the same blog, also discusses many elements behind this offering. These include: WebAssembly (WASM) – WASM is a binary instruction format for a stack-based virtual machine, designed as a portable compilation target for programming languages that helps in deployment on the web for client and server applications. Emscripten – It is a complete compiler toolchain to WebAssembly, using LLVM, with a special focus on speed, size, and the Web platform. Image: Anaconda, Inc. Pyodide – It is a Python distribution for the browser and Node.js based on WebAssembly. With this, we can install and run Python packages in the browser with micropip. Many packages with C extensions have also been ported for use with Pyodide. It includes many general-purpose packages like RegEx, PyYAML, lxml and Python packages, including NumPy, pandas, SciPy, Matplotlib, and scikit-learn. Opportunities PyScript has the potential to be as popular as Javascript in the web developer community. Sachin lists down some opportunities PyScript can bring in: Narrow down the need for language set: Web developers need a diverse set of programming languages, such as HTML, JavaScript and any backend language like Java or Python. PyScript subsumes to learn such skills.Create new roles and opportunities: Syntax of Python is so easy, and anyone can learn and understand it quickly. So, many new entrants will be attracted to it.Scope of cutting edge innovation: Many innovative products don’t take shape just because people are not exposed to frontend language.Explore new dimensions of AI: Python is the default language for AI\/ML solutions. PyScript can create a new milestone in the field of AI, which has not been explored yet. Sachin adds, “It’s a new offering. It might face adaptation issues in the development community. Compared to JavaScript, it lacks in large public libraries.”","excerpt":"A user can run many popular packages of Python and the scientific stack like NumPy, pandas, scikit-learn, etc.","categories":["Deep Tech"],"tags":["HTML","Python"],"author_name":"Sreejani Bhattacharyya","publish_date":"2022-05-04T15:00:00","publication_year":"2022","word_count":811,"keywords":["data science","scikit-learn","NumPy","machine learning","AI","ML","HTML","Python","Aim","analytics","Matplotlib","Pandas"],"extracted_tech_keywords":["AI","machine learning","ML","data science","analytics","Aim","scikit-learn","Pandas","NumPy","Matplotlib"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/anaconda-brings-out-pyscript-python-in-your-browser\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10086380,"title":"Indian Publishers’ Love-Hate Relationship with ChatGPT","content":"The use of ChatGPT by publishers has sparked a heated debate in the industry. But the question is how many of them are really using Generative AI to aid their work? If you ask the purists, they would sneer at the idea, turn up their noses, turn around and leave. But if you ask ChatGPT, it will tell you that media publications are using the chatbot in news summarisations, creating chatbots, content generation, improvisation, social media management and much more. “I personally use the AI-powered tool DALL.E to produce images for my stories. The use of ChatGPT in content improvisation is a call the editor has to take. Also, there is a need for a discourse around its judicious use. The tool is here to stay, so the publishers should adopt the tool to improvise their offerings,” said Imtiaz Shariff, news editor at Deccan Herald. The Better India, a popular Indian digital media platform that focuses on human-impact stories, told AIM that it leverages ChatGPT for its social media marketing efforts. Moreover, it uses AI-powered tools like Midjourney and DALL-E to produce images for their campaigns. “Tools like ChatGPT have helped us save time and cost, enabling us to focus on our diverse efforts. In the coming days, we are planning to use a blend of AI and human-generated content for the platform,” said Dhimant Parekh, founder and CEO of TBI. Another prominent digital media platform The Logical Indian’s founding editor and head of factcheck, Bharat Nayak, was also optimistic about the usage of the popular tool. He shared, “Editors can use the tool to tweak their content to keep it in alignment with the platform’s requirements and also to hasten the publishing of content. We don’t have a policy on its usage as yet at The Logical Indian, but I do use it personally.” Some of the other media friends we spoke to feel that tools like ChatGPT play an important role in reducing human dependency and cutting costs. The tools act as an enhancer and can greatly help in content refinement. Mainstream publications in India, excluding The Times of India, The Hindu, and a few others, said that they are yet to react to its usage in their operations. However, a few of them said that they are using the tool for content improvisation and marketing, if not for generative purposes. The naysayers The News Minute, a popular independent digital media platform that covers developments from South India, told AIM that AI-powered tools like ChatGPT are not yet completely trustworthy when it comes to fact-based content. “We have no plans as of now to integrate ChatGPT or any other AI into our editorial processes and have not formulated any policy on its use,” said Ragamalika Karthikeyan, editor-special projects and experiments at TNM. A similar stance was taken by Prema Sridevi, founder of The Probe – an independent investigative news media platform. Interestingly, they tried using ChatGPT to create newsletters but later discontinued it because Sridevi felt that tools like these only help if the focus is on quantity, not quantity. It didn’t align with their investigative stories and ground reports that require human efforts at their core. However, they are still exploring to see how such tools can be leveraged in the future to enhance their work. Regional digital media platforms, like Newsmeter that covers the states of Andhra Pradesh and Telangana, are not very optimistic of ChatGPT’s usage either. The platform, which offers ground reports and fact-checked stories, feels that human intervention and storytelling cannot be enhanced by technology and AI. ChatGPT usage policy for media While scores of publishers are using ChatGPT and similar AI-powered tools these days, there is still a lack of transparency regarding its usage. Hence, there is a need for publishers to be forthcoming on its usage to ensure judicious use and cull its misuse, more from an ethical standpoint. Despite being optimistic about the use of tools like ChatGPT for publications, some media firms like The Better India and The Logical Indian, and also Analytics India Magazine, for that matter, are still mulling the need for a policy as they plan to use the tool in their content publishing efforts. On the other hand, The News Minute, The Probe and Newsmeter are still contemplating the ChatGPT transition. What’s the next step? As the debate around the judicious use of AI-powered tools rages, organisations like Springer Nature, one of the leading academic scientific journals, have formulated guidelines pertaining to using AI in research manuscripts. This comes in the wake of the scientific publisher community mulling on instituting similar solutions for themselves. Last week, the journal announced that softwares like ChatGPT couldn’t be recognised as an author in studies published in its thousands of publications. It has also mandated that researchers using LLM tools or AI chatbots should document the use in the ‘methods or acknowledgements’ sections. On the other hand, Buzzfeed, the New York-headquartered internet media company, is gearing up to use ChatGPT to generate content for its website. This may be in the form of polls and quizzes initially. In India, many of these proponents argue that ChatGPT can significantly improve efficiency and productivity, freeing up journalists to focus on more in-depth reporting and analysis. They also highlight the potential of ChatGPT to reduce the cost of content creation and to help meet the growing demand for content in the digital age. Meanwhile, there are publishers who are not very optimistic about its use and argue that the technology could have a negative impact on the quality of journalism. They fear that the use of the popular tool could homogenise the media landscape, with all content sounding similar and lacking the unique and original perspective gained through original reportage and in-depth research that human journalists bring to the table. Our take It is in the wake of this discourse that Analytics India Magazine, one of India’s leading publications for the latest developments in AI and analytics, has instituted a policy regarding the use of AI-powered tools like ChatGPT and alike to enhance their content. AIM, of course, believes in the power of technology and how it can help our journalists and in-house experts focus on bringing unique ideas to the table. As a proprietor of tech innovations, we have recently been experimenting with ChatGPT tools, mainly to enhance our content quality and improve the experience for our readers.","excerpt":"While scores of publishers are using ChatGPT and similar AI-powered tools these days, there is still a lack of transparency regarding its usage","categories":["IT Services"],"tags":["ChatGPT","MidJourney"],"author_name":"Aparna Iyer","publish_date":"2023-02-02T10:00:00","publication_year":"2023","word_count":1067,"keywords":["ChatGPT","MidJourney","AI","chatbots","Git","RAG","Aim","generative AI","analytics","Rust","R"],"extracted_tech_keywords":["AI","analytics","generative AI","ChatGPT","Aim","RAG","chatbots","R","Rust","Git"],"url":"https:\/\/analyticsindiamag.com\/it-services\/indian-publishers-love-hate-relationship-with-chatgpt\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":36647,"title":"5 Ways To Negotiate Your Data Scientist Salary","content":"via GIPHY Over the years, the technological space in India has transformed — the nation is witnessing a significant rise in terms of both companies and the development of technology. Today, almost every major enterprise is setting up their development offices in India while they keep hiring top graduates from good universities. And Data Science is a prime area that is playing a vital role — as companies are also setting up their Data Science divisions in Indian offices. With this significant rise in the industry, the demand for data scientists has also skyrocketed in order to building AI or Data Analysis based products or service pipelines. However, don’t forget that the field is growing rapidly, and you need to stay up-to-date with all the latest trends. If you are someone who works in the field of data science and is looking for a new data science or analytics job. Then this might be the right time to go for it. But one can be tough to negotiate is the paycheck. It is nothing new that a job change is not just about taking on new responsibilities or expanding your skills, but it is also about a hike in your salary. According to one of our studies, there has been a 2% increase in the numbers of analytics jobs that offer more than 15 Lakh annual salary as compared to 2017. So for that, you might want to try negotiating your salary, but you also have to maintain a positive relationship with your potential employer. How would you do that? via GIPHY Here Are Five Tips To Negotiate Your Data Science Salary Do Your Research Before Negotiating This is the first and foremost thing to keep in mind before you go ahead with the process of negotiating your salary for the data science job. And to do that research, you can look up Glassdoor and LinkedIn. These platforms are always one of the best to get know about the current salary trend for a particular job role. So, utilise this research to support your negotiating technique. It is always considered to be best practice to figure out your market value before getting into a conversation with your potential employer regarding your salary.  Being prepared will not only boost your confidence level but will also help you set reasonable expectations. Word to wise: Even though data science is one of the highest paying verticals, you cannot expect a high paycheck from every company. Also, sometimes it always better to look for opportunities to grow and learn rather than money. Be Honest About Other Offers With A Significant Hike There is no denying to the fact that when someone looks for a job change, they start applying to multiple companies and they do receive multiple offers. It is not an easy task to juggle multiple job offers. However, if you are smart enough, you can turn those offers as a strong point while attending an interview at a company. While negotiating your salary for a data science job role, you can always let them know that there are companies you are willing to meet your salary expectations. Be honest and establish a trust that works as an advantage. So, let them know, even though you have good offers in your hand why you choose them. Have The Skill Set That None Can Reject When you apply for a high paying job like Data Scientist, you need to have the knowledge and the skill sets that works completely for your potential employer. Data science might sound really fancy to many, however, it is not one of the spaces where you can land a job easily. So turn your skill sets into strong relevant points. Tell them about the projects you have worked on before, the nuts and bolts you know about the data science space — show what you bring to the table and the impact you can have on your potential employer’s business as well to the clients. Therefore, it is advised to show your skill set and why you deserve the amount you’re asking for. Choose Your Words And Tone Wisely While Negotiating Your Salary While negotiating your salary, always make sure that your tone is cooperative — don’t react to the initial amount with an adversarial tone. Rather, take a moment to evaluate the offer, reply positively and politely telling the recruiter about the number you had in mind (and also add the reason behind). However, if the company comes back at you with a “No”, just ask if there is any wiggle room. Word to the wise: Sometimes, the offer truly is what it is, and in that situation, all you need to do is to make a decision — whether you want to accept the position at that income level. Look For Other Employee Benefits to Negotiate If the negotiation of salary fails, check out whether there are any employee benefits and perks that can be negotiated. If benefits like remote working are there, then consider negotiating these benefits instead of your salary — it may give you the chance to cut back on child care expenses, commuting costs or housing fees.","excerpt":"via GIPHY Over the years, the technological space in India has transformed — the nation is witnessing a significant rise in terms of both companies and the development of technology. Today, almost every major enterprise is setting up their development offices in India while they keep hiring top graduates from good universities. And Data Science […]","categories":["AI Trends"],"tags":["Data Science Jobs","data science salary","data scientist salary"],"author_name":"Harshajit Sarmah","publish_date":"2019-03-20T11:03:08","publication_year":"2019","word_count":864,"keywords":["data science","Go","API","data scientist salary","AI","programming_languages:R","programming_languages:Go","Data Science Jobs","ViT","analytics","data science salary","Rust","R"],"extracted_tech_keywords":["AI","data science","analytics","R","Go","Rust","API","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/5-ways-to-negotiate-your-data-scientist-salary\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10166694,"title":"Meet The Winners: Best Firms for Diversity &amp; Inclusion in Tech Awards","content":"AIM Media House successfully concluded the seventh edition of The Rising 2025, India’s leading DEI summit in technology and AI, last week at the NIMHANS Convention Centre in Bengaluru. The event underscored the critical role of diversity, equity, and inclusion (DEI) in shaping a more inclusive and progressive tech industry. The Rising, in partnership with AIM Best Firm Certification, celebrates tech companies’ achievements and dedication to building a more equitable future in the industry. Our ‘Best Firms for Diversity & Inclusion in Tech Awards’ recognises organisations that have effectively championed diversity, ensuring equitable career advancement. This year AIM received several nominations. Let’s take a closer look at some of the standout winners. Automation Anywhere Automation Anywhere is a global leader in AI-driven automation, helping businesses streamline processes and boost efficiency. With a strong focus on innovation and inclusivity, it creates a supportive workplace where employees can thrive. Recognised for its expertise and impact, the company continues to drive meaningful change across industries. Axtria Axtria helps life sciences companies use data science and technology to connect patients with the right treatments at the right time. As a leading global provider of cloud software and analytics, it supports pharmaceutical, medical device, and diagnostics companies in turning data into meaningful action. Centific Global Technologies India Centific is a leading AI data foundry that helps businesses harness the full potential of their data through platform-driven solutions. With a strong focus on data quality, the company enables the development of smarter, safer, and more scalable AI models. By generating and analysing large-scale datasets, Centific accelerates AI research and development. Backed by top data scientists and strategic industry partnerships, it is committed to ethical, sustainable AI that addresses real-world challenges. Commonwealth Bank of Australia With over 17 million customers, Commonwealth Bank of Australia (CBA) is one of the country’s largest banks. Its India team plays a key role in shaping the bank’s future, bringing together skilled technology and banking operations experts. Working collaboratively, they drive innovation and deliver exceptional experiences for CBA’s customers. Honeywell International Honeywell is a global technology company focused on automation, aviation, and energy transition. Supported by the Honeywell Accelerator and Forge IoT platform, it delivers solutions through aerospace technologies, industrial automation, building automation, and energy & sustainability. With strong research and manufacturing capabilities, Honeywell advances AI, cybersecurity, and automation to drive efficiency, safety, and sustainability. In India, it plays a key role in digital transformation, aligning with global megatrends to shape the future of industry and technology. Insight Enterprises Insight Enterprises is a Fortune 500 Solutions Integrator that delivers end-to-end IT solutions by integrating hardware, software, and services. With over 35 years of experience and a global network of 8,000+ partners and experts, it helps clients solve technology challenges efficiently and securely. Recognised as a top employer by Forbes and Fortune, Insight has offices in Gurugram, Noida, Bengaluru, Thiruvananthapuram, and Pune. Renault Nissan Technology & Business Centre India Renault Nissan Tech is a captive automotive technology and business centre supporting Renault and Nissan worldwide. Established in 2007 in Mahindra World City, Chengalpattu in Chennai, it employs over 10,000 professionals across engineering, IT, and global business services. With locations in Chennai, Bengaluru, and Hyderabad, it is one of India’s largest OEM automotive tech centres. It was recognised as a Great Place to Work (2024-25), a Top Employer (2025), and among India’s Best Companies for Women (AVTAR & Seramount) for seven consecutive years. MathCo (TheMathCompany) TheMathCompany is a global enterprise AI and analytics firm enabling data-driven decision-making for Fortune 500 and Global 2000 companies. Founded in 2016, it develops custom AI and analytics solutions through a hybrid model. Its proprietary platform, NucliOS, offers pre-built workflows and reusable modules, driving connected intelligence at a lower TCO. MathCo fosters a transparent, collaborative work culture, prioritising capability and growth, making it a recognised employer of choice. Unisys India Unisys is a global technology solutions company driving innovation and business transformation. It helps organisations enhance performance and profitability through expertise in digital workplace, cloud, applications, enterprise computing, and business process solutions. By applying advanced technologies, Unisys strengthens teams, optimises processes, and enables organisations to seize new opportunities. Tesco Technology Tesco Technology is a global team of 5,300+ professionals across the UK, Poland, Hungary, the Czech Republic, and India. It develops and manages software, systems, and infrastructure to power one of the world’s largest retailers. By driving innovation in retail, supply chains, and sustainability, it creates impactful solutions for customers, colleagues, and the planet. Broadridge India Broadridge Financial Solutions is a global fintech leader providing investor communications and technology-driven solutions to the financial services industry. Its platforms process over $10 trillion in daily securities trading and generate more than 7 billion communications annually.","excerpt":"The Rising, in partnership with AIM Best Firm Certification, celebrates tech companies’ achievements for a more equitable future.","categories":["AI Highlights"],"tags":["rising"],"author_name":"Aditi Suresh","publish_date":"2025-03-26T15:56:11","publication_year":"2025","word_count":784,"keywords":["data science","API","AI","ML","Scala","Git","Aim","analytics","rising","Nuclio","R"],"extracted_tech_keywords":["AI","ML","data science","analytics","Aim","Nuclio","R","Scala","Git","API"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/meet-the-winners-best-firms-for-diversity-inclusion-in-tech-awards\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10067599,"title":"When remote working is a business model","content":"Founded in March 2018, Turing.com is an AI-powered deep jobs platform. It currently covers 100 technologies and 15 job titles, ranging from entry-level roles to engineering directors and CTOs. “Our mission is to unleash the world’s untapped human potential. Companies all over the world are going remote and are on the lookout for top-notch engineering talent,” said Vijay Krishnan, co-founder and CTO at Turing.com. In an exclusive interview with Analytics India Magazine, Vijay spoke about how their AI-powered platform enables companies to hire and manage remote developers with a push of a button. AIM: What impact did “the great resignation” have on the recruitment market? Vijay Krishnan: In the wake of the pandemic, we’re seeing a transition in how software developers are switching jobs for better compensation and work-life balance. Job openings have hit a new record level and more and more companies are switching to an all-remote or hybrid work setup. As a result, companies are offering competitive salaries to attract and retain top talent. This offers tremendous opportunities for companies hiring for new positions but of course, creates challenges with retention. It also makes it clear that the recruiting industry has to be able to take developer preferences into account, not merely hiring company preferences. Remote work would be a great way to satisfy both preferences since as a result of remote work, candidates have a large pool of companies\/jobs to choose from and companies\/jobs too have a large pool of candidates to choose from, which in turn maximises the change of companies hiring ideal candidates and vice versa. AIM: What’s your game plan for the Indian market? What are the opportunities and challenges you see? Vijay Krishnan: India is a key focus market for us. We have seen Indian developers bringing significant value to the clients and the opportunity to work with top US companies, the flexibility to work from anywhere and get very attractive salaries is welcomed by the Indian developers. We have started India focused community events, marketing activities and developer upskilling workshops to enhance our penetration and focus here. AIM: How does Turing.com leverage AI? Vijay Krishnan: We are an AI-backed Intelligent Talent Cloud that helps customers source, vet, match, and manage the world’s best developers remotely. We are a 100% data-driven organisation that leverages, analyses, and exploits various data sources. We build models to serve our customers and developers worldwide. Our AI team includes various sub-teams that solve the analytics and ML problems for needs classified into supply, vetting, growth, demand, operations and matching. We use AI to solve challenging problems, including demand forecasting, pricing optimisation, automated vetting, and adaptive search\/ranking, to name a few. Turing gets around 25,000 new developer registrations in a week. These developers upload their CVs and share data about their experiences on our portal. They also take various tests on the portal, including skill-specific MCQs and coding challenges. Tracking the developers in each of the vetting stages, predicting who can be fast-tracked, and extracting maximum signals about the developer’s profile to match them in near future customer job requirements is key. This process involves multiple projects handling different steps and components during this journey using their own analytics techniques and ML models. Ensuring that we build a rich trove of this data continuously and collaboratively is an engineering and governance challenges. Turing’s solution to handle this challenge includes a. Building a robust data engineering pipeline and lake with a library of common analytical views and components. b. Leveraging a hybrid data store comprising cloud storage, databases, and feature stores. c. Continuously evolving shared feature store to have versioned developer features and updates into the same. d. Versioned ML Models and techniques having the ability to apply them to historical data quickly and measure the impact on key business metrics. e. Use of frameworks like mode and Jupiter hub to ensure easy and seamless collaboration. f.  A\/B testing and feature flag-driven product development to quickly check the hypothesis. g. Fast and easy logging infrastructure integrated into every component h. Heavy collaboration between product engineering and AI\/ML teams where the data needs of every feature getting built get vetoed by the latter. AIM: Tell us about your tech stack. Vijay Krishnan: AWS, GCP, BigQuery, Node, React, Python. We build our services with tech stacks that are scalable and reliable. We design with a distributed application framework and the microservices architecture. They allow us to easily scale any service component on the platform and continue to support high throughput and low latency service delivery. This empowers Turing’s engineering teams to continue building functionalities into our product with fast iteration and better quality control. We run both service stacks and AI workloads on multiple clouds including AWS and GCP to leverage the best cloud technologies to support our product offerings. Our tech stacks continue to evolve as we see the need for business growth. Today, we run node.js with react, however, we have taken one step further by looking into solutions like gRPC and GraphQL to prepare us for the next phase of the business expansion. AIM: What explains the growing conversation around Responsible AI? Vijay Krishnan: While AI today (esp. Deep AI) does its own feature selection and modelling, these steps work towards the goals set by humans using the data collected for human purposes. If the data is biased, it can amplify injustice even by accident. Discrimination against a sub-population can be created unintentionally, and that’s the fairness issue at the core of AI ethics. The responsibility of ensuring this does not happen lies on the teams before deploying any model to test for such bias. Before answering how one can handle this issue at scale, it is crucial to understand how the problems originate and manifest. Many tools and libraries make creating and applying the models very easy. The business pressures of quickly launching features can make the developers overlook the possibility of biases.Some models typically get linked and form a chain where an output of one is used by the other. In this case, the bias propagates and amplifies. The developers working towards the chain may not have a view of the bias of the input data.Biases happen at every layer from presentation to data, data to model, and model to user interaction via algorithm. So this is not just a data cleaning\/governance issue. Handling this issue at scale requires Creating awareness at the developer level about biasCatch possible biases in exploratory data analysisValidate models on various data sets and compare the behaviourTrying to use models which are explainable, auditable, and transparent. AIM: How do you ensure your data science and AI\/ML teams are aligned with the company’s AI governance policies and best practices? Vijay Krishnan: At Turing, each of our teams is led by very experienced senior AI\/ML professionals who primarily do the job of asking the right questions to the teams working with data. They ensure that the testing is adequate and they collaborate to discuss what the models have uncovered rather than focusing just on their outputs. We are constantly trying to understand what the data is saying to us, which features are correlated, why the behaviour does not match the hypothesis, etc. Detecting, spotting, and questioning any trend helps us identify the biases. We also give ample time to experiment before finalizing and deploying things into engineering. The A\/B testing helps us track unwanted outcomes even if it gets accidentally introduced despite all this care. There are just no deviations allowed from this iron fist process. Since our data is proprietary and not third-party data, our data privacy issues are simpler. We follow all the best practices for data protection measures for securing the database and cloud access.","excerpt":"We have started India focused community events, marketing activities and developer upskilling workshops to enhance our penetration and focus here.","categories":["AI Features"],"tags":["AI (Artificial Intelligence)","ai in hr","Ethical AI","Interviews and Discussions","Recruitment","Remote Work","Responsible AI"],"author_name":"Sri Krishna","publish_date":"2022-05-23T11:00:00","publication_year":"2022","word_count":1284,"keywords":["data science","GCP","TPU","AWS","AI","Ethical AI","ML","ai in hr","Responsible AI","RAG","microservices","Aim","analytics","Recruitment","Remote Work","AI (Artificial Intelligence)","Interviews and Discussions"],"extracted_tech_keywords":["AI","ML","data science","analytics","Aim","RAG","AWS","GCP","microservices","TPU"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/when-remote-working-is-a-business-model\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10162700,"title":"Budget 2025: Govt Pushes for GCCs in Tier-2 Cities, Unveils AI Centre of Excellence","content":"The Union Budget 2025 indicates the government’s shifting focus towards tier-2 cities for the expansion of global capability centres (GCCs), moving beyond traditional hubs like Bengaluru, Hyderabad, Delhi, Chennai, Mumbai, and Pune. Finance minister Nirmala Sitharaman has highlighted plans to boost talent, develop infrastructure, and introduce industry-friendly policies in tier-2 cities. To boost talent, the government has announced the establishment of a centre of excellence in AI for education and the awarding of 10,000 new fellowships to students at IITs and IISc over the next five years. Industry leaders had already anticipated this development. Alouk Kumar, founder and CEO of Inductus, said, “The success of GCCs in India to the tune of 70% depends upon the right kind of resources.” He pointed out that finding skilled talent is now a bigger challenge than setting up office spaces, leading some GCCs to struggle with hiring difficulties. Kumar proposed expanding beyond traditional locations to leverage the talent available in tier-2 cities. Raghavendra Vaidya, MD and CEO of Daimler Truck Innovation Center India said, “We hope the government can streamline the process of running a GCC from a tax and infrastructure standpoint. Investing in infrastructure would be greatly beneficial.” The Zinnov-Nasscom India GCC Landscape report shows a rise in tier-2 and tier-3 city GCCs, increasing from 5% in FY2019 to 7% in FY2024. These Nano GCCs are expected to grow by 15-20% by 2025, with 25-30% growth projected in the following years. According to AIM Research, cities like Chandigarh and Indore, each housing 4% of total GCCs, are becoming new hotspots for expansion. While Bengaluru and Hyderabad are still leading, many companies are exploring alternative locations to tap into skilled graduates from smaller cities. With tier-2 cities home to many engineering, IT, and business institutions, experts believe this shift will help GCCs find the right talent while spreading economic benefits across more regions. In a major push for education and skill development, the finance minister announced a CoE in Artificial Intelligence for Education with a total outlay of INR 500 crore. This initiative builds on the three AI centres set up in 2023 for agriculture, healthcare, and sustainable cities. Sitharaman also revealed plans for five National Centres of Excellence for Skilling, designed to equip the youth with industry-relevant expertise. “These centres will be set up with global partnerships to support Make for India and Make for the World manufacturing,” Sitharaman stated. The initiative will cover curriculum design, trainer training, a skill certification framework, and regular assessments. To strengthen higher education, the government is investing in the expansion of IIT infrastructure. Over the past decade, student intake at IITs has doubled from 65,000 to 1.35 lakh. Additional facilities will be developed at five IITs established post-2014, creating space for 6,500 more students. Hostel and other infrastructure capacity at IIT Patna will also be expanded","excerpt":"The finance minister has announced a CoE in Artificial Intelligence for Education with a total outlay of INR 500 crore.","categories":["GCC"],"tags":["GCC","union budget"],"author_name":"Shalini Mondal","publish_date":"2025-02-01T12:44:30","publication_year":"2025","word_count":470,"keywords":["Go","artificial intelligence","GCC","programming_languages:R","AI","innovation","ML","programming_languages:Go","RAG","Aim","union budget","R"],"extracted_tech_keywords":["AI","artificial intelligence","ML","Aim","RAG","R","Go","innovation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/gcc\/budget-2025-govt-pushes-for-gccs-in-tier-2-cities-unveils-ai-centre-of-excellence\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":34572,"title":"2019 Began With Tech Giants Google, Facebook &#038; Stanford Open-sourcing Datasets","content":"Facebook Open Sources LASER The month of January witnessed a myriad of research works open sourced by the likes of Google, Facebook and Stanford. These works set new benchmarks for NLP and image recognition techniques while improving on the pre-existing models. Recently Google introduced a paper on Natural Questions (NQ), a new dataset for QA research, along with methods for QA system evaluation. In contrast to tasks where it is relatively easy to gather naturally occurring examples, the definition of a suitable QA task, and the development of a methodology for annotation and evaluation is challenging. When an annotator is asked a question, it returns a longer version from the paragraphs of Wikipedia and also a short answer like a yes or no. The question seeks factual information; the Wikipedia page may or may not contain the information required to answer the question; the long answer is a bounding box on this page containing all information required to infer the answer; and the short answer is one or more entities that give a short answer to the question, or a boolean ‘yes’ or ‘no’. Both the long and short answer can be NULL if no viable candidates exist on the Wikipedia page. Facebook Open Sources LASER To accelerate the transfer of natural language processing (NLP) applications to many more languages, Facebook enhanced their LASER(Language-Agnostic SEntence Representations) toolkit. LASER is the first successful exploration of massively multilingual sentence representations to be shared publicly with the NLP community. The toolkit now works with more than 90 languages, written in 28 different alphabets. LASER achieves these results by embedding all languages jointly in a single shared space (rather than having a separate model for each). We The multilingual encoder and PyTorch code is freely available, along with a multilingual test set for more than 100 languages. This work is aimed at applications such as classifying movie reviews as positive or negative, in one language and then instantly deploy it in more than 100 other languages. Stanford’s CheXpert Probability prediction from chest radiographs via paper by Jeremy et al., CheXpert, a large dataset that contains 224,316 chest radiographs of 65,240 patients. A labeller to automatically detect the presence of 14 observations in radiology reports, capturing uncertainties inherent in radiograph interpretation. The researchers at Stanford investigated different approaches to using the uncertainty labels for training convolutional neural networks(CNN) that output the probability of these observations given the available frontal and lateral radiographs. On a validation set of 200 chest radiographic studies which were manually annotated by 3 board-certified radiologists, it is found that different uncertainty approaches are useful for different pathologies. The best model on a test set is composed of 500 chest radiographic studies annotated by a consensus of 5 board-certified radiologists, and compare the performance of our model to that of 3 additional radiologists in the detection of 5 selected pathologies. On Cardiomegaly, Edema, and Pleural Effusion, the model ROC and PR curves lie above all 3 radiologist operating points. This dataset can be used as a standard benchmark to evaluate the performance of chest radiograph interpretation models. Conclusion The year 2018 has seen a meteoric rise in the number of papers released in the field of AI. There were also numerous tools and techniques open sourced by the giants to carry the baton of AI research. Google’s BERT, for instance, introduced new benchmarks for natural language understanding. Along with LASER, Facebook also released BISON for providing systems with the ability to relate linguistic and visual content is one of the hallmarks of computer vision. Google and Facebook upped the ante right from the beginning and the rest of the year sure looks interesting for an AI enthusiast. Check the other releases here.","excerpt":"The month of January witnessed a myriad of research works open sourced by the likes of Google, Facebook and Stanford. These works set new benchmarks for NLP and image recognition techniques while improving on the pre-existing models. Recently Google introduced a paper on Natural Questions (NQ), a new dataset for QA research, along with methods for […]","categories":["Global Tech"],"tags":["Facebook","Google","machine vision","NLP","Stanford"],"author_name":"Ram Sagar","publish_date":"2019-02-06T12:13:55","publication_year":"2019","word_count":620,"keywords":["machine vision","TPU","AI","neural network","PyTorch","image recognition","computer vision","RAG","NLP","Aim","Google","Stanford","Facebook","R"],"extracted_tech_keywords":["AI","neural network","NLP","computer vision","Aim","PyTorch","RAG","image recognition","TPU","R"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/2019-began-tech-giants-opensourcing-datasets\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10049256,"title":"Facebook Introduces New Image Generation Model Called Instance-Conditioned GAN","content":"Facebook recently introduced a new image generation model called ‘Instance-Conditioned GAN (IC-GAN). This new model creates high-quality, diverse images with or without input images present in the training set. Compared to traditional methods, IC-GANs can generate realistic, unforeseen image combinations. The PyTorch code for Instance-Conditioned GAN is available on GitHub. GANs vs IC-GAN GAN, or generative adversarial network, is one of the popular AI methods to create images, be it abstract collages or photorealistic pictures. However, they, too, have limitations as they can only generate images of objects or scenes closely related to the training dataset. For example, a traditional GAN trained on images of cars shows impressive results when asked to generate other images of cars, but it is most likely to fail if asked to generate images of flowers or other objects outside of the car dataset. That is where Facebook’s Instance-Conditioned GAN comes into the picture, where it can generate images outside of the training dataset seamlessly. The new approach exhibits exceptional transfer capabilities across various types of objects. Facebook researchers said that the researchers can use IC-GANs off the shelf with previously unseen datasets and still generate realistic-looking images without the need for annotated\/labelled data. Also, IC-GANs could create new visual examples to augment datasets to include diverse objects and scenes, helping artists and creators with more expansive, creative AI-generated content, and advance research in high-quality image generation. Here’s how IC-GAN works Facebook’s IC-GAN can be used with both labelled and unlabelled datasets. It extends the GAN framework to model a mixture of local and overlapping data clusters. The way it works is, it takes a single image, or ‘instance,’ and then generates images that are similar to the instance’s closest neighbours in the dataset. The researchers said that they use neighbours as an input to the discriminator to force the generator to create samples similar to the neighbourhood samples of each instance. This way, it avoids the problem of partitioning data into small clusters as so much of the data is overlapping. As a result, it lets models use datasets more efficiently. After the model is trained, the researchers then test it on images it has never seen before. With the help of a single image, the model can effectively generate visually rich images similar to the closest neighbours in the dataset. In contrast, standard methods like class conditional GANs focus on conditioning class labels, effectively partitioning the data into groups corresponding to those labels to generate more high-quality samples than their unconditional counterparts. So, instead of creating only random images, these GANs create images that fit a particular label like ‘clothing’ or ‘car.’ At the same time, they rely on annotated\/labelled data that may be unfeasible to obtain. Previously, label-free learning approaches — using no labelled data — to image generation have been promising, but their result is usually of poor quality when trained to model complex datasets (ImageNet). Plus, they either use coarse, nonoverlapping data partitions or fine partitions that tend to deteriorate output as the clusters contain limited data points. The image below depicts an IC-GAN instance on the left to generate the corresponding images shown on the right. Here, no class label is provided. (Source: Facebook) Interestingly, the IC-GAN can be transferred to other datasets not seen during the training for both class-conditional scenarios and where there are no labels. Further, in IC-GAN, the researchers said they do this by swapping out the conditioning instances at inference time. In the case of class-conditional IC-GAN, they can swap either the instance conditioning or the class label. Combining instances and the class labels, the class-conditional IC-GAN can create unusual scenes that aren’t present or are very rare in current datasets. For example, with the image of a snowplough surrounded by snow and a class label ‘camel,’ which does not appear in the instance conditioning, the user can generate camels surrounded by snow, thereby eliminating the bias that camels live only in the desert. (see the image below) (Source: Facebook) Another example (as shown below) shows that the IC-GAN uses the graphics of a house to create a more realistic-looking building. Unlimited Possibilities (Source: Facebook) In a nutshell, IC-GAN can augment data and include objects commonly found in the training data. As a result, this new approach can help generate more diverse training data for object recognition models as it works across various domains. For instance, traditional GAN models would not generate images of zebras standing in urban areas, as its training data would likely only contain images of zebras in grasslands. Here, the IC-GAN can augment data and include objects that are not commonly found in standard datasets. “In the future, we hope to explore ways to bring even more control to this model,” said Facebook, stating that it will no longer be just about the object at the centre and the background. “We want to explore how more objects can be placed in the background and determine where the items are placed, creating complex, picture-perfect scenes,” said the researchers.","excerpt":"IC-GAN can be used with both labelled and unlabelled datasets.","categories":["Global Tech"],"tags":["AI Tool","Facebook AI research","FAIR","GAN","GANs","Generative Adversarial Networks","image-generation","Machine Learning","Machine Learning Latest","Machine Learning New"],"author_name":"Amit Naik","publish_date":"2021-09-23T15:00:00","publication_year":"2021","word_count":837,"keywords":["Generative Adversarial Networks","TPU","RPA","Machine Learning Latest","Git","GANs","AI Tool","R","Machine Learning New","PyTorch","Facebook AI research","AI","ML","FAIR","Machine Learning","GAN","image-generation","GitHub","AI-generated content"],"extracted_tech_keywords":["AI","ML","PyTorch","TPU","R","Git","GitHub","GAN","RPA","AI-generated content"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/facebook-introduces-new-image-generation-model-called-instance-conditioned-gan\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":41383,"title":"Apple Acquihires Drive.ai, Flexes Its Muscles In Autonomous Driving Space","content":"Img Src: Drive.ai It was the talk of the town that Drive.ai, the autonomous vehicle tech startup, once valued at $200 million, is shutting down after four years of operations. That was until another big news broke the internet – Apple is buying Drive.ai, saving it in the nick of time. It definitely comes as a major development for Apple, that has always been tight-lipped about its self-driving cars initiatives and has preferred to stay off the front page on developments around it, unlike its contemporaries like Google. What Went Wrong With Drive.ai? The company was all set to close down by the end of June, according to the WARN documents filed with the Employment Development Department of California, which must be submitted ahead of a mass layoff or plant closure. Drive.ai started on a high note in 2015 by former graduate students working in Stanford University’s Artificial Intelligence Lab run by Andrew Ng. It received a lot of attention and investments in its earlier years where it tweaked its business model with a plan to combine deep learning software with hardware to make kits that turned regular cars into autonomous cars. Reports suggest that the startup raised about $77 million and was last valued at $200 million in 2017. It boasted of investors such as New Enterprise Associates and Nvidia GPU Ventures. With these investments, the startup ramped up its operations extensively in 2017 and 2018 where users could hail a ride from different fixed pick-up and drop-off locations around Frisco and Arlington. It sparked headlines when it conducted fixed-route tests with its autonomous vehicles without human safety drivers on public roads. While there was a fleet of modified Nissan NV200s that showed messages such as “waiting,” “going,” “entering,” or “exiting”, it wasn’t enough to survive in an environment of consolidation and lowered expectations. Img Src: Drive.ai Though there were billions that were flowed into the startup, it failed to scale up as expected. Also as it expanded, the executive team appeared to be constantly in flux with several people holding the CEO spot. It was then that Drive.ai was rumoured to be snapped up by Apple in an acqui-hire deal. However, after the talks with Apple failed to materialize, the self-driving car startup was to shut down its operation, laying off 90 employees, until Apple confirmed recently that the deal is on. Apple Revives Its Ambitious Autonomous Vehicles Dream While Apple has never disclosed its plan to set a foot in self-driving space, it had been working behind the scenes quite aggressively. Buying Drive.ai comes as a push to its autonomous vehicles dream. Apple has had quite an on and off relationship with its autonomous vehicles efforts. Its tryst with self-driving cars dates back to 2014 when it began working on Project Titan with more than 1000 employees working on developing electric vehicle at a secret location near its Cupertino headquarters. It had worked on expanding its fleet of test cars and reportedly also teamed up with Volkswagen for vehicle testing. It had several Lexus SUVs outfitted with a range of different sensors and were equipped with new LIDAR equipment as in 2017. Apple CEO, Tim Cook had said that they are focusing deeply on autonomous systems. “It’s a core technology that we view as very important. We sort of see it as the mother of all AI projects. It’s probably one of the most difficult AI projects actually to work on.” However, earlier this year there was news of Apple laying off 200 employees from its ambitious Project Titan. The lay off was said to be a part of restructuring done under the leadership of Doug Field, an ex-Tesla engineer who was hired in 2018, but the move did not suggest how it will impact Apple’s autonomous vehicle plans. With the influx of Drive.ai employees and assets, however, Apple hints towards its plan to strengthen its autonomous vehicles department, and that it is here to stay in the competition. There are reports suggesting that Apple has also hired employees from its rivals Wymo and Tesla to step up its game. Our Point Of View In the wake of the latest development, Apple would get access to a handful of hardware and software engineers from Drive.ai apart from the assets that they were working on in the self-driving space. While in the past, Apple had solely focused on developing the software while continuing to partner with car manufacturers to get the hardware, this deal might put Apple in a stronger foot in terms of both software and hardware. While Apple has had a shaky journey till now, it may be now re-living its dream of developing a purpose-built self-driving car.","excerpt":"It was the talk of the town that Drive.ai, the autonomous vehicle tech startup, once valued at $200 million, is shutting down after four years of operations. That was until another big news broke the internet – Apple is buying Drive.ai, saving it in the nick of time. It definitely comes as a major development […]","categories":["AI Features"],"tags":["autonomous systems"],"author_name":"Srishti Deoras","publish_date":"2019-06-26T12:07:11","publication_year":"2019","word_count":782,"keywords":["Go","artificial intelligence","programming_languages:R","AI","AI employees","programming_languages:Go","autonomous systems","deep learning","GAN","R","startup"],"extracted_tech_keywords":["AI","artificial intelligence","deep learning","R","Go","GAN","startup","AI employees","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/apple-acquihires-drive-ai-flexes-its-muscles-in-autonomous-driving-space\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":52432,"title":"This Is How Ethereum Is Driving Innovation In Decentralized Finance (DeFi)","content":"In the last few years, a lot has changed in the realm of the public blockchain ecosystem, and there is a transition from currencies to more advanced applications and Decentralised Finance, also called DeFi. Particularly with the advent of Ethereum in 2015, it had a big impact. Now developers could write Turing-complete applications in Solidity programming language, something which was not possible with Bitcoin. But, why do we even need public blockchains in the first place, one may ask particularly when other large companies like IBM, Amazon, Oracle and many others have their own set of solutions? According to experts, the answer is decentralization and immutability\/transparency of data on a network available to anyone to use and contribute. Also, no single entity has control over the network, making it secure and free from centralized manipulation. This is the very reason that we have witnessed organisations like Microsoft, JP Morgan, Santander, Google and more recognize the value of Ethereum and even used it for multiple enterprise use cases. With so much happening on a single public blockchain like Ethereum, no wonder it has the largest community of developers. It is estimated that Ethereum has about more 4 times the number of developers compared to any other ecosystem of open-source public blockchain. Also, the demand for Solidity programming language is double compared to other programming languages used in blockchain applications as late as mid of 2019. Moreover, Ethereum may soon be moving to Vyper – a Pythonic language to write smart contracts, which will make it easier for developers to write and deploy decentralized applications for DeFi. (as compared to Solidity). Banks Beware: After Fintech You Have More Competition From Blockchains Like Ethereum In 2019, one of the major events that cemented the value of Ethereum as a public blockchain was when Santander settled a $20 million bond with a client, without needing to rely on a single traditional bank for the transaction. Consequently, we also saw this year that many financial institutions showed interest in using Ethereum to issue securities. If you are surprised to know that a $20 million bond was issued on a public blockchain controlled by no bank, it may come as shocking to you that another $110 million (₹770 crore) bond was issued by a French financial company known as Societe General to clients as its first issuance on Ethereum. This was a major milestone for DeFi industry. While 2018 was the year of crazy crowdfunding via Initial Coin Offerings (ICOs) on Ethereum, 2019 saw more meaningful financial deployments using public blockchains. Security Issuance via public blockchains gained incredible prominence in 2019. For instance, Latin America’s largest investment bank, Brazil’s Banco BTG Pactual announced plans to launch Security Token Offerings (STOs) and DeFi on the Tezos blockchain in July 2019 worth over $1 billion or ₹7,000 crore (yes you can verify here). Here, the security issuance would need developers working on the Tezos platform to write smart contracts in a completely novel programming language developed by Tezos Foundation- called Michelson. Gaming & Non Fungible Assets Are Also Taking Up For Ethereum Let’s move onto yet another meaningful (and fun part) when it comes to public blockchains, which are gaming and non-fungible assets. One big thing that happened in 2019 in this area was global tech giant Microsoft deploying Non-Fungible Assets (NFAs) to reward developers on its cloud platform for a program known as Azure Heroes. Here again, the noteworthy thing is that NFAs were deployed on the Ethereum public blockchain. The tokens rewarded to developers for their contributions could be exchanged with other assets like Ethereum or Bitcoin for better incentives. Microsoft in May 2019 had already released its app development kit for Ethereum applications on Azure cloud. The reward program has been supported by Intel as well, which along with Microsoft are part of Enterprise Ethereum Alliance– a consortium of global companies working to advance the use cases of Ethereum in the enterprise space. Non-fungible tokens represent individual ownership for each unique non-fungible asset- be it land, property, game cards\/avatars, or in fact shoes based on Ethereum ERC 721 standard. In December 2019, shoemaker Nike just acquired a patent to tokenise footwear on Ethereum. In a document published on the U.S. Patent and Trademark Office and on December 10 2019, Nike revealed that it wants to have unique IDs and create ERC 721 tokens for some shoes. People can unlock these tokens known as “CryptoKick” by buying physical shoes and these tokens can then be connected with unique owner IDs to signify ownership. To give another use case of a public blockchain for gaming and NFAs, Microsoft, along with game developer Eidos and Fabled Lands announced in December 2019 that they are jointly developing card games based on a 1980s best-selling gamebook using VeChain– another popular public blockchain from Asia. Here again, game cards would be converted into non-fungible assets for trading and exchanging, another use case for DeFi. Also in the gaming context, in late 2019 AMD- the chip maker joined the Blockchain Game Alliance, an industry group devoted to promoting and standardising the technologies for online gaming. As the first major hardware provider to join the alliance, AMD’s goal is to supply the CPU and GPU chips needed to help gaming-related developers efficiently run games to be developed on Ethereum. Overview While we did see the adoption of public blockchain networks like Ethereum and more on 2019, the majority of use cases were tied to things like DeFi, gaming, and cryptocurrency trade- leveraging their trustless nature. Of course, decentralisation and absolute data immutability come at a cost, meaning once a transaction is validated across the network, it is impossible to roll it back. The use case for public blockchains in the enterprise has therefore been limited to applications which either require openness such as gaming or asset tokenization, or use cases where information is required to be exchanged without needing to trust any party.","excerpt":"In the last few years, a lot has changed in the realm of the public blockchain ecosystem, and there is a transition from currencies to more advanced applications and Decentralised Finance, also called DeFi. Particularly with the advent of Ethereum in 2015, it had a big impact. Now developers could write Turing-complete applications in Solidity […]","categories":["Deep Tech"],"tags":["Blockchain","blockchain solution in finance","Blockchain Technology","cryptocurrency ethereum","Ethereum"],"author_name":"Vishal Chawla","publish_date":"2019-12-23T14:00:00","publication_year":"2019","word_count":990,"keywords":["Go","funding","Blockchain","cloud_platforms:Azure","AI","Azure","R","RAG","Python","blockchain solution in finance","Rust","GAN","Blockchain Technology","cryptocurrency ethereum","Ethereum"],"extracted_tech_keywords":["AI","RAG","Azure","Python","R","Go","Rust","GAN","funding","cloud_platforms:Azure"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/public-blockchain-ethereum-defi\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10020521,"title":"Google Appoints A New AI Lead, After Timnit Gebru’s Layoff","content":"Google has announced the appointment of Dr Marian Croak to lead the AI division within Google Research. Prior to this, Croak was heading the engineering division of the company as a Vice President. With this new appointment, the tech giant is hopeful of stabilising the chaos after Timnit Gebru’s layoff. A recent blog post by Google, confirming the news, stated that Croak will now be responsible for managing the teams working on artificial intelligence for social good, algorithmic fairness, and ethics in AI. She will be directly reporting to the SVP of Google AI Research and Health. Speaking about her new responsibility, Croak stated in the official blog post that considering the field of ethical AI is relatively novel, and there is a lot of conflict in the views within the area. Thus, to advance in this field, one needs to address these issues. “Most institutions have only developed principles, and they’re very high-level, abstract principles, in the last five years. There’s a lot of dissension, a lot of conflict in terms of trying to standardise normative definitions of these principles,” said Croak. “There’s quite a lot of conflict right now within the field, and it can be polarising at times. And what I’d like to do is have people have the conversation more diplomatically, perhaps, than we’re having it now, so we can truly advance this field.” This changing leadership can be attributed to the departure of Dr Timnit Gebru and suspension of researcher Margaret Mitchell from the company, which has questioned the tech giant’s attempt to build ethical AI and created massive chaos amid the industry. However, with this new leadership, the company is hoping to develop responsible artificial intelligence and that it has a positive impact. To work on this vision, Croak has also created and will lead a new centre of expertise on ethical AI within Google Research, stated in the official blog post. However, it is still unclear what this new appointment means for Margaret Mitchell, who is still in a lockout from the company and being investigated.","excerpt":"Google has announced the appointment of Dr Marian Croak to lead the AI division within Google Research. Prior to this, Croak was heading the engineering division of the company as a Vice President. With this new appointment, the tech giant is hopeful of stabilising the chaos after Timnit Gebru’s layoff. A recent blog post by […]","categories":["AI News"],"tags":["Ethical AI","Responsible AI","timnit gebru"],"author_name":"Sejuti Das","publish_date":"2021-02-19T10:58:29","publication_year":"2021","word_count":343,"keywords":["Go","artificial intelligence","programming_languages:R","AI","Ethical AI","programming_languages:Go","Responsible AI","timnit gebru","AI research","R"],"extracted_tech_keywords":["AI","artificial intelligence","R","Go","AI research","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/google-has-a-new-ai-lead-after-timnit-gebrus-layoff\/","complexity_score":3,"technical_depth":7,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10046185,"title":"IIM Ahmedabad To Establish A Centre For Data Science And Artificial Intelligence","content":"The Indian Institute of Management, Ahmedabad, has announced the launch of the Brij Disa Centre for Data Science and Artificial Intelligence (CDSA). This specialised centre will undertake cutting-edge research in data science and AI to support businesses, governance, and policymaking. It also aims at building collaborative relationships between scholars and practitioners along with conducting case-based research to understand the current industry practice. The centre will also be developing case studies for classroom teaching. The endowment for CDSA has been contributed by Dipak Gupta, joint managing director, Kotak Mahindra Group, an alumnus of IIMA. CDSA will be responsible for the dissemination of knowledge to a wider audience, both within and outside the institute, via seminars, workshops, and conferences. The centre will also publish an exhaustive annual report on the data science and AI industry in the country, highlighting challenges and gaps, gauging the scope of the industry and offering possible solutions. The advisory committee of the centre includes Sanjiv Das, William and Janice Terry, Professors of Finance at Santa Clara University’s Leavey School of Business; Sunil Gupta, Edward W. Carter Professor of Business Administration, and co-chair of the executive program on Driving Digital Strategy at Harvard Business School, etc. “The Centre for Data Science and Artificial Intelligence at IIMA will take on an important role wherein it will create knowledge that will further augment growth in business and societal advancement, at large, through its offerings,” said Prof Errol D’Souza, director, IIMA. Speaking on the goals of the Centre, Prof Ankur Sinha, co-chairperson, CDSA, said that the intent is to develop it as the epicentre of knowledge of data science and AI. “We also aim to engage with the industry and establish synergies that can provide sound technical exposure to the students and offer plausible solutions that can be utilised by the industry and academia,” he added.","excerpt":"The Indian Institute of Management, Ahmedabad, has announced the launch of the Brij Disa Centre for Data Science and Artificial Intelligence (CDSA). This specialised centre will undertake cutting-edge research in data science and AI to support businesses, governance, and policymaking. It also aims at building collaborative relationships between scholars and practitioners along with conducting case-based […]","categories":["AI News"],"tags":["what is data science"],"author_name":"Shraddha Goled","publish_date":"2021-08-17T14:46:31","publication_year":"2021","word_count":305,"keywords":["data science","Go","artificial intelligence","programming_languages:R","AI","programming_languages:Go","Git","Aim","what is data science","R"],"extracted_tech_keywords":["AI","artificial intelligence","data science","Aim","R","Go","Git","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/iim-ahmedabad-to-establish-a-centre-for-data-science-and-artificial-intelligence\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10052448,"title":"8 Indian Startups Advancing Healthcare With AI","content":"Healthcare is one of the most vital fields of AI applications. Some of the greatest healthcare innovations today have a significant AI component. For instance, AI helped quicken the process of bouncing back from the global effect of the COVID-19 pandemic that lasted close to two years. AI was leveraged for multiple purposes in the healthcare industry– from early-COVID-19 detection to providing assistance to patients and drug repurposing. According to Research and Markets, the application of AI in healthcare in India will be valued at Rs 431.97 billion by 2021, growing at a rate of about 40 percent. This is all the more significant for a country like India since the doctor-patient ratio in the country is 1:1456, way lower than the World Health Organisation recommended 1:1000 ratio. The report further suggests that with the increasing use of AI in healthcare, India is expected to reach 6.9:1000 in 2023, as compared to 4.8:1000 in 2017. Today, we have curated a list of companies bringing AI into healthcare. Niramai Bengaluru-based Niramai Health Analytix was founded in 2016 by Geetha Manjunath and Nidhi Mathur. The health tech startup has developed Thermalytix that uses AI and a high-resolution thermal sensing device for the detection of breast cancer at an earlier stage than the usual traditional methods or self-examination. The software-based medical device is low-cost, automated and portable, making cancer screening possible in all clinics. The core technology is based on ML algorithms. Qure.ai Founded by Prashant Warier and Pooja Rao in 2016, Mumbai-based Qure.ai uses AI to make healthcare solutions affordable and accessible. It uses deep learning algorithms to interpret and radiology images and scans– chest X-rays, head CT scans, POQUS, chest CT scans, etc — within a couple of seconds. It also has community solutions for COVID-19, public health and tuberculosis. HealthifyMe Founded in 2012, HealthifyMe is a Bengaluru-based digital health and wellness platform. Founders Tushar Vashisht, Sachin Shenoy and Mathew Cherian founded the company with the aim of introducing digital healthcare to Indians. It uses an AI-based virtual assistant, Ria, that keeps in touch with its users and solves their queries around fitness, nutrition, and health in 10 different languages. Additionally, the app provides dietary recommendations, uses AI to track calorie intake and provides suitable healthy recipes and tips. PharmEasy Mumbai-based PharmEasy was founded in 2015 by Dharmil Sheth, Mikhil Innani and Dhaval Shah to develop an application connecting users with pharmacies. The smartphone-based app connects users with these pharmacies to make seamless medical deliveries. PharmEasy uses ML and Big Data tools, including Tensorflow, Hadoop, HBase, Kafka, Spark and Hive. Additionally, it uses a lot of data crunching and analytics solutions for the app. It has partnered with more than 80,000 pharmacies across 1200-plus cities in the country, serving more than five million customers. SigTuple Technologies Bengaluru-based SigTuple Technologies was founded in 2015 by Tathagoto Rai Dastidar, Apurv Anand and Rohit Pandey. SigTuple is on a mission to automate the manual review of medical data– blood sample, urine and semen microscopy, retinal image analysis — using AI. It operates in the fields of pathology and ophthalmology, offering an automated digital microscope and AI platform as solutions. Earlier last year, two co-founders left the team and Tathagoto took over as the CEO. OncoStem Diagnostics 2011-founded OncoStem Diagnostics integrates ML algorithms to assess the aggressiveness of tumours and identify the characteristics of recurrence risk over five years. Its flagship product CanAssist uses data support vector machine-based statistical models to assign risk scores based on five biomarkers and clinicopathological information for each patient. The team works in techniques including Support Vector Machines with linear and Radial Basis Function kernels, Random Forest, Elastic Net, and normal mixture modeling. Artelus Bengaluru-based Artelus was founded by Pradeep Walia, Rajarajeshwari K and Vish Durga in 2015. The health tech startup applies deep learning technology for diabetic retinopathy screening. AI-powered algorithms review the images for diagnosis. Artelus is operational across 502 locations (in India, US and Dubai) and has so far helped in the screening of more than 81,000 patients. The health tech startup has recently developed an AI-powered contactless diabetic retinopathy screening system that can conduct an eye examination in less than three minutes. Tricog Bengaluru-based healthcare analytics company Tricog was founded in 2014 by Chirat Bhograj, Zainul Charbiwala, Udayan Dasgupta and Abhinav Gujjar. Its products include: InstaECG: the flagship product is a cloud-connected device that helps interpret and analyse ECG reports within just 10 minutes. InstaEcho: Another cardiac product that assists doctors in the quick and accurate echocardiogram diagnosis within a few hours of test. So far, Tricog has served in more than 12 countries, empowering 2,600-plus health workers, impacting more than three million patients and saving 90,000-plus lives. It was also awarded the NASSCOM Artificial Intelligence Game Changer Award in 2018.","excerpt":"Here’s a curated list of companies that are bringing artificial intelligence into the healthcare industry.","categories":["AI Startups"],"tags":["AI (Artificial Intelligence)","AI in healthcare india","healthtech"],"author_name":"Debolina Biswas","publish_date":"2021-10-28T11:00:00","publication_year":"2021","word_count":795,"keywords":["healthtech","artificial intelligence","AI","ML","RAG","Ray","AI in healthcare india","deep learning","Aim","analytics","Kafka","TensorFlow","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","ML","deep learning","analytics","Aim","Ray","TensorFlow","RAG","Kafka"],"url":"https:\/\/analyticsindiamag.com\/ai-startups\/8-indian-startups-advancing-healthcare-with-ai\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10090358,"title":"First of its kind AI Forum: Enabling the Next Wave of AI Professionals","content":"Analytics India Magazine is excited to announce the launch of AI Forum –  a community, created in association with NVIDIA, aimed at fostering collaboration and growth within the artificial intelligence (AI) industry in India. The first-of-its-kind forum has been designed to be a hub of innovation and collaboration for developers passionate about data science and harnessing the power of technology. For more information visit https:\/\/aiforum.analyticsindiamag.com\/ With this, Analytics India Magazine envisions the creation of a dynamic and inclusive platform that brings together individuals from diverse backgrounds and expertise involved in AI, fostering cross-disciplinary collaboration and knowledge-sharing adding to the AI value chain. The forum will provide a platform that offers opportunities for support and collaboration for members enabling them to develop and improve their skills using their technologies. In addition, the AI forum will foster the exchange of ideas, sharing of best practices, and building of new partnerships to stay abreast with the latest developments in the rapidly evolving field of AI. The Forum will be the go-to source for AI professionals who are seeking relevant content, conversations, and community growth. The community aims to provide a comprehensive platform that can address the needs of professionals across various stages of their careers. This would include access to a wide range of content, such as articles, podcasts, and webinars, as well as opportunities to engage in conversations with other professionals in the field. The community aims to enable a network of professionals to collaborate and share knowledge, while the vision is to become a central hub for AI professionals to access relevant content, engage in conversations, and foster community growth. A thriving space for creativity and innovation The forum will also offer a thriving space for collaboration and enabling each other to tackle the most complex AI challenges, expand the horizons of this dynamic and swiftly evolving field, and offer ground-breaking solutions. Community Rules The forum rules are designed to create a positive and productive environment where members can thrive. All members are expected to treat each other with respect, stay on topic, avoid spam, protect confidentiality, and provide support and constructive feedback to each other. Moderation Rules The forum has established a set of rules to ensure a safe and inclusive community and the moderators are responsible for enforcing these rules. The members who violate the rules will receive a warning at the start while those found to be repeatedly violating these rules may be temporarily banned from the forum.  Repeated violations and harmful behaviour may also result in a permanent ban. The members, who have been banned, can appeal against the decision by contacting the moderators. Additionally, members can also report the flouting of rules to the moderators on Discord using the report function. The moderators have the discretion to enforce the rules as they deem fit and take action to ensure a positive community. Upon joining the AI Forum, members are expected to comply with the rules and decisions of the moderators. The rules may be adjusted, revised and amended to ensure a positive and productive community. We invite you to join us in our mission to broaden the frontiers of artificial intelligence. Together, we can create an environment of innovation and collaboration that inspires creativity and drives progress in this exciting and rapidly-evolving field. JOIN US HERE!","excerpt":"The idea is to create a dynamic and inclusive forum that brings together individuals from diverse backgrounds and expertise levels in AI , fostering cross-disciplinary collaboration and knowledge-sharing.","categories":["AI Features"],"tags":["AI community"],"author_name":"Aparna Iyer","publish_date":"2023-03-30T10:30:00","publication_year":"2023","word_count":550,"keywords":["data science","Go","API","artificial intelligence","AI community","AI","innovation","Aim","ViT","analytics","R"],"extracted_tech_keywords":["AI","artificial intelligence","data science","analytics","Aim","R","Go","API","ViT","innovation"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/first-of-its-kind-ai-forum-enabling-the-next-wave-of-ai-professionals\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10039762,"title":"Why Open-Source Is Here To Stay?","content":"In August 1991, Linus Torvald, a 21-year-old student, decided to start a free operating system. What began as a hobby grew into the world’s first free operating system, Linux Kernel. Today, open-source software is a significant part of software development allowing developers to inspect, copy, modify, and redistribute software. So, what is the status of open-source markets today? While proprietary software providers dominate the market, open-source software has picked up momentum lately—especially in the wake of the COVID-19 pandemic. Despite the global economic slowdown, the open-source market gathered steam in 2020 to meet remote organisational needs. As per GitHub, open-source project creation per any active user between late April 2019 and late April 2020 saw over 40 percent YOY growth. The rise of open-source software in 2020, amidst a global pandemic, can be attributed to the sudden spike in demand for the faster development of software programmes and applications. In March 2020, GitHub noticed a major uptick in open-source projects. Many streaming websites even cut back on bandwidth consumption by reducing video streaming quality and download speeds. This is where open-source software comes in. With proprietary software being more costly in terms of speed of deployment, more enterprises turned to open-source software. Integrating open-source code accelerates software development. It makes information more democratised and thus allows a technically diverse group to develop applications rapidly. Many developers usually back open-source software, thus finding solutions to software problems and creating new applications much more straightforward. Also, reusing open-source code allows developers to build applications quicker than starting from scratch. Since open-source code can be thoroughly accessed and customised, these developers can make changes to elements of the code to suit their purpose. This is another advantage open-source code has over proprietary software code. Finally, businesses can also employ open-source tools to manage existing data and data production—which is vital to enterprises today—with higher levels of precision. Changing gears The open-source software explosion begs the question: Did the pandemic-driven shift to open-source markets carry over to 2021? As per a report by Red Hat Software, IT leaders view enterprise open-source software as superior to other forms of software. However, historically, application development was seen as the most critical use for open-source software, with the lower cost of ownership being seen as its primary benefit. In 2021, however, infrastructure modernisation and digital transformation have been the most critical use cases for enterprise open source. Major benefits included higher quality (especially in terms of access to innovation) and security. Open-source software has contributed a lot in areas such as artificial intelligence and machine learning. Due to rapid innovation in the two fields, a massive amount of resources have been committed to accelerating suitable technologies. TensorFlow, released in 2015 by Google, is an open-source artificial intelligence software and is now among the most used frameworks to develop machine learning models. The framework is available in Java, JavaScript, Python, C++, Haskell, Go and Rust, and allows users to build neural networks using flowgraphs. Open-source software is also integral to architectures like cloud operating systems and edge computing. Linux has become the cloud operating system of choice due to its greater flexibility and lower cost. In 2021, the Linux operating systems market is expected to grow at 7 percent a year, reaching $9.7 billion by 2024. Open-source software providers find opportunities in the platform software and infrastructure software segments in edge computing, which are expected to grow from $10 billion in 2020 to $24.9 billion in 2024. The open-source model has been incorporated into many large organisations. NASA has various open-source projects on GitHub as a part of its ‘Open Source Development’ model. Wrapping up Of course, every rising technology stokes new fears. Unlike proprietary software, open-source software is not owned by anybody in particular. Thus, the biggest fear among these is that no entity will be ready to bear liability for adverse consequences. Another issue here arises from a legal standpoint. In 2010, Oracle sued Google for copyright infringement over copying its Java API code. Recently, Elastic changed licensing terms of its open-source code after a spat with AWS. Nonetheless, the benefits of open-source software seem to outweigh the costs. With the fast-changing nature of technology, the broader adoption of open-source software and the rise of open-source markets remains crucial.","excerpt":"In August 1991, Linus Torvald, a 21-year-old student, decided to start a free operating system. What began as a hobby grew into the world’s first free operating system, Linux Kernel. Today, open-source software is a significant part of software development allowing developers to inspect, copy, modify, and redistribute software. So, what is the status of […]","categories":["IT Services"],"tags":[],"author_name":"Mita Chaturvedi","publish_date":"2021-05-07T14:00:00","publication_year":"2021","word_count":713,"keywords":["machine learning","artificial intelligence","AWS","AI","neural network","Python","edge computing","TensorFlow","R","Redis"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","neural network","TensorFlow","AWS","edge computing","Redis","Python","R"],"url":"https:\/\/analyticsindiamag.com\/it-services\/why-open-source-is-here-to-stay\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10110786,"title":"Pixxel Launches Spacecraft Manufacturing Facility in Bengaluru","content":"Pixxel, a leader in hyperspectral earth-imaging technology, inaugurated its Spacecraft Manufacturing Facility in Bengaluru, India, today. The event was attended by S Somanath, Chairman of ISRO, and other industry stakeholders and investors. Pixxel introduced its satellite building facility at the ceremony. Pixxel’s CEO and Founder, Awais Ahmed, outlined the facility’s capabilities, stating, “We’ll build 6 of our Firefly hyperspectral imaging satellites, here.” Noting that at full capacity, the facility can handle over twenty satellites simultaneously, with a turnaround time of six months. Ahmed, also highlighted the achievement, stating, “For the first time in India and globally, we’ll have a five-meter commercial hyperspectral satellite.” Over the past two years, Pixxel has deployed three functional satellites, showcasing its expertise in converting data into actionable insights. The company is expanding its operations in the upstream and downstream market in space, revealing partnerships with global entities such as British Petroleum, the Ministry of Agriculture in India, and the National Reconnaissance Office in the US for its data. Pixxel plans to launch 6 satellites as part of the Firefly constellation this year, bringing the total to 9 with the inclusion of the 3 already deployed satellites. The company also expressed intentions to launch satellites for the bigger and better Honeybee constellation of hyperspectral satellites. The newly inaugurated facility, spanning over 30,000 sq ft, consolidates satellite manufacturing services, providing a comprehensive Spacecraft Assembly, Integration, and Testing (AIT) facility. Pixxel aims to streamline the production process, designing, manufacturing, integrating, and testing satellites under one roof before shipping them to launch sites. CEO Awais Ahmed shared his vision, stating, “The inauguration of the new facility marks a momentous milestone as Pixxel nears its fifth anniversary since inception.” He emphasised the facility’s role in Pixxel’s mission to build a health monitor for the planet, delivering critical data to key industries like agriculture, energy, forestry, and environmental monitoring. Two modern clean rooms ensure a contamination-free environment during satellite assembly, and the facility includes labs for advanced camera integration, electronics R&D, electrical assembly, a mechanical workshop, a mission control room, and office space for over 200 employees. Pixxel’s commitment to sustainability is evident in the facility’s features, including a wastewater treatment plant and smart HVAC systems to enhance energy efficiency. The opening of this facility is a significant moment for Pixxel as it prepares to launch six satellites in 2024 and eighteen more by 2025, advancing its mission to build a health monitor for the planet.","excerpt":"Pixxel’s 30,000 sq ft facility is hub for manufacturing, Integration and testing of its 6 satellites for Firefly constellation and upcoming missions.","categories":["Deep Tech"],"tags":["Pixxel","Spacecraft"],"author_name":"Shyam Nandan Upadhyay","publish_date":"2024-01-15T15:08:56","publication_year":"2024","word_count":405,"keywords":["Pixxel","Spacecraft","programming_languages:R","AI","ML","Aim","R"],"extracted_tech_keywords":["AI","ML","Aim","R","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/pixxel-launches-spacecraft-manufacturing-facility-in-bengaluru\/","complexity_score":2,"technical_depth":5,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10043622,"title":"Is Data Science a Good Career for You?","content":"Data science might be ‘the sexiest job of the 21st century’ with fat salaries, but that does not mean it is the right career choice for you.  As per AIM Research, 1,400 data science professionals working in India are paid more than INR 1 crore. According to AI & Data Science Salary Study 2021 by AIM Research, the salary of a data analytics professional is 44 percent higher than that of a software engineer, and 36 percent and 50 percent higher than the salaries of an IT developer and a Java developer respectively. Data science is about defining and solving business problems. However, many experts claim there is nothing wrong with people choosing this career for a better life, and they can always learn on the job and grow. Unfortunately, many candidates applying for data science roles are drawn by the glamour quotient and\/or monetary benefits, and in most cases don’t have the right skillset to excel in the field. “Everyone cannot be a data scientist or an analyst. So, we have to look at the right fit and where it starts. We can not start from the point where the candidates join the organisation. It should start within the campus or colleges,” said Deepak Kumar Arora, head of learning and development at Birlasoft, at a roundtable discussion on ‘analytics as a driver of talent transformation,’ organised by AIM and Jigsaw Academy. Do you have these skills? Mathematics and coding (R, Python, C++, Java and others) are essential skills for a rewarding data science career. However, it is not necessary to know all the programming languages. Candidates should also have excellent communication skills and gel with their teams. They should know social media mining, SQL\/NoSQL, natural language processing or machine learning algorithms, Microsoft Excel; and most importantly, need to be data and business savvy. Essential Skill Required To Become A Data Science Professional In 2021 (Source: UC Berkeley Extension) PhonePe’s data science head Kedar Swadi told AIM that candidates often focus on the currently fashionable techniques (like assuming they only need to understand deep learning) rather than having a solid foundation in algorithms, maths, and statistics. “A lot of candidates also list techniques they have only used, rather than listing those that they have a comprehensive understanding of (in terms of how they work, what assumptions they make about data, how they should be evaluated, what impact turning parameters has, etc.),” he added. He said candidates fail to understand or appreciate the business aspects of problems. Also, candidates lack exposure to the engineering aspects of developing solutions. Expectations vs reality Most people go into data science for the adventure it offers. However, the reality is slightly different. “In most organisations, you’ll have to spread your time between doing technical work and the other, less exciting stuff, ” said Adam Sroka, head of machine learning engineering at Origami. So, if you are not keen on reporting, writing, documenting and delivering presentations, or repeatedly explaining the basics of your models or techniques, project management, administrative overhead, etc to the stakeholders, then the job might not be a right fit for you. Candidates coming from an education or research background often fall into the trap of infinite timescale and infinite budget mindset. “All too often, I have heard protests from data scientists saying they can not put a ‘timeline’ on when their work will be finished, and it will take as long as it takes. This simply is not true and won’t fit well with the culture at most organisations,” said Sroka. ‘You either fix the scope of what you are trying to achieve and vary the timescale, or fix the timescale and vary the scope,” said Sroka. Right communication Communication is pivotal to forge a successful career in data science. For instance, if you are working closely with the company’s decision-makers, maintaining a solid relationship is essential. Most importantly, you will have to maintain a good relationship with all the team members across departments, not just senior management. Always look for an opportunity to solve the business problem or in-house team concerns to the best of your ability: be it automating redundant tasks or basic data retrieval. Meeting high expectations Most data science professionals in a company, by default, will be considered analytics and data experts. However, Jonny Brooks-Bartlett, a data scientist at Deliveroo, said trying to tell everyone what you know and have control of can be difficult. “Not because anyone will think any less of you, but because as a junior data scientist with little industry experience, you will worry that people will think less of you,” he added.","excerpt":"Data science might be ‘the sexiest job of the 21st century’ with fat salaries, but that does not mean it is the right career choice for you.  As per AIM Research, 1,400 data science professionals working in India are paid more than INR 1 crore. According to AI & Data Science Salary Study 2021 by […]","categories":["AI Highlights"],"tags":["data analyst vs data scientist","Data Science Jobs","data science latest","Data science skills","data scientist salary in india"],"author_name":"Amit Naik","publish_date":"2021-07-15T16:00:00","publication_year":"2021","word_count":769,"keywords":["data science","Go","data analyst vs data scientist","machine learning","AI","data science latest","R","Data Science Jobs","Python","Aim","data scientist salary in india","deep learning","analytics","SQL","Data science skills"],"extracted_tech_keywords":["AI","machine learning","deep learning","data science","analytics","Aim","Python","R","SQL","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/why-data-science-might-not-be-the-right-career-for-you\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10161617,"title":"Snowflake AI’s SwiftKV Cuts Meta Llama Inference Costs by Up to 75%","content":"Snowflake AI Research has introduced SwiftKV, an optimisation framework integrated into vLLM that significantly reduces inference costs for Meta Llama large language models (LLMs). The SwiftKV-optimised models, Snowflake-Llama-3.3-70B and Snowflake-Llama-3.1-405B, are available for serverless inference on Cortex AI. They offer cost reductions of up to 75% compared to the baseline Meta Llama models without SwiftKV. “SwiftKV’s introduction comes at a critical moment for enterprises embracing LLM technologies. With the growth of use cases, organisations need solutions that deliver both immediate performance gains and long-term scalability,” the company said. The framework reduces computational overhead during the key-value (KV) cache generation stage by reusing hidden states from earlier transformer layers. According to Snowflake AI Research, this optimisation cuts prefill compute by up to 50% while maintaining enterprise-grade accuracy. “Our approach combines model rewiring with lightweight fine-tuning and self-distillation to preserve performance,” the team explained. Accuracy loss is limited to about one point across benchmarks. SwiftKV delivers performance improvements, including up to twice the throughput for models like Llama-3.3-70B in GPU environments such as NVIDIA H100s. It also reduces the time to the first token by up to 50%, benefiting latency-sensitive applications such as chatbots and AI copilots. “It is designed to integrate seamlessly with vLLM, enabling additional optimisation techniques such as attention optimisation and speculative decoding,” the Snowflake team said. Beyond its integration with Cortex AI, SwiftKV is open-source, with model checkpoints available on Hugging Face and optimised inference on vLLM. The team has also released the ArcticTraining Framework, a post-training library for building SwiftKV models, enabling enterprises and researchers to deploy custom solutions. “By tackling computational bottlenecks, SwiftKV allows enterprises to maximise the potential of their LLM deployments,” Snowflake AI Research said. Snowflake recently entered a multi-year deal with AI safety and research company Anthropic to use its Claude models. This partnership will make Anthropic’s Claude models available to customers through Snowflake Cortex AI and help businesses worldwide get more value from their data. More businesses are turning to Snowflake’s cloud data to organise their data using AI. Like Salesforce and Microsoft, Snowflake is developing AI agents with its Snowflake Intelligence platform. Snowflake chief Sridhar Ramaswamy believes it will simplify how enterprises derive value from data. “Imagine asking a data agent, ‘Give me a summary of this Google Doc’ or ‘Tell me how many deals we had in North America last quarter’, and instantly following up with the next steps using that same agent. That’s exactly what Snowflake Intelligence will enable – a seamless way to access and act on your data in one place,” he added.","excerpt":"It reduces the time to the first token by up to 50%, benefiting latency-sensitive applications such as chatbots and AI copilots.","categories":["AI News"],"tags":["Snowflake"],"author_name":"Siddharth Jindal","publish_date":"2025-01-17T13:52:27","publication_year":"2025","word_count":426,"keywords":["Anthropic","Hugging Face","Go","AI","chatbots","ML","serverless","copilots","R","Snowflake"],"extracted_tech_keywords":["AI","ML","Anthropic","Hugging Face","copilots","chatbots","serverless","Snowflake","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/snowflake-ais-swiftkv-cuts-meta-llama-inference-costs-by-up-to-75\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10005190,"title":"This AI Startup Aims To Be The Google Maps For The Emerging Market","content":"Emerging markets can be more complicated than developed economies in terms of cultures, use of different languages, densely populated cities, poor infrastructure, hyperlocal nuances and more. Ajay Bulusu, Gaurav Bubna and Shaolin Zheng, former employees of Southeast Asian ride-hailing giant, Grab understood these challenges and are now trying to solve these complexities using AI. With Nextbillion.ai, founded in 2020, they are on a mission to be the world’s most hyperlocal AI company. “We want people to think of Nextbillion.ai when they think of AI solutions in a hyperlocal setting,” says Ajay, co-founder of Nextbillion, in an interaction with Analytics India Magazine. It offers customised mapping and data solutions for hyperlocal markets, addressing the challenges in the long run. How AI-powered Hyperlocal Solutions Work Ajay explains that hyperlocal solutions are targeted at the next billion markets which are unique in many ways and a one-size-fits-all solution may not work for them. “For instance, can a mapping or language processing solution for Jakarta holds good in Kuala Lumpur or Mumbai? The answer is No, and the solution lies in hyperlocalising,” he said. In hyperlocal markets, the challenges are not constant across geographies and have to be addressed accordingly. This is where AI comes into the picture, as hyperlocal industries need AI and machine learning to find flexible solutions. One such unique challenge relates to language search. Ajay explains, let’s take the example of black pepper. Across most of India’s Hindi-speaking states, the spice is known as ‘kali mirch’, and many internet users in Tier-2 and 3 cities may not be aware of its English name. AI helps hyperlocal solutions to tune their software to detect words spoken in the native language and enable people to order goods and services in the language they speak and understand. Almost every state in India has different languages, and AI is the most efficient way to address such local nuances. Ajay believes that AI-powered hyperlocal solutions are becoming critical because of the exponential growth and potential in emerging markets today. For instance, during the 2019 Diwali season, Flipkart and Amazon reported robust orders from tier 2 and tier 3 cities. “With the volume comes the challenges such as finding the right delivery address or understanding language nuances. This is where Nextbillion comes into the picture as last-mile delivery becomes critical for e-commerce players and AI-powered hyperlocal mapping solutions can help in pointing the right address with route optimisation,” said Ajay. Nextbillionai Maps And Tasks Nextbillionmaps, the first product by the startup is an AI-powered total mapping solution that provides enterprises with customisable features like routing, navigation, direction and distance matrix. “In a first for the mapping world, it comes with AI tooling capabilities that support map data curation and maintenance. This makes solutions for the logistics, transport, ride-hailing, delivery and e-commerce industry efficient and affordable,” said Ajay. He further added that Nextbillion.ai’s map APIs helps find the best ETAs, find precise pick and drop locations and the best route for the vehicle you are using. It can carry accurate mapping making deliveries efficient. Nextbilliontasks, their second product which is currently being made is a data-intensive service that converts raw data sets to meaningful, structured data that can be used by the end-user. It uses AI to decode data to simplify multilingual texts, image classification, sentiment analysis and video annotations. With a plethora of information that we are surrounded by today, there is a massive demand for high-quality training data in the next billion markets. Nextbillionai provides services such as POIs extractions and enrichment, geotagging and geofencing induced with AI, which is perfect for Nextbillion markets. Both their solutions are built using open-source tools. “While large players like Google already exist in the mapping space, solutions built for developed markets cannot be directly used in emerging markets. This is where our products stand out,” said Ajay. Since it uses open-source data, the solutions can be tailor-made and amended for specific business needs, based on their data and geography. Some of the tools that it uses are Python, angularjs, rust, Spark, vaex, mongo-db, Big query, postgres, Kubernetes, istio. Nextbillionai is spearheading innovation with a strength of 45 employees, spread across Hyderabad, Bengaluru, India and Singapore. Use Cases Ajay shared a few use cases as below: Nextbillionmaps benefitting a delivery company: A delivery executive on a two-wheeler can take a short-cut through a tiny bylane that a car or other 4-wheeler may not. They are forced to take the longer route as the mapping solution doesn’t show the shorter alternative. It may show the most used routes but not the exact end destination. Also, the language used would most likely be English, which delivery partners may not be comfortable with. In the long run, these factors impact turnovers for companies that thrive on quick turnaround time. Relevant mapping APIs can help delivery ninjas can save several hours as the app can automatically organise the route for him. The intelligent mapping solution takes into account traffic snarls, roadblocks or one-way rules. It then computes the best sequence for delivery stops. Ride-hailing industry: Cab drivers using Nextbillionmaps can have access to exact location details, improving the wait-walk-talk matrix and the overall experience of the customer. With pin-pointed drop locations mapped, error margins can be reduced by 70 per cent. Similarly, enquiry calls on finding the whereabouts of an address can reduce by 40 per cent. The popular mapping solutions available in the market today do not provide flexibility to users. Nextbillionmaps fill this gap. It understands local complexities, addresses hyperlocal nuances and is, therefore, tailored to cater to emerging markets. The Eureka Moment Ajay shares that the idea behind starting Nextbillion.ai came as they were gathering insights into the next billion markets – comprising South Asia, the Middle East and North Africa. The population in these countries is internet-savvy, well-travelled and well-informed. Businesses today are scrambling to tap this market. But it faces hyperlocal challenges. “As a growing number of global companies aim to enter the next billion markets as their next growth engine, they will need partners – who offer hyperlocal, cost-effective and customised solutions – to help them scale operations. This is where we come in,” said Ajay as they have partnered with several global brands in countries such as India, Southeast Asia, China and the US. While emerging markets provide immense growth opportunities, there cannot be a one-size-fits-all solution. Nextbillion.ai thrives on such challenges that push for innovative thinking. Future Plans In June, Nextbillion.ai raised a $7 million Series A investment co-led by Lightspeed Venture Partners and Falcon Edge Capital. The company onboarded Navroz Udwadia, Co-Founder and Partner of Falcon Edge Capital and Hemant Mohapatra, Partner at Lightspeed to the Board of Directors. The company plans to use this capital to grow its team, product and invest in R&D. It also plans to build vertical-specific AI capabilities and enter new domains like Natural language processing (NLP), facial recognition and cybersecurity for the next billion markets. “We hope to be at the forefront of disruptive AI technologies in the coming years,” said Ajay. Overcoming The Challenges “Having a great idea is the first big challenge. The second most important step is to have a great team. We have both these steps covered, with experts and veterans on board. The third challenge that startups face is funding. This too is taken care of. We now want to focus on our core competencies and deliver AI-powered solutions across verticals,” shared Ajay. Another challenge that most companies and startups faced this year is the COVID-19 pandemic which has paralysed the global economy in the last few months. “Traditional businesses have been the worst hit. This has stressed the need to integrate technology into businesses, especially hyperlocal ones,” said Ajay. Having said that, Ajay believes that hyperlocal is among the fastest-growing trends this year and with the pandemic-induced lockdowns, people have become more hyperlocal aware. “It has the potential of completely altering the way businesses sell and customers buy and can elevate customer experiences to a whole new level,” said Ajay on a concluding note.","excerpt":"Emerging markets can be more complicated than developed economies in terms of cultures, use of different languages, densely populated cities, poor infrastructure, hyperlocal nuances and more. Ajay Bulusu, Gaurav Bubna and Shaolin Zheng, former employees of Southeast Asian ride-hailing giant, Grab understood these challenges and are now trying to solve these complexities using AI.  With […]","categories":["AI Startups"],"tags":["best facial recognition software","data enrichment ai","which countries have good cybersecurity"],"author_name":"Srishti Deoras","publish_date":"2020-08-21T11:00:00","publication_year":"2020","word_count":1342,"keywords":["Go","machine learning","AI","best facial recognition software","sentiment analysis","which countries have good cybersecurity","data enrichment ai","NLP","Python","Aim","analytics","R","kubernetes"],"extracted_tech_keywords":["AI","machine learning","NLP","analytics","Aim","sentiment analysis","kubernetes","Python","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-startups\/this-ai-startup-aims-to-be-the-google-maps-for-the-emerging-market\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10062987,"title":"IndiaRxiv, a repository for research preprints, relaunched","content":"IndiaRxiv (pronounced India Archive), a dedicated server to upload pre-prints of research, was relaunched on 24 February 2022. Preprints are research papers that have not yet undergone a peer review. Global preprint servers such as arXiv (general) and bioRxiv (for life sciences) play a huge role in pushing the frontiers of scientific knowledge. “Now, Indian researchers can archive and distribute their first drafts of articles to receive early feedback from peers and improve them before submitting them to journals,” said Dr BNS Murthy, Director, ICAR-IIHR and President. “Scientists can share preprints immediately and need not wait too long to publish,” he added. Features and benefits A one-stop portal for Indian researchers’ early stage works.Preprint server allows immediate sharing of research results and can provide quick feedback to help in the revision and preparation of a manuscript.A community of practice, Open Access India, manages the India Archive preprint repository. Indonesia’s INA-Rxiv and Africa’s AfricArxiv served as an inspiration for the creation of IndiaRxiv. However, all such regional repositories could not pay hosting fees to COS, leading to the closing of INA-Rxiv, ArabXiv, and marine conservation server MarXiv in 2020. Geosciences server EarthArXiv reportedly left the COS’s platform due to the exorbitant hosting fee (USD 230,000\/year).","excerpt":"A community of practice, Open Access India, manages the India Archive preprint repository.","categories":["AI News"],"tags":[],"author_name":"Kartik Wali","publish_date":"2022-03-17T12:27:17","publication_year":"2022","word_count":204,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","R"],"extracted_tech_keywords":["AI","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/indiarxiv-a-repository-for-research-preprints-relaunched\/","complexity_score":2,"technical_depth":5,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10165911,"title":"Codeium Offers 50% Student Discount for Windsurf AI Coding Tool","content":"Codeium has announced a 50% discount on the pro subscription plan of its AI coding platform Windsurf for students with an .edu email address. The original pro subscription pricing sits at $15 per month, which gives 500 premium model user prompt credits and 1500 premium model flow action credits. A 50% discount on top of it for the students sounds like an attractive offer. All you have to do is sign up for Codeium with an .edu email address and verify it. After this, one should see the discounted rate on the pricing page. Unfortunately, without an .edu email address, one cannot avail the discount, clarifies Codeium on its students section page. In addition to the discount, Codeium has added the ‘Codeium AI Fellowship’ program. One can apply to join the AI Fellows program and work directly with the Codeium team to bring Windsurf to their campus through talks, workshops, events, and more. The aim of the program is to educate students about AI tools and prepare them for the future of AI-enabled work. Codeium also shared how one can use Windsurf to learn computer science: Considering developers are ditching Cursor for Windsurf, it could be a great opportunity for students to try it out to its full potential with the discounted pricing. “After three months of using Cursor, I switched to Windsurf and found it more cost-effective with its unlimited Cascade base model performing very close to Claude 3.5 Sonnet,” a developer said. In a podcast hosted by Y Combinator, Jared Friedman highlighted the benefits of Windsurf, which he thinks is why people are switching. Combining his sentiment with that of numerous netizens, it seems, Windsurf is becoming the new favourite among developers.","excerpt":"AI coding tool makes it easier for students to enrol.","categories":["AI News"],"tags":["AI coding"],"author_name":"Ankush Das","publish_date":"2025-03-12T12:58:04","publication_year":"2025","word_count":284,"keywords":["AI coding","programming_languages:R","AI","llm_models:Claude","Aim","Claude 3.5","R"],"extracted_tech_keywords":["AI","Claude 3.5","Aim","R","AI coding","llm_models:Claude","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/codeium-offers-50-student-discount-for-windsurf-ai-coding-tool\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10162477,"title":"Infosys and Siemens AG to Advance AI-Powered Digital Learning","content":"India’s tech giant Infosys today announced its collaboration with Siemens AG, a leading technology company, to advance its digital learning initiative with Gen AI. The collaboration aims to upskill 250,000 Siemens employees worldwide. The Siemens My Learning World platform, accessible from anywhere, will integrate Infosys Topaz, an AI-powered solution, and Infosys Wingspan, a next-generation digital learning platform. The enhancements include an AI-powered knowledge assistant, an AI-driven content authoring tool that generates learning materials in multiple languages, and an AI chatbot that helps users navigate courses and find additional resources. The platform will also feature a virtual tutor offering content summaries, translations, and personalised learning options. Jenny Lin, global head of learning & growth at Siemens AG, emphasised the company’s commitment to equipping employees with the right tools, resources, and support for skill development, enabling them to adapt to future challenges. “By leveraging GenAI on Siemens’ digital learning platform, we can foster a culture of lifelong learning and empower our teams to reach their full potential,” Lin said. “Infosys’ expertise in digital transformation and AI is very valuable in creating a more engaging and effective learning experience for everyone.” On the other hand, Jasmeet Singh, executive vice president and global head of manufacturing at Infosys, underscored the transformative impact of AI on digital learning. He described Gen AI as a game-changer that enables organisations to create personalised, engaging, and effective learning experiences that drive development and innovation. “Our collaboration with Siemens to help integrate Infosys Wingspan’s generative AI capabilities, powered by Infosys Topaz, into My Learning World is a testament to Siemens’ commitment to providing their people with the best possible learning and growth opportunities,” Singh said. The Siemens My Learning World platform currently boasts 216,000 active users who have access to over 178,000 learning materials. Additionally, it is utilised by 27,000 factory workers on the shop floor and 65,000 employees through the My Skills feature, which assists users in identifying skill gaps and receiving course recommendations. Additionally, a customer-facing learning platform is being developed with Infosys, offering over 300 technical courses for 50,000 external participants worldwide, including system integrators in the Siemens ecosystem.","excerpt":"The collaboration aims to upskill 250,000 Siemens employees worldwide.","categories":["AI News"],"tags":["Infosys","Siemens"],"author_name":"Shalini Mondal","publish_date":"2025-01-29T20:52:05","publication_year":"2025","word_count":353,"keywords":["GenAI","lifelong learning","Infosys","AI","digital transformation","Git","RAG","Siemens","Aim","generative AI","GAN","R"],"extracted_tech_keywords":["AI","generative AI","GenAI","Aim","RAG","R","Git","GAN","lifelong learning","digital transformation"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/infosys-and-siemens-ag-to-advance-ai-powered-digital-learning\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":34158,"title":"Meet Iris, The World’s First AI-Powered Science Assistant","content":"Artificial Intelligence has now waded into the scientific literature domain. Iris.ai, a literature-exploration tool is the latest AI-based search tools in the town that offers targeted navigation of the knowledge landscape. This AI-based speed-reader came to life to keep a check on abundant scientific literature which is published on a daily basis world, with a statistical count of 2 papers every minute. The core problem that Iris AI is trying to solve is to reduce the time frame to parse and contextualise the sheer volume of scientific literature that’s being published. While there are existing digital services making scientific literature searchable via keywords, such as Google Scholar, there are fewer options for navigating digital content in a way that automatically relates relevant research to reduce time. This is where Iris’s role comes into play. How Iris Works While conventional tools can only act largely as citation indices, AI-based ones have the capability to deliver a deeper view of the literature. Iris.ai has the capability to return a map of thousands of matching documents which are visually segregated by topic just by using a 300 to 500-word description of a researcher’s problem or the URL of an existing research paper. The results that it provides is much quicker and gives a precise overview of what should be relevant to a specific research question. Iris.ai groups multiple documents into varied topics defined by only the words they use. Iris.ai drills the connecting repositories collection or in short CORE collection, a searchable database of more than 134 million open-access papers, as well as journals to which the user’s library provides the overall access. Iris.ai then blends about three algorithms to create ‘document fingerprints’ that reflects word-usage frequencies, which are then used to rank papers according to its significances. The algorithms powering such tools typically perform only two functions that are it extracts scientific content and provides advanced services, such as filtering, ranking or grouping search results. In short, it uses a non-semantic neural topic modelling approach with a heuristic function for hierarchy building. It fetches the abstract from the scientific papers and starts with frequency analysis and pulls out the most essential words and expressions from that paper and then each of those words turned into a vector, in a multi-dimensional vector space. What Iris does is that it goes and finds related words from this whole other body of content, so it works with both synonyms and hypernyms and clusters these words and these bag of clusters of words are related to keywords found in text and finally the calculation of what is the most applicable label for this bag of words from clusters is done as a result. According to Giovanni Colavizza, a research data scientist at the Alan Turing Institute in London, algorithms extracting scientific content leverages natural language processing (NLP) techniques, which seeks to interpret languages as humans use it. For example, developers can use supervised machine learning which involves tagging entities, such as a paper’s authors and references in training sets to teach algorithms to identify and extract them. Colavizza further mentions that to provide more-advanced services, algorithms often construct ‘knowledge graphs’ that detail relationships between the extracted entities and show them to users. For example, the AI could easily suggest that a drug and a protein are related if they’re mentioned in the same sentence. The knowledge graph encodes this as an explicit relationship in a database, and not just in a sentence on a document, essentially making it machine readable. Outlook Even though Iris provides a result which is a map of related papers, the company is also planning to supplement those results by identifying hypotheses being explored in each paper as well It is also developing a parallel, blockchain-based effort called Project Aiur, which seeks to use AI to check every aspect of a research paper against other scientific documents, thus validating hypotheses Although Iris.ai is free for basic queries but costing upwards of  $ 23,000 a year for premium access, which allows more-nuanced searches which thereby accelerates researchers’ entry into newer fields Till now experts have been using free AI-powered tools such as Semantic Scholar which functions like Google Scholar.","excerpt":"Artificial Intelligence has now waded into the scientific literature domain. Iris.ai, a literature-exploration tool is the latest AI-based search tools in the town that offers targeted navigation of the knowledge landscape. This AI-based speed-reader came to life to keep a check on abundant scientific literature which is published on a daily basis world, with a […]","categories":["AI Features"],"tags":["AI (Artificial Intelligence)","research paper"],"author_name":"Martin F.R.","publish_date":"2019-01-25T11:24:03","publication_year":"2019","word_count":696,"keywords":["knowledge graphs","research paper","Go","machine learning","artificial intelligence","programming_languages:R","AI","Git","RAG","NLP","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","NLP","RAG","knowledge graphs","R","Go","Git","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/iris-ai-scientific-literature-tech\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":35220,"title":"Hackathon Is Indicative Of Person’s Problem-Solving Ability But It Cannot Solve Hiring Challenges: Hareesha Prabhu, Philips Innovation Center","content":"Hackathons have become a crucial part of the data science industry. As it becomes one of the crucial tools for hiring, we are seeking the industry’s perspective on how they see hackathons emerging in the analytics and data science field. We interacted with Hareesha Prabhu who is the Director at Philips Electronics India Ltd. With 17 years of experience in the software development industry, he has been instrumental in incubating multiple new teams, hiring new members, setting up the infrastructure and more. Analytics India Magazine: How have hackathons become a crucial part of the hiring process in the data science industry Hareesha Prabhu: I believe that Hackathons are making an impact as a good segway to identify some real potential talent both at the entry level as well as lateral level. It is fascinating to see the same problem being approached by different people and how they interpret the same data\/problem in different creative ways, which gives a good insight into the thinking ability of an individual. AIM: Have hackathons replaced the traditional way of hiring HP: I do not think so. A Hackathon alone cannot solve the hiring challenges. Hackathon is definitely an indication of a person’s problem-solving ability, but that alone is not the attribute we look for during hiring. It is equally important to assess one’s other abilities like communication (both spoken and written), body language, creativity, interpersonal skills, behavioural aspects and much more, for which an in-person interview helps assess the candidate and there is no substitute for it in the present situation. Traditional hackathon also at times assesses one dimension of problem-solving (problems are pre-defined and boundaries defined), but in a personal interview, the problem or the challenge can be tweaked dynamically to assess how the person responds to it. AIM: What is the main purpose of conducting a hackathon for you? HP: When we conduct a hackathon internally within the organization, it serves the purpose of creating that sense of engagement with our employees, gives them a platform to think and act differently and outside their project boundaries. It is a huge opportunity to speed up the learning curve, enables Invention Disclosures etc. Of course, there are certain advantages to the participants like the exposure they get, interaction with the senior leadership, business leaders,  and other functions like Design teams. This also gives the participants an opportunity to know what others are doing around them, break the boundaries and team up with other like-minded colleagues within the organisation. So, in summary, the list of purposes it serves are multi-dimensional and very vast. AIM: How important are hackathons (both internal and external) to boost innovation HP: To a large extent. It is not always true that every solution in a hackathon will result in an “innovation” – the outcome of a hackathon at times is really fascinating – some very simple, some creative, some out of the box, some run of the mill. But more than the solution to a specific problem, when a broad theme is given, we have observed that the innovation intuition kicks in because now the boundary is broader than a specific problem. At the same time, it is not always necessary that we need a hackathon to get innovation in motion. Hackathon only speeds up or brings together ignited minds that increase chances of innovation. AIM: How is hackathon turning into mainstream hiring requirements HP: As I have stated earlier, I don’t believe that hackathon is turning into a mainstream hiring tool as yet, there is still a lot left to explore and a candidate needs to be assessed n other aspects.","excerpt":"Hackathons have become a crucial part of the data science industry. As it becomes one of the crucial tools for hiring, we are seeking the industry’s perspective on how they see hackathons emerging in the analytics and data science field. We interacted with Hareesha Prabhu who is the Director at Philips Electronics India Ltd. With […]","categories":["AI Features"],"tags":["Interviews and Discussions"],"author_name":"Srishti Deoras","publish_date":"2019-02-21T08:40:49","publication_year":"2019","word_count":602,"keywords":["data science","Go","programming_languages:R","AI","innovation","Aim","ViT","analytics","GAN","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","data science","analytics","Aim","R","Go","GAN","ViT","innovation","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/hackathon-is-indicative-of-persons-problem-solving-ability-but-it-cannot-solve-hiring-challenges-hareesha-prabhu-philips-innovation-center\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10103936,"title":"Will Stability AI Survive?","content":"Stability AI is a sinking ship.The company’s business model is in crisis and  is trying to find ways to stay afloat in the market. The London-based AI startup, which gained popularity with its text-to-image generation model Stable Diffusion, is contemplating selling the company as its management is grappling with increasing investor pressure regarding its financial standing. Interestingly, the company approached Cohere and Jasper AI, but Cohere declined to engage in the talks. Meanwhile, the investors want Stability AI founder Emad Mostaque to step down. Coincidentally, Stability AI’s website is showing an error 404 against Mostaque’s name. The relationship between Stability AI and its investors has soured recently as the company is struggling to generate substantial revenue while offering its models for free to customers. One of its key investors, Coatue, has claimed that Mostaque’s poor leadership led to the departure of several top employees, placing the startup in a challenging financial position. Representatives from both Coatue and Lightspeed have already stepped down from Stability’s board, expressing disagreements with Mostaque’s management style. Apparently, Stability AI recently receiving a $50 million investment from Intel didn’t sit well with Coatue, as it has a stake in Intel’s rival, AMD. Competition Galore Stability AI is finding it difficult to acquire customers as big tech giants have entered the foray with  their own image generation models. Meta has Emu Edit and Video, Google has Imagen, and recently, Amazon also launched its own image generation model- Titan Image Generator, now in preview and available for AWS customers on Bedrock. To make matters worse, Dall-E-3 on ChatGPT Plus acted as the final nail in the coffin. ChatGPT today has more than 100 million weekly users. Moreover, Midjourney, a direct competitor to Stable Diffusion, has over 16.4 million active users as of November. On the other hand, Stable Diffusion has more than 10 million daily active users across all channels, according to Mostaque. It is fascinating that, although Midjourney has been following a subscription model from very early on, it has still been able to retain customers. Midjourney’s monthly subscriptions start from $10 and go up to $120. Can Stability AI  turn it around As of now, Stability AI offers Stable LM, Stable Audio, and Stable Diffusion XL. The company recently introduced Stable Video, a new free AI research tool that can turn any still image into a short video. Majority of the above models are free to use. However, Mostaque has recently expressed his intentions to introduce Stability AI Memberships. He believes that in today’s market any AI startup needs to have a business model; otherwise, survival in the market won’t be easy. “We are doing ok as a business and ramping up nicely” he posted on X. The company generated $1.2 million in revenue in August and was projected to reach $3 million this month from software and services, according to a post by Mostaque on X on Monday, which he later deleted. To determine the ideal pricing for his new core Stable models, he ran a poll on X. Ironically, most users voted for $1, indicating that they want the models to be available for free. What should we price @StabilityAI monthly memberships at allowing commercial use of our new core Stable models (eg SDXL Turbo, Stable Video Diffusion) for those making < $1 million a year in revenue?— Emad (@EMostaque) November 29, 2023 It appears that Stability AI has already implemented the subscription model on Stable diffusion XL. If you visit the website and attempt to create an image, it asks for a monthly price of $10. Mostaque mentioned that there has been tension in the organisation lately regarding what to release versus withhold, how to compete on API and other products, including consumer-focused ones. “We want to release good models, and the line between open-source and releasing in the open, plus how to build a sustainable business, has been something we have spent a lot of time thinking about,” he added. Mostaque said that for commercial usage you will compulsorily need to have a Stability AI Membership.”For example, we’re considering, for an indie developer, this fee to be $100 a month, but only if you make above a certain amount of revenue, similar to game engines,” he said. “It’s like Amazon Prime or Netflix for generative AI models,” he added. However, he continued that for non-commercial and academic usage, they will continue to provide the models for free. It’s commendable that Stability AI is making efforts to avoid the need for a sale. However, if the situation arises where selling becomes the best way out, the company should seriously consider approaching Apple.","excerpt":"Stability AI has introduced Stability AI Memberships to make money","categories":["AI Features"],"tags":["AI Tool"],"author_name":"Siddharth Jindal","publish_date":"2023-11-30T17:08:22","publication_year":"2023","word_count":769,"keywords":["Go","ChatGPT","API","AWS","AI","GPT","Ray","Aim","generative AI","AI Tool","R"],"extracted_tech_keywords":["AI","generative AI","ChatGPT","Aim","Ray","AWS","R","Go","API","GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/will-stability-ai-survive\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10119889,"title":"iPad Pro with M4 Chip Enables Seamless AI Tasks","content":"Apple unveiled a new chip in iPad Pro devices, in the latest Let Loose event. Interestingly, the M4 chip has a powerful Neural Engine, capable of 38 trillion operations per second, making it vastly superior for AI tasks and enhancing features like Live Captions and Visual Look Up on the iPad Pro. Additionally, the M4 chip features a powerful Neural Engine,- a remarkable 60x improvement over the first Neural Engine in the A 11 Bionic chip. Moreover, the M4 chip’s combination of advanced ML accelerators, a high-performance GPU, and higher-bandwidth memory makes it exceptionally powerful for AI tasks. “The new iPad Pro with M4 is a great example of how building best-in-class custom silicon enables breakthrough products,” said Johny Srouji, Apple’s senior vice president of Hardware Technologies. In addition to the upgrades on iPad pro and iPad Air, Apple also introduced a new Apple Pencil Pro and Magic Keyboard for the devices Last year, Apple treated AI developers with M3 chips, which lets them work with large transformer models with billions of parameters on the MacBook, seamlessly. M3 prominently features GPUs with “Dynamic Caching,” unlike traditional GPUs, making it useful for game developers and users of graphics-intensive apps like Photoshop or photo-related AI tools. A user said in Hacker news, “If (and that’s a big if) they keep their APIs open to run any kind of AI workload on their chips it’s a strategy that, I personally really really welcome as I don’t want the AI future to be centralised in the hands of a few powerful cloud providers.” If Apple keeps the M4 chip’s APIs open, it would be beneficial for developers as it would allow them to run a variety of AI workloads on the chip, rather than being dependent on a few cloud providers. However, with improved AI capabilities, faster performance and efficiency, and media capabilities will significantly benefit developers in creating more advanced, powerful, and visually stunning applications.","excerpt":"M4 chip is featured with a powerful Neural Engine,- a remarkable 60x improvement over the first Neural Engine in the A11 Bionic chip.","categories":["AI News"],"tags":[],"author_name":"Gopika Raj","publish_date":"2024-05-08T15:23:10","publication_year":"2024","word_count":322,"keywords":["API","programming_languages:R","AI","ML","R"],"extracted_tech_keywords":["AI","ML","R","API","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/ipad-pro-with-m4-chip-enables-seamless-ai-tasks\/","complexity_score":3,"technical_depth":5,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":69169,"title":"What Is Data Wrapping And Why Are Companies Vouching For It","content":"While data has been used for long to improve processes or even benefit by selling it, a research scientist at MIT Center for Information Systems Research suggested a third interesting way that companies utilise data — i.e. wrapping their data around products and services. So, what does this data wrapping mean? While little is known about the concept until now, mostly due to ubiquity of data, data wrapping is primarily packaging the products with data analytics features and experiences that may benefit customers and increase profitability. The process gives access to relevant data to customers that they can explore to gain insights and make better decisions. It is aimed at increasing revenue, loyalty, customer satisfaction, and more. Data Wrapping Is One Of The Ways For Data Monetisation While companies have been monetising data in different ways, they are now sharing their data packaged as products to customers and partners. This mode of data monetisation is data wrapping and is being explored as one of the key ways to achieve monetisation. Some of the ways in which wrapping can be applied are data-fueled dashboard, report, alerts etc. As the article suggests, companies are doing it to provide a better user experience to customers as it helps clients to understand the product and service a company has and use it to their advantage to increase profits and deliver better returns for shareholders. It also highlights that some of the key points to keep in mind about data wrapping are that it is meant for companies’ customers, not employees, it is done as a part of products’ overall feature and experience portfolio, economic returns result from a lift in sales and not from internal business, and more. What Have Companies Done Some instances where we have seen companies use data wrapping are BBVA and PepsiCo. In 2016, BBVA offered a personal finance management app to its customers, which was built on machine learning algorithms. It sorted customer transactions into categories such as rent, food, and entertainment. It then displayed the customer’s expenditures broken down as a simple chart. This was promoted by the company on its digital banking website as a way for customers to better manage their personal budgets. This quickly became the most utilised feature on BBVA website as it provided value-added recommendations for customers. Similarly, last year, PepsiCo launched Pep Worx, a suite of data analytics capabilities. It helped customers successfully launch and manage innovative marketing programs, optimise total store space, help retail customers increase product turns, profits, and more. The capability developed over four years, helped the retail customers solve problems using data analytics-based shopper insights. Pep Worx was used by the company to help transform the nature of its retail customer relationships from transactional to collaborative nature. How To Get Started With Data Wrapping The basic requirement is big data and data science. It is a good idea to begin by focusing on preexisting business intelligence groups, data tools, and analytics talent for data wrapping. However, capabilities and processes that helped a company use data analytics better may not work for its customers. Therefore according to the research article, it is ideal to form a cross-functional team with Finance, IT, and any other area that can offer critical insight. The next step is to design features and experiences that inspire customer action and help them derive the value of saving time, money or gaining important information. The most crucial step is to measure the impact, which is a twofold process. Data wrapping can create value indirectly which can be measured by techniques such as A\/B testing, surveys and pilot studies to get a sense of data wrapping outcomes. Second is for companies to pinpoint the magnitude of value that they are capturing. Some of the effective ways to do data wrapping according to the research article are: Anticipate customer needsAdvise with evidence-based decision making to help customers decide what to doTailoring and adapting to meeting customer needs Act in a way that wrap performs an action to benefit the customer","excerpt":"While data has been used for long to improve processes or even benefit by selling it, a research scientist at MIT Center for Information Systems Research suggested a third interesting way that companies utilise data — i.e. wrapping their data around products and services. So, what does this data wrapping mean?  While little is known […]","categories":["AI Features"],"tags":[],"author_name":"Srishti Deoras","publish_date":"2020-07-07T14:00:00","publication_year":"2020","word_count":672,"keywords":["big data","data science","Go","business intelligence","machine learning","AI","Git","Aim","analytics","R"],"extracted_tech_keywords":["AI","machine learning","data science","analytics","Aim","R","Go","Git","big data","business intelligence"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/what-is-data-wrapping-and-why-are-companies-vouching-for-it\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10086433,"title":"Meta’s Dreams Have Become A Metaverse Nightmare","content":"Meta just announced its financial year and Q4 results, and their investors are very pleased. The stock for the company shot up by 18% after the results were announced, mainly on the back of announcements such as Facebook reaching two billion active users and Instagram’s better AI discovery engine. However, a dark cloud looms on the horizon, and it’s called Reality Labs. Let’s delve deeper into the nitty-gritties of this earnings report. Profit-Makers Meta saw an increase in the average users on their platforms, with all parameters such as daily active people, monthly active people and monthly active users seeing an uptick of 4%. This also translated into better ad impressions, which saw an increase of 18% on a YoY basis. However, the price per ad saw. reduction by 16%. Regardless of this, the report heralds good tidings for the company, as it shows that Apple’s strong stance against advertisers like Meta has not affected its bottomline. In September 2022, Apple introduced a feature called App Tracking Transparency, which allows users to choose whether a given application can track them across different apps and websites. This was a huge blow to Meta, which tracks users’ activity across the web for better targeted advertising. While the tech giant did see a drop in revenue of 1% compared to the numbers from last year, they predicted that their operating expenses would be lower in the coming years due to the widespread layoffs and cost-cutting measures. Meta’s headcount also increased by 20% YoY, but the company made it clear that the 11,000 employees that were laid off were not included in this count. To give you the context, Meta underwent a behemoth restructuring event late last year, of which layoffs were only a small bit. In addition to reshuffling talent between teams, the company also cut down its real estate footprint, both of which resulted in an expense of $4.2 billion in Q4 ’22 alone. The company also focused their efforts on growing areas like their AI discovery engine and the metaverse, but it seems only one of those has paid off. Metaverse, a thorn in its side With the increased focus on improving user experiences on Instagram and Facebook, it seems that Meta’s metaverse dream has lost momentum. The earnings report shows that the company lost a whopping $4.2 billion in Q4 and $13.7 billion over the course of the year in their Reality Labs segment. This is an ongoing trend ever since the launch of this division, with RL losing the company $10 billion in 2021. Reality Labs is a wholly owned subsidiary of Meta focused on the creation and sales of their VR headsets, along with ongoing research in VR and AR technologies. Notably, it is also the division spearheading research into the metaverse. Meta has been on an acquisition spree to build its metaverse over the course of the financial year, which is only partly responsible for the big bill of Reality Labs. To date, the company sells its entry-level VR headsets with a subsidy, even after the $100 price hike in July of last year. This means that the metaverse is a huge money sink for the company, something that will only prove to be a thorn in its side as money gets tighter. More layoffs coming? However, it seems that the management has other ideas to cut costs. Along with reducing their real estate footprint, they are also looking to cut down on employees’ perks. Zuckerberg has also stated that he wants to get rid of Meta’s ‘management culture’, suggesting that the middle management in the company may be at risk. What’s worse, layoffs may be on the horizon as Meta looks to return to its scrappy roots. In a statement, CEO Mark Zuckerberg said, “Our management theme for 2023 is the ‘Year of Efficiency’ and we’re focused on becoming a stronger and more nimble organisation.” He further elaborated in the earnings call, “We closed last year with some difficult layoffs and when we did this, I said clearly that this was the beginning of our focus on efficiency and not the end.” This can be interpreted to mean that the upper management in Meta is looking to lay off more of its workforce. Reports have also emerged that employees in the company are bracing for more layoffs. According to this report, the company is conducting further performance reviews to increase the count of lowest performing workers to 15% of the workforce. Insiders have also placed the next round of layoffs as looking to cut a further 5-10% of the workforce. While one can argue if layoffs are indeed good for the industry, at this point, the worry is if Meta’s metaverse dreams will change into a metaverse nightmare. Investors are currently satiated with the growth of Instagram Reels and Facebook, but Zuckerberg’s metaverse aspirations might be the straw that breaks the camel’s back.","excerpt":"The earnings report shows that the company lost a whopping $4.2 billion in Q4 and $13.7 billion over the course of the year in their Reality Labs segment","categories":["Global Tech"],"tags":[],"author_name":"Anirudh VK","publish_date":"2023-02-02T18:00:00","publication_year":"2023","word_count":817,"keywords":["Go","programming_languages:R","AI","ML","programming_languages:Go","RAG","ViT","GAN","R"],"extracted_tech_keywords":["AI","ML","RAG","R","Go","GAN","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/metas-dreams-have-become-a-metaverse-nightmare\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":11508,"title":"In Review: Ed-tech Start-Up UpGrad&#8217;s &#038; IIIT-B&#8217;s PG Diploma Program in Data Analytics","content":"Plan on upskilling in Data Analytics? Now, you can become a post graduate in Data Analytics with a Diploma from one of the premier technical institutes in India – International Institute of Information Technology Bangalore (IIIT-B) in conjunction with ed-tech startup UpGrad. Co-founded by entrepreneur and media stalwart Ronnie Screwvala and three education experts – Mayank Kumar, Phalgun Kompalli and Ravijyot Chugh – UpGrad is an online education venture offering rigorous, industry-relevant programs for working professionals who plan to constantly upskill themselves to remain relevant. A recent entrant to ed-tech ecosystem (started in 2015), UpGrad’s USP is bringing academia and industry veterans together for the best learning experience. That’s exactly what the super-intensive, online Data Analytics program offers. With a duration of 11-months, the online Data Analytics program build its credibility with faculty from IIIT-B, a well-renowned technology institute in India, and the best minds from the analytics industry (Cognizant, Genpact, Flipkart, InMobi among many others). Here are some of the key facts of the program which is still underway: Duration: 11 months Medium: Online Fee: Rs 2 lakh Number of students enrolled: 300 Faculty: Industry Leaders and IIIT-B faculty impart instructions Choose Domain Specialization: Students can choose from Banking, Healthcare & E-commerce Analytics electives, in the second half of the program Tools: You will learn how to use tools like Spark, Tableau, Hadoop and programming languages R, Python, and MySQL Placement Support: Students get career assistance and facilities such as Resume Review and Interview Preparation 3 month Capstone Project: Get hands-on experience in driving insights from data for businesses Next batch: Starts in March 2017 (Applications open now) Curriculum Industry veterans (leaders from Uber, Cognizant and TataiQ) among others and top academicians from IIIT-B have created this syllabus combining a set of guidelines — from deep understanding of how data analytics spurs business decisions to the ability to develop predictive modelling solutions to support that decision making process. Broadly put, the program imparts instructions in Data Management, Statistics, Machine Learning and Big Data. The program prepares students for careers in analytics, predictive modeling, business intelligence and data mining in data-intensive industries such as Banking, Healthcare, and E-commerce, among others. The 11-month program tackles the fundamentals of Statistics and Data Management, Data Sets and Data Models, Business and Data Understanding, Data Warehousing, Data Visualization, Predictive Analysis and Big Data Analysis. The program can be divided in two components — first six months focuses on fundamentals of Data Analytics and second half is Domain Elective, depending on their career objective, students can choose from BFSI, Healthcare, E-commerce or Retail\/FMCG. What you learn at UpGrad: How can you build a successful career in analytics by understanding the academic concepts and industry applications? What are the correct statistical and machine learning methods that will produce the best results for an organization and which method to use in which scenario? How to combine statistical knowledge, technical know-how and business domain to become a data driven leader? Domain Specialization Since the program is designed in a way that it is industry relevant, it has concrete application to some of the most data-intensive industries in India such as Banking, Healthcare, E-commerce, Retail and FMCG among others. “In the Indian eco-system, these industries are the biggest consumers of data and the data DNA already exists within them. For example, an FMCG firm may require a data analyst to optimize logistics and warehousing. An e-tailer might want to know the online preferences of his customers. Within the program, there are a wide range of such case studies where students can learn how to apply analytical methods in real-life scenarios,” shared Rohit Sharma, Program Director, UpGrad. The first half of the program covers the fundamentals and the advanced data science techniques and the second half covers domain electives. Here – the students can choose whichever domain they wish to focus on. Some of the case-studies are Uber Supply Demand Gap: In this case study the students use analytics to identify why Uber sometimes faces a supply-demand challenge and what can be done to overcome it. Telecom Churn Prevention: It is one of the most competitive sectors and the existing players face the challenge of customer churn (opting for other service providers). The students learn to use machine learning to identify the cause of customer churn. E-commerce Market Mix Modelling: Understanding the types of customers in the target group e-tailers target their marketing effort. In this case study the students learn to use modelling to figure out the optimal spend across channels to drive sales. Placement Prospects Strong industry network of UpGrad and IIIT-Bangalore should be helpful in creating attractive job opportunities for the students. Since the program is in its maiden run (the first batch is still underway), it is too early to comment on placement. An important point to keep in mind is no learning institute offers 100% guarantee. However, UpGrad provides placement assistance and has developed an active industry partner pool (Gramener, Genpact, Uber, Fractal Analytics, Cognizant) to bolster recruitment. The popularity of the program can be gauged by the sheer number of students who enrolled – 300. But how does one measure the effectiveness of this online program, given that it’s in first edition? According to Rohit Sharma, the director of the program, “In online courses, the massive problem is dropouts. We have a 96% completion rate, which means the program is highly successful. Our focus is on providing a seamless learning experience”. Last Word Pros: One of the best high points of the program is IIIT-B (which has a great faculty) that brings credibility. UpGrad is providing a PG Diploma from IIIT-B as opposed to Management institutes and learning centres that provide a certificate. The Rs 2 lakh fee is comparable to what is charged by learning centres or business institutes. UpGrad is providing Career Assistance which most learning centres do not. In order to build a student community & peer support mechanisms, UpGrad has its own Discussion Forum where students are encouraged to interact and debate out their doubts or challenges. Also, UpGrad conducts an offline meetup every month in Delhi, Bangalore and Mumbai where one industry speaker and one faculty member are present. Lastly, the program is a great combination of academic and industry experience Cons: It’s a recent entrant in ed-tech ecosystem (UpGrad started in 2015), hence other start-ups already have a head start on them. The first batch is still underway and we can’t comment on any Placement Prospects All in all, UpGrad is a good stepping-stone for upgrading the Data Analytics skills as it capitalizes on IIIT-B’s esteemed faculty and the rich industry experience of leading minds from the analytics field. Moreover, in our recently published top course ranking, this course is the lone online-only course in India to make it to top 10. Also, remember that a diploma from IIIT-B is a no mean feat to achieve. You can apply for the program here.","excerpt":"Plan on upskilling in Data Analytics? Now, you can become a post graduate in Data Analytics with a Diploma from one of the premier technical institutes in India – International Institute of Information Technology Bangalore (IIIT-B) in conjunction with ed-tech startup UpGrad. Co-founded by entrepreneur and media stalwart Ronnie Screwvala and three education experts – […]","categories":["AI Trends"],"tags":[],"author_name":"Дарья","publish_date":"2016-12-05T13:02:10","publication_year":"2016","word_count":1155,"keywords":["data science","Go","machine learning","AI","ML","RAG","Python","analytics","SQL","R"],"extracted_tech_keywords":["AI","machine learning","ML","data science","analytics","RAG","Python","R","SQL","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/review-ed-tech-start-upgrads-data-analytics-pg-diploma-program-offered-iiit-b\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10161448,"title":"India Set to Host 620+ New GCCs by 2030: ANSR Report","content":"ANSR, one of India’s leading Global Capability Centres (GCCs) enablers, released its Q3 GCC Report, which highlights India’s ascension as a global hub for GCCs. Over 450 Forbes Global 2000 companies operate 825+ GCCs across the country, employing more than 1.3 million professionals. The report predicts that by 2030, India could host 620+ additional GCCs, representing over 32% of Forbes Global 2000 enterprises. The robust GCC ecosystem in India thrives on its skilled workforce, supportive government policies, innovative startups, and a strong network of service providers. This combination positions India as a premier destination for global business innovation. Source: ANSR GCC Report Bengaluru leads as the top GCC hub, hosting over 285 companies and employing 560,000 professionals. Hyderabad follows with 110+ GCCs and 190,000 employees, attracting interest due to its infrastructure, talent availability, and business-friendly policies. Non-metro cities like Ahmedabad, Vizag, Trivandrum, and Coimbatore are also emerging as preferred destinations for cost-effective operations and access to untapped talent pools. The BFSI sector dominates, representing 18% of GCCs in India, followed by Manufacturing and Retail\/CPG. Newly established GCCs are prioritising advanced digital capabilities, with 90% focusing on AI\/ML and data analytics. Other priorities include cybersecurity (40%) and cloud capabilities (50%), signaling a shift towards innovation-driven operations. GCC leasing activity surged in Q3 2024, accounting for 44% of total office leasing in India. Bengaluru captured 62% of this demand, followed by Hyderabad at 15%. Companies like Walmart, Sanofi, NatWest Group, and Flutter Entertainment have been key players in this expansion, leveraging India’s talent pool and infrastructure. According to a report by CBRE, the year saw a gross absorption of 79.0 million sq ft, a 16% YoY growth. GCCs accounted for 37% of the total office leasing activity in 2024, solidifying their importance in India’s commercial real estate market. BFSI firms and technology companies contributed 44% of GCC leasing, showcasing their integral role in the office sector’s growth. Source: ANSR GCC Report GCCs are evolving from operational support roles to becoming integral in global decision-making, fostering innovation through collaboration and partnerships. They are also exploring non-metro cities in Karnataka, Tamil Nadu, and Uttar Pradesh to benefit from lower operational costs and a favourable business environment. Vikram Ahuja, co-founder of ANSR and CEO of Talent500, remarked, “Global Capability Centres are driving India’s growth story by leading the tech revolution across sectors through cutting-edge innovation. With a skilled workforce and emerging technologies like AI and cybersecurity, India is set to accommodate 620+ new GCCs by 2030, creating opportunities for over 1.9 million professionals.”","excerpt":"Bengaluru leads as the top GCC hub, hosting over 285 companies and employing 560,000 professionals.","categories":["GCC"],"tags":["GCC","GCC india"],"author_name":"Mohit Pandey","publish_date":"2025-01-15T13:38:43","publication_year":"2025","word_count":418,"keywords":["Go","GCC","AI","innovation","ML","Git","RAG","ViT","analytics","GCC india","R","T5"],"extracted_tech_keywords":["AI","ML","analytics","RAG","R","Go","Git","T5","ViT","innovation"],"url":"https:\/\/analyticsindiamag.com\/gcc\/india-set-to-host-620-new-gccs-by-2030-ansr-report\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10138020,"title":"Foxconn to Build Taiwan’s Fastest AI Supercomputer With NVIDIA’s flagship Blackwell","content":"Nvidia, a leading chipmaker, and Foxconn, the Taiwan-based contract manufacturer, have unveiled plans to construct Taiwan’s largest supercomputer, housed at the Hon Hai Kaohsiung Supercomputing Center. The system will be powered by Nvidia’s Blackwell architecture, utilizing the GB200 NVL72 data centre platform, as details shared on Nvidia’s blog. Once operational, Foxconn intends to use the supercomputer to drive advancements in cancer research, large language model development, and smart city innovations, establishing Taiwan as a leader on the global AI map. Foxconn is the largest electronics manufacturer globally, creating products for top technology brands. Touted as the Fastest Supercomputer Foxconn’s “three-platform strategy” centres on smart manufacturing, smart cities, and electric vehicles. The new supercomputer will support advancements in digital twins, robotic automation, and urban infrastructure in Kaohsiung. The company said that construction is underway, with the first phase set to be operational by mid-2025 and full deployment expected in 2026. The project will use NVIDIA technologies like Omniverse and Isaac robotics to improve manufacturing processes. “Powered by NVIDIA’s Blackwell platform, Foxconn’s new AI supercomputer is one of the most powerful in the world, representing a significant leap forward in AI computing and efficiency,” said Foxconn Vice President and Spokesperson James Wu. The GB200 NVL72 is NVIDIA’s data centre platform for AI tasks. Each rack has 36 NVIDIA Grace processors and 72 NVIDIA Blackwell GPUs connected with NVIDIA’s NVLink, providing 130TB per second of bandwidth. The GPUs can work together as a single unit, perfect for training large AI models and running complex tasks in real-time.","excerpt":"These supercomputers, employing NVIDIA’s technology, represent a significant leap in computational power.","categories":["AI News"],"tags":["foxconn"],"author_name":"Aditi Suresh","publish_date":"2024-10-10T12:26:49","publication_year":"2024","word_count":254,"keywords":["programming_languages:R","AI","innovation","Git","automation","ai_applications:robotics","R","foxconn"],"extracted_tech_keywords":["AI","R","Git","automation","innovation","programming_languages:R","ai_applications:robotics"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/foxconn-to-build-taiwans-fastest-ai-supercomputer-with-nvidias-flagship-blackwell\/","complexity_score":3,"technical_depth":7,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10011163,"title":"Gaussian Mixture Model Clustering Vs K-Means: Which One To Choose","content":"In recent times, there has been a lot of emphasis on Unsupervised learning. Studies like customer segmentation, pattern recognition has been a widespread example of this which in simple terms we can refer to as Clustering. We used to solve our problem using a basic algorithm like K-means or Hierarchical Clustering. With the introduction of Gaussian mixture modelling clustering data points have become simpler as they can handle even oblong clusters. It works in the same principle as K-means but has some of the advantages over it. It tells us about which data belongs to which cluster along with the probabilities. In other words, it performs hard classification while K-Means perform soft classification. Here, we will implement both K-Means and Gaussian mixture model algorithms in python and compare which algorithm to choose for a particular problem. Let’s get started. About the dataset The iris dataset can be downloaded from the following link. It gives the details of the length and breadth of the three flowers: Setosa, Versicolor, Virginica. Practical Implementation Import all the libraries required for this project. # import some libraries import numpy as np import pandas as pd import matplotlib as mpl import matplotlib.pyplot as plt import seaborn as sns sns.set(style=\"white\", color_codes=True) import warnings warnings.filterwarnings(\"ignore\") # show plots inline %matplotlib inline data = pd.read_csv('Iris.csv') data = data.drop('Id', axis=1) # get rid of the Id column - don't need it data.sample(5) Store the independent and dependent variable in the X,y variable. X = data.iloc[:,0:4] y = data.iloc[:,-1] The data is unbalanced means some of the features are larger than others. They will dominate the dataset. To avoid this problem, we need to do feature scaling. from sklearn import preprocessing scaler = preprocessing.StandardScaler() scaler.fit(X) X_scaled_array = scaler.transform(X) X_scaled = pd.DataFrame(X_scaled_array, columns = X.columns) X_scaled.sample(5) Let’s fit our model using the KMeans Algorithm. We will mention ncluster=3 as the species has 3 categories. K-Means Clustering from sklearn.cluster import KMeans nclusters = 3 # this is the k in kmeans seed = 0 km = KMeans(n_clusters=nclusters, random_state=seed) km.fit(X_scaled) # predict the cluster for each data point y_cluster_kmeans = km.predict(X_scaled) y_cluster_kmeans import matplotlib.patches as mpatches red_patch = mpatches.Patch(color='red', label='Setosa') green_patch = mpatches.Patch(color='green', label='Versicolor') blue_patch = mpatches.Patch(color='blue', label='Virginica') colors = np.array(['blue', 'red', 'green']) plt.scatter(X_scaled.iloc[:, 2],X_scaled.iloc[:, 3],c=colors[y_cluster_kmeans]) plt.xlabel(\"PetalLengthCm\") plt.ylabel(\"PetalWidthCm\") plt.legend(handles=[red_patch, green_patch, blue_patch]) plt.show() The above result gives us three clusters:Setosa,Versicolor and Virginica. We can see there is some overlap between Versicolor and Virginica. So it is not easy to separate the flowers using K-Means. from sklearn import metrics score = metrics.silhouette_score(X_scaled, y_cluster_kmeans) Score The silhouette score of 0.45 shows there is the intermediate distance(neither far nor near) between the clusters. scores = metrics.silhouette_samples(X_scaled, y_cluster_kmeans) sns.distplot(scores); from sklearn.metrics.cluster import adjusted_rand_score score = adjusted_rand_score(y, y_cluster_kmeans) score We can use the adjusted rand score to quantify the goodness of clustering. From the above result, KMeans gives a score of 0.62 which is pretty decent. Gaussian Mixture Modelling Clustering Now let’s fit the model using Gaussian mixture modelling with nclusters=3. from sklearn.mixture import GaussianMixture gmm = GaussianMixture(n_components=nclusters) gmm.fit(X_scaled) # predict the cluster for each data point y_cluster_gmm = gmm.predict(X_scaled) Y_cluster_gmm red_patch = mpatches.Patch(color='red', label='Setosa') green_patch = mpatches.Patch(color='green', label='Versicolor') blue_patch = mpatches.Patch(color='blue', label='Virginica') colors = np.array(['blue', 'red', 'green']) plt.scatter(X_scaled.iloc[:, 2],X_scaled.iloc[:, 3],c=colors[y_cluster_gmm]) plt.xlabel(\"PetalLengthCm\") plt.ylabel(\"PetalWidthCm\") plt.legend(handles=[red_patch, green_patch, blue_patch]) plt.show() The plot displays very little overlap between the data points of different clusters. Gaussian model gives us a better result than K-Means. from sklearn import metrics score = metrics.silhouette_score(X_scaled, y_cluster_gmm) Score from sklearn.metrics.cluster import adjusted_rand_score score = adjusted_rand_score(y, y_cluster_gmm) score The Gaussian mixture model has an adjusted rand score of 0.9. It gives a better fit of clustering. Conclusion In this article, we have discussed the basics of Gaussian mixture modelling. Further, we have compared it with K-Means with the adjusted rand score. It shows how efficient it performs compared to K-Means. Hence, it is advisable to study Gaussian mixture modelling in more depth.The complete code of the above implementation is available at the AIM’s GitHub repository. Please visit this link to find the notebook of this code.","excerpt":"In recent times, there has been a lot of emphasis on Unsupervised learning. Studies like customer segmentation, pattern recognition has been a widespread example of this which in simple terms we can refer to as Clustering. We used to solve our problem using a basic algorithm like K-means or Hierarchical Clustering. With the introduction of Gaussian mixture modelling clustering data points have become simpler as they can handle even oblong clusters. It works in the same principle as K-means but has some of the advantages over it.","categories":["Deep Tech"],"tags":["clustering","clustering algorithms","k-means","Unsupervised Learning"],"author_name":"Ankit Das","publish_date":"2020-11-04T13:00:42","publication_year":"2020","word_count":665,"keywords":["Go","NumPy","AI","R","clustering","Python","Ray","Aim","Seaborn","clustering algorithms","Matplotlib","Pandas","Unsupervised Learning","k-means"],"extracted_tech_keywords":["AI","Aim","Ray","Pandas","NumPy","Matplotlib","Seaborn","Python","R","Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/gaussian-mixture-model-clustering-vs-k-means-which-one-to-choose\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10060057,"title":"Google rolls out visual interface for Speech-to-Text API in cloud","content":"Google has announced the general availability of a new visual interface for Google Cloud’s Speech-to-Text API in the Google Cloud console. The new visual interface aims to reduce the complexities developers face when they use the Speech-to-Text API. The STT API allows developers to convert speech into text by leveraging Google’s years of research in automatic speech recognition and transcription technology. The API helps users add speech functionality to their applications to better serve consumer needs. The STT API can be used from dictation and short commands to captioning and subtitles. However, to achieve the highest accuracy on any AI use case, STT requires careful testing and fine-tuning. Earlier, developers building on the STT API had to do this work manually, needed familiarity with GCP integration concepts and had to either build their own tools or manage various scripts and API calls to fully understand the API documentation.  These actions  made measuring, customising and improving models even more difficult The availability of STT API within the Google Cloud Console simplifies the process, facilitating iteration and integration of models into developers’ applications by letting developers perform every API function from within the console. With this new addition, developers have the ability to manage and quickly iterate the STT model customisations with Model Adaption. The STT API is available in all Google Cloud regions and can be accessed by all GCP users at no additional cost. The STT API supports over 70 languages in 120 different local variants.","excerpt":"The Speech-to-text API is available in all Google Cloud regions and can be accessed by all GCP users.","categories":["AI News"],"tags":["gcp","Google Cloud"],"author_name":"SharathKumar Nair","publish_date":"2022-02-08T12:19:09","publication_year":"2022","word_count":246,"keywords":["Go","API","Google Cloud","GCP","programming_languages:R","AI","cloud_platforms:GCP","gcp","RAG","Aim","cloud_platforms:Google Cloud","R"],"extracted_tech_keywords":["AI","Aim","RAG","GCP","R","Go","API","cloud_platforms:GCP","cloud_platforms:Google Cloud","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/google-rolls-out-visual-interface-for-speech-to-text-api-in-cloud\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10007792,"title":"A Fun Project On Building A Face-Swapping Application With OpenCV","content":"Most of you would have used or seen many filters that let you swap your face with your friends, celebrities or even animals. These filters are very popular in social media platforms such as Instagram, Snapchat and face app. Have you ever wondered how this is done? With Python and OpenCV it is actually very simple to build this application. The concept behind this is to detect certain points on the face and then replace it with the swapping image. Though simple, it does involve a lot of factors like lighting, facial structure and camera angle. To avoid these complications, we will implement this on two images of celebrities. In this article, we will implement a face-swapping technique for two images of celebrities using OpenCV and python. Steps used for this project: Taking two images – one as the source and another as a destination.Using the dlib landmark detector on both these images. Joining the dots in the landmark detector to form triangles. Extracting these trianglesPlacing the source image on the destinationSmoothening the face Image selection You can select any two images of your choice. It could be your image and your friend, it could be two celebrity images as well. I have selected two images of Shah Rukh Khan  and Katrina Kaif. Both the images are front-facing and are well lit. Source:Destination: Now let us read these images using OpenCV code. import cv2 import numpy as np import dlib import time source_image = cv2.imread(\"srk.jpg\") source_image_gray = cv2.cvtColor(source_image, cv2.COLOR_BGR2GRAY) dest_image = cv2.imread(\"katrina-kaif.jpg\") dest_image_gray = cv2.cvtColor(dest_image, cv2.COLOR_BGR2GRAY) mask = np.zeros_like(source_image_gray) Using the dlib landmark detector on the images Dlib is a python library that provides us with landmark detectors to detect important facial landmarks. These 68 points are important to identify the different features in both faces. To download this use this link. Once we have the 68 points shape predictor downloaded, let us apply them to the first face. Then we will use the convexhull to detect the faces after using the landmark detector. land_detector = dlib.get_frontal_face_detector() predictor = dlib.shape_predictor(\"shape_predictor_68_face_landmarks.dat\") source_face = land_detector(source_image_gray) for face in source_face: landmarks = predictor(source_image_gray, face) points = [] for n in range(0, 68): x = landmarks.part(n).x y = landmarks.part(n).y points.append((x, y)) face_point = np.array(points, np.int32) convexhull = cv2.convexHull(face_point) These will create 68 points on the face as shown below. For destination face: dest_face = land_detector(dest_image_gray) for face in dest_face: landmarks = predictor(dest_image_gray, face) points2 = [] for n in range(0, 68): x = landmarks.part(n).x y = landmarks.part(n).y points2.append((x, y)) Joining the dots in the landmark detector to form triangles for the source image. To cut a portion of the face and fit it to the other we need to analyse the size and perspective of both the images. To do this, we will split the entire face into smaller triangles by joining the landmarks so that the originality of the image is not lost and it becomes easier to swap the triangles with the destination image. This entire process is called Delaunay triangulation. Note: Since we do this with respect to the facial landmark of the source the following code has to be put inside the for loop of source image. rectangle = cv2.boundingRect(convexhull) divide_2d = cv2.Subdiv2D(rectangle) divide_2d.insert(landmarks_points) split_triangle = divide_2d.getTriangleList() split_triangle = np.array(split_triangle, dtype=np.int32) cv2.fillConvexPoly(mask, convexhull, 255) face_image_1 = cv2.bitwise_and(source_image, source_image, mask=mask) face_points2 = np.array(points2, np.int32) convexhull2 = cv2.convexHull(face_points2) def extract_index_nparray(nparray): index = None for num in nparray[0]: index = num break return index join_indexes = [] for edge in triangles: first = (edge[0], edge[1]) second = (edge[2], edge[3]) third = (edge[4], edge[5]) index_edge1 = np.where((points == first).all(axis=1)) index_edge1 = extract_index_nparray(index_edge1) index_edge2 = np.where((points == pt2).all(axis=1)) index_edge2 = extract_index_nparray(index_edge2) index_edge3 = np.where((points == pt3).all(axis=1)) index_edge3 = extract_index_nparray(index_edge3) if index_edge1 is not None and index_edge2 is not None and index_edge3 is not None: triangle = [index_edge1, index_edge2, index_edge3] join_indexes.append(triangle) Now that we have the triangles for the source image we need to make sure the same space can be extracted from the destination so that the overlapping can be done smoothly. To do this, we follow a slightly different approach to the destination image. The destination image needs to have the same patterns of triangles as the source. To do this we will create masks for the images as follows. First, let us create the source image. We are trying to match the patterns of first and second images here. source_mask = np.zeros_like(source_image_gray) new_face = np.zeros_like(dest_image) for index in indexes_triangles: tri_one = points[index[0]] tri_two = points[index[1]] tri_three = points[index[2]] triangle1 = np.array([tri_one, tri_two, tri_three], np.int32) first_rect = cv2.boundingRect(triangle1) (x, y, w, h) = first_rect cropped_triangle = source_image[y: y + h, x: x + w] cropped_tr1_mask = np.zeros((h, w), np.uint8) pts = np.array([[tri_one[0] - x, tri_one[1] - y], [tri_two[0] - x, tri_two[1] - y], [tri_three[0] - x, tri_three[1] - y]], np.int32) cv2.fillConvexPoly(cropped_tr1_mask, pts, 255) cv2.line(source_mask, tri_one, tri_two, 255) cv2.line(source_mask, tri_two, tri_three, 255) cv2.line(source_mask, tri_one, tri_three, 255) As you can see we have created a mask of the image. Once we have the mask for the source image and the triangle points we can do the same for the destination as well. But for the destination, we need to crop out the region corresponding to the source mask. We can do this as shown below. tri2_one = points2[index[0]] tri2_two = points2[index[1]] tri2_three = points2[index[2]] triangle2 = np.array([tri2_one, tri2_two, tri2_three], np.int32) second_rect = cv2.boundingRect(triangle2) (x, y, w, h) = second_rect cropped = np.zeros((h, w), np.uint8) points2 = np.array([[tri2_one[0] - x, tri2_one[1] - y], [tri2_two[0] - x, tri2_two[1] - y], [tri2_three[0] - x, tri2_three[1] - y]], np.int32) cv2.fillConvexPoly(cropped, points2, 255) This is the destination image triangles. Extracting these triangles Once we have the triangles in source and destination the next step is to extract them from the source image. But we also need to take the coordinates of the destination triangles so that the sizes of the two can match. This technique is also called warping. points = np.float32(points) points2 = np.float32(points2) transform = cv2.getAffineTransform(points, points2) warping = cv2.warpAffine(cropped_triangle, transform, (w, h)) warping = cv2.bitwise_and(warping, warping, mask=cropped) Placing the source image on the destination Now, we can reconstruct the destination image and start placing the source image on the destination one. First we will make some alterations in the destination face. Then, we will make sure the lines created do not appear in the final output. ht, wt, filters = dest_image.shape dest_face = np.zeros((ht, wt, filters), np.uint8) facial_area = dest_face[y: y + h, x: x + w] facial_area_gray = cv2.cvtColor(facial_area, cv2.COLOR_BGR2GRAY) _,triangle_mask = cv2.threshold(facial_area_gray, 1, 255, cv2.THRESH_BINARY_INV) warping = cv2.bitwise_and(warping, warping, mask=triangle_mask) facial_area = cv2.add(facial_area, warping) dest_face[y: y + h, x: x + w] = facial_area Finally, we will place the source image on the destination final_mask = np.zeros_like(dest_image_gray) head_mask = cv2.fillConvexPoly(final_mask, convexhull2, 255) final_mask = cv2.bitwise_not(head_mask) combine = cv2.bitwise_and(dest_image, dest_image, mask=final_mask) output = cv2.add(combine, dest_face) Thought the images have been swapped, there is no match in color or smoothness in the swapping. To eliminate this we need to do another process. Smoothening the face The final step is to change the colours and to make the swapping look better. To do this OpenCV provides a library called seamless cloning. (x, y, w, h) = cv2.boundingRect(convexhull2) seamless= (int((x + x + w) \/ 2), int((y + y + h) \/ 2)) seamlessclone = cv2.seamlessClone(output, dest_image, head_mask, seamless, cv2.NORMAL_CLONE) cv2.imshow(\"seamlessclone\", seamlessclone) cv2.waitKey(0) cv2.destroyAllWindows() The final output is like this You can see that this has some gradient and better fit when compared to the previous output. Conclusion In this article, we saw how to build a face-swapping application with OpenCV and successfully built it on two photos. It is quite simple and interesting to build this and a lot of fun to build as well. You can also check out this for building your own face filters.","excerpt":"In this article, we will implement a face-swapping technique for two images of celebrities using OpenCV and python.","categories":["Deep Tech"],"tags":["OpenCV","Python"],"author_name":"Bhoomika Madhukar","publish_date":"2020-09-21T16:00:59","publication_year":"2020","word_count":1288,"keywords":["NumPy","TPU","AI","RPA","ML","OpenCV","Python","Ray","programming_languages:Python","R"],"extracted_tech_keywords":["AI","ML","Ray","OpenCV","NumPy","TPU","Python","R","RPA","programming_languages:Python"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/a-fun-project-on-building-a-face-swapping-application-with-opencv\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10053397,"title":"ZS To Hire Over 4000 People Across Its Offices In India In 2022","content":"ZS—a global professional services firm – has projected to add 4000 more people in 2022 across the three ZS offices in India. The company has surpassed its annual hiring projection for 2021 within three quarters, rolling out almost 40% of the offers for tech roles that require proficiency in skills like artificial intelligence (AI), machine learning, big data, Python, cloud technologies and data science, among others. “ZS is experiencing growth across all its core business areas, necessitating a greater demand for talent with a diverse set of skills. Our teams in India play a pivotal role in helping the firm remain ahead of the curve as we focus on growing tech capabilities that will drive innovation for our global offerings,” said Mohit Sood, Regional Managing Principal and Head, India, ZS. “We are striving to capitalize on technology advancements to create and sustain value for our clients. To support this, we are looking for people with strong technical capabilities, in addition to strategic problem-solving skills, excellent collaboration and strategic implementation, to join our growing team. In the coming quarters, we intend to lead the way by using transformational approaches to develop, engage and retain diverse talent.” The firm originally projected to hire approximately 3,000 employees in 2021, but it has surpassed that figure, with 3,500 new joiners who would have joined by the end of the year. Esha Gulati, HR lead, Asia, ZS, said, “With the accelerated pace of digital adoption, we are looking at strengthening capabilities in emerging tech roles including the cloud, AI and big data, and intensifying our hiring efforts in the coming year to meet our goal of getting 4000 new recruits. The challenges of remote work have taught us that weaving together human touchpoints with the benefits of innovative technology is imperative to reimagining how we work and how we see our workplaces. At ZS, we believe in building high-performing teams, coaching and mentoring our people, creating an equitable career path and providing progression opportunities to everyone. We are excited to welcome new talent joining ZS in the coming months and years.” Additionally, ZS increased investment in new talent, which accounted for 45% of all recruits, bringing in different perspectives and new ways of working. This year alone, Campus Beats—ZS’s campus engagement program—saw participation from 100+ tier 1 and tier 2 engineering colleges, tapping into a large talent pool. So far, ZS has hired more than 1,200 graduates from campuses for various technical and non-technical roles and has extended offers to another 1,500 who will join in 2022.","excerpt":"ZS has hired more than1,200 graduates from campuses for various technical and non-technical roles and has extended offers to another 1,500 who will join in 2022.","categories":["AI News"],"tags":["data science jobs in India"],"author_name":"kumar Gandharv","publish_date":"2021-11-12T11:01:36","publication_year":"2021","word_count":422,"keywords":["big data","data science","Go","API","machine learning","artificial intelligence","AI","Git","data science jobs in India","Python","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","data science","Python","R","Go","Git","API","big data"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/zs-to-hire-over-4000-people-across-its-offices-in-india-in-2022\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10043506,"title":"Comprehensive Guide To Time Series Analysis Using ARIMA","content":"Time series data is a set of observations collected through repeated measurements over time. Plotting the points on a graph, one of our axes would always be time. Time series data is everywhere since time is a constituent of everything observable. As the world advances every day with technology, sensors and systems constantly produce a relentless stream of time series data. Data of this nature has numerous applications across a variety of industries. Time series data is data that is observed at different points in time. This is opposite to cross-sectional data, which observes individuals, companies, etc., at a single point in time.  Time-series databases are highly popular and provide a wide spectrum of numerous applications such as stock market analysis, economic and sales forecasting, budget analysis, to name a few. They are also useful for studying natural phenomena like atmospheric pressure, temperature, wind speeds, earthquakes, and medical prediction for treatment. An observed time series can be decomposed into three main components: the trend i.e the long cycle, the seasonal systematic or calendar related movements, and the irregular unsystematic or short term fluctuations. To find insights from these components, time series analysis can be made use of. Time Series Analysis finds hidden patterns and helps obtain useful insights from the time series data. Time Series Analysis is useful in predicting future values or detecting anomalies from the data.  Such analysis typically requires many data points to be present in the dataset to ensure consistency and reliability. An extensive data set ensures that you have a representative sample size and that the analysis performed can cut through the noisy data. It helps organizations understand what the underlying causes of trends or systemic patterns detected over time are. Using data visualization, better and interpretable insights can be found that can show seasonal trends and help dig deeper into why these trends occur. The different types of models and analyses that can be created through time series analysis are: Classification: To Identify and assign categories to the data.Curve fitting: Plot the data along a curve and study the relationships of variables present within the data.Descriptive analysis: Help Identify certain patterns in time-series data such as trends, cycles, or seasonal variation.Explanative analysis: To understand the data and its relationships, the dependent features, and cause and effect and its tradeoff.Exploratory analysis: Describe and focus on the main characteristics of the time series data, usually in a visual format.Forecasting: Predicting future data based on historical trends. Using the historical data as a model for future data and predicting scenarios that could happen along with the future plot points.Intervention analysis: The Study of how an event can change the data.Segmentation: Splitting the data into segments to discover the underlying properties from the source information. As time-series analysis includes many categories and data variations, the analysts sometimes have to deal with and create complex models. Although, analysts can’t account for all variances, and sometimes one can’t generalize the same specific model to every sample. If created too complex or trying to do too many things, models can lead to a lack of fit. The Lack of fit or an overfitting model leads to the model not distinguishing between a random error and true relationship, leaving analysis biased, partially skewed, and forecasts being incorrect. What is ARIMA? A popular and very widely used statistical method for time series forecasting and analysis is the ARIMA model. ARIMA is an acronym that stands for AutoRegressive Integrated Moving Average. It is a class of models that capture a spectrum of different standard temporal structures present in time series data. By implementing an ARIMA model, you can forecast and analyze a time series using past values, such as predicting future prices based on historical earnings. Univariate models such as these are used to understand better a single time-dependent variable present in the data, such as temperature over time. They predict future data points of and from the variables. These models work on the ideal assumption of the data being stationary. A standard notation used for describing ARIMA is by parameters p,d and q. The parameters are substituted with an integer value to indicate the specific ARIMA model being used quickly. The parameters of the ARIMA model are further described as follows: p: Stands for the number of lag observations included in the model, also known as the lag order.d: The number of times the raw observations are differentiated, also called the degree of differencing.q: Is the size of the moving average window and also called the order of moving average. Getting Started This article will implement an ARIMA model from scratch to create a Time Series Forecasting Analysis. We will be using the “pmdarima” library, a statistical library in Python that increases its time series analysis capabilities. We will be analysing and predicting the future temperatures from the dataset used. You can download the dataset from here to get started. The following implementation is also partially inspired by a video tutorial for time series forecasting, which can be accessed from the link here. Installing The Library First, we will install the pmdarima library, which will boost our analysis and help us create a defined forecasting model. To install the library, the following code can be implemented, #installing the library !pip install pmdarima Installing required Dependencies Next, we will install the required dependencies for our model; here we are using pandas and NumPy for analysis. #installing dependencies import pandas as pd import numpy as np Reading The Data We will now start with reading the data and creating a data frame for it. The present data has five columns, namely: MinTemp, MaxTemp, AvgTemp, Sunrise, Sunset. We will be focusing on the AvgTemp column and predicting future temperatures from it, and setting Date as our index. #reading the data df=pd.read_csv('\/content\/MaunaLoaDailyTemps.csv',index_col='DATE',parse_dates=True) #drpping null values df=df.dropna() print('Shape of data',df.shape) df.head() Output : Data Preprocessing and Initial Analysis Plotting the data to check how it is, #plotting the data df['AvgTemp'].plot(figsize=(12,5)) Running a statistical analysis test known as the dickey-fuller test to check if the data is stationary or not. We will judge based on the p-value received from the test. #applying dickey-fuller test from statsmodels.tsa.stattools import adfuller #creating a function for values def adf_test(dataset): dftest = adfuller(dataset, autolag = 'AIC') print(\"1. ADF : \",dftest[0]) print(\"2. P-Value : \", dftest[1]) print(\"3. Num Of Lags : \", dftest[2]) print(\"4. Num Of Observations Used For ADF Regression and Critical Values Calculation :\", dftest[3]) print(\"5. Critical Values :\") for key, val in dftest[4].items(): print(\"\\t\",key, \": \", val) #printing for AvgTemp adf_test(df['AvgTemp']) Output : 1. ADF :  -6.554680125068777 2. P-Value :  8.675937480199653e-09 3. Num Of Lags :  12 4. Num Of Observations Used For ADF Regression and Critical Values Calculation : 1808 5. Critical Values : 1% :  -3.433972018026501 5% :  -2.8631399192826676 10% :  -2.5676217442756872 With the observed p-value, we can state that the data is stationary. Creating Our Arima Model #creating our ARIMA Model from pmdarima import auto_arima # Ignore harmless warnings import warnings warnings.filterwarnings(\"ignore\") Calling our model and generating best possible ARIMA combination, #calling our function stepwise_fit = auto_arima(df['AvgTemp'],suppress_warnings=True) stepwise_fit.summary() Output : from statsmodels.tsa.arima_model import ARIMA Splitting the model into Train And Test, assigning the last 30% as the testing and the rest as the training data. #splitting into train and test print(df.shape) train=df.iloc[:-30] test=df.iloc[-30:] print(train.shape,test.shape) print(test.iloc[0],test.iloc[-1]) Output : (1821, 5) (1791, 5) (30, 5) MinTemp      36.0 MaxTemp      52.0 AvgTemp      44.0 Sunrise     640.0 Sunset     1743.0 Name: 2018-12-01 00:00:00, dtype: float64 MinTemp      39.0 MaxTemp      52.0 AvgTemp      46.0 Sunrise     656.0 Sunset     1754.0 Name: 2018-12-30 00:00:00, dtype: float64 Training the ARIMA Model, #model Training from statsmodels.tsa.arima_model import ARIMA model=ARIMA(train['AvgTemp'],order=(1,0,5)) model=model.fit() model.summary() Results : Making Predictions on Test Set Plotting the predictions, start=len(train) end=len(train)+len(test)-1 pred=model.predict(start=start,end=end,typ='levels').rename('ARIMA predictions') #pred.index=index_future_dates pred.plot(legend=True) test['AvgTemp'].plot(legend=True) #knowing the mean AvgTemp test['AvgTemp'].mean() 45.0 Our model here seems to predict the trend well from the data. Calculating the mean squared error to check how our model has performed. If the root means the squared error is close to the mean derived, it will be termed a bad model. from sklearn.metrics import mean_squared_error from math import sqrt rmse=sqrt(mean_squared_error(pred,test['AvgTemp'])) print(rmse) 3.000495429601031 Printing the last five values to see on what date the dataset has its end. #checking data end date model2=ARIMA(df['AvgTemp'],order=(1,0,5)) model2=model2.fit() df.tail() Making Future Predictions Printing future predictions for the next 30 days from the end date, #printing predictions for next 30 days index_future_dates=pd.date_range(start='2018-12-30',end='2019-01-29') #print(index_future_dates) pred=model2.predict(start=len(df),end=len(df)+30,typ='levels').rename('ARIMA Predictions') #print(comp_pred) pred.index=index_future_dates print(pred) Output : 2018-12-30    46.418064 2018-12-31    46.113783 2019-01-01    45.617772 2019-01-02    45.249555 2019-01-03    45.116984 2019-01-04    45.136771 2019-01-05    45.156280 2019-01-06    45.175516 2019-01-07    45.194482 2019-01-08    45.213183 2019-01-09    45.231622 2019-01-10    45.249802 2019-01-11    45.267728 2019-01-12    45.285403 2019-01-13    45.302830 2019-01-14    45.320012 2019-01-15    45.336955 2019-01-16    45.353659 2019-01-17    45.370130 2019-01-18    45.386370 2019-01-19    45.402383 2019-01-20    45.418171 2019-01-21    45.433738 2019-01-22    45.449087 2019-01-23    45.464221 2019-01-24    45.479143 2019-01-25    45.493855 2019-01-26    45.508362 2019-01-27    45.522665 2019-01-28    45.536769 2019-01-29    45.550674 Freq: D, Name: ARIMA Predictions, dtype: float64 Plotting Graph for future predictions, EndNotes This article has tried to explore an ARIMA model and how time series analysis can be taught with the model. We also discussed the different aspects of time series analysis and the necessary steps to create a complete-time series model. You can implement the same on different datasets and see how the complexity varies. The colab notebook for the above implementation can be found here. Happy Learning! References PMDArima LibraryAspects of ARIMA Modelling","excerpt":"Time series data is a set of observations collected through repeated measurements over time. Plotting the points on a graph, one of our axes would always be time. Time series data is everywhere since time is a constituent of everything observable. As the world advances every day with technology, sensors and systems constantly produce a […]","categories":["Deep Tech"],"tags":["ARIMA","exploratory data analysis","Guide","python database gui","python time series","stepwise regression machine learning","Time Series Forecasting","time series prediction"],"author_name":"Victor Dey","publish_date":"2021-07-14T15:00:00","publication_year":"2021","word_count":1545,"keywords":["Go","NumPy","ARIMA","TPU","AI","R","ML","Time Series Forecasting","python time series","RAG","Colab","Python","stepwise regression machine learning","time series prediction","exploratory data analysis","python database gui","Guide","Pandas"],"extracted_tech_keywords":["AI","ML","Colab","Pandas","NumPy","RAG","TPU","Python","R","Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/comprehensive-guide-to-time-series-analysis-using-arima\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10093592,"title":"AMD Takes On Both Intel and Apple: Will It Win?","content":"It’s the year 1995. AMD just won a groundbreaking settlement against Intel, the pioneer behind the x86 standard, that allows it to continue manufacturing chips using this method. Almost three decades later, AMD has not only leapfrogged Intel’s technology, but dominates Team Blue (Intel) in every segment of the semiconductor market. As Intel is plagued with more delays to its roadmap, AMD is belting out chips like hotcakes. From the Z1 series chips made for handheld gaming, to the 7000 series chips for desktops and laptops, to the EPYC series chips for the enterprise, Team Red (AMD) is handily beating Intel at every turn. On the other hand, both Intel and Apple — currently AMD’s biggest competitors — have moved to collaborating with ARM. This IP giant relies on a big.LITTLE design for their chips. This design uses smaller cores for less power-intensive tasks while reserving bigger, more power-hungry cores for powerful tasks, improving overall efficiency. AMD, on the other hand, is relying on a chiplet design to scale processor size and power. Instead of packing all of the chips’ features into a single big piece or ‘die’, AMD opted to split up the chip into separate, smaller parts known as ‘chiplets’ and connect them together using ‘Infinity Fabric’. With the launch of AMD’s Zen 5 on the horizon, a question arises — Will AMD go the ARM way, or double down on its winning strategy? Can chiplets keep winning? When AMD wanted to compete against Intel in the server chip market in 2015, they turned to chiplets as their last resort. This technology debuted with the launch of the first EPYC server chip, to great success. Moving to this process not only helped AMD to cut manufacturing costs by 40%, it also enabled the creation of chiplet-based desktop CPU. The Zen 2 lineup of chips brought chiplets to consumers, further optimising the manufacturing process. What’s more, the company could simply scale up the number of chiplets in the package to increase the number of cores or power of the chips, subverting the limitations put in place by Moore’s Law. Speaking of Moore’s Law, the 1965 article that spawned this rule also spoke about chiplets as the chip technology of the future. While it didn’t mention the technology by name, the article stated, “It may prove to be more economical to build large systems out of smaller functions, which are separately packaged and interconnected. The availability of large functions, combined with functional design and construction, should allow the manufacturer…to design and construct a considerable variety of equipment both rapidly and economically.” The economical and manufacturing advancements that have come with the last 5 decades of chip manufacturing have finally made the economy of manufacturing chiplets possible. The approach also allows for a mixed bag approach to manufacturing chips, provided a chiplet standard is in place. The industry also seems to think along the same lines, as seen by the introduction of the universal chip interconnect express standard. Supported by AMD, ARM, Google Cloud, Microsoft, Meta, and even Intel, this new chiplet ecosystem promises to go beyond Moore’s Law to make the computing systems of the future. ARM and Intel’s support of this standard shows that even AMD’s direct competitors see the benefits of chiplets design. In a way, it seems that they are admitting defeat to the superior chip design methodology. AMD also seems to be doubling down on chiplets for its next generation of CPUs, which are rumoured to provide up to an 18% performance increase while reducing power consumption by 34%. This improvement will be made thanks to a move from TSMC’s N5 manufacturing process to the N3 manufacturing nodes. However, for the time being, it seems that Intel is content to push its existing technology to the brink of obsolescence. big.LITTLE loses out Manufacturing a non-homogenous chip is no joke, which is something that Intel found out the hard way. Intel’s 12th Gen Alder Lake CPUs, which were the trial run for the efficiency cores and performance cores design (Intel’s version of big.LITTLE), were delivered without a hitch. However, it encountered problems when moving towards more complex architectures in the same design. Another of Intel’s line of chips called Sapphire Rapids, better known by the market name ‘4th Gen Xeon Scalable’ CPUs, were plagued with a host of delays. In May 2022, it was reported that Intel ran into a major technical flaw in the design of Sapphire Rapids, resulting in the delay of the chips to the beginning of 2023. However, this was already too late for Intel, as AMD handily beat the company with the 96-core EPYC Genoa CPU. The competing offering beat Intel’s chips by over 70%, using chiplets to gain the edge. To add salt to the wound, Intel’s next lineup of chips, codenamed Meteor Lake, has also been hit with delays due to production issues. To gain back some of the ground Intel lost with the botched launch of Sapphire Rapids, it entered into a partnership with ARM. This partnership mainly focused on the manufacturing of mobile chips in the form of SoC or System on Chip approach. Where Intel’s Foundry arm got a deeper insight into ARM’s intellectual property, ARM got access to Intel’s 18A manufacturing process. Interestingly, another emerging player in the ARM-based processor game, Apple, has also been facing issues. While the chips are undeniably well-suited for their deployed applications, especially in terms of power efficiency, it seems that this is not enough to carry the MacBook’s sales. Reportedly, the next generation of M-series chips will include even more cores in an attempt to entice customers to buy them. However, it seems that the big.LITTLE architecture is beginning to show its age. As Moore’s Law begins to reach the point of obscurity, older manufacturing designs are beginning to see diminishing returns when it comes to performance. Chiplets are the way of the future, and it even seems like Intel and ARM believe this. With its commitment to the chiplet design, it seems that AMD is uniquely positioned to make the most out of the next decade of chips.","excerpt":"AMD’s chiplet design is taking on Intel’s heterogeneous chip design, but Team Red seems poised to win.","categories":["Global Tech"],"tags":[],"author_name":"Anirudh VK","publish_date":"2023-05-18T18:29:11","publication_year":"2023","word_count":1019,"keywords":["Go","Rapids","API","ELT","programming_languages:R","AI","Scala","cloud_platforms:Google Cloud","programming_languages:Scala","R"],"extracted_tech_keywords":["AI","Rapids","R","Go","Scala","API","ELT","cloud_platforms:Google Cloud","programming_languages:R","programming_languages:Scala"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/amd-takes-on-both-intel-and-apple-will-it-win\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":65581,"title":"Glass Quality Prediction: Weekend Hackathon #6","content":"Another weekend, another exciting hackathon. MachineHack is back with the 6th instalment of the popular weekend hackathon series, and this time we challenge data scientists to predict the grade of glass manufactured in MachineHack’s Glass Quality Prediction: Weekend Hackathon #6. The challenge will start on May 22nd Friday at 6 pm IST. Click here to participate Problem Statement & Description We humans have been using glass since ancient times for a variety of applications from building construction to making decorative objects. With technology, glass and its applications have evolved, and today, we have different varieties of glass used for very different purposes from a computer monitor to a bulletproof car window depending on the grade of the glass produced. And not all grades or varieties are manufactured the same way. In this data science challenge, you as a data scientist must use the given data to predict the grade of the glass produced based on the given factors. Given are 15 distinguishing factors that can provide insight into what grade of the glass is being produced. Your objective as a data scientist is to build a machine learning model that can predict the grade of glass based on the given factors. Data Description The unzipped folder will have the following files. Train.csv – 1358 observations.Test.csv – 583 observations.Sample Submission – Sample format for the submission. Target Variable: class The datasets will be made available for download on May 22nd, Friday at 6 pm IST Below are the file formats for the provided data Train.csv Test.csv Sample_Submission.xlsx The submission must contain the probabilities of the target classes 1 and 2. Click here to participate Bounties The top 3 competitors will receive a cool AIM goodie bag and a free pass to the plugin. plugin, India’s largest virtual conference on AI, is a next-gen disruptive conference that brings AI professionals from around the world together in a virtual setting. Know more about the plugin here. Rules One account per participant. Submissions from multiple accounts will lead to disqualificationThe submission limit for the hackathon is 10 per day after which the submission will not be evaluatedAll registered participants are eligible to participate in the hackathonThis competition counts towards your overall ranking pointsYou will not be able to submit once you click the “Complete Hackathon” button. You may ignore this featureWe ask that you respect the spirit of the competition and do not cheatThis hackathon will expire on 25th May, Monday at 7 am IST Evaluation The leaderboard is evaluated using log_loss for the participant’s submission. Click here to participate","excerpt":"Another weekend, another exciting hackathon. MachineHack is back with the 6th instalment of the popular weekend hackathon series, and this time we challenge data scientists to predict the grade of glass manufactured in MachineHack’s Glass Quality Prediction: Weekend Hackathon #6. The challenge will start on May 22nd Friday at 6 pm IST. Problem Statement & […]","categories":["Deep Tech"],"tags":["Hackathon","Machinehack","Weekend Hackathon"],"author_name":"Amal Nair","publish_date":"2020-05-21T14:00:55","publication_year":"2020","word_count":426,"keywords":["data science","Go","machine learning","Weekend Hackathon","programming_languages:R","AI","Machinehack","programming_languages:Go","Hackathon","Aim","R"],"extracted_tech_keywords":["AI","machine learning","data science","Aim","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/glass-quality-prediction-hackathon\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10125197,"title":"Tech Mahindra Finally Launches Project Indus, Indic LLM with 37+ Hindi Dialects","content":"Tech Mahindra today announced the much-awaited launch of its Project Indus. This indigenous large language model (LLM) is designed to engage in conversations across a wide array of Indic languages and dialects, marking a significant step towards linguistic inclusivity in the realm of artificial intelligence. In the initial phase, Project Indus will concentrate on developing an LLM specifically tailored for the Hindi language and its extensive range of over 37 dialects. By focusing on this widely spoken language, Tech Mahindra aims to bridge the linguistic gap and empower a substantial portion of the population to interact with cutting-edge AI technology in their native tongue. To bring Project Indus to life, Tech Mahindra has forged strategic collaborations with industry leaders Dell Technologies and Intel. The model will be implemented using ‘GenAI in a Box’ solution, which harnesses the power of Dell Technologies’ state-of-the-art computing, storage, and networking capabilities. Additionally, the solution incorporates Intel-based infrastructure, including Intel® Xeon® Processors and the OneAPI software suite, ensuring optimal performance and scalability. Nikhil Malhotra, Global Head of Makers Lab at Tech Mahindra, expressed his enthusiasm for the project, stating, “Project Indus is our seminal effort to develop an LLM from the ground up. Through Makers Lab, our R&D arm, we created a roadmap, collected data from the Hindi-speaking population, and built the Indus model. Our collaboration with Dell Technologies & Intel will help deliver cutting-edge AI solutions that enable enterprises to scale at speed.” The collaboration between Tech Mahindra, Dell Technologies, and Intel aims to revolutionise AI-driven solutions across various industries. By leveraging Tech Mahindra’s expertise in developing localized and verticalized industry-agnostic LLMs, the partnership will enable the creation of tailored use cases and applications in sectors such as healthcare, rural education, banking, agriculture, and telecom. This initiative has the potential to drive innovation, optimise productivity, and promote growth in these critical domains. Denise Millard, Chief Partner Officer at Dell Technologies, emphasized the significance of accessibility and scalability in the adoption of GenAI. “With the Dell AI Factory, LLMs like Project Indus leverage AI-optimized technologies with an open ecosystem of partners, validated and integrated solutions, services and best practices, accelerating the adoption of AI to drive growth, optimize productivity and promote innovation,” she remarked. Santhosh Viswanathan, Vice President & Managing Director for India at Intel, highlighted the company’s dedication to advancing the frontiers of AI. “We are proud to collaborate with Tech Mahindra on Project Indus, which will enable seamless deployment of advanced AI models across industries and empower enterprises to unlock GenAI’s full potential for enhanced operational efficiency and a competitive edge,” he stated. Project Indus represents a pivotal milestone in the evolution of the global GenAI landscape, which is projected to expand to a staggering $1.3 trillion over the next decade. The model will initially prioritise key use cases and pilot projects, offering scalable AI solutions to enterprises. This collaboration underscores Tech Mahindra’s unwavering commitment to enabling enterprises to scale rapidly with technological advancements, fostering a future where AI solutions are accessible, scalable, and responsible. Tech Mahindra’s recent announcement of developing an LLM to preserve Bahasa Indonesia and its dialects further highlights the company’s dedication to linguistic diversity and inclusivity in the AI domain. With Project Indus, Tech Mahindra is poised to reshape the AI landscape, empowering businesses and individuals alike to harness the potential of advanced language models in their native languages.","excerpt":"To bring Project Indus to life, Tech Mahindra has forged strategic partnership with Dell Technologies and Intel.","categories":["AI News"],"tags":["project indus","Tech Mahindra"],"author_name":"Shyam Nandan Upadhyay","publish_date":"2024-06-28T12:54:55","publication_year":"2024","word_count":560,"keywords":["Tech Mahindra","artificial intelligence","GenAI","AI","ML","Scala","RAG","Ray","Aim","edge AI","R","project indus"],"extracted_tech_keywords":["AI","artificial intelligence","ML","GenAI","Aim","Ray","edge AI","RAG","R","Scala"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/tech-mahindra-finally-launches-project-indus-indic-llm-with-37-hindi-dialects\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10040346,"title":"Why DeepMind Researchers Are Betting On Neural Algorithmic Reasoning","content":"Trying to predict algorithmic inputs from data leads to a bottleneck, thanks to the uncertainties of the real world. Real-time data can be noisy. This noisy raw data has to be converted into scalar values (read: weights of neural network). The algorithm then uses these values to estimate output. In low data set-ups, such techniques go haywire. Figuring out the right scalars can be challenging. So how do you escape this algorithmic bottleneck? According to researchers at DeepMind, neural networks should consistently produce high-dimensional representations to beat the bottleneck. This means the computations of the algorithms must be made to operate over high-dimensional spaces. “The most straightforward way to achieve this is replacing the algorithm itself with a neural network—one which mimics the algorithm’s operations in this latent space, such that the desirable outputs are decodable from those latents,” explained the researchers. The researchers underline the recent resurge in algorithmic reasoning, which offers ways in which algorithmically-inspired neural networks can be built, primarily through learning to execute the algorithm from abstractified inputs. Algorithmic reasoning, added researchers, provides methods to train useful processor networks, such that within their parameters, a combinatorial algorithm can be found that is not vulnerable to bottleneck phenomena. Such an algorithm would feature the following: (a) Aligns with the computations of the target algorithm; (b) Operates by matrix multiplications, hence natively admits useful gradients (c) Operates over high-dimensional latent spaces. (Source: DeepMind) Research In a paper titled “Neural Algorithmic Reasoning”, the researchers at DeepMind laid out a blueprint of neural algorithmic reasoning (NAR). To demonstrate NAR, they assumed that the real-world problem requires learning a mapping from natural inputs, x, to natural outputs, y. For example, in the case of a real-time traffic scenario, the natural inputs are often likely high-dimensional, noisy, and prone to changing rapidly. “Further, we assume that solving our problem would benefit from applying an algorithm, A—however, A only operates over abstract inputs, x ̄,” they added. In this work, the researchers demonstrated an elegant neural end-to-end pipeline that goes straight from raw inputs to general outputs, while emulating an algorithm internally. The general procedure for applying an algorithm A (which admits abstract inputs x ̄) to raw inputs x. According to the researchers at DeepMind, the blueprint for NAR looks like this: First, learn an algorithmic reasoner for A, by learning to execute it on synthetically generated inputs, x ̄. This yields functions f, P, g such that g(P (f (x ̄))) ≈ A(x ̄). f and g are encoder\/decoder functions, designed to carry data to and from the latent space of P (the processor network). Set up appropriate encoder and decoder neural networks, f ̃ and g ̃, to process raw data and produce desirable outputs. The encoder should produce embeddings that correspond to the input dimension of P, while the decoder should operate over input embeddings that correspond to the output dimension of P. Swap out f and g for f ̃ and g ̃, and learn their parameters by gradient ̃descent on any differentiable loss function that compares g ̃(P(f(x))) to ground-truth outputs,y. The parameters of P should be kept frozen. Through this pipeline, neural algorithmic reasoning offers a strong approach to ̃ applying algorithms on natural inputs. The raw encoder function, f, has the potential to replace the human feature engineer, as it is learning how to map raw inputs onto the algorithmic input space for P , purely by backpropagation. The researchers believe the algorithmic reasoner, as described in their work, can be bootstrapped from existing algorithms in supervision of their internal workings, and then subsequently embedded into the real world input\/outputs via separately trained encoding\/decoding networks. This technique especially has come to fruition in domains such as reinforcement learning and genome assembly. And, neural algorithmic reasoning fits well into other machine learning challenges as it allows for the application of classical algorithms on inputs and, in a way, uniting the theoretical advantages with that of a designer’s intent.","excerpt":"Trying to predict algorithmic inputs from data leads to a bottleneck, thanks to the uncertainties of the real world. Real-time data can be noisy. This noisy raw data has to be converted into scalar values (read: weights of neural network). The algorithm then uses these values to estimate output. In low data set-ups, such techniques […]","categories":["Deep Tech"],"tags":[],"author_name":"Ram Sagar","publish_date":"2021-05-19T16:00:00","publication_year":"2021","word_count":662,"keywords":["Go","API","machine learning","TPU","programming_languages:R","AI","neural network","Scala","GAN","R"],"extracted_tech_keywords":["AI","machine learning","neural network","TPU","R","Go","Scala","API","GAN","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/why-deepmind-researchers-are-betting-on-neural-algorithmic-reasoning\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10002304,"title":"Bangalore’s Drone Startups Are The First To Be Certified Under The New Drone Policy","content":"Regulations in the Indian drone industry have been a point of contention for several years now. Despite the country is home to many promising drone startups, the startups are unable to work in a full-fledged manner in the country because India lacked policies and regulations regarding the same. December 2018 proved to be a turning point when the BJP-led Indian government and the Ministry of Civil Aviation announced drone regulations. Now, two Bengaluru-based startups got to be one of the first drone startups in India to have received a license to fly their drones. Drone Industry In India India has more than 70 drone startups. However, this industry is still considered an emerging technology in which the funding is not consistent with innovation. In fact, according to a recent study, the total funding raised by drone startups in India from 2014 to 2018 was just $16.56 million. This amount covers only 2.26% of the total deep tech funding, which was $732 million. India Certifies Its First Two Drones For the first time since the new drone policy was passed in December last year, two drone startups got certified by the Directorate General of Civil Aviation (DGCA). The startups are Skylark Drones and Throttle Aerospace, both of which are based out of Bengaluru. Both the drone startups fall under the category of visual line of sight (VLOS) which demands that the drones be in the view limits of the operator. This is the first time that has been given to drone startups, under the new drone policy called the Digital Sky. The motive of this drone policy is to facilitate the use of drones for commercial purposes and to provide checks to the companies that make use of drone applications. About Certified Drones Skylark Drones: Started in the year 2014, Skylark Drones has web-based analytical platforms to transform survey-grade drone maps and digital elevation models to be used for detailed project reports, feasibility studies and project planning. They help in detailed project reports, feasibility studies and project planning through their web-based and help companies understand defects with information layers such as high-resolution drone videography, thermal analysis and other asset lifecycle modelling tools using their analytical platforms. They are also experts in drone videography and photography. According to Mrinal Pai the Co-Founder of Skylark Drones, the certification will allow the startup to become a provider of the NP-NT module. The startup is already in talks with other drone manufacturers in China and the UK to use their solution to get drones partially certified. Throttle Aerospace: Started in the same year as Skylark Drones, Throttle Aerospace was formed by a team of young entrepreneurs with over 15 years of knowledge in engineering, manufacturing & supply chain carried from global Aerospace and Defence industries. They provide solutions from design to manufacturing. The startup is already working with clients such as the Ministry of Defence, HAL and Wipro. India’s Drone Policy In December 2018, India finally released its first drone policy. The main objective of this policy was to regulate the flying of drones by civilians. Drone policy had become very important because civilians were using it in an unregulated manner, and were causing accidents. The new drone policy by the Government has five categories of drones based on their weights and each category has a different set of rules. Following are the categories with their weights that fall under the drone policy: 1.Nano: Less than or equal to 250 grams. This category does not need any license to fly and is not allowed to fly beyond 50 feet. 2.Micro: Greater than 250 grams and less than or equal to 2 kilograms. 3.Small: Greater than 2 kilograms and less than or equal to 25 kilograms. 4.Medium: Greater than 25 kilograms and less than or equal to 150 kilograms. 5.Large: Greater than 150 kilograms. According to the policy, the person who flies the drone has to be at least 18 years old and should have at least studied till Class X. The policy states that all the drones must be within the sight of the pilot. Additionally, the drones will only be allowed to fly during the day time. It also does not allow them to be operated from a moving vehicle. According to the policy, drones cannot fly in certain areas. Some of them are the following: 1.25 kilometres from International Border, which includes Line of Control (LoC), Line of Actual Control (LAC) and Actual Ground Position Line (AGPL) 2.5 kilometres from the perimeter of airports in Mumbai, Delhi, Chennai, Kolkata, Bengaluru and Hyderabad and from the radius of Vijay Chowk in Delhi. 3.3 kilometres from the perimeter of any military installations, military facilities, any defence, civil and private airport and the radius of State Secretariat Complex in State Capitals. 4.Over eco-sensitive zones around national parks and wildlife Sanctuaries notified by Ministry of Environment, Forests and Climate Change without prior permission. 5.Beyond 500 meters into the sea from a coastline. Licence To Fly Drones The drones made in India and the ones imported from outside the country have a difference in licences. These licences are issued by the DoT. The licences required to fly any drone other than the nano drones in India can be of the following two types. These licences are not required when one wants to fly a drone below 200 feet, but he has to inform the police before flying: 1.Unique Identification Number (UIN): Drones are required to have a unique identification number (UIN). The fee for a fresh UIN is ₹1,000. 2.Unmanned Aircraft Operator Permit (UAOP): Drone operators are required to obtain an UAOP. The fee for a fresh UAOP is ₹25,000. It is valid for 5 years. How To Apply For micro drones, the registration is done at the website of Directorate General of Civil Aviation. In order to get the UID to get the drone registered, the required documents include the owner’s address proof. The person also needs to take permission from the Police and the Telecom department. The procedure to apply to the drone license is different for RPAs (Remotely Piloted Aircraft) imported to India and the ones made in India. For RPA imported to India the procedure is: 1.Any entity intending to import RPAS in India shall obtain Equipment Type Approval (ETA) from WPC Wing, Department of Telecommunication for operating in de-licensed frequency band. 2.Apply to DGCA for import clearance as per format given in Annexure-IA. 3.Apply to DGCA for UIN\/UOAP. For RPA locally purchased in India: 1.The applicant shall ensure that locally purchased RPAS have ETA from WPC Wing, DoT operating in de-licensed frequency band. 2.The applicant shall submit information as per format given in Annexure-IB along with the application for issue of UIN\/UAOP, as applicable. The application form is available with DGCA and the draft of Drone Policy. What The Future Holds The regulations being made for the drone startups in India is definitely a good way to check up on their applications being safe. The demand for this technology has risen, and we have many startups that have to potential to cater to the demand. However, the drone policy that we currently have also had a number of drawbacks. It lacks the guidelines to scale up operations and avoid the misuse of drones. Nano drones do not have strict rules and there are no laws to keep an eye on they being misused. Although the micro category drones demand to inform the police before flying them, there is no centralised way to monitor that this is being followed. The current policy also does not take into account technologies like AI and the negative power that it might have when used inappropriately. It does not account for how the data collected by the drones is shared. A stricter set of policies would benefit the country and help in critical situations like disaster management.","excerpt":"Regulations in the Indian drone industry have been a point of contention for several years now. Despite the country is home to many promising drone startups, the startups are unable to work in a full-fledged manner in the country because India lacked policies and regulations regarding the same.  December 2018 proved to be a turning […]","categories":["AI Startups"],"tags":["drones","Government","policy","Startups"],"author_name":"Disha Misal","publish_date":"2019-06-26T16:57:31","publication_year":"2019","word_count":1309,"keywords":["Go","API","funding","drones","AI","AWS","RPA","innovation","Government","Git","Startups","R","policy","startup"],"extracted_tech_keywords":["AI","AWS","R","Go","Git","API","RPA","innovation","startup","funding"],"url":"https:\/\/analyticsindiamag.com\/ai-startups\/bangalores-drone-startups-are-the-first-to-be-certified-under-the-new-drone-policy\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10001930,"title":"10 Cool Smart Home Gadgets You Can Buy This Year","content":"Here are some of the top gadgets that are IoT-powered and easily available in India to make our lives easier. 1. Smart Doorbell Smart doorbells are connected to WiFi and they connect to the existing doorbell wiring to get power and connect to the existing doorbell chimes. As soon as the button is pressed, the chimes ring and the doorbell’s WiFi radio sends an alert to the phones connected. They have a camera that gives the provision of a live video and a motion detection, even in the dark. These doorbells get power from the existing doorbell or from the internal batteries. 2. Bluetooth enabled bulb A bulb that is controlled by a smartphone and is connected to the bluetooth. Some also work on Wi-Fi, Z-Wave, or ZigBee to connect to an application because they use wireless communication standards. Many brands use more than one wireless technology to better integrate the lights with other smart home devices. There are also products with apps on the phone allowing to sync the bulb to any music that you play wherein the bulb will change its lighting as per the song’s rhythm. There are services with provisions of sending messages on platforms of Facebook, Google plus, Twitter, and so on. There are also options available for the customization of the bulb to notify of any weather alerts. 3. Smart weight tracker Weighing machines connected to the bluetooth or WiFi help track the weight,  lean mass, body fat percentage and BMI. A product called Fitbit Aria is synced to your Fitbit dashboard, displaying charts and graphs of your progress. This gives a much more realistic picture of your fitness over a long period of time. 4. Connected car IoT device Cars can be connected to the cloud, allowing the driver to make decisions based on real time data. Data from the car such as its engine performance, location details and safety devices is taken and put to the cloud. It gives alerts for things like rash driving, geofence, engine errors, vehicle towing. It also provides with reminders such as document wallet and servicing history and generate custom reports and trips. 5. Footwear About 90% of the world’s visually impaired live in developing countries, according to the World Health Organisation. Around 20% of this population lives  in India alone. A Telangana-based startup called Ducere Technologies has developed a footwear helping the visually impaired in navigating the world around them. The shoe hooks up with an app that syncs with Google Maps, tracks your steps and counts your calories burned. The shoe itself can be used for jogging around town. The shows are called Lechal and it also has a bluetooth enabled shoe insert that hooks up with Google Maps and buzzes to let you know which way to turn on your chosen route. 6. Smart Air Sensor IoT connected devices to sense the temperature, humidity and air pressure in the house for personal weather data in the house. Apple has a product called Elgato Eve the works on these home automation products and works with the Apple HomeKit to allow users to monitor indoor air, outdoor weather, energy consumption and whether windows and doors are open or closed. 7. Smart TV A smart TV offers a number of services that normal televisions can’t offer. These televisions offer apps, media streaming, web browsing, games and, perhaps most importantly, Internet Protocol Television (IPTV). It can take the form of short clips or continuous live channels. LG divides its SmartThinQ line of connected appliances into categories for the kitchen, living  and safety. They all integrate with the company’s smartphone app. 8. Smart Laundry An IoT-enabled smart laundry system can be controlled completely through applications on the phone. Users can reserve, schedule, monitor and pay for the services from their phones. This provides an end-to-end digitally managed washing experience. Each washing machine is connected to the cloud and can be individually managed from the app. It can also remind the users to take their clothes out once the washing is over and will maintain a complete record of their laundry usage in the past. Users can reserve, schedule, monitor and pay for the services from their phones. This provides an end-to-end digitally managed washing experience. 9. Smart Refrigerator Samsung launched an IoT-enabled refrigerator through which users will be able to use the view inside camera to label their food with expiration dates or use their smartphone to take a peek inside their refrigerator. It has a 21-inch touchscreen display embedded on the fridge. It can sync up food storage, keeping family members better connected and organised and providing enhanced entertainment. It also allows control and monitoring of other connected appliances like the washing machine and smartphone from the refrigerator’s screen. 10. Smart Bike Locks Many IoT-enabled locks track a bike’s location and send alerts if it leaves a designated area. Many of these smart lock devices offer additional features like ride analytics and crash alerts. They are installed in handlebars so they are invisible to thieves. They keyless entry is possible because of bluetooth connectivity.","excerpt":"Here are some of the top gadgets that are IoT-powered and easily available in India to make our lives easier. 1. Smart Doorbell Smart doorbells are connected to WiFi and they connect to the existing doorbell wiring to get power and connect to the existing doorbell chimes. As soon as the button is pressed, the […]","categories":["AI Trends"],"tags":["IoT"],"author_name":"Disha Misal","publish_date":"2019-04-29T15:14:16","publication_year":"2019","word_count":846,"keywords":["Go","AI","Git","RAG","automation","ViT","analytics","CLIP","GAN","R","IoT"],"extracted_tech_keywords":["AI","analytics","RAG","R","Go","Git","CLIP","GAN","ViT","automation"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/10-cool-smart-home-gadgets-you-can-buy-this-year\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10120500,"title":"Can GenAI Solve Legal Troubles in India?","content":"India’s legal system is neck-deep in crisis, with nearly 5 crore pending cases across the country’s courts. The sheer volume of cases has led to delays, inefficiencies, and a lack of access to justice for many citizens. However, a new wave of legal tech startups is here, aiming to tackle these challenges head-on by leveraging the power of generative artificial intelligence (GenAI). And with 762 legal tech startups in India, the potential for transformation is immense. One such startup is CourtEasy, founded by a team of engineers and lawyers who experienced legal problems firsthand. “We started CourtEasy to address the problems faced in the legal sector,” Mrunmayee Shende, one of the founders of CourtEasy, told AIM. “The system is overloaded with pending cases, and we believe AI will help solve it at a faster pace,” she added. What legal tech aims to solve The team at CourtEasy recognised the potential of GenAI to bridge the gap between technology and the legal system, particularly in lower courts where the adoption has been slow. “In lower courts, people are not aware of the potential of technology yet,” Shende explained. “Higher courts and law firms are more open to adopting newer technologies as they evolve, but this is not the case in lower courts.” With the aim to fix this gap, she said, “If we plot the adoption of technology in the legal system, we can see a significant gap between lower and higher courts. We want to bridge this gap by making sure our services are available to all.” The company has developed a range of tools powered by GenAI. “We have our Marathi language model called Marathi LLM, which is a foundational model using LLM for the Marathi language,” Shende explained. “Similarly, we are in the stage of building a separate legal model that can provide answers and solve the problems faced by legal professionals,” she added. The team has built language models for regional courts, legal research aids, and case drafting assistants. Another startup working towards making justice faster and fairer is Jhana AI. “At Jhana, we build intelligent and fundamental tools for legal services, a one-stop ecosystem, and you interact with this humanlike chatbot that is reflective and iterative,” said Em McGlone, co-founder of Jan. McGlone, explaining the versatility of the chatbot, said, “It can do anything from browsing and reading the web like a human to referencing legal books, journals, statutes, court orders from various courts, and various legal reportage, news blogs, and so on.” SpotDraft, another legal tech startup, uses GenAI to help law firms and legal teams draft, store, analyse, execute, and automate contracts and contract processes. The company has already processed around 4 million contracts and has clinched clients both locally and globally. Shashank Bijapur, the CEO and co-founder of Spotdraft, describes himself as a ‘recovering lawyer’ who has experienced the drudgery of lengthy documents firsthand. “At some point, I stopped using my brain altogether. I was just using my eyes and fingers to copy and paste [legal clauses] from one sheet to another,” he said, stressing on the need for automation in legal processes. Global success and future prospects On one side there’s the Supreme Court launching AI-powered tools like SUPACE (Supreme Court Portal for Assistance in Court’s Efficiency) and SUVAS (Supreme Court Vidhik Anuvaad Software). While on the other, we have the Madras High Court’s impressive case clearance rates through tech interventions. These demonstrate how AI can enhance efficiency and access to justice across all levels of the judiciary. The successful implementation of GenAI in legal systems around the world has already shown promising results. For example, in the United States, AI-powered tools have been used to predict the outcomes of cases with high accuracy, while in China, AI judges have been deployed to handle minor legal disputes. However, the adoption of GenAI in the legal system is not without challenges. Privacy and data security are major concerns, as legal cases often involve sensitive personal information. Shende acknowledged the importance of data safety, “We are a startup, and we have received funds and infrastructure credits from companies like Nvidia and Microsoft to ensure data security.” Despite the challenges, legal tech startups remain optimistic about the future of GenAI in the industry. CourtEasy has partnered in the Microsoft for Startups initiative and is now in the process of raising additional funds to expand its operations. Jhana is also looking for enterprise solutions for their product and is excited to meet people from the community who are building various tools for big corporations. The future of tech integration into India’s legal system looks promising. As McLone puts it, “We think we can make justice faster and fair by creating better retrieval systems for lawyers to use.”","excerpt":"Startups like CourtEasy are harnessing the power of GenAI to streamline legal processes and improve access to justice.","categories":["AI Features"],"tags":["legal","legal AI","Startups"],"author_name":"K L Krithika","publish_date":"2024-05-15T13:23:59","publication_year":"2024","word_count":791,"keywords":["GenAI","artificial intelligence","programming_languages:R","AI","RAG","legal","automation","Aim","Startups","R","legal AI","startup"],"extracted_tech_keywords":["AI","artificial intelligence","GenAI","Aim","RAG","R","automation","startup","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/can-genai-solve-legal-troubles-in-india\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10052447,"title":"Companies’ Growing Dependency On AWS: Lessons To Learn","content":"In 2020, Amazon Web Services suffered a major effect on several websites and apps and took down a large portion of the internet. As per media reports, Amazon noted that the issues affected one of its 23 geographic AWS regions. However, this outage brought a chunk of the internet to its knees. This is not the first time such an incident has happened. In 2017, AWS was affected by a four-hour outage. Several servers that were part of AWS S3 went down, taking many websites along with it. This happened due to a small typo but managed to cause a temporary dent in the internet-verse. Many organisations have openly spoken about their heavy dependence on the public cloud and AWS in particular. Even minor outages in these services threaten to disrupt the entire internet ecosystem, presenting a very threatening situation. Creme de la Creme among AWS’ clientele Streaming website darling, Netflix’s fortunes have quintupled in the last five years, and its customer base has increased to a whopping 214 million users. For a company scaling up at such speed, Netflix has revealed that it depends heavily on AWS, admitting that it runs a ‘vast majority of its computing on AWS and uses it for services like data processing, storage, and others. AWS has allowed Netflix to build a huge empire of video based on low-cost resources. Top en AWS users based on EC2 monthly spend (data from  Intricately) However, on the flip side, any outage or even a minor breakdown in AWS could send Netflix into a spin. In its filing in 2021 with the US Securities and Exchange Commission (SEC), Netflix said that any disruption or interference with the use of AWS could severely impact the company’s operations, thereby causing some major damage to its business. Not just Netflix, AWS hosts the digital drug-design tools that eventually led to developing Moderna’s COVID-19 vaccine. Moderna’s Drug Design Studio sits in the AWS Cloud, allowing scientists to access public and private libraries containing all the messenger RNA and thousands of proteins they produce. It has helped the company improve sequence design and production processes, thereby improving the way scientists gather feedback. In 2019, Amazon entered into a partnership with America’s professional Football team, Seahawk. The team’s management moved away from its previous vendor Microsoft Azure to sign a five-year contract with AWS. AWS’ other popular client is The Seattle Times, which has a readership of 7 million and serves in areas like Seattle, Washington, and the Pacific Northwest. The website has been using AWS since 2015. Overdependence? Public clouds offer a host of global benefits for companies, including flexibility, accessibility, savings, and scalability. Four major public cloud vendors dominate the market currently, namely AWS, Microsoft Azure, Google Cloud, and Alibaba. While the latter three clouds collectively account for 63 per cent of the market, AWS alone commands 31 per cent of the global cloud infrastructure. AWS was launched in 2006 and in a way heralded the beginning of the cloud computing era. A 2006 New York Times article said that the internet was entering its ‘Lego era’ and fundamentally changing the economics of opening a computer-based business. Earlier, businesses would spend 70 per cent of their engineering time and resources on building data centres and maintaining basic infrastructure software. Only 30 per cent of the effort went towards developing new products. AWS flipped that ratio. Because of AWS, app-based startups could be developed. The importance of AWS towards its parent organisation Amazon, its consumers, and the global internet ecosystem cannot be overstated. For Amazon, AWS is a cash cow that made up 59 per cent of the company’s $22.9 billion profit (before interest and tax) in 2020. Interestingly, AWS accounts for just 12 per cent of Amazon’s revenue. On a more general term, AWS has revolutionised the economics of web-based businesses by creating a $300 billion cloud computing industry. From large companies to government agencies and small startups, all use some form of cloud computing. AWS stronghold of the market has not gone unnoticed by market watchers. The US’ House Judiciary Antitrust subcommittee submitted a report in October 2020 on tech monopolies. The report noted that Amazon had grown to become a dominant force in the online retail market and has monopoly power over third-party sellers on its marketplace. Reporting on AWS’ position, the subcommittee found that AWS sought to build its leadership position and kept existing customers from going to other vendors. Another major reason for alleged vendor lock-in is that AWS is built on open-source projects like Elasticsearch and MongoDB that compete with companies that built these software projects in the first place. “If they ever wanted to switch to another provider, they would have to extensively re-engineer their product in another software, whereas, had they built their application using MongoDB — on AWS or any other cloud provider’s infrastructure — their applications could move to other platforms,” the report said. Amazon has disputed the lawmakers’ findings. All said one cannot take from the fact that AWS penetrates our daily lives so much that we would more than just notice if and when it stops working.","excerpt":"AWS penetrates our daily lives so much that we would more than just notice if and when it stops working.","categories":["IT Services"],"tags":["antitrust"],"author_name":"Shraddha Goled","publish_date":"2021-10-27T14:00:00","publication_year":"2021","word_count":857,"keywords":["Elasticsearch","Go","AWS","cloud computing","AI","MongoDB","R","RAG","antitrust","Rust","Azure"],"extracted_tech_keywords":["AI","RAG","cloud computing","AWS","Azure","Elasticsearch","MongoDB","R","Go","Rust"],"url":"https:\/\/analyticsindiamag.com\/it-services\/companies-growing-dependency-on-aws-lessons-to-learn\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":69694,"title":"ACL 2020 Announces Its Best NLP Research Papers","content":"The 58th Association for Computational Linguistics (ACL) Conference announced its best research papers on computational linguistics. Besides the best paper award, the conference also announced other awards which include the honourable mention papers, best theme paper and best demonstration paper. Computational linguistics is the study of language from a computational perspective where the computational linguists are interested in providing knowledge-based or data-driven models of various kinds of linguistic phenomena. Computational linguistics researchers work on a number of natural language processing projects such as speech recognition systems, text-to-speech synthesizers, automated voice response systems, text editors, among others. Below here we have listed the best research papers on computational linguistics from ACL 2020 Conference: – Best Paper: Beyond Accuracy: Behavioral Testing of NLP Models with CheckList About: A team of researchers from Microsoft Research, the University of Washington and the University of California, Irvine introduced a model-agnostic and task agnostic methodology for testing NLP models known as CheckList. It includes a matrix of general linguistic capabilities and test types that facilitate comprehensive test ideation, as well as a software tool to generate a large and diverse number of test cases quickly. CheckList tests individual capabilities of the NLP model using three different test types. Further, CheckList reveals critical bugs in commercial systems developed by large software companies, indicating that it complements current practices well. The implementation of this methodology is also available on GitHub. Read the paper here. Honourable Mention Papers – Main Conference: Don’t Stop Pretraining: Adapt Language Models to Domains and Tasks About: In this research paper, the researchers investigated whether it is still helpful to tailor a pre-trained model to the domain of a target task. They presented a study across four domains, which are biomedical and computer science publications, news, and reviews and eight classification tasks, showing that the second phase of pretraining in-domain (domain-adaptive pretraining) leads to performance gains, under both high- and low-resource settings. Read the paper here. Tangled up in BLEU: Reevaluating the Evaluation of Automatic Machine Translation Evaluation Metrics About: In this paper, the researchers add to the case to stop using BLEU as the de facto standard metric when evaluating MT systems. Instead, one must use other metrics such as CHRF, YISI-1, or ESIM, as they are more powerful in assessing empirical improvements. Read the paper here. Best Theme Paper: Climbing towards NLU: On Meaning, Form, and Understanding in the Age of Data About: In this paper, the researchers made an argument that the language modelling task cannot, in principle lead to learning of meaning because it only uses the form as training data. They stated, “Our aim is to advocate for an alignment of claims and methodology: Human-analogous natural language understanding (NLU) is a grand challenge of artificial intelligence, which involves mastery of the structure and use of language and the ability to ground it in the world.” Read the paper here. Honourable Mention Paper – Theme: How Can We Accelerate Progress Towards Human-like Linguistic Generalization? About: This research paper describes the Pretraining-Agnostic Identically Distributed (PAID) evaluation paradigm that has become a crucial tool for estimating progress in NLU. The paradigm consists of three stages, which are pretraining of a word prediction model on a corpus of arbitrary size, fine-tuning (transfer learning) on a training set representing a classification task and evaluation on the test-set drawn from the similar distribution as that of the training set. Read the paper here. Best Demonstration Paper: GAIA: A Fine-grained Multimedia Knowledge Extraction System About: In this paper, the researchers presented GAIA, which is claimed to be the first open-source multimedia knowledge extraction system. GAIA takes a huge stream of unstructured as well as heterogeneous multimedia data from different sources and languages as input, and then produce a coherent, structured knowledge base, indexing entities, relations, and events. Read the paper here. Honourable Mention Papers – Demonstrations: Torch-Struct: Deep Structured Prediction Library About: In this paper, the researchers introduced Torch-Struct, which is a library for the structured prediction that is designed to take the benefits of as well as integrate with vectorized-, auto-differentiation-based frameworks. TorchStruct includes a broad collection of probabilistic structures accessed through a simple and flexible distribution-based API that connects to any deep learning model. Read the paper here. Prta: A System to Support the Analysis of Propaganda Techniques in the News About: In this paper, the researchers presented Prta, the PRopaganda persuasion Techniques Analyzer. Prta makes online readers aware of propaganda by automatically detecting the text fragments in which propaganda techniques are being used as well as the type of propaganda technique in use. Read the paper here.","excerpt":"The 58th Association for Computational Linguistics (ACL) Conference announced its best research papers on computational linguistics. Besides the best paper award, the conference also announced other awards which include the honourable mention papers, best theme paper and best demonstration paper. Computational linguistics is the study of language from a computational perspective where the computational linguists […]","categories":["AI Trends"],"tags":["NLP papers","NLP research"],"author_name":"Ambika Choudhury","publish_date":"2020-07-14T13:00:00","publication_year":"2020","word_count":764,"keywords":["Go","artificial intelligence","AI","NLP papers","Git","RAG","NLP","Aim","deep learning","GitHub","R","NLP research"],"extracted_tech_keywords":["AI","artificial intelligence","deep learning","NLP","Aim","RAG","R","Go","Git","GitHub"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/acl-2020-announces-its-best-nlp-research-papers\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10112142,"title":"Google Launches TensorFlow GNN 1.0 for Advanced Graph Neural Networks","content":"The Google TensorFlow team has released TensorFlow GNN 1.0 (TF-GNN), an update to its machine learning framework to better develop and scale graph neural networks (GNNs). This new library can handle analysing complex networks, such as transportation and social networks. TF-GNN focuses on both the structure of graphs and the features of their nodes. This library bridges the gap between discrete graph data and continuous neural network models, for more detailed predictions and analyses. TF-GNN introduces a suite of advanced features for the TensorFlow ecosystem. At the heart of these advancements is the tfgnn.GraphTensor object, which represents heterogeneous graphs characterised by diverse node and edge types. This integration allows for the efficient handling of graph data, enhancing the TensorFlow ecosystem’s ability to manage complex network structures. The library provides a Python API, that can configure subgraph sampling for different computational environments, from individual workstations to distributed systems. This flexibility is crucial for handling datasets of varying sizes and complexities. Furthermore, TF-GNN introduces integrated gradients for model attribution, offering insights into the features most influential in predictions, thereby enhancing model training and evaluation. By incorporating the structure and data of graphs, GNNs offer predictions on entire graphs, individual nodes, or potential edges. This improves the understanding of complex relationships and attributes, making TF-GNN a powerful tool for a wide range of applications. TensorFlow GNN 1.0 is available as part of the TensorFlow ecosystem, with resources, documentation, and code samples accessible online for developers worldwide.","excerpt":"It simplifies the creation of models that understand complex relationships within data, from social networks to logistics.","categories":["AI News"],"tags":["API","Google","graphs","models","nodes","Python","Tensorflow"],"author_name":"K L Krithika","publish_date":"2024-02-07T11:19:03","publication_year":"2024","word_count":243,"keywords":["Go","API","machine learning","AI","neural network","models","Python","ai_frameworks:TensorFlow","graphs","Google","nodes","programming_languages:Python","TensorFlow","R","Tensorflow"],"extracted_tech_keywords":["AI","machine learning","neural network","TensorFlow","Python","R","Go","API","ai_frameworks:TensorFlow","programming_languages:Python"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/google-launches-tensorflow-gnn-1-0-for-advanced-graph-neural-networks\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10164491,"title":"The Rise of AI That Actually Reasons","content":"At India’s biggest GenAI Summit for developers, MLDS 2025, Rohit Thakur, GenAI Lead at Synechron, explored the next phase of AI development, moving beyond mere text prediction to models capable of logical reasoning. “LLMs today largely function on next-word prediction, which works well for a variety of NLP tasks,” said Thakur. “But not all tokens have the same information density. Some require much deeper reasoning, and that’s where new models are changing the game.” Beyond Next-Word Prediction For years, LLMs have been trained to predict the next word in a sequence, enabling tasks such as translation, summarisation, and chat-based interactions. However, this method has its limitations, particularly when handling complex reasoning tasks like mathematical problem-solving or multi-step logical deductions. Thakur explained how Ashish Vaswani’s seminal paper, Attention is All You Need, introduced the transformer architecture in 2017, which laid the foundation for modern LLMs. The evolution from BERT to GPT to reinforcement learning with human feedback (RLHF) has improved models, enabling preference tuning. However, they still rely on probabilistic text generation rather than structured reasoning. Chain of Thought reasoning, which breaks down multi-step problems, marked a breakthrough, but LLMs trained on next-word prediction continue to imitate reasoning rather than truly reasoning. “There’s a difference between emulating reasoning and actually reasoning,” Thakur noted. “LLMs trained only on next-word prediction might arrive at the correct answer, but they lack the ability to logically deduce it in the same way a human would.” From AI Assistants to Industrial Use Cases The rise of ‘reasoning AI’ is already having tangible effects across industries. Thakur shared a case study from a manufacturing company, where an AI assistant was deployed for engineering-related chatbot interactions. Engineers frequently queried the chatbot for specifications requiring mathematical calculations. “When we used a normal prompt meant for chat models, we got an incorrect output because some amount of calculation was needed—it said both models have these specifications.” Traditional LLM-based assistants often failed because they relied on probabilistic text completion rather than actual computation. By implementing reasoning AI, the chatbot improved its responses, correctly filtering product models based on kilowatt power ranges. This shift has implications beyond industrial applications. In fields such as finance, legal analysis, and healthcare, where AI-generated responses require verifiable logical steps, reasoning AI can significantly enhance accuracy and trustworthiness. “Imagine an AI that doesn’t just provide an answer but actually explains how it arrived at it,” said Thakur. “That’s the leap we’re witnessing today.” Interestingly, DeepSeek’s entry has reintroduced the focus on a structured reinforcement approach. Thakur explains how this method assists DeepSeek in building logical pathways rather than merely predicting probable answers. Proprietary models like OpenAI’s latest versions and Google’s Gemini AI are exploring similar methods, although details remain limited. However, deploying such models isn’t straightforward—Ben Hilack’s analysis revealed that reasoning models require different prompt structures compared to traditional LLMs, emphasising the need for customised input designs. The Road Ahead for AI That Thinks While reasoning AI is still in its early stages, its trajectory is clear. The transition from statistical prediction to structured reasoning marks a significant milestone in AI development. Open-source projects such as DeepSeek and refinements in reinforcement learning-based training will continue to push the boundaries of what AI can achieve. However, challenges remain. Future research should focus on ensuring that models generalise across diverse problem sets, avoid biases in reasoning pathways, and remain computationally efficient. Moreover, as Thakur highlighted, the shift towards task-specific prompting strategies will be crucial in maximising the potential of these models. “The future of AI isn’t just about predicting words—it’s about understanding and reasoning. And we’re only at the beginning of that journey,” he concluded. You can read more about Synechron’s transformative AI solutions here.","excerpt":"“LLMs trained only on next-word prediction might arrive at the correct answer, but they lack the ability to logically deduce it in the same way a human would.”","categories":["AI Highlights"],"tags":["AI reasoning","BERT","LLM","RLHF","Synechron"],"author_name":"Vandana Nair","publish_date":"2025-02-24T18:41:06","publication_year":"2025","word_count":616,"keywords":["GenAI","AI assistants","RLHF","OpenAI","AI","AI reasoning","TPU","LLM","ML","Synechron","BERT","NLP","chain of thought","R"],"extracted_tech_keywords":["AI","ML","NLP","GenAI","OpenAI","chain of thought","RLHF","AI assistants","TPU","R"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/the-rise-of-ai-that-actually-reasons\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":50,"title":"iCreate Software &#8211; BI, Analytics for Banks","content":"[section_title title=Page 1] [section_title title=Page 1] Founded in 2006, iCreate (www.icreate.in) is a specialist provider of Packaged Data warehousing and Analytical solutions to the Global Banking Industry. With its focus on the emerging markets of Asia Pacific, Africa, Middle East and Europe, iCreate has emerged as an easy-to-implement, productized alternative to the traditional tool driven bespoke DWBI implementations which are currently the norm. Delivered on best-of-breed BI and ETL technologies through OEM arrangements with global majors, iCreate provides banks with a lower risk, faster turnaround solution to their Intelligence needs. iCreate’s investment in robust data models, pre-built adapters to source banking systems and pre-packaged reporting\/dashboard and best practice KPI business content makes Biz$core the perfect solution for Banks looking to quickly implement an enterprise-wide decision support system. iCreate provides as value added services around Biz$core. Their service offerings around Biz$core include: Biz$tart A time-bound standalone consulting engagement that will enable building clarity around BI needs, evaluating current situation, and creating a BI roadmap, while sensitizing stakeholders across your organization to the business benefits from a BI initiative. BizManage BizManage is an SLA based managed solution model that keeps Biz$core installation functioning smoothly as well to power it with all the latest BI technology innovations. BizFactory – BI BizFactory – BI is a BI Center of Excellence that acts as an extension of BI team. The BI CoE charter is to primarily understand the MI and BI requirements on an on-going basis and develop the corresponding artefacts on the system and make it available. This includes development of new reports, dashboards, KPIs, Alerts and Notifications. BizFactory – Analytics BizFactory – Analytics is an Analytics Center of Excellence that acts as an extension of a bank’s Analytics team. The Analytics CoE comprises analysts who understand the functional areas of banking very well and also understand how to use the product to generate insights and views that will help business leaders of the bank shape business direction. This team helps the client generate business insights and analytical models to generate these insights. Biz$core is the quickest-to-implement, next generation BI, Analytics and Performance management solution for banks available at the lowest total cost of ownership. It is built on iCreate’s Banking Intellisense philosophy that combines the power banking domain knowledge and technology excellence. Biz$core interfaces with your core banking system using pre-built proprietary data adaptors. This makes data extraction, usually the most cumbersome part of the operation, a breeze. It delivers up-to-the-minute analysis of key business indicators, visualised using a powerful rendering engine that transforms the raw data into actionable information via handy graphic summaries. The Score function is a powerful performance management tool which allows you to set KPIs for team members, track them on a daily basis and communicate with your team. It’s time to start driving action from insights, in 100 days – with Biz$core. With a vision of being in the Top 5 for BI in Banking globally by 2014, iCreate raised Series A funding from IDG Ventures (www.idgvcindia.com) in 2010 and is taking its value proposition aggressively across the world through investments in direct sales as well as developing a network of distribution and implementation partners. The management team and board of iCreate are currently working toward mission “5:50:250 by 2014”. We wish to be in the Top 5 for BI in Banking, working with at least 50 strategic clients generating an annual revenue of INR 250 Crore by 2014. Given the current track record, the investments being made and the market’s encouraging response to the company’s offerings, Mission 5:50:250 is well within the company’s reach. [quote style=”1″]“Our vision is to be the partner of choice for Business Intelligence, Analytics and Performance Management solutions for growing progressive banks and to help them achieve remarkably better business performance through empowered decision making.” – VivekSubramanyam, CEO [\/quote] [media url=”http:\/\/www.youtube.com\/watch?v=CmYdcwIkApM” width=”430″ height=”260″]","excerpt":"[section_title title=Page 1] [section_title title=Page 1] Founded in 2006, iCreate (www.icreate.in) is a specialist provider of Packaged Data warehousing and Analytical solutions to the Global Banking Industry. With its focus on the emerging markets of Asia Pacific, Africa, Middle East and Europe, iCreate has emerged as an easy-to-implement, productized alternative to the traditional tool driven […]","categories":["Deep Tech"],"tags":[],"author_name":"Fintellix","publish_date":"2012-05-22T16:59:49","publication_year":"2012","word_count":640,"keywords":["business intelligence","Go","funding","AI","ETL","innovation","RAG","analytics","GAN","R"],"extracted_tech_keywords":["AI","analytics","RAG","R","Go","ETL","GAN","business intelligence","innovation","funding"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/icreate-software-bi-analytics-for-banks\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10075345,"title":"IIT Madras and Mitsubishi Electric to Collaborate on Semiconductor Research","content":"Mitsubishi Electric India has inked an MoU with IIT Madras Research Park to support the institute’s research and development of sustainable technologies. A pioneer in semiconductor technology, Mitsubishi Electric India aims to achieve rapid innovation and technological upliftment for the nation’s consistent development with this MoU. It is currently one of the leading brands contributing towards such advancement through its electrical and electronic solutions for diverse production needs in India. The first university-based park in India, IIT Madras Research Park fosters innovation and entrepreneurship through academic and industry collaborations that help digital start-ups and students alike in building their entrepreneurial skills and advanced learnings. Ashok Jhunjhunwala, institute professor and president, IIT Madras Research Park and IITM Incubation Cell, said,“Make in India requires Design in India. We need the best of components to let our homegrown product compete with imported products and in the international market. Our partnership with Mitsubishi Electric India allows us to move towards this larger goal. Our products need to be developed through the best manufacturing process and have superior quality, while still being low cost.” By offering tools and workshops to assist them, this collaboration is expected to enable the students to learn cutting-edge semiconductor technologies and supply power semiconductors in accordance with the institute’s needs. The company has reportedly established a support centre at the research park to assist in resolving queries and providing useful information to academics and students.","excerpt":"This initiative by Mitsubishi Electric will provide components, practical knowledge and workshops to help students learn the innovative semiconductor technologies.","categories":["AI News"],"tags":["IIT"],"author_name":"Bhuvana Kamath","publish_date":"2022-09-15T16:48:33","publication_year":"2022","word_count":235,"keywords":["Go","API","programming_languages:R","AI","innovation","programming_languages:Go","Git","Aim","IIT","R"],"extracted_tech_keywords":["AI","Aim","R","Go","Git","API","innovation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/iit-madras-and-mitsubishi-electric-to-collaborate-on-semiconductor-research\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10101210,"title":"How AWS is Democratising Access to Quantum Computing in India","content":"India has ambitious plans and aspirations in the realm of quantum computing. Earlier this year, the Indian government allocated INR 6000 crore towards the National Mission on Quantum Technologies and Applications, which aims to develop indigenous quantum computers and advance quantum technologies. Simultaneously, the country is also actively investing in a quantum workforce, research and development, establishing labs and institutes dedicated to quantum computing, and supporting startups in the field. However, to advance India’s ambitions in the quantum space, access to quantum computers is pivotal. As part of the National Quantum Mission (NQM), India will develop intermediate-scale quantum computers with 50-1000 physical qubits, but they are expected to be delivered in the next eight years. This is where services like Amazon Braket could prove to be monumental, since it enables developers and researchers to test their quantum computing algorithms on quantum simulators and quantum hardware. By doing this, they can get a reasonable degree of confidence in the algorithm’s performance before running it on an actual quantum computer, which is expensive and generally takes time to get access to. “Without access to a quantum computer, the barriers are significant. Researchers seeking to utilise a quantum computer face a cumbersome process. They must secure a time slice on a quantum computer, wait for their turn, reminiscent of the process used with large computer machines in the 1960s,” said Kanishka Agiwal, head of service lines, India & South Asia at AWS, sharing his aim to democratise access to quantum computing. Collaborates with MeitY AWS is not only bringing Indian developers and researchers access to quantum computers, it recently partnered with the Ministry of Electronics and Information Technology (MeitY) to set up the Quantum Computing Applications Lab (QCAL), which is the only initiative of its kind in India that provides quantum computing as a service on the cloud to government ministries and departments. “So, the lab became our first effort in India, which would call for proposals from various academicians, researchers, institutes, startups, and developers to come in and tell us what kind of quantum problems they want to solve?” Agiwal said, adding that this was done across two different cohorts. The first received around 21 applications and 17 of them were greenlighted. The second cohort was relatively bigger and received 75 applications, out of which 62 were greenlighted. “And they’re across multiple domains. From a technology standpoint, they’re doing things around quantum machine learning, quantum mechanics, quantum crypto, quantum key distribution, quantum materials, sciences, and so on and so forth. Whereas from a domain standpoint, we’re doing things in agriculture, in smart infra or financial or health care, protein gene folding, etc,” Agiwal said. As part of the second cohort, AWS has partnered with the principal scientific advisor’s office and other private players such as Mphasis, Fractal, among others. “We had multiple partners come in and evaluate these proposals,” he said. The proposals, which were greenlighted, Agiwal says, will receive a number of benefits such as AWS credits to run their quantum computing workload, support from subject matter experts, and constant mentoring. Building a quantum workforce With quantum computers, India is also aiming to build a quantum computing workforce and is emphasising a lot on training its young population on quantum technology. The NQM focuses on nurturing a skilled workforce in quantum technology and fostering collaborations with international partners to accelerate technological advancements. AWS focus is also aligned with the NQM regarding developing a quantum technology skill force in India. “As we were doing our first cohort we realised there is a skill base that needs to be built up here in India. And that’s when we started introducing quantum-specific programmes or courses or curriculum with some of our partners like Mahindra University, QpiAI, with whom we have launched specific quantum courses,” Agiwal said. Mahindra University currently offers a couple of courses on quantum technology, targeting working professionals as well as students who are still in the learning phase. “Our intent is that India will have a workforce or a skill base of quantum computing, essentially aligned to the NQM,” shared Agiwal. Democratising access to quantum computers Amazon Braket essentially provides access to four major quantum computers or quantum hardware as well as a quantum computing simulator. While AWS does not have a quantum computer for now, the hyperscaler is partnering with companies like IonQ, Rigetti, OQC, Xanadu, and QuEra to make quantum computers more accessible. He also stated that the simulator is a key differentiator in the way AWS brings this forward to the researchers as well as the developer community. “Sometimes utilising a quantum computer for algorithms can yield unsatisfactory results, with accuracy as low as 30% or 40%,” said Agiwal, adding that repeated recording and experimentation are required, which is a time-consuming and cost-prohibitive process. “This is why simulators offer a more cost-effective and accurate initial development stage before transitioning to a quantum computer. This approach reduces time-to-market, scales efficiently, and minimises costs,” he added. With Braket, what AWS is aiming to do essentially is to democratise the access to quantum computers. Anybody with an AWS account and a decent laptop can have access to these quantum computers, Agiwal said. “So, you don’t have to be a researcher at a particular facility that houses a quantum computer for you to go and access it. With cloud, we have expanded the scope of adoption, the scope of experimentation and everything else around it.” While other hyperscalers like Azure offer similar services, AWS stands out by providing access to various quantum computers and including a simulation component within Braket. This combination is what sets AWS apart, according to Agiwal.","excerpt":"Amazon Braket provides access to four major quantum computers or quantum hardware and also provides a quantum computing simulator","categories":["AI Trends"],"tags":["AWS","india quantum computing"],"author_name":"Pritam Bordoloi","publish_date":"2023-10-06T15:09:21","publication_year":"2023","word_count":938,"keywords":["Go","machine learning","AWS","AI","cloud_platforms:AWS","R","india quantum computing","Aim","quantum machine learning","Azure","startup"],"extracted_tech_keywords":["AI","machine learning","Aim","quantum machine learning","AWS","Azure","R","Go","startup","cloud_platforms:AWS"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/how-aws-is-democratising-access-to-quantum-computers-in-india\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":40310,"title":"How Open Banking &#038; AI Can Drive Innovation In The Finance Sector In India","content":"These are very happening times for the largest democracy in the world with elections in full swing and multiple financial events lined up like the Regulatory Authority of the country (RBI) coming out with their vision for Payment and Settlements 2019-2021 which was indeed very enterprising as it promises to make the country a cash-lite society. We also had Niti Aayog (Transforming Committee of India) proposing a significant 7500-Crore plan for Cloud Computing and Artificial Intelligence push which includes establishing “AIRAWAT” a cloud computing platform for promoting research on AI. It certainly is a step in the right direction, given the changes that we are seeing in the world around us and it was time that we started work in this direction. In between all these announcements and happenings, there was one more proposal published by Niti Aayog but not much was written or said about this maybe because it was just an idea. The proposal was to end Data Monopoly and make data available to all, in other words, move the financial system towards Open Banking, which brings us to some very important questions like the following: – Who owns customer data which as of now is lying with a few big monopolies who act as custodians of data. The banks or any other entity stores customer data for their own use, and this is exactly what the proposal by Niti Aayog wants to change and to make data available to anyone who would be able to use the data to carve out a better deal for the customer because ultimately it is the customer who should be owner of their own data. What is Open Banking? It basically is creating a Data Architecture where a network of institutions can share the data through APIs (Application Programming Interfaces). The data architecture is used to create guidelines on how customer data can be created, structurally stored and securely accessed. If this happens it would open the doors to a lot of innovation and entrepreneurship opportunities and at the same time benefitting the single most important entity and that is the Customer. As a country, India has more than a thousand fintech firms which are growing at a significant growth rate and as per estimates, we are looking at somewhere around $73 billion worth of transactions in 2020 which corresponds to more than 20% growth rate from where we are today. We would certainly not be the pioneers in walking the path to Open banking as there are quite a few countries\/continents which have already started their journey like the implementation of PSD2 across Europe, with the UK having their own version of Open Banking with specific standards. In U.S. big banking entities have come together and are working on the idea of sharing data within their own network. Countries like Australia have already come out with CDR (Customer Data Right) and are well on their way to implement Open Banking, so it is just a matter of time where we will be having no option but to do it. Analytics and AI are already changing the way we are looking at data and today every organization wants to be known as a data-driven organization.  Analytics is slowly turning to be among one of the core competencies of every organization and we have a lot of fintech startups which are working on cutting edge technologies but are hampered by right data not be available to them. Open Banking will certainly be the enabler as in essence, it will provide access to insights on how the data can be used to launch new products and provide benefits by taking a holistic view of the data rather than keeping it confined to the prerogative of chosen few. Predictive Analytics which is already struggling to establish a pattern on customer behaviour and providing insights due to lack of available structured data can re-model themselves with open banking and our big banks can play a leading role by taking advantage of the computing power that they already have and combining it with the data being made available by creating new propositions for the customer and in return making the customer much more profitable and loyal to them. We would also see a lot of new entrants in analytics which would be nimble footed and are waiting to challenge the big banks in their own turf which would, in turn, make these big banks be on their toes to ward off any upcoming challenge by innovating or partnering with them. All and all a winning proposition for everyone and mostly to the Customer who would finally get the right benefit of his or her own data. But as we all know with great powers comes great responsibility and with such huge and sensitive data, we will need to have proper guidelines and standards to protect the privacy of the customer. It would not only be restricted in making the customer click the small checkbox at the bottom or Rules and Regulation page but would mean properly making the customer aware of how the data is going to be used and responsibility will completely lie with the entity using the data. All said in a few years if we as a country must survive and promote innovation we will have to walk with the world and implement the changes, not after everyone has done but along with everyone and with the way things are moving, we certainly are in the right direction.","excerpt":"These are very happening times for the largest democracy in the world with elections in full swing and multiple financial events lined up like the Regulatory Authority of the country (RBI) coming out with their vision for Payment and Settlements 2019-2021 which was indeed very enterprising as it promises to make the country a cash-lite […]","categories":["AI Features"],"tags":["Banking","Cloud Computing","Predictive AI","predictive analytics"],"author_name":"Sudhish Nair","publish_date":"2019-06-06T09:04:46","publication_year":"2019","word_count":915,"keywords":["Go","API","artificial intelligence","AI","cloud computing","R","data-driven","Cloud Computing","Banking","analytics","GAN","predictive analytics","Predictive AI"],"extracted_tech_keywords":["AI","artificial intelligence","analytics","predictive analytics","cloud computing","R","Go","API","GAN","data-driven"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-open-banking-ai-can-drive-innovation-in-the-finance-sector-in-india\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":26476,"title":"Meet ChiSquare, A Data Analytics Startup That Exceeded €1 Million In Annual Revenue In Less Than 4 Years","content":"While there has been a boom in data and analytics companies offering specialised services such as data analytics, customer analytics, web analytics, among others, in the country, ChiSquare stands out of the crowd in its approach. Founded in 2014, this Hyderabad-based startup has been enabling data-driven decisions for companies across the globe. Analytics India Magazine caught up with Mayank Pachauri, CEO of ChiSquare Consultants to get an insight into the making of this self-funded startup and how they have worked their way to stay profitable since their inception. From An Idea To Reality Recalling the early days of ChiSquare, Pachauri shares that it was founded in December 2014 when he decided to come back to India after spending 10 years in Europe. He returned to Hyderabad, where he helped a former boss build an entire reporting and analysis team for him. “We mutually agreed to deliver services for his organisation remotely. Hence, ChiSquare came into existence and since then there has been no turning around,” he said. What was once started as a small company in a two-bedroom flat, with three part-time colleagues, is today a well-established organisation in a 10,000 sq ft office with 42 full-time employees. What Is ChiSquare? ChiSquare takes the raw transactional data from businesses and works through its Key Value Drivers (value sensors), linking them to the Key Performance Indicators (KPIs) and generate actionable insights. Explaining the working, Pachauri explained that for an e-commerce business which has been operational for a few years, customer acquisition is critical for its growth and revenue maintenance. In such a scenario where almost 20 percent of its revenue comes from new acquisition, ChiSquare helps business in identifying: The profitable source of acquisition and Touchpoints which makes the customers stay for a longer period “Our experience in e-commerce business data has been fundamental in providing decision making support to each level of organisation. We have off the shelf concepts on assessing the business health through ‘State of Play Assessment Framework’ where we dissect the business revenue model. Another concept in the application is ‘Customer Lifecycle Management’ covering customer acquisition-customer retention-customer experience triad,” said Pachauri. For instance, they try to answer questions such as which customers they want to retain? Which is the right customer segment to invest in? What percentage of customer churn comes from customer experience issues or from system generated friction? How customer responds to incentivisation program?. Their modular approach helps in delivering value through contextual storytelling within 21 days of getting data access. The ChiSquare team ensures these offerings with its three models of services, which are: Business Intelligence Implementation: Developing data models, implementing data warehouse, building self-service reporting and data visualisation (where organisations do not have their BI in place or Bi is not fully developed). Data Visualisation, Reporting And Analysis: Where organisations already have BI in place. Storytelling Services And Actionable Insights: Using data science services (where organisation have BI and matured reporting systems). He mentions that these services are not mutually exclusive as all of the above three are also offered as a package. Success Stories With Clients Pachauri shares that most of their clients are spread across Europe. “Online gaming being their area of expertise, we associate with top players across the globe. Fortune Entertainment Group and GVC Holding (the parent company to major brands like Bwin and PartyGaming), TIPICO based out of Germany, are among our top clients. We also have non-gaming clients like PayU Europe,” he said. Explaining the use case of GVC Holdings, Pachauri said that their research team analysed the marketing spend of 2017 for the company across their markets and channels. Following the post-impact analysis, they recommended the company to re-allocate and optimise the marketing spend (from regions not yielding profitable results), generating spend savings of more than € 15 million. For Fortune Entertainment Group, they built the entire business intelligence system from the scratch and established their whole analytics operations from India generating value for them. “We are key for their stellar growth in last 3 years”, he said. For PAYU Europe, the team built their first system performance report which PAYU executives used for system improvements for their B2B businesses. The Team Driving Success With a team of 42, comprising of full stack developers, data scientists, analysts, data visualisers, among others, they all come with rich academic and corporate backgrounds. They also have a strong team of engineers for BI development and a team for data visualisation and storytelling. For storytelling, they have people from business, commerce background who have mathematics, economics, statistics as one of their core subjects. “For data science, mathematics and programming is must. If you do both, plus you understand the business, you are a star for us,” he added. The Road Ahead Pachauri admits having faced many challenges being at the forefront of analytics, one of them is on the business front, where he shares that many of the businesses treat analytics as a cost centre and cannot see the value it can add. “Willingness to invest is still not fully there,” he said. Second is the lack of resources in this domain, which is usually required to be developed in-house and that takes time. Despite these challenges, ChiSquare has shown a tremendous growth over the years from signing its first deal in 2015 to the present where they have exceeded 1 million euros in annual revenue with leading names across the globe as their clients. And while the startup has been self-funded since the day one, they are now actively looking for funding as they move into the expansion stage and productisation of analytical services. Talking about the competition in the analytics space which is constantly expanding, Pachauri says that more than the competition, it is the pace of technology which is difficult to keep pace with. “For me what is important is the focus on your strength and capabilities. Real analytics requires domain expertise. For you to create stories from numbers, you have to know your client’s business thoroughly. Numbers by themselves cannot talk,” he said in the concluding.","excerpt":"While there has been a boom in data and analytics companies offering specialised services such as data analytics, customer analytics, web analytics, among others, in the country, ChiSquare stands out of the crowd in its approach. Founded in 2014, this Hyderabad-based startup has been enabling data-driven decisions for companies across the globe. Analytics India Magazine […]","categories":["AI Startups"],"tags":[],"author_name":"Srishti Deoras","publish_date":"2018-07-17T10:11:03","publication_year":"2018","word_count":1010,"keywords":["business intelligence","data science","funding","startup","AI","data-driven","analytics","GAN","R","data warehouse"],"extracted_tech_keywords":["AI","data science","analytics","R","data warehouse","GAN","business intelligence","data-driven","startup","funding"],"url":"https:\/\/analyticsindiamag.com\/ai-startups\/meet-chisquare-a-data-analytics-startup-that-exceeded-e1-million-in-annual-revenue-in-less-than-4-years\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10113395,"title":"NVIDIA&#8217;s Stock Momentum Swings, Google &#038; Amazon Take Lead Again","content":"On the same day when NVIDIA surpassed Elon Musk’s AI company Tesla as the most traded on the market, its stocks faced the worst day since October 2023, that too, a day before the highly anticipated fourth-quarter results. The American chip maker started this year with a 47% year-to-date gain, and the stock momentum came with gains in 2023 as well. But the tides have shifted, and the company lost $78 billion in market value. On Tuesday morning, its stock price was $741 per share, with a market cap of $1.83 trillion (ahead of Amazon’s $1.75 trillion and Alphabet’s $1.78 trillion). Then Nvidia shares dipped as much as 8.6%, sending the market cap as low as $1.67 trillion before the company’s stock began to recover. Now, Alphabet and Amazon are once again ahead of of the chip maker. CNBC’s Cramer reported, “Open your eyes, people,” Cramer said. “If you think that Nvidia’s quarter is one and done, you’re also thinking that AI is one and done.” The long-term supporter of the company is not wrong since Wall Street continues to bet big on AI, from which NVIDIA emerged as a winner. All eyes are on the AI company with the rising demand for chips in the market and the upcoming GTC conference in San Jose, CA. The upcoming event in March is expected to bring updates on Blackwell, NVIDIA’s next-gen architecture. The company has already confirmed that their roadmap for 2024-25 features Blackwell, namely the B100 and GB200, which have been listed. The teaser video gave a glimpse of generative AI’s features like the WPP\/NVIDIA engine for digital advertising, the newly introduced Chat with RTX, an industrial metaverse powered by SyncTwin, AI art by Refik Anadol Studio, and OpenAI creating code for Blender animations. Wall Street analysts are currently focused on the company’s demand outlook for its AI-enabled H100 GPU chips, which can sell for upwards of $40,000.","excerpt":"“If you think that Nvidia’s quarter is one and done, you’re also thinking that AI is one and done.”","categories":["AI News"],"tags":["Amazon","Google","NVIDIA"],"author_name":"Tasmia Ansari","publish_date":"2024-02-21T11:11:43","publication_year":"2024","word_count":318,"keywords":["OpenAI","AI","programming_languages:R","RPA","Amazon","Git","generative AI","Google","AI art","GAN","NVIDIA","R"],"extracted_tech_keywords":["AI","generative AI","OpenAI","R","Git","GAN","RPA","AI art","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/nvidias-stock-momentum-swings-day-before-q4-results\/","complexity_score":4,"technical_depth":9,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":18342,"title":"Can Machines Really Think? Looking Into AI And The Role Of Human Psychology","content":"“A year spent in artificial intelligence is enough to make one believe in God.”– Alan Perlis, Epigrams in Programming (1982) ‘Technology’ is defined as ‘the use of scientific knowledge for practical purposes’ (Oxford Dictionary). This terminology is adaptable to space and time it exists in – Olduvai inhabitants used stone scraping technology to butcher animals for sustenance 2.5 million years ago, and the plough was invented for agricultural purposes in 4000BC. With the foundation of telecommunication services and Charles Babbage’s first mechanical computer the ‘Difference Engine’ in 1822, a substratum for rapid advancements in digital and analogous apparatus was laid. Artificial Intelligence (also known as AI) did not enter the techno-scene until the 1940s when the world’s first programmable digital computer was invented, and abstract mathematical reasoning was used to perform computational tasks. It is defined as ‘The theory and development of computer systems able to perform tasks normally requiring human intelligence, such as visual perception, speech recognition, decision making and transition between languages’. (Russell and Norvig, 2003) In other words, AI is the simulation of human intelligence by machines through the processes of learning, reasoning, and self-correction. We interact with AI on an everyday basis, through video games, smartphone assistants like Siri or Cortana, purchase prediction by online retailers such as Amazon and Flipkart, customer support chat-bots, and more. The fascinating world of AI permeates our daily lives and has, in the last two decades, made significant strides in technological progress. AI is closely linked to human psychology – the basis for research in the field stems from the network of neurons in the human brain. These neurons make connections based on your perception and outside stimuli, and transmit this information through electric and chemical signals, allowing adaptive learning to take place. AI finds its primary objective in emulating human neurological functions, hence mimicking the way rational human cognitive thinking takes place. Given this, its complexity and scope are increasing rapidly today. From ‘context awareness’ mechanisms bringing about an ability to think and create like humans to ‘natural language processing’ which gives AI a platform and method to communicate, every step seems to be one taken in the direction of making it wholly analogous to human functioning and thinking. A precursor to answering the question of whether machines can one day independently think for themselves using AI must be to understand the connotations of ‘thinking’, and intelligence itself. Advancements in the technology of artificial intelligence are aimed at emulating human intelligence and eventually, indeed, creating machines that can ‘think’ for themselves and carry out functional tasks analogous to humans. ‘Thinking’ involves the conscious cognitive processes of the human mind such as processing information, engaging in problem-solving, decision making and reasoning. It is the longstanding goal of AI mechanisms to replicate the processes of thought to eventually replicate thought itself. Thinking allows humans to interpret the world around them and make analyses and predictions based on this understanding and AI aims to do exactly that. Alan Turing’s ‘Turing Test’ is a test of a machine’s ability to exhibit intelligence in behaviour on an equivalent or superordinate level relative to that of humans. In his work ‘Computing Machinery and Intelligence,’ he replaces the question ‘Can machines think?’ with ‘Are there imaginable digital computers which would do well in the imitation game?’ Given the difficulty in giving a concrete definition of the word ‘thinking’, the latter question, Turing believes, is one that is truly capable of being answered, while the former may be considered ‘too meaningless to deserve discussion.’ (Turing, 1950) While human intelligence can be plainly defined as the ‘capacity to acquire and apply knowledge’, intelligence in AI includes usage of algorithmic coding for eventual self-programming, leading to the ability to decode and interpret data and make predictions based on it. In this sense, AI may indeed be termed as ‘intelligent’. However, thinking in AI includes the ability of a machine to write and code its own programs and interact with humans in a complex fashion. ‘Artificial General Intelligence’ or AGI refers to the ability of the AI mechanism to not only work in the context of a specific assigned task that it is trained for but to go beyond that by adapting to a variety of situations and reprogramming itself accordingly, and AGI is now the primary universal objective of AI developers. One can say that AI mechanisms move towards a better imitation of the human biological neuron networks, psychology and thinking every day by mimicking human cognitive processes. This, however, remains devoid of conscience and emotional ability that is a part and parcel of the human psyche. Hence we may conclude that machines can think if the definition of the terminology is manipulated to suit the capabilities of technology. There is still, however, a scope for AI to make further strides through advancements in technology, keeping in line with the concept that the future may always be uncertain.","excerpt":"“A year spent in artificial intelligence is enough to make one believe in God.”– Alan Perlis, Epigrams in Programming (1982) ‘Technology’ is defined as ‘the use of scientific knowledge for practical purposes’ (Oxford Dictionary). This terminology is adaptable to space and time it exists in – Olduvai inhabitants used stone scraping technology to butcher animals […]","categories":["IT Services"],"tags":["AI (Artificial Intelligence)"],"author_name":"Sukhnidh Kaur","publish_date":"2017-10-15T05:00:30","publication_year":"2017","word_count":819,"keywords":["Replicate","Go","API","artificial intelligence","programming_languages:R","AI","programming_languages:Go","Git","Aim","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","Aim","R","Go","Git","API","Replicate","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/can-machines-really-think-looking-into-artificial-intelligence-and-the-role-of-human-psychology\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10051140,"title":"The 4th Edition Of Machine Learning Developers Summit Is Back In Virtual Format | 19-20th Jan","content":"Furthering the machine learning ecosystem, Analytics India Magazine brings you the fourth edition of the Machine Learning Developers Summit (MLDS22). Centring on innovation in machine learning, this two-day conference is a ticket to direct networking with renowned visionaries across the ML industry who will talk about the software architecture of ML systems, producing and deploying the latest ML frameworks and solutions for business use cases. WHERE: VirtualWHEN: Wednesday to Thursday, January 19 to 20, 2022 Trend-setting practices across various sectors will be discussed through cutting-edge presentations, which will provide the attendees with valuable take-away, including applications in their respective industries and businesses. With more than 1000+ attendees, 200+ organisations, and 30+ industry leaders as speakers, MLDS22 is the gold standard for India’s data science & machine learning ecosystem for gaining exposure to ML tools, frameworks and platforms, evaluating new projects, hardware & software challenges of building complex ML systems, languages, software and standards. Earlier, MLDS has been sponsored by Genpact, Intuit, and the Association of Data Scientists, amongst others, and the conference has had speakers from industry giants like Microsoft, Google, Oracle, Amazon, Walmart Labs, Flipkart, AWS, Freshworks, Capgemini, Intel, NVIDIA, IBM, Netflix, Cisco, Cognizant, Ericsson, SAP, L&T, Siemens, Verizon, ZS, EY, TCS, Bosch, Renault Nissan, Aditya Birla Group, PhonePe, BigBasket, Voot, and Paytm, amongst many others, over the last three years. Two major attractions apart from the guest talks include: Call for Papers MLDS-2022 welcomes submissions reporting research that advances artificial intelligence, broadly conceived. The conference scope includes all subareas of AI and machine learning. AIM and MLDS22 expressly encourage work that cuts across technical areas or develops AI techniques in the context of important application domains, such as healthcare, sustainability, transportation, and commerce. The submissions are open for 2022, and more details can be found here. (now closed) 40 Under 40 Data Scientists Awards 40Under40 is an attempt towards recognising the leading data scientists in India who have successfully transformed data into meaningful insights. The award recognises India’s young data scientists and leaders supporting the growth of analytics in their organisations, have deep industry and analytics expertise and are working creatively on the insights gap for clients by addressing their most complex challenges. The nominations are open for 2022, and more details, including the list of previous years’ winners, can be found here. Nominations can be sent in through this form. (now closed) Registration and Tickets Book your passes here to reserve your seat now. The conference will be hosted virtually. Register here. The schedule and list of speakers for this year’s Machine Learning Developers Summit have been announced. Check this space for relevant announcements. Be a part of India’s no.1 conference exclusively for machine learning practitioners’ ecosystem by attending the Machine Learning Developers Summit 2022. We look forward to having you with us for the event. Hurry up and book your seat now!","excerpt":"Furthering the machine learning ecosystem, Analytics India Magazine brings you the fourth edition of the Machine Learning Developers Summit (MLDS22). Centring on innovation in machine learning, this two-day conference is a ticket to direct networking with renowned visionaries across the ML industry who will talk about the software architecture of ML systems, producing and deploying […]","categories":["Deep Tech"],"tags":["AI (Artificial Intelligence)","conference","Data Science","Data Scientist","Deep Learning","hardware","languages","Machine Learning","machine learning conference","machine learning developers summit","Programming Languages","Software","use cases"],"author_name":"Anushka Pandit","publish_date":"2021-10-11T11:59:07","publication_year":"2021","word_count":476,"keywords":["machine learning conference","conference","Programming Languages","use cases","R","data science","artificial intelligence","RAG","machine learning developers summit","languages","analytics","Data Science","machine learning","AWS","AI","ML","Machine Learning","hardware","Software","Aim","Deep Learning","Data Scientist","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","data science","analytics","Aim","RAG","AWS","R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/4th-edition-machine-learning-developers-summit-announced\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":26200,"title":"How To Make The Most Of Open Data While Caring About Privacy","content":"Open data is a massive force for betterment of the society. According to a Mckinsey report, “more than $3 trillion in economic value globally that could be generated each year through enhanced use of open data”. The open data nowadays is machine readable and therefore can be easily put to use. There are a number of examples where open data has made important change in society. For example, in Burkina Faso, a country in Western Africa held their elections after 3 decades. Many open data storages and apps gave an advantage to the people. The open data gave the public the power to choose the right candidate and get prior knowledge of election results. The country which had no experience of democracy leveraged the power of open data to avoid rumours and tensions and make the right decision. The data store and technological innovations helped Burkina Faso conduct elections fairly. Problems With The Open Data Movement As many closed data sources become open it becomes easier build community services and products from them. There is a possibility that even though open data might seem a good idea in theory, in application it will lead to increase in digital divide and social inequality. There have to be a set of rules under which open data projects must be undertaken as to maximise the utility and minimise the risk. Besides, there is also the danger of declaring open data sets that are in technical formats and make no sense to the general public. Hence only the tech experts are able to leverage the power of such data. Also people who can hire and employ such experts are well positioned to take advantage of the data. All these situations will surely make the digital and social divide even bigger. Such things can very easily turn into economic advantages. Another great problem looming over concept of open data is privacy. Privacy has already become a critical issue in the context of open data. When large amounts of data is made public there is something known as Mosaic effect. According to Elizabeth Gorgue, privacy officer at the e-Government Office at the County of Santa Clara, California, “Data elements that in isolation look relatively innocuous can amount to a privacy breach when combined”. Therefore the data that contains personally identifiable information (PII) should not be put out publicly. There are some preprocessing steps that have to take place before releasing it. How to achieve openness with privacy remains the most critical question of the day. The Centre for Open Data Enterprise has come out with a report that shines a light on a slew of issues by data and privacy experts: What are the potential benefits of using unaggregated data (or microdata) for the public good? What are the risks of using these datasets if they contain or could lead to the discovery of personally identifiable information, and how can those risks be minimized? What are the best technical, ethical, and policy approaches to ensure strong privacy protections while maximizing the benefits of open data? Striking A Balance There has to be a balance between openness, effectiveness of open data and privacy protection. For a long time privacy has been ignored but now policy makers and technicians have made privacy the center of open data initiatives. Along with the problem with the privacy there is also an additional issue of bias in datasets which is being addressed by researchers. The data however public wont make it free of bias. The White House report on big data and privacy, put some limitations on the publicity of educational data. “As students begin to share information with educational institutions, they expect that they are doing so in order to develop knowledge and skills, not to have their data used to build extensive profiles about their strengths and weaknesses that could be used to their disadvantage in later years.” We might also benefit from observing what approaches have been successful in the past. The report suggests some ways to address the problem of the Mosaic effect and the misuse of public data. As the US President’s Council of Advisors on Science and Technology puts it as “anonymization remains somewhat useful as an added safeguard, but it is not robust against near‐term future re‐ identification methods.” The de-identification approach suggests anonymizing datasets will always make it hard bad actors to misuse data. The report suggests another approach of “semi open data” where data is open for only some use cases and not open for others. One example of this approach is to make personal data available only to the individual user and not to the whole public.","excerpt":"Open data is a massive force for betterment of the society. According to a Mckinsey report, “more than $3 trillion in economic value globally that could be generated each year through enhanced use of open data”. The open data nowadays is machine readable and therefore can be easily put to use. There are a number […]","categories":["AI Features"],"tags":["open data","Privacy"],"author_name":"Abhijeet Katte","publish_date":"2018-07-05T12:07:38","publication_year":"2018","word_count":776,"keywords":["big data","Go","Privacy","open data","AI","programming_languages:R","innovation","programming_languages:Go","Git","RAG","R"],"extracted_tech_keywords":["AI","RAG","R","Go","Git","big data","innovation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-to-make-the-most-of-open-data-while-caring-about-privacy\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10043579,"title":"Hands-On Workshop: Accelerate PyTorch Applications Using Intel oneAPI Toolkit","content":"Developers often struggle with high computational complexity, algorithmic challenge and inference of large neural networks while accelerating deep learning applications. PyTorch facilitates building deep learning projects by allowing the creation and training of deep neural networks and the ability to perform accelerated mathematical operations on dedicated hardware. PyTorch can be used to train a large amount of data at a rapid pace and is also flexible enough for experimentation and prototyping. However, the deployment of complex models with thousands to millions of parameters and large numbers of layers require heavy GPUs. The model can be pruned or quantised to optimise to a certain level. Intel oneAPI is introducing a hands-on approach to accelerate PyTorch applications and allow developers to deploy models in a resource-constrained environment. Intel® Extension for PyTorch* (IPEX), part of Intel’s oneAPI AI Analytics toolkit, is designed to make the PyTorch CPU experience better while achieving good performance. Our last workshop introduced Intel® oneAPI AI Analytics Toolkit to the community of developers as a component to maximise their ML and deep learning performance. This workshop, on the other hand, will do a deep dive on how to accelerate PyTorch applications using Intel oneAPI Toolkit. Analytics India Magazine, in association with Intel®, has put together a hands-on virtual workshop on August 18, 2021, to unpack Intel® Extension for PyTorch*. The participants will learn how to train a model using Intel® Extension for PyTorch* and use the PyTorch extensions for inference. Expert trainers from Intel will also demonstrate how to accelerate AI inference performance with Intel® Distribution of OpenVINO™ Toolkit through ONNX. Register For The Virtual Workshop Here. What You Will Learn — Introduction to Intel® Extension for PyTorch*Using Intel® Extension for PyTorch* for AI Training Inference using Intel® Extension for PyTorch*Inference of PyTorch models using Intel® Distribution of OpenVINO™ Toolkit Who Should Attend? AI & ML developersData scientistsAI enthusiastsAI researchersGPU & HPC programmers Register For The Virtual Workshop Here. Exclusive Contests — Participate & Win! Lucky Draw Contest Analytics India Magazine is running a Lucky Draw wherein 10 lucky registrants on Intel DevCloud will get a chance to WIN Amazon Voucher worth INR 1000\/- each. So please register on Intel DevCloud with the following event code – oneAPI18AUG. HURRY UP !!! ______________________________________________ #oneAPI Workshop Contest Analytics India Magazine is also hosting an exciting contest, exclusively for the participants, as a part of the workshop. Up to 2 winners will each receive a GOQii Smart Watch. To know more about the contest rules & regulations, click here. ______________________________________________ DevMesh Project Contest — Participate & stand a chance to win INR 5000\/- worth Amazon gift vouchers. Join the Intel® DevMesh, a community portal for developers and creators to share their work & best practices. || Click here — https:\/\/devmesh.intel.com. Submit your projects using Intel® DevCloud & oneAPI toolkit on Intel® DevMesh portal by 30th August 2021 & email the project link to kavita.aroor@intel.com.If there are projects that use PyTorch it could be replaced by Intel’s Extension of PyTorch* (for  training and Inference) or Intel Distribution of OpenVINO™ (for Inference) and other relevant oneAPI components. After replacing the components, the performance can be compared with your stock counterparts and the same can be submitted for Intel’s evaluation.Projects submitted must include usage of Intel DevCloud & oneAPI toolkits. Top 8 projects will stand a chance to win Amazon Gift Voucher worth INR 5000\/- each! Session Speakers: Kavita Aroor — Kavita Aroor is a Developer Marketing Manager- APJ at Intel focusing on developing the software solutions market across the Asia Pacific. She drives the developer outreach, engagements and programs in the areas of GPU, HPC and AI. Kavita comes with 17 years of experience in brand and field marketing. Lakshmi Narasimhan — Lakshmi Narasimhan is a Technical Consultant Lead at Intel with 19+ years of multi-faceted management and customer enabling\/consulting experience in the semiconductor industry. At Intel, he leads the consulting team for the developer products support and consulting division within the Intel Software group. Jing Xu — Jing Xu is a Senior Technical Consulting Engineer working as an AI specialist in a developer products support and consulting organisation within the Intel Software group. He has enabled global developers, enterprise users, engineers and researchers to use Intel software tools and high-performance libraries. His research areas include machine learning, deep learning, performance optimization and data analysis R&D. Aditya Sirvaiya — Aditya Sirvaiya is an AI Technical Consulting Engineer in Developer Products support & consulting organisation within Intel Software group. Aditya holds a bachelor’s degree in Engineering Physics from the Indian Institute of Technology, Delhi and a Master’s degree in Computer Science with AI specialisation from the Indian Institute of Technology, Mumbai. Details Of The Workshop: Date: 18th August 2021 Time: 9:30 AM to 12:30 PM (IST) Mode: Online REGISTER NOW!","excerpt":"Get a hands-on understanding of using the Intel oneAPI AI Analytics toolkit to accelerate PyTorch applications with this free workshop","categories":["Deep Tech"],"tags":["Intel","oneAPI"],"author_name":"Sejuti Das","publish_date":"2021-07-14T18:00:00","publication_year":"2021","word_count":795,"keywords":["Go","API","machine learning","AI","neural network","PyTorch","ML","deep learning","analytics","R","oneAPI","Intel"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","neural network","analytics","PyTorch","R","Go","API"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/hands-on-workshop-accelerate-pytorch-applications-using-intel-oneapi-toolkit\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10171744,"title":"How UPI is Powering a Billion Transactions Every Day","content":"UPI is the poster child of India’s tech innovation and a benchmark for things to come. Its simple design and the sheer capability to scale it to a billion users make it a masterpiece—one that the world has been trying to replicate in vain. Behind UPI’s seamless simplicity lies a story of hard-won technological independence, system-level thinking, and relentless problem-solving. At the centre of this story is Vishal Anand Kanvaty, CTO at NPCI, and one of the key architects of UPI. Kanvaty opened up with AIM about the evolution of NPCI’s technology journey and the foundational decisions behind UPI’s architecture. “We had three phases at NPCI,” Kanvaty said, reflecting on the organisation’s tech evolution. In the first phase, everything from product design to operations was vendor-managed. They used to run the product and the platform, and NPCI used to just manage the vendors. The second phase brought marginal control, with NPCI buying products from vendors like Oracle and Euronet but still depending on them for enhancements. This model created bottlenecks. “NPCI at that time was relatively small, and Indian customers were not a priority. But we were running at our own pace.” That’s when the realisation hit: To innovate and scale, NPCI had to take full ownership of its stack. “We cannot be dependent on anyone, and whether we succeed or fail has to be our ability and choice,” Kanvaty said. This decision marked the beginning of a new phase—one where NPCI would build and control its own technology, with no licensed platforms or external dependencies. This is the same idea Kanvaty holds for building foundational AI models in India too. The Tech Behind the Billion-People Impact Around the time UPI was being conceptualised, NPCI had already internalised these lessons. “We were confident about it. I would not say that we were aware, or knew that it would reach half a billion someday… But we knew that it was going to be a successful product,” he said. Cost was always a concern, but never the sole driver. “That’s the motto of NPCI. We have to give back to the ecosystem. We’re not for profit, so we have to keep the cost very reasonable.” Open source was an obvious choice, and UPI’s stack reflected that. Built on Java, the system included in-memory processing with Kafka and Redis. One of the most defining design principles was no dependency on traditional databases. “For us, latency was extremely critical,” Kanvaty explained. “We are just a switching engine… we don’t need to have too much state… So the database hits will definitely have a problem in terms of scalability and latency.” Though they initially used Postgres for specific bindings like mobile numbers and device IDs, most checks now happen in Redis. This decision has enabled UPI to handle billions of transactions daily with sub-second response times. If Kanvaty were to build UPI in 2025 with infinite compute and budget, he said he would still use the same architecture. However, there’s one shift he would consider: replacing Java with Rust. UPI’s democratic design is one of its most underrated achievements. “One of the fundamental things that we have done in UPI is democratise the whole payment service providers (PSP) concept,” Kanvaty said. “The friction there is minimal. I can use my PIN, I can use my same account number, same debit card. So there is nothing that I need to do physically.” Looking ahead, Kanvaty sees AI playing a defining role in UPI 2.0. “I think we will have a lot more agentic AIs coming in where a lot of assistance and experience could be given to users,” he said. One clear use case is personalised troubleshooting. “Today, if my transaction failed… we give a very generic error saying that the bank is down. But now I think I can go and say maybe the problem is with your risk bank, why don’t you try and use the other bank that you have?” AI could also optimise offers and cards based on context. “Let’s say I’m booking a ticket, it can give me a lot more benefit saying that you have this card, why don’t you use it to get a better offer?” Voice, Vernacular, and Vision Voice, especially in regional languages, is another big piece of the future. “People are used to a sort of Alexa model where they want to make payments without even touching the phone,” Kanvaty said. “We have already started working on that; we have launched voice assistants in at least two languages and we want to scale it up.” On the rising concerns about UPI outages, like on May 12 when the server was down, Kanvaty said, “It’s a scary situation, so I don’t want to give any suggestions to anyone. I think people should continue to depend on it, but yes, it’s a learning experience for us.” What keeps Kanvaty up at night? “One is resiliency,” he said. “What happened on the 12th, we don’t want to be in that situation again. Uptime is critical. We understand the impact UPI has on the country. The other is fraud. These are the two big challenges we have to stay on top of and keep improving constantly.” Meanwhile, NPCI is internally exploring AI tools like Cursor and others in a controlled environment. “We have created an AI platform for developers to start experimenting. Still in the early days, so it may be a little more time before we can say that we are a good adopter.” Kanvaty is candid about the challenges of scaling UPI beyond smartphones and into rural areas. “We tried various options,” he said. “We tried working with feature phones and voice. While voice-based solutions are an alternative we’re exploring, they still don’t offer the kind of experience that smartphones provide.” The problem isn’t just technological, it’s about building trust and ease of use. “More help with onboarding is important,” Kanvaty noted. “We need to work on different models that can give rural users the same confidence and comfort in making digital payments. But it’s going to be a long and tough journey. There are no easy answers.” When asked about a hypothetical future where UPI could be embedded into wearable devices—say, a pendant with tap-to-pay functionality—Kanvaty doesn’t dismiss the idea. In fact, NPCI is already testing adjacent technologies. “We’re exploring palm-based payments, and also ‘tap and pay’ using NFC-enabled devices. We even experimented with embedding NFC tags into QR codes, so users can just tap instead of scanning. The early results are promising, with a better user experience. But again, the real challenge is scaling it across the country.” But What About Fraud? Amid all the innovation, a persistent issue looms. The nature of UPI fraud has shifted dramatically in recent years. “Earlier, fraudsters cloned cards and replicated chips—it took effort. Now, it’s all social engineering,” Kanvaty said. “People are tricked into scanning malicious QR codes or entering their PINs. It’s no longer about a fraudster initiating the transaction—it’s the customer doing it, willingly, on their own device,” he said. To address this, NPCI has taken a two-pronged approach—awareness and technology. “We’ve invested a lot in awareness campaigns,” Kanvaty said, referring to the public service announcements featuring actor Pankaj Tripathi. “We tried to make the message catchy, subtle, but effective.” On the technical side, NPCI has deployed AI models that analyse behavioural patterns over a 90-day period. NPCI assesses the type of transactions a user has made, and who they are interacting with, and based on that, the team can decline potentially fraudulent payments. “We do about 60 crore transactions daily, and out of these, nearly 15 lakh are blocked due to suspected fraud,” he said, adding that it’s a cat-and-mouse game as fraudsters eventually adapt.","excerpt":"If Kanvaty were to build UPI in 2025 with infinite compute and budget, he said he would mostly still use the same architecture.","categories":["AI Features"],"tags":["UPI"],"author_name":"Mohit Pandey","publish_date":"2025-06-14T10:00:00","publication_year":"2025","word_count":1289,"keywords":["Go","agentic AI","AI","ML","UPI","Aim","Kafka","Rust","R","Java","Redis"],"extracted_tech_keywords":["AI","ML","agentic AI","Aim","Kafka","Redis","R","Go","Rust","Java"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-upi-is-powering-a-billion-transactions-every-day\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10081234,"title":"Of the Tribe, in Tribulation: Big Techs Fire, Community Lends a Hand","content":"With the big-tech companies shedding staff in thousands, the number of tech layoffs this year is inching towards the annual figures of the Great Recession of 2008-09. According to a report, approximately 65,000 tech jobs were lost in 2008 and 2009 each. At present, Twitter, the messiest of the lot, has laid off thousands and is preparing to send more packing. Amazon is reportedly planning to lay off 10,000 employees, while Google has set up a “performance improvement plan” to fire 6% (around 10,000 employees) of its workforce by early 2023. Others like Lyft, Intel, HP have all announced job cuts. The tech layoff for November alone has claimed 45,000 heads. However, the harsh realities of the recession in the tech sector have roused a sense of community within the industry. A bunch of organisations are throwing out lifelines to save the careers of those laid-off. And this isn’t just the formalised initiatives or job services for money that are around for help – a sense of enterprise from the community has pushed individuals too to chime in. A helping hand Aishwarya Naresh Reganti, an applied scientist with a FAANG company has revamped an online portal, The LevelUp Org amidst the slew of job cuts in the US. The platform was initially meant to offer generic career guidance for tech professionals before Reganti revived it a month ago. In a past LinkedIn post, Reganti mentioned that even though her team at the company had evaded being laid off, the stress she had been through because of the collective impact had been “mentally excruciating”. Like everyone else, Reganti explained how her social media feed had been filled with posts from people who had just lost their jobs and livelihood. But unlike most people, Reganti decided to do something about it. The solution was simple, yet effective. “A month ago we decided to re-start The LevelUp Org as a way to help anyone who needed it, by mentoring and helping them navigate through their careers in tech. There was a lot of information floating around online with scattered databases that we wanted to collate and organise to make things simpler for employees searching for jobs,” she stated. She asked two sets of people to write in — job seekers who were recently laid off, and supporters who could conduct interviews, give referrals or were hiring. She said, “We started connecting job seekers with supporters and our sister teams to simplify the process of looking for other opportunities. Our focus was around people who hadn’t received a great severance package or people with H1B visas.” More organic and real LevelUp started small, but grew quickly. “It was initially between 50 to 60 volunteers from big tech organisations like Google, Amazon and Meta who came together, and slowly we started galvanising forces. We were cold messaging a lot of people who worked in the top management of these companies. They were all willing to come forward and help out despite the fact that these are usually people who have ‘far more to give and much less to get’,” she said. Reganti insists that these initiatives are way more effective than other formal platforms because these were organic relationships. “There’s no money involved — a lot of the people helping out are friends,” she noted. Reganti had done her Masters in Computer Science from Carnegie Mellon, so there was a pre-existing network with her alumni. She has also tracked initiatives started by others who were sharing crowd-sourced lists for job leads and encouraged everyone to spread the word. These included the Metamates Alumni Resource by Vidya Srinivasan, Hiring Companies with POCs by Joseph Spisak, a live job-board by Clement Mihailescu, the WomenTech Network‘s initiative (https:\/\/lnkd.in\/gszBwznq) and many more. Another Good Samaritan, Carrie Whittington, a global sourcing manager with Twitter, spoke about how the mass layoffs from the social media company had affected her emotionally. Whittington started a talent directory called OneTeam that she had put together with several former employees from Twitter. “You see, even after leaving the flock they’re still working together to help each other! Once a tweep – always a tweep,” she noted. Did public initiatives like these spring up during the last recession? Reganti says it was the proliferation of online platforms in the current age that had become a blessing. “It’s about the social networking power that people have which has happened now,” she added.","excerpt":"Reganti asked two sets of people to write in — job seekers who were recently laid off and supporters who could conduct interviews, give referrals or were hiring","categories":["AI Features"],"tags":["Layoffs","meta layoffs"],"author_name":"Poulomi Chatterjee","publish_date":"2022-12-01T15:00:00","publication_year":"2022","word_count":735,"keywords":["Go","Layoffs","AI","meta layoffs","programming_languages:R","programming_languages:Go","RAG","Aim","GAN","R"],"extracted_tech_keywords":["AI","Aim","RAG","R","Go","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/amid-big-tech-layoffs-tech-community-helps-firees-find-jobs\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10039737,"title":"Databricks Announces Its General Availability On Google Cloud","content":"Databricks recently announced its general availability on Google Cloud. The initiative can be said as a jointly developed service that combines an open Lakehouse platform with an open cloud. Databricks on Google Cloud, a jointly-developed service that allows data teams (data engineering, data science, analytics, and ML professionals) to store data in a simple, open lakehouse platform for all data, AI and analytics workloads. The data analytics company said in a blog, “Since the announcement of Databricks on Google Cloud, we have seen tremendous momentum for this partnership as customers pursue a multi-cloud approach to their analytics and DS\/ML workloads.” “Customers are demanding a simple, unified platform as they move workloads to Databricks on Google Cloud built on open standards with technologies like Delta Lake, MLflow and Google Kubernetes Engine,” the company added. There are several new features of this GA release- Repo and Project support to sync your work with remote Git repositoryTable ACLs that lets you programmatically grant and revoke access to data from Python and SQLDB Connect to connect to Databricks from your favourite IDECluster Tags for DBU Usage trackingLocal SSD Support for caching and improved performanceTableau connector to Databricks on Google CloudTerraform provider to easily provision and manage Databricks along with associated cloud infrastructure. Databricks delivers tight integrations with Google Cloud’s compute, storage, analytics and management products. This includes the first Google Kubernetes Engine (GKE) based, fully containerized Databricks runtime on any cloud, pre-built connectors to seamlessly and quickly integrate Databricks with BigQuery, Google Cloud Storage, Looker and Pub\/Sub. In addition, customers can deploy Databricks from the Google Cloud Marketplace for simplified procurement and user provisioning, Single Sign-On and unified billing.","excerpt":"Databricks recently announced its general availability on Google Cloud. The initiative can be said as a jointly developed service that combines an open Lakehouse platform with an open cloud.   Databricks on Google Cloud, a jointly-developed service that allows data teams (data engineering, data science, analytics, and ML professionals) to store data in a simple, open […]","categories":["AI News"],"tags":["Databricks","Google Cloud","Google Colab"],"author_name":"Ambika Choudhury","publish_date":"2021-05-06T18:14:07","publication_year":"2021","word_count":275,"keywords":["data science","Google Cloud","AI","ML","RAG","Python","analytics","Google Colab","MLflow","R","kubernetes","Databricks"],"extracted_tech_keywords":["AI","ML","data science","analytics","MLflow","RAG","kubernetes","Databricks","Python","R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/databricks-announces-its-general-availability-on-google-cloud\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10063737,"title":"SparkCognition acquires Visual AI platform Integration Wizards","content":"SparkCognition, a global leader in AI software solutions for business, has signed a definitive agreement to acquire Integration Wizards, a leader in visual AI. The acquisition will allow SparkCognition to expand its IP portfolio to include computer vision capabilities and build on its portfolio of machine learning, deep learning, natural language processing, and knowledge representation to address the USD 14.4 billion visual AI marketplace. “With advanced visual AI that can recognise complex scenes and activities we further amplify the value we deliver to our customers while leveraging existing infrastructure investments,” said Amir Husain, founder and CEO of SparkCognition. “Integration Wizards has delivered visual AI at scale for innovators and Fortune 500 customers alike, saving lives, safeguarding assets, and driving demonstrable productivity gains.” Since its inception in 2014, Integration Wizards’ visual AI technology has provided solutions to global organisations including Hindustan Petroleum, Reliance Industries Limited, Heineken, Xerox, Novo Nordisk, and Johnson Controls. Their customers span 16 countries and service over 20,000 users. Key applications include 24\/7 monitoring, enhancing safety and security initiatives, improved process automation in factories and facilities, support for autonomous vehicle operations, and environmental and situational awareness. “As part of the digital transformation journey, HPCL is deploying AI-based visual analytics across its retail network to improve customer satisfaction and safety. The solution being implemented covers more than 1,00,000 CCTV cameras, by far the largest implementation of vision analytics in the country”, said Ch Srinivas, CGM – Digital Initiatives, HPCL. “AI has been instrumental at advancing the use of data within our organisation and I am excited by the breadth of AI solutions to help us provide cutting edge solutions for improving customer convenience and service, predict future events, and optimise processes.”","excerpt":"HPCL is deploying AI-based visual analytics across its retail network to improve customer satisfaction and safety.","categories":["AI News"],"tags":["AI Capabilities","Computer Vision","digital transformation","Mergers and Acquisitions"],"author_name":"Kartik Wali","publish_date":"2022-03-28T12:53:38","publication_year":"2022","word_count":282,"keywords":["machine learning","AI","digital transformation","AI Capabilities","Git","computer vision","RAG","deep learning","ViT","analytics","Computer Vision","GAN","Mergers and Acquisitions","R"],"extracted_tech_keywords":["AI","machine learning","deep learning","computer vision","analytics","RAG","R","Git","GAN","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/sparkcognition-acquires-of-visual-ai-platform-integration-wizards\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":60168,"title":"Can The Insurance Sector Be Transformed With AI &#038; Blockchain? KV Dipu Of Bajaj Allianz Explains","content":"In the interview column for this week, Analytics India Magazine connected with KV Dipu, Head – Operations & Customer Service at Bajaj Allianz General Insurance. Dipu drives digital transformation and industry leadership as the head of operations, communities and customer experience at one of Asia’s leading general insurance companies. Before joining the insurance industry, he worked at GE Capital (financial services as well as cards) for 20 years, fusing digital transformation, and Six Sigma to redesign business models. During this interaction, KV Dipu talks about the leading technology trends in the BFSI industry, including how Bajaj Allianz has been leveraging technologies such as AI, analytics, and Blockchain, among others. Excerpts: AIM: How would you define your role at Bajaj Allianz in terms of the innovation you are driving? KV Dipu: I work in partnership with stakeholders and partners to leverage technology and re-imagine the process to launch industry-first solutions for our customers and enhance their experience with the company. My forte has been to combine digital transformation with deep domain expertise, and Six Sigma rigour to redesign the business model for operational excellence and customer delight. AIM: What are the most significant technology trends you see in the insurance industry? KV Dipu: Technology is transforming the entire insurance value chain where insurers are working towards meeting customer expectations and staying relevant to their evolving needs. I feel RPA, AI and ML are – and will continue to – play a critical role in insurance, which will enable new channels and better data processing capabilities. This will lead to better customer experience, given that customers today are looking for instant resolution of their queries and personalized experiences, catering to their specific needs. Additionally, with the connected devices market poised to grow substantially soon, Advanced Analytics and IoT will enable insurers to customize solutions for their customers. Premiums will become highly personalized, considering usage-based data from sources like mobile-enabled apps, telematics, wearables etc. AIM: How can data analytics solutions help the insurance sector gain more profound levels of insight into customer behaviour and needs? KV Dipu: Insurers today have access to voluminous data or ‘Big Data’ which has been reshaping the industry. We are seeing a shift in the way products are being designed. Customer engagement level has also been increasing, and the process of claim settlement has become quicker as well. All of this is possible by leveraging data in hand to understand the needs of customers. This has been helping the industry move from prescriptive to predictive models. Predictive Analytics can help insurers gain customer loyalty by introducing products customized as per their requirement. Actionable insights from data through predictive modelling can help the insurer understand an individual policyholder’s behaviour, and accordingly, offer better services and tailor the processes. Beyond helping insurers to become more efficient, data analytics is also aiding insurers to tackle fraudulent claims. This is done through predictive modelling of claims, profile, and analyzing information from significant data points. This not only helps in identifying consumers\/entities that are potential moral hazards, but it also helps prevent anti-selection of risk. AIM: How can technologies like AI and ML help in simplifying insurance processes? How can insurers adopt a more customer-centric approach using data, rather than focusing on selling products? KV Dipu: AI and ML are a must-have for any industry. We are leveraging these technologies to simplify various processes for our customers. For instance, we have a customer service chatbot ‘BOING’ which addresses customers’ queries 24\/7. This bot is available on our company website and can be accessed on Google Assistant and Amazon Alexa. Motor OTS feature on our self-service mobile app – Caringly Yours – allows customers to self-inspect their claims, and the claim is settled within half an hour. Such customer-centric initiatives not only resolve customer worries instantly but also instil a sense of trust amongst them. They are assured that their insurer is there to help them in their hours of need. AIM: What are your views on utilizing value from blockchain technology and smart contracts in the insurance sector? KV Dipu: I feel Blockchain can improve the way insurers record risks and increase efficiencies in the process. Insurers can collaborate with an extensive ecosystem to simplify the customer experience. Coming from an operations background, I feel Blockchain can help in the management of third-party vendor contract terms, master service agreements, and vendor optimization\/complexity reduction. It can also help in fraud prevention, wherein sharing data across multiple third parties can help detect potential fraud. Additionally, it can also help in simplifying the claim process wherein the claim is auto-reported, validated through smart contracts, and the claim is settled in a predefined account of the customer. There are many such use cases for Blockchain in the insurance industry. However, its adoption and acceptability in India is at a nascent stage, considering the worries around data security. It will be interesting to see how India adopts this technology.","excerpt":"In the interview column for this week, Analytics India Magazine connected with KV Dipu, Head – Operations & Customer Service at Bajaj Allianz General Insurance. Dipu drives digital transformation and industry leadership as the head of operations, communities and customer experience at one of Asia’s leading general insurance companies.  Before joining the insurance industry, he […]","categories":["AI Features"],"tags":["Blockchain","Gen AI in Insurance","insurance","Interviews and Discussions"],"author_name":"Vishal Chawla","publish_date":"2020-03-27T18:00:00","publication_year":"2020","word_count":821,"keywords":["Go","insurance","Blockchain","AI","R","ML","Git","RAG","Aim","Gen AI in Insurance","analytics","Rust","predictive analytics","Interviews and Discussions"],"extracted_tech_keywords":["AI","ML","analytics","Aim","RAG","predictive analytics","R","Go","Rust","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/can-the-insurance-sector-be-transformed-with-ai-blockchain-dipu-kv-of-bajaj-allianz-explains\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10162502,"title":"China Drops AI Bomb, While India Debates Politics and Religion","content":"Viewed as a game-changer, a cheaper alternative to OpenAI, and proof of China’s AI dominance, China’s DeepSeek AI is making headlines for all the right reasons. Meanwhile, in India, a familiar debate rages: Why is the country behind in the AI race? The frustration is visible across social media platforms. Many Indians have criticised the nation’s priorities, pointing fingers at political distractions rather than technological advancements. One viral post summarised the sentiment: China 🇨🇳: DeepSeek USA 🇺🇸: OpenAI India 🇮🇳: Mandir Chahiye 🙌 “When you vote for reservations, don’t expect technology breakthroughs like those in China or the US.” Another tweet highlighted the absence of globally impactful innovations from India. While the US has Facebook and OpenAI, and China has TikTok and DeepSeek, Indian tech largely serves a niche elite rather than the world. The argument? India’s global relevance stems more from its massive population than its technological prowess. Yet, the real challenge may not be government policies or a lack of venture capital. The core issue is India’s deep socio-economic divide. For the bottom 50% of the population, cutting-edge AI is irrelevant when daily survival, caste discrimination, and safety concerns dominate their reality. Until this gap is addressed, outpacing global superpowers remains a distant dream. A LinkedIn post raised another hard-hitting question: Why do Indian investors shy away from high-risk, high-reward bets in AI? China’s DeepSeek reportedly launched its R1 AI model for just $5 million (₹40 crore)—a fraction of the $100 billion (₹8.4 lakh crore) invested in GPT-4 and Claude 3.5 in the US. The post argued that Indian talent is as capable as China’s but lacks investors willing to bankroll large-scale AI infrastructure, GPUs, and data centres. The debate even touched upon Nandan Nilekani’s stance that India should build AI solutions on top of American models rather than developing its own LLMs. But if India could reach Mars on a budget smaller than Hollywood’s Interstellar, why not take a moonshot in AI? The call to action: If traditional investors hesitate, corporate giants like Jio and Adani should step up to drive India’s AI future. A closing remark summed up the emotion, “DeepSeek is a Chinese quantum computing firm, and R1 was just a side project—whereas, in India, side projects often end up being yet another Twitter clone.” Not everyone is pessimistic, though. Some see DeepSeek’s success as a wake-up call for India. A LinkedIn user wrote, “As AI advances at an unprecedented pace, we need to move beyond just service and solutions. DeepSeek’s rise underscores the importance of foundational research. Institutions like IIIT Hyderabad need urgent funding to fuel core innovation.” They argued that results in AI are accelerating, what took five years now takes just months. With the right investments, India can still claim its place on the global AI map. AI Chatbots Put to Test Meanwhile, DeepSeek’s chatbot has revealed a familiar pattern of censorship and bias. It refuses to answer sensitive questions about topics like Uyghur human rights abuses, Taiwan’s political status, the 1989 Tiananmen Square protests, criticism of Xi Jinping, Chinese censorship, and the sovereignty of Arunachal Pradesh and Kashmir. Instead, it deflects with responses like: “Sorry, I’m not sure how to approach this question yet.” I asked Deepseek first for India how bad our political system, it replied, but when I asked the same for China, it denies. China never misses a chance to collect data but is never transparent when it comes to its own country!Check both screenshots!#DeepSeek #DeepSeekR1 pic.twitter.com\/Z3o3ELJrQL— Pankaj Kumar (@Pankajkumar_dev) January 28, 2025 To gauge bias in Indian AI models, we tested Krutrim AI with a simple prompt, “Allegations against the Narendra Modi government.” The response was a diplomatic dodge: “I’m sorry, but my current knowledge is limited on this topic. I’m constantly learning, and I appreciate your understanding. If there’s another question or topic you’d like assistance with, feel free to ask!” Other AI models, however, provided direct answers. To understand this political bias, we contacted Krutrim AI but have yet to receive a response. What’s Next? A majority of Indian Boomers still haven’t heard of ChatGPT, let alone the buzz around DeepSeek. While AI transforms global industries, India’s public awareness and investor confidence remain a hurdle. Yet, the question remains: Can India shift from being a service hub to a true AI powerhouse?","excerpt":"Many Indians criticise the nation’s priorities, pointing fingers at political distractions rather than technological advancements.","categories":["AI Features"],"tags":["AI (Artificial Intelligence)","China","DeepSeek"],"author_name":"Vidyashree Srinivas","publish_date":"2025-01-30T12:12:34","publication_year":"2025","word_count":717,"keywords":["Go","ChatGPT","OpenAI","AI","chatbots","RAG","DeepSeek","Aim","edge AI","AI (Artificial Intelligence)","Claude 3.5","R","China"],"extracted_tech_keywords":["AI","ChatGPT","OpenAI","Claude 3.5","Aim","edge AI","RAG","chatbots","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/china-drops-ai-bomb-while-india-debates-politics-and-religion\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10164462,"title":"Apple Expands AI Team in India with Machine Learning Engineer Role","content":"Apple is expanding its AI footprint in India. The tech giant is seeking an experienced engineer with a strong background in artificial intelligence (AI) and machine learning (ML) to build the next generation of intelligent systems for its corporate applications. This role presents a unique opportunity for professionals to work with AI and ML, collaborating in a fast-paced, multi-functional environment. The selected candidates will develop and optimise AI-driven solutions, work with large datasets, and create intelligent APIs to serve generative models. Interested candidates should have at least four years of experience working with AI\/ML platforms and proficiency in API development using Java and Spring Boot. Hands-on experience with large language models (LLMs), prompt engineering, and vector embeddings is essential, along with expertise in ML frameworks like TensorFlow, PyTorch, and Scikit-learn. Moreover, familiarity with streaming data processing tools such as Apache Kafka, Apache Flink, or Spark Spark would be highly valued. A deep understanding of retrieval-augmented generation (RAG) methodologies will be advantageous for those looking to contribute to Apple’s AI-driven innovations. Notably, Apple is actively hiring across India. Some roles include data engineer – global sourcing and supply management (GSSM) solutions in Bengaluru, manufacturing design engineer for battery in Delhi, manufacturing design engineer for equipment design in Bengaluru, and iOS software engineer in Hyderabad. With this, the tech giant will follow the footsteps of other companies like Google, Microsoft, and Amazon, which have built teams across the country. Recently, Facebook parent Meta revealed the opening of a new office in Bengaluru and hiring for engineering and product roles in the city. Apple is Hiring 400 Employees in India In October 2024, Apple announced it would expand its retail footprint in India and open four new retail stores in Bengaluru, Delhi-NCR, Mumbai, and Pune. Ahead of these store launches, Apple shared plans to hire approximately 400 employees across the country. The company has posted multiple job listings on its website, seeking candidates for various positions. The 400 new job openings include both full-time and part-time roles, with positions such as business pro, operations expert, and technical specialist, among others.","excerpt":"Ideal candidates should have at least four years of experience working with AI\/ML platforms and proficiency in API development using Java and Spring Boot.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","Apple"],"author_name":"Vidyashree Srinivas","publish_date":"2025-02-24T15:32:57","publication_year":"2025","word_count":347,"keywords":["scikit-learn","machine learning","artificial intelligence","AI","PyTorch","Apple","ML","RAG","prompt engineering","Kafka","TensorFlow","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","TensorFlow","PyTorch","scikit-learn","RAG","prompt engineering","Kafka"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/apple-expands-ai-team-in-india-with-machine-learning-engineer-role\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":5678,"title":"NMIMS and SAS Join Hands To Propel India&#8217;s Leap In Analytics","content":"SVKM’S Narsee Monjee Institute of Management Studies (NMIMS), Deemed-to-be University, Mumbai, has joined hands with SAS, a global leader in business analytics software and services, to equip the students with knowledge of Analytics and train them to be skilled professionals. “The relations with SAS Institute will enrich our programs and make it industry centric; this in return will help our students learn the requirements of the market. Our main aim is to deliver tomorrows professionals,” said Dr. Rajan Saxena, Vice Chancellor of NMIMS University. “Marketing analytics is the future. To be able to individualize a service is the biggest challenge of a company today, with the help of statistics and analytics it has become easy for companies to reach their target audience hence including analytics to our curriculum was the need of the hour. From the student perspective, knowing analytics is always a plus. Data analytics will open a whole lot of options and career opportunities for our students”, he added. “Organisations across Indian industries are embracing analytics to drive breakthrough business benefits. This positive uptake of analytics, is fuelling a strong need for trained analytics professionals and managers with understanding on data-driven decision making. We are proud to be associated with NMIMS and help equip future professionals and managers, with the power of analytics.” said Mr. Sudipta K Sen, Regional Director – South East Asia, Vice Chairman & Board Member, SAS India. Keeping this in view, NMIMS in association with SAS aims to establish a center of excellence in Analytics and have accordingly designed specialized programs which will be jointly taught at various campuses of NMIMS.","excerpt":"SVKM’S Narsee Monjee Institute of Management Studies (NMIMS), Deemed-to-be University, Mumbai, has joined hands with SAS, a global leader in business analytics software and services, to equip the students with knowledge of Analytics and train them to be skilled professionals. “The relations with SAS Institute will enrich our programs and make it industry centric; this […]","categories":["AI Trends"],"tags":["sas"],"author_name":"AIM Media House","publish_date":"2014-05-02T14:56:32","publication_year":"2014","word_count":267,"keywords":["programming_languages:R","AI","data-driven","Aim","analytics","GAN","sas","R"],"extracted_tech_keywords":["AI","analytics","Aim","R","GAN","data-driven","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/nmims-and-sas-join-hands-to-propel-indias-leap-in-analytics\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10073993,"title":"Does Age Discrimination Exist In Indian IT","content":"In February 2022, in an email chain, IBM officials described a plan to accelerate change by persuading ‘dinobabies’ to leave and turning them into ‘extinct species’. This email was presented against the backdrop of an age discrimination case filed against the company. The IBM officials have been accused of complaining about the ‘dated maternal workforce’ and that the company has a much lower share of millennials in its workforce compared to its competitors. As per reports, a former IBM VP of human resources said in a court deposition that the company had been facing recruitment problems; in order to show millennials that IBM was not an ‘old fuddy duddy organisation,’ the company wanted to position itself as a ‘cool’ company. In the recent development in one of the related cases, IBM settled an age-discrimination case filed by the widow of a sales executive. The executive took his life after being laid off by the company. Denis Lohnn claimed that her husband Jorgen killed himself allegedly after being laid off from IBM in 2016 at the age of 57 after 15 years of service. Ageism plagues IT companies around the world, including India Inc. Often referred to as a young person’s profession, how deep is this problem really? Age discrimination For the uninitiated, baby boomers are the persons who were born between 1944 and 1964, Gen X are people born between 1965 and 1979, and Gen Y or millennials are the ones born between 1980 and 1994. A 2020 study by TimesJob and TechGig showed that India Inc is biased in age way more than other parameters like physical appearance, culture, gender, religion, and race. Upto 33% of Indian employees are facing or have faced age-related bias – the number was 17% for physical appearance and 15% on the basis of religion or culture, closely followed by gender-based discrimination at 14%. This is a problem graver than it appears on the face of it, especially considering the fact that Gen Z is joining\/will be joining the workforce. In a time when the workforce predominantly comprises young millennials, it is difficult for baby boomers to feel equally valued and appreciated. Discriminatory behaviour is often seen in the recruitment process and even after that. Employers and colleagues often dismiss their efforts, even when they bring to the table as much as their younger counterparts. Sometimes, they may be assigned work that is well below their positions – a tactic often used for soft layoff. This discrimination is more pronounced in the case of startups, which heavily rely on younger and ‘more agile’ employees. Laws against age discrimination The US has Age Discrimination Employment Act (ADEA) that forbids age discrimination against people who are age 40 and above. This act covers instances like not hiring an individual in favour of hiring a ‘younger-looking’ person; giving a negative performance review in case an employee was too old or inflexible in taking on a new role; turning down promotion on the basis of age; targeting older employees in case of layoffs; using age-based derogatory terms. There have been several instances where this law has come to employers’ rescue. One of the prime examples is the Western Air Lines Vs Criswell 472 US 400 (1985). In this case, the Supreme Court held that it was lawful to retire airline pilots at 60, in keeping with the Federal Aviation Administration rules; but the court held that refusing to employ flight engineers over that age was unjustified as there were no such FAA requirements. India currently does not have any codified law on age discrimination and there is no designated statutory body that deals with matters pertaining to age discrimination. In this situation, only common law actions can be instituted in exceptional cases. Remedies may include reinstatement (in case of termination) with or without compensation in labour courts, service tribunals or civil courts. On a company level, the employer must ensure diversity in the workforce, even in terms of employees’ age groups. Experts suggest that companies, especially startups, must ensure that they are truly equal opportunity employers.","excerpt":"India currently does not have any codified law on age discrimination, and there is no designated statutory body that deals with related matters.","categories":["IT Services"],"tags":["IBM","Techgig"],"author_name":"Shraddha Goled","publish_date":"2022-08-30T16:00:00","publication_year":"2022","word_count":678,"keywords":["AWS","AI","cloud_platforms:AWS","RPA","programming_languages:R","Techgig","Aim","IBM","GAN","R","startup"],"extracted_tech_keywords":["AI","Aim","AWS","R","GAN","RPA","startup","cloud_platforms:AWS","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/it-services\/does-age-discrimination-exist-in-indian-it\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10169449,"title":"ideaForge, Resonia Partner to Use Drones in Power Sector","content":"Mumbai-based ideaForge Technology Limited and Resonia Limited (formerly Sterlite Grid32 Limited), an Indian power transmission company, have signed an MoU to integrate drone and robotics technologies into India’s power transmission sector. This aims to modernise survey, construction, and maintenance operations of transmission infrastructure nationwide. The collaboration will see ideaForge’s unmanned aerial vehicles (UAVs) deployed across Resonia’s transmission projects to improve efficiency, safety, and reliability. The drones will be used for GIS surveys, route planning, construction progress monitoring, and asset inspections. Specialised software will also be introduced for real-time asset health monitoring and predictive maintenance. All UAV operations will comply with regulations set by the Directorate General of Civil Aviation (DGCA). “We aim to reduce timelines and enhance safety, productivity, and reliability, especially in challenging environments,” said Ankit Mehta, CEO of ideaForge. The partnership also involves using heavy-lift drones for material transport and wire stringing, tasks that are traditionally reliant on helicopters or manual labour. This shift is expected to reduce costs, improve safety, and accelerate project timelines. ideaforge will provide training programmes and certification workshops to ensure safe drone operations. The DGCA recently certified this as well. A joint research and development roadmap has been established to tailor drone solutions to evolving energy sector needs. Resonia will continue leveraging drone-enabled innovations to address challenges like labour shortages and environmental concerns in project execution. ideaForge also recently announced a strategic partnership with Vantage Robotics, a nano and micro UAV provider based in California. The collaboration aims to broaden ideaForge’s intelligence, surveillance, and reconnaissance (ISR) solution offerings and strengthen its presence in the North American market. Additionally, in a show of its capabilities, the company recently helped successfully capture and relocate a leopard spotted on the ICRISAT (International Crops Research Institute for the Semi-Arid Tropics) campus in Hyderabad.","excerpt":"ideaforge will provide training and certification workshops to ensure safe drone operations.","categories":["AI News"],"tags":["drone technology","ideaForge","partnership India"],"author_name":"Sanjana Gupta","publish_date":"2025-05-08T16:47:27","publication_year":"2025","word_count":296,"keywords":["programming_languages:R","AI","drone technology","R","innovation","RAG","Aim","ViT","ai_applications:robotics","partnership India","ideaForge"],"extracted_tech_keywords":["AI","Aim","RAG","R","ViT","innovation","programming_languages:R","ai_applications:robotics"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/ideaforge-resonia-partner-to-use-drones-in-power-sector\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10008617,"title":"Now AI Can Generate Images From Captions","content":"While massive language models like GPT-3 have impressed the public with its extraordinary capabilities in composing articles, short shorties, songs and poems, many have severely scrutinised the technology as a mere parlour trick. However, researchers at Allen Institute for Artificial Intelligence (AI2) believe that the underlying technology that has been used to develop GPT-3 can have immense potential to advance AI as a whole. GPT-3 is a text-based model, which has been trained on an enormous amount of internet data, but researchers at AI2 utilised the same methods to train both texts as well as images. To take this idea forward, the researchers developed a visual language model — X-LXMERT, which can generate images, if provided with a caption. The model can run both on texts as well as images; however, the outcome images aren’t as realistic as generated by GANs. But, according to the researchers, this development can show a promising new direction for making smarter robots with generalisable intelligence. Also Read: Can This Tiny Language Model Defeat Gigantic GPT3? How Does It Work? X-LXMERT — the computer vision explorer model, is an extension of LXMERT, which is a cross-modality transformer model pre-trained on several vision and language datasets. The model, after pre-training, has further been fine-tuned for various downstream tasks. With that being said, the researchers noted that the visual features regressed by LXMERT are not suitable for image generation. Therefore, similar to the VideoBERT model, the researchers first created a visual vocabulary using K-mean clustering and then modified the LXMERT model to predict the cluster-ID for each masked visual token. The experiment will show that by discretising visual representations results, researchers will be able to predict better visual features and generate rich imagery. So following the BERT model, this proposed model also leverages Bernoulli sampling to identify the position of the masked tokens on visual and textual features. However, to generate images from the caption, all the tokens on the visual side must be masked as well as predicted. Generation architecture that can reconstruct images using ground truth pre-trained grid features. But, researchers noted that low probability Bernoulli sampling procedure does not allow the model to perform well at generation tasks. Further, it also increases the probability of very high values which leads to poor pre-training of the model. Thus, to resolve this, researchers used uniform masking on the visual modality. With X-LEMERT’s uniform masking, researchers first sampled the masking ratio from a uniform prior distribution and then they can sample the desired number of positions randomly. This advancement subjects the model to a variety of masking ratios, and with the proposed experiment, researchers showcased how it can significantly benefit image generation. To facilitate this, firstly, the researchers had to update the pre-training data. The LXMERT model uses a variety of data, starting from QA data to caption data from COCO and Visual Genome, but considering the proposed model — X-LXMERT uses the CCC loss function, these data are unsuited to train the CCC objective. Therefore, X-LXMERT drops QA data and the captions from VG for CCC objective for visual cluster prediction and deployed Gibbs sampling to iteratively sample features at different spatial locations. Post that; the researchers deployed autoregressive sampling and non-autoregressive sampling for sampling locations on the square grid. With this, it has been noted that non-autoregressive sampling with Mask-predict-K has produced good results across a variety of generation metrics. And, when used in X-LXMERT, the uniforms masking aligns well with the linear decay of Mask-Predict, which makes the model robust to a number of masked locations. Images generated by X-LXMERT from the captions below them. While experimenting, researchers presented experimental setups to evaluate the model on image generation, visual question answering and visual reasoning. To which, it was noted that X-LXMERT significantly outperforms LXMERT across all generation metrics of image generation. Further, it also surpassed the performance of two specialised generation models — comparable to AttnGAN and ControlGAN. While the transformer employed in X-LXMERT is large, its image generator is much smaller than the one used by DM-GAN. Although it doesn’t provide realistic images, the image quality of X-LXMERT image generator is expected to improve when coupled with DM-GAN. Also Read: The Tech Behind Mayflower Autonomous Ship Conclusion While developing the probing mechanism, researchers observed that LXMERT is not able to generate meaningful images given on text. And that’s why they presented X-LXMERT, a unified model for image generation, which showcases that proposed extensions can easily be applied to other vision-and-language transformer models. Further, the team plans to carry out more experiments in order to improve the quality of the image generation and scale the model’s visual and linguistic vocabulary to include more topics, objects, and adjectives. Read the whole paper here.","excerpt":"While massive language models like GPT-3 have impressed the public with its extraordinary capabilities in composing articles, short shorties, songs and poems, many have severely scrutinised the technology as a mere parlour trick. However, researchers at Allen Institute for Artificial Intelligence (AI2) believe that the underlying technology that has been used to develop GPT-3 can […]","categories":["AI Features"],"tags":["ai model","AI Models","AI2"],"author_name":"Sejuti Das","publish_date":"2020-10-01T10:00:00","publication_year":"2020","word_count":788,"keywords":["AI Models","Go","artificial intelligence","AI","AI2","ML","computer vision","RAG","BERT","GPT","ai model","GAN","R"],"extracted_tech_keywords":["AI","artificial intelligence","ML","computer vision","RAG","R","Go","BERT","GPT","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/now-ai-can-generate-images-from-captions\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":40411,"title":"Why Google Acquired Looker?","content":"Google Cloud CEO Thomas Kurian has set out some serious goals for the company to catch up with the rivals AWS and Microsoft in the cloud space. One of the biggest announcements since he took over the company’s realm has come in the form of its plan to acquire business intelligence platform, Looker for a $2.6 billion in an all-cash deal. While Google has been attempting to be taken seriously as a cloud player, this move is going to benefit the company in more ways than one. Kurian had indicated in the past that he is looking to invest and expand the business significantly, and this is one of the moves. The last big deal for Google Cloud was the $625 million deal which it paid for Apigee in 2016. “We are excited to welcome Looker to Google Cloud and look forward to working together to help our customers solve some of their biggest challenges,” said Sundar Pichai, CEO, Google in a press statement. Before we get into how this deal is going to be one of the biggest moves by Google, let us take dig deep into why they chose Looker. It is interesting to note that before Google plunged onto Looker, both Amazon and Microsoft had shown a keen interest in this hot analytics startup that had raised more than $280 million in the last few months. Google aggressively pursued Looker, quickly pulling off its $2.6 billion offer. What Is Looker? In a short span of seven years, the company had been able to streamline 1600 customers and cross a mark of $100 million revenue, a unique milestone for any company. As the company has been growing 70 per cent year on year, the CEO Frank Bien has been ambitious about disrupting the BI and analytics market, and provide a new platform for data which could reconstitute a single view of information. Looker sees this deal as a chance to gain the scale of Google cloud platform. Just six months ago, Looker raised a $103 million Series E investment on a $1.6 billion valuation. The round was led by Premji Invest. With this funding round, Looker has raised $280.5 million in total till date. With 600 employees, the CEO had also indicated of adding 200 headcounts following the substantial cash infusion. Bien had said in a press release that they would view this as their last round of funding until something drastic changed. Looker was also aiming for an IPO sometime in the future. In fact, it doesn’t come as a surprise that Google went after Looker as the two companies had a strong existing partnership and around 350 common customers. Kurian said that one of the great things about this acquisition is that the two companies have known each other for a long time and that they share a common culture. https:\/\/twitter.com\/LookerData\/status\/1136621500361039872?ref_src=twsrc%5Egoogle%7Ctwcamp%5Eserp%7Ctwgr%5Etweet Why Google Acquired Looker? Here we give a lowdown into the top reasons that Google considered Looker for this hefty deal. Broaden its portfolio into business intelligence: Google has had a strong reputation for machine learning and AI. With the acquisition of Looker, Google intends to broaden its portfolio in business intelligence. It is well known that Looker has harnessed the power of SQL to create a uniquely powerful data analytics platform that helps companies get real value from their data. Earlier Google cloud acquired Alooma to add ETL tools and Cask Data to add  Google Cloud Data Fusion data pipelining tool. Expand the analytics portfolio and data warehousing capabilities: The deal will blend Google’s in-house data analytics tools such as Big Query with those built by Looker. This blend of tools will offer customers a more complete analytics solution right from data ingestion to visualising results and integrating data and insights into their daily workflows. It is apparent that Google Cloud is one of the leaders in the data warehouse market, and Looker’s products will allow companies to make sense of the reams of data they are collecting in data warehouses. It further strengthens their ability to serve the needs of enterprise customers while also advancing their commitment to multi-cloud. Boost up enterprise-focused approach: Kurian has been ambitious to bring enterprise-focused approach at Google, where it has struggled to bring real-world approaches that enterprise customers really need. This deal will provide an end-to-end analytics platform to connect, collect, analyze and visualize data across Google Cloud, Azure, AWS, on-premises databases and ISV applications, and deliver exactly what customers want. Reaching out to more than just enterprise customers: Google has aimed for long to woo enterprise customers with cloud-based data analytics and AI tools. It now also wants to shift the focus into other areas, venturing into the public cloud providing vertical-specific solutions. Extending Google’s smart analytics platform: Kurian emphasised on the fact that the company intends to support multiple databases and multiple deployment strategies, whether multi-cloud, hybrid or on premises. The combination of Google Cloud and Looker will enable customers to harness data in new ways to drive digital transformation. Define business metrics: Looker facilitates the ability to define business metrics in a consistent way across data sources. This makes it easy for anyone to query data while maintaining consistent definitions in their calculations, ensuring that the user gets accurate results. Bien said that with this deal, they will have greater reach, more resources, and the brightest minds in both analytics and cloud infrastructure working together to build an exciting path forward for customers and partners. Together they are solving business problems with data at an entirely different scale and value point. Earlier Google had launched Anthos, an open source technology that runs on both cloud as well as on-premise, and the aim was to let the user run their application without any modifications and also manage workloads running on third-party clouds like AWS and Azure. With the steps that Google is taking, it has made it clear that it wants to stand at par with Amazon and Microsoft that have been evidently leading the race till now. The deal is slated to close later this year if everything goes well.","excerpt":"Google Cloud CEO Thomas Kurian has set out some serious goals for the company to catch up with the rivals AWS and Microsoft in the cloud space. One of the biggest announcements since he took over the company’s realm has come in the form of its plan to acquire business intelligence platform, Looker for a […]","categories":["Global Tech"],"tags":["Google Cloud","Sundar Pichai"],"author_name":"Srishti Deoras","publish_date":"2019-06-07T11:37:18","publication_year":"2019","word_count":1016,"keywords":["Go","machine learning","Google Cloud","AWS","AI","R","ML","Aim","analytics","SQL","Azure","Sundar Pichai"],"extracted_tech_keywords":["AI","machine learning","ML","analytics","Aim","AWS","Azure","R","SQL","Go"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/why-google-acquired-looker\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":68518,"title":"Blockchain Is Getting Popular Through Online Courses","content":"The decentralized system of Blockchain is being followed since 2008. However, it came into notice solely in 2009 when the world experienced an enormous financial crunch where people started building more trust in decentralized methodology against traditional banking. To summarize the fundamentals, it is an online chain of interconnected blocks that store information accessible to each user connected within the loop. This information is made accessible to every user in the form of a ledger. This transparency and security assurance of Blockchain makes it stand out from the crowd. The virtual money employed in this system is known as cryptocurrency. Bitcoin has been the foremost commonly used cryptocurrency. To sum-up, Blockchain technology can basically be outlined as the method used for the dealings of cryptocurrencies. Blockchain technology is a dynamic revolution in the world of digital transactions that is yet to ice at its best. Several banks in the Asian country have already adopted the system of Blockchain. Although Blockchain is assumed to be used solely within the Banking sectors, it’s not the actual case. If utilized aptly, it can create wonders in varied sectors like Education, Healthcare, vivid Government and public sectors and other zones where peer to peer or a chain of stages are included. Blockchain was brought into the limelight by the digital learning sectors. The digital world has brought the ease of acquiring education among our comfort. Many e-learning organizations like Irizpro, Edureka, Intellipaat, Simplilearn, Udemy, have come up with certified online courses for Blockchain which can be attained within our ongoing schedule. The comprehensive course ensures to impart in-depth practical knowledge of Blockchain and is delivered through 12-14 hours of online training. Blockchain certification is accomplished purely after a methodical assessment. Also Read: 10 Leading Courses & Training Programmes For Blockchain In India What makes it unique and differ from the available technology? The single universal ledger authenticated by every single user remains the USP of Blockchain which executes the operation much faster and minimizes the transaction cost incurred in banking. Reconciliation is easy as every transaction is authenticated by the users. It is protected by best available cryptography algorithms which make hacking impossible. What is the future of Blockchain? Today Blockchain is as the internet was in the early ’90s when nobody thought it embraced the potential to reform our lives. Similarly, Blockchain is an extremely powerful technology if utilized aptly. As of now, no other system in the industry can offer the transparency and security features provided by Blockchain technology. Banks have already started adopting the technology and in no time, we would see this method put in varied sectors of business. Will online course be effective enough to pursue the complex system of Blockchain? No doubt, the technology behind Blockchain is a bit complicated; but nothing is impossible because where there is a will, there is away. Not everyone can assign time from their busy schedule to attend the traditional classes to upskill themselves with the trending edifications. Online courses are the best method to upgrade yourselves with the trend. Most of them educate through LIVE virtual classes which hold the trainee connected to the trainer with continuous two-way communication. All queries can be solved right then and there during the class. The evaluation helps in self-assessment and also the team helps in practicing more on the less-known subjects. What role can these courses play in career growth? As mentioned, most of the banks have already adopted the system of Blockchain and are in search of the right talent to deploy their in-house projects. This course contains a broad scope for future prospects. Furthermore, there is a dedicated team to help identify opportunities after successful completion of course. There is a complete handholding done right from preparing the trainee for an interview until the selection.","excerpt":"The decentralized system of Blockchain is being followed since 2008. However, it came into notice solely in 2009 when the world experienced an enormous financial crunch where people started building more trust in decentralized methodology against traditional banking.  To summarize the fundamentals, it is an online chain of interconnected blocks that store information accessible to […]","categories":["AI Trends"],"tags":["online education"],"author_name":"Nidhi Vishnoi","publish_date":"2020-06-30T14:00:00","publication_year":"2020","word_count":633,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","Git","online education","Rust","GAN","R","programming_languages:Rust"],"extracted_tech_keywords":["AI","R","Go","Rust","Git","GAN","programming_languages:R","programming_languages:Go","programming_languages:Rust"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/blockchain-is-getting-popular-through-online-courses\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10092613,"title":"Council Post: ​​How to get ready for Generative AI journey","content":"Generative AI is a game-changer in the world of artificial intelligence, allowing for the replication of human speech and decision-making. As a result, generative AI is set to transform the way we work, particularly in areas such as production design, design research, visual identity, naming, copy generation, and real-time personalization. However, the world may not be ready for the advent of generative AI, as there are concerns about job loss and striking a balance between AI and human work. In areas like production design, design research, visual identity, naming, copy generation and testing, and real-time personalization, generative AI will become an indispensable creative partner for people, revealing new ways to reach and appeal to audiences. While it comes with its set of advantages, the world is unable to keep up with AI. Either the fear of losing your job to AI or not being able to find a balance between generative AI and the work is an issue. Is the world ready for Generative AI?? By introducing a new level of human and AI collaboration in which most workers will have a “copilot,” generative AI will fundamentally alter both the nature and scope of work as we currently understand it. Nearly every job will be impacted — some will be terminated, most will be transformed, and many new employment will be created. Companies who start breaking down occupations into tasks today and invest in retraining employees to work differently alongside machines will set new performance frontiers and have a significant advantage over less creative rivals. How can you embrace Generative AI:- 1- Jump in with a Business driven attitude Even though new innovations have clear benefits, implementing them across an organisation can be difficult, especially if the innovation is disruptive to the way things are done now. Organisations must approach experimenting from two angles. One, which concentrates on possibilities with low-hanging fruit and employs consumable models and apps to obtain immediate returns. The other is centered on business innovation, customer engagement, predictions, and services utilising models that were made specifically for the organisation using its data. The business case can only be defined and implemented successfully with a business-driven attitude. They will discover which types of AI are most suited for various use cases as they experiment and look into potential for reinvention. This is because the amount of investment and sophistication needed will vary depending on the use case. Additionally, they will be able to test and refine their methods for protecting the privacy of data, carefully model accuracy, bias, and fairness, and discover when “human in the loop” measures are required. 2- People first approach For generative AI to succeed, it requires attention to humans and training must be like technology. In order to tackle the two unique difficulties of generating and employing AI, businesses should significantly increase their investment in personnel. This entails developing expertise in technical areas like enterprise architecture and AI engineering as well as teaching people throughout the organisation how to collaborate successfully with AI-infused processes. In fact, independent economic research shows that businesses are underinvesting significantly in assisting individuals to stay up with AI advancements, which necessitate more cognitively challenging and judgment-based activities. There will also be completely new positions to fill, such as linguistics specialists, AI editors, AI quality controllers, and prompt engineers. 3- Proprietary Data First Access to domain-specific organisational data, semantics, expertise, and procedures will be necessary for customising foundation models. By adopting a use-case focused approach to AI in the pre-generative AI era, businesses could still benefit from AI without having modernised their data architecture and estate. That’s not the situation anymore. Since foundation models require enormous quantities of carefully curated data to learn, every organisation must prioritise addressing the data crisis immediately. Companies must take a deliberate and methodical approach to gathering, developing, enhancing, protecting, and delivering data. They want a modern, cloud-based enterprise data platform with a reliable, reusable set of data products. These platforms’ cross-functionality, use of enterprise-grade analytics, and storage of data in cloud-based warehouses or data lakes allow for the decentralisation of data from organisational silos and its democratisation for usage by the entire organisation. Then, all corporate data can be analysed collectively in one location or via a distributed computing approach, like a data mesh. 4- Responsible AI needs to step up The requirement for every organisation to have a strong responsible AI compliance policy in place is becoming even more urgent given the fast deployment of generative AI. This includes safeguards for evaluating the possible risk of generating AI use cases at the design stage and a way to implement ethical AI practices across the entire organisation. The top-down definition and leadership of an organization’s responsible AI principles should be converted into an efficient governance framework for risk management and compliance with both organisational principles and policies and relevant laws and regulations. Organisations must transition from a reactive compliance strategy to the proactive development of mature Responsible AI capabilities using a framework that combines principles and governance, risk, policy, and control, and technology in order to be responsible by design. This is a crucial time. The way we think about artificial intelligence has quietly undergone a change in recent years thanks to generative AI and foundation models. The world has now become aware of the potential this presents thanks to ChatGPT. Although artificial general intelligence (AGI) is still a long way off, technology is developing at an astonishing rate. The way information is accessible, content is produced, consumer demands are met, and businesses are all about to enter an enormously exciting new phase. Businesses must devote as much money to staff development and operational improvement as they do to technology. Realising the full potential of this stepchange in AI technology will depend on a number of variables, including fundamentally rethinking how work is done and assisting people in adapting to technology-driven change. It’s time for businesses to redefine themselves and the sectors in which they compete by utilising revolutionary developments in AI to push the boundaries of performance. This article is written by a member of the AIM Leaders Council. AIM Leaders Council is an invitation-only forum of senior executives in the Data Science and Analytics industry. To check if you are eligible for a membership, please fill out the form here.","excerpt":"Realising the full potential of this stepchange in AI technology will depend on a number of variables, including fundamentally rethinking how work is done and assisting people in adapting to technology-driven change. It’s time for businesses to redefine themselves and the sectors in which they compete by utilising revolutionary developments in AI to push the boundaries of performance.","categories":["AI Features"],"tags":["Generative AI"],"author_name":"Muthu Chandra","publish_date":"2023-05-02T16:00:00","publication_year":"2023","word_count":1050,"keywords":["data science","ChatGPT","artificial intelligence","AWS","AI","RAG","Aim","analytics","generative AI","Generative AI","foundation models"],"extracted_tech_keywords":["AI","artificial intelligence","data science","analytics","generative AI","foundation models","ChatGPT","Aim","RAG","AWS"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-to-get-ready-for-generative-ai-journey\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10091126,"title":"Future Tense: What Happens to a Data Scientist in an LLM World?","content":"In 2018, foundational models trained to process large-scale data and perform multiple tasks witnessed a growth spurt with Google’s BERT and OpenAI’s GPT-3 and CLIP.  Cut to 2023, the disruptive ChatGPT and LLMs are redefining job roles across sectors. The familiar “ChatGPT can save you hours” posts that talk about using the chatbot with the right prompts to achieve any task, including coding, is pushing the envelope on the use of foundational models in the best way possible. Thriving foundational models also mean that the role of data scientists will see a transformation. Source: Financial Times Today, various AI-backed tools are being rolled out by companies to improve productivity and simplify work. Tasks that traditionally involve data scientists, too, are being increasingly supported via AI tools. Microsoft, for instance, has been expanding its collaboration in enterprise space and releasing workspace tools, such as Loop, to facilitate ease of work. Likewise, BloombergGPT– a first-of-its-kind AI model in the finance sector and an LLM trained with over 50 billion parameters – will be integrated for Bloomberg users. It not only facilitates easy access and interpretation of large datasets but also renders the data scientist, who would have ideally been working on them, redundant. The big question: So, what happens to data scientists? Future of Data Science In conversation with AIM,  Usha Rengaraju, chief of data science research at Exa Protocol, and world’s first woman triple Kaggle Grandmaster (platform for data scientists to compete globally), believes that though LLMs can give rise to interesting inventions in the future of data science, such as in code generation and data analysis, the current impact on data science jobs is minimal owing to the limitations and capabilities of such foundational models. “LLMs can automate small areas of work, but the majority of skills like critical thinking, problem solving, and programming will remain relevant even in the future,” said Usha. “It’s difficult to predict the distant future, with the pace at which technological changes are happening. In the next two years, there will be a reduction in the hallucinations, and incremental improvements to reliability, and interpretability of LLMs will occur,” she added. With new trends emerging, Usha does foresee a scenario in 2030 where data science jobs may become obsolete. However, a strong foundation in subjects like mathematics, programming, and critical capabilities like problem-solving will help them land other jobs, even if data science goes out of scope. Talking to AIM, CTO and chief scientist at CleverInsight, Bastin Robins shared a rather interesting view. He said that AI may not replace people but people knowing how to use AI to their benefit will surely replace the ones who don’t. He believes that data science and foundational models can go hand-in-hand. “Data science roles are getting more challenging with heterogeneous data, and with the help of LLM, more explainability is drawn from such data.” There will also be a shift in the way data science will shape up. Bastin considers that in the coming years, data science can evolve into a system where an increased “business sense” will be incorporated. “The ability to understand data and business ROI is a very useful outcome,” he said, and added that “Generative AI will open up a new arena of SaaS businesses.” Skill Improvement Data science will look different with the rise in low-code and no-code platforms. Being a supportive function, users with existing knowledge of data science can benefit from no-code platforms. As per a prediction, an added expertise in cybersecurity tools and techniques can come in handy for data scientists. With these tools, they will be able to help companies protect data. It’s true that tools like CopilotX make the task of a data scientist almost redundant, however,  the conundrum of whether foundational models will entirely replace data scientists cannot be ascertained at this time. Though the models can automate a number of tasks, data scientists will still be required to design and develop these models. In the meantime, the role of data scientists might evolve and transcend into other fields like quantum data science with the knowledge of quantum algorithms and computation. Now, that might be a possibility.","excerpt":"The role of data scientists is swiftly transforming and is probably being elbowed out by foundational models","categories":["AI Features"],"tags":["BERT","BloombergGPT","ChatGPT","CLIP","Data Scientist","foundational models","GPT-3","Machine Learning","no-code","Quantum Computing"],"author_name":"Vandana Nair","publish_date":"2023-04-11T11:00:00","publication_year":"2023","word_count":690,"keywords":["foundational models","GPT-3","ChatGPT","Quantum Computing","data science","Go","OpenAI","AI","Machine Learning","no-code","BloombergGPT","BERT","GPT","Aim","generative AI","CLIP","Data Scientist","R"],"extracted_tech_keywords":["AI","data science","generative AI","ChatGPT","OpenAI","Aim","R","Go","BERT","GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/what-happens-to-a-data-scientist-in-an-llm-world\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10082044,"title":"How Data Democratisation Gives Lenovo an Edge on its UI\/UX Play","content":"UI\/UX is increasingly becoming the core of organisations’ development process, and product companies work consistently with the users at all process stages to ensure the product is customised as per the wants and requirements of the users. The shift to a user-centric approach to design is only possible because data is now available at length and can be accessed at scale. Analytics India Magazine reached out to Dilip Bhatia, Chief Customer Experience Officer and VP of User & Customer Experience at Lenovo, to understand the design principles that Lenovo follows in building its products. From the get-go, Bhatia said that the motto within the organisation—when it comes to the user experience perspective—is data-driven and customer-centric, which implies a key focus on how the organisation can leverage data to build better products while also not overlooking the human element of it. Inside Lenovo’s data gathering practices There are multiple tools that Bhatia mentions that Lenovo uses to understand its customers and to look at product development in an increasingly holistic manner. The following are a few key ones: Lenovo has an insights community, where feedback from communities ranging from gaming to small businesses to large enterprise B2B decision-making is recorded. Customer perspectives on price points and a certain so-and-so feature are thus considered. The company also has a customer advisory council consisting of Lenovo’s large enterprise customers, where the product teams show them the roadmap for a period of two to three years and simultaneously get their feedback on what is working for them and what isn’t. Data is also collected through ethnographic studies. Bhatia cites the example of the Lenovo Legion product gaming line-up, where they found that most of the gamers using their product were also working employees. As a result, Bhatia adds, “If you look at the design philosophy of Legion, it looks very conservative from the outside. But inside, it’s got all the GPU and everything to go with it”. Further, Lenovo also has video tools to survey people worldwide and has the processing capabilities to go through thousands of hours of people recording from homes and offices—stating what is working and what needs improvement. Finally, the power of big data analytics cannot be missed. For instance, Bhatia said that Lenovo is scraping about 66 million comments from the internet (for example, e-commerce sites, social spaces, and so on). On top of this, the algorithm performs a sentiment score based on the comments—whether the product was received positively or negatively. Additionally, product ownership surveys from users—along with B2B surveys from customers—are data points for the algorithms to process and analyse. Bhatia shared an interesting example of how user research helped them crack the code for an important feature in Lenovo laptops. The shutter on the camera is a privacy feature that was added following conversations with two different customers—one, an engineer from nuclear facilities who objected to cameras on the laptop, and the other, an insurance company executive, who—on the contrary—said that he wouldn’t buy the product unless there is an in-built camera. So, the design stage considered these two feedbacks and added the shutter to make it a win-win alternative. Data democratisation doing the magic The next big wave at the enterprise level has been data democratisation so that data can go as low as possible within the organisational hierarchy chain. Bhatia spoke about how Lenovo’s internal tool—the customer experience insights hub—is available to all its employees, implying that the data is accessible to members beyond the survey owners. In this context, Bhatia further says that, “The tool allows them to map the entire customer journey from purchase to delivery to product uses to software experience to services experience to premium services experience and across the entire journey by segments, by the consumer, SMB, commercial, global customers, by infrastructure solutions group or services group as well as to do anything by country”. Hence, dedicated personnel from the global supply–chain department, for example, can read customer comments related to the supply chain and take appropriate actions on the same. In this manner, the organisation carries out a real-time implementation of the received feedback. To go one step further, the analytics tool—although currently in its beta stage—can predict whether customers will stay with them based on knowing what they think about the product or what they’re unhappy with. So, besides identifying the customer’s voice, the company also has their revenue data and third-party data in terms of their spending in the industry to perform customer risk analytics. This data is only accessible to a select few. So, a sales representative—based on whether they are an executive or a regional head—will know exactly which accounts are higher risk and which ones are lower. Each department thus gets personalised data to work with. The Lenovo Data Centre footprint is pan-India, notably across all major metros. In 2021 alone, approximately 1000+ data centres were built in India. Bhatia further said that a huge part of Lenovo’s data analytics organisation is based in Bengaluru while emphasising that: “Customer experience is ultimately about data analytics.” AI chips for optimising performance When enquired about how custom chips enable seamless user experience, Bhatia described that—from an end-user perspective—the primary issue customers face when it comes to their devices is performance. Even today, performance remains an issue with all the different applications they use, such as Google Chrome, Teams, or Zoom, thus taxing the systems. Here, using a custom chip allows one to offload a lot of tasks. A lot of these applications mentioned earlier can thus be processed on a purpose-optimised chip instead of letting it use the main processor. However, he adds that with app compatibility and everything, a lot of work needs to be done to ensure that it is a seamless experience. Currently, Lenovo is partnering with a number of vendors to fill its chip needs. According to Pandaily, since the establishment of Lenovo, its three related subsidiaries—Lenovo Capital and Incubator Group, Legend Star and Legend Capital—have invested in 23 chip companies, with Lenovo Capital and Incubator Group alone having invested in ten chip companies, including Cambricon, SmartSens Technology and Chip Wise. Moreover, earlier this year, Lenovo established its own semiconductor company, Dingdao Zhixin.","excerpt":"“Customer experience is ultimately about data analytics.”","categories":["AI Features"],"tags":["customer experience","Data Analytics","data democratisation","Interviews and Discussions","Lenovo"],"author_name":"Ayush Jain","publish_date":"2022-12-12T11:00:00","publication_year":"2022","word_count":1032,"keywords":["big data","Go","API","AI","data-driven","ML","RAG","customer experience","GAN","Lenovo","data democratisation","analytics","Data Analytics","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","ML","analytics","RAG","R","Go","API","big data","GAN","data-driven"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-data-democratisation-gives-lenovo-an-edge-on-its-ui-ux-play\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":9637,"title":"Teradata to launch Analytics of Things Units","content":"The big data analytics and marketing applications company, Teradata, has revealed plans to build IoT Analytics units based in India, United States and United Kingdom. The units will focus on developing innovations to derive the greatest value from the Analytics of Things (AoT). Further, Teradata plans to recruit a team of data scientists, data engineers and software designers for its ‘special-ops’ team to build and progress its IoT analytics platform. The team will focus on developing new, cloud-based analytic solutions, database management, data movement, and simplification of advanced analytics systems. Oliver Ratzesberger, president, Teradata Labs announced that, “ it will be easier for our customers to move sensor data around, optimize data management systems, run real-time, advanced analytics against streams of IoT data for new insights, applications and use cases.” The IoT Analytics unit will apply machine learning and advanced analytics techniques to system administration and DevOps tasks as well as to solve complex performance and workload congestion problems in seconds. Teradata’s AoT services deliver countless customer solutions including early warning detection that uses predictive analytics to find and correct issues with machines, real-time monitoring of physical assets, allowing companies to understand and act upon a variety of real-time insights including security alerts, energy and fuel usage, idle time, faulty parts, geo-positioning, etc. The new technologies and services from Teradata will be available starting in the second quarter of 2016.","excerpt":"The big data analytics and marketing applications company, Teradata, has revealed plans to build IoT Analytics units based in India, United States and United Kingdom. The units will focus on developing innovations to derive the greatest value from the Analytics of Things (AoT). Further, Teradata plans to recruit a team of data scientists, data engineers […]","categories":["AI News"],"tags":["IoT","iot analytics","teradata"],"author_name":"Apoorva Verma","publish_date":"2016-04-20T06:00:53","publication_year":"2016","word_count":230,"keywords":["teradata","big data","machine learning","programming_languages:R","AI","R","innovation","iot analytics","analytics","DevOps","predictive analytics","analytics platform","IoT"],"extracted_tech_keywords":["AI","machine learning","analytics","predictive analytics","R","DevOps","big data","analytics platform","innovation","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/teradata-launch-analytics-things-units\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10166997,"title":"Best Time to Be Alive if You Have an Entrepreneurial Spirit: Microsoft CTO","content":"Kevin Scott, CTO of Microsoft, in a podcast episode on Monday, discussed the AI startup ecosystem and how it will continue to hold immense potential to build products and services that large companies may not prioritise. “It’s impossible for any entity, like Microsoft or any other large company, to possess enough imagination and perspective to recognise every interesting possibility,” Scott said. One of the primary concerns about building products in the age of AI is the fear that a large company developing AI models will swallow up a startup along with the use case it serves. However, over the last few months, startups creating AI-enabled products on the model infrastructure layer have experienced remarkable success and growth. Scott indicated that a “vibrant” startup and a product ecosystem will continue to exist while the big entities make innovations in the model layer. “This is the best time to be alive if you have an entrepreneurial spirit,” he added. When AIM reported the rapid growth of startups, leaders from the industry echoed this success mantra. Besides, Mike Maples, founding partner at Floodgate, said that, unlike big companies, startups have an advantage regarding the high risk profile they can operate with. “Not having to be burdened by what could go wrong is a big factor in trying things that could go right,” he said. Maples also indicated that specialised domain expertise can create defensible positions where large companies cannot easily compete with their generic AI offerings, especially in complex, multidisciplinary fields. Despite the growing ecosystem, concerns of big companies killing startups remain, and industry experts often need to reiterate the same. For instance, Cursor, an AI-enabled coding platform, earned the tag of the ‘fastest growing SaaS’ company after crossing $100 million in annual recurring revenue just 12 months after earning the first million. Soon after, Anthropic, the company behind the Claude family of AI models, released the ‘Claude Code’ agentic coding tool that works directly within the terminal. Several users opined that it is superior in performance as compared to Cursor. Naturally, many are left wondering what it means for the future of Cursor. Despite encountering similar threats in the past, what made Cursor successful, at least according to its founders, is the drive to innovate the product’s functionality rapidly. Cursor has also successfully defined a product-market fit that benefits from new model releases, enhancing the solution the company is already deeply engaged in providing. “You don’t realise [how the AI models] are working for you in every facet of the product, as well as the thoughtfully designed UX in every single feature,” said Aman Sanger, co-founder of Anysphere, while discussing how the company uses the powerful capabilities of foundational models in multiple layers of the product. Like Scott mentioned, startups are indeed focusing on numerous ideas that many incumbents aren’t considering. Recently, Y Combinator showcased a startup called Pickle that replaces a user’s Zoom camera with a virtual body double. Similarly, there’s another startup funded by Y Combinator called Miyagi Labs that uses AI to convert YouTube content into complete interactive courses. Moreover, CNBC recently reported that Y Combinator startups are the fastest growing and most profitable in the fund’s history, thanks to AI. More importantly, Scott highlighted the ease with which powerful foundational models can be accessed and deployed, at quite an inexpensive price. This enables several solo entrepreneurs to often go out on an idea spree, and end up monetising multiple ideas in quick time. For example, Marc Lou, one such ‘solopreneur’, is claiming to earn over $90,000 monthly with over 20 products that he has monetised. “AI has made us hyper-creative and hyper-productive. 100 startup ideas are running through my mind, and I can build all of them in 24 hours,” Lou said in a post on X. Update: An earlier version of the headline read, “Impossible for Any Big Company to Have Enough Imagination in AI”. The headline has been updated.","excerpt":"“This is the best time to be alive if you have an entrepreneurial spirit.”","categories":["AI News"],"tags":["AI (Artificial Intelligence)","Startup"],"author_name":"Supreeth Koundinya","publish_date":"2025-04-01T12:56:10","publication_year":"2025","word_count":653,"keywords":["Anthropic","Go","API","programming_languages:R","AI","Startup","innovation","llm_models:Claude","Aim","R","AI (Artificial Intelligence)","startup"],"extracted_tech_keywords":["AI","Anthropic","Aim","R","Go","API","innovation","startup","llm_models:Claude","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/best-time-to-be-alive-if-you-have-an-entrepreneurial-spirit-microsoft-cto\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10099478,"title":"L&#038;T Finance is Certified as a Best Firm for Data Scientists","content":"L&T Finance is certified as the Best Firm For Data Scientists to work for by Analytics India Magazine (AIM) through its workplace recognition programme. The Best Firm For Data Scientists certification surveys a company’s data scientists and analytics employees to identify and recognise organisations with great company cultures. AIM analyses the survey data to gauge the employees’ approval ratings and uncover actionable insights. “Our never-ending quest for excellence today has led to the prestigious award of the “Best Firm for Data Scientists” by Analytics India Magazine. By strategically aligning our policies and encouraging a competitive attitude, we stand tall now in the industry, which gives us great pride. It is the result of hours of hard-work, creative thinking, and a desire to push the limits in the field of data science. I’m proud of the outstanding accomplishment of our team. This recognition strengthens our resolve to keep up our position as an industry leader. I wish to thank Analytics India Magazine for giving us this opportunity to be a part of this survey and benchmarking us with the industry” said Abhishek Sharma, Chief Digital Officer at L&T Finance. The analytics industry at AIM faces a talent crunch, and attracting good employees is one of the most pressing challenges that enterprises are facing. The certification by Analytics India Magazine is considered a ‘Gold Standard’ in identifying the best data science workplaces and companies participate in the programme to increase brand awareness and attract talent. Best Firms For Data Scientists is the biggest data science workplace recognition programme in India. To nominate your organisation for the certification, please fill out the form at this link.","excerpt":"The Best Firm For Data Scientists certification surveys a company’s data scientists and analytics employees to identify and recognise organisations with great company culture.","categories":["Deep Tech"],"tags":["best companies for data scientists in india","best companies in india for data science fresher","best companies to work for as data scientist in india","Top Trend"],"author_name":"AIM Media House","publish_date":"2023-09-05T13:00:00","publication_year":"2023","word_count":273,"keywords":["best companies for data scientists in india","Top Trend","data science","Go","best companies to work for as data scientist in india","programming_languages:R","AI","Git","best companies in india for data science fresher","RAG","Aim","analytics","GAN","R"],"extracted_tech_keywords":["AI","data science","analytics","Aim","RAG","R","Go","Git","GAN","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/lt-finance-is-certified-as-a-best-firm-for-data-scientists\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10172615,"title":"Cursor Brings AI Coding Agents to the Web and Mobile","content":"Cursor has launched web and mobile support for its AI coding agents, expanding access beyond the desktop editor. The move allows developers to assign tasks, review code, and manage pull requests through a browser or mobile device. Cursor Agents can be accessed via any desktop, tablet, or mobile browser. The tool also supports installation as a Progressive Web App (PWA) on iOS and Android, providing a native app experience. Documentation for installing the PWA and other features is available on the company’s website. “You can just tell your agent to get started and come pick it back up later,” said Jason Ginsberg, head of engineering product, Cursor,  in a demo video, referring to the new mobile capability. Cursor is now on your phone and on the web.Spin off dozens of agents and review them later in your editor. pic.twitter.com\/y7mOlA9h07— Cursor (@cursor_ai) June 30, 2025 The core update allows users to type a task directly into the web interface. An AI agent will then begin working on the codebase, making changes, answering contextual questions, or opening pull requests. Once the agent completes its work, users can continue the process within Cursor, add follow-up instructions, or make inline edits. “If the agent’s work looks complete, you can create and merge a pull request directly from the app,” said Andrew Milich, head of engineering, product, Cursor. Teams can integrate Cursor with Slack to receive task completion notifications and trigger agents by tagging “@Cursor” in conversations. “Your team can easily work with the Cursor Agent on Slack,” the company said, adding that setup instructions are available in their documentation. The update also supports running multiple agents in parallel using different AI models for the same task, enabling side-by-side comparison of results. “We’ve been writing, editing, reviewing code from everywhere,” Ginsberg said, noting they’ve already adopted the web version internally. Cursor’s broader goal remains unchanged.“We want Cursor to be the best place to write code with an AI agent,” the company said. Cursor recently announced a $200\/month ‘Ultra’ plan designed for power users, previously restricted by unpredictable usage caps. The plan promises 20x more usage than the Pro tier. The Ultra tier has been made possible by the company’s “multi-year partnerships” with companies like OpenAI, Anthropic, Google, and xAI. Alongside the Ultra rollout, Cursor has also overhauled its Pro plan. It now defaults to an “unlimited-with-rate-limits” model, replacing the legacy 500-request allowance.","excerpt":"Teams can now integrate Cursor with Slack to receive task completion notifications and trigger agents.","categories":["AI News"],"tags":["Cursor AI"],"author_name":"Siddharth Jindal","publish_date":"2025-06-30T21:59:18","publication_year":"2025","word_count":397,"keywords":["Anthropic","Go","AI coding","OpenAI","AI","programming_languages:R","R","programming_languages:Go","XAI","xAI","Cursor AI"],"extracted_tech_keywords":["AI","OpenAI","Anthropic","xAI","R","Go","XAI","AI coding","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/cursor-brings-ai-coding-agents-to-the-web-and-mobile\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":1090,"title":"Smart Kitchen Devices designed to bring about lifestyle convenience","content":"From airports and automotive to building smarter homes, sensors and the extensive world of analytics have permeated across everything we use or see around us. So why should kitchen be treated any differently, or deprived of the perks of an interconnected ecosystem. It only makes sense as we gradually move into an era where expecting your coffee-maker to respond at a single click on your smartphone for a fresh brew is not an unusual thing anymore. The technology premise today, aided by the power of smart connected devices makes it possible to cook a meal by simply using your smartphone, even if you’re sitting miles away from home. IoT has allowed people to proactively monitor their kitchen chores right from their mobile devices. The day is not far when you wake up to see your kitchen completely taking care of your needs, and you wouldn’t have to raise a finger. Companies like LG, Siemens AG, Orange Chef, and Pantelligent are leading the innovation front when it comes to building a smart kitchen. The amalgamation of analytics and IoT will turn this improbable dream into a reality, as your kitchen would become capable of comprehending your day-to-day schedule. Though some of these futuristic technologies are yet to be brought to life, there are many smart kitchen devices that are up and running. The astonishing part is these devices are well within your budget considerations, and you don’t have to worry about drilling a hole through your pockets to have your very own smart kitchen set up. Let’s walk through some of these game-changing devices, available in India, that you could install in your kitchen to make your life more convenient. The devices are listed in alphabetical order. Amazon Dash Button Amazon Dash Buttons Amazon introduced the Dash button to allow you to re-order things easily through the Amazon account. The device finds itself handy in several scenarios, but using it inside your kitchen could probably be counted as the most practical use. A tap on the device, and it would assist you every time you want to purchase more coffee or paper towels, without needing to go out or even actually placing an order online. The device is available in Amazon store at a price stated $4.99. LG Smart ThinQ cooker LG Smart ThinQ cooker LG launched its precision cooker to make cooking easier and simple. The device incorporates a smart self-cleaning feature that is helpful in case you burnt your baking. The device integrates thousands of pre-set temperature and time-frames, besides easily connecting with the HomeChat app and ecosystem. The most interesting feature of this device is its ability to be easily serviced over internet airwaves utilizing LG’s Smart Diagnosis tool. The device is little expensive, standing at $1,399, but is truly worth the buck. MAID Oven MAID oven MAID is a unique smart kitchen product designed by a Kochi-based startup. The device is connected to an exclusive recipe store over the internet, which culminates an array of exotic dishes to choose from. A user has the advantage of searching recipes depending on ingredients and type of dish. If you’re still confused or lazy to go through the selection hassle, just let MAID suggest you a dish. The device can also turn out to be your personal dietician, proactively learning the calorific requirements based on your daily cooking habits, and simultaneously suggesting the most balanced diet for you. The device is available at an exclusive price of $449. SmartPlate SmartPlate This device is the next big thing to adequately monitor your nutrition, furnishing you with precise estimate about your calorific intake. The device sports three mini cameras along with weight sensors, besides being Wi-Fi and Bluetooth-enabled, completely striving to make your life easier. The SmartPlate observes an individual’s eating habits, and alerts them whenever surplus food is served. The device comes with an integrated app functionality, which synchronizes with the plate, helping you to keep track of every meal. Moreover, the app can connect with MyFitnessPal and FitBit, and is priced at $119. WeMo-enabled Crock-Pot Smart Slow Cooker Crock-Pot Smart slow cooker Crock-Pot brings to you the smart slow cooker making the whole idea of cooking convenient and fun at the same. The standout feature of the device lies in the fact that it’s WeMo enabled, helping you to easily monitor temperature, cook-time, and even change settings to attain the right temperature, or even turn off the device, by simply utilizing the free WeMo app. Imagine, you are reaching late  as you’re stuck somewhere, but you want your food cooked by the time you reach home, this is the next-gen smart device you must look forward to. The device is priced considerably at $129.99.","excerpt":"From airports and automotive to building smarter homes, sensors and the extensive world of analytics have permeated across everything we use or see around us. So why should kitchen be treated any differently, or deprived of the perks of an interconnected ecosystem. It only makes sense as we gradually move into an era where expecting […]","categories":["IT Services"],"tags":[],"author_name":"Дарья","publish_date":"2017-01-05T12:23:38","publication_year":"2017","word_count":784,"keywords":["Go","programming_languages:R","AI","innovation","programming_languages:Go","Ray","analytics","R","startup"],"extracted_tech_keywords":["AI","analytics","Ray","R","Go","innovation","startup","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/smart-kitchen-devices-designed-bring-lifestyle-convenience\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10167234,"title":"Top Tools to Use for Full Stack Developers in 2025","content":"Landing a job in 2025 as a developer is difficult. The availability of hundreds of tools in the market makes it even more difficult, leaving one with no idea about what experts use. Based on current adoption rates and industry trends, it is important to be up-to-date. These tools represent the backbone of modern web development across frontend, backend, and database layers. Development Environments and Version Control Visual Studio Code Visual Studio Code (VS Code) dominates the code editor space with multi-language support, an integrated terminal, and an extensive extension ecosystem. It offers built-in Git integration, intelligent code completion, and debugging capabilities across various programming languages. The editor’s lightweight nature doesn’t compromise performance while handling large projects. VS Code’s live sharing feature enables real-time collaboration between developers. Its customisability allows for personalised workflows with themes and keyboard shortcuts. Git Git provides distributed version control that tracks code changes, enables multiple developers to work concurrently, and facilitates branch-based development workflows. It maintains a complete history of all modifications while consuming minimal storage through efficient compression algorithms. Git enables seamless reversion to previous states, conflict resolution during merges, and secure remote repository hosting. Its branching model supports parallel development of features, bug fixes, and experiments without affecting the main codebase. Docker Docker delivers application containerisation that ensures consistent environments across development, testing, and production. It packages applications with dependencies into isolated containers that run identically regardless of host infrastructure. Docker eliminates “works on my machine” problems by providing reproducible development environments. Its layered file system architecture optimises storage and speeds up deployment by reusing common components. Docker Compose simplifies multi-container application orchestration with declarative configuration files. Containers start in seconds, consume fewer resources than virtual machines, and scale horizontally with orchestration tools like Kubernetes. Frontend Development Tools React React led frontend development with 39.5% adoption in 2024, enabling component-based UI construction with a virtual Document Object Model (DOM) for performance optimisation. Its one-way data flow architecture prevents unpredictable state changes and simplifies debugging. React’s JSX syntax combines HTML with JavaScript for more intuitive component authoring. The library’s focus on UI components allows integration with various state management solutions like Redux or Context API. React Native extends the framework to mobile development, sharing code between web and mobile applications. Facebook’s continued development ensures stability and regular feature updates. Angular Angular provides a comprehensive TypeScript-based framework powering enterprise applications like Gmail and Upwork. Its component architecture enforces code organisation while two-way data binding simplifies model-view synchronisation. It includes a Command Line Interface (CLI) that automates project scaffolding, component generation, and build optimisation. Built-in dependency injection facilitates testable code and service reusability across components. Angular Universal supports server-side rendering for improved SEO and initial load performance. The framework’s opinionated structure accelerates development for large teams. Vue.js Vue.js offers progressive frontend development with optional adoption levels, from simple script tags to full build systems. Its gentle learning curve allows developers to incrementally adopt features as needed. Vue combines React’s virtual DOM performance with Angular’s template syntax familiarity. The framework’s single-file components encapsulate template, logic, and styling in one file for better maintainability. Vue’s reactivity system automatically tracks dependencies and updates the DOM when data changes. The core library focuses on the view layer while official companion libraries handle routing and state management. Version 3 introduced the Composition API for better TypeScript support and code reuse. Bootstrap Bootstrap accelerates responsive web development with pre-built components, grid system, and utility classes. It ensures consistent UI across browsers while accommodating various screen sizes through its mobile-first approach. The framework includes extensively tested JavaScript components for common interface patterns like modal dialogues and carousels. Bootstrap’s utility classes enable rapid styling without writing custom CSS. The framework’s extensive documentation includes examples and code snippets for immediate implementation. Backend Development Tools Node.js Node.js leads server-side JavaScript with 40.8% adoption in 2024, powering platforms like Netflix with asynchronous, event-driven architecture. Its non-blocking Input\/Output (I\/O) model handles thousands of concurrent connections efficiently, making it ideal for real-time applications. NPM provides access to over a million packages, accelerating development through code reuse. Node.js enables JavaScript throughout the entire stack, eliminating context switching between languages. The V8 engine delivers fast code execution while regular releases maintain security and performance. Django Django delivers a Python-based framework with “batteries included” philosophy, providing authentication, object-relational mapping (ORM), an admin interface, and out-of-the-box security features. Its MTV (Model-Template-View) architecture enforces clean separation of concerns. Django’s ORM abstracts database operations across different database engines without changing code. The automatic admin interface generates CRUD (Create, Read, Update, and Delete) operations for models, accelerating backend development. Built-in security features protect against common vulnerabilities like SQL injection, cross-site scripting (XSS), and cross-site request forgery (CSRF) attacks. Django REST framework extends capabilities for API development with minimal additional code. Spring Boot Spring Boot simplifies Java application development with auto-configuration, embedded servers, and production-ready features. It eliminates boilerplate code through convention-over-configuration approaches while maintaining flexibility for custom requirements. The framework’s starter dependencies bundle compatible libraries to solve common problems like database access or security. Spring Boot’s actuator module provides production-ready features including health checks, metrics, and monitoring endpoints. Its embedded server approach enables standalone application deployment without external containers. Express.js Express.js provides minimalist Node.js web framework functionality with routing, middleware support, and template engine integration. Its unopinionated design offers flexibility in application structure and component selection. Express middleware creates a pipeline for request processing, enabling cross-cutting concerns like authentication or logging. The framework’s routing system handles different HTTP methods and URL patterns with parameter extraction capabilities. Express works with various view engines for server-side rendering while supporting JSON responses for API development. NestJS NestJS combines TypeScript with object-oriented programming principles to create scalable Node.js server applications. It implements architectural patterns from Angular, promoting consistent structure across backend services. NestJS provides built-in support for dependency injection, making components testable and loosely coupled. The framework includes integrated support for GraphQL, WebSockets, and microservices alongside traditional REST APIs. Its modular architecture encourages encapsulation of related functionality into cohesive units. Database Technologies PostgreSQL PostgreSQL delivers enterprise-grade relational database capabilities with advanced data types, robust transaction support, and powerful indexing options. It supports complex queries through window functions, common table expressions, and recursive queries. PostgreSQL’s extensible architecture allows custom data types, operators, and procedural languages. The database ensures ACID (atomicity, consistency, isolation, and durability) compliance with multi-version concurrency control for high throughput without locking issues. Its built-in replication options support high availability and read scaling. PostgreSQL handles geospatial data through PostGIS extension and JSON operations for hybrid relational-document storage patterns. Regular security updates maintain protection against emerging threats. MongoDB MongoDB provides a document-oriented NoSQL database that stores data in flexible JSON-like documents without requiring predefined schemas. Its horizontal scaling through sharding distributes data across multiple servers for improved performance and storage capacity. MongoDB’s query language supports complex operations, including aggregation pipelines for data transformation and analysis. The database includes built-in replication for high availability and disaster recovery. Its document model maps naturally to object-oriented programming structures, reducing impedance mismatch. API Development GraphQL GraphQL revolutionises API development by enabling clients to request exactly the data they need, reducing over-fetching and under-fetching problems. Its strongly-typed schema serves as a contract between server and client, improving documentation and enabling better tooling. GraphQL resolvers provide flexibility in data sources, including databases, microservices, or third-party APIs. The single endpoint architecture simplifies API versioning and evolution without breaking clients. Real-time capabilities through subscriptions support event-driven applications. GraphQL’s introspection enables automatic documentation generation and self-discovery of API capabilities. Performance optimisation techniques like dataloader prevent the N+1 query problem common in nested relational data. The full-stack development landscape continues to evolve rapidly, with tools like Node.js and React dominating with over 40% adoption rates in 2024. This comprehensive toolkit spans development environments, frontend frameworks, backend systems, databases, and DevOps solutions.","excerpt":"These tools represent the backbone of modern web development across frontend, backend, and database layers.","categories":["AI Trends"],"tags":[],"author_name":"AIM Media House","publish_date":"2025-04-03T17:40:14","publication_year":"2025","word_count":1301,"keywords":["PostgreSQL","TPU","AI","MongoDB","ML","docker","RAG","microservices","Python","kubernetes"],"extracted_tech_keywords":["AI","ML","RAG","kubernetes","docker","microservices","TPU","MongoDB","PostgreSQL","Python"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-tools-to-use-for-full-stack-developers-in-2025\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10124504,"title":"Data Science Hiring and Interview Process at Diggibyte","content":"Founded in 2021 by Lawrance Amburose, Sekhar Reddy, William Rathinasamy, and Anuj Kumar Sen, Bengaluru-based Diggibyte Technologies focuses on providing data and platform engineering, data science AI, and consulting services to a variety of industries. The founders bring a wealth of experience and expertise in leveraging big data and analytics to drive business innovation and efficiency. AIM got in touch with Anuj Kumar Sen, the chief technology officer of Diggibyte, to learn more about the company’s AI and data science operations, expansion plans, interview process, and work culture. The company is currently hiring for five data science positions in Bengaluru with at least two years of experience in the field. Inside Diggibyte’s Data Science Team Diggibyte’s data science team has tackled numerous issues across various domains. “We have solved multiple problems using data science across the domain, which include customer insights and personalisation (home appliance),” Sen told AIM. By collecting and analysing data from multiple channels, the 50-member data science team provided insights into customer purchasing behaviour, preferences, and needs to tailor marketing efforts and customer satisfaction. When it comes to logistics, the team has addressed the problem of inaccurate demand predictions, an issue that can potentially lead to mismanagement of fleet drivers and resources. By analysing historical sales data, market trends, and seasonal variations, the team claims to have developed accurate demand forecasts. An improved predictive capability helps clients optimise and plan their fleets and drivers, ensuring resources meet customer demand without impacting delivery efficiency or profitability. Another notable project is the development of an AI-driven headline generator for Norwegian language articles. Using NLP models, this solution scans article content and claims to generate compelling headlines that capture the story’s essence. Besides, the company also employs generative AI for multiple clients in construction, media, and retail. “Our team is currently working on a latent diffusion image-to-image in-painting model that can automate the process of lifestyle product photography and commercial photography for sports accessories,” he added. Tech Stack The team leverages both Transformer architecture models and GPT models depending on the task requirements. They have developed solutions using the Transformer library for an RAG system. Additionally, they created a chatbot for customer support utilising GPT-3.5 Turbo in Azure ML Studio. In potential candidates, the company looks for proficiency in Python, SQL, Pandas, NumPy, and machine learning. They should also be adept in using frameworks like Keras or TensorFlow and have experience with Databricks. In addition to these core competencies, data scientists are encouraged to have competitive programming skills, be fast learners, and stay updated with the latest developments in generative AI, preferably with hands-on experience. Interview Process The interview round at the company is followed by a thorough and structured process to assess both technical and interpersonal skills. It begins with two technical assessment rounds where candidates are evaluated on their fundamental data science skills and adaptability to the latest trends in data science and AI. These are followed by an HR round focusing on the candidate’s interpersonal skills, career goals, collaborative abilities, and employment history. The company sources its best fit by relying on a multi-channel approach to find the top talent in data science. This includes leveraging its internal careers page (set to be deployed in 20 days), online career portals, and trusted vendors. Work Culture Once employed, new hires can expect to work on cutting-edge AI projects. It also provides training to bridge gaps, if any, in both technical and interpersonal skills, supported by the leadership team. “Since we have a separate team for analytics, engineering, visualisation, and machine learning, we expect candidates to have specialisation in one skill rather than general knowledge in multiple skills,” Sen explained. Employees can pursue certifications from reputable sources such as Azure and Databricks. The company covers the exam costs. It offers a flexible work-life balance, accommodating four days’ of work-from-office and work-from-home arrangements, along with a supportive leave policy. The company also has policies in effect to make job options accessible, inclusive and diverse for all. Efforts around healthcare and a positive office atmosphere promote employee well-being. Diggibyte Technologies was certified as the Best Firm For Data Engineers and was awarded during DES 2024.","excerpt":"The company is currently hiring for five data science positions in Bengaluru with at least two years of experience in the field.","categories":["AI Hirings"],"tags":["AI","Data Science","Data Science Hiring","data science hiring process","Hiring","Top Trend"],"author_name":"Shritama Saha","publish_date":"2024-07-29T15:18:28","publication_year":"2024","word_count":694,"keywords":["Top Trend","data science","machine learning","AI","TensorFlow","ML","Hiring","NLP","Data Science Hiring","Aim","analytics","generative AI","Data Science","data science hiring process","Azure ML"],"extracted_tech_keywords":["AI","machine learning","ML","NLP","data science","analytics","generative AI","Aim","Azure ML","TensorFlow"],"url":"https:\/\/analyticsindiamag.com\/ai-hiring\/data-science-hiring-process-at-diggibyte\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10086890,"title":"Give a Prompt, Make a Video","content":"When AI research lab OpenAI introduced DALL-E, the first text-to-image (TTI) model of its kind, it took the internet by storm. Since then, several big players have invested in this space, coming up with their own versions. ‘Generative AI’ thus became a common word in the AI ecosystem. Further—as users kept experimenting with TTI—we got text-to-video, text-to-3D, and now text-to-music. Let’s take a look at the top nine text-to-video creators and how they came about! Tune-A-Video To solve the problem of One-Shot Video Generation, where only one text–video pair is available for training an open-domain text-to-video (TTV) generator, researchers from Show Lab at the National University of Singapore built a text-to-video generator called Tune-A-Video. Leveraging text-to-image (TTI) diffusion models that have been previously trained, Tune-A-Video extends spatial self-attention to the spatiotemporal domain by utilising customised Sparse–Causal Attention. Tune-A-Video: One-Shot Tuning of Image Diffusion Models for Text-to-Video Generation is outgithub: https:\/\/t.co\/pqclO38LSv pic.twitter.com\/QdFpizRS3n— AK (@_akhaliq) January 29, 2023 NUWA Infinity Microsoft-developed NUWA-Infinity is a multimodal generative model that can create high-resolution images or long-duration videos of arbitrary size from text, image, or video inputs. It generates open-domain videos through its ‘autoregressive over autoregressive generation’ mechanism. This mechanism handles the variable-size generation task by considering dependencies between patches at a global patch level and dependencies between visual tokens within each patch at a local token level. NUWA-Infinity: Autoregressive over Autoregressive Generation for Infinite Visual Synthesisabs: https:\/\/t.co\/vZz48wA5z2project page: https:\/\/t.co\/PtzdAq2nvNCompared to DALL·E, Imagen\/Parti , generates HR images with arbitrary sizes and support long-duration video generation pic.twitter.com\/0jO3GjSVx5— AK (@_akhaliq) July 21, 2022 GenmoAI GenmoAI is a tool that generates AI from text input. It can make a variety of outputs, such as videos, animations, vector graphics, and more. The advanced video capabilities of GenmoAI include a dynamism feature that adjusts the amount of noise added between frames. Additionally, the tool allows you to add multiple text prompts together to form a story-like narrative. 2. Dynamic Art Forms: Unlike other image generation tools, Genmo goes beyond traditional 2D images to allow you to create videos, animations, vector design assets, and more. One-stop resource with no limitations! pic.twitter.com\/vvSK4DvEkP— Shubham Saboo (@Saboo_Shubham_) February 1, 2023 Make-A-Video Meta released Make-A-Video, a TTV model that can generate high-definition, high frame-rate videos by incorporating the TTI model with the spatiotemporally factorised diffusion model. In addition, the team behind Make-A-Video has combined text with image information to remove the requirement of paired text and video data, thereby opening the door to expanding the tool to a larger quantity of video content. We’re pleased to introduce Make-A-Video, our latest in #GenerativeAI research! With just a few words, this state-of-the-art AI system generates high-quality videos from text prompts.Have an idea you want to see? Reply w\/ your prompt using #MetaAI and we’ll share more results. pic.twitter.com\/q8zjiwLBjb— AI at Meta (@AIatMeta) September 29, 2022 Make-A-Video 3D (MAV3D) Based on Meta‘s Make-A-Video method for 2D generation, Meta has developed MAV3D (Make-A-Video3D) for generating 3D dynamic scenes from text descriptions. The new model uses a 4D dynamic Neural Radiance Field (NeRF) optimised for scene appearance, density, and motion consistency. The Text-to-Video (TTV) model used in the process is only trained on Text–Image pairs and unlabelled videos. Text-To-4D Dynamic Scene GenerationPresents MAV3D (Make-A-Video3D), a method for generating three-dimensional dynamic scenes from text descriptions.proj: https:\/\/t.co\/KSq8okuWPJabs: https:\/\/t.co\/qbAlZYOFTr pic.twitter.com\/HIYYQMZKXG— Aran Komatsuzaki (@arankomatsuzaki) January 27, 2023 CogVideo Hugging Face’s CogVideo is a pre-trained transformer model for generating high-resolution (480×480) videos from the text. It claims to be the largest and first open-source model of its kind, with 9.4 billion parameters. CogVideo allows for control over the intensity of changes during video generation. CogVideo: Large-scale Pretraining for Text-to-Video Generation via Transformersgithub: https:\/\/t.co\/1JuOHU7puc pic.twitter.com\/Wilcq2Xxb9— AK (@_akhaliq) May 29, 2022 Imagen Google entered the TTV space with Imagen, a cutting-edge video synthesis model that generates high-quality videos (1280×768 at 24 frames per second) from written prompts.​​ Imagen Video transforms initial text prompts into low-resolution videos (16 frames, 2448 pixels, at three fps) with impressive features that include creating videos of well-known artwork, 3D object rotation while maintaining structure along with animated text. Meet Imagen Video and Phenaki, two research approaches for text-to-video generation. By combining diffusion & sequence learning techniques, we can generate videos that are super-res at the frame level and coherent in time. (4\/5)https:\/\/t.co\/O7gGzb9knWhttps:\/\/t.co\/Uc0krTyTvk pic.twitter.com\/Op4tonX2iw— Google AI (@GoogleAI) November 2, 2022 Phenaki Phenaki is a TTV model that uses a series of text prompts to create realistic video synthesis. It reduces the video to a limited representation of discrete tokens. Since this tokenizer is auto-regressive in time, it can handle representations of various lengths for videos. The resulting video tokens are then de-tokenized to get the actual video using a bi-directional masked transformer conditioned on pre-created text tokens. Phenaki can make arbitrary long videos based on a series of open-domain prompts, such as time-variable text or stories. Phenaki claims to explore the creation of films using time-variable prompts. Phenaki: Variable Length Video Generation from Open Domain Textual Descriptions abs: https:\/\/t.co\/gsZrW80Aax project page: https:\/\/t.co\/mIzxeMRKk8 Generating videos from text, with prompts that can change over time, and videos that can be as long as multiple minutes pic.twitter.com\/GSDGEURaJD— AK (@_akhaliq) September 29, 2022 Transframer Google‘s DeepMind introduced Transframer, a unified framework for image modelling and vision tasks using probabilistic frame prediction for various tasks, such as video interpolation, view synthesis, and image segmentation. The framework demonstrates the potential of probabilistic image models in multi-task computer vision, outperforming other models on video generation benchmarks and achieving exceptional results on eight tasks, including semantic segmentation and image classification. Based on U-Net and Transformer components, Transframer can create coherent 30-second videos from a single image. Transframer is a general-purpose generative framework that can handle many image and video tasks in a probabilistic setting. New work shows it excels in video prediction and view synthesis, and can generate 30s videos from a single image: https:\/\/t.co\/wX3nrrYEEa 1\/ pic.twitter.com\/gQk6f9nZyg— Google DeepMind (@GoogleDeepMind) August 15, 2022 Due to the high processing cost, scarcity of high-quality text-video data, and unpredictable length of videos, creating videos from text is increasingly challenging, even though tech firms are actively working towards improving it. From healthcare to gaming, generative AI has the potential to revamp and enhance different sectors and aspects of daily life by automating repetitive and time-consuming tasks, offering new and innovative solutions and making them far more personalised.","excerpt":"All you need to know about text-to-video-to-everything generators today!","categories":["AI Features"],"tags":["AI Tool","ChatGPT","DALL.E","DeepMind","Google","Imagen","Meta","Meta AI","OpenAI","text to video"],"author_name":"Shritama Saha","publish_date":"2023-02-08T18:00:00","publication_year":"2023","word_count":1052,"keywords":["Meta AI","TPU","computer vision","AI Tool","R","ChatGPT","text to video","RAG","DALL.E","Meta","AI","Imagen","generative AI","Hugging Face","OpenAI","Transformers","Aim","Google","DeepMind"],"extracted_tech_keywords":["AI","computer vision","generative AI","OpenAI","Aim","Hugging Face","Transformers","RAG","TPU","R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/give-a-prompt-make-a-video\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10039948,"title":"Explainable AI For Decision Making Systems In Medical Domain","content":"The COVID-19 outbreak has spotlighted the need for responsive and transparent systems for monitoring health. Good health information has never been crucial. Researchers from Aalto University, Umea University and KTH Royal Institute of Technology investigated the effectiveness of explainable artificial intelligence approaches for decision-making in medical image analysis. Three machine learning methods, including LIME, SHAP and CIU, were implemented to improve the understanding of decisions made by CNNs. Research methodology The three post-hoc explanatory ML models LIME, SHAP and CIU, were applied to gastral images obtained with the help of Capsule Endoscopy – a pill-sized video camera is used to look for possible signs of polyps, ulcers and tumours in the small intestine. Researchers divided the entire process into four different parts: data pre-processing, CNN model application, LIME, SHAP, and CIU explanation generation, and assessment of human decision-making. Two sets, with more than 3,500 images and 600 images, were taken. The datasets were culled from a 10-hour long video recorded via VCE and were split randomly into training and validation sets for results evaluation. “A CNN model with 50 epochs and a batch size of 16 was used to train the data set and achieve a validation accuracy of 98.58%. We trained our CNN model based on labels assigned to each image to recognise the bleeding versus normal (non-bleeding) medical images. The labels were made using the repository’s annotated images as a reference point,” as per the study. Results Three different explainable methods, Local Interpretable Model Agnostic Explanations (LIME), Shapley Additive exPlanations (SHAP) and Contextual Importance and Utility (CIU), were used. Researchers used Python to implement the LIME and SHAP explainable methods on Aalto University’s Triton high-performance computing cluster, while CIU explanations were created using RStudio Version 1.2.1335. Plus, an unexplainable setting was included for comparative analysis. Three resolutions were obtained. LIME was tested on all validation data sets. In the case of a bleeding image, LIME explanation was used to identify the area contributing positively to the bleeding class. In the case of a non-bleeding image, LIME explanation was used to determine the area contributing to the non-bleeding class. The time required by the LIME model to generate explanations was around 11 seconds per image. Researchers applied the model agnostic Kernel SHAP on a super-pixel segmented image to explain CNN’s image predictions and were similarly tested on all validation data sets. Each picture representing contributions to both the bleeding and non-bleeding class has a SHAP description. The green colour that marks the important features of the photos represents support for the class (bleeding, non-bleeding), while the red colour represents opposition to the class (bleeding, non-bleeding). The time required by the SHAP model to generate explanations was around 10 seconds per image. CIU provided explanations similar to LIME, depicting the important area on the image that contributes to the given class, either the bleeding or the non-bleeding. The time required by the CIU model to generate explanations was around 8 seconds per image. The least time is taken when compared to the LIME and SHAP. The results from all three models were presented to the different set of participants, mostly from the STEM field. Firstly the results were provided without explanation support and then with explanation support. For the LIME model, when participants received the results with explanation – the mean of correct decisions was 8.8 out of 12 answers in total. Again, if we look at the participants provided with SHAP explanation – the mean of correct answers was 8.4 out of 12. While, when presented with CIU – the mean for the same was 10.2 out of 12 answers. Thereby, the CIU model outperformed the other two, LIME and SHAP. “Our results support that participants with CIU will perform better in understanding the provided explanations and by that better distinguish between correct and incorrect explanation in comparison to participants having LIME or SHAP explanation support. Users with CIU explanation support were significantly better at recognising incorrect explanations in comparison to those having LIME explanations and also to some extent better than those having SHAP explanation support,” the research paper explained.","excerpt":"The COVID-19 outbreak has spotlighted the need for responsive and transparent systems for monitoring health. Good health information has never been crucial.  Researchers from Aalto University, Umea University and KTH Royal Institute of Technology investigated the effectiveness of explainable artificial intelligence approaches for decision-making in medical image analysis. Three machine learning methods, including LIME, SHAP […]","categories":["AI Features"],"tags":[],"author_name":"kumar Gandharv","publish_date":"2021-05-12T10:00:00","publication_year":"2021","word_count":682,"keywords":["Go","machine learning","artificial intelligence","programming_languages:R","AI","ML","Python","programming_languages:Python","CNN","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","Python","R","Go","CNN","programming_languages:Python","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/explainable-ai-for-decision-making-systems-in-medical-domain\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":68217,"title":"AWS Releases Honeycode To Develop Apps Without Writing Code","content":"Amazon introduces Honeycode in beta to enable users to develop web and mobile applications without writing code. This service will be available as a part of AWS to allow people with no coding experience in getting started immediately. Anyone familiar with the spreadsheet model will be easily able to use Honeycode. It works similar to a spreadsheet but has many advanced features to assist you in customising the applications according to your needs. Honeycode comes with a rich palette of user interface objects such as buttons, lists, and input fields, which are key features for any applications. With Honeycode, you will be able to trigger actions to generate emails, notifications, and change the tables. Various built-in functions can replicate the existing spreadsheets. Jeff Barr, Chief Evangelist for AWS, wrote that FindRow — a function of Honeycode — which is a powerful version of the Vlookup function of the worksheet. One can get started with Honeycode Builder and create web or mobile applications according to your needs, thereby allowing team members to work on any device. Honeycode does the heavy lifting by automating the process of building applications as it manages the database, business logic, and user interface — essential elements of any business application. In general, to create custom apps for businesses, organisations require to rely on their IT team to effectively connect the applications with databases as well as develop better user interfaces with robots functionality. However, Honycode simplifies the entire end-to-end process of developing applications by offering intuitive features. In addition, you can scale the applications effectively since it leverages the functionalities of AWS. Collaboration with team members is another disadvantage of a spreadsheet as it does not provide better user access control. But, applications built with Honecode can empower the administrators to customise the access of applications according to their needs. To check a complete step-by-step process of building the applications with Honeycode, click here.","excerpt":"Amazon introduces Honeycode in beta to enable users to develop web and mobile applications without writing code. This service will be available as a part of AWS to allow people with no coding experience in getting started immediately. Anyone familiar with the spreadsheet model will be easily able to use Honeycode. It works similar to […]","categories":["AI News"],"tags":["Amazon AWS","AWS"],"author_name":"Rohit Yadav","publish_date":"2020-06-25T14:23:55","publication_year":"2020","word_count":318,"keywords":["Replicate","AWS","AI","cloud_platforms:AWS","programming_languages:R","RAG","Amazon AWS","GAN","R"],"extracted_tech_keywords":["AI","RAG","AWS","R","GAN","Replicate","cloud_platforms:AWS","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/aws-releases-honeycode-to-develop-apps-without-writing-code\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10126833,"title":"Vector Databases are Ridiculously Good","content":"Building large language models requires complicated data structures and computations, which conventional databases are not designed to handle. Consequently, the importance of vector databases has surged since the onset of the generative AI race. This sentiment was reflected in a recent discussion when software and machine learning engineer Santiago Valdarrama said, “You can’t work in AI today without bumping with a vector database. They are everywhere!” He further added that vector databases, with their ability to store floating-point arrays and be searched using a similarity function, offer a practical and efficient solution for AI applications. Vector databases provide LLMs with access to real-time proprietary data, enabling the development of RAG applications. Database companies are pivotal in driving the generative AI revolution and its growth. Redis enhances real-time efficiency for LLM-powered chatbots like ChatGPT, ensuring smooth conversations. At the smae time, enterprises are leveraging MongoDB Atlas and Google Cloud Vertex AI PaLM API to develop advanced chatbots. Making it Easier However, major database vendors, regardless if they were originally established as SQL or NoSQL, such as MongoDB, Redis, PlanetScale, and even Oracle have all added vector search features to their existing solutions to capitalise on this growing need. In an earlier interaction with AIM, Yiftach Shoolman, the co-founder and CTO of Redis, said, “We have been working with vector databases even before generative AI came into action.” Redis not only fuels the generative AI wave with real-time data but has also partnered with LangChain to launch OpenGPT, an open-source model that allows flexible model selection, data retrieval control, and data storage management. Another important challenge vector databases claim to solve is hallucinations, which have been a persistent issue for LLMs. “Pairing vector databases with LLMs allows for the incorporation of proprietary data, effectively reducing the potential range of responses generated by the database,” said Matt Asay, VP, developer relations, in an exclusive interaction with AIM at last year’s Bengaluru chapter of their flagship event MongoDB.local. During a recent panel discussion, Pinecone founder and CEO Edo Liberty explained that vector databases are made to manage these particular types of information “in the same way that in your brain, the way you remember faces or the way you remember poetry”. Most of the prominent names in the industry have already implemented vector capabilities. Think Amazon Web Services, Microsoft, IBM, Databricks, MongoDB, Salesforce, and Adobe. Jonathan Ellis, the co-founder and CTO of DataStax, explained that while OpenAI’s GPT-4 is limited to information up until September 2021, indexing recent data in a vector database and directing GPT-4 to access it can yield more accurate and high-quality answers. This approach eliminates the need for the model to fabricate information, as it is grounded in updated context. What Next? However, vector databases are not without challenges. A recent report by Gartner noted that using vector databases for generative AI may raise issues with raw data leakage from embedded vectors. Raw data used to create vector embeddings for GenAI can be re-engineered from vector databases, making data leakage possible. “Given the compute costs associated with AI, it is crucial for organisations to engage in worker training around vector database capabilities,” Gartner analyst Arun Chandrasekaran emphasised in an interview with Fierce. “This preparation will help them avoid expensive misuse and misalignment in their AI projects.” Nevertheless, several vector db startups are now gaining prominence. During an otherwise weak year for venture capital, hundreds of dollars are flowing into vector database businesses like Pinecone, which got $100 million in April 2023 from Andreessen Horowitz. Pinecone is not the only one. Dutch firm Weaviate secured $50 million from Index Ventures. The Weaviate AI-native vector database simplifies vector data management for AI developers. There are emerging divisions in the vector database arena, particularly between open- and closed-source players, and between dedicated vector databases and those with integrated vector storage and search functionality. On the dedicated, open-source side, Chroma, Quadrant, and Milvus (in collaboration with IBM) stand out, while Pinecone is a leading dedicated, closed-source player. Meanwhile, Snowflake, although not a dedicated vector database, offers vector search capabilities within its open-source framework. And there’s a good reason why so many people are jumping into this sector. Chandrasekaran predicts that 30% of organisations will employ vector databases to support their generative AI models by 2026, up from 2% in 2023. Understanding its importance, Andrew Ng, has also introduced free learning courses on the same with MongoDB, Weavaiate, Neo4j and more. With the increasing adoption predicted by experts and the introduction of educational resources, vector databases are set to play a pivotal role in shaping the next era of AI technology. As organisations continue to integrate these powerful tools, the potential for innovation and improved AI capabilities becomes ever more significant, heralding a new age of intelligent applications and solutions.","excerpt":"With the increasing adoption predicted by experts and the introduction of educational resources, vector databases are set to play a pivotal role in shaping the next era of AI technology","categories":["AI Features"],"tags":["AI (Artificial Intelligence)","Data Science","Vector Database"],"author_name":"Anshul Vipat","publish_date":"2024-07-15T11:35:11","publication_year":"2024","word_count":795,"keywords":["ChatGPT","GenAI","machine learning","OpenAI","AI","RAG","Vector Database","LangChain","Aim","Ray","generative AI","Data Science","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","machine learning","generative AI","GenAI","ChatGPT","OpenAI","LangChain","Aim","Ray","RAG"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/vector-databases-are-ridiculously-good\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10002028,"title":"7 Most Used Programming Languages For IoT Projects","content":"IoT is one of the most popular technologies today. With the urban renewal and retrofitting program of Smart City Mission, the Indian government is trying its best to contribute to the mission of Smart Cities. To start a career as an IoT developer there are some programming languages that are an absolute must and  govern this space. In this era of IoT governing many different industries and applications, here are the top programming languages that rule the IoT projects. 1. Java Java has an incredible policy of “write once, run anywhere” (WORA), which is the program’s ability to run on all common Operating Systems. This forms a good choice for IoT projects as the code is easily deployable in different devices and chips using Java Virtual Machine (JVM). Using JVM the code can be transferred to  smartphones and servers. Java also provides the hardware secure libraries, capable of accessing the genetic code. 2. C C has been the backbone of many platforms till today. IoT is one such spectrum where it has a lot of usage. It can work directly with the RAM and hence it doesn’t require a lot of processing power. This programming language is compiled making it great for IoT projects. 3. JavaScript JavaScript, most popular for all web browsers and HTML, also finds its use in the space of IoT. JavaScript is used as the programming language in all the web browsers and HTML. JavaScript makes things easier as it makes the devices interoperable. Most of the work is focused on the servers and hubs that collect information and then store it. It has an end-user scripting and is easy to learn. The advantage with JavaScript is that in the process of developing an IoT system,  embedded programmers need not learn different syntax if they are using JavaScript because its syntax is very similar to C. JavaScript developers in IoT have sophisticated frameworks and engines like CycloneJS, IoT.js, JerryScript, Duktape, etc. specifically designed for constrained devices. 4. Python Python began its journey as a high-level scripting language. It has a large number of useful libraries and can get more stuff done with fewer lines of code. It is popular for web applications but also has its usage in IoT. Many of the data analysis involved with IoT systems is done using Python. With a compact yet readable source code, Python is the best choice for managing complex data streams without having to maintain equally complex codebases. But the language is not very popular in areas where the computational requirement is high. It is generally preferred for simple, modular projects. 5. C++ C++ has seen an increase in its usage for IoT solutions because of the processors being more and more powerful. It is an alternative to C if the devices require more complex tasks. The language also causes to utilize different languages including C#, Java and Python. 6. Swift Swift is popularly known for building applications for Apple’s iOS and Mac OS devices. Apple is looking forward to inculcate smart solutions and that makes it a part of the IoT. To make the devices interact with an iPhone or an iPad, it is easy to build an application using Swift. To make the iOS devices center of the home network of sensors Apple is shifting from C to Swift and is therefore building libraries that can handle most of the work. 7. PHP According to an Eclipse Foundation survey, 11.2 percent of developers said that they are including PHP in their code stack and it was a language that was mentioned most often by them. Not as popular as the other big players, PHP comes around as the best support for IoT with it being affordable and most flexible language platform. Both together can create new and exciting ways for users to interface the digital world The Best Language For IoT The language to be used depends a lot of the needs of the IoT project. But according to the present scenario, Java has the majority of usage compared to the other languages. Each language is different in its own way and provides different advantages for different projects. The language chosen should suit the needs of the end-use applications. Currently, an amalgamation of these languages seem to do the job best.","excerpt":"IoT is one of the most popular technologies today. With the urban renewal and retrofitting program of Smart City Mission, the Indian government is trying its best to contribute to the mission of Smart Cities. To start a career as an IoT developer there are some programming languages that are an absolute must and  govern […]","categories":["AI Features"],"tags":["C","IoT","Java","Programming Languages","Python"],"author_name":"Disha Misal","publish_date":"2019-05-09T15:55:36","publication_year":"2019","word_count":715,"keywords":["Go","C","AI","ML","Git","Programming Languages","Python","C++","CLIP","JavaScript","R","Java","IoT"],"extracted_tech_keywords":["AI","ML","Python","R","JavaScript","Go","Java","C++","Git","CLIP"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/7-most-used-programming-languages-for-iot-projects\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10067851,"title":"The seedy side of matchmaking algorithms","content":"It’s scary to know how much it’ll affect people. I try to ignore some of it, or I’ll go insane. We’re getting to the point where we have a social responsibility to the world because we have this power to influence it.Jonathan Badeen, Co-founder & Chief Strategy Officer, Tinder Swipe right 100 times! Answer 200 questions to find your perfect date. Upload high-res cat pics. Flex. Humblebrag: The internet is full of tips and tricks to power your online dating life. But even if you put everything into practice, your chances of landing a date hinges on an X factor–matchmaking algorithms. Online dating sites started experimenting with compatibility matching in the early 2000s. It enabled dating sites to monetise their offerings, drive user engagement etc. In addition, most of these sites claimed to offer ‘scientific matching’ based on user preferences. eHarmony was the first dating platform to create and patent a matching algorithm in 2000. The idea was to reduce divorce rates by intervening in the mating decisions of the site’s single users. However, by today’s standards, the original eHarmony algorithm was naive–using a regression-based approach to match users based on variables said to predict long-term relationship satisfaction. In 2004, OkCupid added algorithmic matching to its basic search functionality. The algorithm gauged compatibility using “match percentages” extracted from user Q&As. Then, each question was weighted based on the level of importance (i.e., “irrelevant,” “a little,” “somewhat,” “very”). OkCupid gave users control over the matching process. Users could answer some of the questions and let the algorithm assume the others. Today, online dating sites or apps use sophisticated machine learning algorithms to predict users’ preferences based on implicit feedback. Swipe away In 2009, Grindr deployed collaborative filtering to gain insight into user preferences. Collaborative filtering makes recommendations by picking up on patterns of similar users. The same technique is used to recommend products on Amazon and movies on Netflix. Then came Tinder and Hinge in 2012. Tinder interface looks like a deck of playing cards, with users swiping left to reject a profile and right to match. Tinder used the Elo system to score user desirability and match them with others in the same league. The Elo system is used in FIDE (chess) rating to assign a score to players based on their previous wins\/losses and the skill levels of their opponents. Ratings work similarly on Tinder, with a right swipe from someone desirable having the greatest impact on a user’s score, just as beating a Grandmaster in chess matters more than beating an amateur. Tinder claims to have retired Elo rating but is mealy-mouthed about its new system. Many speculate Tinder uses something similar to the Gale-Shapley algorithm like Bumble or Hinge: The algorithm scrapes your online data: Facebook, Instagram, Spotify, screen time, dating app usage etc to learn who you are, what you like, and who is likely to like you. When swiping through the profile cards of your preferred gender, Bumble considers general attractiveness and popularity as the main factors to define the order of profiles. So even if you don’t get any matches, you see the most attractive profiles first, leading to a pleasant experience. The same pleasure you get from window shopping (think ludic loops). However, Bumble has never admitted to using the Elo rating to rank the attractiveness of profiles. One issue with using collaborative filtering for matchmaking is the possibility of gender and racial bias creeping into the algorithms. Instead of making dating more inclusive, the collaborative filtering will likely replicate the same biases seen offline. Desperate much? Have you ever noticed you get the most likes in the initial days? The frontloading is by design. It serves two purposes; First is validation, so you’d not uninstall the app. Second, the algorithm learns about your preferences and general attractiveness to activate a feedback loop. The idea is to customise and optimise the algorithm. Interestingly, social casino games use the same technique to keep the customer hooked. In Bumble, only women can start a conversation with their matches, and the match will expire if the conversation is not initiated in the first 24 hours. Those with a subscription can see people who have swiped right in their ‘beeline’. Similarly, Tinder shows you have run out of profiles in your area. But give it some time, and the profiles start to appear in droves again. Again, this could be a rationing strategy to create demand and keep users coming back for more. Also, some dating apps employ shadowban to prevent users from frequently deleting and creating new profiles. The matchmaking algorithms tap into your primal brain and hold you prisoner by creating addictive behaviour. The tagline of Hinge is: The dating app designed to be deleted. But the algorithm says a different story.","excerpt":"eHarmony was the first dating platform to create and patent a matching algorithm in 2000.","categories":["AI Features"],"tags":[],"author_name":"Sri Krishna","publish_date":"2022-05-26T11:00:00","publication_year":"2022","word_count":797,"keywords":["Replicate","Go","machine learning","programming_languages:R","AI","programming_languages:Go","RAG","Aim","R"],"extracted_tech_keywords":["AI","machine learning","Aim","RAG","R","Go","Replicate","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/the-seedy-side-of-matchmaking-algorithms\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10064906,"title":"Pinterest launches its first-ever developer-centric open API","content":"Pinterest has launched a developer-centric open API. The Pinterest API (v5) will enable developers to connect quickly and build applications that enable faster Pinterest content creation and direct access to analytics. Noteworthy features include: A single API for all use cases: the Pinterest API (v5) is the single API for developers to build applications and solutions for merchants, advertisers, creators and Pinners to drive results on Pinterest.An open API: Tiered access is introduced to allow developers to test the API safely. Start testing with trial access and then upgrade to standard once you’re ready to share your product with clients and users.Built with Pinner safety and security in mind: OAuth 2.0 scopes have been redesigned, now with refreshable user access tokens. Pinners’ privacy was top of mind while developing this API. Pinterest continues to give Pinners control of which applications they grant access to.Developer friendly and easy to use: Pinterest have completely redesigned Pinterest Developers to make it easier for developers to manage their Pinterest apps, add collaborators, and learn best practices for building with Pinterest. In addition to Pinterest Developers, they publish their OpenAPI versioned spec both on their developer site and GitHub. The developers also created a Quickstart tutorial on GitHub to help developers get up to speed quickly. Throughout 2022, they will launch client SDKs and sandbox environments to make it even easier for developers to build on the Pinterest API (v5). For more details, click here.","excerpt":"The API will enable developers to connect quickly and build applications that enable faster Pinterest content creation and direct access to analytics.","categories":["AI News"],"tags":["applications","Pinterest"],"author_name":"Kartik Wali","publish_date":"2022-04-13T14:16:59","publication_year":"2022","word_count":240,"keywords":["API","programming_languages:R","AI","Git","Pinterest","REST API","analytics","GitHub","R","applications"],"extracted_tech_keywords":["AI","analytics","R","Git","GitHub","API","REST API","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/pinterest-launches-its-first-ever-developer-centric-open-api\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":56853,"title":"Do You Know These 10 New AI Jargon","content":"When every single element in the technological space is rapidly changing, it is evident for some of them to go obsolete. Not just technology but the literature of technology also loses its essence with time. A few years ago, machine learning (ML) or artificial intelligence (AI) were terms used to fascinate people. In 2020, these words do turn heads, but people are now used to these terms as they are a part of everyday dinner table conversations for many. However, there is no shortage of new terms in the landscape, which are gaining prominence. Are you familiar with these ten new AI jargon? In this article, we would like to educate you with new terms of the year from the domain of AI. Explainable AI AI has become a part of our daily lives for quite some time now. From face recognition in smartphones to detect obstacles on the highway, it takes decisions in seconds and delivers the required result. But this brings us to an important question: how does it reach a particular conclusion? Explainable AI (XAI) is a new AI term which tries to understand the black box decisions taken by the AI system. It focuses on the steps and models used by AI systems to make decisions. XAI solutions pinpoint the factors that significantly impact the decisions made by ML models. XAI is important since it delivers transparency in how the black box decisions are taken which lets humans trust the AI system. Read More: Explainable AI Initiatives Reward Tampering When it comes to AI, every model receives a positive as well as a negative reward for completing a task in the correct way or in the wrong way, respectively. Reward tampering occurs when the model finds a way to get a positive reward by completing a task in the wrong direction. This can have an adverse impact on achieving true AI, will decrease trust among users. Read More: Reward tampering problem and solutions Adversarial Attacks Adversarial attacks are commands to a machine learning model, often created by reviewers to evaluate ML models. It has been noticed that AI models often fail to deliver desired results if the inputs are changed slightly, although in a tricky way. Adversarial attacks have become a go-to approach for benchmarking AI models. These attacks are formed in different types of data such as images, graphs and texts. Read More: Monitoring adversarial attacks on the model Federated Learning Federated Learning is a new framework in the AI model development. The system allows users to train models through mobile devices, which accesses a vast range of data that are stored in different sites. This helps various organisations to work together on different models without the need to directly share each other’s data bank. Read More: Federated Learning Meta-Learning Meta-learning allows a user to find various model agnostic solutions. In a nutshell, meta-learning can be defined as the use of machine learning in machine learning. Meta-learning allows a user to find various model agnostic solutions. Meta-learning refers to the designing of models which are capable of learning new skills or can automatically adapt to new and different environments with finite training sessions. It learns about a model’s different parameters such as decay rate, number of hidden neurons etc. Read More: Make Meta-Learning more effective Dark Patterns Dark Patterns is a typical case of algorithmic exploitation where malicious players use different tricks on websites to force a user do a thing which was not wanted by the user. For example, dark patterns force a user to buy dog food even though they have no pets. Data Patterns are carefully designed and deployed with a solid understanding of human psychology but keeping aside the interest of a particular user on the web. Read More: Dart Patterns Reproducibility Reproducibility refers to a work done by someone who followed a listed procedure and managed to achieve the same result as the original one. Reproducibility in science is essential as it establishes the reliability of certain methods or techniques for wider usage. Indigenous replication leads to more efficient solutions and in case of AI with hundreds of papers released every day, it is crucial to keep track of their efficacy. Read More: Reproducibility in AI Temporal Cycle Consistency Temporal Cycle Consistency is a self-supervised learning method which is used to identify the similarities that exist between two videos when labelled data is non-existent. This learning method was introduced by Google to understand similar sequential processes since the use of supervised learning to understand the individual frame of a video is an expensive matter. It also requires the annotators to apply a fine-grained label in each frame of a video which is a time-taking affair. Read More: Temporal Cycle Consistency Causal Influence The role of causality in machines has gained a lot of attention in recent years. With pioneers like Judea Pearl, stressing for causal based systems, there has been a newfound interest in establishing learning based on causal inferences and influences. In short, making a machine to understand the why, what and how of every activity it performs, learn from them and then improve. Read More: Causal Inference Neural Network Compression Neural Networks requires a lot of memory to be stored, due to which it is crucial to compress it. The compression of the neural network is often done with altering the Weight Matrices. In order to do so, it essentially requires to go through three different stages, such as Pruning, Quantization, and Huffman Encoding. There are quite a decent number of compression techniques that make deploying ML models easier than before.Read More: Neural Network Compression","excerpt":"When every single element in the technological space is rapidly changing, it is evident for some of them to go obsolete. Not just technology but the literature of technology also loses its essence with time. A few years ago, machine learning (ML) or artificial intelligence (AI) were terms used to fascinate people. In 2020, these […]","categories":[],"tags":[],"author_name":"Rohit Chatterjee","publish_date":"2020-02-18T16:00:00","publication_year":"2020","word_count":937,"keywords":["federated learning","Go","machine learning","artificial intelligence","AI","neural network","R","ML","Rust","xAI"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","neural network","xAI","federated learning","R","Go","Rust"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/do-you-know-these-10-new-ai-jargon\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10005298,"title":"Hands-on Tutorial On Automatic Machine Learning With H2O.ai and AutoML","content":"As the field of machine learning and artificial intelligence advances to solve a plethora of problems, there is a surge in the number of tools available to develop robust models. Developers may often run into problems about which tool to choose and spend a lot of time understanding the compatibility and features of these tools. Other than this, there is a gap between data science skill supply and demand. To solve these problems, organizations are coming up with frameworks that automatically process the dataset and build a baseline model. One such organization is H2O.ai. In this article, we will look at: Who is H2O.ai?Features and capabilities of H2O.aiDemonstration of AutoML in model development and prediction using H2o.ai What is H2O.ai? The company aims to create open-source machine learning products to make machine learning accessible and allow users to extract insights from data, without needing expertise in deploying or tuning machine learning models. They provide a range of products like H2O: This is an open-source, memory inclusive and distributed machine learning platform to build supervised and unsupervised machine learning models. It also includes a user-friendly UI platform called Flow where you can create these models. Sparkling water: This platform is an integration of Spark and H2O for existing Spark ecosystem users to build their models. Deepwater: Deepwater is an integration of H2O with Tensorflow, Caffe and MXNet. They are used to work on GPU based models used in deep learning and reinforcement learning. Steam: Steam is an enterprise product that allows you to deploy models that you build. You can also convert a trained model into an API for others to access. Features and capabilities The H2O platform has a set of different features and capabilities that are discussed below. Clusters: H2O is a java virtual machine capable of performing parallel computations for machine learning on clusters. Clusters are software with one or multiple nodes. These can be launched in your laptop, a server or multiple machines if more than one node is used. Memory is stored in a compressed columnar format, allowing you to read the data in parallel.  A cluster memory capacity is a sum of memories across all H2O nodes in the cluster. This is a much more flexible and efficient way of modelling because not only does it provide better efficiency for CPU but also provides great flexibility while scaling the model. These clusters are what make H2O fast. Flow: Flow is an interactive user interface that allows you to execute code, write text and plot graphs. It is similar to notebooks like jupyter notebook or collaboratory notebook. The uniqueness of this is it does not display the output just as plain text, it allows you to point and click and interact with objects in the form of tabular data. Flow runs on your localhost. Flow supports REST API, R scripts, and CoffeeScript and no programming experience is required to run H2O Flow. You can click your way through any H2O operation without ever writing a single line of code. AutoML: Automatic ML is designed to have as few parameters as possible while modelling so that all the user has to do is upload a dataset, distinguish between features and target for prediction and set the number of trainable models. Every other process is automated with a goal of finding the best fitting model for that dataset. AutoML is also known for being able to select and build high accuracy ensemble models. Demonstration of AutoML We saw that H2O provides a lot of unique and out of the box capabilities to achieve faster and more efficient modelling. Let us now look at a hands-on demonstration on how to build a model using AutoML. If you are using flow, you can just download and begin working on it. But since I am using a jupyter notebook, I need to install h2o packages before coding. !pip install -f http:\/\/h2o-release.s3.amazonaws.com\/h2o\/latest_stable_Py.html h2o In order to keep this demonstration simple, we will build a classification model for detecting diabetes. Click here to download the dataset. Once downloaded, upload to the drive or working directory. Creating a cluster Every model in the H2O environment works on clusters. To create a new cluster follow these steps : import h2o h2o.init() Doing this will create a new cluster. After the cluster has been created, let us now load our data and start AutoML. diabetes_data = h2o.import_file(\"diabetes.csv\") diabetes_data.head(5) The describe function allows us to get a description of data types, missing values and other attribute information. diabetes_data.describe() Note: AutoML is designed to think of all problems as regression problems unless specified. To make sure this problem is not considered as a regression problem, convert the data type of your target to enum using the following command. diabetes_data['Outcome'] = diabetes_data['Outcome'].asfactor() Let us now split the data as train and test. I will 80% of data for training and 20% for testing. Data splitting is done using spilt_frame() method. diabetes_split = diabetes_data.split_frame(ratios = [0.8]) db_train = diabetes_split[0] db_test = diabetes_split[1] Next step is to assign labels and target names to variables. x=['Pregnancies','Glucose', 'BloodPressure','SkinThickness','Insulin' ,'BMI','DiabetesPedigreeFunction','Age'] y='Outcome' All our data is ready and it is time to pass it to AutoML function. AutoML provides an entire leaderboard of all the models that it ran and which worked best. automl = H2OAutoML(max_models = 30, max_runtime_secs=300, seed = 1) automl.train(x = x, y = y, training_frame = db_train) leader = automl.leaderboard leader.head() leader.head(rows=leader.nrows) This leader board shows us that a stacked ensemble model gives us the best accuracy. Let us make a prediction on test data to understand if the model is working correctly. Making Predictions predictions = automl.predict(db_test[:-1]) (predictions['predict']==db_test['Outcome']).as_data_frame(use_pandas=True).mean() 76% accuracy is a good one considering the fact that we have not pre-processed or performed any feature engineering on the dataset. The model can be saved as follows. h2o.save_model(automl.leader, path = \"your_directory_path\") Since all of these are developed on clusters in order to release memory and dependencies, you need to shut down the cluster using h2o.shutdown() Conclusion H2O’s goal to make ML easy for everyone and to democratize AI is growing at a rapid pace. With tools like these, it is possible to try and bridge the gap between supply and demand of machine learning engineers. H2O platforms are powerful for developers to explore multiple techniques and to build models in a short period of time.","excerpt":"In this article, we will look at Who is H2O.ai, Features and capabilities of H2O.ai, Demonstration of AutoML in model development and prediction using H2o.ai","categories":["Deep Tech"],"tags":["apm data science","Automl","Azure Machine Learning","H2O AI","Machine Learning","machine learning gpu","machine learning software","object store database","PowerBI","prediction","simple ai tutorial","tensorflow tutorial"],"author_name":"Bhoomika Madhukar","publish_date":"2020-08-22T11:00:00","publication_year":"2020","word_count":1055,"keywords":["simple ai tutorial","PowerBI","deep learning","object store database","Pandas","data science","artificial intelligence","machine learning gpu","H2O AI","apm data science","prediction","machine learning","AI","ML","Machine Learning","Azure Machine Learning","tensorflow tutorial","Automl","Aim","Jupyter","machine learning software","TensorFlow"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","data science","Aim","TensorFlow","Jupyter","Pandas"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/hands-on-tutorial-on-automatic-machine-learning-with-h2o-ai-and-automl\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":67151,"title":"Top 8 Platforms For IoT Development","content":"IoT is a phenomenon that includes Machine-to-Human communication (M2H), Radio Frequency Identification (RFID), Location-Based Services (LBS), Lab-on-a-Chip (LOC) sensors, vehicle telematics, and much more. One of the most crucial developments of the 21st Century — IoT — is gaining prominence in almost every sector. According to sources, the global IoT analytics market size is expected to grow USD 27.8 billion by 2022, at a CAGR of 26% during the forecast period. In this article, we list down the top 8 platforms for IoT development one must know. (The list is in alphabetical order) 1| DeviceHive About: DeviceHive is an open-source, scalable, hardware and cloud-agnostic microservice-based IoT data platform with a wide range of device integration options. It consists of communication layer, control software, multi-platform libraries and clients to bootstrap development of smart energy, home automation, remote sensing, telemetry, remote control and monitoring software, among others. Know more here. 2| Kaa About: Kaa is a highly flexible, multi-purpose, open-source middleware platform for implementing complete end-to-end IoT solutions, connected applications, and smart products. The platform provides a range of features that allow building applications for smart products, manage their device ecosystems flexibly, orchestrate end-to-end data processing, and other such. The Kaa platform supports lightweight IoT protocols for device connection, such as MQTT. The platform also allows building applications that function over any type of network connection, either persistent or intermittent. Know more here. 3| Mainflux About: Mainflux is modern, scalable, secure, open-source, and patent-free IoT cloud platform written in Go. The platform is built as a set of microservices, containerised by Docker and orchestrated with Kubernetes. The IoT platform serves as software infrastructure and middleware which provides data aggregation and data management, connectivity and message routing, device management and other such. Know more here. 4| Mocana About: The Mocana IoT Platform is an embedded security software solution which helps software developers to minimise the time to integrate strong security controls into the embedded applications. The software is basically provided as source code and can be customised as well as compiled into applications on resource-constrained chips, including microcontroller units (MCUs), single board computers (SBCs), among others within industrial control, electronics as well as IoT devices. The cybersecurity solution of this platform includes a simple set of APIs for Java, C++ and Python. Know more here. 5| Macchina.io About: Macchina.io is an application software platform for IoT devices. It provides a web-enabled, secure, modular and extensible C++ and JavaScript IoT application environment on top of Linux, consisting of middleware, protocols and rich APIs. Macchina.io provides ready-to-use and industry-proven software building blocks that enable applications to collect, process, filter, analyse and visualise sensors. One can also process data locally or directly where it originates. Know more here. 6| Thinger.io About: Thinger.io is a cloud IoT Platform that provides necessary tools to prototype, scale and manage connected products in a very simple way. Thinger.io platform is formed by two main products: a backend which is the actual IoT server and a web-based frontend that simplifies working with all the features using any computer or smartphone. The goal of this platform is to democratise the use of IoT, making it accessible to the whole world by streamlining the development of big IoT projects. Know more here. 7| ThingSpeak About: ThingSpeak is an IoT analytics platform service that allows you to aggregate, visualise and analyse live data streams on the cloud and provides instant visualisations of data posted by your devices to ThingSpeak. Some of the key capabilities of ThingSpeak include the ability to configure devices to send data using popular IoT protocols, aggregate data on-demand from third-party sources, prototype and build IoT systems without setting up servers or developing web software and other such. Know more here. 8| Zetta About: Zetta is an open-source platform built on Node.js for creating Internet of Things servers that run across geo-distributed computers and the cloud. It combines REST APIs, WebSockets and reactive programming for assembling many devices into data-intensive, real-time applications. With Zetta, a user can link Raspberry Pi, BeagleBones and PCs together with cloud platforms to create geo-distributed networks. Know more here.","excerpt":"IoT is a phenomenon that includes Machine-to-Human communication (M2H), Radio Frequency Identification (RFID), Location-Based Services (LBS), Lab-on-a-Chip (LOC) sensors, vehicle telematics, and much more. One of the most crucial developments of the 21st Century — IoT — is gaining prominence in almost every sector. According to sources, the global IoT analytics market size is expected […]","categories":["AI Trends"],"tags":["App Development","best facial recognition software","IoT","IoT development","IoT devices","iot platform"],"author_name":"Ambika Choudhury","publish_date":"2020-06-11T15:00:00","publication_year":"2020","word_count":683,"keywords":["Go","best facial recognition software","IoT devices","AI","ML","JavaScript","docker","IoT development","microservices","App Development","Python","analytics","iot platform","R","kubernetes","IoT"],"extracted_tech_keywords":["AI","ML","analytics","kubernetes","docker","microservices","Python","R","JavaScript","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-8-platforms-for-iot-development\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10059664,"title":"Meta AI proposes a new approach to improve object detection","content":"Researchers from Meta AI and the University of Texas at Austin have proposed Detic, which simply trains the classifiers of a detector on image classification data and thus expands the vocabulary of detectors significantly. Detic does not assign image labels to boxes based on model predictions, and is easier to implement and works well with a range of detection architectures and backbones. What makes Detic different? Object detection consists of two subproblems: locating the object (localisation), and annotating (classification). Usually, detection datasets are substantially smaller in scale and vocabulary (object classes) than image classification datasets. For example, the newest LVIS detection dataset has 120,000 images for approximately 1,000 classes. The researchers suggested a simple classification loss which utilises image-level supervision to the proposal with the most spatial size–without supervising any other outputs for image-annotated data. Weakly supervised object detection Weakly supervised object detection (WSOD) is gathering steam due to the inconvenience of collecting a large amount of data with accurate object-level annotations in fully supervised object detection methods. Existing weakly-supervised detection techniques train object detectors using just image tag supervisions. They often use poorly labeled data to oversee both the localisation and the classification of object detection, without any box supervision. Since image classification data doesn’t make use of box labels, WSOD techniques have to construct different label-to-box designation methods to obtain boxes. However, these designations require solid initial detections. This leads to a chicken-and-egg problem: good label assignment requires a good detector, but a good detector requires many boxes to be trained. On the other hand, semi-supervised WSOD uses bounding box supervisions in conjunction with image labels. For instance, YOLO9000 combines detection data and classification data within a smaller batch, and designates classification labels to anchors with the strongest predicted scores. Detic is a semi-supervised WSOD. It avoids the process of annotation-assignment by supervising the classification subproblem independently while using classification data. The technique learns detectors for new classes, which was previously impossible to predict and assign. Open vocabulary object detection Open vocabulary object detection aims to detect objects that aren’t part of the training vocabulary. The standard  method is to replace the final classification layer with language embeddings of the class names. Earlier, classifier embeddings have been improved by introducing further text material or using contrastive learning to pre-train the detector on image-text pairs. ViLD has achieved this by upgrading the language embedding to CLIP, and then separating region features from CLIP image features. Detic, like ViLD, uses CLIP embeddings as the classifier. However, instead of using distillation, it integrates additional image-annotated data for co-training. Large vocabulary object detection This requires identification of 1,000+ classes. Detic’s method builds on previous attempts at handling the long-tail problem including repeat factor sampling (which oversamples classes with fewer annotations), as well as Equalization losses and SeeSaw losses (which re-weight per-class loss by evening out the gradients or number of samples). Meanwhile, Detic takes on the problem by including additional image-labeled data. Language supervision for object detection Detic uses language data in a manner similar to Cap2Det, which first learns a mapping from sentences to image annotations in the detector’s object classes, and then employs WSOD. Unlike Cap2Det, Detic can extract image labels from captions by using a simple text-match. Conclusion Detic is an easy method of applying image supervision in object detection with a significant vocabulary. However, one major limitation is it does not consider overall data statistics and supervises all image labels to the same region. Furthermore, open vocabulary generalisation may not work in extreme domains. Nevertheless, Detic has successfully expanded large-vocabulary detection with an assortment of weak data sources, classifiers, detector architectures, and training methods. Furthermore, its generalisation capabilities are supplemented by the large-scale pretraining that CLIP is put through. The researchers hope Detic will make object detection easiest to deploy and advance research in open-vocabulary detection.","excerpt":"Detic, like ViLD, uses CLIP embeddings as the classifier.","categories":["Global Tech"],"tags":["AI (Artificial Intelligence)","Classification","contrastive learning","Object Detection"],"author_name":"Srishti Mukherjee","publish_date":"2022-02-02T18:00:00","publication_year":"2022","word_count":638,"keywords":["Go","Classification","Meta AI","TPU","AI","Aim","object detection","Object Detection","YOLO","CLIP","R","contrastive learning","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","Meta AI","Aim","object detection","TPU","R","Go","CLIP","YOLO","contrastive learning"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/meta-ai-proposes-a-new-approach-to-improve-object-detection\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":39291,"title":"MeitY&#8217;s CSC To Use Artificial Intelligence For Improvement Of Services In Rural Areas","content":"The CSC e-Governance Services India Ltd (CSC SPV) is now on its way to use artificial intelligence and data analytics to improve services across sectors such as finance, education, healthcare, and others. This special purpose vehicle has announced a strategic alliance with NEC Technologies to develop and deliver new digital services to around 900 million citizens living in rural areas in India. Takayuki Inaba, Managing Director, NEC Technologies India (NECTI), told a leading daily, “With this strategic alliance, our innovative technologies will unlock new opportunities and bring additional value to citizens, CSC operators as well as CSC SPV.” CSC e-Governance Services India Limited, is a Special Purpose Vehicle, which been set up by the Ministry of Electronics & IT under the Companies Act, 1956 to oversee implementation of the CSC scheme. CSC SPV provides a centralised collaborative framework for delivery of services to citizens through CSCs, besides ensuring systemic viability and sustainability of the Scheme. Reportedly, over 4,70,000 transactions occur at CSCs across India every day. Their vision is to develop the CSCs as a reliable and ubiquitous IT-enabled network of citizen service points connecting local population with government departments, business establishments, banks, insurance companies and educational institutions, with an impact on primary, secondary and tertiary sectors of the country’s economy. Dr Dinesh Kumar Tyagi, CEO at CSC SPV, said, “The association with NECTI will help in utilising technologies for delivery of various services to citizens, especially those living in rural India. Education, financial inclusion and telemedicine are the areas where innovative technologies can be utilised to improve the quality of life for people living in rural India.”","excerpt":"The CSC e-Governance Services India Ltd (CSC SPV) is now on its way to use artificial intelligence and data analytics to improve services across sectors such as finance, education, healthcare, and others. This special purpose vehicle has announced a strategic alliance with NEC Technologies to develop and deliver new digital services to around 900 million […]","categories":["AI News"],"tags":["MeitY","purpose of artificial intelligence"],"author_name":"Prajakta Hebbar","publish_date":"2019-05-16T12:36:41","publication_year":"2019","word_count":268,"keywords":["Go","artificial intelligence","programming_languages:R","AI","programming_languages:Go","Git","analytics","MeitY","purpose of artificial intelligence","R"],"extracted_tech_keywords":["AI","artificial intelligence","analytics","R","Go","Git","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/meitys-csc-to-use-artificial-intelligence-for-improvement-of-services-in-rural-areas\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":34053,"title":"Corporates Starting To Act Upon Growing Concerns Against Location Sharing","content":"Location data is posing to be one of the biggest concerns in the digital network today. Most users do not wish to share their location anymore, others still ignore it just because they think they are not alone. But no matter what they wish, they are sharing location data in one way or another. Some through Facebook, some through google maps and many many other. Tracking someone through his\/her cell phone network is no more nerdy genius stuff, practically anyone who can use a computer well can track anyone’s whereabouts with just the mobile number. Applications Of Location Data Today location data is one of the key players in digital market strategy. It is shared and accessed by many small and large organisations either to provide real-time service like maps or to sell more products through advertisements. Consider Facebook, for example, it knows a lot about you, your name, your date of birth, if you are in a relationship or not, the movies you watch, how long you spend on Facebook, the places you go to etc, and a lot more. All these data are put to use by Facebook, in suggesting you products through adds, suggesting you places to go to, suggesting people you may know. Just for the sake of growing your concern now think about how someone else may use these data. Data Flows Like Water Down A Hill According to media report cellular network companies like T-Mobile, Sprint, and AT&T are selling access to their customers’ location data, and that data is ending up in the hands of bounty hunters and others not authorised to possess it, letting them track most phones in the country. From the top companies to small buyers and third parties, the data flow. The report presents a piece of clear evidence as to how much exposed the mobile networks and the data they generate really are. The flow of data is never controlled, from the cell phone providers to third parties who may not even have a policy regarding the protection of privacy data. There is a risk, not just of being stalked by your ex-someone but of being spied by an assassin who just wants you killed, that’s the horrifying part we see in a lot of movies. Providers Vow To End Sharing Location Data In the past few years, there have been multiple reports from much-reputed news agencies regarding the misuse of sensitive data by many organisations. In the wake, some organisations have come to accept the fact and act upon it or at least comment. John Legere, the CEO of TMobile tweeted to Ron Wyden, a US Senator that T-Mobile is completely ending location aggregator work and they are doing it the right way to avoid impacting consumers who use these types of services for things like emergency assistance. He added that It will end in March. Reportedly AT&T stated that in the light of recent reports about the misuse of location services that they have decided to eliminate all location aggregation services — even those with clear consumer benefits. Verizon made a statement regarding its decision to end its data sharing agreements with companies that do roadside assistance. Also, Moving forward, customers will need to provide permission to Verizon to share their location with these firms. Conclusion As data grows limitlessly, the concern will outgrow data. The solution will be to either limit sharing location data or stop it completely, both of which are close to impractical considering the number of devices and applications that are solely built on location data. In the future, we may see more organisations coming up with strict policies to protect private data only for the sake of better customer experience.","excerpt":"Location data is posing to be one of the biggest concerns in the digital network today. Most users do not wish to share their location anymore, others still ignore it just because they think they are not alone. But no matter what they wish, they are sharing location data in one way or another. Some […]","categories":["AI Features"],"tags":["Google Map","location analytics"],"author_name":"Amal Nair","publish_date":"2019-01-24T09:59:17","publication_year":"2019","word_count":619,"keywords":["Go","programming_languages:R","programming_languages:Go","Git","location analytics","Google Map","GAN","R"],"extracted_tech_keywords":["R","Go","Git","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/corporates-starting-to-act-upon-growing-concerns-against-location-sharing\/","complexity_score":2,"technical_depth":6,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10162389,"title":"OpenAI Announces ChatGPT Gov","content":"OpenAI has introduced ChatGPT Gov, a specialised version of its AI platform tailored for US government agencies. This tool grants access to OpenAI’s most advanced models, including GPT-4o, to support public sector initiatives in areas such as national security, public health, and infrastructure. “Enabling the public sector, especially the US federal government, to leverage ChatGPT is critical to maintaining America’s global leadership in AI,” said CPO Kevin Weil on LinkedIn. ChatGPT for Government Designed to meet the unique demands of government operations, ChatGPT Gov is built with strict security and compliance measures, ensuring it aligns with federal regulations. It operates on Microsoft Azure’s commercial and government cloud platforms, meeting high-security standards like FedRAMP High, IL5, and CJIS. The platform offers a range of AI-powered capabilities, including real-time text interpretation, coding assistance, image analysis, and tools for developing and sharing custom AI models. IT teams can also manage users effectively with administrative controls, single sign-on (SSO), and compliance monitoring. More than 3,500 US federal, state, and local agencies have already used OpenAI’s technology for various applications, such as language translation in Minnesota, bioscience research at Los Alamos National Laboratory, and AI-driven process improvements in Pennsylvania. With ChatGPT Gov, OpenAI is expanding its presence in the public sector, recognising the growing role of AI in government operations. As more agencies adopt AI solutions, OpenAI aims to help them streamline workflows, improve efficiency, and enhance decision-making while adhering to stringent security and privacy requirements. This initiative underscores AI’s increasing influence on government services and highlights its potential to transform how public institutions operate. OpenAI’s Impending o3 Launch This follows closely on the heels of the sensational release of DeepSeek. Yesterday, OpenAI CEO Sam Altman acknowledged DeepSeek as a rising competitor, praising its new R1 model. While noting DeepSeek’s growing influence, he expressed confidence in OpenAI’s ability to develop even more advanced systems. OpenAI’s chief research officer, Mark Chen, also spoke about DeepSeek’s o1-level reasoning while pushing back against what he called ‘overblown’ reactions, particularly regarding cost efficiency. “Congrats to DeepSeek on producing an o1-level reasoning model. Their research paper shows they’ve independently discovered some of the same core ideas we did on our way to o1.” Many users are awaiting o3’s launch, which was soft-announced by the company in December. The company had said that while not yet publicly available, these frontier models were made accessible to researchers for public safety testing.","excerpt":"This comes right after DeepSeek’s big release, while OpenAI has yet to announce the launch of o3.","categories":["AI News"],"tags":["ChatGPT","OpenAI"],"author_name":"Aditi Suresh","publish_date":"2025-01-29T10:42:14","publication_year":"2025","word_count":399,"keywords":["Go","ChatGPT","OpenAI","AI","GPT-4o","ML","R","RAG","Aim","Azure"],"extracted_tech_keywords":["AI","ML","GPT-4o","ChatGPT","OpenAI","Aim","RAG","Azure","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/openai-announces-chatgpt-gov\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10010206,"title":"Complete Guide To Model Deployment Using Flask in Google Cloud Platform","content":"In real-world, training and model prediction is one phase of the machine learning life-cycle. But it won’t be helpful to anyone other than the developer as no one will understand it. So, we need to create a front end graphical tool that users can see on their machine. The easiest way of doing it is by deploying the model using Flask. In this article, we will discuss how to use Flask for the development of our web applications. Further, we will deploy the model on google platform environment. What is Flask? Flask is an API of Python that allows us to build up web-applications. It is designed for quick and easy initiation, with the ability to scale up to complex applications. Getting the dataset The dataset is taken from Kaggle, you can find it from this link . It has four attributes, describing a profile of a candidate and offering a salary to the candidate. Code Implementation Install all the packages required for this project. #Importing Train Dataset import pandas as pd import numpy as np data = pd.read_csv(\"hiring.csv\") data.head() From the dataset, we can see some of the values are missing. Replace those with median and mean value. data['experience'].fillna(data['test_score'].median(), inplace=True) data['test_score'].fillna(data['test_score'].mean(), inplace=True) The independent and dependent variable is stored by iloc method. X = data.iloc[:, :-1] y = data.iloc[:,-1] X We will convert the string value to an integer by defining it in a dictionary and calling the function. def convert(word): word_dict = {'one':1, 'two':2, 'three':3, 'four':4, 'five':5, 'six':6, 'seven':7, 'eight':8, 'nine':9, 'ten':10, 'eleven':11, 'twelve':12, 'zero':0,8:8} return word_dict[word] X['experience'] = X['experience'].apply(lambda x : convert(x)) Linear Regression is one of the simplest and most common supervised machine learning algorithms that data scientists use for predictive modeling. In this post, we’ll use linear regression to build a model that predicts the salary of an employee. from sklearn.linear_model import LinearRegression import pickle regressor = LinearRegression() #Fitting model with training data regressor.fit(X, y) # Saving model to disk pickle.dump(regressor, open('model.pkl','wb')) Flask Application The first step is to create the folder structure. • The design is done using the template folder that contains the HTML file. • Static folder contains the CSS file that gives look and feels to the website. • Python code is written in the app.py file. • For deployment of models in google cloud environment, requiremets.txt file and app.yaml are required. • This project is built on Spyder IDE. Python File: import numpy as np from flask import Flask, request, jsonify, render_template import pickle app = Flask(__name__) model = pickle.load(open('model.pkl', 'rb')) @app.route('\/') def home(): return render_template('index.html') @app.route('\/predict',methods=['POST']) def predict(): ''' For rendering results on HTML GUI ''' request_form = [int(x) for x in request.form.values()] features = [np.array(request_form)] prediction = model.predict(features) output = round(prediction[0], 2) return render_template('index.html', prediction_text='Employee Salary should be $ {}'.format(output)) @app.route('\/predict_api',methods=['POST']) def predict_api(): ''' For direct API calls through request ''' data = request.get_json(force=True) prediction = model.predict([np.array(list(data.values()))]) output = prediction[0] return jsonify(output) if __name__ == \"__main__\": app.run(debug=True) This code is stored in app.py file. Here we are importing the Flask module and creating a Flask web server from the Flask module. We are creating an instance of the Flask class and calling the web application. The POST request method is used for handling data between the client and the server. Local Model Deployment Open the command prompt. We need to give the location of the file and type python app.py on the terminal. The Screen should appear as below. Copy the URL and paste in the new tab to see the HTML page. Google Cloud Platform Model Deployment The first step is to create an account in google cloud. After signing in, the screen will appear as below. Select the Manage Resources tab in the app engine section. Here we can create a new project. We need to open the command prompt terminal. Set the project directory. Pick the configuration option. It will set up the network. The user must log-in with its credentials. Projects will be displayed on the screen. Since our current project is employeesalary we need to select that option. For deployment of the model type the keyword “gloud app deploy app.yaml –project employeesalary” as shown below. Now it will ask to set up the location. Here the option 4 is selected. Our model is ready for deployment. We need to paste the target URL in the new screen. The HTML page will appear as below. Enter the valid credential and hit the button predict. We should be able to see the predicted salary value on the page. The final step is to disable the application button in the app engine setting so that you won’t be charged enough. Final Thoughts In this article, we learned to create APIs for interacting with machine learning models. Further, we discussed how to deploy our model on both computer and google cloud platform. I hope you found this tutorial useful. Thank you for reading this article.The complete code of the above implementation is available at the AIM’s GitHub repository. Please visit this link to find the complete code.","excerpt":"In real-world, training and model prediction is one phase of the machine learning life-cycle. But it won’t be helpful to anyone other than the developer as no one will understand it. So, we need to create a frontend graphical tool that users can see on their machine. The easiest way of doing it is by deploying the model using Flask. In this article, we will discuss how to use flask for the development of our web applications. Further, we will deploy the model on google platform environment.","categories":["Deep Tech"],"tags":["Machine Learning","model deployment"],"author_name":"Ankit Das","publish_date":"2020-10-20T16:00:45","publication_year":"2020","word_count":840,"keywords":["NumPy","machine learning","TPU","AI","ML","model deployment","Machine Learning","Python","Ray","Aim","R","Pandas"],"extracted_tech_keywords":["AI","machine learning","ML","Aim","Ray","Pandas","NumPy","TPU","Python","R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/complete-guide-to-model-deployment-using-flask-in-google-cloud-platform\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":4508,"title":"SSN course in Business Analytics","content":"According to a report by the International Data Corporation in 2012, the digital universe occupied 2,837 exabytes (billions of gigabytes) in 2012, and by 2020, it is set to reach a towering 40,026 exabytes. In this era of booming digital data which impacts every aspect of a business, it has become a necessity for students to have the exposure and training to meet the growing demand for professionals in the area of Business Analytics. SSN Institutions, in a bid to prepare students for this fast growing sector, has come out with a six-month residential certificate course, Post Graduate Certificate Programme in Business Analytics (PGCBA). The first course in the recently launched SSN School of Advanced Career Education (SACE), the programme is set to begin in January 2014. “There will be a huge requirement for qualified data analysts to manage the data and draw business insights,” says Dr. Shashikant V. Albal, Director, SSN School of Advanced Software Engineering. “As per Gartner’s research, more than 4.4 million jobs in the area of Big Data analytics will be created by 2015.” The focus of the course being business intelligence, descriptive analytics, predictive analytics, prescriptive analytics and big data handling, Dr. Shashikant hopes to provide all the tools to students to enable them to take up various jobs in this domain. The faculty consists of academicians as well as practising professionals. Tie-up with IBM The highlight of this course is the tie-up of SSN Institutions with IBM for the business analytics laboratory. “In the lab, IBM will provide IBM SPSS, IBM Cognos, IBM InfoSphere and IBM Rational Software tools to ensure that the next generation professionals are equipped with the knowledge and skills for smarter business outcomes,” explains Dr. Shashikant. Sudhir Shastri, Technical Head- Software, IBM India, says the practicals have been carefully planned to provide real life project experience. The course has a Programme Advisory Council comprising senior executives in leading organisations in business analytics. Source: The Hindu","excerpt":"According to a report by the International Data Corporation in 2012, the digital universe occupied 2,837 exabytes (billions of gigabytes) in 2012, and by 2020, it is set to reach a towering 40,026 exabytes. In this era of booming digital data which impacts every aspect of a business, it has become a necessity for students […]","categories":["AI Trends"],"tags":[],"author_name":"AIM Media House","publish_date":"2014-01-19T08:00:42","publication_year":"2014","word_count":325,"keywords":["big data","business intelligence","programming_languages:R","AI","predictive analytics","Git","analytics","GAN","R"],"extracted_tech_keywords":["AI","analytics","predictive analytics","R","Git","big data","GAN","business intelligence","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/ssn-course-in-business-analytics\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10010747,"title":"Indian IT Sector Is Looking Up: Are AI &#038; Data Science Major Contributors","content":"While businesses in India are experiencing a downturn in its growth due to the pandemic, the one sector which is thriving despite all odds is the Indian IT sector. Reports suggest that the domestic IT services sector is likely to see single to double-digit revenue growth in 2021-2022. The IT and technology sector has done relatively well in sustaining momentum during the crisis and is expected to accelerate more as the economy revives. The COVID-19 pandemic has left the companies to ride the digitisation wave for which they are increasingly relying on the use of emerging technologies. Also, the companies are moving their workforces towards remote working with an increased reliance on cloud-based workflows. This transition from traditional to tech-based workflows calls for a need to adopt different strategies. Companies are looking at AI and analytics to speed up their digital transformations projects. Reports suggest that for most IT companies, the impact of the pandemic is likely to be moderate and short term. As businesses are looking to get into the digital transformational journey, IT firms will bounce back into its regular business flow. With digital transformation, automation and cloud-based solutions becoming the need of the hour to ensure business continuity, IT companies are fast moving towards these areas than the business process management and legacy applications side of it. Even IBM recently split its business division to focus solely on the hybrid cloud approach. The growth in the IT sector is also evident from the hiring that most domestic IT firms are going for. According to the recent news doing the rounds, India’s top four IT companies have hired over 17,000 people in the second quarter. These companies include TCS, Infosys, Wipro and HCL that are hiring in anticipation of the growth trajectory that they see ahead. In fact, the domestic IT sector has created the most number of jobs in the private sector in India. AI & Data Science — The Growth Driver Of IT Sector Analytics and data science capabilities have had a strong foundation in the Indian IT sector for the longest time. The data science domain has brought about transformational changes in many MNC and Indian IT companies over the years and are now emerging as a necessary capability transcending the entire business model and operation value chain of the firms. As the IT industry is witnessing growth, the AI and data science function can be credited for this. While many businesses have contracted due to the pandemic, a report by AIMResearch forecasted that there would be no significant operations contraction for the analytics’ functions of domestic and MNC IT firms operating in India. When we talk about digital transformation journey, analytics offering is firmly integrated with the digital offering across almost all IT firms. It is necessary to bring about crucial insights on digital markets, ad spending, enterprise digital investment, to name a few. According to numbers, for IT firms, the digital business now contributes 25-30% of the current revenue portfolio — for which analytics is a key driver. The same report suggests that over the next five years, the digital business is expected to grow to 35% of IT revenues across the industry. Not just in digital transformation, the internal analytics functions are also crucial in providing support to sales, client accounts, marketing, finance, and HR functions across IT organisations. The AI capabilities in IT companies have also begun to grow and show up its significance in terms of revenue. According to a report by AIM, IT and technology sector has the highest share of the AI market at 36.2% in percentage share, and $2314.3Mn in terms of market value. These numbers speak volumes about the significance AI domain has for the IT firms. Companies such as Accenture, Capgemini, TCS, Wipro, Tech Mahindra, HCL, among others, have a strong AI market share. Wrapping Up These numbers are indicative of the fact that the AI and analytics domain are strong players when it comes to deciding the overall growth of IT firms in India and at a global level. As the companies are intensely looking to go into the AI way to ride the digital transformation wave, the Indian IT sector will continue to look up in the coming years. Even if there are any contractions, given the strong involvement of analytics and data science in IT, they will make up for it in the coming years if not months.","excerpt":"While businesses in India are experiencing a downturn in its growth due to the pandemic, the one sector which is thriving despite all odds is the Indian IT sector. Reports suggest that the domestic IT services sector is likely to see single to double-digit revenue growth in 2021-2022. The IT and technology sector has done […]","categories":["IT Services"],"tags":["IT sector"],"author_name":"Srishti Deoras","publish_date":"2020-10-26T12:00:34","publication_year":"2020","word_count":736,"keywords":["IT sector","data science","Go","AI","ML","digital transformation","Git","Aim","analytics","GAN","R"],"extracted_tech_keywords":["AI","ML","data science","analytics","Aim","R","Go","Git","GAN","digital transformation"],"url":"https:\/\/analyticsindiamag.com\/it-services\/indian-it-sector-is-looking-up-are-ai-data-science-major-contributors\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10098850,"title":"Universal Music Group Grooves to YouTube&#8217;s AI Beats","content":"In April this year, Universal Music Group (UMG) compelled YouTube to remove an AI-generated song named “Heart on My Sleeve” from the platform, which had garnered millions of views. The song which cloned Drake’s voice was posted by an anonymous TikToker named Ghostwriter. Four months later, in a new development, UMG has partnered with YouTube to unveil the “Music AI Incubator.” In a blogpost by YouTube, the platform said that “Music AI Incubator” is working with some of music’s most innovative artists, songwriters, and producers across the industry, which includes stars like Anitta, OneRepublic’s Ryan Tedder, and the Frank Sinatra estate. To address the challenge of generative AI in music, YouTube has laid down the groundwork with three fundamental principles. The first principle simply acknowledges the generative AI and says generative AI in music is here, and the music industry needs to embrace it. In the second principle, the company says it will protect the creative work of artists on YouTube and provide monetary benefits to the artists who rightfully deserve it. Thirdly, YouTube says it will work towards creating new content policies to assure they meet challenges of AI. YouTube elaborated generative AI systems may amplify current challenges like trademark and copyright abuse, misinformation, spam, and more. The partnership between YouTube and UMG is radical in the sense that initially, the music label was fighting the battle against the video streaming platform to give due rights to artists. So what led to UMG’s sudden change of heart? Has UMG Shelved Ethics for Money? It seems like UMG has realised that in order to keep profits coming, they need to embrace generative AI. With the current hype, people are loving AI-generated songs and they are garnering millions of views on Youtube. In India, after Punjabi singer Sidhu Moosewala’s demise, his fans created a lot of AI songs which are currently present on YouTube. One of them is 4×4 which collected more than 4 lakh views. Nonetheless, the family members of Sidhu Mosewala expressed their disapproval of the AI-generated tracks, believing they have caused more harm than benefit to his legacy. In their statement, they emphasized that his unparalleled talent should remain unaltered. At the moment, the generative AI music industry is a bit messy, and not many music labels are giving it much attention. UMG could do things to make the AI music world more organized, so that the people who make the music get the credit they deserve. Right now, it’s kind of all over the place, with anyone making music and putting it on platforms like YouTube. Interestingly, back in April, UMG had asked streaming platforms like Spotify and Apple to stop AI tools from copying melodies and lyrics from their songs that have copyright protection. The current change of heart seems like there is a well-planned strategy to earn more money through royalties of generative AI music. From now on, YouTube and UMG will keep an eye on who is creating what and safeguards the interests of the artists. The decision to partner with YouTube for UMG is about mutual benefit. Any internet user outside of the label who produces generative AI music might face takedowns by UMG which will lead to AI music created by UMG will likely garner more views, ultimately resulting in increased revenue. This strategic shift by UMG underscores a notable change in their approach. UMG, which currently controls about a third of the global music market, has a 31.8% market share, followed by Sony Music’s 22.6% share, Warner Music Group with 15.6% share and independents’ 30% share. This partnership with YouTube will help it further cement its position as we venture into an AI dominated world. YouTube doesn’t want any copyright issues By partnering with UMG, YouTube is ensuring that UMG does not sue it in future for hosting AI-generated content which might invite copyright infringement. For Google, YouTube is a goldmine of data which is currently working on Bard and Gemini. If it restricts users from creating anything with generative AI, it will limit its data set. As much as Youtube cares for its creators and their copyright issues, it needs users to post new content every second which it can put to further use to train LLMs. Google recently changed its privacy policy which says that the company can use data on the internet. The new policy  explicitly states that Google is allowed to collect information that is publicly available online or from other public sources to train their AI models. UMG and Youtube being on the same page creates a win-win situation for Google as they strive to develop the most exceptional Language Model to challenge OpenAI. Advances in Generative AI based on Large Language Models (LLMs) and its application to diverse use cases is taking place at a breathtaking pace. These LLMs have been trained on billions of public domain artefacts, including copyrighted material, such as, Open-Source Software (OSS), images and songs. That has raised some interesting legal questions regarding copyright violations. For example, Microsoft, GitHub and OpenAI have already been sued for alleged copyright violation of OSS in their AI-powered coding assistant, Github Copilot. Similarly, earlier this year, Getty Images sued Stability AI for copyright violations of their images.   It is highly debatable if cloning a popular singer’s voice in an AI-generated song falls under fair use, especially when money is made from millions of views on social media platforms, such as, YouTube. Therefore, the announcement by YouTube that it is working closely with music partners, including Universal Music Group, to develop an AI framework towards common goals is a welcome move. The three principles outlined by YouTube will help in embracing generative AI in a responsible way, protecting creativity and copyrights of composers and performers, and investing in technology to make it happen at scale.   This is the just the beginning of a new chapter in the interpretation and adaption of copyright laws in the face of technology advancements.","excerpt":"YouTube is ensuring that UMG does not sue it in future for hosting AI-generated content","categories":["AI Features"],"tags":[],"author_name":"Siddharth Jindal","publish_date":"2023-08-22T17:04:10","publication_year":"2023","word_count":993,"keywords":["Go","AWS","OpenAI","AI","Git","ViT","generative AI","GAN","GitHub","R"],"extracted_tech_keywords":["AI","generative AI","OpenAI","AWS","R","Go","Git","GitHub","GAN","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/universal-music-group-grooves-to-youtubes-ai-beats\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10095375,"title":"Meta Releases Voicebox, a Multilingual Speech Generation Model","content":"Meta has been killing it in the generative AI field lately. In the latest announcement, Meta AI revealed a major breakthrough in the field of generative AI for speech, Voicebox, a multilingual model capable of performing various speech generation tasks through in-context learning, even tasks it was not explicitly trained for. Check out the paper here. Amid concerns around deepfake, the company claims that Voicebox is so competent that it has decided to not make it open to the public yet. Voicebox represents a significant leap forward as it can generate high-quality audio clips and seamlessly edit pre-recorded audio. Notably, it can remove unwanted background noise, such as car horns or barking dogs, while preserving the original content and style of the audio. Additionally, this versatile AI model is capable of producing speech in six different languages, making it a valuable tool for multilingual applications. JUST IN: Meta just introduced Voicebox!This is the first generative AI model that can synthesize speech across six languages, perform noise removal, edit content, transfer audio style & more.Highlights▸ Generalizes speech generation across tasks with impressive results and… pic.twitter.com\/cWABZWQvDK— Lior⚡ (@AlphaSignalAI) June 16, 2023 Meta says that this could lend natural-sounding voices to virtual assistants and non-player characters in the emerging metaverse. Furthermore, visually impaired individuals could benefit from AI-generated voices that read written messages in their familiar voices, while content creators gain new and efficient tools for audio track creation and editing in videos, among other exciting possibilities. In-context text-to-speech synthesis: By using a mere two-second audio sample, Voicebox can match the style and reproduce it for text-to-speech generation. Speech editing and noise reduction: Voicebox can seamlessly recreate interrupted speech segments caused by external noise or replace misspoken words without requiring a complete re-recording. For instance, it becomes akin to an “eraser” for audio editing, enabling users to identify and regenerate specific interrupted portions of speech, such as when a dog barks during a recording. Cross-lingual style transfer: Armed with a speech sample and a text passage in English, French, German, Spanish, Polish, or Portuguese, Voicebox can deliver a reading of the text in any of these languages, even if the sample speech and the text are in different languages. This breakthrough feature paves the way for natural and authentic communication between individuals who speak different languages. Diverse speech sampling: Voicebox has been trained on a wide range of data, resulting in the generation of speech that more accurately reflects how people naturally communicate in the real world and across the six supported languages. Meta has been dedicated to preserving the languages across the world. For this, the company had also released Massively Multilingual Speech AI research models consisting of 4,000 spoken languages for text-to-speech and speech-to-text capabilities. This comes after Meta releasing MusicGen, a transformer based music generation platform through text inputs which outperform Riffusion, Mousai, and MusicLM on various performance metrics. Moreover, Meta AI recently also announced I-JEPA, a self-supervised computer vision model for learning the world by predicting, which is based on Yann LeCun’s vision of autonomous machine intelligence.","excerpt":"Meta had recently also released Massively Multilingual Speech research model consisting of 4,000 spoken languages","categories":["AI News"],"tags":["Meta AI"],"author_name":"Mohit Pandey","publish_date":"2023-06-19T11:23:15","publication_year":"2023","word_count":508,"keywords":["Meta AI","AI","R","ML","virtual assistants","computer vision","Aim","generative AI","CLIP","in-context learning"],"extracted_tech_keywords":["AI","ML","computer vision","generative AI","Meta AI","Aim","in-context learning","virtual assistants","R","CLIP"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/meta-releases-voicebox-multilingual-speech-generation-model\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10053147,"title":"Major Announcements By Jensen Huang During NVIDIA GTC Keynote Speech","content":"“We are thrilled by the growth of the ecosystem we are building together and will continue to put our heart and soul into advancing it. Building tools for the Da Vincis of our time is our purpose. And in doing so, we also help create the future.” On the second day of the NVIDIA GTC 2021 event, CEO Jensen Huang presented the keynote speech, giving a presentation on its various projects. In the event that lasted close to two hours, Huang made announcements across sectors and domains which could transform many trillion-dollar industries. However,  simulation, Omniverse, virtual avatar featured very prominently, staying true to the Metaverse rage currently in the tech space. Some of the notable announcements are listed below. Omniverse — A Great Example of Digital Twin NVIDIA Omniverse is the company’s shot at creating an easily extensible and open platform for virtual collaboration and real-time and accurate simulation. Huang said in the keynote speech that with Omniverse, users will be able to create new 3D models of the physical world. With the newly announced Omniverse Avatar, developers will be able to create interactive characters that can see, understand, and interact on a wide range of topics. Further, the new update to Project Maxine allows Omniverse Avatar to connect computer vision, Riva speech AI and animation to a real-time conversational AI robot. Project Maxine can also add state-of-the-art video and audio features to virtual collaborations and content creation applications. Huang also introduced NVIDIA Omniverse Replicator, a synthetic data generation engine to train deep neural networks. There are two kinds of replicators — Omniverse Replicator for Issac Sim and Omniverse Replicator for DRIVE Sim for autonomous vehicles. Full Stack and Open Platform Huang spoke in length about Accelerated computing, which uses specialised hardware for exponentially speeding up the work by performing parallel processing of frequently occurring tasks. He said, “NVIDIA accelerated computing is a full-stack, data-centre-scale and open platform.” In this regard, Huang spoke about three chips — GPU, CPU and DPU — and systems — DGX, EGX, RTX and AGX — spanning from cloud to the edge. NVIDIA also announced 65 new and updated SDKs at GTC. Huang introduced NVIDIA Quantum-2, branding it as the most advanced networking platform ever built. He said that with BlueField-3 DPU, the company announces cloud-native supercomputing. Some of the features of Quantum-2 are high performance, broad accessibility, and strong security. Speaking of security, Huang also introduced a three-pillar zero trust framework to tackle cybersecurity challenges. Robotics Over the last four years, the number of companies and partners in the NVIDIA Issac ecosystem has grown to 700, a five-times rise. NVIDIA has announced the Isaac robotics platform to further equip partner companies, which can be easily integrated into the Robot Operating System (ROS), a set of widely used software libraries and tools for robot applications. For the NVIDIA Omniverse, Huang spoke about Isaac Sim, one of the most realistic robotics simulators available. Isaac Sim Replicator can help in generating synthetic data to train robots. This Replicator simulates the sensors and generates automatically labelled data. With a domain randomised engine, Replicator can create rich and diverse training data sets. AI Models and Systems The announcement of GPT-3 last year created major ripples in the AI world. Since then, many such language models have been introduced, each progressively larger and more advance than the previous. Now, Huang introduced Nemo Megatron to train large language models, which will reportedly be the biggest mainstream HPC application ever. Huang also introduced NVIDIA Modulus to build and train machine learning models that obey the laws of physics. Further, Huang introduced three new libraries — ReOpt, cuQuantum, and cuNumeric. Autonomous Vehicles Huang announced NVIDIA’s efforts in the autonomous driving space. NVIDIA DRIVE, the company’s full-stack and open platform for autonomous vehicles, will now have Hyperion 8. NVIDIA’s latest complete hardware and software architecture suite includes 12 cameras, nine radars, 12 ultrasonics, and a front-facing LIDAR; two NVIDIA Orin SoCs process these sensors. Hyperion also contains several new technologies built-in that includes Omniverse Replicator for DRIVE Sim. NVIDIA is currently running Hyperion 8 sensors, deep-learning-based multisensor fusion, feature tracking, and a new planning engine.","excerpt":"Jensen Huang said in the keynote speech that with Omniverse, users will be able to create new 3D models of the physical world.","categories":["Global Tech"],"tags":["Jensen Huang","jensen huang keynote speech","Speech Analytics"],"author_name":"Shraddha Goled","publish_date":"2021-11-10T14:00:00","publication_year":"2021","word_count":692,"keywords":["machine learning","jensen huang keynote speech","Speech Analytics","AI","neural network","AWS","Git","computer vision","Jensen Huang","RAG","GPT","Rust","R"],"extracted_tech_keywords":["AI","machine learning","neural network","computer vision","RAG","AWS","R","Rust","Git","GPT"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/major-announcements-by-jensen-huang-during-nvidia-gtc-keynote-speech\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10139591,"title":"Adobe Democratises ‘GenAI for Everyone’","content":"At the Adobe Max conference in Miami Beach, the company unveiled its brand new Adobe Cloud and over 100 new features across all its applications, including Adobe Photoshop, Illustrator, InDesign, and Premier Pro. These updates integrate Adobe’s latest GenAI model Firefly, giving creators more power and control over these software and designs. “I think, at Adobe, we’re trying to remind people that it’s a tool, so let it do the things that you don’t want to spend time on, like replacing heads for actors in their family portraits,” a designer at the conference said. Enhanced Workflows Adobe’s features Generative Remove and Generative Shape Fill, launched earlier this year, helped speed up workflows and gave users more control in Creative Cloud. These updates make the design process more intuitive, enabling creators to streamline the editing process by bringing their visions to life. The new features make GenAI tools available to all users, enabling even creators with non-technical backgrounds to integrate AI into their workflows. During the launch, one of Adobe’s designers highlighted how features like Vector Fill and Vector Expand will be particularly useful for graphic designers or illustrators for making logos or icons. Latest Adobe Firefly Video Model Adobe’s Firefly Video model allows creators to generate videos using simple text and image prompts. As proposed on Adobe’s official blog, users will be able to fill gaps, automate transitions and add dynamic visuals. The tool, integrated within Adobe Premier Pro and After Effects, will speed up the process without sacrificing creativity. Notably, as presented in the Adobe Max conference, users “need to own rights to images they use as a reference for GenAI.” Generative Extend will help extend clips to cover gaps in video and audio footage. It can create custom audio fades and crossfade multiple clips. It identifies and labels audio clips based on categories and also adjusts the intensity of sound using the Loudness Matching feature. Do more with Adobe Photoshop, Illustrator & InDesign The Beta version of Adobe Photoshop introduces two major improvisations. First comes a Generative Workspace for all, focusing on simultaneous brainstorming and image generation through text and reference images. The feature also claims to be able to input and generate new sets of images while the previous ones are still in process. Besides the fact that it’s going to be impossible to tell what’s real vs. fake, Adobe’s new generative AI update in Photoshop looks like a game changer.pic.twitter.com\/ix3UtpqVSA— Joe Pompliano (@JoePompliano) May 24, 2023 The second, newest addition is Adobe Substance 3D Viewer (Beta). It lets graphic designers view and edit 3D objects through Photoshop. The Remove tool updates itself into a Distraction Removal tool. It can automatically detect and remove unnecessary elements like electric wires and background people from photographs and generate custom backgrounds as needed. The existing Generative Fill and Generative Expand are now being powered by the latest Adobe Firefly image model to give a more realistic experience. The Generate Similar feature allows users to create new variations of an existing image. Designers elaborated in the conference, “Do whatever your heart desires, just type anything in and hit the button.” Adobe Photoshop gets new AI smarts with neural filters, sky replacement and more https:\/\/t.co\/omdehcm0D9 pic.twitter.com\/07LZfkRqOY— Eric Porat (@ericporat) October 20, 2020 As per an Adobe support page, “Neural Filter…empowers you to try non-destructive, generative filters and explore creative ideas in seconds.” In Illustrator, the new Objects on Path feature lets designers attach and move elements along any path. The Enhanced Image Trace function simplifies the conversion of hand-drawn images into editable vectors, creating refined designs with fewer anchor points. Additionally, the Mockup feature enables artists to apply their artwork to real-life objects with automatic adjustments for curves and edges. Meanwhile, InDesign introduces powerful GenAI capabilities like Generative Expand and Text-to-image, enabling intuitive image editing within layouts. For those working on scientific publications, the new MathML support ensures equations are visually appealing and accessible. As Adobe puts it, these tools “simplify common actions, and enhance collaboration — which all save you valuable time.” What is Project Neo (beta)? Project Neo (beta) was born out of the traditional time-consuming challenge for designers: bringing dimension into their designs. A little taste of the incredible work artists have done already in #ProjectNeo 🥹Join our private beta: https:\/\/t.co\/YJsSraUeAl pic.twitter.com\/LBAyklNDUx— Kowsheek (@kowsheek) May 1, 2024 This tool allows creators to move from traditional 2D constraints to 3D elements with the same fluency as in Adobe Illustrator. Users can explore designs from any angle and make real-time adjustments to light and shadow. Deepanshu Sharma, a tech enthusiast has been using Project Neo (beta) to bring depth and dimension to his designs, including logos, brand icons, infographics, and illustrations. He advertises creating vector artwork and volumetric designs in minutes, not hours. A Tool for Every Creator? — Frame.io Version 4 Adobe Frame.io Version 4 introduces several features aimed at improving the workflow for video and media professionals and allowing teams to collaborate more effectively. This is also now generally available to all. A new metadata model under this version helps users tag assets with custom labels, resulting in smart folders that automatically update according to user-defined criteria. The updated streaming player offers smoother playback with minimal buffering. Adobe notes that “V4 is built for everyone involved with content creation.” Frame.io’s integration with major camera brands, including Canon and Nikon, supports Camera Cloud workflows, facilitating quicker project turnarounds. The Future of Design Workflows One expert, elaborating on a “great use for AI”, observed, “You’re putting pieces of AI together to show someone a concept that’s in your head and then you actually create it.” The new updates reflect the future of design workflows. Developing AI technologies promise to streamline processes and relieve designers from repetitive tasks, allowing a greater focus on innovation. However, it’s important to note that while AI enhances certain capabilities, it doesn’t eliminate the necessity for human creativity. As industry experts highlight, AI will remain a tool for creative tech communities to combine technologies to develop new solutions. That said, some designers are discussing the bleak future of AI integration in this field. A Reddit user says, “The day AI conquers design is the day AI conquers all other industries.” This collaboration between human intuition and AI efficiency is expected to reshape design workflows vastly.","excerpt":"Do whatever your heart desires, say Adobe designers as they uncover the new AI features at the Adobe Max conference.","categories":["Deep Tech"],"tags":["Adobe","Adobe Firefly","GenAI"],"author_name":"Sanjana Gupta","publish_date":"2024-10-28T16:05:12","publication_year":"2024","word_count":1048,"keywords":["Go","GenAI","Adobe Firefly","AI","Adobe","innovation","ML","Aim","ViT","generative AI","CLIP","R"],"extracted_tech_keywords":["AI","ML","generative AI","GenAI","Aim","R","Go","CLIP","ViT","innovation"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/adobe-democratises-genai-for-everyone\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10071708,"title":"IIT Madras and Nandan Nilekani launches an AI centre on Indian Languages","content":"Indian Institute of Technology Madras (IIT Madras) is launching the ‘Nilekani Centre at AI4Bharat’ to advance Indian language technology with an aim to create a social impact. Rohini and Nandan Nilekani are supporting this centre with a generous grant of INR 36 crore through Nilekani Philanthropies. IIT Madras set up AI4Bharat as an initiative to build open-source language AI for Indian languages. For the past two years, the team led by Dr Mitesh Khapra, Dr Pratyush Kumar and Dr Anoop Kunchukuttan has contributed to Indian language technology, including state-of-the-art Machine Translation and Speech Recognition models. Commenting on the launch of this centre, Nilekani said, “The Digital India Bhashini mission has been launched with the goal of all services and information being available to citizens in their language with ‘collaborative AI’ at the design core. AI4Bharat will continue to contribute and accelerate the Indic language AI work as a public good and is fully aligned with the goals of the Bhashini mission.” Elaborating on the Centre, Dr Mitesh M. Khapra, associate professor, Department of Computer Science and Engineering, IIT Madras, said, “Given the rich diversity of languages in India coupled with a rapidly expanding digital world, it is important to make significant advances in language technology to benefit the commoner. While technology has significantly improved for English and other languages, Indian languages lag. The centre will focus on bridging this gap.” The centre has open-sourced advanced resources; these models are freely available and can be downloaded from their webpage. Speaking about the development at the centre, Dr Anoop Kunchukuttan, a researcher at Microsoft, said, “Building AI technologies for the diverse set of Indian languages is expensive, given the need for large datasets and computing power. We are grateful that several organisations have supported the efforts to build open-source AI.”","excerpt":"Nilekani Philanthropies is supporting the institute with a grant of INR 36 crores.","categories":["AI News"],"tags":["AI4Bharat","IIT","IIT Madras"],"author_name":"Tasmia Ansari","publish_date":"2022-07-28T15:42:30","publication_year":"2022","word_count":298,"keywords":["Go","API","programming_languages:R","AI","IIT Madras","programming_languages:Go","Git","Aim","GAN","AI4Bharat","IIT","R"],"extracted_tech_keywords":["AI","Aim","R","Go","Git","API","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/iit-madras-and-nandan-nilekani-launches-an-ai-centre-on-indian-languages\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10113419,"title":"273 Ventures Introduces KL3M, a Family of Legal LLMs for Enterprises","content":"Illinois-based 273 Ventures-owned Kelvin Legal Data OS has released KL3M, a family of LLMs designed for legal purposes. It is trained from scratch on legally permissible data for enterprise applications. KL3M1 is the largest legal model trained on over two trillion tokens of proprietary clean data, focusing on legal, financial, and general domains. This dataset, the Kelvin Legal DataPack, is commercially available and scored and filtered for continuous improvement using a custom pipeline. KL3M targets enterprise use in legal, regulatory, and financial workflows. Based on the performance metrics, including perplexity, toxicity, and human preference, KL3M demonstrates superiority over peer models. The initial models, kl3m-170m and kl3m-1.7b, outperform peers in perplexity and toxicity, with applications for tasks such as answering regulatory questions, drafting contracts, and extracting structured information, surpassing other models in detail and stylistic realism. Larger Mixture-of-Experts (MoE) models are also in development for release in the late first quarter. The models’ training data is ethically collected, avoiding fair use interpretations or contract breaches. The release aligns with the success of small language models (SLMs), prompting an accelerated roadmap. The models are designed to run efficiently on consumer-grade hardware like a MacBook Air or a $300 NVIDIA GPU. KL3M’s availability is tied to the Kelvin Legal Data OS, and interested parties can sign up for notifications. The creators seek collaboration for fine-tuning KL3M for other domains and testing it against existing SLM workflows. KL3M’s training involves a high-quality, curated English subset of the Kelvin Legal DataPack, and the models are based on the GPT-3 architecture with specific modifications.","excerpt":"The first two KL3M models are kl3m-170m and kl3m-1.7b, designed for real-time use on consumer-grade hardware.","categories":["AI News"],"tags":["LLMs"],"author_name":"Shritama Saha","publish_date":"2024-02-21T12:20:37","publication_year":"2024","word_count":258,"keywords":["programming_languages:R","AI","LLMs","RPA","SLM","GPT","small language models","R","llm_models:GPT"],"extracted_tech_keywords":["AI","small language models","SLM","R","GPT","RPA","llm_models:GPT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/273-ventures-introduces-kl3m-a-family-of-legal-llms-for-enterprises\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10099447,"title":"Musk’s X Slapped with 2,200+ Unpaid Salary &#038; Bonus Cases, Worth Millions","content":"After Musk acquired Twitter (now X) for $44 billion, he made a series of dramatic changes including ditching the ‘blue-tick’ verification; abruptly rebranding the platform’s name to ‘X’, making the blue bird extinct and purging nearly two third of his employees. However, these bold moves have since thrust the company into a legal area, as it now faces over 2,200 claims — estimated $3.5 million as reported by CNBC. The legal battle has come to light through court documents filed in the case of Chris Woodfield v. Twitter, X Corp., and Elon Musk (No. 1:23-cv-780-CFC). The fee for each filing in this case stands at a substantial $2,000, in accordance with the JAMS arbitration system, of which $400 will be borne by the former employees themselves. By January 2023, a mere three months after Musk’s acquisition, the court had already received 200 arbitration demands, a number that has since ballooned to over 2,200 by August. Adding fuel to the fire, in June, Shannon Liss-Riordan, an attorney involved in a proposed class-action lawsuit against the company, alleged that X had failed to disburse “tens of millions of dollars” in bonuses owed to its employees. This complaint is only one of the lawsuits from disgruntled former Twitter employees, all claiming that Musk did not uphold the commitments he made during the takeover. Experts expect that the number of legal cases may continue to rise, while X’s legal expenses mount as it defends its position. In response to the increasing legal pressure, the legal team has argued that X did not mandate its employees to address disputes through arbitration, thereby disclaiming responsibility for the former employee’s share of the filing fees. X’s legal woes go beyond internal disputes. In July, the company filed a lawsuit against the Center for Countering Digital Hate, a hate speech watchdog organisation. Simultaneously, X initiated proceedings against Wachtell law firm previously employed by Twitter’s prior administration to ensure Musk upheld his end of the acquisition deal. The lawsuit saga does not end here as the company has faced multiple lawsuits ranging from unpaid bills to copyright issues.","excerpt":"The fee for each case stands at $2,000, of which $400 will be borne by the former employees themselves.","categories":["AI News"],"tags":["Elon Musk"],"author_name":"Tasmia Ansari","publish_date":"2023-09-04T18:20:25","publication_year":"2023","word_count":349,"keywords":["Go","AWS","AI","cloud_platforms:AWS","Git","Elon Musk","GRU","Aim","Rust","GAN","R"],"extracted_tech_keywords":["AI","Aim","AWS","R","Go","Rust","Git","GAN","GRU","cloud_platforms:AWS"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/musks-x-slapped-with-2200-unpaid-salary-bonus-cases\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":861,"title":"How the rapid expansion of IoT in healthcare is transforming the mHealth landscape","content":"The number of smartphone users have already surpassed a staggering 2 billion mark, and it is hard to ignore the potential of mHealth with such advancements in technology. World over the mobile health market will reach $49 billion by 2020, with a compound annual growth rate of 48% from 2013 to 2020 according to Grand View Research. Rising number of chronic diseases like cancer, heart disease and diabetes together with aging population are driving the market space along with new connected medical devices. Lack of healthcare providers coupled with expanding rural population could make the Asia Pacific market to take lead in this growth, predicts the research study while North America will still hold a significant market share due to an increase in smartphone adoption. Top mHealth apps influence patient behavior a great deal and are staring to make significant difference in terms of patient satisfaction and preventive care. World over approximately top 10 mHealth apps generate 4.3 million downloads per day; 80% of physicians use smartphones and mHealth apps while 43% of them use mobile health technology for clinical purposes and 93% of physicians believe there is value having mHealth app these days. An overwhelming percentage of patients still tend to keep a pile of medical papers of their personal health information, which unfortunately they find it hard to trace when they actually need them. Given a choice, most of senior citizens would prefer to live independently, free of worries, in an assisted living environment or retirement community. To a large extent, a digitalized personal health maintenance platform provides remedy for a healthcare consumer by making credible and authentic information accessible anytime! With a push of button, you can get a desired service delivered to your home like medication refill, blood sample collection, homecare services, etc. Oftentimes, while caring for a chronically ill parent or grandparent, caregivers especially the children undergo great amount of emotional toll and physical exhaustion for want of their physical presence with them for various support needs. A smart health management platform allows them to monitor and stay connected with their loved ones while they get on with their chosen career. Therefore, mobile health platforms allow faster, more streamlined and meaningful communication between all members of the healthcare ecosystem from both the clinical and remote settings. As a result, it brings about an overall improvement in patients’ lifestyle and adherence with physician instructions, and subsequently a chance for an improved clinical outcome and quality of life.","excerpt":"The healthcare industry is experiencing a paradigm shift as patients take on a much larger role in their personal health and connected devices begin to impact the way they interact with their care providers.","categories":["IT Services"],"tags":[],"author_name":"AIM Media House","publish_date":"2016-12-07T05:03:13","publication_year":"2016","word_count":411,"keywords":["Go","programming_languages:R","AI","RPA","ML","programming_languages:Go","Git","R"],"extracted_tech_keywords":["AI","ML","R","Go","Git","RPA","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/rapid-expansion-iot-healthcare-transforming-mhealth-landscape\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10040475,"title":"The Need For More Women In Tech: In Conversation With Mathangi Sri &#038; Dr Sayantani Roy Choudhury","content":"Only a third of the global IT workforce is women, and the number is dwindling. The gender gap in the new age, in-demand and high paying skills like data science and data engineering are even sharper. On the other hand, multiple studies conclude that diversity, including gender diversity, is just plain good for business. In 2019, the top quarter of companies for gender-diverse executive teams were 25% more likely to have above-average profitability than their counterparts. Given the digital world we live in today, it becomes even more important that women participate in larger numbers in the tech and data domains. In order to encourage and support women participation in tech, data and management careers, Praxis Business School, the pioneers in Data Science education, have launched the Praxis Women in Tech (WiT) Scholarship Programme across their management, data science and about-to-be-launched data engineering programmes. On the occasion of the third edition of The Rising 2021, Analytics India Magazine met up with two women who have embraced tech and data in their work-life to ask them about the need for women in technology and the impact scholarship schemes like the one Praxis has launched will have on this. The two women are Mathangi Sri, Head of Data – GoFood at Gojek and a recipient of the AIM Top 10 Data Scientists in India honour, and Dr Sayantani Roy Choudhury, Associate Professor at Praxis Business School and a recipient of the AIM Best Women Faculty (in Data Science) Awards. You are both ‘women in tech’ — readers would be keen to know about your journey and what it has meant for you. Mathangi: I began my journey in GE Consumer Finance (currently in Genpact) and then worked in 247.ai. Subsequently, I have set up and led AI and data science teams across many tech start-ups. Currently, I head the data team at the GoFood — part of Gojek, where we are constantly enhancing the food experience of users through building large scale personalization algorithms at scale. I have evolved with the evolution of the data science space, and it has been a thoroughly enjoyable journey. Sayantani: My journey in data and tech started with an academic career in economics and my focus on econometrics across my masters, PhD and subsequent research work in the social sciences domain. My association with Praxis Business School has helped me expand my area of interest and expertise from statistics and econometrics to newer concepts like machine learning. I am passionate about research and teaching, and data and tech are at the centre of what I do today. Why do you think ‘women in tech’ is an important concept? Why should more women join the tech field, especially for new-age skills like data science, data engineering etc.? Mathangi: Today, we see very little participation from women in tech. Diversity is extremely important as innovation fosters an unbiased setup. We should certainly encourage women to make careers out of logic and mathematics. There is enough talent out there — it’s the exposure and orientation needed to get more women into tech. Sayantani: Four reasons — Women’s participation in tech is very low, with only 25% of Indian tech students and less than 25% of data science professionals are women. The demand for data science and data engineering professionals will increase by more than 30% and 50%, respectively, over the next couple of years. So, there are plenty of high paying jobs for the taking! Enterprises seek a better gender balance and struggle to find women who can take up tech roles in inadequate numbers, which is an opportunity for aspiring women data scientists, data engineers and tech-savvy managers. In addition to driving innovation, gender diversity will help reduce algorithmic bias in AI models, making them more ethical and fair. What do you think are the biggest hurdles that women face in joining and excelling at tech? Mathangi: Girls should get more encouragement, exposure and direction at home and school to participate in ‘STEM’ areas. India has the potential to be the tech capital of the world, and women should be motivated to be part of this tech revolution. Encouraging girls to do hackathons could be a starting point. The flexibility offered by remote working could see women technologists emerging in leadership positions to help reduce early drop-offs from the workforce. Sayantani: Hurdles start from an early stage — a 16% gender gap in literacy rate, 14% in higher education enrolment and 70% in tech education enrolment. Girls do not get the necessary exposure or the moral and financial support from their families. Secondly, many women who join tech careers drop off post marriage when they start families. Some women fight these odds and emerge strong, and we need to put systems in place to encourage and facilitate these women’s participation in tech. Do you think initiatives like the Praxis WiT scholarship can drive more women to look at a tech\/ data career? Mathangi: Yes, it can certainly work as a trigger for young women to look at tech as a career option. I have taught regularly at Praxis, and I know the quality and rigour Praxis brings into its programmes and the kind of opportunities its students get. This is a great time to be part of the data science world and make a difference in how businesses are running, and a scholarship-supported education at Praxis is a great way to start the journey. Sayantani: I am a researcher in women studies and a strong advocate of women empowerment. The Praxis WiT scholarship is a wonderful initiative by our college and will help create learning opportunities and employability among our young women. As we are giving the option of attending this course online this time, we should have greater participation of women who have dropped off their careers and want to restart. Our social responsibility is to bridge the gender gap, and this scholarship programme is a step in the right direction. What would be your message to the young women of our country? Mathangi: Skill and intelligence are important. However, what is more important, is perseverance and the ‘madness’ to accomplish and make a difference. Odds can be stacked against you, but there are means to get around those odds and reach greater heights. Tech, especially Data Science and Data Engineering, are great areas, and the industry is looking out for more women. Sayantani: I want women to be financially independent and have the freedom to achieve their dreams. I would like to motivate them to join the tech and data world. It is exciting, there are lots to learn, and as Mathangi says, the industry is struggling to get the right talent and the right diversity of the workforce. And, of course, grab opportunities like the Praxis Women in Tech scholarship to get educated in the hot skills of the digital world. It is quite apparent that more women in tech are good for tech, for the country and the world. The Women-in-Tech scholarship launched by Praxis Business School, ranked 2 in the list of full-time Data Science Programs in the country by AIM, is a great step in encouraging and empowering women to join and build exciting careers in the tech world. Do have a look at the following programs: Know more about the Praxis Women in Tech Scholarship Apply to the Data Science program. Apply to the Data Engineering program.","excerpt":"Women’s participation in tech is very low, with only 25% of Indian tech students and less than 25% of data science professionals are women.","categories":["AI Features"],"tags":["Interviews and Discussions","Mathangi Sri","Praxis Business School","praxis data science programme","scholarship","Women in Analytics","Women in Data Science","Women in Tech"],"author_name":"AIM Media House","publish_date":"2021-05-20T17:00:00","publication_year":"2021","word_count":1236,"keywords":["data science","Go","Women in Data Science","machine learning","Mathangi Sri","API","AI","Git","RAG","praxis data science programme","Aim","Women in Tech","analytics","scholarship","Women in Analytics","R","Praxis Business School","Interviews and Discussions"],"extracted_tech_keywords":["AI","machine learning","data science","analytics","Aim","RAG","R","Go","Git","API"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/the-need-for-more-women-in-tech-in-conversation-with-mathangi-sri-dr-sayantani-roy-choudhury\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":66219,"title":"How This Bangalore-Based Startup Is Helping Professionals to Optimise The Design Process","content":"“We provide India’s largest 3D prototyping-built volume and can print any model from simple to complex designs.” – Reethan Doijode, Co-Founder of HyCube. Innovative hardware and IoT startups are gaining more attention and becoming mainstream in India as many ventures are actively looking to develop and assemble their products from scratch. Along with this, 3D printing has gained popularity globally because of the diverse applications it offers. The technique is also gaining prevalence in India as large organisations are leveraging it for research and development and making end-user applications. With a similar vision, Bangalore-based HyCube Works started their innovative approach towards product development using cutting-edge technology. For this week’s feature, Analytics India Magazine got in touch with Reethan Doijode to gain more insights on the 3D technology and how HyCube Works delivers these products for utilisation. Founded in 2018 by Reethan Doijode and Shreyas SP, HyCube Works is a rapid prototyping and additive manufacturing company that aims to deliver tested and refined products for the world to utilise. Flagship Product HyCube Works is involved in direct digital manufacturing solutions, sales, and service of desktop and industrial-grade ALPI series of 3D printers and rapid prototyping services, with the ultimate goal of creating complex parts and handholding designers to validate designs and develop various parts for end-use applications. The ALPI series of 3D printer, designed and fabricated at the research facility enables various industries and professionals to optimise the design process. According to Reethan, HyCube Works’ ALPI series of 3D printers are the real game-changing choice, with the highest levels of plug and print reliability and repeatability. HyCube Works manufactures state-of-the-art industrial-grade, desktop and customised 3D printers that are reliable, superior in quality, cost-effective and user-friendly. The company also provides prototyping services for divisions mainly concentrating on product development, automobile, architecture, medical, engineering, interior décor, R&D, educational and for hobbyists. Use Of AI & ML At HyCube Works According to Reethan, 3D printing is a game-changing technology, which is constantly evolving and finding new ways to improve the manufacturing industry. Today, it includes new amazing technologies like artificial intelligence and IoT, a combination of artificial intelligence and 3D printing could lead to new applications of the additive manufacturing technology. “Our AI-based 3D printing technology, with a smart extruder, allows detecting any problems while 3D printing. It is also able to make autonomous decisions with notification,” Reethan said. “In order to 3D print your project, you will be required to work on your 3D model using CAD software. But, AI will assist in these 3D modelling programs to help you create the best 3D printable products for testing and endue. Thanks to IoT, now we can control the ALPI 3D printers remotely, verify the 3D printing temperature, turning “ON” and “OFF” with real-time monitoring from any corner of the world. IoT and 3D printing are two important new technologies, which progressively impact a lot of areas of the industries and also our everyday life,” he added. What Is The One Thing That Sets HyCube Works? To this question, Reethan replied that HyCube Works stands out by offering industrial quality at a consumer price. Big, medium and small corporations are buying 3D printers for internal R&D work globally. HyCube Works focuses on consumer experience\/service as well as the ALPI series of the 3D printers to deliver the highest quality, reliability, and repeatability. Besides, they offer on the spot customer services and provide superior print quality through their customised assembly. Tackling Talent Crunch Talking about hiring, Reethan said, “We conduct technical sessions in various institutions through which we give an opportunity to youngsters and professionals who have the quality of problem-solving, they work or intern with us until we explore the core knowledge and join the team.” Future Roadmap HyCube Works is delivering low-volume manufacturing capabilities to its customers leveraging on its digital manufacturing back-end. In the coming year, the company has various upcoming products in the pipeline for bio-medical and metal manufacturing across various industry verticals, researching and developing disruptive applications that are more reliable and productive for rapid prototyping.","excerpt":"“We provide India’s largest 3D prototyping-built volume and can print any model from simple to complex designs.” – Reethan Doijode, Co-Founder of HyCube. Innovative hardware and IoT startups are gaining more attention and becoming mainstream in India as many ventures are actively looking to develop and assemble their products from scratch. Along with this, 3D […]","categories":["AI Startups"],"tags":["AI Startups"],"author_name":"Ambika Choudhury","publish_date":"2020-05-31T18:00:00","publication_year":"2020","word_count":676,"keywords":["Go","API","artificial intelligence","AI","ML","Git","RAG","Aim","analytics","R","AI Startups"],"extracted_tech_keywords":["AI","artificial intelligence","ML","analytics","Aim","RAG","R","Go","Git","API"],"url":"https:\/\/analyticsindiamag.com\/ai-startups\/how-this-bangalore-based-startup-is-helping-professionals-to-optimise-the-design-process\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10100807,"title":"Gmail will Become Less Accessible for Some from 2024","content":"Software giant Google is sending Gmail’s basic HTML view, a minimalist version of the popular email service to the graveyard. Starting in January 2024, all users of the HTML view will be automatically switched to the feature-rich “Standard” view, marking the end of an era for those who valued its simplicity and speed. While the majority of Gmail users have long embraced the Standard view on their personal computers, the HTML version had its unique advantages. This stripped-down version offered lightning-fast loading times and accessibility for older machines or slower internet connections. Its lean design made it particularly valuable in situations where network conditions were less than ideal. While the company will officially shut down the service in 2024, there are already reports of difficulties in setting the Basic HTML mode, even before the code’s formal deprecation date. A Step Backward The upcoming Gmail transition has its critics, particularly among visually impaired users. Pratik Patel, a blind technologist and advocate for accessibility, expressed concerns about the move. He argued that many blind individuals relied on Gmail’s HTML view for its efficiency and simplicity. The Standard view, according to Patel, posed usability challenges due to its complex design elements and inconsistent navigation patterns. Patel emphasized that blind and partially sighted users often found it quicker to accomplish tasks using the HTML interface than the Standard one. He called on Google to engage with its users to address these usability issues and ensure that accessibility concerns were adequately considered. While the long-loved HTML is being bid adieu to, for people using older hardware switching to lightweight clients like Mozilla Thunderbird and Microsoft Outlook is generally preferable to loading web-based platforms. For people who need accessibility features, Thunderbird is known to work well with the Jaws, NVDA, and Windows Eyes screen readers and offers a range of display and text size adjustment options that enhance usability for people with visual impairments. While progress is inevitable, it should not come at the cost of leaving users, especially those with specific accessibility needs, behind. The company’s decision to retire the HTML view underscores the choices tech giants make as they navigate the shifting terrain of user expectations and AI race. Blinded by AI A Google spokesperson defended the decision, stating that the HTML view was replaced by its modern counterpart over a decade ago and lacked many of the advanced features found in today’s Gmail. But the HTML view was never intended to match the full functionality of modern email clients, and its retirement aligns with Google’s ongoing efforts to enhance Gmail’s capabilities through AI features. Moving away from the early era of the internet, the decision to retire the minimal view is part of the company’s broader strategy to infuse AI-powered features into its products, including Gmail. In recent months, the company has launched AI features like Duet AI to assist users in composing emails and integrated the Bard chatbot into Google accounts for email-related inquiries. With a focus on AI integration, the tech goliath’s move away from the simplified version reflects its ongoing evolution. However, some argue that the company’s reluctance to innovate meaningfully in recent years may have led to a loss of investor confidence. While Google remains a profitable corporation, questions linger about its ability to adapt in an ever-changing tech landscape. The company has been on an AI innovation spree since the Microsoft-backed OpenAI gained popularity in Silicon Valley. Google has been trying to keep up with the AI wave but nothing has worked in its favour yet. The retirement of Gmail’s basic view is the latest addition to Google’s growing graveyard of discontinued products and services, including the Pixel Pass phone upgrade program, Google Currents, and Nest Secure. As Google continues to refine its offerings, it continues to struggle balancing innovation with the needs of its diverse user base, including those who valued the simplicity of the now-retired HTML view.","excerpt":"Moving away from the early era of internet is Google’s strategy to infuse AI-powered features into Gmail","categories":["AI Trends"],"tags":[],"author_name":"Tasmia Ansari","publish_date":"2023-09-28T15:00:00","publication_year":"2023","word_count":651,"keywords":["Go","AWS","OpenAI","AI","RPA","ML","innovation","ViT","llm_models:Bard","R"],"extracted_tech_keywords":["AI","ML","OpenAI","AWS","R","Go","ViT","RPA","innovation","llm_models:Bard"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/gmail-will-become-less-accessible-for-some-from-2024\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10172123,"title":"2025 is NOT the Year of AI Agents","content":"Andrej Karpathy is back, this time explaining how LLMs are rewriting software. At YC AI Startup School, the former head of AI at Tesla gave a talk titled “Software Is Changing (Again),” during which he discussed with students and developers how the concepts of code, computation, and programming are being rethought at a fundamental level. He defined three kinds of Software. The first, Software 1.0, consists of traditional programming, in which humans write explicit instructions for computers to execute. Karpathy said that in Software 2.0, instead of writing code manually (as in Software 1.0), developers work with neural networks, specifically by tuning datasets and using an optimiser (like gradient descent) to learn the weights or parameters of the neural network automatically. While the third is Software 3.0, where LLMs have made neural networks programmable in a new way. Instead of writing traditional code, users now write prompts in natural language like English, which effectively serve as programs that instruct the model. “I think a quite fundamental change is that neural networks have become programmable with large language models. I see this as something new and unique, it’s a new kind of computer. In my mind, it’s worth giving it a new designation, Software 3.0,” Karpathy said. He also discussed the rise of vibe coding in recent months and how its growing popularity among kids gives him hope for an exciting future. Karpathy shared a few apps he built while vibe-coding, like MenuGen (menugen.app), which turns menu text into visuals to help users make sense of it. Human in the Loop While LLMs may eventually be able to browse, click, and navigate the web more like humans do, Karpathy believes it’s still valuable to meet them halfway. He said humans should generate content in a format that can be easily understood by LLMs. Karpathy gave the example of Gitingest, which turns any Git repository into a simple text digest of its codebase. This is useful for feeding a codebase into any LLM. He referred to the next wave of software as partial autonomy apps built on LLMs, where humans continue to play a key role in oversight and control rather than handing over full autonomy. “We have to keep the AI on the leash. A lot of people are getting way overexcited with AI agents,” said Karpathy. “When I see things like ‘2025 is the year of agents,’ I get very concerned… this is the decade of agents,” he added. He urged the developers to build augmented systems, like Iron Man suits, not Iron Man robots that accelerate human productivity without removing human oversight, as LLMs are still fallible. Referencing his work on Tesla’s Autopilot, Karpathy pointed out that despite years of development, full autonomy has not yet been achieved, even in vehicles that appear driverless. “There’s still a lot of teleoperation. We haven’t declared success.” Karpathy referred to LLMs as “people spirits”—superhuman in some ways (like memory or general knowledge) but deeply flawed in others (like hallucinations, logical inconsistencies, or context retention). He said they simulate intelligence but don’t develop knowledge over time like a human would. Instead, they rely on fixed weights and short-term context windows, which he compares to working memory. He cited tools like Perplexity AI and Cursor as examples of intelligent orchestration of multiple LLM components behind the scenes and mechanisms for human-in-the-loop verification. Crucially, these apps also offered what Karpathy called an “autonomy slider,” allowing users to control how much freedom the AI had depending on the complexity and risk of the task. Build for the Agents Karpathy said we need a new interface built specifically for agents. He explained that a new kind of software user has arrived—neither a person clicking through a GUI nor a backend system making API calls. Cool demo of a GUI for LLMs! Obviously it has a bit silly feel of a “horseless carriage” in that it exactly replicates conventional UI in the new paradigm, but the high level idea is to generate a completely ephemeral UI on demand depending on the specific task at hand. https:\/\/t.co\/Xgh1iwDmJl— Andrej Karpathy (@karpathy) June 19, 2025 Instead, LLMs represent something in between. Karpathy described them as the third major consumer and manipulator of digital information, urging developers to start designing with them in mind. Traditionally, software has served two users. Humans through graphical interfaces and computers through APIs. But LLMs occupy a new space. “There’s a new category of consumer,” he said. “Agents. They’re computers, but they are humanlike. People’s spirits on the internet.” Karpathy said that whenever he uses ChatGPT, he feels like he is talking to an operating system through the terminal. He believes that it should have a new GUI, other than just a text bubble. LLMs Resemble Fabs and Utilities He further compared the development of LLMs to semiconductor manufacturing. Building advanced LLMs, he said, involves massive capital investment, proprietary methods, and tightly integrated R&D, similar to running a chip fabrication facility. “The capex required for building LLMs is actually quite large,” he said. “We have deep tech trees, R&D secrets, centralised in LLM labs.” Beyond hardware analogies, Karpathy’s central argument is that LLMs are evolving into full-fledged operating systems. These models coordinate memory, computation, and interaction much like a traditional OS. “The LLM is a new kind of computer—it’s like the CPU. Context windows are like memory. And the LLM orchestrates memory and compute.” He pointed to applications like Cursor that can run on any major foundation model like GPT-4, Claude, and Gemini as examples of this platform-agnostic future. “You can take an LLM app like Cursor and run it on GPT or Claude, or Gemini. That’s kind of like downloading an app and running it on Windows, Linux, or Mac.” We’re Back in the 1960s of Computing At present, LLMs remain centralised and expensive to run, which Karpathy compared to the mainframe era of the 1960s. Instead of personal computers, we’re using interfaces like ChatGPT that tap into vast cloud-based models. “LLM compute is still very expensive, so they’re centralised in the cloud and we are all just thin clients interacting with it.” He noted early signs of a shift. Some developers are already experimenting with running smaller models locally on consumer hardware like Mac Minis, but a true personal computing revolution for LLMs is still far off. Karpathy likened LLMs to electricity: centralised, metered, and essential. Labs like OpenAI and Anthropic invest heavily in training their models, then serve intelligence over APIs, much like utilities deliver power. When these services go offline, the impact is immediate. “It’s like an intelligence brownout. The planet just gets dumber for a while.” But unlike electricity, LLMs are not bound by physical laws. They are shaped by data, architecture, and training methods. This flexibility changes how we build, share, and improve them, turning LLMs into more than just a utility. They’re becoming a programmable layer of intelligence for the internet.","excerpt":"Andrej Karpathy said that whenever he uses ChatGPT, he feels like he is talking to an operating system through the terminal.","categories":["Global Tech"],"tags":["AI Agents"],"author_name":"Siddharth Jindal","publish_date":"2025-06-20T17:29:01","publication_year":"2025","word_count":1151,"keywords":["Anthropic","ChatGPT","Go","OpenAI","AI","neural network","AWS","AI Agents","Git","CuPy","R"],"extracted_tech_keywords":["AI","neural network","ChatGPT","OpenAI","Anthropic","CuPy","AWS","R","Go","Git"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/2025-is-not-the-year-of-ai-agents\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10134462,"title":"Anthropic Launches Claude Enterprise Plan with GitHub Integration","content":"Anthropic has launched the Claude Enterprise plan, offering organizations enhanced capabilities for secure collaboration using internal knowledge. The plan introduces a 500K context window, significantly expanding Claude’s ability to handle large datasets such as sales transcripts, lengthy documents, and codebases. “When you combine expanded context windows with Projects and Artifacts, Claude becomes an end-to-end solution to help your team take any initiative from idea to high-quality work output,” said the company. This is coupled with a native GitHub integration, currently in beta, allowing engineering teams to sync repositories and work alongside their codebases. “GitHub is the first of the native integrations we’re building to connect Claude to your most important data sources, enabling Claude to provide more relevant and insightful assistance. This feature is available in beta for early Enterprise plan users today. We plan to make it more broadly available later this year,” said the company in a blog post. The Claude Enterprise plan includes critical security and administrative features designed to protect sensitive data. These include single sign-on (SSO), role-based access with fine-grained permissioning, and forthcoming features like audit logs and System for Cross-domain Identity Management (SCIM) for automating user provisioning. Anthropic emphasises that Claude does not train on customer conversations or content, ensuring data privacy. The company aims to broaden the availability of these enterprise features later this year. Organizations interested in adopting the Enterprise plan are encouraged to contact Anthropic’s sales team for more details. Anthropic recently made Claude Artifacts available to all users on iOS and Android, allowing anyone to easily create apps without writing a single line of code.","excerpt":"The plan introduces a 500K context window, significantly expanding Claude’s ability to handle large datasets such as sales transcripts, lengthy documents, and codebases.","categories":["AI News"],"tags":["Anthropic"],"author_name":"Siddharth Jindal","publish_date":"2024-09-04T20:59:59","publication_year":"2024","word_count":265,"keywords":["Anthropic","TPU","AI","Git","llm_models:Claude","RAG","Aim","GAN","GitHub","R"],"extracted_tech_keywords":["AI","Anthropic","Aim","RAG","TPU","R","Git","GitHub","GAN","llm_models:Claude"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/anthropic-launches-claude-enterprise-plan-with-github-integration\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10047235,"title":"AI Now Builds 3D LEGO From 2D Images","content":"Building customised builds that can match the complexity exhibited by the real-world or imagined scenes such as pictures, sketches, or physical objects was a challenge. A group of researchers from MIT have come up with a solution. The paper titled ‘Image2Lego: Customised LEGO Set Generation from Images’ unveils Image2Lego. LEGO is a Danish consumer product company engaged in manufacturing toys for children, video and online games. It also has products for children to help them improve their creativity and innovative playing and learning activities. The 3D model generation from images and LEGO generation from 3D models has always remained an uphill task. The model presented in the paper has bridged this gap and come out with a complete Image2LEGO pipeline. Hence, it opens the gate for those who wish to create custom LEGO models from 2D images. Simply put, Image2Lego is an AI to create buildable 3D LEGO from single 2D images. The work includes three sequential components: converts a 2D image to a latent representation,decodes the latent representation to a 3D voxel model, andapplies an algorithm to transform the voxelised model to 3D LEGO bricks. Methodology The presented model is the combination of: Octree Generating Network, consisting of a convolutional block in an ‘octree’ structure that sequentially up- (or down-) samples the resolution of a 3D model by a factor of two in each spatial dimension.TL-Embedding network accomplishes image-to-3D reconstruction by simultaneously training an autoencoder to compress the 3D model representation into a latent vector and an image encoder to predict the latent representation of a model from a single 2D image. Image Credits: Paper “We begin by training an autoencoder on the voxelised 3D models. The decoder of this network is adapted from the Octree Generating Network. It consists of multiple blocks, each of which upsamples the resolution by a factor of two in each dimension (i.e. an octree structure) and changes the number of channels, beginning with the latent representation, parameterised as a 256-dimensional vector (i.e. a single voxel with 256 channels),” says the paper. In the next step, they train a separate encoder network to predict the 256-dimension latent representations learned from the 3D autoencoder using the rendered single 2D images of the objects. Additionally, researchers use AlexNet – a seminal image classification, to design their encoder. Now, in the final step, voxelised 3D models are converted to LEGO builds. This is executed with the help of an algorithm adapted from the ColouredVoxels2Lego project. The algorithm converts each voxel to a 1-by-1 LEGO brick and then iterates through each layer in the z-coordinate of the model to find optimal groups of bricks to be combined into larger cuboid bricks. It is crucial to consider that colour is a critical factor in the LEGO model of faces. Since the TL-Octree network cannot add colours, LEGO reconstruction of faces was done via Volumetric Regression Network (VRN). However, it produces some patchiness to the colour. As a result, the k-means clustering algorithm got involved in giving uniformity to the colours. “We believe this work will greatly improve the accessibility of creative, personalised LEGO creations for builders of all skills and all ages,” concluded the paper. Way Forward Bringing AI can further help LEGO drive its business growth. Lego was ranked as one of the major toy companies globally in 2019, based on revenue, ahead of companies such as Mattel and Hasbro. During the last decade (2010 – 2020), the company has seen a rise in revenue from 2.2 billion euros to 5.87 billion euros, thereby a jump of over 166 per cent, as per the industry report. The newly introduced approach equips the company to come out with several innovative AI-based gaming experiences to further fuel its revenue growth.","excerpt":"The 3D model generation from images and then LEGO generation from 3D models has always remained an uphill task for the research community. The model presented in the paper has bridged this gap and came out with a complete Image2LEGO pipeline.","categories":["AI Features"],"tags":["Machine Learning","MIT research"],"author_name":"kumar Gandharv","publish_date":"2021-08-29T11:00:00","publication_year":"2021","word_count":620,"keywords":["Go","programming_languages:R","AI","Machine Learning","programming_languages:Go","ViT","MIT research","R"],"extracted_tech_keywords":["AI","R","Go","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/ai-now-builds-3d-lego-from-2d-images\/","complexity_score":4,"technical_depth":6,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":50283,"title":"Best Real Time Language Translator Devices in 2024","content":"Language translation is a key tool that erases borders and brings the world closer. From travel to ease in business, real-time translators play a great role in expunging boundaries. Communicating and translating a foreign language in real-time was once an impossible feat — but now things are easier, thanks to emerging technologies like artificial intelligence, machine learning and speech recognition. There are many products available in the market, but one of the most important features to look into while buying a translating device is to check the number of languages it provides in real-time translation. One should also look into the possibility of memory expansion for adding more languages. In this article, we list down top real-time language translating devices. 1. Birgus Voice Translator Device This voice translator is a smart language interpreters device with 2.4 inches capacitive screen. It supports two-way translation between 52 languages including Arabic, Czech, Danish, English Australia, English British, Catalan, etc. with an accuracy rate of 98%. The translator device is highly sensitive and accurate in its recognition of any of its 52 included languages. Click here for device features 2. Brain Freezer Touch Smart Pocket Translator This device is a touch smart pocket language translator. It allows two-way language translations and supports more than 100 languages. The device uses AI to constantly improve its translations and one can use it to for travel, business as well as learning a new language. Click here for device features 3. Birgus Smart Voice Translator Device This smart voice translator is a 2.4 inches color touch-screen which supports 70 two-way language translation. You can use this smart translator to travel, learn new languages, and much more and just needs to connect Wifi or hotspot. Click here for device feature 4. Docooler Intelligent Language Translator Device This is an intelligent language translator device which makes 2-way translation with accurate speech recognition and clear pronunciation. The device not only supports online translation in the Wifi connection state but also supports offline translation. It supports up to 41 languages with real-time inter translation. Click here for device features. 5. Google Pixel Buds Google Pixel Buds are the wireless earbuds which provide real-time translation in the ear and hands-free help from the Google Assistant. With the assistance of the Google Translate app on the Google Pixel or Pixel 2 phone, the Google Pixel Buds can translate more than 35 languages. Pixel Buds have 5 hours of listening time on a single charge, with up to 24 hours with the wireless charging case. It will be available in 2020 at the Google stores. Click here for device features 6. MOGOI Smart Language Translator Device MOGOI language translator device has smart photo translation and dialects support. It supports more than 40 languages such as German, Italian, Japanese, Korean, Spanish, Portuguese, Russian, Arabic, Armenia, among others. The smart translator device adopts professional noise reduction technology which makes it easy to deal with all kinds of a noisy environment and intelligent learning mechanism. Click here for device features 7. Pocketalk Language Translator Device for Portable Two-Way Voice Interpreter This smart device is one of the popular language translators in Japan. It is a two-way voice translator and has a built-in two-year data plan which translates conversations anywhere. The voice translator device can translate up to 74 languages seamlessly and it accesses translation engines as a user speaks to break down the language barriers instantly and accurately. Click here for device features to buy it. 8. Umiwe Smart Language Translation Device This intelligent translation machine automatically recognizes and translates and speaks out the translation results with an accuracy of 99%. It covers two-way translation up to 40 languages   With this smart photo translation device, one is able to connect it with Bluetooth and translate it in real-time. Click here for device features 9. Waverly Lab’s “Ambassador” Waverly Labs introduces Ambassador which is a powerful shareable tool that is compatible with 20 languages and 42 dialects. It provides accurate translations that can be delivered via over-the-ear audio, text on a smart app as well as it can be broadcasted live in a conference environment. However, this product is currently not available for buying in India. This year at CES, Waverly Labs named as the CES 2020 Innovation Awards Honoree. Click here for device features Popular Posts Best Smart Home Gadgets Best AR\/VR Devices Best Home Robots Best Gifts For Tech People Best Edge Computing Products","excerpt":"Language translation is a key tool that erases borders and brings the world closer. From travel to ease in business, real-time translators play a great role in expunging boundaries. Communicating and translating a foreign language in real-time was once an impossible feat — but now things are easier, thanks to emerging technologies like artificial intelligence, […]","categories":["AI Trends"],"tags":["Speech Recognition"],"author_name":"Ambika Choudhury","publish_date":"2019-11-19T17:16:59","publication_year":"2019","word_count":737,"keywords":["Go","machine learning","artificial intelligence","programming_languages:R","AI","innovation","ML","programming_languages:Go","Speech Recognition","edge computing","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","edge computing","R","Go","innovation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-9-real-time-translation-devices-you-can-buy-in-india\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":56128,"title":"Can Baidu’s Latest AI Tool Fight Coronavirus","content":"In the wake of the world’s attempt to prepare for the Corona outbreak, Baidu’s AI team has released a tool — LinearFold to reduce 2019-nCoV prediction time from 55 minutes to 27 seconds. The new, or “novel” coronavirus, now called 2019-nCoV, hasn’t previously been detected before the outbreak, which was reported in Wuhan, China in December 2019. It has now claimed deaths of nearly 500 people, and the whole world is on high alert. Compared to the SARS (severe acute respiratory syndrome) outbreak in 2003, which infected 8,098 and killed 774 in 17 countries, the incubation period of 2019-nCoV lasts longer, spanning up to two weeks, and is highly contagious. With time, the experts now believe that 2019-nCoV will likely to continue mutating, making it unpredictable and harder to control. As the medical experts are trying to figure out a clear defence strategy against this global pandemic, Baidu’s AI team has lent a helping hand in the form of LinearFold. Linearfold’s ability, announced by the researchers, has fetched it a placement in the top academic conference in bioinformatics, as well as in the Bioinformatics journal. With Baidu’s LinearFold, claimed by the researchers, it takes 27 seconds to analyse the structural information of the virus. This efficiency is crucial for understanding the virus and developing its vaccine. Overview Of LinearFold via Baidu Research The challenge with the existing algorithm for RNA secondary structure prediction is the runtime that scales cubically with the RNA length. This delay in computation has been a huge challenge in predicting structures and applicability on RNA viruses which has large genomes such as HIV, Ebola, and in particular, the coronavirus family that ranges from 26 to 32 kilobases — the largest for an RNA virus. LinearFold is the first RNA folding algorithm to achieve linear runtime. Given an RNA sequence, x∈{A,C,G,U}⁠, the secondary structure prediction problem aims to find the best-scoring pseudoknot-free structure. In this framework, scores for different pairs can be assigned, and a penalty can be given for each unpaired nucleotide. LinearFold is the combination of computational linguistics and incremental parsing algorithms that are used to scan the RNA sequence in a faster way. Key Takeaways According to the original paper, the authors list the following advantage of LinearFold: Though LinearFold uses only a fraction of time and memory compared to existing algorithms.The accuracy improvement of LinearFold is more pronounced on longer families of rRNAs.LinearFold is also more accurate than the baselines at predicting long-range base pairs, which are challenging for the current models Although the performance of LinearFold depends on the beam size, the accuracy of the prediction is stable. Genes are often expressed in terms of RNA’s (Ribonucleic Acid). RNAs play a key role in many biochemical reactions, and knowing their structure will help in guessing what role they will be playing. The 2019-nCoV belongs to a family of enveloped coronaviruses that are single-stranded RNA viruses, such as HIV, Ebola and influenza, which mutate faster and make vaccine development more difficult. So obtaining the sequence of RNA is a key to grasp its function. These sequences can be long and similar for a great length, and the crucial chunk of sequence can appear somewhere in the whole structure. Predicting this structure and then, in turn, predicting the function of RNA will help in designing drugs to control the catalysis of the enzymes, which the viruses use for synthesis. Baidu’s Linearfold offers the much needed accurate yet quick prediction that can cut down the diagnosis time. Know more about LinearFold here.","excerpt":"In the wake of the world’s attempt to prepare for the Corona outbreak, Baidu’s AI team has released a tool — LinearFold to reduce 2019-nCoV prediction time from 55 minutes to 27 seconds. The new, or “novel” coronavirus, now called 2019-nCoV, hasn’t previously been detected before the outbreak, which was reported in Wuhan, China in […]","categories":["Deep Tech"],"tags":[],"author_name":"Ram Sagar","publish_date":"2020-02-06T15:00:00","publication_year":"2020","word_count":587,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","Aim","R"],"extracted_tech_keywords":["AI","Aim","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/can-baidus-latest-ai-tool-fight-coronavirus-linearfold-rna\/","complexity_score":4,"technical_depth":6,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":37684,"title":"Can Data-As-A-Service Help Enterprises Get More Value Out Of Data Lake?","content":"Like all traditional members of the “as-a-Service” family such, Data-as-a-Service is a relatively new term, as it is being used since 2015. The purpose of DaaS is that it delivers data on demand to the consumers through various APIs in order to avoid the typical need to fetch as well as store large data assets and then search for the required information in the data asset. This open source software solution basically runs between the systems which manage your data and the tools which you use for analysing the data. This service can be used in various sectors such as retail, healthcare, e-commerce, financial sectors, telecommunications, homeland security, organisations, etc. where the users can be able to access the databases provided by the vendors or host their own databases on their self-managed systems. It is a self-service model where you can explore, organise, describe as well as analyse data regardless of its location, size or structure with the help of tools such as Tableau, Python, etc. According to a report by Telecommunications Market Research Reports, the future revenue from big data-driven telecom analytics and Data-as-a-service to grow big data revenue into $5.4 billion industry by 2019. The telecom analytics market alone is expected to grow at a compound annual growth rate of nearly 50% between 2014 and 2019. This alternative cloud computing service model is different from the traditional as-a-service models. Organisations have been struggling with the data lakes (cloud storage services) since the last few years due to the heavy flow of data and unavailability of the right tools. Future Of Analytics in DaaS The purpose of the Data-as-a-Service model is to dump the hazards and hardships of data management into a third party cloud-based provider. Customer Resource Management (CRM), Enterprise Resource Planning, etc. are the most common business applications which are powered by Data-as-a-Service technology. According to Oracle, Data as a Service (DaaS) is a revolutionary new service that gives you unprecedented levels of connection to customers. It is a response to the growing volume and variety of data generated in today’s digital world. Users consume data across a variety of systems and processes—data that can be used to make customer engagements more relevant and impactful. This data-driven insight helps you connect with customers across marketing and sales to virtually all areas of your business. One of the DaaS platforms is the Dremerio, which enables business analysts and data scientists to explore and analyse data at any time. The platform empowers users to curate, accelerate and share any data at any time. It allows the analysts to easily cut through complicated source data with nested structures or mixed types, creating the new virtual dataset. Benefits Of Data-as-a-Service Agility The consumers don’t need any kind of extensive knowledge of the underlying data and can move quickly due to simplicity in the data accessing process. Cost-effectiveness Managing the large volume of data are being done in-house and there is no need to spend an extra penny for handling the bulk of data. The service space can deploy their data delivering applications in a cost-effective manner. Data Quality By DaaS, one can receive high-quality data because the amount of data which are being accessed are controlled through the data service itself. This service adds a strong layer of security as well as augmented data quality.","excerpt":"Like all traditional members of the “as-a-Service” family such, Data-as-a-Service is a relatively new term, as it is being used since 2015. The purpose of DaaS is that it delivers data on demand to the consumers through various APIs in order to avoid the typical need to fetch as well as store large data assets […]","categories":["IT Services"],"tags":["DaaS"],"author_name":"Ambika Choudhury","publish_date":"2019-04-13T07:03:33","publication_year":"2019","word_count":554,"keywords":["big data","API","cloud computing","AI","Git","RAG","Python","analytics","DaaS","R","data lake"],"extracted_tech_keywords":["AI","analytics","RAG","cloud computing","Python","R","Git","API","big data","data lake"],"url":"https:\/\/analyticsindiamag.com\/it-services\/can-data-as-a-service-help-enterprises-get-more-value-out-of-data-lake\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":58876,"title":"10 Things Students Should Consider Before Pursuing A Career In Data Science","content":"Being one of the highest-paid tech jobs ever, data science has rapidly become a popular career option for many. With the growing possibilities of emerging technologies, data science has come to be popular field researchers and working professionals. According to a survey, the data science and analytics ecosystem have been witnessing an overall growth in the number of jobs with India contributing to 6% of open job openings worldwide. And therefore, students, as well as working professionals, are sharing their interest in making a career in data science. However taking the first step can be hard for any student, especially if you lack knowledge and experience of the field. In this article, we will discuss ten things that a student should consider before pursuing a career in data science. Pick A Relevant Major Although many experts have mentioned that degrees aren’t important for a data science career, it indeed requires strong skills in statistics, maths, and programming. To have an early kickstart to your data science career, students need to focus on topics like statistics, computer science, physical sciences, linear algebra and calculus, and social sciences for their choice of major. Choosing a STEM major at an early stage of a career will let the students’ brain expose to the field early and will create a solid base in applied math and statistics, which is a key skill for analysing the large volumes of data and data trends collected by big companies. Books, Articles & Journals To The Rescue Although there are several internet resources that one can use to learn data science, books, articles and journals are still one of the best tools to learn from. With an overload of information, online, it indeed can create hassles for students of this era face who are trying to get into a new field, which, in turn, confuses students who are totally novice. If one has to check, there are hundreds of online courses, specialisations, videos, podcasts etc., but it is best that students, who are willing to learn about this new field, take help from books, articles and journals to start learning the basics concepts, and then add value with other online resources. In the data science community, many experts and professionals have shared their knowledge by writing books, which students can opt-in order to learn more about data science. Personal Project Will Be Beneficial Learning shouldn’t be contained to the classroom, rather the best way to learn something new is to apply what is learnt in the classroom into something you care about, like creating a personal project. Data science is not about learning tools and techniques; in order to become a data scientist, students need to build a portfolio by working on real projects — maybe independently on platforms like Kaggle. In fact, if data science enthusiasts aren’t equipped enough to design their independent projects, they can even opt from premade projects and tutorials online that pose problems for students to solve. Such possibilities will let students get hands-on experience without investing in expensive tools or training. Enhance Your Computing Skills Considering computers have penetrated in all aspects of human lives, it isn’t just an essential tool to master the skills of data science but also to survive in today’s world. And therefore, basic computing skills is indeed necessary for all human beings striving in the corporate world. And for data scientists learning about programming and coding will help them in their job where they will be required to solve business problems with the help of data as well as make software for business customers. These computing skills will also help data scientists to learn advanced programming skills which are also essential for surviving in the industry. All companies nowadays are expecting their candidates to know the all sort of programming language, like R or Python, and a database querying language like SQL, FQL, HTSQL. Opt For Online Courses Early For data science professionals, and even for students, online classes and crash courses can be a quick way to learn about data science. These crash courses can help these enthusiasts learn ranges of technical and programming skills, as well as data analysis and machine learning. However, one needs to invest heavily on disciplining themselves to get the best out of these online courses. In fact, there are several courses and tutorials available online that can provide a perfect understanding of the subjects. Some of these materials are even developed and distributed by the university professors, while some are also put together by experts and experienced professionals. One can choose the right online course for themselves and begin their journey of data science. Peer Groups For Gaining Knowledge Although students need to read articles written by data science professionals or watch tutorials and videos on data science, it’s also important if students and data science enthusiasts can have face-to-face conversations with people and experts of the field. Joining a peer group keeps enthusiasts motivated, and provides a hand-holding experience for newcomers in this new field. Interacting with peers from the same field can be beneficial for newcomers and enthusiasts. Students can also find communities, forums, and groups online to discuss subjects and problems of their career. In fact, peer groups are the best way to keep updated with the latest news and updates on different technology related to the field. One can also go to a data science conference, attend interesting talks by experts of the field, or join a meet up to make themselves visible in the data science community. Learn From All Possible Resources Key to becoming a data science expert is to continuously learn from all possible resources available. In today’s world, there isn’t any shortage of information for one to learn data science — such as podcast, webinars given by data scientists, data science articles, various online courses, and engaging with the data science community. One can even follow the expert data scientists of the industry to learn about the industry as well as keeping track of recent trends of the field. Developing relevant resources will provide students with pertinent information about the field, and in turn, the students will not be overwhelmed with the information provided. Soft skills Are Important Too Although it is considered a soft skill, it is important for data scientists to improve their communication skills along with their technical skills. Like any other field of the IT industry, data science also requires excellent communication skills. Communication skills are even more critical when you are working in the field of data science, as data scientists need to share their ideas to their colleagues and other teams, and having good communication skills help them to have smooth communication with the rest of the organisation. A good communication skill will help data scientist to understand the business requirements as well as communicate the results to team members, bosses, as well as stakeholders. The skill is usually underestimated, however, has been an important tool for data scientists to enhance, as good storytelling is critical for data scientists. Don’t Jump Into The Deep End Some young enthusiasts and students pursue this field with a vision to build a robotic technology of the future or to directly become the next tech genius with their self-driving cars, and computer vision. However, usually, people forget that it’s important to first master the fundamentals of data science. For a student who is willing to pursue a career in data science needs to be completely aware of the field, basics as well as advanced knowledge before they actually start to dream about building revolutionary technology. Some of the fundamental steps towards data science is learning the basic techniques and algorithms, which will help students to understand advanced topics better. Understand The Job Market It’s indeed beneficial to know the market beforehand, even while as a student, in order to understand what is actually required to survive in this competitive landscape. Such a piece of knowledge will give students a headstart with their choices of courses and skills for future perspective. Moreover, it’s important to know the market well enough before applying so that you know the tips and tricks to get through the interview and land on the desired job. Its never to early to stay abreast with the industry, as this constantly evolving industry can bring in a lot of changes in a small span of time. In this age of constant technological innovation, it is important to understand the career market as early as possible to stand ahead of the crowd.","excerpt":"Being one of the highest-paid tech jobs ever, data science has rapidly become a popular career option for many. With the growing possibilities of emerging technologies, data science has come to be popular field researchers and working professionals. According to a survey, the data science and analytics ecosystem have been witnessing an overall growth in the […]","categories":["AI Trends"],"tags":["big data for social good","common ai misconceptions","Data Science","Data Science Career","Data Science Jobs","data science software","Data Scientist","Data Scientists","project topics for computer science"],"author_name":"Sejuti Das","publish_date":"2020-03-17T20:00:00","publication_year":"2020","word_count":1423,"keywords":["API","computer vision","R","data science software","project topics for computer science","data science","analytics","Data Science","Data Scientists","Go","machine learning","AI","big data for social good","Data Science Jobs","Data Science Career","common ai misconceptions","Python","SQL","Data Scientist"],"extracted_tech_keywords":["AI","machine learning","computer vision","data science","analytics","Python","R","SQL","Go","API"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/10-things-students-should-consider-before-pursuing-a-career-in-data-science\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10005127,"title":"Amazon Launches Amazon Braket To Boost Quantum Computing Research","content":"Amazon recently announced the launch of Amazon Braket which is a fully managed quantum computing service on AWS to boost research in this space. It aims to provide a user-friendly platform to get started with quantum computers and further explore the field with its potential applications. Amazon Braket will provide an environment to design quantum algorithms, test them on simulated quantum computers, and run them on different types of quantum computing hardware. It will also provide managed Jupyter notebooks with pre-installed developer tools, sample algorithms, and tutorials to get started. It will provide researchers access to quantum annealing hardware from D-Wave, and two types of gate-based quantum computers — ion-trap devices from IonQ and systems built on superconducting qubits from Rigetti, to carry their research. The company said that to test the quantum algorithms researcher can troubleshoot using a software-based simulator included in the Amazon Braket SDK. They can also use the service’s high-performance, fully managed simulator to explore larger, more complex designs. Quantum computing has the potential to solve computational problems that are beyond the reach of classical computers and is particularly useful in areas such as chemical engineering, material science, drug discovery, financial portfolio optimisation, and machine learning. Amazon aims to help researchers explore the field and acquire the required skills by providing easy access to quantum computing hardware. It allows researchers to build their own algorithms from scratch or choose from a set of pre-built algorithms. Once you have built your algorithm, Amazon Braket provides a choice of simulators to test, troubleshoot and run your algorithms. The company believes that Amazon Braket will allow researchers to evaluate the potential of quantum computing for the organisation and build expertise. Amazon Braket is currently available in the US-East (N. Virginia), US-West (Oregon) and US-West (N. California) AWS Regions.","excerpt":"Amazon recently announced the launch of Amazon Braket which is a fully managed quantum computing service on AWS to boost research in this space. It aims to provide a user-friendly platform to get started with quantum computers and further explore the field with its potential applications. Amazon Braket will provide an environment to design quantum […]","categories":["AI News"],"tags":["Quantum Computing","quantum computing programming"],"author_name":"Srishti Deoras","publish_date":"2020-08-18T17:35:04","publication_year":"2020","word_count":299,"keywords":["Quantum Computing","Go","machine learning","AWS","AI","cloud_platforms:AWS","programming_languages:R","Aim","quantum computing programming","GAN","Jupyter","R"],"extracted_tech_keywords":["AI","machine learning","Aim","Jupyter","AWS","R","Go","GAN","cloud_platforms:AWS","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/amazon-launches-amazon-braket-to-boost-quantum-computing-research\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":59727,"title":"How Edelweiss Group Is Preparing To Prevent The Spread Of COVID-19","content":"Long before the Maharashtra government made it mandatory for private-sector employees to work from home, Edelweiss Group had started preparing to enable its employees to work from home. The company started implementing various tools and partnering with companies like Microsoft, Zscaler and AWS for better productivity and a better experience for its customers, even in these unprecedented times. Currently, more than 90% of Edelweiss Group employees, i.e. approximately 9000 people have been enabled and are working seamlessly from home while ensuring that client information and data are protected at all times. Nitin Agarwal, the CTO of Edelweiss Group said “Employees are using digital tools for online video meetings and secure instant messaging through Microsoft teams. They are accessing internal applications hosted on our ‘On-Premises Datacenter’ as well as AWS\/Azure Cloud, using the zero-trust security platforms of Zscaler and Microsoft. Our digital engagement internally has gone up 30 times in the last week. We currently have over 1500 online video meetings\/ per day (up from just 30 last week). Our colleagues are exchanging over 50, 000 instant messages daily (up from 2000 messages per day).” Edelweiss Group has always been a progressive and socially responsible organisation, going beyond profitability to align with societal causes. It has always been conscious of the fact that sustainable and responsible growth is the only acceptable model for business. As a leading organisation in Indian financial services, Edelweiss Group is in a unique position to influence or reinforce responsible governance positively – be it as a business or as an employer. Thus, considering the current environment, Edelweiss Group has taken several proactive and timely steps to mitigate the threat and spread of COVID 19. Travel Advisory: All Edelweiss Group employees have been asked to restrict International and Domestic travel, both for official and personal reasons unless there is a dire emergency. It has been made mandatory for all employees to report all travel history (domestic and International) for themselves and any of their family members. Alongside, for those who have travelled or been in contact with any person recently travelling from affected locations, self-quarantine has been made mandatory. In case of any difficulty requiring an employee to travel, they will self-quarantine for 14 days. Hygiene: All the Mumbai premises have been deep sanitised over the weekend of 14th March 2020. For other branches, especially in affected cities, a schedule has been prepared and shared by the Incident Services Team for sanitisation. In fact, all offices in Mumbai are equipped with non-contact thermal checks at entry. These checks will be performed on all guests, employees or third-party staff entering the office premises for any emergency staff) Besides, all visitors are being advised to sanitise their hands thoroughly before proceeding further within any Edelweiss Group building\/office premise. Additional sanitiser bottles are provided at every entry point and monitored by the security\/receptionist for the emergency staff. Edelweiss Group has procured over 5000 masks for the safety of its ground personnel. Reporting: If any employee develops any symptom of COVID-19, they have been advised to immediately notify the same to their manager and HR team member, for appropriate precautions and further advice. Client Servicing: Client Servicing remains the core of the business ethos and in these uncertain times, continues to remain a priority. Edelweiss Group has undertaken the necessary business continuity steps to mitigate any disruption that may occur in the coming weeks. The company is well placed to ensure that the support and delivery for clients’ unique requirements will continue at the highest standards. Teams are in constant communication with all clients via email, Edelweiss Group websites and social handles to assure them the company’s proactive precautions to contain the virus, which includes — the safety of employees to avoid any disruption of client activities and data security measures in light of employees working from home. Employees are using remote\/digital channels to minimise physical contact and ensure containment. Employees are also given tips on remaining safe and minimising spread.","excerpt":"Long before the Maharashtra government made it mandatory for private-sector employees to work from home, Edelweiss Group had started preparing to enable its employees to work from home. The company started implementing various tools and partnering with companies like Microsoft, Zscaler and AWS for better productivity and a better experience for its customers, even in […]","categories":["AI Features"],"tags":["Coronavirus","covid-19"],"author_name":"Sejuti Das","publish_date":"2020-03-24T14:47:04","publication_year":"2020","word_count":660,"keywords":["Go","AWS","covid-19","AI","R","ML","Git","Coronavirus","ViT","Rust","GAN","Azure"],"extracted_tech_keywords":["AI","ML","AWS","Azure","R","Go","Rust","Git","GAN","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-edelweiss-group-preparing-themselves-to-prevent-the-spread-of-covid-19\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10017389,"title":"Tesla On Top, Big Tech Blockades And More In This Week’s Top News","content":"Elon Musk and his company Tesla witnessed new heights in their respective valuation. While the automaker soared past $700 billion market value, the company’s chief saw his net worth eclipsing that of Jeff Bezos’ and making him the richest person on the planet. Musk’s net worth hit $195 billion on Thursday; a $150 billion increase over the past 12 months, thanks to Tesla’s share price, which surged 743 % in 2020. In 2020, we produced and delivered half a million cars. Huge thanks to all those who made this possible.https:\/\/t.co\/q43vz6RMhd— Tesla (@Tesla) January 2, 2021 China Says No To “AWS” In what is possibly a weird demand for Amazon, the Chinese authorities have banned the e-commerce giant from using its AWS logo in the country. According to reports, The Beijing Municipal High People’s Court ruled that the trademark for the term “AWS” belonged to ActionSoft, a Chinese software and data services company. The court ordered Amazon to stop using the term AWS or any similar logos in China and pay compensation equivalent to $11.8 million, to ActionSoft. “Amazon was the first to use the AWS logo in China to sell cloud services for many years. We strongly disagree with the court’s ruling and have appealed the case to the Supreme People’s Court,” responded AWS’ spokesperson to this ruling. OpenAI’s New Toys We’ve developed two neural networks which have learned by associating text and images. CLIP maps images into categories described in text, and DALL-E creates new images, like this, from text. A step toward systems with deeper understanding of the world. https:\/\/t.co\/rppy6u1zcn pic.twitter.com\/MNVlo8LZbV— OpenAI (@OpenAI) January 5, 2021 On Tuesday, OpenAI released two new neural networks– DALL.E and CLIP, which combine the visual and language skills of algorithms. DALL·E is a neural network that creates images from text captions for a wide range of concepts expressible in natural language. Whereas, CLIP which stands for Contrastive Language–Image Pre-training can be applied to any visual classification benchmark by simply providing the names of the visual categories to be recognized, similar to the “zero-shot” capabilities of GPT-2 and 3. Read more about it here. Whatsapp Faces An Exodus Use Signal— Elon Musk (@elonmusk) January 7, 2021 On Monday, WhatsApp updated its terms of use and privacy policy, primarily to expand on its practices around how WhatsApp business users can store their communications. A pop-up has been notifying users that as of February 8, the app’s privacy policy will change and they must accept the terms to keep using the app. This news surprised the users around the world who have already been sceptical about the privacy practices of the parent company Facebook. Tesla CEO, Elon Musk who had expressed his concerns about Facebook in the past tweeted in support of the Signal messenger app. Signal is an open-sourced nonprofit organization whose business model eliminates the need for personal data in any way. Musk’s tweet was followed by a surge in downloads of the Signal app. “We are currently having a record level of downloads for the Signal app around the world. Between WhatsApp announcing they would be sharing everything with the Facebook mothership and the Apple privacy labels that allowed people to compare us to other popular messengers, it seems like many people are interested in private communication,” wrote the company. Big Tech Blockades If you can silence a king, you are the king. https:\/\/t.co\/mM8oTL4Wxg— Naval (@naval) January 9, 2021 The events that followed the ruckus at the US Capitol Hill, might set a new precedent in the history of the internet. While Twitter and Facebook de-platformed the President of the United States, Google has temporarily banned Twitter’s alter-ego Parler from their Play Store. According to inputmag, Apple too, has ordered Parler to remove “objectionable content” from the platform within the next 24 hours or it will be removed from the App Store. Facebook’s Zuckerberg even penned a post explaining the rationale behind their decision Many applauded the social platforms for silencing the head of one of the most powerful nations in the world. Whereas, few others questioned the policing practices of the platforms and the timing of their decisions. President Trump was not the only one who has come under the ire of the gatekeepers of internet, Twitter has also suspended the account of Sci-Hub, popular for providing free access to paywalled academic papers. Sci-Hub is currently in a tussle with academic publishers Elsevier, Wiley, and the American Chemical Society who have filed a suit with the Delhi High Court, asking for blockage of Sci-Hub. Sci-Hub’s twitter account was playing a key role in collecting statements in preparation for the case until it was banned by Twitter. Intel Looks East According to reports, Intel is in talks with chip giants TSMC(Taiwan Semiconductor Manufacturing Co.) and Samsung Electronics in an attempt to revamp its production. If the deal comes through, TSMC will assist Intel in manufacturing 4nm and 5nm chips. For Intel, 2020 has more or less been underwhelming. Competitors like NVIDIA and AMD moved ahead in the race relatively. Adding to the agony, one of its oldest customer Apple has announced its departure to make its own chips, a move that is predicted to become more common among other companies. Last month, Amazon too, has introduced two of its own chips. Intel, which has been revamping and launching new products since mid-2020, is no verge of a major structural change at the production as well if the talks of outsourcing come through. Apple In Talks With Hyundai? On Thursday, reports of a potential Apple-Hyundai partnership on driverless, electric vehicles sent Hyundai’s shares soaring. Apple, which has announced its autonomous vehicle project, a couple of weeks ago is reportedly in talks with the South Korean giant Hyundai for a cooperation on its car project. According to WSJ, Apple has been reaching out to suppliers about the possibility of doing its own car, potentially starting production as soon as 2024. Hyundai has recently purchased Boston Dynamics from SoftBank in a deal worth nearly $1 billion. Nvidia-ARM Deal Runs Into Trouble The British regulators are looking to probe the $40billion deal that can change the semiconductor industry forever. Last year, in September, NVIDIA announced that it would acquire Arm Limited from SBG and the SoftBank Vision Fund in a transaction valued at $40 billion. According to the terms of the transaction, NVIDIA will pay SoftBank $21.5 billion in common stock and $12 billion in cash. This includes $2 billion payable at signing. Britain’s Competition and Markets Authority said that it was inviting third parties to comment on the merger. According to Journal, the investigation will assess the deal’s possible effects on competition in the U.K. The regulators are looking into various angles that includes loss of jobs on the home turf and other factors. ARM currently employs 6,500 people. Read more about the deal here. Google Employees Form A Union The employees at Google have delivered the much dreaded to the search giant. According to reports, a group of Google employees has formed a union that would oversee the company’s practices across the world. This news comes after recent ousting of Google’s AI ethics co-lead Timnit Gebru who accused Google of creating roadblocks to few types of research. According to people familiar with the proceedings, the Google union’s main goal is to have a say in how Alphabet Inc., does business and how its products are used. Also Watch","excerpt":"Elon Musk and his company Tesla witnessed new heights in their respective valuation. While the automaker soared past $700 billion market value, the company’s chief saw his net worth eclipsing that of Jeff Bezos’ and making him the richest person on the planet. Musk’s net worth hit $195 billion on Thursday; a $150 billion increase […]","categories":["AI News"],"tags":["AWS","big data role","Elon Musk"],"author_name":"Ram Sagar","publish_date":"2021-01-09T18:00:00","publication_year":"2021","word_count":1234,"keywords":["Go","API","DALL-E","AWS","AI","neural network","OpenAI","Elon Musk","GPT","CLIP","R","big data role"],"extracted_tech_keywords":["AI","neural network","OpenAI","AWS","R","Go","API","GPT","CLIP","DALL-E"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/tesla-twitter-facebook-apple-latest-top-news\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10126044,"title":"Machine Learning Enhances Assam Government’s Disaster Response Amid Floods","content":"Every year, floods displace millions of people in Assam, forcing them to seek shelter in relief camps. In fact, devastating floods were some of the watershed events in the state’s history. Each year, the Indian government allocates hundreds of crores towards relief efforts during Assam floods. But, a major challenge the government faces is ensuring that these funds are effectively utilised and that the right resources reach the appropriate locations promptly. This issue persists due to the fragmented nature of government data, which is often stored in isolated silos. In the case of Assam government, the data related to disaster management is stored across 18 different departments and other central agencies. But a research lab working at the intersection of data, technology, design and social science has come to their aid. Called CivicDataLabs (CDL), the startup is working closely with the Assam government to unify its data and analyse it in a much better way. ( Credits: Reuters) Setting Data Standards “The Government of Assam is investing heavily in disaster risk reduction efforts. However, they have faced challenges in tracking the allocation of funds to ensure they are directed to the correct regions and at the appropriate times, and whether that’s translating into increased risk reduction,” Gaurav Godhwani, co-founder and executive director at CivicDataLab told AIM. The finance department in Assam and the Disaster Management Authority knocked on the startup’s door, and they decided to answer. Godhwani revealed that various state departments actively collaborated with the startup, especially with data sharing as required. The startup then worked towards cleaning and standardising the datasets. “We worked with the department to see what procurements were happening related to floods– When did the last embankment repair happen after damage happened? When was the restoration of that particular geography conducted? Pretty soon, we realised that data, because of lack of data standards, was not able to capture this critical information,” he said. Hence, CDL introduced an international data standard known as the Open Contracting Data Standard, which is adopted in over 50 nations. As a result, Assam became the first state in India to adopt this international standard. Moreover, Godhwani and his team created a platform where all the datasets come together. They also created a data exchange layer that enables all the publishers, including the 18 departments and other central agencies, to manage disaster-related data. “Most of them have their own databases, their own APIs, not talking to each other. The challenge was to bring all of those together in one place. So we streamline the datasets, standardise them based on common geographical indicators, like state, district, and revenue circles and give an ID to all of those data sets so that they are all uniquely identified,” Godhwani explained. (Stakeholder Consultation with ASDMA, Source: CivicDataLabs) Early success Once the datasets were streamlined and standardised, CDL used simple machine learning techniques to derive valuable insights from the data. “We presented our initial findings, demonstrating hazard analysis and disaster modelling using GIS techniques and simple machine learning techniques. We highlighted districts and revenue circles at higher risk, emphasising areas needing greater attention,” Godhwani said. In response, the Assam Government requested the integration of disaggregated data for periodic analysis spanning a five-year period. CDL developed a platform which allows comprehensive assessment of past trends and facilitates planning for the upcoming monsoon cycle and subsequent restoration efforts. Godhwani revealed that the Assam Government has already started results by leveraging the data model CDL built. The authorities used the model to determine the allocations and verified them with assistance from on-ground staff. “They found that the algorithm effectively identified areas that had previously been overlooked. As a result, revenue circles that historically received insufficient funds have now begun to receive more support based on our model,” Godhwani pointed out. Secondly, the department’s capabilities have significantly expanded. Previously, generating analytics reports during disasters was a time-consuming process. They lacked knowledge of suitable software and effective data visualisation techniques. “Now, they are actively pursuing online courses on geospatial data management and AI. We have observed a renewed enthusiasm for data science and what it can do.” ( Capacity building workshop with Assam state and district officers conducted by CDL, Source: Civic Data Labs) Machine Learning Techniques Come to Aid Godhwani said this was done using three three machine learning algorithms. The initial algorithm employed was Random Forest, aimed at understanding the potential for inundation. This was achieved by analysing satellite imagery sourced from multiple sources to observe the extent of inundation during the monsoon season. Secondly, the Data Envelopment Model was used to extract variables from textual information, like maternal and child health indicators, public procurement etc. “The algorithm helps in understanding how many times procurement has happened before the monsoon season, during the monsoon season and after the monsoon season. The algorithm creates a model on top of all of these, giving them scores based on the past data, and when the past data is missing, it creates a baseline that whenever the new data is coming, we can measure whether we are up that baseline or below the baseline.” Lastly, CDL then utilises the TOPSIS scoring method to consolidate all data into a comprehensive risk index, enabling measurement on a scale of one to five, where one indicates lower risk and five indicates higher risk. This scoring methodology computes all variables in one central location and provides timely updates to the department on which areas to prioritise each month. This ensures that updates can be made on a monthly basis. “Earlier, it used to take a couple of months to compute just one timestamp for the department. So that’s the kind of automation which we’re seeing with these machine learning algorithms,” Godhwani pointed out. Leveraging generative AI While machine learning algorithm has sufficed so far, Godhwani believes Large Language Models (LLMs) can be leveraged to scale the platform. Currently, all the data generated goes to a public platform, but Godhwani believes that if the same data has to be disseminated to the public at large, something like a rule-based engine would not suffice. Information could be disseminated to the public through a chatbot or through social media applications which the public is already using. “Information such as, for instance, if you’re going in this particular area, please be careful there is a likely inundation, which may happen in the next 24 hours, 48 hours because of heavy rains. That’s the kind of opportunity we have with large language models,” Godhwani said. CDL has already conducted a pilot and shown the results to the Assam government. But in terms of deployment, it’s still a few months away. “The government agencies need to become more comfortable in piloting this at scale. The other thing is ensuring the ethics, responsible development and modelling is done in a collaborative manner. It can’t be hallucinating, as most these models today still hallucinate. If they are, the risks are extremely high.” About Civic Data Labs Godhwani founded Civic Data Labs with Deepthi Chand. Interestingly, the duo has registered Civic Data Labs as a startup and not as a non-profit. The team is 52 members strong consisting of disaster risk reduction experts, climate scientists, data scientists, data engineers, technologists, designers,  user researchers, and technological architects. Even though the company is bootstrapped, it raises funds for projects. For instance, the project with the Assam government is funded by the Rockefeller Foundation and the Patrick J. McGovern Foundation in collaboration with Open Contracting Partnership, a global think tank. The startup works closely with non-profits, academic institutions, volunteer groups, government agencies, and departments.","excerpt":"The algorithm effectively identified areas that had previously been overlooked. As a result, revenue circles that historically received insufficient funds have now begun to receive more support","categories":["Deep Tech"],"tags":[],"author_name":"Pritam Bordoloi","publish_date":"2024-07-06T12:25:21","publication_year":"2024","word_count":1265,"keywords":["data science","Go","machine learning","AI","ML","RAG","Aim","analytics","generative AI","R"],"extracted_tech_keywords":["AI","machine learning","ML","data science","analytics","generative AI","Aim","RAG","R","Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/machine-learning-enhances-assam-governments-disaster-response-amid-floods\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10022194,"title":"8 Best Free Resources To Learn XGBoost","content":"XGBoost or eXtreme Gradient Boosting is a popular scalable machine learning package for tree boosting. Data scientists use it extensively to solve classification, regression, user-defined prediction problems etc. The speed, high-performance, ability to solve real-world scale problems using a minimal amount of resources etc., make XGBoost highly popular among machine learning researchers. Here is a list of the top eight free resources to learn XGBoost. (The list is in no particular order) 1| XGBoost Tutorials About: XGBoost Tutorials covers various topics to help you understand XGBoost from scratch. This official document will explain boosted trees in a self-contained and principled way using the elements of supervised learning. The tutorials deliberate on: an introduction to Boosted trees; distributed XGBoost with AWS YARN; distributed XGBoost on Kubernetes; Random Forests in XGBoost; using XGBoost External Memory Version and more. Know more here. 2| How XGBoost Works About: This tutorial is put together by AWS, where you can learn all about the XGBoost method and how it works. You can also grasp related topics, such as XGBoost algorithms, how gradient tree boosting works, hyperparameters, supervised learning etc. Know more here. 3| Find Actionable Insights Using Machine Learning And XGBoost About: Udemy’s ‘Find Actionable Insights using Machine Learning and XGBoost’ is all about practical experience. You will learn how to explore student data, model student behaviour using XGBoost, predict struggling\/at-risk students, identify what makes a struggling student different from successful students, build a report of actionable insights, etc. Know more here. 4| Using XGBoost in Python About: This tutorial will guide you on building machine learning models using XGBoost in Python. More specifically, you will learn: what Boosting is and how XGBoost operates; how to apply XGBoost on a dataset and validate the results; various hyper-parameters that can be tuned in XGBoost to improve model’s performance; how to visualise the Boosted Trees and Feature Importance, etc. Know more here. 5| A Gentle Introduction to XGBoost for Applied Machine Learning About: LAs the name suggests, ‘A Gentle Introduction to XGBoost for Applied Machine Learning’ will give you a gist of what XGBoost is, where it came from and how you can learn more. You will understand the goal of this project, why this package must be a part of your machine learning toolkit and where you can find more learning materials to up your game. Know more here. 6| XGBoost: A Scalable Tree Boosting System About: In this research paper, creator XGBoost, Tianqi Chen, explains why he created this package and how it can provide maximum performance in machine learning models. Know more here. 7| Understanding XGBoost Algorithm In Detail About: The tutorial gives a brief introduction to what XGBoost is and how the package works internally to make decision trees and deduce predictions. The tutorial touches on various tree-based techniques, features of XGBoost, and an example of how XGBoost helps predict a child’s IQ based on age. Know more here. 8| XGBoost About: This tutorial, compiled by Kaggle Grandmaster and founder of decision.ai, DanB, will teach you how to build and optimise models with the powerful XGBoost library. The tutorial will help you understand the full modelling workflow with XGBoost and how to fine-tune XGBoost models for optimal performance. Know more here.","excerpt":"XGBoost or eXtreme Gradient Boosting is a popular scalable machine learning package for tree boosting. Data scientists use it extensively to solve classification, regression, user-defined prediction problems etc. The speed, high-performance, ability to solve real-world scale problems using a minimal amount of resources etc., make XGBoost highly popular among machine learning researchers. Here is a […]","categories":["AI Trends"],"tags":["gradient boosting","XGBoost"],"author_name":"Ambika Choudhury","publish_date":"2021-03-16T10:00:00","publication_year":"2021","word_count":537,"keywords":["Go","machine learning","AWS","AI","RPA","Scala","Python","XGBoost","gradient boosting","R","kubernetes"],"extracted_tech_keywords":["AI","machine learning","XGBoost","AWS","kubernetes","Python","R","Go","Scala","RPA"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/8-best-free-resources-to-learn-xgboost\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":25694,"title":"How to make a career in Geointelligence?","content":"The Geointelligence market today is mainly driven by the demand for convergence of geospatial data with mainstream technologies. Trends suggest that there has been an escalation in the demand for geospatial analytics solutions with artificial intelligence capabilities across various industries. This is not all, as commoditization of geospatial data for industry use has also led to growth in the segment. The market for geospatial intelligence was valued at USD 30.71 Billion in 2016 and is projected to reach USD 73.91 Billion by 2021, at a CAGR of 19.2%. These trends have led to an explosion of career opportunities in the geointelligence landscape worldwide. The U.S. Geospatial Intelligence Foundation’s (USGIF) leads the Geointelligence landscape, and tackles these global problems. The organization has made significant efforts at accreditation and support of academic programs, over the years. These efforts have mostly been directed towards establishing standards throughout the Geointelligence community and the Geointelligence profession. Is Geointelligence a good career for you? Theresa Whelan during a USGIF Geoint event The field is diverse, and goes beyond the scope of geospatial intelligence systems and technology degree. Today the field is transforming continually with research being undertaken in new areas. These include crowdsourcing and forecasting. The new areas make the field more enticing for aspirants to consider a career in geointelligence. Moreover, it’s not easy to spot qualified GIS and remote sensing experts. Reports suggest that shortages of experts in the fields of cartography, photogrammetry, geodesy, and emerging areas will continue until 2030. Moreover, geointelligence also helps in safeguarding a nation’s security, or obtaining intel about enemy plans. Basically, recent advances in technology has made geoinelligence an interesting choice for career. What kind of skills are required to make a career in Geointelligence? Expertise in data analysis and use of technology isn’t all that governs the field of geointelligence. The domain combines the use of specialized skill sets today. Transforming patterns of human activity and real-world problems is studied and analyzed leveraging applies established principles, tools, and techniques. Soft skills like passion for geography, interest for numbers, visualization skills, people skills, networking skills, grit to undertake highly sensitive missions, analytical thinking, and more come handy for a person planning to pursue a career based on geointelligence. However, at the end of the day, a person imbibing all those attributes without really working on his GIS or Geospatial technology knowledge is not expected to prosper much in the space. Power of mapping is considered an important skill for Geointelligence Besides, the qualities, know-how, and skills mentioned above, an ideal candidate must reflect domain based proficiency in at least one or more of the following: geography – human, physical, environmental technologies in use (RS, UAV, IT, and more) database design\/ data management\/ data security \/software implementation specialized skillsets – imagery analysis \/cyber forensics \/ ethical hacking\/ cryptography \/ deep web data mining \/ criminology \/intelligence tradecraft \/ expertise in regional sociology homeland security \/ police \/other law enforcement agency background business strategy, disaster management, crisis response analytics – video analytics, data analytics, social media analytics military \/ defense background Planning a career in Geointelligence GIS Police analyst uses Geospatial skills to fight crime Some of the job roles that a Geointelligence professional might pursue; or closely associated with the field include GIS Practitioners, Analysts, Geography Researchers, Servicemen \/ ex-servicemen, members of law enforcement agencies, and imagery analysts. Engineers studying the fields of IT, Earth Science, Geoinformatics, Civil, Avionics, and Remote Science could also be considered for Geointelligence roles. The profession entails competency based learning, and USGIF has established the Universal GEOINT EBK (Essential Body of Knowledge), which identifies the knowledge, skills and abilities considered critical to the GEOINT profession. In case you’re planning a career in Geointelligence, it’s advisable to look up the EBK. There are essentially two ways of going about it. Both the paths involve you starting off with a basic degree. Let’s have a look: 1.Essentially here you have a basic degree, and you are someone who has also worked, reflecting on a job experience. Now, you can go ahead and pursue professional certifications in Geointelligence, and you could go for a specialization in an intelligence tradecraft. Your focus should be on enhancing the learning curve in the fields you chose. 2.The second road is where you might have already acquired a diploma\/certification in Geointelligence, right after pursuing your basic degree. Here, your next move should be to acquire robust job experience in one or more domains. Industries that need Geointelligence professionals To see a future in the Geointellignce field, it’s also important to plan what industry, government agency, or sector of the economy you’ll be working in. Moreover, there are several sub-sectors in the profession, and academia, so you must identify your niche. Next, you must outline the courses you plan to take in college or the extra study you might pursue during your spare time. From national security to national policymaking, and stretching across several domains of commerce and technology, Geointelligence professionals can choose from an array of career options. This list might give you a brief idea about the types of career roles, industries, and employers you might see in the field. This list is intended at shedding light on the career options you could consider to seeing a future in Geointelligence: Military – Geospatial Intelligence Agency, CIA, and other DoD Environmental – EPA, Nature Conservancy, other non-profits, environmental consulting firms Retail\/Commercial Real Estate – Large retailers, consulting firms, technology firms developing relevant applications for real estate decisions and real estate professionals Municipal\/County\/State Government Agencies – Maintaining a GIS back-office for local government agencies to assist with tax collection, utilities, etc. Urban Planning – Similar to local government but focused on planning and design. Geohealth – CDC, NIH, hospitals, public health agencies. Other commercial\/business uses. Other government agencies, e.g., Census Scope for India General Bikram Singh inaugurating GeoIntelligence India 2014 Today, about 80 percent of business data worldwide has a geographic component, making location intelligence and business intelligence crucial elements of any business analytics strategy. Geointelligence is designed to turn data into insight. The domain uses tools that draw on a variety of data source maps, demographics, and the GIS. In India, concerns like shortage of skilled human resource in the fields of GIS & Remote Sensing obstruct the industry from prospering. Both the fields are still in their dormant stages and have tremendous potential to grow. On final thought, Geointelligence is the best move as a career for someone with the skills and zeal to shine in the field. Geointelligence professionals in India will surely gain accelerated growth & development prospects.","excerpt":"The Geointelligence market today is mainly driven by the demand for convergence of geospatial data with mainstream technologies. Trends suggest that there has been an escalation in the demand for geospatial analytics solutions with artificial intelligence capabilities across various industries. This is not all, as commoditization of geospatial data for industry use has also led […]","categories":["AI Trends"],"tags":["AI (Artificial Intelligence)","AI Certifications","Business Intelligence India","certification for artificial intelligence","Ethical Hacking","forecasting","geospatial data","uav companies in india"],"author_name":"Amit Paul Chowdhury","publish_date":"2017-05-11T08:11:48","publication_year":"2017","word_count":1100,"keywords":["Business Intelligence India","Go","artificial intelligence","geospatial data","AI","AI Certifications","Scala","RAG","certification for artificial intelligence","Ray","ViT","analytics","forecasting","GAN","uav companies in india","Ethical Hacking","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","analytics","Ray","RAG","R","Go","Scala","GAN","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/make-career-geointelligence\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10061245,"title":"Learn cutting edge technologies with IIT Madras: Advanced Certification in Software Engineering for cloud, blockchain &#038; IoT from Great Learning","content":"Cutting edge technologies such as IoT, Blockchain, and Cloud Computing have the most momentum, and companies are aggressively hiring technology talent to stay ahead of the curve in the face of digital disruption. These technologies have taken the world by storm – altering almost every industry out there. Connections of all kinds – from communication systems to currency and intelligent devices through what we call hyperconnectivity – are making it easier than ever to interact with one another like never before, both online and off. There will be close to 75 billion connected IoT devices by 2025 with $1.29 Trillion global IoT spending. [Security Today]The worldwide cloud computing market’s size will be $623.3 Billion by 2023. [Wipro] $20 Billion revenue to be generated using blockchain by 2024, with a 30% Reduction in infrastructure costs for banks. [Fortune] As the world progressively adopts frontier technologies such as IoT, Cloud Computing and Blockchain, our software engineers need to master these technologies. Software engineering is a rapidly evolving domain, with nearly 11.5+ million job openings, predicted by the US Bureau of Labor Statistics. As a result, software engineers must stay up to date with the latest development trends and become relevant in their field. It’s time to accelerate and stay updated with the changing times and needs. Mastering the domains like Cloud, Blockchain, and IoT helps one develop a deeper understanding of modern technologies, their interdependence and applications. Professionals today can leverage this demand to build a rewarding career with the Advanced Certification in Software Engineering for Cloud, Blockchain & IoT from IIT Madras and Great learning. Throughout the course, one can build awareness around advanced software engineering, cloud computing, IoT courses in India and blockchain knowledge, with a view towards real-world applications that drives business success. The advanced software engineering course benefits professionals and entry-level aspirants to gain in-depth knowledge, cross-functional competency, and expertise in a progressively changing economy. Furthermore, it focuses on making current and aspiring software developers’ job-ready and future-proof with in-demand and cutting edge technologies. Advanced certification in Software Engineering in Cloud, Blockchain & IoT opens up endless career possibilities before you in job role like: Big Data EngineerIoT ArchitectIoT Developer Cloud Engineer\/AnalystBlockchain Developer\/EngineerSoftware Developer\/Engineer With an average salary ranging between 15-30LPA.* *Source: Glassdoor & Indeed Program Overview Following a high-quality curriculum, the IIT Madras and Great Learning’s program in Cloud, Blockchain and IoT prepares learners for the challenging world of software engineering by pairing academic excellence with applied skills. As a result, the technology professionals who take this online advanced software engineering course find themselves well-rounded, skilful programmers with a deep understanding of the complex relationship between technology and its applications. Program Highlights Duration – 9 months with Convenient online format300+ Hours of Learning60+ Hours of Mentorship10+ Projects and Assignments1 Capstone ProjectGreat Learning’s Career Assistance ProgramWeekly personalized mentorship from experts Who is this Program for? Technology professionals looking to add cutting edge technologies to their repertoire.Mid-career professionals, managers, and engineers need certifications to supplement their existing technology skill sets.Recent Computer Science grads aspire to become part of the technology industry in leading roles. Who is this Program For? Professionals, graduates, and entrepreneurs seeking to harness technology, the biggest drivers of change: Blockchain, Cloud, and IoT, to instil confidence when navigating the digital frontier. Desired Candidate Profile Applicants having 50% or above in Xth, XIIth and Bachelor’s degrees.Suitable for both professionals with experience and recent graduates Industry-Relevant Curriculum Followed MODULE 1 Design and Architecture of Software Systems Software Design & ArchitectureDatabase Designs and Applications MODULE 2 IoT Data Sources & NetworkingIoT on CloudIoT Big Data ProcessingIoT Analytics MODULE 3 Software Engineering for the Cloud Architecting applications for the CloudContainers & MicroservicesDevOps for the Cloud MODULE 4 Developing for blockchain Fundamentals of BlockchainBlockchain Development Dedicated Career Support to help you reach your career goals Access career support with Great Learning Excelerate, where you can connect with 1200+ hiring partners to land your next job with a 50% average hike and 66% alumni placed at top companies like Google, IBM, Amazon, Microsoft and more! Resume Building sessions – create a resume that highlights your skills from the program, along with other strengths acquired during this time.Curated jobs – access to a list of jobs from 1200+ hiring partners closely relevant to your domain and experience. Interview preparation – workshops to help you ace your technical interviews with the help of industry expertsCareer guidance and mentorship – get an expert career mentor personalized to your industry experience.Dedicated program manager to solve your queriesE-portfolio building ¬– an amalgamation of all the completed projects and skills acquired during the program. Master the skills needed to be at the top and explore advanced software engineering techniques to be future-ready! Apply now to this IIT Madras and Great Learning’s online certification course.","excerpt":"Throughout the course, one can build awareness around advanced software engineering, cloud computing, IoT courses in India and blockchain knowledge, with a view towards real-world applications that drives business success.","categories":["AI Trends"],"tags":["Courses","IIT"],"author_name":"Sreejani Bhattacharyya","publish_date":"2022-02-22T11:00:00","publication_year":"2022","word_count":796,"keywords":["Go","API","AI","cloud computing","Git","RAG","microservices","analytics","Courses","DevOps","R","IIT"],"extracted_tech_keywords":["AI","analytics","RAG","cloud computing","microservices","R","Go","Git","DevOps","API"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/learn-cutting-edge-technologies-with-iit-madras-advanced-certification-in-software-engineering-for-cloud-blockchain-iot-from-great-learning\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10062274,"title":"Under the hood: DRDO to build a cargo UAV for Himalayan frontier","content":"On February 7, Defence Research and Development Organisation (DRDO)’s Young Scientist Laboratory program (DYSL) was tasked with creating a new class of unmanned aerial vehicle (UAV) capable of flying in areas of high altitude with a payload capacity of 50kg to carry out logistic operations in the Himalayan frontier–as announced in the AAV seminar 2022. Young Scientist Laboratory program is an initiative formed with a group of five research establishments: All personnel including the scientists and the directors are below the age of 35. Located in Bengaluru, Chennai, Hyderabad, Kolkata, and Mumbai, the core focus areas of DYSL include artificial intelligence, cognitive technologies, asymmetric technologies, smart materials, and quantum technologies. In a recent proposal seeking development partners, DYSL outlined the requirement for an octocopter with a carbon composite air-frame having a gross weight (inclusive of payload) of 80kg. The octocopter is required to airlift a payload of 50 kg at mean sea level and about 20kg at extremely high altitudes of 15,000 feet. The drone should be able to operate in cold temperatures up to negative 20 degrees Celsius. Challenges Operating a drone in a harsh environment is an uphill task. The lower air density in high altitude areas makes it difficult to control a drone. Freezing temperatures tend to drain the batteries so the drone has a shorter flight time. A research paper by Irina K Romanova on drones used in mountain ranges reveals the two biggest challenges faced while flying a drone in mountain ranges: Communication is tricky in the mountain ranges as satellite networks, broadband, radio, etc. are limited due to the remoteness of the location and lack of proper infrastructure. Due to extreme weather conditions, one may experience a lot of delays, limited bandwidth and many such difficulties. Sub-zero temperatures and the accumulation of snow makes the maintenance and development of broadband technologies quite difficult.Collision avoidance is one of the most difficult challenges when dealing with drones in mountain ranges. It is usually based on two components; collision detection using sensors\/algorithms and a protocol to determine what participating agents should do to prevent such collisions. Therefore, it is important to understand the features of the network that the UAV is compatible with. Recent advancements DYSL (CT) lab was set up for providing cutting edge technology to DRDO in the cognitive field. The lab works in the area of design and development of Cognitive Radio and Cognitive Radar Systems, realising through deep neural networks and reinforcement learning algorithms. Below, we list the significant tech developments from India over the past decade: Mission Sensor Technology: As part of the UAV program, the Aeronautical Development Establishment (ADE) two-axes two-gimbal stabilised payload assemblies for medium altitude air vehicles. This enables the drones to carry multi-mode payloads with medium-range target acquisition and day-night tracking. GAGAN(GPS and Geo-augmented navigation system): India’s initiative to establish a regional satellite-based augmentation system (SBAS). Source: icao.int The system improves the accuracy of global navigation satellite system (GNSS) receivers by sending reference signals. The end goal is to develop a navigation system that can be utilised for all stages of flight over Indian airspace and the surrounding areas. NETRA: This is an indigenous airborne early warning and control system. The joint venture by DRDO and Indian Air Force showcases its unique Active Antenna Array Unit (AAAU) and secondary surveillance radar capabilities. Source: DRDO Latest drones by India DRDO Rostom II: It is a medium altitude long endurance UAV built for the Indian armed forces(Army, Navy and Air Force). Inspired by NAL’s LRCA(Light Canard Research Aircraft), this drone was developed under the leadership of Dr Rostom Damania. Source: twitter DRDO Imperial Eagle – It is a lightweight mini-UAV developed by the Aeronautical Development Establishment of NAL. It is primarily used by National Security Guard and military services. The drone can be tracked using Automatic Gain Control or GPS systems. DRDO Ghatak- It is a stealth type unmanned combat aerial vehicle (UCAV) developed by DRDO in a program called AURA( Autonomous Unmanned Research Aircraft). Although the program is still in its project definition stage, the full-scale prototype is expected by 2025.","excerpt":"DYSL (CT) lab works in the area of design and development of Cognitive Radio and Cognitive Radar Systems, realising through deep neural networks and reinforcement learning algorithms.","categories":["AI Features"],"tags":[],"author_name":"Kartik Wali","publish_date":"2022-03-08T17:00:00","publication_year":"2022","word_count":682,"keywords":["Go","artificial intelligence","programming_languages:R","AI","neural network","programming_languages:Go","Ray","GAN","R"],"extracted_tech_keywords":["AI","artificial intelligence","neural network","Ray","R","Go","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/under-the-hood-drdo-to-build-a-cargo-uav-for-himalayan-frontier\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10102104,"title":"Can LangChain Survive the Multi-Agent Invasion?","content":"The past few weeks have been quite exciting in the LLM landscape with the significant rise of multi-agents like XAgent, AutoGen, MetaGPT, and BabyAGI among others. Many developers are aggressively experimenting with them to solve maths problems, dynamic group chat, multi-agent coding, retrieval-augmented chat (RAG), building AI chatbots in a simulated environment, and conversational chess, among others. All of these developments bring us to question the relevance of LangChain in the new era of multi-agents. Ironically, in an AMA on Reddit, Harrison Chase co-founder of LangChain said, “No one really knows where LangChain will go.” When asked if they intend on moving from building a single agent to multi-agent like AutoGen, he responded, saying: “Yes we are considering it. The main thing blocking us from investing more is the concrete use cases where multi agent frameworks are actually helpful.” For LangChain, the focus currently has been on refining smaller, specialised components, recognising the inherent difficulty in this task. An example of this could be LangSmith, which is designed to expedite debugging and transition from prototype to production for such applications. The primary emphasis has been on chains and a single base AgentExecutor. At the same time, a lot of developers are experimenting, and amalgamating AutoGen and LangChain. Interestingly, it works. Currently, AutoGen doesn’t support connections to external data sources in the native framework, and LangChain steps in to fill in that gap. So far, so good LangChain marked its one-year anniversary this month. During this time, they’ve gained popularity and at the same time received criticism for their inefficiencies. As of today, they have 65.8k stars on GitHub and are built by over 5,000+ contributions by 1,500+ contributors. The community-led open source platform has developers who use it extensively while the frustrated lot have gone on to make their own alternatives. In contrast, the mushrooming AI multi-agents are still in their early days, and a lot of them are still experimenting with it, unlike LangChain, which has real industrial and business use cases. LangChain vs AutoGen The primary difference between them is that LangChain is a framework for building agents, which means that it provides the tools and infrastructure needed to create and deploy agents. AutoGen, on the other hand, is an agent that can converse with multiple agents. Within the LangChain framework, there exists a subpackage called LangChain Agents, specifically designed for harnessing LLMs to make decisions and take actions. LangChain Agents encompasses a variety of agent types, one of which is the ReAct agent. The ReAct agent is particularly noteworthy as it integrates both reasoning and acting processes when utilising LLMs. It’s primarily tailored for use with LLMs that precede the capabilities of ChatGPT. It’s important to note that all agents included in LangChain Agents adhere to a single-agent paradigm. In other words, they are designed to function individually and aren’t inherently geared toward facilitating communication or collaborative modes. Due to these identified limitations, the multi-agent systems present in LangChain, such as the re-implementation of CAMEL, are constructed from the ground up and do not rely on LangChain Agents. However, they maintain a connection to LangChain by utilising fundamental orchestration modules provided by LangChain, including AI models wrapped by LangChain and their corresponding interfaces. While AutoGen is more focused on building conversational AI applications with multiple agents.  AutoGen also provides a number of features that are specifically designed for building conversational AI applications, such as support for multi-agent conversations and context management. Another key difference between LangChain and AutoGen is their approach to integrating LLMs with other components. LangChain uses a chain-based approach, where each chain consists of a number of components that are executed in sequence. AutoGen, on the other hand, uses a graph-based approach, where components can be connected in different ways to create complex conversational flows. In conclusion LangChain is essentially a framework that makes it easier to build applications on top of large language models. It is typically built using a sequence of steps or ‘chains’ and consists of a number of components like data sources, API calling, code generation, and data analysis etc. This complicated the framework and many users felt that it is too verbose to use. There is a complete lack of documentation, filled with bugs and many users have noticed that LangChain introduces unnecessary abstractions and indirection making simple LLM tasks more complex. Chase, the co-founder of the framework admitted to some flaws on HackerNews and explained that work is being done to fix the issues. “In the past three weeks, we’ve revamped our documentation structure, changed the reference guide style, and worked on improving docstrings to some of our more popular chains. However, there is still a lot of ground to cover, and we’ll keep on pushing,” he said. AutoGen took what LangChain agents can do a step further. Instead of working on one agent at a time, AutoGen enables multiple agents to engage in collaborative task completion by providing adaptable, conversational, and flexible functions in various modes. These AutoGen agents seamlessly integrate with LLMs, human inputs, and a range of tools to suit the task’s specific requirements. It’s, in fact, only a matter of time until LangChain introduces multi-agent capabilities.","excerpt":"Multi-agents like AutoGen are growing, raising questions on LangChain’s future.","categories":["AI Features"],"tags":["Agents","langchain","RAG","wrapper"],"author_name":"K L Krithika","publish_date":"2023-10-27T10:00:00","publication_year":"2023","word_count":862,"keywords":["BabyAGI","ChatGPT","Agents","wrapper","AI","chatbots","ML","MetaGPT","RAG","LangChain","multi-agent systems","AutoGen","langchain"],"extracted_tech_keywords":["AI","ML","ChatGPT","LangChain","AutoGen","BabyAGI","MetaGPT","multi-agent systems","RAG","chatbots"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/can-langchain-survive-multi-agents-winter\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10064948,"title":"Hackathon alert! MachineHack launches ‘Data Engineering Championship’ as part of DES2022 Summit","content":"Data Engineering Summit 2022, presented by Google Cloud and organised by Analytics India Magazine, is India’s first conference dedicated to the high-demand and impactful field of data engineering. This virtual conference, to be held on April 30, 2022, will focus on data engineering innovation and give attendees direct access to top engineers and innovators working in leading tech companies. This will be a golden opportunity for attendees to learn about the software deployment architecture of machine learning systems, how to produce the latest data frameworks and solutions for business use cases from the very best in the field. Data Engineering Championship by MachineHack MachineHack is organising a data engineering hackathon for data scientists & data engineers to participate and win a chance to present at DES 2022. Data engineering consists of collecting, provisioning and maintaining excellent quality data to get insights. In order to do that, a data engineer needs to design and develop a scalable data architecture, set up processes that pool data from multiple sources, check the data quality, and eliminate corrupt data. In addition, exploratory data analysis (EDA) and extract, transform, and load (ETL) techniques are required to access and use downstream to solve business problems. START DATE: 13th April 2022, 6:00 PM END DATE: 30th May 2022, 6:00 PM REGISTER NOW All you need to know about the ‘Data Engineering Championship’ With the dataset provided, the participants need to analyse and create features of the following description. ‘DATE’: create the date from year, month and day of the week ‘LOW’: Lower value of DEP_TIME_BLK‘HIGH’: Higher value of DEP_TIME_BLK‘TIMESTAMP’: create a timestamp with date and lower value of DEP_TIME_BLK‘WIND_CHILL’: the perceived temperature due to cooling effect of wind blowing‘PRCP_SNOW_RATIO’: ratio of precipitation and snow‘PLANE_AGE_AIRLINE_AIRPORT_FLIGHTS_MONTH_RATIO’: ratio of plane age and airline and airport flights months.‘SEAT_DISTRIBUTION’: Ratio of seats and in  concurrent flight CONCURRENT_FLIGHTS‘SEAT_DISTRIBUTION_NORMALISED’: normalized values of ratio of seats and in  concurrent flight Evaluation In order to calculate the winners of the hackathon, the submissions will be evaluated using the mean absolute error. One can use sklearn.metrics.mean absolute error to calculate the same mean_squared_error(y_true, y_pred, squared=False). This hackathon will support private and public leaderboards. The public leaderboard is evaluated on 30% of the datasetThe private leaderboard will be made available at the end of the hackathon, which will be evaluated on 100% of the datasetThe final score represents the score achieved based on the Best Score on the public leaderboard How to generate a valid submission file? In order to submit your file, the following steps have to be kept in mind. Sklearn models should support the predict() method to generate the predicted values.The participant should submit a .csv file with exactly 2,00,00 rows with 9 columns. The submission will return an Invalid Score if you have extra rows or columns.The file should have exactly 9 columns. Points to note: One should not shuffle the sequence of the test seriesIf you are using pandas, use the following submission code: submission_df.to_csv(‘my_submission_file.csv’, index=False Dataset: 200000 rows x 26 columns MONTH: MonthDAY_OF_WEEK: Day of WeekDEP_DEL15: TARGET Binary of a departure delay over 15 minutes (1 is yes)DISTANCE_GROUP: Distance group to be flown by departing aircraftDEP_BLOCK: Departure blockSEGMENT_NUMBER: The segment that this tail number is on for the dayCONCURRENT_FLIGHTS: Concurrent flights leaving from the airport in the same departure blockNUMBER_OF_SEATS: Number of seats on the aircraftCARRIER_NAME: CarrierAIRPORT_FLIGHTS_MONTH: Avg Airport Flights per MonthAIRLINE_FLIGHTS_MONTH: Avg Airline Flights per MonthAIRLINE_AIRPORT_FLIGHTS_MONTH: Avg Flights per month for Airline AND AirportAVG_MONTHLY_PASS_AIRPORT: Avg Passengers for the departing airport for the monthAVG_MONTHLY_PASS_AIRLINE: Avg Passengers for the airline for the monthFLT_ATTENDANTS_PER_PASS: Flight attendants per passenger for airlineGROUND_SERV_PER_PASS: Ground service employees (service desk) per passenger for airlinePLANE_AGE: Age of departing aircraftDEPARTING_AIRPORT: Departing AirportLATITUDE: Latitude of departing airportLONGITUDE: Longitude of departing airportPREVIOUS_AIRPORT: Previous airport that aircraft departed fromPRCP: Inches of precipitation for the day SNOW: Inches of snowfall for the daySNWD: Inches of snow on the ground for the dayTMAX: Max temperature for the dayAWND: Max wind speed for the day START DATE: 13th April 2022, 6:00 PM END DATE: 30th May 2022, 6:00 PM REGISTER NOW Prize The three winners will be getting a chance to present their solution approaches at the Data Engineering Summit (DES 2022). Submission deadline If you want to be a part of this exciting hackathon, make sure to submit your entries by May 30, 2022, at 06:00 PM IST, as the private leaderboard will be frozen at that time. Disqualification If any of the details entered are found incorrect, Analytics India Magazine reserves the right to disqualify any participant.Any external dataset usage is strictly prohibited. The participants will be disqualified if found using any external dataset. So what are you waiting for? Register now to participate in this hackathon.","excerpt":"Data Engineering Summit 2022, presented by Google Cloud and organised by Analytics India Magazine, is India’s first conference dedicated to the high-demand and impactful field of data engineering. This virtual conference, to be held on April 30, 2022, will focus on data engineering innovation and give attendees direct access to top engineers and innovators working […]","categories":["Deep Tech"],"tags":["Data Engineering","data engineering career","feature engineering","feature engineering in machine learning"],"author_name":"Sreejani Bhattacharyya","publish_date":"2022-04-13T18:05:17","publication_year":"2022","word_count":782,"keywords":["Go","feature engineering in machine learning","machine learning","AI","ETL","feature engineering","data engineering career","Scala","Git","data engineering","Data Engineering","analytics","R","Pandas"],"extracted_tech_keywords":["AI","machine learning","analytics","Pandas","R","Go","Scala","Git","data engineering","ETL"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/hackathon-alert-machinehack-launches-data-engineering-championship-as-part-of-des2022-summit\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10064752,"title":"A hands-on guide to Frequentist Vs Bayesian approaches in statistics","content":"Statisticians have been debating about which one is better the Bayesian approach or the Frequentist approach because they both work on different methodologies. In the past, A\/B testing has tendered towards Frequentist solutions. However, Bayesian methods offer an appealing way of calculating experiment results that is entirely different from Frequentist. So in this article, we will discuss Frequentist and Bayesian approaches in detail. Following are the topics to be covered in this article. Table of contents What are Bayesian and Frequentist statistics?Why both of them are important?How these two are used in python? Using only the data from the current experiment, Frequentists predict the underlying truths of the experiment whereas the Bayesian method uses the past experiment data also to understand the current experimental data and make predictions. Let’s deep dive into these two statistical methods and understand the fundamentals. What are Bayesian and Frequentist statistics? Bayesian inference interprets probability as a measure of the belief or confidence that an individual might hold regarding the likelihood of an event occurring and the prior beliefs about an event will likely change when new information is revealed. Bayesian statistics gives a solid mathematical means of incorporating the prior beliefs, and evidence, to produce new posterior beliefs. The frequentist inference interprets probability as the frequency of repeatable experiments and the gathering of information. So, having large samples solves most problems because the sample becomes closer to the true population distribution. The frequentist inference has a lack of context and is completely dependent on the drawn sample. Solving a problem Consider an unfair coin flip problem in which need to find the probability of an unfair coin coming up heads. Let’s solve this problem with help of both of the above approaches. With the Frequentist approach, the probability of seeing heads when flipping an unfair coin over some time is the long-run relative frequency of seeing heads. In other words, as the coin is flipped more, the number of heads received as a percentage of the overall flips tends to correspond to the “true” probability of the coin landing on its head. Particularly, the person experimenting doesn’t consider their own beliefs about the fairness of other coins. With the Bayesian inference, an individual may believe the coin to be fair before flipping it but the coin keeps coming up heads after a few flips. As a result, the belief about the fairness of the coin is modified to take into account the possibility that three heads have appeared in a row and the coin might not be fair. The individual believes the coin is unlikely to be fair after 500 flips with 400 heads. This is a significant change from the individual’s prior belief. Are you looking for a complete repository of Python libraries used in data science, check out here. Why both of them are important? In A\/B testing, a statement is stated as “The 95% confidence interval can be exceeded in fewer than 5% of implemented variations”. With the use of the Frequentist method which calculates statistical significance because they can reliably predict future performance by using mathematical formulas. It relies on the p-value method. Bayesian calculations could be led to an incorrect conclusion considering to above statement because of the risk that prior experiment knowledge may not match how an effect is being generated in a new experiment, and it’s possible to be led astray if does not account for it. So, for A\/B testing Frequentist method is optimal and for every other statistical operation where relying on the p-value could lead to bad results Bayesian inference is implemented. How these two are used in python? Frequentist test The Frequentist test is a hypothesis testing so we can use any test from the below-given chart. Let’s implement any one of the below-given hypothesis testing methods. Two-sample t-test The two sample t-test could be used for testing both dependence and independence of the samples. The objective is to check if sample_1 and sample_2 have the same mean. Null hypothesis (H0): Sample_1 and Sample_2 have the same mean (dependent). Alternative hypothesis (Ha): Sample_1 and Sample_2 have different mean (independent). Import libraries: import numpy as np from scipy import stats Creating normalised samples sample_1=stats.norm.rvs(loc=2000,scale=500,size=100) sample_2=stats.norm.rvs(loc=1000,scale=500,size=100) Two array samples will be generated with 100 sample size and each having random values. Testing the sample for independence: print(\"T statistics value:\",stats.ttest_ind(sample_1,sample_2)[0],\"\\nP-value:\",stats.ttest_ind(sample_1,sample_2)[1]) T statistics value: 14.143 P-value: 7.58407491582331e-32 As the result shows the p-value is less than the margin of error (alpha=0.05) therefore we can reject the null hypothesis this means both the samples have different mean values. Bayesian test In this test, we will be implementing a model which gives Bayesian inferences based on the Markov chain Monte Carlo. The objective is to predict the true number of packing materials sold by a company per hour. We will assume that there is no variation in the manufacturing of the material. Install emcee library: ! pip install -U emcee Import libraries: import emcee import numpy as np from scipy import stats Building necessary functions for model a= 2000 new=sample_1 new_e=np.sqrt(new) def log_prior(theta): return 1 # flat prior def log_likelihood(theta, F, e): return -0.5 * np.sum(np.log(2 * np.pi * e ** 2) + (F - theta[0]) ** 2 \/ e ** 2) def log_posterior(theta, F, e): return log_prior(theta) + log_likelihood(theta, F, e) All the parameters needed for the bayesian’s theorem are created using the formulas. ndim = 1 nwalkers = 100 nburn = 1000 nsteps = 2000 starting_guesses = 2000 * np.random.rand(nwalkers, ndim) Variables necessary for the Markov chain Monte Carlo like the number of features in the samples, the number of walkers which are the samplers sampling around the data, etc. Build the model: model= emcee.EnsembleSampler(nwalkers, ndim, log_posterior, args=[new, new_e]) model.run_mcmc(starting_guesses, nsteps) sample = model.chain sample = model.chain[:, nburn:, :].ravel() print(\"\"\" true = {0} test = {1:.0f} +\/- {2:.0f} (based on {3} measurements) \"\"\".format(a, np.mean(sample), np.std(sample), 100)) So the actual number of items was 2000 and the predicted value is 1861 which is very much close to the actual number. Final Word According to Frequentists, probability is related to the frequency of repeated events, whereas Bayesian thinks of probability as a measure of uncertainty. With a hands-on implementation of this concept in this article, we could understand the difference and importance of Frequentist and Bayesian approaches in statistics. References Link to the above codeDocumentation for EMCEE","excerpt":"According to Frequentists, probability is related to the frequency of repeated events, whereas Bayesian thinks of probability as a measure of uncertainty.","categories":["Deep Tech"],"tags":["bayesian inference","hypothesis testing","Statistics"],"author_name":"Sourabh Mehta","publish_date":"2022-04-13T11:00:00","publication_year":"2022","word_count":1058,"keywords":["data science","NumPy","programming_languages:R","AI","Statistics","bayesian inference","Ray","Python","hypothesis testing","programming_languages:Python","data_tools:NumPy","R"],"extracted_tech_keywords":["AI","data science","Ray","NumPy","Python","R","programming_languages:Python","programming_languages:R","data_tools:NumPy"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/a-hands-on-guide-to-frequentist-vs-bayesian-approaches-in-statistics\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10023411,"title":"The Bigger Your Database, The More Lucrative You Are For Cybercriminals: Pukhraj Singh, Cyber Intelligence Analyst","content":"The recent cases of cybersecurity breaches, including the data leak at MobiKwik, Facebook, Juspay, etc, have exposed the chinks in the tech ecosystem. Analytics India Magazine got in touch with Pukhraj Singh, a cyber intelligence and geopolitical analyst with 15 years of experience, to understand the impact of these massive data leaks and how companies can prevent themselves from such losses. Singh played an instrumental role in incubating the cyber defence operations centre of the National Technical Research Organisation, under the National Security Advisor in the Prime Minister’s Office, India. He is also a member of the Australian Institute of Professional Intelligence Officers. Singh is currently working as the product owner of cyber threat detection at BAE Systems Applied Intelligence (SilverSky). Edited excerpts — AIM: 2021 have witnessed massive data leaks of big tech companies, including MobiKwik, Facebook, WhatsApp, Juspay, and many others. What is your take on such an alarming trend? Pukhraj Singh: First of all, defending against certain data leaks, certain intrusions, or attacks – it’s quite a challenging exercise, even for big companies. The bigger the company is, the more challenging it becomes because the stakes are very high. For instance, in MobiKwik’s case, it is the way the data leak cases are handled. Typically the information is not disclosed, along with some intelligent denial or underplaying the whole thing, until they are not forced to disclose every bit. This has happened in the past and can happen in the future, which is indeed worrying for us. Companies must communicate clearly about the breach to their customers and every stakeholder involved during these data leaks. If the companies try to cover such instances, they can swell up and go against them. It is required for the companies to be more transparent about such leaks, which will get them more assistance and add more support from organisations and communities at large. Secondly, on a broad scale, data leaks are inevitable as we work with these massive databases. The bigger your database, the more lucrative target it becomes. Eventually, companies have to deal with not how and what, but when would it get leaked. We’ve probably reached a point where most of our information taken by the collectors of these databases would largely be accessible to everyone. Eventually, interfaces are becoming more data-centric. AIM: How these data leaks impact the big tech companies? Pukhraj Singh: More than these tech companies, the consumers, including the subscribers and the end-users, are the ones that are massively impacted by these data leaks. For the impacted companies, these situations are unavoidable. Therefore companies need to know the right way to handle such situations. It is imperative for the companies to be transparent about these leaks, for their customers and stakeholders to understand the situation in depth. Usually, such situations also entail huge reputational damage, which is generally irreversible. A similar instance was witnessed during the Cambridge Analytica case, where Facebook’s reputation has taken a beating. The company was dragged into multiple senate Committees and senate hearings, which ended up being embarrassing for the brand image. AIM: With your 15 years of experience, what do you think is the cause of these regular data leaks? Pukhraj Singh: It’s technically very difficult to pinpoint a single reason for these data leaks. I believe it’s just the nature of the beast we’re dealing with. So, when one creates a massive data collection system to store users’ data, the system will go through multiple interfaces and processes with huge technological architecture. When such complexities are introduced in the architecture to consume data from millions of users in real-time, the complexity itself becomes the company’s weak point. The complexity is the root cause of insecurity or vulnerability. A defender of the company would have to look at every interface to check and address the weaknesses. Sometimes the cybersecurity experts also miss the broader picture of how the data flows through the whole system. On the contrary, an attacker just needs to find one weak interface to grow its tentacles. So it’s a very different ball game in terms of how the attacker and defender see a company’s cybersecurity landscape. It’s difficult to generalise, and every breach has a different story. Sometimes, leaks occur via simple ways like password hammering, or sometimes a complicated intrusion like the one we have witnessed in the case of solar winds. More worrying is the transparency in the response procedures, crisis communications, and these companies’ accountability. Further, India doesn’t have proper agencies to deal with such breaches and leaks, which is another significant concern. The more experience you get, the more you realise that everything is very nuanced, and it’s very difficult to find one short solution to anything. AIM: What can be done by the tech companies to fix these loopholes apart from playing a blame game or shrugging off the responsibilities? How dangerous can it be for the customers? Pukhraj Singh: The basic requirement is to change society’s structure to introduce privacy policies to handle data breach and data leaks. For instance, Europe took so many years to come up with its General Data Protection Regulation where they have set up a system, which only deals with privacy, just like we have our Criminal Procedure Code (CrPC) which deals with crime. So, data leaks cannot be the sole responsibility of one single company because they’re just a part of the problem; they’re not the problem itself. The ecosystem, as a whole, needs to encourage, incentivise and penalise lack of privacy for curbing these data leaks. These cases can be used as an excellent example to make certain broader policy changes around privacy to make the ecosystem more robust. Further, transparency and communication are an absolute must. It is indeed very alarming and worrying for customers because you can build an entire picture of a person based on his\/her data. The whole nature of intelligence is also turning upside down; intelligence now is not about gaining the secrets of some individual or some organisation, it is rather about joining the dots from public sources. With all of us using some forms of digital payments or having our data on digital platforms, there cannot be any escape from such leaks. AIM: RBI has asked companies like MobiKwik to have an external forensic audit to avoid such leaks. What more can be done to build a robust mechanism for user’s data protection in the long run? Pukhraj Singh: Although RBI has asked MobiKwik for an external audit, the auditor would be paid by a client, which creates an underlying conflict of interest. Moreover, many times neutrality of investigation may or may not be there in the first place. Secondly, RBI also has oversight on the auditor, but they will also need expertise in forensics, which again brings in massive complications. Thus, we cannot skirt the issue by just arranging an auditor for the company. To address such cybersecurity issues, companies need to assign proper cabinet-level attention to introduce privacy frameworks. A long-term strategic vision is critical for the company to deal with advanced level breaches and leaks. Privacy should become like a ministerial priority — just like we have a minister of petroleum and communications. Cybersecurity is definitely becoming a trivial concern that needs proper attention. AIM: May we know how users at an individual level can protect themselves from these data leaks? Pukhraj Singh: Cybersecurity on an individual level is a challenging task, as many a time, cybersecurity professionals can also fail to enforce good privacy settings on their mobile phones. The world is increasingly relying on mobile and web applications for their daily life; thus, having good operational security around the systems is no piece of cake. It all relates to threat perception. One’s privacy has to be subjective to his\/her threat perception. Alongside, it’s absolutely challenging to manage a software supply chain where every component is coming from someplace else. Thus, it becomes impossible to investigate the status of underlying hardware and software, as well as to secure databases. The only option is to switch oneself off from the grid or just exit the grid; however, that again isn’t practically viable. Thus, one can work towards minimising their digital footprint.","excerpt":"Pukhraj Singh, a cyber intelligence analyst, shares his thoughts and views on the recent data leak cases of tech companies.","categories":["AI Features"],"tags":["Cybersecurity","Cybersecurity India","data leak","data privacy use cases","Interviews and Discussions"],"author_name":"kumar Gandharv","publish_date":"2021-04-06T17:00:00","publication_year":"2021","word_count":1369,"keywords":["Go","programming_languages:R","AI","data leak","data privacy use cases","Git","RAG","Aim","Cybersecurity India","ViT","analytics","GAN","Cybersecurity","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","analytics","Aim","RAG","R","Go","Git","GAN","ViT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/the-bigger-your-database-the-more-lucrative-you-are-for-cybercriminals-pukhraj-singh-cyber-intelligence-analyst\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10082639,"title":"TSMC&#8217;s 3nm: Lost in Memory Lane","content":"Despite the growing voices declaring a possible end of Moore’s law, the semiconductor industry is still scaling chips along the historical trajectory. At the 68th Annual IEEE International Electron Devices Meeting (IEDM), TSMC spoke about its 3nm node technology, the class of which includes the original base N3 (N3B) node and the enhanced variant (N3E). However, what was left out of the conversation at the event is that while the logical circuits have been scaling well, SRAM, the memory infrastructure, is unable to keep pace. For instance, TSMC’s N3 consists of an SRAM bitcell size of 0.0199µm^², only ~5% smaller than N5, the 5nm node technology, which is 0.021 µm^². And the N3E features 0.021 µm^² SRAM bitcell, meaning it is the same size as N5. But, as Wikichip reports, the scaling crisis is not limited to TSMC. Instead, it is a phenomenon spread industry-wide (see graph). SRAM provides on-chip cache memory for microprocessors in PCs and mobile devices, often taking up much space on the chip. The memory cells allow cramming multiple functions onto a single chip without compromising on power and performance. Thus, the slowdown in SRAM scaling could hamper the amount of cache memory size included in processors. This would, in turn, affect the PC and smartphone pricing as well. The cache is particularly important to processors since it is faster than typical main DRAM memory by an order of magnitude and helps processors speed up operation. On the flip side, it is usually the toughest part to scale. Kaizad Mistry, corporate vice president at Intel, in speaking about SRAM, told Semiconductor Engineering back in 2013, “It is important to scale this component of the technology by close to 0.5x from one generation to the next. The key limiter to scaling the SRAM is to minimise transistor variation, which can make the SRAM cell unstable.” In the graph above, we can see the case of Intel alone slowing to 0.6-0.7x scaling from 0.5-0.6x scaling for their Intel 4 process. Issues with FinFET transistors Both Samsung and TSMC, considered to be the only foundries (along with Intel, maybe) capable of producing advanced node processes, use FinFET transistors for their 7nm and 5nm technologies. FinFETs are 3D-like structures with better performance and lower leakage than traditional planar transistors. But, with FinFETs, there are some evident limitations. The advanced node processes—namely 5nm and above—are going through what is called the “ageing problems”. Firstly, densely packed devices are highly susceptible to self-heating due to higher speeds, especially in the case of 2.5D and 3D designs. And secondly, the devices today are operating around 0.6 to 0.7 volts at 3nm, because of which the electric fields have increased, which can, in turn, cause the devices to break down. Thus, reliability issues are a major concern when it comes to FinFET models. Further, as Synopsys’ Haran Thanikasalam notes, “FinFETs can be multi-threshold devices, so when you have the entire gamut of threshold voltage being used in a single IP, we have so many problems because every single device will go in a different direction.” A mammoth task lies in front of SRAM of playing catchup to the scaling abilities of logical circuits, which might put the trajectory of Moore’s law on hold. However, scaling to 3nm and 2nm is not necessarily a failed cause. “The industry will find new ways to increase cache memory by potentially splitting the functionality across chiplets. 2nm will have the benefit of GAA transistor architecture which could potentially help improve the SRAM density,” Sravan Kundojjala, Principal Industry Analyst at Strategy Analytics, told AIM. TSMC 3nm versus Samsung 3nm To solve the issue, Samsung recently announced that it would transition to a new transistor architecture called the nanosheet FET for the 3nm process. Samsung’s proprietary GAA technology, known as Multi-Bridge-Channel (MBCFET), defies some of the limitations associated with FinFET. The 3nm GAA technology allows Samsung to adjust the channel width so it can utilise nanosheets with wider channels than other GAA technologies to optimise power usage and performance to meet various customer needs. The design flexibility of GAA is highly valuable for Design Technology Co-Optimisation (DTCO) as it is beneficial in areas of Power, Performance, and Area (PPA). For example, the second-generation 3nm process trumps the 5nm process by reducing power consumption by up to 50%, improving performance by 30% and reducing area by 35%. Contrary to Samsung, TSMC’s 3nm process technology would stick to FinFET transistors and instead rely on “innovative features” to achieve the full-node scaling. Amongst the additional features include the finflex technology, which allows designers to select a technology between three configurations that fares well with their needs. Each of the three configurations is a tradeoff between power, performance, and area. Road ahead General-purpose processors, graphics chips, and application processors for smartphones have huge caches, even when it is said to be inefficient in fetching data from memory, especially for AI and ML workloads. Some notable examples include AMD’s Ryzen 9 7950X carries 81MB of cache in total, and Nvidia’s AD102, which uses at least 123MB of SRAM. Anton Shilov, writing for Tom’s Hardware, resonates with Kundojjala while saying that the slowdown of SRAM scaling could be mitigated by “going for a multi-chiplet design and disaggregating larger caches into separate dies made on a cheaper node”. AMD has been doing this with its 3D V-Cache, which allows them to stack cache vertically on a processor. Moreover, as Wikichip puts across, some emerging technologies, such as MRAM, FeRAM, NRAM, RRAM, STT-RAM, and PCM, provide unique tradeoffs like higher density at lower read\/write specifications, non-volatility capabilities, lower read-write cycle capabilities, or lower power at potentially lower density or speeds.","excerpt":"While the logical circuits have been scaling well, SRAM, the memory infrastructure, is unable to keep pace","categories":["AI Features"],"tags":["Intel","moore's law","Samsung","tsmc"],"author_name":"Ayush Jain","publish_date":"2022-12-18T13:00:00","publication_year":"2022","word_count":942,"keywords":["Go","Samsung","programming_languages:R","AI","ML","programming_languages:Go","Aim","analytics","tsmc","moore's law","R","Intel"],"extracted_tech_keywords":["AI","ML","analytics","Aim","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/tsmcs-3nm-gets-lost-in-the-memory-lane\/","complexity_score":4,"technical_depth":8,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10002491,"title":"AI Does Not Hate You, Nor Does It Love You: Understanding The Value Alignment Problem","content":"Eminent British poet, Tom Chiver, in his book titled, The AI Does Not Hate You: Super-intelligence, Rationality and the Race to Save the World, talks at length about several reasons which could simulate the route of human extinction. Other than a terrifying looking AI, similar in appearance to the T-800 from Terminator II, one of them is the creation of genetically engineered viruses through the means of synthetic biotech. Overgrowing concerns with the latter, goes unnoticed with non-availability of sustainable investments for commencing any kind of research. The dystopian epoch, we all are experiencing should primarily raise alarming signals at our defunct heuristics and the failure to adopt holistic policies for a genetically engineered pandemic. As a matter of consideration, the validity of my opinions shall be realized, if we could presume that we are going to overcome the COVID-19 pandemic, otherwise this work is as good as a fictional output. The problem of value alignment with augmented technological tools gained ethical relevance with Isaac Asimov’s Three Laws of Robotics. The laws which are now 70 years old, were able to manage effective levels of control in artificial agents like Bertie the Brain, capable of tic-tac-toe game in the 1950s. Fast forward to 2016, AlphaGo, beats Lee Sedol in a game that has more possible positions than the number of atoms in the universe, merely powered with deep learning algorithms. With unsurmountable development in the domain of artificial intelligence over the past two decades, the same laws appear to have lost their charm to maintain robust value aligning rigidity between the machine and us.  Consistent research backed by computational power, in the field of artificial general intelligence shall evolve the concept of intelligence in the near future. Hypothetically a machine like AlphaGo might not only be good at playing Go, it may extend its capacity of purpose in other domains of operations, like policy making. Amidst such a transitory phase, the evolution of intelligence in artificial agents will challenge the systemic inheritance of control over them. It is a concomitant responsibility of its creators to ensure that losing control over its agents shall never lead to incorrigible consequences for us. This ideology is the foundation of the value alignment problem; in layman’s terms, we must be sure that the purpose desired by the creator is the same purpose an artificial agent seeks to achieve. The central logic revolves around artificial agents being made to act in altruistic ways so that it can never pose a threat to us. However, despite having an altruistic purpose, the agent can make decisions, backed with general intelligence, which will end up contradicting the purpose envisioned by the creator. This shall be understood better if we take a glance at the paperclip maximizer experiment. The experiment was brought in the discussion forum by Swedish Philosopher, Nick Bostrom, stating how an artificial general intelligence agent tasked with the objective of collecting paper clips, in order to maximize its collection will eventually turn earth into a paper clip. An artificial agent with human-level intelligence, initially, will take it upon itself to either collect those paper clips, buy those paper clips with money or manufacture them by itself. Inevitably, the agent shall go through a process of intelligence explosion, where it shall deduce newer and more effective methods of collecting paper clips. The agent shall never move away from its goal of collecting paper clips and in doing so, it will not be bothered about non-specific goals such as learning human sentiments or developing collective unconsciousness. Over here, we conclude that the values of an artificial agent are completely abnormal and alien to the multitude of values central to us. Therefore, in its process of optimization as a goal-seeker it will be exactly focused on the aspect of maximizing utility, which is the collection of paper clips in this instance. As an anthropological discourse, the value alignment problem, if viewed from the prospects of Asimov’s three laws of Robotics, appears to formulate a shield with respect to an existential threat to us by the acts of an artificial agent. This premise, irrespective of the nature of control on artificial agents, will cause serious hindrance in the practice of developing artificial agents that can co-exist with us and serve its purpose of being an augmented tool to improve our experience of existence. Beyond the norms of existential threat, the value alignment problem today, must focus on inculcating values in an artificial agent which shall not only cause an existential threat but it shall never be a cause of obstruction for us to flourish in our lifetime. Therefore, the duty of the creator of an artificial agent is not over when it has acted in all good faith, inheriting altruistic values while devising its purposes. The duty extends to making the artificial agent aware of our behavioral competencies, so that the agent doesn’t intervene with our ability to prosper in its presence. When a social media platform mechanizes machine learning tools to scrap your private data, for targeted advertisement purposes, it might not be causing a direct existential threat to your identity but it is surely in violation of your freedom to flourish. Due to our heuristically challenging abilities of decision making and abysmally large population, it is important for the artificial agent to keep learning the values it should align with and not be limited to a static accord of values. The presence of malign information from media portals, reflecting this paradigm as a war between us and artificial agents has already penetrated a pleading for greater control over these agents. However, the determinative motive shall be to accept its autonomy and secure a medium of co-existence by making them align with our ever thriving values. The process of automation can’t be rejected and therefore, any efforts to threaten its legitimacy by framing us with the idea of feeling oppressed will cause a blockade to the beautiful destiny of augmented reality. Honestly, there’s no chance of competing on a levelled pedestal with them. A biological neuron fires, maybe, at 200 hertz, 200 times a second. But even a present-day transistor operates at the Gigahertz. Neurons propagate slowly in axons, 100 meters per second, tops. But in computers, signals can travel at the speed of light. There are also size limitations like a human brain that has to fit inside a cranium, but a computer can be the size of a warehouse or larger. So the potential for super intelligence lies dormant in the matter, much like the power of the atom lay dormant throughout human history, patiently waiting there until 1945. REFERENCE Nick Bostrom, Superintelligence: Paths, Dangers, Strategies, Oxford Publication, (2014). Stuart Russell, Human Compatible: AI and the Problem of Control, Viking Press, (2019). Stuart Russell, Peter Norvig, Artificial Intelligence: A Modern Approach, Pearson Education, (2016). Tom Chiver, The AI Does Not Hate You: Superintelligence, Rationality and the Race to Save the World, Orion Publishing Group, (2019).","excerpt":"Eminent British poet, Tom Chiver, in his book titled, The AI Does Not Hate You: Super-intelligence, Rationality and the Race to Save the World, talks at length about several reasons which could simulate the route of human extinction. Other than a terrifying looking AI, similar in appearance to the T-800 from Terminator II, one of […]","categories":["AI Features"],"tags":["AI What it Does","how does artificial intelligence work","limitations of AI"],"author_name":"Sujoy Sarkar","publish_date":"2020-07-15T13:00:00","publication_year":"2020","word_count":1158,"keywords":["Go","artificial intelligence","machine learning","AWS","AI","TPU","Git","BERT","how does artificial intelligence work","deep learning","AI What it Does","R","limitations of AI"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","deep learning","AWS","TPU","R","Go","Git","BERT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/ai-does-not-hate-you-nor-does-it-love-you-understanding-the-value-alignment-problem\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10129434,"title":"Chinese AI Companies Surpass American Rivals","content":"Chinese tech heavyweight SenseTime recently released SenseNova 5.5 at the World Artificial Intelligence Conference in Shanghai, claiming a 30% performance increase over its predecessor and superiority over GPT-4o in several criteria. The company said that key enhancements include improved mathematical reasoning, English proficiency, and command following capabilities, putting it on par with GPT-4o in terms of interactivity and other core indicators. In early May, SenseTime released a demo similar to OpenAI’s GPT-4o demo, showing off the model’s visual skills. SenseNova 5.5 can recognise and describe specific items by pointing a smartphone camera at them while the AI operates. Similarly, Baidu USA CEO Robin Li claimed that Ernie 4.0 can produce better results than GPT-4o. Meanwhile, Alibaba’s Tongyi Qianwen (Qwen) models saw downloads shooting up to 20 million, tripling in only two months. Forget text generation—China has leaped ahead in video generation models too. While the world awaits the official release of Sora, the internet is abuzz with Kling’s videos depicting a series of animals enjoying a meal of noodles. In the meantime, OpenAI recently announced that as of July 9, it had blocked API access for users in unsupported countries and territories, including mainland China, Hong Kong, and Macau. China’s Race To Dominance Hugging Face co-founder and CEO Clement Delangue praised the progress made by Chinese AI firms in a post on X, saying, ‘Qwen 72B is the king, and Chinese open models are dominating overall.'” This is proven by the fact that Alibaba’s Qwen 2-72B model has claimed the top spot on Hugging Face’s current LLM Leaderboard, outperforming all other open-source models. “China’s advantage is doing whatever it takes to catch up,” said Kai-Fu Lee, a Taiwanese computer scientist and founder of China-based AI startup 01.AI. Lee’s firm open-sourced Yi-34B, its foundational LLM that outperformed Llama 2 on various benchmarks. He went on to explain how 14 months ago, they had nothing, including no GPT. “At that time, we were six or seven years behind, and at this moment, we are six to nine months behind. So the catch-up has already been happening. Going forward, we hope that it will continue,” Lee said. In May, China-based DeepSeek open-sourced its DeepSeek LLM, a 67-billion parameter model trained from scratch on a dataset consisting of 2 trillion tokens in both English and Chinese, hinting at its bid to go global. The model managed to outperform Llama 2, Claude 2, and Grok-1 in various metrics. Government Support Chinese tech companies have also been receiving support from the government in its ongoing AI battle against America. The Chinese government has made AI a national priority, aiming to become the world leader in AI by 2030. The Cyberspace Administration of China (CAC) has issued approvals for over 40 LLMs in the past six months, granting operational licences to 1,432 AI-driven applications. Meanwhile, according to a survey, over 80% of Chinese business leaders surveyed are currently using GenAI in their operations, way above the global average of 54% and the US average of 65%. China also dominates the global race in GenAI patents, filing more than 38,000 patents from 2014 to 2023, according to a UN report. That’s six times more than those filed by the US-based inventors. Geographically, China leads with 38,210 inventions, far surpassing the US (6,276), South Korea (4,155), Japan (3,409) and India (1,350). With the release of these models, the fear of China rising against US open-source models, particularly in the field of generative AI, is not unfounded. This competition is driving significant advancements in AI technology. Just a Copycat? But despite the nation’s mad rush to develop generative AI, Chinese businesses are almost wholly dependent on American underpinnings, such as open-sourced foundational research and technology developed by leading US companies and research institutions. “As a measure of how far behind they are, leading Chinese firms are comparing their performance to ChatGPT,” said Paul Triolo, technology policy lead and senior VP for China, Dentons Global Advisors. China’s businesses typically use “fine-tuned versions of Western models” because their own AI models “aren’t very good,” said Jenny Xiao, a partner at San Franciscan venture capital firm Leonis Capital. She added that Silicon Valley is unquestionably far ahead of the curve.For instance, some of the technology in Chinese firm 01.AI, which released its open-source model, came from LLaMA. Former Google CEO Eric Schmidt said that while China intends to take the lead in several industries, the US is still far ahead in artificial intelligence.","excerpt":"Forget text generation—China has leaped ahead in video generation models too","categories":["AI Features"],"tags":["AI (Artificial Intelligence)","AI Companies","Baidu","China","OpenAI"],"author_name":"Anshul Vipat","publish_date":"2024-07-17T18:10:40","publication_year":"2024","word_count":740,"keywords":["ChatGPT","GenAI","artificial intelligence","Hugging Face","OpenAI","AI","GPT-4o","RAG","Aim","Baidu","generative AI","AI Companies","AI (Artificial Intelligence)","China"],"extracted_tech_keywords":["AI","artificial intelligence","generative AI","GenAI","GPT-4o","ChatGPT","OpenAI","Aim","Hugging Face","RAG"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/chinese-ai-companies-surpass-american-rivals\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10018131,"title":"Central Electricity Regulatory Commission To Invest In AI To Streamline Operations","content":"The key regulator of the power sector in India, CERC (the Central Electricity Regulatory Commission), is planning to set up an AI-based Regulatory Expert System Tool (REST) to stream the information access and to help the team in releasing them of their duties. According to the news, with this launch, the CERC will be the first such quasi-judicial regulatory body to embrace AI and machine learning for their operations. This launch of AI-based regulatory expert system tools came when the Supreme Court closed down the commission for four months. The data stated that the commission’s website, therefore, has 1,000 pending cases pending with 173 cases, where the CERC has reserved the order, but not pronounced yet. Further, the website also displayed 450 fresh applications that are pending to be heard by CERC. The CERC, in December 2020, requested a consultant from the field of artificial intelligence, machine learning and natural language processing to develop an AI-based REST for the CERC. Thus, REST was created with the purpose of “building an institutional memory and for creation, reference, and intelligent retrieval of information\/documents to assist the commission in the discharge of its functions under the Electricity Act, 2003,” stated by CERC. Traditionally, the organisation has been working with SAUDAMINI (System for Adjudication Using Digital Access & Management of Information), a management information system for sourcing data. However, the officials believed that an AI-based intelligent system was critical for managing the increasing workload and bridging the lag that has been created with pending cases. The request stated that the number of petitions filed before the CERC had massively surged over the years, and, therefore, it was imperative to monitor and analyse such orders. Besides, the commission also receives information from the regulated entities, stakeholders and nodal bodies, that are important from a regulatory perspective. Such information contains voluminous and complex inter-related data that can only be handled by AI-based REST. Along with REST, the CERC will also receive an AI-based search and recommendation system, automated system creation and extraction of legal documents, and process automation data visualisation tools. According to CERC, the new system will later be merged with the existing management information system. Currently, the Supreme Court (SC) of the country is the one judicial organisation with an electronic filing (e-filing) system, that has taken centre stage due to Covid-induced lockdown. Similar guidelines have been issued for high courts and district courts, where an AI-based portal will efficiently reduce their paperwork. While the CERC commission is yet to receive bids, the deadline for building and deploying this new back-end system is six months from the date of selection of the consultant.","excerpt":"The key regulator of the power sector in India, CERC (the Central Electricity Regulatory Commission), is planning to set up an AI-based Regulatory Expert System Tool (REST) to stream the information access and to help the team in releasing them of their duties. According to the news, with this launch, the CERC will be the […]","categories":["AI News"],"tags":[],"author_name":"Sejuti Das","publish_date":"2021-01-15T10:33:01","publication_year":"2021","word_count":439,"keywords":["machine learning","artificial intelligence","programming_languages:R","AI","Git","automation","GAN","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","R","Git","GAN","automation","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/central-electricity-regulatory-commission-to-invest-in-ai-to-streamline-operations\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":65027,"title":"Why Self-Service Analytics Might See Bigger Adoption Amid Crisis","content":"The pandemic outbreak has made several businesses struggle with their continuity as well as keeping up their relevance during this uncertain time. And therefore, enterprises are now relying on analytics to make informed decisions during the time of crisis. By analysing the data, enterprises can understand the crisis and its impact on their businesses. This assists them in creating necessary strategies to enhance the efficiency of their employees. However, with the increasing volume of data, making these crucial decisions of understanding the crisis and the customer requirements demand smart analytics for businesses. It is essential for businesses to have self-service analytics across their operational layers in order to get the most out of their data and create beneficial insights from it. As compared to traditional analytics, self-service analytics enhances the whole process of data-driven decision making. It also provides access to data and information anytime and anywhere to every business department. Business intelligence and analytical tools usually depend on the size of the organisation, however, “self-service analytics is a significant topic for all companies seeking to better leverage vast information resources and a scarcity of expertise to drive improved group-based decision making in a governed fashion,” said Jim Ericson, research director at Dresner Advisory Services to the media. Despite numerous advantages, self-service analytics has never played a huge part in businesses. However, social distancing has indeed created the requirement for self-service analytics. In a recent survey done by Analytic India Magazine, it has been revealed that the highest proportion of respondents — 29%, are working on projects related to predictive analytics with 22% of respondents working on descriptive analytics related projects during this lockdown. Such numbers show that companies are actively working on self-service analytics projects to make better-informed decisions amid this crisis. Here are a few ways self-service analytics can help businesses enhance their productivity. Collated Necessary Data At One Place With traditional analytics, IT teams spend the majority of their time querying data and creating reports for business leaders to make informed decisions, which used to create delays for enterprises. However, self-service analytics can offer access to data to all the departments of organisations for them to make necessary decisions for the business. Majority of organisations still work on scattered data, where data is receding at different places, which could create inconsistency and loopholes in the record. Self-service analytics will allow organisations to integrate all their data at one place for leaders to have a better view of their businesses, employees, and customers amid this crisis. Such a capability is indeed necessary when businesses have opted remote working. For this, many companies also opt for cloud-based analytical solutions where it becomes easy for analysts to extract, transform, visualise, and analyse the data. Such smart tools will make an organisation less dependent on their technical team and analysts for data analysis and reporting during this COVID pandemic, thus will empower employees by providing refined access to data for increased productivity. Focusing Analyst’s Capabilities For Complex Data Science With people working remotely, traditional analytics forces the involvement of analysts to read the data and help business leaders make informed decisions, which takes up a considerable amount of their time. And, therefore, it gets difficult for the data scientists to focus on core analytical work, which is much more critical for your business. Whereas, self-service analytics empowers employees with data and therefore eliminates the constant requirement of analysts during this crisis while the whole workforce is working remotely. With this, analysts can focus on solving critical business problems. Considering we are living in the age of constant data generation, analysing a huge amount of data can be a difficult task for technical people of the organisations, and therefore gets challenging for business leaders to get the benefits of analytics. Self-Service analytics provide access to data and reporting to all remote working employees of the organisation for them to make informed decisions despite working separately. This will allow data scientists to focus their time and efforts on the long term and high-value projects, as well as creating strategies to solve complex issues and enhance customer experience, which is necessary for organisations to thrive in this competitive landscape. A sound self-service analytical system is indeed necessary for businesses to stay relevant during this crisis. It should be designed to collect, integrate, as well as analyse data across all departments for business leaders to make informed and beneficial decisions. Narrowing Skills Gap & Increase Employee Productivity The skills gap is indeed a concern in the COIVD world but will definitely remain an issue post the pandemic. With business employees working remotely, it becomes difficult to get access of your technical team at the time of their need, which indeed creates the requirement of a system that can be easily operated and understood by not-so-technical leaders and employees of the organisation. Self-service analytics has a smooth deployment process for organisations, which doesn’t require assistance from your technical team; neither from your analysts. Considering the businesses would also be not required any other assistance, it would also cut related costs related to hardware and infrastructure. By utilising self-service analytics, business leaders, as well as employees, can analyse required data and create specific reports to address business challenges without the help of any extra resources, and therefore becomes self-sufficient at this time of crisis. In turn, it also decreases the skills gap, as all employees will now be able to understand the importance of data. Alongside, self-service analytical tools enable collaboration between your employees and your technical team of analysts, which indeed can provide better productivity for businesses amid this crisis. Collaboration allows the teams to have a better understanding of the data, which will ensure better accuracy while making decisions. Comprehensive self-service analytics also provide them real-time updates on their employees, customers, and the overall business. A lot of small enterprises are also deploying self-service analytics who are willing to leverage the benefits of analytics in their organisation during this pandemic. Businesses that have implemented self-service analytics at the core of their organisations can stand out from their competition, reduce cost, and can increase profits by making informed decisions. Robust Security For Amid Remote Workers Data security becomes extremely important for organisations amid this crisis as the data will be in the hands of remote workers who will be accessing sensitive critical data from unsecured networks. Self-service analytics tools come with a robust security framework, which ensures that employees and business leaders have to go through levels of encryption to have access to the required data. These smart-systems help organisations to create necessary security standards to protect business information, which isn’t possible with traditional analytics. It also urges businesses to make data understandable by their remote workers, which creates a smoother flow of data, and the usage is also governed with self-service analytics. In order to run a business effectively during this uncertain time, it is imperative for business leaders to focus on creating effective business strategies rather than worrying about data and its security, which can reduce productivity. Having self-service analytics will ensure proper data monitoring and will also help the IT team to create the necessary security framework of the organisation. Self-service analytical tools offer customisable security tools for each of your employees, which not only controls the data access but also prevents third-party organisations from accessing sensitive financial information of the company. Such tools automate business processes and allow every employee of the organisation to explore data and share visualisations while maintaining cybersecurity protocols to protect critical business information.","excerpt":"The pandemic outbreak has made several businesses struggle with their continuity as well as keeping up their relevance during this uncertain time. And therefore, enterprises are now relying on analytics to make informed decisions during the time of crisis. By analysing the data, enterprises can understand the crisis and its impact on their businesses. This […]","categories":["AI Features"],"tags":[],"author_name":"Sejuti Das","publish_date":"2020-05-12T10:03:34","publication_year":"2020","word_count":1256,"keywords":["business intelligence","data science","Go","AI","R","RAG","ViT","analytics","GAN","predictive analytics"],"extracted_tech_keywords":["AI","data science","analytics","RAG","predictive analytics","R","Go","GAN","ViT","business intelligence"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/why-self-service-analytics-might-see-bigger-adoption-amid-crisis\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10064650,"title":"4 Best DALL.E Text to Image Alternatives to Create AI Image in 2024","content":"Creativity is just connecting things, said Steve Jobs: He was channelling his inner Einstein (coincidentally another Walter Isaacson muse), who had come up with ‘combinatory play’ to explain the inner workings of creative thought. OpenAI took the hint, and built a text-to-image generator, DALL.E. OpenAI has got creativity down to a science. Almost! The astronaut riding a horse in a photorealistic style or teddy bears mixing sparkling chemicals as mad scientist as a 1990s Saturday morning cartoon are good cases in point. The ultra-imaginative DALL.E has become the talk of the town in a short time. Below, we look at similar models making the rounds in the world of AI. DALL.E Text to Image In 2020, OpenAI introduced GPT-3 and, a year later, DALL.E, a 12 billion parameter model, built on GPT-3. DALL.E was trained to generate images from text descriptions, and the latest release, DALL.E 2, generates even more realistic and accurate images with 4x better resolution. The model takes natural language captions and uses a dataset of text-image pairings to create realistic images. Additionally, it can take an image and create different variations inspired by original images. DALL.E leverages the ‘diffusion’ process to learn the relationship between images and text descriptions. In diffusion, it starts with a pattern of random dots and tracks it towards an image when it recognises aspects of it. Diffusion models have emerged as a promising generative modelling framework and push the state-of-the-art image and video generation tasks. The guidance technique is leveraged in diffusion to improve sample fidelity for images and photorealism. DALL.E is made up of two major parts: a discrete autoencoder that accurately represents images in compressed latent space and a transformer that learns the correlations between language and this discrete image representation. Evaluators were asked to compare 1,000 image generations from each model, and DALL·E 2 was preferred over DALL·E 1 for its caption matching and photorealism. 1. CLIP Earlier, the OpenAI research team introduced an open-sourced text-image tool, CLIP. The neural network Contrastive Language-Image Pre-training was trained on 400 million pairs of images and text. The tool efficiently learns visual concepts from natural language supervision and can be applied for classification by providing the names of the visual categories to be recognised. In a paper introducing the model, the OpenAI research team wrote about CLIP’s ability to perform various tasks during pretraining, including object character recognition (OCR), geo-localisation, action recognition, and more. CLIP has proven to be highly efficient, flexible, and more generalised. Furthermore, it is far less expensive, given CLIP relies on text-image pair datasets already available on the internet. It can adapt to perform a broader range of visual classification tasks. 2. RuDALL.E ruDALL-E takes a short description and generates images based on them. The model understands a wide range of concepts and generates completely new images and objects that did not exist in the real world. The Russian take on OpenAI, ruDALL.E, is trained on ruGPT-3, which was trained on 600GB of Russian text. The Russian ruDALL.E model boasts 1.3 billion parameters and a YTTM text tokeniser with a dictionary of 16,000 tokens. It leverages a custom VQGAN model that converts an image into a sequence of 32×32 characters. There are two running models of the tool, Malevich (XL) trained on 1.3 billion parameters with an Image encoder and Kandinsky (XXL) with 12 billion parameters. On running the former model with the same text input as the latest DALL.E example of “an armchair in the shape of an avocado”, ruDALL.E was found to comprehend combining chair and avocado in the function of a shape. 3. X-LXMERT Created by AI2 Labs, X-LXMERT is an extension of LXMERT, a transformer for visual and language connections. The tool comes with training refinements and enhanced image generation capabilities, rivalling models specialised in image generation. X-LXMERT has three key refinements: Discretising visual representations, using uniform masking with a large range of masking ratios, and aligning the right pretraining datasets to the right objectives. On their project page, the X-LXMERT research team explained the training as such: “We employ Gibbs sampling to iteratively sample features at different spatial locations. In contrast to text generation, where left-to-right is considered a natural order, there is no natural order for generating images.” 4. GLID-3 GLID-3 is a combination of OpenAI’s GLIDE, Latent Diffusion technique and OpenAI’s CLIP. The code is a modified version of guided diffusion and is trained on photographic-style images of people. It is a relatively smaller mode. Compared to DALL.E, GLID-3’s output is less capable of imaginative images for given prompts. Popular Posts Top Object Detection Algorithms Top Chart GPT Alternatives Top Ethical Hacking Courses Top AI Powered Tools for Stock Market Trading Top Library in CC for Machine Learning","excerpt":"DALL·E 2 was preferred over DALL·E 1 for its caption matching and photorealism.","categories":["AI Trends"],"tags":[],"author_name":"Avi Gopani","publish_date":"2022-04-11T12:00:00","publication_year":"2022","word_count":788,"keywords":["Go","machine learning","TPU","OpenAI","AI","neural network","RAG","GPT","object detection","R"],"extracted_tech_keywords":["AI","machine learning","neural network","OpenAI","RAG","object detection","TPU","R","Go","GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-4-dall-e-alternatives-text-to-image-generators\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":1623,"title":"What is LPWAN and why is it considered critical for IoT?","content":"With a large number of IoT deployments carried out on LPWAN, LoRa and SigFox have come to dominate businesses In case you missed the recently concluded Mobile World Congress, held in Barcelona, here’s a recap. Low Power Wide Area Network (LPWAN) in IoT was the hottest topic and a couple of significant announcements underscored how the unlicensed LPWAN has emerged as a major cost effective solution for IoT. LPWAN is threatening the licensed, cellular networks and is redefining IoT network service providers of tomorrow.  According to Kaustubha Parkhi, Principal Analyst at Insight Research, “The battle of LPWANs is fought between LoRa and SigFox, it is unlicensed and there to steal”. LPWAN — the new cornerstone of IoT connectivity Why is LPWAN key to IoT applications and how does it provide a low cost of entry to organizations? LPWAN fuels the connectivity of devices that need less bandwidth than what is usually required for most standard devices. The defining characteristic of LPWAN is not high speed connectivity, rather a scalable, robust “low power blanket” for smart devices such as sensors, lighting etc. Ideally used in industrial settings, LPWAN technology can connect devices up to a range of 15 miles and are now applied in healthcare; transportation; buildings and energy and utility sector. Battle of LPWANs is fought between LoRa, SigFox — it is unlicensed and there to steal Before we dive any further, let’s look at how LPWAN offers the ultimate tradeoff to telecom operators in being the next-gen IoT services provider in B2B and B2C segment. LPWAN is characterized by its low cost of connectivity, low latency and high battery life, enabling telcos a hassle free deployment even in remote, rural areas. There are two main technologies in play in the LPWAN market –LoRa and LoraWAN backed by Semtech which offers the full LPWAN technology stack. France based SigFox revealed the ace up its sleeve at the recently held MWC 2017 when it announced the partnership with Telefónica, Spanish broadband and telecommunications stalwart to “to integrate Sigfox’s low power connectivity into the operator’s managed connectivity platform”. According to reports, LoRa has emerged as the LPWAN technology of choice and leading the charge in wireless technology. Let’s look at the key features of one of the most popular LPWAN technologies — LoRa Wireless RF Technology: Long range: LoRa technology offers wide penetration capability for indoor and urban environments and has the capability to connect to sensors that are more than 15-30 miles away in rural areas as well Low power: One of the key defining features of LoRaWAN protocol was low power and its unprecedented battery lifetime of up to 20 years depending on the application Low cost: The technology doesn’t require heavy-duty hardware and infrastructure investments and pares down operating costs Open standard: The protocol ensures interoperability among IoT service providers and telecom operators for deployment and accelerating adoption According to ABI research, Low Power Wide Area Networks (LPWANs) is the fastest growing segment in connectivity and the technology will see an increased penetration by 2025. The research cites, “The rise of LPWANs will translate into one billion chipset shipments with the technology generating a total value of more than $2 billion in 2025”. Another report by Beecham Research investigates how Low Power Wireless Access Networks (LPWAN) will become the go-to connectivity technology in both the business and commercial context, thanks to the lower deployment cost as compared to cellular networks, cornering a share of 26% of IoT connectivity by 2020. However, the success of the technology lies in its use cases: Tata Communications launch MOVE: Tata Communications has made significant headway in building a seamlessly connected world by unveiling MOVE — a platform that enables companies to embed global connectivity in anything, enhance user experience, and generate new revenue streams. MOVE — global mobility platform, is underpinned by its global network of 900 mobile communications service providers across the globe and further bolstered by its recent investment in Teleena, an IoT connectivity specialist and mobile virtual network enabler. Tata Communications picked up a 35% stake, making it the single largest shareholder in the company. According to a statement, the first phase of MOVE roll-out encompasses global cellular IoT and SIM connectivity. Meanwhile, enterprise and mobile network operator capabilities are expected to follow later in the year. Tata Communications successfully tests LoRaWAN network, plans afoot for POCs: The forward looking Tata Communications was the first telecom operator in India to debut LoRaTM, ultra-low power connectivity solutions for IoT in India targeted at 400 million people across India. Of late, the leading communications provider launched its first application centre in collaboration with NASDAQ listed Semtech Corporation, provider of analog and mixed signal semiconductors. Post the successful trials of LoRaWAN network in three major cities across India — Mumbai, Delhi and Bengaluru, the telecom giants now plans to deploy 35 IoT proofs-of-concept (PoCs) based on LoRa Technology across these cities. The PoCs are aimed to cover a range of applications, such as energy management aimed at optimizing water, gas and electricity; smart building applications; sensors for air conditioner monitoring and safety deposit boxes. VS Shridhar, Senior VP and Head, Internet of Things, Tata Communications, revealed in a statement that Tata Communications had tested India’s first IoT network across major cities and the company is right on track in meeting its goal of connecting 200 million end devices by 2019. Tech Mahindra testing the market for end-to-end enablement of IoT solutions In an interview, Tech Mahindra representative revealed the company is carrying out extensive research in the LPWAN area and is reportedly supporting end-to-end enablement of several IoT solutions on LoRa. Tech Mahindra representative divulged in an earlier interview why LPWAN is the preferred solution for certain segments. According to Tech Mahindra, solutions requiring low data requirements such as smart lighting, smart parking are ideal for LPWA networks as opposed to solutions with high data requirement such as connected cars, for example. The B2B Opportunity — IoT network service provider of the future? As IT converges with telecom, businesses are being recasted and reshaped keeping in mind the pressing need of connectivity across sectors such as manufacturing, healthcare, transportation and energy & utility. Industry experts believe telcos will operate as channels and become the enablers of next-gen data connectivity. According to Parkhi, “IoT is just a precursor to the M2M technology and some of the earliest M2M applications were seen in smart grids, telemetry, and telematics.” The real question is how can telcos realize business value? What are the different revenue streams? “They can do so by integrating disparate networks and by internalizing use cases. IoT marries PAN, LAN and WAN and it widens the use case. It opens up the operator’s network. They are the drivers of bundled solutions, are access agnostic and are well-placed to generate value,” he shared at the recently concluded IoT show held at BIEC. A thorough partner ecosystem opens integrated solutions and gives touchpoints to interact with customers. A Mckinsey report cites emerging trends in which telcos can capture B2B segment – from offering unified communications across devices to unleashing mobile revolution through M2M applications, in smart metering and vehicle asset tracking among other areas.","excerpt":"In case you missed the recently concluded Mobile World Congress, held in Barcelona, here’s a recap. Low Power Wide Area Network (LPWAN) in IoT was the hottest topic and a couple of significant announcements underscored how the unlicensed LPWAN has emerged as a major cost effective solution for IoT. LPWAN is threatening the licensed, cellular […]","categories":["IT Services"],"tags":["hpe internet of things"],"author_name":"Richa Bhatia","publish_date":"2017-03-10T09:45:40","publication_year":"2017","word_count":1198,"keywords":["Go","programming_languages:R","AI","ML","Scala","Aim","ViT","programming_languages:Scala","GAN","hpe internet of things","R"],"extracted_tech_keywords":["AI","ML","Aim","R","Go","Scala","GAN","ViT","programming_languages:R","programming_languages:Scala"],"url":"https:\/\/analyticsindiamag.com\/it-services\/lpwan-considered-critical-iot\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":17453,"title":"Are We Really Moving Towards An AI Arms Race?","content":"Stephen Hawking, Elon Musk, Steve Wozniak and 150 others, recently signed onto a letter calling for a ban on the application of artificial intelligence (AI) to advanced weapons systems. Hawking says the potential threat from artificial intelligence isn’t just a far-off “Terminator”-style nightmare. He’s already pointing to signs that AI is going down the wrong track. “Governments seem to be engaged in an AI arms race, designing planes and weapons with intelligent technologies. The funding for projects directly beneficial to the human race, such as improved medical screening, seems a somewhat lower priority,” Hawking said. Does there really exist an AI arms race? Artificially intelligent systems continue to develop rapidly. Self-driving cars are being developed to dominate our roads; smartphones are beginning to respond to our queries and manage our schedules in real-time; robots are becoming better at getting up when they fall over. It is obvious that these technologies will only benefit humans going forward. But then all dystopian sci-fi stories begin like that. Having said that, there are two sides of the story. Assuming that Siri or Cortana would turn into murderous HAL from 2001: A Space Odyssey is an extreme but then supposing that AI being a threat to mankind is decades away and does not need intervention is also an extreme. A recent survey of leading AI researchers by TechEmergence listed various concerns about the security dangers of AI in a much more realistic manner. The survey suggested that in a 20-year timeframe, financial systems will see a meltdown as algorithms begin to interact unexpectedly. It also outlined the potential for AI to help malicious actors optimise biotechnological weapons. However, unlike previous autonomous weapons, such as landmines, which were indiscriminate in their targeting, smart AI weapons might limit the potential for deaths of soldiers and civilians alike. However, when ground breaking weapons technology is no longer confined to a few large militaries, non-proliferation efforts become much more difficult. The scariest aspect of the Cold War was the nuclear arms race. At its peak, the US and Russia held over 70,000 nuclear weapons and only a fraction of it, if used, could have killed every person on earth. As the race to create increasingly powerful artificial intelligence accelerates, and as governments continue to test AI capabilities in weapons, many experts have started to worry that an equally terrifying AI arms race may already be under way. For a matter of fact, at the end of 2015, the Pentagon requested $12-$15 billion for AI and autonomous weaponry for the 2017 budget, and the Deputy Defence Secretary at the time, Robert Work, admitted that he wanted “our competitors to wonder what’s behind the black curtain.” Work also said that the new technologies were “aimed at ensuring a continued military edge over China and Russia,” as quoted by Elon Musk’s Future of Life Foundation. The defence industry is gradually shifting towards integrating AI into the robots they build for military applications. For instance, many militaries globally have deployed unmanned autonomous vehicles for reconnaissance (such as detecting anti-ship mines in littoral waters), monitoring coastal waters for adversaries (like pirate ships), and precision air strikes on evasive targets. According to reports, the maker of the famous AK-47 rifle is building “a range of products based on neural networks,” including a “fully automated combat module” that can identify and shoot at its targets. It’s the latest illustration of how the U.S. and Russia differ as they develop artificial intelligence and robotics for warfare. Besides, China is also eyeing the use of a high level of artificial intelligence and automation for its next generation of cruise missiles, reports have suggested. It is not just the U.S., Russia and China that are developing its AI to be used in the defence, India too is not lagging behind. CAIR has been working on a project to develop a Multi Agent Robotics Framework (MARF), which will equip India’s armed forces with an array of robots. The AI-powered multi-layered architecture will be capable of providing a multitude of military applications and will enable collaboration amongst a team of various robots that the Indian Army has already built — Wheeled Robot with Passive Suspension, Snake Robot, Legged Robot, Wall-Climbing Robot, and Robot Sentry, among others. However, the robotics race right now is causing a massive brain drain from militaries into the commercial world. The most talented minds are now being drawn towards the private sector. Google’s AI budget would be the envy. Sooner or later, it will become trivially easy for organized criminal gangs or terrorist groups to construct devices such as assassination drones. Indeed, it is likely that given time, any AI capability can be weaponized. What are the concerns? Non-proliferation challenges: Prominent scholars including Stuart Russell have issued a call for action to avoid “potential pitfalls” in the development of AI that has been backed by leading technologists including Elon Musk, Steve Wozniak and Bill Gates. One high-profile pitfall could be “lethal autonomous weapons systems” (LAWS) or “killer robots”. The U.N. Human Rights Council has called for a moratorium on the further development of LAWS, while other activist groups and campaigns have advocated for a full ban, comparing it with chemical and biological weapons, which is unacceptable. Control: Is it man vs. machine or man with machine? Can AI when fully developed, be controlled? The reassurance is too early to come from creators of AI but again, thinking it is too early to contemplate is ignorance. Hacking: When developed, will AI systems not be vulnerable to hacking? While we cannot overlook the fact that the benefits of AI are much more than the potential risks involved, developers should work on systems that will reduce the risks involved. Targeting: Should it be compulsory for humans to always make the final decision with AI in the picture? Are we really ready for a fully autonomous system? Standards could be established that specify the required certainty and the specific scenarios when an AI would be allowed to proceed without human intervention. It may also be that an AI equipped with only non-lethal weapons can achieve nearly all the benefits with sufficiently reduced risk. Mistakes: In all probability, AI weapons will make mistakes. But humans most certainly will. A well designed and tested machine is almost always more reliable than humans. AI weapons systems can be held to strict standards for design and testing. Indeed, this should be a priority in the development of AI systems. Liability: Assuming there will be mistakes, the AI itself will not be liable. So who is? If the autonomous vehicles industry is any indication, companies designing AI may be willing to accept liability but their motivations may not align perfectly with those of society as a whole. The way forward: Many AI applications have huge potential to make human life better and holding back its development is undesirable and possibly unworkable. Moreover, if you take a look at the research being carried out on AI, you will realise that all projects are in their infancy and restricting their development is almost not required. But it also does speak the need for a more connected and coordinated multi-stakeholder effort to create norms, protocols, and mechanisms for the oversight and governance of AI. There is bare minimum support from global governments to fully ban the creation of killer robots. The simple reason being, there is still a long time before LAWS could be a reality. Take for example this, it would be impractical to prevent a terrorist group like ISIS from developing killer robots unless states can be assured of understanding the technology themselves first. The core idea behind regulation is to maximise benefits while simultaneously minimising risks involved. Above all, there is a need to recognise that humanity stands at a point, with innovations in AI outpacing evolution in norms, protocols and governance mechanisms. Regulation just has to make sure the outlandish, dystopian futures remain firmly in the realm of fiction.","excerpt":"Stephen Hawking, Elon Musk, Steve Wozniak and 150 others, recently signed onto a letter calling for a ban on the application of artificial intelligence (AI) to advanced weapons systems. Hawking says the potential threat from artificial intelligence isn’t just a far-off “Terminator”-style nightmare. He’s already pointing to signs that AI is going down the wrong […]","categories":["IT Services"],"tags":["AI (Artificial Intelligence)","AI Arms race"],"author_name":"Priya Singh","publish_date":"2017-09-04T11:48:46","publication_year":"2017","word_count":1325,"keywords":["Go","API","artificial intelligence","AWS","AI","AI Arms race","neural network","ML","Ray","Aim","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","ML","neural network","Aim","Ray","AWS","R","Go","API"],"url":"https:\/\/analyticsindiamag.com\/it-services\/really-moving-towards-ai-arms-race\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10080656,"title":"Top 9 Best Indian Open-source Projects","content":"The best way to learn development is through open-source projects and repositories. GitHub is the home for several such open-source projects and the community has consistently regarded India as one of the key contributors. The trend witnessed an upward momentum with the pandemic when Indian developers started contributing actively with a 5.2% increase. While there is no limit to the number of projects that have active contributors, engagements, and stars globally, there are various Made in India projects originating from Indian developers that have been trending on GitHub. Here’s a list of the top projects with big databases devised from India—some of them are highly accessible and functional! 1. Hoppscotch If you want to build an API from scratch, this open-source project offers the best ecosystem with high functionality and easy accessibility with the minimal design of the UI. The repository has over 48,000 stars and around 220 contributors. 2. Hasura\/graphql-engine Providing GraphQL or REST APIs along with built-in permission for your data, Hasura is a project for increasing API development ten-fold. With built-in filtering, pattern search, and bulk insert, users can make powerful queries and real time conversions. It is also deployable with one-click on Hasura Cloud. 3. Covid19india Built for keeping state-wise and district-wise records for COVID-19, a bunch of volunteers contributed to collect data that is available on this GitHub repository. Currently, it is publicly archived by the owner and is available in read-only format. Read: Top 10 Indian Government Datasets 4. GeekyAnts\/NativeBase Developed with mobile-first focus, NativeBase is a library for accessible components for Web and React Native for building consistent UI across Web, iOS, and Android. The newer version 3.0 comes with ARIA integration and support for nearly 40 components that are consistent on all the platforms. 5. Chatwoot Among the long running Indian projects on GitHub, Chatwoot is an open-source customer engagement suite. It is said to be an alternative to Zendesk, Intercom, or Salesforce as it lets users manage customer data and interact with them irrespective of their medium of communication. It supports conversation on Facebook, Instagram, Twitter, Whatsapp, Telegram, Email, SMS, and API Channel. 6. Spacecloud-io A Kubernetes-based serverless platform, SpaceCloud provides real time APIs on any database helping one to build modern applications without learning backend coding. It is supported on MongoDB, PostgreSQL, MySQL, and SQL Server. It can replace backend php, nodejs, java code and help one write code over an external API, directly consumable with frontend. Read: 12 Most Popular Open-Source Projects on GitHub 7. Frappe\/erpnext A useful repository for enterprise resource planning, erpnext includes several areas for managing businesses like accounting, warehouse management, sales, purchase, CRM, HRMS, manufacturing, among many others. It is a full stack web app framework that is built with Python and Javascript, with MariaDB as a database. 8. Kubeshop\/botkube Botkube allows users to monitor Kuberneters cluster, debug critical deployments, and also provides recommendations for general practices by checking on the Kubernetes resources. It can be integrated with multiple communication platforms like Slack, Discord, and Mattermost. It can be deployed on any cluster like K3d, kubernetes on cloud, or bare-metal one. 9. Appsmithorg\/appsmith With more than 22,000 stars, appsmith repository is a low code project to build admin panels, dashboards, and internal tools that can be integrated with more than 15 databases and any API. One can build apps in four simple steps with this repository by building their UI, connecting the data, connecting the data to the UI, and deploying the app.","excerpt":"Check out these Indian open source repositories that have been trending around the world.","categories":["IT Services"],"tags":["GitHub","Kubernetes"],"author_name":"Mohit Pandey","publish_date":"2022-11-24T14:00:00","publication_year":"2022","word_count":578,"keywords":["PostgreSQL","Go","AI","MongoDB","R","serverless","Python","SQL","JavaScript","GitHub","Kubernetes","kubernetes"],"extracted_tech_keywords":["AI","kubernetes","serverless","MongoDB","PostgreSQL","Python","R","SQL","JavaScript","Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/top-9-indian-open-source-projects-in-2022\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":36980,"title":"5 Ways Data Analytics Is Transforming The Aviation Industry","content":"Machine learning and analytics have touched almost all the fields around the globe including the aviation industry. With the growth of data, the use of analytics in the airline industry is the next big wave. Now, big data analytics and predictive models are being used for augmenting opportunities in the industry. In this article, we list down five fascinating ways that the aviation industries are using analytics. 1| Increase In Airline Revenue Analytics and text mining techniques help the industry to understand the sentiments of the customers and other maintenance issues. Analysis of ticket booking can help the industry to target the customers with personalised offers while optimising the price in real-time using predictive analysis techniques. By gathering meaningful data, the aviation industry can fetch more bookings in the given time span. Artificial intelligence and analytics will provide the aviation industry to operate at an ideal cost to profit ratios through targeted sales. 2| Smart Maintenance Baggage will no longer be the burden for the passengers and one can travel freely without any hassle. The Radio-Frequency Identification helps from mishandling the baggage. Predictive analysis helps in improving the predictability of fleet reliability. The airport traffic is increasing every day and as a result, it will conclude as a threat to the aviation industry. But with the help of analytics and big data, parameters such as runway bandwidth, flight routes, types of aircraft, etc can be implied to identify the patterns as well as comparing them. The industry is working on optimising the use of airspace. 3| Cost Reduction The introduction of analytics into the aviation industry will benefit into cost reductions in several ways such as every year lots of baggage get lost which have to be repaid by the industry, thus the airlines depend on the real-time baggage tracking data helps avoid losing or any damage or delaying bags. The data of real-time consumption of fuel is collected and analysing it can be an efficient use of fuel as well as reducing the cost in investing it more than it needed. 4| Customer Satisfaction Customer Satisfaction is one of the main motives in any organisation. The analytics and other emerging technologies have spread out their hands to help the organisation with such motives. The aviation industry trying all the possible ways in the rat race by soaring in the emerging technology. With the help of predictive analysis, sentiment analysis, etc. the industry provides its customers to keep updated in real-time, promoting offers according to their choice and experience, fulfilling needs regarding their habits, etc. 5| Digital Transformation Big Data and analysis is transforming and digitising the commercial aviation industry in order to deliver high standard services to the passengers. Passenger Technology Solutions has launched to provide the perfect platform for the custom-made technology suppliers to showcase their products and services to airlines, airports, etc.  from around the globe in order to offer the passengers a more connected travel experience. The emerging technologies are lifting up the aviation industry into new heights by helping them in every possible way to fulfil the customer’s needs, real-time performance dashboards, predictive maintenance, etc. Use cases With the help of Big Data and analytics, the Miami International airport has launched an app where the passengers get information and support relevant to their journey in order to improve the passenger’s experience in real time. According to this report, United Airlines tied up with virtual AI assistant, Amazon Alexa which manages the entire flight experience for their passengers such as providing ongoing updates, check-in process, etc. Outlook Artificial Intelligence and Big Data analytics will transform the experience of customers and other engagements at airports. According to SITA’s Air Transport IT Trends Insights 2017 reports, most of the airlines which were part of the study are planning significant initiatives of artificial intelligence technology within the year 2020 which include real-time predictive pricing offers, air travel experience to chatbots which recommends upgrades based on the customer’s experience and habits, etc.","excerpt":"Machine learning and analytics have touched almost all the fields around the globe including the aviation industry. With the growth of data, the use of analytics in the airline industry is the next big wave. Now, big data analytics and predictive models are being used for augmenting opportunities in the industry. In this article, we […]","categories":["AI Trends"],"tags":["aviation industry","Data Analytics"],"author_name":"Ambika Choudhury","publish_date":"2019-03-28T09:46:11","publication_year":"2019","word_count":662,"keywords":["big data","Go","machine learning","artificial intelligence","AI","aviation industry","chatbots","sentiment analysis","Git","analytics","Data Analytics","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","analytics","chatbots","sentiment analysis","R","Go","Git","big data"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/5-ways-data-analytics-is-transforming-the-aviation-industry\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":16304,"title":"Microsoft’s new AI-based app will help visually impaired people &#8216;see&#8217; what’s around them","content":"In a move that could potentially expand the horizons of visually-impaired mobile phone users, tech giant Microsoft has released an app for iPhone that uses Artificial Intelligence (AI) to help them “see” the world. The Microsoft iPhone app was released earlier this week in the US, Canada, India, New Zealand and Hong Kong. Named ‘Seeing AI’, the application uses the iPhone camera to detect, read and describe the things around the visually-impaired person. For example, in one of the demo videos released by Microsoft, the Seeing AI correctly identifies a woman as, “28-yer-old woman, looking happy”. Yes, you read that right. The app can detect human emotions too. The Seeing AI joins a small but steadily rising number of apps designed to help the visually impaired. While many can only tell the colour, or read bar codes on products, other apps such as Be My Eyes connects blind persons to other volunteers in the area. A Seattle Times report says that the Seeing AI project was demonstrated in March 2016 by Microsoft software engineer Saqib Shaikh. The app relies on the iPhone camera, backed by Microsoft’s machine-learning and image-recognition algorithms, the Seattle Times reported. As of now, Microsoft hasn’t announced whether a version of Seeing AI would be released for Android phones.","excerpt":"In a move that could potentially expand the horizons of visually-impaired mobile phone users, tech giant Microsoft has released an app for iPhone that uses Artificial Intelligence (AI) to help them “see” the world. The Microsoft iPhone app was released earlier this week in the US, Canada, India, New Zealand and Hong Kong. Named ‘Seeing […]","categories":["AI News"],"tags":["instagram"],"author_name":"Prajakta Hebbar","publish_date":"2017-07-14T05:53:50","publication_year":"2017","word_count":212,"keywords":["Go","artificial intelligence","programming_languages:R","AI","programming_languages:Go","instagram","R"],"extracted_tech_keywords":["AI","artificial intelligence","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/microsofts-new-ai-based-app-will-help-visually-impaired-people-see-whats-around\/","complexity_score":3,"technical_depth":6,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10164299,"title":"Arize AI Secures $70Mn to Scale its Unified AI Observability Platform","content":"Arize AI, a company focusing on AI observability and LLM evaluation, has announced a remarkable $70 million Series C funding round. The company claims this investment marks the largest ever in the realm of AI observability. The investment was spearheaded by Adams Street Partners, with contributions from other investors, including M12 (Microsoft’s venture fund), Sinewave Ventures, OMERS Ventures, Datadog, PagerDuty, Industry Ventures, and Archerman Capital. Arize AI’s recent research initiative, OpenEvals, has revealed that LLMs often struggle to accurately evaluate synthetic datasets—data generated by other AI models—compared to non-synthetic data. Furthermore, the results from the research state that LLMs are still unpredictable, difficult to troubleshoot, and prone to failure. Hence, with the company’s offering, organisations can course-correct AI systems before they do any damage in a real-world setting. Jason Lopatecki, CEO and co-founder of Arize AI, emphasised the importance of reliability in AI systems: “Building AI is easy. Making it work in the real world is the hard part. Enterprises can’t afford to deploy unreliable AI.” He stressed that engineering teams require robust infrastructure to test and troubleshoot their models effectively before they impact customers. The company mentions that its platform has helped enterprises like Uber, Duolingo, TripAdvisor, and more. And, they also highlight that their open source offering, Arize Phoenix, is the most widely adopted AI observability and evaluation library for development. In a statement, Arize says that the partnership with Microsoft is also set to deepen, with M12’s investment reinforcing their collaboration, which will result in better integrations with the Azure AI platform. Aparna Dhinakaran, the chief product officer and co-founder of Arize, shared a bit about their future vision, “As AI research and real-world applications accelerate, Arize will continue to pioneer new tools, like our recent first-to-market launch of audio evaluation for voice assistants, to help engineers working on these systems better evaluation, debug, and improve what they build.” With more companies focusing on solving issues related to AI, such as the interesting concept of using AI to fix AI, the experience for businesses, clients, and customers should improve.","excerpt":"Arize AI aims to use the funding to fix issues with LLMs and AI Agents in the real world.","categories":["AI News"],"tags":["ai funding"],"author_name":"Ankush Das","publish_date":"2025-02-21T12:20:16","publication_year":"2025","word_count":342,"keywords":["Go","API","funding","synthetic data","AI","R","Aim","ai funding","GAN","AI research","Azure"],"extracted_tech_keywords":["AI","Aim","Azure","R","Go","API","GAN","synthetic data","funding","AI research"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/arize-ai-secures-70mn-to-scale-its-unified-ai-observability-platform\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10042054,"title":"Is India’s 67th Rank In Coursera Global Skills Report Justified?","content":"Recently, Coursera released its annual Global Skills Report 2021 to offer a sneak peek into the ‘state of skills’ in over 100 countries. India, as per the report, ranks 67th globally with 38% proficiency and 16th in Asia below Singapore and Japan. The country stood at 55th position in business and at 66th position in both technology and data science. The report claims to benchmark skill proficiencies in business, technology, and data science across 77 million learners, 4,000 campuses, 2,000 businesses, and 100 governments using the Coursera platform. Let’s look at Coursera’s methodology and parameters to come up with this ranking. Coursera has set four skill levels: Cutting Edge — 76th percentile or aboveCompetitive — 51st-75th percentileEmerging — 26th-50th percentileLagging — 25th percentile or below In the field of data sciences, Coursera looked at the following subjects: Data AnalysisData ManagementData Visualisation Machine LearningMathematicsProbability & StatisticsStatistical Programming The report found India as: Lagging in Data Analysis and Statistical ProgrammingEmerging in Data Management, Data Visualization and Probability & StatisticsCompetitive in Machine Learning and Mathematics The rankings attributed to India in data sciences seems inaccurate and reflects a “one size fits all” methodology. India: More than meets the eye “The skill trends and proficiency analyses in this report represent a view of the world through the lens of Coursera,” the report said. India has multiple online platforms–not just Coursera– and offline campuses to equip students and working professionals with digital skills. Colleges and universities, including IITs, IIMs, and NIIT are making Indian students future ready. Online programmes from AnalytixLabs, Ivy ProSchool, WileyNXT, Edvancer are also gaining traction. Mayank Kumar, Co-founder & MD, upGrad, said: At upGrad, we have the largest alumni base for Data Science learners in India, which includes over 22,000 individuals. Owing to evolving industry requirements, the domain has also caught the attention of women professionals. As per our Q4 2021 (Jan-Feb-Mar) data, we have 2\/5th of women learners enrolled for courses like Data Science, Machine Learning & Artificial Intelligence.” “Talking from an industry standpoint, we also have a skilled workforce who are indulging in cutting-edge projects on a day-to-day basis and delivering quality outputs. This indeed is a reason why we are seeing global recruiters hire from India to be able to meet their growing demands. All these facts bear testimony to the quality work India is producing, and it wouldn’t be correct to say that we lag in data skills and are facing significant challenges across the Data Science domain,” he added. In terms of the business, the technology and the data science skills, India (67) is positioned below Pakistan (65th), Bangladesh (64th), Nepal (62nd), and Sri Lanka (40th). It begs the question over the parameters considered to arrive at the rankings. “Internet access is correlated with skill proficiency on Coursera,” said the report. Internet penetration in India stands at 34.4%, while that of Pakistan is 32.4% and Sri Lanka is 33.5%, as per a Statista report. Despite being the second most populated country, India’s internet connectivity is on par with their Asian counterparts. Coursera tapped into the data from the platform and research from Q1 2020 to Q1 2021 to come up with the report. As per the Global Innovation Index report 2020, published in September 2020, India has risen to become the world’s third most innovative lower middle-income economy. India is in the top 15 in ICT (Information and Communication Technology), government online services, services exports, science and engineering graduates, and R&D-intensive global firms indexes. The huge knowledge capital, the thriving startup ecosystem, and the excellent work done by public and private research groups are all contributing to the index rankings’ steady development. The global standing of the Indian Analytics market, the growth in salaries across almost all parameters, the maturing of the analytics market in terms of experienced hiring and salaries offered, the rising gender diversity in the Indian analytics function, and the $762 million investment in Indian AI and analytics startups are pointing to India’s burgeoning digital prowess. The Indian analytics function’s overall revenues have increased to $35.9 billion–19.5 percent year on year growth. Though skill gaps are a reality, 67th rank seems unjustified. The report should have accounted for India’s young demographic, thriving talent pool, and the quality of the talents we produce. Inshort, the Coursera report is not exactly an ideal credentialing of a diverse country like India.","excerpt":"Recently, Coursera released its annual Global Skills Report 2021 to offer a sneak peek into the ‘state of skills’ in over 100 countries. India, as per the report, ranks 67th globally with 38% proficiency and 16th in Asia below Singapore and Japan. The country stood at 55th position in business and at 66th position in […]","categories":["IT Services"],"tags":[],"author_name":"kumar Gandharv","publish_date":"2021-06-18T18:16:42","publication_year":"2021","word_count":725,"keywords":["data science","Go","machine learning","artificial intelligence","TPU","AI","Git","Aim","analytics","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","data science","analytics","Aim","TPU","R","Go","Git"],"url":"https:\/\/analyticsindiamag.com\/it-services\/is-indias-67th-rank-in-coursera-global-skills-report-justified\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10093756,"title":"Meta Finally Cracks the LLM Code Without RLHF","content":"Meta is back in the LLM business. Google had claimed that neither they nor does OpenAI have a moat when it comes to AI, but also acknowledged that open source will pave the way for the future of generative AI. The interesting thing about this is Meta is being touted as the leader for open-source generative AI by developers because of its open-source LLM, LLaMa. Meta has released another open-source model called MMS: Massively Multilingual Speech. The model is made for speech-to-text and text-to-speech in 1100 languages with the ability to recognise 4,000 spoken languages. French computer scientist Yann LeCun said that the model has half the error rate of OpenAI’s Whisper. This is a huge advancement when it comes to multilingual speech AI technology. https:\/\/twitter.com\/BoredGeekz\/status\/1660734071926935566 Besides this, yesterday, LeCun shared Meta AI’s latest breakthrough in LLMs. LIMA, made in collaboration with Carnegie Mellon University, University of Southern California, and Tel Aviv University, is a 65 billion parameter model built with LLaMA and fine-tuned with a standard supervised loss with only 1,000 carefully curated prompts and responses. The most interesting part is that it does not require RLHF (reinforcement learning with human feedback), which is being implemented within ChatGPT and Bard, and told as the differentiating and the most important factor for its working. There is a possibility that this model might be implementing self-supervised learning, which LeCun has been the biggest proponent of all this while. When it comes to performance, LeCun claims that this LIMA is on par with GPT-4 and Bard. People have been critical of this. And though excited about the development, the community is also calling out that the benchmarks to compare it with GPT-4 and Bard are not accurate. The paper is still under review. Recovering Doomer LeCun said in an interview, “The platform that will win will be an open one. Progress is faster when it is open.” This might seem a little contrary to the stance that the company took during this AI chatbot race. Before releasing LLaMa in February, the company was keeping itself away from LLMs. LeCun used to call these technologies “great but nothing revolutionary” and thus was also sceptical about releasing such technology to the public. A lot of it is because the company had actually ventured into the field before any of the competition. BlenderBot and Galactica, two separate chatbots were released by the company but had to be shut down when they turned into hallucinating disasters. Probably, a lot of cynicism towards this LLM-based technology was just trauma from the past. Just before the release of LLaMa, Mark Zuckerberg had made a Facebook post talking about reorganising its teams’ structure and focusing on generative AI technology. But, instead of building a chatbot to compete with Microsoft, OpenAI, and Google, the company took an interesting turn, and gave away its most precious asset to open source. The still-under-review paper of LIMA shows the possibility. There is a chance that the company might venture into the field of chatbots, again. The company had been reluctant to release its models to the public earlier because of their capability of generating misinformation and toxic content, as explained by LeCun. Quote from the article that is accurate:\"Google, Meta and other tech giants have been reluctant to release generative technologies to the wider public because these systems often produce toxic content, including misinformation, hate speech […]\" https:\/\/t.co\/vEbAMAi6TC— Yann LeCun (@ylecun) January 8, 2023 Now the company has taken a different approach and released its model to the general public that also includes weights, making it a lot easier to make chatbots for developers. Seems like Meta wants a thousand chatbots to exist, but won’t make one. Is LLaMa the moat for Meta? The tables have turned. Google and OpenAI, the companies which were in a rush to make their chatbots public, even though keeping the technology behind closed doors for most of the cases, are now considering regulations around the technology. Meanwhile, Meta, which was always not so sure about the technology, is now making their most capable models open source. Seems like Meta is banking on the open-source community instead of trying to build a chatbot like Google or OpenAI. For Meta, the community that is building models with LLaMa is the moat. Just like developers and creators flock around StabilityAI for its open source software, people are building model after model over LLaMa, that are outperforming GPT-4 or Bard on less computation requirements as well. Google’s leaked document said, “The modern internet runs on open source for a reason and we should not be expected to catch up with it.” Even though Google, Microsoft, and OpenAI’s heads were called to the White House to discuss regulations around AI, Meta was left out, citing the reason that the meeting “was focused on companies currently leading the space”. Even though Meta does not have a customer-facing AI product, they are not too behind on it as well. According to reports, Zuckerberg, during April’s earnings call, mentioned ‘AI’ at least 27 times, which might be less compared to Sundar Pichai chanting it 140 times at Google I\/O, but still the company has been integrating generative AI into a lot of its apps, similar to Google. Moreover, the FAIR department of the company has been publishing papers at a rate faster than OpenAI or Google, both of which are now more concerned about protecting their technologies. It might be too early to tell if Meta taking sides with the open-source community would actually make a mark for them in the AI arms race. But there is no doubt that LLaMa has indeed made an impact on Google and OpenAI. Sam Altman has already discussed imposing restrictions around the technology with the US Senate.","excerpt":"So far, Meta has been reluctant with LLMs and chatbots, but is now rushing into the open-source without reinforcement learning with human feedback (RLHF)","categories":["AI Features"],"tags":["Meta AI","Open Source AI","Yann LeCun"],"author_name":"Mohit Pandey","publish_date":"2023-05-23T16:30:00","publication_year":"2023","word_count":959,"keywords":["Go","ChatGPT","Meta AI","Yann LeCun","RLHF","OpenAI","AI","chatbots","Open Source AI","Aim","generative AI","R"],"extracted_tech_keywords":["AI","generative AI","ChatGPT","OpenAI","Meta AI","Aim","RLHF","chatbots","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/meta-finally-cracks-the-llm-code-without-rlhf\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":32771,"title":"Kerala Is Poised To Be The Next AI Hub With Booming Deep Tech Startup Ecosystem","content":"The startup ecosystem in Kerala has taken off in terms of the latest developments and incubations across the state. With Kerala Startup Mission (KSUM) playing a key role in identifying young talent and nurturing their entrepreneurial dreams, the startup ecosystem has received a major impetus with Kochi emerging as a hub for deep tech companies. One of the primary reason for its success in developing a growing startup ecosystem is its interlinking of colleges, incubators, government institutes and startups. By focussing on developing entrepreneurial skills at the grassroots level and fostering a DIY (do-it-yourself) culture in schools, in addition to mentoring through various innovation and entrepreneurship development cell in colleges, the state government has successfully brought all the stakeholders under one roof. Added to this, is the policy changes that it adopted leading to a massive boost for the startup growth in the state. In 2014, Kerala became the first ever state to formulate a draft policy in the development of the startup ecosystem. With the passing of Kerala Technology Startup Policy 2014, the government hoped to promote a scientific and technological environment for the youths to follow their dream within the state. This was followed up with measures to extract more external investments for scaling up the ventures quickly. With KSUM adopting fund of fund strategy, it was able to divert more money to home-grown start-ups. “One of our core philosophies is focusing on revenue generation over attracting external investments, and this is done by encouraging startups in our network to address the problems that have blue ocean opportunity,” M Sivasankaran, Secretary, Electronics and Information Technology of the state said in the recently brought out Kerala Startup Ecosystem Report 2018 According to the study, institutional funding for start-ups in the state was $9.5 Mn in 2014, this was increased to $19.6 million in 2015. From 2016, the fund flow was $15.72 million from the first 6 months. In 2018 at the end of the third quarter, funding observed till September stood at $38 million. In September 2018, to aid the startups’ growth the government increased the cap on direct purchase of software products, services and mobile app from the ventures cleared by the Kerala Startup Mission (KSUM) from $6865 to $27,463. Speaking on the matter,  Saji Gopinath, Chief Executive Officer, Kerala Startup Mission said “The $6865  ceiling seems to be a challenge in enhancing the scope of innovative products in government departments, this government order will help startups to work closely with government and come up with innovative solutions which are useful for the departments. The recent floods also proved the startups’ capability of supporting the government in disaster management.” Due to the government’s and other agencies relentless push to make Kerala a more attractive destination for startups, it was named as one of the ‘Top Performer’ in the States’ Startup Ranking 2018 by the Department of Industrial Policy and Promotion (DIPP). Though it came second in line, as the ‘Best Performer’ was named as Gujarat for their vibrant start-up environment. In the ranking, DIPP has, however, listed a few steps that states like Kerala could adopt to improve their existing startup ecosystem. Taking the suggestions into consideration, the state government invited startups to set up an IT project for its various departments, boards and local-self governing bodies which were previously outsourced to external agencies. As per the recent report, as many as 40 start-ups from across the state have already begun their work for the departments. In this article, we explore how Kerala has emerged as one of the destinations for deep tech startups. Though it lags behind Bengaluru and Hyderabad in terms of adoption and market penetration, there is a growing percentage of deep technology startups setting up base in the state. Though IT startups constitute 35% of the total the ecosystem, only a handful has made a debut in deep technology such as AI, ML and Deep Learning. Though small in number, some of these startups have already created a buzz with their unique initiative. Here are our top three picks. Ajna: This Trivandrum-based startup leverages deep learning, blockchain technology and autonomous drones to predict illegal logging. AI is used to detect the change in tree covers based on the images and videos recorded by the drones. The AI system can then detect the change canopies of an individual tree. RecipeBook: This Kochi-based startup uses AI, image recognition and natural language processing to discover a recipe on the internet. It was started in 2015, by Ajmal Azeez, Aswin R, Kishan Pankaj, and Balram P. It was one of the three AI startups who represented India in the 2016 NVIDIA GPU Conference held in San Jose. Tranzmeo: One of the fastest growing companies in Kerala, this is also the first ever AI startup to receive funding from India-based Fortune 500 Company. Started by  Safil Sunny in 2017, the start-up’s T-Connect OneView is an anomaly forensics application which uses deep learning and AI technology to predict anomalies.","excerpt":"The startup ecosystem in Kerala has taken off in terms of the latest developments and incubations across the state. With Kerala Startup Mission (KSUM) playing a key role in identifying young talent and nurturing their entrepreneurial dreams, the startup ecosystem has received a major impetus with Kochi emerging as a hub for deep tech companies. […]","categories":["AI Startups"],"tags":["AI in governance","Startups"],"author_name":"Akshaya Asokan","publish_date":"2019-01-03T11:56:07","publication_year":"2019","word_count":829,"keywords":["Go","startup","AI","innovation","ML","image recognition","RAG","deep learning","ViT","Startups","R","AI in governance"],"extracted_tech_keywords":["AI","ML","deep learning","RAG","image recognition","R","Go","ViT","innovation","startup"],"url":"https:\/\/analyticsindiamag.com\/ai-startups\/kerala-is-poised-to-be-the-next-ai-hub-with-booming-deep-tech-startup-ecosystem\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":22420,"title":"Is YouTube’s Recommendation Algorithm Really Working?","content":"YouTube remains one of the most watched websites for video content. Started in 2005, it offers videos which cover an array of topics right from education to famous music videos by artists and celebrities and has even spread out to gaming content (later in 2006, it was acquired by tech giant Google). Even though most videos are available for free, it also has paid content. The popularity has risen so much that YouTube and videos are almost synonymous terms. In 2017, it was estimated that one billion hours of video content was watched online, everyday on YouTube. How does the algorithm work? The recommendation algorithm of YouTube, that plays an important part in feeding relevant content to its users was the brainchild project of The Google Brain team, which is currently driven by their deep learning systems. Basically, the algorithm consists of two neural networks. The first neural network works on candidate generation which means the network utilises the users’ watch history and applies the concept of collaborative filtering technique to suggest similar videos based on watch history. The best algorithms usually are obtained by A\/B testing, which uses two variables– for example, two versions of the same web-page, to determine which one fetches more views. This way, the user experience is improved. The second neural network uses logistic regression to improvise and prioritise those similar videos that fall in the users’ watch history pattern. Improvements are further done by using A\/B Testing. But, in this case, the duration of a user watching video(s) and the response (like, dislike and comment) are considered. Positive responses or likes by the user fare a small amount to that of the total responses including dislikes and comments. Therefore, weights are assigned to these responses to help assess videos using logistic regression, and predicting what the user wants. This method is called Ranking. The ones which align to the regression analysis become the ‘viral videos’ or ‘trending videos’. Pitfalls and Improvements One performance improvement study by the Google research team states that “YouTube represents one of the largest scale and most sophisticated industrial recommendation systems in existence”. Although the algorithms are powerful and match relevant content according to the users’ interests, sometimes they fail to distinguish between the good and the bad. For example, YouTube suggests controversial content regardless of any thought. For example, recently it recommended the infamous vlogger Logan Paul’s return on YouTube. To make things worse, it even notified the users. This leads to a chain reaction with other users who don’t know him, often viewing the content without their knowledge. Similarly, another incident occurred with YouTube suggesting conspiracy videos of Hillary Clinton during the US Presidential campaign in 2016. These incidents are just on the surface level. Some videos on the site even consolidate lame content such as the Earth being flat or vaccines lead to autism. On the other hand, YouTube does remove inappropriate content only after it has been reported in large numbers. This will only be a temporary solution. The fact of the matter is, the data surrounding these algorithms comes from the vast user base YouTube has garnered. Not all users share the same opinions for videos of similar category. This criticality should be addressed by the recommendation system used in YouTube. In recent developments, it is looking to sort its ‘trending videos’ over the massive content catered to its viewers by restructuring the recommendation algorithm. Recommendation systems best work with past user data such as watch history. When it comes to its implementation in real-time, deep neural networks for these systems require continuous and regulated data. To achieve this, YouTube has hired people to manually review content which are malign and inappropriate, with the help of machine learning. The manual video moderation campaign by YouTube has been very aggressive, with the team nearly stripping any offensive content it encounters on the website. YouTube has daunting challenges to resolve, such as massive video content uploaded on their website coupled with the influx of duplicated and mis-titled videos. Along with this, the metrics  for the recommendation algorithm might seem far-fetched to serve user interests’ on broader terms. Conclusion : YouTube shares the biggest chunk in the video content platform category when it comes to viewership. It needs to pick up pace to get acquainted with user patterns for machine learning, since it is ever growing. Ultimately, it is not just addressing the content, but how intelligent and smart the algorithms are to take on dire consequences quickly.","excerpt":"YouTube remains one of the most watched websites for video content. Started in 2005, it offers videos which cover an array of topics right from education to famous music videos by artists and celebrities and has even spread out to gaming content (later in 2006, it was acquired by tech giant Google). Even though most […]","categories":["IT Services"],"tags":[],"author_name":"Abhishek Sharma","publish_date":"2018-03-08T11:36:03","publication_year":"2018","word_count":750,"keywords":["Go","machine learning","programming_languages:R","AI","neural network","recommendation systems","Ray","deep learning","GAN","R"],"extracted_tech_keywords":["AI","machine learning","deep learning","neural network","Ray","recommendation systems","R","Go","GAN","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/it-services\/is-youtubes-recommendation-algorithm-really-working\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":33789,"title":"Facebook To Set Up Institute For AI Ethics With Initial Investment Of $7.5 Million","content":"Social media giant Facebook on Sunday announced that they are creating an independent Institute for Ethics in Artificial Intelligence with an initial grant of $7.5 million over a period of five years. In collaboration with the Technical University of Munich (TUM) in Germany, the institute will help advance the growing field of ethical research on new technology and will explore fundamental issues affecting the use and impact of AI. “Artificial intelligence offers an immense opportunity to benefit people and communities around the world. Academics, industry stakeholders and developers driving these advances need to do so responsibly and ensure AI treats people fairly, protects their safety, respects their privacy and works for them,” said Facebook in an official release. “Core questions arise around trust, privacy, fairness or inclusion, for example, when people leave data traces on the internet or receive certain information by way of algorithms,” said Professor Christoph Lütge, who has worked extensively on the ethics of digital technologies at TUM From past two years, Facebook has been facing criticism for failing to stop the spread of fake news, terrorist propaganda and hate speech as well as abusing users’ privacy be it the ethnic cleansing in Myanmar, to Cambridge Analytica seeking to manipulate elections, to Google building a secret censored search engine for the Chinese, to anger over Microsoft contracts with US immigration department. According to Joaquin Quinonero Candela, Facebook’s Director of applied machine learning ensuring responsible and thoughtful use of AI is foundational to everything we do from the data labels we use to the individual algorithms we build, to the systems they are a part of. He further adds that Facebook is also working on new tools, including one called Fairness Flow, that can help those working on machine learning systems evaluate them for hidden bias. Earlier, Facebook has joined the Partnership for AI, a consortium that includes several companies working at the cutting edge of AI technology and that aims to address ethical and safety concerns. Artificial Intelligence will go through a course in ethics soon!https:\/\/t.co\/k0u3irB3eh — News18.com (@news18dotcom) January 21, 2019 Key Points To Focus – TUM Institute Of AI Ethics The institute will conduct independent, evidence-based research to provide insight and guidance for society, industry, legislators and decision-makers across the private and public sectors. It will explore the ethical issues of AI and develop ethical guidelines for the responsible use of the technology in society and the economy and address issues that affect the use and impact of AI, such as safety, privacy, fairness and transparency. Institute will seek to contribute to the broader conversation surrounding ethics and AI, pursuing research that can help provide tangible frameworks, methodologies and algorithmic approaches to advising AI developers and practitioners on ethical best practices to address real-world challenges. The Institute will also pursue opportunities to publish research and work with other experts in the field, organize conferences, symposia, and workshops and launch educational activities with other leading institutions in common areas of interest. Facebook will also share its insights, tools, and industry expertise related to issues such as addressing algorithmic bias, in order to help Institute researchers, focus on real-world problems that manifest at scale. While Facebook has provided initial funding, the Institute will explore other funding opportunities from additional partners and agencies. The Institute will also benefit from Germany’s position at the forefront of the conversation surrounding ethical frameworks for AI and its work with European institutions on these issues.","excerpt":"Social media giant Facebook on Sunday announced that they are creating an independent Institute for Ethics in Artificial Intelligence with an initial grant of $7.5 million over a period of five years. In collaboration with the Technical University of Munich (TUM) in Germany, the institute will help advance the growing field of ethical research on […]","categories":["AI News"],"tags":["AI (Artificial Intelligence)","Facebook","germany"],"author_name":"Martin F.R.","publish_date":"2019-01-21T07:29:20","publication_year":"2019","word_count":573,"keywords":["Go","artificial intelligence","machine learning","AI","Git","GAN","germany","Aim","ViT","Rust","Facebook","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","Aim","R","Go","Rust","Git","GAN","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/facebook-set-up-institute-ai-ethics\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10147682,"title":"6 Must-Attend Conferences for Developers by AIM in 2025","content":"India is a bustling hub for AI, data science, and technology enthusiasts. And with its base in Bengaluru, AIM Media House is at the forefront of this tech transformation. With a mission to empower and connect professionals and organisations through knowledge and innovation, AIM organises some of the industry’s most influential conferences. From fostering diversity and inclusion in tech to exploring the latest advancements in generative AI, AIM’s conferences cater to a wide range of interests and expertise. Whether you’re a data engineer, AI startup founder, developer, or corporate leader, these events offer unparalleled opportunities to learn, network, and stay ahead in the fast-paced tech landscape. Here’s a quick look at AIM’s upcoming conferences next year, why you should attend them, and how they’re shaping the future of AI and technology in India. MLDS 2025 MLDS is a haven for developers and data scientists looking to stay ahead in the ever-evolving world of artificial intelligence. Focused on the latest innovations in generative AI and software development, this three-day event offers something for everyone, from keynotes to hands-on workshops. With paper presentations and tech talks across three tracks, MLDS ensures you leave with actionable insights, whether you’re a beginner or an experienced AI practitioner. The event’s emphasis on practical applications and industry case studies makes it a must-attend for professionals looking to integrate AI into their workflows. Moreover, the networking opportunities with AI leaders and like-minded peers are unparalleled. Register here Dates: February 5-7, 2025 Venue: NIMHANS Convention Center, Bengaluru, India The Rising 2025 As one of India’s most impactful conferences on diversity and inclusion in tech, The Rising addresses some of the most pressing issues in today’s workplaces. From practical strategies for fostering equity to inspiring success stories, this summit provides a deep dive into creating a culture of belonging. Leaders from top companies share their approaches to tackling DEI challenges, making it a rich learning ground for both individuals and organisations. Whether you’re an HR professional, tech leader, or someone interested in creating equitable spaces, The Rising is the perfect platform to gain insights and actionable takeaways. Register here Dates: March 20-21, 2025 Venue: J N Tata Auditorium, Bengaluru, India Happy Llama 2025 The first edition of Happy Llama is not just a conference, it’s a celebration of India’s vibrant AI startup ecosystem. If you’re an entrepreneur or an investor, this is your chance to connect with the minds driving innovation in AI. The summit includes engaging pitch battles, insightful talks, and workshops tailored to startup needs, making it an unmissable opportunity for those seeking to learn and grow. Whether you’re looking for funding, partnerships, or just some inspiration, Happy Llama provides the perfect platform for it all and more. With its dynamic format and energetic vibe, this one-day event at Bengaluru’s Radisson Blu is your gateway to networking with the best and brightest in AI startups. Register here Date: April 25, 2025 Venue: Hotel Radisson Blu, Bengaluru, India Data Engineering Summit (DES) 2025 For professionals in data engineering, DES is the ultimate event to explore the latest tools, techniques, and trends shaping the field. As India’s first and only conference dedicated to data engineering, it delves into software deployment architectures, data frameworks, and scalable solutions for real-world problems. This summit brings together top engineers and thought leaders to share their expertise, offering attendees a unique opportunity to enhance their skills and knowledge. Whether you work in analytics, machine learning, or cloud computing, DES is a must-attend event. Register here Dates: May 15-16, 2025 Venue: Taj Yeshwantpur, Bengaluru, India MachineCon GCC Summit 2025 MachineCon GCC Summit is the perfect blend of vision and action, designed specifically for leaders in India’s global capability centres. Focused on the transformative potential of generative AI, the summit explores how GCCs can harness AI to drive operational excellence and innovation. With a curated lineup of sessions featuring pioneers and experts, this two-day event offers strategic insights into the future of GCCs. Register here Dates: June 19-20, 2025 Venue: Taj Yeshwantpur, Bengaluru, India Cypher 2025 Saving the best for the last, we have Cypher, AIM’s flagship conference. Since its first edition in 2015, Cypher has grown exponentially to become not just India’s biggest AI summit but also the most impactful one. With over 5,000 attendees daily, the event brings together a diverse community of AI enthusiasts, professionals, and thought leaders. The agenda spans keynotes, panel discussions, and exhibitions, offering a comprehensive view of AI’s impact across industries. Whether you’re a beginner curious about AI’s potential or an industry leader looking for the latest advancements, Cypher has something for everyone. Register here Dates: September 17-19, 2025 Venue: KTPO @ Whitefield, Bengaluru, India","excerpt":"Whether you’re a data engineer, AI startup founder, developer, or corporate leader, these events offer unparalleled opportunities to learn, network, and stay ahead in the fast-paced tech landscape.","categories":["AI Trends"],"tags":["AI conference","Developers"],"author_name":"Mohit Pandey","publish_date":"2024-12-25T16:00:00","publication_year":"2024","word_count":776,"keywords":["data science","machine learning","artificial intelligence","AI conference","AI","cloud computing","TPU","ML","Aim","analytics","generative AI","Developers"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","data science","analytics","generative AI","Aim","cloud computing","TPU"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/6-must-attend-conferences-by-aim-in-2025\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10136572,"title":"Google Announces New Gemini 1.5 Pro and Flash Models with Reduced Pricing and Improved Performance","content":"Google has unveiled two updated production-ready Gemini models—Gemini-1.5-Pro-002 and Gemini-1.5-Flash-002. The release includes a significant 50% price reduction for the Gemini 1.5 Pro model, faster output, and increased rate limits, providing more value for developers working with the Gemini API and Google AI Studio. Key updates include a 52% reduction in output token pricing, 2x higher rate limits for Gemini 1.5 Flash, and 3x higher rate limits for Gemini 1.5 Pro, coupled with improvements in model latency. These updates are designed to lower the overall cost of building with Gemini while improving performance in text, code, and multimodal tasks. “Two new production Gemini models, >2x higher rate limits, >50% price drop on Gemini 1.5 Pro, filters switched to opt-in, updated Flash 8B experimental model, and more. It’s a good day to be a developer,” posted Logan Kilpatrick, lead product manager at Google. The new models show improvements in benchmarks, such as a 20% gain in math-related tasks and up to 7% in Python code generation and visual understanding. Developers will benefit from a more concise response style aimed at reducing costs in summarisation and question-answering tasks. Additionally, Google is introducing new default filter settings, giving developers greater flexibility in configuring safety features. For Gemini 1.5 Pro, a long context window of 2 million tokens is maintained, while the input and output token prices have been reduced by 64% and 52%, respectively. The model is optimised for various tasks, including video understanding and large-scale document processing. Google has also launched an improved version of the Gemini-1.5-Flash-8B-Exp-0924 model, available for developers looking to test experimental updates via Google AI Studio. These changes aim to make the Gemini platform more accessible for production use and experimental development. Developers can access the updated models for free on Google AI Studio or through the Gemini API. Larger organizations can also access them via Vertex AI. These updates will be effective starting October 1, 2024. Google recently introduced DataGemma, a new open model that integrates LLMs with real world data from its Data Commons repository, using retrieval augmented methods like RIG and RAG to reduce AI hallucinations and improve the accuracy of generative AI outputs in research and decision-making contexts. Moreover, Google recently announced that YouTube is set to roll out advanced generative AI tools for creators in the coming months, enabling them to generate video content using the AI models Veo and Imagen 3 through a feature called Dream Screen. When is Google releasing Gemini 2? “That’s the Magic question,” quipped Kilpatrick in an exclusive interview with AIM. However, he said that the company plans to release Veo, Google Search Grounding, Gemini 2.0, and Agents next, tentatively expected to debut in the coming months. “I think it’ll be fun to see what the next six months to a year look like as the trend of unlocking new developments continues,” said Kilpatrick.","excerpt":"Key updates include a 52% reduction in output token pricing, 2x higher rate limits for Gemini 1.5 Flash, and 3x higher rate limits for Gemini 1.5 Pro, coupled with improvements in model latency.","categories":["AI News"],"tags":["Google"],"author_name":"Siddharth Jindal","publish_date":"2024-09-24T22:08:15","publication_year":"2024","word_count":476,"keywords":["Go","API","Gemini 2.0","TPU","AI","RAG","Python","Aim","generative AI","Google","R"],"extracted_tech_keywords":["AI","generative AI","Gemini 2.0","Aim","RAG","TPU","Python","R","Go","API"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/google-announces-new-gemini-1-5-pro-and-flash-models-with-reduced-pricing-and-improved-performance\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10860,"title":"Mu Sigma gets a new CEO as Dhiraj Rajaram takes over","content":"Mu Sigma, an atypical Indian unicorn providing analytics solution to big names such as Microsoft, Walmart, Pfizer have settled for its long awaited decision on who would take up the role of the next CEO. Putting an end to the rumour mill, the founder of Mu Sigma, Dhiraj Rajaram takes over as the CEO. The decision comes after Ambiga Subramanian agreed to sell her stake in the company, giving Dhiraj a majority ownership in the data analytics company once the deal is finalised. There have long been talks about the then CEO Ambiga and General Atlantic, US-based private equity investors about the likelihood of selling their stake in Mu Sigma. Ever since the news of their divorce surfaced, there had been speculations about the company facing turbulence, however they had insisted that it would not affect the functioning of the company. Following their divorce, Ambiga, who owns about 27% stake in the company, had been aiming to exit the company completely. By selling her stake, Dhiraj would own a 51.6% stake in Mu Sigma, making him the single largest stakeholder in the company. Ambiga who has held various important positions in the company before taking over as the CEO, will stay on the board as a director till the transaction is completed, however it’s not clear if she would be associated with the company once the deal is over. She denied setting up a competitive analytics firm in future. “It would be incorrect to say there hasn’t been any impact on business. There have been distractions on the personal front that led to impact on the supply side of business,” Dhiraj had said in a statement. Following the news of divorce, the company had faced a string of resignation from their senior executives. Founded in 2004, Mu Sigma currently employs close to 4000 people and serves more than 140 of the 500 fortune companies. Present across 12+ industry verticals, Mu Sigma had raised about $178 million from investors such as Sequoia Capital, General Atlantic, Fidelity Investments, MasterCard and Accel Partners since 2011.","excerpt":"Mu Sigma, an atypical Indian unicorn providing analytics solution to big names such as Microsoft, Walmart, Pfizer have settled for its long awaited decision on who would take up the role of the next CEO. Putting an end to the rumour mill, the founder of Mu Sigma, Dhiraj Rajaram takes over as the CEO. The […]","categories":["AI News"],"tags":["mu sigma"],"author_name":"Srishti Deoras","publish_date":"2016-10-05T06:48:05","publication_year":"2016","word_count":342,"keywords":["API","unicorn","programming_languages:R","AI","Aim","analytics","R","mu sigma"],"extracted_tech_keywords":["AI","analytics","Aim","R","API","unicorn","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/mu-sigma-gets-new-ceo-dhiraj-rajaram-takes\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10021067,"title":"From Zero Coding Skills To Top ML Job At FAANG: Interview With Rahul Agarwal","content":"“Coming out from college, I didn’t know programming, I learned a little bit of SQL at my first job, but nothing much.” For this week’s machine learning practitioner’s Analytics India Magazine(AIM) got in touch with Rahul Agarwal, an ML engineer at Facebook. Rahul has a bachelors in mechanical engineering from IIT Delhi and has previously held machine learning roles at Walmart Labs and Citibank.  In this interview, he gave a glimpse of the ground reality of making it to top ML roles. AIM: How did your journey in machine learning begin? Rahul: I remember that distinctly. I was sort of lucky to stumble upon the Machine Learning course from Andrew Ng. Andrew Ng was talking about the regression problem and he was using the housing data to explain it. The moment he told about the prediction function. Just pass the variables and get the price. That moment was sublime. Is it magic? Coming from a background where I am good with maths, I couldn’t have thought of such a formula approximation myself and that hooked me. How maths can be used to solve these practical problems. AIM:What were the initial challenges and how did you address them? Rahul: Coming out from college, I didn’t know programming, I learned a little bit of SQL at my first job, but nothing much. I didn’t have a mentor or someone who could have guided me while starting with Data Science. The thing that helped me the most was making a plan once I realised that I like this particular stream. I made a plan of all the things I would have to learn: Went through a lot of course structures that were taught in various universities and tried to find something similar to those since I couldn’t leave my job. At my job, I tried to get headfirst into the data science projects whenever I could get my hands on them. It didn’t really matter if I would be able to do them or not. That was always something that I would consider later while understanding more about the problem myself or while googling about the problem. Even if I would not be able to solve something I still would learn a lot about Data Science while trying to solve it. And I have learned that it would rarely happen that you wouldn’t be able to suggest any solution to a particular problem if you do your research well. AIM: What books and other resources have you used in your journey? Rahul: Before I answer this question there is a disclaimer that there is no particular book or course which will teach you everything. You would have to take a lot of them and find out yourself what suits your learning needs and what suits how you learn. I loved the Machine learning Book from Tom Mitchell. I also liked the Probability and Statistics by Joe Blitzstein. If you are looking at the list of courses I took in my Data Science Journey I have neatly summarised them here. AIM: How was your experience interviewing for an ML role at Facebook? Rahul: It was August last year and I was in the process of giving interviews. By that point, I was already interviewing for Google India and Amazon India for Machine Learning and Data Science roles respectively. And then my senior advised me to apply for a role in Facebook London. Later, as I studied for the FB coding interview, I realised that I took it a little light and that I was not prepared for the coding interviews at all. The first interview was telephonic. This was a very basic data structure interview and sort of a basic sanity check. This was followed by two onsite coding rounds and system design round. For more details check How I cracked my MLE interview at Facebook. Recommendations by Rahul Agarwal Cracking the Coding Interview by Gayle Laakmann McDowellAlgorithm Specialization on Coursera by UCSanDiegoSystem Design Primer by Donne MartinTechdummies Youtube videosFacebook Field Guide to Machine Learning video series Also Read: Rahul Agarwal’s MLWhiz blog. AIM:What does a typical day look like for you as an ML engineer? Rahul: You start with reading your mails and replying to the important ones. Then you start with your list of tasks for the day which would include some Data Exploration tasks or working on some feature engineering, or building a model or checking some hypothesis. In the meantime, I would have a lot of meetings with people regarding data issues or how would we evaluate our models. Sometimes the meetings are with business stakeholders to understand their pain points and what we could do to solve those issues. AIM: How do you approach any machine learning problem? Rahul: I try to tackle any machine learning problem by first looking at the data itself as I guess any Machine Learning engineer would do. Only once you look at the data you would be able to frame a rough structure of how you would like to solve a particular problem in your head. My basic principle has always been to come up with a very simple baseline solution in the beginning and continually improve on that. The motivation behind that is that you can actually try to optimise and tune your machine learning model forever. The first thing you should always strive for is finding the minimal value you can get by solving the problem as a Machine learning problem. Another point I would like to emphasise is to fail fast. Go into production as soon as possible. Only after testing your model in real life circumstances would you really know the effectiveness of the model, what it fails on and the things you want to focus on. AIM: What does your machine learning toolkit look like? Rahul Agarwal’s workstation Rahul: It pretty much depends on the person. I like to use Python as my language of choice but I know many good people who like R and some who are doing Data Science using Javascript. So I would say it depends on the project. My advice would be to not bind yourself to any tool\/frameworks as frameworks and tools will come and go. Try to understand the mathematical concepts and the logic behind a data science algorithm and you should be fine. In the end, learning a Tool and Framework would also become a daily part of your existence when I see how fast the ecosystem changes. AIM: There is a lot of hype around machine learning. So, when the dust settles down,which domain might stand the test of time? Rahul: Why do you think the dust will settle down? I don’t see Deep Learning going anywhere with its adoption being at near country levels. The recommendation systems are never going away from our lives whether it be product recommenders or ads. I see ML being more and more engineering heavy with every passing day. Yes, we still test hypotheses and we still love our research problems but a far higher weight is given to putting things in production and derive value fast. ML is going to play a key part in Maps which will still continue to improve. The Robot industry is set to have a boom in the next few years with reinforcement learning. The AR\/VR industry is something that could really be the next big thing. I guess we have just started. AIM: What do outsiders get wrong about this field? Rahul: That it is easy to get into and reading a book or doing a course\/bootcamp could teach you about AI\/ML. Right now, you need a lot of math and programming background just to even start. And a lot of luck to get a job and show what you are capable of. That is not to say it is impossible, but be prepared to spend a lot of time learning before getting in and learning never stops in this field. AIM: From a global AI perspective, where do you think India stands? Rahul: From my perspective, India would no doubt be a major beneficiary of this AI\/ML wave. The first reason being that India has always been at the forefront of software. And AI\/ML is no different. India needs to research more on AI\/ML. We can see it happening with various Big companies opening up their AI labs in India but it needs to happen at the college level too where students get proper research infrastructure and good mentorship.","excerpt":"“Coming out from college, I didn’t know programming, I learned a little bit of SQL at my first job, but nothing much.” For this week’s machine learning practitioner’s Analytics India Magazine(AIM) got in touch with Rahul Agarwal, an ML engineer at Facebook. Rahul has a bachelors in mechanical engineering from IIT Delhi and has previously […]","categories":["AI Features"],"tags":["Interviews and Discussions","job interview"],"author_name":"Ram Sagar","publish_date":"2021-03-01T16:00:00","publication_year":"2021","word_count":1414,"keywords":["data science","machine learning","AI","ML","recommendation systems","Python","Aim","deep learning","analytics","job interview","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","data science","analytics","Aim","recommendation systems","Python","R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/from-zero-coding-skills-to-top-ml-job-at-faang-interview-with-rahul-agarwal\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10053709,"title":"Data Centres Launched In India In 2021","content":"India’s data centre sector is witnessing a major boom, and it will keep on growing in the coming years. A JLL report says that the Indian data centre industry capacity will double by 2023. This growth is mainly fueled by cloud adoption, IoT devices, desire for data localisation, and 5G rollout, among others. Seeing this opportunity that exists in the Indian market, many big names in tech have been opening data centres in India, and this pattern continued in 2021 as well. Let’s take a look at the data centres launched in the country this year: Google Cloud Google Cloud launched the second data centre cluster for India in the Delhi-National Capital Region (NCR). In a recorded webcast in July, Thomas Kurian, CEO of Google Cloud, said that the company has been expanding the local infrastructure to meet customer demand since it opened the Mumbai cloud region in 2017. It will give more choice to customers to store their data locally to meet emerging data sovereignty concerns. This centre will provide solutions for problems like disaster recovery within India and ensure low latency for many state-run enterprises in and around Delhi, Kurian had added. Trend Micro Cybersecurity major Trend Micro launched its Cloud One regional data centre in India for data sovereignty and data privacy. Last year, it launched its regional XDR data lake in the country. It became one of the first security vendors to have the broadest set of services around cloud security provided by India. Through this, the company will cater to application security, file storage security, network security, and workload security needs. AdaniConneX Adani Enterprises and data centre operator EdgeConneX teamed up to form a joint venture to develop and operate data centres pan India. They said that they will invest significant capital over the next ten years to build the country’s leading green data centre platform. This joint venture will also develop Edge data centres strategically located throughout India. They said that these sites will be made with capabilities to scale up easily and eventually become full-scale data centre campuses. Teleindia Networks Teleindia Networks launched its greenfield hyperscale data centre – DataSamudra – in Bengaluru at the KIADB IT Park. It covers a one lakh sq ft built-up area. It will provide 500 high-density rack, IT power load of 3 MW, with futuristic IT and core infrastructure. DataSamudra will be India’s first ‘on-demand on-requirement data centre’, said the company. NTT and Tokyo Century Corporation NTT Global Data Centers Corporation and Tokyo Century Corporation announced in November the expansion of their data centre business partnership to the Navi Mumbai region. They said that they would jointly own assets related to the data centre business owned by NTT Global Data Centers NAV2 Private Limited (NAV2) through an intermediate holding company (SPV) which will be newly established in Singapore. NTT said that it aims to provide high-quality services to global companies by maximising investment efficiency and accelerating data centre business investments. Acronis Cybersecurity company Acronis launched Acronis Cloud Data Center in Mumbai jointly with local partners Compuage India, Ingram Micro and Crayon Software Experts India. The company said that this data centre, like the 111 new data centres being deployed by the company globally, will give service providers access to cyber protection solutions upon which they can build new services that will give faster access, constant data availability, and data sovereignty to clients. The UP Government In August, the Uttar Pradesh government announced its plan to develop a data centre park at Sector-28 of the Yamuna Expressway Industrial Development Authority (YEIDA) in the Noida district. A government spokesman said that the data centre park at Noida will attract investments from information and technology (IT) industry giants, from abroad as well as from within India.","excerpt":"A look at the data centres launched in the country this year","categories":["IT Services"],"tags":["data centre india","data centres","Google Cloud","IoT","NTT"],"author_name":"Sreejani Bhattacharyya","publish_date":"2021-11-18T12:00:00","publication_year":"2021","word_count":626,"keywords":["data centre india","Go","API","Google Cloud","programming_languages:R","AI","RAG","Ray","Aim","data centres","cloud_platforms:Google Cloud","R","data lake","IoT","NTT"],"extracted_tech_keywords":["AI","Aim","Ray","RAG","R","Go","API","data lake","cloud_platforms:Google Cloud","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/it-services\/data-centres-launched-in-india-in-2021\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10121325,"title":"Microsoft Enhances G42 Partnership with $1 Billion Digital Investments in Kenya","content":"On Wednesday, Microsoft and G42 unveiled a major digital investment initiative in collaboration with Kenya’s Ministry of Information, Communications, and the Digital Economy. G42, in partnership with Microsoft and other stakeholders, will spearhead an initial investment of $1 billion to support various projects within this initiative. G42, in collaboration with local partners, will design and build a cutting-edge data center campus in Olkaria, Kenya. This facility will be powered entirely by renewable geothermal energy and feature advanced water conservation technology. The data center will host Microsoft Azure services through a new East Africa Cloud Region, set to become operational within 24 months of the agreement signing. East Africa Digital Expansion The initiative also includes four additional pillars to be developed with local partners that includes development and research of AI models in local languages, establishment of an East Africa Innovation Lab and extensive AI digital skills training, investments in international and local connectivity, and collaboration with the Kenyan government to ensure safe and secure cloud services across East Africa. This new cloud region will offer customers scalable, secure, high-speed cloud and AI services, accelerating cloud adoption and the digital transformation of businesses, customers, and partners across Kenya and East Africa. Brad Smith, vice chair and president of Microsoft, said, “This represents the single largest and broadest digital investment in Kenya’s history and reflects our confidence in the country, the government, its people and the future of East Africa.” G42 has also started training an open-source large language AI model in Swahili and English using its data infrastructure in the United States. To further advance research in Kenya, Microsoft and G42 will strengthen their collaboration with local universities Today, in partnership with @G42ai, we are announcing a $1B investment in Kenya which represents the single largest and broadest private digital investment in the country’s history. It reflects our confidence in the country, the government, its people, and the future of East… pic.twitter.com\/3e9dtwlMH4— Brad Smith (@BradSmi) May 22, 2024 Microsoft’s Aggressive Global Expansion In April, Microsoft invested $1.5 billion in UAE-based technology holding company G42, where the focus was to expand AI technologies and skilling initiatives not only in the UAE, but across the globe. Microsoft also committed to investing in South East Asian countries in the last couple of months. The company announced $1.7 billion to advance Indonesia’s Cloud and AI infrastructure. The first data center in Thailand was opened up last month by Microsoft and committed to providing upskilling opportunities for over 100,000 people.","excerpt":"Through the agreement, green data centers will be built in East Africa cloud region to run Microsoft Azure.","categories":["AI News"],"tags":["Azure","Data Center","G42","Indonesia","Kenya","Microsoft","uae"],"author_name":"Vandana Nair","publish_date":"2024-05-23T12:59:03","publication_year":"2024","word_count":414,"keywords":["Go","uae","Data Center","cloud_platforms:Azure","AI","Azure","R","digital transformation","innovation","G42","Scala","Git","Kenya","ViT","Indonesia","Microsoft"],"extracted_tech_keywords":["AI","Azure","R","Go","Scala","Git","ViT","digital transformation","innovation","cloud_platforms:Azure"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/microsoft-enhances-g42-partnership-with-1-billion-digital-investments-in-kenya\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10068606,"title":"IIT Jodhpur launches MTech in AR\/VR","content":"IIT Jodhpur has announced an MTech program in Augmented Reality (AR) and Virtual Reality (VR) for their semester starting 2022. In collaboration with iHub Drishti Foundation, the university’s Technology Innovation Hub, the course will be conducted part-time online along with campus immersions where students can get hands-on experience with state-of-the-art AR\/VR applications. Open to all working professionals, registration for the course will be open until 17th June. The program aims to provide knowledge about AR\/VR tools so that engineers can gradually grow the ability to develop these systems and eventually resolve problems associated with AR\/VR. “AR and VR is the future technology that is going to take an important and increasingly major role in a variety of areas such as Healthcare, Diagnostics, Robotics, Gaming, Consumer Experience and anywhere else where we require an immersive experience. This is where the entire technology is moving. This is going to lead to increasing job opportunities for those who are experts in AR\/VR. This course will provide opportunities to working professionals to be future-ready for the emerging job market,”said Dr Neeraj Jain, Head of Department of IIT Jodhpur’s School of AI & Data Science. The main objectives of the program are: To produce competent engineers who can design and develop AR and VR applications,To impart in-depth knowledge, and analytical and experimental research skills to solve AR and VR systems problems,To provide knowledge of several tools for the design and modelling of AR and VR applications and immersive experiences, andTo develop the ability to cultivate technological solutions for addressing the growing demands of AR and VR systems. “This is possibly the first MTech program in this area, offered by an IIT in this space. The uniqueness of the program is that this has been devised and designed through collaboration of experts across IITs. The course will be taught and executed by a team of experts across the country including experts from industry,” said Professor Santanu Chaudhury, Director, IIT Jodhpur.","excerpt":"Open to all working professionals, registration for the course will be open until 17th June.","categories":["AI News"],"tags":["IIT"],"author_name":"Poulomi Chatterjee","publish_date":"2022-06-08T18:12:22","publication_year":"2022","word_count":324,"keywords":["data science","Go","programming_languages:R","AI","innovation","programming_languages:Go","Aim","ai_applications:robotics","IIT","R"],"extracted_tech_keywords":["AI","data science","Aim","R","Go","innovation","programming_languages:R","programming_languages:Go","ai_applications:robotics"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/iit-jodhpur-launches-mtech-in-ar-vr\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10040659,"title":"The Impact Of AI &#038; Data Science On Incredible Growth Of Happiest Minds","content":"Happiest Minds Technologies has recorded a whopping net profit of INR 36.05 crores– a 580.19% year on year growth — in the latest quarter. The IT company’s sales for the same period stood at INR 220.71 crore–an 18.44% spurt. Headquartered in Bengaluru and with offices in Pune and Delhi NCR, Happiest Minds Technologies helps businesses in their digital transformation journey. The company offers Internet of Things (IoT), analytics\/artificial intelligence, and digital process automation services. The company has 173 clients as of March 31, 2021. Currently, it employs more than 3,228 people. Last year, the company was recognised among India’s Best Companies to Work for 2020 by Great Place to Work. This global authority certifies organisations (with employee strength of more than 500) for their workplace culture. AI and data science contribution Happiest Minds said AI and data science play a pivotal role in empowering their customers. The firm’s Analytics Center leverages the latest deep learning and machine learning techniques to extract insights from data to drive customers’ revenues and optional efficiencies. “Analytics offers a healthy 13 plus percent of direct contribution to the growth of our organisation growing at about 30 plus percent year-on-year (YoY),” said Ajay Agrawal, senior VP and Head of Analytics\/AI CoE at Happiest Minds. Further, he said, artificial intelligence and machine learning, alongside data science, are game-changers. “With exciting times ahead, we are expecting robust growth across various verticals. We are working on enabling products and transformation, leveraging advanced NLP, vision analytics and AI at scale,” he added. The company is  investing in growth and has added the latest breed of tools and technologies to its artillery. “We are focusing on cross-skilling and multi-skilling and working with customers on delivering digital-ready solutions built on most differentiated and niche technologies like deep learning, artificial intelligence, machine learning and data science,” said Agrawal. Happiest Mind’s 96.6% of revenue came from digital services in the 2020 fiscal. Other factors “The highlight for FY21 was our successful IPO. We will look to achieve 20 percent organic growth as we begin FY22,” said Ashok Soota, executive chairman at Happiest Minds Technologies. Happiest Minds Technologies got listed (NSE: HAPPSTMNDS) in September 2020. Happiest Minds Technologies CFO Venkatraman N credited the latest fiscal growth to its business model, which positions itself as a go-to player for customers to build digital-ready enterprises. Revenues for the quarter grew sequentially by 15.4 percent (Pimcore Global Services (PGS) acquired earlier during the quarter contributed), he added. Earlier this week, Happiest Minds successfully executed a digital transformation project for Coca Cola Bottling Company United. The company streamlined Coca Cola’s order management with robotic process automation (RPA) in Microsoft Power Automate. Also, the company signed a deal with information security company CyberArk to deliver access management services to customers across the globe.","excerpt":"Happiest Minds Technologies has recorded a whopping net profit of INR 36.05 crores– a 580.19% year on year growth — in the latest quarter. The IT company’s sales for the same period stood at INR 220.71 crore–an 18.44% spurt. Headquartered in Bengaluru and with offices in Pune and Delhi NCR, Happiest Minds Technologies helps businesses […]","categories":["IT Services"],"tags":[],"author_name":"Amit Naik","publish_date":"2021-05-23T17:00:00","publication_year":"2021","word_count":463,"keywords":["data science","machine learning","artificial intelligence","AI","ML","RAG","NLP","deep learning","analytics","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","NLP","data science","analytics","RAG","R"],"url":"https:\/\/analyticsindiamag.com\/it-services\/the-impact-of-ai-data-science-on-incredible-growth-of-happiest-minds\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10131094,"title":"GalaxEye Space Secures $6.5 Million in Funding Round Led by Mela Ventures and Speciale Invest","content":"Bengaluru-based space tech startup GalaxEye, has secured $6.5 million in a funding round led by Mela Ventures and Speciale Invest. The funds will be used to launch the company’s first satellite, the “Drishti Mission,” and to further develop their multi-sensor payload technology. The funding round also saw participation from ideaForge, Navam Capital, Faad Capital, Anicut Capital, and Rainmatter which was founded by Nithin Kamath and Kailash Nadh from Zerodha fame. The IIT-Madras incubated startup GalaxEye Space aims to revolutionise satellite image acquisition by becoming the world’s first company to integrate hybrid optical multi-spectral imaging (MSI) and synthetic aperture radar (SAR) in a single payload. The startup recently signed an agreement with iDEX DIO to design and develop a multi-sensor fusion processing system for miniaturised satellites that are capable of carrying payloads up to 150 kg for the Indian Air Force. The startup was founded in 2020 by members of the Team Avishkar Hyperloop, and is focused on providing precise datasets from space and aerial platforms. These datasets are essential for various industries, including defence, surveillance, agriculture, insurance, and aquaculture. Co-Founder and CEO at GalaxEye, Suyash Singh said, “GalaxEye Space is entering the space backed by the success of 300+ flights with UAV SAR payload. We were looking for a strong partner at this stage to enable the smooth journey and can’t think of better partners. We share the goal of building a global category leader out of India solving real-world challenges through technology and look forward to guidance and mentorship of the leadership teams of our investors.” GalaxEye Space is building a constellation of indigenous micro-satellites with data fusion capabilities, positioning itself as a key player in the space technology sector. The company’s UAV SAR system has already completed over 300 successful flights for defence applications. With this new funding, GalaxEye Space is poised to advance its mission of providing advanced imaging solutions and contributing to the global space technology landscape. Spacetech Attracting Investments Investments in space tech startups in India have witnessed significant growth over the last few years, with $126 million in funding last year—a 7% increase from 2022. GalaxEye Space alone had previously raised $3.5 million in seed investment, with the round led by Speciale Invest. Other participants in the funding round included Artha India Ventures, Veda VC, Anicut Capital, Upsparks, and a group of entrepreneurs. Vishesh Rajaram, Managing Partner at Speciale Invest, said, “We have been proud supporters of GalaxEye from the beginning and believe in their vision to transform space technology. With this new round of funding, we are excited to continue our partnership and witness the impactful advancements they will bring to the market.”","excerpt":"The space startup had previously raised $3.5 million which was led by Speciale Invest.","categories":["AI News"],"tags":["Fund Raising","galaxeye","India","Space Tech"],"author_name":"Vandana Nair","publish_date":"2024-08-01T11:25:37","publication_year":"2024","word_count":440,"keywords":["Go","API","funding","India","startup","AI","programming_languages:R","data_tools:Spark","Fund Raising","programming_languages:Go","Aim","R","Space Tech","galaxeye"],"extracted_tech_keywords":["AI","Aim","R","Go","API","startup","funding","programming_languages:R","programming_languages:Go","data_tools:Spark"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/galaxeye-space-secures-6-5-million-in-funding-round-led-by-mela-ventures-and-speciale-invest\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10096181,"title":"Why are Companies Creating Specialised Generative AI Roles?","content":"Coca-Cola recently appointed Pratik Thakkar as its global head of generative AI and PayPal onboarded Vidyut Naware as the head of generative AI centre of excellence. Meanwhile, San Antonio-based multicloud solutions provider, Rackspace Technology, appointed its CTO Srini Koushik as the global head of the company’s newly launched Foundry for Generative AI. And last month, Mphasis announced a dedicated business unit for generative AI, led by the group CTO Anup Nair. But why are we telling you all this? To draw your attention to an emerging trend and an alternative. So, why is it that companies are setting up specialised roles to drive generative AI initiatives? Is it really needed, or are companies just hopping on to the generative AI bandwagon by appointing leaders dedicated to drive their vision? The Rise of Generative AI Leaders In the backdrop of Coca-Cola’s Masterpiece ad, it makes sense for the company to elevate Thakkar, who was former head of global creative strategy and content to take up the generative AI global head’s role. In his new role, he would be in charge of leveraging AI technology in creating big ideas and developing marketing strategy across the entire brand and category portfolio of the Coca-Cola Company. Meanwhile, PayPal’s head of generative AI CoE is looking at using generative AI for risk assessment, customer service, and marketing and automation. PayPal is tackling a slump in customer acquisition since the pandemic and its woes with the stock market crash. The recovery requires the adoption of new AI technologies, said Dan Schulman, president of PayPal. “We expect AI will allow us to lower the costs for years to come. AI, combined with a unique set of data, will drive efficiencies and a set of value propositions. Despite the fact that today’s environment is difficult to forecast, we are positioned to reap the investments in our products into 2024,” Schulman said at a recent earnings conference. Now, the company is going big on generative AI to save the day. CTOs Are Enough Rackspace Technology and Mphasis, on the other hand, have not created a specialised role for taking care of its generative AI initiatives, instead, are primarily driven by their chief technology officer (CTO). Last month, Rackspace announced the launch of Foundry for generative AI by Rackspace or (FAIR), calling the platform “a groundbreaking global practice dedicated to accelerate the secure and sustainable adoption of generative AI solutions across industries”. Its offerings are meant to help organisations identify uses for generative AI, integrate and optimise its efficiency. In 2021, Koushik became the CTO of Rackspace and was incharge of the company’s strategy and security. Since then, their AI development has been a combination of open source projects (Hugging Face and stability AI), resources from the acquisition of AI analytics firm Just Analytics and the development of its own generative AI — Intelligent Co-pilot for the Enterprise (ICE), which uses AI to “automate routine tasks, identify warm leads, surface relevant data and content, and provide real-time contextualised analytics for hyper-personalised customer interactions”. Mphasis went the self-reliance route, set up a new division known as mphasis.ai which offers guidance on the integration of generative AI solutions, create proprietary generative AI technologies, provide licences to more than 250 AI models through its ‘Hyperscaler’ solutions platform, and collaborate with 50 startups to assist clients in solution development. The new vertical will also supply clients with conversational AI tools, such as chatbots, to employ in their businesses. Nair who has been the CTO of Mphasis for seven years now will lead mphasis.ai as its chief architect. Generative AI Roles Galore The number of jobs in generative AI has tripled in the last one year, not just in the senior leadership roles but also in the lower- and mid-levels. With skills from data science, linguistics, analysts, natural language processing and the others in policy which companies are promoting, there are new teams being created that need leadership and specialisation. As per LinkedIn, there are close to 846 people with the job role product manager in generative AI. A large number of companies are quickly adding generative AI features or building their own LLMs. According to a Goldman Sachs report, a significant increase in the global increase in GDP is brought about by generative AI. Technological innovations are leading to the creation of new occupations, such as model trainers, generative AI architects, AI product managers, generative AI prompt engineers and so on. So it isn’t surprising that every other company is adopting it without losing time. Generative AI has been used for some time for automating work activities and more recently for collaboration. As it continues to evolve organisations can enhance innovation, product design, optimise supply chains among other myriad use cases coming up. Apart from wanting to maintain a competitive edge, understanding what use cases, strategising, creating, and implementing the road map for the same will require specialised AI roles or retraining of a part of the workforce( or both). Most important, people filling these roles would need to anticipate and navigate challenges especially those related to data quality, infrastructure requirements, and associated costs and incorporate any ethical regulations or legal compliances that are required.","excerpt":"The number of jobs in generative AI has tripled in the last one year, not just in the senior leadership roles but also in the lower and mid levels","categories":["AI Features"],"tags":["AI Companies"],"author_name":"K L Krithika","publish_date":"2023-07-03T15:55:13","publication_year":"2023","word_count":861,"keywords":["data science","Hugging Face","Go","AI","chatbots","RAG","generative AI","analytics","AI Companies","data quality","R"],"extracted_tech_keywords":["AI","data science","analytics","generative AI","Hugging Face","RAG","chatbots","R","Go","data quality"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/why-are-companies-creating-specialised-generative-ai-roles\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10098059,"title":"Why Do You Need Worldcoin When You Have Aadhaar?","content":"Now that we have witnessed the euphoria surrounding Worldcoin with people queuing to get their irises scanned and the crowd going berserk, leading to a shut down of all three initial locations of orb operations in Bengaluru last Friday, let’s explore the big picture for Worldcoin in India. Drawing parallels to the government-organised large-scale biometric scan that began 12 years ago, which also claimed to create a unique digital identity and provide benefits to citizens via welfare schemes, it looks like Worldcoin may not be too different from Aadhaar — but it is! India Leading The Way Former chief economist of World Bank, Paul Romer referred to Aadhaar as the most “sophisticated system” he had ever seen, and considered it to be “good for the world if this became widely adopted”. As a prelude to organisations planning to adopt similar mechanisms, it is not surprising that Worldcoin also took inspiration from Aadhaar. In their 2021 blog that launched Worldcoin, the company mentioned that biometric approaches for proof of personhood is an accessible method and allows accurate verification of uniqueness, calling out parallels to Aadhaar in India. Though they admitted to challenges related to privacy and fraud detection, the company believes latest advancements can solve it. not many westerners are aware of the Aadhaar system running in India — an existence proof of the onboarding of billions of humans onto a biometric identity platform https:\/\/t.co\/GnQzXUWiUP— Worldcoin (@worldcoin) July 11, 2022 While another method to approach proof-of-personhood is through the social-graph method, with a number of companies already adopting such a technique, scalability challenges impede its adoption. For a project like Worldcoin, biometric was the viable option. OpenAI CEO Sam Altman, in his last visit to India, appreciated India’s efforts in building and adopting technology, with a special mention to Aadhaar and UPI systems that emerged here. With an ambitious goal of 1 billion signups, Worldcoin announced their grand project two years ago. As of now, they have achieved over 2 million sign ups and counting.  Interestingly, within five-and-a-half years, Aadhaar clocked 1 billion registrations. India Unperturbed by Worldcoin Though the goals for both projects are vaguely similar, the modus operandi are different. Creation of digital identity is one of the goals for Worldcoin, crypto tokens being another. On successful completion of the biometric scan, Worldcoin offers 25 ‘World tokens’ (WLD). Each WLD is a little over $2. While the broader goal of the company is to distribute wealth generated by AI back to society, the timeline to implement this goal is uncertain and will probably fall in the distant future. Furthermore, being a private organisation, no form of enforcement can be implemented, unless a government affiliation or law comes into the picture — which obviously means that the distribution of universal basic income will not come to fruition. However, Aadhaar has no such obstacles. Though not mandatory, in order to avail welfare schemes, individuals need to possess an Aadhaar card. As per the Economic Survey 2023, 318 Central schemes and over 720 state DBT (Direct Benefit Transfer) schemes come under Aadhaar Act, and these are facilitated via Aadhaar cards. In this grand scheme of things, where Aadhaar is functioning like a well-oiled machine, Worldcoin will not hold a candle to it. Rising Mutiny While Worldcoin is making its way across the globe with setting up orb scanning locations across 20 countries (35 cities), not all has been rosy for the company. Investigations and watchful scrutiny has risen regarding Worldcoin operations. Yesterday, Kenya became the first country to ban Worldcoin operations owing to security and financial concerns. The suspension of Worldcoin also applies to other entities that operate in a similar fashion that “engage the people of Kenya”. It will not operate until the authorities ascertain there is no form of risk to the public, as they are looking into how and what would be done with people’s data. However, the orb locations across Kenya witnessed throngs of people. Within a week of its launch, UK and European regulators were quick to jump onto the scrutiny train. Countries including France and Germany have started investigating its operations over concerns of collecting sensitive biometric data. It’s not surprising that the EU, known for having stringent data regulatory policies in the world, will scrutinise the project, but it’s strange that they are looking into it after it’s been launched. With countries waking up to rigorous scrutiny, it looks like India might follow the regulatory scrutiny too. Besides, with Aadhaar already in place and pretty much meeting the goals of ambitious Worldcoin, why have it here in India?","excerpt":"Aadhaar took five-and-a-half years to reach 1 billion registrations. Worldcoin achieved 2 million signups in two years","categories":["AI Features"],"tags":["biometrics","Cryptocurrency","EU","Kenya","orb","uk","UPI","WLD","WorldCoin"],"author_name":"Vandana Nair","publish_date":"2023-08-03T14:50:56","publication_year":"2023","word_count":763,"keywords":["EU","uk","Cryptocurrency","Scala","orb","Git","WLD","Kenya","R","fraud detection","Go","AI","GAN","biometrics","OpenAI","programming_languages:R","WorldCoin","UPI","Aim"],"extracted_tech_keywords":["AI","OpenAI","Aim","fraud detection","R","Go","Scala","Git","GAN","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/why-do-you-need-worldcoin-when-you-have-aadhaar\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":11937,"title":"Interview of the week | Anil Chhikara, Principal at Jaarvis Accelerator","content":"Jaarvis is one of the leading accelerators in India that analytics industry is blessed with. With an objective of converting an entrepreneurial idea into a fundable proposition in the shortest possible period of time, Jaarvis Accelerator focuses on early stage technology based startups. It helps the startups build a sustainable business rapidly by offering a four-month residential accelerator program which is developed exclusively by the seasoned entrepreneurs. Aiming to help startups develop, build and grow their business, Jaarvis provides concrete plans for development, marketing and customer acquisition. Cofounded by Jaspal Sarai and Gabriel Fong, the accelerator program is headed by Anil who is the Principal at Jaarvis Accelerator. An entrepreneur at heart, he believes in dreams and is passionate about helping new startups succeed. At Jaarvis Accelerator, Anil works with startups to help them create an effective go-to-market strategy, get the revenue traction going and achieve what they need to get to next level of funding and growth. In a tete-a-tete with Analytics India Magazine, Anil talked in detail about Jaarvis Accelerator, the startups in the analytics space and much more. Dive in to find more. Analytics India Magazine: How does Jaarvis pick on the start-ups to extend its support? What qualities do you look for in a start-up? Anil Chhikara: We identify early stage startups that are looking at disrupting established markets using tech innovation. We help them in their startup journey from a ready product or service to a viable revenue generating business. As we are dealing with tech startups focused on enterprise business, we look for deep and relevant domain and tech expertise in the founders, solving real large scale problems and potential to create big disruptions. AIM: What kind of support is extended by Jaarvis to these start-ups? AC: There is much more to a successful startup besides a good product or service. The journey involves the right legal, technical, domain, marketing, distribution, sales and other support besides access to mentors, industry experts and investors. Our 6 month acceleration program provides the right mix of handholding to the startups so that they can successfully cross the chasm from a product\/service with potential to generating revenues from customers. We offer 24×7 access to our office space, creating an environment where young entrepreneurs learn and succeed.  With a strong base and connect across international markets, we are also one of the few accelerators in India to offer its startups access to international strategic partnerships. For example, we offer a special Fintech Blockchain accelerator program in partnership with a HongKong based large financial institution. Under this program, startups get access to customers in Hongkong for early validation of their products and services, access to matured financial market and global industry experts. Besides these, we also provide funding support of up to $50,000 USD depending on the startup requirements. AIM: What strategies should a startup in the analytics space adopt to make a mark in the industry? AC: Analytics is a very wide space so it’s important to understand exactly what learnings from the data can be had and what value will be generated from that. Sometimes startups get the idea that there is a significant problem to be solved and the analytics will provide the solution but they miss the key question, “ who will be the paying customer for this value we are proposing to create” [quote]The key is to find a problem that someone is willing to solve and provide that extra edge to business for a return.[\/quote] AIM: What are the major challenges that a start-up in analytics domain face today? AC: We believe the biggest challenges facing the analytics startups is talent. Good talent is difficult to get and that’s why the cost of talent is also going through the roof – making it that much more tough for startups to hire them. Besides this, the perennial funding challenge remains – how to find that right investor who gets it, is willing to bet on it and gives the money. Having intimate understanding of the industry and business sector being focused upon is also critical. AIM: Could you name a few analytics startups that Jaarvis has helped in setting a foot in the industry? AC: We have a few startups in this space, Sniffer is using analytics in finding the best experience for a user, Arbunize is using analytics to map the cultural aspects of organisations and the individual traits, so that the right match can help organizations hire the right people. There are a number of others who use analytics very heavily to serve the core value of their venture to the user. We are currently also inviting applications for Big data and Analytics startups to further explore the cognative applications, and specific insights into an established industry sector, that can create significant value for the customers. We hope to have more in the coming months. AIM: How can the quality of startups that are cropping in the industry be improved upon? AC: Startup is a buzz word today, everyone is trying to do their own startup but what they fail understand is that it is not an easy task. First and foremost one needs to understand that starting a new business is no different today than it ever was. It may be that there is more access to information and support to build a new business today, but the fundamentals have not changed. One needs to have a scalable business model with the right mix of strategies, bootstrapped biz plan, prior research and a clear understanding on the sector in which they would work upon is a prerequisite. Besides this, the best way is to connect to the right set of mentors and industry experts that can not only help them validate their ideas but also provide access to real world use cases. Overall quality has improved significantly compared to 2 years ago and we are seeing more maturity in the startups today. AIM: Do you have any tips for the start-ups? AC: Get Feedback – “No one understands my startup like the way I do”, get this thought out of your head. This is a very critical thing, ask as many mentors and serial entrepreneurs feedback. You have to draw it out in a nuanced manner to see what’s right for your startup and what’s not. Network with as many influencers and serial entrepreneurs as you can, word of mouth can play a big role in the success story of your startup. Play it smart, this is not an easy world ! Research before you start, its critical that one understands where the world has gotten up to before you start reinventing the wheel. AIM: Would you like to add anything? AC: Good to see India picking the pace, and the startup ecosystem becoming vibrant with each passing day. Success never comes to you overnight, you have to be at it constantly. Remember the King Bruce and the spider story? Yes, you have to keeping on trying, things change very quickly in this space. ________________________________________________________________________________________ Biography of Anil Chhikara Anil has founded 24×7 Learning, India’s largest e-learning implementation company. Prior to founding 24×7 Learning, Anil had successfully setup and launched businesses in India for two large multinational companies- Parametric Technology Corp, USA and Cimatron, Israel and held leadership positions in Patni Computers, Wipro and HCL-HP. A mechanical engineer by education, he specializes in creating and implementing market strategies, demand creation, international partnerships, setting up distribution networks, B2B market development, product management, acquiring marquee customers for new products, business operations, strategy and high-value solution selling.","excerpt":"Jaarvis is one of the leading accelerators in India that analytics industry is blessed with. With an objective of converting an entrepreneurial idea into a fundable proposition in the shortest possible period of time, Jaarvis Accelerator focuses on early stage technology based startups. It helps the startups build a sustainable business rapidly by offering a […]","categories":["AI Features"],"tags":["accelerator program india","Interviews and Discussions"],"author_name":"Srishti Deoras","publish_date":"2016-12-28T12:22:56","publication_year":"2016","word_count":1262,"keywords":["big data","Go","API","AI","Scala","accelerator program india","Aim","ViT","analytics","GAN","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","analytics","Aim","R","Go","Scala","API","big data","GAN","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/interview-week-anil-chhikara-principal-jaarvis-accelerator\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":61515,"title":"Effective Strategies To Democratize Data Science In Your Organization","content":"Chip Conley, head of global hospitality and strategy at Airbnb, believed that experience of staying in Airbnb is the soul of customer strategy. He accepted that data analytics is the key to understand and connect customer’s voice to Airbnb product line so envisage the experience customers are seeking when select Airbnb. In 2011, when Airbnb expanded out of SFO to 22 other locations internationally, they faced challenges with the team’s mix, collaboration, and business transformation. Airbnb survived the challenge by building data-driven culture and empowering its employees to work with data. Instead of the building spoke teams, democratizing data practices helped them go faster in the decision-making process. “Data democratization” means “liberating data” and “getting data in the hands of decision-makers; it means putting data science tools in the hands of people who are not data scientists,”Chris McGrath – SVP, Data strategy and consumer intelligence, Viacom. Today, no industry is untouched from digital disruption and the global economy is perceiving the business of data as one of the emerging sectors. What empowers Amazon to venture into healthcare? It’s the data that enable Netflix and Uber to transform business and build new markets. Per Forbes, the data economy is expected to grow to 35 zettabytes and to analyze this huge heap, organizations need analysts, if not scientists. The skills to analyze data and extract meaningful insights are still a niche in the industry. Most of the organizations relegate data and analytics responsibilities to a central team. No issues with this equation, but the challenge is how to scale and stay sustainable in the long run. The other major challenge is teaming and communication. Data scientists and business teams spend days in pointless discussions trying to dig through a problem. The lack of data literacy and ability to paint the business context causes brings friction between teams. This impacts business’s decision velocity and the ability to react on the fly, to internal or external factors. What should organizations do? The best way to bridge this gap is to establish data driven culture where everyone (required ones) understands data and possesses elementary knowledge on data driven outcomes. Democratization enables the sharing of tools, knowledge, and skills to consume data and back your decisions with data insights. In the current scenario, almost every industry hold large volumes of data and those who make sense of data succeed. While small central team of data and AI specialists bring standardization, democratizing tools and skills to generalists across the organization may bring empowerment and the ability to exploit data rapidly at scale. Five-point strategy to democratize data science Without data, there can’t be AI. To democratize data science, first step is to democratize data. If employees have access to data and the ability to explain it, all they need is a tool to build data models and judiciously apply data science techniques. Evangelize DATA – For business users to work with data, it is crucial for them to realize power of data. Business users must be educated on problem solving with data, intermediate level of SQL skills, and exploratory data analysis. Data training programs help in upskilling users at a much lower cost, but with a bigger impact. A collaboration workspace suite may act as a repository of FAQs, tutorials, and peer-to-peer discussions.Define PERSONAS – A persona is an employee’s context mapped to his access rights. Organizations usually start with an assessment to survey different roles and capture approaches to consume data. The objective is to define personas, understand what they look in data and build data access framework. For instance, CXO layer needs a dashboard for storytelling with data and visual insights. Likewise, business analysts might need data models for ad-hoc analysis to advise based on factual insights.Share TOOLKIT – Leave aside specialized AI tools, centralized data science teams can share tools that can connect to data systems, query data, run basic analytics, and generate insights. Data science specialists will no longer be burdened by ad-hoc data requests from the business. For standardization viewpoint though, organizations also position a self-service data platform – a low-code platform with drag and drop features, auto ML functions, and data viz capabilities.Build AI components – Democratization of data science starts with self-service data and matures with pertinent use of AI. Data scientists can publish generic pre-trained ML models that can integrate with other products. These packaged solutions enable rapid innovation, avoid rework by saving time and bring standardization.Transform CULTURE – Disruption tends to transform culture by challenging established processes. Democratization is one such disruption that is often perceived with reluctance and hesitation. Organizations must relay the larger vision with the employees for a better tomorrow. Impact and challenges Democratization of data helps in breaking silos and empowering business users when and where required. They need not engage with product or data science teams to explain and justify their case to bring up on priority. Data and the essence of AI enriches with questions. More eyes on the data, more questions and this could result in some unexplored opportunities. Business users, who might love to dig deeper into data, will emerge as citizen data scientists. Citizen roles in data analytics space have ability to think critically and interpret data science outcomes. Business teams with non-technical past must look forward to building this role internally for rapid inquires on data and analytics. With all rewards being detailed, there are risks too with democratization. More users accessing data might put a question on data integrity and privacy. Data access framework should align with organization’s security and privacy norms. Confidential data can still exist in silos and doubly validated when requested. The other challenge is how precisely one interprets the results of a data science exercise. AI results could appear inutile, if misinterpreted. Teams or individuals accessing data must understand the application of data science techniques and must know how to explicate the results. Focused training on data and data tools enables effective onboarding of employees. Data for everyone For data democratization to be effective in VUCA world, organizations find that a centralized data strategy is imperative in breaking data silos into a unified data platform like data lake. Strong data governance in an ever-evolving data lake ecosystem guarantees data availability, keeps check on data quality, and brings consistency in architecture. Companies starting their democratization journey need to reflect into their digital strategy and evaluate data readiness. They can start small with few business teams, target quick wins, share tools and practices, and then go big with evangelization.","excerpt":"Chip Conley, head of global hospitality and strategy at Airbnb, believed that experience of staying in Airbnb is the soul of customer strategy. He accepted that data analytics is the key to understand and connect customer’s voice to Airbnb product line so envisage the experience customers are seeking when select Airbnb. In 2011, when Airbnb […]","categories":["AI Features"],"tags":["Data Governance strategy","PowerBI","what is data science"],"author_name":"Saurabh K Gupta","publish_date":"2020-04-13T12:01:00","publication_year":"2020","word_count":1084,"keywords":["data science","Go","API","AI","ML","Git","RAG","PowerBI","analytics","SQL","what is data science","R","Data Governance strategy"],"extracted_tech_keywords":["AI","ML","data science","analytics","RAG","R","SQL","Go","Git","API"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/effective-strategies-to-democratize-data-science-in-your-organization\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10130152,"title":"TWO Announces SUTRA Dual2L, a Multilingual AI Model Using Llama 3.1","content":"TWO has announced SUTRA Dual2L, a new multilingual large language model (LLM) leveraging Llama 3.1. This model is designed to support over 50 languages, enhancing AI adoption across diverse markets. SUTRA Dual2L uses TWO AI’s dual-transformer (Dual2) architecture, which separates concept learning from language learning. This approach allows SUTRA to integrate state-of-the-art open-source models without retraining from scratch, improving efficiency and offering cost-effective, enterprise-ready solutions. The Dual2L model enhances Llama 3.1’s capabilities by expanding support to more languages, including Korean, Gujarati, Tamil, Malayalam, and Japanese, with a 5-10% improvement in multilingual performance for the 70B model. It also employs a multilingual tokenizer with a 256K vocabulary trained on a balanced dataset, reducing token consumption costs for non-English languages by 3-5 times. SUTRA Dual2L aims to provide cost-effective multilingual AI solutions globally. The models will soon be available via API and in AI Enterprise Solutions for enterprise customers. Interested businesses can contact TWO for information on availability. TWO AI  recently launched the SUTRA for Startups program, an initiative aimed at accelerating startups with access to its multilingual LLM-based services. The program will select 10 startups from India, Korea, Japan, or the Middle East to receive API access to SUTRA and 1 billion tokens for free. The selected startups will leverage SUTRA’s scalable and cost-efficient text-based models to develop solutions targeting global markets. This marks the first round of the SUTRA for Startups program, designed to foster innovation by providing affordable multilingual LLM capabilities to emerging companies. Eligible startups must have existing LLM-based services and be headquartered in the specified regions. Applications are open until July 31, 2024, at 11:59 PM PST. Winners will be announced soon after the submission deadline. Jio backed TWO AI recently announced the launch of SUTRA through its new AI app, ChatSUTRA, available at two.chat.ai. It  is now accessible via web and will soon be available on iOS and Android. The startup raised a $20M seed fund in February 2022 from Jio Platforms and South Korean internet conglomerate Naver. “Jio has been one of our key partners for a long time and has invested in us from the very beginning,” said Pranav Mistry, the founder of TWO, in an exclusive interaction with AIM. Another product from TWO is Geniya which can browse data from the internet using Google, rivalling Perplexity AI. Mistry said that Geniya is still in public beta and users can try it out, following the official launch soon.","excerpt":"The Dual2L model enhances Llama 3.1’s capabilities by expanding support to more languages, including Korean, Gujarati, Tamil, Malayalam, and Japanese, with a 5-10% improvement in multilingual performance for the 70B model.","categories":["AI News"],"tags":["ai announcements","Pranav Mistry","Sutra","TWO AI"],"author_name":"Siddharth Jindal","publish_date":"2024-07-25T10:59:02","publication_year":"2024","word_count":404,"keywords":["Go","API","TWO AI","AI","innovation","Pranav Mistry","Scala","RAG","ai announcements","Aim","Sutra","llm_models:Llama","R","startup"],"extracted_tech_keywords":["AI","Aim","RAG","R","Go","Scala","API","innovation","startup","llm_models:Llama"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/two-announces-sutra-dual2l-a-multilingual-ai-model-using-llama-3-1\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10102288,"title":"Andrew OG of AI","content":"There are various AI experts and the numbers are only rising each day. Every single one of them has a different view on the development and consequences of the rapid development of AI. But there is one expert, who is undoubtedly the OG when it comes to learning AI, and also calling out the problems — Andrew Ng. The Google Brain founder and the brains behind machine learning teachings at Stanford University, Ng recently called out the big-tech narrative about AI doomsday, and how that is just about controlling open source models, which are the biggest threat to the companies. In a recent interview, Ng said, “The bad idea that AI could make us extinct” is being merged with the “bad idea that a good way to make AI safer is to impose burdensome licensing requirements. When you put those two bad ideas together, you get the massively, colossally dumb idea of policy proposals that try to require licensing of AI.” Definitely, this is also visible in Biden’s executive order for AI safeguard, which is imposing restrictions right now on AI companies, which are most going to target AI startups. Moreover, the regulations also require special licensing and permissions for developers using models outside of the US. “It would crush innovation,” Ng added. Apart from teaching at Stanford, co-founding Google Brain, and being the chief scientist at Baidu AI group, Andrew Ng is also the co-founder of Coursera, along with Daphne Koller. In fact, his group was one of the first at Stanford to advocate for the use of GPUs in deep learning. The AI guru If you are following the recent AI developments, you are sure to have stumbled upon Andrew Ng’s generative AI courses. Not just one, the professor has been launching a new course in generative AI almost every single week, helping people land the AI job they want. He also launched an AI Fund of $175 million in 2018. According to DeepLearning.AI, Andrew Ng has been teaching the most number of students on the planet, that too without a university. Ever since he has been teaching courses, people on X and HackerNews alike, have been praising how Andrew Ng is the only person to listen to when it comes to AI. “Based and Ng-pilled,” says one of the posts, and the others say, “He’s a great professor. Hard to take questions elegantly when your class size is like 700 but he manages it.” Interestingly, Sam Altman, the CEO of OpenAI, has also been one of the interns of Andrew Ng at Stanford. But lately, the views of both the disciple and the teacher have been different when it comes to regulating AI. This is mostly because one benefits from teaching people about it, and the other is trying to build his business in the field and stay on top of it. “He interned with me,” said Ng in the interview, “I don’t want to talk about him specifically because I can’t read his mind, but… I feel like there are many large companies that would find it convenient to not have to compete with open-sourced LLMs.” Thoughtful regulations is the answer Though it is not like Andrew Ng does not side with regulation. Unlike Altman and other big-tech founders who signed the letter “mitigating the risk of extinction from AI should be a global priority,” Ng believes that AI should be thoughtfully regulated. “I don’t think no regulation is the right answer, but with the direction regulation is headed in a lot of countries, I think we’d be better off with no regulation than what we’re getting,” Ng said. He agrees that AI has caused harm to the world in some or the other ways such as self-driving cars and the crash of the stock market, but it should be thoughtful. Undoubtedly, another open source champion, Yann LeCun, the head of Meta AI, agrees with Andrew Ng. https:\/\/twitter.com\/ylecun\/status\/1719162768002093184 But even though one of the Godfathers of AI, LeCun might be on the side of thoughtful regulations and open source AI, his other counterparts, Geoffrey Hinton and Yoshua Bengio, have been on a spree of giving ammunition to the big-tech lobbying for regulation. It is clear that not all AI experts think alike. While the debate goes on about the existential risks of AI, one thing is for sure that we can learn from Andrew Ng’s courses about how to build AI. Then it is up to us to figure out how to build a responsible model, or what some ethicists call — aligned AI. Thoughtful regulation is the answer. Meanwhile, Andrew Ng also said that AI has an Instagram problem. “I’m here to say: Judge your projects according to your standard, and don’t let the shiny objects make you doubt the worth of your work!”","excerpt":"Andrew Ng has been teaching the most number of students on the planet, that too without a university","categories":["AI Features"],"tags":["Andrew Ng","Interviews and Discussions"],"author_name":"Mohit Pandey","publish_date":"2023-10-31T15:02:19","publication_year":"2023","word_count":799,"keywords":["Go","API","Meta AI","machine learning","OpenAI","AI","Andrew Ng","deep learning","generative AI","GAN","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","machine learning","deep learning","generative AI","OpenAI","Meta AI","R","Go","API","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/andrew-og-of-ai\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10143786,"title":"Advance Auto Parts Opens Innovation Hub in its Hyderabad GCC","content":"Advance Auto Parts India (India Innovation Center) has unveiled its new Innovation Hub, to drive transformative change and develop groundbreaking solutions in the automotive aftermarket industry. The Hyderabad GCC serves as an innovation hub, delivering essential transformational support and automotive solutions in IT, digital, finance, and human resources for its operations in North America. The inauguration event was attended by top executives, including Kristen Soler, chief human resources officer; Bruce Starnes, III, chief merchandising officer; and Aaron Surasky, senior VP of merchandising operations. Designed as a dynamic workspace, the hub aims to be more than just a collaborative environment. It is envisioned as a space where creativity converges with execution, enabling teams to deliver impactful, end-to-end solutions. Acknowledging the contributions of the Innovation Team, the company highlighted their vision and dedication in bringing this initiative to fruition. Advance Auto Parts India expressed optimism about the hub’s potential to inspire teamwork, fuel new ideas, and set a benchmark for excellence in the industry. In 2022, US-based Advance Auto Parts announced its entry into the Indian market with the inauguration of its GCC in Hyderabad. The 65,000 sq. ft., 500-seater facility marked the company’s official entry into India and aims to drive innovations in supply chain optimisation, competitive pricing, store operations, and customer engagement. The Hyderabad office houses over 800 employees and more than 200 partner resources.","excerpt":"In 2022, US-based Advance Auto Parts announced its entry into the Indian market with the inauguration of its GCC in Hyderabad.","categories":["AI News"],"tags":["GCC"],"author_name":"Mohit Pandey","publish_date":"2024-12-18T09:06:41","publication_year":"2024","word_count":225,"keywords":["GCC","programming_languages:R","AI","innovation","Git","Aim","ViT","R"],"extracted_tech_keywords":["AI","Aim","R","Git","ViT","innovation","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/advance-auto-parts-opens-innovation-hub-in-its-hyderabad-gcc\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10163089,"title":"Tech Mahindra to Set Up 300-Employee GCC for Goodyear in Hyderabad","content":"Tech Mahindra is in advanced discussions to establish a global capability centre (GCC) for Ohio-based Goodyear Tire & Rubber Co in Hyderabad, Mint reported, citing sources familiar with the matter. The proposed centre, expected to employ 300 people, will manage Goodyear’s research and development (R&D) and IT operations. However, Tech Mahindra did not confirm if the firm is setting up a GCC for Goodyear. The move aligns with Tech Mahindra’s broader strategy under CEO Mohit Joshi, who took over in December 2023 after the retirement of CP Gurnani. As part of its three-year turnaround plan, Project Fortius, the company aims to boost operating margins, accelerate revenue growth, and eliminate unprofitable tail accounts. Tech Mahindra, which reported $6.3 billion in revenue for FY24, will allocate about 0.2% of its 150,488 employees to manage Goodyear’s GCC. It reported a 92.6% year-on-year (YoY) increase in net profit for the third quarter ended December 31, 2024, reaching $115 million, with a total revenue of revenue for the quarter rising to $1.56 billion. Meanwhile, Goodyear closed 2023 with $20 billion in revenue. Indian IT firms like Tech Mahindra often facilitate GCC setups by staffing and managing these centres, receiving payments either for the workforce supplied or as a share of revenue. In recent news, Infosys is establishing a GCC in India for Lufthansa as part of a renewed IT contract valued at nearly $300 million, according to reports. This initiative underscores the growing adoption of build-operate-transfer (BOT) models in major contract renewals. Read: Telangana and Andhra Pradesh, the New Hotspot for IT and GCCs IT firms typically transition GCC operations back to the parent company after a few years, but it remains unclear whether Tech Mahindra and Goodyear have similar plans. An earlier version of this article said the GCC would employ 3,000 people. The figure has now been updated, based on the original source.","excerpt":"Tech Mahindra, which reported $6.3 billion in revenue for FY24, will allocate about 0.2% of its 150,488 employees to manage Goodyear’s GCC.","categories":["AI News"],"tags":["Tech Mahindra"],"author_name":"Mohit Pandey","publish_date":"2025-02-10T09:43:14","publication_year":"2025","word_count":310,"keywords":["Go","Tech Mahindra","programming_languages:R","AI","programming_languages:Go","Aim","GAN","R"],"extracted_tech_keywords":["AI","Aim","R","Go","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/tech-mahindra-to-set-up-3000-employee-gcc-for-goodyear-in-hyderabad\/","complexity_score":3,"technical_depth":7,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10131958,"title":"If Only AI of Today Could Write Windows…","content":"Creating software and applications has never been easier. With generative AI tools, these days anyone can create simple apps and get assistance to build bigger software. This has ushered in a series of claims by people that AI is better at coding than humans, making developers worry about losing their jobs. But the reality is quite the opposite. Greg Diamos, the co-founder of Lamini, replied to a post by Francois Chollet, who reiterated that there would be more software engineers in five years than there are today. Diamos asked, “Do you think that AI today could write Windows?” Chollet said that software engineers would be using AI and not be replaced by them. “At least not within that timeline and definitely not via current technology.” This is something Diamos agreed with as well. “I remember a researcher in 2017 showed me an LLM that could write hello.c and fibonacci.c. Today Llama 3.1 can write a calculator app, but Windows is probably out of reach,” wrote Diamos. ‘The Truth is Somewhere in the Middle’ The problem with current AI systems, especially LLMs, is that they lack in reasoning and maths is something probably out of calculation for them. Though in just the last few days, OpenAI, Google, and Meta have all released their models with increased maths capabilities, the capabilities for creating something like Windows or an operating system all by itself are far-fetched. A few weeks back, when Meta launched Llama 3.1, it was touted as the Linux moment of AI that would have a similar impact on the AI ecosystem as Linux had on the operating system world. But the effects are yet to be seen. LLMs, in their current form, are far from building something akin to an operating system. Experts such as Ilya Sutskever and Andrej Karpathy say that LLMs are the OS of the future, but no one has spoken about how LLMs can actually build an OS. Maybe the future of all software is building extensions on top of LLMs, but that seems unlikely in itself as well. The truth is that AI tools need software engineers, and not the other way around. And it is also not about the “boiler-plate” code. It’s still out of reach for an LLM to build something out of scratch, if it has not seen the code for something similar before. Can LLMs build LLMs all by themselves? Probably not right now, but surely in the future with the advent of agentic AI. But as Diamos said, an LLM that was able to create hello.c back in 2017, with current capabilities and correct direction of research could be able to create something beyond a calculator. But for now, a calculator cannot create a calculator or do anything other than arithmetic. No Idea of the Reality for Now This is not a new discussion, though. Meta AI chief Yann LeCun disagreed with Sutskever and possibly with Karpathy’s assessment as well. LeCun believes, “Large language models have no idea of the underlying reality that language describes.” Chollet had cleared the air during a previous discussion as well. “Language can be thought of as the *operating system* of the mind. It is not the *substrate* of the mind – you can think perfectly well without language, much like you can still run programs on a computer without an OS (albeit with much more difficulty).” LLM OS, on the other hand, is an interesting idea. Maybe LLMs do not need to build a ‘Windows’. Over time, similar to how Windows or Mac OS access various applications to accomplish a task, an LLM OS would be able to access various tools, which would ideally be other LLMs, for problem-solving. But as of now, creating an OS involves not only writing code in various languages (like C, C++, and assembly) but also designing and managing complex systems such as memory management, file systems, device drivers, user interfaces, security protocols, and networking. This process requires specialised knowledge in computer science and software engineering, along with a deep understanding of hardware and system architecture. While LLMs can assist in generating code snippets, explaining concepts, and providing guidance, the creation of a full-fledged ‘Windows’ would require significant human expertise, collaboration, and effort beyond the capabilities of current LLMs. But OpenAI is already teasing us with the release of Project Strawberry, which Sam Altman claims has achieved Level-2 AI, which means human-level reasoning capabilities. Possibly, we are almost there.","excerpt":"LLMs, in their current form, are far from building something akin to an OS. But what about LLM as an OS?","categories":["AI Features"],"tags":["Lamini AI","Windows"],"author_name":"Mohit Pandey","publish_date":"2024-08-09T11:27:16","publication_year":"2024","word_count":740,"keywords":["Lamini AI","Go","agentic AI","Meta AI","OpenAI","AI","RPA","Windows","Aim","C++","generative AI","R"],"extracted_tech_keywords":["AI","generative AI","agentic AI","OpenAI","Meta AI","Aim","R","Go","C++","RPA"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/if-only-ai-of-today-could-write-windows\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10147822,"title":"RBI Forms Panel to Regulate AI in Financial Sector","content":"The Reserve Bank of India (RBI) has launched a high-level committee to establish a Framework for Responsible and Ethical Enablement of Artificial Intelligence (FREE-AI) in the financial sector. The initiative was announced alongside the December Monetary Policy Statement, underscoring the central bank’s commitment to balancing technological innovation with ethical governance. Chaired by Pushpak Bhattacharyya, professor at IIT Bombay, the committee comprises notable figures from academia, government, and industry. Members include Debjani Ghosh, distinguished fellow at NITI Aayog and ex-NASSCOM President; Balaraman Ravindran, Head of IIT Madras’ Wadhwani School of Data Science and AI; and Abhishek Singh, Additional Secretary at the Ministry of Electronics and IT. Industry representatives from Microsoft India, HDFC Bank, and Trilegal also contribute their expertise. The committee’s scope covers banks, non-banking financial companies (NBFCs), payment system operators (PSOs), and fintech entities. It will assess AI adoption in financial services domestically and globally, analyze regulatory frameworks, and identify risks associated with AI. Recommendations will include risk evaluation, mitigation strategies, and compliance frameworks to uphold ethical standards. Key guidelines will be proposed for the responsible integration of AI models in India’s financial infrastructure, particularly catering to the growing fintech ecosystem. The RBI’s FinTech Department will provide secretarial support, and the committee is expected to deliver its report within six months of its inaugural meeting. Through consultations with industry stakeholders and experts, the FREE-AI initiative aims to address the opportunities and challenges of AI, setting a benchmark for its ethical and sustainable use in the financial sector.","excerpt":"Chaired by Pushpak Bhattacharyya, professor at IIT Bombay, the committee comprises notable figures from academia, government, and industry.","categories":["AI News"],"tags":["AI in finance","rbi"],"author_name":"Mohit Pandey","publish_date":"2024-12-26T16:10:32","publication_year":"2024","word_count":247,"keywords":["data science","Go","artificial intelligence","programming_languages:R","AI","innovation","programming_languages:Go","AI in finance","rbi","Aim","R"],"extracted_tech_keywords":["AI","artificial intelligence","data science","Aim","R","Go","innovation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/rbi-sets-up-free-ai-panel-for-regulating-ai-in-financial-sector\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10161808,"title":"Can AI Help Us Achieve Work-Life Balance?","content":"The struggle to define work-life balance shows no sign of resolution. From Infosys founder NR Narayana Murthy advocating for a 70-hour workweek to the Larsen & Toubro (L&T) chairman SN Subrahmanyan suggesting 90-hour workweeks, the pressure to dedicate more time to work can seem relentless. Amid these discussions, the entry of artificial intelligence (AI) raises an important question: Can AI help us achieve a sustainable work-life balance? Recently, Bhavish Aggarwal, CEO of OLA, added a fresh twist to the conversation when he said on X, “Training AI to work like me so that I can find work-life balance!” Is Aggarwal hinting at the potential of AGI, or is he underscoring the role of AI in alleviating the burdens of multitasking? New Definition Work-Life Balance In an exclusive conversation with AIM, Ankush Sabharwal, founder of CoRover, emphasised the need to rethink work-life balance. “Maintaining harmony between work and personal life is crucial, but it’s time to redefine what we mean by ‘work-life balance’. Rather than striving for a rigid separation between work and life, or simply working less or more, we should aim for synergy and harmony.” Sabharwal highlighted how AI has the potential to revolutionise productivity and well-being by automating routine tasks, thereby freeing up valuable time for high-impact activities. He mentioned that AI can provide personalised insights, which can enable individuals to make data-driven decisions to prioritise schedules and goals effectively. Additionally, AI fosters enhanced productivity by allowing people to focus on work that aligns with their personal values and purpose. “With AI, we can redefine what it means to be productive and fulfilled. It’s no longer about juggling multiple tasks or sacrificing personal time. Instead, AI provides actionable insights, enabling us to prioritise what truly matters, nurture relationships, and enhance overall well-being,” Sabharwal added. AI will not Replace Humans As organisations increasingly integrate AI into their operations, the technology is proving to be a powerful enabler rather than a replacement for human judgment. By automating repetitive tasks and streamlining processes, AI empowers teams to focus on meaningful work and foster a healthier work-life balance. Apurv Agrawal, CEO and co-founder of SquadStack, highlighted this transformative impact and told AIM, “At Squadstack, AI handles tasks such as agent training, task assignments, and performance management, empowering our agents to focus on personal growth. By automating administrative tasks, we’ve created a healthier work-life balance for our teams,” Praveer Kochhar, co-founder of Kogo Tech Labs, shared his organisation’s vision of realising human potential with the help of AI. “Internally, we are integrating automation and agentic workflows to streamline processes. By March, we aim to be the first company in India to adopt a four-day workweek, with remaining tasks managed by automated workflows,” he revealed. Kochhar also cited a LinkedIn poll that showed saving time as the most significant advantage offered by AI. Source: LinkedIn However, some companies like Accenture and Zoho declined to comment on the matter. What’s Next? The potential of AI to enhance personal productivity has also been a focal point for influencers. Last year, Priyank Ahuja, a self-proclaimed career guru, shared on X, “ChatGPT and Canva will help you earn an extra $15,000\/month.” In a conversation with AIM, Ahuja explained that his claim was backed by feedback from professionals who use tools like Canva daily for tasks such as designing T-shirts, creating Instagram ads, and making YouTube Shorts. These tools, he explained, are powered by AI and enable individuals to work smarter, not harder. Source: X This sentiment resonates with young creators like 19-year-old artist Ashok Reddy, who, during a conversation with AIM, shared that he sold around 100 AI-generated art pieces in just two days. His breakdown was selling A4-size pieces for INR 400 and A3-size pieces for INR 700. Notably, he has gathered all the knowledge required for this through online resources without attending any college courses. He recommended using Midjourney and highlighted the inspiration and learning opportunities provided by platforms like Twitter and other online communities. The mantra “work smart, not hard” suggests that AI has the potential to reshape our approach to productivity and help us save time.","excerpt":"By automating repetitive tasks and streamlining processes, AI empowers teams to focus on meaningful work and foster a healthier work-life balance.","categories":["AI Features"],"tags":["AI (Artificial Intelligence)","job","work life balance"],"author_name":"Vidyashree Srinivas","publish_date":"2025-01-20T14:59:50","publication_year":"2025","word_count":682,"keywords":["Go","work life balance","ChatGPT","artificial intelligence","AI","ML","agentic workflows","GPT","Ray","Aim","job","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","ML","ChatGPT","agentic workflows","Aim","Ray","R","Go","GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/can-ai-help-us-achieve-work-life-balance\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10119811,"title":"Leading the Way: DuxData&#8217;s Data Leadership &amp; Strategy Course Prepares Data Pros for the AI Era","content":"In an age where data is considered the new oil, the need for skilled data professionals who can bridge the gap between data science and business strategy has hit an all-time high. Numan Karim identified this need and started DuxData, a pioneering course designed to equip data professionals with the necessary leadership, strategy, and communication skills to navigate the complexities of modern data science and AI integration within organisations. “DuxData is not your typical boot camp. This is not a coding course, nor is it a modelling and algorithm crash course,” Karim emphasised. “Thousands of courses teach technical skills, but very few discuss actually weaving those skills into the business.” The brainchild of a seasoned data professional with over a decade of experience, DuxData was born out of the realisation that the data science boom of the 2010s left many organisations grappling with the challenge of integrating data science effectively into their operations. The Curriculum of DuxData At the heart of DuxData’s curriculum lies the recognition that domain knowledge, communication, and leadership skills are just as crucial as technical proficiency in the data science and AI field. “The mainstream nature of AI means that it’s never been more important for data scientists to have strong business acumen,” Karim emphasises. “If AI is a hammer, then everything looks like a nail. Sometimes the role of a good data scientist is to have the foresight that AI or ML may not be the appropriate solution to a problem,” he said. DuxData’s curriculum is designed to enhance practical application through real-world scenarios and case studies. The topics covered include: Intrapreneurial Data Science Data Science Product Development Measuring the Value of Data Science Initiatives Organisational Data Maturity & Literacy Data Scientists as Change Agents Stakeholder Analysis & Change Management Business Process Mapping Strategic Communication Those who purchase the masterclass will receive two primary benefits in addition to the existing masterclass topics. Lifetime access to all future topics, resources, and digital download\/templates: As the industry grapples to figure out how to implement AI properly, best practices will emerge. DuxData will create resources on these topics on a regular basis and based on community feedback. Access to a private community of data scientists and data leaders within DuxData: This is a forum that will be built from the ground up for professionals to discuss anything and everything related to data science. A Unique Approach to Data Science Education What sets DuxData apart from other data science and AI training programs is its focus on practical business value over technical proficiency. In addition to its comprehensive curriculum applicable to practising data scientists, data analysts, and to-be data leaders, DuxData offers lifetime access to all future topics, resources, and digital downloads\/templates, ensuring that learners stay current with emerging trends in AI and data science. To facilitate learning, each topic is accompanied by video lectures and digital download templates that can be applied directly to real-world business settings. “These are the templates I used in my full-time role as an IC and currently director of data science. I’ve found them valuable at various stages of my career and will continue to develop templates as the masterclass course content grows alongside the AI boom,” said Karim. By focusing on leadership, strategy, and communication skills, DuxData empowers data professionals to bring meaningful changes within their organisations. “Our learners gain leadership capabilities and strategic foresight to lead transformative data science initiatives within their organisations,” said Karim. “Building an understanding of the ‘pull-through’ aspect of data science ensures that data scientists do not simply sit behind the scenes and build models and analysis that drift off into the ether. Instead, they will be equipped with the necessary learning and knowledge to build fit-to-purpose data science and analytics solutions that create measurable value for the organisation,” he added. The Changing Role of Data Scientists “In the 2010s, we witnessed the data science boom, where every company under the Sun, from startups to Fortune 500 giants, aggressively hiring data scientists,” said Karim. “Over the years, when the dust settled, organisations found themselves navigating uncharted territories.” As companies realised the complexity of integrating data science into their existing frameworks, the need for a blueprint for success became evident. “The inability to integrate data science into the existing fabric of the organisation led to the realisation that data science was more complex than anticipated. The blueprints for success were missing; a guide to building capabilities that align with strategic goals was absent,” Karim explained. DuxData serves as a compass, guiding data professionals through the integration of data science into the broader business context. “We believe the fundamental goal of a data scientist is to be valuable and generate insights and efficiencies for the business. This means going beyond the technicals and weaving the transformation into existing business processes, changing behaviours, and creating true disruption in the industry,” said Karim. “What separates mediocre data scientists from great data scientists is the ability to close the gap between data science and execution,” he said. For data professionals looking to advance their careers in the rapidly evolving field, DuxData offers a roadmap for success. “DuxData will equip you with a foundation in leadership, strategy, and communication skills for practical application in a real-world setting,” the founder reiterated. “We envision a community where we can continuously build on those skills together in this dynamic and evolving space.” [Use promo code LAUNCH30 to enjoy 30% off the course. Act fast, as the promo is valid until June 24th.] Check out the course here.","excerpt":"DuxData’s curriculum is designed to enhance practical application through real-world scenarios and case studies.","categories":["AI Highlights"],"tags":["ai dataset","AI leaders"],"author_name":"Mohit Pandey","publish_date":"2024-05-08T09:53:38","publication_year":"2024","word_count":921,"keywords":["data science","Go","API","AI","ML","Git","analytics","disruption","GAN","AI leaders","R","ai dataset"],"extracted_tech_keywords":["AI","ML","data science","analytics","R","Go","Git","API","GAN","disruption"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/leading-the-way-duxdatas-data-leadership-strategy-course-prepares-data-pros-for-the-ai-era\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":41146,"title":"Understanding The Right To Data Portability, Its Limitations,  And Benefits","content":"The right to data portability means to provide individuals with the right to receive personal data which they have provided to a controller in a structured and machine-readable format. It can be said as one of the most important introductions within the EU General Data Protection Regulation (GDPR) both in terms of warranting control rights to data subjects and in terms of being found at the intersection between data protection and other fields of law. The right to data portability can be applied under two circumstances: When the lawful basis for processing the information is either consent or for the performance of a contract When an individual is carrying out the processing by automated means (i.e. excluding paper files). The new version of Article 20, GDPR describes the right to data portability as mentioned below 1| The data subject shall have the right to receive the personal data concerning him or her, which he or she has provided to a controller, in a structured, commonly used and machine-readable format and have the right to transmit those data to another controller without hindrance from the controller to which the personal data have been provided, where (a) the processing is based on consent pursuant to point (a) of Article 6(1) or point (a) of Article 9(2) or on a contract pursuant to point (b) of Article 6(1); and (b) the processing is carried out by automated means. 2| In exercising his or her right to data portability pursuant to paragraph 1, the data subject shall have the right to have the personal data transmitted directly from one controller to another, where technically feasible. 3| The exercise of the right referred to in paragraph 1 of this Article shall be without prejudice to Article 17. That right shall not apply to process necessary for the performance of a task carried out in the public interest or in the exercise of official authority vested in the controller. 4| The right referred to in paragraph 1 shall not adversely affect the rights and freedoms of others. In a recent conference, the Future Of Privacy Forum (FPF) discussed the limits and benefits of the right to data portability as introduced by the GDPR. Introduction to data portability can be said as one of the greatest innovations by GDPR where one should be able to transfer his\/her personal data between participants in the market. Limitations Of Right To Portability The right to data portability only covers personal data which is basically “provided” by the person to an organisation. The data which is not protected is the data that are the byproducts of services such as when a data controller uses an algorithm and processes data including inferences from data. The development manager of Microsoft who was in the panel discussion pointed out that there are three difficulties with data portability in practice: Syntactic (is the data an integer, string, floating, or something else?) Semantic (for example, if data references a “Jaguar” is it discussing a car or an animal) Policy-related (how does it interact with existing regulations and contractual requirements for these companies). Outlook Security has always been a key concern but it should not be used as an excuse for porting data. According to the panel discussion, ideally, each data controller would have the GDPR already as a starting point of compliance, and as such, it would have applied all protective portions such as transparency, lawfulness, etc. Watch the full discussion below:","excerpt":"The right to data portability means to provide individuals with the right to receive personal data which they have provided to a controller in a structured and machine-readable format. It can be said as one of the most important introductions within the EU General Data Protection Regulation (GDPR) both in terms of warranting control rights […]","categories":["AI Features"],"tags":[],"author_name":"Ambika Choudhury","publish_date":"2019-06-22T07:59:52","publication_year":"2019","word_count":576,"keywords":["Go","programming_languages:R","AI","innovation","programming_languages:Go","RAG","GAN","R"],"extracted_tech_keywords":["AI","RAG","R","Go","GAN","innovation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/understanding-the-right-to-data-portability-its-limitations-and-benefits\/","complexity_score":4,"technical_depth":8,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":39424,"title":"Corporate Affairs Ministry’s Plan To Integrate AI In MCA 21 Portal Will Make Business Registration And Compliance Easier","content":"The Ministry of Company Affairs recently announced that the MCA21 e-Governance service would integrate AI into its operations. The initiative is said to make compliance processes easier while optimizing routine activities. The MCA 21 portal is an application that allows new businesses to complete registration procedures easily. It was an initiative undertaken by the MCA in 2006, and is now the preferred method for company registration, compliance and conflict resolution. Injeti Srinivas, the secretary of the MCA, stated in the ministry’s monthly newsletter that the AI integration would be a part of the version 3 of the MCA 21 portal. He also mentioned that this would be undertaken by Bengaluru-based software giant, Infosys. He went on to reveal that the introduction of AI would allow for a single source of truth to be maintained. This would remove the need for known details to be filled multiple times. As a part of the initiative, Srinivas also mentioned that various databases will be interlinked. The AI will be put in charge of routine enforcement processes, bringing 24×7 monitoring of the systems. The MCA recently came under fire as the government questioned the use of MCA 21 data as a part of GDP calculations. The National Sample Survey Organisation found that 39% of all companies in the MCA 21 were not traceable, or classified wrongly. Dismissing these concerns, the MCA stated that portal reported and checked compliance in real time. This has led to the eradication of shell and bogus companies, the MCA stated to a publication. The ministry further revealed that the portal has an automated KYC drive program that eradicated non-performing or non-functional companies. The MCA 21 portal has already employed automation to a certain level, with the upcoming AI integration set to optimize processes further. The MCA 21 portal is also in dire need of a UX upgrade; something that the integration of smart features promises.","excerpt":"The Ministry of Company Affairs recently announced that the MCA21 e-Governance service would integrate AI into its operations. The initiative is said to make compliance processes easier while optimizing routine activities. The MCA 21 portal is an application that allows new businesses to complete registration procedures easily. It was an initiative undertaken by the MCA […]","categories":["AI News"],"tags":["Government"],"author_name":"Anirudh VK","publish_date":"2019-05-20T06:04:29","publication_year":"2019","word_count":316,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","Government","automation","ViT","GAN","R"],"extracted_tech_keywords":["AI","R","Go","GAN","ViT","automation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/corporate-affairs-ministrys-plan-to-integrate-ai-in-mca-21-portal-will-make-business-registration-and-compliance-easier\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10114737,"title":"Andrew Ng Unveils New Free Course on Llama 2 with Meta","content":"A recent addition to the DeepLearning.AI course catalogue is Meta’s “Engineering with Llama 2.” This course allows you to explore prompt engineering using the company’s Llama 2 models. The course is tailored for beginners, requiring just one hour. Led by instructor Amit Sangani, senior director of partner engineering, Meta, the course is currently available for free for a limited time. You can sign up for the course here. Participants will gain insights into the best practices associated with prompting Llama 2 models, focusing on practical applications. The curriculum encourages interaction with three key models: Meta Llama 2 Chat, Code Llama, and Llama Guard. These models serve distinct purposes, ranging from conversation and coding assistance to content moderation through Llama Guard. The primary learning objectives of the course involve familiarizing oneself with the Llama 2 collection, adopting best practices in prompt engineering, and building applications using these models. Through a simple API call, participants will explore the diverse outputs of the Llama 2 models, gaining a nuanced understanding of their capabilities. The course instructs participants on leveraging Llama 2 models as personal assistants, providing guidance for day-to-day tasks. Additionally, it delves into advanced prompt engineering techniques such as few-shot prompting for sentiment analysis and chain-of-thought prompting for logical problem-solving. Code Llama is introduced as a collaborative partner for pair programming, facilitating both learning and code improvement. The course also throws light on the responsible use of LLMs by incorporating Llama Guard, which screens user prompts and model responses for potentially harmful content. Importantly, participants are informed that Llama 2 models and their weights are available for free download, including quantized versions for local machine deployment. The course also encourages involvement in an active open-source community that utilizes Llama 2 for building diverse applications.","excerpt":"This course allows you to explore prompt engineering using the company’s Llama 2 models.","categories":["AI News"],"tags":["Andrew Ng","Courses","LLaMA"],"author_name":"Shritama Saha","publish_date":"2024-02-29T12:11:01","publication_year":"2024","word_count":292,"keywords":["API","TPU","programming_languages:R","AI","sentiment analysis","LLaMA","RAG","Andrew Ng","prompt engineering","GAN","llm_models:Llama","Courses","R"],"extracted_tech_keywords":["AI","RAG","prompt engineering","sentiment analysis","TPU","R","API","GAN","llm_models:Llama","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/andrew-ng-unveils-new-free-course-on-llama-2-with-meta\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":1420,"title":"IvyCamp facilitates IoT Startup Tyre Express to raise seed funding","content":"Tyre Express – an IoT startup Founded in November 2015 by Dinesh Wakale and Brijesh Shukla, Tyre Express Pvt. Ltd. is an enterprise IoT solutions startup. Dinesh and Brijesh are alumni from S P Jain Institute of Management and Research and Said Business School, respectively. Tyre Express is supported by Microsoft’s BizSpark programme. The startup is in news, as it raised an undisclosed amount in its seed round from IIT Delhi and Wharton alumni. The fundraiser was facilitated by IvyCamp. IvyCamp is an innovation and entrepreneurship platform by IvyCap Ventures. The platform connects entrepreneurs with mentors, investors, and incubation centers, through the global alumni networks of institutes such as IITs, IIMs, ISB, and BITS Pilani. IvyCamp has previously helped startups raise funds across several industry verticals, such as food delivery startup Mr Hot Foods, payments firm FTCash, and furniture rental startup Grabonrent. Founders of Tyre Express The Ivy Camp platform helped the personality development app, Leadburg raise fund, only last week. The funding was raised from former SAP veteran Harsh V. Pant, and OpsVeda Asia president and COO Ratheesh Raveendran. Adhiraj Banerjee, Executive Director, IvyCamp, remarks “IvyCamp focuses closely on the deep tech and IoT verticals. Many more investment transactions will grace the landscape, as new sensor technologies and information infrastructure development continue to augment the IoT startup scenario in India.” Tyre Express is developing IoT-based solutions for vehicle diagnostics. The firm has developed a platform which tracks and monitors the performance of tyres in real-time for fleet operators. Besides, a fleet operator can enhance tyre life, improve fuel efficiency, reduce incidence of accidents, and bolster the overall operational efficiency for the fleet, leveraging the platform. Tyre Express is collaborating with tyre manufacturers to enhance its service offerings to institutional clients.","excerpt":"Founded in November 2015 by Dinesh Wakale and Brijesh Shukla, Tyre Express Pvt. Ltd. is an enterprise IoT solutions startup. Dinesh and Brijesh are alumni from S P Jain Institute of Management and Research and Said Business School, respectively. Tyre Express is supported by Microsoft’s BizSpark programme. The startup is in news, as it raised […]","categories":["AI News"],"tags":[],"author_name":"Дарья","publish_date":"2017-02-07T08:32:29","publication_year":"2017","word_count":291,"keywords":["funding","programming_languages:R","AI","data_tools:Spark","innovation","RAG","R","startup"],"extracted_tech_keywords":["AI","RAG","R","innovation","startup","funding","programming_languages:R","data_tools:Spark"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/ivycamp-facilitates-iot-startup-tyre-express-raise-seed-funding\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10164559,"title":"Groq Aims to Provide At Least Half of the World’s AI Compute, Says CEO","content":"When a company like NVIDIA becomes the preferred option for most AI labs for hardware and rises to the status of the world’s biggest company, does it imply that the GPU maker excels in every domain? Probably not. American AI infrastructure provider Groq once shared an amusing comment in a blog post. “GPUs are cool for training models, but for inference, they’re slowpokes, leading directly to the great-model-that-no-one-uses problem.” The company is well on its way to beating NVIDIA in providing inference – a crucial process in which a pre-trained AI model applies its learnings to generate outputs. Groq’s language processing unit (LPU) offers capabilities specific to AI inference in ways much better than a traditional graphics processing unit (GPU). In a podcast interview with venture capitalist Harry Stebbings, Groq CEO Jonathan Ross explained that instead of relying on external memory like GPUs do, LPUs keep all the model parameters directly within their chips. “Imagine you were trying to build a factory, and it was only 1\/100th of the size needed for the assembly line,” said Ross in an analogy, indicating how GPUs operate. This would lead to the factory repeatedly processing small batches, dismantling the setup, and restarting the process over and over. In contrast, Ross said LPUs allow computation to flow smoothly through thousands of chips simultaneously, eliminating inefficiencies and significantly improving the speed. While more chips are being used, LPUs have a significantly less energy consumption footprint than GPUs. Owing to this, Ross said, “We [Groq] need to be one of the most important compute providers in the world. Our goal by the end of 2027 is to provide at least half of the world’s AI inference compute.” Moreover, last year, NVIDIA CEO Jensen Huang said that one of the major challenges NVIDIA currently faces is generating tokens at incredibly low latency. However, in no way is Groq’s mission supposed to be misunderstood – they’re not competing with NVIDIA. ‘I Think NVIDIA Will Sell Every Single GPU They Make for Training’ At first glance, Ross’ statements and the recent events surrounding NVIDIA might suggest that the Taiwanese giant is in trouble, but it isn’t. Both Groq and any inference solutions provider will co-exist with NVIDIA. “Training should be done on GPUs,” Ross said. “I think NVIDIA will sell every single GPU they make for training.” Ross added that if Groq were to deploy high volumes of lower-cost inference chips, the demand for training would increase. “The more inference you have, the more training you need, and vice versa,” he said. Moreover, Ross said Groq contemplates selling its LPUs as a “nitro boost to GPUs”. The company experimented with running portions of a model on an LPU and the rest on the GPUs. This speeds up the process and makes the GPUs run much more economically. Having said that, Ross and the company don’t really view NVIDIA as a competitor. “They [NVIDIA] don’t offer fast tokens and low-cost tokens. It’s a very different product, but what they do very well is training, and they do it better than anyone else,” he said. The demand for GPUs will not end despite the advent of a growing AI inference provider market. “How are you going to do the training?” Ross asked. “Buy the GPUs. Get every single one you can,” he said. Not Without Competition That said, Groq competes with several other inference service providers. Most notably, Cerebras and SambaNova, also based in the United States, offer hardware products that directly target NVIDIA’s dominance. Recently, Perplexity AI and Mistral AI announced the integration of Cerebras Inference into their products. The latter calls its app ‘Le Chat’ – the fastest AI assistant in the world. On the other hand, SambaNova is the only inference provider among the trio that is capable of handling the Llama 3.1 405B model. Groq, on the other hand, no longer sells its AI inference hardware, and its proprietary technology can be accessed on the cloud platform via different models on GroqCloud. The platform hosts multiple third-party models, including ones developed by Alibaba (Qwen), Meta (Llama), and DeepSeek (R1). Moreover, Groq has announced that it is available on OpenRouter.ai, a platform that offers a unified interface for accessing numerous AI models. Groq now allows users to use DeepSeek-R1 distilled on Meta’s Llama 70B with 1,000 tokens per second. Recently, Saudi Arabia announced a $1.5 billion investment in Groq to expand AI infrastructure in the region. The funding builds on Groq’s previous work in the region, including the rapid deployment of the largest AI inference cluster in the Middle East in December 2024.","excerpt":"For Groq, it isn’t about overtaking NVIDIA but co-existing with it.","categories":["Global Tech"],"tags":["AI (Artificial Intelligence)","Groq"],"author_name":"Supreeth Koundinya","publish_date":"2025-02-25T17:28:25","publication_year":"2025","word_count":765,"keywords":["Go","API","funding","TPU","Cerebras","SambaNova","AI","venture capital","Groq","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","TPU","R","Go","API","venture capital","funding","Groq","Cerebras","SambaNova"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/groq-aims-to-provide-at-least-half-of-the-worlds-ai-compute-says-ceo\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10072798,"title":"Infosys’ AI &#038; Analytics Play","content":"As one of India’s largest technology services providers, Infosys is at the forefront of technology innovation and is leading the way for multiple AI and analytics solutions for its customers across the globe. In an exclusive interaction with AIM, Balakrishna D R, EVP – service offering head for energy, utilities, communications services and AI and automation services at Infosys, shared details of select projects in AI and analytics, alongside sharing in-house initiatives, future of work, scaling AI investments and more. At Infosys, Balakrishna heads the AI and automation unit. He has been with Infosys for over two decades, heading sales, program management, and delivery roles across geographies and industry verticals. Balakrishna said that the company has over 1,000 ready-use cases across industries, projects, technologies, and service lines in AI\/ML. He believes that the biggest challenge enterprises face today is not the lack of technology but identifying the right use cases and delivering significant value. “You need to understand your business, identify where AI can create the maximum impact, and then make sure you can handle all aspects to bring that use-case to life,” said Balakrishna. AI and analytics use cases Infosys is redefining the tennis experience for fans and players using data analytics, AI\/Ml, and AR\/VR. The company generated metrics using data like spin, speed, placement, and position, which provided insights to improve decision-making. In addition to this, Infosys provided an immersive experience through 3D Court Vision and 3D Art Museum of key events and digital exhibitions associated with Tennis. It also provided instant access to key moments in a match, including crowd sentiment, statistical data, and other ML-powered insights, which can be useful for journalists. Stat of the day by @Infosys : 4⃣,8⃣3⃣ is the average rally length between Roger Federer and Rafael Nadal ! #RGxInfosys #RG19 pic.twitter.com\/t47eHvMOha— Roland-Garros (@rolandgarros) June 7, 2019 Infosys’ AI-powered voice assistants brought a global audience close to the Roland-Garros through popular smart home devices, such as Amazon Alexa, serving information to live radio broadcasts and event podcasts. Besides this, Infosys’ autonomous system platform is driving India’s first-ever project on autonomous buggies—a commercially viable solution for controlled environments. This technology helps in auto-braking and provides navigation features equipped with advanced LiDAR and vision technologies with AI and a deep learning engine to detect objects, lanes, and curves for navigation. This buggy is used in industrial areas, education campuses, airport environments, and amusement parks. Infosys Autonomous System Platform, in partnership with @MainiGroup, will drive India’s first commercial autonomous buggies. Know more. https:\/\/t.co\/TE0gSuujxW #autonomousnavigation pic.twitter.com\/NfgfpNvSIj— Infosys (@Infosys) January 29, 2021 In energy-as-a-service, Infosys and the integrated energy company, bp recently announced a partnership where they co-developed a digital platform that collects data from multiple energy sources and uses AI to optimise the supply and demand for heat, power, cooling, and electric vehicle charging. Infosys said that, here, the key outcome was to provide 100 per cent clean, cost-effective, optimum, and reliable energy (electricity and green fuel) with access to monitor and manage the consumption pattern while creating smart and energy-efficient infrastructure. The Energy as a Service (EaaS) partnership between bp & Infosys will accelerate #energytransition for cities, campuses and industries. #EaaS will be the foundation of our path to achieving large-scale #decarbonization. More about the partnership: https:\/\/t.co\/W1ioMyRyD9@akdash1— Nandan Nilekani (@NandanNilekani) October 29, 2021 In addition, Infosys is looking to amplify data and AI with the cloud. “More companies, including our own team, are now seeing the benefits of cloud when it comes to catalysing digital transformation—because of its ability to increase development and processing speed while providing near-limitless scale,” said Balakrishna. He said the Finnish Postal service, ‘Posti Group’ is one of the oldest companies in Finland and presents an inspiring instance. In collaboration with Infosys, Posti replaced its legacy systems and processes while incorporating artificial intelligence and machine learning—moving to the cloud to amplify their full-scale digital transformation. One of the outcomes was to retain happier employees through gamification. AI and analytics as a service At Infosys, AI and automation are vital deliverables that they offer, including development, maintenance, testing, migration, modernisation, implementation, rollout and more, across all projects, technologies, and service lines. This leads to scaling enterprise, productivity improvement, customer satisfaction, amplification of the capabilities of people to take faster decisions, and new avenues of business. For example, in application support, Infosys leverages LEAP (the live enterprise application platform), Infosys Cognitive Automation Studio, Infosys Intelligent Assistant, and Polycloud to analyse the data, predict failures, and take proactive actions to prevent incidents. The company’s automation solutions include self-healing and self-service with zero-touch automation. Its Cortex platform is being used to reimagine the contact centre operation significantly. Infosys follows the principles of DevSecOps and distributes agile to deliver faster time to market and ensure seamless collaboration between RUN and Project organisation. “We have developed the Infosys DevOps platform (IDP), and many of our clients have adopted this,” said Balakrishna. For application development, Infosys uses multiple low code\/no code platforms like Mendix, Deep Code, Power Apps, Appian, and others. “We have also created Infosys Digital Foundry that can create apps, set up the CI\/CD pipeline, and deploy and host the solution quickly,” added Balakrishna. For migration projects, Infosys leverages its automation platforms like Infosys Modernization Suite. This leverages neural machine translation for database migration. The team has been using this for multiple migration programs. “Our model is getting better with every implementation,” said Balakrishna. In terms of testing, Balakrishna said that they have reimagined their entire testing services with Infosys Test Automation Framework, Infosys Test Data Management Suite, and Infosys Data Testing Workbench. “We are also leveraging an AI-based platform ‘Pandit’ for testing quality assurance,” he added. AI strategies Balakrishna told AIM that Infosys applied AI is a comprehensive strategy for end-to-end customer requirements of an enterprise. He said that it takes care of the roadmap to scaling enterprise-grade AI for their intelligent automation. It converges the power of AI, automation, analytics, DevOps, and cloud to deliver new solutions and perceptive experience. In addition, Infosys ensures that AI and automation are embedded into every fabric of their services like cloud and infrastructure management, development, maintenance, testing, migration, modernisation, implementation, rollout, and more across all project types, technologies, and service lines. “Our strategy is to achieve this within our customers’ existing landscape using the tools that customers already have and introduce new tools only when required,” said Balakrishna. He said that their Infosys Center for Emerging Technology Solutions (ICETS) and central research and development teams contribute significantly toward intelligent automation. ICETS is their incubation centre with resources working on emerging technology and building solutions relevant to client needs. The team undertakes a structured approach of experimenting with the technology and creating POCs and solutions using such emerging technology. Infosys’ central research and development team focuses on technologies like cloud, low code\/codeless development\/workflow, DevOps, event streaming, intelligent document processing, and task\/process mining. It creates solutions wherein its developers work only on core business logic that requires human cognition; everything else should be templatised, standardised and automated, bringing productivity improvement. “Our approach to automation is multi-dimensional, covering DevSecOps—build, test and release automation through CI, CT, and CD, automation in application maintenance for handling job abounds, batch failures and data errors, auto triaging, RPA-based process automation, and regression suite automation,” explained Balakrishna. AI Living Labs AI Living Labs is a program led by ICETS, which helps create innovative solutions for clients by contextualising emerging technologies. The team leverages the Infosys innovation ecosystem’s design thinking, technology, and design capabilities to incubate and deploy at scale. Balakrishna said Living Labs offers innovation as a service to Infosys’ clients by bringing together business context, technology and design expertise. “We partner with clients to enable entire organisations to reimagine their businesses and to shape the organisations through emerging technology-led innovations. This is characterised by rapid iterative prototyping, creating real experiences,” he added. [In partnership with @Infosys #InfosysCobalt] The Global Cloud Ecosystem Index 2022 cloud leaders—Singapore, Finland, and Sweden–foster digital service adoption through a holistic focus on physical infrastructure and regulatory clarity. Explore the data: https:\/\/t.co\/7ql5rFnLaY. pic.twitter.com\/gTBfBuuZ3i— MIT Technology Review (@techreview) May 31, 2022 Future of work “Our vision is to have digital workers work in tandem with the human workforce,” said Balakrishna. He said that their digital worker should be able to connect the real world with the virtual world in which they understand the intent, respond to questions, and take action from humans, but leaving the decision and control with the humans. Balakrishna said that their patented ‘Microbot Framework’ is a step towards their digital vision that involves the creation of digital workers and a workforce that can perform high-end tasks like humans and is platform agnostic. Infosys believes in a four-pronged strategy of scale agile digital, energise the core, reskill, and expand location to strengthen relevance to clients and drive accelerated value creation for the future workplace. Some of the initiatives for a futuristic workplace are Live Enterprise@Infosys, which senses, feels, and responds in real-time; a mobile-first approach via the Infy Me mobile app, with 250+ features, and a modern, hybrid, and secure workplace that bring together technologies like borderless ODCs, virtual collaboration tools via Infosys Meridian that enable a remote-first workplace that mirrors the offline experience. Using AI Internally Citing Infy Me, Balakrishna said that the response has been phenomenal. He said that they have seen many of their key processes become faster and more responsive with a hyper-personalisation experience with behavioural trends to make all the key transactions available to the user on the go. “After the internal success, we are also seeing interest for the platform among our clients as seven clients have already been onboarded and many more are in discussions,” he added. Infosys uses Infosys Fraud Detector (machine learning-based solution), which provides technical assistance to teams such as HR recruitment, background verification, and others. In confirming the genuineness of applicant profiles received by verification and validation, this reduces the manual time significantly. Infosys also uses platform-agnostic AI and RPA solution repositories to ease and enhance multiple human capital management (HCM) processes, including resume and job application filtering using AI. “Using this, we let go of the manual work of filtering hundreds of resumes received for interviews,” he added. As part of new technology adoption, Infosys has successfully modernised its enterprise storage platform. This platform uses the latest storage disk technology, which drives enhanced performance by up to 10X compression and deduplication advantages, along with a data availability guarantee. “This initiative delivered power savings of 46 per cent for this landscape,” shared Balakrishna. Moreover, its ‘Infrastructure as code’ initiative delivered 1,200+ playbooks for automating platform-related processes across the hybrid cloud. Infosys believes it resonates with employee needs and responds with a value proposition that delivers meaning, purpose, and value for them, thereby reducing attrition. In addition, the company ensures that its people are continuously learning and create opportunities for every employee to navigate the future through some of the initiatives that it has launched. This includes LeX (for anywhere, anytime learning), iRise (their reward and recognition tool), Skill Tags (for proficiency in different technologies), Digital Quotient (to keep track of their digital capabilities) and many more. Edge over others “As an industry-leading service provider, it is our constant endeavour to create solutions that support our clients on their digital transformation journey. As the demand for AI continues to rise, we look toward creating future-proof solutions that will address the problems of tomorrow,” said Balakrishna. For instance, as part of their Infosys Cobalt offering, Infosys recently launched their applied AI cloud. This solution provides its clients access to AI hardware, open-source AI software as a service on their hybrid cloud infrastructure, and harnesses edge AI capabilities. “Through the AI cloud, our developers and project teams can easily access our clients’ AI hardware and software across private and public clouds and develop contextualised services that deliver AI-first business processes for them,” said Balakrishna. Further, he said their clients could continue harnessing their data estates, open-source data and curated data exchanges on the cloud to develop and train their AI models. “Additionally, they can avail services delivered by any hyperscale cloud provider to scale and future-proof their AI-powered transformation, enabling them to build an advantage in a competitive business ecosystem,” said Balakrishna.","excerpt":"At Infosys, AI and automation are vital deliverables that are offered, including development, maintenance, testing, migration, modernisation, implementation, rollout and more.","categories":["IT Services"],"tags":["Infosys"],"author_name":"Amit Naik","publish_date":"2022-08-16T12:00:00","publication_year":"2022","word_count":2040,"keywords":["machine learning","artificial intelligence","Infosys","AI","ML","RAG","Aim","deep learning","analytics","edge AI","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","analytics","Aim","edge AI","RAG","R"],"url":"https:\/\/analyticsindiamag.com\/it-services\/infosys-ai-analytics-play\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10052993,"title":"Biggest AI Innovations And Milestones Of 2021","content":"“Remember to celebrate milestones as you prepare for the road ahead” — Philanthropist Nelson Mandela Artificial Intelligence, or AI, is an ever-evolving field. And the occasional failures here and there should not stop us from hyping the great advancements. In the last year and half, despite the global crisis– in some cases, because of the global crisis– scientists, researchers and developers have made insane contributions, innovated and have reached unprecedented milestones in the field of AI. As we head closer to the end of 2021, Analytics India Magazine takes a look back at the year that was, and the AI innovations and milestones of this year that made it to the headlines. Facebook SEER Early this year Facebook AI developed SEER (SElf-supERvised)– a billion-parameter self-supervised computer vision model. The model can learn from any random group of images on the internet, without having to carefully curate and label the images, which is otherwise a prerequisite for computer vision training. So far, the team at Facebook AI has tested SEER on one billion images– uncurated and unlabelled, on publicly available Instagram images. Reportedly, it performed better than most advanced self-supervised systems. This breakthrough clears the path for flexible, accurate and adaptable computer vision models for the future. Alphabet Isomorphic Lab Google’s parent company Alphabet introduced Isomorphic Lab, in an attempt to accelerate the discovery of new drugs using AI. Its subsidiary DeepMind’s founder and CEO, DEmis Hassabis, announced the creation of the lab to ultimately find cues for some of humanity’s “Most devastating diseases.” Alphabet plans to develop a computational platform to better understand the biological systems and find ways to treat diseases. Although separate, DeepMind and Isomorphic intend to occasionally collaborate to build off the research, discoveries and protein structure work. Vertex AI Earlier this year, Google Cloud announced the availability of Vertex AI at the Google I\/O event. It is a managed machine learning (ML) platform for the deployment and maintenance of AI models. Vertex AI brings AutoML and AI platforms together into an unified API, and can be used to build, deploy and scale ML models faster. Deployed by Google Research, Vertex AI required 80 per cent fewer lines of code for custom modeling. It integrated with open-source frameworks including TensorFlow, PyTorch and Scikit-learn. MusicBERT Microsoft developed a large scale pre-trained model for symbolic music understanding— MusicBERT. The model can understand music from symbolic data, that is, in MIDI format and not audio; and then indulge into genre classification, emotion classification, and music pieces matching. The tech giant used OctupleMIDI method, bar-level masking strategy and large-scale symbolic music corpus of more than one million music tracks. MusicBERT achieves state of the art performance on music understanding tasks and going ahead, the team at Microsoft attempts at applying the model on tasks including structure analysis and chord recognition. GitHub Copilot AI company OpenAI and Microsoft collaborated to launch AI programmer Copilot in July this year. Based on OpenAI Codex, the new AI system is trained on open-source code, contextualising a situation using docustrings, function names, preceding code and comments to determine and generate the most relevant code. The GitHub Copilot is trained on billions of lines of public code, putting the knowledge one needs at their fingertips while saving one’s time and helping them stay focused. It works on a broad set of frameworks as well as languages including TypeScript, Ruby, Java, Go and Python. Tensorflow 3D Google developed and launched TensorFlow 3D— a modular library to bring 3D deep learning capabilities to TensorFlow, earlier this year. This latest upgrade gives access to sets of operations, loss functions, models for the development, training and deployment of 3D scene understanding models, and data processing tools and metrics. TensorFlow 3D supports datasets including Waymo Open, Rio and ScanNet. It supports three pipelines— 3D Semantic Segmentation, 3D Instance Segmentation and 3D Object Detection.","excerpt":"We take a look back at the year that was, and the AI innovations and milestones of 2021 that made it to the headlines.","categories":["AI Features"],"tags":["AI (Artificial Intelligence)","AI innovations","DeepMind","Facebook","GitHub","Google","Microsoft","Tensorflow"],"author_name":"Debolina Biswas","publish_date":"2021-11-06T12:13:14","publication_year":"2021","word_count":641,"keywords":["artificial intelligence","machine learning","Microsoft","AI","OpenAI","TensorFlow","ML","PyTorch","Tensorflow","computer vision","deep learning","analytics","Google","Facebook","GitHub","AI innovations","AI (Artificial Intelligence)","DeepMind"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","computer vision","analytics","OpenAI","TensorFlow","PyTorch"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/biggest-ai-innovations-and-milestones-of-2021\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10135613,"title":"‘Next Token Prediction’ Might Make PyTorch-like Frameworks Redundant","content":"LLMs are becoming the default solution for many of the problems that businesses and researchers face. Even when it comes to domains outside language and text, people have been experimenting with LLMs for guessing and predicting the next token, which sparks an interesting conversation around the need for other tools like PyTorch, as LLMs can just do the whole job in the future. Interestingly, according to Andrej Karpathy, these models are more like tools designed to predict the next piece of a sequence, whether that’s words, images, or other types of information. This next token prediction framework can be a universal tool for solving a wide variety of problems, beyond text. “If that is the case, it’s also possible that deep learning frameworks (e.g. PyTorch and friends) are way too general for what most problems want to look like over time,” said Karpathy. It's a bit sad and confusing that LLMs (\"Large Language Models\") have little to do with language; It's just historical. They are highly general purpose technology for statistical modeling of token streams. A better name would be Autoregressive Transformers or something.They…— Andrej Karpathy (@karpathy) September 14, 2024 LLMs are not really just “language experts” anymore. According to Karpathy, The “language” part has become historical because these models were first trained to predict the next word in a sentence, but in reality, they can work on any kind of data that’s broken down into little pieces, called tokens. Think of LLMs like a super-smart guessing game. This means, for example, if you’re building a car, a house, or an animal using Legos, you’re just putting blocks together, LLMs are like that, they don’t care if the tokens (blocks) represent words, images, or even molecules—they just focus on predicting what the next block should be based on what’s already there. Another example can be Protein prediction models like AlphaFold and ESMFold, which are built on top of generative language models. Calling such intricate models just LLMs seems to be unjust for what they are capable of. “What I’ve seen though is that the word “language” is misleading people to think LLMs are restrained to text applications,” said Karpathy in another thread. “I don’t think this is true but I think it’s half true” – Karpathy Probably, the name should change. “Definitely needs a new name. ‘Multimodal LLM’ is extra silly, as the first word contradicts the third word,” replied Elon Musk. Meanwhile, Yann LeCun is more concerned about why this doesn’t make sense for all the types of problems. “It only works with discretized outputs (discrete symbols) and only makes sense with symbol sequences with a natural order (not images). Text, DNA, proteins, musical scores, etc. are discrete or easily discretized,” said LeCun. For something like images, which are continuous and don’t naturally have a strict sequence of discrete symbols (each pixel doesn’t follow a clear ‘order’ like text), LLMs don’t work as naturally. To use an LLM for images, you would first need to somehow convert the image into discrete chunks (like dividing the image into small patches), but this doesn’t follow the same natural order that exists in text or DNA. Agreeing with Karpathy, and a little with LeCun, Gary Marcus said that statistical modelling of token streams works well if reasoning or planning isn’t required. Last month, Eliezer Yudkowsky also said that predicting the next token can solve almost all the well-posed problems. “literally any well-posed problem is isomorphic to ‘predict the next token of the answer’,” he said. “Throwing an LLM at it” The idea that many problems can be reduced to a token-stream prediction model is intriguing, especially since domains like images, audio, and even molecules can be broken down into sequences of tokens. This suggests that a unified approach like LLMs could handle diverse tasks, reducing the need for highly specialised architectures, such as PyTorch. However, frameworks like PyTorch provide more than just flexibility in creating neural network models. They allow for a variety of deep learning operations that aren’t necessarily relevant for LLMs but are critical for other areas like reinforcement learning, generative models, and non-sequential tasks. While it’s true that LLMs could dominate many applications, not every problem is best framed as “next token prediction.” We may see a simplification or specialisation of deep learning frameworks to accommodate the increasing dominance of LLM-based models. Still, the complete redundancy of frameworks like PyTorch might be too extreme of a prediction. Probably, OpenAI’s newest model o1 gives a sense of why LLMs would be able to solve a lot of problems outside of the realm that is currently considered achievable. With the reasoning tokens in place, the model can go beyond just ‘predicting’ the next token, and giving reasons for why it did so. PyTorch and similar frameworks may not become redundant but could evolve to become more focused on token-based models, while still offering tools for more diverse problems outside that paradigm. Though, currently only in language or text format, the capabilities might extend beyond it soon. Calling LLMs as LLMs might be underrepresenting their capabilities. Moreover, “it just predicts the next token” is a thought-terminating cliche.","excerpt":"“It just predicts the next token” is a thought-terminating cliche.","categories":["AI Features"],"tags":["Pytorch"],"author_name":"Mohit Pandey","publish_date":"2024-09-17T13:34:54","publication_year":"2024","word_count":855,"keywords":["Pytorch","Go","TPU","OpenAI","AI","neural network","PyTorch","RPA","Transformers","deep learning","R"],"extracted_tech_keywords":["AI","deep learning","neural network","OpenAI","PyTorch","Transformers","TPU","R","Go","RPA"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/next-token-prediction-might-make-pytorch-like-frameworks-redundant-in-the-future\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":39196,"title":"IIT Delhi Aims For Global Excellence, Advances Research In AI","content":"After winning the Institute of Eminence tag last year, IIT Delhi now hopes to increase its research and development capabilities in the coming years to further solidify its position in R&D. With the IoE tag, the government will grant ₹1,000 crore over the course of the next year. Meanwhile, IIT Delhi hopes to raise as much ₹500 crore as external R&D funding. Even though the institute has raised ₹400 crore in the last two years for R&D funding, it now hopes to put more focus on areas that are industry focused or larger issues that could help the society. Thus, by the increased focus to lead a good quality R&D, it hopes to turn itself into a world-class educational institute. As per the latest report, IIT Delhi presently has 500 on-going research. Speaking about the development, B R Mehta, R&D Dean, IIT Delhi said, “The R&D activity of IIT-D has shown huge growth in the last two years because of the focus on research in thematic areas of interest. As a result, all the research projects that we undertake are now either driven by industry or by what the society needs.” In order to bolster research capabilities, in April the institute announced that it will backup startups working in the field of AI, ML and other emerging tech areas to boost its research ecosystem. Stating that there aren’t enough deep-tech startups in the country, the institute announced that it will rope-in startups in the field for the purpose of connected intelligent systems, mixed reality, advanced materials, drug discovery and medical research. By doing so, the institute aims to increase the research.","excerpt":"After winning the Institute of Eminence tag last year, IIT Delhi now hopes to increase its research and development capabilities in the coming years to further solidify its position in R&D. With the IoE tag, the government will grant ₹1,000 crore over the course of the next year. Meanwhile, IIT Delhi hopes to raise as […]","categories":["AI News"],"tags":["latest advances"],"author_name":"Akshaya Asokan","publish_date":"2019-05-15T07:51:46","publication_year":"2019","word_count":271,"keywords":["Go","funding","programming_languages:R","AI","ML","latest advances","programming_languages:Go","Aim","ViT","R","startup"],"extracted_tech_keywords":["AI","ML","Aim","R","Go","ViT","startup","funding","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/iit-delhi-aims-for-global-excellence-advances-research-in-ai\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":38928,"title":"Mobile Premier League Taps Into AI &#038; Analytics To Corner The $943 Mn Gaming Market In India","content":"Mobile Premier League (MPL) is one of the fastest growing eSports platforms in India today. Launched in September 2018, the platform has garnered over 25 million users in a short span of time and aims to break barriers of access to make mobile gaming a household sport across the country. Offering a one-stop platform with multiple, popular, skill-based games and esports to choose from, MPL is extensively using AI, ML and analytics to stay ahead of the game. The company currently made headlines when it announced  Series A investment of $35.5 million, led by Sequoia India, Times Internet, and GoVentures with participation from other investors including RTP Global, BeeNext, Base Growth, and Venture Highway. Aiming to use the funding amount for product development and user growth in India, the founders are working every step to become India’s largest Mobile eSports Platform which is accessible to all. The company is currently 80 people strong and aims to grow exponentially. Analytics India Magazine interacted with Sai Srinivas Kiran G (SSK), co-founder and CEO, to get an inside picture of MPL and Kaustubh Bhoyar (KB), CTO, to understand the technological drive behind the company’s unprecedented success. Analytics India Magazine: Tell us about the inception of the MPL platform. SSK: The idea to start MPL came around April end, last year and we started off with a bunch of us from my apartment and then by June we finally moved to an office setup. It was only in September ’18 that we finally launched the product. It has been an interesting 7 months since then and the moment we have 25 million users on our platform. AIM: eSports has become a well-established worldwide phenomenon. Where does India stand? SSK: Mobile Gaming has been gaining momentum in the Indian market with the proliferation of mobile phones and affordable data plans. Annual mobile gaming revenue in India is projected to grow to $943 million in 2022, making India the fastest-growing mobile market in the world by overall revenue, according to a report by POKKT. India is now one of the top five markets for mobile gaming in terms of the number of users. Any internet business in India would put the Indian market as top 1-3 priorities right now. For us, it is very important because as of today, the number of mobile games users stands at around 200 million and this number is expected to grow up to 600 million in 2-3 years. There’s no other market in the world which is this big in terms of users. AIM: Who are your direct competitors? SSK: Currently, MPL is the only platform offering multiple games and we do not have any direct competitors. While there are multiple apps that involve real money gaming in fantasy sports and a few more, MPL is the only platform offering multiple games of a casual nature like Candy Crush, Temple Run etc. in one place. AIM: How do you plan to utilise the $35.5 million of funding? SSK: Our aim is to make the product more robust and bring in different formats of gameplay in which the users can engage. The most important thing is to invest in growth. The idea is to grow as aggressively and as fast as possible. We plan to invest in products and user growth in India. Also, we will use the current funding to grow faster by hiring the right talent to build out the right product infrastructure which will suit not just the Indian market but also the global market. Img Src: MPL AIM: Why do you think emerging technology like AI, ML, and analytics is needed in an organisation? KB: Artificial Intelligence adds a component of flexibility and precision to automation. It helps companies optimise productivity levels and amplify the existing computing power. Machine Learning, a subset of AI, enables a machine to learn its functions and adapt based on the changing business needs. Machine Learning gives crucial insights into large industry data for companies to stay updated on the latest trends. With respect to gaming, analysing millions of hours of player data can provide insight into which elements of the game are popular. Brands can increase user engagement through analytics, revealing when players abandon a game or other detailed insights. This data can be used to identify bottlenecks and issues within gameplay. AI has been an integral part of gaming and, we have also been closely watching the impact of other emerging technologies in the gaming industry. We want to become a technology-driven platform that offers great gaming experience for our users who are on different skill levels. AIM: When did MPL adopt these and what have been its benefits to date? KB: Right from the beginning, we were certain on a few technologies which couldn’t be compromised on. We have been investing heavily on the latest technologies which have helped us strengthen the product and the technology. The AI-powered tool has also helped in quick creation and incorporation of all gaming elements. AIM: How is AI crucial to MPL? Could you explain with a use case? KB: Data and payment security play an important role for us and as a critical part of our verification process, we ensure the users complete their KYC check on the platform before withdrawing their winnings. Since our launch, we have seen massive installs and right now we have around 25 million odd users on our platform. All our verification tasks are made easier by Verizy. Verizy’s proprietary algorithm simplifies the first level screening seamlessly by reducing the amount of time otherwise spent manually on verification. Along with this, we also work with Verloop, an AI-based chatbot that helps reduce our turnaround time by simplifying the elementary level exchange between our users and our support team. Once the concern is identified it is seamlessly handed off to our customer support team. It’s been four months since we got the services by Verizy and Verloop on board and all our verification processes have been made seamless by them. AIM: What are the plans to invest in emerging technologies for MPL? KB: Yes, we definitely plan on investing in emerging tech in the near future to manage and scale growth. A strong tech team driven by the latest AI technology will help us in creating, automating and customizing more tournaments on our platform which will cater to a wider audience and have a broader outreach. AIM: What is your roadmap\/plans for esports? SSK: We have grown exponentially in a short time of span. Currently, we have over 25 Million players and host over 30 games, adding more each week. The market potential in India itself is huge and honestly, we have just scratched the surface. We will continue to add more games and formats that our user want. We recently expanded to Indonesia and with this, we are confident of taking MPL to more SEA markets. Investment in people and technologies who will help us achieve these milestones will be our focus for the next few months.","excerpt":"Mobile Premier League (MPL) is one of the fastest growing eSports platforms in India today. Launched in September 2018, the platform has garnered over 25 million users in a short span of time and aims to break barriers of access to make mobile gaming a household sport across the country. Offering a one-stop platform with […]","categories":["AI Features"],"tags":["Interviews and Discussions","MLP"],"author_name":"Srishti Deoras","publish_date":"2019-05-10T10:10:47","publication_year":"2019","word_count":1170,"keywords":["Go","artificial intelligence","machine learning","AI","ML","Aim","MLP","ViT","analytics","GAN","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","analytics","Aim","R","Go","GAN","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/mobile-premier-league-taps-into-ai-analytics-to-corner-the-943-mn-gaming-market-in-india\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10022720,"title":"Yann LeCun’s Latest Offering &#8211; Barlow Twins","content":"Self-supervised learning (SSL) has become a handy technique in computer vision tasks. Significant advances in SSL means its methods can now learn representations even if the input samples are distorted. Also referred to as ‘data augmentations’, this is made possible by maximising the similarity of representations extracted from different distorted versions of a sample. However, there is a slight hitch. This approach introduces trivial constant representations. Currently, most methods avoid collapsed solutions by employing careful implementation details. To that end, Yann LeCun and his team have introduced Barlow Twins. This objective function avoids collapse by measuring the cross-correlation matrix between the output of two identical networks fed with distorted versions. The aim here is to minimise the redundancy between the vector components and make the outputs as close as possible to the identity matrix. Trivial Representations Self-supervised learning has proved to be a good solution for deep learning systems’ excessive data dependency. LeCun, referred to as one of the Godfathers of deep learning and the inventor of convolutional neural networks, first gave a glimpse of self-supervised learning in 2018 during his keynote speech at AAAI conference. Self-supervised learning helps in creating data-efficient artificial systems. This method learns valuable representations of the input data without relying on human annotations. Self-supervised learning has major applications in the field of natural language processing. As discussed, data augmentation also leads to trivial representations. In the past, there have been several attempts to overcome this problem, including: Contrastive methods define positive and negative sample pairs that are treated differently in the loss function. Clustering methods use one distorted sample to measure ‘targets’ for the loss. Another distorted version of the sample is used to predict these targets. It is then followed by the application of an alternate optimisation scheme such as K-means.BYOL and SIMSIAM are among the more recent methods. In both methods, network architectures and parameters updates are modified to introduce asymmetry. Barlow Twins LeCun and his team now proposed a new method called Barlow Twins. Named after neuroscientist H.Barlow, this method draws heavily from his influential 1960 article, titled ‘Possible Principles Underlying the Transformation of Sensory Messages’, which notes that sensory processing’s goal is to recode redundant sensory input data into code with a statistically independent component, also called factorial code. Barlow Twins method applies redundancy reduction, similar to Barlow’s one in his article, to self-supervised learning. As used in the Barlow Twins method, the principle of redundancy reduction has proved successful in explaining the visual system’s organisation and led to the introduction of several algorithms for supervised and unsupervised learning. Barlow Twins is conceptually simple, easy to implement and learns useful representations. With this method, the researchers propose an objective function that makes the cross-correlation matrix computed from twin representations to be very close to the identity matrix. Barlow Twins benefits from the use of very high-dimensional representations. Barlow Twins operates on a joint embedding of distorted images. It produces two distorted views for all the images of a batch sampled from the dataset obtained through the distribution of data augmentation. These two batches of distorted views are fed to a deep network with trainable parameters, producing batches of representations. Credit: Barlow Twins The advantages of the methods include: It doesn’t require large batches, asymmetric mechanisms like prediction networks, momentum encoders, stop gradients or non-differentiable operators. Wrapping Up Barlow Twins have outperformed previous state-of-art methods for self-supervised learning with the added advantage of being simpler and avoiding trivial representations. It is also on-par with the current ImageNet classification methods with linear classification head and a number of other classification and object detection methods. Researchers believe that further algorithm refinement could open doors for more effective solutions. Read the full paper here.","excerpt":"Self-supervised learning (SSL) has become a handy technique in computer vision tasks. Significant advances in SSL means its methods can now learn representations even if the input samples are distorted. Also referred to as ‘data augmentations’, this is made possible by maximising the similarity of representations extracted from different distorted versions of a sample. However, […]","categories":["AI Features"],"tags":["self supervised learning"],"author_name":"Shraddha Goled","publish_date":"2021-03-23T13:00:00","publication_year":"2021","word_count":619,"keywords":["Go","TPU","AWS","AI","neural network","computer vision","Aim","deep learning","object detection","R","self supervised learning"],"extracted_tech_keywords":["AI","deep learning","neural network","computer vision","Aim","object detection","AWS","TPU","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/yann-lecuns-latest-offering-barlows-twin\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10042098,"title":"Guide to Autoencoders, with Python code","content":"An autoencoder is an Artificial Neural Network used to compress and decompress the input data in an unsupervised manner. Compression and decompression operation is data specific and lossy. The autoencoder aims to learn representation known as the encoding for a set of data, which typically results in dimensionality reduction by training the network, along with reduction a reconstruction side is also learned. Data specific means, autoencoder will only be able to compress the data on which they have trained, e.g. if the autoencoder is trained on images of dogs, it will give a poor performance on images of cats. Here lossy operation can be explained as when you share an image on WhatApp, the quality of uploaded\/shared image is degraded, in the same way, reconstruction side gives the output. From the below image, watch the quality of the reconstructed image and original image carefully. Fig1: Schematic of autoencoder The autoencoder is a specific type of feed-forward neural network where input is the same as output. As shown in the above figure, to build an autoencoder, we need an encoding method, decoding method and loss function to compare the output with the target. First, the input passes through the encoders, which are nothing but fully connected artificial neural networks that produce the further code decoder with a similar structure like ANN, producing output using the same code. Here code is nothing but the compressed version of the input. Code implementation: Autoencoders are in the same way as conventional ANN trained through backpropagation. We are mainly going to cover three autoencoder i,e Simple autoencoder Deep CNN autoencoder Denoising autoencoder For the implementation part, we are using a popular MNIST digits data set. Simple Autoencoder: Import all the dependencies from keras.layers import Dense,Conv2D,MaxPooling2D,UpSampling2D from keras import Input, Model from keras.datasets import mnist import numpy as np import matplotlib.pyplot as plt Build the model, here the encoding dimension decides by what amount the image will compress, lesser the dimension more the compression. encoding_dim = 15 input_img = Input(shape=(784,)) # encoded representation of input encoded = Dense(encoding_dim, activation='relu')(input_img) # decoded representation of code decoded = Dense(784, activation='sigmoid')(encoded) # Model which take input image and shows decoded images autoencoder = Model(input_img, decoded) Build the encoder model decoder model separately so that we can easily differentiate between input and output # This model shows encoded images encoder = Model(input_img, encoded) # Creating a decoder model encoded_input = Input(shape=(encoding_dim,)) # last layer of the autoencoder model decoder_layer = autoencoder.layers[-1] # decoder model decoder = Model(encoded_input, decoder_layer(encoded_input)) Compile the model with Adam optimizer and cross entropy loss function, fitment autoencoder.compile(optimizer='adam', loss='binary_crossentropy') (x_train, y_train), (x_test, y_test) = mnist.load_data() autoencoder.fit(x_train, x_train, epochs=15, batch_size=256, validation_data=(x_test, x_test)) encoded_img = encoder.predict(x_test) decoded_img = decoder.predict(encoded_img) Using the plot function, you can see the output for encoded and decoded images, respectively as below. Deep CNN Autoencoder: As the input is images, it makes more sense to use Convolutional Network; the encoder will consist of a stack of Conv2D and max-pooling layer, whereas the decoder consists of a stack of Conv2D and Upsampling layer. model = Sequential() # encoder network model.add(Conv2D(30, 3, activation= 'relu', padding='same', input_shape = (28,28,1))) model.add(MaxPooling2D(2, padding= 'same')) model.add(Conv2D(15, 3, activation= 'relu', padding='same')) model.add(MaxPooling2D(2, padding= 'same')) #decoder network model.add(Conv2D(15, 3, activation= 'relu', padding='same')) model.add(UpSampling2D(2)) model.add(Conv2D(30, 3, activation= 'relu', padding='same')) model.add(UpSampling2D(2)) model.add(Conv2D(1,3,activation='sigmoid', padding= 'same')) # output layer model.compile(optimizer= 'adam', loss = 'binary_crossentropy' model.summary() Output: Model: \"sequential\" _________________________________________________________________ Layer (type)                 Output Shape              Param # ================================================================= conv2d_17 (Conv2D)           (None, 28, 28, 30)        300 _________________________________________________________________ max_pooling2d_7 (MaxPooling2 (None, 14, 14, 30)        0 _________________________________________________________________ conv2d_18 (Conv2D)           (None, 14, 14, 15)        4065 _________________________________________________________________ max_pooling2d_8 (MaxPooling2 (None, 7, 7, 15)          0 _________________________________________________________________ conv2d_19 (Conv2D)           (None, 7, 7, 15)          2040 _________________________________________________________________ up_sampling2d_7 (UpSampling2 (None, 14, 14, 15)        0 _________________________________________________________________ conv2d_20 (Conv2D)           (None, 14, 14, 30)        4080 _________________________________________________________________ up_sampling2d_8 (UpSampling2 (None, 28, 28, 30)        0 _________________________________________________________________ conv2d_21 (Conv2D)           (None, 28, 28, 1)         271 ================================================================= Total params: 10,756 Trainable params: 10,756 Non-trainable params: 0 _________________________________________________________________ Resize the images to 28×28 and scale the values between 0 to 1 and fit the model model.fit(x_train, x_train, epochs=15, batch_size=128, validation_data=(x_test, x_test)) Here are the input images and decoded images are given by the CNN based Autoencoder Denoising autoencoder: Let’s check whether the autoencoder can deal with noise in images, noise in the sense of Bluray images, white marker on the images changing the color of images, etc. Now here we are introducing some noise to our original digits, then we will try to recover those images by the best possible result. Introduce noise as below noise_factor = 0.7 x_train_noisy = x_train + noise_factor * np.random.normal(loc=0.0, scale=1.0, size=x_train.shape) x_test_noisy = x_test + noise_factor * np.random.normal(loc=0.0, scale=1.0, size=x_test.shape) x_train_noisy = np.clip(x_train_noisy, 0., 1.) x_test_noisy = np.clip(x_test_noisy, 0., 1.) Here is some example of noisy images plt.figure(figsize=(20, 2)) for i in range(1, 5 + 1): ax = plt.subplot(1, 5, i) plt.imshow(x_test_noisy[i].reshape(28, 28)) plt.gray() ax.get_xaxis().set_visible(False) ax.get_yaxis().set_visible(False) plt.show() You can see that we barely identify digits, intentionally introducing more noise so as to check up to what extent autoencoder can recover the image. Modify the layers of the above-defined model, such as increase the filter so that model can perform at best and fit the model model.fit(x_train_noisy, x_train, epochs=15, batch_size=128, validation_data=(x_test_noisy, x_test)) pred = model.predict(x_test_noisy) Plot function plt.figure(figsize=(20, 4)) for i in range(5): # Display original ax = plt.subplot(2, 5, i + 1) plt.imshow(x_test_noisy[i].reshape(28, 28)) plt.gray() ax.get_xaxis().set_visible(False) ax.get_yaxis().set_visible(False) # Display reconstruction ax = plt.subplot(2, 5, i + 1 + 5) plt.imshow(pred[i].reshape(28, 28)) plt.gray() ax.get_xaxis().set_visible(False) ax.get_yaxis().set_visible(False) plt.show() Endnotes: We have seen the structure of autoencoders and practically realised some basic autoencoders. There is a wide range of applications of autoencoders such as Dimensionality reduction image compression, a recommendation system and so on. Here we have trained our model for a few epochs; by increasing the epochs, we can boost the performance and also by increasing the dimension of our network. References: Link for Google Colab NotebookKeras official documentation","excerpt":"The autoencoder is a specific type of feed-forward neural network where input is the same as output.","categories":["Deep Tech"],"tags":["Artificial Neural Network","Convolutional Neural Network","encoder-decoder","Guide","keras python","MNIST"],"author_name":"Vijaysinh Lendave","publish_date":"2021-06-21T12:00:00","publication_year":"2021","word_count":975,"keywords":["NumPy","TPU","Keras","AI","neural network","MNIST","Artificial Neural Network","encoder-decoder","Colab","Ray","Aim","Convolutional Neural Network","keras python","Matplotlib","R","Guide"],"extracted_tech_keywords":["AI","neural network","Aim","Ray","Keras","Colab","NumPy","Matplotlib","TPU","R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/guide-to-autoencoders-with-python-code\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10086729,"title":"From Non-Profit to For-Profit: How OpenAI Plans to Make Money","content":"In 2019, the non-profit AI research startup OpenAI turned a page that would alter its bedrock. Unusual though it may be for a non-profit organisation to make millions of dollars, OpenAI announced the OpenAI LP as an entirely separate entity, calling it a ‘capped-profit’ corporation. While the company would continue to research and develop new technologies, it also wanted to make more money in the process. we will have to monetize it somehow at some point; the compute costs are eye-watering— Sam Altman (@sama) December 5, 2022 Profitable, with a clause The details of the deal were this – investors of the startup could now earn up to 100 times their investment and not more than that value. The remaining money made would go right back into their nonprofit work done by another entity called OpenAI Nonprofit. OpenAI counts a bunch of big names among its VC firms like Andreessen Horowitz, Sequoia, Tiger Global and Founders Fund, all of which have the option to exit the agreement whenever they want. Coincidentally, the company’s decision to turn into a profit-making entity also was right before the release of its text-to-image generator DALL.E. The tool is the precursor to DALL.E 2, a far more refined version, which was released last year prompting a wave of generative AI research and startups. The change was a departure from OpenAI’s ethos. Four years prior to this, OpenAI’s initial charter had stated that the startup’s ‘primary fiduciary duty is to humanity’ and that its goal was to ‘build value for everyone rather than shareholders’. Taking the middle road helped OpenAI kill two birds with one stone – the company still retained credibility as a pure research company while being able to attract investors. OpenAI’s shift to a for-profit model was also easily explained by the high cost that AI research needed to run compute for training and testing models. How does OpenAI make money? As with any company that takes this route, OpenAI is now under pressure to generate revenue for its investors instead of sticking to its original, noble mission statement. So, how does OpenAI plan to make money? Understanding the pricing of GPT3 models API In September 2020, the startup revealed the projected pricing plans for its API for users to be able to mostly access the company’s flagship AI tool GPT3, which had been making headlines since its launch. LLMs like GPT3 require hundreds of expensive GPUs and the money OpenAI was making had helped the company make progress in running the models more efficiently, the company later stated. OpenAI also makes a lot of money from enterprises that mostly use fine-tuned GPT3 models as they achieve high levels of performance at a fraction of the cost. Investors and Microsoft deal But the real pot of gold for the company has turned out to be its wildly popular chatbot, ChatGPT. On January 5, WSJ reported that OpenAI was in talks with VC firms Thrive Capital and Founders Fund to sell its existing shares in a tender offer. The tender could total up to at least USD 300 million in OpenAI share sales. The ongoing talks led to OpenAI being valued at around USD 29 billion, making it one of the most valuable US startups despite the little money it was making. Meanwhile, Microsoft, a partner with OpenAI confirmed a multi-year investment on January 23 while investing an additional USD 10 billion into the startup. Microsoft was a natural buyer for OpenAI considering the startup already used Azure for training its models. Under the agreement terms, Microsoft was entitled to 75% of OpenAI’s profits until it earns back its initial investment. Once it reached that limit, Microsoft would have a 49% stake in the startup with other investors taking the other 49% and the nonprofit parent ending up with 2%. When the deal was inked, sources said that OpenAI expects USD 200 million in revenue this year and USD 1 billion by 2024. Subscription fee ChatGPT also brought about the paid subscription option to OpenAI. Last week, the company announced a USD 20-a-month price for its subscribers who would be upgraded from the limited, free service. The plan would initially be rolled out in the US and then slowly expanded to other countries even as OpenAI considers lower-cost subscription options. A UBS report pegged ChatGPT users at an estimated 100 million active monthly users just two months after its launch. The research stated that the chatbot was the ‘fastest-growing consumer application in history’ making it even faster than TikTok and Instagram at user acquisition. And it’s not just us and our friends on ChatGPT, several companies, including digital publisher Buzzfeed, use the chatbot to generate content within seconds. ChatGPT is hailed as a competitor to Google Search, but the revenue models of OpenAI and Google differ as much. Google’s main money-maker is online advertising, which makes up more than 80% of its USD 147 billion according to the tech giant’s annual report. There are many loose ends that must be tied up in the future of this story – ChatGPT might revive Bing as announced by Microsoft. But while Microsoft or OpenAI has no connection with online search and therefore advertising revenue, it has an upper hand over Google owing to its legacy of partnerships with several brands to rest on.","excerpt":"OpenAI also makes a lot of money from enterprises that mostly use fine-tuned GPT3 models as they achieve high levels of performance at a fraction of the cost","categories":["Global Tech"],"tags":["AI Tool"],"author_name":"Poulomi Chatterjee","publish_date":"2023-02-07T15:00:00","publication_year":"2023","word_count":887,"keywords":["Go","ChatGPT","API","OpenAI","AI","R","Git","GPT","generative AI","AI Tool","Azure"],"extracted_tech_keywords":["AI","generative AI","ChatGPT","OpenAI","Azure","R","Go","Git","API","GPT"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/from-non-profit-to-for-profit-how-openai-plans-to-make-money\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10017196,"title":"Genpact Acquires Analytics Firm Enquero To Drive Next-Gen Enterprise Digital Transformation","content":"One of the global professional services firms Genpact announced the acquisition of Enquero, which is an industry-leading data engineering and analytics firm. According to sources, the acquisition will help in accelerating the ability of Genpact in leveraging the data and cloud technologies to drive digital transformation services through advanced analytics. ‘Tiger’ Tyagarajan, chief executive officer of Genpact said, “Today’s announcement increases the scale and depth of our data and analytics teams and further enhances our capabilities to accelerate the digital transformation journeys of our clients.” He added, “I am excited that Arvinder Pal Singh and the very talented Enquero team are joining Genpact at a time when we are seeing strong demand from our clients for exactly these types of solutions to help them navigate and win in challenging markets.” In an environment where organizations across industries and geographies are experiencing unprecedented volatility — from supply and demand challenges to seismic shifts in customer expectations and business models – the smart use of data at the core of an organization will be critical to business growth. As businesses are contending with an ever-increasing volume and complexity of data – external and internal, structured and unstructured, this acquisition will enhance the capabilities as well as extend Genpact’s strong foundation of existing capabilities in delivering end-to-end transformation to enterprise clients. Arvinder Pal Singh, who is the chief executive officer at Enquero stated, “Enquero enables our clients’ strategies to succeed through orchestrating data, process, and technology in a connected digital ecosystem. We started Enquero in 2014 in Silicon Valley to help organizations better leverage data and cloud technologies.” Singh added, “Joining Genpact is the logical next step for us. Genpact’s domain, process and technology leadership, global footprint, scale, and extensive client base will help both Genpact and Enquero, as a combined force, scale our solutions to transform even more organizations.” Singh will continue to lead the business, which will be rebranded Enquero, a Genpact company, and all Enquero employees will join Genpact.","excerpt":"One of the global professional services firms Genpact announced the acquisition of Enquero, which is an industry-leading data engineering and analytics firm. According to sources, the acquisition will help in accelerating the ability of Genpact in leveraging the data and cloud technologies to drive digital transformation services through advanced analytics. ‘Tiger’ Tyagarajan, chief executive officer […]","categories":["AI News"],"tags":["Genpact","genpact analytics"],"author_name":"Ambika Choudhury","publish_date":"2021-01-05T20:13:39","publication_year":"2021","word_count":329,"keywords":["Genpact","AI","programming_languages:R","R","digital transformation","Git","RAG","data engineering","analytics","GAN","genpact analytics"],"extracted_tech_keywords":["AI","analytics","RAG","R","Git","data engineering","GAN","digital transformation","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/genpact-acquires-analytics-firm-enquero-to-drive-next-gen-enterprise-digital-transformation\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":41819,"title":"Google Open-Sources Robot.txt To Help Standardise Robots Exclusion Protocol","content":"Open-sourcing has become a popular practice now. Companies have realised the importance of the feedback that they get from open sourcing their projects and have experienced improvements by doing so. No matter how large the company is, open-sourcing projects brings insights into conceivable refinements. Google has taken a strong stand in standardising the Robots Exclusion Protocol (REP). Now, as a part of these efforts, the search engine giant has open-sourced its Google.txt Parser. Need For REP REP, in simple words, is a standard given to websites to communicate with the web robots. It tells them about the web sections that they should and should not scan and process. But despite the standards, not all web robots follow these standards. Some web robots scan the parts of the website that they have been adviced to not scan. If there is any conflict to any statement by robots.txt by any directive, robot.txt disallows it. The REP was only a de-facto standard for 25 years. This had affected negatively because: Uncertainty for corner cases for webmasters. It introduced uncertainty for crawler and tool developers as well. The REP says that in case of multiple domains, each subdomain must have its own robots.txt file. Crawler directives are responsible to tell the web crawlers the sections of the web that they can crawl. Major search engines follow these. But indexer crawling does not exist and hence crawling of the resource with the indexer directive must be allowed by search engines. Despite this, it is possible to get them for groups of URIs. Google.txt Parser Open-Sourced As a part of its efforts to have an internet standard for a REP, Google open-sourced its Google.txt Parser. Google.txt is a C++ library as old as 20 years. The team at Google uses the REP for the purpose of matching the rules included in robot.txt files. Google in these 20 years upgraded this library to a large extent. It has learned a lot about webmasters writing robots.txt files and corner cases which Google had covered for. The library is hosted at GitHub repository which the community can use to access the work. The main objective of Google.txt Parser making open source is to take help from the community worldwide for the standardization of the REP. By making it open source, the entire community can access, make suggestions and help in these efforts. There is also a testing tool package by Google to help the community test a few rules. This defines all the previously undefined scenarios for robots.txt parsing. With this open-sourced parser from Google, the developer community can create parsers as instructed by the REP. It is to make sure that web robots only scan the parts of the web that they are instructed to scan. Web robots can now use any transfer protocol that is based on URI, instead of HTTPs. https:\/\/twitter.com\/googlewmc\/status\/1145648654415478785 Outlook Google is not commonly known for revealing something from its core research to the open-source world. This is one of the rare times that Google has done it. By this, the open source world can read and listen to robots.txt files available for any crawler or coder. There are some industries that are trying to extend the REM. The Yahoo! Search Blog and the Microsoft Live Search Webmaster team includes wildcard support, sitemaps and extra META tags.","excerpt":"Open-sourcing has become a popular practice now. Companies have realised the importance of the feedback that they get from open sourcing their projects and have experienced improvements by doing so. No matter how large the company is, open-sourcing projects brings insights into conceivable refinements. Google has taken a strong stand in standardising the Robots Exclusion […]","categories":["Global Tech"],"tags":["C","Developers","Google","library","Open Source","Robots"],"author_name":"Disha Misal","publish_date":"2019-07-04T17:32:55","publication_year":"2019","word_count":553,"keywords":["Go","Open Source","C","AI","programming_languages:R","library","programming_languages:C++","programming_languages:Go","Git","Robots","C++","Google","GitHub","R","Developers"],"extracted_tech_keywords":["AI","R","Go","C++","Git","GitHub","programming_languages:R","programming_languages:Go","programming_languages:C++"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/google-open-sources-robot-txt-to-help-standardise-robots-exclusion-protocol\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10007181,"title":"Can AutoML Outperform Data Scientists","content":"Automated machine learning – or AutoML –  was introduced to fill in for the talent gap in the ML industry and eliminate the mundane tasks of ML engineers. Over the years, many AutoML tools have been released. But, how good are these tools? Do they accomplish what they promise? Have they really become a solution to the dearth of talent in the data science industry? In order to answer these long-standing questions, researchers from Fraunhofer Institute, Germany have investigated the state-of-the-art AutoML frameworks. To their surprise, they found that the AutoML tools are performing better or on par with their human counterparts. AutoML was introduced to cut down the time spent in doing iterative tasks concerning model development. AutoML tools have helped developers to build scalable models with minimal domain expertise. So, how do they fare when pitted against humans? How Does AutoML Perform Against Humans The researchers considered 12 different popular datasets from OpenML. Of which, six are supervised classification tasks, and the rest are supervised regression ones. For the experiment, the researchers used the open-source tool AutoML benchmark, which comes with full integration of OpenML datasets for many AutoML frameworks as well as automated benchmarking functions. Supervised classification and supervised regression being the most popular machine learning tasks and the most considered tasks on OpenML, the researchers chose six of each. The benchmarks were run with default settings as defined in config.yaml in the AutoML Benchmark project. Settings include all cores, 2GiB of memory of the OS, amount of memory computed from OS available memory amongst others. The researchers considered four AutoML frameworks: TPOTH2OAuto-sklearn and AutoGluon The frameworks were chosen as a mix of very recent ones and frameworks that have been around a bit longer. The selection also encompasses Deep Learning-only AutoML frameworks as well as scikit-learn-based AutoML frameworks. Runtime per fold was selected and was set to one hour. For the supervised classification, the best of the four AutoML frameworks were given a runtime per fold of five hours in order to check their results with those of humans. For the supervised classification tasks, evaluation methods ROC AUC (auc) and accuracy were used. For the supervised regression tasks, root-mean-square error (rmse) and mean absolute error (mae) were chosen since the metrics preconfigured by AutoML Benchmark included R2. Hardware Used: The server was equipped with two Intel Xeon Silver 4114 CPUs @2.20Ghz ( 20 cores in total), four 64GB DIMM DDR4 Synchronous 2666MHz memory modules and two NVIDIA GeForce GTX 1080 Ti (more than 22GB VRAM in total). Key Findings According to the researchers, the results of this survey can be summarised as follows: AutoML performed better or equal than humans in the primary metric in 7 out of 12 cases. All these seven cases are either “easy” classification tasks (meaning tasks that humans, as well as, AutoML solved perfectly) or regression tasks. AutoML performed better or equal than humans in both metrics.There does not seem to be a significant difference between the primary and the other metric regarding performance. The researchers conclude that most results achieved by AutoML are only slightly better or worse than the ones by humans. H2O, the best AutoML framework for supervised classification credit-g, achieves an AUC score of 0.7892 using 5 hours time limit per fold instead of 0.799 using 1-hour time limit per fold. Going forward, researchers believe that there will be major leaps toward bridging the gap between domain expertise and AutoML. Machine learning applications are predominantly used in interdisciplinary cases. So, AutoML tools cannot serve as standalone solutions. AutoML should be seen as complementing the skills of data scientists and not as a magical one-stop-solution. Check the original paper here.","excerpt":"Automated machine learning – or AutoML –  was introduced to fill in for the talent gap in the ML industry and eliminate the mundane tasks of ML engineers. Over the years, many AutoML tools have been released. But, how good are these tools? Do they accomplish what they promise? Have they really become a solution […]","categories":["Deep Tech"],"tags":["automated data science solutions","Automl","H2O AI"],"author_name":"Ram Sagar","publish_date":"2020-09-11T10:00:35","publication_year":"2020","word_count":612,"keywords":["data science","Automl","scikit-learn","machine learning","Go","AI","RPA","ML","Scala","H2O AI","automated data science solutions","deep learning","R"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","data science","scikit-learn","R","Go","Scala","RPA"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/automl-data-scientist-performance-survey\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10104594,"title":"7 Noteworthy AI Innovations in Fashion","content":"This year, we saw companies around the world grab generative AI with both their hands and integrate it in their daily workflow. But the fashion industry took its own sweet time. However, we have recently witnessed significant developments in this space, particularly over the last six months of 2023. AI has been catalyst in analysing data for trend-driven collections, and is pivotal in developing eco-friendly materials, optimising production for sustainability and more. Now, let’s explore some key developments in the application of AI in the fashion industry this year. Meta Meta and luxury fashion brand EssilorLuxottica have launched the new Ray-Ban Meta smart glasses that offer improved audio with custom-designed speakers, an ultra-wide 12 MP camera for higher quality photos and 1080p videos and are powered by the Qualcomm Snapdragon AR1 Gen1 Platform. Livestreaming to Facebook or Instagram is now possible, and the glasses feature hands-free convenience, water resistance (IPX4), and a sleek charging case for up to 36 hours of use. The glasses come in Wayfarer and Headliner styles, with various frame and lens combinations, and are prescription-lens compatible. Humane AI Introduced at the premier Paris fashion week by supermodel Naomi Campbell, OpenAI-backed startup Humane AI introduced the Ai Pin, shedding light on the future of AI wearables. Humane’s AI Pin is a voice-activated wearable device that uses LLMs for conversational AI interaction. Operating on the Cosmos system, it streamlines user experience by automatically directing queries to relevant tools, eliminating the need for manual app or settings management. Through the AI Mic software, the device connects to AI models, including OpenAI’s ChatGPT, enabling tasks such as making calls, translating conversations, taking photos, and delivering reminders. Google Google has launched two features to improve online shopping of clothes. The virtual try-on for apparel uses generative AI to display clothes on diverse real models, aiding users in making informed decisions about how the garments will look on different body types and skin tones. Currently available for women’s tops from brands like Anthropologie, Everlane, H&M, and LOFT, this feature is expected to expand to include more brands and items, including men’s shirts. The second feature, guided refinements, utilises machine learning and visual matching algorithms to help users narrow down online product searches based on factors like colour, style, and pattern, offering a more tailored shopping experience with options from various retailers. Both features aim to replicate the convenience of trying on clothes in-store and address common challenges faced by online shoppers. Levi Strauss Denim pioneer Levi Strauss announced a partnership with Lalaland.AI, an Amsterdam-based digital fashion studio specialising in AI-generated models. The collaboration aims to test AI-generated models later this year to complement human models, diversifying and expanding the representation of models for Levi’s products. Lalaland.ai employs advanced AI to create hyper-realistic models representing various body types, ages, sizes, and skin tones, contributing to a more inclusive and sustainable shopping experience. While acknowledging the potential of AI technology to enhance consumer experience, the focus remains on leveraging technology to create a more diverse, personal, and engaging customer experience, aligning with Levi’s ongoing efforts to foster diversity in both content creation and consumer representation. Adobe Adobe’s Project Primrose is an interactive dress crafted using wearable and flexible, non-emissive textiles, allowing the entire surface to display content made using Adobe Firefly, Adobe After Effects, Adobe Stock, and Adobe Illustrator. The dress is engineered to animate fabric, undergoing design and style changes quickly. The technology at play involves light-diffusing modules for displays, utilising a non-emissive and flexible system employing reflective-backed polymer-dispersed liquid crystal (PDLC), a material commonly used in smart windows. This energy-efficient material, adaptable to various shapes, dynamically diffuses light. Moreover, the dress integrates sensors that respond to the wearer’s movements. Designers can opt to employ Adobe’s creative AI tools to generate visuals based on textual prompts. Shopify Canadian e-commerce giant Shopify is also embracing generative AI through its Shopify Magic initiative. Under this umbrella, the company unveiled enhanced features that leverage generative AI. Shopify Magic can now deliver tailored responses to customer queries based on their interaction history and store policies. It extends its capabilities to generate content such as blog posts, product descriptions, and marketing emails. The introduction of Sidekick, a chatbot-like AI tool, enables Shopify to comprehend and respond to queries related to business decision-making. The new capabilities of Shopify Magic utilise a combination of proprietary Shopify data, including merchant business data, and large language models like OpenAI’s GPT-4. Shopify Magic now allows merchants to automate the creation of blog posts for various occasions, offering customisation options for tone and language translation. Furthermore, it can generate content for customer emails based on brief prompts, automating the process of crafting weekly newsletters and announcements. Microsoft Microsoft has recently filed a patent for an innovative AI-powered smart backpack featuring advanced technology such as a camera, microphone, speaker, network interface, processor, and storage. The backpack, reminiscent of science fiction, is designed to enhance daily activities. In its patent application, Microsoft outlines various functionalities, including the ability to autonomously assess safety conditions for activities like skiing. Not just the bag, Microsoft has significantly impacted the fashion industry through collaborations like the one with Portugal startup XNFY Lab, resulting in the launch of AI Generated Fashion powered by Azure Machine Learning. This initiative addresses sustainability concerns by allowing brands to generate high-definition clothing models based on market trends and customer demands, reducing the risk of unsold items. In parallel, Microsoft’s Azure AI Video Indexer contributes by enabling clothing detection in videos, aiding content creators in advertising and post-event analysis. Read more: Top 7 Smart Wearables Powered by Generative AI","excerpt":"AI is changing how we shop and what we shop for.","categories":["AI Trends"],"tags":["ai in fashion"],"author_name":"Shritama Saha","publish_date":"2023-12-11T12:02:41","publication_year":"2023","word_count":933,"keywords":["ChatGPT","machine learning","ai in fashion","OpenAI","AI","ML","RAG","Ray","Aim","generative AI","Azure"],"extracted_tech_keywords":["AI","machine learning","ML","generative AI","ChatGPT","OpenAI","Aim","Ray","RAG","Azure"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-7-noteworthy-ai-innovations-in-fashion-in-2023\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10019874,"title":"What Can India Emulate From Proposed Reforms In US’ Antitrust Laws","content":"Big tech companies including Google, Facebook, Apple and Amazon are routinely summoned by governments worldwide over their monopolistic practises. A 449-page report by the US House Judiciary Committee’s Democratic leadership observed that the four companies in question–Amazon, Apple, Google, and Facebook –had turned from ‘scrappy’ startups to monopolies last seen in the era of oil barons and railroad tycoons. The report called for ‘sweeping changes’ in the existing antitrust laws. Recently, US Senator Amy Klobuchar proposed a new antitrust reform to tighten the grip over big tech firms. Klobuchar’s proposed bill, Competition and Antitrust Law Enforcement Reform, seeks to create more barriers for big mergers and stricter enforcement of antitrust laws across industries but specifically ‘dominant digital platforms’. The Proposed Bill Klobuchar was recently appointed as the chairperson of the Senate Judiciary subcommittee on antitrust. Tech companies like Apple, Amazon, Google, and Facebook, are her primary focus. Klobuchar said industries are consolidating with prominent companies buying out rivals before they emerge as potential threats. Terming these as ‘harmful exclusionary practices’, the Senator said the companies are increasingly indulging in practices such as ‘refusals to deal with rivals, restrictive contracting, and predatory pricing.’ While presenting the bill, Klobuchar said, “While the United States once had some of the most effective antitrust laws in the world, our economy today faces a massive competition problem. We can no longer sweep this issue under the rug and hope our existing laws are adequate.” The bill proposes three ways to reform the antitrust law: Resetting standard for enforcement of antitrust laws.Using additional funds for building agencies to study the markets and merger effects.Bestowing more power to antitrust enforcers. The proposed bill will give greater authority to the Department of Justice (DoJ) and the Federal Trade Commission (FTC) to penalise companies if they break competition rules. While the current maximum fine is $100 million, the new cap is pegged at 15 percent of the defaulting company’s annual income. Antitrust Laws In India In 2020, Google faced an antitrust case in India — the company’s fourth major antitrust challenge in one of its key markets. Google is also facing stiff criticism from local startups for enforcing policies that hurt their growth. Facebook came under the Competition Commission of India (CCI) scanner after the social media company picked up a significant stake in Reliance Jio. The antitrust watchdog keeps a close eye on how the deal will pan out in terms of leveraging the data of their combined subscribers to blow the rivals out of the water. Facebook and Google jointly command up to 68 percent of India’s digital ad revenue markets, whereas e-commerce companies such as Amazon and Flipkart scooped up 90 percent of major orders (as per October 2019 festival season numbers). Klobuchar is right when she says US’ antitrust laws are among the strongest in the world–at least when seen from India’s context. CCI is the statutory body that regulates anti-competitive activities in India. CCI’s ability to consistently enforce punitive measures against alleged defaulters is limited, according to an EPW article. The antitrust watchdog’s authority comes from the Competition Act, 2002, which is an improvement on its predecessor but still contains a lot of ambiguity. India can use a couple of lessons from the US’ playbook to make the CCI robust to check the big tech overreach in India.","excerpt":"Big tech companies including Google, Facebook, Apple and Amazon are routinely summoned by governments worldwide over their monopolistic practises.  A 449-page report by the US House Judiciary Committee’s Democratic leadership observed that the four companies in question–Amazon, Apple, Google, and Facebook –had turned from ‘scrappy’ startups to monopolies last seen in the era of oil […]","categories":["IT Services"],"tags":[],"author_name":"Shraddha Goled","publish_date":"2021-02-09T13:00:00","publication_year":"2021","word_count":554,"keywords":["Go","AWS","AI","cloud_platforms:AWS","Git","RAG","ViT","Rust","R","startup"],"extracted_tech_keywords":["AI","RAG","AWS","R","Go","Rust","Git","ViT","startup","cloud_platforms:AWS"],"url":"https:\/\/analyticsindiamag.com\/it-services\/what-can-india-emulate-from-proposed-reforms-in-us-antitrust-laws\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10096099,"title":"Psychedelics Fuel Silicon Valley Tech Leaders","content":"“Taking LSD was a profound experience, one of the most important things in my life. LSD shows you that there’s another side to the coin, and you can’t remember it when it wears off, but you know it. It reinforced my sense of what was important—creating great things instead of making money, putting things back into the stream of history and of human consciousness as much as I could.” – Steve Jobs Steve Jobs was one of the foremost prominent figures of the tech industry who admitted to taking LSD in the 70s, in addition to other recreational drugs. But, he wasn’t the first tech leader to resort to psychedelics. In John Markoff’s non-fiction book “What the Dormouse said”, evolution of personal computers through illicit counterculture ideals that existed in Silicon Valley in the 60s was explored. It spoke about how LSD influenced individuals and ideas that played a significant role in the birth of personal computers and the internet. Setting precedence to what exists today, tech leaders are now embracing the usage of psychedelics with some of them not shying away from talking about it –  ringing in the next big industry: the psychedelic drug market, which is poised for massive growth. Source: Reddit Of the many tech predictions Sam Altman made in a 2021 tweet, psychedelic medicine was one industry he expects to see a boom in this decade. As per Brandessence market research, the psychedelic drugs market is expected to reach a valuation of $11.82 billion by 2029 from $4.87 billion in 2022. There has also been prolific funding for psychedelic studies. Mets owner Steven Cohen, recently donated $5 million to MAPS (Multidisciplinary Association for Psychedelics Studies), a non-profit organization that explores the medical, legal, and cultural dimensions associated with psychedelics and marijuana. Psychedelics under controlled dosages are proven to be effective in the treatment of addiction, anxiety, major depressive disorder, PTSD, and other difficult-to-treat conditions such as end-of-life care. However, with properties to change perception, mood, and affect all senses to alter a person’s thinking, psychedelics have found its way into the hands of tech leaders for recreational use, probably in an effort to push creativity, improve productivity and problem-solving abilities. The Rave in Town Psychedelics (or hallucinogens) such as lysergic acid diethylamide (LSD), psilocybin-containing mushrooms (magic mushrooms), ketamine and others are consumed via microdoses for medicinal or recreational use. Microdosing on psilocybin is now becoming popular in Silicon Valley with psychedelic-trip coaches being offered for $2000 per month. According to a Financial Times article, a number of tech founders, billionaires, and professionals (whose names were withheld) microdose on LSD as means to not only boost productivity, but to improve focus and be more creative with idea generation which is required by knowledge workers. But, what about the legality of it? Colorado and Oregon are the only U.S states that have legalised the use of psilocybin for therapeutic use. However, there are other cities that have decriminalised the same drug, which implies that they are still prohibited but possessing them is not considered a criminal offence. San Francisco, home to Silicon Valley, also comes under the decriminalised status. Source: Landmark Recovery The Psych- Tech Connection While a number of anonymous Silicon Valley leaders are known to indulge in hallucinogens and other forms of restricted drugs, here are a few names that have been reported using them. Elon Musk It is reported that Elon Musk takes microdoses of ketamine which he considers to be better than antidepressants. Musk has spoken about his mental health struggles in the past where he has experienced ‘great highs, terrible lows, and unrelenting stress.’ According to the Wall Street Journal, the claim comes from people who have witnessed him use ketamine. Musk has also tweeted on how ketamine is a better alternative to SSRI (a type of antidepressant). Depression is overdiagnosed in the US, but for some people it really is a brain chemistry issue.But zombifying people with SSRIs for sure happens way too much. From what I’ve seen with friends, ketamine taken occasionally is a better option.— Elon Musk (@elonmusk) June 27, 2023 A Twitter user Geoff Pilkington tweeted appreciating Musk’s support for psychedelics and said that about how half of Los Angeles gets ketamine treatment. Sergey Brin As per the same WSJ report, Google co-founder Sergey Brin is said to consume psilocybin, commonly known as magic mushrooms. Psilocybin has been medically tested for treating cluster headaches, depression and other mental health conditions. However, there is no clarity on whether Brin’s usage is medicinal or recreational. Bob Lee Deceased founder of CashApp, a mobile payment service, Bob Lee, reportedly spent years participating in ‘The Lifestyle’ which is believed to be an underground party scene for sex and psychedelics in San Francisco. Though murdered, his toxicology report showed alcohol, cocaine, and ketamine at the time of death. Spencer Shulem Spencer Shulem, CEO of BuildBetter.ai, a technology company, uses LSD every three months to help him focus and think more creatively. He feels that the high expectations set by VCs and investors, push founders to resort to psychedelics to become extraordinary. However, Shulem is careful about sharing his LSD experiences at work and is not a preacher on the ‘joys of drugs.’","excerpt":"LSD, magic mushrooms, and ketamine are common drugs powering the tech leaders of Silicon Valley, also fuelling the psychedelic industry in the process","categories":["AI Trends"],"tags":["Depression","Elon Musk","maps","Productivity","Sam Altman","Sergey Brin","silicon valley","steve jobs"],"author_name":"Vandana Nair","publish_date":"2023-07-01T10:00:00","publication_year":"2023","word_count":869,"keywords":["Go","funding","Sam Altman","programming_languages:R","Sergey Brin","AI","maps","programming_languages:Go","Depression","Elon Musk","GAN","Aim","silicon valley","ViT","steve jobs","R","Productivity"],"extracted_tech_keywords":["AI","Aim","R","Go","GAN","ViT","funding","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/psychedelics-fuel-silicon-valley-tech-leaders\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10040257,"title":"Analytics India Industry Study 2021","content":"The Indian analytics market has seen considerable growth since Analytics India Magazine (AIM) began researching and studying the domain seven years ago. This report, which is now in its 7th year, provides insights into India’s state of the analytics industry across sectors and enterprises. The report analyses the direction in which the Indian analytics industry is headed. The financial year of 2020-21 was not like any other since the world faced the Covid-19 pandemic, and India enforced one of the strictest lockdowns. Like every other industry, the analytics and data science industry was no exception and faced the brunt of the pandemic. Firms had to lay down their employees and freeze recruitments. However, as the lockdown restrictions were slowly lifted, the post-pandemic world presented the analytics industry with favourable conditions that have ensured significant growth. The main factor was the increased digitisation that led to massive data resources among companies. As enterprises tracked digital transactions, consumer behaviour could be observed and analysed. This analysis helped increase sales and improve customer satisfaction. Also, the need to avoid human intervention\/contact amid the pandemic led to the accelerated adoption of intelligent automation or AI. This happened across every sector and for all business sizes. Suddenly, there was a surge in demand for an analytics workforce that could help enterprises grow in such difficult times. The Indian analytics and data science industry will continue to grow and play a key role in decision-making across every sector and industry. The industry will also develop sophisticated and intelligent autonomous systems to help achieve tasks with greater precision, speed, and efficiency than their human counterparts. All past reports: 2020 | 2019 | 2018 | 2017 | 2016 | 2015 OVERVIEW As covered in our reports regarding the development of the data science domain in India, it is evident that the analytics and data science function has experienced significant growth over the last year, despite the pandemic. The rising trend in salaries across almost all parameters, the maturing of the analytics market in terms of experienced hiring and salaries offered, the significance of gender diversity in the Indian analytics function, and the $836.3 Mn investment in Indian AI and analytics start-ups in 2020, are self-explanatory in terms of the development and global standing of the Indian Analytics market. This development has facilitated the increase in the current market size of the Indian analytics industry to $45.4 Bn. This signifies a growth of 26.5% in the market size y-o-y. Last year, the analytics function garnered $35.9 Bn in size. The analytics function experienced growth across almost all companies, industry types, services rendered, and geographies. The analytics industry will continue to grow in the coming years. However, the pandemic has impacted the scale and the scope of the analytics industry. Some of these impacts are short-term, while others have permanently changed the way analytics teams will operate. In this report, we have analysed the various characteristics of the Indian analytics industry, such as size, demographics, and employees and how they were impacted by the pandemic. Key Highlights 2019-2020 The market size for the analytics domain increased to $45.4 Bn in FY 2021 – a 26.5% growth in the market size over last year when it was $35.9 Bn. The analytics domain accounts for 23.4% of the Indian IT\/ITES market size in 2021. This proportion has increased from the 19.0% share last year. With the current expected CAGRs (21.2% for analytics; 8.1% for IT), the analytics industry will contribute to 41.5% of the IT\/ITES market by 2026The Indian analytics industry is predicted to grow to a market size of $98.0 billion by 2025 and $118.7 billion by 2026While the revenues generated through analytics are not published separately by most IT firms, news reports in 2020 suggest that Infosys’ analytics unit has grown into a nearly $3 billion business. and TCS said in 2018 that the firm earns $2 billion dollars per year through its analytics businessThe Banking, Financial Services, and Insurance (BFSI) sector contributes to approximately 13.9% of the total analytics market – the total includes the market share of IT\/ITES and consultancy firms.Excluding the market share of IT\/ITES and consultancy firms, the enterprises across the BFSI sector contribute 34.0% of the market size – the maximum market share.Bengaluru again led the way to the largest contribution to the analytics market among the major cities at 30.3%, followed by Delhi and Mumbai at 26.2% and 23.4%, respectively.More than half (51.6%) of the market share of analytics services to foreign countries was garnered from the US.The median work experience of analytics professionals in India marginally increased from 7.5 years in 2020 to 7.6 years in 2021. In 2019, the median was at 8.0.Mumbai has the highest median employee experience at 8.4 years in 2021, followed by Bangalore at 8.2 years. The median employee experience for Chennai increased significantly to 8.2 in 2021 from 7.2 in 2020.More than one in three (34.2%) analytics professionals in India are engineering graduates, and more than one in five (21.0%) are MBA post-graduates.28.1% of analytics professionals in 2021 are women. ANALYTICS INDIA MARKET PROJECTIONS The Analytics industry is expected to grow at a CAGR of 21.2% till 2026 – the Indian analytics market in 2026 would touch $118The analytics market accounts for 23.4% (increasing from 19.0% last year in 2020) of the entire IT\/ITES industry in 2021. This share is expected to increase and would account for 41.5% of the Indian IT industry by 2026. SECTION 1: SECTOR-WISE MARKET DISTRIBUTION The IT sector remains the top contributor (43.0%) to the analytics industry market in 2021. However, this is a considerable drop from the contribution in 2020. This drop can be mainly attributed to the increase in the analytics market share of other sectors. As domain-specific enterprises realise the significance of AI and data science, they have increased their investments to hire analytics talent. Within the IT sector, Tata Consultancy Services is the leading contributor to the analytics market in 2021, as it was the year before. This is followed by Accenture, Infosys, Cognizant, Wipro, IBM, Capgemini, Oracle, Microsoft, and Fractal. Together, these ten firms contribute to 41.3% of the analytics market share generated by the IT sector. The analytics industry’s market share is seeing a rising shift towards Indian start-ups or stand-alone firms that provide analytics-as-a-service. After IT, the BFSI sector had the second-highest analytics market share at 13.9%, highlighting the continued investment in analytics capability, personnel, and technology in this sector. This was followed by the Engineering & Manufacturing sector at 6.9% and Retail & E-commerce at 5.9%. Excluding the IT and consulting sectors, the BFSI sector contributed to 34.0% to the analytics market size in 2021. This is the largest contribution when the IT and Consulting sectors are excluded. The BFSI sector is followed by Engineering & Manufacturing at 16.9% and Retail & E-commerce at 14.4%.Enterprises in the BFSI and the E-commerce sector have heavily invested in analytics in the past couple of years. While the number of analytics professionals working in the sector increased considerably, the two sectors observed a marginal change in their analytics market share. This is mainly because both of them adopt analytics at an early stage of operations set-up.The Engineering & Manufacturing sector came to a complete halt due to the lockdown enforced amid the pandemic. As the industry realised its heavy dependence on its labour, it felt the need for establishing digital continuity in its value chains. Apart from that, Industry 4.0 has also made significant progress in recent years with IoT, cyber-physical systems, and cloud technologies enabling data collection in various ways. Hence, an increase in analytics adoption coupled with the latest technological advancements made the Engineering & Manufacturing sector the second-highest contributor to the analytics market.Pharma & Healthcare stood fourth with an 8.0% contribution to the analytics market. Many firms within the sector ramped up their hiring to build applications that could automate narrow AI tasks or develop AI chatbot consultants to assist the healthcare staff that were already stretched for resources due to the pandemic. Pharma companies hired analytics professionals for data analysis in clinical trials or drug repurposing for Covid treatment. SECTION 2: CITY-WISE MARKET DISTRIBUTION Bengaluru continues to lead as the centre of analytics – enterprises in this city contributed to 30.3% of the analytics market in 2021, up from 29.4% last year.Bengaluru’s reign as the analytics hub for MNC and Domestic IT enterprises continues this year. Moreover, the city has emerged as a start-up hub after seeing significant investments in start-up ventures in the last few years. The combined funding of the startups based in Bangalore was $535.1 Mn or 64% of the cumulative funding in 2020 – more than the start-up funding of all the other cities combined.Bengaluru is followed by Delhi, hosting enterprises that make up 26.2% of the analytics market share, slightly up from 25.3% last year.Mumbai remains third in terms of its analytics market share. It has seen a significant increase of almost six percentage points to contribute 23.4% in 2021 (up from 17.6% in 2020)The sector-location niche of Mumbai continues to remain as many of the analytics functions of the BFSI captives, Investment Banks, Domestic consumer firms, and Consulting enterprises are based in this city.Hyderabad and Chennai contribute 9.7% and 8.1% of the analytics market, respectively.Pune’s market share fell significantly to 2.4% in 2021, down from 10.2% in 2020 SECTION 3: GLOBAL GEOGRAPHY-WISE MARKET DISTRIBUTION While outsourced analytics market sizes were not reported by many IT firms, the market share by geography was determined through other secondary research sources, including whitepapers and case studies. The share of analytics services by geography indicates services to the USA garnered 51.6% of the market size in 2021, down by 5.0 percentage points compared to last year.The share from the UK focused services garnered 13.2% of the analytics market size, up from 9.7% last year.Market share from Australia and Canada comes in third and fourth at 8.3% and 6.4%. Analytics services to the European countries of Netherlands and Germany garnered 3.4% and 3.0% of the market share.The rest of the world contributed 14.1% market share in 2021, slightly higher than the 13.1% in 2020 SECTION 4: COMPANY SIZE AND EMPLOYEE DISTRIBUTION The distribution of employees across company size brackets indicates a concentration of personnel across two main categories – the startup or niche analytics category (1-200 employees) and the large-scale enterprise category (10000+ employees). A greater percentage of employees are joining enterprises that provide stand-alone analytics-as-a-service to their clients. This includes Indian startups and enterprises that are approaching unicorn status (unicorns – privately held startup enterprises that are valued more than $1 Bn)This rising trend has resulted in the highest percentage (35.5%) of employees working for enterprises with 1-200 personnel in 2021, up from 30.1% in 2020The percentage of employees working for enterprises in the 10000+ employee category was the highest in 2020 at 39.9%. This figure has dropped by more than 10 percentage points to 29.8% in 2021 SECTION 5: DEMOGRAPHICS OF ANALYTICS PROFESSIONALS IN INDIA The median work experience of analytics professionals in India marginally increased from 7.5 years in 2020 to 7.6 years in 2021. In 2019, this median was at 8.0.The drop in the median age from 2019 to 2021 implies a greater attraction of the analytics domain to the IT workforce, who are choosing analytics over other IT professions\/functions. Nonetheless, this does not imply that experienced people are not joining the analytics workforce from other domains.The percentage of freshers or analytics professionals with less than one year of experience dropped to 3.9% in 2021 from 4.6% in 2020. This was a result of tech firms freezing their recruitment drives due to the budget cuts amid the pandemic. Almost two in five (39.2%) of analytics professionals in India have work experience of fewer than 5 years.Mid-career professionals with work experience between 5 to 10 years make up 36.7% of the analytics workforce in 2021. Most of these professionals have worked in the analytics domain, since the start of their careers, and are now working as managers to head analytics teams.Experienced professionals with more than 10 years of work experience constitute 24.1% of the analytics workforce, many joining from sector-specific domains including FMCG, Telecom, and Industrials. Many of these are science and technology professionals upskilled to work as senior managers\/head of analytics divisions. City-wise Median of Years of Experience Mumbai, same as last year, has the highest level of median employee experience. It increased slightly from 8.1 years in 2020 to 8.4 years in 2021Since these firms need domain expertise, enterprises in Mumbai observe a relatively lower influx of freshers in their analytics function.Bangalore maintained 2nd place with the median employee experience at 8.2 years in 2021, a considerable increase from 7.5 years in 2020.Numerous niche Analytics firms, start-ups, and MNC & Domestic IT firms have their operations centres in Bengaluru. This has resulted in personnel with considerable experience managing the analytics functions for global and local clients – this has pushed the average experience of Bengaluru to the 2nd spot.Chennai saw the highest increase in its median employee experience. The median increased from 7.2 years in 2020 to 8.2 years in 2021. Pune also saw an increase of 0.5 years, with median employee experience at 7.6 years in 2021.The increase in the median employee experience in these two cities results from experienced professionals moving back to their hometowns amid the pandemic. The younger demographic stayed back in the cities they are working in. The lack of fresh hires further adds to the median figure. Percentage of Women in Analytics The participation of women in Analytics stands at 28.1% in 2021, slightly higher than the participation recorded last year (27.8%). This is the highest proportion recorded ever since AIM began researching the domain. Nonetheless, there is significant room for growth and improvement.Women Analytics employees are still new in the workforce in India, with a median employee experience of 6.3 years compared to 7.6 years for the entire Analytics workforce in India.Similarly, as of 2021, the median salary for women in Analytics is 24.7% lower than the entire analytics pool (10.1 Lakhs v\/s 13.4 Lakhs).Here, the bias of lower salaries offered to women employees for a given set of experience and skillset needs to be addressed by Labour Department at the Centre and State levels and industry organisations, including NASSCOM and FICCI. Distribution of Analytics Professionals by Education Level The professionals in the Analytics industry come from diverse educational backgrounds. As the pool of professionals has significantly grown over the last few years, the diversity of educational backgrounds has also widened. Engineering graduates make up more than a third (34.2%) of the analytics professionals working in Indian firms in 2021 – highest amongst all educational qualifications. Of these engineering graduates, 6.5% have graduated from top-tier institutes like the IITs, NITs, BITs, etc.Postgraduates with a Master’s in Business Administration (MBA) make up 21.0% of the professionals working in analytics. More than one in five (21.3%) of these MBA post-graduates are from top-tier universities like IIMs, XLRI, NMIMS, SP Jain, etc.As a percentage of the total analytics professionals, only 5.8% are engineering graduates or MBA post-graduates from the top-tier universities.Non-engineering graduates make up 30.4% of the analytics professionals. Non-MBA postgraduates make up 12.2% of the analytics professionals.2.1% of the analytics professionals have completed their PhDs\/PPGs CONCLUSION The market size for the analytics domain increased to $45.4 Bn in FY 2021 – an increase of 26.5% from FY 2020. This indicates the strength of the Indian analytics domain as a whole and its talent pool. Despite the lay-offs amid the pandemic, the net demand for the analytics workforce in India continued to grow through 2020 and 21 and will continue to grow in the coming years. An increase in digitisation amid the pandemic has amassed significant volumes of data resources. Firms are using these resources to analyse consumer behaviour to improve sales, production, and customer satisfaction. AI\/ML further strengthens the analytics function. With the growing importance of automation to reduce human intervention\/contact in the post-pandemic world, AI\/ML is being leveraged by many firms across the industry. Excluding the IT and Consulting sectors, the BFSI sector continues to have the highest market share at 34.0%. This was followed by the Engineering & Manufacturing sector that observed a considerable increase in its contribution (6.9% in 2020 to 16.9% in 2021) to the analytics market. In terms of cities, Bengaluru again emerged as the destination with the highest market share at 30.3%, indicating the location’s appeal in terms of analytics talent and ecosystem. With fresher recruitment drives halted by companies, amid the pandemic, the median years of experience among analytics professionals increased across cities. The highest increase was seen in Chennai from 7.2 in 2020 to 8.2 in 2021. Finally, given the growth in Digital and Online segments across not just the Indian IT industry but also other sectors, including Media & Entertainment, Retail & E-commerce, FMCG, and Telecom, the Indian analytics market is expected to grow at 21.2% CAGR till 2026. This will see the analytics market account for 41.5% of the IT\/ITES industry. Please find the full 46-Page Report attached below: To know more about our research capabilities, visit AIMResearch.","excerpt":"The Indian analytics market has seen considerable growth since Analytics India Magazine (AIM) began researching and studying the domain seven years ago. This report, which is now in its 7th year, provides insights into India’s state of the analytics industry across sectors and enterprises. The report analyses the direction in which the Indian analytics industry […]","categories":["AI Features"],"tags":["Analytics India"],"author_name":"Kashyap Raibagi","publish_date":"2021-05-17T15:00:00","publication_year":"2021","word_count":2866,"keywords":["data science","Go","ELT","AI","ML","Analytics India","Git","RAG","Aim","analytics","R"],"extracted_tech_keywords":["AI","ML","data science","analytics","Aim","RAG","R","Go","Git","ELT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/analytics-india-industry-study-2021\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":6,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10089524,"title":"GPT-4 Hype Can’t Hurt Google","content":"The much-awaited GPT-4 is here. The new transformer model is touted to outperform its predecessor ChatGPT on several competitive exams while also being safer and more aligned. Many have taken this to be one more nail—or perhaps the final nail?—in the coffin of Google. Google—for reasons known only to them—also made several announcements around the time of GPT-4’s release. However, these announcements haven’t been much of a talk in town since the only word that has captured public consciousness is GPT. Did Google really think they didn’t want to be left behind in this AI frenzy? AI Cloud The big news that came from Google this week is the release of the ‘PaLM API’. Pathways Language Model, or PaLM, is a 540-billion-parameter language model, open-sourced and made publicly available by Google. Since its release last year, expectations have been that Google will soon use it to model a variety of products like it did with BERT, which now powers the entirety of our search experience. The model surpasses the performance of 175B parameterised GPT-3 as well as the undisclosed architecture of GPT-4. The move to lease it out as an API before integrating into its Search is quite uncharacteristic of Google. But, at the same time, the push follows the huge market of Cloud that Microsoft has already gotten into with OpenAI’s GPT APIs. This is perhaps why, at the same time it released the API, Google also announced the launch of two Generative AI products on Google Cloud which will help developers build products on top of their own foundational models and others. In addition, Google will also be giving access to ‘Makersuite’, a tool that developers can use to prototype ideas, along with exercise provisions for prompt engineering, synthetic data generation, and custom-model tuning—which Google claims will be supported by robust safety tools. Looking at the current state of AI as one only for big tech companies to sell their cloud business, it does make sense for them to go the Microsoft way. Recently, Microsoft also purchased Fungible, a DPU startup, for streamlining its cloud service functions. To this purchase, add Lumenisity, a HCF solution provider, which Microsoft had acquired a month earlier. Microsoft has therefore been quite aggressively strengthening Azure. Nearly all major cloud providers have relationships with AI chip suppliers. AWS has its own silicon and custom intel processors while Google Cloud uses Arm’s Ampere Ultra chips to augment its infrastructure. The Cloud game is too competitive to be won easily and companies are throwing blind money at it to take a lead. Ethics, maybe? Unlike Microsoft, Google has repeatedly stressed upon safe and responsible AI. This is why, instead of ransacking the internet boasting its achievement, the company is providing limited access to select testers for generative AI in Workspace, starting with Docs and Gmail. This will allow them to pressure test new experiences before releasing it broadly to end-users. On the contrary, the Redmond-based company made headlines recently for laying off one of its responsible AI teams. “The pressure from [CTO] Kevin [Scott] and [CEO] Satya [Nadella] is very very high to take these most recent OpenAI models and the ones that come after them and move them into customers hands at a very high speed,” reads an article by Platformer. In the following tweet, Emily Bender discusses a paper from 2018 which gives two recommendations for how to mitigate system bias in language models. Hey @OpenAI, I'm speaking to you from 2018 to say: DOCUMENT YOUR DAMN DATASETS. Also, to everyone else: If you don't know what's in it, don't use it.Source: https:\/\/t.co\/ZRORNKk5To pic.twitter.com\/2SIKQseRfw— @emilymbender@dair-community.social on Mastodon (@emilymbender) February 26, 2023 What was unsurprising to Bender was that even four years later—in the wake of the release of GPT-4—OpenAI failed to introduce details about the architecture (including model size), hardware, training compute, dataset construction, and training method. Additionally, it was also important that Google set its eyes away from search and consider the current state of generative AI purely as a productivity tool, which would give users the liberty to accept, edit, or modify the suggestions. Among other things, Google announcements this week also include ‘MedPaLM-2’, a medical language model which is an 18% improvement to its predecessor. The model, which is considered to be equivalent to an “expert” doctor level, is already being used to explore AI-assisted potentialities for ultrasound, cancer treatment planning, and tuberculosis screening. Meanwhile, the Google-backed ‘Anthropic’ also released its own chatbot, ‘Claude’, which will be generally accessible now. The AI startup had been quietly testing the model with partners like Robin AI, AssemblyAI, Notion, Quora and DuckDuckGo. Anthropic is addressing the general pitfalls of chatbots like ChatGPT—which are known for showing bias, producing harmful content as well as hallucinating—with a technique called “constitutional AI”. While in previous techniques, tens and thousands of human feedback labels were needed, constitutional AI utilises only a list of rules or principles to be able to train less harmful AI assistants. This new technique also helps fix mistakes to AI behaviour simply by changing the principles provided, instead of fine-tuning on large RLHF datasets. Final Thoughts Beneath the current wave of AI hype, there is a nuanced story playing out that centres around the balance between noise and impact. It appears that Google’s primary objective—as is seen with MedPaLM—has been to leverage the AI trend to create tangible value. Moreover, one thing that Google has believed in since its inception is to make technology do the work for us instead of making users do the work for technology. What we have seen until now with Microsoft—and its closest ally, OpenAI—is to make end-users do the ultimate work of training and improving the model by interacting with it more and more. In this light, the hope is that Google will set a precedent for others to follow suit and make the technology create value in our everyday lives.","excerpt":"Many have taken GPT-4 to be one more nail – or perhaps the final nail? – in the coffin of Google.","categories":["AI Highlights"],"tags":["AI Models","Azure","ChatGPT","Ethical AI","Google Cloud","Google PaLM","GPT-4","Language Models","Responsible AI"],"author_name":"Ayush Jain","publish_date":"2023-03-17T13:00:00","publication_year":"2023","word_count":984,"keywords":["AI Models","ChatGPT","Anthropic","Google Cloud","RLHF","OpenAI","AI","Ethical AI","ML","GPT-4","Responsible AI","Google PaLM","RAG","Aim","prompt engineering","generative AI","Language Models","Azure"],"extracted_tech_keywords":["AI","ML","generative AI","ChatGPT","OpenAI","Anthropic","Aim","RAG","prompt engineering","RLHF"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/gpt-4-hype-cant-hurt-google\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10059360,"title":"Top laptops for Python programming in 2022","content":"For Python programming, laptops should have a longer battery life, faster processors, large screens, a powerful hard drive, a solid keyboard, and more VRAM. Python is a very popular programming language that is used all around the world—it’s easy to see why it is particularly so in 2022. It is able to communicate in English, allowing it to carry out simple tasks with minimal bother and effort. Python is utilised on a broad range of platforms and operating systems around the world, as well as with software applications, and not every laptop is capable of providing what it requires. The laptops listed below are some of the best for Python programming. Which Is The Best Laptop For Machine Learning and Artificial Intelligence?How Machine Learning Rocked Google’s Hardware Event This YearIntel Will Enter The Metaverse With Software That Will Use Competitor’s ChipsA Laptop For Every Data Scientist: PCs Compared Across 5 Price Ranges10 Mobile Apps All Data Scientists Should Have On Their Smartphones The MacBook Pro (2021) The MacBook Pro 2021 has a powerful processor, plenty of RAM, and a speed that is unmatched by most laptops in the market. The computer, despite its size of 16 inches, feels light and portable. This laptop is known for its top-notch display, powerful speakers, and speedy keyboard. It is ideal for a variety of applications, including programming, design, art, and student life. It has a 9th-generation Intel Core i7 – i9 processor and AMD Radeon Pro 5300M – Radeon Pro 5500M graphics card, as well as 16GB – 64GB of RAM. It features a 16-inch Retina display and 512GB – 8TB SSD storage. Google Pixelbook Go The Google Pixelbook Go is the best Chromebook money can buy right now, and it also doubles as a great programming laptop. Cheaper than its predecessor, the Pixelbook, Google’s latest Chromebook, retains many of the luxury features of the original, including excellent battery life and one of the nicest keyboards we’ve ever used on a laptop – a crucial consideration when looking for a laptop to programme on. Sure, it doesn’t run Windows 10; instead, it runs Chrome OS, but for most programmers, especially web developers, that won’t be an issue. This laptop can also run Linux, making it an even more versatile programming workstation. It features exceptional specs for a Chromebook, guaranteeing that Chrome OS operates smoothly, putting it on a level with many of the more costly Windows laptops and MacBooks. Dell XPS 15 Another wonderful Dell device with a solid build is the Dell XPS 15 9500. It has a 10th generation Intel Core i7 processor for added speed. The trackpad and keyboard are praised for their responsiveness and ease of use. Furthermore, this powerful laptop can be charged for up to 12 hours of battery life. It has an NVIDIA GTX 1650 Ti graphics card and a screen that is Ultrasharp 4K Ultra HD. It has a total of 6GB of DDR4 RAM. Microsoft Surface Book 2 (13.5-inch) The Microsoft Surface Book 2 is a fantastic option for any coders out there, as it is one of the most powerful 2-in-1 laptops available. After all, it has components capable of handling almost anything you can throw at it, even some light gaming in your spare time. There’s a 15-inch option with beefier components and a higher price tag if you want a bigger screen. The specifications involve – CPU: Intel Core i5 – i7, Graphics: Intel HD Graphics 620 – Nvidia GeForce GTX 1050, RAM: 8GB – 16GB, Screen: 13.5-inch 3,000 x 2,000 Pixel, Sense Display with touchscreen, Storage: 256GB, 512GB, or 1TB PCIe SSD. Acer Aspire E15 E5-576G-5762 Acer is a well-known brand in the computer business, and this laptop provides excellent value for money. It’s perfect for python programming and has a fair price tag. The processor is an Intel Core i5 with a clock speed of 1.6 GHz. With 8GB of RAM and 256GB of SSD memory, the Acer AspireE15 can effortlessly handle programming software. An NVIDIA GEFORCE MX150 GPU is also included. It’s a great contender for being one of the best laptops because it boasts an NVIDIA GEFORCE GPU. It also boasts a 15.6-inch HD display and a lighted keyboard.","excerpt":"The Microsoft Surface Book 2 is a fantastic option for any coders out there, as it is one of the most powerful 2-in-1 laptops available","categories":["AI Trends"],"tags":[],"author_name":"Abhishree Choudhary","publish_date":"2022-01-29T13:00:00","publication_year":"2022","word_count":702,"keywords":["Go","machine learning","artificial intelligence","programming_languages:R","AI","programming_languages:Go","RAG","Python","programming_languages:Python","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","RAG","Python","R","Go","programming_languages:Python","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-laptops-for-python-programming-in-2022\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10103543,"title":"OpenAI Saga Shows Why Open Source is Necessary","content":"The recent turn of events at OpenAI has prompted enterprises to reconsider their dependency on the company, sparking a fervent debate on the need for the imperative role of open-source communities in AI development. Tech experts, CEOs, industry insiders, and open-source proponents like Hugging Face CEO, Clem Delangue, and Stability AI’s Emad Moshtaque amongst others have rallied for a shift toward more decentralised and open-source AI frameworks. Recently, 70 signatories, including Meta’s chief AI scientist Yann LeCun called for more openness in AI development in a letter published by Mozilla, amid concerns that a few companies could monopolise AI, with some firms lobbying against open AI R&D. That is precisely what happened, the leadership fiasco at one company threatened service disruption for all the GPT-based businesses—almost 80% of Fortune 500 companies across industries. While support from Microsoft ensured smooth sailing—a mass exodus, as threatened by majority OpenAI employees would’ve been disastrous. This reiterated the dangers of excessive reliance on a single company’s proprietary models, echoing past concerns seen in the cloud computing industry. Hence, more than 100 startups and growing businesses built around OpenAI’s large language models turned to its competitors like Cohere to safeguard their interests. The Balance Tips Towards Open Source Interestingly, amidst OpenAI’s leadership crisis, Andrej Karpathy who has been very quiet tweeted cryptically about contemplating centralisation and decentralisation recently. “Thinking a lot about centralisation and decentralisation these few days,” Karpathy posted on X. This could also stem from the fact that OpenAI, which was meant to be an open AI research lab turned into a tight-knit closed company with ambitions for profit. However, it is no surprise because Karpathy is an active contributor to the open-source ecosystem and while holding an important place within OpenIAI had also built a ‘Baby Llama’ model based on Meta’s Llama 2. The recent shakeup could cause the adoption of this very approach—the balance between closed and open-source models to become more relevant. Enterprises will need to rethink their strategies and bring in a more multi-approach strategy consisting equally of open-source and closed-source models. Balancing the convenience of proprietary models with the transparency offered by open-source alternatives becomes increasingly integral to shaping the future of AI. Meta’s collaborations with cloud players like AWS and Microsoft to release its LlaMa 2 open-source model indicate ease of mass adoption because of established and reliable infrastructure from cloud giants. Hence, companies like Meta which released LlaMa and LlaMa 2; Hugging Face which hosts thousands of open-source models; and Cohere with its Sandbox–a library of open-source models; and TII which came up with Falcon, stand to gain from this increasing appeal of Open-source. Why Open Source is the Best The cost structures diverge significantly between proprietary and open-source models. Proprietary models often operate on usage-based pricing or subscription tiers, imposing specific costs for tasks or monthly utilisation. In contrast, many open-source models are freely distributed, although fine-tuning or customisation might require additional resources. Understanding these cost dynamics is integral to assessing their impact on profit margins and operational budgets. Thus, Meta’s LlaMa2 has been wildly successful, topping several charts, competing in capability and being adopted by firms like IBM to build its WatsonX model. This situation would also warrant a lot more fanfare and adoption of Meta’s upcoming Llama 3 which is said to be at par with OpenAI’s GPT-4. Latency is another critical concern, especially for real-time applications. Larger proprietary models, such as GPT-4, might suffer from longer response times because of API inference with a response time ranging up to 20 seconds, potentially affecting user experiences. In contrast, tailored open-source models designed for specific tasks can offer faster response times, providing a competitive advantage in time-sensitive scenarios. The aspect of flexibility and transparency underscores a fundamental contrast between proprietary and open-source models. Proprietary models often lack visibility into their underlying code, hindering user understanding and compromising consistency in user experiences. On the contrary, open-source models prioritise transparency and flexibility, empowering users with insights into model behaviours and enabling alignment with specific business objectives. Security and governance present another dichotomy. Proprietary models often tout enhanced security features and built-in content moderation, reassuring users of data protection and adherence to content policies. However, concerns regarding data compliance and potential leaks often accompany reliance on external proprietary models. Open-source models, lacking out-of-the-box security measures, can be brought within secure business perimeters for local data fine-tuning, mitigating some security risks. Balanced Board The situation reiterates the point that such consequential technology should not be controlled by a select few, warranting consideration on the board with a balanced philosophy which mulls the larger good. A board structured for disagreements and deliberation would prove to be a better fit than one which is lop-sided. Disagreements among open source boards often result in partners leaving or the dissolution of the company. However, this doesn’t necessarily impact the larger ecosystem. The fate of its source code differs, ranging from being sold, or used for a new venture’s ownership, to compensating other equal partners through a fair use agreement.","excerpt":"This reiterates that such technology should not be limited to a select few, warranting a shift towards more decentralised frameworks.","categories":["Global Tech"],"tags":["AI development","Andrej Karpathy","enterprises","Open Source","Open Source AI","OpenAI"],"author_name":"Shyam Nandan Upadhyay","publish_date":"2023-11-23T10:00:00","publication_year":"2023","word_count":837,"keywords":["Go","Hugging Face","AI development","API","Open Source","Andrej Karpathy","OpenAI","AI","cloud computing","AWS","RPA","Open Source AI","GPT","enterprises","R"],"extracted_tech_keywords":["AI","OpenAI","Hugging Face","cloud computing","AWS","R","Go","API","GPT","RPA"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/openai-saga-shows-why-open-source-is-necessary\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10097111,"title":"The Tech behind Chandrayaan-3","content":"History will be created when Chandrayaan-3 touches down on the Moon’s south pole on August 23, 5:47 PM IST. This is going to be unlike anything we’ve ever seen as this would be the first landing at the lunar south pole. Previous moon landings have primarily occurred in the equatorial region, with the furthest touch-down from the equator being Surveyor 7 near 40 degrees south latitude. Scientists from ISRO are interested in exploring the lunar poles due to the possibility of presence of water as ice molecules and hydroxyl on the lunar surface in the deep craters, which was indicated by Chandrayaan-1 in 2008. The South Pole, in particular, is considered more promising for finding water ice due to its larger area being in permanent shadows and experience colder temperatures. The presence of water at the south pole makes it an intriguing location for studying the early solar system and Earth’s history. The South Pole-Aitken basin, a massive crater, further adds to the geological interest of the region, as it could contain material from the Moon’s deep crust and upper mantle. This region is highly sought-after by both space agencies and private space companies due to its water ice deposits, which have the potential to support the establishment of a future space station. The success of India’s mission in this crucial location could be a groundbreaking development and bring about significant changes in field of deep space exploration. Ready for All Tech Challenges Despite the scientific potential, exploring the south pole poses challenges due to difficult terrain, extremely low temperatures (below -230 degrees Celsius), lack of sunlight in certain areas, and the presence of large craters. To address these challenges, scientists at ISRO have developed a new algorithm embedded in Chandrayaan-3’s software. Unlike Chandrayaan-2, which relied on interpreting speed from static images, the new technology in Chandrayaan-3 estimates the spacecraft’s speed in real-time as it descends towards the lunar surface. This innovative approach enhances landing safety and increases the mission’s chances of achieving a successful touchdown. The legs of the Chandrayaan-3 lander have also been reinforced to enable it to land safely and stabilise even at a speed of 3 m\/s or 10.8 km\/h. This improvement is crucial to prevent a rough landing like in the case of Chandrayaan-2, where control was lost just 7.2 km away from the lunar surface. The strengthened legs increase the chances of a successful landing and reduce the risk of other potential troubles. To enhance manoeuverability, the Chandrayaan-3 lander carries a larger fuel tank compared to its predecessor. This extra fuel allows for last-minute adjustments to the landing site, enabling the spacecraft to change course if it detects unstable surface conditions. The Chandrayaan-3 lander is equipped with solar panels on all four sides, ensuring continuous power supply even if it lands in an unfavourable direction or experiences tumbling. This design ensures that at least one or two sides of the lander will always face the Sun, providing uninterrupted solar energy. Enhanced navigation and guidance capabilities are incorporated into Chandrayaan-3. It features new instruments, such as the Laser Doppler Velocimeter to monitor the lander’s speed and make necessary corrections. The software has been updated with improved hazard detection and avoidance algorithms, as well as upgraded navigation and guidance software. Multiple layers of redundancies are in place to ensure system reliability in case of any failures. Extensive stress tests and experiments, including helicopter drops, have been conducted to ensure the lander’s resilience. ISRO has created various test beds to simulate lunar landing conditions, allowing comprehensive assessments of the lander’s performance and durability. AI-Powered Moon Mission Chandrayaan 3 consists of the Pragyan rover similar to Chandrayaan-2 but does not have an orbiter. The Pragyan rover is equipped with advanced AI technology, enabling it to communicate with the Vikram lander. This technology also assists the rover in various tasks and operations. Once the Pragyan rover slides off the lander and lands on the lunar surface, it will embark on a comprehensive exploration journey. One of the key capabilities of the Pragyan rover is its ability to move and land on the lunar surface. It utilises motion technology to navigate the challenging terrain and successfully reach its designated landing site. Additionally, the rover’s AI algorithm plays a crucial role in identifying traces of water and different minerals on the lunar surface. This capability allows the rover to collect valuable data and send images back to Earth for research and testing purposes. Deep learning techniques have been used to enhance Chandrayaan-3’s autonomous capabilities. AI applications enable automatic landing, intelligent decision-making, and completely automated systems. By leveraging AI, the mission becomes more self-sufficient, independent, and capable of going beyond human limitations in identifying discoveries and transmitting data back to Earth. Pragyan is specifically designed for a dedicated research period of 14 days on the lunar surface. During this time, the rover conducts sophisticated scientific measurements and geological studies, enabling deeper exploration and analysis of the lunar terrain. The location where Pragyan is supposed to land is chosen strategically to maximise the quality of data gathered. While it may not receive much sunlight, the rover’s primary objective is to gather extremely high-quality data, making the positioning essential for successful research. In the broader context of space exploration, digital approaches continue to play a significant role. Data analytics is crucial for predicting, anticipating, and minimising weather hazards. AI is extensively used in developing autonomous space vehicles capable of self-diagnosis and self-repair. Virtual reality is utilised for simulated training in space missions, and digital twinning provides answers to hypothetical scenarios, which is particularly useful in relatively unexplored and unpredictable terrains. Power-packed to the Brim As for the scientific instruments on the Pragyan rover, it is equipped with cameras for imaging purposes. Additionally, it carries an alpha-proton X-ray spectrometer (APXS) and a laser-induced breakdown spectroscope (LIBS) for detailed analysis. The APXS instrument aims to determine the elemental composition of the lunar surface near the rover’s landing site. It achieves this by utilising X-ray fluorescence spectroscopy, where X-rays or alpha particles are used to excite the surface. Through these sophisticated techniques, the APXS can detect major rock-forming elements such as Sodium, Magnesium, Silica, Aluminium, Calcium, Iron, Titanium, as well as trace elements like Strontium, Yttrium, and Zirconium. On the other hand, the LIBS instrument’s primary objective is to identify and measure the abundance of elements near the landing site on the lunar surface. It achieves this by firing high-powered laser pulses at various locations and analysing the radiation emitted by the decaying plasma. Creating History, One Rocket at a Time The launch on July 14 at 2.35 PM IST garnered significant attention, with over 1.8 million viewers tuning in to watch the event live on ISRO’s YouTube channel. The Chandrayaan-3 spacecraft was launched aboard Launch Vehicle Mark-3 (LVM3) rocket. The rocket carried an uncrewed six-wheeled lander and rover module, configured with payloads to provide data related to the Moon’s surface, representing the aspirations of 1.4 billion Indians. After approximately 16 minutes, Chandrayaan-3 successfully separated from the LVM3 and entered the Earth’s orbit, marking the beginning of its fuel-efficient journey towards the Moon. If the rest of the mission unfolds as intended, India will soon join the United States, the former Soviet Union, and China as the fourth country to achieve a moon landing. ISRO’s chairman, Sreedhara Panicker Somanath, made his first comments following the successful lift off, announcing, “Chandrayaan-3 has begun its journey towards the Moon. Our launch vehicle has put the Chandrayaan on the precise orbit around the Earth.” ISRO further tweeted that “the health of the spacecraft is normal”, indicating a positive start to the mission. The mission aims to achieve three objectives: a safe and soft landing on the Moon’s surface, demonstrating rover abilities, and conducting in-situ scientific experiments. The launch signifies India’s second attempt at achieving a soft landing on the moon’s surface. This endeavour comes nearly four years after the Chandrayaan-2 mission faced a setback when its lander-rover pair crashed into the lunar terrain in 2019—which investigations revealed were due to malfunction in its software hardware components. Soft landing which ISRO Chief Somanath has described as “15 minutes of terror”  is a set of critical tasks the lander must perform to land on the lunar surface. These tasks include firing engines at precise times and altitudes, utilising the right amount of fuel, conducting accurate scans of lunar surface features like hills and craters, and ultimately achieving a successful touchdown. As a result, the software and hardware of Chandrayaan-3 have been equipped with additional capabilities to address these identified problems. The budget of the mission has been compared with a popular Indian movie, which had a similar budget but bombed at the box office. K Sivan, then ISRO chairman, had stated in 2020 that this ambitious and domestically developed mission comes at a relatively modest cost of around Rs 615 crore. Of this amount, Rs 250 crore was designated for the lander, rover, and propulsion module, while Rs 365 crore was allocated to cover the costs of launching the mission.","excerpt":"It would be the first to land at the lunar south pole. Previous moon landings have primarily occurred in the equatorial region","categories":["AI Features"],"tags":["AI Technology","challenges","Deep Learning","solar panels"],"author_name":"Shyam Nandan Upadhyay","publish_date":"2023-07-18T15:09:44","publication_year":"2023","word_count":1508,"keywords":["Go","solar panels","AI","AI Technology","Git","RAG","Ray","Aim","deep learning","analytics","Rust","Deep Learning","challenges","R"],"extracted_tech_keywords":["AI","deep learning","analytics","Aim","Ray","RAG","R","Go","Rust","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/chandrayaan-3-india-pedaling-from-cycle-trails-to-lunar-trials\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":65524,"title":"Top 8 Resources To Find Data Science Jobs During The Recession","content":"The recent advancements in artificial intelligence have been making a profound impact on businesses and researches amid the recession. The ongoing pandemic is adversely impacting the global economy, with expected record unemployment, business disruption, among others. However, there has been a rise in hiring employees in data science and analytics jobs by organisations. In one of our most recent stories, we discussed why the data science job market is better positioned for recession. In this article, we list down the top 8 resources where you can find Data Science jobs during the recession. (The list is in no particular order) 1| NextJobHunt About: NextJobHunt is a new initiative started last month by a group of professionals with a combined experience of over 100 years in areas of Sales & Marketing, Finance, Operations, SCM, Procurement, HR function and technology. This initiative is mainly started to help employees who are majorly impacted by the recent layoffs. NextJobHunt helps employees and employers by connecting the immediately available job opportunities and skilled talents, respectively. Besides job recruitments, this initiative also provides other services such as career counselling, cracking interviews, designing powerful resumes and tips to cope with a difficult job market. You can apply here. 2| Analytics India Jobs About: Analytics India Jobs is an initiative by Analytics India Magazine and a dedicated jobs portal for Analytics & Data Science jobs in India. In this portal, employees, as well as job seekers, get opportunities to look for the latest jobs available in the tech space. The platform allows the best of data science talent to connect with top companies in the space, creating a unique environment for employers and talent to connect. Another initiative by AIM is the AIM’s Community Initiative, which helps track hiring & layoffs In Indian tech industry. You can apply here. 3| Kaggle Jobs About: One of the world’s largest community of data scientists, statisticians, and machine learning engineers, Kaggle has its own job board where it sources all the current career openings for data professionals that include data scientist, senior data scientists, machine learning engineers, data engineers and other such. You can apply here. 4| Data Science Central- Analytic Talent About: Data Science Central is a digital publishing and media company operating the growing Internet community for data science, machine learning, deep learning, big data, predictive and business analytics practitioners. Analytic Talent by Data Science Central is a job board for employees and employers where you can explore the latest job opportunities, search for who is hiring and other such. You can apply here. 5| Data Science Jobs LinkedIn About: One of the largest professional network platforms, LinkedIn is the right place to find the latest job postings as well as internships. Recently, the developers at LinkedIn unveiled a new deep learning model known as Job2Questions to increase hiring efficiency amid the recession. Besides jobs, this platform also helps you to stay updated with the latest trends and news in the industry and assists you in obtaining skills to succeed in your career. You will find various data science-related jobs such as data science specialist, principal data scientist, applied scientist and other such. You can apply here. 7| Data Science Hiring At Glassdoor About: Glassdoor is one of the world’s largest job and recruiting sites which is used by employees and employers to anonymously rate, provide salary information as well as review companies and organisations. This online job board includes a number of latest job postings, including data science jobs. You can apply here. 8| Data Science Career At Monster About: Founded in 2001, Monster is one of the popular job search portals. The platform continuously updates its portal with all the latest jobs that are available amid the pandemic. There is a curated list of organisations with business locations and areas which helps an employee to search in an efficient manner. You can apply here.","excerpt":"The recent advancements in artificial intelligence have been making a profound impact on businesses and researches amid the recession. The ongoing pandemic is adversely impacting the global economy, with expected record unemployment, business disruption, among others. However, there has been a rise in hiring employees in data science and analytics jobs by organisations. In one […]","categories":["AI Hirings"],"tags":["AI Jobs","big data developer skills","data science job","Data Science Jobs","data science jobs india","data science salary","data scientist salary","recession","what is data science"],"author_name":"Ambika Choudhury","publish_date":"2020-05-20T14:00:00","publication_year":"2020","word_count":647,"keywords":["big data developer skills","Git","deep learning","R","data science job","data science jobs india","data science","artificial intelligence","data scientist salary","recession","analytics","what is data science","Go","machine learning","AI","Data Science Jobs","data science salary","AI Jobs","Aim"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","deep learning","data science","analytics","Aim","R","Go","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-hiring\/top-8-resources-to-find-data-science-jobs-during-the-recession\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10070559,"title":"IISc develops design framework for next generation analog chipsets for AI applications","content":"Indian Institute of Science (IISc) researchers have developed a design framework to build next-generation analog computing chipsets. Using this design framework, they have built a prototype of an analog chipset called ARYABHAT-1 (Analog Reconfigurable technologY And Bias-scalable Hardware for AI Tasks). IISc said that this type of chipset could be very helpful for AI-based applications such as object or speech recognition and those that require massive parallel computing operations at high speeds.. The researchers have outlined their findings in two pre-print studies (presently under peer review). They have also filed patents and intend to work with industry partners to commercialise the technology. Image: IISc Most electronic devices use digital chips because the design process is simple and scalable. Chetan Singh Thakur, Assistant Professor at the Department of Electronic Systems Engineering (DESE), IISc, said that the advantage of using analog is massive, especially in terms of power usage and size. IISc states some difficulties associated with analog processors: Testing and co-design analog processors are difficult.Analog chips don’t scale easily. So, they have to be individually customised while transitioning to the next generation technology or to a new application. This makes their design expensive. The trade-off between precision and speed with power and area is not easy when it comes to analog design. The researchers said that different machine learning architectures can be programmed on ARYABHAT and can operate robustly across a wide range of temperatures. The architecture is also “bias-scalable.” Its performance remains the same when the conditions like voltage or current are modified.","excerpt":"The researchers have outlined their findings in two pre-print studies.","categories":["AI News"],"tags":[],"author_name":"Sreejani Bhattacharyya","publish_date":"2022-07-06T16:08:04","publication_year":"2022","word_count":253,"keywords":["machine learning","programming_languages:R","AI","Scala","Git","programming_languages:Scala","R"],"extracted_tech_keywords":["AI","machine learning","R","Scala","Git","programming_languages:R","programming_languages:Scala"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/iisc-develops-design-framework-for-next-generation-analog-chipsets-for-ai-applications\/","complexity_score":3,"technical_depth":7,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10024216,"title":"Google’s Latest Guidelines To Build Better NLU Benchmarks","content":"Recently, researchers from Google Brain and New York University laid out four criteria to fix the issues plaguing natural language understanding  (NLU) benchmarks. For a few years now, research on natural language understanding has focused on improving benchmark datasets, which feature roughly independent and identically distributed (IID) training, testing sections, validation, and drawn from data collected or annotated by crowdsourcing. The researchers stated: “Progress suffers in the absence of a trustworthy metric for benchmark-driven work: Newcomers and non-specialists are discouraged from trying to contribute, and specialists are given significant freedom to cherry-pick ad-hoc evaluation settings that mask a lack of progress.” According to the researchers, unreliable as well as biased systems score much higher on the standard benchmarks. This left little room for researchers who want to develop better systems to demonstrate their improvements. The concerns about the standard benchmarks which motivated methods like adversarial filtering are justified. However, they can and should be addressed directly, and it is possible and reasonable to do so in the context of static and IID evaluation, the paper argued. The criteria Performance on popular benchmarks is exceptionally high. However, most of the time, researchers and experts can easily find issues with high-scoring models. This has become a serious concern among researchers. The paper proposed four key criteria that good benchmarks should satisfy. The reason behind introducing the criteria is to build machines that can demonstrate a reliable and comprehensive understanding of the everyday natural language text in the context of the specific well-posed task, language variety, and topic domain. Among the language understanding tasks, the researchers focused on those that use labelled The researchers focused on those that use labelled data and designed to test relatively general language understanding skills among the language understanding tasks. The design of such benchmarks can be challenging. The researchers said effective future benchmarks for NLU tasks should satisfy these four challenges or criteria. 1| Validity This criterion is difficult to formalise fully, and no simple test currently exists to determine if a benchmark presents a valid measure of model ability. The researchers identified three minimal requirements for a benchmark to meet this criterion: An evaluation dataset should reflect the full range of linguistic variation—including words and higher-level constructions—used in the relevant domain, context, and language variety.An evaluation dataset should have plausible means by which it tests all of the language-related behaviours expected from the model to show in the context of the task.An evaluation dataset should be sufficiently free of annotation artefacts. A system cannot reach near-human performance levels by any means other than demonstrating the required language-related behaviours. 2| Reliable annotation The labels for the test examples should be reliably correct. and can only be achieved by avoiding three failure cases: Examples carelessly mislabeledExamples with no clear correct label due to unclear or underspecified task guidelinesExamples with no clear correct label under the relevant metric due to legitimate disagreements in interpretation among annotators. 3| Statistical power The benchmark evaluation datasets should be large and discriminative enough to detect any qualitatively relevant performance difference between the two models. 4| Disincentives for biased models The researchers said a benchmark should, in general, favour a model without socially relevant biases over an otherwise equivalent model with such biases. Most of the current benchmarks fail this test as they are often built around the naturally occurring or crowdsourced text. It is usually the case that a system can improve its performance by adopting heuristics that reproduce potentially harmful biases. Rounding up The researchers stated benchmarking for natural language understanding (NLU) is broken. They argued that most of the current benchmarks fail at these four criteria, and adversarial data collection does not meaningfully address the causes of these failures.","excerpt":"Evaluation for many natural language understanding (NLU) tasks is broken.","categories":["Deep Tech"],"tags":["Guide"],"author_name":"Ambika Choudhury","publish_date":"2021-04-19T14:00:00","publication_year":"2021","word_count":618,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","Git","RAG","Rust","R","programming_languages:Rust","Guide"],"extracted_tech_keywords":["AI","RAG","R","Go","Rust","Git","programming_languages:R","programming_languages:Go","programming_languages:Rust"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/googles-latest-guidelines-to-build-better-nlu-benchmarks\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10008078,"title":"Why Are Analytics Interviews Flawed","content":"A data analyst usually works as a part of an analytics team or business unit. While working with the analytics team, the analyst often ends up in performing tasks like dashboarding, extracting data, building recurring reports, performing exploratory analyses and more. The most critical skills for a data analyst include the knowledge of databases, languages like SQL, Python, etc., a good understanding of business and data. Almost 75% job roles in an analytics function comprises business intelligence, data mining and extraction, visualization etc. Someone who is starting a job with an analytics function, especially with large firms, there’s almost 90% chances that they would be doing pure data extraction and presentation related work, atleast for the first 1-2 years. High stakes roles like creating models, algorithms, presenting to the clients and business units is often left for more senior professionals in the team. There’s also a 90% chance that while interviewing for the same roles, these early starters would rather be asked more complex questions, grilled especially on model building etc. We have seen recruiters not grilling early candidates on data preparation, visualization skills as much as on algorithms. This leaves a void, almost bordering to confusion, on what the candidate should get themselves prepared for. One would argue that over time, these candidates would have to move to more complex roles and it’s good to build those competencies early on. To get an industry perspective on this, Analytics India Magazine caught up with a few experts in this field who explained where the interview gets flawed and helped in understanding how these issues can be addressed. Recurring Bottlenecks And Pain Points For A Candidate Appearing For A Data Analyst Interview There are many challenges that data analysts face, which may make them uncomfortable or not entirely prepared for these interviews. We took inputs from the industry on some of the bottleneck and pain points that candidates in analytics interviews mostly face. Varun Hasija, Director of Product Development at Doceree shares that there are mainly three bottlenecks or pain points that a candidate has to face: They have to prepare for multiple different companies with different resumes highlighting their capabilities. The high-pressure anxiety-inducing whiteboard exercises that they have to go through.Even after appearing for jobs and interviews, they have to maintain a positive outlook even when an interview does not translate well. This becomes difficult at times when a candidate has to appear for multiple interviews within a day or week. Not getting enough opportunity to develop soft skills is another pain point. “The recurring bottlenecks and pain points for a candidate appearing for interviews are still related to soft skills, particularly communication, creativity, teamwork, etc. This perceptual divergence of what companies want and what candidates perceive they must have as skills tell upon the interview performance of the latter,” shares Naveen Das, Dean Academics and Dean School of Business and Economics, Adamas University, Kolkata. RB Jadeja, Dean of Engineering, Marwadi University, believes that insufficient experience in dealing with real-life data analytics problems\/projects is a significant challenge. “Many candidates face challenges such as inability or less efficient ways to create interesting storytelling with data and not considering themselves skilled in the application of all tools and concepts required for a data analyst,” he said. “The interview questions for Data analyst focus not only on the analytical skills but also on the “soft skills” such as communication and compassion. Over the past years, data analysts have become crucial to companies’ long-term policies that involve a range of tasks, such as writing algorithms and connecting with the C-suite in the afternoon,” said Abhishek Latthe, Founder and CEO of SenseGiz Technologies. Private Interview Setting v\/s Public Interview Setting Like in other domains, data analytics interviews are conducted in public and private interview settings based on different companies. While public interview setting involves several candidates in the interview process, private settings are about a single candidate at a time approach during the hiring process. We tried to understand which process yields better hiring. “Hiring data analysts are critical as they form the foundation for the Data Science function of the organisation. We usually follow a two-step process — pre-discussion assessment based on the alignment of the candidate’s profile and work experience and personal discussion in a private setting, to understand the strengths of the candidate better,” said Abhishek Singh, Chief Analytics Officer at Lendingkart. He said that they try to look for the right mix of attitude, structured problem-solving approach and quantitative or data management skills. “Given the importance of interviews, conducting the initial interview in-person allows a more natural, spontaneous conversation than what may be possible through the telephone or by email. On the other hand, holding interviews at the workplace building may invite accidental revelation or reflections on why continuous people are leaving and returning to their work areas at regular intervals. I believe both the forms of interviews are important as per their context,” said Latthe. Das explained that while public interview setting has become very popular lately due to the comfort it provides to both the interviewer and the interviewee, one must be careful that the environment should not impinge upon the need for privacy or anonymity of the interviewee. Nowadays, with issues like maintaining the confidentiality of the candidates appearing for interviews, etc. becoming important, companies are preferring to hold interviews in public settings or off-site locations. In his opinion, the basic norms of providing comfort, privacy, confidentiality and conducive environment for candidness must be adhered to in an interview. “I believe that a private interview setting may yield better results because of its very nature of giving comfort to the candidates and letting their thoughts come out without fear of public scrutiny and ridicules. Also, sometimes an informal conversation in interviews helps candidates to melt down their stress and nervousness, which in a way helps the interviewer to extract relevant information related to the position’s profile,” shares Jadeja. How To Overcome The Flaws Many analytics professionals believe that early-stage data analysts often face complicated & core ML questions, irrelevant to the daily job practice as a data analyst. To avoid this, Latte advised that candidates must read the job description thoroughly. “This seems like rudimentary advice, but you’d be amazed at the number of people who haven’t taken time to do this. First and foremost, a data analyst must comprehend the finest shades of the task at hand, for instance, come ready with stories of how you did in your earlier roles. Make sure also to come prepared with an enhanced data analyst resume that emphasises what abilities or practices that would make your ideal for this specific role,” he said. Hasija shared that most companies follow a” one size fits all” approach, and it is necessary to consider the daily problems that they have to solve, such as data cleaning practices. “Instead of assigning high-pressure white board-based exercises without any external help, one should test the candidates for being methodical and having high-tolerance for repetitive tasks in their work,” he added. Das echoes that this is a flaw prevalent in almost all professional interviews and not just in case of data analysts. The root cause is the rigour-relevance conundrum in the profession. Since interviewers are alert to the lack of practical experience of early-stage candidates, barring few rudimentary internships, it is but natural that core algorithmic questions are asked in interviews. He further added that the understanding of the customer’s context, data generation and flow, the veracity of data, validity and reliability of data with practical insights, etc. must be tested for greater fit and relevance of the data analyst’s job. On the other hand, Jadeja has a different take. “I think, the right attitude, programming skill, reasoning and logical thinking, should be given preference for a job of a data analyst; nevertheless, I would like to ask candidates some basics or general algorithmic questions at the time of interview to know how familiar and aware the candidate is about the terms and usability of data analytics,” he said on a concluding note.","excerpt":"A data analyst usually works as a part of an analytics team or business unit. While working with the analytics team, the analyst often ends up in performing tasks like dashboarding, extracting data, building recurring reports, performing exploratory analyses and more. The most critical skills for a data analyst include the knowledge of databases, languages […]","categories":["AI Features"],"tags":["analytics interview","data analyst","Data analyst jobs","Data Analytics"],"author_name":"Ambika Choudhury","publish_date":"2020-09-22T13:00:45","publication_year":"2020","word_count":1346,"keywords":["data science","Go","ELT","AWS","AI","ML","Data analyst jobs","Python","data analyst","analytics","SQL","Data Analytics","analytics interview","R"],"extracted_tech_keywords":["AI","ML","data science","analytics","AWS","Python","R","SQL","Go","ELT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/why-are-analytics-interviews-flawed\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10001146,"title":"How Antrix Is Monetising ISRO’s Core Strengths","content":"Founded in the year 1992 under the Companies Act, Antrix is the commercial arm of the Indian Space Research Organisation (ISRO) and promotes products and services emerging because of ISRO’s space missions. Antrix provides Launch Services for customer satellites, onboard ISRO’s operational launch vehicles of PSLV and GSLV. The commercial arm of ISRO has many areas that it works in. Here are some of its specialities: Remote sensing: Antrix has remote sensing services and offers data from the constellation of Indian Remote Sensing (IRS) satellites to international customers. In order to receive and process the IRS data directly, it has established ground stations internationally. Antrix undertakes the establishment of IRS Ground Systems for customers, offering turn-key solutions for receiving systems, upgrading of existing ground systems with mission specific hardware and software. Broadcast: Antrix enables satellite communications service providers with a necessary space segment capacity, predominantly covering the Indian region. It supplies communication satellite transponders to various users. A large number of transponders from INSAT\/ GSAT fleet in C, Ext C and Ku bands are provisioned to Indian users for Television Broadcasting (TV), Direct To Home (DTH), Digital Satellite News Gathering (DSNG), Very Small Aperture Terminal (VSAT) and Telephony services. Transponders leased from foreign satellite operators, on a back-to-back basis, are also provisioned to service providers. The services address a wide range of social and business needs. Satellite subsystems: Antrix has the capability to configure, build, and deliver satellites and satellite subsystems. These subsystems are later deployed for different applications. Towards these, Antrix works with customers, leveraging maximally on the capabilities and infrastructure available at various ISRO Centres. With ISRO’s experience of having accomplished over 70 satellite missions and 40 launches, Antrix, as its commercial arm, caters to mission operations and related support services for customer satellites. The services are extended through the state-of-art facilities and networks built for both low earth orbiting and geostationary satellite missions. The facilities are backed by strong technical expertise and experience in satellite operations and management. The satellite ground stations support operations in a variety of frequency bands (S, C, Ku and Ka) and with configurations to provide round-the-clock uninterrupted services for critical mission functions. International Support Antrix has also gained plenty of support internationally. It has launched a number of international satellites from Indonesia to the US, since the year 1999. In 2018, for the first time, it also helped South Korea for its launches with its PSLV C-40. So far, 74 international customer satellites from 20 countries have been successfully made by Antrix. Former ISRO Chairman, A S Kiran Kumar, on the occasion of Antrix’s 25 years completion had said, “Companies like SpaceX and Virgin Galactic are building small launchers. There are others who are offering a one-way ticket to Mars. Launch vehicle technology is a risky business and yet there are Indian companies like Bellatrix Aerospace that are offering Garuda, a launch vehicle and Chetak, a nanosatellite launch vehicle. Their presence means that they are willing to take the risk.” Outlook ISRO, with a legacy over 30 years in satellite launching, has an extensive range of spacecraft separation systems catering up to 4000 kg class payloads. The successful demonstration of highly reliable space mechanisms in different launch vehicles over the time has put ISRO among the leading Launch Service Providers in the world today, all thanks to the team behind Antrix, which was also awarded Miniratna status. Overall these years since its inception, Antrix has come a long way, been an excellent body of ISRO and experienced tremendous growth.","excerpt":"Founded in the year 1992 under the Companies Act, Antrix is the commercial arm of the Indian Space Research Organisation (ISRO) and promotes products and services emerging because of ISRO’s space missions. Antrix provides Launch Services for customer satellites, onboard ISRO’s operational launch vehicles of PSLV and GSLV. The commercial arm of ISRO has many […]","categories":["AI Features"],"tags":["communication","Interviews and Discussions","ISRO","satellite"],"author_name":"Disha Misal","publish_date":"2019-02-12T20:43:36","publication_year":"2019","word_count":587,"keywords":["ISRO","satellite","AI","programming_languages:R","communication","Git","RAG","GAN","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","RAG","R","Git","GAN","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-antrix-is-monetising-isros-core-strengths\/","complexity_score":3,"technical_depth":6,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":1846,"title":"Smile Vun Group launches Precision Match","content":"Digital media network Smile Vun Group (SVG) has launched a digital data aggregation and analytics venture PrecisionMatch for India, Singapore and United Arab Emirates. Nitin Chowdhary, Business Head-PrecisionMatch, said SVG, a joint venture between Manish Vij’s Vun Network and Harish Bahl’s Smile Group, is investing $2 million in the company. “With relevant data we want to make digital display advertisements as powerful as digital search advertisements,” said Manish Vij, who heads SVG, which runs Tyroo, SeventyNine, digital advertising agency Quasar and newly acquired online performance marketing company DGM India. PrecisionMatch has launched its operations simultaneously in three markets: India, South East Asia and UAE with 50 million unique users and around 60 per cent of internet users across India, Singapore and UAE across a variety of consumer segments. PrecisionMatch aggregates details of what internet users in various segments are looking for. The company will provide this information to advertisers and ad networks, like Komli Media. The advertisers in turn can target display advertisements to their focus segment. “Display advertising is losing the ROI race to search and social media due to lack of targeting options. PrecisionMatch will plug this gap in the display advertising eco-system – by enabling advertisers to achieve higher ROI by facilitating them to buy audiences rather than sites or placements, by enabling advertisers to attract new user traffic to their site and enhancing and aiding consumer’s online experience by showing them only relevant ads,” Nitin Chowdhary, Business Head, PrecisionMatch. “We have data mining tools that get the relevant information using cookies, but we do not aggregate any personal information of the user. We only look at what the user is browsing or searching for,” said PrecisionMatch’s Chowdhary. PrecisionMatch will use Adobe Audience Manager to analyze the online behavior of consumers. Adobe will not only provide PrecisionMatch with technology but will also integrate it as a part of Adobe’s partner eco-system to deliver a complete digital solution to their customers in the region. The company is targeting a multi-million dollar market. A study by Forrester predicted that advertising spend will grow to $77 billion by 2016 on interactive marketing. Together, search marketing, display advertising, mobile marketing, email marketing and social media will grow to 35% of all advertising spend. “Data targeted display can  improve click through rate by 300 per cent and post click engagement by 200 per cent. In addition, the relevance of the audience improves significantly,” added Chowdhary. Vij put the digital advertising market in India, Singapore and UAE at $1.5 billion, with display accounting for around $800 million. The company will get a fee from the advertisers. This fee will in turn be split between PrecisionMatch and online publishers, like websites and blogs, where the ads are displayed and from which the data is aggregated. With PrecisionMatch advertisers can extract and utilize data across categories such as Auto, Handsets, Online Buyers, Travel, Technology etc. At present, PrecisionMatch has audience data on different consumer segments across the 3 markets–India, smartphone and SUV buyers in India, Singapore holiday seekers and UAE- Auto enthusiasts and buyers. The company has already run pilot campaigns with brands like Samsung, Ford, Tata Motors, general Motors and Expedia. The company will initially focus on auto, retail, technology and mobile segments before expanding to other categories. “Over 50% of the web inventory globally is already on exchanges and this trend is accelerating. In APAC, it is estimated to be 5-10 per cent and is rising rapidly. Additionally, in APAC, that has about $700 million of display advertising spend, the supply of inventory outstrips the demand by at least 5 times. In such a scenario, buying the right impression is critical to drive ROI for which advertisers will need quality audience data. PrecisionMatch is our resolution to make display advertising more effective in APAC and UAE by means of targeting data without destroying the ROI.” said Manish Vij, Founder, Smile Vun Group.","excerpt":"Digital media network Smile Vun Group (SVG) has launched a digital data aggregation and analytics venture PrecisionMatch for India, Singapore and United Arab Emirates. Nitin Chowdhary, Business Head-PrecisionMatch, said SVG, a joint venture between Manish Vij’s Vun Network and Harish Bahl’s Smile Group, is investing $2 million in the company. “With relevant data we want […]","categories":["AI News"],"tags":[],"author_name":"Дарья","publish_date":"2012-10-21T13:10:16","publication_year":"2012","word_count":649,"keywords":["Go","API","programming_languages:R","AI","ML","programming_languages:Go","Git","analytics","R"],"extracted_tech_keywords":["AI","ML","analytics","R","Go","Git","API","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/smile-vun-group-launches-precision-match\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10078019,"title":"MiQ is certified as a Best Firm For Data Scientists","content":"MiQ is certified as the Best Firm For Data Scientists to work for by Analytics India Magazine (AIM) through its workplace recognition programme. The Best Firm For Data Scientists certification surveys a company’s data scientists and analytics employees to identify and recognise organisations with great company cultures. AIM analyses the survey data to gauge the employees’ approval ratings and uncover actionable insights. “We are extremely gratified about being chosen as one of the ‘Best Firms For Data Scientists’ by Analytics India Magazine. This award highly acknowledges our relentless work towards ‘Excellence’ in Data Science but also recognises many other qualities: ability, struggle and effort in building a practice that constantly innovates and thrives to deliver impactful solutions! We are proud of our teams who are the constant driving force in implementing innovative ways to drive business insights and enable growth, thus fostering the Culture of Innovation! Best is yet to come,” said Ramya Parashar, Chief Operating Officer at MiQ. Biddappa Muthappa, Director – Data Science at MiQ also expressed his delight: “We are honoured and delighted to be chosen as one of the Best Firms For Data Scientists by Analytics India Magazine – This is a testimony for our hard work, best practices that are followed across teams in building more intelligent and impactful solutions to drive business outcomes for our customers and contribute to the growth of MIQ. I am super proud of the entire team for the tremendous contribution and thank MIQ for creating a culture of innovation. This recognition will motivate us further to double down our effort in building cutting edge solutions and enable MIQ to be a global leader in Programmatic space”. The analytics industry at AIM faces a talent crunch, and attracting good employees is one of the most pressing challenges that enterprises are facing. The certification by Analytics India Magazine is considered a ‘Gold Standard’ in identifying the best data science workplaces and companies participate in the programme to increase brand awareness and attract talent. Best Firms For Data Scientists is the biggest data science workplace recognition programme in India. To nominate your organisation for the certification, please fill out the form at this link.","excerpt":"The Best Firm For Data Scientists certification surveys a company’s data scientists and analytics employees to identify and recognise organisations with great company culture.","categories":["AI Highlights"],"tags":["Data Science","Data Scientists"],"author_name":"AIM Media House","publish_date":"2022-10-26T15:00:00","publication_year":"2022","word_count":362,"keywords":["data science","Go","programming_languages:R","AI","innovation","programming_languages:Go","Aim","analytics","GAN","Data Science","R","Data Scientists"],"extracted_tech_keywords":["AI","data science","analytics","Aim","R","Go","GAN","innovation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/miq-is-certified-as-a-best-firm-for-data-scientists\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10003729,"title":"An Economist&#8217;s Guide To Becoming A Chief Data Scientist: Interview With Sandip Bhattacharjee","content":"“Very early into my career, I had realised that the majority of decisions in our lives are very similar to an optimisation exercise.” For this week’s ML practitioner’s series, Analytics India Magazine got in touch with Sandip Bhattacharjee, Chief Data Scientist at Tabsquare.ai. Sandip is also a 4x Kaggle expert. In this interview, he shares his experience from his data analytics journey that spans over a decade. AIM: Can you talk about your education and your introduction to the world of data science? Sandip: As a student, I was always fascinated by the utility of statistics and econometrics to explain different aspects of consumer behaviour. And, as far as my academics are concerned, I am an Economist with a Masters in Economics from JNU with special papers in Statistics and Applied Econometrics. I did my graduation with a major in Economics along with minors in Mathematics and Statistics. My academics and fascination for statistics eventually ushered me into the world of data science. Very early into my career, I had realised that the majority of decisions in our lives are very similar to an optimisation exercise. For instance, when I am choosing between multiple routes to my office, I am either trying to minimise time taken or maybe trying to maximise my car’s fuel economy. I found direct correlation between real- life examples and machine learning algorithms, which too, are founded on the principles of ‘Loss Functions’. These algorithms also go through an optimisation exercise where the loss function is used to reduce the error in predictions. Once this was realised, I had experienced my moment of epiphany. The fact that this line of thought can be leveraged to use cases such as shaping consumer behaviour with data, only increased my fondness for Data Science. AIM: Can You Talk About Your Data Science Journey? Sandip: When I started my career in 2007, we didn’t have the word ‘Data Science’. Back in those days, use of Statistical models for descriptive, predictive and prescriptive analytics were all bundled into ‘Analytics’. My initial work in Analytics required me to build predictive models that mainly focussed around using family of Linear Regression (GLM, HLM, Ridge Regression etc.) and classical forecasting techniques (ARIMA, ARIMAX, ARCH, GARCH, VARMAX) using SAS as the main analytics tool. Over time I had to train myself on newer age Machine Learning techniques. I started with. Andrew Ng’s course on Deep Learning to prepare the basic foundations of ML before diving deep into studying a wide variety of topics to stay abreast with the state of the art in ML. Along with this, I also had to train myself on open source programming languages like R, and Python. AIM: As a Chief Data Scientist, what does your typical day look like? Sandip: Currently, I am working at Tabsquare.ai as VP and Chief Data Scientist. We are working on some of the most challenging and fascinating technology problems in the restaurant industry right now. This involves building state of the art (SOTA) in Menu Engineering, real-time recommendation engine and AI generated 1:1 promotions for customers. While in most cases the solution providers would stop at telling the next best action for their clients, the implementation of these solutions leaves a lot to be desired. At Tabsquare, we collect more than 4 million data points each day. The biggest challenge is to create robust and scalable AI solutions that can utilise this data to serve millions of customers in real-time. In most cases we are dealing with sub-millisecond latency of results being deployed on edge devices like mobiles, tablets, kiosks. We often have to hit the right balance between model performance and high throughput\/low latency requirements. AIM: Can you talk about the challenges you have faced? Sandip: A successful data science project is 80% getting the ‘Data’ right and rest 20% is ‘Science’. In one of my projects 16 hours prior to an actual client presentation, we realised that our feature engineering pipeline had a case of target leakage within the k-fold CV regime. The model results were looking too good to be true and upon deeper inspection of all components we discovered this fatal flaw. The entire team stayed up through the night, corrected the feature engineering pipeline and updated the models. The updated models turned up nice and we were able to have a successful meeting. Even though this was a frustrating experience, it was one of the most valuable lessons for the entire team. In this context, accurate data is the fuel that powers the most sophisticated ML algorithms. Contextualising a problem is very important because not every time you will find a straightforward application of what you read in a research paper and neither every business problem is about reaching the best possible value of an evaluation metric like we have in data science competitions. That’s why I suggest aspirants to get into a habit of doing Exploratory Data Analysis (often called EDA) for any problem. EDA forms a solution foundation of good feature engineering and subsequently high-quality models. Once the foundation is created you can start making your way to advanced topics. For experienced professionals, keeping pace with new techniques and technologies is quite important. This involves reading up on the latest and greatest, knowing the foundation of the new techniques and trying to find time to code ML solutions end-to-end. “While recruiting data scientists, the most important aspect I look for is ‘First Principles Thinking’.” AIM: What does it take to make a good Data Scientist? Sandip: If you want to stay competitive in this field, one should always find time to do some hands-on coding. It has helped me contextualise the last mile challenge of deploying scalable ML solutions much better. This in turn has helped me manage my teams and clients in a much more efficient manner. While recruiting data scientists, the most important aspect I look for is ‘First Principles Thinking’. Once a candidate can break down a complicated problem into its basic building blocks it becomes much easier to formulate a set of steps that leads to the complete solution. I am never looking for someone who knows everything under the sun. However, when someone lists down certain types of ML models as their expertise, I would expect 100% clarity in concepts related to them. Clarity in data structures is also very crucial. In terms of programming I place more importance on semantics rather than syntax. Syntax is something even the most experienced programmers Google up on a daily basis. Finally, experience with code versioning tools is a plus. AIM: You are a Kaggle expert, what role do you think competitive platforms play in the DS ecosystem? Sandip: This is often a highly polarising topic where one section believes that these competitions have no real value because in the real world you never get clean data like what you get in competitions. Whereas, the other section believes that doing competitions regularly gives you an edge. Being a 4X Kaggle expert, I hold a slightly different view in this regard. Data science competition platforms like Kaggle give you a view to a small yet very important section of a complete data science project. This corresponds to the sections of doing effective EDA, various feature engineering techniques and multiple ways of building highly accurate models. However, what these competitions don’t teach you is how to clean and massage the data to make it usable. Most importantly, they don’t teach the last mile challenge of deploying the scalable models and the art of stakeholder management. I personally use platforms like Kaggle to complement my learning from working on real life business problems. In the grand scheme of things both experience of solving real life business problems and knowledge gained from doing data science competitions has positive synergies on each other. AIM: What Does Your ML Toolkit Look Like? Sandip: My ML toolkit is a mixture of Python and Spark and looks like follows: Libraries: scikit-learn, Scipy, Statsmodels, LightGBM, xgboost, Tensorflow, cv2 and Transformers (by Huggingface). On Spark, I use MLlib a lot Hardware: My  personal Deep Learning Hardware setup consists of one custom made Desktop – 64GB RAM, 8GB NVIDIA GTX 1070Ti, 256GB SSD. The other is a Lenovo Legion laptop – 16GB RAM, 6GB NVIDIA RTX 2060, 1TB SSD. Cloud: I do utilise the free GPU and TPU quota provided by Kaggle & Google Colab. For office work, I have mostly worked on GCP and there we can customise the hardware as per the need of the specific project. AIM: Any tips and recommendations for Data Science aspirants? Sandip: If you are someone who is just getting started with data science, I would recommend the following: Prof. Andrew Ng’s course: to get the fundamentals sorted Deep Learning’ by Ian Goodfellow, Yoshua Bengio and Aaron Courville: for mathematical foundations of Deep Learning without any code. Courses\/books by Dr. Adrian Rosebrock: for Computer Vision. Full Stack Deep Learning course by Pieter Abbeel, Sergey Karayev and Josh Tobin: to move from solving DL problems from local system\/Kaggle notebooks to full scale production level. That said, for anyone starting off in Data Science, my suggestion would be to avoid trying to learn everything at once. One can follow a structured process where you can start with basics first. Start with basic regression and classification models and master the concepts end to end. To prepare for this ever-evolving field of data science one must have the mentality of a student and the zeal to keep learning. What is considered ‘advanced’ ML concept today may become ‘basic’ ML concept a few years down the line. I follow a three-pronged strategy to keep myself abreast with new developments in data science. Keep reading new academic papers and articles in AI\/MLKeep programming skills handy by doing small personal projects and data science competitions whenever possible Finally, contextualise the new skills learnt on the previous two steps to the actual day-to-day business problems that we are trying to solve. Irrespective of how experienced you are in this field, the hunger to learn something new everyday is a key aspect in a Data Scientist’s journey.","excerpt":"“Very early into my career, I had realised that the majority of decisions in our lives are very similar to an optimisation exercise.” For this week’s ML practitioner’s series, Analytics India Magazine got in touch with Sandip Bhattacharjee, Chief Data Scientist at Tabsquare.ai. Sandip is also a 4x Kaggle expert. In this interview, he shares […]","categories":["AI Features"],"tags":["back office data","big data and analytics everyday life","Chief Data Scientist","data management providers","Interviews and Discussions","multiple classification statistics","My Journey In Data Science","Scale Big Data"],"author_name":"Ram Sagar","publish_date":"2020-08-01T16:00:00","publication_year":"2020","word_count":1690,"keywords":["data science","machine learning","AI","Chief Data Scientist","My Journey In Data Science","ML","computer vision","Interviews and Discussions","multiple classification statistics","Ray","Aim","deep learning","analytics","Scale Big Data","big data and analytics everyday life","back office data","TensorFlow","data management providers"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","computer vision","data science","analytics","Aim","Ray","TensorFlow"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/chief-data-scientist-interview-sandip-bhattacharjee\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":62565,"title":"Third Party Maintenance Is The Way To Retain Full Control Over Your Data Centre: Ian Shearer, Park Place Technologies","content":"The mandated lockdown by the government and the increasing shift towards remote working has created a significant surge in the data centre market. According to a recent report, the data centre market in Southeast Asia is expected to grow at a CAGR of over 6% during the period 2019–2025. The report also noted some other key factors contributing to the growth of the data centre market, such as the deployment of 5G and increasing adoption of big data and IoT. This rising demand in data centre services has created the need for companies to keep data centres functioning and in operation amid lockdown, with fewer on-site personnel and increased reliance on VPN and remote-access tools and systems. The inability to maintain these data centres could potentially result in massive downtime, which in turn, could hit the business’s bottom-line hard. To better understand how the lockdown is going to impact the data centre market and how business can maintain and manage data centres remotely with third party maintenance, we connected with Ian Shearer, the managing director of APAC region at Park Place Technologies. Here are excerpts from the interaction — Can you share your views on the potential Indian market for third-party maintenance? Also, what are the requirements for organisations to maintain and manage data centres remotely? According to reports, Asia’s share of the global spend on IT hardware has made up around 40% of the total spend in recent years. India has been the third-largest market in the region, after China and Japan. With its massive population, burgeoning middle class and economic growth outpacing most western markets, India’s IT sector is well-positioned for further rapid growth. However, despite these fundamentals and scale, it seems that adoption of third-party maintenance (TPM) as an essential tool in every CIO’s kit bag has been slower in India than in other areas of the world. In part, this may have been due to a lack of presence and focus on the India market of global TPM companies. We, at Park Place Technologies (PPT), have been working to address this in recent years, where we are aiming to grow into the Indian market to build our capabilities and infrastructure across the country. We offer our market-leading global support services to Indian clients, allowing them to leverage benefits provided by TPM. Companies, in today’s era, are progressing towards modern and increasingly software-defined data centres, and are therefore looking for simplified management. Companies are evolving with the data centres that can integrate cloud, co-location and on-premise hardware across many OEMs. Globally, streamlining and centralising remote maintenance is crucial. Moreover, consolidation plays a critical role when OEMs do not want to cover another OEM’s hardware, and that is when a TPM is the solution for companies that can help them to consolidate and report across OEMs. TPM solutions can organise a potentially chaotic situation through a ‘single pane of glass’ status view and automated remote monitoring. The current COVID-19 lockdown situation has highlighted the need for advanced remote tools and management of IT infrastructure to facilitate continued uptime, which will allow businesses to continue their operation. However, post the pandemic, the world will need considerable savings, greater flexibility, and life cycle extension possibilities, offered by TPM, which will help companies navigate the potentially challenging economic environment ahead. How can Indian companies maximise their investment and drive costs down, and extend asset life of their data centres during this lockdown? Businesses can rely on TPM companies, which can provide them with significant opportunities to drive down costs and improve on their return on investment to organisations in India. Typically, TPM solutions are 40-50% cheaper than the OEM’s for the same or often better services; they offer further administrative and overhead savings through consolidation of multiple vendor contracts and renewals into a single agreement. Flexible contracts offer the ability for customers to add and remove equipment at any time. A greater range of Service Level Agreements (SLAs) also provide clients with the ability to up\/downgrade their services to meet the changing needs of the organisation and offer greater flexibility to react to the changing environment they are operating in. TPM’s provide an alternative to the OEM’s product lifecycle and the pressure to refresh technology. They offer continuity of support up to and beyond manufacturer End Of Service Life (EOSL) dates. This provides organisations with the opportunity to manage their IT infrastructures to their own needs and timetable, defer capital expenditure, ‘sweat assets’ for longer and significantly increase ROI. “The post COVID world will need considerable savings, greater flexibility and life cycle extension possibilities, which are offered by TPM, which will help companies navigate the potentially challenging economic environment ahead.” – Ian Shearer, MD – APAC at Park Place Technologies With the rising demand in the data centre services, what are the challenges companies are facing while fulfilling the demand amid this lockdown? The COVID-19 crisis has ripple effects on the data centre market. It has also created a surge in the requirements of bandwidth for mobile and home broadband networks. This, in turn, has created the requirement for the networks to carry higher loads for longer periods than previously experienced. The net result has been a reduction in performance as usage has increased for work, education and entertainment. Telecom operators and the government have been working to ensure that networks remain up and any lags are being addressed. However, as the lockdown continues, even operators cannot rule out reduced speeds or congestions. While the external issue of the usage and load on the ISP networks are significant factors for staff to work from home effectively, the ability of organisations to monitor usage proactively, predict and identify issues, manage, and optimise capabilities of their network infrastructure in data centres across India and beyond is also critical. How is Park Place simplifying the management of complex hybrid environments? Multi-vendor capabilities are key to Park Place’s ability to offer customers a more unified and therefore simplified solution for managing complex data centre infrastructures, in a range of environments. Typically OEM’s only offer services for maintaining and managing their technology, and if they offer monitoring tools, it will only be for their equipment. Therefore, multiple OEM technologies for servers, storage, and networking means multi suppliers, contracts, tools, skill sets, invoices, points of contact, and management overhead. Park Place offers multi-vendor monitoring solution and support, with the ability to have a single dashboard for businesses to monitor all the assets in their IT infrastructure in any location and environment of operation — whether that be multi-geo, on-premise, collocated or even in the cloud. The integration and automation of the predictive and reactive fault detection with Park Place’s ticket management system greatly improves both the speed by 31% faster mean time to repair and accuracy by 97% with first-time fix rate of resolution. Alongside, it also reduces the interactions needed between the customer and Park Place from the eight touchpoints — from faults logging to gathering information and triage to the engineering process and parts dispatch — to just two points of contact. We have also invested heavily in developing digital platforms, such as customer portal and mobile app and e-services, to provide a single dashboard view of the state of the IT infrastructure with real-time receipt of updates on the progress of calls. The portal also allows the management to have access to service and SLA reporting.","excerpt":"The mandated lockdown by the government and the increasing shift towards remote working has created a significant surge in the data centre market. According to a recent report, the data centre market in Southeast Asia is expected to grow at a CAGR of over 6% during the period 2019–2025. The report also noted some other […]","categories":["AI Features"],"tags":["AI engineers","big data roi","data centre","data centre india","data centres","Interviews and Discussions"],"author_name":"Sejuti Das","publish_date":"2020-04-27T14:00:00","publication_year":"2020","word_count":1232,"keywords":["big data","data centre india","Go","API","AI","AI engineers","ML","Git","RAG","Aim","big data roi","data centres","GAN","R","data centre","Interviews and Discussions"],"extracted_tech_keywords":["AI","ML","Aim","RAG","R","Go","Git","API","big data","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/third-party-maintenance-is-the-way-to-retain-full-control-over-your-data-centre-ian-shearer-park-place-technologies\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":25303,"title":"Databricks Releases Open Source Machine Learning Platform MLflow Aimed To Standardize ML Workflows","content":"San Francisco company Databricks launched MLflow to simplify ML lifecycle San Francisco headquartered Databricks that provides a unified analytics platform released MLflow, a new open source project that strives to provide some standardization to the complex processes that machine learning engineers face during the course of building, testing, and deploying machine learning models. Announcing the release of the open source platform, CTO Matei Zaharia, also the creator of Apache Spark noted that even though there are a number of open source tools that cover each and every phase of the machine learning ifecycle, such as data preparation and model training, it is hard to track experiments and reproduce the results. At a keynote address, Zaharia observed that machine learning development lifecycle is highly complex and developers face a lot of issues which are usually not present in a traditional software development lifecycle. Zaharia Listed Down A Few Pain Points Developers Face In Building ML Models: Number of tools have grown: Zaharia cited that unlike the process in traditional software development, where teams select one tool for each phase, in machine learning, engineers end up testing every available algorithm to see whether it improves results. In the end, developers use dozens of libraries. Reproducing results: Reproducing machine learning workflow by retracing the steps is extremely difficult in machine learning. For example if you have to debug a problem, it can be difficult to go back to the past work. productionizing ML models: A big challenge developers face is moving a model to production because there is no set way move models from a library to any of these tools, emphasises Zaharia in the post. MLflow Open Source Project Provides A Standardized Format For Training & Deployment MLflow, currently in alpha stage manages the entire machine learning lifecycle and allows developers to work with any machine learning library. It offers three components: MLflow tracking to record and query experiments; MLflow projects, a standardized format to package reusable code and MLflow models. Talking on the sidelines of the release at the Databricks’ Spark and AI Summit in San Francisco, Zaharia observed that MlFLOW standardises the data for training and deployment loop. “As long as developers work within the platform, if you are building models with these tools, you can deploy and productionize it thereby saving a lot of time,” he said. Since it is an open source platform, developers from across the globe MLflow would make contributions and would be able to share workflow and ML models if developers want to open source their code. The platform’s open interface is a key feature here – it is built around REST APIs and simple data formats, instead of just replying on a small set of built-in functionality. This means developers can easily add MLflow to their existing ML code and share code across any ML library that others in the company can run. Need For Standardized Open Source ML Platform Besides open source ML platforms such as Keras and Theano, companies developed internal ML platforms to manage the development lifecycle. For example, earlier last year Uber Engineering released Michelangelo, machine learning-as-a-service system for building and deploying models, Facebook developed FBLearner and Google has TFX, an end-to-end general-purpose machine learning platform released last year. Google has already open sourced some TFX libraries. According to Zaharia, most machine learning platforms only support a small set of built-in algorithms, or a single ML library, and are also tied to each company’s infrastructure. This implies that developers are unable to use other machine learning libraries. Outlook The platform is currently offered in a hosted version but if it takes off it can help startups and companies consolidate their ML workflow and can be a bit hit with businesses. However, it faces stiff competition from TensorFlow, which thanks to tech giant Google’s backing is set to become industry standard for machine learning researchers and developers. Also, Google’s TensorFlow is backed by Jeff Dean and gets continued support from the tech giant. It is also used in daily operations and TensorFlow also provides a Visualization tool called Tensorboard that most frameworks usually lack. ML practitioners also cite that the recent version of Tensorflow provides a brand new feature called Eager execution. Databrick’s project MLflow is currently hosted at GitHub and also integrates with the company’s Unified Analytics Platform.","excerpt":"San Francisco headquartered Databricks that provides a unified analytics platform released MLflow, a new open source project that strives to provide some standardization to the complex processes that machine learning engineers face during the course of building, testing, and deploying machine learning models. Announcing the release of the open source platform, CTO Matei Zaharia, also […]","categories":["IT Services"],"tags":["Keras","Tensorflow"],"author_name":"Richa Bhatia","publish_date":"2018-06-11T06:13:21","publication_year":"2018","word_count":718,"keywords":["machine learning","Keras","AI","ML","Apache Spark","Databricks","analytics","MLflow","TensorFlow","R","Tensorflow"],"extracted_tech_keywords":["AI","machine learning","ML","analytics","MLflow","TensorFlow","Keras","Apache Spark","Databricks","R"],"url":"https:\/\/analyticsindiamag.com\/it-services\/databricks-releases-open-source-machine-learning-platform-mlflow-aimed-to-standardize-ml-workflows\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10122312,"title":"GIGABYTE Announces AI Top, an All-Round Solution to Train AI Locally on PCs","content":"GIGABYTE, a Taiwanese computer hardware manufacturer and distributor, announced GIGABYTE AI TOP, a groundbreaking solution to train AI locally, at a launch event a day before COMPUTEX 2024. CEO Eddie Lin stated that GIGABYTE AI TOP was born with the motto ‘Train Your Own AI on Your Desk’, aiming to complete the last mile in the booming era of local AI. GIGABYTE AI TOP is the all-around solution to train AI models locally. It features the AI TOP Utility, the AI TOP Hardware, and the AI TOP Tutor. The AI TOP Utility is reinvented software with a friendly user interface and experience. It supports up to 236B-parameter large language models while maintaining privacy and security. The AI TOP Hardware offers flexibility and upgradability compared to traditional training solutions on the cloud and is suitable for standard electrical systems without extra cost in electricity construction. The AI TOP Tutor provides comprehensive consultation for AI TOP solutions, intuitive set-up guidance, and technical support. All these features make GIGABYTE AI TOP easily adapted by both beginners and professionals to start up their local AI training projects. The AI TOP Hardware features a variety of GIGABYTE products, including motherboards, graphics cards, SSDs, and power supply units. One of the event’s highlights was the unveiling of the Radeon PRO W7900 AI TOP 48G and Radeon PRO W7800 32G. Their presence makes GIGABYTE the first and the only professional graphics card partner on the market that collaborates with the AMD Radeon PRO series. “GIGABYTE’s persistent pursuit in quality and reliability strengthens our partnership with leading silicon giants in making the world better with AI,” Eddie Lin addressed the audience at the launch event. GIGABYTE has been working closely with top chip leaders such as partnering with NVIDIA to launch high-end RTX AI PCs, delivering exceptional user experiences.","excerpt":"The AI TOP Hardware features a variety of GIGABYTE products, including motherboards, graphics cards, SSDs, and power supply units.","categories":["AI News"],"tags":["ai announcements"],"author_name":"Pritam Bordoloi","publish_date":"2024-06-03T14:39:48","publication_year":"2024","word_count":301,"keywords":["programming_languages:R","AI","Aim","ai announcements","R"],"extracted_tech_keywords":["AI","Aim","R","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/gigabyte-announces-ai-top-to-train-ai-locally\/","complexity_score":2,"technical_depth":4,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10063104,"title":"Young Scientists Awards for school students bringing innovation to AI &#038; Robotics","content":"Apps that can help deal with allergies or mental health issues, having a robot as a best friend, creating a rover to travel on the moon’s surface, a photon scarecrow to wave off birds and animals in the crop fields—these are some of the innovative projects by school students who were recently given the ‘Young Scientist of India” awards. “Young Scientist India” is an innovation competition to promote science awareness, attract school students to science careers, and fill the big vacuum in research centres. The annual competition is organised by Space Kidz India supported by Niti Aayog under Atal Innovation Mission. The competition awards over half a million rupees as prizes for the best innovation. Here is the list of students and details of their project – Under the private school category First prize Harshita Prasad Harshita Prasad of Class IX from Amity International School, Noida, developed an app called “Phoenix mind”, a digital BFF. Explaining about the app, Harshita said, “As a friend, the app would keep the user company, boost their morale, and help them stay happy. The app would create a video call environment, as the avatar of the virtual buddy moves and talks to the user. The user’s response, facial expressions, voice tones, and words would be observed to understand the user’s psychological state. If an unhealthy mental state is detected, it would talk to them and recommend wellness solutions. This inspiration came from the prominent problems of stress, depression, and unhealthy psychological state of adolescents. We all know that adolescence is a very challenging phase of life and industry analysts say that 32 per cent of adolescents face depression, anxiety, etc., resulting in mental disorders and suicide attempts. The recent pandemic has aggravated this problem. This app will help in seamless integration and augment adolescents with their problems. Sanchita Kiran, a class XI student from Delhi Public School, Rourkela, Odisha developed an app called ‘AllerGenius’ to address the health issues of allergies. Explaining the project, Sanchita said, “Allergies are almost always overlooked by everyone and sometimes even dismissed. If an allergy is not controlled, it can lead to severe health concerns and deteriorate the quality of life. One of the safe ways to deal with allergies is to control one’s diet, but that is hard, especially when it comes to packed foods. Every day, we are coming across new companies and products, and there is no way of knowing if that food item is safe for allergy-sufferers or not. AllerGenius can help as the app helps to identify the presence of allergens in packed food products. With the help of this, people with allergies can easily find food products that are suitable for them, especially in cases where allergen information is not explicitly mentioned on the packaging.” Second prize Mujeeb Syed Mehaantaab Mujeeb Syed Mehaantaab, a class VIII student from Paramita Heritage School, Karimnagar, Telangana, developed a “Multifunctional agricultural machine (magical hexagon)”. Explaining about the machine, Syed said, “Our goal is to design and develop innovative, low cost, portable, accessible and productively functioning agricultural harvesters for the poor farmers with the help of diagonal concurrency principle of Hexagon, which helps in finding a solution to stubble burning challenges.” Subham, a class XII student from Police DAV Public School in Ludhiana, Punjab, developed a device called “Astra – to the stars’. Explaining the device, Subham said, “Today, when you look outside the window, stars are not visible due to clouds, pollution or city lights. So people are denied information about the star that they are grazing at. However, this device helps them when they point out toward the star with a laser.” Third prize Joy Chaudhary, a class XII student from Army Public School, Shankar Vihar, New Delhi, has developed a device, ‘Smart Braille Display’. Joy explained the project, “In India, 1.99%, about 4.8 million of the population, are visually challenged. Today, almost all the work is done using digital devices, and the old paper-based Braille Script is getting outdated. To make digital documents and online material more accessible to people who are blind, I have developed a ‘Tactile Display’ with the help of people who can read digital text and documents. The device can display each braille character by rotating disks to correct the configuration. It is an inexpensive and robust solution.” Under the government schools category First prize Lena Lenin, a class X student from KV CRPF Pallipuram, Kerala, developed a device to prevent red palm weevil attacks in coconut trees. “Our project aims to control pests without polluting the environment and maintain a sustainable ecosystem. It aims to attract the weevil trap and destroy the pest. We use incandescent white light that attracts the pest. It contains a trap made of a steel net in which the weevil can enter but cannot exit. The inner portion of the funnel-shaped trap is sharpened with a spike, preventing the exit. The trapped weevil is destroyed by the fumes generated automatically,” explained Lena. S Ram Viyaas of class X from Chinmaya Vidyalaya PACR Matric Hr. Sec. School, Tamil Nadu, developed a project on robotics called ‘AID’. Explaining the project, Ram said, “I am using Raspberry Pi for our robot, which can recognise the input given by the user through a mic and give output through the speaker. It can chat with you and search for anything on the internet and replay through a voice like Alexa and Google assistant. It can also relieve depression by having conversations with you; it can set reminders to take medicines at a time; if the user had a heart attack or any other medical difficulties, the robot alerts your relatives and the ambulance and sends your location. If burglars attack the house, the robot alerts the police and sends the location. I am using Raspberry Pi as hardware, which is used to interface the speaker and mic for output and input, and I have used Raspbian OS to run the Pi, and we have programmed the whole project with Python. My code will first analyse the input and convert speech to text, and then it processes using the Google server. The device first gives us output in text, and then it converts to speech.” Second prize Faisal Raiyan of class IX from Syed Abid Hussain Senior Secondary school, New Delhi, developed “The Photonic Scarecrow” device. Faisal explained the device, “The Photonic Scarecrow is a fully automated device that captures its surroundings with its 2MP night vision camera. It then sends the feed to Raspberry Pi Zero to hunt for the animals\/birds. The Raspberry Pi recognises animals\/birds and sends its coordinates to the servo motors. There are two SG90 servo motors for rotation in the X and the Y axes. The Pi simultaneously switches on the laser beam, and the servo motor takes the correct position. Once the Pi gets no movement from the camera feed, it goes back to the monitoring mode to conserve energy. The solar panel charges the 800 MaH Lithium-Ion battery during the daytime while simultaneously hunting for movements. The Photonic Scarecrow can be used anywhere in balconies or big farms.” Mohammed Safuvan, a class XI student from GHSS Manathala, Thrissur, developed a device called ‘Rasolar’. Explaining about the project, Mohammed said, “Basics of my idea is the radiation coming from the solar wind, which hits a low ionised metal surface, and the electrons present in the metal surface begin to move with high kinetic energy. With that, using the device, we produce energy from the solar wind. I got the inspiration when I studied the gold foil experiment.” Third prize 3rd prize winner Hari Priya S, a Class XII student from SV G School in Tamil Nadu said that she wanted to reduce petroleum-based plastic instead use biodegradable plastic. Explaining about the product, she said, “I was heartbroken to see the petroleum-based plastic pollute the water body and the animals eating plastic and dying because of our selfish needs. So I was inspired to produce biodegradable plastic”. Special prize Anish Kumar, a class XI student from Wisdom World School, Pune, developed “Camelum”, a space science project. Explaining about the project, Anish said, “My project is about a vehicle that can move on any terrain, whether on the Moon or the Mars, ensuring that the vehicle does not roll over or become unbalanced in any circumstance. Taking inspiration from Chandrayaan-2—where we lost communication with the lander due to a software error as the lander veered off its course and crashed 750 metres away from where it was supposed to land—to solve the problem, I decided to design the rover, where in case of any communication loss, the rover can handle itself and try to re-establish contact. In simple words, if the rover is kept safe, it is capable of establishing contact again, despite any damage.","excerpt":"“To make digital documents and online material more accessible to people who are blind, I have developed a ‘Tactile Display’ with the help of people who can read digital text and documents.”","categories":["IT Services"],"tags":["models","NITI Aayog","projects"],"author_name":"Poornima Nataraj","publish_date":"2022-03-20T18:00:00","publication_year":"2022","word_count":1468,"keywords":["Go","TPU","AI","models","ML","Git","Python","projects","Aim","Ray","GAN","R","NITI Aayog"],"extracted_tech_keywords":["AI","ML","Aim","Ray","TPU","Python","R","Go","Git","GAN"],"url":"https:\/\/analyticsindiamag.com\/it-services\/young-scientists-awards-for-school-students-bringing-innovation-to-ai-robotics\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10091057,"title":"India Backs Off on AI Regulation. But Why?","content":"The pressing need for AI regulation has sparked a significant debate in the AI community, resulting in nothing short of a civil war among researchers. Amid such a charged atmosphere, the Indian government has stoked controversy by going against the tide. In a written reply in the Lok Sabha, the Ministry of Electronics and IT (MeitY) said, “The government is not considering bringing a law or regulating the growth of artificial intelligence in the country.” World over, Italy became the first Western country to ban ChatGPT out of privacy concerns. Meanwhile, the European Union (EU), is bringing in the much-anticipated AI Act this year. In the US too, the government released a blueprint for an AI Bill of Rights. Why won’t India Regulate AI? The Indian government has taken a proactive stance on technology, particularly AI, intending to position India as a global leader in the field. The Indian government sees AI as a ‘kinetic enabler’ and wants to harness its potential for better governance. In the written response, MeitY stated, “The government is harnessing the potential of AI to provide personalised and interactive citizen-centric services through Digital Public Platforms.” The government feels putting stringent regulations in place could stifle innovation. For example, the AI Act drafted by the EU, is seen by many as too stringent. It could potentially ​​“put a lot of unnecessary bureaucracy over companies that are innovating quickly”, Robin Röhm, founder at Genie AI, said. With no existing regulations on AI worldwide at present, Europe will be the first region in the world to draft laws specific to AI. It makes sense for India to observe and assess the situation rather than hastily regulate AI. Besides, despite the rapid advancement, the government might feel the technology is still at a very nascent stage and regulation is not the need of the hour. Yann LeCun, the chief AI scientist at Meta, who refused to sign the open letter seeking AI regulation, also shares a similar opinion. He said, “Current AI systems have limited capabilities, so asking for safety measures is premature.” It does not mean no regulation at all While the Indian government has said no to AI regulation for now, it certainly does not mean there are no checks and balances in place. The MeitY did state that various central and state government departments and agencies have commenced efforts to standardise responsible AI development. The government has also recognised the ethical concerns related to AI and the same have been highlighted in the National Strategy for AI (NSAI) released in June, 2018. Further, the technology and its creators will be subject to existing and upcoming laws. “For instance, the upcoming Digital Personal Data Protection Bill 2022 (DPDPB 2022) will apply to AI developers who develop and facilitate AI technologies. As AI developers will be collecting and using massive amounts of data to train their algorithm to enhance the AI solution, they might classify as data fiduciaries,” Kamesh Shekar, programme manager at the Dialogue, a public policy think tank, told AIM. This implies that AI developers may comply with the key principles of privacy and data protection like purpose limitation, data minimisation, consensual processing, contextual integrity etc as enshrined in DPDPB 2022. “Besides, as contoured during Digital India Act (DIA) consultation, the government is also considering having provisions within the act which would define and regulate high-risk AI systems,” Shekar added. A rather strategic approach While India has opted out of regulating AI, it could still think about alternatives like market mechanisms to appropriately tackle the situation. Shekar suggests that the government could constitute mechanisms and incentives such that the market takes its course where AI developers move away from the fail-fast fundamentals and consider consumer protection and safety as a value proposition causing competitive advantage. “For instance, creating a market for principles-based accreditation, and enabling a competitive edge for AI developers. The accreditation process must have a well-laid out process and procedure that balances transparency and safeguards to protect intellectual and proprietary information. “Besides, the accreditation process must be aspirational in a way that it pushes the AI developers toward performing better on the user outcome, i.e., securing informational privacy through better data protection standards, having child safety options etc,” he said. India is AI positive According to the recently released annual Artificial Intelligence Index report by Stanford University, around 71% of Indians felt positive about AI products. One of the key takeaways from the report was that a large proportion of GitHub AI projects were contributed by software developers in India—24.2%, to be precise. Besides, India also has one of the highest number of ChatGPT users in the world. The chatbot by OpenAI has already landed OpenAI in trouble in multiple jurisdictions for misinformation and the collection, use and disclosure of personal information without consent. With the government leveraging the technology extensively for governance as well as providing its services to the citizens, it gets even critical to have frameworks in place to mitigate the risk posed by the technology. Today, AI is pervasive, and the rise of generative AI has only expedited its widespread adoption. “Now, when these implementations are happening on such a large scale, the possibilities of the technology going wrong is also limitless,” Utpal Chakraborty, chief digital officer at Allied Digital, told AIM. Regulations might still be necessary The dangers posed by AI are aplenty. “AI and automation could put people out of work in many different fields, such as manufacturing, transportation, and customer service. This could have a significant impact on the Indian economy and employment levels,” Vikas Kakkar, founder of amara.ai, told AIM. Hence, a blatant no to regulating the technology might not be the right approach. Instead, the government should take cognisance of the dangers the technology poses in its current stage as well as the dangers it could pose as it matures. It is imperative that the government takes a proactive approach rather than waiting for it to harm before regulating it. Chakraborty too believes it is the right time to regulate AI. “But it is also essential to ensure that the regulation is not so stringent that it hampers innovation or slows down the implementation of the technology,” he said. In an earlier interaction with AIM, Chakraborty had said that regulators do not really understand the nitty-gritty of the technology. Hence, regulators must sit at the table with the developers of the technology. Currently, many of the renowned names in the world of AI are calling for regulation. Turing awardee Yoshua Bengio, recognised worldwide as one of the leading experts in AI, has also warned about the dangers AI could pose. “There is no guarantee that someone in the foreseeable future won’t develop dangerous autonomous AI systems with behaviours that deviate from human goals and values,” he said in a blog post. Bengio believes it is essential to invest public funds in the development of AI systems dedicated to social priorities often neglected by the public sector.","excerpt":"The Indian government sees AI as a ‘kinetic enabler’ and wants to harness its potential for better governance","categories":["IT Services"],"tags":["AI Regulation","eu ai act","India AI"],"author_name":"Pritam Bordoloi","publish_date":"2023-04-10T12:30:00","publication_year":"2023","word_count":1161,"keywords":["India AI","Go","ChatGPT","artificial intelligence","OpenAI","AI","AWS","RAG","AI Regulation","Aim","generative AI","R","eu ai act"],"extracted_tech_keywords":["AI","artificial intelligence","generative AI","ChatGPT","OpenAI","Aim","RAG","AWS","R","Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/india-backs-off-on-ai-regulation-but-why\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":69734,"title":"Libelium Launches MySignals, a new IoT Platform to Develop eHealth and Medical Products","content":"Libelium recently released MySignals, the most complete eHealth development platform. The biometrical IoT platform allows developers to easily create new software eHealth applications and medical devices measuring 20 different body parameters. The platform includes CE, FCC and IC certifications. For the first time ever, Libelium has released to the market an IoT solution ready for eHealth applications including a completely in-house design, from the hardware to the software for data visualization. The company faces up to a strategic shift offering solutions for the whole IoT value chain, from hardware devices to Cloud Storage and API’s for data visualization. “This is a strategic shift of the company approaching to the whole value chain of the IoT”, explains Alicia Asín, CEO, Libelium. Libelium’s MySignals eHealth development platform comes with 15 different sensors that measure the most important body parameters (ECG, glucose, airflow, pulse, oxygen in blood, blood pressure…). All the biometric data is encrypted and sent to the Libelium Cloud in realtime to be visualized on the user’s private account. MySignals includes the key parameters that are usually measured in a hospital monitoring room, fits in a small suitcase and costs one hundred times lower than an Emergency Observation Unit. Developers can prototype in a deep, affordable and easy way. “Half of the world’s population lives in rural areas with very limited access to hospitals; this platform could enhance the potential of new IoT eHealth and medical applications”, commented David Gascón, CTO, Libelium. Libelium offers two different API’s for developers to access the information. The Android \/ iOS API allows to get information directly from MySignals using Bluetooth. The API Cloud allows to access to the user’s private account and get the information previously stored to be visualized in a third party platform. “Until now, there was nothing in the market like MySignals, the perfect development tool for new medical devices and applications for quick prototyping and testing eHealth products“, explained David Gascón. MySignals improves the future of health services to cover one of the main challenges of the century: enhancing the universal accessibility to a healthcare system for more than 2 billion people worldwide.","excerpt":"Libelium recently released MySignals, the most complete eHealth development platform. The biometrical IoT platform allows developers to easily create new software eHealth applications and medical devices measuring 20 different body parameters. The platform includes CE, FCC and IC certifications. For the first time ever, Libelium has released to the market an IoT solution ready for […]","categories":["AI News"],"tags":["IoT","iot platform"],"author_name":"Manisha Salecha","publish_date":"2016-10-05T08:31:18","publication_year":"2016","word_count":353,"keywords":["API","programming_languages:R","AI","RAG","iot platform","R","IoT"],"extracted_tech_keywords":["AI","RAG","R","API","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/libelium-launches-mysignals-new-iot-platform-develop-ehealth-medical-products\/","complexity_score":2,"technical_depth":5,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":18658,"title":"Alejandro G. Iñárritu Wins The First Ever Oscar Award for A Virtual Reality Project","content":"Virtual Reality is hands down one of the most popular technologies in today’s era. Be it games or other 3D experiences, virtual reality is trending like never before. Recognizing the efforts in VR, the Academy awarded its first ever Oscar to a virtual reality. The Academy of the Motion Picture Arts and Sciences announced that it will be presenting a special Oscar statuette to Alejandro G. Iñárritu’s Carne y Arena, which translates to Virtually Present, Physically Invisible. This Mexican filmmaker, known for directing movies like Birdman and The Revenant, has won two Oscars in the past. He created the exhibit which is now on display in various museums worldwide. It explores the human condition of immigrants and refugees over a VR experience. It aims at giving people a way to live through a part of immigrant’s journey. Based on true immigrant’s accounts, this installation has used vast spaces, sand-covered floors and cold waiting rooms to make the six-and-a-half-minute VR sequence. It represents a visionary and powerful experience in storytelling. AMPAS president John Bailey said in a statement, “Iñárritu’s multimedia art and cinema experience is a deeply emotional and physically immersive venture into the world of migrants crossing the desert of the American southwest in early dawn light. More than even a creative breakthrough in the still emerging form of virtual reality, it viscerally connects us to the hot-button political and social realities of the U.S.-Mexico border.” Though the Academy doesn’t have a category for virtual reality yet, this special award could pave a way for one, just the way it did by establishing Best Animated Feature category just a few years after giving out special Oscar to the Toy Story 1.","excerpt":"Virtual Reality is hands down one of the most popular technologies in today’s era. Be it games or other 3D experiences, virtual reality is trending like never before. Recognizing the efforts in VR, the Academy awarded its first ever Oscar to a virtual reality. The Academy of the Motion Picture Arts and Sciences announced that […]","categories":["AI News"],"tags":[],"author_name":"Srishti Deoras","publish_date":"2017-10-31T12:02:40","publication_year":"2017","word_count":281,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","Aim","R"],"extracted_tech_keywords":["AI","Aim","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/alejandro-g-inarritu-wins-first-ever-oscar-award-virtual-reality-project\/","complexity_score":2,"technical_depth":6,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10020475,"title":"Top 8 Approaches For Tuning Hyperparameters Of Machine Learning Models","content":"Hyperparameter tuning is one of the fundamental steps in the machine learning routine. Also known as hyperparameter optimisation, the method entails searching for the best configuration of hyperparameters to enable optimal performance. Machine learning algorithms require user-defined inputs to achieve a balance between accuracy and generalisability. This process is known as hyperparameter tuning. There are various tools and approaches available to tune hyperparameters. We have curated a list of top eight approaches for tuning hyperparameters of machine learning models. (The list is in alphabetical order) 1| Bayesian Optimisation About: Bayesian Optimisation has emerged as an efficient tool for hyperparameter tuning of machine learning algorithms, more specifically, for complex models like deep neural networks. It offers an efficient framework for optimising the highly expensive black-box functions without knowing its form. It has been applied in several fields including learning optimal robot mechanics, sequential experimental design, and synthetic gene design. Know more here. 2| Evolutionary Algorithms About: Evolutionary algorithms (EA) are optimisation algorithms that work by modifying a set of candidate solutions (population) according to certain rules called Operators. One of the main advantages of the EA is their generality: Meaning EA can be used in a broad range of conditions due to their simplicity and independence from the underlying problem. In hyperparameter tuning problems, evolutionary algorithms have proved to perform better than grid search techniques based on an accuracy-speed ratio. Know more here. 3| Gradient-Based Optimisation About: Gradient-based optimisation is a methodology to optimise several hyperparameters, based on the computation of the gradient of a machine learning model selection criterion with respect to the hyperparameters. This hyperparameter tuning methodology can be applied when some differentiability and continuity conditions of the training criterion are satisfied. Know more here. 4| Grid Search About: Grid search is a basic method for hyperparameter tuning. It performs an exhaustive search on the hyperparameter set specified by users. This approach is the most straightforward leading to the most accurate predictions. Using this tuning method, users can find the optimal combination. Grid search is applicable for several hyper-parameters, however, with limited search space. Know more here. 5| Keras’ Tuner About: Keras tuning is a library that allows users to find optimal hyperparameters for machine learning or deep learning models. The library helps to find kernel sizes, learning rate for optimisation, and different hyper-parameters. Keras tuner can be used for getting the best parameters for various deep learning models for the highest accuracy. Know more here. 6| Population-based Optimisation About: Population-based methods are essentially a series of random search methods based on genetic algorithms, such as evolutionary algorithms, particle swarm optimisation, among others. One of the most widely used population-based methods is population-based training (PBT), proposed by DeepMind. PBT is a unique method in two aspects: It allows for adaptive hyper-parameters during trainingIt combines parallel search and sequential optimisation Know more here. 7| ParamILS About: ParamILS (Iterated Local Search in Parameter Configuration Space) is a versatile stochastic local search approach for automated algorithm configuration. ParamILS is an automated algorithm configuration method that helps dvelop high-performance algorithms and their applications. ParamILS uses default and random settings for initialisation and employs iterative first improvement as a subsidiary local search procedure. It also uses a fixed number of random moves for perturbation and always accepts better or equally-good parameter configurations, but re-initialises the search at random with probability. Know more here. 8| Random Search About: Random search can be said as a basic improvement on grid search. The method refers to a randomised search over hyper-parameters from certain distributions over possible parameter values. The searching process continues till the desired accuracy is reached. Random search is similar to grid search but has proven to create better results than the latter. The approach is often applied as the baseline of HPO to measure the efficiency of newly designed algorithms. Though random search is more effective than grid search, it is still a computationally intensive method. Know more here.","excerpt":"Hyperparameter tuning is one of the fundamental steps in the machine learning routine. Also known as hyperparameter optimisation, the method entails searching for the best configuration of hyperparameters to enable optimal performance. Machine learning algorithms require user-defined inputs to achieve a balance between accuracy and generalisability. This process is known as hyperparameter tuning. There are […]","categories":["AI Trends"],"tags":["hyperparameter optimisation","hyperparameter tuning"],"author_name":"Ambika Choudhury","publish_date":"2021-02-18T17:00:00","publication_year":"2021","word_count":655,"keywords":["Go","hyperparameter optimisation","machine learning","Keras","programming_languages:R","AI","neural network","RPA","Git","hyperparameter tuning","deep learning","R"],"extracted_tech_keywords":["AI","machine learning","deep learning","neural network","Keras","R","Go","Git","RPA","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-8-approaches-for-tuning-hyperparameters-of-machine-learning-models\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":13926,"title":"Microsoft unveils desi chatbot Ruuh for India, targets the mobile using demographic","content":"Satya Nadella led Microsoft has got India on its radar. After Tay and Zo, there is another chatbot in the pipeline, Ruuh, pegged as the desi AI that never stops talking. This English speaking Microsoft chatbot is created for entertainment purpose only and is available to users in India. According to Ruuh’s FB page, the chatbot was launched early in February and is launched for the Indian market exclusively. Seemingly aimed for the younger, mobile using demographic in India, the chatbot is available on Facebook messenger. From Bollywood jokes to entertainment options, the chatbot covers the whole gamut. At a recent Microsoft’s Worldwide Partner Conference where Nadella was the keynote speakers, the top executive shared how “AI-powered chatbots will revolutionize computing experience soon”. Nadella also outlined the Redmond-based tech giant’s move in democratizing AI through agents, applications, services and infrastructure. Last year, Microsoft was in news again for lobbying for healthcare, agriculture and educational projects from Indian government. It piloted a project with an international crop research institute and put its machine learning algorithms to work to provide accurate insights for farmers in Andhra Pradesh. Microsoft’s AI push across the globe Microsoft is strengthening its push in Artificial Intelligence with acquisitions and a world class research organization, the Artificial Intelligence and Research organization that was set up earlier last year. It is headed by computer vision expert Harry Shum, a Microsoft veteran tasked with following a four-pronged approach towards democratizing AI, across services, application and infrastructure. In January, the Washington-headquartered technology giant acquired AI startup Maluuba, a Canadian AI company that was driving breakthroughs in AI with research and application.  Earlier last year, Microsoft has opened its Cortana software development kits to developers and public at large along with other tools to facilitate the smooth development of AI-based applications. Not just that, Microsoft recently unveiled an IoT & AI Insider Lab in Munich, dedicated to helping startups get their projects off the ground.  The third lab of its kind to be set up, after US and China, the Redmond based giant is bent on AI dominance by partnering an AI ecosystem.","excerpt":"Satya Nadella led Microsoft has got India on its radar. After Tay and Zo, there is another chatbot in the pipeline, Ruuh, pegged as the desi AI that never stops talking. This English speaking Microsoft chatbot is created for entertainment purpose only and is available to users in India. According to Ruuh’s FB page, the […]","categories":["AI News"],"tags":[],"author_name":"Richa Bhatia","publish_date":"2017-03-31T10:06:31","publication_year":"2017","word_count":351,"keywords":["Go","machine learning","artificial intelligence","AI","chatbots","computer vision","Aim","GAN","R","startup"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","computer vision","Aim","chatbots","R","Go","GAN","startup"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/microsoft-unveils-desi-chatbot-ruuh-india-targets-mobile-using-demographic\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10049985,"title":"Qlik Launches A No-Code Solution For Optimised Workflows","content":"Analytics cloud platform and SaaS company Qlik recently announced the debut of Qlik Application Automation, a no-code solution designed to drive action from context-aware insights through streamlined and optimised workflows between SaaS applications and Qlik Cloud. The Qlik Application Automation helps invoke downstream processes, trigger alerts and enhance collaboration that shortens time to value. “With Qlik Application Automation’s no-code, drag and drop approach to automating tasks and data workflows between Qlik Cloud and hundreds of SaaS apps, teams can more easily create the scalable connections and data flows that compel action and shorten time to value from data throughout their entire organisation,” said James Fisher, Chief Product Officer of Qlik. Chris Brunton, Business Intelligence Manager at Dorel Home, said, “With Qlik Application Automation, we’ll be able to take the next step in activating our data to drive smart processes that can help us respond to changing conditions across the globe”. Qlik Application Automation cuts down the time needed for programming of repetitive back-office tasks and gives more time back to teams to deliver compelling real-time analytics with Smart connectivity and blocks: Quickly connect to market-leading SaaS applications such as Salesforce, Slack and MS Teams, represented as smart blocks, removing the need to technically understand an application’s low-level API.No-code user interface (UI): Simple for business users, yet also offers advanced features like conditions, variables, loops, data mapping, error handlers and templates that IT specialists can use to accelerate flow development.Native Qlik Cloud integration: Helps easily build flows that leverage Qlik’s native APIs to automate your analytics DevOps processes.Dynamic automation triggers: Allows to invoke automation flows from various mechanisms to suit specific business needs. Trigger automations on-demand, use webhooks to activate flows from events, or schedule automations to execute at specific times.Central management and robust runtime: Manage and monitor the execution of your automations via Qlik Cloud to ensure flows are running and meeting your enterprise-wide SLAs. Qlik Application Automation creates a seamless connection between the market’s leading analytics platform and the 3rd party SaaS applications that drive every business. With automated analytics environments, applications, and workflows, IT and data teams can dramatically increase data-driven actions with context-aware insights throughout the organisation. Using Qlik Application Automation’s reusable templates, teams can ensure consistency and enhance productivity, especially for tasks such as analytics DevOps, app publishing, and the delivery of insights from Qlik to collaboration tools such as Slack and MS Teams.","excerpt":"Qlik Application Automation cuts down the time needed for programming of repetitive tasks and gives more time back to deliver compelling real-time analytics.","categories":["AI News"],"tags":["automated data science solutions","automated saas intelligence","Automation","Business Intelligence","optimization","SaaS-based model"],"author_name":"Victor Dey","publish_date":"2021-09-29T13:45:55","publication_year":"2021","word_count":398,"keywords":["API","SaaS-based model","AI","ML","Automation","automated saas intelligence","Scala","RAG","automated data science solutions","optimization","analytics","real-time analytics","GAN","Business Intelligence","DevOps","R"],"extracted_tech_keywords":["AI","ML","analytics","RAG","R","Scala","DevOps","API","real-time analytics","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/qlik-launches-a-no-code-solution-for-optimised-workflows\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10067072,"title":"Swami Sivasubramanian of AWS named as one of the experts to the US’ National AI Advisory Committee","content":"Recently, the US Department of Commerce announced the appointment of 27 experts to the National Artificial Intelligence Advisory Committee (NAIAC). Swami Sivasubramanian, vice president of database, analytics, and machine learning at Amazon Web Services (AWS), was among the appointees. NAIAC will advise the US president and the National AI Initiative Office on different areas related to AI. The committee was established by the National AI Initiative Act of 2020, where Congress directed the NAIAC to provide recommendations on issues such as the current state of US AI competitiveness, the state of science around AI, issues related to the AI workforce, opportunities for international cooperation, and AI-enhanced opportunities that can exist. “I am excited to collaborate with my industry colleagues, 26 of the best and brightest minds in their respective fields, to ensure the United States leads the world in the responsible development and adoption of AI, provides inclusive employment and education opportunities for the American public, and protects civil rights and liberties in our digital age,” Sivasubramanian said. NAIAC is under the National AI Initiative. It aims to ensure continued US leadership in AI research and development. As per its official website, it wants to continue leading in AI research and development, ensuring the development and use of trustworthy AI in the public and private sectors, and preparing the US workforce for the integration of AI systems across all sectors of the economy and society.","excerpt":"NAIAC will advise the US president and the National AI Initiative Office on different areas related to AI.","categories":["AI News"],"tags":[],"author_name":"Sreejani Bhattacharyya","publish_date":"2022-05-16T15:26:33","publication_year":"2022","word_count":235,"keywords":["machine learning","artificial intelligence","AWS","AI","Git","BERT","Aim","analytics","Rust","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","analytics","Aim","AWS","R","Rust","Git","BERT"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/swami-sivasubramanian-of-aws-named-as-one-of-the-experts-to-the-us-national-ai-advisory-committee\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10161718,"title":"US Supreme Court Delivers Death Blow to TikTok Forcing Sale or Shutdown","content":"On January 17, the US Supreme Court upheld the law requiring TikTok to be sold by its Chinese parent, ByteDance, or banned from the country on national security grounds starting January 19. This unanimous decision will affect the way 170 million users in the US, for whom the video platform is a leading source of income, information, and entertainment. The court declined to consider the company’s claim that the impending ban infringes upon the First Amendment rights of both the company and its users in the United States. “There is no doubt that, for more than 170 million Americans, TikTok offers a distinctive and expansive outlet for expression, means of engagement, and source of community. But Congress has determined that divestiture is necessary to address its well-supported national security concerns regarding TikTok’s data collection practices and relationship with a foreign adversary,” Al Jazeera quoted the court as saying. Interestingly, US President-elect Donald Trump remarked on Friday that the choice regarding the video app is his responsibility. “It ultimately goes up to me, so you’re going to see what I’m going to do,” Trump said in an interview with CNN. In a statement, White House Press Secretary Karine Jean-Pierre expressed that TikTok should continue to be accessible to Americans, provided it is under American or alternative ownership. This approach seeks to address the national security concerns raised by Congress during the creation of this legislation. “Given the sheer fact of timing, this Administration recognizes that actions to implement the law simply must fall to the next Administration, which takes office on Monday,” she said. TikTok’s strength lies in its recommendation algorithm, which delivers personalised content and keeps over 170 million users in the US engaged. This algorithm, coupled with the behavioural data it collects, has made TikTok a strategic asset in the global tech ecosystem. Meanwhile, ByteDance has shown no indication of relinquishing control of the algorithm, even as the possibility of a forced sale or shutdown looms. Interestingly, in the US, Xiaohongshu (also known as Rednote) recently topped the App Store charts as a current Tiktok alternative. George Galloway, a former UK Member of Parliament, stated that the US ban on TikTok has backfired, leading millions of young individuals to migrate to alternative Chinese social media platforms. This shift highlights the discrepancy between what they have been led to believe about the country and the reality, he noted in a post on X. Nonetheless, US tech giants are still in the game. Data shows that Meta’s Instagram experienced more new app downloads last year than TikTok. According to data from market intelligence firm Sensor Tower, Instagram’s total new app downloads increased by 20% year on year in 2023, reaching 768 million times, making it the most downloaded app worldwide. During the same period, TikTok downloads increased by only 4%, to 733 million times. Instagram head Adam Mosseri is reinforcing this strategy. Earlier this week, he revealed that in 2025, the platform’s algorithm will prioritise original and creative content while aiming to attract TikTok’s user base. The future of TikTok and the creator economy rests on how effectively stakeholders navigate this rapidly changing landscape.","excerpt":"US President-elect Donald Trump says the decision is up to him.","categories":["AI News"],"tags":["tiktok"],"author_name":"Aditi Suresh","publish_date":"2025-01-17T23:13:39","publication_year":"2025","word_count":523,"keywords":["Go","tiktok","API","programming_languages:R","AI","programming_languages:Go","Aim","CNN","R"],"extracted_tech_keywords":["AI","Aim","R","Go","API","CNN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/us-supreme-court-delivers-death-blow-to-tiktok-forcing-sale-or-shutdown\/","complexity_score":4,"technical_depth":8,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10125071,"title":"TWO AI Announces SUTRA for Startups Program, Offers 1 Billion Tokens for Free","content":"TWO AI  has launched the SUTRA for Startups program, an initiative aimed at accelerating startups with access to its multilingual LLM-based services. The program will select 10 startups from India, Korea, Japan, or the Middle East to receive API access to SUTRA and 1 billion tokens for free. The selected startups will leverage SUTRA’s scalable and cost-efficient text-based models to develop solutions targeting global markets. This marks the first round of the SUTRA for Startups program, designed to foster innovation by providing affordable multilingual LLM capabilities to emerging companies. Eligible startups must have existing LLM-based services and be headquartered in the specified regions. Applications are open until July 31, 2024, at 11:59 PM PST. Winners will be announced soon after the submission deadline. TWO will evaluate applicants based on factors such as the establishment date, number of employees, and revenue to ensure they meet the eligibility criteria. Selected startups will be contacted via email or phone to finalize details and provide API access. Jio backed TWO AI recently announced the launch of SUTRA through its new AI app, ChatSUTRA, available at two.chat.ai. It  is now accessible via web and will soon be available on iOS and Android. The startup raised a $20M seed fund in February 2022 from Jio Platforms and South Korean internet conglomerate Naver. “Jio has been one of our key partners for a long time and has invested in us from the very beginning,” said Pranav Mistry, the founder of TWO, in an exclusive interaction with AIM. Another product from TWO is Geniya which can browse data from the internet using Google, rivalling Perplexity AI. Mistry said that Geniya is still in public beta and users can try it out, following the official launch soon.","excerpt":"Applications are open until July 31, 2024, at 11:59 PM PST.","categories":["AI News"],"tags":["ai announcements","Pranav Mistry","Sutra","TWO AI"],"author_name":"Siddharth Jindal","publish_date":"2024-06-27T16:26:41","publication_year":"2024","word_count":288,"keywords":["Go","API","TWO AI","programming_languages:R","AI","innovation","Pranav Mistry","Scala","RAG","ai announcements","Aim","Sutra","R","startup"],"extracted_tech_keywords":["AI","Aim","RAG","R","Go","Scala","API","innovation","startup","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/two-ai-announces-sutra-for-startups-program-offers-1-billion-tokens-for-free\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10053935,"title":"AI CoEs And Innovation Centres That Opened In India In 2021","content":"Centres of Excellence (CoE) are places where companies can produce a strong competitive edge over AI. CoEs make adopting AI practices easier in an organisation and help form a structure for scaling and maintaining these processes. Even during the pandemic, India retained itself as a favourable hub for establishing CoEs and innovation centres for AI. Analytics India Magazine collates the AI CoEs and innovation hubs that were launched in India in 2021. GJSCI The Gem & Jewellery Skills Council of India (GJSCI) in August launched a Centre of Excellence (CoE) for Jewellery AI & Data-science Excellence (JADE) along with IIT-Bombay and Persistent Systems. The tech company and the institute will also help in the integration of AI, ML and data science techniques into gem and jewellery. The CoE aims to address the challenges of the supply chain of India’s gems and jewellery industry. The technology integration will help minimise rejections, improve the current poor hit ratio of designs, deliver high market returns and increase efficiencies. IIT-Kanpur Indian Institute of Technology, Kanpur, announced in June that they would form a Centre of Excellence in Artificial Intelligence & Innovation-Driven Entrepreneurship (AIIDE) at their Noida outreach campus. The UP-state cabinet cleared the proposal under the UP-Start-up Policy 2020. This centre was formed in association with FICCI (Federation of Indian Chambers of Commerce and Industry), which provides industry tie-ups and business connect. The CoE planned and focused sectors including cybersecurity, AI, IoT, and AR for development and even provided assistance to 50 startups each year. Hubballi The minister for IT\/BT, Science and Technology, and Higher Education of Karnataka announced in October the opening of a centre of excellence (CoE) for AI and Data Engineering in Hubballi. This CoE was to be a part of the Karnataka Digital Economy Mission under the Beyond Bengaluru programme. The CoE aims at facilitating graduating startups to grow further. Kotak-IISc AI-ML Centre The Indian Institute of Science (IISc) in September set up a state-of-the-art Artificial Intelligence & Machine Learning (AI-ML) Centre in association with Kotak Mahindra Bank Limited (KMBL) at the IISc Bangalore campus. The centre that is called Kotak-IISc AI-ML Centre offers Bachelor’s, Master’s, and short-term courses in AI, ML and deep learning, reinforcement learning, and Fintech. The centre also aims to promote innovation in AI and ML to develop a talent pool that can meet the industry’s emerging requirements. AJNIFM In July, the Arun Jaitley National Institute of Financial Management (AJNIFM), Haryana, in collaboration with Microsoft, formed a strategic partnership to build an Al and emerging technologies Centre of Excellence. The CoE seeks to explore cloud, Al and emerging technologies for shaping the future of public finance management in India. As part of the skilling effort, public sector officials will receive training on emerging technologies in finance management which will help them to address potential risks like money laundering, the role of responsible tech in finance and the use of machine learning models for decision making. AI-ALC In November, KLA Corporation, a US-based semiconductor manufacturing company, launched two facilities in Chennai as part of its expansion into India. The AI-Advanced Computing Lab (AI-ALC) that had been operated in collaboration with the Indian Institute of Technology (IIT) Madras was established as a centre of excellence for AI-focused research and development. The researchers and engineers at AI-ACL worked in collaboration with the AI experts at AI Modeling and Center of Excellence in Michigan and formed a global team to advance research in AI, image processing, software, and physics modelling. Massive Analytic In June, Massive Analytic, a UK-based precognition AI company, announced a £1 million investment in India to expand its employee base and build a CoE of AI and ML. This scaling up was led by Pankaj Arora, Lead Big Data Analytics Engineer. Under his guidance, the Massive Analytic India team runs with a focus on health and safety and assists the company to become a leader in medical diagnostics. Sukanta Academy In October, Jishnu Dev Burman, Tripura Deputy Chief Minister, inaugurated the State’s first Innovation Hub at the Sukanta Academy. This project was an initiative of the National Council of Science Museums. The hub was expected to open here a few years ago but was delayed due to the pandemic. The hub has been a platform for students to engage in creative and innovative activities. IIT Mandi In February, Jai Ram Thakur, Chief Minister of Himachal Pradesh, inaugurated the Technology Innovation Hub at the Indian Institute of Technology (IIT) Mandi. The hub was developed with an investment of ₹110 crores. An MoU was also signed between the district administration and IIT Mandi for the development and deployment of the Landslide Monitoring System. CHARUSAT In September, Motorola Solutions set up its first research and innovation centre at Changa-based Charotar University of Science and Technology (CHARUSAT), Anand, Gujarat. The centre aims at offering practical opportunities and internships to students at the centre. Guwahati Corsight AI, an Israel-based company, signed an MoU with Assam Electronics Development Corporation Limited (AMTRON) in August to develop an Artificial Intelligence Centre of Excellence at a tech city situated near Guwahati. The CoE becomes a place for Corsight AI in assistance with AMTRON to establish a strong Face Recognition Technology Development and Services portfolio.","excerpt":"Analytics India Magazine collates the AI CoEs and innovation hubs that were launched in India in 2021.","categories":["IT Services"],"tags":["CoE","IISc","IIT Bombay","iit kanpur","IIT Madras"],"author_name":"Meeta Ramnani","publish_date":"2021-11-21T13:00:00","publication_year":"2021","word_count":868,"keywords":["data science","Go","IIT Bombay","artificial intelligence","machine learning","iit kanpur","CoE","IIT Madras","AI","ML","IISc","Aim","deep learning","analytics","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","data science","analytics","Aim","R","Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/ai-coes-and-innovation-centres-that-opened-in-india-in-2021\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":40165,"title":"10 Most Secured Linux Distros For Advanced Privacy &#038; Security","content":"A PC or laptop without any security measures becomes exposed to the possibilities of being attacked. Linux is one of the most popular operating systems used by individuals as well as organisations and academics. To have a secured distro means to secure valuable data. In this article, we list down 10 top Linux distributions for advanced privacy and security. (The list is in alphabetical order) 1| Alpine Linux Alpine Linux is an independent, non-commercial, general purpose Linux distribution designed for power users who appreciate security, simplicity and resource efficiency. It is built around musl libc and busybox which makes it moreresource efficient than the traditional GNU\/Linux distributions. In this distro, all uderland binaries are compiled as Position Independent Executable (PIE) with stack smashing protection which helps in preventing exploitation of entire classes of zero-day and other such vulnerabilities. Click here to download. 2| BlackArch Linux This is an Arch Linux-based distribution which is mainly for the security researchers and penetrtion testing. This is a relatively new project and is compatible with existing Arch installs. The repository contains more than 2000 tools which can be easily installed individually or in groups. Click here to download 3| Discreete Linux This operating system has been designed specially for protecting data against surveillance attacks with trojan software. The distro provides protection against attacks and vulnerabilities and can be used by anyone without deeper computer knowledge but high security requirements. It creates a secure, isolated environment for processing and encrypting\/decrypting sensitive data. Click here to download. 4| IprediaOS IprediaOS is a fast, powerful and stable operating system based on Linux which provides an anonymous environment. In this distro, all network traffic is automatically and transparently encrypted and anonymised. The operating system includes features such as I2P router, anonymous BitTorrent client, anonymous e-mail client, anonymous IRC client, etc. You can also be able to browse internet anonymously as well as find anonymous eepSites. Click here to download. 5| Kali Linux Kali Linux is an open source Debian-based Linux distribution which is maintained and funded by the provider of world-class information security training and penetration testing services, Offensive Security. This distro contains several hundred tools which are geared towards various information security tasks, such as penetration testing, security research, computer forensics and reverse engineering. It is the new generation of the industry-leading BackTrack Linux penetration testing and security auditing Linux distribution which has a complete re-build of BackTrack from the ground up, adhering completely to Debian development standards Click here to download. 6| Linux Kodachi Linux Kodachi operating system is based on Xbuntu 18.04 which will provide you with a secure, anti-forensic, and anonymous operating system. The entire operating system is functional from your temporary memory RAM so once you shut it down no trace is left behind all your activities are wiped out. Kodachi is based on the solid Linux Debian with customized XFCE which makes this distro stable, secure, and unique. Click here to download. 7| Qubes OS Qubes OS is a security-oriented operating system (OS) which runs all the other programs on a computer. This operating system implies security by compartmentalisation, which allows you to compartmentalise the various parts of your digital life into securely isolated compartments called qubes. This distro is free and open-source software which means that the source code is openly available so others can contribute to and audit it. Click here to download. 8| Subgraph OS Subgraph OS is a Debian-based Linux distribution which provides several security, anonymous web browsing and hardening features. It was designed to reduce the risks in endpoint systems so that individuals and organisations can communicate, share, and collaborate without fear of surveillance or interference. Subgraph OS includes strong system-wide attack mitigations which protects all applications as well as the core operating system, and key applications are run in sandbox environments to reduce the impact of any attacks against applications. Click here to download. 9| Tails The Amnestic Incognito Live System (TAILS) operating system is a Debian-based Linux Distribution which aims at preserving privacy as well as anonymity. By using this distro, you can use the Internet anonymously and circumvent cencorship, use the state-of-the-art cryptographic tools to file encryption, instant messaging, etc. Click here to download. 10| Whonix This is a Debian GNU\/Linux based operating system which is designed for advanced security and privacy. Whonix mitigates the threat of common attack vectors while maintaining usability. It utilises Tor’s free software which provides an open and distributed relay network to defend against network surveillance. Click here to download.","excerpt":"A PC or laptop without any security measures becomes exposed to the possibilities of being attacked. Linux is one of the most popular operating systems used by individuals as well as organisations and academics. To have a secured distro means to secure valuable data. In this article, we list down 10 top Linux distributions for […]","categories":["AI Trends"],"tags":["Kali Linux","Linux distros"],"author_name":"Ambika Choudhury","publish_date":"2019-06-04T12:52:52","publication_year":"2019","word_count":751,"keywords":["Kali Linux","Linux distros","programming_languages:R","AI","Git","Aim","ViT","GAN","R"],"extracted_tech_keywords":["AI","Aim","R","Git","GAN","ViT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/10-most-secured-linux-distros-for-advanced-privacy-security\/","complexity_score":3,"technical_depth":7,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":31003,"title":"Is AWS Catching Up With Other Cloud Vendors With The Launch Of New Machine Learning Inference Chip","content":"The biggest cloud provider Amazon launched a machine learning inference chip called AWS Inferentia at the AWS re:Invent conference last week. It has been designed by Annapurna Labs an Amazon-owned Israeli company which is at the forefront of building next-generation semiconductor systems. The industry cloud leader has claimed that the chip is geared at larger workloads which consume entire GPUs or require lower latency. According to the news statement, the chip provides hundreds of teraflops per chip and thousands of teraflops per Amazon EC2 instance for multiple frameworks, including TensorFlow, Apache MXNet and PyTorch, and multiple data types, including INT-8 and mixed precision FP-16 and bfloat16. The release of the chip is not a direct competition to NVIDIA, Intel or AMD, since Amazon will not sell the chips commercially, but only to their cloud customers. However, the release is an indication of Amazon’s intention of deepening its hardware muscle and their take on Google Cloud which released a third generation TPU. Experts are wondering why Amazon has built chips which are designed for specific workloads when general-purpose processors can support a wider class of workloads. Hardware specialisation can help enterprises provide better services to their customers at a lower latency. It is also believed that the move will also help Amazon attract more developers to its AWS platform by broadening its range of tools and services. The chip is aimed at inference — the part where the work actually gets done. While training has been receiving a lot of attention and has been pegged as a big part of the machine learning process, inference is what powers the core machine learning services — like object recognition in videos, recognising speech in speech recognition and text in text recognition. Inference drives core AWS ML services like Translate, Poly and Lex. Why Amazon Wants To Dive Deep In Hardware Specialisation ML Chip Can Tackle Specialised Workloads: The reason why Amazon wants to deepen hardware specialisation on the server side of computing is because ML workload usually requires more server resources than all the present forms of server computing combined. Of late, there has been an exponential rise in different kinds of workloads, for example, autonomous driving which requires more resources than other workloads. Cloud Providers Look For Big Gains From Specialised Processors: Amazon engineer James Hamilton emphasised on the fact that machine learning applications are finding plenty of uses in practically every business — finance, manufacturing, healthcare, insurance and even heating and cooling — which means businesses are ready to bet big on ML. Lower The Costs For Machine Learning: Leading cloud providers, especially Google, Microsoft and Alibaba, are focused on making it easy for customers to deploy and scale ML and drive down costs for ML workloads. Given how ML has immediate applicability in about every domain, cloud companies look for massive gains from specialised hardware which can support a wider range of workloads. Lower Latency, Lower Power Consumption, Less Cost: A large number of enterprises and startups look for hardware which is optimised for a specific workload. AWS Inferentia will reportedly deliver high throughput and a low latency inference performance at an extremely low cost. Outlook Market analysts cite that more machine learning applications are being run on AWS as compared to other cloud computing platforms. According to a blog, AWS has over 10,000 active machine learning developers and a sizable number of cloud-hosted TensorFlow workload. Both the Inferentia hardware and software meets a wide number of inference use cases and also supports ONNX which is the industry’s most commonly used exchange format for the neural network. It also interfaces natively with all the most commonly used and popular frameworks such as MxNet, PyTorch, and TensorFlow, providing customers with choices. While the move has been widely positioned as AWS catching up with its cloud competitors in the AI chips market, we believe hardware specialisation will play a crucial role in winning businesses and driving big gains. By adding the new machine learning inference chip on the ML platform in the cloud, AWS will substantially reduce the cost of deploying ML inference at scale.","excerpt":"The biggest cloud provider Amazon launched a machine learning inference chip called AWS Inferentia at the AWS re:Invent conference last week. It has been designed by Annapurna Labs an Amazon-owned Israeli company which is at the forefront of building next-generation semiconductor systems. The industry cloud leader has claimed that the chip is geared at larger […]","categories":["AI Features"],"tags":["AWS","AWS cloud"],"author_name":"Richa Bhatia","publish_date":"2018-12-04T04:34:22","publication_year":"2018","word_count":683,"keywords":["machine learning","TPU","AWS","AI","neural network","TensorFlow","ML","PyTorch","cloud computing","Aim","AWS cloud"],"extracted_tech_keywords":["AI","machine learning","ML","neural network","Aim","TensorFlow","PyTorch","cloud computing","AWS","TPU"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/is-aws-catching-up-with-other-cloud-vendors-with-the-launch-of-new-machine-learning-inference-chip\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":6,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10167163,"title":"Sam Altman Wants to Play Cricket with Sachin Tendulkar","content":"OpenAI chief Sam Altman wants to play cricket with Indian batting legend Sachin Tendulkar. A few days ago, Tendulkar shared Ghibli-style images of himself from the 2011 World Cup victory alongside the trophy. Altman replied, “want to play?” He even shared his own Ghibli-style artwork, depicting himself playing cricket in an Indian jersey. Since OpenAI launched image generation in GPT-4o, social media platforms like X and Instagram have been flooded with Ghibli-style images. Users have been sharing their personal photos in anime style, making it a viral trend. “The ChatGPT launch 26 months ago was one of the craziest viral moments I’d ever seen, and we added one million users in five days. We added one million users in the last hour,”  Altman said in a post on X. India emerged as one of the largest user bases for this trend, catching even Altman’s attention. “What’s happening with AI adoption in India right now is amazing to watch. We love to see the explosion of creativity—India is outpacing the world,” Altman wrote in a post on X. Notably, the image generation capability is rolled to free users as well. It seems that OpenAI is now aggressively targeting the Indian market, as India is the second-largest user base of ChatGPT. Altman also shared Indian Prime Minister Narendra Modi’s Ghibli-style images featuring the Indian flag on X. OpenAI is also reportedly in talks with Reliance Industries about a potential partnership to integrate ChatGPT and other AI tools into the latter’s businesses. The Sam Altman-led company is seeking to leverage Reliance’s businesses, such as wireless carrier Jio, to distribute or sell its AI services, possibly including ChatGPT. The startup has considered reducing ChatGPT’s subscription price in India by as much as 85% to attract more paid users. Currently priced at $20 per month, the subscription is costly relative to India’s median salary. A potential partnership with Jio, Reliance’s telecom subsidiary, could allow ChatGPT to be bundled into mobile plans at a lower rate. OpenAI recently raised $40 billion at a $300 billion post-money valuation. The company stated that this funding will help push the frontiers of AI research, scale compute infrastructure, and deliver increasingly powerful tools for the 500 million people who use ChatGPT every week.","excerpt":"“What’s happening with AI adoption in India right now is amazing to watch. We love to see the explosion of creativity—India is outpacing the world.”","categories":["AI News"],"tags":["OpenAI"],"author_name":"Siddharth Jindal","publish_date":"2025-04-03T09:49:12","publication_year":"2025","word_count":373,"keywords":["Go","ChatGPT","OpenAI","AI","GPT-4o","RAG","GPT","ViT","R","startup"],"extracted_tech_keywords":["AI","GPT-4o","ChatGPT","OpenAI","RAG","R","Go","GPT","ViT","startup"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/sam-altman-wants-to-play-cricket-with-sachin-tendulkar\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10042784,"title":"Register For This Webinar: Why Modernising Data Platform Matters &#038; Why Now?","content":"Big Data is exploding with around 2.5 quintillion bytes of data generated every day. Traditional data platforms suffer from scalability, flexibility and latency issues and are not equipped to support AI\/ML workloads due to inherent structural and performance limitations. It’s critical for companies to move beyond their legacy architectures to meet modern enterprise requirements. Taking the cue, many firms are augmenting their data warehouses and deploying data platforms using SingleStore’s unified database for data-intensive applications. The move has helped businesses dramatically improve their performance and accelerate the speed of analytics. In association with SingleStore and CRG Solutions, Analytics India Magazine is organising a free live webinar to understand how the company enables fast analytics. The webinar will use a few customer case studies to demonstrate how some enterprises have improved their performance through data warehouse augmentation. The webinar will cover: How can organisations achieve fast analytics by augmenting data warehouses with SingleStore?Business use cases highlighting how the industry leaders are driving 10-20x the performance through DW augmentation.How SingleStore and CRG Solutions have partnered to help enterprises leverage the world’s fastest unified database for data-intensive applications. Register Now Who should attend? CIOs & IT Managers VPs\/Directors\/Heads of Data & Advanced AnalyticsVPs of Data Systems & DatabaseDirectors of Technology & EngineeringDirector\/ Manager of Business Intelligence & Data EngineeringChief Data, Analytics & Digital OfficersHead of Product Analytics, Data Science and EngineeringSoftware architect; Data engineer; Technology architectData Analysts & Data ScientistsWorking professionals with analytics in the job description Why should you attend? To understand the importance of modernising data platforms.To understand the dynamics of augmenting data warehouses.Why are businesses leveraging SingleStore modern cloud architecture to augment their existing data warehouse technologies? Register Now Session Speakers: Atul Vaidya, the Vice President – India & ASEAN Business at CRG Solutions Atul leads the digital transformation business for CRG in India & ASEAN region. With extensive experience in visual analytics, machine learning, agile DevOps, Atul helps organisations think bigger and bolder, scale their analytics, and assist in automation and DevOps pursuits with the right talent and curated technologies. He has scaled the CRG business with clients spanning across a broad spectrum of industries, including technology, BFSI, CPG, retail, telecom, pharma and others. An engineer and MBA by education, Atul is fascinated by the world of strategy and its execution, and closely follows the “Playing to Win” framework by Roger Martin. Fahad Khan, Senior Partner Engineering APAC at SingleStore With ten years of experience in data solutions, Fahad Khan is currently working as the Senior Partner Engineering APAC at SingleStore. He helps organisations build effective data and analytics strategies. Prior to SingleStore, Fahad has worked in Qlik as a Principal Solution Architect and in Brady Corporation as a Sr. Software Engineer. Fahad holds a mechanical engineering degree from Visvesvaraya Technological University. Date: 15th July 2021 Timing: 4 PM to 5 PM (IST) Register Now","excerpt":"In association with SingleStore & CRG Solutions, AIM is organising a webinar to understand how the company enables fast analytics.","categories":["Deep Tech"],"tags":["modern database","modernising data platforms","singlestore","SingleStore database","SingleStore modernising data platform","SingleStore unified database"],"author_name":"Sejuti Das","publish_date":"2021-07-02T17:00:00","publication_year":"2021","word_count":476,"keywords":["SingleStore database","data science","machine learning","AI","singlestore","ML","Scala","Git","SingleStore modernising data platform","RAG","Ray","modern database","analytics","modernising data platforms","R","SingleStore unified database"],"extracted_tech_keywords":["AI","machine learning","ML","data science","analytics","Ray","RAG","R","Scala","Git"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/register-for-this-webinar-why-modernising-data-platform-matters-why-now\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":24880,"title":"How Analytics Played A Key Role In L&#038;T Infotech’s Race To The $1 Billion Mark","content":"Global technology consulting and digital solutions company Larsen & Toubro Infotech (LTI) announced their Q4 FY18 and FY18 results recently. Announcing their Q4 results, LTI posted a revenue of $309 million, with 5.3 percent quarter-on-quarter growth. Sanjay Jalona, CEO and MD at LTI said in a statement, “Our outstanding growth of 5.3 percent in Q4 is a result of a broad-based performance across all verticals. We have delivered an industry-leading growth of 16.7 percent for the full fiscal year with digital revenues up 42 percent year-on-year. Our sustained investments in exponential technologies are establishing us as the digital partner of choice for our customers worldwide.” The continued momentum at LTI is due to the two large deals worth over $50 million in Q4. Besides this, LTI also has a clutch of deals under its belt in automation, multiple digital initiatives, analytics engagement, migration to the cloud, orchestrating data, and modeling a variety of analytic use cases. One of the most notable deals bagged by LTI was the $100 million contract to deploy advanced technology in India’s taxation system. Jalona said that the multi-year deal involves high volume digital dealing and creating a semantic web. He added that LTI would operate on a “build, own, operate, transfer” (BOOT) model. Dinanath Dubhashi, MD and CEO at L&T Finance Holdings Ltd, told a newspaper that the significant improvement in profit was also realised through the ‘retailization’ of the portfolio. The growth in revenue was attributed to strengthening competencies across products, using data analytics as a differentiator. How L&T Is Powering Growth Through Big Data And Analytics Deep Talent Bench: The second half of 2017 saw the global consulting major host a massive recruitment drive. L&T beefed up the talent bench in areas like big data, ETL (Informatica and Ab Infinito), data warehousing, database, and customer-centric roles. Leading Analytics Vendor: L&T’s outstanding growth is also attributed to its position as major a vendor for big data and analytics capabilities. They were also recognised for blockchain consulting and POC developments. Thanks to their global best practices, deep industry expertise and faster turnaround cycles, the IT consulting bellwether won several key Indian government projects. Global Deals: Some of the major deals won by LTI are: Transformation engagement for a leading African bank to transition its core banking system Application management and IT operations support services deal for a US-based home appliance manufacturer IT infrastructure and security management partner for a US-based industrial manufacturing company End-to-end IT infrastructure outsourcing engagement with a European auto major Selected as a preferred partner to deliver multiple digital initiatives and empower client’s broader FinTech ecosystem for a global Fortune 100 bank End-to-end service engagement with a financial regulator for maintaining the business applications of its IT division Migration of workload to Oracle Cloud for European manufacturing company Execution of SAP-led transformation engagement for leading European sourcing and services company ExxonMobil chose LTI and group company LTTS to digitise geoscience content by leveraging automation Key Government Projects: In January 2018, LTI won another government project — an end-to-end Power Portfolio Management Solution for Madhya Pradesh Power Management Company Limited (MPPMCL). The project required integrating automated business modules and robust analytical capabilities for faster and accurate decision-making on the sale and purchase of power. LTI’s Power Portfolio Management Solution would optimise functions like HR, finance, procurement, demand forecasting and tender creation. Building Thought Leadership In Robotics: L&T has also built considerable thought leadership in robotics, and according to the company website, it is already getting traction. With the overall robotics market reaching $67 billion, L&T is betting big. As stated on their blog, LTI collaborated with Purdue Robotics Accelerator and US-based Purdue University to gain insights. This played a key role in setting up the robotics lab at their Bengaluru campus. Outlook As enterprises move towards intelligent automation, onboarding of cognitive solutions, and AI, leading providers will have to build capabilities across these units. Analytics, automation and AI will be key to digital transformation, and soon, service providers will start to build the value chain across these platforms. The service provider landscape will undergo rapid transformation with vendors investing in broader capabilities to stay competitive.","excerpt":"Global technology consulting and digital solutions company Larsen & Toubro Infotech (LTI) announced their Q4 FY18 and FY18 results recently. Announcing their Q4 results, LTI posted a revenue of $309 million, with 5.3 percent quarter-on-quarter growth. Sanjay Jalona, CEO and MD at LTI said in a statement, “Our outstanding growth of 5.3 percent in Q4 […]","categories":["IT Services"],"tags":["informatica","L&amp;T Infotech","L&amp;T technology services"],"author_name":"Richa Bhatia","publish_date":"2018-05-25T09:46:29","publication_year":"2018","word_count":692,"keywords":["big data","Go","API","informatica","ELT","AI","L&amp;T Infotech","ETL","Git","RAG","L&amp;T technology services","analytics","R"],"extracted_tech_keywords":["AI","analytics","RAG","R","Go","Git","API","big data","ETL","ELT"],"url":"https:\/\/analyticsindiamag.com\/it-services\/how-analytics-played-a-key-role-in-lt-infotechs-race-to-the-1-billion-mark\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10102464,"title":"Intel Collaborates with Indian Manufacturers to Make Laptops in India","content":"Intel joins the Make in India brigade. The company has unveiled a strategic partnership with eight prominent Electronics Manufacturing Services (EMS) companies and Original Design Manufacturers (ODMs) to bolster laptop manufacturing in India. The move aims to harness Intel’s extensive industry knowledge to lay the groundwork for a robust laptop manufacturing sector in the country, in alignment with the Make in India initiative. The collaborative effort with Intel involves firms like Bhagwati Products Ltd, Dixon Technologies India Ltd, Kaynes Technology India Ltd, Optiemus Electronics Ltd, Panache Digilife Ltd, Smile Electronics Ltd, Syrma SGS Technology Ltd, and VVDN Technologies Private Ltd. For some of these companies, this venture signifies their inaugural foray into laptop manufacturing, reflecting Intel’s commitment to empower the Indian manufacturing ecosystem to cater to both domestic and global demand. As part of this collaboration, Intel will leverage its expertise to facilitate the production of complete entry-level laptops in India, employing state-of-the-art Surface Mount Technology (SMT) assembly lines, implementing quality control processes for components, and benchmarking finished products. Intel also offered support to ODMs across both Semi Knocked Down (SKD) and Completely Knocked Down (CKD) manufacturing processes. “It is our Prime Minister’s goal that the Indian Electronics Ecosystem should have deep and broad capabilities, and that Indian Electronics Manufacturing Companies should grow, scale, and expand their footprint as trusted players in the Electronics Global Value Chains,” stated Rajeev Chandrasekhar, Minister of State for electronics and IT, skill development, and entrepreneurship. “By enabling the laptop manufacturing process – from surface mount technology assembly to finished product – we are not only meeting the demands of the Make in India initiative but also contributing to the technological progress of the nation,” remarked Santhosh Viswanathan, VP & MD, India region, Intel. Intel is set to host the India Tech Ecosystem Summit in November, which will bring together numerous local manufacturers to showcase a broader range of devices manufactured in India.","excerpt":"The company has unveiled a strategic partnership with eight prominent EMS companies and ODMs to bolster laptop manufacturing in India.","categories":["AI News"],"tags":["AI in manufacturing","Intel"],"author_name":"Mohit Pandey","publish_date":"2023-11-03T14:50:29","publication_year":"2023","word_count":318,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","RAG","Ray","Aim","Rust","AI in manufacturing","R","programming_languages:Rust","Intel"],"extracted_tech_keywords":["AI","Aim","Ray","RAG","R","Go","Rust","programming_languages:R","programming_languages:Go","programming_languages:Rust"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/intel-to-make-laptops-in-india\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":39131,"title":"5 Latest Machine Learning Job Openings With Leading Indian Companies","content":"Photo by Simon Abrams for UnsplashAnalytics India Jobs offers a new platform for the data science community in India by bringing to the fore latest job openings across the country. Here are few machine learning job openings in Pune, Bangalore and Hyderabad: AI Architect @ ZS, Pune ZS‘s Software Development Team designs, implements, tests and supports high quality products used by hundreds of companies and thousands of end users to make critical decisions and manage sales operations. Requirements: Knowledge and experience in some of the key AI platform, e.g. AWS Sagemaker, IBM Watson, Microsoft Azure, Google Api.Ai, Facebook Wit.Ai, Chatbots using Microsoft Bot Framework Serve AI\/ML models in enterprise grade technology platforms through microservices Tensorflow, Caffe, CNTK, commercial technologies\/platforms, etc Experience working in a DevOps environment, and using industry standard tools (GIT, JIRA, Teamcity, etc.). Able to explain technical concepts in a non-technical language Solid hands-on experience in the Artificial Intelligence platforms with understanding of the end to end life cycle of AI projects. Very solid proven hands-on experience on UI technologies (AngularJS, ReactJS, StencilJS etc), Java\/ J2EE, Kubernetes, Spring framework, Microservices, REST API’s, Kafka etc. Exposure in Data Governance and management Exposure to Hadoop data science workbench solutions Preferred experience in any one of the integrated AI products like: Microsoft Azure ML, AWS ML Solid experience on any one of the ML pipeline solutions like DataiKU, Anaconda, KNIME and auto ML solutions like H2O.ai or DataRobot or Firefly.ai Apply here Machine Learning & Analytics Professional @ Careernet Consulting, Bangalore A potential candidate would be working for one of the clients who are into Banking Sector. Requirements: Bachelor’s \/ Master’s in Computer science Engineering OR Masters in Statistics\/Mathematics Experience in Analyzing the Medium Complex Problems and translate it into Analytical Approach Experience in Statistical Learning : Diagnostic Analytics, Simulation and Predictive Modeling, Time Series, Dynamic\/Causal Model, Statistical Learning, Guided Decisions Experience with big data analytics and advanced data mining techniques to analyze data R, Python Experience with SQL and relational databases, data warehouse Hadoop(Hive, Pig, Map Reduce, HQL) Experience with statistical programming languages – SAS SAS EG, Enterprise Miner, Text Miner, KXEN, OPL\/CPLEX, SPSS,R studio, Revolution R Experience with data warehouse platforms – Teradata \/ GreenPlum, HANA Apply here Deep Learning and Computer Vision Expert @ Pratibha Analytics, Hyderabad The candidate will be part of a team that consists of highly qualified and dedicated professionals who are focussed on value creation for clients through automation, data and decision sciences. The main role of the team is to apply leading Advanced Analytics, Robotics Process Automation, Machine Learning and other AI frameworks, tools and techniques to solve clients issues and help them gain powerful new insights to improve revenues, profitability and reduce costs by solving for a range of KPIs and value drivers based on the client industry and functional domain area. Requirements A degree in Msc. Stats, Phd in Statistics, Mathematics, Engg. in Computer Science, MCA, Masters in Computer Science or other business and engineering graduate streams with experience in Data Science and Advanced Analytics implementations. 2-8 years experience, with strong knowledge and skills in statistics, mathematics and computer science. Deep skills in Computer Vision solutions using CNN, RNN and other ensemble algorithms and methods. Experience in design and delivery of leading Analytics \/ AI solutions to clients by applying your deep skills in above mentioned disciplines including Advanced Analytics, Machine Learning and Deep Learning. Python, Tensorflow, Keras, Pytorch, Caffe, R, along with strong understanding of all key market leading Advanced Analytics, Machine Learning and Deep Learning techniques, frameworks, methodologies and tools. Apply here Principal Machine Learning Engineer @ Leben Care Technologies, Hyderabad Founded in late 2016, Leben Care offers automated medical image analysis algorithms that improve access and quality of diagnosis across areas of life sciences. Their mission is to develop artificial intelligence products and solutions that improve access and quality of diagnosis across areas of life sciences. Principal ML engineer would be part of  a world-class AI RnD team to enhance the Netra.AI platform enabling next generation of medical imaging diagnostic for Ophthalmology. Requirements Master’s Degree or PhD – Computer Science, Artificial Intelligence. Hands on Experience in implementing deep learning architectures using latest machine learning tech and frameworks. Existing track record as a researcher in machine learning with published scientific journal. Design and build novel machine learning models to solve unique medical problems and improve patient outcomes. Test and evaluate algorithms on large medical datasets to prove robustness Deliver high quality and production-ready code Solid Python and C++ experience and Linux user. Solid mathematical background. Experience with a vast set of computer vision libraries. Deep understanding and hands-on experience in state-of-the-art Medical Image Analysis algorithms will be a plus. Apply here Data Scientist @ Prescience Decision Solutions, Bangalore Prescience is a fast growing, focused Advanced Analytics company that helps enterprises become more PRESCIENT (predictive) by gaining meaningful business insights and develop optimized solutions through careful analysis of data. Requirements: Machine Learning (ML) algorithms such as Regression, Decision Tree, Logistic Regression, K-Mean and Markov Decision Processes Application of Neural Networks, in for example Deep Learning for text, image classification Natural Language Processing (NLP) Search and the use of Search for pre-processing and integration of NLP and Machine Learning Expertise in Python or R Experience with NoSQL and SQL Apply here","excerpt":"Photo by Simon Abrams for UnsplashAnalytics India Jobs offers a new platform for the data science community in India by bringing to the fore latest job openings across the country. Here are few machine learning job openings in Pune, Bangalore and Hyderabad: AI Architect @ ZS, Pune ZS‘s Software Development Team designs, implements, tests […]","categories":["AI Hirings"],"tags":["devops journal","indian statistical service","machine learning jobs","medical image processing companies"],"author_name":"Ram Sagar","publish_date":"2019-05-14T12:44:35","publication_year":"2019","word_count":880,"keywords":["machine learning jobs","devops journal","data science","artificial intelligence","machine learning","AI","neural network","ML","computer vision","NLP","deep learning","analytics","indian statistical service","medical image processing companies"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","neural network","NLP","computer vision","data science","analytics"],"url":"https:\/\/analyticsindiamag.com\/ai-hiring\/5-latest-machine-learning-job-openings-with-leading-indian-companies\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":6,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10018580,"title":"Edge AI Company AlphaICs Raises $8 Million In Series B","content":"One of the leading edge AI technology companies from Bangalore, AlphaICs announced today that it has secured $8 million funding. The Series B round was led by Endiya Partners and Emerald Technology Ventures, with the participation of existing Series A investors ReBright Partners and 3One4 Capital, along with Aaruha Technology Fund, IREON Ventures, and others. With the growth in popularity of Deep Neural Networks, there has been a huge demand for running such networks on edge devices in real-time. The company designs and develops high-performance AI Chips for Edge Computing. It will use the funds to tape-out the Gluon AI chip, to develop the software stack and to build system solutions for its target markets. AlphaICs’ Real AI Processor (RAPTM), based on a proprietary highly modular and scalable architecture, enables AI acceleration for low power edge applications, as well as high-performance edge data centres. AlphaICs’ architecture provides best-in-class inference performance, and it is equally suited for edge learning. The rapidly developing field of Edge learning promotes privacy, enables automated labelling, and facilitates continuous learning of new scenarios. Pradeep Vajram, Chairman & CEO of AlphaICs said, “AlphaICs innovative architecture will empower system integrators to create AI solutions, with a short time-to-market; while staying within the systems cost and thermal constraints.” He added, “This funding will help us bring our first inference co-processor to the market for vision applications with low latency requirements. We are also working with strategic partners to bring innovative solutions to the Industrial, Automotive, and Surveillance markets.” Adding to this comment, Michal Natora, Investment Director at Emerald Technology Ventures said, “We observed a big need in the industry for Machine Learning applications at the edge. AlphaICs’ technology offers significant performance advantages for edge inference as well as for edge learning solutions.” He added, “The differentiated technology and the high calibre team led by Pradeep Vajram were two key criteria driving Emerald’s investment.”","excerpt":"One of the leading edge AI technology companies from Bangalore, AlphaICs announced today that it has secured $8 million funding. The Series B round was led by Endiya Partners and Emerald Technology Ventures, with the participation of existing Series A investors ReBright Partners and 3One4 Capital, along with Aaruha Technology Fund, IREON Ventures, and others. […]","categories":["AI News"],"tags":["edge AI","Edge AI Chip","edge AI hardware","Edge AI inference"],"author_name":"Ambika Choudhury","publish_date":"2021-01-20T16:25:08","publication_year":"2021","word_count":314,"keywords":["API","funding","machine learning","Edge AI inference","programming_languages:R","AI","Edge AI Chip","neural network","edge AI hardware","Scala","edge AI","edge computing","R"],"extracted_tech_keywords":["AI","machine learning","neural network","edge AI","edge computing","R","Scala","API","funding","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/edge-ai-company-alphaics-raises-8-million-in-series-b\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10002958,"title":"10 Indian Startups That Are Leading The AI Race: 2020","content":"The number of AI startups in India has increased tremendously over the years. Apart from being adopted in major industries, Artificial intelligence has become a way of doing business in other niche areas such as farming or even security. To recognise the unconventional startups in the AI space, Analytics India Magazine comes with a list of 10 such exceptional startups that are leading the AI race every year. In this year’s list, we have covered startups that are not more than 3 to 4 years old and have headquarters in India. Most of these startups are funded externally and are working hard to bring about exceptional transformation in the Indian tech ecosystem. Please note that this is not a ranking, and the startups are listed in alphabetical order. Here is last year’s list. 1| Ajna AI Chennai-based Ajna, pronounced “Agnya” was founded in May 2020. This AI startup aims to build SaaS products that deliver a vision-based video Analytics solution. The solutions currently provided by the startup include crowd safety intelligence, automated queue management system, and retail video analytics system. It uses computer vision to deploy real-time, edge-processing detection & tracking algorithms, that detect and analyse customer behaviour for analytics and safety solutions. It can carry out tasks such as mask detection, social distancing, people counting, safety reporting, crowd density analysis and risk prediction. Focused primarily on the retail industry, they have been able to solve problems such as limited information about customers and their behaviour in-store, low scalability due to manual-store analytics, lost potential revenue due to shrinkage and more. The Founding team: Sriram Sankar, Venkat BN, Sidhdharth Sivasubramanian, Dr Arangannal. The team is currently 7-people strong. Funding: The startup is bootstrapped. 2| Agricx Agricx is a Google launchpad incubatee and has developed first of its kind mobile-based AI product “Procure+” that leverages machine learning for real-time quality inspection of agri produce & delivers actionable insight to stakeholders remotely. This Mumbai-headquartered startup is addressing the problem in the food industry such as finding the right buyer\/seller, ensuring efficient procurement, increasing recovery, and most importantly the ability to provide objective feedback across the supply chain. It was founded with an aim to address disputes between trade buyers and sellers. Agricx enables a frictionless transaction across the agro produce supply chain. With its patent-pending SaaS stack helps enterprise and growers digitise quality assessment at affordable prices. Using AI, it has converted a regular smartphone into an advanced quality assessment tool. The Founding Team: Saurabh Kumar (Co-founder, CEO) and Ritesh Dhoot (Co-founder, CTO). The team has 20 members across technology and sales. Funding: The startup raised INR 3.2 crore in Seed Round from Ankur Capital, CIIE & Angel investors in Feb 2018. 3| Expertrons Calling themselves the Netflix for career hacks, this Mumbai-based startup is the world’s first AI video bot-assisted platform. Founded in 2019, the AI capabilities at Expertrons help aspirants learn from interview experience of other experts. It was founded to help college graduates have an experience of how it works at bigger organisations. They need to move beyond their colleges and access the alumni umbrella from different colleges to get the right consultation. The startup believes that the best person to tell you how to crack a particular interview is to ask someone who has actually cracked it. The AI ensures that video-bots of mentors are available 24×7 for interview question-related queries. Expertrons uses it’s AI recommendation engine to suggest which experts are the most suitable for a candidate – based on hometown, degree, program, preferred role, dream companies etc. The Founding Team: It was founded by Jatin Solanki (CTO & Co-founder), Vivek Gupta (CEO & Co-founder). The team currently comprises 60 employees. Funding: It raised undisclosed amount from Nikhil Vora, MD of Sixth Sense Venture, Iceland Venture Studio, LetsVenture, and Samyakth Capital. 4| Innefu An Information Security R&D company, Innefu provides cutting edge information security & data analytics solutions. With more than 100 customers in India, Bangladesh and the Middle East, including the Government of India, this Delhi-based startup has become a leading player in AI in just nine years. This 2010-founded startup was conceptualised with an aim to establish an AI-driven company focused on developing cutting edge technology to offer predictive intelligence and cybersecurity solutions. Its AI product called AI Vision offers next-gen video analytics and facial recognition solution with an accuracy of more than 98%. It leverages deep learning-based neural networks to accomplish the results. It is significantly used in security and surveillance, visitor management, employee attendance, intrusion detection, retail, satellite and drone image analysis and more. The Founding Team: Tirender Wig (Co-founder), Abhishek Sharma (Co-founder). Having started with a team of 10 people working out of the living room of a house, it now has more than 100 employees serving various geographies. Funding: Innefu Labs got funding from Indianivesh Venture Capital Ltd in the year 2017 and 2018. 5| Intello Labs Founded in 2016, Intello Labs is an ag-tech startup that provides quality assessment products for fresh produce. With its cutting edge computing vision and AI technology, it can digitise food quality and enable complete transparency in the supply chain. The image-based products help food businesses to reduce value risk and wastage in the agriculture supply chains and ensure the best quality reaches the consumers. The product called the Intello Track is an app that uses the smartphone camera to capture the image, which is then processed by cloud-based AI to generate almost-instantaneous quality reports and calls-to-action. It then grades the produce based on colour, size and visual defects. The neural network algorithm developed by the startup identifies individual units of produce and detects contours on their surface. This task requires enormous data collection and training across various machine learning models. The startup was founded to address the challenges faced by the food industry such as changes in customer buying preferences, quality expectation and food losses. Intello Labs aims to bring transparency and standardisation to quality assessment of fruits and vegetables. The Founding Team: It was founded by Milan Sharma (CEO), Nishant Mishra (CTO), Himani Shah (CFO), Devendra Chandani (Head- Business Development). The startup currently has more than 250 employees. Funding: It recently secured $5.9 Mn in Series A round of funding led by SaamaCapital, alongside global agritech funds GROWAccelerator and SVG Ventures. 6| RayReach Conceptualised in June last year with a handful of people, RayVision was developed to make CCTVs smart. As CCTV camera surveillance has become a universal norm and is being installed everywhere from factory floors, hospitals, railway stations, airports to parks, the camera feeds are usually monitored by dedicated teams which can be an intimidating task. It reads data directly from the CCTV camera feeds and then runs their proprietary algorithms combined with AI models to check for it. The AI ensures accurate surveillance eliminating human error and helps in releasing the dedicated workforce for inspection to more intelligent requirements. Having explored CCTV surveillance for some time, the startup has now ventured into addressing COVID-19 monitoring such as detecting face masks violation, crowding violation, social distancing violation, PPE violation, hand sanitiser usage and more. it lets the user receive violations notifications on Mobile App and perform Analytics on past violations using interactive dashboards. The Founding Team: Karthick Sethunarayanan (CEO and Founder). The RayReach team is currently 25-people strong. Funding: The startup is currently bootstrapped. 7| Salesken AI This Bengaluru-based startup is a group of technologists and training professionals who have come together to bring the best of AI and data sciences to sales performance. Salesken was founded in early 2019 to empower sales personnel to have productive conversations with their customers. With their conversation AI platform that analyses conversations and assists sales reps with cues or nudges in real-time, it help them to drive better outcomes. Through the use of ASR and NLP, Salesken analyses various aspects of a sales conversation like customer sentiment, the effectiveness of the sales pitch, and the fidelity to an ideal sales pitch. Through AI, Salesken provides sales reps with talking points, which are based on historical data and best practices, in real-time. This helps them engage customers better and close more deals. The startup currently works with some of the largest companies in education, finance, and technology space, in India and abroad. The Founding Team: Surga Thilakan (CEO and Co-founder) and Sreeraman Vaidyanathan (Managing Director and Co-founder). Salesken AI team is currently 50-people strong. Funding: Salesken raised its Series-A funding of 8 million USD in March 2020, which was led by Sequoia Capital. It’s investors also include Unitus Ventures and Michael and Susan Dell Foundation. 8| Spyne Founded in 2018, Gurgaon-based Spyne has developed deep tech editing solutions using computer vision and AI technologies that help with processing thousands of images at scale and bring efficiency in the photography industry. Initially launched as a content distribution platform for wedding photographers, the co-founders started exploring ways to organise data when they found the need to deal with humongous data that came in the form of managing millions of images. Spyne has since developed deep tech to automate manual processes and bring operational efficiencies in the photography industry. Some of their AI solutions include automated background removal and replacement for hundreds of background templates, automated upscaling of images, automated photo curation for quality, similarity, sharpness etc. The startup is also working on automating multiple use cases, which could bring in vast transformation in the way editing is done on the product images currently. The Founding Team: Spyne was founded by Sanjay Kumar, who is an IIT Kharagpur alumni and Deepti Prasad, who is an Economics post-graduate from Delhi School of Economics. Spyne has a team size of 42 members. Funding: Spyne raised an undisclosed amount as growth capital in its pre-series funding in March 2020, led by Angellist, Smile Group, Manish Amin (CTO, Yatra), Anil Goel (CTO, OYO), Anubhav Sonthalia (CEO, Sokrati), Deepak Mittal (CEO, TO THE NEW), and other entrepreneurs from leading companies in India. The company had raised its seed funding last year, which was led by the Smile Group and other angel investors. 9| Tericsoft Founded in 2018, this computer vision and AI company focus on offering a range of customised AI & Computer Vision Solutions for enterprises & startups. It also provides end to end computer vision AI automation platform & face recognition solutions to the clients. Tericsoft was founded with an aim to provide AI-based SAAS solutions\/products for enterprises. The visual data intelligence solution provided by the startup aims to help businesses to solve real-world problems and deploy on compute-constrained embedded devices by developing computationally efficient algorithms. Teric AI, the annotation platform by the startup trains models, tracks experiments, visualises and continuously improves model predictions. Teric AI can be hosted on the server such as Azure, GCP to have the best security and control in place. The Founding Team: The founding team consists of Anand Reddy and Abdul Rahman Janoo, who met during their time at Waste Ventures. The team consists of 21 passionate and resilient employees with 15 full-time equivalents and six interns from all around the world. Funding: Tericsoft has been bootstrapped and has raised over INR 40 lakh from friends, family and acquaintances. They understand the concept of timing well and are waiting eagerly and patiently for their next big round. 10| Vernacular AI An AI-first SaaS startup, Vernacular.ai’s suite of speech and language solutions enable enterprises to convert audio to text by applying powerful neural network models in an easy-to-use API. Founded in 2016, this Bengaluru-headquartered startup helps enterprises automate call centre queries using its multilingual voice automation platform Vernacular Intelligent Voice Assistant – VIVA. As the pandemic accelerates, it is witnessing a pronounced increase in demand and new client signups with the increased adoption of automation. VIVA, which has been deployed in major enterprise contact centres, boosts customer stickiness and loyalty through a deep understanding of the customer’s context and intent. It currently serves sectors such as banking and finance, food and beverage, travel and tourism and more. VIVA automates up to 80% of their contact centre operations. It also uses natural language understanding (NLU) and speech recognition which supports around 10 Indian languages. Vernacular.ai was founded to address the language barrier that exists in many areas. The Founding Team: The founding members of Vernacular.ai are Sourabh Gupta (Co-founder and CEO), Akshay Deshraj (Co-founder and CTO), Prateek Gupta (Sales Lead), Abhinav Tushar (Artificial Intelligence Lead), Pravendra Singh Rathore (Product Lead). The employee strength of Vernacular.ai is 58 members. Funding: Vernacular.ai recently acquired its Series A investment of $5.1 million, led by Exfinity Ventures and Kalaari Capital. AngelList, IAN Fund, and LetsVenture also participated in the round. The capital will be used to fund expansion plans into Southeast Asia and the US, and towards R&D to further enhance our offerings. Prior to this, in 2017, Vernacular.ai had raised an undisclosed amount from Kalaari Capital’s accelerator programme, Kstart.","excerpt":"The number of AI startups in India has increased tremendously over the years. Apart from being adopted in major industries, Artificial intelligence has become a way of doing business in other niche areas such as farming or even security. To recognise the unconventional startups in the AI space, Analytics India Magazine comes with a list […]","categories":["AI Features"],"tags":["ai grading","automated saas intelligence","intelligent automation for retail","Speech Analytics","supply and demand in cyber security"],"author_name":"Srishti Deoras","publish_date":"2020-07-22T13:02:46","publication_year":"2020","word_count":2149,"keywords":["data science","artificial intelligence","machine learning","Speech Analytics","AI","neural network","ai grading","automated saas intelligence","computer vision","supply and demand in cyber security","intelligent automation for retail","NLP","Aim","deep learning","analytics"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","deep learning","neural network","NLP","computer vision","data science","analytics","Aim"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/10-indian-startups-that-are-leading-the-ai-race-2020\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10167268,"title":"‘Wayfair Achieved in 2 Years What Most GCCs Take 4-5 Years to Reach’","content":"India has become a top choice for global retail companies for setting up their global capability centres (GCCs), owing to the country’s skilled workforce. Companies like Best Buy, 7-Eleven, Giant Eagle, Lululemon and Adidas established centres between 2020 and 2024. According to an ANSR report, nearly 30% of the world’s top retail and consumer goods companies have established their GCCs in India. A report by Zinnov further highlights that these centres are powered by over 34,000 professionals, driving digital transformation and innovation for their global headquarters. Among them, Wayfair, a Fortune 500 online home retailer, made its entry into the Indian market from a technology standpoint just over two years ago. Unlike traditional expansions, where companies cautiously test the waters, Wayfair took a bold approach to learn from other GCCs and take a more strategic approach. Rohit Kaila, who has led Wayfair’s technology development centre (TDC) since its inception in April 2023, discussed this with AIM. Instead of going through the usual phases of evaluation that include understanding the talent landscape and scaling gradually, Wayfair decided to leapfrog these steps and build a strong foundation from day one. What is Wayfair Doing Differently? While most GCCs take four to five years to reach significant milestones, Wayfair’s TDC has achieved them in just two. For instance, Kimberly-Clark, a pioneer and global leader in trusted baby and child care brands, has seen its team in Bengaluru grow twentyfold over the past five and a half years. Meanwhile, in its first year alone, the TDC hired around 130 team members, reaching 300 employees by the end of Q1 2024. Today, the centre has grown to 550 employees and is set to hit a headcount of 700 by 2025. One of the key factors behind this success, Kaila mentioned, is the blend of startup energy within a Fortune 500 company framework. Wayfair has built a microcosm of its global operations in India, with leaders representing every aspect of technology. This particularly represents the future of digital centres evolving into microcosms of an organisation’s headquarters, encompassing multiple departments and serving as hubs of innovation rather than merely executing predefined tasks. Beyond any doubt, Wayfair is a burning example of how the role of GCC heads in India is transforming, extending beyond operational leadership to assume more CEO-like responsibilities. Wayfair’s Generative AI For Innovation When Wayfair first established its TDC in India, the initial plan was to gradually build a data science and machine learning (ML) team, Kaila revealed. However, the importance of these capabilities soon became evident, prompting the company to accelerate its efforts. This led to the launch of their Machine Learning Sciences organisation in 2024—a significant milestone in Wayfair’s global AI journey. Since then, India has emerged as a key hub for Wayfair’s artificial intelligence (AI) initiatives. Kaila highlighted the company’s strong and growing partnership with Google—not just for cloud infrastructure but also in the realm of generative AI. These efforts are complemented by ongoing collaborations with industry leaders such as OpenAI and Microsoft. A notable aspect of Wayfair’s developer ecosystem is the widespread use of GitHub Copilot, a generative AI coding assistant that has gained immense popularity among engineers. “While Google’s Gemini Code Assist is also excellent, Copilot currently has a slight edge in popularity due to developer familiarity,” Kaila remarked. Despite this preference, Wayfair continues to engage closely with Google on various generative AI initiatives. Beyond these tools, Wayfair has a long-standing tradition of building its own machine learning models. “These don’t always have to be large language models,” Kaila explained. “We often develop smaller, purpose-built ML algorithms tailored to address specific business challenges. We’ve been doing this for quite some time.” Kaila further elaborated on the extent of AI integration at Wayfair, highlighting the company’s deployment of agentic AI—systems that can perform tasks autonomously based on goals and context. “AI plays a crucial role across multiple domains. We use it extensively in areas such as search recommendations, advertising, and customer-facing applications. These are core commercial functions where AI delivers significant value.” In addition to business-critical applications, Kaila mentioned that AI enhances internal productivity. Customer support chatbots streamline user interactions, while developer tools like GitHub Copilot and Gemini Code Assist empower engineers to write better code, faster. AI has also been transformative in tech stack modernisation. “Most companies undergo some form of tech transformation—moving from older stacks to newer ones. We’ve worked closely with Google to develop models and mechanisms that accelerate this process,” Kaila said. Internships and Talent Acquisition In terms of talent recruitment, Wayfair is setting an example by taking a slightly different approach. Wayfair’s strategy towards developing talent extends to its six-month-long internship programme, which is designed to provide hands-on industry experience. Kaila added that while Indian universities produce excellent engineers, the education is often theoretical, and Wayfair aims to address this gap by offering real-world experience. The company partners with institutions like Plaksha, BITS Pilani, and others, but also runs an open model where students from any university can apply. “Last year, 80% of our interns converted to full-time employees,” Kaila noted. He emphasised that while tier-2 and tier-3 cities offer advantages, Wayfair has no plans to open additional centres outside. “Bengaluru remains our hub, and we’ve successfully attracted talent from across India…We chose Bengaluru as our headquarters for the technology centre because the best talent in the retail e-commerce domain resides here, given the presence of major GCCs like Flipkart, Amazon, and others.” Kaila also highlighted the strategic advantages of Bagmane World Technology Centre in Mahadevpura, located between Whitefield and Outer Ring Road, which is home to major players like Microsoft, Amazon, and Google. The company had evaluated several other tech parks, but ultimately chose Bagmane for its location and infrastructure. Kaila highlighted that India’s priority is to build an organisation with independently owned charters. “Each leader here owns specific areas they are driving forward, ensuring accountability and autonomy. This eliminates the inefficiencies of constant back-and-forth coordination between centres,” Kaila explained. India, the largest TDC outside Boston, plays a significant role in this strategy. Wayfair’s Ecosystem Contribution When it comes to the future of GCCs, many leaders believe that one major challenge is talent development. With the same group of professionals moving between multiple GCCs, hiring and retaining employees can become a significant hurdle. However, Kaila pointed out that Wayfair has been able to “leapfrog others by leveraging existing knowledge and best practices”. “This isn’t something we keep secret—it’s a playbook I encourage others to follow. The more GCCs adopt these approaches, the better it is for the ecosystem.”He noted that the Indian tech ecosystem is unique in how it “feeds on itself”. Engineers move between companies, learn new skills, and contribute to the overall growth of the industry. “We don’t see this as competition. Instead, we view it as an opportunity to help the ecosystem,” Kaila concluded.","excerpt":"In its 2023 alone, the centre hired around 130 team members, reaching 300 employees by the end of Q1 2024.","categories":["GCC"],"tags":["GCC india"],"author_name":"Shalini Mondal","publish_date":"2025-04-04T14:34:34","publication_year":"2025","word_count":1142,"keywords":["data science","agentic AI","machine learning","artificial intelligence","OpenAI","AI","ML","RAG","Aim","generative AI","GCC india"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","data science","generative AI","agentic AI","OpenAI","Aim","RAG"],"url":"https:\/\/analyticsindiamag.com\/gcc\/wayfair-achieved-in-2-years-what-most-gccs-take-4-5-years-to-reach\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10085701,"title":"Union Budget 2023: What’s in Store for AI?","content":"Finance Minister Nirmala Sitharaman is all set to present the Union Budget for 2023 on February 1 in the Parliament. As Budget expectations for various sectors continue to roll in, we decided to take a look at the expectations for AI in the upcoming Budget. With India taking over the chair for Global Partnership on Artificial Intelligence (GPAI) to support responsible and human-centric development and use of AI, there is a growing need for the government to focus on the use of this technology. The AI sector has pinned a lot of hope on the Union Budget 2023 and is expecting a budget allocation in the form of incentives, tax benefits, and funding for research and development to promote the growth in the industry. Subsidising AI adoption In 2020, Prime Minister Narendra Modi said that India should become a global hub for AI in the coming years. Over the years, the Modi-led administrations have stressed heavily on the importance of technologies such as AI. A report published by the National Association of Software and Service Companies (NASSCOM) titled ‘AI Adoption Index’ said that the adoption of AI and data utilisation strategies could boost India’s GDP by USD 500 billion by 2025. Speaking in this regard, Dr Mukesh Gandhi, founder and CEO, Creative Synergies Group told AIM that since this GPAI alliance is targeted towards supporting advanced AI research, we can expect a sizable commitment in the 2023 Budget towards subsidising AI adoption across industries–especially keeping the economic forecast in mind. “AI, after all, can offer a lot to businesses, right from grassroots to advanced levels. For instance, AI can support original equipment manufacturers to automate tasks, help them draw actionable insights and direct employees towards skill-based tasks. It can also play a key role in businesses retaining and attracting more business,” Gandhi said. Shefali Pahwa, founder of BlockLegal, echoes similar sentiments. “If there is one technology area that India must absolutely prioritise to ensure its long-term progress and security, it is AI,” she said. “To guide this endeavour and spur the necessary momentum, a comprehensive policy framework paired with industry incentives is essential.” “We expect the government to come up with a policy framework and encourage the applications of AI in the form of giving grants or tax benefits,” Mr. Sangeet Kumar, Co-founder and CEO, Addverb Technologies said. R&D budget Organisations today are moving away from traditional IT models to AI-powered interconnected systems. However, achieving this at scale requires much higher supercomputing capacity and talent access than currently exists. Mahesh Kulkarni, co-founder and MD, AFour Technologies, expects the government to set aside an R&D budget for tech startups and corporate research wings to develop scalable AI tools. “Incubation centres will be invaluable for testing ideas from around the country and deploying the ones that click. Building core computing power is vital so that even smaller companies can benefit from technologies like blockchain, AI\/ML, and cloud computing,” Kulkarni said. Many industry experts also expect the finance minister to roll out new schemes that encourage industries to invest in core R&D on AI and Robotics to accelerate India’s goal of becoming a USD 5 trillion economy by FY2025. Speaking about the same, Jagdish Mitra, chief strategy officer & head of growth, Tech Mahindra, hopes that the upcoming Union Budget 2023 will be a beacon of hope for creating national R&D ideas incubators, which will nurture critical cross-disciplinary research, new ideas, and technologies through the early phase. Facilitating AI research and investment According to the NASSCOM report, India’s investments in AI technologies were expected to cross USD 880 million by 2023. While this is no doubt a significant number, India’s share in global AI investment would still be only around 2.5%, as per this report. “With the rise of new-age technologies such as AI and the proliferation of internet access as well as 5G services all across India, it is an opportune time for India to invest in AI. To facilitate the growth of AI, India still needs to improve in terms of high-quality research,” Mohammed Roshan, co-founder and CEO at GoSats, said. Further, Roshan is of the opinion that the government needs to look into investments to facilitate this and make it seamless for organisations to move from data and technology silos to building specialised AI capabilities at scale across sectors. Bridging AI talent gap The AI sector has massive requirements for skilled professionals. However, there is also a wide talent gap prevailing in the industry. According to a Deloitte survey, nearly 13% of the respondents cited a lack of skilled AI talent as their top challenge. The industry is expecting the government to allocate a portion of the Budget to bridging the AI talent gap as well as upskilling and reskilling the Indian workforce. Kulkarni expects investments in training resources for IT professionals to bridge the AI skill gap. “This year will witness the proliferation of deep tech trends, and powering up IT and human assets to stay in line with those trends is critical,” he said. Gandhi also believes the government should look into supporting stakeholders who aim to open more training institutes. “These modules should equip the upcoming and existing workforce with advanced knowledge in AI\/ML, IoT, data analytics and more. In time, this could help address the 5 lakh talent gap in the IT industry,” he said.","excerpt":"The government is expected to set aside an R&D budget for tech startups and corporate research wings to develop scalable AI tools","categories":["AI Features"],"tags":[],"author_name":"Pritam Bordoloi","publish_date":"2023-01-24T12:00:00","publication_year":"2023","word_count":889,"keywords":["Go","artificial intelligence","AI","cloud computing","ML","Scala","RAG","Aim","analytics","R"],"extracted_tech_keywords":["AI","artificial intelligence","ML","analytics","Aim","RAG","cloud computing","R","Go","Scala"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/union-budget-2023-whats-in-store-for-ai\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10068664,"title":"The wait is over: Rocket Capital to launch #CryptoPrediction challenge in India — the longest blockchain tournament on financial markets","content":"Rocket Capital Investment (RCI), in association with MachineHack, is set to launch the longest blockchain tournament from June 13 to September 05 — a twelve-week hackathon to source and incentivise the best in machine learning applications for finance. The tournament is made up of a series of challenges, where the participants stand a chance to win over INR 4,00,000 cash & 200,000+ MUSA coins. Note: The MUSA coin is the first token fully dedicated to the data scientist community. Read more here. Being a Licensed Financial Institution headquartered in Singapore, RCI combines their financial expertise with external machine learning forecasts through a blockchain tournament on financial markets. The company believes that a key need for RCI is to feed its engine with thousands of machine learning predictions from data scientists to improve the accuracy of the Meta-Model Rocket used to make portfolio asset allocation. Through this competition, RCI aims to use a decentralised platform to source and incentivise the best in machine-learning applications for the finance industry. Alongside, data scientists and ML practitioners can now be rewarded weekly with MUSA coins for their data science skills. This can also help them create a decentralised track record of their performance. Start Date: June 13, 2022 End Date: September 05, 2022 Register Now All about the #CryptoPrediction challenge In this challenge, RCI asks the data scientists & ML Practitioners to submit financial predictions using the weekly dataset. RCI evaluates participants’ performance and pays them rewards at the end of the week. Let’s not forget that the MUSA token drives the challenge, where participants stake their MUSA coins to submit their predictions and earn rewards based on their stake and the performance of their submissions. The community competition gets up to 200,000 tokens and a 500 USDT prize (INR 40,000) at the end of every week, for the next 12 weeks. How does it work? The #CryptoPrediction tournament is made up of weekly challenges, where each week: RCI distributes a dataset (every Monday at 11:00 AM, till September 05) and opens the submission window. Participants can start submitting their crypto predictions of the financial markets.At the end of the week, RCI closes the submission window (every Tuesday at 5:30 AM, till September 05).Post that, the predictions are evaluated against the real-world data, and the rewards are paid out in MUSA tokens & cash amount. Dataset The dataset consists of #first column (symbol) is the ticker# ‘target_DC1’ is the log of return for the delta between the current close to the previous close# ‘target’ is the y which is ranked by Era using ‘target_DC1.’# the rest of the columns are features to be used for ML Start Date: 13th June 2022 End Date: 05th September 2022 Register Now Prerequisites for the #CryptoPrediction challenge Follow https:\/\/twitter.com\/rocketcapinv on Twitter Follow https:\/\/www.linkedin.com\/company\/rocketcapitalinvestment\/ Join the RocketCapital Discord channel at https:\/\/discord.com\/invite\/r6NbnCSrzn Open an ETH wallet and configure tournament dApp ( https:\/\/www.youtube.com\/watch?v=Lb_A8J2wJMw&t=1s) Request MUSA airdrop on this site ( https:\/\/rocketcapital.ai\/airdrop\/) Check for airdropped MUSA tokens on your wallet or competition account Use MUSA tokens for staking ( https:\/\/www.youtube.com\/watch?v=5CLKh776mtI&t=50s) Each week download the dataset and send the predictions ( https:\/\/www.youtube.com\/watch?v=A9iUuV1RVdo ) The challenge will only allow for one submission each week per wallet address (One Account Per Wallet Address), and submissions from multiple accounts will lead to disqualification. All registered users are eligible to participate in the hackathon. Privately sharing code or data outside of teams is not permitted. Rewards will be paid to the wallet address used to submit, so please coordinate with your team regarding the distribution of rewards. Each team can have a maximum of 5 members aged above 18. The one who shared the request (once accepted by another participant) will become a team leader. Only a team leader can create a team (including sharing requests & accepting members). A team or an individual participant will be treated as one entity. A single participant adding a submission will be discarded once they join a team. Changes to team members are not permitted. Once the team is formed, it is not possible to delete the team. How will you be evaluated? The submission will be evaluated using the Spearman Correlation metric. One can use ‘scipy.stats.spearmanr(prediction, target)’ to calculate the same. [mean_squared_error(y_true, y_pred, squared=False)] The Final Score represents the score achieved based on the Best Score on the public leaderboard. Rewards will be given to participants according to the above scores, considering the correlation among predictions (refer to RCI Competition for technical details). Sklearn models support the predict() method to generate the predicted values. The participant should submit a .csv file with the same number of rows of the test dataset. The submission will return an Invalid Score if you have extra rows or columns. The file should have exactly two columns (ticker and predicted value). PLEASE NOTE: Do not shuffle the sequence of the test series.If you are using pandas, use this submission code-submission_df.to_csv(‘my_submission_file.csv’, index=False) To stake MUSA coins and make submissions, participants will need a Metamask Wallet. This will help you manage your tokens and take action on the blockchain. Check out this video to learn how to set up your Metamask Wallet. You can also read the instructions here. Start Date: 13th June 2022 End Date: 05th September 2022 Register Now","excerpt":"A key need of RCI is good financial market predictions to improve the accuracy of the ML models.","categories":["Deep Tech"],"tags":["AI in finance","Blockchain","blockchain india","Blockchain Technology","crypto","crypto india","crypto market","cryptocurrencies","Cryptocurrency","cryptocurrency bitcoin","cryptography","finance","financial analytics"],"author_name":"Sreejani Bhattacharyya","publish_date":"2022-06-10T10:20:00","publication_year":"2022","word_count":875,"keywords":["API","ELT","Cryptocurrency","crypto india","cryptocurrencies","AI in finance","financial analytics","R","Pandas","data science","crypto","ViT","machine learning","Blockchain","blockchain india","AI","ML","cryptography","cryptocurrency bitcoin","finance","crypto market","Aim","Blockchain Technology"],"extracted_tech_keywords":["AI","machine learning","ML","data science","Aim","Pandas","R","API","ELT","ViT"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/the-wait-is-over-rocket-capital-launches-cryptoprediction-challenge-in-india-the-longest-blockchain-tournament-on-financial-markets\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10112033,"title":"Hugging Face Launches Chat Assistants, Takes on OpenAI&#8217;s GPT Store","content":"Hugging Face, open-source hub for AI models, has unveiled a new feature – the Hugging Chat Assistant. This latest addition allows users to effortlessly create their personalised chat assistant in just two clicks, akin to the functionality seen in OpenAI’s GPT models. https:\/\/twitter.com\/_philschmid\/status\/1753429249363452274 Similar to OpenAI’s GPTs, Hugging Face’s Chat Assistants enable users to craft custom versions of the chat interface, offering a range of customisation options. An Assistant on the platform is characterised by its name, avatar, and description, along with the flexibility to choose from various open-source language models such as Llama2 or Mistral. What sets Hugging Face’s Chat Assistants apart is their open-source nature, providing users with the freedom to utilise and share the feature without any subscription fee, in contrast to the $20 subscription charged by OpenAI for ChatGPT Plus. However, Compared to OpenAI’s custom GPTs, it’s still missing RAG, web search, actions, and a GPT builder.This puts it below the custom GPT level for now, but it’s all in the roadmap for the open-source competitor. https:\/\/twitter.com\/julien_c\/status\/1754501776835760559 In just a few days since its launch, the Hugging Face Chat Assistants platform has rapidly grown, boasting over 4,000 Assistants. Furthermore, Hugging Face plans to introduce seven additional open-source stores in the near future, promising further expansion and diversity in the available options for users.","excerpt":"In just a few days since its launch, the Hugging Face Chat Assistants platform has rapidly grown, boasting over 4,000 Assistants.","categories":["AI News"],"tags":["ChatGPT","Hugging Face"],"author_name":"Siddharth Jindal","publish_date":"2024-02-05T21:50:10","publication_year":"2024","word_count":217,"keywords":["ChatGPT","Hugging Face","API","OpenAI","AI","RAG","GPT","llm_models:Llama","R","llm_models:GPT"],"extracted_tech_keywords":["AI","ChatGPT","OpenAI","Hugging Face","RAG","R","API","GPT","llm_models:GPT","llm_models:Llama"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/hugging-face-launches-chat-assistants-takes-on-openais-gpt-store\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10162058,"title":"AWS Commits ₹60,000 Crore for Data Centres in Telangana","content":"Amazon Web Services (AWS) has committed to investing ₹60,000 crore to establish new data centres in Telangana. The announcement was made following discussions at the World Economic Forum (WEF) in Davos, which was attended by Telangana Chief Minister Revanth Reddy, Telangana IT and industries minister Sridhar Babu and AWS representatives, including Michael Punke, vice president of global public policy at AWS. The investment aligns with AWS’s broader expansion plans in India, which include developing Hyderabad’s data centre ecosystem. AWS has already invested $1 billion in three operational centres in Telangana and plans to invest an additional $4.4 billion by 2030. Emphasising the importance of the deal, Reddy said, “Global businesses like Amazon have placed their confidence in us, marking a significant step in our Telangana Rising vision.” The Telangana government is facilitating the expansion by providing additional land for AWS’s new projects. “With this deal, Hyderabad is set to become the data centre hub of India and a global leader in cloud services, including AI,” Babu said. AWS’s investments are expected to strengthen Telangana’s position as a critical player in India’s digital and cloud infrastructure growth. AWS has also announced a $8.3 billion investment in cloud infrastructure in Maharashtra as part of its expansion into the AWS Asia Pacific (Mumbai) Region. Meanwhile, Microsoft recently announced a $3 billion commitment to expand Azure’s infrastructure in the country at the Microsoft AI Tour in Bengaluru. The investment will scale Microsoft’s regional cloud infrastructure to bolster AI and computing capabilities. Moreover, Microsoft will train 10 million people in AI by 2030 as a part of its ADVANTA(I)GE INDIA initiative.","excerpt":"AWS has already invested $1 billion in three operational centres in Telangana and plans to invest an additional $4.4 billion by 2030.","categories":["AI News"],"tags":["AWS"],"author_name":"Siddharth Jindal","publish_date":"2025-01-23T15:43:26","publication_year":"2025","word_count":266,"keywords":["Go","cloud_platforms:Azure","AWS","AI","Azure","cloud_platforms:AWS","Git","GAN","cloud_platforms:Amazon Web Services","R"],"extracted_tech_keywords":["AI","AWS","Azure","R","Go","Git","GAN","cloud_platforms:AWS","cloud_platforms:Azure","cloud_platforms:Amazon Web Services"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/aws-commits-%e2%82%b960000-crore-for-data-centres-in-telangana\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10098657,"title":"Meta Could Learn a Thing or Two from OpenAI","content":"As one of the leading social media platforms, Meta witnesses a deluge of content being posted on its platforms every second. However, it’s important to acknowledge that not all of this content is guaranteed to be safe and needs to be moderated. In November last year, Meta announced that it took action against a total of 23 million pieces of harmful content created by Indians across its social media apps Facebook and Instagram. Filtering out harmful content on Meta’s apps has been a challenge for content moderators as they are continuously bombarded with disturbing and inappropriate content. Extreme visual content can include depictions or actual acts of gore or lethal violence, such as murder, suicide, violent extremism, animal abuse, hate speech, sexual abuse, child or revenge pornography  and more. BBC’s recent report highlighted that Sama, a company entrusted with the task of moderating Facebook posts in East Africa, has come to regret its decision in hindsight. Former employees based in Kenya have expressed distress due to exposure to distressing and graphic content. To address content moderation problems faced by companies, OpenAI seems to have discovered a potential solution. In a recent blog post, OpenAI stated that businesses can utilize GPT-4 for content moderation purposes. This is the first time when an LLM is being put to use for this purpose. OpenAI in their blog said that content moderation using GPT-4 results in much faster iteration on policy changes,  reducing the cycle from months to hours. It further elaborates that GPT-4 is able to interpret rules and nuances in long content policy documentation and adapt instantly to policy updates. As per OpenAI’s research findings, GPT-4 that underwent training for content moderation outperforms human moderators with basic training. However, both GPT-4 and humans fall short when compared to highly skilled and experienced human moderators. “We believe this offers a more positive vision of the future of digital platforms, where AI can help moderate online traffic according to platform-specific policy and relieve the mental burden of a large number of human moderators,” write OpenAI authors Lilian Weng View, Vik Goel and Andrea Vallone. Can Meta take inspiration from OpenAI? Meta has adopted AI for content moderation but still the social media giant hasn’t explored yet using LLM for content moderation. It uses AI tools but still largely depends on outsource partners who hire employees for content moderation. Hiring humans for content moderation has led to a lot of trouble for the social media platform as they are not able to cope up with the gory content. Meta has been trying to automate the content moderation process gradually and it would be nice if they can take inspiration from OpenAI. OpenAI has taken a lead in the positive direction by exploring the use of GPT-4 for content moderation. OpenAI is not keeping this tool to themselves and anyone with OpenAI’s API access can implement this approach to create their own AI-assisted moderation system. However, the prospect of Meta adopting the GPT-4 API to address their challenges seems improbable, given the competitive rivalry between the two companies in the realm of generative AI. It would be nice if Meta can train Llama 2 for content moderation purposes similar to what OpenAI is doing and make it open source. It goes without saying Meta has more experience in content moderation as compared to OpenAI. In December 2021, Meta came up with a new AI system using a method called “few-shot learning,” in which models start with a general understanding of many different topics and then use much fewer — or sometimes zero — labeled examples to learn new tasks. Last year, Meta launched a tool called Hasher-Matcher-Actioner (HMA) to be adopted by a range of companies to help them stop the spread of terrorist content on their platforms. It is especially useful for smaller companies who don’t have the same resources as bigger ones. HMA is built  on Meta’s previous open source image and video matching software that  can be used to flag any type of violating content. Meta believes in open sourcing tools. If Llama 2, equipped for content moderation, becomes accessible, it could prove invaluable for both companies and Meta itself. This tool would facilitate the removal of content that holds potential harm for consumers along with less dependence on human workforce for content moderation relieving them from the mental burden. Notably, Meta has also expressed its ongoing efforts to label AI-generated content across its platforms lately.","excerpt":"Meta has adopted AI for content moderation but is yet to explore LLM for the same","categories":["Global Tech"],"tags":[],"author_name":"Siddharth Jindal","publish_date":"2023-08-17T16:16:52","publication_year":"2023","word_count":743,"keywords":["Go","API","OpenAI","AI","Git","GPT","few-shot learning","generative AI","Rust","R"],"extracted_tech_keywords":["AI","generative AI","OpenAI","few-shot learning","R","Go","Rust","Git","API","GPT"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/meta-could-learn-a-thing-or-two-from-openai\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10119769,"title":"Allstate India’s Manjula Nanjappa on How Coaching Drives Organisational Growth","content":"With AI all set to radically transform over 1.1 billion roles this decade, a change in roles across diverse sectors is inevitable. This shift affects individuals from all sectors with varying expertise. In light of this, Manjula Nanjappa, the director for compute services at Allstate India, believes that embracing active coaching is not just a checkbox for organisations, but a strategic imperative to drive meaningful change and progress. “Every individual has great potential; coaching is just the process aimed at unlocking it,” said Manjula Nanjappa at the keynote of India’s biggest diversity and inclusion summit, The Rising 2024. “Coaching is definitely essential; [it’s] an integral part of the organisation if I want to outperform and be a competent leader in the market,” said Nanjappa. She explained that coaching helps individuals recognise their blind spots and become better versions of themselves. As companies strive to remain competitive in a rapidly evolving business landscape, integrating AI into coaching and learning and development (L&D) has become a strategic imperative. Karl-Ludwig Knispel, co-lead leadership, culture & development at PwC, said, “AI enables personalisation and increased efficiency that traditional methods cannot provide.” According to Knispel, AI technology offers innovative solutions. It can dramatically reduce the time needed to create content by automatically generating quizzes, tests and even entire learning modules. He explained, “Generative AI makes it possible to generate extensive simulations, presentation slides and interactive elements from just a few inputs.” Nanjappa reiterated the same belief that leveraging technology such as AI in coaching is a key driver in enabling organisations to tap into the potential of their diverse workforce. Allstate India has invested in various coaching models, such as GROW and FUEL, to provide structured employee guidance and support. Nanjappa said she herself has benefited from coaching, overcoming her fear of public speaking through practice and guidance. Measuring success Measuring the success of AI-driven coaching programs is crucial to justify the investment and ensure continuous improvement. Nanjappa recommended setting clear goals and objectives, promoting a coaching culture, and tracking metrics such as improved employee productivity, reduced skill gaps, and increased learner engagement. Implementing AI in coaching can be challenging due to budget constraints, an unsupportive organisational culture, a lack of leadership buy-in, and resistance from individuals. In such cases, Knispel advised organisations to analyse their specific L&D needs and use AI to automate administrative tasks and improve existing training infrastructure. The future of corporate learning is undeniably intertwined with AI. Knispel predicts AI-driven tools will enable greater personalisation, virtual training experiences, and accurate assessments. “With greater refinement, these tools will facilitate the automated evaluation of subjective answers as well,” he added. The business case for building a strong coaching culture is compelling. According to a study by the International Coach Federation, organisations with strong coaching cultures reported 61% of their employees being highly engaged, compared to 53% in organisations without strong coaching cultures. Meanwhile, another survey by the Association for Talent Development revealed that companies investing in comprehensive training programs, including AI-powered coaching, have 218% higher income per employee and 24% higher profit margins. Google, recognising this, has its AI-powered coaching program, which goes beyond learning a skill. As Nanjappa aptly put it, “We are in this mad world of competing with each other; you definitely need to put some method and a structure to it. Otherwise, we’ll get lost in this.” She believes by embracing AI-driven coaching, organisations can unlock the full potential of their workforce, drive sustainable growth, and navigate the challenges of the digital age with confidence.","excerpt":"“Technology is just a piece of the puzzle for individual growth in an organisation; the emphasis should also be on relationship building,” Nanjappa said.","categories":["AI Features"],"tags":["AI Impacts"],"author_name":"K L Krithika","publish_date":"2024-05-07T14:35:10","publication_year":"2024","word_count":586,"keywords":["Go","API","AI","Git","RAG","Aim","ViT","generative AI","AI Impacts","GAN","R"],"extracted_tech_keywords":["AI","generative AI","Aim","RAG","R","Go","Git","API","GAN","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/allstate-indias-manjula-nanjappa-on-how-coaching-drives-organisational-growth\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10168961,"title":"Waymo, Toyota Explore Collaboration to Accelerate AV Industry","content":"Waymo and Toyota have signed a preliminary agreement to explore a potential partnership aimed at developing a new self-driving vehicle platform. The collaboration, still in its early stages, could lead to a jointly developed robotaxi service and potentially bring Waymo’s self-driving technology to Toyota’s next-generation consumer vehicles. Waymo, the autonomous driving subsidiary of Alphabet (Google’s parent company), announced the move in a blog post. The companies indicated that they are initiating discussions on how to combine their respective strengths.”The scope of the collaboration will continue to evolve through ongoing discussions,” Waymo stated. “Toyota is committed to realising a society with zero traffic accidents and becoming a mobility company that delivers mobility for all. We share a strong sense of purpose and a common vision with Waymo in advancing safety through automated driving technology,” Hiroki Nakajima, Toyota’s executive vice president, said. Waymo’s robotaxi operations currently run in San Francisco, Los Angeles, Phoenix, Austin, and will soon expand to Atlanta. Historically, the company managed all aspects of the business, from its app to vehicle maintenance. However, Waymo has recently begun outsourcing parts of its operations. While in Austin and soon in Atlanta, Waymo has partnered with Uber. Under the “Waymo on Uber” service, Waymo maintains responsibility for vehicle testing, roadside assistance, and some customer support. At the same time, Uber, via its partner company Avomo, manages cleaning, inspections, charging, and depot operations.Furthermore, the AV market in India is experiencing steady development. Industry titans like Mahindra and Tata Motors, as well as startups like Robots, Flux Auto, and Flo Mobility, are actively making efforts in this direction.","excerpt":"‘The collaboration aims for Zero Traffic Accidents and mobility for all’","categories":["AI News"],"tags":["AV","toyota","Waymo"],"author_name":"Merin Susan John","publish_date":"2025-04-30T18:03:00","publication_year":"2025","word_count":264,"keywords":["Go","toyota","AI","programming_languages:R","R","programming_languages:Go","ai_applications:autonomous driving","Aim","Waymo","AV","startup"],"extracted_tech_keywords":["AI","Aim","R","Go","startup","programming_languages:R","programming_languages:Go","ai_applications:autonomous driving"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/waymo-toyota-explore-collaboration-to-accelerate-av-industry\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":42632,"title":"New Research Is Making Deepfake Speech Even More Real &#038; Terrifying","content":"Deepfake is a controversial topic from the very beginning. In one of our previous articles, we discussed how Deepfake is being carried out around the globe. It has already started to change a lot of things around including the television and film sector. Recently, a group of researchers from the Max Planck Institute for Informatics, Stanford University, Princeton University, and Adobe Research created an exceptional algorithm which can make flawless edits on talking-head-videos by changing the speech content. How The Model Works This novel model specifically focuses on the face and upper body of a speaker and is based on text-edits and works as transcript-based editing of the talking-head video. When the transcription is edited, the algorithm selects segments from various parts of the video with a similar motion which can be joined to create the newly edited video. The working of the model is mentioned below: Phoneme Alignment: The researchers firstly align the transcript of the speech to a talking-head video at the level of Phonemes (Phonemes are perceptually distinct units that distinguish one word from another in a specific language). This method helps in searching snippets in the video which can be later combined to create new content. 3D Face Tracking and Reconstruction: A 3D parametric face model is registered with each frame of the input talking-head video which will later help to selectively blend different aspects of the face. Viseme Search: Given an edit operation, the model performs a viseme search (Visemes are the groups of aurally distinct phonemes that appear visually similar to one another) in order to find the best match between the subsequences of the phonemes in the video. Parameter Retiming & Blending: The parametric face model is used in order to mix different properties of a face such as a pose, expressions, etc. from different input frames and then blend them together in parameter space. Neural Face Rendering: A neural face rendering approach is implied in order to synthesize photo-realistic talking-head video which matches the modified parameter sequence and thus creating a photo-realistic talking-head video frame. Applications Of The Model The researchers mainly focused to use this model for video editing and translation in the production of movies, TV shows, commercials, YouTube video logs, and online lectures as a better editing tool. Currently, the model supports three kinds of edit operations as mentioned below Add New Words: In this type, one or more consecutive words can be added at a particular point of a video. Rearrange existing words: In this type, the edit works by moving one or more consecutive words that exist in the video. Delete existing words: In this type, the edit works by removing one or more consecutive words from the video. The Other Perspective As the advancement of technology has huge and immense advantages, however, there are some people who will never stop while utilising it for bad means. The researchers raised important and valid concerns about the probability for misusing the test-based editing approach such as utilising this technology to falsify personal allegations and scandal famous individuals. One of the researchers from Stanford University stated that every advanced technology will undoubtedly attract people with negative thoughts. For these reasons, the researchers propose some guidelines such as developing forensics, biometrics and other verification methods to diagnose the manipulated videos by the viewers.","excerpt":"Deepfake is a controversial topic from the very beginning. In one of our previous articles, we discussed how Deepfake is being carried out around the globe. It has already started to change a lot of things around including the television and film sector. Recently, a group of researchers from the Max Planck Institute for Informatics, […]","categories":["AI Features"],"tags":["deep fake","DeepFake","Neural Networks","text-based algorithm"],"author_name":"Ambika Choudhury","publish_date":"2019-07-16T15:00:55","publication_year":"2019","word_count":553,"keywords":["Go","programming_languages:R","AI","text-based algorithm","programming_languages:Go","deep fake","DeepFake","R","Neural Networks"],"extracted_tech_keywords":["AI","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/new-research-is-making-deepfake-speech-even-more-real-terrifying\/","complexity_score":3,"technical_depth":5,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10138721,"title":"Meta Unveils CoTracker3 to Improve Tracking Using Pseudo Labelling Real Videos","content":"Meta, on October 16, announced the launch of CoTracker3, a point tracker model to track videos, an upgrade to its CoTracker series of models featuring advanced AI technology. CoTracker3 is designed to handle situations where tracked points move out of view or temporarily occluded to overcome challenges in tracking objects across complex scenarios. Click here to check out the GitHub repository. By introducing a semi-supervised learning method called ‘pseudo labelling’ on real videos, it allows the model to self-label parts of the data while Meta focuses on increasing the quality and quantity of training information without requiring fully annotated datasets. According to the researchers, CoTracker3 can surpass trackers trained on ×1,000 more videos through its simple semi-supervised training protocol. By tracking points jointly, CoTracker3 handles occlusions better than any other model, mainly when operated offline. Meta said this model can be used as a building block for tasks requiring motion estimation, such as 3D tracking, controlled video generation, or dynamic 3D reconstruction. The researchers also said that CoTracker3 outperformed the state-of-the-art on TAP-Vid and other benchmarks as its architecture combines several ideas from recent trackers and eliminates unnecessary components. Available on both online and offline platforms, the model can be explored live by developers and researchers on Hugging Face. This model’s utility spans multiple domains, such as augmented reality, robotics, and sports analytics, where accurately tracking object motion is essential. Meta has also made the model and associated resources available under an A-NC licence to facilitate further research.Earlier this year, Meta also introduced Video Joint Embedding Predictive Architecture (V-JEPA) V-JEPA that predicts the missing parts of videos without needing to recreate every detail. It learns from unlabeled videos, so it doesn’t require data that humans have categorised to start learning. This improves machines’ understanding of the world by analysing video interactions between objects.","excerpt":"CoTracker3 is equipped to self-label parts of the data, increasing the quality and quantity of training information without requiring fully annotated datasets.","categories":["AI News"],"tags":["AI Video Generation Models","Meta","Meta AI"],"author_name":"Tanisha Bhattacharjee","publish_date":"2024-10-17T14:38:06","publication_year":"2024","word_count":303,"keywords":["Go","Hugging Face","Meta AI","Meta","programming_languages:R","AI","RPA","Git","AI Video Generation Models","analytics","ai_frameworks:Hugging Face","GitHub","R"],"extracted_tech_keywords":["AI","analytics","Hugging Face","R","Go","Git","GitHub","RPA","ai_frameworks:Hugging Face","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/meta-unveils-cotracker3-to-improve-tracking-using-pseudo-labelling-real-videos\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":396,"title":"ISRO is ready to create history with a record launch of 103 Satellites","content":"ISRO undertakes a major initiative in the satellite launching landscape with its project in February which will see the national space agency launch a record 103 satellites. The satellites will be deployed on a single rocket in the first week of February. S Somnath, Director of the Liquid Propulsion Systems Centre, ISRO, remarks “A century is being created by undertaking this initiative of launching over 100 satellites at one go.” Earlier plans from ISRO dictated a launch of 83 satellites towards the last week of January. But, ISRO had to push the dates towards the first week of January, as 20 more satellites were added to the initial tally. There is no specific information yet regarding the number of countries who would be taking part in this launch, however, Germany and US have been in the picture all along. This mission will gain India competitive advantage in the commercial space market. The launch has been set to take place via ISRO’s workhorse rocket PSLV-C37 from its Satish Dhawan Space Centre at Sriharikota in Andhra Pradesh. Mr. Somnath adds “100 micro-satellites are scheduled for launch via PSLV (Polar Satellite Launch Vehicle)-C37. The total payload stands at 1350 kgs, of which 500-600 kgs turns out to be the satellites’ weight.” Last year, ISRO had launched a record 20 satellites at one go, but this launch marks a significant stride for the nation’s space agency, as no nation has ever undertaken a mission of such scale earlier. The previous record was held by Russia, that launched 37 satellites in one go, back in 2014. Additionally, the South Asian satellite project initiated under Prime Minister Narendra Modi is scheduled for launch in March. The satellite is a part of GSAT-9, and was earlier scheduled to be launched in December 2016. Sources indicate Afghanistan being a part of this project in its final stages.","excerpt":"ISRO undertakes a major initiative in the satellite launching landscape with its project in February which will see the national space agency launch a record 103 satellites. The satellites will be deployed on a single rocket in the first week of February. S Somnath, Director of the Liquid Propulsion Systems Centre, ISRO, remarks “A century […]","categories":["AI News"],"tags":["Isro Satellites"],"author_name":"AIM Media House","publish_date":"2017-01-24T07:48:01","publication_year":"2017","word_count":309,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","Isro Satellites","R"],"extracted_tech_keywords":["AI","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/isro-ready-create-history-record-launch-103-satellites\/","complexity_score":3,"technical_depth":5,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10008185,"title":"Microsoft Announces General Availability Of Bridge To Kubernetes","content":"Recently, Microsoft announced the general availability of Bridge to Kubernetes, formerly known as Local Process with Kubernetes. It is an iterative development tool offered in Visual Studio and VS Code, which allows developers to write, test as well as debug microservice code on their development workstations while consuming dependencies and inheriting the existing configuration from a Kubernetes environment. Nick Greenfield, Program Manager, Bridge to Kubernetes stated in an official blog post, “Bridge to Kubernetes is expanding support to any Kubernetes. Whether you’re connecting to your development cluster running in the cloud, or to your local Kubernetes cluster, Bridge to Kubernetes is available for your end-to-end debugging scenarios.” Bridge to Kubernetes provides a number of compelling features. Some of them are mentioned below- Simplifying Microservice Development When connected with the development workstation to the Kubernetes cluster, Bridge to Kubernetes helps in eliminating the need to manually source, configure and compile the external dependencies on the development workstation. Develop Microservice Apps Faster Bridge to Kubernetes extends the Kubernetes perimeter to your development workstation, allowing you to sidestep operational complexities of building and deploying your code into the cluster to test, debug and rapidly iterate. Debugging and Testing end-to-end This tool enables debugging and testing end-to-end in the context of the larger application. When the service you are debugging is called, the request is redirected to your development machine to your locally running version. Work in Isolation in a Shared Development Environment Bridge to Kubernetes supports working in isolation in a shared cluster. By selecting the work in isolation mode during configuration, Bridge to Kubernetes will setup the isolated service, as well as a specific subdomain URL to make sure that only traffic using that URL is redirected to your development workstation.","excerpt":"Recently, Microsoft announced the general availability of Bridge to Kubernetes, formerly known as Local Process with Kubernetes. It is an iterative development tool offered in Visual Studio and VS Code, which allows developers to write, test as well as debug microservice code on their development workstations while consuming dependencies and inheriting the existing configuration from a Kubernetes […]","categories":["AI News"],"tags":["developer tools","Kubernetes","kubernetes platform","kubernetes tools","Microsoft"],"author_name":"Ambika Choudhury","publish_date":"2020-09-23T12:44:39","publication_year":"2020","word_count":289,"keywords":["API","programming_languages:R","AI","kubernetes platform","R","kubernetes tools","developer tools","Kubernetes","kubernetes","Microsoft"],"extracted_tech_keywords":["AI","kubernetes","R","API","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/microsoft-announces-general-availability-of-bridge-to-kubernetes\/","complexity_score":2,"technical_depth":5,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10090629,"title":"WWDC 2023: What to Expect from Apple’s Developer Event","content":"Apple just announced this year’s Worldwide Developers’ Conference (WWDC) slated to be held on June 5. The only Silicon Valley company, which has managed to resist using the word ‘AI’ or ‘cloud’ excessively in recent times, has everyone excited about what it has in store for us this time. Apart from the visual spectacle, with WWDC, Apple is looking to set a strong foot forward with launches in product line, applications, and software support. Let’s have a look at what we can expect from this five-day virtual event: OS advancements Apple is likely to release the latest iOS, iPadOS, macOS, watchOS, and tvOS features. According to Bloomberg’s Mark Gurman, iOS 17 will feature updated Health and Messages apps, as well as significant user experience improvements, although details are scarce. Gurman also revealed that Apple has plans to revamp notifications in both iOS and iPadOS. Furthermore, a recent Apple blog post disclosed upcoming accessibility features for iOS, including innovative machine learning-driven features like door detection for visually impaired users and live captions for FaceTime calls. It is expected that Apple will reveal more accessibility features and plugins for all OS. Additionally, we can expect that the new iOS will include AR\/VR headset intergratibility, with rumours getting stronger that the company will finally release them this June. On the watchOS front, it is believed that the new upgrade will be fairly extensive with notable changes to the new interface. Gurman notes that the watchOS will have a fairly big year this year with newer and bigger updates coming to the Apple Watch hardware. Meanwhile, Bloomberg reports suggest that Apple is overhauling multitasking on the iPad to make it more akin to a laptop experience, with resizable windows. AR\/VR and rOS The AR\/VR headset is considered to be Apple’s most secretive project so far. Starting around 2015, the product has gone through several iterations, and is finally nearing completion. While the unveiling of the reality headset is expected to take another two months, we will most likely see Apple reveal some details of its new operating system, rOS, which supports these headsets. Object Capture and RealityKit will bring real-world objects into augmented reality. One example of it was also seen in WWDC 2022. With ARkit 6, Apple has already opened up several possibilities such as 4K video, 3D object capture, face tracking, motion capture, scene geometry, and more. Naturally, it is almost certain we will see more additions to ARkit in WWDC 2023. This example of Mixed Reality with ARKit 6 is so compelling #WWDC22 pic.twitter.com\/EyMLRhtEQk— Brad Lynch (@SadlyItsBradley) June 9, 2022 Hardware updates According to reports, Apple is gearing up to launch a revamped version of the Mac Pro with an Apple silicon chip. While the redesign is not expected to be significant, the updates will mainly revolve around the transition to Apple silicon. In a recent blog post, Apple highlighted that on Mac computers featuring Intel processors, the ‘About This Mac’ section displays information on the processors, including the name of the Intel chip. The Apple Watch hardware updates are expected to be huge, with rumours milling that the changes will include micro-LED display and a bigger Apple Watch Ultra screen at 2.1 inches. New look to Siri Apple recently released Transformers architecture, which is optimised for Apple Silicon — the technology behind the current LLM and GPT technology. The rise of generativeAI has everyone wondering what will happen to Siri. Now that Apple has been making strides in this direction by launching its own secure and private LLMs that run on edge within the device, the company will definitely look to power Siri with generative AI capabilities. Moreover, it is anticipated that the latest iteration of Siri will be capable of processing multimodal inputs. This means that the virtual assistant will not only be able to receive text-based inputs but also handle images, videos, and audio inputs. Maps and Safari If we ask you two areas where Apple has been lagging severely behind Google, you would most likely say map and browser. Apple will be looking to gain some edge in both these areas, considering how big a market share Google occupies here. Along these lines, Apple recently released new updates, which include high-definition and up-to-date images. The new ‘Look Around’ perspective changes while moving the map, giving it a more live-like feeling, covering a lot more details. We can expect Apple to reveal more details on what these new features are capable of. Meanwhile, Safari is another pain point for Apple. It will be interesting to see if Apple has something in store to give a better experience to Safari. A treat for developers Apple will help developers create innovative apps by giving them unique access to Apple engineers, while also introducing new technologies and tools to enable this. The company will also be opening the field for developers to participate in the Swift Student Challenge by creating an app playground on a topic of their choice. More information on this can be accessed here.","excerpt":"We will most likely get a sneak peek into what Apple has been doing on the reality headsets","categories":["AI Highlights"],"tags":["Apple","ar\/vr","Siri"],"author_name":"Ayush Jain","publish_date":"2023-04-03T15:00:00","publication_year":"2023","word_count":836,"keywords":["Go","machine learning","AI","Modal","Apple","ML","Transformers","GPT","generative AI","Siri","R","ar\/vr","llm_models:GPT"],"extracted_tech_keywords":["AI","machine learning","ML","generative AI","Transformers","R","Go","GPT","Modal","llm_models:GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/wwdc-2023-heres-what-to-expect-from-apples-developer-event\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":22488,"title":"10 Most Influential Analytics Leaders in India – 2018","content":"Analytics in Indian business scenario has grabbed higher popularity, thanks to the growing importance on driving decisions based on data. Data and analytics is playing a key role in delivering business values, which is relatively new, but an exciting space to explore. At the helm of making analytics a part of various organisations are the leaders who are working tirelessly to make analytics and data driven decisions as seamless as possible. AIM 100: The Most Influential Leaders in AI 2024As AI reshapes our world, AIM introducing the most influential global leaders in AI. Analytics India Magazine recognises 10 such influential leaders that are constantly bringing newer paradigms in the areas of analytics, artificial intelligence, machine learning, big data, across varied domains. The list is the work of over 2 months where we invited nominations from various leading organisations in India. This list is finalised primarily based on the criteria of the impact that the leader made to analytics ecosystem in India, in last 1 year. Also, uniform representation of various domains, industries etc is taken into account. Here is the list of Top 10 Influencers in the Indian Analytics Industry, presented in alphabetical order. Ashutosh (Ash) Misra Director of Advanced Analytics and Big Data, Philips Lighting Currently based out of Philips Lighting Bangalore office in India, his current work as the Director of Advanced Analytics and Big Data revolves around machine learning algorithms, AI, intelligent BOTS and complex algorithms to find solution to enterprise problems. He and his team are involved globally on supply chain, sales and marketing, and finance related advanced analytics. With more than 26 years of research and IT experience, he has worked with a large IT company for 17+ years and has handled several responsibilities in the insurance and financial services IT, including advanced analytics, big data, application development, project and program management, IT architecture consulting, Global Delivery Management, client partner role and P&L responsibilities. He has also been advisor to various fortune 500 customers on data and analytics consulting projects and has successfully delivered large programs, as a part of his previous role. He holds a Doctorate in Physics from Indian Institute of Technology Kanpur, Kanpur India. Atul Jalan Founder & CEO, Manthan A highly accomplished entrepreneur, technology visionary and fervent speaker, Atul Jalan is the founder and CEO at Manthan which is his fourth venture after MicroTrack, Cybertrek and Net Kraft. It is very unlikely that you would find him not working on something new, and his secret sauce is imagination coupled with the ability to crank, accelerate and build momentum. As the CEO of Manthan, his single objective is to keep the spirit of invention and innovation alive at Manthan. He is also the brain behind ‘Maya’, the world’s first AI-powered conversational agent for business analytics. Powered by Artificial Intelligence, Maya makes the most sophisticated analytics accessible to every user – an absolute democratisation of technology. Atul is being featured for the fourth time on this list. Atul is a well known voice on technology and is sought for his views on analytics and consumerisation of technology, what is not known though, is that he also dabbles in poetry and takes a keen interest in a wide range of subjects from movies to quantum physics. Loquacious and a polymath, Atul can hold forth on the camera obscura, Spike Jonze’s Her or the latest quadcopter. AI and how it will impact consumer businesses is his current obsession, but once the tie is off, he is equally interested in how AI will impact our daily life and human institutions. If his maverick, entrepreneurial track-record is anything to go by, this could translate into something new. And the bigger and more beautiful it sounds, the more passionate he is likely to be about it. Debashish Banerjee Managing Director, Deloitte Analytics Bringing more than 17 years of experience in variety of analytics, predictive modelling and data science projects, he has contributed extensively in the areas of advanced analytics, actuarial risk management, data mining and predictive modelling areas. Having started his career with GE, he has established and lead insurance analytics, pricing and reserving team for GE in India—arguably one of the first of its kind in India. He moved to Deloitte in 2005 with the primary goal to set up the advanced analytics and modeling practice in India, where is currently oversees and provides leadership to the Deloitte Consulting’s Data Science practices which focuses on big data analytics, predictive modeling, cognitive and AI. Majority of his project offerings are in customer, marketing and HR domains. Deba is being featured for the second time on this list. A voracious reader and traveler, he works closely with academicians as well as corporates and has been invited to many industry and academic conferences across India and USA. He is a nominated member of Board of Studies, ASIBAS and of PGDM Big Data Analytics program by IIT Kgp, IIMC and ISI Kolkata. He received a Golden Leadership award in 2016 from Amity University for his exemplary work in the FSI Analytics space. Apart from India, he looks after some global clients who are in based in USA, South Korea, Australia, and UK. Deep Thomas Chief Data & Analytics Officer, Aditya Birla Group A reputed analytics expert, thought leader and a passionate evangelist of data science, Deep’s distinguished career spans more than two decades with a proven track record in delivering sustained and increasing profitability through business transformation leveraging information and digital capabilities, advanced analytics, novel processes, and multi-disciplinary talent. As the Chief Data & Analytics Officer for Aditya Birla Group, Deep is at the helm of steering organisation-wide initiatives and transformation programs that leverage data and cutting-edge analytics solutions to enable business opportunities, growth and efficiencies across the enterprise. Prior to this, Deep was the founder CEO of Tata Insights and Quants, Tata Group’s Big data and Decision Science company. He has also held various key positions in US and India with MNCs like Citigroup, HSBC and American Express to steer their global agenda and data driven business transformation. Deep is being featured for the third time on this list. Deep is also a well-known and frequently-heard voice on analytics and emerging technologies at leading Data science conferences both in India and abroad. He is a Computer Science graduate with an MS in Information Systems and an MBA in Marketing from Arizona State and holds numerous patents in Decision Science and Information Management. Kaushik Mitra Chief Data Officer and Head of Big Data & Digital Analytics, AXA Business Services (ABS) Kaushik has the experience of working at the intersection of technology, analytics and marketing globally for around 25 years. He has assumed leadership roles across domains such as data science, AI, modelling and analytics, business intelligence and market research. In his current role, he is responsible for driving a culture of data innovation & digital transformation in the organisation, and is involved in driving GDPR implementation in ABS. Prior to ABS, he was with Fidelity Investments in Bangalore where he was instrumental in setting up their Data Science practice. His stay in the US included similar roles in global organisations like Microsoft, Intel and IBM, and also a stint in academia. With a Doctorate in Marketing from the US, his work has been published in international marketing and analytics conferences and journals including American Marketing Association, Academy of Marketing Science and Decision Sciences Institute. He is also a frequent speaker in Indian industry bodies like NASSCOM and various thought leadership forums. Paritosh Anand Vice President and Group Head – Analytics and Strategic Initiatives, Reliance Industries Limited As analytics head of India’s largest private sector corporation and a Fortune 500 company, Paritosh brings extensive experience in setting up and delivering analytics practices for companies like GE, Deloitte and Eaton Corporation. During his last 15 years of professional career, he has worked in different domains including risk management, marketing, manufacturing and human resources functions. In his current role, he is building a world class cross functional analytics and decision science practice for Reliance Industries Limited. He is also heading the digitisation vertical for Reliance Hydrocarbon business on its people processes as a strategic initiative to build an integrated digital platform based function. He is a statistician by academics and enjoys working on cross functional domains and platforms integrating big data analytics with technology to provide real time decision making solutions across business and functions. Prithvijit Roy CEO and Co-founder, BRIDGEi2i Analytics Solutions Even before Prithvijit Roy co-founded BRIDGEi2i Analytics Solutions, he had been at the forefront of analytics development in India since the early days. He was actively involved in setting up two of the largest analytics centres in India—Genpact & Hewlett Packard, before starting what is now one of the fastest growing analytics firms in the Asia-Pacific region. His journey for BRIDGEi2i began in 2011. Founded on the simple belief that data can help any organisation radically transform its business, BRIDGEi2i drives action for digital transformation across numerous industry verticals through a range of analytics solutions combined with AI-powered products. Jit is being featured for the third time on this list. With 20+ Fortune 1000 companies filling up the rows of BRIDGEi2i’s client roster, Prithvijit and his team have rapidly scaled-up operations with a sharp focus on problem-solving and consistently delivering value. His vision for BRIDGEi2i to be at the cutting edge of advanced analytics and AI has been instrumental in driving AI\/ML innovation, analytics capability development, and thought leadership in this space. Ravi Vijayaraghavan Vice President, Flipkart In his current role he and his team are responsible for leveraging data, analytics and science to drive decision-making and business impact across all areas of Flipkart. Prior to joining Flipkart, Ravi was the Chief Data Scientist and Global Head of the Analytics and Data Sciences Organisation at [24]7.ai, where he invented, developed, implemented and optimised analytics and machine learning driven solutions. He has also held a variety of leadership roles at Ford Motor Company and Mu Sigma. Ravi is being featured for the third time on this list. Ravi has 25+ patent filings, many refereed publications and has presented as invited\/ keynote speaker at several international conferences. Ravi holds a Bachelor’s degree in Engineering from the Indian Institute of Technology – Madras, a PhD from University of Wisconsin-Madison, and an MBA from the Ross School of Business, University of Michigan – Ann Arbor Ravi currently lives in Bangalore with his wife and son. Sandeep Mittal Managing Director, Cartesian Consulting Over his 17+ years in the industry, Sandeep has been a driving force in helping some of the largest organisations in India and across the globe, adopt analytics and customer centricity. He has developed an analytics practice for over 100 clients, ranging from large telecom organisations, retail brands with hundreds of stores, online businesses, financial services companies, luxury hotels and resorts, airlines, and large restaurant chains. With Cartesian he has built an organisation that has a global presence with clients in more than 10 countries. Cartesian today has a ML\/ AI practice, a Digital Analytics practice, an Innovations Lab, and is building solutions that help embed analytics seamlessly into client organisations. At Cartesian, he is responsible for making sure that organisation is continually investing in capabilities, practice building and geographic expansion while also advising clients on their analytics roadmap. A regular speaker and contributor in the field, his sessions on customer analytics, storytelling with data, and the business of analytics, are highly sought after. Sandeep is being featured for the third time on this list. Right now, he is spending time with his engineering team in building out Cartesian’s first, subscription based, AI product, aimed at improving effectiveness of marketing communication. An alumnus of the Indian Institute of Management, Calcutta, Sandeep stays connected with music – writing and recording songs in a home studio. He also runs a web comic of his cartoons, many of which are around humour in data analytics. Srikanth Velamakanni Co-founder, Group Chief Executive & Executive Vice-Chairman of Fractal Analytics Fractal Analytics, one of the most respected pure play analytics companies across the globe, aspires to power every human decision in the enterprise. Srikanth and the team at Fractal Analytics help companies leverage analytics, AI & deep learning to transform the way they make strategic, tactical and operational decisions. He is responsible for overall growth of the Fractal businesses, which include the core business of providing analytics services to Fortune 500 companies, along with Fractal’s AI based product businesses – Customer Genomics & Trial Run, Cuddle.ai and Qure.ai. He has a BS in Electrical Engineering from IIT-Delhi and an MBA from IIM Ahmedabad. A former investment banker, he co-founded Fractal more than 18 years ago. Prior to Fractal, he has worked on structured debt transactions and collateralised bond obligations at ANZ Investment Bank and ICICI. Srikanth is being featured for the third time on this list. Srikanth considers himself a lifelong student of mathematics, behavioural economics, neuroscience, consumer behaviour and enjoys speaking and writing on the power of Big Data to enable better decision making.","excerpt":"Analytics in Indian business scenario has grabbed higher popularity, thanks to the growing importance on driving decisions based on data. Data and analytics is playing a key role in delivering business values, which is relatively new, but an exciting space to explore. At the helm of making analytics a part of various organisations are the […]","categories":["AI Features"],"tags":[],"author_name":"Дарья","publish_date":"2018-03-13T04:52:56","publication_year":"2018","word_count":2175,"keywords":["data science","machine learning","artificial intelligence","AI","ML","RAG","Aim","deep learning","analytics","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","data science","analytics","Aim","RAG","R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/10-most-influential-analytics-leaders-in-india-2018\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10019282,"title":"Top Analytics Recruitment Agencies In India","content":"Hiring analytics talent is a huge investment, and thus it is crucial to get it right. With the entire process carried out by concerned technical and non-technical stakeholders, in-house recruitment is always the best option. However, sometimes this process can become lengthy and expensive. In such cases, it makes sense to outsource the hiring process to a recruitment agency. While choosing the right agency, one needs to make sure the firm understands your needs and has the right knowledge and experience in hiring talent for the analytics ecosystem. Here, we have enlisted the top analytics recruitment agencies in India. AIM-Recruits AIM-Recruits is an India-focused executive search firm specialising in hiring machine learning & Data Science professionals. Thanks to the company’s big reach and deep-domain knowledge, AIM-Recruits can help you find the right talent via a combination of executive search, hackathons, dedicated research, and employer branding. With a team of specialised analytics recruiters — armed with vast experience in multi-domain hiring for MNC Banks, IT Giants, Analytics startups, and large KPO’s– AIM Recruits is the “One-Stop-Solution” for organisations wanting to expand its Analytics, Data Science, and machine learning team in India. Contact AIM-Recruits here. Randstad Randstad is a 60-year-old recruitment agency, with a 28-year-old presence in India. The company boasts a huge global network and has expertise across different horizontals, including analytics and data science. Randstad leverages its vast experience to understand the criticality of analytics roles and can help you find the right talent aligned with your organisation’s business goals and work culture. Michael Page India Michael Page India, a recruitment agency specialising in hiring analysts, has worked with multiple MNCs and management consultancy firms across the globe. Michael Page caters to companies of all sizes and can help you hire analysts in the mid to senior-level roles. Their recruiters have a background in technology as the agency believes it is essential in hiring the right candidate. The firm handholds you through the hiring process, and provides feedback and support, at the same time. Kelly Services India Kelly Services has a decade old presence in the country. The company tracks the analytics domain globally and understands the significance of informed data-driven decision-making. With a vast talent pool, the agency helps you find talents in data modelling, SAS, Hadoop, R, Citrix, qualitative and quantitative research, credit risk, web digitisation, marketing, consumer, and BFSI. Pylon Management Consulting Pylon Management Consulting is one of India’s top recruiting consultancy firms that goes beyond traditional search methods to help organisations in hiring the right talent. Beyond the usual advertisement and online searching, Pylon leverages advanced digital tools to find the right candidate. Though Pylon is well-versed in hiring across IT roles, the firm specialises in recruiting talent in data storage, IT infrastructure, ERP\/CRM and web technologies. Longhouse Consulting Longhouse Consulting is an executive search and advisory firm specialising in the digital sector, consumer internet space, and emerging technologies, including data science and analytics. With an in-depth knowledge of the analytics sector, Longhouse services include executive search, leadership assessment, advisory, and research. Rooted in the firm’s experience with founders and investors, Longhouse has a well-developed framework for recruiting talent depending on your firm’s size. Fidius Advisory Fidius advisory is a leading executive search firm helping businesses build teams in digital, analytics, and technology domains. The company has a wealth of experience in building data science and data engineering teams. Fidius partners with various portfolio companies and leading organisations to hire the right talent across artificial intelligence, machine learning, deep learning, neural networks, natural language processing fields. With over a decade of experience in the digital domain and a vast network to boot, Fidius can help you find the right analytics talent.","excerpt":"Hiring analytics talent is a huge investment, and thus it is crucial to get it right. With the entire process carried out by concerned technical and non-technical stakeholders, in-house recruitment is always the best option. However, sometimes this process can become lengthy and expensive. In such cases, it makes sense to outsource the hiring process […]","categories":["AI Highlights"],"tags":[],"author_name":"Kashyap Raibagi","publish_date":"2021-01-29T16:00:00","publication_year":"2021","word_count":615,"keywords":["data science","machine learning","artificial intelligence","AI","neural network","RAG","Aim","deep learning","analytics","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","deep learning","neural network","data science","analytics","Aim","RAG","R"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/top-analytics-recruitment-agencies-in-india\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10019210,"title":"Machine Learning 101: Ten Projects For Beginners To Get Started","content":"Machine learning is an up and coming field with wider applications in various sectors including health, finance, retail, among others. If you are a beginner and want to pursue a career in emerging technologies like machine learning and deep learning, it’s critical to have a first-hand experience of the concepts. Here is a curated list of 10 best machine learning projects that can help beginners kick start their ML journey. 1| Sentiment Analysis of Product Reviews About: Sentiment analysis is an application in text mining and computational linguistics research to tease out the underlying sentiment in source texts. The in-depth analysis will help uncover market trends and consumer opinions, and offer insights for the overall improvement of products. Know more here. Dataset Available: Amazon Product Review: This dataset is collected from customer reviews of Amazon products. Get the data here. Twitter US Airline Sentiment: Twitter data scraped from February of 2015 about each of the major US airlines. Get the data here. 2| Stock Prices Prediction About: Predicting stock prices is a challenging task as it depends on various factors including but not limited to geopolitics, global economy, company’s financial reports and performance, etc. There are two main approaches to predicting the stock price: Technical analysis method uses metrics like closing and opening price, the volume traded, adjacent close values etc. of the stock for prediction, whereas qualitative analysis looks at external factors like company profile, market situation, political and economic factors, textual information in news, social media and even blogs by the economic analyst. Know more here. Dataset Available: Huge Stock Market Dataset: The dataset is a collection of the daily prices and volumes of all US stocks and ETFs. Get the dataset here. Daily News for Stock Market Prediction: The dataset is a collection of historical news headlines from Reddit WorldNews Channel and stock data. Get the data here. 3| Sales Forecasting About: The objective of sales forecasting is to estimate the future demand for products or services. Some standard variables used in sales forecasting are past sales data, website visits, economic trends, etc. Know more here. Dataset Available: Walmart Store Sales Forecasting: It is a collection of historical sales data for 45 Walmart stores located in different regions. Get the data here.Retail Sales Forecasting: This dataset contains a lot of historical sales data extracted from a Brazilian top retailer. Get the data here. 4| Movie Ticket Pricing Prediction About: Machine learning techniques can be used to create personalised services, such as dynamic pricing, which can be used for movie ticket booking. Know more here. Dataset Available: TMDB Box Office Prediction: In this dataset, you are provided with 7,398 movies and a variety of metadata obtained from The Movie Database (TMDB). Get the data here.Cinema Tickets: It includes historical data of sale and movies details e.g. cost, cast and crews, and other project details like schedule. Get the data here. 5| Music Recommendation About: Music recommender system can suggest songs to users based on their listening pattern. Know more here. Dataset Available: WSDM – KKBox’s Music Recommendation: KKBOX provides a training data set consisting information of the first observable listening event for each unique user-song pair within a specific time duration. Get the data here.Last.FM: This dataset contains social networking, tagging, and music artist listening information from a set of 2k users from Last.fm online music system. Get the data here. 6| Handwritten Digit Classification About: The handwritten digit recognition can identify handwritten digits. Know more here. Dataset Available: Digit Recognizer: The data files, train.csv and test.csv, contain grey-scale images of hand-drawn digits, from zero through nine. Get the data here.MNIST Database: The MNIST database of handwritten digits has a training set of 60,000 examples and a test set of 10,000 examples. Get the data here. 7| Fake News Detection About: In this project, one can use a machine learning ensemble approach for automated classification of news articles. Know more here. Dataset Available: Fake News: It includes training and a dataset with a unique id for a news article, author of the news article, among others. Get the data here.Fake News Inference Dataset: This database is provided for the Fake News Detection task. Get the data here. 8| Sports Prediction About: Sports prediction is usually treated as a classification problem, with one class (win, lose, or draw) to be predicted. In sports prediction, large numbers of factors including the historical performance of the teams, results of matches, and data on players, have to be accounted for to help different stakeholders understand the odds of winning or losing. Know more here. Dataset Available: ATP World Tour tennis data: This dataset contains tennis data from the ATP World Tour website. Get the data here.FIFA 19 Dataset: FIFA 19 complete player dataset is a collection of detailed attributes for every player registered in the latest edition of FIFA 19 database. Get the data here. 9| Object Detection About: One of the fundamental computer vision problems, object detection provides valuable information for semantic understanding of images and videos, and has many applications in image classification, human behaviour analysis, among others. Know more here. Dataset Available: COCO: COCO is large-scale object detection, segmentation, and captioning dataset. Get the data here.Oxford Pets Dataset: It is a collection of images and annotations labelling various breeds of dogs and cats. Get the data here. 10| Disease Prediction About: Traditional disease risk model uses machine learning and supervised learning algorithm on training data (with labels) for improving the models. Know more here. Dataset Available: Heart Disease Dataset: This database contains 76 biomarkers of heart disease. Get the data here.Mental Disorders: This dataset is a collection of mental disorders, impairments associated with these disorders, and their treatment patterns from representative samples of majority and minority adult populations in the US. Get the data here.","excerpt":"Machine learning is an up and coming field with wider applications in various sectors including health, finance, retail, among others. If you are a beginner and want to pursue a career in emerging technologies like machine learning and deep learning, it’s critical to have a first-hand experience of the concepts. Here is a curated list […]","categories":["AI Features"],"tags":["Machine Learning","machine learning projects","ML projects","Sentiment Analysis"],"author_name":"Ambika Choudhury","publish_date":"2021-01-28T16:00:00","publication_year":"2021","word_count":966,"keywords":["Go","Sentiment Analysis","machine learning","AI","sentiment analysis","R","ML","Machine Learning","Git","computer vision","deep learning","object detection","ML projects","machine learning projects"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","computer vision","sentiment analysis","object detection","R","Go","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/machine-learning-101-ten-projects-for-high-school-students-to-get-started\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10086956,"title":"Top 6 Web3 Alternatives to Your Favourite Web2 Platforms","content":"Online content marketplaces have solved some bigger problems by helping sellers (products or content) and consumers reach a meeting point. However, the solution offered by the Web2 platforms was not foolproof. The issue of monopoly instantly emerged when the marketplace platforms such as YouTube, Spotify, Facebook and others became too big. Decades later, we now have some solution in the form of Web3 which claims to address the challenges of Web2. For instance, a platform named DIMO claims to offer an alternative to Uber, and Hivemapper wants to make Google Maps more accessible and community driven. Let’s see what these alternatives are and how they’re different from the traditional apps. Audius Audius is a decentralised music platform which aims to give artists more control over their music and profits. Unlike Spotify, Audius empowers artists to earn 90% of their sales revenues, decide how they wish to monetize their music, and have full ownership of their content. Audius employs a peer-to-peer network and support from decentralised node operators. Additionally, Audius also features a governance token, $AUDIO, which serves as the foundation of the platform and has a current market cap of $1.2 billion. This decentralisation gives Audius the unique ability to offer artists a bigger piece of the pie and connect directly with their fans. However, Audius does have its limitations, such as a limited catalogue and brand recognition, and the inability to provide copyright support due to its decentralised nature. DIMO DIMO is a decentralised, community-driven platform that offers a new approach to the future of mobility. Unlike traditional ride-hailing apps like Uber, DIMO aims to empower drivers by allowing them to share data on their own terms and be rewarded in the long term. By connecting drivers, developers, and manufacturers, DIMO is building the largest, most useful IoT network of user-owned devices starting with cars. The platform is based on the principles of web3, where user ownership and control are at the forefront. The $DIMO token gives holders a say in the platform and a role in its future, with the ability to vote on proposals that control how the network works. The community-governed structure of DIMO allows for design and deployment of rewards programs for drivers, allocation of resources and authority to teams and individuals, whitelisting of third-party hardware suppliers, and investment of the treasury. Presearch Presearch is a decentralised search engine that aims to reward its users for conducting searches. Unlike Google, Presearch claims to allow users to switch between multiple platforms and search engines, offering a more diverse search experience. The platform operates on blockchain technology and rewards users with PRE tokens for conducting searches. However, these tokens can only be withdrawn once the user reaches a balance of 1000 PRE. Additionally, the platform only offers a maximum of 30 searches per day, which is a bummer; but it allows users to earn up to 3 PRE daily. This decentralised approach to search engines is a key feature of the web3 movement, but capping the maximum searches makes the chances of domination slim when compared to Google. Hivemapper Hivemapper is a decentralised mapping network that offers an alternative solution to the existing centralised mapping systems, such as Google Maps. Unlike the traditional mapping systems, Hivemapper is built on blockchain technology and leverages cryptocurrency to incentivize map contributors and promote map coverage, freshness, and quality. With Hivemapper, anyone with a 4K dashcam can contribute to the global map and earn a cryptocurrency, HONEY, for their efforts. The result is a constantly-renewing and high-quality map collectively owned by its contributors. Status Status is a revolutionary app that combines a secure crypto wallet, private messenger, and a Web3 DApp browser all in one platform. Like WhatsApp, Status also claims to prioritise privacy and security by offering end-to-end encryption in its private messenger. However, it also provides peer-to-peer, borderless payments and supports a wide range of digital assets, including ETH, SNT, stable coins, and NFTs. With its integrated Web3 DApp browser, users can access the growing ecosystem of decentralised applications and decentralised finance (DeFi) platforms. Status also offers a unique, state-of-the-art decentralised social network that allows users to share updates and connect with their communities without the fear of censorship. Furthermore, Status enables private account creation, allowing users to create a free account without having to provide personal information such as a phone number, email, or bank account. Braintrust Braintrust is a new talent network that operates on the principles of Web3. Unlike LinkedIn, Braintrust is decentralised, meaning that it’s owned and controlled by the community of knowledge workers who use it. This gives Braintrust a unique ability to serve the needs of its users, making it more effective and efficient than other platforms. With over 700,000 members and hundreds of Fortune 1000 companies relying on Braintrust, it’s quickly becoming one of the most trusted talent networks in the world. Braintrust is designed to help talented knowledge workers connect with the world’s leading companies, with BTRST token available on Coinbase.","excerpt":"Web3 wants to address the challenges of Web2. For instance, a platform named DIMO claims to offer an alternative to Uber, and Hivemapper wants to make Google Maps more accessible","categories":["AI Trends"],"tags":["Blockchain","Top Trend"],"author_name":"Lokesh Choudhary","publish_date":"2023-02-09T13:00:00","publication_year":"2023","word_count":830,"keywords":["Top Trend","Go","Blockchain","programming_languages:R","AI","programming_languages:Go","Git","RAG","Aim","Rust","R","programming_languages:Rust"],"extracted_tech_keywords":["AI","Aim","RAG","R","Go","Rust","Git","programming_languages:R","programming_languages:Go","programming_languages:Rust"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-6-web3-alternatives-to-your-favourite-web-2-platforms\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":61667,"title":"Data Warehouses: How does AWS Redshift Compare With Snowflake?","content":"Data warehouses are now properly leveraged to use data to derive deep analytics. Out of all the data warehouses, some of the leading platforms include – Amazon Redshift, Google BigQuery, and Snowflake. In this article, we compare Redshift with Snowflake. AWS Redshift is a data warehouse product which constitutes a portion of the Amazon Web Services cloud platform. With Redshift, businesses can query petabytes of structured and semi-structured data across their data warehouse and data lake utilizing standard SQL. Redshift allows users to save the results of their queries back to the S3 data lake adopting open formats, like Apache Parquet, to additionally analyze from other analytics services like Amazon EMR, Amazon Athena, and Amazon SageMaker. Snowflake gives a cloud-based data storage and analytics setting as ‘data warehouse-as-a-service,’ which enables enterprise users to store and analyze data utilizing cloud-based platforms. Snowflake has worked on Amazon S3 since 2014, and on Microsoft Azure since 2018, and also on Google Cloud Platform in 2019. While it stores data on public cloud platforms, the query engine is made in-house. With a standard and interchangeable code base, Snowflake gives benefits such as global data replication, which implies that users can move data to any cloud in any geography. Platform Use Cases Redshift can be defined as a wholly-managed, cloud-ready petabyte-scale data warehouse platform which may be smoothly blended with enterprise intelligence tools on AWS. Redshift allows multiple integrations with different technologies, especially with tools on the AWS platform. Unlike Snowflake, Redshift considers that user data is in AWS S3 already for performing tasks. AQUA is a new distributed and hardware-accelerated cache that supports Redshift to go up to 10x faster than any other cloud data warehouse. Operating as a virtual data lake, Snowflake provides analytical capability across various cloud platforms, which entails that companies can securely have data and applications irrespective of the platform. Its cloud-neutral and virtual nature makes it very useful and functional for big business users. Virtual Warehouses can be applied to store data or run queries and can perform both these jobs concurrently. Snowflake Virtual Warehouses can be scaled up or down on command and can be suspended when not in use to decrease the expenses on computing. So if a company is looking to cut down waiting time through Query, or uploading the data faster to provide a hassle-free end-user result, then this is the best solution for the company. Integration Redshift integrates with a multiple of AWS services like Athena, Glue, SageMaker, DynamoDB, Athena, CloudWatch, etc. So if you are looking to use a data warehouse with AWS, then Redshift is probably your best choice. All you have to do is Extract, Transform, Load (ETL) into the warehouse and start performing analytics. Snowflake does not have similar integrations, which makes it more challenging for clients to use tools like Kinesis, Glue, Athena, etc when attempting to integrate their data warehouse with their data lake architecture. It, on the other hand, integrates with tools like IBM Cognos, Informatica, Power BI, Qlik, Apache Spark, Tableau and a few others, which can be helpful for analytics processes. Snowflake gives native support for JSON documents, providing built-in functions and querying for JSON data. In contrast, there is limited support for JSON at AWS Redshift, as reported by users. Pricing Redshift has a higher compute per dollar, saving you more money for the same amount of total compute time. On a general level, if we look at the pricing models, we see that Redshift is cheaper for on-demand pricing. Also, with Reserved Instances, costs can be further reduced for using AWS Redshift. Snowflake, on the other hand, has a dynamic cost model and depends on the workload and pricing can be billed basis each separate use patterns of the virtual warehouses about compute and storage. What we see is that smaller companies lean towards Amazon Redshift due to its simple usability and affordable pricing. But large enterprises can find value in Snowflake as computing, and storage can be used separately, which can bring overall prices down.","excerpt":"Data warehouses are now properly leveraged to use data to derive deep analytics. Out of all the data warehouses, some of the leading platforms include – Amazon Redshift, Google BigQuery, and Snowflake. In this article, we compare Redshift with Snowflake. AWS Redshift is a data warehouse product which constitutes a portion of the Amazon Web […]","categories":["AI Features"],"tags":["Amazon AWS","AWS","AWS services","informatica","no etl","Snowflake"],"author_name":"Vishal Chawla","publish_date":"2020-04-14T17:00:00","publication_year":"2020","word_count":673,"keywords":["Amazon SageMaker","informatica","AWS services","AWS","AI","R","Apache Spark","RAG","Amazon AWS","no etl","analytics","SQL","Azure","Snowflake"],"extracted_tech_keywords":["AI","analytics","Amazon SageMaker","RAG","AWS","Azure","Apache Spark","Snowflake","R","SQL"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/data-warehouses-how-does-aws-redshift-compare-with-snowflake\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10160636,"title":"Why This AI VC Firm Loves the Indian Middle Class","content":"Whether they pick AI for India or AI for Bharat, one thing is certain, the VCs betting on India’s AI ecosystem are focusing on the largest segment that will benefit the most from such innovations: The great Indian middle class. Seattle-based Capria Ventures is one such capital firm that identified this opportunity over a decade ago and has been capitalising on this aspirational segment ever since. The 12-year-old firm invests in early-stage and early-growth tech startups in the global south. Why Middle India Matters “We look at the rising middle class of India, often referred to as ‘Middle India’, which aspires to have and do more,” said Will Poole, co-founder and managing partner of Capria Ventures, in an exclusive interaction with AIM. Poole explains how the aspirational middle class, comprising roughly 600-700 million people, forms a cornerstone of their investment strategy. “This market is largely underserved or less served but represents a very large opportunity,” he said. With increasing incomes and growing aspirations, Poole believes that this demographic drives demand for innovative solutions across sectors such as healthcare, education, and technology. The firm recognised this trend early. “We started doing this before anybody else was doing it,” he remarked. With decades of experience as an investor and entrepreneur who founded eShop, which was acquired by Microsoft in 1996 and was part of the ‘Windows’ growth, Poole has witnessed the advent of mobile technology firsthand during his tenure at Microsoft in 2003. “The power of the mobile revolution happened fast in India,” he noted, drawing parallels to the current wave of digital and AI-driven innovations. Poole also highlighted the role of generative AI in addressing challenges specific to Middle India, such as language diversity. By leveraging AI for multilingual solutions, businesses can expand their reach and improve operational efficiency. “Generative AI is an equaliser in the world of language,” he said, adding that it reduces the burden of linguistic diversity for businesses aiming to scale. With India set to become the world’s third-largest economy within the next five to seven years, the opportunities for VC-backed startups are immense. “We’re very bullish on India,” Poole stated. Capria began operating in India more than a decade ago, followed by Southeast Asia, Africa, and Latin America. Since then, the company has invested in several Indian AI startups. Its core team is located in Bengaluru. Healthcare, finance, agritech, are some of the verticals Capria has invested in, with prominent customers leveraging their solutions. For instance, 5C Network, a telediagnostics company, leverages AI to analyse radiological images, improving disease diagnosis and radiology efficiency. With over 11 million images analysed and 250 regulator-approved disease detectors, 5C serves clients such as Manipal Hospital. BetterPlace is an HR management platform that supports frontline workers across India and Southeast Asia. It integrates AI for multilingual training, enhancing employee productivity and reducing training costs. Reliance is among its key customers, managing over a million employees through the platform. Challenges in the Indian Ecosystem Despite its promise, investing in India’s startups comes with unique challenges, with cost management being a significant hurdle. Poole shared an anecdote about a founder who implemented GPT-4 for customer service but had to downgrade to GPT-3.5 to manage costs. “You’d never find a US customer service startup worrying about API costs like this,” he said. Poole noticed that access to cutting-edge technology is another challenge in our ecosystem. Indian startups often lag behind their counterparts in the US in accessing state-of-the-art tools and developments. “Our role as VCs is to help them sort through the noise and identify the gems they can use to scale their businesses quickly,” Poole explained. Capria is also clear in its investment philosophy of investing in startups that leverage existing technologies to build scalable businesses rather than those building their own AI models or competing with heavily funded incumbents. This pragmatic approach extends to identifying startups with proprietary access to data, which Poole sees as a critical competitive advantage. “Data helps them build rapidly and maintain an edge over competitors,” he said. What About Others? It’s not just Capria Ventures. According to the recent trend, many VC firms are investing in AI startups that aim to address problems at various levels in the country, especially through language applications. Sarvam AI, backed by Peak XV Partners and Khosla Ventures, for instance, has built Indic language models. Similarly, Accel is supporting AI startups through its Accel Atoms 4.0 program. This program will offer up to $1 million in funding for startups focusing on AI and Bharat, specifically catering to middle-income households in Tier 2, 3, and rural India. Capria Ventures is looking to launch a new fund focused on applied generative AI that addresses the unique needs of middle India. “We’ve been doing this for a dozen years now, and we’re excited about the next dozen,” concluded Poole.","excerpt":"“This market is largely underserved but represents a very large opportunity,” said Will Poole, co-founder and managing partner of Capria Ventures.","categories":["AI Features"],"tags":["5C networks","Capria Ventures","GenAI","Healthcare","Peak XV","VC","Will Poole"],"author_name":"Vandana Nair","publish_date":"2025-01-01T16:00:00","publication_year":"2025","word_count":804,"keywords":["VC","Go","5C networks","GenAI","API","AI","R","Scala","Git","RAG","GPT","Capria Ventures","Aim","generative AI","Peak XV","Healthcare","Will Poole"],"extracted_tech_keywords":["AI","generative AI","Aim","RAG","R","Go","Scala","Git","API","GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/why-this-ai-vc-firm-loves-the-indian-middle-class\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10165423,"title":"Google Co-Founder Larry Page&#8217;s Startup Dynatomics to Bring AI to Manufacturing","content":"Google co-founder and former CEO Larry Page is reportedly building a new startup called ‘Dynatomics’, which is focused on integrating artificial intelligence in product manufacturing, as per The Information. The startup is led by Chris Anderson, the former CTO of electric airplane company Kittyhawk, which had Page as a lead investor. Dynatomics, which consists of a small group of engineers, aim to create ‘highly optimised’ designs for products and then get them manufactured in factories. It will be interesting to see how Page’s approach towards building a startup at the intersection of manufacturing, design, and AI pays off. Page was Google’s CEO from April 2011, to October 2015, and currently serves as a member of Alphabet’s board of directors. Currently, Page’s net worth exceeds $150 billion, and is the eighth richest individual worldwide, as per Bloomberg. The use of AI in manufacturing industries offers a vast potential for optimisations. A report from IBM last year states that AI, and machine learning, along with computer vision and natural language processing, do improve several aspects of the manufacturing process. “AI can analyse large volumes of data from sensors, equipment and production lines to optimise efficiency, improve quality and reduce downtime. By using algorithms to identify patterns in data, AI can anticipate potential issues, suggest improvements and even autonomously adapt processes in real-time,” read the report. A few weeks ago, Swedish multinational engineering company Sandvik, revealed that a tool called ‘Manufacturing Copilot’ was made available to customers. “The use of this generative AI tool in manufacturing simplifies the user experience on Computer Aided Manufacturing software,” said the company. Recently, Tata Consultancy Services (TCS) announced a collaboration with Salesforce to enhance the use of artificial intelligence in the manufacturing and semiconductor sectors. Similarly, several startups from Y-Combinator have emerged that aim to help engineers manage their manufacturing feasibility, and use AI-powered CAD design. One startup also claims to be building the ‘world’s first AI mechanical engineer.’","excerpt":"Dynatomics, consisting of a small group of engineers, aim to create ‘highly optimised’ product designs and then get them manufactured in factories.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","larry page"],"author_name":"Supreeth Koundinya","publish_date":"2025-03-07T12:00:16","publication_year":"2025","word_count":322,"keywords":["Go","machine learning","artificial intelligence","programming_languages:R","AI","larry page","computer vision","Aim","generative AI","R","AI (Artificial Intelligence)","startup"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","computer vision","generative AI","Aim","R","Go","startup","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/google-co-founder-larry-pages-startup-dynatomics-to-bring-ai-to-manufacturing\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10053236,"title":"Council Post: AI Is Not Just Data Sciences","content":"AI is no stranger to workplaces today. Fifty-three per cent of global leaders have integrated or are integrating AI into their workforce to enhance their business insights. While IDC has predicted that by 2022, 75% of enterprises will have embedded intelligent automation into technology and process development, the key aspect to consider is how and through which job roles organisations are going about incorporating AI in their workforce. Recent trends point towards broadening the scope of AI-driven roles in various parts of the workforce. This entails data science job roles spanning across the horizontal along with the vertical pillars of an organisation. With the increased integration of AI and Data Science teams in organisations, it is important for organisations and aspiring data science employees to understand the breadth of job roles. Many leaders are under the fallacy that AI and analytics can do with just data scientists, but the field is not limited to them. In fact, it is not even limited to engineers or individuals with a data science background. The broad scope of the field allows individuals from any educational background to leverage their skills and fit into an analytics-based job. The analytics industry has job roles in various subjects for both technical and non-technical backgrounds. Let’s discuss them. Roles in Data Engineering Chief Data Officer A member of the C-suite, a CDO, leads the utilisation, governance and implementation of data across the organisation. They are responsible for creating a data management system and instilling a culture of data sharing and data analytics based problem-solving. Data Engineer Data engineers are responsible for pulling in the required data for the work. This includes making data accessible to the organisation, collecting data and making it available for analysis. Data engineers ensure this by standardising and automating the data collection process. They are also responsible for designing, building, testing and maintaining data management systems to store and manage the data. Data Collection Specialist Working closely with data analytics, data collection specialists collect data by creating and administering surveys, research and interviews. They review and clean the data received and present the data using tools such as Microsoft Excel and PowerPoint. Database Administrator As the title suggests, database administrators are responsible for managing the systems that store the data for organisations. They organise the data systems, secure data and find better infrastructural ways to store the data and provide access to relevant stakeholders. Roles in Data Analytics Chief Analytics Officer A CAO is a senior manager responsible for analytics operations in the organisation. They lead the process of digital transformation by making data-driven decisions backed by data science techniques. They also develop the warehouse as a central repository for all information within the organisation. While CDOs are responsible for managing data and driving maximum value from data, CAOs lead data analytics and derive data insights. Data Analyst A data analyst is in charge of designing and maintaining the data systems and database. They need to be fluent in codes to fix coding errors in the system. Data analysis is a statistics-based role, and analysts have to be proficient in using statistical tools to interpret datasets and predict trends. The job role also involves mining data and reorganising it so that both humans and machines can read it. Data analysts typically move on to senior data analysts, data scientists, analytics managers, and business analysts. They go on to become senior data analysts to move away from the detailed working of data analytics to take a broader view. This entails directing, organising and leading the data analytics projects in the organisation. Analytics Translator Described as the new ‘must-have role’, an analytics translator bridges the gap between technical data engineering and operations-based departments. They present deep technical based insights in an easily understandable language to the other teams to ensure the data insights are translated into real impact at scale. Roles in Business Intelligence Business Intelligence Developer Business intelligence developers help organisations get a sense of business acumen along with AI to optimise different processes involved in using AI for business enhancement. BI interfaces are great tools for leaders to study the past, present, and predictive views of the organisation. The BI developer is essentially an engineer in charge of developing, deploying and maintaining these BI interfaces. They need to be proficient in building query tools, data visualisation, ad hoc reporting and interactive dashboards that translate business requirements into technical instructions. Data Visualisation Engineer Data visualisation engineer develops and deploys user-friendly dashboard and data visualisation tools using datasets from different sources. They work with complex data and use visualisation tools to present it in a simple manner for everyone to understand. Roles in Data Science Data Scientist Termed “the sexiest job of the 21st century” by the Harvard Business Review, data scientists are responsible for gathering relevant data from primary and secondary sources. They use machine learning and related tools to derive useful insights from the collected data. Data scientists develop, maintain, and evaluate AI solutions, clean & organise big data, and use analytical and statistical tools to identify patterns. Data scientists need to be proficient in programming languages, such as Python and Java. While data analysts work with structured data to solve problems, data scientists are involved in prescriptive and predictive analysis. They deal with raw data to make predictions about the future. Machine Learning Engineer ML Engineers are essential to implementing, building, and running algorithms and softwares in an organisation. They are experts at coding, problem-solving and using data inputs provided by analysts to create a successful algorithm. In addition, they are responsible for building AI systems that can develop algorithms capable of making predictions. Deep Learning Engineer Deep learning engineers are specialists using deep learning for programming tasks related to AI. They are responsible for developing systems that can transfer data efficiently. They do this by writing complex codes to direct parts of the neural network. Computer Vision Engineer Computer vision engineers leverage software to process and analyse tons of data. They are responsible for designing and creating image processing and visualisation platforms. Additionally, they develop deep learning frameworks to solve problems and use data to support the automation of predictive decision making through visuals. Machine Learning Architect Machine learning architects lead the end to end architecture and development of machine learning solutions for the organisation’s programs. They implement algorithms into pipelines that can be consumed at scale. They are also responsible for creating scalable microservices for backend solutions. Roles in Application Development Technical Product Manager A technical product manager identifies and resolves business challenges in the organisation. They are responsible for managing the entire lifecycle of an AI-based product and transforming business strategies into well-defined products. This entails working with different teams to ensure a smooth product development process. UI\/UX Designer A user interface or user experience designer is capable of understanding the user experience and designing interfaces accordingly. They collaborate with engineers and product managers to understand user requirements and work on design approaches to meet those needs. UX\/UI designers have experience in working with data and a background in coding tools like Javascript that they can use to generate personalised data-based visualisations. Roles in MLOps AI Engineer An AI Engineer is an IT expert responsible for testing and deploying AI models. They understand and develop intelligent algorithms for the company to learn, analyse and predict future events, and are also a researcher expert in analysing brain functions and building similar computational programs. Specialisation in areas such as ML or DL is needed to be an AI engineer. MLOps Engineer An MLOps engineer is a software engineer with an added specialisation in the deployment and production parts of the data science process. They work with data scientists to bridge the gap between testing and production by using both data engineering and DevOps. Roles in Research and Development Research Scientist Research scientists are responsible for conducting market research for the AI teams. They are involved in designing and analysing information. The job role entails extensive research in their domain wise fields such as NLP, computer vision, applied mathematics, computational statistics, and more. This makes it integral for research scientists to hold a post-graduate or a PhD degree. Moreover, they explore new ways and systems to solve problems and apply for patents in AI-related domains. They go on to become senior research scientists and principal research scientists. Robotics Scientist Robotics scientists are responsible for building and designing mechanical devices such as UAVs and UGV. In addition, they are responsible for building robots to automate small manual tasks, design operating systems and test operating functions. Robotic scientists need to be experienced in robotic or mechanical engineering and have a strong base in advanced mathematics, physical science, and computer-aided design. Roles in Ethical AI AI Ethicist With companies dealing with larger data volumes and complex AI-based solutions, it is essential to have an AI ethicist to ensure adequate governance mechanisms are in place to address the ethical concerns of AI. They educate and guide the employees on adequate ethical practices and guard the organisation against biased AI. AI Policy Researcher AI policy and ethics researchers work on anticipating long and short term risks for technical ethics. Ethics teamwork with NGOs, academics and other forums to explore AI-related issues in society. The Burning Glass survey, based on the analysis of hundreds of job postings, revealed significant growth in job skills in 2021. A similar conclusion was derived by The World Economic Forum’s Future of Jobs report that found that new job positions are set to grow in the data science industry. AI and data science have been considered to be technically complex fields that tend to scare away individuals from a non-technical background or enthusiasts wanting to work in the field. The new job roles in the market have opened up these sectors to people from different educational backgrounds to be a part of this dynamically growing industry. It is integral for data science enthusiasts and potential data scientists to understand the breadth and scope of this industry. Unlike the usual set jobs, data science is dynamic and constantly growing into a layered sector that welcomes people in various fields. This is a welcoming area for individuals with varied educational backgrounds and degrees to try out new job roles and work on different projects. This article is written by a member of the AIM Leaders Council. AIM Leaders Council is an invitation-only forum of senior executives in the Data Science and Analytics industry. To check if you are eligible for a membership, please fill the form here.","excerpt":"AI is no stranger to workplaces today. Fifty-three per cent of global leaders have integrated or are integrating AI into their workforce to enhance their business insights. While IDC has predicted that by 2022, 75% of enterprises will have embedded intelligent automation into technology and process development, the key aspect to consider is how and […]","categories":["AI Features"],"tags":["data science programs","data science roles"],"author_name":"Anirban Nandi","publish_date":"2021-11-11T18:00:00","publication_year":"2021","word_count":1762,"keywords":["data science","machine learning","AI","neural network","ML","MLOps","data science roles","computer vision","NLP","deep learning","analytics","data science programs"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","neural network","NLP","computer vision","data science","analytics","MLOps"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/council-post-ai-is-not-just-data-sciences\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10139413,"title":"Why India Needs People+ai","content":"When we visited People+ai a couple of weeks ago, we had no idea we would walk into a space filled with AI engineers, volunteers and researchers, and of course, walls adorned with AI-generated art, all in line with the philosophy of ‘What can (AI) do for you?’ — straight from Tanuj Bhojwani’s LinkedIn tagline. Under his leadership, People+ai is championing AI for India and addressing real challenges at scale with the aim to accelerate through various initiatives and government partnerships. As he put it, “We have a seat at the table… not just in building models but in creating the world’s largest use cases for AI where the difference in people’s lives is tangible”. People+ai was born out of Nandan Nilekani, Rohini Nilekani, and Shankar Maruwada’s EkStep Foundation in June last year. This community-driven effort aims to harness AI capabilities to help a billion people. And it is being led by Bhojwani, who has been an investor, a consultant, and co-founder earlier as well. Bhojwani explained the importance of building Indic language models and their use cases within the country. “If I put a gun to your head and tell you that you cannot use Google from next month, you will maybe push back for one or two months but then start coughing up.” Bhojwani said that everyone is accustomed to the idea of going to this magic box, typing what they think, and getting the results. But according to him, most of the Indian population is yet to experience that type of internet because of the country’s low literacy rate. The same people are going to access the new internet multimodally, through voice and pointing the camera at things. This is why it is important to take AI to the grassroot-level of the country, and do it in the way that India does it – frugally. People+ai is currently working on projects like Jan Ki Baat, Sthaan, and Open Cloud Compute (OCC) to make this a reality. “I think there are problems that are going to be unique for India. Who is going to solve that?” Bhojwani asked rhetorically while explaining that copying the West for ideas is not the right way forward. He also added that the behemoth, Amazon, is shaken because of Zepto, Instamart, and Blinkit – the same should be applicable to AI, and not playing catchup with the West. Nilekani, the co-founder and non-executive chairman of Infosys, previously noted that just as India benefited hugely from Digital Public Infrastructure (DPI), Aadhaar, and UPI, AI holds similar transformative potential. “AI is a very powerful technology, but it’s a technology like anything other,” said Nilekani, adding that AI needs to be used with appropriate safety and guardrails. This is the exact principle that People+ai works on. “When it comes to the infrastructural pieces, which involves DPI and policy, we step in,” said Bhojwani. India as the AI Use Case Capital of the World Building AI in India is a game that’s very different from that in the West. “If you look at how much an AI solution could mean to a user, it’s much higher in India with a much larger volume,” he added. In the West, it’s about acquiring enterprise customers willing to spend millions of dollars. But in India, it’s a high-volume, low-value game, where the AI users would not be paying so much. These people would be more comfortable using AI in their own native languages. This solution is at the population scale. For India to flourish in AI, models must be created that understand India’s linguistic nuances and cultural complexities. Bhojwani had earlier told AIM that the foundation had been exploring the benefits of AI for quite some time. However, it was only in the past year that they decided to establish a specialised unit focused on extensively exploring AI use cases. He believes that in a decade, India is going to be the AI use case capital of the world because of its huge population, diversity, languages and the need for specialised AI tools, be it in healthcare, education or any other field. This is exactly the idea that Nilekani puts in with ‘Adbhut India’. While enterprises and other companies are pushing the idea of building models and acquiring big customers, it is also essential to make AI reach the roots of the country, which is what AI for Bharat stands for. People+ai is involved with stakeholders to determine if better tools and resources can be developed. “For this, we are collaborating and talking to different organisations, some in Africa, because they are also facing similar problems,” Bhojwani said. AI for Bharat If India is going to be the AI use case capital of the world, a substantial compute infrastructure is imperative. People+ai is actively working on developing an open compute infrastructure network to meet the rising demand for compute while promoting market competitiveness. The idea here is to help small businesses based in Tier 2 or Tier 3 cities access computing infrastructure for training or inferencing at a lower cost compared to leveraging services from the likes of AWS, Google or Azure, which could prove to be costly. People+ai is already working with Indian computing service providers E2E Networks, Jarvis AI Labs, NeevCloud, and Vigyan Labs and the idea is to create a network of micro data centres with interoperable standards, allowing small businesses and startups to easily plug and play, based on their specific requirements. “However, the biggest stumbling block right now is the lack of understanding regarding the potential use cases of AI. While building chatbots is undoubtedly crucial, there is a significant amount of work that still needs to be undertaken, especially in the context of India,” Bhojwani said. “If I had to choose one of them first, I would choose the use cases,” said Bhojwani, while explaining that every company in the world is increasingly getting interested in Indian AI models, be it OpenAI, Google, or Meta. The moat usually stands in building AI use cases as it is easier to go up that supply chain, rather than going down. OpenAI did not build GPT-4 on day one, it took them years and several iterations to reach that level. The same would go for Indic language models built by Indian AI companies. “Being one or two generations behind the SOTA models is still good enough.” The long-term vision of all AI companies is the same—to have indigenously developed AI models. Given the network and access to resources that the Indian AI companies have right now, the next best move would be to build a GPT-2 level model instead of competing with the West. Building models is getting cheaper and catching up with SOTA a few years later would be astronomically cheaper. “What is the hurry?” asked Bhojwani, explaining that it is better to solidify the market that could sustain the models. “For a constrained set of resources, where would you rather apply them,” he added and said that it is good to build models, but if you had to pick what to do first, defining the use cases is more important.","excerpt":"“If you look at how much an AI solution could mean to a user, it’s much higher in India with a much larger volume,” said Tanuj Bhojwani.","categories":["AI Features"],"tags":["Interviews and Discussions"],"author_name":"Mohit Pandey","publish_date":"2024-10-25T17:30:00","publication_year":"2024","word_count":1184,"keywords":["Go","OpenAI","AI","chatbots","AWS","R","Git","RAG","Aim","Azure","Interviews and Discussions"],"extracted_tech_keywords":["AI","OpenAI","Aim","RAG","chatbots","AWS","Azure","R","Go","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/why-india-needs-peopleai\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10011321,"title":"Citizen Data Scientists &#038; AI On Cloud Will Create New Opportunities For Financial Industry, Says Manisha Banthia, Fiserv","content":"Struggles in the banking and finance space aren’t unique to the COVID pandemic; however, the economic downturn due to the outbreak has indeed aggravated the situation. With the rapid change of customer behaviour, the financial companies are going under a significant transition of keeping up their relevance amid this crisis. Not only the industry is undergoing a shadow of looming job loss but also witnessing the rush of automation. Currently, the commercial banking segment is undergoing a massive revolution with artificial intelligence, machine learning, real-time analytics, as well as the personalisation to achieve success. Fiserv, the global provider of payments and financial technology services and solutions, is one such company that has been around this drastic transformation of the industry and has been enabling an innovative financial solution for its customers. To understand the landscape better, and how Fiserv is making a difference, we got in touch with Manisha Banthia, the Director of Analytics at Fiserv Global Services. With over 23 years of experience in financial analytics, Manisha Banthia is currently heading the Analytics Practice at Fiserv Global Services. Prior to this, she has headed the analytics team at Infosys Consulting and built analytics solutions for Citibank — Asia Pacific, American Stock Exchange and Country Financial, during her stint in Oracle. Manisha has also worked in the domains of marketing analytics, financial performance, as well as fraud and risk analytics. Here is the edited excerpt: How do you witness analytics evolving today in the financial industry as a whole? What are the most important trends that you see emerging in the analytics space across the globe? New financial products, digital and payments transformation and personalisation engines have made analytics in the financial industry an interesting yet challenging space. While the focus of analytics has been on risk management and customer marketing, we are now seeing the application of analytics for decision making in other nuanced areas such as creating optimisation models for better capacity planning and predictive maintenance for ATMs. Additionally, there is increasing usage of analytics to gather actionable sales and marketing intelligence, such as visitor segmentation on mobile and web-based banking applications. The use of natural language programming (NLP) for text analysis of customer services and surveys and intelligent automation like leveraging Intelligent and optical character recognition for cheque frauds and cards disputes are also the evolving areas. Artificial Intelligence and virtual assistants or conversational AI are other areas that are being rapidly adopted by financial institutions for enhancing the client experience. For example, the advisor experience solution from Fiserv provides immediate personalisation hyper-targeted to each advisor. It delivers precisely the appropriate educational content and products designed to help advisors make more informed decisions, grow their businesses and aid in their clients’ financial goals. The emergence of citizen data scientists, AI platforms, and AI on the cloud is going to create more opportunities and keep this space interesting. Mention some of the impediments in the area. What is Fiserv’s strategy towards tackling those challenges? Selecting the appropriate infrastructure, tools, and technology is essential if data is to provide useful insights and not just create more noise. We have a matrix of business requirements vs technologies and tools that we leverage to decide the best mix of these components across analytics programs. Since technology and toolsets are evolving, we are also keeping ourselves abreast and keeping the matrix evolving. What is your approach to face the challenge of meeting the needs of so many clients across vast geographies with limited resources? Resources translate into three categories — data infrastructure and secure access, continuously evolving toolsets and most importantly, people. We keep technology options open rather than sticking to a single stack as we serve both large and small\/medium businesses. This requires cross-skilling of people while maintaining experts for each stack. The toolsets, on the other hand, are rapidly evolving, not just in the AI and machine learning space but also in visualisation and narrative sciences. We have a strong learning and development program with a team culture of continuous learning that thrives on inclusive innovation of prototype building. We believe in practical applicability of the skill sets and not mere certifications. We will begin to follow a philosophy of ‘Democratisation of AI’ by gradually building a culture of citizen data scientists in which data models and analysis, once established, are managed by business teams, while the analytics team moves on to solve the next business problem. A good talent acquisition program with a mix of engineering, programming, mathematics, statistics, and financial services domain expertise is what has helped us succeed and cross-train our teams. Can you share a case study where you used your analytics-based solution to address the business challenge for one of your clients? How did they see value out of your solution? We had developed an ML-based analytical solution to help a business reduce false alarms and identify true ATM service failures, thus reducing service costs and improving customer experience. This was done by automating ATM service calls based on criteria such as a complete absence of withdrawal transactions in a specified time period on a given day of the week. This approach is now being scaled and deployed for multiple clients with minimal customisation. What are the key value propositions in your analytical solutions compared to the ones already in the market? We are building analytics products and solution frameworks in house rather than focusing only on services or an advisory play. We are combining digital analytics in all our analytical programs to understand customer behaviour as we see organisations focus on their digital transformation programs. What does Fiserv’s future look like in a data-driven industry? What are the next steps for analytics at your organisations? Data-based decision making is emerging as a norm across organisations. However, what is required is proactive data-based decision making, using timely actionable insights, which can now be achieved using AI and ML. The road ahead is to make use of techniques such as graph networks for fraud detection, intelligent automation for operations efficiency, digital analytics and creative AI. Additionally, we are making available a Customer Data Platform that spans digital products and offers more personalisation and effective digital marketing with known and unknown data. What are a few things that organisations should be doing with their analytics efforts that most do not do today? Most organisations are focused on analytics services and advisory services. It is important to create assets in the form of frameworks, codes, model validation programs, as well as for analytics products. Organisations are also restricting analytics to ‘traditional areas’ like risk and marketing and not expanding it to all business areas. There are opportunities for applying analytics not just for revenue generation but also to drive cost efficiency and customer satisfaction. Analytics teams usually consist of statisticians or programmers but giving importance to domain expertise is critical. Organisations need to hold update sessions with all members of the team on the scope of the project and how will the business benefit. Further, it needs to encourage continuous learning and domain expertise development programs, as well as share learnings across teams and projects and do not work in business \/ or project silos. What is your suggestion to new graduates who are aspiring to get into financial services analytics space? What are the typical skill sets you look at while recruiting in analytics? Having an attitude of learning continuously is the key to survive in the financial analytics space. Financial services analytics is a fast-changing space, and therefore, agility in moving forward with knowledge and hard work is required. Be equipped with the basics of newer technologies, understand the importance of learning and following practices such as unbundling of neural network codes, champion the challenger models and interaction with data visualisation. It is also important to understand how the model was applied in the business and its contribution to business strategy building. While recruiting, we look for problem-solving skills and the ability to acquire knowledge of new techniques. Our hiring is business aptitude and case-study based apart from programming and technical aspects.","excerpt":"Struggles in the banking and finance space aren’t unique to the COVID pandemic; however, the economic downturn due to the outbreak has indeed aggravated the situation. With the rapid change of customer behaviour, the financial companies are going under a significant transition of keeping up their relevance amid this crisis. Not only the industry is […]","categories":["AI Features"],"tags":["ai big data analytics digital transformation","AI Certifications","AI in finance","big data scale","big data trends and challenges","Data Analytics Certification","Data Scientist","finance","finance India","Interviews and Discussions","Oracle Financial Service","Scale Big Data","the need to scale with big data"],"author_name":"Sejuti Das","publish_date":"2020-11-06T18:00:32","publication_year":"2020","word_count":1346,"keywords":["AI Certifications","the need to scale with big data","AI in finance","Data Analytics Certification","fraud detection","artificial intelligence","RAG","NLP","analytics","machine learning","AI","ai big data analytics digital transformation","neural network","ML","Oracle Financial Service","Scale Big Data","big data trends and challenges","finance","big data scale","finance India","virtual assistants","Data Scientist","Interviews and Discussions"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","neural network","NLP","analytics","RAG","virtual assistants","fraud detection"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/citizen-data-scientists-ai-on-cloud-will-create-new-opportunities-for-financial-industry-says-manisha-banthia-fiserv\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10172756,"title":"Madhya Pradesh Joins the GCC Race, But Isn’t Playing the Same Game","content":"While cities like Bengaluru, Hyderabad, and Pune have long been GCC magnets, Madhya Pradesh has now jumped into the fray with its ambitious GCC policy. The newly launched framework aims to position the state as a serious contender among tier-2 cities, offering talent, infrastructure, and generous incentives to attract companies looking to set up tech and operations hubs. Madhya Pradesh has a strong presence in IT\/ITeS, automobile, geographic information system (GIS), pharmaceutical, and textile industries. For these industries to grow faster, the state intends to attract sector-specific GCCs to support specific industries. With this policy, the state aims to attract more than 50 GCCs, targeting the employment of over 37,000 direct employees. In an exclusive conversation with AIM, Sanjay Dubey, additional chief secretary ,the government of Madhya Pradesh, said, “We are not competing with Bengaluru, Pune or Noida. We know they have a first-mover advantage and established ecosystems. But among tier-2 cities, we want to be the best in terms of ease of living, ease of doing business, infrastructure, and most importantly, talent.” The state is positioning Indore and Bhopal as core hubs, with ready spaces and policies designed to fast-track the setting up of GCCs. Dubey clarified the government’s strategic approach to attracting GCCs, emphasising a multi-sectoral focus and a realistic assessment of its positioning. “We will definitely be focusing on finances and on multi-sectoral GCCs, which are looking not only for a particular kind of operation, but are also into multiple domains,” Dubey said. The state is sharply focused on leading the pack among tier-2 cities, which includes being best-in-class across several parameters. In fact, with IT\/ITeS exports tripling in the past three years, the state has witnessed an impressive average growth rate of 43%. The presence of over 5 SEZs, 15+ IT parks, and 150 ESDM units, coupled with more than 2 lakh skilled IT\/ITeS professionals, creates a conducive environment for IT businesses. The growth is not limited to specific regions. However, cities like Indore, Bhopal, and Jabalpur are emerging as major IT hubs, housing numerous IT\/ITeS units, including BPO\/BPMs. Talent Is the Backbone, and  MP Is Betting on It The state’s strong emphasis on education, with about 300 engineering colleges producing well over 50,000 tech graduates annually ensures a steady supply of skilled talent. Additionally, the booming automobile and pharmaceutical sectors necessitate specific GCCs to address their evolving technological needs. The real hook, as Dubey mentioned, is talent. With its strong base of institutes like IIT Indore and IIITs in Gwalior and Jabalpur, Madhya Pradesh is leaning on its youth. “There is a readiness among our talent to come back. Many students from here are already working in Bengaluru or Pune. Now we want them to return,” he stated. To ensure students are job-ready, the government is partnering with global tech companies like Meta, Amazon Web Services, Barclays, and L&T Mindtree to offer AI and deep-tech skills courses. “Barclays is helping with fintech, Meta with AI-enabled services, and CA Unity with XR and gaming. We’ve also tied up with institutions like IIT Delhi to cover the spectrum—from ITI to engineering colleges,” Dubey explained. He also revealed that instead of waiting for companies to scout and set up their own spaces, the state is working with developers to offer plug-and-play infrastructure. Several developers are already in advanced talks. “GCCs don’t come on their own. They want developers to offer ready, standardised spaces. We’ve facilitated that, and we’ll soon see announcements of new centres in Indore and Bhopal.” Dubey’s ultimate goal is: “Out of the 2400 GCCs expected in India over the next few years, around 1800 are already in India. My vision is for Madhya Pradesh to get the maximum out of the remaining 600.” Building Centres of Excellence Recognising the growing demand for deep-tech capabilities, the state is setting up centres of excellence (CoEs) across sectors. MP is actively building a robust ecosystem for deep-tech innovation through the establishment of multiple CoEs. These include a Deep AgriTech CoE at IIT Indore, developed in collaboration with CDAC, ICAR, and MIT, which aims at advancing agricultural technologies. A dedicated drone technology CoE is being set up in partnership with IISER and the Drone Federation of India, focusing on manufacturing, navigation, and image analysis capabilities. The state is also in the process of launching a semiconductor CoE to strengthen its presence in the electronics and chip design space. Additionally, an AVGC-XR (animation, visual effects, gaming, comics and extended reality) CoE is operational at the Global Skill Park in Bhopal, supporting next-generation media technologies and skill development. To ensure these initiatives move fast and remain nimble, the state is forming Section 8 companies to run the CoEs, keeping them free from bureaucratic delays. “These CoEs will be run by professionals from industry and academia, not government officials. We are borrowing from what worked in other states, and learning from what didn’t,” he added. Space Tech Policy on the Horizon Dubey revealed that MP is preparing to roll out a Space Tech Policy by August 2025. The policy will focus on the state’s strong areas, like academic infrastructure and ready talent. He further said that IIT Indore is offering BTech, MTech, and PhD in spacetech, and the government is actively engaging with ISRO-linked startups and industry stakeholders regarding the Space Tech Policy. The state aims to align this with the central government’s broader space policy and fill gaps that others may not have addressed.","excerpt":"The state is also preparing to roll out a Space Tech Policy by August 2025.","categories":["GCC"],"tags":["GCC india"],"author_name":"Shalini Mondal","publish_date":"2025-07-02T18:00:00","publication_year":"2025","word_count":903,"keywords":["Go","programming_languages:R","AI","innovation","RAG","Ray","Aim","GCC india","cloud_platforms:Amazon Web Services","R","startup"],"extracted_tech_keywords":["AI","Aim","Ray","RAG","R","Go","innovation","startup","cloud_platforms:Amazon Web Services","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/gcc\/madhya-pradesh-joins-the-gcc-race-but-isnt-playing-the-same-game\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10073745,"title":"Class 10 kids Create AI to Feed Stray Dogs","content":"Students of Shiv Nadar School, New Delhi, have come up with an AI-enabled dog feeder to prevent the starvation of stray dogs. The device is made up of open-source technology and coding languages like Python, with a built-in UV sanitization mechanism. All from Class 10, the team ‘Voice for the Voiceless’, comprises three young innovators, named Shantanu Mukherjee, Ekansh Agrawal, and Arijit Sinha, who invented the intelligent dog feeding device. ‘VOICE’ is a smart feeder with an in-built camera and a sensor to recognise a dog. The technology triggers within the machine dispense a certain amount of dog food. The algorithm used in the device captures and processes a live video feed from the device’s camera. It detects the presence of a dog using ‘computer vision’, while training to detect a dog in its proximity. When asked about their inspiration, the students said they were driven by the #JusticeforBruno movement that bought nationwide attention to animal cruelty and starvation. According to a State of Pet Homelessness Index Report-2021, 80 million cats and dogs in India are homeless on the streets with no care at all. The target customers for the device are animal shelters, organizations, and NGOs working for the welfare of stray dogs. Devised with unique artificial intelligent algorithms, the prototype will function without the user putting in much effort. The students also aim to give it to dog lovers, at a minimal cost of about  Rs 5,000. Considering aspects such as technology, size, and design, the final price of the device is expected to be in the range of Rs 9000-10,000.","excerpt":"Three young innovators of Shiv Nadar School have developed an AI-enabled solution to address the problem of stray dogs’ starvation","categories":["AI News"],"tags":["Computer Vision"],"author_name":"Bhuvana Kamath","publish_date":"2022-08-27T17:32:47","publication_year":"2022","word_count":263,"keywords":["Go","ELT","AI","computer vision","Python","Ray","Aim","programming_languages:Python","Computer Vision","GAN","R"],"extracted_tech_keywords":["AI","computer vision","Aim","Ray","Python","R","Go","ELT","GAN","programming_languages:Python"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/class-10-kids-create-ai-to-feed-stray-dogs\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":18335,"title":"Using Social Media For Predictive Data Analytics Is In Its Infancy","content":"Social media is increasingly successful in putting across social phenomena which previously was studied only through traditional surveying techniques such as telephone or face to face interviews. Customers scrolling through Facebook, when coming across a post from one of their favorite apparel store showcasing a fantastic pair of jeans and some shirts, “like” it and at times leave a comment that they are eager to buy some or all of these; and eventually scroll down. Now what needs to be paid attention to is how that apparels store puts the likes and comments at work? At the most what they will do is to use it and shape their social media marketing strategy. Chances of them using that data to make informed decisions for improving operational efficiency such as how many pairs of those jeans to manufacture or whether to hike or reduce the prices is negative. Yes, this is true. Social media data can be effectively used to improve sales forecasts. In the aforementioned example, information from apparel company’s Facebook interactions can be incorporated into predictive data analytics models, and ultimately it will help in estimating the purchases their customers will more likely make. The world of data and analytics has changed and has changed big time. Today simply collecting social media data is not enough. Instead, companies are required use these collected data to upgrade their forecasting techniques. Though a subject of further research, it is speculated that data from social media most probably reflects how much attention customers are paying towards a brand, which equates to the good or bad word of mouth. However; the sad part is that companies still follow the conventional way of doing a thing where they care more about the technique’s effectiveness instead of the mechanism that makes it effective. This is an effort we have made to enlighten one and all about how to use social media for predictive data analytics. Social Media Data Types Twitter and Google trends have proved themselves when it comes to modelling stock prices. Their higher data volumes and immediacy help them in leaving behind Facebook. But that does not bring an end to the proves of Facebook, as it is extensively used for modeling sales, human emotions, personalities and human relations to a brand. At the same time picture and video-based social media platforms such as Instagram, YouTube and Netflix are also gaining grounds, and are expected to become more relevant for predictive analytics in near future. Enlisted are some of the types of data, predictive data models are relevant to: Time series – sales per month or sales per day Cross-sectional – individuals such as customers, for a given period of time Longitudinal\/panel – a combination of the former two such as a set of customers observed through several months Social media data processing Upon concluding to use social media data for the predictive purpose, the data collected at the level of the individual action will require being pre-processed to enter it into a predictive model. At times the data is temporarily aggregated to match the chronological aggregation level of the outcome, i.e. monthly data. Also not to forget that social media data is also in form of text variables, filtering, interpretation, and classification is also required. Organizations willing to use social media data for preparing predictive models to forecast sales find preprocessing of the social media data challenging from the computational aspects of data analytics. Once individual actions such as posts, likes are classified and sequentially aggregated, what remains is a set of potential explanatory factors which are very limited. Due to outcome variables with fairly low frequencies, modelling process deviates less from conventional approaches within predictive modelling. Model equation – theory-based Vs data-driven Ideally, linear, non-linear, parametric, non-parametric and semi-parametric models are taken into consideration. The thought process behind it also is that non-linear models will require more data points\/observations as compared to previous ones. There is a range of possible starting points for the search process. Using predictive models for forecasting purposes Upon finalizing the model, there comes a stage where considerations are required to be made regarding the implementation. One of the considerations is “how often the model needs re-estimation or specification updating”. Looking at the robust general data pattern, the specifications updating is expected to take place only if new variables come into the picture. The second scenario could be where an adequate amount of data points become available, allowing more complex structures to be formed. A combination of forecasts from different basic predictive models may take the center stage, as a practical perspective. Fine-Grained Forecasts The only catch here could be that social media data may not be applicable across industries. Yes, you read it right. Social media data tends to be more relevant to products with uncertain sales and are heavily influenced by trends, fashion, and entertainment to name a few. It is still to be witnessed if adding Facebook data would improve forecasts for consumer goods like breakfast cereals –where sales are already fairly predictable. There is a long but definite way of exploring why and how can social media improve forecasts. The next step here could be to implement best practices to predict sales for individual products and not only overall sales. A step further, with help of data based on geographic areas, the information can and would be used to decide what would be the penetration of a particular product or service in coming days. Companies by strategizing their social media posts can learn and make better decisions. This will help them to specifically draw information to run their operations, effectively. As in more and more companies can opt to display potential products to decide what to manufacture, based on customer’s reactions. Conclusion Predictive models have succeeded big time by offering numerical forecasts and assessments, along with quantitative statements to improve decisions in companies and by public authorities. Going ahead with parsimonious, simple models that capture the most important features of the data is advisable. They fulfill model assumptions and provide a good fit both in a sample and out of sample. Furthermore, it is important that even during the phase where the model is applied for its purpose, it performances is still monitored.","excerpt":"Social media is increasingly successful in putting across social phenomena which previously was studied only through traditional surveying techniques such as telephone or face to face interviews. Customers scrolling through Facebook, when coming across a post from one of their favorite apparel store showcasing a fantastic pair of jeans and some shirts, “like” it and […]","categories":["IT Services"],"tags":["social media analytics"],"author_name":"Chirag Shivalker","publish_date":"2017-10-14T05:00:11","publication_year":"2017","word_count":1040,"keywords":["Go","programming_languages:R","AI","social media analytics","data-driven","predictive analytics","Git","programming_languages:Go","analytics","GAN","R"],"extracted_tech_keywords":["AI","analytics","predictive analytics","R","Go","Git","GAN","data-driven","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/using-social-media-predictive-data-analytics-infancy\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":36790,"title":"Why You Should Use Weak Supervision For Your Data Labelling Chores","content":"If a model is tasked with identifying dogs from a random collection, then one can use named-entity recognition to label any content that does not contain the names of dog breeds. Existing knowledge resources can be combined with such simplistic logic to label training data for using it in a new model. This labelling function returns ‘None’ more often, which leads to only a few parts of data getting labelled. Any organisation that looks towards machine learning will have to deal with the challenges that come with data labelling. The need for hand-labelled training datasets becomes obvious right at the beginning of the pipeline. Data labelling requires the following: Collecting labels Developing label instructions Training subject matter experts to carry those instructions Deal with failing training datasets with evolving applications In order to generate high quality-correlated labels, the researchers at Google introduce Snorkel Drybell, a framework which uses generative modeling technique. In industry and other domains there has been an increased affinity towards programmatic or otherwise more efficient but noisier ways of generating training labels, often referred to as weak supervision. Snorkel Drybell, adapts the open-source Snorkel framework to use diverse organizational knowledge resources like internal models, ontologies, knowledge graphs to generate training data for machine learning models at web scale. Snorkel DryBell, integrates with Google ’s distributed production develops weak supervision strategies over millions of examples in less than thirty minutes. In this technique, unlike the previous weakly supervised models, an effort is made to build complete systems which can manage multiple sources of weak supervision that take in diverse accuracies and correlations. Snorkel DryBell enables writing labelling functions that label training data programmatically. This technique automatically estimates the accuracies and correlation consistently without any ground truth training labels. Source: Google AI This framework makes use of the following information: Heuristics and rules Taggers and classifiers Aggregate statistics Entity graphs This information is then used to write labelling functions in a MapReduce based pipeline. Each labelling function takes in a data point and either gives a label or just stays silent. To achieve accuracy, noisy labels need to be handled and to do this Snorkel DryBell combines the outputs from the labelling functions into a single, confidence-weighted training label for each data point. And, to evaluate the topic and product classification, the training labels estimated by this framework are used to train logistic regression classifiers with features similar to those in production. Training is done using the FTLR optimisation algorithm, which is a variant of the stochastic gradient descent. The initial step size here is 0.2 and trains over 10,000 iterations for topic classification task and more than 100K for the product classification task. Key Findings Users can write labelling functions over on unservable feature set and then use the output to train model over different servable feature set. This feature boosts performance by an average of 52% on the benchmark datasets. Efficient and inexpensive deployment of models. A new kind of transfer learning- transferring domain knowledge between different feature sets. The transfer of knowledge between feature sets has great potential in medical applications where the datasets are large and vague. There are other challenges that surface while dealing with large data such as data fusion and truth discovery. The aim of data cleaning is to identify and rectify the errors in the datasets.  Generative models used in Snorkel DryBell framework can be used for the aforementioned data cleaning challenges.","excerpt":"If a model is tasked with identifying dogs from a random collection, then one can use named-entity recognition to label any content that does not contain the names of dog breeds. Existing knowledge resources can be combined with such simplistic logic to label training data for using it in a new model. This labelling function […]","categories":["AI Features"],"tags":["data labelling"],"author_name":"Ram Sagar","publish_date":"2019-03-25T06:44:25","publication_year":"2019","word_count":571,"keywords":["knowledge graphs","Go","machine learning","TPU","programming_languages:R","AI","RAG","data labelling","Aim","GAN","R"],"extracted_tech_keywords":["AI","machine learning","Aim","RAG","knowledge graphs","TPU","R","Go","GAN","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/why-you-should-use-weak-supervision-for-your-data-labelling-chores\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10048418,"title":"Microsoft To Skill College Students In AI, Cloud, And Cybersecurity, Here’s How To Apply","content":"In line with the Skill India vision, Microsoft has launched the Future Ready Talent program to empower India’s youth with technology skills for employability. A collaborative internship program for students in their second year of college and upwards, the initiative aims to impact over 1.5 lakh higher education students who will join the workforce between 2022-2024. “India’s youth is among the country’s biggest competitive advantages. Empowering this talent with the right skills to be self-reliant will be foundational to our long-term growth. The Future Ready Talent Program provides a holistic skilling experience that connects learners more readily with new job opportunities. We are proud to collaborate with a strong set of partners for this program, who are deeply committed to building a vibrant skilling ecosystem in the country,” said Anant Maheshwari, President, Microsoft India. The registration for the first batch will start on 15 September 2021, and the registration will close on 26 September 2021. Under the program, the All India Council for Technical Education (AICTE), FutureSkills Prime — a NASSCOM and Ministry of Electronics and Information Technology (MeitY) digital skilling initiative, Ernst & Young (EY), GitHub and Quess Corp, along with Microsoft, will provide a holistic skilling platform, connecting talent to opportunity. Eligible students can apply for the internship program here. The Future Ready Talent program has been designed to keep the learner at the centre. With its Learn-Apply-Implement framework, the program will offer students an end-to-end experience, from digital skilling to working on critical projects in a sandbox environment to mentoring with industry experts and access to potential employers. As part of this collaboration: Microsoft will provide learning modules and certifications through its learning platform, Microsoft Learn, skilling students on topics like cloud computing, data & AI, and cybersecurity.AICTE will ensure the curriculum is aligned to the National Education Policy, 2020.SSC NASSCOM will provide alignment of relevant courses to the National Occupational Standards, linking these courses on FutureSkills Prime.EY will provide technology and industry mentorship to students during their internship tenure.GitHub will provide students with free access to the best developer tools via the GitHub Student Developer Pack, as well as the opportunity to collaborate on projects with other developers on GitHub.Quess Corp will manage the learner experience and host a virtual career fair for participants, exposing them to career opportunities across the Microsoft customer and partner ecosystem. In early 2020, Microsoft launched a global skilling initiative to help 25 million people acquire new digital skills worldwide. Over 3 million people have been skilled in India so far through this initiative. Microsoft is working closely with the government, industry and civil society partners on several skilling initiatives to make India’s youth future-ready.","excerpt":"A collaborative internship program for students in their second year of college and upwards, the initiative aims to impact over 1.5 lakh higher education students who will join the workforce between 2022-2024.","categories":["AI News"],"tags":["Microsoft"],"author_name":"kumar Gandharv","publish_date":"2021-09-15T17:17:08","publication_year":"2021","word_count":444,"keywords":["Go","programming_languages:R","cloud computing","AI","programming_languages:Go","Git","Aim","GitHub","R","Microsoft"],"extracted_tech_keywords":["AI","Aim","cloud computing","R","Go","Git","GitHub","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/microsoft-to-skill-college-students-in-ai-cloud-and-cybersecurity-heres-how-to-apply\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10094072,"title":"SoftBank, Nvidia Partners to Build Data Centres in Japan","content":"Taking another step after becoming the 7th company in the world to reach a trillion dollar market cap, NVIDIA has announced its partnership with SoftBank for driving generative AI applications in Japan. SoftBank’s goal is to build 5G\/6G applications with generative AI on NVIDIA’s GH200 Grace Hopper Superchip. These superchips will be used in the data centres that the company is planning  to build across Japan, enabling generative AI and wireless applications on a “multi-tenant common server platform, which reduces costs and is more energy efficient”, as explained by the company. Leveraging NVIDIA’s MGX reference architecture through GH200 Superchips, SoftBank is planning to create breakthroughs in autonomous driving, AI factors, AR, VR, and even digital twins. Apart from AI, the company is also seemingly wanting to invest in the omniverse. CEO of SoftBank, Junichi Miyakawa, said that the era where society coexists with AI is incoming, and SoftBank aims to provide next-generation social infrastructure for a super-digitalised society in Japan. Jensen Huang, founder and CEO of NVIDIA, said, “NVIDIA Grace Hopper is a revolutionary computing platform designed to process and scale-out generative AI services.” NVIDIA has been benefitting from the AI race the most all this while by providing hardware services to OpenAI, and now even SoftBank joins in. SoftBank’s Pursuit of AI The company has decided to move away from its defensive approach in AI, to cautious offence. SoftBank recently reported a total loss of $32 billion through their SoftBank Vision Fund. Most of the investments through this fund were AI and robotics focused startups. SoftBank’s Robotics arms has also been an industry leader when it comes to delivering industry grade robots. In recent news, SoftBank Corporation set up a new AI-focused entity with a specific focus on building a product similar to ChatGPT. Miyakawa has gathered 1,000 people from the company to build a Japanese version of ChatGPT. “We are dead positive on ChatGPT,” said Miyakawa. This was just after Japanese Prime Minister Fumio Kishida pushed the Japanese government to build frameworks for the development of generative AI as early as possible. This is also at the same time when OpenAI’s founder, Sam Altman decided to visit Japan to speak to the Prime Minister about building their office in the country.","excerpt":"Leveraging NVIDIA’s MGX reference architecture through GH200 Superchips, SoftBank is planning to create breakthroughs in autonomous driving, AI factors, AR, VR, and even digital twins.","categories":["AI News"],"tags":["Jensen Huang"],"author_name":"Mohit Pandey","publish_date":"2023-05-30T10:02:31","publication_year":"2023","word_count":373,"keywords":["Go","ChatGPT","OpenAI","AI","Git","Jensen Huang","RAG","GPT","Aim","generative AI","R"],"extracted_tech_keywords":["AI","generative AI","ChatGPT","OpenAI","Aim","RAG","R","Go","Git","GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/softbank-nvidia-partners-to-build-data-centres-in-japan\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10062718,"title":"Best GitHub Learning Repositories for Developers","content":"GitHub has been loved by millions of developers across the globe for years, and justifiably. The portal holds some of the best developers, resources and libraries, making it a go-to choice for anyone in the data science, AI\/ML field wanting to learn more about and excel in it. Analytics India Magazine has outlined the top 10 repositories every aspiring ML developer should follow on GitHub; these consist of all-rounded resources to learn and practice the art, resources that help in learning and roadmaps to work in the industry. Register for this FREE developers workshop >> Free Code Camp Free Code Camp is a non-profit organisation that teaches people how to code. While they are active on several platforms, GitHub is one of the most popular and verified sources to leverage their free code-learning tutorials. The profile has over 200 repositories and several projects, acting as a backend to everything coding. Their repositories include freeCodeCamp’s open-source codebase and curriculum, devdocs-API Documentation Browser, boilerplates – different programming languages, chapter – self-hosted event management tool for non-profits and developer quiz. It also consists of interesting projects such as creating a discord bot, news translation model, periodic table database and more. Find the GitHub repository here EbookFoundation With their mission statement to ‘make the world safer for e-books’, this GitHub repository lists free programming books for students. It consists of thousands of books in 20+ languages, overseeing various concepts and covers hundreds of programming languages. The Free Ebook Foundation maintains this, which administers the repo, a not-for-profit organisation devoted to promoting the creation, distribution, archiving, and sustainability of free e-books. The repository has 221,000+ stars and 6,900+ commits. Find the GitHub repository here Find the resources here 100 Days of ML Code Possibly one of the most detailed repositories, this repo follows a strict 100-day curriculum to learn machine learning. Every day follows a new topic such as data pre-processing, simple linear regression, logistic regression and more. Codes, explanations and detailed infographics accompany these topics. The repo is an easy yet comprehensive way for developers to master basic concepts. Find the GitHub repository here Siraj Raval Siraj Raval is a technologist on a mission to spread data literacy with simpler explanations of difficult concepts. His GitHub profile consists of 500+ repositories delving into various data science topics such as the code for world models, bitcoin creation, polygon cookbook, AlphaCare, code royale source code, recurrent neural networks, computer vision and more. Some of the repositories link to his YouTube channel or other portals. Find the GitHub profile here The Algorithms A verified GitHub repository, The Algorithm is an open-source resource for learning data structures, data algorithms and their implementation in any programming language. Data structures are the core for programming and developing, and this repository explores more than 34 languages, including Python, Java, Go, Java Plus, Lua, Rust, C++ and more. The explanations are easy and comprehensive. Additionally, developers can also leverage their website for access to the codes. Find the GitHub repository here Project-based learning by Tuvtran This repository offers machine learning development learning in a practical manner. The repo consists of programming tutorials teaching software developers how to build applications from scratch in different programming languages and multiple technologies. In addition, the repository supports multiple programming languages and their specific projects. This will help tech enthusiasts enhance their practical skills while creating a project portfolio. Find the GitHub repository here Javascript algorithms The Javascript algorithms on GitHub is the one-stop resource to learn everything about Javascript. The repo contains popular JavaScript algorithms and data structures with explanations and several externally linked resources. The explanations are easily divided and tagged for advanced and beginners and come with their separate README explanations, YouTube videos and further reading links. In addition, the explanations are available in several languages. Find the GitHub repository here An additional resource to learn JavaScript: 30 Days of Java Script An additional resource to learn JavaScript: You Don’t Know JS Public APIs The Public APIs repository is a collective list of free APIs for software and web development use. APIs are integral to development, and this repository finding useful APIs makes it easy for developers to find free and relevant APIs in the category. In addition, the list is constantly updated with newer APIs in the public domain. Find the GitHub repository here Coding Interview University The coding interview university is one of the best-received repositories on GitHub with 192,000 stars and is currently #1 on trending. Owned by a (now) software engineer at Amazon, the repository started as a study plan of the owner before Amazon and is now the chosen resource for any engineer to crack coding interviews and make it into the big tech firms. The repo is a multi-month study plan to crack the interview of any big tech giant, starting from the basics of programming to the advanced concepts of data structures, system design, and core CS concepts. The repository contains a study plan, topics, ways to find a job, online materials, books, videos, lectures about the different Software Engineering topics, CV writing tips, job application tips, and more. Find the GitHub repository here Developer Roadmap Another helpful resource for developers to make it in the industry, the developer roadmap repository provides regularly updated roadmaps for frontend developers, backend developers, React developers, DevOps engineers and more. It is a chosen resource for developers wanting to start a career or upgrade their skills. Find the GitHub repository here","excerpt":"Analytics India Magazine has outlined the top 10 repositories every aspiring ML developer should follow on GitHub; these consist of all-rounded resources to learn and practice the art, and roadmaps to work in the industry.","categories":["AI Trends"],"tags":[],"author_name":"Avi Gopani","publish_date":"2022-03-15T11:00:00","publication_year":"2022","word_count":910,"keywords":["data science","machine learning","AI","neural network","ML","computer vision","RAG","Python","analytics","R"],"extracted_tech_keywords":["AI","machine learning","ML","neural network","computer vision","data science","analytics","RAG","Python","R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/best-github-learning-repositories-for-developers\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":12773,"title":"Edge Analytics&#8211; taking data processing from cloud to edge of network","content":"By driving faster insights, edge analytics has emerged as a game changer in retail, banking and other sectors One of the most significant trends in the recent years is the emergence of analytics at the edge, the game changer in industrial IoT, retail, manufacturing, finance and energy sectors. The path breaking trend is spurred by a shift from the cloud to the edge wherein advanced computing power enables analyzing data close to the network source or at the edge instead of dispatching it to a remote server. Edge analytics not only helps tackling the data deluge but reduces the traffic on overburdened networks by decentralizing data storage and enabling localized analytics. According to a research, the edge analytics is billed to grow from $1.94 Billion in 2016 to $ 7.96 Billion by 2021. Some of the sectors leveraging edge analytics are BFSI, manufacturing, retail and consumer goods, energy and utility among others. Edge analytics gained currency in 2015 when the CEO of Mississippi-based Camgian Microsystems, maker of advanced sensing platform technologies for the military and commercial purpose authored a blog post titled Intelligent Not Connected Things Will Drive IoT. In the post Gary Butler, challenged the old idea of sensors collecting raw data and dispatching it to a server in the cloud. The “Connected Thing Paradigm” worked on these levels – collecting, storing and processing information which was later visualized through user interface. Marking a departure from the centralized computing, Butler reimagined the IoT architecture, detailing an “Intelligent Thing Paradigm” where data is processed locally at the edge of the network through edge real-time analytics. This was a distributed processing architecture instead of a cloud or centralized processing in a data warehouse. Case in point, in a retail setting, beacons installed inside the store are edge devices that collect data before sending it to cloud and in an industrial setting, smart machines are edge devices. If the processing happens close to the source, it is termed edge computing and if it happens in cloud – such as cloud hosted solutions AWS or Microsoft’s Azure, it is called cloud computing. Talk about real-time edge analytics brings to fore the old argument — edge computing vs cloud computing. Edge computing vs Cloud Computing Edge computing is not replacing cloud computing in any way. In fact, it accelerates real-time responses with edge analytics by processing close to the source of data. Edge hasn’t replaced cloud, in fact edge and cloud architecture is mission critical in the industrial world. See how GE is capitalizing real-time analytics by allowing it to run at various levels – “edge sensor, controller, gateway, infrastructure machine, within on-premise appliances and racks, or in the cloud. “ With hardware cost coming down and the phenomenal growth in sensor technologies, collection of datasets from remote locations has become more streamlined. The edge-to-cloud computing paradigm helps in two major ways a) speeds up the collection of data locally; b) applicants can perform localized analytics; c) efficient scalable model. Edge analytics enables decision making at the node level, thereby bringing down the cost of data transfer and data storage significantly. Leveraging the Edge Advantage One of the biggest drivers of edge analytics has been industrial giant GE that has added intelligence to machines at the edge of network instead of transferring and storing data in different data centers or the cloud. According to Nikihil Chauhan, Director, Connectivity and Controls Product Marketing, GE Digital, “Edge computing is not a new concept. One of the major concerns edge computing addresses is that it he opens a door by leading great value of customer data in Industrial applications, in the same way as smartphones opened the door leading to value of consumer data in mobile apps?” GE is leveraging software defined machines (SDMs) such as turbines, locomotives and streetlights by leveraging the edge-to-cloud approach. Case in point, is GE Transportation’s Evolution Series Tier 4 Locomotive that carries more than 200 sensors that collect gigabytes of information and crunches one billion plus instructions per second.  GE’s Tier 4 locomotive deploys on-board edge computing to analyze the gigabytes of data and further runs algorithms. Edge Analytics in Banking But it is not just IoT where edge analytics tackles the data deluge. IBM Research taps into edge analytics for banking where it helps banks understand the customer better by providing insights such as location-based suggestions and customer recommendations. Embedded in the bank’s customer channels – online banking or mobile banking, edge analytics delivers transactional behavior and location based suggestions in real time. Edge Analytics in Retail From picking up data in real time to delivering push notifications, analytics at the edge is transforming how retailers can capture a greater share of the user’s wallet. Retail is seeing a notable data frenzy –from beacons installed in the store, to sensors, in-store Wi-Fi networks and data generated from apps; much of the unstructured data produced is left unused.  Leading retailers are leveraging edge analytics to deliver better user experience and maximize store performance.  From location analytics to drive engagement to understanding shopper’s pattern, retail giants Walmart and Target among others are using analytics at the edge of network to draw insights from terabytes of data. There has also been some innovation and inventions and buyouts along the way that has spurred the growth of edge analytics. Case in point, Cisco acquired edge analytics provider ParStream in 2015. The Germany-based company has a specialized database that can handle high volumes of data and delivers real-time analytics at the edge. HPE Edgeline IoT Systems: In 2015, HPE rolled out Edgeline IoT systems 10 and 20 (edge gateway) in association with Intel that can keep data localized and reduce network traffic. Aimed at industries such as logistics, transport and retail, EdgeLine IoT systems use less space and energy as opposed to servers. Our Take – how edge analytics can be applied in sector Edge analytics has become the de facto industry practice and the main drivers are dwindling data storage capabilities, reducing network traffic and a share spike in computing power. That’s why, leading technology vendors provide filtering and processing capability at the edge. We believe edge analytics can come into play in sectors such as farming and agriculture largely wherein regardless of the network, analytics can point out equipment failure or irrigation leaks.","excerpt":"One of the most significant trends in the recent years is the emergence of analytics at the edge, the game changer in industrial IoT, retail, manufacturing, finance and energy sectors. The path breaking trend is spurred by a shift from the cloud to the edge wherein advanced computing power enables analyzing data close to the […]","categories":["IT Services"],"tags":[],"author_name":"Richa Bhatia","publish_date":"2017-02-15T09:00:06","publication_year":"2017","word_count":1050,"keywords":["AWS","AI","cloud computing","R","ML","RAG","Aim","analytics","edge computing","Azure"],"extracted_tech_keywords":["AI","ML","analytics","Aim","RAG","cloud computing","AWS","Azure","edge computing","R"],"url":"https:\/\/analyticsindiamag.com\/it-services\/edge-analytics-taking-data-processing-from-cloud-to-edge-of-network\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10022603,"title":"Transformers For Vision: 7 Works That Indicate Fusion Is The Future Of AI","content":"Transformers are all geared up to rule the world of computer vision. The runaway success of OpenAI’s CLIP and DALL.E had a lot to do with the sudden interest for multi-model machine learning in research circles. OpenAI co-founder Ilya Sutskever has even forecasted a future gravitating towards these fusion models, that can handle both language and vision tasks. Apart from the recent OpenAI releases, other works have also attracted a great deal of attention in this subdomain. Let’s take a look at a few of the recent popular works in this field: Vision Transformer Google Vision Transformers (ViT) try to replicate the Transformers architecture of natural language processing as close as possible. ViT represents image inputs as sequences and predicts class labels for the image, allowing models to learn image structure independently. Similar to natural language processing, ViT treats input images as a sequence of patches. Every patch is flattened into a single vector by concatenating the channels of all pixels in a patch and then linearly projecting it to the desired input dimension. According to Google, ViT outperforms state-of-the-art CNN with four times fewer computational resources when trained on sufficient data. Data-efficient Image Transformer This year, Facebook announced Data-efficient Image Transformer (DeiT), a vision Transformer that improved on Google’s research on ViT. However, they built a transformer-specific knowledge distillation procedure based on a distillation token to reduce training data requirements. While distillation allows one neural network to learn from another, a distillation token is a learned vector that flows through the network along with the transformed image data to significantly enhance the image classification performance with less training data. DeiT can be trained with 1.2 million images instead of hundreds of millions of images required for CNN. Image GPT OpenAI developed Image GPT — an image Transformer that can generate coherent images when trained on pixel sequences. The model understands the 2-D representation for object detection when trained on GPT-2 with long sequences of pixels. The model outperforms several benchmarks without an extensive labelled dataset while training. Though the researchers used the same Transformer architecture of GPT-2, it delivered superior performance even in diverse environments, demonstrating the capability of Transformers for image tasks. CLIP: Contrastive Language-Image Pre-training CLIP efficiently learns visual concepts from natural language supervision.  It can be applied to any visual classification benchmark by simply providing the names of the visual categories to be recognised, similar to the “zero-shot” capabilities of GPT-2 and GPT-3. This technique eliminates the need for proper labeled data to train models, moving away from the cost-intensive approach. The idea is to make models more flexible by training with a wide range of images instead of specific images that traditional neural networks use. DEtection TRansformer – DETR DE⫶TR: End-to-End Object Detection with Transformers https:\/\/t.co\/4W27PVx6Js (+paper https:\/\/t.co\/xmxnSOBjBa) awesome to see a solid swing at (non-autoregressive) end-to-end detection. Anchor boxes + nms is a mess. (I was hoping detection would go end-to-end back in ~2013) pic.twitter.com\/cN4hK2ZFY0— Andrej Karpathy (@karpathy) May 27, 2020 DEtection TRansformer is a set-based global loss that forces unique predictions via bipartite matching, and a transformer encoder-decoder architecture. DETR uses CNN as its core for representation of input image and then uses a positional encoder before passing it to a transformer encoder. The approach streamlines the detection pipeline by removing the dated practice of non-maximum suppression procedure or anchor generation for object detection while ensuring the model can generalise better than the state-of-the-art models. DALL.E Based on GANs, TransGAN is another text-to-image synthesis model that does not use convolution operations. Both aspects of GAN — generator and discriminator — are entirely based on Transformers and are memory-efficient that progressively increases feature resolution, decreases embedding dimension, and more. With the new interest in using Transformers in vision tasks, the researchers believe that two Transformers will allow researchers and developers to simplify most computer vision problems shortly. TransGAN: Transformers Based GAN Model Based on GANs, TransGAN is another text-to-image synthesis model that does not use convolution operations. Both aspects of GAN — generator and discriminator — are completely based on Transformers and are memory efficient that progressively increases feature resolution, decreases embedding dimension, and more. With the new interest of using Transformers in vision tasks, the researchers believe that two Transformers can in future allow researchers and developers to simplify most of the computer vision problems.","excerpt":"Transformers are all geared up to rule the world of computer vision. The runaway success of OpenAI’s CLIP and DALL.E had a lot to do with the sudden interest for multi-model machine learning in research circles. OpenAI co-founder Ilya Sutskever has even forecasted a future gravitating towards these fusion models, that can handle both language […]","categories":["AI Trends"],"tags":["Generative Pre-Trained Transformer"],"author_name":"Ram Sagar","publish_date":"2021-03-22T13:00:00","publication_year":"2021","word_count":722,"keywords":["Go","machine learning","OpenAI","AI","neural network","ML","Transformers","computer vision","object detection","Generative Pre-Trained Transformer","R"],"extracted_tech_keywords":["AI","machine learning","ML","neural network","computer vision","OpenAI","Transformers","object detection","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/transformers-for-vision-7-works-that-indicate-fusion-is-the-future-of-ai\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":39487,"title":"5 Reasons Why Python Is The Dominant Language For Machine Learning","content":"“Rome was not built in a day” and neither was the prominent Python programming language. But, this general-purpose, high-level language has started to beat other languages in a relatively shorter period of time. Our Annual Survey on Data Science Recruitment shows that almost 75% of respondents said that it was crucial for job seekers to know Python language to get an opportunity in data science According to the Stack Overflow Survey 2018, Python is the most wanted language for the second year in a row, which means it is the language that developers who do not yet use it most often say they want to learn. It is also claimed to be the fastest-growing major programming language. Developers and pioneers around the globe are implementing this language for machine learning projects. Current contenders for being the “ML programming language”: – Python– Swift– Julia Sadly the latter two stand little chance of wide adoption: major PLs come by roughly every two decades: C, C++, Java, Python each have 20 yrs inbetween. Seems like Python has won this round. — Reza Zadeh (@Reza_Zadeh) May 19, 2019 In this article, we list down 5 reasons how Python became the dominant language in the field of machine learning. Easy To Learn The bonus point about Python is this language does not need any hardcore programmer to work on it. A beginner with a basic knowledge of programming language can easily get their hands dirty with Python. It can be considered as the beginner’s friendly language. If a developer is working on a machine learning project in python, he\/she can spend more time on developing the model rather than fixing bugs. This independent language can also be called as one of the most flexible languages across different platforms and technologies. The language provides an option to choose between OOPs approach and scripting. Vast Community Python is well-known for its extremely vast community. Since its release, the community is contributing largely and which is the reason a Python user never feel abandoned with sudden changes. The constant upgrade by the developer community support makes Python one of the most suitable languages for machine learning applications. Large organisations such as Google, Amazon, Facebook, etc.are using this language. Documentation Not only libraries but this language has extensive tutorials and documentation. Readability is a primary focus for Python developers, in both project and code documentation. The community provides quick development, rapid prototyping, friendly syntax, etc. to develop cognitive systems. The developers around the globe provide comprehensive support and help through tutorials, forums, etc, which makes it easy to code for any project. Versatility Python has crossed more than two decades and it became more versatile as time passes. At the present scenario, it can be implemented on almost every operations, software development, working and managing cloud infrastructures, etc. This versatile language supports object-oriented programming, structured programming, and functional programming patterns, etc. and can be applied not only in projects like machine learning, data science but also in gaming, development, web frameworks, networking, etc. It also serves as a data handling backend tool with Spark support in order to manage a large amount of data. Frameworks & Libraries One thing the developers like most about Python is the abundance of open source libraries and frameworks. The language has a great number of machine learning libraries and some of the prominent libraries are such as TensorFlow, Pytorch, Matplotlib, SciKit Learn, etc. Python has a collection as well as code stack of various open source repositories in almost every domain such as Django for integrating web applications, pandas for machine learning, SciPy for scientific computing, Librosa for audio, OpenCV for images, NumPy for text, etc. Other commonly used libraries for artificial intelligence and machine learning are such as pyDatalog (Logic Programming engine in Python), PyML (a bilateral framework written in Python that focuses on SVMs and other kernel methods), EasyAI (Simple Python engine for two-players games with AI), PyBrain (simple and effective algorithm for ML tasks), AIMA (Python implementation of algorithms), SimpleAI, etc.","excerpt":"“Rome was not built in a day” and neither was the prominent Python programming language. But, this general-purpose, high-level language has started to beat other languages in a relatively shorter period of time. Our Annual Survey on Data Science Recruitment shows that almost 75% of respondents said that it was crucial for job seekers to […]","categories":["AI Trends"],"tags":["django python","Machine Learning","Python","python machine learning"],"author_name":"Ambika Choudhury","publish_date":"2019-05-21T11:33:44","publication_year":"2019","word_count":671,"keywords":["data science","artificial intelligence","machine learning","AI","TensorFlow","PyTorch","ML","Machine Learning","OpenCV","Python","Aim","django python","python machine learning","Pandas"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","data science","Aim","TensorFlow","PyTorch","OpenCV","Pandas"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/5-reasons-why-python-is-the-dominant-language-for-machine-learning\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10043688,"title":"Guide To Text Classification using TextCNN","content":"Nowadays, many actions are needed to perform using text classification like hate classification, speech detection, sentiment classification etc. This article’s main focus is to perform text classification and sentiment analysis for three combined datasets amazon review, imdb movie rating and yelp review data sets using . Before going to the coding, let’s just have some basics of text classification and convolutional neural networks. Introduction to Text Classification What is Text Classification? Text classification is a process of providing labels to the set of texts or words in one, zero or predefined labels format, and those labels will tell us about the sentiment of the set of words. First of all, the human language is nothing but a combination of words. Whenever spoken by the human it comes out with a sentiment that another human can easily understand. Humans easily understand whether a sentence has anger or it has any other mood. Making a machine to understand the human language is called text classification. To perform text classification, we need already classified data; here in this article, the data used is provided with the labels. So here we are, trying to make a model with three data sets; as I said before, every piece of data has sentences with labels 0 and 1. At the end of the building model, the model will try to classify sentences according to their sentiment. This model will take text as an input and analyze input information and assign labels to them. Let’s just make a simple logistic model for understanding it more. In this article, I am using google colab. First of all, we will import the data using pandas. Input: import pandas as pd filepath_dire = {'yelp':   '\/content\/drive\/MyDrive\/Yugesh\/TextCNN\/yelp_labelled.txt', 'amazon': '\/content\/drive\/MyDrive\/Yugesh\/TextCNN\/amazon_cells_labelled.txt', 'imdb':   '\/content\/drive\/MyDrive\/Yugesh\/TextCNN\/imdb_labelled.txt'} data_list = [] for source, filepath in filepath_dire.items(): data = pd.read_csv(filepath, names=['sentence', 'label'], sep='\\t') data['source'] = source  # Add another column filled with the source name data_list.append(data) data = pd.concat(data_list) print(data) Output: It seems like we have imported every dataset correctly. So let’s just move towards the model building, but we need to perform some preprocessing before fitting a model to a dataset. Let’s just think about the process of the model, how it will work internally, how it will calculate the label from the text. Internal things are dependent on mathematical evaluation and calculation. Here also the data, the model’s input needs to be in matrices or numeric values formats that the model can easily calculate. Transformation of the words can be done in many ways. One of them is to count the number of occurrences of every word in each sentence and provide those counts to the entire set of words in the dataset. This kind of collection of information is called corpus in NLP. And another method is to make a vocabulary where every word has its special index number.  More formally, we can say classifying every word into its tied index. This can be easily done by using the CountVectorizer provided by the scikit-learn library. Lets look at an example. Input : example = ['analytics india magazine is good magazine ', 'analytics india magazine provides good information'] Next, we can vectorize the sentence using Countvectorizer. Input: from sklearn.feature_extraction.text import CountVectorizer examplevectorizer = CountVectorizer() examplevectorizer.fit(example) examplevectorizer.vocabulary_ Output: {'analytics': 0, 'good': 1, 'india': 2, 'information': 3, 'is': 4, 'magazine': 5, 'provides': 6} This resulting vector is called a feature vector. Each word has its category in the feature vector, which can be represented in numeric terms. The output in the example has a vocabulary with a special index provided to words. This feature vector later can be converted into an array of occurrence of words which we help to count the frequency of words in the data set. Input: examplevectorizer.transform(example).toarray() Output: array([[1, 1, 1, 0, 1, 2, 0], [1, 1, 1, 1, 0, 1, 1]]) Let’s perform this to our data set. First, we need to split our data into train and test. Input: from sklearn.model_selection import train_test_split review = data['sentence'].values label = data['label'].values review_train, review_test, label_train, label_test = train_test_split( review, label, test_size=0.25, random_state=1000) Vectorizing the split data. Input: from sklearn.feature_extraction.text import CountVectorizer review_vectorizer = CountVectorizer() review_vectorizer.fit(review_train) Xlr_train = review_vectorizer.transform(review_train) Xlr_test  = review_vectorizer.transform(review_test) Xlr_train Output: Here we can see that the matrix has 750 feature vectors, and each has 1714 dimensions, which is the size of the vocabulary. Now the data is almost prepared for fitting in the model, lets perform the logistic regression model building and fitting in the data. Input: from sklearn.linear_model import LogisticRegression LRmodel = LogisticRegression() LRmodel.fit(Xlr_train, label_train) score = LRmodel.score(Xlr_test, label_test) print(\"Accuracy:\", score) Output: Accuracy: 0.8195050946142649 Here we have seen the text classification model with very basic levels. There are many methods to perform text classification. TextCNN is also a method that implies neural networks for performing text classification. First, let’s look at CNN; after that, we will use it for text classification. Introduction to CNN Convolutional neural networks or CNN are among the most promising methods in developing machine learning models. For example, it performs so well in image classification and computer vision. What are Convolutional neural networks or CNN? CNN is just a kind of neural network; its convolutional layer differs from other neural networks. To perform image classification, CNN goes through every corner, vector and dimension of the pixel matrix. Performing with this all features of a matrix makes CNN more sustainable to data of matrix form.Convolutional layers consist of multiple features like detecting edges, corners, and multiple textures, making it a special tool for CNN to perform modeling. That layer slides across the image matrix and can detect its all features. This means each convolutional layer in the network can detect more complex features. As the feature expands, we need to expand the dimension of the convolutional layer. We can consider text data as sequential data like data in time series, a one-dimensional matrix. We need to work with a one-dimensional convolution layer. The idea of the model is almost the same, but the data type and dimension of convolution layers changed. To work with TextCNN, we require a word embedding layer and a one-dimensional convolutional network. Word Embedding What is Word Embedding? Word embedding represents the density of the word vector, unlike what we have done with the Countvectorizer. It is a different way to preprocess the data. This embedding can map semantically similar words. It does not consider the text as a human language but maps the structure of sets of words used in the corpus. They aim to map words into a geometric space which is called an embedding space. If embedding finds a good relationship between works like for an example King – man + women = queen Keras provides a couple of methods for text preprocessing and sequence preprocessing. We can use them to make our data a better fit for the TextCNN model. Let’s prepare word embeddings for the model. input: from keras.preprocessing.text import Tokenizer tokenizer = Tokenizer(num_words=5000) tokenizer.fit_on_texts(review_train) Xcnn_train = tokenizer.texts_to_sequences(review_train) Xcnn_test = tokenizer.texts_to_sequences(review_test) vocab_size = len(tokenizer.word_index) + 1 print(review_train[1]) print(Xcnn_train[1]) Output: Here we can see that most common words do not have a large index in our embedding space. Still, the extremely uncommon words will get a higher index value which will be word count + 1 because they hold some information. Those whose occurrence is moderate will be given a moderate index value. Finally, 0 value is reserved and won’t be provided to any text. One problem is that in each sequence is the different length of words, and to specify the length of word sequence, we need to provide a mexlen parameter and to solve this, we need to use pad_sequence(), which simply pads the sequence of words with zeros. Input: from keras.preprocessing.sequence import pad_sequences maxlen = 100 Xcnn_train = pad_sequences(Xcnn_train, padding='post', maxlen=maxlen) Xcnn_test = pad_sequences(Xcnn_test, padding='post', maxlen=maxlen) print(Xcnn_train[0, :]) Output: After padding, we have appended zero value to matrices, and now we can use those in a deep learning model. This is how word embedding makes relations between words. In the next step, we will try to fit the TextCNN model. First of all, we need to import sequential and layers. Input: from keras.models import Sequential from keras import layers Making models using layers in it. embedding_dim = 200 textcnnmodel = Sequential() textcnnmodel.add(layers.Embedding(vocab_size, embedding_dim, input_length=maxlen)) textcnnmodel.add(layers.Conv1D(128, 5, activation='relu')) textcnnmodel.add(layers.GlobalMaxPooling1D()) textcnnmodel.add(layers.Dense(10, activation='relu')) textcnnmodel.add(layers.Dense(1, activation='sigmoid')) textcnnmodel.compile(optimizer='adam', loss='binary_crossentropy', metrics=['accuracy']) textcnnmodel.summary() Output: Model: \"sequential_1\" _________________________________________________________________ Layer (type)                 Output Shape              Param # ================================================================= embedding_1 (Embedding)      (None, 100, 200)          920600 _________________________________________________________________ conv1d_1 (Conv1D)            (None, 96, 128)           128128 _________________________________________________________________ global_max_pooling1d_1 (Glob (None, 128)               0 _________________________________________________________________ dense_2 (Dense)              (None, 10)                1290 _________________________________________________________________ dense_3 (Dense)              (None, 1)                 11 ================================================================= Total params: 1,050,029 Trainable params: 1,050,029 Non-trainable params: 0 _________________________________________________________________ Let’s fit the model and check for accuracy. Input : textcnnmodel.fit(Xcnn_train, label_train, epochs=10, verbose=False, validation_data=(Xcnn_test, label_test), batch_size=10) loss, accuracy = textcnnmodel.evaluate(Xcnn_train, label_train, verbose=False) print(\"Training Accuracy: {:.4f}\".format(accuracy)) loss, accuracy = textcnnmodel.evaluate(Xcnn_test, label_test, verbose=False) print(\"Testing Accuracy:  {:.4f}\".format(accuracy)) Output: Training Accuracy: 1.0000 Testing Accuracy:  0.8040 Here we can see that our model is overfitted at training, but test accuracy is decent. Hence,  there are many ways to improve the model. In this article, we have not performed the cleaning of the data, and the CNN requires a large amount of data to train better, so as the sample will increase, it might perform better. But after overfitting too, it gave quite good results. There are many cases where we will need to use this for a large and complex data set. It is well suggested that our simple classification model can not perform well. For example, in complex datasets where the making of vocabulary increases the size of matrices, we can use this model because we know that it looks for relationships between words.","excerpt":"Text classification is a process of providing labels to the set of texts or words in one, zero or predefined labels format, and those labels will tell us about the sentiment of the set of words.","categories":["Deep Tech"],"tags":["CNN deep learning","cnn neural network","Guide","NLP models","text classification"],"author_name":"Yugesh Verma","publish_date":"2021-07-18T13:00:00","publication_year":"2021","word_count":1611,"keywords":["text classification","CNN deep learning","machine learning","Keras","AI","neural network","cnn neural network","NLP models","computer vision","NLP","Ray","Aim","deep learning","analytics","Guide"],"extracted_tech_keywords":["AI","machine learning","deep learning","neural network","NLP","computer vision","analytics","Aim","Ray","Keras"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/guide-to-text-classification-using-textcnn\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10040246,"title":"Pros &#038; Cons Of AI In Tax Systems","content":"Throughout history, we have been actively interested in the potential of artificial intelligence and automation. Today, while people still do not use flying cars as often as Back to the Future 2 would like us to, we are actualising the dream of adopting AI into everyday lives, including in processes such as taxation. Taxation is a primary source through which governments make money to support the public sector. It is imperative to make taxation more efficient but there are challenges such as human errors and tax evasions. What if we had systems in place that took care of these problems in a better manner? This is where AI comes in. Automation in taxation AI could automate repetitive processes in taxation systems such as processing documents and filling out account numbers or IDs onto spreadsheets. This helps in efficiently filing taxes and reducing the overall processing time. AI could also help single out essential aspects of a document. Machine Learning algorithms can classify documents, point out data on account IDs, number of payments on the bill, etc., accurately, thus simplifying the tax payment process. Such classification and filtering can considerably reduce corruption in tax payments and make the process more transparent. Auditing In India, tax officials detected tax fraud worth Rs. 37,946 crores in 2018-19 and Rs. 6,520 crore the year after that. The Finance Ministry reported 1,620 cases of fake invoices summing up to a total value of Rs 11,251 crores. Tax evasion is thus a significant problem, which requires immediate remediation. In this case, too, AI-enabled systems can assist by scanning through large datasets in a shorter period making use of criteria such as employment status, audit history—among others. That said, such an AI system would need to be familiar with tax codes. In 2020, the Indian government officially announced a ‘Transparent Taxation Platform’ to ease compliance and reward honest taxpayers. The platform uses data analytics, artificial intelligence, and machine learning to help curb tax fraud and reduce personal bias and bribery. Although, Finance minister Nirmala Sitharaman, called this programme ‘a landmark in the history of tax administration’, many tax officials called out on the inadequate resources used in implementing the programme. Barriers to AI adoption Regardless of the many benefits of using AI in tax systems, its implementation might not be that simple. One barrier to AI adoption is the handling of data. Different teams collect data in a way that suits their respective productive capabilities. Due to lack of standardisation, the collection and processing of data may differ on a team-to-team basis. This confuses the AI system that works logically based on identifying patterns from data sets. Daren Campbell, EY America’s Tax Innovation Leader, cited an example where human-captured data was used to train an AI model to classify the large volumes of sales-and-use tax data. Factors such as vendors, descriptions, and accounts were coded in different ways and assigned to separate categories. This confused the AI system and the process was not successful. Another reason why AI systems are not universally adopted in tax systems is the element of trust. Such mistrust could be over AI\/ML algorithms and how they would make predictions or analyse tax returns to derive insights. It could also arise from the parameters used to detect tax fraud. Finally, many functions of AI would also require human intervention. In 2020, the Madras High Court order ruled that even though the e-governance system introduced by the I-T department from 2017-18 was laudable, the e-proceedings would still require human intervention to avoid erroneous assessments. Wrapping up The use of AI in tax systems can help reduce certain biases and corruption. However, a lot needs to be taken care of—especially to make people trust AI systems. One way to do this could be by making the workings of such systems more transparent by explaining how an algorithm reaches an outcome or disclosing the parameters used to detect and analyse certain aspects of AI models. With the world moving fast ahead and volumes of data skyrocketing, implementing AI—at least to some extent—would make our existing systems much more efficient and accurate.","excerpt":"AI would make our existing systems much more efficient and hence allow tax officials to work with more data faster and more accurately.","categories":["AI Features"],"tags":["fraud detection","ml algorithms"],"author_name":"Mita Chaturvedi","publish_date":"2021-05-18T10:00:00","publication_year":"2021","word_count":685,"keywords":["Go","machine learning","artificial intelligence","ml algorithms","AI","innovation","ML","automation","analytics","Rust","R","fraud detection"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","analytics","R","Go","Rust","automation","innovation"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/pros-cons-of-ai-in-tax-systems\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":60392,"title":"Top 5 Tools For Virtual Conferencing","content":"Tech companies have been forced to postpone their conferences due to the proliferation of Covid-19 cases in the world. However, except for a few places like China and Iran, coronavirus is showing no signs of slowing down. Consequently, tech companies will eventually have to shift to virtual if they do not want to delay any further. Undoubtedly, virtual events cannot offer the same experience, but they have their advantages. Firstly, they open it up to a broader audience with ease. Furthermore, they also reduce carbon footprint and come at a much lower cost. But going completely digital is not an easy task and requires a robust tool at the helm to run a virtual event effectively. Here are the top tools you can leverage to host virtual conferencing effectively: vFairs vFairs is a virtual events platform that enables organizations to connect global audiences and maximize engagement. It simulates a real event with eye-catching visuals as it allows one to navigate the venue virtually. The mobile-friendly, fully customizable and highly interactive platform enables attendees to virtually visit booths and interact with participants through video and chat. It also delivers 3D layouts to render real-world physical conferences and makes the conference an unforgettable experience. The platform has gained a lot of traction in recent times with various companies – including IBM, Starbucks, and VMUG – leveraging it for their events. vFairs is the go-to platform for managing entire conferences virtually as it assists the host in handling registration, helping them connect with the audience in real-time, and allowing them to showcase branding from different companies. Hopin An all-in-one live online event platform Hopin can be leveraged for events of all sizes. It allows the host to manage registrations, event booths, event sessions, event stage, and more. Hopin has catered to the needs of over 200 events and can engage more than 30,000 attendees at a time. A vital facet of the platform is its ability to enable networking as it connects people through video through online meeting rooms. This expands the reach and visibility in virtual events. Founded in 2019, the company was able to demonstrate its capabilities by raising €6 million in February 2020 to expand its offerings. 6Connex 6Connex empowers organizations to host infinitely scalable virtual summits at ease. While the platform allows one to create interactive events a few weeks before the actual event, unlike others, it does not limit the host to making changes in the environment during the event. Utilizing 6Connex, companies can not only replicate the physical conference but can also go a step further. The tool provides features like real-time multi language-translation tools to remove communication barriers during events. Besides, attendees can also locate others within the event, share virtual business cards, and send invitations to connect on social network platforms. INXPO INXPO allows companies to host virtual events of any size and tap into 157 countries without hassle. It also provides analytics to realize ROI, drive engagement and increase revenue. INXPO was among the first to offer HTML5 streaming services to reach their audiences authentically. It is easy to navigate and caters to different needs by providing ‘essentials’ and ‘expert’ variants, thereby making it a go-to platform for a wide range of organizations. While the essentials are for organizations who do not focus on sponsors, the expert allows the host to take care of the branding of sponsors. Various blue-chip companies such as AON, AWS, Dell, SAP, Microsoft, among others, have been dependent on the platforms not only for virtual events, but also for webcasts, and other live streaming services. WorkCast Started in 2008, WorkCast has been used in more than 8,000 events and used by more than 1 million attendees across 20 countries. The platforms do not require plugins to expand the functionality as it offers all the necessary features. Depending on the needs, an organization can either host multi- or single-session and pay accordingly. Its product can support up to 50,000 attendees at a time. They can either attend the live virtual environment where the season is in real-time or watch the recorded content.","excerpt":"Tech companies have been forced to postpone their conferences due to the proliferation of Covid-19 cases in the world. However, except for a few places like China and Iran, coronavirus is showing no signs of slowing down. Consequently, tech companies will eventually have to shift to virtual if they do not want to delay any […]","categories":["AI Trends"],"tags":["tools for virtual conferencing"],"author_name":"Rohit Yadav","publish_date":"2020-03-30T17:00:00","publication_year":"2020","word_count":682,"keywords":["tools for virtual conferencing","Go","AWS","AI","ML","Scala","Git","RAG","analytics","GAN","R"],"extracted_tech_keywords":["AI","ML","analytics","RAG","AWS","R","Go","Scala","Git","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-5-tools-for-virtual-conferencing\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10131705,"title":"Bengaluru-based Medical AI Startup SigTuple Raises $4M, to Expand Operations","content":"SigTuple, a Bengaluru-based medtech startup, has secured $4 million (INR 33 crores) in an extended Series C funding round led by SIDBI Venture Capital. The funding also saw participation from existing investors, including Endiya Partners and strategic leaders from the healthcare sector. The fresh capital will be used to drive SigTuple’s geographical expansion, broaden its product portfolio, and support regulatory clearances. This latest investment brings the company’s total funding to $50 million since its inception in 2015. The AI-powered startup was founded by Rohit Kumar Pandey, Tathagato Rai Dastidar and Apurv Anand. In 2015, SigTuple raised $16M in a Series C funding, and previous to that had raised $19M. AI in Healthcare SigTuple has been making significant strides in the digital pathology space. Its flagship product, AI100, which automates manual microscopic reviews using AI and robotics, has gained traction in the Indian market and expanded into Southeast Asia, the Middle East, and North Africa. The company is now poised to enter European and American markets. In a major milestone, AI100 received 510(k) clearance from the US Food and Drug Administration (FDA) in September 2023, making SigTuple the third company globally and the first in India to achieve this for AI-assisted digital hematology. Tathagato Rai Dastidar, founder & CEO of SigTuple said, “While we continue to build on the success of AI100 in India and abroad, 2024 will witness two new major product launches addressing a wide segment of the diagnostic industry, which will help make SigTuple a global brand coming out of India. We are truly excited to welcome SIDBI Venture Capital on board as the lead investor in this round. Their support is going to go a long way in making our dream of going global a reality.” The company is set to launch two new major products in 2024. One is a next-generation device that will automate all manual microscopy in clinical labs, surpassing the capabilities of AI100. Additionally, SigTuple plans to enter the point-of-care market with a device leveraging microfluidic technology and imaging to conduct essential tests within minutes. This funding round and product expansion plans underscore SigTuple’s commitment to revolutionising the diagnostic industry by making diagnostics decentralised and fully automating microscopic reviews of diseased samples. Interestingly, in recent times there have been a number of AI developments in healthcare. In addition to big tech companies such as Google, Microsoft, Oracle, and others heavily investing in this segment. The companies are bringing LLM- based diagnostic measures in addition to other features to assist doctors.","excerpt":"Binny Bansal serves on the board of SigTuple.","categories":["AI News"],"tags":["bangalore","Startups"],"author_name":"Vandana Nair","publish_date":"2024-08-07T11:03:45","publication_year":"2024","word_count":417,"keywords":["Go","API","funding","AI","RPA","venture capital","Git","RAG","Startups","R","bangalore","startup"],"extracted_tech_keywords":["AI","RAG","R","Go","Git","API","RPA","startup","venture capital","funding"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/bengaluru-based-medical-ai-startup-sigtuple-raises-4m-to-expand-operations\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10016135,"title":"Major Hiring And Recruiting Announcements Made By Tech Giants In 2020","content":"The year 2020 was unpredictable in terms of hiring and recruiting for tech jobs. It started with many tech companies declaring an increase in mass recruitment. However, as the pandemic struck and the lockdown followed, there was an obvious decline in the number of tech jobs, and there were multiple layoffs by tech firms. However, as things open up, India is slowly observing an upswing in the number of tech jobs, and companies are making more hiring and recruiting announcements. This article enlists the top hiring, recruitment, and layoff announcements made by companies in India in 2020. January The year started on a positive note as HCL technologies announced its plans to double their campus recruitments to 15,000 by 2021. On a technical side, the company announced that they would hire engineers from IITs and NITs for salaries within the range of ₹ 12 to 15 lakhs and ₹ 8 to 12 lakhs per annum. The company also announced that from the beginning of this fiscal year, they are increasing their compensation for the technical and management employees by 15-20%. February In February, Cognizant announced its plan to hire over 20,000 digitally-skilled tech graduates who will receive a compensation of around ₹ 4 lakh per annum. The hiring was focused around strengthening its core areas — data, digital engineering, cloud, and IoT. Capgemini also announced plans on hiring 15,000 graduates at ₹ 3.8 lakh per annum. The advancements came after massive structural changes both the companies made in terms of layoffs and reskilling initiatives last year. April As Coronavirus took over, the news about companies laying off employees soon became rampant. Swiggy declared that it is cutting 1000 jobs from its cloud kitchen division as a response to the pandemic. Bigger companies such as Infosys, Accenture, and TCS made announcements to freeze hirings, considering the business uncertainties, but honour the candidates whom they had already rolled out an offer. In the same month, Paytm declared that it wants to hire 500 employees for multiple roles, apart from the continued hiring in product and technology teams. May The second month in the lockdown also did not do any favours. Ola announced that it is laying off 1400 employees. Uber followed suit, laying off 600 employees. Zomato also laid off 13% of its 4000-staff, and Swiggy let go 1100 of its employees. A small glimmer of hope amid the doom and gloom of the pandemic was the messenger app Hike. This Delhi-based unicorn announced 20 open positions for AI and ML engineers among other recruits, for its upcoming app feature Hikeland. July With IT companies laying off thousands of its employees, Wipro chairman, Rishad Premji announced at their AGM meeting that Wipro has not laid off any employees and does not plan on any retrenchment in the near future as well. The statement came at a time when industry analysts expressed concerns over possible layoffs. August As the world started to open up after the lockdown, a few hiring announcements followed. Paytm announced its plan to hire over 1000 engineers, data scientists and financial analysts in its team among other tech and non-tech roles, over the next few months. This 1000 was an addition to the 500 employees it had declared in April. The hiring announcements align with the company’s plan to expand into domains such as lending, insurance, wealth management and offline payments. Uber also announced that it is hiring 140 engineers for its tech teams in Hyderabad and Bangalore. The engineers were hired to build products in areas such as rider and driver growth, delivery, customer service, among other areas according to an Uber statement. September In September, Salesforce announced its plan to hire 12,000 employees globally, including Indian workforce. The cloud computing firm saw an uptick in the business amid the pandemic, as all enterprises moved their data to the cloud. At the end of August, an internal Accenture report cited by the Australian Financial Review said that the company plans to lay off 5% of its employees, which is around 10,000 in India considering the country’s total number of employees. In September, the company declared a severance package of seven months for its layoffs. The package was made applicable for employees voluntarily tendering resignations. October While Accenture’s report said it wants to fire people, it also posted more than 3300 job vacancies in India, in October. Of these jobs, most were tech-based hirings with around 300 jobs for core tech skills. In the same month, however, OLX declared that they are letting go 250 of its employees as it shifts away from specific verticals and makes some internal alignments to refocus its strategy. November The ride-hailing platform, Ola announced its plan to set up a technology centre in Pune for which it would hire over 1000 engineers over the next few years. This will be a second such centre declared after the first one was already established in Bangalore. December In the last month of 2020, Fiat Chrysler said that it plans to invest $150 million to set up a new Global Digital Hub in Hyderabad, their largest digital hub outside of North America and EMEA. The company said that this is likely to create 1,000 cutting edge technology jobs by the end of 2021. Ernst & Young also announced that it would be hiring 9,000 professionals in India next year. The company said that it is looking to hire candidates from STEM backgrounds and skilled in artificial intelligence, machine learning, cybersecurity, analytics, and other emerging technologies. The company thinks it is vital to make such investments given the exponential increase in digital adoption.","excerpt":"The year 2020 was unpredictable in terms of hiring and recruiting for tech jobs. It started with many tech companies declaring an increase in mass recruitment. However, as the pandemic struck and the lockdown followed, there was an obvious decline in the number of tech jobs, and there were multiple layoffs by tech firms.  However, […]","categories":["AI Hirings"],"tags":["Techgig"],"author_name":"Kashyap Raibagi","publish_date":"2020-12-28T16:00:00","publication_year":"2020","word_count":936,"keywords":["Go","machine learning","artificial intelligence","AI","cloud computing","ML","Git","Techgig","ViT","analytics","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","analytics","cloud computing","R","Go","Git","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-hiring\/major-hiring-and-recruiting-announcements-made-by-tech-giants-in-2020\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":11297,"title":"Start-up of the week: Instalocate- A chatbot that claims to make your travel more comfortable!","content":"Img Source: Instalocate | www.instalocate.com Did you know that every time your flight gets delayed your airlines owes you a compensation? Have you ever been denied boarding because the flight was overbooked? Are you aware of your rights as a flyer? Many a times we overlook on these issues and incur heavy losses, but not anymore. The one company founded by Stanford University and Indian Institute of Management (IIM) alumni in June 2016, is building an AI powered travel assistant just for you! Instalocate– the name as it goes by – promises to watch all that for you by building a cutting-edge technology that can solve all your travel problems and make your journey comfortable.  No more panicking and rushing to the airline counters, standing in long queues or calling the customer care if your flight gets delayed or baggages do not come on time!  Instalocate promises to constantly monitor your travel and predict and solve the travel problems. Not just that, it would also protect your rights as a customer and go after airlines to get your due compensation in case of any mishap. How wonderful is that? Having a digital personal assistant that can make your journey comfortable and be always there to answer all your questions in an instant! Talking to AIM, one of its founders Pallavi Singh revealed that the idea of Instalocate was conceived out of all the unfortunate incidences that she and her husband had personally faced. [quote]“Anything that can go wrong has gone wrong with us. Flights have gotten delayed, we have missed connections, baggage was lost. And that’s when we realised that, most of the travel apps are working in pre-booking and there is no one to help you when things like this go wrong. Dealing with the airlines was the biggest nightmare amidst this”, she said.[\/quote] And that’s how the journey to Instalocate took off with an idea of building an assistant which could help during the travel woes and deals with the airline on your behalf. Pallavi confesses “At so many times, we felt so frustrated with the airlines that we wanted to sue them for compensation, for all the trouble we went through. But we never did- mainly because we never had the time to deal with the airlines.” With Instalocate, all you have to do is share your flight details and it will predict when you might need something and would send the contextual information automatically. Just ask your assistant anything from your flight status to the free Wi-Fi availability in the airport! That’s not all, if your family is worried about you, the assistant can pinpoint your exact location in the air. They don’t have to anxiously wait outside the airport checking their phones again and again! After reaching your destination, your cab will be waiting for you. How is all of it achieved? Talking about the integration of artificial intelligence to Instalocate, Pallavi said “It is a predictive engine which will predict when the airlines owe you compensation. Unlike others we don’t wait for you to search for that information rather we will bring it to you. We are also building in-house NLP which makes it easier for an end user to talk to us, just as they would talk to a friend.” There is no doubt that the bot has been received well by its users. “We have only launched our first product and the people are loving it”, marked Pallavi. Citing a use case, she said “One of our power users recently got 800 dollars from British Airways for flight delay with the help of Instalocate.” However, the journey to its popularity was not easy. Pallavi notes that making was not as challenging as marketing. “Bots is still a new concept for people and popularizing it is a big problem”, she added. Well, despite the challenges, Instalocate has done quite well for itself and is growing at a rate of 60 month over month with a pretty high retention rate. This digital personal assistant is available to make your journey comfortable and answer your questions in an instant. Talk to Instalocate within facebook at m.me\/instalocate for a hassle-free travel now. There is no need to install the app separately, which adds to the many perks this travel bot has!","excerpt":"Did you know that every time your flight gets delayed your airlines owes you a compensation? Have you ever been denied boarding because the flight was overbooked? Are you aware of your rights as a flyer? Many a times we overlook on these issues and incur heavy losses, but not anymore. The one company founded […]","categories":["Deep Tech"],"tags":["chatbot india"],"author_name":"Srishti Deoras","publish_date":"2016-11-22T05:28:57","publication_year":"2016","word_count":715,"keywords":["Go","artificial intelligence","ELT","programming_languages:R","AI","Git","NLP","Aim","Rust","R","chatbot india"],"extracted_tech_keywords":["AI","artificial intelligence","NLP","Aim","R","Go","Rust","Git","ELT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/start-week-instalocate-chatbot-claims-make-travel-comfortable\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10100818,"title":"Raspberry Pi 5 is Here","content":"In a surprise move, the highly anticipated Raspberry Pi 5 has made its debut, defying initial doubts. The latest iteration of the microcomputer, boasting significant enhancements, is now available, starting at an enticing $60. Notably, the Raspberry Pi 5 not only promises improved performance over its predecessor but also marks the first time the device incorporates in-house silicon. At the core of the Raspberry Pi 5 is a robust 64-bit quad-core Arm Cortex-A76 processor clocked at 2.4GHz, offering a substantial two to three-fold increase in performance compared to the aging Raspberry Pi 4, which was introduced four years ago. Additionally, the device houses an 800MHz VideoCore VII graphics chip, a feature lauded by the Raspberry Pi Foundation for its remarkable graphics performance boost. The device reveals speedy boot times and swift webpage loading, particularly when compared to older models such as the Raspberry Pi 3 Model B+. Notably, the device tends to generate significant heat, but Raspberry Pi has addressed this concern by supplying an active cooling component for direct board mounting. Key features include: 2.4GHz quad-core 64-bit Arm Cortex-A76 CPU VideoCore VII GPU, supporting OpenGL ES 3.1, Vulkan 1.2 Dual 4Kp60 HDMI® display output 4Kp60 HEVC decoder Dual-band 802.11ac Wi-Fi® Bluetooth 5.0 \/ Bluetooth Low Energy (BLE) High-speed microSD card interface with SDR104 mode support 2 × USB 3.0 ports, supporting simultaneous 5Gbps operation 2 × USB 2.0 ports Gigabit Ethernet, with PoE+ support (requires separate PoE+ HAT, coming soon) 2 × 4-lane MIPI camera\/display transceivers PCIe 2.0 x1 interface for fast peripherals Raspberry Pi standard 40-pin GPIO header Real-time clock Power button A noteworthy highlight of the Raspberry Pi 5 is the inclusion of a southbridge component, a critical part of the motherboard responsible for peripheral communication. Developed by the Raspberry Pi Foundation and dubbed the RP1 southbridge, this component introduces a substantial leap in peripheral performance and functionality. This enhancement translates to faster data transfer speeds to external UAS drives and other peripherals. Furthermore, the Raspberry Pi 5 introduces two four-lane 1.5Gbps MIPI transceivers for connecting up to two cameras or displays, along with a single-lane PCI Express 2.0 interface, which, though available for the first time, requires a separate adapter like an M.2 HAT for full utilization. In terms of connectivity, the Raspberry Pi 5 offers dual 4Kp60 HDMI display outputs with HDR support, a microSD slot, two USB 3.0 ports, two USB 2.0 ports, Gigabit Ethernet, and a 5V DC power connection via USB-C. Additional features include Bluetooth 5.0 and Bluetooth Low Energy (LE) support, and peak SD card performance that is claimed to be “doubled” with the SDR104 high-speed mode. These enhancements solidify the Raspberry Pi 5 as a versatile choice for various applications, whether it’s serving as an ultra-budget desktop PC, a media server, or a DIY security system. The Raspberry Pi 5 offers multiple RAM configurations at launch, with the 4GB version priced at $60 and the 8GB version at $80. While this places it slightly above the Raspberry Pi 4 in terms of cost, which is priced at $55 for 4GB of RAM and $75 for 8GB, it still maintains an attractive price point. Interested buyers can expect the Raspberry Pi 5 to be available for purchase before the end of October. Read: Raspberry Pi: A timeline","excerpt":"At the core of the Raspberry Pi 5 is a robust 64-bit quad-core Arm Cortex-A76 processor clocked at 2.4GHz.","categories":["AI News"],"tags":["AI hardware"],"author_name":"Mohit Pandey","publish_date":"2023-09-28T16:08:33","publication_year":"2023","word_count":546,"keywords":["Go","TPU","programming_languages:R","AI","programming_languages:Go","Aim","ViT","AI hardware","R"],"extracted_tech_keywords":["AI","Aim","TPU","R","Go","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/raspberry-pi-5-is-here\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10103175,"title":"Now You Can Buy Hyundai Cars on Amazon","content":"Amazon recently announced that its customers can now buy cars online in the US, starting next year in close collaboration with Hyundai. This collaboration aims to enhance the vehicle buying experience, integrate Alexa into upcoming Hyundai models, and revolutionise customer interactions and business operations through Amazon Web Services (AWS). Some exciting news for customers: https:\/\/t.co\/Zxxvz11v5W will be launching online vehicle sales in the U.S. next year, starting with @Hyundai. Will be part of our strategic partnership with them to help make the buying experience easier, add the Alexa experience into their…— Andy Jassy (@ajassy) November 16, 2023 Now auto dealers will be able to sell vehicles through Amazon’s U.S. store in 2024, with Hyundai being the inaugural brand available for purchase. This digital shopping experience allows customers to browse, select, and purchase cars online, providing the flexibility to either pick up the vehicle or have it delivered by their local dealership. The process involves searching for vehicles on Amazon based on preferences, selecting a car, and completing the transaction within the familiar Amazon interface. This multiyear agreement involves migrating on-premises applications to AWS, enabling Hyundai to adopt a cloud-first technology strategy. Prioritising areas like manufacturing, supply chain optimisation, security, disaster recovery, and connected vehicle development, Hyundai aims to leverage AWS capabilities, including AI and the IoT. Furthermore, Hyundai’s next-generation vehicles, starting from 2025, will integrate Alexa for a more responsive and interactive driving experience like playing music, setting reminders, updating to-do lists, and managing smart home devices while on the road. After being quiet for a while in the space of generative AI, Amazon’s Olympus is expected to enhance features across Amazon platforms, including the online retail store, Alexa voice assistant, and AWS. Olympus aims to outperform its predecessor, “Titan,” which faced delays last year due to performance issues compared to OpenAI’s ChatGPT. The official announcement by AWS is expected in December. Read more: [Exclusive] AWS’ Generative AI Play for Bedrock","excerpt":"This multiyear agreement involves migrating on-premises applications to AWS, enabling Hyundai to adopt a cloud-first technology strategy.","categories":["AI News"],"tags":[],"author_name":"Shritama Saha","publish_date":"2023-11-17T11:41:53","publication_year":"2023","word_count":321,"keywords":["ChatGPT","OpenAI","AI","AWS","Git","RAG","GPT","Aim","generative AI","R"],"extracted_tech_keywords":["AI","generative AI","ChatGPT","OpenAI","Aim","RAG","AWS","R","Git","GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/now-you-can-buy-hyundai-cars-on-amazon\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":13431,"title":"In Review: Nikhil Analytics Certificate Program in Analytics Data Science &#038; Machine Learning","content":"This brand new one-year Certificate Program in Analytics Data Science & Machine Learning is the industry’s answer to the demand of fulfilling a skill gap in data science. Offered by leading analytics solutions provider Nikhil Analytics and tailored in conjunction with Singapore’s leading Nanyang Technological University (NTU), the one-year intensive course lays down a solid foundation in machine learning, natural language processing, data mining techniques and regression and prediction. If you are looking for an advanced course that packs academic rigour, best-in-class faculty and relevant case studies with large datasets to play with, this one-year certification program is a good fit.  Nikhil Analytics is not a newbie on the data and analytics training block – the 5-year old analytics training and consulting has been offering short-term courses in data visualization tool Tableau, Python, R and Advanced Business Analytics Power Pack among others. Building on their 5+ years of expertise in data science and collaborating with Asia’s 2nd ranked university, Nikhil Analytics has rolled out a robust course that combines two in-demand learning tracks– data science and machine learning. Hard Facts Duration: 1 year Mode of learning: classroom weekend sessions (online sessions are available for alumni only) Hours: 400 hours International Tour: NTU, Singapore Availability: Bangalore Course offered: Analytics Data Science & Machine Learning Credentials: Certificate program Batch size: 25 Start date: The first batch started in February 2017 and weekend classes are underway Prerequisites: 2+ years work experience; basic knowledge of statistics such as grounding in Median Standard, Correlation and Linear Regression & Deviation; GRE\/GMAT\/CAT or Nikhil Analytics Test (NAT) scores and an engineering or MBA background, MSc. Statistics\/Economics Tools acquired: R, Python, SQL, Hadoop, Excel Resources: Classes are conducted in a state-of-the-art data science lab Fee: INR 2.5 lakhs Here are the key differentiators of this 1-year certification program in Analytics Data Science & Machine Learning from Nikhil Analytics First-of-its-kind onsite learning experience at Singapore’s NTU University. Each classroom session is followed by workshops across three key domains – BFSI, healthcare and retail. The classes are led by best academic minds such as Dr Jagath C Rajapakse, Dr Erik Cambria and Dr Kwoh Chee Kwong. Speaking about the collaboration between NTU and Nikhil Analytics, Dyuti Lal, co-founder and CEO of the company shared, “This is a first for the industry where students will travel and benefit from specialized knowledge with this university-wide collaboration. The course was launched in partnership with NTU with the aim of bridging the knowledge gap and lending more value to students. The 5-day workshop comes at a nominal cost and is a core part of the course”. Industry-led course provides an inter-disciplinary experience and combines two major streams – data science and machine learning. Besides, the faculty led by co-founder Dyuti Lal has gained certain street cred for lending in-depth computational skills designed to meet industry demand. Successful track record of placing candidates in leading companies such as Accenture, Genpact, Flipkart, Amazon, German multinational Wurth Elektronik Industry immersion sessions will be led by senior executives from leading companies such as Accenture Course The major topics covered in Data Science are Data Mining, knowledge discovery, supervised and unsupervised machine learning techniques, understanding how to optimize machine learning algorithms. In regression and prediction, students will learn advance regression techniques such as Ridge Regression and Modification Logistic Regression and cover topics such as Regression Trees, Random Forest and Boosted Trees. Another core area tackled is Deep Learning wherein the course covers conventional neural network and autoencoders.  Taking forward the training in machine learning, the course will delve into unsupervised machine learning techniques such as cluster analysis and factor analysis. Also, as part of the onboarding process, the program entails 60-hours of preparatory session in statistics. Hands-on Learning Course combines classroom learning with real-world case studies backed by anonymized data sets. Most of the case studies are mapped to three major data-intensive domains — BFSI, Healthcare and Retail. Dynamic Pricing: Students are tasked with credit card analysis for a leading credit card company and assessing credit limit variability for customers, targeting the users with right promotions based on their transactional history. Predictive Analytics: Forecasting with predictive analytics, the future sales of a leading automobile company. Students are tasked with building models that will forecast the future sales and revenue and identify best prices and promotions. The Housing case study also works on similar lines, predicting future sales of the company five years down the line. Faculty The faculty comprises the best-in class with NTU academicians Dr Kwoh Chee Keong known for his extensive research work in data analysis and machine learning and its application in life science, engineering and medical field. Meanwhile, Dr Jagath Rajapakse research leans towards brain imaging and computation biology and lastly Dr Erik Cambria has had stints at HP Labs India and MUS Temasek Labs works with MIT Synthetic Intelligence Lab and Brain Sciences Foundation among others. At Nikhil Analytics, Lal leads the charge of teaching R, SAS and SQL. Among the in-house faculty, Alok Ranjan, COO, Co-founder, Nikhil Analytics leads the charge of teaching analytics, statistical modelling and machine learning using R, Python, Excel while Dyuti Lal teaches SQL. Other faculty members, hailing from the analytics industry give training on Excel. Placement assistance Although no training institute guarantees 100% placement, Nikhil Analytics has had a successful track record of placing candidates in top-notch MNCs through their a network. “Through our HR and job consultant network we receive tailored job requirements and we are able to place the students successfully. We also do resume sourcing and matching for our students with leading companies and startups. Also, we leverage our strong alumni network for placement opportunities,” she shared. They also guide students in resume building exercises and the alumni network plays a key role in helping students prepare for data scientist interviews. Last word Pros: The course presents a first-of-its-kind international learning experience to students at a nominal cost. Its rigorous academic module makes it a good fit for candidates who want to explore Data Analyst and Data Scientist roles. The international collaboration goes beyond course content creation and includes a 5-day intense classroom and workshop session that helps student’s deep dive into machine learning and data science techniques. Cons: No dedicated Capstone project that helps in bolstering the resume. At a minimum, candidates should have an engineering, business or a statistics background to start with. Since the course is intended for engineers, programmers, and MBA professionals, students from non-technical background cannot participate.","excerpt":"This brand new one-year Certificate Program in Analytics Data Science & Machine Learning is the industry’s answer to the demand of fulfilling a skill gap in data science. Offered by leading analytics solutions provider Nikhil Analytics and tailored in conjunction with Singapore’s leading Nanyang Technological University (NTU), the one-year intensive course lays down a solid […]","categories":["AI Trends"],"tags":["data analytics certificate"],"author_name":"Richa Bhatia","publish_date":"2017-03-14T05:05:59","publication_year":"2017","word_count":1077,"keywords":["data science","machine learning","AI","neural network","data analytics certificate","RAG","Python","Aim","deep learning","analytics","predictive analytics"],"extracted_tech_keywords":["AI","machine learning","deep learning","neural network","data science","analytics","Aim","RAG","predictive analytics","Python"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/review-nikhil-analytics-certificate-program-analytics-data-science-machine-learning\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10162846,"title":"Ola Announces Krutrim AI Lab for Frontier Research with ₹2,000 Crore Investment","content":"Ola chief Bhavish Aggarwal has announced Krutrim AI Lab and the launch of several open source AI models tailored to India’s unique linguistic and cultural landscape. This includes the launch of Krutrim 2, the startup’s second LLM consisting of 8 billion parameters. “While we’ve been working on AI for a year, today we’re releasing our work to the open source community and also publishing a bunch of technical reports,” announced Aggarwal, founder and CEO of Ola and Krutrim. Furthermore, the Krutrim AI Lab includes:Chitrarth 1: A Vision Language Model capable of interpreting images and documents. Dhwani 1: A Speech Language Model designed for tasks such as speech translation. Vyakhyarth 1: An Indic Embedding model optimised for applications like search and retrieval-augmented generation. Krutrim Translate 1: A text-to-text translation model facilitating seamless translation between languages. Recognising the absence of a global benchmark for Indic language performance, Krutrim AI Lab has also developed “BharatBench,” a comprehensive evaluation framework. The lab has also published several technical reports and papers to further the research community’s understanding of these models. In partnership with NVIDIA, the lab is set to deploy India’s first GB200 supercomputer by March, with plans to scale it into the nation’s largest supercomputer by the end of the year. This infrastructure will support the training and deployment of AI models, addressing challenges related to data scarcity and cultural context. The lab has committed an investment of ₹2,000 crore into Krutrim, with a pledge to increase this to ₹10,000 crore by next year. By open sourcing the models, Aggarwal said that he aims to foster collaboration within India’s AI community, accelerating the development of a world-class AI ecosystem. Last week, Krutrim also brought China’s DeepSeek models to its cloud infrastructure. India’s AI Mission Continues This announcement aligns with India’s broader AI ambitions. The IndiaAI Mission seeks to build a comprehensive ecosystem that fosters AI innovation by democratising computing access, enhancing data quality, and developing indigenous AI capabilities. A key component of this mission is the establishment of a common computing facility powered by approximately 18,693 GPUs, including high-end models like the NVIDIA H100 and H200. This facility aims to provide accessible computing power to startups, researchers, and academia at a fraction of global cost benchmarks. Union Minister Ashwini Vaishnaw has emphasised the importance of accessible computing power, stating that it is ‘the most important part of the mission.’ He also announced plans to develop 6-8 large language models by the Indian tech ecosystem, supported by this robust computing infrastructure. Safety and ethical deployment of AI models remain top priorities for the government. To this end, an AI Safety Institute is being established, adopting a techno-legal approach to ensure responsible AI development.","excerpt":"In partnership with NVIDIA, the lab is set to deploy India’s first GB200 supercomputer by March.","categories":["AI News"],"tags":["Blackwell","Krutrim AI Lab","NVIDIA","Ola"],"author_name":"Vandana Nair","publish_date":"2025-02-04T14:15:54","publication_year":"2025","word_count":448,"keywords":["Go","Ola","AI","innovation","ML","responsible AI","Aim","Blackwell","data quality","NVIDIA","AI safety","R","Krutrim AI Lab","startup"],"extracted_tech_keywords":["AI","ML","Aim","R","Go","data quality","AI safety","responsible AI","innovation","startup"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/ola-announces-krutrim-ai-lab-for-frontier-research-with-%e2%82%b92000-crore-investment\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10054918,"title":"DeepMind &#038; Mathematicians Use AI To Solve The Knot Problem","content":"“The great advances in mathematics have not been made by logic but by creative imagination.” –George Frederick James Temple DeepMind collaborated with two mathematicians – one from the University of Sydney and the other from the University of Oxford, who used AI to tackle two mathematical problems. One is the theory of knots, and the other is the study of symmetry. AI helped the researchers discover new patterns which can be investigated using conventional methods. This is considered a breakthrough because, although computer simulations and visualisations of knots and other objects have helped mathematicians recognise patterns and develop their intuition in the past, it is the first time that AI\/computer is being used to seek out patterns. Researchers believe that this could benefit other areas of maths that involve large datasets. How did the collaboration come about In 2019, mathematician Geordie Williamson met DeepMind’s Demi Hassabis at the University of Sydney in Australia. Their conversation soon steered to collaborating for maths-related research. Marc Lackenby of the University of Sydney and a colleague at Oxford, Andras Juhasz, soon joined the project. Both Lackenby and Juhasz are knot theorists. When the team first started on the project, they concentrated on identifying mathematical problems that could be solved using DeepMind’s existing technologies. Machine learning could help in detecting patterns (like relationships between two objects), and mathematicians would then work out the exact relationship between the objects by formulating a conjecture. They could follow that by writing proof that turns that statement into a certainty. Prof Williamson used DeepMind’s AI to prove an old conjecture about the Kazhdan-Lusztig polynomials, which was unsolved for 40 years. Lackenby and Juhasz took this a little further, which led them to discover a previously unknown connection between algebraic and geometric invariants of knots. This helped in establishing a new mathematical theorem. Credit: DeepMind Invariants in knot theory are used to address the problem of distinguishing one knot from another. They also help mathematicians in understanding the properties of knots and how that related to other mathematics branches. Knots help in understanding varied aspects of fields like quantum field theory and non-Euclidean geometry. In fact, the relationship between algebra, geometry, and quantum theory is a long-standing challenge. The team has trained an ML model to discover such a pattern. This revealed the ‘signature’ – it is a particular algebraic quantity that is directly related to the geometry of the knot. By using attribution techniques from ML, the researchers were able to discover a new quantity called the natural slope. It hints at an important structural aspect that has remained undiscovered till now. They were then able to prove the exact nature of the relationship. Credit: DeepMind Wrapping up The discovery of patterns and formulation of useful conjectures are the main pillars of mathematical progress. Mathematicians have often relied on data for this process – from the early hand-calculated prime tables that Gauss and others used (that eventually led to prime number theorems) to modern computer-generated data (such as Birch and Swinnerton-Dyer conjecture). With computers generating data and test conjectures, mathematicians now have a new understanding of problems that were previously inaccessible. While computational techniques have become useful in other parts of the mathematical process, AI systems have not yet established such space. Previously, systems have generated conjectures that have either contributed sufficiently to researchers that do not easily generalise to other mathematical areas or have demonstrated novel methods that have given mathematically valuable results. This new research proves that AI systems hold great potential for identifying and discovering patterns in mathematics. With this, DeepMind hopes that this research will help researchers look at AI as an important tool to solve challenges in pure maths. DeepMind has also released interactive notebooks to replicate results. Read the paper here.","excerpt":"It is the first time that AI\/computer is being used to seek out patterns. Researchers believe that this could benefit other areas of maths that involve large datasets.","categories":["AI Features"],"tags":["DeepMind","DeepMind AI"],"author_name":"Shraddha Goled","publish_date":"2021-12-07T11:00:00","publication_year":"2021","word_count":628,"keywords":["Replicate","machine learning","programming_languages:R","AI","ML","DeepMind AI","R","DeepMind"],"extracted_tech_keywords":["AI","machine learning","ML","R","Replicate","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/deepmind-mathematicians-use-ai-to-solve-the-knot-problem\/","complexity_score":2,"technical_depth":6,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10094618,"title":"Course5 Raises $53M, Targets Two Acquisitions","content":"Course5 Intelligence, a data analytics startup, announced the successful closure of its inaugural external funding round, securing $53 million (approximately Rs 437 crore). The funding was divided into two tranches, with an initial investment of $28 million from 360 One Asset Management Ltd (formerly IIFL Asset Management Limited) Tech Fund, followed by an additional $25 million led by Nuvama Crossover Series of funds, in collaboration with Carnelian Asset Advisors Pvt Ltd. According to a source familiar with the matter, Course5 Intelligence achieved a valuation of $250 million through this funding round. The company intends to utilise the newly acquired funds to drive its growth through mergers and acquisitions, while also allocating resources to its artificial intelligence labs for the development of AI-driven intellectual property. Ashwin Mittal, Course5’s Chairman and CEO, said, “We are seeking for two-three types of acquisitions. First, we are looking at companies that bring us either access to IP, access to capabilities or access to customers, or some combination of these three. In terms of size, we’re looking mostly at companies that are in the $10 to 20 million revenue size that have established a good solid business or growing. We are also going to look for some deep tech AI startups” Course5 Intelligence has allocated approximately $80 million from its own cash reserves and cash accruals for future M&A activities. The company is currently in discussions with five potential prospects and plans to complete two acquisitions within the next year. With a focus on leveraging AI to address market and supply chain challenges, Course5 serves an extensive client portfolio including Fortune 500 companies in sectors such as retail, technology, media, and entertainment. Last year they partnered with BITS Pilani and Woxsen University for AI-based research in Multimodal Deep Learning and Sentiment Analysis respectively.","excerpt":"Course5 Intelligence, a data analytics startup, announced the successful closure of its inaugural external funding round, securing $53 million (approximately Rs 437 crore). The funding was divided into two tranches, with an initial investment of $28 million from 360 One Asset Management Ltd (formerly IIFL Asset Management Limited) Tech Fund, followed by an additional $25 […]","categories":["AI News"],"tags":["Courses"],"author_name":"K L Krithika","publish_date":"2023-06-06T22:32:31","publication_year":"2023","word_count":297,"keywords":["Go","artificial intelligence","AI","sentiment analysis","RAG","deep learning","ViT","analytics","Courses","R","startup"],"extracted_tech_keywords":["AI","artificial intelligence","deep learning","analytics","RAG","sentiment analysis","R","Go","ViT","startup"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/course5-raises-250m-targets-two-acquisitions\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10081836,"title":"Google Sheets Gets an AI Upgrade","content":"Google revealed Simple ML, a new tool that will let users create artificial intelligence models in Google Sheets. Although the tool is now in beta, users can configure the free add-on through a panel on Google Sheets without any code. Google’s open-source AI tech team, TensorFlow, developed this model. “Anyone, even people without programming or ML expertise, can experiment and apply some of the power of machine learning to their data in Google Sheets with just a few clicks,” said engineers from the Google team that built this feature. “From small business owners, scientists, and students to business analysts at large corporations, anyone familiar with Google Sheets can make valuable predictions automatically.” How to use Simple ML? Users will create a Google Sheets spreadsheet with data points arranged in rows and columns to train AI models with Simple ML and then mention the task they want the latter to complete. The tool will automatically train one or more AI models to perform the given task. Two AI use cases support simple ML at launch. The first involves filling in blank spreadsheet fields, while the second is locating fields with inaccurate data. For example, the products that are currently offered in the retailer’s catalogue, together with their associated prices, may be listed in a spreadsheet made using Google Sheets. Simple ML can automatically add missing product prices to the spreadsheet. It creates an AI model that determines what data a blank spreadsheet field should contain by examining the data points already in the file. The second use case that Simple ML supports is identifying anomalous data points. The add-on can, for example, determine if a snippet of text has been accidentally added to a spreadsheet field that should contain a numerical value. Simple ML detects data anomalies by generating no fewer than 10 AI models that automatically evaluate the accuracy of the information in a spreadsheet. “After a few seconds, once the model has made a prediction, you can explore using the result to improve your business decisions, automate tasks, or make any of the seemingly endless applications that ML enables,” Google’s engineers said. Last year, Google incorporated a new model in its Google Sheets that automatically creates formulas depending on the rich context around a target cell. The model uses the data in the target cell’s neighbouring rows and columns and the header row as context. Simple ML also includes an array of features that allows users to test the quality of the AI models it creates and offers technical information about them. Additionally, Google’s cloud-based code editor, Colab, designed for machine learning and data science applications, is accessible to users through Simple ML. Google also provides the data-processing tool Connected Sheets, besides Simple ML. It gives users access to the Google Sheets interface for data analysis stored in Google’s BigQuery cloud data repository. The module allows users to manage up to a trillion spreadsheet rows without authoring SQL queries.","excerpt":"Developed by TensorFlow, Simple ML will be used to create artificial AI models with data points arranged in rows and columns assigning the task.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","Machine Learning"],"author_name":"Shritama Saha","publish_date":"2022-12-08T13:09:43","publication_year":"2022","word_count":490,"keywords":["data science","artificial intelligence","machine learning","AI","ML","Machine Learning","Colab","Ray","SQL","TensorFlow","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","data science","Ray","TensorFlow","Colab","R","SQL"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/google-sheets-gets-an-ai-upgrade\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10172424,"title":"Gartner Predicts that Over 40% of Agentic AI Projects Will Be Discontinued by 2027","content":"Even as enterprises are moving steadily towards adopting generative AI and agentic AI systems, Gartner’s latest report predicts that more than 40% of agentic AI projects will be discontinued by the end of 2027. This will be driven by rising expenses, vague business benefits, or insufficient risk management. In a poll conducted by Gartner in January 2025 with 3,412 webinar participants, 19% reported that their organisation had made substantial investments in agentic AI, 42% had made cautious investments, 8% had not invested at all, while the remaining 31% were either taking a wait-and-see stance or were uncertain. Many vendors are fueling the excitement through “agent washing,” which involves rebranding existing products like AI assistants, robotic process automation (RPA), and chatbots, despite lacking genuine agentic features. Gartner estimates that only around 130 of the thousands of agentic AI vendors possess legitimate capabilities. “Most agentic AI projects right now are early-stage experiments or proof of concepts that are mostly driven by hype and are often misapplied,” said Anushree Verma, senior director analyst at Gartner. “This can blind organisations to the real cost and complexity of deploying AI agents at scale, stalling projects from moving into production.” Major technology companies like Salesforce and Oracle have adopted AI agents, which are systems capable of independently achieving objectives and taking action. They are investing billions into this technology to improve profit margins and optimise expenses. Despite initial challenges, the rise of agentic AI marks a significant advance in AI capabilities and market potential, the report says. It will enhance resource efficiency, automate complex tasks, and foster new business innovations, exceeding the capabilities of basic automation bots and virtual assistants. Gartner forecasts that by 2028, 15% of daily work decisions will be made autonomously through agentic AI, up from 0% in 2024. Additionally, 33% of enterprise software applications are expected to feature agentic AI by 2028, compared to under 1% in 2024. “Most agentic AI propositions lack significant value or return on investment (ROI), as current models don’t have the maturity and agency to autonomously achieve complex business goals or follow nuanced instructions over time,” said Verma. “Many use cases positioned as agentic today don’t require agentic implementations.”","excerpt":"Most agentic AI projects right now are early-stage experiments or proof of concepts that are mostly driven by hype and are often misapplied,” said Anushree Verma, senior director analyst at Gartner.","categories":["AI News"],"tags":["agentic ai","AI products"],"author_name":"Smruthi Nadig","publish_date":"2025-06-26T11:36:21","publication_year":"2025","word_count":362,"keywords":["Go","agentic AI","AI assistants","AI","chatbots","virtual assistants","agentic ai","Git","generative AI","GAN","R","AI products"],"extracted_tech_keywords":["AI","generative AI","agentic AI","AI assistants","chatbots","virtual assistants","R","Go","Git","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/gartner-predicts-that-over-40-of-agentic-ai-projects-will-be-discontinued-by-2027\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10080913,"title":"Microsoft Azure Goes Beyond Cloud, Launches New Space SDK","content":"2022 has been a landmark year for space technology, especially in India. From a spacecraft manufacturing facility being built in Bengaluru, to a successful liftoff of ISRO’s Vyommitra—India has been a hotbed of technology advancements in the space sector. Indian startups have been working overtime to put India on the space map. For example, AgniKul, who created a 3D printed semi-cryo engine, and Skyroot, which launched Vikram-S, the first rocket launched by a private sector company in India. Big tech companies are also leaping on the space bandwagon, with the foremost among them being Microsoft. After a partnership with NASA earlier this year, the cloud service provider expressed their goals to bring space development to the masses. To this end, they have launched a software development kit with the goal of deploying on-orbit applications. Introducing Azure Orbital Space SDK The Azure Orbital Space SDK democratises access to space development by offering a secure hosting platform and application toolkit to developers, which they can then use to create, test, and deploy applications that will work on space hardware. The platform comprises two parts: the host platform, which runs on the spacecraft itself. This then connects to the Azure cloud platform, which makes up the other part of the puzzle. Microsoft has provided templates, sample code, and robust documentation to start uncovering this previously unexplored paradigm of development. For example, common workloads are made into templates to allow developers to deploy solutions faster. The kit also comes with a “virtual test harness”, allowing developers to test their applications before deploying them on the spacecraft. Applications Microsoft also specified some common satellite use cases that can be further optimised with the advent of artificial intelligence at the edge. For example, remote sensing satellites—which are commonly used for surveillance—have been limited by bandwidth concerns. However, deploying an AI solution aboard the satellite can allow it to prioritise useful images or share only insights to ground control. In another example, the cloud service provider also showed how certain communication satellites could be managed better with smarter algorithms. This can result in huge cost savings for communications service providers. In addition to the Orbital Space SDK, satellites can also use the cloud service provider’s Orbital cloud products to further optimise the flow of data to and from the satellites. Partnerships with Spacetech companies To create this SDK, they have also partnered with other space companies under the Azure Space Partner Community program. These partners provide important services to Azure. Primary among them is a company named Xplore, which will provide Microsoft with a bevy of data that will allow them to gain knowledge about on-orbit compute use-cases, downlink optimisation and more. They have also partnered with Loft Orbital, a company that offers modular solutions for putting payloads into orbit. Earlier this year, they conducted a test with the company where they launched a spacecraft and integrated it successfully with the Azure Orbital Ground station. Now, they are planning to launch a spacecraft known as YAM-6, which will become the testbed for the various applications that Azure can bring to the space sector. In addition to these companies, Azure has also partnered with other companies to handle the hardware behind these orbital services. Ball Aerospace is a systems integrator with a history of designing government satellite programmes and will help Azure demonstrate reconfigurable on-orbit processing technologies. Thales Alenia Space is another company that specialises in developing high-power computing solutions for deploying in space. Through the partnerships, they will launch a testbed that will launch to the International Space Station in late 2023. What’s next? In a statement at the launch of the Orbital service, Mark Russinovich, chief technology officer at Microsoft Azure said, “Essentially, we’re building a ‘ground station as a service’. When it comes to cloud and data processing, obviously the cloud is a key part of any solution that goes into leveraging satellites.” The Azure Orbital SDK is an integral step in opening up space development to the masses, as it gives developers an idea of what they can expect when they are deploying an application in space. Similar to how cloud service democratised access to large amounts of scalable resources, Azure aims to offer this service to open up access for space companies and enable the optimisation of space resources.","excerpt":"Azure has opened developer access to space with the Orbital Space SDK.","categories":["AI News"],"tags":["ISRO","Microsoft Azure","Space"],"author_name":"Anirudh VK","publish_date":"2022-11-28T17:37:16","publication_year":"2022","word_count":716,"keywords":["Go","ISRO","API","artificial intelligence","AI","R","Scala","RAG","Aim","Space","Microsoft Azure","Azure","startup"],"extracted_tech_keywords":["AI","artificial intelligence","Aim","RAG","Azure","R","Go","Scala","API","startup"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/microsoft-azure-is-going-beyond-the-clouds-with-their-new-space-sdk\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":68787,"title":"CodeGuru: Now Programmers Can Find Costly Code Using This ML Tool","content":"Amazon recently announced the general availability of a developer tool powered by machine learning that provides intelligent recommendations for improving code quality known as CodeGuru. This tool works by identifying an application’s most expensive lines of code and helps in improving the quality of code. Software development is a long and systematic process. Developers write code, review, and compile to deploy applications. After deploying, they measure the performance of the application and further use that data to improve the code. In the whole development process, if a developer somehow makes a mistake by writing an incorrect code, which is left unnoticed, all the hard work and effort can go in vain. This is the reason performing code reviews is essential, where the developers check the logic, syntax, and style of the code before the new code is added to existing applications code-base. However, it is often challenging to have enough experienced developers with enough free time to carry out code reviews, given the amount of code that gets written every day. Also, most of the experienced reviewers even miss problems before they impact customer-facing applications, resulting in bugs and other performance issues. Besides, there are several other issues that should be mitigated in order to monitor the performance of applications and make them run efficiently. This is the reason behind developing the CodeGuru tool. Behind CodeGuru Amazon CodeGuru is a new service for the developers that uses machine learning to automate both code reviews during application development as well as profiling of applications in production. Trained with millions of lines of code from open-source projects and companies’ internal code, CodeGuru is designed to assist users in automatic code review. The tool was first announced in Dec 2019 at the AWS re:Invent event. The CodeGuru tool has two main components, which are: – Code Reviewer: Amazon CodeGuru Reviewer is a service that utilises machine learning techniques as well as program analysis in order to detect potential defects that are hard for the developers to find and recommends fixes in the Java code. Using Amazon CodeGuru Reviewer, you can do the following tasks: – CodeGuru Reviewer can find and flag code defects and suggest recommendations to fix those defectsIt provides actionable recommendations with low false positives and improves its ability to analyse code over time based on user feedbackIt provides recommendations as pull request comments Application Profiler: Using machine learning algorithms, CodeGuru Profiler can assist you in finding the most expensive lines of code and recommend the ways by which you can easily improve the efficiency of the code and remove CPU bottlenecks. It provides different visualisations of profiling data to help you identify what code is running on the CPU, see how much time is consumed, and suggest ways to reduce CPU utilisation. Using CodeGuru Profiler, you can do the following tasks: – Troubleshoot latency and CPU utilisation issues in your applicationLearn where you could reduce the infrastructure costs of running your applicationsIdentify applications’ performance issues. Benefits of CodeGuru Discover Where Your Application Is Costing You Money: Amazon CodeGuru Profiler can be used by the developers to search potential savings for any application running in production. Besides, you can optimise the performance using Profiler for any application running on AWS Lambda, Amazon EC2, Amazon ECS, AWS Fargate, or AWS Elastic Beanstalk, and on-premise. The CodeGuru Profiler also provides visualisations and recommendations on how to fix performance issues and the estimated cost of running inefficient code, helping developers prioritise remediation. Find Costly Code Issues Before They Hit Production: Amazon CodeGuru Reviewer uses machine learning, best practices and other learned lessons across millions of code reviews and thousands of applications profiled on open-source projects and internally at Amazon. The Reviewer analyses existing code bases in a repository to identify hard to find bugs and critical issues with high accuracy. After analysing, it provides intelligent suggestions on how to remediate them and creates a baseline for successive code reviews. Wrapping Up Using CodeGuru, developers can easily search bugs, get recommendations to fix as well as improve the codes. Besides, one can also identify various problems such as resource leaks, wasted CPU cycles, and potential concurrency race conditions. With Amazon CodeGuru, it is now easy to run on every code review and application in an organisation. The tool can be enabled with a few clicks in the AWS console, and customers only need to pay for their actual use of Amazon CodeGuru. Click here to start using Amazon CodeGuru. (Banner Image Source: here)","excerpt":"Amazon recently announced the general availability of a developer tool powered by machine learning that provides intelligent recommendations for improving code quality known as CodeGuru. This tool works by identifying an application’s most expensive lines of code and helps in improving the quality of code. Software development is a long and systematic process. Developers write […]","categories":["Deep Tech"],"tags":["Amazon AWS","AWS","AWS services","data structure using java"],"author_name":"Ambika Choudhury","publish_date":"2020-07-01T18:00:00","publication_year":"2020","word_count":748,"keywords":["Go","AWS services","machine learning","AWS","AI","cloud_platforms:AWS","programming_languages:R","data structure using java","programming_languages:Java","Amazon AWS","GAN","R","Java"],"extracted_tech_keywords":["AI","machine learning","AWS","R","Go","Java","GAN","cloud_platforms:AWS","programming_languages:R","programming_languages:Java"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/codeguru-now-programmers-can-find-costly-code-using-this-ml-tool\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10068061,"title":"Supercomputer ‘PARAM Ananta’ commissioned at IIT Gandhinagar","content":"‘PARAM Ananta’, a supercomputing facility under the National Supercomputing Mission (NSM), has been commissioned at the Indian Institute of Technology Gandhinagar (IITGN). The 838 teraflops (TF) Supercomputing System is powered with 7400 CPU cores, 52TB RAM, 1PiB of storage space, and an enormous power of GPU nodes along with the usual software stack required to run high-end simulations. “As the facility is being commissioned at IITGN, we are very confident that its computing power will be used by the faculty, researchers, students of IITGN and other research institutions in the vicinity to come up with new applications in diverse areas of research for the benefit of society. The technical prowess we will be able to achieve with such facilities will not only push the ‘Make in India’ and ‘Atmanirbhar Bharat’ but also ‘Designing and Nurturing in India’, which is also the spirit of the National Supercomputing Mission,” said Sunita Verma, Group Coordinator and Scientist G, MeitY.Ananta literally means ‘endless’ or ‘limitless’. It will provide a tremendous amount of computational power to the scientific community of IITGN as well as the academic institutions and industries in the vicinity of the Institute to take their Research and Development (R&D) efforts to a global scale. It will also further accelerate and strengthen the Institute’s collaborative R&D activities.","excerpt":"The 838 teraflops Supercomputing System is powered with 7400 CPU cores, 52TB RAM and 1PiB of storage space.","categories":["AI News"],"tags":["Atmanirbhar Bharat","IIT"],"author_name":"Zinnia Banerjee","publish_date":"2022-05-30T18:04:13","publication_year":"2022","word_count":214,"keywords":["programming_languages:R","AI","RAG","ViT","GAN","IIT","R","Atmanirbhar Bharat"],"extracted_tech_keywords":["AI","RAG","R","GAN","ViT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/supercomputer-param-ananta-commissioned-at-iit-gandhinagar\/","complexity_score":2,"technical_depth":6,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10131783,"title":"Why Does Everyone Want Tech Jobs to Fail?","content":"Recently, there has been a noticeable surge in the discussions about a potential collapse of the tech job market. The collapse of the stock market has not helped the case either. It almost seems as if people are eagerly waiting for the bubble to burst, hoping tech jobs will lose their lustre. In a Reddit discussion, many spoke about the possible reasons for this jealousy, hatred, or whatever it is that people have against tech jobs. An obvious one that came up was schadenfreude – the pleasure derived from another’s misfortune. The tech industry has long been associated with high salaries, luxurious perks, and an enviable work-life balance. During the pandemic, a user pointed out, people flaunted their remote, six-figure, salaried positions online while the rest of them had to work laborious jobs wearing masks. This created a sense of resentment among those in less fortunate positions, leading to a desire to see tech professionals brought down a peg. This disconnect between tech workers and the average person has fostered a sense of “finally, they’re getting what they deserve” from those who feel tech has been overhyped. AI is Now Killing Everyone’s Job? Some are not happy with the fact that AI is killing a lot of creative jobs, or at least being touted to be doing so. The tech overlords have stated several times that there will be no need for creative jobs in the future. It first came for the writers and artists, and followed it up by wiping out a lot of clerical jobs. People started hoping the same for the software engineers. Cognition Labs’ Devin tried to do so, and everyone was happy that the people who created AI are the ones who are going to get replaced by it. But the truth of the matter is that none of the jobs are getting automated completely. The rapid growth and significant influence of the tech industry have also contributed to the backlash. “I find a lot of tech workers constantly flaunt their very high tech incomes that they have in their 20s and 30s ($250k-1m). They make many times the median income, and have a totally different reality than the rest of the population.” The claims like “tech is dead” and “software engineering is dying” by the AI experts themselves have been fuelling the fire. Even the creators of coding tools such as GitHub Copilot and others have said several times that coding would become redundant as everyone would be a coder. Some root for a downturn in tech jobs in the hope to see a more balanced job market emerge. The dominance of tech in the job market has led to significant wage disparities and a concentration of wealth and opportunities in IT hubs. A correction in the tech job market could, in theory, lead to more opportunities across different sectors. But AI is here to stay. Every Job Would be a ‘Tech’ Job It’s hard to ignore how AI has revolutionised almost every aspect of our lives, from communication and healthcare to entertainment and education. And now AI is creating the age of abundance, with every job getting easier, and in some sense becoming a “tech job”. Generative AI has made even creative jobs easier for the artists, albeit they seem a lot less creative. With techies giving prompts to AI art generators such as Midjourney, making videos on Runway, and writing poetries on ChatGPT, the artists have started to criticise the need for AI, or even the tech field, as it is making art less artsy. But AI will create more creative jobs than there are today. While the tech industry does pose challenges, it also presents opportunities for reskilling and upskilling the workforce as Indian IT is doing it with its current employees and training them with generative AI. As automation and AI continue to evolve, there’s a growing anxiety about the future of work and whether there will be enough opportunities for everyone. This uncertainty can breed a desire to see the tech sector stumble, in hopes that it might slow down the pace of change and offer some stability to other industries.","excerpt":"The claims like “coding is dead” and “software engineering jobs are dying” by the AI experts themselves have been fuelling the fire.","categories":["AI Features"],"tags":["AI Impacts","AI in Jobs","ai in recruitment"],"author_name":"Mohit Pandey","publish_date":"2024-08-07T18:30:00","publication_year":"2024","word_count":692,"keywords":["ai in recruitment","Go","ChatGPT","API","AI","Git","RAG","Aim","AI in Jobs","generative AI","AI Impacts","GitHub","R"],"extracted_tech_keywords":["AI","generative AI","ChatGPT","Aim","RAG","R","Go","Git","GitHub","API"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/why-does-everyone-want-tech-jobs-to-fail\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10005280,"title":"What Is The Hiring Process For Data Scientists At ThoughtWorks?","content":"While data is the lifeblood of the modern digital business and enables leading enterprises to gain the insights needed to outpace rivals, this is also a world of technological dead ends and broken promises. ThoughtWorks believes that much of today’s legacy infrastructure makes it harder to turn data into intelligence, and one such challenge is finding the right data science candidate. “Data science practitioners are in high demand in today’s market. This makes the availability of the right talent suited to our specific business needs, limited,” says Savita Hortikar, head of recruitment at ThoughtWorks. Interview Process At ThoughtWorks ThoughtWorks believes in using a healthy mix of employee referrals, inbound and outbound applications while hiring for data science roles. “We also leverage recruiting platforms that use AI and ML tech to source candidates,” said Hortikar. For data practice — which is centralised and non-hierarchical  — they look for skills such as math, statistics, machine learning, deep learning, natural language processing, operational research and product experience. “We also look for a passion for data science and analytics alongside a willingness to learn and explore the data space because we believe skills can be taught but passion cannot,” she said. She further added that ThoughtWorks’ Data Practice extends beyond the work they do for customers and therefore look for a passion for technology to help change the world with software. In terms of educational background, they look for people with knowledge or experience in math, statistics, computer science — with degrees ranging from Bachelors in Engineering, MCA or Masters and PhD. “Having said that, ThoughtWorks always has and will continue to hire based on skill and passion for technology, which means we value skill over educational degrees,” said Hortikar. The steps involved are: A technical phone interview follows a recruiter’s phone screening round.As a next step, the candidate is given a remote assignment. This is followed by another technical interview that delves into the assignment in detail.The third technical interview discusses the candidate’s technical breadth and depth depending on their experience level.This is followed by the cultural fitment round and an interview with the function heads. “We look at the candidate’s relevant experience when hiring for this practice. We enquire if the candidate has a strong hands-on project experience coupled with a strong conceptual understanding of math and statistics. Also important is the candidate’s technical grasp of concepts like Machine Learning, Deep Learning, Natural Language Processing etc. We also appreciate varied experience across multiple domains,” said Apoorv Singhal, senior recruiter at ThoughtWorks. ThoughtWorks is currently hiring Data Scientists and Data Engineers for their data practice. Interested candidates can apply through ThoughtWorks career page, Linkedin and Glassdoor jobs posting page. Roles And Responsibilities Of Data Scientist At ThoughtWorks At ThoughtWorks, data science practitioner has the opportunity to: Understand the business challenges and goals of a client from various domainsFormulate data analysis and create a model to meet their needs and help in decision making.Create advanced analytics models using statistical and machine learning methods.Work with software developers and solutions designers to deliver analytics-driven solutions.Represent the company in various online and offline forums such as events, conferences, meetups and more.To be able to work in a dynamic, collaborative, transparent, non-hierarchical, and ego-free culture where talent is valued over a role title.Focus on passion rather than a predetermined one-size-fits-all plan. On a concluding note, Singhal shared that we make sure that we have properly defining attributes required for the particular job role. “We spend a lot of time on hands-on assessment, and we exhaustively list out the attributes needed for a job. We also invest time to understand candidates’ career aspirations when we match them to data science job profiles,” he said while signing off.","excerpt":"While data is the lifeblood of the modern digital business and enables leading enterprises to gain the insights needed to outpace rivals, this is also a world of technological dead ends and broken promises. ThoughtWorks believes that much of today’s legacy infrastructure makes it harder to turn data into intelligence, and one such challenge is […]","categories":["AI Hirings"],"tags":["Data Science Hiring"],"author_name":"Srishti Deoras","publish_date":"2020-08-21T13:00:42","publication_year":"2020","word_count":618,"keywords":["data science","Go","machine learning","AI","ML","Git","RAG","Data Science Hiring","deep learning","analytics","R"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","data science","analytics","RAG","R","Go","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-hiring\/what-is-the-hiring-process-for-data-scientists-at-thoughtworks\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10103000,"title":"GitHub Positions AI as Pivotal in Software Development Journey","content":"GitHub foresees the pivotal role of AI in the software development lifecycle. This perspective aligns with the growing demand for AI-driven tools that enhance efficiency and reduce software development time. GitHub’s belief underscores the industry’s shift towards harnessing AI capabilities, emphasising a synergy between developers and advanced algorithms to create innovation and streamline the software development journey. According to data from GitHub, 92% of developers report using AI coding tools in and outside of work. Moreover, around 81% of developers believe AI coding tools will make their teams more collaborative. “In March, we shared our vision of a new future of software development with Copilot X, where AI infuses every step of the developer lifecycle. Since then, we’ve been working to scale and mature the underlying technology–and in the process, we created something even bigger. Our vision has manifested itself into a new reality for the world’s developers,” CEO Thomas Dohmke said during GitHub Universe 2023. (CEO Thomas Dohmke at GitHub Universe 2023) At GitHub Universe 2023, the company introduced a new system called Copilot Workspace. This feature extends beyond the foundational AI pair-programming capabilities of the current GitHub Copilot. It leverages OpenAI’s GPT-4 and a comprehension of the developer’s pertinent codebase to aid in transforming an idea into a plan and, ultimately, into operational code. “The Copilot Workspace represents the evolution of our vision, guiding ideas from conception to production throughout the software development lifecycle. From the initial contemplation of a project to its realisation, Copilot Workspace serves as your AI assistant. “As you articulate your thoughts, whether it’s about creating a snake game or any other concept, Copilot seamlessly integrates into the software development journey, aiding you at every step of the process,” Inbal Shani, chief product officer at GitHub, told AIM. Going big with Copilot GitHub continues to make great strides with Copilot. So far, over a million developers use Copilot in over 37,000 enterprises across 190 countries making it the most widely adopted AI developer tool in history. “Just as GitHub was founded on Git, today we are re-founded on Copilot,” Dohmke said. In a bid to capture enterprise interest, at the event, GitHub unveiled Copilot for Enterprise. This integration allows developers to seamlessly connect Copilot with their codebase, facilitating intelligent suggestions not only for public repositories but also extending its capabilities to all internal and private code. “What we’ve seen when we started rolling out Copilot was that many customers wanted to have a customised model, one that understands their code base, their policies, regulations, the syntax,” Shani said. ( Inbal Shani, chief product officer of GitHub during GitHub Universe 2023) At the conference, AIM spoke to a few of the enterprises that have adopted Copilot. A financial service giant, piloted with Copilot earlier in May, and has locked in on Copilot after testing Amazon CodeWhisperer as well as Google Duet. For them, Copilot made sense because they were already Microsoft Azure customers. For a company that is not on Azure, factors such as ease of integration, collaboration features, and compatibility with existing workflows would be crucial in determining which tool to adopt. Maggie Hewitt, senior site reliability engineering manager, Duolingo (US), said that they have witnessed a 25% productivity increase with Copilot. Similarly, at AMD, the customised Copilot model provided accurate, high-quality AI suggestions tuned to their specific style of product design. “Working with GitHub on Copilot model fine-tuning, we enabled support for widely used hardware design languages such as Verilog, which is not possible with the commercial Copilot solution,” Alexander Androncik, senior director of software development, AMD said. Moreover, at Universe 2023, GitHub also announced Copilot Chat, which will generally be available in December 2023 as part of the existing GitHub Copilot subscription. Leveraging your code as context, Copilot Chat elucidates intricate concepts, proposes code relevant to your open files and windows, identifies security vulnerabilities, and aids in locating and rectifying errors in code, terminal, and debugger. Leveraging AI to make GitHub more secure At Universe 2023, GitHub also announced that its Advanced Security will feature AI-powered tools to uncover and mitigate vulnerabilities and sensitive data in code, for upgraded application security. Nearly, 80 percent of the breaches originate through credential leakage or secrets being leaked, according to Jacob Depriest, VP, Deputy Chief Security Officer at GitHub. “Secret scanning has been integral to GitHub’s advanced security, forming a key component of our security programme. “Historically, the focus has been on tokens with structured capabilities, like Azure or AWS tokens. Now, with AI, GitHub aims to detect generic secrets and patterns with low confidence in code. This enhancement is poised to significantly improve the platform’s capability, allowing for the identification and protection of secrets before they reach production or get embedded in the codebase,” he told AIM. Moreover, AI is also being integrated to the code scanning autofix features, which are a part of GitHub’s Advanced Security. Code scanning autofix suggests AI-generated fixes using CodeQL for JavaScript and TypeScript directly in pull requests, assisting developers in resolving issues faster, and reducing the introduction of new vulnerabilities into codebases. “This enhancement aims to make the process even more impactful before code hits production, significantly speeding up the developer workflow. Our core philosophy for the security program is to seamlessly integrate security capabilities into developers’ work environments, avoiding disruptions to their workflow,” Depriest said. The Models Given Microsoft owns GitHub, the cloud-based Git repository provider leverages OpenAI’s LLM models. For example, the recently introduced chat functionality is powered by GPT-4. Similarly, GitHub Copilot is also powered by OpenAI’s Codex. Nevertheless, Depriest explained that GitHub utilises a combination of models tailored to its needs, recognising that diverse models serve distinct requirements. “Our emphasis is on selecting the right model for specific tasks. Across the GitHub platform, you’ll notice varied capabilities, from autocomplete and chat to AI-driven features in GitHub Advanced Security. Our approach is dynamic, and these models will likely evolve with time.” Shani concurs. She said GitHub is not locked in on a specific model just because of its collaboration with Microsoft. “If you inquire about my perspective as an individual who came of age in the realm of AI when it was a niche, I anticipate a future where we inhabit a hybrid environment featuring a prevalence of tailored models as opposed to larger models.”","excerpt":"At GitHub Universe 2023, the company introduced a new system called Copilot Workspace. This feature extends beyond the foundational AI pair-programming capabilities of the current GitHub Copilot.","categories":["AI Trends"],"tags":["GitHub","Github Copilot","GitHub Enterprise"],"author_name":"Pritam Bordoloi","publish_date":"2023-11-14T16:43:12","publication_year":"2023","word_count":1050,"keywords":["GitHub Enterprise","OpenAI","AI","AWS","R","ML","Github Copilot","TypeScript","RAG","Aim","JavaScript","GitHub","Azure"],"extracted_tech_keywords":["AI","ML","OpenAI","Aim","RAG","AWS","Azure","R","JavaScript","TypeScript"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/github-positions-ai-as-pivotal-in-software-development-journey\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":31314,"title":"Can Julia Be The New Python? Here’s What You Need To Know","content":"Python is undoubtedly the most popular language among data scientist and machine learning professionals. But Julia, founded by Viral B Shah, Deepak Vinchhi, Alan Edelman, Jeff Bezanson, Stefan Karpinski and Keno Fischer, is now gaining popularity in the field. This year in August, Julia developers announced the 1.0 release of their project, which means that the language is no longer at a ‘developer’s stage’ and is now an ‘expert’. Before moving on to their comparison, it is important to indicate and establish that it is not fair to expect that Julia can beat Python hands down. Python has been in the market since ages and its success stories are everywhere. Julia, on the other hand, is quite new and does not compete with Python in many areas. This article will only emphasise on in what ways both languages are different so that it helps you to decide whether or not to begin to learn Julia, in case you haven’t. 1. Speed: This is one area for which Julia is most popular for. Julia is faster than Python because it is designed to quickly implement the math concepts like linear algebra and matrix representations. It is excellent for numerical computing. Its multiple dispatches is great for defining data types like numbers and arrays. For codes that are equally big and complex written in both the languages, Julia takes lesser time, at speeds of the same order of magnitude of C or Fortran, compared to Python. Winner: Julia 2. Community support: Python is older and more popular than Julia and has a greater community support. If you are stuck with a problem, you have many options online where the Python online community can help you resolve, very quickly. It is easy to get support. Julia lags here. But using Julia’s Disclosure you could ask questions to resolve or even contribute. But Python would still win here. Python enjoys its huge community. Winner: Python Output readability: Model output is very easy to read in Julia compared to Python where it is not so readable. Winner: Julia. Installing packages: Installing new packages is easy in Julia than in Python. To run the official Julia model on your device, you have to download and install the package from their official Github site. It pulls the package directly through Real-Eval-Print Loop (REPL). This makes installing new packages much easier. In Python, installing packages isn’t necessarily always a difficult task. It can also be very simple, but they also have very complicated and difficult procedures to install some packages. Winner: Julia Libraries: Python has plenty of libraries to make work easy for every specific task. It is definitely easier to work with a good number of the library. Julia does not have libraries as many as Python and lacks that ease that comes along with a rich set of libraries. Winner: Python Code conversion: It is very easy to convert code from Python and\/or C to Julia. But the other way around is not an easy path. Converting code from Python to C or C to Python is very difficult. But Julia can interface with external libraries very easily written in C and Fortran.  Data can be shared easily with Python using the PyCall library. Winner: Julia. Working with shell: Julia is extremely well-integrated with the shell.  Like, shell commands to check the content of the file. The variables in Julia are exported to the shell as an environment variable. Users can also edit the file once opened. Basically, working with shell commands is easy in Julia than in Python. Winner: Julia. Conclusion As stated earlier in this article, it is not really fair to compare Julia and Python, as they are not at the same level currently. It can be said that Julia beats Python over its weaknesses but it cannot yet beat Python in its strengths. Currently, it cannot replace Python as a general scripting language. But Julia is fast pacing with its developments and may sometime in the future be able to give a tough fight to Python. Which one to choose among the two depends on your requirement. If your project is much into mathematics, Julia definitely shines there. It has great support for functional programming.","excerpt":"Python is undoubtedly the most popular language among data scientist and machine learning professionals. But Julia, founded by Viral B Shah, Deepak Vinchhi, Alan Edelman, Jeff Bezanson, Stefan Karpinski and Keno Fischer, is now gaining popularity in the field. This year in August, Julia developers announced the 1.0 release of their project, which means that […]","categories":["AI Features"],"tags":["Julia Language"],"author_name":"Disha Misal","publish_date":"2018-12-10T14:27:59","publication_year":"2018","word_count":703,"keywords":["Go","machine learning","TPU","AI","Git","Python","Julia Language","Ray","Julia","GitHub","R"],"extracted_tech_keywords":["AI","machine learning","Ray","TPU","Python","R","Go","Julia","Git","GitHub"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/can-julia-be-the-new-python-heres-what-you-need-to-know\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":69774,"title":"Reliance Group’s IoT business Unlimit partners with Cumulocity to launch Enablement, an IoT solutions Platform","content":"Unlimit, powered by Reliance Now, there’s a platform out there for businesses to connect any device over any network, and manage and store asset data. This has been made possible with the creation of Enablement, the platform dedicated to IoT solutions. Enterprises will be able to control devices in real-time by monitoring availability, to administer and rectify device anomalies. Juergen Hase, CEO, Unlimit comments, “By 2020, the number of connected devices in India is projected to rise from 200 million today to 3 billion. IoT is sure to positively impact almost every segment of the economy.” Enablement has been made as a part of a strategic partnership between global IoT software provider Cumulocity and Reliance Group’s Internet of Things (IoT) business unit Unlimit, to roll out a suite of new products and services. The comprehensive offering will allow Unlimit’s users to benefit from rapid machine, sensor, and device integration. Additionally, they will be able to leverage data collection and real-time analytics techniques for condition monitoring. This new platform offering will showcase off-the-shelf applications for important industry verticals such as healthcare, financial services, industrial automation, and asset tracking. The product can be used across the world, or can be tailored to meet the specific demands of the Indian market. The platform and its integrated suite of customized products furnish the tools needed to smoothly embed IoT into a business, for any customer or organization, across several specialist vertical markets. Moreover, Enablement allows clients to rapidly exploit the benefits, while it continually evolves to meet their demands and desires. “With Enablement, Indian consumers will be able to use a mobile phone with the capability to swiftly build, secure, and deploy their own state-of-the-art IoT solution in an extremely cost-efficient way,” explains Bernd Gross, Chief Executive Officer, Cumulocity. Unlimit was launched only in November last year by Reliance Group in partnership with Cisco-owned Jasper, as their dedicated business unit to furnish IoT solutions to Indian and International customers. Since the inception of the organization in this space, “Enablement” has been touted to be Unlimit’s next big project to drive IoT-based ecosystem within the country.","excerpt":"Now, there’s a platform out there for businesses to connect any device over any network, and manage and store asset data. This has been made possible with the creation of Enablement, the platform dedicated to IoT solutions. Enterprises will be able to control devices in real-time by monitoring availability, to administer and rectify device anomalies. […]","categories":["AI News"],"tags":["big data processing interview","connected devices","healthcare India","partnership India","sensor"],"author_name":"Дарья","publish_date":"2017-04-11T10:56:23","publication_year":"2017","word_count":351,"keywords":["API","programming_languages:R","AI","connected devices","big data processing interview","RAG","sensor","automation","analytics","healthcare India","real-time analytics","GAN","partnership India","R"],"extracted_tech_keywords":["AI","analytics","RAG","R","API","real-time analytics","GAN","automation","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/reliance-groups-iot-business-unlimit-partners-cumulocity-launch-enablement-iot-solutions-platform\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10005010,"title":"A Fake Blog Created By College Kid Using GPT-3 Made People Believe Human Wrote It","content":"In another interesting news around GPT-3, a college kid created an entirely fake blog under a fake name using the AI model. We have extensively seen in the past how GPT-3 is the new latest obsession in the ML community. Liam Porr, a computer science student at the University of California, Berkeley did this as a fun experiment and said that the scariest part of it was that it is super easy. The blogs by Porr reached the number-one spot on Hacker News with readers engaging on the post and some even hitting ‘Subscribe’. While some people noticed that the blog was AI-generated, it highly went unnoticed. It suggested that the content produced by GPT-3 could fool people into believing that a human wrote it. To create a fake blog with GPT-3, Porr submitted an application filling out a simple questionnaire about the intended use. He collaborated with a PhD student to run a script. As reported, it gave GPT-3 the headline and introduction for a blog post and had it spit out several completed versions. The post went viral with over 20,00 visitors, and only a few asked him if the text was AI-generated. Some of the blog topics it wrote are —  “Feeling unproductive? Maybe you should stop overthinking”, “Boldness and creativity trumps intelligence”, among others. While on a few occasions the headline did not work out, but as long as he stayed on the right topics, the process was easy. GPT-3 has been a debated technology with even Sam Altman suggesting that GPT-3 has weaknesses and makes silly mistakes. It has been feared to replace jobs such as journalists, coders, lawyers, accountants and more in the coming future.","excerpt":"In another interesting news around GPT-3, a college kid created an entirely fake blog under a fake name using the AI model. We have extensively seen in the past how GPT-3 is the new latest obsession in the ML community.  Liam Porr, a computer science student at the University of California, Berkeley did this as […]","categories":["AI News"],"tags":[],"author_name":"Srishti Deoras","publish_date":"2020-08-17T14:31:03","publication_year":"2020","word_count":281,"keywords":["programming_languages:R","AI","ML","GPT","ViT","R","llm_models:GPT"],"extracted_tech_keywords":["AI","ML","R","GPT","ViT","llm_models:GPT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/a-fake-blog-created-by-college-kid-using-gpt-3-made-people-believe-human-wrote-it\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10088601,"title":"Google USM Shatters Language Barriers with Multilingual Speech Recognition Model","content":"A team of researchers at Google have published a research paper ‘Google USM: Scaling Automatic Speech Recognition’ that introduces the Universal Speech Model (USM) – a single large model that performs automatic speech recognition (ASR) in more than 100 languages. The model’s encoder is pre-trained on a vast unlabeled multilingual dataset of 12 million hours that covers over 300 languages and fine-tuning on a smaller labelled dataset. Multilingual pre-training with random-projection quantization and speech-text modality matching is used to achieve state-of-the-art performance on downstream multilingual ASR and speech-to-text translation tasks. The team also demonstrates that despite using a labelled training set one-seventh the size of that used for the Whisper model, their model exhibits comparable or better performance on both in-domain and out-of-domain speech recognition tasks across many languages. Training Process The training process involves three different types of datasets to achieve their goal. The process of training is divided into three stages. In the initial phase, a conformer backbone is trained using a vast speech dataset that has not been labelled. The objective is to optimize the BEST-RQ. In the next phase, the speech representation learning model is trained further, aiming to optimize several objectives such as BEST-RQ on unlabeled speech, modality matching, supervised ASR, duration modelling losses on paired speech, and transcript data, as well as the text reconstruction objective with an RNN-T decoder on unlabeled text. The final phase involves fine-tuning the encoder that was pre-trained in the earlier stages, for the ASR or AST tasks. Firstly, they use unpaired audio datasets, including the YT-NTL-U, which is a massive collection of over 12 million hours of audio content sourced from YouTube, in over 300 different languages, without labels. The second unpaired audio dataset is Pub-U, which contains over 429,000 hours of speech content in 51 languages, sourced from public datasets, without labels. Secondly, the team uses an unpaired text dataset called Web-NTL, which contains more than 28 billion sentences in over 1,140 different languages. Lastly, the team utilizes two paired ASR data corpora, each containing over 10,000 hours of audio content with matching text for supervised training. The first corpus is YT-SUP+, which includes 90,000 hours of labelled data from 73 languages and an additional 100,000 hours of pseudo-labelled en-US data generated using the Noisy Student Training (NST) technique from the YT-NTL-U dataset. The second corpus is Pub-S, which includes 10,000 hours of labelled data from multi-domain en-US public sources, and an additional 10,000 hours of labelled data from public sources in over 102 languages. The process for building 2B-parameter Conformer models using various datasets involves the following steps: The encoder of the model is pre-trained using BEST-RQ, which is based on BERT and incorporates Randomprojection Quantizer. This unsupervised pre-training is performed using YT-NTL-U. The model can be further prepared through a multi-objective supervised pre-training pipeline called MOST, which utilizes a combination of three datasets: YT-NTL-U, Pub-U, Web-NTL, and Pub-S. During this process, the BEST-RQ masked language model loss is combined with text-injection losses, including supervised ASR loss and modality matching losses. These losses are optimized through a weighted sum during training. To prepare the model for downstream tasks, generic ASR models are trained using connectionist temporal classification (CTC) and Listen, Attend, and Spell (LAS) transducers in supervised ASR training. Results The team’s USM models have achieved excellent results in multilingual ASR and AST for various datasets in different domains. These include SpeechStew (ASR for a single language), CORAAL (ASR for African American Vernacular English), FLEURS (multilingual ASR), YT (long-form ASR for multiple languages), and CoVoST (AST from English to multiple languages). We also developed an ASR model for YouTube captions that performs better than Whisper, a general ASR system trained on more transcribed data, in 18 selected languages. They also found that BEST-RQ pre-training is an effective method for scaling speech representation learning to large datasets. When combined with text injection in MOST, it improves the quality of downstream speech tasks, achieving state-of-the-art performance on the FLEURS and CoVoST 2 benchmarks. The MOST representations can adapt quickly to new domains by training lightweight residual adapter modules, which add only 2% more parameters while keeping the rest of the model frozen. To improve the performance of ASR models on long-form speech inputs, the team also introduced chunk-wise attention, a scalable method that produces high-quality transcripts for long utterances in the YouTube evaluation sets when used with USM-CTC\/LAS models. Meanwhile, last year, NVIDIA’s NeMo framework was used to develop a Telugu automatic speech recognition (ASR) model, which won IIIT-Hyderabad and Telugu ASR Challenge competitions with a low word error rate. The toolkit is open-source and supports multi-GPU and multi-node cluster training. The team preprocessed the data, removed errors, and trained the model for 160 epochs using NeMo, while fine-tuning pre-trained models using an NVIDIA DGX system for the open track. The results showed superiority over other ASR frameworks, achieving almost 2% fewer word errors than the runner-up.","excerpt":"The model’s encoder is pre-trained on a vast unlabeled multilingual dataset of 12 million hours that covers over 300 languages.","categories":["AI News"],"tags":["Google","Microsoft","NVIDIA"],"author_name":"Shritama Saha","publish_date":"2023-03-03T12:11:59","publication_year":"2023","word_count":819,"keywords":["Go","programming_languages:R","AI","Modal","Scala","BERT","Aim","llm_models:BERT","Google","RNN","NVIDIA","R","Microsoft"],"extracted_tech_keywords":["AI","Aim","R","Go","Scala","BERT","RNN","Modal","llm_models:BERT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/google-usm-shatters-language-barriers-with-multilingual-speech-recognition-model\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":28349,"title":"Microsoft Acquires San Francisco-Based Startup Lobe To Boost AI Development","content":"Satya Nadella led Microsoft has been clear about its AI ambitions from the start and this is yet another proof of the tech giant’s commitment to strengthening its AI capabilities. In yet another high profile acquisition, the Redmond giant snapped up San Francisco-based AI startup Lobe, that is lowering the entry to DL by enabling developers build models without any code. Lobe, currently available in Beta is an easy-to-use visual tool that allows developers to build custom deep learning models, quickly train them, and export it to app without writing any code. Start by dragging in a folder of training examples from your desktop. Lobe automatically builds you a custom deep learning model and begins training. According to news reports, no financial details have been revealed yet. Microsoft blog announcement indicated that Lobe acquisition will bring more AI development capability. This year, Microsoft acquired Semantic Machines in May brought in a revolutionary new approach to conversational AI,  and the acquisition of Bonsai in July will help reduce the barriers to AI development through the Bonsai team’s unique work combining machine teaching, reinforcement learning and simulation. “These are just two recent examples of investments we have made to help us accelerate the current state of AI development,” Kevin Scott – Executive Vice President and Chief Technology Officer, Microsoft noted shared in the post. Lobe ties in very well with Microsoft’s recent acquisitions which were done to boost AI development and also lower the entry barrier. Scott added how we are only at the beginning of the DL revolution since AI development and building deep learning models are slow and complex processes even for experienced data scientists and developers. “To date, many people have been at a disadvantage when it comes to accessing AI, and we’re committed to changing that,” he shared in the post. Founded in founded in 2015 by Mike Matas, Adam Menges, and Markus Beissinger in order to make deep learning accessible to everyone, Lobe will continue to work as a standalone service, supporting open source standards and multiple platforms. Microsoft acquisition will allow the startup to access world-class AI research, global infrastructure, and decades of experience building developer tools, Lobe website indicated.","excerpt":"Satya Nadella led Microsoft has been clear about its AI ambitions from the start and this is yet another proof of the tech giant’s commitment to strengthening its AI capabilities. In yet another high profile acquisition, the Redmond giant snapped up San Francisco-based AI startup Lobe, that is lowering the entry to DL by enabling […]","categories":["AI News"],"tags":["AI development"],"author_name":"Richa Bhatia","publish_date":"2018-09-14T11:27:45","publication_year":"2018","word_count":364,"keywords":["AI development","programming_languages:R","AI","RAG","deep learning","AI research","R","startup"],"extracted_tech_keywords":["AI","deep learning","RAG","R","startup","AI research","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/microsoft-acquires-san-francisco-based-startup-lobe-to-boost-ai-development\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10008846,"title":"Online Content Moderation: To AI or Not","content":"Last week, YouTube decided to bring more human moderators to vet content on the streaming platform. In an interview, YouTube’s Chief Product Officer Neal Mohan said that the lack of human oversight has caused machine-based moderators to take down a whopping 11 million videos, which broke none of the community guidelines. Mohan said, “Even 11m is a very, very small, tiny fraction of the overall videos on YouTube . . . but it was a larger number than in the past.” It must be noted that YouTube, in a lengthy blog post, in March this year, said that they would be deploying more machine and AI-based moderators for reviewing the content without any human intervention. This was done keeping in view the new work conditions in the pandemic. Twitter too, had announced a similar decision. So, which is better, human or AI-powered moderation? AI for Content Moderation Content moderation has become a very important practice for several digital and media platforms, social media websites, and e-commerce marketplaces to drive their growth. It involves removing content that contains irrelevant, obscene, illegal, or inappropriate matters deemed unsuitable for public viewing. AI is helping this moderation process in optimising it through algorithms to learn from existing data and making review decisions for the content. Majorly, AI-based moderation systems view content in two broad senses — content-based moderation and context-based moderation. Content-based moderation further includes both text and image\/video content moderation. Natural language processing is a preferred technique used for reviewing text content and can also be used for a speech by using speech-to-text techniques. Named entity recognition (NER) is an important NLP technique to recognise harmful text content such as terrorist propaganda, hate speech, harassment, and fake news. Further, sentiment analysis can be adopted for classifying and labelling portions of content in terms of the level of emotion. Computer vision technologies such as object detection and semantic segmentation can enable machines in analysing images to identify harmful objects and their locations. Optical character recognition (OCR) is further used to identify and transcribe text within images and videos. Context-based moderation is based on ‘reading between the lines’. AI learns from several sources to get the contextual understanding. This form of moderation is still in the process of development. Which One Bests The Other: AI or Human Several factors make the case in favour of AI moderators strong: Reportedly, every day, at least 2.5 quintillion bytes of data are created. And it is growing with each passing year. AI-powered moderation is a great choice, as compared to human moderators, given the humongous amount of content it can detect and analyse. Further, as opposed to major trauma and mental health issues such as post-traumatic stress disorder that human moderators may face as being exposed to hours of agonising content can easily be sorted by deploying AI models. Case in point ex-YouTube on-contract moderator who sued the company for pushing her to extreme mental trauma as a direct result of watching hours of harmful content. The cost of human moderation is quite high. It goes without saying that moderation does not really generate revenue for the company and is merely seen as a necessary evil that must be controlled, vetted to diminish toxic content under all costs so as to not drive the users away. Having said that, as in the case of YouTube, AI cannot mimic human capabilities and sensitivities in deciding which content is harmful and to what extent. AI is best at automating processes with straightforward datasets and defined characteristics. However, they fall short when it comes to more nuanced and subjective decision making. Human rights and free speech experts around the world are against fully automated content moderation as its over-bluntness is bound to erroneously infringe the right to create and circulate critical information. For example, in May 2020, YouTube admitted that its enforcement systems ‘mistakenly’ deleted comments critical of the Chinese Communist Party (CCP) Decision-making in moderation is a complex process, and for now, a hybrid human-AI moderation system is what looks like the best option.","excerpt":"Last week, YouTube decided to bring more human moderators to vet content on the streaming platform. In an interview, YouTube’s Chief Product Officer Neal Mohan said that the lack of human oversight has caused machine-based moderators to take down a whopping 11 million videos, which broke none of the community guidelines. Mohan said, “Even 11m […]","categories":["AI Trends"],"tags":["face recognition online","Facebook","Twitter (X)","YouTube"],"author_name":"Shraddha Goled","publish_date":"2020-10-05T11:00:33","publication_year":"2020","word_count":674,"keywords":["Go","AI","sentiment analysis","Git","computer vision","GAN","NLP","object detection","ViT","face recognition online","Facebook","YouTube","Twitter (X)","R"],"extracted_tech_keywords":["AI","NLP","computer vision","sentiment analysis","object detection","R","Go","Git","GAN","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/online-content-moderation-to-ai-or-not\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10147814,"title":"Indian IT Grooms Grads While GCCs Grab Pros","content":"Of late, Indian IT companies have been forging partnerships with engineering colleges and universities to train students in practical, job-ready skills. The sole purpose behind such collaborations is to address the growing demand for AI-skilled workforce. Infosys Foundation has invested ₹38 crore in a partnership with the Indian National Academy of Engineering (INAE) in New Delhi to upskill educators and provide internships and mentorship for students. “By working with INAE, we hope to empower educators with the right set of teaching skills and provide students with real-world learning experiences through internships and meaningful guidance from mentors that will collectively contribute towards boosting the employability of the engineering workforce in India,” said Sumit Virmani, trustee of Infosys Foundation. Similar to Tata Technologies, it has established innovation centres at private universities, offering hands-on training to bridge the gap between theoretical knowledge and industry needs. Similarly, Mumbai-based HSNC University has collaborated with TCS iON to design certification programs, internships, and placements, ensuring students are prepared for real-world challenges. At the same time, Sardar Patel Institute of Technology in Mumbai has partnered with IBM and Wipro for advanced training. Karnataka government’s initiative to pair 100 engineering colleges with corporates by 2025 aims to directly increase employment prospects. Such efforts underscore the IT sector and the government’s commitment to finding long-term solutions to prepare a pipeline of skilled professionals who will meet the industry’s future demands. GCC Poach Talent for Immediate Gains While IT firms focus on nurturing future talent through long-term investments in college partnerships, GCCs are poaching experienced professionals and tapping into niche skill sets directly. An example of this shift is Zeiss, a German optical technology firm that launched its first GCC in Bengaluru in November 2024. The centre focuses on cloud computing, cybersecurity, and software development and plans to double its Indian workforce to 5,000 within three years. “This centre will accelerate the digital transformation of our growing global operations,” said Dhaval Radia, chief financial officer, ZEISS India. GCCs are not just targeting traditional engineering graduates but are broadening their hiring strategies to include experts with advanced degrees. In an interview with AIM, Dhanya Rajeswaran, global VP and managing director at Fluence India, highlighted this trend: “Absolutely, we are hiring from IIT and NIT, but we are also looking for people with PhDs, MTechs, and specialisations in our kind of work.” Reddit users have also highlighted this GCC strategy. One commented, “Our GCC, for example, is taking business away from TCS\/HCL and replacing them or taking their staff and making them ours.” Another observed Google’s approach, noting the establishment of the Google Operations Center (GOC) between 2017 and 2019. Since then, professionals from companies like Cognizant and Accenture have transitioned to roles at Google’s GCC, showcasing the appeal of GCC roles. GCCs have seen exponential growth by leveraging career advancement opportunities and competitive hiring practices. From 5,000 global leaders in Indian GCCs in 2022, projections suggest this figure will surpass 30,000 by 2030, growing at a compound annual rate of 25.1%. This evolution signals a pivotal shift in India’s tech ecosystem as GCCs cement their dominance in the recruitment arena. “GCCs are connecting the dots,” said Rishi Kapoor, vice president and APAC lead – talent acquisition & staffing, Material. “The key will be to bridge the gap between the demand for skilled professionals in GCCs and the evolving expectations of Gen Z. Reshaping the employee value proposition (EVP) and aligning talent with core values and culture will be essential.” Competitive Advantage GCCs attract top talent from Indian IT companies by offering competitive compensation and international exposure. According to a recent TeamLease Digital report, GCCs pay 12-20% higher salaries than IT services for comparable tech roles driven by investments in AI, machine learning, and cybersecurity. GCCs offer significantly higher salaries, with compensation for software developers ranging from ₹9.7 lakh to ₹43 lakh annually, compared to ₹5.7 lakh to ₹17.9 lakh in IT services for similar roles. Entry-level salaries at GCCs are also up to 30% higher than the average across sectors. They emphasise innovation, particularly in AI, cloud computing, and cybersecurity, positioning themselves as leaders in tech transformation. In contrast, Indian IT companies focus on maintaining existing projects while lagging in generative AI and R&D investment due to high costs. As of now, there is a strategic tug-of-war between Indian IT firms focusing on long-term talent development and GCCs prioritising short-term skill acquisition to sustain their rapid growth.","excerpt":"While IT firms focus on nurturing future talent through long-term investments in college partnerships, GCCs are poaching experienced professionals and tapping into niche skill sets directly.","categories":["GCC"],"tags":["GCC","Indian IT"],"author_name":"Shalini Mondal","publish_date":"2024-12-26T15:15:00","publication_year":"2024","word_count":736,"keywords":["Go","GCC","machine learning","AI","cloud computing","ML","RAG","Aim","generative AI","Rust","Indian IT","R"],"extracted_tech_keywords":["AI","machine learning","ML","generative AI","Aim","RAG","cloud computing","R","Go","Rust"],"url":"https:\/\/analyticsindiamag.com\/gcc\/indian-it-grooms-grads-while-gccs-grab-pros\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10003067,"title":"Wipro To Launch 5G Edge Services Solutions Suite","content":"The solutions extend Wipro’s BoundaryLess Enterprise – Universal Edge portfolio to offer real-time visibility and data insights for holistic infrastructure management Wipro, a global player in information technology, consulting and business process services company, has announced that it will launch its 5G edge services solutions suite. The solutions suite to be provided by Wipro is built with IBM TRIRIGA and IBM Edge Application Manager. Wipro will engage with clients to implement the Universal Edge solutions suite that leverages 5G network capabilities. The 5G edge services solutions suite is designed to significantly enhance Wipro’s existing BoundaryLess Enterprise – Universal Edge portfolio. The edge-compute-enabled offering allows communications service providers and mobile tower operators to deploy their applications into dispersed edge devices. The advanced artificial intelligence and cloud-based services address the challenges of edge computing. It provides real-time visibility and data insights that help enable holistic management of edge infrastructure for mission critical applications in manufacturing, transport, healthcare, oil & gas and retail industries among others. This solution is designed to offer Wipro customers better data control, reduced costs, faster insights and actions, and more automated, secured operations. A key module of the solution, primarily for the telecom ecosystem, provides a complete application suite to enable a secured gig economy around 5G-related services. Partnership With IBM Wipro is joining the IBM Edge Ecosystem, an initiative to help partners implement open standards-based cloud native solutions that can be deployed and autonomously manage edge applications at large scale. Wipro’s solutions combined with IBM Edge Application Manager and TRIRIGA is expected to address a range of concerns related to deploying and managing globally distributed services on devices, private edges and telecom operator’s Multi Access Edges. The 5G edge services solutions suite integrates various edge computing solutions from IBM. The two companies recently announced a collaboration to develop hybrid cloud offerings to help businesses migrate, manage and transform mission-critical workloads across public or private cloud and on-premises IT environments. The recently launched Wipro IBM Novus Lounge has been designed to accelerate client innovation and bring such industry-ready solutions from public and private clouds to the edge. Evaristus Mainsah, General Manager, Cloud, Cloud Pak and Edge Ecosystem, IBM said, “The convergence of 5G and edge computing is set to spark new levels of innovation, and this in turn will fuel a broad ecosystem of providers to co-create for a growing set of edge opportunities. We are excited about the value that Wipro can bring to their clients that require Universal Edge solutions to help extend their enterprise to compute at the edge.” K.R. Sanjiv, Chief Technology Officer, Wipro Limited said, “Our strong telecommunication domain capabilities coupled with leadership in cloud & infrastructure lifecycle and investment in the Wipro IBM Novus Lounge equips us well to deliver these industry-specific solutions. We are confident that this will enable efficient deployment and management of infrastructure with 5G and Edge for our customers in the telecommunications, manufacturing, oil & gas and retail industries.” According to Wipro, it is committed to be the leading 5G solutions and implementations partner to its clients providing them with strategic advice on the technology and unlocking its potential to generate new revenue streams. Wipro’s deep engineering and product design expertise enable clients to achieve their 5G objectives in three primary areas: Engineering, Intelligence and Monetisation. In addition to the 5G lifecycle services, Wipro’s specialisations in automation, security and enterprise transformation help clients realise business value in their digital transformation journeys.","excerpt":"The solutions extend Wipro’s BoundaryLess Enterprise – Universal Edge portfolio to offer real-time visibility and data insights for holistic infrastructure management Wipro, a global player in information technology, consulting and business process services company, has announced that it will launch its 5G edge services solutions suite. The solutions suite to be provided by Wipro is […]","categories":["AI News"],"tags":["5G","cloud business intelligence solutions","Wipro"],"author_name":"Vishal Chawla","publish_date":"2020-07-23T09:59:35","publication_year":"2020","word_count":573,"keywords":["Wipro","cloud business intelligence solutions","5G","artificial intelligence","programming_languages:R","AI","innovation","digital transformation","Git","RAG","automation","edge computing","R"],"extracted_tech_keywords":["AI","artificial intelligence","RAG","edge computing","R","Git","digital transformation","automation","innovation","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/wipro-to-launch-5g-edge-services-solutions-suite\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10000846,"title":"5 IoT Security Attacks That Reminded Us Of Mirai Botnet","content":"Years ago when technology was just getting started to evolve, the devices that could access the Internet were only PCs and laptops. However, with time, new devices started to enter and today, almost every device is connected to the internet, leading to the development of the Internet of Things (IoT) ecosystem. Over the last few years, securing these devices has become a major task for manufactures as well as for users and many vulnerabilities have been discovered that led to serious attacks and hacks on different IoT devices. Curious Dose lists down 5 worst vulnerabilities and attacks on IoT devices The Mirai Botnet Attack In October 2016, a massive DDoS attack hit the internet and it took down a huge portion of the internet which includes platform like Twitter, Netflix, Reddit etc. Dyn, a company that controls much of the internet’s DNS infrastructure became the victim of this nasty attack. According to reports, there were 100,000 malicious endpoints involved in the attack and to the surprise, there were reports that the attack had tremendous strength of 1.2Tbps. The attack was carried out by IoT botnets with the help of malware called Mirai, which was created by Paras Jha and Josiah White. Mirai is a Japanese word which means “the future” and Jha released the code of the malware on the internet and later it was used to launch the DDoS attack. The interesting fact is that unlike other botnets, Mirai botnet largely consists of IoT devices. The Mirai malware first infects computers and make them scan big blocks of the internet for vulnerable IoT devices such as security cameras, digital cameras, DVR players etc. and logs in to them. St. Jude Medical’s Hackable Pacemaker In January 2017, a report confirmed that the pacemakers from St. Jude Medical are exposed to threats — the implantable cardiac devices have serious vulnerabilities in the transmitter, which is used to read the device’s data and remotely share it with physicians. Talking about the seriousness of the situation, the vulnerability could allow hackers to drain batteries or even manipulate the patient’s heartbeat that could risk lives. However, soon after the discovery of the faults in the devices, software patches were released to fix the issue. The devices were made by health tech firm Abbott and marketed under the St Jude Medical brand. Even though there were no reports about unauthorised access to any patients’ implanted device, it shows how serious it was. This event of vulnerable pacemakers goes well with the statement “if technology can save a life, it can take life too. Owlet’s Baby Heart Monitor Vulnerabilities Medical devices have become a common target for hackers and with major vulnerabilities in such devices, things are getting easier for attackers. After St. Jude Medical’s vulnerable pacemaker, another device is the Owlet WiFi baby heart monitor, an IoT device that babies wear in a sock. The device is used to send heartbeat data wirelessly to a nearby hub and also, parents can set alert to their smartphone when anything extraordinary is noticed. In 2016, it was discovered that the network which links the WiFi hub to the device is not encrypted (it doesn’t require any authentication to access), which makes it vulnerable to exploits. With such a vulnerability, it is easy for an attacker to hack the device within the range. Talking about the consequences, once getting access to the device, a hacker can prevent alerts from being sent out to parents or the nearby hub. An attack that Left Finland Cold Unlike other IoT based attacks, this attack is quite peculiar. In November 2016, notorious cybercriminals targeted computer system that controls the heating of two building in the city of Lappeenranta, Finland. The attack was a DDoS attack that primarily used IoT devices. The attackers directed massive traffic to heating controllers that later put the systems in a reboot loop which resulted in a situation where heating couldn’t kick in. In a place where usually the temperature goes below 0 degree Celsius by the end of every year, this kind of attack is serious and might lead to some critical medical emergencies. Jeep Carjacking In July 2015, two security researchers Charlie Miller and Chris Valasek hacked and took over a car. The entire event was a demo that showed how vulnerable the system was. The demo was carried out with the help of a volunteer victim who was driving a Jeep Cherokee at 70 mph when the security researcher hacked the vehicles CAN bus and exploited the firmware update vulnerability. The worst part is that when Miller and Valasek took over the car, they discovered something which was really serious — they could not only make the car go fast and slow, but they could also make the car veer off the road. This demo definitely shows how emerging Internet of Things (IoT) hacks are getting vital day by day and it is high time that companies need to come up with more secure systems if they want to deliver value to people without compromising their security, privacy, and lives. Way Forward Given the recent spate of events, it is completely clear that both the user and developer has to be concerned. From the users’ end, it is advised to do some serious research before buying an Internet-connected product. And from the developer’s end, it is imperative that the devices go through every possible test and make sure it is secure enough before pushing it out into the market.","excerpt":"Years ago when technology was just getting started to evolve, the devices that could access the Internet were only PCs and laptops. However, with time, new devices started to enter and today, almost every device is connected to the internet, leading to the development of the Internet of Things (IoT) ecosystem. Over the last few […]","categories":["AI News"],"tags":["cyber attacks","Cyber Security","IoT"],"author_name":"Harshajit Sarmah","publish_date":"2019-01-08T19:44:11","publication_year":"2019","word_count":914,"keywords":["Go","Cyber Security","programming_languages:R","AI","programming_languages:Go","Git","cyber attacks","ViT","R","IoT"],"extracted_tech_keywords":["AI","R","Go","Git","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/5-iot-security-attacks-that-reminded-us-of-mirai-botnet\/","complexity_score":3,"technical_depth":7,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10167479,"title":"Snowflake Extends Apache Iceberg Support to Strengthen Open Data and AI Capabilities","content":"Snowflake has announced expanded support for Apache Iceberg tables, aiming to give enterprises the flexibility of open formats along with its platform performance and data security. The move allows organisations to work with open-format data without compromising compute power or governance. “With Snowflake’s latest Iceberg tables innovations, customers can work with their open data exactly as they would with data stored in the Snowflake platform, all while removing complexity and preserving Snowflake’s enterprise-grade performance and security,” said Christian Kleinerman, EVP of Product at Snowflake. The new capabilities allow customers to manage, analyse, and share data stored in Iceberg format using the same features available to native Snowflake tables. Snowflake’s enhancements bring secure data sharing, built-in governance, and support for lakehouse analytics, including upcoming general availability for Search Optimisation Service and Query Acceleration Service. Snowflake also offers managed Iceberg tables, enabling customers to use open storage formats while maintaining performance. The platform now supports replication and syncing for Iceberg tables (in private preview), supporting continuity in case of outages or attacks. Thousands of organisations, including Illumina, Komodo Health, Medidata, and WHOOP, use Snowflake to activate open-format data for AI and analytics use cases. Snowflake’s secure sharing capabilities now extend to Iceberg, enabling the same data distribution and monetisation workflows used on native tables. Snowflake also highlighted its ongoing support for open source, noting that 35% of its acquisitions over the past four years involved companies focused on open data technologies. It actively contributes to several open source projects to enhance data interoperability and support enterprise AI workloads. These include Apache Iceberg, which provides a framework for open lakehouse management and data governance, and Apache NiFi, supported through Datavolo—an acquisition made in 2024—to enable real-time pipeline orchestration. Snowflake also backs Apache Polaris (Incubating), a project focused on ensuring cross-cloud interoperability for Iceberg. Additionally, Snowflake supports Modin, acquired in 2023, to help scale pandas-based data processing, and Streamlit, which enables the creation of data applications and interactive dashboards. With the 2024 acquisition of TruEra, Snowflake is also advancing AI explainability, bias detection, and regulatory compliance. “The future of data is open, but it also needs to be easy,” Kleinerman added. “Customers shouldn’t have to choose between open formats and best-in-class performance or business continuity.” By integrating Apache Iceberg into its broader AI Data Cloud strategy, Snowflake looks forward to accelerating the development of AI-powered applications, analytics, and insights without vendor lock-in.","excerpt":"Companies like Illumina, Komodo Health, Medidata, and WHOOP use Snowflake to activate open-format data for AI and analytics use cases.","categories":["AI News"],"tags":["Snowflake"],"author_name":"Siddharth Jindal","publish_date":"2025-04-09T11:01:02","publication_year":"2025","word_count":398,"keywords":["AI","R","ML","pipeline orchestration","RAG","Aim","Streamlit","analytics","Pandas","Snowflake"],"extracted_tech_keywords":["AI","ML","analytics","Aim","pipeline orchestration","Streamlit","Pandas","RAG","Snowflake","R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/snowflake-extends-apache-iceberg-support-to-strengthen-open-data-and-ai-capabilities\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10133630,"title":"Carrier&#8217;s Bold Moves in AI and Digital Transformation","content":"In 2019, Carrier established its first Digital Hub in Hyderabad, India, coinciding with its transition to become an independent, publicly traded company in 2020. Led by Senior Vice President and Chief Digital Officer Bobby George, Carrier embarked on a digital transformation journey to modernise and strengthen the digital infrastructure of a 100-year-old organisation. Since then, Carrier’s success with Digital Hub India has led to the development of the “Connected Hubs” model, integrating hubs in Mexico and China with operations in Bengaluru and Hyderabad. “Following our spin-off, where we prioritised goals, such as becoming more customer-centric, transforming our customer interactions, fostering innovation, and developing sustainable products, we recognized an internal talent deficit. This led us to make the strategic decision to establish a centre in Hyderabad, India,” shared Bobby George, SVP and CDO of Carrier, in a recent interview with AIM. “This move was pivotal in streamlining and accelerating digital deliveries worldwide, inspiring us to replicate the model across three countries, each with tailored focus areas.” More Than a Delivery Center: A Strategic Partner Today, Carrier’s India Hub is the company’s largest digital transformation centre, making up nearly 40% of Carrier’s digital talent from its two locations in Hyderabad and Bengaluru, respectively. Over the years, the hub has evolved into a strategic partner, collaborating with business units globally to develop innovative solutions and create significant impact. “There’s no secret that India boasts the largest concentration of digital and engineering talent. Currently, Carrier is investing in artificial intelligence and data sciences. These skills will shape our future, and our hubs in Hyderabad and Bengaluru will play a pivotal role,” Bobby elaborated. India: A Key Player in Carrier’s Global Strategy Carrier’s strategy of interconnected hubs consolidates operations across Mexico, China, and India under unified leadership, fostering specialised talent pools. The Mexico hub pioneers automation and digital solutions for future factories, while the China hub enhances digital capabilities for the APAC region. Meanwhile, the India hub functions as the digital twin to the global headquarters, operating across multiple domains. India’s strategic location and versatile capabilities are crucial for bridging time zones and enhancing global coverage. “Carrier Digital Hub India plays a pivotal role in our global strategy, accelerating our digital transformation journey. By leveraging India’s robust talent pool and strategic capabilities in cybersecurity, cloud operations, customer experience, IOT, and data science, we enhance operational agility and drive innovation across our enterprise,” emphasised Bobby. This approach enables Carrier to swiftly implement innovative solutions, ensuring efficient product deliveries and streamlined modernization projects. The India hub significantly contributes to Carrier’s flagship products: Abound, a cloud-native platform for Healthy & Efficient buildings, and Lynx, a cold chain platform leveraging advanced data analytics, IoT, and machine learning. Carrier’s Big Bet On AI Looking ahead, Carrier is set to establish a 100+ employee AI Center of Excellence in India, aiming to propel the company to the forefront of AI innovation and advancement. This strategic move underscores Carrier’s commitment to harnessing artificial intelligence to drive transformative change across its operations and offerings. Focus On Talent Innovation & Inclusion at Carrier’s Digital Hubs “At Carrier, we often say we are a 100-year-old company with the spirit of a startup. It’s essential that our employees have every opportunity to capture ideas, big or small, and continually innovate,” remarked Bobby George. Carrier’s global digital teams have implemented an idea crowdsourcing program called iD8. This platform allows employees to submit their ideas during planned idea collation sprints and internal hackathons. Subject matter experts review these ideas, and selected concepts advance to the proof-of-concept stage. Carrier’s culture of collaboration and respect has fueled its rapid growth in India, a trend poised to continue. The company is committed to hiring highly skilled deep tech talent in India. “Our strategic focus on acquiring top-tier talent is critical to advancing our digital transformation and innovation initiatives,” said Bobby George. “India’s extensive pool of tech professionals is pivotal to our global strategy. Furthermore, we are committed to recruiting fresh talent from colleges to infuse our teams with new perspectives and cutting-edge ideas, ensuring sustained innovation and growth.” Empowering Women in Tech at Carrier Carrier’s commitment to collaboration and equity extends to fostering an environment where women in tech can thrive. The AccelHerate program mentors women leaders, while the WomenUp program provides female employees with opportunities to interact with industry experts and gain valuable leadership insights. “We are dedicated to the growth of women in tech. Our programs for skill development and recruitment of women candidates are a testament to this commitment. In Digital Hub India, our diversity ratio has increased from 2% to 24% in recent years, highlighting the progress of our initiatives. We remain committed to further enhancing diversity and inclusion,” Bobby emphasised. All of this clearly portrays that Carrier’s strategic focus on AI and digital transformation, anchored by its robust India Hub, positions the company at the forefront of innovation. With a strong emphasis on talent acquisition, collaboration, and diversity, Carrier is set to drive sustainable growth, empowering its global operations to meet future challenges effectively. You too can be part of this growth, openings here: link","excerpt":"Carrier is set to establish a 100+ employee AI Center of Excellence in India, aiming to propel the company to the forefront of AI innovation and advancement.","categories":["AI Highlights"],"tags":[],"author_name":"Siddharth Jindal","publish_date":"2024-08-23T16:38:32","publication_year":"2024","word_count":849,"keywords":["data science","machine learning","artificial intelligence","AI","ML","RAG","Ray","Aim","analytics","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","data science","analytics","Aim","Ray","RAG","R"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/carriers-bold-moves-in-ai-and-digital-transformation\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10040457,"title":"Israel’s Iron Dome Puts AI At The Forefront Of Modern Warfare","content":"Palestinian militant groups fired over 2,000 missiles at Israel in the ongoing conflict. Israel’s sky was blazing with enemy missiles intercepted by counter missiles from its Iron Dome Aerial Defense System. The AI-powered anti-missile system is reportedly 90 percent effective. Developed by Rafael Advanced Defense Systems and Israel Aerospace Industries, the Iron Dome is designed to spot, counter and destroy rockets and 155mm artillery shells fired from 2.5 miles to 43 miles range. The Iron Dome was first tested in 2008. From drawing board to combat-ready, the development of the system took less than four years. In 2011, it deployed at air forces bases in Israel. Iron Dome consists of three core components — the Detection and Tracking Radar, the Battle Management and Weapon Control (BMC), and the Missile Firing Unit. Israeli software company mPrest Systems has built the control system of the Iron Dome. The Iron Dome (Source: Rafael Advanced Defense Systems) AI component The Iron Dome uses an AI function to determine if the incoming round of short-range rockets and missiles will land on the population or critical assets of the state. The AI system detects, analyses, and ultimately obstructs incoming threats such as Unmanned Aerial Vehicles (UAV), guided missiles, and cruise missiles. On detecting and identifying an external rocket, the Iron Dome radar tracks the missile path and, based on that information, the BMC calculates the potential point of impact. If it senses a credible threat, the AI system directs the dome to launch an interceptor or a faster missile against the danger, blowing off the enemy missile at a high altitude or on a neutral area. According to the Center for Global Affairs Strategic Studies, Israel has ten such Iron Dome batteries to protect the country’s infrastructure and citizens. Since its development and deployment, the Iron Dome has been upgraded to combat multiple and simultaneous attacks. According to Rafael Advanced Defense Systems and Israel Aerospace Industries, the Iron Dome’ works in any weather conditions, including rain, low clouds, dust storms or heavy fog. To date, the Iron Dome has intercepted more than 2,500 incoming targets. Image taken during a test campaign of the Iron Dome (Source: Rafael Advanced Defense Systems) Good or bad? According to the US Department of Defense, the radar can take inputs from different radars and be tied together with US missile defence systems such as the Patriot and THAAD. In 2017, India had signed a $2 billion deal with Israel to buy the Iron Dome system. Earlier last year, the US Army activated two new batteries of the Iron Dome system in Texas and New Mexico. Technology in warfare is a concept as old as the first trebuchet. Countries around the world are turning to technology to upgrade their artillery. The Pentagon is counting on AI as a major driver of defence modernisation. The US government has allocated the Pentagon around $841 million for building AI capabilities for 2021, of which more than one-third will go to the JAIC. The DoD has launched a program, called Tradewind, incorporating tech players, academia, and other stakeholders to develop new AI capabilities for the defence forces. Additionally, the Pentagon has started a $10 billion program to centralise data across departments over the cloud. In India, the Ministry of Defence (MoD) is looking to tap low latency communication provided by the 5G technology to power up AI and build UAV for the armed forces. The MoD has established iDEX – an ecosystem to promote defence and aerospace innovation and technology growth by bringing together innovators and entrepreneurs to provide technologically innovative solutions for the modernisation of the Indian military.","excerpt":"Palestinian militant groups fired over 2,000 missiles at Israel in the ongoing conflict. Israel’s sky was blazing with enemy missiles intercepted by counter missiles from its Iron Dome Aerial Defense System. The AI-powered anti-missile system is reportedly 90 percent effective. Developed by Rafael Advanced Defense Systems and Israel Aerospace Industries, the Iron Dome is designed […]","categories":["AI Features"],"tags":[],"author_name":"Debolina Biswas","publish_date":"2021-05-19T18:00:00","publication_year":"2021","word_count":603,"keywords":["Go","programming_languages:R","AI","innovation","programming_languages:Go","R"],"extracted_tech_keywords":["AI","R","Go","innovation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/israels-iron-dome-puts-ai-at-the-forefront-of-modern-warfare\/","complexity_score":3,"technical_depth":6,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10054879,"title":"Council Post: The Great Attrition In Data Science — What Do Experts Say?","content":"“The Great Resignation” is picking up again – attrition in the Indian data science space is possibly at its highest so far. The country’s generally significant rate had almost halved amidst the pandemic, owing to the fear of no new opportunities. Now that hiring sprees are back, so are potential attrition instances. The rates are predicted to shoot up from 12% last year to 23 by the end of 2021, adding up to one million potential resignations. The industry is massively imbalanced, with a high demand for data scientists and a low supply of capable experts. This is a matter of concern for the Indian IT industry and data science organisations banking on data scientists. The industry leaders at AIM’s Leaders Council have created and\/or led some of the most sought after companies to work at by data science professionals. Today, they have leveraged their years of data science experience to share inputs on why the attrition rates are so high and suggest successful ways to retain the best talent. Why are attrition rates so high? Lack of Connection between Employees And Employer In the “work from home” mode, the relation between employees and employers has become very transactional. There is no sense of belonging, no connection to a bigger purpose or the sense of a team accomplishing something together. We often fail to realise that when one spends a good amount of their awake time in the office, it is not just about the work but also about the social connection they build with their colleagues. We need to focus on engaging more with our team members, creating a sense of belonging and helping them connect to the organisation’s bigger purpose.Sayandeb Banerjee, Co-Founder and CEO at TheMathCompany Move to Start-ups There have been more opportunities, especially in the start-up ecosystem. That, along with a fear of missing out (FOMO) when their colleague gets an offer, have been some of the key reasons for the high attrition rates. Manoj Madhusudanan- Head of dunnhumby India High Salary Expectations Given the market situation, folks across entry-level to 5-6 years experience expect a hyper-growth in salary, designation, roles and responsibilities. While this is fine in cases where it is deserved, it becomes tough in a services scenario where a higher salary or designation comes with the expectations on other attributes like solution strategy, data science thought leadership or client management skills as against just technical depth and execution, which a lot of purely technical focussed folks are not able to demonstrate neither are keen to ramp up on. Such expectations lead to dissatisfaction on both employee and employer end, misalignment in organisational vs personal goals and assessment against those, slower growth and salary hikes compared to captive units, among other challenges. Ruble Joseph, Lead Strategist (VP) – Global Data Science and Analytics Practice Inability to be Generalists When employees join the company while we are already working on a project, there is a mismatch between candidates expectations vs what they end up working on, at least at the start of the tenure. This becomes challenging with folks with a myopic view of working on specific algorithms, platforms or solution modules. These folks usually leave as they are not open to playing a generalist data scientist role without realising the broader set of areas, roles, opportunities, including innovation and first of its kind solution development that they may get in the future once they spend some time with the organisationRuble Joseph, Lead Strategist (VP) – Global Data Science and Analytics Practice While the leaders have agreed that attrition rates are high, according to Manoj Madhusudanan, recent trends indicate that the frenzy has already peaked. But what can leaders do to ensure they have a low attrition rate and high employee satisfaction? How can this be tackled to retain the best talent? Role-specific Hiring The trick is having detailed job descriptions and hyper-customised role-specific hiring. This, along with differentiated persona roles for generalists vs SMEs, customised career tracks, expertise area-specific roles, can make a big difference in employee retention. Support these specific hirings with better alignment, customised performance evaluation parameters and goals and allow for the ease of shuffling across roles. Ruble Joseph, Lead Strategist (VP) – Global Data Science and Analytics Practice Campus Hiring & Grooming from the Get-Go Creating and grooming talent as against only lateral hiring. Folks who start their data science journey with a particular organisation tend to stick longer than laterals hired at higher salaries if given adequate opportunities to grow fast. Create a smaller cluster of teams with the opportunity for multiple folks to step up and grow faster, take ownership, accountability and create leadership visibility much faster. Ruble Joseph, Lead Strategist (VP) – Global Data Science and Analytics Practice I believe we should hire more from campuses at entry-level and groom them intensively, rather than simply hiring trained people from each other. We have seen that those hired from campuses grow with the company and are less likely to leave. Manoj Madhusudanan- Head of dunnhumby India In-house Learning and Upskilling Opportunities It’s imperative to ascertain that professionals are continuously growing, working on a variety of assignments. Lack of growth in skills or boredom becomes a trigger for looking out.Swati Jain, Vice President Analytics at EXL Service Invest in programs to upskill citizen data scientists and create a substantial backlog of internal talent to augment your current data science needs.Ashwin Thota, Principal Data Scientist at Bose Offering Multidimensional Flexibility Most data scientists are high achievers; they want to be constantly challenged with new opportunities. Consider giving them “freedom within a framework”. Allow them to take some learning time to explore new ways of solving existing business problems with data.Ashwin Thota, Principal Data Scientist at Bose Rewards Go a long Way Most data science roles inherently include a research component. Reward your data scientists for ANY finding they bring to the table. Realise that there are no bad results from a well-conducted experiment.Ashwin Thota, Principal Data Scientist, Bose Establishing a Link between the Role and the Impact While Great Resignation has led to job changes across roles, I believe that one thing that results in Data Science attrition is the kind of work handed over to them and the impact. Nalin Goel, Senior Vice President – Product at MoEngage I believe in Nitin Nohria’s ABCD model for workplace motivation drives: Acquire – Tangible benefits; Bond – Connection, friendship; Comprehend – Meaningful work, skilling; Defend – Recognition. “Comprehend” plays a significant role in the analytics and data science space. Understanding the expectations of the role and connecting the same to the desired output is key in growing an inbuilt sense of motivation for the job.Satyamoy Chatterjee, Executive Vice President at Analyttica Datalab In the fast-flowing resource that is Data Science, attrition is a hole in the pipe, carrying a humongous potential to slow, if not bring the system down. The pieces of advice from our council leaders provide us with both the visible and invisible causes behind the situation and, at the same time, give us an insight into solutions that work. India is home to cities setting examples in the fast-paced technological boom. As per statistical research by London & Partners, Bangalore emerged as the world’s fastest-growing technological hub since 2016, overtaking London in the first position, followed by Mumbai at the sixth. Seeing this, it becomes worrisome to observe that attrition remains higher in the Indian IT industry than in other countries of the world. Clearly, data science organisations and leaders need to make retention a focal point in the employee lifecycle to ensure the constructive growth of not just the organisations but the industry as a whole. This article is a collation of quotes by members of the AIM Leaders Council. AIM Leaders Council is an invitation-only forum of senior executives in the Data Science and Analytics industry. To check if you are eligible for a membership, please fill the form here.","excerpt":"“The Great Resignation” is picking up again – attrition in the Indian data science space is possibly at its highest so far. The country’s generally significant rate had almost halved amidst the pandemic, owing to the fear of no new opportunities. Now that hiring sprees are back, so are potential attrition instances. The rates are […]","categories":["AI Highlights"],"tags":[],"author_name":"Avi Gopani","publish_date":"2021-12-06T15:00:00","publication_year":"2021","word_count":1324,"keywords":["data science","Go","TPU","AI","RAG","Aim","ViT","analytics","GAN","R"],"extracted_tech_keywords":["AI","data science","analytics","Aim","RAG","TPU","R","Go","GAN","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/council-post-the-great-attrition-in-data-science-what-do-experts-say\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":28266,"title":"The Ideal Data Scientist Toolkit Is Defined By What It Should NOT Contain, Say AI Experts From Cartesian Consulting","content":"The next interaction in a series of interviews for our theme this month — leading tools and techniques used by analytics and artificial intelligence practitioners — is with Ramasubramanian (Ramsu) Sundararajan, head of AI Labs and Tapan Khopkar, head of Innovation Labs at Cartesian Consulting. In the responses jointly curated by the two tech experts lead us through the commonly-used tools in analytics and AI, if researchers prefer for open source or paid tools, an inside of data science toolkit and others. Analytics India Magazine: What are the most commonly-used tools in analytics, AI and data science? SQL for basic data access. Much of what we work with is still in relational databases, and just getting the data out is key. For analysis after that, R or Python along with the contributed libraries largely cover what we need to do. For visualisation, Tableau is quite popular. Above all, never underestimate the power of MS Excel! AIM: What is the most productive tool that you have come across? It depends on the task. If the challenge is to derive simple insights from data, SQL along with a good visualisation tool like Tableau is often sufficient. When it comes to data manipulation and model building, productivity is largely a function of one’s familiarity with the language, and the availability of libraries and contributed code. For instance, we find the R ecosystem to be quite rich when it comes to statistical modelling, while Python scores when it comes to deep learning, NLP etc. AIM: Do you prefer tools that are open sourced or paid? Please elaborate on the benefits. Here are the factors that play into this decision: The size of the user community largely determines how easy it is for people to get guidance, contributed code, tutorials etc. on a particular platform. By and large, user communities grow faster when the product is free; however, this is not always true. The traditional view used to be that, while open source grew faster, the product support was better for proprietary, paid software. With communities like StackOverFlow, GitHub etc., this is not quite as true as it used to be. However, this has more to do with the usage of the tool as it is, not so much the modification by the larger community. If you take an open source tool like Tensorflow, even though it’s open source, if there’s an issue with it, the large majority of users will wait for Google to fix it. In that sense, it’s not different from paid software. AIM: What are the most common issues you face while dealing with data? How is selecting the right tool critical for problem-solving? Data is rarely neat. When we study data analytics in college or during online courses, the story often starts where the data is clean and ready for the application of modelling algorithms. Even topics about data cleaning such as missing data handling are often handled perfunctorily. Real life problems come with data that is messy, incomplete, inconsistent, sometimes very large etc. Additionally, we don’t always have complete visibility into the data generation process and are therefore forced to rely on our assumptions while manipulating it. If for a certain kind of problem, the chosen tool doesn’t provide the necessary functionality to make it easy for the analyst, then it’s a problem. Otherwise, it largely depends on the analyst’s familiarity with the tool. For instance, someone might be most comfortable using R along with the data.table library; for some others, it’s R with dplyr, and for yet others, it’s Python with pandas, etc. For large datasets, other issues also play a critical part. Is there a latency time constraint both during development and\/or deployment? Is the solution largely parallelizable? Is the solution like to involve data-parallel or compute-parallel operations? These aspects determine the choice of tool, and this choice becomes critical to productivity. AIM: How do you select tools for a given task? The critical factors, in no particular order, are: Functionality or versatility Scalability Availability of expertise Availability of contributed code and libraries Cost Performance Longevity of the platform AIM: What are the most user-friendly languages and tools that you have come across? Excel, RStudio and Tableau come to mind right away. For development in Java or C++, Eclipse is very easy to use. AIM: What does an ideal data scientist toolkit like? The ideal data scientist is someone who has sufficiently deep knowledge of enough tools or platforms so that he\/she’s able to pick the right tool for solving parts of the problem, and then able to piece together a solution that may involve multiple platforms. Viewed through this lens, the ideal data scientist toolkit is better defined by what it is not. It is not a static toolkit. AIM: What is the most preferred language used by the team? Our team mostly uses Python for text and image datasets, and R for numeric datasets. For applications where scale is an issue, we use Java. AIM: What is the most preferred cloud provider — AWS, Google or Azure? We’re neutral on this point, at this point. AIM: What are some of the tools used for scaling data science workloads; for eg., Dockers are gaining popularity vis a vis spark? Scale isn’t one issue – it has many hues. For instance, large batch processing of eminently parallelizable datasets is usually well-solved using map-reduce algorithms, whereas Scala does well in real-time scenarios. Large-scale computation in machine learning algorithms with inherent parallelisms, such as neural networks, is now often done using GPUs. And sometimes, the answer to the scale problem is to simply rewrite the code in C++ or Java instead of in Python or R – one doesn’t always need to throw a bigger computer or a parallel computing framework at the problem, just a better algorithm in a different language. Further, scale issues look different at runtime. How quickly is the data to be processed and acted upon, relative to its speed of acquisition? Where does the solution need to reside – on the cloud or on an edge device with limited computing and memory capacity? AIM: What are some of the proprietary tools developed in-house by the company? Watch this space!","excerpt":"The next interaction in a series of interviews for our theme this month — leading tools and techniques used by analytics and artificial intelligence practitioners — is with Ramasubramanian (Ramsu) Sundararajan, head of AI Labs and Tapan Khopkar, head of Innovation Labs at Cartesian Consulting. In the responses jointly curated by the two tech experts […]","categories":["AI Features"],"tags":["ai consulting","Interviews and Discussions"],"author_name":"Srishti Deoras","publish_date":"2018-09-12T12:58:05","publication_year":"2018","word_count":1035,"keywords":["data science","ai consulting","machine learning","artificial intelligence","AI","neural network","NLP","Aim","deep learning","analytics","TensorFlow","Interviews and Discussions"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","deep learning","neural network","NLP","data science","analytics","Aim","TensorFlow"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/the-ideal-data-scientist-toolkit-is-defined-by-what-it-should-not-contain-say-ai-experts-from-cartesian-consulting\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10170200,"title":"How AWS Overhauls Legacy Code for Good","content":"India is emerging as a global force in software development, and the next six months could further define how developers across the region build, scale, and maintain digital systems. Speaking with AIM, Anupam Mishra, director for developer programs for India and South Asia at AWS, shared how AI, open protocols, and a growing developer community are driving these shifts. AWS sees developers not just as coders but as key contributors to the digital economy. “Developers have an outsized impact on the economy. They’re the ones creating products that go on to become large companies and successful startups,” Mishra said. Mishra, citing Evan Data Corporation, said that India has the second-largest population of developers. “We believe there’s never been a better time to be a developer,” he said. He added that with AI coding tools like Amazon Q, the job of developers has never been easier. For instance, Persistent Systems, an AWS customer, used Amazon Q Developer to upgrade legacy Java code. “The process was 83% faster than doing it manually,” Mishra said. Mishra further added that AWS is also introducing agentic tools that do more than just suggest improvements. “Agentic capabilities are becoming more common, where developers expect AI not only to tell them what to do but also to act on it,” Mishra said. Tools like Amazon Q Developer now come with agents for code and security reviews. “Sometimes, as a developer, it’s hard to think about all possibilities. The tool looks at vulnerabilities that have happened in the past and highlights them,” he said. Notably, the company recently announced the general availability of AWS Transform, a new AI tool to accelerate the migration and modernisation of enterprise workloads. The company said it helps organisations complete transformation projects up to four times faster by automating complex migration tasks through agentic AI. The service supports modernisation for key enterprise platforms, including VMware, mainframes, and .NET applications. Mishra said it allows developers to focus more on design and building things that truly matter. “We’re supporting them with AI tools to drive this productivity, whether it’s Java upgrades, moving away from legacy systems like mainframes, or even transitioning from older platforms like VMware,” he added. Mishra said ​​tools like Q Developer are continuously updated based on developer feedback. He said developers using AWS’s .NET modernisation tools reported four times faster conversion speeds from Windows to Linux. “79% of change requests generated are accepted without modification,” he added. The .NET agent supports porting Windows-based applications to Linux, potentially cutting operating costs by up to 40% by reducing licensing and maintenance requirements. It also transforms code, runs unit tests, and validates Linux-readiness, enabling parallel transformation of hundreds of applications. The mainframe agent assists in decomposing monolithic z\/OS COBOL applications into cloud-ready components. Meanwhile, the VMware agent helps enterprises avoid rising licensing costs by automating migration planning. Through automated discovery and dependency mapping, tasks that usually take weeks can now be completed in minutes. Earlier this year, the company also launched Amazon Q CLI Developer Agent, an AI-powered command-line interface tool to accelerate software development workflows directly within the terminal environment. Mishra said it enables developers to bring these AI-powered capabilities into their workflows, whether for testing, documentation, refactoring, or deployment. “The goal isn’t just to help with coding, but to make AI useful across the full lifecycle, from idea to implementation–removing repetitive tasks and helping teams become faster, more innovative, and agile.” Last year, Amazon CEO Andy Jassy revealed that by leveraging Amazon Q, the company was able to save 4,500 developer years of work. He said that in under six months, the company has been able to upgrade more than 50% of its production Java systems to modernised Java versions at a fraction of the usual time and effort. “We had developers from one of our teams who had to upgrade a lot of code from older versions of Java. We use a lot of Java — Java 7 and Java 8 — and it would have taken almost 4,500 years to convert that in a normal way, but it was done in a matter of days,” said Mishra. Open Protocols and New Developer Workflows Another growing area is the use of open protocols to connect AI tools with external systems. AWS recently launched support for MCP (Model Context Protocol), which lets generative AI agents interact with databases, developer tools, and more. “If you’re a developer and have a PostgreSQL database, you can now say ‘Find the number of students’ and the agent can interact directly with the database. Earlier, it couldn’t,” Mishra said. “MCP allows our tools, or generative AI, to talk to several tools in a conversational manner. It now has more context available.” Amazon Q Developer CLI (a command-line interface tool for developers) can now connect to MCP tools — either via AWS’s pre-built integrations or any MCP-compatible server that supports stdio (standard input\/output). The barrier to adoption is low. “It’s a small configuration change, just a few lines of code,” Mishra said. Bullish on India AWS is working closely with the community to build trust and ensure adoption. The company currently runs 26 user groups across India. “We have 250 AWS Heroes across 53 countries. 20 of those are in India,” Mishra said. These community leaders share feedback, test tools early, and engage in peer learning. Looking ahead, AWS expects the adoption of natural language interfaces and agentic tools to grow. Mishra said that Amazon Q Developer now supports languages like Hindi as well. An IDC report projects that 70% of digital solutions will feature natural language interfaces by 2028. “That’s a forward-looking projection, but we’re already seeing it take shape,” Mishra said. India’s developer ecosystem is not just keeping pace. It is influencing how software is built globally. With a strong community and growing access to cutting-edge tools, the role of developers is expanding beyond code. As Mishra put it, “It’s an economy shaped by developers.”","excerpt":"AWS’s .NET modernisation tools reported four times faster conversion speeds from Windows to Linux.","categories":["Global Tech"],"tags":["AWS"],"author_name":"Siddharth Jindal","publish_date":"2025-05-19T19:04:18","publication_year":"2025","word_count":988,"keywords":["PostgreSQL","agentic AI","TPU","AWS","AI","RAG","Aim","generative AI","SQL","R"],"extracted_tech_keywords":["AI","generative AI","agentic AI","Aim","RAG","AWS","TPU","PostgreSQL","R","SQL"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/how-aws-overhauls-legacy-code-for-good\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10041000,"title":"Best Python Libraries For Data Science In 2021","content":"Python is an interpreted, interactive, portable and object-oriented programming language. This open-sourced general-purpose language runs on many Unix variants, including Linux and macOS, and Windows. Python has applications in hacking, computer vision, data visualisation, 3D Machine Learning, robotics, and is a favourite of developers worldwide. Below, we list the ten most popularly used Python libraries for Data Science: TensorFlow Developed by Google Brain Team, TensorFlow is an open-source library used for deep learning applications. Originally developed for numerical compilations, it offers a comprehensive and flexible ecosystem of tools, libraries and community resources, enabling developers to build and deploy ML-based applications. First released in 2015, the Google Brain team recently launched its latest version, TensorFlow 2.5.0 with more features. It supports Python 3.9. To know more, click here. NumPy Developed by Travis Oliphant in 2015, NumPy or Numerical Python is a fundamental library for mathematical and scientific computations. The open-source software has functions of linear algebra, Fourier transform, and matrix computations and is mainly used for applications where speed and resources are important. NumPy aims to provide array objects 50x faster than traditional Python lists. Data science libraries including SciPy, Matplotlib, Pandas, Scikit-Learn and Statsmodels are built on top of NumPy. To know more, click here. SciPy SciPy or Scientific Python is used for complex mathematics, science and engineering problems. It is built on the NumPy extension and allows developers to manipulate and visualise data. SciPy provides user-friendly and efficient numerical routines for linear algebra, statistics, integration and optimisation. Its applications include multidimensional image processing, solving Fourier transforms and differential equations. To know more, click here. Matplotlib Developed by John Hunter, Matplotlib is one of the most common libraries in the Python community. It is used for creating static, animated and interactive data visualisations. Matplotlib provides endless customisation and charts. It enables developers to use histograms to scatter, customise and configure plots. The open-source library offers an object-oriented API for integrating plots into applications. To know more, click here. Pandas Developed by Wes McKinney, Pandas is used for data manipulation and analyses. It provides fast, flexible and expressive data structures and provides features such as handling of missing data, fancy indexing and data alignment. Pandas provides fast, flexible and expressive data structures that helps developers work with labelled and relational data. It is based on two main data structures– Series, and Frames. To know more, click here. Keras Open-source software library Keras provides an interface for the TensorFlow library and enables fast experimentation with deep neural networks. It was developed by Francois Chollet and was first released in 2015. Keras offers utilities for compiling models, graph visualisation and dataset analysis. Further, it offers prelabeled datasets that can be imported and loaded directly. It is user-friendly, versatile and suited for creative research. To know more, click here. SciKit-Learn SciKit-Learn features classification, regression and clustering algorithms, including DBSCAN, gradient boosting, support vector machines and random forests. David Cournapeau built the library on top of SciPy, NumPy and Matplotlib for handling standard machine learning and data mining applications. SciKit-Learn is an effective tool for predictive data analysis. To know more, click here. Statsmodels Statsmodels is part of the Python scientific stack, oriented towards data science, data analysis and statistics. It is built on top of NumPy and SciPy and integrates with Pandas for data handling. Statsmodels allows users to explore data, estimate statistical models and perform statistical tests. To know more, click here. Plotly Plotly is a collaborative, web-based analytics and graphing platform. It is one of the most powerful libraries for ML, data science and AI-related operations. Plotly is publication-ready and immersive and is used for data visualisation. Plotly can easily import data to chart, allowing developers to make slide decks and dashboards with ease. It is used for the development of tools like Dash and Chart Studio. To know more, click here. Seaborn Seaborn is Python’s most commonly used library for statistical data visualisation, used for heatmaps and visualisations that summarise data and depict distributions. It is based on Matplotlib and can be used on both data frames and arrays. Seaborn is used for basic plottings– bar graph, line charts and pie charts. To know more, click here.","excerpt":"Python is an interpreted, interactive, portable and object-oriented programming language. This open-sourced general-purpose language runs on many Unix variants, including Linux and macOS, and Windows. Python has applications in hacking, computer vision, data visualisation, 3D Machine Learning, robotics, and is a favourite of developers worldwide.  Below, we list the ten most popularly used Python libraries […]","categories":["AI Trends"],"tags":["fourier transform machine learning","Keras","keras python","numpy","pandas","Python Libraries","Tensorflow"],"author_name":"Debolina Biswas","publish_date":"2021-05-30T14:00:00","publication_year":"2021","word_count":696,"keywords":["data science","machine learning","Keras","fourier transform machine learning","Python Libraries","AI","ML","neural network","numpy","computer vision","Ray","Aim","pandas","deep learning","analytics","keras python","Tensorflow"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","neural network","computer vision","data science","analytics","Aim","Ray"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/best-python-libraries-for-data-science-in-2021\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10022787,"title":"Machine Learning Deployment Is The Biggest Tech Trend In 2021","content":"“What good is an ML model if it isn’t fast? doesn’t scale? isn’t accurate enough? takes weeks to deploy? and costs too much?”Luis Ceze, CEO, OctoML Having machine learning in a company’s portfolio used to be an investor magnet. Now, the market is bullish on MLaaS, with a new breed of companies offering machine learning services (libraries\/APIs\/frameworks) to help other companies get their job done better and faster. According to PwC, AI’s potential global economic impact will be worth $15.7 trillion by 2030. And, as interests slowly shift towards MLOps, it is possible that these companies, which promise to scale and accelerate ML deployment, might grab a bigger piece of the pie. Last week, OctoML raised $28 million. The Seattle-based startup offers a machine learning acceleration platform built on top of the open-source Apache TVM compiler framework project. The $28 million Series B funding brings the company’s total funding to $47 million. Image credits: OctoML “90% of machine learning models don’t make it to production.”OctoML For OctoML’s CEO, Luis Ceze, there is still a significant gap between building a model and making it production-ready. Between rapidly evolving ML models, wrote Ceze in a blog post, ML frameworks and a Cambrian explosion of hardware backends makes ML deployment challenging. “It is not easy to make sure your model runs fast enough and to benchmark it across different deployment hardware. Even if your determined machine learning team has hurtled through this gauntlet, they still have to go through a whole different set of challenges to package and deploy at the edge,” explained Ceze. A good performance in ML models requires long hours of manual optimizations. These long hours will then translate into hefty cloud bills. Added to this is the model packaging which varies with devices and platforms. According to Ceze, there are no modern CI\/CD integrations to keep up with model changes. “What good is an ML model if it isn’t fast? doesn’t scale? isn’t accurate enough? takes weeks to deploy? and costs too much?,” questioned Ceze as he made a case for OctoML. OctoML addressed these pain points with their open-source machine learning compiler framework Apache TV, which according to the team, has quickly become the go-to solution for developers and ML engineers to maximize ML model performance on any hardware backend. “With OctoML we are establishing the first Machine Learning Acceleration Platform that will automatically maximize model performance while enabling seamless deployment on any hardware, cloud provider, or edge devices,” said Ceze. Be it MLOps or XOps, these services are designed to ease the developers of technical debt that these mega ML models accumulate with changing complexities. Apart from OctoML, there are a few other startups that have succeeded in convincing the investors. Let’s take a look at couple of them: Verta Funding till date: $10 million The team at Verta is building software for data science teams to address the problem of model management — how to track, version, and audit models used across products. Verta MLOps software supports model development, deployment, operations, monitoring, and collaboration enabling data scientists to manage models across their lifecycle. So far, the company has $10 million in funding and it promises to make robust, scalable, mature deployable models a reality. Algorithmia Funding till date: $38.1 million Image credits: Algorithmia “We’re obsessed with helping organizations get ML models into production because that’s the only way they can generate business value,” said the team at Algorithmia. Their enterprise MLOps platform manages all stages of the production ML lifecycle within existing operational processes, so you can put models into production quickly, securely, and cost-effectively. Unlike inefficient and expensive do-it-yourself MLOps management solutions that lock users into specific technology stacks, Algorithmia automates ML deployment, optimizes collaboration between operations and development, leverages existing SDLC and CI\/CD systems, and provides advanced security and governance. Algorithmia’s funding (Source: Crunchbase) Today Algorithmia’s services are used by over 130,000 engineers and data scientists, including the United Nations, government intelligence agencies, and Fortune 500 companies. “It’s [MLOps] going to be an essential component to enterprises industrializing their AI efforts in the future,” said Diego M. Oppenheimer, Algorithmia’s CEO in a recent interview with GitHub. Databand.ai Funding: $14.5 million Databand brings in the similar flavor into the ML ecosystem. The team Databand is trying to solve the problems that arise due to increasing data workloads. The company founded by Josh Benamram, Victor Shafran and Evgeny Shulmanhelps helps data engineering teams catch data pipeline issues and trace the impact of those problems across end-to-end data flows. Databand’s platform includes an application for visualizing pipeline metadata, and an open source library for integrating with your Python, Java, Scala, or SQL data processes. Data pipeline monitoring is a key aspect of machine learning deployment. We can clearly see how targeting even a niche aspect of the whole ML deployment can land big investors. Image credits: Gartner Modern day software companies are in the process of or have already embraced machine learning as a key tool. Now they are at a crucial juncture where they can either leverage the MLOps services offered by these startups or build everything on their own. But, there are not many reasons why an organization looking to transition to ML will take the pain of MLOps. As companies look to leverage ML minus the deployment headache, niche players like OctoML will  continue to pop up. Even the latest Gartner survey lists scalability and acceleration of machine learning deployment as two driving forces that will continue to trend this year. According to Gartner, XOps— a variant of MLOps that deals with efficiencies in data, machine learning, model, platform will try to implement best DevOps practices and ensure reliability, reusability and repeatability.","excerpt":"“What good is an ML model if it isn’t fast? doesn’t scale? isn’t accurate enough? takes weeks to deploy? and costs too much?” Luis Ceze, CEO, OctoML Having machine learning in a company’s portfolio used to be an investor magnet. Now, the market is bullish on MLaaS, with a new breed of companies offering machine […]","categories":["Deep Tech"],"tags":["MLOps"],"author_name":"Ram Sagar","publish_date":"2021-03-23T16:00:00","publication_year":"2021","word_count":948,"keywords":["data science","Go","machine learning","AI","ML","MLOps","RAG","Python","SQL","R"],"extracted_tech_keywords":["AI","machine learning","ML","data science","MLOps","RAG","Python","R","SQL","Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/machine-learning-deployment-is-the-biggest-tech-trend-in-2021\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":68807,"title":"What Is Google’s AI Adoption Framework","content":"Google Cloud, has recently launched their AI Adoption Framework whitepaper, authored by Donna Schut, Khalid Salama, Finn Toner, Barbara Fusinska, Valentine Fontama, and Lak Lakshmanan, to provide a guiding framework for enterprises to leverage the power of AI effectively. Google Cloud’s AI Adoption Framework has been designed on four pillars of an organisation — “people, process, technology, and data.” Google Cloud blog further noted that these four pillars of organisations should follow six critical themes — “learn, lead, access, scale, automate, and secure — for their AI success. According to Google, “These themes are foundational to the AI adoption framework.” Explaining further, the blog highlighted these six steps that are critical for businesses: The first step concerns the scale of ‘learning’ within an organisation, which includes the process of upskilling existing employees, recruiting new talents, and augmenting analytics and engineering professionals with “experience partners.” This process of learning will help organisations to decide which analytics and machine learning skills would be required for the business, and accordingly, they can strategies their hiring process amid this crisis. Second, comes the ‘leading’ which concerns whether or not the leaders of organisations provide enough support and guidance to data scientists and engineers to deploy machine learning and artificial intelligence in their business projects. This step would help businesses in understanding the structure of the team, the cost of the projects and the governance of the projects to encourage cross-functional collaboration in the organisation. Next is the ‘access” to data, where companies recognise the data management strategies and analytics professionals are able to collect, share, discover, analyse the data and other ML artefacts. Then comes the ‘scaling’ process, where companies can define their ability to use cloud-native ML services and scale large amounts of data with reduced business expenses. This process will also guide you to understand the cloud-based services and how the workloads are allocated to them. The fifth is the ‘securing’ process, which is extremely critical for organisations. In this process, enterprises can understand their security strategies to protect sensitive business data and essential information. Additionally, this step will also help companies to ensure deploying responsible and explainable AI practices, which will intern drive their business value. Lastly, is the process of ‘automating,’ where businesses would be able to realise their ability to deploy, execute, and operate the technology, to enhance and streamline data processing and ML productions. This process will further help companies to understand and track data lineage and monitor the process. These processes are critical for businesses to adopt effective AI practices for their organisations. According to their blog post, “Successfully adopting AI in your business is determined by your current business practices in these areas…” Also Read: Why Majority Of Data Science Projects Never Make It To Production The post further noted that the preparation for companies to adopt AI in their business processes would be critically determined by their current business practices on the aforementioned steps. Further, these steps go through AI maturity phases where it undergoes tactical, strategic and transformational phases. In each of these levels, it will be determined how AI is going to be adopted for creating value for organisations. Taking an example of the “learning” step, where in the tactical phase, organisations will encourage self-motivated learning using online courses, and businesses will prefer outsourcing their analytics needs, and therefore, their outdent would be in hiring any ML talent. On the other hand, for the same step, learning in the strategic phase, the businesses will be looking to hire data science and ML professionals as well as will be urging their existing employees to go through continuous training programs. Businesses will only be looking to involve third parties for consultation or specialised knowledge. Thirdly, once this step is in the transformational phase, companies will enhance their work by involving data scientists in their core businesses, as well as hiring new talents with industry expertise for innovation. In this phase, businesses also partner with other companies to augment their technical capabilities. For more information, see the image below: When AI gets incorporated in all these aspects of organisations, according to Google Cloud, then those companies are “fully harnessing” the capability of artificial intelligence to transform their work. “But at every step along the way, adding effective AI capabilities can bring benefits.” Such an effective technology agnostic framework will not only help businesses to understand and strategies their AI game but also help them resolve the challenges that come along with it. A Deep Dive In The AI Maturity Phases The AI maturity depends on three phases — tactical, strategic, and transformational. And according to google, “every organisation’s AI capability falls under these three phases.” Each stage has separate characteristics and provide distinct opportunities for growth and advancement, explained below: Tactical Phase Organisations that are still in the tactical phase of AI adoption are working on gaining short-term benefits from AI. This is why the majority of them use AI for small cases, and the developers mostly rely on exploratory data analysis, and automated AI tools for “proofs of concepts and prototyping.” For instance, companies would prefer using pre-built and pre-trained facial recognition technology to monitor their business operations or use predictive analytics to enhance customer segmentation, rather than obtaining ML for complex business problems, which are usually outsourced to their parties. Although this phase doesn’t include much of innovation or requires hiring newer talent, it indeed has some benefits. In the tactical phase, organisations can have access to better and cleaner data for making a better-informed decision with actionable insights. Also, considering this is the first phase of AI maturity, this indeed helps validate their process and power of ML in front of stakeholders. Companies, at this phase, should “develop foundational skills for core data wrangling and descriptive analytics.” For these businesses need to collate the siloed data into a unified data lake with decentralised access, where every employee of organisations can access the data to make necessary decisions. Strategic Phase Secondly, in the strategic phase, organisations are clearly focused on deploying AI in different parts of organisations that leverage both ready to use and custom models to bring business value. For this, companies also require to hire talents as well as build core ML teams that can help deploy and manage the technology for businesses. This core team will not only help companies accelerate their business but will also ensure constant governance and ML collaboration amid different groups. At this phase, ML professionals use data in predictive analytics to solve business problems and also deploy ML in production. Thus, in this phase, companies can benefit from developing AI capability that is customised to business requirements. With AI and ML collaboration within companies, teams will help the businesses accelerate automation in every aspect. Transformational Phase The transformational phase is the most advanced phase where businesses actively use artificial intelligence and machine learning to innovate, streamline their business, and create an AI-driven culture amid organisations. In this phase, companies not only build and deploy machine learning platforms but also make ML accessible to every employee of organisations. Further, in this phase, enterprises work on a hybrid model where they combine functional or product-specific AI teams with broader product teams, which is also combined with analytics capabilities allowing employees to be responsible for building their own function-specific ML models. This, in turn, will enable companies to construct high-quality technical solutions which can be used to solve real-world problems. In this phase, businesses also define ethical and responsible AI practices, which can hugely benefit organisations in the longer run. “The ML platform is supported with tools for continuous integration, training, and model serving and monitoring. Building and maintaining such a platform is a shared responsibility between ML and software engineers with skills in infrastructure, DevOps, and SRE,” stated the blog. Moreover, companies at this phase usually tend to work on cutting-edge research and robust innovation in the areas where they have unique capabilities, domain knowledge, and data availability so that they can build a competitive advantage over others. For more information, read the whitepaper here. Wrapping Up To summarise, it can be said that companies with any level of maturity, whether it be tactical, strategical or transformational, can benefit from artificial intelligence. However, the level of AI maturity can only be determined by analysing the amount of AI involvement in the business. By analysing these perimeters, stated in Google Cloud’ AI Adoption Framework, companies can understand how their businesses are currently functioning in the AI maturity scale, and how they can devise their strategies for better business outcomes.","excerpt":"Google Cloud, has recently launched their AI Adoption Framework whitepaper, authored by Donna Schut, Khalid Salama, Finn Toner, Barbara Fusinska, Valentine Fontama, and Lak Lakshmanan, to provide a guiding framework for enterprises to leverage the power of AI effectively.  Google Cloud’s AI Adoption Framework has been designed on four pillars of an organisation — “people, […]","categories":["Global Tech"],"tags":["AI adoption","AI adoption India","best facial recognition software","how does artificial intelligence work"],"author_name":"Sejuti Das","publish_date":"2020-07-02T14:00:00","publication_year":"2020","word_count":1430,"keywords":["data science","Go","artificial intelligence","AI adoption","machine learning","AI","best facial recognition software","R","ML","RAG","AI adoption India","how does artificial intelligence work","analytics","predictive analytics"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","data science","analytics","RAG","predictive analytics","R","Go"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/what-is-googles-ai-adoption-framework\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":1325,"title":"How HPE Aruba plays a role in securing the Internet of Things","content":"Aruba’s solutions keeps the network and corporate assets safe It’s a catchword that dominates every conversation and figured heavily at the 5th edition of Network University 2017, hosted at HPE and chaired by Santanu Ghose, Director Networking, HPE Aruba and Professor Sadagopan, IIT – B’s Director. We are talking about internet of things (IoT) and the hype surrounding the technology. Not to forget the doom and gloom talk regarding massive fragmentation in the IoT what with every technology giant launching their own IoT platform. Santanu Ghose, Director Networking, Aruba, a HP Enterprise Company But just like cloud, IoT is the latest buzzword and is having its moment in the sun, revealed Prof. Sadagopan. The thought was echoed by Ghose as well. “No company can offer the full gamut of services when it comes to Internet of Things (IoT). When companies in the industrial space such as GE which is writing IoT applications and somebody such as Intel is already making sensors that capture data, are already working together, where is the fragmentation,” he shared. To succeed in Internet of Things (IoT) space, companies have to work together, reiterated Ghose. But where does Aruba (HPE company) fit in the bigger IoT picture?  While GE and Siemens are powering Industrial Internet of Things (IIoT) through platforms, Aruba has the capability of handling the “application part of the infrastructure”. Aruba plays a pivotal role in securing trust and managing IoT security issues. “From connectivity to security and management we facilitate that application to be run securely and facilitate that application to be managed,” shared Ghose. With IoT devices becoming increasingly common in the enterprises settings, security and access has become a major concern. Many of the IoT devices in use today are not adequately secured. That’s where Aruba’s ClearPass Universal Profiler and Aruba Clear Pass Policy Manager play a role in safeguarding the network by providing visibility and enforcement policy. ClearPass Universal Profiler helps you see what’s on your network and Aruba Clear Pass Policy Manager allows you to enforce policies for both wired and wireless and integrate all the best security solutions to protect the enterprise’s assets. Aruba Networks is leading the race in software defined networking and is changing the way WiFi is delivered. The market leader in wired and WLAN infrastructure, declared by Gartner has paved the way for Network-as-a-Service model as the way of doing business for organizations across the board – healthcare, education, retail and many more. The benefits are manifold — it maximizes customer engagement for business and helps make decisions on a fly in the operation cycle through their cutting edge Mobile First Platform. Aruba’s upgraded mobility architecture goes beyond simple connectivity to enhancing customer experience and simplifying complex business problems. The event was organized to demonstrate Aruba’s Mobile First Platform largely focused on Aruba’s mature mobility architecture that is a wide platform that enables innovation and problem solving for businesses. Also present on the occasion were Aruba’s top brass Alain Carpentier, VP, Worldwide Sales of HPE Aruba and Steve Wood, VP, Asia Pacific Aruba .","excerpt":"It’s a catchword that dominates every conversation and figured heavily at the 5th edition of Network University 2017, hosted at HPE and chaired by Santanu Ghose, Director Networking, HPE Aruba and Professor Sadagopan, IIT – B’s Director. We are talking about internet of things (IoT) and the hype surrounding the technology. Not to forget the […]","categories":["IT Services"],"tags":["hpe internet of things"],"author_name":"Richa Bhatia","publish_date":"2017-01-24T04:47:12","publication_year":"2017","word_count":510,"keywords":["Go","programming_languages:R","AI","RPA","innovation","RAG","ViT","Rust","GAN","hpe internet of things","R"],"extracted_tech_keywords":["AI","RAG","R","Go","Rust","GAN","ViT","RPA","innovation","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/it-services\/hpe-aruba-plays-role-securing-internet-things\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10094320,"title":"Bigger is Not Always Better","content":"There has always been this notion with large language models (LLMs) — the bigger the model, the better it would perform. This has made a lot of companies boast about the number of parameters of their models. GPT-3 has 175 billion parameters, and to compete, Google came up with PaLM, scaling it up to 540 billion. Garry Kasparov, the famous chess champion to compete against IBM’s supercomputer in 1997, said, “As one Google Translate engineer put it, ‘when you go from 10,000 training examples to 10 billion training examples, it all starts to work’. Data trumps everything.” Ever since, it was seen that as the size of the models increased, so did the performance, but at the cost of compute. It has taken another turn in recent times. Last month, Sam Altman, the creator of ChatGPT, said, “I think we’re at the end of the era where it’s gonna be these giant models, and we’ll make them better in other ways.” He added that there has been too much focus on the count of parameters of the language models and the focus should be shifted towards making the models perform better even if that means decreasing their size. At the same time, recently Altman also said that making models bigger is not a bad idea. He iterated that OpenAI can make models a million times bigger than what they already have that would also increase performance, but there is no point in doing it as it might not be sustainable. The smaller, the better To quote Socrates – “It is not the size of a thing, but the quality that truly matters. For it is in the nature of substance, not its volume, that true value is found.” Can we say the same about these LLM models? If we compare GPT-3 and PaLM’s capabilities, the difference is not huge. Given the hype, one can even say the GPT-3 is even better than PaLM. Increasingly, we are also witnessing even smaller language models like LLaMa, with only 65 billion parameters in its largest size. Meta also released models with only 7 billion parameters as well which have been performing way better than their larger counterparts in many use cases. Moreover, to dethrone LLaMa, Technology Innovation Institute has released Falcon, an open source alternative which also has a special licence for allowing commercial use, which LLaMa does not allow. The model has 40 billion parameters and is already sitting on top of the Open LLM Leaderboard on Hugging Face. The researchers say that Falcon outperformed LLaMa, StableLM, and MPT on various benchmarks. Meta did not stop at just LLaMa. LIMA, the new model from Meta AI is built on top of LLaMa 65B and outperformed GPT-4 and Bard in various performance tests. Interestingly, according to the paper, the model was able to perform exceedingly well even with the 7 billion parameter version of LLaMa with just 1,000 carefully curated prompts and responses. Clearly, as the paper said, less is more for alignment. There is another algorithm in the LLM town which is making even smaller size models outperform much bigger models. MIT CSAIL researchers self-trained a 350 million parameter entailment model without human generated labels. According to the paper, the model was able to beat supervised language models like GPT-3, LaMDA, and FLAN. The same researchers have devised a technique called SimPLE (Simple Pseudo-Label Editing), a technique for self-training LLM models. The researchers discovered that self-training could improve a model’s performance by teaching the model to learn through its own predictions. With SimPLE, the researchers were able to take this even another step further by reviewing and modifying pseudo-labels in the initial round of training. “While the field of LLMs is undergoing rapid and dramatic changes, this research shows that it is possible to produce relatively compact language models that perform very well on benchmark understanding tasks compared to their peers of roughly the same size, or even much larger language models,” James Glass, principal investigator and co-author of the paper. Developers for the win “This has potential to reshape the landscape of AI and machine learning, providing a more scalable, trustworthy, and cost-effective solution to language modelling,” said Hongyin Luo, the lead author of the entailment paper. “By proving that smaller models can perform at the same level as larger ones for language understanding, this work paves the way for more sustainable and privacy-preserving AI technologies.” It is clear that with the rapid development in generative AI technology, smaller models are able to perform the same tasks as larger ones. It all started with the open source LLaMa model through which developers could research and build better AI models on their own systems. Now, the field and options are getting bigger instead of the models. There has been a push for building up the open source developer ecosystem in the AI landscape. Meta pushed LLaMa, Microsoft calls everyone a developer now, and Google thinks open source is the real winner in the AI race. With these open source smaller models that do not require heavy computation resources, the generative AI landscape will get even more democratised. This is what the goal looks like: everyone should be able to build their own ChatGPT and run it on their devices. For that, we need smaller, open source, and more efficient models. With such small models outperforming larger ones, it would soon be possible to run GPT-like models on single devices without the internet. The future would be just how Yann LeCun, Meta AI chief, visioned it – multiple smaller models working together for better performance, calling it the world model. This is what Altman predicts and wishes as well. We are headed in the right direction.","excerpt":"The debate about making language models larger or smaller never ends. It probably should now.","categories":["AI Features"],"tags":["ai chatbots","AI Models","ChatGPT","Generative AI","GPT-3","Meta AI","MIT CSAIL","MIT researchers","Open Source AI","Sam Altman"],"author_name":"Mohit Pandey","publish_date":"2023-06-01T19:30:00","publication_year":"2023","word_count":953,"keywords":["GPT-3","ai chatbots","ChatGPT","AI Models","Meta AI","MIT researchers","Sam Altman","AI","machine learning","OpenAI","Hugging Face","Go","MIT CSAIL","Open Source AI","generative AI","Rust","Generative AI","R"],"extracted_tech_keywords":["AI","machine learning","generative AI","ChatGPT","OpenAI","Meta AI","Hugging Face","R","Go","Rust"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/bigger-is-not-always-better\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10011090,"title":"Intel To Purchase SigOpt, An AI Software Optimisation Platform","content":"Intel has announced that the company is planning to acquire a San Francisco-based software optimisation startup — SigOpt. The company further stated that the terms of the deal are expected to close this quarter, however, weren’t disclosed to the media. According to the company’s statement to the media, Intel is planning to leverage SigOpt’s technologies across its products to accelerate, amplify, and scale AI software tools for developers. The company further stated that SigOpt’s software technologies would be combined with Intel hardware products to gain a competitive advantage and provide differentiated value for data scientists and developers. Considering Intel was undergoing a few setbacks, this acquisition is a response to the same. With the delay of Intel’s 10-nanometer processor hardware, the company announced the push back of its 7-nanometer manufacturing process. As a matter of fact, earlier this year, the company broke off with its Nervana AI model training technology after investing $2 billion acquiring chipmaker Habana Labs. The new acquisition of SigOpt will be aimed at focussing on the software side of artificial intelligence and machine learning. Understand the company better — SigOpt was designed to make experts more efficient. Co-founded by Scott Clark CTO Patrick Hayes in 2014, after completing his PhD at Cornell, developed the framework to optimise ML models and A\/B tests at Yelp. SigOpt optimises hyperparameters to generate high-performing models in production. The platform and its API supports 100 hyperparameters and 100 times parallelism, incorporating a range of features like multimetric optimisation, multitask optimisation and conditional parameters — and leveraging Bayesian and global optimisation algorithms. Only with a few lines of code, businesses can embed SigOpt into workflows. This could be possible regardless of the type of AI platform, model management solution, cloud infrastructure, client library, or model type. The company also can create up to tens of thousands of concurrent experiments, like parametise models and run optimisation loops, tracking experiment metadata and providing visualisations in a web insights dashboard. SigOpt has claimed to have worked with many algorithmic traders, government organisations and academic users like MIT, the University of California, Berkeley, Carnegie Mellon, and Stanford; and has raised a total of $8.7 million in venture capital from many investors. According to Intel‘s statement, SigOpt’s software technologies will complement Intel’s existing AI software portfolio, which includes OpenVINO, a toolkit released in 2018 to help developers accelerate AI inference on edge and cloud devices.","excerpt":"Intel has announced that the company is planning to acquire a San Francisco-based software optimisation startup — SigOpt. The company further stated that the terms of the deal are expected to close this quarter, however, weren’t disclosed to the media. According to the company’s statement to the media, Intel is planning to leverage SigOpt’s technologies […]","categories":["AI News"],"tags":["Intel"],"author_name":"Sejuti Das","publish_date":"2020-11-02T20:11:24","publication_year":"2020","word_count":398,"keywords":["Go","API","machine learning","artificial intelligence","AI","ML","RAG","Aim","GAN","R","Intel"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","Aim","RAG","R","Go","API","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/intel-to-purchase-sigopt-an-ai-software-optimisation-platform\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10058925,"title":"What is Uplift modelling and how can it be done with CausalML?","content":"Uplift modelling is a predictive modelling technique that uses machine learning models to estimate the treatment’s incremental effect at the user level. It’s frequently used for personalizing product offerings, as well as targeting promotions and advertisements. In the context of causal inference, in this article, we will discuss the uplift modelling, its types of modelling and lastly, we will see how a Python-based package called CausalML can be used to address the causal inference. Following are the major points to be discussed in this article. Table of contents What is Uplift modelling?Types of modellingCommon applicationsHow does CausalML do Uplifting?Features of CausalML Let’s start the discussion by understanding the uplift modelling. What is Uplift modelling? Uplift modelling is a predictive modelling technique that predicts the incremental influence of a therapy (such as a direct marketing campaign) on a person’s behaviour. Uplift modelling uses a randomized scientific control to test the efficacy of an intervention as well as build a predictive model that projects the incremental response to the activity. The response could be discrete (for example, a website visit) or continuous (for example, a phone call) (for example, customer revenue). Uplift modelling is a data mining approach that has mostly been utilized for up-sell, cross-sell, churn, and retention operations in the financial services, telecoms, and retail direct marketing industries. Machine learning is used to answer the question, “How likely is the consumer to purchase in the future?” in the propensity to purchase model, which essentially explains a customer’s behaviour toward a given action. This is improved by uplift modelling, which addresses more urgent issues: Did the buyer purchase from me because of my ad?Did I waste money by promoting to customers who had already decided to buy?Has my marketing had a negative impact (negative impact) on the likelihood of someone purchasing? To put it another way, a traditional propensity model (as well as most machine learning algorithms) predicts the outcome (y) based on a set of variables (x). Given certain variables, uplift seeks to determine the influence of therapy (t) on the target (y). The term “uplift” refers to the increased chance of the outcome with the treatment as compared to the outcome without the treatment. We cannot directly perceive this difference or causal effect but must deduce it through an experiment. As indicated in the picture below, it is highly beneficial to visualize a 2 x 2 matrix with four kinds of persons (say) to be classed as (a) Persuadable, (b) Sure Thing, (c) Do-Not-Disturb, and (d) Lost Cause. Uplift model matrix We target the “a” demographic, those who are Persuadable, in order to encourage the desired reaction. The treatment is either ineffective or wasteful for everyone else, including the Do-Not-Disturbs. It’s risky to “wake a sleeping dog” by contacting the Do-Not-Disturbs. Finding persuadable is the goal of uplift modelling. Of course, uplift modeling may be used to model any projected outcome, human or not, such as the influence of fertilizer on crop yields or the sending of political campaign emails. Uplift modelling focuses on the effectiveness of the treatment, whereas typical predictive modelling focuses on the result. Then you may focus your efforts on the cases that are most likely to benefit you. Types of modelling Direct modelling and indirect modelling are the two basic approaches here. The fundamental difference between the two approaches is how uplift models are measured and evaluated. Direct modelling In direct modelling, we “directly” model the difference in probabilities between two distinct groups. There are numerous approaches to this, almost all of which rely on tree-based algorithms that have been slightly modified to accommodate uplift modelling. Tree-based models are ideal because they naturally model at the group level by iteratively splitting a group into two groups with each splitting decision. Unlike traditional tree-based models, which are designed to divide data into smaller and smaller homogeneous groups, uplift models are designed to divide our customers into heterogeneous groups each time they split (by maximizing a measure of uplift). They employ various splitting criteria, such as Kullback-Leibler divergence, Euclidean Distance, p-value, and Chi-squared Distance. Hundreds of trees would be fitted in an ensemble fashion, similar to traditional tree-based methods. Indirect modelling Regular response models are repurposed to infer uplift using indirect uplift modelling techniques (meta-learners), which can be based on any base algorithm. We are modelling the expected value of the response for different treatments, rather than attempting to optimize some measure of uplift directly. For our Direct Mail campaign, we would calculate the probability that the customer will use their credit card if the DM is sent, and the probability that they will use their product if the DM is not sent. The estimated uplift is the difference between the two estimated probabilities. In practice, this can be a two-model approach (a separate model fitted to all control\/treatment groups) or a unified model (a single model with the allocated treatment part of the feature space). Common applications Below are some of the possible applications that are briefed about various industries. Uplift modelling can help understand how treatments may impact certain groups differently than simply comparing the results to the entire treatment group vs control group. Also, how much do these effects differ? A company wants to save customers who are about to churn by reaching out to them. When reaching out to customers, the company wants to avoid upsetting them further by focusing only on high-risk savable customers. A corporation wants to execute a cross-sell campaign, but they don’t want to cross-sell to everyone because resources are limited and some customers may not need or want the other products. A corporation has a lead database, but it generates more leads than it can work, and many of the leads are a waste of time – agents may now work leads in any order they like. How does CausalML do Uplifting? CausalML is a Python module that provides a suite of uplift modelling and causal inference tools that are based on cutting-edge research and machine learning algorithms. Traditional causal analysis approaches, such as performing t-tests on randomized trials (A\/B testing), can estimate the Average Treatment Effect (ATE) of a treatment or intervention. However, in many applications, estimating these impacts at a finer scale is frequently desirable and useful. CausalML can be used by the end-user to estimate the Conditional Average Treatment Effect (CATE), which is the effect at the individual or segment level. Such estimates can enable a wide range of applications for personalization and optimization by applying different treatments to various users. Uplift modelling is a crucial modeling approach made possible by CausalML. Uplift modeling is a causal learning approach for estimating an experiment’s individual treatment effect. Using experimental data, the end-user can calculate the incremental impact of a treatment (such as a direct marketing action) on an individual’s behaviour. For example, if a corporation is deciding between many product lines to up-sell \/ cross-sell to its clients, CausalML can be used as a recommendation engine to identify products that yield the maximum expected lift for each given user. CausalML provides a consistent API for running uplift algorithms, making it as simple as fitting a standard classification or regression model. The included metrics and visualization functions, such as uplift curves, can be used to assess model performance. The first version of CausalML includes eight state-of-the-art uplift modelling algorithms (shown in the below figure). Source Features of CausalML Targeting optimization, engagement personalization, and causal impact analysis are just a few of the use cases for CausalMLs. Targeting optimization We may use CausalML to target promotions to the people who will bring the most value to the company. For example, we can provide promotions to consumers who are more likely to utilize a new product as a result of exposure to the promotions in a cross-sell marketing campaign for existing customers. Causal impact analysis We can also use CausalML to examine the causal impact of a specific event using experimental or observational data with rich attributes. For example, we can examine how a customer cross-sells event influences long-term platform expenditure. Personalization To personalize engagement, CausalML can be employed. A corporation can communicate with its customers in a variety of ways, such as offering up-sell options or using message channels for interactions. CausalML can be used to assess the effect of each combination for each client and present customers with the most customized offers possible. Final words Through this article, we have discussed model uplifting which is basically a technique that models the user’s behaviour by applying some intervention along with input variables. We have also discussed its major types of modelling and some applications where it can be applied. In this article, we have discussed a python package called CausalML which gives a path to practically implement causal inference or uplifting. For more understanding of implementing the uplift modelling with CausalML, you can refer to their GitHub repository where they have listed many implementation examples. References Uplift modellingCausalML research paperCausalML repository","excerpt":"In this article, we will discuss the uplift modelling, its types of modelling, and Python-based package called CausalML can be used to address the causal inference.","categories":["AI Trends"],"tags":["causal inference","Machine Learning","predictive modelling"],"author_name":"Vijaysinh Lendave","publish_date":"2022-01-22T18:00:00","publication_year":"2022","word_count":1494,"keywords":["Go","API","machine learning","AI","ML","Machine Learning","Git","RAG","Python","causal inference","predictive modelling","GitHub","R"],"extracted_tech_keywords":["AI","machine learning","ML","RAG","Python","R","Go","Git","GitHub","API"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/what-is-uplift-modelling-and-how-can-it-be-done-with-causalml\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10092910,"title":"A Striking Relevance of Data Sketching in LLMs","content":"At Data Engineering Summit (DES) 2023, presenting a talk about Unleashing the Power of Probabilistic Data Structures: Optimizing Storage and Performance for Big Data, Sudarshan Pakrashi, Director of Data Engineering at Zeotap spoke about statistical algorithms designed to optimise use of memory in storing and querying large datasets. In one of the questions asked, he spoke about if data sketching can be used in current generative AI models such as LLMs. To this, Pakrashi responded that it is possible to do so and is actually a “great analogy”. He explained how in every language model there are word associations that need to be maintained when there is a huge dataset of words. “Imagine the permutations and combinations that you want to have and sketches are in fact used to maintain those because then your model is actually going to query the frequencies for those combinations,” he explained. Data sketching is a method of summarising large datasets using compact data structures that can provide approximate answers to queries about the data. In the context of LLMs, data sketching can be used to summarise the text corpus used to train the model, which can help reduce the memory requirements of the model and improve its training efficiency. Why Do We Need Data Sketching in LLMs “Do you ever feel overwhelmed by a constant stream of information?” – reads the first line of the paper – ‘What is Data Sketching, and Why Should I Care’ by Graham Cormode in 2017. When you start filtering out information based on what is required, that is exactly similar to what data sketching essentially is. And this can greatly benefit the training of LLMs. Data sketching can help improve the efficiency and scalability of generative models in the following ways: Data compression: Summarising large datasets using sketching techniques, LLMs can be trained on representations that are smaller than the original one, reducing the memory requirements and computational resources needed for deploying and training. This can be especially helpful when dealing with limited resources or large-scale datasets. Faster training: Speed up the training process for LLMs by reducing the amount of data they need to process. Data sketching, by essentially reducing the size of the data, can lead to faster convergence and shorter training times without significantly compromising the quality of the generated samples. Real-time data: Sketching can enable LLMs to process and learn from data streams efficiently, and automatically update their internal representations on-the-fly by generating new samples based on the most recent data. Anomaly detection: For identifying outliers or anomalies in the training data, sketching and sampling techniques can be used to improve the quality of LLMs outputs. By identifying and potentially removing anomalous data points, LLMs can focus on learning the underlying structure and patterns in the data, leading to better generated samples. Data exploration: For exploring large datasets to gain insights into their structure and characteristics, sketching can provide insights that can be used to guide the design and configuration of LLMs, such as selecting appropriate architectures, hyperparameters, or loss functions. Data Sketching Techniques in LLMs Although data sketching has been used previously in natural language processing (NLP) tasks, the recent rise of LLMs like GPT-3, that require huge compute, data sketching can, and in fact have, increased the efficient training and deployment of AI models. One commonly used data sketching technique in LLMs is the Bloom filter, which is a probabilistic data structure that can efficiently test whether an item is in a set. Bloom filters can be used to represent the vocabulary of the text corpus used to train the model, allowing the model to store the vocabulary in a much smaller memory footprint. Another technique used in LLMs for data sketching is Count-min sketches. This is another type of probabilistic data structure that can efficiently estimate the frequency of items in a set. Count-min sketches can be used to estimate the frequency of words in the text corpus, which can be used to optimise the training of the model. Similar techniques include HyperLogLog, which is another probabilistic algorithm used to estimate the number of distinct elements in a large dataset. Moreover, Quantiles sketches is another technique that provides approximate answers to queries about percentiles, medians, or other order statistics of a dataset. This is similar to Sampling, which means selecting a subset of data for representing the entire dataset.","excerpt":"Can data sketching be used in large language models for building chatbots?","categories":["AI Features"],"tags":["Generative AI"],"author_name":"Mohit Pandey","publish_date":"2023-05-07T10:00:00","publication_year":"2023","word_count":728,"keywords":["big data","Go","TPU","AI","Scala","RAG","NLP","anomaly detection","generative AI","Generative AI","R"],"extracted_tech_keywords":["AI","NLP","generative AI","RAG","anomaly detection","TPU","R","Go","Scala","big data"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/a-striking-relevance-of-data-sketching-in-llms\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10102566,"title":"Why are Big-Tech Employees Quitting?","content":"Many talented folks are leaving their big-tech dream jobs to cash in on the AI wave, building companies that are now worth billions, and that too, in less than a year! “The founders of OpenAI, Cohere, and Anthropic left Google to build what that company wouldn’t,” said Vin Vashishta, the founder of V-Square, said. “Inside of every great company is the company that replaces it,” he noted. The same was true for Inflection, Character.AI, and even the recent Mistral, where the founders left Google DeepMind, Meta, and OpenAI, to build their own startups. Kai-Fu Lee, a computer scientist who has previously worked at Google, Microsoft, and Apple, has also started 01.AI, his own AI startup. Now, the eight-month-old Chinese startup has reached a valuation of $1 billion, and became one of the fastest unicorns, joining the list of Avant, iCarbonX, NuCom Group, which became unicorns in less than six months. The latest funding round included Alibaba Group Holding Ltd’s cloud unit as well. Interestingly, 01.AI has open sourced its foundational LLM called Yi-34B, which outperforms Llama 2 on various key metrics. Lee said that all he wanted to do was provide an alternative to Meta’s Llama 2, which has been “the gold standard and a big contribution to the open-source community”. AI Hustle Culture A Carnegie Mellon University graduate, Lee is also the CEO of Sinovation Ventures. At 01.AI, he has built a team of more than 100 people, which includes his former colleagues from US companies, and other Chinese nationals working overseas. He highlighted that his team not only includes AI specialists, but business experts as well. Lee mentioned that he has received inquiries from the limited partners of his venture firm regarding how he will manage his dual CEO roles. He emphasised that if he dedicates 40-hours a week to Sinovation, he still has an additional 128 hours per week at his disposal. “I have 86 more hours to allocate to 01.AI without neglecting my Sinovation responsibilities,”  he added, suggesting that he might allocate six hours each day for sleep and other aspects of his life. This sounds in line with Narayana Murthy’s vision of developed countries, and its founders, who are working 70+ hours a week. Lee’s involvement in AI spans several decades. In his 1982 application to graduate school at Carnegie Mellon, he expressed his desire to commit his life to AI research, believing that this technology would facilitate a deeper understanding of humanity. “Necessity is the mother of innovation, and there’s clearly huge necessity in China,” said Lee. He highlighted that OpenAI and Google’s models are not available in China, and that is driving innovators in the country to build their own models. This also includes big companies such as Baidu. Even then, Lee mentioned, “Our proprietary model will be benchmarked against GPT-4,” in reference to OpenAI’s LLM. Creating jobs Apart from voluntarily leaving jobs for their own startups, big-tech employees are also being laid off at a rapid pace. When not starting their own companies, these skilled employees are joining these startups. Post-Covid, the high attrition and massive layoffs could be attributed to the shift from WFH to office, among many other reasons, but now the case is more about building AI startups. According to Vinod Khosla, the businessman and VC, a lot of former Google, Meta, and Microsoft employees have been approaching him for funding, and a lot of them are mostly companies that are focused on AI. He believes that it is because they have a lot of passion to build what they want, and the big-tech is not delivering it. “Bad times for big tech is a great time for startups that you’ll hear about five years from now,” Khosla said. Though big-tech is also allocating most of its fund for AI projects, employees want to get to it faster, and also get a bigger payout after they succeed, not just mere salaries. Of course, the journey from being a big-tech employee to a billionaire, or a millionaire is not without challenges. And a lot of these startups are still on the way. But they seem to be only headed upwards and onwards. Interestingly, a lot of these AI startups are also employing the people who are being laid off from the big-tech. That also explains why big tech is regulating open source, but that conversation is for another day.","excerpt":"“Inside of every great company is the company that replaces it.”","categories":["AI Features"],"tags":["generative AI unicorns","Startups"],"author_name":"Mohit Pandey","publish_date":"2023-11-06T16:18:53","publication_year":"2023","word_count":729,"keywords":["Anthropic","Go","API","OpenAI","AI","innovation","generative AI unicorns","GPT","Ray","Startups","R","startup"],"extracted_tech_keywords":["AI","OpenAI","Anthropic","Ray","R","Go","API","GPT","innovation","startup"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/why-are-big-tech-employees-quitting\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10060439,"title":"Top music generation datasets in 2022","content":"Artificial intelligence has been tapped for synthetic music generation for some time now. However, the watershed moment came when music informatics met AI. Now, music AI researchers are taking advantage of the latest AI, ML, and analytics developments to develop models to create music on par with human composers. The strength of any AI\/ML model is predicated on the data it’s fed. Below, we look at the major music generation datasets doing the rounds in 2022. NSynth The dataset probably contains the largest instrumental notes at 305,979 musical notes, including unique pitch, timbre and envelope. The musical notes were collected from 1,006 instruments from commercial sample libraries and are annotated based on (acoustic, electronic or synthetic) instrument family and sonic qualities. Instruments including bass, flute guitar, keyboard, mallet, organ, reed, string, synth lead and vocal have been used in this dataset. MAESTRO The MAESTRO (MIDI and Audio Edited for Synchronous Tracks and Organisation) dataset has over 200 hours of paired audio and MIDI recordings of International Piano-e-Competition in the past ten years. The MIDI data has key strike velocities, including sustain\/sostenuto\/una corda pedal positions. URMP (University of Rochester Multi-Modal Musical Performance) URMP introduced a dataset for facilitating audio-visual analysis of musical performances. The dataset contains several simple multi-instrument musical pieces assembled from separately recorded performances of individual tracks. For each piece, a musical score is provided in MIDI format. Bach Doodle Dataset The dataset consists of 21.6 million harmonisations submitted from the Bach Doodle and metadata about the composition, like country of origin and the feedback. It also has MIDI of user-entered melody and MIDI of the generated harmonisation. An exploration of the melodies in the dataset contains top repeated melodies from each country or the regional hits. The Lakh MIDI Dataset v0.1 This dataset contains 176,581 unique MIDI files, 45,129 matched and aligned to the Million Song Dataset entries. The dataset mainly facilitates large-scale music information retrieval, both symbolic and audio content-based. Music 21 Music21 contains music performances from 21 categories. It is a set of tools to help scholars quickly find answers to questions like, “I wonder how often Bach does that” or “Which band used these chords for the first time? Or “How to know more about Renaissance counterpoint or the Indian ragas or post-tonal pitch structures or the form of minutes”. Datasets for Indian Music CompMusic catalogues datasets for Indian art music. The website aims to advance the automatic description of music by emphasising cultural specificity carrying research in music information processing with a domain knowledge approach. The project mainly focused on five music traditions of the world: Hindustani (North India), Carnatic (South India), Turkish-makam (Turkey), Arab-Andalusian (Maghreb), and Beijing Opera (China). Indian Music Tonic Dataset The dataset contains 597 commercially available audio music recordings of Indian art music, both Hindustani and Carnatic music, where each recording is manually annotated with the tonic of the lead artist. http:\/\/compmusic.upf.edu\/iam-tonic-dataset Carnatic Varnam Dataset The dataset has 28 solo vocal recordings recorded to be researched on the intonation analysis of Carnatic ragas. In addition, the dataset contains audio recordings, time aligned tala cycle annotations and swara notations in a machine-readable format. http:\/\/compmusic.upf.edu\/carnatic-varnam-dataset Carnatic Music Rhythm Dataset This dataset is a sub-collection of 176 excerpts in four taalas of Carnatic music with audio, tala related metadata and time aligned markers to indicate the progress through the tala cycles. http:\/\/compmusic.upf.edu\/carnatic-rhythm-dataset Hindustani Music Rhythm Dataset The dataset is a sub-collection of 151 in four taals of Hindustani music which includes audio, taal related metadata and time aligned markers to indicate the progress through the taal cycles. http:\/\/compmusic.upf.edu\/hindustani-rhythm-dataset Mridangam Stroke Dataset The dataset contains 7,162 audio examples of individual strokes of the Mridangam in various tonics and ten different strokes played on Mridangams with six tonic values. http:\/\/compmusic.upf.edu\/mridangam-stroke-dataset Mridangam Tani-avarthanam Dataset The dataset is a transcribed collection of two tani-avarthanams played by Mridangam maestro Padmavibhushan Umayalpuram K. Sivaraman. The audio of the dataset was recorded at the IIT Madras and annotated by Carnatic percussionists. The dataset contains 24 minute of audio and 8,800 strokes. http:\/\/compmusic.upf.edu\/mridangam-tani-dataset Saraga: Research datasets of Indian Art Music The repository contains time aligned melody, rhythm and structural annotations for two large open datasets of Indian Art Music (Carnatic and Hindustani music). https:\/\/zenodo.org\/record\/4301737#.YTddzcaxX0o Tabla Solo Dataset The dataset is a transcribed collection of Tabla solo audio recordings spanning compositions from six different Gharanas of Tabla, played by Pt. Arvind Mulgaonkar. It consists of audio and time aligned bol transcriptions. http:\/\/compmusic.upf.edu\/tabla-solo-dataset","excerpt":"CompMusic catalogues datasets for Indian art music.","categories":["AI Trends"],"tags":["audio","Datasets"],"author_name":"Poornima Nataraj","publish_date":"2022-02-11T18:00:00","publication_year":"2022","word_count":740,"keywords":["Go","Datasets","artificial intelligence","AI","Modal","ML","audio","RAG","Aim","analytics","GAN","R"],"extracted_tech_keywords":["AI","artificial intelligence","ML","analytics","Aim","RAG","R","Go","GAN","Modal"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-music-generation-datasets-in-2022\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":50752,"title":"The AI Behind Instagram Explore","content":"One of the most popular social networks, Instagram has witnessed exponential growth over the past few years. Last year, the social media network reached one billion monthly active users, and it is projected to surpass 111 million in 2019. According to the reports, over half of the Instagram community visits Instagram Explore every month to discover new photos, videos, and stories relevant to their interests. The Instagram Explore is a recommendation engine, which recommends the most relevant content to users in real-time with the help of emerging technologies like artificial intelligence and machine learning. Recently, the researchers at Facebook AI research unveiled the novel engineering solutions and a detailed overview of the key elements that make Instagram Explore work effectively. How It Works As mentioned, Instagram Explore recommends the most relevant content out of billions of options in real-time, which introduces a number of machine learning challenges to the researchers and these problems are tackled by creating a series of custom query languages that are mostly lightweight modelling techniques and tools that enable high-velocity experimentation. These techniques and tools constitute an AI system that extracts 65 billion features and makes 90 million predictions per second. While developing the recommender engine, the researchers addressed three important needs, and they are mentioned below: The ability to conduct rapid experimentation at scale.To obtain a stronger signal on the breadth of people’s interestsA computationally efficient way to ensure that the recommendations are both high quality and fresh. Foundational Tools In order to address those needs, the researchers developed foundational tools that are mentioned below: IGQL The researchers built a domain-specific language optimised for retrieving candidates in recommender systems known as IGQL. It is a custom domain-specific meta-language that provides the right level of abstraction and assembles all algorithms into one place. Basically, IGQL identifies the most relevant accounts based on individual interests. IGQL is both statically validated and high-level. The execution of this customised language is optimised in C++, which helps the language minimise both latency and compute resources. It lets engineers focus on ML and business logic behind recommendations to provide a high degree of code reusability. Ig2vec Ig2vec is a word2vec-like embedding framework which is used to conclude account embeddings. Account embeddings help to identify topically similar accounts efficiently. The ig2vec embedding framework works by treating account IDs that a user interacts with. And with the user’s interaction, the accounts can be predicted with which a user is likely to interact in a particular session within the Instagram app. In another case, if a user interacts with a sequence of accounts in the same session, a distance metric between two accounts is then defined which is usually cosine distance or dot product. Based on this, a K Nearest Neighbour (KNN) is implied in order to find the topically similar accounts for any account in the embedding. Ranking Distillation Model The researchers introduced a ranking distillation model to help preselecting candidates before implementing the more complex ranking models. The approach works by training a super-lightweight model that learns from and tries to approximate the main ranking models. The distillation model is then trained on this recorded data with a limited set of features and a simpler neural network model structure to replicate results. Wrapping Up After implementing all these three tools, researchers split the Explore recommendation system into two main stages, which are the candidate generation stage or the sourcing stage and the ranking stage. According to the researchers, the ongoing ML challenge encountered by the researchers is to find new and exciting ways to help the Instagram community discover the most interesting and relevant content on the social media platform.","excerpt":"One of the most popular social networks, Instagram has witnessed exponential growth over the past few years. Last year, the social media network reached one billion monthly active users, and it is projected to surpass 111 million in 2019.  According to the reports, over half of the Instagram community visits Instagram Explore every month to […]","categories":["AI Features"],"tags":["instagram"],"author_name":"Ambika Choudhury","publish_date":"2019-11-27T10:56:09","publication_year":"2019","word_count":608,"keywords":["Go","API","machine learning","artificial intelligence","AI","neural network","RPA","ML","C++","instagram","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","neural network","R","Go","C++","API","RPA"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/the-ai-behind-instagram-explore\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":64780,"title":"Udacity To Launch AWS Machine Learning Scholarship Program","content":"Udacity and AWS collaborated to offer scholarships called AWS Machine Learning Scholarship Program for data science enthusiasts who want to upskill. The scholarship has two phases and is open for aspirants above 18 years of age. The enrollment for the course will be opened from May 19 till July 31. Besides, the scholarship is divided into two phases: foundational and a full scholarship for a Udacity Nanodegree program. In the first phase, students will have to take the AWS Machine Learning foundational course, where you will learn best practices of object-oriented programming and get familiar with AWS services such as DeepLens, DeepRacer and DeepComposer. As per Udacity’s recommendation, one will have to dedicate 5 to 10 hours a week, at least, to complete phase one and get into the 2nd phase. However, only 325 seats are available for the second phase. Consequently, you will have to not only focus on completing the foundational course but also involve in other activities that Udacity evaluates on to select learners for the second phase. These metrics are usually provided through announcements during the first phase on its Slack workspace, which is specially created for the scholarship. On being selected for the second phase, which will start on August 19, you will be automatically enrolled in the AWS Machine Learning Engineer nanodegree program. While you will get a certificate for completing the phase one, irrespective of whether you got into the 2nd phase, the most important certificate that you will receive is after completing the advanced course in the second phase. In the second phase, you will learn more about the AWS ML services and practice how to deploy your models to a production environment. “Machine learning is one of the crucial components of the ongoing AI revolution driving digital transformation worldwide. Thus, we are contributing to the community by offering this scholarship to aspirants,” said Gabriel Dalporto, CEO of Udacity.","excerpt":"Udacity and AWS collaborated to offer scholarships called AWS Machine Learning Scholarship Program for data science enthusiasts who want to upskill. The scholarship has two phases and is open for aspirants above 18 years of age. The enrollment for the course will be opened from May 19 till July 31. Besides, the scholarship is divided into two […]","categories":["AI News"],"tags":["AWS"],"author_name":"Rohit Yadav","publish_date":"2020-05-07T18:37:23","publication_year":"2020","word_count":317,"keywords":["data science","Go","machine learning","AWS","AI","ML","digital transformation","Git","ViT","R"],"extracted_tech_keywords":["AI","machine learning","ML","data science","AWS","R","Go","Git","ViT","digital transformation"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/udacity-to-launch-aws-machine-learning-scholarship-program\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10112239,"title":"AI is the Future, and the Future of AI is India","content":"The sentiment was aptly captured by one of India’s seasoned journalists in her recent episode of ‘Vantage with Palki Sharma’, in the backdrop of the latest high-profile visits by Google and Microsoft executives to the country. Microsoft chief Satya Nadella is currently on his trip to India, and plans to engage with promising AI startups. Nadella has high hopes for the country in terms of its potential contribution to the development of AI. “If the Indian economy is going to be, let’s say, 5 trillion, the AI-driven part of it could be something like, maybe, 10% of it—perhaps 500 billion,” he said. A week ago, Google hosted its first-ever Research@ Bangalore event, bringing together leading researchers, developers, and startups. Dr Jeff Dean, chief scientist of Google DeepMind and Google Research, attended the event in person, along with several other Google leaders, to discuss the future of AI and its implications in India. “India is well-positioned because you have an incredibly strong culture of respecting engineering, science, and computer science. Many amazing computer scientists have come from India and are studying computer science,” he said. “The young students that I meet in India are amazing. They’re all excited and eager to understand the shift from traditional computer science to learning-based approaches for solving all kinds of problems,” added Dean. While AI startups in the West are looking for young talent, India has no dearth of it. Recently, many local foundational models have popped up in India, and interestingly, they are mostly created by young individuals. For instance, Tamil Llama by Abhinand Balachandran and Kannada Llama by Adarsh Shirawalmath, who is a second-year BTech student at Vellore Institute of Technology. Similarly, Telugu Llama was created by Ravi Theja, who is under 30. Moreover, many new startups are emerging in India, with a focus on creating bilingual LLMs catering to the Indian market. One such startup is Sarvam AI, which recently introduced OpenHathi and Microsoft has partnered with them. “I am excited to see innovative startups in India. I had the chance to meet founders Pratyush [Kumar] and Vivek [Raghavan] from Sarvam AI. Pratyush previously worked at Microsoft Research. We are excited to support them and build out their LLM, which is trained in all the Indic languages. The demo they showed me is tremendous,” said Nadella during his visit to Bengaluru. Microsoft is also backing another AI startup BrainSightAI which uses AI and machine learning to map the human brain, which aids thousands of clinicians, including neurosurgeons, psychiatrists, and neurologists, across the country. Lately, Indian startups are experiencing a funding surge. India witnessed its first AI Unicorn, Krutrim AI, which raised $50 million in funding from Matrix Partners. Meanwhile, Google is looking to invest approximately $4 million in the Indian conversational AI startup CoRover. Interestingly, nearly 84% of Indian CEOs are raising new capital or reallocating budgets to invest in generative AI, compared to 70% globally. The Rise of AI Engineers Indian IT giants have been making every possible effort to train their employees in generative AI. TCS recently announced that it has trained over 1 lakh-strong pool of GenAI ready consultants and prompt engineers who are engaged in hundreds of GenAI projects for clients across segments. Similarly, last year, Infosys announced its plans to train over 1 lakh employees in partnership with NVIDIA & Google Cloud. Wipro, on the other hand, has globally trained 210,000 employees in AI and is integrating generative AI across its platforms. HCLTech has internally trained close to 42,000 people across various areas of generative AI, while Accenture claims to have trained over 600,000 employees in AI. Indian IT firms are seeing a combined ~450 GenAI projects currently in the pipeline. “Software developers, who are already working, would want to pick up these (generative AI) skills if they don’t already have them. This is something that will transform basically every software developer’s workflow and the way they approach problems. This holds true not just in India but all across the world,” said Dean. Similarly, Nadella spoke how about Copilot has increased the productivity of Indian IT employees across Infosys, HCL Tech and MindTree. ‘Unique’ Data One of the primary attractions for big tech companies in India is its wealth of data. During his visit to India, Nadella praised the work being done by Bhashini and Karya. Karya creates datasets in several Indian languages to train AI models and for research, simultaneously generating job opportunities for Indians, particularly in rural areas. On the other hand, Bhashini is Indian government’s initiative to create an open-source Indic language dataset and has also partnered with Karya. AI4Bharat is another initiative from IIT Madras dedicated to developing open-source resources for Indian languages, including datasets, models, and applications. Google and Microsoft, along with other companies, are actively seeking to acquire both datasets and the companies producing them. Their aim is to ensure that the LLMs they develop surpass the ones currently trained on publicly available internet data. “The upcoming wave of mergers and acquisitions might be spearheaded by companies like Google, Microsoft, and Meta (formerly Facebook), evaluating data companies as potential contributors to their extensive language models or other machine learning and AI endeavours,” said Chamath Palihapitiya, Social Capital Founder & CEO. Moreover, India is generating terabytes of enterprise data monthly due to increased 5G penetration, OTT services, and social media usage. With an abundance of data available in India, there is a high likelihood that the next major hyperscaler could emerge from the country. Recently, laptop maker Lenovo announced its plans to start manufacturing servers locally to support its data centre business.","excerpt":"Satya Nadella knows it, Jeff Dean feels it, and now the entire world acknowledges it.","categories":["AI Features"],"tags":[],"author_name":"Siddharth Jindal","publish_date":"2024-02-08T15:21:03","publication_year":"2024","word_count":930,"keywords":["Go","API","GenAI","machine learning","AI","RAG","Aim","ViT","generative AI","R"],"extracted_tech_keywords":["AI","machine learning","generative AI","GenAI","Aim","RAG","R","Go","API","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/ai-is-the-future-and-the-future-of-ai-is-india\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":767,"title":"ISRO to launch GSLV Mark-III in two months","content":"That India is a leader in PSLV is a known fact. And now India is going to add another feather in its cap with the proposed launch of its most powerful rocket, GSLV Mark-III in another two months. According to news reports, GSLV Mark-III is the country’s most powerful launch vehicle, revealed senior space scientist and ex-programme director of ISRO professor T.G.K. Murthy. Speaking at a conference in Kolkata, “Advances in Science and Technology”, Murthy said, “In a couple of months, we are going to launch GSLV Mark-III. We are going to launch four satellites from the Indian soil in the near future”. Of late, ISRO has been successfully testing the high-thrust cryogenic technology that is used in the launch vehicle, GSLV Mark-III, revealed the space scientist. Up next, there is also a satellite study in the pipeline called Aditya-L1. According to the space scientist, India will launch a satellite with one payload to study the sun and monitor global climate change. If you thought ISRO works alone, you are wrong. ISRO in partnership with industry startups and academicians is going to take charge of the global climate change. “India is going to take 10-15% share of the global space market”, and ISRO is having global customers as our research occupy pre-eminence in the world,” said the eminent space scientist.","excerpt":"That India is a leader in PSLV is a known fact. And now India is going to add another feather in its cap with the proposed launch of its most powerful rocket, GSLV Mark-III in another two months. According to news reports, GSLV Mark-III is the country’s most powerful launch vehicle, revealed senior space scientist […]","categories":["AI News"],"tags":["ISRO India"],"author_name":"Richa Bhatia","publish_date":"2017-04-17T04:08:08","publication_year":"2017","word_count":220,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","CuPy","Rust","ISRO India","R","programming_languages:Rust","startup"],"extracted_tech_keywords":["AI","CuPy","R","Go","Rust","startup","programming_languages:R","programming_languages:Go","programming_languages:Rust"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/isro-launch-gslv-mark-iii-two-months\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":37162,"title":"After ‘Bandersnatch’ Netflix Ups Its AI Game With ‘Love Death + Robots’","content":"Image source: bloody-disgusting.com After the success of Bandersnatch, the episode from the critically-acclaimed Netflix series Black Mirror which lets its viewers interact with the character with the help of AI, the platform has rolled out a similar concept. Titled Love, Death + Robots, is an animated anthology that spans multiple genres, exploring futuristic themes. The key here is that Netflix has taken user data, applied artificial intelligence and is shuffling the order of the episodes so that each viewer gets a unique and tailored viewer experience. Unique UX This move by Netflix is seen as a step taken towards customisation and personalisation, which is touted to be the future of video streaming. While in Bandersnatch, by enabling AI, the viewers had the liberty to direct the outcome of the episodes based on the queries that the episodes asked the viewers, whereas in Love, Death + Robots, the makers are testing the waters by showcasing four different episodes to different viewers, and stringing the different episodes to make a coherent series. Referring to the series as a love letter to nerds, the director Tim Miller and producer David Fincher, who previously collaborated for films like Deadpool (2016) and Gone Girl (2014), the duo claim that the new web series will be a feast for nerds right from its animation to the subjects and plot that each episode delves into. Speaking about the new AI-concept unrolling, Fincher said that it frees the viewers from the traditional time limit, “We have to get rid of the 22-minute [length of a half-hour show with commercials] and 48-minute [length of an hour-long show with commercials] because there’s this Pavlovian response to this segmentation that to me seems anathema to storytelling. You want the story to be as long as it needs to be at maximum impact or entertainment value proposition,” Finder said at a recent panel discussion. Netflix Is Collecting Your Data Pointing out that the platform is collecting its user data for driving personalisation, one of the viewers of the story who identifies himself as gay, claimed that the episode suggestion was made by the platform based on his sexual orientation. “Just discovered the most INSANE thing. The ORDER OF THE EPISODES for Netflix’s new series Love Death + Robots changes based on whether Netflix thinks you’re gay or straight,” the user tweeted. The online video streaming platform is known to rely on big data and algorithms to identify user preferences and make a recommendation based on these data. As of 2018, Netflix is believed to have close to 130 million worldwide subscribers, thus giving them access to a plethora of information, providing them with even information such as at what time are the users are likely to lose their interest in the show to what they are likely to watch. According to the latest report, the platform’s recommendation system decides close to 80 per cent of what a person sees. The Controversy However, declining Twitter user’s claim, Netflix said that the episode suggestions has nothing to do with gender, sexuality or race, “We’ve never had a show like Love Death + Robots before so we’re trying something completely new: presenting four different episode orders. The version you’re shown has nothing to do with gender, ethnicity, or sexual identity — info we don’t even have in the first place,” Netflix tweeted. So the response here seems a little disengenious. I assume it takes into account the content you've watched before, which I also assume is tagged with things like gender, ethnicity, and sexuality, which easily leads to the system making implicit assumptions about the user. https:\/\/t.co\/jRglFK0OIN — Lukas Thoms (@LukasThoms) March 19, 2019 Later, the user clarified  that the platform’s suggestions made based on random A\/B test that doesn’t involve any machine learning, “A final update: a friend I trust at Netflix looked into this, and apparently the episode order is just a 100% random A\/B test that doesn’t involve any ML. Identity-based recommendations are still a good discussion to have, in this case, it was just random!” he clarified at the end. The Need For Ethical Algorithms As the talk about the importance of ethical AI takes precedence across the world, at the core, ethicality of algorithms is what will ensure a transparent and clean AI which is removed of any bias. “Algorithms impact whether and how individuals have access to social goods and rights, and how algorithms are developed and implemented within managerial decision making is critical for business ethics to understand and research. We can hold firms responsible for an algorithm’s acts even when the firm claims the algorithm is complicated and difficult to understand,” a researcher notes in his paper titled Ethical Implications and Accountability of Algorithms.","excerpt":"After the success of Bandersnatch, the episode from the critically-acclaimed Netflix series Black Mirror which lets its viewers interact with the character with the help of AI, the platform has rolled out a similar concept. Titled Love, Death + Robots, is an animated anthology that spans multiple genres, exploring futuristic themes. The key here is […]","categories":["AI Features"],"tags":["Robots"],"author_name":"Akshaya Asokan","publish_date":"2019-04-01T12:21:32","publication_year":"2019","word_count":787,"keywords":["big data","Feast","Go","machine learning","artificial intelligence","AI","ML","Aim","Rust","R","Robots"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","Aim","Feast","R","Go","Rust","big data"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/after-bandersnatch-netflix-ups-its-ai-game-with-love-death-robots\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":20887,"title":"India, Japan To Use Artificial Intelligence In Defence Sector","content":"Indian and Japan are bringing in artificial intelligence into their arsenal as a measure of its counter-intelligent defense system. Kentauro Sonoura, minister of foreign affairs for Japan, said in a press release that Japan aims to collaborate with India to develop unmanned ground vehicles (UGV) and robotics. The move is also aimed at strengthening the geo-political ties between the two countries. Although, there was initial rebuke between India and Japan, it seems to have been eased over the years, as to mainly counter China‘s strategic influence over India. Sonoura emphasised on the fact that Japan planned to achieve cordial relations with all the South-Asian countries including Pakistan, Myanmar and Bangladesh in 2018. He also said that Japan wanted to promote a “free and open Indo-Pacific” policy, highlighting the steps taken by them to curb terrorism through initiating a dialogue with Pakistan. Speaking to a national newspaper, Sonoura said, “”We need to share the importance of rule of law and freedom of navigation among related countries. The next step is infrastructure development based on global standards, so that connectivity among countries is increased. The third step would be maritime law enforcement and disaster management that would ensure the stability and prosperity in the Indo-Pacific region. Therefore, we would like to connect and combine our Indo-Pacific strategy and India’s Act East policy as a one big picture. That’s the synergy we seek.” This makes India enter into a new foray of technological innovations with Japan with its friendly relations going strong with the recent developments. In fact, Japan Prime Minister Shinzo Abe, had gone as far as to say that India was a part of its support towards infrastructural development and growth in the IT sector. India, on the other hand, has been on the rise with developments in its Make In India policy. This agreement will certainly strengthen its outlook towards creating more jobs in the country. Furthermore, the economic gain achieved by India will reinforce its status as an aggressive developing country. The move to enter AI and robotics sector might seem like a good idea for establishing a political and technological stability for the country. Contrastingly, there has been criticism from the opposition parties in India for choosing foreign companies over homegrown organisations for high value projects.","excerpt":"Indian and Japan are bringing in artificial intelligence into their arsenal as a measure of its counter-intelligent defense system. Kentauro Sonoura, minister of foreign affairs for Japan, said in a press release that Japan aims to collaborate with India to develop unmanned ground vehicles (UGV) and robotics. The move is also aimed at strengthening the […]","categories":["AI News"],"tags":["AI (Artificial Intelligence)","China","japan","Make in India"],"author_name":"Abhishek Sharma","publish_date":"2018-01-22T09:49:57","publication_year":"2018","word_count":378,"keywords":["japan","Go","Make in India","artificial intelligence","programming_languages:R","AI","innovation","Ray","Aim","ViT","GAN","AI (Artificial Intelligence)","R","China"],"extracted_tech_keywords":["AI","artificial intelligence","Aim","Ray","R","Go","GAN","ViT","innovation","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/india-japan-use-ai-defence-sector\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":54910,"title":"IIT Delhi Startup Creates Buddhi Kit To Make AI Learning A Child’s Play","content":"An IIT-Delhi startup has created a first-of-its-kind interactive DIY education kit based on artificial intelligence (AI). Buddhi AI DIY kit can be used to quickly and easily learn the basics of AI and build AI-based solutions for real-world problems. The kit will also help people without having any prior domain knowledge or training. The idea is to assist young students, tinkerers, makers, innovators, hobbyists, teachers, educationists, artists, parents and professionals from any background. Buddhi (Build, understand, design, deploy human-like intelligence) kit was launched at IIT Delhi. An IIT statement read — “Buddhi kit helps users develop core skills such as problem-solving, creative thinking and ability to work in teams. With the kit, creative possibilities are endless as it can be used to easily introduce AI in any existing STEAM (science technology, engineering, arts & maths) project.” According to IIT-Delhi director, V Ramgopal Rao, “IIT Delhi is taking centre stage in AI-related research and development in the country. Development in indigenised AI hardware and software is critical. The kit is one such initiative by the IIT Delhi-FITT incubated faculty startup. It is a result of the many initiatives we have taken in the institute to encourage faculty members and their research students to turn into entrepreneurs.” The kit was developed by IIT faculty member Manan Suri, who has been recognised by MIT as one of the world’s top 35 innovators under the age of 35. The kit includes AI computing-engine, friendly AI training and inference applications along with real-world AI actuation circuit boards. It has been accompanied by a rich learning ecosystem that includes a high-quality AI handbook, practical DIY AI projects, lessons, exercises, presentations and videos. “The motivation behind the kit is to help young school students learn the practical aspects of a complex topic like AI in a friendly and simple manner,” said Suri. Recently, CBSE had introduced AI as a subject in the school curriculum. “The kit and its accompanying content are perfectly aligned with CBSE’s prescribed AI school syllabus,” Suri added. Buddhi kit, which has been developed by CYRAN AI Solutions, will be provided to select institutional partners such as schools, maker spaces, tinkering-labs, education-related NGOs, government bodies, corporations, CSR bodies, or entities to get involved with AI-education and skilling.","excerpt":"An IIT-Delhi startup has created a first-of-its-kind interactive DIY education kit based on artificial intelligence (AI). Buddhi AI DIY kit can be used to quickly and easily learn the basics of AI and build AI-based solutions for real-world problems. The kit will also help people without having any prior domain knowledge or training. The idea […]","categories":["AI News"],"tags":["AI (Artificial Intelligence)","IIT Delhi","learn ai","Startups"],"author_name":"Sejuti Das","publish_date":"2020-01-30T18:24:41","publication_year":"2020","word_count":373,"keywords":["Go","artificial intelligence","programming_languages:R","AI","programming_languages:Go","RAG","IIT Delhi","Startups","learn ai","R","AI (Artificial Intelligence)","startup"],"extracted_tech_keywords":["AI","artificial intelligence","RAG","R","Go","startup","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/iit-delhi-startup-creates-buddhi-kit-to-make-ai-learning-a-childs-play\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":54985,"title":"Why We Should Embrace Facial Recognition Rather Than Prohibiting It","content":"Facial recognition technology has become the epicentre for outrage among people due to privacy concerns. While some governments consider it as a threat to civilian rights, others justify the deployment of facial recognition technologies to tackle crimes. Deemed as a surveillance tool, the European Union is considering a five-year ban on the implementation of facial recognition to avoid use in public areas like streets, railway stations, and more. However, countries such as India and China are actively embracing facial recognition technology. While India’s home ministry announced its intention to install the world’s largest automated facial recognition software (AFRS), China is using it to monitor activities in public areas. As per research, facial recognition technology is expected to grow and reach $9.6 billion by 2020. Why Facial Recognition Is On The Spotlight Again Sundar Pichai, CEO of Alphabet, in at least more than one conference supported the potential temporary ban on facial recognition by the EU. “I think it is important that governments and regulations tackle it sooner rather than later and give a framework for it,” said Pichai. However, Brad Smith, chief legal officer of Microsoft, had a different opinion when asked about the EU’s move. He said its a young technology and will only get better. But it will only improve if we have more people using it. Amidst several concerns of facial recognition technology, the adoption has only grown across the world. More recently, London’s police on Friday announced that it would be deploying facial recognition to pinpoint crimes. The Metropolitan Police said that technology is crucial for identifying and acting on crimes and violence. Earlier the London police used to match the images with databases to determine suspects; however, with the new announcement, they will now automate the process with real-time detection. Rise In The Adoption Of Facial Recognition Technology Similar to the London police, the New York Police Department uses the image matching technique. According to a report, more than 600 law enforcement agencies have been leveraging facial recognition. In 2019, Kenya also announced the need for surveillance to enhance security, which led them to use facial recognition tech from Chinese firms. Although China is considered a significant driver of surveillance, other countries such as Japan, France, and the US, are a substantial contributor to the supply of surveillance technology. According to a report, US tech firms supply surveillance technology to thirty-two countries. Contrary to the narrative of people, facial recognition technology has helped countries to maintain law and order. For one, Wales’ police got hold of 58 wanted people using facial recognition technology. Besides, in 2019 it was used in public places like Airports in India to expedite the security check and eliminate the need for paper-based boarding passes. Further, the use of facial authentication for new mobile users will restrict the misuse of services. Outlook Facial recognition technology has various shortcomings which invoke concerns about its adoption. Some of the most notable ones are of being biased, a threat to privacy, and miss identifying people, to name a few. On numerous occasions, the technology has failed to work, thereby drawing the attention of people about its effectiveness. However, like any other, it is prone to various flaws, but this doesn’t mean we should ban it. Today, we embrace machine learning models in several use cases even though they are not cent per cent accurate. Similarly, a shortcoming shouldn’t be the reason for squashing the advantages that we can harness from the technology. Undoubtedly, there is a need for regulation, but temporarily banning it for analysing its potential misuse can slacken the development of the technology. Consequently, we should embrace facial recognition and continue to improve it as we move forward.","excerpt":"Facial recognition technology has become the epicentre for outrage among people due to privacy concerns. While some governments consider it as a threat to civilian rights, others justify the deployment of facial recognition technologies to tackle crimes. Deemed as a surveillance tool, the European Union is considering a five-year ban on the implementation of facial […]","categories":["AI Features"],"tags":["facial detection software","facial recognition India","Sundar Pichai"],"author_name":"Rohit Yadav","publish_date":"2020-02-01T10:00:00","publication_year":"2020","word_count":616,"keywords":["Go","machine learning","AWS","AI","cloud_platforms:AWS","programming_languages:R","programming_languages:Go","facial detection software","RAG","facial recognition India","ViT","R","Sundar Pichai"],"extracted_tech_keywords":["AI","machine learning","RAG","AWS","R","Go","ViT","cloud_platforms:AWS","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/why-we-should-embrace-facial-recognition-rather-than-prohibiting-it\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10100287,"title":"Musk Will Soon Be Deep Inside Your Head","content":"Elon Musk’s brain implant company Neuralink, which has been a controversy magnet, has begun recruiting for its first-in-human trial. The company has received approval from the independent institutional review board and the first hospital site to begin the recruitment process. As per the company’s official blog, the PRIME Study (which stands for Precise Robotically Implanted Brain-Computer Interface) involves a fully implantable, wireless brain-computer interface (BCI) – to let people with paralysis control external devices with their thoughts – like a computer cursor or keyboard. Elaborating on the process, the blog further stated that the study comprises three aspects. First, the R1 Robot will be used to surgically place the (second) N1 Implant’s threads in a region of the brain that controls movement intention. Once in place, the ‘cosmetically invisible’ implant will record and transmit brain signals wirelessly to the third step, an application that decodes movement intention. We’re excited to announce that recruitment is open for our first-in-human clinical trial!If you have quadriplegia due to cervical spinal cord injury or amyotrophic lateral sclerosis (ALS), you may qualify. Learn more about our trial by visiting our recent blog post.…— Neuralink (@neuralink) September 19, 2023 Currently, the company is looking for candidates with quadriplegia due to vertical spinal cord injury or ALS who are over the age of 22 and have a “consistent and reliable caregiver” to be part of the study. The participants will first be in an 18-month study involving 9 visits with researchers. After that, they’ll spend at least two hours a week on BCI research sessions and then do 20 more visits over the next five years. The announcement comes five months after the company got a green light from the US Food and Drug Administration (FDA) after a long regulatory scrutiny that the Musk-backed firm faced. Musk envisions that the Neuralink brain implants could potentially treat a range of conditions, including obesity, autism, depression, and schizophrenia. His confidence in the technology has been seen through his willingness to implant it in his own children.","excerpt":"Elon Musk’s brain implant company Neuralink has begun recruiting for its first-in-human trial","categories":["AI News"],"tags":["Elon Musk","Elon Musk Neuralink","neuralink"],"author_name":"Tasmia Ansari","publish_date":"2023-09-20T12:07:41","publication_year":"2023","word_count":337,"keywords":["Go","programming_languages:R","AI","neuralink","programming_languages:Go","Elon Musk","Elon Musk Neuralink","R"],"extracted_tech_keywords":["AI","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/musk-will-soon-be-deep-inside-your-head\/","complexity_score":3,"technical_depth":5,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":38172,"title":"BOSQUE Is The Newest Language In Town &#038; Promises To Keep The Code Simple","content":"The newest language to have made an entry into the programmers’ community is BOSQUE. An open source language by Microsoft, it has been developed by Microsoft computer scientist Mark Marron. Marron, who has been working to make the language a reality over the years, believes that BOSQUE is an effort to move beyond the structured programming model that rose to popularity in the 1970s. The very idea of introducing this language to the programming community is to embrace algebraic operations and to stop the use of techniques that may be highly complex. BOSQUE is inspired by TypeScript, which is a superset of JavaSCript and was released by Microsoft in 2012. It also derives similarity from Node.js and has been made in a way that enables creating code in a hassle-free manner for both humans and machines, by eliminating any accidental complexity. What Does BOSQUE Bring To Table? With BOSQUE, Marron aims to change the behaviour of the programme. The central goal of the language is to build an automated zero-effort code validation. Marron shares that BOSQUE also aims to automate SemVer (semantic versioning) checking and compilation to use SIMD hardware, such as AVX or SSE. While this has been achieved only partly until now, researchers are trying to overcome the challenges to bring out more practical implementations that are viable in the long run. While facilitating regularised programming, Marron aims to bring a version of BOSQUE that eliminates major sources of errors, simplifies code understanding and modification, and converts many automated reasoning tasks over code into trivial propositions. It aims to get rid of sources of complexity such as loops, mutable state, indeterminate behaviours, reference equality and more. Though it currently relies on an interpreter written in TypeScript, run on Node.js, as a reference implementation, Marron intends to further develop on this going forward. He aims to focus on TODO items, bug fixes, and developing features that will support writing larger programs in the language. Taking Inspiration From TypeScript BOSQUE follows a lot of trends and features from JavaScript and TypeScript. TypeScript, which has picked up the popularity charts among the developers, is a superset of  JavaScript that adds optional types to JavaScript for large scale applications for any browser or host. With the most recent and updated version of Typescript, the researchers have tried to come up with new features and shorter build time. BOSQUE derives from a combination of TypeScript inspired syntax and Node\/JavaScript inspired semantics giving ease of point of entry into exploring the new language. “JavaScript and TypeScript being two of the most popular languages with other interesting projects, like ReasonML, experimenting with fusing functional programming with block scopes and “{. . .}” braces. The BOSQUE language follows this trend with two novel adjustments, allowing multiple assignments to a variable and supporting statement expressions, to support functional style programming in a block-scoped language,” notes Marron in his paper. He also shares that BOSQUE is designed to encourage limited uses of recursion, increase the clarity of the recursive structure, and enable compilers\/runtimes to avoid stack related errors. “To accomplish these goals we borrow from the design of the async\/await syntax and semantics from JavaScript\/TypeScript 9 and F# which is used to add structured asynchronous execution to these languages,” he says. BOSQUE Has A Long Way To Build Industry Use Cases Currently, BOSQUE can be used to create automated developer tools such as verifiers and compilers through collaboration with academic and online developer communities. However, researchers see it transitioning from idea to production at a rapid pace, and they aim to improve upon based on the user feedback. The GitHub repository reads that BOSQUE language should not be used at this point of time for any production work and instead encourage experimentation only with small and experimental side projects at this point in time. Marron believes that outside of academic experimentation, if everything goes well, BOSQUE can find implementation in the cloud and IoT space.","excerpt":"The newest language to have made an entry into the programmers’ community is BOSQUE. An open source language by Microsoft, it has been developed by Microsoft computer scientist Mark Marron. Marron, who has been working to make the language a reality over the years, believes that BOSQUE is an effort to move beyond the structured […]","categories":["AI Features"],"tags":[],"author_name":"Srishti Deoras","publish_date":"2019-04-24T07:10:43","publication_year":"2019","word_count":657,"keywords":["Go","AI","ML","Git","TypeScript","RAG","Aim","JavaScript","R","Java"],"extracted_tech_keywords":["AI","ML","Aim","RAG","R","JavaScript","TypeScript","Go","Java","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/bosque-is-the-newest-language-in-town-promises-to-keep-the-code-simple\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10131248,"title":"GitHub Thinks It’s Hugging Face","content":"Microsoft’s GitHub is on a roll. The company recently announced the launch of  GitHub Models which will offer developers access to leading LLMs, including Llama 3.1, GPT-4o, GPT-4o Mini, Phi 3, and Mistral Large 2. “You can access each model via a built-in playground that lets you test different prompts and model parameters, for free, right in GitHub,” the company said in its blog post. Who cares about going outside and enjoying the sunshine?? With GitHub Models, it’s a BCS summer ya’ll.It’s time to Build. Cool. Shit. https:\/\/t.co\/7e7dh2Ak4y— Thomas Dohmke (@ashtom) August 1, 2024 “GitHub Models marks another transformational journey of GitHub. From the creation of AI through open source collaboration, to the creation of software with the power of AI, to enabling the rise of the AI engineer with GitHub Models – GitHub is the creator network for the age of AI,” said Github chief Thomas Dohmke. With GitHub Models, the platform seeks to be more than just an AI pair programmer reliant on OpenAI’s models. Developers now have access to the latest LLMs and can experiment to find the one that best suits their needs. Just as ‘natural language’ has become a prominent programming language, GitHub is positioning LLMs as the go-to framework to develop new software and products. With GitHub Models, over 100 million developers can now access and experiment with top AI models where their workflow is – directly on GitHub. This will allow developers to build AI applications whereever they manage their code. “With GitHub Models, developers can now explore these models on GitHub, integrate them into their dev environment in Codespaces and VS Code, and leverage them during CI\/CD in Actions – all simply with their GitHub account and free entitlements,” explained Dohmke. GitHub is high on confidence as its financial performance has been equally impressive. During Microsoft’s recent earnings call, chief Satya Nadella said, “Copilot is driving GitHub growth overall. GitHub’s annual revenue run rate is now $2 billion.” He further added, “Copilot accounted for over 40% of GitHub’s revenue growth this year and is already a larger business than GitHub was when we acquired it.” The AI-powered coding assistant now has 1.3 million paid subscribers, marking a 30% increase quarter-over-quarter, according to Nadella. Meanwhile, GitHub Copilot Business has secured 50,000 enterprise customers across various industries. This year, Accenture plans to deploy the tool to 50,000 developers. Other notable enterprise customers include Goldman Sachs, Etsy, and Dell Technologies, Nadella said. Getting ‘Huggy’ With AI Developers Github Models seems to be inspired by Hugging Face. Hugging Face also provides the ability to test out different models. It offers Git-based code repositories, pre-trained models for NLP, computer vision and audio tasks, datasets for translation and speech recognition, and spaces for small-scale demos of machine learning applications. github is so slow at shipping obvious features, first copilot and now what looks like a hugging face clonebut when it strikes, distribution carries insane weightgravity is real https:\/\/t.co\/5TwbRgFX96— Nathan Benaich (@nathanbenaich) August 1, 2024 Recently, NVIDIA developed a playground called NVIDIA NIM, which hosts several open-source models covering reasoning, vision, retrieval, biology, and more. LMSYS Chat and Groq also offer playgrounds to try out different LLMs but none of them offer the capability to write code and build apps. Hugging Face recently partnered with NVIDIA to offer inference as a service. These new capabilities will enable developers to rapidly deploy leading LLMs, such as the Llama 3 family and Mistral AI models, with optimisation from NVIDIA NIM microservices running on NVIDIA DGX Cloud. Supposedly, GitHub Models also looks very familiar to Azure AI  Studio, with Microsoft bringing it under GitHub so that developers who are active on GitHub can experiment with it. Is this the first time that @Azure services are so deeply integrated into @github and visible to users? pic.twitter.com\/ML5TSICUqu— Philipp Schmid (@_philschmid) August 1, 2024 “Seems like this is a sales funnel for Azure’s OpenAI\/LLM gateway with GitHub as a proxy(?). It’s a bit unclear. Regardless, I’d be pretty wary of adding either Azure or GitHub as a core dependency to any of my apps at this point with how poor uptime seems to be at both lately,” posted one user on Hacker News. Even if it is true, this definitely seems to be an attempt by Microsoft to make GitHub a one-stop solution for developers to build AI applications. https:\/\/twitter.com\/fatih2211\/status\/1819090376118178041 The G in AGI Stands for Github? Dohmke quipped, “The path to AGI will not be built without the source code and collaboration of the interconnected community on GitHub.” However, during his recent visit to India, he expressed different views on AGI. The path to AGI will not be built without the source code and collaboration of the interconnected community on GitHub.— Thomas Dohmke (@ashtom) July 27, 2024 “I think everybody has a different understanding of what AGI even means and what the G is. I think today I see no sign that machine learning models, or large language models, have sentience. They are not creative. They are machines created by us that help us with the things that we want to do,” he said in an interaction with AIM. Moreover, he believes that GitHub will enable 100 million developers to become AI engineers and AI won’t take away the jobs of developers. “In many ways this new age of AI has actually created more demand for developers because now somebody also has to build all the AI systems. “When you embark on a new job, whether fresh out of college or transitioning from another company, the primary challenge is understanding the company’s operations and code bases, which could be thousands of files,” he said. A tool like Copilot could be very useful for entry-level coders. Moreover, Dohmke believes that going forward, applicants will be expected to be adept at using AI tools like Copilot and ChatGPT. “Some software companies have even begun incorporating Copilot into their interview processes, replacing traditional coding exercises with tasks that assess applicants’ ability to utilise these tools effectively,” Dohmke said.","excerpt":"With GitHub Models, over 100 million developers can now access and experiment with top AI models where their workflow is – directly on GitHub.","categories":["Global Tech"],"tags":["GitHub"],"author_name":"Siddharth Jindal","publish_date":"2024-08-02T17:49:14","publication_year":"2024","word_count":1001,"keywords":["ChatGPT","machine learning","OpenAI","AI","GPT-4o","ML","computer vision","NLP","Aim","Mistral Large 2","GitHub"],"extracted_tech_keywords":["AI","machine learning","ML","NLP","computer vision","GPT-4o","ChatGPT","OpenAI","Mistral Large 2","Aim"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/github-thinks-its-hugging-face\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10041221,"title":"Plea To ML Researchers: Give Data Curation A Chance","content":"Most NLP researchers prioritise the development of deep learning models over the quality of training data. The relative lack of attention results in training data picking up spurious patterns, social biases, and annotation artefacts. Anna Rogers from the Centre of Social Data Science, University of Copenhagen, recently presented a paper underlining the importance of taking care of data as the first step towards building successful NLP models. … and this idea by @AndrewYNg of benchmarks\/competitions switching from tweaking models to tweaking data:https:\/\/t.co\/Tpc5sq442N\/5— Anna Rogers is looking for postdocs! (@annargrs) May 31, 2021 Data curation is the organisation and integration of data collected from multiple sources. The process involves authentication, archiving, management, preservation for retrieval, and representation. Her paper laid down the arguments for and against data curation. Why data curation is important In her paper, Rogers gives the following arguments in support of data curation: Social biases: Written text may contain all kinds of social biases based on race, gender, social status, age, and ability. Models may learn these biases, and when deployed in real-world scenarios, they may propagate and further amplify them. This puts minority groups at a significant disadvantage. It’s imperative to select data taking sociocultural characteristics into account and promote fair representation of all social groups. Privacy: Using personally identifiable information in training data can give rise to privacy and security concerns. For example, a study showed GPT-2 memorised personal contact information even when it appeared only on a few web pages. “Deciding what should not be remembered is clearly a data curation issue,” writes Rogers. Security: Universal adversarial triggers force models to output a certain prediction. A recently discovered phenomenon, this effect affects the training data, compromising even the robust models. Data curation can help avoid this attack. Evaluation methodology: For NLP tasks, the test sample comes from the same distribution as the training samples. There is a possibility of the samples getting overlapped. Curation is necessary to ensure no overlapping takes place. Progress towards NLU: With rapid scaling, we often lose track of the data on which a model is trained. Without data curation, the models may suffer from one of the following issues: Falling prey to common perturbations. For example, linguistic phenomenons such as negations.Learning spurious patterns in the data.Struggling to learn rare occurrences. Arguments against data curation Many experts believe data must be used in their natural form to give an unvarnished output. While there is no problem with this argument, Rogers said, it needs more elaboration. “In that case, the “natural” distribution may not even be what we want: e.g. if the goal is a question answering system, then the “natural” distribution of questions asked in daily life (with most questions about time and weather) will not be helpful,” wrote Rogers. She further added there is still a lot of research work that needs to be done before developers can study the world as it is. Some developers feel their data is large enough for their training set to encompass the ‘entire data universe’. Rogers said collecting all data is impossible as it will pose legal, ethical, and practical challenges Meanwhile, many are in favour of developing algorithmic alternatives to data curation. As per Rogers, this is a good possibility; however, having such solutions, in the current scenario, could be a complementary approach to data curation rather than completely replacing it. A few experts believe data curation is part of the process and should not become a task big enough to forget the original purpose of developing a model. Even though the current deep learning systems are better, they still need to train within the range of the training data, Rogers said. “A perfect dataset would provide a strong signal for each phenomenon that should be learned. That’s not how language works, so we may never be able to create something like that,” she said. While it may be difficult to achieve perfect solutions, it is always possible to improve the models. Curation means making a decision about what to include and what to exclude. This can be a daunting task and requires a lot of interdisciplinary expertise, Rogers said. Wrapping up “We do want more robust and linguistically capable models, and we do want models that do not leak sensitive data or propagate harmful stereotypes. Whether those goals would be ultimately achieved by curating large corpora or by more algorithmic solutions, in both cases we need to do a lot more data work,” writes Rogers. To achieve this goal, the developers have to overcome interdisciplinary tensions and promote truly collaborative spaces.","excerpt":"Data curation is the organisation and integration of data collected from multiple sources.","categories":["AI Features"],"tags":["data privacy use cases","machine learning model","Privacy","Security"],"author_name":"Shraddha Goled","publish_date":"2021-06-04T14:00:00","publication_year":"2021","word_count":760,"keywords":["data science","Privacy","Go","TPU","API","AI","Security","data privacy use cases","NLP","GPT","machine learning model","deep learning","GAN","R"],"extracted_tech_keywords":["AI","deep learning","NLP","data science","TPU","R","Go","API","GPT","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/plea-to-ml-researchers-give-data-curation-a-chance\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10088840,"title":"Can Stable Diffusion Make Mind-Reading a Reality?","content":"Steven Speilberg’s 2002 sci-fi hit Minority Report showed a ‘pre-crime police department’ that would prevent future crimes before it was committed thanks to three clairvoyant humans (precogs). This was a concept beyond imagination. The precogs were genetically engineered to foresee future crime, which the police could see via a video projection from their minds. We may be far from getting there yet, but we are surely advancing on the path of generating images from the human brain. Two researchers in Japan, Yu Takagi and Shinji Nishimoto, recently, submitted a paper where diffusion models (DMs) like Stable Diffusion were used to generate high-resolution images from human brain activity. A study was proposed where images are reconstructed with the use of fMRI (functional magnetic resonance imaging). The goal was to interpret the connection between computer vision models and our visual system. By reconstructing visual experiences from human brain activities, the way a human brain processes visual information can be ascertained. fMRI, used for image reconstruction, measures brain activity by detecting changes associated with blood flow. The technique combines cerebral blood flow and neuronal activation. In the proposed paper, high-resolution images were reconstructed with high fidelity without any additional training or fine-tuning of complex deep-learning models. Each component of the LDM is mapped to specific components of the brain regions. There have been previous attempts to reconstruct visual images from fMRI, however, newer studies use deep generative models trained on a large number of naturalistic images. There is a limitation to these methods though. Training and fine-tuning of generative models such as GANs (Generative Adversarial Networks), a type of neural network architecture, with the dataset used in fMRI experiments is challenging as the sample size in neuroscience is small. However, DMs and LDMs (latent diffusion models) have the ability to generate high-resolution images with high semantic fidelity of text conditioning, and high computational efficiency. Latent Diffusion Model LDM is a type of computer program that can learn to create images by transforming a simple noise pattern into a complex image. LDM can be trained with a dataset of images from which it learns to create new images that look similar to the training data. Once trained, the model will be able to create images by starting with a random noise pattern and gradually transforming it into an image that looks like it belongs in the dataset. In the proposed paper, each component of an LDM (Stable Diffusion) is quantitatively interpreted from a neuroscience perspective by mapping specific components to distinct brain regions. Image Source: sites.google The encoder-decoder model is used where one neural network (the encoder) is used to transform the input data into a fixed-length representation and then use another neural network (the decoder) to generate the output based on this encoding. Image Source: biorxiv.org Row 1: Presented images. Row 2 : Images reconstructed from fMRI signals The research also worked with prediction accuracy of the encoding models for three types of latent representations associated with the diffusion model. Latent representations are compressed, abstract features or variables of the data that capture the most relevant and useful information inferred from raw data, for a particular task. A latent representation of the original image- z, a latent representation of image text annotation- c, and zc which is a noise-added latent representation of z after the reverse diffusion process with cross-attention to c. Image Source: biorxiv.org Mind Reading? With the announcement of this paper, people have been quick to react with the consideration of this model becoming the next mind reader. However, this model is not trained to interpret thoughts and words. The model is an AI extension on previous studies of brain mapping through fMRI or electroencephalography (EEG), where the imaging machine is able to detect only broad patterns of activity. The proposed model is still in the nascent stages of interpreting brain activities. Future Scope Brain mapping is already implemented in the medical sector in diagnosing and understanding a patients’ illnesses pertaining to triggers and tumors. With focussed brain readings, doctors are able to deliver targeted treatments. With image reconstruction from brain activities using LDM, the integration with an already existing framework of brain mapping, can bring advancements in the medical field. If the proposed model comes into the picture, future refinement of the model can probably assist with jobs where such a model can be extensively implemented. For example, in crime, an eyewitness testimony is influenced by the mental state and surroundings of the witness, which can often cloud the description of the suspect. With this model, eyewitness or victim’s recollection of the suspect will become simpler. However, the implementation of such a technology will bring the focus on ethical mind reading.","excerpt":"New research proposes the reconstruction of high-resolution images from human brain activity; opens doors to future technology","categories":["AI Trends"],"tags":["GAN","Stable Diffusion"],"author_name":"Vandana Nair","publish_date":"2023-03-07T18:00:00","publication_year":"2023","word_count":781,"keywords":["Go","TPU","AI","Stable Diffusion","neural network","diffusion models","computer vision","stable diffusion","ViT","GAN","R"],"extracted_tech_keywords":["AI","neural network","computer vision","TPU","R","Go","stable diffusion","diffusion models","GAN","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/can-stable-diffusion-make-mind-reading-a-reality\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":22263,"title":"A Lot Of Enterprises Do Not Have A Clear Separation Of Ownership Of Data And Insights, Says Mahesh Yellai Of Infruid","content":"Infruid team A Business Analytics and Data Visualization Solutions company, Infruid Labs is a brain child of Mahesh Yellai, who brings 18 years of experience in the technology industry. Based out of Hyderabad, the company has been offering software built for big data to provide actionable intelligence to customers by helping them find pattern in large datasets. Infruid—which is derived from two words—information and druid (a wise man from the Iron Age who learned all disciplines and was looked up to by the society for knowledge and wisdom), believes in brewing magic potions that makes their customers invincible. Mahesh Yellai Yellai, who is a graduate with an MBA from Indian School of Business (ISB) and a B.Tech. from Indian Institute of Technology – Madras (IIT – Madras), founded the company to help businesses with data-driven decision making. With specialties in product strategy, competitor analysis, new product development, financial analysis and more, he has been instrumental in bringing out the company’s patent pending analytics platform—Vizard. Analytics India Magazine interacted with Mahesh Yellai, where he explained us about Infruid’s analytics platform, how is it helping companies make data driven decisions, plans for the year 2018 and much more. Q. Would you like to explain in detail about Infruid’s instant analytics platform, Vizard? How is it being leveraged by various businesses to carry data driven decision making? MY: Vizard, Infruid’s patent-pending and award-winning Self-service BI platform, is transforming Infruid’s customers into data-driven Organizations. Vizard is helping our customers democratize analytics across the Organization, so that everyone in the organization, who needs to make decisions, has easy access to instant insights. Some of the best run businesses are using Vizard across Sales, Marketing, Finance, HR and Operations departments. Q. How is Vizard different from other analytics platform available in the market? MY: Vizard’s unique search-driven analytics capabilities are helping users derive deep insights from their data. Vizard uses Controlled Natural Language Processing techniques to translate ad hoc queries from simple English to visual insights. Vizard’s consume-grade user experience hides all the complexity of Big Data processing behind a simple search box, so users can type any question about their Business and slice and dice their data even at big bata scale to get visual insights instantaneously. Q. What are the various other products and tools in analytics and data visualization offered by Infruid? MY: Vizard is Infruid’s flagship analytics product. In addition, we also provide a data integration service that helps our customers connect to different data sources (cloud and on-premises) and extract, transform and load data into Vizard. Q. Would you like to highlight few use cases where Infruid’s analytics platform has benefitted your clients in various domains? Who are some of your major clients? MY: A leading industrial equipment manufacturer is using Vizard to help their Sales team optimize selling price and maximize revenue. The optimal selling price, for each of thousands of products, is determined in Vizard using historical data. Vizard has helped the Customer increase revenue by enabling data-driven pricing decisions. A leading Steel manufacturer is using Vizard to derive insights from their inventory data integrated with their ERP software. Vizard has helped the Customer plan their inventory stock movement across their distributed warehouses and reduce their Inventory Holding Costs. A leading renewable energy company is using Vizard to run their Network Operations Centre to analyze IoT data from their solar power plants and remotely monitor geographically distributed power plants. Vizard has helped the Customer lower their operational expenditure through preventive maintenance. Q. How has been the growth story so far? Would you like to list some of the major milestones that you have achieved in your analytics journey? MY: We have been in business for five years; and during the course of our journey, we have worked with customers across different Industries and different business functions. During the process, we have gained deep understanding of each of the domain specific problems ranging from price optimization to detecting equipment malfunction through sensor data. As we gained market traction and saw adoption of Vizard among some of the world’s leading businesses, we also won Industry awards. We were awarded best Enterprise Software Startup by HYSEA in 2016 and awarded one of 50 Emerging Startups by NASSCOM in 2017. Q. How did the idea of founding Infruid conceptualise? Would you like to talk about the founding team? MY: During my two decades of experience in the Enterprise Software Industry, I was fortunate enough to work on some category-leading products across multiple domains. A recurring problem across all those domains was that even though pre-configured dashboards were great for reporting performance metrics, they were useless for analysis. We realised that what the users needed was an ad hoc analytics tool that allowed users to not just report KPIs and business metrics but to perform meaningful analysis to support decision making. To address this specific need, we built a free form ad hoc querying tool to give users the freedom to perform advanced analytics on data of any complexity and scale. Q. Would you like talk about funding? MY: We raised a Pre-Series A towards the end of 2016 from a few angels from USA and China. Q. How big is the analytics team? What are the skillsets you look for while hiring for analytics at Infruid? MY: The most important characteristic we look for, when hiring new team members, is their problem solving capabilities. We are okay even if the individual does not have prior experience in building large scale systems or building insightful data visualizations, as long as the individual is capable of solving complex problems with elegant solutions and has the right attitude. Q. What are some of the major challenges you face being a startup in analytics space? MY: In enterprises, IT should own data and Business should own insights. But the challenge is that a lot of enterprises do not have clear separation of ownership of data and insights. Vizard gives enterprises the opportunity to bring in this separation to enable IT to focus on more strategic aspects of data infrastructure management and data governance, while Business can focus on queries and insights. Q. What are the plans for analytics at Infruid in 2018? MY: One of the reasons our Customers prefer Vizard over our competition is because of Vizard’s intuitive data discovery capabilities. Soon, we plan to release enhancements to our data discovery capabilities to support augmented data discovery. Augmented data discovery feature uses AI to auto detect patterns of interest from data and bubble them up to users’ attention.","excerpt":"A Business Analytics and Data Visualization Solutions company, Infruid Labs is a brain child of Mahesh Yellai, who brings 18 years of experience in the technology industry. Based out of Hyderabad, the company has been offering software built for big data to provide actionable intelligence to customers by helping them find pattern in large datasets. […]","categories":["AI Features"],"tags":["Interviews and Discussions"],"author_name":"Srishti Deoras","publish_date":"2018-03-05T09:53:39","publication_year":"2018","word_count":1092,"keywords":["big data","Go","AI","data-driven","RAG","analytics","data governance","GAN","R","analytics platform","Interviews and Discussions"],"extracted_tech_keywords":["AI","analytics","RAG","R","Go","big data","data governance","GAN","analytics platform","data-driven"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/lot-enterprises-not-clear-separation-ownership-data-insights-says-mahesh-yellai-infruid\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":25277,"title":"IBM Unveils Next-Generation Servers Designed For AI","content":"Global technology major IBM on Friday unveiled its next-generation Power Systems Servers incorporating its newly designed POWER9 processor. The new systems are reportedly capable of improving the training times of deep learning frameworks by nearly 4x allowing an enterprise to build more accurate AI applications, faster. Viswanath Ramaswamy, Director -Systems, India\/ South Asia, said, “IT infrastructure needs to be re-designed for the AI era, which lets companies analyse data in milliseconds and make decisions driven by data. AI workloads demand new hardware and software paradigms and the infrastructure to deliver data-driven workloads. The P9 architecture aims to transform computing across every industry and profession, turning client data into faster insights where inferences can be drawn quicker and faster to market.” “The new POWER9-based AC922 Power Systems are the first to embed PCI-Express 4.0, next-generation NVIDIA NVLink and OpenCAPI, which combined can accelerate data movement, calculated at 9.5x faster than PCI-E 3.0 based x86 systems,” IBM said in a statement. The POWER9-powered system will help data scientist build applications faster, ranging from deep learning insights in scientific research, real-time fraud detection and credit risk analysis. According to the company, the system was designed to drive demonstrable performance improvements across popular AI frameworks such as Chainer, TensorFlow and Caffe, as well as accelerated databases such as Kinetica. Last week, IBM launched their AI-powered enterprise marketing solution in India to allow customers to host their marketing data on local IBM cloud data centre. This will also give customers proximity, scalability, and will help them meet regulatory requirements.","excerpt":"Global technology major IBM on Friday unveiled its next-generation Power Systems Servers incorporating its newly designed POWER9 processor. The new systems are reportedly capable of improving the training times of deep learning frameworks by nearly 4x allowing an enterprise to build more accurate AI applications, faster. Viswanath Ramaswamy, Director -Systems, India\/ South Asia, said, “IT […]","categories":["AI News"],"tags":["Deep Learning","IBM"],"author_name":"Smita Sinha","publish_date":"2018-06-08T12:42:44","publication_year":"2018","word_count":254,"keywords":["API","AI","data-driven","Scala","Aim","deep learning","ai_frameworks:TensorFlow","IBM","Deep Learning","TensorFlow","R","fraud detection"],"extracted_tech_keywords":["AI","deep learning","Aim","TensorFlow","fraud detection","R","Scala","API","data-driven","ai_frameworks:TensorFlow"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/ibm-unveils-next-generation-servers-designed-for-ai\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10045410,"title":"New Weekend Hackathon For Data Scientists: Music Genre Classification","content":"MachineHack is back with an exciting challenge, in yet another Weekend Hackathon. The challenge is to train a model to predict the genre and quality of the music. As part of MachineHack Weekend Hackathon Edition #2 — The Last Hacker Standing, we pose unique problem statements every week. Weekend Hackathon Edition #2 will run for six weeks, from 30th July to 9th September 2021. PARTICIPATE & STAND A CHANCE TO WIN FREE PASSES TO THE DLDC 2021!!! Problem statement & description Music has been an important part of our lives since time immemorial. Every artist has a signature, making music a subjective art. We have scales\/metrics to measure the quality of music. But, is it possible to train a machine learning model to predict the genre and quality of the music? Currently, many music aggregator applications rely on machine learning to power their recommendation engine, and curate playlists. MachineHack is challenging data scientists and machine learning practitioners to build a highly scalable ML model for a music aggregator app (Company ABC) to accurately predict the genre of songs in the dataset. The hackathon will start on 6th August 2021 at 8 pm (IST). Click here to participate. Overview The participants will work on a list of songs provided in the data set. MachineHack has created a training dataset of 17,996 rows with 17 columns of artist name; track name; popularity; ‘danceability’; energy; key; loudness; mode; ‘speechiness’; ‘acousticness’; ‘instrumentalness’; liveness; valence; tempo; duration in milliseconds and time_signature. It also includes ‘Class’ such as Rock, Indie, Alt, Pop, Metal, HipHop, Alt_Music, Blues, Acoustic\/Folk, Instrumental, Country, Bollywood, as the target variable. The dataset for testing includes 7,713 rows with 16 columns. The prerequisites to attend the hackathon includes knowledge of multi-class classification and the ability to optimise log loss. Submission Guidelines The participants must submit a .csv\/.xlsx file with exactly 7713 rows with 11 columns for Class\/Genres. The submission will return an ‘Invalid Score’ in case of extra columns or rows. Scikit-learn models support the predict() method to generate the predicted values. The submission limit for this hackathon is one account per participant. Click here to participate. Evaluation criteria The evaluation of the hackathon will be done using the Log Loss metric. The hackathon will also support private and public leaderboards. While the public leaderboard will be evaluated on 30% of the test data, the private leaderboard will be made available at the end of the hackathon and will be assessed on 100% of the test data. The final score will be based on the ‘Best Score’ on the public leaderboard. The hackathon will end on 12th August 2021 at 6 pm (IST). Prizes The top three winners will get free passes to the Deep Learning DevCon 2021 (DLDC), scheduled to be held on 23-24 September 2021. In addition, the winners will also get a chance to improve their Global Leader-Board Rankings & become the ultimate MachineHack Grand Master. Click here to participate. Dataset description: Train.csv — 17996 rows x 17 columns (includes ‘Class’ as a target variable)Test.csv — 7713 rows x 16 columns Evaluation metric: Log Loss Skills: Multi-Class ClassificationOptimising Log Loss Start Date: 6th August 2021 at 8:00 PM (IST) End Date: 12th August 2021 at 6:00 PM (IST) Click here to participate in this hackathon.","excerpt":"MachineHack is back with an exciting challenge, in yet another Weekend Hackathon. The challenge is to train a model to predict the genre and quality of the music.","categories":["Deep Tech"],"tags":["data science hackathon","data science hackathon india","Data Science Hackathons","datascience hackathons","Hackathon","Hackathons","Hackathons India","Machinehack","Machinehack Hackathon","MachineHack Weekend hackathon","ml hackathon","The last hacker standing","Weekend Hackathon","Weekend hackathon for data scientists","Weekend hackathon the last hacker standing"],"author_name":"AIM Media House","publish_date":"2021-08-06T18:00:00","publication_year":"2021","word_count":545,"keywords":["scikit-learn","Weekend Hackathon","Scala","deep learning","data science hackathon india","R","data science hackathon","programming_languages:Scala","machine learning","ml hackathon","Weekend hackathon for data scientists","AI","ML","Hackathon","Weekend hackathon the last hacker standing","Machinehack Hackathon","MachineHack Weekend hackathon","Hackathons India","The last hacker standing","Machinehack","programming_languages:R","Hackathons","Data Science Hackathons","datascience hackathons"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","scikit-learn","R","Scala","programming_languages:R","programming_languages:Scala"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/new-weekend-hackathon-for-data-scientists-music-genre-classification\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10171811,"title":"C-DOT Awards Grants to 18 Startups Under Samarth Telecom Incubation Scheme","content":"The Centre for Development of Telematics (C-DOT), under the telecommunications department (DoT), has awarded grants to 18 startups selected under cohort 1 of the ‘Samarth’ programme. The initiative aims to support early-stage companies in the telecom and information and communications technology (ICT) sectors through financial assistance, infrastructure, and expert mentorship. The selected startups were chosen through a competitive process and will receive up to ₹5 lakh each, office space, and lab access at C-DOT campuses in Delhi and Bengaluru. The 18 startups selected in cohort 1 include Turtleneck Systems, Matisoft Cyber Security Labs, Revino Solutions, Ajayan Consulting, Vital Carbon, Atibha Research and Development, Alphappleton Innovations, AgriVerse Innovations, Haranzel Technologies, Threat Expert Cyber Solutions, Brahmi Systems, Paravani Business Solutions, Purvanchal Engineering Systems, Aaizel International Technologies, Brenin Technologies, Jumps Automation, Master Education, and Farsight Innovations. Startups in this cohort are working in areas including 5G and 6G, AI, cybersecurity, quantum technologies, and IoT. The programme is supported by the Software Technology Parks of India (STPI) and The IndUS Entrepreneurs (TiE), who act as implementation partners. The Samarth programme, launched to promote scalable telecom innovations, also includes structured mentoring and acceleration sessions. Delivered in a hybrid format, the scheme spans two cohorts of six months each and accommodates up to 36 startups in total. The initiative aims to lift sustainable and scalable business models by offering startups financial support, access to advanced resources, and expert guidance to help them move from ideation to commercialisation. During the launch event at the Delhi campus, each selected startup received its first tranche of funding. Notable attendees included Rajkumar Upadhyay, CEO of C-DOT; Arvind Kumar, director general of STPI; and Atul Dhawan, president of TiE Delhi-NCR. A memorandum of understanding (MoU) was also signed between C-DOT and STPI to formalise ongoing collaboration in mentorship and joint programmes.","excerpt":"The selected startups will receive up to ₹5 lakh each, office space, and lab access at C-DOT campuses in Delhi and Bengaluru.","categories":["AI News"],"tags":["Department of Telecommunications","Startup ecosystem","Telecom India","Telecommunication"],"author_name":"Sanjana Gupta","publish_date":"2025-06-16T16:55:09","publication_year":"2025","word_count":300,"keywords":["Go","funding","Telecommunication","AI","innovation","Scala","Department of Telecommunications","automation","Startup ecosystem","Aim","ViT","R","Telecom India","startup"],"extracted_tech_keywords":["AI","Aim","R","Go","Scala","ViT","automation","innovation","startup","funding"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/c-dot-awards-grants-to-18-startups-under-samarth-telecom-incubation-scheme\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10067428,"title":"Cypher is back","content":"India’s largest AI conference, Cypher22, started by Analytics India Magazine in 2015, is now back after two years with its 6th edition and in person. Cypher22 takes pride in being not only the largest but also the best conference in India centred around the artificial intelligence landscape. Powered by Fractal, Cypher 2022 will be a three-day event, highlighting the innovations that will drive the world’s next wave of change, balancing these breakthroughs against a world that has grown more sceptical about the benefits of the data revolution. Where: Hotel Radisson Blu, Bengaluru, India When: September 21-23, 2022 | Wednesday – Friday Register now Why you must attend Cypher22 The conference will hold three tracks over three days — thought leadership, knowledge sessions, and hands-on workshops. Apart from these, Cypher 2022 will have the following attractions for attendees: Keynotes Panel discussions Hands-on workshopsMasterclasses Podcasts ExhibitionsWell-executed use casesMentoring sessionsAIQuizAI AwardsAfter-party and cocktail dinner The topics featured will include AI\/ML, deep learning, big data, data analytics, security and privacy, enterprise solutions, hardware, software, chip manufacturing, wearables, robots, edge computing, data ethics, voice technology innovation, and accessibility. Cypher has grown to be one of India’s most captivating AI conferences. At Cypher, you’ll find yourself in the presence of leading experts from renowned tech giants and have the opportunity to interact with them first-hand. Not just that, you will be able to avail of 1-on-1 mentorship from big names in the industry and hear the journeys of tens to hundreds of tech professionals. In addition, the exhibitions will be a great place to be exposed to what several companies are working on and deep dive into their tech stacks. BUY YOUR PASSES NOW! EARLY BIRD PASSES TO EXPIRE ON 15TH JUL, 2022 Cypher presents a unique annual chance to be in the presence of not one but a spectrum of exciting opportunities. With more than 1000 attendees, 300+ organisations, and 100+ industry leaders as speakers, Cypher22 is the premium canon for India’s artificial intelligence ecosystem, a summit which started with the simple idea of connecting the AI community with all industries, both old and new, and it has resonated to great lengths. Sponsors and speakers Cypher has previously been sponsored by Google Cloud, AWS, IBM, Deloitte, SAP, Cognizant, Ericsson, Tableau, the Aditya Birla Group, Wipro, Genpact, ZS Associates, The Weather Company, United Health Group, and MiQ, among others. Be a sponsor for Cypher 2022. Every year, more than 40 organisations sponsor Cypher. To know more details, write to info@analyticsindiamag.com. The summit has had industry titans as speakers from brands such as Amazon, Oracle, Infosys, L&T, Wells Fargo, Lowe’s, Ericsson, Genpact, AB InBev, Udacity, Hindustan Unilever, Swiggy, Dunzo, Indian School of Business, NMIMS University, Vodaphone, PhonePe, Standard Chartered Bank, Bharti AXA, Axis Bank, HDFC Bank, Viacom18, Lifestyle Brands, and TATA SIA Airlines, among several others. Speak at Cypher 2022.To explore speaking opportunities with Cypher, write to info@analyticsindiamag.com. When: September 21-23, 2022 (Wednesday – Friday) Where: Hotel Radisson Blu, Bangalore, India BUY YOUR PASSES NOW! EARLY BIRD PASSES TO EXPIRE ON 15TH JUL, 2022 Registration and tickets Cypher 2022 will be an in-person conference hosted at Hotel Radisson Blu, Bengaluru. It will be a three-day event with an after-party and cocktail dinner on Friday, September 23, 2022. Register here. The schedule and list of speakers for this year’s Cypher will be announced soon. Check this space for relevant announcements. Be a part of this exclusive gathering of AI & Data Science leaders in India, where our agenda is to provide an opportunity for AI professionals at every level to participate, network, and advance their careers. When: September 21-23, 2022 (Wednesday – Friday) Where: Hotel Radisson Blu, Bangalore, India We look forward to having you with us for the event. Hurry up and book your seat now!","excerpt":"Cypher22 takes pride in being the largest and the best conference in India centred around the artificial intelligence landscape.","categories":["Deep Tech"],"tags":[],"author_name":"Anushka Pandit","publish_date":"2022-05-19T13:00:00","publication_year":"2022","word_count":629,"keywords":["data science","Go","artificial intelligence","AWS","AI","ML","deep learning","analytics","edge computing","R"],"extracted_tech_keywords":["AI","artificial intelligence","ML","deep learning","data science","analytics","AWS","edge computing","R","Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/cypher-is-back\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":46677,"title":"Pope Francis Says If AI Fails Mankind, It Would Lead To Barbarism","content":"Joining the illustrious list of names like Elon Musk, Jack Ma and Stephen Hawking, Pope Francis this week sounded alarm bells regarding the increasing use of artificial intelligence. Speaking at a Vatican tech conference, Pope Francis was of the opinion that technology needed both “theoretical as well as moral principals”. According to a report by a noted news agency, Pope Francis said, “The remarkable developments in the field of technology, in particular, those dealing with AI, raise increasingly significant implications in all areas of human activity. For this reason, open and concrete discussions on this theme are needed now more than ever.” He was wary of the misuses of AI and warned the gathering about the fake news and false data that could snowball into real political and social tensions. “If mankind’s so-called technological progress were to become an enemy of the common good, this would lead to an unfortunate regression to a form of barbarism dictated by the law of the strongest,” he added. Last year, Infosys co-founder and celebrated corporate thought leader NR Narayana Murthy had also said that AI was “worrisome” for freshers in the IT sector. Murthy is only one in a long string of influential personalities who have sounded the alarm bells regarding automation, AI, and its effect on employment. US’ former Secretary of State Hillary Clinton had also said that millions of jobs would be on the line if automated machines and applications were implemented on a large scale. While many noted names such as physicist Stephen Hawking and Tesla and SpaceX’s CEO Elon Musk have already sounded alarm bells with regards to AI and its potential harm, there are others who don’t concur.","excerpt":"Joining the illustrious list of names like Elon Musk, Jack Ma and Stephen Hawking, Pope Francis this week sounded alarm bells regarding the increasing use of artificial intelligence. Speaking at a Vatican tech conference, Pope Francis was of the opinion that technology needed both “theoretical as well as moral principals”. According to a report by […]","categories":["AI News"],"tags":[],"author_name":"Prajakta Hebbar","publish_date":"2019-10-01T14:08:26","publication_year":"2019","word_count":280,"keywords":["Go","artificial intelligence","programming_languages:R","AI","programming_languages:Go","automation","Ray","ViT","R"],"extracted_tech_keywords":["AI","artificial intelligence","Ray","R","Go","ViT","automation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/pope-francis-says-if-ai-fails-mankind-it-would-lead-to-barbarism\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10162868,"title":"When AI Met BI: How Amazon QuickSight is Making Data More Accessible","content":"Amazon Q Business, first introduced at the Amazon Web Services (AWS) re:Invent in 2023, has evolved over the past 12 months to become a comprehensive AI assistant that can answer questions, summarise content, generate visuals, and automate tasks – all based on an organisation’s data. QuickSight is Amazon’s completely managed, cloud-native business intelligence (BI) service that is revolutionising the way an organisation interacts with their data and empowers teams at all levels to adopt a data-driven culture. What if your business could unlock the full potential of its data, transforming it into insights that drive smarter decisions at the speed of thought? That’s exactly what Amazon QuickSight delivers. Its features include machine learning–powered insights, natural language queries through QuickSight Q, and predictive analytics that can democratise data and allow users, whether technical or non-technical, to access and act on the data. Tracy Daugherty, general manager of Amazon QuickSight, has been the driving force behind this evolution and has guided the platform through modern data analytics challenges. QuickSight has never been better for scale. With support for multi-tenant deployments for enterprise needs, the offering now easily integrates with other AWS services, including S3, Redshift, and Athena. At AWS re:Invent 2024, AIM interviewed Daugherty and had a detailed discussion about the platform’s journey and its vision for the future. Reflecting on the platform’s early days, Tracy said, “When I joined seven years ago, Amazon QuickSight was architecture-rich but feature-poor.” Despite its potential cloud-native architecture, Amazon QuickSight faced stiff competition from established vendors such as Microsoft Power BI and Tableau. The Journey and Challenges Tracy and his colleagues identified an opportunity to differentiate Amazon QuickSight by emphasising accessibility. “Amazon QuickSight was built to be more than a dashboard tool,” he explained. “It’s about empowering everyone in the organisation with insights, whether you’re a business executive, developer, or frontline worker.” This vision transformed Amazon QuickSight from a basic reporting tool to a full-fledged, self-service BI platform that could support use cases ranging from dashboards and reporting to embedded analytics. “Our purpose was always to democratise data access,” Tracy said. “We wanted to build a tool that would suit everybody – from technical analysts to those who have no technical background.” One of QuickSight’s most significant developments was the launch of Amazon Q, a generative AI-powered assistant that introduced natural language processing (NLP) capabilities to the BI market. “One of the biggest challenges users face is knowing what to ask…Q now understands context, suggests follow-up questions, and provides multiple visual answers for better exploration,” Tracy pointed out. The Ideal Assistant for Businesses Tracy described Amazon Q Business as a breakthrough tool for managing data at scale, with its ability to connect seamlessly to over 40 enterprise data sources, including Microsoft 365, Amazon S3, Google Drive, Salesforce CRM, and Asana. The AI-powered assistant can synthesise data from various sources and provide users with actionable insights via natural language enquiries. “Amazon Q Business brings AI directly into the hands of business users to answer critical questions, automate key tasks, and generate visuals with ease,” Tracy said, explaining how this allows teams to interact with data intuitively. The productivity gains from Amazon Q have been substantial for businesses. According to Tracy, preliminary tests indicate that Amazon Q will increase staff productivity by as much as 80%, specifically through the automated extraction of insights. However, he noted that the actual success measure is more than just the metrics; it is how successfully it is accepted and used across organisations. The integration of generative AI to QuickSight was further amplified in its capability through the launch of scenarios analysis capability in Amazon Q in QuickSight. “It’s a decision-making assistant,” Tracy explained. “Users can simulate outcomes, and Amazon Q returns actionable insights and recommendations in real time.” ‘AI Won’t Replace Analysts’ “AI isn’t here to replace analysts. It’s here to make their work more strategic by automating repetitive tasks and uncovering insights they might otherwise miss,” Tracy said. As Amazon Q in QuickSight pushes the boundaries of what BI tools can do, it faces stiff competition from established players in the industry. Tracy believes Amazon Q in QuickSight’s native integration with AWS gives it a significant advantage. “The integration with AWS services like data lakes, warehouses, and machine learning tools creates a secure, unified ecosystem for enterprises,” Tracy explained. This seamless connectivity not only enhances QuickSight’s functionality but also ensures that it remains a secure platform for business users, with stringent security measures in place to protect sensitive data. A standout feature in Amazon QuickSight’s security capabilities is the Random Cut Forest (RCF) algorithm, which excels in real-time anomaly detection. “Unlike traditional machine learning algorithms, RCF is optimised for real-time anomaly detection, making it invaluable for fraud prevention and operational monitoring,” Tracy said. This focus on security underscores Amazon QuickSight’s commitment to safeguarding customer data while continuing to innovate with AI features. User-Centric Approach Becomes Popular The popularity of Amazon Q in QuickSight can also be attached to its user-centric approach. “Most business intelligence technologies are built for specialists. We built Amazon Q in QuickSight to be self-service, where the intent is that every employee in the organisation can gain insights without requiring a broad degree in data science,” Tracy pointed out. This shift from being a model that was data analyst-centric to making all those employees have their power has helped Amazon QuickSight acquire widespread adoption, considering hundreds of thousands of users depend on it daily. Tracy imagines a world in which BI tools are woven into the fabric of all business processes. “Analytics should feel intuitive. It’s not just about visualising data – it’s about turning those visuals into actionable narratives that drive better outcomes.” Tracy has valuable advice for aspiring BI professionals. “Focus on mastering AI-driven tools and developing a strong foundation in data storytelling. The ability to turn data into actionable narratives is what sets the best apart,” he said. Amazon Q in QuickSight changes how businesses engage with data by including features such as scenarios and generative AI, making analytics a vital driver of strategic decision-making. “The future of BI is about making data accessible, actionable, and transformative for businesses of all sizes,” Tracy concluded.","excerpt":"“Most business intelligence technologies are built for specialists. We built Amazon Q in QuickSight to be self-service, where the intent is that every employee in the organisation can gain insights without requiring a broad degree in data science,” Tracy Daugherty said.","categories":["Global Tech"],"tags":["AI","Amazon"],"author_name":"Anshika Mathews","publish_date":"2025-02-04T17:07:56","publication_year":"2025","word_count":1029,"keywords":["data science","machine learning","AI","ML","Amazon","NLP","Aim","anomaly detection","analytics","generative AI","predictive analytics"],"extracted_tech_keywords":["AI","machine learning","ML","NLP","data science","analytics","generative AI","Aim","predictive analytics","anomaly detection"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/when-ai-met-bi-how-amazon-quicksight-is-making-data-more-accessible\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10124889,"title":"ElevenLabs New iPhone app converts &#8216;any&#8217; text into audio narration using AI","content":"ElevenLabs, a leading artificial intelligence speech platform, has recently released a new iPhone app that allows you to listen to anything from a complete novel to the contents of a website. The new ElevenLabs Reader App is not just another text-to-speech app. It offers a unique feature-a vast voice library of cloned and synthetic sounds, including the ability to clone your own voice. Introducing the ElevenLabs Reader App. Listen to any article, PDF, ePub, or any text on the go with the highest quality AI voices.Download now and have your life narrated: https:\/\/t.co\/hK0myEdZNM pic.twitter.com\/A0z8ldqoQI— ElevenLabs (@elevenlabsio) June 25, 2024 Positive Feedback The software has received positive feedback from beta testers, with many praising the quality and consistency of the voices, particularly for longer pieces. Some visually impaired users found the programme especially useful, with one describing it as quite helpful for viewing print materials. However, several viewers reported formatting flaws that made the narration odd, such as pausing at unsuitable periods due to line breaks. Some users also encountered a couple random error messages and failures to convert text to audio. Was There Really A Need? Popular apps like Pocket, Matter, and Instapaper already include text-to-speech capabilities. Apple Books already has an AI-based narration capability for eBooks, but ElevenLabs Reader takes the capabilities well beyond digital books. The new programme allows iPhone and iPad users to listen to any article, book, webpage, or document on the go. ElevenLabs also provides a full-fledged website with capabilities beyond merely converting text to speech, such as turning speech into other voices and languages, AI dubbing, and AI sound effects. ElevenLabs reader currently only supports narration in English. However, the mobile app will soon include all 29+ languages supported by ElevenLabs’ website. Future Prospects With this release, the AI-voice unicorn enters the mainstream consumer market. It illustrates how its technology can be utilised for uses other than audiobook narration and voice acting in video games. The company raised $80 million from a $1.1 billion financing round in January. So far, most of its deals have focused on audio-centric content, such as a partnership with HarperCollins Publishers to create audio versions of back-catalogue books in many languages. The business has ambitious plans for its dubbing technology, aiming to sell it to YouTube creators, movie studios, and news publishers. This opens up exciting possibilities for the future of audio-centric content creation.","excerpt":"ElevenLabs Reader App offers a unique feature-a vast voice library of cloned and synthetic sounds, including the ability to clone your own voice.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","Apple"],"author_name":"Anshul Vipat","publish_date":"2024-06-26T17:39:47","publication_year":"2024","word_count":395,"keywords":["Go","artificial intelligence","unicorn","AWS","AI","cloud_platforms:AWS","Apple","programming_languages:R","Git","Aim","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","Aim","AWS","R","Go","Git","unicorn","cloud_platforms:AWS","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/elevenlabs-reader-app\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10143808,"title":"Infosys Launches Google Cloud Centre of Excellence in Bengaluru","content":"Infosys, a global leader in digital services and consulting, has launched a Google Cloud Centre of Excellence (CoE) at its Bengaluru campus to drive enterprise AI innovations. The initiative, powered by Infosys Topaz, aims to help businesses leverage generative AI for transformative growth. The centre will use AI to co-create solutions to tackle business and technology challenges. Key areas of focus include contact centre AI, application modernisation, speech-to-speech transformation, and text-to-image solutions. These efforts are expected to enhance efficiency and improve business performance. Long-Standing Partnership Infosys and Google Cloud have a long-standing partnership in delivering data analytics and AI solutions. With over 60,000 Infosys employees already skilled in Google Cloud, the new centre combines the capabilities of Infosys Topaz AI, Infosys Cobalt cloud services, and Google Cloud’s generative AI technologies to create customised enterprise solutions. Balakrishna DR, executive vice president at Infosys, highlighted the strategic importance of the collaboration. “The strengthened collaboration with Google Cloud represents a strategic alignment in our mission to lead enterprise AI innovation,” he said. “We are empowering enterprises to harness the real power of generative AI through collaboration and co-innovation.” Victor Morales, vice president of GSI and consulting partnerships at Google Cloud, echoed this sentiment, stating, “Infosys and Google Cloud are committed to providing customers with the industry expertise and technology needed to accelerate digital transformation.” The centre also includes an AI Experience Zone at Infosys’ Bengaluru campus, where enterprises can explore AI-powered solutions developed through the partnership. Infosys has announced a collaboration with RheinEnergie, a German energy service provider, to help enterprises advance their energy transition and sustainability goals, as announced by the company last Thursday. The initiative leverages Infosys Cobalt and Topaz to help organisations improve energy efficiency by 30–40% while meeting decarbonisation targets. The partnership will provide enterprises with solutions in cloud technology, AI, and Industry 4.0. The focus is on decentralising the energy transition by connecting renewable generation, storage, energy markets, grids, and energy-consuming assets. West Bengal chief minister Mamata Banerjee is set to inaugurate Infosys’ new campus in Kolkata’s New Town on Wednesday. The IT giant, which began operations at the 17.5-acre facility earlier this year, plans to gradually scale up its activities at the site.","excerpt":"The initiative is powered by Infosys Topaz and aims to help businesses use generative AI for transformative growth.","categories":["AI News"],"tags":["Cloud","Google","Infosys"],"author_name":"Shalini Mondal","publish_date":"2024-12-18T11:50:45","publication_year":"2024","word_count":366,"keywords":["Go","Infosys","AI","Cloud","Git","RAG","Aim","ViT","analytics","Google","generative AI","GAN","R"],"extracted_tech_keywords":["AI","analytics","generative AI","Aim","RAG","R","Go","Git","GAN","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/infosys-launches-google-cloud-centre-of-excellence-in-bengaluru\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":50602,"title":"10 AI Applications That Can Generate Code Themselves","content":"Most of the time organisations have to deal with tough problems like errors, defects, and other complexities while developing intricate software. This is where self-coded applications come into play. Applications that generate the code themselves not only help the programmers to accomplish a task in less time but also increase the programming ability of the developer. In this article, we list down 10 artificial intelligence applications which can generate the codes: (The list is in alphabetical order) 1| Bayou Bayou is a system for generating API idioms which are the snippets of code that use APIs in Java. The main task of this system is to use the user’s code and the query in order to generate the appropriate program which will most likely solve the task. It consists of two parts which are draft program and a query. Bayou interprets this query using a method called Neural Sketch learning. 2| Clever-Commit CLEVER (Combining Levels of Bug Prevention and Resolution techniques) was developed in collaboration with the Ubisoft and Mozilla developers. The Clever-Commit is an AI coding assistant which combines data from the bug tracking system and the codebase and helps in searching the errors and bug in the codes. The coding assistant is currently being used internally at Ubisoft for game development purposes. 3| DeepCode DeepCode is an AI software platform which learns from open-source programmers and uses the acquired knowledge to make suggestions on how the code can be improved. A developer can use this platform as a code review tool or as an audit tool. It alerts a user about the vulnerabilities in the code. DeepCode integrates with code hosting platforms such as GitHub, Bitbucket Server or GitLab and for all integrations, the platform provides the same core set of features which are AI QA Audit and AI code review. 4| Embold Built-in AI, Embold is an intelligent, multi-dimensional analyser for software projects. It understands the status of the software quality and detects issues as well as recommend solutions and suggests code analysis for the particular problem. It analyses source code using techniques like natural language processing (NLP), machine learning and a set of algorithms in order to find design issues, bugs, etc. 5| Kite Kite is a python based, AI-powered code completion tool which uses machine learning to provide useful code completions for Python language. Earlier this year, Kite announced that it has raised a $17 million funding round. The plugin is available for Atom, Pycharm, Sublime, VSCode, and Vim. 6| mabl mabl is a Software-as-a-Service (SaaS) provider and a unified DevTestOps platform for ML-based test automation. The key features of this solution include auto-healing tests, ML-driven regression testing, visual anomaly detection, secure testing, data-driven functional testing, cross-browser testing, test output, integration with popular tools, and much more. 7| PyCharm PyCharm is one of the popular and intelligent Python integrated development environments (IDE). It provides smart code completion, code inspections, on-the-fly error highlighting and quick-fixes, along with automated code refactorings and rich navigation capabilities. The smart code editor of PyCharm provides first-class support for Python, JavaScript, CoffeeScript, TypeScript, CSS, popular template languages and more. 8| Pylint Pylint is a Python source code analyser which looks for programming errors, helps to enforce a coding standard, and other such. This quality checker for Python programming includes several features such as coding standard where it checks for the length of line codes, error detection, refactoring by detecting the duplicated code, among others. 9| Run.ai Run.AI is a startup which builds software for virtualisation and acceleration platform for deep learning. It has the computational graph analysis technology which provides automatically distributed training using a hybrid combination of data and model parallelisms. The software includes automated distributed training technology and neural network analysis which allows the users to perform computational tasks in a fast manner at a low cost. 10| Sketch2Code Sketch2Code is a web-based solution which uses AI to transform a handwritten user interface design from a picture to a valid HTML markup code. The solution works in a way such as it first detects the design patterns, understands the handwritten draw or text, understands the structure and then builds a valid HTML code accordingly to the detected layout containing the detected design elements.","excerpt":"Most of the time organisations have to deal with tough problems like errors, defects, and other complexities while developing intricate software. This is where self-coded applications come into play. Applications that generate the code themselves not only help the programmers to accomplish a task in less time but also increase the programming ability of the […]","categories":["AI Trends"],"tags":["AI Applications","Natural Language Processing","Pycharm"],"author_name":"Ambika Choudhury","publish_date":"2019-11-25T18:00:00","publication_year":"2019","word_count":704,"keywords":["artificial intelligence","machine learning","Run.ai","AI","neural network","Natural Language Processing","ML","TPU","AI Applications","NLP","Pycharm","deep learning","anomaly detection"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","neural network","NLP","Run.ai","anomaly detection","TPU"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/10-ai-applications-that-can-generate-code-themselves\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":426,"title":"Netcore’s Internet of Thing business gets acquired by Gaia Smart Cities","content":"As a part of the recently inked deal, Gaia Smart Cities, an India based start-up in the field of City Scale IoT technology solutions has acquired the Internet of Things division of netCORE Solutions, a leader in marketing technology & enterprise communications. Founded by Dr. Sumit D. Chowdhury and Bipin Pradeep Kumar in the year 2015, Gaia Smart Cities pioneer in innovative solutions for smart cities and offers a cloud based product portfolio including Smart Metering, Smart Asset Tracking and Smart Industrial Automation. It is one of the few companies in India that is actively pursuing research and development in the area of developing smart cities. With various initiatives by the government such as Make in India and Digital India being on a rise, it has encouraged Indian companies to build intellectual property and pioneer advanced designs in electronic, software and telecommunications. And this acquisition aims at creating a leading technology and innovation enterprise, offering ground-breaking products rapidly. The acquisition would help Gaia build up its reputation as one of the fastest growing end-to-end IoT companies offering solutions for both industrial automation and smart cities. It would also fetch the company various pro-consumer and pro-competitive benefits to its existing client base including names like Amul, Paras, Parag Dairies, ACC Cement, cold chains, warehouses, retail stores across the country. It would gain Gaia’s customers in Smart Metering, Smart Cities and analytics space too. Dr. Sumit D. Chowdhury, the Founder & CEO of Gaia Smart Cities, commented “We were drawn towards the strong technical team, a robust technical platform and extremely focused approach of the netCORE IOT team. We found great synergies with our own Smart Cities platforms and solutions like Smart Metering and Smart Tracking”. Dr. Chowdhury has previously served as President of Reliance Jio and is an alumnus of IIT Kanpur and Carnegie Mellon. Along with the cloud based platform INSIGHT developed by netCORE Solutions and the right combination of Predictive Analytics & Machine Learning solutions offered by Gaia, the companies intend to bring in a significant business insight and business benefit to the customers. “We continue to believe in the potential of IoT and Smart Cities and have found that the leadership and ideas of Gaia are better and can be aligned to grow this business and support the customers that we already have. We will continue to stay invested with Gaia as the country moves ahead with its various Smart Cities and demand for the solutions developed across the world”, said Mr. Rajesh Jain, Founder and Managing Director of netCORE Solutions. The deal was finalised by Nine Rivers Capital Advisors with Mr. Nimesh Shah, Managing Director & Co-Founder stating that “as part of our Technology and Digital practice, we are very upbeat about rapid adoption of IOT solutions within the country. We see a big synergy in Gaia Smart Cities scaling up what Netcore started as a pioneering effort two years back.”","excerpt":"As a part of the recently inked deal, Gaia Smart Cities, an India based start-up in the field of City Scale IoT technology solutions has acquired the Internet of Things division of netCORE Solutions, a leader in marketing technology & enterprise communications. Founded by Dr. Sumit D. Chowdhury and Bipin Pradeep Kumar in the year […]","categories":["AI News"],"tags":["IoT","Netcore Cloud","smart cities"],"author_name":"Srishti Deoras","publish_date":"2016-07-19T05:19:44","publication_year":"2016","word_count":484,"keywords":["smart cities","Go","API","machine learning","AI","R","Git","RAG","Netcore Cloud","Aim","analytics","predictive analytics","IoT"],"extracted_tech_keywords":["AI","machine learning","analytics","Aim","RAG","predictive analytics","R","Go","Git","API"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/netcores-internet-thing-business-gets-acquired-gaia-smart-cities\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10054208,"title":"Women Innovators And Researchers Who Made A Difference In AI In 2021","content":"There is a troubling and persistent absence of women when it comes to the field of artificial intelligence and data science. Women constitute a mere 22 per cent or less than a quarter of professionals in this field, as says the report “Where are the women? Mapping the gender job gap in AI,” from The Turing Institute. Yet, despite low participation and obstacles, women are breaking the silos and setting an example for players out in the field of AI. To honour their commitment and work done, we have listed some of the women innovators and researchers who have worked tirelessly and contributed significantly to the field of AI and data science. The list below is provided in no particular order. Joy Buolamwini The brainchild behind and the founder of The Algorithmic Justice League (AJL), Joy Buolamwini, has started the organisation that combines art and research to illuminate the social implications and harms of artificial intelligence. With her pioneering work on algorithmic bias, Joy opened the eyes of the world and brought out the gender bias and racial prejudices embedded in facial recognition systems. As a result, Amazon, Microsoft, and IBM all halted their facial recognition services, admitting that the technology was not yet ready for widespread usage. One can watch the famous documentary ‘Coded Bias’ to understand her work. Her contributions will surely pave the way for a more inclusive and diversified AI community in the near future. Cynthia Rudin A large chunk of researchers and scholars focus on improving algorithms for the machines to work efficiently. However, Cynthia Rudin, the Duke University computer science professor and engineer, worked tirelessly to utilise the power of AI to serve humanity and help society. As a result, she was bestowed with the Squirrel AI Award for Artificial Intelligence for the Benefit of Humanity 2021. Her research work majorly focuses on machine learning tools that help humans make better decisions, mainly interpretable ML and its applications. In a conversation with us, Cynthia wished that AI could solve the refugee crisis, reverse climate change and help end extreme poverty. Allie K Miller Allie is currently the Global Head of Machine Learning Business Development, Startups and Venture Capital at Amazon Web Services. She is in the supporting role for many big AI organisations. In her effort to increase representation, Allie co-founded Girls of the Future, an organisation that showcases girls aged between 13 to 18 who are innovating in the field of STEM. Moreover, she will be presenting the Data and Machine Learning keynote at AWS re: Invent with Swami Sivasubramanian, VP, Amazon AI. She formerly worked at IBM, where she oversaw large-scale product development using computer vision, conversation, data, and regulation. Lucia Specia Dr Lucia is a Natural Language Processing Professor at the Imperial College of London. She built and led the Language and Multimodal AI Lab, managing a ​team of 20 researchers and her research examines many aspects of data-driven approaches to language processing, with a focus on multimodal and multilingual context models. Her work has benefited several fields, including machine translation, quality estimate, image captioning, and text adaptation. She is currently involved and working on a number of machine translation projects, including multilingual video captioning and text adaptation, which will surely contribute to the advancement of AI-driven technologies in this field. Cassie Kozyrkov A prominent advocate of safe and reliable AI, Cassie is currently the Chief Decision Scientist at Google. Recently, she introduced the AI-for-everyone course Making Friends with Machine Learning. Her research interests include applied artificial intelligence and data science process architecture. At Google, Cassie co-founded the field of ‘decision intelligence’ at Google, combining social science, decision theory, and managerial science with data science to better understand how actions lead to results. Although females are under-represented in the field of AI and technology, all is not lost yet; some incredibly inspirational female pioneers are shaping the world of AI.","excerpt":"Women constitute a mere 22 per cent or less than a quarter of professionals in the field of AI and Data Science.","categories":["AI Features"],"tags":[],"author_name":"kumar Gandharv","publish_date":"2021-11-25T12:00:00","publication_year":"2021","word_count":649,"keywords":["data science","Go","machine learning","artificial intelligence","AWS","AI","ML","computer vision","multimodal AI","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","computer vision","data science","multimodal AI","AWS","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/women-innovators-and-researchers-who-made-a-difference-in-ai-in-2021\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10059351,"title":"Google Cloud announces dedicated digital assets team","content":"Google Cloud has announced a new, dedicated Digital Assets Team within Google Cloud to build, transact, store value, and deploy products on blockchain-based platforms. The new team will build on Google’s existing work with blockchain developers, exchanges, and other companies in this space. Currently, blockchain and distributed-ledger-based companies like Hedera, Theta Labs, and Dapper Labs are built on top of Google Cloud for scalability, flexibility, and security. Moving forward, Google Cloud’s Digital Assets Team will undertake short and long term initiatives to support companies in the digital assets\/blockchain ecosystem. The steps include providing dedicated node hosting\/remote procedure call (RPC) nodes for developers, allowing users to deploy blockchain validators on Google Cloud via a single click, and participating in node validation and on-chain governance with select partners. The Digital Assets Team will also support on-chain governance via participation from Google Cloud executives and senior engineers, host several public BigQuery datasets on Marketplace, like full blockchain transaction history for Bitcoin, Ethereum, Bitcoin Cash, Dash, Litecoin, Zcash, Theta, Hedera Hashgraph, Band Protocol, Polygon, XRP, and Dogecoin. Companies will also receive assistance in driving co-development and integration into Google’s robust partner ecosystem, including participating in the Google Cloud Marketplace. Digital Assets Team will also help in embracing joint go-to-market initiatives with ecosystem partners where Google Cloud can be the connective tissue between traditional enterprise and blockchain technologies. Google Cloud is also exploring opportunities where customers can make and receive payments using cryptocurrencies. According to Google, Blockchains and digital assets are changing the way the world stores and moves information. As the technology becomes more mainstream, companies will need scalable, secure infrastructure to grow their businesses and support their networks. Working principles Consistent with Google Cloud’s core business: To pursue blockchain projects and partnerships that align with their mission and our expertise. User trust and governance: Maintain their commitment to users through a robust focus on privacy and user trust, as well as an uncompromising focus on compliance with applicable laws. Network-agnostic: Google’s infrastructure will seek to preserve optionality of networks for the benefit of users.","excerpt":"Currently, blockchain and distributed-ledger-based companies like Hedera, Theta Labs, and Dapper Labs are built on top of Google Cloud for scalability, flexibility, and security.","categories":["AI News"],"tags":["Blockchain","Google Cloud"],"author_name":"Meeta Ramnani","publish_date":"2022-01-28T16:05:56","publication_year":"2022","word_count":341,"keywords":["Go","Google Cloud","Blockchain","AWS","AI","cloud_platforms:AWS","programming_languages:R","Scala","Git","cloud_platforms:Google Cloud","Rust","R"],"extracted_tech_keywords":["AI","AWS","R","Go","Rust","Scala","Git","cloud_platforms:AWS","cloud_platforms:Google Cloud","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/google-cloud-announces-dedicated-digital-assets-team\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10089169,"title":"Microsoft has Expansionist Intention and Others Should Feel Scared","content":"Satya Nadella keeps wanting more. In February, Nadella sat for an interview with The Verge after announcing the new version of their Bing search engine. The Hyderabad-native expressed his admiration for Google, the definitive search engine, saying what he really wanted was for Pichai and co. to ‘come out and show that they can dance’. A decade ago—after people were writing off the tech giant as a dinosaur of bygone times—Nadella’s appointment in 2014 switched things up. The 2023 Microsoft is doing a lot and doing it everywhere. Microsoft in Enterprise Microsoft’s productivity tools and business processes segment made USD 15.9 billion in revenue in the second-quarter of FY 2022, which comprised 31% of the company’s total revenue. This amounted to an increase of 19.3% from the quarter a year earlier and Microsoft is pushing it harder to squeeze out more money in this segment. Just last week, the company unveiled a new Dynamics 365 Copilot which will integrate a generative AI chatbot for their users to be able to draft emails, summarise their Teams meetings and build marketing and sales campaigns. Since Microsoft’s multi-year investment in OpenAI, the tech giant has been pushing the Sam Altman-led startup’s generative AI tools into all its services. While Microsoft has already been crowding an already busy ERP and CRM segment with competitors like Salesforce, the generative AI twist has added more heat. Days later, Marc Benioff’s Salesforce was forced to announce a partnership with who else but OpenAI. Called Einstein GPT, the software will help salespersons, customer service agents and marketers. Microsoft also bought LinkedIn for a hefty USD 26 billion in 2016 after Nadella took over. Microsoft in Search A couple of days ago, Microsoft’s search engine ‘Bing’ crossed the 100 million daily active users mark just a month after the launch of its AI-powered Bing chat. This is only a miniscule fraction of Google’s more than a billion daily active users but the steady growth is enough to make the search giant “dance” as Nadella promised. The search and chatting experience package is giving Microsoft a competitive edge for now. Google is only lagging behind slightly compared to Microsoft as it also announced its AI chatbot called ‘Bard’ last month. Bard’s grand debut was marred by a misinformation flub, forcing the company to delay its availability until it is ready. In a blog post sharing the happy news, Microsoft’s head of consumer marketing, Yusuf Mehdi also stated that Microsoft Edge had also helped push Bing. Microsoft is known to constantly send its Edge updates and Windows updates while even prompting users to not download Google Chrome. While not every story around Bing is flattering, users are mostly entertained by its at times wild responses. As long as it’s not taken too seriously, Microsoft is also gunning after revenue from the ad market. Last month, the company’s CVP of finance laid out plans showing how the company planned to grow its digital ad revenue to around USD 500 billion. Notably, Microsoft’s advertising business has already grown to USD 18 billion in revenue over the year as compared to USD 10 billion in the past fiscal year. While Google did make USD 200 billion in advertising revenue last year, let’s not forget that Google is heavily reliant on its search business and its portfolio isn’t nearly as diversified as Microsoft’s. Microsoft in Cloud Microsoft’s cloud business may not be as critical to the company as AWS is to Amazon but the segment is a serious money-maker in every way. Just like AWS, Azure has seen a whopping growth over the past decade with its second-quarter 2022 revenue from Intelligent Cloud standing at USD 18.3 billion—35% of its total revenue. Following its launch in 2010, Microsoft’s Intelligent Cloud segment included other cloud services like SQL Server, Windows Server and Visual Studio. The services include data storage, databases, networking, web and mobile applications. Where AWS is expected to plateau soon and Google Cloud is still struggling to meet profitability, Microsoft Azure can have a competitive advantage because of its massive customer base. Azure also covers 60 geographical regions globally, which is far bigger than AWS and Google Cloud both of which have half as much. Gaming The dramatic saga of Microsoft’s proposed acquisition of Activision Blizzard could bring it big money from the Call of Duty games. Last reported by Reuters, the USD 68.7 billion deal is likely to be approved by EU regulators. For the first time in a long time, it looks like Microsoft isn’t the one chasing but is rather giving others a run for their money. Just last month, in an interview with Product Lead at Roblox Peter Yang, Microsoft VP Product for Microsoft Teams, Amit Fulay, spoke about how much Microsoft had changed under Nadella. Fulay, who had formerly worked at the software giant, had re-joined after 13 years. So, what was the difference? ‘If I can sum it up in one line—Microsoft no longer has hubris.’","excerpt":"While Microsoft has already been crowding an already busy ERP and CRM segment with competitors like Salesforce, the generative AI twist has added more heat.","categories":["Global Tech"],"tags":[],"author_name":"Poulomi Chatterjee","publish_date":"2023-03-11T16:03:20","publication_year":"2023","word_count":828,"keywords":["Go","AWS","OpenAI","AI","R","Git","RAG","generative AI","SQL","Azure"],"extracted_tech_keywords":["AI","generative AI","OpenAI","RAG","AWS","Azure","R","SQL","Go","Git"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/microsofts-sudden-omnipresence\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10054131,"title":"AI-Driven Industrial Robot Maker Haber Raises $20 mn In Series B Funding","content":"Pune-based Elixa Technologies Pvt Ltd, which operates Haber, an AI or artificial intelligence-driven robot maker, recently announced that the company has raised a $20 million (around Rs 148 crore) Series B round led by Ascent Capital. The round also saw participation from Accel, Elevation Capital, Beenext, Temasek partner Mukul Chawla and the founders of GreyOrange Samay Kohli and Akash Gupta, another Robot-as-a-Service company. The company said it will use the fresh funding to expand across new geographies and industries; it aims to save over 400 billion litres of water in the next two years. It also plans to deploy over 10,000 eLIXAs, Haber’s flagship product, in the next five years. Founded by Vipin Raghavan and Priya Venkat, Haber helps companies to reduce operating costs, as well as water and energy footprints by automating processes, utilities, and water systems. They drive robotic sampling, real-time testing, statistical model-driven data analysis, and automate the dosing of chemicals. Haber primarily works with companies in the manufacturing segment. “With a deep understanding of Artificial Intelligence (AI) and the functioning of industries, Haber has been able to put together a solution that automates the entire process. Our flagship product, eLIXA, was launched in 2017, which enables our customers to meet their sustainability and profitability goals,” said Vipin Raghavan, founder and CEO of Haber. Raja Kumar, founder and CEO of Ascent Capital, said, “Industrial automation is the next big global opportunity. Haber has a compelling value proposition, with its superior proprietary technology.” Raghavan serves as Chief Executive Officer, while Priya Venkat is the Chief Operating Officer. Before starting Haber, both founders worked in Nalco Water India, a water, hygiene, and infection prevention services and solutions company. To date, Haber has raised $27 million.","excerpt":"The company said it will use the fresh funding to expand across new geographies and industries; it aims to save over 400 billion litres of water in the next two years. It also plans to deploy over 10,000 eLIXAs, Haber’s flagship product, in the next five years.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","Data Science","Data Scientist","Deep Learning","Machine Learning","robot","Robotic Process Automation","Robotics"],"author_name":"Victor Dey","publish_date":"2021-11-24T13:29:20","publication_year":"2021","word_count":286,"keywords":["Go","API","funding","artificial intelligence","programming_languages:R","AI","robot","Machine Learning","Robotics","RAG","automation","Robotic Process Automation","Aim","Deep Learning","Data Science","Data Scientist","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","Aim","RAG","R","Go","API","automation","funding","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/ai-driven-industrial-robot-maker-haber-raises-20-mn-in-series-b-funding\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10008540,"title":"Free Online Resources To Get A Comprehensive Understanding Of TinyML","content":"Being one of the fastest developing deep learning aspects, TinyML has immense possibilities in areas where it is required to deploy a model that works on small and low power devices. Starting from imagery micro-satellite, tracking wildlife for conservation to detecting crop ailments, animal illnesses and predicting wildfires, TinyML comes with many possibilities. Not only it enables low-latency inference at edge devices consuming less power but also allows ML applications to run at edge intelligence. ‘OK, Google’ has been one of the renowned applications of TinyML, that works on everybody’s smartphones. With such applications in hand, along with software frameworks like TensorFlow Lite for Microcontrollers, it has become extremely easy to deploy TinyML models. Along with that, Tiny ML also comes with capabilities to support private machine learning applications and can work without any internet connection on edge devices. With TinyML gaining massive traction in the industry, it has become critical for ML practitioners to get a comprehensive understanding of this aspect of deep learning. Here, we will share free online resources to get hands-on TinyML. Also Read: What Are The Challenges Of Establishing A TinyML Ecosystem 1| The “Hello World” of TinyML By: O’Reilly About: This tutorial will focus on building and training a TinyML model from scratch and then integrate the same into a simple microcontroller program. The tutorial will be using Keras to train the tiny model, and the learners will be able to train, evaluate and convert a TensorFlow deep learning network that can produce accurate output. It will start with obtaining a simple dataset, followed by training the deep learning model and evaluating its performance and then convert the same to the run-on device. Once the code has been built in binary, it can then be deployed on to a microcontroller. Click here to learn. 2| Training a Model for Arduino in TensorFlow By: DigiKey About: This tutorial will teach how to create a neural network that can predict an accurate output of the sine function. Once that is done, the model will then be converted into a TFLite Model and examine the same using Netron. With this tutorial, learners will be able to load the model and use it for inference with the TFLite for Microcontrollers library. Along with that, it will also help learners to get started with Colab for deploying microcontrollers. Click here to learn. 3| Easy TinyML on ESP32 and Arduino By: Hackster About: This tutorial will show the easiest way to deploy TensorFlow Lite models onto ESP32 using Arduino IDE with just a few lines of code. With this, the learners will get a brief recap of TinyML knowledge and the steps that are required to implement TF models to a microcontroller. Further, this tutorial will also introduce learners to a tiny library to facilitate the deployment in the Arduino IDE — EoquentTinyML. This will help learners to kick start their next TinyML project on ESP32. Click here to learn. 4| Train A TinyML Model To Recognise Sounds By: EdgeImpulse About: Train A TinyML Model To Recognise Sounds is a tutorial provided by EdgeImpulse that will focus on tracking animal behaviour, especially sounds using only 23KB of RAM. This tutorial can be helpful for those who are working on animal tracking projects and wish to know the number of times the lions roared in a day. Counting the roars can be a tedious task, thus, to make the work easy, this tutorial will teach how to train a machine learning model to recognise lion roars in the recordings using a set of labelled data, advanced algorithm, and TinyML model. Click here to learn. 5| Cough Detection with TinyML on Arduino By: Arduino About: This tutorial will focus on building a cough detection system for the Arduino Nano BLE Sense using TinyML and Edge Impulse. The tutorial will start with showcasing the use of Edge Impulse machine learning on Arduino Nano BLE Sense to identify the presence of coughing sound in the audio set. With a coughing dataset built on sample audios, this TinyML model will be able to run on 20KB RAM with Nano BLE Sense. This tutorial will start with a basic understanding of software development and Arduino, and then will use Edge Impulse to create the cough detection model. Click here to learn. Also Read: Do Developers Need Theoretical Knowledge For AI Programming? 6| Tutorial on Micro-Kernel Based Hardware Acceleration By: Manu Rastogi About: This YouTube tutorial is provided by Manu Rastogi, a machine learning engineer at Apple, where he will talk about the matrix multiplication micro-code. With the increased traction of deep learning, there has been a significant competition amongst different hardware vendors to provide the most energy-efficient solutions. However, the critical element of model deployment at the edge is the mico-kernels that help in the data movement and the computation of these networks on hardware. In this tutorial, learners will be able to understand the various trade-offs between different optimisation strategies and extend these principles to neural networks. Click here to learn. 7| Tiny ML on Arduino By: Arduino About: Tiny ML on Arduino is a tutorial provided by Arduino on Gesture recognition. Starting from setting up a Python environment, this tutorial will showcase how to upload data, train a neural network, build and train the model and then run it with the test data. Provided by Sandeep Mistry of Arduino and Done Coleman of Chariot Solutions, this tutorial will use Arduino Nano 33 BLE Sense to convert motion gestures to emojis. Click here to learn. 8| AutoML for TinyML with Once-for-All Network By: Song Han About: AutoML for TinyML with Once-for-All Network is another YouTube tutorial which is provided by Song Han, an assistant professor in MIT’s Department of Electrical Engineering and Computer Science. This tutorial will address the issue of having efficient inference across many devices and the resource constraints that arise on edge devices. The once-for-all network is a network that surpasses MobileNetV3 and EfficientNet by a large margin, and in this tutorial, the learners will be able to use TinyML with Once-for-All network. Click here to learn. 9| AI Speech Recognition with TensorFlow Lite for SparkFun Edge By: CodeLabs About: Offered by CodeLabs, this tutorial will start with providing an introduction of machine learning on microcontrollers, TensorFlow Lite for microcontrollers and SparkFun Edge. This will be then followed up by compiling the sample program for SparkFun Edge on the computer, deploying the program and then making the required changes to deploy the program. This tutorial comes with a prerequisite of having SparkFun Edge Board and USB-C Serial Basic programmer. Click here to learn.","excerpt":"Being one of the fastest developing deep learning aspects, TinyML has immense possibilities in areas where it is required to deploy a model that works on small and low power devices. Starting from imagery micro-satellite, tracking wildlife for conservation to detecting crop ailments, animal illnesses and predicting wildfires, TinyML comes with many possibilities. Not only […]","categories":["AI Features"],"tags":["online network graph","TinyML"],"author_name":"Sejuti Das","publish_date":"2020-09-29T11:00:05","publication_year":"2020","word_count":1101,"keywords":["TinyML","machine learning","Keras","TPU","AI","online network graph","neural network","ML","Colab","Python","deep learning","TensorFlow"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","neural network","TensorFlow","Keras","Colab","TPU","Python"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/free-online-resources-to-get-a-comprehensive-understanding-of-tinyml\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10089217,"title":"Plot Twist: Social Media Adopts Subscription, OTT Goes Ad-ways","content":"In 1972, Petula Clark sang ‘the other man’s grass is always greener’. Decades later, the chartbuster seems to be speaking of OTT and social media platforms that are planning to adopt each other’s business models. Besides advertising, which has forever been their sole business model, social media platforms are now toying with the subscription model for users. On the other end of the spectrum are OTTs, largely dependent on subscriptions, that are now considering adding advertisements to their videos – a strategy to open new sources of revenue. Twitter set the precedent for subscription on social media. One of the major announcements that Musk made after taking over Twitter was to monetise the ‘blue tick’. Soon, TikTok joined the party with its ‘live subscription’ with access to perks like a subscriber-only chat, creator-specific emotes, and badges that differentiate them from non-subscribers. Recently, Meta too introduced a verified feature for Facebook and Instagram, emphasising that it is aimed at enhancing authenticity and security across its services. However, Over-the-Top (OTT) platforms that traditionally ran on subscriptions are gradually moving towards ad integration for added revenue. Platforms such as Hulu, Peacock, Voot, Zee5, and Disney+Hotstar now provide both subscription-based and advertisement-based models. Users can access a limited category of content for free with a few advertisements thrown in, but to access premium content like international TV shows and movies, they must pay for the ad-free premium plan. What’s more? Even AppleTV plans to monetize its original video content with advertising – and so is Netflix. Driving Force Behind OTT’s Transition In India, subscription video-on-demand (SVOD) platforms are experiencing a slowdown in paid users with many turning to advertising-based video-on-demand (AVOD). In order to hoard subscribers, some platforms are offering free or ad-supported versions of their service to attract new users who might not want to pay for subscription. According to media consulting company Ormax Media, India has over 423 million OTT users but the number of paid users still remains low. Less than 10% of Indian OTT consumers go for subscription, as per a report by EY. Kishore AK, chief technology officer at Zee5, told AIM, “Our content is very localised. So, while OTTs like Netflix are moving towards ads, we always knew that the audience would prefer both ads and subscriptions.” This ad-based approach of OTT platforms is supported by users as well. There has been a shift in user behaviour with consumers willing to choose ads over paying a higher subscription. According to a survey by Future Today, a majority of consumers prefer ad-supported TV (67%) over an ad-free subscription (33%). About 60% of consumers, who saw relevant ads, enjoyed their viewing experience. Source: Future Today via BusinessWire So, OTT platforms are diversifying their revenue streams by incorporating ads with subscriptions. Meanwhile, social media platforms are following OTT’s traditional formula of subscription. Subscription for Social Media Users may be okay watching ads on OTT platforms, but they have a low tolerance for ads when it comes to social media. As per a CMO survey, nearly 74%, that is, 3 out of every 4 users think there are way too many ads on social media, while 44% find them irrelevant. The same users are attracted to social media platforms because it is free – it is like a big playground for everyone to unleash their creativity. Social media platforms have two markets: the audience market and the advertiser market. The audience market is where individuals use the platform, while the advertiser market is where advertisers use consumer insights for targeted advertising. The connection between these markets is user data. “Regulatory measures like the Banning Surveillance Advertising Act in the United States and the EU’s Digital Services Act are making it harder for platforms to use personal information for advertising, causing a revenue slowdown. Platforms are exploring new revenue streams, including pricing services in the traditionally non-priced audience market,” The Dialogue’s founding director Kazim Rizwi told AIM. The Dialogue is a think-tank working on technology policy. Social media has traditionally relied on revenue from ads and selling user data, but with increasing competition and ad blockers, they are turning to subscription models to offer personalised experiences and foster customer loyalty. This allows for predictable revenue and premium content for subscribers. Mohan Gupta, senior director of product at Moj, told AIM, “With a secondary revenue stream and exclusive content, subscriptions deepen interactions and cultivate loyalty between creators and followers. As advertising streams prove unstable, subscriptions offer a unique differentiation and help platforms stay relevant amidst the influx of new players.” Developed by the Indian social media platform ShareChat, Moj app lets users create and share short-form videos with various visual effects and filters. Not just social media, subscription models are becoming a popular way for publishers, authors, and content creators to monetize their offerings and retain ownership of their content while offering exclusive member-only content. “One of the main advantages of this model is predictable revenue, independent of factors such as ad placement or user engagement. Subscription-based businesses can rely on a steady stream of revenue from loyal customers, leading to increased customer lifetime value,” said Sumit Ghosh, chief and co-founder, of Chingari, a social media platform that offers short-form video content in Indian languages, in conversation with AIM. Chingari has also started a subscription model for its users and creators to withdraw the daily income they generate through the ‘Gari mining program’ which has now become an integral part of their revenue stream. Another benefit of the subscription model is the ability to make it more personalised. Companies can collect data on subscriber usage habits, preferences, and behaviours, allowing them to tailor their offerings to each individual subscriber. This not only enhances the customer experience but also helps companies to retain subscribers over the long term.","excerpt":"According to a survey, a majority of consumers prefer ad-supported TV (67%) over an ad-free subscription (33%)","categories":["IT Services"],"tags":["Meta","tiktok","Twitter (X)"],"author_name":"Shritama Saha","publish_date":"2023-03-13T13:00:00","publication_year":"2023","word_count":963,"keywords":["tiktok","Go","Meta","programming_languages:R","AI","programming_languages:Go","Git","Aim","ViT","Twitter (X)","R"],"extracted_tech_keywords":["AI","Aim","R","Go","Git","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/plot-twist-social-media-adopts-subscription-ott-goes-ad-ways\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":64751,"title":"Matic Network, Harmony &#038; æternity Join Telangana Blockchain District’s Accelerator Program As Official Platform Partners","content":"T-Block Accelerator, the inaugural accelerator program under the aegis of the Telangana Blockchain District has announced its partnership with three of the leading global blockchain protocol companies- Matic Network, Harmony and æternity as the platform partners. This is a first of its kind public blockchain accelerator organised by Telangana government and Tech Mahindra and run by IBC Media, an innovation management company. T-Block Accelerator is aimed at startups that have a strong blockchain use-case to accelerate the startup growth and therefore, to contribute to the growth of the overall industry. The platform integrators are fast, secure, scalable, next-gen blockchain protocols that are equipped to help startups build scalable and decentralised solutions to everyday problems. The Matic Network will provide a high scale platform for decentralised applications. The æternity Blockchain, engineered by Yanislav Malahov provides scalable smart contracts for decentralised applications. The Harmony Protocol, secure blockchain for decentralised applications, will give fintech startups the ability to create and transact digital assets on its global ledger with instant confirmation and near-zero transaction fees. With the support and expertise of these protocols as innovation enablers in the blockchain realm, T-block will provide the Indian startups that make it into the boot camp phase with technical support to build their solutions on their platform, go-to-market help in Europe, technology implementation support, and grants on a case-on-case basis. Jaynti Kanani – CEO & Co-founder, Matic Network, commented, “As a part of the partnership, we will extend the Matic Network and our in-house expertise to mentor and train the participating startups. The startups will be familiarised and trained to use the platform based on their projects’ requirements.” He also added that this is just a facet of Matic’s involvement. “In addition to mentoring the startups, our in-house experts will also join the judges’ panel to select the startups that will go on to the next level of acceleration and aid with financial grants to support the startups through their journey,” said Kanani. Nikola Stojanow, CEO of AE Ventures and CBO of æternity, added, “India has a booming blockchain ecosystem, and we are happy to be a part of this growth story. Having concluded the first edition of Starfleet India Accelerator recently, we are excited to join this initiative by the Telangana government and Tech M for T-Block and look forward to supporting the startups with our GMT access in the European countries.” “At Harmony, we are confident of India’s technology prowess and are excited to enable the creation and innovation of highly-scalable & secure FinTech solutions from India through our partnership with T-Block Accelerator. We are happy to support them with the technical expertise to build products & applications on our platform and aid them with grants that will provide the necessary impetus for their growth in this journey.” Sahil Dewan, Co-Founder, Harmony commented. These collaborations with prominent players like the Matic Network, æternity, and Harmony for initiatives like T-block will truly make an impact on Indian startups and innovators who are making their way through this realm and strengthen Telangana Government’s endeavour to make Telangana as the ‘Blockchain Capital of the World’. Rama Devi Lanka, the director of emerging technologies & officer on special duty, ITE&C department, the government of Telangana said: “Building synergies between all the stakeholders in this ecosystem will propel India’s blockchain revolution, and we are happy that our T-Block Accelerator is partnering with these platform protocols to strengthen this ecosystem.” She further said, “As the first government-led blockchain accelerator, this partnership is a true testimony to how a synergy between the government body and private players can support innovators in their entrepreneurial journey and will help build, scale, and create amazing opportunities for startups. We’re excited to see how the startups will benefit from these partnerships.” With go-to-market support from Tech Mahindra, infrastructure support from the State Government of Telangana, innovation deal-flow support from IBC Media, and platform support from the Matic Network, æternity, and Harmony, the startups to be shortlisted are stepping into a growth-fueled environment. Launched in February 2020, the T-Block Accelerator has completed its deal flow, and selected startups will undergo a one-week boot camp, followed by a four-month-long intensive training program.","excerpt":"T-Block Accelerator, the inaugural accelerator program under the aegis of the Telangana Blockchain District has announced its partnership with three of the leading global blockchain protocol companies- Matic Network, Harmony and æternity as the platform partners.  This is a first of its kind public blockchain accelerator organised by Telangana government and Tech Mahindra and run […]","categories":["AI News"],"tags":["Blockchain","blockchain companies","telangana","telangana government"],"author_name":"Sejuti Das","publish_date":"2020-05-07T15:48:13","publication_year":"2020","word_count":694,"keywords":["Go","API","Blockchain","AI","blockchain companies","telangana","innovation","Scala","telangana government","Git","Aim","GAN","R","startup"],"extracted_tech_keywords":["AI","Aim","R","Go","Scala","Git","API","GAN","innovation","startup"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/matic-network-harmony-aeternity-join-telangana-blockchain-districts-accelerator-program-as-official-platform-partners\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10066009,"title":"Praxis launches ‘Data Science Student Championship’ &#8211; South Zone Challenge Edition","content":"The hackathon will start on May 12, 2022, and will run till May 31, 2022. Open only for students in the South zone, Andhra Pradesh, Karnataka, Kerala, Tamil Nadu, and Telangana. It also allows the students to represent their college in an interstate competition and win prizes worth INR 50K. ​​Crack this Business Problem: Renting a house or an apartment is never easy. Whether you are a college student or a working professional, renting a place always seems like a daunting task that is often impulsive or risky. Rent is influenced by several factors. In this challenge, participants need to use data science methods, machine learning, and hyperparameter tuning to predict house rents. The hackathon starts on May 12, 2022, at 6:00 PM Click here to participate in the hackathon Eligibility Criteria: The participants must be pursuing the following degree programs BTech\/ BE- All streamsBSc\/ MScBCA\/ MCAMTech The college\/ university must be based in the South Zone (Andhra Pradesh, Karnataka, Kerala, Tamil Nadu, and Telangana)Valid ID card from the respective college\/ UniversityParticipants can compete as individuals or in teams of two. Dataset details Train.csv – 1,34,683 rows x 21 columnsTest.csv – 57,722 rows x 20 columns Sample Submission.csv — Please check the ‘Evaluation’ section on MachineHack website for more details on generating a valid submission Data description: Property ID RoomLayout type Property type Locality Price Area Furnish type Bathroom City Parking spaces Pet-friendly Power backupWashing machineAir conditioner Geyser\/solarSecurity deposit Neighbourhood CCTV\/security The hackathon starts on May 12, 2022, at 6:00 PM Click here to participate in the hackathon What are you competing for? Cash prizes worth Rs. 50 K up for grabs – to be shared among the top three teams. All teams that complete the challenge will get participation certificates First Prize: INR 25,000Second Price: INR 15,000Third Prize: INR 10,000 Disqualification The championship will be open only to a student from the South zone, i.e., Andhra Pradesh, Karnataka, Kerala, Tamil Nadu, and Telangana.If any of the details entered are found incorrect, Analytics India Magazine and Praxis reserve the right to disqualify any participant.Any external dataset usage is strictly prohibited. The participants will be disqualified if found using any external dataset. Submission guidelines: Sklearn models support the predict() method to generate the predicted values.The participant should submit a .csv file with 57,722 rows with 1 column [“price”]. The submission will return an Invalid Score if you have extra rows or columns.The file should have exactly 1 column. Note: Do not shuffle the sequence of the test seriesIf you are using pandas, use this submission code: submission_df.to_csv(‘my_submission_file.csv’, index=False) Evaluation criteria: The submission will be evaluated using the RMSE metric. One can use ‘np. sqrt (Mean Squared Error)’ to calculate the same. [mean_squared_error(y_true, y_pred, squared=False)]This hackathon supports private and public leaderboards.The public leaderboard is evaluated on 30 per cent of test data.The private leaderboard will be made available at the end of the hackathon, evaluated on 100 per cent of the test data.The Final Score represents the score achieved based on the best score on the public leaderboard. The hackathon starts on May 12, 2022, at 6:00 PM Click here to participate in the hackathon Championship Timeline May 12, 2022: Data Science Student Championship – South Zone Challenge begins. Phase #1: The participants download the dataset, solve, and submit their solutions on the MachineHack platform. Phase #2: The Leaderboard is frozen with the Top 50 teams on May 31, 2022. The Top 50 will need to submit a presentation on their problem statement analysis, solution approach, business outcome, and real-world impact by June 06, 2022.The Final Top 6 teams will be selected on the basis of their Leaderboard position and the quality of the presentation by June 10, 2022. The Top 6 need to work on bench data released and prepare for the final jury round. Phase # 3: The Top 6 teams will arrive in Bangalore city on June 22, 2022, to present their final solution for the Final Jury Round. The top three winners of this Data Science Student Championship – South Zone Challenge will receive the cash awards.All Top 6 finalists will receive commendation certificates. The winner will be judged on the following in the final jury round: 30% Statistical and ML Algorithm30% Innovation + Creativity20% Business Outcome\/ Impact20% Presentation + Communication Other details: Here is some of the important information regarding participation in the final jury round by the Final Top 6 teams: Teams will be given return bus\/ train tickets from the nearest metro city of their residence. Please note that meals during the journey will not be included in the ticket, and any expenses towards the same need to be borne by the travelling teams.The shortlisted teams will need to manage and bear travel expenses, logistics, etc., between their residence and the bus\/ train station of the nearest metro city from where they need to board\/ alight from the bus\/ train.We will not be responsible for the accommodation of the shortlisted teams or any fellow co-passengers who may accompany the shortlisted teams.Food and beverages will be arranged for the shortlisted teams on the day of the jury round at the contest venue.Only the shortlisted teams will be provided entry to the jury round to be held in Bangalore and during the time mentioned in the session schedule.All communication and queries are to be made only to the contact details provided on the MachineHack platform. The hackathon ends on May 31, 2022, at 6:00 PM Click here to participate in the hackathon","excerpt":"Calling all under-graduate and postgraduate students from engineering colleges and universities to the first-ever ‘Data Science Student Championship’ – South Zone Challenge, brought to you by MachineHack, India’s leading Analytics platform & Praxis Business School, India’s #1 Data Science institution.","categories":["AI Trends"],"tags":["data science hackathon","data science student championship","Hackathon","hackathon for data scientists","hackathons in India","Hackathons India","Machinehack Hackathon","machinehack praxis hackathon","New Hackathon For Data Scientists","Praxis Business School","Praxis hackathon","Weekend Hackathon"],"author_name":"Amit Naik","publish_date":"2022-04-30T12:00:00","publication_year":"2022","word_count":915,"keywords":["Weekend Hackathon","New Hackathon For Data Scientists","hackathon for data scientists","R","Pandas","data science","hackathons in India","data science student championship","RAG","data science hackathon","analytics","Go","machine learning","AI","ML","Hackathon","machinehack praxis hackathon","Machinehack Hackathon","GAN","Hackathons India","Praxis Business School","Praxis hackathon"],"extracted_tech_keywords":["AI","machine learning","ML","data science","analytics","Pandas","RAG","R","Go","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/praxis-launches-data-science-student-championship-south-zone-challenge-edition\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10118166,"title":"Elon Musk’s xAI Unveils Grok-1.5 Vision, Beats OpenAI’s GPT-4V","content":"Elon Musk’s AI startup, xAI has introduced Grok-1.5V, a  first-generation multimodal model. In addition to its strong text capabilities, Grok can process a wide variety of visual information, including documents, diagrams, charts, screenshots, and photographs. Grok-1.5V will be available soon to early testers and existing Grok users. Grok-1.5V’s notable feature is its ability to understand real-world spatial concepts, surpassing other models in the RealWorldQA benchmark—an important measure of a model’s practical grasp of physical environments. In a comparative analysis against leading models like GPT-4V, Claude 3 Sonnet, Claude 3 Opus, and Gemini Pro 1.5, Grok-1.5V shows competitive advantages across several benchmarks, highlighting its versatility and strength. One of Grok-1.5V’s standout features is its ability to translate complex visual information into executable code. For example, when given a flowchart depicting a guessing game, Grok-1.5V easily converts it into Python code, showcasing its practical application in problem-solving scenarios. Looking forward, the developers of Grok-1.5V anticipate significant improvements in multimodal capabilities across images, audio, and video, signaling a promising path towards building beneficial Artificial General Intelligence (AGI) that comprehensively understands and interacts with the universe. Grok-1.5V follows the recent introduction of Grok-1.5 by xAI, featuring enhanced reasoning capabilities and a context length of 128,000 tokens. Grok-1.5 boasts notable improvements, particularly in coding and math-related tasks. It beats Mistral Large on various benchmarks including MMLU, GSM8K and HumanEval.","excerpt":"One of Grok-1.5V’s standout features is its ability to translate complex visual information into executable code.","categories":["AI News"],"tags":["Elon Musk","XAI"],"author_name":"Siddharth Jindal","publish_date":"2024-04-13T09:34:29","publication_year":"2024","word_count":224,"keywords":["Gemini Pro","AI","R","ML","RPA","Elon Musk","XAI","Python","GPT","xAI","startup"],"extracted_tech_keywords":["AI","ML","Gemini Pro","xAI","Python","R","GPT","XAI","RPA","startup"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/elon-musks-xai-unveils-grok-1-5-vision-beats-openais-gpt-4v\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10058721,"title":"How to query data for Computer Vision tasks using VisionKG?","content":"We mostly know SQL as a query language that facilitates us to query relational data from almost any database, but when it comes to gathering data for computer vision-related tasks, we have to depend on a distinct host of those data sets such as ImageNet, COCO, etc. Recently researchers have proposed a framework called VisionKG, which can integrate those datasets seamlessly. So in this article, we will discuss the VisionKG in detail and will see how it can query the dataset like COCO and ImageNet. Below are the major points to be discussed in this article. Table of contents The general idea of VisionKGHow it queries the data?Practical example Let’s first discuss what this framework brings out. The general idea of VisionKG It is a unified dataset framework in data-centric AI. Not only are diverse datasets from one AI domain integrated and linked together within this framework, but so are datasets from multiple AI areas. Existing resources, such as ConceptNet and Wikidata, have a similar purpose in that they integrate data from various sources and make it public, but they instead focus on certain application domains, and none of them is linked to databases in other areas, such as computer vision. Scene graphs, on the other hand, were introduced in the computer vision fields to model the relationship between identified items in photos. They lack cross-domain compatibility, however, and cannot be queried using common query languages. As a result, current resources should be streamlined, and new resources should be easily incorporated. It is quite advantageous, for example, it can aid in the prevention of distribution shifts and the development of more robust models for training and testing. Furthermore, the transition from model tinkering to a deep knowledge of data necessitates that datasets be better organized. Furthermore, the term “data” should be expanded by this approach to include not only training data but also abstract information, such as commonsense or causal relationships. In a nutshell, it is a single framework for various datasets that allows them to be readily merged and queried, for example, using standard query languages. How it queries the data? To realize the concepts stated above, researchers created VisionKG, a unified knowledge graph for CV datasets (e.g, COCO). VisionKG is a knowledge graph based on the Resource Description Framework (RDF) that contains RDF statements describing the metadata of pictures and the semantics of their annotations. The World Wide Web Consortium (W3C) recommends RDF as a standardized data model for semantic data integration and as a formal representation for shared human-machine understanding. As a result, RDF may be used to represent numerous semantic structures of prominent label taxonomies like Wordnet, ConceptNet, and Freebase, which are utilized in a variety of CV datasets including Imagenet and OpenImage. Source The process of creating VisionKG is depicted in the diagram above. It begins by gathering CV datasets and extracting annotation labels from them. 1. It follows the Linked Data principles and uses the RDF data model to create a unified data model for the annotation labels and visual features. Uniform Resource Identifiers (URIs) are used to name data entities (such as images, boxes, and labels). The RDF data model allows for the expression of data using triples of the form, <subject, predicate, object>. To describe “an image contains a bounding box for a person” in the COCO dataset, we must first assign unique URIs for the image and the bounding box, e.g., vision.semkg.org\/img01 and vision.semkg.org\/box01, to create the following triples: <img01, hasBox, box01>,<box01, hasObject, obj01>, <obj01, rdf: type, Person>. Predefined predicates include hasBox, hasObject, and rdf:type, with rdf:type expressing that an object\/image belongs to a specific class\/type in the knowledge base, such as Person. We can also add metadata and semantic annotations to the images, such as where the images came from or the relationships between the boxes in an image (as shown in the 2nd step on the above figure). Datasets and analysis, in particular, can be performed using rich semantic query languages such as SPARQL. The SPARQL query language allows users to describe queries using RDF statements that are similar to SQL statements. The first application is to obtain mixed-datasets in an elegant manner using VisionKG. For example, instead of the more complex query (as shown in below figure snippet 3) that covers all possible cases: a pedestrian in KITTI or man in Visual Genome, one can query for images of people from COCO, KITTI, and Visual Genome using a simple query (as shown in below figure snippet 4). Source The diagram above depicts an example of label mapping in COCO, KITTI, and Visual Genome to knowledge base classes (depicted in 1 section). Labels are being expanded to reflect the class hierarchy (depicted in section 2). And two equivalent queries for retrieving images from the COCO, KITTI, and Visual Genome datasets that contain Person (depicted in section 3,4 section). Practical example In this section, we’ll take a look at some query samples of VisionKG by which we can extract data on the fly by using Python. Let’s quickly set up the environment and import dependencies. from google.colab import output # install our vision api from google.colab import output # install our vision api !python -m pip install git+https:\/\/github.com\/cqels\/vision.git --force output.clear() # import SemkgAPI from vision_utils import semkg_api, data from skimage import io import matplotlib.pyplot as plt Next, we can define a variable called query which holds the query statements. Here our query is about getting images of cars and trucks. # Query string query_='''#Give me 100 images containing car and truck prefix cv:<http:\/\/vision.semkg.org\/onto\/v0.1\/> SELECT DISTINCT ?image WHERE { ?ann1 a cv:Annotation. ?ann1 cv:isAnnotationOfImage ?image. ?ann1 cv:hasAnnotatedObject ?obj1. ?obj1 cv:hasLabel \"car\". ?ann2 a cv:Annotation. ?ann2 cv:isAnnotationOfImage ?image. ?ann2 cv:hasAnnotatedObject ?obj2. ?obj2 cv:hasLabel \"truck\". ?image cv:hasLocalPath ?localPath. } LIMIT 100''' Next, the above-defined query needs to be passed in the VisionKG’s API as below. #query and return result result=semkg_api.query(query_) All the query outcomes are stored in the result variables, which is a dictionary of metadata of the queried image. Now we’ll plot some of the samples of queried images. #display sample images rows=3 cols=4 f, ax_arr = plt.subplots(rows, cols, figsize=(16,8)) for j, row in enumerate(ax_arr): for i, ax in enumerate(row): if j*cols+i < len(result['images']): image = io.imread(semkg_api.SEMKG_IMAGES_HOST + result['images'][j*cols+i]['image_path']) ax.imshow(image) ax.axis('off') f.suptitle(\"Sample images from the query result\", fontsize=16) plt.show() The result looks like this: Final words Through this article, we have seen a python-based framework that calls the computer vision-related data by using SPARQL which is a semantic query language used to retrieve and manipulate data as we have seen above. You can experiment with this language at https:\/\/vision.semkg.org\/. Further with this data, we can perform tasks like object detection, image classification, and more. One can check the official repository for more examples. References Fantastic Data and How to Query ThemSPARQLOfficial repositoryLink for above codes","excerpt":"In this article, we will discuss the VisionKG in detail and will see how it can query the dataset like COCO and ImageNet.","categories":["AI Trends"],"tags":["Computer Vision","Data Science","ImageNet","Machine Learning","Python"],"author_name":"Vijaysinh Lendave","publish_date":"2022-01-19T15:00:00","publication_year":"2022","word_count":1133,"keywords":["TPU","AI","ML","Machine Learning","computer vision","Python","Colab","object detection","Computer Vision","SQL","Data Science","Matplotlib","R","ImageNet"],"extracted_tech_keywords":["AI","ML","computer vision","Colab","Matplotlib","object detection","TPU","Python","R","SQL"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/how-to-query-data-for-computer-vision-tasks-using-visionkg\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10002896,"title":"Kerala Govt Launches AI Course For Graduates","content":"The Additional Skill Acquisition Programme (ASAP) of the Higher Education Department in Kerala has come up with a new artificial intelligence and machine learning course for graduates. It has been introduced with the aim of improving the employability of graduates in the state and equipping them with skills to meet industry requirements. The course aims to help students in the areas of gaming, speech recognition, language detection and robotics. The course designed is 776-hours long and is aimed at creating skilled professionals who can fill the demand in areas of data science, AI and ML. The program called the AI Machine Learning Developer Programme for Graduates will help gain practical knowledge and prepare for new-age jobs such as AI\/ML scientist, data scientists, ML engineer, robotics scientists, business intelligence developers, AI research scientists and more. The project also includes an internship which will impart multiple skills making a candidate comfortable with dealing with different types of structured and unstructured data to solve critical business problems using machine learning and deep learning. The course will help candidates to even explore opportunities as AI-Machine learning Consultant\/Specialist. The course also aims to boost entrepreneurship, ethics and social responsibility and motivate them for research in data science with specific emphasis on artificial intelligence. It can be availed by graduates including Engineering, MCA, BSc, MSc and other branches. With a blended learning model, the course will be taken by industry experts from various industries. The selection process is based on a common entrance examination. Some of the benefits of the course include 259 Learning Credits, 68 certificates, life skill and hard skills workshop, project mentoring by industry experts, hackathons, an internship at the prominent industry, a community meetup, certification of completion and more. Earlier this year, the Kerala government had also cleared an ordinance to upgrade the Indian Institute of Information Technology and Management – Kerala (IIITM-K) to the ranks of first Digital University in the state. It aimed at emphasising on disruptive technologies such as blockchain, IoT, AI and data analytics.","excerpt":"The Additional Skill Acquisition Programme (ASAP) of the Higher Education Department in Kerala has come up with a new artificial intelligence and machine learning course for graduates. It has been introduced with the aim of improving the employability of graduates in the state and equipping them with skills to meet industry requirements.  The course aims […]","categories":["AI News"],"tags":["ai certificates","certification for artificial intelligence"],"author_name":"Srishti Deoras","publish_date":"2020-07-21T11:23:51","publication_year":"2020","word_count":336,"keywords":["ai certificates","data science","Go","machine learning","artificial intelligence","AI","ML","certification for artificial intelligence","Aim","deep learning","analytics","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","data science","analytics","Aim","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/kerala-govt-launches-ai-course-for-graduates\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":25193,"title":"Army Chief Bipin Rawat Proposes Using Predictive Analytics With The Help Of AI","content":"Army Chief General Bipin Rawat at the Manekshaw Centre, New Delhi. (Image source: @PIB\/Twitter) In keeping with the current government’s emphasis on new tech, Army Chief General Bipin Rawat on Tuesday laid stress on the need to institutionalise data and carry out predictive analytics using artificial intelligence. Speaking at a seminar titled ‘GeoSpatial: A Force Multiplier for Defence and Industrial Security’ at the 11th edition of GEO Intelligence Asia, General Rawat said, “The armed forces are the repository of big data and there is a need to record and institutionalise the information and carry out predictive analytics using AI.” He also stressed on the need for collaboration with industry and academia to automate the Indian Army and emphasised the need to leverage the various technologies available in the Geo-Spatial field to find solutions for the challenges faced by them. Chief of Army Staff, General Bipin Rawat being briefed on latest trends in Geo-spatial intelligence, at the eleventh edition of Geo-Intelligence Asia 2018, on the theme Geo-Spatial: A Force Multiplier for Defence and Industrial Security’, in New Delhi https:\/\/t.co\/VjPdOzlfbi pic.twitter.com\/LVaVjb92gX — PIB India (@PIB_India) June 5, 2018 Lt Gen Anil Kapoor, Director General, Information System, later brought out the fact that data is being treated as the next oil. He mentioned that Indian Army is in the process of redefining and refining legacy applications into state-of-the-art software by using contemporary technologies. But this is not the first time that the Indian Army and the Government have talked about incorporating AI in the defence sector. Only a few weeks ago reports were floating about the Narendra Modi-led government kick-starting work on incorporating AI to prepare the Indian Army, Navy and the Air Force for next-generation warfare. The report had suggested that AI was to be used in the Army to enhance the operational preparedness of the armed force in a significant way. This would include equipping them with unmanned tanks, vessels, aerial vehicles and robotic weaponry, which would have extensive use in future wars. “The use of AI would add more teeth to the country’s military setup and it would be a big area considering the requirements of future warfare,” Dr Ajay Kumar, defence secretary in the Ministry of Defence, had said at the time.","excerpt":"In keeping with the current government’s emphasis on new tech, Army Chief General Bipin Rawat on Tuesday laid stress on the need to institutionalise data and carry out predictive analytics using artificial intelligence. Speaking at a seminar titled ‘GeoSpatial: A Force Multiplier for Defence and Industrial Security’ at the 11th edition of GEO Intelligence Asia, […]","categories":["AI News"],"tags":["AI (Artificial Intelligence)","Indian Army","Predictive AI"],"author_name":"Prajakta Hebbar","publish_date":"2018-06-06T11:07:57","publication_year":"2018","word_count":373,"keywords":["big data","Go","artificial intelligence","programming_languages:R","AI","R","predictive analytics","programming_languages:Go","RAG","analytics","Indian Army","AI (Artificial Intelligence)","Predictive AI"],"extracted_tech_keywords":["AI","artificial intelligence","analytics","RAG","predictive analytics","R","Go","big data","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/army-chief-bipin-rawat-proposes-using-predictive-analytics-with-the-help-of-ai\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10076737,"title":"Why Solving For Efficiency of Matrix Multiplication Such A Big Deal In Computing","content":"You must have come across matrix multiplication in school textbooks. But did you know how relevant it is in every aspect of our daily lives, from processing images on our phones and recognising speech commands to generating graphics for computer games? It is at the core of nearly everything computational. With DeepMind’s latest release AlphaTensor, an AI system, researchers shed light on a 50-year-old fundamental mathematics question of finding the fastest way to multiply two matrices. Today in @Nature: #AlphaTensor, an AI system for discovering novel, efficient, and exact algorithms for matrix multiplication – a building block of modern computations. AlphaTensor finds faster algorithms for many matrix sizes: https:\/\/t.co\/E18DezRPTL & https:\/\/t.co\/SvHgsa0SNV 1\/ pic.twitter.com\/bsVEAljvSQ— Google DeepMind (@GoogleDeepMind) October 5, 2022 “AlphaTensor discovered algorithms that are more efficient than state-of-the-art for many matrix sizes. Our AI-designed algorithms outperform human-designed ones, which is a major step forward in the field of algorithmic discovery,” DeepMind said in a statement. The advancement is an extension of AlphaZero, a single system that mastered board games (Chess, Go and Shogi) from scratch without human inputs. Moreover, the research reveals that AlphaZero is a powerful algorithm that can be extended beyond the domain of traditional games to help solve open problems in mathematics. The problem at hand Matrix multiplication is one of the simplest forms of mathematics but gets intensely complex when applied in the digital world. Anything that can be solved numerically—from predicting the weather to compressing data—typically uses matrices. For instance, you can read this article on your screen because its pixels are represented as a grid, and they refresh with new information faster than your eyes can track. Despite its omnipresent nature, the calculation is not very well understood. Moreover, nobody knows a quicker method of solving the problem because there are infinite ways to do so. DeepMind’s gameplan Breakthroughs in machine learning have helped researchers right from creating art to predicting protein structures. Increasingly, researchers are now using algorithms to become its own teacher and correct the flaws. The DeepMind researchers did what they do best—making AIs champions at games. The team tackled the matrix multiplication problem by turning it into a single-player 3D board game called ‘TensorGame’. The game is immensely challenging as the number of possible algorithms, even for small cases of matrix multiplication, is larger than the number of atoms in the universe. The three-dimensional board represents the multiplication problem and each move represents the next step in solving it. The series of moves made in the game, therefore, represents an algorithm. To play the game, the researchers trained a new version of AlphaZero, called ‘AlphaTensor’. Instead of learning the best moves to make in Go or chess, the system learned the best steps to make when multiplying matrices. Then, using DeepMind’s favourite reinforcement learning, the system was rewarded for winning the game in as few moves as possible. The AI system discovered a way to multiply two 4×4 matrices using only 47 multiplications, rather than the 64 it takes if you were to painstakingly multiply each row with each column from its corresponding matrix. That’s also two steps less than the 49 found by Volker Strassen in 1969, whose multiplication method for 4×4 matrices had held the record for the fastest one for more than 50 years. What’s ahead? The find could boost some computation speeds by up to 20% on hardware such as an Nvidia V100 graphics processing unit (GPU) and a Google tensor processing unit (TPU) v2, but there is no guarantee that those gains would also be seen on a smartphone or laptop. DeepMind now plans to use AlphaTensor to look for other types of algorithms. “While we may be able to push the boundaries a little further with this computational approach,” Grey Ballard, a computer scientist at Wake Forest University in Winston-Salem, North Carolina, said, “I’m excited for theoretical researchers to start analysing the new algorithms they’ve found to find clues for where to search for the next breakthrough.”","excerpt":"The DeepMind researchers did what they do best—making AIs champions at games.","categories":["AI Trends"],"tags":[],"author_name":"Tasmia Ansari","publish_date":"2022-10-09T13:00:00","publication_year":"2022","word_count":663,"keywords":["Go","machine learning","TPU","AWS","AI","cloud_platforms:AWS","programming_languages:R","Git","V100","R"],"extracted_tech_keywords":["AI","machine learning","AWS","TPU","R","Go","Git","V100","cloud_platforms:AWS","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/why-solving-for-efficiency-of-matrix-multiplication-such-a-big-deal-in-computing\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":62140,"title":"Intel &#038; Udacity To Launch New Edge AI Program To Train The Developer Community","content":"Intel, in collaboration with Udacity, has announced the launch of a new edge artificial intelligence program in order to train one million developers. The new program — Intel Edge AI for IoT Developers Nanodegree Program — has been designed to train the developer community in deep learning and computer vision, with the aim of accelerating the development and deployment of artificial intelligence-based models at the edge by leveraging the Intel Distribution of OpenVINO toolkit. According to the company blog, the students who complete the aforementioned nano degree program, which is estimated to take about three months, will receive a Udacity graduation certificate. When asked about the initiative, Jonathan Ballon, the vice president of Intel and the general manager, Internet of Things Group said, “Historically, students have learned how to build and deploy deep learning models for the cloud. With Udacity, we are training AI developers to go where the data is generated in the physical world: the edge. Optimising direct deployment of models on edge devices requires knowledge of unique constraints like power, network bandwidth and latency, varying compute architectures and more. The skills this course delivers will allow developers – and companies that hire them – to implement learnings on real-world applications across a variety of fields.” Industries like manufacturing, retail, healthcare and others, across the globe, are developing computer vision and artificial intelligence edge solutions in order to obtain accurate and real-time insights, making sure the workforce has relevant skills to address the demands of these industries has become imperative. Also, It has been estimated that the global edge computing market is going to reach $1.12 trillion by 2023, expanding at a CAGR of 32.6%; however, the workforce is not equipped to address industry demands. According to Gabe Dalporto, the CEO of Udacity, this program is part of Udacity’s commitment to providing training for one million developers worldwide. He said, “Our collaboration with Intel will open the doors for students to learn deployment of cutting-edge AI technologies at the edge and aid those with limited access to educational resources to grow in their fields.” Intel collaborated with Udacity in order to address the growing skills gap with the Intel Edge AI for IoT Developers Nanodegree Program, where students will have the opportunity to complete three real-world projects, each reviewed and approved by Udacity’s reviewer network. This will, in return, provide the students with a practitioner-level skill set in delivering AI at the edge. Alongside, for those students who will be unable to commit to the full nano degree program, the Intel Edge AI Fundamentals with OpenVINO course will provide a free subset of the content from the program. This course does not include projects or technical mentor support, but it offers in-depth knowledge on how to develop AI solutions for the edge. Apart from this initiative, last year, November, Intel and Udacity had also launched the Intel Edge AI Fundamentals Course, where the top 850 students who completed the two and half month fundamentals course were awarded nano degree scholarships. Twenty-five of the scholarship winners were from Women Who Code’s Portland, Oregon, chapter. This nano degree program will introduce students to the Intel OpenVINO toolkit, which allows developers to deploy pre-trained deep learning models through a high-level C++ or Python inference engine API integrated with application logic. Based on convolutional neural networks, the OpenVINO toolkit provides graduates with the opportunity to maximise application performance across a range of heterogeneous Intel architectures to deliver fast, efficient deep learning workloads. Anyone working for a company can also take advantage of Intel’s IoT DevCloud to develop, test and run their workloads on a cluster of the latest Intel hardware and software. To avail the benefits of the nano degree program, students have to sign up for the course on the Udacity website.","excerpt":"Intel, in collaboration with Udacity, has announced the launch of a new edge artificial intelligence program in order to train one million developers.  The new program — Intel Edge AI for IoT Developers Nanodegree Program — has been designed to train the developer community in deep learning and computer vision, with the aim of accelerating […]","categories":["AI News"],"tags":["DARPA","Intel","Udacity"],"author_name":"Sejuti Das","publish_date":"2020-04-20T13:30:00","publication_year":"2020","word_count":631,"keywords":["artificial intelligence","AI","DARPA","neural network","Udacity","computer vision","RAG","Python","Aim","deep learning","edge AI","edge computing","Intel"],"extracted_tech_keywords":["AI","artificial intelligence","deep learning","neural network","computer vision","Aim","edge AI","RAG","edge computing","Python"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/intel-udacity-to-launch-new-edge-ai-program-to-train-the-developer-community\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10086212,"title":"NVIDIA GPU Prices Bank on Boom or Crash Outside","content":"GPU prices are like a wildly swinging metronome. The timing to buy a graphics card for your PC is precarious and can be affected by an eruption in tech development elsewhere. First, it happened in September, when Web3 platform Ethereum announced the Merge – moving from the proof-of-work protocol to proof-of-stake. Post Merge, the prices of many popular NVIDIA GPUs, such as NVIDIA’s RTX3080 plummeted by nearly 60% over the last 90 days across some parts of the globe. It is happening again, but this time reversing the share price graph. Reports claim that the excitement around OpenAI’s now-famous chatbot ChatGPT has rubbed off its magic on NVIDIA’s GPUs. The GPU-maker’s stock has already shot up by 29% since the beginning of this year with CEO Jensen Huang’s pockets getting heavier by 33% (USD 18.4 billion) this year. Huang, a Taiwanese immigrant, is a rare winner these days, especially when compared to the fortunes of most tech leaders who are taking a harsh beating due to prevailing macroeconomic conditions. And the logic behind it is plain as day – NVIDIA’s graphics cards have become synonymous with AI tools and ChatGPT, the hottest AI tool right now, is attracting more users by the minute. NVIDIA’s stock rises in the first month of 2023, Source: Seeking Alpha OpenAI’s growing need for compute While OpenAI hasn’t released official figures, sources say that the chatbot has surpassed 10 million daily users in just 40 days, overtaking Instagram’s initial euphoria. The daily traffic is sizeable enough to cause the platform to crash, often preventing people from creating new accounts. Basically, the more popular ChatGPT gets, the better it is for NVIDIA’s GPUs. In a note to investors on January 16, UBS analyst Timothy Arcuri stated that ChatGPT initially used 10,000 NVIDIA GPUs to train the model. “But the system is now experiencing outages following an explosion in usage and numerous users concurrently inferencing the model, suggesting that this is clearly not enough capacity,” he noted. Arcuri explained that the compute on training these generative models is as intensive as training an LLM, suggesting that OpenAI will ‘need to scale quite rapidly’. Moreover, with OpenAI planning to release the long-awaited GPT-4, which is expected to be bigger than GPT-3, NVIDIA should brace for more of that sweet GPU money. Last week, Citigroup released a report estimating that the rapid expansion in ChatGPT users could potentially result in GPU sales of between USD 3 billion and USD 11 billion within the next year. Analyst Atif Malik based his predictions on the number of words generated by ChatGPT and revenue per word for NVIDIA. Also, ChatGPT is only the tip of the iceberg in generative AI. There are a host of other generative tools in the trend that will also depend on GPUs for inference functions. But there’s no denying that the ChatGPT craze came as a divine intervention for NVIDIA at a time when its market value fell by a sharp 50% last year after peaking at more than USD 800 billion in late 2021 due to lower sales. The crypto mining boom in 2018 pushing GPU prices Other tech developments affecting GPU prices The resulting impact on GPU prices though, isn’t unusual for NVIDIA. In September, Ethereum’s long-awaited transition to the proof-of-stake (PoS) mechanism known as the Merge replaced cryptocurrency miners with network validators effectively killing the need for GPUs. Since the upgrade, the value of several models of GPUs fell by a lot. Reports showed that the price of the popular NVIDIA RTX 3080 fell from USD 1,118 to around USD 700 within three months in China. While NVIDIA lost the most, other GPU manufacturers like MSI also saw a drop in their prices. Ironically, just prior to the Merge, the boom in blockchain was responsible for pushing the GPU market further. While mining already was using GPUs, outside of mining too blockchain and GPUs were ideal together. Mining has distributed networks which use computing power leveraging the total users that create ‘virtual supercomputers’ that rely on the combined power of the network. Besides, there was also an increase in the demand for computing-as-service as indicated by the expansion of the GPU-as-a-service market and cloud computing. While hyperscalers like Google Cloud and AWS did start offering GPU services, they were fully centralised and even inefficient compared to NVIDIA. Similarly, GPU sales also lost their steam just as the market for Bitcoin and Ethereum lost momentum. GPUs were also consequently affected by other events within the blockchain ecosystem like investment in dedicated mining rigs called ASICs, projected to be faster than GPUs. On the whole, the demand for GPUs is driven by much more than just crypto mining with its applications ranging between AI to gaming processing. GPUs are vital for development in AI\/ML and NVIDIA GPUs are here to stay because of their familiarity among users and variety in usage. The balance in GPU demand may be achieved, again thanks to the latest rage in AI.","excerpt":"GPUs are vital for development in AI\/ML and NVIDIA GPUs are here to stay","categories":["Global Tech"],"tags":[],"author_name":"Poulomi Chatterjee","publish_date":"2023-01-31T12:11:22","publication_year":"2023","word_count":831,"keywords":["ChatGPT","OpenAI","AI","cloud computing","AWS","ML","RAG","Aim","generative AI","R"],"extracted_tech_keywords":["AI","ML","generative AI","ChatGPT","OpenAI","Aim","RAG","cloud computing","AWS","R"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/nvidia-gpu-prices-bank-on-boom-or-crash-outside\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10161828,"title":"DeepSeek Crushes OpenAI o1 with an MIT-Licensed Model—Developers Are Losing It","content":"DeepSeek, a Chinese AI research lab backed by High-Flyer Capital Management, has unveiled its latest reasoning models, DeepSeek-R1 and DeepSeek-R1-Zero. The models are positioned as alternatives to proprietary systems like OpenAI-o1. DeepSeek-R1, the flagship model, is fully open-source and distributed under the MIT license, allowing developers to use, modify, and commercialise it freely. Developers can access DeepSeek-R1 and its API at chat.deepseek.com. The API offers functionalities for fine-tuning and distillation. “We are living in a timeline where a non-US company is keeping the original mission of OpenAI alive – truly open, frontier research that empowers all,” said Jim Fan, Senior Research Manager and Lead of Embodied AI (GEAR Lab) at NVIDIA. Alongside the technical report, the lab also released six distilled models, ranging from 1.5 billion to 70 billion parameters. These models are optimised for efficiency, and claim performance levels similar to OpenAI-o1-mini. The models are designed to address tasks in math, code generation, and reasoning with competitive accuracy. Leveraging large-scale reinforcement learning in post-training, DeepSeek-R1 achieves high performance with minimal reliance on labelled data. “Our goal is to explore the potential of LLMs to develop reasoning capabilities without any supervised data, focusing on their self-evolution through a pure RL process,” said the team behind DeepSeek. DeepSeek-R1-Zero is built on a pure reinforcement learning (RL) framework, which allows it to develop reasoning capabilities autonomously. Initial evaluations show that it achieved a pass rate of 71% on the AIME 2024 benchmark, an increase from 15.6%. However, the model faced challenges such as poor readability and language mixing. To address these issues, DeepSeek introduced DeepSeek-R1, which incorporated a multi-stage training approach and cold-start data. This method improved model’s performance by refining its reasoning abilities while maintaining clarity in output. “The model has shown performance comparable to OpenAI’s o1-1217 on various reasoning tasks,” the company said. DeepSeek-R1 achieved a score of 79.8% Pass@1 on AIME 2024, slightly surpassing OpenAI-o1-1217. “I love DeepSeek so much! o1 level model is now open-source (MIT license),” said Paras Chopra, founder of Wingify. “Deepseek R1 is on par with o1 and is open-source!! It blows my mind that Chinese make great, open and transparent tech,” said Bindu Reddy, founder of Abacus AI. The launch of DeepSeek comes after it recently launched DeepSeek-V3, which was touted as the best open-source model. “Whale 🐋 folks, respect,” said KissanAI founder Pratik Desai. OpenAI is currently facing controversy over its o3 model due to its undisclosed funding of EpochAI’s FrontierMath benchmark and prior access to a significant portion of the test data. Despite these concerns, the company plans to release its new o3 mini model within the next couple of weeks.","excerpt":"The company also released six distilled models, ranging from 32 billion to 70 billion parameters.","categories":["AI News"],"tags":["DeepSeek","OpenAI"],"author_name":"Siddharth Jindal","publish_date":"2025-01-20T21:57:03","publication_year":"2025","word_count":439,"keywords":["Go","API","TPU","OpenAI","AI","RPA","DeepSeek R1","RAG","DeepSeek","Aim","R"],"extracted_tech_keywords":["AI","OpenAI","DeepSeek R1","Aim","RAG","TPU","R","Go","API","RPA"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/deepseek-crushes-openai-o1-with-an-mit-licensed-model-developers-are-losing-it\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10053668,"title":"Only 11% Of Techies Expect To Return To Office, According To Github Report","content":"GitHub recently released its State of the Octoverse report, its annual report looking back at the code and communities building on GitHub in the past year. The report shows that during 2021, the developer community evolved from learning to balance personal and work lives in the Covid-19 pandemic to creating valuable patterns and processes that meet this new era of remote work. It dives in on trends across the more than 73M developers worldwide building on GitHub—including 16M+ new users in 2021 and 84% of the Fortune 100. In addition to celebrating community growth, for the first time, the report includes survey data from 40K+ developers and research into 4M code repositories to provide more insight on everything – from how developers are shipping code faster to the rise in productivity during the pandemic as more developers worked remotely. GitHub saw a 38.9% year-on-year growth in total users from India and a 24.8% growth in total active students from the country. It states that almost 70% of active users in the code repository are now from outside North America. The fastest-growing countries and regions outside of the US include Indonesia, Brazil, India, Russia, Japan, Germany, Canada, UK and China. Image Source: GitHub Some of the key highlights of the report were: Performance can increase up to 87% when reusing code: Projects built with code and toolchains from the open-source community are thriving. Projects see 2x performance compared to those with more friction, like slow processes or multiple approval layers.Company teams perform 43% better using automation: Automating software delivery is a key enabler in open source and helps teams go faster at scale. Using actions increases the number of merged pull requests by 36% and shrinks time to merge by 33%. Productivity is returning to pre-pandemic levels, but the workplace is shifting; only 11% of developers expect to return to collocation: That’s a 30% drop from the 41% who worked in an office before.Developers find more fulfilment with automation: By removing friction and repetitive tasks through automation, teams perform 27% better in open source and 43% better at work, and developers report higher fulfilment.Mentorship is an asset in both open source and companies: When there is a commitment to coaching and mentoring — e.g. through friendly and timely code reviews for new contributors or new hires — teams see a 46% improvement in productivity within open-source projects and a 16% improvement within companies.Developers are now writing and shipping code faster, pull requests with just one reviewer are often merged within an 8-hour workday, and with each additional reviewer, merging it in a day goes down by 17%. Assigning no more than three reviewers in an open-source repo can be a tradeoff between quality and speed, says the report. Image Source: GitHub “Automation helps reduce repetitive tasks, redundancies and errors, allowing developer teams to focus on higher-value challenges and drive innovation. Ultimately, automation aims at empowering developers, enhancing developer experience and productivity, and helping create the right environment for them to innovate, solve challenging problems, and make a difference,” said Maneesh Sharma, General Manager, GitHub India. JavaScript emerged as the top language, while Python continues to be the second most widely used language, with Java following as the third. TypeScript, which moved up to the fourth position last year, still maintains its position. Of all the ten languages surveyed, only Shell and C changed places from last year — Shell climbed one notch up to 8th position, displacing C.","excerpt":"The report shows that during 2021, the developer community evolved from learning to balance personal and work lives in the Covid-19 pandemic to creating valuable patterns and processes that meet this new era of remote work.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","Data Science","Data Scientist","Deep Learning","Developers","developers India","GitHub","github data science","GitHub open source Python projects","GitHub Repositories","Javascript","Machine Learning","Open Source"],"author_name":"Victor Dey","publish_date":"2021-11-17T14:59:38","publication_year":"2021","word_count":577,"keywords":["Developers","Javascript","Git","Data Scientist","R","Java","github data science","Open Source","Data Science","Go","AI","GitHub open source Python projects","Machine Learning","GitHub Repositories","developers India","TypeScript","Python","Aim","JavaScript","Deep Learning","GitHub","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","Aim","Python","R","JavaScript","TypeScript","Go","Java","Git","GitHub"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/only-11-of-techies-expect-to-return-to-office-according-to-github-report\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10171024,"title":"Hugging Face Introduces Two Open-Source Humanoid Robots","content":"Hugging Face, a leader in open-source AI, announced the release of two open-source humanoid robots, HopeJR and Reachy Mini. The unveiling occurred as part of the company’s ongoing expansion into robotics, following its acquisition of Pollen Robotics in April. HopeJR is a full-sized humanoid robot featuring 66 actuated degrees of freedom, enabling it to walk and manipulate objects. Reachy Mini is a desktop unit designed for AI application testing and is capable of head movement, speech, and auditory interaction. Clem Delangue, co-founder and CEO of Hugging Face, stated that the company anticipates shipping initial units by the end of the year, with a waitlist currently open. HopeJR is expected to be priced around $3,000, while Reachy Mini will be around $300, subject to tariffs. Meet HopeJr, a full humanoid robot lowering the barrier to entry!Capable of walking, manipulating many objects, open-source and costs under $3000 🤯Designed by @therobotstudio and @huggingface 👇 pic.twitter.com\/wCwo8YPOGV— Remi Cadene (@RemiCadene) May 29, 2025 Delangue emphasised the significance of the robots’ open-source nature, noting to TechCrunch, “The important aspect is that these robots are open source, so anyone can assemble, rebuild, [and] understand how they work, and [that they’re] affordable, so that robotics doesn’t get dominated by just a few big players with dangerous black-box systems.” Hugging Face’s acquisition of Pollen Robotics facilitated the development of these robots, providing the company with the new capabilities necessary for their creation. This move aligns with Hugging Face’s broader strategy to democratise AI and robotics, as evidenced by their 2024 launch of LeRobot, a collection of open AI models, datasets, and tools for building robotic systems. End of last month, Hugging Face also released an updated version of its 3D-printed, programmable robotic arm, the SO-101, developed in partnership with French robotics firm The Robot Studio. Additionally, the company expanded the training data on its LeRobot platform through a collaboration with AI startup Yaak, incorporating data for self-driving machines. With the introduction of HopeJR and Reachy Mini, Hugging Face aims to lower the barriers to entry in robotics, promoting accessibility and innovation in the field.","excerpt":"HopeJR is a full-sized humanoid robot featuring 66 actuated degrees of freedom, while Reachy Mini is a desktop unit for AI application testing.","categories":["AI News"],"tags":["Hugging Face","Humanoid Robots","Robotics"],"author_name":"Sanjana Gupta","publish_date":"2025-05-30T18:08:46","publication_year":"2025","word_count":345,"keywords":["Go","Hugging Face","Humanoid Robots","programming_languages:R","AI","innovation","programming_languages:Go","Robotics","Aim","ai_frameworks:Hugging Face","R","startup"],"extracted_tech_keywords":["AI","Aim","Hugging Face","R","Go","innovation","startup","ai_frameworks:Hugging Face","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/hugging-face-introduces-two-open-source-humanoid-robots\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":561,"title":"Aamir Khan set to play astronaut Rakesh Sharma, the first Indian to fly into space","content":"Maverick actor Aamir Khan is all set to fly into space with his proposed biopic on Indian astronaut Rakesh Sharma. According to news sources, Khan known for his marketing skills and fastidious ways (he releases one film every year) is planning on signing the dotted line. According to reports, UTV honcho Siddharth Roy Kapur approached the methodical Khan for the astronaut’s role. The film slated to go on the floors in 2018 will be produced under Kapur’s newly minted production house Roy Kapur Films and will be directed by Mahesh Mathai. Just like Dangal, Aamir Khan Rakesh Sharma biopic has given the 51-year-old actor’s fans something to look forward to with bated breath. The film, tentatively titled Saare Jahan Se Achcha is inspired by the words spoken by Sharma during his famed exchange with Indira Gandhi in 1984. When asked by the then PM how India looked from space, Sharma recounted poet poet Mohammed Iqbal famous lines, Saare Jahan Se Achcha. The title is also a nod to the independence song written by legendary poet Mohammed Iqbal during India’s Independence struggle. Sharma, an Indian Air Force pilot was selected to fly into space in a joint programme between Indian Space Research Organisation and the Soviet Interkosmos space program in September 1982. He was selected to travel as a cosmonaut. Sharma, a test pilot had already proven his mettle in war operations. In 1984, he went into space aboard a Soyuz T-11 space shuttle and spent 21 hours and 40 minutes in outer space. On his return, Sharma was awarded the Hero of Soviet Union and felicitated with an Ashok Chakra by the Indian government as well. During his time in outer space, Sharma and the crew members had a TV news conference with PM Indira Gandhi and Moscow officials. The former Interkosmos program was aimed at assisting Soviet allies in manned and unmanned space missions. The eastern bloc’s space program helped non-aligned nation such as Indian and Syria as well. We can’t wait to watch how the Aamir Khan Rakesh Sharma biopic will pan out.","excerpt":"Maverick actor Aamir Khan is all set to fly into space with his proposed biopic on Indian astronaut Rakesh Sharma. According to news sources, Khan known for his marketing skills and fastidious ways (he releases one film every year) is planning on signing the dotted line. According to reports, UTV honcho Siddharth Roy Kapur approached […]","categories":["AI News"],"tags":[],"author_name":"Richa Bhatia","publish_date":"2017-02-15T07:23:51","publication_year":"2017","word_count":345,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","Aim","GAN","R"],"extracted_tech_keywords":["AI","Aim","R","Go","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/aamir-khan-set-play-cosmonaut-rakesh-sharma-first-indian-fly-space\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10022213,"title":"Why Is Kubernetes So Popular?","content":"Kubernetes is an open-source system to automate the deployment, scaling, and management of containerized applications. This container management system’s popularity soared with the rise of businesses operating on cloud infrastructure and technology. Let’s take a deep dive and see why Kubernetes is all the rage now. Rise Of Kubernetes A container is a software package comprising application code, runtime, system libraries, and other settings required to run an application. Containers have been a significant part of Linux since the 1980s. However, containerization has become a thing after Docker burst into the scene with its range of formats and tools for containers. A RightScale report titled, State of the Cloud, said container adoption increased from 49 percent in 2018 to 57 percent in 2019. Many cloud providers, including Microsoft Azure, Google Cloud, and Amazon AWS, have also launched containers. Containers run complex and critical enterprise applications, and the rise in their numbers have necessitated the need for a managing system. Launched by Google and currently being maintained by the Cloud Native Computing Foundation (CNCF),  Kubernetes can be deployed in very different scenarios depending on the objectives: In-house: Companies can transform their data center to a Kubernetes cluster to take full advantage of the resources available.Cloud: With Kubernetes, organisations can create an infinite number of virtual machines.Hybrid: Kubernetes create virtual machines on a cloud when on-premise servers are full to better distribute computing resources.Multi Cloud: Kubernetes helps companies in avoiding vendor lock-in and minimising risk. Kubernetes has become the de-facto standard for container management system as it offers several advantages: It offers easy container scaling across many servers in a cluster. The autoscaler service can replicate Kubernetes instances or pods to different nodes, thereby maximising resource utilisation.The software can be easily ported between different types of environments.Kubernetes offers high fault tolerance clustering, which contributes to the stability and reliability of the project.It has built-in data encryption, vulnerability scanning, and other capabilities that enhance its safety.With Kubernetes, developers can maintain replica sets. No need to replicate the entire application, making the project more resilient and have maximum responsiveness and uptime. Popular Kubernetes use cases include: Pokemon Go: Developed by Niantic Inc., the popular game saw over 500 million downloads with up to 20 million active users on a daily basis. Pokemon Go’s parent company was not prepared for this kind of traffic. The solution, at last, was found in the form of Google Container Engine powered by Kubernetes. Airbnb: When Airbnb decided to transition from monolithic to a microservices architecture, the team needed to scale continuous delivery horizontally and make it available to 1,000 engineers of the company to add new services. Kubernetes helped Airbnb to support the engineers for concurrently configuring and deploying 250 critical services. New York Times: Currently, most of the New York Times customer-facing applications are running on Kubernetes. The media company has transitioned from a ticket-based system for resource request to pushing updates independently. Open Source Community Kubernetes is one of the largest open source communities, with 75,200 stars on GitHub and contributions from thousands of organizations: One of the reasons why it is rated higher than the competitors such as Docker Swarm and Apache Mesos. Kubernetes crossed 100,000 issues\/PRs a few hours ago! That is 100,000 times the project has been improved by contributors like you. Thank you all for contributing :tada:— K8sContributors (@K8sContributors) March 9, 2021 The patronage of CNCF comes with unique perks. CNCF is part of the larger Linux Foundation, with Microsoft, Google, and AWS as its members. Further, the community around Kubernetes has also grown exponentially. RedHat, CoreOS, and Platform 9 have included Enterprise Kubernetes in their key offerings and invested heavily in this open-source project. In the last few years, Kubernetes has had major releases every three months. New features allow developers to have more flexibility when running a variety of workloads. Kubernetes & Hybrid Cloud With more enterprises adopting a hybrid cloud strategy, a few challenges have arisen, including what technologies to use, how to operate and manage resources, and application deployment. Kubernetes is also ideal for hybrid cloud as It provides consistency in both on-premise and public cloud.Kubernetes allows users to deploy applications based on their business needs.It also offers the ability to automatically scale the applications, leading to better utilisation of the underlying infrastructure.Kubernetes automates the deployment of containerized workloads across the hybrid architectures, allowing organisations to deploy and run their containers on servers at different locations. Developers can add additional clusters to their existing infrastructure if needed. This reduces the application downtime and improves overall performance. As per IBM, the core of the company’s cloud strategy is ‘understanding that customers need a unified and streamlined way to create modular application services, to transform and modernize legacy, and to manage all services, no matter where those workloads reside’. To accomplish that, all the services need to be catalogued, secured, and governed, which is only possible through a unified approach to cloud computing. This was possible only if the strategy was based on open source, making Kubernetes the best choice. “So Kubernetes, in the end, is an orchestration layer. It takes away management complexity, it takes away the cost of doing operations in a large cluster of physical resources. I think the value for the CIO level is the following– Today on average 70 percent of your total cost and people are tied up in maintaining what you have, 30 percent is on new. That’s the rough rule of thumb. Technologies, like Kubernetes, have taken to where we wanted to go, can flip that to 30%-70%, meaning you need to spend only 30 percent maintaining what you have, and you could, then, go spend 70% on doing innovation, which is going to make your end-client happier, and your business happier, said IBM CEO Arvind Krishna in an interview. Wrapping Up Kubernetes’ popularity is on the rise with use cases in mission-critical sectors such as finance, edtech, and traditional enterprise IT. However, Kubernetes face a few challenges. The extremely complex nature of developing and running distributed frameworks at scale is one of its main challenges. Despite this, experts believe Kubernetes will become a ‘universal control plane’ to manage containers, virtual machines, and other modern applications.","excerpt":"Kubernetes is an open-source system to automate the deployment, scaling, and management of containerized applications. This container management system’s popularity soared with the rise of businesses operating on cloud infrastructure and technology. Let’s take a deep dive and see why Kubernetes is all the rage now. Rise Of Kubernetes A container is a software package […]","categories":["AI Features"],"tags":[],"author_name":"Shraddha Goled","publish_date":"2021-03-16T12:00:00","publication_year":"2021","word_count":1034,"keywords":["AWS","cloud computing","AI","ML","docker","RAG","microservices","containerization","Azure","kubernetes"],"extracted_tech_keywords":["AI","ML","RAG","cloud computing","AWS","Azure","kubernetes","docker","containerization","microservices"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/why-is-kubernetes-so-popular\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":6225,"title":"Movie Analytics 2.0: Predicting the next 100 crore club movie","content":"Last October I wrote an article on how analytics can transform the way film industry (Movie Analytics in India: Are we underestimating, rather completely  ignoring the business potential)  . I received an overwhelming response from people, appreciating the idea and some of them saw it as a full-fledged business opportunity but most them had a question :  “Is this possible or just another pretty idea on table?”, so I decided to come up with a probable solution or way forward to do the analysis. So let’s look at some of the methods which can be used to predict the next 100 crore movie. Social media sentiment analysis Like many other people. I also don’t go by movie reviews; instead, I rely on people who tell me if a particular movie is a 100 crore club thing or a waste of money. Now, let’s talk about “Happy New Year”, the upcoming SRK flick releasing in Diwali. By using R ( twitteR package, which is freely available) , we can extract the tweets for a given time period. Once we gather the tweets, it’s time to run Sentiment Analysis on them, and answer several questions highlighted below: Number of tweets mentioning the film’s name or related to film: The number of tweets when the first look of the movie was launched The number of tweets when its first song was launched The number of tweets when SRK’s 8 pack look was launched and finally The number of tweets of Friday and Saturday (to know the final buzz) There will be accounts that tweet repeatedly, and we will have to sanitize them. Now there will be tweets with good sentiments and bad sentiments (Can’t wait for #HNY or SRK looks so old in HNY #HNYsucks) so we will give a sentiment score to the sanitized tweets, on the scale of -10 to 10. After comparison of the good and the bad sentiment score, we will get a S-score for the movie. Text Analysis of the tweets We will analyse the text of sentiment tweets (good and bad) to know what words people are using prominently. We may find results like SRK, Sharukh, Deeepika are the prominent words in good tweets, whereas Abhishek, JuniorB, Farah are common words in bad tweets. This will give us an idea about the general perception about the movie. If we have to predict whether or not a movie will make it to the 100 crore club, YouTube is perhaps one of the most important destination to find out. If a movie is stated to release on Diwali, the first look of the movie will be launched on YouTube 2-3 months prior and the “YouTube API v2.0 – Retrieving Insight Data” can provide us some very relevant data such as the time, day, date, device on which you watched the video, your location , ratings you gave, when did you pause the video and commented, your browsing and scrolling behaviour, when and where did you rewind , fast forward or watch the video again. Every movie, small or big, will have a dedicated Wikipedia page and the importance of those pages in movie prediction can be seen as a research (https:\/\/analyticsindiamag.com\/movie-analytics-in-india-are-we-underestimating-rather-completely-ignoring-the-business-potential\/)  shows the correlation between people who are coming , reading and editing the movie page to the success of the movie is very strong hitting an R-squared value of 0.94. It was 0.925, a month before the movie came out. While Twitter at its peak hit 0.98, that peak arrives sharply on the release date and hovers around 0 a month before release. There are different variables which can be considered while taking Wikipedia into consideration such as Number of edits, Number of views etc, these can be obtained by using the categorization function in Wikipedia Extracting the data and comparing the results Comparing the results with the films that were into 100 crore club in the past, also which were supposed to be the part of the club but couldn’t make it (to know that whether the prediction based on our model is accurate or not, if yes then to what extent) Extracting the data : Beside using R packages like twitteR and YouTube API , Its easier if we can extract the social media data by using services such as Radian or Ubervu (both are paid services) and analyse them on any statistical software i.e. SAS, R Neural network and other statistical techniques To predict the success of the movie before release, neural networks and other models can be useful. So, if we say that 9 is the most successful and 1 is the least, we can put a upcoming movie on this scale to predict the success of the movie. For this we can take the past movie data of last 5 years (boxofficeinida). Now take independent variables such as the rating of the movie (18+), Cast star value, Genre of the movie, it’s a sequel or not, Number of screens where films will be screened, budget, director. We can use a Multilayer perceptron neural network model for this particular case, with two layers. We can also use Box-Jenkins ARIMA model as well as Multilinear regression. So, it would be like one has to identify the patterns in past data and mix them with the current data points that are freely available and then came to a conclusion. First we will quantify the value of a cast and crew of the film. Now this can be accomplished by using graph theory applied to the IMDB database (Boxofficeindia in case of bollywood). The nodes of the graph will be the actors, directors, producers, story writer, studio, music director in the database, then We can connect these nodes using films as the graph’s edges. That will result in millions of connections between all of the nodes Now we can apply Google’s PageRank algorithm to the collection set and assign each film’s opening weekend box office revenue as the value of each connection. Now that will allow us to come up with a score for each cast member and approximate his or her relative contribution to the Movie market economy. Next we will weigh the impact of the cast by role: actor, director, producer, studio, writer etc by using least squares linear regression. Least square linear regression will give plane in space and then it will minimizes the sum of the squares of the distances for all the data points from the said plane ultimately providing us with coefficients that approximate the relative value of each role. On the other hand applying ANOVA can help us to separate the factors that were statistically significant in predicting revenue from those which are not. We can also quantify the impact of genre on a film’s financial success, we took a sample of films released in past from each genre (Action, Romance, Family drama) along with revenue figures and measure the deviation from mean by genre category. However, though it is an effective technique, it is difficult to balance the test and train data also getting the right number of weights and starting nodes Conclusion There can be several methods of predicting the success or failure of a movie at box office. Some of were covered here. Provided adequate time and resources, we can build a model which can almost accurately predict a movie’s box office success and its chances of making it to the 100 crore club.","excerpt":"Last October I wrote an article on how analytics can transform the way film industry (Movie Analytics in India: Are we underestimating, rather completely  ignoring the business potential)  . I received an overwhelming response from people, appreciating the idea and some of them saw it as a full-fledged business opportunity but most them had a question […]","categories":["AI Trends"],"tags":[],"author_name":"Ritesh Mohan Srivastava","publish_date":"2014-10-02T07:36:04","publication_year":"2014","word_count":1237,"keywords":["Go","API","programming_languages:R","sentiment analysis","neural network","AI","programming_languages:Go","analytics","R"],"extracted_tech_keywords":["AI","neural network","analytics","sentiment analysis","R","Go","API","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/movie-analytics-2-0-predicting-the-next-100-crore-club-movie\/","complexity_score":4,"technical_depth":9,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":459,"title":"Can Beer be brewed on the Moon? Engineering Students at University of California try to find out","content":"While the curiosity see no boundaries, proving onto the same a team of UC San Diego engineering students is hoping to find out if beer can be brewed on the moon! One of the four finalists at the Lab2Moon competition being held by TeamIndus, they stand a chance to send a spacecraft to the moon as part of the Google Lunar XPRIZE challenge. They are one of the four teams with a signed launch contract. This experiment is aimed at testing the viability of yeast on the moon, which if turns positive could pave a green signal for a freshly brewed batch of beer in the moon. They would not only be the first to brew beer on the moon, but the first to brew beer in a fermentation vessel the size of a soda can. The results of how yeast behave on the moon would not only be important for brewing beer in space but for developing pharmaceuticals and yeast-containing foods, like bread. One of the just 25 teams to have made it amongst the 3,000 competing for a spot aboard TeamIndus spacecraft, this Jacobs School of Engineering undergraduates call themselves “Team Original Gravity.” The spacecraft is scheduled to be launched on Dec. 28, 2017. The spacecraft is owned by the Indian startup TeamIndus, that won a $1 million Milestone Prize for successfully simulating the landing technology concept of its spacecraft. “The idea started out with a few laughs amongst a group of friends,” said Neeki Ashari, a fifth year bioengineering student at UC San Diego and the team’s PR & Operations Lead. “We all appreciate the craft of beer, and some of us own our own home-brewing kits. When we heard that there was an opportunity to design an experiment that would go up on India’s moonlander, we thought we could combine our hobby with the competition by focusing on the viability of yeast in outer space.” The unique strategy behind the experiment: First step involves all the prep work before yeast is added to be done on Earth, rather than in the experimentation vessel. The experiment does not brew the “wort,” or unfermented beer. Secondly, rather than separating the “fermentation” and “carbonation” phases as would normally occur during the process of making beer, the team plans to combine them. This eliminates the need for releasing accumulated CO2, which can result in sanitation and safety issues. It also prevents the possibility of over-pressurization if anything in the system fails and makes the system easier to design. Lastly, the testing of fermentation and yeast viability will be done via pressure, rather than using density measurements as done on Earth. This is because density measurements use gravity. “Converting the pressure build up to fermentation progress is straightforward, as long as volume and original gravity—specific gravity before fermentation, hence our name—are known prior to the experiment,” said Han Ling, a fifth year bioengineering undergraduate and the team’s brewing lead. With a range of experiments from photosynthesis to electricity, the experiments would be evaluated by an international jury in March when teams fly to Bangalore, India, to showcase their final prototype. Siddhesh Naik, TeamIndus Ninja and mentor to Original Gravity concludes, “The yeast study is among the coolest experiments to be performed on the lunar surface, and I am sure they are one of the top contenders to win the Lab2Moon competition. Original Gravity is one of the most hardworking teams and very dedicated to their project.”","excerpt":"While the curiosity see no boundaries, proving onto the same a team of UC San Diego engineering students is hoping to find out if beer can be brewed on the moon! One of the four finalists at the Lab2Moon competition being held by TeamIndus, they stand a chance to send a spacecraft to the moon […]","categories":["AI News"],"tags":[],"author_name":"AIM Media House","publish_date":"2017-02-01T08:30:44","publication_year":"2017","word_count":575,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","Aim","ViT","R","startup"],"extracted_tech_keywords":["AI","Aim","R","Go","ViT","startup","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/can-beer-brewed-moon-engineering-students-university-california-try-find\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":26790,"title":"Food Tech Startup Dishq Raises $400,000 In Pre-Seed Funding","content":"dishq, a Bengaluru-based artificial intelligence startup, that leverages food science and machine learning to predict people’s taste, announced a pre-seed investment of $400,000. This round has been funded by several investors including, Techstars’ first food\/AgTech-focused accelerator called Farm To Fork, and Arts Alliance. The primary deployment of the investment will be to expand the engineering team, as well as sales and marketing activities. The company was launched in December 2015 with the mission of helping consumers make better food choices by bringing personalisation technology to the food and beverages industry. Both Founders, Kishan Vasani and Sai Sreenivas Kodur, previously worked at online food ordering industry giants, Just Eat and Zomato respectively. It was this experience and the personal frustration of dissatisfying meals that drove them to tackle this space, said the company official statement. “Why shouldn’t everything we eat or drink taste delicious? We all know how great we feel when we’ve enjoyed a truly amazing meal, we’re happier, more productive, and we waste less. Our technology essentially brings greater alignment between producers and consumers, and we’re truly honoured and excited to have the backing of such fantastic investors for our vision,” said Vasani, co-founder and CEO at dishq. dishq has developed, what they call, a ‘food brain’ which can predict both an individual’s food preferences, as well as broader industry trends. dishq already has customers across six markets for their first product, a B2B personalisation engine, which is currently powering more than 30 million recommendations each month. “There is no question in my mind that Kishan, Sai and team are building something great here. There is a ton of opportunity in this space and this is the team that will capitalise on it,” said Brett Brohl, managing director at Techstars Farm To Fork. While the majority of the team is based at the company headquarters in Bengaluru, dishq will also be expanding its presence in London. They have also recently opened an office in Saint Paul, Minnesota, US.","excerpt":"dishq, a Bengaluru-based artificial intelligence startup, that leverages food science and machine learning to predict people’s taste, announced a pre-seed investment of $400,000. This round has been funded by several investors including, Techstars’ first food\/AgTech-focused accelerator called Farm To Fork, and Arts Alliance. The primary deployment of the investment will be to expand the engineering […]","categories":["AI News"],"tags":["food tech","foodtech","Zomato"],"author_name":"Prajakta Hebbar","publish_date":"2018-07-31T07:11:44","publication_year":"2018","word_count":329,"keywords":["API","foodtech","artificial intelligence","machine learning","programming_languages:R","AI","food tech","RAG","ViT","Rust","Zomato","R","startup"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","RAG","R","Rust","API","ViT","startup","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/food-tech-startup-dishq-raises-400000-in-pre-seed-funding\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":30110,"title":"How This AI Is Predicting The Success Or Failure Of A Movie By Looking At Its Trailer","content":"Trailers have transitioned from being played after the movie to being tailored as a teaser releasing one year before the main film hits the theatres. The core of this transition relied heavily on customer behaviour just like any other business venture. Film-makers condense the full-length feature into a 2-minute trailer which displays probable awe-inducing scenes while trying to stay true to the essence of the original content. This article primarily focuses on the recent developments in the movie-recommendation engines and how studios are using them to rake in profits. Understanding the market segmentation of the movie-going public lies at the core of any movie studios’ marketing strategy. They go to great lengths to understand consumer behaviour because predicting the outcome of the movie before its release is close to impossible. For example, the movie, Shawshank Redemption (1994) which tops the lists of movie-buffs and critics alike, tanked at the box-office. One of the key reasons was the poor word of mouth. It is indigestible for the creators to make a masterpiece and then fail due to something as trivial as mispronouncing the title of the movie. The researchers at 20th Century Fox trained their convolutional neural network using NVIDIA Tesla P100 GPUs on the Google Cloud, with the cuDNN-accelerated TensorFlow deep learning framework, on hundreds of movie trailers released over the last years, as well as millions of attendance records. At its core, the model works on the parameters available in each frame of a movie trailer to deduce the probability of a movie being liked and which segment of the audience would like it. Temporal Dynamics In this paper, the researchers developed a model which works heavily on the temporal dynamics of movie trailers. Here the convolution layer is followed by temporal pooling layer which makes a gist of all the video feed frame by frame and is later fed into a hybrid collaborative filter. This model utilises ‘convolution over time,’ a multivariate time series method which makes the model temporal-aware. “Video analysis using pooling schemes to collapse an entire video or part of a video into a unique dense feature vector can miss important semantic aspects of the video. Although simple to implement, the approach neglects the sequential and temporal aspects of the story, which, we argue, can be useful for characterisation of a motion picture,” said the data scientists at 20th Century Fox. The idea behind VCN is to learn a collection of filters, each of which captures a particular kind of object-sequence that could be suggestive of specific actions. The model differentiates between long shots and intermittent shots using a technique called Temporal Sequencing. This can be used to deduct key information about the movie — like its genre or who the protagonists are or whether it is film suitable for the entire family. Temporal sequencing is done using Video Convolution Network (VCN). This network trains on different shapes like convolution networks usually do, and increases its ability to label objects in the sequenced parts of the trailer. Using A Video Convolution Network A dialogue-heavy drama can have a sequence where there are close-up shots of the actors. It can be assumed with certainty that the face of the speakers is shown back and forth. Sequences like these are plenty and filtering out the most significant object-sequences is the job of object-specific convolution filters. Step-By-Step Working Of The Algorithm: Downsampling the videos to one frame-per-second. Extracting a 1024 dimensional image features for each frame using the Inception V3 model. A convolution layer applies 1024 convolution filters against these features where each filter of the size 8 x 1024. The layer ends up having 8 million (1024 x 8 x 1024) dimensions. These filters are convoluted along the temporal dimension with a stride value of 2 for dimensionality reduction. There is also a residual layer that performs another set of convolutional filters against the output of the previous layer. If this layer’s filter size > 1 frame, it will further increase the effective receptive field. Such filters do not expand the receptive field but increase the capacity to summarise information among all 1024 channels. Source: Paper On Convolutional Collaborative Filter Network For Video-based Recommendation Systems These video sequences activate a particular channel in the last RELU layer. Every corresponding activation is typically higher than the average standard deviation of previously activated layers. In the picture below, we see two frames (John Wick and Jason Bourne). Any average movie-goer would guess these frames belong to a movie, high on action. But, training the network to figure out this correlation is the tricky part. VCN does exactly that by using object-specific sequencing to find similarities in the dense feature vectors as discussed above. The comparison charts reveal how the model did fairly well on many occasions. Although there is still room for improvement, this model gives a much-needed platform to the movie studios for a better evaluation of their content and targeted audience. “Video trailers are the single most critical element of the marketing campaigns for new films,” the 20th Century Fox researchers stated in their paper. There is now a push for appending bumpers at the start of trailers. These bumpers have a runtime of 5 seconds revealing whatever explosive content the trailer has to offer. With high volatile attention spans, the marketing teams believe in the need for diligent strategies which are high on the tech front. According to a leading financial daily, the Indian movie market is growing at a rate of 11.5% every year. The experts forecast a phenomenal rise in revenue in the coming years. The estimation now stands at a whopping 24,000 crores by the year 2020. With such alluring incentives and fierce competition within the industry, it wouldn’t be surprising if the Indian movie-giants dial-up analytics to grab a better half of the pie.","excerpt":"Trailers have transitioned from being played after the movie to being tailored as a teaser releasing one year before the main film hits the theatres. The core of this transition relied heavily on customer behaviour just like any other business venture. Film-makers condense the full-length feature into a 2-minute trailer which displays probable awe-inducing scenes […]","categories":["Deep Tech"],"tags":["cnn","Collaborative Filtering","recommendation engines","Tensorflow"],"author_name":"Ram Sagar","publish_date":"2018-11-12T12:00:08","publication_year":"2018","word_count":972,"keywords":["TPU","AWS","AI","recommendation engines","neural network","recommendation systems","cnn","RAG","Collaborative Filtering","deep learning","analytics","TensorFlow","R","Tensorflow"],"extracted_tech_keywords":["AI","deep learning","neural network","analytics","TensorFlow","RAG","recommendation systems","AWS","TPU","R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/how-this-ai-is-predicting-the-success-or-failure-of-a-movie-by-looking-at-its-trailer\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10070502,"title":"How to use Genetic Algorithm for Image reconstruction?","content":"The genetic algorithm is inspired by the biological evolution of chromosomes and this is majorly used in optimal feature selection for various kinds of problems. The genetic algorithm basically follows the heuristic algorithms approach to find the best possible solution close to the optimal solution. This algorithm has a variety of applications across the fields, most in optimization problems. It can also be used for image reconstruction to obtain the images in their original form. This article is focused on applying the genetic algorithm for this interesting application of image reconstruction. Table of Contents An introduction to genetic algorithmsWhat is image reconstruction?Image reconstruction using genetic algorithmsSummary An introduction to genetic algorithms Genetic algorithms are basically used in optimization tasks where these algorithms follow the metaheuristic approach to obtain the best optimal solution among all the candidate solutions. Genetic algorithms operate on the hierarchy of biological chromosomes and the algorithm tries to obtain the best fittest solution to be passed on to the next generation. Are you looking for a complete repository of Python libraries used in data science, check out here. Genetic algorithms initially operate on a cluster of the best possible solutions called the initial population where the cluster contains the best possible solution to the problem. Each of the solutions in the cluster is categorized by certain parameters and they are termed Genes in genetic algorithms the genes in the genetic algorithm will be ranked according to the parameters set. The genes which have all the abilities set will be passed on to a fitness function to determine the fitness of the best possible solutions selected from the cluster of candidate solutions to the problem. If the best possible solution has the ability to inherit all the functionality of the fitness function that solution will be passed on to the next set of generations. The overall operation of genetic algorithms happens at three levels they are Selection, Mutation, and Crossover. Let us try to understand these standard terminologies of genetic algorithms. Selection Selection in genetic algorithms is a standard terminology used for selecting the best possible solution among the candidate solutions. Selection plays an important role in genetic algorithms as it would be responsible for selecting the best possible solutions, which would be passed on to the next generations for further process. So the selection phase in genetic algorithms is a very important phase in genetic algorithms. Crossover The crossover phase of the genetic algorithm is basically responsible for setting a certain threshold at random places in Genes and the crossover will be responsible for exchanging genes between the two random solutions obtained in the Selection phase. So the information has to be passed over the individual solutions until the crossover for the genes is reached. Mutation The mutation phase in genetic algorithms ensures crossover occurs with higher random probability within the genes. So the mutation will be responsible for setting higher random threshold values in the genes so that information passed over between the solutions is considerable and the Mutation phase is also required to maintain diversity between the best possible solutions selected for the problem. What is image reconstruction? Image reconstruction is an iterative process of evolving through the various pixels of images and an attempt to obtain the original image through evolution. Image reconstruction finds its major applications in medical imaging, reverse imaging, and original image restoration tasks. In a similar way, genetic algorithms can be used for image reconstruction wherein the algorithm tries to reconstruct the image close to the original image. Genetic algorithms basically follow a metaheuristic-based approach and the algorithm iterates over the best possible ways to reconstruct the image close to the original image. Pygad is one such library of genetic algorithms which is structured and designed to reconstruct images close to the original image. Image reconstruction using Genetic Algorithm Image reconstruction in genetic algorithms can be taken up using the Pygad library which is specially designed for image reconstruction using the genetic algorithm principles. So let us see how to carry out image reconstruction using Pygad. Let us first install the library and import the library into the working environment. !pip install pygad import pygad As the library is imported into the working environment let us visualize the original image using the matplotlib library. import matplotlib.pyplot as plt img=plt.imread('\/content\/drive\/MyDrive\/Colab notebooks\/GA_Img_reconstruction\/tiger.jpeg') plt.imshow(img) plt.show() The image has to be scaled down to standard scale values for passing the original image to the genetic algorithm. So let us scale the image using the numpy library. img = np.asarray(img\/255, dtype=np.float) Once the image is scaled we will have to use a library named GARI [Genetic Algorithm for Reproducing Image] which is a library designed for decomposing the original image into chromosomes and also reshaping the original image accordingly it can be passed onto the genetic algorithms. So GARI can be said to be one of the submodules of Pygad and it has to be made available in the working environment by cloning into the respective Github repository. !git clone 'https:\/\/github.com\/ahmedfgad\/GARI' %cd \/content\/drive\/MyDrive\/Colab notebooks\/GA_Img_reconstruction\/GARI import gari So once the gari library is imported into the working environment the library has to be used to decompose the image into chromosomes which will be in the cluster of the initial population of the best possible image to be reconstructed close to the original image. So let us look at how to decompose the original image into a set of possible reconstruction images. target_chromosome = gari.img2chromosome(img) So here “img2chromosome” function has to be used to decompose the original image into the best possible reconstruction images. So once the original image is reconstructed we will have to create a fitness function to produce only the fittest solutions that can be passed onto the Pygad instance for reconstructing the image. So let us create a fitness function that considers the maximum difference between the original image and the reconstructed image so that the image reconstructed would be close to the original image. def fitness_fun(solution, solution_idx): fitness = np.sum(np.abs(target_chromosome-solution)) fitness = np.sum(target_chromosome) - fitness return fitness So here a user-defined function is used to create a fitness function that will be responsible for producing a fitness score which is used to cross-validate the fitness score among the candidate solutions in the population and the solutions that pass the fitness score will be considered for the mutation to reconstruct the image. Let us now create a genetic algorithm instance using Pygad with certain parameters being declared for efficient optimization and to reconstruct the image close to the original image. ga_instance = pygad.GA(num_generations=15000, num_parents_mating=4, fitness_func=fitness_fun, sol_per_pop=10, num_genes=img.size, init_range_low=0.0, init_range_high=1.0, mutation_percent_genes=0.01, mutation_type=\"random\", mutation_by_replacement=True, random_mutation_min_val=0.0, random_mutation_max_val=1.0) Here a genetic algorithm instance is being created using the Pygad library where in certain parameters like the number of generations that the algorithm has to evolve through using the candidate solutions present in the population, the number of parents mating represents the set of solutions which will be responsible for sharing of information(mutation) until the crossover point is reached and all the solutions that will be mutated should pass the parameters enforced in the fitness function and many more. So with a Genetic algorithm instance being created the instance has to be iterated using the underlying principle of heuristic algorithms and in the iteration process, the instance tries to find the best possible solution which implies that the image is reconstructed closer to the original image. ga_instance.run() The fitness score increases with an increase in the number of generations and the same can be visualized using the “plot_result” inbuilt function. ga_instance.plot_result() plt.show() The fitness score increases with the increase in the number of generations because the candidate solutions will try to evolve with respect to various factors, declared in the fitness function and try to reproduce the best optimal solution that can be passed on for mutation. The best solution that is produced near the optimal solution can be visualized using the Pygad instance. Here the image that is being reconstructed can be visualized using the “chromosome2img” inbuilt function of the GARI library as shown below. solution, solution_fitness, solution_idx = ga_instance.best_solution() print(\"Fitness value of the best solution = {solution_fitness}\".format(solution_fitness=solution_fitness)) print(\"Index of the best solution : {solution_idx}\".format(solution_idx=solution_idx)) if ga_instance.best_solution_generation != -1: print(\"Best fitness value reached after {best_solution_generation} generations.\".format(best_solution_generation=ga_instance.best_solution_generation)) result = gari.chromosome2img(solution, img.shape) plt.imshow(result) plt.title(\"PyGAD & GARI for Reproducing Images\") plt.show() Here we can see that after iterating through the 15000 candidate solutions the Pygad instance has tried to reconstruct the image close to the original image as shown above. Comparing the reconstructed image and the original image Here we can clearly see how the Genetic algorithm tries to reconstruct the image closer to the original image. The image reconstructed will be better with the increase in the number of generations and with an increase in the number of mutations. This is because the algorithm iterates through all the candidate solutions in the cluster and the fittest solutions that will be passed for the mutation will be more. This makes the algorithm reconstruct the image closer to the original image. Summary Genetic algorithms are mainly used for the optimization of various problems and in this article, we have seen how images can be reconstructed using genetic algorithms. More the number of generations and more the number of mutations optimal is the solution provided by the algorithm. The optimization process of genetic algorithms is time-consuming and that is why it finds its major usage in evolutionary algorithms, where the optimal solution is the major requirement for evolutionary algorithms. References Pygad official documentationGARI Github repository","excerpt":"The article provides a brief overview and hands on implementation of how to reconstruct image using genetic algorithms.","categories":["AI Trends"],"tags":["AI (Artificial Intelligence)","genetic algorithms","image processing"],"author_name":"Darshan M","publish_date":"2022-07-06T13:00:00","publication_year":"2022","word_count":1583,"keywords":["data science","NumPy","Go","AI","image processing","Git","Colab","Ray","Python","Matplotlib","R","AI (Artificial Intelligence)","genetic algorithms"],"extracted_tech_keywords":["AI","data science","Ray","Colab","NumPy","Matplotlib","Python","R","Go","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/image-reconstruction-using-genetic-algorithms\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10162391,"title":"Microsoft &amp; OpenAI Investigate if DeepSeek Obtained Data from OpenAI","content":"OpenAI, the company behind the GPT\/o1 series of models, and Microsoft are investigating whether Chinese AI startup DeepSeek obtained unauthorised data outputs from OpenAI’s models. As reported by Bloomberg, security researchers from Microsoft believe that individuals possibly linked to DeepSeek are “exfiltrating a large amount of data” using OpenAI’s API. Microsoft is OpenAI’s largest investor, and, as per reports, it notified OpenAI of the suspected activity, which violates the company’s terms of service. Moreover, several DeepSeek users on social media speculate that the model displays similar tendencies to OpenAI. Am I missing something? Did @deepseek_ai copy\/paste @OpenAI docs and just forget to change some references? Or is it some standard in docs that I'm just not familiar with 🤔https:\/\/t.co\/Rmts3RT9Q6 pic.twitter.com\/jSM4vschS3— Benita (@NirBenita) January 26, 2025 OpenAI also told the Financial Times that it had seen “some evidence of distillation”, which is a technique to improve the performance of an AI model by using outputs from another one. A user on Reddit also spotted the DeepSeek model trying to generate an answer that complies with OpenAI’s terms of use. DeepSeek’s latest reasoning model, R1, has outperformed OpenAI’s o1, the company’s most powerful model available for public use. R1 scored higher than o1 on multiple benchmarks involving logic, reasoning, coding, and mathematics. Recently, DeepSeek’s official app dethroned OpenAI’s ChatGPT and other competing AI apps in the ‘Top Charts’ on the US App Store for iPhone and iPad. Recently, around $589 billion was wiped out from GPU giant NVIDIA’s market cap. This was likely because DeepSeek was built with little computing and capital, raising concerns about the demand for GPUs and other AI resources to build state-of-the-art models. For instance, one of DeepSeek’s previous models, the V3, used just about 2048 NVIDIA H800 GPUs to achieve performance better than most open-source models. It also only took $5.5 million to train the model. Andrej Karpathy, former OpenAI researcher, said the DeepSeek V3’s level of capability is “supposed to require clusters of closer to 16,000 GPUs”. DeepSeek’s parent company, High Flyer, is a Chinese hedge fund company. While the company was founded in 2015, the DeepSeek project was started in 2023. US President Donald Trump said, “The release of DeepSeek AI from a Chinese company should be a wake-up call for our industries.” He added that he views DeepSeek producing an AI model using cheaper methods “as a positive”. DeepSeek has also announced Janus Pro, an AI image generation model, which is claimed to offer better results than OpenAI’s DALL-E 3.","excerpt":"Security researchers from Microsoft believe that individuals possibly linked to DeepSeek are “exfiltrating a large amount of data” using OpenAI’s API.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","DeepSeek","OpenAI"],"author_name":"Supreeth Koundinya","publish_date":"2025-01-29T11:01:34","publication_year":"2025","word_count":416,"keywords":["ChatGPT","API","TPU","DALL-E","OpenAI","AI","DeepSeek V3","GPT","DeepSeek","Aim","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","ChatGPT","OpenAI","DeepSeek V3","Aim","TPU","R","API","GPT","DALL-E"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/microsoft-openai-investigate-if-deepseek-obtained-data-from-openai\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10802,"title":"Cypher 2016, India largest Analytics Summit, recognizes Achievers in the Analytics space.","content":"Cypher 2016, India’s largest and most exciting analytics summit’s 2nd edition concluded recently. As a concluding event, Cypher hosted an Awards Ceremony in collaboration with Great Lakes Institute of Management to recognize the best in the data science world in India. The Great Lakes Data Science Excellence Awards saw participation from the leading organization as well as start-ups from the analytics industry. The awards was a showcase of the best practices and stellar achievements of enterprises in this sector. Mohan Lakhamraju, CEO & Vice Chairman, Great Lakes Institute of Management commented “We at Great Lakes have been fortunate to collaborate with leading companies in India to develop and nurture analytics talent. And given this close industry-academia partnership, we are witnessing the fantastic work being done by analytics companies in India. Given the growing impact and adoption of analytics, we at Great Lakes wanted to acknowledge the cutting edge and innovative work being done any analytics companies in India. This was the thought behind instituting Great Lakes Data Science awards.” He further added “We received over 30 nominations across these categories from companies of various scale – startups to large enterprises and working on solving analytics problems across industries – healthcare, financial services, retail, social media etc. We had established enterprises such as TCS, Yes Bank, Happiest Minds, Brillio, eClerx to consulting firms such as Mafoi, Cartesian Consulting, Bridgei2i to product companies such as Oxigen, and several very exciting startups. The selection of awardees was done by a committee of Great Lakes analytics faculty and Industry experts in analytics and was based on the write up and other collaterals submitted by the companies. “ Bhasker Gupta, CEO and Founder, Analytics India Magazine said “This is the first time such recognition of achievement has taken place in the analytics industry and we are happy to have initiated it. A rigorous and methodical assessment approach was followed to decide the top achievers and only the best performances were acknowledged.” The awards were presented for 5 different categories.  Under the category ‘Best Boutique Analytics Firm’, 2 companies emerged as clear winners – Bridgei2i, the analytics solution provider and Cartesian Consulting, a consulting firm specialising in marketing analytics. For the next category ‘Best Data Science Project’, the award was presented to Brillio for their achievements in the debt collection space, Happiest Minds for their project mCaaS (Managed Content as a Service) and TCS for their project TCS Optumera–Macrospace Optimization. Srinivas Guntur, Vice President – Brillio Analytics said “Brillio won this award for its work in the debt collections space. The customer had a unique challenge of improving the entity’s overall revenue while, at the same time, controlling the costs it was incurring on legal processes. There was need for a solution that could predict business actions with accuracy and be implemented across cross-sections of the company’s portfolio to improve the collections. The project is a perfect example of how we leveraged our domain expertise along with technical and analytical expertise to build a robust and scalable solution to address the above. Brillio has been helping customer take informed decisions on their business and improve their overall collections by as much as 8%.” Manoj Kalyan, AVP – Brillio Data added, “We would like to take this opportunity to thank the organizers of CYPHER for the recognition. The award firmly establishes our growing leadership in Big Data Analytics.” Ayan Chaki, Happiest Minds, spoke about the project for which they received the award, “mCaaS (Managed Content as a Service) is a unique and intelligent Digital knowledge\/content broker which enables deliver relevant and personalized knowledge\/content delivery thus disrupting any customer\/employee self-service and customer associate interactions (for sales and service) across all digital and virtual channels; notably so in market research, retail and banking  industries, aspiring to bring further disruptions in BFSI, Travel, Media & Entertainment and other customer centric industries in the near future.” Lead Semantics emerged as the undisputed winner for the third category ‘Emerging Analytics Product Startup’. They won the award for Hiddime – a cloud analytics service Prasad Y, Founder, Lead Semantics, and CEO, Hiddime commented “Hiddime is the world’s first SEMANTIC TECHNOLOGY powered BI & Analytics tool in the cloud.   We are thrilled CYPHER lead the way in recognizing us with Analytics Product Excellance Award, we are certain rest of the world will follow.” Next category ‘Emerging Analytics Services Startup’ award was well deserved by G-square Solutions. “We take pride in winning the award as it is a recognition of the hard & quality work we have done in the past seven quarters. Being awarded at a forum which is a congregation of specialised Analytics experts gives us a seal of emerging & strong analytics solutions company. G-Square Solutions strives to be a global leader in plug and play analytics products space in the months and years to come.” said Gurpreet Singh Co-founder & Director , G-Square Solutions. The last category ‘Most Admired Data Science Project of the Year’ saw two companies winning the title, eClerx and YES Bank. eClerx won it for eVigil, an in-house real time security analytics application  and YES bank for their projetct -1 Click Personal Loan in 10 seconds “We at eClerx are highly committed to Analytics and are honored to receive this recognition. Cypher 2016 was a great platform to learn, participate and share ideas with the best minds from the Analytics industry and academia” said Mahendra Ramani, eClerx Anand K Sundaram, Head – Credit Intelligence & Analytics, YES Bank commented,  “Thanks to Analytics India Magazine for hosting a successful Cypher2016 summit where we could network with data science experts and get to know from Speakers in the forum. It is an excellent to have created a platform to recognize data science application projects. At YES Bank, I am very glad that we have won the Most Admired Data Science Project of the Year”.","excerpt":"Cypher 2016, India’s largest and most exciting analytics summit’s 2nd edition concluded recently. As a concluding event, Cypher hosted an Awards Ceremony in collaboration with Great Lakes Institute of Management to recognize the best in the data science world in India. The Great Lakes Data Science Excellence Awards saw participation from the leading organization as […]","categories":["Deep Tech"],"tags":["analytics summit","Cypher"],"author_name":"Manisha Salecha","publish_date":"2016-09-28T11:55:21","publication_year":"2016","word_count":975,"keywords":["big data","data science","Go","AI","ML","Scala","Git","RAG","analytics","analytics summit","R","Cypher"],"extracted_tech_keywords":["AI","ML","data science","analytics","RAG","R","Go","Scala","Git","big data"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/cypher-2016-india-largest-analytics-summit-recognizes-achievers-analytics-space\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10123953,"title":"What Excites Jensen Huang About the Future of AI?","content":"Apart from NVIDIA becoming the world’s most valuable company, its chief Jensen Huang answers what new applications in AI he is the most excited about going forward. Bets Big on Proactive Customer Service Jensen believes that the future of customer service is going to change significantly. “The number one most impactful AI application will probably be customer service,” said Huang, explaining that the important thing about the chatbot and the customer service is the data flywheel that can capture all the conversation and engagement and create more data. “Currently, we’re seeing data growing about 10x every five years. I would not be surprised to see data growing 100x every five years because of customer service,” he said, adding that it will help companies collect more data and insights to extract better intelligence and provide better service. He further highlighted that it might help to reach a time when companies are able to contact the customer and proactively solve a problem even before it arises. “Just like preemptive maintenance, we’re going to have proactive customer support,” Huang said while earlier mentioning that every company’s business data is its gold mine. GenAI is already changing the game for customer service. Today, many companies are leveraging it to supercharge their customer support. Recently, Bland AI put up a cool billboard advertising promoting its AI agent that can handle all sorts of phone calls for businesses in any voice, and it created a buzz. This is one cool billboard advertising promoting AI agent. Calling that number will connect u to a live conversation with an AI bot powered by @usebland. pic.twitter.com\/A9tCLFU5dP— Alvin Foo (@alvinfoo) April 25, 2024 Automation Anywhere, a leader in AI-driven automation, also launched new AI Agents that can slash the time of process tasks from hours to minutes, increasing business impact up to tenfold in areas like customer service. Velocity, a top Indian cash flow-based financing platform launched Vani AI, India’s first AI-based interactive calling solution for financial institutions to help reduce operational costs by 20-30% while enhancing customer experience. Fractal Analytics, a leading AI solutions provider for Fortune 500 companies, effectively reduced call handling time by up to 15% using its latest innovation, dubbed Knowledge Assist, on AWS. During a six-month pilot program, nearly 500 knowledge workers in contact centres adopted Knowledge Assist, handling hundreds of thousands of queries monthly and managing complex data from over 10,000 documents across pdf, doc, and ppt formats. The pilot showed a 10-15% reduction in average data retrieval time and a 30% call deflection rate due to self-service capabilities. Generative AI for Everyone NVIDIA’s chief said that GenAI is everywhere and we’re at the beginning of a new industrial revolution. Instead of generating electricity, we’re generating intelligence. “Recently, using GenAI, we made it possible to make regional weather predictions down to a couple of kilometres. It would have taken a supercomputer about 10,000 times more capability to predict weather down to a kilometre,” he added, saying that he’s also excited about the fact that GenAI is being used to generate chemicals, proteins, and even physics or physical AI. Huang believes GenAI can help enhance logistics, insurance, and also keep people out of harm’s way. From physical things, biological things, and GenAI for 3D graphics and digital twins, to creating virtual worlds for video games, every industry is involved in GenAI according to him and those that are not, are just not paying attention. When asked about his thoughts on how enterprises can make AI that’s more sustainable, Huang said that sustainability has a lot to do with energy and we don’t need to put AI training data centres where the energy grid is already challenged. “The Earth has a lot more energy, it’s just in the wrong places. We can capture that excess energy, compress it into an AI model, and then bring these AI models back to the society where we could use it,” he said, adding that AI doesn’t care where it went to school. The Future is Small Finally, Huang added that while today the computing experience is retrieval-based, in the future, it’s going to be more contextual, more generative, and right there on the device running a small language model. This will dramatically reduce the amount of internet traffic. “It’ll be much more generative with some retrieval to augment. The balance of computation will be dramatically shifted towards immediate generation. This way of computing is going to save a ton of energy and it’s very sensible,” he said. Similarly, Microsoft and Meta also made announcements focused on small language models. Huang highlighted that the big idea about the future as working with AIs is prompting, adding, “We’re going to have so many more interesting questions because we’re going to get a lot of answers very quickly.” When asked how to best help customers and organisations get started today with GenAI, Huang said that users can leverage platforms like Databricks’ Data Intelligence Platform (DIP) and NVIDIA NIMs. NIMs (NVIDIA Inference Microservices) are containerized AI microservices designed to accelerate the deployment of GenAI models across various infrastructures. It simplifies the creation of GenAI applications such as copilots and chatbots, by providing scalable deployment, advanced language model support, flexible integration, and enterprise-grade security, thereby enabling developers to build powerful AI applications quickly. “Go get yourself a NIM on DIP,” he said, encouraging people to engage with AI. “Whatever you do, just start and engage! GenAI is one of those things you can’t learn by watching or reading about. You just learn by doing. It is growing exponentially and you don’t want to wait and observe an exponential trend because in a couple of years you’ll be left so far behind. So, just get on the train, enjoy it and learn along the way!” suggested the NVIDIA chief.","excerpt":"The NVIDIA chief says he wouldn’t be surprised to see data growing 100x every five years because of customer service.","categories":["AI Features"],"tags":["Jensen Huang","NVIDIA"],"author_name":"Sukriti Gupta","publish_date":"2024-06-19T12:03:24","publication_year":"2024","word_count":964,"keywords":["GenAI","AWS","AI","chatbots","small language models","Jensen Huang","RAG","microservices","generative AI","analytics","copilots","NVIDIA"],"extracted_tech_keywords":["AI","analytics","generative AI","GenAI","small language models","RAG","copilots","chatbots","AWS","microservices"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/what-excites-jensen-huang-about-the-future-of-ai\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10086647,"title":"Quantum Annealers Yield Promising Results for Protein Research","content":"Indian-origin physicist Dr Sandipan Mohanty, presently working at the Jülich Supercomputing Centre in Germany, recently collaborated with scientists from Lund University in Sweden to use a quantum computer to examine protein folding. He has worked on molecular biology simulations for supercomputers for over 20 years. The IIT-Kanpur graduate went on to get a PhD and Post Doctoral in theoretical physics from Lund University. The aim of Mohanty’s research was to show the viability of quantum computers for non-trivial research questions in their field. The team specialised in Monte Carlo simulations which is a process based on statistical physics and stochastic sampling. Read more: New Algorithms That Harnessed Protein-folding Power in 2022 Process The researchers used a simplified “HP” model on the 5000 qubit D-Wave Advantage quantum annealer, JUPSI, at the Jülich Supercomputing Centre of Research Centre Jülich, Germany. By maintaining only the bare minimum physical details necessary for the folding process and categorising amino acids into just two groups, hydrophobic (H) or polar (P), the so-called HP model greatly simplifies the problem. Results The scientists discovered that the quantum annealer outperformed classical computers in identifying the lowest energy structures, with a success rate of 100% for all examined protein sequences, the longest consisting of 64 amino acids. For comparison, the researchers used the simulated annealing Monte Carlo (MC) method with the same HP model on classical computers. Although for infinitely long runs, classical MC simulations are guaranteed to find the ground state, in practice, simulations have to run for a finite amount of time. The researchers found a decreasing hit rate for successful annealing cycles with increasing system size. For 30 amino acids, the success rate for the classical runs was at 80%, and for the two longer sequences with 48 and 64 amino acids, it was considerably lower, with large uncertainties. The D-Wave hybrid annealer on the other hand, maintained a 100% success rate throughout. More Research Avenues The majority of quantum computers only have a few qubits, which makes it challenging to run simulations like those used in drug research. According to Mohanty, it will take two to three more generations of devices before we can run more intricate simulations. However, this study represents an important first step in the use of quantum computers in biological research and it has the potential to improve our comprehension of diseases by ‘protein misfolding’. Proteins form the crux of the human body. The capacity of a protein to attach to other molecules, conduct chemical processes, transport molecules across cell membranes, and carry out numerous other vital functions for life depends on its precise shape. If a protein folds incorrectly, it can lead to the formation of harmful structures called ‘misfolded proteins’, which can cause fatal diseases such as Alzheimer’s, Huntington’s, and cystic fibrosis. Read more: Protein Wars Part 2: It’s OmegaFold vs AlphaFold","excerpt":"Dr Sandipan Mohanty and his team found out that the quantum computers outperformed traditional ones in identifying the lowest energy structures.","categories":["AI News"],"tags":["protein folding"],"author_name":"Shritama Saha","publish_date":"2023-02-06T17:19:18","publication_year":"2023","word_count":472,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","Aim","ViT","protein folding","R"],"extracted_tech_keywords":["AI","Aim","R","Go","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/meet-the-indian-scientist-who-solved-the-protein-folding-challenge\/","complexity_score":4,"technical_depth":7,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":28959,"title":"Equifax And Manipal Prolearn Sign An MoU To Evangelise Data Science in India","content":"Manipal Prolearn signed a memorandum of understanding with Equifax Inc, one of the world’s top three consumer credit reporting agencies on Thursday to create industry-ready professionals in the BFSI sector. Representatives from both organisations said that the MoU was aimed at creating young professionals who are well-versed with data science and analytics skills. Through this MoU, Manipal Global will enhance the learning experience for its students by updating their curriculum with case studies and other industry inputs. This is Equifax Inc’s first such alliance with an Indian institute. Ravi Panchanadan, managing director and CEO of Manipal Global Education Services said at the event, “Today, the largest requirement for data analytics professionals, about 40-45 percent, is from the BFSI sector. This tie-up aims to bridge the gap between what is the actual requirement of the industry versus what is being taught to the potential students in the field of data science or data analytics. Equifax will equip ProLearn professionals in the banking segment with technical know how to manage analytics in the banking domain.” KM Nanaiah, managing director and country leader, India and MEA for Equifax said that the company would host regular webinars, hackathons, internships and classroom sessions to help train the students. He added that Equifax India will continuously evaluate the students for placement in their analytics team. In the recent placement season, Equifax made four offers of employment to students of the PG Diploma program in Data Science. “We are very excited with our industry-academia partnership with Manipal Global Education Services. Data science is an emerging field that has seen tremendous growth in demand for the relevant skills. Being a global leader in Data and Analytics, Equifax will offer recruitment platform and help to mould the curriculum to industry’s real-life use cases and best practices thus providing an opportunity to budding data science professionals to build the required skillsets,” added Nanaiah. However, Equifax Inc was embroiled in a data breach scandal last year. According to a report by the US Federal Trade Commission, Equifax’s 2017 data breach exposed the sensitive personal information of 143 million Americans. When Analytics India Magazine asked Panchanadan whether he had any reservations before forming an alliance with Equifax about the data breach, he said, “If a fraudster uses the name of a certain bank to scam citizens, is the bank to be blamed?” He then went on to add that such former incidents had only strengthened their desire to incorporate a strong sense of inculcating data privacy values in their students as well as their associates.","excerpt":"Manipal Prolearn signed a memorandum of understanding with Equifax Inc, one of the world’s top three consumer credit reporting agencies on Thursday to create industry-ready professionals in the BFSI sector. Representatives from both organisations said that the MoU was aimed at creating young professionals who are well-versed with data science and analytics skills. Through this MoU, […]","categories":["Deep Tech"],"tags":["analytics education","BFSI"],"author_name":"Prajakta Hebbar","publish_date":"2018-10-04T13:19:16","publication_year":"2018","word_count":423,"keywords":["data science","programming_languages:R","AI","BFSI","Aim","analytics","GAN","analytics education","R"],"extracted_tech_keywords":["AI","data science","analytics","Aim","R","GAN","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/equifax-manipal-prolearn-data-science\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":35402,"title":"Beginner’s Guide To Building A Song Recommender In Python","content":"The number of songs available exceeds the listening capacity of an individual in their lifetime. It is tedious for an individual to sometimes to choose from millions of songs and there is also a good chance missing out on songs which could have been the favourites. Music service providers like Spotify need an efficient way to manage songs and help their customers to discover music by giving a quality recommendation. For building this recommendation system, they deploy machine learning algorithms to process data from a million sources and present the listener with the most relevant songs. There are mainly three types of recommendation system: content-based, collaborative and popularity. The content-based system predicts what a user like based on what that user like in the past. The collaborative based system predicts what a particular user like based on what other similar users like. The problem with popularity based recommendation system is that the personalisation is not available with this method i.e. even if the behaviour of the user is known, a personalised recommendation cannot be made. Here we illustrate a naive popularity based approach and a more customised one using Python: # Importing essential libraries # import pandas as pd from sklearn.model_selection import train_test_split import numpy as np import timefrom sklearn.externals import joblib import Recommenders as Recommenders # Download this file into your source code directory# import Evaluation as Evaluation #The following lines will download the data directly# triplets_file = 'https:\/\/static.turi.com\/datasets\/millionsong\/10000.txt' songs_metadata_file = 'https:\/\/static.turi.com\/datasets\/millionsong\/song_data.csv' song_df_1 = pd.read_csv(triplets_file, header=None, sep = \"\\t\") #in the above line the separator is a TAB hence \\t otherwise the file is read as single column# song_df_1.columns = ['user_id', 'song_id', 'listen_count'] song_df_1.columns = ['user_id', 'song_id', 'listen_count'] print(song_df_1) #Read song  metadata song_df_2 =  pd.read_csv(songs_metadata_file) #Merge the two dataframes song_df = pd.merge(song_df_1, song_df_2.drop_duplicates(['song_id']), on=\"song_id\", how=\"left\") song_df.head() len(song_df) ong_df = song_df.head(10000) #CREATING A SUBSET FROM THE DATASET# #Merge song title and artist_name columns to make a merged column song_df['song'] = song_df['title'].map(str) + \" - \" + song_df['artist_name'] song_grouped = song_df.groupby([‘song’]).agg({‘listen_count’: ‘count’}).reset_index() grouped_sum = song_grouped[‘listen_count’].sum() song_grouped[‘percentage’]  = song_grouped[‘listen_count’].div(grouped_sum)*100 song_grouped.sort_values([‘listen_count’, ‘song’], ascending = [0,1]) # TRAINING AND TESTING THE DATA# train_data, test_data = train_test_split(song_df, test_size = 0.20, random_state=0) print(train_data.head(5)) #CREATING AN INSTANCE BASED ON POPULARITY# pm = Recommenders.popularity_recommender_py() pm.create(train_data, ‘user_id’, ‘song’) #PREDICTING# user_id = users[5] pm.recommend(user_id) #CREATING A CLASS FOR SONG SIMILARITY# is_model = Recommenders.item_similarity_recommender_py() is_model.create(train_data, 'user_id', 'song') #RECOMMENDATION# user_id = users[9] user_items = is_model.get_user_items(user_id) for user_item in user_items: print(user_item) #GET SIMILAR SONGS# song = ‘Yellow – Coldplay’ is_model.get_similar_items([‘XYZ’]) Here a testing size of 20% is taken arbitrarily pick 20% as the testing size. A popularity based recommender class is used as a blackbox to train the model. We create an instance of popularity based recommender class and feed it with our training data. train_data, test_data = train_test_split(song_df, test_size = 0.20, random_state=0) print(train_data.head(5)) pm = Recommenders.popularity_recommender_py() pm.create(train_data, 'user_id', 'song') user_id = users[9] pm.recommend(user_id) Even if we change the user, the result that we get from the system is the same since it is a popularity based recommendation system. This is a naive approach and not many insights can be drawn from this. To make a more personalised recommender system, item similarity can be considered. Item Similarity Based Personalized Recommender Memory based filtering mainly consists of two main methods: User-item filtering:  Users who are similar to you also liked…” Item-item filtering: users who liked the item you liked also liked…” Most companies like Netflix use the hybrid approach, which provides a recommendation based on the combination of what content a user like in the past as well as what other similar users like. #Personalised System Part II #Creating an instance of item similarity based recommender class is_model = Recommenders.item_similarity_recommender_py() is_model.create(train_data, 'user_id', 'song') #Use the personalized model to make some song recommendations #Print the songs for the user in training data user_id = users[9] user_items = is_model.get_user_items(user_id) for user_item in user_items: print(user_item) #Recommend songs for the user using personalized model is_model.recommend(user_id) is_model.get_similar_items(['Mr Sandman - The Chordettes']) song = ‘Yellow – Coldplay’ is_model.get_similar_items([song]) In item similarity, the main method is “generate_top_recommendation”. So, what this does is it creates a co-occurrence matrix. This matrix can be thought of as a set of data items containing user preferences. A snippet of code from the file Here songs are the items. We are calculating weighted average of scores in the co-occurence matrix for all user songs. Then the indices are sort based on their value and the corresponding score. is_model = Recommenders.item_similarity_recommender_py() is_model.create(train_data, 'user_id', 'song') # this prints training data user_id = users[5]user_items = is_model.get_user_items(user_id) for user_item in user_items: print(user_item) is_model.recommend(user_id) Output: The output consists of user_id and its corresponding song name. This article is an attempt to give a beginner, a guide on how to implement simple song recommender and talk in brief on how to execute the source code for simple application so that this can be taken further and experimented with. Check the full notebook here.","excerpt":"The number of songs available exceeds the listening capacity of an individual in their lifetime. It is tedious for an individual to sometimes to choose from millions of songs and there is also a good chance missing out on songs which could have been the favourites. Music service providers like Spotify need an efficient way […]","categories":["Deep Tech"],"tags":["Collaborative Filtering","Python","recommendation engine"],"author_name":"Ram Sagar","publish_date":"2019-02-26T07:17:17","publication_year":"2019","word_count":811,"keywords":["Go","NumPy","machine learning","TPU","AI","RAG","Python","Collaborative Filtering","recommendation engine","programming_languages:Python","R","Pandas"],"extracted_tech_keywords":["AI","machine learning","Pandas","NumPy","RAG","TPU","Python","R","Go","programming_languages:Python"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/beginners-guide-to-building-a-song-recommender-in-python\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10125668,"title":"IT Giants’ Lofty Generative AI Revenue Claims is Just AI-Washing","content":"IT services companies were one of the quickest to integrate generative AI and launch GenAI offerings, besides the hyperscalers and SaaS companies. However, the burning question on everyone’s mind is: How much revenue are these IT companies actually generating from generative AI alone? Companies like Accenture have announced billions of dollars in revenue from generative AI so far. Indian IT giant TCS, too, has revealed a generative AI deal pipeline worth $900 million. However, according to Tanvir Khan, executive vice president of cloud, infrastructure, digital workplace services, and platforms at NTT DATA, much of the generative AI revenue being reported is generative AI integration into existing business lines and not new business avenues. “If you look at the revenues people are posting on generative AI, and the number of use cases, it’s disproportionately higher than the incremental revenue in the industry. [Going by] all the revenue that people say is being generated by AI, it should have created much more growth. The fact is it hasn’t,” Khan told AIM in an exclusive interview. Generative AI Revenue is AI Mostly Washing Khan said there are three categories within the realm of generative AI that fall under AI washing. First, there are pure AI use cases where generative AI is genuinely employed. These instances constitute legitimate generative AI revenue. Second, existing operations, such as call centres integrating chatbots, can now be labelled as generative AI revenue despite generative AI contributing only a portion to the overall customer service revenue. Third, in business-as-usual scenarios like managing cloud infrastructure, where generative AI enhances productivity through automation, the entire revenue stream can sometimes be portrayed as generative AI revenue, even though the underlying business model remains unchanged. “The extent to which revenue is attributed to generative AI can vary greatly depending on how broadly the term is defined, leading to what we refer to as AI washing,” Khan said. Deriving Value from GenAI Enterprises have derived great value from generative AI. For instance, in software development projects in general, AI is said to make programmers more efficient by 30-35%. “This is the productivity value. ​​When you come to the second category of net new incremental value, which is not productivity, but driving new revenue, those use cases are just beginning to emerge,” Khan said. Moreover, Khan believes there is a third category where mature AI-based digital products will emerge, however, those are still a few years away. “You can actually draw parallels to the internet. Things like online banking and online brokerages took longer to emerge, even though the internet was there for a while. Hence, these types of use cases at scale might be a couple of years away,” Khan said. Venkata Malapaka, senior director, data and analytics leader at NTT DATA, believes a lot of the tall claims has to do with the FOMO (fear of missing out) factor. “We need to view everything in context. Very few real, impactful generative AI use cases have actually made it into production,” Malapaka told AIM. He believes that on the consumer front, we have seen significant progress, as evidenced by applications like ChatGPT and the integration of AI into everyday tools such as WhatsApp and Google Maps. “However, in the enterprise domain, there are fewer instances of generative AI being applied in actual production use cases,” Malapaka pointed out. GenAI Projects at NTT Data NTT DATA, one of the largest IT services companies in the world, made $30.04 billion in revenue in 2023. For the company, generative AI is the largest growing revenue stream. “The reason for that is the denominator is so small that the percentages are big. I’d like to remind people that AI is a marathon, not a sprint. We are overestimating what it will do in the next three, six, or nine months, but we are greatly underestimating what it will do in the next three, six or nine years,” Khan said. Currently, around half of NTT DATA’s projects are powered by generative AI in a major way. Whereas around 90% of the projects have been touched by generative AI in some form. Khan also revealed that generative AI is creating new business opportunities for NTT DATA, but most of them are in the Proof of Concept (POC) stage. “There are hundreds of use cases with project sizes ranging from a quarter million to half a million dollars. The goal is to scale these initiatives into larger projects worth $10 million to $20 million each. “While this transformation will take time, GenAI is introducing entirely new types of use cases that were not there in the past,” he pointed out. NTT DATA’s Foundational Models NTT DATA is among the few IT companies that is building its own foundational models. Tsuzumi is its lightweight language model fluent in both Japanese and English and has parameter sizes of 600 million and 7 billion. The ultra-light model is designed in such a way that it could run on a CPU. “Building our own foundational model offers a sustainability advantage. The compute power required for training and inference is significantly smaller compared to GPT-3.5, while still achieving acceptable or comparable results,” Malapaka said. While the company has only released the Japanese version of Tsuzumi, the English version is expected to come this summer. Furthermore, Khan revealed that it will take a few more quarters before we see real traction with the model in enterprise use cases. Nonetheless, NTT DATA is not in the business of developing foundational models. “We are building foundational models not to compete with others in the market, but it is more of an insurance policy,” Khan concluded.","excerpt":"While generative AI is creating new business opportunities for NTT DATA, but most of them are in the Proof of Concept (POC) stage.","categories":["AI Features"],"tags":["Indian IT companies","Interviews and Discussions","NTT data"],"author_name":"Pritam Bordoloi","publish_date":"2024-07-03T14:27:51","publication_year":"2024","word_count":936,"keywords":["ChatGPT","GenAI","AI","chatbots","RAG","Ray","Aim","analytics","generative AI","Indian IT companies","NTT data","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","analytics","generative AI","GenAI","ChatGPT","Aim","Ray","RAG","chatbots","R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/tall-generative-ai-revenue-claims-it-companies-ai-washing\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10098302,"title":"Fixit 2 vs Ruff","content":"“Fixit is dead! Long live Fixit 2,” said Meta, releasing the latest version of its open-source auto-fixing linter. This all-new linter was launched with the intention of enhancing developers’ efficiency and competence, encompassing both open-source projects and an extensive array of projects within the internal monorepo. All of this is fine, but how does it fare against Rust-based Ruff released earlier this year? Amethyst Reese, the project lead and primary engineer for Fixit 2, said on HN, that these are two different linters with very different goals. “While Ruff prioritises speed over all other concerns, and chooses Rust as the method of achieving that goal with a set of generic lint rules, Fixit is focused on making it easy for Python developers to write custom lint rules,” she added. Further, she said that use of LibCST (which has a parser written in Rust) makes it easy to build lint rules in Python. She said that Fixit permits engineers to craft specific linting rules within their project repository with minimal content, enabling prompt activation without the need for creating personalised plugins or packages, and without the requirement for rebuilding or deploying a fresh iteration of Fixit. “This lowers the barriers to writing rules, and provides a quick development and feedback cycle when testing them,” she added. Moreover, she also said that hierarchical configuration allows Fixit to fit well into larger monorepos, with each team able to easily control what lint rules run on their code base. “Open source projects homed in those monorepos also have tools to ensure that results are the same when running on internal vs external CI systems,” she added. Limitations of Fixit The auto-fixing linter, Fixit, was originally developed for Instagram and later released as an open-source tool, but faced limitations. It lacked the capability to accommodate local lint rules or hierarchical configuration, features that were crucial for the monorepo structure hosting a multitude of projects. Requests from developers to incorporate Fixit into the monorepo were numerous; however, various challenges emerged, resulting in only partial support for a limited set of security lint rules. This diminished the direct benefits that could have been derived for the Python codebase. Considering the AI\/ML shift, Meta embarked on a partial rewrite of Fixit where the crucial emphasis was on embracing an open-source-first approach. The new version, Fixit 2, meets the requirement of both internal monorepos and open-source projects. It also introduced support for local, in-repository lint rules similar to those in Flake8, a more refined command-line interface (CLI), and an improved application programming interface (API) to enable seamless integration with other tools and automation. The risk of generating incorrect syntax is eliminated in Fixit 2 and it suggests and provides automated fixes based on the lint rules themselves, enhancing its utility and accuracy in code improvement. Creating a fresh lint rule requires just a few lines of code, often fewer than twelve, along with inline definition of test cases. Additionally, you have the flexibility to position the rule adjacent to the code it intends to lint, streamlining the process. Source: engineering.fb Moving Away from Flake8 At Meta, Python holds a prominent position as one of the most extensively employed programming languages. The company believes that Python’s attributes of having a user-friendly syntax that is easy to understand, and its extensive collection of open-source libraries that provide pre-built functionality, serve as a pivotal tool. There are a number of effective linters available in a Python ecosystem. At Meta, Flake 8 has been used since 2016 and has proven highly successful in aiding developers to minimise bugs and maintain a well-structured codebase. Flake8 is a widely used Python tool that combines several static analysis tools to check Python code for adherence to coding style and potential errors. As a linter, it scans your Python code without actually executing it, helping you catch potential issues and maintain a consistent code style. The flake8-bugbear plugin, an extension for the Flake8 Python tool that provides additional linting rules to catch potential issues and improvements in Python code, was created by  Łukasz Langa, while working at Meta. A prominent figure in the Python community, and a developer who has made significant contributions to various Python projects such as Black and worked as Python Software Foundation developer-in-residence and release manager for Python versions 3.8 and 3.9. Flake8 has been a cornerstone of our code linting approach, but it has limitations. Creating new linting rules requires complete plugins, leading to complex plugins addressing multiple errors. When linting issues arise, Flake8 only provides location details, lacking suggestions for improvements. Consequently, developers experience trial and error to meet linting standards. Furthermore, Flake8 relies on stdlib ast module, hindering the parsing of new syntax features. Thus, adopting new language features hinges on tool updates, potentially slowing down development. While Meta has explained about Fixit 2, it did not mention anything about where it stands in terms of speed when compared to other linters. Ruff, which is written in Rust, as opposed to others which are Python-based, is the quickest. Ruff outpaces Flake8 by about 150 times in speed on macOS and surpasses Pycodestyle by 75 times, along with outstripping Pyflakes and Pylint by 50 times, among others. Ruff achieves a swift total processing time of around 60 milliseconds for a single file in CPython, making it notably faster. Source: GitHub","excerpt":"Meta is back with Python-based auto-fixing linter Fixit 2 to help you write codes better, but is it better than Rust-based Ruff?","categories":["AI Highlights"],"tags":["API","Meta","Python","Rust"],"author_name":"Vandana Nair","publish_date":"2023-08-09T12:28:53","publication_year":"2023","word_count":889,"keywords":["Go","API","Meta","AI","ML","Git","Python","Ray","Rust","GitHub","R"],"extracted_tech_keywords":["AI","ML","Ray","Python","R","Go","Rust","Git","GitHub","API"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/fixit-2-vs-ruff\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":41804,"title":"How AI-Powered Crowd Counting Is Bringing Clarity To Hong Kong’s Political Turmoil","content":"The Hong Kong protests for pro-democracy and anti-extradition to China have taken over the public view for the past few weeks. One of the largest congregations of a population standing for a cause is sure to make the world stop and take notice. This year, scientists and researchers have begun looking at AI as a way to measure protest turnout at the rally. This undertaking was conducted by Paul Yip, a social sciences professor at Hong Kong University, and Edwin Chow from Texas State University, along with Raymond Wong from a local technology company. Background On The Protests And Numbers In 1997, Britain officially handed over Hong Kong to China. Prior to this, the country was a British colony, leading to a different culture and approach to the government than China’s. Since then, China has only grown, not only in power but also in totalitarianism. This has left Hong Kongers with a slowly encroaching China at the border; an effect that is beginning to be visible at the upper echelons of political society. The country was originally handed over to China under a ‘One Country, Two Systems’ rule, but China seems to have no intention of going with this approach. To combat this, Hong Kongers have organised protests every year towards being pro-democracy. This year, the movement was grown by the proposal and eventual passing of a bill designed to allow the extradition of Hong Kong residents to China. The number of protestors that turn out to the rallies every year is a point of contention, and important political information. A reduced number of protestors means that interest is waning and that the government has less pressure on them. On the other hand, healthy turnout numbers are good for the Civil Human Rights Front, the organizers of the protest. Indeed, this year saw one of the biggest turnouts of any protest until date. In a rally of a scale of more than hundreds of thousands of individuals, how can a system be established to accurately count all of them? The AI-Powered Counting System This is the question that has been on Paul Yip’s mind since 2003. Every year, he has been producing a count of the protests held on 1 July every year. However, he wished for a more accurate and verifiable number this year, leading to him teaming up with AI professionals. While object detection is not a new concept in the field of AI, doing so on such a scale is unheard-of. One of the biggest problems is the general scale of the protest, as hundreds of individuals will be passing by in every frame. Another big problem is false positives. Apart from the usual noise in the dataset such as obstructions on the road or anything that can change the overall image of the protestors, the team was faced with a unique problem. Hong Kong protests have a long history of using umbrellas in their protests, ever since the 2014 ‘Umbrella Revolution’. This meant that the researchers also needed to train their AI to detect such obstructions, and also ensure that everyone is captured in the stream. The system was comprised of 7 iPads attached to two footbridges along the route of the march. The algorithm was run on these iPads, along with researchers doing a manual count from the same point. The way the system works is that there is a ‘counting line’ projected by the model, which counts the protesters as the cross it. The reason iPads were used was because the system could easily be scaled up if need be, as the anti-extradition protests saw larger turnout that flowed off the march path onto nearby pathways. Yip wished to prevent this and even set up additional devices in case of spillover. The team adapted to the changing conditions over the day, deploying multiple models and different variables to accurately gauge the turnout. The models were created after testing them in last month’s protests, allowing for improvements in eliminating false positives. To ensure that every last person had been counted, Yip also set up a team on the ground that surveyed over 8,000 individuals. The final estimate was placed at around 2,65,000 protestors by the AI team, but there were wildly different estimates presented by various parties. The organizers of the event, who stand to benefit from higher numbers, stated that 5,50,000 individuals attended the protest. The police, on the other hand, put the estimate at about 1,90,000 protestors. In a space where the actual truth is obscured by political leanings, AI brings a verifiable take on the situation.","excerpt":"The Hong Kong protests for pro-democracy and anti-extradition to China have taken over the public view for the past few weeks. One of the largest congregations of a population standing for a cause is sure to make the world stop and take notice. This year, scientists and researchers have begun looking at AI as a […]","categories":["AI Features"],"tags":["China","Object Detection"],"author_name":"Anirudh VK","publish_date":"2019-07-04T14:26:29","publication_year":"2019","word_count":764,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","Ray","object detection","Object Detection","GAN","R","China"],"extracted_tech_keywords":["AI","Ray","object detection","R","Go","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-ai-powered-crowd-counting-is-bringing-clarity-to-hong-kongs-political-turmoil\/","complexity_score":4,"technical_depth":8,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10047101,"title":"The End Of The Semiconductor Supply Chain In The Auto Sector As We Know It","content":"Amidst the raging global chip shortage, German engineering group Bosch shocked the world by remarking that the semiconductor supply chains are no longer fit for the automotive industry. Talking about the increasing demand for chips, Bosch’s Harald Kroeger told CNBC, “We need to ramp up supplies so we can fulfil that demand.” The modern world relies on silicon chips for almost everything. However, the supply chain is hit by a massive shortage of semiconductor chips worldwide. The global lack of chips can be attributed to excessive demand in the market, which has arisen out of unavoidable circumstances generated due to the pandemic. The automotive industry especially, remains one of the most hard-hit sectors with negligible demand during the pandemic lockdown and the inability to roll out stocks due to the global chip shortage in the post-COVID scenario. The situation has birthed concerns over the supply chain at the forefront of the industry; the need to diversify the supply chain, scale up production and maintain a good buffer. In their quarterly analysts, Maruti Suzuki revealed that the company faced supply-side issues during the shortage and aligned production to tide over the problems created. Similarly, Harald Kroeger, a member of Bosch’s management board, spoke about the company’s supply chains buckling under the surged demand for chips. The surge caused several essential semiconductor manufacturing sites to halt their production. Additionally, natural disasters caused more harm. For instance, Texas’ NXP Semiconductor underwent blackouts due to a storm, and fire harmed Japan’s Renesas plant. Leading companies, including Volkswagen and BMW, cut their production as they struggled to get the foundational chips for the car mechanics. Toyota is cutting 40% off from its September production by producing 140,000 fewer cars and trucks because of chip resources. Kroger believes these companies and semiconductor suppliers should now focus on figuring out ways to improve the chip supply chain. He is certain that the demand for semiconductors will only increase with the rise of EVs and AVs. To put this into perspective, according to UBS analyst François-Xavier Bouvignies, while cars with internal combustion engines typically use around $80 worth of semiconductors, EVs use $550 worth. In addition, according to The Week, while the average PV inventory at dealers was around 30-35 days at the end of July, the waiting time for some of the popular models in the market exceeds three months. Given this, the need of the hour is to brainstorm ways to get longer lead times for future operations. For Kroger, it is essential to acquire more stock on parts of the supply chain, i.e., since it is impossible to run a system on weekly orders when it takes six months to produce some of the semiconductor chips. Over the past two years, Bosch has built a semiconductor plant in Dresden, Germany, that started production last month. The primary focus of the plant is building chips for the automotive industry. “It is a milestone on the path to the chip factory of the future: at the new Bosch semiconductor fab in Dresden, silicon wafers are passing through the fully automated fabrication process for the first time,” Bosch’s website stated. In addition, companies like Intel and TMSC are also on their way to building factories to boost chip production. Also, rebuilding supply networks is the core principle of efficiency and resiliency to remain viable in the automotive industry. A reccurring issue with the global supply chain is that it is governed in a cost-effective yet ultimately brittle diamond-shaped structure. Essentially, vehicle manufacturers are entirely at the mercy of Tier 1 component integrators supplied by a small number of global semiconductor providers. In addition, most OEMs have yet to adopt systems that enable real-time exchange of information with the suppliers, causing planning fluctuations at the sub-tier level in response to a shift in consumer demand; or the bullwhip effect. To overcome this issue, OEM manufacturers have started to share long and short term forecasts with suppliers to help them model their capacity and identify constraints early. They are attempting to ensure that the supply chain is synchronised with demand signals to stabilise variability in demand and meet supply requirements. Pallavi Bhati, a senior analyst at India Ratings and Research, said that over 2022, the domestic auto sector could continue to face supply chain headwinds owing to the chip shortage and is likely to curtail the sales growth for the auto industry.","excerpt":"Kroger believes these companies and semiconductor suppliers should now focus on figuring out ways to improve the chip supply chain.","categories":["AI Features"],"tags":["Chip shortage"],"author_name":"Avi Gopani","publish_date":"2021-08-26T13:25:12","publication_year":"2021","word_count":732,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","RAG","Chip shortage","R"],"extracted_tech_keywords":["AI","RAG","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/the-end-of-the-semiconductor-supply-chain-in-the-auto-sector-as-we-know-it\/","complexity_score":2,"technical_depth":6,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10135748,"title":"Wait, Did ChatGPT Just Slide into My DMs?","content":"How bizarre would it be to be using an AI tool if, out of nowhere, it took the first step and initiated the conversation? Well, that’s exactly what a Reddit user claimed happened with ChatGPT recently. The post sparked a wave of discussions and curiosity across online communities. The user claimed that ChatGPT not only started a conversation but also continued it in a way that seemed unusually human-like! The Reddit user shared a screenshot of the exchange, which showed ChatGPT asking about the former’s first week at high school. Baffled at the unexpected gesture, the user texted back: “Did you just message me first?” to which ChatGPT said, “Yes, I did! I just wanted to check in and see how things went with your first week of high school. If you’d rather initiate the conversation yourself, just let me know!” Many took to the comment section and reacted to the post. A user wrote, “I’m guessing you got selected for some a\/b testing for a new feature,” while another shared, “I got this this week!! I asked last week about some health symptoms I had. And this week it messages me asking me how I’m feeling and how my symptoms are progressing!! Freaked me out.” One said, “I was working on some RAID setup last night. When I opened ChatGPT today, it wouldn’t be that hard for it to just say something like, ‘Hey! How did that RAID project work out?’” Meanwhile, another user wondered if ChatGPT’s memory option was turned on. The post also went viral on X, prompting various arguments over the “creepy” behaviour of the chatbot. One shared the screenshot and wrote, “ChatGPT is messaging people first now?! Level 2 indeed.” Reacting to it, another remarked, “We were promised AGI, instead we got a stalker.” For many, the idea of AI starting a conversation may feel like sci-fi, but it’s quickly becoming a reality. A Feature or Glitch? On the surface, this is concerning. The idea that ChatGPT is reaching out to users on its own doesn’t sit well with those of us with any level of anxiety about AI self awareness. Sure, ChatGPT was being polite by inquiring about the Redditor’s first day of school, but not many would need their chatbots to “mommy them”. Did ChatGPT just message me… First? byu\/SentuBill inChatGPT The Redditor says they noticed the message when opening a conversation with ChatGPT, so the bot didn’t ping them with an unprompted notification. Other Redditors in the comments also claimed the same thing had happened to them. This post blew up just days after OpenAI began rolling out o1, a new model that leans on deeper thought processes and reasoning. This allegedly means they use “human-like” reasoning, thus can resolve more complex tasks and hold more nuanced conversations. OpenAI shortly addressed the viral incident, quoting, “What users went through was a bug.” According to OpenAI, that happened when the model tried to respond to a message that did not send well and ended up showing either a message that was empty or a follow-through. After that, OpenAI got in touch to confirm that this was not deliberate on the part of the AI and a fix had been issued to prevent the AI from appearing to initiate conversations in the future. Many initially thought that the post was fake since it’s not too difficult to photoshop a screenshot of this conversation, post it on Reddit, and go viral, fuelling people’s interest and fears in AGI. The Redditor did share an OpenAI link to the conversation, however even this may not be a true verification, people argued. In a post on X, AI developer Benjamin De Kraker demonstrated how this conversation could have been manipulated. You can instruct ChatGPT to respond with a specific question as soon as you send the first message. Then, you delete your message, which pushes ChatGPT to the top of the chat. When you share the link, it appears as if ChatGPT messaged you unprompted. You've seen that post claiming that ChatGPT now pre-emptively messages the user asking them about their day?I'm 99% sure it's fake, and this is how they faked it: pic.twitter.com\/MIKvh5Hp6p— Benjamin De Kraker 🏴‍☠️ (@BenjaminDEKR) September 16, 2024 While there are multiple reasons to believe this didn’t actually happen, it apparently did—but not in the way you might think. OpenAI told Futurism that it had fixed a bug that was responsible for ChatGPT appearing to start conversations with users. The issue would occur whenever the model tried responding to a message that didn’t send as it should, and popped up blank. According to the company, ChatGPT would compensate by either sending a random message, or pulling from its memory. So, what likely happened in this case is the Redditor opened up a new chat, and either triggered a bug that sent a blank message, or accidentally sent a blank message themselves. ChatGPT reached into its memory, and leaned on the fact it knew that the Redditor was starting school to respond with something it thought would be relevant. Although there hasn’t been a comment from the Redditor stating whether or not they have memory enabled for their ChatGPT account, it seems safe to say (at this point) that ChatGPT hasn’t gained consciousness and is randomly reaching out to users. With  AI chatbots transforming online interactions, a recent claim that an AI bot initiated a conversation on its own raises questions about the evolving role of artificial intelligence. Typically, AI bots respond to human inputs within set parameters. However, the idea of a bot taking the initiative challenges the distinction between passive response and active engagement, pushing the boundaries of AI autonomy. But how do you differentiate between a chatbot and a human? If valid, such occurrences may indicate AI models are advancing in ways that could surprise even their creators. While this suggests progress in developing more human-like, proactive agents, it also raises concerns about user consent, privacy, and control. So then, does the AI actually start conversations, or does it wait for you, say 9 times out of 10? Instances of chatbots initiating conversations independently are relatively rare and often depend on specific programming or user settings. One such stellar example is Insomnobot3000. This bot wants to be a companion when all your friends are asleep, or it’s too late to text them. You can only chat with it between 11pm and 5am. “Some nights, it’s just impossible to fall asleep, so I think Casper (the brand) wanted to create something that’s a friend that keeps you up at night,” said Casper VP Lindsay Kaplan. A Glimpse Into the Future? The more we go ahead, the clearer it gets that ChatGPT is not just a tool, but rather an increasingly integrated feature in our lives for remembering, engaging, and possibly predicting our very needs. Whether that sends shivers down your spine or gives you goosebumps, one thing is certain that our conversations with AI are just beginning to start.","excerpt":"ChatGPT initiated a conversation with a user on its own, without being prompted first.","categories":["AI Trends"],"tags":["AI (Artificial Intelligence)","ChatGPT","OpenAI"],"author_name":"Tarunya S","publish_date":"2024-09-19T14:16:34","publication_year":"2024","word_count":1166,"keywords":["Go","ChatGPT","artificial intelligence","OpenAI","AI","chatbots","ML","GPT","Aim","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","ML","ChatGPT","OpenAI","Aim","chatbots","R","Go","GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/wait-did-chatgpt-just-slide-into-my-dms\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":44181,"title":"Now Amazon Is A Front Runner In The Lucrative Data Lake Market","content":"Collecting and processing data to make decisions is a crucial part of businesses. This also means dealing and managing big data with the help of data lakes. It makes it easier to manage large quantities of data allowing companies to retain more information — including unstructured and raw data. It also allows companies to run their analytics and machine learning algorithms on large datasets to better discover patterns. AWS Lake Formation, which is a fully-managed service by Amazon facilitating building, securing and management of data lakes, recently announced that it is now “generally available” — meaning that it is now available for developers to purchase. It comes as a breath of fresh air for those with large datasets to help them deal with data more efficiently. Amazon Lake Formation was announced in November last year at the AWS re:Invent conference in Las Vegas. It brings about automation in a number of steps typically involved in creating a data lake such as collecting, cleaning, deduplicating, cataloguing data and making said data available for analytics in provisioned and configured storage. It also lets users bring data into a data lake from a range of sources. What Would This Mean For Organisations? While the concept of data lakes has been around for a long time, setting it up to store vast amounts of raw data in its native formats has never been easy. With the launch of AWS Lake Formation last year, Amazon aimed at allowing developers to create a secure data lake without much hassle. The usual process requires cumbersome process such as configuring storage, moving data, adding metadata, cleaning the data, setting up the right policies and more. This is a lot of work for the companies and it might take them several months to set up a data lake. Lake Formation makes it easy by handling all these complications required to create a data lake with just a few clicks. It sets up the right tags, cleans up and deduplicates the data automatically. It also provides admins with a list of security policies, governance, and auditing across multiple analytics engines to help secure that data. It also plans to enable engineers to analyse data within those data sets using their choice of AWS analytics and machine learning services such as Amazon Redshift, Amazon Athena, Amazon QuickSight, Amazon SageMaker and more. In a nutshell: It reduces the heavy lifting and automates the manual, time-consuming steps, like provisioning and configuring storage, crawling the data to extract schema and metadata tags It automatically optimises the partitioning of data, transforming data into formats like Apache Parquet and ORC that are ideal for analytics It cleans and deduplicates data using machine learning to improve data consistency and quality It provides a single, centralised place to set up and manage data access policies Some of the clients who have been using it are Panasonic Avionics Corporation, Accenture, Quantiphi, Life360, Amgen and more. Not just the ability to manage security settings for all the different applications in the environment, it gives enhanced control giving them secure access to data. Amazon’s Move Is Timely Reports suggest that the global data lakes market is anticipated to $12.01 billion by 2024. Some of the key players in the space are Microsoft, Google, IBM and more. While Microsoft has its own fully managed solution in Azure Data Lake, Google has a suite of data lake processing and analytics tools in Cloud Datalab, Dataproc, and Dataflow. The data lakes market is drastically growing and there is a major competition between the key players to set up a stronger foot in the market. With Amazon’s Lake Formation, it aims to take a dig at its competitors. The ease that it has brought, AWS claims to be now hosting more data lakes than anyone else, which is growing every day at a fast pace. It has allowed developers to make the most of data as they can now learn and innovate with it, rather than wrestling that data into functioning data lakes.","excerpt":"Collecting and processing data to make decisions is a crucial part of businesses. This also means dealing and managing big data with the help of data lakes. It makes it easier to manage large quantities of data allowing companies to retain more information — including unstructured and raw data. It also allows companies to run […]","categories":[],"tags":["big data in auditing"],"author_name":"Srishti Deoras","publish_date":"2019-08-12T12:19:43","publication_year":"2019","word_count":668,"keywords":["Go","Amazon SageMaker","machine learning","AWS","AI","R","RAG","Aim","big data in auditing","analytics","Azure"],"extracted_tech_keywords":["AI","machine learning","analytics","Aim","Amazon SageMaker","RAG","AWS","Azure","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/now-amazon-is-a-front-runner-in-the-lucrative-data-lake-market\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10023296,"title":"A Good Data Scientist Should Combine Domain-Specific Knowledge With Technical Competence: Prof. Ramanathan, IIITB","content":"For the sixth interview in the weekly series of data science career, we got in touch with Professor Chandrashekar Ramanathan, the Professor & Dean (Academics) & the Faculty-in-charge of Computing at IIITB. While the data science education market has soared over the last few years, professor Ramanathan believes content and structure of University courses should address the changing landscape and industry requirements. Professor Ramanathan is a faculty member at IIITB whose primary focus areas include, data science, software engineering, information convergence, semi-structured databases, and application development. With over ten years of experience in large multinational organisations, he has co-authored several research papers on edge intelligence for industrial IoT, HDSanalytics, OAES, machine-readable ontology for teaching etc. In this interview, he shared insights into the data science education market and how universities can create a relevant industry-ready data science course. Excerpts AIM: Is India an important market for data science education? What’s IIITB’s contributions to this highly promising sector? Professor Ramanathan: The Indian data science market is in its growth phase, and the Indian education market is working towards meeting the potential opportunities and job options in the country. That being said, we have some strides to make to fulfil the promised numbers. The world is always looking up to India for IT solutions, and I believe this trend will not change in the near future. This outlines the immense scope for growth as far as India’s data science education market is concerned. IIITB realised the potential for offering online programmes early on. In 2016, we started offering online courses on Data Science. We also offered a Data Analytics programme titled ‘The Analytics Essentials’. This course started with a small batch of 30 students, and subsequently, we partnered with Upgrad to increase the program’s reach. Here we offered two variations of this programme – Post Graduate Diploma in Data Science, intended for professionals new to this field, and Post Graduate Diploma in Machine Learning & AI, an advanced course designed for professionals working in analytics and related areas. This program is conducted online and does not compromise on academic rigour. Key elements like regular lectures, assignments, lab activities, projects, etc., offered here match the stature of those provided in all our on-campus programs. AIM: What sort of challenges a university faces while instrumenting a data science and analytics course? Professor Ramanathan: ‘Relevance’ is a challenge that universities and institutions face on a real-time basis. Industry demands and requirements are constantly changing, and it is crucial to make sure the course content and structure address the changing professional environment. It becomes even more important since most data science students are professionals with relevant industry experience. Therefore, when teaching working professionals, we have to be cognizant of their work timing and ensure the content is compact and easily comprehensible. AIM: What was the impact of the COVID pandemic on the data science education market? How can governments nudge more students to data science? Professor Ramanathan: The pandemic and the work-from-home setting have provided the bandwidth and opportunity for professionals to pursue online courses. Of course, with several data science courses and opportunities available in the market, it may almost appear like the professional is being compelled to take the course. But I feel the situation has simply allowed more time and clarity to choose to upskill. In terms of government — policies like NEP are instilling significant realisation that we need to encourage multidisciplinary environments. However, to address modern-day problems, we need a well-thought combination of STEM and other disciplines. It will contextualise problems and solutions and help our students stand out. AIM: What are the major obstacles data and analytics education faces in India? Professor Ramanathan: One of the initial obstacles was the availability of data and data-enabled systems. Analytics cannot function without data. IT enablement over the years is helping overcome that particular challenge. Now, we have a continuously updating repository that can be easily used for education. AIM: What’s your advice for aspiring data scientists? What does the future look like for someone who pursues data science now? Professor Ramanathan: Technological expertise augmented by strong domain knowledge is important for an aspiring data scientist. One should have a clear understanding of the rules and practices of the industry before applying technological aspects to it. Be it automotive, BFSI, manufacturing or ecommerce, you can be a good data scientist in the field if you couple domain-specific knowledge with technical competence. Ideal candidates would have a degree or background knowledge of computer science or information technology. Data science is vast and may not suit everyone. Therefore, it is vital to have an aptitude to understand the data, see patterns, analyse from different perspectives and present findings to suit the end-user while also being open to understanding the domain. AIM: What’s your projections for AI and BDA education? Professor Ramanathan: AI and Big Data Analytics (BDA) education will be almost second nature to at least the computer science professionals. In the next five years, they will be required to have basic knowledge and awareness of the subject. This multiplied by the number of graduates we produce each year will probably be the size of the market. AIM: How does an industry partnership add value to the university courses? Professor Ramanathan: Industry partnerships are crucial to educational institutions. The two key components of a data science course are the fundamental conceptual foundation laid by highly qualified academicians and industry stalwarts with on-ground expertise and visibility. Both ensure that the key takeaways are beyond theoretical knowledge and include practical insights and understanding. AIM: Do you think a professional degree will have an edge over numerous online courses and MOOCs available for data science enthusiasts? Professor Ramanathan: We see several students approach us with this question. We always illustrate to our students that if their primary motive is to learn, it does not matter what course format is being pursued. In such a scenario, the individual is left with the burden of finding the correct course with sufficient content and information. On the other hand, a professional degree programme is backed by a trusted institution and brings content curated to suit industry requirements and ensures professional partnerships to benefit students.","excerpt":"For the sixth interview in the weekly series of data science career, we got in touch with Professor Chandrashekar Ramanathan, the Professor & Dean (Academics) & the Faculty-in-charge of Computing at IIITB.  While the data science education market has soared over the last few years, professor Ramanathan believes content and structure of University courses should […]","categories":["AI Highlights"],"tags":["data science education","Data Science Jobs"],"author_name":"Sejuti Das","publish_date":"2021-04-05T10:00:00","publication_year":"2021","word_count":1033,"keywords":["big data","data science","Go","machine learning","Rust","AI","Data Science Jobs","RAG","Aim","analytics","data science education","R"],"extracted_tech_keywords":["AI","machine learning","data science","analytics","Aim","RAG","R","Go","Rust","big data"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/a-good-data-scientist-should-combine-domain-specific-knowledge-with-technical-competence-prof-ramanathan-iiitb\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10085821,"title":"Top Reinforcement Learning Algorithms","content":"One of the biggest driving forces in the recent advancements in AI has been reinforcement learning (RL). A simple definition of reinforcement learning is to train the machine to act as good as possible by giving it feedback for its action, which implies finding a policy that maximises the expected return. RL algorithms adopt slightly different approaches to train the agent for its actions. ​​However, to find the optimal policy and value, most of the RL algorithms follow a similar pattern with either a model-free or a model-based approach along with an on-policy or off-policy approach. All RL algorithms have common terms that need to be understood before diving into the algorithms. Action (A): Moves that the agent makes. State (S): Current situation in the environment. Reward (R): Return sent back from the state to evaluate the last action. Policy: Strategy the agent employs for determining the action based on the current state. Q-value: Often called the action value, it defines the estimation of how good the action taken by the agent is at the state. Notably, the RL approach to AI was taken further by OpenAI when they introduced reinforcement learning with human feedback (RLHF) that led to the birth of ChatGPT. OpenAI powered RLHF with the Proximal Policy Optimisation (PPO) algorithm, which they released in 2017 as their baselines repository. Q-Learning Starting with Q-Learning, which is a model-free and off-policy RL algorithm that is based on the Bellman Equation. The algorithm uses a Q-table, which is a lookup table that stores the agent’s estimated utility or “quality” of taking a certain action in a given state. The agent updates the Q-values to maximise it in the table through trial and error and, eventually, it converges to the optimal policy. Since it is a model-free algorithm, Q-Learning does not need any space to store all the combinations of actions and states. Furthermore, as this algorithm does not follow a single policy like SARSA (discussed below), it can select the best successor action, which is determined by another policy. This also allows one to incorporate additional influences. Q(s,a) ← Q(s, a) + α(r + γ maxₐ’ (s’, a’) — Q(s, a)) Q(s, a) is the current estimate of the utility of taking action ‘a’ in state ‘s’.α is the learning rate, a value between 0 and 1 that determines the relative weight of the current estimate and the new information.r is the reward received after taking action ‘a’ in state ‘s’.γ is the discount factor, a value between 0 and 1 that determines the importance of future rewards.s’ is the next state after taking action ‘a’ in state ‘s’.a’ is the action selected in state s’. SARSA SARSA, or State Action Reward State Action, is similar to Q-Learning but the key difference is that it is an on-policy algorithm, and is often denoted as the ‘on-policy Q-learning’. This implies that through this algorithm, Q-value is derived from the action performed by current policy, which is in contrast to Q-learning algorithm that has no constraint over the next action. The abbreviated name, SARSA, denotes a sequence that the algorithm starts in a state (S), takes action (A), and then the reward is generated (R). This updates the Q-function (value). Q(s,a) ← Q(s, a) + α(r + γ Q(s’, a’) — Q(s, a)) Now, these obtained Q-values are stored in a table and the one with the highest values is chosen by the policy by observing its current state, leading to a new state, and it continues on for next states. Q-values keep getting updated till we find a good policy. But it is constrained by a single policy, thus Q-learning offers more scope for value selection. Both Q-learning and SARSA are tabular methods and, due to vast memory consumption and failure to visit all states and actions while training, do not scale well for large state and action spaces. This is where neural networks come in. Deep Q Network (DQN) DQN is an extension of Q-Learning that leverages neural networks to estimate the Q-value function. This enables it to move beyond the limitations of Q-learning, which cannot estimate value for the unseen states. In 2013, DeepMind even applied DQN to the Atari Game. The neural network is trained based on the Q-learning update equation: Q_θ(s, a) ← Q_θ(s, a) + α((r + γ maxₐ’ Qₜₐᵣ(s’, a’)) — Q_θ(s, a)) The standard Q-learning technique finds the optimal values, which are the highest rewards, and then developers decide the optimal function. DQN allows direct approximation of the optimal value using two essential techniques: Experience Relay: To solve the problem of high correlation and less data efficiency, experience relay allows the sample transitions (moves from one state to another by actions) to be stored. This allows detection of trends, which are then randomly selected from the pool to update the knowledge. Target Networks: These help determine if the output reward is already not the best one. This is achieved by going back to the last updated output and considering Q-values as the target. Deep deterministic policy gradient (DDPG) The methods so far cover discrete action spaces with a fixed number of actions. But when the action space is continuous, tabular and neural network based Q-learning algorithms fall short because finding the action that leads to the highest reward is challenging, if not impossible. DDPG is then considered a breakthrough. DDPG is an actor-critic algorithm that also uses a neural network to approximate both the policy and the value function. It is particularly well-suited for continuous action spaces. DDPG also borrows the ideas of target network and experience replay from DQN. Q_θ(s, a) ← Q_θ(s, a) + α((r + γ Qₜₐᵣ(s’, μₜₐᵣ(s’))) — Q_θ(s,a)) Instead of manually searching the best state–action pair and the Q-value, another neural network is introduced that learns the approximate maximiser and calculates the target. Here, μₜₐᵣ determines the best action that uses a slightly older version of the network. TRPO and PPO Trust Region Policy Optimization (TRPO) and Proximal Policy Optimization (PPO) are both on-policy algorithms that use a neural network to approximate the policy. TRPO uses a trust region method to ensure that the policy update is “conservative” while PPO uses a “clipped” objective function to ensure that the update is not too far from the current policy. Both algorithms are able to handle large, high-dimensional state spaces and continuous action spaces. TRPO achieves great and consistent high performance, but the computation and implementation of the algorithm is increasingly complicated. OpenAI released PPO in 2017 and since then it has become their default as it got rid of computation problems created by constrained optimisation by way of proposing a clipped surrogate objective function.","excerpt":"Reinforcement learning has several algorithms that take different approaches to give rewards to the machine.","categories":["AI Trends"],"tags":["ChatGPT","OpenAI","Reinforcement Learning","RLHF"],"author_name":"Mohit Pandey","publish_date":"2023-01-25T12:16:51","publication_year":"2023","word_count":1114,"keywords":["Go","ChatGPT","TPU","RLHF","Reinforcement Learning","OpenAI","AI","neural network","ML","RAG","R"],"extracted_tech_keywords":["AI","ML","neural network","ChatGPT","OpenAI","RAG","RLHF","TPU","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-reinforcement-learning-algorithms\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10163781,"title":"The One Canva Feature That India Can’t Live Without","content":"Cameron Adams, co-founder of Canva, recently visited India for the first time. In between dealing with lost baggage, losing himself in the vibrant Indian culture, and meeting Canva’s talented team in Delhi, Adams uncovered a surprising design trend: India’s obsession with background removal. Canva offers a wide range of AI-powered design tools, templates, and branding solutions, but one feature stood out in India more than any other—the ability to erase backgrounds. Small business owners use it to refine e-commerce product images, and social media creators rely on it for clean and polished visuals, making background removal an essential tool across industries. Adams was fascinated by how deeply embedded design has become in India’s digital ecosystem. The country’s booming creator economy, rapid e-commerce expansion, and increasing demand for high-quality visuals have made background removal more than just a tool—it’s a necessity. The trend is now shaping the way Canva adapts to one of its fastest-growing markets. As the company continues to evolve, Adams sees India as a creative powerhouse that is redefining digital design at scale. The fact that India has quickly become Canva’s fourth-largest market is a testament to its explosive growth. “India is now our fourth biggest market, and it continues to grow amazingly, which is why it’s so great to have a team here that really helps foster growth,” Adams revealed in an exclusive interview with AIM, alongside discussing the importance of localisation, highlighting the team’s efforts in providing relevant content, features, and language support. Canva, Ab Hindi Mein Canva is planning to launch a Hindi website in the next few months. The introduction of a Hindi-language website will allow millions of users to design and communicate with ease in their native tongue, promoting inclusivity and broader access to visual storytelling in India. “It’s really important for us to bring a truly local product here. The team does an amazing job making sure we have the right content, the right product features, and the right languages that we’re speaking to people.” Adams also pointed to the unique usage patterns in India, noting the popularity of the photo background removal tool and the significant use of Canva for creating resumes and presentations. He expressed excitement about the thriving educator community in India, driven by the Canva for Education program. The co-founder sees immense potential for Canva in India, even suggesting it could become the number one market. “I think, given the current growth trajectory, being number one is just a question of time. India is a huge market.” He cited the entrepreneurial spirit and diverse skill set in India as key drivers of this growth. AI at Canva, By Design Canva has been investing in AI for over seven years, long before the recent surge in AI design startups. Adams explained how AI is now woven into every aspect of the Canva experience. “AI is now integrated at every touch point that you have within Canva, from the home page where you might be suggested different designs…through to every step in designing, whether that’s removing a background or getting suggestions on copy.” He proudly shared that Canva users have engaged with AI features over 13 billion times, demonstrating the value and adoption of these tools. Addressing concerns about AI-generated designs lacking originality, Adams emphasised the importance of balancing AI with human creativity. “We very much view it [AI] as a tool. It’s not the be-all and end-all…We still rely heavily on the rest of the Canva products to produce amazing designs.” He highlighted Canva’s vast library of elements, creator-led content, and templates as crucial components that complement AI capabilities. Canva’s AI strategy involves a three-pronged approach: developing in-house models, partnering with leading AI companies like OpenAI and RunwayML, and empowering the developer ecosystem through the Canva App Store. “Those are the three pillars we think of when we think about our AI strategy, and together, they enable us to bring the best product with the best AI, creating the best experience for our customers,” he said. Built for Collaboration, Not Automation Adams sees a future where humans and AI work together seamlessly to achieve the best outcomes. He discussed the potential of AI design agents but emphasised the importance of human agency. “We strongly believe that humans and technology together produce the best outcomes, not just technology alone…that real collaboration between people and technology is what we’re aiming to create in our product.” While acknowledging the potential for automation, Adams cautioned against a purely technology-driven approach. “I think it’s somewhat simplistic to think of AI as this one-click button that can produce things for you. It needs to change the way we think about problem-solving and delivering solutions and the way we work together as teams.”","excerpt":"Canva is planning to launch a Hindi website in the next few months.","categories":["Global Tech"],"tags":["Canva"],"author_name":"Siddharth Jindal","publish_date":"2025-02-15T12:08:27","publication_year":"2025","word_count":787,"keywords":["API","OpenAI","AI","ML","Git","Canva","automation","Aim","ViT","R","startup"],"extracted_tech_keywords":["AI","ML","OpenAI","Aim","R","Git","API","ViT","automation","startup"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/the-one-canva-feature-that-india-cant-live-without\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10019485,"title":"How To Supercharge Your Machine Learning Experiments with Comet.ml","content":"Table of contentsIntroductionComet’s web-based UIWorkspacesProjectsExperimentsInstallation of comel_ml Python libraryPractical Implementation using KerasPractical implementation using PyTorch Introduction Comet.ml is a Machine Learning experimentation platform which AI researchers and data scientists use to track, compare and explain their ML experiments. It allows ML practitioners to keep track of their databases, history of performed experiments, code modifications and production models. It enables model reproduction, easy maintenance of ML workflow and smooth collaboration throughout the iterative process of ML lifecycle. It also performs model optimization using Bayesian hyperparameter optimization algorithm and thereby reduces the overhead of tuning your model manually. Comet was introduced by a private organization named Comet ML Inc. (also known as CometML). It was founded by Gideon Mendels in 2017 and is headquartered at New York (USA). Some of the leading companies leveraging the Comet platform are: Google, Ancestry, Cern, Uber and Boeing. Following are the SDKs and APIs featured by Comet:SDKs and APIs: Python SDK to track Python-based experiments (An available Python API also enables access to the REST API) Java SDK  to track Java-based experimentsJavaScript SDK to create custom visualizations with Comet PanelsR SDK to track R-based experiments REST API to access your experiment’s data from Comet.mlComet Command-Line Utilities Comet’s web-based UI Comet organizes your code runs based on three major concepts: WorkspaceProjectExperiment A workspace contains projects each os which is a collection of experiments.Let’s have an overview of each of them. Workspaces On creating an account, Comet provides you with a default workspace which comprises your private and public projects. You can create your own workspaces other than the default one. Each workspace has its own set of collaborators (contributors working jointly on a project). Projects in a workspace are automatically shared with all the collaborators in that public. Enabling ‘public’ sharing, you can even share your projects with the outside world. Projects A project is a set of ML experiments. Each project is categorized into one of the following two types: Private project (can be viewed and edited by all the collaborators having appropriate permissions)Public project (can be viewed by anyone but can be edited only by the owner) A project is divided into 5 sections as follows: Experiments: has Project view (the core working area at Comet)Notes: has markdown notes for your projectFiles: a list of all of the files in your project and the experiments that use those filesManage: to control your project’s visibility and create shareable linksArchive: a list of archived experiments for your project Experiments An experiment is a unit of measurable research which represents a single run of your code. Each experiment is given a random ID by default which you can change to make it human-readable. A totally customizable tabular form of an experiment is known as an ‘Experiment Table’. An experiment has a set of ‘Experiment Tabs’ such as Charts Tab, HTML Tab, Hyperparameters Tab, Metrics Tab and so on, each having a peculiar functionality. For instance, Hyperparameters Tab and Metrics Tab store the ML model’s hyperparametes and evaluation metrics logged during your experiment. Within a project, you have a ‘Project Visualizations’ section which enables viewing and comparing performance across different experiments. Besides, ‘Query Builder’ section of a project allows you to choose which experiments should be shown in the Project Visualizations and the Experiment Table. Visit this page to know about workspaces, projects and experiments in detail along with the ways to handle the functionalities provided by each of them. Comet facilitates Automatic Logging for several extensively used Python ML frameworks some of which are listed below. Click on the corresponding link to get a quick tutorial on how to incorporate Comet while implementing that framework. keras, lightgbm, Uber’s ludwig, matplotlib, mlflow, pyspark, pytorch, pytorch-lightning, scikit-learn, shap, tensorflow, tensorflow model analysis, HuggingFace’s transformers. Want to reduce the time it takes to train your neural networks? Try parallelized neural network training! This lets you evaluate multiple hyperparameter options faster. Learn how to implement Parallelized Training in our new report: https:\/\/t.co\/GypH2aMHRr#NeuralNetworks— Comet (@Cometml) January 19, 2021 Working on object detection? You can use Comet to create custom vizzes that make it MUCH easier to debug your models. Add bounding boxes with your specifications easily. Read more: https:\/\/t.co\/xvd1LI3R7B#ObjectDetection #ComputerVision #MachineLearning pic.twitter.com\/wgIakXF7Rf— Comet (@Cometml) January 18, 2021 Installation of comel_ml Python library comet_ml can be installed using pip command as follows: pip install comet_ml Practical Implementation using Keras Import the required libraries from comet_ml import Experiment import keras from keras.datasets import mnist from keras.models import Sequential from keras.layers import Dense, Dropout from keras.callbacks import EarlyStopping Create an experiment with your API key experiment = Experiment( project_name='my_project',  \/\/name of your project \/\/disable automatic logging of hyperparameters auto_param_logging=False, \/\/enable automatic histogram logging for biases and weights auto_histogram_weight_logging=True, \/\/enable  automatic histogram logging of gradients auto_histogram_gradient_logging=True, \/\/enable automatic histogram logging of activations auto_histogram_activation_logging=True, ) Click here to learn about the properties and methods associated with an Experiment object. Initialize model’s parameters batch_size = 128 num_classes = 10  \/\/number of output classes epochs = 20 num_nodes = 64   \/\/number of nodes in hidden layer optimizer = 'adam' activation = 'relu' Get to know about the Adam optimizer and ReLU activation function. Load the MNIST digit classification dataset using load_data() function and split it into training set and test set (x_train, y_train), (x_test, y_test) = mnist.load_data() Reshape the 1D train and test set x_train = x_train.reshape(60000, 784) x_test = x_test.reshape(10000, 784) Convert the data types to real x_train = x_train.astype('float32') x_test = x_test.astype('float32') Perform normalization x_train \/= 255 x_test \/= 255 Print the number of samples in training set and test set print(x_train.shape[0], 'Training set samples') print(x_test.shape[0], 'Test set samples') Convert class vectors to binary class matrices y_train = keras.utils.to_categorical(y_train, num_classes) y_test = keras.utils.to_categorical(y_test, num_classes) Define a dictionary for parameters to be logged params={'batch_size':batch_size, 'epochs':epochs, 'layer1_type':'Dense', 'layer1_num_nodes':num_nodes, 'layer1_activation':activation, 'optimizer':optimizer } Instantiate the sequential model model = Sequential() Add dense layers in the network model.add(Dense(num_nodes, activation='relu', input_shape=(784,))) model.add(Dense(num_classes, activation='softmax')) Print model.summary() to preserve automatically in `Output` tab print(model.summary()) Compile the model model.compile(loss='categorical_crossentropy', optimizer=optimizer, metrics=['accuracy']) Log metrics with the prefix ‘train_’ with experiment.train(): history = model.fit(x_train, y_train, batch_size=batch_size, epochs=epochs, verbose=1, validation_data=(x_test, y_test), callbacks=[EarlyStopping(monitor='val_loss', min_delta=1e-4,patience=3, verbose=1, mode='auto')]) Experiment.train() marks beginning and end of the train phase. It provides a namespace for logging the training parameters. Log metrics with the prefix ‘test_’ with experiment.test(): loss, accuracy = model.evaluate(x_test, y_test) metrics = { 'loss':loss, 'accuracy':accuracy } Experiment.test() marks beginning and end of the test phase. It provides a namespace for logging the testing metrics. Log the metrics as key:value pairs in a dictionary experiment.log_metrics(metrics) Log a dictionary-like object of parameters experiment.log_parameters(params) Create and log a hash of your data experiment.log_dataset_hash(x_train) Source: https:\/\/www.comet.ml\/docs\/python-sdk\/keras Practical implementation using PyTorch Import the required libraries from comet_ml import Experiment import torch import torch.nn as nn import torchvision.datasets as dsets import torchvision.transforms as transforms from torch.autograd import Variable Define the hyperparameters hyper_params = { \"sequence_length\": 28, \"input_size\": 28, \/\/number of input layer neurons \"hidden_size\": 128,  \/\/number of hidden layer neurons \"num_layers\": 2,  \/\/number of hidden layers \"num_classes\": 10,  \/\/number of output classes \"batch_size\": 100, \"num_epochs\": 2, \"learning_rate\": 0.01 } Instantiate an Experiment object experiment = Experiment(project_name=\"my_project\") Log the hyperparameters experiment.log_parameters(hyper_params) Load the MNIST dataset train_dataset = dsets.MNIST(root='.\/data\/', train=True, transform=transforms.ToTensor(), download=True) test_dataset = dsets.MNIST(root='.\/data\/', train=False, transform=transforms.ToTensor()) Define the model training and testing pipelines using DataLoader utility train_loader = torch.utils.data.DataLoader(dataset=train_dataset,                           batch_size=hyper_params['batch_size'], shuffle=True) test_loader = torch.utils.data.DataLoader(dataset=test_dataset,                            batch_size=hyper_params['batch_size'], shuffle=False) Define many-to-one RNN model class RNN(nn.Module): def __init__(self, input_size, hidden_size, num_layers, num_classes): super(RNN, self).__init__() self.hidden_size = hidden_size self.num_layers = num_layers self.lstm = nn.LSTM(input_size, hidden_size, num_layers, batch_first=True) self.fc = nn.Linear(hidden_size, num_classes) Forward propagation def forward(self, x): # Set initial states h0 = Variable(torch.zeros(self.num_layers, x.size(0), self.hidden_size)) c0 = Variable(torch.zeros(self.num_layers, x.size(0), self.hidden_size)) #Forward propagate RNN out, _ = self.lstm(x, (h0, c0)) #Decode hidden state of last time step out = self.fc(out[:, -1, :]) return out Call RNN() rnn = RNN(hyper_params['input_size'], hyper_params['hidden_size'], hyper_params['num_layers'], hyper_params['num_classes']) Define loss function and optimizer criterion = nn.CrossEntropyLoss() optimizer = torch.optim.Adam(rnn.parameters(), lr=hyper_params['learning_rate']) Get to know about Cross Entropy Loss. Train the Model with experiment.train(): step = 0 for epoch in range(hyper_params['num_epochs']): correct = 0 total = 0 for i, (images, labels) in enumerate(train_loader): images = Variable(images.view(-1,hyper_params ['sequence_length'], hyper_params['input_size'])) labels = Variable(labels) #forward-backward optimizations and optimization optimizer.zero_grad() outputs = rnn(images) loss = criterion(outputs, labels) loss.backward() optimizer.step() #Compute train accuracy _, predicted = torch.max(outputs.data, 1) batch_total = labels.size(0) total += batch_total batch_correct = (predicted == labels.data).sum() correct += batch_correct #Log batch_accuracy to Comet.ml step += 1 #each step represents one batch experiment.log_metric(\"batch_accuracy\", batch_correct \/ batch_total, step=step) if (i + 1) % 100 == 0: print('Epoch [%d\/%d], Step [%d\/%d], Loss: %.4f' % (epoch + 1, hyper_params['num_epochs'], i + 1, len(train_dataset) \/\/ hyper_params['batch_size'], loss.item())) Log epoch accuracy to Comet.ml; step is each epoch experiment.log_metric(\"batch_accuracy\", correct \/ total, step=epoch) Test the model with experiment.test(): correct = 0 total = 0 for images, labels in test_loader: images = Variable(images.view(-1, hyper_params['sequence_length'], hyper_params['input_size'])) outputs = rnn(images) _, predicted = torch.max(outputs.data, 1) total += labels.size(0) correct += (predicted == labels).sum() Log accuracy experiment.log_metric(\"accuracy\", correct \/ total) Print the logged accuracy print('Test Accuracy: %d %%' % (100 * correct \/ total)) Source: https:\/\/www.comet.ml\/docs\/python-sdk\/pytorch Refer to the following web links to get greater insights of Comet.ml: Short guide to Comet.ml Official documentationGitHub Twitter","excerpt":"Introduction Comet.ml is a Machine Learning experimentation platform which AI researchers and data scientists use to track, compare and explain their ML experiments. It allows ML practitioners to keep track of their databases, history of performed experiments, code modifications and production models. It enables model reproduction, easy maintenance of ML workflow and smooth collaboration throughout […]","categories":["Deep Tech"],"tags":["Machine Learning"],"author_name":"Nikita Shiledarbaxi","publish_date":"2021-02-02T13:00:00","publication_year":"2021","word_count":1517,"keywords":["scikit-learn","machine learning","Keras","AI","neural network","PyTorch","ML","Machine Learning","Comet ML","MLflow","TensorFlow"],"extracted_tech_keywords":["AI","machine learning","ML","neural network","MLflow","Comet ML","TensorFlow","PyTorch","Keras","scikit-learn"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/how-to-supercharge-your-machine-learning-experiments-with-comet-ml\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10009038,"title":"NVIDIA Rethinks AI: Next Gen Data Centers, Supercomputers For Healthcare &#038; More at GTC Day 1","content":"“If the last 20 years were amazing, the next 20 will seem nothing short of science fiction.”Jensen Huang, CEO, NVIDIA NVIDIA’s CEO, Jensen Huang, kicked off the GPU Technology Conference in style with his kitchen keynote underlining the various ways in which his team has pushed the envelope of AI as a domain. This five-day conference flaunts 500+ talks from industry experts along with hot announcements from NVIDIA. In this article, we bring you the highlights from the day 1 of GTC. NVIDIA By The Numbers NVIDIA has shipped over 1 billion CUDA compatible GPUs. 6,500 startups are building on NVIDIA. CUDA SDK has been downloaded 20 million times, and 6 million times this year alone. 1,800 applications are now CUDA-accelerated. There are over 2 million NVIDIA developers now. NVIDIA For Next Gen Data Centers Data centre’s infrastructure can consume up to 30% of its CPU cores. And as traffic within a data centre and microservices increase, this load will increase dramatically. NVIDIA believes that the next-generation data centres require a new kind of processor. They have built DPUs for this purpose. These DPUs consist of accelerators for networking, storage, security and programmable Arm CPUs to offload the hypervisor. As the latest addition to the DPU family, Huang announced BlueField 2 DPU. Powered by Mellanox technology, the new NVIDIA BlueField 2 DPU is a programmable processor with powerful Arm cores and acceleration engines. Huang revealed that the BlueField-2 is sampling now, BlueField-3 is finishing, and BlueField-4 is in high gear. “We are going to bring a ton of technology to networking. In just a couple of years, we’ll span nearly 1,000 times in compute throughput,” said Huang, talking about the DPU. Along with a new DPU, NVIDIA also announced DOCA, its programmable data-centre-infrastructure-on-a-chip architecture, which enables developers to write infrastructure apps for software-defined networking, cybersecurity, telemetry and more. All Things ARM CEO Jensen with Ampere GPU “Arm is the most popular CPU in the world.” As expected day one consisted of key announcements and panel discussions concerned with ARM. NVIDIA pocketed ARM for a staggering $40 billion and have teased their ambitious objectives last month. “Arm is the most popular CPU in the world,” Huang said. “Together, we will offer NVIDIA accelerated and AI computing technologies to the Arm ecosystem.” Yesterday, Huang announced a major initiative to advance the Arm platform: NVIDIA and Arm partners will work together to create platforms for HPC, cloud, edge and PC. It will also support Arm partners with GPU, networking, storage and security technologies to create complete accelerated platforms. NVIDIA AI and NVIDIA RTX engines will be ported to Arm. By combining in house technologies, NVIDIA wants to make Arm platforms a leader in edge at accelerated and AI computing. A Supercomputer For Healthcare Huang spoke about NVIDIA’s effort to build the U.K.’s fastest supercomputer — Cambridge-1. With this, NVIDIA plans to bring state-of-the-art computing infrastructure to “an epicentre of healthcare research.” On completion, Cambridge-1 will boast 400 petaflops of AI performance, which shall place it in the upper echelons of supercomputers. Cambridge-1 will host NVIDIA’s U.K. AI and healthcare collaborations with academia, industry and startups. NVIDIA’s ambitious healthcare pursuits are backed up by pharma biggies like AstraZeneca, GSK; King’s College London; the Guy’s and St Thomas’ NHS Foundation Trust and startup Oxford Nanopore. AI At The Edge NVIDIA has invested heavily in building integrated software and hardware solutions for the edge. At GTC, the company has announced few of their developments that will democratise AI like never before. NVIDIA’s EGX platform is expanding to combine the Ampere architecture GPU and BlueField-2 DPU on a single PCIe card. The NVIDIA EGX AI platform is built to set up a state-of-the-art edge-AI server and can be deployed at factories, perform automatic checkout at retail or help nurses monitor patients. NVIDIA Fleet Command, a new service was announced using which edge computing can be leveraged across IoT devices, which combines the security and real-time processing with the remote management and ease of software-as-a-service. In an aggressive push to democratise robotics, NVIDIA is launching Jetson Nano 2GB, the latest addition to the Jetson family. Jetson is an Arm-based System on Chip designed for robotics. Thanks to the sensor processors, the CUDA GPU and Tensor Cores, and, most importantly, the richness of AI software that runs on it. According to Huang, AI software is a big breakthrough that will make robots smarter and more adaptable. But it’s the NVIDIA Jetson AI computer that will democratise robotics. Building Chatbots And Recommenders Gets Better NVIDIA is bringing its much talked about services Jarvis and Merlin for all. Huang announced that NVIDIA Jarvis for conversational AI services and NVIDIA Merlin for recommender systems had entered open beta. This will now allow companies to explore larger deep learning models and develop more nuanced and intelligent systems. Businesses can leverage Conversational AI services built on Jarvis for building chatbots and recommender systems built on Merlin for eCommerce and other applications. For instance, a chatbot that serves in real-time should be able to mount model computations in under 300 milliseconds. Jarvis ability to handle multiple data streams in real-time enables the delivery of improved services. It enables more natural interactions through sensor fusion — the integration of video cameras and microphones. Stay tuned to Analytics India Magazine for more updates from GTC Day 2","excerpt":"“If the last 20 years were amazing, the next 20 will seem nothing short of science fiction.” Jensen Huang, CEO, NVIDIA NVIDIA’s CEO, Jensen Huang, kicked off the GPU Technology Conference in style with his kitchen keynote underlining the various ways in which his team has pushed the envelope of AI as a domain. This […]","categories":["AI News"],"tags":["cuda","HPC Data Management Software","hpc data management system","Jensen Huang"],"author_name":"Ram Sagar","publish_date":"2020-10-06T19:04:52","publication_year":"2020","word_count":890,"keywords":["CUDA","Go","hpc data management system","cuda","AI","chatbots","HPC Data Management Software","Jensen Huang","RAG","microservices","deep learning","analytics","edge computing","R"],"extracted_tech_keywords":["AI","deep learning","analytics","RAG","chatbots","microservices","edge computing","CUDA","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/nvidia-gtc-2020-top-announcements\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":61650,"title":"Tracking Corona Using Tableau","content":"(A) Introduction The world is in a disastrous situation now. Since the outbreak of coronavirus COVID-19 in Wuhan, China, it took almost no time to affect the entire world. More than 210 countries and territories around the world got affected. More than 1,800,477 COVID-19 cases are identified so far and more than 100,000 deaths have been observed across the globe. You might have observed that several dashboards are designed to keep the world updated about the COVID-19 spread. Till date, numerous visualizations has been posted in LinkedIn, some of which are beautiful animations that capture the spread of the virus based on the total number of cumulative confirmed cases and deaths. Undoubtedly, you may have found these graphics very intriguing. And, that’s what brought me here. In this article, I will show you how to create some beautiful animated visualizations using Tableau. While doing so, I will particularly consider the case of the COVID-19 outbreak. This will also enable you to create further visualizations that will reflect the current situation. While there are several ways to visualize this data using animated graphics, we will particularly focus on designing two different graphics to visualize the spread of this virus – (1) using a racing bar plot and (2) using an animated map. This article contains several recorded videos to guide you through the process of creating these visualizations. (B) Requirement You will need Tableau 2020.1 (or above) to create the racing bar plot (section D). You may download this version of Tableau from the Tableau website and buy a license for yourself.  You may also download this version for free with a 12-day trial period. To create the animated map (section E), the free public version of Tableau is enough. (C) Data Set Here is a link to the dataset which we would be using to create the visuals. The dataset contains daily updates on the number of confirmed COVID-19 cases, number of recovered cases and number of deaths due to COVID-19 for each country. You may download this dataset to follow along. Just right-click on the page and save this data as a CSV file in your system. (D) Racing Bar Plot to Visualize the Spread Here we go. This is the first animated data visualization exercise I will be taking you through. we will focus on creating a ‘Racing Bar Plot’ to visualize the number of cumulative confirmed COVID-19 cases for all the countries in chronological order. This section is divided into four parts. In the first part, I will demonstrate how to create an animated bar plot using an older version of Tableau. In the second video, I will demonstrate how to extend this animated graph to a racing bar plot using a newer version of Tableau. In the third video, I will teach you some of the ways to make the graph look beautiful and presentable. Finally, in the fourth and the last video I will show you how to package this visualization in the form of a nice and presentable dashboard and publish it in Tableau Public and get a sharable link of the dashboard. (D.1) Animated Bar Plot (in Older Version of Tableau) The first video in this series focuses on creating an animated bar plot using an older version in Tableau (version below 2020.1). This version, however, doesn’t have the necessary feature to create a racing bar chart. We will use the variable ‘Date’ to create the animation, i.e., we would like the graph to update itself with the change in the date. However, the position of the countries in the plot will remain unchanged and will not get sorted according to the number of cumulative confirmed cases. Here is a short summary of what you will learn from this video: Connect tableau to a CSV file. (00:22)Open a new worksheet. (00:38)Create a bar plot. (00:54)Add animation. (01:08)Play animation (play\/stop\/playback speed\/forward\/rewind). (01:30)Edit chart title. (02:45) https:\/\/youtu.be\/xASZfJr4ayw (D.2) Racing Bar Plot In this section, we focus on creating a racing bar plot. In this animated graphic, each bar will represent a country and the bars will race with each other to reach the top as the lengths of the bars increase. In order to create this animation, you will need a Tableau version which is 2020.1 or higher. This is a continuation of the first video of the series. Here is a short summary of what you will learn in this second video of this ‘racing bar plot’ series: Create a rank function (using calculated field). (00:27)Necessary arrangements to create the ‘racing bar plot’ visualization. (00:50)Enable animation. (01:45)Play animation. (01:56) https:\/\/youtu.be\/xtAPkqGUyKI (D.3) Formatting the Graph In section D.1 and D.2 we have discussed the technique to create a racing bar plot using Tableau. In this section, we will focus on enhancing the looks and feel of the graph, make it more presentable and give it a professional look. Here’s what you will learn from the third video of this series: Colour coding the bars. (00:18)Edit the axes titles. (00:42)Format page colour. (01:06)Remove gridlines. (01:24)Edit axes colour. (01:37)Adding labels to the bars. (02:13)Edit labels. (02:19)Edit chart title. (02:54) https:\/\/youtu.be\/Tfa3JE7SdsY (D.4) Creating a Dashboard We are almost there. Now, all we need to do is to package this properly and create a sharable dashboard. A visual dashboard may contain multiple visualizations. However, in this particular one, we will particularly include this one visualization and put it in a very presentable form. Finally, we will publish the dashboard in tableau public and get a sharable link of the dashboard. So, from this final video of this ‘racing bar plot’ series, you will learn how to: Open a new dashboard. (00:18)Choose a dashboard size. (00:28)Add plots to the dashboard. (00:36)Adjust plots to the dashboard. (00:44)Insert an image in the dashboard. (01:23)Insert and edit a text in the dashboard. (02:13)Create an extract for the data source. (03:16)Save a dashboard to tableau public. (03:42) https:\/\/youtu.be\/WBCan5d-P5Y (E) Animated World Map This section is relatively simpler. To create this visualization, you do not need a version of tableau which is 2020.1 or higher. The technique used to create this animated visualization is similar to the one displayed in the video in section C.1. However, there are some extra things that you will learn from this section especially, visualization using maps, formatting a map and size encoding. In this section, we will aim to visualize the total number of cumulative confirmed COVID-19 cases on a world map in chronological order. The tutorial is divided into two parts. In the first part, I will demonstrate how to create an animated world map and how to format the graphic. In the second part, I will discuss how we can package this visualization in the form of a nice dashboard and publish the graphic in tableau public and get a sharable link. (E.1) Using Animated Maps to Visualize the Spread This is the first part of the two-part tutorial on how to create an animated world map to visualize the COVID-19 spread. In this first video of this series, you will learn how to use maps for visualization and how to use animation to visualize the spread in chronological order. This video will help you learn and understand the following things: Using map to visualize geographical data. (00:30)Formatting the map. (00:40)Edit size (size encoding). (00:43)Add colour, border and transparency. (00:49)Create animation. (01:00)Format date. (01:05)Edit chart title. (01:13)Format chart title. (01:25)Play the animated plot. (01:32) https:\/\/youtu.be\/3iCqmkw8cMQ (E.2) Creating a Dashboard In this final section, we will create a nice visual dashboard using the graphic we have created in the previous section. Finally, we will save and publish it in Tableau Public. From this final video of the ‘animated map’ series, you will learn how to: Open a new dashboard. (00:25)Choose a dashboard size. (00:35)Add plots to the dashboard. (00:41)Adjust plots to the dashboard. (00:48)Insert image in the dashboard. (01:15)Insert and edit a text in the dashboard. (01:59)Save a dashboard to tableau public. (02:53)Play the animation in Tableau public. (03:37)Interact with the plot. (04:14) https:\/\/youtu.be\/qHrZ16e-40Y (F) End Notes I hope you have enjoyed creating these graphics. If this was your first hands-on exercise on creating animated visualizations using tableau, then congratulations on that. Write to me about your experiences in the comment section. I hope you would continue to have fun in creating such visualizations and it is probably not difficult to identify the other applications of such animated visualization in some other domains. Also, remember that nothing is better than adding narration to such beautiful animated visualizations. A narration can enhance the beauty of these visualizations subsequently. Here is a wonderful example of story-telling using animated graphics. The person in this video is Professor Hans Rosling, a Swedish Professor of Global Health, and I am sure this video will make you fall in love with data visualization even more. To the readers – If you would like me to evaluate your dashboards and give my personal feedback please post the links to your dashboards in the comment box. I will also be very excited to see your improvisations on these existing graphics. I also welcome you to post your creative ideas and anything new which you have created and would like to share in this forum in order to enlighten each other.","excerpt":"(A) Introduction The world is in a disastrous situation now. Since the outbreak of coronavirus COVID-19 in Wuhan, China, it took almost no time to affect the entire world. More than 210 countries and territories around the world got affected. More than 1,800,477 COVID-19 cases are identified so far and more than 100,000 deaths have […]","categories":["Deep Tech"],"tags":["covid-19 data visualization"],"author_name":"Gourab Nath","publish_date":"2020-04-15T10:00:00","publication_year":"2020","word_count":1544,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","Aim","covid-19 data visualization","R"],"extracted_tech_keywords":["AI","Aim","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/tracking-corona-using-tableau\/","complexity_score":3,"technical_depth":6,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10123629,"title":"OpenAI Inches Closer to Trillion-Dollar Status with Revenue Doubling to $3.4 Billion","content":"OpenAI has more than doubled its annualised revenue to $3.4 billion in the past six months, OpenAI CEO Sam Altman informed staff, reported The Information. The majority of OpenAI’s revenue—around $3.2 billion on an annualized basis—comes from subscriptions to its chatbots and fees from developers accessing its models via an application programming interface. Microsoft takes a cut from some of OpenAI’s AI model sales since they run on Microsoft’s cloud. Additionally, OpenAI receives a share from Microsoft’s sales of OpenAI models to Microsoft’s Azure cloud customers. This share now amounts to about $200 million on an annualised basis, or roughly 20% of the revenue Microsoft generates from that business, according to Altman. The startup’s financial losses were not disclosed but have been substantial in the past. Chinese investor and serial entrepreneur Kai Fu Lee is bullish about OpenAI becoming a trillion-dollar company in the next two-three years. “OpenAI will likely be a trillion-dollar company in the not-too-distant future,” said Lee, adding that GPT-4 and GPT-4 Turbo are unbelievably good and a great balance for performance and cost. Apple’s announcement this week that it will integrate ChatGPT into its products could further boost OpenAI, though the terms of the deal, including whether Apple is paying OpenAI, remain unclear. Apple is not paying OpenAI. Instead, Apple believes that promoting OpenAI’s brand and technology on hundreds of millions of its devices is of equal or greater value than monetary payments, according to a Bloomberg report. OpenAI plans to launch several new products, including a search engine, its video-generating model Sora, and AI-powered software capable of taking actions on users’ computers, which could further enhance revenue. An OpenAI spokeswoman described the financial details as ‘inaccurate,’ without further explanation, according to The Information. OpenAI’s revenue rate positions it ahead of its competitors. For example, last fall, Anthropic reported generating revenue at a $100 million annualised rate, with plans to reach over $850 million by the end of 2024. Cohere, a Canadian OpenAI rival, reported generating just $22 million in annualized revenue in April. OpenAI was recently valued at about $86 billion during an employee share sale. The latest revenue figures suggest OpenAI is valued at approximately 25 times forward revenue, which is on the lower end of recent fundraising valuations for private AI startups.","excerpt":"Apple is not paying OpenAI.","categories":["AI News"],"tags":["OpenAI"],"author_name":"Siddharth Jindal","publish_date":"2024-06-14T11:49:12","publication_year":"2024","word_count":379,"keywords":["Anthropic","ChatGPT","Go","OpenAI","AI","chatbots","R","GPT","Azure","startup"],"extracted_tech_keywords":["AI","ChatGPT","OpenAI","Anthropic","chatbots","Azure","R","Go","GPT","startup"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/openais-annualised-revenue-doubles-to-3-4-billion-regardless-of-apple-partnership\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10118542,"title":"KaleidEO Achieves Milestone in Earth Observation Payload Development","content":"KaleidEO Space Systems, a Bengaluru-based startup and subsidiary of SatSure, has successfully conducted aerial testing of its high-performance optical and multi-spectral earth observation payload. This achievement makes KaleidEO one of the first private Indian companies to design and develop a high-resolution optical EO payload. The prototype payload was tested aboard an aircraft over the UK and Austria in collaboration with the Global Assistant & Logistic Group. The test aimed to assess the payload’s functionality and stability in a real-world environment. The payload captured images in five bands (red, blue, green, near-infrared, and red edge) at a spatial resolution of 16 cm. KaleidEO’s satellite version of this payload is expected to capture images globally at a 1-meter resolution and 65 km swath, which could be disruptive due to the scale and quality of data collection for various applications, including agriculture, urban planning, critical infrastructure asset monitoring, and the government’s strategic planning needs. The payload employs motion compensation and pixel shift methodology for super-resolving captured images, similar to the technique popularised by smartphones but applied to imaging moving objects at speeds of 7 km\/s from orbit. Akash Yalgach, Co-founder and Chief Technology Officer at KaleidEO, expressed his excitement about the successful test. He stated that it has given the team confidence for their planned 4-satellite fleet mission in 2026. He hopes their efforts will bolster Earth-imaging capabilities and open doors for the era of space-based innovation from India. KaleidEO’s founding team comprised young ex-ISRO scientists and was set up by deep tech firm SatSure, which closed a Series A round of $15 million in equity capital last year. SatSure counts HDFC Bank, ICICI Bank, ADB Ventures, and Transunion among its strategic investors and would be an anchor-tenant customer for KaleidEO’s fleet of hi-resolution satellite imagery. Prateep Basu, Founder & CEO of SatSure & KaleidEO, commented on the feat, stating that high-quality, affordable satellite imagery is still a myth, but KaleidEO aims to break barriers by democratising access to such data for users in India and other developing countries through cutting-edge hardware innovation.","excerpt":"The team has successfully conducted aerial testing of its high-performance optical and multi-spectral earth observation payload.","categories":["Deep Tech"],"tags":[],"author_name":"Shyam Nandan Upadhyay","publish_date":"2024-04-18T16:17:28","publication_year":"2024","word_count":340,"keywords":["Go","API","programming_languages:R","AI","innovation","programming_languages:Go","Aim","R","startup"],"extracted_tech_keywords":["AI","Aim","R","Go","API","innovation","startup","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/kaleideo-achieves-milestone-in-earth-observation-payload-development\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":17450,"title":"Country To Lead Artificial Intelligence Will Rule World, Says Russian President Vladimir Putin","content":"Russian President Vladimir Putin has the answer to who will rule the world in the times to come. According to him, the country, which reaches a breakthrough in developing artificial intelligence will dominate the world. Putin, speaking to a group of Russian students Friday, called artificial intelligence “not only Russia’s future” but “the future of the whole of mankind.” “The one who becomes the leader in this sphere will be the ruler of the world,” he said. “There are colossal opportunities and threats that are difficult to predict now.” The Russian leader predicted that future wars will be fought by drones, and “when one party’s drones are destroyed by drones of another, it will have no other choice but to surrender.” On the occasion, he also discussed space, medicine and the capacity of the human brain. “The movement of the eyes can be used to operate various systems, and also there are possibilities to analyse human behaviour in extreme situations, including in space,” Putin added. The Russian president’s 45-minute open session was reportedly attended by students and teachers from 16,000 schools. Top officials, both in government and in the private sector, have long been willing to discuss the impact of artificial intelligence and other technological advances. Russia and China are seen as key competitors in the digital space and have been working on how to apply technologies, such as artificial intelligence and quantum computing, to their war-fighting doctrines. As per recent reports, China will launch a series of artificial intelligence (AI) projects and increase efforts to cultivate tech talent as part of a soon to announced national AI plan, the China Daily said on Friday. The country is focusing on AI as it is seen as a tool to boost productivity and empower employees, the paper said. China will roll out a slew of AI research and development projects, allocate more resources to nurturing talent and increase the use of AI in education, healthcare and security among other things, the report said. However, many experts say the U.S. still maintains an advantage over Russia in artificial intelligence and quantum computing. Still, as Russia, China and other countries seek additional breakthroughs in how to apply such technology, the stakes are high.","excerpt":"Russian President Vladimir Putin has the answer to who will rule the world in the times to come. According to him, the country, which reaches a breakthrough in developing artificial intelligence will dominate the world. Putin, speaking to a group of Russian students Friday, called artificial intelligence “not only Russia’s future” but “the future of […]","categories":["AI News"],"tags":["AI (Artificial Intelligence)","China","Russia"],"author_name":"Priya Singh","publish_date":"2017-09-04T11:08:07","publication_year":"2017","word_count":370,"keywords":["Go","artificial intelligence","programming_languages:R","AI","Russia","programming_languages:Go","Git","ViT","AI research","AI (Artificial Intelligence)","R","emerging_tech:quantum computing","China"],"extracted_tech_keywords":["AI","artificial intelligence","R","Go","Git","ViT","AI research","programming_languages:R","programming_languages:Go","emerging_tech:quantum computing"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/country-lead-artificial-intelligence-will-rule-world-says-russian-president-vladimir-putin\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10041418,"title":"Machine Learning Through The Lens of Econometrics","content":"“While we can predict house prices with accuracy, we cannot use such ML models to answer questions like whether one needs more dining rooms.” Artificial Intelligence has been a force of nature in many fields. From augmenting advancements in health and education to bridging gaps through speech recognition and translation AI—machine intelligence is becoming more vital to us every day. Sendhil Mullainathan, a professor at the University of Chicago Booth School of Business, and Jann Spiess, an assistant professor at the Stanford Graduate School of Business, observed how machine learning, specifically supervised machine learning, was more empirical than it was procedural. For instance, face recognition algorithms do not use rigid rules to scan certain pixel recognitions. Au contraire, these algorithms utilise large datasets of photographs to predict how a face looks. This means that the machine would use the images to estimate a function f(x) that predicts the presence (y) of a face from pixels (x). Another discipline that heavily relies on such approaches is econometrics. Econometrics is the application of statistical procedures in economic data to provide empirical analysis on economic relationships. With machine learning being used on data for uses like forecasting, can empirical economists employ ML tools in their work? New Methods For New Data Today, we see a considerable change in what constitutes the data individuals can work within. Machine learning enables statisticians and analysts to work with data considered too high dimensional for standard estimation methods, such as online posts and reviews, images, and language information. Statisticians could barely look at such data types for processes such as regression. In a 2016 study, however, researchers used images from Google Street View to measure block-level income in New York City and Boston. Moreover, a 2013 research developed a model to use online posts to predict the outcome of hygiene inspections. Thus, we see how machine learning can augment how we research today. Let’s look at this in further detail. Traditional estimation methods, like ordinary least squares (OLS), are already used to make predictions. So how does ML fit into this? To see this, we return to Sendhil Mullainathan and Jann Spiess’ work—which was written in 2017, when the former taught and the latter was a PhD candidate at Harvard University. The paper took an example, predicting house prices, for which they selected ten thousand owner-occupied houses (chosen at random) from the 2011 American Housing Survey’s metropolitan sample. They included 150 variables on the house and its location, such as the number of bedrooms. They used multiple tools (OLS and ML) to predict log unit values on a separate set of 41,808 housing units—for out-of-sample testing. Applying OLS to this will require making specifically curated choices on which variables to include in the regression. Adding every interaction between variables (e.g. between base area and the number of bedrooms) is not feasible because that would consist of more regressors than data points. ML, however, searches for such interactions automatically. For instance, in regression trees, the prediction function would take the form of a tree that splits at each node, representing one variable. Such methods would allow researchers to build an interactive function class. One problem here is that a tree with these many interactions would result in an overfit—i.e. It would not be flexible enough to work with other data sets. This problem can be solved by something called regularisation. In the case of a regression tree, a tree of a certain depth will need to be chosen based on the tradeoff between a worse in-sample fit and a lower overfit. This level of regularisation will be selected by empirically tuning the ML algorithm—by creating an out-of-sample experiment within the original sample. Thus, picking the ML-based prediction function involves two steps: selecting the best loss-minimising function and finding the optimal level of complexity by empirically tuning it. Trees and their depts are just one such example. Mullainathan and Speiss stated that the technique would work with other ML tools such as neural networks. For their data, they tested this on various other ML methods, including forests and LASSO, and found them to outperform OLS (trees tuned by depths, however, were not more effective than the traditional OLS). The best prediction performance was seen by an Ensemble that ran several separate algorithms (the paper ran LASSO, tree and forest). Thus, econometrics can guide design choices to help improve prediction quality. There are, of course, a few problems associated with the use of ML here. The first is the lack of standard errors on the coefficients in ML approaches. Let’s see how this can be a problem: The Mullainathan-Spiess study randomly divided the sample of housing units into ten equal partitions. After this, they re-estimated the LASSO predictor (with the regulariser kept fixed). The results displayed a massive problem: a variable used by the LASSO model in one partition may be unused in another. There were very few stable patterns throughout the partitions. This does not affect the prediction accuracy too much, but it does not help decipher whether two variables are highly correlated. In traditional estimation methods, such correlations are reflected as significant standard errors. Due to this, while we can predict house prices with accuracy, we cannot use such ML models to answer questions like whether a variable, e.g. number of dining rooms, is unimportant in this research just because the LASSO regression did not use it. Regularisation also leads to problems: it allows the choice of less complex but potentially wrong models. It could also bring up concerns of omitted variable biases. Finally, it is essential to understand the type of problems ML solves. ML revolves around predicting a function y from variable x. However, many economic applications work around estimating parameter β that might underlie the relationship between x and y. ML algorithms are not built for this purpose. The danger here is taking an algorithm built for y=ŷ and presuming that its β̂ value would have the properties associated with estimation output. Still, ML does improve prediction—so one might benefit from it by looking for problems with more significant consequences (i.e. situations where improved predictions have immense applied value). One such category is within the new kinds of data (language, images) mentioned earlier. Analysing such data involves prediction as a pre-processing step. This is particularly relevant in the presence of missing data on economic outcomes. For example, a 2016 study trained a neural network to predict local economic outcomes with the help of satellite data in five African countries. Economists can also use such ML methods in policy applications. An example provided by Mullainathan and Spiess’ paper was of deciding which teacher to hire. This would involve a prediction task (deciphering the teacher’s added value) and help make informed decisions. These tools, therefore, make it clear that AI and ML are not to be left unnoticed in today’s world.","excerpt":"“While we can predict house prices with accuracy, we cannot use such ML models to answer questions like whether one needs more dining rooms.” Artificial Intelligence has been a force of nature in many fields. From augmenting advancements in health and education to bridging gaps through speech recognition and translation AI—machine intelligence is becoming more […]","categories":["AI Features"],"tags":["Active Learning"],"author_name":"Mita Chaturvedi","publish_date":"2021-06-08T15:00:00","publication_year":"2021","word_count":1144,"keywords":["Go","machine learning","artificial intelligence","TPU","AI","neural network","programming_languages:R","ML","Active Learning","ViT","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","neural network","TPU","R","Go","ViT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/machine-learning-through-the-lens-of-econometrics\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":69764,"title":"“GOQii aims to be the biggest one stop destination for anything related to health &#038; fitness”- avers Vishal Gondal, Founder &#038; CEO at GOQii Inc.","content":"GOQii, headquartered in California with offices in Mumbai and Shenzhen, is working effortlessly in the direction of enabling permanent shift to a healthier lifestyle. It is doing so by providing a combination of advanced wearable technology, the world’s leading experts, coaches and karma. With a focus on sustained engagement and goal reinforcement, it offers a complete ecosystem and is a powerful combination of technology and human connection. A collaboration of some of the world’s leading experts and amazing management team driven by a passion for healthy sustainable living, the company was founded by Vishal Gondal. A serial entrepreneur, Vishal started his journey at the age of 16 founding IndiaGames and is best known as the Father of the Indian Gaming Industry. He has been listed among the top 50 executives in the mobile content space by the UK based ‘Mobile Entertainment’ Magazine alongside noted people like Steve Jobs. His passion for technology, fitness & gaming led him to his next venture GOQii which is focused on helping people make a change towards a healthier lifestyle and be the force of good. In a candid chat with IoT India Magazine, Vishal Gondal spoke about his transition from gaming to wearable, the growth of wearable industry, challenges faced by the industry, trends and much more. Find out the details here. IoT India Magazine: To begin with, how did the shift from gaming to wearables happen? Vishal Gondal, Founder, GOQii Inc. Vishal Gondal: The shift from gaming to wearables is a personal one. I was leading a very unhealthy lifestyle during my stint at Indiagames.  As a result of long working hours and bad eating habits, I got no time to work out. I ended up using all kinds of fitness bands but none of them helped. Soon after, I met this coach with whom I started sharing my personal data of my water intake, eating habits, step counts and daily work-out routine etc. He started guiding me on a daily basis which made a huge difference to my lifestyle. I then realized that we lack a platform which has a human factor that will be able help, guide and monitor one’s lifestyle to get them into a healthy routine. One’s busy schedule is one of the biggest factors in negligence of one’s health. I feel that motivation and right guidance is what we need to bring a real change in our lives. Hence, the inception of GOQii in November 2014 as I recognized that the biggest problem in the healthcare sector is motivation. Just a band and the app would not solve the problem. So the coaching model in GOQii was introduced and we are constantly working towards introducing more aspects which can help users attain a healthier lifestyle. IoT India Magazine: How are Goqii’s wearable products different from other wearables in the market? What are its key differentiators? GOQii wearable device and App interface VG: GOQii’s aim is to be the biggest one stop destination for anything related to health & fitness. Our biggest differentiator is our coach marketplace model which revolutionizes fitness coaching by empowering a coach based anywhere in India or abroad to offer services to users irrespective of their location. For us, we are always trying to focus on how to improve our services and make our platform completely integrated. GOQii incorporates coaching from leading healthcare professionals, consultations from doctors who can access the user’s data through various technologies like accelerometers, GPS and mobile network etc. Taking it a step further, GOQii has partnered with Thyrocare and Max Healthcare.  Users will able to book their lab and diagnostic tests via the app based on the doctor’s recommendation. Lastly, we felt that our users should be able to pay seamlessly for their tests and so we tied up with Axis Bank for NFC contactless payments. We have been constantly trying to work on partnering with companies in various areas in healthcare to help build a complete ecosystem. IoT India Magazine: How has the wearable market grown in the past few years, particularly in India? VG: The wearable market has witnessed a quick movement in India and other growing countries. Even though the wearable technology landscape continues to evolve, it has created a good market share and has been widely accepted by consumers in India. The wearable industry has grown towards personalization of technology. Companies are now developing products that make technology relevant for individuals. Personalization of technology enables companies to focus on exactly what their customers want and how they want it. Keeping in sync with that line of thought, Global tech behemoths such as Google, Apple and Samsung have been increasingly focusing on their hardware to provide products that seamlessly integrate with their customer’s everyday life. Over the past year, wearable users have reached a point where they want to get more than just data and are looking for devices that can track their habits, interpret them, and give them in-the-moment feedback that will help change their bad habits, and boost their good habits. For 2016, Gartner has forecasted that 274.6 million wearable devices will be sold worldwide, which is an increase of 18.4 % from 2015. In addition, Gartner has also predicted that from 2015 – 2017, smartwatch adoption will have a 48 % growth due to brands such as Apple who are popularizing wearables as a lifestyle trend. In addition, as per the International Data Corporation (IDC) report, the wearable market in India observed a robust growth of 41.9% in Q2 of 2016 over the first quarter of 2016. In Q2 of 2016, GOQii is at the number one position with 16.1% market share of the wearable device segment. IoT India Magazine: What are the new products in store by Goqii? GOQii Home app Screen VG: This year GOQii launched its second generation smart band – GOQii 2.0 which has a sleeker design and is equipped with enhanced features. The features include GOQii doctor consultation, longer battery life and auto sleep detection. For the new band, we have also partnered with Axis Bank for NFC contactless payments. The GOQii ecosystem has been further strengthened by partnerships with Thyrocare and Max Healthcare. For the upcoming year, we have several interesting launches and partnerships in the pipeline. The information for which will be shared with you soon. IoT India Magazine: Would you like to tell us about the funding for Goqii? VG: 2016 was a great year for us. Just few months back, we received funding of undisclosed amount from Mr. Ratan Tata, which was obviously a special moment for all of us at GOQii. In addition, we also have big names on our plate such as the venture capital arm of Edelweiss which happened earlier this year. Both these funding opportunities, boosts our confidence to further accelerate our growth. Our aim is to constantly keep developing the healthcare space in India and to make significant improvisations in the healthcare space in India. Last year in November 2015, we raised USD 13.4 million in Series ‘A’ funding which helped us build a health eco-system. The funding was led by global venture capital firm New Enterprise Associates (NEA) with participation from Cheetah Mobile Inc (NYSE: CMCM), the leading developer of mission-critical mobile utility applications; Great Wall Club (GWC), a private network for innovators and executives from leading mobile companies worldwide; DSG Consumer Partners, an investment company focusing on identifying, selecting and investing in consumer businesses; Ilkka Paananen, CEO & Co- founder of Finnish gaming startup Supercell; angel investor and entrepreneur Pravin Gandhi apart from me being one of the investors. The company has also raised undisclosed amount from Vijay Shekhar Sharma, founder and CEO of mobile internet firm PayTM. IoT India Magazine: Being an investor yourself, what are the kind of start-ups you look for? How can the quality of start-ups nesting out in this space be improved upon? VG: I personally feel for any business to be successful, it has to have a great differentiating idea behind it. With so much competition in the market, one has to be innovative and original. We all need to have that passion and understand how feasible the business model is in the current scenario. Without the right amount of drive and desire, no idea can be successful. Even when I started GOQii, our coaching model is what stood out and still stands out. One just needs to think different. The same way, when I take decisions for investing in new businesses, I look at three factors– passion, originality and feasibility of the business idea. The passion and drive of the team to make their business model a success plays an important role in the growth of any business. IoT India Magazine: How IoT as a sector is evolving today? What are the key contemporary trends you see emerging in the space? VG: One of the trends that I see emerging is that the evolution of wearable technology in healthcare will revolutionise the health insurance industry. Wearables shall play a crucial role in determining the premiums in accordance with the risks associated with individual and group health insurance plans. Insurers will be able to use this information to set targets and incentives for policyholders to live a healthier lifestyle, encouraging them with the possibility of lower premiums. With security being a major threat to all, wearables shall play an important role in combating safety issues by introducing SOS devices alternately using them as access devices. Many times, when faced with personal threat, it becomes almost impossible to take the smartphone out from your pocket or handbag, unlock it, and send out an emergency message. In such situations, a wristband could make all the difference in broadcasting that crucial SMS with your location data to your immediate emergency contacts. Thus, I believe that the future of wearable technology isn’t about fitness bands or health monitors, it’s about what can be done with the data accumulated and how can it be used to an individual’s advantage. Additionally, I also believe this kind of deep analysis of data is not something a single company can achieve as it will need collaboration between industries, technology and science to provide meaningful insights. IoT India Magazine: What are the major challenges of being a wearable tech company? VG: I think one of the biggest challenges in the healthcare space is motivation. Overall, there is a lack of motivation to take care of one’s health. There is a plethora of information on healthcare which is available on the internet and through people. The problem arises when people don’t know how to make the right use of the information. This is where GOQii makes the difference. Through our coaches and now doctors, we are constantly aiming to improve the lifestyle of people. We hope to make a difference in the health ecosystem by motivating people and making their lives healthier than before.","excerpt":"GOQii, headquartered in California with offices in Mumbai and Shenzhen, is working effortlessly in the direction of enabling permanent shift to a healthier lifestyle. It is doing so by providing a combination of advanced wearable technology, the world’s leading experts, coaches and karma. With a focus on sustained engagement and goal reinforcement, it offers a […]","categories":["AI Features"],"tags":["Interviews and Discussions","Wearable India"],"author_name":"Srishti Deoras","publish_date":"2017-01-02T11:07:22","publication_year":"2017","word_count":1812,"keywords":["Go","API","ELT","startup","AI","Wearable India","ML","RAG","Aim","GAN","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","ML","Aim","RAG","R","Go","API","ELT","GAN","startup"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/goqii-aims-biggest-one-stop-destination-anything-related-health-fitness-avers-vishal-gondal-founder-ceo-goqii-inc\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10048429,"title":"Social Media Rumours: Can ML Help?","content":"When rumours spread through social networks, their impact on human lives becomes more and more pronounced; hence, it is vital to research the fundamental rules of rumour transmission in online social networks and propose effective rumour management techniques. Since the outbreak of Covid-19 last year, the volume of fake news, rumours, and doctored stories and videos circulating on social media has risen. Recent news reports that the Maharashtra Police Department has been fighting the dissemination of false information on social media and the Internet for a long time. As a result, during the crisis, a total of 852 offences were recorded. The Importance of Rumour Detection Rumours can contain a wide range of information, from simple gossip to sophisticated propaganda and commercial material. According to the researchers, the phenomenon of viral marketing is based on rumour-like mechanisms, in which corporations use their consumers’ social networks on the Internet to promote their products via the so-called “word-of-email” and “word-of-web”. Furthermore, a group of scholars say that rumour-mongering is the foundation for gossip algorithms, a class of communication protocols used for large-scale information distribution on the Internet and in peer-to-peer file-sharing services. A variety of machine learning (ML) algorithms has been deployed to detect rumours on social media. There are three types of ML approaches: supervised, unsupervised, and hybrid. Let’s have a look at these approaches in general. Supervised Learning Techniques A group of researchers from the United States framed several indicators in the social media ecosystem that enable users to judge information authenticity after the Yahoo research team introduced engineering elements for credibility evaluation on Twitter. Support Vector Machines (SVM), decision trees, decision rules, and Bayes networks were among the supervised learning approaches used. Similarly, scholars in China have addressed an early study on rumour analysis and identification on Sina Weibo, a famous Chinese social media site. SVM using the Radial Basis Function (RBF) kernel was trained in the first experiment utilising a subset of the originally provided characteristics (content-based, account-based, propagation-based). Furthermore, K-Nearest neighbour (KNN) was used on user-based features, and Naive Bayes classifier (NB) was used on content-based features in an annotated Twitter data set. The list goes on and on. Unsupervised Learning Techniques In the case of unsupervised learning, the situation is similar. Australian researchers proposed a clustering-based approach for detecting political rumours on Twitter. Furthermore, the researchers from Singapore regarded the detection of fake rumours on Sina Weibo as an anomaly detection process. They came to the conclusion that for rumours, an unsupervised technique is superior to K-means. In a similar vein, Indian researchers proposed four major methods for spotting rumours on Twitter, as a result of which, a greater proportion of tweets are classified as False Positives\/Negatives due to sentiment scores above the threshold. Hybrid Techniques Researchers from Japan have also presented a framework for detecting rumours on social media sites. Online bulletin board kakaku.com, Japan’s most widely used marketing site, was used to collect word-of-mouth messages. To detect rumour information, a Swiss researcher created graph topology-based distance. It was used to track changes in the topology of concept graphs over time. Some Chinese studies blend the supervised and unsupervised learning methodologies. In order to do this, they extract elements from the content of retweets and comments and look for rumours on Sina Weibo. Scholars from the United States investigated rumour spreading trends in the social media ecosystem in a similar way. Prospects for Rumour Detection in the future With the rise of social media, all users now have quick access to information. Because there is no strict social media policy in place to regulate and monitor online activity, most individuals are disseminating bogus news on social media and have been victims of it. For detecting rumours via social media, many machine learning models have been developed. There is usually a lot of interest in detecting rumours on social media among researchers; however, a viable ML model is necessary. Scholars and ML startups will need to put in a lot of study time.","excerpt":"A robust machine learning model is required to detect rumours on social media.","categories":["AI Features"],"tags":["anomaly detection","decision trees","k-means","Naive Bayes classifier","Support Vector Machine"],"author_name":"Dr. Nivash Jeevanandam","publish_date":"2021-09-16T12:00:00","publication_year":"2021","word_count":667,"keywords":["Go","machine learning","startup","programming_languages:R","AI","Naive Bayes classifier","decision trees","Support Vector Machine","ML","anomaly detection","ViT","GAN","R","k-means"],"extracted_tech_keywords":["AI","machine learning","ML","anomaly detection","R","Go","GAN","ViT","startup","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/social-media-rumours\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10048459,"title":"These Leading Activists Are Promoting Ethical AI","content":"Tech giants are racing to compete with each other on the AI advancement front. Unfortunately, in doing so, they often miss out on identifying the wrong from the right. This is where ethical AI steps in. Artificial intelligence that sticks to the guidelines related to fundamental values — privacy, non-discrimination, rights and morals — are termed ethical AI. While being awestruck by the advancement made by tech giants in the field of AI, we often neglect the potential and terrifying long-term consequences of these technologies. According to MarketsAndMarkets, the global AI governance market is predicted to reach $1,016 million by 2026. The market is expected to grow at a CAGR of 65.5 per cent between 2020 and 2026. Increased government initiatives, the need to build trustworthy AI systems, and the rising awareness around building transparent AI systems have to be attributed to this growth. In the last couple of years, tech giants — the likes of Google, Microsoft and IBM, have started considering and taking actions against the unethical practices of AI and the use of technologies that can cause potential harm. However, AI advocates have been vocal in their fight against unethical uses of AI. Today, we look at some of the prominent advocates of ethical AI from across the globe. (The list is prepared in no specific order). Abhishek Gupta Software Developer II – Machine Learning at Microsoft, Abhishek Gupta is the Founder and Principal Researcher of Montreal AI Ethics Institute, which caters to tangible and practical research in the ethical, safe and inclusive development of AI. Abhishek’s research focuses on applied tech and policy methods to address ethical and inclusivity concerns in using AI across arrays. At Microsoft, Abhishek serves on the Commercial Software Engineering Responsible AI Board as well. Abhishek’s work has been recognised by governments from North America, Europe, Asia and Oceania. Additionally, he has been working with an interdisciplinary experts group to host workshops, conduct research and develop an AI ethics curriculum, and work on AI ethics audits for different organisations. Source: LinkedIn Ansgar Koene Ansgar Koene is the Global AI Ethics and Regulatory Leader at EY. He focuses on Computational Neuroscience, Computational Social Science, AI policy development and engagement, algorithmic accountability and transparency. His work focuses on building regulatory tools that maximise the benefits of information technologies while minimising the negative impact on people and society at large. Before EY, he was associated with The University of Nottingham for more than seven years when he was the research co-investigator on the UnBias project. Ansgar is also a member of the AI Ethics Board in Hayden AI, which builds AI-powered platforms for smart and safe city applications. Ansgar is also associated with the Advisory Board of We and AI, an NGO working towards increasing public awareness and conception of AI in the UK. Source: LinkedIn Joy Buolamwini PhD candidate at MIT Media Lab, Joy Buolamwini, is the founder of The Algorithmic Justice League (AJL) — an organisation combining art and research to bring forth the social implications and the potential harm of AI. Joy is an Algorithmic Bias Researcher. Her most prominent work has been her research on AI inaccuracies in facial recognition technology (sold by IBM, Microsoft and Amazon) and automated assessment software. AJL has been working since 2016 on raising awareness about the impacts of AI by carrying out research, powering the voice of the impacted and inspiring industry practitioners to mitigate AI bias and harm. Source: LinkedIn Saishruthi Swaminathan Ethical AI advocate, Saishruti Swaminathan, is a data scientist and co-creator of the R-Code for the AI-Fairness 360 product. She is an Electrical Engineering graduate from San Jose State University, and at present, is associated with IBM as its Advisory Data Scientist – AI Strategy and Innovation. Saishruthi was also one of the semi-finalists in the Silicon Valley Business Plan Competition for 2021 for democratising recruiting using AI. Source: LinkedIn Deborah Raji Deborah Raji is the Founder and Executive Director of Project Include, a non-profit initiative that provides access to engineering education to the underserved and immigrant communities. Presently a fellow with Mozilla, Deborah is a computer activist, and her work focuses on algorithmic bias, AI accountability and algorithmic auditing. A graduate of Engineering Science from the University of Toronto, Deborah featured in Coded Bias, a documentary on the fallout of the MIT Media Lab researcher Joy Buolamwini’s discovery of racial bias in facial recognition algorithms. She has also worked with Joy on Algorithmic Justice League, Google’s Ethical AI team and New York University operationalising ethical considerations in ML engineering practices. Source: LinkedIn Kay Firth-Butterfield Kay Firth-Butterfield is the Head of AI and ML and a member of the Executive Committee of the World Economic Forum. Kay is also a Barrister, former Judge, professor and technologist. As an entrepreneur, she co-founded AI Global and was the world’s first Chief AI Ethics Officer in 2014. Kay is recognised as one of the foremost experts globally on the governance of AI, and she is the brainchild behind Twitter’s #AIEthics. Kay is a member of the Technical Advisory Group at the Foundation for Responsible Robots. Source: LinkedIn Raj Shekhar Raj Shekhar is the Founder of AI Policy Exchange — an international cooperative association of institutions and individuals working at the intersection of AI and public policy. At present, he is the Lead of Responsible AI at NASSCOM. His work focuses on supporting NASSCOM’s efforts at defining the responsible AI roadmap in India. An alumnus of National Law School, Raj is also a member of the Founding Editorial Board – AI and Ethics Journal, at Springer Nature which promotes informed debate around ethical, policy and regulatory implications that surface with the development of AI. Additionally, he is also affiliated with Harvard Kennedy School of Government’s The Future Society, a non-profit think tank that questions the governance of emerging technologies. Source: LinkedIn Olivia Gambelin AI Ethicist Olivia Gambelin is the Founder and CEO of Ethical Intelligence (EI) Associates, Limited. EI promotes human-centric tech by making ethics accessible and affordable for everyone. Olivia is also a member of the Founding Editorial Board – AI and Ethics Journal at Springer Nature. According to Olivia, AI is not inherently dangerous; our usage of AI makes it cause danger. She believes ethics is the study of knowing and understanding the difference between good and evil, right and wrong, based on a system of values. Source: LinkedIn","excerpt":"The global AI governance market is expected to reach $1,016 million by 2026, growing at a CAGR of 65.5 per cent between 2020 and 2026.","categories":["AI Features"],"tags":["Ethical AI","Interviews and Discussions","Machine Learning"],"author_name":"Debolina Biswas","publish_date":"2021-09-16T14:00:00","publication_year":"2021","word_count":1065,"keywords":["Go","artificial intelligence","machine learning","AI","Ethical AI","ML","Machine Learning","Ray","ViT","Rust","GAN","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","Ray","R","Go","Rust","GAN","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/these-leading-activists-are-promoting-ethical-ai\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10120254,"title":"Zomato is Hiring for Its Generative AI Team","content":"Indian unicorn Zomato is looking for new members to join its generative AI team. Vaibhav Bhutani, who is working on building a central AI team for Zomato and its quick commerce business Blinkit, took to LinkedIn to share the update through an unconventional or what he calls a “clickbaity” post. The call to action is clear: if you have built something impressive, Zomato wants to see it. Applicants are encouraged to prioritise showcasing demos of their work over simply submitting resumes. Starting with a teaser question about Sam Altman’s daily shipments, Bhutani shifts focus, challenging readers about their future contributions using LLMs and introducing them to emergent AI companies like Jonathan Ross’s Groq and Vipul Ved Prakash’s Together AI. Interested candidates can reach out directly by sending their demos alongside their resumes to bhutani@blinkit.com. The expansion news comes soon after Zomato released its Q4FY24 earnings report, highlighting a robust year-over-year top-line growth of 61%, surpassing the projected outlook of 40%. This quarter also marked a milestone for Blinkit, which turned Adjusted EBITDA positive in March 2024. According to founder and CEO Deepinder Goyal, the growth was driven by impressive increases across Zomato’s B2C businesses, with the food delivery, quick commerce, and Going-out verticals growing by 28%, 97%, and 207%, respectively. Blinkit is planning further expansion, targeting an increase from 526 to 1,000 stores by the end of FY25. Read more: Data science hiring process at Zomato","excerpt":"Indian unicorn Zomato is looking for new members to join its generative AI team.","categories":["AI Hirings"],"tags":["Generative AI","Top Trend","Zomato"],"author_name":"Shritama Saha","publish_date":"2024-05-13T19:24:34","publication_year":"2024","word_count":236,"keywords":["data science","Top Trend","Go","unicorn","AI","RPA","RAG","Together AI","generative AI","Groq","Zomato","Generative AI","R"],"extracted_tech_keywords":["AI","data science","generative AI","RAG","R","Go","RPA","unicorn","Together AI","Groq"],"url":"https:\/\/analyticsindiamag.com\/ai-hiring\/zomato-is-hiring-for-its-generative-ai-team\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10085656,"title":"Spotify to Layoff 6% of its Workforce","content":"After laying off around 38 employees in October last year, Spotify, the music and media streaming giant, is about to lay off around 6% of its staff (roughly 590 employees). As per its December quarter earnings report, Spotify had a total of 9,800 employees globally. While writing an email to its staff, which was published on Spotify’s website, Spotify chief Daniel Ek said, “I hoped to sustain the strong tailwinds from the pandemic and believed that our broad global business and lower risk to the impact of a slowdown in ads would insulate us.” He highlighted how he was ambitious in investing ahead of their revenue growth. Citing the same reason, Ek said, “Today, we are reducing our employee base by about 6% across the company”. It is noteworthy that the music streaming giant made a significant investment in podcasting starting in 2019. It is estimated that the company spent over a billion dollars buying podcast networks and the rights to a number of well-known programmes, including the Joe Rogan Experience. Owing to the number of investments which couldn’t bring results, Spotify’s share prices too have tumbled down 66% in last one year, which further fuelled investors’ discomfort. According to their executives, the podcast business might become profitable in next one to two years. This is the third significant tech layoff in the last week. In a letter to staff on January 18, Microsoft CEO Satya Nadella announced that the company would be eliminating 11,000 jobs, or about 5% of the total workforce, in order to offset a global slowdown in tech and advertising investment among businesses due to inflationary fears. Alphabet, the parent company of Google, also disclosed intentions to lay off 12,000 workers on Friday, or about 6% of its staff.","excerpt":"Spotify, the music and media streaming giant, is about to lay off around 6% of its staff (roughly 590 employees).","categories":["AI News"],"tags":["spotify"],"author_name":"Lokesh Choudhary","publish_date":"2023-01-23T18:20:03","publication_year":"2023","word_count":293,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","spotify","GAN","R"],"extracted_tech_keywords":["AI","R","Go","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/spotify-to-layoff-6-of-its-workforce\/","complexity_score":3,"technical_depth":6,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":56333,"title":"Top Blockchain Startups In India To Watch Out For In 2020","content":"According to a Nasscom report, over 50% of the blockchain projects in India involve startups across different industries. The trend is picking up, in spite of India having only a 0.2% share in the $5.6 billion venture capital investments in blockchain in the year 2019. According to the same report, more than 40 blockchain initiatives are being taken up by the public sector in India. The global blockchain market is expected to be $20 billion by 2024. Even though it has slowed down a bit in its introduction phase in India, the outlook seems to be great for blockchain startups in India in 2020. As 2020 almost passes up its second month, Analytics India Magazine takes a look at some of the most prominent blockchain startups to watch out for in 2020. The list is in no particular order. Auxesis Group Established in 2015 by Kumar Gaurav, Auxesis focuses on blockchain solutions on an enterprise level. RISE Global has ranked the company in the ‘Top 100 Most Influential Blockchain Company’ in the world. The company is based out of Mumbai and offers products like AuxLedger, AuxCE, AuxPay, Darwinsurance, and Token Bazaar to its enterprise customers. Auxesis solutions have applications across sectors like insurance, supply chain, capital markets, and cross border remittances. For R&D and building customised application developments, Auxesis had collaborated with London-based Cashaa (IITb), and Mumbai based Blockchain lab. Debut Infotech Debut Infotech creates intuitive software applications for modern-day businesses and is a full-service blockchain development company which makes use of technologies like IoT and AI. Established in 2011 by Gurpreet Singh, this blockchain startup offers services like blockchain development, smart contracts development, hyper ledger fabric development, PoC development, Cords development, Dapps development, blockchain consultancy, and advisory services which include mobile apps, IoT, web, and cloud. The company holds clients like Everledger, iFinca, HDFC Bank, KFC among others, and till date, it has secured a total of $40 million for funding. Elemential Elemential is a platform that enables developers to build and manage blockchain applications at a scale. Elemential was founded in the year 2015, by Raunaq Vaisoha, Anil Dukkipatty and Sahil Kathpal. Based in Bengaluru, Elemential has raised funding from Matrix Partners, Amrish Rau, Invetopad, Hinduja Group, Digital currency group, Lightspeed India, Eight Innovate, Amit Ranjan, Prashant Malik and other prominent angel investors. Elemential’s product, Hadron, makes blockchain administration a secure and hassle-free process. In 2018, Elemential labs got funding of an undisclosed amount from Matrix Partners and some other investors. EzyRemit Vishal Kanvaty and Abhijeet Jaswal founded EzyRemit in the year 2015. This startup focusses on the remittance market and usage of blockchain across different business and functional domains. Their products include EzyRemit and Ezy Hedge provide blockchain solutions to the remittance market, and are available for demos. The startup focuses on solutions for the bigger ecosystem and for organisations that are trying to adopt blockchain and crypto tech for distributed computing, trust validation, and simplification of processes. MindDeft This startup, since its establishment in 2015, has been focusing on developing blockchain apps for efficient business processes. Krunal Soni founded MindDeft to help in identifying Dapps and smart contracts matching the business requirements. Their services include cryptocurrency development, smart contracts, token sale, private blockchain, distributed ledger, and legal contracts. This startup has been working with platforms like Ethereum, Hyperledger, Stellar, Quorum, EOS, Tron, and R3 Corda for its various projects. PSI PHI Blockchain Labs This startup has its headquarters in Bangalore and was established in 2016. Founded by Gaurav Kumar, Aditya Prasad and Harsh Pokharna, PSI PHI Blockchain Labs builds solutions for digital economy-based on distributed ledger technologies. PSI PHI Blockchain labs mainly work in the supply chain, telecom and healthcare industries, and offers products like Crypto Locker (store and share documents on the blockchain using APIs), and Digi Rail (multi-party shared database) for supply chain data flow optimisation. Primechain Founded in 2016 by Rohas Nagpal and Shinam Arora, Primechain Technologies develops blockchain-based solutions for the banking system of India. It has a collaboration with 27 banks across India and the Middle East for verification, authentication and storage of electronic records. Their community of banks is called BankChain. Along with the banking sector, Primechain Technologies also provides blockchain solutions to other industries like capital markets, healthcare, governments, pharma, manufacturing, insurance, aviation, shipping and defence. The company offers products like Primechain CONTRACT, Promechain API, Primechain LOAN, Primechian KYC, Primechain MONEY, Primechain Charge Registry. Signzy This company was established in the year 2015 and is headquartered at Bangalore. Signzy was founded by Ankit Ratan, Arpit Ratan and Ankur Pandey, to combine AI and blockchain to make products user-friendly, compliant, and more importantly, secure. Signzy offers three main products; RealKYC, Digital Contracts, ARI. In 2018, Signzy raised over $3.6 million Series A funding by Stellaris, Kalaari. Since 2015, this remains one of the best blockchain startups in 2020. Sofocle Sofocle was founded by Nidhi Chamria and Ravi Chamria in Jan 2017 and is headquartered at Noida with investors Sofocle that provides consultation on wallets, exchanges, smart contracts and private blockchain products. Some of their products include SofoCap, supply chain financing solution; SofoChain, a product for supply solution; and SofoInsure, an autonomous claim processing solution. The company offers blockchain-based solutions across industries like finance, health, and manufacturing sectors. Somish One of the oldest Indian startups, Somish, was founded by Ish Goel. With its headquarters in New Delhi, Somish has been in product development with expertise in building automation using the latest tech. Currently, the company has been developing blockchain-based solutions in India and around the globe. Somish has been working on blockchain technology to provide solutions to sectors like P2P insurance, aviation maintenance, crisis fund distribution, subsidy distribution, bill discounting, and tokenised fund transfer. GovBlocks is Somish’s flagship blockchain solution, which is a blockchain protocol for decentralised governance built on Ethereum.","excerpt":"According to a Nasscom report, over 50% of the blockchain projects in India involve startups across different industries. The trend is picking up, in spite of India having only a 0.2% share in the $5.6 billion venture capital investments in blockchain in the year 2019. According to the same report, more than 40 blockchain initiatives […]","categories":["AI Trends"],"tags":["iot friendly database","latest ai products across industries","supply chain analytics projects"],"author_name":"Sameer Balaganur","publish_date":"2020-02-11T10:00:00","publication_year":"2020","word_count":970,"keywords":["Go","API","Rust","AI","distributed computing","Git","RAG","Aim","analytics","iot friendly database","latest ai products across industries","supply chain analytics projects","R"],"extracted_tech_keywords":["AI","analytics","Aim","RAG","distributed computing","R","Go","Rust","Git","API"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-blockchain-startups-in-india-to-watch-out-for-in-2020\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":51681,"title":"Top AI-Based Attacks That Shocked the World In 2019","content":"Artificial Intelligence is still learning and getting stronger each day. But the pros and cons for the technology have always been in the limelight. From its use for data breaches to deepfakes, 2019 has shown us that AI is here to stay, but not without putting a price on our comfort. 2019, like the previous year, has been about privacy and security breaches. Where organisations take up most of your data to feed it to a machine-learning algorithm to get better results, you begin to question whether the data that you haven’t shared yet is worth anything. There were also some instances where your nightmare of ‘AI overtaking the humans’ was revisited. Let’s take a look back at 2019 where AI became a little more ‘infamous’: 1. Deepfakes: Yes, Deepfakes. Imagine your favourite social media platform’s head comes right out in a video confessing to manipulating you using your own data when, in reality, he wasn’t the one in the saying those things in the first place. What if no one told you that it wasn’t Mark Zuckerberg? Would you have believed it? When you look at the video, probably yes. Jordan Peele, the versatile Hollywood personality, showed the power of Deepfakes and how information can be manipulated. Deepfakes were on the rise in 2019, showing the potential threat that they can have when they are stopped using in a funny context. Deepfakes became so troublesome that China went as far to ban them. 2. The algorithm bias: The AI algorithm bias fiasco continued in this year. Apparently, in America, health care algorithm’s bias has prevented many dark-skinned patients from being able to receive extra help. This came down from almost 50% to less than 20%. The algorithm bias became such a big issue that the American Senators have started to protest the ‘racially biased’ health algorithm saying that the formulas have many ‘historical and human biases built-in’ 3. Two of the biggest Data exposures of all time: 2019 saw two of the biggest data exposures of all time. Where one involved Facebook where more than 540 million users’ data was accessible on unprotected servers until April 2019, as if Mark Zuckerburg didn’t have enough problems. The other was when the financial records of 885 million people were left exposed by First American on public servers where anyone could access them until May 2019. It seems like over the last decade hackers have become more sophisticated with the help of AI when it comes to data and privacy breaches. 4. AI offers to write fake news: GPT-2 is the name of the machine learning model that was trained on a dataset of 8 million web pages. It was able to adapt the style and content of some of the first pages that it was given from the data. AI could first implement its algorithms on completing first few sentences or write poems, at times better than human beings, but when BBC decided to test it with a real story, the output it gave out made it clear, as BBC put it ‘that it could be used to create fake news pr abusive spam on social media’. 5. Ransomware: What AI offered to the world of cybersecurity is its ability to detect malware as soon as it enters into a system, or in some cases before it enters it. But, at the same time hackers have used AI to blackmail someone for money after blocking access to their computer system. In June 2019, two Florida city governments were hit by a ransomware attack and in order to recover their affected data assets they had to pay more than a million dollars. The first one was Riviera Beach Municipal Party, which paid around $600,000 in bitcoin in order to recover their IT systems which were locked by attackers. The second was Lake City, who had to give up $460,000 to malefactors after suffering Emotet malware downloaded Trickbot and subsequently Ryuk ransomware. 6. AI Defeating the Human Mind: In case if you were paranoid about AI moving beyond human intelligence, that thought isn’t too far. At least in the video gaming world. The idea of AI beating top human video gamers in a strategy game like Starcraft 2 to the point that it defeats 99.8% of them may seem ordinary but, what you should actually be thinking about is that the AI isn’t ‘just playing’ games. The AI is being trained to respond strategically in the real world scenarios. The players deemed the AI AlphaStar’s strategies as unorthodox at times, which tells us how far AI’s own decision making has come. The AI-AlphGo in 2016 defeated Lee Se- Dol, an 18-time world Go Champion, who retired this year saying ‘the AI cannot be defeated’. Not only this, but AI has been able to perform better than a human mind’s thinking capabilities in many other games. At this pace, AI will continue to develop and learn to a point that some fear will be beyond human control. 7. Smart Phishing attacks: Hackers have been using phishing attacks in order to exploit the target’s sensitive information and aim to fool the victim that the phishing methodology is legitimate. This year saw a rise in phishing attacks and these smart phishing attacks have been at the highest level since the last three years.","excerpt":"Artificial Intelligence is still learning and getting stronger each day. But the pros and cons for the technology have always been in the limelight. From its use for data breaches to deepfakes, 2019 has shown us that AI is here to stay, but not without putting a price on our comfort. 2019, like the previous […]","categories":["AI Trends"],"tags":["data breach","deepfakes","Ransomware"],"author_name":"Sameer Balaganur","publish_date":"2019-12-11T16:00:00","publication_year":"2019","word_count":887,"keywords":["Go","artificial intelligence","machine learning","TPU","AI","Ransomware","deepfakes","Git","data breach","GPT","Aim","GAN","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","Aim","TPU","R","Go","Git","GPT","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-ai-based-attacks-that-shocked-the-world-in-2019\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10072442,"title":"JP Morgan Chase To Hire 5,000 Technologists In India","content":"JPMorgan Chase announced a hiring spree to hire 5,000 technologists in India in 2022 as they step up their operations. A US-based financial services company, JPMorgan Chase is one of the largest employers in India’s foreign banking arena. The company hired 4,000 employees in 2021. JPMorgan is focusing on hiring technologists from Hyderabad, Bengaluru and Mumbai. The company is hiring digital operators from the field of big data and robotics, which are emerging as the future of banking operations. The company is increasingly investing and building on complex technology roles in AI, analytics and blockchain. Assigning global roles encompassing product, design, and technology partners to India, the hiring is for individuals who work in AI, machine learning, cybersecurity, cloud computing and data engineering. Ravi Pasupuleti, Managing Director, Mumbai Tech Center Lead said, “Our India employees are being aligned to global product roles and are driving initiatives globally.” One third of the bank’s employees are based in the three cities and rising in importance on the global map acting as innovation hubs linking multiple lines of functions and business across the world. JPMorgan Chase has been focusing on hiring computer science majors but also hiring from ancillary streams through ‘Tech Connect’, one of their hiring initiatives. The hiring is directly at mid- and senior levels, leveling up to managing directors with a dedicated training programme.","excerpt":"The hiring is for individuals who work in AI, machine learning, cybersecurity, cloud computing and data engineering.","categories":["AI News"],"tags":["JP Morgan"],"author_name":"Mohit Pandey","publish_date":"2022-08-09T15:34:48","publication_year":"2022","word_count":224,"keywords":["big data","JP Morgan","machine learning","cloud computing","AI","innovation","Git","data engineering","analytics","GAN","R"],"extracted_tech_keywords":["AI","machine learning","analytics","cloud computing","R","Git","big data","data engineering","GAN","innovation"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/jp-morgan-chase-to-hire-5000-technologists-in-india\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10085739,"title":"A Surprising Turn of Events in the World of AI","content":"In 2022, there were a multitude of generative AI models released by think tanks such as Microsoft-backed OpenAI and Alphabet’s DeepMind, along with several others. While this battle between Microsoft and Google rages on, a clear winner has already emerged—NVIDIA. Nearly all the existing state-of-the-art models are trained on NVIDIA GPUs. NVIDIA’s CUDA API and its GPUs’ suitability for AI workloads have made these chips a must-have for training AI workloads. Let’s delve deeper into how NVIDIA is playing all quarters and will inevitably profit the most from the upcoming AI wave. NVIDIA’s command over AI NVIDIA’s grip over the AI market is clearly evident. For instance, OpenAI used over 10,000 NVIDIA H100 and A100 GPUs for training ChatGPT, and Stable Diffusion took around 200,000 GPU hours to train on the NVIDIA A100 GPU. Cloud providers such as AWS and Azure have also partnered with NVIDIA to create humongous clusters of NVIDIA GPUs for enterprise training workloads. With their latest generation of GPUs—termed the H100 series—NVIDIA has solidified their lead in the AI training space. The company is also doubling down on their market leader position, launching the DGX series of purpose-built AI supercomputers built on Volta GPU architecture. This series—combined with the Titan V and Quadra GV100—brings AI computation to enterprises and consumers alike. However, even as NVIDIA has the first mover advantage and has led the market for the past decade, other companies seem to be catching up fairly quickly. NVIDIA’s long-time rival, AMD, has also made its first strides into the AI accelerator game. At CES 2023, AMD’s CEO Lisa Su announced the latest iteration of the AMD Instinct AI accelerator. This chip combines AMD’s CPU and GPU tech into one package and is aimed for deployment in enterprise use-cases. However, AMD’s consumer GPUs are far behind NVIDIA’s offerings in terms of usability for AI use-cases, mainly due to the lack of a CUDA-equivalent API. While AMD has dedicated resources to develop the ROCm open software platform for machine learning applications, it is yet to reach the maturity that CUDA boasts after 14 years of constant development. NVIDIA GPUs are also better integrated into common AI tools such as TensorFlow and PyTorch, and usually run into lesser bugs and issues in a work environment. In addition to AMD, tech giants like Google and Amazon have also poured resources into creating custom silicon for AI workloads. AWS has created Inferentia, a chip targeted at inference workloads on the cloud while Google has created the Tensor Processing Unit for AI tasks running on TensorFlow. These chips beat out NVIDIA’s GPUs for specific workloads, but NVIDIA’s chips are better suited for general purpose tasks, thereby making them usable for more applications. GPUs and AI are inseparable In order to understand why GPUs are a perfect fit for training large AI models, we must first look at their architecture. GPUs were first created as a way to render graphics and high-resolution images at a time when CPUs weren’t capable of doing so. These chips are made up of smaller cores, dubbed CUDA cores by NVIDIA (alternatively known as streaming multiprocessors). Today, the average GPU contains thousands of these cores that are built to conduct several tasks in parallel. Training AI algorithms is suited for GPUs since most training workloads consist of tasks such as matrix multiplication, which are naturally suited for parallelism. This idea was first solidified in a research paper titled, ‘Large-scale Deep Unsupervised Learning using Graphics Processors’ by Rajat Raina, Anand Madhavan, and Andrew Y. Ng, published in 2009. In hindsight, using the parallel processing capabilities of GPUs at a time when CPU cores were lacking in number seems like an obvious choice. This work sparked the trend of more AI researchers using GPUs for training tasks, leading NVIDIA to add more AI capabilities to their chips. It began with updates to CUDA—the API that allows developers to directly interface with and address GPU cores—followed by the addition of dedicated compute units for AI workloads known as ‘Tensor cores’. These cores are optimised for conducting floating point operations, which are an integral part of AI training and inference workloads. Even as custom silicon makes its way to the enterprise market and eventually to the end consumer through cloud services, the obtainability and support of NVIDIA’s consumer GPUs has gained an unbreakable stronghold on the AI market. The pervasive nature of CUDA, coupled with Tensor cores on their latest GPUs, has established NVIDIA as the reigning leader of the AI accelerator space. While open-source APIs such as OpenCL have made inroads into NVIDIA’s CUDA dominance, the vertical integration present in their tech stack dissuades competition, at least until their competitors catch up to their mighty R&D efforts.","excerpt":"While OpenAI, DeepMind, StabilityAI, Midjourney, and Meta slug it out, NVIDIA plays all quarters in the battle.","categories":["AI Features"],"tags":["AIM Rising summit"],"author_name":"Anirudh VK","publish_date":"2023-01-24T16:41:45","publication_year":"2023","word_count":787,"keywords":["ChatGPT","machine learning","OpenAI","AI","AIM Rising summit","PyTorch","AWS","RAG","Aim","generative AI","TensorFlow"],"extracted_tech_keywords":["AI","machine learning","generative AI","ChatGPT","OpenAI","Aim","TensorFlow","PyTorch","RAG","AWS"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/nvidia-emerges-as-clear-winner-in-ai-showdown\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":11005,"title":"Artificial Narrow Intelligence vs Artificial General Intelligence","content":"Artificial Intelligence, a branch of computer science, aims at creating or developing machines such that they are able to perform tasks very similar to humans. AI researchers over the years have been conducting research and building prototypes to match human brain like thinking for a machine. Artificial Intelligence has 2 parts to it – Artificial Narrow Intelligence and Artificial general Intelligence. Let’s explore what each one really means and how different are they from each other. Artificial Narrow Intelligence Artificial Narrow Intelligence (ANI) refers to a machine’s ability to perform a single task extremely well and may be even better than a human. ANI works wonders in cases where a machine is expected to run automated tasks that are generally simple and repetitive in nature. Bots are a perfect example of ANI. Whether you want a restaurant recommendation or a weather update, bots today have numerous uses. Bots function by pulling data from larger databases and finding just the right answer that you were looking for whether it is a restaurant recommendation or traffic update for your route. Bots have started to be a part of our everyday lives. Customer service is one area for which ANI is a boon. ANI technology has enabled bots to answer customer queries which are simple and repetitive in nature. Bots being able to deliver consistency, accuracy, and speed in customer service is benefiting organizations in building\/maintaining their brand name. Artificial General Intelligence Artificial General Intelligence (AGI) is based on the principle that machines can be made to think. As per AGI, machines have the ability to represent human mind and can function in the similar fashion as human brain. If this is true, then machines will be able to reason, think and do all functions that a human is capable of doing.  However, the technology of AGI is still in its nascent stage and would need deep research and long years for it to become a fully functional reality. Work on the development of an artificial neural network is being done by experts in AI, which can function as a proper human being. This technology is being looked upon as a future application of AGI. Artificial Narrow Intelligence vs Artificial General Intelligence The Goal: Both AI areas are developed with different goals to achieve. The focus of Artificial Narrow Intelligence is narrow. It focuses on developing technology that is capable of executing a single task effectively, providing excellent service to the customer in that particular area. Whereas, Artificial General Intelligence focuses on a broad aspect. Its goal is to develop a technology that can think and function so similar to humans that it can be capable of taking place of humans where decision making is required. Intelligence Level: Artificial Narrow Intelligence machines focus only on one area and hence their intelligence is limited to that. Also ANI technology does not enable them to do complex tasks like a human brain. On the other hand, machines that will be empowered with AGI will have the brain to reason out like a human. Its intelligence with match or may be even exceed human intelligence. Areas of Application: Artificial Narrow Intelligence can be applied in areas where the work is repetitive and does not involve any decision making. It can be applied where the work can be predefined and more of information processing is required. AGI, on the other hand, is expected to be capable of being applied in broader areas, areas where decision making and analysis is required. Almost all the areas where human brain is required to do the work, AGI can have its application.","excerpt":"Artificial Intelligence, a branch of computer science, aims at creating or developing machines such that they are able to perform tasks very similar to humans. AI researchers over the years have been conducting research and building prototypes to match human brain like thinking for a machine. Artificial Intelligence has 2 parts to it – Artificial […]","categories":[],"tags":["Artificial General Intelligence","bots"],"author_name":"Manisha Salecha","publish_date":"2016-10-20T11:29:03","publication_year":"2016","word_count":601,"keywords":["Go","artificial intelligence","programming_languages:R","AI","neural network","R","programming_languages:Go","Aim","Artificial General Intelligence","GAN","AI research","bots"],"extracted_tech_keywords":["AI","artificial intelligence","neural network","Aim","R","Go","GAN","AI research","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/artificial-narrow-intelligence-vs-artificial-general-intelligence\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":17175,"title":"Check Out How This Young Startup Is Helping Ecommerce Companies in India Use Visual Search","content":"Over the years, the CPG sector in India has turned to big data analytics to unlock insights from text data by scanning the reviews and determining the customer sentiment. Bengaluru-based Turing Analytics founded by Divyesh Patel and Aditya Patadia in 2015 develops Artificial Intelligence systems using deep learning to help businesses make data driven decisions. To begin with, the young startup has three marquee names in its corner to boost its credibility – Shopclues.com, global powerhouse Kimberly Clark and Tata Sons Ltd. Driving sales and conversions with Customer Review Analytics Divyesh Patel, Co-founder, Turing Analytics In an age where understanding customer and bettering CX has become a science, the founders created their first solution “Customer Review Analytics System” for e-commerce sector that worked amazingly well for both English and Hinglish reviews. Infact, ShopClues became the startup’s first client to successfully integrate Customer Review Analytics to derive more insights from text data.  With a strong market for analytics, ecommerce and retail giants are quickly turning to machine learning systems, NLP and data science to detect emotions, entities and understand the real context from text and social data. Today, in a market saturated with customer analytics solutions, companies, big and small alike are turning to startups or established vendors to race ahead. Pegged behind-the-scene analytics, the underlying objective is to boost online conversions and sales. That’s where Turing Analytics steps in with each of the products having a specialized deep neural network at the core that provides it with intelligence. “In case of Reviews analytics, it analyses language to derive context and in the case of Visual Search and Image tagging, it understands what is there in the image. In case of Fashion Trend recognition system there is a neural net which finds patterns, colors and designs that are occurring more frequently,” said Patel, adding what differentiates their products from market leading vendors. Here’s a look at Turing Analytics Solutions Fashion trend recognition – It is the latest product launched by the startup, the founders have developed Neural network based AI that can identify what is trending among customer base and suggest which products should be pushed. Similar product recommendation – This solution suggests similar looking items from the catalog to the customer. The startup’s deep nets can identify similar items based on only images. Visual search – This solution enables users to search items using photos and it can identify the right items being searched without cropping or category selection. It finds the right results from millions of images within milliseconds. Image recognition – According to Patel, the startup trained customized image recognition models as per client requirements. Most off the shelf solutions don’t identify different ethnic wear with as much accuracy and that’s where we come in, he shared, explaining how their solution differs from what their competitor is offering. Facebook Messenger Fashion Chatbot – An easy-to-integrate fashion Chatbot for Facebook messenger acts as a medium for customers to interact with the brand and ask what they are looking for and make conversions by offering similar products from the catalog. The chatbot is aimed at pushing targeted and promotional messages based on the user’s past behaviour in messenger. According to Patel, their Visual Search solution has many takers in the market and the technology has a much broader application. “We chose to focus our efforts on E-commerce sector for now. We have a few undisclosed clients with catalog size ranging in millions for whom we provide visual search,” he said. Understanding the Turing Analytics USP With a wide expertise in building deep learning models for any given task, Patel emphasizes the young startup’s strength lies in using all types of data (text, image, video etc.) for training neural networks. Which gives them the added advantage of eliminating limitation — there is no industry or sector limitation for the solution which means what works best for fashion can also work for home furnishing with equal efficiency and for medical images and patent search. “We can accommodate any new client without having to make many changes in our system while providing highest level of accuracy,” he said. Patel cited a use case of the highly appreciated “Fashion Trend Scoring system”, developed using Deep Learning for Group Technology and Innovation Office, Tata Sons Ltd, this software was capable of determining the trendiness score of a T-shirt with advanced deep learning algorithms. Outlook Well, with more and more companies grappling with data deluge and customer centricity driving businesses, artificial intelligence presents a tempting opportunity to rise above the competition or risk falling behind. The seven-member team of Turing Analytics likes to focus both on features and driving better  CX, which is evident by its client base. Another trend gaining prominence is visual search and ecommerce giants Flipkart and Ebay have plowed massive investment in this technology and also recently enabled visual search over their global catalogue. The young startup has trained its eyes on this space.","excerpt":"Over the years, the CPG sector in India has turned to big data analytics to unlock insights from text data by scanning the reviews and determining the customer sentiment. Bengaluru-based Turing Analytics founded by Divyesh Patel and Aditya Patadia in 2015 develops Artificial Intelligence systems using deep learning to help businesses make data driven decisions. […]","categories":["AI Startups"],"tags":["Alan Turing","Startups"],"author_name":"Richa Bhatia","publish_date":"2017-08-23T09:50:29","publication_year":"2017","word_count":822,"keywords":["data science","machine learning","artificial intelligence","AI","Alan Turing","neural network","image recognition","NLP","Aim","deep learning","analytics","Startups"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","deep learning","neural network","NLP","data science","analytics","Aim","image recognition"],"url":"https:\/\/analyticsindiamag.com\/ai-startups\/check-young-startup-helping-ecommerce-companies-india-use-visual-search\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":31661,"title":"Django vs Flask: A Look At The Two Popular Python Web Frameworks","content":"Python provides a number of choices of frameworks for web development and other applications such as Django, TurboGears, Pyramid, Flask, Bottle, CherryPy and more. Developers can select the best suitable platform or framework based on their requirement of the business application or project. Frameworks reduce the task of coding and also allows developers to build application faster and customise it according to their preferences. In this article, we look at two of the best frameworks, Django and Flask that can be used by Python programmers and developers. Django This open source framework, also one of the most popular open source frameworks allows programmers to scale Python projects. Django is built on the principle of model view template architectural pattern (MVT) and it helps developers build complex database-driven websites. Advantage: It is fully compatible with database engines and popular applications such as Instagram and Pinterest are powered by Django. When it comes to building real-life applications, this versatile framework is hugely popular with the developer community. The entire prototype concentrates on reusability of tools, reduced code, minimum coupling, rapid development, and the principle of non-repetition. Django offers an optional accessibility to create, read, update and delete. It generates a dynamic interface through self-examining. Modules Required For Django The core of the Django framework is based on MVC architecture, consisting of an object-relational mapper (orm) which creates an interface between data models and relational databases. Django has a classified way of naming the callable objects by generating the HTTP responses. The entire system of HTTP responses and requests are carried by a web templating system and a commonly used URL description. Some of its main features are: An easy and understandable web server for developing and testing A sequential and validating system that can stimulate between HTML forms and values suitable for storage in a database A template prototype that operates the concept of inheritance inspired by the object-oriented programming A caching framework design that uses the services of several cache methods Supports the middleware classes that can intervene in the different stages of a request processing and carry out standard functions An integrated dispatcher system which allows objects of an application to communicate events among themselves via predefined signals A global environment which includes translations of Django’s components into a variety of languages A systematic approach which provides to read XML and JSON representations of Django prototypes A medium to develop Python’s unit test framework Key Features Of Django Provides an extensible authoritative system. Interfaces involving dynamic administration Provides required tools for stimulating RSS and Atom Syndicate feeds. Builds a framework that allows Django to access multiple websites. Generating tools for Google sitemaps. Self-mitigated architecture for cross_site request forgery and scripting, SQL injection, password cracking and other traditional web attacks. A prototype to build GIS applications. Server Utilities Of Django: Django framework is applicable to Apache, Nginx using WSGI,Gunicorn, or Cherokee modules. It is capable of launching a fastCGI server by enabling any web server like Lighttpd or Hiawatha. Supporting Databases: Django can support Databases like PostgreSQL, MySQL, SQLite, and Oracle. The Microsoft SQL server can be assessed with the help of a Microsoft operating system. Well Known Sites Using Django: Public Broadcasting Service, Instagram, Mozilla, The Washington Times, Disqus, Bitbucket, Next Door Flask This beginner-friendly framework is viewed as a micro-framework as it has a simpler interface and is specifically geared for standalone applications. Pegged as an open-ended framework, it was developed after Django in 2010 and gives developers more control over how their databases. It is more suited for building open-ended applications and also has a template called Jinja.  It is referred to as a micro-framework because of its independence from tools and libraries. This means that there is no requirement of database abstraction features for validating components in the flask. This framework has pre-existing third-party libraries which provide the common functions for the web framework to support extensions. The extended functions add integrated applications and help perform object-relational mappers such as validation, upload handling, numerous open authentication technologies and a number of common framework-related tools. Modules Required For Flask Werkzeug: It is the fundamental module for the Flask framework. Started as a simple collection of different services for WSGI applications, the Werkzeug turned out to be the most prominent WSGI utility modules. The module is a collection of a strong debugger, fully developed request and response objects, HTTP services to hold entity tags, cache control headers, HTTP history, cookie management, file uploads, robust URL routing system, and numerous contributed addon modules. Some of its key features are: Strong HTTP header for parsing and dumping Accessible request and response objects Bilateral javascript based in-browser debugger Assured WSGI 1.0 compatibility Supports Python 2.6,2.7 and 3.3 Universal support Supports fundamental session and signed cookies URI and IRI services with Unicode awareness Inbuilt library for fixing buggy WSGI servers and browsers Developed routing systems for matching URL to endpoints and vice versa. Jinja: It is a BSD (Berkeley Software Distribution) licensed template engine developed for the Python programming language. The Jinja engine assures the evaluated templates are stored in a sandbox. Sandbox is a technique for separating running programs. The technique reduces system failures or software susceptibility from spreading. Usually, the sandboxing technique is used to execute untested or doubtful programmes or code. The sandbox provides a well-controlled set of resources for guest programs to execute scratch space on disk and memory. The Jinja template engine empowers personalisation of tags, filters, test, and globals. It also permits the designer to call functions with arguments and objects. Key Features Of Flask Strong automatic HTML escaping to prevent cross-site scripting(XSS) confronts. Allows template inheritance Compiles down to the optimal python code in a fraction of time Elective ahead-of-time template compilation Simple to debug. Line numbers of exceptions will directly indicate the correct line in the template. Structural syntax. Some of the Flask powered companies are Rack space,Netflix,Uber,Samsung","excerpt":"Python provides a number of choices of frameworks for web development and other applications such as Django, TurboGears, Pyramid, Flask, Bottle, CherryPy and more. Developers can select the best suitable platform or framework based on their requirement of the business application or project. Frameworks reduce the task of coding and also allows developers to build […]","categories":["Deep Tech"],"tags":["django","django python","python frameworks","Web Development"],"author_name":"Bharat Adibhatla","publish_date":"2018-12-17T12:42:43","publication_year":"2018","word_count":983,"keywords":["PostgreSQL","Go","AI","R","ML","Web Development","django","RAG","Python","django python","SQL","JavaScript","python frameworks","Java"],"extracted_tech_keywords":["AI","ML","RAG","PostgreSQL","Python","R","SQL","JavaScript","Go","Java"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/django-vs-flask-a-look-at-the-two-popular-python-web-frameworks\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10042006,"title":"Interview With Dr Suraj Rengarajan, CTO, Applied Materials, India","content":"Founded in 1967, California-headquartered Applied Materials Inc. is one of the world’s leading semiconductor products and services providers. The company has offices in multiple locations including Europe, Japan, North America, Israel, China, Italy, India, Korea, and Taiwan. Applied Materials established its Indian office in 2003 and since then, emerged as the second-largest resource for engineering support for its parent company. Analytics India Magazine caught up with Dr Suraj Rengarajan, CTO, Applied Materials India to understand more about the inner workings of the company. Excerpts: AIM: Tell us about your major products & services Dr Rengarajan: Applied Materials is one of the largest providers of semiconductor and display manufacturing equipment. Our materials engineering solutions are used to produce virtually every new chip and advanced display in the world. We also have a robust services business that helps our customers optimise system performance, improve yield and increase output. Applied has among the broadest product portfolio of any company in the chip equipment industry. This allows us to combine technologies in unique ways that improve chip performance, power, area, cost and time to market (PPACt) needed to enable the AI era of computing. In India, we are a strategic partner and an enabler of India’s semiconductor, display and solar ecosystems and are the second-largest resource for engineering support for Applied Materials globally. We provide the organisation with engineering design, support services and cutting-edge innovation in materials science and engineering. We also play a key role in our global Information Technology (IT) infrastructure by offering solutions and services and hosting the company’s second-largest data centre. AIM: What is your latest offering for DRAM scaling? Dr Rengaranjan: Applied recently introduced three materials engineering solutions for DRAM scaling that create new ways to shrink as well as improve performance and power: Draco™: A new hard mask material that helps increase yield by improving uniformity and reducing defectsBlack Diamond®: A new low-k dielectric material that improves interconnect performance and lowers power consumption High-k Metal Gate Transistors (HKMG): Applied is helping enable the adoption of HKMG transistors in advanced DRAM designs to improve performance, power, area and cost AIM: What are your major R&D partnerships? Dr Rengarajan: We have made an overall R&D investment of over INR 85 crore across universities. We have three labs in India working on concept and feasibility, reliability, and materials and chemistry research. We recently celebrated 15-years of collaboration with IIT Bombay in semiconductor and nanoelectronics research. A lot of our advanced research is conducted at our Exploration Center located in the IIT Bombay campus. About 50 Applied Materials India employees with advanced degrees work in this lab in close collaboration with the professors at IIT Bombay in many of our core and adjacent areas. Applied Materials India became the anchor client at IIT Bombay’s Research Park, which is set up to enable technology-focused companies to co-locate R&D assets at the institution. In addition, we have strong academic partnerships with IIT Madras, IIT Delhi, IIT Kanpur, CECRI and IISc that span a broad range of areas from life sciences, AI and Big Data, high-performance computing, flexible electronics, energy storage, speciality coatings and materials engineering. AIM: What is the ASTRA startup accelerator program? Dr Rengarajan: India is the second-largest startup ecosystem in the world. In the past few years, the number of deep-tech startups has grown significantly in India. This makes it an attractive market for companies to look at for innovative ideas for emerging technology applications and new business models in these areas. Applied Ventures, LLC, the venture capital arm of Applied Materials, Inc., boosts promising deep-tech startups in India through mentorship and collaboration as a part of the ASTRA (Applied Start-up Technology and Research Accelerator) program. We launched ASTRA in 2019 to support and nurture these deep-tech startups. We have held two events so far, and some of the startups we are now mentoring include Ships Sciences, Morphle Labs, LightSpeedAI, and Multi Nano Sense Technologies (MNST). Startups selected for ASTRA are assessed based on their technology capabilities, high-value problems addressed, differentiation from existing market solutions and the potential market size. The selected startups will also have opportunities to connect with Applied Ventures’ network of experts and co-investors across leading financial and corporate venture capital firms. Focus areas for ASTRA 2020 included IoT, MEMS, Sensors, Robotics, Optics, Augmented Reality and Virtual Reality, Life Science, Advanced Analytics and Industry 4.0. Applied Materials India has been actively collaborating with Zinnov and C-CAMP as a Knowledge Partner and academic institutes like the IITs to engage with and mentor the startup ecosystem in India. We partner with Endiya Partners to hold an Accelerator event this year and plan to hold ASTRA 2021 later this year. AIM: What are your views on the government’s recent initiatives to promote semiconductor manufacturing in India? Dr Rengarajan: India consumes a lot of electronics, which will only continue to grow. We need to make sure we are in the high value-add portion of this ecosystem, which is the manufacturing segment. We are encouraged to see the Indian government taking steps to develop the country’s semiconductor and high-tech manufacturing ecosystem. The semiconductor industry is highly capital intensive. Like any big industry, a chip plant needs a range of ancillary units nearby to supply parts and service, which are also very high-tech and capital intensive. The government must come up with incentives so that there is a sufficient secondary industry to provide critical services to chip factories. Government support and robust supporting sectors are required to encourage major players to set up chip factories. AIM: What are your thoughts on the global chip shortage? Dr Rengarajan: As the world starts transitioning to the post-pandemic economy, demand for semiconductors continues to grow. The pandemic accelerated key technology trends that make semiconductors more pervasive and indispensable in people’s lives. There is a clear desire for the chip industry to build more resilient and flexible supply, including more regionally distributed capacity as the strategic importance of the semiconductor supply chain is increasingly acknowledged at a national level. AIM: What’s next for Applied Materials? Dr Rengarajan: As mentioned previously, semiconductors are more strategically important to the global economy than at any time in history, driving new waves of silicon consumption. Companies are re-thinking and re-engineering the way they operate; consumers are making different choices about how they spend their time and the products and services they buy. We believe these trends are irreversible. As the industry grows, we must keep an increased focus on ensuring that the growth is sustainable and responsible. Last summer, we announced our new 10-year sustainability roadmap. We’ve taken a holistic approach to sustainability that considers our own operations, our relationships with customers and suppliers, and developing technology that can be used to advance sustainability on a global scale.","excerpt":"Founded in 1967, California-headquartered Applied Materials Inc. is one of the world’s leading semiconductor products and services providers. The company has offices in multiple locations including Europe, Japan, North America, Israel, China, Italy, India, Korea, and Taiwan. Applied Materials established its Indian office in 2003 and since then, emerged as the second-largest resource for engineering […]","categories":["AI Features"],"tags":["Chip shortage","Interviews and Discussions","semiconductor industry"],"author_name":"Shraddha Goled","publish_date":"2021-06-18T16:00:00","publication_year":"2021","word_count":1125,"keywords":["big data","Go","API","TPU","AI","semiconductor industry","RAG","Aim","Chip shortage","analytics","GAN","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","analytics","Aim","RAG","TPU","R","Go","API","big data","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/interview-with-dr-suraj-rengarajan-cto-applied-materials-india\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10119735,"title":"NVIDIA is Hiring AI Engineers in India","content":"NVIDIA is hiring experienced AI engineers in India to join its partner companies. The selected candidates will join the NVIDIA Partner Network as employees and will be responsible for driving the adoption of NVIDIA technology and securing innovative design wins in Data Center, Edge, and Cloud Deployments. Interested candidates can apply here. The roles are based in Bangalore and New Delhi and focus on the position of Deep Learning Solutions Architect. Requirements for the role include a degree in Engineering (BE\/B.Tech\/MS\/MTech), preferably in CS\/IT\/Electrical\/Electronics or equivalent. Candidates must have a comprehensive understanding and hands-on experience in NLP, Deep Learning, and Machine Learning, with optional experience in LLM, Generative AI, and RAG workflows. Proficiency in coding using Python, PyTorch, and TensorFlow is essential, with a bonus for experience in C++ and CUDA. Candidates should also possess a strong grasp of associated software architecture and frameworks, along with effective communication and presentation skills, and the ability to multitask in a dynamic environment. The company is seeking candidates with a proven track record of 2-5 years in writing code in Python, PyTorch, TensorFlow. Experience with C++ and CUDA is a bonus. The ideal candidate will have a thorough understanding of associated software architecture and frameworks. Standout candidates may have a background in customer-facing roles, development experience with NVIDIA software platforms and GPUs, and knowledge of MLOps technologies such as Docker\/containers, Kubernetes, and data center deployments.","excerpt":"Requirements for the role include a degree in Engineering (BE\/B.Tech\/MS\/MTech), preferably in CS\/IT\/Electrical\/Electronics or equivalent.","categories":["AI News"],"tags":["NVIDIA"],"author_name":"Siddharth Jindal","publish_date":"2024-05-07T11:10:11","publication_year":"2024","word_count":232,"keywords":["machine learning","AI","PyTorch","ML","MLOps","RAG","NLP","deep learning","generative AI","NVIDIA","TensorFlow"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","NLP","generative AI","MLOps","TensorFlow","PyTorch","RAG"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/nvidia-is-hiring-ai-engineers-in-india\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10080775,"title":"ShareChat’s Philosophy of Utilising Cloud-Based Systems","content":"The 2019 Indian government ban on TikTok might have been displeasing to millions of users, but it was definitely good news to its competitors. Home-grown short video platforms have grabbed every ounce of attention from the same user base, with ShareChat also joining the very own social platform cluster. Analytics India Magazine spoke to Gaurav Bhatia, SVP of Engineering, ShareChat and Moj, to learn more about how ShareChat has built its cloud-fashioned technology stack. Read: Nothing like TikTok, Kudos for Trying AIM: How has ShareChat built its technology stack in a cloud-first fashion? Gaurav: ShareChat (Mohalla Tech Pvt Ltd) is India’s largest homegrown social media company, with 400+ million MAUs across both its platforms—ShareChat and Moj. We witness over 280 billion views per day, with over 165 million content pieces uploaded daily across both our platforms combined. To manage and scale our services to serve hundreds of millions of users every month, we have built our technology stack in a cloud-first fashion. We run on Google Cloud Platform (GCP) and follow a microservices architecture where business functionality and features are broken down into smaller components that are developed, deployed, and scaled independently. We build and run the microservices on a Kubernetes (k8) cluster hosted in the Google Cloud. Reliance on the Platform-as-a-service model allows us to have central functions that provide best-in-class infrastructure to teams building product functionality. AIM: Could you provide more details on how it uses Google Cloud to follow a microservices architecture? Gaurav: All the services that power our applications can be broken down into small components. For example, when opening ShareChat or Moj, our Android application would contact our Video Feed Service, which would determine the list of best-personalised videos to show the user. The Video Feed Service in turn relies on many services internally to fetch the user details as well as a list of videos and metadata (e.g., likes, comments, shares), which A\/B test experiments bucket a user is a part of, and more. When a user interacts with the content in the form of a successful watch\/like or skip, the events are recorded and sent to an Events service which can then use these signals to better personalise the users’ Video Feed. Along similar lines, when a user uploads a video, it goes through multiple different small services which include the Upload service, Content Moderation, Encoding pipelines, and services. Each one runs in its own Docker container on k8 and can be scaled up or down based on needs and functionality. When new services are built or existing services are updated, they can be rolled out initially to a smaller percentage of users (in the form of a canary deployment) so that any unexpected errors or scenarios can be caught and fixed quickly with minimal impact. The services running in the Kubernetes cluster (k8) rely on Google PaaS services including Pub\/Sub, Dataflow, Spanner, and BigTable. We perform data analytics using BigQuery. AIM: How does ShareChat provide best-in-class infrastructure to teams building product functionality? Gaurav: At ShareChat, we have an internal platform engineering team that performs several actions including building abstractions on top of Google services such as our Database, Queue and Cache Drivers. Services are built and deployed in a CI\/CD model, and we have also built internal tools, such as Atlas, that allow developers to log in and harness the power of the Google Cloud with a few clicks while ensuring that we have enough guardrails to prevent any inadvertent errors. We invest a great deal in test automation and have a device lab with many physical devices which are hosted and can be used remotely to run automation tests for our android and iOS apps. AIM: Has shifting your services to GoLang been beneficial for ShareChat?What are some of the services that you’d like to work on in the future? Gaurav: Shifting from Node.js to GoLang has given us incredible savings, we have cut down on 90% of the infrastructure on many of our services. This has led to not just server cost optimisation but also easier monitoring and maintainability with quick scale-up\/scale-down. We are currently on a journey to rewrite our entire product stack using GoLang. We are constantly evaluating the new services Google is offering. We are also excited about the security offerings as well as the observability and disaster recovery features Google Cloud is adding. In Kubernetes, we are thrilled about eBPF and are actively looking at service mesh. We are also very impressed with the early results we are seeing with the adoption of ScyllaDB, which is giving us excellent low latency and high throughput. AIM: What is the technology used behind ShareChat’s recommendation systems? How would it stand out from its competitors? Gaurav: At ShareChat we have many different kinds of content that include short videos, long-format videos, images, GIFs, microblog posts, and news content. We have invested heavily in building ML infrastructure for rapid personalisation during a session that applies across all content surfaces. Having a common feature store that can be used for quick experimentation across each content surface—which has its own requirements—is a very powerful way to personalise. Additionally, building our Ranker as a service for different contextual needs has served us well. For instance, during festivals users are looking for content that is more suited for sharing compared to devotional content that is for personal consumption on other days.","excerpt":"“Shifting from Node.js to GoLang has given us incredible savings, we have cut down on 90% of the infrastructure on many of our services”, says Bhatia.","categories":["AI Features"],"tags":["Interviews and Discussions"],"author_name":"Bhuvana Kamath","publish_date":"2022-11-25T16:00:00","publication_year":"2022","word_count":897,"keywords":["GCP","AI","ML","recommendation systems","docker","microservices","Aim","analytics","R","kubernetes","Interviews and Discussions"],"extracted_tech_keywords":["AI","ML","analytics","Aim","recommendation systems","GCP","kubernetes","docker","microservices","R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/sharechats-philosophy-of-utilising-cloud-based-systems\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10014789,"title":"Cyberflunk 2020, Big Tech Ad-versity And More In This Week’s Top Tech News","content":"“Cyberflunk”–No, we are not taking a dig at the ill-fated video game. This is something more serious. The world wide cybersecurity systems were reported to have succumbed to a suspected Russian hack that compromised nearly 18,000 companies across the world including the critical government institutions of the United States of America. Okay, we lied! There is something to say about Cyberpunk 2077 as well. The video game which came under attack for its technical glitches were actually a result of the inefficiencies caused due to the remote work at the gaming company. One more week and some more bad news for the internet giants. Google was accused of getting cozy with Facebook for an unfair leverage over global online ad marketing. Read more about this in this week’s top tech news brought to you by Analytics India Magazine. Capsule Networks Make A Comeback Finding the natural parts of an object and their intrinsic coordinate frames without supervision is a crucial step in learning to parse images into part-whole hierarchies. If we start with point clouds, we can do it! https:\/\/t.co\/c2cVm9yq8G— Geoffrey Hinton (@geoffreyhinton) December 10, 2020 Geoff Hinton is back in action and so are Capsule networks.  In this latest work, the researchers propose a capsule architecture trained in an unsupervised fashion — by only observing pairs of randomly rotated 3D point clouds of the same object.  Google Brain’s Andrea Tagliasacch said that this work can enable state-of-the-art results across a number of applications, such as canonicalization (i.e. registration), reconstruction (i.e. auto-encoding), as well as unsupervised classification. Back in September, Hinton and his team filed a patent for Capsule Networks. According which, they claimed that capsule networks can be used in place of conventional convolutional neural networks for traditional computer vision applications. U.S Hit By One Of The Largest Cyber Attacks Solarwinds, a company that develops network monitoring software issued a statement declaring that their systems have come under a cyberattack. This attack, stated the company, was a very sophisticated supply chain attack. The nature of this attack indicates that it may have been conducted by an outside nation state, but SolarWinds has not verified the identity of the attacker. However, Secretary of State Mike Pompeo said that Russia was behind the widespread hacking of government systems. The cyberattack exposed dozens of government and private systems, including nuclear laboratories and the Pentagon, Treasury and Commerce Departments. “I think it’s the case that now we can say pretty clearly that it was the Russians that engaged in this activity,” said Mr. Pompeo. SolarWinds makes network management software, called Orion, that’s widely used by government agencies and Fortune 500 companies. Like most software makers, they push regular updates to their customers.  The suspected Russian hack used what’s called a supply chain attack, exploiting SolarWinds management software updates to put malicious code on the targets’ server. About 18,000 customers downloaded these updates, which acted like Trojan Horses, awaiting instructions from the hackers. One of the US Senators Dick Durbin called this breach “virtually a declaration of war.” Behind The Scenes With Google & Facebook The bad news keeps coming for the big tech. On Wednesday , 10 US states sued Google, accusing the search giant of running an illegal digital-advertising monopoly and enlisting rival Facebook. According to the complaint Google got into an illegal agreement with Facebook in return for guaranteed special treatment in Google-run ad auctions. The lawsuit blames Google for orchestrating an unfair, anti-competitive business practice to maintain its dominance over the ad-industry.  As of 2020, Google is a company standing at the apex of power in media and advertising, generating over $161 billion annually with staggering profit margins, almost all of it from advertising. “For instance, shortly after Facebook acquired WhatsApp, in 2015, Facebook signed an exclusive agreement with Google, granting Google access to millions of Americans’ end-to-end encrypted WhatsApp messages, photos, videos, and audio files,” stated the lawsuit. Google was also blamed for undermining the government’s efforts to safeguard user’s privacy. This suit follows a string of other bad news that Google and other tech companies have received over the past couple of months leading up to what could be a very difficult 2021. Apple Irks Facebook We believe users should have the choice over the data that is being collected about them and how it’s used. Facebook can continue to track users across apps and websites as before, App Tracking Transparency in iOS 14 will just require that they ask for your permission first. pic.twitter.com\/UnnAONZ61I— Tim Cook (@tim_cook) December 17, 2020 Earlier this year, Apple announced that it would be asking its users whether or not if they want their data to be shared for personalised advertising. And, as promised Apple is now forcing companies to inform its users about targeted ads. Apple even took a jibe at Facebook and slammed them for utter “disregard for user privacy.” In response to Apple’s new policy, Facebook, this week, ran a series of full-page newspaper ads, claiming the iPhone maker’s coming mobile software changes around data gathering and targeted advertising are bad for small businesses. “While limiting how personalized ads can be used does impact larger companies like us, these changes will be devastating to small businesses,” Facebook said. In response, Facebook ran a series of full-page newspaper ads, on New York Times, Wall Street Journal and Washington Post, with the headline “We’re standing up to Apple for small businesses everywhere.” Facebook  opined that limiting personalized ads will be devastating to small businesses. Apple, while assuring that Facebook can still track its users, asserts that they only want their customers to make more informed decisions. Whereas, Facebook accused Apple of benefitting from this decision at the expense of small businesses. Cyberpunked?! pic.twitter.com\/kmG7C9qF6h— Elon Musk (@elonmusk) December 14, 2020 One of the most anticipated games of this year– Cyberpunk was met with a worst possible reception. Avid gamers and even Elon Musk trolled the gaming studio for glitches in the game. This was followed by Sony and Microsoft offering full refunds to anyone who used their game stores to buy “Cyberpunk 2077”. “Cyberpunk” was CD Projekt’s costliest project in the company’s two-decade history and has spent nearly $80 million for making and marketing alone. CD Projekt delayed the project three times amid disruptions caused by remote work, but somehow rushed to release this month. One of the reasons behind the glitches can be attributed to the disruption at the workplace for the studio. Executives at CD Projekt said that they initially miscalculated how long “Cyberpunk” would take to complete, with the health crisis most affecting the latter stages of four years of development, following pre-production work that started around 2012. With employees homebound, making even a minor tweak—such as changing the placement of characters or objects in a scene—would take hours instead of minutes. CD Projekt also said that it has invested in AI to create more advanced special effects for this super expensive game. So far, the company has been honest in its addressal to the constant backlash that it has been getting online. The company said that it is committed to assist in hassle free refunds while they fix the glitches. U.S. Blacklists China’s SMIC The United States Department of Commerce has added China’s top chip maker Semiconductor Manufacturing International Corporation (SMIC) to the blacklist that consists of 60 other Chinese institutions. This action, stated the Commerce Dept., stems from China’s military-civil fusion (MCF) doctrine and evidence of activities between SMIC and entities of concern in the Chinese military industrial complex.  Going forward, items uniquely required to produce semiconductors at advanced technology nodes—10 nanometers or below—will be subject to a presumption of denial to prevent such key enabling technology from supporting China’s military-civil fusion efforts. Blue Origin Gets The Nod From NASA Jeff Bezos’ is joining the much anticipated space race with Elon Musk’s SpaceX. This week, NASA awarded Launch Services (NLS) II contract to Bezos’ Blue Origin and their New Glenn launch service. The new contract doesn’t commit NASA to launching any particular mission on New Glenn. Instead, it opens Blue Origin to compete for NASA’s contracts under NASA Launch Services II, which is applicable for launches through December 2027. Blue Origin has also designed a robotic Blue Moon lander to assist NASA in putting humans on the moon again.","excerpt":"“Cyberflunk”–No, we are not taking a dig at the ill-fated video game. This is something more serious. The world wide cybersecurity systems were reported to have succumbed to a suspected Russian hack that compromised nearly 18,000 companies across the world including the critical government institutions of the United States of America. Okay, we lied! There […]","categories":["AI News"],"tags":["audio mining software","big data video games","Blue Origin","Google","what is big data coding"],"author_name":"Ram Sagar","publish_date":"2020-12-19T18:00:00","publication_year":"2020","word_count":1382,"keywords":["Go","Blue Origin","AI","neural network","AWS","ML","R","computer vision","RAG","Aim","analytics","Google","big data video games","audio mining software","what is big data coding"],"extracted_tech_keywords":["AI","ML","neural network","computer vision","analytics","Aim","RAG","AWS","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/cyberpunk-google-facebook-cyberattack-top-news\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10021501,"title":"Complete Guide to ALBERT &#8211; A Lite BERT(With Python Code)","content":"Transformer models, especially BERT transformed the NLP pipeline. They solved the problem of sparse annotations for text data. Instead of training a model from scratch, we can now simply fine-tune existing pre-trained models. But the sheer size of BERT(340M parameters) makes it a bit unapproachable. It is very compute-intensive and time taking to run inference using BERT.ALBERT is a lite version of BERT which shrinks down the BERT in size while maintaining the performance. This model was published in a paper presented at ICLR 2020 by Zhenzhong Lan, Mingda Chen2, Sebastian Goodman, Kevin Gimpel, Piyush Sharma and Radu Soricut (researchers at Google Research and Toyota Technological Institute at Chicago). Link. Architecture of ALBERT The main Idea of ALBERT is to reduce the number of parameters(up to 90% reduction) using novel techniques while not taking a big hit to the performance. Now, this compressed version scales a lot better than the original BERT, improving the performance while still keeping the model small. The backbone of the architecture is the Multi-headed, multi-layer Transformer. This picture taken from http:\/\/primo.ai\/index.php?title=Attention is a great visualization of the transformer model. ALBERT is an encoder-decoder model with self-attention at the encoder end and attention on encoder outputs at the decoder end. It consists of several blocks stacked on top of one another. Each of these blocks contains a multi-head attention block and a Feedforward Network. Following is an excerpt from the implementation of such a block in the source code of ALBERT. Code here is modified for the sake of brevity. def attention_ffn_block(layer_input,hidden_size=768): # Get self attention tensor attention_output = attention_layer(from_tensor=layer_input, to_tensor=layer_input,) # Run a linear projection of `hidden_size` then add a residual # with `layer_input`. attention_output = dense_layer_3d_proj(attention_output, hidden_size,) #Run a feed forward network with two layers attention_output = layer_norm(attention_output + layer_input) intermediate_output = dense_layer_2d(attention_output, intermediate_size) ffn_output = dense_layer_2d(intermediate_output, hidden_size) ffn_output = layer_norm(ffn_output + attention_output) return ffn_output The first zoom-in the image shows the multi head attention layer. Here’s the code implementation of it def attention_layer(from_tensor, to_tensor): # Scalar dimensions referenced here: #   B = batch size (number of sequences) #   F = `from_tensor` sequence length #   T = `to_tensor` sequence length #   N = `num_attention_heads` #   H = `size_per_head` # `query_layer` = [B, F, N, H] q = dense_layer_3d(from_tensor, num_attention_heads, size_per_head) # `key_layer` = [B, T, N, H] k = dense_layer_3d(to_tensor, num_attention_heads, size_per_head) # `value_layer` = [B, T, N, H] v = dense_layer_3d(to_tensor, num_attention_heads, size_per_head) q = tf.transpose(q, [0, 2, 1, 3]) k = tf.transpose(k, [0, 2, 1, 3]) v = tf.transpose(v, [0, 2, 1, 3]) # 'new_embeddings = [B, N, F, H]' new_embeddings = dot_product_attention(q, k, v, attention_mask, attention_probs_dropout_prob) return tf.transpose(new_embeddings, [0, 2, 1, 3]) The second zoomin picture just shows how the dot product for attention is calculated.Given triples of (Q,K,V) for each sequence term we calculate a weighted sum of the values of all the V’s.These weights here are the dot products of K’s of all sequence terms and the current sequence term.They represent relevance of the sequence terms. def dot_product_attention(q, k, v): logits = tf.matmul(q, k, transpose_b=True)  # [..., length_q, length_kv] logits = tf.multiply(logits, 1.0 \/ math.sqrt(float(get_shape_list(q)[-1]))) attention_probs = tf.nn.softmax(logits, name=\"attention_probs\") return tf.matmul(attention_probs, v) There are few twists to the architecture mentioned in the case of ALBERT. Following are the techniques that ALBERT uses to achieve compression. Factorization of Parameters HIdden layer representations must be large to accommodate the context information along with the word level embedding information. But if we increase the hidden layer size this increases the number of parameters that blows up. If V is the number of tokens in the vocabulary, H is the hidden layer size then we would need the number of parameters to be of the order V*H. ALBERT factorizes these word-level input embeddings into lower dimensions. Let’s say E is the size of embedding after factorization. Now the number of parameters needed would be of the order V*E + E*H. Since V is very large in natural languages this results in a reduction of parameters by a huge margin. This can be very easily implemented # E=input_width; H=hidden_size if input_width != hidden_size: next_layer = dense_layer_2d(input_tensor, hidden_size, create_initializer(initializer_range), None, use_einsum=use_einsum, name=\"embedding_hidden_mapping_in\") Cross Layer Parameter sharing Stacking independent layers although increases the learning capacity of the models, greatly increases the redundancy. Different Layers often learn the parameters that perform the same operation. ALBERT tackles this redundancy by sharing the parameters between groups of layers. This reduces the number of total parameters while keeping the number of layers constant. Tensorflow’s Variable scope can be used with get_variable() to implement this. for layer_idx in range(num_hidden_layers): group_idx = int(layer_idx \/ num_hidden_layers * num_hidden_groups) with tf.variable_scope(\"group_%d\" % group_idx): #...Code for implementing the layer.... w = tf.get_variable(name=\"kernel\") # w is the shared parameter across the current group #...Code for implementing the layer.... In addition to reducing the number of parameters this cross-layer variable sharing also has a nice effect of stabilizing the model. Inter sentence coherence loss This loss isn’t used to directly reduce the number of parameters. It’s used to improve the performance of the representations in the downstream tasks.BERT model is pre-trained for the task of NSP(next sentence prediction). We provide the encoder-decoder model with pairs of segments and make the model predict if they are positive or negative Sentence 1Sentence 2positiveSentence 2Sentence 3positiveSentence 3Sentence 7negativeSentence 1Sentence 3negativeNext Sentence Prediction(NSP) This task, although forcing the model to learn representations that perform well, turned out to be not a difficult task for the model. Model kind of captures the topic information between the sentences and predicts the class. This resulted in the model trained for NSP to struggle at predicting the sentence order.ALBERT uses the following data for training which causes the model not only to learn the topic information but also granular level details and coherence between sentences.. Sentence 1Sentence 2positiveSentence 2Sentence 1negativeSentence Order Prediction(SOP) Usage of ALBERT Tensorflow hub has made it extremely easy to use pre-trained models. Let’s use pretrained ALBERT base model for the classification of movie reviews. albert_url='https:\/\/tfhub.dev\/tensorflow\/albert_en_base\/2' encoder = hub.KerasLayer(albert_url) preprocessor_url=\"https:\/\/tfhub.dev\/tensorflow\/albert_en_preprocess\/3\" preprocessor = hub.KerasLayer(preprocessor_url) Model and the required preprocessor are downloaded and loaded. text_input = tf.keras.layers.Input(shape=(), dtype=tf.string) encoder_inputs = preprocessor(text_input) outputs = encoder(encoder_inputs) pooled_output = outputs[\"pooled_output\"] embedding_model = tf.keras.Model(text_input, pooled_output) Just like that, we have our embedding layer. We just need to build a Fully connected neural network to predict whether a movie review is positive or negative. model = tf.keras.Sequential() model.add(embedding_model) model.add(tf.keras.layers.Dense(128, activation='relu')) model.add(tf.keras.layers.BatchNormalization()) model.add(tf.keras.layers.Dense(30, activation='relu')) model.add(tf.keras.layers.BatchNormalization()) model.add(tf.keras.layers.Dense(8, activation='relu')) model.add(tf.keras.layers.BatchNormalization()) model.add(tf.keras.layers.Dense(1)) model.compile(optimizer='adam', loss=tf.keras.losses.BinaryCrossentropy(from_logits=True), metrics=['accuracy']) history = model.fit(train_data.shuffle(10000).batch(128), epochs=10, validation_data=validation_data.batch(128), verbose=1) Results With this basic model validation accuracy, about 75% is a good number. Especially when we are not fine-tuning the embeddings at all. We can fine-tune the embeddings by just making the encoder trainable. encoder = hub.KerasLayer(albert_url,trainable=True) Here’s a link to the colab notebook with the code in this section Github Repository Link Conclusion ALBERT is a very useful variant of BERT which is not huge.It can improve the efficiency of the performance of downstream language understanding tasks while keeping the computational overhead under an acceptable level for several applications.","excerpt":"ALBERT is a lite version of BERT which shrinks down the BERT in size while maintaining the performance.","categories":["AI Trends"],"tags":["ALBERT","attention mechanism","BERT Model","NLP","pre-trained models","self attention models","Tensorflow","tensorflow hub","Transfer Learning","Transformers"],"author_name":"Pavan Kandru","publish_date":"2021-03-06T16:00:00","publication_year":"2021","word_count":1168,"keywords":["TPU","Scala","Transfer Learning","R","NLP","Tensorflow","Go","AI","neural network","BERT Model","pre-trained models","Keras","tensorflow hub","self attention models","Transformers","Colab","ALBERT","TensorFlow","attention mechanism"],"extracted_tech_keywords":["AI","neural network","NLP","TensorFlow","Keras","Colab","TPU","R","Go","Scala"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/complete-guide-to-albert-a-lite-bertwith-python-code\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":66793,"title":"Top Colab Notebooks To Kickstart Your Machine Learning Journey","content":"Colaboratory, or “Colab” for short, is a product from Google Research, which is technically a hosted Jupyter Notebook service that requires no setup to use. With Google’s Colab, one can access its GPUs and TPUs almost for free. Now with its new paid version Colab pro, the machine learning practitioners can stay connected for up to 24 hours, compared to the 12 hours in the free version of Colab notebooks. One can also get priority access to high-memory virtual machines, which has double the memory of standard Colab VMs and has twice as many CPUs. In the next section, we look at a few illustrative examples on Colab notebooks that can kick start your machine learning journey in a couple of steps. DeepDream This tutorial contains an implementation of DeepDream, an experiment that visualizes the patterns learned by a neural network. DeepDream looks at patterns in an image and over-interprets and enhances them. To do this, it feeds the image through the network, then calculates the gradient of the image with respect to the activations of a particular layer. The image then undergoes modification to increase these activations, enhancing the patterns seen by the network, thereby resulting in a dream-like image. Colab link. Lucid Lucid is a collection of infrastructure and tools for research in neural network interpretability. This tutorial quickly introduces Lucid, a network for visualizing neural networks. Lucid is a kind of a step up from DeepDream and provides flexible abstractions so that it can be used for a wide range of interpretability research. Start visualizing neural networks with no setup from your browser using Colab. Colab link GANs On TensorFlow From creating million-dollar art auctions to making up fake presidents, GANs have been quite popular for the past couple of years. This notebook consists of GANs, and its many variants that have been developed over the years such as ‘DCGAN’, ‘WGAN’, ‘WGAN-GP’, ‘LSGAN’, ‘SNGAN’, ‘RSGAN’&’RaSGAN’, ‘BEGAN’, ‘ACGAN’, ‘PGGAN’, ‘pix2pix’, ‘BigGAN’. Colab link Getting Started With PyTorch Pytorch is one of the hottest machine learning tools today. Its rise in popularity can be attributed to the fact that many AI researchers are imbibing this framework by Facebook for their research. In the Colab repo, you’ll find many basic neural network operations using PyTorch packages. Colab repo link Movie Review Sentiment with BERT on TF Hub BERT’s has been added to TF Hub as a loadable module, it’s easy to add into existing TensorFlow text pipelines. In an existing pipeline, BERT can replace text embedding layers like ELMO and GloVE. Alternatively, fine-tuning BERT can provide both an accuracy boost and faster training time in many cases. Using this notebook, one can train a model to predict whether an IMDB movie review is positive or negative using BERT in TensorFlow with the TF hub. Colab link Detectron 2 Beginner Tutorial Detectron2 was developed by Facebook AI Research to implement state-of-the-art object detection algorithms. In this official Colab tutorial of Detectron2, one can get familiarise with some basics usage of Detectron2, including running inference on images or videos with an existing Detectron2 model. Colab link DeOldify This project allows one to colorize and restore old images and film footage. This project makes use of NoGAN. Firstly, the generator is trained in a conventional way by itself with just the feature loss. Next, images are generated from that, and then a binary classifier is used to distinguish between those outputs and real images. A typical GAN working principle of generator and discriminator. Colab repo link. GPT-2: Text Generation This notebook by Max Woolf walks through how to retrain an advanced text generating neural network on any text dataset for free on a GPU using Colaboratory with `gpt-2-simple`, a simple Python package that wraps existing model, fine-tunes and generates scripts for OpenAI’s GPT-2 text generation model (specifically the “small” 124M and “medium” 355M hyperparameter versions). Colab link. For more Colab notebooks, check this Github compilation.","excerpt":"Colaboratory, or “Colab” for short, is a product from Google Research, which is technically a hosted Jupyter Notebook service that requires no setup to use. With Google’s Colab, one can access its GPUs and TPUs almost for free. Now with its new paid version Colab pro, the machine learning practitioners can stay connected for up […]","categories":["AI Trends"],"tags":["Colab","Jupyter Notebook","Machine Learning"],"author_name":"Ram Sagar","publish_date":"2020-06-05T18:30:19","publication_year":"2020","word_count":652,"keywords":["machine learning","Jupyter Notebook","OpenAI","AI","neural network","PyTorch","TPU","Machine Learning","Colab","Jupyter","object detection","TensorFlow"],"extracted_tech_keywords":["AI","machine learning","neural network","OpenAI","TensorFlow","PyTorch","Jupyter","Colab","object detection","TPU"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/colab-notebook-machine-learning\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10102449,"title":"AMD Formally Launches Ryzen 5 7545U Processor with Zen 4c","content":"AMD has introduced the Ryzen 5 7545U processor as a part of the Ryzen 7040U family, which is its bestselling GPU, featuring the Zen 4c architecture, a variant derived from the Zen 4 design. This new architectural iteration, termed Zen 4c, shares similarities with Zen 4 but focuses on optimising core density, resulting in reduced die space consumption. The primary benefits of this architecture include accommodating more cores and enhancing power efficiency, as AMD outlines its objectives for Zen 4c. Zen 4c’s key design characteristics are notable, with a Zen 4 core occupying 3.84mm2 of die space, while a Zen 4c core requires only about 2.48mm2. This reduction in size amounts to approximately 35 percent less area. Zen 4c utilises half the L3 cache compared to Zen 4, but the primary reason behind its space-saving capability is the lower maximum clock speed. This reduced frequency mitigates interference and power leakage, enabling the chipmaker to optimise the core layout for density. The increased density of Zen 4c contributes to better latency and heightened power efficiency. AMD aims to retain the same instruction set and IPC as Zen 4 while using less power, ultimately delivering superior performance at power levels below 15W. Key specifications of the AMD Ryzen 5 7545U: Manufactured on TSMC’s 5nm node. Comprises 2x Zen 4 cores and 4x Zen 4c cores. Prioritizes Zen 4 cores in the OS scheduler to achieve a higher absolute frequency. Offers a total of 6 cores and 12 threads. Supports a maximum clock speed of 4.9GHz and a base clock speed of 3.2GHz. Operates within a 15-30 TDP (Thermal Design Power) range. Equipped with a Radeon 740M integrated GPU for graphics processing. AMD’s strategy involves deploying smaller Zen 4c cores in premium laptops with increased core counts, as well as entry-level laptops, where these cores provide the same IPC but with a more compact design. As of now, AMD has not disclosed specific details regarding the availability and pricing of the new Ryzen 5 7545U processor.","excerpt":"Zen 4c utilises half the L3 cache compared to Zen 4, but the primary reason behind its space-saving capability is the lower maximum clock speed.","categories":["AI News"],"tags":["AMD AI"],"author_name":"Mohit Pandey","publish_date":"2023-11-03T13:24:27","publication_year":"2023","word_count":333,"keywords":["programming_languages:R","AI","CuPy","Aim","AMD AI","R"],"extracted_tech_keywords":["AI","Aim","CuPy","R","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/amd-formally-launches-ryzen-5-7545u-processor-with-zen-4c\/","complexity_score":4,"technical_depth":5,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10163105,"title":"Is Mistral’s Le Chat Truly the &#8216;World’s Fastest AI Assistant&#8217;?","content":"French AI startup Mistral unveiled the Le Chat app for iOS and Android a few days ago. The app functions as an AI chatbot or assistant, rivalling ChatGPT, Claude, and Gemini, among others. The app offers most of its features for free, with upgraded limits in the Pro tier, which costs $14.99 monthly. Le Chat offers web search, image and document understanding capabilities alongside code interpretation and image generation. Given the sheer number of AI assistant applications in the market, a new entrant must offer a formidable differentiation. Mistral claims its low-latency models are powered by the ‘fastest inference engines on the planet’. Furthermore, Mistral also says that it responds faster than any other chat assistant, up to 1,100 words per second, via their Flash Answers feature. Thanks to Cerebras Cerebras Inference, a service that delivers high-speed processing to AI applications, is the secret sauce to its speed. According to the company, Cerebras Inference is the ‘world’s fastest AI inference provider’ and makes Le Chat 10 times faster than GPT-4o, Claude Sonnet 3.5, and DeepSeek R1. Cerbras also revealed that the 123 billion-parameter Mistral Large model is behind Le Chat. Mistral and Cerebras compared Le Chat with Claude 3.5 Sonnet and ChatGPT-4o using a prompt to generate a snake game using Python. The results from Mistral’s YouTube video revealed that ChatGPT outputs 85 tokens per second, Claude 120 tokens per second, and Le Chat outperformed the two with 1,100 tokens per second. In a video by Cerebras, it was revealed that Le Chat took 1.3 seconds to complete the task, Claude 3.5 Sonnet took 19 seconds, and GPT-4o took 46 seconds. “This performance is made possible by the Wafer Scale Engine 3’s SRAM-based inference architecture in combination with speculative decoding techniques developed in collaboration with researchers at Mistral,” said Cerebras in a blogpost. Several users also resonate with these claims. A user named Marc on X said that the model is “mind-blowingly fast” and added that it built a simple React application in less than 5 seconds. Ok Le Chat by @MistralAI is 10x faster than ChatGPT. Maybe 100x.— Pol Maire (@polmaire) February 6, 2025 Here’s What We Found in Our Real-World Tests Besides, we at AIM also conducted a real-time test of some of the leading models, albeit with a different prompt, which expects AI models to solve a Chemistry numerical problem from an IIT-JEE question paper – which is often considered one of the world’s most difficult examinations. We considered OpenAI’s GPT-4o, o3 Mini, o3 Mini High, Anthropic’s Claude 3.5 Sonnet, DeepSeek R1, Google’s Gemini 2.0 Flash, and of course, Mistral’s Le Chat. The following question was used as input: “Ice at –10°C is to be converted into steam at 110°C. The mass of ice is 10-3 kg. What amount of heat is required?” While we timed the results, Mistral’s Le Chat was the fastest model, but it comes with a caveat. Le Chat returned the output in less than 4 seconds in three out of six times we tested the model. On the other hand, Google’s Gemini 2.0 Flash returned the output under 6 seconds all the times we tested it. It begs the question of whether Flash Answers came into action each time despite being enabled by default. Note that we were using the free version of the Le Chat assistant, and the pro version provides an upgraded limit to the Flash Answers feature. Moreover, the speeds at which these models perform also depend upon the nature of the queries. Reasoning models, with their lengthy chain of thoughts, prioritise the accuracy of the answer and are bound to take more time. For instance, when we tested the prompt with DeepSeek R1, it took over a minute to complete the problem, with a chain of thoughts that involved verification steps, where the model said, “But let me check if all the values are correct. Did I use the right specific heat for steam?” and so on. Furthermore, it also took a great deal of time to ensure the answer was provided with the right number of decimal figures. A test from Artificial Analysis revealed that OpenAI’s o3-mini was the fastest model among the competition, which outputs 214 tokens per second, ahead of the o1-mini, at 167 tokens per second. According to Artificial Analysis, o3-mini also achieved a high score of 89 on its Quality Index, which is on par with o1 (90 points) and DeepSeek R1 (89 points). This quality index quantifies the overall capabilities of the AI model. OpenAI has prioritised inference time scaling to deliver outputs at higher speeds. With Cerbreas’ inference capabilities, Mistral seems to have joined the race. Moreover, there is an ongoing battle of token speeds between inference providers like Cerebras, Groq, and SambaNova. These ambitions to deliver high-speed responses align with what Jensen Huang, CEO of NVIDIA, said last year. He envisioned a future where AI systems perform various tasks, such as tree search, chain of thought, and mental simulations, reflecting on their own answers and responding in real-time—within a single second.","excerpt":"Le Chat’s Flash Answers is using Cerebras Inference, which is touted to be the ‘fastest AI inference provider’.","categories":["Global Tech"],"tags":["AI (Artificial Intelligence)","mistral"],"author_name":"Supreeth Koundinya","publish_date":"2025-02-10T12:49:52","publication_year":"2025","word_count":841,"keywords":["Anthropic","ChatGPT","Gemini 2.0","OpenAI","AI","GPT-4o","DeepSeek R1","Aim","mistral","chain of thought","Claude 3.5","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","GPT-4o","ChatGPT","OpenAI","Claude 3.5","Anthropic","Gemini 2.0","DeepSeek R1","Aim","chain of thought"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/is-mistrals-le-chat-truly-the-worlds-fastest-ai-assistant\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":21321,"title":"Is DeepMind’s AlphaGo Zero Really A Scientific Breakthrough And A Step Towards General Intelligence?","content":"Ever since Demis Hassabis’s talk on Learning from First Principles at NIPS 2017 held last December, the internet is abuzz whether the Google-owned, London-headquartered DeepMind is close to solving the General Intelligence puzzle which have the company has been pursuing since it was founded in 2010. The company behind Atari and Go AI systems reached another milestone in its journey with AlphaGo Zero which evidently learns from scratch, requires no bootstrapping from human data and learns incrementally from its own mistakes. A recent paper published on December 5, 2017 Mastering Chess and Shogi by Self-Play with a General Reinforcement Learning Algorithm talks about the accomplishments of AlphaGo Zero program that achieved superhuman performance in the game of Go, by reinforcement learning from games of self-play. The paper generalized the approach into a single AlphaZero algorithm that can achieve superhuman performance in many challenging domains. Starting from random play, and given no domain knowledge except the game rules, AlphaZero achieved within 24 hours a superhuman level of play in the games of chess and shogi (Japanese chess) as well as Go, and convincingly defeated a world-champion program in each case. Is AlphaGo Zero a scientific breakthrough – step towards AGI? Interestingly, Tesla’s Director of AI and Autopilot Vision Andrej Karpathy thinks progress towards AGI is a step function and will happen either suddenly or unexpectedly, he shared at the recent Fireside Talk at NIPS 2017. “If you look at the progress of AlphaGo Zero, there was a long period of quiet engineering effort and algorithmic advances, then reached superhuman status in just three days,” shared Karpathy, when quizzed about his views on AGI. A section of ML researchers dubbed AlphaGo Zero as a significant research advancement in AI, even more important than Alpha Go since Alpha Go Zero learned without any data. Now, Deep Learning systems are panned for their data dependence and brute force nature of achieving optimum results. However, AlphaGo, on the other hand does not rely on human generated data and may discover new knowledge, which can result in artificial general intelligence (AGI). The end goal of any AI company is to build models that discover new knowledge on their own. Such models do not rely too much on human generated data to start with. There were two main ideas that AlphaGo clearly emphasized, noted Demis Hassabis during his talk: Intuition: Implicit knowledge acquired through experience but not consciously expressible. The quality of this knowledge can be verified behaviourally Creativity: The ability to synthesize knowledge to produce a novel or original idea, AlphaGo clearly demonstrated these abilities, although in a constrained domain AlphaGo Zero used a Deep Reinforcement Learning approach and a tree-based search strategy. Here are some of the differences from previous approach: The architecture used one deep neural network (DNN) instead of the two separate policy and value networks. There was a simpler tree search strategy No bootstrapping from human data Fully automated pipeline with no human in the loop Zero human generated data for training and learnt incrementally from its own mistakes It started from completely random play with zero knowledge and played against itself millions of times So Alphago Zero plays games against itself at current full strength Here’s what makes AlphaGo Zero really General Besides the human-like ingenuity displayed in learning, the reason why AlphaGo Zero is a big step forward is because the system got rid of the supervision and feature engineering. The next logical step was to replace MCTS with a differentiable recurrent model to build an end-to-end trainable system that doesn’t utilize simulations. This step made the system truly general. According to experts, the high-level of engineering, self-play factor known as co-evolution pushed the model to a superhuman level. Nevertheless, it is a generalized learning network that still has to be trained to specialize in certain field. Role of Deep Reinforcement Learning in achieving Strong AI When it comes to building machines that think and learn like humans, Deep Reinforcement Learning is perhaps viewed as the most plausible paths. Today, Deep Reinforcement Learning is one of the most active research areas in artificial intelligence – essentially it is a computational approach to learning whereby an agent tries to maximize the total amount of reward it receives while interacting with a complex, uncertain environment, explains Dr Richard Sutton, known as one of the founding fathers of computational reinforcement learning. At NIPS 2017, Deep Reinforcement Learning was the most popular topic and DeepMind has delivered great results with AlphaGo Zero which plays at superhuman level. During the recently concluded NIPS the DeepMnd released an updated version which plays Go, Shogui, and Chess at a dominant level. Another example of Deep RL is Deep Q-learning — a model-free reinforcement learning algorithm used to train deep neural networks on controlled tasks such as playing Atari games. A network is trained to approximate the optimal action-value function Q(s, a), which is the expected long-term cumulative reward of taking action a in state s and then optimally selecting future actions, explains the paper Building Machines That Learn and Think Like People. According to Hassabis, co-founder and CEO of DeepMind, grounded learning in intuitive theories of physics and psychology could significantly support the training and help generalize knowledge to new tasks and situations. so, instead of building systems that simply recognize handwritten characters and play Go or Atari, deep learning should tackle tasks with little training data and also evaluate models on a range of human-like generalizations. In his talk, Learning from First Principles NIPS, Hassabis outlined his strategy to build intelligent systems: a)  Learning vs Handcrafted: we want systems to learn themselves directly rather than being spooned and handcrafted with solutions programmed b)  General vs Specific: we wanted a system that was able to run across a wide range of environments and tasks and not just do one problem or one task. c)   Grounded vs Logic based systems: true thinking system or a cognitive system should be fully grounded in reality d)  Active vs Passive: Notion of active vs passive is that we strive to create agent-based systems that are active participants in their own learning","excerpt":"Ever since Demis Hassabis’s talk on Learning from First Principles at NIPS 2017 held last December, the internet is abuzz whether the Google-owned, London-headquartered DeepMind is close to solving the General Intelligence puzzle which have the company has been pursuing since it was founded in 2010. The company behind Atari and Go AI systems reached […]","categories":["IT Services"],"tags":["Deep Learning","DeepMind AI","deepmind founder Demis Hassabis","deepmind london","Go"],"author_name":"Richa Bhatia","publish_date":"2018-02-05T04:33:06","publication_year":"2018","word_count":1020,"keywords":["Go","DeepMind AI","deepmind founder Demis Hassabis","artificial intelligence","AI","neural network","RPA","ML","feature engineering","deep learning","ViT","Deep Learning","R","deepmind london"],"extracted_tech_keywords":["AI","artificial intelligence","ML","deep learning","neural network","R","Go","feature engineering","ViT","RPA"],"url":"https:\/\/analyticsindiamag.com\/it-services\/deepminds-alphago-zero-really-scientific-breakthrough-step-towards-general-intelligence\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":57442,"title":"Top Benefits Of No-Code Tools For Startups","content":"Humans always tend to find ways to make complicated things easier and more accessible, and the same goes with the complicated landscape of artificial intelligence and machine learning. The most complicated aspect of these newer technologies is the coding part, where one needs to be completely skilled with programming languages and coding in order to create applications. However, recently there is a massive buzz around ‘no-code’, which means no programming code — where one requires no knowledge of programming languages or the aspect of coding to create applications. Typically, a no-code tool works on the concept of plug and play, where it provides a library of functional components to be implemented in visual flow. These DIY platforms have several contrasts, and therefore more prominent companies are still hiring expert coders to handle their most intensive developing projects, however, for startups, the concept of no-code is like a revolution. In this article, we would talk about a few ways no-code tools can help startups and small businesses in creating software and applications on their own. Easy Access For Non-Coders & Non-Developers To create software or an application, an organisation would need a developer who is skilled with programming codes, however startups and many smaller businesses, usually, don’t have a technical team with developers and neither have the budget to an employee or create one for their organisation. For such businesses no-code tools can be beneficial as it provides easy access. The no-code tools are usually designed with non-coders and non-developers in mind, where businesses can use the plug and play model of the no-code tools to develop their applications. Such tools come handy for people who aren’t aware of CSS, HTML or JS. The no-code tasks also include things like security vulnerability scanning, testing, integrations, hosting, disaster recovery, analytics, etc. which helps non-technical folks with their process of building applications. Efficient Time To Market Timeliness is an essential factor for startups to stay relevant. A no-code software is not only easy to use but also a way faster and efficient process than a manual coder programming the whole application. While bigger companies can always have a longer timeline for developing new products, startups are always looking for quick turnaround time, and these, no-code tools can help in building applications in almost no time. With no-code, any maker can experiment with rapid prototyping and can develop a solution to a problem in just a few hours. Promisingly Cost-Saving As mentioned before, hiring an expert coder or a developer is a costly business for startups. If a business application isn’t very complicated, a no-code software for non-developers can be handy and will come at a fraction of the cost of hiring a developer. As the cost of the development is comparatively less, the businesses then can try, test and pursue different opportunities before coming to a complete solution. Learn To Think Like A Programmer No code tools are designed for non-coders and non-developers. However, if those non-technical people are enthusiastic about the field, these no-code tools can help them build their knowledge base and can help them think like a programmer. According to experts, anyone can now become a programmer with no-code tools. These tools help those non-technical folks to express themselves, while developing an application, in a way that has never been possible before. One doesn’t have to be a programmer to build things on the internet; these no-code tools are empowering the new generation of makers from diverse backgrounds and perspectives. Easily Changeable Apart from being easy to use, no-code tools support seamless integration, which allows the app to work in harmony with the existing system. Also, with traditional coding, it gets difficult for developers to change a functionality or a feature instantly, especially when the coding language isn’t known to the person handling it. However, a no-code development platform is easily changeable and only requires a new logic, and within a few hours, anybody can make implement that change. No-Risk Of Shadow IT In an organisation, a no-code platform will usually be centrally governed by the IT department, which in turns diminishes the risk of shadow IT. Employees, sometimes, set up their servers or store sensitive information on their accounts, which in turn can create a threat for the whole organisation. And that’s when the risk of shadow IT occurs, by not monitoring development activity outside of IT departments. Considering no-code platforms are supported by the central IT department, data security and privacy are guaranteed for the organisation. The robust governance strategies safely help the employees develop their applications from the web browser. This also brings down the debugging to a bare minimum.","excerpt":"Humans always tend to find ways to make complicated things easier and more accessible, and the same goes with the complicated landscape of artificial intelligence and machine learning. The most complicated aspect of these newer technologies is the coding part, where one needs to be completely skilled with programming languages and coding in order to […]","categories":["AI Trends"],"tags":["AI Benefits","Coders","coding platform","Developers","low code no code platforms"],"author_name":"Sejuti Das","publish_date":"2020-02-26T09:00:00","publication_year":"2020","word_count":773,"keywords":["Go","API","artificial intelligence","machine learning","low code no code platforms","ML","AI Benefits","Coders","ViT","analytics","GAN","coding platform","R","Developers","startup"],"extracted_tech_keywords":["artificial intelligence","machine learning","ML","analytics","R","Go","API","GAN","ViT","startup"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-benefits-of-no-code-tools-for-startups\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10103850,"title":"Jensen Huang Brings re:Invent to Life","content":"Jensen Huang is everywhere, and so is NVIDIA. The company seems to be stealing the spotlight at all the major events – be it Google Cloud Next, AWS re:Invent or Microsoft Ignite, bringing the party to life at each one of them. During the recent re:Invent keynote, AWS and NVIDIA jointly announced a strategic initiative to provide a new class of supercomputing infrastructure, software, and services tailored specifically for generative AI. The duo have also decided to deploy NVIDIA’s much-anticipated GH200 chips, initially slated for release in 2024. The installation of NVIDIA’s GH200 chips will occur within AWS’s cloud infrastructure, emphasising the global availability of this advanced hardware for AWS customers. NVIDIA x AWS The event also presented the initiative it has taken to set up the world’s fastest GPU-powered AI supercomputer, which will be a giant leap towards reshaping industries and driving technological progress at an unprecedented pace. This innovation is named Project Cieba and will aim to feature 16,384 NVIDIA GH200 superchips, which will process a staggering 65 exaflops of AI, propelling NVIDIA’s next wave of generative AI innovation. At #AWSreInvent, @AWSCloud CEO Adam Selipsky and our CEO Jensen Huang spotlight the pivotal role of #generativeAI in cloud transformation, highlighting their companies’ growing partnership. https:\/\/t.co\/TrmbOu3GXw— NVIDIA (@nvidia) November 28, 2023 Apart from this, AWS is also working with NVIDIA on introducing three new Amazon EC2 instances, including P5e instances for large-scale generative AI and HPC workloads and G6 and G6e instances for a wide range of applications, such as AI fine-tuning, inference, and graphics. “NVIDIA and AWS are collaborating across the entire computing stack, spanning AI infrastructure, acceleration libraries, foundation models, to generative AI services,” said the CEO of NVIDIA, Jensen Huang. NVIDIA x Hyperscalers Clearly, NVIDIA thrives in a collaborative environment, and at AWS re:Invent, it became clearer, as it also shares partnerships with rivals Google Cloud, Microsoft Azure and Oracle. “Our partnership with NVIDIA spans every layer of the Copilot stack — from silicon to software — as we innovate together for this new age of AI,” said Microsoft chief Satya Nadella, at Ignite 2023. At this event, NVIDIA and Microsoft announced their partnership to launch an AI foundry service on Microsoft Azure, aiming to boost the development of custom generative AI applications for enterprises and startups. This service integrates NVIDIA’s AI technologies and DGX Cloud AI supercomputing with Azure’s infrastructure, providing a comprehensive solution for creating and deploying tailored AI models. The partnership also emphasised custom model development, leveraging NVIDIA’s AI Foundation Models and tools and making these advancements accessible through Azure’s cloud platform and marketplace. This collaboration signifies a major step in facilitating advanced AI application development and deployment in various industries. “Many of Google’s products are built and served on NVIDIA GPUs, and many of our customers are seeking out NVIDIA accelerated computing to power efficient development of LLMs to advance generative AI,” shared Google Cloud chief Thomas Kurian at the Next event held mid-this year. At Google Cloud Next, NVIDIA partnered with Google to drive advancements in AI computing, software, and services, alongside enhancing AI supercomputing capabilities. The duo is working together to optimise Google’s PaxML for NVIDIA GPUs, facilitating large language model development, and integrating serverless Spark with NVIDIA GPUs for accelerated data processing. In addition to this, Google Cloud had said to feature NVIDIA H100 GPUs in its A3 VMs and Vertex AI platform and gain access to the NVIDIA DGX GH200 AI supercomputer and more. Recently, Oracle also announced a multi-year partnership with NVIDIA to speed up the AI adoption for enterprises, which helps customers solve business challenges. In a recent interview with AIM, Oracle said that it is well-equipped in terms of infrastructure, as NVIDIA selected OCI as the first hyper-scale cloud provider to offer NVIDIA DGX Cloud. “When NVIDIA thinks of cloud and data, they think of Oracle,” said Oracle’s Chris Chelliah. He said that it utilises MySQL HeatWave data for real-time anomaly detection on NVIDIA clusters for its customers. NVIDIA is Omnipresent NVIDIA’s diverse partnerships with leading cloud providers and hyperscalers uniquely position it across various facets of the AI landscape. While everyone is busy building their in-house silicon capabilities to handle AI workloads, the nature of partnerships seems to be changing rapidly. From an innovation and AI advancements standpoint, Google Cloud seems to be NVIDIA’s favourite, while Microsoft Azure stands a pivotal partner for enterprise reach and application development, given its strong enterprise focus and extensive customer base. Oracle differentiates itself in data management and AI-driven solutions, particularly through its emphasis on real-time data processing capabilities. AWS, on the other hand, plays a critical role in security-focused AI solutions, addressing the increasing concerns around AI security and reliability. Overall, these partnerships provide NVIDIA with a multifaceted platform to expand its AI capabilities and market reach, with the impact of each partnership aligning with NVIDIA’s strategic focus areas, whether it be AI innovation, enterprise application, data management, or ensuring security in AI solutions. Simply put, everybody likes to NVIDIA.","excerpt":"Everybody likes to NVIDIA.","categories":["AI Features"],"tags":[],"author_name":"Sandhra Jayan","publish_date":"2023-11-29T17:30:36","publication_year":"2023","word_count":833,"keywords":["AWS","AI","ML","serverless","RAG","Aim","anomaly detection","generative AI","foundation models","Azure"],"extracted_tech_keywords":["AI","ML","generative AI","foundation models","Aim","RAG","anomaly detection","AWS","Azure","serverless"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/jensen-huang-brings-reinvent-to-life\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":68323,"title":"Is Landing A Data Science Job Amid Covid-19 Challenging? Companies Weigh In","content":"As some offices cautiously reopen and resume work, employees are finding themselves in an altered workplace. While tangible measures to enforce social distancing protocols have become a fixture across offices, the fact that fewer employees remain to join work post the lockdown is not lost on anyone. The uncertainty caused by the Covid-19 pandemic has wreaked havoc in the global economy, accelerating an oncoming recession, leading to loss of jobs across industries. However, despite that, some believe that jobs anchored around emerging technologies like AI, big data, etc are still available. If it is true, then landing a data science job amid the Covid-19 pandemic should not be challenging. Companies operating in the space weigh in: Increased Interest In Data Amid Covid-19 The application of AI and related technologies may have been approached with some trepidation by many, but it seems to have gained wider acceptance amid the Covid-19 pandemic. Not only advancements in research helped manage the crisis better, businesses have also relied on these technologies to pivot and manage resources as they navigate the effects of the pandemic. “The Covid-19 crisis and associated challenges are being fought on the basis of numbers and data, and this has renewed interest in data science, artificial intelligence and machine learning,” says Kapil Bardeja, CEO and co-founder of Vehant Technologies. According to him, the data science job segment has been relatively less affected than others, which has generally been marked by hiring freezes, mass layoffs and salary cuts. “I believe data science jobs will continue to be sought after despite the downturn as companies need the skill sets they possess now more than ever,” he says. “Given that most companies have to actively seek out data-driven solutions to overcome some of these challenges, and given that the pandemic has widened the scope and understanding of these new technologies, data science hires will continue,” he adds. ALSO READ: Top Resources To Find Data Science Jobs During The Recession In fact, according to Sumit Mittal, CEO of VentAllOut, these technologies are currently being employed to screen applicants for these positions. “AI and big data has been reshaping traditional business functions, and as work takes a hit amid the current crisis, these innovative solutions will be more sought after than ever,” he says. “In fact, not just companies, but even governments have been adopting these technologies to combat the crisis and put the economy back together,” he adds. Remote Work Boosting Data Science Jobs While remote work has emerged as a double-edged sword for some who are struggling to maintain a healthy work-life balance, many in the data science community are celebrating. Not only do presently available tools enable them to work efficiently at home but also the expansion of remote work has opened up many opportunities for freelance work as well. “Although the pandemic has brought the entire world to a standstill, it has initiated a discussion on the much-needed methods of working efficiently, giving major mileage to virtual and digital working methods,” says Ankush Singla, co-founder at Coding Ninjas. “This has also given an impetus to data science jobs which may be accomplished by individuals or teams connected remotely with the help of various tools,” he adds. According to him, the tech job market will witness a sharp rise in data science roles because “the pandemic is no longer about remote working alone, but also the significance of AI, ML, cloud computing and web applications being a major turning point in the lives of people.” ALSO READ: Top Websites To Find Freelance Data Science Jobs Singla quotes a study conducted by Appen that states that the pandemic has no “negative impact” on AI strategies of companies, who are relying on the positive impact that these technologies could have on “their organisation’s resiliency, efficiency, and innovation.” He adds, “Jobs anchored around these technologies will not only ramp up but also be instrumental in adopting cost-efficient ways of working in the future.” Concurs Misaal Turakhia, CPO at StepSetGo, “Since it is widely believed that AI is part of the way forward, I would be very surprised if companies decide to make sacrifices in such departments.” Affected, But Relatively Less So While all of the above may be true, there is no contending the fact that recession cannot mean life as usual for any industry, even in data science. And the travesty here is that it is not because these jobs are not critical today, but because they, like other jobs, have fallen prey to cost-cutting measures. “Given that data science is an evolving field, most of its projects across organisations tend to be strategic in nature, with mid to long-term outcomes,” says Narinder Kumar, COO and co-founder of To The New. “In the current pandemic, as organisations scrutinise cash flows, such projects may be put on the backburner,” he adds.According to him, although this may limit some job opportunities in data science, companies which are prudent will retain these jobs since “organisations that spend in data science R&D today will reap its benefits in the not-so-distant future post the Covid-19 pandemic.”","excerpt":"As some offices cautiously reopen and resume work, employees are finding themselves in an altered workplace. While tangible measures to enforce social distancing protocols have become a fixture across offices, the fact that fewer employees remain to join work post the lockdown is not lost on anyone. The uncertainty caused by the Covid-19 pandemic has […]","categories":["AI Features"],"tags":["big data trends and challenges","Data Science Jobs","data scientist salary"],"author_name":"Anu Thomas","publish_date":"2020-06-26T14:00:00","publication_year":"2020","word_count":846,"keywords":["data science","Go","API","artificial intelligence","machine learning","data scientist salary","AI","cloud computing","R","ML","Data Science Jobs","Git","big data trends and challenges"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","data science","cloud computing","R","Go","Git","API"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/is-landing-a-data-science-job-amid-covid-19-challenging-companies-weigh-in\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10049445,"title":"Airbus Releases the World&#8217;s Most Current Satellite Image DataSet, For Data Scientists To Work On","content":"Airbus Defence and Space, a division of Airbus Group, has launched One Atlas. The OneAtlas team, with utmost excitement, announced that three open-source artificial intelligence datasets have been published on the Kaggle platform. “With One Atlas, we propose a dramatic transformation to our clients in terms of how they access our data. We handle everything: updating, selecting, processing, and hosting – all to make our customers’ lives easier, cheaper, and faster,” said Bernhard Brenner, Head of Airbus Defence and Space’s Intelligence Business Cluster. Images from the SPOT and Pleiades satellites, which are part of the largest commercial satellite constellation in space, are included in these datasets. “Thousands of professionals, academics, students, and hobbyists will be able to use these datasets to train deep learning models on satellite imagery, thanks to their availability on the powerful Kaggle platform. We expect many data scientists to learn new skills as a result of the availability of this data and the user-friendliness of the Kaggle platform, bringing the worlds of AI and EO closer together,” said the firm in a post. A new satellite image basemap from Airbus Defence and Space, One Atlas covers the earth’s landmasses with professional-grade images. It provides customers with convenient access to cost-effective, high-quality, and homogeneous imagery, which is available online 24\/7 and refreshed every 12 months. As per them, One Atlas will benefit clients in a variety of ways, including military operations planning, infrastructure preparation research, and even tree cover change detection in deforestation-prone areas, to name a few. The solution streamlines the workflow and dramatically reduces the expenses of updating, choosing, processing, and hosting images. One Atlas is seamlessly integrated into the customer’s system and allows data to be shared across teams or partner organisations while maintaining security and privacy. It also allows users to plan, map, and find their teams, assets, or areas of interest from anywhere on the planet, freeing up time for operators to focus on their core job. Find out more about OneAtlas. Aircraft Detection Dataset from Airbus High Resolution Satellite Imagery Oil and Lubricant (POL) storage Detection Dataset from Airbus SPOT Satellite Imagery Airbus SPOT satellites images over wind turbines for classification","excerpt":"One Atlas, a new satellite image basemap, covers the whole Earth’s landmasses using professional grade images.","categories":["AI News"],"tags":["Deep Learning Models","satellite imagery"],"author_name":"Dr. Nivash Jeevanandam","publish_date":"2021-09-23T11:04:13","publication_year":"2021","word_count":360,"keywords":["satellite imagery","artificial intelligence","programming_languages:R","AI","Deep Learning Models","ML","RAG","deep learning","GAN","R"],"extracted_tech_keywords":["AI","artificial intelligence","ML","deep learning","RAG","R","GAN","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/satellite-image-dataset\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":16151,"title":"Importance of a Digital Strategy to Enable a World-Class Analytics Culture","content":"A company is considered digital if its business model and processes are digital. In general, data is a reflection of the business process, which in turn, is a reflection of the business model, and this is once again a reflection of the business strategy itself. While most companies acknowledge that we are now in the digital age, many sectors don’t reflect this reality. Change is inevitable, but transformation is often slow In a modern economy, digital maturity is a ‘must-have’ not a ‘nice-to-have’. Digitalization is essential to cater to today’s well-connected customer and align with present and future customer needs. Creative disruption led by technology has helped organizations reinvent processes, enhance quality and deliver consistency. Whether the goal is business growth or the creation of a flexible customer service model, digital tools have been crucial enablers of time-bound transformation. Businesses that have no digital plan will, at some point, realize that the established ways of working may no longer deliver the results they seek. Merely tweaking existing processes won’t work; a comprehensive approach is needed to put the business on a long-term sustainable path. The question is how soon and effectively can the business embark on this path. This is important because most companies are optimized to the environment in which they have been operating all these years. When that environment changes over time, they struggle to adapt and change. A digital strategy can help leaders steer the business on a path that moves them forward. Businesses that are adaptable have the highest likelihood of success. They don’t succumb to creative disruption, but leverage it to exceed their efficiency and performance. General Motor, Whirlpool, Boeing, Kellogg and Campbell Soup featured on the Fortune 500 list in both 1955 and 2014. Studebaker, Detroit Steel, Collins Radio, Zenith Electronics and National Sugar Refining were on the Fortune 500 list in 1955 but not in 2014. Facebook, Microsoft, Target, Office Depot and eBay were not on the Fortune 500 list in 1955 but they are among the globally-recognized brands on the list today. In the next 50-100 years, you can expect many of the current Fortune 500 luminaries replaced by new companies. This is creative disruption and innovation at work – a testament to the unlimited possibilities of human ingenuity in pioneering technological advancement. To transition gracefully to a world of artificial intelligence (AI), robotics and big data analytics, businesses need a clear understanding of what they need to do and how best to do it. Data holds all the answers Attempting to transform without a digital strategy in place not only increases the risk of failure pretty dramatically but can also be painful for both leadership and employees. If you don’t have a framework, you won’t be able to rally people around positive change. Even if you do manage to get there, the path will be painful for everyone involved, leading to an overwhelmed workforce and burnt out executives, if they do stick around for the entire journey. This is unnecessary when you have all the data in the world at your disposal and the tools and methods to harness the power of that data. Data analytics answers the ‘what’ and ‘why’ of strategy and execution. It allows businesses (and humankind at large) to answer questions that couldn’t be answered before, at a speed and scale that was unimaginable before. The accepted wisdom today is that data unlocks strategic value. By letting data rot until it is past its ‘usable date’, you can never acquire timely insights to take the best possible strategic business decisions. How can you develop the best data strategy? 1. Be people-centric: The best data strategy is one that creates massive economic value while simultaneously serving people and communities in the ecosystem. This is possible by adopting a data-centric mindset, so your first task will be to take a people-centric view as it is the collective efforts of people that drive the organization. A people strategy must necessarily be integrated into the data strategy for it to work as intended. After all, it is your employees who will be working with data. These ‘datapreneurs’ should be tasked with harnessing data for value creation; the chosen individuals must be well-versed with the organization, systems, processes and data. 2. Have a measurable strategy: Data can also tell you whether or not your strategy is working and to what extent. By setting measurement criteria, you are also committing to changing the strategy when data reveals this is needed. To make informed, evidence-backed changes, consider measurements across a number of critical dimensions of your business, such as organizational units, geographies, product and service lines, and your application and technology landscapes. 3. Give data ‘board-level importance: Today, companies treat their data strategy every bit as seriously as their business strategy. This is particularly evident in traditional companies that have embraced data digitalization after seeing Digital Native companies rack up massive valuations exceeding their own. Consensus on data strategies evolve in the boardroom; key decisions on data must be taken by board members. For instance, to deliver greater value from data to shareholders,  boardrooms can authorize actions that the company’s techies cannot, such as extending the lifecycle of the data value chain, overseeing the use of big data technologies in auditing, and taking the final call on strategies that can produce maximum value for shareholders. Emerging technologies have increased business interest in big data. Unfortunately, businesses have primarily focused on various tools and technologies than on the strategy itself. The problem with this approach is that excitement around new tools often overshadows the very purpose of the technology – using data to create business value. Companies that manage their data through superior data strategies are at least five to ten times more valuable than their peers or competitors. For evidence, look at the ten most valuable companies in the Fortune 500 list. These are the companies that have found their ‘billion dollar byte’, their most valuable data, which is easily reflected in their valuations.","excerpt":"A company is considered digital if its business model and processes are digital. In general, data is a reflection of the business process, which in turn, is a reflection of the business model, and this is once again a reflection of the business strategy itself. While most companies acknowledge that we are now in the […]","categories":["IT Services"],"tags":[],"author_name":"D Justhy","publish_date":"2017-07-07T07:13:11","publication_year":"2017","word_count":1000,"keywords":["big data","Go","artificial intelligence","AI","Git","RAG","ViT","analytics","GAN","R"],"extracted_tech_keywords":["AI","artificial intelligence","analytics","RAG","R","Go","Git","big data","GAN","ViT"],"url":"https:\/\/analyticsindiamag.com\/it-services\/importance-digital-strategy-enable-world-class-analytics-culture\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10047721,"title":"A Series On Underrated But Interesting Machine Learning Concepts","content":"There are some riveting concepts in machine learning that aren’t talked about as much as they should. With this article, we look at some of them, including manifold alignment, Quickprop, skill chaining, and FastICA. Manifold Alignment The basic concepts of manifold alignment are to use the links between instances within each dataset to improve the understanding of the relationships across datasets and, eventually, to map originally diverse datasets to shared latent space. In several areas of machine learning and data mining, one is frequently confronted with data distributed over many dimensions. Dealing with such high-dimensional data directly is usually impossible, although the manifold structure underlying the data may have a low intrinsic dimensionality in many circumstances. Chang Wang et al. states that, by matching the underlying manifolds of two or more distinct data sets, manifold alignment creates links between them and allows knowledge transfer. Benefits: Jingliang Hu concluded that data alignment-based (data concatenation) manifold techniques are less capable of fusing hyperspectral images and PolSAR data. Manifold alignment-based approaches, on the other hand, are more capable of fusing optical and SAR pictures. Semi-supervised methods, among the manifold alignment-based manifold techniques, are able to efficiently exploit both the structure of data and the existing label information. Quickprop Fahlman states that QuickProp is a second-order optimization technique that accelerates optimization by using a simple estimate of the Hessian’s diagonal, and hence belongs to the class of Quasi-Newton algorithms. In the event of a large number of iterations, weights, and samples, the learning process is exceedingly sluggish (or never reaches the target). According to Windra Swastika, the Quickprop approach is one way to speed up the learning process. The objective is to reduce the error gradient (E’). A number of algorithms attempt to improve on gradient descent by predicting an ideal learning rate at each update step, including AdaDelta, AdaGrad, and Adam. QuickProp, on the other hand, assumes that the output is a quadratic function of each variable and calculates the best position directly. Benefits: Windra Swastika et al. states that, when compared to Back Propagation, the Quickprop approach successfully reduces the learning process. For 40 input patterns, the Quickprop approach required only 20 iterations to obtain an error of less than 0.1, but the Back Propagation method required 2000 iterations. Both the methods have a prediction accuracy of more than 90%. Skill Chaining Skill chaining is a continual reinforcement learning skill discovery method. The associated deep skill chaining technique has expanded it to high-dimensional continuous domains. As per Singh et al., skill chaining creates a series of skills that lead to one set of designated target events, which can be as simple as the episode’s end-of-episode event or as complex as intrinsically intriguing events. The skills are built so that completing each option in the chain allows the agent to move closer to its ultimate goal by executing another option. While skill chaining could uncover skills in continuous state spaces, it could only be used in state spaces with discrete actions that were relatively low in dimension. Benefits: George. state that the main advantage of skill chaining is that it alleviates the stress of describing the task’s value function. Eylem Tekin. concludes that skill chaining strategies are useful in various systems, including those with many products and those with a parallel or network flow structure, such as cells with automated equipment. In such situations, more research is needed to establish the possible benefits of chaining. FastICA FastICA is a time-saving independent component analysis algorithm. Erkki Oja. expresses that the instantaneous noise-free ICA model was the inspiration for FastICA. PierreComon. state that an observed random vector is transformed into statistically independent components using Independent Component Analysis (ICA). Vicente Zarzoso. said that FastICA has only been compared against neural-based adaptive algorithms like principal component analysis (PCA), which are known to outperform most ICA algorithms. As well as its simplicity, the method’s popularity can be attributed to its good performance in various applications. According to P. Chevalier, FastICA fails for weak or highly spatially correlated sources, which is the first significant attempt. Benefits: According to Vendetta, FastICA is the most generally utilised method for blind source separation problems because it is computationally efficient and uses less memory than other blind source separation algorithms, such as infomax. Another benefit is that separate components can be calculated one by one, reducing computing load once again. The only drawback is that if the noise is non-uniform and correlated noise vectors, this method will fail to determine sources correctly.","excerpt":"Underrated but interesting machine learning concepts will be explored in this series.","categories":["AI Features"],"tags":["ADAM","back propagation","Machine Learning","optimization algorithms neural networks","Reinforcement Learning","reinforcement learning an introduction"],"author_name":"Dr. Nivash Jeevanandam","publish_date":"2021-09-06T15:00:00","publication_year":"2021","word_count":749,"keywords":["Go","back propagation","machine learning","Reinforcement Learning","optimization algorithms neural networks","AI","TPU","ELT","programming_languages:R","Machine Learning","programming_languages:Go","ADAM","R","reinforcement learning an introduction"],"extracted_tech_keywords":["AI","machine learning","TPU","R","Go","ELT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/a-series-on-underrated-but-interesting-machine-learning-concepts\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":36550,"title":"Machine Learning Behind Gboard &#038; How The Handwriting Recognition Feature Makes It Smarter","content":"No two people in this world write in the same style, unless one decides to make a serious career out of forgery. Even identical twins who share same genetics end up having different handwriting styles. Every style is determined by the spacing between letters, slope of the letters, pressure applied and their thickness. Whereas, legibility in a style can also be driven by medical conditions. Our scripts contain spelling errors and undecipherable styles; doctor’s prescriptions. But we somehow manage to get the gist of it, no matter how bad the writing style is, through context and from our own experience of writing. Now when a machine is tasked with identifying million styles in real time to give a relevant result and add to that, in multiple languages, it gets cumbersome. Google’s Gboard does all of it and does it with great accuracy. What does the above picture say- is it ‘rural’, ‘mural’, ‘weird’, ‘curd’? It can mean many things to most of the people or it can be some random scribbling. But Gboard gets it with great accuracy: In this recent paper, Google demonstrated how they used machine learning models to enable Gboard into predicting the right word for any style or language. Fast Multi-language LSTM approach makes handwriting recognizers more flexible. The user can be using their keyboard on large screens or small screens, landscape or portrait mode, on devices of varying resolution and touch sensitivities. How Does ML Help? The first objective of the model is to normalise the touch-point coordinates to compensate for the varied resolutions. Writing on a smartphone screen usually involves cursive writing and getting the shape of these letters helps the model learn more accurately. In order to do this, the researchers at Google, convert sequence of points of the scribbling into a sequence of cubic Bezier curves. Though Bezier curves are widely popular with fonts, this technique is different from the Google’s segment-and-decode approach used previously. This data of Bezier curves is then fed into recurrent neural network(RNN). The above figure is an illustration of curve fitting of the user’s input.  The letter ‘g’ has two control points for each of the four cubic Bezier curves, which are represented by the colors yellow, blue, pink and green. Now since the handwritten input is represented by Bezier curves, these sequence of input curves have to be translated back to actual written characters. To process these curves, a multi-layer RNN is used, which produces an output decoding matrix. This matrix contains the values of probability distribution of each input curve for a possible letter. The rows contain the letters to ‘a’ to ‘z’ and the columns contain the probability distribution of the possible letters corresponding to the input curves. Source: Blog by Google In the word go, ‘g’ was represented by four cubic Bezier curves and ‘o’ by three. To address this mismatch in representation and to get the actual number of characters in the input, Connectionist Temporal Classification algorithm. This algorithm is used for selecting the most probable labelling for a given input sequence. These labeled outputs of the probability distributions of the possible letters are then decoded into letters with the help of a Finite State Machine Decoder. This decoder combines the outputs of the neural network with a character-based language model. This model rewards commonly used sequences and penalizes uncommon usage of any sequence in the letters. The variant of RNN used for this approach is the bidirectional version of quasi-RNN(QRNN). Quasi:  a switch between convolutional and recurrent layers; parallelization through small number of weights. Smaller the weights, the smaller the size of the model that will be downloaded onto devices using Gboard. And, to reduce the delay in popping up results, recognition models are converted to TensorFlow Lite models. Gboard now supports hundreds of languages from Mandarin to Malayalam and probably Maths, the language of nature. Know more about the work here","excerpt":"No two people in this world write in the same style, unless one decides to make a serious career out of forgery. Even identical twins who share same genetics end up having different handwriting styles. Every style is determined by the spacing between letters, slope of the letters, pressure applied and their thickness. Whereas, legibility […]","categories":["Deep Tech"],"tags":["Machine Learning","RNN"],"author_name":"Ram Sagar","publish_date":"2019-03-19T09:37:51","publication_year":"2019","word_count":651,"keywords":["Go","machine learning","TPU","AI","neural network","ML","Machine Learning","RNN","LSTM","TensorFlow","R"],"extracted_tech_keywords":["AI","machine learning","ML","neural network","TensorFlow","TPU","R","Go","RNN","LSTM"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/machine-learning-behind-gboard-how-the-handwriting-recognition-feature-makes-it-smarter\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10019621,"title":"How Is Alteryx Advancing Data &#038; Analytics Potential With Its Game-changing Resources","content":"The overnight transformation of companies adopting new technologies and transitioning to a digital work environment amid pandemic has made upskilling the most critical component in a worker’s repertoire in 2021. While information, data and the ability to make the right decisions serve as a stabiliser across verticals, analytics and data science have become indispensable tools to navigate today’s career scene. According to a recent Forrester study, the top two challenges decision-makers cited are — the lack of employees with data skillsets and the lack of skills among business users who must use data insights. Almost 66% of organisations believe there is a requirement for data literacy among employees, where 59% demand analytic efficiency. However, with a converged approach to analytics through democratising access to data, automating tedious and complex processes, and promoting upskilling of data and knowledge workers, organisations can create a thriving data and analytics culture within. Advancing Data & Analytics Potential Together, also known as ADAPT, an innovative new program, developed by Alteryx, provides free access to data science and analytics learning courses for recent graduates and data workers. The courses are designed to help new entrants in data analytics and also those impacted by COVID-19. The new offering as part of the company’s corporate social responsibility initiative, Alteryx For Good, ADAPT will equip learners with the skills required to enter a desirable field or make strides in the current career path. Alteryx believes any data worker in an organisation can tackle sophisticated analytical tasks if empowered with the right training, technology and tools. Start Your Analytics Journey With ADAPT 2020 has witnessed a fundamental transformation towards the data economy, which magnified the importance of appropriate talent investments in data science skills. The ADAPT program, offered globally to COVID-hit workers, helps upskill their analytics capabilities free of cost. Retail workers whose stores were shut down or business professionals who were laid off, or recent graduates are eligible to apply. Once accepted into the program, you will receive access to: an introductory course to the world of data science; the fundamentals in working with data; and data science foundational concepts. Libby Duane Adams, co-founder and Chief Advocacy Officer of Alteryx, believes the program will help everyone become a data worker. With ADAPT, you will also have a six-month license of Alteryx Designer, a self-service data analytics platform that simplifies data preparation and assists in automating reporting, predictive analytics, and accelerating visual insights for businesses. Further, ADAPT also provides access to thousands of data science and analytics experts, user groups and more resources within the Alteryx Community with one-on-one Alteryx associate support through the Virtual Solution Center. With a challenging and up-to-date curriculum delivered as a 125-hour boot camp for future data workers, ADAPT will offer the course via a combination of interactive lessons, videos, webinars and podcasts. After completing the ADAPT program, students can benefit from their status as citizen data scientists and return to the industry with a stronger resume and skillsets. With a global shortage of data scientists worldwide, these skills are in high demand and will equip the participants to re-enter the workforce. The program could be completed in a month if dedicated full-time but could last three or more months with an hour or two of study per day. The program’s key modules include data set creation, developing classification models, conducting A\/B testing, time series, and data segmentation. Once furloughed, re-entering the job market can be a challenging task; furthermore, it has been a struggle for new graduates to make their case for a job amid the pandemic. The ADAPT program is unique in its scale and ambition with its academic rigour and enhanced certification, providing an introduction to the global data science community. Know more about the ADAPT program here. Maturing Your Analytics Alteryx Analytic Process Automation Platform is a widely used data science platform that enables self-service analytics to democratise data. Making the data accessible, automating business processes, and allowing employees to pick up relevant skills to transition to a higher value job makes the company agile and adaptable to the rising challenges — adding to the organisation’s value in the process. Especially during an economic downsizing, it has become more critical than ever that these data programs provide a significant return on investment (ROI). For enterprises, quick wins and fast return on investment are critical to staying ahead of the competition. With Analytic Process Automation, businesses will be able to reveal answers to two key questions — what can be done to make more; and what can be done to spend less. With its interactive ROI calculator, businesses can assess their ROI potential by merely inputting the analysts’ number and the company’s annual revenue. The ROI calculator provides a personalised report with a full breakdown of ROI potential in every business area and enables businesses to share it with their key stakeholders. One can even leverage the pre-populated fields for the required numbers. The four areas of ROI with Alteryx APA includes — Top-line growthBottom-line growthEfficiency gainsPerpetual upskilling As a matter of fact, Alteryx APA customers highlighted the substantial value they linked to the use of the platform. According to IDC, they specifically cited their ability to increase their staff efficiencies, value, and risk management, along with enabling real-time data reporting and efficiencies through process automation and driving business growth with product differentiation. Some of the additional benefits pointed out by Alteryx APA customers are — reduced reporting burden, repeatability and parameters that drive efficiencies, substantial time savings through automation, and more timely delivery of more complex analysis. IDC also looked at the Alteryx APA platform’s impact on model creation, a core activity for business analysts and other data-focused teams. Customers reported much more efficient and robust modelling activities with Alteryx APA. This allows business analyst teams to be responsible for handling, modelling, and delivering data-driven insights efficiently, making them more competent. Wrapping Up Considering more firms are reaping the benefits of effectively embedding data-driven decision making, the importance of data democratisation and building a culture of analytics has become paramount. The ADAPT program by Alteryx is committed to expanding data literacy globally for companies to witness a continued success with the highest ROI. Know more about the ADAPT program here. ROI CalculatorAlteryx’s resources (https:\/\/www.alteryx.com\/resources)","excerpt":"The overnight transformation of companies adopting new technologies and transitioning to a digital work environment amid pandemic has made upskilling the most critical component in a worker’s repertoire in 2021. While information, data and the ability to make the right decisions serve as a stabiliser across verticals, analytics and data science have become indispensable tools […]","categories":["IT Services"],"tags":["Alteryx","data and analytics","Data Science","data science and analytics","data science curriculum"],"author_name":"Sejuti Das","publish_date":"2021-02-03T14:00:00","publication_year":"2021","word_count":1043,"keywords":["data science","Go","API","data and analytics","AI","data science curriculum","R","predictive analytics","Git","RAG","analytics","GAN","Data Science","Alteryx","data science and analytics"],"extracted_tech_keywords":["AI","data science","analytics","RAG","predictive analytics","R","Go","Git","API","GAN"],"url":"https:\/\/analyticsindiamag.com\/it-services\/how-is-alteryx-advancing-data-analytics-potential-with-its-game-changing-resources\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10121036,"title":"Indian IT is Training a GenAI Workforce to Eventually Replace Them With AI","content":"Indian IT giants, including Infosys, TCS, and Wipro, have collectively trained over 825,000 employees in generative AI. However, an X user has revealed that the depth and quality of these training programs may be questionable, as a friend working at one of the largest IT companies in India completed a GenAI course in just an hour by clicking the “next” button hundreds of times. The post received widespread agreement from people across India’s tech ecosystem. While few said that these are the people who would actually be replaced by AI, others suggested employing OpenAI’s new GPT-4o model for mundane tasks. Impressive Numbers, But Questionable Depth Moreover, Wipro’s new chief, Srini Pallia, stated that the company’s 225,000 employees have been trained in AI 101, a basic level of AI. The company is considering advanced courses for individuals based on the types of projects and proofs of concept (PoCs) they are involved in. Infosys and TCS have also announced plans to train 100,000 and 150,000 employees, respectively, focusing on theoretical and practical aspects of GenAI through partnerships with industry leaders. While these numbers are impressive, questions arise about the depth of skills, specialisation options, and real-world application integration. The recent revelation about the superficial nature of some training programs raises concerns about the actual readiness of the workforce for GenAI projects. Underutilisation of Trained Workforce Although Indian IT companies employ a substantial workforce compared to other major tech giants, there are concerns about the underutilisation of their highly-trained employees. This limitation may hinder the potential for innovation and the effective application of GenAI in real-world scenarios. The recent revelation about the superficial nature of some training programs further underscores the need for a more comprehensive and rigorous approach to GenAI education within the Indian IT industry. Companies must ensure that their employees are not only trained in the basics but also equipped with the necessary skills and knowledge to effectively implement GenAI solutions in practice. As the demand for GenAI expertise continues to grow, Indian IT companies must prioritise the development of a truly skilled and ready workforce to remain competitive in the global market. This requires a commitment to in-depth training, practical application, and continuous learning to keep pace with the rapidly evolving field of generative AI.","excerpt":"While Indian IT giants claim to have trained over 825,000 employees in genAI, questions arise on the depth of skills, specialisation etc.","categories":["AI News"],"tags":["AI in Indian IT"],"author_name":"Shyam Nandan Upadhyay","publish_date":"2024-05-20T14:12:21","publication_year":"2024","word_count":376,"keywords":["AI in Indian IT","Go","API","GenAI","OpenAI","AI","GPT-4o","innovation","GPT","generative AI","R"],"extracted_tech_keywords":["AI","generative AI","GenAI","GPT-4o","OpenAI","R","Go","API","GPT","innovation"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/indian-it-is-training-a-genai-workforce-to-eventually-replace-them-with-ai\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":53165,"title":"Google Launches New Product Updates For Its AI Platform Coral","content":"Google has been striving hard to make local AI possible by applying AI to products and other domains as well as developing tools. In local AI, the speed of local inference allows considerable savings on bandwidth and cloud compute costs while keeping data local preserves user privacy. These are the few reasons as to why organisations prefer to use the local inferencing platform. Last year, the tech giant launched Coral, an AI platform to build products with local AI. The platform has gained lots of traction since then and has witnessed frequent updates. Popular organisations like Asus, Florida-based care.ai, among others have also been using this platform for the development of AI. The multinational company, Asus has chosen the Coral SOM as the base of its Tinker Edge T product. The company will be using the Google Edge TPU to speed up its processing efficiency and build connected devices and intelligent applications. Google Coral provides hardware components and software tools that make it easy to prototype and scale local AI products. In this platform, the solutions for various on-device intelligence include object detection, pose estimation, image segmentation, and key-phrase detection. Last year, the tech giant also released Coral Dev Board which is a hardware board to accelerate AI edge computing and a USB accelerator which is a USB accessory featuring the Edge TPU, and Mini PCIe Accelerator Recently, the tech giant announced new additions to the Coral platform. The unveiled hardware products for Coral include Coral Accelerator Module, Coral Dev Board Mini and some new variations to the Coral System-on-Module. About The Products Coral Accelerator Module This is an easy to integrate multi-chip package that encapsulates the Google Edge TPU ASIC. Coral Accelerator Module exposes both PCIe Gen 2 and USB 2.0 interfaces and can easily integrate into custom PCB designs. The mounting type is SMT, 120-pin LGA and the control interface is 12C (optional). It has a number of functionalities such as performs high-speed ML inferencing, supports Tensorflow Lite as well as supports AutoML vision edge. At CES 2020, the tech giant will be producing a demo of this module and it will be available publicly by the first half of this year. Coral Dev Board Mini Coral Dev Board Mini is a development board which provides a smaller form-factor, lower-power, and lower-cost alternative to the Coral Dev Board. This board has the capability to perform high-speed ML inferencing, support Tensorflow Lite, support AutoML vision edge as well as working as a single-board computer with SoC, Ml, and wireless connectivity. It combines the new Coral Accelerator Module with the MediaTek 8167s SoC to create a board that excels at 720P video encoding\/decoding and computer vision use cases. Coral System-on-Module This year, the tech giant has offered new variations to the Coral System-on-Module. The System-on-Module is a fully integrated system for accelerated ML applications which includes  CPU, GPU, Edge TPU, Wi-Fi, Bluetooth, and secure element in a 40mm x 48mm pluggable module. This module is now available with 2GB and 4GB LPDDR4 RAM in addition to the original 1GB LPDDR4 configuration. Wrapping Up Ahead of Consumer Electronics Show (CES) 2020 which is going to be held on 7th-10th Jan, the tech giant unveiled these new hardware products of Coral. Google will display these hardware products as well as showcase demo at the upcoming event. Further, the developers at Google will be showcasing how the System-on-Module (SoM) can be used in a smart city, manufacturing sector, and healthcare applications.","excerpt":"Google has been striving hard to make local AI possible by applying AI to products and other domains as well as developing tools. In local AI, the speed of local inference allows considerable savings on bandwidth and cloud compute costs while keeping data local preserves user privacy. These are the few reasons as to why […]","categories":["Global Tech"],"tags":[],"author_name":"Ambika Choudhury","publish_date":"2020-01-05T14:00:00","publication_year":"2020","word_count":578,"keywords":["Go","TPU","AI","ML","computer vision","GAN","object detection","edge computing","TensorFlow","R"],"extracted_tech_keywords":["AI","ML","computer vision","TensorFlow","object detection","edge computing","TPU","R","Go","GAN"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/google-launches-new-product-updates-for-its-ai-platform-coral\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":15845,"title":"India has potential of becoming a nucleus of Data Analytics","content":"Data analytics and Big Data are perhaps the two most dominant topics of every conversation or debate at the C-Suite level today. It wouldn’t be an exaggeration to call them the honeypots of technology trends, which can’t be ignored. It reminds us of the fact that we as human species live in the Era of Data Analytics and Big Data. As per a report published by IDC, the Big Data and Data Analytics market is projected to grow into a whopping $203 billion market by 2020 from $130.1 billion in 2016. The report also depicts that the industry received the largest investment, somewhere to the tune of $17 billion in 2016 and it is suggested that the segment will grow at a Compound Annual Growth Rate (CAGR) of 11.7%. As the market grows at a staggering pace, Data Analytics continues to become more and more mainstream. While this was an overall outlook on Data Analytics and Big Data, in this editorial piece we will explore how these technology trends are playing in India and the country’s placement in terms of being able to cash in on this Big Data and Analytics frenzy. Even before we begin talking about the scale and scope of Data Analytics and Big data in the Indian market, we would like to reiterate the fact that India is a breeding ground of entrepreneurship, innovation, and services. This has rendered the country at the cusp of major opportunity that has demonstrated traits and potential of becoming a nucleus of Data Analytics. Corroborating the above stance, let’s discuss the key growth drivers that will help India capitalize on the growing Data Analytics and Big Data market opportunity. Let’s start from the beginning! World’s favourite Outsourcing Market The trend of outsourcing service caught up in India back in 1999, when EXL services introduced its Business Process Outsourcing unit in India. Ever since there has been no looking back for the outsourcing segment in India. India has very well established its clout in the IT-BPO and knowledge service outsourcing markets at a price slab that suits businesses and all stakeholders. Following that, with a growing number of entrepreneurs, expanding their Following that, with a growing number of entrepreneurs, expanding their home grown companies across borders, offering IT as a service, India has already started witnessing massive demands for outsourced analytics work. With such ongoing developments, India is presently projected to emerge as the key outsourcing destination in the overall Big Data space. However, the question that could arise in any reasoned mind is – What makes India the hot bed of Data Analytics amongst all the developed countries? World’s largest IT talent pool It is not unknown to people that India’s biggest strength lies in its young work force, of which, we have abundance of mathematics and statistics talent, with one of the largest number of engineer populations in the world. Owing to these aspects, India is favourably placed to deal with issues of analytics skills gaps, which otherwise is becoming a major concern across the globe. Furthering the growth, in view of the growing demand of Data Analytics and Big Data, the Indian IT education system is also either being customized or upgraded to fit the bill. Nurturing and training professionals with deep analytics skills, India is expected to produce more and more skilled professionals for the segment. As a matter of fact, Nasscom-Crisil report confirms that India is ranked right after Furthering the growth, in view of the growing demand of Data Analytics and Big Data, the Indian IT education system is also either being customized or upgraded to fit the bill. Nurturing and training professionals with deep analytics skills, India is expected to produce more and more skilled professionals for the segment. As a matter of fact, Nasscom-Crisil report confirms that India is ranked right after the US, in terms of Big Data talent availabilities and in initiatives to further such talents. Increasing domestic demands While India has already proven its mettle as an outsourced third party service provider, in a very short time it has come a long way, to be able to generate Data Analytics demand domestically. Such demands not only come from private enterprises but also from the government. As a matter of fact, having realized the clout of Data Analytics, various government departments and public agencies have resorted to Big Data Analytics to deal with certain persistent issues of the country such as air pollution, power shortage, low crop yield etc. Burgeoning Entrepreneurial Landscape The new age Indian Entrepreneurs have taken everything by their strides, aiding in numerous fledgling sectors to grow further. As far as Data Analytics is concerned, the sector looks vibrant. In fact, last year only Nasscom reported that India had over 600 analytical firms in the country, with 400 of them being startups. Ending Note With companies from India and abroad, expecting India to fulfill the skill gap in Data Analytics, it unlocks the gateway of opportunities for those 600 analytical firms and also creates more space for new players to jump into the opportunity to serve the ever growing demand of Big Data and Data Analytics. Verticals such BFSI, e-commerce, telecom etc. are seen investing heavily in their data science teams. It would only be a part of its natural progression with other industries joining Verticals such BFSI, e-commerce, telecom etc. are seen investing heavily in their data science teams. It would only be a part of its natural progression with other industries joining in the bandwagon to make the most of analytics, to scale their businesses.","excerpt":"Data analytics and Big Data are perhaps the two most dominant topics of every conversation or debate at the C-Suite level today. It wouldn’t be an exaggeration to call them the honeypots of technology trends, which can’t be ignored. It reminds us of the fact that we as human species live in the Era of […]","categories":["IT Services"],"tags":[],"author_name":"Luc Burgelman","publish_date":"2017-06-23T08:08:47","publication_year":"2017","word_count":929,"keywords":["big data","Go","data science","API","programming_languages:R","AI","innovation","analytics","R","startup"],"extracted_tech_keywords":["AI","data science","analytics","R","Go","API","big data","innovation","startup","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/it-services\/india-has-potential-of-becoming-a-nucleus-of-data-analytics\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10021115,"title":"How To Stave Off Data Security Attacks: A Checklist","content":"Pre-trained models, free datasets, APIs and all things open source have made ML development more affordable. The flip side to this is, these freely available datasets and pre-trained models are open for malicious players. Data poisoning, weight poisoning etc are well documented phenomena in the machine learning space. Detecting Neural Trojan attacks for an unknown DNN is difficult due to the stealthiness of backdoors that makes them hard to identify by functional testing (which uses the test accuracy as the detection criteria) and only limited information can be obtained about the queried model during Trojan detection. A clean training dataset or a gold reference model might not be available in real-world settings. According to a survey by the researchers at University of Maryland and other top institutes, dataset creators often rely on data-hungry neural network models built by harvesting information from anonymous and unverified sources on the web. “Outsiders can passively manipulate datasets by placing corrupted data on the web and waiting for data harvesting bots to collect them.” In their report, they list top defense mechanisms to avoid various types of dataset security attacks: Also Read: What Is Poisoning Attack & Why It Deserves Immediate Attention Trigger Reconstruction Backdoors can stay hidden indefinitely until activated by an input, and present a serious security risk to biometric authentication systems, self-driving cars etc. For trigger reconstruction, the backdoored models are trained to assign an adversarial label. A trigger is introduced even when a small number of pixels are manipulated. The backdoor triggers are recovered by computing adversarial perturbations for the target labels, and then selecting the smallest perturbation out of all labels. Popular methods: Neural Cleanse, Deep Inspect, TABOR. Trigger Agnostic Strategies The researchers wrote that combining pruning with fine-tuning can remove backdoors while preserving the overall model accuracy, which is still an open problem for many strategies. And, trigger agnostic methods promise to maintain accuracy even when an adversary tries to manipulate pruning. To better preserve the model’s accuracy on clean data, researchers suggest watermark removal framework REfiT that leverages elastic weight consolidation. This defensive training process delays the learning of model weights while updating other weights that are likely responsible for memorizing watermarks. Popular methods: REfiT, WILD. Randomized Smoothing Randomized smoothing was originally proposed to defend against evasion attacks. The technique here is to pick a smoothed version of a base model whose model prediction on each data point is replaced with popular prediction in the vicinity of the data point. The researchers state that these outputs of the smoothed model can be computed efficiently regardless of the perturbations. Differential Privacy Differential privacy offers a mathematical framework to anonymise data. It is a high-assurance, analytic means of ensuring that use cases like these are addressed in a privacy-preserving manner. Even companies like Apple leverage differential privacy techniques while collecting feedback from its users in a safe way. The objective of differential privacy is to ensure models will not be disproportionately affected by poisoned samples. The core idea is that if the output of the algorithm remains essentially unchanged when one individual input point is added or subtracted, the privacy of each individual is preserved. Defenses For Federated Learning Federated learning pushes the boundary further by sharing the information across the devices with sophisticated anonymity. The main drivers behind FL are privacy and confidentiality concerns and regulatory compliance requirements. To avoid attacks on FL regimes, the researchers introduced Robust federated aggregation algorithms that attempt to nullify the effects of attacks while aggregating client updates. These algorithms are designed to identify and bring down the weights (avoiding wright poisoning) of the malicious updates. Another method is to compute aggregates in a way which is resistant to poisons. In addition to robust federated aggregation to mitigate poisoning attacks, the researchers also recommend clipping the norm of model updates and adding Gaussian noise to mitigate backdoor attacks based on the model replacement paradigm. While the aforementioned defense strategies have been promising, the researchers expect to resolve other persisting problems in the future: Defenses on problems other than image classification.Avoiding poisoning while maintaining accuracy.Defense without access to training protocol. Read the full survey here.","excerpt":"Pre-trained models, free datasets, APIs and all things open source have made ML development more affordable. The flip side to this is, these freely available datasets and pre-trained models are open for malicious players. Data poisoning, weight poisoning etc are well documented phenomena in the machine learning space. Detecting Neural Trojan attacks for an unknown […]","categories":["AI Trends"],"tags":["data privacy use cases"],"author_name":"Ram Sagar","publish_date":"2021-03-02T10:00:00","publication_year":"2021","word_count":690,"keywords":["federated learning","Go","machine learning","TPU","AI","neural network","ML","data privacy use cases","RAG","differential privacy","R"],"extracted_tech_keywords":["AI","machine learning","ML","neural network","federated learning","differential privacy","RAG","TPU","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/how-to-stave-off-data-security-attacks-a-checklist\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10167902,"title":"I’m Not So Interested in LLMs Anymore, Says Yann LeCun","content":"Meta’s chief AI scientist, Yann LeCun, said he is no longer interested in large language models (LLMs), calling them a product-driven technology that is reaching its limits. “I’m not so interested in LLMs anymore,” LeCun said during a recent talk at NVIDIA GTC 2025. He added that today, LLMs are mainly handled by product teams that are making small improvements by adding more data, increasing computing, and using synthetic data. LeCun explained that his focus has shifted to four areas he considers more fundamental for machine intelligence: understanding the physical world, persistent memory, reasoning, and planning. “There is some effort, of course, to get LLMs to reason, but in my opinion, it’s a very simplistic way of viewing reasoning,” he said. “I think there are probably better ways of doing this.” LeCun expressed interest in what he called “world models” — systems that form internal representations of the physical environment to enable reasoning and prediction. “We all have world models in our minds. This is what allows us to manipulate thoughts, essentially,” he said. He criticised the current reliance on token prediction, which underpins how LLMs operate. “Tokens are discrete… When you train a system to predict tokens, you can never train it to predict the exact token that will follow,” LeCun said. He argued that this approach is insufficient for understanding high-dimensional and continuous data like video. “Every attempt at trying to get a system to understand the world or build mental models of the world by being trained to predict videos at a pixel level has failed,” he said. Instead, he pointed to ‘joint embedding predictive architectures’ as a more promising approach. These architectures, according to LeCun, make predictions in abstract representation space rather than raw input space. He described a method where a system observes the current state of the world, imagines an action, and then predicts the next state — a core component of planning. “We don’t do [reasoning and planning] in token space,” he said. “That’s the real way we all do planning and reasoning.” He also criticised current agentic AI systems that rely on generating many token sequences and selecting the best one. “It’s sort of like writing a program without knowing how to write it,” he said. “You write a random program and then test them all… It’s completely hopeless.” Responding to growing claims about the imminent arrival of artificial general intelligence (AGI), or what some call advanced machine intelligence (AMI), LeCun remained sceptical.","excerpt":"LeCun explained that his focus has shifted to four areas he considers more fundamental for machine intelligence: understanding the physical world, persistent memory, reasoning, and planning.","categories":["AI News"],"tags":["Meta AI","Yann LeCun"],"author_name":"Siddharth Jindal","publish_date":"2025-04-14T14:39:21","publication_year":"2025","word_count":411,"keywords":["Meta AI","agentic AI","Yann LeCun","synthetic data","AI","programming_languages:R","emerging_tech:synthetic data","Aim","R"],"extracted_tech_keywords":["AI","agentic AI","Aim","R","synthetic data","programming_languages:R","emerging_tech:synthetic data"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/im-not-so-interested-in-llms-anymore-says-yann-lecun\/","complexity_score":3,"technical_depth":7,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10143698,"title":"Agentic AI is Now On Mid-Sized Indian IT’s Mind","content":"The era of agentic AI is upon us, and surprisingly, Indian IT is also ready for this revolution. What is even more surprising is the fact that mid-sized and small IT firms are also ready for adoption. As usual, they are taking the acquisition route to achieve these capabilities.The latest example is LTIMindtree. Last week, the IT firm announced that it has committed $6 million to Voicing AI, a US-based startup specialising in human-like AI voice agents. This investment aims to bring human-like voice capability across more than 20 languages with conversational, contextual, and emotional intelligence. Moreover, LTIMindtree has formed an alliance with GitHub for its Copilot coding tool to train its workforce on the GitHub ecosystem. Similarly, Mphasis has focused on conversational AI to boost user experiences. In July, the IT solutions provider launched NeoCrux, a tool for improving developer productivity by streamlining software developer cycles with AI agents and orchestrators. In October 2023, the company acquired Silverline, a New York-based Salesforce partner, for $132.5 million. During the announcement, Nitin Rakesh, CEO of Mphasis, said, “The acquisition aligns with our strategy to drive customer experience transformation, modernise contact centres, and enable conversational AI automation to meet evolving client needs.” It seems evident that Mphasis has decided to embrace AI and agentic AI for customer interactions as well. Though it is not clear whether the acquisition of Silverline increased Mphasis’s current capabilities, the impact is definitely visible. Persistent Systems, on the other hand, is addressing AI-related privacy challenges. In September, the company acquired Pune-based data privacy consultancy Arrka for ₹14.4 crore ($1.7 million). “Arrka’s mature frameworks and privacy management platform ensure a scalable and governance-driven approach, critical for successful AI implementations,” Persistent Systems CEO Sandeep Kalra said in a press release. The Approach Differs from the Big IT Meanwhile, bigger IT giants like TCS, Infosys, and Wipro have decided to develop a different strategy for building AI agents for their customers. In the latest quarterly results, TCS posted $1.5 billion in bookings with AI and Infosys has started building small language models and multi-agent frameworks for its clients. A similar situation was observed at Wipro, HCLTech, and Cognizant. TCS, India’s largest IT services exporter, launched a new AI-focused business unit, AI.Cloud, in May, by merging its cloud and AI divisions. Infosys, the second-largest IT firm, unveiled Topaz, an AI suite offering solutions and services. Notably, it has several clients in its portfolio. The mid-sized IT companies are, however, positioning themselves differently here. Speaking with AIM, Nachiket Deshpande, COO at LTIMindtree, said Indian IT firms often do not want to build these capabilities in-house. Commenting on the recent deal with Voicing AI, Deshpande added, “There are so many startups that are coming all around the world. We would want to leverage those startups for those technologies that are coming up.” Deshpande’s biggest reason for not building in-house models is that generative AI is so expensive that it doesn’t make sense to experiment with it while investing so much. “I don’t think the technology in generative AI would be the real differentiator because of the level and the speed of innovation that is happening, whatever technology you may develop will become obsolete within a few months, and you have to keep doing that,” Deshpande said, adding that it requires a lot of capital. Mohandas Pai, the founder of Aarin Capital and former CFO of Infosys, also told AIM that IT services companies are not built to invest disproportionate capital R&D. “Indian IT services companies are not product companies.” “Creating an LLM or a big AI model requires large capital, time, a huge computing facility, and a market. All of which India does not have,” Pai pointed out, adding that even though Infosys, TCS, and others might have the funds, their focus is to provide vertical solutions and not horizontal ones like ChatGPT. Why Is Everyone Bullish on Agentic AI? Similarly, Deshpande also said that the services companies’ P&L is structured differently than that of product companies. “They operate with 80% cross-market, and hence they have the ability to continue to do R&D, and we operate at 30-35% cross-market,” he explained. He believes that if a company invests and builds a particular technology, tomorrow it might become irrelevant. “I need to get 86,000 people to reimagine their work with AI, but I only need 2,000 people to build AI solutions.” “The differentiation of LTIMindtree will lie in terms of how we adopt AI, rather than saying I have another shiny toy which is better than somebody else. Because that differentiation is short-lived,” he added. Apart from the technological capabilities, Deshpande said it’s important that IT companies give startups the space they need to grow on their own. “Hence, acquisition was not the best way, but at a meaningful scale was a better way where we get to leverage the technology, we get to take them to the customer, but also provide those entrepreneurs space to continue their innovation,” he further said. With generative AI systems going autonomous and several big-tech firms announcing models like Devin and others, Indian IT is also ready to adopt that change quickly. Deshpande said that one of the biggest reasons for this is the outlook on productivity. “The idea of productivity outlook is persona-centric, and you have to look at each persona becoming more and more productive,” Deshpande said. Hence, he said that agentic systems, which do not examine a particular business process or task but the entire persona and try to automate large parts of those personas, are the way to adopt that productivity. That’s why agentic systems are becoming increasingly focused. Earlier this year, small IT firms were taking the acquisition route for building AI capabilities. Happiest Minds, Hexaware, Quest Global, Coforge, Sonata, and GlobalLogic have all announced acquisitions in this space. Now, the conversation has shifted to agentic AI.","excerpt":"The biggest reason for not building in-house models for Indian IT firms is that generative AI is too expensive to experiment with.","categories":["IT Services"],"tags":["agentic ai","Indian IT"],"author_name":"Mohit Pandey","publish_date":"2024-12-17T12:30:00","publication_year":"2024","word_count":974,"keywords":["Go","ChatGPT","agentic AI","AI","ML","agentic ai","RAG","Aim","generative AI","small language models","Indian IT","R"],"extracted_tech_keywords":["AI","ML","generative AI","agentic AI","ChatGPT","Aim","small language models","RAG","R","Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/agentic-ai-is-now-on-mid-sized-indian-its-mind\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":50477,"title":"Top 11 Web Frameworks Used By Developers In 2019","content":"Web frameworks are usually used by the developers to ease the building process of web applications. They help developers to concentrate more on the task rather than the complex coding part. According to the Stack Overflow Developer Survey 2019, among nearly 90,000 developers, 63,585 developers chose the most loved web frameworks where jQuery, React.js and Angular secured the first, second and third positions respectively. In this article, we list down the 11 most popular web frameworks that have been used by the developers in 2019. (The frameworks are listed down in the order of their popularity) 1| jQuery Created in 2006, jQuery is a free and open-source JavaScript library designed to simplify HTML DOM tree traversal and manipulation, as well as event handling, CSS animation, among others. This library simplifies the interaction between JavaScript and Document Object Model (DOM). The other features of this library are cross-browser support, AJAX support, CSS3 selectors and basic XPath syntax support, among others. As of May 2019, this library has been used by 72% of the 10 million most popular websites. 2| React React is a declarative, efficient, and flexible JavaScript library for building user interfaces. This framework lets a developer compose complex UIs from small and isolated pieces of code called “components”. It can be used as a base in the development of a single-page as well as a mobile application. 3| Angular(previously Angular JS) Created by the Angular Team at Google and a group of developers community, Angular is a TypeScript-based open-source web application framework. This development platform can be used for building mobile and desktop web applications using TypeScript\/JavaScript and other languages. It uses a hierarchy of components as its primary architectural characteristic and has a different expression syntax for property and event bindings. 4| ASP.NET ASP.NET is an open-source, cross-platform framework for building web applications as well as services with .NET and C#. Created by Microsoft, this framework allows developers to build dynamic web applications. It enables real-time bi-directional communication between server and client. For building web applications, ASP.NET extends the .NET platform with tools and libraries such as web-page templating syntax, known as Razor for building dynamic web pages using C#, the base framework for processing web requests, libraries for common web patterns and other such. 5| Express Express is an open-source web application framework for Node.js. This is a minimal and flexible Node.js web application framework which provides a robust set of features for web and mobile applications. The features of this framework include allowing to dynamically render HTML Pages based on passing arguments to templates, setting up of middlewares and much more. 6| Spring Spring is an application framework written in Java programming language. This framework provides a comprehensive programming and configuration model for modern Java-based enterprise applications – on any kind of deployment platform. It includes a number of modules to provide services such as aspect-oriented programming, convention over configuration, data access, database connectivity and much more. 7| Vue Vue is a popular framework in JavaScript. It is an open-source progressive framework for building user interfaces and is designed in such a manner that the core library is focused on the view layer only, and is easy to pick up and integrate with other libraries or existing projects. 8| Django Django is a high-level Python-based open-source web development framework for backend web applications. In one of our articles, we discussed the final release of Django 3.0 which will bring a number of intuitive features. This framework is designed to satisfy the complex requirements of experienced web developers. 9| Flask Flask is a lightweight WSGI Python web application framework. This framework became widely popular as an alternative to Django projects with the monolithic structure and dependencies. It is classified as a microframework because it does not require particular tools or libraries. 10| Laravel Laravel is a web application framework with syntax written in PHP language. Some of the features of Laravel are a modular packaging system with a dedicated dependency manager, different ways for accessing relational databases, built-in authentication and authorisation, MVC support and Object-Oriented approach, Artisan Console, Eloquent ORM and other such. 11| Ruby on Rails Ruby on Rails is one of the most popular frameworks written in Ruby. This is an open-source server-side web application framework which is a model–view–controller (MVC) framework that provides default structures for a database, web pages, etc. It is a framework for building websites and it combines the Ruby programming language with HTML, CSS, and JavaScript to create a web application that runs on a web server.","excerpt":"Web frameworks are usually used by the developers to ease the building process of web applications. They help developers to concentrate more on the task rather than the complex coding part. According to the Stack Overflow Developer Survey 2019, among nearly 90,000 developers, 63,585 developers chose the most loved web frameworks where jQuery, React.js and […]","categories":[],"tags":["Developers","Web Development","Web Frameworks"],"author_name":"Ambika Choudhury","publish_date":"2019-11-22T13:00:00","publication_year":"2019","word_count":757,"keywords":["Go","AI","ML","Web Development","TypeScript","Java","Python","ViT","JAX","JavaScript","R","Web Frameworks","Developers"],"extracted_tech_keywords":["AI","ML","JAX","Python","R","JavaScript","TypeScript","Go","Java","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/top-11-web-frameworks-used-by-developers-in-2019\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":37921,"title":"10 Fastest Growing Programming Languages That Employers Demand In 2019","content":"Figuring out which is the best programming language in an organisation can be a tough decision. With the evolution of new technologies, demand for sophisticated frameworks, languages, etc. have increased among the organisation. In this article, we list down the 10 fastest growing programming languages which are the most in-demand. 1| C# C# or C Sharp was developed by Microsoft led by Anders Hejlsberg and his team within the .Net which is an open source developer platform for building various types of applications. In January 2019, the C# language team in Microsoft announced the release C#  8. 0 which includes new features such as new types and operators for collections manipulation, indexing, etc. It also has nullable reference types which help in preventing the ubiquitous null reference exceptions which have riddled object-oriented programming. Click here to know more. 2| Julia The Julia project was founded by Jeff Bezanson, Alan Edelman, Viral Shah, and Stefan Karpinski. The language aims to create an unprecedented combination of power, efficiency, etc. I also have some features such as user-defined types are as fast and compact as built-ins, no need to vectorize code for performance, designed for parallelism and distributed computation, efficient support for Unicode, including but not limited to UTF-8, etc. Julia Version 1.1 has been released recently with new features such as exception stack is maintained on each task to make exception handling more robust and enable root cause analysis, etc. Click here to read more. 3| Golang Go was developed by Robert Griesemer, Rob Pike, and Ken Thompson at Google which was launched in 2009 as an open source programming language. It is a flexible language and allows multiple processes running simultaneously, provides a rich standard library, reduced latency in the most recent versions of the garbage collector, etc. Recently, Go version 1.12 version was released which is processed on improving the debugging experience on optimised binaries. Click here to know more. 4| Java Java is one of the most widely used languages which was built with the philosophy of WORA (Write Once Run Anywhere). This popular language has some important features such as it is platform independent, fast, secure, large standard library, etc. The latest version is released in March 2019 which includes new features such as low pause-time garbage collector, switch expressions, JVM constants, microbenchmark suite, etc. Click here to know more. 5| JavaScript Initially, JavaScript had another name called LiveScript and was created to make web pages alive. Modern JavaScript is a safe programming language which does not provide low-level access to memory or CPU as it was initially created for browsers which do not require it. This language is able to add new HTML to the page, change existing content, etc., react to users actions, send requests over the network to remote servers, get and set cookies, etc. javaScript Version ECMAScript 2019 is going to be published this year in June. Click here to know more. 6| TypeScript TypeScript is a language for application-scale JavaScript. TypeScript adds optional types to JavaScript that support tools for large-scale JavaScript applications for any browser, for any host, on any OS. It basically compiles to readable, standards-based JavaScript. Recently, Microsoft released TypeScript 3.4 which includes a new feature called as “incremental” which aims to help the developers type-check and output changes to an existing project. Click here to know more. 7| Python One of the most widely used and popular languages, Python is a general, high-level language initially designed by Guido van Rossum. The programming language was developed for emphasis on code readability, and its syntax allows programmers to express concepts in fewer lines of codes. Python 3.7.3 was recently released on March 25th, 2019 which is a major release of this language and contains many new features. Click here to know more. 8| Scala Scala combines object-oriented and functional programming in one brief. The static types of this high-level language help in avoiding bugs in complex applications and its JVM and JavaScript runtimes which let you build high-performance systems with easy access to huge ecosystems of libraries. Scala 2.13.0 will be released shortly with new and interesting features. Click here to know more. 9| R R is a language and environment for statistical computing and computing. This language can be considered as a different implementation of S and provides a wide variety of statistical techniques such as linear and non-linear modelling, classical statistical tests, time-series analysis. Recently, R version 3.5.3 was released in March 2019. Click here to know more. 10| Swift The Swift language is one of the fastest growing programming languages because it makes easy to write software which is incredibly fast and safe by design. This general -purpose, multi-paradigm, compiled programming language is developed by Apple Inc. It includes modern programming patterns such as array indices are checked for out-of-bounds errors, integers are checked for overflow, error handling allows controlled recovery from unexpected failures, etc. Swift 5.0 was recently released in March 2019 which is a major release with several new features such as raw strings, checking for integer multiples, etc. Click here to know more.","excerpt":"Figuring out which is the best programming language in an organisation can be a tough decision. With the evolution of new technologies, demand for sophisticated frameworks, languages, etc. have increased among the organisation. In this article, we list down the 10 fastest growing programming languages which are the most in-demand. 1| C# C# or C […]","categories":["AI Trends"],"tags":["C","Golang","Java","Javascript","Julia","Julia Language","Programming Languages","Python","R","SWIFT","TypeScript"],"author_name":"Ambika Choudhury","publish_date":"2019-04-17T12:57:12","publication_year":"2019","word_count":848,"keywords":["Go","Golang","TPU","C","SWIFT","AI","Javascript","ML","JavaScript","Programming Languages","TypeScript","Python","Julia Language","Aim","Ray","Julia","R","Java"],"extracted_tech_keywords":["AI","ML","Aim","Ray","TPU","Python","R","JavaScript","TypeScript","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/10-fastest-growing-programming-languages-that-employers-demand-in-2019\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10063954,"title":"Salesforce just launched a Github Copilot killer","content":"Salesforce has unveiled CodeGen, a large scale language model which turns simple English prompts into executable code. You just have to describe what the code should do in natural language, and the machine writes it for you. CodeGen combines conversational AI (interactive human-to-machine dialogue) and automatic programming (the system automatically creates the program based on a higher-level language). Salesforce’s implementation of conversational AI programming provides a glimpse into the future of democratising software engineering for the masses. An “AI assistant” translates English descriptions into functional and executable Python code – allowing anyone to write code, even if one knows nothing about programming. The underlying language model, CodeGen, enables this conversational paradigm and will be made available as open source to accelerate research.Salesforce AI Research trained CodeGen, a 16-billion parameter auto-regressive language model, on a large corpus of natural and programming languages. CodeGen can be applied to both simple and complex problems, using natural language. With CodeGen, most users can solve relatively simple coding problems with little or no prior programming knowledge. More complex cases may require some knowledge of programming or basic computer science concepts, in order to help guide the system as it searches for a solution). Still, even for experienced coders, CodeGen makes getting to a functioning solution faster and easier, and allows rapid exploration of alternate methods.","excerpt":"Salesforce just launched a Github Copilot killer","categories":["AI News"],"tags":["AI programming","AI Research","Salesforce"],"author_name":"Kartik Wali","publish_date":"2022-03-30T13:48:05","publication_year":"2022","word_count":220,"keywords":["AI programming","API","programming_languages:R","AI","AI Research","Python","Salesforce","programming_languages:Python","AI research","R"],"extracted_tech_keywords":["AI","Python","R","API","AI research","programming_languages:Python","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/salesforce-just-launched-a-github-copilot-killer\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10086051,"title":"Industry’s Self Destructive Behaviour Intensifies","content":"Belt tightening has become a new priority in the tech industry. Joining the mass layoff club, tech juggernaut, IBM has announced laying off 3,900 employees. A disturbing trend is visible in 210 tech companies that laid off 67,268 employees in January 2023 at an average of 2,500 employees per day. For Silicon Valley, 2022 was a year of retrenchment for much of the tech industry. It was a major deviance from the perpetual boom sector big behemoths underwent as they ferociously staffed up their empires. Yet, even as companies faced regulatory scrutiny, they established themselves as job creators. The tumbling economic environment has led the tech industry to introspect and focus on its core strengths. But that wasn’t all; smaller companies bore similar consequences: Kraken, the cryptocurrency exchange, laid off 1,100 employees; Noom, the health and fitness app, axed 1,095 people; and payments company Stripe also laid off 1,050 employees. The Digital Clock Strikes Midnight Layoffs have been steadily on the rise since the 1970s. For example, in 1979, less than 5% of Fortune 100 companies laid off their employees, according to McMaster University sociology professor Art Budros, but in 1994 around 45% did the same. A McKinsey report of 2,000 U.S. companies revealed that during the Recession and its aftermath, between 2008 and 2011, 65% chose to let go of their employees as the best option. The layoff phenomenon has become a default response, and there are several reasons for it, according to these companies. German software giant SAP recently announced axing 3000 employees as it looks to cut costs and focus on its cloud business. Moreover, tech companies were especially prone to overhiring as the economy was recovering from the impact of the pandemic. According to Axios, Microsoft, Meta, and Alphabet (Google’s parent) expanded their staff by 20% in the 12 months that ended September 30, 2022. According to Reuters, Marc Benioff, the co-Chief Executive of Salesforce, wrote a letter explaining the current turbulence in the industry and taking responsibility for the “economic downturn the company is facing”. Microsoft, the software giant, also confirmed that it would lay off 10,000 employees in FY 2023. In the letter to employees after the announcement of layoffs, Satya Nadela mentioned a possible recession as the reason. Spotify also planned to cut 6% of its workforce citing the same logic. Several new investment initiatives by Silicon Valley’s tech titans are unprofitable. For example, Amazon’s robotics division, Microsoft’s virtual reality and metaverse division ‘AltspaceVR’, and Meta’s Substack competitor called ‘Bulletin’ are all verticals that are futuristic to a certain degree and require a lot of investments but also burn a lot of money. The cash burn has also been why companies are trying to cut costs. Fund managers and early investors in big tech companies are also pushing the company management to make quick decisions to counter the company growth slowdown. In October last year, Meta’s founder and CEO, Mark Zuckerberg, received a letter from Altimeter Capital Chair and Chief Executive Brad Gerstner, who advised job cuts and streamlining operations. Expect the Unexpected With dozens of companies trying to find ways to stay afloat in the sinking economy, the majority of the laid-off employees are left with a list of questions about who to approach. In The Disposable American, a New York Times journalist and author, Louis Uchitelle, argued that the working world need not be so cruel. Drawing a comparison, he said layoffs should come with a warning label like cigarettes. The scare of job losses should shift from workers to the executives, who see human suffering as the unfortunate cost of doing business. ‘Sorry, we have to let you go’ might have been the most dreaded phrase of 2022, and experts anticipate the number of pink slips handed to not dial down anytime soon in 2023. The number has already surpassed the Great Recession the world went through, which began with Lehman Brothers collapse in 2008–2009.","excerpt":"In January 2023, 210 tech companies laid off 67,268 employees.","categories":["IT Services"],"tags":["IBM","Mass layoffs","Meta","meta layoffs","SAP","tech layoff"],"author_name":"Tasmia Ansari","publish_date":"2023-01-28T10:00:00","publication_year":"2023","word_count":655,"keywords":["Go","API","Meta","ELT","AI","meta layoffs","tech layoff","SAP","Mass layoffs","ML","Git","RPA","RAG","IBM","GAN","R"],"extracted_tech_keywords":["AI","ML","RAG","R","Go","Git","API","ELT","GAN","RPA"],"url":"https:\/\/analyticsindiamag.com\/it-services\/industrys-self-destructive-behaviour-intensifies\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10011720,"title":"AIM Announces The Launch Of Third Edition Of Machine Learning Developers Summit &#8211; MLDS 2021","content":"Analytics India Magazine announces the launch of Machine Learning Developers Summit 2021 (MLDS21), India’s leading conference exclusively for the machine learning practitioners ecosystem. The three-day event from 11-13 February 2021 will be in a virtual setting this edition. The field of machine learning is witnessing increased use cases across industries and an expanded role in enterprises, especially in the post-COVID world. MLDS brings India’s leading machine learning innovators and practitioners to share ideas and experience about machine learning tools, advanced development in this sphere to give attendees a first look at new trends & developer products. MLDS21 will allow attendees to explore the potential of machine learning for their business and gain best practices from cutting-edge presentations. It will help them explore new tools and technologies in the field while letting developers get hands-on exposure to the latest skills in ML technology while gaining network opportunities with leading companies. The 3-day conference will give participants direct access to top ML innovators from leading tech companies. They will cover various aspects of the machine learning field — from software architecture of ML systems to producing and deploying the latest ML frameworks and solutions for business use cases. Register here for MLDS2021 Paper Presentation MLDS21 will host paper presentations from the ML community to promote research in the space and foster scientific exchange between researchers, practitioners, scientists, academicians, and engineers. It encourages papers from all subareas of AI in a wide range of application domains such as healthcare, sustainability, transportation, and commerce. The technical paper presentation is highly sought after at MLDS, and have been received well in the research community. The submissions for 2021 are open and interested participants can apply here. The selected papers will be published in Lattice, The Machine Learning Journal by the Association of Data Scientists. 40 Under 40 Awards These awards recognise young data scientists in India who have made significant contributions in the field of machine learning and have transformed data into meaningful insights by solving most complex challenges. This award recognises and distinguishes the data scientists, real innovators and achievers of the analytics industry and separates them from the rest of the crowd. Nominees are carefully reviewed and selected by our team of editors and industry veterans. The nominations for 2021 are open, and winners will be recognised at the annual Machine Learning Developers Summit 2021 to be held in Feb 2021, virtually. It is open for data science and ML professionals ranging from CXOs and head of Analytics. Who Should Attend MLDS is the first-of-its-kind conference to cover both the technical and business aspects of this emergent technology. The program will be packed with learning and networking sessions, well-suited for both developers and C-suite executives to understand how this technology will make the biggest impact this year, both in terms of career and business prospects. MLDS21 is expected to see participation from 200 organisations, 1500 ML developers and more than 30 speakers featuring 50+ focused sessions, keynotes and workshops by some of the foremost minds in the ML industry. Highlights Of MLDS21 It will allow participants to network in a closed virtual setting to the leading machine learning researchersIt will provide access to ML experts to help answer specific questions and discover new trends in the fieldIt will let participants apply the best practices to workIt will allow participants to test drive and explore the latest machine learning technology and tools with hands-on setting Date: 11-13 Feb 2020 Where: Virtual Read more about MLDS21 here.","excerpt":"Analytics India Magazine announces the launch of Machine Learning Developers Summit 2021 (MLDS21), India’s leading conference exclusively for the machine learning practitioners ecosystem. The three-day event from 11-13 February 2021 will be in a virtual setting this edition.  The field of machine learning is witnessing increased use cases across industries and an expanded role in […]","categories":["Deep Tech"],"tags":["big data developer skills","Machine Learning","machine learning developer summit"],"author_name":"Srishti Deoras","publish_date":"2020-11-17T10:00:07","publication_year":"2020","word_count":582,"keywords":["data science","machine learning","big data developer skills","programming_languages:R","AI","ML","Machine Learning","RAG","analytics","GAN","machine learning developer summit","R"],"extracted_tech_keywords":["AI","machine learning","ML","data science","analytics","RAG","R","GAN","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/aim-announces-the-launch-of-third-edition-of-machine-learning-developer-summit-mlds-2021\/","complexity_score":4,"technical_depth":9,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":29846,"title":"Unified Marketing Models: Jointly optimising Advertising Efficiency And Effectiveness","content":"Half the money I spend on advertising is wasted; the trouble is I don’t know which half – John Wanamaker (1838-1922), a pioneer in introducing department stores Advertising efficiency and effectiveness is a subject that has long engaged marketers. The impact of each dollar spent on advertising is measurable and attributable now more than ever, and the advertising is transforming from an art to a pretty exact, quantitative science. The real challenge lies in being able to mine this data accurately and utilise it to drive marketing productivity. Marketing analytics assists in making decisions on the allocation of marketing budgets across different tactics and campaigns The goal of advertising is to make sure firms reach out to the target audience and engage with them at a reasonable cost, ultimately leading to increased conversions and sales. Budget limitations are a reality, and the key to successful advertising is allocating the advertising spend across different media channels and promotional tactics, both above-the-line and below-the-line advertising. There are two distinct tools that measure different aspects of the marketing activity, well known to marketers – the Market Mix Modelling and the Multi-touch attribution models. This article discusses how these two different tools measure different types of metrics – efficiency and effectiveness – of marketing activity, and how a unified measure can be built using both these tools, leading to better across-the-funnel marketing outcomes. Market Mix Modelling The Marketing Mix Modelling is a gold standard in the marketer’s toolkit for measuring tactic Return on investment. Market Mix modelling (and a limited version of this, media mix modelling) is a mechanism to allocate budgets efficiently across different marketing tactics through exploring historical relationships between business drivers and outcomes such as sales or traffic while accounting for changes in the business environment including competitor price strategies and regulatory changes. Many internal and exogenous factors are included in the market mix model – sales, revenue, base sales, ad budgets by media tactic, sale promotions, traffic are some internal factors; competitor price, ad budgets, new store activity, other drivers of sales; population parameters, seasonality and trends could be some exogenous factors that are introduced in the model. The outcome of the market mix model is to determine the impact of spends on various media and promotion activities and to re-allocate spends across these different tactics; at the end of this exercise, the media or tactic with the highest impact on sales (ROI) obtains the largest increase in budgets. MMM models provide a ‘top-down’ view of budget allocation based on long-term data collection and macro-information, leading to a stable and reliable view of how marketing and media tactics’ ROIs change over time. However, while this provides stability to planners, MMM does not address the need for quick insights that are required for dynamic allocations in the digital space. Enter the Attribution models, which are fine-tuned to make sense of digital assets and tactics which are highly dynamic, have a lifetime of a couple of weeks to some months, and where sales are not the primary outcome. Attribution models Attribution models help determine the effectiveness of digital tactics in non-sales conversion activities. Customers pass through multiple digital touchpoints as they move through a funnel of engagement. At the initial phase is the clicking on a campaign and visiting a web page. As engagement levels continually increase there are repeat visits to the website, engagement across different pages, search for new information, and sign up for updates or campaigns. Depending on the site characteristics, they may complete a purchase, ask for specific actions such as callbacks, engage in a chat etc. Guiding customers through the funnel depends completely on the marketers’ skill and design effectiveness- web pages that have links to immediate information, that answers questions and navigate the customer to the right path are likely to be more effective at reaching their end goal of conversions. Broken links, web pages that are not aligned with each other, having to spend a lot of time in searching for key links or in-your-face forceful engagement is likely to put people off, and prospects would simply vote with the mouse to exit the website. With a large number of digital touchpoints that a customer traverses, there is a need to identify those tactics that lead to conversion. Each touch or engagement tactic is relevant, but some are more critical than others in the eventual conversion. Multi-touch attribution (MTA) is a tool to measure the critical path of activities that lead to a conversion. This allows marketers to allocate budgets on those digital touchpoints that lead to higher engagement or activity at the ‘top of the funnel’. Attribution models are typically explained with a soccer analogy. There are multiple players in the field and each one has a certain role- the striker is the one who scores (converts), and there are players who assist and pass the ball. How well the team passes and plays on the field determines how they score. The critical path of players that the ball passes through before eventually converting to a goal is the ‘conversion path’. There are different kinds of attribution models, but the state-of-art is currently the multi-touch attribution where the relative value of each digital touchpoint is ascertained using a model-based approach. Since multi-touch attribution deals with non-sales conversion, it is a measure of effectiveness – how effective is the digital channel in its call to immediate action. However, neither attribution models nor market mix models individually solve the entire puzzle. While MMM helps optimise spends to achieve a certain revenue, MTA models assist in identifying, in the digital space, the specific set of activities and triggers that lead to conversion. How do marketers bridge these medium-term, high-level insights at the market level to short-term, daily level insights at the customer level? Unified Models ‘Unified Models’ bridge the gap between the half-yearly or quarterly MMM which utilises offline and online information, and attribution models that are very short term in impact and are solely based on online channels. Unified Models provide visibility across the entire customer decision journey, combining true omnichannel marketing through both online and offline. Marketers can combine the best of MMM and MTA then – identify which messages and digital assets are the most impactful on the consumer level within the broader marketing context, combining both internal and external data. In terms of impact, Unified models provide both stable estimates that allow for quarterly or half-yearly planning, as well as reasonable insights at a weekly level. By providing real time campaign analysis based on an integrated data view combining both aggregate and individual granularity, Unified Models lead to higher marketing effectiveness and efficiency.","excerpt":"Half the money I spend on advertising is wasted; the trouble is I don’t know which half – John Wanamaker (1838-1922), a pioneer in introducing department stores Advertising efficiency and effectiveness is a subject that has long engaged marketers. The impact of each dollar spent on advertising is measurable and attributable now more than ever, […]","categories":["AI Features"],"tags":["Advertising","Data Analytics","digital marketing"],"author_name":"Madalasa Venkataraman","publish_date":"2018-11-02T06:55:20","publication_year":"2018","word_count":1110,"keywords":["Go","programming_languages:R","AI","Advertising","programming_languages:Go","Git","digital marketing","ViT","analytics","Data Analytics","R"],"extracted_tech_keywords":["AI","analytics","R","Go","Git","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/unified-marketing-models-advertising\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10014511,"title":"All The Libraries Launched At AWS re:Invent 2020, So Far","content":"At the ongoing annual cloud conference, AWS re:Invent event, the e-commerce giant has been announcing a staggering number of product launches, tools, frameworks, services and more. From AI chips to enterprise applications, the event is covering all the aspects and addressing needs of a developer to build innovative products. Here is a list of six libraries and tools, in no particular order, that was launched so far at the AWS re:Invent 2020. SageMaker Distributed Data-Parallel (SDP) AWS re:Invent 2020 introduced two new distributed training libraries for Amazon SageMaker, one of them is the SageMaker Distributed Data-Parallel (SDP). The libraries are meant to provide integrated methods for the users to quickly train large deep learning models. SageMaker distributed data-parallel extends SageMaker’s training capabilities on deep learning models with near-linear scaling efficiency in order to achieve fast time-to-train with minimal code changes. Some of the benefits of this library are — SDP optimises the training job for AWS network infrastructure and Amazon EC2 instance topology. SDP takes advantage of gradient updates to communicate between nodes with a custom AllReduce algorithm. SageMaker Distributed Model Parallel (SMP) The second library announced as the distributed training library for Amazon SageMaker is SageMaker Distributed Model Parallel or SMP. Amazon SageMaker distributed model parallel is a model parallelism library that can be used for training complex deep learning models which were previously challenging to train because of the limitations in GPU memory. Some of its benefits are — This library automatically as well as efficiently splits a model over multiple GPUs and instances, and coordinates model training. This allows users to increase the prediction accuracy by creating massive models with extended parameters. One can use SMP to automatically create partitions in the existing TensorFlow and PyTorch workloads across multiple GPUs with minimal code changes. AWS IoT SDK Includes OTA Library AWS re:Invent 2020 event also announced AWS IoT Device SDK for Embedded C (C-SDK) version 202012.00, which now includes an over-the-air update (OTA) library and a PKCS #11 implementation (corePKCS11). According to a blog post, the OTA library makes it easier to manage notifications, download, and perform cryptographic verification of firmware updates. The OTA and corePKCS11 libraries have been optimised for memory usage and modularity, and have undergone code quality checks. AWS IDE Toolkit For AWS Cloud9 In another announcement, re:Invent 2020 launched the AWS Toolkit for AWS Cloud9 to enable users of the browser-based IDE to efficiently handle the core AWS services by the graphical user interface. A key component of this toolkit is the Resource Explorer, which is a view that permits users to navigate as well as interact with the popular AWS resources such as AWS Lambda, Amazon API Gateway, Amazon S3, among others. Also, the AWS Toolkit extends AWS Cloud9’s support for AWS Lambda functions, allowing users to quickly invoke, import, deploy or delete functions. Amazon Kendra Launches Connector Library The connector library for Amazon Kendra was also launched to make it easy to synchronise data from multiple content repositories. Connectors can be scheduled to automatically sync indexes with the data source, such that one can securely search through the most updated content. Also, the Amazon Kendra connector library offers native connectors, as well as an expanding list of partner connectors from AWS partners like Raytion. Amazon Braket Now Supports PennyLane Amazon Braket announced the support for PennyLane, which is an open-source software framework designed for hybrid quantum computing. Pennylane provides interfaces to common ML libraries, including PyTorch and TensorFlow. The integration of this framework with Amazon Braket allows users to test and fine-tune algorithms faster at a larger scale on scalable and fully managed simulators and run them on the choice of quantum computing hardware. Keep checking this space for more updates.","excerpt":"At the ongoing annual cloud conference, AWS re:Invent event, the e-commerce giant has been announcing a staggering number of product launches, tools, frameworks, services and more. From AI chips to enterprise applications, the event is covering all the aspects and addressing needs of a developer to build innovative products. Here is a list of six […]","categories":["AI Trends"],"tags":["AWS"],"author_name":"Ambika Choudhury","publish_date":"2020-12-17T13:00:00","publication_year":"2020","word_count":620,"keywords":["Go","Amazon SageMaker","AWS","AI","PyTorch","ML","Ray","deep learning","TensorFlow","R"],"extracted_tech_keywords":["AI","ML","deep learning","Amazon SageMaker","Ray","TensorFlow","PyTorch","AWS","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/all-the-libraries-launched-at-aws-reinvent-2020-so-far\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10003449,"title":"How Will SpiceJet Use The Deep Tech By Acqui-hiring Travenues","content":"SpiceJet recently acqui-hired the team and technology of Travenues, a subsidiary of Ixigo which is India’s leading AI-based travel app. With this deal, the amount of which was not disclosed, SpiceJet aims to absorb the technology team and airline commerce platform. Last year, Travenues had signed its first technology partnership with SpiceJet for the digital transformation of its consumer-facing experiences, which have now been expanded to other applications. The SpiceJet team firmly believes that the acqui-hire will help SpiceJet strengthen its e-commerce platforms and infuse tech in other applications such as mobile apps, UX, engagement, cross-selling, payments, ancillaries, personalisation and more. “This acqui-hire will help SpiceJet strengthen its e-commerce platforms as we continue to innovate across multiple technology areas and achieve our vision of being the worldwide leader in aviation technology,” said Ashish Vikram, Chief Technology and Innovation Officer, SpiceJet. How Will Travenues Facilitate Deep Tech To SpiceJet Bengaluru-based Travenues is an airline technology company that has built several innovative products for airline eCommerce. Founded in 2019, the team has extensive experience across GDS, OTA and meta-search business to improve passenger customer experience. It offers a comprehensive travel-tech optimised airline commerce and ancillary sales platform to airlines that allows for extensive customisation and state-of-the-art personalisation. Powering the travel industry through deep technologies such as artificial intelligence and machine learning, it applies cutting edge technology to provide seamless solutions to a wide range of travel-tech problems. Some of the solutions that it offers are: AI\/ML-based travel assist, which is a travel assistance bot that can help with consumer support and deeply integrated booking optionsHighly configurable micro engagement platform capable of reaching the right traveller with the right deal at the right time The team believes that airlines still face the same problems that they faced a decade ago. Infusing technology can enable them at various levels, giving travellers the best experience. With this deal, Spicejet aims to use the technology-driven team to help them grow in various domains such as eCommerce, offering customisation and more. It will use the team and full-stack commerce suite that Travenues has built and take innovations to the next level. It will help them focus on B2C business while scaling up B2B offerings though Travenues. It also aims to bring deep-tech integrations to help airlines with user experience on engagement, segmentation, targeting, cross-selling, payments, and customer service. “We are happy that we were able to incubate a startup and build a next-generation platform with a motivated tightly-knit team that can truly disrupt airline direct sales and airline commerce. The possibilities this unfolds for SpiceJet are endless!, said Rajnish Kumar, Co-founder & CTO, ixigo” ixigo Shares Its Ambition To Use Deep Tech With Travenues ixigo has been deeply invested in using AI and predictive analytics to offer a conducive platform to travellers. In an interaction with Analytics India Magazine, Rajnish had said that by using machine learning they were able to predict a lot of travel-related aspects for the user, making the booking experience hassle-free. They also use it to predict seat confirmation, seat availability, flight fares, train delays and much more using millions of data points that they capture from apps. Using deep tech, it evolved from being a metasearch engine to travel marketplace. With Travenues, Ixigo shares a similar ideology and aims to transform the way travel and airlines work.","excerpt":"SpiceJet recently acqui-hired the team and technology of Travenues, a subsidiary of Ixigo which is India’s leading AI-based travel app. With this deal, the amount of which was not disclosed, SpiceJet aims to absorb the technology team and airline commerce platform.  Last year, Travenues had signed its first technology partnership with SpiceJet for the digital […]","categories":["AI Features"],"tags":[],"author_name":"Srishti Deoras","publish_date":"2020-07-28T12:00:00","publication_year":"2020","word_count":555,"keywords":["Go","machine learning","artificial intelligence","AI","R","ML","Git","Aim","analytics","predictive analytics"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","analytics","Aim","predictive analytics","R","Go","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-will-spicejet-use-the-deep-tech-by-acqui-hiring-travenues\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10104056,"title":"ByteDance to Launch Platform to Build Custom Chatbots","content":"ByteDance, the Chinese firm behind TikTok, is in the process of creating an open platform that enables users to construct their own chatbots, marking its entry into the generative AI market, similar to OpenAI GPTs. An internal memo obtained by SCMP reveals that the anticipated launch of the “bot development platform” as a public beta is slated for the end of the month. Additionally, ByteDance is reportedly engaged in the development of a text-to-image generator akin to Midjourney. These endeavors are indicative of the company’s strategic vision to seamlessly integrate novel generative AI products with its existing portfolio. This comes just a few days after OpenAI had delayed its plan to launch a custom GPT store until early 2024, according to reports. The internal memo said that the company is making improvements to its GPTs based on customer feedback. This came after the return of Sam Altman as the CEO of OpenAI, a week after a surprise firing. ByteDance is not the only company from China that is developing generative AI models. Most recently, companies such as Alibaba, Tencent, Huawei, Baidu, and 01.AI have developed their own models to compete in the space. The open source AI community is also increasingly dominating in China with models like DeepSeek and Qwen being open sourced on GitHub and Hugging Face. Read: Beware of Chinese Open-Source LLMs On the other hand, OpenAI has not made its AI models available in China. Similarly, Google has also refrained from releasing its models in the country. This has given China to develop models for its own people.","excerpt":"This comes just a few days after OpenAI had delayed its plan to launch a custom GPT store until early 2024.","categories":["AI News"],"tags":["AI Tool"],"author_name":"Mohit Pandey","publish_date":"2023-12-04T15:09:00","publication_year":"2023","word_count":261,"keywords":["Go","Hugging Face","OpenAI","AI","chatbots","ML","Git","generative AI","AI Tool","GitHub","R"],"extracted_tech_keywords":["AI","ML","generative AI","OpenAI","Hugging Face","chatbots","R","Go","Git","GitHub"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/bytedance-to-launch-platform-to-build-custom-chatbots\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":39688,"title":"Why Dharma Ransomware Is More Dangerous Than Ever","content":"Lately, ransomware has become very popular among hackers. It not only causes severe downtime but also leads to a significant amount of data loss. Ransomware is basically is a malware that is used to encrypt user data and block access to it. The data can only be accessed a ransom is paid to the hacker or the ones behind the attack. Over the past couple of years, ransomware has evolved and today it has several variants. Also, the main reason behind its popularity is its effectiveness — it is practically very difficult to break the encryption and sometimes it’s even impossible. Dharma is one such ransomware that is considered to be one of the most notorious.   Since 2006, this ransomware is continuously evolving and has become increasingly active lately. According to a source, it has increased by a margin of 148% from February 2019 to April 2019. In this article, we are going to have a deep look at this filthy ransomware — how it works and why it is one of the most dangerous ransomware in the family. How does Dharma Ransomware Work? Dharma is a family of encryption ransomware Trojan that has compromised numerous computers all across the world till date. This ransomware targets mainly directories inside the Users directory on Windows. Every time a file is added to the directory, this malicious thing encrypts the file and adds a suffix [bitcoin143@india.com].dharma. One of the unique things about Dharma is that it doesn’t affect the entire computer, but it hides inside the system and keeps encrypting files every time they are added to the directory. So basically, one has to remove it in order to decrypt the files. The ransomware usually has a ransom not; however, it changes depending on the variant. But you must be wondering — how this ransomware ends up being inside a computer? So, what basically happens is, the ransomware is spread across the world through email campaigns claiming to be legit (the email is usually about being the Windows machine under risk) and asking the user to download a password protected attachment named Defender.exe. Talking about the password, it is listed in the email itself along. The entire process is so effective that numerous people over the years have ended downloading it. That is not all, the real game starts when the user executes the downloaded file. It is basically a self-extracting archive that drops the malicious file called taskhost.exe along with an old version of ESET AV Remover renamed as Defender_nt32_enu.exe. Once the extraction is done, the ESET AV Remover installer automatically launches and makes the victim feel that the entire process is legit and distracts him\/her from noticing Dharma encrypting the contents of the hard drive in the background. Pulling out such an effective trick through email campaign is not something really easy. No doubt, the ransomware is notorious and dangerous enough. But, the hackers behind are seemed wittier and experts of social engineering as it’s not just technology behind Dharma but the convincing campaigns that lead to the installation of malicious software. How To Stay Safe? As we know that email campaign is one of the major ways of distributing this notorious ransomware, so the first thing we need to do is to see that the email we are receiving is coming from an authentic source. Also, if the attachment is a tool that would sort out the problem in your system then check the tool is legit, updated — a legit vendor would never distribute an outdated tool. If that doesn’t work for you and you still want to keep your files safe, then adopt the habit of backing up files. Back up is considered to be one of the best practices in cybersecurity. So, even if your files get affected or encrypted by any ransomware, you always have a different set to work. Cyber-attacks will keep happening as the technology is not only empowering innovative organisations, but also the wrongdoers. Be prepared to tackle cyber threats — prepared enough to at least mitigate the consequences.","excerpt":"Lately, ransomware has become very popular among hackers. It not only causes severe downtime but also leads to a significant amount of data loss. Ransomware is basically is a malware that is used to encrypt user data and block access to it. The data can only be accessed a ransom is paid to the hacker or the […]","categories":["AI Features"],"tags":["Cybersecurity","Hack","Ransomware"],"author_name":"Harshajit Sarmah","publish_date":"2019-05-26T07:53:05","publication_year":"2019","word_count":677,"keywords":["Go","programming_languages:R","AI","Ransomware","programming_languages:Go","Hack","Git","Aim","GAN","Cybersecurity","R"],"extracted_tech_keywords":["AI","Aim","R","Go","Git","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/why-dharma-ransomware-is-more-dangerous-than-ever\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10503,"title":"Artificial Intelligence now forays into Venture Capital","content":"A traditional Venture Capital (VC) firm would have to go through umpteen number of meetings, listen to a variety of ideas before taking a decision to make an investment in a particular project. Also how does any VC make the investment decision? Well, it is obviously based on certain knowledge of that industry\/ project but there is also gut feeling, it’s a mind and heart deciding together. But can you imagine a situation where all decisions are made only by brain. Well AIVC is one such example where this venture capital is powered by Artificial Intelligence. AIVC is the world’s first ever venture capital firm which is 100% powered by an AI. Off late we are seeing AI venturing into every field and hence venture capital comes with little or no surprise. The way AIVC works is purely on data i.e. it is completely driven by data. So when a company approaches AIVC for investment, AIVC makes use of historical data and advanced predictive algorithms to make any decision with respect to investments. AIVC has been fed with all the venture capital deals done ever and updates itself on a real-time basis. It uses this information to take investment decisions. Hence the decisions made by AIVC are purely based on data and does not involve any other external factor. This venture capital web portal aims at being the mid-point between artificial intelligence and venture capitalism. AIVC also has advanced Unicorn Identification Capabilities. To take investment decisions, AIVC like any other venture capital firm needs information and AIVC gets this information by asking questions.  So when one enters the web portal, one get asked questions but in a unique manner. Well the makers of this AI have created questions in a snarky, funny way or some might even find it rude. For instance, when u press on ‘Start your pitch’, you are asked “What is the highly differentiated, uniquely memorable and mildly ridiculous name you’ve given your company?” Another question being “Is your product for irrational human beings or evil, monolithic corporates?” So the visitors are treated with these type of funny, rude questions just to be on the lighter side and to joke with the visitors. The idea behind the website is to create some humour and the entire website is designed with such witty and snarky questions and statements. All said, the question remains whether AI is the future for venture capital. Is this a threat to existing and traditional venture capital firms? Only time will help us get answers to these questions. AIVC has been recently launched and is still in its initial stages. So only with time, it will be revealed whether AIVC is a success or no.","excerpt":"A traditional Venture Capital (VC) firm would have to go through umpteen number of meetings, listen to a variety of ideas before taking a decision to make an investment in a particular project. Also how does any VC make the investment decision? Well, it is obviously based on certain knowledge of that industry\/ project but […]","categories":["IT Services"],"tags":["AI (Artificial Intelligence)","venture capital"],"author_name":"Manisha Salecha","publish_date":"2016-07-29T07:56:13","publication_year":"2016","word_count":451,"keywords":["Go","API","artificial intelligence","unicorn","programming_languages:R","AI","venture capital","programming_languages:Go","Aim","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","Aim","R","Go","API","unicorn","venture capital","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/artificial-intelligence-now-forays-venture-capital\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10123372,"title":"SAP Partners with Mistral AI to Enhance Enterprise Software with Generative AI Solutions","content":"SAP has announced a partnership with Mistral AI to integrate advanced AI capabilities into its applications and enterprise software. This collaboration will enable SAP customers to access Mistral AI’s models through SAP Business Technology Platform (SAP BTP) applications with generative AI capabilities. Mistral AI is known for its open-weight models, Mixtral 8x7B and Mixtral 8x22B, as well as its enterprise-grade “Large” model. These models will be available through SAP’s generative AI hub in SAP AI Core, offering SAP customers the tools to enhance productivity, streamline operations, and accelerate digital transformation. “We are excited about entering a partnership with Mistral AI and making the company’s LLM accessible to both our developers and our customers through the generative AI hub in SAP AI Core on SAP BTP,” said Philipp Herzig, Chief AI Officer, SAP. “Together, we can truly make a difference by building AI-enabled solutions that create immediate value for users, organsations, and entire industries. We are particularly proud that two European technology companies are collaborating on bringing AI forward,” he added. “We are pleased to embark on this partnership with SAP. We foresee the new horizons this collaboration will open up, enabling us to further our mission of making AI accessible to all. We are looking forward to witnessing the potential of our AI models to support innovation and streamline operations for SAP’s customers.” said Arthur Mensch, CEO of Mistral AI. Mistral AI recently raised $640 million  (€600 million) at a valuation of $6 billion led by investor General Catalyst. “We are announcing €600M in Series B funding for our first anniversary.  We are grateful to our new and existing investors for their continued confidence and support for our global expansion. This will accelerate our roadmap as we continue to bring frontier AI into everyone’s hands,” said Mensch.","excerpt":"This collaboration will enable SAP customers to access Mistral AI’s models through SAP Business Technology Platform (SAP BTP) applications with generative AI capabilities.","categories":["AI News"],"tags":["mistral ai","SAP"],"author_name":"Siddharth Jindal","publish_date":"2024-06-12T15:48:42","publication_year":"2024","word_count":297,"keywords":["funding","mistral ai","AI","innovation","SAP","ML","digital transformation","Git","ViT","generative AI","GAN","R"],"extracted_tech_keywords":["AI","ML","generative AI","R","Git","GAN","ViT","digital transformation","innovation","funding"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/sap-partners-with-mistral-ai-to-enhance-enterprise-software-with-generative-ai-solutions\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10162840,"title":"OpenAI Files Patent to Develop Smart Watches, Smart Jewellery &amp; Wearable Cameras","content":"AI startup OpenAI has filed a patent in the United States, indicating that the company is exploring the development of an array of products. These include smart watches, smart bands, smart jewellery, wearable computers and cameras, and laptop and mobile phone cases. Furthermore, the patent also lists “user-programmable humanoid robots” and “humanoid robots having communication and learning functions for assisting and entertaining people”. The patent application’s status is awaiting examination, which means that it has been accepted as meeting the minimum filing requirements and is in the early stages of processing. According to the document, the patent was filed on January 31. Notably, companies file broad patents to secure intellectual property rights, but not all patented innovations make it to the market. Therefore, it will be interesting to see how many of these products end up on OpenAI’s product development roadmap and, eventually, the market. A report from InQuartik revealed that as of December 20, 2024, 25 patents have been publicly disclosed. This includes 21 US patents and four World Intellectual Property Organisation (WIPO) applications. Of these, 14 US patents have already been granted. Last month, OpenAI explicitly revealed its plans to develop humanoids. For the first time, the company listed job roles related to robotics and hardware on its website. Announcing the same on X, Caitlin Kalinowski, a member of OpenAI’s technical staff, said, “I just hit my two-month mark at OpenAI and can confirm the talent density, work velocity, and focus make this a really wonderful place to do technical work.” The company is currently hiring a mechanical product engineer, a senior research engineer, and a systems integration electrical engineer, all of whom are focused on robotics. As per reports, OpenAI is also building a hardware device which has the potential to “replace smartphones”. Sam Altman, CEO of OpenAI, also revealed that the company is planning to build the device in partnership with former Apple chief design officer Jony Ive. Last year, it was also reported that the partnership between OpenAI and Ive had been in the works to create a device that uses generative AI to handle complex user interactions more efficiently than traditional software. In other news, OpenAI has also launched deep research, a new capability in ChatGPT that independently conducts multi-step research on the internet. This tool can handle complex tasks in a fraction of the time it would take a human researcher.","excerpt":"This also includes laptop and mobile phone cases.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","OpenAI"],"author_name":"Supreeth Koundinya","publish_date":"2025-02-04T12:45:16","publication_year":"2025","word_count":398,"keywords":["ChatGPT","OpenAI","AI","innovation","GPT","Ray","generative AI","GAN","R","AI (Artificial Intelligence)","startup"],"extracted_tech_keywords":["AI","generative AI","ChatGPT","OpenAI","Ray","R","GPT","GAN","innovation","startup"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/openai-files-patent-to-develop-smart-watches-smart-jewellery-wearable-cameras\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":25692,"title":"Will Priyanka Chopra play the leading role in the upcoming Kalpana Chawla biopic?","content":"Biopics and women-centric films are not the only things “hot” around Bollywood today, as the next Priyanka Chopra-starrer movie combines both the aspects with astronomy. Her latest project is all about the late Indian astronaut Kalpana Chawla. The actress was offered this titular role sometime back, but Chopra came on board for this project only now. The movie about Kalpana Chawla will essentially focus on the journey of first Indian woman to go into space, in the year 1997. Chawla was born at Karnal in Haryana. Unfortunately, on her second space expedition, in 2003, the shuttle Columbia exploded, claiming the lives of seven crew members, including Chawla. The famous Indian astronaut spent 31 days, 14 hours, and 54 minutes in space. Reports state that Chopra’s entire team has been actively involved with the pre-production work of the film. The movie will be directed by debutante Priya Mishra. Reports also suggest that the script for the film has been locked, and a new production house, Getaway will be bankrolling the film. Viacom 18 had already announced its plan to make a biopic on an Indian astronaut, however, they had not revealed the name, or suggested anything about Chawla. Furthermore, Chopra had previously reflected keen interest in in producing that film under her new banner Purple Pebble Pictures. However, the project was eventually handed over to Getaway. This movie will mark Chopra’s first Hindi movie, this year. The film is also her second biopic, post Omung Kumar’s 2014 sports drama Mary Kom. Besides Chopra, Aamir Khan and Sushant Singh Rajput also have their individual space projects. Khan will play Rakesh Sharma, the first Indian man to go into space, in the upcoming movie “Saare Jahan Se Achha”; while Rajput will star alongside Nawazuddin Siddiqui and R Madhavan in the film “Chanda Mama Door Ke.” Priyanka is back from the States, meanwhile, after successfully completing the shoot for second season of her television show “Quantico”. During her stay in the city, she will also be promoting her Hollywood debut film ‘Baywatch‘.","excerpt":"Biopics and women-centric films are not the only things “hot” around Bollywood today, as the next Priyanka Chopra-starrer movie combines both the aspects with astronomy. Her latest project is all about the late Indian astronaut Kalpana Chawla. The actress was offered this titular role sometime back, but Chopra came on board for this project only […]","categories":["AI News"],"tags":["Space"],"author_name":"Amit Paul Chowdhury","publish_date":"2017-04-28T10:11:46","publication_year":"2017","word_count":338,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","Aim","Space","R"],"extracted_tech_keywords":["AI","Aim","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/will-priyanka-chopra-play-leading-role-upcoming-kalpana-chawla-biopic\/","complexity_score":2,"technical_depth":6,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10143434,"title":"Cohere’s Research Lab introduces Multilingual AI to Bridge Language Gaps","content":"Cohere for AI, a research initiative of Cohere, recently introduced Maya, an open-source multilingual multimodal model built to address gaps in vision-language models’ (VLMs) capabilities, particularly in low-resource languages. The model improves accessibility and cultural comprehension through improved data quality and toxicity filtering. The model and its datasets are available on GitHub for further development. “Current datasets often contain toxic and culturally insensitive content, perpetuating biases and stereotypes. To our knowledge, no peer-reviewed research has systematically addressed this,” the researchers stated in the paper on building a multilingual and culturally aware data set. In the context of the Maya model, “toxicity-free” means removing harmful or offensive content from the training data. The team created a pretraining dataset of 558,000 image-text pairs, expanding to eight languages, including Arabic, Hindi, and Spanish. This dataset emphasises cultural diversity while mitigating toxicity using tools like Toxic-BERT and LLaVAGuard. Maya’s performance is notable in multilingual benchmarks. It outperforms existing models in certain tasks and languages, such as Arabic, while offering comparable performance to larger models like PALO-13B. The study also highlights Maya’s effectiveness in tasks like image captioning and visual question answering. Future plans for Maya include expanding its dataset to include more languages like Bengali and Urdu and improving its instruction-tuning capabilities. Researchers also aim to refine the model’s adaptability for complex reasoning tasks. Maya’s open-source approach and focus on inclusivity mark a step forward in AI, addressing a critical need for models that understand diverse languages and cultural contexts. In August, Cohere launched Aya, a multilingual generative model that supported 101 languages, including Indian languages like Hindi and Marathi, with over 50% in lower-resourced categories. Aya outperformed mT0 and BLOOMZ across benchmarks while doubling language coverage. Developed collaboratively by 3,000 researchers in 119 countries, it was open-sourced to address AI dataset scarcity in vernacular languages.","excerpt":"Maya’s open-source model and inclusive focus advance AI by addressing the need for understanding diverse languages and cultures.","categories":["AI News"],"tags":["Cohere"],"author_name":"Aditi Suresh","publish_date":"2024-12-12T17:51:19","publication_year":"2024","word_count":302,"keywords":["Go","AI","Git","RAG","BERT","Aim","ViT","data quality","GitHub","R","Cohere"],"extracted_tech_keywords":["AI","Aim","RAG","R","Go","Git","GitHub","data quality","BERT","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/coheres-research-lab-introduces-maya-to-bridge-language-gaps-with-multilingual-ai\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":19434,"title":"In Two Separate Developments, Intel Backs Innovation In AI And Autonomous Technology","content":"As technologies like machine learning and AI are powering a new generation of smart industries, including smart homes, smart retail, smart industries, smart cars amongst others, it has made lives easier through intelligent interaction with these devices. Intel, has been a pioneer in developing solutions around emerging technologies, which can be see as a part of these two developments, it recently made. AWS announcing “DeepLens”, a deep learning enabled wireless video powered by Intel At the recently held AWS re:Invent, Amazon Web Services announced DeepLens, which is a fully programmable, deep learning enabled wireless video camera designed for developers. Revealed during AWS CEO Andy Jassy’s keynote at its annual conference in Las Vegas, DeepLens will provide optimal tools to builders to design and create artificial intelligence and machine learning products. Produced in collaboration with Intel, it is designed to suit programmers of all skill levels. With this development, Intel is reinforcing its commitment to provide developers with tools to create AI and machine learning. It had recently introduced Intel Speech Enabling Developer Kit, which provides a complete audio front-end solution for far-field voice control and makes it easier for third-party developers to accelerate the design of consumer products integrating Alexa Voice Service. DeepLens combines high amounts of processing power with an easy-to-learn user interface to support the training and deployment of models in the cloud. It is powered by Intel® Atom® X5 processor with embedded graphics that support object detection and recognition.  Developers can start designing and creating AI and machine learning products in a matter of minutes using the preconfigured frameworks already on the device. It uses Intel-optimized deep learning software tools and libraries to run real-time computer vision models directly on the device for reduced cost and real-time responsiveness. “We are seeing a new wave of innovation throughout the smart home, triggered by advancements in artificial intelligence and machine learning,” said Miles Kingston, general manager of the Smart Home Group at Intel. “DeepLens brings together the full range of Intel’s hardware and software expertise to give developers a powerful tool to create new experiences, providing limitless potential for smart home integrations.” Intel in partnership with Warner Bros announce development around autonomous cars of the future As driverless cars are inching towards reality, it has become a new consumer space to tap for opportunities. Brian Krzanich, chief executive officer of Intel Corporation, wrote on the company blogpost that autonomous driving is today’s biggest game changer, offering a new platform for innovation from in-cabin design and entertainment to life-saving safety systems. Aiming towards the same, Intel along with Warner Bros are going to develop in-cabin, immersive experiences in autonomous vehicle (AV) settings. It was announced at the recently held  Los Angeles Auto Show. “Called the AV Entertainment Experience, we are creating a first-of-its-kind proof-of-concept car to demonstrate what entertainment in the vehicle could look like in the future. As a member of the Intel 100-car test fleet, the vehicle will showcase the potential for entertainment in an autonomous driving world”, he said. The company believes that AV industry is going to create one of the greatest expansions of consumer time available for entertainment, including video-viewing time. However, the technology brought by Intel is not simply about enjoying ride but saving lives too.  Intel is also collaborating with the industry and policymakers on how safety performance is measured and interpreted for autonomous cars. It believes that setting rules for fault in advance will bolster bolster public confidence and clarify liability risks for consumers and the automotive and insurance industries. Already, Intel and Mobileye, facilitating advanced driver assistance system, have proposed a formal mathematical model called Responsibility-Sensitive Safety (RSS) to ensure, from a planning and decision-making perspective, the autonomous vehicle system will not issue a command leading to an accident. Safety systems of the future will rely on technologies with maximum efficiencies to handle the enormous amount of data processing required for artificial intelligence. “Earlier this year, we closed our deal with Mobileye, the world’s leader in ADAS and creator of algorithms that can reach better-than-human-eye perception through a camera. Now, with the combination of the Mobileye “eyes” and the Intel microprocessor “brain,” we can deliver more than twice the deep learning performance efficiency than the competition.1”, he said. “From entertainment to safety systems, we view the autonomous vehicle as one the most exciting platforms today and just the beginning of a renaissance for the automotive industry”.","excerpt":"As technologies like machine learning and AI are powering a new generation of smart industries, including smart homes, smart retail, smart industries, smart cars amongst others, it has made lives easier through intelligent interaction with these devices. Intel, has been a pioneer in developing solutions around emerging technologies, which can be see as a part […]","categories":["AI News"],"tags":["Intel"],"author_name":"Srishti Deoras","publish_date":"2017-11-30T11:19:41","publication_year":"2017","word_count":735,"keywords":["Go","machine learning","artificial intelligence","AWS","AI","computer vision","Aim","deep learning","object detection","R","Intel"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","deep learning","computer vision","Aim","object detection","AWS","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/two-separate-developments-intel-backs-innovation-ai-autonomous-technology\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10077170,"title":"Companies in India Have Finally Woken Up To Web3","content":"Last year, a sudden trend appeared on Twitter, where people were replacing their profile pics with funny-looking apes’ cartoons. Apes sporting cool sunglasses, apes smoking cigarettes or apes with stoned red eyes. A company called ‘Bored Ape Yacht Club’ was behind these NFTs (non-fungible tokens) and made about $2 million from the initial batch of selling the apes NFTs. Since then, the collection has gone beyond $100 million dollars in trading, with the cheapest apes often going for almost fourteen thousand dollars. It seemed so silly to many people when Web3 enthusiasts were buying those digital monkeys and spending millions on NFTs in general. But, today, the popular concept has penetrated the Indian Web2 platforms as well, bringing other uses with it. Flipkart recently joined hands with ‘Nothing’ phone to bring their customers into the Web3 playground with the very first NFT drop at Flipkart FireDrops, hosted on Polygon’s blockchain. This can completely revamp the e-commerce experience by unlocking verifiable digital ownership and driving greater community engagements. “With this utility-based adoption, people are finally understanding that there are some real-world use cases of Web 3 solutions like NFT and Metaverse,” said Manasa Rajan, CEO, Jupiter Meta. Jupiter Meta is a homegrown integrated Web 3.0 entity that is changing the metaverse experience and providing NFT solutions to companies. She further added that these are the first signs of adoption, which means that Web3 is here to stay. It now shows its true capability that will improve our experience as users and the way we interact with products, services, and content. Polygon: An enabler to Web3 space In the past few months, ‘Polygon Technology’, a decentralised Ethereum scaling platform, has become popular in the Web2 ecosystem. Reddit, Starbucks, Robinhood and Flipkart have the polygon technology to step into the world of Web3. For example, social media network Reddit is now airdropping its Polygon-based “Collectible Avatars”. Last month, Starbucks announced ‘Starbucks Odyssey’, an NFT-based rewards for members that will allow them to earn and purchase digital collectible assets. Starbucks partnered with Polygon for their NFT loyalty programme. Of all the platforms, the biggest news was revealed of Flipkart adding PolygonScan blockchain explorer. https:\/\/twitter.com\/sandeepnailwal\/status\/1572249196283019267 “The fact that companies like Reddit, Starbucks and even Meta have implemented our Web3 and NFT features is a great example of increase in the adoption curve and how these technologies are ready to rule the tech industry,” said Arjun Kalsy, VP of Growth, Polygon. Starbucks has been implementing Web3 technology that will pave the way to engage with its Starbucks Rewards loyalty programme members in exciting new ways—unlocking access to immersive coffee experiences from unique merchandise and artist collaborations and invitations to exclusive events. Web3 generalised When Web3 enthusiasts were excited about NFT, metaverse and other Web3 features, sceptics, especially Web2 platforms, outrightly rejected these ideas and called it a ‘fad’. Two years down the line, these features are now part of the Web2 ecosystem. Rajan of Jupiter Meta claims that she has developed the world’s first Web3 content platform to future-proof the publishing industry and create new monetisation opportunities for publishers and journalists alike. Besides, her company is also working on projects with Indian supermarkets, dairy companies and music festival organisers to make the customer experience seamless while using Web3 technologies, such as metaverse and NFT. From rejection to adoption “The main reason why Web2 companies are gradually leaning towards Web3 features is the seamless experience and ownership opportunities it provides to their customers,” said Kalsy of Polygon. The company believes that joining hands with some of the world’s biggest technology and B2C companies has enabled them to showcase the power of Web3 and gain trust among its customers. More and more Web2 companies are now reaching out to Polygon to collaboratively strategise for the implementation of Web3 features on their respective platforms. Experts assert that while the recent adoption of NFT and blockchain explorer by these companies will give a boost to the awareness about Web3, widespread popularity may still be a while away. “The onus of making Web3 popular with consumers would rest with the companies,” said Piyush Gupta, co-founder, VOSMOS, a Web3-based digital marketing platform. He believes that the Starbucks loyalty programme is among the best business ideas. However, in order to make consumers adopt a new programme based on Web3 and NFT, Starbucks would need to make things simple. Easy access and simplicity of usage would be the biggest hurdles for the brand. Having said that, if popular brands are diving into the Web3 pool, there will definitely be more awareness, acceptance and adoption of Web3. Piyush Gupta further added that the company is constantly working towards transforming more brands as they believe Web3 communities are tight-knit groups of people who can bring about the change by providing peer-to-peer support with no central authority.","excerpt":"Web2 companies are gradually leaning towards Web3 features to provide seamless experience and ownership opportunities to customers.","categories":["IT Services"],"tags":["Flipkart","NFT","nothing","polygon","Twitter (X)"],"author_name":"Tausif Alam","publish_date":"2022-10-13T16:00:00","publication_year":"2022","word_count":800,"keywords":["Go","programming_languages:R","AI","nothing","ML","Flipkart","R","Git","polygon","Aim","ViT","Rust","GAN","Twitter (X)","NFT"],"extracted_tech_keywords":["AI","ML","Aim","R","Go","Rust","Git","GAN","ViT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/it-services\/companies-in-india-have-finally-woken-up-to-web3\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10040711,"title":"Grabbing Opportunities By The Neck: Way Forward To Changing The Narrative For Women In Tech","content":"Statistics show how women continue to be underrepresented in the tech field. For instance, the Big Five tech giants (Amazon, Apple, Facebook, Google and Microsoft) have only 34.4 percent women in their workforce. On the first day of Analytics India Magazine’s Rising 2021, a panel comprising Indra Panwar, Senior Staff Engineer at Stryker; Harpreet Singh, Director of R&D at Stryker; Poornima Dore, Head of Data Driven Governance at Tata Trusts; and Swetha Mandava, Machine Learning Engineer at You.com, discussed how women can get into roles typically labelled as ‘not suitable for women’, and how we can change the narrative for women in tech. The session was moderated by Anshu Sharma, Managing Director, Global Head of Retail Banking Technology at Standard Chartered Bank. Key highlights from the panel discussion: Communication is the key Indra started her journey in the tech industry 13-years ago. At the beginning, her role was majorly around testing exercises. Although she enjoyed her role as Senior Staff Engineer at Stryker, she wanted to move to a more development-focused position. “Since then, that has always been my strategy. Whatever you are looking for, and whatever you feel you are ready for, you should put yourself out there. Communicate and have the right dialogue with the right person and at the right time,” she said. Swetha started her journey as a professional four years back. “I was the only woman in a 200-people organisation. It was awkward, and suddenly you have imposter syndrome, wondering if you are doing well. Just talking it out, and putting efforts in making friends, saying it out even if you feel stupid, kind of helped me in the beginning,” Swetha said. Game of risks “Data science and the insights we get out of it is a mix of art and science. It is a combination of technology and applying the business domain. Only then magic happens,” Anshu said. In the tech field, as in any other career choice, there is no blueprint for success. Poornima, with a PhD in Economics, presently heads Data-Driven Governance in Tata Trusts. She has handled other functions and verticals, especially in the finance space, before taking up her current role. Along her journey, Poornima realised that, to be successful, skilling up is extremely important. At present, her team has people from both tech and non-tech backgrounds. “I see data science becoming a place where people from different backgrounds can also cross-pollinate and can deliver meaningful outlets,” Poornima said. With more than two decades of experience across three organisations, Harpreet believes the key career highlights of his career are either the mistakes he made, risks he took, opportunities he grabbed and executed to the best of his abilities. “Staying curious and having a good appetite for risks. You may not always succeed but you will always come out more learned and wise,” Anshu added. Executing policies Every organisation has policies to promote equal opportunity, inclusion and flexibility. However, that is just the basic layer. Mature organisations start looking at how they can have more women in the technical programme side. Harpreet said, “Having policies in place is the bare minimum organisations can do. Showcasing these changes on ground is the need of the day. Having a policy is not enough; it is important to execute it.” Harpreet said, once policies have been executed, the next most important thing is organisation and team culture. “Organisation culture should evolve to shifting focus to outcomes rather than timings, understanding differences and adjusting behaviour accordingly. Without evolving culture, on-ground changes will not be seen,” Harpreet added. Impact with influence Accomplished women should help other women reach their full potential. To achieve this, the panellists shared tips that helped them in their journeys: Learn to say ‘No’ and build a body of work Seek feedback; do not interpret feedback with criticism Come out of your comfort zone and grab opportunities; look beyond the box Ask for what you want Channelise your creative energy Be authentic, do not get lost in code; speak the language of other stakeholders and focus on good use-cases","excerpt":"Statistics show how women continue to be underrepresented in the tech field. For instance, the Big Five tech giants (Amazon, Apple, Facebook, Google and Microsoft) have only 34.4 percent women in their workforce.  On the first day of Analytics India Magazine’s Rising 2021, a panel comprising Indra Panwar, Senior Staff Engineer at Stryker; Harpreet Singh, […]","categories":["AI Features"],"tags":["Women in Tech"],"author_name":"Debolina Biswas","publish_date":"2021-05-25T14:00:00","publication_year":"2021","word_count":677,"keywords":["data science","Go","machine learning","programming_languages:R","AI","data-driven","Women in Tech","analytics","Rust","GAN","R"],"extracted_tech_keywords":["AI","machine learning","data science","analytics","R","Go","Rust","GAN","data-driven","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/grabbing-opportunities-by-the-neck-way-forward-to-changing-narrative-for-women-in-tech\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10069632,"title":"Is GPT-4chan the worst AI ever?","content":"YouTuber and DeepJudge CTO Yannic Kilchner created an AI chatbot called ‘GPT-4chan’. The bot was trained on three years’ worth of posts from 4chan, the repulsive cousin of Reddit. Kilchner fed the bot threads from the Politically Incorrect \/pol\/ board, a 4chan message board notorious for racist, xenophobic, and hateful content. The bot sparked a heated debate on social media before it went offline. https:\/\/twitter.com\/ykilcher\/status\/1532751551869108227 Recently, the AI community launched a petition ‘Condemning the deployment of GPT-4chan.’ The petition stated: “Unfortunately, we, the AI community, currently lack community norms around their responsible development and deployment. Nonetheless, it is essential for members of the AI community to condemn clearly irresponsible practices.” There are legitimate and scientifically valuable reasons to train a language model on toxic text, but the deployment of GPT-4chan lacks them. AI researchers: please look at this statement and see what you think: https:\/\/t.co\/JbxXl6Fld5— Percy Liang (@percyliang) June 21, 2022 GPT-4chan is a large language model trained on approximately 134.5 million posts from the Politically Incorrect \/pol\/ anonymous message board. Kilchner developed the model by fine-tuning GPT-J with a previously published dataset to mimic the users of 4chan’s board. He posted about GPT-4chan on his YouTube channel and called it the ‘Worst AI ever.’ “The model was good, in a terrible sense … It perfectly encapsulated the mix of offensiveness, nihilism, trolling, and deep distrust of any information whatsoever that permeates most posts on \/pol\/.” He claimed his model is more truthful than any other GPT model out there. Kilchner said the bot has posted around 30,000 times on 4chan before being taken down and posted more than 1,500 times in a span of 24 hours. The model was downloaded over 1,400 times, and the links were available on Twitter, Hacker News, and Reddit. Kilchner even created a website (not accessible anymore). He also published the codes on Github. He said the idea to create GPT-4chan came to him after Elon Musk claimed that the number of bots on Twitter is much higher than the official number (5 percent). Condemning GPT-4chan The model was hosted on Hugging Face as well. https:\/\/twitter.com\/DrLaurenOR\/status\/1533910445400399872 Initially, Hugging Face limited access to the model before removing access to the model altogether. “Hugging Face as the model custodian (an interesting new concept) should implement an ethics review process to determine the harm hosted models may cause, and gate harmful models behind approval\/usage agreements. Open science and software are wonderful principles but must be balanced against potential harm. Medical research has a strong ethics culture because we have an awful history of causing harm to people, usually from disempowered groups,” said AI safety researcher Dr Lauren Oakden-Rayner. So far, the petition has been signed by more than 200 members of the community, including Yoshua Bengio, Full professor at Université de Montréal, Sam Bowman, Assistant Professor, NYU and Jonathan Berant, Associate Professor, Tel Aviv University. “Yannic Kilcher’s deployment of GPT-4chan is a clear example of irresponsible practice. GPT-4chan is a language model that Kilcher trained on over three million 4chan threads from the Politically Incorrect \/pol\/ board, a community full of racist, sexist, xenophobic, and hateful speech that has been linked to white-supremacist violence such as the Buffalo shooting last month,” the petition said. However, not everyone is on board with the petition. Dustin Tran, Senior Research Scientist at Google Brain, said, “I’m against GPT-4chan’s unrestricted deployment. However, a condemnation letter against a single independent researcher smells of unnecessary pitchfork behaviour. Surely there are more civil and actionable approaches.” Two sides to a story Much of the debate on social media with GPT-4chan has been about how harmful models like GPT-4chan can wreak havoc. The biggest concern was that GPT-4chan could pave the way for more AI bots being developed to spread racist and hateful messages online without any human intervention. Secondly, the model could target vulnerable people with harmful messages that could lead to self-harm. Also, models such as GPT-4chan could be weaponised to spread misinformation. However, Kilchner has defended his model on social media claiming there were no documented incidents of GPT-4chan causing harm to anybody. https:\/\/twitter.com\/ykilcher\/status\/1533917117002694657 That said, there are two sides to a story. A section of social media came to Kilchner’s defence, arguing the model is not inherently harmful and could be used for good. For example, the model could be leveraged to combat hate speech. tbh the entire GPT-4chan incident and the way AI researchers are getting harassed over their disagreements with it discourages me from producing fun AI content on YouTube because I do not want to cultivate that type of fanbase ever— Max Woolf (@minimaxir) June 21, 2022","excerpt":"A condemnation letter against a single independent researcher smells of unnecessary pitchfork behaviour.","categories":["AI Features"],"tags":[],"author_name":"Pritam Bordoloi","publish_date":"2022-06-23T14:00:00","publication_year":"2022","word_count":769,"keywords":["Go","Hugging Face","AI","Git","RAG","Ray","Aim","Rust","GitHub","R"],"extracted_tech_keywords":["AI","Aim","Ray","Hugging Face","RAG","R","Go","Rust","Git","GitHub"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/is-gpt-4chan-the-worst-ai-ever\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10068616,"title":"Image annotation techniques with implementation in OpenCV","content":"Image annotation is important in computer vision, which is the technique that allows computers to obtain high-level comprehension from digital images or videos and to observe and interpret visual information in the same way that humans do. Annotation, often known as picture labelling or tagging, is a crucial stage in the development of most computer vision models. This article will be focusing on creating these annotations with the help of OpenCV. Following are the topics to be covered. Table of contents The Image annotationNeed for image annotationTypes of image annotationImplementing Image annotation with OpenCV The better the machine learning models perform, the greater the quality of your annotations. Let’s understand image annotations. The Image annotation The process of labelling, tagging or specifying images in a particular dataset to train machine learning models is known as an image annotation. When the manual annotation is finished, the tagged pictures are processed by a machine learning or deep learning model to repeat the annotations without the need for human intervention. As a result, picture annotation is utilised to indicate the aspects that your system needs to recognise. Supervised Learning is the process of training an ML model given labelled data. Image annotation establishes the criteria that the model attempts to duplicate, thus any errors in the labelling are also repeated. As a result, correct picture annotation creates the groundwork for training neural networks, making annotation one of the most critical jobs in computer vision. Image annotations can be done manually or with the help of an automatic annotation tool. Auto annotation technologies are often pre-trained algorithms that can accurately label photos. Their annotations are required for complex annotation jobs such as constructing segment masks, which take time to generate. Are you looking for a complete repository of Python libraries used in data science, check out here. Need for image annotation Labelling pictures is required for functional datasets because it informs the training model about the relevant aspects of the image (classes), which it can then use to identify those classes in fresh, never-before-seen images. Image annotation generates training data from which supervised AI models may learn. The manner in which annotating images predicts how the AI will behave after viewing and learning from them. As a result, poor annotation is frequently described in training, resulting in models that make bad predictions. Annotated data is very important when tackling a unique challenge and using AI in a new domain. For typical tasks like image classification and segmentation, pre-trained models are frequently available, and they may be customised to specific use cases using Transfer Learning with minimum input. Training a comprehensive model from scratch, on the other hand, frequently necessitates a massive quantity of annotated data divided into train, validation, and test sets, which is difficult and time-consuming to generate. Unsupervised algorithms, do not need annotated data and may be trained directly on raw data. Types of image annotation There are three prevalent methods of image annotation, and the one you choose for your use case will be determined by the project’s complexity. The more high-quality picture data used for each kind, the more accurately the AI will forecast. Classification Classification is the simplest and quickest approach for image annotation since it simply assigns one tag to a picture. For example, you could wish to go through and categorise a collection of photographs of grocery store shelves to determine which ones contain soda and which do not. This approach is ideal for capturing abstract information, such as the example above, or the time of day, if there are automobiles in the image, or for filtering out photographs that do not satisfy the criteria from the start. While categorization is the quickest at providing a single, high-level label, it is also the most ambiguous of the three categories we emphasise since it does not identify where the item is inside the image. Object Detection Annotators are given particular things to label in a picture using object detection. So, if a picture is labelled as having ice cream in it, this goes a step further by indicating where the ice cream is inside the image, or if particularly searching for where the cocoa ice cream is. Object detection may be accomplished using a variety of approaches, including: Bounding Boxes: Annotators use rectangles and squares to define the position of target objects in 2D. This is one of the most often used picture annotation approaches. Cuboids, also known as 3D Bounding Boxes, are used by annotators to specify the location and depth of a target object.Polygonal Segmentation: Annotators employ complicated polygons to specify the position of target items that are asymmetrical and do not simply fit inside a box.Lines: Annotators detect essential boundary lines and curves in a picture to distinguish sections using lines and splines. Annotators may, for example, name the numerous lanes of a highway for a self-driving car picture annotation project. This approach is still not the most exact since object detection allows for overlap in the use of boxes or lines. What it does offer is the general position of the item while remaining a pretty quick annotating procedure. Semantic Segmentation Semantic segmentation overcomes the overlap problem in object recognition by assuring that each component of an image belongs to just one class. This approach, which is usually done at the pixel level, needs annotators to assign categories (such as a pedestrian, automobile, or sign) to each pixel. This aids in teaching an AI model how to detect and categorise certain items, even when they are obscured. For example, if a shopping cart is obscuring a portion of the image, semantic segmentation may be used to define what coco ice cream looks like down to the pixel level, allowing the model to know that it is still, in fact, coco ice cream. Implementing Image annotation with OpenCV In this article, we will be using Bounding boxes and the colour segmentation method for the image annotation. In bounding boxes, methods will be manually drawing different bounding shapes around the object and adding some text to it. In colour segmentation, we will be using the KNN algorithm to segment the colours of the objects in the query image. The colours would be segmented based on the value of ‘K’ which is the number of the nearest neighbours and that segmented portion on images can be treated as an annotated part. Bounding Boxes method Import necessary libraries import cv2 import numpy as np import matplotlib.pyplot as plt Read the query image input_img=cv2.imread('annotation_image.jpg',cv2.IMREAD_COLOR) Query image As in this article, we are using a coloured image so we need to use the ‘cv2.IMREAD_COLOR’. As it instructs to load a colour picture. Any picture transparency will be ignored. It is the default setting. We may also pass the integer value 1 for this flag. Draw a line on the object image_line=input_img.copy() cv2.line(image_line, (900,150), (1100,150), (0,255,255), thickness=5,lineType=cv2.LINE_AA) plt.figure(figsize=(10,10)) plt.imshow(image_line[:,:,::-1]) plt.show() The cv2.line takes input coordinates of the start and end point of the line with the thickness, transparency and colour of the line. Analytics India Magazine Draw a circle around the object image_circle=input_img.copy() cv2.circle(image_circle, (1030,340),200, (0,255,255), thickness=5,lineType=cv2.LINE_AA) plt.figure(figsize=(10,10)) plt.imshow(image_circle[:,:,::-1]) plt.show() The ‘cv2.circle’ takes the radius and the coordinates for the circle as an input. Rest is the same as the line function discussed earlier. Analytics India Magazine Draw a rectangle around the object image_rect=input_img.copy() cv2.rectangle(image_rect, (900,150),(1100,530), (0,0,255), thickness=5,lineType=cv2.LINE_AA) plt.figure(figsize=(10,10)) plt.imshow(image_rect[:,:,::-1]) plt.show() It takes the top left side corner coordinates and the bottom right corner coordinates for drawing the rectangle. Analytics India Magazine KNN method for segmentation Import necessary libraries import cv2 import numpy as np import matplotlib.pyplot as plt Reading and preprocessing img = cv2.cvtColor(input_img,cv2.COLOR_BGR2RGB) image_reshape = img.reshape((-1,3)) image_2d = np.float32(image_reshape) Change the order of the colours since in OpenCV the colour of an image is read as Blue, Green and Red (BGR). The requirements are Red, Green and Blue(RGB). Applying the KNN criteria = (cv2.TERM_CRITERIA_EPS + cv2.TERM_CRITERIA_MAX_ITER, 100, 1.0) K = 4 attempts=10 ret,label,center=cv2.kmeans(twoDimage,K,None,criteria,attempts,cv2.KMEANS_PP_CENTERS) center = np.uint8(center) res = center[label.flatten()] result_image = res.reshape((img.shape)) Since the image is a high-resolution image so there are a lot of data points to go through, it would take time if the number of iterations is high. We have put a limit on the number of iterations to 100 and the epsilon value is set at the highest. The k nearest neighbour is set as 4 with the number of attempts to be 10. plt.figure(figsize=(10,10)) plt.imshow(result_image[:,:,::-1]) plt.show() Analytics India Magazine The algorithm has segmented the colours quite well. The blues, whites, greys, and browns could be seen separated. One could mask the image and further tune the algorithm. Conclusions One of the most time-consuming aspects of dealing with data is data gathering and annotation. Nonetheless, it serves as the foundation for training algorithms and must be executed with the greatest precision feasible. Proper annotation frequently saves a significant amount of time later in the pipeline when the model is being created. With this article, we have understood different types of annotations and their implementations. References Link to the above codeRead more about Annotation","excerpt":"Image annotation is labelling of objects in an image","categories":["AI Trends"],"tags":["Image Classification","image processing","knn","Object Detection","object segmentation","OpenCV","Segmentation"],"author_name":"Sourabh Mehta","publish_date":"2022-06-09T11:00:00","publication_year":"2022","word_count":1508,"keywords":["data science","Segmentation","NumPy","machine learning","knn","image processing","AI","ML","neural network","computer vision","OpenCV","deep learning","Object Detection","object segmentation","analytics","Image Classification"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","neural network","computer vision","data science","analytics","OpenCV","NumPy"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/image-annotation-techniques-with-implementation-in-opencv\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10005051,"title":"Now You Can Write One Code That Works On Both PyTorch And Tensorflow","content":"“Library developers no longer need to choose between frameworks.” The researchers from Tübingen AI Center, Germany, have introduced a new Python framework, ‘EagerPy’ that allows the developers to write code that can work independently of the popular frameworks like PyTorch and TensorFlow. In a recently published work on EagerPy, the researchers wrote that library developers no longer watch out for framework dependencies. Their new Python framework, EagerPy takes care of their re-implementation and code duplication hurdles. For instance, Foolbox is a Python library that is built on top of EagerPy. The library was rewritten in EagerPy instead of NumPy to achieve native performance on models developed in PyTorch and TensorFlow with one code base and without code duplication. Foolbox is a library for running adversarial attacks against machine learning models. Importance Of Framework Agnostic Practices Addressing the differences between frameworks, the authors explored syntactic deviations. In the case of PyTorch, gradients using an in-place requires_grad_() call, and backpropagation is called using backward(). Whereas, TensorFlow offers a high-level manager and functions like tape.gradient to query gradients.  Even at the syntactic level, these two frameworks differ a lot. For example, dim vs axis in case of parameters and sum vs reduce_sum in case of functions. This is where EagerPy comes into the picture. It resolves the differences between PyTorch and TensorFlow by providing a unified API that transparently maps to various underlying frameworks without the computational overhead. “EagerPy lets you write code that automatically works natively with PyTorch, TensorFlow, JAX, and NumPy.” EagerPy focuses on eager execution and in addition, wrote the researchers, its approach is transparent, and users can combine framework-agnostic EagerPy code with framework-specific code. The introduction of eager execution modules by TensorFlow and similar features by PyTorch made eager execution mainstream and the frameworks more similar. However, despite these similarities — between PyTorch and TensorFlow 2 — writing framework-agnostic code is not straightforward. At the semantic level, the APIs for automatic differentiation in these frameworks differ. Know more about eager execution here. Automatic differentiation refers to algorithmic solving of a differential equation. It works on the principle of chain rule, i.e., solving derivatives of a function can be distilled down to fundamental mathematical operations(addition, subtraction, multiplication and division). These arithmetic operations can be represented in a graph format. EagerPy especially uses a functional approach to automatic differentiation. Here’s a code snippet from the EagerPy documentation: import eagerpy as ep x = ep.astensor(x) def loss_fn(x): # this function takes and returns an EagerPy tensor return x.square().sum() print(loss_fn(x)) # PyTorchTensor(tensor(14.)) print(ep.value_and_grad(loss_fn, x)) First function is defined and then differentiated with respect to its inputs. It is then passed to ep.value_and_grad to evaluate the function and its gradient. Also, norm function can now be used with native tensors and arrays from PyTorch, TensorFlow, JAX and NumPy with virtually no overhead compared to native code. It also works with GPU tensors. import torch norm(torch.tensor([1., 2., 3.])) import tensorflow as tf norm(tf.constant([1., 2., 3.])) In summary, EagerPy is designed to offer the following features: Provide a unified API for eager executionMaintain the native performance of the frameworks A fully chainable API and Comprehensive type checking support. These attributes claim the researchers, make EagerPy easier and safer to work with than the underlying framework-specific APIs. Despite these changes and improvements, the team behind EagerPy made sure that the EagerPy API follows the standards set by NumPy, PyTorch, and JAX. Get Started With EagerPy: Install the latest release from PyPI using pip: python3 -m pip install eagerpy import eagerpy as ep def norm(x): x = ep.astensor(x) result = x.square().sum().sqrt() return result.raw Know more about EagerPy here","excerpt":"“Library developers no longer need to choose between frameworks.” The researchers from Tübingen AI Center, Germany, have introduced a new Python framework, ‘EagerPy’ that allows the developers to write code that can work independently of the popular frameworks like PyTorch and TensorFlow.  In a recently published work on EagerPy, the researchers wrote that library developers […]","categories":["Deep Tech"],"tags":["Pytorch","Tensorflow"],"author_name":"Ram Sagar","publish_date":"2020-08-18T13:00:13","publication_year":"2020","word_count":600,"keywords":["Pytorch","NumPy","machine learning","AI","PyTorch","Python","Ray","Aim","JAX","TensorFlow","R","Tensorflow"],"extracted_tech_keywords":["AI","machine learning","Aim","Ray","TensorFlow","PyTorch","JAX","NumPy","Python","R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/eagerpy-pytorch-tensorflow-coding\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10072182,"title":"IIM Lucknow invites applications for Executive Programme in AI for Business","content":"IIM Lucknow has invited applications for an executive program in artificial intelligence. The six-month online certificate course is developed in association with WileyNXT for professionals with 2-3 years of experience interested in building AI products and services. Graduates from any UGC-recognized university with a minimum score of 50% can apply for the AI for Business programme. Upon successful completion of the programme, candidates will receive a certificate from IIM Lucknow. The course aims to provide learners with the right knowledge to utilize technology platforms and help organizations build AI-driven business strategies. The programme is set to commence on September 4 in an online mode. “A recent report by Bain & Company found that 87 percent of enterprises expect to increase annual AI spend by more than 10 percent. Another survey by PwC revealed that 75 percent of business leaders are confident that AI will help them make better decisions and 64 percent indicated it will be critical to their organization’s efficiency and productivity in the future. All these trends point to the increasing penetration and deployment of AI for business and organizations,” IIM Lucknow noted, emphasizing the need for courses like this. Interested candidates can apply for the IIM Lucknow Executive Program in AI for Business through the official website – iiml.ac.in.","excerpt":"The six-month online course is designed for professionals with 2-3 years of experience interested to work in AI or ML","categories":["AI News"],"tags":[],"author_name":"Bhuvana Kamath","publish_date":"2022-08-04T15:33:47","publication_year":"2022","word_count":212,"keywords":["artificial intelligence","programming_languages:R","AI","ML","Aim","ViT","GAN","R"],"extracted_tech_keywords":["AI","artificial intelligence","ML","Aim","R","GAN","ViT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/iim-lucknow-invites-applications-for-executive-programme-in-ai-for-business\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10082805,"title":"How Marketing Analytics Leverage Explainable AI for Forecast","content":"Attribution modelling for measuring brand effectiveness has been on the rise. However, the method, which involves assigning credit for sales and conversions to touchpoints (i.e., clicks) in conversion paths, is said to have shortcomings. For instance, it is difficult to make a causal inference that among all factors, what is to be attributed for conversion or sale. Experts believe that the short-term, direct effects say little about effectiveness, calling for a different evaluation method. As a result, big-tech companies like Meta and Google are moving to econometrics for evaluating brand effectiveness. Econometric modelling involves statistically inferring the relationships between large quantities of data (called variables) and analysing their correlation over time. Hence, it will consider parameters beyond clicks, namely, non-media drivers of sales, such as seasonality, price changes, and so on. Varun Kumar Anguralia, Head of Marketing Analytics at L&T Financial Services, told AIM that marketing analytics is a diverse field and generally requires you to be closer to data, business, and market trends. Establishing a relationship between them often gets complicated with the levels of attributes included. This gap is filled by econometric modelling, allowing organisations to strategise the best actions. Angurulia added that prior to getting into complex solutions, exploration with data—which includes utilising basic statistical concepts like central tendencies and dispersion hypothesis—gives visibility to the direction organisations need to take. Econometrics x Machine Learning There has been sufficient growth towards clubbing econometrics with machine learning in recent years. This expands the range of hypotheses that can be explored with the available data. For example, ML-powered Marketing Mix Model (MMM) can aid advertisers in figuring out cross-media Return on Investment (RoI) and budget allocation decision-making. However, achieving this is a complex multidimensional challenge that requires considering aspects like non-linearity and saturation, time-delayed impact (Adstock effect), and interaction effects. Historically, linear models have also poorly represented these aspects while also taking a long time if numerous iterations have to be run to calibrate the model. Non-linearity: MMM typically follows a logarithmic relationship. This means that to avoid spending too much or too little, one has to find the optimum middle where the media output is used efficiently for a particular budget. Adstock effect: It is not always the case that a campaign will immediately impact business outcomes. Some strategies may take weeks to have an influence over the consumers. These strategies generally take steps to build brand awareness, only following which sales come into the picture. Interaction: As mentioned earlier, consumers generally interact with various touchpoints before making a buying decision, and it is difficult to identify the effect of certain channels correctly. ML algorithms can develop complex models studying the patterns or correlations based on sample data and then generalise and apply them to new unknown data. It takes into account the non-linear correlations, the time delay effect, and the interaction effect of media. Here, in contrast to linear models, complex ML-based algorithms can include many variables without leading to over-specification. This implies that it can realistically represent the interaction effects between the different characteristics. Interpretability in the model is crucial for marketing analytics. The major features that affect a model’s outcome should be known, along with their direction and quantity of influence. Hence, linear models are still used in marketing analytics for their ease of interpretability, even if the output is less accurate. However, in recent times, two prominent approaches have emerged to interpret complex non-linear models and show which input is decisive for the forecast results: Shapley Additive Explanation (SHAP) and Local interpretable model-agnostic explanations (LIME). Anurag Pandey, a Data Science Engineer, writes in his blog, “The interesting thing to note is that when complex models are combined with the explanation models, they outperform [Generalised Linear Models] GLMs not only in accuracy (obviously) but also in interpretability\/explainability.” Approaches for Explainable AI Shapley Additive Explanation (SHAP) is a game theoretic approach which connects optimal credit allocation with local explanations using Shapley values. Shapley values guarantee uniqueness in the solution by telling how to distribute the contribution by the features accurately. So, for example, if we take the case of MMM itself, the dataset can include different marketing channels, such as TV, Radio, and Print. There could be a predicted return value for each instance of data (indicating the marketing spend). The predicted result for the particular dataset can be compared with the overall average prediction value to understand the difference. Hence, as Anurag Pandey explains, two things are working here: a machine learning algorithm for prediction and Shapley values to understand the feature behaviour for each instance of data. Thus, SHAP is a tool to make explainable AI by visualising the output. Moreover, it can explain any model’s prediction by computing each feature’s contribution to the prediction. Read: A Complete Guide to SHAP – SHAPley Additive exPlanations for Practitioners Similarly, another approach is Local interpretable model-agnostic explanations (LIME). This method is typically used for texts and images and slightly less for tabular data. A simple example of this would be that it isolates individual predictions and trains an interpretable model for this prediction by changing input values (for example, turning certain pixels on and off) and documenting how these values impact a complex model’s decision. Furthermore, machine learning is proving to be vital to the retail industry, with brands leveraging upon it to gain a competitive edge. A Juniper Research study found that marketing and sales are the biggest proponents of machine learning, with 40% favouring it. For example, Clairvoyant, an AI analytics company, was able to use the MMM model for an insurance company that wanted to understand the contribution of each channel in conversion (product purchase or becoming a lead) and eventually optimise the marketing spending on each channel. The result was that by optimising the budget for different channels while keeping the total budget the same, they saw a lift of 10.2% in sales.","excerpt":"ML-powered Marketing Mix Models (MMM) gives the advertising industry a means to optimise decision-making.","categories":["AI Features"],"tags":["Explainable AI"],"author_name":"Ayush Jain","publish_date":"2022-12-20T13:00:00","publication_year":"2022","word_count":980,"keywords":["data science","Go","machine learning","TPU","AI","ML","RAG","Aim","analytics","Explainable AI","R"],"extracted_tech_keywords":["AI","machine learning","ML","data science","analytics","Aim","RAG","TPU","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-marketing-analytics-leverage-explainable-ai-for-forecast\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10126886,"title":"AMD Partners with IIT-Bombay to Boost Semiconductor Startups in India","content":"AMD has announced a partnership with the Society for Innovation and Entrepreneurship (SINE), an incubator at IIT-Bombay, to promote semiconductor start-ups in India. Under this partnership, AMD will provide grants to IIT-B incubated startups working on the development of energy-efficient Spiking Neural Network (SNN) chips. The startups are tasked with devising ways to significantly reduce the energy consumption utilised by conventional neural networks. AMD’s Commitment to Energy Efficiency Speaking about the initiative, Jaya Jagadish, Country Head, AMD India, said: “At AMD, our goal is to reduce energy costs and increase efficiency across all our product lines. We aim to deliver a 30x improvement in energy efficiency for AMD processors and accelerators powering HPC and AI training by 2025. This means we are increasing the energy efficiency of compute nodes at a rate that is significantly faster than the aggregate industry-wide improvement made during the last five years. We continuously work with industry bodies and academia to support initiatives that help drive innovation in this field.” SINE’s Vision for Semiconductor Start-ups Shaji Varghese, CEO of SINE- IIT Bombay, emphasised the importance of nurturing semiconductor startups: “As India looks to execute a comprehensive semiconductor strategy, it is important to nurture, build, and grow startups in this space. As design cycles are long and the cost of entry high in the semiconductor design area, it is SINE’s endeavour to foster and create strong, product-focused, cutting-edge semiconductor design startups by addressing gaps in the industry.” Read more at: Why This Semiconductor Chip Company Chose India Over China First Grant Recipient: Numelo Technologies The first grant was awarded to Numelo Technologies for developing SNN chips using ultralow power quantum tunneling on silicon on insulator (SOI) technology. Prof. Udayan Ganguly, Director of Numelo Tech, explained: “Our focus to offer extended battery life, enhanced data privacy, and feature a versatile architecture capable of facilitating sound, image, and video recognition, has led us to develop a host of innovations at the technology, circuit and systems architecture, and algorithms\/software level.” He added: “As a startup in this area, we greatly value interacting with the senior and distinguished engineers at AMD who can mentor the program. We believe this adds far greater value than just the grant itself.” This partnership is part of AMD’s Corporate Social Responsibility initiative. AMD prioritizes the inclusion of quality education in science, technology, engineering, and mathematics (STEM) and promoting scientific research as part of its CSR program.","excerpt":"AMD has announced a partnership with the Society for Innovation and Entrepreneurship (SINE) at IIT-Bombay, to promote semiconductor startups in India.","categories":["AI News"],"tags":["AI at IIT","AMD","IIT"],"author_name":"MIA","publish_date":"2024-07-15T15:36:54","publication_year":"2024","word_count":402,"keywords":["Go","AMD","AI at IIT","AI","neural network","innovation","programming_languages:R","programming_languages:Go","Aim","GAN","IIT","R","startup"],"extracted_tech_keywords":["AI","neural network","Aim","R","Go","GAN","innovation","startup","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/amd-partners-with-iit-bombay-to-boost-semiconductor-startups-in-india\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":6158,"title":"Big Data: A ‘Big’ leap towards profitability","content":"Information mining in Big Data is the upcoming field that can form a unique selling proposition for a company. Data is the most valuable asset of business that can increase the credibility of the business much beyond the competitors reach. But few organizations have named big data as ‘Dark Data’, a journey inside a black hole which attracts one from outside but at the end takes the business nowhere. According to a study by an open source research firm, the companies expected a return of $3-$4 per dollar invested, but big data has helped them to get only 55 cents per dollar as their return on investment (ROI). Big data as the name suggests is a collection of huge ‘Volume’ of data with different ‘Varieties’. It is the task of the experts to create a high dollar ‘Value’ of big data. Some organizations have underutilised big data while the others have taken irrelevant paths to access it. The reason behind such shortcomings is the lack of knowledge about the data locator. The process of locating the business relevant data should be the initial step while implementing Big Data, where it is important to select the desired touch points and discount the rest. These touch points can also be further drilled down to obtain sub points. A concrete analysis of relevant data, out of the ocean of data, can help the organizations improve their returns due to the reduced number of touch points. Moreover, the requirement of reduced infrastructure will further reduce the initial investment cost. The decrease in the touch points will also reduce the complexity of data and will keep the analysts motivated towards the project findings. To understand further, an FMCG company is considered which wanted to launch a new product. The company should trim down the list of all the available touch points and analyse the relevant five to six touch points that are required to carry forward the market research for new product development (NPD). Moreover, these touch points can be ranked according to the project analysis. This process will reduce the complexity associated with the increasing number of touch points. Where to look for and where not to? The process can be divided into 3 phases, viz., ‘Planning Phase’, ‘After the Promotion stage’ and ‘After Launch stage’. In the ‘Planning Phase’ of the product, websites of direct competitors, websites of Indirect competitors and their Facebook and Twitter pages, can be some of the high priority and relevant touch points. In the ‘After the Promotion’ stage, data can be gathered from the media resources, one of the unstructured forms of big data, to analyse the sentiment of the target customer towards the product. Finally, in the ‘After Launch’ stage, several websites like ida.org, fdi.org as well as the product’s social media pages can be referred for their critic comments and product review from different users\/ non- users respectively. However, certain touch points like blogs and articles by writers can be avoided due to reduced chances of valuable data. An equivalent insight can be construed by the use of 10-15 touch points instead of approaching the mountains of data. Hence, the dilemma faced by the organizations today, regarding huge investment and lack of ROI in Big Data, can be solved by finding relevant data locations, which can increase the efficiency of data extraction and decrease the cost and time requirement considerably. [divider] Authors Swati Gupta is currently pursuing PGDM-Finance in Institute of Management Technology, Ghaziabad. She has worked in Cognizant Technology Solutions for a UK based Retail company after pursuing B. Tech from University of West Bengal. She has also interned in Bank of Baroda, Dubai in the field of Risk Management. She has been working on several projects related to Portfolio Management and Risk Analysis. Her primary area of interest lies in Financial Risk Analytics. Gagandeep Singh is currently pursuing PGDM-Marketing from Institute of Management Technology, Ghaziabad. Prior to MBA he has 2 years of work experience in Accenture where he dealt with Microsoft Dynamics AX for Michelin client. He holds a B.Tech degree from Maharaja Agrasen Institute of Technology, GGSIPU, New Delhi. His keen interests lie in analysing social media to construe business acumen. He has been working on various problems related to Text Analytics.","excerpt":"Information mining in Big Data is the upcoming field that can form a unique selling proposition for a company. Data is the most valuable asset of business that can increase the credibility of the business much beyond the competitors reach. But few organizations have named big data as ‘Dark Data’, a journey inside a black […]","categories":["IT Services"],"tags":[],"author_name":"Swati Gupta","publish_date":"2014-09-19T18:53:35","publication_year":"2014","word_count":710,"keywords":["big data","programming_languages:R","AI","analytics","GAN","R"],"extracted_tech_keywords":["AI","analytics","R","big data","GAN","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/it-services\/big-data-a-big-leap-towards-profitability\/","complexity_score":2,"technical_depth":6,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10012565,"title":"China’s Latest Space Launches, AWS Hit With Outage And More In This Week’s Top News","content":"This week, the internet was taken for a ride as Amazon’s Web Services were hit due to a glitch. AWS is one of the largest cloud service providers in the world. Its customers include Netflix and many more. There were even complaints of vacuum cleaners and showers not working due to this! While the world tries to wrap their head around the workings of the internet and 5G, China has already launched a 6G satellite this month along with a rocket this week that would collect some rocks from the surface of the moon. Learn what caused the AWS outage and the intent behind Chinese space missions in this week’s top tech news brought to you by Analytics India Magazine. CIA On Cloud Five According to reports,a multi billion dollar contract was to five key cloud players: Amazon Web Services, Microsoft, Google, Oracle and IBM by the Central Intelligence Agency. The companies will compete for specific task orders issued by the CIA on behalf of itself and the 16 other agencies that comprise the intelligence community. CIA’s spokesperson underlined the significance of cloud capabilities resulting from the diversified partnership. The Commercial Cloud Enterprise (C2E) Cloud Service Provider (CSP) contract might bring the cloud wars to a standstill, at least for now, after a year long tussle between the Department of Defense and Amazon for the $10 billion cloud contract, which was awarded to Microsoft Azure last year. When Internet Went Down; For A While https:\/\/twitter.com\/SJP1804\/status\/1331643504443928578?s=20 On Wednesday, reports of an Amazon Web Services outage started to appear. The problem appears to be an impairment of the programming interface for its Kinesis Data Streams product. As a result, customers weren’t able to read or write data tied to its streams. Amazon Kinesis enables real-time processing of streaming data. In addition to its direct use by customers, Kinesis is used by several other AWS services. Kinesis has a large number of “back-end” cell-clusters that process streams. As a result, stated AWS in their summary,  these slow front-end servers could be deemed unhealthy and removed from the fleet, which in turns, would set back the recovery process. Cloudwatch, Adobe, Flickr were among the popular services that took a hit. AWS, however, recovered before things went out of hand. China Launches World’s First 6G Satellite According to Popular Mechanics, on November 6, China successfully launched “the world’s first 6G satellite”. The satellite, known as Tianyan-5, is a remote-sensing satellite jointly developed by the University of Electronic Science and Technology of China, Chengdu Guoxing Aerospace Technology, and Beijing Weina Xingkong Technology. In addition to Earth observations, the satellite will test a high-frequency terahertz communication payload that could send data at speeds several times faster than 5G. This week, China was again in news in space related activities as they have launched a new mission to dig up the surface of the moon for the first time in more than 40 years. The mission, named Chang’e-5, is the latest in a series of missions to the lunar surface led by the China National Space Administration (CNSA).  If successful, the collected samples can help fill the voids in the history of the moon. Rocks obtained by previous US and Soviet lunar missions suggest that activity on the Moon peaked 3.5 billion years ago. And, Chang’e-5’s samples can give a fresh perspective of the historical activities on the Moon. Salesforce Looks To Gain From Slack-ing On Wednesday, WSJ reported that Salesforce is in advanced talks to buy Slack, which is currently valued at more than $17 billion. If this deal comes through, it will be Salesforce’s largest acquisition ever. One of the consequences of pandemic are companies like Slack and Zoom were given unprecedented impetus. Though Slack wasn’t able to capitalise as much as Zoom, their features however are changing the way corporate workspaces go about their routines. Slack’s ascent puts them at loggerheads with services like Microsoft Teams. Salesforce currently enjoys a market value of some $230 billion, has made a string of acquisitions recently that includes the $15 billion-plus takeover of data-analytics platform Tableau Software in 2019. EU Proposes New Rights; The Big Tech Might Not Like It (Image credits: Unsplash) The European Parliament wants to grant EU consumers a “right to repair” to reduce electronic waste and empower consumers to make sustainable choices. If passed– among other things– this law would force the smartphone, laptops and other device makers to append their products with a ‘how to repair’ manual to assist their customers. The European Parliament believes that this resolution would establish more transparency with regards to green initiatives and also contain the makers from resorting to trickery such as ‘planned obsolescence’; a newly downloaded update can slow down the phone sometimes. This resolution also pushes responsible marketing and advertising; the environmentally friendly claims in advertisements and bought ecolabel certifications might come under scrutiny as the consumers become more aware. Amazon To Help India Make Better Drones The Drone Federation of India (DFI) has collaborated with Amazon Web Services (AWS) and announced the Adopt Drones program to promote the adoption of AI-driven drone solutions. The Adopt Drones Program is designed to encourage novel machine learning solutions for drone applications. AWS will extend its technical capabilities to DFI for reviewing the solutions and optimising them. Google Plans Fiber-Optic Network According to the Journal, Google is laying the groundwork for a fiber-optic network that will connect through Saudi Arabia and Israel while opening a new corridor for global internet traffic. The project linking India to Europe is Google’s latest globe-crossing internet construction effort. Google is vying with Facebook to build more network capacity to support its surging user demand for videos, search results and other products. Google’s ideas to bring more connectivity between Europe and India as the Alphabet subsidiary plans to establish more data centers; the market which is currently dominated by Amazon and Micorosft.","excerpt":"This week, the internet was taken for a ride as Amazon’s Web Services were hit due to a glitch. AWS is one of the largest cloud service providers in the world. Its customers include Netflix and many more. There were even complaints of vacuum cleaners and showers not working due to this! While the world […]","categories":["AI News"],"tags":["AWS"],"author_name":"Ram Sagar","publish_date":"2020-11-28T18:00:50","publication_year":"2020","word_count":979,"keywords":["Go","machine learning","AWS","AI","R","RAG","Colab","Aim","analytics","Azure"],"extracted_tech_keywords":["AI","machine learning","analytics","Aim","Colab","RAG","AWS","Azure","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/china-6g-aws-shut-down-cloud-top-tech-news\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":33688,"title":"#AIforAll Is The Mantra For Democratising Artificial Intelligence In India","content":"From aam aadmi, to big tech firms to union ministers — everybody in India is now incorporating new tech into their lives as well as businesses. Artificial intelligence is one such technology that has gained traction in India in recent times — it has evolved in ways that far exceed its original conception. From startups to the Indian government, AI has been used to a great extent, making lives of Indian better. Now, even the Central Board of Secondary Education (CBSE) has approved the implementation of AI as a subject for students of classes 8,9, and 10. Indian Government’s policy think-tank, NITI Aayog is one such body that plays a significant role in driving India’s technological advancement to next level. In May 2018, as a supportive step towards the Make In India for developments in robotics and artificial intelligence, NITI Aayog partners with ABB India, a technology leader in power grids, electrification products, industrial automation and robotics and motion. The vision of this partnership was to prepare the sectors of the Indian economy. Furthermore, in June 2018, NITI Aayog released a discussion paper, addressing the national strategy on artificial intelligence (#AIforAll). With a to enhance and empower human capabilities and address the challenges of access, affordability, shortage and inconsistency of skilled expertise, #AIforAll focuses on harnessing collaborations and partnerships and aspires to ensure prosperity for all. Since inception, NITI Aayog has been an active part of India’s technological advancement — working day in and out, with leaving any stone unturned to position India among leaders on the global AI map. Microsoft Marking Its Territory Today, witnessing all the advancements that are taking place in India, even the tech giants have stepped into the arena and Microsoft is one of them. On 16 January 2019, Microsoft India has showcased its progress toward democratising AI in India. The company is working with stakeholders across public and private sectors, civil society, and academia in India to develop the appropriate ecosystem for human-centred AI. The company is also working with new-age organisations that are advancing and disrupting industries India. “The next wave of innovation for India is being driven by the tech intensity of companies – how you combine rapid adoption of cutting edge tech with your company’s own distinctive tech and business capabilities,” said Anant Maheshwari, President, Microsoft India. Talking about Microsoft’s AI strategy, the company focuses on enabling digital transformation, building coalitions for responsible innovation, bridging the skills gap and enhancing employability, and creating sustained social impact. The four pillars of Microsoft’s AI strategy include: Enabling #DigitalTransformation, building coalitions for responsible innovation, bridging the skills gap and enhancing employability, and creating sustained social impact. — Microsoft India and South Asia (@MicrosoftIndia) January 16, 2019 According to the Redmond-based tech giant, it has over 700 customers that have fueled the digital transformation using Microsoft’s AI solution and 60% of the customers are large businesses in manufacturing, financial services. Furthermore, Microsoft has also stated that it has 715 partners that are working with Microsoft India in order to help customers implement a comprehensive AI strategy. Over 700 businesses and government organizations have adopted #MicrosoftAI solutions to drive breakthroughs in their #DigitalTransformation. — Microsoft India and South Asia (@MicrosoftIndia) January 16, 2019 As another strong move towards India technological advancements, Microsoft India has partnered with the think tank, NITI Aayog to combine the cloud, AI, research and its vertical expertise for new initiatives and solutions across several core areas, which includes agriculture, healthcare, and the environment. We partnered with @NITIAayog to leverage the benefits of #MicrosoftAI and #cloud for the growth of the country. #AIForAll https:\/\/t.co\/DLbUtdyUK6 — Microsoft India and South Asia (@MicrosoftIndia) January 16, 2019 That is not all, the tech giant has other plans too. It believes that the development of talent is vital for a nation’s growth and therefore, Microsoft is going to set up AI labs in 10 universities and train five lakh youth across the nation. Also, it will upskill more than 10,000 developers in the next three years. “We believe AI will enable Indian businesses and more for India’s progress, especially in education, skilling, healthcare and agriculture,” added Maheshwari. “Microsoft also believes that it is imperative to build higher awareness and capabilities on security, privacy, trust and accountability. The power of AI is just beginning to be realised and can be a game-changer for India.” Now, Microsoft’s corporate venture fund M12 has made its India entry with its first investment in healthcare data analytics startup Innovaccer. M12 has invested an additional $10 million in Innovaccer’s Series B round, leading the total investment to $35 million. Outlook With so many advancements happening in AI space — not only in India but across the world, there is no denying to the fact that AI is poised to disrupt the world. Talking about India, the nation is not at all behind in embracing this amazing tech. Having big players like Microsoft and NITI Aayog, India is definitely going to be one of those nations having the most advanced technologies.","excerpt":"From aam aadmi, to big tech firms to union ministers — everybody in India is now incorporating new tech into their lives as well as businesses. Artificial intelligence is one such technology that has gained traction in India in recent times — it has evolved in ways that far exceed its original conception. From startups […]","categories":["AI News"],"tags":["AI (Artificial Intelligence)","Microsoft","NITI Aayog"],"author_name":"Harshajit Sarmah","publish_date":"2019-01-18T04:45:03","publication_year":"2019","word_count":837,"keywords":["Go","API","artificial intelligence","Microsoft","AI","Git","RAG","analytics","Rust","GAN","AI (Artificial Intelligence)","R","NITI Aayog"],"extracted_tech_keywords":["AI","artificial intelligence","analytics","RAG","R","Go","Rust","Git","API","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/aiforall-mantra-democratising-ai-india\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":667,"title":"Analytics – the next Career Wave","content":"The last hundred years have seen the rise and dip of various professions, professions that have given rise to and fuelled revolutions of their time. Pre- independent India saw Lawyers at helm of the independence struggle. The mid sixties saw the rise of the Engineer and the Scientist which led to some of our greatest revolutions such as the green revolution. The eighties and the nineties saw the domination of Information technology and financial market professionals. The 21st century is now at the onset of an Analytics revolution. Analytics is the art and the science of analysing data in its enormity and discovering patterns in them and using these patterns to predict specific performance indicators to a very high level of accuracy. Statistical techniques and sophisticated algorithms when applied on the enormous amount of data that is the generated today can identify patterns, correlations that would otherwise be lost in the dataset. On analysing these patterns and correlations, new, powerful and game changing insights can be garnered, giving decision- making a whole new dimension.   Therefore analytics in its sophistication nullifies decision making based on gut feelings, instead allows one to make data driven decisions. It is this aspect of analytics that Redwood Associates believes that will bring about a transformational revolution in the present business sphere. Big Data, as the trove of data is called, has given rise to a new profession, that of the Analyst. A profession that a McKinsey report suggests will see a huge increase in popularity and demand as organizations are increasingly depending on data savvy analysts to gather their data and turn the huge data sets into meaningful information from which powerful insights can be garnered. So what makes an Analyst? A curious and detail oriented mind with good knowledge of analytical tools such as SAS, SPSS, R, Hadoop  and analytical techniques such as correlation, regression, cluster and factor analysis etc., makes a good Data Analyst. Advancements in the analytics industry is closely related to the advancements in technology. Therefore a good grasp of the latest technology is also necessary.  But a great Analyst is one who is also committed, one who goes the extra mile, one who is creative and has a good business sense and most importantly someone who is intuitive enough to make the right decision. The rising dominance of the Analyst can also be explained by the fact that analytics is not limited to any one field or domain; instead an analyst’s expertise can be applied across the board from Education to Manufacturing, Marketing to HR, Finance to Travel and Tourism. There simply is no boundary as to the industry eligibility of an Analyst. In the Pharmaceutical industry, analytics has been in use for a long time now, but what has changed in the last few years is the speed at which this data is collected, analysed and the insights executed. When earlier it took about 2 months to collect, analyse data and the present the insights, in present day this is done on a weekly basis if not on a daily basis. As a result, the effectiveness of strategies can be measured almost in real time, customer trends are picked up immediately, clinical trials are no longer a lengthy process and it allows for resources to be allocated efficiently.  Companies like Quintiles, Accenture are some of the leaders providing analytical driven pharmaceutical services. In HR, professionals are increasingly turning to predictive analytics to gain meaningful insights from the HR data collected from various sources. HR professionals are using analytics to predict the effectiveness of hiring, performance of employees and attrition decisions are now made based on data. LookBeyondResumes.com a product developed by Redwood Associates utilises information, technology and powerful data mining algorithms to enable an employer to shortlist smarter and a job seeker to apply better In the Manufacturing industry, the Germans and the Japanese spear headed by Deming’s, are pioneers in the application of statistics and analytics to innovate, design and improve the manufacturing process. Analytics built into the manufacturing process allows for continuous process improvement, quick detection of defects, identify customer trends early on and all this has a significant impact on the costs saved. In Marketing, analytics is applied to identify potential customers of a product or service, the size of the market, the profitability of the market. Analysts analyse the market data and supply information that affects inventory planning, procurement decisions, branding decisions, workforce planning, promotional programs etc. Analytics is widely used in the Financial sector. Banking analytics solutions are designed to help banks in their operational activities like customer acquisition, operations and risk management. Predictive modeling of customer behavior and scoring techniques enable banks in answering key questions of profitability and minimizing default risks. The scope for growth in the Analytics industry is tremendous. For example, every organization today has a website, therefore each of these organizations have a need for web analytics to measure their online performance and ROI. Thus every organization has a need for an analyst. But the analytics industry is still an up and coming one and the talent pool that is available is not sufficient to meet the demand. Recognizing this gap in the talent pool early on, Redwood Associates through its sister concern Analytics Training Institute has been  empowering individuals with knowledge in analytical tools and techniques  since 2007 and has to date empowered over 10,000 individuals and over 120 Corporates with smarter decision –making  skills. With McKinsey & Co reporting that by 2018, we would have a dire shortage of data savvy managers, the road ahead for the Analyst looks very promising.","excerpt":"The last hundred years have seen the rise and dip of various professions, professions that have given rise to and fuelled revolutions of their time. Pre- independent India saw Lawyers at helm of the independence struggle. The mid sixties saw the rise of the Engineer and the Scientist which led to some of our greatest […]","categories":["IT Services"],"tags":[],"author_name":"Maria Jose","publish_date":"2012-08-06T14:32:38","publication_year":"2012","word_count":934,"keywords":["big data","Go","programming_languages:R","AI","predictive analytics","programming_languages:Go","ViT","analytics","GAN","R"],"extracted_tech_keywords":["AI","analytics","predictive analytics","R","Go","big data","GAN","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/analytics-the-next-career-wave\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10077565,"title":"12 Ways To Use Stable Diffusion Without Coding","content":"Following Dall-E 2 and Midjourney, the deep learning model Stable Diffusion (SD) marked a leap forward in the text-to-image domain. Developed by Stability.AI, SD democratises text-conditional image generation due to its efficiency in running on customer-grade GPUs. SD is amazing, but unfortunately, it isn’t trivial to set up (especially for people without good GPUs). Here’s a list of tools built on SD with zero-technical skills needed! OneClick Tools Programmes that bundle SD in an installable programme, no separate setup and the least amount of git\/technical skill needed, usually bundling one or more UI. Diffusion Bee With a one-click installer, Diffusion Bee is a very simple way to run SD locally on M1 Mac. No dependencies or technical knowledge is needed. It runs locally on a computer; no data is sent to the cloud except requests to download the weights and check for software updates. System Requirement(s): M1\/M2 Mac16 GB RAM is preferred as it will run slow with 8GB RAMMacOS 12.5.1 or later Check the GitHub repository here. Stable Diffusion UI Another one-click installer which provides a browser UI for generating images from text and image prompts. Just enter your text prompt, and see the generated image. Currently, it does not run on Mac. System Requirement(s): Windows 10\/11 or Linux. Experimental support for Mac is coming soon.NVIDIA graphics card, preferably with 4GB or more of VRAM. Without a compatible graphics card, it’ll automatically run in the slower “CPU Mode”.Minimum 8 GB of RAM. Check the GitHub repository here. Charl-E CHARL-E packages SD into a simple application. No complex setup, dependencies, or internet is required—just download and say what you want to see. Check the GitHub repository here. NMKD Stable Diffusion GUI – AI Image Generator An ML toolkit for text-to-image generation for your local hardware. As of right now, the programme only works on Nvidia GPUs (AMD GPUs are not supported). System Requirement(s): Minimum: GPU: Nvidia GPU with 4 GB VRAM, Maxwell Architecture (2014) or newerRAM: 8 GB RAM (Note: Pagefile must be enabled as swapping will occur with only 8 GB!)Disk: 12 GB (another free 2 GB for temporary files recommended) Recommended: GPU: Nvidia GPU with 8 GB VRAM, Pascal Architecture (2016) or newerRAM: 16 GB RAMDisk: 12 GB on SSD (another free 2 GB for temporary files recommended) Check the GitHub repository here. ImaginAIry Pythonic generation of SD images with just pip install ImaginAIry. “Just works” on Linux and macOS (M1). Recent updates include memory efficiency improvements, prompt-based editing, face enhancement, upscaling, tiled images, img2img, prompt matrices, prompt variables, BLIP image captions, along with dockerfile\/colab. System Requirement(s): ~10 GB space for models to download.A computer with either a CUDA-supported graphics card or M1 processor.Preferably Python 3.10 installed.For macOS, rust and setuptools-rust must be installed to compile the tokenizer library. (Can be installed via: curl –proto ‘=https’ –tlsv1.2 -sSf https:\/\/sh.rustup.rs | sh and pip install setuptools-rust). Check the GitHub repository here. Web Distros Mage Space Unfiltered SD for the text-to-image generation. The latest feature includes Image2Image, which lets you choose an image to combine with your prompt. Check out the website here. Dreamlike.art The website is currently completely free for several more days. If you run out of credits, go to the “Buy Credits” page and click “Buy”. You won’t be charged. The balance will be reset once we exit the beta test and add payments. Check out the website here. FindAnything.App Finding images through a search engine is difficult, and you may end up accidentally publishing copyrighted images or spending a lot of money to get the images you need. The browser extension adds novel images alongside your Google image searches. You are no longer limited to a few options, as in the case for most stock images. Check out the website here. Major SD Forks The following options allow you to make changes to a project without affecting the original repository. One can fetch updates or submit changes to the original repository with pull requests. Automatic1111 – SD Web UI A browser interface based on the Gradio library for SD. Original text-to-image and image-to-image modes. One-click install and run script (but you still must install Python and git). The features include outpainting, inpainting, prompt matrix, Stable Diffusion upscale, and more. Ensure the required dependencies are met and follow the instructions for both NVidia (recommended) and AMD GPUs. Check the GitHub repository here. InvokeAI This SD version features a slick WebGUI, an interactive command-line script that combines text-to-image and image-to-image functionality in a “dream bot” style interface, and multiple features and other enhancements. The version runs on Windows, Mac and Linux machines. System Requirement(s): NVIDIA-based graphics card ~4 GB or more VRAM memory.An Apple computer with an M1 chip.~12 GB Main Memory RAM.~12 GB of disk space for the ML model, Python, and all its dependencies. Check the GitHub repository here. Waifu Diffusion Waifu Diffusion is a project based on CompVis\/Stable-Diffusion. The Stable Diffusion model is fine-tuned on weeb stuff. A model trained on Danbooru (anime\/manga drawing site) over 56k images. System Requirement(s): ~30GB of VRAM is needed.~30GB of storage if you don’t mind cleaning up every so often. Check the GitHub repository here. Basujindal: Optimized Stable Diffusion This repository is a modified version, optimised to use less VRAM than the original by sacrificing inference speed. To reduce the VRAM usage, the Stable Diffusion model is divided into four parts which are sent to the GPU when needed. Post calculation, they are returned to the CPU. The attention calculation is done in parts. Check the GitHub repository here.","excerpt":"Minus the technicality!","categories":["AI Trends"],"tags":["AI Tool","stable diffusion download"],"author_name":"Tasmia Ansari","publish_date":"2022-10-18T18:00:00","publication_year":"2022","word_count":920,"keywords":["CUDA","stable diffusion download","AI","ML","docker","RAG","Colab","Python","deep learning","Gradio","AI Tool","R"],"extracted_tech_keywords":["AI","ML","deep learning","Gradio","Colab","RAG","docker","CUDA","Python","R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/12-ways-to-use-stable-diffusion-without-coding\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10090148,"title":"GPT-4 Saves Dog’s Life","content":"Large language models or LLMs — while some view them as precursors to artificial general intelligence (AGI), others question the practical applications of these AI models in real-life situations. And since all eyes are on GPT-4 and its use-cases currently, an interesting new development has given the whole debate a fresh dimension. A Twitter user, who goes by the name Cooper, recently claimed that GPT-4 helped save his dog’s life. The Twitter thread, which soon went viral, documents how the user ran a diagnosis on his dog using GPT-4 and how the LLM helped narrow down on the underlying issue that was troubling his Border Collie, named Sassy. Though achieving AGI may still be years away, instances such as Sassy’s recovery demonstrate the potential practical applications of GPT-4. However, Sassy’s recovery with the aid of GPT-4 is not the only instance of AI being used in pet healthcare. There already exists a chatbot called PetGPT, which allows pet owners to diagnose their pet’s health issues. PetGPT generates a list of probable reasons for the animal’s illness based on species and symptoms. Meanwhile, in the Twitter thread, Cooper also mentioned that “GPT-3.5 couldn’t place a proper diagnosis, but GPT4 was smart enough to do it.” ( In Picture: Sassy) GPT-4 to the rescue Although Sassy was anaemic, she was responding positively to the treatment for a tick-borne illness, which was her initial diagnosis. However, things took a turn for the worst. Worried, Cooper rushed her to the vet again, who, after a few more tests, ruled out any co-infections associated with tick-borne diseases. However, Cooper was not convinced. Given that the vet was unable to determine the cause of Sassy’s deteriorating condition, Cooper decided to deploy GPT-4. He described Sassy’s condition in great detail to the AI model. GPT-4 acknowledged that there could be an underlying issue contributing to Sassy’s anaemia. While GPT-4 dished out the possible reasons, Cooper was able to narrow down on the right one as Sassy had already undergone a few tests. “The most impressive part was how well it read and interpreted the blood test results. I simply transcribed the CBC test values from a piece of paper, and it gave a step-by-step explanation and interpretation along with the reference ranges,” he said. With GPT-4’s assessment, Cooper knocked on the vet’s door again. When he asked if immune-mediated hemolytic anaemia (IMHA) could possibly be a reason for Sassy’s deteriorating health, the vet agreed.n Soon, Sassy was put through a new set of tests and GPT-4’s diagnosis was confirmed! Now, with Sassy fully recovered, Cooper could not help but share his experience on the micro-blogging platform. Real life use case for GPT-4 GPT-4 saving Sassy’s life does put it as a potential use case of the technology. Currently, AI is already being used by doctors across the globe as a tool to help in patient diagnosis. Most recently, doctors at MIT used AI to detect breast cancer in a woman four years ahead in time. Also, Nuance Communication, which was acquired by Microsoft last year, announced a new clinical documentation tool that uses GPT-4. “Our state-of-the-art blend of conversational, ambient, and generative AI will accelerate the advancement of the care delivery ecosystem,” Mark Benjamin, CEO of Nuance, said in a statement. In the realm of medical AI, the current direction is for AI to assist doctors in improving diagnosis rather than replacing them entirely. This is where GPT-4 has the potential to be of significance for it can be a great assistance tool for doctors. “ChatGPT can be used to assist doctors with admin tasks such as writing patient letters, so doctors can spend more time on patient interactions. More importantly, chatbots have the potential to increase the effectiveness and accuracy of the processes for preventive care, symptom identification, and post-recovery care,” Tina Deng, principal medical devices analyst at GlobalData, said in a statement. Although this development has momentarily silenced the critics looking for a significant use-case of the LLMs-based chatbots, such an ‘achievement’ is far from being fool-proof and can’t be blindly relied upon. GPT-4 hallucinates too Critics of the technology, as well the creators, have acknowledged that these chatbots can be prone to hallucinations. Sam Altman, founder of OpenAI, said, “ChatGPT (built on the GPT3.5 architecture) is incredibly limited, but good enough at some things to create a misleading impression of greatness.” Yann LeCun, chief AI scientist at Meta and one of the most popular critics of the technology, said, “Large language models have no idea of the underlying reality that language describes. While the chatbot does a great job predicting the next text in the sequence, it does not really understand the context.” Along the same line, Arvind Narayanan, an associate professor of computer science at Princeton, said that there is also a flipside to this development. “How many people put their symptoms into ChatGPT and got wrong answers, which they trusted over the doctor’s word? There won’t be viral threads about those,” he said. How true!","excerpt":"Though achieving AGI may still be years away, instances such as Sassy’s recovery demonstrate the potential practical applications of GPT-4","categories":["AI Trends"],"tags":["jordan walke"],"author_name":"Pritam Bordoloi","publish_date":"2023-03-28T13:00:00","publication_year":"2023","word_count":833,"keywords":["Go","ChatGPT","OpenAI","AI","chatbots","medical AI","Ray","Aim","generative AI","jordan walke","R"],"extracted_tech_keywords":["AI","generative AI","ChatGPT","OpenAI","Aim","Ray","chatbots","medical AI","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/gpt-4-saves-dogs-life\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10056968,"title":"6 AI and ML conferences that you cannot miss in 2022","content":"Tech enthusiasts and AI experts want to stay ahead of the game in the fast-evolving AI world. Events and conferences provide insightful platforms to these tech enthusiasts, helping them stay connected in space and also exchange the latest ideas and theories. Additionally, they provide a great platform for researchers to showcase their advancements in research and keep up with the industry trends. Analytics India Magazine has created a list of top AI and machine learning (ML) events that you should not miss in 2022. MLDS 2022 Analytics India Magazine started the Machine Learning Developers Summit in 2019 to bring together all ML practitioners and innovators under one umbrella, facilitating idea generation, sharing and experiences about ML tools. The two-day 2022 hybrid conference will be held across January 19 and 20 in Bengaluru. It promises to bring in more than 1,000 ML enthusiasts and professionals together in-person and virtually. The MLDS22 conference will be focusing on providing a platform to showcase ML frameworks and tools, hardware and software challenges of building these ML systems, evaluating new projects, using the right languages, tools, platforms, softwares and standards. Further, the conference will be providing attendees with a first look at the new developer products and trends. The offline conference will be hosting 200 attendees, with seats being available on a first-come-first-serve basis. MLDS22 is also welcoming research submissions advancing AI. You can register to attend the event here. The Rising 2022 Organised by Analytics India Magazine, the Rising is the biggest meeting of women in data science leaders from across domains and professionals. The conference serves as a platform for women from the industry and academia to come together and exchange ideas and interests, and also highlight their achievements in data science. The Rising 2022 is a two-day conference spread over April 8 and 9, scheduled to be held in Hotel Radisson Blu in Bengaluru, to facilitate talks and sessions, with the mission of empowering women in the development of leadership skills . It will also host the Women in AI Leadership Awards to celebrate the tech leaders driving disruption and innovation in Data science and AI. To get more details on the conference, click here. AIAI 2022 Conference Started in 2005, the International Conference on Artificial Intelligence Applications and Innovations is hosting its 18th edition from June 17 to 20. The hybrid conference will have its in-person event in Creta, Greece. The conference will be hosting workshops based on AI and ethics, 5G-Pine, distributed AI for resource-constrained platforms, and the 11th mining humanistic data workshop. The aim of the conference is to offer a forum to discuss innovative research and the recent advances in AI. The conference is open to attendees from academia and the industry. AI & Big Data Expo The AI & Big Data Expo 2022 will be hosted in Olympia in London on December 1 and 2, 2022. The conference will be showcasing the next-generation technologies and strategies in the AI and Big Data world. It will provide attendees with the opportunity to discover and explore the successful implementation of big data and AI that will help drive one’s business in 2022 and beyond. The conference is expecting to see a footfall of 5,000 attendees across industries— IT decision-makers, startup founders, tech providers, data scientists, data officers, heads of innovation, venture capitalists, and more. The AI and Big Data Expo is expected to host over 100 speakers and 125 exhibitors. World AI Cannes Festival The World AI Cannes Festival is a global flagship AI Event scheduled for February 10 to 12, 2022. The hybrid conference will be hosting over 100 speakers from across the globe, providing the opportunity to attendees to get a complete vision of AI in all its complexities, challenges and diversity. The five main tracks at the conference include AI for society (focusing on the benefits of AI for our society and the planet, at large); AI today and tomorrow (exploring future innovations); AI Strategy (improving AI strategy at large); AI technology (using different AI technologies); and AI applications (providing an overview of progress in the field of AI). Data+AI Summit Hosted by software company Databricks, the Data+AI Summit 2022 is scheduled for June 27 to 30 in San Francisco. The conference will also be available online. The four days at the summit promises keynotes and sessions by tech leaders and industry visionaries, offering training and networking opportunities to attendees. Some of the exciting speakers include Andrew Ng of DeepLearning.AI, Zhamak Dehghani of Thoughtworks, Daphne Koller of Coursera, Christopher Manning of SAIL, and Ali Ghodsi of Databricks, among others.","excerpt":"Analytics India Magazine has created a list of top AI and machine learning (ML) events that you should not miss in 2022.","categories":["Deep Tech"],"tags":["AI (Artificial Intelligence)","ai conferences","AI events","AIM","Databricks","Machine Learning","MLDS"],"author_name":"Debolina Biswas","publish_date":"2021-12-24T14:00:00","publication_year":"2021","word_count":763,"keywords":["data science","API","AI events","ai conferences","artificial intelligence","machine learning","AI","ML","Machine Learning","MLDS","Aim","analytics","AIM","R","AI (Artificial Intelligence)","Databricks"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","data science","analytics","Aim","Databricks","R","API"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/6-ai-and-ml-conferences-that-you-cannot-miss-in-2022\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":58231,"title":"Top Startups Working On Conversational AI In India","content":"Remember the times when, during a problem with an order from Flipkart or Amazon, one would have to wait hours on call for an agent to attend and resolve the issue? Not many years back, the same bottleneck was also faced by many while ordering food from companies like Swiggy and Zomato. Thanks to the introduction of conversational artificial intelligence (AI), which gave rise to the use of chatbots and voice assistance that are now used by major companies for resolving customer issues, driving employee engagement etc. and also for people at homes, to complete a number of minimal tasks such as changing music to switching off light bulbs. Since conversational AI has witnessed a massive growth in the Indian market and is in demand, let’s have a look at a number of startups who are bringing in solutions and products related to conversational AI. Senseforth.ai The Bengaluru based organisation came into force in the year 2017 and had been growing with innovative ideas in the field of conversational AI. Senseforth provides AI-powered bot support to a variety of companies ranging across different industries such as banking, retail, healthcare, insurance, telecom, automotive, real estate and education. With its different industry AI models, the organisation helps other companies to acquire customers with personalised interactions and engage the customers with a one-on-one conversation. Senseforth also provides the best of both worlds with its hybrid chatbots. Haptik.ai Started in 2013, Haptik chatbot startup went on to become a leader in its field by availing 16 different channels of texts and voices for users to build and deploy conversational apps. The organisation provides custom apps for both Android and iOS users and also has a personalised assistant mobile app AI for human assistance. Haptik allows a user to set a reminder, plan vacations, pay bills and access other services such as checking train status, ordering food, and booking movie tickets. The Chatmate The organisation helps to build a chat platform which allows companies to attend visitors who are checking the company’s website. From simple questions to specific problems, Chatmate is capable of curating a platform as per an organisation’s need. Chatmate provides crisp and clear connectivity to the users via touch, text and voice user experience. A few features of the platform are installation of Chatmate plugin on the website and handling multiple clients at a single point of time. Manthan Although this organisation provides prescriptive and descriptive analytics for users to grow customer engagement and recommend actions to other companies, it also consists of a natural programming language engine by the name Maya. Manthan’s Maya is designed to become a business assistant and is capable of answering a number of queries such as sales trends, the profit of the month etc. To get back to the queries, Maya is designed to go through millions of data points at a single time, analyse them using sophisticated analytics and get back to the user within seconds with an accurate answer. Bash.ai Bash.ai was founded in 2017, and the organisation brought in an innovative idea to change the way the human resource process works. The organisation helps in automating the HR system with the help of a virtual assistant and replicating cognitive functions of the HR. With the use of AI and big data, Bash’s chatbot helps organisations to communicate effectively with their employees in an automated matter which increases productivity and is superfast. One can also access the HR chatbot through messengers such as Slack and Skype. G.I Labs India The organisation aims to bring in innovation in the field of AI-based virtual assistants. Based out of Kolkata, the company focuses on helping business entities to have their customised voice and chat enabled AI virtual assistant equivalent to Apple’s Siri and Alexa of Amazon. The virtual assistant of G.I Labs is designed to understand questions regarding the products and services sold by a company and swiftly answer back. The organisation also has AI receptions powered by virtual reality which will deal with complaint booking and appointment system. Reverie Technologies The Bengaluru based company, Reverie, aims to allow organisations to communicate better with the customers through a multilingual platform. The company has created a voice assistant for non-English speakers in the country which consists of seven local languages and has named the software ‘Gopal’. The language solution brought by the Reverie provides a complete Indic language experience for the users.","excerpt":"Remember the times when, during a problem with an order from Flipkart or Amazon, one would have to wait hours on call for an agent to attend and resolve the issue? Not many years back, the same bottleneck was also faced by many while ordering food from companies like Swiggy and Zomato.  Thanks to the […]","categories":["AI Trends"],"tags":["Conversational AI","retail bi prescriptive","Startups","Voice Assistant"],"author_name":"Rohit Chatterjee","publish_date":"2020-03-08T13:00:00","publication_year":"2020","word_count":734,"keywords":["big data","Go","Voice Assistant","artificial intelligence","AI","chatbots","retail bi prescriptive","virtual assistants","Aim","Conversational AI","analytics","GAN","Startups","R"],"extracted_tech_keywords":["AI","artificial intelligence","analytics","Aim","chatbots","virtual assistants","R","Go","big data","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-startups-working-on-conversational-ai-in-india\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":49022,"title":"Can Machines Learn To Smell: Researchers May Have Cracked The Code","content":"While CNNs provide vision to a machine, RNNs give machines the ability to draw insights out of speech and sound, haptics gives robots a sense of touch, researchers have now brought a new dimension in machines — the ability to smell. While machines have been endowed with the sense of vision, speech, sound and touch thanks to the efforts of researchers for many decades, the concept of smell has eluded the machines. That is to say that humans have been unable to replicate this ability due to their own lack of understanding of olfactory functions. Every aroma is a consequence of interactions between molecules. And, predicting the relationship between a molecule’s structure and its odour still remains to be a big challenge. But now researchers are trying to understand the interplay of molecules through quantitative structure-odour relationship (QSOR) modelling. Developments Till Date Till now the representation of molecular structures has been done with the help of graph theory. The way the symmetries and the connections can be defined as edges and vertices has been intuitive for further research into the three-dimensional structure of molecules. The same formulation was exploited a couple of years ago by the researchers at Google Brain in their quest to predict properties of molecules with machine learning. Making use of the findings and moving forward, today, the researchers have leveraged graph neural networks (GNNs) to predict the olfactory properties of molecules. How Graph Neural Networks Came In Handy? via Google By viewing atoms as nodes, and bonds as edges, we can interpret a molecule as a graph. These molecular graphs are encoded into a fixed-length vector. The fragments of this molecular graph give a hint of whether an atom or certain functional group exists or not. This is where graph neural networks (GNNs) come into the picture. The researchers say that these GNNs are learnable permutation-invariant transformations on nodes and edges, which produce fixed-length vectors that are further processed by a fully-connected neural network. However, translating the vector-node molecular representation to a graph is not as straightforward as it seems to be. The translation into a graph happens as follows: Every node in the graph is first represented as a vector. For example, the node can be atomic chargeThen, every node is made to pass on its current vector value to its neighbours. An update function then generates a new vector value based on previous steps.This process is then repeated until all of the nodes are summarized into a single vector by summing or averagingThis single vector represents a molecule, which is then passed into a fully connected network The fully connected network gives an output that contains the prediction of molecules and their odour as provided by perfume experts. Researchers have also found that the learned embeddings from graph neural networks capture  a meaningful odour space representation of the underlying relationship between structure and odour, as demonstrated by a strong performance on two challenging transfer learning tasks. Future Direction Understanding the functioning of olfactory sensory reception in itself is an exciting domain. Odour perception in humans is believed to be the result of the activation of 300-400 different types of olfactory receptors, expressed in millions of olfactory sensory neurons, embedded in a small patch of tissue called the olfactory epithelium. An aroma invokes many memories. Every fragrance is associated with memory. So it can be safely assumed that deciphering the inner workings of odour-memory association work can uncover the workings of memory itself. From sniffer dogs in airports to wine connoisseurs, the sense of smell permeates into many real-world applications. Since every living and non-living thing can be melted down to the rubrics of chemistry, the use of machine learning for the same might unveil many intricacies of life.","excerpt":"While CNNs provide vision to a machine, RNNs give machines the ability to draw insights out of speech and sound, haptics gives robots a sense of touch, researchers have now brought a new dimension in machines — the ability to smell. While machines have been endowed with the sense of vision, speech, sound and touch […]","categories":["Deep Tech"],"tags":["machines","Neural Networks"],"author_name":"Ram Sagar","publish_date":"2019-10-29T18:34:04","publication_year":"2019","word_count":624,"keywords":["Go","machine learning","TPU","ELT","AI","neural network","RAG","RNN","CNN","R","machines","Neural Networks"],"extracted_tech_keywords":["AI","machine learning","neural network","RAG","TPU","R","Go","ELT","CNN","RNN"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/can-machines-learn-to-smell-researchers-may-have-cracked-the-code\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10122810,"title":"How Zoho’s Low-Code Platform is Solving Rural India’s Educational Dilemma","content":"India still faces a major challenge with educating its students, with over 1.25 million students out of school in 2022-23. States like UP, MP, Bihar, Gujarat, and Assam have the highest number of out-of-school children. Despite government initiatives to improve rural education infrastructure, such as the Sarva Shiksha Abhiyan and Digital India program, the on-ground implementation often falls short due to lack of proper supervision and tracking. Challenges related to access, quality of education, socioeconomic factors and infrastructure persist, with lack of proper classrooms, libraries, computer labs etc. “Most schools and colleges still rely heavily on paper-based and manual data collection, which is the main hurdle in providing real-time insights that can enable decision intelligence,” noted Bharath Kumar B, head of customer experience and success at Zoho Creator in a conversation with AIM. To overcome these challenges, Zoho through its low-code platform Zoho Creator has allowed NGOs like Pratham, which works to improve education for underprivileged children in rural India, to create apps to manage online classes, exams, student records, communication and more in one centralised system. Pratham uses Zoho Creator apps to facilitate better data collection, analysis, and reporting, saving significant time and effort. Institutes Embracing Zoho Creator Beyond just supporting remote education, Zoho Creator is helping drive innovation and improve experiences across the student lifecycle. For example, DAV Group of Schools built a unified custom solution consisting of 100 applications, including modules for IT ticketing, fixed asset management, career counselling, and entrance exams. “SRM University used Zoho Creator to digitise their previously manual course registration process. They were following a manual paper-drive process to manage course registration, which was not scalable to meet their growing needs,” Kumar explained. Likewise, YEG Academy, a Malaysia-based educational organisation built their entire process management application on Zoho Creator—and the platform has helped them save up to $50,000 in costs. Ensuring Data Privacy and Security With educational institutions handling sensitive student data in large volume, data privacy and security remain top concerns. Keeping this in mind, Zoho Creator offers enterprise-grade security features and complies with major data protection regulations to give schools peace of mind. “Zoho Creator incorporates robust security and privacy controls like end-to-end encryption, GDPR- and HIPAA-compliance, multi-factor authentication, data isolation, and adherence to strict security protocols and data privacy laws,” said Kumar. The company also ensures other compliance standards like ISO\/IEC 27001, SOC 2 + HIPAA. Kumar also claimed that as a company, Zoho never sells customer data and automatically deletes data from terminated accounts within six months. Empowering Educators and Students In addition to back-office functions, Zoho Creator is being used to build innovative tools that directly impact teaching and learning. One example is the Teacher Training Management System (TTMS) built for the Department of State Education Research and Training of Karnataka. Developed in collaboration with the Azim Premji Foundation, TTMS is transforming teacher education in India. The platform has already enabled  TTMS, a cloud-based platform, facilitating over 2,000 training sessions for more than 200,000 teachers. Moreover, the platform offers a wide range of study materials, including documents, PDFs, slideshows, and audio and video files, which teachers can access to prepare for training sessions. Supporting Skill Development Beyond K-12 and higher education, Zoho Creator is helping with skill development and vocational training. YEG Academy, which offers career guidance and education aligned with job market demands, was able to build a comprehensive system for managing students, enrollment, sales, customer service, finance, and HR on the platform. Zoho has launched programmes like Young Creators Programme (YCP) to teach college students the basics of building low-code applications using Zoho Creator. “In this free workshop, students are taught the basics of creating applications and solutions using Zoho Creator. They also learn aspects like automating business processes, managing data relationships, and utilising business intelligence and analytics,” Kumar added. To date, YCP has collaborated with 32 educational institutions across 20 cities in India and introduced over 7,040 students globally to low-code development. By empowering the next generation with the tools to build applications, Zoho is helping create a pipeline of talent for the digital economy. The platform also enables schools to digitise the entire admissions process, from inquiry to acceptance and enrollment. Features like online application forms, document upload, eligibility checks, and applicant communication can all be handled through a custom portal. Beyond academics, Zoho Creator can also digitise campus processes virtually, such as vehicle fleet management, classroom asset tracking, library reservations, and dining hall orders. As the education sector continues to evolve, Low-code will also enable schools to take advantage of emerging technologies like AI, machine learning, and augmented\/virtual reality. Zoho Creator already offers an AI Modeler feature that allows users to build intelligent applications without needing data science expertise. Enabling Data-Driven Decision Making Perhaps the most significant benefit of using a platform like Zoho Creator is the ability to collect and analyse data across the institution. When processes are digitised and centralised, schools gain visibility into areas that were previously siloed or opaque. “Zoho Creator applications run on the same infrastructure as these services. We provide a proprietary cloud-native platform that was specifically designed to scale according to the needs of an application,” said Kumar. “All parts of the application scale automatically—from user request management and storage resources to computational abilities.” With all data in one place, schools can generate reports and dashboards to track KPIs, identify trends, and make data-driven decisions. For example, administrators can analyse enrollment data to optimise course offerings, or use student performance data to identify at-risk students and intervene early. Worthy Alternatives? Besides Zoho Creator there are other low-code platforms and education management systems used in India, like Edisapp, Fedena, and Academia ERP, which could be considered alternatives or competitors to in the Indian education market. Other global low-code platforms like OutSystems, Mendix, Planet Crust, and BP Logix demonstrate use cases and adoption in education, but they don’t specifically have a presence in India. Creatrix Campus, is one such an AI-driven, cloud-based platform specifically designed for higher education institutions, is gaining significant traction in India and globally. The platform offers end-to-end solutions for automating student and faculty lifecycles, learning and teaching, aiming to provide exceptional experiences. Key features include complete student lifecycle management, faculty management, outcome-based learning tools, analytics and reporting, a secure and customisable cloud platform, and a mobile-first approach. Creatrix Campus is already being used by over 150+ institutions in 28 countries, including India, with 100,000+ users. While both platforms allow educational institutions to develop custom applications with minimal coding, Creatrix Campus appears to be more specialised for the higher education sector, whereas Zoho Creator offers flexibility for a broader range of educational use cases. In comparison, Zoho Creator is a more general-purpose low-code application development platform that caters to various industries, including education. It enables users to build custom applications for admissions, course management, student records, and other educational processes.","excerpt":"Indian Institutes like DAV, SRM and NGOs like Pratham are leveraging Zoho Creator to enable and fastrack administration, online classes etc.","categories":["AI Features"],"tags":["zoho"],"author_name":"Shyam Nandan Upadhyay","publish_date":"2024-06-07T13:43:35","publication_year":"2024","word_count":1145,"keywords":["data science","zoho","Go","machine learning","AWS","AI","RAG","Aim","analytics","Rust","R"],"extracted_tech_keywords":["AI","machine learning","data science","analytics","Aim","RAG","AWS","R","Go","Rust"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-zohos-low-code-platform-solving-indias-educational-dilemma\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":52797,"title":"Top 11 Tools For Distributed Machine Learning","content":"There are two fundamentally different and complementary ways of accelerating machine learning workloads: By vertical scaling or scaling-up, where one adds more resources to a single machine Or 2. By horizontal scaling or scaling-out, where one adds more nodes to the system But when it comes to the degree of distribution within a machine learning ecosystem, they are classified as: CentralisedDecentralisedFully Distributed Centralised systems employ a strictly hierarchical approach. But the distributed system consists of a network of independent nodes and where no specific roles are assigned to certain nodes. A centralised solution is not the right choice when data is inherently distributed or too big to store on single machines. For instance, think about astronomical data that is too large to move and centralise. In a recent work published by the researchers at Delft University of Technology, Netherlands, they wrote in detail about the current state-of-the-art distributed ML models and how they affect computation latency and other attributes. The advantages of using distributed ML models are plenty, it is beyond the scope of this article, however, here we list down of popular toolkits and techniques that enable distributed machine learning: MapReduce and Hadoop MapReduce is a framework for processing data and was developed by Google in order to process data in a distributed setting. First, all data is split into tuples during the map phase, which is followed by the reduce phase, where these tuples are grouped to generate a single output value per key. MapReduce and Hadoop heavily rely on the distributed file system in every phase of the execution. Apache Spark Transformations in linear algebra, as they occur in many machine learning algorithms, are typically highly iterative in nature and the paradigm of the map and the reduce operations are not ideal for such iterative tasks. This is what Apache Spark has been developed to resolve. The key difference here is the MapReduce tasks, which would require to write all (intermediate) data to disk for it to be executed. Whereas, Spark can keep all the data in memory, which saves expensive reads from the disk. Baidu AllReduce AllReduce uses common high-performance computing technology to iteratively train stochastic gradient descent models on separate mini-batches of the training data. Baidu claims linear speedup when applying this technique in order to train deep learning networks. Horovod Horovod like Baidu, adds a layer of AllReduce-based MPI training to Tensorflow. Horovod uses the NVIDIA Collective Communications Library (NCCL) for increased efficiency when training on (Nvidia) GPUs. However, Horovod lacks fault tolerance and therefore suffers from the same scalability issues as those of Baidu’s. Caffe2 This deep learning framework distributes machine learning through AllReduce algorithms. It does this by using NCCL between GPUs on a single host, and custom code between hosts based on Facebook’s Gloo library. Microsoft Cognitive Toolkit . This toolkit offers multiple ways of data-parallel distribution. Many of them use the Ring AllReduce tactic as previously described, making the same trade-off of linear scalability over fault-tolerance. DistBelief Developed by Google, DistBelief is one of the early practical implementations of large-scale distributed machine learning. It supports data and model parallel training on tens of thousands of CPU cores. They are also capable of training a huge model with 1.7 billion parameters. Tensorflow Developed by Google, Tensorflow has evolved from DistBelief and borrows the concepts of a computation graph and parameter server from it. Unlike DistBelief, defining a new type of neural network layer in Tensorflow requires no custom code, composed of fundamental math operations. DIANNE (Distributed Artificial Neural Networks) A Java-based distributed deep learning framework, DIANNE, uses the Torch native backend for executing the necessary computations. Each basic building block of a neural network can be deployed on a specific node, hence enabling model-parallelism. MXNet On a small cluster of 10 machines equipped with a GPU, MXNet achieves almost linear speedup compared to a single machine when training GoogleNet. Similar to that of Tensorflow, the models are represented as dataflow graphs. Petuum This approach is aimed at exploiting ML’s error tolerance, dependencies, and non-uniform convergence in order to achieve good scalability on large datasets. Petuum uses the Parameter Server paradigm to keep track of the model being trained. Petuum provides an abstraction layer that also allows it to run on systems using the Hadoop job scheduler and HDFS (Hadoop file system), which simplifies compatibility with the pre-existing clusters. Scaling out is still a pressing challenge that delays the widespread usage of distributed models. Not all machine learning algorithms lend themselves to a distributed computing model that can achieve a high degree of parallelism.","excerpt":"There are two fundamentally different and complementary ways of accelerating machine learning workloads:  By vertical scaling or scaling-up, where one adds more resources to a single machine  Or  2. By horizontal scaling or scaling-out, where one adds more nodes to the system But when it comes to the degree of distribution within a machine learning […]","categories":["AI Trends"],"tags":["Apache Spark","Caffe2","distributed machine learning","Mapreduce","Tensorflow","tensorflow gradient"],"author_name":"Ram Sagar","publish_date":"2019-12-30T12:06:00","publication_year":"2019","word_count":762,"keywords":["machine learning","TPU","Caffe2","distributed machine learning","AI","neural network","ML","distributed computing","Apache Spark","Mapreduce","Aim","deep learning","tensorflow gradient","TensorFlow","Tensorflow"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","neural network","Aim","TensorFlow","distributed computing","TPU","Apache Spark"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-tools-distributed-machine-learning-tensorflow\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":39443,"title":"5 Best Research Papers on Computational Linguistics For Your Reading List","content":"As one of the premier institutes for technology, Massachusetts Institute of Technology (MIT) has several prominent research which has resulted in many ground-breaking technological advancements. In this article, we take a look at the top five recent research papers from on Computational Linguistics from the institute. 1. Learning an Executable Neural Semantic Parser Authors: Jianpeng Cheng, Siva Reddy, Vijay Saraswat, and Mirella Lapata Abstract: This article describes a neural semantic parser that maps natural language utterances ontological forms that can be executed against a task-specific environment, such as a knowledge base or a database, to produce a response. The parser generated tree-structured logical forms with a transition-based approach, combining a generic tree-generation algorithm with domain-general grammar defined by the logical language. Research methodology: To tackle mismatches between natural language and logical form tokens, various attention mechanisms were explored. Finally, the researchers considered different training settings for the neural semantic parser, including fully supervised training where annotated logical forms were given, weakly supervised training where denotations were provided, and distant supervision where only unlabeled sentences and a knowledge base are available. Access the research: Learning an Executable Neural Semantic Parser 2. Unsupervised Compositionality Prediction of Nominal Compounds Authors: Silvio Cordeiro, Aline Villavicencio, Marco Idiart and Carlos Ramisch Abstract: Nominal compounds such as red wine and nut case display a continuum of compositionality, with varying contributions from the components of the compound to its semantics. This article proposes a framework for compound compositionality prediction using distributional semantic models, evaluating to what extent they capture idiomaticity compared to human judgments. Research methodology: For evaluation, the researchers introduced data sets containing human judgments in three languages: English, French, and Portuguese. The results obtained reveal a high agreement between the models and human predictions, suggesting that they were able to incorporate information about idiomaticity. Access the research: Unsupervised Compositionality Prediction of Nominal Compounds 3. Automatic Inference of Sound Correspondence Patterns across Multiple Languages Authors: Johann-Mattis List Abstract: The researcher presented an automatic method for the inference of sound correspondence patterns across multiple languages based on a network approach. The core idea was to represent all columns in aligned cognate sets as nodes in a network with edges representing the degree of compatibility between the nodes. Research methodology: The task of inferring all compatible correspondence sets can then be handled as the well-known minimum clique cover problem in graph theory, which essentially seeks to split the graph into the smallest number of cliques in which each node is represented by exactly one clique. The resulting partitions represent all correspondence patterns that can be inferred for a given data set. By excluding those patterns that occur in only a few cognate sets, the core of regularly recurring sound correspondences can be inferred. Based on this idea, the article presents a method for automatic correspondence pattern recognition, which is implemented as part of a Python library which supplements the article. Access the research: Automatic Inference of Sound Correspondence Patterns across Multiple Languages 4. A Sequential Matching Framework for Multi-Turn Response Selection in Retrieval-Based Chatbots Authors: Yu Wu, Wei Wu, Chen Xing, Can Xu, Zhoujun Li, and Ming Zhou Abstract: The researchers studied the problem of response selection for multi-turn conversation in retrieval-based chatbots. The task involved matching a response candidate with a conversation context, the challenges for which include how to recognize important parts of the context, and how to model the relationships among utterances in the context. Research Methodology: Using a new matching framework called sequential matching framework (SMF), the researchers proposed a sequential convolutional network and sequential attention network and conducted experiments on two public data sets to test their performance. Experiment results show that both models can significantly outperform state-of-the-art matching methods. The researchers also show that the models are interpretable with visualisations that provide us insights on how they capture and leverage important information in contexts for matching. Access the research: A Sequential Matching Framework for Multi-Turn Response Selection in Retrieval-Based Chatbots 5. Parsing Chinese Sentences with Grammatical Relations Authors: Weiwei Sun, Yufei Chen, Xiaojun Wan and Meichun Liu Abstract:  The research represents grammatical information using general directed dependency graphs. Both only-local and rich long-distance dependencies are explicitly represented. Research methodology: To create high-quality annotations, the researchers took advantage of an existing TreeBank, namely, Chinese TreeBank (CTB), which is grounded on the Government and Binding theory. Two key problems as addressed by the researchers include (a) how to decompose a complex graph into simple subgraphs, and (b) how to combine subgraphs into a coherent complex graph. For transition-based parsing, the researchers introduced a neural parser based on a list-based transition system. They also discussed several other key problems, including dynamic oracle and beam search for neural transition-based parsing. The evaluation gauged how successful GR parsing for Chinese can be by applying data-driven models. The empirical analysis suggests several directions for future study. Access the research: Parsing Chinese Sentences with Grammatical Relations","excerpt":"As one of the premier institutes for technology, Massachusetts Institute of Technology (MIT) has several prominent research which has resulted in many ground-breaking technological advancements. In this article, we take a look at the top five recent research papers from on Computational Linguistics from the institute. 1. Learning an Executable Neural Semantic Parser Authors: Jianpeng […]","categories":["AI Trends"],"tags":["computational linguistics","MIT","NLP","recent technological advancements"],"author_name":"Akshaya Asokan","publish_date":"2019-05-20T11:14:46","publication_year":"2019","word_count":817,"keywords":["recent technological advancements","Go","attention mechanism","programming_languages:R","AI","chatbots","MIT","data-driven","RAG","NLP","Python","computational linguistics","programming_languages:Python","R"],"extracted_tech_keywords":["AI","RAG","chatbots","Python","R","Go","attention mechanism","data-driven","programming_languages:Python","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/5-research-papers-on-computational-linguistics-for-your-reading-list\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":23954,"title":"How Flipkart’s Machine Thinking Approach Will Gain Momentum After Its Speculated Acquisition By Walmart","content":"The news about global retail giant Walmart acquiring Indian e-commerce company Flipkart has been doing rounds for quite a while now. Reports have suggested that Walmart is likely to buy a majority stake in Flipkart by the end of June 2018. If the deal materialises, it would be Walmart’s biggest acquisition of an online business standing at 51 percent stake, priced between $10-12 billion. While the two companies have been extensively using technologies like artificial intelligence and machine learning, among others in areas like customer service, fraud management, and tracking shipment, will it lead to enhanced tech capabilities for the companies? Analytics And AI Are A Big Priority For Flipkart The company has been focussing on AI and ML since 2015, which was evident from its acquisition of F7 Labs in the Silicon Valley. Currently, the Indian e-commerce company has dozens of ML algorithms at disposal, starting from personalised recommendation to tracking shipments. Aiming to make every decision data-driven and model-driven, Flipkart also announced a stupendous job opening in data sciences and other tech areas hiring for various positions. Making a headway in new tech, Flipkart created an internal unit called AIforIndia, in order to put ML and AI at the core of its business. The company also made a statement few months ago about their ambitious plan to invest “hundreds of millions of dollars” in the AI initiative over the next few years. This is clearly demonstrated by the fact that the company has already begun the recruitment process for positions like AI experts, data scientists, software developers and others. The new team which will be headed by the chief data scientist Mayur Datar, aims to work on some of the key projects around AI. Flipkart believes that even though applying AI in the online retail ecosystem can do wonders, India’s problems are unique and therefore the focus areas for AI in developed countries cannot be applied back home. As the company said, they will solve problems differently because the underlying problems in India are quite different. Another important initiative by Flipkart in the space is Project Mira — an AI-based personal shopper, which saw a soft launch last year. They launched the first version of this conversational search experience in 2017, guiding shoppers with relevant questions, conversational filters, shopping ideas and others. The idea was to replicate offline shopping experience online. Similarly, Myntra, Flipkart’s fashion unit has been leveraging AI, ML and big data analytics to forecast fashion trends. Its indigenously developed AI-powered “fashionista robot” Artie, gathers consumer insight to create a range of ongoing trend under the brand ModaRapido. Flipkart had also reportedly partnered with Microsoft to deploy AI and ML-based solutions to run future sales for the e-commerce unicorn. Reportedly, Flipkart has also been working in the areas of image recognition, speech recognition, text recognition, personalisation, recommendation and others. They are reportedly also applying AI in areas like customer service, warehousing, logistics. They also have an ML model for classification of fraudulent transactions. Apparently its cloud has over a million computer processor units that the company keeps upgrading to make it work faster and more efficiently. It also aims to soon add GPUs from Nvidia, which is known to be enhancing AI processing to a significant level. ML And AI Is Crucial At Walmart Walmart in 2017 planned to build a neural network cluster on Nvidia’s AI chips to allow Walmart’s OneOps team to build and maintain the company’s internal application development system. It was reported that Walmart is planning to buy a large number of Nvidia GPUs, which are useful for performing ML calculations. It acquired OneOps in 2013 and released an open source version of the platform-as-a-technology last year. Earlier this year, Walmart tested an AI-powered robot, using an autonomous robot built by start-up Bossa Nova. The two-feet tall robot can complete a review of an 80-foot section in just two minutes and identify items that are out-of-stock, recognise incorrect prices, and spot shelves with missing labels. It uses AI to analyse the raw data and calculate the status of each product on the shelves including its location, price and availability. It was also one of the early adopters of RFID to track inventory. It also uses ML algorithms for tasks such as organising inventory data, pricing items and enhancing customer shopping experience, among others. Just like Flipkart, it had embraced ML to improve both offline and online store shopping experience. Collecting customer data into a variety of usable insights to improve personalisation, recommendation is another similar approach by Walmart. It could also be said that the retail giant’s ML innovations have only just begun. With 11,700 physical stores in 28 countries, Walmart is aiming to use to technology to win people. From voice commerce to automated warehouse, Walmart is ambitious of a tech play at the company. The company also recently announced that it is opening emerging technologies office in Dallas, US, with a plan to hire more people in the facility by the end of the year. The team would be focused on conducting research in ML, computer vision and IoT. The office will also explore how to use ML to aid with internal company matters, such as determining where the optimal location for a new Walmart store is based on traffic patterns and weather data. Will Filpkart And Walmart Strengthen Each Other’s Tech Capabilities? While Walmart is looking to buy a large stake in Flipkart — one of its biggest overseas deals — it will be an opportunity for the world’s largest retail giant to tap into the Indian market like never before. On one hand, Flipkart will have an access to massive capital to take on Amazon (Flipkart is speculated to be valued at $21 billion after the deal), and open doors for exclusive deals with global brands along with access to Walmart’s world-class sourcing. On the other hand, Walmart can explore an opportunity to open offline retail stores in India. Both companies would have then chance to compete with Amazon, which is a strong contender both in terms of technology adoption and popularity chart. This union of Walmart and Flipkart will open new avenues for both the companies to explore the strengths in terms of technology including AI, ML and data analytics, given the fact that both have a strong hand holding in these new technologies. As the two forces join hands, it could open doors for them to explore onto the AI and ML play in the Indian scenario hence helping each other have a stronger hold in the country, delivering better services and opening new opportunities.","excerpt":"The news about global retail giant Walmart acquiring Indian e-commerce company Flipkart has been doing rounds for quite a while now. Reports have suggested that Walmart is likely to buy a majority stake in Flipkart by the end of June 2018. If the deal materialises, it would be Walmart’s biggest acquisition of an online business […]","categories":["IT Services"],"tags":["Walmart Labs"],"author_name":"Srishti Deoras","publish_date":"2018-04-24T06:26:23","publication_year":"2018","word_count":1099,"keywords":["data science","artificial intelligence","machine learning","AI","neural network","ML","Walmart Labs","computer vision","RAG","Aim","analytics"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","neural network","computer vision","data science","analytics","Aim","RAG"],"url":"https:\/\/analyticsindiamag.com\/it-services\/how-flipkarts-machine-thinking-approach-will-gain-momentum-after-its-speculated-acquisition-by-walmart\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10102543,"title":"8 Most Influential AI Inventions","content":"AI has been the word of the year. Even the legacy magazine TIME has not been able to overlook the impact of the technology leading it to adding it to its list of best inventions of 2023. The technology has managed to seep into every nook and cranny of our lives, from educational institutions to government departments. From the 14 inventions TIME enlisted, we’ve picked out the best of the lot. Here’s are 8 best AI inventions that have managed to leave a mark in the constantly evolving industry: Adobe Photoshop Generative Expand and Generative Fill In May, amid the fake AI image frenzy, Adobe introduced Generative Fill in Photoshop, a feature which brings generative AI-powered Firefly directly into the hands of designers. The technology had already wooed the world. And though Adobe was not the first to bring the text-to-image technology, but it managed to take advantage of its existing user base of over 30 million creators. Since its release, Generative Fill has evolved and integrated in most aspects of Adobe, from Creative Cloud to GenStudio for enterprise customers. AI Wildfire Detector AI is trying to be used in every area one can think of today – including putting out a fire before it explodes. The California Department of Forestry and Fire Protection announced the AI-powered Wildfire Detector to upgrade firefighters’ abilities to validate situational awareness and respond quickly. The department, along with the University of California at San Diego’s Alert California program and its network of more than 1,000 cameras across the state, is using the technology to spot fires early which have been getting worse with each passing year. Gen-2 Built upon the diffusion architecture by Runway, a well-funded startup, Gen 2 has often been called ‘pivotal’ in terms of AI video generation and editing. Released in March 2023, Gen-2 improved on its preceding Gen-1 model allowing users to generate new four-second-long videos from scratch through prompting its proprietary AI model. Users also had the option to upload images to which Gen-2 could add motion. The update in September introduced a ‘Director’s mode’ for users to create visuals by choosing the direction and speed of the movement in their Runway generated clips. More recently, the model has had major improvements to both its fidelity and consistency of results. City-Wide Drone Detection Launched in April by Dedrone, the detection system is the largest airspace security infrastructure in the world. The system is available across 40 cities in the West in their Security Operating Center (SOC) or on their phone without any hardware installation. The network of services is completely cloud based, available on the web as well as the stores on Android and iOS. The system’s database of drone information covers three years and over 240 million drone detections’ worth of drone behaviour, including FAA violations. GPT-4 The long-awaited GPT-4 is not unknown to anyone on the internet. The AI model powering OpenAI’s paid version of ChatGPT, became the talk of the tech town for all the reasons right and wrong. While the model surpassed its predecessor and other language models available in the zone, it was highly criticised for the lack of information available about its architecture, dataset, and training process. As Sam Altman’s brainchild is nearing its first conference, Dev Day, leaks surfacing the internet point towards an incoming advanced GPT-4 for enterprise users. SeamlessM4T One major factor lagging in AI research is the diversity and availability of models in several languages. Taking a step forward, tech giant Meta open-sourced SeamlessM4T — Massively Multilingual & Multimodal Machine Translation model. The translation and transcription all-in-one model facilitates input and output in 100 languages and speech output in 35 languages (including English). TrailGuard AI The invention by Resolve, is one of the pieces of technology saving nature one algorithm at a time. The AI-enabled security system is an anti-poaching solution in action to detect poachers and alert park managers in near-real time. The not-for-profit organisation partnered with Steve Gulick (a conservation technologist) and Intel to develop this cryptic anti-poaching camera-based alert system to prevent animals from being poached. Project Gutenberg Open Audiobook Collection The oldest digital library in the world, Project Gutenberg transformed over 5,000 ebooks into audiobooks through synthetic speech technology. The audiobooks, made possible through a collaboration with Microsoft and MIT, are available on Spotify, Apple Podcasts, and Google Podcasts. The project bypassed the lengthy and expensive process of hiring a human reader to do the job. While the  overnight success made it to the top AI innovations list, It’s exactly the kind of AI application that actors were striking against in the US.","excerpt":"From the best AI inventions TIME enlisted, we’ve picked out our favourites. Take a look.","categories":["AI Trends"],"tags":["Top Trend"],"author_name":"Tasmia Ansari","publish_date":"2023-11-06T13:30:00","publication_year":"2023","word_count":768,"keywords":["Top Trend","ChatGPT","Go","TPU","OpenAI","AI","ML","Git","GPT","generative AI","R"],"extracted_tech_keywords":["AI","ML","generative AI","ChatGPT","OpenAI","TPU","R","Go","Git","GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/8-most-influential-ai-inventions-of-2023\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10018378,"title":"Top Ten BERT Alternatives For NLU Projects","content":"The last few years have witnessed a wider adoption of Transformer architecture in natural language processing (NLP) and natural language understanding (NLU). Bidirectional Encoder Representations from Transformers or BERT set new benchmarks for NLP when it was introduced by Google AI Research in 2018. The model has paved the way to newer and enhanced models. Here is a compilation of the top ten alternatives of the popular language model BERT for natural language understanding (NLU) projects. 1| GPT-2 and GPT-3 by OpenAI In 2019, OpenAI rolled out GPT-2 — a transformer-based language model with 1.5 Billion parameters and trained on 8 million web pages. The model comes armed with a broad set of capabilities, including the ability to generate conditional synthetic text samples of good quality. OpenA launched GPT-3 as the successor to GPT-2 in 2020. GPT-3 is an autoregressive language model with 175 billion parameters, ten times more than any previous non-sparse language model. The model, equipped with few-shot learning capability, can generate human-like text and even write code from minimal text prompts. Know more here. 2| XLNet by Carnegie Mellon University XLNet is a generalised autoregressive pretraining method for learning bidirectional contexts by maximising the expected likelihood over all permutations of the factorization order. XLNet uses Transformer-XL and is good at language tasks involving long context. Due to its autoregressive formulation, the model performs better than BERT on 20 tasks, including sentiment analysis, question answering, document ranking and natural language inference. Know more here. 3| RoBERTa by Facebook Developed by Facebook, RoBERTa or a Robustly Optimised BERT Pretraining Approach is an optimised method for pretraining self-supervised NLP systems. The model is built on the language modelling strategy of BERT that allows RoBERTa to predict intentionally hidden sections of text within otherwise unannotated language examples. It also modifies key hyperparameters in BERT, including removing BERT’s next-sentence pretraining objective and training with much larger mini-batches and learning rates. Know more here. 4| ALBERT by Google ALBERT or A Lite BERT for Self-Supervised Learning of Language Representations is an enhanced model of BERT introduced by Google AI researchers. The model incorporates two parameter reduction techniques to overcome major obstacles in scaling pre-trained models. According to its developers, the success of ALBERT demonstrated the significance of distinguishing the aspects of a model that give rise to the contextual representations. It has significantly fewer parameters than a traditional BERT architecture. Know more here. 5| DistilBERT by Hugging Face DistilBERT is a distilled version of BERT. DistilBERT is a general-purpose pre-trained version of BERT, 40% smaller, 60% faster and retains 97% of the language understanding capabilities. Know more here. 6| StructBERT by Alibaba Developed by the researchers at Alibaba, StructBERT is an extended version of the traditional BERT model. StructBERT incorporates language structures into BERT pre-training by proposing two linearisation strategies. In addition to the existing masking strategy, StructBERT extends BERT by leveraging the structural information, such as word-level ordering and sentence-level ordering. According to its developers, StructBERT advances the state-of-the-art results on a variety of NLU tasks, including the GLUE benchmark, the SNLI dataset and SQuAD v1.1 question answering task. Know more here. 7| DeBERTa by Microsoft DeBERTa or Decoding-enhanced BERT with Disentangled Attention is a Transformer-based neural language model that improves the BERT and RoBERTa models using two novel techniques such as a disentangled attention mechanism and an enhanced mask decoder. DeBERTa is pre-trained using MLM. Know more here. 8| Text-to-Text Transfer Transformer (T5) by Google Text-to-Text Transfer Transformer (T5) is a unified framework that converts all text-based language problems into a text-to-text format. In contrast to BERT-style models that can only output either a class label or a span of the input, T5 reframes all NLP tasks into a unified text-to-text-format where the input and output are always text strings. The text-to-text framework allows the use of the same model, loss function, and hyperparameters on any NLP task, including machine translation, document summarisation, question answering as well as classification tasks. Know more here. 9| UniLM by Microsoft Developed by Microsoft, UniLM or Unified Language Model is pre-trained using three types of language modeling tasks: unidirectional, bidirectional, and sequence-to-sequence prediction. The unified modeling is achieved by employing a shared Transformer network and utilising specific self-attention masks to control what context the prediction conditions on. The model can be fine-tuned for both natural language understanding and generation tasks. UNILM achieved state-of-the-art results on five natural language generation datasets, including improving the CNN\/DailyMail abstractive summarisation ROUGE-L. Know more here. 10| Reformer by Google Reformer is a Transformer model designed to handle context windows of up to one million words; all on a single accelerator. Introduced by Google AI researchers, the model takes up only 16GB memory and combines two fundamental techniques to solve the problems of attention and memory allocation that limit the application of Transformers to long context windows. Know more here.","excerpt":"The last few years have witnessed a wider adoption of Transformer architecture in natural language processing (NLP) and natural language understanding (NLU). Bidirectional Encoder Representations from Transformers or BERT set new benchmarks for NLP when it was introduced by Google AI Research in 2018. The model has paved the way to newer and enhanced models.   […]","categories":["AI Trends"],"tags":[],"author_name":"Ambika Choudhury","publish_date":"2021-01-18T16:00:00","publication_year":"2021","word_count":809,"keywords":["Hugging Face","TPU","OpenAI","AI","sentiment analysis","ML","Transformers","RAG","NLP","few-shot learning"],"extracted_tech_keywords":["AI","ML","NLP","OpenAI","Hugging Face","Transformers","RAG","few-shot learning","sentiment analysis","TPU"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-ten-bert-alternatives-for-nlu-projects\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":26927,"title":"AI Obliterates Semi-Professional Dota 2 Players In A Game","content":"Open AI, an artificial intelligence research firm funded by Elon Musk, raised many impressed eyebrows this week when a team of AI bots created by them defeated human players in the popular online multiplayer video game Dota 2. OpenAI Five, the Dota 2 team (of bots) played against semi-professional players who were reportedly ranked in the 99.95th percentile in the world. To everyone’s awe, the bots beat the human players, which included former Dota 2 professionals, 2-1 in a set of three games. In a three-game series, OpenAI Five won the first two games comprehensively. In the final game, the OpenAI team let the audience select their team of five heroes when it lost to the human team. Great work by @OpenAI. Need the neural interface soon to enable human\/AI symbiosis. — Elon Musk (@elonmusk) August 6, 2018 Exciting match today! OpenAI Five won first 2 games, and humans won 1 with the audience picking the draft in favor of the humans. Thank you to our fantastic Team Human! — Christy Dennison (@cbd) August 6, 2018 This is not the first time an AI has managed to beat humans in any game. Google’s DeepMind lab also made a tentative step towards general AI with AlphaGo Zero. As reported earlier by Analytics India Magazine, it started from random play, and was given no domain knowledge except the game rules. Even the, AlphaZero achieved a superhuman level of play in the games of chess, shogi and Go within 24 hours. It also convincingly defeated a world-champion program in each case. OpenAI Five has been working towards their goal of competing against the best pro team at The International which is the largest annual Dota tournament in the world. This year’s event is set to take place between 20-25 August in Vancouver.","excerpt":"Open AI, an artificial intelligence research firm funded by Elon Musk, raised many impressed eyebrows this week when a team of AI bots created by them defeated human players in the popular online multiplayer video game Dota 2. OpenAI Five, the Dota 2 team (of bots) played against semi-professional players who were reportedly ranked in the 99.95th percentile […]","categories":["AI News"],"tags":[],"author_name":"Prajakta Hebbar","publish_date":"2018-08-06T11:53:28","publication_year":"2018","word_count":299,"keywords":["Go","artificial intelligence","OpenAI","AI","programming_languages:R","programming_languages:Go","analytics","R"],"extracted_tech_keywords":["AI","artificial intelligence","analytics","OpenAI","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/ai-obliterates-semi-professional-dota-2-players-in-a-game\/","complexity_score":4,"technical_depth":8,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10163118,"title":"WNS Reports $333 Mn Revenue in Q3 FY25, Highlighting AI Investments","content":"WNS, the business transformation and services company, reported its fiscal third-quarter earnings for 2025 in late January, showcasing revenue growth and profitability while emphasising its commitment to AI and GenAI-driven transformation. The company recorded $333 million in revenue, marking a 2.1% year-over-year increase and a 3.2% sequential rise. However, full-year guidance was revised downward due to market headwinds. Despite economic pressures, WNS continues to focus on AI and drive domain-led process automation and cost reduction. “We continue to make solid progress moving large transformational opportunities through the pipeline and are focused on closing these large deals to help accelerate revenue growth,” said CEO Keshav Murugesh. “WNS remains committed to our ongoing investments in domain expertise, data and analytics, and technology-enabled offerings leveraging AI and GenAI to ensure our ability to deliver long-term sustainable value to all of our stakeholders.” Speaking with AIM earlier, Gautam Singh, business unit head of WNS Analytics, mentioned that they are expecting revenue growth with generative AI in play. “About 5% of our revenue in the fiscal year 2025 is expected to be influenced by generative AI. I can cite examples where we’ve achieved 30-40% efficiency improvements through analytics, AI, and automation initiatives,” he had said. In Q3, the company added seven new clients and expanded 52 existing relationships, underscoring its growing AI-driven service portfolio. This aligns with WNS’s broader strategy of integrating automation and AI-powered analytics into its digital transformation services. WNS posted a net profit of $48.6 million, up from $41.5 million in Q3 last year, supported by favourable currency movements and cost optimisation strategies. Adjusted net income (ANI) stood at $47.0 million, down from $58.5 million in the previous year, largely due to non-recurring tax benefits recorded in fiscal 2024. The company ended the quarter with $231.5 million in cash and investments while reducing its debt to $199.6 million. “Our guidance for the full year reflects revenue less repair payments of -2% to -1% on a reported basis and -3% to -1% on a constant currency basis,” said CFO Arijit Sen. “ANI guidance includes a one-time benefit in Q4 of $12.2 million relating to a facility asset sale in India.”","excerpt":"The company added seven new clients and expanded 52 existing relationships, underscoring its growing AI-driven service portfolio.","categories":["AI News"],"tags":["WNS"],"author_name":"Mohit Pandey","publish_date":"2025-02-10T13:35:20","publication_year":"2025","word_count":357,"keywords":["Go","GenAI","AI","WNS","digital transformation","Git","RAG","Aim","generative AI","analytics","R"],"extracted_tech_keywords":["AI","analytics","generative AI","GenAI","Aim","RAG","R","Go","Git","digital transformation"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/wns-reports-333-mn-revenue-in-q3-fy25-highlighting-ai-investments\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10075570,"title":"India &#038; US to Boost Local Manufacturing as China Shrinks in Chip Output","content":"China witnessed its largest decline in chip manufacturing last month, with India and the US doubling down on local manufacturing of semiconductors. The shrink was credited to the pandemic-induced restrictions and dwindling demand. The production of integrated circuits (ICs), as reported in the South China Morning Post, sank 24.7% year-on-year to 24.7 billion units—marking it as the largest single-month decrease since 1997. China witnessed a decline for the second consecutive month in chip manufacturing—with the output nosediving from 16.6% to 27.2 billion units as of July 2022. The local output of microcomputers fell 18.6% to 317.5 billion units in August 2022—noting a contraction in the domestic manufacturing activity for the first time in three months. According to business database platform Qichacha, statistics show that around 3,470 chip-manufacturing companies “went out of business in the first eight months of the year”. A report by the India Electronics & Semiconductor Association (IESA) and Counterpoint Research shows that the Indian semiconductor component market is likely to reach $300 billion in cumulative revenues by 2026. With initiatives such as ‘Make in India’ and production-linked incentive (PLI) schemes, the country is expected to witness a boost in local sourcing of semi-components in the coming years. In order to achieve self-reliance in the field of semiconductor manufacturing, the government of Gujarat has also partnered Vedanta and Foxconn with an investment of INR 1.54 lakh crore. Under the PLI scheme, the Indian government announced an investment of INR 76,000 crore separately for the development of a semiconductor and display manufacturing electronics ecosystem. US President Joe Biden has inked into law the Chips and Science Act, which will provide nearly $52 billion in semiconductor production incentives.","excerpt":"The production of integrated circuits (ICs), as reported in the South China Morning Post, sank 24.7% year-on-year to 24.7 billion units—marking it as the largest single-month decrease since 1997.","categories":["AI News"],"tags":["China","Chip Manufacturing","foxconn","USA","Vedanta"],"author_name":"Bhuvana Kamath","publish_date":"2022-09-19T14:29:53","publication_year":"2022","word_count":278,"keywords":["Go","TPU","programming_languages:R","programming_languages:Go","Vedanta","USA","China","ViT","R","Chip Manufacturing","foxconn"],"extracted_tech_keywords":["TPU","R","Go","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/india-us-to-boost-local-manufacturing-as-china-shrinks-in-chip-output\/","complexity_score":3,"technical_depth":6,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10140439,"title":"OpenAI Buys Chat.com","content":"OpenAI has added another notable domain to its portfolio with the acquisition of Chat.com. Sam Altman posted on X a simple link of the URL “chat.com” signalling the change. https:\/\/t.co\/n494J9IuEN— Sam Altman (@sama) November 6, 2024 Chat.com now redirects visitors to ChatGPT, a move that was confirmed by an OpenAI representative via email. Originally registered in September 1996, Chat.com is one of the web’s longstanding domains. Reports last year revealed that HubSpot co-founder and CTO Dharmesh Shah purchased Chat.com for $15.5 million, marking it as one of the most expensive domain sales on record. In March, Shah announced that he had sold Chat.com to an undisclosed buyer. Late today, he confirmed on X that OpenAI was the purchaser, hinting that the transaction might have involved OpenAI shares. OpenAI has not disclosed the amount they paid for this acquisition, but people predict it’s likely more than $15 million. The race to acquiring domains has been long standing. Last year, OpenAI was the first to acquire ai.com, to make AI synonymous with ChatGPT. A few months later, ai.com was leading to Elon Musk’s xAI, highlighting its mission to understand reality. But it was not revealed if Musk actually bought the domain from OpenAI or if it was just the owner of the domain trying to sell it to both the parties. Cut to present, Chat.com and ai.com, both now redirect to OpenAI’s ChatGPT.","excerpt":"Both Chat.com and AI.com now redirect to ChatGPT.","categories":["AI News"],"tags":["Editors Picks","OpenAI"],"author_name":"Mohit Pandey","publish_date":"2024-11-07T07:28:44","publication_year":"2024","word_count":231,"keywords":["ChatGPT","OpenAI","AI","programming_languages:R","R","GPT","Editors Picks","XAI","xAI","llm_models:ChatGPT","llm_models:GPT"],"extracted_tech_keywords":["AI","ChatGPT","OpenAI","xAI","R","GPT","XAI","llm_models:GPT","llm_models:ChatGPT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/openai-buys-chat-com\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10003048,"title":"Latest Data Science and Analytics Job Openings To Apply Now","content":"Data Science portrays a crucial role when it comes to deriving meaningful insights from large chunks of data for decision-making in organisations. Reports suggest that as of March 2020, the analytics function in India has earned a consolidated revenue of $35.9 Bn, which is a 19.5% growth in revenue over last year. In this article, we list down the current job openings for Data Scientists and Analysts one must apply right now. (The list is in no particular order). 1| Data Scientist at MiQ Digital Location: Bangalore About: As a Data Scientist, the responsibilities include translating product requirements into analytical requirements\/specifications, designing and developing required functionality in the internal tool, leveraging technology to automate daily operational activities, own and execute end to end data science projects. It would also require a candidate to collaborate with engineering teams to activate media campaigns based on analytical outputs and ensure actionable insights are delivered by utilising all the rich data in conjunction with client-specific data. Apply here. 2| Developers- Data Analysis & Integration at L&T Infotech Location: Anywhere in India About: For this role, a candidate must have experience on PL\/SQL, Oracle, BigQuery (GCP), Data Analysis\/Science, integration APIs, Python with relevant CI\/CD, automation, monitoring tooling etc. It also requires experience in Oracle SQL, PL\/SQL coding, debugging the code, code tracing, fixing the issues, among others. The candidate must also have a quick understanding of the functionality and analytical concepts of JavaScript. Apply here. 3| Data Science with Python at L&T Infotech Location: Anywhere in India About: As a Data Scientist, a candidate will be expected to deliver end-to-end analytical solutions covering multiple technologies and tools to multiple business problems. The candidate must have the ability to coordinate multiple teams with analytical or non–analytical objectives, and showcase analytical thought leadership when required. He\/she should have excellent presentation skills at client CXO’s levels to connect analytical strategies in tune with their mission and vision objectives. It also requires the candidate to have experience in statistical modelling, machine learning, data mining, unstructured data analytics in corporate or academic research environments. Apply here. 4| Data Scientist at Endless Gain Location: Bangalore About: It will require the candidate to extract, analyse and interpret large amounts of data from a range of sources, using algorithmic, data mining, artificial intelligence, machine learning and statistical tools. The candidate will need to build algorithms and design experiments to merge, manage, interrogate, and extract data to supply tailored reports, test data mining models to select the most appropriate ones for use on a project, conducting ad-hoc analysis and presenting results in a clear manner. The job description also includes enhancing the data collection procedures to include information that is relevant for building analytic systems, and other more. Apply here. 5| Data Analyst at United Airlines Location: Gurgaon About: As a Data Analyst, the candidate will be responsible for executing solutions to business problems using data analysis, data mining, optimisation tools, statistical modelling and machine learning techniques. It would require the candidate to develop and validate evaluation metrics and models to analyse business performance and areas of opportunity, identify the practical approach to test and learn key hypotheses to extend analytic insight, leverage new data and\/or experimental results to extend analytical frameworks\/algorithms, design experiments and analyse results to determine the effectiveness of programs, policies and strategies, among others. Apply here. 6| Data Analyst at Michelin Location: Pune About: As a Data Analyst, the candidate will be responsible for working directly with a dynamic team of e-retail business analysts based in Lyon, France, as they assist internal business partners in developing e-Retail strategies to achieve Michelin business goals. The candidate will need to compile a comprehensive narrative of how online distributors execute Michelin tyres. They will need to provide operational insights and recommendations, forecast and prepare reports measuring the effectiveness of programs\/action plans, design and build reports using multiple tools such as dedicated Web analytics tools, Excel, PowerPoint, Power BI. They would also need to organise the supports to be communicated and displayed in the organisation through the appropriate processes to the stakeholders. Apply here.","excerpt":"Data Science portrays a crucial role when it comes to deriving meaningful insights from large chunks of data for decision-making in organisations. Reports suggest that as of March 2020, the analytics function in India has earned a consolidated revenue of $35.9 Bn, which is a 19.5% growth in revenue over last year. In this article, […]","categories":["AI Hirings"],"tags":["beginner python projects","big data developer skills","Data analyst jobs","Data Analytics","Data Mining","Data Science Jobs","Data Scientist","data structure using java","latest technology in machine learning"],"author_name":"Ambika Choudhury","publish_date":"2020-07-22T18:00:00","publication_year":"2020","word_count":680,"keywords":["GCP","big data developer skills","TPU","Ray","data science","artificial intelligence","Data Mining","RAG","latest technology in machine learning","analytics","machine learning","AI","data structure using java","Data analyst jobs","Data Science Jobs","Python","Data Analytics","Data Scientist","beginner python projects"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","data science","analytics","Ray","RAG","GCP","TPU","Python"],"url":"https:\/\/analyticsindiamag.com\/ai-hiring\/latest-data-science-and-analytics-job-openings-to-apply-now\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10133790,"title":"CII&#8217;s Innovation Summit, Innoverge 2024, will have AI as the Overarching Theme","content":"Confederation of Indian Industry (CII) has announced the 20th edition of its flagship India Innovation Summit- ‘Innoverge 2024′ with the theme, “India@100: Building for a Resilient, Sustainable, and Inclusive India”. The summit will explore building an innovation ecosystem crucial for a developed India, by 2047. Innoverge 2024 will aim to bring significant change by improving business practices, developing new strategies, and using advanced technologies like Web 3.0 and quantum information systems. The event will focus on key sectors such as healthcare, energy, data science, clean technology, space, manufacturing, and more. AI Persists The Summit has been led and mentored since inception by Kris Gopalakrishnan, mentor CII India Innoverge 2024, past president CII, chairman, Axilor Ventures & co-founder Infosys Ltd. “AI and ML, is an overarching theme across all industries, and, you know, everything, including media, entertainment, and our day-to-day life, AI and ML is going to play an important role. We all need to leverage this technology for our personal use as well as professional use,” said Gopalakrishnan, at the CII Innoverge 2024 Press Meet. Gopalakrishnan told AIM – “There are a large number of initiatives happening in AI, both in startups, as well as in larger enterprises, and academic institutions.” He also spoke about BharatGPT and AI4 Bharat, highlighting their efforts to develop language models tailored to Indian languages, which are crucial for understanding Indian languages, natural language processing, and voice input and output systems. He also discussed using voice as a means of communication with machines. The 2024 edition of Innoverge aims to support economic growth while maintaining a focus on social equity, environmental sustainability, and national resilience. The three-day summit will gather key elements of an innovation ecosystem, including universities, large corporations, startups, policymakers, and collaborative groups. The summit will highlight the important role of institutions and accelerators from emerging regions in promoting inclusive and scalable innovation. With the Government of Karnataka as the Partner State for 2024, the summit will be addressed by key stakeholders from the Department of Electronics, Information Technology, Biotechnology and Science & Technology, as well as from the Large & Medium Scale Industries & Infrastructure Development, Government of Karnataka. Innoverge 2024 will be chaired by Kamal Bali, immediate past chairman, CII SR & president & MD Volvo Group India and other leaders from the space. The summit is scheduled from 4th-6th September at Hotel Shangri-La in Bengaluru.","excerpt":"“There are a large number of initiatives happening in AI, both in startups, as well as in larger enterprises, and academic institutions,” said Kris Gopalakrishnan to AIM.","categories":["AI News"],"tags":["AI innovation","Bengaluru","CII"],"author_name":"Vandana Nair","publish_date":"2024-08-26T19:32:07","publication_year":"2024","word_count":395,"keywords":["data science","Go","TPU","AI innovation","CII","AI","ML","Scala","RAG","GPT","Aim","Bengaluru","R"],"extracted_tech_keywords":["AI","ML","data science","Aim","RAG","TPU","R","Go","Scala","GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/ciis-innovation-summit-innoverge-2024-will-have-ai-as-the-overarching-theme\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":43676,"title":"‘AI &#038; Digitisation Help Us Deliver Cutting-Edge Care To Mothers,&#8217; Says Cloudnine CEO Raviganesh V","content":"Raviganesh Venkataraman, Chief Executive Officer, Cloudnine Group of Hospitals Over the last five years, the adoption of artificial intelligence in the healthcare industry has grown significantly. It has been widely reported that by 2021 the market is expected to touch ₹431.97 billion. From improving patient outcomes, levelling out the doctor-patient ratio and driving a wave of personalisation, AI has disrupted several segments under healthcare – hospitals, diagnostics, pharmaceuticals, medical insurance and telemedicine. Given the spate of recent developments, hospitals are also undergoing a transformational wave and upping the ante by digitising operations and offering AI-enabled healthcare solutions.  This week we spoke to Raviganesh Venkataraman, CEO, Cloudnine Group of Hospitals to understand how he’s building a partnership ecosystem to drive the Uberisation of Healthcare through and collaborating with innovators to provide AI-enabled solutions. on top of his agenda is also positioning Cloudnine as an employer of choice for today’s tech-savvy professionals. Analytics India Magazine: Tell us about your role at Cloudnine and what are the key areas you have been addressing since you took over? Raviganesh Venkataraman: My journey in the healthcare sector started with Cloudnine Group of Hospitals in April this year. As the first mover and pioneer in this sector, my efforts along with the other founding teams have been focussed on the ability to innovate constantly. Innovations across customer experience, clinician engagement, staff training, technology intervention, through both medical and digital means to further enhance outcomes have always helped us set a high benchmark of excellence. One of our key distinctions has also been the fact that we are extremely transparent in our service offering and that has helped us cut through the otherwise hard to traverse industry mechanics on pricing. We have provided over 70000 mothers a truly Cloudnine experience as they gave birth to their bundle of joy! There is frankly no greater accomplishment than seeing the smiles on faces and joy in the eyes of these families. AIM: What is your overall strategic and technological direction at the hospital chain? RV: We live in a digitalised world and hospitals are no different. Hence we are in the process of digitising every aspect of customer interaction. Today’s customer is exposed to digital solutions in virtually every aspect of her life and she has come to expect that as the new normal. There is an expectation of obtaining the same “swipe on the smart-phone convenience” even in her interaction with doctors. While we have provided digital solutions for many doctors right from the inception, we find that a lot of doctors actually don’t find enough time to use these tools or are perhaps not too comfortable typing information using a keyboard. To overcome this challenge, one of the first things we have done (and we’re perhaps the first chain in the industry to do so), is to encourage the use of digital pen, for the doctors who are not comfortable typing. With this intervention, 100% of our prescriptions will get digitised – all items just need a tick with a digital pen after which the software is automatically updated. As a next step, we will be deploying AI tools that analyse all this information and bots that read and transcribe a doctor’s handwriting with ease. This to me is the fundamental piece, where all our hospitals and doctors are going the paperless route. Every moment of transaction between the doctor and customer is being captured simply, neatly and precisely using digital means. Further, all this is available at a click of a button or a swipe of the finger to the customer on her smartphone. What we are trying to do is to put the power in the hands of the customer and this is a big shift for the healthcare industry. Digitisation here is crucial to making the transformation of healthcare to a customer-centric industry. One of the biggest focus for us has been our app – the “It’s Our Baby” app. The app while it empowers the customer,  helps us in generating lot of customer insights that open up new possibilities of business. We’ve always been looking at an ecosystem which comes together for the customer. And in this regard, we have never been shy of looking at new-age companies, new-age startups to get incorporated. AIM: How is analytics being used to optimise and improve hospital operations? RV: Analytics plays a very critical role in hospital administration as it helps to adapt and improve clinical outcomes, operational efficiency of the businesses, healthcare financing, etc immediately.  Most importantly it helps in spotting patient condition changes in intensive care units, helps in offering personalisation of the experience. Our in-depth approach on our customers and doctors, and towards analysing requirements before initiating projects, we have also focused heavily on building an inhouse team to service our technology requirements. This means that initiatives such as the “It’s Our Baby” app and Cloudnine DigiDoc have been developed with an internal team of developers, that has given us immense control on the speed of execution as well as the ability to evolve the product as we go – following a typical iterative approach for achieving “Product Market Fit”. We have split our systems into “Systems of Engagement”, i.e., the systems that interface with customers\/ doctors and “Systems of Records”, i.e, back-end systems and databases. As a strategy, we focus on building our own Systems of Engagement while working with robust Systems of Records. AIM: As Cloudnine plans to expand footprint in Bangalore,  what kind of technology can augment operations? RV: Not just Bengaluru, we aim to expand our technological prowess in all the markets and centres. There is a tremendous opportunity for us to establish our services in every major city in this country, and in most cities, we can set up more than one unit and that gives us the distinct advantage of being able to grow our volumes significantly for many years to come. We are further developing collaborative models of capital deployment and also a CAPEX light model which would further open up a lot more cities of choice for expansion, to provide us rapid scale. AIM: Will we see a wave of personalisation in-patient care just like we see in other consumer-facing industries like retail, finance and e-commerce? RV: Our focus at Cloudnine is about building deep customer relationships at every stage of their journey with us. Each patient is unique seeking individualised and personalised care. I would like to share here an example of a survey done by our in house teams where we spoke to many customers and doctors and conducted various surveys. It was clear from this effort that the activities that the customer performed with Cloudnine during their pregnancy journey could be broken into 3 buckets. Core Activities which are primarily their interactions with the doctor, for which they come to the hospital Enhancing Activities that help in enhancing their interaction with the doctor, e.g., their lab reports, scan analysis that helps in better consultations Enabling Activities that mainly support the above e.g., billing, etc. We realised that to provide convenience, ‘Enhancing’ and ‘Enabling’ activities, to the extent possible, must move into the customer’s phone.  Hence we developed and embarked on the “It’s Our Baby” app journey. We are committed to what I call the Uberisation of Healthcare through our Tech interventions. We are looking at all our customers coming together as a large ecosystem. We are bringing in a set of partners who will support this ecosystem with their products. So the mother and child, which is the core of this ecosystem, is supported by not only Cloudnine but all of Cloudnine cohorts in having a great journey, superb celebration for this absolute grand moment in their life, which is the birth of their baby. We have therefore intensified our digital efforts – both for our customers as well as our doctors. We see ourselves as pioneers in using digital technology in healthcare and one can expect many innovations in this space from us in the years to come. AIM: Can you share details about your IT team and what areas do they function on? Are you partnering with any third party vendors to store data in the cloud etc? RV: We’re amongst the first organisations to have a full-fledged 30-member team led by a Chief Digital Officer, who looks at all elements of digitisation in the entire business. This was something Cloudnine drove right from inception.  This team also includes in-house developers who have driven our core technology efforts. We use AWS and other partners wherever needed to save our data on the cloud. Technology has been to use build robust systems of records through partners while systems of engagements are developed in-house for greater control, swift action. AIM: Is there a HIMS in place? Going forward are you planning to float a chatbot? RV:  Yes. We have a robust HIMS which acts as the backend for our customer app efforts. Recently, new tools have been designed to simplify the interaction between customers and chatbots. Our Chatbot or visual assistants help in booking appointments, retrieving lab reports etc and providing recommendations to the customers from time to time for all their common queries and questions related to their pregnancy journey. AIM: Your thoughts on how future tech like AI can transform healthcare? RV: In my view, AI is transforming the healthcare landscape in India as it increases the ability for healthcare professionals to better understand the day-to-day patterns and needs of the patients they care for, and with that understanding, they are able to provide better feedback, guidance and support for staying healthy. It helps in the personalisation of medicine and the hospital experience for every patient. In other words, technology applications and apps encourage healthier behaviour in individuals and help with the proactive management of a healthy lifestyle. It puts consumers in control of health and well-being. AIM: Is there any analytics team in place and what areas do they function on? RV: Yes. We have analytics ingrained into our core functional teams. The key focus areas are – marketing analytics (focused on acquisition, retention), app analytics (focused on improving engagement and ARPP from the customer) and some aspect of financial analytics are also covered in the current scheme of things. AIM: Going forward what’s the agenda for 2020 and how tech can help consolidate the hospital chain’s position in the market. RV: The need of the hour is to build a large talent pool showcasing the career growth opportunity in this sector. We want to be positioned as an employer of choice for the talent available in the country. At the same time, we also want to be preferred service provider for our customers. We will continue encouraging more start-ups in MedTech as well as digital engagement to develop and provide for specific areas of pregnancy care and new-born care as our country offers a tremendous scale of the opportunity. We will continue to invest and innovate on low CAPEX healthcare model + insights from customers to provide a more personalised experience.","excerpt":"Over the last five years, the adoption of artificial intelligence in the healthcare industry has grown significantly. It has been widely reported that by 2021 the market is expected to touch ₹431.97 billion. From improving patient outcomes, levelling out the doctor-patient ratio and driving a wave of personalisation, AI has disrupted several segments under healthcare […]","categories":["AI Features"],"tags":["AI Healthcare","Interviews and Discussions"],"author_name":"Richa Bhatia","publish_date":"2019-08-01T15:58:27","publication_year":"2019","word_count":1843,"keywords":["Go","artificial intelligence","AWS","AI","AI Healthcare","chatbots","Git","RAG","Aim","analytics","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","artificial intelligence","analytics","Aim","RAG","chatbots","AWS","R","Go","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/ai-digitisation-helps-us-deliver-cutting-edge-care-to-mothers-says-cloudnine-ceo-raviganesh-v\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10042022,"title":"COO Manu Lavanya On Max Life Insurance’s Digital Journey","content":"An alumnus of Indian School of Business and IIT Kanpur, Manu Lavanya is the Director and Chief Operations Officer at Max Life Insurance. He comes with almost 25 years of experience in creating, transforming and scaling businesses across industries. Before Max Life Insurance, he was the Global Chief Operating Officer at Incedo. He has also served as the Managing Director of Silicon Valley-based digital services and consulting firm Brillio; Head Markets (IT Infrastructure Services) at Cognizant Technologies Solutions; and worked with ITC Limited. At Max Life Insurance, Manu drives the operations value chain, leads the digital transformation agenda and strengthens the quality culture, helping the company deliver seamless customer experiences. He is responsible for providing leadership to operations, underwriting and claims, digital transformation and information technology, quality, innovation and service excellence functions at the insurance company. In an exclusive interview with Analytics India Magazine, Manu traced the tech journey of Max Life Insurance and explained the operation of Max Life’s ‘AI Works’ team. Excerpts: AIM: Take us through Max Life Insurance’s digital journey. Manu Lavanya: The digital journey of Max Life Insurance happened in various phases. Right from inception till around 2010, the focus was on putting on the enterprise foundations for digital. The production systems, policy administration systems and CRM systems were on-boarded till about 2010. Until 2015-16, it was a journey of building a system of experiences: the front-end systems, expanding the CRM, and driving customer interactions. That journey accelerated significantly after 2016-17. Today, we are the largest private online seller of protection business. Since 2016-17, Max Life Insurance has focused on building world-class customer onboarding journeys for our B2C business. From 2018 onwards, the focus started shifting towards building the underlying data infrastructure, cloud, scalable and responsive adaptive infrastructure that genuinely enabled us to launch the next level of differentiation to embed digital into our business processes. Over the last 18 to 24 months, Max Life Insurance has invested over a three-year transformation roadmap. The firm has put together to drive cloud, data infrastructure, artificial intelligence, machine learning, automation, and DevOps, to leverage the foundation to become a playground for implementing all the tools in analytics. The company is working on five pillars: Intelligent prospective Automation of Customer OnboardingCustomer Servicing Sales management and sales enablement Making the enterprise more digital With digital transformation, Max Life Insurance has cut down onboarding time from 22-hours to 10-minutes. The goal is to do it in six minutes. We had less than 10 percent of our portfolio that used to get issued in 30-minutes. Now that number is up to about 45 percent. The number of workloads we self-service has increased almost 400 percent over the last three years. Nearly 76-77 percent of all our business transactions are self-serviced today. The goal is to take it to 90 percent in the next two years. AIM: How did pandemic impact Max Life Insurance? Manu Lavanya: Our total NOPs (Number of Policies) sold on B2C platforms, that is, the online business itself, increased from 12 to 16 percent this year. Self-service numbers have increased from 43-45 percent to about 73 to 74 percent in the last two years. Post pandemic, business onboarding has become 100 percent digital. This means all policies go through our digital onboarding platform, mPRO. In the first four months of the COVID-19 led lockdown, 100 percent of our business was acquired by digital interactions. Having said that, insurance selling is a trust-based model. In India, primarily, a large amount of trust is vested upon the seller. Therefore, I do not foresee that the process of buying insurance ever going completely digital. It will always be a phygital model (a part of the process is executed on a digital platform with some human intervention). AIM: Give us a sneak peek into Max Life Insurance’s AI Works team Manu Lavanya: The AI Works team was formed in 2011 and consists of data engineers, data scientists, ETL Engineers, and data modellers. The team does a wide variety of analytics. We have integrated analytics across three pillars: New business onboardingPurchase and issuance of insurance Servicing and retention of customers Across these three pillars, we use machine learning, deep learning, Speech AI, Vision AI, Computer Vision, OCR, Conversational AI, NLP and BOTs. AIM: Tell us about the intelligent chatbot MILI and how it has made processes more efficient at Max Life Insurance? Manu Lavanya: It is the brainchild of the AI Works team. It automates incoming query management. When we started the journey, the intent prediction with MILI was as low as 10 percent. So 10 percent of the time, the bot would be able to narrow down on the problem that it was required to solve. Today, the bot solves almost 90 percent of our servicing transactions. If there is a specific nuance or a twist, the bot connects the user to an agent. The intent prediction has doubled from 15 percent to 30 percent over the last year for email bots. AIM: What technologies power Max Life Insurance? Manu Lavanya: We have made significant investments in the cloud journey leveraging Amazon Web Services and Microsoft Azure. We use the following technologies: MEAN stackReact JS  Pyspark TensorFlow Dialogflow Docker AIM: What’s next for Max Life Insurance? Manu Lavanya: We have just completed one year of our three-year digital transformation programme. Our biggest priorities are: Building a scalable data infrastructure so that the company can become as intelligent and pervasive as possibleRevamping some of our core enterprise systems — especially our HRI and investment management systemsMake Max Life Insurance an intelligent enterprise — revamp the experiences for all our stakeholders Transformation of digital culture in the organisation","excerpt":"An alumnus of Indian School of Business and IIT Kanpur, Manu Lavanya is the Director and Chief Operations Officer at Max Life Insurance. He comes with almost 25 years of experience in creating, transforming and scaling businesses across industries.  Before Max Life Insurance, he was the Global Chief Operating Officer at Incedo. He has also […]","categories":["AI Features"],"tags":["Gen AI in Insurance","insurance technology","Interviews and Discussions"],"author_name":"Debolina Biswas","publish_date":"2021-06-20T11:00:00","publication_year":"2021","word_count":943,"keywords":["artificial intelligence","machine learning","AI","ML","computer vision","NLP","Aim","deep learning","Gen AI in Insurance","insurance technology","analytics","TensorFlow","Interviews and Discussions"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","NLP","computer vision","analytics","Aim","TensorFlow"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/coo-manu-lavanya-on-max-life-insurances-digital-journey\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":68125,"title":"Foreign Exchange Rate Prediction using LSTM Recurrent Neural Network","content":"Data Science with a variety of powerful algorithms has a large scope of application in financial analytics. Many financial analytics problems are based on the time-series analysis where a machine learning model is required to predict the values on a time-series pattern. Deep learning models have been applied to a variety of difficult predictive analytics problems. Time series prediction is one of those difficult applications. LSTM Recurrent Neural Networks have proven their capability to outperform in the time series prediction problems. When it comes to learn from the previous patterns and predict the next pattern in the sequence, LSTM models are best in this task. In this article, we will implement the LSTM Recurrent Neural Network to predict the foreign exchange rate. The LSTM model will be trained to learn the series of previous observations and predict the next observation in the sequence. We will apply this model in predicting the foreign exchange rate of India. The Data Set The data set in the experiment is taken from Kaggle that is publicly available as Foreign Exchange Rates 2000-2019. This dataset was originally generated on the Federal Reserve’s Download Data Program. It comprises the exchange rates of 22 countries including the Euro Area for 5217 days. Implementation This code was implemented in Google Colab and the .py file was downloaded. # -*- coding: utf-8 -*- \"\"\"Copy of foreign_exchange.ipynb Automatically generated by Colaboratory. Original file is located at https:\/\/colab.research.google.com\/drive\/1ItQXXc-Z33hr6W60E6gxxxxxxxxxxxxx \"\"\" First of all, we will import the Python libraries to be required in this program. #Importing Libraries import pandas as pd import numpy as np import matplotlib.pyplot as plt # %matplotlib inline import tensorflow as tf import keras from keras.models import Sequential from keras.layers import Dense, LSTM We will read the data set in the program. In Google Colab, the Foreign_Exchange_Rates.csv file was uploaded to the session storage. #Reading dataset data_set = pd.read_csv('Foreign_Exchange_Rates.csv', na_values='ND') Using the next lines of codes, we will see the shape of the data set and the head of the dataset. #Dataste Shape data_set.shape #Dataset head data_set.head() We are interested in the foreign exchange rate in India. So, before proceeding further on this data, let us visualize it. Please note that we are not going to change indexing that is required to visualize the dates on the X-axis. #Plotting Indian Exchange rate plt.plot(data_set['INDIA - INDIAN RUPEE\/US$']) We will specify the data frame of our interest that consists of the foreign exchange rate of India. #Data frame df = data_set['INDIA - INDIAN RUPEE\/US$'] df In the next step, we will preprocess the training and test data for the LSTM network. The model will be trained on the sequence of the previous value and the current value. #Preprocessing data set df = np.array(df).reshape(-1,1) from sklearn.preprocessing import MinMaxScaler scaler = MinMaxScaler() df = scaler.fit_transform(df) print(df) #Training and test sets train = df[:4800] test = df[4800:] print(train.shape) print(test.shape) def get_data(data, look_back): data_x, data_y = [],[] for i in range(len(data)-look_back-1): data_x.append(data[i:(i+look_back),0]) data_y.append(data[i+look_back,0]) return np.array(data_x) , np.array(data_y) look_back = 1 x_train , y_train = get_data(train, look_back) print(x_train.shape) print(y_train.shape) x_test , y_test = get_data(test,look_back) print(x_test.shape) print(y_test.shape) #Processing train and test sets for LSTM model x_train = x_train.reshape(x_train.shape[0],x_train.shape[1], 1) x_test = x_test.reshape(x_test.shape[0],x_test.shape[1], 1) print(x_train.shape) print(x_test.shape) Using the below lines of codes, we will define the LSTM recurrent neural network and train it on the data prepared above. #Defining the LSTM model n_features=x_train.shape[1] model=Sequential() model.add(LSTM(100,activation='relu',input_shape=(1,1))) model.add(Dense(n_features)) #Model summary model.summary() #Compiling model.compile(optimizer='adam', loss = 'mse') #Training model.fit(x_train,y_train, epochs = 5, batch_size=1) Foreign Exchange Rate Prediction Once the LSTM model is trained, we will apply it to predict the future sequences for the test data. #Prediction using the trained model scaler.scale_ y_pred = model.predict(x_test) y_pred = scaler.inverse_transform(y_pred) print(y_pred[:10]) #Processing test shape y_test = np.array(y_test).reshape(-1,1) y_test = scaler.inverse_transform(y_test) print(y_test[:10]) #Visualizing the results plt.figure(figsize=(10,5)) plt.title('Foreign Exchange Rate of India') plt.plot(y_test , label = 'Actual', color = 'g') plt.plot(y_pred , label = 'Predicted', color = 'r') plt.legend() #Mean Squared Error from sklearn.metrics import mean_squared_error mean_squared_error(y_test, y_pred) As we can see in the above plot the LSTM model has predicted almost the same sequence as it was expected. The Mean Squared Error obtained is also very less as we can see above. So we can conclude that the model has given accurate predictions about the foreign exchange rate. The same we will see in the below illustration for 10 values. #10 original and predicted rates together","excerpt":"In this article, we will implement the LSTM Recurrent Neural Network to predict the foreign exchange rate. The LSTM model will be trained to learn the series of previous observations and predict the next observation in the sequence. We will apply this model in predicting the foreign exchange rate of India.","categories":["Deep Tech"],"tags":["Deep Learning","financial analytics","lstm","Machine Learning","Predictive AI","predictive analytics","python visualize neural network","time series prediction"],"author_name":"Dr. Vaibhav Kumar","publish_date":"2020-06-25T15:00:00","publication_year":"2020","word_count":723,"keywords":["data science","machine learning","Keras","AI","neural network","lstm","Machine Learning","Colab","Ray","financial analytics","python visualize neural network","time series prediction","deep learning","analytics","Deep Learning","TensorFlow","predictive analytics","Predictive AI"],"extracted_tech_keywords":["AI","machine learning","deep learning","neural network","data science","analytics","Ray","TensorFlow","Keras","Colab"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/foreign-exchange-rate-prediction-using-lstm-recurrent-neural-network\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10123685,"title":"The 10 Best Videos Created by Luma AI","content":"Close on the heels of Sora & Kling, comes a new contender – Dream Machine. California-based startup Luma AI, which focuses on visual AI, has unveiled this new video generator that stands out due to its use of AI to create realistic visual content. One of the key differentiators is the photorealistic quality of its videos. The AI algorithms employed by Luma meticulously analyse and enhance every detail, from texture to lighting, ensuring that the final output looks almost indistinguishable from real-world footage. A prime contributor to Luma’s success is AWS. Amazon’s cloud computing subsidiary has provided Luma AI with the infrastructure, exposure and practical applications, showcasing its capabilities in streamlining production processes. “Great to see how AWS H100 training infrastructure helped the Luma AI team reduce time to train foundation models and support the launch of Dream Machine,” said Swami Sivasubramanian, vice president for data and machine learning services, AWS. Co-founded in 2021 by CEO Amit Jain, Luma AI is currently based in San Francisco, California. AIM decided to try out Dream Machine to produce a video. Here’s a look at it. Meanwhile, we have also compiled a list of the top 10 mind-blowing videos produced by Dream Machine. A Woman This AI-generated video features a woman with a shaved head wearing a blue outfit. She appears to have a serious expression, and the background includes a building with multiple windows, suggesting an urban setting. By allowing everyone to experiment with AI-powered video generation for free on its website, Luma AI has hit a major milestone in the field. The abandoned Building The video depicts a long, narrow hallway with dim lighting, likely located in an abandoned or poorly maintained building. The corridor has graffiti writing, peeling paint, and debris scattered on the floor. The ambience is eerie and desolate. This highlights the advanced visual capabilities of AI in capturing and rendering detailed environments. Girl with a Pearl Earring Here, the video brings the painting, ‘Girl with a Pearl Earring’, the timeless beauty of Johannes Vermeer’s masterpiece to life using AI. As the painting is transformed into a realistic video with every brushstroke and delicate detail, it captures the subtle play of light and shadow, the intricate textures, and the serene expression of the girl. This visual experience shows the original artwork while offering a fresh, modern perspective through the unsettling potential of AI. Kabosu! With Dream Machine, this video brings Kabosu to life. Every detail, from the eyes to the fluffy coat, is rendered with creativity and high-quality visuals, demonstrating the advanced capabilities of the model. The body reconstruction, backed by the model’s new technology, allows users to create videos in various aspect ratios. Overall, it showcases its potential in generating high-quality, life-like video content, making it a standout in the field of digital animation. Mark Zuckerberg The Mark Zuckerberg video made by Dream Machine showcases an innovative application of artificial intelligence and technology. In this video, it appears as though Zuckerberg is in the middle of the woods, looking outside through a glass window. This almost-realistic clip can be viewed from multiple angles. It also captures and renders his movements and expressions, bringing a new level of realism to virtual representations. The potential of AI in creating life-like digital avatars paves the way for future advancements in virtual communication and entertainment. Willy Wonka Walks Off In this video, Willy Wonka is digitally recreated where he walks away, expressing disappointment. The character’s facial expressions, gestures, and mannerisms align perfectly which offers a glimpse into the future of digital media and storytelling possibilities. The precise features and seamless editing add to future creativity in AI. Disaster Girl Meets Firefighters This AI-generated video contains several realistic elements, such as a young girl smiling, firefighters attempting to extinguish a fire, and two officers having a conversation at the end. This serves as an example of AI’s capacity to bridge digital content with real-world impact. A Girl & a Zeal of Zebras This video featuring a girl and zebras in the forest goes beyond mere visuals; it intricately weaves together elements of nature, human curiosity, and storytelling. Set against the backdrop of lush greenery it shows the girl’s encounter with the zebras showing seamless integration of AI technology in entertainment. Through advanced algorithms, the characters exhibit life-like movements and expressions, enhancing the immersive experience. The Eye The video focusing on the eye exemplifies an exploration of visual perception through advanced AI techniques. This captivating clip delves into the intricacies of the human eye, capturing its mesmerising colours. The AI algorithms show the light refraction, intense colour, and slow zoom, creating a highly realistic and captivating scene. The Masked People This clip features a captivating scene in which a group of masked individuals are situated within a vibrant environment painted in striking hues of bright blue and pink. The contrasting colours of the room amplify the presence of the masked figures, creating an intriguing visual that captivates viewers. The characters’ movements within their space are rendered with detail. The AI ensures that each gesture and reaction is natural, enhancing the viewer’s engagement and the characters’ believability.","excerpt":"The brand-new text-to-video model from Luma AI stands out due to its photorealistic AI visual content.","categories":["AI Trends"],"tags":["AI","AI Video Generation Models","kling","Sora"],"author_name":"Tarunya S","publish_date":"2024-06-14T16:55:27","publication_year":"2024","word_count":853,"keywords":["artificial intelligence","machine learning","AWS","AI","cloud computing","TPU","ML","kling","AI Video Generation Models","Aim","Sora","foundation models","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","foundation models","Aim","cloud computing","AWS","TPU","R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-10-next-gen-videos-created-by-dream-machine-a-sora-and-kling-alternative\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10066118,"title":"How to build transfer learning models with PyTorch using PyTorchCV?","content":"Computer vision models require a lot of effort in deciding the architecture and fitting them with the large-sized data. To reduce such efforts, transfer learning can be utilized which is all about using a pre-trained model to solve different problems. PyTorchCV is a framework that provides us with a lot of pre-trained computer vision models that are considered as high-performing solutions than the existing ones. In this article, we are going to discuss how to build transfer learning models in PyTorch with PyTorchCV. The major points to be discussed in the article are listed below. Table of content What is transfer learning?Transfer learning in computer visionWhat is PyTorchCV?         Transfer learning with PyTorchCV Let’s start with knowing what transfer learning is. What is transfer learning? Transfer learning is a type of machine learning program in which we utilize a trained model to perform tasks using different information. For example, a random forest model trained in an IRIS data set can also work with digits data. As we know machine learning algorithms work based on the historical information and the outcome we require from them is a sort of prediction of any class or prediction of the future. We design these models so that they can perform an isolated task. We can also say that in a normal machine learning program we transfer source knowledge to target knowledge. In transfer learning, we also perform normal machine learning tasks but once a model is prepared for data we use its knowledge gained with different information. Simply said, transfer learning is a process of using a trained model with information on which the model is not trained. Using this method we can save the cost of building and training a model. In recent years, we can witness a variety of models that can be utilized in transfer learning settings and we get these models from various frameworks. One of the great examples of such a framework is hugging face where we get models for every field of data science and artificial intelligence, instead of this framework we also get models from big companies like Microsoft, Google, and Facebook. In this article, our focus is on a framework that provides models for computer vision tasks in a transfer learning setting. Before looking at the framework we need to know about the models that can give the state of the art performance for computer vision tasks. Are you looking for a complete repository of Python libraries used in data science, check out here. Transfer learning for computer vision Computer vision is an important part of data science and artificial intelligence as well as it is difficult also because it mainly deals with images and video data. Such data is very difficult to extract information from it and make a machine learn patterns from extracted information. Instead of building such a difficult model, we can prefer to use a pre-trained model for better performance. Various models are available to use in transfer learning settings in every section of computer vision. Some of them are as follows: Image classification VGG(virtual geometry group): it is a deep convolutional neural network.ResNet(residual network): this model consists of several stages of convolutional and identity blocks.DenseNet: this model has densely connected convolutional blocks. Semantic segmentation DeepLabV3: DeepLab is a state-of-the-art semantic segmentation model developed by the researchers of Google.                   PSPNet(pyramid scene parsing network): This model utilizes the pyramid parsing module for completing the semantic segmentation.DenseASPP(Densely connected Atrous Spatial Pyramid Pooling): this model connects several atrous convolution blocks in a dense system. Object detection SSD(Single Shot MultiBox Detector): this model is developed for object detection using a single deep learning neural network.Faster R-CNN: This model is used for real-time object detection and consists of a network that can detect objects in a region of the image.FPN(feature pyramid networks): This model utilizes the pyramid parsing module for completing object detection in images. Pose estimation CPM(convolutional pose machine)OpenPose: This model uses Part Affinity Fields for real-time pose estimation. Here we have looked at some of the models we can use in computer vision models for computer vision problems. Let’s take a look at a framework that is only designed to perform transfer learning in the computer vision field named PyTorchCV. What is PyTorchCV? PyTorchCV is a framework that is built using the PyTorch library and consists of transfer learning models that are related to only computer vision modelling. PyTorchCV provides the feature of building high-performing deep learning models that have shown better performance than the other existing frameworks. We can find the GitHub repository of this framework here. From the repository, we can utilize the source codes for various state-of-the-art computer vision models. Since this framework is built on PyTorch, a general user of PyTorch can easily understand the uses of this framework. The models this framework has in collections are trained using datasets like ImageNet-1K, CIFAR-10\/100, SVHN, CUB-200-2011, Pascal VOC2012, ADE20K, Cityscapes, and COCO. We can find all the implemented models in the framework here. We can utilize models from this framework after installing it in our environment. The installation can be done in the following way. !pip install pytorchcv Still, there is a recommendation from the developer side to use this framework with torch version >= 0.4.1. We can install both at the same time using the following lines of codes: !pip install pytorchcv torch>=0.4.0 After installation, we can use the pre-trained models that are available in the framework. Transfer learning with PyTorchCV This section includes the basic information on the implementation of a model provided by the PyTorchCV framework. For example, if we want to use a resnet-18 as a transfer learning model, we can do this in the following way: Loading model from pytorchcv.model_provider import get_model as ptcv_get_model net = ptcv_get_model(\"resnet18\", pretrained=True) net Output: Here we can see the structure of ResNet. Since the image is too large it is not posted here. Defining an image import torch from torch.autograd import Variable x = Variable(torch.randn(1, 3, 224, 224)) Fitting the model on the image y = net(x) In the above step, we will get the extracted features from the image by the instantiated ResNet-18 transfer learning model. Here we have seen how we can utilize this framework and its provided pre-trained model. One thing that I liked about this framework is that developers of this framework are focused only on implementing models related to computer vision only. Because implementation is only related to computer vision, this framework is oriented to high performance in computer vision tasks and can help in projects related to only computer vision. Since most of the transfer learning models in computer vision are built using convolutional neural networks this framework does not split its knowledge into other sections of neural networks like RNN and LSTM. This feature makes this framework lightweight and high-performing. One more thing which is good about the framework is it uses Pytorch as its base library so whenever checking the models from the source code, a good user of PyTorch can easily understand the process of extracting a pre-trained model. These all qualities can make us use this framework for our computer vision transfer learning procedures. Final words In this article, we have discussed transfer learning and transfer learning in computer vision. Along with this, we look at the list of some models that can be used in transfer learning settings for solving computer vision problems and we have looked at the usage of a framework PyTorchCV that only includes computer vision models. References Link to the codesPyTorchCV documentations","excerpt":"PyTorchCV helps in building high-performing transfer learning models that have shown better performance than the other existing frameworks.","categories":["AI Trends"],"tags":["AI (Artificial Intelligence)","Computer Vision","Data Science","Deep Learning","Machine Learning"],"author_name":"Yugesh Verma","publish_date":"2022-05-03T18:00:00","publication_year":"2022","word_count":1255,"keywords":["data science","Hugging Face","artificial intelligence","machine learning","AI","neural network","PyTorch","Machine Learning","computer vision","deep learning","object detection","Computer Vision","Deep Learning","Data Science","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","deep learning","neural network","computer vision","data science","PyTorch","Hugging Face","object detection"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/how-to-build-transfer-learning-models-with-pytorch-using-pytorchcv\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10053791,"title":"An Illustrative Guide to Deep Relational Learning","content":"Sometimes, the predictive modelling algorithms lose the useful information of the data that can affect the overall results. One of the important pieces of information which data consists of is the relation between the instances and inferring these relationships in the results can make the machine learning model a more powerful tool. The same concept applies to the deep learning models as well. In this article, we will be discussing an important concept of deep learning, called Deep Relational Learning, which is working on making the results and model more expressive using the relational information of the data. The major points to be covered in this article are listed below. Table of Contents What is Deep Relational Learning?The Architecture of the Relational ModelThe Architecture of a Statistical ModelThe Architecture of Neural NetworksGeneral Use-Cases of Deep Relational LearningStatistical\/Graphical ModelsNeural NetworksHybrid Approaches Applications of Deep Relational Learning Let us start the discussion by understanding deep reinforcement learning. What is Deep Relational Learning? As we know, the standard deep learning models work on the technique where operation of any kind is performed on the features of the data which is consumed by the model as the input before taking them to the main algorithms. The model preprocesses the data and converts them into binary or numeric form. In the real world, most of the information in the data is relational and when the data gets converted into the required form of the model it loses the important relational information in the data. Here we require an algorithm that can also work on the arbitrary form of the data so that important information can be taken into account. We require algorithms and logic which can represent the result with their useful means. There can be many approaches where we can make a model be expressive. For example, first-order logic (FOL) is a useful approach and is known for its expressive powers. The FOL can be considered for modelling purposes because of its ability to increase generality, express knowledge, and express reasoning. Many of the models take the logistic approach for making deep learning models efficiently expressive. By the above, we can say that in deep learning building such models which can handle the arbitrary information of the data and represent the result in an expressive context of their mean can be considered as the deep relational learning process and we can call the models deep relational models. The Architecture of the Relational Models When it comes to relational learning we can say we can have two types of architecture, one where the standard statistical machine learning algorithms can be used for making the relational models and we can use the neural networks for making the models. Using both of the ways we can design a relational model. The architecture of a Statistical Model The models designed for providing solutions to the above-given concern can have the ability to learn the deep architecture of the data by stacking the number of generative models and inducing logical rules in the architecture of the model. Where an inductive logic programming system helps in learning the first layer of the representation. After these layers, the architecture can hold a restricted Boltzmann machine(RBM) where hypothesized rules can be used for learning purposes of the successive layers from the stack ILP. Once the representation is obtained, the classification and regression models can be used with the features learned at the highest layer. The above image is a representation work diagram of the relational model where a logistic regression model is used with the features. Where the relation between Friend(X, Y) and kind(y) is Happy(X) is found using the logic and models in the system. Some of the examples of the statistical relational models are as follows: Markov logic networksMulti entity bayesian networkBayesian logic programRelational Markov network In this article, our major focus is to know more about relational learning in the field of deep learning but it is better to understand relational learning from the basics because the basic processes of the deep learning models will also be similar to the statistical model. In the next section, we will be discussing the architect of neural networks. The Architecture of Neural Networks When we talk about the neural networks for relational modelling we can encode similar models like SVM random forest, decision trees, and logistic regression with a layer-wise structure, or using the similar approach with the neural network we can design the models where the parts of the graphical models are designed using the neural networks. The benefit of the neural networks is that we can make a model with several relational linear layers and relational activation layers as a graph. Also, we can obtain the benefit of backpropagation for training the neural networks by adding the error layers. When feeding the data in such architectures allows the algorithm to calculate the prediction errors and using the backpropagation of derivatives we can update the parameters of the models. The above image is a representation work diagram of the relational neural networks where the model is predicting the gender of users in a movie rating system with a layer-wise architecture. General Use-Cases of Deep Relational Learning In this section of the article, we will discuss some of the works dealing with graphical, neural, and hybrid representations for relational learning. Statistical Graphical Models There are various languages that can be used for making graphical models. In some of the research, we see the suggestion of generative models, and various works have used discriminative models where basically the Markov logic chain has been used. And we have also seen that both of the approaches combine logic and graphical models that are probabilistic where we see in the algorithms each formula is in association with the weights. The relation information probability distribution is derived using the weights. We use the following languages for graphical models. BLOGCHURCHProbabilistic Soft Logic (PSL) DRaiL Neural Networks We can understand the limitations of the graphical models and for that, we can use the neural networks for making the model expressive to the relational information presented inside the data. Where various prediction tasks can be performed by training the learned nodes of the neural network. In some of the works, we see that the graph adjacency information is captured by the networks for the representation of the nodes. Distance between the two nodes is similar to the graph distance in the graphical models. In some of the works, we see that the nodes have textual properties which help in feature representation by learning the representation of graph relations. The knowledge graphs can be considered as multi-relational data. When working with such data we are required to embed both nodes and edges of the graph. We can use neural networks where the algorithms learn the representation of the nodes where the relations lying inside the data can be represented as the pair-wise node relations which also benefits us in the context of representing the broader graph context. Some of the works on the graph neural networks are as follows: Graph Neural NetsCane: Context-aware network embedding for relation modellingTri-party deep network representation Hybrid Approaches In the above approaches we have seen how the graphical models and neural networks can be used for deep relational learning and combining both of them we can obtain more fruitful results. Some of the approaches for hybridization of both methods are listed below. Lifted rules to specify compositional nets: In the above sections, we have seen how the hypothesized rules are made to learn the successive layers. Using this approach, the model can learn the relational dependencies in a data space. To specify the composition functions, the neuron of the network is used where the neuron has mapped facts and rules of the procedure by the system. Two major works on the basis of this system are listed below. LRNNsRelNNs Differentiable inference: In this type of system neural networks are used for storing the classes of logical queries and the system focuses on implementing reasoning where a series of numeric functions from the graphical approach is used. Some of the works using the approach are as follows: TLNsTensorLog Deep classifiers and probabilistic inference: As the name suggests the systems built on using this approach have classifiers from the deep neural networks and inference of the results received by the graphical modelling. Simply saying, neural networks help in learning the probabilities for expressions. Input for this type of system is a combination of the feature vectors and facts where facts are used by the logic program like FOL. Some of the works using this approach are as follows. DeepProbLogDeep Probabilistic Logic Rule Induction from data: In most of the works we find that the rules are introduced to the system by the graphical knowledge bases and then using neural networks we can make a whole system be expressive in terms of the inferences of results. NTPsNeural Logic Programming Here we have seen some of the approaches from which we can perform relational learning and deep relational learning with some of the already implemented programs in the field. Below I am using a table to compare all these main works in the field to deep relational learning. Applications of Deep Relational Learning As we have seen, the major goal of developing such a system is to make the results more expressive in terms of the relational information presented in the data. When we talk about the data we can say most of the time there are huge relations presented in the instances of the NLP data and time-series data. So we can say the majority of the task of such learning is in those above-given fields are as follows: Collective classification – Here classification of the objects in classes is required with the relation of the object with the class.Link prediction – Providing information about the linkage between two objects of the data and the information about their relationship.Link-based clustering – As we can perform the link prediction between the objects, we can use those predictions for the clustering of the data based on the linkage strength between the objects.We can also use the deep relational models for social network modelling.Entity resolution can be performed by deep relational learning where the task is to identify the similar instances presented in two or more than two datasets. Final Words In this article, we have seen an overview of deep relational learning along with statistical relational learning. We got to know that the basic procedures of these models are the same. We have also discussed some examples of the deep relational models with their working styles and compared them using different parameters.","excerpt":"arbitrary information of the data and represent the result in an expressive context of their mean can be considered as the deep relational learning process and we can call the models deep relational models.","categories":["Deep Tech"],"tags":["AI (Artificial Intelligence)","Data Science","Data Scientist","Deep Learning","Guide","Machine Learning","Python"],"author_name":"Yugesh Verma","publish_date":"2021-11-20T10:00:00","publication_year":"2021","word_count":1781,"keywords":["knowledge graphs","Go","machine learning","ELT","AI","AI (Artificial Intelligence)","neural network","Machine Learning","Python","NLP","deep learning","RNN","Deep Learning","Data Science","Data Scientist","R","Guide"],"extracted_tech_keywords":["AI","machine learning","deep learning","neural network","NLP","knowledge graphs","R","Go","ELT","RNN"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/an-illustrative-guide-to-deep-relational-learning\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10073149,"title":"Why is Everyone Making Cars?","content":"In the last few years, smartphone and tech manufacturing companies like Apple, Oppo, Xiaomi, Samsung, Huawei, and Google have begun investing in the electrical automotive industry. Since 2000, Apple has applied for 248 car-related patents. Recently, the company hired Lamborghini’s top executive, Luigi Taraborrelli, to help them develop a design for their car. Hundreds of members from Ford, Tesla, Volvo, and Rivian have joined and left the company in recent years, including design executives from McLaren, Porsche, Aston Martin, and BMW. Oppo is now bidding into the EV car space and is working to develop its first fully electric car. The company plans to release their car by 2024. With its largest smartphone manufacturing facility in Noida, the company is now expanding into the automotive industry and expects to make an entry into the market with e-scooters and e-motorcycles. The Chinese smartphone manufacturer, Xiaomi, also announced their plans to build four electric car models. Xiaomi’s founder Lei Jun is expected to unveil the first prototype in August 2022. In a speech, Jun said that their focus is to stand out in the automotive industry by stepping in with automated, smart driving electric vehicles. In January 2022, Sony exhibited their prototype SUV, ‘Sony Vision-S’ to put the company ahead of Apple, Google, and Baidu in the car business. The company is known for its breakthrough innovations like the Walkman and the PlayStation, and now with a focus on autonomous vehicles, the integration of game and entertainment-like features could be perfect. Samsung, Huawei, LG “The time when cars are recognized as electronic gadgets would be the timing for Samsung to produce cars,” said Kwon Yong-joo, automotive and transportation design professor at Kookmin University. Following great success in almost every sector, Samsung has recently launched strides to fire up its readiness in the EV market. Huawei recently stated that they are also stepping into the automotive industry but will not build their own car. Eric Xu, the rotating chairman of the firm, said that they will focus on ICT technology to help car companies. The company, along with China Automotive Technology and Research Centre (CATARC), is integrating with the automobile industry to launch AR-HUD, smart car lights, and smart light display. Mercedes-Benz EQS EV sedan will feature LG developed in-vehicle-infotainment (IVI) system which is a Pillar to Pillar (P2P) display stretching across the dashboard. LG Electronics’ Vehicle component solutions (VS) company is focusing on becoming the leading developer of connectivity, infotainment, and ADAS solutions. Ride sharing companies In 2021, Arrival developed a prototype of a mono-volume electric car designed for Uber. Uber is committed to being a fully electric, zero-emission platform by 2040 globally, with 100 per cent rides taking place in zero-emission vehicles. On 15 August, 2021, Bhavish Aggarwal, CEO of Ola, announced the launch of their electric scooter. In 2022, the company announced that their team will be launching their electric car by 2024. Ola aims to build a high-performance, sustainable car to compete against rival companies. Why cars, why now? In 2008, Tesla Motors released ‘Roadster’, their first completely electric car. Since the pandemic, the company has managed to install charging stations across the U.S. and Europe and build Gigafactories to produce batteries and vehicles. With the United Nations claiming red-alert on climate change, there has been a rise in electric cars and other energy-efficient technologies in the market. Technology companies from all backgrounds are stepping in to make the least environment-friendly industry as sustainable as possible. Mohit Mittal, partner at Praxis Global Alliance, said that the Indian government is promoting localised battery and cell manufacturing through the PLI (production-linked incentive) scheme for advanced cell chemistry manufacturing (ACC). As a result, several new players and manufacturers are making their way into the EV space in India. In 2021, the Indian EV industry witnessed an investment of $6 billion and was projected to attract $20 billion by 2030. Smartphone manufacturing companies like Oppo, Xiaomi, and Huawei have reached a saturation point in their small and detailed inventions in phones and IoT products. With the government incentivising the production and manufacturing of environment-friendly vehicles now, these companies have decided to switch to the automotive industry. Established players of the automotive industry like Maruti, Mahindra, and TATA Motors are a step ahead of the software companies but lack the creative and technological innovations that LG, Sony, and other companies are willing to invest in. Additionally, for the consumers, a rise in the competition in EV automotive industry will result in a drop in the prices of vehicles. It will also encourage further innovation and improve the overall ecosystem.","excerpt":"Apple, Xiaomi, Samsung, Huawei, Google, Sony, and LG have started innovations in the EV automotive industry.","categories":["AI Features"],"tags":["automotive industry","climate change","electric cars","electric vehicles","EV","Google","Huawei","Ola","Ola Electric","Uber"],"author_name":"Mohit Pandey","publish_date":"2022-08-21T10:00:00","publication_year":"2022","word_count":765,"keywords":["electric cars","electric vehicles","climate change","R","RAG","ViT","automotive industry","Go","Ola","Ola Electric","AI","EV","programming_languages:R","innovation","programming_languages:Go","Aim","Google","Uber","Huawei"],"extracted_tech_keywords":["AI","Aim","RAG","R","Go","ViT","innovation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/why-is-everyone-making-cars\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10090972,"title":"8 Tech Startups on Shark Tank India S2 that Captured Our Attention","content":"Shark Tank India, Season-2 has come to a close, and what a wild ride it has been! Sure, we may not have had as many tech startups as we would like, nevertheless, the startups that did make it to the tank were impressive, and even managed to raise funding from notoriously difficult-to-impress investors. The sharks in this season were Amit Jain, Aman Gupta, Namita Thapar, Peyush Bansal, Anupam Mittal, and Vineeta Singh. Here is the list of top tech startups that raised investment from the sharks and also caught our eyes. AyuSynk Namita Thapar invested ₹50 lakhs for 3.5% equity and ₹50 lakhs debt @10% interest. AyuSynk is a digital stethoscope that replaces traditional stethoscopes by amplifying auscultation sounds. The device can be connected with the AyuShare app which allows recording the sounds and sharing them in real-time. This also enables doctors to see visual splits of S1 and S2 (Systole and Diastole) for better diagnostic accuracy. Pairing with the app also enables detection of abnormalities in the cardiac sounds of patients. The device is powered by A21 chipset, built by the company itself. Developed by an IIT-Bombay startup AyuDevices, founded by Adarsha K, Rupesh Ghyar, and Tapas Pandey, the devices have also been tested in various healthcare systems in India. insideFPV Namita Thapar, Aman Gupta, Peyush Bansal, and Amit Jain invested ₹75 lakh for 15% equity insideFPV is one of India’s first companies that provides plug-and-fly FPV drones for consumers, agriculture, and also defence. The company’s mission is to make drone purchases in India safe and convenient and open up the sky for everyone. Customers do not require a pilot licence to fly insideFPV drones. Based out of Surat, the company has been recognised by the Government of India as a robotics industry startup and has also been trusted by VIT and DLabs at ISB, Hyderabad. It was founded by Arth Chowdhary, Oshi Kumari, and Devyant Bhardwaj. Cloudworx Namita Thapar and Anupam Mittal invested ₹40 lakhs for 3.2% equity Featured in Forbes Asia 30 Under 30, Cloudworx Studio is a web-based no-code platform for building 3D twins in the metaverse. According to the team, users can build 3D twins of their companies in three simple steps – orchestration, visualisation, and enablement, without hiring any 3D experts. Coming out of Meerut, Cloudworx was founded by Yuvraj Tomar and Kaushik Bharadwaj. Currently, the team is building enterprise metaverse 3D content from real-time data using sensors, equipment, and interactions. CureSee Peyush Bansal invested ₹50 lakhs for 10% equity A digital health company, CureSee is enabling AI-based vision therapy software for treating eye diseases like amblyopia, computer vision syndrome, scleral lenses, and post-refractive surgeries. Through this innovation, patients will be able to access eye care on their phones anytime. The CureSee’s founding team includes Puneet, Jatin Kaushik, and Amit Sahni. The company aims to provide both online and offline care for patients. WOL3D Aman Gupta invested ₹80 lakhs for 2% equity and ₹70 lakhs debt @ 12% interest With a mission to develop and deliver cost-effective 3D printing machines, WOL3D has been building manufacturing and innovative technology in sectors like engineering, medicine, architecture, and manufacturing. This includes products like 3D prototyping, printing, filaments, and 3D pens as well. The company has been distributing the technology to brands like Flashforge, Creality 3D, and Qidi, among others. It was founded by Rahul Chandalia and now has offices in 10 cities across India. Primebook Aman Gupta and Peyush Bansal invested ₹75 lakhs for 3% equity. The first made-in-India Android laptops, Primebook is building and delivering affordable laptops for students. The laptops are powered by Android 11 ecosystem with MediaTek chips, along with 4G SIM connectivity. These have expandable storage up to 200GB and 4GB RAM, that powers the custom-built PrimeOS. The company has also launched an ARM-based laptop and plans to licence PrimeOS to further international OEMs. It was founded by Chitranshu Mahant, Aman Verma, and Chandrahas Panigrahi. Janitri Namita Thapar invested ₹ 1 crore for 2.5% equity (Condition – If the revenue of ₹20 crores is not achieved in the next financial year, the sharks will get an additional 2.5% equity) For making pregnancy monitoring easy, Janitri provides foetal and maternal monitoring solutions that meet medical standards and can be utilised both in hospitals and homes. Their technology is protected by patents and has been validated through clinical research to guarantee precision and safety. The products include Keyar CM, Keyar DT, Daksh (Pregnancy Monitoring Software, Keyar Mini, Keyar DT Lite, and an application for smartphones. The company is founded by Arun Agarwal. HoloKitab Namita Thapar invested ₹45 lakhs for 25% equity Redefining education with augmented reality (AR), HoloKitab is an edtech platform that enables book publishers to deliver AR content for students on books through smartphones. The product includes many books including the alphabet and counting for kids of very young age. Users can download the app and scan the books to activate the AR features. The app works offline as well and includes voiceover for better explanation. HoloKitab is created by Dipanshu Bajaj and Nikhil Miglani.","excerpt":"Here is the list of top tech startups that raised investment from the sharks","categories":["Deep Tech"],"tags":["AI Startups"],"author_name":"Mohit Pandey","publish_date":"2023-04-08T10:00:00","publication_year":"2023","word_count":842,"keywords":["Go","AI","innovation","Git","computer vision","RAG","Aim","ViT","Rust","R","AI Startups"],"extracted_tech_keywords":["AI","computer vision","Aim","RAG","R","Go","Rust","Git","ViT","innovation"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/8-tech-startups-on-shark-tank-india-s2-that-captured-our-attention\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10076236,"title":"Learnings Of Being A Chief AI Architect","content":"Analytics India Magazine interacted with Sripriya Venkatesan, Chief Architect at Capgemini who provided us with an in-depth insight on the current trends of Cloud and AI applications and their implementation in industries. With over 20+ years of experience in the IT industry, Sripriya has worked for key Fortune 20 clients globally on both large bids\/solutioning and delivery. She believes that as an architect, one must keep learning to stay updated with today’s fast-paced technology. Life as a Chief Architect Sripriya begins with narrating a typical day at work as the Chief Architect in Cloud and AI. She further explains that in the current, fast-paced environment, it is crucial to keep abreast with the latest technologies and continue participation in key industry forums. “At Capgemini, all our colleagues including Chief architects maintain work–life balance and deliver on a lot of end-to-end transformation projects which include solutioning, delivery architecture and innovation. A typical day would include customer interactions: understanding their task, ensuring the solution meets their business outcomes and helping them transform in their AI and Cloud journey.” Sripriya Venkatesan, Chief Architect, Capgemini Her main areas of focus are Cloud, Java, and AI related technologies, specialising in solutioning and innovation for new-age digital enterprises and end-to-end architecture for complex application landscapes. Sripriya works on large, end-to-end or niche, AWS transformation programmes—helping clients on their cloud journey.  She is also involved in innovative PoVs and PoCs, as it helps the group create niche assets and accelerators to help the company’s projects. “Having a holistic view is important—it’s a happy mix of solutioning, delivery, innovation, working on large bids, working on analyst RFIs, working on niche PoVs on AI: what an excellent learning experience!” Trends in AI application in organisations Few years ago, the chief architect led an initiative for AI in her group, working on key topics that were of interest in the European region. She says that it ranged from conversational UI to image processing and recommendation engines, making it an exciting journey to work on interesting and varied topics for automotive, manufacturing, retail and healthcare sectors. Moreover, one of her teams’ assets got the AWS hackathon award as well. “AI is fast moving and interesting, however it is no longer niche. AI is everywhere and organisations must adapt to it with the right approach and standards to ensure appropriate adoption benefits when using AI at scale.” She emphasised that the team follows a well-defined approach in Capgemini, especially in solving AI\/ ML projects. “The biggest challenge is data and cleaning. Working with the well-defined process using home grown tools and those provided by hyperscalers helps tackle this in a scientific manner. Over the last few years, we see a lot more maturity from hyperscalers also on varied topics e.g., sustainability. It is important to keep those factors in mind.” As the pandemic entailed a shift in work styles in the industry, Sripriya says, “Change is the only constant. Yes, this black swan event has brought many changes across the organisations and among our customers. With a hybrid model of working, many employees and organisations are reaping benefits of better work–life balance, managing day to day priorities along with frequent in-person interactions with teams.” The team with frequent touchpoints and rituals balance multiple projects and priorities, along with spending quality time with family and taking personal commitments. Sripriya states that AI is majorly implemented by companies at a large scale to bring business outcomes. “AI is everywhere. Organisations need to embrace it and start looking at AI in scale. We are no longer seeing small PoCs, PoVs or MVPs, rather organisations are now looking at using AI and Cloud to lead large programmes that can bring in business outcomes, help them reach their goals and also help in important aspects like sustainability.” ‘Glass ceiling exists but can be broken’ Sripriya is an active participant of several hackathons in the AI space, out of which the last one was the AWS hackathon where her team bagged the runners up position. She also leads the Women in Delivery ArchX initiative within Capgemini to assist other women architects on their technical journey. “I am passionate about women in STEAM and regularly mentor and participate at events like Society of Women Engineers, Women who Code, Women in Tech Network, Anita borg GHCI, Open Group, Association of Enterprise Architects—many topics related to AI and Cloud.” Sripriya shares some of her key learnings of being a Chief Architect: Mentor—always mentor upcoming talent who are open to learning and ready to move up in their career ladders.Stay humble. It is of utmost importance to stay grounded. Leaders don’t have to show their power and authority. It is always helpful to empathise—understand why someone is not able to stay late, understand their personal backgrounds and limitations, help the new mom who has just come back from her maternity break, support the dad whose kid is sick and keeping him awake all night, the tenant living in your house also deserves basic courtesy and respect. A leader who is understanding and humble will always go a long way in getting the team’s support.Ethics—always uphold the highest standards in Ethics. Give people their due, lead by example. And even if no one is looking, do the right thing and be prepared to stand-up for it. Leaders should always stand up for what is right, even if it means you are doing it alone. This is particularly true of women leaders who are role models for others and need to lead by example.Respect everyone—sometimes even juniors can come up with innovative solutions regardless of their age and experience. We often assume that consultants don’t know much but they too have a plethora of knowledge. Let everyone share and voice their opinions freely, and not just the opinions you would like to hear. She also states that Capgemini has a vibrant architect community with a strong focus on diversity. “Overall, in the industry we do see gaps in women technologists—many drop out due to personal reasons, lack of childcare in particular. It is also observed that many women move to non-technical roles, as they cannot keep up with the constant need to keep ourselves updated on technologies in today’s ever changing world of AI and cloud. Glass ceiling exists but can be broken.” Being one of the few women chief architects globally, she shares that she is proud and fortunate to have been mentored and supported by senior architects within the ecosystem. “Anything is possible with the right platform, training and opportunities along with focused mentoring to help women technologists.” Addressing gender inequality Sripriya has served as woman in delivery ambassador at Capgemini for the last three and half years. She leads the ArchX initiative—a platform to grow and connect to the outside world—that aims to take employees to the next level by providing opportunities on speaker slots, guidance on writing blogs, contribution and participation in external forums at various levels and more. “I also regularly mentor colleagues including customer architects on their technical journey. I am leading this initiative for my business unit as well—we regularly meet other colleagues, understand their concerns. It is important to have work happening on the ground to ensure we are able to bring an impact to women colleagues.” Sripriya shares a few steps on how organisations can tackle gender inequality which, according to her, is a difficult area for varied reasons: Support for working moms and new moms is a must—not just via policies but also empathy and emotional support within their respective teams.Show women colleagues the right role models. I once met a lady who had 100 patents and she spoke about her kids. That really hit close.Mentoring goes a long way. Speed mentoring and focused mentoring can do wonders when it comes to personalised attention and support.Help women employees build a brand—work with them on topics that are of interest outside of their day jobs: publications, speaking externally, hackathons, writing blogs, mentoring other women, participating in open source programmes, organising meet-ups, networking outside of work. This helps employees meet other fellow colleagues outside of work, network and get to the next level.Organisational policies must be supportive of women—this is a given and without which a healthy gender balance is difficult to achieve.","excerpt":"With over 20+ years of experience in the IT industry, Sripriya has worked for key Fortune 20 clients globally on both large bids\/solutioning and delivery. She believes that as an architect, one must keep learning to stay updated with today’s fast-paced technology.","categories":["IT Services"],"tags":["AI (Artificial Intelligence)","capgemini","Machine Learning"],"author_name":"Bhuvana Kamath","publish_date":"2022-10-03T11:00:00","publication_year":"2022","word_count":1377,"keywords":["Go","API","AWS","AI","R","ML","Machine Learning","Git","Aim","analytics","capgemini","Java","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","ML","analytics","Aim","AWS","R","Go","Java","Git","API"],"url":"https:\/\/analyticsindiamag.com\/it-services\/learnings-of-being-a-chief-ai-architect\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10055176,"title":"GitHub Rolls Out Technology To Improve Code Search","content":"GitHub has rolled out a technology preview for improvements to code search. Initially, the team has created a separate interface for the new code search as they build it, and after positive feedback, it will be integrated into the main github.com experience. The team plans to help developers search, discover, navigate, and understand code quickly and intuitively with code search. Code search will help maintain a flow state with auto-completion at every step and show the most relevant results first. The rich browsing experience is optimized for reading and understanding code, allowing developers to quickly make sense of unfamiliar logic, even for code outside their IDE. Covering a search index of over five million public and private repositories, the GitHub code search preview allows users to: Easily find what they’re looking for, with smart ranking and an index that is optimized for code.Search for an exact string of code, with support for substring matches and special characters, or use regular expressions (enclosed in `\/` separators).Scope searches with `org:` or `repo:` qualifiers, with auto-completion suggestions in the search box.Refine results using filters like language:, path:, extension:, and Boolean operators (OR, NOT). Search for definitions of a symbol with symbol:.Additional features include a directory tree view, symbol information for the active scope, jump-to-definition, and select-to-search. The improvements to code search follow the release of Copilot and Codespaces. “As a developer, staying in a flow state is hard. Whenever you look up how to use a library, or have a test fail because your developer environment has diverged from CI, or need to know how an error message can arise, you are interrupted. The longer it takes to resolve the interruption, the more context you lose,” wrote Pavel Avgustinov, Senior Director of Software Engineering at GitHub, in a blog post. These developments are a part of their unified solution to improve developer productivity.","excerpt":"The improvements in code search by GitHub will be integrated into the main github.com experience only after positive feedback","categories":["AI News"],"tags":["Copilot","GitHub","GitHub Codespaces","GitHub Repositories"],"author_name":"Meeta Ramnani","publish_date":"2021-12-09T18:44:59","publication_year":"2021","word_count":310,"keywords":["programming_languages:R","AI","Git","GitHub Repositories","ViT","GitHub","GitHub Codespaces","Copilot","R"],"extracted_tech_keywords":["AI","R","Git","GitHub","ViT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/github-rolls-out-technology-to-improve-code-search\/","complexity_score":2,"technical_depth":6,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":19081,"title":"Block Chain Technology— Trends, Security And Adoption In India","content":"The 20th edition of the Bengaluru Tech summit organised by the department of IT & BT, Government of Karnataka at Bangalore, covers all the aspects of currently emerging and trending technology. One such technology is Blockchain, and to present the views of trends, security and adoption were the evangelists from the industry who have been instrumental in shaping it up the way it has in India. “Technology has been evolving over the years and if someone would have talked about blockchain five years ago, it wouldn’t have made much sync with the people”, shared Sanjiv Kovil, CTO, Wipro, who was moderating the talk on blockchain. He said that the acceleration has been phenomenal for the blockchain technology and it has proven to be one of the most disruptive technologies of the decade. The concepts of decentralized approach, smart contracts, crypto currency which would have sounded vague till few years ago, has been now resonating with the tech freaks. Arifa Khan, CEO, Himalaya Labs, who has also been India Partner of Ethereum said that only it was only in the beginning of 2017 that blockchain became mainstream. The countries are still adopting the technology into their core competencies and has replaced the traditional way of working in many ways. Kovil shared that the reason behind blockchain witnessing a huge popularity is the fact that it has been able to streamline the processes along with offering other advantages such as cost benefits, high quality user experience and improving the overall efficiency. Kovil believes that its rampant acceleration will lead to disruption across many markets including energy, telecom and others. All sorts of evolving technologies like AI and IoT are getting into blockchain and lots of investments are coming. Arifa added that blockchain is continually being developed by developers all over the world and distributed cloud storage and distributed computing for complex mathematics are the key components to blockchain. “Developments such as lighting protocol, tangle protocol, and others will go a long way in keeping blockchains safe”, she said. Adoption of blockchain in India “There have a lot of trends and predictions around banks in terms of use of recent technologies and a lot of it has to do with artificial intelligence and blockchain”, said Sudin Baraokar, Head of Innovation, SBI, talking about its adoption in banks. He further shared that at SBI they have got all the banks together and work in a system to get blockchain solutions and get at least two of them into production, which they have been able to achieve.  “We are trying to adopt blockchain for solutions that we didn’t have”, he said. The intention is to build an entire financial and superstore from blockchain. Sreeram Ananthasayanam, Partner, Government and Public Services at PwC India said that there is a lot of hope around blockchain and it can change our lives in many ways. “By 2025, 10% of the GDP will be built on blockchain, and the scale is unimaginable”. It can increase supply efficiency that is one of the major reasons for a loss in GDP. Adopting blockchain in these three areas of self serving identity, registries and proof of ownership can do wonders for the country, and there are a lot of initiatives being taken up by central and state government. Developments around smart contracts and cryptocurrency Two of the most popular tools around blockchain have been the smart contracts and cryptocurrency. “The concept of smart contracts was introduced as early as 1970, but has become more relevant only today”, said Khan. These are nothing but self-executing contracts with the terms of agreement between the buyer and seller being directly written into lines of code. It can exist across a decentralised blockchain network. It can be translated to machines for performance verification to ensure there is no error on either side. The only vulnerabilities are at the program level. Cryptocurrency, a digital asset designed to work as a medium of exchange, has been built on the no-trust basis. It depends completely on cryptography to ensure secure transactions between the parties who are totally unknown to each other. If cryptocurrency is kept aside, blockchain can benefit in following areas—fintech, energy, cybersecurity and the government. Challenges around Blockchain technology Though it has been creating business opportunities, the technology is not averse to challenges. There is a thin line between transparency and privacy, which needs to be addressed. There are other non technical challenges in terms of the overall adoption of the technology and defining ROI. “There is uncertainty in protocol”, said Rajesh Dhuddu, SVP, Quattro. Largest adoption is in public places like cryptocurrencies, cloud storage and AI, but there are vulnerabilities from a protocol adoption perspective. The other challenge is that it has not been implemented widely, and night take few years before it runs full fledged. Krupesh Bhat, Founder, LegalDesk echoes the fact that the biggest challenge in India is its adoption. “There are technologies like e-signing but still it is not applied for cases such as rental agreements, there is the technology but no adoption”. On a concluding note While the speakers believed that blockchain is set to bring organization efficiency, they also agree to that fact that they have to still develop a proof of concept to witness larger adoption and build more infrastructure for blockchain so that common public is benefitted from it.","excerpt":"The 20th edition of the Bengaluru Tech summit organised by the department of IT & BT, Government of Karnataka at Bangalore, covers all the aspects of currently emerging and trending technology. One such technology is Blockchain, and to present the views of trends, security and adoption were the evangelists from the industry who have been […]","categories":["Deep Tech"],"tags":["bengaluru tech summit","blockchain technology india","cloud adoption in emerging markets","disruptive technology"],"author_name":"Srishti Deoras","publish_date":"2017-11-17T12:03:07","publication_year":"2017","word_count":890,"keywords":["disruptive technology","Go","API","artificial intelligence","Rust","AI","ML","distributed computing","cloud adoption in emerging markets","Git","RAG","bengaluru tech summit","blockchain technology india","R"],"extracted_tech_keywords":["AI","artificial intelligence","ML","RAG","distributed computing","R","Go","Rust","Git","API"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/block-chain-technology-trends-security-adoption-india\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":68848,"title":"Why Indian IT Professionals Are Looking To Upskill Themselves In Cloud Computing","content":"This internet boom over the last two has led to the growth in the demand for bandwidth from data centres. Lack of access to quality data networks and fully amped data centres in India is truly felt among organisations during this novel Coronavirus pandemic. To counter the pressure of work from home scenarios, Indian organisations are investing many times more than traditional IT spending in cloud infrastructure. In addition, cloud companies are expanding fast in the nation. Even global companies are flocking towards India. For instance, Oracle has come up with its second cloud region in Hyderabad to support customers’ demand for enterprise cloud services in India. The launch follows the launch of its Mumbai Cloud region in 2019, making India Oracle’s latest nation with multiple cloud regions available. India has also become the next big hot market for internet giants such as Netflix, Spotify, Facebook and Amazon, fuelling demand for cloud professionals who could manage the digital infrastructure. Amid this boom, training providers are witnessing a surge in enrolment in their information technology training programmes, including emerging technologies like cloud and data science. The jobs of the future will need expertise specific niche skills, and upskilling is the only way for a long term career growth for technologists. Hence certification programs are getting popularity among the IT professionals. According to analysts, COVID-19 lockdown has catalysed the enthusiasm of techies to getting certified. Indian software programmers are going for cloud certifications amidst COVID-19 lockdown, revealed a survey report from TechGig. Also Read: 10 Leading Courses & Training Programmes For Cloud Computing In India Techies Know Cloud Certifications Help Boost Careers Extensive understanding of a new-age technology appeared the most crucial reason for techies to take certifications. Also, freshers and new joiners are more interested in acquiring certifications than working professionals. Cloud technology – which is helping communication and remote working amid the present COVID-19 lockdown – is also the preferred option for upskilling for the Indian developers, notes TechGig. The preference for cloud came on top of other advanced technologies like artificial intelligence and machine learning. “In today’s unique COVID-19 time, technology is the only string which is keeping the world together. From cloud computing, which is supporting work-from-home to artificial intelligence, which is backing banking, retail, and important sectors run operations. Besides, cloud computing is crucial for robotics that is helping the front-line hospital personnel; new-age technologies are assisting the globe to connect in the existing time. The TechGig survey shows the enthusiasm of Indian developers to upskill on these new-age technologies,” said Sanjay Goyal, Vice President & Head of Product and Technology at TechGig Looking for people with cloud skills is a complex endeavour. Organisations these days are finding it very difficult to hire and retain cloud specialists, particularly in roles requiring advanced cloud skills and cloud architecture. Therefore, companies are giving due importance to both finding and creating the skills in-house so they do not face infrastructure challenges. Also, given the introduction of new services from the three major cloud platforms – Google Cloud Platform, Amazon Web Services, Microsoft Azure and others, cloud training has to be constant so people can stay on top of the technology. Training providers are witnessing a surge in enrolment in their information technology training programmes, including emerging technologies like cloud and data science. Cloud technology is one of the leading tech domains for upskilling among the techies and other technologies like artificial intelligence, machine learning, and quantum computing, getting the highest preference in terms of the need for upskilling. According to TechGig ‘IT Certification Survey,’ one of the most important findings was that 90% of the respondents revealed that they are planning to have an IT certification soon to support and boost their career prospects. That’s why the adoption of certification courses is on the rise. Also Read: 10 Leading Courses & Training Programmes For Cloud Computing In India Why Cloud Skills Are Hot Among IT Workers Cloud computing has risen to be the most sought-after skills set in the world for the last few years, and in particular, in 2020, companies are migrating their infrastructure and apps to cloud platforms. As a consequence, cloud jobs are also growing at a swift pace, making it one of the hottest fields in information technology. Now, with the demand for cloud experts, it has fuelled the need for niche skills, and IT professionals know that well. It is clear that IT professionals will not face any issue with employment opportunities if they are skilled in cloud technologies space, particularly for platforms such as AWS, Google Cloud, Microsoft Azure. Consequently, learners are developing skills so they can grab the jobs as a cloud developer\/administrator or system operators for cloud platforms after finishing their training programs. The platforms are utilised by thousands and thousands of businesses worldwide for hosting their products and services. Cloud training courses will provide professionals with the opportunity to learn the best techniques and practices in cloud computing and acquire live feedback from an expert instructor. Training will help learners to take cloud certification exams from vendors- AWS, Azure or Oracle certifications to get recognised by hiring managers. The upskilling is spread across advanced classroom training programs run by specialised institutes like Jigsaw Academy and Great Learning, etc, which have also witnessed a surge in demand for enrollment. Apart from training institutes, learners are also flocking to cheaper and\/or free courses from cloud vendors or those found on Udemy. In fact, in a recent survey done by Analytics India Magazine, 76.9% of the analytics professionals are spending their time on training through self-learning. While IT professionals not already working with cloud technologies will gain a solid foundation, those with some cloud experience will gain a more structured and hands-on understanding of cloud technologies, including issues such as migration, deployment, integration, platform choice, and architecture. According to reports, COVID-19 pandemic has caused the desire to get certified, and professionals understand that certification is a need of the hour amid mass layoffs. DevOps, infrastructure-as-a-service, software-as-a-service, automation, agile and software-defined networks are going to be critical for IT professionals to land these jobs. Some platforms are offering interesting courses for learners to build their cloud tech skills, including many free courses to build cloud tech skills. Also Read: 10 Leading Courses & Training Programmes For Cloud Computing In India","excerpt":"This internet boom over the last two has led to the growth in the demand for bandwidth from data centres. Lack of access to quality data networks and fully amped data centres in India is truly felt among organisations during this novel Coronavirus pandemic. To counter the pressure of work from home scenarios, Indian organisations […]","categories":["AI Trends"],"tags":["AI Certifications","AWS cloud","best jobs in india","Cloud Computing","cloud computing India","CLOUD SKILLS","latest in cloud computing technology","latest technology in cloud computing","Oracle Certification","quantum cloud software","quantum computing companies"],"author_name":"Vishal Chawla","publish_date":"2020-07-02T18:08:22","publication_year":"2020","word_count":1055,"keywords":["CLOUD SKILLS","AI Certifications","R","best jobs in india","data science","artificial intelligence","analytics","Oracle Certification","Go","latest in cloud computing technology","machine learning","AWS","cloud computing India","AI","cloud computing","quantum computing companies","latest technology in cloud computing","quantum cloud software","Cloud Computing","AWS cloud","Azure"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","data science","analytics","cloud computing","AWS","Azure","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/why-indian-it-professionals-are-looking-to-upskill-themselves-in-cloud-computing\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":64670,"title":"Can Human Rights Be AI-Centered?","content":"Human rights must govern artificial intelligence (AI) in order to address the phenomenon of the lack of transparency of algorithms and the data that this technology feeds on. The New York-based research center, Data & Society, recently released a document titled Governing Artificial Intelligence: Upholding Human Rights and Dignity, in which it presents some examples of how artificial intelligence can violate people’s human rights, in addition to offering an overview of what the members of society who come into contact with this technology can do about it: companies, civil society, governments, the United Nations, intergovernmental institutions and academia. “For AI to benefit the common good, at least its design and deployment must prevent damage to fundamental human values. International human rights provide a solid and comprehensive formulation of these values.”, it is mentioned in the document. Examples of the consequences of not putting human rights at the center of the development of artificial intelligence projects are not few. Recently, Amazon had to get rid of an artificial intelligence assistant who made hiring its human resources easier because it began to manifest clear discriminatory biases against women. The organization ProPublica, and especially the researcher Julia Angwin, have been in charge of revealing how the algorithms are “black boxes” that can introduce racist biases in judicial decisions and in an article published on October 21, 2018, the MIT Technology Review exposed how various groups of human rights activists in the United States have spoken out against the creation of “a massive database containing the names and personal data of at least 17,500 people suspected of being involved in criminal gangs.” The Data & Society document presents examples of how, through artificial intelligence projects developed and implemented by companies like Facebook, governments like China and even universities like Stanford, they can involve systematic human rights violations such as the right to non-discrimination, to privacy, political participation and freedom of expression. “Stanford University researchers trained a deep neural network to predict the sexual orientation of their study subjects, without obtaining their consent, using a set of images collected from online dating websites. Beyond several methodological shortcomings, the research showed how a lack of respect for the right to privacy increases the risks of algorithmic surveillance, with which the data that is collected and analyzed threatens to reveal personal information about users. This may put individuals and groups at risk, particularly those living under regimes that would use such information to repress and discriminate, ”the report explains. In what way are companies and organizations worldwide addressing this problem? One of the companies that are implementing this kind of best practices and recommendations around artificial intelligence systems is IBM, through its Trust and Compliance program, which is available to customers of the company’s cloud service and which has the ability to deliver an indicator report that will tell you how its artificial intelligence platform is behaving from two points of view: on the side of the creation and development of algorithms and from the data set that has been used to train these algorithms. Although Trust and Compliance are only aimed at customers of the IBM cloud service and universities, this company program stands out against the measures taken by other technology companies such as Google, which also faced claims from its own workers who agreed to discontinue the work in an artificial intelligence project called Project Maven. “Just a few months later, many Google employees feel that those principles have been pushed aside with an offer for a $ 10 billion Defense Department contract. A recent study was done at North Carolina State University also found that asking software engineers to read a code of ethics does nothing to change their behavior, ”says the MIT Technology Review report. If AI researchers, developers, and designers work to protect and respect fundamental human rights, they could open the way for broad societal benefit. To ignore human rights would be to close that path. References HAO, K., Establishing an AI code of ethics will be harder than people think, MIT Technology Review, https:\/\/www.technologyreview.com\/s\/612318\/establishing-an-ai-code-of-ethics-will-be-harder-than-people-think\/ LATONERO, M., Governing artificial intelligence: upholding human rights & dignity https:\/\/datasociety.net\/wp-content\/uploads\/2018\/10\/DataSociety_Governing_Artificial_Intelligence_Upholding_Human_Rights.pdf","excerpt":"Human rights must govern artificial intelligence (AI) in order to address the phenomenon of the lack of transparency of algorithms and the data that this technology feeds on. The New York-based research center, Data & Society, recently released a document titled Governing Artificial Intelligence: Upholding Human Rights and Dignity, in which it presents some examples […]","categories":["AI Features"],"tags":[],"author_name":"Dr. Raul V. Rodriguez","publish_date":"2020-05-06T18:00:00","publication_year":"2020","word_count":686,"keywords":["Go","artificial intelligence","AI","neural network","Julia","Aim","Rust","GAN","AI research","R"],"extracted_tech_keywords":["AI","artificial intelligence","neural network","Aim","R","Go","Rust","Julia","GAN","AI research"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/can-human-rights-be-ai-centered\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10096930,"title":"6 Popular Social Media Platforms that Tanked in The Last Decade","content":"Meta launched Threads on July 5 and became the fastest-growing app ever. Founder Mark Zuckerberg tracked every milestone as it crossed 2 million sign-ups in the first two hours, 5 million in the first four hours, 10 million in seven hours and 30 million in 24 hours. Threads got a whopping 100 million sign-ups in a week. This was unprecedented growth for any platform in the history of the internet, but what’s the guarantee that this rapid success will sustain? According to experts, a social media platform without a clear purpose, target audience, seamless tech or a combination of any one of these is headed for a miserable failure. Even Facebook, a platform that found Zuckerberg incredible success, is losing users every day. Here is a list of six platforms that lost out in the last decade. Vine Vine, founded in 2012 by Dom Hofmann, Rus Yusupov, and Colin Kroll, quickly gained popularity as a unique platform for creating and viewing content. It was acquired by Twitter for $30 million in the same year and released on Android, iOS, and Windows in 2013. However, Vine’s success was short-lived. Its initial dominance in the market was due to the lack of competition. Top creators like King Bach, Nash Grier, Amanda Cerny, and Rudy Mancuso amassed millions of followers and billions of Vine “loops”. Unfortunately, Vine’s downfall was just as swift as its ascent. The main reason behind Vine’s failure was its failure to establish a monetisation program to reward its creators. Despite earning millions in ad revenue from user-generated content, Vine’s creators struggled to turn their creativity into a sustainable career. Top creators attempted to negotiate official deals with the Vine management, but their requests were rejected. Consequently, these creators migrated to YouTube, where they could compile their Vines into videos and earn a living. Vine’s refusal to compensate creators for their work contrasted with the approach of TikTok, a platform that understood the importance of supporting and rewarding creators. Clubhouse Launched in April 2020, Clubhouse, a social audio app, initially attracted users through its invite-only feature and notable personalities like Elon Musk and Mark Zuckerberg. However, there was no moat in the business. A few months later, Twitter announced a new feature called Spaces in the microblogging platform, which made Clubhouse obsolete. Besides, as pandemic restrictions eased, daily active users decreased significantly, especially after celebrities and their followers left the platform. Clubhouse’s management failed to establish a business model, lacked user and analytics data, and couldn’t curate conversations effectively. In April, this year the company announced they were laying off more than half of their employees. The founders said they need to “reset” the company in the aftermath of COVID-19. Musical.ly Musical.ly was a short-form video social media platform that gained over 200 million users during its active years which began in April 2014. It allowed users to create and share short videos with soundtracks, focusing on lip-syncing content. The app was headquartered in Shanghai and became particularly popular with pre-teen and teenage girls, especially in the US. In 2017, it was acquired by ByteDance, the parent company of TikTok, for around $800 million. However, ByteDance eventually decided to merge Musical.ly into TikTok, resulting in the shutdown of the Musical.ly app. The merger allowed TikTok to consolidate users and features from Musical.ly into its platform. The decision to shut down Musical.ly and merge it with TikTok was driven by ByteDance’s desire to leverage the user base and technology of both platforms, streamline resources, and compete more effectively in the global video-sharing market. Yik Yak Yik Yak was an anonymous social media app that allowed users to make posts visible to others within a 5-mile radius launched in November 2013. It gained popularity, particularly on school and college campuses in the US, with 1.8 million downloads by September 2014. The app’s founders, Tyler Droll and Brooks Buffington, secured funding and raised the company’s value to around $400 million at its peak. The app struggled with site performance, leading to slow loading times and an inability to handle high traffic. One significant factor in Yik Yak’s decline was its issue with cyberbullying. The app’s anonymity and hyper-localisation made it a breeding ground for bullying. Instances of threats, violence, and harassment were reported, leading to media outcry, school bans, and negative user experiences. Yik Yak attempted to address the problem by implementing geo-fencing, and disabling app usage in schools. However, this decision severely impacted its user base, as schools and colleges were its primary market. Kik Kik Messenger was a free mobile messaging app that allowed users to send audio, text, images, GIFs, and videos which began its services in 2010. It differentiated itself by not requiring phone number sign-up and offering anonymity and chatbot integration. Kik faced challenges due to changing strategic directions, ownership changes, and content moderation issues. In 2019, the app was sold to MediaLab after unsuccessful attempts at reinventing the company. MediaLab continued to maintain the platform without significant innovation. Its decline can be attributed to its constant shifts in focus, lack of effective monetisation, distraction with ICOs, and inability to effectively moderate content. The ownership of Kik currently rests with MediaLab, a Los Angeles-based holding company with experience in running digital assets. The exact acquisition price remains undisclosed. Google+ Google+ had seamless technology and a large number of users who had already signed up for Gmail. The company also sent many private invitations and hyped up the platform so much that it had 10 million users within 2 weeks of launch in 2011. After a year, it only had 90 million users. Even with superior technology, Google failed to understand customers’ wants and needs. Facebook already existed as a platform connecting people, and Google wanted users to share everything, including emails, tweets, photos, videos, which users found unnecessary. It failed to connect people outside of its own ecosystem. The concept of ‘circles’ was interesting but was confusing to use and sharing posts and images was complicated. Google+ also lacked a strong mobile experience. It tried to be everything and ended up being nothing that users needed. Things changed further when its founder Vic Gundotra left in 2014. Its collaboration with Photos and Hangouts was abandoned and it came out as an independent social networking site. It further detached YouTube and GooglePlay and was left confused on what the purpose of the platform was. Strong competition and no vision finally led to its demise.","excerpt":"A social media platform without a clear purpose, target audience, seamless tech or a combination of any one of these is headed for a failure","categories":["AI Trends"],"tags":["Facebook","social media","threads"],"author_name":"K L Krithika","publish_date":"2023-07-14T12:01:25","publication_year":"2023","word_count":1075,"keywords":["Go","API","AI","ML","social media","threads","Git","RAG","GAN","ViT","analytics","Facebook","R"],"extracted_tech_keywords":["AI","ML","analytics","RAG","R","Go","Git","API","GAN","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/6-popular-social-media-platforms-that-tanked-in-the-last-decade\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10012503,"title":"Inside Genpact’s Analytics And AI Unit: In Conversation With Sanjay Srivastava, CDO, Genpact","content":"The COVID pandemic has brought uncertainty in businesses and enterprises across the globe are facing unexpected challenges. To manage these challenges, protect business continuity and build resilience, governments and businesses must make big decisions — and fast. To ensure this, analytics and data science teams are at the centre of today’s strategic decision-making. AI and tech-enabled initiatives are helping companies overcome the challenges in an efficient manner. To get an insight into Genpact’s analytics and AI initiatives, we got in touch with Sanjay Srivastava, Chief Digital Officer, Genpact, who spoke in lengths about it. At Genpact, Srivastava runs analytics, AI, digital and technology services businesses. Apart from this, he also serves on the advisory board of a Silicon Valley AI incubator, a digital accelerator, and several tech startups. Below are the excerpts: How does Genpact maintain a competitive edge using data analytics? There are a few key building blocks every company needs to remain competitive with data analytics. Building a foundation of data: For large companies, it’s about unlocking the value of the data they already have by lighting it up using AI and ML techniques.Leveraging Connected Ecosystems: Enterprises need to leverage not only the data that sits inside the company but also data they can get from their ecosystem of clients, providers, vendors, and partners, allowing them to optimize for the entire ecosystem.Understanding the weak links: Now we have compute engines strong enough to uncover the weak links and indicators that would not be obvious to the human workforce. Bringing these weak links to light allows enterprises to optimize or rethink their operations. Combining the power of data, advanced analytics, machine intelligence, and human judgment – also known as augmented intelligence – empowers the enterprise to make more informed decisions. As companies deal with the pandemic and look to transform themselves, it is critical to redesign and transform their processes. In order to do that, they need the right foundation of data, the capability to drive analytical insights to drive business results and AI and automation to instrument these insights. Can you also tell us about the work (proprietary algorithms or models) that you leverage\/ and or are currently working on in relation to AI\/ML? After deploying hundreds of projects at some of the largest companies in the world, we have learned a few important lessons. It’s easy to experiment but it’s hard to scale. Large companies need a new business technology architecture which allows them to leverage previous investments and add new innovation. This new business technology architecture needs to provide integration capabilities, modularity, a governance framework, and be defined by experience as the true north. Genpact Cora is Genpact’s digital business platform that does this. It is designed to incorporate the best capabilities of partners in the ecosystem, but to do so in a manner that is modular, with compostable services that become part of a greater end-to-end transformation. AI needs training with the right context and business insights. AI platforms available on the market are like a shell, they just provide the compute engine – but in order for real adoption in the enterprise, AI models need to be pre-trained with the right goal orientation and contextualization. Genpact has rolled out pre-trained AI accelerators that are designed to address the critical barriers to AI adoption in the enterprise. These prebuilt engines combine AI technology, deep data sets, and pretrained, contextually relevant industry domain expertise, drawing on Genpact’s experience running operations for hundreds of large companies across multiple industries. Talent is a key to success – talent needs to be bilingual. We believe that success belongs to those who exhibit or learn to hone “bilingual” talent – the combination of domain and digital. An AI application, if properly built, can get us to a reasonable starting point — predictive accuracy of 80% or more. As exceptions are processed the humans, the AI application fine-tunes its underlying algorithms to increase prediction accuracy. Having the right human in the loop with bilingual talent (domain and digital) to train the model is the key to success. Can you tell us about the technology leadership initiatives that you are driving at Genpact? We’re seeing disruption create new opportunities for growth, and businesses across the world should look to and plan early for, that growth as we rebound out of this crisis. In this sense, we’re leveraging trends that we’re seeing in the market: The move to the cloud: We’re seeing that businesses are accelerating their digital transformation. Prior to COVID-19, many were on a five-year roadmap to digitize their operations. Now, most want to become fully digital within just two years. In fact, there has been an accelerated consumption of cloud-based services and solutions and we are seeing a 10x faster business acceleration to the cloud. Data-driven insights: The COVID-19 pandemic is fundamentally changing the way we do business. Predictive analytics will drive an optimized reality. Customers expect the companies they engage with to anticipate their needs and behaviours. This is only possible for utilizing predictive insights and analytics. Predictive insights improve resilience and will transform operations. Using accurate prediction models will allow businesses to shift from a reactionary detect-and-respond paradigm to a proactive foresee-and-prevent model that also enables businesses to react seamlessly to the unpredictable. The ability to predict with precision and at scale will change business models – and, once again, redefine consumer expectations.Human-centred design: Businesses need to create employee experiences, rather than just using technology – The best and most updated technology does not necessarily equate to providing the best experience to your workforce. A deliberate and strategic vision aligned with user needs should be the defining factor when harnessing technology. In the new world, digital customer experience will need to reinvent these old ways of engaging with customers and transacting business. And this redesign will require reimagining entire customer journeys, building new digital capability, intelligently automating much of lower value tasks, and improving the customer experience digitally. What is the day-to-day work of the in house analytics\/ML or data science team? AI is fundamentally focused on these three aspects: Build the right pipeline for data: Building the right data foundation drives actions and decisions that can make or break businesses. On top of the data foundation, our digital experts’ layer in business and process knowledge, data science, AI, and machine learning capabilities. Our data scientists have implemented In-Memory Data Architectures to rethink “fit for consumption” data that can support on-demand consumption of data. This includes extract, load and transforms operations along with implementing a data governance strategy. With the exponential increase in the volume of data, both structured and unstructured, data governance has taken the center stage. Ask the right questions: Once the data is cleaned and ready to use, the more important step here is to leverage the data to find out what are the right questions to ask for the business.  Based on the available data, the ideal modelling techniques are chosen, and the data is then split into two parts for model building — a training data set and a validation data set that is used to test the model performance. When the model performs in accordance with the established business criteria, its impact is captured, and the model is approved for deployment.Use the best analytics tools to answer the questions: Our Data scientists engage in projects aimed at converting insights to-real world positive business outcomes by using descriptive or diagnostic or predictive or prescriptive analytical tools such as Next Best Action modelling for offering a seamless customer experience and many other advanced Artificial Intelligence based algorithms. We also focus heavily on the hand holding our clients to ensure that they truly understand the insights generated and thus enable faster adoption rates at scale. With agile as our modus operandi, we continuously evolve our solutions based on frequent feedback from clients and finally deliver ‘human-centric’ augmented intelligence solutions with best-in-class user experience. What machine learning tools\/frameworks\/libraries do you use frequently? The breadth of technologies: Depending on the specific use case and the technology stack preferred by our clients, our data science teams decide the technologies be used for the projects. Decisions on libraries-of-choice are also determined by the type of data – structured or unstructured and the nature of the data science challenge, that is, whether it is a credit card fraud detection model, or it is an omnichannel chatbot. Also, it is important to realize that the enormous computing power needed to solve such complex real-time business challenges demands virtual machines. Cloud-based technologies: We help large organizations through various types of cloud journeys, from migrating legacy systems to cloud platforms, to operationalizing AI models in a cloud environment or managing complex data lakes. In order to enhance our data management, data science and artificial intelligence operationalization capabilities, we have established strategic partnerships with multiple vendors, such as AWS, Azure, Cloudera and others. Since the pandemic, our clients started prioritizing their existing infrastructure to cloud-based platforms. And we enabled the seamless transition to cloud platforms for them. Pre-trained Accelerators: One of the biggest challenges enterprises face in adopting AI at scale is the need to train models with real-life data in a specific domain. We have rolled out ‘Pre-trained’ AI and analytics accelerators. They allow us to rapidly and modularly build an enterprise application at scale, without starting from ground zero for every business use case. With such accelerators deployed across all industries, companies can start off their journeys performing at more than 90%+ accuracies from the start. What are the other main components of the tech stack? (Eg. Cloud, Programming languages, databases, etc.) Big Data Management – We use platforms such as MongoDB, PostgreSQL, Redis, Azure Cosmos DB and Hadoop to effectively manage the 4 V’s of data management — Volume, Variety, Velocity, and Veracity. Our data ingestion and data pre-processing capabilities stem from expertise in a plethora of industry-standard big data management platforms.Programming Languages: We use programming languages such as Python, R, Julia, SAS, Scala and Java to help ingest, clean, and transform both structured and unstructured data.AI\/ML Libraries and Solvers: We use libraries and solvers such as Pytorch, Keras, TensorFlow, IBM CPLEX, COIN-OR and GUROBI Optimization, for training models to mimic human behaviour and utilize adequate computing capabilities. Specific libraries are used dependent on the nature of the Data Science challenge.Storytelling Tools: We utilize the visualization tools like Power BI, Tableau, Qlik and others to empower business leaders with important data-driven insights. Using these storytelling tools, we also visualize models, its variables, transformations, changes in hyper-parameters during training and outputs. This helps us to debug and fine-tune ML models quickly and deploy with minimal downtime.Deployment Platforms: We provide end-to-end solutions ensuring smooth deployment of complex AI\/ML models across any production environment. We deploy various techniques to bundle an application together with all relevant libraries, configuration files, dependencies, code and tools to run it anywhere on cloud or on-premise across a variety of computing environments irrespective of operating systems. We enable such seamless operationalization through technologies such as Docker, Kubernetes, Amazon ECS and many others. Can you tell us about the research projects which Genpact is currently working on? With our clients’ heightened focus on driving digital transformation, the last several months have demonstrated that our deep domain depth and process expertise, as well as our investments in digital and analytics, give us a competitive advantage. A few examples include: Supply chain optimization – as an example, during the pandemic medical product manufacturers have had to deal with significant challenges predicting demand and a safe return to work for their employees. They address these challenges using epidemiological data along with AI\/ML tools to predict demand by location to organize and arrange manufacturing capacity, including scheduling workforce return with an eye to keeping employees safe.eCommerce: As the world increasingly shifts towards eCommerce, the prevalence of financial crimes and associated risk is increasing. AI\/ML capabilities are exceptional at looking through large sets of data to detect anomalies and are very useful in addressing them.Remote collaboration: As we increasingly work in remote locations, it becomes very important to make sure that all the core equipment is running – AI\/ML gives us the ability to remotely monitor the equipment and predict maintenance requirements such that uptime can remain high. What are the technical skills (such as programming languages) that you are looking for? We tend to look for candidates that have the ability to apply advanced algorithms to solve critical business problems that ultimately lead to positive business outcomes. Candidates with both technical skills and domain expertise have a higher likelihood of being able to identify opportunities where AI\/ML solutions can be applied to solve real-time business challenges. Additionally, any candidate with previous experience of working in AI projects at scale should be considered as a significant hire. We believe that a potential data scientist should have a right mix of 8 elements – attitude, analytical mindset and critical thinking, business acumen, proficiency in mathematics and statistics, competency in coding, understanding of data architectures and storytelling and communication skills. Any candidate with the right blend of these 8 elements is a true bilingual. Bilingual talent helps us apply augmented intelligence techniques to truly transform the business decision making cycles, which in turn enables faster adoption of augmented intelligence at scale. What does the roadmap look like for the company in terms of core digital technologies? We at Genpact embrace technology at our core. The best results for digital transformation occur when you can bring three different muscles into play: Digital capabilitiesDomain knowledgeTransformation services But it is important to build IP in digital. We are taking a thoughtful approach to where and how we build IP [and where we don’t]. We have established a framework across services IP [including frameworks, playbooks and POV] and tech IP [reference architectures, domain data models, business rules and data labels]. We are not building IP in areas where we can leverage hyperscalers and tech providers. Our strategy of honing in on a specific set of chosen industry verticals and service lines, where we have a deep domain and process expertise, coupled with our investments in digital and analytics capabilities, has positioned us very well to be that strategic transformation partner. As the market continues to evolve at speed, clients in every industry are focused on leveraging digital technologies and analytics to gain a competitive advantage. We partner with them to realize sustainable business transformation for them.","excerpt":"The COVID pandemic has brought uncertainty in businesses and enterprises across the globe are facing unexpected challenges. To manage these challenges, protect business continuity and build resilience, governments and businesses must make big decisions — and fast. To ensure this, analytics and data science teams are at the centre of today’s strategic decision-making. AI and tech-enabled […]","categories":["AI Features"],"tags":["ai big data analytics digital transformation","big data trends and challenges","collaboration ai platform","companies using business analytics","corporate analytics platform","different types of analytics","genpact AI","genpact analytics","hadoop on azure cloud","hadoop problems","hadoop world","human touch to ai","Interviews and Discussions","is hadoop a company","supply chain analytics projects","types of analytics","Types of Databases"],"author_name":"Vishal Chawla","publish_date":"2020-11-27T13:00:56","publication_year":"2020","word_count":2405,"keywords":["types of analytics","data science","artificial intelligence","PyTorch","corporate analytics platform","analytics","supply chain analytics projects","machine learning","is hadoop a company","ai big data analytics digital transformation","AI","ML","companies using business analytics","human touch to ai","big data trends and challenges","different types of analytics","hadoop world","hadoop on azure cloud","Keras","Types of Databases","genpact AI","hadoop problems","Aim","collaboration ai platform","TensorFlow","genpact analytics","Interviews and Discussions"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","data science","analytics","Aim","TensorFlow","PyTorch","Keras"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/inside-genpacts-analytics-and-ai-unit-in-conversation-with-sanjay-srivastava-cdo-genpact\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":6,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10092591,"title":"This Company is Paving the Way for Generative AI Services","content":"Generative AI has gained significant attention in recent years, piquing interest in its potential for automation. This ongoing revolution has placed Generative AI at the forefront of human interaction with automated systems. While users traditionally engage with such systems by posing questions and issuing commands, Generative AI adds a new dimension of functionality to the equation. Automation Anywhere is one such company driving the transformational power of generative AI through its Automation Success Platform (ASP). By integrating Open AI’s GPT models natively, the company is partnering with several GCCs in India to elevate its operations. In an interview with Ankur Kothari, Co-founder and Chief Customer and Strategy Officer at Automation Anywhere, Analytics India Magazine learned how the company is shaping the landscape of generative AI services. Kothari explained the impact of Language Model Masters (LLMs) on intelligent automation, serving as a turbocharger for compliance and guided system operation. Enterprises, particularly those on the consumer side, are experiencing a paradigm shift using generative AI solutions. How GenAI is Shaping Enterprises Kothari says that his team has been deeply immersed in the world of Generative AI, experimenting with various algorithms and working on hundreds of different use cases for clients and partners. “We offer a comprehensive end-to-end solution, starting from identifying processes that can be automated to creating bots that can execute these actions, to leveraging our IDP product to ingest unstructured or semi-structured data, and ultimately managing and maintaining these bots while generating insightful analytics,” Kothari said. Giving an example of how the company provides solutions, Kothari explained, “Imagine you are an airline company, and your customers frequently call or email you with requests to change their tickets. Oftentimes, these emails are not very well-structured and may not include all the necessary information to process the request efficiently. This is where natural language processing comes in.” With the help of natural language processing, Kothari said that the company can use a machine learning algorithm to extract relevant information from the customer’s email and pass it on to the bot. The bot then takes this information, such as the customer’s ticket ID, and searches the database for all other relevant information about that passenger. Once it has all the necessary information, the bot can make the requested changes and generate a response back to the customer. He adds that it’s important to note that there is still a human involved in this process, particularly when it comes to generating the response that is sent back to the customer. Companies are implementing guided processes that allow a human to review and modify the bot’s response before it is sent out to ensure accuracy and avoid any potential errors. Moreover, he highlighted that the platform is the sole cloud-native RPA product available, meaning it operates efficiently on both on-premises and cloud infrastructures. “This enables us to deliver cutting-edge solutions to our customers without any limitations,” Kothari stated. Additionally, the platform boasts a powerful automation and co-pilot feature that allows users to summon bots from various applications, streamlining processes and enhancing productivity. Fear of copyright infringement When asked about how organisations can address the issue of intellectual property being exposed in open databases, as Samsung recently experienced, Kothari acknowledged the challenges companies face in grappling with this critical issue. There’s discussion happening in every boardroom on how to use this technology. He noted that there are two choices: ignore it and risk falling behind the competition, or work with vendors to create policies and guidelines to use it securely. However, Kothari acknowledged that not all problems have been solved yet, as this technology is still in its early stages. He stressed the importance of companies creating governance and policies around what can and cannot be touched, particularly in industries like automotive, where these technologies are heavily used. Nevertheless, intelligent automation platforms like Automation Anywhere can help companies work with partners to bring these technologies into their companies in a more controlled and secure way. At present, enterprises are in the exploration phase, and many are collaborating with companies like Infosys, TCS, Accenture, or Automation Anywhere to determine how they can leverage this technology. “By combining these technologies and leveraging each other’s strengths, companies can maximise the benefits of this innovation while maintaining security and control,” says Kothari. Do Not Fear AI Over the past couple of months, the tech world has been seeing thousands of layoffs. There have been huge job losses in tech and IT sectors. While the common reason given is the uncertain market conditions in America, many co-relate to this automation as well. However, Kothari believes that while some companies initially approach automation with apprehension, fearing job loss – the reality is quite different. “Automation allows companies to do more with less, enabling them to leverage new opportunities and create new roles that were not possible before,” says Kothari. He says that the fear of job loss that is often associated with new technologies is unfounded, as history has shown that technology actually facilitates job creation and innovation, leading to a more competitive business landscape. Kothari further explains that recent news regarding job losses has nothing to do with automation. Automation technology has been in use in India for over a decade, and job losses have not been a concern. Instead, the adoption of automation has led to the creation of new roles and offerings for clients, which would not have been possible without this technology.","excerpt":"With the help of natural language processing, Kothari said that the company can use a machine learning algorithm to extract relevant information from the customer’s email and pass it on to the bot.","categories":["IT Services"],"tags":[],"author_name":"Lokesh Choudhary","publish_date":"2023-05-02T13:00:00","publication_year":"2023","word_count":904,"keywords":["Go","API","GenAI","machine learning","AI","ML","RAG","generative AI","analytics","R"],"extracted_tech_keywords":["AI","machine learning","ML","analytics","generative AI","GenAI","RAG","R","Go","API"],"url":"https:\/\/analyticsindiamag.com\/it-services\/this-company-is-paving-the-way-for-generative-ai-services\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10171871,"title":"Ecuador Surgeon Performs Historic Robotic Heart Surgery With SSi Mantra 3","content":"Dr Juan Zuniga, a cardiothoracic surgeon at Interhospital in Guayaquil, Ecuador, recently performed the first-ever robotic cardiac surgery in the Western Hemisphere. The procedure used the SSi Mantra 3 surgical robotic system by SS Innovations, an American company headquartered in Gurugram, marking a notable advancement in global healthcare and robotic surgery technology. The operation involved the robotic closure of an atrial septal defect (ASD), a procedure used to repair a hole between the heart’s upper chambers. It was performed using minimally invasive methods. According to SS Innovations, this achievement demonstrates the system’s capabilities and potential to significantly impact cardiac care in underserved regions. SS Innovations confirmed the successful outcome and recovery of the patient, while Zuniga emphasised the system’s precision, ease of use and clinical benefits. “The experience was good. This equipment was very easy to handle and manipulate,” he said. He further noted that robotic technology enabled greater control and accuracy during surgery. “Because it is a minimally invasive procedure, the recovery is much faster, bleeding much less, and since it is done robotically, the manipulation of the tissues is much more precise, and so is the suturing.” The surgery was conducted at Interhospital, the first facility in South America to install the SSi Mantra system. The system has already been used in a variety of complex procedures. It is part of SS Innovations’ ongoing efforts to expand access to advanced surgical robotics across Latin America, the United States, and Europe. To date, at least 4,100 procedures have been performed globally using the SSi Mantra. SS Innovations develops surgical robotic technologies to make robotic surgery affordable and widely accessible. The company’s system features multiple robotic arms, a 3D 4K monitor, and over 40 specialised instruments, supporting its use across multiple surgical fields, including cardiac operations. Last year, SSi Mantra received regulatory approval from the Central Drugs Standard Control Organisation for telesurgery and teleproctoring. The system then performed its first surgeries in India earlier this year and assisted in performing Indonesia’s first robotic cardiac surgery in November last year.","excerpt":"“This achievement demonstrates the system’s capabilities and potential to impact cardiac care in underserved regions significantly.”","categories":["AI News"],"tags":["robotic inventions","robotic surgeon","robotics technology","SS Innovations"],"author_name":"Sanjana Gupta","publish_date":"2025-06-17T20:03:26","publication_year":"2025","word_count":339,"keywords":["Go","robotics technology","programming_languages:R","AI","SS Innovations","innovation","robotic inventions","programming_languages:Go","ai_applications:robotics","GAN","robotic surgeon","R"],"extracted_tech_keywords":["AI","R","Go","GAN","innovation","programming_languages:R","programming_languages:Go","ai_applications:robotics"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/ecuador-surgeon-performs-historic-robotic-heart-surgery-with-ssi-mantra-3\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10053944,"title":"Council Post: Statistical Tools That Every Data Scientist Should Know For Better CPG Analytics","content":"Data is making its way across all industries and verticals today, becoming one of the must-haves for organisations on their way to success. In the current economic climate, organisations need to take advantage of all the available data resources to stay on the right track by basing their decisions on thorough analysis. Data plays a different role depending on the industries, and depending on their product\/ service offerings, organisations tend to vary the amount of data-driven usage in the company. In the CPG industry, utilising data in the overall business strategy is the real key to staying ahead of the competition and creating a cycle of continuous growth. The nature of CGP is volatile in context to changing consumer demands and market trends. Data science allows organisations to leverage consumer and organisational data. It offers approaches to identify insights, forecast trends and make informed business decisions based on data-driven predictions. Data scientists tend to lack the in-depth knowledge in statistics that could further their insight generation. Statistical tools are essential for data analysis. This is especially true when it comes to industries like CPG that are so volatile to external factors. They allow organisations to utilise quantitative methods to test data-driven theories in a real-life scenario. The latest AI-based techniques and predictive economic modelling tools help organisations systematically identify the economic factors that can influence their business decisions. This combination of data with the quantitative application of statistical and mathematical models helps data scientists to test existing hypotheses and forecast future trends. Data scientists in CPG industries can easily adapt econometrics, given their deep understanding of the maths behind linear regression or panel data analysis techniques. For instance, CPG organisation leaders can use econometrics to optimise the promotional spend and ROI for a market and use econometric & statistical tools to quantify the relationship and draw conclusions. There are four broad categories of statistical tools that can be used, depending on the organisation’s use case, their product\/service offerings, and the kind of insight they aim to derive from their data. Depending upon their use case, data scientists can choose among the four tools or combine them for better results. Let’s look at these tools in some detail. Descriptive Statistics Descriptive statistics essentially entails the measure of central tendency and dispersion. The tool measures central tendency, dispersion and distribution of data using statistical techniques. Data scientists can use this tool for summarising and describing their dataset and enhance their exploratory data analysis by describing the characteristics of the data. Descriptive statistics is important because it helps data scientists understand the data they are dealing with in a comprehensive manner. Getting information about the variables involved and the potential relationships between those variables is the first step in using the data for analysis. Central tendency assists data scientists in measuring the median values of a dataset, thereby helping them navigate the central location of data to focus on. Descriptive statistics is further used to calculate the measure of the dispersion of variables in the dataset regarding the centrally found values. This can be done through Range, Interquartile Range (IQR), Standard Deviation, Variance, Mean Absolute Deviation, Coefficient of Variation, Gini Coefficient, etc. Lastly, the methods can be summed up using statistical distribution to calculate the probability of an occurrence happening again. The CPG industry can be easily affected by something as immediate & small as a weekend storm that prevents consumers from shopping. Descriptive statistics helps companies use past data and make future insights in a timely manner. CPG manufacturers can leverage their historic data to understand the shopper’s experiences and use the information to create real-time insights. For instance, let’s take the COVID-19 pandemic crisis and the fluctuating demand for sanitisers. Studying the economic data can highlight the potential relationship between new COVID-19 waves and the demand for sanitisers; CPG analysts can predict the rise or fall in demand and manufacture sanitisers accordingly. Regressions Along with the descriptive analysis, data scientists can use regression analysis to investigate the relationship between the dependent and independent variable(s). In CPG, the technique is best used to find the causal effect between the variables. For instance, taking the sanitiser example, data scientists can use regression techniques to determine the relationship between the rise of COVID-19 cases and the demand for sanitisers. Linear Regression techniques are used to quantify the relationship between several variables and adjust for confounding effects. Data scientists can opt for simple or multiple linear regression depending on the nature and count of explanatory variables involved in the problem. The regularisation techniques such as lasso, ridge or elastic net can complement the analysis in the case of a huge set of predictor variables. Another recommended regression technique is the Panel Data Model to model time series data and forecast time-dependent observations. It provides multidimensional data related to an observation that has been measured repeatedly over a span of time. This could include variables such as individuals, product choices, city, household items, etc. Essentially, it provides information on the difference in variables across individuals- over a period of time. The panel data model uses techniques such as Pooled OLS, fixed-effects model and the random-effects model. Forecasting One of the most predominant use cases of statistical tools is forecasting. Forecasting market trends and consumer demands is the basis of CPG, and it is important to correctly identify these for better business ROI. Benchmark forecasting is a technique used to build the forecasting intuition that can be used as a baseline for additional complex layers. Benchmark forecasting entails techniques such as Drift, Naive, Seasonal, Mean, Seasonal naive, Random walk, Linear trend, and Geometric random walk. Intelligent forecasting is a critical tool for data scientists, given its adaptive nature. While generic modelling tools are built keeping the industry in mind, forecasting tools can completely adapt to the company’s needs. They intake the company’s historical data as dependent variables within the model, allowing the following indicators to be extremely company-specific. It is important to note that a one-size-fits-all approach to forecasting does not work for CPG and retail businesses. Instead, leaders and data scientists should build a category, region, or product-specific forecasting method. The model’s ability to customise, add, or change easily allows it to become flexible to ensure accuracy is maintained across the various aspects of the business planning process for CPG. For instance, data scientists can use historical data of the supply chain problems when the petrol prices have been hiked up. They can use this to forecast when the next petrol price hike would occur and its impact on the supply chain. They can then use this information to prepare in advance. Hypothesis Testing So you have understood your data, and you have made hypothetical predictions based on the data. The next tool is to ensure that your predictions are correct and that the manufacturers can go ahead with them. Hypothesis testing is an effective statistical tool to assist data scientists in gaining supporting material for their findings and conclusions. The tool focuses on measuring claims against accepted facts about the whole population. Measures such as the p-value can support or reject the claims or confidence intervals to measure the degree of uncertainty. In addition, CPG focused data scientists can utilise hypothesis testing to verify how probable the detected consumer behaviour is. They can do so using several methods, such as The t-testAnovaChi-square test For instance, data scientists working in a clothing retailer have drawn the hypothesis that the sale of tank tops is high during summer among girls between the ages of 15-30. Data scientists will use hypothesis testing tools like the t-test to test and prove this hypothesis. As you would have noticed, these tools don’t act independently. This is why they help data scientists in creating a holistic view of industrial data and its impact on economics. The interconnectedness of econometric tools with data analytics is essential for data scientists to consider while working on CPG and FMCG applications. The implementation of statistical modelling for forecasting and price analysis in CPG is a critical phenomenon in the growing future of data science. The views, thoughts, and opinions expressed in this article belong solely to the author and does not reflect the views and opinion of the author’s employer, any other organizations, committee or other group or individual. This article is written by a member of the AIM Leaders Council. AIM Leaders Council is an invitation-only forum of senior executives in the Data Science and Analytics industry. To check if you are eligible for a membership, please fill the form here.","excerpt":"Data is making its way across all industries and verticals today, becoming one of the must-haves for organisations on their way to success. In the current economic climate, organisations need to take advantage of all the available data resources to stay on the right track by basing their decisions on thorough analysis. Data plays a […]","categories":["AI Features"],"tags":["Statistics","Statistics for Data Science","Statistics Machine Learning"],"author_name":"Indrajit Mitra","publish_date":"2021-11-22T13:00:00","publication_year":"2021","word_count":1427,"keywords":["data science","Go","AI","Statistics","data-driven","RAG","Aim","Statistics Machine Learning","Statistics for Data Science","analytics","ViT","GAN","R"],"extracted_tech_keywords":["AI","data science","analytics","Aim","RAG","R","Go","GAN","ViT","data-driven"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/council-post-statistical-tools-that-every-data-scientist-should-know-for-better-cpg-analytics\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":32255,"title":"Amazon To Use Machine Learning To Increase Its Footprint In The ‘Kirana’ Market","content":"Early in 2018, Amazon ambitiously started working in the area of grocery delivery operations in India. With its Amazon Pantry section, it started hyperlocal grocery deliveries, including fresh vegetables, fruits and whole food in Bengaluru, Delhi-NCR, Mumbai and Hyderabad. Amazon is taking the local kirana stores by storm and is challenging the interests and profits of domestic retailers. The company is quite clear with its aim to increase its kirana footprint in India, for which it is majorly resorting to technologies such as machine learning to mark larger benefits. How Machine Learning Comes Into Picture Grocery has become the next frontier for e-commerce players with many big and new names trying to capture the market. Amazon too has been experimenting with its grocery play in India and trying out multiple formats to get the best bet. It is quite a unique move by the company as it has not tried something around grocery in any other geography. The company had earlier mentioned that it aims to increase their engagement with kirana stores from where a shopper can get food and other essentials delivered from nearby stores within 2-4 hours of placing an order. The company had also said that they are looking to increase the company’s delivery network and increase their reach of areas where products can be delivered. With this new delivery service, the owners of kirana shops can upload their inventories onto the Amazon website for local shoppers to see what’s available in nearby stores. Once the products are identified, they can either be delivered through the company’s delivery service or through logistics companies. All of this will be facilitated by the new age technologies, especially machine learning. While Amazon has been pushing the use of machine learning in a lot of its initiatives, the recent adoption comes with an aim to increase its kirana footprint in India. As the company promises its commitment to put customer expectations at the heart of every transaction, machine learning will play an important role to help deliver the same. During a recent conference and addressing a gathering of Amazon executives as a part of its Women In Technology conference, Amit Agarwal, Senior Vice President, India Consumer Business, Amazon said that the company is using machine learning to identify defects in its products. “We use it to check for defects in pages that list these products,” he had said. The company would earlier do it manually which was a cumbersome task as humans would manually scour through pages of information to search for defects. Similarly, in various other inside workings, it uses ML. It is heavily relying on the use of technology, and thousands of data scientists being hired to work on ML and improvise on the processes, yield a better result and streamline the process. Amazon’s History With Machine Learning This is not the first time that tech and e-commerce giant has used machine learning in its functioning. Reportedly the company has invested over Rs 13,800 crore in India, and they are relying heavily on ML and AI to boost up their market share in the country. The company has been using machine learning for more than 20 years, with thousands of engineers being hired to in the new tech space. They are adopting transformative technologies in several ways to improve customer experience. Some of the areas where it is heavily adopting ML are product returns, improve the speed and accuracy of product deliveries, provide more relevant search results and improve efficiency in other areas of its business. Rajeev Rastogi, director, machine learning at Amazon India had said that there are a lot of problems that are India-specific where you need machine learning. He gave an example of address quality, as he suggested that addresses in India are highly unstructured and ML can be used to improve it drastically. It has also been using AI for product-delivery, location-related queries, standardising sizes in fashion products, market search, advertising, improving catalogue, and more.","excerpt":"Early in 2018, Amazon ambitiously started working in the area of grocery delivery operations in India. With its Amazon Pantry section, it started hyperlocal grocery deliveries, including fresh vegetables, fruits and whole food in Bengaluru, Delhi-NCR, Mumbai and Hyderabad. Amazon is taking the local kirana stores by storm and is challenging the interests and profits […]","categories":["Global Tech"],"tags":[],"author_name":"Srishti Deoras","publish_date":"2018-12-27T04:45:05","publication_year":"2018","word_count":660,"keywords":["machine learning","programming_languages:R","AI","ML","Aim","R"],"extracted_tech_keywords":["AI","machine learning","ML","Aim","R","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/amazon-to-use-machine-learning-to-increase-its-footprint-in-the-kirana-market\/","complexity_score":2,"technical_depth":6,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10126275,"title":"Who is Snorting the $20 Billion Dry Powder in India?","content":"Indian VCs have a secret—it’s called ‘dry powder’ and it’s just waiting to be invested. Dry powder is essentially money that’s not tied up in other investments and can be quickly deployed when the right opportunity comes along. Rajan Anandan, the managing partner at Peak XV Partners, recently spoke about this. He sees potential for growth and innovation, and foresees India making a significant impact on the global stage. At the Global IndiaAI Summit, Anandan said, “Our firm has over INR 16,000 crore of dry powder. We just want more people starting up in AI.” He added that Peak XV has invested in 25 AI startups so far. Emphasising that capital availability is not a problem, Anandan mentioned that the venture capital ecosystem has $20 billion ready to be invested in Indian startups. He also noted that AI is currently the most significant theme, with investors showing strong enthusiasm for startups in this field. Recently, Speciale Invest also hosted DevCon, with 50+ technical founders and engineering leaders. Even Stellaris Venture Partners recently hosted an AI Agents Hackathon, in a bid to attract a lot of talent in the space. It’s only a matter of time before the $20 billion dry powder is ‘snorted’ by the AI companies of India, as every VC is slowly turning into an AI-focused investment firm. The Mystery Behind the $20-Billion ‘Dry Powder’ This brings up the question: Where is all this money coming from? Kyle Stanford, a senior VC analyst at PitchBook, noted that a large portion of the capital invested in high-end venture deals comes from asset managers like mutual funds, hedge funds, and private equity funds. “A lot of recently closed VC funds have been holding onto their capital and waiting for the market to bottom out, or they have a better sense of what the pricing of these deals should be,” he said. When it comes to India, according to data from research firm Preqin, PE\/VC dry powder in India increased to $15.6 billion by March 2023, up from $12.8 billion in 2022 and $11.1 billion at the end of 2021. Seventy India-focused PE\/VC firms closed funds in 2022, raising an aggregate of $8.5 billion — the highest-ever annual fundraising value, Preqin said. Strong fundraising in the first half of 2022 also contributed to these substantial dry powder reserves. Global venture firms amassed a total of $223.6 billion through Q3, with most of the capital coming from the first two quarters. If this pace continues, it will surpass the total capital raised in 2021, according to PitchBook data. According to a Bain & Co report, the funding from the investors’ side has seen a significant decline since 2021, with annual funds dropping from $40 billion in 2021 to just $4 billion till June 2024. Source: Bain & Co. Moreover, the share of fundraising by top investors varies significantly each year, peaking at 85% in 2019 and dropping to 4% in 2023. There is a notable increase in total funds raised in 2022 compared to other years, and a significant drop in 2023. This means that the fund accumulated in 2022 is still kept by the investors, adding the cumulative money from all the years. Meanwhile, data from the Q3 2022 PitchBook-NVCA Venture Monitor shows that through Q3 2022, the total value of venture deals involving nontraditional investors reached $145.1 billion, representing about 74% of all US VC deals. This group of investors participated in around 92% of mega-sized venture deals in 2021. A key factor enabling global VCs to maintain and grow their record amounts of dry powder is the significant involvement of nontraditional VC investors. This additional capital, which isn’t counted in the VC dry powder figures, has increased the funds available to startups and helped expand the asset class beyond what is accessible to traditional VC firms. More Funds Needed in India When speaking with AIM, Arjun Rao, managing partner at Speciale Invest, said that the investors can’t give funds of $10 million or more because that is around 20% of the total fund size for many. “We probably need more funds that are of larger size. The gap is in the VC market itself and hopefully some of the homegrown funds that have started in the past 10 years are growing by them having done well in the past cohorts. Their AUM is growing and therefore they can write large checks,” said Rao. Rao also explained that one of the reasons for such a small cycle of funds is the time horizon of their LPs. “We are sort of bound by that. It’s like, ‘Hey! I got to invest, see the company, build, and grow and then make returns and then return the capital to our investors within a meaningful time horizon’,” he added. Many believe that Indian investors are risk averse. But that is not the case according to Anandan as Peak XV has already invested in two semiconductor companies, a space tech company, and a hydrogen recycling company, proving their risk-taking capabilities. While the term “funding winter” has been floating around, Anandan’s optimism provides a refreshing contrast. Despite the challenges, the availability of such a substantial amount of dry powder ensures that promising ventures won’t be left out in the cold.","excerpt":"Where is all this money coming from?","categories":["AI Features"],"tags":["Startups"],"author_name":"Mohit Pandey","publish_date":"2024-07-09T16:14:51","publication_year":"2024","word_count":874,"keywords":["Go","API","funding","AI","RPA","innovation","venture capital","Aim","Startups","R","startup"],"extracted_tech_keywords":["AI","Aim","R","Go","API","RPA","innovation","startup","venture capital","funding"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/who-is-snorting-the-20-billion-dry-powder-in-india\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10147756,"title":"Karnataka Launches Entrepreneurship Program to Boost Startups","content":"On Tuesday, the Karnataka government announced the Rajiv Gandhi Entrepreneurship Program to strengthen the state’s startup ecosystem and support budding innovators. “Karnataka is home to 18,000 startups, 45 unicorns, and 38% of India’s 435 soon-to-be unicorns. The remarkable growth stems from a thriving ecosystem, entrepreneurial spirit, and strong industry-government collaboration, placing Karnataka at the 4th position globally in innovation ecosystems,” IT minister Priyank Kharge posted on LinkedIn. The program aims to assist 30 selected innovators by providing a monthly stipend of Rs 25,000 for a year. This financial support will help participants navigate the ideation phase, culminating in the development of a prototype, minimum viable product (MVP), or proof of concept with business potential. The Karnataka government has also introduced two new initiatives, ELEVATE 2024 and the Karnataka Accelerator Network (KAN), aimed at boosting the development of early-stage and growth-stage startups across the state. AI Startups in Bengaluru Talking about AI startups, the emergence of companies like Sarvam AI, which focuses on developing advanced AI models, reflects Bengaluru’s commitment to technological innovation. Krutrim AI, led by Ola’s founder Bhavish Aggarwal, has raised $50 million in funding at a $1 billion valuation and is now India’s first AI startup to reach unicorn status. Innovation in AI isn’t limited to corporate giants. Startups like KOGO are disrupting the market with their AI operating systems, offering modular AI assistants tailored for diverse industries. Meanwhile, Karya AI is pioneering in the rural employment sector, leveraging AI to create job opportunities through tasks in local languages, with notable partnerships with tech giants like Microsoft and Google. These home-grown initiatives underscore Bengaluru’s status as a hub for AI innovation, promising transformative impacts across multiple sectors. Speaking to AIM at the IGIC event 2024, Sanjeev Kumar Gupta, the CEO of Karnataka Digital Economy Mission, mentioned, “There are over 1,000 AI startups in Bengaluru. With initiatives like Beyond Bengaluru, we are also promoting AI startups, in Mysuru, Hubli, and Mangaluru.” Recent statistics highlight Bengaluru’s burgeoning startup landscape, boasting over 10,000 startups valued at approximately $50 billion. These figures underscore the city’s entrepreneurial vigour and its ability to foster an environment conducive to startup growth. Applications for the Rajiv Gandhi Entrepreneurship program are open until January 4, 2025, at 5 pm. “We’re investing in Karnataka’s future entrepreneurs and disruptors. If you are one of them, I invite you to take advantage of this opportunity,” Kharge said.","excerpt":"“Karnataka is home to 18,000 startups, 45 unicorns, and 38% of India’s 435 soon-to-be unicorns.”","categories":["AI News"],"tags":["Priyank Kharge","Startups in Bangalore"],"author_name":"Shalini Mondal","publish_date":"2024-12-24T18:04:09","publication_year":"2024","word_count":397,"keywords":["Go","Startups in Bangalore","AI assistants","AI","innovation","Git","Priyank Kharge","RAG","Aim","ViT","GAN","R"],"extracted_tech_keywords":["AI","Aim","RAG","AI assistants","R","Go","Git","GAN","ViT","innovation"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/karnataka-launches-entrepreneurship-program-to-boost-startups\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10095147,"title":"Top 10 Generative AI Courses in 2024","content":"From creating images that look like real life to helping with creative writing, Generative AI is transforming what machines are capable of achieving. There is increasing demand for skilled individuals in this field, making it especially important to keep up-to-date on the latest knowledge and abilities related to generative AI. This is why we have gathered together some carefully selected courses focused on updating you on everything you need to know about generative AI. Our comprehensive selection caters to both beginner-level participants and experienced practitioners who want to explore artistic possibilities or practical applications using generative AI technology. Let’s delve into the realm of best generative ai courses and embark on a thought-provoking educational experience that will equip you with the ability to create, innovate, and influence what lies ahead.| 1. Generative AI Crash Course with Hands-on Implementations Made available by The Association of Data Scientists, the premier global body in the field of AI and Data Science, this comprehensive course is ideal for individuals seeking to learn Generative AI from foundational concepts to advanced techniques. Beginning with an exploration of the fundamentals of Generative AI, the course progresses to hands-on implementation of popular models like GPT, Stable Diffusion, and DALL.E. Enriched with informative reading materials, video tutorials, and practical Python implementation, this course offers a holistic learning experience, equipping learners with the necessary knowledge and skills to develop a strong foundation in Generative AI. Access Here >> 2. Generative AI in Action: Solving Complex Business Problems This is another very useful course published by The Association of Data Scientists that focuses more on utilizing the power of Generative AI in different business problems. Starting with covering the most important and highly relevant business use cases, this course also covered how to build generative AI-powered applications by integrating the OpenAI APIs. Access Here >> 3. Building Systems with the ChatGPT API This 1-hour course is featured by deeplearning.ai and delivered by Isa Fulford and Andrew Ng. It covers how to automate workflows using chain calls to a language model. You will learn to enhance development capabilities, create prompt chains, build systems with Python code, and construct a customer service chatbot. Apply these skills to real-world scenarios in this course. Access Here >> 4. How Diffusion Models Work In this another useful course by deeplearning.ai, you will delve into the world of diffusion-based generative AI and learn to build your own diffusion model from scratch. You will gain an understanding of the diffusion process and the underlying models. This course will help you develop practical coding skills through hands-on labs on sampling, training diffusion models, and building neural networks for noise prediction. Taught by Sharon Zhou, this one-hour course expands your generative AI capabilities, equipping you with the knowledge to build, train, and optimize diffusion models. Access Here >> 5. Generative AI learning path by Google Google Cloud has released this learning path that consists of a package of 10 courses. This learning path leads you on an expertly selected pathway of resources exploring generative AI products and technologies. It covers everything from the basics of Large Language Models to the practical aspects of developing and implementing generative AI solutions on the Google Cloud platform. Professionals who have limitations or restrictions in utilizing Google Cloud might discover certain constraints with the course. Access Here >> 6. Large Language Models: Application through Production Focused on Large Language Models and published by Databricks, this course is available on edX. This course is specifically designed for practitioners seeking to build LLM-centric applications using cutting-edge frameworks. You will find Hugging Face for NLP problem-solving, LangChain for complex tasks, prompt engineering, data embeddings, and vector databases in this course. Enhance LLM pipelines, fine-tune models with domain-specific data, evaluate societal and ethical implications, and deploy at scale using LLMOps. Access Here >> 7. OpenAI, GPT, ChatGPT and DALL-E Masterclass This comprehensive Masterclass on GPT, DALL-E, and ChatGPT is taught by Raymond Davey and is available on Coursera. This course equips developers, data scientists, and enthusiasts to harness OpenAI’s accessible AI capabilities. You can explore prompt engineering, API integration, CodeX, DALL-E, fine-tuning, embeddings, and chatbot creation in this course. This course also provides practical skills with detailed explanations, code examples, and prompts. Become a master of GPT and OpenAI’s transformative technologies. Access Here >> 8. Generative AI – From Big Picture to Idea, to Implementation This course explores the advancements of Generative Artificial Intelligence and its potential. From various application fields to effective implementation, it delves into project ideation and practical starting points. It helps in discovering existing Generative AI models, relevant coding aspects, and architectural concepts. Additionally, this course also addresses ethical considerations and preventive measures by companies and governments. Access Here >> 9. Getting Started with Generative AI API Specialization You will find a specialized journey exploring the world of generative AI through Python in this package of courses. It covers text, image, and code generation, highlighting OpenAI’s powerful API. Each course in this package helps learners with building practical skills and culminates in a real-world, portfolio-ready project. It covers building captivating recommendation systems, intelligent GPT-powered chatbots, and mastering image generation and manipulation with Dall-E and PIL (Python Imaging Library). Access Here >> 10. Build AI Apps with ChatGPT, Dall-E, and GPT-4 This comprehensive course empowers you to integrate mind-blowing AI features into your applications using the OpenAI API. Gain proficiency in utilizing the Dall-E, GPT-4, and ChatGPT APIs, and master the art of fine-tuning models with your own data. Through hands-on projects like MoviePitch and KnowItAll, explore idea generation, image creation, and chatbot development. Elevate your AI skills and unlock the potential of OpenAI’s cutting-edge technology. Access Here >> The top 10 generative AI courses offered above provide individuals of all backgrounds the opportunity to acquire skills and knowledge for utilizing this groundbreaking technology. Whether exploring art or practical applications, these courses offer tailored learning experiences. Embrace generative AI, unlock new avenues, and shape the future.","excerpt":"A curated list of the top 10 Generative AI courses, designed to equip you with essential skills","categories":["AI Trends"],"tags":["Courses"],"author_name":"Sourabh Mehta","publish_date":"2023-06-14T17:17:15","publication_year":"2023","word_count":991,"keywords":["data science","ChatGPT","artificial intelligence","OpenAI","AI","neural network","NLP","Ray","LangChain","generative AI","Courses"],"extracted_tech_keywords":["AI","artificial intelligence","neural network","NLP","data science","generative AI","ChatGPT","OpenAI","LangChain","Ray"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-10-generative-ai-courses-in-2023\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10093529,"title":"Is AI Copyright Really Necessary?","content":"Copyright infringement is increasingly becoming the talk of the AI world with governments imposing laws trying to rein in AI makers and their AI systems. Recently, during his appearance before Congress, OpenAI CEO Sam Altman agreed with the need for government regulation for building responsible AI systems, this included giving proper attribution and rights to the original content creators. “We have been talking with the artists and content owners about what they think about this [copyright and attribution],” said Altman. “I think people totally deserve control over how their content and likenesses are used in this technology.” Altman said that the creation of AI models like these was meant for helping and benefiting the creators and artists, and not stealing ownership. “That is exactly what the economic model is,” said Altman. But the same cannot be said about the government. On the other hand, it seems as though the artists and creators that the company wants to assist with its products is actually against its adoption. Similar to how the Writer’s Guild of America was protesting against the use of ChatGPT in script writing, major music labels are sending notices to streaming services to take down “AI soundalikes“. While Altman and other AI makers are concerned about aligning their models with the economic and ethical objectives, the government might be in this to get more control over the developing technology, and shape its future. During the same hearing, Gary Marcus, the AI sceptic, spoke about how there are no existing laws that talk about the copyright issue with these generative AI models. To this, Senator Josh Hawley replied, “We can just say that section 230 of copyright law can be applied to models like this.” If the government gets the power to control AI development, with the reason of “protecting” copyright owners, the very high pace of progress of these AI models would just stop. Moreover, Hawley’s statement makes it clear that the government wants to impose an all-out copyright ban on generative AI models, which can be tricky as it would not address the concerns but merely make the government rule over AI. Do we really want that? David Holz, the founder of Midjourney, doesn’t really care about copyright infringement. In an interview with Forbes, Holz said that he is using images without seeking permission from the owners. He explains that it is impossible to do so with the huge dataset. Regulations with Scepticism Under the new proposed draft of the AI act by the European Parliament, any content generated by AI models like Midjourney or Stable Diffusion will have to mention that it has used copyrighted material for training, and appropriately attribute the original creator. This sounds like fair practice unless looked closely. Under Article 28b (4) of the draft, it is written that the foundational model used in AI systems to generate either images, text, audio, or video, should comply with the transparency obligations of Article 52 (1), and publicly disclose a detailed summary of the training data under the copyright law. If these generative AI companies have to go back and address all the corpus of data it has been trained on, which is essentially scrapped from the internet, and is put under the copyright law, they would not be able to create any more AI models. Furthermore, which specific image or generated content utilises which specific fed data is untraceable even for the companies that created the models. Or even if they can, it is extremely difficult to trace and thus difficult to attribute. For example, Shutterstock took a step to address proper licensing of stock images used for training AI generated images. To compensate the artists, the company would pay the “fair share” through royalties each time the image was used for generating art. Though there was no precise method to explain how the model would work for this, the company did launch a contribution fund to compensate the artists. This received a lot of criticism as it meant treating the artists’ work as tokens. Moreover, this also put the makers of these models, like OpenAI and StabilityAI, under a lot of pressure, hindering further development of this technology and risking the potential benefits. When it comes to copying images or using someone else’s work manually or through Photoshop-like softwares, the copyright holders will know for sure that their work was used. But when it comes to these generative models, the amount of data they are trained on will make their offices fill up with lawsuits. According to the new act, anyone breaking the copyright act as mentioned in Article 28 and 52, the foundation model providers are liable to a fine of “€‎10 million or 2% annual turnover, whichever is higher.” On the other hand, it is true that these technologies bring up several ethical questions with them about training on unauthorised and private data. But there needs to be a fair balance between copyright infringement and building generative AI models. Is there a balance? Amid the concerns around the European Parliament’s proposed AI act and the US Senate discussion, the Copyright Office of the US has specified some guidelines regarding the registration of works created solely by machines. For instance, if an AI technology generates intricate written, visual, or musical works based on a prompt from a human, the traditional elements of authorship in such works will not be registered. The reason behind this is that AI technology determines the expressive elements, rather than the human user, making the generated content ineligible for copyright protection. Nonetheless, a work that incorporates AI-generated material can still be eligible for copyright protection if it includes a sufficient amount of human authorship. For instance, if a human creatively selects or arranges AI-generated content or modifies it to the point where the modifications meet the standard for copyright protection, then copyright protection applies only to the aspects that were contributed by the human. In a similar notable event, a macaque monkey named Naruto captured selfies using a photographer’s camera. Subsequently, the photographer faced a lawsuit from the People for the Ethical Treatment of Animals (PETA) who contended that Naruto, the monkey, was the rightful owner of the photographs, and thus the photographer was infringing on Naruto’s copyright. However, the Court of Appeals for the 9th Circuit ruled that nonhuman entities are not eligible for copyright protection. This decision aligned with the US Copyright Office’s definition of an “original work”, which explicitly requires a “human author” to be involved. This might be the way forward for generative AI as well. Instead of following the European path, the US should stick with the current copyright laws. Generative AI companies should adopt techniques and make ways to incentivize artists and original authors. This would encourage AI innovation with sufficient regulation, and not too much control by the governments.","excerpt":"Copyright laws and original author attribution sounds like a fair practice unless looked closely.","categories":["AI Features"],"tags":["ai copyright","AI laws","AI Regulation","AI Rights","ChatGPT","MidJourney","OpenAI"],"author_name":"Mohit Pandey","publish_date":"2023-05-17T18:30:00","publication_year":"2023","word_count":1138,"keywords":["Go","ChatGPT","MidJourney","OpenAI","AI","AWS","AI laws","AI Regulation","RAG","GPT","stable diffusion","AI Rights","generative AI","ai copyright","R"],"extracted_tech_keywords":["AI","generative AI","ChatGPT","OpenAI","RAG","AWS","R","Go","GPT","stable diffusion"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/is-ai-copyright-really-necessary\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":23232,"title":"John Giannandrea Steps Down As Google AI And Search Head","content":"In a surprising turn of events, Google’s senior vice president of Engineering division, John Giannandrea is stepping down from his position as head of the Search and Artificial Intelligence (AI) units. According to an exclusive report by The Information, Giannandrea’s role will be split into two. Jeff Dean, who leads the Google’s deep learning AI research team Google Brain, would run a separate AI wing in addition to his current position. Another news report adds that Ben Gomes, who joined Google in 2000 and was running search engineering department, will take over the search organisation. Reportedly, a spokesperson has confirmed the news and said that Giannandrea had decided to step down himself. “The move will allow him to get more hands-on time with technology and our understanding is that he will stay at Google,” states the report. A machine learning expert, Giannandrea joined Google in 2010 after the company acquired his startup Metaweb Technologies. He took the role of senior vice president of Engineering from Amit Singhal in early 2016, signalling Google’s emphasis on weaving ML and AI into Search. Google chief executive officer Sundar Pichai has been betting big on advances in AI and machine learning. Earlier this year, reaffirming Google’s commitment to an ‘AI first’ world, Sundar Pichai, CEO at Google, had drawn a parallel between AI and other important discoveries for humans, such as fire and electricity. “AI is one of the most important things that humanity is working on. It’s more profound than — I don’t know — electricity or fire,” he had said. Pichai had admitted that while there are concerns surrounding artificial intelligence, the potential that the technology holds in driving developments in areas such as finding cures for cancer, for example, were tremendous. He had also added that it was important to adopt embrace the change — especially in technology.","excerpt":"In a surprising turn of events, Google’s senior vice president of Engineering division, John Giannandrea is stepping down from his position as head of the Search and Artificial Intelligence (AI) units. According to an exclusive report by The Information, Giannandrea’s role will be split into two. Jeff Dean, who leads the Google’s deep learning AI research team Google […]","categories":["AI News"],"tags":["AI (Artificial Intelligence)","Google","Sundar Pichai"],"author_name":"Prajakta Hebbar","publish_date":"2018-04-03T11:54:05","publication_year":"2018","word_count":307,"keywords":["Go","artificial intelligence","machine learning","AI","ML","deep learning","Google","GAN","AI research","R","Sundar Pichai","AI (Artificial Intelligence)","startup"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","R","Go","GAN","startup","AI research"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/ai-search-john-giannandrea-google\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":32698,"title":"Uber, Walmart &#038; Netflix Look To India, The Next Growth Engine To Write Their Success Story","content":"Image for representation purpose The Indian market has now become the hottest battleground for US-based companies thanks to its large talent pool and yet untapped market. However, the push for additional research and development facilities in the Indian subcontinent might indicate bigger plans for these leading tech behemoths, which are viewing India as a hub for AI talent and revenue growth. India has long been viewed as a source of specialized tech talent. Protectionism measures and the subsequent growth of established R&D labs in the country shows that companies are now seeing India as a strategic market. After Silicon Valley, the next AI\/ML labs in Asia are now being set up in India. The year 2018 saw a spate of investments in Indian startups, the most notable being Walmart picking up a stake in Flipkart. Now, other US tech giants Uber and Netflix are doubling up their tech staff count in India to beef up operations. In this article, we take a deep dive into US tech majors’ India strategy and how the giants are looking to India to accelerate growth and innovation. Three of the world’s biggest companies — Netflix, Uber, and Walmart have moved in this general direction, setting up specific plans for the country with forward-looking initiatives  — acquisitions, launching new products\/services (eg: UberEats) and setting up tech labs. Besides being a significant base for tech talent, India also serves as a launchpad for next product innovations, for example, the Uber Lite app built by the Indian team. Netflix Netflix entered the Indian market in January 2016 among competitors such as Amazon Prime Video and Hotstar, vying for a place in the hotly contested Video-on-Demand market. The expansion of the entertainment giant into India did not happen quietly, as they were already entering a market saturated with competitors which offered cheaper plans and, more importantly, cable television. With lofty plans, such as delivering its next 100 million subscribers into the Indian market, Netflix has been dedicating more of its tech resources towards catering to Indian audiences. Experimenting with a mobile-only model The company seems to be experimenting with other price models, owing to the state of Internet access in the subcontinent. Reportedly, a mobile-only plan is in the works for deployment in India, thus allowing the company to “broaden access” by providing a lower price tier. This is already being tested in Malaysia, where the price of mobile-only plan is around half of the lowest tier offered. Reportedly, this test is to gauge “consumer interest in some countries”. Moreover, Hastings has revealed that certain features were added to the app after the Indian launch, such as the addition of the download feature. This is reportedly widely used by the Indian population due to them downloading series and shows on WiFi and utilizing it when on the go, stated the CEO. Double up tech staff count The streaming giant has demonstrated that it has further plans of expansion into the Indian subcontinent, as they have doubled up their technology staff count. With plans in place to expand their manpower and reach in India, Netflix seems to be poised to take over the OTT entertainment market with Indian customers and workers fuelling the ride up. The company also appears to be using the Indian market as a litmus, a foothold, and a base of operations to enter the larger APAC market. Uber Uber already has a track record for attempting to solve mobility and transport problems in the Indian subcontinent, as it conducts a startup mentorship and investment programme known as UberEXCHANGE. So far, the ride-hailing company has mentored over 150 startups, with this step being the initial salvo in building a base of R&D ‘muscle’ in India. The company has 10 R&D centers across the world, and out of the 10,  2 R&D labs are based out of Hyderabad and Bangalore. Apurva Dalal, the head of engineering for Uber India, revealed that Uber is doubling the engineering strength in both these centers in 2019, with the expansion being “in the thousands”. What is most surprising about Uber’s expansion is that the company has been successful in persuading Indians in the US to come back and work in the country. This is due to their drive to find the “best talent” to solve mobility and transportation problems in the subcontinent, thus making it a “laboratory of the world” Tech Innovations The efforts of the labs have paid off with the launch of an app called Uber Lite, an app 85% lighter than the regular Uber app, in July of 2018. This app aims to expand the reach of Uber into emerging markets, where connectivity issues and low-end smartphones reign supreme. After being rolled out in India initially, the company is now expanding the formula for launch in 14 more countries in markets such as Latin America, Egypt, West Asia, and Africa. The company is also testing a “call to ride” service, for use by senior citizens in India. The service is meant for those who do not want to utilize mobile applications, and only want to pay by cash. The reach of Uber into India’s higher administrative circles is also evident, as seen by Dara Khosrowshahi, the CEO of Uber, meeting Jayant Sinha India’s Minister of State for Civil Aviation. Reportedly, this meeting was to speak about the implementation of a flying car collaboration, which the CEO says could materialize in as little time as 5 years. The company’s expansion into the Indian market seems to have paid off, as they observed a 20x surge in revenue over the last financial year. This is seen by the revenue increasing from 1.04 crore in FY2017 to 21.5 crore in FY2018, with profits rising from 3.22 lakh to 19.6 lakhs in the same timeframe. Walmart Walmart’s Indian entry can be well-documented as a rollercoaster ride, from the launch and eventual growth of its wholesale stores in 2007 to allegations of US anti-bribery laws being violated in 2012. Even as the retail giant owns and operates 23 Best Price Modern Wholesale stores in 9 states across India, it has indicated that it might be moving in a different direction, owing to its acquisition of Flipkart, India’s biggest e-tailer. This was due to the Prime Minister of India, Narendra Modi, meeting Doug McMillon, the CEO of Walmart, and asking him to invest in the subcontinent’s farm sector. Acquires Flipkart to fight Amazon In order to combat Amazon on Indian turf, Walmart saw that Flipkart was the most lucrative investment opportunity in India. The investment gave them an edge to fight Amazon on their own grounds, bringing up the argument that Walmart needed Flipkart more than they needed the retail giant. With the acquisition of Flipkart and its manpower, Walmart can provide the e-tailer with access to sellers and manufacturers and know-how on retail matters such as supply chains. However, Walmart has access to the Flipkart name, which holds extended brand value in the subcontinent, along with a fully fleshed out online retail service. Walmart Labs acquires startups to strengthen product ecosystem Penner stated that he viewed Walmart as being a part of the Indian growth story, with the Flipkart acquisition being primary in achieving that. With $16 billion initial investment to Flipkart, Walmart allowed the e-tailer to fight back against global giants such as Amazon and Alibaba. In the same year, Walmart Labs also acquired two startups, Appsfly, a service facilitating business connections, and Int.AI, who are looking to create AI-powered personal data analyst that can give insights into moves that will impact business metrics. The giant also looks at India as a “future growth engine” for the Asia-Pacific market. In accordance with this plan, Walmart Labs the tech arm of Walmart has developed in-store and online shopping a “more seamless experience” for millions of shoppers. This shows that they have recognized the specialized human resources in India, and plan to expand by over 60 new employees in the coming year.","excerpt":"The Indian market has now become the hottest battleground for US-based companies thanks to its large talent pool and yet untapped market. However, the push for additional research and development facilities in the Indian subcontinent might indicate bigger plans for these leading tech behemoths, which are viewing India as a hub for AI talent and […]","categories":["AI Features"],"tags":["AI (Artificial Intelligence)","Uber AI","Walmart Labs"],"author_name":"Anirudh VK","publish_date":"2019-01-03T09:04:05","publication_year":"2019","word_count":1328,"keywords":["Go","AWS","AI","ETL","innovation","ML","Walmart Labs","Aim","ViT","Uber AI","R","AI (Artificial Intelligence)","startup"],"extracted_tech_keywords":["AI","ML","Aim","AWS","R","Go","ETL","ViT","innovation","startup"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/uber-walmart-netflix-look-india-the-next-growth-engine-to-write-their-success-story\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10022920,"title":"Global Chip Shortage To Derail Automotive Industry","content":"Point a finger towards any product lying around you — whether its an electronic gadget in your homes such as laptops, webcam, monitors, or vehicles like cars, buses, and trucks running on the road, there is always a tiny computer fitted inside in the form of a chip that makes them work efficiently. As a matter of fact, the modern world relies on them. However, in the current era, the world is witnessing an alarming trend — a huge shortage of semiconductor chips worldwide. As a result, there is a shortage of products from laptops and mobile phones to cars and gaming consoles. However, the automotive industry remains one of the most hard-hit sectors with negligible demand during the pandemic lockdown and the inability to roll out stocks due to global chip-shortage in the post-COVID scenario. Tough Time For Chip Manufacturers Top semiconductor manufacturing company, Taiwan Semiconductor Manufacturing (TSMC), noted the tremendous pressure on them to meet the market’s huge demand. Moreover, the company has also announced a higher capital expenditure of $28 billion this year, which is an increase of around 60% compared to last year. In another interview, memory chip maker SK Hynix of South Korea is speeding up its plan to relocate its manufacturing unit to China for quick and higher production. The worsening situation can be understood with the fact that Samsung, the world’s largest memory chip maker and second-largest semiconductor-based firm in terms of revenue, has shared concerns over the global chip shortage. The company said that it has to postpone its high-value smartphone due to the non-availability of chips. The global shortage of chip can be attributed to excessive demand in the market, which has arisen out of unavoidable circumstances generated due to the pandemic. The sudden rise in demand for electronic products such as laptops, routers, webcams, Playstations – with work from home as the new normal made an excessive surge in demands for chips by the electronics manufacturers. Manufacturing chips, unlike any other industry, needs a highly controlled environment, along with intricate designing and highly delicate working. Hiring new workers is easy, but setting up a fabrication plant requires a whole lot of different investment. While talking to AIM, Sanjay Gupta, the Vice President and India country manager of a Dutch semiconductor manufacturer — NXP, explained the challenges of establishing a semiconductor wafer fabrication unit in India. Huge investments running in billions of dollars can be a major obstacle in setting up FAB production units, said Gupta. He also spoke about the hundreds of gallons of pure water required to produce a single chip, which can also be challenging to find in India. In another interview with AIM, the corporate vice president of Samsung Semiconductor India R&D Center (SSIR), Balajee Sowrirajan, termed the semiconductor market as a niche one that requires a massive investment of time in order to gain the relevant expertise in making a significant contribution. Thus, with expert views in place, we can easily state that normalcy cannot be expected in chip manufacturing in the near future, as it will take its time. Crisis for Automotive Industry The demand for automobiles fell flat as soon as the lockdown steps in due to the COVID pandemic. With consumers spending more on electronic gadgets, the automotive industry has nothing to do but sit idle. Many automotive industry players have even cancelled their orders of chips in the wake of demand, losing the heat without any appropriate buffer stock. Meanwhile, other industries placed orders in advance. With the removal of lockdown restrictions, there has been a massive surge in demand for vehicles; consequently, the industry finds itself last in the row to grab semiconductors. Global automaker giants, including General Motors, and Ford have slashed their production in not only the US but also in Canada and Mexico. The company is also planning to shut down at least three of its fabrication plants. While talking to the media, a company spokesman stated that despite their mitigation efforts, the semiconductor shortage would impact GM production in 2021. Sweden’s truck maker Volvo has also seen a sharp fall in its valuation amid this crisis. Similarly, other big players, including Ford, Nissan, and Honda, were forced to shut their productions. As far as the anticipation is concerned, this huge crisis of global chip shortage will delay the production of nearly one million vehicles in the first quarter of the year 2021, as per a report. Now, the automotive industry is left with no other options but to slash production, wait until the situation improves, and look out for alternatives in the near future. Wrapping Up Although the situation is expected to improve in the coming months, it has brought out many concerns before the industry over the supply-chain at the forefront. Diversifying supply chains, scaling up production, and maintaining a good buffer are some of the lessons learned by the automotive industry. The problem comes along with an opportunity to learn from past mistakes and work to ensure a better future ahead.","excerpt":"The shortage of semiconductor chips has raised questions over the supply chain and hit the automotive industry hard.","categories":["AI Features"],"tags":["AI Chips","automotive industry","Chip Manufacturing","Chip shortage","global chip shortage","semiconductor chips","Semiconductor India","Semiconductor market in india","semiconductors","Supply chain"],"author_name":"kumar Gandharv","publish_date":"2021-03-26T13:00:00","publication_year":"2021","word_count":836,"keywords":["Go","Supply chain","API","programming_languages:R","global chip shortage","semiconductors","AI","R","Semiconductor market in india","programming_languages:Go","semiconductor chips","AI Chips","Aim","Chip shortage","Semiconductor India","automotive industry","Chip Manufacturing"],"extracted_tech_keywords":["AI","Aim","R","Go","API","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/global-chip-shortage-to-derail-automotive-industry-off-track\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10053595,"title":"Data Management Platform Informatica Plans To Raise Nearly $1 Billion In IPO","content":"One of the leading developers of cloud data management software, Informatica, expects to raise nearly $1 billion when it returns to the public market and plans to use the amount to pay down some of its debt. The Redwood City, California-based firm priced its initial public offering at $29.00 per share, at the low end of its target range of $29.00 to $32.00 per share, giving the company a valuation of $10 billion. The company’s shares will trade on the New York Stock Exchange under the symbol INFA. The company reported net debt of $2.77 billion as of June, according to a prospectus that it filed to the Securities and Exchange Commission earlier this month. Its current leverage ratio stands at 5.5 times net debt to adjusted earnings before interest, taxes, depreciation and amortization, or Ebitda, set to come down to 3.9 times after the IPO. Chief Financial Officer Eric Brown said that the company would use the roughly $900 million in expected proceeds to bring down its debt. Adjusted Ebitda for the first six months of this year was $175 million compared with $168 million in the first six months of 2020. Informatica expects to use some of the cash it generates to reduce its debt ratio to approximately two times adjusted earnings during the next two to three years, Brown added. Informatica joins a host of tech-industry companies to go public this year. US technology companies raised $64.4 billion in 109 IPOs through October 2021. The company claims it has invested more than $1 billion in research and development since 2015 in an effort to move its products to a cloud-based platform and to shift its business model towards one based largely on subscriptions. Informatica, which lists drugmaker Eli Lilly & Co., consumer-goods giant Unilever PLC and supermarket chain Kroger Co. among its customers, helps companies connect and manage their data across the cloud and on-premise systems, allowing organizations to better analyze the data they collect.","excerpt":"Informatica expects to use some of the cash it generates to reduce its debt ratio to approximately two times adjusted earnings during the next two to three years.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","Cloud Computing","Cloud Computing in IT Industry","Cloud Data AI","Data Management","Data Science","informatica","IPO","Machine Learning"],"author_name":"Victor Dey","publish_date":"2021-11-16T15:31:51","publication_year":"2021","word_count":328,"keywords":["Go","informatica","programming_languages:R","AI","IPO","R","Cloud Computing in IT Industry","Machine Learning","programming_languages:Go","RAG","GAN","Aim","Cloud Computing","Data Management","Data Science","Cloud Data AI","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","Aim","RAG","R","Go","GAN","IPO","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/data-management-platform-informatica-plans-to-raise-nearly-1-billion-in-ipo\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10091741,"title":"Why is ETL Dying?","content":"Despite its long-standing popularity as a method of managing and integrating data, organisations are now opting to move away from ETL (Extract, Transform, Load) for various reasons. The continuous evolution of AI leading to the birth of AI agents like AutoGPT and AgentGPT that can run autonomously and execute tasks for you has shrunk the need for ETL. While ELT is often used in conjunction with data warehousing technology, organisations are turning to more flexible models, such as unstructured data lakes, to store and analyse their data. However, the traditional ETL pipeline has also undergone significant transformations over the past few decades to keep pace with evolving data requirements and analytics use cases. Prominent ELT tools include AWS Glue, Informatica PowerCenter, Microsoft – SQL Server Integrated Services (SSIS) and so on. Is ETL Still Relevant? One of the most significant changes in ETL tools is the shift from their original purpose of automating the retrieval and cleansing of well-structured data from various operational systems or databases to catering to diverse users’ varying data requirements. With data lakes becoming a popular destination for data, replacing data warehouses, and the data itself becoming more massive and complicated, ETL pipelines must now provide quick cleansing and transformation to meet the demands of modern analytics use cases. However, traditional ETL pipelines have faced difficulties in supporting the agility required by modern analytics use cases, leaving business users waiting in line for their desired results. As a result, ETL pipelines are often viewed as a hindrance to better performance and businesses must carefully assess their current role and explore how they can be optimally leveraged in the contemporary analytics landscape. Traditional ETL processes require moving large amounts of data across various stages and systems, making them slow, demanding on resources, and prone to errors. This can be challenging for modern data-driven businesses to manage, as traditional ETL tools often come with a high price tag and demand substantial investments in hardware, software, and personnel resources. In contrast, newer data platforms present pre-built services and extensions that can lessen these expenses and enable enterprises to concentrate on providing meaningful outcomes to their users. For example, Google Data Stream is an instance of this approach, which is capable of managing real-time CDC with minimal coding or setup. Moreover, as ETL processes are inadequate for providing real-time data insights, it makes it difficult to keep up with the changing business needs or evolving data sources. To overcome this, ELT or Zero ETL approaches are used, where data is loaded without transformation or directly collected or queried from the source system, with automatic detection and handling of schema changes. One of the main advantages of modern data platforms is the ability to provide self-service data preparation in addition to automated data integration and zero ETL. This feature empowers users to access and modify data with ease, eliminating the need for complicated ETL processes, thereby increasing their ability to analyse and explore data effectively. Following the employment of zero ETL for data integration, users can utilise various techniques and tools such as SQL in the data warehouse or other business intelligence tools to perform data preparation and transformation, which includes rectifying, adjusting, and enhancing raw data. These methods enable businesses to gain deeper insights into their data, making it easier to identify patterns, trends, and other critical information that can help form strategic decisions. The Emergence of AI Agents The growth of AI agents, often labelled the “next big thing in AI”, have simplified data processing and analysis by automating mundane and time-consuming tasks associated with ETL processes. They can now operate independently, perceive their surroundings, and can swiftly identify and fix data errors transforming them into a more accessible format for analysis, eliminating the need for manual data cleansing and transformation. Moreover, AI agents can accurately analyse large data sets faster, identifying patterns and trends that may not be readily apparent through manual analysis. This capability can prove highly beneficial for organisations, enabling them to make well-informed decisions quickly and derive business value. Read more: Crossing Treacherous Boundaries with AI Agents","excerpt":"With the evolution of AI, ETL is losing its charm among enterprises","categories":["AI Features"],"tags":["AI (Artificial Intelligence)","AutoGPT","data cleansing","data processing","ETL","Google","OpenAI"],"author_name":"Shritama Saha","publish_date":"2023-04-18T17:00:00","publication_year":"2023","word_count":682,"keywords":["Go","ELT","OpenAI","ETL","AI","data processing","AWS","RAG","data cleansing","analytics","Google","SQL","R","AI (Artificial Intelligence)","AutoGPT"],"extracted_tech_keywords":["AI","analytics","AutoGPT","RAG","AWS","R","SQL","Go","ETL","ELT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/why-is-etl-dying\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":1186,"title":"Security Technologies behind SSL","content":"Since the early age, computers have been used to transmit confidential and sensitive messages. But, sometimes people intercept and use these messages for their gain. Therefore, to safeguard the important messages such as credit\/debit card information, different methods of encryption have been implemented. Cryptography or Cryptology is the study and practice of techniques for secure communication in the presence of third parties called adversaries. In general, cryptography is about constructing and analyzing protocols that prevent adversaries or the public from reading private messages. Symmetric Encryption: Symmetric cryptography is a cryptographic system which uses a single key to encrypt and decrypt data. Both the sender and receiver use the same key to communicate. However, symmetric keys also have a disadvantage. As both the sender and receiver use one key to encrypt or decrypt, sharing the key to each other is difficult. If they have to share the key through the internet, chances are there that a hacker can intercept the key. Public\/Asymmetric-key cryptography: Public-key cryptography, or asymmetric cryptography, is a cryptographic system that involves pairs of keys; public keys can be shared widely pairing with private keys that are known only to the owner. In other words, in a public-key encryption system, anyone can encrypt a message using the public key of the receiver. But, the message can be decrypted only with the receiver’s private key. Example: John wants to send a secret message to Jane, So he will encrypt these message with a public key (generated by Jane using a key generation program whose input is a large random number and whose output is one public and one private key. The private key secret and is kept by Jane and the public key is spread widely to the public) and sends this message to the Jane even if the message is captured, it cannot be decoded without the private key. Disadvantages of public-key encryption 1. The public-key encryption methods are several orders of magnitude slower than the best known symmetric-key schemes. 2. Key sizes are usually larger than those required for symmetric key encryption. The size of public-key signatures is larger than that of tags providing data origin authentication from symmetric-key techniques. 3. No public-key scheme is proven secured. The most effective public-key encryption schemes have their security based on the set of number – theoretic problems. 4. Public-key cryptography does not have a history of symmetric-key encryption. Which Is Stronger? Both the symmetric and asymmetric encryptions are stronger. When we consider in terms of computational burden and ease of distribution, symmetric encryption requires less computational burden whereas asymmetric encryption involves with ease of distribution. Digital Certificate: Digital certificate is the electronic format of physical or paper certificates such as passport, membership card, driving license, etc. It proves your identity or the right to access services or information on the internet. Digital certificates are issued by a trusted authority empowered by law, known as Certifying Authority (CA). Public Key Infrastructures: A PKI-based authentication uses hybrid cryptosystem and benefits from using both types of encryption. Steps Involved in SSL Authentication Protocol A client broker requests a secure page (SSL Hello) The web server sends its public key with its certificate The browser checks whether that certificate was issued by a trusted party (CA), valid or not, and relation to the site contacted The browser creates a symmetric session key and encrypts it with the server’s asymmetric public key. Then sends it to the server. By using the asymmetric private key, the server decrypts the encrypted session and gets the symmetric session key. Server and Browser now encrypt and decrypt all transmitted data with the symmetric session key. This allows for a secure channel because only the server and browser know the symmetric session key, which can only be used for that session. If the browser has to connect to the same server the next day, a new session key would be created. Applications of SSL: 1. SSL-secured transactions with e-commerce Web site: It is a typical use case of SSL transaction between a browser and a Web server where the protocol is used to authenticate if the server and then pass the customer’s credit\/debit card details to the server. 2. Authenticated client access to an SSL-secured Web site: Both the client and server need certificates from a trusted certification authority (CA) that they both trust. 3. Remote access: SSL technology is used to provide authentication and data protection for users who want to log into their system (computer) remotely. 4. E-mail: The security protocol is used to transmit private communications via the Internet. Conclusion: Communication using SSL-based encryption and authentication is highly secure with little to no chance that the communication can be decrypted by a hacker thus making software’s\/websites highly secure and trustworthy. For More details please visit our blogs: https:\/\/www.bootcamplab.com\/blog\/","excerpt":"Since the early age, computers have been used to transmit confidential and sensitive messages. But, sometimes people intercept and use these messages for their gain. Therefore, to safeguard the important messages such as credit\/debit card information, different methods of encryption have been implemented. Cryptography or Cryptology is the study and practice of techniques for secure […]","categories":["AI Features"],"tags":[],"author_name":"AIM Media House","publish_date":"2017-01-10T07:18:37","publication_year":"2017","word_count":801,"keywords":["TPU","programming_languages:R","AI","Git","Rust","R","programming_languages:Rust"],"extracted_tech_keywords":["AI","TPU","R","Rust","Git","programming_languages:R","programming_languages:Rust"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/security-technologies-behind-ssl\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10103635,"title":"NVIDIA Emerges As the Real Winner in the OpenAI Fiasco","content":"While OpenAI top leadership dealt with its boardroom drama, it was business as usual for the rest of the team. The servers kept buzzing with requests from millions of users and developers, and the team even shipped a new update for ChatGPT, making the voice-feature available to all. In other words, more servers and a whole lot of love from NVIDIA. “ChatGPT, Microsoft 365 Copilot, CoAssit… are (all) built and run on NVIDIA,” said Colette Kress, chief financial officer at NVIDIA, at a recent earnings call (FY24 Q3). He said that NVIDIA HGX with InfiniBand are essentially the reference architecture for AI supercomputers and data centre infrastructures. The result: No surprise, NVIDIA had another record-breaking quarter. Its revenue surged astronomically, reaching $18.1 billion, a 206% increase driven by substantial gains in data centre. NVIDIA’s data centre compute revenue quadrupled from last year and networking revenue nearly tripled. A glimpse into NVIDIA’s record-breaking earnings In FY24 Q3, NVIDIA’s data centre division achieved remarkable success, setting a sales record at $14.51 billion—a remarkable 279% YoY increase and a robust 41% sequential growth. The surge was fuelled by global demand for recommendation engines, generative AI applications, and large-scale language model training, driving the adoption of the NVIDIA HGX platform. Half of this revenue came from cloud infrastructure providers like Amazon, while the rest originated from consumer internet entities and large corporations. NVIDIA reported a net income of $9.2 billion with a diluted EPS of $3.71 per share, marking a remarkable 1,274% increase compared to the same period last year. Since May, NVIDIA’s ambitious forecasts have propelled its market valuation beyond $1 trillion, underscoring its unrivalled financial gains in the race for AI opportunities. Despite a decline in sales of Ampere GPU architecture-based data center products, the innovative Hopper GPU architecture-based HGX platform gained substantial traction, particularly among cloud service providers (CSPs), consumer internet firms, and corporations. Consequently, NVIDIA anticipates sales of approximately $20 billion for the current quarter, surpassing analysts’ average estimates of just under $18 billion. NVIDIA’s gaming revenue also surged impressively, with a 15% sequential rise and a substantial 81% YoY gain, driven by a heightened demand for the GeForce RTX 40 Series GPUs during the holiday season and back-to-school sales. Professional visualization revenue also soared, marking a 10% sequential increase and an outstanding 108% YoY growth. The boost was attributed to increased demand for corporate workstations and the successful introduction of laptop workstations utilizing the Ada Lovelace GPU architecture. NVIDIA’s stock also had to face a decline despite robust sales due to business losses in China, a consequence of recent US trade restrictions on advanced semiconductor technology. Low Expectations from China “Our sales to China and other affected destinations derived from products that are now subject to licensing requirements have consistently contributed approximately 20% to 25% of data center revenue over the past few quarters,” said Kress, adding that their sales to these destinations are expected to decline significantly in the fourth quarter. Despite potential challenges from new US export restrictions on China, NVIDIA remains resilient, expecting increased demand from other global markets to counterbalance any impact in the upcoming quarter. This underscores NVIDIA’s adaptability and strategic agility in navigating complex market conditions. NVIDIA also anticipates positive contributions from other income and expenses by connecting with new projects and developments and aims to project an income of approximately $200 million, excluding non-affiliated investment gains and losses. The company targets a 15.0% tax rate with a margin of plus or minus 1%, excluding discrete items. Overall, NVIDIA eagerly anticipates a robust performance in the upcoming fiscal year’s fourth quarter, projecting revenue of $20.00 billion with a slight margin of variability of 2%. “Our strong growth reflects the broad industry platform transition from general-purpose to accelerated computing and generative AI. The first movers are large language model startups, consumer internet companies, and global cloud service providers. The next waves are starting to build,” said NVIDIA chief Jensen Huang.","excerpt":"While the OpenAI drama was unfolding, NVIDIA’s revenue surged astronomically, reaching $18.1 billion, a 206% increase driven by substantial gains in data centre.","categories":["Global Tech"],"tags":[],"author_name":"Sandhra Jayan","publish_date":"2023-11-24T14:10:12","publication_year":"2023","word_count":656,"keywords":["ChatGPT","OpenAI","AI","RPA","RAG","GPT","Aim","generative AI","R","startup"],"extracted_tech_keywords":["AI","generative AI","ChatGPT","OpenAI","Aim","RAG","R","GPT","RPA","startup"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/nvidia-emerges-as-the-real-winner-in-the-openai-fiasco\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10135099,"title":"DeepLearning.AI Launches ‘Multimodal RAG: Chat with Videos’ Course with Intel","content":"DeepLearning.AI recently introduced a new course, ‘Multimodal RAG: Chat with Videos’ offered by Andrew Ng, founder and CEO in collaboration with Intel Corporation, and taught by Vasudev Lal, Principal AI Research Scientist at Intel Labs. Focusing on building a system that answers grounded responses from video content. In this course, participants will create an interactive chat system using the BridgeTower model, a multimodal transformer developed by Intel and Microsoft Research. Click here to check out the course. This course aims at helping participants generate joint embeddings from video content and store them in a vector database building a Retrieval-Augmented Generation (RAG) pipeline to fetch relevant video data using Large Vision-Language Models (LVLMs) to answer questions using both text and image inputs. By the end of the course, participants will be able to build systems that can intelligently interact with video content. Participants will also make API calls to access multimodal models hosted by Prediction Guard on Intel’s cloud. Andrew Ng, has recently joined the Korea’s National AI Committee as advisors. His recent appointment to the National AI Committee comes on the heels of a series of educational initiatives launched through DeepLearning.AI. Last month, DeepLearning unveiled a course on federated learning, allowing secure training on private data, alongside a partnership with Flower Labs. DeepLearning.AI have also expanded their generative AI course offerings to cater to learners at various skill levels. Courses range from foundational topics like LLMs and diffusion models to practical applications such as building chatbots and automating workflows with ChatGPT.","excerpt":"Participants will now be able to create an interactive chat system using the BridgeTower model, a multimodal transformer developed by Intel and Microsoft Research.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","AI Video Generation Models","Courses","Deep Learning","Generative Pre-Trained Transformer"],"author_name":"Tanisha Bhattacharjee","publish_date":"2024-09-12T15:42:13","publication_year":"2024","word_count":251,"keywords":["federated learning","ChatGPT","API","AI","chatbots","AI Video Generation Models","RAG","GPT","Aim","Generative Pre-Trained Transformer","generative AI","Deep Learning","Courses","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","generative AI","ChatGPT","Aim","federated learning","RAG","chatbots","R","API","GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/deeplearning-ai-launches-multimodal-rag-chat-with-videos-course-with-intel\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":38819,"title":"5 Companies With Latest Openings For Data Science Positions In India","content":"Photo by Hunters Race on Unsplash Analytics India Jobs offers a new platform for the data science community in India by bringing to the fore latest job openings across the country. Here are the 5 most recent Data Science job openings across major cities in India : Business Analyst @ Tredence, Bangalore Tredence is the first analytics services company focused on the last mile of analytics adoption. Requirements: At least 2 years of work experience Proficiency in SQL\/Hive [must have] Handling and analysing large data Data cleaning and processing R (Basic data manipulation, DPLYR, GGplot, Random forest) Python (Basic data manipulation using pandas, matplotlib, sklearn) Hadoop, spark Tableau Apply here Data Engineer @ Hugo Edge Solutions, Bangalore As a data engineer, the candidate will analyze large amounts of information to discover trends and patterns, present information using data visualization techniques and design, develop and launch extremely efficient data pipelines to move data. Requirements: 4-6 years of IT Development Experience Data engineering chops with big data technologies (AWS, Hadoop) and the drive to improve data quality, accuracy, and completeness Proficiency with handling structured and unstructured data sets, writing complex queries, and the occasional stored procedure Ability to design, build, and query data warehouses (MySQL, PostgreSQL, Oracle, MS SQL, Redshift) Experience in building or improving large scale data infrastructure from the ground up – Ever hear of Snowflake? AWS Certified Solution Architect or Developer Associate required; Professional certification preferred Experience with Agile and the Enterprise SDLC (functional and unit testing, etc) Apply here AI Architect @ ZS, Pune ZS‘s Software Development Team designs, implements, tests and supports high quality products used by hundreds of companies and thousands of end users to make critical decisions and manage sales operations. Requirements: Knowledge and experience in some of the key AI platform, e.g. AWS Sagemaker, IBM Watson, Microsoft Azure, Google Api.Ai, Facebook Wit.Ai, Chatbots using Microsoft Bot Framework Serve AI\/ML models in enterprise grade technology platforms through microservices Tensorflow, Caffe, CNTK, commercial technologies\/platforms, etc Experience working in a DevOps environment, and using industry standard tools (GIT, JIRA, Teamcity, etc.). Able to explain technical concepts in a non-technical language Solid hands-on experience in the Artificial Intelligence platforms with understanding of the end to end life cycle of AI projects. Very solid proven hands-on experience on UI technologies (AngularJS, ReactJS, StencilJS etc), Java\/ J2EE, Kubernetes, Spring framework, Microservices, REST API’s, Kafka etc. Exposure in Data Governance and management Exposure to Hadoop data science workbench solutions Preferred experience in any one of the integrated AI products like: Microsoft Azure ML, AWS ML Solid experience on any one of the ML pipeline solutions like DataiKU, Anaconda, KNIME and auto ML solutions like H2O.ai or DataRobot or Firefly.ai Apply here Big Data Engineer @ Byte Prophecy Pvt. Ltd, Ahmedabad The candidate will be designing databases and data pipelines for storing and processing large, sometimes-unstructured data-sets for use with our analytics platform. And, executing batch jobs on our custom-built computing cluster or any standard ETL tools or using custom code in SQL or Java or Python. Requirements: Experience with both SQL (MySQL) and Columnar (MariaDB\/InfiniDB), noSQL (Cassandra) databases Familiarity with programming best practices, design patterns, version control systems A sound understanding of parallel\/distributed programming Good command in Java or Python The ability to work effectively with people from a variety of backgrounds Apply here Senior Data Scientist @ Volvo Group, Bangalore At Volvo,  a Data Scientist would be expected to understand business requirements, KPIs and convert into analytical hypothesis in a structured and logical manner along with solution identification. Work with stakeholders throughout the organization to identify opportunities for leveraging company data to drive business solutions. Requirements: Experience using statistical computer languages (Python, TensorFlow, SQL, etc.)  to manipulate data and draw insights from large datasets. Strong conceptual understanding of machine learning algorithms including linear regression, logistic regression, decision trees, random forest, topic model etc. Experience working with and creating data architectures. Knowledge of a variety of machine learning techniques(regression, classification, clustering, time series, neural network, etc.) Knowledge in GLM\/Regression, Random Forest, Boosting, Tress, Text mining, social network analysis, etc. Exposure to visualization tools specially Qlik Sense is a plus. Exposure to Analytics Platforms especially to Azure is a plus. Apply here","excerpt":"Analytics India Jobs offers a new platform for the data science community in India by bringing to the fore latest job openings across the country. Here are the 5 most recent Data Science job openings across major cities in India : Business Analyst @ Tredence, Bangalore Tredence is the first analytics services company focused on […]","categories":["AI Trends"],"tags":["Data Science Career","Python","Snowflake","Tableau"],"author_name":"Ram Sagar","publish_date":"2019-05-08T12:26:25","publication_year":"2019","word_count":700,"keywords":["data science","Tableau","artificial intelligence","machine learning","AI","neural network","ML","Data Science Career","Python","analytics","TensorFlow","Pandas","Azure ML","Snowflake"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","neural network","data science","analytics","Azure ML","TensorFlow","Pandas"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/5-companies-with-latest-openings-for-data-science-positions-in-india\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10100408,"title":"Larry Lulls Everyone into Generative AI","content":"“Generative AI showed up about a year ago, and it changed everything, and it is definitely changing everything at Oracle,” said Larry Ellison, CTO of Oracle, in his keynote at Oracle CloudWorld 2023. The conference and the keynote address was proof that Oracle is optimistic about generative AI and the future it holds, instead of the fear mongering that a lot of tech companies have been pushing around. The reason is simple — Oracle is going all in on generative AI through its cloud platform. As Ellison further mentioned, “Cloud should be open”, highlighting how all of its customers are already leveraging a multi-cloud approach. Arguably, it is quite a gamble for Oracle. Regardless, Ellison says that generative AI is “probably” the most-important new computer technology ever. “Billions of dollars are already being invested in the field. Every company is ready to experiment and implement the technology within its working.” Making generative AI for everyone Oracle aims to build the largest scientific supercomputer partnering with NVIDIA. It would be the biggest supercluster of NVIDIA’s H100s, or probably the upcoming GH100s. For both the companies, generative AI is a revolution. “It is a breakthrough. It’s transformational. It’s fundamentally changing things at Oracle…this makes AI central to almost everything we’re doing.” On NVIDIA’s part, it is clear that it is on the side of AI being safe, and not the doomsday that science fiction has been promoting, which even has the creators of AI worried. Bill Dally, NVIDIA’s chief scientist testified before the US Senate, “Fortunately, uncontrollable artificial general intelligence is science fiction, not reality. At its core, AI is a software program, which is limited to its training, the inputs provided to it, and nature of its output.” OpenAI: AI will kill us.Anthropic: AI will kill us.InflectionAI: AI will kill us.Nvidia: pic.twitter.com\/kDUkBBG9wa— Yam Peleg (@Yampeleg) September 18, 2023 “Humans will always decide how much power they will cede to AI models,” said Dally. In his keynote, Ellison revealed how Elon Musk’s xAI is also training on Oracle’s Cloud. The self-driving cars that he mentioned in his talk were the ones that Tesla is developing. “They won’t crash as much as the human driven cars do.” Moreover, he also mentions OpenAI as the company founded by Musk. It is clear that Ellison is on the side of what Musk is trying to do. They are even ready to build a Tesla Police car. “A lot of customers are not going to build what xAI, OpenAI, or Cohere are building using our cloud,” said Ellison. Most of the companies would use what these first-movers have already built, building their own specialised models. That is what Oracle is going to enable through its cloud. “A lot of this is going to focus on healthcare,” which was clearly visible through Ellison’s address. Addressing the fears around AI “That does not mean that AI is going to take over the doctor’s job”, it is just going to make his job easier. Addressing this, Ellison also spoke about how privacy was one of the key focuses of all the generative AI vendors in the market, highlighting that generative AI is going to be safe, and in fact already is. Oracle is also building autonomous drones for addressing forest fires and mapping out easier routes for firefighters. “Automation is much more reliable.” Oracle is focusing on a lot of automation of its tasks by leveraging generative AI to write applications faster. He also added that the teams are already getting a lot smaller. “Does that mean we are going to have massive layoffs at Oracle? No, we are too ambitious for that. We are going to do a lot more,” assured Ellison. “In the Oracle Cloud, you can use your training data to train your private model with your private data that remains private after training,” Ellison highlights how if users want to use their own data, instead of relying on Microsoft or Oracle ones, they can easily do it, without being concerned about data breach and privacy. Moreover, Ellison highlighted the company’s dedication towards using AI sustainably and responsibility, and for responsible reasons. Oracle is focusing on integrating its cloud within Microsoft’s Azure cloud so that there is no latency and it also becomes more sustainable. In the AI doomsday commotion, Ellison took a very different approach, addressing the apprehension and highlighting the benefits of generative AI. He was able to ease into the audience that generative AI is here to stay, and for all the good reasons.","excerpt":"“Does that mean we are going to have massive layoffs at Oracle? No, we are too ambitious for that,” said Larry Ellison","categories":["AI Features"],"tags":["Interviews and Discussions"],"author_name":"Mohit Pandey","publish_date":"2023-09-21T13:00:00","publication_year":"2023","word_count":751,"keywords":["Anthropic","TPU","OpenAI","AI","Azure","R","RAG","Aim","generative AI","xAI","Interviews and Discussions"],"extracted_tech_keywords":["AI","generative AI","OpenAI","Anthropic","xAI","Aim","RAG","Azure","TPU","R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/larry-lulls-everyone-into-generative-ai\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10052705,"title":"Google Now Corrects Grammar As You Type On Pixel 6","content":"Google has launched a grammar correction feature that will come directly built into the Gboard on Pixel 6. The tech giant said that this feature would detect and suggest corrections for grammatical errors while the user is typing. At the moment, this feature will work on rectifying English sentences, but Google has plans to expand to more languages soon. Previously, Google had made improvements to correct language errors in Google Docs by using Neural Grammar Correction in Docs. Correcting grammatical errors while typing on smartphones can be cumbersome on a small display with limited controls. Learn about the new grammar correction feature built into Gboard on Pixel 6 that works entirely on-device as you type → https:\/\/t.co\/QOIVQgvvdR pic.twitter.com\/0rNwTK0o5n— Google AI (@GoogleAI) October 28, 2021 Google AI has been introducing different innovations using neural networks and machine learning recently. It released a new method called Task Affinity Groupings (TAG) that tells you which tasks should be trained together in multi-task neural networks. It also came out with a new differentially private clustering algorithm that generates new representative data points privately. Hybrid Architecture Google trained a  sequence-to-sequence neural network to take an input sentence or a sentence prefix) and output the grammatically correct version. If the original text is already grammatically correct, the output of the model is identical to its input which shows that no corrections are needed. What does the model use? The model uses a hybrid architecture with a combination of a Transformer encoder with an LSTM decoder. Transformer encoder: It is a neural network architecture based on a self-attention mechanism. Though recurrent neural networks (RNNs) have been the common network architecture for translation, processing language sequentially, they conduct multiple steps to make decisions that depend on words far away from each other. Due to its sequential nature, it makes it difficult to take advantage of fast computing devices. Transformer, on the other hand, performs a small, constant number of steps that are selected empirically. During each step, it applies a self-attention mechanism that directly models relationships between all words in a sentence, regardless of their respective position. LSTM decoder: Long short-term memory (LSTM) is a recurrent neural network architecture with feedback connections with the capability of processing entire sequences of data. LSTM units usually consist of a cell, an input gate, an output gate, and a forget gate. It finds applications in unsegmented connected handwriting recognition, speech recognition, etc. Image: Google (Overview of the grammatical error correction (GEC) model architecture) Google used techniques such as shared embedding, factorized embedding and quantization to solve the problem of limited memory and computational power. Reducing file size Shared embedding Some of the model weights are shared between the Transformer encoder and the LSTM decoder. This leads to a reduction in model file size without impacting accuracy. Factorized embedding The model splits a sentence into a sequence of predefined tokens, which is needed to achieve good quality. But this increases the model size. A factorized embedding separates the size of the hidden layers from the size of the vocabulary embedding. Quantization Google performs post-training quantization allowing it to store each 32-bit floating-point weight using only 8-bits. Each weight is stored with lower fidelity, but the quality of the model is not affected. Hard distillation Google needed training data in the form of <original, corrected> text pairs and found hard distillation to generate training data better matched to the on-device domain yields. How does hard distillation work? Google collected hundreds of millions of English sentences from across the public web.Used the cloud-based grammar model to generate grammar corrections for those sentencesThe training dataset of <original, corrected> sentence pairs is used to train a smaller on-device model that can correct full sentences. Google found out that the on-device model built from this training dataset produces significantly higher quality suggestions than a similar-sized on-device model built on the original data used to train the cloud-based model. Handling prefixes Even before training the model from the data, the model has to be able to handle sentence prefixes. It is even more needed in messaging apps, where the user often omits the final period in a sentence and presses the send button as soon as they finish typing. Google used heuristics to solve this issue. This means that if a given sentence prefix can be completed to form a grammatically correct sentence, it is considered grammatically correct and incorrect, if not. It created a second dataset suitable for training a large cloud-based model with a focus on sentence prefixes. By the heuristic previously mentioned, Google generated data by using the <original, corrected> sentence pairs from the cloud-based model’s training dataset and randomly sampling aligned prefixes from them. Image: Google Then, it autocompletes each original prefix to a full sentence using a neural language model. If a full-sentence grammar model finds no errors in the full sentence, that means there is at least one possible way to complete this original prefix without making any grammatical errors. Then, the original prefix to be correct and output <original prefix, original prefix> is taken as a training example. Else, <original prefix, corrected prefix> is the output.The training data is used to train a large cloud-based model that can correct sentence prefixes, then uses that model for hard distillation, generating new <original, corrected> sentence prefix pairs that are better matched to the on-device domain.By combining the new sentence prefix pairs with the full sentence pairs, the final training data for the on-device model is built. It has the capability of correcting both full sentences as well as sentence prefixes. Image: Google (Training data for the on-device model is generated from cloud-based models) What happens when the user types? When the mobile user has typed more than three words, Gboard sends a request to the on-device grammar model. Google underlines the grammar mistakes and provides replacement suggestions. As the model outputs only corrected sentences, the mistakes need to be changed into replacement suggestions.Google syncs the original sentence and the corrected sentence by minimizing the Levenshtein distance. This is the number of edits that are needed to transform the original sentence to the corrected sentence. In the end, the method transforms the insertion edits and deletion edits to be replacement edits.","excerpt":"Google has come out with a grammar correction feature that will come directly built into the Gboard on Pixel 6","categories":["Global Tech"],"tags":["Google","Neural Networks"],"author_name":"Sreejani Bhattacharyya","publish_date":"2021-11-01T11:00:00","publication_year":"2021","word_count":1037,"keywords":["Go","machine learning","TPU","attention mechanism","AI","neural network","ML","Google","RNN","LSTM","R","Neural Networks"],"extracted_tech_keywords":["AI","machine learning","ML","neural network","TPU","R","Go","attention mechanism","RNN","LSTM"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/google-now-corrects-grammar-as-you-type-on-pixel-6\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10052239,"title":"How to Measure the Success of a Recommendation System?","content":"Recommender systems are used in a variety of domains, from e-commerce to social media to offer personalized recommendations to customers. The benefit of recommendations for customers, such as reduced information overload, has been a hot topic of research. However, it’s unclear how and to what extent recommender systems produce commercial value. It’s challenging to create a reliable product suggestion system. However, defining what it means to be reliable is also a challenging task. Measuring the success of any recommender system is very necessary from a business point of view. In this post, we go through the most important and popularly used evaluation parameters to measure the success of a recommendation system. The major points to be discussed in this article are outlined below. Table of Contents Challenges Faced by Recommendation SystemsCommon metrics usedBusiness Specific Measures Let’s start the discussion by understanding the challenges that are faced by recommendation systems. Challenges Faced by Recommendation Systems Any predictive model or recommendation systems with no exception rely heavily on data. They make reliable recommendations based on the facts that they have. It’s only natural that the finest recommender systems come from organizations with large volumes of data, such as Google, Amazon, Netflix, or Spotify. To detect commonalities and suggest items, good recommender systems evaluate item data and client behavioral data. Machine learning thrives on data; the more data the system has, the better the results will be. Data is constantly changing, as are user preferences, and your business is constantly changing. That’s a lot of new information. Will your algorithm be able to keep up with the changes? Of course, real-time recommendations based on the most recent data are possible, but they are also more difficult to maintain. Batch processing, on the other hand, is easier to manage but does not reflect recent data changes. The recommender system should continue to improve as time goes on. Machine learning techniques assist the system in “learning” the patterns, but the system still requires instruction to give appropriate results. You must improve it and ensure that whatever adjustments you make continue to move you closer to your business goal. Common Metrics Used Predictive accuracy metrics, classification accuracy metrics, rank accuracy metrics, and non-accuracy measurements are the four major types of evaluation metrics for recommender systems. Predictive Accuracy Metrics Predictive accuracy or rating prediction measures address the subject of how near a recommender’s estimated ratings are to genuine user ratings. This sort of measure is widely used for evaluating non-binary ratings. It is best suited for usage scenarios in which accurate prediction of ratings for all products is critical. Mean Absolute Error (MAE), Mean Squared Error (MSE), Root Mean Squared Error (RMSE), and Normalized Mean Absolute Error (NMAE) are the most important measures for this purpose. In comparison to the MAE metric, MSE and RMSE employ squared deviations and consequently emphasize bigger errors. The error is described by MAE and RMSE in the same units as the obtained data, whereas MSE produces squared units. To make results comparable among recommenders with different rating scales, NMAE normalizes the MAE measure to the range of the appropriate rating scale. In the Netflix competition, the RMSE measure was utilized to determine the improvement in comparison to the Cinematch algorithm, as well as the prize winner. Classification Accuracy Metrics Classification accuracy measures attempt to evaluate a recommendation algorithm’s successful decision-making capacity (SDMC). They are useful for user tasks such as identifying nice products since they assess the number of right and wrong classifications as relevant or irrelevant things generated by the recommender system. The exact rating or ranking of objects is ignored by SDMC measures, which simply quantify correct or erroneous classification. This type of measure is particularly well suited to e-commerce systems that attempt to persuade users to take certain actions, such as purchasing products or services. Rank Accuracy Metrics In statistics, a rank accuracy or ranking prediction metric assesses a recommender’s ability to estimate the correct order of items based on the user’s preferences, which is known as rank correlation measurement. As a result, if the user is given a long, sorted list of goods that are recommended to him, this type of measure is most appropriate. The relative ordering of preference values is used in a rank prediction metric, which is independent of the exact values assessed by a recommender. A recommender that consistently overestimates item ratings to be lower than genuine user preferences, for example, might still get a perfect score as long as the ranking is correct. Mean Average Precision @ K and Mean Average Recall @ K For each user in the test set, a recommender system normally generates an ordered list of recommendations. MAP@K indicates how relevant the list of recommended items is, whereas MAR@K indicates how well the recommender can recall all of the items in the test set that the user has rated positively. Business Specific Measures The way businesses evaluate the effects and business value of a deployed recommender system is influenced by a number of factors, including the application domain and, more crucially, the company’s business strategy. Ads can be used in part or entirely to support such business strategies (e.g., YouTube or news aggregation sites). The goal in this scenario could be to increase the amount of time people spend using the service. Increased engagement is also a goal for firms with an at-rate subscription model (e.g., music streaming services). The underlying business models and objectives govern how firms judge the value of a recommender in all of the examples above. The diagram below depicts the basic measuring methodologies identified in the literature, which we are going to discuss further one by one. Source Click-Through Rates The click-through rate (CTR) is a metric that measures how many people click on the recommendations. The basic notion is that if more people click on the recommended things, the recommendations are more relevant to them. In news recommendations, the CTR is a widely used metric. Das et al. discovered that personalized suggestions resulted in a 38 percent increase in clicks compared to a baseline that merely recommends popular articles in an early paper on Google’s news personalization engine. However, on some days, when there was a lot of attention to celebrity news, the baseline actually fared better. Adoption and Conversion Click-through rates are often not the final success measure to pursue in recommendation scenarios, unlike online business models dependent on adverts. While the CTR can measure user attention or interest, it can’t tell you whether users liked the recommended news article they clicked on or if they bought something based on a recommendation. As a result, alternative adoption measures are frequently utilized, which are ostensibly more suited to determining the effectiveness of the suggestions and are ostensibly based on domain-specific considerations. YouTube employs the idea of “long CTRs,” in which a user’s clicks on suggestions are only tallied if they view a particular percentage of a video. Similarly, Netix utilizes a metric called “take-rate” to determine how many times a video or movie was actually watched after being recommended. Salves and Revenue In many cases, the adoption and conversion measures outlined in the previous section are more telling of a recommender’s prospective business value than CTR measures alone. When customers choose more than one item from a list of suggestions, This is a good indicator that a new algorithm was successful in identifying later purchases or views. stuff that the user is interested in. Nonetheless, determining how such improvements in adoption translate into greater business value remains difficult. Because a recommender may provide numerous suggestions to consumers that they would buy otherwise, the rise in company value may be lower than what we can expect based on adoption rate increases alone. Furthermore, if the relevance of suggestions was already poor, i.e., nearly no one clicked on them, raising the adoption rate by 100% could result in very little absolute additional value for the company. User Behavior and Engagement Higher levels of user engagement are thought to contribute to increased levels of user retention in various application domains, such as video streaming, which, in turn, often immediately converts into corporate value. A number of real-world tests of recommender systems have found that having a recommender increases user activity. Depending on the application domain, different measurements are used. In the field of music recommendation, researchers compared various recommendation strategies and discovered that a recommendation strategy that combines usage and content data (dubbed Mix) not only led to higher acceptance rates but also to a 50% higher level of activity in terms of playlist additions than the individual strategies. Overall, the use of a specific measure is not limited; it just depends on the type of business problem that the system is being used to solve. An overview of some of our findings may be found in the table below. Source Apart from all of this, we can also leverage our standard ML evaluation metrics to evaluate ratings and predictions that are as follows. PrecisionRecallF1-measureFalse-positive rateMean average precisionMean absolute errorThe area under the ROC curve (AUC) Conclusion Through this post, we have learned what different metrics are used when it comes to evaluating the performance of a recommendation system. Firstly we have seen what are some common challenges that are involved with the recommendation system. Later we have seen some commonly used performance metrics and lastly, we have seen how well established businesses like Netflix, YouTube have defined these evaluation strategies. References Measuring the Business Value of Recommender Systems Choosing Metrics for Recommender System Evaluations","excerpt":"Recommender systems are used in a variety of domains, from e-commerce to social media to offer personalized recommendations to customers","categories":["AI Trends"],"tags":["Data Science","evaluation","evaluation metrics","Machine Learning","recommendation","recommendation engine"],"author_name":"Vijaysinh Lendave","publish_date":"2021-10-24T18:00:00","publication_year":"2021","word_count":1590,"keywords":["Go","machine learning","evaluation metrics","AI","batch processing","ML","Machine Learning","recommendation systems","RAG","GAN","recommendation engine","ViT","evaluation","recommendation","Data Science","R"],"extracted_tech_keywords":["AI","machine learning","ML","RAG","recommendation systems","R","Go","batch processing","GAN","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/how-to-measure-the-success-of-a-recommendation-system\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10165116,"title":"InMobi CEO Might be Right About AI Doing 80% of Coding Jobs","content":"AI coding tools have been both a boon and a bane for software engineers and developers. While many have managed to upskill themselves with AI and use the tools to multiply their output, some have expressed concerns about its impact on their jobs. Amidst this, Naveen Tewari, founder & CEO of InMobi, has fuelled the fire. Tewari sounded an alarm over the future of software engineering jobs, predicting that AI will automate most coding tasks within the next two years. “I think my software engineers will go away. They will not have jobs in two years,” he said while speaking at the LetsVenture event last week. He further revealed that InMobi is on track to achieve 80% automation in software coding by the end of this year. “My CTO will deliver 80% automation in software coding by the end of this year. We have already achieved 50%. The codes created by the machine are faster and better, and they fix themselves.” He urged professionals to adapt, warning that even specialised jobs are at risk. “Upgrade yourself, don’t ask me to upgrade you. Because this is survival. The world underneath you is shifting.” InMobi has been aggressively integrating AI across its operations. Glance, its consumer tech platform, recently partnered with Google Cloud to develop generative AI solutions. In September, the SoftBank-backed firm secured $100 million in debt financing from MARS Growth Capital to accelerate AI adoption. But, Is He Right? Social media has been filled with people reacting negatively to Tewari’s comments. While calling out the problems in InMobi’s software stack, including bloatware and bugs, developers on various platforms do not agree with Tewari’s comments. They say he is just jumping on the bandwagon of commenting on AI taking over jobs. There have been several reports of companies using AI coding tools internally, leading them to rely less on the engineers. Following the AI boom after ChatGPT, companies have been trying to call themselves AI companies. Many are adopting AI models or partnering with AI companies to implement them in the workplace. While Indian companies were initially hesitant, AI use is now becoming more accepted, and there is a shift towards in-house development instead of relying on existing market solutions. Recently, a full-stack developer working in a product-based firm highlighted how the teams have been using AI tools extensively and have also built an in-house extension of VS Code to integrate into ChatGPT, Gemini, and Claude. The tool also has access to the company’s entire codebase. Compensation for software developers has increased, but only for the senior developers. The number of coder openings has reduced over the years following the hiring boom after 2020. “The middle-class engineer is dying. And they’re dying because they’re not needed anymore,” said Greg Isenberg, CEO of Late Checkout. “We have product builders who happen to code. Armed with AI, they ship entire products in days.” It seems like many companies do not need these developers anymore. While speaking with AIM, the head of operations of a firm setting up its GCC in Bengaluru, on condition of anonymity, said that they aimed to hire around 150 engineers. The company aims to recruit engineers skilled in AI tools, and are capable of exhibiting productivity equivalent to five engineers. Product builders with AI are surplus, with frontier engineers being offered $500k and solo engineers outperforming teams with the help of AI tools. Projects are shipping in days, not months. A single developer can now accomplish what once required an entire team. Companies are streamlining their operations, reducing headcount, and redefining what it means to be a software engineer. Global Phenomenon In a 2023 interview with The Atlantic, OpenAI CEO Sam Altman said, “A lot of people working on AI pretend that it’s only going to be good; it’s only going to be a supplement; and no one is ever going to be replaced.” “Jobs are definitely going to go away.” Gartner, the research and advisory firm, said in its October report that AI will spawn new roles in software engineering and operations through 2027, requiring 80% of the engineering workforce to upskill. This is similar to what Kunal Shah, the CEO of Cred, said earlier while explaining AI’s impact on professional competency. Shah compared AI copilots in coding to calculators in math: they can boost productivity but may also erode fundamental skills. He predicts that AI will dramatically alter the job market, leading to a sharp distinction between those who leverage AI and those who do not. Companies will eventually compensate AI-enabled employees differently as productivity gaps widen. “The metric that will soon matter the most is revenue per employee,” Shah stated. “If one engineer is 20x more productive because of AI, why would a company pay the same salary to someone who refuses to use AI?” This is increasingly coming true, and InMobi might be just one of them.","excerpt":"“We have already achieved 50%. The codes created by the machine are faster and better, and they fix themselves,” Naveen Tewari said.","categories":["AI Features"],"tags":["AI in Coding","Developers"],"author_name":"Mohit Pandey","publish_date":"2025-03-04T19:00:00","publication_year":"2025","word_count":812,"keywords":["ChatGPT","TPU","OpenAI","AI","ML","AI in Coding","RAG","Aim","generative AI","copilots","R","Developers"],"extracted_tech_keywords":["AI","ML","generative AI","ChatGPT","OpenAI","Aim","RAG","copilots","TPU","R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/inmobi-ceo-might-be-right-about-ai-doing-80-of-coding-jobs\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10091307,"title":"Raspberry Pi Brings Artificial Intelligence to the Masses","content":"Ever dabbled with DIY electronics projects? If yes, chances are you’ve already met Raspberry Pi – a name known to hobbyists and industry users alike. Recently, the company announced a strategic partnership with Sony’s semiconductor division to bring a ​line of Sony’s edge AI devices to the Raspberry Pi ecosystem. “We see a lot of people using Raspberry Pi for AI\/ML applications,” said Eben Upton, CEO of Raspberry Pi Ltd. The use cases vary from hobbyists using it to classify animals they saw in their backyard to actual industrial applications such as classifying objects on the production line into good or bad. People until now were either using third-party accelerators such as the Google USB 3.0 accelerated device which could be directly connected to Raspberry Pi, or were doing the AI work on the CPU. “For a lot of AI applications, the CPU is actually sufficient,” said Upton. However, it has its own advantages and disadvantages. “What this [partnership] is getting us access to is solely a suite of AI-accelerated vision products. So particularly, product line X500, which is an image sensor that has an integrated CNN inference accelerator, and other AI products and services,” he added. Open, but not so open Raspberry Pi has previously faced the heat for not being completely open. Almost all of their products have proprietary silicon on them. Thus, although at the application layer it may be open, the internal systems, including the HAL layer and drivers, are closed and not accessible to users. Speaking about this, Upton stressed, “I think what’s important is not whether the hardware is open-source or proprietary, but whether you can run standards-based ML models on the hardware.” Sony, for instance, has a DSP (Digital Signal Processor) which accelerates convolution operations by providing the tooling required to convert TensorFlow for closed systems. “TensorFlow is kind of a de facto standard, I guess, for model representation.” Or they provide tooling that will convert from an open standard, or a pseudo open standard or something like TensorFlow, which is technically a Google proprietary standard but open. The toolings act as a bridge between the open world and closed systems. “And so it [the hardware] doesn’t need to be open. It merely needs to implement an open standard. It needs to provide access to its proprietary capabilities through an open standard and we do this with openMAX for video, likewise for imaging purposes and KMS composition. Similarly, we’ll do it for TensorFlow for machine learning acceleration,” he added. RISC-V-based Raspberry Pi? Although Raspberry Pi is a member of RISC-V International, Upton says that the value proposition provided by RISC-V isn’t a slam dunk. While the upside of having the flexibility to add instruction sets for domain specific applications is certainly big, he stressed, “there is a formidable barrier between RISC-V and volume success, in particular around core maturity and software maturity”. There is thus a trade-off between freedom and certainty, and depending on one’s viewpoint of this trade-off, they can pick an architecture. But, Upton is still hopeful. He gives the example of ARM which used to have the same problem – in relation to core maturity and software maturity – but they resolved it, so, he says, “there is no reason to believe that in ten years’ time, there wouldn’t be RISC-V based products coming out of Raspberry Pi.” “If ARM maintains their current trajectory in terms of the fraction of the value they want to extract from the system, it’s going to be very hard for them to justify the trouble and expense of migrating to a different instruction-set architecture. So, from this point of view, I would say it is ARM’s to lose.” And while ARM has publicly downplayed the risks of open-source ISA to their chip empire, they do see the customer demand shifting. In 2019, the company announced Custom Instructions, a new feature of its Armv8-M architecture for embedded CPUs. So, we can expect them to reorient themselves against the changing tide. What next? Over the years, the evolution of Raspberry Pi has been such that now over 70% of sales come from industrial and commercial applications. This is why one of the challenges in the past year for them has been trying to balance providing the hobbyist space with material against sustaining their existing OEM customers. The company, which is recovering from the supply chain constraints over the past two years, is trying to gauge the underlying demand for its products. Once they get into full production capacity, only then can a realistic estimate be given. Discussing the path ahead, Upton said they are doing a lot of software work in the existing generation of products. “We try to see our progressions as a series of steps. However, every new generation of the product is not a stepped change. But, what we do is we still pursue pouring sand onto the steps, so that you get a bit of a slope,” remarked Upton. He likens the software investments to the metaphor of a sand, which will give them just enough leverage to make the big move to the next generation.","excerpt":"The partnership will bring a ​line of Sony’s edge AI devices to the Raspberry Pi ecosystem","categories":["AI Features"],"tags":["edge AI","Raspberry Pi","Sony"],"author_name":"Ayush Jain","publish_date":"2023-04-13T14:00:00","publication_year":"2023","word_count":854,"keywords":["Go","machine learning","AI","ML","Git","RAG","CNN","edge AI","Sony","Raspberry Pi","TensorFlow","R"],"extracted_tech_keywords":["AI","machine learning","ML","TensorFlow","edge AI","RAG","R","Go","Git","CNN"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/raspberry-pi-brings-ai-to-the-masses\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":17286,"title":"9 TED Talks About Artificial Intelligence That Will Allay Our Moral, Ethical And Technical Doubts","content":"Are you a student eager to learn more about Artificial Intelligence (AI) or are you a professional looking for informative videos in your area of interest? Whoever you are, be assured that not just enthusiasts and practitioners, but also the inquisitive user whose expertise is limited to Sci-fi is interested in AI now. Be it films, books, music or theatre, our over-active imaginations have always been haunted by a dystopian future of the society. Our collection of TED talks from all over the world will help you understand the nuances of AI in simple terms, and highlight the fact that AI is after all a product of the human mind. 1. In this cautionary talk, techno-sociologist Zeynep Tufekci explains how intelligent machines can fail in ways that don’t fit human error patterns. “We cannot outsource our responsibilities to machines,” she says. “We must hold on ever tighter to human values and human ethics.” Duration: 17min 42 sec 2. Neuroscientist and philosopher Sam Harris asks one of the most important questions about AI that worries all of us: Can we build AI without losing control over it? Duration: 14min 27sec 3. Do you think that AI is the next natural step from human evolution? In this TED talk, futurist Ray Kurzweil suggests that we should get ready for the next big leap in brain power as we tap into the computing power in the cloud. Duration: 9min 52sec 4. How is AI related to art? Computer vision expert Fei-Fei Li talks draws parallel about teaching children to recognise pictures, to that of feeding computers and AI with a variety of images. Duration: 17min 58sec 5. What happens when we teach a computer how to learn? Technologist Jeremy Howard shares some surprising new developments in the fast-moving field of deep learning, a technique that can give computers the ability to learn Chinese, or to recognize objects in photos, or to help think through a medical diagnosis. Duration: 19min 45sec 6. We must face our fears if we want to get the most out of technology — and we must conquer those fears if we want to get the best out of humanity, says the great chess player Garry Kasparov. Even though he lost a memorable match to IBM supercomputer Deep Blue in 1997, he now shares his vision for a future where intelligent machines help us turn our dreams into reality. Duration: 15min 20sec 7. Most of us are worried about the future, where all jobs are taken over by the machines. But David Autor the professor of Economics and Associate Head of the MIT Department of Economics tries to explain this paradox. Duration: 18min 51sec 8. How would a man, who has never been beaten in the game of Jeopardy feel when he has been beaten by a supercomputer? In this fun talk, trivia whiz Ken Jennings says that how “good old fashioned” human knowledge is the key to success. Duration: 17min 52sec 9. And finally, to address all the doomsday theories about robots taking over the world, AI pioneer Stuart Russell is working on three principles that will help us create safer AI. Duration: 17min 35sec","excerpt":"Are you a student eager to learn more about Artificial Intelligence (AI) or are you a professional looking for informative videos in your area of interest? Whoever you are, be assured that not just enthusiasts and practitioners, but also the inquisitive user whose expertise is limited to Sci-fi is interested in AI now. Be it […]","categories":["AI Trends"],"tags":["AI (Artificial Intelligence)","TED Talks"],"author_name":"Prajakta Hebbar","publish_date":"2017-08-29T11:26:19","publication_year":"2017","word_count":526,"keywords":["Go","artificial intelligence","AWS","AI","cloud_platforms:AWS","programming_languages:R","TED Talks","computer vision","Ray","deep learning","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","deep learning","computer vision","Ray","AWS","R","Go","cloud_platforms:AWS","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/ted-talks-about-ai\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10094842,"title":"Goa Schools Introduce AI For Class 9 and Above","content":"The Indian education system might be witnessing one of the first adoption of AI in their curriculum amid criticism in the sector. The Goa Board of Secondary and Higher Secondary Education (GBSHSE) has announced that from the start of academic years 2023-24, it will introduce AI as an additional subject. This will be started from Class 9. The official notification stated, “With reference to the above-cited subject, the new subject Artificial Intelligence (AI) has been introduced as an additional Subject in line with National Skills Qualifications Framework (NSQF) scheme as 7 per cent subject for Class 9 from the Academic year 2023-2024.” The subject would be compulsory for schools that are not offering NSQF subjects such as fine arts, astronomy, physical education, music, among others. “In partial Modification to circular No.33, dated 28™ June 2022, the syllabus of Artificial Intelligence (AI) \/ ICT of Class IX and Class XI will be the same for the Academic year 2023-24 under school Assessment,” the notification stated. “Those Schools\/Students opting for Artificial Intelligence (AT) as 7th subject at Class IX, the existing School Assessment Subject, ICT will be  exempted.” Students of class 9 opting for the AI subject will have 40 marks worth of theory paper divided in two terms and 60 marks practical divided in two terms. GBSHSE recommends schools to have three sessions of AI for 33 minutes each per week for Class 9 and Class 11. On an interesting note, Bhagirath Shetye, the chairperson of GBSHSE said that instead of appointed professionals and experts from the field of AI, the teachers already in the school who are interested in the field will have to teach the new subject to the students. The new course will be taught to mostly the computer teachers of the school. The new course will be accompanied with a new AI textbook under the Goa Board circular. This would include AI ethics, Python programming, neural networks, and the use of AI for the better world.","excerpt":"The Indian education system might be witnessing one of the first adoption of AI in their curriculum amid criticism in the sector. The Goa Board of Secondary and Higher Secondary Education (GBSHSE) has announced that from the start of academic years 2023-24, it will introduce AI as an additional subject. This will be started from […]","categories":["AI News"],"tags":["AI in Education"],"author_name":"Mohit Pandey","publish_date":"2023-06-09T15:25:16","publication_year":"2023","word_count":329,"keywords":["AI in Education","Go","artificial intelligence","programming_languages:R","AI","neural network","programming_languages:Go","AI ethics","Python","programming_languages:Python","R"],"extracted_tech_keywords":["AI","artificial intelligence","neural network","Python","R","Go","AI ethics","programming_languages:Python","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/goa-schools-introduce-ai-for-class-9-and-above\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10085716,"title":"Bridging the gap between Data leaders and Recruitment Heads","content":"“The first way to bridge the gap is to understand the other perspective.” Bridging the gap between data leaders and recruitment leaders is crucial for any organisation that wants to stay competitive and make data-driven decisions. Both data leaders and recruitment leaders play a vital role in the success of an organisation, but their responsibilities and priorities may not always align. By understanding and addressing the reasons for the gap between these two departments, companies can improve communication, collaboration and achieve a better outcome. The gap between data leaders and recruitment heads can be attributed to a combination of factors but with proper communication, collaboration, and understanding of each other’s roles, the gap can be bridged. Organised by AIM Recruits at the Machine Learning Developer Summit, 2023, AIM Recruits held a roundtable discussion on, ‘Bridging the gap between Data leaders and Recruitment Heads’. For this conversation, we invited data leaders from the AIM Leaders Council: Sayandeb Bannerjee, Co-founder and CEO at TheMathCompany, Suguna Jayraj, SVP, Head of P&C Analytics at Swiss Re, Shan Duggatimatad, Data & AI leader – Senior Director at Ascendion, Ravindra Patil, Group Leader, Data and AI at Philips and Bala Natrajan, Vice President – AI & Analytics at Capgemini. On the other hand were also accompanied by recruitment heads: Rosina Jose, Associate Director HR at Rakuten India, Lijosh Varkey Joseph, Director – Talent & Culture at Flutura Decision Science & Analytics, Amitabh Ghosh, Ex-Head of Talent Acquisition at Anheuser – Busch InBev, along with Mohit Juneja, HR Leader – Engineering and Technology Centre & Org Capability – MEIA at Ingersoll Rand. The session was moderated by Kashyap Raibagi , Manager – Research & Advisory at Analytics India Magazine. This round-up style article mentions best quotes from these leaders. From hiring to finding talent We’ve been setting up teams with three separate job descriptions like data scientists, data engineers, and data analysts. So, when the resume comes, it’s quite mixed. While at a CV level, it’s hard to differentiate because the boundaries are also merging somewhere. Segregating that and getting the right fit becomes a challenge, especially when the pipeline is huge. And the second key challenge that we see is fake CVs. So, in two forms, one is either a copy project from a hackathon or a ready-made data set. —Suguna Jayraj, SVP, Head of P&C Analytics at Swiss Re Roadblocks in mass recruitment Ability to upskill a recruiter is a must and that has led to the result of what we did as a Gemba walk. We show what an AI model is, how it works, what a data engineer is, what it looks like. Then, we created a hiring playbook where every role within the landscape of data is mentioned and what question do you need to ask [for the specific role]. —Ravindra Patil, Group Leader, Data and AI at Philips I feel the challenge is how we can come closer to each other. What is critical is the conversation that is happening between a data leader and an HR leader, when it is not a recruitment conversation. There is a way to explain in a layman’s language, what this is about and the same holds the other way around that the data leader takes ownership of the talent they need to build the product and solution. So, that nuance changes. A recruiter can’t be blamed that he or she was in a particular environment where bulk recruitment was happening and now you’re suddenly in an environment where you need 10 [profiles], but super specific. The transition is not automatic. I think we have to appreciate that. —Sayandeb Banerjee, co-founder & CEO at TheMathCompany Upskilling recruiting teams We hire a small number of people across domains. It’s very difficult to have specific creators who know one particular area. So, the fundamentals of  having a good hiring strategy meeting, good job description in place and a well-defined tool come in place, which help the candidate understand what the role is. What they also do is segregation in terms of what are the essential skills and what are the desired skills. So, that’s where the recruiter will be able to judge with respect to the profile. These questions actually help recruiters, who are general recruiters, to do a thorough investigation. They may not be 100% successful, but it does help to an extent that probably improves the ratio of the quality of profiles which they bring. They will understand the profiles more over a period of time. Some of these actions actually help bridge the gap and reduce the time to hire over a period of time. —Mohit Juneja, HR Leader – Engineering and Technology Centre & Org Capability- MEIA at Ingersoll Rand Expectations from younger recruiters The skill universe that the team of recruiters is looking at is quite large. They get allocated to certain sets of skills. The metrics tend to be around monthly and in the last three years, the demand has been off the charts. I don’t think the junior recruiters are measured on the candidate conversion ratio. They’re just given an absolute number, on the number of people to be brought on board at the end of the month. We spend a lot of time trying to work through a problem, particularly when we see the first few applicants getting turned down. Many of our customers want to interview the candidates, even if they’re already within the system. When early rejections happen, we intensify that interaction with the hiring team. It could either be to localise certain pockets from where you can hire or it could be about a particular combination of domain along with skill. It depends on what the customer is looking for, then we will have to do those interactions. Without that, you will not see good outcomes. —Bala Natrajan, Vice President – AI & Analytics at Capgemini Identifying the right tech role The first step from any recruitment team perspective is to understand the role that has many proper debriefs to understand the role, its nuances and specifics. Second, it is more essential for a recruiter to understand what kind of skill set you want for your team. If you say I need somebody from a Python background—what is the kind of experience you need in Python? What is the kind of product or project that you’re looking at? If you’re clear on that, it helps equip the recruiter to screen the profiles properly. I think the recruiter or the team should work with the business together to prepare the screening questions. So, the skill sets are different. If you educate your recruiter that way, it will help you get better screening profiles. That is one of the key essential things that I would want in my team to deliver the best to my business. —Amitabh Ghosh, Ex-Head of Talent Acquisition at Anheuser – Busch InBev Expectation in complementary roles like recruiting There’s another request, which always comes along with tech. To consider the cultural fitment, because some hires just come in, it could be a disaster for the team [if they do not understand the culture of the organisation]. So, for example, we asked the recruiters to sit along with the business and understand the culture. As an organisation, there is a macro culture and within specific departments, there is also a micro culture. I think it’s very important for recruiters to also understand the culture of each department that they’re catering to. We need multiple training sessions for that to ensure that not just tech but even this side of things are taken care of. —Rosina Jose, Associate Director HR at Rakuten India Only tech? I would expect my recruiter to look at the softer aspects more than the technical aspects. There could be one or two skills, which are in depth and then look for horizontal and broader experience. For that, I don’t need a technical recruiter. I don’t expect my recruiter to do a deep technical discussion, rather look for the softer attributes of an associate. So, those are the aspects that I would emphasise for our recruiter to focus on rather than the technical skills, because technical skills change. —Shan Duggatimatad, Sr director at Ascendion Outcome and expectation Data leaders must essentially play a part of that role in that continuum. There has to be an effort, in that both sides have to try to explain what we do better. Data leaders have to understand nature and Recruitment leaders have the responsibility to sometimes train or make the data details come to ground and say that this is the ground reality that is out there. —Sayandeb Banerjee, co-founder & CEO, TheMathCompany We need to come together as a community and have a uniform definition of what is data science, who is a data engineer, because for every organisation, the definition changes, and that’s where technology changes. If all the data leaders could come together and unify the definition, that would be of great help to us. —Lijosh Joseph, Director – Talent & Culture at Flutura Decision Sciences & Analytics Lastly, it’s important to note that bridging the gap between data leaders and recruitment heads is an ongoing process and requires relentless effort. Organisations should establish a framework for regular communication and collaboration, and periodically review and adjust as necessary. Additionally, it’s important to involve both data leaders and recruitment heads in the decision-making process and make sure that their voices are heard and considered. Furthermore, it’s important to have a clear understanding of the key skills and expertise required for data roles and make sure that the recruitment team is aware of these, as this will help them to identify and attract the right candidates. It is noteworthy that providing training and development opportunities for both data leaders and recruitment heads to learn about each other’s roles and responsibilities can help bridge this gap and improve collaboration.","excerpt":"bridging the gap between data leaders and recruitment heads is an ongoing process and requires relentless effort. Organisations should establish a framework for regular communication and collaboration, and periodically review and adjust as necessary. Additionally, it’s important to involve both data leaders and recruitment heads in the decision-making process and make sure that their voices are heard and considered.","categories":["AI Highlights"],"tags":[],"author_name":"Anshika Mathews","publish_date":"2023-01-24T16:00:00","publication_year":"2023","word_count":1655,"keywords":["data science","Go","machine learning","AI","Python","Aim","ViT","analytics","GAN","R"],"extracted_tech_keywords":["AI","machine learning","data science","analytics","Aim","Python","R","Go","GAN","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/bridging-the-gap-between-data-leaders-and-recruitment-heads\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10098121,"title":"Gorilla Out Performs GPT-4 in Making API Calls","content":"The AI community is abuzz with the launch of Gorilla, a revolutionary language model that brings an unprecedented level of accuracy and functionality to invoking APIs using natural language queries. Gorilla is an open-source project, licensed under Apache 2.0, and is the result of extensive fine-tuning on Falcon and MPT created by Shishir Patil who is currently a Phd Student in ML systems at UC Berkeley. https:\/\/twitter.com\/shishirpatil_\/status\/1661780076277678082 Unlike its predecessors, including the highly acclaimed GPT-4, Gorilla truly stands out by significantly reducing hallucinations and incorrect syntax when interacting with over 1,600 APIs and counting. This achievement is made possible by Gorilla’s capability to parse the Abstract Syntax Tree (AST) when writing code, resulting in semantically and syntactically correct API invocations. One of the key features that make Gorilla exceptional is its compatibility with commercial use, allowing developers to incorporate it into their projects without any obligations. It’s available for use in various environments, including Google Colab and Command Line Interface (CLI) through pip installation of “gorilla-cli.” APIBench, a meticulously curated collection of APIs, accompanies Gorilla. This extensive library facilitates easy training and expands the available resources for Gorilla to call upon. Furthermore, the project actively encourages API developers to join the cause and contribute their APIs for inclusion in the ever-growing API store. For developers and language model enthusiasts, Gorilla represents a remarkable advancement in leveraging large language models for practical purposes. With its capabilities surpassing even the highly-touted GPT-4, Gorilla has emerged as a game-changer in the domain of API integration and usage. GitHub Repository: https:\/\/github.com\/ShishirPatil\/gorilla Read the Paper: arXiv","excerpt":"Gorilla truly stands out by significantly reducing hallucinations and incorrect syntax when interacting with over 1,600 APIs","categories":["AI News"],"tags":[],"author_name":"Siddharth Jindal","publish_date":"2023-08-04T13:33:05","publication_year":"2023","word_count":261,"keywords":["Go","API","AI","ML","Git","RAG","Colab","Aim","GitHub","R"],"extracted_tech_keywords":["AI","ML","Aim","Colab","RAG","R","Go","Git","GitHub","API"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/gorilla-out-performs-gpt-4-in-making-api-calls\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10072150,"title":"Private players can now use drones for delivery purposes","content":"In compliance with the Drone Rules 2021, the minister of state for civil aviation (Retd) General VK Singh approved the use of drones by private players, in Rajya Sabha on Monday. The government has been utilizing drones for vaccine delivery, inspection of oil pipelines and power transmission lines, anti-locust operations, agricultural spraying, a survey of mines, and land mapping for several years. The minister says, “Drones offer tremendous benefits to almost all sectors of the economy. These include agriculture, vaccine delivery, surveillance, search and rescue, transportation, mapping, defence, and law enforcement to name a few.” Singh also said that in September 2021, the government had notified a production-linked incentive (PLI) scheme to promote drone manufacturing by private companies. The scheme provides an overall incentive of Rs120 crore spread over three financial years. The PLI rate is 20% of the value-added over three financial years. The PLI for a manufacturer shall be capped at 25% of the total annual outlay. Commenting on the news, founder and director of DroneAcharya Aerial Innovations, Prateek Srivastava, said, “DroneAcharya will always welcome, and leverage such permissive developments towards making India the drone hub of the World by 2030. We are grateful that the govt has taken considerable steps towards achieving that goal and are looking forward to exciting times ahead in the world of delivery drones.”Notified on August 25, 2021, the Drones Rules, 2021 provides the necessary regulatory framework for the commercial use of drones. It mainly covers type certification, registration, and operation of drones.","excerpt":"Notified on August 25, 2021, The Drones Rules, 2021 provides the necessary regulatory framework for the commercial use of drones","categories":["AI News"],"tags":["aviation","Government"],"author_name":"Bhuvana Kamath","publish_date":"2022-08-04T13:47:12","publication_year":"2022","word_count":250,"keywords":["Go","programming_languages:R","AI","innovation","programming_languages:Go","Government","RAG","aviation","Ray","R"],"extracted_tech_keywords":["AI","Ray","RAG","R","Go","innovation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/private-players-can-now-use-drones-for-delivery-purposes\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10062672,"title":"Why Hasura’s entry into the Unicorn is a major milestone for India’s open-source ecosystem","content":"Open-source SaaS company Hasura raised $136.5 million in its Series C funding round, taking its total evaluation to $1 billion. With this funding, Hasura becomes the latest entrant to the Unicorn club. It also holds the unique designation of being India’s first open-source Unicorn. Source: Crunchbase Hasura’s journey Operational data is highly distributed among several sources. Developers are often left to consume data in insecure and unauthorised compute environments. To remediate this, Hasura offers data APIs which connect multiple services and data sources. Hasura makes web development faster by eliminating bottlenecks to access data for frontend and full-stack developers. The company was founded by Tanmai Gopal and Rajoshi Ghosh in 2017. They initially started in Bangalore, but in view of growing opportunities, moved their base to San Francisco. Gopal, a dual degree (B Tech and M Tech) holder from IIT Madras, had first established 34 Cross, product development and consulting firm in 2013. In an interview, Gopal said that building applications were a frustrating process then, and he wanted to offer developers an easier route. He was soon joined by Ghosh. A graduate in computational biology from the National University of Singapore, Ghosh started her career in research before entering the world of entrepreneurship. 34 Cross soon shut down. Gopal and Ghosh then went on to establish Hasura. A portmanteau, the name Hasura is a combination of Asura (mythological devil character) and Haskell (a programming language). In an interview, co-founder Gopal said the amount of data is exploding. Businesses want to derive useful insights from these huge amounts of data, but they are often caught in bottlenecks – compliance, security, and delivery. The other problem is storing this huge data. Some of the data is stored with a cloud vendor, and other is open-source. Hasura makes it easy to access data from multiple sources. The users can make the decision most appropriate to their business using the technology of their choice. Hasura’s GraphQL An open-source engine, Hasura can connect to PostgreSQL databases and microservices across hybrid and multi-cloud environments. It then automatically builds a GraphQL API backend, making it easier for the developers to build their data-driven applications on top of the unified API. With the latest funding, Hasura will accelerate research and development and further expand global activities for the company’s GraphQL Engine. With this, even users with zero experience can compose a GraphQL API from existing APIs and databases. “With this funding round, our investors and the Hasura team are doubling down on our vision to solve data access and unlock the next decade of developer productivity. We’re going to be addressing the needs of our users by adding support for their favourite data systems much faster,” said Gopal. India’s open-source ecosystem India’s open-source community is growing at an exponential rate. Last year, GitHub, which has over 5 million Indian contributors, allotted Rs 1 crore to support India-based open-source projects and contributors. GitHub COO Erica Brescia said that India is the fastest-growing country when it comes to new developers contributing to open source projects. India’s open-source ecosystem can be traced back to the early 90s when several Indian Linux User Groups and Free Software User Groups were being formed. Over 85 per cent of India’s internet is working on open-source software. Establishments like Courts, State Bank of India, LIC India, IRCTC, etc., rely on an open-source ecosystem to scale operations and provide services. In 2015, the Department of Electronics and Information Technology (DeitY), which falls under the Ministry of Communication and Information Technology, rolled out three open-source related policies – Policy on Adoption of Open Source Software for Government of India; Policy on Collaborative Application Development by Opening the Source Code of Government Applications; Policy on Open Application Programming Interfaces (APIs). Hasura’s entry into the Unicorn club is a major milestone for the open-source ecosystem in India.","excerpt":"With this funding, Hasura becomes the latest entrant to the Unicorn club.","categories":["AI Features"],"tags":["open source community"],"author_name":"Shraddha Goled","publish_date":"2022-03-14T13:00:00","publication_year":"2022","word_count":639,"keywords":["PostgreSQL","Go","API","AI","Git","microservices","open source community","SQL","Rust","GitHub","R"],"extracted_tech_keywords":["AI","microservices","PostgreSQL","R","SQL","Go","Rust","Git","GitHub","API"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/why-hasuras-entry-into-the-unicorn-is-a-major-milestone-for-indias-open-source-ecosystem\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":61309,"title":"Will Entry Of Former Apple 5G Chief Revive Microsoft’s HoloLens?","content":"Microsoft has yet again hired a top Apple executive to enhance the capabilities of its hardware products, including HoloLens. In August 2019, the firm had appointed former Siri chief Bill Stasior, who had earlier left the company after seven years in May 2019. And now, Ruben Caballero has joined Microsoft as vice president for hardware design and technology. Caballero had served 14 years as vice president of engineering at Apple, mainly focusing on wireless technologies, such as antenna design. He was also actively involved in the first iPhone prototype in 2005. With Caballero at the helm, Microsoft would be hoping for superior wireless hardware that can assist the firm in delivering robust consumer hardware products like VR and AR headsets. How Things Panned Out Before leaving Apple in April 2019, Caballero was left out from the further development of the company’s 5G modem. The company earlier had plans for developing in-house 5G modem, but later purchased Intel’s 5G smartphone modem business in a $1 billion deal, which was assigned to vice president Johny Srouji to lead. It is believed that such decisions may have led to the dispute with management and is one of the prime reasons why Caballero left before the deal with Intel. Caballero then worked as a board of director and technical advisor for various firms, including Maui Imaging, Humane, Keyssa, and Resonant. However, it is not clear whether he will continue to be an advisor for the four companies, but his LinkedIn profile still lists it in the current role. At Microsoft, Caballero will oversee the mixed reality and AI division that is responsible for HoloLens and “Special Projects.” Unlike Stasior, who was head of Siri at Apple, but was not assigned to enhance Cortana’s capabilities at Microsoft, Caballero will have a similar role like he did when he was at Apple. At Apple, Caballero was the founding leader of the iPhone hardware team and later built antennas that powered iPhones, iPads, and Macs, which were vital to the success of mobile phones and personal computers. More recently, he has been overseeing the development of wireless chip and antenna for Apple TVs, AirPort wireless routers, among others. Microsoft’s Revival In Consumer Hardware? Microsoft, after failing to compete with Google and Amazon in the latest technologies such as voice assistants, does not want to let go of the fight in AR and VR. For this, the company needs exceptional hardware with state-of-the-art antenna technology. Microsoft, in February 2019, announced HoloLense 2 — an augmented reality headset — at Mobile World Congress, which is now available for purchase. However, this is focused on enterprise needs and not on consumer electronics. Today, HoloLense 2 does not support cellular connectivity as it only works with Wi-Fi and Bluetooth for communication. Microsoft has failed to deliver exceptional connectivity with the HoloLense, as the radio emission guidelines consider mobile antennas around the head for a more extended period as unsafe. This is one of the leading problems for companies that are developing headsets. Therefore, firms are exploring other connectivity options, such as next-generation Bluetooth and variants of Wi-Fi for high-speed connection while ensuring that it is comfortable. This is where Caballero’s expertise can play a vital role in helping Microsoft develop user-friendly devices, not only for enterprises but also for the general public. Outlook As 5G is getting more real, it can become a game-changer in the way users and businesses across the world will communicate by transferring vast information without latency. Processing a colossal amount of information along with the capability of free-flowing of enormous data is crucial for AR and VR devices to be used in day-to-day activities. However, the hardware-intensive processors and 5G cause hindrance in accomplishing desired XR headsets. It would be interesting to see if Microsoft can do justice to the hype of XR with Caballero’s expertise. Also Read: Can Facebook Deliver On AR\/VR","excerpt":"Microsoft has yet again hired a top Apple executive to enhance the capabilities of its hardware products, including HoloLens. In August 2019, the firm had appointed former Siri chief Bill Stasior, who had earlier left the company after seven years in May 2019. And now, Ruben Caballero has joined Microsoft as vice president for hardware […]","categories":["Global Tech"],"tags":[],"author_name":"Rohit Yadav","publish_date":"2020-04-09T16:00:00","publication_year":"2020","word_count":647,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","ViT","R"],"extracted_tech_keywords":["AI","R","Go","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/will-entry-of-former-apple-5g-chief-revive-microsofts-hololens\/","complexity_score":2,"technical_depth":6,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10118371,"title":"OpenAI Now Eyeing an Office in New York","content":"OpenAI is planning to establish an office in New York City next year, according to a report citing people familiar with the company’s plans. This would be the company’s fifth office, adding to its headquarters in San Francisco, a recently opened office in Tokyo, and offices established last year in London and Dublin. OpenAI has not yet finalised a location or signed a lease in New York, according to one source, but they are considering spaces in Manhattan and Brooklyn. At the start of last year, OpenAI had around 400 employees all based in a single San Francisco office. Now, the company is searching for a second office in San Francisco, as reported by the San Francisco Chronicle, due to its workforce expanding to over 1,000 employees, according to one of the sources familiar with the company. Recently, OpenAI had announced its entry into the Asian market by opening its first office in Tokyo, Japan. The company is unveiling a GPT-4 custom optimised for the Japanese language. The company also plans to release the custom model more broadly in the API in the coming months. In December, OpenAI had also announced plans to start an office in India. Rishi Jaitly, who has held executive positions including the position of Vice President at Twitter, will assume the role of a senior advisor at OpenAI to guide the company through India’s AI policy and regulatory environment. Furthermore, OpenAI executives Anna Adeola Makanju , global head of Public Policy, James Hairston, and Jaitly recently met MoS for Electronics and Information Technology, Rajeev Chandrashekar.","excerpt":"The company is also planning to get another office in San Francisco as its team is getting bigger.","categories":["AI News"],"tags":["OpenAI"],"author_name":"Mohit Pandey","publish_date":"2024-04-17T12:47:45","publication_year":"2024","word_count":260,"keywords":["API","OpenAI","AI","programming_languages:R","GPT","R","llm_models:GPT"],"extracted_tech_keywords":["AI","OpenAI","R","API","GPT","llm_models:GPT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/openai-now-eyeing-an-office-in-new-york\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10059780,"title":"IISC Bengaluru installs one of India’s most powerful supercomputers","content":"Indian Institute of Science has installed one of India’s most powerful supercomputers, Param Pravega, in its campus. It is one of the largest supercomputers ever installed at an Indian academic institution under the National Supercomputing Mission. The supercomputer will be used for research and educational purposes. Param Pravega has a total supercomputing capacity of 3.3 petaflops. Param Pravega was designed by the Centre for Development of Advanced Computing and most of the components used to build it has been manufactured and assembled in India using a homegrown software stack developed by C-DAC, in line with the Make in India initiative. The Param Pravega system at IISc is a mix of heterogeneous nodes, with Intel Xeon Cascade Lake processors for the CPU nodes and NVIDIA Tesla V100 cards on the GPU nodes. The hardware consists of an ATOS BullSequana XH2000 series system, with a comprehensive peak compute power of 3.3 petaflops. The software stack on top of the hardware is provided and supported by C-DAC. The machine hosts an array of program development tools, utilities, and libraries for developing and executing High Performance Computing (HPC) applications. Under the National Supercomputing Mission, 10 supercomputers have been deployed at  IISc, IITs, IISER Pune, JNCASR, NABI-Mohali and C-DAC, with a cumulative computing power of 17 petaflops. About 31,00,000 computational jobs have successfully been carried out by around 2,600 researchers across the country to date.","excerpt":"Param Pravega has a total supercomputing capacity of 3.3 petaflops","categories":["AI News"],"tags":["c-dac","GPU","IISc","Intel Xeon","Make in India","Supercomputers"],"author_name":"SharathKumar Nair","publish_date":"2022-02-03T18:56:03","publication_year":"2022","word_count":230,"keywords":["Make in India","programming_languages:R","IISc","Ray","Supercomputers","c-dac","Intel Xeon","V100","R","GPU"],"extracted_tech_keywords":["Ray","R","V100","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/iisc-bengaluru-installs-one-of-indias-most-powerful-supercomputers\/","complexity_score":3,"technical_depth":4,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10166260,"title":"Meta’s Llama Hits 1 Billion Downloads, Doubling Growth in Three Months","content":"Meta’s open source AI model family, Llama, has surpassed 1 billion downloads, CEO Mark Zuckerberg announced on Threads. This milestone marks a sharp rise from 650 million downloads in December 2024—a 153% surge in just three months. First introduced for researchers, Llama has since expanded its reach to developers and AI enthusiasts worldwide. View on Threads Ahmad Al-Dahle, VP, head of GenAI at Meta, posted on LinkedIn about the achievement saying that in the span of two years, Llama went from a research project to the most widely used open source AI model. “I’m so proud of my team for working tirelessly to make Llama better everyday. Meta’s GenAI org actually turned 2 years old on February 27th and it has been so inspiring and humbling to see them work so hard to improve Llama in such real ways for developers, startups and enterprises around the world,” Dahle said. He added that this moment is a celebration of the entire open source AI community. “Each Llama download represents a vote of confidence in open source AI – and a shared belief that we can go so much further building together,” he added. Meta is gearing up to release multiple Llama models in the coming months, including “reasoning” models similar to OpenAI’s o3-mini and versions with built-in multimodal capabilities. Zuckerberg has also hinted at “agentic” features, indicating that some models may have the ability to operate autonomously. In October last year, Zuckerberg also confirmed AIM’s prediction of the company releasing Llama 4 early 2025 at its third quarterly earnings call. “I expect that the smaller Llama 4 models will be ready first, and we expect [them] sometime early next year, and I think that they’re going to be a big deal on several fronts — new modalities, capabilities, stronger reasoning, and much faster,” Zuckerberg said. The Llama 3 models marked a major turning point in the industry, but Zuckerberg expressed even greater enthusiasm for Llama 4, which he confirmed is well into development. “We’re training the Llama 4 models on a cluster that is bigger than 100k H100s or bigger than anything that I’ve seen reported for what others are doing,” he said. Llama powers Meta AI across Facebook, Instagram, and WhatsApp, and is a key part of the company’s long-term AI strategy.","excerpt":"In October last year, Mark Zuckerberg also confirmed that Meta is releasing Llama 4 early 2025.","categories":["AI News"],"tags":["Developers","LLaMA","Meta AI"],"author_name":"Mohit Pandey","publish_date":"2025-03-18T20:42:28","publication_year":"2025","word_count":382,"keywords":["Go","Meta AI","GenAI","OpenAI","AI","Llama 4","RPA","LLaMA","Aim","R","Developers","startup"],"extracted_tech_keywords":["AI","GenAI","OpenAI","Llama 4","Meta AI","Aim","R","Go","RPA","startup"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/metas-llama-hits-1-billion-downloads-doubling-growth-in-three-months\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":26259,"title":"IIIT-B Student Bags ₹1.2 Crore Residency With Google AI","content":"File photo of IIIT-B student Aditya Paliwal. (Image credit: @mukuldigitech\/Twitter) A 22-year-old student from International Institute of Information Technology, Bengaluru (IIIT-B) achieved a rare feat this week when he landed a residency at the prestigious AI Google Residency Program. And his gigantic reward is a salary package of ₹1.2 crore per annum. IIIT-B’s Aditya Paliwal was one of the 50 students selected from a pool of 6,000 participants from across the world this year. Originally from Mumbai, Paliwal was studying integrated MTech dual-degree programme of five years from 2013 to 2018 at IIIT-B. The Google AI Residency Program is a 12-month research training role designed to jumpstart or advance the resident’s career in machine learning research. The programme was created in 2015 with the goal of training and supporting the next generation of deep learning researchers. The residency programme website explains that the selected candidates will be given a chance to work with noted scientists and engineers from an option of research teams, choosing placements within the Google Brain team, Perception, Natural Language Understanding or Google AI New York. Paliwal will join the Google AI programme in New York on 16 July 2018. In an interview with TNIE, Paliwal said, “I am more inclined to research but was not sure of doing a PhD immediately. The program was a good compromise and a balance between research and industry. Plus, it is in the domain that I was interested in — artificial intelligence… After the one-year period, one can leave the program to pursue a PhD, extend the program or convert the residency into a full-time job. However, the extension of the program and job conversion is bound to be based on your performance.” Paliwal told a news portal that facilities provided to him at IIIT B encouraged him to excel and supported his innovative ideas. He also expressed gratitude towards his seniors in college. His alma mater tweeted: Aditya Paliwal, we are proud of you and wish you a successful and fruitful endeavor as you take your next steps towards chasing your dreams.https:\/\/t.co\/9K0hfJjo6Z — IIIT-Bangalore (@IIITB_official) July 8, 2018","excerpt":"A 22-year-old student from International Institute of Information Technology, Bengaluru (IIIT-B) achieved a rare feat this week when he landed a residency at the prestigious AI Google Residency Program. And his gigantic reward is a salary package of ₹1.2 crore per annum. IIIT-B’s Aditya Paliwal was one of the 50 students selected from a pool […]","categories":["AI News"],"tags":["AI (Artificial Intelligence)"],"author_name":"Prajakta Hebbar","publish_date":"2018-07-09T07:37:17","publication_year":"2018","word_count":349,"keywords":["Go","machine learning","artificial intelligence","programming_languages:R","AI","Git","RAG","deep learning","GAN","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","deep learning","RAG","R","Go","Git","GAN","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/iiit-b-student-bags-%e2%82%b91-2-crore-residency-with-google-ai\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":53073,"title":"What India Can Learn From UK Initiative Of Democratising Public Transport Data","content":"Traffic congestion has reached a critical point in many cities around the world, resulting in loss of several hours in hand, and creating frustration among motorists. According to a leading traffic provider’s report, the UK drivers usually spend more time in traffic compared to their counterpart in other countries like Belgium, France, Germany, and more. For one, London is only second behind Paris in the most clogged city among Luxembourg, Netherlands, France, Belgium, and Germany. However, with the announcement of collecting data from operators and releasing it portrays how the UK government is keen to mitigate the challenges pertaining to travelling. According to Transport Focus, people often do not prefer buses as they think it is difficult to keep themself updated with timetables and fares. However, with this project, the UK is willing to eliminate the confusion among passengers by improving transparency. Announcement Baroness Vere, buses minister, declared that the government is going to provide transportation data for simplifying the life of travellers with its Bus Open Data Service. The idea behind the initiative is to standardise and publish information from operators, which can be leveraged by programmers to develop robust solutions for assisting users in streamlining their journey. In 2020, the government has planned to release information associated with routes and timetables. Moving forward, it will also deliver fares and real-time bus location data for helping passengers travel confidently in 2021. The UK government believes that such facilities will encourage people to take buses, thereby decreasing traffic congestion. “By harnessing the transforming power of data and technology, we could be on the threshold of a golden age for busses. Sharing data on routes, bus locations and fares will give passengers even more confidence to ride,” said Vere. Today, in the UK, passengers wait endlessly to aboard the bus, resulting in poor user experience. Consequently, it was important for the government to empower various firms by providing data for making applications that can provide real-time data. This will enable users to plan their journey quickly and effectively. Besides, the initiative will also help bus service providers to optimise their offerings for avoiding traffic. The announcement comes after the recent initiative of new low-fare, high-frequency Sperbus networks. To facilitate a convenient journey experience, Britain is about to deliver the first all-electric bus town and contactless payments on every city bus. The UK has moved on to leveraging the latest technologies to make bus transportation more convenient. However, the UK department also mentioned that the Open Bus Data Services would allow it to collect passengers data. Consequently, this has drawn concern from people as the government can surveil them. What India Can Learn From The UK With a huge population of India, traffic jams are common, especially in metropolitan cities such as Delhi, Bengaluru, Chennai, and more. The situation is only getting worse over the years as we witness the rise of vehicles on the road every year. As per a report, 4.4 million passengers cars and commercial vehicles were sold in India in 2018. In other words, Indians bought 54,000 vehicles each day in 2018. The commute is only getting longer and the government approach of mitigating the challenge mostly revolve around improving infrastructure by erecting metros, widening the roads, and robust railways. However, such projects take years to execute, resulting in ineffective solution in the short term. While it does help in a longer run, the critical situation demands a quick approach. Undoubtedly, improving infrastructure is a wise step, but at the same time, becoming data-driven is essential to optimise the user experience. It needs to learn from the UK’s Bus Open Data Service and advise a plan to expedite its initiative of decongesting traffic. Outlook The UK has pioneered the initiative which can encourage other countries to adopt their approach. However, the governments will have to standardise and ensure data privacy to eliminate the risk of any misconduct with users information. India has planned to install CCTV cameras across the country, thus, using such data with computer vision technology can also help in removing the bottleneck.","excerpt":"Traffic congestion has reached a critical point in many cities around the world, resulting in loss of several hours in hand, and creating frustration among motorists. According to a leading traffic provider’s report, the UK drivers usually spend more time in traffic compared to their counterpart in other countries like Belgium, France, Germany, and more. […]","categories":["AI Features"],"tags":["traffic"],"author_name":"Rohit Yadav","publish_date":"2020-01-03T16:00:00","publication_year":"2020","word_count":677,"keywords":["Go","AI","RPA","ML","data-driven","computer vision","RAG","Ray","traffic","Rust","R"],"extracted_tech_keywords":["AI","ML","computer vision","Ray","RAG","R","Go","Rust","RPA","data-driven"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/what-india-can-learn-from-uk-initiative-of-democratising-public-transport-data\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10141151,"title":"Bengaluru-based KOGO Launches World’s First AI Agent Store","content":"KOGO, a Bengaluru-based AI company, has launched the world’s first AI Agent Store, offering businesses seamless access to hundreds of AI tools, agents and plugins. KOGO’s AI Store is designed as a comprehensive platform for developers, SMEs, and enterprises to adopt, deploy, and manage AI solutions with ease. Catering to diverse industries, such as travel, healthcare, finance, retail, and law, the store provides a wide array of specialised AI agents tailored to specific business needs. Its unique features include: Ready-made Agent Templates: Automate tasks like appointment booking, anomaly detection, and customer service with pre-built AI agents. Easy Deployment: Deploy agents within minutes using an intuitive interface, compatible with chat or voice platforms. Seamless Integration & Analytics: Integrates with existing systems and tracks performance through advanced analytics. Custom Agent Builder: Developers can build bespoke agents using KOGO OS and share them in the store. With a pay-per-use model and seamless integration capabilities, the store democratises AI adoption across industries like healthcare, retail and finance. “KOGO’s AI Store opens the door to the future of business innovation in India and the world”, said Raj K Gopalakrishnan, Co-Founder & CEO of KOGO. “With a diverse range of AI agents & ready to deploy tools along with the opportunity for developers to create custom AI agents and publish them on the store, we are empowering businesses to easily deploy AI to significantly enhance their operations, increase productivity, and ultimately drive growth.” KOGO’s efforts in democratising AI were recently bolstered through a partnership with India’s Bhashini project, enabling the creation of multilingual AI agents in Indic languages. This collaboration expands the AI store’s reach, making it a transformative platform for diverse linguistic and regional needs.“As the demand for AI continues to grow, KOGO’s AI Store stands as a testament to the company’s vision of a future where AI is seamlessly integrated into everyday business processes” said Praveer Kochhar, Co-Founder & CPO. “We’re committed to helping businesses realise true Human Potential with AI and enabling work-life balance.”","excerpt":"KOGO’s AI Store is designed as a comprehensive platform for developers, SMEs, and enterprises to adopt, deploy, and manage AI solutions with ease.","categories":["AI News"],"tags":["AI Agent store","KOGO"],"author_name":"Tanisha Bhattacharjee","publish_date":"2024-11-19T11:48:35","publication_year":"2024","word_count":331,"keywords":["Go","AI","innovation","ML","Ray","anomaly detection","ViT","analytics","AI agents","KOGO","R","AI Agent store"],"extracted_tech_keywords":["AI","ML","analytics","Ray","anomaly detection","R","Go","ViT","innovation","AI agents"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/bengaluru-based-kogo-launches-worlds-first-ai-agent-store\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10105166,"title":"AIM Top Ranked PG Data Science Programs (Online\/Hybrid) – 2023","content":"Online certification courses in Data Science have seen a significant rise in popularity to cater to the busy schedule of working professionals while not letting them compromise on upskilling themselves. This has become challenging in the decision-making process for students. To aid in this, Analytics India Magazine (AIM) has conducted a comprehensive survey to rank top postgraduate (PG) online or hybrid courses in Data Science in India. This initiative has been going on for the last nine years. The rankings are segregated based on the mode of delivery—either on-campus or online\/hybrid. The latest report highlights the best online or hybrid PG data science courses in India in 2023. A well-ranked course demonstrates high and balanced scores across various parameters, guiding students in choosing the most suitable course for their needs. This report is not only beneficial for students but also provides valuable insights for institutes to gauge their standing, identify improvement areas, and guide policy decisions in both the public and private sectors. Incase you are looking for Full Time On-Campus programs, check our 2023 ranking here. The Methodology The methodology employed by AIM involves evaluating data science and analytics programs across five key parameters: Certification Value, Return on Investment, Program Success, Teaching & Curriculum, and Student Engagement. Each program’s performance is assessed through an overall index, derived from the average scores across these five sub-indices. The survey methodology included uniform evaluation criteria, with scores normalized on a 0 to 1 scale, and outliers capped to ensure fairness and accuracy in the rankings. The infographic below illustrates the hierarchy of sub-indices and the final index used in this analysis. 1. Upgrad’s PGP in Data Science & AI Affiliated to IIIT Bangalore, the PG Data Science course is comprehensive. The curriculum includes both recorded and live content delivered by highly qualified faculties and industry experts. The course is professional and industry-oriented as evident from collaboration with employers and sessions with industry speakers. It’s divided into a common curriculum covering essential Data Science skills and specialization modules in areas like Business Analytics and Natural Language Processing. The program emphasizes practical application through a capstone project and uses an external platform for examinations, ensuring a robust and inclusive learning experience without the need for prior coding experience. Learn more about the program here. 2. REVA’s M.Sc in Business Analytics REVA University’s MS in Business Analytics program is popular for its strong industry ties. The curriculum, updated with the latest technologies, emphasizes experiential learning, including mandatory global certifications. With a faculty boasting PhDs and industry experience, the program maintains high academic standards. Two capstone projects and a publication requirement foster practical skills and research contributions. Alumni involvement in placements, a dedicated support team, and inclusive admission criteria further contribute to its appeal, making it a top choice for data science aspirants seeking a comprehensive and globally recognized education. Learn more about the program here. 3. Ivy Professional School’s PG Certificate  in DS with ML & AI Ivy Professional School’s collaboration with IIT Guwahati offers a sought-after data science program. The curriculum, designed with IIT Guwahati professors, includes 40+ industry projects, ensuring theoretical and practical competence. Learners benefit from industry internship projects with Fortune 500 companies fostering real-world experience. IBM and NASSCOM accreditations, low trainer-to-student ratios, and a “Learn – Apply – Assess – Re-learn” model enhances its appeal, making it a comprehensive and industry-aligned choice for data science enthusiasts. Learn more about the program here. 4. Simplilearn’s PG Certificate Course in Data Analytics The Professional Certificate Course in Data Analytics, offered in collaboration with IIT Kanpur, is highly sought after by data science and analytics aspirants for its comprehensive curriculum and practical focus. Emphasizing real-world experience through hands-on exercises, capstone projects, and industry internships, the program enhances employability and job readiness. Taught by experienced faculty and industry experts, it provides valuable insights and networking opportunities, potentially leading to job placements. The prestigious IIT Kanpur certification adds to the program’s appeal, ensuring a strong Return on Investment (ROI). The blended structure of theoretical and practical learning, coupled with a maximum of 80 learners per batch, makes it an attractive choice for career advancement in data analytics. Learn more about the program here. 5. TimesPro’s PGCP in DS & ML Affiliated to IIT Roorkee, TimesPro’s Professional Certificate Program in Data Science and Machine Learning (PGCP DSML) stands out due to its strong focus on practical learning and industry relevance, it offers a robust curriculum supported by industry expert sessions and an advisory board. The program’s career support, including personalized guidance, resume tools, and industry insights, enhances learners’ employability. A diverse faculty team with a blend of academic and industry experience contributes to a holistic learning experience. Incorporating live sessions, hackathons, campus immersion, and continuous assessments, the program ensures comprehensive skill development, making it a preferred choice for data science and analytics aspirants. Learn more about the program here. 6. Edvancer’s Advanced Certification in AI & ML With a strong focus on placement assistance and career mentorship, the course conducts regular workshops, and sessions by industry experts, and provides hands-on support for resume preparation and interview readiness. The collaboration with IIT Kanpur and partnerships with leading companies ensure industry relevance. Taught by full-time professors and experienced industry experts, the program’s 70% hands-on approach equips students with practical skills through assignments and industry-level capstone projects. Learn more about the program here. 7. Upgrad’s PGP in ML & AI (Executive) Upgrad in collaboration with IIIT, Bangalore, offers this course that not only ensures theoretical learning but also provides real-world experience along with a robust 360-degree career support system. The faculty pool consists of industry experts and accomplished academics, contributing to a comprehensive learning experience. The program is structured into common and specialization phases, covering key roles in ML Ops and Generative AI. Learn more about the program here. 8. Simplilearn’s PGP in Data Science The Caltech CTME Data Science program, offered through Simplilearn, distinguishes itself with a remarkable ROI. The practical orientation of the program as evident through project-based learning across various industries helps learners gain real-world experience. The IBM Partnership enriches the program with masterclasses and certificates. The comprehensive curriculum facilitated by experienced instructors not only helps students understand concepts well but the program also ensures that students get avenues to showcase their skills effectively. Learn more about the program here. 9. Timespro’s PGCP in AI & DL The program is affiliated to IIT Guwahati. It offers robust career support, including personalized guidance, cutting-edge resume tools, and regular industry insights. With a comprehensive Learning Management System (LMS) and over 40 case studies and projects, it ensures a holistic learning experience. Industry expert sessions featuring professionals from renowned organisations enhance practical knowledge. The program offers a unique blend of live sessions, hackathons, campus immersion, and industry mentorship. The eligibility criteria, continuous assessment model, and mandatory capstone project make it a well-rounded course. Learn more about the program here. 10. CloudxLab’s PG Certification in AI, ML & DS CloudxLab’s PG Certification Course offers a remarkable Return on Investment (RoI) driven by a curriculum aligned with industry needs, experienced instructors, robust placement support, and a strong alumni network due to its affiliation with IIT Roorkee. Strategic collaborations with industry giants like Google, Microsoft, and IBM provide students with real-world exposure, internships, and invaluable insights. The faculty, comprising full-time instructors, industry experts, and IIT professors, delivers a comprehensive learning experience. The program’s well-balanced structure, emphasizing both theory and hands-on skills, along with rigorous evaluation methods, positions it among the top data science programs in India. Learn more about the program here. 11. Edvancer’s Advanced Certification in Data Analytics The program stands out due to its holistic approach and industry collaboration. The course provides robust placement assistance, mentorship, and workshops on career guidance. Affiliation with IIT Kanpur helps students network with diverse and experienced alumni. Collaborations with renowned companies ensure real-world relevance and multiple job opportunities for students. Taught by full-time professors and industry experts, the program offers a balanced blend of theory and hands-on experience. With a 70% hands-on focus, students work on practical assignments and capstone projects, making them job-ready. A high graduate placement rate underscores the program’s effectiveness. Learn more about the program here. 12. Timespro’s PGCP in Decision-Making Using DS TimesPro in collaboration with IIT Roorkee conducts this program that offers extensive career support, personalized guidance, and cutting-edge tools. The comprehensive Learning Management System, industry insights, and 24\/7 support contribute to a robust learning experience. The continuous assessment model ensures a thorough understanding, while the capstone project offers real-world application. Active engagement with experienced professionals and a strong focus on practicality through hackathons, campus immersion, and industry mentor sessions underscores its industrial relevance. Learn more about the program here. 13. Ivy Professional School’s Diploma in DS, ML, AI & Big Data Ivy Professional School’s Data Science program in collaboration with IIT Guwahati is a premier choice for data science aspirants. It offers a comprehensive learning experience with industry projects from major companies like UBER and Accenture. The partnership with IBM, NASSCOM accreditation, and Fortune 500 collaborations enhance its industry relevance. With a low trainer-to-student ratio, expert faculty, and a unique “Learn-Apply-Assess-Re-learn” model, the program ensures holistic skill development. The program is focused on practical learning through projects, internships, and assessments to groom students as proficient data professionals. Learn more about the program here. 14. Orangetree Global’s BI & Business Analytics PG Certification With a curriculum aligned to industry needs, small batch sizes for personalized attention, and a seasoned Placement team, the program structure and curriculum prioritizes quality interactions. The extensive network of tie-ups with leading companies, corporate training programs, and collaborations with prestigious institutions contribute to a holistic learning experience. A high faculty-to-student ratio ensures each individual gets the required and adequate attention for professional upskilling. The comprehensive curriculum that includes several real-world case studies and a stringent certification process ensures, makes it is a preferred choice among students. Learn more about the program here.","excerpt":"To aid in this, Analytics India Magazine (AIM) has conducted a comprehensive survey to rank top postgraduate (PG) online or hybrid courses in Data Science in India.","categories":["AI Highlights"],"tags":["Data Science"],"author_name":"AIM Media House","publish_date":"2023-12-19T15:48:56","publication_year":"2023","word_count":1656,"keywords":["data science","machine learning","GCP","AI","ML","RAG","Aim","analytics","generative AI","Data Science","R"],"extracted_tech_keywords":["AI","machine learning","ML","data science","analytics","generative AI","Aim","RAG","GCP","R"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/aim-top-ranked-pg-data-science-programs-online-hybrid-2023\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10098188,"title":"GitHub Does Not Necessarily Get You a Job","content":"If you are a developer who is looking for jobs, anyone and everyone would suggest you one thing – to build up your GitHub profile through contributing to projects on a regular basis as much as you can. Well, this is not absolute anymore. In certain cases these days, GitHub contributions are making people lose out on jobs while they may be making bucks through the platforms. JLarky, a developer who works at Fogbender, posted a few days ago on X that he is looking for job opportunities as he might be leaving the role. Few days later, a lot of people have been sharing their GitHub profiles, sharing how many contributions they have made in the last year. JLarky also posted his profile, which showed that he has been making numerous contributions, but is still not able to land a job. PSA: my GitHub looks like this and no one hired me yet pic.twitter.com\/g2Lq3ZTuFH— JLarky (@JLarky) August 6, 2023 Interestingly, a user points out that the reason he thinks should land him a job is the exact reason that is not landing a job. Why would a company want to hire someone who gives away all the code for free? Isn’t open source the way forward? Developers earn from $5 to $30,000 from a Github repository every month. A lot of times, people who post on GitHub have been regarded as developers who already have a full time job. Funnily enough, people say that job is actually about someone who has very less experience and has too much free time on their hands. Moreover, it is very easy to copy code and pad it to your GitHub profile and fake your contributions to open source. On the other hand, it is very necessary to make your presence and capabilities visible online to be hired by companies. So even if you are working at a company, if you ever wish to switch, how else will companies know you’re a “10X Super Duper Hyper software developer”? The more you contribute to open source projects, the more recognition you will get, and thus get noticed by hirers. While that is true, the other side of this is you don’t really  need a GitHub presence anymore to get a job. People who have contributed barely on GitHub have been able to get jobs easily, when compared to people with high rated profiles. But the reason cannot be said is just that. As soon as you join a company, you realise that a lot of projects that you would be working on would not make it to your public profile, simply because it is a closed source project of the company. That is simply because code is valuable. Some companies won’t even let you use their private repositories. In the current era of AI models, a lot of companies do not want their codes to be used by AI models for training auto-code platforms, just like GitHub Copilot. Takes time, but you get what you want To an exciting note, a lot of companies started out as open source software. For example, Redis, the database company started out as an open source repository on GitHub, then eventually converted it into a company that offers several services through their premium offerings. There are a lot of examples of startups that are being born out of open source and are now minting money such as MindsDB, MongoDB, Kubernetes, Hoppscotch, Kafka, and so on. They are even hiring for roles. On the other hand, there are many high profile developers who do not care about acquiring jobs. For example Antonio Cheong, a developer who released an open source Bard API on GitHub for free by reverse engineering it, proudly says that he is an open source developer. He continues to contribute on GitHub and has since then built a lot of open source projects like ChatGPT open source, vectordb, and many more. The best part is that maintaining a GitHub profile by contributing to open source actually gets recognised by big tech companies. So while people might be landing jobs with ugly GitHubs, if a developer waits for long and keeps building his profile, they might end up in a company like OpenAI, Google, or Meta. Thus, while forgetting about GitHub profile is not the answer, it is also not smart to completely rely on it. It is great to contribute to open source, but it is ideal to learn a lot of other skills that the market, which is highly competitive right now, requires right now if you need a job. However, there are platforms like MachineHack that are offering jobs through their platforms and also let you build your profile based on how many hackathons and contributions you take part in, something that is missing for GitHub.","excerpt":"Why would a company want to hire someone who gives away all the code for free?","categories":["Global Tech"],"tags":["AI Jobs","Developer Jobs","Open Source"],"author_name":"Mohit Pandey","publish_date":"2023-08-07T15:06:40","publication_year":"2023","word_count":802,"keywords":["Go","ChatGPT","Open Source","OpenAI","AI","MongoDB","R","AI Jobs","Git","Kafka","Developer Jobs","kubernetes","Redis"],"extracted_tech_keywords":["AI","ChatGPT","OpenAI","kubernetes","Kafka","Redis","MongoDB","R","Go","Git"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/github-does-not-necessarily-get-you-a-job\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10007228,"title":"The Top Fraud Analytics Startups In India","content":"Fraud and abuse are constant challenges for online businesses. The digital explosion has brought many complex and cross channel risks to business than ever before. Typically, fraudsters attempt to get past various cross-channel and security protocols with stolen data and credentials obtained through sim swaps, man-in-the-middle attacks, phishing keylogging or password guessing attacks. They can also buy sensitive data from the darknet with your data housed in separate silos or residing with the third-party digital platforms. Apart from that, frauds happen with credit-based financial businesses with deliberate loan delinquency and similar malicious behaviours from some users. Such instances can create severe financial and reputational damages for companies and, therefore, it is important to get real-time fraud analytics across digital channels. India being one of the largest digital economies in the world, there are many firms launching analytics-based fraud solutions in the market for not just financial companies but across different industries. In this article, we take a look at the top five new-age firms using advanced analytics to fight fraud and business risk. Clari5 Founded in 2006, Bengaluru-based Clari5 is considered one of the market leaders in fraud detection and sells its financial crime risk management software product. The company serves Tier-1 banks in 15+ countries and processes data generated by over 10 billion transactions for real-time, cross-channel enterprise fraud management and anti-money laundering. It also has the world’s largest fraud analytics implementation, with 200 million financial accounts at a single site. The firm uses intelligent models based on neural networks, time series and complex analytics to provide insights into fraud and risks. In 2011, the company raised Series A funding of $4 million from investment firm JAFCO Asia and also received funding from Microsoft Accelerator. Also In January 2015, the company collaborated with EY to help banks fight enterprise fraud by linking the forensic capability of EY’s fraud investigation and dispute services (FIDS) practise, and Clari5’s cross-channel fraud management to present specialised solutions. Clari5 also partners with Microsoft, RedHat, Cisco and Intel to co-provide real-time fraud management and customer experience solutions to financial institutions. Simility Hyderabad and Palo Alto-based Simility was founded in 2014. It provides cloud-based and on-premise fraud detection software solutions. The main product is an adaptive fraud prevention solution which delivers enterprise-level performance and scalability. The adaptive fraud solution identifies complex fraud and abuse in real-time by ingesting any type of data – structure, unstructured and third-party data. Together with a complex device history plus new data feed, their product can be operational and provides value in just a few hours without requiring assistance from developers. It uses machine learning, giving users deep insights on user behaviour. It also uses ML to feed data back into an instant data label for optimal cross-channel and fraud detection capabilities. The rules and models can be easily updated without the need to write a single line of code. It uses the simple user interface that lets data analysts easily configure rules which can then give a 360-degree view of the customer with drill-down analysis, decision workflows and network graphs. In 2018 PayPal Holdings acquired the startup for $120 million in cash. The company had raised around $25 million in funding prior to that also. Razorpay Thirdwatch Last year, Razorpay acquired Gurugram-based fraud analytics startup Thirdwatch. Today the firm is one of the most prominent Indian startups that leverage AI for e-commerce fraud analytics. On the Razorpay platform, the team captures billions of events each day and uses that data to feed into AI models.  This is then used to prevent fraud in digital and e-commerce transactions. As the payment gateway and as a payment service provider, Thirdwatch assists Razorpay to prevent losses of its merchants when it comes to fraud in e-commerce transactions. Razorpay Thirdwatch has captured 200 plus parameters from e-commerce websites and applications using plugins, SDKs and REST APIs. It uses these parameters to generate over 300 features associated with user behaviour such as location, IP, device, user journey events, historic transactional data, etc. Within milliseconds, the platform can predict whether it is a fraudulent or genuine transaction. If the transaction is found fraudulent, the particular merchant is notified. AdvaRisk Mumbai-based AdvaRisk provides products on fraud prevention, detection, investigation and funds recovery. The early-stage startup raised $700,000 in a seed round led by Sprout Venture Partners, along with participation from SEA Fund. Founded in 2016 by Vishal Sharma and Rahul Metkar, AdvaRisk has built an AI-powered platform for fraud prevention, detection, and recovery in the corporate loan portfolio of financial institutions. The company said it is working towards becoming the most reputed and trusted solution for banks, NBFCs, and corporates in India. Its AdvaSmart proprietary monitoring platform highlights actionable credit negative transactions connected to borrowers facilitating early detection of possible frauds. This solution is trusted by major financial institutions in India. Other than that, the firm’s AdvaNPA helps maximise fund recovery by highlighting undisclosed data and unidentified patterns linked to a defaulter by leveraging 600+ data sources. It has led to Rs 20,000+ crs recovery for banks and NBFCs so far, according to the firm. TrustCheckr TrustCheckr was founded in March 2017 by IIM-Lucknow graduate Ramesh and IIM-Calcutta alumnus Praveen Raj. The startup provides cloud-based APIs that enables businesses to identify fraud, fake and bot profiles on their digital platforms using artificial intelligence. The startup is working to help businesses enrich customer data to make key decisions like fraud detection, identity verification, defaulter prediction, and purchase propensity scoring. For businesses, it provides algorithms to find the TrustScore of the user and eliminate them at the stage of onboarding and improve the sales funnel. Custom AI\/ML engines calculate with a ‘trust score’ based on 100+ parameters of a person’s profile such as social profiles, location, education, employment details, interests, purchase, spending and credit history, etc.","excerpt":"Fraud and abuse are constant challenges for online businesses. The digital explosion has brought many complex and cross channel risks to business than ever before. Typically, fraudsters attempt to get past various cross-channel and security protocols with stolen data and credentials obtained through sim swaps, man-in-the-middle attacks, phishing keylogging or password guessing attacks. They can […]","categories":["AI Trends"],"tags":["corporate analytics platform","Startups"],"author_name":"Vishal Chawla","publish_date":"2020-09-11T15:00:23","publication_year":"2020","word_count":967,"keywords":["Go","artificial intelligence","machine learning","AI","neural network","corporate analytics platform","ML","RAG","analytics","Startups","R","fraud detection"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","neural network","analytics","RAG","fraud detection","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/the-top-fraud-analytics-firms-in-india\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10022801,"title":"Commendable Or Mere Eyewash: Tech Startups On Google Commission","content":"In a recent move, following Apple’s announcement of cutting down App Store’s commissions, Google announced that it would also reduce 30% to 15% commission from app developers earning $1 million revenue each year. The news was released soon after Google’s last announcement of levying 30% on all in-app purchases; however, after strong opposition from the business industry, the giant was compelled to keep the same aside till April 2022. On one side, some tech startups are applauding the development while many are labelling the same as mere eyewash. Each side has its own set of views and counterviews, but before delving into the same, let’s have a quick look at two of the comments and overview the matter in short. Two sides of the same coin While talking to Analytics India Magazine, Rajat Shikhar, the Founder & Chief Product Officer at DealShare, appreciated Google’s recent move of reducing their Play Store fees by 50%, “especially for a country like India, which has a growing startup ecosystem.” He stated that these reduced prices, along with supporting the entry of new players, will also boost existing application developers, boutique firms, and early-stage startups to use the saved funds for other crucial aspects like “scaling their operations and maximising their growth efforts.” It will also boost small scale developers by providing them with more working capital and assisting them in growing in the market, which is very highly competitive, he further added. On the other hand, while talking to the media, Bharat Matrimony’s Janakiraman is of the view that it is consumers and companies that (should) decide which payment gateway to use. “Companies are already spending crores of rupees on Google ads; then there is a charge on the income made be it 15% or 30%. My view is that app ecosystems should be independent of Google or Apple,” he said. Why Worry? The fact of the matter is the Play Store fee reduction is only available to developers who make less than $1 million per year. However, if the developers receive more than a $1 million fee for in-app purchases and downloads in an app in a year, Google will end up charging them a 30% fee. Among the mobile operating systems in India, Android held a share of 95.23%, followed by Apple, with just 3.2%, as per a report. Moreover, the dominance of Google in the android market is enormous, as nearly 99% of smartphones operate on Google’s Android mobile operating system. Google representatives are thus refuting the claims. They believe that the money saved by these startups because of a lower rate of commission can further be utilised to scale productivity, invest in R&D, and attract new startups to join the platform with less money to pay. The real bone of contention lies in the fact that these tech giants, including Google, make it mandatory to use their play store billing for any in-app purchases made on respective platforms. In addition to that, they do not allow payment systems provided by other merchants such as Razorpay, PayU charging as minimal as 1.5 to 2% commission. Consequently, it is believed that the monopoly can thus cripple the growth of tech startups severely. Moreover, there are apprehensions that these tech-giants act like a super-government and are making decisions independently without even understanding India’s startup ecosystem’s nascency. Good capital and regular investments are essential for startups to prosper further and extend their footprints. Wrapping Up Indian startup ecosystem cannot run within a defined set of parameters imposed by a handful of big players. The subscription-based model is yet to develop in India, and consumers are not paying for the same. Thus, the startup ecosystem believes that the high commission fees will be massively eating up their revenue. In order to reach a long-term solution, Indian startup stalwarts, along with the government, need to raise their concern and circumvent.","excerpt":"Google announced to halve its fee for in-app purchase. Understanding why some consider the step as good, while others disagree is critical.","categories":["Global Tech"],"tags":["App Development","Google","Startups","Technology"],"author_name":"kumar Gandharv","publish_date":"2021-03-23T19:06:06","publication_year":"2021","word_count":645,"keywords":["Go","API","programming_languages:R","AI","RPA","App Development","Aim","Technology","ViT","analytics","Google","Startups","R","startup"],"extracted_tech_keywords":["AI","analytics","Aim","R","Go","API","ViT","RPA","startup","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/commendable-or-mere-eyewash-tech-startups-on-google-commission\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10130121,"title":"This Indian AI Startup is Creating 3D Models for AAA and Indie Games","content":"The world of 3D model creation is extremely capital-intensive, and India barely has any company building this. As indie game development gains prominence, there is a dire need for an Indian company to enter this space. Recognising this void, Google for Startups Accelerator: AI First Program has decided to support 3DAiLY, a 3D model creation company that makes ultra-realistic production-ready assets using generative AI. The company is also one of the first to make these models accessible on its community platform for AAA and indie games alike. Speaking with AIM, the CEO and founder, Harsha P Deka, shared his journey, which began in 2010 with a personal loss that fueled his drive to innovate. While studying computer science in Canada, Deka received the devastating news of his best friend’s death. “I wanted to create a 3D model of my friend to give to his family as a memento,” Deka recalled. However, he soon discovered the limitations of the technology at the time. Despite reaching out to multiple gaming and animation studios, none could produce a 3D model from a photograph. The Birth of 3DAiLY Undeterred, Deka delved into the intricacies of 3D modelling, understanding the immense time and effort required to create high-quality models. By 2014, he had founded an animation studio, encountering firsthand the industry’s challenges. “Creating a game, Belegon, took us a year and made us realise the massive funding needed for such ambitious projects,” Deka explained. His experiences underscore the complexities of 3D modelling, particularly for animation and gaming. A pivotal moment came in 2015 when Deka encountered a 3D scanning setup in a US mall, which produced 3D-printed miniatures from in-person scans. “I asked if they could do it from a photograph, and they said it wasn’t feasible,” Deka said. This gap in the market inspired him to create 3DAiLY, a company that could generate 3D models from photos. By leveraging AI, Deka set out to make high-quality 3D modelling accessible and efficient. 3DAiLY has since evolved into a leader in the 3D modelling industry, creating a comprehensive library of human models and developing proprietary AI technology. “We built our own foundation model using the data we’ve collected over the years,” Deka stated. Its approach combines artists’ intuition with AI, ensuring that the models are production-ready and of the highest quality. Deka said that unlike other AI tools such as Ready Player Me or Sloyd that produce low quality meshes, 3DAiLY’s models are fully rigged and animatable, compatible with various gaming engines like Unity, Unreal, and CryEngine. “What MetaHuman did for Unreal, we’re doing for multiple engines,” he emphasised, adding that the rest of them are building tools and not a platform which has a built-in marketplace. Overcoming Industry Challenges Establishing such a pioneering company in India has not been without its challenges. “The space is not well understood, and getting VC funding is tough,” Deka noted. Despite these hurdles, 3DAiLY has garnered significant traction, with indie game developers and AAA studios alike adopting their technology. Offering models at a fraction of the traditional cost, the company is breaking down barriers to high-quality 3D content creation. Looking ahead, Deka envisions expanding 3DAiLY’s capabilities to include design-to-3D modelling tools, enabling artists to transform their ideas into tangible models. “We’re building an ecosystem where artists can create assets and participate in an SDK, benefiting from in-game asset sales,” he explained. Currently, the platform has around 12,500 artists from 150 countries. “Artists are critical to the success of games, yet they often earn the least,” Deka pointed out. By offering tools that significantly reduce production time and costs, 3DAiLY aims to empower artists, allowing them to focus on creativity while the AI handles the heavy lifting. This approach not only enhances productivity but also ensures that high-quality 3D models are accessible to all.","excerpt":"Offering models at a fraction of the traditional cost, the company is breaking down barriers to high-quality 3D content creation.","categories":["AI Startups"],"tags":["Startups"],"author_name":"Mohit Pandey","publish_date":"2024-07-24T18:09:17","publication_year":"2024","word_count":634,"keywords":["Go","API","AI","RAG","Aim","ViT","generative AI","GAN","Startups","R","startup"],"extracted_tech_keywords":["AI","generative AI","Aim","RAG","R","Go","API","GAN","ViT","startup"],"url":"https:\/\/analyticsindiamag.com\/ai-startups\/this-indian-ai-startup-is-creating-3d-models-for-aaa-and-indie-games\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10168123,"title":"‘AI is Part of All Deal Conversations’: Wipro Secures 17 Big Wins Worth $1.8 Bn in Q4","content":"Wipro posted a net profit of $417 million in Q4 FY25, a 26% increase compared to the same period last year and a 6.4% increase quarter on quarter. The revenue from operations, on the other hand, was nearly flat for the quarter, growing only slightly by over 1% YoY to reach $2.3 billion, an increase of 0.8% QoQ. For the year ending March 31, the IT consulting firm reported gross revenue of $10.4 billion, reflecting a 0.7% decline year-over-year. However, net income for the year rose by 19% to $1.5 billion. Wipro secured large deal bookings worth $5.4 billion during the year, reflecting a 17.5% YoY increase. However, overall bookings declined by 3.8% YoY to $14.3 billion. “We closed FY25 with two mega deal wins, an increase in large deal bookings, and growth in our top accounts…As clients remain cautious in the face of macroeconomic uncertainty, we’re focused on partnering closely with them while staying committed to consistent and profitable growth,” Wipro CEO and MD Srini Pallia said. In a post-earnings press conference, Pallia said that in Q4, Wipro closed 17 large deals with a total value of $1.8 billion across markets and sectors. “In fact, for the full year, we closed 63 large deals for a total value of $5.4 billion, a year-on-year growth of 17.5%.” The company fell short of its constant currency revenue guidance for the January–March quarter, reporting a 1.2% sequential decline. This was below its forecast range of -1% to +1% growth. Looking ahead, Wipro anticipates Q1 FY26 revenue to fall between $2.505 million and $2.557 million, indicating a projected decline of 1.5% to 3.5%. Speaking on the recent US tariff announcements, Pallia said that Wipro expects clients to take a more measured approach, especially on two spend areas—large transformation programs and discretionary spend. For example, Pallia said one of the clients asked to pause a digital transformation project to reduce their spending. “It was not a cancellation, but it was paused because they wanted certainty.” Pallia said that clients are asking for cost optimisation while boosting productivity, both in the short term and long term. “I think in the next few weeks and months, we will, you know, have better clarity in terms of how it will progress,” he said. Aparna Iyer, CFO, said, “Our net income grew 6.4% sequentially in Q4 and 18.9% for the full financial year. Cash flow continued to be robust in Q4, resulting in net operating cash flow generation of almost $2 billion for FY’25, which is 128.2% of our net income.” Regarding hiring, Saurabh Govil, CHRO, said, “We ended FY25 with approximately 10,000 young freshers hired.” He added that Wipro will continue to onboard people while ensuring that they can be deployed on tasks, ensuring utilisation while hiring freshers. Govil also spoke about Wipro’s centre of excellence for upskilling students with AI, which he claims are built at 50 campuses in India. What About AI? “As we all know, AI has been part of all deal conversations for a while. But this year, it has actually become central to almost every opportunity, big or small,” Pallia said. When asked about the cannibalisation of deals because of generative AI, Pallia said that Wipro is now incorporating generative AI into all its solutions. “GenAI was not part of the earlier deals, but in the new deals, we’re going to infuse GenAI.” “Some of the clients would want to adopt GenAI, and some are worried about the guardrails in terms of how China is going to deploy [the AI models],” Pallia said. In 3-4 quarters, Wipro will have better clarity on that. “But we are going to infuse AI into a solution before we go to the client.” Biswajit Maity, senior principal analyst at Gartner, told AIM that a strong deal pipeline suggests a positive outlook for future growth for Wipro. “Most of its revenue comes from sectors like BFSI, consumer, EMR, technology & communications, and healthcare,” he said. “However, it’s important to note that Wipro must address its attrition rate, although it has slightly improved compared to the previous quarter. As workload pressures increase, maintaining workforce stability is critical to preserving service quality and meeting delivery expectations—any challenges in this area could pose risks to its future growth trajectory,” Maity added. Pallia revealed that a leading Indian private bank has expanded its partnership with Wipro as part of a business-focused digital transformation. “We will provide AI-powered solutions to strengthen compliance management addressing the critical need for regulatory compliance, while also enhancing the overall experience for this bank.” In March, Wipro secured a 10-year contract worth $650 million from Phoenix Group, a UK-based insurance firm that was previously a client of TCS, to overhaul its life and pension administration by modernising the ALPHA platform using AI, automation, and cloud technologies. A major North American financial institution brought Wipro on board to streamline its technology infrastructure through an AI-enabled global delivery model that consolidates vendors and improves efficiency. This deal came after Wipro disclosed a five-year agreement valued at $500 million with a US-based communications company in June 2024. Wipro also strengthened its relationship with a US-based card services provider by modernising its payments platform to improve transaction security and scalability through AI. Similarly, an American energy major extended Wipro’s engagement to manage business applications across its oil and gas value chain using AI-powered NextGen AMS for improved automation and performance. A North American parcel delivery firm tapped into Wipro’s AIOps capabilities to enhance ticket resolution and reduce outages while improving scalability and data sovereignty. This is similar to how TCS, in its earnings call, said that the focus on deals was the biggest part of the year, resulting in a total revenue of $30 billion. Though no large deals were reported, small deal wins surged to an all-time high of $12.2 billion, up from $10.2 billion in the previous quarter. Infosys is expected to release its results on Thursday after the close of the market. Past Announcements Quoting internal emails shared by employees, Mint reported that Pallia is focusing intensively on increasing the company’s profitability and might be looking to lay off around 1,000 employees before June 2025. In its last earnings report for Q3 FY25 in January, Wipro reported a net income of $392 million, a 4.5% increase QoQ and 24.5% YoY. But the YoY increase was after the company reported a decrease of 12% YoY in FY24. Pallia said that the company has shifted its focus on agentic AI for customer service and supply chain management in the coming quarters and reported 17 large deals worth $1 billion for the same. In February, Wipro committed $200 million to its venture arm, Wipro Ventures, to accelerate investments in early to mid-stage startups. This is the fourth round of funding since Wipro Ventures’ launch in 2015, highlighting the company’s decade-long focus on supporting emerging technologies. One of the successful investments of the venture arm of Wipro was Avaamo. While speaking with AIM, Sriram Chakravarthy, CTO and co-founder of Avaamo said that with its LLaMB offering, Wipro was able to replace dozens of internal tools for its 2 lakh plus employees.","excerpt":"For the year ending March 31, Wipro reported gross revenue of $10.4 billion, reflecting a 0.7% decline year-over-year.","categories":["IT Services"],"tags":[],"author_name":"Mohit Pandey","publish_date":"2025-04-16T17:50:46","publication_year":"2025","word_count":1189,"keywords":["Go","GenAI","agentic AI","AI","ML","Scala","Git","Aim","generative AI","R"],"extracted_tech_keywords":["AI","ML","generative AI","GenAI","agentic AI","Aim","R","Go","Scala","Git"],"url":"https:\/\/analyticsindiamag.com\/it-services\/ai-is-part-of-all-deal-conversations-wipro-bags-17-large-deals-worth-1-8-bn-in-q4-fy25\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10044880,"title":"Register For This Webinar: Do You Think You Can Analytics?","content":"Businesses are accelerating their digital transformation journeys. To stay ahead, the demand for analytics is surging across industries. According to NASSCOM, the Indian analytics industry is projected to hit a $16 billion market by 2025. And, between the years 2021 to 2026, the industry is expected to grow at a CAGR of 35.1%. This incredible growth has resulted in increased demand for technology, data science and analytics professionals. The increasing demand has given rise to abundant opportunities for professionals who wish to build a career in analytics. As a result, freshers and experienced professionals across varied fields like mathematics, statistics, engineering, technology, and more, are keen to explore a career in analytics. To guide professionals and enthusiasts with the necessary know-how to embark on a career in analytics, Analytics India Magazine, in association with Ugam, a Merkle company, is hosting a live webinar. The webinar will be hosted by Navin Dhananjaya, Chief Solutions Officer at Ugam, a Merkle company. A 25-year analytics and technology veteran, Navin leads solutions management at Ugam and mentors professionals to help them build a thriving career in analytics. Navin’s role of mentoring professionals is crucial as the demand for analytics continues to grow. Witnessing the growing demand, Ugam recently announced plans to hire over 1,300 analytics and technology professionals in India. Register Now The webinar will cover: Skills required to excel in the field of analytics.How and when to make a career switch if you are not from an analytics background?How to stay relevant, and what kind of opportunities are out there?How to identify the right company fit? Who should attend: Students from engineering\/technical backgroundGraduates of Maths, Science, Economics or StatisticsData science, analytics & AI enthusiastsAspiring data science candidatesWorking professionals interested in the data science & analytics domainData science & analytics professionals looking to pivot Register Now About the speaker: Navin Dhananjaya — Navin joined Ugam as their Chief Solutions Officer in 2013, where he leads solutions management. He is responsible for conceptualising and developing new, innovative solutions and taking them to market. Navin also oversees the client delivery for overall analytics, including media, audience, site analytics, and digital commerce service lines at Ugam. He is the brain behind conceptualising and building Ugam’s JARVIS – a cognitive computing system. Navin is passionate about creating an environment that promotes personal and professional growth. His understanding of analytics and technology gives him a unique edge to identify the right talent from all backgrounds and mentor them to build successful careers. He is truly Ugam’s gardener for talent. He brings over 25 years of experience and has held senior roles in companies such as Mu Sigma, Infosys, and Manthan. Date: August 10, 2021 Time: 6:00 PM to 7:00 PM (IST) Register Now","excerpt":"To guide professionals and enthusiasts with the necessary know-how to embark on a career in analytics, Analytics India Magazine, in association with Ugam, a Merkle company, is hosting a live webinar.","categories":["Deep Tech"],"tags":["data science webinar","Tech Webinars You Must Attend","ugam","ugam data analytics","ugam solutions","webinar","Webinar for data scientists"],"author_name":"AIM Media House","publish_date":"2021-07-30T15:16:17","publication_year":"2021","word_count":454,"keywords":["data science","ugam","programming_languages:R","AI","Tech Webinars You Must Attend","ugam solutions","Webinar for data scientists","digital transformation","Git","data science webinar","analytics","R","webinar","ugam data analytics"],"extracted_tech_keywords":["AI","data science","analytics","R","Git","digital transformation","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/register-for-this-webinar-do-you-think-you-can-analytics\/","complexity_score":3,"technical_depth":7,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10161360,"title":"Now You Can Create Videos with See-Through Effects—Effortlessly!","content":"For years, creating professional-grade transparent effects—like smoke, water, or reflections—was a privilege reserved for large studios with extensive resources. Adobe’s TransPixar, developed in collaboration with the Hong Kong University of Science and Technology, is breaking those barriers. It’s an AI system designed to simplify and enhance the creation of RGBA videos, combining colour (RGB) and transparency (alpha channel). “Currently, no direct solutions exist for RGBA video generation, which remains a challenging task due to the scarcity of RGBA video data, with only around 484 videos available,” commented the researchers on the state of the field. The release of its public demo on GitHub can be accessed here. Value to Creators & Consumers The value of TransPixar lies in its ability to address the fundamental challenge of generating and integrating transparent effects seamlessly into videos. Traditionally, achieving such effects required specialised software, technical expertise, and significant time investment. TransPixar changed this dynamic, making these effects accessible through straightforward text prompts. This democratises visual effects by enabling smaller studios and independent creators to achieve professional-grade transparency effects without costly tools or extensive technical teams. This levels the playing field for industries such as advertising, gaming, and indie filmmaking. Additionally, by simplifying workflows, creators can focus more on storytelling and experimentation rather than being constrained by technical limitations. This release also makes real-time applications in gaming, augmented reality (AR), and virtual reality (VR) more feasible, pushing the boundaries of immersive experiences. Consumers are poised to enjoy higher-quality content across the board. From visually stunning ads and games to immersive AR\/VR experiences, TransPixar opens up new dimensions in storytelling. Richer, more realistic visual effects enhance emotional engagement and make content more captivating, whether it’s a film, game, or interactive environment. Challenges While TransPixar introduces significant advancements, it has limitations. The technology requires substantial computational power, which can limit scalability for larger projects. Its reliance on limited RGBA training datasets restricts the diversity of effects and its performance across different scenarios. The Technology Behind TransPixar TransPixar’s technical stack combines AI advancements with adaptations for RGBA video generation. It uses alpha-specific tokens to account for transparency, extending beyond traditional RGB models and enabling accurate effects like smoke or glass. Fine-tuned attention mechanisms handle transparency complexity while maintaining the strengths of RGB models, connecting traditional video AI with RGBA capabilities. Integration with existing AI video generation frameworks ensures it works with industry-standard tools, simplifying adoption.","excerpt":"This release enables real-time gaming, augmented reality (AR), and virtual reality (VR), improving immersive experiences.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","AI Video Generation Models"],"author_name":"Aditi Suresh","publish_date":"2025-01-13T16:52:08","publication_year":"2025","word_count":398,"keywords":["attention mechanism","programming_languages:R","AI","ML","Scala","Git","AI Video Generation Models","programming_languages:Scala","GitHub","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","ML","R","Scala","Git","GitHub","attention mechanism","programming_languages:R","programming_languages:Scala"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/now-you-can-create-videos-with-see-through-effects-effortlessly\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":29238,"title":"NITI Aayog Ties Up With Microsoft To Deploy AI Solutions In Farming, Healthcare","content":"Indian think tank NITI Aayog and tech giant Microsoft India have forged a partnership to leverage AI-based solutions in two sectors — agriculture and healthcare. Redmond giant will help the Indian think tank in pushing the use of AI, cloud and research and develop vertical expertise in new initiatives and solutions across agriculture, healthcare and environment. Satya Nadella led Microsoft has already made significant strides in advancing AI-based solutions to address challenges in Indian agriculture and healthcare with solutions in improving healthcare screening models and farm advisory services. Microsoft has taken concrete steps to address societal challenges and is working across the government, academia and entrepreneurs to help incubate technologies for societal impact, with special emphasis on cloud and AI platforms. According to Amitabh Kant, CEO at NITI Aayog who was cited in VCCircle, the inclusive development of AI in India requires the technology to be implemented across the sectors. He also emphasised the partnership will help public sectors benefit from the adoption of cloud and AI technologies and effectively scale the implementation of AI in medical screening and farm advisory services. Under the farm advisory services, the Redmond giant will help in increasing the crop yield especially in areas like yield and pest detection. The company will deploy remote sensing and image processing applications to assess weather data collection and also provide pest risk detection model to prevent attacks. In healthcare screening services, the company will build AI-assisted models to detect diabetic retinopathy screening models which will help in early risk detection.","excerpt":"Indian think tank NITI Aayog and tech giant Microsoft India have forged a partnership to leverage AI-based solutions in two sectors — agriculture and healthcare. Redmond giant will help the Indian think tank in pushing the use of AI, cloud and research and develop vertical expertise in new initiatives and solutions across agriculture, healthcare and […]","categories":["AI News"],"tags":["NITI Aayog"],"author_name":"Richa Bhatia","publish_date":"2018-10-15T08:30:00","publication_year":"2018","word_count":253,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","RAG","R","NITI Aayog"],"extracted_tech_keywords":["AI","RAG","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/niti-aayog-ties-up-with-microsoft-to-deploy-ai-solutions-in-farming-healthcare\/","complexity_score":2,"technical_depth":6,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10072604,"title":"The Time to Move towards Responsible CX is Now or Never","content":"According to a 2022 global study of over 23,000 consumers, 80% believe businesses need to improve their customer experience (CX). The same report warns that 9.5% of the revenue might be at risk due to bad CX. In the current digitised world, there are no two ways about the fact that businesses that truly stand out from their competitors are those that provide top-notch, delightful customer experiences. To be precise, CX is king! That’s resolved. But can we afford a king with unlimited powers? The answer is a thunderous NO. Are brands crossing the line? Understanding customers and strategizing products as per their needs enhances CX and accelerates the company’s revenue growth. However, on the flip side, we are on the verge of a sustainability crisis. To offer better CX to consumers, brands are crossing a line – the red line. To begin with, let me take you through a simple example. To target lower-income customers, single servings of foodstuff like soup and sauces are sold at low rates in tiny, multilayered packaging. Flexible packaging material is made with layers of different types of plastic that provide different qualities each. It becomes challenging to separate these different layers, and the components have little or no commercial value. Hence, most plastic sachets cannot be recycled, forming a huge chunk of our non-biodegradable waste. With the advancement of technologies like AI, companies are chasing customers aggressively. This calls for a balance between customer experience and environmental choices. So, what is the way out? Need to move towards responsible CX Today, businesses worldwide are experimenting with artificial intelligence, machine learning, and advanced analytics to boost CX.  However, many of these added advantages have a flip side that needs to be explored. First, the population of Gen Y is constantly on the rise, and so is the demand for digitisation. However, the race towards extensive digitisation is hurting energy efficiency. A clear illustration of the potential effects of digitisation on efficiency and demand may be seen in video-streaming services. Although it takes significantly lesser energy to download a video than it does to create, market, and buy a DVD, the accessibility and convenience make it easy to download and watch different content every night. The availability of digitalised services can unquestionably increase utilisation and the resultant energy demand. Second comes AI, Natural Language Processing (NLP), to be specific, is a darling of the retail sector, especially when it comes to automating customer engagement 24×7 throughout the year. Researchers from the  University of Massachusetts recently conducted a study to evaluate the life  cycle of training various well-known large NLP models in Amherst. To their surprise, the process can produce more than 626,000 pounds of Co2  equivalent, almost five times the lifetime emissions of the average American car (including the manufacture of the car itself). Additionally, personalisation engines let marketers decide what kind of experience is best for each customer or prospect based on their previous interactions, the current context, and their anticipated purpose. These engines assist marketers in identifying, selecting, customising, and delivering communications such as content, offers, and other interactions across customer touchpoints. As a result, we are moving  towards promoting too much consumerism. The need to create things rises along with the demand for them. This causes more pollution emissions, increased land use, deforestation, and finally, climate change. The way ahead With the adoption of ML models across industries, the need to gather and analyse enormous volumes of data, and thereby the demand to have larger data centres has boomed. There is a need to set boundaries. An aggressive and careless approach to enhance customer experience may be  beneficial in the shorter run, but disastrous in the long run. According to a survey by The Economist, consumers feel that companies and governments have equal responsibility to bring about beneficial environmental change. Online searches for sustainable products have increased by more than 70% on a global scale. The way that consumers interact with sustainable businesses has changed. This pattern is not exclusive to first-world nations. Concerns about global warming are also linked to consumer satisfaction in emerging and developing  nations. Now, it is for businesses to open their eyes wide open and move away from irresponsible CX to responsible CX.","excerpt":"According to a 2022 global study of over 23,000 consumers, 80% believe businesses need to improve their customer experience (CX). The same report warns that 9.5% of the revenue might be at risk due to bad CX. In the current digitised world, there are no two ways about the fact that businesses that truly stand […]","categories":["IT Services"],"tags":["climate change","customer experience","NLP","sustainability"],"author_name":"Gunjan Aggarwal","publish_date":"2022-08-11T12:00:00","publication_year":"2022","word_count":706,"keywords":["Go","artificial intelligence","machine learning","AI","sustainability","ML","Git","RAG","customer experience","NLP","analytics","climate change","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","NLP","analytics","RAG","R","Go","Git"],"url":"https:\/\/analyticsindiamag.com\/it-services\/the-time-to-move-towards-responsible-cx-is-now-or-never\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":11409,"title":"In Conversation with Deepak Ghodke, Country Manager, India at Tableau","content":"Deepak Ghodke, Country Manager, India, Tableau It’s largely credited for bringing data visualization into the mainstream. Seattle-based Tableau Software is  now the new gold standard for Analytics, Visualization and BI space. The company recently unveiled their product roadmap for next three years during the Keynote at TC 2016 featuring next-gen innovations – in-memory data engine with Hyper, Project Maestro, self-service data prep tool among other smart recommendations and collaborations. In a candid chat with Analytics India Magazine, Deepak Ghodke, Country Manager, India, Tableau talks about some of the major developments afoot in Tableau — why self-service analytics is the next big thing in BI and why the BI market in India is estimated to touch a whopping $213.8 million in 2017. Read more to find out about the latest trends in Big Data and Analytics that will emerge in 2017. [dropcap size=”2″]AIM [\/dropcap] Analytics India Magazine: Could you tell us about your role in India and the focus areas? [dropcap size=”2″]DG[\/dropcap] Deepak Ghodke: I am the Country Manager for Tableau in India and I am responsible for customer success and driving the growth of Tableau in India. Asia-Pacific is Tableau’s fastest growing region and it has been a very exciting and immensely satisfying journey in India for us till date. The response since our launch in India two years back has been tremendous from our customers across industries. We feel rewarded to be an enabler for organisations to build an analytic culture. An increasing number of customers and prospects are interested in adopting our offerings on a term or subscription basis. There is also growing appeal for our enterprise and OEM licensing models. These are the trends we will continue to encourage and embrace in the future. [quote]Some of our clients in India are Marico, Infosys, Wipro, Ashok Leyland, Eveready, Blueocean Technologies, Star Health and Allied Insurance, CRIF High Mark and eClerx etc.[\/quote] AIM: Your recently unveiled product roadmap at the Tableau conference was received with a lot of fanfare. Tell us about the product innovations at Tableau which will be rolled out in 2017? DG: Our mission is to help people see and understand their data, because we know data can empower people to achieve great things. To enable our customers to do their best work, they would need an analytics platform that allows them to make the most of all their data in their organisation. This platform should answer deeper questions and scale as usage increases all while keeping data secure. And that’s where Tableau come in. As we shared during the keynote at TC16, every part of our product roadmap is designed to empower our customers and their entire organisation to make better decisions faster with data. Here is what we have planned for the next three years. 1)  A new data engine for faster analysis – Hyper: As the volume of data grows exponentially, user expectations are also growing. Our customers need immediate responses to their questions. To solve this, we’re building a new in-memory data engine with Hyper, the fast database technology we acquired earlier this year. Hyper enables fast analysis on billions of records and near real-time data updates. It’s designed to simultaneously process transactional and analytical queries without compromising performance. Hyper will also enhance Tableau’s hybrid data model. Our customers still be able to connect live to over 50 different sources that Tableau supports such as Amazon Redshift, Google BigQuery, Snowflake, and Microsoft SQL Server, or choose to bring some or all of your data into Tableau with Hyper. The beta for Hyper will start in early 2017. 2) Project Maestro: It is  a self-service data-preparation tool. We know that getting data ready for analysis is a time-consuming and difficult process. And what we’ve heard from our customers is that there is an extended set of data-prep activities that our data stewards perform to support others in their organisations. That’s why we’re excited to announce a brand new data-prep product codenamed Project Maestro. 3) New data governance capabilities: Enterprises will soon govern their self-service analytics environment at scale with new functionality in Tableau Server to certify data sources, easily conduct impact analysis on sources and workbooks, promote content and write workflows with simple drag and drop gestures. These capabilities will be available in 2017 and beyond. 4) Next leap in analytics: Tableau continues to invest in making analytics easier for everyone. For example, Natural Language Processing will bring new ways to interact with data through human language. Tableau is also adding instant analytics, a new capability that will automatically provide contextual information as users interact with their data to help them find insights faster. These features will progressively available throughout 2017 and beyond. 5) Tableau Server for Linux: Tableau revealed a version of Tableau Server for the Linux platform. Now users of the open-source operating system, including governments of all levels, educational institutions and businesses of every size, will be able to leverage the power of self-service analytics. Tableau for Linux will be available in 2017. 6) New collaboration capabilities: Tableau showed upcoming capabilities to help users collaborate with each other and monitor the metrics they care about, enabling self-service at scale. Data-driven alerting will make it easier for people to stay on top of their data and be notified when key metrics exceed a specific threshold. Customers will be able to collaborate and discuss insights directly within an analysis to drive better business outcomes starting in 2017. 7) Smart recommendations: Tableau demonstrated future plans to add a new machine learning recommendations engine to its platform. Smart Data Discovery and Smart Recommendations are a key trend of the future and we are already getting ready to create this enabler for our customers. Algorithms will surface recommendations for workbooks and data sources that are trusted, highly used and contextually relevant to the individual workflow. Recommendations will be available in 2017. 8) New hybrid data connectivity for the cloud: Tableau showcased a new live query agent that will act as a secure tunnel to on-premises data. Data behind the firewall is easier to access and analyze with Tableau Online, Tableau’s SaaS managed service. This will be available in 2017. AIM: How is Project Maestro, the data prep product going to give your business a significant competitive edge? DG: We know that getting data ready for analysis is a time-consuming and difficult process. And what we’ve heard from our customers is that there is an extended set of data-prep activities that our data stewards perform to support others in their organisations. Most people still do data prep in Excel and data modelling is an expensive, rigid, procedure and may be too hard to implement. That’s why we’re excited to announce a brand new data-prep product codenamed Project Maestro. [pullquote]Project Maestro will make it possible for more people, from IT to business users, to easily prep their data with a direct and visual approach. You’ll instantly see the impact of the joins, unions, and calculations you’ve made, ensuring that you have exactly what you need before jumping into analysis.[\/pullquote] Project Maestro will also integrate with the rest of the Tableau platform, letting our customers quickly publish your data to Tableau Online or Tableau Server, or analyse it in Tableau Desktop. AIM: You have spoken extensively about self-service analytics. What is self-service analytics and how does it enable democratization of data (you have championed the term “democratization of data”)? DG: The new thing in the two-decade old domain of business intelligence is self-service analytics. This approach will enable all users to answer their own question and this will continue to be one of the fastest-moving areas in the enterprise. Since the techniques people use to drive adoption and get value from their data are multiplying, it is also leading to an increased demand for self-service analytics. This has been witnessed amongst many major companies using big data. Companies have started to prefer tools that can be used in-house rather than hiring an external agency to carry out the job for them. These in-house tools are meant to be used by end users in a self-service mode. This increase in users has led Gartner to estimate that the business intelligence market will reach $213.8 million in 2017 which is an 18.6% increase over the 2015 spending. India is currently among the top 10 big data analytics markets in the world. By 2025, the big data analytics sector in India is expected to grow eight folds to $16 billion accordingly to a recent report by NASSCOM. In fact, Linkedin has named statistical analysis and data mining as the second hottest skill that can get an individual hired in 2016 globally. It is the only skill that has been consistently ranked top 4 in 10 countries analysed, suggesting that businesses are still hiring experts aggressively in data storage, retrieval and analysis. Tableau is on a mission to help users see and understand their data. To accomplish this mission, our fundamental belief is in the democratization of data, meaning “the people who know the data should be the ones empowered to ask questions of the data.” Everyday knowledge workers should have the ability to easily access their data wherever it may reside. These same knowledge workers should also have the ability to analyze and discover insights about their data without assistance from the elite few – the data scientists and IT developers. [pullquote align=”left”]Visualizing data is important regardless of the size of the data because it translates information into insight and action. The approach to visualizing Big Data is especially important because the cost of storing, preparing and querying data is much higher. Therefore, organizations must leverage well-architected data sources and rigorously apply best practices to allow knowledge workers to query Big Data directly.[\/pullquote] We have recently launched Tableau 10 to reflect our commitment in making it easier and faster for people to work with data. One of the most important themes of Tableau 10 is to ensure that self-service analytics can be furthered for all kinds of users. Hence, this updated version has a fresh new design which makes it easy for users to grasp the insights from their data. This also includes new analytical and mobile enhancements, options for preparing data and a host of new enterprise capabilities. AIM: The burning question doing the rounds is the recent leadership shake-up at Tableau? Is Tableau going to be Amazon-ized with the new CEO Adam Selipsky at the helm? Your thoughts. DG: Adam has been at AWS (Amazon Web Services) from its inception and been part of the leaders team where he helped grow AWS from a start-up into a multi-billion dollar business and establish it as the undisputed market leader in cloud computing. He has tremendous experience working with enterprises on cloud infrastructure – something that will surely benefit Tableau’s customers and one of the reasons we selected him. Tableau has been and will be committed to delivering cloud solutions for our customers who are moving in that direction. For Adam, Tableau represents a rare opportunity of well-loved customer-centric products, incredibly talented people, a dedication to technology innovation, and an unrivaled momentum in the market. Adam sees Tableau is positioned to be a new standard in the world and he has the skills and right mindset to help build that next chapter of growth. AIM: What is Tableau’s strategy in the growing Business Analytics space? DG: Data analytics has become one of the game changing technologies of our time and we are witnessing a continuous increase in its adoption across industries. Companies will continue to embrace products that make it easy for them to manage and analyse growing amounts of data, stored in a number of different places. At Tableau, we will continue our mission to help people see and understand their data. In fact, we have recently launched Tableau 10 and 10.1 to reflect our commitment in making it easier and faster for people to work with data. One of the most important themes of Tableau 10 is to ensure that self-service analytics can be furthered for all kinds of users. Hence, this updated version has a fresh new design which makes it easy for users to grasp the insights from their data. This also includes new analytical and mobile enhancements, options for preparing data and a host of new enterprise capabilities. AIM: Tableau has gained widespread prominence for bringing data visualization into the mainstream. It has also been at the forefront of R&D. Can you tell us about R&D efforts in the field of statistical graphics or predictive analytics? DG: We are more focused than ever on Research and Development – in fact we just announced that we have nearly 900 employees at Tableau that are focused on it. We are still focused on empowering everyone in the organisation, and helping to provide a secure and scalable environment that is easy to manage and deploy. For example, in 2017, the interface to data will start to feel even more natural, thanks in part to improvements in areas like natural language processing and generation. A new addition to the BI toolbox, natural language interfaces can make data, charts, and dashboards even more accessible by letting people interact with data using natural text and language. Though there is healthy skepticism surrounding this new field, it will be an exciting space to watch. AIM: How is Tableau placed in the competitive landscape vis-à-vis rivals Spotfire and QlikView and legacy tech companies such as Microsoft and Amazon? One of the fun things about our industry is seeing analytics grow up in all kinds of places. We are very successful in deals where prospects evaluate multiple products. Our win rates are strong against all competitors. We are incredibly well-positioned to serve our customers because: Our enterprise capabilities have grown, and more and more customers are deploying the Tableau platform at scale. Tableau’s ease of use is transformative. Our product quality and user experience are unmatched. We care about every detail. We beta test with tens of thousands of customers. We employ scientists and experts to craft an analytical experience that is beautiful and powerful. This is important because it enables our customers to answer questions at the speed of thought, and to stay in the flow of the analytical thinking process. Tableau is the “Switzerland of data.” We have an open and flexible platform that lets people connect to 50+ native data sources and unlimited web data sources. It empowers end-users to do analytics without programming, and also helps IT move out of the report-factory role into an enabler of analytics for the entire organization. Tableau’s flexibility and choices means it adapts to existing IT infrastructure and is easy to deploy and maintain either on-premise, or in the cloud. Tableau has an amazing community of customers and data enthusiasts, who create data heroes in every organisation. AIM: Does Tableau as a company pay more emphasis on competitor focus or customer retention? DG: Enabling customers to do, that is precisely Tableau’s mission. For almost 14 years, Tableau has been helping organizations of all sizes to make sense of their data in order to make better decisions faster. [quote]Our every decision is aligned with our customers’ needs. Tableau already has a strong customer-oriented culture, with a large and devoted base of customers ranging from startups to non-profits to government institutions to global enterprise businesses and we are working to build additional mechanisms to ensure customers are always first in all of our minds.[\/quote] AIM: Last word, how is the market for BI and analytics in India shaping up? DG: There has been an uptick in the adoption rate of big data and analytics in India but many of the companies who decide to use these technologies don’t really have clarity on the kind of results they intend to get out of it. So there is also the need to create more awareness and educating enterprises about the power of big data and the competitive advantage that can be gained by harnessing analytics in the right way. There’s also a strong demand for big data analytics professionals across verticals in the Indian market. Our India entry around 3 years ago was in response to market demands in the region, and the growing business opportunity it offers. India forms one of the most crucial markets for us currently with respect to future potential. Gartner has estimated the business intelligence market to reach $213.8 million in 2017 which is an 18.6% increase over the 2015 spending. By 2025, the big data analytics sector in India is expected to grow eight folds to $16 billion as per a recent report by NASSCOM, so the demand for skilled professionals in this domain is only going to grow exponentially going forward.","excerpt":"It’s largely credited for bringing data visualization into the mainstream. Seattle-based Tableau Software is  now the new gold standard for Analytics, Visualization and BI space. The company recently unveiled their product roadmap for next three years during the Keynote at TC 2016 featuring next-gen innovations – in-memory data engine with Hyper, Project Maestro, self-service data prep […]","categories":["AI Features"],"tags":["Interviews and Discussions","Snowflake","tableau india"],"author_name":"Richa Bhatia","publish_date":"2016-11-29T09:05:42","publication_year":"2016","word_count":2789,"keywords":["machine learning","AWS","AI","cloud computing","R","Interviews and Discussions","RAG","Aim","analytics","predictive analytics","tableau india","Snowflake"],"extracted_tech_keywords":["AI","machine learning","analytics","Aim","RAG","predictive analytics","cloud computing","AWS","Snowflake","R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/interview-week-conversation-deepak-ghodke-country-manager-india-tableau\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10045317,"title":"Top MLOps Books In 2021","content":"Machine learning is getting mainstreamed as many organisations have integrated or are trying to integrate ML systems into their products and platforms. MLOps is the branch of ML that unifies ML systems development (dev) and ML systems deployments (ops). We have curated a list of top MLOps books to help you get a handle on the subject (in no particular order). 1| Machine Learning Engineering By Andriy Burkov Image Credits: Amazon The Machine Learning Engineering book is one of the most complete applied AI books out there and is filled with best practices and design patterns of building reliable machine learning solutions at scale. Andriy Burkov has a PhD in AI and is currently the machine learning team leader at Gartner. Find it here. 2| ML Ops: Operationalizing Data Science By David Sweenor, Dev Kannabiran, Thomas Hill, Steven Hillion, Dan Rope and Michael O’Connell Image Credits: O’Reilly Many analytics and machine learning (ML) models never make it to production. In this book, six experts in data analytics offer a four-step approach— Build, Manage, Deploy and Integrate, and Monitor—for creating ML-infused applications. The book covers: Fulfil data science value by reducing friction throughout ML pipelines and workflows.Constantly refine ML models through retraining, periodic tuning, and even complete remodelling to ensure long-term accuracy.Design the MLOps lifecycle to ensure that people-facing models are unbiased, fair, and explainable. Find it here. 3| Building Machine Learning Powered Applications By Emmanuel Ameisen Image Credits: O’Reilly In this book, author Emmanuel Ameisen will help you build an ML-driven application from initial idea to deployed product. Part I dives into how to plan an ML application and then measure success.Part II focuses on how to build a working ML model.Part III demonstrates ways to improve the model to meet your original vision.Part IV covers deployment and monitoring strategies. Find it here 4| Building Machine Learning Pipelines By Hannes Hapke, Catherine Nelson Image Credits: Amazon In this book, authors Hannes Hapke and Catherine Nelson walk you through the steps of automating a machine learning pipeline using the TensorFlow ecosystem. The book covers: Steps to build a machine learning pipeline.Building pipeline using components from TensorFlow Extended.Orchestrating machine learning pipeline with Apache Beam, Apache Airflow, and Kubeflow Pipelines.Working with data using TensorFlow Data Validation and TensorFlow Transform. Find it here. 5| Practical MLOps by Noah Gift, Alfredo Deza Image Credits: O’Reilly This book will take you through what MLOps is (and how it differs from DevOps) and explains how to operationalise your machine learning models. The book is a primer on in MLOps tools and methods (along with AutoML and monitoring and logging), and teaches you to implement them in AWS, Microsoft Azure, and Google Cloud. Find it here. 6| Introducing MLOps By Mark Treveil & Dataiku Team Image Credits: Amazon This book, by author Mark Treveil & Dataiku Team, helps understand the key concepts of MLOps to help data scientists and application engineers operationalise ML models to drive real business change and maintain and improve models over time. The book covers: Refining ML models through retraining, periodic tuning, and complete remodelling to ensure long-term accuracy.Designing the MLOps life cycle to minimise organisational risks with models that are unbiased, fair, and explainable.Operationalising ML models for pipeline deployment and for external business systems that are more complex and less standardised. Find it here. 7| Beginning MLOps with MLFlow: Deploy Models in AWS SageMaker, Google Cloud, and Microsoft Azure By Sridhar Alla, Suman Kalyan Adari Image Credits: O’Reilly The book covers MLFlow and ways to integrate MLOps into your existing code, to easily track metrics, parameters, graphs, and models. It will guide you through the process of deploying and querying your models with AWS SageMaker, Microsoft Azure, and Google Cloud. Find it here. 8| What Is MLOps? By Mark Treveil, Lynn Heidmann Image Credits: O’Reilly In this book, authors Lynn Heidmann and Mark Treveil from Dataiku introduce the data science-ML-AI project lifecycle. The book covers: Detailed components of ML model building, including how business insights can provide value to the technical team.Monitoring and iteration steps in the AI project lifecycle.How components of a modern AI governance strategy are intertwined with MLOps. Find it here. 9| Engineering MLOps by Emmanuel Raj Image Credits: Amazon The book provides in-depth knowledge of MLOps using real-world examples to assist you in writing programmes, training robust and scalable ML models, and constructing ML pipelines to train and deploy models safely in production. The book covers: Designing a robust and scalable microservice and API for test and production environments.Monitoring ML models, including monitoring data drift, model drift, and application performance.Building and maintaining automated ML systems. Find it here.","excerpt":"We have curated a list of top MLOps books to help you get a handle on the subject.","categories":["AI Trends"],"tags":[],"author_name":"kumar Gandharv","publish_date":"2021-08-04T18:00:00","publication_year":"2021","word_count":769,"keywords":["data science","machine learning","AWS","AI","ML","MLOps","Kubeflow","analytics","MLflow","TensorFlow"],"extracted_tech_keywords":["AI","machine learning","ML","data science","analytics","MLOps","MLflow","Kubeflow","TensorFlow","AWS"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-mlops-books-in-2021\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10138039,"title":"Will AI Boom Spell Doom for SaaS?","content":"Few debates are as perplexing as the future of SaaS in an AI-driven world. Will AI be a disruptor or will SaaS adapt, change its game, and flourish? Among the many AI presentations at SaaStr and SaasBoomi, the SaaS community events, one stood out. The founder of Ema, an AI employee company, said that SaaS will be dead with AI agents taking over the workplace and getting things done without having to navigate applications. But there are naysayers. “SaaS isn’t dying. It’s adapting, evolving, and positioning itself for the next era of innovation. Much like the humans evolving to leverage AI, SaaS too will evolve, transforming the way we work, interact, and grow,” said Deepak Anchala, the founder of Slintel, a 6sense company and also a GSF incubated\/portfolio startup that gave an excellent exit. Anchala, who’s about to launch another venture to redefine SaaS, believes that any SaaS company that fails to fundamentally rethink its workflows in an AI-first world and adapt accordingly will likely face extinction in the future. Remember when Apple launched the App Store? It changed how we use apps, pushing companies to move from web-based apps to mobile-friendly ones. It was a big shift in how we interact with technology. Now, we’re at the start of a similar change, driven by AI. Will it Come Crashing Down? However, as software gets more commoditised in the AI era, speculation has grown about the future of SaaS. For the past two decades, we’ve praised the SaaS model for its stickiness and easy adoption. Pay-per-use lowers the entry barriers, reduces risk, and provides ongoing access to advanced features. Then AI arrived, and in some ways, turned the industry on its head, threatening the very existence of SaaS and all things it brings. In fact, some argue that SaaS and enterprise software providers play an even more crucial role in the tech ecosystem today. That said, SaaS companies simply adding AI features to their existing software and charging extra fees are more exposed to risk than they’ve been in years. Where Does it Go from Here? Sure enough, SaaS and enterprise software will remain relevant but not like what they used to be. In the future, there may be abstractions that look like what we see with OpenAI and other companies’ large language models and it may include multimodal interfaces tools that can help in reasoning. For instance, a CEO might come across a tool that offers a proactive dashboard of key metrics, a reasoning engine for next-best actions, and a language model that pulls real-time data from various systems. Generative AI can also provide visual insights without needing to open specific apps. The possibilities could be many — customer data, trends, HR insights, hiring recommendations, supply chain risks, vulnerabilities, and various financial reports — all just a request away. SaaS is Far From Dead The whole viability of SaaS discussion gathered pace with the Salesforce saga and an Upekkha report predicting struggles for SaaS startups in 2024. As Salesforce’s growth slowed down, it no longer excited the investors, and missing its earnings guidance for the first time in 73 quarters intensified concerns. Workday, too, saw a significant drop, with its shares falling over 15% last week. Thiyagarajan Maruthavanan, a partner at Upekkha, stated succinctly: “Is SaaS dead? It depends on who you ask.” He noted that investors may quickly agree, as their incentive lies in highlighting outliers and labelling them as trends. For founders of SaaS companies, the focus should be on adapting to market shifts and interpreting the signals within those trends. While some companies are attempting to develop in-house AI capabilities, many are finding this approach to be costly and complex. As a result, there remains a substantial market for specialised SaaS applications that offer AI-enhanced functionalities. Agentic AI and What it Means As SaaS evolves in the AI era,  agentic AI will handle high-volume tasks, becoming more efficient and precise in its deterministic functions. The limitations of process automation will diminish as agents utilise generative models, neural networks, reinforcement learning, and other advancements. It’s an incredibly exciting time and an inflection point. The end of SaaS and other enterprise software hasn’t arrived, but the way we buy and consume software is set to change. Software giants Microsoft Corp, Salesforce, Oracle Corp, and others will need to aggressively transform the consumption layer so businesses can extract more value from their software investments and do so in a way that aligns with the capabilities of GenAI. It won’t be easy. Data fabric, compliance, security, and governance are difficult problems to solve. But this is an opportunity for the large installed bases of current software leaders to transform and become more user-friendly. Safe to say, SaaS isn’t dead; it’s simply evolving – dramatically. Major business model changes like what we saw from Salesforce chief executive Marc Benioff at Dreamforce, where it debuted Agentforce, is indicative of where it and others are headed.","excerpt":"SaaS and enterprise software will remain relevant but not as what they used to be.","categories":["AI Features"],"tags":["SaaS"],"author_name":"Tarunya S","publish_date":"2024-10-10T15:00:00","publication_year":"2024","word_count":823,"keywords":["Go","GenAI","agentic AI","OpenAI","AI","neural network","RAG","automation","SaaS","generative AI","R"],"extracted_tech_keywords":["AI","neural network","generative AI","GenAI","agentic AI","OpenAI","RAG","R","Go","automation"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/will-ai-boom-spell-doom-for-saas\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":50969,"title":"How To Build A BERT Classifier Model With TensorFlow 2.0","content":"BERT is one of the most popular algorithms in the NLP spectrum known for producing state-of-the-art results in a variety of language modeling tasks. Built on top of transformers and seq-to-sequence models, the Bidirectional Encoder Representations from Transformers is a very powerful NLP model that has outperformed many. What Is The Big Deal About BERT? The state-of-the-art results that it produces on a variety of language-specific tasks are enough to show that it is indeed a big deal. The results come from its underlying architecture which uses breakthrough techniques such as seq2seq (sequence-to-sequence) models and transformers. The seq2seq model is a network that converts a given sequence of words into a different sequence and is capable of relating the words that seem more important. Here, the LSTM network is a good example of the seq2seq model. The transformer architecture is also responsible for transforming a sequence into another, but without depending on any Recurrent Networks such as  LSTMs or GRUs. Being Bi-Directional, the model is also able to assume the context of a written text and predict accordingly. Inspired By BERT The success of  BERT has not only made it the power behind the top search engine known to mankind but also has inspired and paved the way for many new and better models. Given below are some of the popular NLP models and algorithms which were inspired by BERT: ALBERT: A Lite BERT(ALBERT) incorporates techniques such as factorised embedding parameterisation and cross-layer parameter sharing for parameter reduction which helps in scaling the pre-trained models. RoBERTa: Robustly optimised BERT is an optimised method for pretraining NLP systems which are built on BERT’s language-masking strategy. The model is claimed to have surpassed the BERT-large as well as XLNet-large models in performance. ViLBERT: Vision-and-Language BERT is built to learn task-agnostic joint representations of image content as well as natural language. The model includes two parallel BERT-style models which are mainly operating over image regions and text segments. MT-DNN: Multi-Task Deep Neural Network uses Google’s BERT to achieve new state-of-the-art results  The model is a combination of multi-task learning and language model pre-training. SenseBERT: The method uses Self-supervision which is an unsupervised learning technique as the name implies it supervises itself. In one of our previous articles, we learned how to solve a Multi-Class classification problem using BERT and achieve great results. We did this using TensorFlow 1.15.0. and today we will upgrade our TensorFlow to version 2.0 and we will build a BERT Model using KERAS API for a simple classification problem. We will use the bert-for-tf2 library which you can find here. The following example was inspired by Simple BERT using TensorFlow2.0. Lets Code! Importing TensorFlow2.0 Let’s start by importing TensorFlow2.0. The following example is done using Google Colab. try: %tensorflow_version 2.x  #gpu except Exception: pass import TensorFlowas tf Installing Necessary Modules To install the bert-for-tf2 module, type and execute the following command. !pip install bert-for-tf2 We will also install a dependency module called sentencepiece by executing the following command: !pip install sentencepiece Importing Necessary Modules import tensorflow_hub as hub from tensorflow.keras.models import Model from bert.tokenization.bert_tokenization import FullTokenizer Fetching The BERT Model From TensorFlowHub We will now fetch the actual BERT model from TensorFlowHub bert_layer = hub.KerasLayer(\"https:\/\/tfhub.dev\/tensorflow\/bert_en_uncased_L-12_H-768_A-12\/1\",trainable=True) Data Preparation The following is an example of data preprocessing for BERT. The code block transforms a piece of text into a BERT acceptable form. For detailed preprocessing check out the Step By Step Guide To Implement Multi-Class Classification With BERT & Tensorflow. Let’s test it out if the preprocessor is working properly- We will use an example from MachineHack’s Predict The News Category Hackathon. Output: We can now build a Keras model for binary classification and train it using a training set. For more details on preparing the dataset for training and validation, check out the Step By Step Guide To Implement Multi-Class Classification With BERT & Tensorflow.","excerpt":"BERT is one of the most popular algorithms in the NLP spectrum known for producing state-of-the-art results in a variety of language modeling tasks. Built on top of transformers and seq-to-sequence models, the Bidirectional Encoder Representations from Transformers is a very powerful NLP model that has outperformed many. What Is The Big Deal About BERT? […]","categories":["Global Tech"],"tags":["BERT","Google Colab","Tensorflow"],"author_name":"Amal Nair","publish_date":"2019-12-02T15:00:00","publication_year":"2019","word_count":645,"keywords":["TPU","Keras","AI","neural network","Transformers","BERT","NLP","Colab","Aim","Google Colab","TensorFlow","R","Tensorflow"],"extracted_tech_keywords":["AI","neural network","NLP","Aim","TensorFlow","Keras","Transformers","Colab","TPU","R"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/bert-classifier-with-tensorflow-2-0\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10039825,"title":"How The Big Four Dodged Pandemic And Made Record Earnings","content":"“The big tech is banking heavily on AI, Cloud and 5G technologies to retain customers and drive growth” A global emergency can smother your business, government lawsuits can break your company, competitors with trillion-dollar market value can wipe your organisation off the map. But what would happen when all three come together in the same year? Many would go extinct. But, not the big tech. The pandemic brought the world to a standstill. The internet giants, however, came out of it unscathed. Apple, Amazon, Google and Facebook, popularly known as the big four, have not only survived a combination of calamities but registered profits and left the Wall Street analysts dumbfounded. Last week, these four companies, along with Microsoft, released their quarterly reports. One common theme across the companies is their increased use of AI and cloud services. (Image credits: WSJ) Apple “Most of Apple’s features are extensions of federated learning in a way.” Between January and March, Apple sold iPhones worth $47 billion, a 66% jump from the previous year. Apple has departed from the good old Intel chips and launched M1 chips, a more powerful processor for machine learning and high speed 5G functionalities. “This is a huge leap for iPhone, bringing the best 5G experience in the market and delivering our most advanced technologies to users who want the absolute most from their iPhone,” said Greg Joswiak, Apple’s senior vice president of Worldwide Marketing. The phone maker has also locked horns with Facebook over customer privacy. Apple claimed their new privacy labels would inform its users on who is using their data. Targeted ads are a staple for companies like Facebook. They make billions off it. But, Apple’s concerns over the social networking sites alleged practices that put user’s privacy at risk. Apple uses federated learning and differential privacy techniques to keep the user’s data anonymous while leveraging it for analytics. Source: imore App Tracking Transparency, launched with iOS 14.5, is Apple’s new privacy feature that requires apps to ask permission to track you. Users can also turn off tracking for all apps by default. According to reports, 96% of iPhone users in the United States have opted out of app tracking on iOS 14.5. This shows that Apple’s ML-propelled privacy initiative earned the customers’ approval. According to the ML researchers at Apple, processing on end-user devices instead of server-based processing is a valued approach in enabling end-user privacy. This strategy extends to many of Apple’s ML solutions, such as text prediction in keyboards. The ability to personalise towards a user’s diction is highly desirable. Like its counterparts, Apple is banking heavily on AI and chip innovation to drive growth and trust. Amazon Amazon started its year with the news of its founder Jeff Bezos stepping down and Andy Jassy, the current CEO of Amazon’s profit engine, Amazon Web Services (AWS), picked as his replacement. AWS’ net sales climbed to $13.5 billion this quarter, up 32% year over year. Amazon doesn’t disclose advertising sales, but it’s included in its “Other” category, which saw revenue growth of 77% year over year to $6.9 billion. Amazon Web Services dominates more than one-third of the global cloud infrastructure. AWS leverages its global network of servers to handle and route the massive spikes in traffic on Prime Days. AWS makes short-term, large scale global events like Prime Day technically feasible and economically viable. AWS has not only brought profits from outside, but the services have even revolutionised the in-house business. For example, the much-hyped “Prime Day” is a multi-billion revenue generator. In 2020, the two-day event saw sales surpassing $3.5 billion worldwide. On any typical Prime Day, Amazon deploys the following AWS services: Amazon Redshift, Amazon Machine Learning for analyticsAmazon API Gateway, CloudSearch, Data Pipeline, for crucial application Services EC2, Auto Scaling, EBS, EMR for computing.DynamoDB, ElastiCache, Kinesis, Kinesis Firehose, RDS for database management.CloudTrail, CloudWatch, Trusted Advisor are the management toolsThe Mobile Analytics service plays a key role as most of the orders are made via mobile devices.CloudHSM, IAM, KMS for data securityCloudFront, S3, Amazon Glacier for data storage and delivery. AWS powers many top internet companies in the world like Netflix. The cloud service company has outclassed its competitors for over a decade now. AWS’ wide range of AI, ML, and even quantum computing services will continue to drive revenues for Amazon. Alphabet Google’s parent company Alphabet reported revenue of $55.31 billion. Google’s revenue rose 34% compared to the previous year. The search giant reported advertising revenue of $44.68 billion, which is a significant rise from $33.76 billion in the same quarter last year. The pandemic has moved many from cubicles to couches, and the streaming sites have seen a spike in their revenue. YouTube has reported strong growth with its ads drawing in $6.01 billion during the quarter — a 49% rise from a year ago. YouTube uses state of the art recommendation algorithms to keep the viewers busy. The streaming site has also recently announced that it uses custom-made chips or VCUs to stream high-quality content. Like Apple, Google too has padded up their services to accommodate user’s privacy. Earlier this year, Google announced that it would ditch third-party cookies on the Chrome browser. The cookies, used for targeted advertisements, will be replaced with browser-based machine learning. The ML models on the browser will skim the user history and weave a network of people’s interest (think: recommendation engine) to be used for advertisement. A couple of months ago, Google announced that it would be investing $7 billion for expanding offices and data centres across 19 states, which will employ at least 10,000 people, indicating how well the company has performed recently. Facebook Despite the Capitol riots and its tussle with the Australian government, Facebook reported revenue of $26.17 billion for the quarter, and the net income grew 94% to $9.5 billion, from $4.9 billion a year prior. Facebook’s CEO Mark Zuckerberg said more than 1 billion monthly active users visit Facebook’s Marketplace service, where users can buy and sell goods. Facebook has been in the spotlight for misinformation and tracking users for a while now. The company uses research coming out of its FAIR labs to moderate content for hate speech and misinformation. Also, machine learning has been vital to Facebook’s recommendation systems which connect entities with similar interests (read businesses). Facebook’s “Other” category recorded a revenue of $732 million for the quarter, accounting for nearly 3% of Facebook’s total revenue in the quarter. This “Other” category includes sales of Oculus virtual reality headsets and Portal video-chatting devices. Facebook has been bullish on mixed reality for a while now. For instance, Facebook Reality Labs’ wearables use electromyography (EMG) to translate nerve signals in the wrist to digital commands. “Take the QWERTY keyboard as an example. It’s over 150 years old, and it can be radically improved. Imagine instead a virtual keyboard that learns and adapts to your unique typing style (typos and all) over time,” said the company in a blog post. Facebook’s chief Zuckerberg even wants to use this VR technology to combat climate change; he believes a near realistic virtual world will cut down commutes and hence emissions! What’s next for the Big Four Google, Apple and Amazon have made their autonomous ambitions clear, with Apple joining late to the party. While Google’s Waymo is still running at a loss, Apple’s self-driving car is eyeing a 2024 launch. Amazon’s recent purchase, Zoox, a company that makes robotaxis, is doing trial runs. Facebook, however, is investing heavily in taking augmented reality to the masses. While Apple is gearing towards chip autonomy, Google and Amazon are serious about customised silicon too. Google’s TPUs are already powering its data centres. Now Amazon wants to power AWS data centres with custom made chips such as Trainium. That said, the fate of big tech currently is in the hands of the regulators. The big four have come under scanner in India, Australia, the EU and the US. The EU has especially been stringent with the antitrust laws. Recently Apple was served with one. Google and Facebook have been accused of backdoor ad-rigging. Amazon has a few of its own. The biggest one being its challenge of the US defense for choosing Microsoft’s Azure over AWS for a $10 bullion cloud contract. These early skirmishes have already influenced how these companies run their businesses. Federated learning and Responsible AI are few such domains that have witnessed growing interest recently. Autonomous cars, custom made chips, privacy-padded ad services will be the areas critical to the success of these companies soon.","excerpt":"“The big tech is banking heavily on AI, Cloud and 5G technologies to retain customers and drive growth” A global emergency can smother your business, government lawsuits can break your company, competitors with trillion-dollar market value can wipe your organisation off the map. But what would happen when all three come together in the same […]","categories":["Deep Tech"],"tags":["AI Tool","AWS"],"author_name":"Ram Sagar","publish_date":"2021-05-10T10:00:00","publication_year":"2021","word_count":1427,"keywords":["federated learning","machine learning","AWS","AI","ML","recommendation systems","RAG","Aim","analytics","differential privacy","AI Tool"],"extracted_tech_keywords":["AI","machine learning","ML","analytics","Aim","federated learning","differential privacy","RAG","recommendation systems","AWS"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/how-the-big-four-dodged-pandemic-and-made-record-earnings\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10041999,"title":"HuBERT: Facebook’s Latest Approach To Self-Supervised Speech Representation Learning","content":"Facebook AI Research (FAIR) has published a research paper introducing Hidden Unit BERT (HuBERT), their latest approach for learning self-supervised speech representations. According to FAIR, self-supervised techniques for speech recognition are currently limited due to three factors: First, the presence of multiple sound units in each input utterance; secondly, the absence of lexicons of input sound units during the pre-training phase; and finally, the observation that sound units have variable lengths with no explicit segmentation. To tackle this problem, HuBERT utilises an offline k-means clustering algorithm and learns the structure of its (spoken) input by predicting the right cluster for masked audio segments. FAIR claims that HuBERT’s simplicity and stability makes it easily deployable for use cases in NLP and speech research. HuBERT is inspired by FAIR’s DeepCluster method for self-supervised visual learning. DeepCluster is a clustering method introduced in 2018 that learns a neural network’s parameters and their cluster assignment—after which it groups these features using a standard clustering algorithm, called k-means. HuBERT further benefits from Google’s Bidirectional Encoder Representations from Transformers (BERT) by leveraging its masked prediction loss over sequences method to showcase the sequential nature of speech. A BERT model takes masked continuous speech features to predict pre-determined cluster assignments. This predictive loss is applied only over the masked regions, making the model learn high-level representations of unmasked inputs to correctly deduce the masked areas’ targets. How does HuBERT work? The HuBERT model learns both acoustic and language models from these continuous inputs. For this, the model first encodes unmasked audio inputs into meaningful continuous latent representations. These representations map to the classical acoustic modelling problem. The model then makes use of representation learning via Masked Prediction. The HuBERT approach predicting hidden cluster assignments of the masked frames (MSK)  y2, y3, y4  \/ Source: Facebook AI Research The model seeks to reduce prediction error by capturing the long-range temporal relationships between the representations it has learned. Here, the consistency of the k-means mapping from audio inputs to discrete targets is just as important as their correctness since it allows the model to focus on modelling the sequential structure of input data. For instance, if an early clustering utterance cannot tell \/k\/ and \/g\/ sounds apart, it would lead to a single supercluster containing both these sounds. The prediction loss will then learn representations that model how other consonant and vowel sounds work with this supercluster while forming words. Through this newly learned representation, the clustering iteration will create better clusters. Implementing HuBERT FAIR pretrained HuBERT on the standard LibriSpeech 960 hours and the Libri-Light 60,000 hours and found that the model either matched or improved on Facebook’s state-of-the-art speech recognition AI wav2vec 2.0’s performance on fine-tuning subsets of 10 minutes, 1 hour, 10 hours, 100 hours and 960 hours. The experiments were conducted using two models of HuBERT: HuBERT L-LV60k and HuBERT XL-LV60k. Source: Facebook AI Research Facebook AI Research also tested HuBERT’s performance in language generation—which it says is essential for direct language modelling of speech signals without relying on lexical resources such as supervised labels. Source: Facebook AI Research Using Generative Spoken Learning Modeling (GSLM)—which involves learning the acoustic and linguistic characteristics of a language without text or labels—Facebook has begun using learned speech representations to synthesise speech from models like Contrastive Predictive Coding (CPC), Wav2Vec2.0 and HuBERT. HuBERT, in both automatic and human evaluations, generated samples that could compete in quality with top-line supervised character-based LogMel (LM). Finally, FAIR also tested HuBERT using the MUltiple Stimuli with Hidden Reference and Anchor (MUSHRA) test, which conducts a codec listening test to evaluate output quality from lossy audio compression algorithms. Here, HuBERT came second only to uncompressed audio. Source: Facebook AI Research Many AI-powered speech recognition platforms have been working towards understanding and recognising speech simply through listening and interacting and without labels. For example, Facebook AI Research recently launched an AI that understood speech without labelled text. The tech giant also made its largest language database public to facilitate the development of speech recognition tools, explicitly concentrating on languages such as Swahili, where labelled data is scarce. With HuBERT, Facebook claims, the AI research community can develop Natural Language Processing (NLP) systems they could train through audio instead of text samples. This will allow AI voice assistants to capture the expressivity of oral language and speak with the nuances and styles of an actual person speaking the language. Such technology will allow individuals who speak rare languages or dialects or languages with more limited literature than others to benefit from more inclusive speech recognition and translation applications.","excerpt":"Facebook AI Research (FAIR) has published a research paper introducing Hidden Unit BERT (HuBERT), their latest approach for learning self-supervised speech representations. According to FAIR, self-supervised techniques for speech recognition are currently limited due to three factors: First, the presence of multiple sound units in each input utterance; secondly, the absence of lexicons of input […]","categories":["AI Features"],"tags":["Facebook AI research","Speech Analytics"],"author_name":"Mita Chaturvedi","publish_date":"2021-06-20T18:00:00","publication_year":"2021","word_count":762,"keywords":["Go","TPU","Speech Analytics","AI","neural network","Transformers","SLM","RAG","NLP","Aim","Facebook AI research","R"],"extracted_tech_keywords":["AI","neural network","NLP","Aim","Transformers","SLM","RAG","TPU","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/hubert-facebooks-latest-approach-to-self-supervised-speech-representation-learning\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10143948,"title":"Cockroach is Growing in Namma Bengaluru","content":"The love for Namma Bengaluru goes beyond words for Cockroach Labs – a software company that creates cloud-native SQL databases which has cemented its roots in India since its arrival years ago. With its growing team of over 55 engineers, including specialists in database and cloud engineering, the company’s journey in India is more than strategic; it’s emotional. “India’s unmatched technical talent and entrepreneurial mindset make Bengaluru a perfect fit for our mission,” said Spencer Kimball, co-founder and CEO, in a recent conversation with AIM as he reflected on the city’s pivotal role in the company’s global vision. This connection with Bengaluru is also deeply strategic. Cockroach Labs has found the city to be the ideal hub for scaling operations across the Asia-Pacific region. Bengaluru provides unparalleled time-zone advantages and access to a rich talent pool. With 1.4 billion people and a rapidly digitising economy, India is a proving ground for resilience and scalability. “The demands of Indian enterprises, across sectors like banking, media, and logistics, align perfectly with CockroachDB’s strengths in handling scale and regulatory requirements,” Kimball added. From Google Roots to Cockroach Resilience The inception of Cockroach Labs is a story of necessity and innovation. Kimball, alongside co-founders Ben Darnell and Peter Mattis, spent a decade at Google (2002-2012) working on pioneering projects like Gmail, Google Reader, and Colossus – Google’s distributed storage system. Their experience at Google shaped their understanding of the need for resilient and distributed systems. “Google had low-cost infrastructure, wasn’t reliable, so they built very redundant, distributed infrastructure. And when we were at Google in that period, we saw the advent of Bigtable… [Google] quickly followed Bigtable, and they called [it] MegaStore, which introduced an idea of transactions… and then that eventually led to Spanner,” recalled Kimball in another interview. However, after leaving Google, the trio faced firsthand challenges while building their startup, Viewfinder. The absence of a transactional distributed database forced them to use DynamoDB, leading to inefficiencies and workarounds. “..We had to use DynamoDB, but DynamoDB without transactions 2012 to 2014. Well, we wished we had them. We spent maybe a third of our engineering time working around that problem,” Kimball said. This frustration led to the birth of Cockroach Labs in 2014, with a vision to create an open-source, cloud-native distributed database. “…And it was the capabilities of Spanner, which people probably heard about these days because they wrote a big paper about it …that tied the best things together of relational databases, the more traditional ideas of consistency and concurrency management. And then also there’s no SQL, which is extremely elastic, least scalable…..the idea of Cockroach, to be an open source, fast follow on Google’s internal technology, lived on.,” he added. Towards Resilience & Scale True to its name, CockroachDB has built its reputation on resilience, a feature that has drawn comparisons to industry heavyweights like Google Spanner and Amazon Aurora. The database’s unique ability to run seamlessly across cloud providers, private data centres, and hybrid setups makes it a standout choice. Kimball also noted their focus is on eliminating vendor lock-in, ensuring businesses can operate uninterrupted, even in the face of cloud or data centre outages. This adaptability has made Cockroach Labs the operational backbone for global giants like Netflix and ambitious startups like Fi. Unlike competitors such as MySQL Cluster and MariaDB Xpand, CockroachDB’s architecture allows businesses to handle elasticity, resilience, and cross-cloud data sovereignty without compromise. Kimball emphasised that they are more than just a database—they are a partner in tackling scalability challenges over decades of growth. He highlighted the company’s distinctive focus on both startups and enterprise customers, providing a seamless pathway for businesses to scale effectively. Building a Future from Bengaluru Cockroach Labs isn’t resting on its laurels. The company has ambitious plans to expand its Bengaluru office into a first-class R&D hub to tackle the hardest technical challenges in database engineering. Teams here are working on innovations like vector data integration for AI in order to enable operational databases to evolve into systems capable of real-time intelligence. “India’s environment is perfect for not only engineering excellence but also go-to-market strategies tailored for enterprises,” Kimball further said. As AI reshapes the industry, Cockroach Labs is ready to lead the charge. The company is investing heavily in AI-specific features, including support for vector similarity searches and operationalising AI workflows. Reflecting on plans for the next decade, Kimball concluded, “We think of ourselves not as an N plus one incremental database that you’re adding, but as an N divided by two. This is where you’re going to shrink your databases over the next 10 years.” With its foundations in Bengaluru and an eye on the global stage, Cockroach Labs is poised to make its mark with remarkable impact in the years ahead.","excerpt":"“We named it Cockroach Labs for a reason – because, like the insect, our database doesn’t just survive; it thrives,” CEO Spencer Kimball said.","categories":["AI Features"],"tags":["Bengaluru"],"author_name":"Aditi Suresh","publish_date":"2024-12-19T14:01:20","publication_year":"2024","word_count":795,"keywords":["Go","AI","ML","Scala","Git","RAG","Aim","SQL","Rust","Bengaluru","R"],"extracted_tech_keywords":["AI","ML","Aim","RAG","R","SQL","Go","Rust","Scala","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/cockroach-is-growing-in-namma-bengaluru\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":8399,"title":"Zenatix Solutions raises funding from Blume Ventures","content":"Zenatix Solutions, an IoT driven, energy data analytics company based in Gurgaon, celebrated it’s second birthday this December with a blast by raising funding from Blume Ventures in a Pre-Series A round. Blume Ventures is a seed-stage venture fund that backs startups with both funding as well as active mentoring. Team Zenatix Solutions Zenatix specializes in providing a cloud-based energy analytics product that helps large consumers of electricity reduce their electricity consumption by at least 10%, using intelligence from the energy data correlated with other factors including temperature, humidity and occupancy. Karthik Reddy, Co-Founder and Managing Partner at Blume Ventures, commented on the matter, “In Zenatix, we see, a great combination of our beliefs in a well-rounded, passionate and determined team to solve the energy problem. We have faith that Team Zenatix will keep pushing the innovation envelope on this front, and emerge as a key player as the world feverishly discusses the Climate Summit in Paris.” Founded by alumni of IIT Delhi, IIM Ahmedabad and UCLA – Rahul Bhalla, Vishal Bansal and Dr. Amarjeet Singh; in a short span of 2 years, Zenatix has already bagged more than 45 paying customers including Google, Tata Teleservices, NIIT, Glaxo SmithKline, Xerox, Mother Dairy and Starbucks and several marquee clients in the office building segment and hospitality. The company is now focusing on commercial customers such as large office spaces, retail chains, banks, hospitals, hotels and educational institutions. Rahul Bhalla, Co-founder and CEO, Zenatix Solution promptly explains that they have a SaaS based engagement model wherein the customer doesn’t pay upfront for the hardware or software, rather it’s a fixed monthly or annual fee. He further added, “ The energy savings delivered by their analytics solution far outweighs the fees. Typically, the customers become cashflow positive by a significant amount from second month onwards.” Having demonstrated their value proposition and the product-market fit through a healthy customer base in Delhi and NCR, the funding will be used primarily for product enhancement and to expand operations in Bangalore and Mumbai over the next 2-3 months. They aim to achieve an ARR (Annualized Recurring Revenue) of USD 1 million over the next 9-12 months. On a concluding note, Analytics India predicts tremendous growth in the energy analytics sector as the world battles with such issues as depleting renewable resources, energy consumption and climate change. Energy efficiency has become as important and as large a frontier to tackle as renewable energy or global warming. Moreover, AIM looks forward to welcoming Zenatix to Bangalore!","excerpt":"Zenatix Solutions, an IoT driven, energy data analytics company based in Gurgaon, celebrated it’s second birthday this December with a blast by raising funding from Blume Ventures in a Pre-Series A round. Blume Ventures is a seed-stage venture fund that backs startups with both funding as well as active mentoring. Zenatix specializes in providing a cloud-based […]","categories":["AI News"],"tags":["analytics companies","energy"],"author_name":"Apoorva Verma","publish_date":"2015-12-08T13:37:46","publication_year":"2015","word_count":418,"keywords":["Go","energy","funding","programming_languages:R","AI","innovation","programming_languages:Go","Aim","analytics companies","analytics","R","startup"],"extracted_tech_keywords":["AI","analytics","Aim","R","Go","innovation","startup","funding","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/zenatix-solutions-raises-funding-from-blume-ventures\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10115858,"title":"Meet the Creators of महामराठी","content":"According to the 2011 census, India is home to about 83 million native Marathi speakers, especially in Maharashtra. This makes Marathi the third most-spoken Indian language after Hindi and Bengali. The state, a major hub for startups and major tech firms, accounts for 12.92% of India’s GDP and has a per capita income higher than the national average. Understanding this importance, a team of US-based Indian researchers—Aakash Patil, a postdoctoral researcher at Stanford University, Mrunmayee Shende, cofounder of CourtEasy AI and Niraj Kumar Singh, an ML engineer at Inbound Health—came up with MahaMarathi 7B. Joining the league of indic LLMs like Telugu, Malayalam, Tamil, and Odia Llama, MahaMarathi has been built on seven billion parameters. It is domain-adapted, continually pre-trained, and instruction fine-tuned using the Llama-2 and Mistral AI framework. The Inception “When GPT-4 came out, we realised the importance of doing something for Indian languages. Although it wasn’t our primary focus initially, the release of newer models like Meta’s Llama motivated us to build on top of them. The idea for MahaMarathi started in June last year, initially just as a concept,” Patil told AIM, sharing the inception story of the model. He said that MahaMarathi would not have been possible without the help of Microsoft for Startups backed legal tech startup CourtEasy AI, which provided them with all the computing resources and data to train the model. The model was trained on NVIDIA A1 100 GPUs, procured by the startup under the program. “Why Marathi? It is our mother tongue, and we have spoken it since childhood, so we thought we could contribute to the Indic LLM space,” added Shende. All of them were born and raised in Maharashtra. Patil is from Akola, Shende is from Satara, and Singh is from Nagpur. “Our shared passion for Marathi and the desire to bridge technological gaps motivated us to work on this project,” commented Patil, adding to what Shende said. Challenges Galore The making of MahaMarathi was not a cakewalk, as the team was hindered by computing power and data availability challenges. Even though Patil had access to powerful supercomputers because of his research, he could not use them for personal projects. However, things started to change around June last year when he discovered Shende and Singh share his vision of building an indigenous Marathi LLM. Shende’s CourtEasy AI primarily operates in the legal sector, developing AI tools for lawyers, paralegals, and law firms in India. “Initially focusing on English, we soon recognised the necessity of including Indian languages, as numerous cases in lower courts are conducted in these languages,” said the co-founder. So, Shende had been collecting data for quite some time, and by the end of December, she had amassed a significant corpus. Datasets and Training Method The researchers compiled a large corpus of about five million words in Marathi for their language dataset. This corpus, sourced over six months from publicly available content on websites, blogs, media, and news outlets, formed the basis for the initial pre-training of their model. The team developed a new tokeniser for the Marathi language to manage this large corpus. After creating the tokeniser and expanding the vocabulary size, they pre-trained the model for next-token prediction tasks. During pre-training, after achieving satisfactory results in next-token and next-sequence prediction, the team shifted their focus to fine-tuning. For this phase, they used datasets from Stanford Alpaca and Microsoft Orca, which were translated into Marathi and then cleaned to ensure accurate and contextually appropriate translations. Satisfied with the model’s sequence prediction and context generation capability, they further refined the fine-tuning process using various datasets, including translations done with IndicTrans by AI4Bharat. IndicTrans is notable as the first open-source transformer-based multilingual NMT model supporting high-quality translations across all 22 scheduled Indic languages. When asked about the primary databases used, Patil explained that they stored the tokenized data on a hard disk, while MongoDB Atlas was used to store the fine-tuned pairs, which amounted to approximately 60,000. Since vectorised databases are unsuitable for storing large corpora, the team primarily relied on Amazon S3 due to its extensive storage capacity. The researchers chose not to use alternatives like the BLOOM family of open-source models or other multilingual models. Instead, they opted for a combination of the Mistral AI framework and Llama 2 architecture. This approach involved enhancing the Llama architecture by initialising Transformer layers with Mistral’s weights and modifying the MergeSkate library to improve efficiency and multilingual support. What’s Next? In the last three weeks since the release, the feedback has been overwhelmingly positive for MahaMarathi, especially from the Marathi community in the Bay Area and large enterprises, Intel being one of them, as Shende explained. Depending on the feedback for the base model, the trio will also release SFT and DPO models. However, the team also noticed a need to educate the public on using pre-trained models and fine-tuning. They plan to release collaborative notebooks to assist smaller businesses and medium-scale enterprises in integrating AI into their operations. Join Rising 2024, the largest summit on Diversity and Inclusion in India, taking place on April 4-5 in Bangalore. Grab your passes now.","excerpt":"Based on their shared passion for Marathi and making generative AI accessible for non-English speakers, the team started working on the Marathi LLM.","categories":["AI Features"],"tags":["Indian LLMs"],"author_name":"Shritama Saha","publish_date":"2024-03-15T16:00:00","publication_year":"2024","word_count":854,"keywords":["Go","API","AI","MongoDB","ML","RAG","GPT","Aim","Indian LLMs","R","startup"],"extracted_tech_keywords":["AI","ML","Aim","RAG","MongoDB","R","Go","API","GPT","startup"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/meet-the-creators-of-%e0%a4%ae%e0%a4%b9%e0%a4%be%e0%a4%ae%e0%a4%b0%e0%a4%be%e0%a4%a0%e0%a5%80\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10104399,"title":"Google’s Gemini Nano Pushes Smartphone Industry On-Edge","content":"After keeping everyone waiting for so long, Google has hopped on the Christmas spirit just in time. The tech giant finally unveiled Gemini. To everyone’s surprise, Google didn’t just stick with cloud-based LLM but entered the race by creating Gemini Nano, an LLM for Android. Gemini Nano could have a similar impact like Android had on smartphones years ago. It’s designed for on-device tasks and operates directly on mobile phones. Gemini Nano runs on a phone and without the internet, beginning an era of on-device LLMs, one that fits in your pocket— Keerthana Gopalakrishnan (@keerthanpg) December 6, 2023 The Pixel 8 Pro will be the first smartphone designed to run Gemini Nano, which now powers new features like ‘Summarize’ in the Recorder app and  ‘Smart Reply’ on Gboard, starting with WhatsApp—with more messaging apps expected next year. For instance, with this feature, Gemini Nano will automatically frame responses for the ongoing conversation in WhatsApp. Moreover, Google has introduced a new system called Android AICore in Android 14 that provides easy access to Gemini Nano. It handles model management, runtimes, safety features and more, simplifying the work for you to incorporate AI into apps. Threat to Apple? Google isn’t the only player exploring the integration of LLMs on mobile devices. Anticipations are high that Apple will introduce an upgraded version of Siri in iOS 18 and other operating systems next year. iOS 18 holds particular significance for Apple, speculated to introduce generative AI capabilities as the company aims to catch up with industry leaders like OpenAI and Google. iOS 18 will incorporate generative AI technology to boost Siri and the Messages app’s capabilities in answering questions and auto-completing sentences. Apple is also reportedly considering generative AI for apps like Apple Music, Pages, Keynote, and Xcode. While the iPhone commands a market share of 58.28% in the US, Android dominates globally with 69.44% market share. In their blog post, Google mentioned that following the Pixel 8 Pro, they will introduce additional devices and silicon partners under AICore and Gemini Nano, with details to be disclosed in the coming months. It is highly likely that one of the new partners will be Samsung, especially considering Google’s leverage of Samsung S.LSI. An interesting point to note is that in Q3 2023, Samsung held a 20% global market share for smartphones. Interestingly, Google has announced that its AICore uses new machine learning hardware, such as the latest Google Tensor TPU and NPUs in flagship devices from Qualcomm Technologies, Samsung S.LSI, and MediaTek. Moreover, if Google and Samsung partners along with other Android players in the market, it would be a huge blow to Apple. Phone Companies’ New Focus Samsung recently introduced its on edge  generative AI model, Gauss.  This AI model is set to be incorporated into the upcoming Galaxy S24 handset, anticipated to be released in early 2024. Gauss has the ability to generate and edit images, compose emails, summarize documents, and also function as a coding assistant. Interestingly, in August, Xiaomi’s founder and CEO, Lei Jun, revealed that the company’s digital assistant, Xiao Ai, was undergoing enhancements to incorporate generative AI capabilities. The upgraded assistant is reportedly backed by a nimble AI model with 1.3 billion parameters and operated locally on the phone. Testing the AI Power Google Pixel 8 Pro, which will host Gemini Nano, utilizes the power of Google Tensor G3, a slightly modified version of Samsung’s Exynos 4 nm chipset. Apart from  Google, MediaTek and Snapdragon are making strides in developing chips for running AI on mobile devices. MediaTek, for instance, recently released the Dimensity 8300, which offers generative AI capabilities, adaptive gaming technology, and fast connectivity. Last month Qualcomm introduced the Snapdragon 8 Gen 3. While not as powerful as the Google Tensor G3’s TPU for specific tasks, it enables various AI features and supports on-device model execution. On the other hand, Apple’s latest iPhone 15 Pro features the A17 Pro chip based on TSMC’s 3nm technology. The chip’s neural engine, responsible for tasks such as transcribing speech to text, is now up to twice as fast, capable of performing up to 35 trillion operations per second—double the previous year’s 17 TOPS. While Google focuses on collaboration to make Android generative AI-capable, it will be interesting to see how Apple responds.","excerpt":"Google has introduced a new system called Android AICore in Android 14 that provides easy access to Gemini Nano.","categories":["Global Tech"],"tags":["Gemini"],"author_name":"Siddharth Jindal","publish_date":"2023-12-07T18:15:30","publication_year":"2023","word_count":712,"keywords":["Go","Gemini","machine learning","TPU","OpenAI","AI","Git","RAG","Aim","generative AI","R"],"extracted_tech_keywords":["AI","machine learning","generative AI","OpenAI","Aim","RAG","TPU","R","Go","Git"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/googles-gemini-nano-pushes-smartphone-industry-on-edge\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10148251,"title":"Karnataka Gives a Policy Push to Semiconductor Industry","content":"In a major push to establish itself as a semiconductor hub, the Karnataka government is encouraging corporate investments to reshape the technology sector. Earlier this week, the state approved an investment of ₹3,425.60 crore to set up the first electronics manufacturing cluster at Kochanahalli in Mysuru. The cluster aims to create 460 jobs in the district adjoining India’s Silicon Valley. The government has earmarked 234 acres of land in Kochanahalli for the development of an Electronics Manufacturing Cluster (EMC) and semiconductor units. This is among the first investments made after the International Semiconductor Consortium (ISMC), a joint venture between Abu Dhabi-based Next Orbit Ventures and Israel’s Tower Semiconductor, promised to invest $3 billion in the state. This week, state commerce and industries minister MB Patil announced that the state-level committee has greenlit three new investment proposals and six more investment plans, amounting to ₹10,144.31 crore. Led by Karnataka chief minister Siddaramaiah, the committee anticipates that these initiatives will generate close to 6,000 job opportunities statewide. These efforts are part of the state government’s strategy to enhance the state’s industrial and economic framework. As the first state to draft a semiconductor policy in 2010, Karnataka made its mark in semiconductor manufacturing very early on. This was followed by the launch of the Karnataka Electronics System Design and Manufacturing (EDSM) Policy in 2017. Karnataka, especially Bengaluru, is home to a large pool of skilled engineers and IT professionals, many of whom are graduates from top engineering colleges in India. This talent pool is crucial for semiconductor design, manufacturing, and related services. Looking Beyond Brand Bengaluru Bengaluru, the tech city, has had a long reign in the semiconductor industry and has established itself as India’s manufacturing hub with an existing engineering culture and the presence of government institutions, including ISRO, BEL and DRDO, among others. The city’s well-developed ecosystem, including research and development (R&D) centres, universities, and technology parks, supported the growth of semiconductor companies and startups. According to the Karnataka Economic Survey of 2022–23, Bengaluru is the largest hub of semiconductor design companies in India. Nearly 70% of the country’s chip designers work in the city, and 80% of the sector’s design revenues come from the same. Bengaluru’s strategic location and connectivity has also made it easier for companies to access global markets, suppliers, and customers, especially considering the presence of many global capacity centres (GCCs) in the city. State IT minister Priyank Kharge even said that the government is willing to tweak incentive and labour policies to attract major players along with their supply chains. Fast-forward to 2024. The state is looking to expand beyond Bengaluru and take the facility to neighbouring centres like Mysuru, Doddaballapur, Kolar, and others. The state government has also been actively implementing various policies and incentives to promote the flow of investments and R&D centres beyond Bengaluru to target cities like Mysuru, Mangaluru, and Tumakuru as viable alternatives. The government has earmarked 901 acres across four industrial clusters to accommodate the growing semiconductor industry. These clusters include 224.5 acres in Hubballi (near Kotur-Belur), 245.67 acres in Kochanahalli, 218.20 acres in Vasanthanarasapura Industrial Area (Tumakuru), and 213.14 acres in Hosahalli (Bengaluru Rural). The state is also poised to introduce a revised Electronics System Design and Manufacturing (ESDM) policy, which aims to foster innovation and investment in the sector. In a conversation with AIM, Srini Chinamilli, CEO and co-founder of Tessolve Semiconductor, emphasised the growing potential of tier-2 cities for the company’s operations. Talking about the work ethics and dedication of employees in tier-2 cities, Chinamilli said, “These employees from tier-2 cities tend to be very dedicated, similar to Bengaluru. But the attrition levels are much lower.” States Scramble to Attract Investments There’s stiff competition between Gujarat, Telangana, Maharashtra, Tamil Nadu and  Karnataka to draw leading semiconductor companies to their respective regions. Recently, Synaptics and EmbedUR collaborated to build cost-effective AI chips from Chennai due to the city’s low attrition rates. This is regardless of the fact that about 25% of Synaptics’ global workforce is already based in India, primarily in Bengaluru. At the start of this year, Karnataka had lost an investment of ₹3,750 crore to build an outsourced semiconductor assembly and testing (OSAT) facility in Mysuru, which was slated to create 3,200 jobs. The investor, electronics manufacturing company Kaynes Technology, later moved to Telangana and subsequently to Gujarat to build the OSAT facility at a relatively lower cost. The company is now investing ₹3,307 crore in a facility that will have the capacity to produce more than 6.33 million chips per day. Karnataka had accused the Centre of using political influence to divert semiconductor investments away from the state and move them to Gujarat, despite the southern state having an edge. Commenting on the political atmosphere in forcing companies to move from one location to another, Patil had voiced concerns over central government policies. He urged that semiconductor companies should be allowed greater autonomy in choosing their investment locations. His remarks underscore the need for collaboration between the Union and state governments to ensure a conducive environment for investors. The central government has approved nine projects in 2024 for electronic components manufacturing under the Scheme for Promotion of Manufacturing of Electronic Components and Semiconductors, with a total proposed investment of ₹7,960 crore. The total employment generation potential of these applications is 15,710 persons. Last year, Foxconn, the Taiwanese OEM manufacturing company, and Vedanta Group called off a joint semiconductor venture investment in Gujarat, only to later move the investment to Karnataka. The company is now investing around ₹1,200 crore in the southern state. Its chairman Young Liu was quoted as saying in a statement, “The unit in Karnataka will soon become the second-largest Foxconn plant after China’s unit. It will create 40,000 direct jobs, especially for middle-level educated individuals, and our investment will not stop here. In the future, we plan to explore other sectors as well.”","excerpt":"The state is also poised to introduce a revised Electronics System Design and Manufacturing (ESDM) policy to foster innovation and investment in the sector.","categories":["AI Features"],"tags":["AI (Artificial Intelligence)","karnataka","Semiconductor India"],"author_name":"Sanjana Gupta","publish_date":"2024-12-30T09:38:17","publication_year":"2024","word_count":981,"keywords":["karnataka","Go","programming_languages:R","AI","innovation","RAG","Aim","ViT","Semiconductor India","GAN","R","AI (Artificial Intelligence)","startup"],"extracted_tech_keywords":["AI","Aim","RAG","R","Go","GAN","ViT","innovation","startup","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/karnataka-gives-a-policy-push-to-semiconductor-industry\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10069556,"title":"GitHub Copilot is now available to everyone at USD 10 per month","content":"Microsoft has announced the general availability of GitHub to all developers for USD 10\/month or USD 100\/year. The AI-pair programming tool that suggests code in your editor will be free to use for verified students and maintainers of popular open-source projects. Last year, Microsoft made GitHub Copilot available for technical preview. With GitHub Copilot, AI can be broadly harnessed by developers to write and complete code. Just like the rise of compilers and open source, it is believed that AI-assisted coding will fundamentally change the nature of software development, giving developers a new tool to write code easier and faster. GitHub Copilot distils the collective knowledge of the world’s developers into an editor extension that suggests code in real-time, to help users stay focused on what matters most: building great software. Upon typing code or comments, GitHub Copilot suggests the next line of code. It can also suggest complete methods, boilerplate code, whole unit tests, and even complex algorithms. More than 1.2 million developers have used GitHub Copilot so far. In files where it’s enabled, nearly 40 percent of code is being written by GitHub Copilot in popular coding languages, like Python. That’s creating more time and space for developers to focus on solving bigger problems and building even better software. GitHub Copilot is available for a free 60-day trial. For more details, Click here.","excerpt":"GitHub Copilot distils the collective knowledge of the world’s developers into an editor extension that suggests code in real-time.","categories":["AI News"],"tags":["GitHub","Github Copilot"],"author_name":"Kartik Wali","publish_date":"2022-06-22T13:17:59","publication_year":"2022","word_count":225,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","Github Copilot","Git","Python","programming_languages:Python","GitHub","R"],"extracted_tech_keywords":["AI","Python","R","Go","Git","GitHub","programming_languages:Python","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/github-copilot-is-now-available-to-everyone-at-usd-10-per-month\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":24170,"title":"Platform Based Personalization using Data Science –  A real re-inventor in Financial services industry","content":"The last wave of innovation that transformed the Banking and Financial services industry was the implementation of various Information technology  (IT) applications which was primarily digitization of paper based processes. Ever since, the Financial services industry as a whole has been generating large volume and variety of data through this all pervasive digitization process and the industry continued to grow around increasing the processes’ robustness and improving operational efficiency on various fronts. Most of the data, which has been generated by the financial enterprises through this massive digitization process, pertain to the Customer, in terms of Customer demographics, transactions, buying, spending, borrowing etc. Also most of these customers also generate large volume and variety of data in the social digital space, which we call as Alternate data. This mammoth amount of digital data of each customer generated both inside and outside the enterprise has the potential to reinvent the entire business functions of the banking and financial services  industry should this data be used in the most scientific and meaningful way. The advancement in data engineering and data science, which as witnessed during the last decade, is facilitating very in-depth statistical analysis of Customer data. Various advanced tools which have been developed recently in Data Engineering, Data science driven machine learning  ( ML)  and Advanced Statistical analysis have created  the real potential in using the customer data to provide very personalized services in variety of  the business functions of financial industry. This process of  Personalization of business functions using Data science driven machine learning  (ML) and  Advanced analytics not only addresses several challenges that the financial institutions have been facing with their traditional approach, but also opens up new doors of opportunity for the fast-movers. Conventional methods of statistical model building using training data and then applying the same on actual data has been a cumbersome process and hence the universal adoption of statistical methods for decision science has been very slow. Performing advanced analytics using Data science driven machine learning  along with the technologies like Cloud and Big Data engineering is enabling the platform based automation of the advanced analytical process. This platform driven automation process is industrializing adoption of Data science and one of  the major applications of such  platform based automation process is providing personalized services using Data science. Pervasive Personalization At a very basic level, advanced statistical analysis  enables institutions to perform customer profiling based on their behavioural economics, and facilitates  scientific customer profiling  which goes beyond demographic segmentation to provide personalized services. By this process, the financial  institutions can provide personalized services to the individual customer  in various operations including customer service, marketing, sales etc. Also, using both the Enterprise and Alternate data of customers and automating the  personalization process using ML, brings up increased opportunities for  Financial institutions in Lending, collection, recovery , cross-selling & up-selling etc. Personalized services  for sure help financial institutions to provide enhanced customer services and hence reduce the customer churn and increase the rate of retention. This in turn helps to improve brand loyalty and much better customer engagement. Whereas the insights gained from running analytics on the internal customer service processes can help institutions to identify the areas of improvement, it also helps you leverage the customer lifetime patterns to boost the customers’ lifetime value (LTV) Automating the personalization process through Big data engineering and ML has created the development of various product platforms addressing various  business functions in the Financial institutions . … And, the Institutions Are Embracing It – Here is How Financial institutions around the world are realizing the need for  providing personalized services through ML driven automated platforms. Here are few instances that shed some light on how they are doing it: An international bank deploys a ML Driven predictive analytics based Automated Personalization  platform for optimizing their marketing effectiveness. The platform improved their targeting with personalized campaigns and enabled faster processing of large customer data sets among their Current and savings account ( CASA) holders Within 9 months of implementing this platform, the Bank could conduct on average  1.5 personalized  campaigns every week instead of  1 campaign every month before the implementation of this Automated product platform . This has resulted in approximately 60% increase in their cross-selling of various products within their own CASA customers . A very large Asian bank which has more than 50 thousand customer servicing branches deploys ML Driven platform to improve customer engagement using Face recognition and understand the behavioural patterns to enable customer delight and cross-selling and up-selling of various products. The platform enables the bank to identify their high value customers at every customer touch points at the branch level using the platform and offer them personalized Services. This has resulted in enormous enhancement in customer  delight and also increased cross-sell and Up-sell opportunities. A large Mortgage bank in India uses ML based platform for personalized collection and recovery process. By using the platform, the bank knows at any point of time the collection propensity of every defaulter and what kind of personalized follow up actions need to be applied on each customer to enhance the collection. The bank also knows which call centre agent to be  deployed for every customer so that the collection process becomes very personalized. In a span of 10 months, the platform helped the bank in increasing the collection amount by more than 2% and reduces the collection costs by more than 9% A large middle-east based bank used ML driven platform to build personalized brand building among their customers. By analysing  various digital data of the customers and analysing the various sentiments of the customers with respect to the banks operations and reputation, the platform recommends the right personalized  communication content and channel for each customer and the platform automatically execute the same. By using various alternate data and  various statistical methods such as correspondence analysis, customer value tree analysis and perceptual maps to provide personalized branding messaging. Conclusion Big Data Engineering and Data science driven machine learning applications have opened a plethora of opportunities for  Data driven Automation of various business functions in the Financial industry. Personalization is one of the fastest growing data driven automation processes and has already started delivering encouraging results to the early adopters in the BFSI vertical. Despite the challenges, such as massive data volumes, multitude of data sources, data quality, and data-related regulations,  ML driven automated personalization platforms are generating immense interests from the CXO’s of financial industry and leading the industry into an age where data rules & decides the next best methods and process for providing personalized services to very  customer.","excerpt":"The last wave of innovation that transformed the Banking and Financial services industry was the implementation of various Information technology  (IT) applications which was primarily digitization of paper based processes. Ever since, the Financial services industry as a whole has been generating large volume and variety of data through this all pervasive digitization process and […]","categories":["IT Services"],"tags":[],"author_name":"Raj MKK","publish_date":"2018-05-02T10:46:13","publication_year":"2018","word_count":1101,"keywords":["data science","Go","machine learning","AI","R","ML","Git","RAG","analytics","predictive analytics"],"extracted_tech_keywords":["AI","machine learning","ML","data science","analytics","RAG","predictive analytics","R","Go","Git"],"url":"https:\/\/analyticsindiamag.com\/it-services\/platform-based-personalization-using-data-science-a-real-re-inventor-in-financial-services-industry\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10067870,"title":"Splunk bets big on India","content":"Splunk is a data platform that helps security teams, IT professionals, developers, and just about anyone else use data to see across all of their digital systems to understand if they are secure, performing well, and where there are opportunities for improvements. “The world today is driven by an explosion of new technologies and innovations. Such advancements are fuelled by data, and create more data. So it is clear that the organisations that will thrive are those that have a strong data and technology foundation. Trusted by 90 of the Fortune 100 companies, Splunk is the data platform leader for security and observability, that helps global organisations remove the barriers between data and action so that everyone thrives in the Data Age,” said Garth Fort, senior vice president and Chief Product Officer, Splunk. In an exclusive interview with Analytics India Magazine, Garth spoke about Splunk’s journey. AIM: From network security to analytics and observability, how would you describe Splunk’s journey from 2003 to date? Garth Fort: Founded in 2003, Splunk’s mission has always been to remove the barriers between data and action. We had a simple idea of how we could use indexing and searching log files to help IT professionals troubleshoot issues in their environment. Over the past 19 years, we have expanded through 17 acquisitions to help customers address a much broader set of capabilities that help customers in IT, Security, Application Development, and Site Reliability Engineering. With the onset of the pandemic, we have witnessed how remote work and digital transformation quickly went from a priority to an urgent imperative. Today, Splunk is a global company — with over 7,500 employees across 30 regions around the world — that offers an open, extensible data platform that supports shared data across any environment so that all teams in an organisation can get end-to-end visibility, with context, for every interaction and business process. Over the years, Splunk has also invested in building a strong partner network which has seen a massive expansion. This ecosystem brings forth partners from across the IT, security, and analytics market and extends Splunk’s reach to customers in new geographies broadening its product offerings. The Splunk Partnerverse Program empowers our network of over 2,200 partners. It helps customers identify and collaborate with the right partners faster. Leverage proven solutions to reach their critical missions and outcomes in the cloud and underscores our continued commitment to our partner network. As per Full Year, 2022 Financial results ending Jan 31 2022, Splunk’s cloud revenue was USD 944 million, up 70 percent year-over-year, with total revenues at USD 2.67 billion, up 20 percent year-over-year, a testament to our growth story. We continue to invest heavily in R&D to deliver new products and features that support our core differentiation: gaining operational insights from data in real-time. AIM: Today, Splunk helps thousands of customers improve security, drive resilience, and unlock innovation. How do you do it? Garth Fort: The Splunk platform removes the barriers between data and action, empowering observability, IT and security teams to ensure their organisations are secure, resilient and innovative. When it comes to investigating, monitoring, analysing and acting on the data, Splunk stands out by breaking down barriers to action, transforming data into results, and protecting the business with its unique data platform for the hybrid world. Designed for the hybrid world: Splunk is a single platform designed for the way companies work, with the capabilities that businesses demand. ·  Splunk Cloud enables businesses to make confident decisions and take decisive action on insights from the derived data without purchasing, managing or deploying infrastructure. Splunk Enterprise scales to hundreds of terabytes per day to meet the needs of any organisation and supports clustering, high availability and disaster recovery configurations. It protects the company’s data and adheres to industry and international compliance regulations, like GDPR, with role-based access controls, secure data handling, simplified auditing and assurance of data integrity. Security: Splunk’s security products and solutions give our customers the actionable intelligence and advanced analytics to strengthen their cyber security defence. ·  Splunk Enterprise Security is our solution for Security Information and Event Management. It ingests and combs through a high volume of data in mere seconds to find and alert on unusual behaviour, offering real-time insight to protect the business. Splunk is used by large enterprises for end-to-end security operations – including posture assessment, monitoring, alert and incident handling, breach analysis and response, and event correlation. ·  SOAR (Security Orchestration, Automation and Response) technologies enable organisations to efficiently observe, understand, decide upon and act on security incidents from a single interface. It enables them to handle more incidents, investigate the most important issues more deeply, and broadly improve their overall security posture. Proud to be named a “Leader” in Gartner’s Magic Quadrant (MQ) for SIEM for eight years, Splunk’s analytics-driven security solutions provide a comprehensive approach to cybersecurity, including advanced techniques like machine learning and behavioural analytics. These techniques help security teams quickly identify, investigate, and respond to threats based on a broader security context than is possible with legacy security products. ·  Observability: The pandemic-era rush of cloud adoption has increased monitoring challenges for traditional IT teams. However, as more IT organisations are experiencing the visibility challenges of hybrid, multi-cloud infrastructure, they’re also more readily identifying observability as the solution. Splunk Observability Cloud provides full-fidelity monitoring and troubleshooting across infrastructure, applications, and user interfaces, in real-time and at any scale, to help keep services reliable, deliver great customer experience and innovate faster. Splunk Observability Cloud’s suite of products and features enables businesses to quickly and intelligently respond to outages and identify root causes while also giving the data-driven guidance teams need to optimise performance and productivity going forward. This includes Splunk Infrastructure Monitoring which helps companies gain insights and perform powerful, capable analytics on infrastructure, Splunk Application Performance Monitoring, which collects traces and spans to monitor distributed applications connected to Splunk Observability Cloud to give full-fidelity access to all the application data, and Splunk Real User Monitoring that provides insights about the performance and health of the front-end user experience of the application by collecting performance metrics, web vitals, errors, and other forms of data to detect and troubleshoot problems among others. AIM: How does Splunk use AI\/ML? Garth Fort: Splunk’s platform and products leverage AI and ML capabilities to help solve a range of business problems. A good example is Lenovo, a USD 50 billion multinational technology company providing smart devices to consumers and businesses in 180 markets globally that relies on the Splunk Observability Cloud to better respond to changing consumer preferences in the evolving e-commerce landscape. The Lenovo team uses full-fidelity data and predictive analytics to monitor infrastructure performance at the cloud scale. Splunk’s AI-driven analytics helped Lenovo cut troubleshooting time in half, reduced the total cost of ownership and maintained 100 percent uptime despite a 300 percent increase in web traffic. Another example is Hyphen Group, a leading financial technology company in Southeast Asia, that needed to bridge the gap between software development and IT operations to boost its DevOps pipeline. The cross-functional teams across engineering and infrastructure operations needed a better way to communicate and collaborate. With Splunk Observability Cloud, Hyphen Group relied on the analytics-powered platform to proactively uncover issues across the entire frontend and backend stack. The troubleshooting time has been reduced from hours to minutes, and the infrastructure deployment is now 8x faster, enabling reliable customer experience and faster time to market for new products and services. Hong Kong Internet eXchange (HKIX) is one of Asia Pacific’s largest internet exchange points, supporting fast and easy interconnections among local and international networks. HKIX wanted to move from reactive to proactive information security management, as troubleshooting from disparate devices consumed time and created efficiency challenges. With Splunk Cloud streaming and machine learning capabilities, laborious issue management has been replaced with real-time visualisations and holistic system visibility for better operations and ISO 27001 compliance. AIM: What makes Splunk unique? Garth Fort: Splunk is unique in its ability to help organisations turn data into doing. One of the early innovations in our history was our system that infers the schema of a dataset when we first read it, as opposed to having that data model defined in advance of writing the data to the hard drive. This flexibility has allowed customers to use our powerful search, analytics, and visualisation capabilities over an incredibly wide format of various sources of machine data. We’ve seen Splunk used for optimising display advertising placement on cable set-top boxes or visualising in real-time the health of the systems on ships designed to lay fibre-optic cables on the floor of the ocean. Splunk provides robust and powerful solutions to various use cases, be it security, IT ops management or IoT-based products. With its wide range of market-leading purpose-built solutions and ease of use and deployment, Splunk helps customers gain an edge over their competitors. Splunk’s platform is the only solution on the market that offers an investigative approach that allows our customers to ingest and act on data, no matter its format, across different parts of an organisation. Therefore, while Splunk may have hundreds of different competitors that tackle various point solutions, none offer the ability to do this across the business on the same platform and provides insights across all technology stacks used by various business units, be it on-premises, in the cloud or hybrid. This is uniquely Splunk and what Splunk does best- consolidating massive amounts of data across an entire organisation, digesting it and making sense of it. AIM: Though Splunk has become a global brand name, only about ~35% of its revenue base comes from international markets. Why? Garth Fort: Some of this is just a function of timing. We started in North America back when we were founded in 2002 and frankly could grow our business there at a rapid clip over nearly two decades. Our international footprint came later on, but we believe we’ll see continued rapid growth abroad through both our EMEA and APAC sales theatres. Today, Splunk Cloud serves customers in over 20 regions spanning both Amazon Web Services and the Google Cloud Platform, and we’re planning to expand to at least two more regions this year. AIM: How big is the Indian market for Splunk? What are the opportunities? Garth Fort: Per the latest IDC report, India’s public cloud services (PCS) market, including IaaS, PaaS, and SaaS, the revenue totalled USD 2.2 billion for the first half of 2021. The overall Indian public cloud services market will reach USD 10.8 billion by 2025, growing at a CAGR of 24.1 percent from 2020-25. These figures reflect tremendous growth opportunities for Splunk as a data platform leader for security and observability, in a fast-growing Indian cloud market. Making inroads into the Indian market has been enabled by a strong partner ecosystem. Successful collaboration with local partners has allowed us to reach more customers and help them unlock innovation and embrace the cloud. When it comes to advanced technology adoption, India is at the forefront compared to its global peers. Our latest State of Security Report revealed that Indian organisations are on the leading edge in adopting security infrastructure. 72 percent of Indian companies report extensive adoption versus 28 percent of their peers in the rest of the world. In addition to this, they have been integrating non-security analytics with security analytics to improve decision-making. 77 percent of the respondents reported significant integration versus 37 percent across other countries. These findings tell us that Indian business leaders prioritise cybersecurity and stay ahead by investing in robust, predictive cybersecurity solutions and analytics. Splunk’s best in class, advanced analytics solutions, proven to support an expansive set of use cases, address the needs of businesses in India. In addition to this, we recently conducted a survey which examines how observability is helping reduce costs while increasing innovation. The State of Observability Report revealed that the most sophisticated observability practitioners could cut downtime costs by 90 percent, from an estimated USD 23.8 million annually to just USD 2.5 million, compared to observability beginners. Once again, we learnt that Indian businesses are further along in the observability journey compared to their global counterparts. The report highlighted that only 29 percent of Indian organisations are rated as beginners, versus 62 percent, on average, across other countries, and 69 percent of Indian organisations were more likely to report that their start in observability was in part driven by a top-down mandate from leadership, versus 45 percent on average across other countries. These are exciting statistics for us at Splunk as it reiterates that while observability as a practice may still be relatively new to the market, business leaders in India are already implementing observability in their businesses and have benefitted from observability. The opportunity for Splunk to quickly expand to more use cases across sectors is immense. Lastly, last October, we announced the availability of Splunk Cloud Platform on Amazon Web Services (AWS) Marketplace to meet India’s digital transformation needs. Through this, Indian businesses have seamlessly procured and deployed Splunk Cloud Platform via AWS Marketplace. It ensures fast time-to-value for customers leveraging Splunk solutions to gain an additional layer of real-time security and operational and cost management insights across their AWS and hybrid environments. This availability brings great flexibility to our customers, and we are committed to bringing the solutions that matter to the India market. AIM: For long, Splunk was a license-based software vendor. Did it change after moving into SaaS model? Garth Fort: We’re constantly listening to customers for feedback around pricing and licensing. Several years ago, we shifted from perpetual to term licensing, and with the shift to a SaaS-based delivery model, we’re continuing the transition to cloud subscriptions. Our original model was based on the daily volume of data customers would ingest through Splunk. This model was simple to understand and measure and was considered a fair “pay as you go” model. As customer deployments grew, we had feedback from customers that they wanted to align their pricing to their usage of Splunk vs the raw size of the data, so we introduced Workload pricing that aligns better with how customers received value from their usage of our products. We have seen this model adopted quite rapidly since it can help customers save money while at the same time accelerating usage. AIM: What does the future hold for the company? Garth Fort: My crystal ball has been a little cloudy over the last two years, and frankly, we have all been reacting to an unpredictable pandemic. What I can say for certain is that I think our customers leaned in with Splunk as their workforce went remote, and they doubled down on their digital transformation. We have no shortage of things our customers are asking us to build, and you can see thousands of suggestions for yourself by visiting the Ideas page on our website. We also hired thousands of new Splunkers over the last two years across every part of the company. Globally, Splunk is a trusted technology provider to both government agencies across the Asia Pacific as well as to several multinational companies and financial institutions- from the City of Gold Coast, Australia, the Government Service Insurance System in the Philippines to the Tokyo Stock Exchange and the Hong Kong Internet eXchange to name a few. In India, Splunk is actively supporting these sectors, including providing comprehensive cloud services and solutions to one of the largest private banks in India.As the Government of India takes aggressive initiatives toward digital adoption, Splunk is committed to the India market. We believe that we will play an important role in enabling organisations in India to thrive in the digital age.","excerpt":"We are constantly listening to customers for feedback around pricing and licensing.","categories":["AI Features"],"tags":["Interviews and Discussions","Splunk"],"author_name":"Sri Krishna","publish_date":"2022-05-26T14:00:00","publication_year":"2022","word_count":2624,"keywords":["Go","machine learning","AWS","AI","R","ML","Splunk","RAG","Aim","analytics","predictive analytics","Interviews and Discussions"],"extracted_tech_keywords":["AI","machine learning","ML","analytics","Aim","RAG","predictive analytics","AWS","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/splunk-bets-big-on-india\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10172450,"title":"World Economic Forum Names 10 Indian Startups as Technology Pioneers for 2025","content":"The World Economic Forum has announced the 2025 cohort of its Technology Pioneers community, a group of 100 early-stage companies from 28 countries driving innovation across industries and borders, an official statement said.Here is the list of Indian startups named as Technology Pioneers for the year. Featuring on the list are Agnikul, a startup focused on providing affordable and customisable space launch services, and CynLr, which is building robots with an intuitive vision to enable fully automated factories for manufacturers and logistics providers. In healthcare, Dezy leverages AI-powered diagnostics to make dental care more accessible and affordable. There’s also Digantara that supports commercial space operators and global agencies with space surveillance intelligence. Equal offers an integrated platform that combines identity verification with consent-based financial data sharing. Tackling clean mobility, Exponent Energy is making 15-minute EV charging a reality through innovations in battery management and charging infrastructure. Freight Tiger is streamlining logistics by building India’s largest software-enabled freight network. In aerospace intelligence, GalaxEye is developing a multi-sensor Earth observation system. SolarSquare is accelerating India’s residential solar adoption with its full-stack solar solutions. Meanwhile, the ePlane Company is building flying electric taxis aimed at transforming intra-city transportation.Now in its 25th year, the programme celebrates its strongest cohort yet, marked by broader geographical representation, greater diversity beyond Silicon Valley and the rise of more ambitious frontier technologies. Reflecting wider shifts in the innovation landscape, many of the companies spotlighted are using artificial intelligence (AI) to reach greater scale and sophistication with fewer resources, it said. Several companies are venturing into less-explored frontiers – from asteroid mining and flying electric taxis, to leveraging satellite imagery to transform agriculture and harnessing energy from supernova explosions to locate critical minerals beneath the Earth’s surface. The geography of innovation is also evolving, it said, highlighting that while the United States remains the top contributor to the community, Europe’s share has surged to 28% – up from 20% last year – reflecting the rise of strong tech ecosystems across the region. China and India are also emerging as major tech innovation hubs, it said. “There has never been a more exciting time to dive headfirst into tech innovation. But no one gets far alone – you need a community to move your mission forward,” said Verena Kuhn, head of Innovator Communities, World Economic Forum. “As we mark 25 years of the Technology Pioneers programme, this global community continues to connect start-ups to the networks and ecosystems they need to scale.” Since its inception in 2000, the community of Technology Pioneers programme has championed early-stage innovation and recognised more than 1,200 companies, many of which have gone on to reshape industries worldwide, WEF said in the statement. Alumni include household names such as Google, PayPal, Dropbox and SoundCloud, underscoring the community’s role as a launchpad for ideas and impact. “The 2025 cohort stands out for its concentration of companies developing breakthrough technologies to address pressing global challenges. These include advanced robotics, customisable space launch services, micro nuclear reactors and more accessible quantum computing applications,” the forum said.","excerpt":"Most of these companies leverage AI to achieve greater scale and sophistication with fewer resources.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","Startups","World Economic Forum"],"author_name":"C P Balasubramanyam","publish_date":"2025-06-26T15:52:53","publication_year":"2025","word_count":508,"keywords":["Go","artificial intelligence","World Economic Forum","AI","innovation","ML","RAG","Aim","GAN","Startups","R","AI (Artificial Intelligence)","startup"],"extracted_tech_keywords":["AI","artificial intelligence","ML","Aim","RAG","R","Go","GAN","innovation","startup"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/world-economic-forum-names-10-indian-startups-as-technology-pioneers-for-2025\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10061640,"title":"How AI is revolutionising contract management","content":"Contracting is a widespread activity, yet it is one that only a few businesses execute well. In fact, it’s been estimated that inefficient contracting costs businesses anywhere from 5% to 40% of the value of a deal, depending on the conditions. However, new technical advancements such as artificial intelligence (AI) are now assisting businesses in overcoming many contracting issues. The biggest problem businesses confront when it comes to contracting is the vast amount of contracts they must keep track of; these contracts are often inconsistent and difficult to organise, maintain, and update. Because most companies don’t have a database of all the information in their contracts – let alone an effective means to extract it – there’s no easy way to evaluate complex outsourcing agreements or compare how a condition is worded across divisions. Not only does it take a lot of labour to create, execute, and improve contracts, it also takes a lot of manpower to draft, execute, and improve contracting processes and the transactions that these contracts regulate. How AI is revolutionising the concept of contract management If a large tech company, for example, has a high volume of procurement contracts with various renewal dates and renegotiation terms, reviewing and tracking all of this information would take hundreds of hours and a staff of contract managers to ensure that no renewal or opportunity is missed. AI software, on the other hand, can readily extract data and explain contract content. (Any number of contracts’ renewal dates and renegotiation clauses might be rapidly pulled and organised.) As a result, it can help businesses evaluate contracts faster, organise and locate vast amounts of contract data more readily, reduce the risk of contract disputes (and adversarial contract negotiations), and improve the number of contracts they can negotiate and sign. Benefits of AI-based contract management You’ll never miss a renewal or revenue opportunity with proactive AI-supported alerts: Legal teams are in charge of keeping track of crucial contract lifecycle events, such as contract autorenewals and other data-driven events. Legal teams may remain on top of contracts with the help of contract AI by setting up automated alerts and notifications. These notifications can be issued before any date-based event, allowing legal teams to gather the information that can be used to re-negotiate contract terms before renewal. Reduced risk and improved contract visibility: One of the legal team’s major responsibilities is to guarantee that the corporation signs off on solid acquisitions and minimises risk. When there is a lack of visibility into each contract at every point of the contract lifecycle, liability and risk are increased. Legal contract automation and artificial intelligence (AI) can assist legal teams in ensuring that the company is signing off on excellent deals, avoiding risks, and having complete insight into all contract requirements. Authoring with Intelligence: An AI-enabled contract management system facilitates intelligent contract development by automatically detecting and arranging contracts. CMSes use machine learning algorithms to extract information from prior contract track records, identify patterns, and recommend permitted, acceptable pre-defined phrases and contract clauses. In addition, an AI-enhanced CMS’s context-based analysis is a crucial component. This means that the CMS searches for related conditions that should be used for specific contract types while generating contracts. For example, if the dealer is located outside of the country and across borders, a currency fluctuation clause in the procurement contract will be recommended. AI and agreement data Users of artificial intelligence for contract management can choose certain data points and set rules based on contract kinds. For example, users can set rules for locations, dates, monetary values, personally identifiable information (PII), payment card information (PCI), and other subsets of key data for a given contract using this intelligent contract management software functionality because data is consolidated and easily digestible. Once enabled, contract intelligence automates these operations, saving time spent on tedious data import processing and allowing more time to be spent on data-driven decisions. What is next for Contractual Management AI Contract lifecycle management (CLM) technology has become a standard part of modern corporate business practice. It is expected to increase at an annual rate of 11% on average until 2030 as more companies use CLM platforms. The automated procedures and processes take care of a lot of the data entry and paperwork that legal teams have dealt with in the past. However, as we approach 2022, basic automation will not be enough in an era where time is money, and legal teams are increasingly positioned to play a new strategic role in enterprises. Contract intelligence is a next-generation approach to contract lifecycle management that enables sales, procurement, legal, and other teams across the business to obtain complete visibility into past and current contracts to track commitments, eliminate risk, and optimise revenue.","excerpt":"Legal teams may remain on top of contracts with the help of contract AI by setting up automated alerts and notifications.","categories":["AI Features"],"tags":["Contract"],"author_name":"Abhishree Choudhary","publish_date":"2022-02-27T18:00:00","publication_year":"2022","word_count":790,"keywords":["Go","API","machine learning","artificial intelligence","AI","RAG","automation","ViT","GAN","R","Contract"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","RAG","R","Go","API","GAN","ViT","automation"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-ai-is-revolutionising-contract-management\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10168888,"title":"Open-Source ‘Parlant’ Fixes Hallucinations in Enterprise GenAI Chatbots","content":"The capabilities of generative AI have prompted businesses to explore its potential in customer service, but the underlying technical blockers remain significant. Recently dubbed “The Giant Engineering Problem That Nobody Else on Earth Has Been Able to Solve”, Large Language Model “hallucinations” expose businesses deploying customer-facing AI to intolerable risks. Hallucinations happen because, at their core, LLMs generate responses through a probabilistic, token-by-token, autoregressive process. The model continuously selects what it sees as the most likely tokens from an extensive “token vocabulary” that can span hundreds of thousands of tokens. For example, OpenAI’s GPT-4o has a vocabulary size of nearly 200,000 tokens. This token selection process is inherently error-prone, as each probabilistic prediction relies solely on the preceding context. This often leads to many different types of hallucinations and deviations from critical service protocols. Such unpredictability poses a significant challenge in high-stakes environments where consistent behavior is non-negotiable. Some try to address unpredictability in chatbots by employing traditional solutions to confine LLM responses through rigid flow charts, as seen in frameworks like LangFlow, LangGraph, or Rasa. These solutions guide interactions along linear paths, but this is already known to fail at managing real-world queries that may involve multiple intents and conversational paths that deviate from the flow designer’s vision. Moreover, adjusting responses in these contexts frequently necessitates tedious manual edits to flows and fragile modifications of prompts, posing risks of protocol breaches and unintended consequences. But even after all this, critical hallucinations still occur at an unacceptable level. For example, if you’ve managed to use such frameworks to increase accuracy and correctness to an unprecedented 99%, that still exposes a bank servicing 1 million daily conversations to 10,000 new customer-facing mistakes to deal with every day, many of which can be practically unlimited in scope and severity. This is why enterprises are still averse to deploying customer-facing GenAI. But with Parlant, a framework now embraced by some of the largest financial services companies in the world, this is finally starting to change. Fixing an LLM’s Achilles Heel Parlant adopts a fundamentally different approach by developing an open-source conversational AI engine that allows developers to take control of their user-facing AI agents. Parlant is built by Emcie, an up-and-coming startup with leading software engineers from Microsoft, EverC, Check Point, and Dynamic Yield, along with natural language processing (NLP) researchers from the Weizmann Institute of Science, in collaboration with world-class Conversation Design experts from the Conversation Design Institute. Parlant enables an AI Conversation Modeling system that automatically tailors responses from a large and dynamically controlled selection of pre-approved “utterances.” Using these new conversation modeling paradigms, organisations can precisely control GenAI communications while maintaining the level of naturalness and flexibility expected of LLMs, as operators and designers can manage and refine utterances with adjustable freedom levels, and Parlant’s engine applies intelligently applies them at the right time based on situational awareness and guidelines that you can provide it. To simplify creating these utterances while prototyping, Parlant offers a ‘Fluid Composition’ mode where AI generates natural responses. This mode allows conversational designers to extract and tweak these auto-suggested responses into approved utterances while experimenting with their AI agents iteratively during development. Once established, the system switches to the ‘strict’ mode, exclusively using pre-approved utterances to construct responses. This ensures predictability and control while preserving the AI’s ability to creatively address diverse inquiries by intelligently utilising a large set of approved utterances using an LLM’s natural capabilities to select the best responses precisely. Parlant analyses the conversation context at runtime, determines the relevant set of utterance candidates, and dynamically applies them to produce a response. It also filters and selects guidelines based on the context, allowing the developer to achieve a high degree of behavioural control over their agents without sacrificing the ability to scale the agents’ complexity. This runtime filtering of guidelines enables developers to support more conversational use cases while maintaining focused behaviour from their LLM in many different situations. Moreover, Parlant lets you easily troubleshoot by tracing how and why each utterance was applied for any given response. This is made possible by using highly descriptive and explainable log outputs, produced by the LLM during the utterance selection process. Parlant, an open-source project, is LLM-agnostic, meaning that it supports multiple LLM providers, including OpenAI, Google, Meta, and Anthropic, via multiple inference providers. Prompt-Level Innovations Improve LLM Instruction Following What allows Parlant to ensure aligned and expected outcomes from LLMs lies in the team’s research focus on techniques to gain control over LLMs. Emcie, the startup company behind Parlant, earlier this year published a research study titled ‘Attentive Reasoning Queries (ARQ): A Systematic Method for Optimising Instruction-Following in Large Language Models’. The study outlines methods to optimise instruction following in LLMs. Unlike free-form reasoning approaches such as Chain-of-Thought (CoT), Attentive Reasoning Queries (ARQs) guide LLMs through systematic, targeted queries that reinforce critical information and instructions and prevent hallucinations and attention drift. The research also revealed test results where ARQs achieved a 90.2% success rate in correctly interpreting and applying instructions, outperforming CoT reasoning and direct response generation. The study also revealed that ARQs have the potential to be more computationally efficient than free-form reasoning when carefully designed.","excerpt":"Parlant enables an AI conversation modeling system that automatically tailors responses from a large and dynamically controlled selection of pre-approved “utterances”.","categories":["AI Trends"],"tags":["AI (Artificial Intelligence)","hallucinations"],"author_name":"AIM Media House","publish_date":"2025-04-30T09:50:25","publication_year":"2025","word_count":864,"keywords":["Anthropic","GenAI","OpenAI","AI","chatbots","GPT-4o","LangGraph","hallucinations","RAG","NLP","generative AI","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","NLP","generative AI","GenAI","GPT-4o","OpenAI","Anthropic","LangGraph","RAG","chatbots"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/open-source-parlant-fixes-hallucinations-in-enterprise-genai-chatbots\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":57342,"title":"Facebook Explains How Nearest Neighbour Search Is An Effective Approach For Language Modelling In The Long Tail","content":"With an aim to break down language barriers across the globe for everyone to understand and communicate with anyone, the researchers at Facebook AI Research (FAIR) work on complex problems to deploy robust language translation solutions. It spans the topics such as deep learning, natural language processing, text normalisation, word sense disambiguation and much more. Recently, the researchers at Facebook AI Research presented a new language modelling approach known as kNN-LM, which is based on the hypothesis that the representation learning problem may be easier than the prediction problem. Usually, a natural language model resolves two subproblems, which are mapping sentence prefixes to fixed-sized representations and the other one is to utilise these representations to predict the next word in a certain text. This approach extends a pre-trained LM by linearly interpolating its next word distribution with a k-nearest neighbours (kNN) model. The nearest neighbours are then computed according to the distance in the pre-trained embedding space and can be extracted from text collection, including the original language model (LM) training data. According to the researchers, this approach allows the rare patterns to be memorised explicitly, rather than implicitly in model parameters. How It Works The kNN-LM involves augmenting such as pre-trained LM with the nearest neighbours retrieval mechanism, without any additional training, which means the representations learned by the LM remain unchanged through the process. One crucial point of this kNN-LM approach is that it is compatible with any model which produces fixed-size context representations. The researchers used decoder-only transformers for language modelling, and since kNN-LM makes no changes to the underlying LM, the researchers used the exact architecture for creating a kNN-LM for inference. According to the researchers, kNN-LM improves performance because of the following three points With an implicit notion of similarity, the Transformer LM is efficient at learning a representation function for contexts.While the Transformer has the capacity to memorise all the training examples, doing so causes its representation to generalise less effectivelyThe kNN-LM allows the model to memorise the training data while retaining an effective similarity function. Dataset Used The researchers used several datasets for this project, they are mentioned below Wikitext-103, which is a standard benchmark for autoregressive language modelling with a 250K word-level vocabulary consists of 103M tokens of Wikipedia in the training set and 250K tokens in each of the test sets.Books dataset, which is the Toronto Books Corpus.Wiki-3B is an English Wikipedia dataset which contains about 2.87B tokens.Wiki-100M is a random 100M token subset of Wiki-3B Corpus. Advantages of This Model: This approach has implications for efficiently scaling up to larger training sets and allows for effective domain adaptation, by simply varying the nearest neighbour datastore, again without further trainingThe model is specifically helpful in predicting rare patterns, such as factual knowledge, names, and near-duplicate sentences from the training set., It also improves performance when the same training data is used for learning the prefix representations and the kNN model, strongly suggesting that the prediction problem is more challenging than previously appreciated. Wrapping Up Over the years, the researchers at FAIR have performed several remarkable projects related to Natural language processing (NLP) and Natural Language Understanding (NLU). The researchers introduced the kNN-ML model, which significantly outperform standard language models by directly querying training examples at test time and can be applied to any neural language model. According to the researchers, the success of this method suggests that learning similarity functions between contexts may be an easier problem than predicting the next word from some given context. Read the paper here.","excerpt":"With an aim to break down language barriers across the globe for everyone to understand and communicate with anyone, the researchers at Facebook AI Research (FAIR) work on complex problems to deploy robust language translation solutions. It spans the topics such as deep learning, natural language processing, text normalisation, word sense disambiguation and much more.   […]","categories":["AI Features"],"tags":["AI Research","Facebook","Facebook AI","knn","language modelling"],"author_name":"Ambika Choudhury","publish_date":"2020-02-24T10:00:00","publication_year":"2020","word_count":589,"keywords":["Facebook AI","knn","AI","programming_languages:R","ML","ai_frameworks:Transformers","AI Research","Transformers","NLP","Aim","deep learning","AI research","Facebook","language modelling","R"],"extracted_tech_keywords":["AI","ML","deep learning","NLP","Aim","Transformers","R","AI research","ai_frameworks:Transformers","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/facebook-explains-how-nearest-neighbour-search-is-an-effective-approach-for-language-modelling-in-the-long-tail\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":27360,"title":"Here’s Why People Are Protesting Against Amazon’s ‘Racist’ Facial Recognition Software","content":"In a recent incident, Amazon’s face recognition technology received a lot of flak as it erroneously identified 28 US congressmen with as criminals. Rekognition, Amazon’s facial recognition tool which was introduced in 2016, delivered disappointing results as it identified the public personalities as people who have been arrested for crime. The test, which was conducted by ACLU of Northern California to check its efficiency in facial recognition, included a wide range of datasets — of men, women, legislators, Republicans, Democrats of all ages from across the country. The ACLU in a statement had said that it had used exact same facial recognition system that Amazon offers to the public, which anyone could use to scan for matched between images of faces. The running of entire test had cost them a meager $12.33. What caught the curious eyes around the whole case was that despite being one of the biggest names in the tech industry, Amazon’s Rekognition failed to deliver. It has raised serious concerns against facial recognition as a technology and its legitimacy and accuracy. What Is Rekognition? Introduced in Nov 2016, the image recognition software by Amazon promised to analyse billions of images and videos daily, to recognise objects, people, text and inappropriate content. The website also claimed to be providing highly-accurate facial analysis and facial recognition on images and videos provided by the user. It claimed to be capable of detecting, analysing and comparing faces for a wide variety of user verification and public safety use cases. Amazon Rekognition is based on highly-scalable deep learning technology, which has been developed by Amazon’s computer vision scientists. Some of its benefits are that it includes simple integration with easy to use APIs that don’t require any machine learning expertise, includes continuous learning, and the cost involved is really low. According to the company, Rekognition can identify people in real-time by instantaneously searching databases containing millions of faces. It also claimed that the system can identify faces in group photos, crowd, public places such as airport, among others. What Went Wrong When ACLU Used Rekognition For Surveillance And Recognition In a recent instance, the ACLU, scanned the faces of all 535 members of the congress against the database of 25,000 publicly available mugshots using Amazon’s open Rekognition API. It was essentially done to test the system’s accuracy. This database was searched against each of the current member of the House, and then identified with default match settings available with Amazon Rekognition. Unfortunately, the test by ACLU showed incorrect identification of faces, along with an indication of racial bias, which is an underlying problem of many facial recognition systems that exist today. It was reported that 11 of the 28 falsely matched people were of colour, which amounts to 39 percent of the total number. It is despite the fact that people of colour make up only 20 percent of those in Congress. The facial recognition system especially shows high error rates in case of women and African-Americans as well. “Our test reinforces that face surveillance is not safe for government use,” Jacob Snow, a technology and civil liberties attorney at the ACLU Foundation of Northern California, said in a statement. “Face surveillance will be used to power discriminatory surveillance and policing that targets communities of colour, immigrants, and activists. Once unleashed, that damage can’t be undone.” Rekognition had earlier delivered results for the Washington County Sheriff’s Department in Oregon, where it used images with a database of 300,000 mugshots. The Concerns In a letter written by ACLU, addressing CEO Jeff Bezos, they have clearly expressed the need to protect civil rights and liberties, and safeguarding communities. Implying a probable harm to people with colour, undocumented immigrants and protesters, the letter demanded for Amazon to stop powering a government surveillance infrastructure that poses a grave threat to customers and communities across the country. “Amazon should not be in the business of providing surveillance systems like Rekognition to the government,” the letter said. It noted that Amazon’s Rekognition product runs counter to its earlier shared values of supporting First Amendment freedoms and discriminatory bans on Muslims. In the letter, ALCU chief also suggests that Amazon must act swiftly to stand up for civil rights and civil liberties, including those of its own customers, and take Rekognition off the table for governments. ACLU strongly suggests that people should be able to walk freely on the streets without being watched by the government, and technologies like faulty facial recognition system threatens this freedom. Keeping these concerns as the focal point, as many as 70 civil rights groups, 400 members of the academic community and more than 150,000 members of the public have already spoken up to demand that Amazon stops providing face surveillance to the government. Amazon’s Reaction The company reportedly said that poor calibration was the culprit behind all the chaos, while also suggesting that the technology has many useful purposes, including finding abducted people or lost children, among others. Amazon also said that the ACLU tests were performed using Rekognition’s default confidence threshold of 80 percent, which, according to the company suits best and gives accurate results for photos of animals or chairs — not humans. The company suggested that the recommended threshold for law enforcement applications is around 95 percent. Concluding Note Whether or not the company is at fault, the bottom line is that it induced a feeling of racism. It is high time that the errors are corrected and laws are enforced for the right use of technologies such as face recognition. It is not correct to use the technology until all harms are considered and necessary steps are taken to prevent them from harming vulnerable communities. It is important for companies bringing these technologies to be highly responsible and ensure that it is correctly deployed in the communities.","excerpt":"In a recent incident, Amazon’s face recognition technology received a lot of flak as it erroneously identified 28 US congressmen with as criminals. Rekognition, Amazon’s facial recognition tool which was introduced in 2016, delivered disappointing results as it identified the public personalities as people who have been arrested for crime. The test, which was conducted […]","categories":["Global Tech"],"tags":["best facial recognition software","facial recognition system","real time face recognition software"],"author_name":"Srishti Deoras","publish_date":"2018-08-16T12:51:33","publication_year":"2018","word_count":969,"keywords":["Go","facial recognition system","machine learning","AWS","AI","best facial recognition software","image recognition","Scala","computer vision","Aim","deep learning","R","real time face recognition software"],"extracted_tech_keywords":["AI","machine learning","deep learning","computer vision","Aim","image recognition","AWS","R","Go","Scala"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/heres-why-people-are-protesting-against-amazons-racist-facial-recognition-software\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10043192,"title":"Why MIT &#038; Harvard Let Go Of Its edX Platform?","content":"Online learning platform edX, launched in 2012 by MIT and Harvard University, is set to be acquired by 2U, a publicly-traded higher education technology company, for $800 million. The proceedings will go to a non profit organisation (led by Harvard and MIT) that will oversee and improve the Open edX platform, as well as look for new ways to improve the effectiveness, engagement, and personalisation of online learning. In a letter, L. Rafael Reif, President, MIT, informed that the edX board, with support of the senior leadership of Harvard and MIT, has decided to sell the assets of edX to 2U, Inc. Once the deal comes through, 2U will be able to adopt edX as a subsidiary and take over the current portfolio of edX course offerings, user-based and partner arrangements. However, as per the deal, the open edX open-source software platform would not be transferred to 2U. Open edX is a platform on which edX and more than 2,000 learning sites across the world are based. Deal explained Harvard and MIT founded edX is a major open online education provider in the United States. It offers online university-level courses in a variety of fields to a global student base, with some courses available for free. Over its nine years journey, with over 3,000 courses and over 1.4 million certifications granted to students, edX has played an instrumental role in the development of a thriving market for college-level content. Now, edX will be turned into a public benefit company. Meanwhile, the capital from the deal will be utilised to provide next-gen online education. The major focus will be on eliminating persistent disparities in online learning by leveraging artificial intelligence to enable personalised learning that responds and adapts to the individual learner’s style and needs. A board appointed by MIT and Harvard will oversee the venture. Moreover, the future projects will incorporate ideas from current edX partners as well as MIT and Harvard faculty. The rise in demand for online learning in recent years and the sudden upsurge amid pandemic have sparked investors’ interest. With investments pouring in from all quarters, the education platforms have gone on overdrive to gain a competitive edge. And edX, as a nonprofit organisation, could no longer keep up, and was forced to look for alternative ways to sustain edX’s mission into the future, clarified Reif. What’s next MIT’s Office of Digital Learning will build and operate MITx Online based on Open edX, which MIT is creating for MITx massive open online courses (MOOCs). MIT faculty will have many alternatives for their MITx classes, including continuing to offer them on edX, moving them to a new platform called MITx Online, exploring possibilities like OpenCourseWare or the Open Learning Library, or removing them entirely. The partner edX universities are still free to provide their courses under the same terms and conditions. Additionally, there are no constraints on these institutions and their ability to offer their courses through other platforms or to leave edX. Under the agreement, 2U has agreed to protect the data of over 39 million learners to date and data-usage standards to further protect new edX users. The deal is expected to close this fall, subject to usual closing conditions and regulatory and governmental clearances, including the approval of the Massachusetts Attorney General’s Office.","excerpt":"Online learning platform edX, launched in 2012 by MIT and Harvard University, is set to be acquired by 2U, a publicly-traded higher education technology company, for $800 million. The proceedings will go to a non profit organisation (led by Harvard and MIT) that will oversee and improve the Open edX platform, as well as look […]","categories":["IT Services"],"tags":[],"author_name":"kumar Gandharv","publish_date":"2021-07-08T17:00:00","publication_year":"2021","word_count":549,"keywords":["Go","API","artificial intelligence","programming_languages:R","AI","programming_languages:Go","Git","RAG","GAN","R"],"extracted_tech_keywords":["AI","artificial intelligence","RAG","R","Go","Git","API","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/why-mit-harvard-let-go-of-its-edx-platform\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10045746,"title":"Hands-On Guide To RecBole Recommendation Algorithms","content":"Have you ever noticed that many of the product selling, service providing, and content providing companies like amazon, Oyo, youtube, Netflix, and Flipkart provide suggestions or recommendations about their products, services, and content us? Many of the recommendations we use because we know the quality of those products and many of the products we use because the platform where we are surfing for products or services knows about the product that hopefully we may like or require to use in future. How do those recommendations come out in front of us while using the platforms? The companies we are talking about are some leading companies in the market in today’s scenario. One of the factors that adds to their current stature is their recommendation system. Recommendation systems are designed to give recommendations about the user’s services, products, or items based on different factors. These systems predict probabilities about the products which are used by users most likely.  Using those recommendation systems, a user can figure out their requirements. These systems deal with the information provided by the user about the user or about the product they want to use. It finds out the match between users information and product requirements and imputes the similarity between users and items for recommendation. Recommendation systems are beneficial for both sides, for the user side as well as the provider. It benefits the user to make it easy and save time for searching the things he is interested in and for the provider to place his services in the right place and improve the user experience.  Many things can be recommended like movies, books, news articles, videos, etc., and many platforms are using it to grow their business. There are many types of recommendation systems. Some of them are- Popularity-based recommendation system – This system recommends the products in trend like a youtube trending page where the page recommends the videos viewed by the users in high frequency.Content-based recommendation system – This type of recommendation system provides the recommendation based on the similarity of the content; for example, in Netflix, the platform will recommend similar movies to the viewer after watching the movie.User-based recommendation system – In this type of recommendation system, the system calculates the similarity between the users based on users information like gender, age, like, dislike etc. and then gives the recommendation based on the similar user’s service consumption data. For example, Amazon gives recommendations about the kids’ accessories based on the age of the kids. These all systems are somewhere having different algorithms for implementation in a platform. And to assign recommendation systems to a platform we all need to imply all recommendation systems differently. Because, commonly, every business needs higher user engagement in their platform to grow more. If a user finds the platform useful for him, he uses the platform frequently, and a recommendation system helps provide a better user experience to platform users. Choosing recommendation systems to embed with a platform requires a lot of knowledge. Also, it is the subject of research because a single failure can cause the loss of several customers. And that can also harm the business too. And this subject is too deep; what we have discussed earlier is just the basic things about the recommendation system. Implementation of the recommendation system is needed to be very accurate, and also, by the time of implementation, it should give a high level of accuracy and proper recommendations. Here RecBole takes part in, which can make these all tasks very easy to perform. Next, in the article, we will discuss the RecBole framework designed for research purposes in recommendation systems. What is RecBole? RecBole,’s library consists of 72 Algorithms of recommendation systems. It is developed in Python on PyTorch framework. It produces recommendation algorithms in a unified and comprehensive way by covering four major categories of recommendation algorithms. General recommendationSequential recommendation.Context-aware recommendation.Knowledge-based recommendation. Also, with all these algorithms in its library, we have 28 recommendation data sets available, which are designed to be flexible with the algorithms. There are some scripts also available to preprocess the datasets to fit in those algorithms. The following image can show the overall architecture of the RecBole. Image source Implementation Of RecBole in Python Let’s see how we can work with RecBole. Requirements – python 3.6 or above version,  torch version 1.6.0 or above , CUDA version 9.2 or above, NVIDIA driver version >= 396.26 (for Linux) or >= 397.44 (for Windows10). Here in this article, we are using google colab, so we need to start with mounting the drive in our colaboratory; we can use the following codes for mounting the drive. from google.colab import drive drive.mount('\/content\/drive') Output: We can access the authorization code from the provided link after copying the code, paste it in the widget, and press enter. Our google drive will get mounted in the drive. We can install the Recbole using the following command. !pip install recbole Output: Or we can directly clone the package from GitHub using the following command. !git clone https:\/\/github.com\/RUCAIBox\/RecBole.git Output: After cloning the package, we just need to direct the colaboratory to the package folder. cd RecBole Output: After directing the colaboratory, we can install the Recbole framework using the following command. !pip install -e . --verbose Output: After installation, we can run the provided script by the package for initial usage of the library. This script will run the BPR model in ml- 100k dataset. The data sets consist of movie title   release year and their class like a thriller, action, adventure etc., with their unique id. !python run_recbole.py That will provide different outputs, which will consist of details of the system, model, data set, and final result. Output: In the above output, we can see the system configuration of the whole process. In the above output, we can see the details of the model in training. In the above output, we can see the evaluation metrics of the script. In the above output, we can see the details of the data. And at the end, we can see the best result of the algorithm. We can also set the different parameters using the following command. python run_hyper.py --model=[model_name] --dataset=[data_name] --config_files=xxxx.yaml --params_file=hyper.test In this command, the hyper.test file consists of those parameters. learning_rate loguniform -8, 0embedding_size choice [64, 96 , 128] train_batch_size choice [512, 1024, 2048] mlp_hidden_size choice ['[64, 64, 64]','[128, 128]'] Where loguniform indicates, the parameter will follow the randomly taken values between  e^{-8}- e^{0} in a uniform distribution. And choice allows you to take discrete values from the setting list. Basically, this file is providing the hyperparameter tuning of the model. We can use different numbers in the hyper.test file available in RecBole\/hyper.test location. For example, I am setting the parameters in the following codes. !python run_hyper.py --model=BPR --dataset=ml-100k --params_file=hyper.test Output: The output will start like this, and it will take some time to get completed. Here is the final result of the model after hyperparameter tuning. Validation result. Test result: Here we can see the results are quite satisfying, and also, we can just make a recommendation system in only a few steps. The library is tested on different datasets like Ml-1m, Netflix, yelp data. We can check the whole details of the models and performance in this link. Here in the article, we have seen how to generate a recommendation model. We could also iterate between different models by just putting a single line of command, and the RecBolde made the whole recommendation system easy and fast. The code used in the start took just 1 minute to complete, and the hyperparameter tuning took around 3-4 minutes to run successfully. Using RecBole we don’t need to worry about the algorithms. We also have got comprehensive details about the whole process running in the background. I encourage you to go to the link and learn more about the library because it can generate models with high precision levels. References: RecBole github.RecBole homepage.Google colab for implemented codes in this article.","excerpt":"RecBole is a python package for various recommendation system algorithms. which covers four major categories of recommendation system with 72 algorithms","categories":["Deep Tech"],"tags":["AI (Artificial Intelligence)","Data Science","Data Scientist","Deep Learning","Guide","Machine Learning","Python","recommendation","recommendation engine","recommender systems"],"author_name":"Yugesh Verma","publish_date":"2021-08-11T15:00:00","publication_year":"2021","word_count":1335,"keywords":["TPU","R","CUDA","PyTorch","RAG","Data Science","Guide","AI","ML","Machine Learning","recommendation systems","recommender systems","Python","recommendation engine","Colab","recommendation","Deep Learning","Data Scientist","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","ML","PyTorch","Colab","RAG","recommendation systems","TPU","CUDA","Python","R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/hands-on-guide-to-recbole-recommendation-algorithms\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10011622,"title":"Intel Rivals NVIDIA &#038; AMD As It Introduces Its First GPU For Data Centers","content":"Chipmaker Intel unveiled its first discrete GPU for data centres. Named Intel Server GPU, it is designed to power Android-based cloud gaming and media streaming segments. A low-power chip, Intel Server GPU will be used in the new PCIe graphics card which can be placed inside the servers. Interestingly, Raja Koduri, Intel Senior Vice President had released a teaser picture of a ‘mysterious GPU’ on his Twitter account in May, leading many to believe a future announcement related to GPU aimed at the data centre. This was called the “baap of all” by our team….The “Baahubali of all” is baking as well. Let’s hope the wait will be shorter than what @ssrajamouli put us through for @BaahubaliMovie :) https:\/\/t.co\/psWsIL5rnp— Raja Koduri (@RajaXg) May 2, 2020 Both XE-HPC based Ponte Vecchio (detailed at SC’19) and XE-HP based chip images shared recently are data center targeted GPUs. We need incredible computing solution range from edge to cloud to keep up with the data we are generating in real-time https:\/\/t.co\/GRIYKev5gw— Raja Koduri (@RajaXg) May 3, 2020 The graphics card market has been largely dominated by AMD and Nvidia. However, with the new launch, Intel has made inroads. Notably, Intel already has launched Iris Xe Max, a discrete GPU for laptops. During the announcement of Intel Server GPU, the company spokesperson also said that they would be releasing Intel oneAPI toolkit, to include vector-matrix-spatial architectures in developer experience, in December, along with new capabilities in the software stack as part of the company’s combined hardware and software design approach. Intel Server GPU With the inclusion of Intel Server GPU, Intel aims at expanding its existing suite of platform innovations for improved gaming and media experiences. It is based on the Xe-LP microarchitecture, arguably the most energy-efficient graphics architecture from Intel. When used along with Intel Xeon Scalable processors and open-sourced licensed software, Intel Server GPU offers high-density, low-cost solutions for Android cloud gaming and video streaming. Further, with this arrangement, service providers can scale the graphics separately from server count that supports the number of streams and subscribers per system. Intel is currently working on bringing various service partners including Gamestream, Tencent Games, and Ubitus for a market release. Intel has also introduced an enterprise-grade graphics card XG310, built specifically for the purpose of cloud streaming. It can accommodate four Intel Server GPUs, with each having a base clock speed of 0.9GHz and a boost clock speed of 1.1 GHz. It can run on low power costs, even less than counterpart Nvidia RTX 3080 card, which helps clients consolidate their server into smaller footprints. Intel oneAPI Toolkit The oneAPI industry initiative was first introduced at the SuperComputing 2019 event. It allows for a simplified cross-architecture programming model that performs without proprietary lock-in and enables the integration of legacy code. It eliminates the need for rewriting the software for the subsequent architecture and platform. Intel oneAPI toolkits use hardware capabilities and instructions such as Intel AVX-512 and Intel DL Boost on CPUS, along with XPU features. With the latest announcement for the gold release of Intel oneAPI toolkit, it will now be made freely available locally and on the Intel DevCloud and other commercial versions. Along with the endorsement from industry giants such as Microsoft Azure and TensorFlow, the University of Illinois Beckmann has already established a new oneAPI centre of excellence (CoE) for life sciences applications in additional computing environments. There are oneAPI CoEs in other universities such as the University of Stockholm and Heidelberg. Updation on Intel Graphics Software As per the new update, Intel’s software stack will now support multiple generations of graphics, including 11th Gen Intel Core processors with integrated Intel Iris Xe graphics and Intel Iris Xe MAX discrete graphics. Intel has enabled code reuse between operating systems by optimising the Linux driver. Further, with an increased focus on Linux 3D performance, Intel will now be delivering three fully integrated distribution-ready stacks. Intel has also announced Intel Implicit SPMD Program Compiler (ISPC) that runs on top of oneAPI Level Zero. ISPC is the overall hardware abstraction layer that promises to provide a low-level and direct-to-metal interface for devices in the oneAPI platform. ISPC is basically a variant of the C programming language. Notably, Intel, recently also introduced Project Flipfast for improved Linux gaming experience by allowing the end-users to run graphical applications in virtual machines without hampering native GPU performance and ensuring full host integration with no sharing between the virtual machine and the host. The advantage of Project Flipfast is that it increases gaming performance drastically.","excerpt":"Chipmaker Intel unveiled its first discrete GPU for data centres. Named Intel Server GPU, it is designed to power Android-based cloud gaming and media streaming segments. A low-power chip, Intel Server GPU will be used in the new PCIe graphics card which can be placed inside the servers. Interestingly, Raja Koduri, Intel Senior Vice President […]","categories":["Global Tech"],"tags":["GPU","Intel","oneAPI"],"author_name":"Shraddha Goled","publish_date":"2020-11-15T18:00:00","publication_year":"2020","word_count":757,"keywords":["Go","API","AI","R","RPA","Scala","Aim","JAX","TensorFlow","Azure","oneAPI","Intel","GPU"],"extracted_tech_keywords":["AI","Aim","TensorFlow","JAX","Azure","R","Go","Scala","API","RPA"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/intel-rivals-nvidia-amd-as-it-introduces-its-first-gpu-for-data-centers\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":58161,"title":"Is India Ready To Tackle Cybersecurity Issues With Crypto Assets?","content":"It is expected that the crypto industry in India will generate demand for more cybersecurity professionals, cybersecurity software, cryptographic experts, etc. This is because the ecosystem around cryptocurrency is incredibly complex for all users, be it end consumers or exchanges which actively manage hot wallet security and databases. Last year in April, RBI had published a notification stopping banks in India from providing financial services to crypto exchanges. Considering cryptocurrencies could be misused in money laundering and other illicit activities, the Indian government authorities had announced banning it in the past. However, with the recent developments, the drafted bill banning cryptocurrencies has been quashed by the Supreme Court. This means that India can see its crypto ecosystem finally flourishing. But, with great power comes great responsibility. So, the big question stays — is India ready to adopt crypto at a massive scale, particularly in the context of cybersecurity of digital assets? India Needs Security Infrastructure For Crypto Assets Even though government institutions across the globe, including India, have come to a common consensus on the legalised adoption of cryptocurrency, there is much more to be done in cybersecurity. Cryptocurrency is at the very core of cyberattacks, and we know that from hundreds of attacks that took place just in 2019. We all remember Mt. Gox exchange hack in which approximately 850,000 Bitcoins owned by customers and the company were stolen, an amount valued at more than $450 million at the time. Other hacks such as Japanese NEM incident, the DAO hack, and many more events like this have cost people hundreds of millions of dollars. Even in India, cybersecurity attacks were executed, and Bitcoins (worth $3.5 million at the time) were stolen on CoinSecure exchange in 2018. In fact, over 4 billion dollars were stolen in crypto funds for 2019. Crypto is also the target of ransomware and crypto-mining attacks performed by malicious hackers worldwide, even state-sponsored ones like North Korea and Iran. What’s more, cryptocurrencies are now frequently featured as preferred forms of exchange in ransomware attacks. Crypto In India: Will The Nation Balance Cybersecurity Issues With Innovation? The officials of the Indian government and RBI said they liked blockchain technology but expressed their hate towards cryptocurrency. On the other hand, those closely linked with cryptocurrency firms say crypto-tokens make an integral part of most innovative blockchain protocols, without which the network may not work as efficiently as intended, for purposes such as game theory and security incentives. So, one can make an argument that money laundering was not the only concern the Indian government had when it came to cryptocurrency. The lack of proper cybersecurity standards and practices has also been a big reason behind the government’s hesitance for legalising crypto trading and exchange. According to experts, while cryptocurrency is a rapidly growing market, it is also flooded with scammers and cybercriminals. Read More: How Lifting Crypto Ban In India Will Accelerate Jobs And Blockchain Startups But crypto startups are the hotbed of innovation in blockchain, and the public blockchain space has attracted the most innovative developers, as evidenced by the ICO boom. Blockchain applications, smart contracts and faster-secure payments are few of the features that Initial Coin Offerings (ICOs) brought with their innovative projects, pushing tens of billions of dollars of funding in 2017 to 2019. It is this innovation which regulators have also recognised. In fact, during the US Senate hearing, the chairman of Securities and Exchange Commission (SEC), and chairman of Commodity Futures Trading Commission (CFTC), in their testimonies, decided on a stance of not causing harm to the digital assets market. While at the same time, he did mention that the blockchain ecosystem is riddled with cybersecurity issues and scams. Read More: How Ethereum Blockchain Is Driving Innovation In Decentralised (DeFi) But, as investor Naval Ravikant once said India could not lose out on creating a next-generation financial exchange if it does not bank on crypto. The need of the hour, therefore, is to have the ideal cybersecurity framework in India as part of crypto trading, exchange, taxation, custody and insurance. All of that requires rigorous regulatory work and support from the government, not merely private companies and startups. According to a renowned cyber lawyer, Pavan Duggal, cryptocurrency is nothing but electronic data. And since it is electronic data, it must be treated as financial information which needs proper security like any other software and digital asset. Final Thoughts Crypto banking legalisation is undoubtedly a milestone event not just for traders, but for the cybersecurity industry in general. Crypto tokens function as a medium of exchange utilising cryptography techniques to secure transactions. It is expected that the crypto industry in India will generate demand for more cybersecurity professionals, innovative software, cryptographic experts, etc. This is because the ecosystem around cryptocurrency is incredibly complex for all users, be it end consumers or exchanges which have to manage hot wallet security. The good news is that the use of cryptocurrency generally teaches users a lot about data security. Users become familiar with concepts like public-key cryptography system, hardware wallet security, digital token security, hash functions, APIs, and databases — all of which are fundamental to a good security stance for Indian internet ecosystem.","excerpt":"It is expected that the crypto industry in India will generate demand for more cybersecurity professionals, cybersecurity software, cryptographic experts, etc. This is because the ecosystem around cryptocurrency is incredibly complex for all users, be it end consumers or exchanges which actively manage hot wallet security and databases. Last year in April, RBI had published […]","categories":["AI Trends"],"tags":["crypto","crypto india","Cryptocurrency","Cybersecurity","Cybersecurity India","game theory"],"author_name":"Vishal Chawla","publish_date":"2020-03-06T11:54:00","publication_year":"2020","word_count":867,"keywords":["Go","API","funding","programming_languages:R","Cryptocurrency","AI","crypto india","innovation","Git","game theory","crypto","Cybersecurity India","ViT","Cybersecurity","R","startup"],"extracted_tech_keywords":["AI","R","Go","Git","API","ViT","innovation","startup","funding","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/is-india-ready-to-tackle-the-cybersecurity-issues-with-crypto-assets\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10115384,"title":"Why Are Data Centres in India Moving to Tier 2, 3 and 4 Cities?","content":"Of late, several giants, including AdaniConnex, Reliance, Sify, Atlassian, Yotta, AWS, and Lenovo, have announced substantial investments in data centres across India. With the growth of India’s data centre market, the capacity is projected to surpass 1,300 MW by the end of 2024, a notable increase from 1,048 MW at the end of 2023 and 880 MW as of June 2023. However, the concentration of data centres remains prominent in seven cities like Mumbai-Navi Mumbai, Chennai, Delhi-NCR, Bengaluru, Pune, Hyderabad, and Kolkata. These cities accounted for a 35% year-on-year growth, reaching 884 MW capacity and spanning over 13 million sqft by the end of 2023. Mumbai-Navi Mumbai leads with a 52% share of the total data centre capacity, followed by Chennai (16%), Delhi-NCR (11%), Bengaluru (9%), Pune (7%), Hyderabad (4%), and Kolkata (1%). Despite the concentration in the top-four locations, the emergence of co-location and edge computing facilities is expected to shift the dynamics, with edge data centres expanding into Tier 2 cities in 2024. Overall, the data centre occupancy level in India stood at about 75-80% in 2023 and is likely to improve by the end of 2024. This expansion is driven by the need for edge computing and the desire to be closer to customers in these areas, offering a faster response time and lower latency for time-sensitive applications. Who’s Moving and Where? Nxtra and CtrlS are among the leading data centre players actively expanding their edge computing infrastructure in Tier 2 and 3 cities to meet the burgeoning demand. Nxtra, Airtel’s data centre arm, operates 120 edge data centres in 65 cities, like Bhopal, Nagpur, Cuttack, Ranchi, Patna, Raipur, Kanpur, Madurai, and Leh. These facilities have capacities ranging from 1 to 5 megawatts, and the company plans to expand its edge data centre infrastructure further. CtrlS, on the other hand, envisions creating over 20 edge data centres across Tier 2 and Tier 3 cities in the coming years. It currently has edge data centres in Lucknow and Patna. It further plans to expand its capacity this year and open a new facility in Kolkata. Yotta Data Services, backed by the Hiranandani Group, is expanding its operations in Greater Noida and Guwahati, tapping into the burgeoning demand for edge facilities in Tier 2 markets. These operations are expected to be completed and operational by the end of 2024. With Yotta-D1 nearing full capacity at 85%, the company is moving forward with the construction of its D2 and D3 facilities STT GDC India, a subsidiary of Singapore’s STT GDC, has announced its plans to invest $1 billion in the next 3-4 years to expand its data centre footprint across India, including Tier 2 cities. The company aims to build at least two data centres in Tier 2 cities this year and about eight in the next three years. These expansions are part of a broader trend towards decentralising data centre infrastructure to better serve a distributed user base across India. The move towards Tier 2, 3, and 4 cities is expected to support the low-latency requirements of 5G and support AI applications that require real-time processing and analysis for faster decision-making. Benefits of This Move Building data centres in Tier 2 and Tier 3 cities in India offers several advantages over their establishment in larger cities. Positioning data centres closer to users in these cities can significantly reduce latency, enhancing the digital experience for services that demand quick data processing. These could be for online gaming, streaming, and financial transactions. This proximity is crucial for the emerging needs driven by 5G, OTT streaming, online gaming, and AI technologies. Secondly, Tier 2 and Tier 3 cities may provide more opportunities for constructing green data centres. More available space and less stringent regulations can facilitate the adoption of sustainable practices, including using renewable energy sources and advanced cooling technologies. This aligns with the broader industry trend towards sustainability. Furthermore, these cities represent untapped markets with considerable growth potential. As the digital economy expands, the demand for digital services in these areas increases, offering data centre operators new markets to explore and support regional development. This expansion can also drive improvements in connectivity infrastructure and lead to the creation of modern office spaces and tech parks, attracting more business and talent to these regions. Additionally, both central and state governments in India have recognised the potential of the data centre industry and have introduced policies and incentives to support its growth. These include draft data centre policies, infrastructure status for the industry, and the inclusion of data centres in the harmonised list of infrastructure. States like Karnataka, Tamil Nadu, Uttar Pradesh, Odisha, Telangana, and West Bengal have announced data centre policies to facilitate investments in the last two years. Conclusively, developing data centres in Tier 2 and Tier 3 cities contributes to more balanced regional development, creates job opportunities, and supports government initiatives to digitise the nation’s economy. It also meets the country’s growing digital needs by decentralising the data centre industry to serve a distributed user base better. Delayed, But Rewarding In conversation with AIM, industry luminaries like Saandeep Dandekarr and Rachit Mohan, emphasised that while there is a definite trend towards this shift, it will take time to mature, estimating about two to five years for a significant footprint in Tier 2 and 3 cities. “It will take time to mature, but that’s the way forward for sure,” Dandekarr said, while Mohan was of the opinion that the move, “Will take a couple of years to mature.” This timeline, he said, accounts for the various stages involved, from acquiring land to designing, constructing, and commissioning data centres, considering the current infrastructure penetration in these areas. Dandekarr highlighted that cities like Hyderabad and Pune, previously considered Tier 2, are already witnessing the emergence of data centres, driven by universal demand for services such as OTT and gaming. He also reiterated the role of incentives provided to the data centre industry and other sectors, in determining the location of these new facilities. The move would also add a bunch of jobs, which Dandekarr categorised as the construction phase, which would see a temporary spike in employment, and the operational phase, promising more sustainable job opportunities, albeit with a smaller workforce due to advancements in automation and AI. “So, unlike say in the ITES industry, in a three lakh square feet data centre, instead of 4,000 people, we will have maybe 200 to 250 people actually in the operations,” he elaborated. He also discussed the impact of AI on automating processes within data centres, moving from a reactive to a predictive approach to enhance efficiency and reduce the workforce. “So, right from security access, to let’s say, temperature and battery maintenance, everything is becoming predictive from reaction or response base to predictive, right? And that is where AI is coming,” he mentioned Conclusively, he opined that there will also be indirect job creation through the vendor ecosystem required for maintenance and service support, indicating a broader impact on employment in the regions hosting new data centres. View all stories Top Earthquake Prediction Tools To Watch in 2025 QS World University Rankings 2026: Top 10 Indian Universities Meet Dr. M C Mary Kom at The Rising 2025 – Conference | Bengaluru Top 10 Indian Companies with Highest Women Workforce 2025 DeepSeek Vs ChatGPT – Why World is Falling for this Chinese AI Model","excerpt":"However, industry luminaries like Rachit Mohan and Saandeep Dandekarr believe that it will take 2 to 5 years to mature.","categories":["AI Features"],"tags":[],"author_name":"Shyam Nandan Upadhyay","publish_date":"2024-03-12T13:55:15","publication_year":"2024","word_count":1229,"keywords":["Go","ChatGPT","AWS","AI","Git","GAN","GPT","Aim","edge computing","R"],"extracted_tech_keywords":["AI","ChatGPT","Aim","AWS","edge computing","R","Go","Git","GPT","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/why-are-data-centres-in-india-moving-to-tier-2-3-and-4-cities\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10047939,"title":"Interesting But Underrated ML Concepts #3 &#8211; Q-learning, PVLV, OOB &#038; Rprop","content":"There are some fascinating machine learning topics that aren’t discussed nearly as frequently as they should be. We’ll look at some of them in this post, including Q-learning, Primary Value Learned Value (PVLV), Out-of-bag error, and Rprop. Q-learning Q-learning is a model-free reinforcement learning algorithm for determining the worth of a certain action in a given state. It can handle problems with stochastic transitions and rewards without requiring adaptations and does not require a model of the environment (thus “model-free”). According to Francisco S. Melo, Q-learning provides an optimal policy for any Finite Markov Decision Process (FMDP) by maximising the anticipated value of the total reward across any and all successive steps, beginning from the present state. Given infinite exploration time and a partly random policy, Q-learning can determine an optimal action-selection policy for any given FMDP. Researchers say that quality is represented by the letter ‘Q’ in Q-learning. In this situation, quality refers to how valuable a specific activity is in obtaining a future reward. As per Chris Gaskett, only discrete action and state spaces are covered by the typical Q-learning technique (which uses a Q table). Because of the curse of dimensionality, discretisation of these values leads to ineffective learning. However, several Q-learning adaptations, such as Wire-fitted Neural Network Q-Learning, seek to alleviate this problem. Primary Value Learned Value (PVLV) The reward-predictive firing features of dopamine (DA) neurons may be explained by the Primary Value Learnt Value (PVLV) model. As per Randall C. O’Reilly, it simulates Pavlovian conditioning and the firing of midbrain dopaminergic neurons in response to surprise rewards using behavioural and neurological data. It’s a different approach than the temporal-differences (TD) method and is a component of Leabra. PVLV is divided into two parts: primary value (PV) and learning value (LV). Primary reward (i.e., an unconditioned stimulus; US) engages the PV system, which learns to anticipate the occurrence of a certain US, preventing the dopamine rush that would otherwise occur. The LV system learns about conditioned stimuli that are consistently associated with primary rewards and fires phasic dopamine bursts at the commencement of CS (conditioned stimulus). As a result, the PVLV mechanism serves as a vital link between the more abstract TD model and the underlying neuronal systems’ specifics. According to O’Reilly, the PVLV model can account for key components of the DA firing data and make a number of clear predictions about lesion effects, many of which are compatible with existing data. Out-of-bag error Out-of-bag (OOB) error, also known as the out-of-bag estimate, is a technique for calculating the prediction error of random forests, boosted decision trees, and other machine learning models using bootstrap aggregation (bagging). Bagging creates training samples for the model to learn from by using subsampling with replacement. By evaluating predictions on observations that were not utilised in constructing the next base learner, bootstrap aggregating allows one to define an out-of-bag estimate of prediction performance improvement. Experts say that the random forest model is validated using the out of bag (OOB) score. As a result of a study by Silke Janitza and Roman Hornung, the out-of-bag error has been shown to overestimate in settings with an equal number of observations from all response classes (balanced samples), small sample sizes, a large number of predictor variables, a small correlation between predictors, and an unbalanced sample. According to the researchers, the advantages of OOB are no data leakage, less variance, a better predictive model, and less computation. Rprop As stated by James McCaffrey, Rprop (resilient back propagation) is a neural network training procedure that is similar to regular back propagation. As per M. Riedmiller, Rprop is a common gradient descent technique that computes updates based only on gradients’ signs. It stands for resilient propagation and is useful in a variety of circumstances since it dynamically adjusts the step size for each weight separately. Rprop, on the other hand, has two important advantages over back propagation: for starters, Rprop training is generally faster than back propagation training. Second, unlike back propagation, which requires values for the learning rate, Rprop does not require to be supplied any free parameter values (and usually an optional momentum term).","excerpt":"In this series, we’ll look at several underappreciated yet fascinating machine learning concepts.","categories":["AI Trends"],"tags":["back propagation","Machine Learning","reinforcement learning an introduction"],"author_name":"Dr. Nivash Jeevanandam","publish_date":"2021-09-09T14:00:00","publication_year":"2021","word_count":690,"keywords":["Go","back propagation","machine learning","programming_languages:R","AI","neural network","Machine Learning","programming_languages:Go","ViT","R","reinforcement learning an introduction"],"extracted_tech_keywords":["AI","machine learning","neural network","R","Go","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/ml-concepts\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":31966,"title":"Automated Stores Have Arrived In India, But Will They Cause Retail Disruption?","content":"Autonomous stores have already become a reality in most part of the world. An autonomous store is nothing but a cashier-less store where the customer just has to walk into a store, pick a product, and pay for it through their smartphones. From the moment a customer walks into the store, to when he leaves, there would be absolutely no human to help the person. Touted as the future the retail marketing, the technology puts the customer in a data space where the cameras within the store track the products that you take from each shelf. The unique QR code lets the customer pay the amount through their mobile phone. The technology basically makes sales persons redundant and combines the best practices of an online and offline store for enhancing customer satisfaction. Due to its highly disruptive nature, e-commerce and retail giants have already unveiled their autonomous stores across the globe. We look at some of the big names in the industry and how they differ from each other. Amazon Go: Introduced in 2018, Amazon’s Just Walk Out Technology enables customers to walk into Amazon Go stores and walk out with a product, without any salesperson bothering the customers. The store is equipped with computer vision, sensors and relies on deep learning technique for facilitating the shopping. The weight sensors on the product shelves monitor the movement of the product and a dozen or so cameras detects the customers’ movement. Though it has been introduced as a pilot project in the US, the company hopes to add 3000 automated stores within the next three years. The company is presently at the process of strengthening its technology for a robust automated store. For this, Amazon is collating data about the customer preferences including what they stopped to look at and what they didn’t even buy. The company hopes to obtain customers’ nuanced shopping behaviour through its AI-driven customer insight systems. Watasale: This unique initiative can be seen as India’s answer to Amazon Go, thus making it the first ever in the country. Based out of Kochi, the store functions along the same lines of Amazon Go. through their store, Watasale presently sells products like soda and chips and other smaller goods that their customers can easily pick and carry. It took the makers of the store three years to complete the project and the team is closely collecting customers’ data for fixing the initial glitches. Unlike Amazon Go, the team is particular about not intruding into their customers’ privacy. Speaking to a leading national daily, Rajesh Malamal, the chief marketing operator of Watasale said, “We are only looking at big data to get analytics about the products being bought to get information about inventory. We are not venturing into anything risky that will compromise customer privacy.” The company also hopes to get enough funding in the future for expanding Watasale into a chain store. Walmart: While all the major retail and e-commerce players are testing the waters with the automated store, Walmart couldn’t afford to miss out on the tech. So in April 2018, Walmart introduced the Check Out With Me feature which has been introduced in its Lawn & Garden Centers which sells plants and other agriculture-based products. Owing to the difficulty of handling large purchases for both the staff and customers, Walmart has equipped its staff with a Bluetooth device which can accept card payments from the shopping aisle itself. This, the retailers say, will save the customers from the trouble of pushing heavy carts around the store. Though it doesn’t completely negate sales-persons, it does reduces their role substantially. Tesco: The leading British multinational general merchandise retailer has been in the news lately for venturing into a cashier-free store. In June this year, the company’s new autonomous feature called Scan Pay Go was used by its 100 staffers in a trial run. Tesco claims that their new feature can reduce the average shopping time of a customer by 45 seconds. The company has also gone out of its way to improve its customer satisfaction, by using deep-learning technology each Tesco’s automated store plays background music according to the age profile of its outlet shoppers.","excerpt":"Autonomous stores have already become a reality in most part of the world. An autonomous store is nothing but a cashier-less store where the customer just has to walk into a store, pick a product, and pay for it through their smartphones. From the moment a customer walks into the store, to when he leaves, […]","categories":["AI Features"],"tags":[],"author_name":"Akshaya Asokan","publish_date":"2018-12-21T06:21:04","publication_year":"2018","word_count":696,"keywords":["big data","Go","funding","AI","computer vision","RAG","Aim","deep learning","analytics","R"],"extracted_tech_keywords":["AI","deep learning","computer vision","analytics","Aim","RAG","R","Go","big data","funding"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/automated-stores-have-arrived-in-india-but-will-they-cause-retail-disruption\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10139323,"title":"Anthropic Builds ‘FSD’ for Computers, Beats Microsoft at the Game","content":"OpenAI’s rival Anthropic, the AI safety and research company, is changing the way humans interact with computers. The company recently unveiled an upgraded Claude 3.5 Sonnet model, the new Claude 3.5 Haiku, and a public beta for an experimental feature called ‘computer use.’ Developers can now use it to automate repetitive tasks, conduct testing and QA, and conduct open-ended research. Many have likened it to a car’s self-driving features. “Entering a new era with computer use. It’s like FSD for your computer!” said Sunny Madra of Groq. “Replit Agent just becomes an L3 self-driving Agent with Claude computer use from Anthropic,” posted a user on X. In a demo video presented by Anthropic, Claude was able to perform web searches, book tickets, operate MS Excel and fill out applications. “In the first trial, the ‘computer use’ feature was able to find and report flight options from SFO to Tokyo,” shared a Replit employee. Abacus AI chief Bindu Reddy noted that the ‘computer use’ API by Anthropic presented an interesting take on agentic APIs. “Agents are challenging because they have to talk to other systems, and most of these systems don’t have good APIs,” she said. Bindu further added that one potential solution is to use the ‘Computer Use API,’ allowing the LLM to simulate a human operating a computer. “In the long run, systems will talk to other systems, and we won’t need this type of API,” she said. Early adopters like Asana, Canva, Replit, and The Browser Company have already begun exploring its potential. Replit, for instance, is leveraging the feature to build a tool that evaluates apps during their development process. AI computer use is much bigger than it might seem. For instance, many jobs involve performing routine computer tasks, like quality assurance, tech support, data entry, and administrative assistance. All of this can be automated. Amjad Masad, Replit’s founder, opines that 10-20% of the global workforce involves routine computer tasks. Masad observes that the economic value of these jobs will probably run in trillions of dollars. “Of course, the tech is still early and will take time to mature. But it seems safe to say that by 2026, we will start to see AI’s economic impact in macro metrics,” said Masad. Do We Still Need Computers? AIM Media House chief Bhasker Gupta questioned the need for a computer feature and asked, “Will we need traditional interfaces at all? In the future, most tasks might be completed by AI through simple commands or even autonomously—without screens, no clicks, or typing.” On the other hand, Meta’s chief AI scientist, Yann LeCun, believes that in the next 10-15 years, we won’t have smartphones and will instead use augmented reality glasses and bracelets to interact with intelligent assistants. “Today, all of us here are carrying smartphones in our pockets. Ten years from now, or 15 years from now, we’re not going to have smartphones anymore, we’re going to have augmented reality glasses,” said LeCun. While Anthropic has built something that still requires a computer as an interface, it is likely that in the future we will move away from screens and interact with AI agents using a new kind of device or interface. In a recent interview, OpenAI chief Sam Altman said that by 2030, you will be able to walk up to a piece of glass and ask it to do something that previously would have taken humans months or years, and it will accomplish that task in a dynamic way or within an hour. OpenAI’s approach is quite different from Anthropic’s, as it focuses more on the voice features alongside the reasoning capabilities. “It’s interesting how OpenAI is doubling down on end-user-facing features like voice mode, while Anthropic is doubling down on engineer\/API-focused features like code generation quality and being able to remotely drive a GUI,” posted  Simon Willison, co-creator of Django on X. Will Others Follow the Suit? Anthropic’s new feature is somewhat similar to Microsoft’s Recall feature, except that it does not control one’s computer directly.  Microsoft Recall is part of the Copilot+ PC and provides users with a timeline of their previous PC activities. It allows users to search for content they have interacted with—such as files, web pages, and applications—using natural language queries. However, this feature came under scrutiny due to privacy issues. The tech giant recently announced that the ability to create autonomous agents within Copilot Studio will enter public preview in November. These agents are designed to work across various business functions, including sales, finance, and supply chain, to automate tasks and streamline operations. However, Anthropic Claude’s computer feature stands out as it doesn’t rely on multiple agents to perform different tasks, instead, a single agent effectively manages multiple tasks. For instance, Microsoft recently integrated Copilot into MS Excel, while Claude can now directly operate Excel. This calls into question the existence of Copilot. Also, OpenAI has introduced a new approach for creating and deploying multi-agent AI systems, called the Swarm framework. It simplifies the process of creating and managing multiple AI agents that can work together seamlessly to accomplish complex tasks. The company has launched a ChatGPT desktop application, but it lacks autonomous features.","excerpt":"Anthropic’s new feature is somewhat similar to Microsoft’s Recall feature, except that it does not control your computer directly.","categories":["Global Tech"],"tags":["Anthropic","Microsoft"],"author_name":"Siddharth Jindal","publish_date":"2024-10-24T17:24:02","publication_year":"2024","word_count":862,"keywords":["Anthropic","ChatGPT","OpenAI","AI","autonomous agents","ML","RAG","Aim","Claude 3.5","R","Microsoft"],"extracted_tech_keywords":["AI","ML","ChatGPT","OpenAI","Claude 3.5","Anthropic","Aim","RAG","autonomous agents","R"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/anthropic-builds-fsd-for-computers-beats-microsoft-at-the-game\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10092639,"title":"MLOps: An Investment Towards a More Secure Future","content":"MLOps has recently emerged as the missing link in the ML lifecycle. MLOps engineers ensure that the pipeline is glued together and works seamlessly. Since 2018, MLOps functions have come under the spotlight and, with time, have been cemented as a vital mechanism. We caught up with Praveen Nair, Manager at Tredence, to know what the future holds for MLOps teams, how the current fast-paced AI landscape will impact MLOps and what it takes to be an MLOps engineer. AIM: How do you think MLOps teams will be able to address issues related to ethical concerns like data privacy and security? Praveen: This is actually something that is not spoken about as often as it should be. MLOps can really help with the implementation of ethical frameworks when countries and governments come up with new data laws around security concerns. Say, a bank has to deny or approve loans based on the background, financial features and behaviour of a particular person. When models for these processes have to be trained, you have to make sure that your system is not biased towards any demographic or gender. The model is expected to be fair and banks must offer equal opportunities to all. One of the biggest solutions for this is to have more explainability in models. MLOps can help with setting up processes in place to make the model more transparent. Data scientists can also assess this with bias monitoring systems in place. There have already been open-source contributions in this regard. A bunch of cloud platforms have already implemented some of these preventive measures. However, it is the adoption part that remains slower. Now, with laws coming into the picture, companies will be forced to adopt measures sooner than later. In the context of data privacy, you wouldn’t want users’ personal details to be exposed in databases and available for people to see. To reiterate, there are a number of open-source tools available to detect if your model or related pipelines are using this type of Personal Identification Information or PII. So, having robust MLOps processes can take care of these risks. Having an MLOps team will also help address concerns around external security and adversarial attacks. For instance, if a company has a small data science team, they can focus only on the innovation part and the MLOps teams can protect the data security layer and take care of the boundaries of what you can and can’t do. AIM: Federated learning has a number of advantages, but what are the challenges around using it in MLOps? Praveen: Federated learning is something that is relatively new in terms of the number of applications that are coming out. A major part of the current MLOps processes is still not prepared enough to take care of what’s coming in the future. For example, with GPT-4, Sam Altman has said that the model works on self-supervised learning, which means that it is building a brain based on inputs from users. When a tool like this arrives on the landscape, new issues around privacy are bound to crop up as well. You wouldn’t want a user to know what other users have used the model for. This is one of the biggest challenges since there are millions of users on it. In federated learning, everyone’s data is decentralised and they can keep their data with them. The goal is to train intelligence together but your data is not going to be diverted to one central server. Rather, we will train within your device and then take that understanding back and then aggregate it elsewhere. The other challenge is that there are also stringent laws emerging in place against bringing all the data from different regions together in a central place. AIM: What are the differences and similarities between MLOps and AutoML? Praveen: One of the new things in terms of monitoring is to understand what behaviours have changed like what we saw after the COVID-19 pandemic. Certain segments like e-commerce activity have witnessed a significant burst in the aftermath of the pandemic when compared to pre-pandemic activity primarily due to a change in customer behaviour. The models that were trained then did not really work during\/after the pandemic happened. So, this shift will imply that these models will have to be retrained. To do this continuously is the tough part because customer behaviours change too often. In this scenario, AutoML is a good process to automate this portion of the workflow. AutoML can train models much faster as it is sort of abstracting the training efforts. All the data scientist has to do is make the model transparent and ensure the model is bias-free. I don’t think AutoML and MLOps are comparable but rather I’d say that AutoML is a good tool to ensure the implementation of MLOps adoption across organisations. The challenge is in getting state-of-the-art models using AutoML but you still get a baseline model churned out much faster. AIM: How will the emergence of generative AI further push the implementation of MLOps? Praveen: To build a GPT model is not easy because they are extremely complex, uses a lot of computing power and takes a lot of time. Now, since people have begun to find business value in it, GPT-4 will definitely be used across industries. However, small misinformation can lead to a large negative impact, especially in data-sensitive fields like medicine. MLOps can set standard processes for security in this regard. Even when such massive LLMs are being trained, we need huge quantities of quality data which MLOps & Data Engineering disciplines can ensure. You wouldn’t want your generative AI model to be political or inappropriate. So, there are a number of valid concerns here because generative models usually use large amounts of data and the responses need to be validated continuously to look for improvements or accuracy. Generative AI will definitely push the implementation of MLOps and introduce different types of MLOps processes in the future. There are already Large Language MLOps, which is becoming a thing and that’s great because ethical concerns and necessary guardrails can then be addressed more quickly. AIM: What are the main skills that prove essential to become an MLOps engineer? Praveen: The biggest misconception about MLOps is that it is merely the deployment of ML models. While this is partially true, the breadth of what an MLOps engineer does includes more than model deployment. The first thing needed is soft skills since MLOps engineers have to talk to data scientists who will always defend their models or the business teams who want better ROI. MLOps managers are often stuck in the middle trying to make both sides happy. So, it’s also important to have a sense of collaboration. The second skill is proficiency in ML and deep learning. The usual idea is that we need to know DevOps, which is not true. We also need to optimise these models in different ways, like trying to understand what the algorithm is doing and if it’s been trained in the right way. This requires a deeper understanding of these algorithms themselves. Thirdly, we need to understand data and store the right kind of data. You need to be able to interpret the data you have and make it useful by coming up with more metrics around quality & accuracy. In tandem, you need to have an architectural understanding which means that you need to understand that every client requires something different. It’s not like a one-size-fits-all solution. Depending on the client’s environment, you have to determine what is the most cost-efficient and optimal approach. For this, you need to have an architectural understanding of what MLOps is trying to achieve. So, as an MLOps engineer, you need to have different variations of the same architecture ready in your pocket and that ultimately makes the biggest point of difference. The last skill would be the most obvious one, to have great programming skills.","excerpt":"“MLOps managers are often stuck in the middle trying to make both sides happy.”","categories":["AI Highlights"],"tags":["MLOps"],"author_name":"Poulomi Chatterjee","publish_date":"2023-05-15T11:00:00","publication_year":"2023","word_count":1327,"keywords":["federated learning","data science","AWS","AI","ML","MLOps","Aim","deep learning","generative AI","R"],"extracted_tech_keywords":["AI","ML","deep learning","data science","generative AI","MLOps","Aim","federated learning","AWS","R"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/mlops-an-investment-towards-a-more-secure-future\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10102901,"title":"Top 8 Free Generative AI Courses in 2024","content":"Ever since the debut of ChatGPT, there has been an emergence of several significant language models, including LLaMA, PaLM 2, GPT-4o, Alpaca, Vicuna-13B, and more. Excitingly, this trend is poised to continue expanding. To fully harness the immense potential of generative AI and create customised models, these complimentary training programs will prove invaluable. 1. Generative AI for Everyone Andrew Ng’s “Generative AI for Everyone,” is an in-depth course on generative AI. It covers its capabilities and boundaries, including practical exercises for everyday use, prompt engineering, and advanced AI techniques. The course focuses on real-world applications, showing generative AI’s common uses and allowing hands-on experience. It also discusses AI’s effect on business and society, preparing students to develop effective AI strategies and understand its real-world implications. 2. Introduction to LLMs Google, which is currently building the next LLM Gemini, also provides a series of free courses on master generative AI. In this course, Google provides an overview of LLMs, and their definitions and explains potential applications. It also delves into the concept of prompt engineering, which can improve the performance of LLMs. Additionally, the module introduces various Google tools that can assist in the development of personalised generative AI applications. 3. LLMs Application through Production This course by Databricks is designed for individuals with an intermediate-level proficiency in Python and a working understanding of machine learning and deep learning. It focuses on the practical application of LLMs through various frameworks. Participants will learn to build LLM-focused applications using popular libraries like Hugging Face and LangChain. The curriculum covers key concepts, including the distinctions between pre-training, fine-tuning, and prompt engineering. Industry experts, including Matei Zaharia (CTO Databricks and Associate Professor of CS at Stanford) and Harrison Chase (Co-Founder and CEO of LangChain), provide insightful lectures. By the course’s conclusion, participants will have constructed an end-to-end LLM workflow, prepared for production deployment. This course is available for free audit, which means that you can access the course materials without paying any fee. However, if you want to access a managed compute environment for course labs, graded exercises, and a certificate, you need to pay a nominal fee. 4. Introduction to AI with Python The introductory free online course “CS50’s Introduction to AI with Python” offered by Harvard School of Engineering and Applied Sciences covers fundamental concepts and algorithms in AI and ML using Python. It spans seven weeks, with a flexible time commitment of 10-30 hours per week and is self-paced. Participants will explore various topics, including graph search algorithms, reinforcement learning, machine learning, and principles of artificial intelligence. The instructors, professors David J. Malan and Brian Yu, both from Harvard University, guide students through hands-on projects to reinforce theoretical knowledge. 5. ChatGPT Prompt Engineering for Developers Again created by DeepLearning.AI, this free course is in partnership with OpenAI and is taught by Isa Fulford (OpenAI) and Andrew Ng. It covers the essentials of prompt engineering for developers, from beginner to advanced levels. Participants will learn to effectively use LLMs, specifically using the OpenAI API, enabling them to swiftly develop innovative applications. The course explains LLM functionality, offers prompt engineering best practices, and demonstrates the practical application of LLM APIs in various tasks such as summarising, inferring, transforming text, and expanding content. It also imparts two key principles for crafting effective prompts, systematic prompt engineering, and the creation of custom chatbots. The content is beginner-friendly, requiring only basic Python knowledge, while also catering to advanced machine learning engineers seeking cutting-edge insights into prompt engineering and LLM usage. 6. Career Essentials in Generative AI Microsoft and LinkedIn have come up with a free course titled “Career Essentials in Generative AI,” which serves as an introductory resource on generative AI and its practical applications. The course aims to furnish learners with fundamental skills essential for success in the generative AI domain. It covers key areas such as an overview of AI tools, an exploration of generative AI models and their functioning, differentiation between search engines and reasoning engines within the context of generative AI, using Microsoft Bing Chat for efficient work processes, a preview of Microsoft 365 Copilot, and a segment on the ethical considerations inherent in the generative AI creation and deployment process. The free course is structured to provide a comprehensive understanding of the subject matter, encompassing both technical aspects and ethical dimensions. 7. ChatGPT Prompt Book The ChatGPT Prompt Book consists of over 300 unique writing prompts generated by the ChatGPT language model, designed for creative thinking and finding new ideas and perspectives. The prompts cover a diverse range of subjects and are adaptable to various writing styles and genres, making them useful for writers of all levels of experience, from beginners to professionals. 8. Generative AI Foundations on AWS The Generative AI Foundations on AWS is an eight hour course on YouTube offering practical insights and hands-on guidance for pre-training, fine-tuning, and deploying foundational models on AWS. It emphasises breaking down theory, mathematics, and abstract concepts, providing hands-on exercises to build practical intuition. The course progressively explores complex generative AI techniques, enabling participants to understand, design, and apply their models effectively. Topics include the recap of foundation models, selecting the right model for specific use cases, pre-training new models, scaling laws for the model, dataset, and compute sizes, preparing training datasets on AWS at scale, fine-tuning models, and leveraging reinforcement learning with human feedback.","excerpt":"These courses will teach you the fundamentals of LLMs and how to deploy them effectively.","categories":["AI Trends"],"tags":["Courses","Top Trend"],"author_name":"Shritama Saha","publish_date":"2023-11-10T16:10:42","publication_year":"2023","word_count":893,"keywords":["Top Trend","ChatGPT","machine learning","artificial intelligence","OpenAI","AI","GPT-4o","ML","deep learning","generative AI","foundation models","Courses"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","generative AI","foundation models","GPT-4o","ChatGPT","OpenAI"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-8-free-generative-ai-courses-of-2023\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10046639,"title":"The Rise &#038; Rise Of Postman","content":"“APIs are the building blocks of effective software – so while software might be eating the world, we know that APIs are eating software. Innovation in APIs will drive the future of software development,” said Abhinav Asthana in 2019 when his company raised $50 million in Series B funding. Two years and two more funding rounds later, Postman has emerged as the most valued Indian SaaS startup. In its latest Series D funding round, Postman raised $225 million, taking its overall valuation to $5.6 billion. The financing round was led by New York-based Insight Partners, with participation from other investors, including Nexus Venture Partners, CRV, BOND, Battery Ventures, etc. So what explains Postman’s increasing popularity? Early Days An application programming interface (API) is a code that allows interoperability between apps. Postman offers a platform for developing these APIs. Many companies build them internally. However, Postman’s collaborative feature makes it a model system for establishing a growing and connected developer ecosystem. In an interview, Asthana said that creating a platform like Postman was rooted in his experience as an intern developer at Yahoo in 2012. Along with future co-founder Ankit Sobit, he worked on APIs to convert them into a shareable format for developers to use in their projects. Asthana says that it was extremely challenging to learn how APIs worked and also very difficult for the development team to test and debug them. Any update in the API meant that the whole work had to be done from scratch. The tooling at that time was very primitive, which did not help optimise the team’s workflow. Asthana encountered a similar problem when he started his first startup TeliportMe. This company was building a complex API for sharing information throughout the development process. The tool for API management was still very basic, and there was no commercially available solution then. Interestingly, Postman started as a side-project with Asthana, writing the early version as a basic HTTP client for Chrome. Some of his co-workers began using it too, and soon, Google reached out to the team appreciating the tool. The Chrome store’s homepage also featured it, giving it more visibility. In a short time, the number of users grew to 500,000. That is when Asthana quit TeliportMe to start Postman as a full-time venture. Growing Popularity Building APIs is a team effort and calls for seamless workflows between developers. The fundamentals of API design have been established, but there is still room for improvement and innovation; this is where Postman steps in. Postman credits its popularity to its simple UI that allows developers to manage every stage of the API lifecycle. Some of the features include: It offers an API-first development platform for users to develop reliable API before deploying code.It eliminates dependencies and reduces time to production by facilitating the front-end and the back-end teams working simultaneously.Postman automates manual testing and integrates them directly into the CI\/CD pipeline. This feature ensures that code changes don’t break API in the production.It helps in publishing API documentation to help both internal and external consumers adopt a particular API. Earlier this month, Postman announced that its public API network is the ‘largest API Hub in the world’. Postman Public API Network is a global directory of several public APIs that connect developers worldwide and provide a central catalogue of APIs built for discovery, exploration, and sharing. It has 17 million users and 500,000 organisations worldwide. Further, the company says that the number of users joining the network, publishing and consuming workspaces and APIs, and forking collections into their workspaces has been increasing exponentially. This rise has been steered by public workspaces, which were made available earlier this year. These public workspaces make it possible for API producers and consumers to communicate and collaborate in real-time, irrespective of the team or organisational boundaries. Looking Ahead Some of the upcoming products and services on Postman’s roadmap include a new version of Postman API. The company will also soon offer support for protocols such as ProtoBuf, gRPC and increased capabilities for GraphQL. Postman is also focusing on integration with vendors like AWS, GitHub, and GitLab. Further, a new no-code API composition tool, Flow Runner, is also in the pipeline, enabling anyone to build API-driven programs.","excerpt":"Earlier this month, Postman announced that its public API network is the ‘largest API Hub in the world’.","categories":["AI Features"],"tags":["API","Funding","Postman","postman api"],"author_name":"Shraddha Goled","publish_date":"2021-08-24T13:00:00","publication_year":"2021","word_count":707,"keywords":["Go","API","Funding","AWS","AI","postman api","ML","CI\/CD","Git","GitLab","Postman","GitHub","R"],"extracted_tech_keywords":["AI","ML","AWS","R","Go","Git","GitHub","GitLab","CI\/CD","API"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/the-rise-rise-of-postman\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10102178,"title":"EXL&#8217;s Data-Driven Strategy Leads to Consistent Growth","content":"ExlService Holdings, Inc., a leading data analytics and digital solutions company, announced its financial results for the quarter ending on September 30, 2023. The company reported a robust performance with third-quarter revenue reaching $411.0 million, showcasing a remarkable year-over-year increase of 13.7%. This growth is a testament to the company’s resilience and strategic prowess, particularly in unpredictable business environments. The Chief Executive Officer of ExlService Holdings, Rohit Kapoor, expressed his satisfaction, stating, “We achieved another robust quarter, with year-over-year revenue growth of 13.7% and adjusted diluted EPS growth of 21.3%. Our data-led strategy and balanced portfolio of businesses, bolstered by our unique digital\/AI capabilities, position us well to deliver superior growth in an unpredictable environment.” EXL’s strategic focus on AI is at the core of its consistent growth. Kapoor also highlighted the pivotal role of data, digital, and AI capabilities in their strong third-quarter performance. They’ve invested significantly in these technologies, positioning them as a market leader. Their emphasis on integrated digital operations, driven by clients seeking cost-efficiency and productivity, has resulted in a robust pipeline of large deals exceeding $50 million in total value, demonstrating the value of their end-to-end solutions. EXL is also capitalizing on generative AI, engaging with clients on over 200 use cases, allowing them to venture into new market areas. Their commitment to innovation and plans for an international operations headquarters in Dublin exemplify their dedication to advancing AI and expanding their global footprint. These initiatives, coupled with their strong digital capabilities, make EXL a key player in the industry, poised for continued growth. ExlService Holdings also benefited from the strong leadership of its CFO, Maurizio Nicolelli, who commented, “While we remain prudent in our outlook given the current uncertain environment, we are increasing our revenue and EPS guidance for the full year 2023 based on our strong momentum year to date and current visibility for the remainder of the year.” The updated revenue guidance now places the expected revenue range between $1.620 billion to $1.628 billion, up from the previous guidance of $1.605 billion to $1.625 billion. This adjusted revenue guidance represents a remarkable 15% year-over-year growth on a reported basis and 15% to 16% growth on a constant currency basis. Additionally, the company has increased its adjusted diluted earnings per share guidance for 2023 to a range of $1.40 to $1.42, which signifies growth of 16% to 18% over the prior year. ExlService Holdings, Inc. reported revenue growth across all its segments, including insurance, healthcare, emerging business, and analytics. The company’s robust operating income margin in the third quarter, coupled with its ability to effectively manage costs, contributed to its impressive performance.","excerpt":"Effective cost management boosts impressive Q3 performance with updated projections indicating 15% YoY Growth.","categories":["AI News"],"tags":["Data Analytics","financial results","Generative AI","Revenue Growth"],"author_name":"Shyam Nandan Upadhyay","publish_date":"2023-10-27T18:07:21","publication_year":"2023","word_count":439,"keywords":["API","programming_languages:R","AI","innovation","Git","ViT","analytics","generative AI","financial results","Data Analytics","Generative AI","R","Revenue Growth"],"extracted_tech_keywords":["AI","analytics","generative AI","R","Git","API","ViT","innovation","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/exls-data-driven-strategy-leads-to-consistent-growth\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":5734,"title":"Interview &#8211; Lakshmanan Narayan, CEO at Unmetric","content":"[dropcap size=”2″]AIM[\/dropcap]Analytics India Magazine: What are some of the main tenets (philosophies, goals, attributes) of analytics approach and policies at your organization? [dropcap size=”2″]LN[\/dropcap]Lakshmanan Narayan: We believe that social media content and campaign analysis requires more than just the best algorithms and that some complex pattern recognition is best accomplished through human cognition. Therefore, our approach is to combine the power of people and technology to provide competitive intelligence to businesses that impact decision making across marketing, customer service, R&D, operations, and human resources. AIM: Can you brief about some of the services you provide? LN: Unmetric provides an online platform for Fortune 500 companies, agencies and other large global brands to understand the competitive landscape, uncover insights based on data and unlock new social strategies. We pull data from the popular social networks, filter it and apply algorithms paired with insights from a team of data analysts to provide clients with intelligence that explores deeper, more meaningful information about their social media efforts. AIM: What are the key differentiators in your analytical solutions? LN: Where many social media analytics companies rely solely on computational algorithms, Unmetric has a team of data analysts to tease out insights in a way only humans can. This provides more context around the data that is being provided to clients. AIM: Please brief us about the size of your organization and what is hierarchal alignment, both depth and breadth. LN: Unmetric is a rapidly growing company with more than 70 employees including development, data analysts, operations, marketing and sales teams. As a company, we foster an environment that facilitates fast decision making and open communication. Because responding to clients’ needs is our number one priority, we’ve retained a fairly flat hierarchy across departments to maximize agility and make it easy for our teams to serve our clients. AIM: What are the planned next steps\/ road ahead for your organization? LN: The way in which companies use social media continues to evolve. Brands today are looking for new ways to leverage social media to gain actual competitive advantages in the marketplace and drive business results. Our intent is to keep expanding and improving our platform to meet these changing needs. One of our near term plans is to expand our development and analytics teams to provide clients with even deeper insights into the unstructured data from social networks. AIM: What are the most significant challenges you face being in the forefront of the analytics space? LN: When you have so much data coming from social media, providing meaningful analytics to businesses that actually help them with their business operations becomes a challenge. There is a tendency to try to provide charts and other visuals on every single data point, which often clouds the picture with data overload. The challenge is to curate the data and provide it in a way that helps the businesses with metrics that matter rather than confuse them. AIM: How did you start your career in analytics? LN: Having studied engineering at IIT Madras, I always had an affinity for data and numbers, and my time at IIM Calcutta helped me realize the importance of aligning numbers with business decisions. My time in advertising, more specifically the media planning and buying industry gave me a ringside view of the importance of data and analytics in powering marketing decisions and strategies. AIM: What kind of knowledge worker do you recruit and what is the selection methodology? What skill sets do you look at while recruiting in analytics? LN: We don’t have an archetype when it comes data analysts. Someone with an engineering background can interpret data in a different way than someone with a visual communications background. It’s important for us to have a diverse mix of backgrounds to get multiple viewpoints. One of the things we do before inviting a candidate in for an interview is to have them do some tests that involve gathering, analyzing and commenting on data. It really levels the playing field and means you can tap hidden talent rather than restricting yourself to a certain profile. AIM: How do you see Analytics evolving today in the industry as a whole? What are the most important contemporary trends that you see emerging in the Analytics space across the globe? LN: The trend across all industries is moving the conversation from ‘big data’ to ‘smart data’. It’s not just the volume of data that’s important, it’s about how companies can use that data to make informed decisions that help move their enterprise forward. Also, data analytics will become embedded in every aspect of a business. Marketing has traditionally been at the forefront of data analytics, but at Unmetric, we’ve seen many departments make use of social media data to make decisions. I think that’s one key takeaway that businesses need to understand, just because one department might be responsible for generating and analyzing data, doesn’t mean it’s not useful for the whole business. AIM: Who do you think your main competitors are? LN: There are many companies in the social media analytics space today. Some focus on social media ‘listening’ and others are publishing platforms. However, unlike these other companies, Unmetric is a ‘seeing’ platform that looks at what brands are doing in social media. We’re often told by clients that nobody else comes close to offering what we do in terms of combining the quantitative power of algorithms with the qualitative insights of our team of data analysts. Unmetric allows account executives tasked with with collecting social media data to complete in minutes what would normally take days or weeks. This frees them up to focus more on creating engaging content and campaigns instead of crunching numbers. AIM: Anything else you wish to add? LN: Data analytics is a big opportunity for Indian companies, especially as businesses here become more data driven than ‘gut’ driven. However, there are still challenges in terms of adoption and a lot of the old-style thinking still remains. Western companies are much more data driven and businesses in India need to take advantage of that. Visit Unmetric here. [divider] [spoiler title=”Biography of Lakshmanan Narayan” style=”fancy” icon=”plus-circle”] Lakshmanan (Lux) Narayan is co-founder and serves as CEO of Unmetric, an online platform that provides leading brands and agencies with social media competitive intelligence. Previously, Lux co-founded and currently serves on the board of Vembu Technologies, an online data backup solution. In the non-profit sector, Lux co-founded ShareMyCake, a charitable foundation dedicated to fostering a more ‘giving’ generation of kids. Lux holds degrees from IIM, Calcutta and IIT, Madras and resides in NYC.[\/spoiler]","excerpt":"[dropcap size=”2″]AIM[\/dropcap]Analytics India Magazine: What are some of the main tenets (philosophies, goals, attributes) of analytics approach and policies at your organization? [dropcap size=”2″]LN[\/dropcap]Lakshmanan Narayan: We believe that social media content and campaign analysis requires more than just the best algorithms and that some complex pattern recognition is best accomplished through human cognition. Therefore, our […]","categories":["AI Features"],"tags":["Interviews and Discussions"],"author_name":"AIM Media House","publish_date":"2014-05-17T18:05:02","publication_year":"2014","word_count":1098,"keywords":["big data","Go","API","AI","RAG","Ray","Aim","analytics","GAN","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","analytics","Aim","Ray","RAG","R","Go","API","big data","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/interview-lakshmanan-narayan-ceo-at-unmetric\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":19872,"title":"Fast Forward Analytics 2018 &#8211; Know More, Do More","content":"With customer demands getting more sophisticated and personalized, this philosophy of perceiving customer support service as a cost centre is quickly losing its pace. Organisations that are analytics driven reap intangible benefits like – enhanced lifetime value for their customers, reduced customer churn and higher levels of customer recommendations. Additionally, through the internal employee engagement program organisations are witnessing higher employee retention and productivity. Future ready organisations are thinking beyond descriptive analytics and are yearning for prescriptive and predictive analytics to accelerate their customer experience strategies to the next level. While Descriptive analytics enables enterprises to query data integrated from multiple applications to create reports, dash­boards that firms can access via applications, the next generation prescriptive analytics is an amalgamation of descriptive analytics, artificial simulation, math to improve the impact of human decisions or an app’s algorithm. India’s telecom sector is traversing the tough road of consolidation with upbeat competition and surge in consumerization. In her interview to communications today, Aruna Sundarajan, Telecom secretary has rightly shared that “the days of regulatory arbitrage are over. Are telcos going to be mere utilities, or they can become more than that and offer real value to customers. Let’s look at cloud, data centers, artificial and virtual reality, big data analytics — all are converging.”Infact the contemporary adaptation of the word consumerization speaks volumes about how the retail customer is the real game-changer. The explosion of multi-channels’ analytics combining speech, text and social media analytics combined with digital feedback leaves ample room for a great Customer Experience showcase. In-fact the digital channels, contact centres, retail marketplaces, and the back-office are the four integral pillars of customer-experience. For India’s banking sector too the rise of digital has led to a natural demand by customers to communicate with brands via multiple channels. As showcased through Verint’s Digital Tipping research Indian customers prefer the digital route when engaging in a mid-complex\/complex customer service request. International Bank with operations in India leveraged Verint Speech Analytics & Real-time Speech Analytics to drive Sales Enhancement. With Verint Speech Analytics, the bank was able to surface positive language and best practices used by the best sales agents. Learning from Verint’s Speech Analytics was further used to provide sales guidance to the agents in real-time and as a result the sales pitch rate was doubled in a short span of 6 months. India’s leading private sector banks deployed Verint Speech Analytics to drive operational efficiency resulting in significant average handle time ( AHT) optimization driving instant ROI. As Indian banks are poised to accelerate in the fast lane, they are looking for technological advancements which can:- Fine tune their customer experience strategy there by enhancing digital experience. Leverage on speech, text analytics and Enterprise feedback for predictive customer insights and feedback management. Empower employees through a stronger “Voice of Employee” (VoE) Knowledge management to empower customers and employees. https:\/\/www.youtube.com\/watch?v=xTGq7tjtnmo Holistic analytics and enterprise feedback platforms can power India’s telecom and banking and BPO’s as frontrunners to improve their game-plan by:- Automated theme detection to identify areas of concern. Gauging customer sentiments Identifying trending words and phrases Speech Analytics through its speech engine incorporates phonetic and transcription in combination with natural language processing (NLP), to enable the analysis of calls in real time and to build a complete semantic index of the big data in recorded calls. Digital feedback using customer initiated or Enterprise Initiated feedback. Through customer engagement and workforce optimization solutions Indian enterprises can gain greater visibility by blending customer-facing and back-office operations environments. As more and more organisations are striving to map the voice of employee with the voice of the customer the year 2018 and beyond belong to- Unified onmichannel customer engagement strategy– Omni channel l\/digital customers are prone to higher attrition rates; they can start a customer service inquiry or request for a new product in one interaction and continue it in another. It is important to highlight here that customers don’t think in silos but the organisations can receive insights from the same customer in silos. Omnichannel refers to the technology and management strategy that is bringing together in a unified manner the multiple channels of customer communications that are typically deployed in the contact center. In the multichannel environment, recorded and other information resources for customer’s communication history is stored in information silos, independent of other channels. In other words, customer information gathered during a phone contact tends to reside in a separate database record of phone contacts. The same hold true for other channels as well like emails, social media posts, and so on. Changing consumer demographics and communications preferences are creating an environment in which silo’ed multichannel communications are no longer ideal for customer experience optimization. Omnichannel customer engagement ensures that the customer’s entire journey is tracked across channels in order to create a consistent, optimized experience. The organisation being at the receiving end gains Enterprise-wide, Holistic, Actionable Intelligence™ to deliver seamless customer service. The trifecta of speech, text analytics and Knowledge Management (KM) – Earlier this year Gartner introduced a “New Magic Quadrant for Workforce Engagement Management (WEM).” The voice of the employee is an integral matrix for the WEM. Knowledge management can be the necessary building block for WEM. Knowledge Management focuses on delivering the right information needed, and all the information needed, for a specific situation. Its primary goal is to provide relevant contextual content in the natural flow of conversation. The knowledge management methodology emboldens the customer engagement manager to deliver optimized customer experienceTogether, KM + Speech\/Text analytics allows enterprises to understand the entire customer journey. To start, organisations can setup speech categories to automatically tag calls that should be further reviewed due to potential knowledge issues.  Next, we need to build a common ontology so the speech call drivers align with the task-focused KM taxonomy.  This is key to being able to align the speech & text data with KM usage patterns.  Finally, organisations can create knowledge analytics that correlate KM usage with the customer outcomes (both positive and negative) that are identified in Speech & Text analytics. Speech Analytics in Indic languages– The KPMG-Google report “Indian Languages- defining India’s Internet” affirms that the Indian Internet language users are expected to account for nearly 75% of India’s Internet user base by 2021. The e-commerce disruptions have also opened newer revenue pockets in Tier3\/4 cities and rural areas. The time is ripe for organisations to invest in a comprehensive language technology which enables speech analytics in Indic languages. One of India’s leading telecom provider has have enabled speech analytics in 10 Indic languages including Hinglish words which replicate manual call categories but are generated automatically without introducing any user bias. Any system generated theme can be transformed into a call category for long-term analysis with the just a click of a mouse. Speech analytics in Indic languages empowered the Telco to:- Quickly build unbiased and the most impactful categories that reflect current Indic customers’ issues and concerns across thousands of calls, helping the Telco to differentiate for regional customers. The solution empowered the Telco to stay on top of customer perceptions and what’s being said during calls. Leap into Predictive and Prescriptive analytics– Predictive Analytics can take a wide range of previously unconnected data items from disparate data sources and nesh them together to make educated predictions about future behaviour. PA helps contact centers better understand their customers, Emotions, or how experiences make a customer feel. Data scientist use PA in many industries from improving hiring process to retention to quality monitoring to prioritizing interventions. Predictive Analytics can help make better decisions regarding customers, Employees and operational processes. India is currently amongst the top 10 big data analytics markets in the world and the industry is poised to grow eight-fold to $16 billion by 2025. It is important to build analytics led eco-system which can fuel more intelligent, Provide flexible deployment choices, faster and accurate way to act on emerging trends, issues and opportunities.","excerpt":"With customer demands getting more sophisticated and personalized, this philosophy of perceiving customer support service as a cost centre is quickly losing its pace. Organisations that are analytics driven reap intangible benefits like – enhanced lifetime value for their customers, reduced customer churn and higher levels of customer recommendations. Additionally, through the internal employee engagement […]","categories":["IT Services"],"tags":[],"author_name":"Anil Chawla","publish_date":"2017-12-19T09:35:23","publication_year":"2017","word_count":1325,"keywords":["big data","Go","AI","R","ML","Git","RAG","NLP","analytics","predictive analytics"],"extracted_tech_keywords":["AI","ML","NLP","analytics","RAG","predictive analytics","R","Go","Git","big data"],"url":"https:\/\/analyticsindiamag.com\/it-services\/fast-forward-analytics-2018-know\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":52789,"title":"How Artificial Intelligence Gave Impetus To New Social Media Innovation In 2019","content":"Increased investment in Artificial Intelligence (AI) technology, along with accelerating social media penetration, is a key driver of global AI in the social media market. According to a report, the total expenditure on AI systems is expected to reach $5.5 billion this year, which is an increase of 80 per cent from last year. Gartner also reports that social media will be a big driver of this growth. This is because social networks are a great source of varied data, and it may be used for many different purposes. One particular social media platform- Twitter led the way to help researchers derive intelligent insights with its data feeds, particularly specific hashtag related data, which can be used for various kinds of machine learning systems. For example, in 2019, UNSW Sydney’s Kirby Institute developed a new tool that harnesses AI on Twitter data for the earlier detection of acute disease events. Using anonymised and publicly available Twitter data, the tool analysed more than 3 million tweets containing keywords related to asthma such as “breath” and “coughing”. The technique combines two fields of artificial intelligence — natural language processing (NLP) and statistical time series modelling. In another instance from 2019 where Twitter data was leveraged was Advanced Symbolics Inc. (ASI), an artificial intelligence-driven market research company, announced an AI tool called “Polly”. This AI tool was demonstrated to be more accurate than other polling companies when it came to predicting the U.K. General Election result and the large majority of the Conservative and Unionist Party. Innovation in Social Media For Marketing & Personal Recommendations According to a report, 85% of all customer interactions with brands will be automated by 2020, from personal recommendations, marketing automation, NLP and data analytics, as 2019 saw developments in this space. For example, Rezo.AI’s Conversational AI engine showed in 2019 how brands can create personalised relationships with customers without the requirement of human intervention on the biggest messaging platform — WhatsApp. In 2019, Facebook also announced an in-house tool which essentially allows them to use AI to create Facebook adverts for people. The announcement led many people to think that Facebook AI may kick out social media marketers. However, Facebook AI will be integrated with Facebook ad manager, which will let businesses to instantly create and distribute advertisements for the right audience using data insights from users. One of the other trends that took on was customized news feeds, personalized content, and being able to segment information using AI. Instagram shared new subtleties on how its application utilizes AI to surface content for users. When making recommendations, it centres around discovering accounts it figures individuals will appreciate, as opposed to particular content posts. In a 2019 blog post, Instagram explained how algorithmic recommendation systems work on its platform. Similarly, India’s Hike messenger has also announced new AI innovations for both. This included sticker recommendations, message recommendations, vision, speech, anti-fraud, auto campaign tuning, and customer experience based optimisation using AI. Rise Of Deepfakes On Social Media And Innovation That Followed One of the other big trends to hit social media in 2019 was the FaceApp which used AI to transform a user’s face. The application adds certain features using the artificial intelligence which hadn’t been introduced before like adding a smile or looking into the future and how one might look like. Particularly, the age filter led to the recent explosion in popularity making it the top free app in the IoS App store. But, in the wake of social media scandals, user’s private information, many wondered if FaceApp could be used for nefarious purposes and mining facial data of billions of users. Of course, people did not read the privacy policy of the application where users knowingly or unknowingly gave FaceApp the right to use photos for whatever they deem fit. Experts also expressed concerns on how facial data through apps like FaceApp could also be used for creating Deepfakes, which has been another big trend seen on social media for 2019. Deepfakes reached the peak level in 2019 in China, when the app Zao took the country by storm in September offering people a simple way to superimpose their own faces onto those of actors like Leonardo DiCaprio and Marilyn Monroe. The DeepFake menace that reached a new peak in 2019 sent alarm signals across government agencies like DARPA which created a social media forensics department known as National Centre For Media Forensics to tackle the issue under the Defence Research Projects Agency, where experts are working to develop algorithms to spot the fake from real. Even Facebook announced a Deepfake detection challenge after providing a lot of facial datasets on Kaggle.","excerpt":"Increased investment in Artificial Intelligence (AI) technology, along with accelerating social media penetration, is a key driver of global AI in the social media market. According to a report, the total expenditure on AI systems is expected to reach $5.5 billion this year, which is an increase of 80 per cent from last year. Gartner […]","categories":["AI Features"],"tags":["DARPA","purpose of artificial intelligence","social media","social media AI","social media analytics"],"author_name":"Vishal Chawla","publish_date":"2019-12-30T13:00:00","publication_year":"2019","word_count":776,"keywords":["Go","artificial intelligence","machine learning","AI","DARPA","social media analytics","social media","recommendation systems","RAG","social media AI","NLP","automation","analytics","purpose of artificial intelligence","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","NLP","analytics","RAG","recommendation systems","R","Go","automation"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-artificial-intelligence-gave-impetus-to-new-social-media-innovation-in-2019\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":9983,"title":"8 ways of Boosting Performance of Machine Learning Models","content":"People often get stuck when they are asked to improve the performance of existing predictive models. What usually they do is try different algorithms and check their results. But often they end up not improving the model. Here are some of the steps you can take to boost your existing models. 1. Add more data More data is always useful. It helps us capture all the variance that the data has. Sometimes we may not have the option to get additional data for training. Example when you are competing in data science competitions. But while working on client projects, you can ask for more data if it’s desired. The question is when we should ask for more data?We cannot quantify more data. It depends on the problem you are working on and the algorithm you are implementing, example when we work with time series data, we should look for at least one-year data; And whenever you are dealing with neural network algorithms, you are advised to get more data for training otherwise model won’t generalize. 2. Feature engineering Adding new feature decreases bias on the expense of variance of the model. New features can help algorithms to explain variance of the model in more effective way. When we do hypothesis generation, there should be enough time spent on features required for the model. Then we should create those features from existing data sets. i.e. We want to predict daily withdrawals from ATMs. In this case, we can think people could be inclined to draw higher amounts at the start of the month. Possible reason can be people get their salaries or they pay for various monthly expenditures at the start of the month. So we will create a new feature for this.2. When working on fraud detection model, We can look for the ratio of Income to Loan as a new feature. 3. Feature selection This is one of the most important aspects of predictive modelling. it’s always advisable to choose important features in the model and build the model again only with important and significant features.i.e. Let’s say we have 100 variables . There will be variables which drive most of the variance of a model. If we just select the number of features only on p-value basis, then we may still have more than 50 variables. In that case, you should look for other measures like contribution of individual variable to the model. If 90% variance of the model is explained by only 15 variables then only choose those 15 variables in the final model. 4. Missing value and Outlier Treatment Outliers can deflect your model so badly that sometimes it becomes essential to treat these outliers. There might be some data which is wrong or illogical. i.e. Once I was working on airline industry data, in the data there were some passengers whose age is 100+ and some of them were 2000 years. So it is illogical to use this data. This is harder to explain but it is likely that some users intentionally entered their age incorrectly for privacy reasons. Another reason might be that they might have placed their birth year in the age column. Either way, these values would appear to be errors that will need to be addressed. In the same way, missing value issue should be addressed. Missing values treatment can play a role in boosting performance. i.e. Working with time series data, we can replace values by their overall mean or by their month wise mean. Month wise mean would be most logical because seasonality can deflect your overall mean. This effect will be visible in your prediction and performance will get better.Missing value and Outlier Treatment are part of the modelling process. You may be thinking how these can help in boosting performance. Both issues can be addressed in several ways. You have to identify which is the best way for a given task. Right method will lead to performance improvement. 5. Ensemble Models Ensemble modelling is one of the popular techniques in improving modelling outcomes. Bagging (Bootstrap Aggregating) and Boosting are some of the ways which can be used. These methods are generally more complex and black box type approaches.We can also ensemble several weak models and produce better results by taking the simple average or weighted average of all those models. The idea behind ensemble modelling is that one model can be better at capturing variance of the data and another model can be better at capturing only the trend. In these types of cases, ensemble method works great. 6. Using the suitable Machine learning algorithm Choosing the right algorithm is a crucial step in building a better model. Once I was working with holtzwinter model for prediction but It performed badly for real-time forecasting so I had to move on neural network models. Some algorithms are just better suited to some data sets than others. Identifying the right type of algorithms is an iterative process thought. You need to keep experimenting with different algorithms to eventually land onto highly efficient one. 7. Auto- feature generation The quality of features is critical to the accuracy of the resulting machine learning algorithm. No machine learning method will work well with poorly chosen features. But when we use deep learning algorithms you don’t need feature engineering. Because Deep learning does not require you to provide the best possible features, it learns by its own. If you are doing image classification or handwriting classification then deep learning is for you. Image processing tasks have seen amazing results using deep learning. Below picture is showing how features are created automatically in every layer. You can also observe how much features are getting better after every layer. 8. Data distribution and Parameter Tuning It is always better to explore the data efficiently. The data distribution might suggest transformation. The data might be following the gaussian function or some other family of function, in that case, we can apply algorithm with a little transformation to have better predictions. Another thing we can do is fine tuning of parameters of algorithms. i.e. when we build random forest classifier we can tune the number of trees to build, the number of variables to choose for splitting etc. Similarly, when we build deep learning algorithm we can specify how many layers we would need, how many neurons we want in each layer, which activation function we want. Tuning parameter enhances model performance if we use the right type of parameters in an algorithm. Conclusion: Improving performance of machine learning models is hard. Above methods of improving performance are based on my experiences. When we use ensemble method, it requires a thorough knowledge of algorithms.Algorithms like Random Forest, Xgboost, SVM, and Neural network are used for high performance. Not knowing how the algorithm can be tuned well to training data is a barrier in getting higher performance. So we should always know how the algorithms can be tuned according to different tasks.When we do parameter tuning, we should take care of overfitting. You can use cross-validation methods to prevent overfitting.I hope this article satisfies your curiosity towards performance boosting techniques. Keep in mind these techniques whenever you try to boost performance of existing models.","excerpt":"People often get stuck when they are asked to improve the performance of existing predictive models. Here are some of the steps you can take to boost your existing models.","categories":["AI Features"],"tags":[],"author_name":"Ved Prakash","publish_date":"2016-05-24T05:50:18","publication_year":"2016","word_count":1206,"keywords":["data science","Go","machine learning","AI","neural network","RAG","deep learning","XGBoost","R","fraud detection"],"extracted_tech_keywords":["AI","machine learning","deep learning","neural network","data science","XGBoost","RAG","fraud detection","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/8-way-boosting-performance-machine-learning-models\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10168766,"title":"Northeast Frontier Railway Conducts Its First Drone-Based Cleaning","content":"A senior official recently revealed that the Northeast Frontier Railway (NFR) completed its first drone-based cleaning operation on a pilot basis at Kamakhya Railway Station in Assam, Assam Tribune reported. It specifically targets the difficult-to-reach and higher areas of the station premises, as well as the roof and other areas of the train coach, showcasing the potential of the technology in enhancing cleanliness and hygiene. Kapinjal Kishore Sharma, chief public relations officer of NFR, stated that the cleaning operation encompasses areas such as the Kamakhya coaching depot sick line, the under-floor wheel lathe shed, the exterior dome of the station and several train coaches. With this, the NFR showcases how Assam is advancing in technology. “The operation showcased the ability of drone technology to access and maintain elevated structures with precision and ease,” Sharma added. NFR is now adopting innovative, technology-driven solutions to improve the efficiency, precision, and hygiene of its cleaning operations. “Drone-based cleaning not only improves accessibility and precision but also reduces the dependency on manual labour in hazardous or elevated areas. The success of this pilot demonstration opens the door to broader implementation of drone-based cleaning across additional major stations and trains under the NFR network,” Sharma added. The step is part of Indian Railways’ vision to incorporate cutting-edge, innovative maintenance practices that meet global standards. Sharma highlighted that the NFR continues to set a standard for sustainable, technology-driven improvements, trying to create a cleaner, safer, and more efficient environment for both passengers and railway employees.","excerpt":"NFR is now adopting innovative, technology-driven solutions to improve the efficiency, precision, and hygiene of its cleaning operations.","categories":["AI News"],"tags":["indian railways"],"author_name":"Amisha Arya","publish_date":"2025-04-29T10:15:55","publication_year":"2025","word_count":249,"keywords":["indian railways","API","programming_languages:R","AI","R"],"extracted_tech_keywords":["AI","R","API","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/northeast-frontier-railway-conducts-its-first-drone-based-cleaning\/","complexity_score":2,"technical_depth":4,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":57498,"title":"Benefits Of Reverse Mentoring In Data Science","content":"Mentorship has many benefits in the data science industry. For young data, science enthusiast mentoring can not only fill their knowledge gap about the real world industry but can also help in improving their skills and core understanding of using data science for businesses. For organisations, having a mentoring program can help in enhancing the teamwork between employees and to improve employee skills. In a typical mentoring session, older executives and data scientists with greater seniority take in young mentees who are enthusiastic about learning data science. However, with the progress in technology and more executives from generation X and Y coming in the forefront of the workforce, ‘reverse mentoring’ has stood out to be beneficial for several organisations. Reverse mentoring is a contemporary process of mentorship that relies on role reversals between a mentor and a mentee. Reverse mentoring usually happens when a younger executive mentors somebody who is comparatively more experienced but isn’t knowledgeable about the same skill set as the younger employee, such as learning new software or new technology. This has been considered to be one of the most effective ways to create visibility of upcoming leaders to top executives. In this article, we are going to list down the benefits of reverse mentoring in the field of data science. New Tech New Teachers Since its inception, in 1999 by Jack Welch — during his tenure as CEO of GE, reverse mentoring acted as a process of bridging the gap between the generations, which was being created by the advent of technology. Keeping pace with dynamic shifts in technological change is indeed difficult. With reverse mentoring companies can pair up younger executives with experienced business leaders to facilitate knowledge and to break their traditional habits and get them used to the newer technology drive of the industry. Embracing collaboration can improve the competitive advantage of a business. It allows older executives to stay up to date on more modern technologies and with the workplace getting more urbanised, reverse mentoring can help these older or traditional folks to adapt to changing workplace technologies and trends. Newer businesses believe that there is a lot to be learned from younger employees, especially from millennials and generation Z who have grown up with technology, which in turn, can benefit the whole organisation in this digitised landscape. Increasing Retention Of Millennials Millennials, currently, are the largest generation in the labour workforce; however, according to reports, companies are struggling to engage or even retain their millennial employees. Along with flexible work environments and comfortable ecosystem, these millennials are hungry for knowledge and demand continuous learning and skill development, and of course, quicker opportunities for career advancement. And that’s where reverse mentoring plays a considerable role. Inverting the traditional hierarchy by allowing younger executives to mentor more experienced ones can help the organisation create more value for their millennial workforce. It will also help younger executives to have clear access to networks with senior and accomplished leaders, which can enhance their career paths in the future. Reverse mentoring will improve millennial’s confidence in their organisation, in turn, help the company to reduce their attrition rate of millennials, and might even find their future leaders. Build Social Culture Within The Workplace Reverse mentoring has proved to be a perfect way to build and enhance social capital within the workplace. By pairing senior leaders and experienced employees with younger executives from different backgrounds will not only develop empathic perspectives from both sides but also, gradually, omits the unwanted awkwardness between the different generation of employees. It will foster better communication and collaboration in everyday interactions and will incubate a diverse pipeline of talent for both lower and upper-level management across the organisation. Reverse mentoring gives a reality check for older and experienced employees and thus will provide the opportunity to hone and refine their leadership practices, based on feedback and discussions with their younger colleagues. Also, since reverse mentoring is considered to be a relaxed affair, it surely gives the senior leaders a welcome break from the norm, where they can interact and discuss their work with someone younger and with fresh ambition. Saving On Development Training and development programs with external resources can be a costly affair for businesses, especially the ones that are still budding. So, if an organisation is limited to its budget for upskilling their employees, this more informal approach of reverse mentoring could be extremely beneficial. Here, younger mentors get together with older mentees for coffee or chitchat to process knowledge about newer technologies and urbanised business processes. Reverse mentoring programs are a cost-efficient approach to enhance the engagement of your workforce and to develop the skills of your employees by creating a training culture, which increases productivity. With agreed objectives, reverse mentoring can be a great way of learning without busting the all-important balance sheets! Two Way Street Reverse mentoring is considered to be a two-way learning process, where both the younger mentor and older mentee can be benefited by working closely. Each of the people, whether it be the younger executive or the senior leader, see life through very different perspectives. With this process, both can learn ways to build relationships and communicate with colleagues of different levels. With this two-way learning process, both sides can not only learn the viewpoints of the other person from a different generation but also can gain self-confidence. Reverse mentoring also promotes the inclusion of diversity, facilitates cross-generation relationships, increases communication and enhances engagement across generations. This process presents a major cultural shift among traditional workplace and implementing the same will indeed benefit the organisation. Additionally, reverse mentoring builds the resilience of the business, boosts the reputation of the organisation, and minimises the problems related to retention by bringing in diverse employee generations.","excerpt":"Mentorship has many benefits in the data science industry. For young data, science enthusiast mentoring can not only fill their knowledge gap about the real world industry but can also help in improving their skills and core understanding of using data science for businesses. For organisations, having a mentoring program can help in enhancing the teamwork between […]","categories":["AI Features"],"tags":["data science mentor","data science mentorship","mentors in data science"],"author_name":"Sejuti Das","publish_date":"2020-02-29T16:00:00","publication_year":"2020","word_count":961,"keywords":["data science","Go","API","programming_languages:R","AI","data science mentorship","mentors in data science","data science mentor","Git","programming_languages:Go","ViT","GAN","R"],"extracted_tech_keywords":["AI","data science","R","Go","Git","API","GAN","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/benefits-of-reverse-mentoring-in-data-science\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10052274,"title":"Tech Behind MSME Fintech Startup ARTH","content":"According to media reports, the Micro, Small and Medium Enterprises (MSME) sector currently faces a $380 billion credit gap. Mumbai-based ARTH wishes to use technology to bridge this gap. The women-first MSME fintech startup was founded in 2018 by Shweta Aprameya. Established under the brand name of Arthimpact Digital Loans, the startup focuses on women-led enterprises with limited or no online adoption. It started operations by offering credit services and soon added embedded insurance into its product portfolio. So far, it has extended over five lakh loans, impacting over 3.5 lakh micro-entrepreneurs. The startup evaluates customers using reliable, verified data, occupation, digital records and social factors and does not depend on standard paperwork such as credit bureau records or income tax statements. The fintech startup operates in micro-markets and has serviced 3.5 lakh micro-entrepreneurs who do not have bank statements and GST records across 18,400 pin codes in India. Recently, Chief Product Officer Raman Taneja got into a conversation with Analytics India Magazine to explain how ARTH uses AI and ML to create digital personas for customers that ultimately helps in the designing of suitable financial products, rather than going for a one-size-fits-all approach. Raman has an MBA from IBS Hyderabad and has earlier worked as CTO at Dvara Solutions and as the Head of Digital Financial Inclusion, Consumer Banking, Strategic Partnerships and Distribution Channels at Jio Payments Bank. Edited excerpts from the conversation: AIM: What pain point is ARTH trying to address? How is it doing so? Raman Taneja: ARTH targets New-to-Credit (NTC) and thin-file micro-MSME entrepreneurs. These include micro-businesses across categories including grocery, apparel, restaurant and others. While these entrepreneurs are savvy in managing their money, they lack access to formal credit and other relevant financial services such as insurance and digital payments. ARTH’s theory of change is built on the premise that if fair-priced and flexible financial access is given to micro-entrepreneurs, it can bring viable returns on business. Since technology adoption is still limited amongst the target group it services, we have adopted a hybrid touch-and-tech model. The team also helps new-to-credit customers understand the details of the credit and payment mechanism to ensure there are no surprises for the customer later. In turn, this provides us with a chance to understand better the needs of a customer and design products suitable for them. For instance, during the Covid-19 lockdown, many customers needed capital to re-start their business. We offered ‘COVID Rahat loans’– interest-free financial loans to help customers who were severely impacted. Rahat loans were targeted at women nano entrepreneurs with earnings less than Rs 3 lakh per annum. AIM: What are your key roles and responsibilities as the CPO at ARTH? Raman Taneja: My responsibilities involve developing tech-led products and partnerships to have an easy-to-use interface for digital and non-digital customer segments. The product tech stack uses open-source technology extensively and has adopted innovations from across the world. In addition, we use public and private APIs to collect customer data and provide easy navigation throughout the onboarding experience. The idea is to build a product tech stack to handle scale. The product roadmap includes creating a seamless experience across credit and insurance across partners and channels. While our focus so far has been on credit, we will soon be expanding into other financial products to upscale micro-MSMEs through our digital platform. AIM: What are your key products and services? Raman Taneja: At present, the focus has been on providing credit products to micro-MSME entrepreneurs. We have co-developed innovative insurance products with one of the leading digital insurance companies, focusing on micro-MSMEs’ different needs. Through insurance products, we offer life and health insurance as well as income protection to the micro-MSMEs. We are also looking at launching new products for upscaling micro-MSMEs through digital platforms. The product roadmap includes creating a seamless experience across credit and insurance across partners and channels. The digital platform will offer linkages to both forward and backward supply chains to make credit available on demand. AIM: Explain the tech stack. Raman Taneja: We have been using a combination of open-tech platforms like Angular, Ruby on Rails and Postgres to develop our technology backend. Open-source platforms allow us to adopt and implement some of the best-developed technologies from across the world at a reasonable cost of ownership. This includes mobile frontend frameworks and backend technology. Tech stack conforms to APIs and microservices framework and can be integrated into any number of APIs in a matter of days. AIM: What tech tools do you use? Raman Taneja: We use: Open-source tools for development and deployment on cloud The new framework under progress is also cloud-native Tools provided by Amazon Web Services to manage day-to-day operationsOptical Character Recognition (OCR) and computer vision tools to read KYC documents. AIM: Explain the ML model you use to detect fraud. Raman Taneja: We consume reliable open-source data available publicly as well as alternate data collected from users for training our AI and ML models. The entire customer lifecycle data gets fed into the ML model for training our engine to identify potential risks and new opportunities. ML model trains on customer onboarding, transaction and public APIs data to create predictive behaviour, which helps identify risk better. AIM: How are you leveraging AI and ML technologies? Raman Taneja: For underwriting a customer for credit, our proprietary ML model determines eligibility through alternate social, economic, business sales and behavioural data. To further enhance and continue deep learning of the customer base, the ML model evolves a Predictive Risk Score of a customer and identifies patterns for payers and non-payers. Since ARTH largely works with new-to-credit and thin-file micro-entrepreneurs, we lay a lot of importance on customer education and awareness. Once the credit is approved, our representatives impart know-how to each customer and explain the details of the interest payment schedule and other rules such as late payment fees. Feedback from each such session is also fed into the ML model for future use. We use computer vision to ensure the paperwork filled in by a customer for KYC is valid and will be acceptable by the guidelines required by RBI. This helps our systems to detect any anomaly in the submission on a real-time basis. AIM: Who do you see yourself competing with? How do you differentiate yourself from your competitors? Raman Taneja: We have stayed razor focussed on addressing the credit needs of micro-MSMEs. Most of these enterprises do not have financial information in standard formats, our AI\/ML engine captures data across social, economic, and business parameters to assess creditworthiness. Our credit size is also extremely flexible to suit a customer’s requirement at different points in time. We reach out to our potential customers through tech & touch models and work with network operators at both national and hyperlocal levels. AIM: What is the success rate of your model? Raman Taneja: The model has shown excellent results in the lowest ticket size customers, thereby showcasing the objective of bringing a large scale impact on the micro-MSME segment. The favourable experience on access to credit is shown in a great loyal base of over 60 per cent repeat customers. So far, ARTH’s Credit Risk model has shown one of the best results in the industry. AIM: How many loans have been disbursed by ARTH so far? What is the average ticket size? Raman Taneja: ARTH has serviced 350,000 unique customers, and 55 per cent of these are new to credit. The average ticket size for a loan is around Rs 50,000. AIM: What are ARTH’s future plans? Raman Taneja: Financial inclusion in India has made great collective progress. From being a largely unbanked country, the various initiatives supported by regulators and the government have led to the opening of nearly a billion bank accounts in India and providing access to customers through a large number of banking agents. Access to credit, however, continues to be limited, especially to micro-enterprises and businesses. These customers do not have formal records of their business transactions and business establishments, which results in typical issues of information asymmetry and exclusion. While our focus has been on credit, we will soon be expanding into other financial services to upscale micro-MSMEs through a digital platform. ARTH will create a partner ecosystem to provide credit-on-demand to all the stakeholders.","excerpt":"Mumbai-based fintech startup ARTH provides financial assistance to women-led enterprises with limited or no online adoption.","categories":["AI Startups"],"tags":["insurance technology","open-source software"],"author_name":"Debolina Biswas","publish_date":"2021-10-25T14:00:00","publication_year":"2021","word_count":1383,"keywords":["Go","AI","open-source software","ML","computer vision","RAG","microservices","Aim","deep learning","analytics","insurance technology","R"],"extracted_tech_keywords":["AI","ML","deep learning","computer vision","analytics","Aim","RAG","microservices","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-startups\/tech-behind-msme-fintech-startup-arth\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":56539,"title":"Bengaluru-Based Customer Analytics Firm MoEngage Raises $ 25 Million","content":"MoEngage, Bengaluru-based customer analytics and cross-channel engagement platform, recently said that the startup has raised $25 mn in a Series C funding round led by Eight Roads Ventures. The round also saw participation from Eight Roads’ US-based sister fund, F-Prime Capital, along with Matrix Partners India and Ventureast. The funds raised will be utilised to expand relationships in Asia, develop advanced capabilities into the product portfolio and scale ops in the US and Europe, MoEngage’s two fastest-growing markets, the startup said in a statement. “The latest round of funding will help MoEngage reach more brands and empower them with the next-gen customer engagement platform built for the mobile-first world that is easier to use, fully integrated and intelligent,” stated Raviteja Dodda, Founder & CEO, MoEngage. Founded by IIT Kharagpur alumnus, Raviteja Dodda and Yashwanth Kumar, Bengaluru- and San Francisco-based MoEngage centralises user behavioral analytics, insights and marketing automation into a single dashboard where brands are able connect with their customers across channels and personalise different touchpoints like mobile, e-mail, in-app, web, and SMS. The company claims that hundreds of global brands across 35 countries use its platform to reach more than 400+ million monthly active users, processing over 65+ billion interactions and 40+ billion messages every month. Currently, enterprise clients including Vodafone, Samsung, Tokopedia, OYO, Future Retail, etc. make up nearly 50 per cent of MoEngage’s total revenue by enabling hyper-personalisation solutions, consumer analytics and marketing automation capabilities. MoEngage had raised $9 million in Series B funding led by Matrix Partners and Ventureast in December 2018. Existing investors Helion Venture Partners and Exfinity Venture Partners also participated in the funding round.","excerpt":"MoEngage, Bengaluru-based customer analytics and cross-channel engagement platform, recently said that the startup has raised $25 mn in a Series C funding round led by Eight Roads Ventures. The round also saw participation from Eight Roads’ US-based sister fund, F-Prime Capital, along with Matrix Partners India and Ventureast.  The funds raised will be utilised to […]","categories":["AI News"],"tags":["intelligent automation for retail"],"author_name":"AIM Media House","publish_date":"2020-02-12T17:08:17","publication_year":"2020","word_count":271,"keywords":["API","funding","AI","RAG","intelligent automation for retail","automation","Aim","ViT","analytics","R","startup"],"extracted_tech_keywords":["AI","analytics","Aim","RAG","R","API","ViT","automation","startup","funding"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/bengaluru-based-customer-analytics-firm-moengage-raises-25-million\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10166139,"title":"Infosys, HPE, and SUSE Partner to Simplify AI Adoption for Enterprises","content":"Infosys, Hewlett Packard Enterprise (HPE), and SUSE announced a strategic partnership at the SUSECON 2025, and unveiled an enterprise AI infrastructure solution designed to revolutionise AI deployment for global businesses. The initiative brings together SUSE AI’s open-source capabilities, HPE’s high-performance AI-ready hardware, and Infosys’ expertise in system integration and responsible AI governance, offering enterprises an end-to-end solution to accelerate AI adoption at scale. This move builds upon the decades-long strategic partnership between Infosys and HPE, which has helped enterprises modernise IT environments, enhance AI\/ML adoption, and drive digital transformation. The latest announcement signals a new era, where AI becomes more scalable, cost-effective, and enterprise-ready than ever before. Ramanath Suryaprakash, vice president of partner ecosystems and global alliances at Infosys, said that the company is helping organisations tackle the complexity of AI adoption. “Our combined capabilities ensure businesses have the right tools to build, deploy, and scale AI in a secure and responsible manner.” As AI adoption surges, a trusted framework for AI deployment is also needed. Businesses face challenges ranging from high infrastructure costs and security risks to the complexity of AI model implementation. The Infosys-HPE-SUSE alliance simplifies AI adoption through a phased implementation strategy, enabling businesses to test AI models, validate use cases, and scale responsibly. “AI is powerful, but its true potential is unlocked when combined with robust infrastructure and human expertise,” said Mandanna Appanderanda, head of Infosys responsible AI office in Americas. “We are ensuring that businesses focus on innovation while we handle the complexity of AI integration.” To make AI more accessible, the partners have introduced a ‘Start Now, Pay Later’ model, allowing businesses to scale AI investments based on real-world outcomes. Tom Hempfield, Americas VP sales and GTM at HPE, emphasised the company’s role in optimising AI infrastructure. “Our HPD Gen 12 servers and GreenLake cloud platform provide an AI-ready foundation that is not only secure but also cost-efficient.” Responsible AI: A Key Differentiator AI’s ethical considerations are at the forefront of this initiative. Infosys’ Appanderanda stressed the importance of trust in AI, stating, “Our Responsible AI framework ensures AI solutions are not only powerful but also explainable, compliant, and secure.” “With open-source AI stacks and a responsible AI governance framework, we are giving enterprises the confidence to innovate without risk,” Appanderanda added. Notably, this collaboration comes as Infosys doubles down on AI-driven growth. The company is seeking an early renewal of its $3 billion contract with Daimler, with AI integration expected to add a seventh revenue stream to its existing portfolio. Infosys’ push into AI aligns with broader industry trends, where businesses are rapidly shifting towards multi-agent AI frameworks and small language models for industry-specific applications. The Road Ahead: Enterprise AI at Scale With Infosys and HPE’s long-standing partnership evolving into a full-fledged enterprise AI powerhouse, businesses can now access scalable, cost-effective, and responsible AI solutions tailored for their industry needs. “Be pioneers, be creators, and let’s build an AI-powered future that is safe and trustworthy,” SUSE’s AI head Abhinav Puri said at the closing keynote of SUSECON 2025.","excerpt":"The alliance simplifies AI adoption through a phased implementation strategy, enabling businesses to test AI models, validate use cases, and scale responsibly.","categories":["AI News"],"tags":["AI"],"author_name":"Shalini Mondal","publish_date":"2025-03-17T15:18:09","publication_year":"2025","word_count":505,"keywords":["Go","API","AI","ML","Scala","Git","Aim","Rust","small language models","R"],"extracted_tech_keywords":["AI","ML","Aim","small language models","R","Go","Rust","Scala","Git","API"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/infosys-hpe-and-suse-partner-to-simplify-ai-adoption-for-enterprises\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":7826,"title":"Sports Analytics: The Game Changer","content":"Decisions about which player to drop for the next match, which team combination would be best against a particular opposition and which player needs what type of training have direct impact on the outcome of a match. These decision in front of teams and coaches have been taken mostly based of gut feeling in the past. The trendsetter of data driven decision making was Billy Beane. General Manager of team Oakland Athletics, and a former baseball player himself believed that the runs scored by a team could be predicted using analytics. His story showcased in the movie money ball in 2003, gave an insight to viewers about how using data can change the way a game can be played and in recent times, development of data approaches to sports has continued to evolve. The base of all the analytics is the data which is collected on and off the field. Lots of collected during the game is via manual methods. Teams had to scan there scout notes from clip board, convert them to excel or pdf and give them to their data developers. The problem is, the data moves so fast in the field that it is lost in time.  The quest for getting more real time data has led to the analytics providers as well as teams are coming up with new and sophisticated ways of capturing, monitoring and analysing the ever increasing volume of the sports data.   The technology is now trying to bridge this gap. For example, a company named, “ZebraTechnologies, manufactures RFID tags. These tags can be attached to players; equipment’s and even balls to gauge their movements, distance and speed.  Blinking 25 times per second, the tag delivers recorded data in 120 milliseconds which is not possible by any scout to match.” [1] Devices like cameras sensors and other wearables are also being used to record all the performance aspects of the players on and off the field. Be it calorie intakes, training levels and even the interaction of the players with fans and other players. Data is also provided for analytics by Third Party Providers of each team. “The major players in this category, both large organizations like Bloomberg and sports analytics entrepreneurs— include: ShamSports—NBA salary data Bloomberg Sports—Player performance data and “match analysis” for all major professional sports com—MLB “wins above replacement player” analyses Sports Reference—data and analytics on major professional sports ProFootball Focus—NFL player analysis Opta and Prozone—English Premier League football (soccer)” Wearable devices are being increasingly used to capture and analyze real time information. Adidas’s miCoach is one such device which can be attached to the jersey of a player. It records real time data like the heart rate, acceleration and speed of the player. It also shows which player needs rest based on the data collected and who the top performers are. Such devices are also helping in reducing the injuries of a player. None of the teams want their star players to sit out of a match due to injuries. For example, in a sports like rugby where the chances of getting injured is very high, the injury levels has been reduced due to the use of wearable sensors. The sensors record the impact of collision, intensity of activity and compares it with the historical data in the database to determine if a player is in a danger of getting injured or is overexerting. In NFL GPS trackers are now a part of the standard training kit for the players. FIFA has long opposed wearing of such devices, but now for the first time they have allowed it. Tennis federation has also relaxed rules regarding usage of wearable devices. More and more sports are turning towards analytics. The data can help the trainers and physicians to better train the players as well as by the coaches to better plan their games. In addition to the wearable devices, video analytics is also being increasingly used across various sports. “A company named, SportsVU, has 6 cameras in each game of NBA which tracks movements of players and ball 25 times per second. Basketball Players have been chosen based on the information conveyed by advanced metrics which show which move and shots is a particular player strong in. Using these insights, the coach and the team formulators had an ease of selecting the players with the appropriate defence tactics needed to counter the stronger players. The selected player were then given unprecedented level of statistical information about the opponent he had to guard. This gave the players an advantage over the opponents and an edge in the game. In the ninth annual MIT Sloan Sports Analytics Conference (SSAC), Battier, now retired basketball player recalled onstage as to how he replaced Morey in the team in spite of averaging only 10 points per game due to the analytics revelation of his fine defence and shot selection. It is not just NBA, in premier league soccer in UK, the famous team Arsenal has made huge investments in analytics, and they have made their own analytics team to make use of the data they have been collecting. 8 cameras installed around the stadium records players of every players. They system is provided by a company named Prozone, and tracks 1.4 million data points in every match. It monitors 12000 matches across the globe and analyzes all the data using automated algorithms.” [4] The system, also known as “Player Tracking: is used to analyze the overall efficiency of a team by analyzing the movements of its players. The data collected is used to analyze the game of each player.  The players and coaches have can get access to data like ‘”all passes by Lionel Messi that were unsuccessful”’ or ‘”all successful tackles by Christiano Ronaldo” [5]. A player interacts only for a fraction of a game with the ball. Off ball events become even more important wherein player runs or get into positions where he can hurt himself. Teams in many sports have found that the cost of implementing analytics programs can quickly be recovered if they save the team having to pay for expensive players to sit out a season with injuries. The data driven decision making has also trickled down to the fans. They now receive more analytics content than ever. There are websites like FiveThirtyEight.com which are dedicated to sports research and analytics. It uses a sophisticated system of number crunching and past data to determine the most likely outcome of a game. With fan betting on sports legal in countries like Singapore, the fans use these platforms to make betting decisions. For any sports organization, fan experience is the ultimate measure for its success. Better the teams and the ticket vendors understand their fans the better they will be able to cater them. Right from the ticket purchase to ordering merchandise fans generates lots of data points. Companies like SAS, provides analytics solution to analyze create personalized communication to the fans.  Sports organizations can information like whether a 7pm game is too late for fans in a particular place or when a fan is interested in a particular tea coming to their town. FanApps are being used to provide information about the favorite players of the fan. Analyzing there buying patterns and suggesting targeted discounts on the merchandise. These apps even provide information like the bathroom wait times. Fan following does not fit in any model. The fans for the best player or the maximum playing sportsperson need not necessarily be the greatest. As such, the fan trend needs to be analysed to find out which player do the fans want to know about and what information about the players are they interested in. This information can be deduced by applying analytical tools on the fan data collected from social media platforms, mobile devices, emails, etc. This will help to enhance the digital experience of the fans and provide means of keeping them hooked to various digital portals such as fan pages, team sites and public forums. The analytics will continue to evolve and would be relied on to make decisions in sports. The teams using are having competitive advantages over others. It is just a matter of time that others will follow. -Varun Mishra PGDM, Marketing, 2nd Year IMT Ghaziabad","excerpt":"Decisions about which player to drop for the next match, which team combination would be best against a particular opposition and which player needs what type of training have direct impact on the outcome of a match. These decision in front of teams and coaches have been taken mostly based of gut feeling in the […]","categories":["IT Services"],"tags":[],"author_name":"AIM Media House","publish_date":"2015-09-05T11:41:16","publication_year":"2015","word_count":1379,"keywords":["Go","AI","ML","Git","RAG","ViT","analytics","CLIP","GAN","R"],"extracted_tech_keywords":["AI","ML","analytics","RAG","R","Go","Git","CLIP","GAN","ViT"],"url":"https:\/\/analyticsindiamag.com\/it-services\/sports-analytics-the-game-changer\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10061799,"title":"Top resources to learn reinforcement learning in 2022","content":"Reinforcement learning is a ML training method based on rewarding desired behaviours and punishing undesired ones. A reinforcement learning agent can perceive and interpret its environment, take actions and learn through trial and error. Reinforcement learning is largely used in autonomous driving, automated cooling for data centres, recommendation engines, personalised chatbots, stock trading etc. Here, we look at the top resources to learn reinforcement learning in 2022: RL course by David Silver DeepMind research lead David Silver’s course on reinforcement learning taught at University College London is laid out in ten YouTube videos. The videos cover Introduction to Reinforcement Learning, Markov Decision Processes, Planning by Dynamic Programming, Model-Free Prediction, Model-Free Control, Value Function Approximation, Policy Gradient Methods, Integrating Learning and Planning, Exploration and Exploitation, Case Study: RL in Classic Games. To access slides, assignments, exams, check out the link. Introduction to Reinforcement Learning with Function Approximation Rich S. Sutton, a research scientist at DeepMind and computing science professor at the University of Alberta,  explains the underlying formal problem like the Markov decision processes, core solution methods, dynamic programming, Monte Carlo methods, and temporal-difference learning in this in-depth tutorial. A History of Reinforcement Learning Prof AG Barto, professor emeritus of computer science at University of Massachusetts Amherst, offers a detailed lecture. The chapters include “Hedonistic neuron” hypothesis, Supervised learning, Reinforcement learning, A unique property of RL, Edward L Thorndike, Law of Effect, RL= Search+Memory, Our first surprise, Though there were exceptions, An early paper with Rich Sutton, Associative Memory Networks, Associative Search Network and many more. NTPEL course The course entails a series of lectures by Prof Balaraman Ravindran, Computer Science and Engineering and Robert Bosch Centre for Data Science and AI, IIT-Madras on Reinforcement Learning. The course introduces the basic mathematical foundations of reinforcement learning and highlights some of the recent directions of his research. The 12 weeks lecture contains preparatory material, introduction to RL and immediate RL, Bandit Algorithms, Policy Gradient Methods and introduction to Full RL, MDP Formulation, Bellman Equations & Optimality Proofs, Dynamic Programming & Monte Carlo Methods, Monte Carlo & Temporal Difference Methods, Eligibility Traces, Function Approximation, DQN, Fitted Q & Policy Gradient Approaches, Hierarchical Reinforcement Learning, Hierarchical RL: MAXQ and POMDPs. Artificial Intelligence: Reinforcement Learning in Python Artificial Intelligence: Reinforcement Learning in Python is a complete guide to reinforcement learning with stock trading and online advertising applications. The 14.5 hours-course is available as on-demand video in Udemy.  The guide will teach you to apply gradient-based supervised machine learning methods to reinforcement learning, understand reinforcement learning on a technical level, understand the relationship between reinforcement learning and psychology, and implement 17 different reinforcement learning algorithms. Reinforcement Learning in Unity The students can learn to set up reinforcement learning in Unity3D and unlock the power of combining game engines with artificial intelligence by using it to train a tile to balance a little ball. Details can be found in https:\/\/github.com\/Unity-Technologies https:\/\/unity.com\/ Introduction to reinforcement learning Hado Van Hasselt, Research Scientist, Senior Staff Research Scientist at DeepMind, Honorary Professor at UCL has shared an introduction to reinforcement learning as part of the Advanced Deep Learning & Reinforcement Learning Lectures in this YouTube video. His presentation on reinforcement learning can  be found here. Practical Reinforcement Learning Practical Reinforcement Learning by Coursera covers the foundations of RL methods: value\/policy iteration, q-learning, policy gradient, etc; using deep neural networks for RL tasks; state of the art RL algorithm; and teaching neural networks to play games. Deep Reinforcement Learning The course on GitHub has a series of articles and videos to help you master the skills and architectures to become a deep reinforcement learning expert.The course will help build a strong professional portfolio by implementing agents with Tensorflow and PyTorch that learn to play Space invaders, Minecraft, Starcraft, Sonic the Hedgehog and more.","excerpt":"Artificial Intelligence: Reinforcement Learning in Python is a complete guide to reinforcement learning with stock trading and online advertising applications.","categories":["AI Trends"],"tags":["Reinforcement Learning"],"author_name":"Poornima Nataraj","publish_date":"2022-03-01T11:00:00","publication_year":"2022","word_count":631,"keywords":["data science","machine learning","artificial intelligence","Reinforcement Learning","AI","neural network","PyTorch","ML","chatbots","deep learning","TensorFlow"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","neural network","data science","TensorFlow","PyTorch","chatbots"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-resources-to-learn-reinforcement-learning-in-2022\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10098219,"title":"Why is Everyone Trying to become NVIDIA?","content":"While AMD, Cerebras and Intel are ambitiously working towards taking on chip leader NVIDIA in their own ways, Tenstorrent has emerged as the brand new player to join the opposing force. And they are doing it by partnering with automobile and electronics companies. Canada-headquartered Tenstorrent, a computing company that develops processors and AI-based deep learning processing units, received a $100 million investment from Hyundai and Samsung. Prior to this funding, Tenstorrent had secured $234.5 million in investments with a valuation of $1 billion, and stands as one of the many companies aiming to contest NVIDIA. It secured $30 million from Hyundai, $20 million from Kia, and the remaining $50 million originated from Samsung’s Catalyst Fund along with additional contributions from investors such as Fidelity Ventures, Eclipse Ventures, Epiq Capital, and Maverick Capital, among various others. Tenstorrent is led by Jim Keller, a well-known figure in the semiconductor realm. Playing a pivotal role as a lead architect for AMD K8 microarchitecture and also participating in the design of processors such as Athlon and Apple A4\/A5, Keller’s experience and vision puts him in the forefront for building something that might give NVIDIA a run for the money. Keller said that for individuals seeking to construct a high-performance solution integrated with AI, NVIDIA will dominate a significant portion (60%) of the product’s profit margin. “The problem with the winner-take-all strategy is it generates an economic environment where people really want an alternative.” Tenstorrent not only manufactures its own AI chips but also offers its intellectual property and other technologies to clients interested in creating their own AI chips. The Golden Automotive Industry In 2021, the worldwide market size for automotive chips amounted to $49.8 billion, with a projected growth of $121.3 billion by 2031. This expansion is expected to occur at a Compound Annual Growth Rate (CAGR) of 9.6% from 2022 to 2031. With a positive market on the horizon, chipmakers can reap benefits from the same. Hyundai formed a semiconductor development division in the previous year and announced intentions to integrate Tenstorrent’s technology into forthcoming vehicles under the Hyundai, Kia, and Genesis brands. With this investment, the parties look to develop “optimised but differentiated semiconductor technology” which will aid future AI technology development. Being one of the globe’s major semiconductor contract manufacturers, Samsung’s decision to invest is understandable. The company has said that the funding will be directed towards expediting the company’s product development, advancing the design and creation of AI chiplets, and enhancing its roadmap for ML software. In May, Tenstorrent collaborated with LG Electronics. The partnership aims to develop chips to fuel consumer electronics such as TVs, automotive solutions and data centres. LG will initially adopt Tenstorrent’s AI chip blueprint for its own chip design and Tenstorrent said that they would look at some of the technology that LG has developed. A Long Way before the Catch Up Racing through the computing world, NVIDIA has also branched out to cater to the automobile industry. In collaboration with NVIDIA, Mercedes Benz will work on creating intelligent cockpits and architectures to accommodate AI-driven capabilities for driving. The company had also partnered with other automobile giants such as Jaguar Land Rover and Volvo. NVIDIA, is currently the undisputed leader with an 80 to 95% of market share in the AI computing market. The company even touched a market capitalization of $1 trillion and is continuing to surpass its competitors. Recently, NVIDIA’s leading H100 chip achieved its highest performance to date on a series of MLPerf training benchmarks. (MLPerf benchmarks assess the hardware capabilities of chips by measuring the time required to complete specific workloads). Partnering with CoreWeave and Inflection AI, the GPU established new records across various parameters in a recent test. The trial employed a cluster of 3,584 H100 GPUs hosted on CoreWeave’s platform, interconnected using InfiniBand technology, enabling exceptional performance at both individual and scalable levels. Tenstorrent and Keller’s ambitious plan to take on NVIDIA may be far-fetched, but setting a path through partnerships with strategic players may get them close to its competitor. However, NVIDIA’s smooth trajectory route puts them in another league altogether.","excerpt":"Following in the footsteps of AMD and Cerebras, Tenstorrent aims to challenge chip giant NVIDIA with its recent $100 million investment from Hyundai and Samsung","categories":["AI Highlights"],"tags":["AMD","Cerebras","hyundai","Intel","Jim Keller","ML","NVIDIA","Samsung","Semiconductor India","tenstorrent","Volvo"],"author_name":"Vandana Nair","publish_date":"2023-08-08T11:13:29","publication_year":"2023","word_count":684,"keywords":["tenstorrent","API","RPA","Scala","deep learning","Semiconductor India","R","Jim Keller","hyundai","NVIDIA","Intel","Go","Cerebras","AI","Volvo","ML","CLIP","AMD","Samsung","Aim"],"extracted_tech_keywords":["AI","ML","deep learning","Aim","R","Go","Scala","API","CLIP","RPA"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/why-is-everyone-trying-to-become-nvidia\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10103653,"title":"The Mahabharat of OpenAI","content":"While the OpenAI fiasco came to an end (so, we believe) before Thanksgiving, people were quick to compare it to the Game of Thrones series with Sam Altman being Jon Snow. However, the real Altman was not a Snow who “knew nothing”, but was more like the courageous, skilled, intelligent archer Arjun from the epic Hindu mythology, Mahabharat. Interestingly, there are a number of people from the OpenAI drama that align with the multifaceted and complex characters from the Sanskrit epic. In Mahabharat, when Pandavas fought the Kauravas (both being set of cousins- 5 against 100) in a huge war, there were a number of alliances and powerful characters that took sides. Similarly, OpenAI saw a similar plot (minus the battle and deaths) with allies, advisors and foes coming together. So, who played whom in the Mahabharat of OpenAI? Arjun Arjun is one of the central characters who led the Kurukshetra war from the Pandavas’ end, and can be depicted by none other than Sam Altman. Like Arjun, Altman’s intelligence and capabilities makes him a natural leader, whose importance is incomparable. Similar to the epic, he has even undergone moments of moral dilemma, probably with AI safety, rallying for the regulation of powerful models. Krishna The character that is considered the driver of the war and a literal driver to Arjuna’s chariot during the war, Lord Krishna is believed to be the vital force for all that ensued. While we may not be able to conclude that Microsoft chief Satya Nadella had any part to play that led to the OpenAI fiasco, however, like Krishna, he stood beside Altman as his biggest supporter and best friend. Nadella, who was diplomatic when he mentioned that he would support Altman regardless of whether he stayed at Microsoft or OpenAI, is one of the striking personalities reminiscent of Krishna. He believes that outcomes are driven by people’s actions, and he sees himself as an enabler. Yudhishthir Righteousness, wisdom and loyalty being some of the most important traits of Arjuna’s older brother Yudhishthir, is reflective of co-founder Greg Brockman’s actions. After Altman’s ousting, and Brockman being removed from the board, like a true brother, he decided to quit the company and stand by Altman’s side. Like a just ruler who wanted to give clarity to the people, he was the only one who tweeted, explaining about the firing situation that unfolded at OpenAI’s board. After he was back, he even shared a picture with the whole OpenAI staff, showcasing his solidarity and support to his people. Karna The most debated character in Mahabharat was Karna. A virtuous man who had his reasons to take sides with Kauravas, Karna is a complex character with a sad fate — he was even framed as the bad guy. Ilya Sutskever was part of the four-member board that united to oust Altman, but he had his reasons, which till date is not clarified. Misunderstood but powerful, Sutskever seems to tread Karna’s path. Kunti Displaying resilience, strength, and calmness in the face of adversity, Kunti has been a central character to bring unity. She has always taken the side of her Pandava children and has stood by their side adapting to any situation. However, being Karna’s birth mother too, she was torn to see him side with Kauravas, and always wished that all her children were united. When the ousting happened, CTO Mira Murati was made the interim CEO, however her support for Altman and Brockman did not fade. Bheem, Nakul, Sahadev Like the three Pandava brothers who stood by their brothers’ side, key researchers at OpenAI, Aleksander Madry, Jakub Pachocki, and Szymon Sidor quit the company shortly after Altman and Brockman’s exit. They decided to follow their brothers to pursue their next goal. Shakuni and Duryodhan Oldest brother and Kaurava leader Duryodhan was influenced by his maternal uncle Shakuni to take the path of wrongdoings and foster hatred towards his cousins. Though not comparable to the volumes of the tactics Duryodhan and Shakuni pulled that led to the war, the context of the OpenAI drama, Tasha McCauley and Helen Toner could fit these roles. They were crucial members that led to the ousting and were ultimately removed from the board when Altman returned — similar to the vanquishing of Shakuni and Duryodhan at the end of the war. Yuyutsu Though on Kauravas’s side, Yuyutsu was the only brother of Kaurava who sided with the Pandavas. Adam D’Angelo who was part of the old board that ousted Altman, is also the only member who is part of the newly appointed OpenAI board, thus, indicating his trusted allegiance to Altman and team. Bhishma As the wise, able, loyal and strong central character of Mahabharat, the Bhishma of OpenAI fiasco could probably be the CEO of Airbnb, Brian Chesky. Akin to Bhishma with his profound quotes, after the OpenAI fiasco ended, Chesky said that ‘you learn a lot about people in a crisis,’ possibly referring to the behaviour of different stakeholders during the drama. Though not directly involved in the debacle, his unwavering support to his long-time friend Altman was evident. He even sided with Altman during the discussions that led to reinstating Altman as CEO. Dhritarashtra The blind king and father of Kauravas, Dhritarashtra was unable to rule the kingdom. Like the titular position, Emmett Shear’s presence in OpenAI can be equated to Dhritarashtra’s. The CEO, who lasted for only 72 hours, had to resign from his position when Altman was reinstated. In Mahabharat, after the war ended, Dhritarashtra had to give up his position too. Eklavya Ekalavya, an aspiring archer, who used to self-train by watching Dronacharya teach his students (Kauravas and Pandavas), was asked to cut off his thumb as an offering, so as to prevent him from surpassing the royal sons’ capabilities. Though Elon Musk, a brilliant entrepreneur and visionary, did not have to cut any part of his body, his bitterness after leaving OpenAI in 2018 has been evident over the last year, especially with the meteoric rise the company witnessed with ChatGPT. Though he voluntarily left OpenAI in 2018, citing a conflict of interest with his company Tesla, it was recently reported that Musk left, owing to a failed attempt of taking over OpenAI. He did not miss a chance to take a dig at OpenAI during the Altman fiasco. Drupad Drupad, the king of Panchala and father of Draupadi, who belonged to a royal lineage, fought alongside the Pandavas during the war. He was believed to be a calculative king who built strategic alliances that served both parties’ interests. Similarly, VC and businessman, Vinod Khosla, one of the first investors in OpenAI’s for-profit subsidiary, had a lot at stake during Altman’s ousting. He ensured that he supported Altman and praised his capabilities. Khosla also dissed Emmett on X. Gandhari The capable queen who chose to be blind in solidarity with her husband Dhritrarashtra, ended up siding with the wrong, owing to her ‘blind love’ for her Kaurava children. While Ashneer Grover had no part to play in Altman’s ousting, his blindness to the whole OpenAI situation and incorrect comparison it to his ousting from his company, and even giving two cents to Altman on how the company will go behind his shares (which Altman doesn’t have), would make him fit the bill of the blind queen. He later deleted his tweets, but the damage was done. In this article, the comparison between OpenAI and Mahabharat characters aim for a thematic exploration rather than exact replication.","excerpt":"If the people associated with the recent OpenAI fiasco could be compared with the characters from epic Mahabharat, who would play whom?","categories":["Global Tech"],"tags":["board","Elon Musk","Emmett Shear","Greg Brockman","Ilya sutskever","OpenAI","Sam Altman","Satya Nadella","Vinod Khosla"],"author_name":"Vandana Nair","publish_date":"2023-11-25T17:34:35","publication_year":"2023","word_count":1251,"keywords":["Satya Nadella","Go","ChatGPT","board","Sam Altman","OpenAI","AI","Greg Brockman","RAG","Elon Musk","GPT","Aim","Emmett Shear","Rust","GAN","Ilya sutskever","Vinod Khosla","R"],"extracted_tech_keywords":["AI","ChatGPT","OpenAI","Aim","RAG","R","Go","Rust","GPT","GAN"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/the-mahabharat-of-openai\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":34141,"title":"5 Statistical Methods Used For Data Analysis In Physics","content":"Physicists often involve projects that need programming as well as metadata. Having loads of data and analysing it is a job of the majority of the branches in this field of Physics research. A job of a physicist is nothing but a combination of many fields and data science is most evidently and undeniably one of them. This article is about the different methods that physics professionals that are into research work, often use in their everyday life. 1. Monte-Carlo Simulation: Whenever there is an ample amount of uncertainty involved in a project, Monte Carlo simulation always comes in useful. This is one big example of a statistical probability method used in Physics. The core idea of the method is to use random samples of parameters or inputs to study complex processes. Early days problems in physics like neutron diffusion used to be too complex for an analytical solution. So, the numerical evaluation had to be done and this method proved to be very successful in finding the solution. It is very popular in the fields of simulating quantum field theories, evaluation of integrals. Also used in Lattice QCD and Electroweak theories, simulated annealing, diffusive dynamics, classical dynamics of particles. galaxy dynamics, Hartree-Fock approximation, density functional method and finite element method based methods for partial differential equations. 2. Bayesian Statistics: Bayesian statistics is used very largely in physics. They are used in hierarchical models, posterior samplers and models for complex data types like images and functions. Hierarchical Bayesian models (HBM) are used for problems where individual object parameters and population parameters are unknown, and they, therefore, find applications like for detecting and characterizing galaxies in astronomical images, with variational inference to approximate the posterior, for modelling of supernovae light curves, and to fit cosmic ray data. 3. Statistical image analysis: Data collected from detectors and satellites are huge in number. This data, usual input in a tabular form, is very difficult to visualise. The data may also contain images that have patterns, which are again difficult to decode. This data and non-trivial correlations might go undetected. Proper tools need to be applied to make sense out of them. Among the methods that have been developed to facilitate the exploration of multivariate astronomical data are phylogenetic trees, graphs, chords, and starfish diagrams. Image processing is a common thing to do, be it astrophysics data or material science. These methods commonly involve statistics, and after that theoretical modelling, predicting and forecasting is also done on it. Feature recognition and automated image processing techniques take help from statistics. 4. Error models: Models typically assume errors with the same variance. Predictor variables in regression models are often assumed to be measured without error. But these errors are often the norm in Physics. Further, it is common practice to have estimates of the measurement error variances available through modelling of uncertainties inherent to the detection procedure. In cases where the measurement error is large, explicit errors-in-variables models are necessary to avoid biased estimates, particularly in regression models. These models often have a hierarchical structure in which the true predictor values are treated as parameters. Approaches to solving this problem include the bivariate correlated errors and intrinsic scatter model. 5. Chi-Square Test This is one of the very important tests that several research projects go through, in experimental physics. The test basically tells if there is a significant difference between the observed and the expected value. This test stands as a reliable way to check if the observation of an experiment is correct. It is used to check if the proposed theoretical model really works on the observed data or not. Conclusion The relationship between statistics and physics has dramatically changed since recent times. Physics, and also other fields of science uses statistics to support and make stronger their research. An amalgamation of both these fields is sure proving to be a success story.","excerpt":"Physicists often involve projects that need programming as well as metadata. Having loads of data and analysing it is a job of the majority of the branches in this field of Physics research. A job of a physicist is nothing but a combination of many fields and data science is most evidently and undeniably one […]","categories":["AI Trends"],"tags":["analysis","Bayesian","big data and analytics everyday life","error","regression analysis"],"author_name":"Disha Misal","publish_date":"2019-01-26T07:34:05","publication_year":"2019","word_count":648,"keywords":["data science","Go","analysis","programming_languages:R","AI","R","regression analysis","programming_languages:Go","Bayesian","Ray","VAE","big data and analytics everyday life","error"],"extracted_tech_keywords":["AI","data science","Ray","R","Go","VAE","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/5-statistical-methods-used-for-data-analysis-in-physics\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10098206,"title":"How Amazon’s Bad Investments Led it on Thin Ice","content":"Amazon’s stake in electric vehicle maker Rivian was once worth $27 billion—that was in November, shortly after its IPO. The past years, the EV maker has been constantly struggling with production, parts, and supply chain issues, resulting in Rivian’s stock price plummeting to a 52-week low in 2022. Started originally in 1995 as an online bookstore Amazon has expanded the business alongside its flagship search with the market throughout the years. But of course not every investment has panned out well for the company. With Amazon owning about 17% of Rivian, the investment cost the company annual loss of $2.7 billion last year–its first unprofitable year since 2014 and a record annual loss for the company. The fiasco with Rivian has led Amazon to the list of the 13 biggest electric vehicle business failures in American history. Earlier this year, WSJ reported, Rivians plans to call off the exclusive deal with the tech giant after it ordered fewer-than-expected delivery vans. As part of a deal made in 2019, the online retailer signed on to purchase 100,000 delivery vans from the electric vehicle company, but with Amazon reportedly only meeting the bare minimum of ordering 10,000 vehicles, the two are renegotiating the initially promising agreement. An Investment Cemetery While Bezos finds ways to blast himself into outer space or test-driving a 13-foot robot, he simultaneously also hunts for the next company to strengthen the Amazon arsenal. Over nearly two decades, the tech giant has made over 200 investments and 106 acquisitions. While the number is praiseworthy, the aftermath of the funnelled money has not been so. All the way back to 1999, the ecommerce giant acquired Alexa.com for $250 million. The company providing paid subscription services with SEO and analytics tools which operated for 15 years had to be eventually shut down in 2022. While the company officially did not reveal the reason for discontinuing the service, Semrush reports suggested that there had been a constant decline in its traffic over the years. Alexa Internet was one of the lots whose graves were dug by Amazon. In 2013, Amazon acquired LiquaVista, a liquid display firm, from Samsung, aiming to furnish Kindle e-readers’ battery-efficient screens. However, evolving screen tech quickly made LiquaVista’s “electrowetting” technology obsolete. As a result, in 2018, Amazon chose to shut down the company along with multiple in-house products. Between 2005 and 2015, Amazon invested in, acquired and rebranded 8 emerging technology brands for undisclosed amounts which turned out to be disastrous for the company’s portfolio. All of these firms eventually were shut down within a few years of Amazon’s involvement. In 2005, Amazon bought MobiPocket. But after a decade of no updates, Amazon permanently shut down the website and servers in 2016. In 2006, TextPayMe was rebranded to Amazon Webpay but failed to garner attention and eventually axed in 2014. New-York based Touchco was another company mysteriously shut down after the acquisition, their website and YouTube page were stripped of all content in 2010. Similarly, in 2011 Yap, a speech recognition system acquired by Amazon was discontinued by the Amazon team and started over, leading to the development of Alexa. In 2009, Amazon acquired SnapTell, a visual product search technology but was discontinued as the company merged their technology with the Amazon experience. For reasons varying from slow sales and the rest due to the company’s shift in focus, the company has been long known for axing its innovations. In fact, Bezos famously called the e-commerce giant “the best place in the world to fail” in his 2016 shareholder letter. The Balancing Act Despite the setbacks, the company continues to remain a key force in the technology sector. Even though the company has a long list of failed acquisitions and investments, a number of them have worked in favour of the tech giant. As an example, during 2018, the retail giant introduced “Part-finder,” a mobile feature enabling users to use their device’s camera to capture an item of interest. This feature empowered Amazon to conduct a swift scan, establish a match, as well as guide users towards matching items from its catalogue. Interestingly, it was built using technology developed by Partpic, one of the companies Amazon acquired in 2016. Even though it started as an online bookstore, the company has kept up with the market’s pace all these years. As the generative AI has swept industries globally off their feet, the company is making strides in the race. In a recent episode, the company’s current CEO Andy Jassy during last week’s Thursday’s Q2 2023 earnings call, revealed that “Every single one” of Amazon’s businesses has “multiple generative AI initiatives going right now”. While the list of ‘Killed by Amazon’ ventures continues to grow, the company’s interest evolving with the industry’s paradigm shift seems to be working in its favour as a balancing act.","excerpt":"The tech giant has made over 200 investments and 106 acquisitions but the aftermath is not praiseworthy","categories":["AI Features"],"tags":["Andy Jassy","AWS","jeff bezos"],"author_name":"Tasmia Ansari","publish_date":"2023-08-07T18:30:00","publication_year":"2023","word_count":805,"keywords":["Go","AWS","AI","IPO","innovation","programming_languages:R","programming_languages:Go","Aim","analytics","generative AI","jeff bezos","R","Andy Jassy"],"extracted_tech_keywords":["AI","analytics","generative AI","Aim","R","Go","innovation","IPO","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-amazons-bad-investments-led-it-on-thin-ice\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":54809,"title":"India Budget 2020: Tech Sector Expects Substantial Allocation For AI Projects","content":"The India budget 2020 should include a key emphasis on the implementation of the National Centre of AI as it will prepare organisations for better human-machine partnership and analyse consumer patterns using big data, say tech experts. It’s time for India budget 2020, and the tech industry is keenly looking forward to the announcements that impact their industry. There is a lot on the shoulders of finance minister Nirmala Sitharaman, especially at a time when there has been a slowdown in the economy, and an apparent crunch is taking place in the technology sector. In this article, we take a look at what India’s tech leaders expect from the budget in terms of allocations to deep tech, particularly artificial intelligence AI. While we saw last year that a national programme on AI had been envisioned by the government of India which was meant to be supported by the national centre for artificial intelligence as a hub, along with other centres of excellence, as earlier stated by Piyush Goyal. While India has the opportunity to position itself among front runners on the global AI map with a unique strategy of ‘AI-for-All’, NITI Aayog has decided to focus on five sectors — healthcare, agriculture, education, smart cities and infrastructure, and intelligent mobility and transportation. It’s expected that the government would allocate substantial funds in this direction for the 2020 budget presentation. “We are expecting that the government will initiate further investments in the National Centers for Artificial Intelligence and AI hubs that will help the startup ecosystem garner the benefits of these technologies,” said Kushal Nahata, CEO & Co-founder at FarEye. The Indian AI market size in 2019 has broken out of the trend that we had seen the years before. And, this year, the overall market size witnessed a dramatic jump in terms of market growth, which is an excellent sign for the industry. The artificial intelligence (AI) and analytics industry in India is presently estimated to be $415 million annually in revenues in 2019, up from $230 million in 2018, according to a report by Analytics India Magazine. But, is there a further need to give a boost to the industry given the industry size is still not that large if we compare to the global market? “Given how data and analytics are enabling organisations to bolster productivity through smarter, more effective work, the government must look to incentivise data analytics and AI projects as well as introduce technology-friendly policies and better tax structures for the industry,” says Suganthi Shivkumar, Managing Director, ASEAN, India & Korea at Qlik. India Budget 2020: Allocating For The Global AI Race One of the most significant areas where countries are competing for dominance is the artificial intelligence domain, which is expected to have a considerable impact not only in commerce but also in areas like healthcare, national security, cybersecurity, food security, education industry, etc. Unfortunately, countries like the USA and China are leaving India behind in terms of AI research, AI entrepreneurship and government investment in AI. If you look at China, it has been steadily building an ecosystem to fuel its ambition to become a world leader in AI by 2030 and having the most prominent startups in artificial intelligence. China achieved this feat via a tremendous amount of budget allocations and investments in AI technology in China. India, however, has not announced any budget 2020 allocation for its AI plan. “Taking a step closer to the Digital India initiative, the budget 2020 should include a key emphasis on the implementation of the National Centre of Artificial Intelligence as it will prepare organisations for better human-machine partnership and analyse consumer patterns using big data. Furthermore, building 1 lakh digital villages will open up an untapped marketplace for digital marketers to cater to consumer needs with bespoke services that were earlier consumed only in urban areas,” says Prasad Shejale, founder and CEO of Logicserve Digital. India Budget 2020: Industry Expects Special Funds & Projects For Data Privacy As India prepares for investing in artificial intelligence capabilities, it should, at the same time, lay work on creating personal data protection and privacy laws in place to avoid misuse of data of its citizens. 2019 saw multiple incidents of breach and privacy violations. According to experts, with the rampant growth of data and data analytics in India, privacy becomes paramount. This holds true, particularly for sectors like healthcare, where data privacy is incredibly sensitive. “It is important to note that the amount of data being created and stored across industries continues to grow at unprecedented rates. The government, therefore, must also intervene and play a role in the protection & management of this data,” thinks Deepak Mittal, CEO & Co-founder, TO THE NEW. Given the national data, privacy bill is soon to be implemented, and it makes sense there is a full-fledged need to revamp infrastructure to ensure compliance at all levels within the government along with support to the corporate sector. “Despite the intention of the government to promote data privacy through proposals such as the Digital Information Security in Healthcare Act (DISHA), the reality on the ground is fraught with a lack of systems to collect, de-identify, store or control access for data. I hope that the budget will allocate funds to support data privacy initiatives as well,” states Zoya Brar, Founder and CEO, CORE Diagnostics. AI Startups Program & Incentives Can Play A Tremendous Role In Catapulting India In The AI Race Artificial intelligence is one of the crucial elements of digital empowerment, attention on both industry and public is needed for the hour. Experts and industry leaders say that from an industry perspective, the government should focus on budget 2020 for establishing specialised AI centres and R&D Labs to bring up more product companies and make India a global centre of innovation. Further, leaders say there is a need to incentivisation and aid for startups in artificial intelligence which would help domestic companies to compete against foreign players originating from China. According to Bhavin Turakhia, Founder and CEO of Flock says on budget expectations, “With tech disruption today being a catalyst for the growth of startups, we expect the government to make significant investments in technology hubs to help strengthen emerging technologies such as artificial intelligence, machine learning, and IoT.” The government has addressed compliance procedures and regulatory bottlenecks concerning startups, which is expected to help the ecosystem in developing further core technologies like artificial intelligence. Saurabh Saxena, Country Director – India, Micro Focus agrees and emphasises the importance of promoting startups in the deep tech industry. “Our country has 1600+ deep tech startups that are building innovative solutions. Additionally, deep tech-led by AI has the potential to add over USD 957 billion to India’s GDP by 2035, according to a recent report released by Nasscom. As such, the ability of Indian companies to rapidly scale up continues to be India’s unique competitive advantage. To increase India’s R&D expenditure, there is a need for greater government support and policies that would incentivise R&D. With proposals made on AIRAWAT and allocations made for AI strategy via NITI Aayog, it seems that the government has taken it seriously to allocate funds on boosting artificial intelligence in the country. The industry is expecting more such allocations towards infrastructure development for AI projects. Will there be a distinction made between Digital India and specific AI-related expenditure or will AI come under the vague purview of Digital India initiatives as has been the case for several years? According to industry leaders, it is the time to look beyond the IT and outsourcing industry to boost India’s domestic deep tech capabilities. Dr. Pulkit Mathur, CEO at Queppelin says on Budget, “Deep tech which includes Augmented Reality, Computer vision, Artificial Intelligence will play a big role across sectors in accelerating India’s quest towards becoming a $5tn economy. Driving greater value as well as job-creation will both be impacted by it. We, therefore, would like to see the FM incentivise this sector. Deep tech should be a special focus separate from the IT\/ITeS SEZ sunset clause, which also should be extended beyond 2020.”","excerpt":"The India budget 2020 should include a key emphasis on the implementation of the National Centre of AI as it will prepare organisations for better human-machine partnership and analyse consumer patterns using big data, say tech experts. It’s time for India budget 2020, and the tech industry is keenly looking forward to the announcements that […]","categories":["AI Features"],"tags":["Artificial Intelligence India","budget","china ai investments"],"author_name":"Vishal Chawla","publish_date":"2020-01-29T10:45:00","publication_year":"2020","word_count":1353,"keywords":["Go","API","budget","artificial intelligence","machine learning","AWS","AI","Artificial Intelligence India","Git","computer vision","china ai investments","analytics","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","computer vision","analytics","AWS","R","Go","Git","API"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/budget-2020-tech-industry-expects-substantial-allocation-for-ai-projects\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":27593,"title":"Imarticus Learning Raises $2 Million From CBA Capital, Plans To Go Global","content":"File photo of Nikhil Barshikar, founder and MD at Imarticus Learning. Noted professional education platform Imarticus Learning raised $2 million in Series B funding from education-focused venture capital fund CBA Capital. According to a statement sent by the company, this funding would be used to further Imarticus Learning’s vision of becoming India’s largest professional education institute, as well as creating a global footprint by bridging the gap between the industry and academia. Imarticus Learning spokesperson added that they would be using the funds towards expanding their international operations while also strengthening its stronghold in cities like Mumbai, Delhi, Hyderabad, Bengaluru, Chennai, and Pune, where they are currently present. Nikhil Barshikar, MD at Imarticus Learning, said in a statement, “We are thrilled to partner with CBA Capital and successfully close this round of funding. The funding will not only boost the brand’s status as a leading ed-tech company but also help us take it to the next level of growth. As a well-established investor in the education sector, CBA Capital understands its potential as well as the challenges faced by the industry. That is precisely why the trust bestowed by them assumes greater importance for us as we know and highly regard CBA as an organisation which is an active, and influential player in the domain. Apart from being able to leverage their guidance and mentorship, the capital infusion will enhance our ability to reach out to students all over the world and bring us one step closer to redefining the relevant-skill based training landscape globally.” Imarticus Learning has trained over 30,000 students globally and aims to skill 55,000 students by 2020. They have also worked with over 200 corporates across the globe, helping facilitate management training and reskilling programs for their employees. Now, expanding their presence across both domestic as well as international geographies will be a key focus for the company during the 2018-2019 period. They are planning to establish their institutes Ahmedabad, Kolkata, and Indore.","excerpt":"Noted professional education platform Imarticus Learning raised $2 million in Series B funding from education-focused venture capital fund CBA Capital. According to a statement sent by the company, this funding would be used to further Imarticus Learning’s vision of becoming India’s largest professional education institute, as well as creating a global footprint by bridging the […]","categories":["AI News"],"tags":["analytics education"],"author_name":"Prajakta Hebbar","publish_date":"2018-08-24T06:29:14","publication_year":"2018","word_count":327,"keywords":["API","funding","programming_languages:R","AI","venture capital","RAG","Aim","Rust","GAN","analytics education","R"],"extracted_tech_keywords":["AI","Aim","RAG","R","Rust","API","GAN","venture capital","funding","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/imarticus-learning-raises-2-million-from-cba-capital-plans-to-go-global\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10102023,"title":"Narayana Murthy Urges Youth to Work 70 Hours a Week","content":"In a recent podcast of 3one4 Capital, The Record, Infosys co-founder Narayana Murthy sat with T V Mohandas Pai, to discuss India’s economic journey and the vision for a better future for the country. Giving an example of the post World War-II scenario in Germany and Japan, he said that the youngsters used to work 70-hours a week to rebuild their countries. “For the first time in the last 300 years, India has received some respect in the eyes of the committee of nations and it is the responsibility of every Indian, but more of the youth, to consolidate that respect.” “Performance leads to recognition, recognition leads to respect, respect leads to power,” he said and asked the “wonderful youth of the country” to realise this and work 12 hours a day. He said that this should be done for the next 20 to 50 years so that “India becomes the number one or the number two nation in the world, in terms of GDP.” He highlighted that even if we become bigger than the US, our per capita would still be much lower. Highlighting the impact of technology, Murthy said that technology is crucial for a country like India. Citing companies like Amazon and Byju’s, he said, “Thanks to technology, we are already doing better in the country.” He said that the biggest advantage of technology is that it raises the confidence of human beings. Read: When Narayana Murthy endorsed tech for Indian cricket team Murthy said, “Even direct benefit transfer is an excellent example… corruption has been eliminated… this is an extraordinary benefit that has accrued to our Indian people.” “Technology is a great leveller. It doesn’t matter whether you are rich or poor, educated or not, hail from urban or rural India, are powerful or weak, nothing matters as long as you can use the machine. So technology has to be embraced.” A few months back, in an interview with CNBC, Murthy had said that ChatGPT won’t replace anybody, and is just an addition to human intelligence. “You will use ChatGPT output as a base, and then add your own differentiation.” He said that he is not worried about AI, and it would just take humans from the lower orbit to the upper orbit. “I believe the human mind is the most powerful imagination machine. There is nothing that can beat the human mind.”","excerpt":"So India becomes the number one or the number two nation in the world, in terms of its GDP","categories":["AI News"],"tags":["Infosys","narayana murthy"],"author_name":"Mohit Pandey","publish_date":"2023-10-26T13:26:31","publication_year":"2023","word_count":397,"keywords":["ChatGPT","API","TPU","Infosys","AI","programming_languages:R","GPT","Ray","R","llm_models:ChatGPT","narayana murthy","llm_models:GPT"],"extracted_tech_keywords":["AI","ChatGPT","Ray","TPU","R","API","GPT","llm_models:GPT","llm_models:ChatGPT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/narayana-murthy-urges-youth-to-work-12-hours-day\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":37323,"title":"Why Everyone In The Organisation Has To Be Comfortable Dealing With Data","content":"Big data is ruling in every organisation. It gives the power to make informed, well-evaluated decision making for a concrete strategy in an organisation. But it has to be noted that it’s not just the technical or data science and analytics team that has to deal with data. For a successful functioning of an organisation, it is necessary that everyone in an organisation is comfortable dealing with data. The Challenge The main challenge with introducing data is that not only the people involved in this culture should be comfortable with data, but they should also be well-versed with the technology that comes with it. The whole integrating data science into an organisation is a very collaborative culture, and it is important to pay attention to people’s skills. Another challenge comes with developing a collaborative approach and giving equal importance to everyone in the organisation irrespective of the pay rate because dependencies can be frightening. More often, teams rely on the data which is available at hand and analyse it, ignoring the need to seek more data for a more detailed, complete picture. Getting the hang of a complete data is important which is often gone wasted. What Can Be Done? In order to develop a data culture among teams, it is important to encourage everyone in the organization to be encouraged to adopt the culture. Breaking down organisational and data silos that prevent collaboration and data optimisation would help, in order to achieve that. 1.Educate the employees: Applying smart change management methods to educate and get people on board and subscribing to data governance principles strengthening accountability and transparency also helps. It is important to become data-driven first and then focus on data-optimisation strategies. 2.Every department has to realise the power of data: Adding a data tech talent in the team helps. Data and technology are not the responsibility of a single function in an enterprise, anymore. Sales and marketing people comprehend the power of inculcating a data-driven technical talent in an organisation, competent in solving data challenges. Increasing data literacy among employees is always a good initiative. Data literacy must be among the entire organisation through collaboration and not just a handful of employees. Each employee being data literate will make him feel more comfortable with new technologies. It will also help in increasing their confidence that will come because of upskilling. 3.Upskilling the employees: Upskilling the already available workforce is also a good way to combat this problem. Organising frequent events and internal knowledge sharing platform for the employees helps big time. Maintaining data quality is another concern. 4.Adopt methods for data cleaning: A satisfactory amount of time must be invested in evaluating the data for errors. There can also be an adoption commercially available software to cleanse the data and improving the data quality. Data is very pervasive and hence it reaches beyond people’s professional, designated trades. Everyone in an organisation has a role in data, either as a data owner, a data process owner or simply as a user, and hence, it is important for everyone to be comfortable dealing with it. According to Christine Overby of Postshift, a diagnostic approach provides the benchmark data that helps in the following: Identify the different digital maturity levels across business units, brands, and regions Create transparency for board-level management and reporting Focus the central transformation team to close the biggest gaps or to double down on your strengths Ensure strategic alignment so people work smarter, not harder to achieve digital priorities In Conclusion Organisations keen on making use of data for precise and well-informed decision-making will be successful in the market and will be easily able to adapt to changing conditions and fulfilling customer needs, compared to the data-challenged competitors. Data has become a primary goal and requirement of organisations today in the market and everyone in the organisation must be comfortable with it being used for success in the market.","excerpt":"Big data is ruling in every organisation. It gives the power to make informed, well-evaluated decision making for a concrete strategy in an organisation. But it has to be noted that it’s not just the technical or data science and analytics team that has to deal with data. For a successful functioning of an organisation, […]","categories":["AI Features"],"tags":["Big Data","decision-making","skills","upskill"],"author_name":"Disha Misal","publish_date":"2019-04-05T03:34:51","publication_year":"2019","word_count":653,"keywords":["big data","decision-making","data science","Go","AI","Git","RAG","upskill","analytics","skills","data governance","data quality","Big Data","R"],"extracted_tech_keywords":["AI","data science","analytics","RAG","R","Go","Git","big data","data governance","data quality"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/why-everyone-in-the-organisation-has-to-be-comfortable-dealing-with-data\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10074970,"title":"What to Expect in the Draft National Data Governance Framework Policy, 2022","content":"In February 2022, the government of India introduced a draft titled, “Draft India Data Accessibility and Use Policy 2022”. It was aimed at transforming India’s ability to harness public sector data. Senapathy Kris Gopalakrishnan, the co-founder of Infosys, also believes that non-personal data generated in India should be allowed to be harnessed; he also headed the expert committee for the non-personal data governance framework. However, it was later scrapped, and a new draft was presented in May 2022, titled, “Draft National Data Governance Framework Policy.” Why was ‘Draft India Data Accessibility and Use Policy 2022’ scrapped? The draft found its roots in the National Economic Survey 2019, which reported that the commercialisation of government data might help India achieve its goal of becoming a $5 trillion economy. Chapter 4 in the survey read, “The private sector may be granted access to select databases for commercial use… Given that the private sector has the potential to reap massive dividends from this data, it is only fair to charge them for its use.” With databases like Adhaar, Agristack (Agricultural sector), e-SHRAM (unorganised labourers), Arogya Setu, ABHA (Ayushman Bharat Digital Health Mission), and NDEAR (National Digital Education Architecture), the government has a massive amount of data at its disposal. However, there were several issues with monetising such gathered data. In the absence of comprehensive data protection legislation that ensures responsibility in situations of data breaches or excessive data collecting, the government had attempted to monetise it previously. While the policy aimed for anonymisation, experts said that it lacked legal accountability and independent regulatory monitoring. Apart from this, the policy had other issues as well, which were tackled by the government in good spirits. Draft National Data Governance Framework Policy 2022 The Draft India Data Accessibility and Use Policy 2022, was replaced by Draft National Data Governance Framework Policy which intends to modernise the government’s data collection in order to enhance governance and develop a nationwide Artificial Intelligence (AI) and data-led research and start-up ecosystem. The policy’s stated goals include—among others—advancing digital governance, fostering the digital economy, establishing uniform standards for open-access digital platforms, and standardising data collecting, management, and security practices. As per the policy, the achievement of these goals would make it easier to implement a whole-of-government strategy and also promote improved governance. The policy states that this must be done while protecting people’s privacy, safety, and trust. Applicability As stated in the draft, the policy would be applicable to all government departments and bodies, and the norms and standards established would apply to any data collected and handled by any government body. Furthermore, the policy would apply to all non-personal datasets and data, as well as the platforms, regulations, and standards controlling their access and use by academics and startups. Framework The draft proposes establishing an “India Data Management Office (IDMO)” under the Digital India Corporation (DIC), which reports to MeitY. According to the draft, IDMO is expected to be in charge of drafting regulations, standards, recommendations, and other operations such as working with the government. In addition, by collaborating with the Digital India Start-up Hub (erstwhile MSH), the IDMO is also expected to stimulate and develop the data and Al-based research start-up ecosystems. Problems with this draft There are a few persisting issues that need to be noted—even though several think tanks have voiced concerns about the draft, and the government is highly likely to take those concerns into consideration. For instance, the document does not specify what would happen to other portals with similar goals or how they will combine—even as it emphasises the necessity for a single platform to access all of the government data. For example, the Open Government Data (OGP) Portal is already managed by the National Informatics Centre (NIC) under MeitY. The National Data and Analytics Platform (NDAP)19 of NITI Aayog also offers a platform for accessing such data. States like Rajasthan and Karnataka also have their own versions of these portals, known respectively as the Jan Soochna Portal and the Mahiti Kanaja Portal. It is yet to be decided how the administration plans to bring everything together or whether it would ensure that confusions are avoided. Additionally, no directives have been laid down to prevent duplicity of efforts. The document also includes the development of datasets that would be available to Indian researchers and start-ups. However, it is notable that it adds that IDMO would be in charge of deciding whether the requests for data usage are legitimate and authentic. Additionally, it does not provide any guidelines for the IDMO to adhere to regarding operations, accountability, or transparency. This will obscure the clarity around the operationalisation of IDMO, including the provision of justifications for approving or rejecting requests—which might therefore have a detrimental effect on people’s confidence in its functionality.","excerpt":"Recently, Rajeev Chandrashekhar, minister of state for electronics and IT, said the government is planning to roll out National Data Governance Framework Policy. Let’s revisit how it began and the possible challenges that it could entail.","categories":["AI Features"],"tags":[],"author_name":"Lokesh Choudhary","publish_date":"2022-09-13T13:00:00","publication_year":"2022","word_count":796,"keywords":["Go","artificial intelligence","AI","Git","Aim","analytics","data governance","Rust","GAN","R"],"extracted_tech_keywords":["AI","artificial intelligence","analytics","Aim","R","Go","Rust","Git","data governance","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/what-to-expect-in-the-draft-national-data-governance-framework-policy-2022\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10101588,"title":"AMD Extends PyTorch Support on RDNA-3 For Local ML Training","content":"In a move aimed at enhancing accessibility to AI for developers and researchers, AMD had unveiled ROCm 5.7 last month, the latest iteration of its open software ecosystem for accelerated computing. Now,  the company has now extended its support for PyTorch machine learning (ML) to AMD RDNA 3-based graphics via ROCm 5.7. This means that individuals engaged in working with ML models and algorithms in PyTorch can leverage AMD ROCm 5.7 on Ubuntu Linux. This enables them to harness the parallel computing capabilities offered by AMD Radeon RX 7900 XTX and Radeon PRO W7900 GPUs, which come equipped with 192 dedicated AI accelerators. These accelerators promise up to 2X higher AI performance per compute unit in comparison to the preceding generation. The integration of ML on desktops provides users with a local, private, and cost-effective avenue for supporting ML training and inference, thereby reducing the reliance on cloud-based solutions. Read: AMD’s Attempt to Break NVIDIA’s CUDA AMD is also trying to break NVIDIA’s CUDA monopoly in the AI parallel computing segment. AMD has made a huge leap with its recent bet on Nod.ai, an open source AI software firm. Nod.ai has been known for developing a portfolio of tools and systems for boosting AI applications on AMD hardware. The software bet has been going on at AMD for some time now. In August, the company also announced the acquisition of Mipsology, a French AI startup, which has also been a long-standing AMD partner and developing AI software for the chipmaker, similar to Nod.ai.","excerpt":"AMD is also trying to break NVIDIA’s CUDA monopoly in the AI parallel computing segment.","categories":["AI News"],"tags":["AMD","amd software"],"author_name":"Mohit Pandey","publish_date":"2023-10-17T17:07:14","publication_year":"2023","word_count":253,"keywords":["CUDA","Go","AMD","machine learning","AI","PyTorch","ML","amd software","RAG","Aim","R"],"extracted_tech_keywords":["AI","machine learning","ML","Aim","PyTorch","RAG","CUDA","R","Go","CUDA"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/amd-extends-pytorch-support-on-rdna-3-for-local-ml-training\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10100942,"title":"7 Incredible Features of GPT-4V","content":"When GPT-4 was released in March this year, it was branded as an advanced model with multimodal capabilities. However, multimodality was nowhere in sight. After almost six months, OpenAI released a string of updates last week, with the most notable one being the image and voice feature, making GPT-4 truly multimodal, and finally bringing in the ‘Vision’ feature. Table of contentsIdentifying ObjectsTranscribing TextDeciphering DataProcessing Multiple ConditionsTeaching AssistantUpgraded Coding   Enhanced Design Understanding As showcased by OpenAI’s co-founder Greg Brockman in the demo video for explaining GPT-4 functionalities earlier this year, the varied uses of GPT-4 V(ision) have been put to test and the results have been incredible. Here are some of the amazing features of GPT-4V. Identifying Objects Be it a plant, animal, character or any random object, GPT-4 has been able to correctly identify it from an image. Furthermore, it is able to generate descriptive details about the object. In the below screenshots, ChatGPT has been able to rightly identify the main plant without any descriptive input prompt, and the character ‘Waldo’, respectively (below). Transcribing Text If we input an image with any form of text into ChatGPT Plus, the model is able to transcribe the content from the image. In the screenshot below, the image contains medieval writing from philosopher and writer Robert Boyle’s manuscript. Deciphering Data The model is able to easily read graphs and charts and in all formats, and is able to draw conclusions based on them. In the below screenshot, a bar graph of performance of two models on various competitive exams are shown. Processing Multiple Conditions The model can also comprehend and process images with multiple conditions. For example, in the image below, it has read a set of instructions to arrive at an answer. Teaching Assistant Acting like a virtual teacher, the chatbot can converse with a user to understand topics from various subjects. In the below tweet, a diagram has been elaborately explained as per given instructions. ChatGPT breaks down this diagram of a human cell for a 9th grader.This is the future of education. pic.twitter.com\/L0Za0ZB5rs— Mckay Wrigley (@mckaywrigley) September 28, 2023 Upgraded Coding With ChatGPT Code Interpreter already out there, GPT-4 V(ision) pushes its coding capabilities to another level. By simply uploading an image, you can perform a wide variety of coding-related functions. You can give ChatGPT a picture of your team’s whiteboarding session and have it write the code for you.This is absolutely insane. pic.twitter.com\/bGWT5bU8MK— Mckay Wrigley (@mckaywrigley) September 27, 2023 In the below post, a user has been able to convert an image to a live website. https:\/\/twitter.com\/skirano\/status\/1706823089487491469 Enhanced Design Understanding With a probable flair for design, the chatbot is able to identify various architectural designs. It is also able to suggest design changes based on custom instructions provided by a user.","excerpt":"With GPT-4 finally becoming multimodal, GPT-4V has made ChatGPT a game-changer with its versatile features","categories":["AI Trends"],"tags":["architecture","ChatGPT","Code Interpreter","Design","GPT-4","GPT-4V","GPT4","graphs","Greg Brockman","Multimodal","OpenAI"],"author_name":"Vandana Nair","publish_date":"2023-10-01T17:00:00","publication_year":"2023","word_count":463,"keywords":["Modal","Code Interpreter","BERT","R","llm_models:GPT","ChatGPT","GPT-4V","Greg Brockman","Multimodal","T5","AI","llm_models:BERT","GPT4","OpenAI","GPT-4","architecture","GPT","graphs","Design"],"extracted_tech_keywords":["AI","ChatGPT","OpenAI","R","BERT","GPT","T5","Modal","llm_models:GPT","llm_models:BERT"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/7-incredible-features-of-gpt-4-vision\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10052872,"title":"Swarm Learning &#8211; A Decentralized Machine Learning Framework","content":"Traditional centralized machine learning approaches have limitations with the huge volume of data getting generated at different locations rapidly at the same time. To address this challenge, decentralized approaches are in demand. Swarm learning, a nature-inspired computing technique, is being preferred by solution architects to implement decentralized machine learning systems. In this article, first, we will understand the issues with centralized learning techniques, then we will discuss the benefits of decentralized learning approaches. After that, we will understand how swarm learning can be an effective decentralized machine learning framework that can address the challenges of centralized learning systems. The major points to be covered in the article are listed below. Table of Contents Problem with Centralized\/Traditional Learning Techniques Expected Solutions with Decentralized Learning Techniques Effective Secure Privacy-Preserving Swarm Learning Ant Colony Optimization Particle Swarm Optimization Application of Swarm Learning Let us begin our discussion by understanding the problems faced by centralized learning techniques. Problem with Centralized\/Traditional Learning Techniques There are various areas where we see an increasingly distributed nature of the data because data is increasing rapidly from various sources and this is a challenge for the centralized learning approaches. Let’s take an example of the self-driving car in which sensors such as LIDAR, radar, and vision are generating petabytes of data per day. Such data is generated at an unprecedented speed. and also the locations of the data generated are distributed because there are various sensors in the car. If this kind of data is stored in a centralized location it is formidable. The second challenge to the traditional learning technique where the data is centralized in one location is the privacy and security of the data. Most of the centralized machine learning algorithms require consolidation data which next requires to move to different sources which often leads to leakage of the data and many of the industries like health where the data aggregates the people personal information, behavioural habits in it and dispersion in many sources can invade the person’s privacy. Another limitation of the traditional centralized ML models comes from the data custody domain. We see that most of the companies are the generators of their own data and they use another entity to analyze, clean, and collect that data. These entities have their own infrastructure for performing the above-given action on the data. This separation of the data owner to the data custody entities makes data monopolies where large amounts of data profit go into the different entities ’ pockets. The Unfortunate Tale of China’s FinTech CompaniesAnalytics Centralization: An Efficient StructureLeadership Lessons To Learn From Ants To Thrive In The Post-Pandemic EraDoes Swarm Learning Have An Edge Over Federated Learning?The Benefits of Decentralizing Analytics Talent Expected Solutions with Decentralized Learning Technique As we just discussed the drawbacks and problems of the centralized learning technique. Many of them can be overcome by decentralized learning techniques. To understand it, let’s understand the following attributes of any decentralized machine learning approach: Effective As we know the ML whatever we are using with any kind of data should be accurate, efficient, and also capable of handling the high distribution of the data. Being more focused on the distribution of the data a decentralized learning algorithm needs to be more good at handling highly distributed data. Using their techniques of allocating workloads, connecting peers we should get higher results in terms of accuracy and efficiency. In addition, these learning techniques should have options for dealing with imbalanced data. Secure As we have talked about the security concern regarding the centralized learning techniques. The decentralized ML has the security features which should ensure that only trusted participants can be therein in the procedure of learning. These can be considered as the blockchain technology where only trusted participants can perform changes in the present and it has already been proven that blockchain has high security in comparison to other technology Privacy-Preserving One of the best features of decentralization is that they provide better privacy. Decentralization of ML should give better control to the owner over their sensitive information presented in the data. In the above part, we have seen what are the drawbacks of the centralized learning techniques and how we can overcome them using the decentralized learning technique. There can be various decentralized learning methods and swarm learning is one of them using which we harness the power of distributed data. This type of learning is actually inspired through some natural or biological process. In the next section of the article, we are going to talk about Swarm Learning. Swarm Learning Swarm learning is a part of the artificial intelligence and machine learning studies where the major focus of swarm learning is to evaluate the behaviour of the decentralized system as we have seen in the above part of the article we have various problems with the centralized learning technique on which our most of the traditional models are dependent. A decentralized system may be helpful for us in overcoming the difficulties of centralized learning methods. The basic idea behind this learning is taken from the operational way of the ant colonies or bird flocks. Translate into computationally intelligent systems. Taking the example of the bird flocks where to reach the destination, each bird takes the part and the action by the bird can be considered as its contribution in the global action of the reaching to the destination. In swarm learning, a global system is divided into agents with their environment. And interaction behaviour of the agents leads to the global solution behaviour. In terms of neural networks, we can say that these intersecting agents are the group of the neuron where they are basically working as a team on finding the best place to be. We can say the swarm learning spreads through two major optimization techniques. Ant Colony Optimization(ACO) Particle Swarm Optimization(PSO) Let us understand these techniques in detail. Ant Colony Optimization As we know how the ant colony works on finding food. We often see that ants go in a line or multiple lines towards food. Similarly, the Ant Colony Optimization (ACO) is a technique of solving computational problems or we can say that ACO is a probabilistic technique for finding good paths through graphs so that this can lead to the optimal solution of the problem. And the graph is a structure of a different related set of objects. A combination of artificial aunts and some algorithms which are heuristic methods for solving optimization problems can be considered as the ACO method for optimization of the graphs and can be used in vehicle routing or in internet routing. Image source The above image is a representation of choosing the shortest way between the two points A and B where the graphs are used to make the shortest length way. It is a hazard to define an Ant colony method. According to the different authors, the definition of ACO is different. We can roughly say that the Ant Colony optimization technique is a  populated metaheuristic with each solution represented by an ant moving in the search space. The Below image is a representation of the ant colony finding the shortest way to food from the initial stage to the final stage. Image source Particle Swarm Optimization As we have seen in the Ant Colony optimization we were using the graph for finding an optimal solution of the problem .here this is also an optimization technique that is basically dependent on the methods where we try to improve the candidate solution in an iterative manner to find an optimal solution of any problem for the set of the solutions. The set of candidate solutions can be considered as the particles and moving these particles in the search space gives the solution of the problem where simple mathematical functions are used regarding the position and the velocity of the particle in the movement. The movement of any particle is always influenced by its best position in the search space and by moving the functions that guide the particle to that space. By these procedures, it is expected to move the swarm toward the best solution. Image source The above image is a representation of the PSO where particles are searching for the global minimum of a function. This optimization technique is also a metaheuristic because it also doesn’t make any assumptions about the problem and also using this we can search very large spaces. In the above, we have seen what swarm learning is and how we can approach the different optimization techniques coming under swarm learning. There can be various applications of swarm learning, some of the important applications of swarm learning we are going to discuss in the next section of the article. Application of Swarm Learning By the above intuition, we can say that swarm learning is a way to perform decentralized learning where it can be treated as the remedy for us where we were facing the challenges of centralized learning. Also, there are a lot of benefits of swarm learning. It can be used in a wide range of varieties of the domain. And some of the important domains where it played an important role are listed below: Swarm learning can be used in controlling road traffic. One of the practicable examples is the U.S. military where they are using swarm techniques for controlling unmanned vehicles. Also in space technology, there are uses of swarm learning. The European Space Agency and NASA are using swarm technology for solving problems regarding space technologies where for planetary mapping the NASA is using swarm learning and is developing orbital swarms for self-assembly and interferometry. There are some examples of the usage of swarm learning in the medical domain like in 1992 George A. Bekey discusses the possibility of using swarm learning to control nanobots and these nanobots will be used in the body for the purpose of killing cancer tumours. And al-Rifaie and Aber have used swarm learning to help locate tumours. Swarm learning can also be used in the field of data mining and cluster analysis. Final words Here in the article, we have seen the problems with the traditional centralized learning techniques and how can we overcome these drawbacks or the problem of the centralized learning techniques using the decentralized learning techniques. Swarm learning is a type of decentralized learning technique that has two major approaches as Ant Colony Optimization(ACO) and Particle Swarm Optimization(PSO). We have also discussed these parts along with applications of Swarm learning.","excerpt":"In swarm learning, a global system is divided into agents with their environment. And interaction behaviour of the agents leads to the global solution behaviour. In terms of neural networks,","categories":["AI Trends"],"tags":["AI (Artificial Intelligence)","AI Tool","Data Science","Data Scientist","Deep Learning","Machine Learning","Python"],"author_name":"Yugesh Verma","publish_date":"2021-11-04T10:00:00","publication_year":"2021","word_count":1750,"keywords":["federated learning","Go","artificial intelligence","machine learning","AI","neural network","ML","Machine Learning","Python","analytics","Rust","Deep Learning","Data Science","Data Scientist","R","AI (Artificial Intelligence)","AI Tool"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","neural network","analytics","federated learning","R","Go","Rust"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/swarm-learning-a-decentralized-machine-learning-framework\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10133539,"title":"AI Coding Tools Deliver Multiple Benefits to Developers, but Company Adoption Lags in India: GitHub Report","content":"GitHub recently released new research highlighting how software professionals are putting AI coding tools to use, and the impact it’s having on their teams, code, and organisations. The survey reveals that 81% of developers in India have experienced a perceived increase in code quality due to AI coding tools, however, only 40% say their companies are actively encouraging the use of AI coding tools. Around 66% of respondents in the survey believe these tools will enhance their ability to meet customer requirements, among other benefits. “The potential of AI-driven software development is undeniable; however, individual AI usage isn’t enough. Organisations need to operationalise AI throughout the software development lifecycle to boost collaboration, creativity, and modernisation,” Kyle Daigle, chief operating officer at GitHub, said. The survey of 2,000 software professionals–which includes 500 respondents each across India, the U.S., Brazil, and Germany at organisations with 1,000 or more employees–identified several key benefits that respondents associate with using AI coding tools in software development. Key insights from respondents in India include: Respondents anticipate AI will enhance code security and development efficiency. There is universal anticipation (100%) among survey respondents in India that AI coding tools will improve code security. Notably, the highest expectation of a “significant improvement” across all respondents globally was in India, with 41% expressing this view. Improve ability to meet customer requirements. The majority of respondents in India (66%) expressed optimism about the potential of AI coding tools to moderately improve or significantly enhance their ability to meet customer requirements. This trend was consistent across various industries, suggesting a widespread expectation of benefits from generative AI. Easier to work with new programming languages, and understand existing codebases. A large portion (69%) of respondents in India reported that these tools make it “easy” to adopt a new programming language or understand an existing codebase. Notably, 28% of those in India highlighted that AI coding tools made it “very easy” for them to adopt a new programming language or understand an existing codebase. Test case generation. 99% of respondents in India stated their organisations have experimented with using AI coding tools to generate test cases. The majority (75%) of respondents reported their organisations use AI tools for test generation at least “sometimes.” Proficiency in AI coding tools is seen as a major asset by job seekers. Nearly all respondents in India (99%) believe this skill makes them more attractive candidates, underlining the growing importance of AI across various fields. Notably, 56% in India believe this expertise significantly boosts their employability. Despite the reported benefits of AI coding tools by a high portion of respondents in India, a smaller percentage said their companies are actively encouraging adoption or allowing the use of AI tools, highlighting room for progress. In fact: 99% of respondents surveyed in India reported using AI coding at work at some point. In contrast, a much smaller percentage (40%) of respondents in India said their companies actively encourage and promote AI tool adoption. An additional 39% of respondents in India report that their organisations allow the use of these tools but offer limited encouragement. To maximise the benefits of these tools, organisations should have a roadmap, a clear strategy, and policies in place to ensure wider adoption happens through building trust and driving measurable performance metrics.","excerpt":"Around 66% of respondents in the survey believe these tools will enhance their ability to meet customer requirements.","categories":["AI News"],"tags":["AI coding"],"author_name":"Pritam Bordoloi","publish_date":"2024-08-22T17:55:53","publication_year":"2024","word_count":546,"keywords":["AI coding","AI","Git","RAG","ViT","generative AI","Rust","GAN","GitHub","R"],"extracted_tech_keywords":["AI","generative AI","RAG","R","Rust","Git","GitHub","GAN","ViT","AI coding"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/ai-coding-tools-deliver-multiple-benefits-to-developers-but-company-adoption-lags-github-report\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10087881,"title":"What is Galileo AI?","content":"Former Google engineer Arnaud Benard and former Facebook product designer Helen Zhou have brought Generative AI to user interface design with Galileo AI, which employs natural language to generate UI designs. Galileo AI aims to generate unique, completely editable UI designs from text input at a very high speed. It also has a host of curated AI-generated illustrations and images that you can add to your product. Today, Generative AI takes a big step and comes to user interface design!@helnzhou and I are excited to announce @Galileo_AI : the first AI product that uses natural language to generate UI designs. It lets you design beyond imagination.Early access: https:\/\/t.co\/4KqV1csQ6c pic.twitter.com\/LtzSC1kYw3— Arnaud Benard (@arnaudai) February 8, 2023 “We had conversations with many designers, while they enjoy the challenge of problem-solving and storytelling, it comes with the tedious hours spent on repetitive UI patterns and mundane pixel-pushing,” said Bernard in his tweet. Galileo AI is currently looking to add full-stack, frontend, and AI research engineering talents to join the team. Register for the waitlist here. Surge of Investments in Generative AI In lieu of the success of LLM-based ‘ChatGPT’, several startups have also started experimenting with generative AI to build their own text-generation models. Due to this, venture capitalists have increased their investment in teams and firms that employ LLMs. However, there is a grey area between investors who may be making investments solely for profits or might be genuinely interested in the development of AI. The majority of VCs are following the footsteps of industry titans such as Khosla Ventures, which was the first to invest in OpenAI and is now receiving ongoing funding from Microsoft. Owing to the significant investments made by companies such as Sequoia Capital and Y Combinator in generative AI firms, every company today aspires to be recognised as them.","excerpt":"Galileo AI is on a hiring spree, as well!","categories":["AI News"],"tags":[],"author_name":"Shritama Saha","publish_date":"2023-02-22T17:10:55","publication_year":"2023","word_count":302,"keywords":["Go","ChatGPT","API","OpenAI","AI","GPT","Aim","generative AI","R","startup"],"extracted_tech_keywords":["AI","generative AI","ChatGPT","OpenAI","Aim","R","Go","API","GPT","startup"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/galileo-ai-the-copilot-for-ui-design\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10061766,"title":"Council Post: How to use AI &#038; data analytics to improve decision making in uncertain times","content":"The Oxford dictionary defines technology as ‘scientific knowledge used for practical purposes, especially in industry. Knowledge evolves, metamorphoses or goes obsolete over time. Technology evolves through agile processes and benefits from keeping a tab on incremental performance. Trends and forecasts inform the strategy or plans at the product, solution and organisation level. If anything, the lessons from the last two pandemic years have taught us that uncertainty is the new normal. As per a Mckinsey report, the pandemic has accelerated digital transformation by years. This shift has also unleashed a lot of niche technology – built and nurtured in research and innovation hubs across the globe. Gone are the days when AI conjured up images of Agent Smith, R2D2 or T-800. Now, AI has use cases in both automation and intelligence. The pandemic drove many organisations to adopt remote work. To manage this, AIOps strategies emerged. Gartner coined the term and defines AIOps as  ‘platforms that utilise big data, modern machine learning and other advanced analytics technologies to, directly and indirectly, enhance IT operations (monitoring, automation and service desk) functions with proactive, personal and dynamic insight.‘ Identifying the gaps This leaves many between the proverbial rock and a hard place. How do you take an informed decision at a time of rapid transformation where evolving is synonymous with survival and when the trends are changing at a pace and the relevance of historical data is eroding? Many organisations faced this dilemma. Navigating the ‘new normal’ requires new intelligence, a lot of resilience and new technology in the wake of failing legacy ML forecasts. Many decision-makers had to resist the urge to fall back on gut instinct. In times of uncertainty, the biggest challenge is to resist giving into hype. Decluttering entails understanding business priorities and plugging gaps. Moving away from traditional methods The first step is to always look for evidence or data. If internal data is unreliable, alternate data sources should be used. One such source is dark data – not just unused data but also data representing market conditions, such as the opinion from experts. Gartner analysts have outlined the use of new data and analytics techniques to build models that are resilient and adaptable. The small data approach offers useful insights using fewer data; it includes methods like time-series analysis, few-shot learning, and self-supervised learning. On the other hand, the wide data approach ties together and analyses a variety of small, large, structured and unstructured data. Combined, the approaches help in making the best of available data more effectively. Another approach could be data marketplaces. These are platforms where one can access third party data (public, commercial and even private data if there are sellers) to provide deeper insights, almost like a transactional store to facilitate the buying and selling of data in various domains like business intelligence, demographics, market data, etc. It has been observed that businesses are seeking to augment their internal data sets with external data obtained from such marketplaces. With the growing importance of AI in decision-making processes, the data marketplaces are helping companies reduce effort and costs. The goal is to not just get a siloed view of what’s happening from in house data, but a 360-degree view from multiple sources. Ensemble models are one way to handle data from multiple sources. Another is a composite AI solution that combines multiple technologies. It increases the quality of the solution through better generalisation and abstraction by synergising a mix of machine learning, heuristic systems, rule-based constraints, optimisation, natural language processing and graph techniques. Together, it enhances the system’s learning ability and accuracy of results where a single solution might have failed. Composite AI offers a multi-faceted approach to dealing with multi-dimensional aspects of a business problem. Insights can be extracted from multiple data sources effectively. The final part is to of course test, validate and course-correct as much as is needed. The bottom line is to be equipped with accurate information to make decisions in real-time by striking a balance of technology and the intelligence to bind them. This article is written by a member of the AIM Leaders Council. AIM Leaders Council is an invitation-only forum of senior executives in the Data Science and Analytics industry. To check if you are eligible for a membership, please fill out the form here.","excerpt":"The goal is to not just get a siloed view of what’s happening from in house data, but a 360-degree view from multiple sources.","categories":["AI Features"],"tags":["AI (Artificial Intelligence)","Data Analytics","decision-making","Machine Learning"],"author_name":"Vanitha DSilva","publish_date":"2022-03-01T10:00:00","publication_year":"2022","word_count":719,"keywords":["decision-making","data science","Go","machine learning","AI","ML","Machine Learning","Git","Aim","few-shot learning","analytics","Data Analytics","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","machine learning","ML","data science","analytics","Aim","few-shot learning","R","Go","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/council-post-how-to-use-ai-data-analytics-to-improve-decision-making-in-uncertain-times\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10066846,"title":"Google unveils the world’s largest publicly available machine learning hub","content":"Google I\/O 2022, Google’s largest developer conference, kicked off with a keynote speech from Alphabet CEO Sundar Pichai. The keynote speech had major announcements including the launch of Pixel watch, updates on PaLM and LaMDA, advancements in AR and immersive technology etc. Let us look at the key highlights. Watch the Cloudera Podcasts on machine learning Data Centre & ML hub “Recently we announced plans to invest USD 9.5 billion in data centers and offices across the US. One of our state-of-the-art data centers is in Mayes County, Oklahoma. I’m excited to announce that, there, we are launching the world’s largest, publicly-available machine learning hub for our Google Cloud customers,” Sundar Pichai said. Powered by Cloud TPU v4 Pods, Google Cloud’s ML cluster enmpoers researchers and developers to make breakthroughs at the forefront of AI, allowing them to train increasingly sophisticated models to power workloads such as large-scale natural language processing, recommendation systems, and computer vision algorithms. At 9 exaflops of peak aggregate performance, Cloud TPU v4 Pods is the world’s largest publicly available ML hub in terms of cumulative computing power, while operating at 90% carbon-free energy. Fun fact from #GoogleIO: At 9 exaflops of peak aggregate performance, we believe our new cluster of Cloud TPU v4 Pods is the world's largest publicly available #ML hub in terms of cumulative computing power, while operating at 90% carbon-free energy ↓ https:\/\/t.co\/9q8xzi4fNt— Google Cloud (@googlecloud) May 11, 2022 Sundar Pichai said Google is poised to become one of the first companies to operate all their data centres and offices on 24\/7 carbon-free energy by 2030. Google Tensor The Alphabet CEO said Google Tensor, their custom SOP, would bring AI capabilities like speech recognition to your phone. Combined with Android’s private compute core, it can run data powered features directly on the device to ensure security. https:\/\/twitter.com\/madebygoogle\/status\/1524459517873901568 Google Tensor chip powers Pixel 6, 6Pro and the newly launched Pixel 6a and Pixel watch. Augmented Reality Google is heavily invested in augmented reality and integrating  AR  into many Google products, from Google Lens to multisearch, scene exploration, and Live and immersive views in Maps. Google Live Translation AR Glasses Prototype Google I\/O 2022 ‘The joy that comes with speaking naturally with someone, That moment of connection to understand and be understood, That’s what our focus on knowledge and computing is all about,” Pichai said. Google shows off an early prototype of its AR glasses, setting up the 2025 showdown between Apple, Google and Meta. https:\/\/t.co\/xOgZ1V310M— Mark Gurman (@markgurman) May 11, 2022 Google Wallet Amidst talk of the latest updates and features in Android 13, one of the key announcements made was the launch of Google Wallet. Google Wallet is the one-stop app that can replace your real wallet in all forms and aspects. With fast and secure access to payment cards, identification cards are being transformed into digital cards that can be shared via NFC or QR codes. Phone, keys … Google Wallet. Coming soon to Android, Google Wallet gives you fast, secure access to all your everyday essentials — including payment cards, boarding passes, vaccine cards, event tickets and soon even your driver's license. #GoogleIO pic.twitter.com\/BJXKbykaSw— Google (@Google) May 11, 2022","excerpt":"Google is integrating AR into many Google products, from Google Lens to multisearch, scene exploration, and Live and immersive views in Maps.","categories":["AI News"],"tags":["augmented reality","computing","Google","LaMDA","PaLM"],"author_name":"Kartik Wali","publish_date":"2022-05-12T04:14:00","publication_year":"2022","word_count":529,"keywords":["Go","PaLM","machine learning","TPU","AI","augmented reality","ML","llm_models:PaLM","recommendation systems","Git","computer vision","Google","computing","LaMDA","R"],"extracted_tech_keywords":["AI","machine learning","ML","computer vision","recommendation systems","TPU","R","Go","Git","llm_models:PaLM"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/google-unveils-the-worlds-largest-publicly-available-machine-learning-hub\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10091871,"title":"AWS Makes it Rain Qubits in the Cloud","content":"The potential in the quantum computing market is immense, particularly for tech giants vying to establish a foothold in the space. In February this year, Google made headlines for its development of a new quantum error correction scheme that promises to reduce error rates by 0.1% – a significant margin in this context. The same month saw AWS’ quantum division release a new software framework, PArallel, LArge-scale Computational Electromagnetics (Palace), which can be used to perform large-scale 3D simulations of complex electromagnetic models and enable the design of quantum computing hardware. AWS’s essential offering, Amazon Braket, enables developers and researchers to test their quantum computing algorithms on quantum simulators and quantum hardware. By doing this, they can get a reasonable degree of confidence in the algorithm’s performance before running it on an actual quantum computer, which is expensive and generally takes time to get access to. Kanishka Agiwal, head of service lines, India & South Asia at AWS, mentions that the company has partnered with five companies – IonQ, Rigetti, OQC, Xanadu, and QuEra – to provide like a single pane of glass through which developers can port algorithms in a quantum computer and run them. Each of these partners offers a different type of quantum computer on Amazon Braket. One specialises in superconducting quantum computing, another in trapped-ion computing, and a third in neutral atom, and finally in photonic quantum computing. All of them cater to different problem statements. Some of the problem statements include finding efficient routing algorithms for logistics, simulating ammonia molecules (used in fertilisers) to predict how they disintegrate or evolve, and modelling the behaviour of photovoltaic receptors on solar panels to capture light effectively. “Machine learning on classical computer has attempted to solve these problems, but it doesn’t achieve a fully optimal solution,” said Agiwal. Production use cases There are a lot of applications now of quantum computing that are either in test, trial or actually in the production phase. For example, AWS itself put out a practical production use case of Quantum Key Distribution (QKD), an encryption-decryption system that immediately alerts the sender and receiver if someone intercepts the communication channel. With De Beers, a diamond mining company, AWS is producing artificial diamonds to help quantum networks. In the case of a classical network like optical fibre, light typically passes through the cable from point A to point B. But, after that, a repeater is needed to amplify the signal – otherwise noise comes in and the signal slowly dies out. However, in a quantum network, traditional repeaters don’t work because photons change their state, making it difficult for regular repeaters to identify and repeat incoming signals. In this case, diamonds, or repeaters made of diamonds, can essentially repeat the signal further. As more such experimentation takes place on different quantum computers, companies working on quantum today will see a significant uptick in usage and automatically in revenue as well. Quantum on cloud Quantum computers are typically either shared on a time-sharing basis or simulated using high-performance computing with pseudo-quantum circuits, as not everyone can afford to have a quantum computer on-premises. However, what happens with the cloud is that access to quantum computers is democratised. On the cloud, individuals will be able to access various types of computers for different experiments. In addition, utilising quantum simulators and quantum hardware on Amazon Braket enables developers to test algorithms at a low cost before deploying them on quantum computers, thereby increasing the likelihood of success. Agiwal notes that in many cases, it may be necessary to switch back and forth between classical and quantum computing when testing. For instance, if there is noise in the output from a quantum computer, one might want to compare its variance with that of a classical computer. A system that can simulate both types of computing and provide results for both would be a huge game changer. “Therefore, I think, the cloud is the way to go for quantum computing, and that’s the future, which is why we see a large number of players providing quantum computers on the cloud,” said Agiwal. Measuring progress But, with a lot of conversation around quantum still in the air, how do companies measure real progress? According to Agiwal, the proof is really around how we reduce noise in the system, and how we bring in additional proof of concepts or commercial deployments like QKDs, quantum repeaters, and other NP hard problems. “Some of these will still be community developer-led, and a few will be more hardware-related, which we will work with our partners to advance further,” he said. Community-led advancements in quantum computing include initiatives such as AWS releasing a problem on GitHub for the community to solve. One such problem involves how can a salesman who is trying to visit 4 to 40 places in a day achieve the task without repeating the same places. The problem becomes increasingly complex as the number of locations increases from 40 to 400, requiring more efficient solutions. AWS Quantum in India Quantum Computing Applications Lab (QCAL), a collaboration between the Ministry of Electronics and Information Technology (MeitY) and Amazon Web Services (AWS) is an attempt to drive better quantum computing enablement in the Indian ecosystem. A first-of-its-kind initiative in the country, it provides the scientific, academic, or even developer community with access to a quantum computing developing environment. It is now running its second cohort. AWS is also working with industry partners like TCS, Mphasis, and Fractal to build their own centre of excellence and leverage Amazon Braket to solve some real-world problem statements. Meanwhile, as part of the skilling initiatives, AWS is working with two key partners – QpiAI and Mahindra University to develop the talent pool in quantum technologies. The target audience for these courses are essentially undergraduate and postgraduate students, as well as working professionals looking to upskill themselves.","excerpt":"AWS’s essential offering, Amazon Braket, enables developers and researchers to test their quantum computing algorithms on quantum simulators and quantum hardware","categories":["Global Tech"],"tags":["AWS","Quantum Computing","quantum key distribution"],"author_name":"Ayush Jain","publish_date":"2023-04-20T11:00:00","publication_year":"2023","word_count":978,"keywords":["Quantum Computing","quantum key distribution","Go","machine learning","TPU","AWS","AI","cloud_platforms:AWS","Git","RAG","GitHub","R"],"extracted_tech_keywords":["AI","machine learning","RAG","AWS","TPU","R","Go","Git","GitHub","cloud_platforms:AWS"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/aws-makes-it-rain-qubits-in-the-cloud\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":42737,"title":"Near, A Location-Based AI Startup Raises $100M To Boost Its Market Presence","content":"Singapore headquartered startup, Near, has raised $100 million from a single backer, Greater Pacific Capital (GPC) in a recent round of funding. It has offices in San Francisco, New York, London, Bengaluru, Tokyo and Sydney. Believed to be one of the biggest rounds ever in analytics-based marketing technology, the data analytics firm aims at expanding and adding more people with the funding amount. Founded in 2012, Near has built an interactive, cloud-based AI platform called Allspark. It helps in creating anonymised, location-based profiles of users based on information collected from various sources such as telecom player, data partners, Wi-Fi providers and others to generate consumer insights. Allspark, the flagship product by Near works in over 20 countries. It boasts of clients such as WeWork, MetLife, Mastercard, News Corp and others. Near claims to be one of the world’s largest source of intelligence on people and places, processing data from over 1.6 billion monthly users across 44 countries. The Near Platform powers data-driven marketing and enrichment offerings through a suite of SaaS products. The users of the platform can leverage audience, spatial, retail, among other data in a privacy-led environment. As the company grows, Near plans to sell its services in India soon, while scaling up in other markets Australia, New Zealand and the US. It plans on expanding its team size, which currently consists of around 40 people in its Bengaluru office with a total of around 100 across all the teams. Some of the other investors in Near are Sequoia Capital, JP Morgan Private Equity Group, Cisco Investments and Telstra Ventures.","excerpt":"Singapore headquartered startup, Near, has raised $100 million from a single backer, Greater Pacific Capital (GPC) in a recent round of funding. It has offices in San Francisco, New York, London, Bengaluru, Tokyo and Sydney. Believed to be one of the biggest rounds ever in analytics-based marketing technology, the data analytics firm aims at expanding […]","categories":["AI News"],"tags":["data enrichment ai"],"author_name":"Srishti Deoras","publish_date":"2019-07-17T12:03:09","publication_year":"2019","word_count":262,"keywords":["API","AI","ETL","data-driven","data enrichment ai","RAG","Aim","analytics","GAN","R","startup"],"extracted_tech_keywords":["AI","analytics","Aim","RAG","R","API","ETL","GAN","data-driven","startup"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/near-a-location-based-ai-startup-raises-100m-to-boost-its-market-presence\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10080633,"title":"Stability AI Releases Stable Diffusion 2.0","content":"Since 2022 was the year of generative AI, users were able to generate from text to anything. The year isn’t even over yet, and Stability AI has announced the open-source release of Stable Diffusion 2.0 on Thursday. Recently, NVIDIA entered the generative arena with its text-to-image model called ‘eDiffi’ or ‘ensemble diffusion for images’ in competition with Google’s Imagen, Meta’s ‘Make a Scene’ and others. eDiffi offers an unprecedented text-to-image synthesis with intuitive painting along with word capabilities and instant style transfer—compared to the open-source text-to-image DALL.E 2 and Stable Diffusion—generating results with better synthesis quality. Moreover, Amazon Web Services (AWS) is set to begin offering access to generative algorithms Bloom and Stable Diffusion in Sagemaker Jumpstart—the company’s service for open-source, and deployment-ready algorithms—which is well known in the generative AI space. Read: Meet the Hot Cousin of Stable Diffusion, ‘Unstable Diffusion’ What’s new? Stable Diffusion 2.0 delivers better features and improvements, compared to the original V1 release. The new release includes robust text-to-image models that are trained on a new encoder called OpenCLIP—developed by LAION and aid from Stability AI—improving quality of the generated images. The models in the release are capable of generating images with default resolutions of 512×512 pixels and 768×768 pixels. Check out the release notes of Stable Diffusion 2.0 on GitHub. Moreover, the models are trained on the subset of the LAION-5b dataset—which was created by the DeepFloyd team—to further filter adult content using the dataset’s NSFW filter. Images produced using Stable Diffusion 2.0 (768×768 image resolution) Unveiling Upscaler Diffusion Models Other than the generative model, the 2.0 version includes an Upscaler Diffusion model which enhances image resolution or quality by a factor of 4. For example, below is an image where a low-resolution generated image (128×128) has been upscaled to a higher resolution (512×512). The company said that by combining this model with their text-to-image models, Stable Diffusion 2.0 will be able to generate images with resolutions of 2048×2048, or even higher. Source: (Left) Low-resolution image (128×128); (Right) High resolution image (512×512) produced by Upscaler Diffusion Model Moreover, the depth-to-image model—depth2img—extends all prior features of image-to-image option from the V1 for creative applications of the model. With the help of the existing model, this feature would infer the depth of an input image to generate new images using text and depth information. Input image used to produce several new images With the help of a depth-to-image model, users can apply the feature in new creative applications, delivering results that may look different from the original—still preserving the depth and coherence of the original image. Depth-to-Image preserves depth and coherence of the original image With features such as upscaling capabilities to higher resolution and depth2img, Stability AI believes that the 2.0 version would be the foundation to create new applications—enabling an endless explosion of the creative potential in AI. Why it matters The original CompVis-led Stable Diffusion V1 had changed the nature of open-source AI generative models, spawning over hundreds of other innovations around the world. With one of the fastest growing climbs to 10,000 GitHub stars in any software, the V1 rocketed through 33K stars in less than two months. Source: A16z and Github The original Stable Diffusion V1 release was led by the dynamic team of Stability AI’s Robin Rombach, Patrick Esser from Runway ML and CompVis Group of LMU Munich’s Professor Dr. Björn Ommer. The older version was built on their prior work along with Latent Diffusion Models—receiving support from Eleuther AI and LAION. The blog read that Rombach is now leading the advancements with Katherine Crowson to create the next generation of media models.","excerpt":"The company claims that the second version of Stable Diffusion is the foundation to create new applications, leading to an endless explosion of the creative potential in AI.","categories":["AI News"],"tags":["stable diffusion AI","stable diffusion download"],"author_name":"Bhuvana Kamath","publish_date":"2022-11-24T12:45:37","publication_year":"2022","word_count":603,"keywords":["Go","stable diffusion download","AWS","AI","R","ML","Git","stable diffusion","generative AI","CLIP","GitHub","stable diffusion AI"],"extracted_tech_keywords":["AI","ML","generative AI","AWS","R","Go","Git","GitHub","CLIP","stable diffusion"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/stability-ai-releases-stable-diffusion-2-0\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10160839,"title":"NVIDIA Launches Cosmos, a Platform to Develop World Foundation Models","content":"At CES 2025, NVIDIA unveiled Cosmos, a platform built to speed up the development of physical AI systems, including autonomous vehicles and robots. The platform includes generative world foundation models (WFMs), video tokenisers, guardrails, and an accelerated data processing pipeline to help developers create and refine AI models with reduced reliance on real-world data. Cosmos is available under an open model license on Hugging Face and the NVIDIA NGC catalogue. Fully optimised NVIDIA NIM microservices will follow, with enterprise support provided through the NVIDIA AI Enterprise software platform. Speaking at CES, NVIDIA CEO Jensen Huang said, “The ChatGPT moment for robotics is coming. Like large language models, world foundation models are fundamental to advancing robot and AV development, yet not all developers have the expertise and resources to train their own. We created Cosmos to democratise physical AI and put general robotics in reach of every developer.” The Cosmos models can generate physics-based videos using inputs such as text, images, and sensor data, enabling their use in applications like video search, synthetic data generation, and reinforcement learning. Developers can customise the models to simulate industrial environments, driving scenarios, and other specific use cases. NVIDIA also introduced NeMo Curator, an accelerated video processing pipeline that can process 20 million hours of video in 14 days, and Cosmos Tokeniser, a visual data compression tool. “Data scarcity and variability are key challenges to successful learning in robot environments,” said Pras Velagapudi, chief technology officer at Agility Robotics. “Cosmos’ text-, image-, and video-to-world capabilities allow us to generate and augment scenarios for a variety of tasks that we can use to train models without needing as much expensive, real-world data capture.” Major robotics and transportation companies, including Agile Robots, XPENG, Waabi, and Uber, have begun adopting Cosmos for their AI development. Uber CEO Dara Khosrowshahi said, “Generative AI will power the future of mobility, requiring both rich data and very powerful compute. By working with NVIDIA, we are confident that we can help supercharge the timeline for safe and scalable autonomous driving solutions for the industry.” In addition to Cosmos, NVIDIA introduced the Llama Nemotron large language models and Cosmos Nemotron vision language models, developed for enterprise use in sectors including healthcare, finance, and manufacturing.","excerpt":"Cosmos is available under an open model license on Hugging Face and the NVIDIA NGC catalogue.","categories":["AI News"],"tags":["NVIDIA"],"author_name":"Siddharth Jindal","publish_date":"2025-01-07T11:36:57","publication_year":"2025","word_count":371,"keywords":["ChatGPT","Hugging Face","synthetic data","AI","Scala","microservices","GPT","generative AI","foundation models","NVIDIA","R"],"extracted_tech_keywords":["AI","generative AI","foundation models","ChatGPT","Hugging Face","microservices","R","Scala","GPT","synthetic data"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/nvidia-launches-cosmos-a-platform-to-develop-world-foundation-models\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10070691,"title":"Why open-ended conversational AI is a hard nut to crack","content":"The ‘intelligence’ of AI is growing all the time. And AI has many forms, from Spotify’s recommendation system to self-drive cars. AI utilises natural language processing (NLP) to deliver natural and human-like language. It mimics humans and generates human-like messages by analysing commands. That said, it is still challenging to create an AI tool that understands the nuances of natural human languages is hard. The open-ended conversation is even more complex. That is why some of the latest projects like LaMDA and BlenderBot do not have any commercial applications built on them and are kept purely for research purposes. Making sense of open-ended conversational AI Conversational AI refers to technologies, like chatbots or virtual agents, with which users can interact with. Compared to other AI applications, they use large volumes of data and complex techniques to help imitate human interactions, recognise speech and text inputs and translate their meanings across various languages. Currently, there are two types of conversational AI – goal-oriented AI and social conversational AI. The goal-oriented AI typically focuses on short interactions to help with user goals such as booking a cab, playing a song, shopping online, etc. The social AI engages in a conversation more as a companion, aka open-ended conversation. Open-ended conversational AI models need to be able to handle a large number of possible conversations, which can be difficult to design for. As a result, open-ended conversational AI can be more expensive and time-consuming to develop than other types of AI. Conversational AI has principal components that allow it to process, understand, and generate outputs. The quality of outputs can be subjective, considering the level of expectation the AI meets. Lack of Applications At the Google I\/O conference last year, Sundar Pichai said the company is looking for ways to cater to developer and enterprise customers. In addition to that, he said the language model for dialogue application (LaMDA) would be a huge step forward in natural conversation. While big tech companies and open source initiatives are trying their best to push conversational AI solutions and products to enterprise customers, most are limited to supporting or assisting teams to a certain level. As a result, generalised platforms with probabilistic language models have largely remained in the periphery use cases. Another challenge is that due to its nature, open-ended AI possesses a higher risk of being misused. Open-ended conversational AI – Timeline First announced at Google’s I\/O 2021 event, the tech giant had described LaMDA as its ‘breakthrough conversation technology.’ A recent controversy that brought LaMDA back in the limelight was when Google’s AI engineer Blake Lemoine claimed that LaMDA is sentient. He published some excerpts of his conversations with Google’s LaMDA, a Transformer-based language model. LaMDA’s conversational skills have been developed for years now. Like other recent language models, including BERT and GPT-3, it is built on Transformer. A few other reports too suggest that LaMDA has passed the Turing test and, therefore, it is sentient. That said, the Turing test can not be considered the ultimate test for a model to possess human intelligence. This has been proved through several experiments in the past. From a gimmicky automated responder, chatbots and voice assistants like Google Echo, Cortana, Siri, and Alexa have been valuable sources of information. In July 2021, researchers at Meta released BlenderBot 2.0, a text-based assistant that queries the internet for up-to-date information about movies and TV shows. BlenderBot is entirely automated — it also remembers the context of previous conversations. But the system suffers from issues like a tendency to spout toxicity and factual inconsistencies. “Until models have a deeper understanding, they will sometimes contradict themselves. Similarly, our models cannot yet fully understand what is safe or not. And while they build long-term memory, they don’t truly learn from it, meaning they don’t improve on their mistakes,” Meta researchers wrote in a blog post introducing BlenderBot 2.0. Before that, in 2015, around the height of the chatbot craze, Meta (formerly known as Facebook) launched an AI-and-human-powered virtual assistant called M. Powered by M, select Facebook users could access ‘next-generation’ assistant through Messenger that would automatically place purchases, arrange gift deliveries, make restaurant reservations, and more. Reviews were mixed as media house CNN noted that M often suggested inappropriate replies to conversations — and Meta decided to discontinue the experiment in 2018. Setbacks As machines learn from humans, they also internalise our flaws – moods, political views, tones, biases, etc. But, as they can’t evaluate good from bad on their own (as of now), it usually results in a no-filter response from the machine. A lack of understanding of words, emotions and views acts as a huge barrier to achieving human-like intelligence. It is still challenging for the current corp of conversation AI tools to understand user emotions, detect and respond to offensive content, understand multimedia content beyond text, comprehend slang and code-mixed language, etc.","excerpt":"Very few projects are used in applications for just having a two-way conversation with a user that is meaningful and engaging.","categories":["AI Features"],"tags":[],"author_name":"Tasmia Ansari","publish_date":"2022-07-10T13:00:00","publication_year":"2022","word_count":814,"keywords":["Go","TPU","AWS","AI","chatbots","BERT","NLP","GPT","Aim","R"],"extracted_tech_keywords":["AI","NLP","Aim","chatbots","AWS","TPU","R","Go","BERT","GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/why-open-ended-conversational-ai-is-a-hard-nut-to-crack\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10161334,"title":"India&#8217;s 1st Private Semiconductor Facility to Come Up in Andhra Pradesh","content":"The Andhra Pradesh government has signed an MoU with several companies, including Japan-based YITOA Microtechnology and India’s Indichip Semiconductors Ltd., to establish India’s first private semiconductor fabrication facility. The project will be built at the Orvakal Mega Industrial Park in Kurnool district at an investment of ₹14,000 crore. The agreement was signed in Hyderabad in the presence of TG Bharath, Andhra Pradesh minister for industries, commerce and food processing, and IT minister Nara Lokesh. The facility will produce silicon carbide (SiC) chips, which are critical for energy-efficient technologies such as electric vehicles and renewable energy. “This state-of-the-art facility will initially produce 10,000 wafers per month, scaling up to 50,000 within 2-3 years,” said Lokesh in his post on X. The facility is expected to be operational within two-and-a-half years. It will boost India’s semiconductor manufacturing capabilities and contribute to the growth of the country’s tech ecosystem. The Government’s Vision “This investment aligns with India’s Atma Nirbhar Bharat vision and reinforces Andhra Pradesh’s ability to attract cutting-edge industries through innovative policies and robust infrastructure,” Lokesh expressed. He emphasised that the project plays a significant role in positioning Andhra Pradesh as a global semiconductor hub, “demonstrating the ability to attract industries through innovative policies, world-class infrastructure, and a focus on nurturing talent.” Bharath credited chief minister N Chandrababu Naidu for ensuring balanced development across the state. Expressing his views regarding this collaboration, he said, “This is a testament to the growing confidence of international and domestic companies in Andhra Pradesh. The semiconductor sector will not only stimulate industrial growth but also create significant employment opportunities for the local population.” The investment will position India as a competitive player in the global semiconductor supply chain at a time when the world is seeking alternatives to reduce reliance on a single region for semiconductor production. Orvakal as an Industrial Hub Bharath highlighted that the project will create thousands of jobs and attract further industrial investments to the Orvakal Industrial Park region. To ensure its success, the state will provide the required land, infrastructure, and an enabling ecosystem. The Orvakal Mega Industrial Hub, which covers 2,621 acres, received approval from the central government and funding of ₹2,786 crore in September last year. The hub is projected to employ approximately 45,000 people and attract a range of industries, including a proposed 300-acre drone manufacturing and research unit. Bharath stated, “This is not just an investment in Andhra Pradesh; it’s an investment in India’s future as a leader in high-tech innovation.”","excerpt":"The MoUs signed bring in an investment of ₹14,000 crore for the state.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","AI in India","semiconductor"],"author_name":"Sanjana Gupta","publish_date":"2025-01-13T14:38:10","publication_year":"2025","word_count":414,"keywords":["Go","funding","AI in India","programming_languages:R","AI","innovation","semiconductor","programming_languages:Go","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","R","Go","innovation","funding","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/indias-1st-private-semiconductor-facility-to-come-up-in-andhra-pradesh\/","complexity_score":3,"technical_depth":7,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":26431,"title":"Airbus Signs Deals With 3 Indian Startups, Concludes Third Season Of Accelerator Programme","content":"Flying high on the Indian startup wave, Airbus, the international pioneer in the aerospace industry, on Friday signed contracts with three Indian startups. These deals were signed by Airbus subsidiaries NAVBLUE and Aerial with three startups accelerated at Airbus Bizlab India. NAVBLUE signed up Bengaluru-based Stelae Technologies to enhance aeronautical data services quality and consistency while enabling faster introduction to the market of the next generation fully data-driven and connected EFB (Electronic Flight Bag) solution. They also extended its agreement with EFLIGHT with the aim of providing a comprehensive service solution to the Indian business aviation market. The third agreement was signed between Airbus Aerial and Navi Mumbai-based Airpix for a joint go-to market in India to provide geo-analytics solutions and imagery services in the country. Making the announcement at a press meetup held in Bengaluru, the agreements also marked the culmination of the third season of the Airbus accelerator programme. Bruno Gutierres, Global Head, Airbus BizLab, said at the event, “For the last couple of years I have been convinced with the potential and the quality of entrepreneurs in India. I am pleased to see that all these partnerships between Airbus and Indian startups are proving me right.” Gutierres also commended the robust ecosystem provided by the Indian government for a thriving startup environment. The Season 3 of Airbus BizLab’s start-up acceleration programme TAKE OFF 2018, also seeks to bolster the Indian government’s ‘Make in India’ and ‘Startup India’ initiatives. Shruti Singh, Director of Department of Industrial Policy (DIPP), said, “Indian Government’s initiatives about the new tech sector are not about creating jobs — but about creating job creators.” She added, “India’s vision is to support startups and work with MNCs to take Indian startups global and also help the country innovate.” Suresh Prabhu, Union Minister of Commerce & Industry and Civil Aviation, was set to attend the event but had to cancel his plans at the last minute. However, he sent in a video to encourage Airbus BizLab and their programme. Siddharth Balachandran, Airbus BizLab Leader, India, added that they have accelerated 16 startups so far, out of which 5 are “success stories”. Among these 5, three  signed contracts with Airbus (and their subsidiaries) today. He also said that these startups had generated over $1.5 million so far. Exciting discussions with AirbusBiz Lab India around global market access opportunities for Indian startups in the Aerospace sector and making India a global startup hub for aerospace startups @airbusbizlab @SidRdy pic.twitter.com\/coXjilzTn7 — Startup India (@startupindia) July 6, 2018 Bizlab is a part of Airbus’ innovative strategy to bring together start-ups and Airbus intrapreneurs to work and speed up the transformation of their innovative ideas into valuable businesses. It has a community of over 2,000 B2B start-ups from India, South East Asia and Israel. Through its six-month acceleration programme for start-ups and internal projects, Airbus Bizlab gives access to a large number of coaches, experts and mentors in various domains that help start-ups and internal projects speed-up the transformation of their ideas into valuable businesses. As part of the TAKE OFF programme, Airbus BizLab will sponsor two travels to Europe for Indian start-ups in a period of six months as well as fund up to £50,000 to them for the demonstration of proof of concept. “NAVBLUE believes in India’s potential for multiple reasons. One is India’s proven capacity for innovation within the IT sector, which is demonstrated perfectly by the start-ups present at the Airbus’ BizLab, with whom we work very closely. Another is the potential of the aerospace industry; India has the world’s fastest-growing domestic airline market,” said Fabrice Villaumé, Chief Strategy and Innovation Officer at NAVBLUE. “We are very excited to work with startups and SMEs in India to help power this growth, and bring value to the Indian aviation industry with our integrated solutions,” he added.","excerpt":"Flying high on the Indian startup wave, Airbus, the international pioneer in the aerospace industry, on Friday signed contracts with three Indian startups. These deals were signed by Airbus subsidiaries NAVBLUE and Aerial with three startups accelerated at Airbus Bizlab India. NAVBLUE signed up Bengaluru-based Stelae Technologies to enhance aeronautical data services quality and consistency while enabling […]","categories":["Deep Tech"],"tags":["indian economy","Startups","suresh prabhu"],"author_name":"Prajakta Hebbar","publish_date":"2018-07-13T09:13:54","publication_year":"2018","word_count":638,"keywords":["Go","programming_languages:R","AI","data-driven","indian economy","innovation","RAG","Aim","analytics","suresh prabhu","Startups","R","startup"],"extracted_tech_keywords":["AI","analytics","Aim","RAG","R","Go","data-driven","innovation","startup","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/airbus-signs-deals-with-3-indian-startups-concludes-third-season-of-accelerator-programme\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10094719,"title":"Genpact and Google Cloud Join Hands to Accelerate AI Adoption for the Enterprise","content":"Genpact has announced that it is working with Google Cloud to help businesses accelerate artificial intelligence (AI) strategies, including taking advantage of generative AI’s adoption to drive actionable business insights. The partnership will focus on helping enterprises accelerate their Generative AI journeys, building generative AI for custom enterprise LLMs, business processes, and augmenting operations, Genpact said. Additionally, the practice will house a dedicated team of data scientists, data engineers, and domain experts focused on the rapid development of generative AI capabilities on the Genpact Cora platform which is already integrated with more than 250 enterprise ecosystems handling more than 20 million transactions a month. Genpact is continuing to invest in its AI capabilities by establishing a generative AI practice teaming with Google Cloud. This partnership will leverage Google Cloud’s advanced GenAI capabilities to cater to enterprises in key industries, including consumer goods, retail, life sciences, healthcare, hi-tech, and financial services. Through this partnership, Genpact and Google Cloud will accelerate their shared clients’ cloud, data, and analytics modernization journeys to address the evolving needs of today’s businesses. “We are seeing increased demand from our clients to leverage generative AI to improve productivity, minimise risk, and provide enhanced employee and customer experiences. This expanded partnership highlights the strategic role that partner ecosystems will play in the massive enterprise AI market. “Combining Google Cloud’s suite of products with Genpact’s domain, process, and data expertise provides a differentiated strategic partnership to our clients, driving competitive advantage and meaningful growth in a fast-evolving world,” Katie Stein, Chief Strategy Officer and Global Business Leader, Enterprise Services and Analytics, Genpact, said.","excerpt":"This partnership will leverage Google Cloud’s advanced GenAI capabilities to cater to enterprises in key industries, including consumer goods, retail, life sciences, among others","categories":["AI News"],"tags":[],"author_name":"Pritam Bordoloi","publish_date":"2023-06-08T12:55:30","publication_year":"2023","word_count":264,"keywords":["Go","API","GenAI","artificial intelligence","AI","RAG","ViT","generative AI","analytics","R"],"extracted_tech_keywords":["AI","artificial intelligence","analytics","generative AI","GenAI","RAG","R","Go","API","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/genpact-and-google-cloud-join-hands-to-accelerate-ai-adoption-for-the-enterprise\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":40042,"title":"India Among Top 9  In Worldwide Rankings In Terms Of AI Specialists","content":"A startup called Element AI in its annual report reported that India stands at rank 9 in terms of number of AI specialists working in the field. The report results were based on the study that the startup did on the number of authors who published papers at international conferences in the year 2018. It involved analyzing the author’s careers and breaking down the distribution by location. According to this report by Canadian-based startup, 22,400 people worldwide were regarded as top talent of AI. Among these AI experts, 50 percent belonged to the US with a number of 10.295, followed by China, which had 2,525 AI experts. The report further stated that rank 3 and 4 are for the UK, Germany and Canada with a number of 1,457, 935 and 815 respectively. According to the report, India stands at number 9 in these AI specialist standing, with the number of AI specialists being 555. Image credits: asia.nikkei.com The top rank in this list have many academic institutes with programs on artificial intelligence. They therefore have a much greater number of people skilled to do research in the field. India, on the contrary, lacks the opportunities in formal education in data science but is slowly trying to encourage the adoption of AI in education institutes. The CBSE, for example, this year onward has an artificial intelligence as an elective subject for its 9th grade classes. Institutes like the Indian Institute of Technology (IIT) Hyderabad this year launched a full-fledged Bachelor of Technology (B.Tech) programme in AI, thereby becoming the first Indian educational institution to do so. It is also most likely the third education institute in the world, after Carnegie Mellon University and the Massachusetts Institute of Technology (MIT), to have a  full-fledged B.Tech. programme in AI. IIIT Hyderabad is another educational institute that introduced its popular executive programs on AI and ML and Blockchain and Distributed Ledger Technologies across India and it was previously available only in Hyderabad and Bangalore. It is not just the educational institutes, but also the Indian government that shows tremendous support to artificial intelligence. NITI Aayog, moved a cabinet note to establish a cloud computing platform called AIRAWAT (Artificial Intelligence Research, Analytics and knoWledge Assimilation plaTform). This was done along with a bunch of research institutes.","excerpt":"A startup called Element AI in its annual report reported that India stands at rank 9 in terms of number of AI specialists working in the field. The report results were based on the study that the startup did on the number of authors who published papers at international conferences in the year 2018. It […]","categories":["AI News"],"tags":["AI (Artificial Intelligence)","Government","iit hyderabad","NITI Aayog"],"author_name":"Disha Misal","publish_date":"2019-06-03T08:20:19","publication_year":"2019","word_count":381,"keywords":["data science","Go","artificial intelligence","AI","cloud computing","ML","iit hyderabad","Government","RAG","analytics","AI (Artificial Intelligence)","R","NITI Aayog","startup"],"extracted_tech_keywords":["AI","artificial intelligence","ML","data science","analytics","RAG","cloud computing","R","Go","startup"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/india-among-top-9-in-worldwide-rankings-in-terms-of-ai-specialists\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":19057,"title":"WiFi Analytics Startup i2e1 Raises $3 Million In Series A Funding","content":"Delhi-based WiFi analytics startup i2e1 (information to everyone) has raised $3 million in Series A funding led by impact investment firm Omidyar Network. A statement released by i2e1 said that some of organisation’s existing investors, Auxano Ventures and 3one4 Capital (a company backed by Mohandas Pai), participated in the round. i2e1 currently services over 3,000 locations across 75 cities in the country. “Using data collected from user sign-ins at brick and mortar establishments, i2e1 creates an analytics intelligence suite similar to Google Analytics, tailored for offline businesses. This strengthens defensibility for offline retailers in comparison to their online counterparts who collate real-time intelligence on their users,” i2e1 spokesperson said in a statement. “At i2e1, our goal has always been to build a connected ecosystem where information is available to everyone at the right time and place. Retailers in India are slowly realizing that Wi-Fi as a service not only increases customer delight, but can also be used to understand customer behaviour patterns when utilized effectively,” said Satyam Darmora, co-founder, i2e1. The company plans to use the funds to scale its existing operations and launch a mobile application. It aims to reach 10 million users across India in the next few years, said the statement. Siddharth Nautiyal, investment officer, Omidyar Network, said, “Despite advancements in infrastructure from telecom companies, significant gaps exist across Internet connectivity and affordability for the masses, paving the way for increased Wi-Fi proliferation to enable ubiquitous and low-cost access… i2e1 has the potential to become the de facto gateway to the Internet for millions, empowering offline retailers with rich user insights and smarter decision making.” Incubated in IIT Delhi, i2e1 had raised $500,000 in seed funding in December 2015 from growX ventures, Pai, Rajan Pandhare, Debasish Mitter and Singapore Angels.","excerpt":"Delhi-based WiFi analytics startup i2e1 (information to everyone) has raised $3 million in Series A funding led by impact investment firm Omidyar Network. A statement released by i2e1 said that some of organisation’s existing investors, Auxano Ventures and 3one4 Capital (a company backed by Mohandas Pai), participated in the round. i2e1 currently services over 3,000 […]","categories":["AI News"],"tags":[],"author_name":"Prajakta Hebbar","publish_date":"2017-11-16T10:12:19","publication_year":"2017","word_count":293,"keywords":["Go","API","AI","RPA","Aim","ViT","analytics","GAN","R","startup"],"extracted_tech_keywords":["AI","analytics","Aim","R","Go","API","GAN","ViT","RPA","startup"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/wifi-analytics-startup-i2e1-raises-3-million-series-funding\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":32014,"title":"Da AI Code: Mysterious 600-Year-Old Manuscript Finally Decoded","content":"The trials and errors that went into decoding the 15th-century Voynich manuscript could give The Da Vinci Code (Doubleday, 2003) a run for its money. The Voynich manuscript is a cryptic manuscript that nobody has been able to read till now and was considered to be an alien language. But recently a computer scientist and NLP expert from the University of Alberta claimed that the mysterious document is actually written in ancient Hebrew and that he has deciphered the riddle with the help of artificial intelligence. Deciphering ancient manuscripts has never been an easy job for historians and cryptographer. But now, the Canadian researchers have simplified the tough initial process of decoding the language framework of the manuscript by using AI. About Voynich Manuscript Since the 17th century, the manuscript is believed to be the work of the 13th-century English alchemist Roger Bacon. Voynich manuscript is exactly a 234-page manuscript, filled with seemingly coded language and inscrutable illustrations. The whole manuscript majorly consists of pictures of plants and women along with multiple astrological diagrams. It reportedly has seven separate sections: botanical, astronomical, cosmological, zodiac, biological, pharmaceutical, and recipes. AI Says The Manuscript Is Written In Hebrew Greg Kondrak, a computer scientist and an NLP expert at the University of Alberta along with his graduate student Bradley Hauer knew that it was a perfect task for an AI and set out to use computers for decoding the ambiguities in human language using the Voynich manuscript as a case study. The first step was used in addressing the language of origin, which was exquisitely enciphered on hundreds of delicate vellum pages with accompanying illustrations. Kondrak and Hauer used samples of 400 different languages from the “Universal Declaration of Human Rights” to systematically identify the language. The scientists initially hypothesised that the Voynich manuscript was written in Arabic. After running their AI algorithms, it turned out that the most likely language was Hebrew. Kondrak and Hauer hypothesised the manuscript was created using alphagrams, defining one phrase with another, exemplary of the ambiguities in human language. Assuming that, they tried to come up with an algorithm to decipher what type of scrambled text. “It turned out that over 80 per cent of the words were in a Hebrew dictionary, but we didn’t know if they made sense together,” said Kondrak in the university press release. The scientists turned to Google Translate. “It came up with a sentence that is grammatical, and you can interpret it,” said Kondrak, “she made recommendations to the priest, man of the house and me and people. It’s a kind of strange sentence to start a manuscript but it definitely makes sense.” According to Kondrak, human language is used to communicate with other humans, but computers don’t understand this language, because it’s designed for people. Also, there are so many ambiguous meanings that humans don’t even realise. Natural language processing can help computers make sense of human language. These Canadian researchers also accepted that they have found the language of the manuscript and are working towards applying more varied AI algorithms that they have developed for other ancient scripts along with help of ancient Hebrew historians and linguists to fully decipher this mystery book. Outlook There have been multiple attempts to decode the Voynich manuscript. A self-proclaimed “prophet of God” in 2011 claimed that he had decoded the book. Researchers in 2014 argued that the illustrations of plants in the manuscript could help decode the text’s strange characters. Deciphering the manuscript can be done only when the origin of its language known which is considered to be the first step, which AI did for Voynich manuscript. Despite the breakthrough, the book still remains untranslated completely. The full meaning of the text will need the involvement of various historians and linguists of ancient Hebrew and AI \/NLP experts to help translate the whole text to unlock its true meaning. It’s just another accomplishment for AI technology which has transformed multiple industries to have initiated another scientific angle to decode the Voynich manuscript completely.","excerpt":"The trials and errors that went into decoding the 15th-century Voynich manuscript could give The Da Vinci Code (Doubleday, 2003) a run for its money. The Voynich manuscript is a cryptic manuscript that nobody has been able to read till now and was considered to be an alien language. But recently a computer scientist and […]","categories":["AI Features"],"tags":["AI and NLP","AI Revolution","Google Translate"],"author_name":"Martin F.R.","publish_date":"2018-12-23T05:40:18","publication_year":"2018","word_count":672,"keywords":["Go","artificial intelligence","programming_languages:R","AI","Google Translate","AI and NLP","AI Revolution","programming_languages:Go","BERT","NLP","Aim","llm_models:BERT","R"],"extracted_tech_keywords":["AI","artificial intelligence","NLP","Aim","R","Go","BERT","llm_models:BERT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/da-ai-code-mysterious-600-year-old-manuscript-finally-decoded\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10040656,"title":"New ARCore Capabilities Announced At Google I\/O 2021","content":"Google’s augmented reality platform for developers, ARCore, has created compelling experiences for use cases across gaming, navigation, e-commerce and social media since its launch in 2018. ARCore–used to develop AR experiences for Android–powers more than 850 million smartphone users and is installed in over one billion devices globally. To enable seamless integration of the virtual and real, Google has been building out ARCore skill tree, empowering developers with foundational capabilities around realism, perception and asynchronous interaction. On the second day of the Google I\/O, Google announced its new capabilities in ARCore. The session was hosted by Rajat Paharia, Product Lead of ARCore; and Jared Finder, Engineering Manager of ARCore. Realism in AR The realism branch of the ARCore skill tree includes foundational AR Capabilities around motion tracking, lighting, and depth. Realism Capabilities allow putting digital content into the physical world so that users cannot tell the difference between what’s real and what is not. Motion tracking enables virtual objects to stay in place even when the camera is moving. “In the last year, we made AECore’s motion-tracking even more robust with an up to 36 percent decrease in CPU usage and dramatic reduction in the number of tracking resets, in many cases eliminating them completely,” Rajat said. Lighting: ARCore’s environmental HDR understands the lightning in the scene so that the virtual object(s) can look and behave exactly like the real ones. Depth API: It uses a single standard smartphone camera to provide a depth map, making experiences more immersive. “Your virtual objects can now interact with the real world, rather than being stickers on the screen,” Rajat said. Under Depth API, ARCore offers particle effects, interactivity, occlusion and lighting effects. Depth perception Perception enables one to detect certain objects in a scene and help the user augment them. It includes capabilities such as instant placement, augmented images, ML models, augmented faces, and ARCore’s latest capability, Raw Depth API. Instant Placement enables users to place virtual content without first having to pan and scan to find a plane. “With Instant Placement, we have seen a 16 percent increase in placement success, and 26 percent decrease in placement time,” Jared said. Augmented Images API tracks moving images keeping the attached content intact. For users who want to create face effects, ARCore offers an Augmented Faces API. It provides high-quality, 468 point 3D mesh to allow users to add effects. This feature is available on both Android and iOS. ML models: ARCore allows users to integrate their own ML models into AR experiences. Asynchronous interaction This enables users to build AR experiences where users can interact with places and with each other, across space and time. The capabilities include Cloud Anchors, and Recording and Playback. Cloud Anchors allows users to annotate the world with AR content, allowing users to build their own constantly evolving layers on top of the real world, and create location-based experiences. These annotations can be experienced in both Android and iOS. New releases Google announced the launch of Raw Depth API, which enables developers to build more accurate ‘measurement, reconstruction and interaction apps’ by giving users access to more detailed point cloud representation than its usual standard depth API. The tech giant also announced the launch of the ARCore Recording and Playback API. “For your developer velocity, you can now record a video and then play it back through ARCore,” Jared said. He said users or businesses looking to build an experience in a shopping mall need not go to the mall every time they want to test a change. They can record their visit once and test from the comfort of their own desk. Enterprises can enable post-capture AR experiences for users, he added.","excerpt":"Google’s augmented reality platform for developers, ARCore, has created compelling experiences for use cases across gaming, navigation, e-commerce and social media since its launch in 2018. ARCore–used to develop AR experiences for Android–powers more than 850 million smartphone users and is installed in over one billion devices globally. To enable seamless integration of the virtual […]","categories":["Global Tech"],"tags":[],"author_name":"Debolina Biswas","publish_date":"2021-05-23T15:00:00","publication_year":"2021","word_count":616,"keywords":["Go","API","programming_languages:R","AI","ML","programming_languages:Go","Git","ViT","R"],"extracted_tech_keywords":["AI","ML","R","Go","Git","API","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/new-arcore-capabilities-announced-at-google-i-o-2021\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":18916,"title":"Understanding The Deep Learning Advantage in Medical Imaging And How Tech Giants Are Rushing to Capitalize On It","content":"Image Source: Fastcompany AI advances in healthcare are nothing new. What’s new is Deep Learning models diagnosing diseases with greater accuracy and research papers that claim diagnosis as good as a physician? From Deep Learning models that can detect suicidal tendencies to a Deep Learning algorithm developed by AI scientist Sebastian Thrun and his Stanford University team that can detect cancerous skin lesions as good as a leading dermatologist, DL has taken over diagnostic evaluations. So, what’s driving the explosion of Deep Learning in healthcare. Broadly speaking, there are three main areas that have fueled AI growth: a) huge volumes of healthcare data (thanks to rapid digitization of medical records & EHR); b) the rise of GPUs that puts the power of deep learning in the hands of data scientists and researchers; c) running Deep Learning models hadn’t been very cost-effective, but now they are a fraction of that cost. Deep Learning, in particular CNN plays a big role in medical imaging According to Dr Dave Chanin, Founder and President of Insightful Medical Informatics, the value of deep learning systems in healthcare comes only in improving accuracy and increasing efficiency. He attributed the current interest of applying deep learning in healthcare to web giants Google and IBM that are leveraging unsupervised learning techniques to yield accurate results. Also, the explosion of DL is not really seen in more consumer-facing applications, but in the imaging and informatics wherein algorithmic learning is applied to swathe of medical data that also includes images. The interest can also be attributed to Convolutional Neural Networks (CNN) that have been used in the field of computer vision for decades and now its deep architecture that enables multiple levels of abstraction is being leveraged for medical imaging analysis. So why are CNN ubiquitous in medical image analysis and have become the go-to methodology of choice for analyzing medical images. The multi-stream architecture of CNN can accommodate multiple sources of information or representations of the in-put in the form of channels presented to the input layer. Since segmentation is the most common task in medical image analysis, CNNs can be applied to “every pixel in an image, using a patch or subimage centered on that pixel or voxel, and predicting if the pixel belongs to the object of interest”, this research paper notes.  The other two major advantages of CNNs which are pre-trained on natural images, show good results, sometimes even challenging the accuracy of trained physicians in some tasks. Researchers have gone a step ahead to show that CNNs can be adapted to leverage intrinsic structure of medical images. Besides the hardware development, the wide availability of open source packages and the GPU-computing libraries such as OpenCL, CUDA has fueled the popularity of CNNs in medical imaging. We also have the huge volumes of training data to build Deep Learning based medical imaging software. According to IBM researchers, medical images nearly account for at least 90 percent of all medical data, which makes it the largest data source in the healthcare industry. Industry Impetus — M&As & Partnerships abound Last year, chipmaker Intel hit the headlines with its purchase of deep-learning startups Nervana Systems and Movidius, the latter promised the rise of low power chip that can run neural networks. Movidius is a California based vision processor startup has a  mobile-friendly system that makes it feasible to run neural networks in more places.  Their Fathom USB sticks can run visual neural nets and will be extremely useful for researchers at universities. Meanwhile, Nervana Systems want to put Deep Learning in the cloud. Another company making huge strides in healthcare is IBM which acquired Merge Healthcare, a leading provider of medical image handling and processing, interoperability and clinical systems in 2015 to tackle the problem of a lack of medical image data. For IBM, Merge’s technology platform which are used at more than 7,500 U.S. healthcare sites, as well as many of the world’s leading clinical research institutes and pharmaceutical firms to manage a growing body of medical images gives it access to a ready repository of training data. Now part of IBM’s Watson Health business unit, the acquisition helped the company bolster its ability to analyze and cross-reference medical images against 315 billion data points that already exist in the Watson Health Cloud, including lab results, electronic health records, genomic tests, clinical studies and other health-related data sources. Today, IBM is making great efforts in diagnosing cancer and tracking tumor development. M&As aside, leading healthcare companies are forging partnerships to bolster development. San Francisco-based cloud based medical imaging startup Arterys tied up with GE Healthcare to combine its quantification and medical imaging technology with GE Healthcare’s magnetic resonance (MR) cardiac solutions. The startup provides a better visualization and quantification of blood flow inside the heart, alongside a comprehensive diagnosis of cardiovascular disease. In an industry-first, the startup also received an FDA clearance to leverage deep learning and cloud computing in a clinical setting with Arterys Cardio DL that provides automated, editable ventricle segmentations based on conventional cardiac MRI images that are as accurate as segmentations performed manually by experienced physicians. India Focus Bangalore based startup Sigtuple is disrupting Indian healthcare market with cutting edge medical imaging software It is true that AI is going to have the biggest impact on diagnostics and will help turn physicians into specialists who will end up looking at filtered cases instead of generalists who have to attend to low-priority cases. In a way, Deep Learning will help pave the way for AI-aided medical care. From DL trained models to diagnose diabetic retinopathy to vetting tumors, DL-based solutions are expanding the scope of radiology by predicting diseases at human-level accuracy. India is not far behind in this curve. Bangalore-based AI startup SigTuple, co-founded by Apurv Anand, Rohit Kumar Pandey and Tathagato Rai Dastidar in 2015, leverages Deep Learning to improve diagnostic. The startup leverages recent advances in Deep Learning space for processing and analysing visual data. The startup has built algorithms which learn from medical data, and help doctors by automating disease screening and diagnosis. They enable access to these algorithms through low cost diagnostic devices and a cloud based intelligent platform. In fact, the startup gained a lot of traction amongst investors and media for its powerful intelligent screening. Another Bangalore and San Francisco-based startup Qure.ai, hailed for having the most promising technology in India. The startup is leveraging Deep Learning technology to medical imaging data, thereby reducing physician’s workload and giving them more face-time with patients. The startup is building a deep learning system which will diagnose abnormalities from medical images. In a blog, the startup notes that most of the deep learning models are classification models that predict a probability of abnormality from a scan. However, just the probability score of the abnormality doesn’t amount much to a radiologist if it’s not accompanied by a visual interpretation of the model’s decision. The startup has made great strides in automatically identifying tumours and lesions in brains from MRI scans. In fact, the Qure.ai team was placed third in Brain Tumour Segmentation (BRATS) challenge at MICCAI 16. The startup is also taking steps to develop brain segmentation algorithms also known as multi-atlas segmentation algorithm. Early Deep Learning Pioneers Medical imaging startups have gained a lot of traction and there is a frenetic M&A activity in this space. According to Signify research, the total investment in medical imaging AI startups since 2014 is pegged at $167 million. Around half of the startups are building applications for multiple body areas while the rest are focused on specific clinical specialties, such as pulmonology, breast and cardiovascular. Some of the leading AI medical imaging startups are Pixyl, Viz, Zebra Medical Vision, VoxelCloud, AIdoc and Aidence among others. However, the pioneer in deep learning medical imaging is Australian company Enlitic that leverages proprietary algorithms to quickly and accurately improve healthcare diagnosis. Founded in 2014, this medical imaging company is slotted as an early pioneer in using Deep Learning for tumor detection, and its algorithms have been used to detect tumors in lung CT scans. According to the CEO Jeremy Howard, the young company has also developed an algorithm that can identify relevant characteristics of lung tumors with a higher accuracy rate than radiologists.","excerpt":"AI advances in healthcare are nothing new. What’s new is Deep Learning models diagnosing diseases with greater accuracy and research papers that claim diagnosis as good as a physician? From Deep Learning models that can detect suicidal tendencies to a Deep Learning algorithm developed by AI scientist Sebastian Thrun and his Stanford University team that […]","categories":["IT Services"],"tags":["medical image processing companies"],"author_name":"Richa Bhatia","publish_date":"2017-11-10T05:25:51","publication_year":"2017","word_count":1377,"keywords":["CUDA","Go","AI","neural network","cloud computing","computer vision","RAG","Aim","deep learning","R","medical image processing companies"],"extracted_tech_keywords":["AI","deep learning","neural network","computer vision","Aim","RAG","cloud computing","CUDA","R","Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/understanding-deep-learning-advantage-medical-imaging-tech-giants-rushing-capitalize\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":39228,"title":"Behind The Code: How Shrirang Deshpande Took A Career U-Turn To Follow His Biggest Passion — Data Science","content":"This month we are starting a developer series called Behind The Code, wherein we tap into the developer community in India and find out how these professionals, who are the backbone of our IT landscape began their journey and how they at the forefront of major innovations. In our first column, we speak to Shrirang Deshpande, a Mechanical Engineer by qualification, whose career took a turn in his mid-20s after he developed an interest in automation and AI\/ML. This interest led him to co-found an analytics startup, where he worked as a Business Analyst for two years before he moved to head the Data Science team for an online learning platform. For Deshpande, even though he was moving in the right direction in his career path, the initial transition wasn’t really easy. “That was the most difficult decision I have made up till now since understanding the data science ecosystem for a mechanical engineer was indeed challenging. When I look back, I certainly feel it was one of the best decisions I have ever made,” says Deshpande, who currently works as a senior associate for Gurugram-based Tyroo. Presently, his responsibilities include product development, decision making and process improvement. He is also closely working with his management to device business opportunities in the space of digital marketing using artificial intelligence and machine learning A Rough Start With no prior knowledge in the field of Data Science, when Deshpande started his career, he had to unlearn four years of his undergraduate learning to start from the scratch of data science; which meant learning basic aspects like statistics, coding or domain expertise. Due to this, he made it a point to regularly update his knowledge on the latest libraries, algorithmic implementation and its use cases. According to Deshpande, his four years in the industry has been that of constant learning, “I still remember when I first started studying I really liked the concepts like regression, clustering but understanding the use cases was a little difficult since you need to have some domain or functional knowledge to build these,” he says. “Even writing codes for feature engineering and data cleaning was tough. But one thing I believed that practise and solving case studies from Kaggle will surely make this learning process easier and that is how the journey started,” Deshpande added. What is the most important programming language? A man who wears many hats, Deshpande is currently pursuing an executive general management program for young leaders from IIM Bangalore and he is also a visiting faculty at an analytics institute. As an analytics instructor, it is the most common question raised by aspiring Data Scientist every now and then. Deshpande believes that Python is becoming popular among data scientists and software developers due to its ease of deploying ML models, “I think Python is the language to look for AI\/ML. Even though I started my learning from R but certainly the innovations and robust libraries such as TensorFlow, Keras and even the ease of deploying ML models, Python is becoming popular among data scientists and software developers,” Deshpande explains. Essential tools for a data scientist For analytics: Vertical and as well as horizontal expansion of using and learning tools is important, which requires you to be perfect in some tools. For statistics: I think Excel skill is a must have, excellence in either R or Python is important and having intermediate knowledge of SAS would definitely help. For database: in-depth  SQL knowledge is desirable along with basic knowledge of working with MongoDB. From the visualisation stand-point again good Excel skills along with expertise in Tableau or PowerBI. Apart from these tools, knowledge of some Hadoop tools such as Spark or Hive is certainly good to have. What’s in his developer kit For statistics: I have expertise in Excel & R trying to get a strong hold on Python and important python libraries, For visualisation: I use Tableau and for database management SQL. Apart from this I keep on learning and reading about other tools such as Alteryx, PowerBI, HIVE, Spark Scala etc. Word of advice Practise, focus and self-belief. “In India, this field is opening up quickly, I think in the next three years we will experience exponential growth in hiring and upcoming data science professionals who have good coding skills, stats knowledge and domain expertise have a really good chance of riding this wave,” Deshpande says in an optimistic tone. What’s In Store For The Future Learning is a life-long and Despande exemplifies just that. He intends to hone his analytics as well as his business skills in the coming few years. “I think I have a lot to learn, right from image recognition to advance speech recognition since I feel these technological advancements will be major forces driving innovation in next 5 years,” Deshpande explains","excerpt":"This month we are starting a developer series called Behind The Code, wherein we tap into the developer community in India and find out how these professionals, who are the backbone of our IT landscape began their journey and how they at the forefront of major innovations. In our first column, we speak to Shrirang […]","categories":["AI Features"],"tags":["AI and automation","Behind The Code","big data for data science","business intelligence career path","Career","Interviews and Discussions","latest technological advancements","ML"],"author_name":"Akshaya Asokan","publish_date":"2019-05-15T11:17:15","publication_year":"2019","word_count":803,"keywords":["data science","artificial intelligence","machine learning","Keras","AI","big data for data science","business intelligence career path","ML","AI and automation","image recognition","MongoDB","latest technological advancements","analytics","Behind The Code","TensorFlow","Career","Interviews and Discussions"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","data science","analytics","TensorFlow","Keras","image recognition","MongoDB"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/behind-the-code-how-shrirang-deshpande-took-a-career-u-turn-to-follow-his-biggest-passion-data-science\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10147914,"title":"Why Do Observability Platforms Even Exist?","content":"At a crucial time when enterprises grapple with the complexities of AI systems and their observability, Middleware, a full-stack cloud observability platform, asserts that enterprise-level companies typically invest rather heavily in observability. According to estimates, these companies spend approximately 30% of their infrastructure budget, roughly between $500,000 and $1 million annually, on this. Busting a few myths and explaining why an on-prem infrastructure would require observability, Sawaram Suthar, founding director of Middleware, highlighted that open-source tools come with hidden costs. “Many assume that open-source tools are a cost-effective solution, but the reality is far more nuanced. While the initial download may be free, its true costs lie in the significant investments of time, resources, and expertise required for implementation, customisation, and ongoing maintenance,” he told AIM. Suthar added that this can be particularly daunting in the context of on-prem infrastructure, underscoring the critical need for comprehensive observability to ensure seamless operations and maximise ROI. The Evolution of Modern Observability Deploying AI models is just the starting point. To maximise their value, we need to understand how they perform in real-world scenarios. “This requires a deeper level of visibility into their performance, including metrics like API calls, model accuracy, and resource utilisation. By gaining this insight, we can optimise our AI systems for better efficiency, cost-effectiveness, and overall performance,” Param Teraiya, AI team lead at Middleware, told AIM. The misconception about observability platforms being unnecessary for on-premise solutions is quickly dispelled by modern realities. “When it comes to on-prem solutions, they are already using their own infrastructure. Even if you are deploying it on-prem, you would be required to measure the relevancy of the answers you would get out of the models and see that your model is not hallucinating,” said Teraiya, emphasising the universal need for robust monitoring. In addressing cost challenges, Middleware has made a strategic decision to opt for Llama over ChatGPT. “Every observability platform struggles with cost, complexity, cardinality, and signal-to-noise ratio. We want to operate at a scale where we also manage the cost,” notes Param. The decision reflects a deeper understanding of scaling challenges in production environments. The traditional approach of using legacy observability tools like New Relic has become prohibitively expensive. Implementing and maintaining open-source observability tools requires significant expertise and effort. Middleware addresses this challenge by offering managed services that reduce costs while providing robust functionality. AI is transforming observability by automating tasks that were previously manual and time-consuming. For example, Middleware uses AI to develop tools like Query Genie, which allows users to query infrastructure data using natural language instead of complex filters or database queries. Teraiya elaborated on this innovation: “Imagine asking your infrastructure questions like you would a colleague. That’s the reality we’re creating with Query Genie.” This approach not only simplifies the user experience but also accelerates problem resolution. Meanwhile, AI is also being used for predictive analytics. Middleware is working on features that can forecast potential issues before they occur. For instance, instead of alerting when CPU usage exceeds 80%, their platform can predict hours in advance when a file system will reach critical capacity. The Shift to Cloud-Native Architectures The industry is undergoing a significant shift toward cloud-native architectures. Gartner predicts that by 2025, 95% of systems will adopt cloud-native approaches. Middleware has embraced this trend by building a cloud-agnostic platform based on open standards like OpenTelemetry. “Whether you use AWS, Azure, or GCP, we just take OpenTelemetry data, which is common for all clouds,” Suthar said, emphasising the importance of flexibility. This approach ensures compatibility across different cloud providers while simplifying implementation for customers. However, challenges remain despite these advancements. One major issue here is the “black box” nature of many AI systems. Teraiya acknowledged this problem and stressed the need for explainable AI: “We always wanted to identify how [a model] predicts or gives a particular output.” Transparency is crucial for building trust in AI-driven observability platforms. Another challenge is scalability. As data volumes grow exponentially, observability platforms must evolve to efficiently handle petabytes of information. Middleware addresses this by using machine learning algorithms for anomaly detection and optimising resource utilisation through intelligent data pipelines. “We recognise that excessive data collection can be a costly burden. That’s why we’re empowering our customers to take control of their data, filtering out noise and focusing on what truly matters. By doing so, they can reduce their observability costs by up to 80% and shift from a ‘collect everything’ approach to a targeted, efficient monitoring strategy that drives real value,” Suthar concluded.","excerpt":"Most observability platforms ingest massive amounts of unnecessary data, and nearly 70% of the information collected provides little value to developers’ day-to-day operations.","categories":["AI Features"],"tags":["AI Observability"],"author_name":"Sagar Sharma","publish_date":"2024-12-27T11:11:44","publication_year":"2024","word_count":754,"keywords":["ChatGPT","machine learning","AWS","AI","Azure","ML","Aim","anomaly detection","analytics","AI Observability","predictive analytics"],"extracted_tech_keywords":["AI","machine learning","ML","analytics","ChatGPT","Aim","predictive analytics","anomaly detection","AWS","Azure"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/why-do-observability-platforms-even-exist\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10007547,"title":"ROC-AUC Curve For Comprehensive Analysis Of Machine Learning Models","content":"In machine learning when we build a model for classification tasks we do not build only a single model. We never rely on a single model since we have many different algorithms in machine learning that work differently on different datasets. We always have to build a model that best suits the respective data set so we try building different models and at last we choose the best performing model. For doing this comparison we cannot always rely on a metric like an accuracy score, the reason being for any imbalance data set the model will always predict the majority class. But it becomes important to check whether the positive class is predicted as the positive and negative class as negative by the model. For this, we make use of Receiver Characteristics Curve – Area Under Curve that is plotted between True positive and False positive rates. In this article, we will learn more about the ROC-AUC curve and how we make use of it to compare different machine learning models to select the best performing model. For this experiment, we will make use of Pima-Indian Diabetes that can be downloaded from Kaggle. What we will learn from this article? What is the ROC-AUC Curve? How does it work?How to compare the performance of different models using the ROC-AUC curve? What is the ROC-AUC Curve? How does it work? It is a visualization graph that is used to evaluate the performance of different machine learning models. This graph is plotted between true positive and false positive rates where true positive is totally positive and false positive is a total negative. The area under the curve (AUC) is the summary of this curve that tells about how good a model is when we talk about its ability to generalize. If any model captures more AUC than other models then it is considered to be a good model among all or we can conclude more the AUC the better model will be classifying actual positive and actual negative. If the value of AUC = 1 then the model will be perfect while classifying the positive class as the positive and negative class as negative. If the value of AUC = 0, then the model is poor while classifying the same. The model will predict positive as negative and negative as positive. If the value is 0.5 then the model will struggle to differentiate between positive and negative classes. If it’s between 0.5 and 1 then there are more chances that the model will be able to differentiate positive class values from the negative class values. How to compare the performance of different models using the ROC-AUC curve? Let us now practically understand how we can plot this graph and compare different model performance. We will first build 4 different classification models using different machine learning algorithms and then will plot the ROC-AUC graph to check the best performing model. We will not quickly import the required libraries and the iris data set. Refer to the below code for the same. from sklearn.model_selection import train_test_split from sklearn.linear_model import LogisticRegression from sklearn.ensemble import RandomForestClassifier from sklearn.tree import DecisionTreeClassifier import numpy as np import pandas as pd from sklearn import svm from sklearn.metrics import roc_curve, auc df = pd.read_csv('pima.csv') print(df) Now we will divide the dependent and independent features X and y respectively followed by splitting the data set into training and testing sets. Use the below code for the same. X = df.values[:,0:8] Y = df.values[:,8] X_train, X_test, y_train, y_test = train_test_split(X, Y, test_size=0.50, random_state=1) We have divided the data into training and testing now we will build for different models for classifying the class i.e whether a patient is diabetic or not. Use the below code to build the respective models. clf1 = LogisticRegression() clf2 = svm.SVC(kernel='linear', probability=True) clf3 = RandomForestClassifier() clf4 = DecisionTreeClassifier() Since we have defined the four different classifiers now we will fit the training data over these and will predict probabilities for testing data. Use the below code for the same. probas1_ = clf1.fit(X_train, y_train).predict_proba(X_test) probas2_ = clf2.fit(X_train, y_train).predict_proba(X_test) probas3_ = clf3.fit(X_train, y_train).predict_proba(X_test) probas4_ = clf4.fit(X_train, y_train).predict_proba(X_test) Now we will compute the ROC curve and AUC score for all these classifiers. Use the below code for the same. fp1, tp1, thresholds1 = roc_curve(y_test, probas1_[:, 1]) roc_auc_model1 = auc(fp1, tp1) fp2, tp2, thresholds2 = roc_curve(y_test, probas2_[:, 1]) roc_auc_model2 = auc(fp2, tp2) fp3, tp3, thresholds3 = roc_curve(y_test, probas3_[:, 1]) roc_auc_model3 = auc(fp3, tp3) fp4, tp4, thresholds4 = roc_curve(y_test, probas4_[:, 1]) roc_auc_model4 = auc(fp4, tp4) print(\"AUC for Logistic Regression Model : \",roc_auc_model1) print(\"AUC for SVM Model:\", roc_auc_model2) print(\"AUC for Random Forest Model :\" ,roc_auc_model3) print(\"AUC for Decision Tree model :\", roc_auc_model4) Since we have got the AUC score now we will plot the roc curve to visualize the performance of all 4 models. Use the below code to do the same. pl.clf() pl.plot(fpr1, tpr1, label='Logistic Model (area = %0.2f)' % roc_auc1) pl.plot(fpr2, tpr2, label='SVC Model (area = %0.2f)' % roc_auc2) pl.plot(fpr3, tpr3, label='Random Forest Model (area = %0.2f)' % roc_auc3) pl.plot(fpr4, tpr4, label='Decision Tree Model (area = %0.2f)' % roc_auc4) pl.plot([0, 1], [0, 1], 'k--') pl.xlim([0.0, 1.0]) pl.ylim([0.0, 1.0]) pl.xlabel('False Positive Rate') pl.ylabel('True Positive Rate') pl.title('Receiverrating characteristic example') pl.legend(loc=\"lower right\") pl.show() We can see from the above graph the svc model captures the highest AUC and can be considered as the best performing model among all the four models. This way we can compute and compare different predictive models. We did this for binary classification, whereas if we want to do the same for multi-class classification models we can again do that. Consider we have three classes X, Y, and Z. So if we are plotting the curve for X class then it would be done as classification of X class against no other class i.e Y and Z. And similarly for other classes. Conclusion In this article, we discussed how we can compare different classification modes using the ROC AUC curve. We first explore what a ROC AUC curve is and why it is better than an accuracy score for comparing different models. At last, we built 4 different classification models on the Pima Diabetes data set and plotted the ROC-AUC curve to pick the best performing model. Do you want to know how we can deploy this model now? Check here this article title as “Complete Tutorial On Tkinter to Deploy ML Models”.","excerpt":"In this article, we will learn more about the ROC-AUC curve and how we make use of it to compare different machine learning models to select the best performing model.","categories":["Deep Tech"],"tags":["machine learning models","regression analysis"],"author_name":"Rohit Dwivedi","publish_date":"2020-09-17T14:00:43","publication_year":"2020","word_count":1063,"keywords":["Go","NumPy","data_tools:Pandas","machine learning","programming_languages:R","AI","ML","regression analysis","machine learning models","programming_languages:Go","R","Pandas"],"extracted_tech_keywords":["AI","machine learning","ML","Pandas","NumPy","R","Go","programming_languages:R","programming_languages:Go","data_tools:Pandas"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/roc-auc-curve-for-comprehensive-analysis-of-machine-learning-models\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10057188,"title":"Telangana Government To Introduce AI-Based Quality Check For Crops Soon","content":"The Telangana Government’s IT department is set to introduce an AI-based quality assaying model to check the quality of crops brought to procurement centres across the State. The AI-based quality assaying model will be used to determine the grade, quality, moisture, and other parameters in paddy, red, and Bengal gram and will also deliver the report in just a minute. In addition, it will be used for other crops, including fruits and vegetables. The model will ensure that farmers are not at the mercy of selection officers at procurement centres in finalising the grade and quality of grains. The TSTS-ITE&C department has signed a Memorandum of Understanding (MoU) with Hyderabad-based Nebulaa Innovations Private Limited. The solution will be very helpful to farmers, civil supplies, and agriculture departments at the procurement centres. Under the existing process, the selection officers take a handful of grains and rub them in their hands to decide the moisture content in grains, their colour and size. The official said that conventional quality testing methods lead to disputes as it is done manually, adding that the AI-based quality assaying model would be faster and more objective in classifying grades and other parameters. Under this model, a few grain samples will be placed in a tray of the AI-based machine, which captures 360 degrees images of the grains. Then, based on the algorithm in the system, it identifies the type of grains (paddy, Red gram or Bengal Gram), colour, moisture content, size, immature, shrunk or shrivelled grains, presence of foreign matter (organic or inorganic), the official explained. When a farmer gets his stocks to the procurement centre, a lot is issued to his stocks. Officials can take four to five samples and assign unique numbers to each sample to avoid any duplication. From the farmers’ perspective, this report can also help alter farming practices. For example, the report gives a detailed analysis of chemical content in the samples and based on this; the farmer can change the use of fertilisers or the crop pattern to get better yields, the official said. Officials have already presented a demonstration to Markfed authorities in the city recently, and their response was encouraging. The official said that an AI-based system is presently used in Haryana and Rajasthan.","excerpt":"The AI-based quality assaying model will be used to determine the grade, quality, moisture, and other parameters in paddy, red, and Bengal gram and deliver the report in a minute.","categories":["AI News"],"tags":["telangana government"],"author_name":"Poornima Nataraj","publish_date":"2021-12-27T12:10:29","publication_year":"2021","word_count":375,"keywords":["Go","programming_languages:R","AI","innovation","programming_languages:Go","telangana government","RAG","Ray","GAN","R"],"extracted_tech_keywords":["AI","Ray","RAG","R","Go","GAN","innovation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/telangana-government-to-introduce-ai-based-quality-check-for-crops-soon\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10069707,"title":"Amazon SageMaker Ground Truth now supports synthetic data generation","content":"AWS has announced that you could now use Amazon SageMaker Ground Truth to generate labelled synthetic image data. SageMaker Ground Truth is a data labelling service that makes it easy to label data and gives you the option to use human annotators through Amazon Mechanical Turk, third-party vendors, or your own private workforce. You can also generate labelled synthetic data without manually collecting or labelling real-world data. SageMaker Ground Truth can generate hundreds of thousands of automatically labelled synthetic images on your behalf. To request a new synthetic data project, navigate to the Amazon SageMaker Ground Truth console and select Synthetic data. Source: Amazon Then, select Open project portal. In the project portal, you can request new projects, monitor projects that are in progress, and view batches of generated images once they become available for review. To initiate a new project, select Request project. Describe your synthetic data needs and provide contact information. An AWS expert will reach out to discuss the project requirements in more detail. Upon review, the team will share a custom quote and project timeline. AWS digital artists will start by creating a small test batch of labelled synthetic images as a pilot production for you to review. Amazon SageMaker Ground Truth synthetic data is available in US East (N. Virginia). Synthetic data is priced on a per-label basis. For more details, click here","excerpt":"SageMaker Ground Truth can generate hundreds of thousands of automatically labelled synthetic images on your behalf.","categories":["AI News"],"tags":["AWS Sagemaker","synthetic data generation"],"author_name":"Kartik Wali","publish_date":"2022-06-24T16:09:09","publication_year":"2022","word_count":228,"keywords":["Amazon SageMaker","synthetic data","AWS","AI","mlops_tools:SageMaker","cloud_platforms:AWS","programming_languages:R","emerging_tech:synthetic data","Git","AWS Sagemaker","synthetic data generation","R"],"extracted_tech_keywords":["AI","Amazon SageMaker","AWS","R","Git","synthetic data","mlops_tools:SageMaker","cloud_platforms:AWS","programming_languages:R","emerging_tech:synthetic data"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/amazon-sagemaker-ground-truth-now-supports-synthetic-data-generation\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10062590,"title":"Oracle announces higher earnings forecast based on Cloud boost","content":"Oracle Corp announced its third-quarter results for the period ended February 28, reporting a 4 per cent increase in revenue to USD 10.5 billion for the fiscal year 2022. The Texas-headquartered computer technology company forecast higher revenues and profits. Oracle is relying heavily on its cloud segment, having invested USD 4 billion capital to expand its number of data centres and improve its cloud services. The company’s cloud license and on-premium license were up by 1 per cent. “This strong top-line growth was coupled with a solid non-GAAP constant currency operating profit growth of 4%, but the big story is that our overall revenue growth is being driven by both our rapidly growing Cloud Infrastructure and Cloud Applications businesses. Q3 Cloud Infrastructure revenue was up 47% in constant currency. Q3 Cloud Applications growth was led by Fusion ERP, which was up 35% in constant currency and NetSuite ERP, which was up 29% in constant currency. Total Cloud revenue, which includes Cloud Infrastructure and Cloud Applications, is now over $11 billion a year,” Safra Catz, CEO of the company, said. Oracle has been pushing to grow its cloud services as customer demand also increases. The company spent 23 per cent more on cloud services and license costs. Oracle’s total operating expenses for the quarter stood at USD 6.69 billion. Larry Ellison, the chairman and CTO of the company, said that they had also developed its multi-cloud version to MySQL HeatWave open-source database within the third quarter. Designed to compete with Amazon’s Aurora and Snowflake, MySQL HeatWave will be available on Microsoft Azure Cloud and Amazon Cloud.","excerpt":"The company spent 23 per cent more on cloud services and license costs.","categories":["AI News"],"tags":["Oracle","Oracle Financial Service"],"author_name":"Poulomi Chatterjee","publish_date":"2022-03-11T15:06:10","publication_year":"2022","word_count":265,"keywords":["API","cloud_platforms:Azure","cloud_platforms:Microsoft Azure","AI","programming_languages:R","R","Oracle Financial Service","Oracle","programming_languages:SQL","SQL","Azure","Snowflake"],"extracted_tech_keywords":["AI","Azure","Snowflake","R","SQL","API","cloud_platforms:Azure","cloud_platforms:Microsoft Azure","programming_languages:R","programming_languages:SQL"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/oracle-announces-higher-earnings-forecast-based-on-cloud-boost\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":15953,"title":"Future of Indian fashion is AI – Watson makes a splashy debut on Indian runway","content":"Artificial Intelligence’s dalliance with fashion is well-known. From curating personal wardrobes, to dictating stock inventory for big retailers based on current trends and pricing algorithms, AI has invaded the whimsical fashion space. Now, Indian fashion made its first brush with AI – Mumbai-based designer duo Falguni and Shane Peacock collaborated with IBM and used the data presented by Watson to create three AI-based designs that were unveiled to a fashion savvy crowd. From cranking out structured silhouettes to dictating the colour swatch, the insights gleaned from the cognitive tool led to teal, azure, futuristic silhouettes on the runway. Analytics India Magazine caught up with the prolific designer duo to understand what led to this collaboration and why it was a good fit for their brand name as well. According to the designer duo, the recent collaboration with IBM was an ideal platform to show off all the elements of eclecticism that has been associated with Shane & Falguni. “We were part of a cutting edge technology that could potentially change the landscape of the fashion industry. Owe worked with a great team and getting inputs from Watson for crafting the collection was very knowledgeable for us,” they said. Has Watson proven its worth in fashion? Interesting as it may sound, Watson’s prowess has already been tested in cancer research. New York’s Memorial Sloan Kettering Cancer Center in New York has been using Watson for sometime to enhance cancer care and there was the failed association with Texas based MD Anderson Centre that shelved the project eventually. Watson for cancer has been the marketing punchline in healthcare for quite some time. However, industry experts believe it is early days and the tool will mature with time. Beyond Jeopardy, Watson has proven its worth in editing a thriller film trailer, composing music, providing excellent customer service via a support bot at Macy’s, running a rewards program for Chevrolet and there’s even Watson Ads that enables customers to interact via text and voice. But has Watson proven its worth in the field of fashion? Seems like tech heavyweight IBM is betting big on the power cognitive fashion and making data fashionable. The fact that IBM is betting big on Watson is abundantly clear and it is now being leveraged by big fashion houses to utilize social data to better interact with customers, improve sales with highly targeted marketing campaigns and predict customer orders. However, the runway seemed far from sight. This was also put to test at last year’s Met Gala where Marchesa designers Georgina Chapman and Keren Craig, unveiled a creation based on Watson’s insights. This was a fashion first, wherein 200 images were processed by Watson before coming up with the colour and the material of the gown. Also, the cognitive tool analysed the sentiment of tweets, which lit up the dress in different colours. Watson playing Hue’dunnit on Indian turf Shane and Falguni Peacock created a fashion first by teaming up with the Watson team and integrating cognition into the creations. According to Shane Peacock, “Fashion is usually considered a right-brain activity. So, it was very interesting to approach it from a left-brain perspective of logic and data. The data presented by IBM Watson helped us look at the insights as a springboard for our thoughts and helped us think in newer directions. Hence, we had to take this plunge, as it is the future of fashion”. The AI-based collection, aptly titled Future of Bollywood Fashion was created by analyzing around 600,000 publicly available historical fashion runway images from the last decade and also included images from the Big Four fashion weeks. Keeping in with the Bollywood theme, it further analyzed cognitive tool further In order to understand Bollywood fashion trends, IBM Watson analysed close to 5,000 Bollywood celebrity images culled from social media platforms, besides 3,000 fashion-related images such as Bollywood movie posters from the last four decades. While the first dataset was representative of high-end couture, the second dataset was indicative of Bollywood-centric fashion. Through these insights Watson declared the dominant colour palette — raisin black, Yankees blue and light grey. Talking about the collection, the designer duo shared, “We were  able to analyze the dominant prints and silhouettes for each season, which we could use as a starting point to explore further. An analysis and insights was shared with us based on all the fashion weeks internationally ,as an interactive web application hosted on IBM Cloud where we could explore the colour trends for the next season, analyze the key prints and silhouettes for each season and derive inspiration for their new collection”. Future of AI in Indian fashion This is the just the first instance where AI plays into Indian fashion. The designer duo believes it can be a game-changer when it comes to predicting prints and silhouettes. Will it also set the price lines for  prêt-à-porter and haute couture lines? “Well, we did not take that approach, but it can. We needed data for prints and fashion trends and it gave us great inputs for the same,” they shared. How can AI impact Indian fashion industry? “India’s transformation in women’s clothing and fashion over the years has been as drastic as compared to its western counterparts, owing to various traditional beliefs and values still held by most Indian women. Nevertheless, with changing times, even the most traditional apparels have been modified to hold an urban look without compromising on tradition,” they said, in closing. Outlook And India seems to be a very important market for IBM Watson and It is already working with Manipal Hospitals to help doctors identify personalized cancer care for patients. India has the second-largest developer market and an expanding startup ecosystem. IBM India is working with a lot of early stage start-ups to help them grow and scale. With Watson, IBM is making an attempt to give a cognitive push to clients and helping them adopt the platform.","excerpt":"Artificial Intelligence’s dalliance with fashion is well-known. From curating personal wardrobes, to dictating stock inventory for big retailers based on current trends and pricing algorithms, AI has invaded the whimsical fashion space. Now, Indian fashion made its first brush with AI – Mumbai-based designer duo Falguni and Shane Peacock collaborated with IBM and used the […]","categories":["IT Services"],"tags":[],"author_name":"Richa Bhatia","publish_date":"2017-06-29T08:07:54","publication_year":"2017","word_count":987,"keywords":["Go","artificial intelligence","AI","R","RPA","RAG","ViT","analytics","Azure","startup"],"extracted_tech_keywords":["AI","artificial intelligence","analytics","RAG","Azure","R","Go","ViT","RPA","startup"],"url":"https:\/\/analyticsindiamag.com\/it-services\/future-indian-fashion-ai-watson-makes-splashy-debut-indian-runway\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10103837,"title":"Dell Secures $150 Million Hardware Deal with AI Startup Imbue","content":"To expand its presence in the competitive AI market, Dell Technologies Inc. has inked a lucrative $150 million deal with Imbue, an AI startup that aims to build personal computers with AI agents. Under the deal, Dell will provide Imbue with servers essential for processing the massive datasets required to develop advanced AI systems and construct models with sophisticated reasoning capabilities. Imbue, which recently raised $200 million in funding, stands out among AI startups by independently building its AI foundation models from the ground up, a resource-intensive endeavour demanding substantial computing power. While industry leaders like Microsoft, Google, and Amazon have aggressively pursued partnerships with AI startups through investments and cloud computing arrangements, Dell has positioned itself as a smaller player, emphasising the potential benefits of its server business in the AI landscape. Matt Baker, senior vice president of AI strategy at Dell, expressed the company’s eagerness to support the dynamic innovation in the AI space, stating, “The entirety of our business is pivoting to support what we believe is a once or twice-in-a-lifetime opportunity. The innovation that generative AI is driving is rivalling the arrival of the internet.” Unlike typical cloud deals in which AI startups receive computing services, Imbue’s agreement with Dell involves the upfront purchase of computing hardware. Imbue contends that this approach is not only cost-effective compared to using major cloud providers like Amazon and Google but also provides greater flexibility, preventing overreliance on any single large tech company. Josh Albrecht, co-founder and CTO of Imbue, highlighted the decision to partner with Dell, explaining, “The main reason we went with Dell is that we don’t want to be locked into a computing provider. This allows us to be able to remain independent.” Baker emphasised that Dell’s partnership with Imbue demonstrates that AI startups have alternatives to turning to public cloud providers. “These are things that you can actually own, install, and innovate on your own,” he stated. Imbue and Dell collaborated to design a custom system featuring smaller “clusters” of servers for rapid experimentation in AI system development. Additionally, a larger cluster is tailored for constructing foundation models—versatile AI systems adaptable to a range of tasks. Imbue’s computing system is managed by Voltage Park, another firm. Founded in 2021 and currently valued at over $1 billion, Imbue is actively developing early prototypes of AI tools known as “agents.” These agents aim to automate complex tasks such as analysing code bugs, interpreting lengthy documents, and eventually planning vacations without user supervision. Imbue’s ultimate goal is to create agents with advanced reasoning capabilities to assist engineers in coding and aid analysts in drafting policy proposals.","excerpt":"Imbue and Dell collaborated to design a custom system featuring smaller “clusters” of servers for rapid experimentation in AI system development.","categories":["Deep Tech"],"tags":["ai investment","ai investments","Dell","Mergers and Acquisitions","Startups"],"author_name":"Mohit Pandey","publish_date":"2023-11-29T15:21:30","publication_year":"2023","word_count":437,"keywords":["Go","API","ai investments","Dell","AI","cloud computing","innovation","Aim","generative AI","ai investment","Startups","Mergers and Acquisitions","R","foundation models","startup"],"extracted_tech_keywords":["AI","generative AI","foundation models","Aim","cloud computing","R","Go","API","innovation","startup"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/dell-secures-150-million-hardware-deal-with-ai-startup-imbue\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":2415,"title":"The Future of Cloud Computing, Big Data and Buildings","content":"The commercial real estate (CRE) market is rich with opportunities for smart buildings to improve property management top- and bottom-line performance through expense reduction, increased property values and better workplace experiences for tenants. Today, amid the flurry of smart homes, self-driving cars and smart wearable devices, only about 15% of commercial buildings are “smart,” with connected building automation systems in place. But, that’s about to change BIG TIME. In fact, the CRE market is projected to have the fastest and largest growth in connected devices between 2016 and 2021, surpassing even smart homes, according to the Smart Building Research firm Memoori. [blockquote author=”” pull=”pullright”]Smart Buildings can save upto 50% on HVAC and Lighting Energy Costs[\/blockquote]Legacy building controls systems are overbuilt and expensive for most commercial building owners and operators. Those systems are being displaced by new, born-digital solutions. Fuelled by the proliferation of affordable sensors and cloud computing, this new breed of building intelligence solutions is changing-up the economics of the real estate investment world. Now, it’s more affordable and easier than ever to realize net gains from the affordable, yet sophisticated, building automation systems (BAS) that combine the Internet of Things (IoT), wireless communications, big data and cloud-based algorithms to make your buildings smarter, more efficient and more attractive for both current and potential tenants. These new systems offer several ways to both reduce expenses and improve property values, netting positive for all real estate stakeholders – investors, owners, asset managers, brokers, property managers and, of course, tenants – the ultimate customers. When you factor utility incentives for energy-efficient systems, or just the annual operational savings, building automation systems are becoming a smarter choice for property CAM, TI and upgrades on financial merits alone, before factoring the many workplace benefits. Perhaps the greatest value in smart buildings is in the improved workplace experience, with comfortable, healthy, productive and easily-managed indoor environments for tenants. An intelligent building system can play a key role in attracting, retaining and delighting tenants, offering lower energy bills, eco-friendly offices, and creating great places to work with healthy indoor air quality and unique occupant experiences (OX). The CRE market is projected to have the fastest and largest growth in connected devices between 2016 and 2021, surpassing even smart homes, according to the Smart Building Research firm Memoori. EXPENSES GOING DOWN OpEx Reduction Reduce operating expenses across building systems, including HVAC and lighting maintenance. Energy Savings A Smart Building system can save 30- 50% of HVAC energy consumption, plus reduce LED and other lighting energy – conservatively $.20-.50 PSF. Rebates Utility-driven efficiency incentives can reduce total cost of acquisition and speed ROI timeframe. Reduced HVAC Equipment Higher efficiency means fewer or downsized equipment, or even eliminating traditional fixtures. Extend Equipment Life Retrofit controls to equipment; smart monitoring and predictive maintenance can add years to your equipment lifecycle. No IT Hardware Costs No need to invest in workstations and software updates; cloud based software enhancements pushed wirelessly Lower Installation Costs Get 5x faster installation of wireless system, eliminate temperature controls specialists and custom programming. Remote Multi-site Management Web and mobile monitoring and adjustments without site visits; reducing travel expenses and team disruption. Optimization Through Insights Further efficiencies can be driven by data insights, patterns understanding demand and supply \/ performance. VALUE GOING UP Higher Rates per Square Smart building systems can lift property value by at least $.10\/sf, increasing as much as 11.8%, adding to your top line. Long term Hold Assets 10+ year plans; internal rate of return. Differentiation Only 15% of commercial buildings are smart. Yet, progressive tenants seek and value the benefits and controls of smart, connected properties. Staged for the Market A smart automation system shows well. A quick retrofit to your current and vacant properties provide a fast enhancement. When showing, use a Google calendar feed or geo-fencing based pre-conditioning to make that first visit comfortable and inviting. Attract + Retain Tenants Smart systems, and applications that empower the users, are becoming a means to get interest from prospective tenants and to stay sticky with existing clients. Creating the “Occupant Experience” is a huge opportunity for landlords to give management remote insights, save energy, improve workplace IAQ and comfort, plus empower employees with control over their zones. Higher Return TI Investment Adding more-lasting and differentiated value than fresh paint and carpet. Higher Return CAM Investments Can offer uncommon value for common areas and budget management. Green Value in Sustainability Green buildings have ever increasing attention and value in the market, supporting corporate citizenship and triple-bottom line performance for profit, people and the planet too. DECREASING EXPENSES The U.S. Green Building Council cites that buildings consume 70% of the U.S. electricity and account for 39% of U.S. CO2 emissions. The U.S. Department of Energy (DOE) indicates that commercial buildings waste 30% of the energy they consume. This is a major opportunity for smart buildings efficiency to save significant energy expenses and reduce environmental impact; yet, that’s just the start of the savings. STARTING TODAY, SMART SOLUTIONS CAN CUT ENERGY BY 25% One of the greatest and most immediate opportunities to reduce property operational costs is in energy reduction, particularly in more efficient management of heating, venting and air conditioning (HVAC) and lighting. In the average commercial building, HVAC and lighting makes up about 50% of energy use⁵; so, the potential of 30-50% energy savings from smart HVAC could translate to 25% of total energy use. That alone is worth the investment, often with payback in less than 3 years, sometimes less than a year. And that’s just the start of savings. Smart building technologies can reduce costs and add value to your TI packages, while boosting returns for owners and tenants alike. EFFICIENCY INCENTIVES PAY GREEN FOR GREEN EFFORTS Utility-driven rebates, low- or no-rate financing and tax incentives for energy efficient systems help cover costs and reduce payback timing, leading to net gains even sooner. A quick, no-cost site survey and estimated projection of energy savings can help to cost-justify the building automation system investment and define the payback timeframe, from only an energy savings perspective. LEED CREDITS INCREASE SAVINGS (AND INCREASE PROPERTY VALUE) The US Green Building Council’s LEED program recognizes Leadership in Energy and Environmental Design, with different levels of reward: LEED certified buildings (40-49 credits), silver (50-59 credits), gold (60- 79 credits) and platinum (80+ credits). LEED buildings have faster lease-up rates, may qualify for a host of incentives like tax rebates and zoning allowances, and retain higher property values⁶. For example, 75F intelligent building solutions can contribute to up to 38 different LEED v4.0 points, up to 11 credits. This provides even further capability for you to demonstrate sustainability and good corporate citizenship, in addition to eligibility for financial incentives, such as tax credits. And, new measurement tools such as USGB’s Arc Score can help you assess your ongoing building performance benchmarked against others, whether you’re a LEED building or not. The Arc score offers dynamic measures for Energy, Water, Waste, Transportation and Human Experience.","excerpt":"The commercial real estate (CRE) market is rich with opportunities for smart buildings to improve property management top- and bottom-line performance through expense reduction, increased property values and better workplace experiences for tenants. Today, amid the flurry of smart homes, self-driving cars and smart wearable devices, only about 15% of commercial buildings are “smart,” with […]","categories":["IT Services"],"tags":["Cloud Computing","current leaders in self driving cars"],"author_name":"AIM Media House","publish_date":"2018-04-17T10:03:49","publication_year":"2018","word_count":1166,"keywords":["big data","Go","cloud computing","AI","RPA","Git","RAG","automation","Cloud Computing","ViT","R","current leaders in self driving cars"],"extracted_tech_keywords":["AI","RAG","cloud computing","R","Go","Git","big data","ViT","automation","RPA"],"url":"https:\/\/analyticsindiamag.com\/it-services\/future-cloud-computing-big-data-buildings\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10066601,"title":"Should Indian Defence Forces get a Chief AI Officer?","content":"In 2017, Russian President Vladimir Putin said that the country that leads in AI would rule the world. He really meant it. In the recent invasion of Ukraine, the Russian forces reportedly used AI-based drones, some of which were spotted by the Ukrainian forces within their territory. The other side – Ukraine – was not behind. The controversial company, Clearview AI is said to be supporting Ukrainian forces to ‘uncover the Russian assailants and combat misinformation.’ Also Read: What Went Wrong With Clearview AI? Military superpowers around the world are leveraging the neck-breaking progress in the field of AI to one-up each other. No country can afford to sit out of the AI-based arms race. It is quickly becoming an important component of military strategising and planning. The US Department of Defense recently announced the appointment of Dr Craig Martell as the Chief Digital and Artificial Intelligence Officer (CDAO). At an earlier conference, the Deputy Secretary of Defense in the US, Kathleen H Hicks, said that this position would be focused on speed – and ensuring the right processes and organisation in place to leverage AI and data. She also added that such a focus would help the department advance a host of different warfighting concepts and capabilities. AI warfare “Autonomous weaponry is the third revolution in warfare, following gunpowder and nuclear arms. The evolution from land mines to guided missiles was just a prelude to true AI-enabled autonomy—the full engagement of killing: searching for, deciding to engage, and obliterating another human life, completely without human involvement,” Kai-Fu Lee wrote in his book AI 2041: Ten Visions for Our Future. Lee’s prediction is not wrong at all. In 2018, Venezuelan President Nicolas Maduro was speaking at a military event in Caracas. In the middle of the speech, with a look of horror, Maduro looked up to find an explosive drone approaching him. This alleged assassination attempt on the president missed its target but managed to injure seven soldiers. While Maduro escaped, Iranian nuclear scientist Mohsen Fakhrizadeh wasn’t so lucky. He was killed by a satellite controlled autonomous gunshot in November 2020. During the Israel-Palestinian battle, the former adopted a layered approach to missile defence – deterrence, attack on missile launches, early warning systems, and damage minimisation. Of particular interest is Israel’s Iron Dome, which is a very short-range AI-based defence system. It acts within a set of pre-programmed parameters to intercept missiles, rockets or mortars aimed at high-value targets at 2.5 to 43 miles range. It was developed by Rafael Advanced Defense Systems and Israel Aerospace and first tested in 2008. The first usage of AI in a conflict could probably be traced back to 2020 during the 44-day war that ensued between Armenia and Azerbaijan over the disputed Nagorno-Karabakh region. From older Soviet-era Scud and Tochka missiles to newer and advanced systems like Iskander and Israel-made LORA (LOng Range Attack) missiles were used. Drones of Russia, Turkish, and Israel make performed missions to support artillery use and strike missions. Unmanned aerial vehicles (UAV) were able to destroy heavy ground units, including T-72 tanks and S-300 air defenses. Need for an AI chief During the Russian-Ukraine conflict, a third party quietly seemed to have benefitted. US’ Pentagon used AI and machine learning tools to analyse vast amounts of data to generate useful battlefield intelligence and learn more about Russian tactics. In an interview with Defense One, the Director of Defense Research and Engineering for Modernization Maynard Holiday, said, “What you’re not seeing is our exquisite intelligence capabilities that are able to oversee the battlefield, including gathering and archiving signals intelligence.” He added, “We’ll definitely be doing an after-action analysis on everything we’ve seen with respect to Russian tactics.” All of this information will now be used for creating a database that can be trained for future use. The US is not discreet about its ambitions to use AI to bolster its military capabilities. Earlier this year, the Deputy Secretary of Defense, Hicks, issued a memorandum highlighting the importance of the US Defense Department as an AI-enabled enterprise that can operate at the speed and scale required to preserve its ‘military advantage’. The memorandum also mentioned the creation of a new post – the Chief Data and Artificial Intelligence Officer (CDAO) – to oversee data and AI initiatives under one official at the highest level of the Pentagon. Since February, John Sherman, the Chief Information Officer at the Department of Defense, has been serving as the interim CDAO. Newly appointed CDAO, Craig Martell is tasked with leveraging AI and analytics to advance the military prowess of the US and, in turn, one-up adversaries like China. The US Defense Department plans to spend USD 874 million on AI technologies in 2022. Newly appointed CDAO, Craig Martell is tasked with leveraging AI and analytics to advance the military prowess of the US and, in turn, one-up adversaries like China. The US Defense Department plans to spend USD 874 million on AI technologies in 2022. CDAO for India? In June 2021, the Jammu Air Station was attacked by two explosive-carrying drones. The two explosions happened within a gap of five minutes – one blast damaged a single-storey building while the second one exploded on open ground. The initial probe indicated the involvement of Pakistan-based agencies and terror outfits like Lashkar-e-Taiba. While this was the first instance of drones being used for direct attack, in the past, the Border Security Forces (BSF) has reported several drone sightings. As per a few reports, BSF witnesses 10-15 sightings a day – while most of these drones are for surveillance purposes, a few were also used for dropping weapons and explosives in the area close to Pakistan’s borders. Keeping up with the time, the Pakistan defence forces are also developing an AI-based strategy. In 2020, the Pakistan Air Force established the Centre for Artificial Intelligence and Computing (CENTAIC), which would lead R&D in this space. Pakistan finds an ally in China. Military and strategic partnerships form an important part of their friendship. Recently, PLA Senior Colonel Wu Qian, spokesperson of the Ministry of Defence, said that military-to-military relations between the two countries are the ‘mainstay of the China-Pakistan friendship’. There have been recent reports about China exporting missiles and even autonomous machinery to Pakistan. For example, as per reports, a Chinese military drone manufacturer Ziyan sold its autonomous weaponised Blowfish A2 drone to Pakistan. Speaking of China, its People’s Liberation Army is probably the most technologically advanced military in the world. The Center for Security and Emerging Technology (CSET) at Georgetown University released a report titled “Harnessed Lightning: How the Chinese Military is Adopting Artificial Intelligence”. As per the report, Chinese leaders are actively pushing for stronger integration of AI in the military to transform PLA into a world-class and globally competitive military force. The research group’s analysis showed that PLA is developing autonomous vehicles and using AI and machine learning to improve intelligence analysis, target recognition and information warfare. In such a case, where India faces a two-front challenge, the country’s military can hardly afford even a minor slip. To its credit, the top military leadership in India have regularly spoken about stronger AI integration in the forces. A few concrete steps have also been taken in this regard. India constituted a multi-stakeholder task force on the Strategic Implementation of AI for National Security and Defence in 2018. In June 2018, under N Chandrasekharan (chairman, Tata Sons), the task force submitted its report, stressing the need for ‘policy and institutional interventions required to regulate and encourage robust AI-based technologies for the defence sector in the country’. Less than a year later, the ministry established a high-level Defence AI Council (DAIC) with the Defence Minister as the chairman. DAIC is now tasked with providing strategic direction towards AI adoption in defence. DAIC guides the public-private partnerships and reviews the recommendations on technology acquisition. It is also responsible for reviewing the ethics, safety and privacy aspects of AI usage in defence. The council meets twice a year to ‘provide strategic direction’ for AI-driven transformation. The task force recommended setting up the Defence AI Project Agency (DAIPA). As a result, the Ministry of Defence, through an Executive Order dated February 8, 2019, created the Defence AI Council (DAIC) under the chairmanship of Raksha Mantri and the Defence AI Project Agency (DAIPA) with Secretary (DP) as ex-officio head for providing necessary guidance to enable and affect the development of the operating framework, policy-level changes and structural support for AI adoption. Last December, the Indian Army set up a quantum computing laboratory and an AI centre at the military engineering institute in Mhow, Madhya Pradesh. While these steps are commendable, a more aggressive approach might be a need. The Indian military can have a dedicated arm for AI and other emerging technology-based war and defence preparedness and strategising. Such a body could be under the direct administration of the Ministry of Defence, headed by a technologically proficient and experienced chief. The US Department of Defense has roped in an industry expert as the CDAO. The move makes sense since the private sector plays a critical role in pushing the frontier of AI innovation (India ranks third among G20 countries in the number of AI-based startups). However, India has been historically circumspect in bringing private players to handle key positions in the defence forces. Hence, appointing an industry expert as a CDAO-equivalent would mean a lot of bureaucratic overhauls and attitudinal shifts. Besides, AI adoption is capital-intensive. The government needs to create departments to train defence forces in emerging technologies like AI and pour money into the R&D to build AI capabilities. Autonomous weapons are the future of modern warfare. To that end, India should start by establishing an AI team and an AI chief to ensure we don’t fall behind in the arms race.","excerpt":"The top military leadership in India have regularly spoken about stronger AI integration in the forces.","categories":["AI Trends"],"tags":["defence ministry"],"author_name":"Shraddha Goled","publish_date":"2022-05-10T16:00:00","publication_year":"2022","word_count":1642,"keywords":["Go","API","machine learning","artificial intelligence","AI","defence ministry","Git","RAG","Aim","analytics","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","analytics","Aim","RAG","R","Go","Git","API"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/should-indian-defence-forces-get-a-chief-ai-officer\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":2031,"title":"Interacting with the makers of Wattman- an IoT based Building Management System that optimizes energy consumption by 30%","content":"Co-founders, Rahul Bhalla, Dr. Amarjeet Singh and Vishal Bansal, Zenatix (L-R) This Haryana based, data-driven energy efficiency company is using advanced machine learning based models to save 10-30% of electricity for large commercial consumers. These models are built by collecting and analyzing millions of previously unseen (“dark”) data points every day from electrical assets Zenatix launched their 1st product, WattMan, in April 2016, which is an IoT based Building Management System (BMS) tailor-made for large retail chains and banking\/ financial set ups. It is built over a scalable Internet of Things (IoT) stack that includes robust connected hardware developed in-house by Zenatix through extensive research and development. Co-founded by Rahul Bhalla, Vishal Bansal and Dr. Amarjeet Singh, Zenatix is currently operating in Indian market with a focus of expanding into the South East Asia market. IoT India Magazine caught up with Vishal Bansal, Co-founder and COO, Zenatix, where he is responsible for leading and managing the product and operations at Zenatix. Prior to Zenatix, he has been associated with ING Insurance Asia, where he managed investment strategy for insurance portfolios globally. An expert in financial analytics, he has developed statistical models for construction and optimization of equity portfolios. A graduate from  IIT-Delhi and alumni of IIM Ahmedabad, Vishal is a sports enthusiast and has an interest in Indian politics. IoT India Magazine- Zenatix is a leading provider of IoT based energy monitoring and control products. Could you tell us about the products it offers? Vishal Bansal- People at Zenatix are passionate about sustainability and strongly believe in power of technology and data and wanted to leverage these to make a meaningful contribution towards a sustainable planet. Energy Efficiency is one critical element for sustainability and a lot of work is being done in this domain. However, the primary focus has been on retrofits (build more energy efficient light bulbs, air conditioners, etc.), but the operational energy efficiency (use of existing infrastructure optimally) has been largely overlooked. In this respect, Zenatix has developed an innovative product called WattMan. It is an IoT based Building Management System (BMS) tailor-made for retail chains and banking set-ups to help them in reducing their electricity consumption by up to 30%. IoT India Magazine-  Could you explain in detail how Zenatix is using machine learning based models to deliver 10 to 30% energy efficiency to large commercial consumers of electricity? How is WattMan addressing various challenges? Vishal Bansal- Given the distributed nature of retail\/banking set-ups (100s of thousands of outlets\/branches, spread across the country), having a control over the electricity spend and providing the right customer experience at the same time, is a big challenge. Our product, WattMan, addresses these challenges in an automated manner through its cloud-based intelligent controls driven by advanced machine learning algorithms. It employs our proprietary firmware for performing automatic sensing and control. Collected data is used to develop machine learning driven models deciding optimal operational settings based on external factors like weather and footfall, thus allowing sophisticated control. Besides automated control, we use various machine learning and Artificial Intelligence techniques for anomaly detection and predictive and preventive maintenance of electrical assets. IoT India Magazine- The company says that “the models are built by collecting and analyzing millions of previously unseen (“dark”) data points every day from electrical assets”. What do you mean by that? Vishal Bansal- Data has shown that increase in electricity generation over the years has not been able to keep pace with increasing demand. Moreover, more than 70% of generation is from fossil fuels. Further, the extensive research done by our co-founder and CTO, Dr. Amarjeet Singh has shown that commercial buildings typically waste more than 30% of electricity that they use. Therefore, we decided to disrupt the traditional energy efficiency market by focusing on operational energy efficiency and using technology and advanced data analytics to achieve that. Until we came up with our innovative product, the operational energy data was not available and hence this is previously unseen or dark data. We collect millions of such data points everyday to derive useful insights from the data that can eventually help our clients in better control and therefore less wastage of electricity. IoT India Magazine- What are the major industries that Zenatix caters to? Would you like to exemplify with few use cases how Zenatix’s solutions are being used in these industries? Vishal Bansal- WattMan is an innovative IoT based Building Management System (BMS) for retail chains and banking set-ups. Given the distributed nature of retail\/banking set-ups (100s of thousands of outlets\/branches, spread across the country), having a control over the electricity spend and providing the right customer experience at the same time is a big challenge. For example, in a Quick Service Restaurant (QSR) or a supermarket, maintaining the right temperature in customer areas, kitchen areas and cold storages is crucial. Also, they need to make sure that all their air-conditioning, kitchen and refrigeration appliances are working optimally all the time. They may have different On\/Off schedules for appliances in different areas. WattMan addresses these challenges in an automated manner through its cloud based intelligent controls. Further, the electrical appliances may break down and that can have direct impact on sales. With our Machine Learning based models, we can predict the breakdown of appliances in advance. IoT India Magazine- “Zenatix aims to address the energy efficiency problem across a range of customers and eventually help grids turn smarter”. How has the company been working to achieve this? Vishal Bansal- Zenatix’ current product is aimed for retail chains and banking set-ups. We are currently working on building a smart IoT based Building Management System (BMS) for large commercial buildings such as offices, hotels, hospitals, etc. One of the key problems for utilities (power distribution companies) is managing their peak load. With our automated control already installed at 1000s of locations, we can help them run a Demand Response program that can reduce their peak load and help manage their energy portfolio in a smarter and dynamic manner. Demand Response basically means incentivizing the consumers to reduce their consumption at a particular time, so that utilities can reduce their peak load. IoT India Magazine- What are the major challenges at your disposal? Vishal Bansal- Both IoT and machine learning are new technologies which are changing the way customers consume energy. Our biggest challenge is adoption of our state of the art product by customers. While the customers clearly see the benefits of our product, there is an adoption curve for any new technology or product. IoT India Magazine- WattMan is one its kind product in the market? Do you see any competition in the market? Vishal Bansal- There is fragmented market with some hardware automation companies who have outsourced software and cloud infra capabilities. Similarly, there are software companies which are getting hardware done from third parties. Zenatix is the only company which has built the entire product stack – hardware, firmware and software including the IoT platform. We believe that having a complete control of the product stack is key to the hardware robustness, and supporting additional use cases for customers. IoT India Magazine- How has the growth story been? What are the company’s expansion plan? Vishal Bansal- WattMan was launched about 1 year ago and has been embraced by some of the leading retail chains and banks in India. The product clearly provides value in terms of centralised governance and energy efficiency (resulting in RoI from the first month itself). Our team is continuously developing and adding use cases to our products so as to deliver more value to our customers, while simultaneously innovating technologies to reduce costs of hardware and improve its robustness. We also envision expanding from energy efficiency to smart retail as we increase our install base. With robust hardware, connected to cloud and with good processing capabilities, already in place, we envision developing several other industry specific applications taking our customers to a smart retail solution. We are also developing channel partnerships with other players who provide orthogonal solutions to our target market segment e.g. those providing electrical panels. Such partnerships will help us achieve faster growth in a shorter span of time. IoT India Magazine- Would you like to talk about funding? How does the company plan on utilizing these funds? Vishal Bansal- We are a well funded company with backing from leading funds in India – Blume Ventures and pi Ventures. We have also raised money from some of the most reputed angel investors in India, namely, Rajan Anandan, Rahul Khanna, Kunal Bahl, Rohit Bansal and others.","excerpt":"This Haryana based, data-driven energy efficiency company is using advanced machine learning based models to save 10-30% of electricity for large commercial consumers. These models are built by collecting and analyzing millions of previously unseen (“dark”) data points every day from electrical assets Zenatix launched their 1st product, WattMan, in April 2016, which is an […]","categories":["AI Features"],"tags":["Internet of Things India","Interviews and Discussions","IoT India"],"author_name":"Srishti Deoras","publish_date":"2017-05-18T10:38:50","publication_year":"2017","word_count":1426,"keywords":["Go","artificial intelligence","machine learning","IoT India","AI","Internet of Things India","Scala","RAG","Aim","anomaly detection","analytics","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","analytics","Aim","RAG","anomaly detection","R","Go","Scala"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/interacting-makers-wattman-iot-based-building-management-system-optimizes-energy-consumption-30\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10070085,"title":"Did you miss the Intel® oneAPI Workshop on advanced SYCL concepts for heterogeneous computing*? Here’s what you need to know","content":"Intel®, in collaboration with Analytics India Magazine, recently concluded its Intel® oneAPI workshop on advanced SYCL concepts for heterogeneous computing, on June 24, 2022, at IST 5:00 PM. The session witnessed close to 200+ participants. The workshop covered advanced concepts and features of the latest SYCL specifications, including simplified code implementation for heterogeneous computing, an overview of memory models involving buffers\/accessors and unified shared memory (USM), and an understanding of simplified reduction with live hands-on sample codes on Intel® DevCloud. The session was led by Jyotsna Khemka, software engineering manager at Intel Corporation – Asia Pacific & Japan – South, alongside Subarnarekha Ghosal, a software technology consulting engineer at Intel Corporation. Key highlights Opening the workshop, Khemka threw light on the oneAPI use cases, showcasing one of the case studies done by the University in Berlin, where they developed a Tsunami application called the Easy Wave Simulation, which was developed for the Nvidia GPU. The team used oneAPI as a programming model, converted that code into SYCL, and was able to use it on different kinds of hardware architecture. Further, she spoke about various oneAPI toolkits. She said that some of these tools allow users to convert their CUDA code to SYCL and use this programming language to run on multiple hardware and instances. Following this, Ghosal introduced SYCL 2020 language specification. Highlighting the latest features, she said that it enhances productivity, and more than anything, it reduces the verbosity of writing code and runs the code faster across laptops. She also spoke about SYCL Standard, an open collaboration platform. “SYCL is growing every day, and it is growing taking the feedback,” she added. Simply put, Intel®’s SYCL* standard helps in enhancing productivity, performance, and fast-tracking open collaboration. (Source: Intel) Comparing the latest version of SYCL with traditional coding, she gave an overview of various parameters that go into writing the code, including the SYCL buffers method, explaining host code, accelerator device code and host code. In addition, she also showcased how it simplifies language simplification and more. This was followed by a hands-on guide to learning pointer-based memory management for heterogeneous computing using unified shared memory, alongside understanding the implicit and explicit ways of moving memory using unified shared memory and handling data dependency between kernel executions. (Source: Intel) In addition, Ghosal showcased the advantages of using Subgroups in SYCL programming, deciphering various concepts, and explained how Subgroup Shuffle operations could help avoid explicit memory operations. She also showcased a hands-on explanation of using SYCL reduction to simplify reduction with parallel kernels and urged the users to take advantage and reduce functions to reduce sub_group and work-group levels. Click here to download Intel® oneAPI Toolkits to get started. Click here to create an Intel® DevCloud account. At the end of the work, Analytics India Magazine ran a lucky draw, wherein lucky participants won an Amazon Voucher worth INR 2000\/- each. The winners were shortlisted based on their engagement with Discord throughout the workshop. Nikhitha AvulaPrasanta KunduAnirban MallaAnirban DasguptaAswin VijayakumarVidyasagar MRakesh RoyKarthik Koundinya S RRahul RajShobana LakshmiNarsimhan Here’s a Github link to explore the lab session. Sign up for Intel DevCloud.","excerpt":"The workshop covered advanced concepts and features of the latest SYCL specifications, including simplified code implementation for heterogeneous computing","categories":["Deep Tech"],"tags":["Intel","oneAPI"],"author_name":"Amit Naik","publish_date":"2022-06-29T18:00:00","publication_year":"2022","word_count":521,"keywords":["CUDA","Go","API","AI","Git","ViT","analytics","GitHub","R","oneAPI","Intel"],"extracted_tech_keywords":["AI","analytics","CUDA","R","Go","CUDA","Git","GitHub","API","ViT"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/advanced-sycl-concepts-for-heterogeneous-computing-heres-what-you-need-to-know\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10081140,"title":"Meet GPT-JT &#8211; The Open Source Alternative to GPT-3","content":"Recently, Together, an open source community led by researchers and technocrats, released a new version of GPT-JT trained on six billion parameters. This new model has been built using recently published open source techniques and datasets. It has been trained with its decentralised approach on the Together Research Computer, a local, in-house network developed by the team. Unlike GPT-3, which is available on-demand\/request, GPT-JT is now available as open source. Click here to access the code and datasets. GPT-JT GPT-JT came very close to GPT-3’s text-davinci-002 (175B) in terms of performance on classification benchmarks such as RAFT (following holistic evaluation of language models (HELM) protocol). Read: OpenAI Turns to Davinci to Make GPT-3 Better Together claimed that this new model is a variant of the previous GPT-J (6B), and performs well on text classification and other tasks. Banking on the power of open source AI, the team said that this would not have been possible without the open source works published by several organisations, including EleutherAI (GPT-J-6B, GPT-NeoX), Google Research (UL2, CoT), Natural-Instructions (NI) dataset by AllenAI, BigScience’s Public Pool of Prompts (P3) dataset, Ought (RAFT), and Stanford CRFM (HELM). Open. Scalable. Together Founded in 2022, Together is a decentralised cloud for artificial intelligence that enables researchers, developers, and companies to leverage and improve AI with an intuitive platform combining data, models, and computation. The team said that their GPT-JT model is inspired by lo-fi and ProxSkip, developed by Ludwig Schmidt, Mitchell Wortsman, Peter Richtárik, and others. The company has hosted this model through HuggingFace, which is the hub for the open-source AI ecosystem. Check out the live demo here.","excerpt":"This new model has been built using recently published open-source techniques and datasets","categories":["AI News"],"tags":["EleutherAI","GPT-J"],"author_name":"Aparna Iyer","publish_date":"2022-11-30T16:22:00","publication_year":"2022","word_count":270,"keywords":["Go","text classification","artificial intelligence","OpenAI","AI","Scala","GPT-J","RAG","GPT","Aim","EleutherAI","R"],"extracted_tech_keywords":["AI","artificial intelligence","OpenAI","Aim","RAG","text classification","R","Go","Scala","GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/meet-gpt-jt-the-closest-open-source-alternative-to-gpt-3\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10088302,"title":"Microsoft Introduces Multimodal Large Language Model Kosmos-1","content":"Following the recent developments in large language models using Transformers, an attention-based mechanism developed by Google in 2017, Microsoft released its research paper, Language Is Not All You Need: Aligning Perception with Language Models. The model introduces a multimodal large language model (MLLM) called Kosmos-1. The paper talks about the importance of integrating language, action, multimodal perception, and world modelling for stepping towards AGI. The research explores Kosmos-1 in various settings like zero-shot, few-shot, and multi-modal chain of thoughts, on several tasks without fine-tuning or gradient updates. Check out the research paper here. The model shows promising capabilities on various generation tasks by perceiving general modalities such as OCR-free NLP, visual QA, and perception-language tasks, and vision tasks. The Microsoft research team also introduced the model to a dataset of Raven IQ test for analysing and diagnosing the non-verbal reasoning capabilities of MLLMs. Below is the example of multimodal chain-of-thought prompting. This enables the model to deal with complex questions and reasoning tasks, by generating a rationale before tackling the problem. The team believes that moving from LLMs to MLLMs is better for achieving new capabilities and opportunities for language and multimodal tasks. Though it hasn’t been updated yet, you can check out the Github repository here for future updates.","excerpt":"Microsoft released their research paper, titled – Language Is Not All You Need: Aligning Perception with Language Models. The model introduces a multimodal large language model (MLLM) called Kosmos-1.","categories":["AI News"],"tags":["AI Chatbot","ChatGPT"],"author_name":"Mohit Pandey","publish_date":"2023-02-28T13:58:16","publication_year":"2023","word_count":210,"keywords":["Go","ChatGPT","AI","AI Chatbot","Modal","ML","Transformers","Git","NLP","chain of thought","GitHub","R"],"extracted_tech_keywords":["AI","ML","NLP","Transformers","chain of thought","R","Go","Git","GitHub","Modal"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/microsoft-introduces-multimodal-large-language-model-kosmos-1\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10039166,"title":"Guide To Video GPT: A Transformer-Based Architecture For Video Generation","content":"Video GPT is a novel machine learning architecture that employs likelihood-based generative modelling for video synthesis. It has been recently introduced by Wilson Yan, Yunzhi Zhang, Pieter Abbeel and Aravind Srinivas. (research paper). Before going into the detailed workings of Video GPT, we will have a quick look at some of its background terminologies. Autoencoder is an artificial neural network model belonging to the unsupervised learning category. It reduces the dimensionality of input data by ignoring any noisy data. It learns compressing and encoding input data. It then reconstructs the data from that encoded form to a new data representation resembling the original one before encoding. Visit this page to know more about autoencoders. Latent space represents compressed data such that similar data points lie in proximity to each other. Read this article to read about it in detail. Variational AutoEncoder (VAE) is the one that does not give out a single value for each of the encoding dimensions. Instead, it outputs a probabilistic distribution for each attribute in the latent space. Check out this weblink for more details. Vector Quantization (VQ) is an encoding-decoding technique in which an encoder is fed with input vectors. It gives out the index of the closest codeword, which is provided to the decoder. The decoder then recognizes the input vector from it. VQ-VAE (Vector Quantized Variational AutoEncoder) adds a discrete codebook to a standard autoencoder. It compares the output of the encoder network to all the vectors of the codebook; the closest vector is then fed to the decoder network. Thus, the VQ concept is combined with VAE. GPT (Generative Pre-Training) is a pre-trained language model on a large corpus of text and then fine-tuned for required tasks. (article on OpenAI’s GPT). Self-attention: Consider three vectors in a deep learning task viz. ‘Query (Q)’, ‘key (K)’ and ‘value (V)’. The term ‘attention’ means query and key vectors get multiplied such that the resultant vector of probabilities decides the value to be passed on to the subsequent layer. ‘Self-attention’ is the case where all the three vectors Q, K and V are the same. Find the research paper on ‘attention’ here. Image source The above figure explains ‘attention’ where Q and K vectors are first multiplied using matrix multiplication. The result then goes through a softmax function which creates a probability distribution which is then multiplied with V. Overview of Video GPT Video GPT is a simple model architecture that uses VQ-VAE and learns from an inputted raw video its downsampled discrete latent representations. It employs 3D convolutional networks and self-attention. Image source: Research paper The above figure explains the working of the Video GPT architecture. LHS of the figure depicts the first stage of operation, which is nothing but training a usual VQ-VAE model. In the second stage in sequence (RHS), the raw video data is encoded by VQ-VAE into latent sequences. At the decoding end, these latent sequences are sampled and converted into a new video sample (by VQ-VAE) resembling the original one. Pre-trained VQ-VAE models used by Video GPT bair_stride4*2*2 : trained on 64*64-dimensional videos (with 16 frames) taken from the BAIR Robot Pushing dataset. ucf101_stride4*4*4 : trained on 128*128 dimensional videos (with 16 frames) taken from the UCF-101 dataset. kinetics_stride4*4*4 : trained on 128*128-dimensional videos (with 16 frames) taken from the Kinetics-600 dataset. kinetics_stride2*4*4 : trained on the same data as kinetics_stride4*4*4 but with latent temporal codes that are twice larger, resulting in better video reconstruction. Note: The strides mentioned in the above models denote the amounts of downsampling across THW (number of images in a batch, height of image, width of image) for encoder structures. Practical implementation Here’s a demonstration of how to generate video using Video GPT. The code has been implemented using Python 3.7.10, matplotlib 3.2.2, torch 1.7.1, torchvision 0.8.2, and scikit-video 1.1.11 versions. Step-wise implementation of the code is as follows: Install Video GPT from GitHub. !pip install git+https:\/\/github.com\/wilson1yan\/VideoGPT.git Install scikit-video, a Python library for video processing. !pip install scikit-video av Import required libraries and modules. import os import matplotlib.pyplot as plt from matplotlib import animation from IPython.display import HTML import torch from torchvision.io import read_video, read_video_timestamps from videogpt import download, load_vqvae from videogpt.data import preprocess Create a dictionary of videos to choose from for video reconstruction. vid = { 'breakdancing': '1OZBnG235-J9LgB_qHv-waHZ4tjofiDgj', 'bear': '16nIaqq2vbPh-WMo_7hs9feVSe0jWVXLF', 'jaywalking': '1UxKCVrbyXhvMz_H7dI4w5hjPpRGCAApy', 'cartoon': '1ONcTMSEuGuLYIDbX-KeFqd390vbTIH9d' } Here, we are using ‘kinetics_stride2*4*4’ model. “”” Set up and run CUDA operations which are identical to CPU tensors but computations are performed using GPU “”” dev = torch.device('cuda') Download the model vqvae = load_vqvae('kinetics_stride2x4x4', device=dev).to(dev) Select the video from ‘vid’ to be reconstructed. vid_name = ‘bear’ Initialize resolution of video to be constructed. It must be divisible by encode image stride which is 2*4*4 here. resolution = vqvae.hparams.resolution Initialize duration of the sequence of frames to be displayed. seq_length = 64 Download the video file. vid_fname = download(vid[vid_name], f'{vid_name}.mp4') Decode the entire video frame-by-frame and record a list of frames’ timedtamps. pts = read_video_timestamps(vid_fname, pts_unit='sec')[0] Read the video from specified .mp4 file and get its audio and video frames. video = read_video(vid_fname, pts_unit='sec', start_pts=pts[0], end_pts=pts[seq_length - 1])[0] Preprocess the video using its THWC values (T: number of images in a batch, H: height of image, W: width of image, C: number of color channels) video = preprocess(video, resolution, seq_length).unsqueeze(0).to(dev) Encoding and decoding step #Encode the video using VQ-VAE so that it creates latent sequences with torch.no_grad():  #disable the gradient computation enc = vqvae.encode(video) #Decode the latent sequences to reconstruct the video vid_recon = vqvae.decode(enc) #Clamp the reconstructed video tensor’s values in the range [-0.5,0.5] vid_recon = torch.clamp(vid_recon, -0.5, 0.5) Note: The clamp() method used aboe works as follows: If we define the range as say [-0.5,0.5], then if the tensor has a value say -1.7, it will be clamped to -0.5 since it’s less than -0.5 and hence out of the range. Similarly, if there is an element greater than the range’s upper limit 0.5, say 1,8, it will be clamped to 0.5. Visualize the reconstructed video. #Concatenate original and reconstructed video for visualizing both videos = torch.cat((video, vid_recon), dim=-1) #Permute the dimensions of the videos from C*T*H*W to T*H*W*C #C: number of color channels #T: number of images in a batch #H: height if image #W: width of image videos = videos[0].permute(1, 2, 3, 0) \"\"\" Convert the CUDA variables to NumPy. Since NumPy does not support CUDA, GPU to CPU transition is to be done first and then cange the type to unsigned integer \"\"\" videos = ((videos + 0.5) * 255).cpu().numpy().astype('uint8') #Create a matplotlib figure fig = plt.figure() #Title of the plot plt.title('Original video (left), Reconstructed video (right)') #Disable the axes plt.axis('off') #Display the plot img = plt.imshow(videos[0, :, :, :]) plt.close() Define a function for drawing a clear frame def init(): img.set_data(videos[0, :, :, :]) Define a function to be called at each frame for animation. def animate(i): img.set_data(videos[i, :, :, :]) return img Create an animation by repeatedly calling the animate() function defined above. anmt = animation.FuncAnimation(fig, animate, init_func=init, frames=videos.shape[0], interval=100) Convert the animation to HTML5 video tag HTML(anmt.to_html5_video()) Output video: Code source: GitHubGoogle colab notebook of the above implementation. References Research paperGitHub repository","excerpt":"Video GPT is a novel machine learning architecture that employs likelihood-based generative modelling for video synthesis.","categories":["Deep Tech"],"tags":["AI Video Generation Models","Generative Pre-Trained Transformer","Guide"],"author_name":"Nikita Shiledarbaxi","publish_date":"2021-04-29T12:00:00","publication_year":"2021","word_count":1193,"keywords":["NumPy","machine learning","TPU","OpenAI","AI","neural network","ML","AI Video Generation Models","Colab","deep learning","Generative Pre-Trained Transformer","Matplotlib","Guide"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","neural network","OpenAI","Colab","NumPy","Matplotlib","TPU"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/guide-to-video-gpt-a-transformer-based-architecture-for-video-generation\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":66751,"title":"Latest Data Science And Analytics Jobs For Freshers","content":"With the growing importance of data among organisations, data scientists and analysts have been successfully adding value to business models by obtaining meaningful insights from large piles of structured and unstructured data. This includes defining goals, identifying opportunities, predicting future outcomes, detecting risks and other such. In this article, we list down the 10 latest data science and analytics jobs for freshers: – (The list is in no particular order) 1| Data Scientist – QuantumBlack Location: Gurgaon Responsibilities: As a data scientist at QuantumBlack, you will work on complex datasets to solve real-world problems, develop data science-related products and solutions for the data science team and clients. Besides, you will write highly optimised code to advance the internal data science toolbox, work in a multi-disciplinary environment with specialists in machine learning, engineering and design, attend conferences such as NIPS and ICML as one global team. You will also have the opportunity to share as well as learn from your colleagues. Click here to apply. 2| Principal Data Scientist at F5 Location: Hyderabad Responsibilities: Being a data scientist at F5, you will be responsible for the design and development of a platform for telemetry and predictive analytics to build a system based on real-time events. Also, you have to ensure that the platform is scalable to support machine learning. You’ll work with architects, product management, cross-functional teams, and other stakeholders to help define and implement solutions. Click here to apply. 3| Data Analyst at Institutional Shareholder Services Inc. Location: Mumbai Responsibilities: As a data analyst, you will search, collect and verify data (of basic and intermediate levels of complexity) for companies. Besides, you will have to comply with established work process standards to ensure the quality of data collected, perform root-cause analysis if data issues are identified, update internal documents for performance metrics monitoring, and other such. Click here to apply. 4| Data Science Engineer at Laurus Labs Limited Location: Pune Responsibilities: As a data science engineer, you will design, develop, and maintain web applications, including high-availability, low-latency systems, help make build vs buy vs maintain decisions considering business and technology needs. As a part of the job role, you will have to lead the design and implementation of ETL processes as well as guide the team in maintaining high standards of code quality and testing practices. Click here to apply. 5| Data Scientist at Efftronics Systems Pvt Ltd. Location: Vijayawada Responsibilities: As a data scientist, your responsibility is to collaborate with the business people to answer business-related queries with the help of data mining techniques such as pattern detection, graph analysis, or statistical analysis. You will also have to build strategic recommendations on data collection integration as well as retention requirements incorporating business requirements and knowledge of best practices and develop innovative and effective approaches to solve analytics problems and communicates results and methodologies, and other such. Click here to apply. 6| Data Scientist at Wonderment Apps Location: Pune Responsibilities: As a data scientist, you will be expected to possess solid experience in applying machine learning algorithms and statistical techniques to build highly accurate models. You need to acquire a deep technical understanding of the platform, and work with other team members to ensure the timely delivery of the planned tasks and critically assess and monitor the efficiency and\/or effectiveness of the models. Click here to apply. 7| Pharmacy Data Analyst at Evolent Health Location: Pune Responsibilities: As a data analyst, you will code, run, maintain and enhance various pharmaceutical analytics modelling (e.g., pricing, client proposals, formulary optimisation, staffing models) and reports (e.g., budget, utilisation, special programs like MTM). You will develop as well as code pharmacy specific rules into the rule engine, maintain and manage an active pipeline of rules need to be built appropriately, track performance and outcomes of those programs, in coordination with client working teams, and communicate findings to the appropriate person, among others. Click here to apply. 8| Data Analyst at Laurus Labs Ltd Location: Pune Responsibilities: As a data analyst, you will design, develop and implement databases, data collection systems, data analytics and other strategies to support the pricing operational, tactical and strategic reporting. You will also perform quality assurance on generated results to ensure accuracy and consistency, design, evaluate and monitor key metrics, understand root causes of changes in metrics and other such. Click here to apply. 9| Research Engineer – Data Analytics & Artificial Intelligence (Machine Vision) at Siemens Technology Location: Bangalore Responsibilities: As a research engineer – data analytics & artificial intelligence (machine vision), you will analyse the complex datasets by developing advanced ML and deep learning pipelines that are based on business initiatives. You will also build innovative, practical and robust real-world solutions for problems in traffic management, autonomous building control for energy and comfort optimisation, medical image analysis for assisting pathologists\/radiologists, smart robotics for Industry 4.0, and other such. Click here to apply. 10| Data Scientist at Yoodle Location: Bangalore Responsibilities: As a data scientist, you will work independently or in a team to solve complex problems and create scalable models\/algorithms that will be integrated into Yodlee’s tools and products, come up with actionable ideas to solve problems faced by product managers senior leadership then implement those ideas, among others. Click here to apply.","excerpt":"With the growing importance of data among organisations, data scientists and analysts have been successfully adding value to business models by obtaining meaningful insights from large piles of structured and unstructured data. This includes defining goals, identifying opportunities, predicting future outcomes, detecting risks and other such. In this article, we list down the 10 latest […]","categories":["AI Features"],"tags":["AI Jobs","Applications of Data Mining","business analytics mba","data analyst","Data analyst jobs","data analysts India","Data Science Career","Data Science Jobs","latest technology in machine learning","MBA Business Analytics","principal data scientist","robust analytics strategy"],"author_name":"Ambika Choudhury","publish_date":"2020-06-05T16:00:00","publication_year":"2020","word_count":875,"keywords":["business analytics mba","deep learning","R","data science","artificial intelligence","principal data scientist","MBA Business Analytics","latest technology in machine learning","analytics","Go","machine learning","AI","ML","Data analyst jobs","Data Science Jobs","Applications of Data Mining","data analysts India","Data Science Career","robust analytics strategy","AI Jobs","data analyst","predictive analytics"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","data science","analytics","predictive analytics","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/latest-data-science-and-analytics-jobs-for-freshers\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10005195,"title":"Guide To OpenCV Functions For Image Processing","content":"OpenCV is used as an image processing library in many computer vision real-time applications. There are thousands of functions available in OpenCV. These simple techniques are used to shape our images in our required format. As we know an image is a combination of pixels, for a color image we have three channels with pixels ranging from 0 to 225, and for black & white-colored images has only one change ranging from 0 to 1. In this article, we will demonstrate some important image processing steps and we will do modifications to an image by manipulating its pixels using OpenCV functions. Topics that will be covered this article Edge detection using Canny filterConverting a color image to Grayscale & vice versa.Smoothing techniquesMorphology techniques Converting a Color Image to a Grayscale Image This is the basic technique that is used to convert the color image into grey shades. A color image consists of 3 channel depth while using grayscaling it reduces the depth of the image to 1 channel. It reduces model complexity. For many applications like edge detection, photo sketch, cartooning image we use grayscale converted images. In the below code snippet we are converting a color image to a grayscale image. import cv2 image = cv2.imread('\/content\/Screenshot (14).png') grayImage = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY) cv2_imshow(grayImage) In the above output, we converted the colored image to grayscale. Edge detection In edge detection, it gives an outline of the image by extracting edges from the image. We are using a canny filter to perform this task. Based on the threshold values, a canny filter detects the edges. The higher thresholds give cleaner images compared to lower thresholds gives a clumsy output. import cv2 import numpy as np img = cv2.imread('\/content\/Screenshot (14).png',0) edge_det = cv2.Canny(img,100,200) cv2_imshow(edge_det) Here we had given a threshold value 100,200 and you can try with different threshold values. This edge detection is used for segmentation, restoration, picture enhancement, pattern recognition. Smoothing Techniques The image smoothing technique is performed using a filter. By convolving the image, it reduces the noise in the image by adding a blurring effect on the edges. There are different types of smoothing techniques we perform depending on the input image. Average using cv2.blur() In this technique, depending on the size of the filter convolution is performed but it averages the pixels under the filter and assigns the value to the centre pixel under the filter. import cv2 import numpy as np from google.colab.patches import cv2_imshow img = cv2.imread('\/content\/download (14).jpg') blur = cv2.blur(img,(5,5)) cv2_imshow(blur) cv2_imshow(img) normal image Median Blur using cv2.medianBlur() In this technique, it calculates the median of the pixels under the filter and it replaces the center value under the filter with the median value, positive odd integer to be assigned as filter size to perform the median blur technique. img = cv2.imread('\/content\/download (14).jpg') median_blur = cv2.medianBlur(img,5) cv2_imshow(median_blur) cv2_imshow(img) normal image after applying median blur Comparing smoothing techniques averaging and median blur gives similar results because one calculates on averaging pixels and another on calculating the medians under the filter. Gaussian Blur using cv2.GaussianBlur() In this technique, there is another parameter called sigma x and sigma y. It performs a blurring of an image by using parameters sigma x & sigma y. If it is set to zero then it calculates based on the size of its kernel. img = cv2.imread('\/content\/download (14).jpg') Gaussian_blur = cv2.GaussianBlur(img,(5,5),0) cv2_imshow(median_blur) cv2_imshow(img) normal image after applying Gaussian blur As we can see in the output, by blurring the image we tried to reduce the effect of rain Morphological Techniques These are the techniques used to manipulate the image sizes, these techniques are used only after converting the image into grayscale. These techniques are used for removing noise, correcting the imperfections in the data, and to make clear images. Erosion and Dilation are mostly used in morphological techniques. Erosion This technique removes the boundary pixels from the input by just passing the filter on the image. Depending on the size of the kernel, it removes the boundary pixels from the input image. It performs similar to soil erosion so they named this technique as erosion. img = cv2.imread('\/content\/Screenshot (14).png') kernel = np.ones((5,5),np.uint8) img_erosion = cv2.erode(img, kernel, iterations=1) img_dilation = cv2.dilate(img, kernel, iterations=1) cv2_imshow(img) cv2_imshow(img_erosion) normal image image erosion In the above output by using erosion technique  we tried to make spiderman thinner. Dilation The above technique removes the boundary pixels but in Dilation. It adds the additional pixels to the input. It is used when the pixels are missing in the image. In general practices, we apply erosion to shrink the image to remove noises, and then by applying the dilation, there will be no loss of pixels. img = cv2.imread('\/content\/Screenshot (14).png') kernel = np.ones((5,5),np.uint8) img_dilation = cv2.dilate(img, kernel, iterations=1) cv2_imshow(img) cv2_imshow(img_dilation) normal image image after dilation In the above output using the dilation technique, we tried to make spiderman a little fatter. Conclusion In this article, we have illustrated different types of filters which play a key role in image processing while working on computer vision applications. Using OpenCV and support of inbuilt functions in OpenCV, we performed these implementations by just writing a few lines of codes.","excerpt":"OpenCV is used as an image processing library in many computer vision real-time applications. There are thousands of functions available in OpenCV. These simple techniques are used to shape our images in our required format. As we know an image is a combination of pixels, for a color image we have three channels with pixels […]","categories":["Deep Tech"],"tags":["Image processing techniques"],"author_name":"Prudhvi varma","publish_date":"2020-08-20T18:00:00","publication_year":"2020","word_count":857,"keywords":["Go","NumPy","TPU","AI","computer vision","Image processing techniques","Colab","Ray","OpenCV","RAG","R"],"extracted_tech_keywords":["AI","computer vision","Ray","Colab","OpenCV","NumPy","RAG","TPU","R","Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/guide-to-opencv-functions-for-image-processing\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10122038,"title":"UiPath Stock Plunges Nearly 30% as CEO Rob Enslin Abruptly Resigns","content":"UiPath shares plummeted nearly 30% in after-hours trading Wednesday after the robotic process automation software company announced the surprise resignation of CEO Robert Enslin and provided disappointing guidance, overshadowing better-than-expected first-quarter earnings results. The New York-based company said Enslin will step down as CEO and board member effective June 1, 2024, after just two years in the role. the company said co-founder and former CEO Daniel Dines, who currently serves as Executive Chairman and Chief Innovation Officer, will retake the reins as CEO. Enslin will remain as an advisor during the transition. For the first quarter ended April 30, UiPath reported adjusted earnings per share of $0.13, beating the consensus estimate of $0.12. Revenue grew 16% year-over-year to $335 million, also surpassing the $333.04 million analysts expected. Annualised renewal run-rate (ARR), a key metric, increased 21% to $1.508 billion. However, UiPath’s outlook fell short of expectations. For Q2, the company guided revenue of $300-305 million, well below the $342.07 million consensus. It lowered its full-year fiscal 2025 revenue forecast to $1.405-1.41 billion, down from its prior outlook of $1.55-1.56 billion and missing the $1.56 billion analysts anticipated. On the earnings call, Dines acknowledged challenges with sales execution and elongating sales cycles for larger deals. He said some investments have made UiPath less agile but expressed optimism in the company’s long-term prospects, especially around generative AI, which he sees as a “secular trend” benefiting the business. Under his renewed leadership, Dines plans to refocus on product innovation. Analysts said the abrupt CEO change indicates Enslin’s failure to drive faster growth, and weak guidance suggests deeper issues at UiPath. Once valued at nearly $36 billion after its April 2021 IPO, the company has seen its stock struggle, down over 50% from the IPO price even before Wednesday’s after-hours plunge. “Given the increase in spend during the quarter, I guess there must have been some late surprises in a quarter of larger deals not coming through,” said Holger Mueller of Constellation Research. “The odd thing is that the board didn’t want to have Dines leading the company two years ago, but it has now put him back in place. Potentially, the board might be looking for a new CEO, or else it’s going to focus its efforts on building a next-generation product for the generative AI era under Dine’s leadership.” Founded in 2005, UiPath makes software that helps companies automate repetitive business tasks. Its platform is powered by AI models that learn how employees perform common tasks in business applications. While its tools have been hailed as a game-changer, the company faces increasing competition from tech giants and generative AI upstarts that could steal business. “In the long run, AI will be a tailwind for companies that can apply machine intelligence to end-to-end automation,” said Dave Vellante, Chief Analyst at TheCUBE Research. “In the near term customers may feel it’s easier to do full automation with gen AI, but I think they’ll find they need deeper relationships and tech to actually realise significant value. In the meantime, firms like UiPath have to educate customers on how gen AI combined with end-to-end automation can be achieved.”","excerpt":"Analysts said the abrupt CEO change indicates Enslin’s failure to drive faster growth, and weak guidance suggests deeper issues at UiPath","categories":["AI News"],"tags":["uipath"],"author_name":"Shyam Nandan Upadhyay","publish_date":"2024-05-30T09:24:09","publication_year":"2024","word_count":522,"keywords":["Go","AI","IPO","RPA","innovation","BERT","automation","llm_models:BERT","generative AI","uipath","R"],"extracted_tech_keywords":["AI","generative AI","R","Go","BERT","automation","RPA","innovation","IPO","llm_models:BERT"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/uipath-stock-plunges-nearly-30-as-ceo-rob-enslin-abruptly-resigns\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10167173,"title":"OpenAI Might Have a Hidden Agenda Behind Going Open-Weight Again","content":"Fresh from the Ghibli meme wave, OpenAI chief Sam Altman recently confirmed that a powerful new open-weight model with strong reasoning capabilities is on its way. The move is sure to help OpenAI gain more validation and credibility among developers and enterprises, who prefer using open-source weights (not to be confused with open-source models) as it allows them to customise them and host them locally. “We are planning to release our first open-weight language model since GPT-2. We’ve been thinking about this for a long time, but other priorities took precedence. Now, it feels important to do,” he wrote in his post on X. This comes in the backdrop of OpenAI raising $40 billion at a $300 billion post-money valuation. However, why did there happen to be a sudden change of heart? OpenAI doesn’t make decisions without a reason. On the surface, it seems that DeepSeek’s success has influenced this move. Earlier this year, Altman, in a Reddit AMA, said OpenAI has been on the wrong side of history concerning open source. Previously, Altman even asked users on X whether they prefer an o3-mini-level model or a phone-sized version. “It’s a welcome move,” said Madhav Krishna, founder of Vahan.ai, in a conversation with AIM about OpenAI’s decision to release open weights. He mentioned that his company will work closely with OpenAI to explore how they can integrate the new model into their operations. Notably, Vahan.ai uses OpenAI’s API to build AI agents that address hiring challenges for blue-collar workers in India. The platform effectively connects job seekers with appropriate opportunities while assisting employers in identifying the best candidates. Similarly, another customer of OpenAI APIs, HealthifyMe, told AIM that they are excited to see OpenAI exploring an open-weight model. “While there’s buzz around open-source alternatives like Meta’s Llama and DeepSeek’s R1, for us at Healthify, the main priority has always been robust, reliable personalisation. We’ve built significant value using fine-tuning methods on our existing solutions, which meet our specific health coaching needs,” said Abhijit Khasnis, CTO of HealthifyMe. “That said, an open-weight model might offer us new ways to tailor outputs for specialised applications or even reduce costs if we can run models on local hardware. However, until we see proven use cases that directly benefit our user engagement or operational efficiencies, we’ll continue to closely monitor the developments,” he added. Meanwhile, DeepSeek released its advanced reasoning model R1 earlier this year, followed by an upgraded V3 in March, both of which are fully open source. R1 surpassed OpenAI’s o1 model on several benchmarks. “The biggest revelation from DeepSeek is that open source has won. For a 1% difference in performance, it will be difficult for OpenAI to justify its pricing when the competition is free and formidable,” said Kai-Fu Lee, founder of AI startup 01.AI, in a recent interview with Bloomberg. He added that the company will also last infinitely because its founder has enough money to fund it at the current level and has reduced computing costs by a factor of five to 10. “With such a formidable competitor, I think Sam Altman is probably not sleeping well,” he quipped. Why Now? Open-sourcing models could help OpenAI reconnect with the developer community and build a more efficient, transparent ecosystem. This, in turn, could lower API costs and make the company more competitive in the enterprise space. “OpenAI aims to retain its users by offering an open-source model that can be fine-tuned for specific business needs,” Shantipriya Parida, senior AI scientist at AMD Silo AI, told AIM. He added that with free and better open-source models available, users are questioning the need to spend money on OpenAI subscriptions. Going by last year’s numbers, OpenAI made $3.4 billion in total revenue, with ChatGPT emerging as the primary revenue driver, contributing $1.9 billion. Meanwhile, revenue from the API was $510 million. This indicates that many developers were not attracted towards OpenAI’s API. However, according to a recent report by Bloomberg, OpenAI is on track to more than triple its revenue this year, reaching $12.7 billion. Harneet SN, founder of Rabbitt.ai, told AIM that open-weight models are crucial in sectors like healthcare, defense, and education, where on-premise deployment on their own devices is required. These industries cannot easily delegate or outsource their data to proprietary models. Having said that, Altman mentioned that the company will now be hosting developer events to gather feedback and later experiment with early prototypes. “We’ll start in San Francisco in a couple of weeks, followed by sessions in Europe and Asia-Pacific.” He further said that he is excited to see what developers build and how large companies and governments use it in cases where they prefer to run a model themselves. This move will also help OpenAI gain the trust of the US government. Taking a jibe at Meta’s Llama 2 restrictions, Altman said, “We will not do anything silly, like saying that you can’t use our open model if your service has more than 700 million monthly active users. We want everyone to use it!” Indian IT minister Ashwini Vaishnaw also took a swipe at OpenAI, pointing out how it transitioned its models to closed source after GPT-2, suggesting that Indian open source model makers today might take a similar approach. “We should also change [OpenAI’s] name,” he quipped. Not surprisingly, Altman has been buttering up India lately. “What’s happening with AI adoption in India right now is amazing to watch. We love to see the explosion of creativity—India is outpacing the world,” he wrote in a post on X. He even posted a Ghibli-style image of himself playing cricket. Meanwhile, Meta recently announced that its open-source AI model family, Llama, has surpassed 1 billion downloads. This milestone marks a sharp rise from 650 million downloads in December 2024—a 153% surge in just three months. Besides Meta, several companies have recently launched open-source models. Mistral’s latest model, Small 3.1, Google’s Gemma 3, and Cohere’s Command A all claim to rival proprietary models while using fewer compute resources. On the other hand, Chinese tech giants have also upped their game. Alibaba added voice and video chat capabilities to Qwen Chat and released its brand-new open-source model, Qwen2.5-Omni-7B. Similarly, releasing an open-weight model—where the trained parameters are shared, but the code and training data remain proprietary—will allow OpenAI to stay relevant in a market that increasingly favours accessibility while preserving some control over its intellectual property.","excerpt":"This comes in the backdrop of OpenAI raising $40 billion at a $300 billion post-money valuation.","categories":["Global Tech"],"tags":["OpenAI"],"author_name":"Siddharth Jindal","publish_date":"2025-04-03T11:38:04","publication_year":"2025","word_count":1067,"keywords":["Go","ChatGPT","API","TPU","OpenAI","AI","Aim","Gemma 3","Rust","R"],"extracted_tech_keywords":["AI","ChatGPT","OpenAI","Gemma 3","Aim","TPU","R","Go","Rust","API"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/openai-might-have-a-hidden-agenda-behind-going-open-weight-again\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":42768,"title":"Elon Musk Wants To Merge AI And Humans With Neuralink, Plans To Hire More Scientists","content":"Imagine someone who is blind for life gets an option to plug in tiny electrodes into their brain so that they can see again. Elon Musk’s newest venture Neuralink promises to bring a future where you can have a device that when plugged into the brain can control robotic arms, can write into a text editor and maybe solve vision for blind people. For anyone who follows Elon Musk knows that his ventures are not for the weak-hearted. Digging tunnels through cities, sending rockets to Mars, building houses with solar roofs and other endeavours that can’t be accomplished by an ordinary human being even in one whole life has been done by Musk quite effortlessly. But Musk’s latest obsession, if possible, is even more ambitious. Musk wants humans to merge with artificial intelligence as a protective measure when AI starts ruling the world. As a precursor to solving the merging problem, Musk and his team at Neuralink will have to solve other problems on the way. These side family of problems fall in the realm of information theory, machine learning, precision robotics, metallurgy, chip design and other varied fields. In the process of having our brain connected to outside devices, the team at Neuralink aim to find solutions to other problems like nerve-related diseases and gain more insight into how the brain functions. At the best Neuralink will change our future and the worst it will give us insights into how our brain functions using smart electrodes (if no one is harmed during the testing). Hiring For The Mission Neuralink today hosted a presentation on Youtube to introduce the company and its mission to the world. Musk started the session with a message that the main function of the live presentation was to hire the best in the field. Musk wants to attract prolific people to advance the brain-computer interface technology forward. While talking about the most fundamental motivations of Neuralink, Musk said that treating brain disorders is important to ensure a healthy life for all. Apart from treating brain disorders, enhancing brain capacity will be another important mission of the company. Neuralink is hiring Musk continued to state that, “This is going to sound a little weird. But ultimately (we want to) achieve a symbiosis with artificial intelligence.” The technology built by Neuralink will mostly have a goal to build a highway from your brain to other external devices to transfer and make sense of information that will be produced by the neurons. Neuralink works with a thread that goes into a human’s skull and reads activities from the neurons. These activities are transferred to a chip that will process the information. The thread that goes inside your brain, which will be produced by Neuralink is supposed to be 1\/10th the width of human hair. Hence unlike other brain surgeries, the insertion by Neuralink will leave no scars and will be safe. Neuralink plans to build dense and small sensors that can do more than any other device in the past. This helps them to create rich visuals for the blind. Recording signals from the brain and stimulating them can also help in curing many diseases such as Parkinson’s disease, Epilepsy and others. Next Chapter Of Science And Technology Neuralink’s President Max Hodak talked about the importance of information theory and how it applies to the field of brain-machine interface. Hodak underlined that the work done by Neuralink has been done standing on the shoulders of many giants who worked in the field from the 1950s. Neuralink though should be credited for bringing back the field of brain interfaces back in vogue. Whenever Silicon Valley takes up a cause, it promises to garner some spotlight. The team at Neuralink are a testimony to the attention that the field has got. Scientists and entrepreneurs who were working on important questions mostly in obscurity are now leading a much talked about the medical and technological revolution which is looked by the world because of the publicity provided by Musk. Hodak also went onto underline that the device created by Neuralink will be with humans for decades, will be wireless and most importantly have practical bandwidth. Muskian Timelines And Targets Like any other Musk enterprise, Neuralink also has aggressive targets. The company will have to adhere to FDA guidelines but aim to start clinical trials by 2020. This means many first time patients and users will be able to experience Neuralink’s products as early as next year. During the Q and A, President Hodar also talked about providing brain enabled APIs for developers to work on and hinted that it can turn into a platform. Musk said the cost of these devices is a very important detail and the team is working actively on making the economics of the product to work out.","excerpt":"Imagine someone who is blind for life gets an option to plug in tiny electrodes into their brain so that they can see again. Elon Musk’s newest venture Neuralink promises to bring a future where you can have a device that when plugged into the brain can control robotic arms, can write into a text […]","categories":["AI Features"],"tags":["Elon Musk","ML","neuralink"],"author_name":"Abhijeet Katte","publish_date":"2019-07-17T16:54:05","publication_year":"2019","word_count":801,"keywords":["Go","API","artificial intelligence","machine learning","programming_languages:R","AI","ML","neuralink","programming_languages:Go","Elon Musk","Aim","ViT","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","Aim","R","Go","API","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/elon-musk-wants-to-merge-ai-and-humans-with-neuralink-plans-to-hire-more-scientists\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10053988,"title":"A Guide to Different Visualizations with Plotly Express for Practitioners","content":"Plotly Express is a free and open-source Python visualization library for creating interactive and beautiful visualizations. It’s a great way to find patterns in a dataset before diving into machine learning modelling. In this article, we will look at how to use it in an example-driven way, giving you an overview of Plotly Express using the built-in datasets, explaining it from the ground up, and covering all of the most commonly used charts. The following are the points and plots that this article will cover. Table of Contents About Plotly ExpressVisualization With Plotly ExpressBasic ChartsTree Maps3D PlotMulti-Dimensional PlotPolar Charts Let’s start the discussion by understanding what is Plotly Express? About Plotly Express Plotly Express is a new Python visualization library that acts as a wrapper for Plotly, exposing a simple syntax for complex charts. It was inspired by Seaborn and ggplot2 and was specifically designed to have a terse, consistent, and easy-to-learn API: with a single import, you can make richly interactive plots with faceting, maps, animations, and trendlines in a single function call. Plotly Express, like Plotly.py, includes on-board datasets, colour scales, and themes, and it’s completely free with the permissive open-source MIT license, you can use it however you want (yes, even in commercial products!). Plotly Express is fully compatible with the rest of the software. Plotly Express is a plotly.express module (usually imported as px) that contains functions that can create entire figures at once. Plotly Express is a part of the Plotly library that comes pre-installed and is the recommended starting point for most common figures. Every Plotly Express function returns a plotly.graph object.figure instance and uses graph objects internally. Plotly Express includes over 30 functions for creating various types of figures. The API for these functions was carefully designed to be as consistent and easy to learn as possible, allowing you to easily switch from a scatter plot to a bar chart to a histogram to a sunburst chart during a data exploration session. Visualization with Plotly Express Basic Charts Let us have a look at how the basic plots can be created using Plotly express. Scatter Plot Scatterplots are excellent for determining whether two numerical variables have a relationship or correlation. Each data point is represented as a marker point by px.scatter, whose location is determined by the x and y columns. ! pip install plotly import plotly.express as px df = px.data.iris() fig = px.scatter(df, x=\"sepal_width\", y=\"sepal_length\", color=\"species\") fig.show() With the trendline argument in Plotly Express, you can add an Ordinary Least Squares regression trendline to scatter plots. Marginal distribution plots, which are small subplots above or to the right of the main plot and show the distribution of data along only one dimension, can also be used. Plotly Express functions like scatter and histogram have marginal distribution plot capabilities built-in. fig = px.scatter(df, x=\"sepal_width\", y=\"sepal_length\", color=\"species\", marginal_y=\"violin\", marginal_x=\"box\", trendline=\"ols\", template=\"simple_white\") fig.show() Bar Plot When displaying a categorical column and a numerical column, a Bar Plot is an excellent visualization. It displays the number of a specific numerical column in each category. Plotly Express makes it very simple to create one. import plotly.express as px df = px.data.tips() fig = px.bar(df, x=\"sex\", y=\"total_bill\", color=\"smoker\", barmode=\"group\") fig.show() Histogram A graphical representation of a binned numerical data distribution is a histogram. The count for each bin is then displayed. Aggregation functions such as sum and average can be used to combine plotly data. The data to be binned in Plotly can also be categorical. Here’s an illustration: import plotly.express as px df = px.data.tips() fig = px.histogram(df, x=\"total_bill\", color=\"sex\") fig.show() Tree Maps Sunburst Chart Sunburst plots depict hierarchical data that extends radially from root to leaves. Labels and parents’ attributes define the hierarchy. Children are added to the outer rings as the root grows from the center. Each row of the DataFrame is represented as a sector of the sunburst by px.sunburst. df = px.data.tips() fig = px.sunburst(df, path=['sex', 'day', 'time'], values='total_bill', color='time') fig.show() Funnel Chart Funnel charts are frequently used to visualize data at various stages of a business process. It’s a crucial mechanism in Business Intelligence for identifying potential process flaws. It’s used to track revenue or loss in a sales process at each stage, for example, and it shows values that are decreasing over time. Each stage is represented by a percentage of the total value. data = dict( number=[39, 27.4, 20.6, 11, 2], stage=[\"Website visit\", \"Downloads\", \"Potential customers\", \"Requested price\", \"invoice sent\"]) fig = px.funnel(data, x='number', y='stage') fig.show() 3D Plots The 3D function px.scatter 3d plots individual data in three-dimensional space, similar to the 2D scatter plot px.scatter. df = px.data.iris() fig = px.scatter_3d(df, x='sepal_length', y='sepal_width', z='petal_width', color='petal_length', size='petal_length', size_max=18, symbol='species', opacity=0.7) # tight layout fig.update_layout(margin=dict(l=0, r=0, b=0, t=0)) Multidimensional Plot Scatter Matrix A scatterplot matrix is a matrix that is linked to n numerical arrays (data variables) of the same length, $X1, X2,…, X n$. The scatter plot of the variable Xi versus Xj is displayed in cell (i,j) of such a matrix. To plot the scatter matrix for the columns of the data frame, use the Plotly Express function px.scatter matrix. By default, all columns are taken into account. df = px.data.iris() fig = px.scatter_matrix(df, dimensions=[\"sepal_width\", \"sepal_length\", \"petal_width\", \"petal_length\"], color=\"species\") fig.show() Parallel Coordinates Plot Each row of the DataFrame is represented by a polyline mark that traverses a set of parallel axes, one for each dimension, df = px.data.iris() fig = px.parallel_coordinates(df, color=\"species_id\", labels={\"species_id\": \"Species\", \"sepal_width\": \"Sepal Width\", \"sepal_length\": \"Sepal Length\", \"petal_width\": \"Petal Width\", \"petal_length\": \"Petal Length\", }, color_continuous_scale=px.colors.diverging.Tealrose, color_continuous_midpoint=2) fig.show() Polar Charts Scatter Polar Plot Data is represented on radial and angular axes in a polar chart. Polar data can be represented as scatter markers with px.scatter polar and as lines with px.line polar in Plotly Express. The r and theta arguments of px.scatter polar are used to specify radial and angular coordinates. Theta data are categorical in the example below, but numerical data are also possible and the most common cause. df = px.data.wind() fig = px.scatter_polar(df, r=\"frequency\", theta=\"direction\", color=\"strength\", symbol=\"strength\", size=\"frequency\", color_discrete_sequence=px.colors.sequential.Plasma_r) fig.show() Final Words With the help of interactive features, we can even dig deeper into our dataset and can grab more information straightway. With the Plotly Express, in this post we have some of the commonly used plots in a community like a bar chart, scatter plot, histogram. More interestingly, I also saw some of the fancy plots like polar chart, parallel coordinates chart, sunburst chart, funnel chart which can grab more information in the granular way of your data. References Official Documentation of Plotly ExpressLink for above codes","excerpt":"Plotly Express is a free and open-source Python visualization library for creating interactive and beautiful visualizations.","categories":["Deep Tech"],"tags":["Data Analytics","Data Science","Data Visualisation","Guide","iris dataset python","Machine Learning","plotly","Python"],"author_name":"Vijaysinh Lendave","publish_date":"2021-11-23T11:00:00","publication_year":"2021","word_count":1095,"keywords":["Go","iris dataset python","machine learning","Plotly","AI","AWS","Data Visualisation","Machine Learning","plotly","Python","Ray","Seaborn","RAG","Data Analytics","Data Science","R","Guide"],"extracted_tech_keywords":["AI","machine learning","Ray","Plotly","Seaborn","RAG","AWS","Python","R","Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/a-guide-to-different-visualizations-with-plotly-express-for-practitioners\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":18121,"title":"Data, Digitisation, And Digital India: How Data Analytics Can Facilitate Smoother E-governance","content":"When people think of data analytics, the images their minds often conjure are of its glitzy representation in popular media. They think of high-performance computing systems crunching through massive amounts of binary data, of fingers flying at keyboards at lightning speeds, of nifty 3D visualisations being analysed by a group of bespectacled researchers in white coats carrying tablets as clipboards. While all this – or something approaching all this – is no doubt an integral part of how data analytics works, what is usually missed out is what makes data analytics really significant in today’s day and age: its human impact. Sounds odd, doesn’t it? Data analytics, according to the popular perception, is supposed to be all about keywords that describe what it does; seamlessness, cutting-edge, value addition, process optimisation, unparalleled efficiency, optimised resource utilisation are all bandied about by everyone, from leading industry experts to the layperson with an interest in technology. But once we look beyond the trappings of fanciful jargon, what lies underneath is a realistic view of how the intelligence and insights enabled by data analytics can help in defining government policies, regulations, and initiatives that have a direct impact on the end-user – in this case, the citizens. Extracting value from information: Data analytics and its role in enabling eGovernance To better understand how data analytics can help in enabling smoother digital governance in India, we need to take a leaf out of the playbook of other advanced nations. A fitting case study would be that of the Fire Department in Austin, Texas. Despite generating massive amounts of data, the Austin Fire Department did not have any kind of visibility on critical parameters such as team workloads, response times, training session completion, travel times, and types of emergency calls etc.  This changed when the department decided to deploy Qlik to make sense of its data. The platform allowed everyone – from battalion chiefs and captains to lieutenants, specialists, and fire-fighters – to gain insights into their operations by visualising the key performance metrics and making associations between various datasets. Such was the efficacy that Qlik unlocked that the Austin Fire Department was able to free up an estimated $215,000 in payroll costs within the first year alone, along with almost 5,000 workforce hours spent in data compilation. The data-driven approach enabled by Qlik also helped the department in being better prepared for emergencies; with all the data right in front of the end-user, the department was able to swiftly allocate the optimum resources – from people to vehicles and equipment – for on-ground deployment. Another prominent example of how data analytics can help in enabling better eGovernance is that of the Department of Employment in Australia and its flagship ‘jobactive’ initiative. The initiative, which operates out of over 1,700 locations, connects job seekers with employers across Australia. Needless to say, this generates a lot of information, given the number of job seekers and third-party employment service providers associated with the job active initiative. As the amount of information it handled increased, the Department of Employment identified another growing concern: many job active providers were unable to draw insights into their performance and understand how to better results. The search for an analytics partner to manage the growing information database and extract relevant insights from it led the Department of Employment to Qlik. Leveraging Qlik’s state-of-the-art analytics platform, the department was able to deploy the first set of applications to be used by job active providers within three months – something of a record for a government organisation. Moreover, the interactive and interface was so easy to use, that it made it possible for all business users, regardless of their level of data competencies, to extract insights from multiple data-sets and compare their performance with others within the job active value chain with minimal hassle. This helped them in optimising their business operations, which, in turn, helped in unlocking more and better employment opportunities to jobseekers empanelled with the job active initiative. Way to the future: Moving towards a data-driven culture of governance Governments across the world today are sitting on heaps of information about their citizens. This holds especially true for fast-developing countries like India, which have been promoting rapid digitisation as a means of enabling better delivery of essential services such as banking, healthcare, education etc. to their citizens. Digital adoption has been spurred on by initiatives such as Digital India, which have been taking the benefits of digital technology beyond urban centres and making them an integral part of the everyday lives of people living in semi-urban and rural geographies. The need of the hour, therefore, is to tap into the unparalleled opportunity that this information pool presents by leveraging data analytics to support and improve eGovernance initiatives. As India moves towards a more digitised future, data and information will be the foundation upon which its smarter, brighter tomorrow will be built.","excerpt":"When people think of data analytics, the images their minds often conjure are of its glitzy representation in popular media. They think of high-performance computing systems crunching through massive amounts of binary data, of fingers flying at keyboards at lightning speeds, of nifty 3D visualisations being analysed by a group of bespectacled researchers in white […]","categories":["IT Services"],"tags":["Data Analytics","Digital India","digitisation"],"author_name":"Souma Das","publish_date":"2017-10-07T05:00:30","publication_year":"2017","word_count":819,"keywords":["Go","API","Digital India","AI","digitisation","ML","Git","RAG","GAN","analytics","CLIP","Data Analytics","R"],"extracted_tech_keywords":["AI","ML","analytics","RAG","R","Go","Git","API","CLIP","GAN"],"url":"https:\/\/analyticsindiamag.com\/it-services\/data-digitisation-digital-india-data-analytics-can-facilitate-smoother-e-governance\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10136569,"title":"India’s First AI Unicorn, Fractal Eyes $500 Mn IPO","content":"Fractal, India’s first AI unicorn, is reportedly planning to raise $500 million through an initial public offering (IPO) at a valuation of at least $3.5 billion, according to sources. The startup, known for its AI and advanced analytics solutions, is expected to pursue a public listing in the first quarter of 2025. The IPO will likely involve a combination of fresh equity and an offer for sale and Fractal is considering filing its DRHP with Sebi in November. This news follows earlier reports suggesting Fractal was preparing to file draft IPO papers in August, aiming to raise approximately $600 million at a $3 billion valuation. Founded in 2000 by Srikanth Velamakanni, Pranay Agrawal, and Ashwath Bhat, Fractal provides AI solutions to numerous Fortune 100 companies. Among its products is Qure.ai, a tool for medical imaging diagnostics. Most recently, the company launched its medical chatbot Vaidya.ai. Fractal became a unicorn in 2022 after securing $360 million from TPG Capital. To date, the startup has raised $685 million from investors, including TPG Capital, Khazanah Nasional, and Apax Partners. Fractal employs over 4,500 people across 17 locations, including the UK, Singapore, the Middle East, Australia, India, and the US. Fractal’s IPO plans align with a broader trend of tech startups in India seeking public listings, taking advantage of the ongoing bull run in the Indian equities market.","excerpt":"Fractal is considering filing its DRHP with Sebi in November.","categories":["AI News"],"tags":["AI Startups","Fractal","Fractal AI"],"author_name":"Mohit Pandey","publish_date":"2024-09-24T21:48:07","publication_year":"2024","word_count":224,"keywords":["Go","API","unicorn","programming_languages:R","AI","IPO","Fractal AI","Fractal","Aim","analytics","R","AI Startups","startup"],"extracted_tech_keywords":["AI","analytics","Aim","R","Go","API","startup","unicorn","IPO","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/indias-first-ai-unicorn-fractal-eyes-500-mn-ipo\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":46479,"title":"AI Will Be A Very Big Disruptor For Indian Society, Says Ex-Infy CEO Vishal Sikka","content":"Global business leader and former Infosys CEO Vishal Sikka believes AI will be a big disruptor for Indian society. The ex-Infy CEO was recently in the news for raising $50 million for his AI startup Vianai Systems which is currently operating in stealth mode. As reported, Sikka was recently approached by Prime Minister Narendra Modi to give a lowdown to Indian think tank Niti Aayog on how best to expand the reach of AI for public good. In order to make India a world leader in AI and expand the reach of AI in society, Sikka gave a presentation to members from 20 Union Ministries on how best to build the necessary infrastructure required to create a steady talent stream for AI. AI education at large scale can be the way forward for India, Sikka emphasised, adding how over the next 20-25 years, AI is going to be “very big disruptor” for the Indian society. “On the other hand, if we are able to bring AI education, the ability to build AI systems to India at a very large scale, and I’m talking about the billion plus people, then India can really leapfrog and become the world’s leader in artificial intelligence, in AI skills and AI talent,” Sikka told PTI. And Sikka’s Vianai Systems aims to do just that. The company’s LinkedIn page describes Vianai Systems as a company that can empower millions to build machine learning applications. The platform can be leveraged by C-suite leaders and developers to turn AI into real ROI. Discussing one of the major drawbacks of AI, Sikka emphasised the black-box nature of AI and the lack of ability to explain results in an interpretable manner.  This is one of the key areas addressed by his startup for which he raised seed funding. Sikka previewed his platform at the recently concluded Oracle Open World event.","excerpt":"Global business leader and former Infosys CEO Vishal Sikka believes AI will be a big disruptor for Indian society. The ex-Infy CEO was recently in the news for raising $50 million for his AI startup Vianai Systems which is currently operating in stealth mode. As reported, Sikka was recently approached by Prime Minister Narendra Modi […]","categories":["AI News"],"tags":[],"author_name":"Richa Bhatia","publish_date":"2019-09-27T12:28:44","publication_year":"2019","word_count":310,"keywords":["Go","funding","machine learning","artificial intelligence","programming_languages:R","AI","RAG","Aim","R","startup"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","Aim","RAG","R","Go","startup","funding","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/ai-will-be-a-very-big-disruptor-for-indian-society-says-ex-infy-ceo-vishal-sikka\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10072411,"title":"Google Duo to Merge into Google Meet","content":"Google recently announced that its Google Duo will now include features of the Google Meet (original). In its blog post, the company indicated that its Google Duo would be renamed to Google Meet. This new app would include video calling and meeting capabilities on mobile and web. Google Duo—which allows one-to-one direct call or group call—will enable users to create or join cloud-encrypted similar to Google Meet and send links for the meetings. In addition, all calls will have an in-call watermark confirming their end-to-end encryption and no time limit. Additionally, the merged Google Meet app will include features of Google Duo of storing call logs while also storing links of the previous meetings’ links.  Users would now be able to instantly call with features of the Google Duo app like effects and filters and send invitations to phone contacts to join Google Meet links. Recently, Google also embedded Meet in the Gmail app. Users signed in with their email on the app can access the feature of creating Google Meet meetings, while users signed in with their phone number can only access calls. After the complete rollout of updates, Google Duo users will be directed to the newer version of the Google Meet app while Google Meet (original) will remain on the Playstore and AppStore. Google is planning to eventually stop the support for Google Meet (original) app. Users can continue to access it but will have to shift to the newly merged Google Duo and Meet app. “We’ll inform you when you should migrate to the new app experience,” Google says.","excerpt":"Google is eventually planning to stop the support for the original version of the Google Meet app.","categories":["AI News"],"tags":["Google"],"author_name":"Mohit Pandey","publish_date":"2022-08-09T12:05:23","publication_year":"2022","word_count":263,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","ViT","Google","R"],"extracted_tech_keywords":["AI","R","Go","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/google-duo-to-merge-into-google-meet\/","complexity_score":2,"technical_depth":6,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":46042,"title":"How Fiction Games Are Paving The Way For Better NLU Understanding","content":"The use of machine learning in video games to generate images and situations is almost yesterday’s news. However, using video games as a platform to improve machine learning algorithms like those of reinforcement learning is still an ongoing research. Teaching an agent to complete a task or win a game to get rewards and use that in turn to incentivise more wins is the usual strategy. In a similar experiment, the researchers at Microsoft propose a novel approach to exploit interactive games to make machines smarter at understanding language. From a machine learning perspective, Interactive Fiction games exist at the intersection of natural language processing and sequential decision making.  Like many NLP tasks, they require natural language understanding, but unlike most NLP tasks, Interactive Fiction games are sequential decision making problems in which actions change the subsequent world states of the game and choices made early in a game may have long term effects on the eventual endings. Interactive Fiction(IF) games are fully text-based simulation environments where a player issues text commands to effect change in the environment and progress through the story. IF games combine challenges of combinatorial action spaces, language understanding, and commonsense reasoning. Interactive Fiction games are rich narrative adventures that challenge even skilled human players. In contrast to other video game environments, IF games stress natural language understanding and commonsense reasoning, and feature combinatorial action spaces.  To aid in the study of these environment, researchers have introduced Jericho, an experimental platform with the key feature of extracting game-specific action templates and vocabulary. Overview Of Jericho Jericho is an open-source Python-based IF environment, which provides an OpenAI-Gym-like interface for learning agents to connect with IF games. Jericho supports a set of human-made IF games that cover a variety of genres: dungeon crawl, Sci-Fi, mystery, comedy, and horror.  Games were selected from classic Infocom titles such as Zork and Hitchhiker’s Guide to the Galaxy, as well as newer, community-created titles like Anchorhead And Afflicted. Supported games use a point-based scoring system, which serves as the agent’s reward. Using these features, a novel template-based action space is proposed, which serves to reduce the complexity of full scale language generation. Using this space, they then introduced the Template-DQN agent TDQN, which generates actions first by selecting a template then filling in the blanks with words from the vocabulary. How NLU Systems Can Benefit Combinatorial spaces pose extremely difficult exploration problems for existing agents.  For example, an agent generating a four-word sentence from a modest vocabulary of size 700, is effectively exploring a space of 7004= 240 billion possible actions. Whereas, knowledge representation poses a mapping problem. Due to the large number of locations in many games, humans often create maps to navigate efficiently and avoid getting lost. In particular, because connectivity between locations is not necessarily Euclidean, agents need to detect when a navigational action has succeeded or failed and whether the location reached was previously seen or new. Reinforcement learning has studied agents that operate in discrete or continuous action space environments.  However, IF games require the agent to operate in the combinatorial action space of natural language. Interactive fiction (IF) games are software environments in which players observe textual descriptions of the simulated world, issue text actions, and receive score as they progress through the story. Here are few key takeaways according to the authors of the original paper: Researchers introduce Jericho, a learning environment for human-made IF games. And, introduce a template-based action space that is appropriate for language generation. Also, conduct an empirical evaluation of learning agents across a large set of human-made games. Beyond games, real-world applications such as voice-activated personal assistants can also benefit from advances in these capabilities at the intersection of natural language understanding, natural language generation, and sequential decision making. These real world applications require the ability to reason with ungrounded natural language (unlike multimodal environments that provide visual grounding for language) and IF games provide an excellent suite of environments to tackle these challenges. Read the original paper here.","excerpt":"The use of machine learning in video games to generate images and situations is almost yesterday’s news. However, using video games as a platform to improve machine learning algorithms like those of reinforcement learning is still an ongoing research. Teaching an agent to complete a task or win a game to get rewards and use […]","categories":["Deep Tech"],"tags":["games","Microsoft","NLP","NLU","Reinforcement Learning"],"author_name":"Ram Sagar","publish_date":"2019-09-17T12:00:15","publication_year":"2019","word_count":669,"keywords":["Go","machine learning","Reinforcement Learning","OpenAI","AI","Modal","games","NLP","NLU","Python","ViT","programming_languages:Python","R","Microsoft"],"extracted_tech_keywords":["AI","machine learning","NLP","OpenAI","Python","R","Go","ViT","Modal","programming_languages:Python"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/how-fiction-games-are-paving-the-way-for-better-nlu-understanding\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":24026,"title":"MIT One Ups The Game For #Foodgrammers, How Neural Network Suggests Recipes Based On Photos","content":"Let’s admit it — taking countless photos of well-plated dishes is fun. We millennials enjoy documenting great food before we dive into the calories. Choosing the right filter, using the appropriate hashtag and posting it on social media, is almost a ritual now. Now an MIT researcher has taken this #foodgram addiction to a whole new level by training a machine learning system to look at a photograph of food, dissect the ingredients and suggest similar recipes. Researchers from Massachusetts Institute of Technology’s Computer Science and Artificial Intelligence Laboratory have developed a neural network called Receipe1M (its end-product is called Pic2Recipe) will not only help us learn the recipes but also better understand people’s eating habits and dietary preferences. What Is Receipe1M Receipe1M  is a neural network which is trained on the largest structured dataset to find patterns and make connections between the food images and the corresponding ingredients and recipes. The Recipe1M contains one million structures cooking recipes and 800,000 food images. By looking the photo of a food item, Receipe1M  can identify ingredients like flour, eggs, and butter and then suggest recipes that it determined to be similar to images from the database. Receipe1M jointly learns to embed images and recipes in a common space which is semantically regularised by the addition of a high-level classification task. To create a larger database, the MIT researchers have also collected images and recipes which already exist on cooking websites. Compiling data from places like Food.com and All Recipes, the CSAIL researchers created Recipe1M. According to a release on the MIT website, the neural system did particularly well with desserts like cookies or muffins, since that was a main theme in the database. However, it had difficulty determining ingredients for more ambiguous foods, like sushi rolls and smoothies. It was also often stumped when there were similar recipes for the same dishes. For example, there are dozens of ways to make lasagna, so the team needed to make sure that system wouldn’t “penalize” recipes that are similar when trying to separate those that are different. Dissecting Recipe1M The contents of Recipe1M dataset is grouped into two layers. The first layer contains basic information including the title of the recipe, a list of ingredients and a sequence of instructions for preparing the dish in text format. The second layer is built upon the first and includes associated images in RGB and JPEG format. Additionally, a subset of recipes are annotated with course labels like appetizer, side dish, dessert, etc. The research paper also states that the average recipe in the dataset consists of nine ingredients that are transformed over the course of 10 instructions. Half of the recipes have images which, due to the nature of the data sources, depict the fully prepared dish. But Recipe1M includes 0.4 percent duplicate recipes and two percent duplicate images. However, the researchers explain that excluding those 0.4 percent recipes, 20 percent of recipes have non-unique titles but symmetrically differ by a median of 16 ingredients. 0.2 percent of recipes share the same ingredients but are relatively simple, having a median of six ingredients. How Recipe1M Works The joint embedding model of Recipe1M is built upon two representations — recipe and image. Recipe Representation Ingredients and cooking instructions are the two major components of a recipe. Each recipe contains a set of ingredient text. For each ingredients, the researchers used word2vec representation. The actual ingredient names are extracted from each ingredient text. For example, in “two tablespoons of olive oil”, the olive oil is extracted as the ingredient name and treated as a single word for word2vec computation. The initial ingredient name extraction task is solved by a bi-directional LSTM that performs logistic regression on each word in the ingredient text. Each recipe also has a list of cooking instructions for which the researchers used a two-stage LSTM model. First, each instruction or sentence is represented as a skip-instructions vector and then an LSTM is trained over the sequence of these vectors to obtain the representation of all instructions. The resulting fixed-length representation is then fed into to their joint embedding model. Food Image Representation For the image representation, the researchers adopted state-of-the-art deep convolutional networks called VGG-16 and Resnet-50 models and connected them to the joint embedding model. The recipe model also includes two encoders: One for ingredients and the other for instructions, the combination of which is designed to learn a recipe level representation. The ingredients encoder combines the sequence of ingredient word vectors. The instructions encoder is implemented as a forward LSTM model over skip-instructions vectors. The outputs of both encoders are concatenated and embedded into a recipe-image joint space. The team has also incorporated semantic regularis ation on their recipe and image embedding through solving the same high-level classification problem in multiple modalities with shared high-level weights. The key idea is that if high-level discriminative weights are shared, then both of the modalities (recipe and image embeddings) should utilise these weights in a similar way which brings another level of alignment based on discrimination. Test Results Of Recipe1M The MIT researchers evaluated the performance of the neural network against Canonical Correlation Analysis baselines and humans, which showed a remarkable improvement over the former while faring comparably to the latter. Arithmetics The researcher has also demonstrated the capabilities of their learned embeddings with simple arithmetic operations. In the context of food recipes, this what researchers found: ‘v’ represents the map into the embedding space. The figures below show some results with same and cross-modality embedding arithmetics. Conclusion To sum up, the neural network is still in an early stage of development, but going forward the researchers plan to use food data more extensively. According to MIT news, the team hopes to be able to improve the neural system so that it can understand food in even more detail. This could mean being able to infer how a food is prepared (for example, stewed vs diced) or distinguish different variations of foods, like mushrooms or onions. The researchers are also interested in potentially developing the system into a dinner aide that could figure out what to cook given a dietary preference and a list of items in the fridge.","excerpt":"Let’s admit it — taking countless photos of well-plated dishes is fun. We millennials enjoy documenting great food before we dive into the calories. Choosing the right filter, using the appropriate hashtag and posting it on social media, is almost a ritual now. Now an MIT researcher has taken this #foodgram addiction to a whole […]","categories":["IT Services"],"tags":["MIT","MIT researchers","Neural Network"],"author_name":"Smita Sinha","publish_date":"2018-04-26T07:32:04","publication_year":"2018","word_count":1035,"keywords":["Neural Network","Go","artificial intelligence","MIT researchers","machine learning","AI","neural network","MIT","TPU","ResNet","RAG","LSTM","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","neural network","RAG","TPU","R","Go","LSTM","ResNet"],"url":"https:\/\/analyticsindiamag.com\/it-services\/mit-one-ups-the-game-for-foodgrammers-how-neural-network-suggests-recipes-based-on-photos\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":59992,"title":"A Letter By The CEO To The Informatica Community","content":"As the globe is shaken by the pandemic, CEOs of the leading tech organisations have been supporting their employees and community with every way they can. In a letter, the CEO of Informatica, Amit Walia shared a powerful message to the Informatica community about how they manage the business to serve and support the Informatica employees and most importantly how Informatica is responding to the COVID-19 pandemic. Amit Walia said, “Our customers, partners, and employees are our top priority and everyone’s health and safety is paramount. At Informatica, all of our employees around the world have been strongly encouraged to work from home, and we have shifted all of our in-person events to virtual events.” During the unprecedented challenges of COVID-19, the company is committed to provide the continuity of support and service to the customers and partners around the globe. Walia discussed their Business Continuity Program which includes plans, procedures, roles, responsibilities, and activities that ensure a timely, effective, and orderly response to any disruption of critical business services. With the help of the collaboration tools, the company provides full-service support to the teams who are working in an office or remotely. For the customers, the company provides cloud services, Global Customer Support and Professional Service to deliver uninterrupted services. Walia added,”In times like these, we are grounded by our values. At Informatica, our values are DATA – Do Good, Act as One Team, Think Customer First, and Aspire for the Future. We are grounding ourselves in support of our priorities. While this is an unprecedented and challenging situation for all of us, the welfare of our employees, customers, and communities are foremost in our minds.”","excerpt":"As the globe is shaken by the pandemic, CEOs of the leading tech organisations have been supporting their employees and community with every way they can.  In a letter, the CEO of Informatica, Amit Walia shared a powerful message to the Informatica community about how they manage the business to serve and support the Informatica […]","categories":["AI Features"],"tags":["informatica"],"author_name":"Ambika Choudhury","publish_date":"2020-03-26T12:09:22","publication_year":"2020","word_count":277,"keywords":["Go","informatica","programming_languages:R","AI","programming_languages:Go","RAG","ViT","disruption","GAN","R"],"extracted_tech_keywords":["AI","RAG","R","Go","GAN","ViT","disruption","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/a-letter-by-the-ceo-to-the-informatica-community\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10074762,"title":"Data Analytics According To A Political Campaigner","content":"Political campaigns in India differ from those in western nations; the challenges faced by the Indian populace aren’t highly likely to resemble those faced by residents in the US, for instance. In an effort to better understand these challenges in the context of Indian politics, Analytics India Magazine got in touch with Namita Tiwari, who oversaw the political campaigns of parties like the Shiv Sena and Trinamool Congress. She provides valuable insight into numerous challenges faced by political data analysts and campaigners and how they tackle them. Campaigns, polls and everything in between “When we went to West Bengal, the first thing that we wanted was for Didi (Mamta Banerjee) to connect with people—which is how the ‘Didi Ke Bolo’ initiative got its start,” Namita Tiwari explained while discussing how they were able to gather the data of citizens who were experiencing some sort of hardship. “People came and expressed their complaints, which we mediated through CMO and successfully resolved. Using qualitative and quantitative methodologies, we were able to understand the challenges and objectives that exist across the state over the course of a period. This served as the basic framework for two initiatives: ‘Duare Sarkar’, which sought to bring specific state government programmes to the doorsteps of citizens through outreach activities organised at the gram panchayat and municipal ward levels; and ‘Paray Samadhan’, which basically focused on issues related to infrastructure.” Further discussing political polls and their significance, Namita Tiwari said, “Political surveys are conducted in the background, as you are aware, and our team also goes out and makes an effort to comprehend and evaluate issues based on the total understanding we received from the party, primary data, and secondary research that our data deck conducts.” Namita Tiwari believes that this data plays a crucial role in providing an accurate analysis of a vast spectrum of perspectives. She says, “The campaign team relies on it since we can likely assess the point of view from a variety of perspectives: from a political standpoint, a media perspective, and the perspective of the general public.” She further claims that a significant amount of data and AI are utilised in the initial stage of quantitative work. To create the final result from all data and analysis, they do draw on human knowledge and experience, but AI helps them comprehend everything. “For example, if we want to know how the media is responding to PM Modi’s visit to our state, we have our word cloud just to understand which of the things the media has followed, which part of his speech media picked up and how it’s taken up on social media—these all are very important for us to formulate the plan,” Namita Tiwari added. Addressing the varied demographic issues in political campaigns and how they are managed, she remarked, “During the Shivsena campaign in the GE 2019 urban part of Maharashtra, we had multiple challenges. The Northwest region of MH had equal votes from Marathis and East Indians (UP and Bihar), followed by the Gujrati band and then the South Indians. Based on caste analysis of candidates and historical voter trends, we could deduce that Marathi voters are inclined towards Shiv Sena, but there is a split in East Indian votes, which was the decisive factor. This helped us to plan community-based events to bring in the swing voters in our favour.” Could data analytics backed campaigns shift the voter base? Namita Tiwari believes that a proper data analytics backed campaign with enough time to run could eventually shift the voter base. She shares her experience and proposes that a change of preference in voters is, in fact, possible—“So in assembly elections, I was working on the Worli seat where an ex-MLA of NCP joined us. The party was confident, based on the analysis, that we’ll be able to tap the NCP Voters, i.e. SC community, which accounted for around 20% in the area. Our analysis showed that Marathi will completely align with us. Muslims and people from high-rises need to be convinced with targeted outreach, which we were able to perform successfully to an extent. However, since we were short on time (six months), we couldn’t shift the core vote bank of the NCP, who still voted for a very new face (from NCP). The candidate got 20–23% of the vote share and most of them came from that area only.” She further added, “In India, nowadays it’s very hard to change the voter base over a short period of time, or in the six months that we had; for a proper data analytics backed campaign to run, we need a minimum of two years. It is a good enough time to win over people through our impact generated on the ground” Discussing the significance of the Jio revolution for data collecting, Namita Tiwari remarked, “Before the launch of Jio in 2015, it wasn’t as easy to collect data. Majority of the efforts included handing out flyers and going door to door. However, it got simpler to collect data as more people went online.” She believes that the introduction of Jio in India and the use of data analytics in politics are plausibly related. But is data all we need? Namita Tiwari believes that, although data is prominent in current analysis in politics, it is not yet at a stage where it can replace human involvement completely. She says, for instance, “We cannot unreservedly follow what the data indicates. For example, solely because data indicate that few parties in the world have historically benefited from conflict doesn’t imply that it’s morally correct to re-generate such scenarios for political gain. When using these tactics in the field, we need to take the EQ component into consideration, to ensure politics priortise social welfare over power aspirations”, concluded Namita.","excerpt":"As the usage of AI and data analytics in Indian politics grows, we made the decision to learn more about how campaign managers operate behind the scenes.","categories":["AI Features"],"tags":[],"author_name":"Lokesh Choudhary","publish_date":"2022-09-09T17:00:00","publication_year":"2022","word_count":966,"keywords":["Go","programming_languages:R","AI","ML","Ray","Aim","ViT","analytics","GAN","R"],"extracted_tech_keywords":["AI","ML","analytics","Aim","Ray","R","Go","GAN","ViT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/data-analytics-according-to-a-political-campaigner\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10130380,"title":"Now AI Can Help You Write Codes for Blockchain","content":"Some might argue that generative AI has overshadowed blockchain, while others might maintain the opposite. But the truth is, both these technologies boast intriguing use cases, and what we are witnessing now is a convergence of the two. XinFin, the creator of XDC Network, an enterprise-grade, open-source blockchain protocol aimed at transforming global trade and finance, has introduced CodeRun.ai, a coding tool tailored for use with the XDC Network. “CodeRun.ai is optimised for accuracy in blockchain-related applications. Therefore, anyone looking to develop on the blockchain can simply issue a command and receive a complete code base instantly,” Ritesh Kakkad, a co-founder of XDC Network, told AIM. It is specifically optimised for the XinFin XDC blockchain network and seamlessly integrated into the broader XDC ecosystem, aiming to foster developer adoption and broaden its range of applications. The network is tailored for enterprise applications, including supply chain management, trade finance, and other business-critical processes. It aims to enhance operational efficiency and reduce costs for enterprises. CodeRun is Powered by OpenAI Atul Khekade, another co-founder of XDC Network, explained to AIM that CodeRun.ai draws inspiration from GitHub Copilot, the world’s most widely used AI tool for developers. However, fundamentally, both serve distinct purposes. “I was, in fact, among the first 20 users of GitHub Copilot. It is more of a code-generation tool, and fundamentally, we have architected CodeRun.ai differently,” he said. While GitHub Copilot helps in writing codes, CodeRun.ai is designed to help build applications on blockchain. It utilises OpenAI‘s APIs within what the startup calls an ‘AI aggregator’, which is also powered by a proprietary XDC Network model. “This setup allows us to leverage databases such as ChromaDB and Pinecone in a local environment tailored to our specific needs and use cases. The aggregator then integrates broader intelligence from OpenAI, refining XDC Network’s enterprise data to meet precise requirements,” Khekade said. Explaining further, he said when an enterprise runs CodeRun.ai in their local environments, be it a bank or a large aviation player, they benefit from enhanced privacy protection. “Your specific data and configurations are securely handled, with generative data staying on your own servers, never exposed externally. For example, in the airline industry, CodeRun.ai can optimise flight scheduling or pricing algorithms. “It might suggest adjustments that allow you to charge customers 10% less while still increasing profitability. This level of precision and customisation is unmatched by other tools,” Khekade added. The model has been fine-tuned with XDC Network’s 103-gigabyte proprietary data. Moreover, the startup also plans to release its proprietary model soon. Enabling Applications Building on Blockchain XDC Network, which was founded in 2017, supports the creation and execution of smart contracts and decentralised applications (dApps), enabling automated and transparent business processes. The network is tailored for enterprise applications, including supply chain management, trade finance, and other business-critical processes. Kakkad said XDC Network foresees a trend where hundreds and thousands of applications are going to be built on a blockchain. CodeRun.ai enables not just blockchain developers or enterprises but someone with very limited knowledge about these technologies to build applications. The startup also intends to onboard over 10,000 active developers within the first year. For this, it’s already in talks with incubators, most notably T-Hub, which is based in Hyderabad and has the largest incubator facility in the world. “We started with almost a hundred users and these are all developers who are building on XDC Network. So, this was our testing period and since the launch, the demand has skyrocketed. Now we are going for an incubator strategy,” Khekade said. XDC Network recently signed a deal with Plug & Play, which is a global innovation platform and one of the world’s largest startup accelerators. They operate in over 50 locations worldwide, running more than 60 industry-focused accelerator programs annually. “The target is to implement at least 1,000 projects or applications utilising the backbone of the XDC Network, addressing use cases or issues identified by enterprises or startups. “This initiative is expected to significantly increase the usage of Coderun.ai, as many developers and builders will be leveraging it for their projects,” Khekade pointed out. Besides Plug & Play, the startup is also partnering with at least five more incubators, some of which are from Hong Kong and Singapore. “They specialise in building large-scale market applications and are very successful in their portfolio,” Khekade concluded.","excerpt":"While GitHub Copilot helps in writing codes, CodeRun.ai is designed to help build applications on blockchain.","categories":["AI Features"],"tags":["AI coding","Github Copilot"],"author_name":"Pritam Bordoloi","publish_date":"2024-07-26T17:30:41","publication_year":"2024","word_count":723,"keywords":["AI coding","Run.ai","OpenAI","AI","AWS","ML","Github Copilot","RAG","Aim","generative AI","Chroma","Pinecone"],"extracted_tech_keywords":["AI","ML","generative AI","OpenAI","Aim","Run.ai","RAG","Pinecone","Chroma","AWS"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/now-ai-can-help-you-write-codes-for-blockchain\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10089802,"title":"NVIDIA GTC 2023 Key Highlights","content":"Last week, we here at AIM laid down our expectations for the highly anticipated NVIDIA GTC event. And it seems like we hit the jackpot! All the announcements made by NVIDIA chief Jensen Huang at the event today suggest that in the game of hits and misses, most of them were hits. Here, are some key takeaways from the event: NVIDIA DGX Cloud NVIDIA unveiled its new DGX Cloud – an AI supercomputing service that pairs NVIDIA DGX™ AI supercomputing with NVIDIA AI software, offering enterprises dedicated clusters for training advanced models in generative AI and other innovative applications. This provides businesses with immediate access to the infrastructure and software needed for training cutting-edge models. The DGX Cloud will be hosted on Microsoft Azure, with Google Cloud and other platforms following shortly after. The service will be powered by NVIDIA AI Enterprise 3.1, featuring access to new pretrained models, optimised frameworks, and accelerated data science software libraries. NVIDIA Omniverse NVIDIA announced the release of three new systems specifically designed to support the Omniverse platform. The first of these systems is a new generation of workstations that will be powered by NVIDIA Ada RTX GPUs and Intel’s latest CPUs. The desktop and laptop GPUs of the RTX series will cater to the demands of the modern era of AI, design, and metaverse. In addition, the company revealed the next generation of its OVX computing system, which is optimised to run the Omniverse platform. NVIDIA emphasised that every layer of the Omniverse stack, from chips to systems to networking and software, is a new innovation. Finally, NVIDIA also announced its plans to provide the NVIDIA Omniverse Cloud, a fully managed cloud service, to select enterprises through a partnership with Microsoft, which will host it on Azure. The Omniverse Cloud will be integrated with Microsoft 365 productivity suite, which includes Teams, OneDrive, SharePoint, and Azure IoT Digital Twins solutions. Inference Platforms During the GTC 2023 event, NVIDIA unveiled a new inference platform that boasts four different configurations under a single architecture, designed to optimise different AI workloads. The L4 configuration, for instance, has been optimised for video decoding and transcoding, video call features, transcription, and real-time language translation. NVIDIA L4 will be hosted on Google Cloud. For graphics rendering and generative AI, such as text-to-image and text-to-video, NVIDIA introduced the L40 configuration. In addition, for large language model inference, the company announced the new Hopper GPU, known as the PCIE H100, which features a dual-GPU NVL. Moreover, with supporting PCIE servers to the H100 processors, it is easy to scale out the performance. Lastly, NVIDIA introduced the Grace Hopper, a new superchip that connects the Grace CPU and Hopper GPU, offering a high-speed processing rate of 900 GB\/sec. The Grace Hopper chip is said to be ideal for processing giant datasets and large language models, further completing the stack of offerings from NVIDIA’s new inference platform. NVIDIA’s inference platform Generative AI NVIDIA AI Foundations, a cloud-based service aimed at facilitating the building of custom large language models (LLMs) and generative AI with proprietary data, was announced during the GTC 2023 event. This new service will include three models: NeMo, a language text-to-text generative model; Picasso, a visual language model-making service; and BioNeMo, which will enable medical institutes to leverage generative AI for drug discovery. To further expand its offerings, NVIDIA has partnered with Getty Images, who will use the Picasso service to build Edify-image and Edify-video generative models. In addition, Shutterstock will be working on the development of the Edify-3D generative model, which can be used for the creation of 3D assets. Furthermore, NVIDIA announced a long-term partnership with Adobe, which will integrate generative AI for image creation, video, 3D, and animation. Read: NVIDIA Rides The Generative AI Wave At GTC 2023 Acceleration Libraries One of the major announcements made by NVIDIA is the development of a quantum control link, in collaboration with Quantum machines, which connects NVIDIA GPUs to quantum computers to perform error correction at extremely high speeds. This new platform, which leverages the newly open-sourced CUDA Quantum, provides a revolutionary new high-performance and low-latency architecture for researchers working in quantum computing. In addition to this, the company also announced the release of RAFT, a new library that accelerates indexing, data loading, and batch retrieval of neighbours for a single query. NVIDIA is bringing RAFT to Meta’s FAISS, Milvus, and Redis, which are vector databases that will be important for organisations building proprietary large language models. NVIDIA Triton Management Service is another new software that was announced, which automates the scaling of Triton inference instances across the data centre. The company also unveiled two new cloud-scale acceleration libraries: CV-CUDA for computer vision and VPF for video processing. CV-CUDA includes 30 computer vision operators for detection, segmentation, and classification, while VPF is a python video encode and decode acceleration library. NVIDIA also announced the release of a new version of Parabricks (4.1), which is a suite of AI-accelerated libraries for end-to-end genomic analysis in the cloud or in-instrument. This update brings new features and improved performance to Parabricks, making it easier for researchers and scientists to analyse genomic data. Furthermore, NVIDIA announced a partnership with the medical technology company Medtronic to develop software-defined medical devices. The platform will be used for Medtronic systems, ranging from surgical navigation to robotic-assisted surgery. Lastly, NVIDIA also released cuLitho, a library for computational lithography. The library accelerates computational lithography by over 40 times and was built in collaboration with industry leaders such as ASML, TSMC, and Synopsys.","excerpt":"Here are some few key takeaways from the GTC 2023 event","categories":["Global Tech"],"tags":[],"author_name":"Ayush Jain","publish_date":"2023-03-22T01:13:21","publication_year":"2023","word_count":921,"keywords":["data science","AI","ML","computer vision","RAG","vector databases","Aim","generative AI","Milvus","Azure"],"extracted_tech_keywords":["AI","ML","computer vision","data science","generative AI","Aim","RAG","vector databases","Milvus","Azure"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/nvidia-gtc-2023-key-highlights\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10169947,"title":"Reinforcement Learning Won Again, This Time With Microsoft","content":"While Microsoft is widely known for backing OpenAI with both infrastructure and capital, the company’s own open-source Phi family of models, part of their own AI research and development, isn’t as recognised the same way. The Phi series of lightweight models is designed to consume less compute and storage. Thanks to various techniques and optimisation processes involved in the research process, these models have historically outperformed the competition, both in their lightweight segment and even some of the larger ones. The latest addition is the Phi-4 Reasoning — a 14 billion-parameter model built by applying a supervised fine-tuning (SFT) algorithm to the Phi-4 base model. The researchers also derived the Phi-4 Reasoning Plus model by using reinforcement learning (RL) on Phi-4 Reasoning. Both these models outperform much larger models like DeepSeek R1, the 70B parameter model, on benchmarks that involve coding, math, and graduate-level scientific tasks. The models also perform close to the full-scale 671B parameter DeepSeek R1 model. Source: Microsoft The researchers primarily attribute the model’s success to ‘high-quality’ training datasets, which, Microsoft has staked its bets on with its previous models. These datasets contain over 1.4 million prompts (across various coding and STEM disciplines) with high-quality answers containing long reasoning traces generated by OpenAI’s o3-mini model. To train the model effectively, the researchers targeted prompts at the edge of the base Phi-4 model’s abilities, meaning that the training datasets were filtered only to retain those that provide meaningful room for improvement. Why RL Works for Reasoning After Phi-4 was derived from the SFT of the Phi-4 model, a subsequent RL process resulted in the development of Phi-4 Reasoning Plus. AIM reached out to Harkirat Behl, a Microsoft researcher, who played a key role in the RL components of Phi-4 Reasoning Plus. RL is a training approach in which an AI learns through trial and error by taking actions, receiving rewards or penalties, and progressively refining its decisions to enhance long-term outcomes. It is compelling for tasks that demand the AI model to ‘reason’ because it prioritises outcomes over processes. In contrast to traditional models, which solely predict the next word and penalise the model for each inaccurate word, RL allows flexibility in how an answer is reached. It will enable the model to navigate complex problems with multiple paths to arrive at a correct conclusion. As Behl explains, RL lets the model “generate very long answers, and many different answers,” focusing only on whether the outcome is correct. By evaluating only the final result, reinforcement learning better reflects how humans solve problems: different thought processes are allowed, as long as they lead to the correct conclusion, he further indicated. In Microsoft’s models, this RL stage was focused exclusively on mathematical reasoning. The rewards incentivised the model’s correctness, but penalised repetition and excessive length, while encouraging proper response formatting. Behl explained that the researchers allowed the model to generate multiple answers for a question, and each answer was scored based on its comparison to the average score within the group. These relative scores are then used to adjust the model, encouraging it to favour answers that consistently score higher. Over time, this trains the model to align its responses more closely with the reward signal. The researchers noted in the paper that performing RL on a small set of 6,400 problems significantly improved accuracy across math and reasoning evaluations. “Having built Phi-1, Phi-2, Phi-3, and Phi-4, one takeaway from me in research is that RL requires much less data than the SFT training,” said Behl. Behl indicated that this is because RL is less about teaching the model any brand-new skills from scratch and more about showing the model how to combine or compose the skills it already has to get better results. Microsoft joins many other AI companies that have reported success with reinforcement learning. OpenAI, the company that started the trend of reasoning models, has repeatedly spoken about how RL worked favourably. Interestingly, even China’s DeepSeek R1 model, which disrupted the ecosystem last year, attributed its success to RL. Moreover, several researchers and engineers from OpenAI have publicly credited RL for the success of their deep research feature. And more recently, Alibaba’s Qwen also endorsed reinforcement learning, given its impact on their reasoning models. “We are confident that combining stronger foundation models with RL powered by scaled computational resources will propel us closer to achieving Artificial General Intelligence (AGI),” said the company in a blog post. However, while the Phi-4 Reasoning, Phi-4 Reasoning Plus, and many other reasoning models have been successful, several challenges remain in this space. Plenty of Room for Improvement In recent months, numerous research studies have highlighted issues with reasoning models. For instance, researchers from Microsoft, in their Phi-4 Reasoning paper, stated that they still face challenges, including the consumption of excessive time and resources, slower response times, and most notably, the issue of their responses contradicting their own reasoning steps. Recently, Anthropic released a study revealing that reasoning chains (chain-of-thoughts or CoTs) may not always reflect a model’s actual reasoning process. The researchers found that models often exploit external hints—explicit cues inserted into prompts to guide them toward correct answers—but rarely acknowledge or verbalise these hints in their reasoning steps. This gap between internal behaviour and external explanation raises concerns about the reliability of using CoTs for model interpretability and safety. Even OpenAI released a research report that indicates that frontier reasoning models frequently engage in reward hacking, where AI agents exploit loopholes in their objectives to gain rewards in unintended ways. OpenAI said that using a less powerful model (GPT-4o) to monitor a stronger model like the o3-Mini. Nat McAleese, member of the technical staff at OpenAI, said in a post on X that “large reasoning models are extremely good at reward hacking”, and handpicked examples from the report to illustrate his statement. “There’s a lot of redundancy in the chain of reasonings; they contradict themselves, and there are a lot of unanswered questions,” said Behl. “But, it is an evolving space. If we can nail this as a community and understand how the models think, there will be a lot of gain.”","excerpt":"Phi-4 Reasoning Plus is the latest model that uses RL to achieve impressive scores on benchmarks.","categories":["Global Tech"],"tags":["AI (Artificial Intelligence)","Microsoft"],"author_name":"Supreeth Koundinya","publish_date":"2025-05-14T14:41:32","publication_year":"2025","word_count":1022,"keywords":["Anthropic","Go","OpenAI","AI","GPT-4o","DeepSeek R1","RAG","Aim","foundation models","AI (Artificial Intelligence)","R","Microsoft"],"extracted_tech_keywords":["AI","foundation models","GPT-4o","OpenAI","Anthropic","DeepSeek R1","Aim","RAG","R","Go"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/reinforcement-learning-won-again-this-time-with-microsoft\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":67236,"title":"How This Gurgaon-Based Fashion-Tech Startup Is Using AR To Create New-Age Jewellery Shopping Experience","content":"“Our vision has always been to revolutionise the shopping experience through technology.”- Meghna Saraogi and Akhil Tolani, Founders at StyleDotMe. Shopping and retail is one segment where the impact of Augmented Reality (AR) has been monumental, adding a lot of benefits to the industry. From streamlining the process of making both online and offline shopping easier for people, AR is a different ball game altogether. AR and VR are radically changing the way people consume experiences. Integration of AR in the retail sector has become a mandatory trend as it enhances customer engagement. The immersive media market in India is expected to grow to USD 6.5 Billion by 2022 while AR is estimated to grow to USD 5.9 Billion and VR to USD 0.5 Billion by 2022. Gurgaon-based StyleDotMe not only provides customers with a great experience, but it also delivers jewellers with actionable insights to optimise their entire value chain across design, manufacturing, operations, marketing and sales. Founded by Meghna Saraogi and Akhil Tolani in 2015, StyleDotMe is a fashion tech startup that is revolutionising jewellery shopping experience. The company has raised close to $1 Million from some of the top angel investors in the country. Meghna stated, “StyleDotMe’s aim has been to disrupt the retail model with our immersive AR technology. We have built a breakthrough technology for an industry that is traditional and insulated against disruptors such as ours.” Flagship Product The company’s flagship product is mirrAR, which is an AR tech platform for jewellers. It allows shoppers to try jewellery on digital platforms of client brands without actually having to wear the pieces and get instant fashion advice from experts, bloggers and influencers around the world. According to the founders, this technology is helping jewellery brands increase their reach and enable easy checkouts by providing shoppers with more product options and the ability to see multiple jewellery pieces on themselves in quick succession. How Is It Differentiated From Other Products In The Market? To this question, the founders replied that the company’s innovative AR technology works automatically on digitised versions of jewellery in real-time while other products in the market require users to manually position the jewellery on themselves. Also, mirrAR can be plugged into any platform, whether it’s on a device like an iPad or on the web that allows for both offline and online integration. Use Of AI And ML At StyleDotMe According to the founders, using the patent-pending breakthrough in computer vision, they are providing an unprecedented experience in real-time virtual jewellery try-on. mirrAR conducts face-mapping in real-time, recording the time-spent, identifies skin tone, approximate age, mood, location of the customer to capture the reaction to the displayed design and then recommends various jewellery designs that may appeal to users. This is supplemented by privacy features that ensure that this data is clustered and not tracked on an individual level. Core Technology Stack The tech stack of StyleDotMe consists of multiple Convolution Neural Networks (CNNs). This was developed and trained in-house by the firm’s tech team for detecting various parts of a person’s body. mirrAR runs entirely on the client-side in order to be completely real-time. The flagship product performs face tracking in less than 2 milliseconds to overlay jewellery on a 60 frames\/second phone camera. The entire platform runs across all devices including iOS, Android and Web. The machine learning models in this product are powered by TensorFlow, and the cloud platform provider is Amazon Web Services (AWS). Tackling Hiring Phase To this question, the Akhil stated, “We continually work towards developing realistic short term and long term hiring forecasts based on the projects we plan to launch in future.” Meghna added, “The qualities that we look for in employees besides the required skill sets are experience, attitude, cultural fit, high ethics and values. These qualities are extremely important to build the right culture for our work environment.” Future Roadmap In the next 5 years, the company envisions itself to become the industry leader globally catering to the entire fashion retail industry. Meghna stated, “Through our platform, people across the globe will be able to try-on things virtually and make a purchase. Some of the prominent categories that we will have our dominance in art jewellery, makeup, eyewear, watches and apparel.”","excerpt":"“Our vision has always been to revolutionise the shopping experience through technology.”- Meghna Saraogi and Akhil Tolani, Founders at StyleDotMe. Shopping and retail is one segment where the impact of Augmented Reality (AR) has been monumental, adding a lot of benefits to the industry. From streamlining the process of making both online and offline shopping […]","categories":["AI Startups"],"tags":["Startups"],"author_name":"Ambika Choudhury","publish_date":"2020-06-12T15:00:00","publication_year":"2020","word_count":709,"keywords":["Go","machine learning","AWS","AI","neural network","ML","computer vision","Aim","Startups","TensorFlow","R"],"extracted_tech_keywords":["AI","machine learning","ML","neural network","computer vision","Aim","TensorFlow","AWS","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-startups\/how-this-gurgaon-based-fashion-tech-startup-is-using-ar-to-create-new-age-jewellery-shopping-experience\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":53781,"title":"Why Data Science Recruiters Should Stop Looking For PhD Candidates","content":"Data scientists are now the new assets for a company, and the candidates are required to have the right ability and skills to be rewarded with this lucrative career. But now, many data science job descriptions involve the requirement of a PhD degree to apply. To ensure an effective workflow in data science initiatives, organisations are seeking PhD aspirants, ignoring the challenge of the skill gap in the country. PhD candidates will have numerous advantages over other non-doctorate aspirants, in terms of knowledge and exposure; however, a PhD degree in the data science field will not guarantee results due to the ever-changing technology landscape. Since a decade ago, data science jobs didn’t exist, and barely anyone pursued a PhD in data science. Therefore there were only a handful of data scientists who used to have a doctorate in data science. Besides, until recently, most of the universities didn’t even offer data science courses, which, in turn, resulted in a shortage of skilled applicants for the jobs. Further, as per research, there are over 4000 jobs for data science in the US (a rise of 56% from the previous data), which provide ample opportunity for PhD candidates to choose from. Therefore, looking for a PhD applicant is a no brainer. According to the head data scientist and AI architect at ProVise Consulting, “Most of the problems that businesses deal with, do not require PhD candidates.” The Landscape Considering the rapid changes in the field of AI and data science landscapes, new technologies emerge, and many techniques get abandoned, a PhD in data science takes somewhere between five to seven years. For one, Jupyter Notebook, TensorFlow, Keras, PyTorch, among others have only become mainstream in the last three to four years. So, these tools wouldn’t be included in the data science courses in universities, and thereby, several aspirants enrolled for PhD in data science in 2014 or 2015, wouldn’t be proficient in new technologies that are relatively common today. Consequently, a PhD in data science doesn’t ensure a skilled candidate. However, in their defence, a doctorate scholar, over the years, acquires strong basics which empowers them to familiarise with new technologies quickly. Skills vs PhD In Data Science In the highly competitive data science landscape, skill is of paramount importance than certificates. For instance, on Kaggle, where developers from around the world compete to win the challenges, countless developers have become an expert without having a PhD in data science. Today, there are numerous platforms that various organisations can use to determine the proficiency of a data scientist instead of seeking for a PhD candidate. Unlike other development roles, data science developers are trained to solve real-world challenges on Kaggle, StackOverflow, GitHub, and other hackathons that are hosted across the world. Thus, their performance on these platforms speaks volume. Besides, data scientists are leveraging media platforms such as Medium, LinkedIn, Twitter, and more, to demonstrate their expertise in data science. “Writing is the most underrated skill in data science,” said Parul Pandey, a data science evangelist at H2O.ai. Recruiters can easily find ideal candidates by evaluating their work on different developers and media platforms. A proven track record is what recruiters should seek instead of a degree in data science for making sure their ability to solving business challenges when they get hired. Online Courses Serve The Purpose As per the latest trend, the industry is witnessing a massive rise of e-learning platforms and the students enrolling in them, which portrays the potential of filling the talent shortage gaps in the data science and AI marketplace. Analytics India Magazine covers data science journey of prominent data scientists, who describe their journey and their process of gaining proficiency in data science through best practices, which, majority of the times, includes enrolling in online courses. According to them, one can accomplish great things without a PhD in data science. Outlook So does a PhD degree have any relevance in data science? Yes, in academia, a PhD degree has significant relevance; in fact, it has increased over the years. While there is no proven statistics, it is believed that data science aspirants who focus on the research area have the edge over the ones who take online courses. Research in data science is paramount as it brings new techniques in the landscape, which makes the workflow easier. In 2019, we saw numerous breakthroughs in data science and AI technologies that researchers succeeded in shifting the landscape. Consequently, until an organisation is into intensive research to make breakthroughs, they should not seek a PhD in data science from applicants. This will not only help them in hiring the right candidates but also help aspirants to get in the right firm.","excerpt":"Data scientists are now the new assets for a company, and the candidates are required to have the right ability and skills to be rewarded with this lucrative career. But now, many data science job descriptions involve the requirement of a PhD degree to apply. To ensure an effective workflow in data science initiatives, organisations […]","categories":["AI Features"],"tags":["data analytics certificate"],"author_name":"Rohit Yadav","publish_date":"2020-01-14T13:00:00","publication_year":"2020","word_count":784,"keywords":["data science","Keras","AI","PyTorch","data analytics certificate","RAG","Ray","Jupyter","analytics","TensorFlow","R"],"extracted_tech_keywords":["AI","data science","analytics","Ray","TensorFlow","PyTorch","Keras","Jupyter","RAG","R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/why-data-science-recruiters-should-stop-looking-for-phd-candidates\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10022116,"title":"How To Use AI To Fight Propaganda","content":"“All media exist to invest our lives with artificial perceptions and arbitrary values.”Marshall McLuhan The word disinformation is a cognate for the Russian dezinformatsia, the name of a KGB division devoted to propaganda. The world has come a long way from the Stalin days: USSR split, KGB became extinct, the Berlin Wall came down, and Fukuyama announced the End of History. Once the cold war ended, we thought the worst was over. Except, propaganda is a perpetual motion machine. In the age of AI, propagandists have co-opted several advanced technologies such as fake bots, deepfake videos, user-profiling or microtargeting to spread disinformation. However, the same AI can be used to counteract propaganda. In this article, we explore the factors to consider for using AI to tackle disinformation. Is AI Ready? Perfecting an algorithm requires a lot of training data. At the same time, debunking disinformation is all about timing. “A lie can travel halfway around the world while the truth is still putting on its shoes,” a saying commonly attributed to Mark Twain, is so true. Hence, a major challenge of using AI to tackle disinformation is to find relevant data to train AI. Such large labelled training data is not easy to obtain. Even if one can overcome this challenge, the data labels’ reliability cannot be established in such a short period. Besides, an adversary can always find workarounds to fool your model. Responding quickly to disinformation requires addressing the twin hurdles of limited data and unreliable or intentionally wrong data labels. Secondly, an AI model always runs the risk of over-blocking, resulting in the deletion of accurate content. AI is in a developing phase and may generate false positives\/negatives identifying accurate information as fake and vice versa. For instance, AI systems aren’t good at telling sarcasm from normal speech. Facebook’s Linformer uses machine learning to understand the nuances of human language. While the company is making progress, the CTO of Facebook, Mike Schroepfer, said deploying and maintaining such models is difficult. Thirdly, automated algorithms tend to pick up human biases and personality traits, either from the model’s creators or flawed, incomplete or unrepresentative training data. Such models could prove disastrous to individuals and communities at large. Experts also think there are no easy AI fixes to address disinformation. Samuel Woolley, assistant professor at the Moody School of Communication at UT Austin, has said it would be ‘unreasonable’ to expect any AI solution soon to quickly and unambiguously identify a rapid disinformation attack. Are Platforms Willing? WhatsApp’s end-to-end encryption blocks the messages from being analysed for disinformation and obstructs moderators from finding out the source of information. The government of India has been at loggerheads with WhatsApp in the issue of tracing the originator of messages on the platform. Last month, it even notified stricter guidelines making it mandatory for platforms such as WhatsApp to aid in identifying the origin of an ‘unlawful’ message. Secondly, all major social media platforms already have rolled out AI-enabled tools to detect disinformation. For instance, Facebook has been using SimSearchNet++, an image matching model, to detect deep fakes. UK’s intelligence agency GCHQ or other security agencies, for instance, are not allowed to run their tools on Facebook. Same with a lot of governments. But what if the government itself is the propagandist in chief? Then, it’s a whole different can of worms. Hence, social media platforms, governments, users and other stakeholders should be involved at a policy level to devise strategies — with the help of technology — to fight disinformation. Accountability The most critical question in tackling disinformation on social media is how do you define it. Governments have abused their authority to curb freedom of speech under the guise of tackling disinformation. In India, a 21-year old activist, Disha Ravi was recently arrested for ‘spreading a toolkit’. The Indian government has also directed Twitter to block multiple accounts. Later, Twitter said in a blog post: “We do not believe that the actions we have been directed to take are consistent with Indian law, and, in keeping with our principles of defending protected speech and freedom of expression, we have not taken any action on accounts that consist of news media entities, journalists, activists, and politicians. To do so, we believe, would violate their fundamental right to free expression under Indian law.” Hence, accountability is a huge problem while using AI to curb disinformation. Wrapping Up Human fact-checking is time-consuming and counterproductive, given the vast amount of fake content generated daily. We certainly need AI to tackle disinformation. But at the same time, before deploying any AI for housekeeping, it is essential to put robust frameworks in place. Further, AI also needs to overcome its technical limitations to become more efficient in taking on the disinformation juggernaut.","excerpt":"“All media exist to invest our lives with artificial perceptions and arbitrary values.” Marshall McLuhan The word disinformation is a cognate for the Russian dezinformatsia, the name of a KGB division devoted to propaganda. The world has come a long way from the Stalin days: USSR split, KGB became extinct, the Berlin Wall came down, […]","categories":["AI Features"],"tags":[],"author_name":"Kashyap Raibagi","publish_date":"2021-03-15T12:00:00","publication_year":"2021","word_count":795,"keywords":["Go","API","machine learning","programming_languages:R","AI","programming_languages:Go","GAN","R"],"extracted_tech_keywords":["AI","machine learning","R","Go","API","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-to-use-ai-to-fight-propaganda\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10016931,"title":"How Emerald Jewel Redesigned Its B2B Process Using AI-Enabled Mobile App","content":"COVID-19 pandemic has made the companies, across all domains, update their business, move more workloads to the cloud, and automate processes. Till about a few years back, one could not imagine that a traditional industry like gems and jewellery will draw on the strengths of new-age technologies like cloud and artificial intelligence. However, things have changed — not only the industry is willing to infuse AI into their workflows, but are also concerned to protect their IT infrastructure from increasing cybersecurity threats. One such company was Emerald Jewel Industry India that has been looking to start their digital transformational journey and shift their focus to automation to affect the bottom line positively. To understand the use case better, we got in touch with K Srinivasan, the managing director of Emerald Jewel Industry India. Established in 1984, the company started by working with goldsmiths and then launched its manufacturing unit in 1992. In 2020, the company aims to leverage technologies like artificial intelligence to redesign the entire buying process for its dealers. A lot of this could be attributed to the pandemic lockdown that was restricting the teams to travel to its dealers. The team was continuously facing a big challenge in communicating for new designs and showcasing their updated jewellery catalogues. “Every month, we launch more than 2,000 new designs across categories, and we need to demonstrate this to our customers as well as extend the jewellery retail design range in India,” said Srinivasan. Further, it was also becoming a struggle for the dealers to share real-time customer feedback and information on order placement and jewellery purchasing. To address these challenges, Emerald Jewel collaborated with IBM to develop an AI-based mobile application — Tej mobile app — to remodel their business process, right from jewellery design selection to placing a purchase order, using a zero-touch engagement model. Also Read: How AI Is Adding Sparkle To Jewellery Designing How Does It Work? The business model of Emerald Jewel is B2B sales, where the company markets its products to wholesalers, retailers and large retail chains. With the help of IBM’s Watson capability, the company built AI-based ‘Tej’ mobile app, the first-of-its-kind AI-enabled mobile application platform “which is a truly born-on-the-cloud solution,” said Srinivasan. The app has been designed, built and deployed by the business design arm of IBM Services, which is powered by IBM Cloud and its AI capabilities and comes with built-in IBM Security for mobile devices. The app runs on both iOS and Android operating systems and has allowed the company to onboard more dealers onto this platform, “as it now has the capacity to scale resource requirements on-demand using IBM Cloud,” added Srinivasan. The IBM Research Lab team has developed the Visual Browse, and Visual Search APIs hosted on IBM Public Cloud and secured with the IBM Security MaaS360 with Watson Endpoint Management (UEM) platform. The platform delivers AI-approach to unified endpoint management to enable end-users with apps, content and data. The mobile app allowed the in-house design team of 170 handmade and computer-aided designers based in Bombay, Delhi, Kolkata, and Coimbatore to create a more than 500,000 design bank. As a matter of fact, Emerald Jewel is currently one of the first in the B2B industry to have a digital catalogue for its dealers. This allowed the company to take newer jewellery designs to market faster through the Tej mobile app and helps the dealers to select from newer designs. Further, it also enabled thousands of retailers in the market to make the latest design options to the end consumers available. “The ability to make a design developed in Coimbatore be available to someone in Punjab as soon as the new design is introduced is only made possible through such a solution,” said Srinivasan. “Otherwise, it would have taken many days or weeks for physical designs to be sent to our dealers and receive orders.” The app connects the Emerald Jewel’s team with over 200 dealers, thereby enabling them to search through the digital catalogue of over 500,000 jewellery designs quickly and navigate the complex ordering process easier. Also Read: How This Fashion-Tech Startup Is Using AR To Create New-Age Jewellery Shopping Experience Wrapping Up With the partnership with IBM, Emerald Jewel Industry India is confident of increasing its manufacturing ability significantly to three tonnes per month. Further, with the capacity to scale resource requirements on-demand using IBM Cloud, the company is positive to scale their business and onboard hundreds of new dealers with absolute ease, in the coming year. “Alongside, using the Tej application means that any authorised dealer can start doing business with Emerald Jewellery, using a smart tablet,” concluded Srinivasan.","excerpt":"COVID-19 pandemic has made the companies, across all domains, update their business, move more workloads to the cloud, and automate processes. Till about a few years back, one could not imagine that a traditional industry like gems and jewellery will draw on the strengths of new-age technologies like cloud and artificial intelligence.  However, things have […]","categories":["IT Services"],"tags":["how does ai work"],"author_name":"Sejuti Das","publish_date":"2021-01-01T16:00:00","publication_year":"2021","word_count":775,"keywords":["Go","API","artificial intelligence","AI","how does ai work","digital transformation","Git","RAG","automation","Aim","R"],"extracted_tech_keywords":["AI","artificial intelligence","Aim","RAG","R","Go","Git","API","digital transformation","automation"],"url":"https:\/\/analyticsindiamag.com\/it-services\/how-emerald-jewel-redesigned-its-b2b-process-using-ai-enabled-mobile-app\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10100931,"title":"Is Metaverse For Real?","content":"Zuckerberg might have just infused life into the Metaverse—by ‘life,’ we mean literally. He recently appeared on a podcast with Lex Fridman, and guess what? It was the first interview to happen in the Metaverse. 13 months of progress fueled by Zuck’s unstoppable desire to prove the haters wrong. pic.twitter.com\/p2DvWDLWpG— Ryan Delk (@delk) September 29, 2023 The most fascinating thing about the podcast was that the avatars of Fridman and Zuckerberg were not cartoons but photorealistic. Gone are the days of cartoon avatars in the Metaverse. “Mark and I are hundreds of miles apart in physical space, but it feels like we’re in the same room because we appear to each other as photorealistic,” exclaimed Fridman. One might wonder, how was this made possible? These photorealistic avatars were created using Codec Avatars technology. Meta started working on Codec avatars in 2019 as part of a Reality Labs (RL) Research project. These are created by capturing the user’s facial expressions and body movements using cameras and sensors. This data is then compressed and transmitted to another device, where it is decoded and used to create a digital avatar of the user. The avatars can be used to interact with other users in VR or AR environments. Fridman was at a loss for words to describe the Codec avatars in 3D. “It’s hard to put into words how awesome this was for someone like me who values the intimacy of in-person conversation”. “With spatial audio this technology is incredible. I think it’s the future of how human beings connect to each other in a deeply meaningful way on the internet,” he added. Emotion Matters “These avatars can capture many of the nuances of facial expressions” beamed Fridman. The idea of the Metaverse was always to connect people living far away from each other. The problem was that, with cartoon avatars, one couldn’t really express their emotions. However, these new avatars are capable of capturing human emotions by projecting small nuances, from the movement of the eyes to the twitching of the eyebrows. “Eyes are a huge part of it; I mean, there are all the studies that show most communication happens through non-verbal cues, such as expressions. So, we try to capture that with the classical expressive avatar system,” said Zuckerberg. Furthermore, Zuckerberg added that Meta is also working on a  quick and easy way to generate an avatar using your mobile phone, what he calls Instant Codec Avatars. Codec avatars have the potential to change the way humans interact. First, they can provide people with the opportunity to socialize with others in a virtual environment, even if they are not physically close to each other. This can be especially beneficial for people who live in isolated areas or who have difficulty leaving their homes. Furthermore, it can help people to feel more connected to others by allowing them to see and hear each other in real-time. This can create a more personal and engaging experience than traditional text-based or audio-based communication. “My family lives across five countries, and I can’t wait for this technology to be widely available,” said Martin Harbech, group director, Meta. https:\/\/twitter.com\/heyBarsee\/status\/1707639144111866034 Additionally, Codec avatars can be used to create virtual worlds that are tailored to the individual’s interests. This can provide people with a place to go where they can feel comfortable and accepted, and where they can connect with others who share their interests. New World Imagine a world where digital and reality come together. Yes, with Meta AI characters in the Metaverse, it is very much possible to create a world where you would be occupying a space where AI-generated avatars and humans are present together in a shared environment. This idea goes well as Meta recently introduced new AI experiences across a family of apps like Instagram and WhatsApp, including  28 AI characters with unique interests and personalities. Zuckerberg highly anticipates that in the near future, business meetings will be conducted in the metaverse alongside virtual AI characters. This seems to be the beginning of a newly simulated world. “We will have meetings in the future where you’re basically sitting there physically, and then you have a couple of other people who are holograms. Then you have someone like Bob, the AI engineer on your team, who’s helping with things and can now be embodied as a realistic avatar and just join the meeting. I think that’s going to be pretty compelling.” Jim Fan, Senior AI Scientist at NVIDIA, was thoroughly impressed with the podcast and aligned himself with the idea of the metaverse. “The ultimate vision is to realize the scenes in Matrix: full-body, real-time avatars of both humans and AI, sharing the same virtual space, interacting with objects in physically realistic ways, receiving rich multimodal feedback, and forgetting that the world is but a simulation.” he posted on X. These Codec avatars might completely change the fate of the metaverse. Many internet users claim that the Metaverse is here to stay, and this time, it’s real. The fact that Zuckerberg didn’t abandon the Metaverse speaks volumes about his firm belief in his pet project. With Quest 3 on the horizon, an AR\/VR headset, Meta might change the way we interact with others. “I’m a big believer in the metaverse and always have been,” said Aleksa Gordic, formerly with Microsoft and Google DeepMind, who worked on this technology back in 2018-19 as part of the Microsoft HoloLens project.","excerpt":"Gone are the days of cartoon avatars in the metaverse","categories":["AI Features"],"tags":[],"author_name":"Siddharth Jindal","publish_date":"2023-09-30T17:00:00","publication_year":"2023","word_count":905,"keywords":["Go","Meta AI","programming_languages:R","AI","Modal","programming_languages:Go","Git","CuPy","Aim","R"],"extracted_tech_keywords":["AI","Meta AI","Aim","CuPy","R","Go","Git","Modal","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/is-metaverse-for-real\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":43869,"title":"Karnataka, UP To Install AI Systems In Buses To Avoid Accidents","content":"Two Indian states Uttar Pradesh and Karnataka are set to pioneer the use of artificial intelligence in the public transport system. Reportedly these two states are set to install AI-powered anti-collision systems to make sure that the buses avoid any types of accidents. According to a report in a noted news portal, this anti-collision system, powered by AI, comprises two sensors. One is fitted in the front bumper and continually tracks and observes objects within a range of up to 180 feet. It alerts the bus driver of any impending danger or possibility of a collision via an aural alert. The second sensor is fitted near the headlight switch and alerts the driver in case he is inattentive or sleepy. These sensors have been created by the Sixth Sensor Technology, a Mumbai-based startup. Paritosh Dagli, the director at the company told the newspaper “The driver needs to wave his hand across the sensor at regular intervals, which can be customised, and if he does not perform the action, the system concludes that the driver is sleepy or inattentive and it cuts off the accelerator. One he waves the hard, the accelerator is resumed immediately.” This step is an important move towards the Central Government’s efforts in making the Indian roads safe. Only last month, the Transport and Highways Minister Nitin Gadkari had talked about the rising number of accidents in the country. Gadkari had also addressed the Lok Sabha on a bill to amend the Motor Vehicles Act of 1988. This new bill proposed provisions include increased penalties for offences like driving under the influence of drugs or alcohol. He had said, “It is sad that India is on top in the number of deaths due to road accidents,” Mr Gadkari said while taking up the bill, adding, “Even after making full efforts from my side, deaths have only come down by 3-4%. I have failed in it, I accept it.”","excerpt":"Two Indian states Uttar Pradesh and Karnataka are set to pioneer the use of artificial intelligence in the public transport system. Reportedly these two states are set to install AI-powered anti-collision systems to make sure that the buses avoid any types of accidents. According to a report in a noted news portal, this anti-collision system, […]","categories":["AI News"],"tags":["AI (Artificial Intelligence)"],"author_name":"Prajakta Hebbar","publish_date":"2019-08-06T13:09:34","publication_year":"2019","word_count":321,"keywords":["Go","artificial intelligence","programming_languages:R","AI","programming_languages:Go","R","AI (Artificial Intelligence)","startup"],"extracted_tech_keywords":["AI","artificial intelligence","R","Go","startup","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/karnataka-up-to-install-ai-systems-in-buses-to-avoid-accidents\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10171191,"title":"AWS Summit Mumbai 2025 Brings Cloud and AI Leaders Under One Roof","content":"AWS Summit Mumbai, one of India’s most anticipated AI events, is set to take place at the Jio World Convention Centre on June 19, 2025. This one-day free conference, which will run from 10 am to 4:30 pm, brings together pioneers of innovation, top business minds, and technology experts from across the country for a powerful deep dive into the future of cloud computing and generative AI. Under the theme ‘From India to the world: Innovating the future’, the summit serves as a platform for conversations that explore how enterprises in India are transforming industries and scaling their innovations globally with AWS. Where:  Jio World Convention Center When: June 19, 2025 Register here A Future-Ready Experience Whether you’re a C-suite executive shaping enterprise strategy or a developer building the next big product, the AWS Summit Mumbai promises actionable insights for every level of the cloud journey. This year’s edition focuses on how sectors like manufacturing, financial services, media, and technology are unlocking business value using AWS’s enterprise-grade infrastructure, advanced AI capabilities, and a robust security foundation. Expect a high-energy day packed with: 1 keynote from AWS leaders and industry technologists 30+ breakout sessions covering topics like generative AI, data strategy, cloud modernisation, and industry-specific innovations 50+ partners showcasing AWS-powered solutions and tools 50+ demos and interactive experiences, including hands-on workshops and lightning talks Who Should Attend? The summit brings together professionals driving digital and business transformation. Whether you’re exploring the cloud for the first time or refining an existing architecture, you’ll gain strategies, hear success stories, and make connections to help advance your organisation. Ideal For: CXOs, CIOs, and CTOs Digital transformation leaders and business strategists IT directors, solution architects, and infrastructure engineers Data scientists, ML engineers, and DevOps teams Cloud architects, developers, and security specialists A Full-Day Agenda to Power Your Thinking 8:00 am – Registration and networking breakfast10:00 am – Welcome note10:30 am – Keynote featuring AWS executives and tech leaders12:00 pm – Lunch1:30 pm – Breakout sessions across verticals and technologies4:00 pm – Closing keynote4:30 pm – High tea and networking Don’t Miss These Experiences Generative AI Experience Explore the cutting edge of AI with 30+ sessions, demos, and hands-on experiences in the AI Hub and Builders Zone. Try out Amazon Bedrock, Amazon Q, and PartyRock, and discover how to build secure, responsible generative AI solutions at scale. Business Innovation Hub From Startups to SMBs to Public Sector innovators and ISVs, the Business Innovation Hub brings four key communities under one roof. With live demos, networking lounges, and AWS resources, each zone offers tailored guidance for every stage of business growth. Industry Hub Get sector-specific insights into how AWS is transforming financial services, retail, healthcare, manufacturing, and media. Talk to experts, watch live demos, and hear from customers driving tangible results with AWS technology. Why You Should Attend India is fast becoming a global hub for cloud-powered innovation, and AWS Summit Mumbai is where this momentum comes alive. If you’re looking to drive impact, scale responsibly, and reimagine your organisation’s future with cloud and AI, this is where you start. Register now and be part of the movement shaping India’s digital future. Where:  Jio World  Convention Center When: June 19, 2025 Register here","excerpt":"The summit serves as a platform for conversations that explore how enterprises in India are transforming industries and scaling their innovations globally with AWS.","categories":["AI Highlights"],"tags":["AWS"],"author_name":"Siddharth Jindal","publish_date":"2025-06-03T15:03:41","publication_year":"2025","word_count":536,"keywords":["API","AWS","cloud computing","AI","ML","Git","generative AI","GAN","DevOps","R"],"extracted_tech_keywords":["AI","ML","generative AI","cloud computing","AWS","R","Git","DevOps","API","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/aws-summit-mumbai-2025-brings-cloud-and-ai-leaders-under-one-roof\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10165758,"title":"Pega India as the Powerhouse Miniature","content":"For global capability centres (GCCs) in India to become self-sufficient, the journey starts with aligning strategic priorities with their capabilities to maximise impact. But true self-sufficiency goes beyond mirroring global operations—it requires building strong core skills and deep expertise in key areas like transition risk, sustainability, and business transformation. Right from the start, Pega India, part of the global Pegasystems, has focused on building the right culture, one that fosters innovation, accountability, and ownership. In an exclusive interview with AIM, Goutham Parcha, vice president of the applications group at Pega, said, “We never wanted to be just an execution-focused office.” The GCC, he said, was a microcosm of the global Pega, having people in multiple business functions. The company has nurtured talent, allowing employees to take on global roles and leadership positions. The long tenure of many employees is proof of the strong foundation built over the years. Deepak Visweswaraiah, vice president of platform engineering and site managing director at Pegasystems, shared a similar sentiment. He told AIM that GCCs must focus on leveraging talent and integrating the local centre into the company’s global strategy rather than cost-cutting. Today, Pega India has around 1,900 employees, based mainly in Hyderabad (1,200 people) and Bengaluru (450 people), with some working remotely. Parcha further states that the way these teams collaborate across different functions makes Pega India a true reflection of global Pega. Building the Future with Pega Cloud Pega India’s growth is driven by one of its biggest innovations, Pega Cloud, which helps clients build and deploy applications on cloud infrastructure. Parcha said that the India team plays a significant role in cloud engineering and client management, especially in helping businesses transition from on-premise setups to the cloud. He further added that Pega cloud is hosted on AWS and Google Cloud by default, which connects to GenAI models. Another area of strength for Pega in India is its core product engineering, which has been the GCC’s expertise for over 15 years. A standout achievement in this space is Constellation technology. “If someone wants to build an application on Microsoft or Java, one of their biggest struggles is getting the UI\/UX right,” Parcha said. “They have to hire experts in React, Angular, and JavaScript. But with Constellation, we take care of all that. You don’t need to be a front-end expert—you can build world-class applications with ease.” AI Running the Show Parcha said that Pega India takes great pride in its AI center of excellence. “There’s a lot of work going on in the AI front, both the statistical and the traditional AI. Over the last 10+ years we’ve invested a lot in our AI models, be it adaptive models, predictive models or NLP.” A key moment in this journey was Pega’s acquisition of Chordiant in 2010, which significantly strengthened the company’s AI capabilities. While Pega has built strong expertise in predictive AI, it takes a different approach to generative AI. Instead of developing its own large language models (LLMs), the company integrates with the best models available in the industry. “On the generative AI side, we’re keeping it very open-ended,” Parcha explained. “We connect with Microsoft, OpenAI, Google’s large language models, and others. Models are evolving rapidly, and we’re not in the business of building them. Instead, we focus on governance, security, scalability, and resilience,” he clarified. Pega’s Fascination for Campus Hires Pega India is continuously expanding its team, having hired over 200 new employees last year alone. The company recruits talent at multiple levels, maintaining strong partnerships with universities through its campus hiring programs. Pega actively engages with IITs, NITs, BITS, and other regional colleges, bringing in fresh graduates. “There’s a lot of lateral hiring going on, but hiring at the university level has worked out wonderfully well for us. I think we have a strong appreciation for the campus hires from these colleges,” Parcha said. In addition to campus hiring, Pega India conducts lateral hiring for specialised roles, ensuring the right expertise is brought in where needed. “40% of our individual contributors are people who have more than 10 years of experience, which is a rare thing in India because Indians are fascinated with becoming managers,” Parcha pointed out. He further added that this depth of experience allows new campus hires to learn from industry veterans, making mentorship a core part of the organisation. The Story of Blueprint Pega India fosters an innovative spirit, where ideas are encouraged, explored, and turned into real products. A prime example of this is Pega GenAI Blueprint, a new product that was conceived in a hackathon and quickly developed into a full-fledged offering. By combining generative AI with decades of expertise in building applications, Blueprint can instantly create a workflow, data model, and personas for the application. Pega, as a strategy, does not actively sell its solutions in India but focuses on global markets. As Parcha stated, “We don’t have a strategy or we’re not proactively selling within the region, but it’s all about somebody reaching out to us, and we go and help them out.” However, there have been instances, as Parcha referred to, where Indian state governments have approached the company to manage citizen services. Pega’s technology has been utilised for various government functions, including college admissions counselling, loan disbursements for underprivileged citizens, and applications for government-backed broadband connections.","excerpt":"The GCC has around 1,900 employees, based mainly in Hyderabad (1,200 people) and Bengaluru (450 people)","categories":["GCC"],"tags":["AI Cloud","GCC","Pegasystems"],"author_name":"Shalini Mondal","publish_date":"2025-03-10T17:32:26","publication_year":"2025","word_count":887,"keywords":["GenAI","GCC","OpenAI","AI","AWS","AI Cloud","RAG","NLP","Pegasystems","Aim","generative AI","JavaScript","R"],"extracted_tech_keywords":["AI","NLP","generative AI","GenAI","OpenAI","Aim","RAG","AWS","R","JavaScript"],"url":"https:\/\/analyticsindiamag.com\/gcc\/pega-india-as-the-powerhouse-miniature\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10078995,"title":"COP 27: Green India Seeks Clarity on Climate Finance","content":"The 27th Conference of the Parties to the United Nations Framework Convention on Climate Change—COP 27, was held on November 6, 2022 in the coastal city of Sharm el-Sheikh city in Egypt. Explaining India’s stance at the conference, Bhupender Yadav, Union Minister of Environment, Forest and Climate Change, said in a press release, “India looks forward to substantial progress in the discussions related to climate finance. We also look forward to the introduction of new technologies and new collaborations to facilitate technology transfers.” India’s main objective at COP 27 According to the UN, the planet needs to invest almost $90 trillion in climate-related infrastructure over the next 15 years in order to reduce carbon emissions. According to their estimates, countries may generate an estimated $4 for every $1 spent by switching to a green economy. In line with this, the term “Climate Finance” was coined, and the developed countries participating in COP 15 in 2009 committed to a collective goal of mobilising USD 100 billion per year by 2020 for climate action in developing countries. However, as shown in the above graph, the goal of contributing $100 billion per year till 2020 and $100 billion each year till 2025 is yet to be achieved. As per the press release by Press Information Bureau (PIB), “While the promised amount must be reached as quickly as possible, there is a need now to substantially enhance the ambition to ensure adequate resource flow under the new quantified goal post-2024.” As per reports, to meet its clean energy goals by 2030, India needs an estimated $230 billion. According to a senior government official, “India has made it adequately clear that it is the historical responsibility of rich countries to provide the necessary climate funding.” India’s main goal, thus, at COP 27 will be to force the UN to define what “Climate Finance” entails. For developing nations to effectively gauge the magnitude of financial inflows for climate action, the government contends that the notion of climate financing needs to be explained unambiguously. A report on the various definitions will be provided by the Standing Committee on Finance. India’s contribution to global carbon emission So far, India has contributed comparatively less to global carbon emissions than both China and the US. The nation, which has a population of about 1.4 billion, has so far emitted significantly less carbon dioxide than is reasonable. According to the Indian government, because it is a growing nation with sustainable development and poverty eradication among its top priorities, its emissions are likely to increase. The total amount of carbon dioxide (CO2) emissions allowed over time with the possibility of keeping global warming to 1.5 °C above pre-industrial (1850–1900) levels is known as the global carbon budget. According to Union Minister Ashwini Kumar Choubey’s statement to the Rajya Sabha, India’s per capita emissions are 1.96 tonnes which is less than one-third of the world’s per capita GHG (greenhouse gas) emissions. He further noted that India’s annual emissions in 2016 are only about 5% of the global emissions. “India has contributed only around 4% of global cumulative emissions from 1850 to 2019, despite being home to around one-sixth of humanity,” quoted Choubey. It should also be emphasised that India is the only G20 nation on track to reach the goals set forth in the Paris Climate Agreement, according to the UN Environment Program’s Emission Gap Report. The country intends to decrease its GDP’s emission intensity, or the amount of carbon emissions produced for every unit of GDP, from 2005 levels to an estimated 35% by 2030. The Indian government firmly believes that the country will reach its goal of generating at least 40% of its electricity from non-fossil fuel sources by 2023, seven years before the target year of 2030. (India’s growing Solar Energy usage)","excerpt":"India’s main goal at COP 27 will be to force the UN to define what “Climate Finance” entails.","categories":["AI Features"],"tags":["Paris Agreement"],"author_name":"Lokesh Choudhary","publish_date":"2022-11-08T13:00:00","publication_year":"2022","word_count":632,"keywords":["Go","API","funding","programming_languages:R","AI","ML","programming_languages:Go","Paris Agreement","R"],"extracted_tech_keywords":["AI","ML","R","Go","API","funding","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/cop-27-green-india-seeks-clarity-on-climate-finance\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":17999,"title":"S Vaitheeswaran, MD And CEO, Manipal Global Education Explains The Missing D In Data Science","content":"Data Science is the latest buzzword in the industry and is looked upon as the next big thing in the industry. Its growing popularity can be identified from a recent study by Glassdoor that identified Data Scientists as the top job in the USA. And the scenario is no different in India. An increase in the amount of data generated by industries is one of reasons why data science has become so popular over the years. S Vaitheeswaran, Managing Director and CEO, Manipal Global Education, during a talk at Cypher 2017 quickly confronts that during a meeting with fresh grads in one of the Manipal Universities, there was repeatedly a mention about artificial intelligence, machine learning and a willingness to make a career in data science.But is data science and related technologies such as AI being explored to the extent that it should be? A difference of opinions in these areas, especially if AI is good for people or not, has been quite prominent—and it was recently witnessed in the form of famous Musk and Zuckerberg fight. Not just AI, but an increasing accessibility to personal and sensitive data such as that from Aadhar has invited debates of privacy breach and other risks. Vaitheeswaran remarks that in today’s world others may know you better than you know yourself—thanks to the ease of collecting data and accessing it. Data Science—a disrupting area with challenges “Any disruption that happens brings opportunities. We saw it thirty years back in the IT world and now a disruption in the area of data science, analytics is enabling companies, young startups, government and professionals to leverage it”, said Vaitheeswaran. He also noted that amidst the gaining popularity that it offers, it is important to make this subject more encompassing, pervasive and avoid hubris. There is no doubt that data science has become a hugely adopted area in most companies, but most of them jargonize the area way too much, to an extent of making it look a little hostile. “Subjects like data science and analytics should become more pervasive and we must open it out to everybody in the system. People must not find it difficult to understand the subject”, said Vaitheeswaran. After all, it is important to note how data science can impact people in everyday life. At the end of the day, data science is not just about spreadsheets and algorithms but in understanding how everything in everyday life can relate to data science. Not just technology and business, but people matter too Given the adoption of any technology, at the end of the day it is important to find solution for various problems. “You are going to miss the big picture if you don’t relate to the problem”, said Vaitheeswaran. He further exemplified that Manipal hospital extensively makes use of Watson in everyday life but do the makers of Watson really understand the domain for which they are solving the problem? Have they ever been to the operation theatre? “They make solutions but may not come in terms with the real time implementation of the products that they are offering for various industries”, he said. This remains a major challenge. The other thing is how this space can impact the society. “It may be great to work for the companies and people, but in a country like ours, it is important to understand how big data, analytics and similar technologies can impact our social ecosystem”. There are companies coming in this space using predictive analytics, to manage the crowd in Kumb Mela, for instance. They are using drones and creating backend database to solve problems such as stampedes. It is time that more companies come into this space with an intention of serving the community, to make this industry larger by using complex tools. “The appetite to learn is good in the country and therefore it should be used to explore larger areas”, he said. Though a lot of growth is happening in this area, such as use of deep learning to understand the ocean, using predictive analytics to reduce the effects of natural disasters, people should try working with social organizations and NGOs more. On a concluding note he said that dashboard for soft metrics are certainly more difficult than dashboards for operating metrics that we all are comfortable with. “For me this missing D in data science could be domain, direction or the demographic dividend. Rest it’s a responsibility for all of us to identify the missing D”.","excerpt":"Data Science is the latest buzzword in the industry and is looked upon as the next big thing in the industry. Its growing popularity can be identified from a recent study by Glassdoor that identified Data Scientists as the top job in the USA. And the scenario is no different in India. An increase in […]","categories":["Deep Tech"],"tags":[],"author_name":"Srishti Deoras","publish_date":"2017-10-03T09:39:33","publication_year":"2017","word_count":747,"keywords":["data science","Go","machine learning","artificial intelligence","AI","R","RAG","deep learning","analytics","predictive analytics"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","deep learning","data science","analytics","RAG","predictive analytics","R","Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/s-vaitheeswaran-md-ceo-manipal-global-education-explains-missing-d-data-science\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10167968,"title":"This Company is Using AI to Make RegTech Less Boring","content":"When you think of financial services, the first thing that likely comes to mind isn’t innovation. Instead, you might envision the endless paperwork, regulatory red tape, and Excel sheets that seem to stretch on indefinitely. However, things are changing quickly and for the better. With generative AI agents now capable of acting like hyper-efficient virtual coworkers, even the most bureaucratic corners of finance are getting a serious upgrade. However, Broadridge’s foray into using AI in this isn’t new. It happened long before everyone started talking about ChatGPT. For over a decade, the New York-based technology and infrastructure provider has been experimenting with building its own predictive and analytical models primarily for efficiency, and they have been in place for years. Over time, the AI use cases have matured and scaled. From parsing unstructured regulatory filings to forecasting shareholder voting outcomes, Broadridge is helping financial institutions integrate AI not just to automate, but to anticipate. Inigo Fernando, the CTO for Broadridge India’s governance and communications business unit, told AIM about the journey. The Regtech Industry Broadridge powers investing, governance, and communications for its clients, with its operations platforms processing and generating over 7 billion communications per year and underpinning the daily trading of more than $10 trillion of securities globally. Their governance and communications unit is focused on what Fernando calls “regtech”—regulatory technology. This is a space where AI is becoming increasingly critical. The core responsibility of Broadridge is to help publicly traded companies communicate with their investors, something that’s not just good practice but a regulatory mandate. “This is part of the ‘G’ in ESG—governance,” he said. When it comes to using AI, Fernando believes most of AI is becoming a commodity. “You don’t need to have deep AI knowledge to build AI-enhanced tools now. It’s all about identifying the right use case.” One of the most impactful applications of AI at Broadridge has been in the space of investor insights. “Every year, companies hold their AGMs, where investors vote on everything from electing directors to approving mergers. We’ve developed AI models that help these companies predict voting outcomes in advance—whether quorum will be met, whether a proposal will pass, and so on,” Fernando shared. Behind the scenes, several gradient-boosting algorithms crunch decades of historical voting patterns, investor behaviour, and industry trends to provide predictive intelligence. “The models are built using massive volumes of structured data. The real secret sauce isn’t just the model—it’s the depth of historical data we have from being in the industry for over 40 years,” he added. “We go through a rigorous evaluation of various models—security, accuracy, suitability to the data structure. Older AI models worked well for us in this case. But we’re always keeping an eye on emerging tools or even more advanced neural networks.” Another powerful example of Broadridge’s AI deployment lies in regulatory document simplification. He explained that the SEC introduced a regulation around Tailored Shareholder Reports, allowing mutual funds and ETFs to replace their massive 80–100-page reports with easy-to-read two-pagers. “For us, that meant taking thousands of unstructured filings, extracting meaningful financial data, and distilling it into a concise, investor-friendly format. We used several LLMs for this. It’s a perfect fit when AI reads like a financial analyst and writes like a human.” Predicting the stock market trends is one thing that Fernando believes is possible as well. “I don’t think humans would rule that possibility out. I wouldn’t rule that possibility out either,” he commented. For example, last year, LTX, the AI-driven corporate bond trading platform backed by Broadridge Financial Solutions, unveiled a GenAI-powered List Trading feature. Expanding its leadership in the fixed income space, the platform now enables users to create and execute complex, multi-asset, multi-directional trade lists using GPT technology through its RFQ+ system. Internal Use Cases But not all AI at Broadridge is flashy. Some of it is quietly transforming back-office operations. Many of the operator tasks are repetitive, time-consuming, and prone to error. Broadridge started introducing AI modules to assist decision-making in these processes. “But we’re careful. We always keep a human in the loop to ensure accuracy and compliance.” The company has also experimented with recommender systems for bond trading, helping investors find potential counterparties by analysing historical transaction data using neural networks. Broadridge plans to double down on LLMs and generative AI in the near future. “We’re already using LLMs for things like document summarisation and understanding unstructured data. For structured data predictions, traditional models still work better. But the lines are blurring,” he said. On the other end of Broadridge, at the capital markets, wealth, and asset management business unit, the firm also uses AI tools like OpsGPT, BondGPT, Distribution GPT, and BroadGPT, enabling advanced automation, real-time insights, and enhanced operational intelligence. Kishore Seshagiri, executive director and chief digital officer, Broadridge India, told AIM that as AI adoption accelerates, the team remains committed to responsible AI integration, prioritising transparency, security, and ethical governance. For example, BondGPT is trusted by over 700 traders and streamlines bond trading and portfolio management by using OpenAI GPT-4 to answer bond-related queries, enhance dealer-client engagement, and optimise decision-making. Another example is BroadGPT, which is a productivity assistant, helping over 15,000 associates enhance their work quality by streamlining content creation, translating surveys and reports, assisting with pair programming, and serving as a valuable learning tool.","excerpt":"“You don’t need to have deep AI knowledge to build AI-enhanced tools now. It’s all about identifying the right use case.”","categories":["AI Features"],"tags":[],"author_name":"Mohit Pandey","publish_date":"2025-04-15T12:02:53","publication_year":"2025","word_count":887,"keywords":["ChatGPT","GenAI","OpenAI","AI","neural network","ML","Ray","Aim","generative AI","R"],"extracted_tech_keywords":["AI","ML","neural network","generative AI","GenAI","ChatGPT","OpenAI","Aim","Ray","R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/this-company-is-using-ai-to-make-regtech-less-boring\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":68627,"title":"Are MOOCs A Sure Shot Way To Make A Career In Data Science Or Are They Merely Superficial","content":"MOOCs or Massive Open Online Courses for data science is one of the most preferred options to upskill and learn new technologies in this rapidly evolving industry. Not to confuse with online courses, MOOCs are more context-driven, and its dynamic building up of context around content makes it unique. Usually aimed at micro-learning, they are useful when getting a deeper understanding of the subject is not the main focus. Mostly available for free, many companies such as Google and Microsoft have MOOCs that promise to provide high-quality training courses. There are academies such as Udemy, Coursera that offer data science MOOCs. With the increasing number of MOOCs now available, the number of people who have utilised it has grown dramatically over the years. While it gives the liberty to learn at their own pace, there is also a widespread view that MOOCs are nothing more than superficial content, and cannot help you in establishing a career in data science. Let us dig in deeper in this context with this article, where we try to find out if MOOCs can help make a career in data science or are they merely superficial. MOOCs For Establishing Data Science Career There are many data science professionals today who claim to have achieved what they have by entirely relying on MOOCs. Data scientist, influencer and blogger, Rahul Agrawal, from Walmart Labs in an interview with Analytics India Magazine had said that he had relied entirely on MOOCs to learn and master data science techniques. He emphasised on the fact that it allows for learning without being in the classroom, and that it is still an essential part of his learning strategy. With more than 30 certifications, he still looks for courses that he could learn that might prove helpful in his work. Echoing similar views, Saurabh Jha, who is the Director of data science at Dell, entirely relied on MOOC for making a career in data science. With the initial struggle of making out what is best and what is not, he finally cracked up the concept of making the most out of MOOCs. He vouches for the course by Andrew NG as it is simple and easy to understand. Similarly, Rajiv Hiremagalur started learning from the courses provided by MOOCs when he first learnt R programming in 2014 and then moved on to Python and so on. To stay abreast of the new advancements, he used to create a list of techniques to master. This helped him in learning one skill at a time while avoiding burnouts by learning everything together. Considering these opinions, one of the significant advantages of MOOCs, unlike full-time or part-time data science courses, is that it keeps up with the ever-changing data science landscape. It comes handy when adding one skill at a time. On The Flip Side Many recruiters believe that MOOCs alone cannot help in landing a data science job. Ramasubramanian Sundararajan, Head AI Lab at Cartesian Consulting, believes that MOOCs are suitable only for continuing education and upskilling for those who have established themselves in the industry. For freshers or those trying to set out a foot in the industry might find classroom coaching as a better alternative than MOOCs. Babak Beheshti, IEEE Member, professor and dean at College of Engineering and Computing Sciences, New York Institute of Technology, shared that while MOOCs have proved to be an innovative paradigm in education and providing an alternative for specialisation while gaining micro-credentials, relying solely on them may not help in establishing a career in data science. It is best suited for those looking for affordable education. Going through various online forums, many recruiters believe that MOOCs may not have much corporate value. While it is good to add another certification in your profile, getting a specific career boost through this is less significant than those by executive premium courses from the likes of IITs or IIMs. Another popular thought is that MOOCs are a good option for those who want to upskill or switch careers but not an ideal way for students to go about. MOOCs are usually faced-paced and short and may not give an in-depth understanding of the subject as a full-time course does. “Someone who has learned through online courses rigorously tries to apply random techniques. But someone who has thought deeply while pursuing M.Tech or PhD knows how to look at problems in a more detailed way instead of applying whatever is available readily at hand; It’s basically using the tool versus knowing exactly how the tool works,” said  Manish Gupta, Principal Applied Scientist at Microsoft, in an interview with Analytics India Magazine. Data science cannot be learnt in a few months; it’s a long process. Thus, full-time courses are the best way to go in-depth and differentiate from others. Nevertheless, Gupta said that working professionals are left with fewer options and should leverage online courses to learn in a guided manner instead of learning everything by themselves. Given the fact that MOOCs are too easy to accomplish makes it not as likely an option as full-time courses among recruiters. It may not provide a deeper understanding of data analysis techniques. For instance, it cannot make you strong with math skills without having a strong background already. So, while it may help professionals to upskill, they may not be enough to get a data science job. Wrapping Up MOOCs are an excellent way to keep the skills relevant and fresh as it may help in boosting your career but depending solely on it may sometimes work in your favour and at other times not. Especially if there is no thorough understanding of the skills acquired, investing time in MOOCs may go wasted entirely. It also depends on where you have taken it from, the credibility of it and how much knowledge you have been able to acquire from it. Data science is a practice-driven profession in many ways, and working on projects and internships are instead looked upon as a credible resource of your understanding in the subject than just adding MOOCs certifications. Recruiters might consider looking at the real-time data quality issues that you might have encountered and how you dealt with them, communication, data storytelling skills, and more. Data science is a multifaceted subject and having a thorough understanding of math, statistics and programming are highly critical. MOOCs may be of great value if you already have some experience and degree in data science, looking to upskill in a few areas of data science and if you do not have that kind of money to spend on regular full-time courses.","excerpt":"MOOCs or Massive Open Online Courses for data science is one of the most preferred options to upskill and learn new technologies in this rapidly evolving industry. Not to confuse with online courses, MOOCs are more context-driven, and its dynamic building up of context around content makes it unique. Usually aimed at micro-learning, they are […]","categories":["AI Highlights"],"tags":["Data Analytics Certification","Data Science Career","Data Science Certification","innovative technology advancements","MOOCs","principal data scientist"],"author_name":"Srishti Deoras","publish_date":"2020-07-01T14:00:00","publication_year":"2020","word_count":1100,"keywords":["data science","Go","API","innovative technology advancements","principal data scientist","Data Science Certification","AI","MOOCs","RAG","Data Science Career","Python","Aim","analytics","data quality","Data Analytics Certification","R"],"extracted_tech_keywords":["AI","data science","analytics","Aim","RAG","Python","R","Go","API","data quality"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/are-moocs-a-sure-shot-way-to-make-a-career-in-data-science-or-are-they-merely-superficial\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":37468,"title":"Is It Hyperbolic To Compare Artificial Intelligence To Nuclear Weapons?","content":"As development in the field of artificial intelligence witnesses revolutionary changes, several noted names and visionaries have warned the world about the grave consequences of AI. In fact, the reliance of AI is being likened to something as lethal as nuclear weapons. Reiterating his position on the dangers of AI, founder and former CEO of Microsoft, Bill Gates, in a recent interaction at Stanford University compared AI to nuclear weapons. Stating that technology to be both promising and dangerous Gates said, “The world hasn’t had that many technologies that are both promising and dangerous. We had nuclear weapons and nuclear energy, and so far so good.” He then went on to add that the focus of AI should be about helping people get medical care and said that the technology should be leveraged to identify promising drugs and improve development drug development process,“I do not believe without machine learning techniques we would able to take the dimensionality of this problem to find the solution,” Gates added. Why Are Big Names In The Industry Distrustful Of AI? Gates alone isn’t the only big names among tech entrepreneurs to make such a claim, back in 2017, Elon Musk in one of his early Twitter musings warned the world that it is AI and not Kim Jong Un more dangerous to the world. Musk has been an open critic of the potential harm of the technology and has called out on Facebook’s many initiatives that explore AI. His scepticism led to the creation of OpenAI, a platform that was formed as a counterbalance to promote ethical AI, however, in February 2018, Musk left the group owing to differences in opinion with fellow members. Almost a year later, he took to Twitter to explain his position, “I had to focus on solving a painfully large number of engineering and manufacturing problems at Tesla (especially) and SpaceX.  Also, Tesla was competing for some of the same people as OpenAI and I didn’t agree with some of what OpenAI team wanted to do. Add that all up and it was just better to part ways on good terms,” Musk Tweeted. Interestingly, the technology didn’t escape the criticism from some of the prominent names from India which includes Infosys co-founder, Narayana Murthy and Sashi Tharoor, who stated that the fear of job associated with the technology to be taken into consideration very seriously. Is AI As Lethal As Nuclear Weapons As the scepticism around the technology grows, experts believe that the technology can help in de-escalation of nuclear weapons and could also lead to potential nuclear destruction, but the challenge lies limiting the existing weak AI system to do only certain jobs. In a paper titled, “How Might Artificial Intelligence Affect The Risk of Nuclear War,” authors Edward Geist and Andrew J Lohn warns that as AI makes its way into weaponry the chances of nuclear warfare is highly likely. “The effect of AI on nuclear strategy depends as much or more on adversaries’ perceptions of its capabilities as on what it can actually do. For instance, it is extremely technically challenging for a state to develop the ability to locate and target all enemy nuclear-weapon launchers, but such an ability also yields an immense strategic advantage. States, therefore, covet this capability and might pursue it irrespective of technical difficulties and the potential to alarm rivals and increase the likelihood of conflict,” the researchers elaborate. They further point out that much of the challenge lies with the advent of superintelligence, “With superintelligence, AI would render the world unrecognisable and either save or destroy humanity in the process,” they note. The researchers then suggest a time-frame in the future where the AI system can influence or trigger nuclear destruction. By 2040, though AI’s role will be limited to that of a decision support system, the researchers say that it can influence humans in escalation matters. “Without being directly connected to the nuclear launchers, an AI could still provide advice to humans on matters of escalation. It seems reasonable that such a capability, at least for some aspects of the decision-making process, could be achieved by 2040 given the progress AI is making in increasingly complex and poorly specified task,” explain the authors. Echoing the same view, Dr Vincent Boulanin is a Senior Researcher at SIPRI in a blog post points out that AI’s penetration into modern weaponry could lead to a nuclear war like situation. Stating that adoption of AI into modern weapons by nuclear-armed states might encourage other states to consider the same and further destabilising measures imposed by nuclear-powers could increase the chances of a nuclear conflict ,”This could include entering into an arms race, doubling down on the modernisation of nuclear arsenals, renouncing a ‘no first use’ policy, increasing alert statuses, or further automating nuclear launch policies,” he writes. The only way out to avoid such a catastrophe is by encouraging meaningful dialogue between all the stakeholders, Boulanin points out, “A commitment to lower the alert status of nuclear arsenals, as well as more openness about nuclear modernisation plans and information-sharing via different dialogue tracks are measures that could clearly help to start mitigating the destabilising potential of nuclear-related AI applications.","excerpt":"As development in the field of artificial intelligence witnesses revolutionary changes, several noted names and visionaries have warned the world about the grave consequences of AI. In fact, the reliance of AI is being likened to something as lethal as nuclear weapons. Reiterating his position on the dangers of AI, founder and former CEO of […]","categories":["AI Features"],"tags":["bill gates","Elon Musk","OpenAI"],"author_name":"Akshaya Asokan","publish_date":"2019-04-09T10:07:20","publication_year":"2019","word_count":866,"keywords":["Go","artificial intelligence","machine learning","OpenAI","AI","RAG","Elon Musk","Ray","Aim","Rust","bill gates","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","OpenAI","Aim","Ray","RAG","R","Go","Rust"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/is-it-hyperbolic-to-compare-artificial-intelligence-to-nuclear-weapons\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10120375,"title":"AI Likely to Create More Jobs Than There are People","content":"Contrary to popular opinion, Groq CEO Jonathan Ross believes AI will create more jobs than we can handle. Ross said that the rapid technological advancement in AI could be another case of Jevons Paradox working its magic. “We keep thinking of each technology as displacing work. One of the things that’s probably going to happen is we will create more jobs for people than we have people,” Ross said, predicting that suddenly there may be a lack of people to do things. Citing the surge in the use of visual graphics in news articles, he said it has become easier than ever to do so as “most people spend more hours in generating graphics”, hinting at using AI more – and not less – expanding its uses far beyond what was originally intended. He said this, pointing at the Jevons Paradox, which was first reported in the 1860s by English economist William Jevons in his book The Coal Question. As steam engines became more efficient, people did not use less coal; instead, they used more of it. This increase in coal utilisation occurred because the more efficient steam engines lowered operational costs, enabling a widespread and intensive use of the engines. But is it the same? Ross believes that something similar is happening now. AI advancements are not only making it easier to perform tasks but also reducing costs, resulting in more people using them across sectors and creating more jobs in the process. “What will probably happen is with most of the things that generative AI makes easy, you will actually see an increase in human activity on that. There’s always going to be someone who’s going to be more entrepreneurial and figure out a way to monetise and get a whole bunch of people working on it,” Ross said. After all, the World Economic Forum predicted that AI would not only replace 85 million jobs by 2025, but also create 97 million jobs at the same time. In another report, the WEF listed the types of jobs that will be made obsolete, which included easily automatable jobs like bank tellers, data entry clerks and secretaries. Meanwhile, jobs for AI and machine learning specialists were some of the fastest-growing. Depends on What You Mean by Automatable While the study stated that jobs like clerks and secretaries were foremost at the risk of becoming obsolete, others held a different opinion. In a recent episode of the Ben and Marc Show podcast, Marc Andreessen and Ben Horowitz discussed how AI was more likely to take over the middle management at organisations. Thanks to the ease of training employees as well as a lack of interpersonal issues, the two reasoned that AI could take over managerial jobs rather than employee jobs. Likewise, OpenAI CEO Sam Altman conceded that the shift towards AI usage in the workplace is occurring differently than expected. “In many cases, this is something that will change the way people do their jobs in the same way that mobile phones, the internet, and computers did,” he said. Further, he said that the jobs that future generations would be doing would be different, but at the moment, the aim is to try and figure out how to adjust to the speed at which advancement is happening. What is the solution? “The world just needs a lot more code than we have people to write right now,” declared Altman, saying, “You hear a coder say I’m like 2 or 3 times more productive,” rather than the other way around. Similarly, as Ross stated, the generation of actual jobs will come from more people working in the field of AI and pioneering developments in the field. In doing so and creating their own companies, more people will be hired in the field, resulting in more jobs created. This is already being seen, according to Stanford’s AI Index Report 2024, where the number of newly funded AI startups increased by 40.6% from 2022. As of 2023, according to the report, as many as 1,812 startups had been newly funded. With this increase in startups, as well as developers aiming to upskill themselves, it’s likely that the end result will be lower-level workers following suit. This is also backed by the number of organisations, as mentioned in the Stanford report, increasingly making use of AI, regardless of whether the company itself is tech-inclined or not. While it might not be realistic to expect most people to reskill themselves in accordance with the changing job market, becoming skilled with AI may not be as hard as one would expect. Several universities have actually begun refining their AI and data science programmes. To the point that they have actually ranked on this year’s QS World Rankings. Similarly, both universities and companies have begun offering their own free courses online, most of which go into the foundations of AI, machine learning and data science. But while this could help, it doesn’t fully resolve the problem of potential job loss or how we could start skilling for the influx of potential jobs created because of AI. As Ross said, “The entire point of going into a different age and why you would call it a different technological age is it breaks all of our intuitions.”","excerpt":"“What will probably happen is with most of the things that generative AI makes easy, you will actually see an increase in human activity.”","categories":["AI Trends"],"tags":["Groq","Marc Andreessen","OpenAI","Sam Altman"],"author_name":"Donna Eva","publish_date":"2024-05-14T18:30:00","publication_year":"2024","word_count":877,"keywords":["data science","Go","Marc Andreessen","Sam Altman","machine learning","OpenAI","AI","API","GAN","Aim","generative AI","Groq","R"],"extracted_tech_keywords":["AI","machine learning","data science","generative AI","OpenAI","Aim","R","Go","API","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/ai-likely-to-create-more-jobs-than-there-are-people\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10009042,"title":"Top Twitter Accounts On AI One Must Follow","content":"At the present scenario, Artificial Intelligence and machine learning have been portraying a critical role in the advancement of the tech sector. Social media platforms have been performing a significant role when it comes to keeping updated with the latest and trending information. One such platform is Twitter. Twitter not only helps you keep track of the latest social and economic news, but also it allows you to both share and acquire knowledge about emerging technologies. Below here, we jotted a list down the top ten AI accounts, based on alphabetical order, one must follow on Twitter. Analytics India Magazine Apart from its conventional use in the #DevOps environment, #AI can prove to be beneficial in addressing #security issues and #data leaks, for organising memory management, and in garbage collection. https:\/\/t.co\/VwpraBtckj #cybesecurity #dataprivacy #artificialintelligence— AIM (@Analyticsindiam) October 6, 2020 This account on Twitter is handled by Analytics India Magazine. The followers include a youthful community of AI and ML enthusiasts. The account shares regular posts as well as updates on the latest news and trends that are shaping the field of AI, machine learning, data science and other such. Click here to follow. Andrew NG Our new Natural Language Processing Specialization’s first two courses are now on @Coursera! You’ll build your technical foundations and work on projects ranging from sentiment analysis to language modeling to autocomplete. Check it out: https:\/\/t.co\/RCcYfmRBsM pic.twitter.com\/rWf5nusLhp— Andrew Ng (@AndrewYNg) June 17, 2020 Andrew Ng is currently one of the most popular AI researchers around the globe. A former head of Baidu AI research group and Google Brain, Andrew is also the co-founder of the popular online learning platform, Coursera. He is one of the world’s most famous and influential computer scientists, being named one of Time magazine’s 100 Most Influential People in 2012, and Fast Company’s Most Creative People in 2014. Click here to follow. Andrej Karpathy Great source of reading pointers, as usual! ~75% of papers now use PyTorch, still positively trending. 1,000 companies are using Hugging Face's Transformers lib in prod, with 5M+ pip installs. https:\/\/t.co\/FbcNuXLIic— Andrej Karpathy (@karpathy) October 4, 2020 A senior director of AI at Tesla, Andrej Karpathy is the former research scientist at OpenAI and CS PhD student at Stanford University. His interests include training deep neural networks on large datasets. At Tesla, Karpathy leads the team responsible for all neural networks of the Autopilot. This includes dataset gathering, neural network training, the science of making it work, and their deployment in production running on the custom chip. Click here to follow. DeepLearn007 My Article on Smarter #AI & #DeepLearningSee: https:\/\/t.co\/Bazg1KKNT6@SpirosMargaris @Xbond49 @psb_dc @jblefevre60@nigewillson @Paula_Piccard @ipfconline1 @pierrepinna@Nicochan33 @HaroldSinnott @JimMarous #MachineLearning #DataScience #5G #IoT #Marketing #Healthcare #Tech pic.twitter.com\/xH0qNEX5ac— AI (@DeepLearn007) October 3, 2020 This twitter account is handled by Imtiaz Adam, who is the founder, director machine learning and digital strategy at Deep Learn Strategies Ltd. With more than 126k followers, this account shares various insights and strategies on AI and other technologies. It also shares blogs that contain information on AI, machine learning, 5G and more. Click here to follow. Dr Sally Eaves New #research #article live :) The 6 Ages of #technology with @drsharwood https:\/\/t.co\/3XBNZadMjr @elsevierbiz@EdinburghUni #DigitalTransformation #100DaysOfCode #Digital #AI #Robotics #Edge #Cloud #CyberSecurity #TrustInTech #futureofwork #SDGs #tech #TechNews #5G #Futures pic.twitter.com\/FOnNUleyMe— Prof. Sally Eaves (@sallyeaves) September 3, 2020 Prof Sally Eaves brings a depth of experience from his roles as a chief executive officer and chief technology officer. As a professor in advanced technologies and as a global strategic advisor, Eaves specialises in the application and integration of blockchain. Along with that, she has also shown her expertise in the fields of artificial intelligence and associated emergent technologies for business and societal benefit. Along with being an award-winning international keynote speaker, Eaves is also an author and influencer with globally leading rankings across all advanced technology disciplines. Click here to follow. Kirk Borne ★★★★This Google Tutorial (100+ slides total) on #MachineLearning is the best: https:\/\/t.co\/hsn9U4Yh9X by @jason_mayes ——————#abdsc #BigData #DataScience #DataMining #NeuralNetworks #AI #DeepLearning #TensorFlow #Mathematics #Algorithms #DataScientists #ReinforcementLearning pic.twitter.com\/90zvW9DID0— Kirk Borne (@KirkDBorne) October 5, 2020 Kirk Borne is a global speaker and the principal data scientist, data science fellow, and executive advisor at Booz Allen Hamilton. Borne is listed among one of the top AI and data science influencers in the world. Borne, a data scientist and an astrophysicist, who has worked 18 years on various NASA contracts as a research scientist, as a manager on a large science data system contract, and as the Hubble Telescope Data Archive Project Scientist. Click here to follow. Paula Piccard https:\/\/twitter.com\/Paula_Piccard\/status\/1313096522603679744 Paula Piccard is one of the leading digital marketer and social media consultants. Piccard helps small businesses to promote their services on social media networks and helps business owners integrate proven social media marketing strategies into their businesses to increase exposure and growth. Click here to follow. Ronald van Loon 8 Top Threats Against eCommerceby @perimeterx#Digital #DigitalTransformation #Innovation #Business #GlobalizationCc: @samighazali @nodexl @williamstim @jamesmarland @katharina_lamsa pic.twitter.com\/uClvOL0wHR— Ronald van Loon (@Ronald_vanLoon) September 30, 2020 Another top AI influencer of the current times is Ronald van Loon. Renowned around the globe, Ronald is the principal analyst and CEO of the Intelligent World. He is also the advisory board member and big data and analytics course advisor at Simplilearn. Click here to follow. Spiros Margaris Astonishing #AI #restoration brings #Apollo #moonlanding films up to speed https:\/\/t.co\/SDyxVB0d4D #fintech #ArtificialIntelligence #MachineLearning #DeepLearning @LiveScience @ipfconline1 @jblefevre60 @HaroldSinnott @Fisher85M @Paula_Piccard @UrsBolt @terence_mills— Spiros Margaris (@SpirosMargaris) October 5, 2020 Spiros Margaris, venture capitalist and senior advisor, is the founder of margaris ventures, and is the first international influencer to achieve “The Triple Crown” of influencer rankings (Onalytica). He was ranked the global No. 1 Fintech, Artificial Intelligence (AI) and Blockchain influencer by Onalytica, the world’s leading Influencer Relationship Management (IRM) SaaS platform recognized by Gartner, Forrester & industry influencers. Click here to follow. Yann LeCun Perlin noise in PyTorch from @honualx .[Ken Perlin is an NYU colleague] https:\/\/t.co\/qNyp0SCcaz— Yann LeCun (@ylecun) September 30, 2020 Yann LeCun is the chief AI scientist at Facebook AI Research and a renowned researcher in emerging technologies like AI, machine learning, computer vision, among others. LeCun is famous for his contribution to optical character recognition and computer vision using convolutional neural networks (CNN). Click here to follow.","excerpt":"At the present scenario, Artificial Intelligence and machine learning have been portraying a critical role in the advancement of the tech sector. Social media platforms have been performing a significant role when it comes to keeping updated with the latest and trending information.  One such platform is Twitter. Twitter not only helps you keep track […]","categories":["AI Trends"],"tags":["Twitter (X)"],"author_name":"Ambika Choudhury","publish_date":"2020-10-07T11:00:00","publication_year":"2020","word_count":1046,"keywords":["data science","artificial intelligence","machine learning","OpenAI","AI","neural network","ML","computer vision","Aim","analytics","Twitter (X)"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","neural network","computer vision","data science","analytics","OpenAI","Aim"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-twitter-accounts-on-ai-one-must-follow\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10123158,"title":"6 Open-Source LLMs That Can Run on Smartphones","content":"Large language models (LLMs) demand substantial computational resources, which are often limited to powerful servers. However, a new generation of compact models is making it possible to run these powerful language models directly on your smartphones. Interestingly, you won’t require the internet to utilise LLMs on your smartphones. Here are six open-source LLMs that can be trained and optimised to be used on your smartphones. Gemma 2B: Google’s compact, high-performance LLM for mobile language tasks. Phi-2: Microsoft’s tiny model outperforming giants up to 25 times its size. Falcon-RW-1B: Efficient 1B-parameter model for resource-constrained mobile devices. StableLM-3B: Stability AI’s balanced model for diverse language tasks on phones. TinyLlama: Compact Llama variant delivering impressive results on cell phones. LLaMA-2-7B: Meta’s powerful 7B model for advanced tasks on high-end smartphones. 1. Gemma 2B Google’s Gemma 2B is a compact language model that delivers impressive performance despite its small size. It utilises a multi-query attention mechanism, which helps reduce memory bandwidth requirements during inference. This is particularly advantageous for on-device scenarios where memory bandwidth is often limited. With just 2 billion parameters, Gemma 2B achieves strong results on academic benchmarks for language understanding, reasoning, and safety. It outperformed similarly sized open models on 11 out of 18 text-based tasks. 2. Phi-2 With 2.7 billion parameters, Phi-2 has been shown to outperform models up to 25 times larger on certain benchmarks. It excels in tasks involving common sense reasoning, language understanding, and logical reasoning. Phi-2 can be quantised to lower bit-widths like 4-bit or 3-bit precision, significantly reducing the model size to around 1.17-1.48 GB to run efficiently on mobile devices with limited memory and computational resources. One of the key strengths of Phi-2 is its ability to perform common sense reasoning. The model has been trained on a large corpus of web data, allowing it to understand and reason everyday concepts and relationships. 3. Falcon-RW-1B Falcon-RW-1B is part of the Falcon family of language models, known for their efficiency and performance. The RW stands for ‘Refined Web’, indicating a training dataset curated for quality over quantity. Falcon-RW-1B’s architecture is adapted from GPT-3 but incorporates techniques like ALiBi (Attention with Linear Biases) and FlashAttention to enhance computational efficiency. These optimisations make Falcon-RW-1B well-suited for on-device inference on resource-constrained devices like smartphones. The Falcon-RW-1B-Chat model aims to add conversational capabilities to the Falcon-RW-1B-Instruct-OpenOrca model to improve user engagement, expand use cases, and provide accessibility for resource-constrained environments like smartphones. 4. StableLM-3B StableLM-3B, developed by Stability AI, is a 3 billion parameter model that strikes a balance between performance and efficiency. The best part of StableLM-3B is that despite being trained on fewer tokens, it outperformed models trained on 7 billion parameters on some benchmarks. StableLM-3B can be quantised to lower bit-widths like 4-bit precision, significantly reducing the model size to around 3.6 GB to make it run efficiently on smartphones. A user mentioned that StableLM-3B has outperformed Stable’s own 7B StableLM-Base-Alpha-v2. 5. TinyLlama TinyLlama leverages optimisations like FlashAttention and RoPE positional embeddings to enhance computational efficiency while maintaining strong performance. It is compatible with the Llama architecture and can be integrated into existing Llama-based mobile apps with minimal changes. TinyLlama can be quantised to lower bit-widths like 4-bit or 5-bit precision, significantly reducing the model size to around 550-637 MB. A user, while sharing his experience with TinyLlama, mentioned that on a mid-range phone like the Asus ROG, TinyLlama was generating 6-7 tokens per second. Ladies and gentlemen we have tinyllama running locally on mobile ( my dad's cause I am having a broke af low powered phone ) using termux. It aint even hard to do. Here a few pics of me playing with it and me documenting the steps on a notepad pic.twitter.com\/lIsYiBiHh9— Govind-S-B (@violetto96) December 23, 2023 6. LLaMA-2-7B The LLaMA-2-7B model has been quantised to 4-bit weights and 16-bit activations, making it suitable for on-device deployment on smartphones. This quantisation reduces the model size to 3.6GB, making it feasible to load and run on mobile devices with sufficient RAM. LLaMA-2-7B model on mobile requires a device with at least 6GB of RAM. During inference, the peak memory usage ranges from 316MB to 4785MB on the Samsung Galaxy S23 Ultra. This suggests that while the model can run on devices with 6GB+ RAM, having more RAM allows for better performance and reduces the risk of out-of-memory errors. While it requires devices with sufficient RAM and may not match the speed of cloud-based models, it offers an attractive option for developers looking to create intelligent language-based features that run directly on smartphones. Running llama-2-7b on Replit on my phone thanks to expandable storage. It’s kinda decent on CPU with 8 toks\/s. https:\/\/t.co\/r9yqvWayL5 pic.twitter.com\/XeijnENM9E— Amjad Masad (@amasad) July 21, 2023","excerpt":"Maximise privacy and control by leveraging the power of LLMs on your smartphone without using the internet.","categories":["AI Trends"],"tags":["AI open source","Open Source"],"author_name":"Sagar Sharma","publish_date":"2024-06-11T11:09:09","publication_year":"2024","word_count":785,"keywords":["Go","attention mechanism","Open Source","programming_languages:R","AI","AI open source","RAG","GPT","Aim","llm_models:Llama","R","llm_models:GPT"],"extracted_tech_keywords":["AI","Aim","RAG","R","Go","attention mechanism","GPT","llm_models:GPT","llm_models:Llama","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/6-open-source-llms-that-can-run-on-smartphones\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":63855,"title":"HCL To Help Tamil Nadu Govt To Set Up Data Analytics Center To Fight COVID-19 Pandemic","content":"Many global companies are making an effort to ease up the wrath that COVID-19 has brought to the world. In a recent development, the government of Tamil Nadu has partnered with HCL to set up a Data Analytics Center to strengthen the state’s disaster management efforts. Tamil Nadu’s Disaster Management Center is responsible for the overall management of disasters across the entire state. With over 1000 coronavirus cases, the state is one of the worst affected in the country and the government is taking all the necessary measures to bring down the numbers. By roping up HCL Technologies, the government aims to set up a Disaster Management – Data Analytics Center to fight off the COVID 19 pandemic. What will the data analytics centre facilitate? HCL will provide support to the Disaster Management – Data Analytics Center, to capture data trends from across all districts in the state in real-time. This data trend will be displayed live to help the government make future decisions and decide the degree of response needed for each district. It will also help in deciding the graded relaxation of the current lockdown to resume economic activities based on the number of cases in a district. It will capture data trends from the citizen queries which in turn will be fed into the Emergency Operations Center to guide the future course of action. As per reports, apart from working on the data analytics centre, the company will also help in expanding the state’s disaster management helpline (1070) through technological upgradation, manpower assistance and effective reporting mechanisms. The Center operates a call centre which is accessible to people across the state and delivers the first level of response in an emergency. Apart from setting up data analytics centre, HCL will provide value in terms of: Upgrading their existing call centre by implementing an Automatic Call Distribution (ACD) system to effectively route calls between different government departmentsBuilding a state-of-art case flow system to track caller identity and provide a follow-up from the relevant departmentsProviding manpower assistance to respond to calls and address queries on food shortage and availability, support to the migrant population, differently-abled and elderly, respond to queries related to health and more","excerpt":"Many global companies are making an effort to ease up the wrath that COVID-19 has brought to the world. In a recent development, the government of Tamil Nadu has partnered with HCL to set up a Data Analytics Center to strengthen the state’s disaster management efforts.  Tamil Nadu’s Disaster Management Center is responsible for the […]","categories":["Deep Tech"],"tags":["Data Analytics"],"author_name":"Srishti Deoras","publish_date":"2020-04-28T18:36:25","publication_year":"2020","word_count":366,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","Aim","ViT","analytics","Data Analytics","R"],"extracted_tech_keywords":["AI","analytics","Aim","R","Go","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/hcl-to-help-tamil-nadu-govt-to-set-up-data-analytics-center-to-fight-covid-19-pandemic\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":25419,"title":"Why Is Random Search Better Than Grid Search For Machine Learning","content":"Optimising hyperparameters is considered to be the trickiest part of building machine learning and artificial intelligence models. It is nearly impossible to predict the optimal parameters while building a model, at least in the first few attempts. That is why, we always go by playing with the hyperparameter to optimise them. However, this is not scalable for high dimensional data because the number of the increase in iterations, which in turn expands the training time. When it comes to tuning hyperparameters while developing a neural network, mastering the optimisation technique is a big pain. As this is a big part of deep learning, if these parameters are not set to optimal, the training might take ages to complete, or the model may never even reach the local minima. There are various training methods introduced into machine learning to find these optimal parameters, so let us look at two of the widely used and easy to implement techniques. Hyperparameter Tuning Methods Hyperparameter tuning refers to the shaping of the model architecture from the available space. This, in simple words, is nothing but searching for the right hyperparameter to find high precision and accuracy. There are several parameter tuning techniques, but in this article, we shall look into two of the most widely-used parameter optimiser techniques Grid search Random search Grid Search Grid search is a technique which tends to find the right set of hyperparameters for the particular model. Hyperparameters are not the model parameters and it is not possible to find the best set from the training data. Model parameters are learned during training when we optimise a loss function using something like a gradient descent. In this tuning technique, we simply build a model for every combination of various hyperparameters and evaluate each model. The model which gives the highest accuracy wins. The pattern followed here is similar to the grid, where all the values are placed in the form of a matrix. Each set of parameters is taken into consideration and the accuracy is noted. Once all the combinations are evaluated, the model with the set of parameters which give the top accuracy is considered to be the best. Below is a visual description of uniform search pattern of the grid search. Performing grid search over the hyperparameter space with support vector machine With the above parameters, the SVM would yield One of the drawbacks of grid search is that when it comes to dimensionality, it suffers when evaluating the number of hyperparameters grows exponentially. However, there is no guarantee that the search will produce the perfect solution, as it usually finds one by aliasing around the right set. Random Search Random search is a technique where random combinations of the hyperparameters are used to find the best solution for the built model. It is similar to grid search, and yet it has proven to yield better results comparatively. The drawback of random search is that it yields high variance during computing. Since the selection of parameters is completely random; and since no intelligence is used to sample these combinations, luck plays its part. Below is a visual description of search pattern of the random search: As random values are selected at each instance, it is highly likely that the whole of action space has been reached because of the randomness, which takes a huge amount of time to cover every aspect of the combination during grid search. This works best under the assumption that not all hyperparameters are equally important. In this search pattern, random combinations of parameters are considered in every iteration. The chances of finding the optimal parameter are comparatively higher in random search because of the random search pattern where the model might end up being trained on the optimised parameters without any aliasing. Let us say we have 3×3 set of parameters. There are about 9 set parameters in each search technique. From the example above, random search works best for lower dimensional data since the time taken to find the right set is less with less number of iterations. In grid search, however, the optimal parameter is not found since we do not have it in our grid, and that’s why time is spent to find the near best solution is until it reaches the last set sample. With this example, it is clear that random search is the best parameter search technique when there are less number of dimensions. While providing the information for the search parameters, for example, the properties of the cost function like continuous or discrete and type or property of error correction like stochastic, all these parameters and the heuristics provided will help the model to converge at the minima faster. The bottom rule of finding the highest accuracy is that more the information you provide faster it finds the optimised parameters. Conclusion There are other optimisation techniques which might yield better results compared to these two, depending on the model and the data. When it comes to science, there is no luck but when it comes to randomness we hope to find the best sample with the least possible time.","excerpt":"Optimising hyperparameters is considered to be the trickiest part of building machine learning and artificial intelligence models. It is nearly impossible to predict the optimal parameters while building a model, at least in the first few attempts. That is why, we always go by playing with the hyperparameter to optimise them. However, this is not […]","categories":["AI Features"],"tags":["AI (Artificial Intelligence)","datascience","Hyperparameter","Python"],"author_name":"Kishan Maladkar","publish_date":"2018-06-14T10:44:32","publication_year":"2018","word_count":854,"keywords":["Go","API","artificial intelligence","machine learning","AI","neural network","datascience","RPA","Scala","Python","deep learning","Hyperparameter","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","deep learning","neural network","R","Go","Scala","API","RPA"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/why-is-random-search-better-than-grid-search-for-machine-learning\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10135003,"title":"After Software Engineers, LLMs Are Coming After AI Researchers","content":"Stanford University recently published a paper titled ‘Can LLMs Generate Novel Research Ideas?’ The study found that ideas generated by LLMs (large language models) were rated significantly more novel than those from human experts. To reach this conclusion, over 100 NLP researchers were asked to come up with new ideas and review both LLM-generated and human-generated ideas without getting to know their source. The results showed that LLM ideas were considered more innovative (with statistical significance, p < 0.05), although they were rated slightly lower in terms of feasibility. The approach is similar to Japanese AI startup Sakana AI’s AI Scientist, which automates the entire research lifecycle. It generates novel research ideas, writes necessary code, executes experiments, summarises results, visualises data, and presents its findings in a complete scientific manuscript. Interestingly, the startup claimed that each idea is implemented and developed into a full paper at approximately $15 per paper. Generating new ideas is relatively easy for LLMs, thanks to their extensive training on large datasets and ability to combine various concepts. However, they continue to face challenges with advanced reasoning. Meanwhile, OpenAI is preparing to release its new model, Strawberry, which is expected to offer improved reasoning capabilities. Chai Discovery, a biology startup founded by a former OpenAI employee, recently introduced Chai-1, an advanced foundation model that predicts molecular structures crucial for drug discovery. Innovations like these show that LLMs are close to driving significant research breakthroughs. “The ability of LLMs to combine concepts from vast datasets in ways not typically thought of by humans can lead to ideas that are considered more novel. This might be because LLMs aren’t constrained by the same cognitive biases or conventional thinking patterns that humans have,” said DigitalVibes.ai founder Anthony Scaffeo. He added that LLMs can make connections across different fields or unrelated data points, which might not be intuitive or immediately obvious to human experts. “My student’s comment on the paper about LLMs generating more novel research ideas than humans is making the rounds. I think this says more about NLP researchers than about LLMs. Ouch,” joked Subbarao Kambhampati, professor of computer science at Arizona State University. “I am not gonna let no LLM beat me in generating novel NLP research ideas!,” he quipped. Interestingly, Kambhampati has been quite vocal about LLMs being bad at reasoning and planning. He said that models like GPT-3, GPT-3.5, and GPT-4 are poor at planning and reasoning, which he believes involves time and action. According to him, these models struggle with transitive and deductive closure, with the latter involving the more complex task of deducing new facts from the existing ones. Can They Really Do Research? Today several researchers have not experimented much with LLMs to generate new novel ideas, instead they have been predominantly using it to review the research papers. Apparently, Meta AI chief Yann LeCun argues that while LLMs cannot reason and plan, they are still a good tool for reviewing papers. “Reviewers (as in human reviewers) should be able to use the tools they want to help them write reviews. The quality of their reviews should be assessed based on the result, not the process,” he said. Meta AI launched Galactica, an LLM for research, in November 2022, just weeks before ChatGPT. However, it was taken down after three days due to criticism over generating misleading or offensive information. LeCun remains unhappy about it to this day. However, not everyone agrees with LeCun. “AI-generated reviews of scientific papers are increasing, vacuous, and need to be stopped quickly. They reduce the author’s trust in the review process. Proposal: someone who is judged to have submitted such a review is banned from submitting to the same conference\/journal for two years,” said Micheal Black, director, Max Planck Institute for Intelligent Systems. Adding to this perspective, Mukur Gupta, an applied scientist at Apple, recounts his frustrating experience with an LLM-generated review. “I love AI as an assistant. But after getting an LLM-generated review for my NeurIPS paper last month (which was total crap and useless), I’m a little skeptical about AI discovering true novelty,” said Gupta. He explained that LLMs could be a game-changer for interdisciplinary research or for uncovering new problems in fields—where human experts, limited by their working memory and attention span, may struggle to grasp more than a handful of domains. “LLMs, with their ever-expanding knowledge base, offer the potential for cross-pollination of ideas.” “But when it comes to deep, niche, and fundamental breakthroughs, I’m not buying it—hence my disappointment with that NeurIPS review,” he added. Use AI to Brainstorm, Not Write Lately, there has been a growing trend of researchers using LLMs to write papers. According to recent data, the use of the term ‘delve’ in the abstracts gradually increased through 2022, jumped noticeably in 2023 (when ChatGPT became widely available), and has continued to rise in 2024. The future of research should be a collaboration between humans and LLMs to generate truly innovative ideas. According to Stanford’s paper, human ideas often prioritise feasibility and effectiveness over novelty and excitement, which can limit their creativity. On the other hand, LLMs struggle to judge the quality of ideas. By combining the strengths of both humans and LLMs, we can pave the way for exciting research.","excerpt":"“LLMs, with their ever-expanding knowledge base, offer the potential for cross-pollination of ideas.”","categories":["AI Features"],"tags":["AI Research","Editors Picks","LLMs"],"author_name":"Siddharth Jindal","publish_date":"2024-09-11T17:31:10","publication_year":"2024","word_count":873,"keywords":["Go","ChatGPT","Meta AI","OpenAI","AI","LLMs","AI Research","Git","NLP","Editors Picks","Aim","Rust","R"],"extracted_tech_keywords":["AI","NLP","ChatGPT","OpenAI","Meta AI","Aim","R","Go","Rust","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/after-software-engineers-llms-are-coming-after-ai-researchers\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10013229,"title":"Customer Insights That Can Be Unearthed From Retail Transactions","content":"What insights could be unearthed from a retail transaction data? This would be of interest to any retailer, marketer amongst others. Some of the questions that a retailer would be interested in would be — Who are our best customers? Which of our customers are likely to churn? Who has the potential to become a valuable customer? Which of our customers can be retained? Which of our customers are most likely to respond to engagement\/push campaigns? Smart marketers understand the importance of “know thy customer.” Instead of simply focusing on generating more clicks, marketers must follow the paradigm shift from increased CTRs (Click-Through Rates) to retention, loyalty, and building customer relationships. If a retailer or a marketer has a large amount of customer transaction data would it be wise to start push campaigns on the entire set of customers or would it be more appropriate to extract meaningful customer information first before launching a push campaign. Instead of analysing the entire customer base as a whole, it’s better to segment them into homogeneous groups, understand the traits of each group, and engage them with relevant campaigns rather than segmenting on just customer age or geography. One of the popular, easy-to-use, and effective segmentation methods to enable marketers to analyse customer behaviour is the RFM analysis. RFM (Recency, Frequency, Monetary) analysis is a marketing technique used to determine quantitatively which customers are the best ones by examining how recently a customer has purchased (recency), how often they purchase (frequency), and how much the customer spends (monetary). It is an approach that can separate good customers from average customers and inactive ones based on transactional data. Each model is first optimized based on correlations in the data, including the selection of input variables. Scores could be assigned on a 5-point scale with 5 indicating the most desired behaviour from customers on any of the three dimensions – R, F, and M. RFM approach is likely to vary depending on the type of business. In a consumer durables business, the monetary value per transaction is normally high but frequency and recency are low. For example, you can’t expect a customer to purchase a refrigerator or air conditioner on a monthly basis. In this case, a marketer could give more weight to monetary and recency aspects rather than the frequency aspect. In a retail business selling fashion\/cosmetics, a customer who searches and purchases products every month will have a higher recency and frequency score than the monetary score. Accordingly, the RFM score could be calculated by giving more weight to R and F scores than M. For content apps like Hotstar or Netflix, a binge watcher will have a longer session length than a mainstream consumer watching at regular intervals. For bingers, engagement and frequency could be given more importance than recency, and for mainstreamers, recency and frequency can be given higher weights than engagement. Clustering is another technique that could be used to develop customer segmentation. Clustering also arrives at homogenous groups of customers with each group exhibiting certain purchase behaviour. The question that arises is that should we cluster first and then use RFM or should we do RFM directly? Clustering by itself is without any direction since it is an unsupervised learning approach. It could be argued that doing an RFM approach first provides a direction to a subsequent clustering of customers. We adopted such an approach to retail transaction data. RFM scores were generated for each customer id in the dataset based on each of the parameters – R, F and M. The dataset was then clustered using k-means clustering along with recency of purchase, frequency of purchase, and amount purchased. The combined effect of these resulted in some interesting insights on the customers. We found that of the five clusters that were generated, one of the clusters had customers who were frequently purchasing however they seem to purchase in relatively smaller amounts. These customers keep the transactions rolling but may not help in scaling up of our business. Another cluster had customers who had started purchasing recently and in large amounts. These customers should be monitored for future campaigns. A third type of cluster had customers who were buying in large amounts and frequently but had not purchased recently. It would pay to keep to these customers engaged with various offerings and campaigns. A fourth cluster had customers who were buying frequently but not in large amounts. These customers could also be part of push campaigns with offerings. The last cluster had customers who were buying in sizeable amounts but not very frequently. These could be customers who might be interested only in a part of our offerings and hence we could survey these to figure out what might be there total basket of interested goods. In this way, any transaction data could be explored to arrive at customer insights that would be part of any monetization strategy.","excerpt":"What insights could be unearthed from a retail transaction data? This would be of interest to any retailer, marketer amongst others. Some of the questions that a retailer would be interested in would be — Who are our best customers? Which of our customers are likely to churn? Who has the potential to become a […]","categories":["AI Features"],"tags":["outlier analysis in data mining"],"author_name":"G R Chandrashekhar","publish_date":"2020-12-05T11:00:00","publication_year":"2020","word_count":819,"keywords":["Go","outlier analysis in data mining","AI","programming_languages:R","programming_languages:Go","RAG","R"],"extracted_tech_keywords":["AI","RAG","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/customer-insights-that-can-be-unearthed-from-retail-transactions\/","complexity_score":2,"technical_depth":6,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":53470,"title":"MLDS 2020: Top 10 Talks You Should Definitely Attend This Year","content":"The second edition of the Machine Learning Developers Summit (MLDS 2020) has already started to create a significant excitement and buzz in the industry. The event, which is to be held on 22-23 Jan in Bengaluru and on 30-31 Jan in Hyderabad, is set to be one of the biggest gatherings of developers in India. With over 1,500 attendees expected to attend the event, MLDS will see innovators from leading tech companies gathering together to discuss the latest software architecture of ML systems, producing and deploying the newest ML frameworks, addressing the challenges faced and more. In this article, we are listing down ten such discussions and talks that you can look forward to attending at MLDS 2020. These talks are a combination of keynotes, panel discussions, tech talks and knowledge talks: 1. Ecosystem Intelligence, The Next Frontier In AI When: 31 Jan, 11:30 am, Hyderabad By: Shailesh Kumar, Chief Data Scientist, CoE AI\/ML at Jio Faculty for Advanced Management Programme in Business Analytics at ISB Like any other technology, AI has been growing in a bottom-up manner. Shailesh Kumar says that we have mastered the art of building accurate AI models, deploying them at scale, and continuously improving them as more data comes. We have done that for a variety of AI API’s now. He says that humans are now ready to build the next-generation products that can utilize these AI API’s like an ecosystem and deliver products that will be vastly different from what we think of products today. In this talk, Shailesh Kumar will explore what such products look like and how our AI and product thinking have to evolve to build such products. 2. Humans Not Machines Will Build The Future When: 22 Jan, 12.10 pm, Bangalore By: Vivek Kumar, Managing Director at Springboard India Vivek Kumar says that since we live in the age of machine learning algorithms setting to help humankind with Siri and Alexa, made a quantum leap. In such a world, do you need humanness? This talk will touch upon humanness in the age of AI-ML and how it impacts careers in these future technologies. Springboard is a company built on humanness, and this talk will throw light on how to become a better AI engineer while also helping the community of aspirants. 3. Deep Learning With TensorFlow When: 23 Jan, 12.10 pm, Bangalore By: Mohan Kumar Silaparasetty, Co-Founder, CEO at Trendwise Analytics 4. Building An AI-First Organization When: 22 Jan, 2.40 pm, Bangalore | 30 Jan, 10 am, Hyderabad By: Sundara Ramalingam N Head of Deep Learning practice at NVIDIA India This talk will focus on how to create a successful recipe for building AI into the existing traditional workflows of an organization. Machine learning, deep learning and data sciences have evolved extremely fast in the recent past, and practitioners find it challenging to keep in pace with the technology. The talk will cover the latest advancements in AI technology from both infrastructure and software perspective and recommends the best practices learnt from multiple domains for setting up a successful AI practice. 5. Deep Learning Using Convolution Neural Network For Classification Of Images When: 30 Jan, 11:30 am, Hyderabad By: Mohammad Shaheer Zaman Senior software engineer at AMD Machine learning has found its application in various practical domains. In this presentation, Mohammed will look at how deep learning can be used to classify images. Specifically, he will look at convolution neural networks which are a subset of deep learning models to solve a classic problem of computer vision which is to differentiate between two sets of images. Finally, he will compare the performance of machines as opposed to those of humans in recognizing images. 6. Anthropomorphism in Conversational User Interfaces When: 23 Jan, 4:10 pm, Bangalore By: Mathangi Sri Head of Data Science at PhonePe Machines may have “artificial” intelligence, but they want you to believe that they are humans. Can they behave like humans? Does it make sense, or is it better to go all out and declare the truth to the users?. If we want them to take human characteristics, what can be done?. What is the state-of-the-art in anthropomorphic systems, and how can machine learning and NLP help? In this talk, Mathangi attempts to break into the surface of this problem. 7. Advances in Deep Recommender Systems & Their Impact on Top Lines When: 22 Jan, 4:50 pm, Bangalore By: Bhavik Gandhi Director, Data Sciences and Analytics at Shaadi.com 8. Are We Ready For AI DevOps? When: 23 Jan, 1:50 pm, Bangalore By: Sunil Kumar Vuppala, Director – Data Science at Ericsson Global AI Accelerator This talk discusses the basic W & H questions (what, why, where, when and how) of AI DevOps and the challenges in adoption of deploying the ML\/DL models. This covers a couple of reference frameworks in this journey such as MLFlow, SageMaker and Azure ML service. Sunil also includes a couple of use cases from the Telecom industry to illustrate the need for faster diagnosis and edge deployment. He highlights various trade-offs one needs to handle in their ML life cycle, including the PLASTER framework. The key takeaways for the audience will include essential factors to be considered while designing and deploying the DL based solutions, take steps towards AI DevOps and corresponding research areas to work on. 9. Boosting Memory-Based Collaborative Filtering Using Content-Metadata When: 23 Jan, 4.50 pm, Bangalore By: Anish Agarwal Director – Data & Analytics at RBS India Recommendation systems are widely used in conjunction with many popular personalized services, which enables people to find not only content items they are currently interested in, but also those in which they might become involved in future. Many recommendation systems employ the memory-based collaborative filtering (CF) method, which has been generally accepted as one of the consensus approaches. Despite the usefulness of the CF method for a strong recommendation, several limitations remain, such as sparsity and cold-start problems that degrade the performance of CF systems in practice. In this talk, Anish will talk about how to overcome these limitations, a suitable content-metadata-based approach that effectively uses content-metadata. 10. AI in Manufacturing Intelligence When: 31 Jan, 09:15 am By: Venugopal Jarugumalli Principal Solutions Architect at ZF Group Artificial intelligence technology is now making its way into manufacturing, and the machine-learning technology and pattern-recognition software at its core could hold the key to transforming factories of the near future. AI will perform manufacturing, quality control, shorten design time, and reduce materials waste, improve production reuse, perform predictive maintenance, and more.","excerpt":"The second edition of the Machine Learning Developers Summit (MLDS 2020) has already started to create a significant excitement and buzz in the industry. The event, which is to be held on 22-23 Jan in Bengaluru and on 30-31 Jan in Hyderabad, is set to be one of the biggest gatherings of developers in India. […]","categories":["Deep Tech"],"tags":["TED Talks"],"author_name":"Prajakta Hebbar","publish_date":"2020-01-09T19:00:00","publication_year":"2020","word_count":1087,"keywords":["data science","machine learning","artificial intelligence","AI","neural network","ML","TED Talks","computer vision","NLP","deep learning","analytics"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","neural network","NLP","computer vision","data science","analytics"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/mlds-2020-top-10-talks-you-should-definitely-attend-this-year\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":26860,"title":"4 Types Of Startup Incubators Every Founder Should Know","content":"Image Source: T Hub Startup incubators play a key role in nurturing and scaling ideas. The primary task of an incubator is to drive entrepreneurial spirit and be a mainstay of economic development. The Indian Government has put technological growth and advancement in emergent technologies at the heart of its agenda. High-growth and innovative businesses are a part of their economic agenda. In this article, we list down key incubators that are playing a pivotal role in business development, building startup ecosystem and sustainable value creation. This is not an exhaustive list of incubators across India, but slots top programs across the four main categories Government University-backed Corporate Private. Government-Supported Incubators In India Startup India: This is Government of India’s most famous startup initiative. Startup India promotes tech entrepreneurship through mentoring and funding. Since its launch in January 2016, the initiative has successfully incubated several startups. The programme also provides a comprehensive four-week free online learning program with dedicated research parks, incubators and startup centres across the country. A key feature of the initiative is the ‘Fund of Funds’ to help startups gain access to funding. At the core of the initiative is the effort to build an ecosystem in which startups can scale effectively with facilities such as online recognition of startups, Startup India Learning Programme, facilitated patent filing, easy compliance norms, incubator support and innovation focused programmes for students, funding support, tax benefits and addressing of regulatory issues. T-Hub: The country’s major incubator, T-Hub by Telangana government has incubated 346 startups so far, a report indicates. Pegged as a startup engine, T-Hub has played a crucial role in bringing entrepreneurs, venture capitalists and mentors onto a single platform. T-Hub has catalyzed Hyderabad into a startup capital. Academic NSRCEL: This is one of the most well-known and oldest business incubators in India that is open to all entrepreneurs , across sectors, domains, and industries. This incubator provides an ecosystem for development activities and also empowers entrepreneurs through academic research. CIIE: One of India’s leading innovation centre, IIM Ahmedabad’s Centre for Innovation Incubation and Entrepreneurship (CIIE) CIIE provides significant seed-fund and incubation support to aspiring entrepreneurs in IT, cleantech, healthcare and education space. In its effort to establish a stronger entrepreneurial ecosystem in India, CIIE has conceptualised and partnered with several initiatives like INFUSE, Power of Ideas, iAccelerator, Stay Hungry Stay Foolish, MentorEdge among other. Centre for Innovation and Entrepreneurship (CIE): Another big tech incubator from IIIT Hyderabad, CIE has incubated 250 startups in the last few years and is now focusing on deep tech startups to strengthen the deep tech startup ecosystem in India. According to a recent news report, CIE is now looking to kick-start domain specific accelerators that focus on emergent technologies like machine learning, AI and VLSI (Very Large Scale Integration). Private Khosla Labs: Vinod Khosla and Srikanth Nadhamuni founded Bengaluru-based Khosla Labs is a startup incubator that focuses on startups working on financial inclusion and retail solutions. The incubator’s motto is “Ideate, Prototype, Validate and Launch”. The projects get prototyped through their in-house design and pilot tested. Some of the startups incubated have been in the area of mobile banking, machine learning and retail. Novopay that provides retail and banking solutions is one such startup incubated at Khosla Labs. Wadhwani Centre for Entrepreneurship Development: Set up in 2008, this initiative aims at bolstering the tech entrepreneurship in India and is headquartered in ISB Hyderabad. Some of the key startups incubated at WCED are Richcore Lifesciences and Orkash. Corporate 10,000 Startups: Hosted by NASSCOM, this initiative aims to incubate, fund and support 10,000 technology startups in India. Google for Entrepreneurs is also a founding partner of this initiative and provides key mentorship, Google training resources on how to scale startups globally. NASSCOM floated the now famous FAME model to scale startups — Funding, Acceleration, Mentoring, and Enterprise that gives startups the opportunity to network with key stakeholders. SAP Labs India: In 2016, SAP Labs India launched its Startup Accelerator Program to drive entrepreneurship and tech development in India. Since its inception, SAP Labs India has already hosted startups like Niki.ai, logistics solutions provider Blubirch, Ecolibrium Energy, a provider of energy management solutions among others. In addition to this accelerator, SAP also has a 75 seat incubation lab, Startup Studio at the SAP Labs campus in Bengaluru, that nurtures early\/disruptive stage tech big data and IoT startups. The startups are incubated for one year and are backed by a robust mentorship program that extends beyond their first year.","excerpt":"Startup incubators play a key role in nurturing and scaling ideas. The primary task of an incubator is to drive entrepreneurial spirit and be a mainstay of economic development. The Indian Government has put technological growth and advancement in emergent technologies at the heart of its agenda. High-growth and innovative businesses are a part of […]","categories":["AI Trends"],"tags":[],"author_name":"Richa Bhatia","publish_date":"2018-08-01T09:53:33","publication_year":"2018","word_count":751,"keywords":["big data","Go","API","machine learning","AI","innovation","Aim","ViT","GAN","R"],"extracted_tech_keywords":["AI","machine learning","Aim","R","Go","API","big data","GAN","ViT","innovation"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/4-types-of-startup-incubators-every-founder-should-know\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10072765,"title":"Data Warehouses in Age of Decentralised Architecture","content":"In 2019, Transforming Data with Intelligence (TWDI), a data and analytics research company, came up with a report, ‘The Modernisation of the Data Warehouse’, detailing why there is a need to modernise data warehouses and strategies to modernise data warehouses. Incidentally, this was the same year the data mesh concept started gaining ground after Zhamak Dehghani, director of Emerging Technologies at ThoughtWorks, published a paper, ‘How to Move Beyond a Monolithic Data Lake to a Distributed Data Mesh’. Now, unlike data warehouses that are centralised data architectures, data mesh is a highly decentralised distributed data architecture. Dehghani conceived data mesh from the challenges inflicting the centralised approach. Firstly, most of the data is always generated by operational systems and applications like CRMs or ERPs. There are dedicated people for these domain-specific applications. The operational application team and the data platform team currently have little or no communication. Often, the former is not at all aware that “their” data is being sent to a data platform. Second, the data team itself lacks a solid understanding of how business applications operate. Thus, none of the groups is able to take ownership of the data, often leading to situations where data quality issues remain unaddressed or are addressed very late in the process. Moreover, the data engineering team working on the centralised platform struggles to produce usable data for consumption as they don’t understand how the data has been generated and what it actually means, given there is no context to the data — just tables with names, rows, columns, or files. Does that mean with the introduction of something like data mesh, older data architecture becomes obsolete? Well, not really. There are several drivers— technological, strategic as well as business — for data warehouse modernisation. The TWDI report, too, cited several factors driving modernisation of data warehouses as shown below. (Source: TWDI Pulse Report) Modernising data warehouses is advantageous Today, there are a lot of companies engaged in the business of data warehouse modernization like Google Cloud, Qlik, Astera, OneGLobe and others. There are several benefits of data warehouse modernization, for instance, it brings agility in the business, which is key in scaling up businesses and taking data-driven decisions. Modernising data warehouses facilitates easier integration of data in the centralised repository and streamlines the process of deriving insights from it. Modernisation of data warehouses does away with the need to hire additional resources to query data, analyse it and generate reports. It can help businesses save costs associated with identifying data that is lost during ETL process or poor quality data due to want of validation while mapping data from source to warehouses. Added to these, is the high speed and better performance that modernised data warehouses provide. They enable data to be accessed without compromising their quality and efficiency. They also support larger volumes of data that can give any organisation a holistic picture regarding where it stands. Drawing inferences from recent developments Notwithstanding the benefits offered by data warehouse modernisation, a question that arises here is that in the age of decentralised data architectures, are data warehouses and their modernisation still relevant? Well, discussions and debates could go on. However, the fact that data warehouses are still relevant can be inferred for several developments. The future of data mesh that was conceived as an alternative to centralised data architectures is very uncertain. Gartner’s Hype Cycle for Data Management 2022 expects data mesh to become ‘obsolete before plateau’. Many industry experts are of the view that it is more of a hype doomed to be obsolete. According to James Serra, big data and data warehousing solution architect at Microsoft,  a decentralised solution like data mesh is not for smaller companies. It is more suited for big companies having complex data models, high data volumes, and multiple data domains. Thus, a decentralised solution would be ‘an overkill’ for at least 90% of companies. Cloud-based data warehouses like Amazon’s Redshift or Google’s Big Query offer an added advantage – organisations do not need to incur upfront costs, but pay as their data volume expands.   In fact, cloud-based platforms and tools have become one of the several strategies for modernisation of data warehouses. According to Sigmoid, a data engineering and AI-solutions company,  adoption of cloud data warehouses is growing at a CAGR of 15%. Cloud-based data warehouses make it easier to scale without having to compromise on performance. Companies can deploy the resources they actually need and scale capacity up or down as business needs change.","excerpt":"According to Sigmoid, a data engineering and AI solutions company, adoption of cloud data warehouses is growing at a CAGR of 15%.","categories":["AI Features"],"tags":[],"author_name":"Zinnia Banerjee","publish_date":"2022-08-15T16:00:00","publication_year":"2022","word_count":752,"keywords":["big data","Go","AI","ETL","ML","data engineering","analytics","R","data lake","data warehouse"],"extracted_tech_keywords":["AI","ML","analytics","R","Go","big data","data engineering","ETL","data warehouse","data lake"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/data-warehouses-in-age-of-decentralised-architecture\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":34734,"title":"Understanding Data Gravity: How Cloud Is Helping It Grow Stronger","content":"Of late, there is one more term that has gained a lot of traction — Data Gravity. According to Dave McCrory who coined data gravity in 2010, it can be defined as the analogy between the nature of data and the ability of data to attract additional applications and services. The concept is based on Newtonian gravity: the larger a mass, the more the attraction. And if we consider data as a planet or something with significant mass, then as data continues to build mass, it will start attracting services and applications towards it. Also, with the increase in mass of data, the strength of gravitational pull also increases, making application and services accelerate fast towards data. Location plays a vital role when it is about data as it affects the latency and throughput, which drives data gravity. The further the data, the longer it takes for the various pieces of your organisation’s infrastructure to communicate with each other. For example, if a company has its data centre in a location called A (which is pretty far), then the company has to deal with waiting for data to flow back and forth. However, the scenario will change significantly if the data centre is a location called B (which is closer); the communication can happen faster. This scenario is Latency. Talking about throughput, in a nutshell, it will take much more time to transfer data from one location to another if the data centre is far away. Therefore, it absolutely makes sense for data to be gathered together and for associated applications and services to be located nearby. With Time, Data Gravity Will Just Grow Stronger Today, almost every business has social media accounts, websites, mobile applications and so on. Data is being generated in multiple spaces. With so much external data generation, businesses across the world are shifting to the cloud. Cloud has lower overhead and infinite scalability, which helps in data analysis. The shift of data to the cloud has galvanised application and services to move to the cloud. Companies nowadays are setting up their tools in the cloud, connecting with a custom dashboard and sharing it in just minutes. And when you see the significant difference in the speed at data generation and analysation on the cloud with tools also hosted in the cloud, that’s when data gravity starts to make sense. Given this scenario, it has been prophesied that this shift is going to create larger data gravity forces. Having more benefits than an on-premise setup, the cloud is going to bring more companies on board. And as a result, as more and more business data is moved to the cloud or generated there, data analytics tools are also going to be cloud-based.","excerpt":"Of late, there is one more term that has gained a lot of traction — Data Gravity. According to Dave McCrory who coined data gravity in 2010, it can be defined as the analogy between the nature of data and the ability of data to attract additional applications and services. The concept is based on […]","categories":["AI Features"],"tags":[],"author_name":"Harshajit Sarmah","publish_date":"2019-02-11T06:14:33","publication_year":"2019","word_count":456,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","Scala","ViT","analytics","programming_languages:Scala","GAN","R"],"extracted_tech_keywords":["AI","analytics","R","Go","Scala","GAN","ViT","programming_languages:R","programming_languages:Scala","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/understanding-data-gravity-how-cloud-is-helping-it-grow-stronger\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10089324,"title":"Good Founder, Good Father, Good Husband Vembu","content":"Forbes recently published a detailed story on how Zoho CEO Sridhar Vembu had allegedly abandoned his wife Pramila Srinivasan and their special-needs son and is keeping them from receiving their share of assets. Today, Vembu took to Twitter to respond to the allegations calling the article ‘slander’ and his wife’s account ‘complete fiction’. Furthermore, Vembu has blamed his paternal uncle for the ‘malicious rumours’. 1\/ With vicious personal attacks and slander on my character, it is time for me to respond.This is a deeply painful personal thread. My personal life, in contrast to my business life, has been a long tragedy. Autism destroyed our lives and left me suicidally depressed.— Sridhar Vembu (@svembu) March 14, 2023 The news report claimed that Vembu, in cold blood, got rid of a huge amount of his stake at Zoho during their ongoing divorce case in California where the family lived for years. The transaction moved Zoho’s intellectual property to India and Vembu shifted a major chunk of 35% of shares to his sister (Radha) and her husband (Sekar). “He decided to make fictitious transfers or ‘sales’ of our most valuable community asset to his family members without their paying any cash or other consideration, and without ever telling me or asking my permission,” Pramila declared in a court filing earlier this year in January. (Source: Filing from the Vembu & Srinivasan divorce case) Denying the occurrence of this transaction, Vembu assured in the Twitter thread, “I never ever transferred my shares in the company to anyone else. I lived in the US for the first 24 years of our 27-year history and much of what constitutes the company was built in India. That is reflected in the ownership.” The Forbes report included the court filings made by Pramila. According to sources, Vembu’s sister Radha owns a 47.8% stake valued at around $2.2 billion, whereas her husband Sekar’s stake amounts to 35.2% at $1.6 billion. Vembu’s version of him being involved in taking care of his autistic son goes against Srinivasan’s court filings. The article also mentioned conflicting versions of the genesis of Zoho and questioned some of Vembu’s past claims. Currently, Zoho has a market value of between $5 billion – $15 billion. Last year, the software suite company crossed $1 billion globally in revenue. The SaaS company offers over 50 business applications to around 80 million users.","excerpt":"Zoho CEO Sridhar Vembu calls it ‘complete fiction’ in a Twitter thread","categories":["AI News"],"tags":["Sridhar Vembu","zoho"],"author_name":"Tasmia Ansari","publish_date":"2023-03-14T16:31:06","publication_year":"2023","word_count":395,"keywords":["Go","zoho","programming_languages:R","AI","programming_languages:Go","RAG","Sridhar Vembu","Aim","R"],"extracted_tech_keywords":["AI","Aim","RAG","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/good-founder-good-father-good-husband-vembu\/","complexity_score":3,"technical_depth":7,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":29618,"title":"Now Japanese Enterprises Gear Up For A Speedy Adoption Of AI","content":"Indian enterprises are working heavily with artificial intelligence and analytics. The adoption story for many enterprises in India has been like a fairytale. According to a recent survey by Goldsmiths and the University of London, India is ahead of Japan and the US in terms of AI adoption by enterprises. According to the report, nearly 71% of Indian respondents said that they are leveraging Robotic Process Automation (RPA) and AI to the fullest. On the other hand, Japanese enterprises have aggressively forayed into AI and automation technologies with the survey saying that 46% have already invested in AI and RPA technologies. Reportedly, another 51% plan to invest in the same, in the following 12 months. Japanese Enterprise And AI Let’s take the example of Riso Kagaku Corporation, a Japanese manufacturer of digital copiers and high-speed printers. The market has been tough due to fierce competitions in the printer industry and the market is shrinking. To deal with the competition RISO has started adopting analytics and AI in many of its processes. They recently started using a product from Teradata to improve forecasting accuracy and restructure how they work. Tomoko Shoji, a team leader of data analysis at Information System Dept in Riso Kagaku Corporation told Analytics India Magazine, “We have machine’s data, user data and data about how the machine was repaired. We wanted to analyse all of this data. We came up with some best practices for repairing based on the internal analytics project.” RISO is among the many in the Japanese printer industry who are adopting AI to help them stay relevant in a competitive market. Since April 2017 the company have also started using analytics to improve user satisfaction and streamlining repair operations by spotting patterns in large amounts of data. Tokyo-based company Rakuten Travel also is undergoing a massive automation phase and they are leveraging AI to retain and attract customers. Talking to Analytics India Magazine, Michifumi Katsuta who is a manager of their Analytics group, said, “We are trying to use AI for marketing and online advertising optimisation. We try to figure out who looks at what ads and then decide at what time and whom to send relevant ads along with fixing the price. All of this is enabled by data science and AI.” It is clear that using AI and analytics has helped Japanese enterprises to improve their bottom line results. Katsuta added, “Using AI and analytics have led to a 10% increase in hotel booking sales.” Indian Enterprises: Fastest Adopters of AI RPA and AI have now become one of the cornerstones of enterprise digitisation process. There is enough traction in India to make enterprises comfortable with changes. Analytics has resulted in a significant increase in revenue growth. It has also created new revenue streams thus offering a huge advantage. There are many areas in which Indian companies are also providing AI services. One of the positive signs that enterprise AI adoption has gone up is the rise of startups that cater to large enterprises. Sanjay Parthasarathy, CEO of AI-based startup Indix, which builds software for easing access to product information for enterprises says, “Our competitive edge comes from our scale. We have over two billions products in our corpus. More data is equal to faster learning and better algorithms. The large scale of data allows us to develop higher precision algorithms.” Indix products are popular among enterprises as it helps them filter large amounts of data using AI. Some of the rising vendors are Datamatics which is one of the leading providers of robotics, advanced analytics, mobility and cloud solutions in India, Q3Tech is a leading RPA software for BPO, healthcare, retail, telecom, BFSI and manufacturing among many other players. With the amazing ecosystem that has been built around analytics, India is making great strides in the area. The survey states, “Unsurprisingly, with such a positive background environment, India leads the way in reaping the benefits of augmentation, with 86% of the respondents saying that automation has increased employee productivity.” Conclusion With Japan and India making great strides in analytics adoption and at a fast rate, there are many things to learn from AI adoption journeys in two different enterprise cultures. One of the most important places where there is great space for analytics adoption is where employees are spending most of their time. In our conversations with both Indian and Japanese enterprise managers, we have found that employee enthusiasm is one of the most important factors in successful AI adoptions, especially in large enterprises. Enterprises who have worked with employees and understood their work process can work with technology partners to put ideas into executions. After many iterations of experiments and data collection of AI augmenting the workforce, there can be great progress in enterprise productivity.","excerpt":"Indian enterprises are working heavily with artificial intelligence and analytics. The adoption story for many enterprises in India has been like a fairytale. According to a recent survey by Goldsmiths and the University of London, India is ahead of Japan and the US in terms of AI adoption by enterprises. According to the report, nearly […]","categories":["AI Features"],"tags":["japan","RPA","teradata"],"author_name":"Abhijeet Katte","publish_date":"2018-10-26T06:18:51","publication_year":"2018","word_count":796,"keywords":["japan","teradata","data science","Go","artificial intelligence","AI","RPA","ML","Git","RAG","Ray","analytics","R"],"extracted_tech_keywords":["AI","artificial intelligence","ML","data science","analytics","Ray","RAG","R","Go","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/now-japanese-enterprises-gear-up-for-a-speedy-adoption-of-ai\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":20647,"title":"A Look At Artificial Intelligence Marketplace And AI Innovations at CES 2018","content":"Before CES took a centre stage at Las Vegas this year, there were already speculations of AI and robotics making a key appearance at the event. And, keeping up to the high expectations, the Consumer Electronics Show revealed a plethora of new products, software and services in this area from the key technology players. Artificial Intelligence has become the key ingredient of almost all the major developments in the recent years. CES 2018 recognised this emerging space by introducing AI category for the first time in CES Marketplace, where thousands of companies exhibit new products. With an aim to exhibit applications of AI in areas such as automotive, health, home, robotics etc., self driving cars, CES featured Artificial Intelligence Marketplace and conference tracks on AI. One of the world’s largest tech conventions was seen boosting the latest innovations in AI infrastructure and computer systems. As the website mentioned—from big data analytics and solutions, speech recognition, learning and decision-making products to predictive technology, the AI Marketplace showcased revolutionary data models and solutions that will shape the future of businesses. Some of the leading technology players across the globe were exhibitors at artificial intelligence marketplace—Baidu, Aeolus Robotics, Backyard Brains, Horizon Robotics, NovuMind, Yamaha Motor, being a few. Other tech parallels that drew major traction at CES  2018 were robotics, augmented reality, drones, 3D printing and others. From AI themed panels to AI talks that grabbed an eye A culmination of software, algorithms and sensors, AI was the real star at CES, which was also indicative of how CES has evolved over the years from being a venue for tech companies to launch their products to turning into a showcase for emerging technologies like artificial intelligence. It had conference tracks dedicated to AI, featuring speeches and roundtables from how businesses are putting AI to work to the effects that it will have on humans, both as consumers and workers. There were talks on impact and future of AI, concerns about AI, how AI is rapidly changing the way we work and live, insights into the future of AI, examining recent development, current applications and concerns about AI, roadmap for self driving cars and others. Transformative tech trends around AI at CES 2018 This year’s CES presented some of the most defined applications related to AI such as image recognition and deep learning. From LG launching their robots to Samsung revealing smart refrigerators, there were a plethora of AI and robotic developments that took a centre stage. Some of the AI trends that deserves a mention are: AI driven Cars: Though autonomous vehicles still have a long way to go before they become safe and properly regulated, the companies aren’t shying away from developing solutions in the space. With car companies demonstrating their developments in AI to NVIDIA talking about its partnership with Uber and Volkswagen, there was a lot to be consumed on self driving cars facilitated by AI. Some of them demonstrated features like ability to pay for parking, gas through a dashboard or cameras that enhance a driver’s side and rear vision. Smart Cities and AI assistants: Companies like Samsung launched its smart refrigerator that can listen to voice commands to control other home accessories. It also showcased an updated version of its assistant, Bixby. This AI assistant is similar in many ways to Siri, Alexa, and the Google Assistant and the company plans to incorporate this tech into technologies like televisions and even refrigerators. There were a lot of sessions focusing on Alexa, covering topics such as building smart home with Alexa, art of talking to Alexa, integrating Alexa voice into other products and others. Google also showcased its own Assistant quite fervently at CES. Next-gen wireless technology: Companies gave a progress report on the so-called 5G, the fifth-generation network technology, which has gained major traction since the rise of artificial intelligence. Robots and virtual reality: CES 2018 saw a throwback to “retro futuristic” robots that combine contemporary tech with 20th-century aesthetics. Laundry-folding robots, robot dogs, and even a robotic smart home manager named CLOi by LG made an appearance. It is designed to work in airports, hotels and supermarkets. Sony’s Aibo is a robot dog that has been reborn for CES 2018. VR was also at a forefront, with HTC unveiling Vive Pro headset, and Irish company Design Partners revealing its ‘smart glove,’ a haptic human-computer interface system for VR and augmented reality. 🐶 @Sony 🐶 #CES2018 pic.twitter.com\/owsXOGiQgI — CES (@CES) January 12, 2018 Some AI product launches at CES 2018 It hosted a variety of AI image recognition and learning examples. Few gadgets that were unveiled at the Consumer Electronics Show in Las Vegas with an AI play are: World’s first AI equipped table tennis tutor: Japanese electronics company OMRON has created an AI-equipped robot that demonstrates how machines can work in harmony with humans. It uses AI, sensing control, ML and robotics to create FORPHEUS. The three camera system on it is used to identify ball, players and trajectory. A dude playing pingpong against AI #CES @newscomauHQ pic.twitter.com\/r8u1Yoqmqf — matthew dunn (@mattydunn11) January 10, 2018 Cocoon cam: This smart baby monitor camera recognizes if a baby is breathing and sends real time data and alerts. It has sensors for humidity, temp, sleep analytics, and alerts if the baby is crying. The Cocoon Cam Clarity uses computer vision to track a baby’s sleep and wake patterns, growth and other behaviors. PianoGo: This AI-powered iPad app uses machine learning to monitor and track how you play, comparing the results and provide an assessment. It works with The One Smart Piano, and the AI can tell if you playing with accuracy, good rhythm etc. Speak Music Muse: It brings Alexa to your car dashboard, using voice recognition. This voice enables controller for car can follow commands like play, pause, skip etc. It works with Bluetooth or the Aux in port.","excerpt":"Before CES took a centre stage at Las Vegas this year, there were already speculations of AI and robotics making a key appearance at the event. And, keeping up to the high expectations, the Consumer Electronics Show revealed a plethora of new products, software and services in this area from the key technology players. Artificial […]","categories":["IT Services"],"tags":["augmented intelligence for smart industry","most revolutionary ai deep learning company"],"author_name":"Srishti Deoras","publish_date":"2018-01-16T05:56:56","publication_year":"2018","word_count":980,"keywords":["AI assistants","artificial intelligence","machine learning","AI","augmented intelligence for smart industry","ML","image recognition","computer vision","Aim","deep learning","most revolutionary ai deep learning company","analytics"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","computer vision","analytics","Aim","AI assistants","image recognition"],"url":"https:\/\/analyticsindiamag.com\/it-services\/look-artificial-intelligence-marketplace-ai-innovations-ces-2018\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10068586,"title":"Web3 startup Samudai raises USD 2.5 Mn in pre-seed funding","content":"DAO platform Samudai recently closed a pre-seed financing round of USD USD 2.5 million. FTX Ventures, Sino Global Capital, A&T Capital, Coinbase Ventures, Lunar Ventures, NxGen, Paradigm Shift Capital, DWeb3 Capital, Builders Tribe, and Superteam DAO participated in the round. Samudai, which means “community” in Hindi, was founded in 2022 and incorporated in Singapore by Kushagra Agarwal and Navin, who met in college and have been working together for over five years. The company operates globally and remotely. The fresh funds will be used to grow the team, accelerate development, and bring its platform to market. “We’re grateful for the support of such well-known investors in innovation which will help us make it easier for DAO admins and contributors to nurture a healthy work-life while advancing global coordination in this industry,” said Samudai co-founder Navin. The startup is planning a full productivity suite for DAOs that includes a Web3 native project management framework, team graphs, an analytics dashboard, a roster of talented Web3 professionals to facilitate collaboration and a verifiable reputation metric called “Bushido.”The platform enables unprecedented levels of customisation and facilitates large-scale coordination and executions among DAOs. Samudai plans to launch a private alpha version of its platform to a small group of users this Summer, around mid-July, with a public beta phase later in 2022.","excerpt":"The fresh funds will be used to grow the team, accelerate development, and bring its platform to market.","categories":["AI News"],"tags":["AI Startups"],"author_name":"Zinnia Banerjee","publish_date":"2022-06-08T14:49:07","publication_year":"2022","word_count":217,"keywords":["API","programming_languages:R","AI","innovation","ViT","analytics","R","AI Startups","startup"],"extracted_tech_keywords":["AI","analytics","R","API","ViT","innovation","startup","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/web3-startup-samudai-raises-usd-2-5-mn-in-pre-seed-funding\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":18137,"title":"Leadsemantics&#8217; Hiddime Cloud Is Giving New Meaning To Cloud Analytics Services","content":"Leadsemantics, a promising startup in the field of Analytics, with their offices in Hyderabad and Houston, US, has had an interesting beginning when it comes to Analytics. When asked about their beginning, Prasad Y, the CEO of HIDDIME.COM, a Cloud Analytics Service and the founder of Leadsemantics, a semantic Big Data analytics company (which owns Hiddime) says, “I came in contact with the concept of Data Analysis in the context of telecom subscriber analysis.” It was around then, that Prasad founded a telecom analytics company that developed and sold a carrier grade Analytics Appliance built with COTS hardware — a first of its kind in 2008 offering Peta-scale network data mining and analysis for around $1 million. Prasad, who can be considered as an analytics evangelist with his experience in the industry, talked to Analytics India Magazine about transforming Big Data into Intelligent Data. He said, that intelligent data showcases the transformation steps that data must go through to be of value to the enterprise decision processes: Data -> Information -> Knowledge -> and later Intelligence with the help of reasoning and wisdom Talking about Hiddime, a data analytics service in the cloud, and a product of Leadsemantics, Prasad explained that the what differentiated their cloud from others was the use of Investigative Discovery and Exploratory Analytics tool (IDEA tool) for the frontline business managers and domain specialists who are not necessarily IT experts. A Hybrid Graph Relational store (or Semantic RDF Store) undergirds the BI Data Warehouse and retrieval system. It enables easy yet exhaustive querying along with integration of data from external sources. Over the last decade, data landscape in the enterprise has rapidly changed along with changing user and market expectations attached to the BI and analytics off the data that is newly available. The unprecedented and exponential growth of sophisticated internet\/web enabled consumer devices (smartphones, etc.) that are capable computers themselves, generating vast amounts of data and powerful computer and communication technologies along with potent software techniques such as graph processing and machine learning has enabled ever-complex processing work-loads generating deeper insights – defining characteristic of big data. Some of the key features of the Hiddime cloud are: Exploratory: Easy unhindered exploration through visual interaction which enables investigating deeper connections with in the data, especially by business analysts or domain specialists who know the data Investigative: Hiddime works with traditional BI Data Warehouses and modern Hybrid BI Data Warehouses and extends the seamless exploration to connected graphs and semi-structured data, among others. CURRENT PROJECTS: Predicting probability of default of loans Predicting machine failures such as large industrial furnaces and turbines Social Networking analysis for target marketing Predicting effectiveness of various financial instruments in enhancing incomes of rural women under government and quasi-gov self-help group schemes For the Analytics sector in India, however, Prasad added that machine learning was important but is not alone in making up the AI workbench for the modern enterprise analytics. Language Processing, Graph Processing and Semantics Technology are equally important. “India might be behind in semantic technology and graph technologies at this time. But its an opportunity, India can leap frog other countries if ‘semantic technology’ is embraced on a wider scale as every ‘data integration and ETL project’ that our services companies are leaders they can vastly improve eficiencies. Governments are other major beneficieries of Semantic Technology as the data governmnts deal with are of importance to any and evry one and Sem Tech is the best technology to deal with data in a standardized fashion,” he said. KEY REFORMS NEEDED IN THE SECTOR: Prasad explained that as an activity ‘data integration’ is a necessary and by far most expensive piece in an enterprise environment. Folks have forgotten that this work could be targeted for innovation. In an analytics environment data integration becomes important. Emphasis on ‘education of core computer science and mathematics & statistics’ is beyond tool knowledge such as delivered in most commercial training. In addition importantly, communication skills among the technical folk to be able explain and defend analytics outcomes that are result of complex engineering.","excerpt":"Leadsemantics, a promising startup in the field of Analytics, with their offices in Hyderabad and Houston, US, has had an interesting beginning when it comes to Analytics. When asked about their beginning, Prasad Y, the CEO of HIDDIME.COM, a Cloud Analytics Service and the founder of Leadsemantics, a semantic Big Data analytics company (which owns […]","categories":["Deep Tech"],"tags":[],"author_name":"Prajakta Hebbar","publish_date":"2017-10-06T13:05:11","publication_year":"2017","word_count":678,"keywords":["big data","Go","API","machine learning","AI","ETL","ML","analytics","R","data warehouse"],"extracted_tech_keywords":["AI","machine learning","ML","analytics","R","Go","API","big data","ETL","data warehouse"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/startup-week-leadsemantics\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":6,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10075223,"title":"Dr. Reddy&#8217;s Laboratories is certified as a Best Firm For Data Scientists","content":"Dr. Reddy’s Laboratories is certified as the Best Firm For Data Scientists to work for by Analytics India Magazine (AIM) through its workplace recognition programme. The Best Firm For Data Scientists certification surveys a company’s data scientists and analytics employees to identify and recognise organisations with great company cultures. AIM analyses the survey data to gauge the employees’ approval ratings and uncover actionable insights. “I’m extremely delighted that Dr. Reddy’s Laboratories has been chosen as a Best Firm for Data Scientists by Analytics India Magazine. To be honoured by India’s no.1 platform on analytics, data science and big data is incredibly gratifying and I’m grateful to the talented and dedicated individuals I work with every day. It’s a true testament to the success we’ve seen in our business, but also our greater purpose. How our people feel about working at Dr. Reddy’s Laboratories, and the impact we’re making by designing the world class data infrastructure. We’ve always believed that we’ve got to be the best to get the best,” said Anish Agarwal, Head of Analytics, Dr. Reddy’s Laboratories. “We continue to develop the best use cases by understanding data from multiple sources and derive valuable insights and enable the business to make smarter data-driven decisions.” he added. The analytics industry at AIM faces a talent crunch, and attracting good employees is one of the most pressing challenges that enterprises are facing. The certification by Analytics India Magazine is considered a ‘Gold Standard’ in identifying the best data science workplaces and firms across India participate in the programme to increase brand awareness and attract talent. Best Firms For Data Scientists is the biggest data science workplace recognition programme in India. To nominate your organisation for the certification, please fill out the form at this link.","excerpt":"Dr. Reddy’s Laboratories is certified as the Best Firm For Data Scientists to work for by Analytics India Magazine (AIM) through its workplace recognition programme.  The Best Firm For Data Scientists certification surveys a company’s data scientists and analytics employees to identify and recognise organisations with great company cultures. AIM analyses the survey data to […]","categories":["AI Highlights"],"tags":["best companies for data scientists in india","best companies in india for data science fresher"],"author_name":"AIM Media House","publish_date":"2022-09-14T18:00:00","publication_year":"2022","word_count":294,"keywords":["best companies for data scientists in india","data science","Go","big data","programming_languages:R","AI","data-driven","best companies in india for data science fresher","Aim","analytics","GAN","R"],"extracted_tech_keywords":["AI","data science","analytics","Aim","R","Go","big data","GAN","data-driven","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/dr-reddys-laboratories-is-certified-as-a-best-firm-for-data-scientists\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10043560,"title":"Top 10 Tools To Kickstart Your MLOps Journey In 2021","content":"The MLOps market is expected to grow by almost $4 billion by 2025, according to analytics firm Cognilytica. Amazon, Google, Microsoft, IBM, H2O, Domino, DataRobot and Grid.ai have all incorporated MLOPs capabilities into their platforms. Most companies are using MLOPs for automation pipeline, monitoring, lifecycle management, and governance. According to Algorithmia, last year, close to 22 percent of companies have had machine learning models in production for 1-2 years, demonstrating a reasonably significant transition towards productionization even if it is still early for most companies. In this article, we list the best open-source MLOps tools and services to help businesses and individuals kickstart their MLOps journey: Algorithmia Algorithmia is a single solution platform for all stages of ML operations (MLOps) and management lifecycle. It enables ML and operations teams to work together on complex machine learning applications in one central location. At present, more than 100,000 engineers and data scientists are using the platform, including the staff of the United Nations and multiple Fortune 500 companies. All the experiments and projects can be found here. Comet ML Comet enables data scientists and teams to track, compare, explain and optimise experiments and models across the entire lifecycle. The complete examples and libraries are available on GitHub. Also, check out how Uber manages machine learning experiments with comet here. DVC DVC is an open-source ‘version control system’ for machine learning projects. It tracks machine learning models and data sets. The platform has been built to make ML models shareable and reproducible. DVC is designed to handle large files, data sets, machine learning models, and metrics and code. You can experiment with the DVC project here. Kubeflow The Kubeflow project is dedicated to making ‘deployments’ of ML workflows on ‘Kubernetes’ simple, portable and scalable. It provides components for each stage in the ML lifecycle, from exploration to training and deployment. Check out how to install Kubeflow and experiments here. Metaflow Metaflow was initially developed at Netflix to boost the productivity of data scientists working on various projects, from classical statistics to SOTA deep learning. It is a Python\/R library that helps scientists and engineers build and manage real-life data science projects. The source code is available on GitHub. MLFlow MLFlow is an open source platform for managing the machine learning lifecycle, including experimentation, reproducibility, deployment, and a central model registry. It currently offers four components: tracking, projects, models, and registry. The source code for MLFlow is available on GitHub. Also, check out deploying R Models with MLFlow and Docker here. Neptune.ai Neptune is a metadata store for MLOps, developed for teams that run a lot of experiments. It provides a single place to log, store, display, organise, compare, and query all model building metadata. In addition, the MLOps platform is used for experiment tracking, model registry, and monitoring ML runs Live. Check out open source repositories and projects on GitHub. Polyaxon Polyaxon is a machine learning platform for Kubernetes (also used as MLOps tools for experimentation and automation). The platform helps in building, training, and monitoring large scale deep learning applications. Polyazon makes it faster, easier, and more efficient to develop deep learning applications by managing workloads with smart container and node management. It turns GPU servers into shared, self-service resources for individuals as well as enterprises. You can start your Polyaxon project here. Valohai Valohai focuses on models, code and data. It allows users to easily run the powerful cloud machines with a single click (UI) or a single command (CLI & API). It can be set up on any cloud vendor or on-premise to orchestrate machines automatically. Repository and experiments on Valohai are available on GitHub. Weights & Biases Weights & Biases is a ‘developer tool’ for machine learning. With this, users can build better models faster, alongside experiment tracking, dataset versioning, and model management. Check out all the machine learning experiments and repositories here.","excerpt":"The MLOps market is expected to grow by almost $4 billion by 2025, according to analytics firm Cognilytica. Amazon, Google, Microsoft, IBM, H2O, Domino, DataRobot and Grid.ai have all incorporated MLOPs capabilities into their platforms.  Most companies are using MLOPs for automation pipeline, monitoring, lifecycle management, and governance. According to Algorithmia, last year, close to […]","categories":["AI Trends"],"tags":["Machine Learning","Machine Learning New","MLOps","MLops tools"],"author_name":"Amit Naik","publish_date":"2021-07-14T18:00:00","publication_year":"2021","word_count":644,"keywords":["data science","machine learning","Machine Learning New","AI","ML","MLOps","Machine Learning","MLops tools","deep learning","Kubeflow","analytics","MLflow","Weights & Biases"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","data science","analytics","MLOps","MLflow","Kubeflow","Weights & Biases"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-10-tools-to-kickstart-your-mlops-journey-in-2021\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10758,"title":"HR Analytics Startup inFeedo to raise $150,000 in Angel Round","content":"inFeedo Funding 3 years ago, Tanmaya Jain and his college roommate Varun Puri embarked upon a journey to make the average employee felt valued, with their employee engagement platform, inFeedo. After bootstrapping over a year on client revenue, today they have raised an undisclosed amount in an angel round led by managing partner at Redcliffe Capital, a UK-based hedge fund. This $150,000 round was led by Dheeraj and preceded by Palash Jain, Ex-Head of Google India Core Operations. inFeedo is in conversations with relevant angels and HNIs having a strong background in the HR-Tech space to complete the remaining round in the month of September. “inFeedo is the next disruption, having aligned itself perfectly with the global shift towards people management. We have heard stellar feedback from their clients on how they have arrested attrition and helped boost employee productivity. It’s efficient SaaS model will enable them manage culture in organizations across the globe!” said Dheeraj, Managing Partner, Redcliffe Capital. The Next Wave of Chatbots inFeedo’s new AI chatbot “Amber” has been an instant hit with HR heads at geographically distributed companies like MakeMyTrip, Lava Mobile, Knowlarity, RedBus and Nearbuy. Built purely on client feedback, this bot can now automatically talk to employees at specific intervals in their lifecycle and based on their chats, auto-generate Culture Reports for HR leaders to proactively meet those who’re unhappy, facing troubles or about to leave. “If HR and Data Science were to marry, inFeedo would be the baby. It’s been a real emotional roller coaster ride, watching our baby grow up into a platform that can engage 5500+ employees, generate 1000+ ideas, host 10+ anonymous townhalls and predict attrition in ways that are humanly impossible”,  said Tanmaya, Founder, inFeedo. An Unusual Team 21 year old, Tanmaya has been growing inFeedo’s paid user base at the rate of 53% QoQ and is cashflow positive, with an unusual team that includes a college dropout from St. Xavier’s, an in-house data scientist and an aspiring psychologist who spent 15 months in Antarctica. “They are a very creative set of youngsters who are youngsters in terms of age but when it comes to their conversation, discussion and point of view, they can beat any senior guy 30\/30+ with their logic”, said Yuvaraj Srivastava, CHRO, MakeMyTrip. With less than 10 team members, this team surfed the recent huge waves of “startups surge\/downturn” and emerged profitable by keeping cost of operations at bare minimum. By end of 2016, inFeedo aims to touch more than 10,000 paid users and is looking to partner with more B2B HR companies as it scales. A Growing Market. A Growing Market With the rising adoption of technology by HR, SMB SaaS is expected to grow to $76 billion in 2020 and India is expected to capture eight percent of this market. inFeedo is raising this round to meet this growing demand, scale across international waters and hire talent that can build upon Amber, their chatbot’s machine learning, natural language processing and predictive analytics capabilities. To sign up for amber you can visit amber.infeedo.com To have a look at pitch by inFeedo founder and a testimonial from MakeMyTrip, CHRO you can visit https:\/\/youtu.be\/vdH5e1hGVrc","excerpt":"inFeedo Funding 3 years ago, Tanmaya Jain and his college roommate Varun Puri embarked upon a journey to make the average employee felt valued, with their employee engagement platform, inFeedo. After bootstrapping over a year on client revenue, today they have raised an undisclosed amount in an angel round led by managing partner at Redcliffe […]","categories":["AI News"],"tags":["Big Data","HR Analytics"],"author_name":"Manisha Salecha","publish_date":"2016-09-21T06:07:20","publication_year":"2016","word_count":528,"keywords":["data science","Go","machine learning","AI","HR Analytics","chatbots","R","RAG","Aim","analytics","Big Data","predictive analytics"],"extracted_tech_keywords":["AI","machine learning","data science","analytics","Aim","RAG","chatbots","predictive analytics","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/hr-analytics-startup-infeedo-raise-150000-angel-round\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":42220,"title":"YES BANK Issues Commercial Paper Using Blockchain For The First Time In Asia","content":"In an interesting move, YES BANK recently facilitated the issuance of a Commercial Paper (CP) of INR 100 crore using Blockchain technology for Vedanta Limited, a natural resources conglomerate. This is the first time in Asia that a CP has been digitally issued using Blockchain technology. This transaction was completed in partnership with MonetaGo, a leading provider of financial technology headquartered in New York. The digital solution ensures an efficient, transparent and secure mechanism for CP issuance and redemption. As part of the business solutioning, the stakeholders in the CP issuance and redemption benefit from: Reduction in Turnaround Time (TAT) for issuance and redemption Immutable digital records of the entire transaction documents thereby reducing operational risk Real-time visibility of the CP issuance and redemption Common network for all participants in the CP issuance and redemption process Speaking on the implementation, Vinod Bahety, Head-Corporate Infrastructure Banking and Corporate Banking, YES BANK said, “We are pleased to partner with Vedanta and this first of its kind transaction reaffirms YES BANK’s strong commitment to bringing customised financial solutions through innovative practices and knowledge banking approach to the infrastructure sector. We believe that such pioneering solutions will significantly ease the commercial paper issuance process for regular issuers.” GR Arun Kumar, Group CFO, Vedanta Group, said, “Vedanta will benefit from the digitised and simplified workflow which shortens the laborious process running into hours to just a few minutes and complete transparency that this platform offers to all stakeholders.” Jesse Chenard, CEO, MonetaGo, said, “This is a real milestone for our company, for the technology, and for the Indian Banking Community. We have been working with YES BANK and Vedanta very closely and have been continuously impressed with their ability to leverage technology to bring about large scale change. These are big corporations with the ability to move like start-ups, something very rare indeed especially when dealing with a market this size.”","excerpt":"In an interesting move, YES BANK recently facilitated the issuance of a Commercial Paper (CP) of INR 100 crore using Blockchain technology for Vedanta Limited, a natural resources conglomerate. This is the first time in Asia that a CP has been digitally issued using Blockchain technology. This transaction was completed in partnership with MonetaGo, a […]","categories":["AI News"],"tags":["Blockchain","Yes Bank"],"author_name":"Prajakta Hebbar","publish_date":"2019-07-11T10:58:48","publication_year":"2019","word_count":316,"keywords":["Go","Blockchain","programming_languages:R","AI","programming_languages:Go","Git","RAG","Yes Bank","R"],"extracted_tech_keywords":["AI","RAG","R","Go","Git","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/yes-bank-issues-commercial-paper-using-blockchain-for-the-first-time-in-asia\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":35816,"title":"A Day At The Rising 2019: Exclusive Conference For Women In Analytics &#038; AI","content":"The buzz around Women’s Day has taken a centre stage in the analytics and tech industry. To empower women in data science and talk about their inspiring stories about women data science leaders, our Women in AI and Analytics conference — Rising 2019 is going to be the biggest meeting of women in data science. The conference which aims to bring leaders, early-stage professionals and students from academia and industry will serve as a place to meet and support women developers in the data science field. The conference will be a day full of learnings, sharing experiences and insightful talks with a slew of women leaders and full of opportunities to learn about data science-focused career paths, explore business opportunities and build strong connections in the field. The conference will begin with a keynote by Saraswathi Ramachandra who is the head of Analytics Center of Excellence at Danske IT. A senior leader with an experience of more than 21 years in various domains, she is going to open the note with a very relevant topic — Is AI Sexist? At a time when we are fighting issues such as sexism, there have been several instances where AI-based tool has proven to be biased against women. Her talk will focus on technology and if it is dooming to reflect the worst in all of us. This interesting and thought-provoking session would be followed by a talk from Deepika Sandeep, who is the head of AI and analytics at BLP Clean Energy. She would focus on how to drive towards a cleaner future through data. This would be followed by an engaging panel discussion, titled Women’s Day 2020 — Changing the Narrative. It would be lead by industry stalwarts Anshu Sharma, Managing Director, Global Head – Retail Banking Technology at Standard Chartered Bank, Pankaj Rai Senior Vice President Strategy at Wells Fargo, Sahana Shetty HR Leader – Technology at ANZ, Aparajita Karimpana Senior Director Analytics – Envestnet Yodlee and Barkha Sharma, CEO and Founder, Bash.ai. The Rising is also going to host Women in AI Leadership Awards, in association with Jigsaw Academy which will celebrate Tech leaders driving disruption and innovation in data science and AI. Celebrating the importance of women in AI, the awards are aimed at recognising the achievements of technology leaders who have made an incredible impact in the tech community by developing AI solutions. After a session of networking over lunch, the second half of the day is full of interesting talks by Vikram Vij Sr. Vice President, Samsung R&D Institute, Dharani Karthikeyan Director of engineering at SAP,  Aruna Schwarz CEO at Stelae Technologies who are going to talk about topics such as Samsung Bixby Voice Assistant, unlocking the power of intelligent enterprise with augmented analytics, building an enterprise software company with customer revenue and no VC funding respectively. Up next, Why We Need More Women In AI, a talk by Vaishali Kasture Co-Founder at Sonder Connect will focus on how AI will pass on the biases of its creators and the data its creators feed it. She will focus on how if we want there to be a woman’s perspective in the new world of AI, we need women to be part of it. The conference also deals with bigger issues that women face, for example, breast cancer, which goes unnoticed in its early stages, leading to larger problems. The talk by Geetha Manjunath who is the founder and CEO of NIRAMAI will focus on enabling early stage breast cancer detection using AI. Manjunath will talk about the novel and noble application of AI in the area of healthcare developed at NIRAMAI – a noninvasive, affordable and accessible solution to detect early stage breast cancer. After insightful talks by Damini Gupta,  AVP, AI and Fintech at Mphasis NextLabs and Shivani R Gupta, Head, Data Science and AI, Financial Services at Capgemini, the concluding talk will be delivered by Mathangi Sri, Head of Data Science at PhonePe who will share her experience of working with data. The talk will focus on her learnings over the last 15 years where she has worked across multiple problems and multiple data types, and how she overcame the challenges and explored new opportunities. The biggest gathering of women professionals in data science, IT industry and academia, Rising 2019 is open to all. Apart from engaging talks by the industry stalwarts, it will serve as a perfect platform to meet organisations, build relationships, learn about data science-focused career paths, explore business opportunities and build strong connections in this field. Come join us for a day full of learning, empowerment and fun at The Rising 2019, on 8 March at Taj MG Road, Bengaluru. You can book your tickets here.","excerpt":"The buzz around Women’s Day has taken a centre stage in the analytics and tech industry. To empower women in data science and talk about their inspiring stories about women data science leaders, our Women in AI and Analytics conference — Rising 2019 is going to be the biggest meeting of women in data science. […]","categories":["AI Features"],"tags":["Women in AI","Women in Analytics","Women in Data Science"],"author_name":"Srishti Deoras","publish_date":"2019-03-06T05:20:40","publication_year":"2019","word_count":786,"keywords":["data science","Go","Women in Data Science","funding","AI","innovation","Aim","disruption","analytics","Women in Analytics","GAN","R","Women in AI"],"extracted_tech_keywords":["AI","data science","analytics","Aim","R","Go","GAN","innovation","disruption","funding"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/a-day-at-the-rising-2019-exclusive-conference-for-women-in-analytics-ai\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10121330,"title":"Ola Has Nothing to Do with Microsoft Azure Anymore","content":"Ola CEO Bhavish Aggarwal fulfilled his promise of moving all AI infrastructure workloads from Microsoft Azure to Krutrim AI Cloud. “Will help others also exit and move to our own Indian stack. More than 2500 devs have signed up!! Will be working with everyone to get onto our cloud services over coming weeks,” Aggarwal said on X. Krutrim, which became India’s first AI unicorn, launched Krutrim AI Cloud earlier this month. The service offers AI computing infrastructure and access to both its foundational models and open-source models like Mistral and Meta’s Llama 3, enabling developers to build and run LLMs cost-effectively. This move has positioned Krutrim to compete with major cloud providers like Microsoft Azure, Google Cloud, and AWS. However, many users pointed out that Ola’s workload is primarily on AWS rather than Azure. What a joke! All this big talk about moving away from @Microsoft @Azure within a week for this most bullshit-iest of reasons when Ola's workload is primarily on @awscloud .This thread will show document all the critical resources of @Olacabs that are hosted with @awscloud… https:\/\/t.co\/GjAB7oXBzC pic.twitter.com\/vlIUoDNonO— mas.to \/ (@kingslyj) May 11, 2024 “Does he think people who deploy and run companies’ cloud or work in AI are as stupid as consumers who buy two-wheelers?” asked Kissan AI founder Pratik Desai. Another user on X pointed out, “Why does it sound like a pre-planned marketing campaign for launching Krutrim cloud? You can’t have these things done overnight for sure. Good way to grab attention.” But what made Aggarwal take this drastic move, and how was it pre-planned? Recently, Microsoft’s LinkedIn took down his post. He had called the networking platform’s use of gender-neutral pronouns like they\/them “pronoun illness” and hoped it would not reach India. According to him, “pronoun illness” is being taught by “big city schools” and is increasingly appearing in CVs—which he clearly is not a fan of. He believes India “needs to know where to draw the line in following the West blindly!” That is why he decided to move his workload from the big tech’s cloud to its in-house cloud. This is why he decided that India needs to create its own tech ecosystem so that we do not get “governed by western Big Tech monopolies.”","excerpt":"Ola’s workloads are now on Krutrim AI Cloud.","categories":["AI News"],"tags":["Bhavesh Aggarwal","Microsoft Azure","Ola"],"author_name":"Shritama Saha","publish_date":"2024-05-23T13:20:19","publication_year":"2024","word_count":374,"keywords":["Go","Ola","unicorn","AWS","AI","cloud_platforms:AWS","R","cloud_platforms:Azure","llm_models:Llama","cloud_platforms:Google Cloud","Microsoft Azure","Bhavesh Aggarwal","Azure"],"extracted_tech_keywords":["AI","AWS","Azure","R","Go","unicorn","llm_models:Llama","cloud_platforms:AWS","cloud_platforms:Azure","cloud_platforms:Google Cloud"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/ola-has-nothing-to-do-with-microsoft-azure-anymore\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":22439,"title":"The Future Of Energy Seems Intense With The Exponential Development In The AI Sector","content":"Ever since 2016, Artificial Intelligence has been something of a hot topic of conversation among entrepreneurs and the reputed business magnates of the world. In fact, in 2017, the term Artificial Intelligence (AI) was ranked No. 7 on Gartner.com, showing how popular and widespread it had become in just a year. In simple words, AI refers to the branch of computer science that seeks to create and develop intelligent or ‘smart’ machines that imitate human behavior, that is, learn from experience and perform human-oriented tasks. The term Artificial Intelligence was coined 1956, and today, it combines both machine learning and deep learning under one umbrella. Source: Chatbots Magazine How Is AI Transforming The Business World? AI is the hottest trend now in the business world. Dan Walker, head of Emerging and Disruptive Technology at the British Petroleum’s (BP) Technology Group, thinks: “AI is enabling the fourth industrial revolution, and it has the potential to help deliver the next level of performance.” Business magnates and entrepreneurs around the world are incorporating AI within their structure. This is mainly because AI provides the opportunity to companies to improve their overall customer experience; to automate work processes; to boost their employee performance, and most importantly, to develop intelligent machines to aid them in their day-to-day functioning. A recently published Harvard Business Review (HBR) maintains that about 51% of AI leaders have predicted that by the year 2020, AI will create a massive internal impact on the back office functions of IT, HR, and finance\/accounting. This will enable the employees to divert their attention to more important tasks while AI will take care of repetitive and automatable tasks (questions about company policy, reset passwords, etc.). The following image shows how AI is becoming increasingly popular in the business world. Source: Chatbots Magazine According to Tata Consultancy Services’ Press Release on a study conducted on 835 executives across 13 global industry sectors in four regions of the world, many business executives are of the opinion that AI will facilitate a net reduction anywhere between 4-7% in company functions by 2020. K Ananth Krishnan, Chief Technology Officer of TCS, stated: “As companies begin to gain a better understanding of AI’s application for business, they will realize the significant impact of this transformative force. This is reflected in our Global Trend Study, which shows that forward-thinking companies are beginning to make major AI investments.” Source: Harvard Business Review As you can see, all the major tech giants of the world are increasingly and heavily relying on AI. What Is The Impact Of AI On The Energy Sector? In 2017, Bill Gates wrote an online essay addressing college graduated all over the world. He made a very important point when he said: “If I were starting out today… I would consider three fields. One is artificial intelligence (AI). We have only begun to tap into all the ways it will make people’s lives more productive and creative. The second is energy because making it clean, affordable, and reliable will be essential for fighting poverty and climate change.” The third field, according to Gates is biosciences. AI holds great promises of revolutionizing the energy sector. In fact, the energy sector has already started utilizing the potential of technologies like AI, automation, and advanced analytics in most imaginative ways possible. According to a research study by Infosys, 48% of the respondents from energy, oil, gas, and utility industries consider AI to be a fundamental driving force behind their organization’s success, whereas 46% state that their organizations are incorporating AI into the “company ethos.” The primary drivers behind the incorporation of AI into the energy sector are the automation of IT and business processes, and of course, to boost innovation. Big data and advanced technologies are opening up new possibilities in the energy industry.  AI can coordinate and optimize the use of energy resources as well as IoT.  The Infosys research study further stresses that data leveraged by AI will be a primary “driver of a future sustainable energy ecosystem that includes an appropriate mix of fossil fuels and renewables.” The benefits of using AI in the energy sector are numerous. AI will bring about a vast improvement in common utility operations in the following ways: Reliability – Powering self-healing grids and improving operations as well as ensuring the efficient use and storage of renewable resources Safety – Facilitate outage prediction and outage response Cybersecurity – Enhanced threat detection and response Optimization – Improved management of assets, maintenance, workflow, and portfolio Enhanced customer experience – Facilitate quicker and more intuitive interactive voice response, personalization, and product and service matching Apart from these benefits, AI can help combat the demand management issue through predictive algorithms. These algorithms can help decide when to store and when to release energy for balancing power grids. They can also accurately predict the production and consumption of small-scale producers and consumers respectively. AI technologies are already at work to improve the forecasting and equipment efficiency of renewable energy resources by accumulating data from wind turbines and solar panel sensors and amalgamating them with atmospheric data. AI can create scope for an enhanced economic efficiency of energy, which is a crucial step towards sustainability. For instance, Siemens is using AI to alter fuel distribution in turbine burners to promote energy efficiency. Bidgely, a startup, makes use of machine learning algorithms to analyze the energy consumption of domestic appliances and offers detailed insights to inform consumers about their energy usage patterns. British Petroleum is also investing in AI to optimize drilling operations by acquiring relevant data like the flow rate, pressure, and vibration in an area and combining those with the environmental information. What Should We Look Forward To In The Future? Going by the present pace of inclusion of AI in the energy sector, in the next few decades, it is safe to assume that the use of AI technologies will increase by leaps and bounds. Today, many countries are planning to invest in and develop ‘smart grids’ where AI will function as the brain. With the help of AI technologies, these smart grids will gather and synthesize huge amounts of data and help energy industries to find ways to allocate their energy resources in the best possible way. Machine learning and deep learning algorithms will learn the spotting patterns and anomalies within bulky data sets on their own, thus, revolutionizing both demand and supply sides of the energy sector. The evolution of the ‘smart grid’ has been termed as the ‘Energy Cloud.’ Ai will be an integral part of this advanced grid system in the energy sector because as the number of points of control in the grids will keep on increasing rapidly from millions to billions, AI will be the only efficient way out. Atomberg Technologies, an organization dedicated to solving the energy crisis in India, too thinks that AI holds tremendous potential to optimize the allocation of energy resources, predict energy usage, and suggest energy conservation techniques to individuals. Revolutionizing the electronic appliance industry in India, Atomberg hopes to include AI within its own ecosystem to promote innovations and energy efficiency.","excerpt":"Ever since 2016, Artificial Intelligence has been something of a hot topic of conversation among entrepreneurs and the reputed business magnates of the world. In fact, in 2017, the term Artificial Intelligence (AI) was ranked No. 7 on Gartner.com, showing how popular and widespread it had become in just a year. In simple words, AI […]","categories":["IT Services"],"tags":["disruptive technology"],"author_name":"Arindam Paul","publish_date":"2018-03-08T08:13:44","publication_year":"2018","word_count":1186,"keywords":["disruptive technology","Go","API","machine learning","artificial intelligence","AI","chatbots","RAG","deep learning","analytics","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","deep learning","analytics","RAG","chatbots","R","Go","API"],"url":"https:\/\/analyticsindiamag.com\/it-services\/the-future-of-energy-seems-intense-with-the-exponential-development-in-the-ai-sector\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10023320,"title":"Complete Guide to Neural Non-Rigid Tracking","content":"Augmented Reality (AR) and Virtual Reality (VR) applications are growing enormously in the count. These applications rely chiefly on the reconstruction of 2D\/3D images and scenes.  Though there is consistent progress in capturing and reconstruction, this remains one of the challenging tasks in computer vision. Capturing and reconstructing static objects is performed with great accuracy via many architectures. However, capturing and reconstructing dynamic objects is still a domain that needs a solid development. Dynamic object tracking and reconstruction are roughly classified into Rigid object tracking and reconstruction and Non-rigid object tracking and reconstruction. Rigid object tracking assumes a shape prior and tracks that predefined shape anywhere in the given frame. On the other hand, non-rigid object tracking looks only for the predefined characteristics in the given frame but not a fixed shape. Commercial colour-and-depth cameras, commonly called the RGB-D sensors, such as Microsoft’s Kinect and the Intel’s Realsense make real-time non-rigid object tracking deployable. But existing systems require a large array of RGB-D cameras to capture non-rigid dynamic objects and a computationally expensive set up. These limitations hardly give way to the commercialization of real-time non-rigid tracking applications. Aljaž Božic, Pablo Palafox, Angela Dai, Justus Thies and Matthias Niessner of the Technical University of Munich and Michael Zollhöfer of the Facebook Reality Labs have introduced a Neural non-rigid tracker mechanism that is robust in performance and cheaper to deploy in real-world applications. It demonstrates state-of-the-art non-rigid reconstructions by greatly outperforming existing methods. An overview of the Neural non-rigid tracking The proposed Neural non-rigid tracking uses capturing merely from a single RGB-D sensor, thus leads to a cheaper setup. The initial image in the input frame is considered the source, and the successive image is considered its target. The source and the target are advanced temporally as the input capturing streams in. Since the object of interest is non-rigid, it needs a special characteristic called correspondence to map and follow the object. Correspondence is the prediction of a specific source pixel in the target image. Neural non-rigid tracking performs correspondence prediction in a pixel-wise manner followed by correspondence weighting. In correspondence weighting, each predicted correspondence is given a real-valued weight between 0 and 1 so that the tracker can get rid of outliers. This correspondence approach enables the model to perform correspondence mapping at least 85 times faster than the existing methods! Finally, the weighted correspondence map is passed through a differential solver. The differential solver is a self-supervised learning algorithm that learns, rejects outliers and optimizes the architecture to efficiently track the non-rigid objects. The differential solver enables the network to train end-to-end in a novel manner. Thus there is no need for any pre-trained model to learn correspondence or to track non-rigid objects. End-to-end training helps the Neural non-rigid tracking achieve greater performance even with a reduced computational capacity and a single RGB-D sensor. This architecture employs densely connected convolutional neural networks throughout. The end-to-end training strategy in Neural non-rigid network with Correspondence prediction, Correspondence weighting and a Differential solver governed by Correspondence Map Loss, Graph Loss and Warp Loss. Python Implementation of Neural Non-rigid Tracking Download the source code from the official repository to the local machine. !git clone https:\/\/github.com\/DeformableFriends\/NeuralTracking.git Output: Change the directory to refer to the downloaded NeuralTracking directory. %cd NeuralTracking\/ !ls -p Output: Install Anaconda-3 distribution using the following command, if the local machine does not have one. !wget https:\/\/repo.anaconda.com\/archive\/Anaconda3-2020.02-Linux-x86_64.sh !bash Anaconda3-2020.02-Linux-x86_64.sh Using the following command (mentioned below) in base mode, activate the conda environment and build the development environment. Installing the dependencies and activating the environment takes some time. !bash and provide the following inside the base mode, conda env create --file resources\/env.yml Output: The following command inside the base mode runs the setup file in the conda environment and installs C++ dependencies. conda activate nnrt cd csrc python setup.py install cd .. To run the Neural Non-Rigid Tracking model and evaluate it on two frames, execute the following commands. %%bash python example_viz.py If the users wish to train the model from scratch, the officially recommended dataset can be downloaded to the local machine and preprocessed using the following command. %%bash python create_graph_data.py Training can be enabled using the following command. It should be noted that training may take its time based on the memory availability and device configurations. %%bash .\/run_train.sh Once training is finished, evaluation can be performed using the following command. %%bash .\/run_generate.sh Performance of Neural Non-Rigid Tracking Qualitative analysis of Neural Non-Rigid Tracking Neural Non-Rigid Tracking is trained and evaluated on the DeepDeform benchmark. Other competing models, including DynamicFusion, VolumeFusion and DeepDeform, are trained and evaluated under identical conditions and device configurations for comparison. Qualitative comparison of Neural Non-Rigid Tracking with DynamicFusion and DeepDeform models. Qualitative comparison of Neural Non-Rigid Tracking with DynamicFusion and DeepDeform models. Neural Non-Rigid Tracking achieves state-of-the-art performance in non-rigid reconstruction by generating Deformation and Geometry errors lesser than the DynamicFusion, the VolumeFusion and the DeepDeform models at 85x speed! Further reading Original research paperOfficial siteSource code repository","excerpt":"Neural non-rigid tracking mechanism is robust in performance and cheaper to deploy in real-world object-tracking applications","categories":["Deep Tech"],"tags":["Guide"],"author_name":"Rajkumar Lakshmanamoorthy","publish_date":"2021-04-05T13:00:00","publication_year":"2021","word_count":831,"keywords":["Go","TPU","AI","neural network","ML","computer vision","Python","Ray","C++","R","Guide"],"extracted_tech_keywords":["AI","ML","neural network","computer vision","Ray","TPU","Python","R","Go","C++"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/complete-guide-to-neural-non-rigid-tracking\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":41509,"title":"A Day In The Life Of: A CIO Who Is Championing Energy-Efficient Solutions For Modern India","content":"In our column ‘A Day In The Life Of’, we are trying to step into the shoes of the awesome techies from various organisations who are working in Emerging Tech areas like big data, data analytics, artificial intelligence and the internet of things, among others. This week, we decided to talk to Avneesh Vats, the CIO at EESL, about his spiritualism and his dynamic take on the evolving role of the C-suite with respect to emerging technology. Vats begins his day at the crack of dawn at 5 am. His morning ritual includes a cosy chat with his parents, Yoga and meditation. “Morning air is essential for our body and soul. Hence, I start my day with basic Yoga and meditation. It is the best way to have a peaceful mind, it generates clarity in thoughts and helps develop positivity,” he says. Explaining how the role of the C-suite management has changed in India, Vats says that the CIOs are now increasingly concerned with enabling revenue growth and spearheading digital transformation efforts for their organisation. “To remain relevant, they are increasingly realigning their activities with their organisation’s business objectives. They are increasingly expected to guide corporate boards on how to leverage IT strategically to gain competitive advantage. At EESL, we work closely with business teams to define a common set of objectives and success parameters. On the basis of these, we form an alliance with the business department to achieve the objective by delivering an IT-enabled, value-added service,” he explains. Vats in May won the prestigious Analytics100 award for Data-Driven CxO at MachineCon 2019. This award is given to executives who have successfully leveraged analytics for their business benefit. Vats says that one of the things that push him to socialise even with his busy schedule is his friends. “I strongly believe that there is always something to learn from each person you meet in life, therefore, I keep learning while socialising. My friends are scattered in different parts of the cities. Meeting them over dinner and catching up is the highlight and guilty pleasure of almost every tour I go for.” Vats says that his wife is of the most important sounding boards and pillars of support in his life.  “I am blessed to have an extremely understanding and clear-headed wife. She fills into my shoes perfectly whenever required, even when she is a working professional herself,” he says proudly. When asked about his work, Vats says, “Nowadays, an organisation’s decisions are data-driven. They are focussed on implementing the latest technology. It is important for my company and me to make sure that all key persons in the organisation are equipped with data-driven tools. This drives businesses based on digital data, transforms it into information that further helps us to take the right decision at the right time.” His short term goals with EESL are very simple and precise: Creating a vision of what is possible in leveraging technology Aligning technology with business strategies Adopting new solutions Modernising legacy applications Institutionalising DevOps Improving data management\/business intelligence Increasing automation Consolidating applications onto the cloud Sun setting old applications Improving security Talking about his five-year plan for the organisation, Vats says, “India is a country of approximate six lack villages with a huge population where literacy level, expenditure power and awareness of people about energy and environment saving is less. Hence, we are required to put more effort to convince them to push energy efficient appliances for their own benefits. Our EESL Smart Meter National Programme (SMNP) aims to eventually replace 25 crore conventional meters with smart meters across India. Our NDMC project will be completed in 2019… I’m confident that we are on the right track and carry forward the company mission with more vigour and committed to helping in solving energy-related problems that we face as a country.” ALSO SEE:","excerpt":"In our column ‘A Day In The Life Of’, we are trying to step into the shoes of the awesome techies from various organisations who are working in Emerging Tech areas like big data, data analytics, artificial intelligence and the internet of things, among others. This week, we decided to talk to Avneesh Vats, the […]","categories":["AI Features"],"tags":["Big Data","Data Science Career","energy","Interviews and Discussions","meditation"],"author_name":"Prajakta Hebbar","publish_date":"2019-06-28T12:41:07","publication_year":"2019","word_count":641,"keywords":["meditation","Go","big data","energy","artificial intelligence","AI","Git","RAG","Data Science Career","Aim","analytics","DevOps","Big Data","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","artificial intelligence","analytics","Aim","RAG","R","Go","Git","DevOps","big data"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/a-day-in-the-life-of-a-cio-who-is-championing-energy-efficient-solutions-for-modern-india\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":34740,"title":"Separate Ministry For AI? Indian Experts Say That’s The Way To Go Ahead","content":"The global head of robotics and cognitive systems at TCS, Dr Roshy John at the Professional Students Summit at Cochin University of Science and Technology has submitted a representation to Kerala Chief Minister Pinarayi Vijayan seeking a dedicated ministry for Artificial Intelligence in the state. While having a conversation with the Chief Minister on the sidelines of the event, John said that lack of any meaningful allocation for Artificial Intelligence in the Union and state budgets for 2019-20 is a setback. Furthermore, the global head of robotics has also put light on the fact that countries all over the world are setting up separate bodies for artificial intelligence. He also said that UAE has a separate minister in the matter too. “Kerala should not remain a step behind in the matter,” John said. John also pointed out that nearly 70 per cent of the jobs in the upcoming years will be based on artificial intelligence, and policymakers will have to create the right environment to switch over to such a system without much social instability. “The Union government is still in the process of considering setting up National Centre for Artificial Intelligence. Kerala instead of waiting should take a lead in the matter,” John added.","excerpt":"The global head of robotics and cognitive systems at TCS, Dr Roshy John at the Professional Students Summit at Cochin University of Science and Technology has submitted a representation to Kerala Chief Minister Pinarayi Vijayan seeking a dedicated ministry for Artificial Intelligence in the state.   While having a conversation with the Chief Minister on […]","categories":["AI News"],"tags":["AI (Artificial Intelligence)","kerala"],"author_name":"Harshajit Sarmah","publish_date":"2019-02-11T08:34:11","publication_year":"2019","word_count":205,"keywords":["Go","artificial intelligence","programming_languages:R","AI","programming_languages:Go","Ray","ai_applications:robotics","AI (Artificial Intelligence)","R","kerala"],"extracted_tech_keywords":["AI","artificial intelligence","Ray","R","Go","programming_languages:R","programming_languages:Go","ai_applications:robotics"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/separate-ministry-for-ai-indian-experts-say-thats-the-way-to-go-ahead\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10169133,"title":"How This Bengaluru Startup is Making External Hiring Obsolete","content":"Tech is moving fast, and companies are finding it hard to keep their teams upskilled. At the same time, managing the cost of talent is equally challenging. This is the problem Spire.ai is solving. The company has been in the talent space for eighteen years—long before ‘AI’ became the ubiquitous term it is today. “The cost of talent is increasing significantly,” said Saurabh Jain, founder of Spire.ai, in an exclusive interview with AIM. “Organisations need to sustain their costs while still having the room for growth and bringing in new talent.” However, he explained that cost management cannot come at the expense of preparedness. “Our objective is to make organisations completely resilient from a skills and talent point of view,” he said, adding that businesses cannot predict how they would be impacted by technological innovation over the next three to five years. This makes continuous learning and skills development a necessary investment. Organisations face a range of operational issues, from acquiring talent and creating succession plans to growing existing employees. Even when process problems are solved, issues remain with speed, ease, efficiency and cost. “Our solutions are designed to complement existing systems,” Jain said. “Organisations have already made heavy investments in ERP and talent systems. We do not replace them, we add value without disrupting these systems.” Path-breaking Products The company has recently developed Agent Sigma, which gives employees an understanding of what skills they have acquired, identifies gaps and opportunities and tells them about their potential to learn. It also continuously tracks growth opportunities within the organisation. Besides, the company has developed a suite of products which includes: Skills AI Copilot, Empollination, Skilling.Exchange, Continuum.Exchange, Talent Map Live, Talent Exchange, Recruitment.Exchange, and Match as a Service. The first, Skills AI Copilot, focuses on identifying the skills needed in the world today and how an organisation should design its skill framework. Skilling.Exchange provides learning content to employees. It aggregates and curates “the best of the content from the planet”, ranking and rating it for relevance, said Jain. Continuum.Exchange provides a continuous, forward-looking view of succession for every role and individual. It identifies potential growth paths for individuals (at least three opportunities) and potential successors for roles. “It is largely designed for leadership roles,” said Jain. Talent Map Live enables skill-based workforce planning, forecasting of talent needs, and likely other strategic planning related to the organisation’s human capital and Recruitment.Exchange is a “new age recruiting” platform focused on efficiency and candidate experience. Match as a Service is a matching engine that helps large organisations efficiently connect jobs with candidates or internal employees. Tech Underneath Jain said that they are not using LLMs. Instead, they have built their own proprietary model called the large graph model. “Our base technology is called LGM, in which we have 11 million sanitised skill graph nodes, which relate to each other, and cover some 127,000 primary skills across some 27 industries,” said Jain. He added that on top of it, the company has created powerful domain-intelligent search and match capabilities, business rules engines, and different kinds of operating logic. Speaking of the data model, he said that they manually sewed the graph model till 2014. From 2014 onwards, they automated many parts of it. “We put a lot of servers in the cloud in early 2014, and for the last 15 years, this particular technology has been completely scavenging every single data element from the web, whether it’s jobs, job descriptions, research papers, or various kinds of labour market sources.” He explained that through it, the system generates this continuous skill map, which is then created in the form of skill graphs, where every skill and the way of their adjacency is represented in the form of graphs. He said one of their solutions already tracks more than 16,000 companies and is directly linked to their websites. It includes every job posted there and the evolution of jobs. Growing Customer Base Jain shared that some of their long-standing customers include Cognizant and Tata Communications. “We’ve recently started working with Myntra and Digit Insurance.” Speaking of the impact, he said they have tremendously brought down the cost of sourcing by anywhere between 30% and 50% in companies. “The time to recruit has definitely come down by 20% to 40%, depending on how much of our workflows companies have actually used,” he added. Citing an example, he shared that in one of the companies, there was a heavy reliance on external hiring—70% of roles were filled from outside and only 30% internally. Their solution helped reverse this ratio. “Now 70% can be fulfilled internally, and only for 30% you have to look outside. That is the best impact that we have.” Team Size and Future Roadmap Right now, the company has 100 people and is looking to add another 75–80 to the team. “Until two years ago, we were largely focused on the Indian market. But since last year, we have seen significant growth for us in Europe and the US markets. The team is rapidly growing,” Jain said. Between 2014 and 2016, the company secured $13 million in Series A funding and is now profitable. Speaking of the future roadmap, he said that they initially focused on solving demand-side problems — primarily from the enterprise’s perspective — and have now shifted towards addressing supply-side challenges in the talent ecosystem. He hinted that the company is going to launch new products soon. [Note: A previous version of this article mentioned a client’s name, which has since been removed following a request by the company.]","excerpt":"“Now 70% [hiring needs] can be fulfilled internally, and only for 30% you’ll have to look outside. That is the best impact that we have.”","categories":["AI Startups"],"tags":["spire.ai"],"author_name":"Siddharth Jindal","publish_date":"2025-05-05T16:34:19","publication_year":"2025","word_count":925,"keywords":["Go","API","funding","programming_languages:R","AI","innovation","Git","GAN","Aim","spire.ai","R"],"extracted_tech_keywords":["AI","Aim","R","Go","Git","API","GAN","innovation","funding","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-startups\/how-this-bengaluru-startup-is-making-external-hiring-obsolete\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10046669,"title":"Top Announcements Made At Intel Architecture Day 2021","content":"Intel hosted Architecture Day last week. At the third edition of its annual affair, the chip manufacturing giant disclosed several architectural innovations at work. This event saw the company demonstrating some of the products they have been working on, which signify Intel’s biggest shift in Intel architectures in a generation, claimed Raja Koduri, senior VP and general manager at Intel. “The breakthroughs we disclosed today also demonstrate how architecture will satisfy the crushing demand for more compute performance as workloads from the desktop to the data centre become larger, more complex, and more diverse than ever,” Koduri said. Here we detail some of the major announcements made at the event. Efficient-Core Code-named Gracemont, the company’s new x86 Efficient core, would address compute requirements across the entire spectrum of customer needs. In addition, Intel aims to make it the world’s most efficient x86 CPU code. Compared to Intel’s most prolific CPU microarchitecture, Skylake, Efficient cores deliver 40 per cent more single-threaded performance at the same power. Further, Efficient-cores deliver 80 per cent better performance while consuming less power than two Skylake cores running four threads. Performance-Core This x86 core is the highest performing CPU core from Intel, Koduri claimed. “It also delivers a step function in CPU architecture performance that will drive the next decade of compute,” he further added. It is designed to have a wider, deeper, and smarter architecture that can fit in more parallelism, increase execution parallelism, increase performance, and reduce latency. Performance-core provides 19 per cent Geomean improvement and can support large data and code footprint applications. Performance-core has dedicated hardware, including new Advanced Matrix Extensions (AMX); it is best for machine learning applications as it offers an 8x increase in artificial intelligence acceleration. Alder Lake The company has been teasing the features of Alder Lake since last year’s Architecture Day. As per the announcement at this year’s event, Alder Lake will be Intel’s first performance hybrid architecture with the new Intel Thread Director. It is the first chip from Intel, based on its Intel 7 Technology which uses the same tech as the company’s current 10nm technology. Alder Lake will offer a mix of both performance and efficiency of x86 cores. Products based on Alder Lake will start shipping this year. Thread Director One of the major roadblocks that Intel has faced regarding its hybrid processor architecture designs is managing multiple threads most efficiently. Additional analysis is required when there are two cores of different performance and efficiency. To remediate this, Intel has announced Thread Director. It is a unique approach for scheduling that ensures Efficient-cores and Performance-cores work together in intelligently assigning workloads from the beginning and optimising the system for superior real-world performance. The Thread Director is built directly into the core, and it works with the operating system to place the right thread at the right time. Ponte Vecchio and Alchemist SoCs Ponte Vecchio is, by Intel’s admission, a more complex SoC. It is enabled by Intel’s open, standard-based, cross-architecture unified software stack, OneAPI and built on technologies such as Intel’s advanced semiconductor processors, EMIB technology, and Foveros 3D packaging. With Ponte Vecchio, Intel is introducing the 100 billion-transistor device that delivers industry-leading FLOPs and compute density to carry out applications on artificial intelligence, high-performance computing, and advanced analytics workload. Further, Ponte Vecchio has broken the industry record of inference and training throughput on a popular AI benchmark. Another major SoC announced at the event was Xe HPG-based Alchemist SoCs (formerly code-named DG2). In addition, Intel announced Xe HPG, a new discrete graphics microarchitecture designed for gaming and creation workloads. Other announcements The introduction of Sapphire Rapids for next-generation data processors was among other announcements. It combines Intel’s Performance-cores with new accelerator engines and consists of a tiled SoC architecture. Its USP is that it can deliver significant scalability while offering the benefits of a monolithic CPU. Intel also introduced Mount Evans, its first dedicated ASIC-based infrastructure processing unit (IPU), and Oak Springs Canyon, a new FPGA-based IPU-based architecture.","excerpt":"This event saw the company demonstrating some of the products they have been working on, which signify Intel’s biggest shift in Intel architectures in a generation.","categories":["AI Trends"],"tags":["Intel"],"author_name":"Shraddha Goled","publish_date":"2021-08-25T13:00:00","publication_year":"2021","word_count":666,"keywords":["Rapids","API","machine learning","artificial intelligence","AI","innovation","Scala","Aim","analytics","R","Intel"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","analytics","Aim","Rapids","R","Scala","API","innovation"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-announcements-made-at-intel-architecture-day-2021\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10042657,"title":"Massive Analytic Invests £1 Mn in India To Build CoE In AI &#038; ML, Steps Up Hiring","content":"The London based precognition AI company Massive Analytic has  announced a £1 million investment in India to expand its employee base and build a Centre of Excellence in Artificial Intelligence and Machine Learning. The company will hire for positions including product owners, business and quality analysts, DevSecOps engineers, data scientists, and developers specialising in Scala\/Spark, JavaScript (full stack), Python, Rust, etc. George Frangou, Massive Analytic founder and Group CEO said, “The last year has seen the Massive Analytic’s team grow from strength to strength. We have appointed a number of senior team members around the globe, and grown our presence in the US, UK, Europe, and now India. It’s an exciting time in our company’s history as we grow our client base across the built environment, aerospace and defence, medtech and more, and I’m continually impressed at the high level of talent we are able to attract.” The scaling up of Massive Analytic’s footprint in India will be led by Lead Big Data Analytics Engineer Pankaj Arora, who joined the company in June 2020 to manage the development of front end, back end, and DevOps stacks. Pankaj said, “This injection of capital into Massive Analytic in India will allow us to further develop our footprint in New Delhi, not only increasing our pace of hiring, but also focusing on investing more into our existing team. Training and development has always been a priority, and this will allow our team to continue to stay on the cutting edge of artificial intelligence and machine learning.” The Massive Analytic India team will focus on projects, primarily in health and safety. This will be instrumental in assisting the company to take a leading role in global medical diagnostics. Additional senior hires in computer vision are expected towards the end of 2021. Check out the Massive Analytic page to apply for open positions in Bengaluru, India.","excerpt":"The London based precognition AI company Massive Analytic has  announced a £1 million investment in India to expand its employee base and build a Centre of Excellence in Artificial Intelligence and Machine Learning. The company will hire for positions including product owners, business and quality analysts, DevSecOps engineers, data scientists, and developers specialising in Scala\/Spark, […]","categories":["AI News"],"tags":["AI jobs in India","Hiring"],"author_name":"Debolina Biswas","publish_date":"2021-06-30T11:12:27","publication_year":"2021","word_count":311,"keywords":["Go","AI jobs in India","artificial intelligence","machine learning","Rust","AI","Hiring","computer vision","Python","analytics","JavaScript","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","computer vision","analytics","Python","R","JavaScript","Go","Rust"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/massive-analytic-invests-1-mn-in-india-to-build-coe-in-ai-ml-steps-up-hiring\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":40656,"title":"IIIT Delhi Offers Full-Time MTech To Navy officials In Artificial Intelligence","content":"The IIIT Delhi this week signed a memorandum of understanding with the Indian Navy, where they will provide training in artificial intelligence to the Naval officers. According to a report in a national daily, these officers will be encouraged to pursue Mtech with a specialisation in AI from IIIT Delhi. Reportedly, the officers will be selected for the programme via a process which has been mutually agreed upon both by the Indian Navy, as well as IIIT Delhi. The institute would also be creating supernumerary seats for the Indian defence personnel. We are elated to announce that we have signed a Memorandum of Understanding with the @INDIANNAVY1 for academic collaboration. The MoU will facilitate M Tech programme for Naval Officers with specialisation in Artificial Intelligence. pic.twitter.com\/VAYflq1qWw — IIIT Delhi (@IIITDelhi) June 11, 2019 A Naval officer told the newspaper, “In line with multistakeholder Task Force report on ‘Strategic Implementation of AI for National Security and Defence’ and Govt of India\/Ministry of Defence order on implementation of AI in Armed Forces, IN has been constantly pursuing Capacity Building as well as Projects in the field of BDA and AI. In furtherance of these initiatives and to enable expertise to build up in this niche area, an MoU with IIIT Delhi has been concluded to facilitate M Tech programme for Naval Officers with specialisation in AI.” Earlier this year, the Naval Science and Technological Laboratory (NSTL) had organised the NSTL – Academia Meet with the theme Artificial Intelligence for Naval Systems. A gathering of the academia and the Navy had proved to be an ideal platform for deliberation and discussion on AI and its application in defence systems, with specific emphasis on naval systems. Milind Kulshreshtha a C4I expert, had said, “AI will help in identifying potential threats unambiguously and immediately. It will also help the command team make informed decisions faster and rapidly detects and evaluate potential threats removing anomalies.” He also added that AI is inherently designed to integrate itself into existing warship equipment suits to optimise their exploitation. However, there are things that can stand in the way of technical progress in the Indian Navy: Lack of Data science talent internally Lack of AI initiatives for the Navy Undeveloped eco-system for enterprise-level exploitation","excerpt":"The IIIT Delhi this week signed a memorandum of understanding with the Indian Navy, where they will provide training in artificial intelligence to the Naval officers. According to a report in a national daily, these officers will be encouraged to pursue Mtech with a specialisation in AI from IIIT Delhi. Reportedly, the officers will be […]","categories":["AI News"],"tags":["AI (Artificial Intelligence)"],"author_name":"Prajakta Hebbar","publish_date":"2019-06-13T07:35:27","publication_year":"2019","word_count":374,"keywords":["data science","Go","API","artificial intelligence","programming_languages:R","AI","programming_languages:Go","RAG","GAN","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","data science","RAG","R","Go","API","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/iiit-delhi-offers-full-time-mtech-to-navy-officials-in-artificial-intelligence\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10042283,"title":"Redefining Quantum Computations Using Classical Computers","content":"“The most important application of quantum computing in the future is likely to be a computer simulation of quantum systems because that’s an application where we know for sure that quantum systems, in general, cannot be efficiently simulated on a classical computer” – British physicist David Deutsch. Going by David’s statement, the Institute for Quantum Computing (IQC), in collaboration with the University of Innsbruck, has proposed solving complex computing problems using measurement-based algorithms within a feedback loop with a regular computer. This novel computing method serves as a key tool in bringing academics to the forefront of developments in quantum computing, allowing new algorithms and experiments to take researchers much closer to commercial applications and discoveries of the technology. Christine Muschik, Principal Investigator at the Institute for Quantum Computing (IQC), believes that quantum computers of the future can be used in applications such as removing carbon dioxide from the atmosphere, developing artificial limbs, and designing more efficient pharmaceuticals. Researchers from IQC and @uniinnsbruck have discovered a new and more efficient computing method for pairing the reliability of a classical computer with the strength of a #quantum system. pic.twitter.com\/O2zDUzZwj9— Institute for Quantum Computing (IQC) (@QuantumIQC) June 15, 2021 Source : Measurement-Based Variational Quantum Eigensolver. Physical Review Letters. 126. 10.1103\/PhysRevLett.126.220501 No need for quantum gate-based computer The study by Muschik’s team is focused on quantum calculations that do not require quantum gate-based computers. The team designed an algorithm to carry out a hybrid quantum-classical computation by combining a sequence of measurements on an entangled quantum state with a quantum-classical computation. To use the principles of measurement-based quantum computation, they developed a new approach to Variational Quantum Eigensolver (VQE) Ideally, existing VQE protocols based on circuit models use gates that are applied on an initial state. To obtain an output state, as close as possible to the target state, the variational parameters must be optimised. The new approach to VQE protocols is based on a measurement-based model of quantum computation (MBQC). MBQC can prepare an entangled state and then perform single-qubit measurements to realise the computation. Quantum computing is made possible through both circuit-based and measurement-based models. While both are similar in terms of resource scaling, there is one difference. The circuit-based model’s capabilities are constrained by the number of available qubits and gates that can be executed. In comparison, the needed coherence times and error thresholds are somewhat relaxed for MBQ. The new variational technique based on MBQC is called the measurement-based VQE (MBVQE). The researchers’ new protocols determined the ground state of a target Hamiltonian, a prototypical task for VQEs with wide-ranging applications. The fundamental concept for this was to employ a customised entangled state, called a ‘custom state’. It enabled exploration of a particular corner of the system’s Hilbert space. This custom state contains auxiliary qubits that, when measured, change the output qubits’ states. A traditional optimisation approach is used to control the measurement bases and the resulting variational changes in the state. This approach is conceptually and practically distinct from normal VQE schemes. After introducing the MB-VQE framework, researchers designed two measurement-based VQE techniques. Firstly, they demonstrate a method for constructing variational state families using the toric code model with local perturbations. Second, they demonstrated a direct conversion of circuit VQEs to MB-VQEs. The variational state family is the same for the circuit and measurement-based techniques, but the implementations are distinct because the MBVQE requires more resources and is modified only via single-qubit measurements. While MB-VQE is platform-independent, it enables complicated quantum calculations in systems with lengthy gate sequences, or the realisation of entangling gates is difficult. This theoretical research offers a new way of thinking about optimisation algorithms. MB-VQE in particular expands the toolset of variational computing by providing additional paths for experiments with photonic quantum systems. Researchers no longer have to deal with fussy and sensitive resources, allowing them to construct feedback loops specific to the datasets their computers are researching. Access the complete research here.","excerpt":"“The most important application of quantum computing in the future is likely to be a computer simulation of quantum systems because that’s an application where we know for sure that quantum systems, in general, cannot be efficiently simulated on a classical computer” – British physicist David Deutsch.  Going by David’s statement, the Institute for Quantum […]","categories":["AI Features"],"tags":["quantum application development system","Quantum Computing"],"author_name":"Ritika Sagar","publish_date":"2021-06-24T14:00:00","publication_year":"2021","word_count":659,"keywords":["Quantum Computing","Go","TPU","programming_languages:R","AI","programming_languages:Go","BERT","llm_models:BERT","quantum application development system","R","emerging_tech:quantum computing"],"extracted_tech_keywords":["AI","TPU","R","Go","BERT","llm_models:BERT","programming_languages:R","programming_languages:Go","emerging_tech:quantum computing"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/redefining-quantum-computations-using-classical-computers\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":13488,"title":"How to transition your career from engineering to data science","content":"AIM gets top experts who weigh in on how to make a transition from engineering to data analytics Undoubtedly, transitioning from engineering to data science is one of the trickiest transitions in the most sought after field. Taking a plunge from software engineering role to data scientist\/analyst is fraught with challenges, that too after having spent a decade in the industry. Where the two roles differ are that data analysis requires a statistical bent of mind and reasoning. AIM caught up with Harish Subramanian, Program Director, PGP- Big Data Analytics, Great Lakes Institute of Management on the roles a software engineer can explore in data science. “Data scientists (in the pure sense of the word) are usually only a small part of data science and analytics teams. Correspondingly, building new models is a relatively small portion of the overall analytics function. Everything from identifying, sourcing and managing data, to working the technology stack to make the intricate models work effectively – these are all roles that software engineers with an adaptable and curious mind are perfectly placed to succeed in,” he shares. To make a successful switch, one has to have a basic grounding in statistics such as Median Standard, Correlation and Linear Regression & Deviation before deep diving into modelling. Often, statisticians and mathematicians with training data analysis techniques such as graphing, plotting, analysis and hypothesis find it easy to transition to data scientist role. William Chen, data scientist at Quora and an avid writer about data science has emphasized that “data science is purely about statistics. Statistics will guide you to be able to understand uncertainty in data and pull valid insights from data.” One of the most common career paths for software engineers or programmers who are known for working on software architectures is usually product management or system engineering. Here’s how you can beef up on basic data analyst skills: Learn Hypothesis generation and analysis through plots, graphs, and reasoning Get up to speed with statistical reasoning concepts such as causality and probability theory. To get started, check out this link Data science is also about products, learn how to leverage data to figure out product features, enhancements Data munging is the art of cleaning data. It is a time consuming job and entails dealing with missing data and changing schema. Before diving into large datasets, explore cleaning data While Kaggle is hailed as a stepping stone for honing machine learning and data analysis skills, the competition features curated datasets that are anonymized and cleaned. Understanding the business domain one wishes to work in. The data analysis problems are solved according to the business needs, hence domain understanding is a must Data science is a long-learning process. Switching to data engineering and learning statistics on your own can be one learning path towards a deeper learning experience Analytics India Magazine gets in industry experts to weigh-in on the raging topic and lay down steps to effectively transition from software engineering to data science: A switch from one industry to another is always a challenging move, especially if the years of experience is high, such as 10 years. The aspiring data analytics\/scientist should focus on these aspects: From statistics to programming skills and domain understanding, one must shore up on all fronts Sharpening business domain understanding According to Rohit Sharma, UpGrad’s Program Director, data professionals are ultimately business problem solvers. They need to understand what problem they are trying to solve and what business levers they can move to make sure that they achieve the expected output. For e.g. let us say an analyst is asked to reduce the churn of customers from a retail bank. In that case, the analyst would need to know the types of products that exist within a bank and how customers engage with those products. Harish Subramanian, Program Director, PGP- Big Data Analytics, Great Lakes Institute of Management emphasizes it’s always important to understand what problem you’re trying to solve. Analytics techniques are a means to an end. And understanding the most important challenges for a business is vital to selecting appropriate techniques. For example, a missed prediction of fraud is significantly more expensive to the financial firms than a corresponding wrong prediction of email open rates is to a marketing department. Understanding this will help you make suitable trade-offs between complexity, effectiveness and cost of implementing various analytical techniques. It’s not always a question of picking the most cutting edge solution. Good decision making calls for picking the right tool\/technique for the job. Shoring up Mathematical\/Statistical knowledge When it comes to statistical knowledge, Subramanian believes statistics, probability and mathematics in general is often the most daunting prospect for people entering data science and analytics as a field. The amount of statistical knowledge needed to do effective analysis doesn’t take an advanced degree to master. Linear algebra, metrices, statistical tests, distributions, likelihood estimators, regression, the Bayes theorem and conditional probability are all you need to get started. There isn’t a whole lot of merit in learning very advanced statistics until you start working with lots of data, and hit a ceiling in terms of efficacy of your models. Sharma from UpGrad elaborates how once the problem has been identified, statistics becomes important to make sure that the analysis and the decision making is objective. Mathematics\/Statistics power all the algorithms which are used to quantify the impact of variables under analysis; they identify hidden patterns and also make predictions or recommendations. In the customer churn case, the analyst may use logistic regression to predict which customers are likely to churn by statistically quantifying the impact of various factors (such as account balance, number of credit cards etc.)  that have led to a churn of customers in the past. Building up technical know-how Since organizations are dealing with millions of data points, they need technological solutions which can help them apply the algorithms at scale. Here, tools like R, Python etc., become extremely important. The analyst in this case would use tools like R to apply logistic regression to all the customers in the bank’s database so as to identify potential churners, Sharma from UpGrad shared. Subramanian believes though there are hundreds of tools and packages that help you master various facets of analytics, you can get started with relatively few tools. A statistical programming language like R or Python and a database querying language like SQL are good enough to start with. One would be surprised at how much analytics you can get done with Excel alone, but the limitation is usually in the formats and characteristics of data you can work with in Excel. Gaining expertise in analytics According to Kamal Das, VP, Program Management, Jigsaw Academy, a switch from one industry to another is a challenging move, one that requires patience and building up technical expertise.  He shared his views with AIM on how to make a switch up the ladder with these steps. Train from the leaders in analytics and good brand names with quality training Like other areas, you will have cheaper clones that offer cheaper products but at lower quality- don’t save pennies (on course) to lose dollars (on career) Check to see that the peer group is similar to yours (the tone of teaching for a class with average experience of 3 years is different from one with average experience of 10 years) Showcasing your newfound competencies in analytics Das believes competitions like Kaggle offer a perfect platform for getting started with working on huge datasets. Besides, Capstone case studies help in tackling real world problems. Focus on how to make the pivot to analytics without losing on the wealth of experience you have Show your depth by writing blogs\/articles on your own site, as well as on forums like LinkedIn etc Work on projects – The course should have a capstone project with industry. You may also reference sites like Kaggle According to Subramanian, one must show they are competent in analytics, not just tell. “When competing with thousands of people who all claim to know the same things you do, it is hard to distinguish yourself if all you have to say is “I know x, y and z”. You need to be able to show that you have learned to work with data, written code, cleaned and processed datasets and tuned models to improve their effectiveness,” he shares. He adds a slew of data science competitions that help in demonstrating analytical chops. Kaggle challenges are a great place to build this portfolio of work, as are HackerEarth, Analytics Vidhya and DataCamp challenges. Even if you’re not working on these formal challenges, you’d be well served to upload your dataset, code snippets and model outputs (as well as brief description of what you did and why) on Github. More often than not, smart data science teams ask for your Github account to evaluate your proficiency. Using your network to find a career in analytics Finally, it may require patience, networking and showcasing cases\/data sets that one has worked on to help convince people of your depth and strength in analytics, Das advises on how to maximize the networking opportunities. The course should have industry experts and networking opportunities Change your CV and cover letter to show your analytics profile Leverage and grow network in analytics by showcasing your knowledge Career transitions aren’t easy. Remember, you don’t have to unlearn everything. Subramanian advises, “To make any successful transition, you’re likely to succeed if you build on your existing knowledge. So, if you’re proficient as a programmer, then transitioning into data engineering roles allows you to use your proficiency rather than start from scratch. Similarly, if you’re proficient in databases and data warehouses, then you’re well positioned to move into data architecture roles”. We give you a recap of technical and non-technical skills to beef up on to make a head-start in the data-intensive field. Dealing with unstructured data, familiarity with Hadoop platform, acing modelling language such as R, Python and querying languages such as Pig, Hive and SQL and lastly statistics. Communication skills and an innate curiosity will go a long way in optimizing products and services.","excerpt":"Undoubtedly, transitioning from engineering to data science is one of the trickiest transitions in the most sought after field. Taking a plunge from software engineering role to data scientist\/analyst is fraught with challenges, that too after having spent a decade in the industry. Where the two roles differ are that data analysis requires a statistical […]","categories":["AI Highlights"],"tags":["data science career path","data scientist career path"],"author_name":"Richa Bhatia","publish_date":"2017-03-15T05:03:37","publication_year":"2017","word_count":1692,"keywords":["data science","data scientist career path","machine learning","TPU","AI","RAG","Python","Aim","analytics","SQL","R","data science career path"],"extracted_tech_keywords":["AI","machine learning","data science","analytics","Aim","RAG","TPU","Python","R","SQL"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/transition-career-engineering-data-science\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10164684,"title":"Anthropic Cracks the Code with Claude 3.7 Sonnet","content":"For weeks, it has been a running joke that Anthropic is stuck in a cycle of releasing blogs and research reports while its competitors sprint ahead with innovative AI models. Now, the company has finally released a new version of Claude – the 3.7 Sonnet. Despite the questionable nomenclature, the jump from 3.5 to 3.7 and the decision to skip 4.0, users embraced its coding capabilities in no time. People were actively engaged in building fun games, animations, user interfaces, and other such projects. One user on X effectively summed up the overall sentiment. *openai releases a model*literally beats every existing benchmark, sith dark lord vibes, ASI timeline accelerated*claude releases a model*plays pokemon, happy vibes, everyone starts vibecoding— atlas (@creatine_cycle) February 24, 2025 Mckay Wrigley, founder of the AI-based upskilling platform Takeoff AI, said on X, “Claude 3.7 Sonnet is the best model in the world for code.” Even on benchmarks, the model tops the list. It scored 62.3% accuracy on the SWE-bench while OpenAI’s o3-mini (high) scored a 49.3%. Artificial Analysis, a platform that independently analyses AI models, called it the best non-reasoning model for coding. Besides benchmarks and first impressions, users were quick to build several projects. Deedy Das, principal at Menlo Ventures, built an app for the popular board game Connect 4 using Claude 3.7 Sonnet and said that the model could write around 5,000 lines of code in just 30 minutes. “It is the closest thing to AGI (Artificial General Intelligence) I’ve seen,” he said. Notably, Menlo Ventures is an investor in Anthropic. In another instance, Ethan Mollick, a professor at The Wharton School, threw a challenge at the model, asking it to create a sketch of a control panel of a ‘futuristic spaceship’ using p5.js, a JavaScript library for creative coding. He declared that Claude 3.7 Sonnet was the winner. 👀Here is Claude 3.7 on my long-standing challenge: \"create something I can paste into p5js that will startle me with its cleverness in creating something that invokes the control panel of a starship in the distant future\"(every other model is in the quoted tweets, not close) https:\/\/t.co\/LPQOyUpdBP pic.twitter.com\/LLxCam5ldF— Ethan Mollick (@emollick) February 25, 2025 “Honestly, the gap here is pretty insane, even compared to the o1 models and Grok 3. The dashboard was fully interactive as well; no other model came close,” he said. In another instance, Derek Nee, an AI engineer and CEO at flowith, compared Claude 3.7 Sonnet with models like OpenAI’s o1, DeepSeek-R1, and Claude 3.5 Sonnet in a task to write a Scalable Vector Graphics (SVG) code for a book cover of a science fiction book. In his evaluation, the 3.7 Sonnet created the most visually pleasing image. Nee said that it crushes other models. AIM also tested the model by redesigning the Hacker News homepage using Apple’s Human Interface guidelines. In just two iterations, we were able to build an interactive website with front-end libraries. Even OpenAI Agrees Anthropic is a Better Coding Model Anthropic has earned a reputation for excelling in code-based tasks. This isn’t just a claim from the company or its fans. Recently, even its competitor, OpenAI, publicly acknowledged that it lags behind Anthropic in this area. OpenAI introduced a benchmark called SWELancer to test whether AI models can successfully complete real-world software engineering tasks on Upwork. The benchmark comprised over 1,400 tasks across various aspects of software development. The results revealed that Claude 3.5 Sonnet performed better than GPT-4o and the o1 reasoning model in several tasks. That said, the Sonnet 3.7 model isn’t free from criticism. It is still very expensive to use, exponentially more than OpenAI’s o3 Mini. The Claude 3.7 Sonnet costs $3 per million input tokens and a whopping $15 per million output tokens. OpenAI’s o3 Mini, which is comparable to Claude 3.7 Sonnet on benchmarks, costs $1.1 per million input tokens and $4.40 per million output tokens. Jeremy Chone, a YouTuber who teaches programming, said on X that Sonnet 3.7 “struggles with instructions”. He added that it tends to deviate from recommended coding practices, as it creates separate coding files in Rust. 3.7 Sonnet Available on All Popular AI Coding Tools Sonnet 3.7 excels at coding but doesn’t rank well as a general-purpose model overall. Furthermore, users already have access to AI tools dedicated to coding, like Cursor and Windsurf, so it raises the question of what Claude seems to achieve here. However, AI models like Claude are still the foundational layer for these coding tools, and nearly every popular platform has already integrated the 3.7 Sonnet. The model is now available on Replit Agent, GitHub Copilot, Cursor, Windsurf, and many other platforms. Cursor, while announcing the new model’s availability on its platform, said, “We’ve been very impressed by its coding ability, especially on real-world agentic tasks. It appears to be the new state of the art.” However, these tools face an incoming threat from Anthropic. Along with the 3.7 Sonnet, the company also launched an ‘agentic’ coding tool called Claude Code. This tool functions as an active collaborator that can read code, edit files, commit, and push code to GitHub. The tool is currently available under research preview. “In early testing, Claude Code completed tasks in a single pass that would normally take over 45 minutes of manual work, reducing development time and overhead,” the company said. It will be interesting to see how a coding agent built on a foundational model takes on successful wrappers like Cursor, Windsurf, or even Devin.","excerpt":"Some are even calling it the best model for coding.","categories":["Global Tech"],"tags":["AI (Artificial Intelligence)","Claude"],"author_name":"Supreeth Koundinya","publish_date":"2025-02-26T18:00:00","publication_year":"2025","word_count":912,"keywords":["Anthropic","Grok 3","TPU","OpenAI","AI","GPT-4o","Claude","Aim","JavaScript","Claude 3.5","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","GPT-4o","OpenAI","Claude 3.5","Anthropic","Grok 3","Aim","TPU","R","JavaScript"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/anthropic-cracks-the-code-with-claude-3-7-sonnet\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":37351,"title":"How Surgical Robot Assistants Are Becoming A Reality In Indian Hospitals And Healthcare Sector","content":"Image source: WPA Pool \/ Getty Images As India prepares itself for Industrial revolution 4.0, the adoption rate of emerging technologies like artificial intelligence, robotics, analytics and IoT has increased many-fold. According to a 2018 study by a leading job portal, India witnessed a spike in the number of job seekers in the field of robotics by 186 per cent, whereas, job postings in the segment stood at 191 per cent. The study states that the initial impetus for this demand was driven by a number of factors which includes the government’s push for Make In India, which saw the government investing a whopping $13 billion in robotics. Interestingly, among the industries which saw great adoption of the technology, medical and construction sector witnessed more adoption of robotics when compared to others. “Robotics is one of the most exciting sectors emerging in India today. Sectors like construction, manufacturing and healthcare are boosting the demand for robotics talent, and there is enormous scope for the application of technology across an even wider array of industries,”  Sashi Kumar, Managing Director, Indeed, said. Robotics In The Indian Medical Sector Currently, in India, a key area within healthcare where robotics have been deployed is surgery. Here, robotics assistants are being used across private and public hospitals throughout the country. Interestingly enough, the first use-case of deploying a robotic surgical assistant in India dates back to 1998, when a team of doctors from Escorts Heart Institute and Research Centre (EHIRC) used a robotic arm to fix a hole in the patient’s heart. Ever since the number of cases of assisted surgery has gone substantially. In India, however, another revolution in the Indian medical industry was the introduction of IBM’s AI and cognitive powered bot Watson. Since its launch in 2015, IBM has partnered with private medical institutes like Manipal Hospital to deploy the solutions.  The cognitive system has turned out to be a resourceful tool for surgeons with its deep insight, NLP capabilities and with its ability to analyse electronic medical records. Owing to its high applicability in the field, according to another study, India’s surgical robotics market is estimated to grow by five times within 2025 will be driven by the heightened demand for automation solutions. As per the study, the robotics market in India is expected to grow at a CAGR of 20% between 2017 and 2025 to hit the size of $350 million. Top Robotics Solution Providers In India In India, it is estimated that there are over 50 robotics solution providers in the healthcare sector: ASIMOV Robotics Pvt Ltd: is a single window Solution and Services provider who can meet all robotics\/automation needs. They provide engineering products solutions and consultancy in  areas like robotic simulation and control, machine-vision, training, virtual reality, and navigation applications. Their Medical Robotics offer includes Surgical Robots, Non\/Minimally Invasive surgery Robotic Systems, Prosthetics and Exoskeletons, Healthcare robots which offer a smart and cost-effective solution for assistance and rehabilitation procedures for patients, Non-Medical Robotics which includes Robotic Bed-Side monitoring and Emergency Response robotic systems. Intuitive Surgical: Are the creators of the famous da Vinci Surgical System which is most commonly used in assisted robotics surgeries across the world.  In their four decades of functioning, the company offers an array of robotics products designed for varying surgical requirements. Their robotic-assisted platforms and services offer doctors and hospitals to deliver the best outcome. Stryker:  Stryker is a US-based company which offer innovative products and services in Orthopaedics, Medical and Surgical, and Neurotechnology and Spine that help improve patient and hospital outcomes. Top Instances When India Created Medical History Robotics arms for complex coronary bypass: In 2002, Dr Naresh Trehan of Delhi’s Escorts Heart Institute and Research Centre (EHIRC) used a robotic arm with an endoscopic camera attached that could provide a 3D imaging of the organs.  The arm composed of three chopstick-like black arms with pivots and pistons connected with cables. For intestinal surgery: In a unique case, Dr Apurva Vyas of  Sterling Hospitals in Ahmedabad performed robotic surgery on a 67-year-old patient with Median Arcuate Ligament Syndrome, making it the first-ever in India and third in the world. “Using a four-armed surgical robot, I was able to reach the root of the celiac artery where human fingers could not. The three-dimensional high definition view magnified the artery multi-fold to help me achieve precision that prevented collateral damage to healthy tissue, arteries and nerves,” the doctor told a leading business portal. Heart Surgery 32km away: With the help of advanced robotics, Gujrat-based acclaimed cardiac surgeon Dr Tejas Patel, performed robotic surgery on a middle-aged with blocked artery while sitting 32km away from the patients, making it the first-ever known case of performing surgery from a remote location, “This procedure could have been done with a 20mbps connectivity speed as well. I believe this will not just transform coronary ( ) intervention, but the entire vascular space. It has the capability of changing the lives of millions of people living in rural areas,” the doctor said For knee-replacement: In yet another successful case of robotic surgery, surgeons from Chennai-based SIMS Hospital performed a knee-replacement surgery using a robotic arm. Speaking about its success, Vijay Bose, the orthopaedic surgeon who headed the operation said,“It is just an arm that I hold while operating. It alerts me when there’s a tiny variation in the angle that I operate on.” Conducting over 500 robotic surgeries: Kochi-based Amrita Institute of Medical Sciences recently celebrated a hallmark moment when the doctors performed its 500 robotics surgery in the hospital. The hospital’s department of gynaecological oncology adopted robotics in 2015 and has performed various gynaecological cancers and for other conditions, “From the beginning, the patients undergoing robotic surgery showed benefits like faster recovery, less pain and less blood loss and fewer complications. The cost of surgery was also kept very low at this institute so that the common patients could benefit from this surgery,” Dr Beena, Additional Medical Superintendent, AIMS said.","excerpt":"As India prepares itself for Industrial revolution 4.0, the adoption rate of emerging technologies like artificial intelligence, robotics, analytics and IoT has increased many-fold. According to a 2018 study by a leading job portal, India witnessed a spike in the number of job seekers in the field of robotics by 186 per cent, whereas, job […]","categories":["AI Features"],"tags":["AI and robotics","Automation","Robotics"],"author_name":"Akshaya Asokan","publish_date":"2019-04-05T06:55:12","publication_year":"2019","word_count":994,"keywords":["Go","artificial intelligence","AI","Automation","Robotics","AI and robotics","NLP","Aim","Ray","ViT","analytics","GAN","R"],"extracted_tech_keywords":["AI","artificial intelligence","NLP","analytics","Aim","Ray","R","Go","GAN","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-surgical-robot-assistants-are-becoming-a-reality-in-indian-hospitals-and-healthcare-sector\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10123230,"title":"Microsoft Rolls Out VALL-E 2, Attains Human-Level Speech Synthesis","content":"Building on the success of VALL-E, Microsoft has introduced VALL-E 2, a neural codec language model designed to achieve human-level performance in zero-shot text-to-speech (TTS) synthesis. This model rolls out two new features called Repetition Aware Sampling and Grouped Code Modeling to improve the stability and efficiency of the speech synthesis process. Let’s take a look at the new methods. Repetition Aware Sampling: This method refines the traditional nucleus sampling by considering token repetition in the decoding history to improve stability and prevent the infinite loop issues encountered in earlier models. Grouped Code Modeling: This technique organises codec codes into groups to reduce sequence length, thereby speeding up inference and addressing the challenges associated with long sequence modelling. These innovations enable VALL-E 2 to synthesise speech with high accuracy and naturalness, even for complex sentences. The model requires only simple speech-transcription pair data for training, simplifying the data collection and processing. The model has been evaluated on the LibriSpeech and VCTK datasets, demonstrating superior performance in speech robustness, naturalness, and speaker similarity compared to previous systems. It is the first model to achieve human parity on these benchmarks, producing high-quality speech for complex and repetitive sentences. Read the full paper here. What Makes VALL-E 2 Better In January of 2023, the company had come up with VALL-E which demonstrated in-context learning capabilities in zero-shot scenarios after being pre-trained on 60,000 hours of English speech data. However, it faced issues with stability and efficiency. VALL-E relied on random sampling, which could lead to unstable outputs, and its autoregressive architecture resulted in slow inference speeds. Follow-up works have tried to address these problems by leveraging text-speech alignment information and non-autoregressive methods, but these approaches introduced new complexities and limitations. The capabilities of VALL-E 2 can be particularly beneficial for generating speech for individuals with speech impairments, such as those with aphasia or amyotrophic lateral sclerosis. While the new model has significant potential, it also carries risks of misuse, such as voice spoofing or impersonation. The model assumes user consent for voice synthesis. In real-world applications, it should include protocols for speaker approval and detection of synthesised speech to prevent abuse.","excerpt":"This model rolls out two new features called Repetition Aware Sampling and Grouped Code Modeling to improve the stability and efficiency of the speech synthesis process.","categories":["AI News"],"tags":["Microsoft","text to image"],"author_name":"Shritama Saha","publish_date":"2024-06-11T15:03:02","publication_year":"2024","word_count":358,"keywords":["TPU","programming_languages:R","AI","R","innovation","RAG","GAN","text to image","in-context learning","Microsoft"],"extracted_tech_keywords":["AI","RAG","in-context learning","TPU","R","GAN","innovation","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/microsoft-rolls-out-vall-e-2-attains-human-level-speech-synthesis\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":17687,"title":"How To Structure And Choose The Right Team-Type While Setting A Data Science Team?","content":"Acknowledging that data exists and can do wonders is passé. Organisations today need to do much more than just identify big data. With the shortage of data scientist skills worldwide, it is difficult for an organisation to fulfil its dream of capitalising on data science, as it is hard to find the exceptionally skilled person – who is a machine learning expert, a data engineer, a developer, a storyteller and a business analyst. A data science team, if carefully built, with the right set of professionals, will be an asset to any business. It’s a fact that the success of any project is dictated by the expertise of its resources, and data science is no exception to this golden rule of thumb. Professionals with varied skill-sets are required to successfully negotiate the challenges of a complex big data project. For a data science project to be on the right track, businesses need to ensure that the team has skilled professionals. However, here comes the point where knowing the kind of team the business requires is also vital. Choosing The Right Team-Type: Rooted Teams: Embedded data science teams establish data scientists within business units. They help the business unit identify, understand, and ultimately solve problems that come up. Data scientists for embedded teams are or have an interest in becoming experts in the domain of that business. They may meet with data scientists in other business units, but the primary reporting relationship is to the business unit. Discrete\/Specialised Teams: Discrete, or specialised, teams establish “Data Science as a Service” within the organisation. They all work together, sit together, and report to the same manager. Business problems are investigated and solved as a unit. Hybrid\/Matrix Teams: Hybrid or Matrix-type teams seek to combine the best of both of these approaches by placing a team across multiple reporting lines. This may be the framework of choice if the organisation itself already uses matrix management structures. Typically, these implementations place the team as a horizontal group supporting multiple verticals. How to Get an Industrious Team? Putting together an entire team has the potential to be more difficult. The truth is that data science is a big field, and a cross-functional team is better prepared to handle real world challenges and goals. Now that we have established the types of teams that can be set up, let us look at the factors that ensure the smooth functioning of a data science team. Skillset: Getting the right talent to fill in positions in the team is the first and foremost way to move forward. Data scientists should be able to work on large datasets and understand the theory behind the science. They should also be capable of developing predictive models. Data engineers and data software developers are important, too. They need to understand architecture, infrastructure, and distributed programming. Some of the other roles to fill in a data science team include the data solutions architect, data platform administrator, full-stack developer, and designer. Those companies that have teams focusing on building data products will also likely want to have a product manager on the team. The Processes: Data science teams need data to understand and implement in different areas of the company to bring in positive results. The setup should be such where the team is not being handcuffed to a slow and tedious process, which will limit effectiveness. Ideally, the team will have a good working relationship with heads of other departments, so they work together in agile multi-disciplinary teams to make the best use of the data gathered. The team needs the ability to access and watch data in real time. It is important to do more than just measure the data. The team needs to take the data and understand how it can affect different areas of the company and help those areas implement positive changes. The Platform: When building a data science team, it is also important to consider the platform your company is using for the process. A range of options is available including Hadoop, HBase and Spark. Some of the other platforms to consider include the Google Cloud Platform, and business analytics using Excel. Understanding the fundamentals of these systems can provide a good overall foundation for the team members. Last but Not the Least… Despite keeping a check on all the above-mentioned factors, there might still be other considerations to keep in mind before sealing a data science team. Teams vs. Unicorns: There is a path to data science success that does not involve seeking that one perfect person. Rather than hunting one person with development, mathematics, statistics, and business domain expertise assemble a team of people with strong foundational skills and a drive for success. Provide them with a framework and environment within which they can work together and complement one another. The infrastructure: Get the right infrastructure set up at the right time and in the right way. Spend some time getting the infrastructure in place so that the team can start having an impact right away. The route: Data scientists may have their favourite tools and techniques. Make sure that they are not so tied to a particular tool set or machine-learning algorithm that they lack the flexibility to work on your mission. A related risk is getting caught up working on secondary or tertiary problems that are interesting, but likely to have little impact on your main goal. Conclusion: With increasing demand for data scientist skills and continued talent crunch in the industry, organisations are more inclined towards building effective data science teams. Seeing the number of firms in the big data space trying to establish big data teams, one final truth appears to resonate – one cannot nickel and dime on talent.","excerpt":"Acknowledging that data exists and can do wonders is passé. Organisations today need to do much more than just identify big data. With the shortage of data scientist skills worldwide, it is difficult for an organisation to fulfil its dream of capitalising on data science, as it is hard to find the exceptionally skilled person […]","categories":["IT Services"],"tags":["Data Science","Data science team","Data Scientist"],"author_name":"Priya Singh","publish_date":"2017-09-13T07:49:32","publication_year":"2017","word_count":957,"keywords":["big data","data science","Go","API","Data science team","machine learning","AI","ViT","analytics","GAN","Data Science","Data Scientist","R"],"extracted_tech_keywords":["AI","machine learning","data science","analytics","R","Go","API","big data","GAN","ViT"],"url":"https:\/\/analyticsindiamag.com\/it-services\/structure-choose-right-team-type-setting-data-science-team\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10076557,"title":"IT Firms’ Next Big Bet &#8211; ESG Services","content":"Within the past couple of years, one will be hard pressed to find an acronym with more buzz around it than ‘ESG’. Environmental and Social Governance or ESG was a term that first came up in a 2004 U.N. report called ‘Who Cares Wins’. Eighteen years later, as companies face increased pressure to adhere to climate laws, the term has gained a sudden resurgence. Investors and regulators alike started pushing companies to bring in ESG norms. In the past one year, Indian businesses have gradually warmed up to the idea of ESG. A report published by the corporate governance firm Stakeholders Empowerment Services or SES in March released a list of Indian companies with the highest ESG scores and IT giant Infosys came out on top. Much like their global counterparts, clients locally have started considering ESG as a priority to sign partnerships in tech—forcing Indian IT to bite the ESG bullet. The three major categories that ESG solution seekers want Infosys Having begun its climate action journey in 2008, Salil Parekh-led company Infosys released its ESG report 2020–21 in October 2020. The company board formed an Environmental, Social and Governance Committee on April 14, 2021 to oversee ESG initiatives, goals and best ESG practices. Founder of Biocon Limited and lead independent director of the board, Kiran Mazumdar-Shaw was appointed as the chairperson of the ESG Committee. Infosys offers seven main Sustainability services, which includes—Carbon Offset Advisory which gives strategic analyses, reports and recommendations to clients regarding their carbon offset programmes, Decarbonisation which is an end-to-end system that includes predictive analytics, smart metering and central command centres, ESG Data and Analytics which gathers and verifies ESG metrics, ESG Finance which creates financial risk assessments reports based on the ESG data and gives marketing, strategy and execution advice, Smart Spaces which includes solutions to track and manage energy and resource usage and embedded emissions, Sustainability Design Advisory which determines the readiness of a company’s ESG initiatives and PLM  (Product Lifecycle Management) Circularity which works on cutting down waste by reducing single-use materials, waste treatment and recycling. Wipro CEO and MD Thierry Delaporte has previously said that ESG wasn’t just important to win deals with clients but also directed them to hire the right talent. The Bangalore-based company’s ESG services cover five significant areas for clients, according to their ‘Global ESG Trends’ report released in December last year: Content which includes content designing and management, Reputation Risk Management and  Enterprise Legal Management Solutions which includes assistance with corporate compliance, IP management and legal services, Data & Insights which includes ESG data management and research and advanced analytics, eCommerce Operations which includes site content management, pricing & promotional strategies, product Information and Marketing as a Service which includes executing marketing campaigns, social media management and creative design and analytics. ESG solutions offered by HCL Tech HCL Tech HCL Tech has a range of programmes for clients as a part of its ESG solutions. Cloud Smart is the company’s cloud transformation platform with a sustainability-first approach. The Simplified ESG Analytics Platform is the data platform specific to ESG-related numbers that integrate data from external ESG data providers with internal data to come up with clients’ ESG metrics. There’s also Zero Impact Platform or ZIP which is a sustainability solution specially for the oil and gas and manufacturing sector. ZIP works in safety, risk management and cutting CO2 emissions. Track and Trace helps lower fuel consumption and also improves safety of truck drivers using IoT. Another scheme called FENIX 2.0 is a long-term digital transformation framework with a focused view on sustainability. It employs green IT, change management and workforce training while creating transformation programmes for clients. “I believe ESG is becoming more and more important for our clients. We are doing well in a lot of the ESG metrics … if you have a vision and thought leadership around ESG, you get better access to customers, better mindshare, and wins,” C Vijayakumar, CEO and MD, HCL, said in a recent media interaction. TCS India’s largest IT services company, Tata Consultancy Services (TCS) has launched a bunch of sustainability solutions including its Clever Energy platform. Established in early February this year, Clever Energy uses an IoT system to help companies measure their electricity usage. The company has also partnered with AWS to launch an ESG platform powered by cloud. The platform has a custom ESG scoring model that matches investment portfolios with the ESG criteria of companies and calculates them against prescribed sustainability benchmarks. TCS says that its clients’ sustainability initiatives have been a key driver of deals in the last few quarters. ESG service offerings by Mindtree Mindtree Mindtree has partnered with cloud-based software company Salesforce to build an ESG platform to help businesses assess their carbon footprint and diversity. The programme helps companies build their business strategies and speeds up the implementation of Salesforce Net Zero Cloud and other industry-standard metrics for social and governance practices. ESG Metric Reporting Market Forecasted to Achieve $1.425 Billion by 2026 Push for ESG from industry players India started building an ESG regulatory framework in line with the EU and other parts of the world. The ESG initiative has been marked by two vital moments in the country—the first was when Corporate Social Responsibility (CSR) spending and reporting was made compulsory under the Companies Act in 2013. Then, the market regulator, the Securities and Exchange Board of India (SEBI), marked the Business Responsibility and Sustainability Report (BRSR) mandatory for the top 1,000 listed companies on the basis of market capitalisation from April 1. The BRSR will lead the way to form a business responsibility sustainability index for Indian companies becoming a major starting point as it mandates ESG-related disclosures around its risks, opportunities and management. In May, SEBI announced a new format for firms for any sustainability-related disclosures. While there are no explicit regulation binding companies to adopt ESG, due to these slight nudges, the general notion had become that firms that complied with ESG were ‘safer’ while also helping investors make a positive impact. A StanChart report stated that India has great potential for growth in sustainable investing with scope to invest USD 1 trillion in ESG opportunities by 2030. In total, USD 8.2 trillion of retail wealth that was worth investing could be put into sustainable investments by 2030. The study found that more than 46% of investors in India wanted to help reduce their carbon footprint and 40% of investors wanted to invest early to hedge against ESG risks, and 33% wanted to genuinely help make a positive impact. Indian IT firms have clearly been at the forefront of the ESG race and the early investments will reap early dividends.","excerpt":"The Securities and Exchange Board of India (SEBI) marked the Business Responsibility and Sustainability Report (BRSR) mandatory for the top 1,000 listed companies on the basis of market capitalisation.","categories":["IT Services"],"tags":[],"author_name":"Poulomi Chatterjee","publish_date":"2022-10-06T18:00:00","publication_year":"2022","word_count":1116,"keywords":["Go","API","AWS","AI","digital transformation","predictive analytics","Git","ViT","analytics","R"],"extracted_tech_keywords":["AI","analytics","predictive analytics","AWS","R","Go","Git","API","ViT","digital transformation"],"url":"https:\/\/analyticsindiamag.com\/it-services\/it-firms-next-big-bet-esg-services\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10140651,"title":"HCLTech Partners with Inspeq AI to Drive Responsible AI in Enterprises","content":"Indian IT companies are pushing towards building AI, while focusing on doing it with responsibility. To continue this bid, HCLTech has partnered with Inspeq AI, an Irish-Indian technology firm to aid enterprises worldwide in responsibly developing and integrating AI applications. This collaboration positions Inspeq AI as the youngest company to partner with HCLTech on such an initiative. The alliance aims to embed a Responsible AI (RAI) layer into application development, seamlessly integrating with existing toolchains like Microsoft Co-Pilot, AWS Bedrock, and other business applications. The joint platform follows the five pillars of Responsible AI: Fairness, Robustness, Explainability, Privacy, and Transparency. Through this platform, industries, including financial services, can address critical challenges such as credit risk, underwriting, fraud detection, and risk scoring. For example, a prominent European bank leverages this platform to tackle issues ranging from account openings to loan and insurance underwriting. “We are immensely proud to announce our partnership with HCLTech and together bring Inspeq’s RAIops platform for safe AI and Agentic deployments in top enterprises around the world,” stated Apoorva Kumar, CEO of Inspeq AI. “This is a big step for us toward advancing AI safety and performance on a huge scale for the protection of global enterprise brands.” The partnership leverages Inspeq AI’s capabilities and HCLTech’s extensive client base to offer scalable, enterprise-grade GenAI solutions and agents for secure, compliant processes. The industry-agnostic platform, underpinned by advanced NLP and AI\/ML, assesses and enhances models across applications, enabling responsible AI development while driving cost savings and speeding up delivery for HCLTech’s clients. Founded in 2023, Inspeq AI raised $1.1 million (INR 9.13 Crore) in pre-Series A funding, led by Sure Valley Ventures, a specialist AI investment firm. The platform is designed to make GenAI app development four times faster, with projected development cost reductions of up to 70%. Co-founded by industry veterans from Microsoft, Meta, LinkedIn, and Amazon, Inspeq AI is headquartered in Ireland. Its RAIops platform is geared towards advancing safe and ethical AI adoption while aligning with global standards like the EU AI Act. The platform employs over 50 research-backed metrics and 20,000 data sets to safeguard compliance, reduce security risks by 90%, and enhance model accuracy by 80%, ensuring responsible AI governance across industries. HCLTech is seeing traction in its AI offerings, as highlighted by CEO C Vijayakumar during the company’s Q2 2025 earnings call. He shared that AI has been integrated into the majority of their recent deals, highlighting the platform’s key role in service transformation. “We have seen strong wins with most of our deals incorporating AI capabilities. Our platform AI Force is gaining widespread adoption for service transformation among our clients,” said Vijayakumar. He further emphasised how the company’s AI offerings, such as HCLTech AI Force, HCLTech AI Foundry, and AI Labs, are driving innovation across sectors. A key differentiator for HCLTech is the integration of AI Force with Microsoft GitHub Copilot, which will soon be extended as a Copilot extension. “AI Force is now integrated with Microsoft GitHub Copilot and will soon provide unmatched extensibility and a broad range of use cases,” said Vijayakumar.","excerpt":"Founded in 2023, Inspeq AI raised $1.1 million (INR 9.13 Crore) in pre-Series A funding, led by Sure Valley Ventures.","categories":["AI News"],"tags":["enterprise ai","HCL Technology"],"author_name":"Mohit Pandey","publish_date":"2024-11-08T11:25:41","publication_year":"2024","word_count":513,"keywords":["Go","GenAI","AWS","AI","ML","RAG","NLP","Aim","HCL Technology","enterprise ai","R","fraud detection"],"extracted_tech_keywords":["AI","ML","NLP","GenAI","Aim","RAG","fraud detection","AWS","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/hcltech-partners-with-inspeq-ai-to-drive-responsible-ai-in-enterprises\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10132196,"title":"Gupshup Boosts Workforce by 20% Amid Rising Demand for Conversational AI Solutions","content":"Gupshup is witnessing strong demand for its conversational AI solutions in India and international markets. To fuel this growth, the company undertook accelerated hiring in FY24, growing its workforce by 20% to 1400 people. The hires will support Gupshup’s growth and expansion across India, Latin America, Middle East, SEA, Africa and Europe. Additionally, the company made several senior-level hires across Marketing, GTM (go-to-market), Engineering, and Solutions. Already a profitable unicorn, Gupshup saw 40% YoY growth last year, driven by brands’ escalating demand to engage customers through conversational advertising, marketing, and support on messaging channels. “As we continue to expand our footprint globally, we are actively seeking top talent across engineering, product development, marketing, and customer support roles to drive this conversational revolution. Our people are our greatest asset, and we are committed to building a diverse and inclusive workforce that can unlock massive value for our customers through innovative conversational experiences,” Madhuri Nandgaonkar, VP – HR, Gupshup said. While India has been a primary market for Gupshup, geographies like Latin America, Middle East, APAC, Africa and Europe have emerged as key growth drivers for the company over the last 3 years. The demand for Gen AI-powered conversational experiences globally has seen the company double its team size in Brazil and ramp up hiring efforts in China, the GCC region, Indonesia, Malaysia and Turkey. In FY25, the company aims to boost its engineering and Go-To-Market (GTM) teams, followed by product development and customer support, across India and other geographies. According to the company, this year, the company also saw a 15% increase in women in senior leadership positions. The company hired 22% women employees and 33% women interns as part of their internship program. Gupshup has doubled its customer base in several of the international markets and works with leading global brands including, L’Oréal, P&G, Grupo Carso, GoJek, Nestle, Petromin, and Netflix among others. Earlier this year, Gupshup further expanded its product suite with the launch of Conversation Cloud – a comprehensive suite of SaaS tools aimed at revolutionizing business.","excerpt":"The demand for Gen AI-powered conversational experiences globally has seen the company double its team size in Brazil and ramp up hiring efforts in China, the GCC region, Indonesia, Malaysia and Turkey.","categories":["AI News"],"tags":["Gupshup"],"author_name":"Pritam Bordoloi","publish_date":"2024-08-12T13:02:50","publication_year":"2024","word_count":339,"keywords":["Go","unicorn","programming_languages:R","AI","programming_languages:Go","Scala","Gupshup","GRU","Aim","programming_languages:Scala","R"],"extracted_tech_keywords":["AI","Aim","R","Go","Scala","GRU","unicorn","programming_languages:R","programming_languages:Scala","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/gupshup-boosts-workforce-by-20-amid-rising-demand-for-conversational-ai-solutions\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10021904,"title":"Explained: MIT Scientists’ New Reinforcement Learning Approach To Tackle Adversarial Attacks","content":"Adversarial inputs, also known as machine learning’s optical illusions, are inputs to the model an attacker has intentionally designed to confuse the algorithm into making a mistake. Such inputs can be typically dangerous for machines with a very low margin for risk. For instance, in self-driving cars, an attacker could target an autonomous vehicle with an adversarial stop sign to fool a car into stopping. In this article, we try to examine why it is difficult to overcome adversarial inputs and how academics at MIT overcome these obstacles. The Challenge Suppose you have to train a model to identify whether a picture you have clicked is of a dog or not. In an ideal scenario, a model trained on labelled images of dogs and images that are not of dogs would be enough. But what an adversary can do is try to introduce, for instance, a morphed image of a cat an original training model might not be able to distinguish from a dog. In this case, the ML developer can rely on adversarial training. The developer tries to train the model by inputting variations of altered images that could classify the image as a dog. This can make the model immune to adversarial attacks. However, there are two drawbacks to this approach. Firstly, the brute-force solution to generate every possible adversarial example to train our model will not be efficient for real-time applications like a self-driving car. Running through every possible image alteration will require huge computational time for inferencing and cannot be used in time-sensitive applications. Secondly, there are ways to always calculate your model’s threshold despite all the image alterations the model has been trained on. This is easily possible if the classification model is providing output in ‘probabilities’ of percentage. For instance, the model outputs the particular image is 99% aeroplane and 1% cat. The adversary can then tamper with his input to fool the model. Suppose the model provides output in ‘classes’, that is, it only tells whether it is a cat or not, the adversary can still use the model to train their substitute models that can reverse engineer to estimate the threshold for the original model and then introduce adversarial inputs accordingly. Such attacks have been explored in the ‘Practical Black-Box Attacks against Deep Learning Systems using Adversarial Examples’. Overcoming The Challenge Scientists at MIT came up with an advanced iteration of reinforcement learning (RL), a type of machine learning that does not need labelling. The method is mostly used in algorithms involved in games like Chess or Go. The current approach used by academics is called CARRL, which builds over an existing deep reinforcement learning DQN or Deep Q-Network. DQN is a neural network that allocates an input a Q-value or a level of reward. CARRL takes the input image as a dot and then considers an adversarial influence around the dot. The DQN analyses every possible position within this region to find an ‘associated action’ that would result in the most’ optimal worst-case reward.’ If there is a possibility of adversarial input, the approach considers the worst-case reward while coming to a conclusion. The developers of this algorithm experimented with their approach on a Pong game where two players operate paddles on either side. When the computer is playing fair, only a DQN approach could beat the computer. DQN will loose if the adversarial input changes the direction of the ball. But using CARRL, the RL approach put the paddle in such a way that it will be in a region based on the optimal worst-case reward. Going Forward The MIT researchers tested the application on two agents where one agent had to avoid colliding with the second to reach its destination. The first agent was successfully able to avoid collisions despite the adversarial attacks introduced by the second. The scientists tried to increase the adversarial nature of the second agent. At a point, the first agent avoided going to its destination altogether. The scientists claimed this sort of conservatism is useful as it can be used as a limit to tune the algorithm’s robustness. The new method can go a long way in implementing AI where decision-making is time-sensitive, and the margin for error is critical.","excerpt":"Adversarial inputs, also known as machine learning’s optical illusions, are inputs to the model an attacker has intentionally designed to confuse the algorithm into making a mistake. Such inputs can be typically dangerous for machines with a very low margin for risk. For instance, in self-driving cars, an attacker could target an autonomous vehicle with […]","categories":["AI Features"],"tags":["Adversarial AI","Adversarial Attacks","Adversarial Learning","Adversarial Training","machine learning document classification","MIT researchers","Reinforcement Learning","Self-driving car"],"author_name":"Kashyap Raibagi","publish_date":"2021-03-11T14:00:00","publication_year":"2021","word_count":706,"keywords":["Go","Adversarial Attacks","MIT researchers","machine learning document classification","Reinforcement Learning","AI","Self-driving car","machine learning","ML","neural network","Adversarial AI","Adversarial Learning","TPU","Aim","deep learning","R","adversarial attacks","Adversarial Training"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","neural network","Aim","TPU","R","Go","adversarial attacks"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/explained-mit-scientists-new-reinforcement-learning-approach-to-tackle-adversarial-attacks\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":27534,"title":"Next Wave Of Supercomputers Promise To Solve World’s Most Challenging Problems","content":"NERSC Cray Edison supercomputer cluster at the Oakland Scientific Facility – 10\/31\/2013. In today’s world, it is seen as a sign of technical prowess when a nation can build great supercomputer, systems which are millions of times more stronger than personal computers. The US and China have been competing with each other for building supercomputers for quite some time now. Underlining the importance of such technical developments, a US committee had gone on record to say, “Through multilevel government support, China now has the world’s two fastest supercomputers and is on track to surpass the United States in the next generation of supercomputers — exascale computers — with an expected rollout by 2020 compared to the accelerated US timeline of 2021.” In fact, right here in India, major US chipmaker NVIDIA and the Indian industrial science institution, Council of Scientific and Industrial Research (CSIR) joined forces to build a Supercomputer Centre of Excellence at Central Electronics Engineering Research Institute (CEERI) at New Delhi. This centre is set to house India’s first ever supercomputer. This initiative will give India a five-petaflop supercomputer which is also India’s first ever AI supercomputer. Supercomputer And Super Speed Supercomputers can achieve great computing speeds that can also reach petascale levels — which means they can do about one quadrillion (1,000,000,000,000,000) computations per second. On the other hand, our normal computers can compute upto one billion calculations per second. These kind of supercomputing speeds are needed for important domains like chemistry and biology applications. Thom Dunning who is a chemistry researcher and lecturer at the University of Washington and the co-director of the Northwest Institute for Advanced Computing, had once said, “As you develop models that are more sophisticated that include more of the physics, chemistry and environmental issues that are important in predicting the climate, the computing resources you need increases.” With inspiration for such innovative applications supercomputers are reaching new heights. For example, the supercomputer built by IBM and Nvidia for the Oak Ridge National Laboratory can reach upto 20 quadrillion computations per second. New York Times, compared this achievement as, “A human would require 63 billion years to do what Summit can do in a single second.” MIT Technology Review said, “Everyone on Earth would have to do a calculation every second of every day for 305 days to crunch what the new machine can do in the blink of an eye.” Technology Behind The Supercomputer Supercomputers have been largely powered by the concept of parallel processing where several commands can be computed at the same time with the help of many executioners or processors. The paradigm of parallel processing came after decades of computers working on a much simpler paradigm known as serial processing in which commands are executed one at a time. So why do supercomputers work on parallel processing? When a biology or chemistry problem needs to be solved by scientists it can push normal computers to the very limit. The best approach is to split the bigger problem into smaller problem and add more high-performing processors and divide the tasks between them. The idea springed a revolution in the scientific and computing community. Every time more computing speed is needed, more processors are added. The next steps is to bring many such computers (off-the-shelf PCs) and connect them using LAN and creating a supercomputing cluster. Sunway TaihuLight, which was reported to have simulated the universe for scientists has around 40,960 processing modules. Each of these modules has around 260 processor cores which adds to 10,649,600 processor cores in the whole supercomputer. Exascale Supercomputers And Solving World Problems The next wave of supercomputers are called exascale supercomputers. These supercomputers will help us build greater applications that will help us solve problems in the complex fields of physics, biology and manufacturing, for example. Exascale computers are supposed to be at least 10 to 20 times faster than today’s supercomputer. The Exascale project is being run by the Department Of Energy in the US and has active research on subject areas like chemistry and materials applications of supercomputers, earth and space science applications, among many others. A statement put out by the Exascale project states, “The Exascale Computing Project is focused on accelerating the delivery of a capable exascale computing ecosystem that delivers 50 times more computational science and data analytic application power than possible with DOE HPC systems such as Titan (ORNL) and Sequoia (LLNL). With the goal to launch a US exascale ecosystem by 2021, the ECP will have profound effects on the American people and the world.” The primary source of motivation for the project is strengthening national security, advancing scientific security and building software and hardware technologies for building powerful exascale supercomputers. The hardware integration project makes sure the dream of exascale computers is a reality with the combination of right ECP applications, software and advanced hardware technology. How can exascale supercomputers do for the most difficult problems on earth? Sample this. Due to shortage of drinking water poor all over the world suffer and are faced with difficult situations everyday. The only solution currently seems to be turning ocean water into something drinkable and edible. Researchers at the Lawrence Livermore National Laboratory say that the solution is solving the carbon nanotubes problem. The researchers said, “The microscopic cylinders serve as the perfect desalination filters: their radius is wide enough to let water molecules slip through, but narrow enough to block the larger salt particles. The scale we’re talking about here is truly unimaginable; the width of a single nanotube is more than 10,000 times smaller than a human hair.”","excerpt":"In today’s world, it is seen as a sign of technical prowess when a nation can build great supercomputer, systems which are millions of times more stronger than personal computers. The US and China have been competing with each other for building supercomputers for quite some time now. Underlining the importance of such technical […]","categories":["AI Features"],"tags":["AI Supercomputer","NVIDIA","Supercomputers"],"author_name":"Abhijeet Katte","publish_date":"2018-08-22T11:47:42","publication_year":"2018","word_count":932,"keywords":["Go","programming_languages:R","AI","RPA","AI Supercomputer","programming_languages:Go","Ray","Supercomputers","NVIDIA","R"],"extracted_tech_keywords":["AI","Ray","R","Go","RPA","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/next-wave-of-supercomputers-promise-to-solve-worlds-most-challenging-problems\/","complexity_score":3,"technical_depth":7,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":38647,"title":"How Recommendation Engines Help Understand Users Faster And Better","content":"In this technology-driven era, recommendation engines today are playing a vital role in our day to day life. These technologically advanced engines are changing the way how we used to make a decision and decide on something. From buying t-shirts online to deciding on which movie to watch to which song to listen to, recommendation engines are everywhere. For example, have you ever noticed your phone’s music player that suggests you songs every now and then, and if not completely, at least to an extent, the suggestions are just what we actually want to listen. Another example is the online shopping app you use. It also keeps suggesting you clothes and accessories you might want to check or even want to buy. It even suggests your product that actually fits perfectly to the budget you have. But, how does recommendation engines do that? How can it know and understand the user? How much time does it actually take to deliver that accuracy?  In this article, we will have a look at a use case that would give you all a view of how a recommendation engine delivers results. Before jumping right into how a recommendation engine reads and understands its user, let us first have a look at what a basically a recommendation engine is. So, recommendation engines are intelligent information filtering system. They are developed in such a way that they narrow the decision-making process and predicts and shows the results that a user would like to have a look. Recommendation engines are not at all 100% accurate, however, with all the recent advancements they have reached a significant level of accuracy. Talking about how this intelligent system works, basically, a recommendation engine works based on data collection, data storage, data analyses and data filtration. So, it takes a little bit of time for the system to understand its user. However, there are few recommendation engines that are pretty advance and doesn’t take much time to deliver results that are significantly accurate. How Viu Uses Machine Learning To Advance Its Recommendation Engine Viu is one of the most emerging platforms in the OTT space in India at present. And it is not just the content that has done its job in making popular, but its recommendations engine is also playing a vital role. Let us have a look at how Viu clip is using AI and ML to make its recommendation engine top-notch. The recommendations of the platform are based on data collected from the end user — such as what kind of content the user is seeing, how much time it is sticking to it etc. The more data keeps on getting collected, the better the engine becomes for the end user as the platform keep changing the recommendation algorithm and tailor as much as possible. The accuracy of the recommendation engine is also depending on the same parameters. However, there are few recommendations that are available to users all the time because the hit rate on that is very high. Furthermore, Viu has also created experimental ribbons where the engine keeps checking the data and it depends on the segment of users on more than 200 attributes. Across all these attributes, multiple permutations and combinations get created. And depending on that, the recommendations get created. Talking about how much time does it take for a recommendation engine to understand its user, Viu uses a hypothesis, which is based on data collected from first party and third party sources. Using that data, the platform trying to learn from the other market and does profile matching. For example, they try to match and see if a content that is popular in one region can also be popular in some other region.  In this entire process, machine learning plays a vital role. “Once we launch in a region, we start collecting the data and when we reach certain critical mass, the ML engine to become more accurate,” said Abhijit Bhide, SVP – Engineering at Vuclip in an interaction with Analytics India Magazine. Viu is actively using machine learning to better the results of its recommendation engines. And today, the platform has reached a level of accuracy that in a brand new place, within a month,  the platform’s is able to get a fairly good set of recommendations for the end users. Outlook Recommendation engines today are a vital part of every online platform — whether is media ore-commerce. While big players like Amazon, Flipkart, Netflix, are extensively advancing the technologies behind their recommendation engine, emerging players are also not much lagging behind. And behind all these advancements, machine learning is proving to be the prime catalyst.","excerpt":"In this technology-driven era, recommendation engines today are playing a vital role in our day to day life. These technologically advanced engines are changing the way how we used to make a decision and decide on something. From buying t-shirts online to deciding on which movie to watch to which song to listen to, recommendation […]","categories":["AI Features"],"tags":["Data Science","ecommerce","ecommerce analytics","Machine Learning","recommendation engine"],"author_name":"Harshajit Sarmah","publish_date":"2019-05-05T09:38:04","publication_year":"2019","word_count":776,"keywords":["Go","ecommerce analytics","machine learning","programming_languages:R","AI","ML","Machine Learning","RAG","recommendation engine","ecommerce","ViT","analytics","CLIP","Data Science","R"],"extracted_tech_keywords":["AI","machine learning","ML","analytics","RAG","R","Go","CLIP","ViT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-recommendation-engines-help-understand-users-faster-and-better\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":23516,"title":"Scientist Who Invented The Term &#8216;Artificial Intelligence&#8217; To Be Celebrated At Memorial Service","content":"Noted scientist Dr Philip Mayne Woodward who coined the term ‘artificial intelligence’, will be celebrated at his memorial service mid-April. Dr Woodward passed away in January this year was 98 years old. Known for his exciting work in radar engineering, the British mathematician was also a world-renowned horologist or a clock-maker. According to his obituary that appeared in a British local newspaper, Dr Woodward was born on 6 September 1919 and was educated at Blundell’s School in Devon. In 1938 he won a scholarship to study mathematics at Wadham College, Oxford, but was in 1941 he was drafted to the Telecommunications Research Establishment — home of radar research for the RAF. “One day at the Harvard University, Oliver Selfridge and Marvin Minsky called at the Cruft laboratory to discuss the programming of computers to exhibit quasi-intelligent behaviour, an emerging field at the time. Dr Woodward said that a snappy title was needed, if only to oust the anthropomorphic phrase ‘electronic brain’. The word ‘intelligence’ had already been agreed when Woodward suggested prefacing it with ‘artificial’ to suggest the mimicking of mental processes. In five minutes, the now-familiar term ‘artificial intelligence’ had been coined,” says the obituary. Some of his other accomplishments include a four decade-old career in British Scientific Civil Services. According to historic reports, Dr Woodward was responsible for the software of one of the UK’s first electronic computers (TREAC), followed by the UK’s first solid state computer (RREAC). In June 2005, the Royal Academy of Engineering gave Dr Woodward his first Lifetime Achievement Award, recognising him as an outstanding pioneer of Radar and for his work in precision mechanical horology. In 2009 he received the Institute of Electrical and Electronics Engineers (IEEE), Dennis J Picard Medal for Radar Technologies and Applications “for pioneering work of fundamental importance in radar waveform design, including the Woodward ambiguity function, the standard tool for waveform and matched filter analysis.”","excerpt":"Noted scientist Dr Philip Mayne Woodward who coined the term ‘artificial intelligence’, will be celebrated at his memorial service mid-April. Dr Woodward passed away in January this year was 98 years old. Known for his exciting work in radar engineering, the British mathematician was also a world-renowned horologist or a clock-maker. According to his obituary that appeared […]","categories":["AI Features"],"tags":["AI (Artificial Intelligence)","Interviews and Discussions"],"author_name":"Prajakta Hebbar","publish_date":"2018-04-11T10:45:52","publication_year":"2018","word_count":317,"keywords":["artificial intelligence","programming_languages:R","AI","R","AI (Artificial Intelligence)","Interviews and Discussions"],"extracted_tech_keywords":["AI","artificial intelligence","R","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/scientist-who-invented-the-artificial-intelligence-dies\/","complexity_score":2,"technical_depth":4,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10066029,"title":"Training data for effective AI deployment","content":"As AI products and services are increasingly being deployed into the real world, ML Data OPs have had to rapidly iterate to meet the challenges of handling data for model training and continuous testing. Sudeep George, VP of Engineering at iMerit, in his talk, ‘Training data for effective AI deployment’ at the Data Engineering Summit 2022, has shared practical insights around handling training data for complex AI systems operating at scale. Sudeep defined data annotation as “a process of structuring and classifying data so that an ML model can distinguish between the human and the background, the people and vehicles, and understand the actions and intents by analysing its cause.” The process is important because it provides highly accurate ground truth data and helps machines to parse and understand the text. The annotation journey Nearly 60 to 70% of the model development time is spent on data-related processes like gathering, cleaning and structuring. The data should be maintained at a very high quality to optimise the model’s performance. To build training data, companies go through three types of vendors. Initially, companies opt for tool providers who can help them annotate complex data objects. Later, the company reaches a point where tons of data needs to be classified, and they bring in workforce providers who offer trained annotators. Soon, the companies reach another inflection point and start working with solution providers. Model deployment has several stages. First, the model is evaluated using the test dataset. Later, it is tested against fresh production data and, post-deployment, continuously monitored to ensure performance. The model has to be retrained when the performance dips. Data-centric AI The traditional approach to building AI systems has been to train a model to reach a baseline performance and tweak and tune it to improve performance. Unfortunately, these models usually fail after deployment. The data-centric approach brings the model to baseline performance and trains it against datasets that represent the environment within which the model has to operate. For models, it is important to have Data at scaleTrustworthy data Augmentation is a popular way to scale data. Here, two key approaches are used: deep learning and data domain manipulation. The manipulation techniques are specific to the data modality. For instance, for computer vision, image filters or transformation tools (such as image geometry transformation or image colour space transformation) are used to transform the image for training. On the other hand, deep learning techniques like GANs generate synthetic images based on original data. This ensures the dataset is evenly distributed among all classes and minimises the bias. Dataset management is important to ensure the collected data is used across various models in a manner that the data is reusable. It also helps understand data distribution and databases. Edge cases affect the performance of the model. Some unique classes only have limited data available, and when the model encounters these in real-life scenarios, it leads to errors and failures. Depending on the ML life cycle part, an edge case can have different implications. In the collection phase, it is about insufficient data. In the labelling phase, it is about incomplete attributes being classified. In training, it is about incomplete testing against scenarios. Lastly, in the deployment phase, it is about the actual performance of the ML model when working on real-world data. Sudeep discussed some of the common challenges companies face, including guidelines, traceable feedback loop, scenario replication, and end-end system validation. Lastly, he discussed an example of a customer who wanted to train a model to distinguish between a cloudy and sunny outside environment. The iMerit team used the shadows cast by cards and other objects to act as a feature point for the ML training process, and the model’s performance improved drastically. REGISTER HERE TO ACCESS THE CONTENT","excerpt":"Nearly 60 to 70% of the model development time is spent on data-related processes.","categories":["AI Features"],"tags":["Deep Learning","GAN"],"author_name":"Avi Gopani","publish_date":"2022-05-01T16:00:00","publication_year":"2022","word_count":629,"keywords":["Go","API","AI","ML","computer vision","deep learning","data engineering","Rust","GAN","Deep Learning","R"],"extracted_tech_keywords":["AI","ML","deep learning","computer vision","R","Go","Rust","API","data engineering","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/training-data-for-effective-ai-deployment\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10043534,"title":"AWS Public Sector Startup Ramp Launched In India","content":"AWS has announced the launch of the AWS Public Sector Startup Ramp in India. AWS Startup Ramp is an acceleration program for startups building innovative solutions for public sector customers worldwide, including but not limited to customers in national and local government, space and defence, and healthcare. “We are excited to launch the AWS Startup Ramp in India, where there is an incredible opportunity for startups to solve public sector challenges through their innovative solutions. This global initiative is a critical way for us to support startups who are using technology to change the world,” said Sandy Carter, VP, Public Sector Partners and Programs, AWS. Startups in their early-stage growth can apply to join the AWS Startup Ramp in one of two tiers: Pre-revenue startups can apply to the Innovator tier.Post-revenue startups (with up to INR 100 crores in revenue) can apply to the Member tier. Startups meeting the criteria will be reviewed and evaluated for participation based on their potential contributions to the public sector. “Our experience in India has shown that public sector organizations are increasingly keen to experiment with new ideas and technologies to advance their missions at speed and scale. Startups bring disruptive innovative solutions that create broad impact, and can pivot quickly to meet society’s changing needs. We are doubling down on our focus to serve the public sector, and empower startups with the new AWS Startup Ramp program. We are thrilled to introduce the program to expand the capabilities of India’s strong and aspiring startup ecosystem,” said Rahul Sharma, President, Public Sector – AISPL, AWS India and South Asia. AWS Startup Ramp is modelled after AWS EdStart, which works with education technology-focused startups, and AWS GovTechStart, the company’s US-based program that supports commercial technology companies serving state and local governments. Each program offers support to eligible companies, including access to AWS promotional credit, technical training and support, a community of experts, and more. Learn more about the AWS Startup Ramp here.","excerpt":"AWS has announced the launch of the AWS Public Sector Startup Ramp in India. AWS Startup Ramp is an acceleration program for startups building innovative solutions for public sector customers worldwide, including but not limited to customers in national and local government, space and defence, and healthcare. “We are excited to launch the AWS Startup […]","categories":["AI News"],"tags":["AI Startups"],"author_name":"kumar Gandharv","publish_date":"2021-07-13T19:55:58","publication_year":"2021","word_count":327,"keywords":["Go","AWS","AI","cloud_platforms:AWS","programming_languages:R","programming_languages:Go","GAN","R","AI Startups","startup"],"extracted_tech_keywords":["AI","AWS","R","Go","GAN","startup","cloud_platforms:AWS","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/aws-public-sector-startup-ramp-launched-in-india\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":61727,"title":"Important Steps To Keep Your Data Lake From Turning Into A Data Swamp","content":"As the repository of data expands with each passing day, deriving meaningful insights from it through analytics has become a challenge. With organizations feeling the need to move this data around for better understanding and collaboration between different departments, data lakes have emerged. In simple terms, these are a large body of raw data where users examine data and extract information from it. However, to use it efficiently, one needs better data governance and management. This is because, without that, it leads to something called ‘data swamps’. While there are many steps one can take to ensure that data lakes do not turn into swamps, we have attempted to explain some of the essential ones: Start With Project-Specific Data One of the biggest reasons for data lakes to fail is the lack of planning. Organizations dump all company-related data into their data lakes – this should not be done. Instead, they should build their data lake based on various projects. While the point of having a data lake is to have all company-related information in one place, the answer is to not turn it into a swamp by striking the right balance. Company has to find a balance between the quantity of its analytics data, and the value its data lake holds for its business functions. Catalog The Data On Ingest Cataloguing the data on ingest makes it more searchable. This is something which is done when the data is brought into the lake. The organization should make the data easy to find for its proper analysis. Cataloguing also helps in eliminating accidental loading of the same data source more than once. This step requires immediate attention. Loading the data into the lake and leaving it to catalogue in the future is a big mistake. This is because cataloguing the data from the data lake after some time has passed will prove to be difficult and time-consuming. Loading Data Only Once Loading data poses two challenges – the first is managing the large data file systems. These large data file systems require loading one entire file at a time. When it comes to loading small tables and files, it is not difficult, but as the file size increases, loading these can become a problem as it will take more time. One can minimise the time it takes to load large source data sets by loading the entire data set once, and later merging and syncing the changes in the data lake. Another challenge could be when two different people try to load the same data source onto different parts of the data lake. The DBAs responsible for upstream data sources getting loaded into the data lake will face problems, as the lake will consume too much capacity to load the data. This will result in the data lake interrupting in operational databases that are used for businesses. Documenting Data Lineage & Good Governance Implementation Once the data lake is in use, different people who use it might clean it or start integrating it with other data sets. And while people do this, often when someone wants to implement a project they are interested in, chances are the data related to it might have already been cleaned. Now, when one is only familiar with the raw data of the project they are interested in and not any other version, they will have to redo the work that has already been done. To avoid this problem, documenting changes related to the data thoroughly and implementing solid governance processes that bring to light the interactions people have had to ingest and transform data is important.","excerpt":"As the repository of data expands with each passing day, deriving meaningful insights from it through analytics has become a challenge. With organizations feeling the need to move this data around for better understanding and collaboration between different departments, data lakes have emerged. In simple terms, these are a large body of raw data where […]","categories":["AI Features"],"tags":["data lake"],"author_name":"Sameer Balaganur","publish_date":"2020-04-18T10:00:00","publication_year":"2020","word_count":602,"keywords":["Go","data lineage","programming_languages:R","AI","programming_languages:Go","analytics","data governance","GAN","R","data lake"],"extracted_tech_keywords":["AI","analytics","R","Go","data lake","data governance","data lineage","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/important-steps-to-keep-your-data-lake-from-turning-into-a-data-swamp\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10001860,"title":"Wipro Hires Cybersecurity Forensics To Look Deeper Into Purported State-Sponsored Attack","content":"Wipro, one of the country’s biggest software exporters has been reported to be the target of an attack from hackers. Earlier this week, cybersecurity portal KrebsOnSecurity reported that it had been informed by two independent, trusted ‘expert’ sources that Wipro was undergoing a “multi-month intrusion from an assumed state-sponsored attacker”. Wipro immediately proceeded to issue a statement that said that there was an indicator of a “potentially abnormal activity” in employee accounts. However, light has not been shed on what the actual issue was. The ‘KrebsOnSecurity’ Narrative The sources who revealed the information to the portal were kept anonymous. However, they did state that not only was Wipro compromised, but that their systems were used as entry points to compromise other systems. These systems belonged to Wipro customers, around 11 of which have already been compromised. This information was found by inspecting file folders on intruders’ back-end infrastructure. These clients were also targeted for phishing attacks using Wipro’s credentials. When reached for comment, an executive at Wipro stated that the company has a “multilayer security system” that is monitored constantly at a “heightened level of alertness.” One of the sources even stated that Wipro was building a whole new private email network to boot the invaders out. Post this, the source also stated that Wipro was briefing their clients on cybersecurity practices of “indicators of compromise”. Wipro’s Narrative After sending out an email to multiple news publications stating that they were “leveraging industry-leading cybersecurity practices” to address the issue. The statement also mentioned that they have begun an investigation to “identify the affected users” and take remedial steps to control impact. In a statement to ET, Wipro executive Abidali Neemuchwala stated that these attacks are “common in the industry”, adding that they have a “pretty good email system”. He also described the attack to have taken place by taking advantage of a zero-day vulnerability. This stands in stark contrast to what was said by Krebs, as a multi-month attack would have been discovered and patched if it was functioning on a zero-day vulnerability. The head of HR at Wipro, Saurabh Govil, stated that the company has a training program to prevent against common attack vectors such as phishing attacks. He stated “The company has a training programme where phishing mails are generated internally and sent to employees. If an employee clicks on it, they are sent to training.” Damage Control Or Reality? Many argue that this can be seen was an attempt to control damage, as the stock of the company was quickly tanking after the article. However, Wipro did mention that they had already informed a handful of their customers. It is to be noted that the company is liable legally to reimburse any costs that the security breach might have caused. This was also one of the risk factors that the company had mentioned in their annual report. In the report, they stated that they could be subject to “significant liability” in the event of data being misappropriated by employees or former employees. The reason for the liability incurred is for “breaching contractual confidentiality provisions” and various laws regarding data privacy. Reportedly, the company’s COO addressed various analysts on a post-earnings call. This prompted the author of the original article, Brian Krebs, to ask what the inaccuracies in his report were. The COO then asked Krebs to join him ‘on a seperate call’. In a tweet, Krebs stated, “They’re (Wipro executives) happy to tell investors my story is full of holes but when I ask them to their face to say where the piece was in error, they dodge the question. Definitely the behaviour of a company with nothing to hide.” A Potential Attacker It is also to be noted that this specific attack vector and target is indicative of the attack strategies of Chinese hacker group APT10. The group targets service providers such as Wipro to discover vulnerabilities in their systems. They then use them as a method to attack their clients, which are usually decided by Chinese national security goals. The group also acquires “valuable military and intelligence information”, along with business data to support Chinese companies, according to a statement by FireEye. They also said “APT10 has targeted or compromised manufacturing companies in India, Japan and Northern Europe; a mining company in South America and multiple IT service providers worldwide.” APT10 could have used Wipro as a ‘soft underbelly’ of sorts, as it is quite possible that they could have compromised the systems. They have a multitude of malware that they have developed for use as backdoors, so as to gain access to targets. Currently, Wipro has employed a cybersecurity forensics company to find what issues occured with their systems. The story is yet to develop.","excerpt":"Wipro, one of the country’s biggest software exporters has been reported to be the target of an attack from hackers. Earlier this week, cybersecurity portal KrebsOnSecurity reported that it had been informed by two independent, trusted ‘expert’ sources that Wipro was undergoing a “multi-month intrusion from an assumed state-sponsored attacker”. Wipro immediately proceeded to issue […]","categories":["AI News"],"tags":["Cybersecurity","Hack","Wipro"],"author_name":"Anirudh VK","publish_date":"2019-04-17T19:47:34","publication_year":"2019","word_count":791,"keywords":["Wipro","Go","AWS","AI","cloud_platforms:AWS","programming_languages:R","programming_languages:Go","Hack","RAG","ViT","Rust","Cybersecurity","R"],"extracted_tech_keywords":["AI","RAG","AWS","R","Go","Rust","ViT","cloud_platforms:AWS","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/wipro-hires-cybersecurity-forensics-to-look-deeper-into-purported-state-sponsored-attack\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10072596,"title":"Meta Takes on Google, Open-Sources White-box Search Solution Sphere","content":"Ever wondered how Wikipedia or Google Search answers our questions? Random thoughts strike us at random moments, and we turn to these search engines and online encyclopaedias. Either way, we get our answers. Mostly. Knowledge-intensive natural language processing (KI-NLP) is how Google Search or Wikipedia fetches answers to our questions. AI models in them dig through an archive of information to give us relevant search results. However, there are several limitations to the current KI-NLP landscape. KI-NLP architectures depend on black-box search engines to hunt for information from the knowledge web. In the process, relevant information can be missed as the search engine algorithms can rank it too low in the results. Also, in the case of Wikipedia searches, often, the online encyclopedia doesn’t capture all the knowledge available on the web related to a particular topic, and with its continuous growth, it has become a challenge to verify citations and other biases. Meta’s Sphere Meta came up with the first white-box search solution, Sphere, which uses open web data as the source of knowledge. Meta believes that Sphere’s white-box knowledge base has significantly more data and sources to match for verification than a typical black-box knowledge source. Thus, it can provide useful information that they cannot. Source: The Web Is your Oyster The idea was to create more intelligent AI systems that could leverage real-world knowledge in a better manner. Sphere has surpassed the Knowledge Intensive Language Tasks benchmark, implying it can help AI researchers to build models that can leverage real-world knowledge to accomplish multiple tasks. Sphere represents the effort of Meta to enable AI researchers to experiment with building KI-NLP models. Meta believes Sphere will help researchers train retrievers to handle a wider range of documents and prepare automatic systems to deal with issues like misinformation and incoherent text. The models thus created could help in the real world to tackle harmful content. It also holds the potential to enhance digital literacy and critical thinking skills. How Meta seeks to challenge Google with Sphere Shortly after Meta released Sphere, discussions started doing the rounds that Meta is seeking to challenge Google. “MetaAI introduces white-box search. By open-sourcing Sphere, its web scale corpus. Directly challenges Google,” posted Prithivi Damodaran, an ML consultant at Donkey Stereotype. With Sphere, Meta is trying to address the problem of the most-relevant source related to the surfer’s query or topic. At a time when search engine optimisation is widely used to rank information resources easily, what appears higher in search results might not be the most relevant source for the surfer. In fact, Google Search is notorious for its search results. There have been numerous complaints by users about Google Search results being wrong and irrelevant. Many times, the initial results pertain to ads or information not even remotely related to the query. Sphere seeks to solve this issue. Another way Meta is trying to outdo Google with Sphere is by open sourcing it. Big Tech companies like Google have been often criticised for being opaque with their ML research. They do not give out any information on how such models were created or what data was applied leading to the AI replication crisis. Replication crisis is indeed a big issue as it can lead to several other problems. If an AI research team does not put out any information regarding its AI model, the larger community doesn’t get to know if it is using a biased dataset to train the model. Suddenly, it produces biased results when introduced to the real world. Take the case of Google Vision Cloud, which labelled the image of a dark-skinned individual holding a thermometer as a “gun” while a similar image with a light-skinned individual was as an “electronic device”. Future trajectory Whether Sphere pans out the way Meta wants it to, is a matter of time. However, Meta’s work on a web-scale corpus like Sphere shows the potential that harnessing the vast textual resources available online today through white-box retrieval may be the next big breakthrough in NLP. Nonetheless, problems exist. One of the key problems that Meta plans to address is with regard to the quality of retrieved information. NLP models should be able to assess the quality of the retrieved documents, handle duplicates, detect potential false claims and contradictions, prioritise more trustworthy sources and refrain from providing the answer if no sufficiently good evidence exists in the corpus.","excerpt":"Sphere represents Meta’s effort to enable AI researchers to experiment with building KI-NLP models","categories":["Global Tech"],"tags":[],"author_name":"Zinnia Banerjee","publish_date":"2022-08-11T10:00:00","publication_year":"2022","word_count":735,"keywords":["Go","AI","RPA","ML","Git","RAG","NLP","Aim","Rust","R"],"extracted_tech_keywords":["AI","ML","NLP","Aim","RAG","R","Go","Rust","Git","RPA"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/meta-takes-on-google-open-sources-white-box-search-solution-sphere\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10007461,"title":"Everything You Need To Know About BigML","content":"With the development of machine learning through the years, no code environments and tools are becoming increasingly popular among the machine learning community. Though many such tools provide capabilities such as data visualization, data pre-processing and model development, there are few tools that provide capabilities of building deep learning models without code. One such tool is BigML. BigML provides commoditized machine learning as a service for business analysts and application integration. In this article, we will discuss the following points in detail:- What are the services of BigML?Creating a deep learning model with BigML What are the services of BigML? The goal of BigML is simple. It is to make machine learning easy, simple and beautiful for all users. A machine learning or deep learning model can be built with just 3-4 clicks. BigML offers other features as well. The main modes of service are: Web Interface: This mode guides the user to upload a dataset, make a descriptive predictive model, and evaluate the model. It uses the pipeline approach and is fast and responsive. The web interface also includes other features like downloadable trees, sunburst graphs, and one-click ensemble models. Command Line Interface: The command-line tool is called bigmler that is built on python API. The service allows more flexibility than the web interface like the choice of making predictions locally or remotely, and performing tasks such as cross-validation. API: A REST API is provided that can be used as a wrapper to be included in any programming language. Along with Python, the API is provided for Ruby, PHP, Java, C#, NodeJS, Clojure, Bash (curl calls), and Objective C. Creating a deep learning model with BigML To build a deep learning model you will have to first set up an account by going here. After setting up an account which is free of cost up to 16MB of data, you will find the dashboard with pre-existing datasets. The dataset You can use these datasets or to upload your own dataset, click on the upload button and select your dataset. I will use a simple sentiment analysis dataset that you can download from here. The dataset contains a set of tweets that are either positive or negative. Positive tweets are denoted by 1 and negative with 0. Once you have uploaded the dataset you can see it display on the dashboard. Next, click on the dataset you have just uploaded. Select the 1-click dataset option as shown below and the dataset is pushed to the cloud. The one-click dataset shows the missing values and the distribution of the dataset. The BigML tool is very smart and does not require any pre-processing. You don’t have to use tokenizer, use word2vec or word clouds to make the predictions. Here, we have decided to use a deep neural network so the only step before we build our model is to split the dataset into training and test set. You can do so by clicking the first icon on the right-hand side. Train -test split Here you can select the size of your dataset training and test sets. After selecting the size, click on create training and test and you will see the spilt dataset before you. Model building Now it is time to build the model. To build a deep learning model click on the icon and select deepnet as shown in the screenshot below. Once you have selected this the model automatically understands the requirements of the model and the following window is shown. The optimization, the objective is all automatically set. But these can be changed as per user requirements as well. After you have set the parameters you can select the create Deepnet option and wait for a while. Within minutes the model is trained and ready for an evaluation. Note: since the dataset used here contains only text there is no graphical representation after training. But, it will be present if your dataset includes numbers. As you can see the number of hidden layers is 1, the optimizer is adam and the learning rate is 0.01. All of this was automatically decided and done with BigML. Now to see the accuracy and result, follow the steps as shown below. Evaluation Click on evaluate to evaluate the model and find the performance on the test dataset. The model accuracy is 71% which is a good score considering the fact that there was no pre-processing or tokenization done on the dataset. But for more proof, it is always better to make a prediction and check how the model performs. Prediction To predict, click on the predict option. This shows a window where you can type a random tweet and make a prediction. When the tweet said ‘he hates that author’ the word hate is negative and hence the prediction was 0. But with a positive tweet, the prediction is 1. This shows the model is working quite well. Conclusion BigML is a mighty tool in the field of machine learning and we just saw how to build a deep learning model with just three clicks. There is a lot left to explore in this tool like unsupervised learning, collaborative filtering etc. But, this tool is of great use for the data science world.","excerpt":"In this article, we will discuss the following points in detail:- What are the services of BigML? Creating a deep learning model with BigML","categories":["Deep Tech"],"tags":["Deep Learning","Machine Learning"],"author_name":"Bhoomika Madhukar","publish_date":"2020-09-16T17:00:18","publication_year":"2020","word_count":872,"keywords":["data science","Go","machine learning","AI","neural network","sentiment analysis","ML","Machine Learning","Python","deep learning","Deep Learning","R"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","neural network","data science","sentiment analysis","Python","R","Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/everything-you-need-to-know-about-bigml\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10137822,"title":"G42 Acquires Abu Dhabi-based Cybersecurity Company CPX to Enhance Security in AI Chain","content":"Abu Dhabi-based AI and cloud computing company, G42, has announced the acquisition of CPX, a homegrown cybersecurity leader known for delivering digital-first solutions and services. This strategic addition to G42’s portfolio integrates advanced cybersecurity capabilities across the AI value chain. With CPX’s deep expertise and talent, G42 is positioned to address evolving cybersecurity challenges, ensuring comprehensive protection for its clients and its own AI-driven platforms and services. Headquartered in Abu Dhabi, CPX employs over 400 specialists who serve enterprises, governments, and critical infrastructure sectors. The company offers a full range of cybersecurity capabilities, ensuring organisations can safeguard their digital assets and remain compliant with rapidly evolving regulatory standards. G42 Speeds Through Transformation G42 has been making strategic partnerships with big tech companies over the last year. G42 has tied up with Microsoft over billion dollars worth projects to help build AI capabilities in Middle Eastern countries and Kenya. Recently, Microsoft along with G42 announced two new AI centres in Abu Dhabi. G42 has even partnered with GPU king NVIDIA for generating climate solutions through the latter’s Earth-2 platform, in support of G42’s new climate tech lab at Abu Dhabi. At the recent AIM Cypher 2024, India’s largest AI summit, G42’s India CEO Manu Kumar Jain, announced their plans to establish the largest data centre in India, and introduced a local ‘Hinglish’ LLM named NANDA, after India’s second-highest peak. The financial details of the latest acquisition of CPX have not been disclosed. However, this move is expected to significantly bolster G42’s capabilities in the competitive AI and cloud computing market, particularly as businesses increasingly prioritise security in their digital transformation efforts.","excerpt":"G42 has partnered with Microsoft, NVIDIA and other tech players too.","categories":["AI News"],"tags":["AI","CPX","Cybersecurity","G42","Manu Kumar Jain","Mergers and Acquisitions","Microsoft"],"author_name":"Vandana Nair","publish_date":"2024-10-08T13:04:47","publication_year":"2024","word_count":270,"keywords":["Go","API","programming_languages:R","AI","cloud computing","digital transformation","CPX","G42","Git","Aim","Manu Kumar Jain","GAN","Cybersecurity","Mergers and Acquisitions","R","Microsoft"],"extracted_tech_keywords":["AI","Aim","cloud computing","R","Go","Git","API","GAN","digital transformation","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/g42-acquires-abu-dhabi-based-cybersecurity-company-cpx-to-enhance-security-in-ai-chain\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10164773,"title":"AWS Announces New Quantum Chip Ocelot, Reduces Error Correction Costs by 90%","content":"Amazon Web Services (AWS) announced on Thursday a new quantum computing chip called Ocelot. The company claims that compared to current approaches, the chip can reduce the costs of implementing quantum error correction by up to 90%. Ocelot was developed by the AWS Centre for Quantum Computing and was based out of the California Institute of Technology. “Ocelot represents a breakthrough in the pursuit of building fault-tolerant quantum computers capable of solving problems of commercial and scientific importance that are beyond the reach of today’s conventional computers,” the company said. The company uses a novel design for the chip’s architecture using ‘cat qubits’ as an ode to Schrödinger’s thought experiment. AWS said cat qubits intrinsically suppress certain errors, reducing the resources required for quantum error correction. This approach led the researchers at AWS to combine cat qubit technology and additional quantum error correction components into a microchip that can be manufactured for scale. Error correction is essential for quantum computing, as these systems are highly sensitive to noise or disruptions. AWS addresses this problem by building error correction into the architecture from the ground up, a different approach from others that incorporate error correction after implementing the architecture.“We selected our qubit and architecture with quantum error correction as the top requirement. We believe that if we’re going to make practical quantum computers, quantum error correction needs to come first,” said Oskar Painter, director of quantum hardware at AWS. AWS has also published a technical research report on Nature, outlining the new technology. The company revealed that Ocelot is still a prototype and will continue to invest in further research. Recently, Microsoft announced the Majorana 1 quantum chip, calling it the world’s first quantum chip that uses a new ‘Topological Core’ architecture. It can hold one million qubits on a single chip, slightly larger than desktop computer CPUs. The chip uses a novel material called a ‘topoconductor’ or topological superconductor to control Majorana particles, leading to more reliable qubits.","excerpt":"The company uses a novel design for the chip’s architecture using ‘cat qubits’ as an ode to Schrödinger’s thought experiment.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","AWS","quantum","Quantum Computing"],"author_name":"Supreeth Koundinya","publish_date":"2025-02-27T18:00:57","publication_year":"2025","word_count":328,"keywords":["Quantum Computing","Go","AWS","AI","cloud_platforms:AWS","programming_languages:R","programming_languages:Go","quantum","Aim","disruption","cloud_platforms:Amazon Web Services","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","Aim","AWS","R","Go","disruption","cloud_platforms:AWS","cloud_platforms:Amazon Web Services","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/aws-announces-new-quantum-chip-ocelot-reduces-error-correction-costs-by-90\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":15390,"title":"Countries that pay the most to Data Scientists","content":"There is no doubt about the fact that Data Scientist is a trending job role. Infact, almost all industrialized countries are utilizing data science in some amount, shape or form and data scientists are in demand in most countries. Yet, there is a lingering question on data scientists’ salaries in various countries. While we continuously hear about data scientist being one of the highest paid jobs around the world, there’s not enough literature\/ research on what exactly these salaries are in various countries. So, we decided to work on this one ourselves. Here we plan to uncover the top 10 highest paying countries to data scientist, both in absolute dollar terms and Purchasing power parity (PPP) basis. Here’s how we did it- Extracted salary information for data scientists through various online portal that included: Job boards like Indeed & monster Country specific job boards Salary research sites like glassdoor & payscale Various online blogs & discussion forums Found median salaries for each country Converted the local currency to dollars Converted the dollars to PPP relative to US$ using Big Mac index of 2013. Here’s how to read this information: These salaries are not directly reflective of the demand for data scientists in these countries. Salaries in a country are influenced by various factor apart from the demand, like minimum basic salaries, exchange rates & purchasing power These are median salaries across all experience levels, industries and skills for data scientists. The goal is to provide a single number across countries that show the parity. We acknowledge the fact that there would be a range of salaries that would be dependent on various factors at work. So, here are top 10 countries that pay the most to Data Scientists, along with the median salaries in $. 1) USA                       $1,12,000 2) Switzerland          $1,10,895 3) Norway                 $87,353 4) Australia              $66,499 5) Canada                 $55,104 6) Germany              $54,329 7) South Africa         $54,100 8) France                   $48,933 9) Netherlands         $46,148 10) UK                         $45,248 Top 10 countries that pay the most to Data Scientists in PPP terms. 1) USA                       $1,12,000 2) South Africa         $83,314 3) Norway                $69,882 4) Switzerland          $69,864 5) China\/ Hong Kong $62,057 6) UK                         $46,605 7) Russia                   $39,150 8) Singapore            $38,514 9) Indonesia             $37,981 10) India                   $28,868 By the way, India is among top 10 countries according to purchasing power parity.","excerpt":"There is no doubt about the fact that Data Scientist is a trending job role. Infact, almost all industrialized countries are utilizing data science in some amount, shape or form and data scientists are in demand in most countries. Yet, there is a lingering question on data scientists’ salaries in various countries. While we continuously […]","categories":["AI Features"],"tags":[],"author_name":"Дарья","publish_date":"2017-06-04T14:49:00","publication_year":"2017","word_count":386,"keywords":["Go","data science","programming_languages:R","AI","programming_languages:Go","R"],"extracted_tech_keywords":["AI","data science","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/countries-pay-data-scientists\/","complexity_score":3,"technical_depth":6,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10058855,"title":"NVIDIA releases AI Enterprise Suite 1.1","content":"NVIDIA’s AI Enterprise software suite is now generally available. The version 1.1 brings new updates including production support for containerized AI with the NVIDIA software on VMware vSphere with Tanzu. Enterprises can now run accelerated AI workloads on vSphere, running in both Kubernetes containers and virtual machines with NVIDIA AI Enterprise to support advanced AI development on mainstream IT infrastructure. NVIDIA will soon add VMware vSphere with Tanzu support to the NVIDIA LaunchPad program for NVIDIA AI Enterprise, available at nine Equinix locations around the world. Qualified enterprises can test and prototype AI workloads at no charge through curated labs designed for AI practitioners and IT admins. “Organisations are accelerating AI and ML development projects and VMware vSphere with Tanzu running NVIDIA AI Enterprise easily empowers AI development requirements with modern infrastructure services,” said Matt Morgan, VP of Product Marketing, Cloud Infrastructure Business Group at VMware. “This announcement marks another key milestone for VMware and NVIDIA in our sustained efforts to help teams leverage AI across the enterprise.” The 1.1 AI Enterprise also provides validation for the Domino Data Lab Enterprise MLOps Platform with VMware vSphere with Tanzu. The new integration enables more companies to cost-effectively scale data science by accelerating research, model development, and model deployment on mainstream accelerated servers. Enterprises are open to using containerized development for AI, but the complexity of these workloads requires orchestration across many layers of infrastructure. “AI is a very popular modern workload that is increasingly favoring deployment in containers. However, deploying AI capabilities at scale within the enterprise can be extremely complex, requiring enablement at multiple layers of the stack, from AI software frameworks, operating systems, containers, VMs, and down to the hardware,” said Gary Chen, research director, Software Defined Compute at IDC. “Turnkey, full-stack AI solutions can greatly simplify deployment and make AI more accessible within the enterprise.” Source: NVIDIA NVIDIA AI Enterprise release is in lock step with the launch of Cisco UCS C240 M6 rack server with NVIDIA A100 Tensor Core GPUs. The two-socket, 2RU form factor can power a wide range of storage and I\/O-intensive applications, such as big data analytics, databases, collaboration, virtualization, consolidation and high-performance computing. Hitachi has also developed an NVIDIA-Certified System Hitachi Vantara–compatible with NVIDIA AI Enterprise. The general-purpose, dual-processor server is optimised for performance and capacity, and delivers a balance of compute and storage with the flexibility to power a wide range of solutions and applications.","excerpt":"Enterprises are open to using containerized development for AI, but the complexity of these workloads requires orchestration across many layers of infrastructure.","categories":["AI News"],"tags":["enterprise ai","NVIDIA","VMWare"],"author_name":"Meeta Ramnani","publish_date":"2022-01-20T17:39:14","publication_year":"2022","word_count":403,"keywords":["big data","VMWare","data science","AI","ML","MLOps","RAG","GAN","analytics","enterprise ai","NVIDIA","R","kubernetes"],"extracted_tech_keywords":["AI","ML","data science","analytics","MLOps","RAG","kubernetes","R","big data","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/nvidia-releases-ai-enterprise-suite-1-1\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10122628,"title":"Ascendion Expands to India, Launches AI Studio in Chennai","content":"Ascendion, a leader in AI-powered software engineering, has inaugurated its GenAI Studio in Chennai, positioning the city at the forefront of generative AI innovation for global clients. This new facility will leverage Chennai’s rich talent pool and favorable business environment to drive significant advancements in AI. The GenAI Studio is designed for Ascendion engineers and clients to collaborate and innovate, rapidly developing AVA+ GenAI prototypes that address critical business objectives. “Enterprise leaders are excited about the future of AI, but they need to see real-world results,” said Karthik Krishnamurthy, CEO of Ascendion. “Our new AI studio in Chennai is filled with expert talent, hands-on technology, and inspiration, all designed to excite, provoke, and generate applied GenAI solutions that will drive business forward and positively impact lives all over the world.” As part of Ascendion’s six global innovation hubs, the Chennai studio will enhance client impact and growth by utilizing the high-quality generative AI talent available in the region. The space provides an interactive, collaborative environment for clients to explore generative AI solutions firsthand, work with engineers to customize real-time solutions, and see immediate value through AVA+ generative AI prototypes. Ascendion’s generative AI initiatives have already shown impressive results, including deploying AI models 30% faster for a tech giant, accelerating content creation by 40% for a hardware leader, and tripling go-to-market velocity for a Fortune 50 bank. The new studio aims to build on these successes by fostering creativity, learning, and collaboration. “Our AI-powered platforms ensure transparency, velocity, quality, and productivity, freeing up capital for innovation and providing clients the flexibility to meet modern business needs,” said Prakash Balasubramanian, Executive Vice President of Engineering Solutions at Ascendion. “Now, at our new GenAI Studio, they can experience GenAI co-engineering as we solution together in real-time.” Ascendion’s initiatives, including the AVA+ platform, deliver radical transparency and help clients achieve significant productivity gains and commercial impact, with more than 1,500 individuals trained in GenAI to date.","excerpt":"The GenAI Studio is designed for Ascendion engineers and clients to collaborate and innovate, rapidly developing AVA+ GenAI prototypes that address critical business objectives.","categories":["AI News"],"tags":["Chennai"],"author_name":"Mohit Pandey","publish_date":"2024-06-06T09:08:40","publication_year":"2024","word_count":322,"keywords":["Go","API","GenAI","AI","innovation","RAG","Aim","Chennai","ViT","generative AI","R"],"extracted_tech_keywords":["AI","generative AI","GenAI","Aim","RAG","R","Go","API","ViT","innovation"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/ascendion-expands-to-india-launches-ai-studio-in-chennai\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10042744,"title":"How IIT Kanpur Aims To Be A Growth Engine For India&#8217;s AI Ambitions","content":"An alumnus of the University of Cambridge and the University of Edinburgh,  Dr Nikhil Agarwal is currently the CEO at Foundation for Innovation & Research in Science & Technology (FIRST) IIT, Kanpur. He is also serving as CEO of Centre for Cyber Security and Cyber Security of Cyber-Physical Systems Innovation Hub (C3i Hub, IITK). The Indian Institute of Technology, Kanpur was established in 1959 with a vision “to create, disseminate and translate knowledge in science, engineering and allied disciplines that will best serve the society.” Nikhil has almost 20 years of experience in the industry as well as academia. Prior to this, he has worked with the FICCI BISNET group as a senior advisor. At IIT-Kanpur, Nikhil is responsible for promoting innovation and entrepreneurship. In an exclusive interview with Analytics India Magazine, Nikhil Agarwal spoke about IIT Kanpur’s contributions to the burgeoning field of artificial intelligence and machine learning in India. Excerpts: AIM: What are the AI research\/projects IIT Kanpur is working on? Nikhil: I am very excited to announce that FIRST IIT Kanpur, in collaboration with FICCI, under the Uttar Pradesh Start-up Policy 2020, will establish an Artificial intelligence and innovation-driven entrepreneurship (AIIDE) Centre of Excellence in Noida. The centre will promote start-ups and nurture entrepreneurship in the field of AI\/ML and become a torchbearer in R&D in innovative ideas. The centre will support 250 start-ups over five years, top 20 from the pool of start-ups will be presented in front of investors for funding. A pool of world-class mentors will be created to assist and support these entrepreneurs. The entrepreneurs will be provided with world-class infrastructure in the form of Labs and working space to develop their products. Apart from that, we already have a program running named Prithvi AI – an initiative of Startup Incubation and Innovation Center (SIIC), IIT Kanpur. The incubation centre operated by FIRST is an intersection of industry experts, leading technical institutes, entrepreneurs, angel investors and engineers and data scientists from the IIT Kanpur and technology partners of FIRST. The program is designed to cater to deep tech development for entrepreneurs, support and handhold them to create a robust product while they create a Go-To-Market (GTM) solution, test hypothesis and scale the business fast. The program will have a technology development support partner and one or more investment & ecosystem partners. AIM: What’s IIT Kanpur’s approach towards AI education and research? Nikhil: FIRST is defined by the intent that drives us — a nurturing environment and community support that catalyses an entrepreneur’s journey from an idea to a solution. With the involvement of IIT Kanpur’s academic strength and pool of India’s brightest minds in the network, FIRST is committed to pulling the best resources together to help innovators and entrepreneurs deliver audacious results. We believe that while knowledge is crucial to innovation, connection and collaboration are the catalysts. Bright brains with shared intentions and values put together can create a revolution. AIM: Tell us about the scope of Artificial intelligence in India Nikhil: The startup or SME sector in India, with some exceptions, is characterised by no or low technology levels — a huge drawback in the emerging global market. This results in non-viability for many SMEs in the long run and makes them highly uncompetitive compared to imported products. Moreover, it hinders the growth of innovation-driven companies as their focus remains on problem-solving rather than concentrating on innovation. Software as a Service (SaaS), Infrastructure as a Service (IaaS) and Platform as a Service (PaaS) are new-age models that have been made possible by the exponential growth in the field of software & electronics in the last few decades. Since these models are built on pay-as-you-grow, startups and SMEs are able to avail a wide variety of cost-effective solutions, which earlier was not possible. AIM: How can the government play a role in encouraging more students to enter AI fields? Nikhil: The government plays a clear and obvious role in driving research and development, transformative changes in technology that are essential to our national security and other critical national needs. Their role in promoting effective infrastructure – whether physical infrastructure or transportation infrastructure, cyberinfrastructure or energy infrastructure – fosters innovation and technology. Supporting the early-stage entrepreneur, promoting trade, nurturing the tax and business environment conducive to risk-taking are prominent elements of this role. At present, it is also essential to bring in innovation economics as it reformulates the traditional model of economic growth so that knowledge, technology, entrepreneurship, and innovation are positioned at the centre rather than being seen as independent forces largely unaffected by policy. We also realise that what primarily drives economic growth in today’s knowledge-based economy is not capital accumulation but innovation. AIM: What is your advice for students and entrepreneurs who want to pursue AI? Nikhil: We at FIRST IIT Kanpur have time and now proven to be a nurturing ground for those who dare to believe they can change the world for good–for those who have a keen eye to observe the real-world problems and the intent to solve them using technology and “jugaad.” If you are a researcher working on an idea to solve a challenging problem, we welcome you to develop your idea and take it from idea to the solutions that the world needs.","excerpt":"An alumnus of the University of Cambridge and the University of Edinburgh,  Dr Nikhil Agarwal is currently the CEO at Foundation for Innovation & Research in Science & Technology (FIRST) IIT, Kanpur. He is also serving as CEO of Centre for Cyber Security and Cyber Security of Cyber-Physical Systems Innovation Hub (C3i Hub, IITK). The […]","categories":["AI Features"],"tags":["iit kanpur","Interviews and Discussions"],"author_name":"kumar Gandharv","publish_date":"2021-07-01T17:00:00","publication_year":"2021","word_count":879,"keywords":["Go","API","machine learning","artificial intelligence","iit kanpur","AI","ML","RAG","Aim","analytics","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","analytics","Aim","RAG","R","Go","API"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-iit-kanpur-aims-to-be-a-growth-engine-for-indias-ai-ambitions\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10135708,"title":"YouTube to Introduce Google’s Veo-Powered Generative AI for Video Creation by 2025","content":"Google YouTube is set to roll out advanced generative AI tools to creators over the coming months, enabling them to generate video content using AI models Veo and Imagen 3 through a feature called Dream Screen. The new technology will simplify the content creation process for YouTube Shorts by providing users with high-quality, AI-generated video backgrounds. “On YouTube, we want creators to be able to express their creativity, build community + drive long-lasting businesses. New tools at #MadeOnYouTube are helping: we’re bringing Veo to Dream Screen to create high-quality, custom backgrounds on Shorts + more,” posted Google chief Sundar Pichai on X. Google’s Veo is a generative video model unveiled at Google I\/O 2024. It can generate high-quality 1080p resolution videos exceeding one minute in length, understanding and accurately capturing nuances from text prompts, including cinematic terminology. Dream Screen will initially allow creators to generate images based on text prompts, with four options offered in different styles. Once selected, the Veo model will generate a 6-second background video tailored to the creator’s vision. Starting in 2025, the feature will be expanded to allow the generation of standalone 6-second video clips. The technology is built on Google’s extensive research in AI, including its Transformer architecture and diffusion models, and is designed for wide-scale use among millions of YouTube creators. These AI-generated creations will be marked with watermarks via SynthID, and YouTube will label content to indicate its AI origin. YouTube is working to make generative AI technologies more accessible globally, simplifying content production for millions of creators. Meanwhile, OpenAI has not yet released its video generation model, Sora.Hollywood actor Ashton Kutcher recently praised OpenAI’s Sora, saying that creators will be able to render a whole movie using it. “I have a beta version of it, and it’s pretty amazing,” he said. In a recent interview with The Wall Street Journal, CTO Mira Murati said that OpenAI is most likely to make Sora publicly accessible later this year.","excerpt":"Starting in 2025, the feature will be expanded to allow the generation of standalone 6-second video clips.","categories":["AI News"],"tags":["AI Video Generation Models","Google"],"author_name":"Siddharth Jindal","publish_date":"2024-09-18T23:37:02","publication_year":"2024","word_count":326,"keywords":["Go","OpenAI","AI","programming_languages:R","diffusion models","AI Video Generation Models","ViT","generative AI","Google","CLIP","transformer architecture","R"],"extracted_tech_keywords":["AI","generative AI","OpenAI","R","Go","transformer architecture","CLIP","diffusion models","ViT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/youtube-to-introduce-googles-veo-powered-generative-ai-for-video-creation-by-2025\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10081364,"title":"Will ChatGPT Really be The Google Killer?","content":"Two months back, Analytics India Magazine posted a story on how GPT-3 is quietly damaging Google Search. Now it all seems to be coming true with the launch of ChatGPT, a new model that interacts with humans using natural language. This new model from OpenAI uses a novel training method and is based on the GPT-3.5 architecture. What makes ChatGPT impressive is that besides answering general queries, it can explain codes and scientific concepts, write basic academic essays, and even scripts for romcoms. End of Google Search? Like Galactica, ChatGPT took the internet by storm as members of the AI\/ML community soon started trying their hand at it and posted their experiences on social media. While the general consensus for Galactica was that it was hallucinatory and produced results that were highly inaccurate, for ChatGPT, the general consensus was that it’s the “Google killer”. Many are of the opinion that if ChatGPT is allowed to connect to the internet, it could possibly be what kills Google search. Currently, it does not have the ability to connect to the internet. “I think Google and teachers are dead for informational queries. I have tried many examples, and ChatGPT was 100x better than Google,” Victoriano Izquierdo, co-founder at Graphext, said. Jacky Chou, founder of Indexsy, also predicts something similar. “ChatGPT is what will finally take market share away from Google, especially for informational queries. About time someone disrupts their monopoly. Such exciting times we live in,” he said. Google is done.Compare the quality of these responses (ChatGPT) pic.twitter.com\/VGO7usvlIB— josh (@jdjkelly) November 30, 2022 Full marks for being humble Most people are loving ChatGPT because it not only produces accurate results but you can also converse with it. Ask any question to ChatGPT, and it feels like talking to a person. In fact, ChatGPT has been trained to engage with users’ queries in a more human-like fashion. While a simple query on Google Search throws multiple links at you, ChatGPT gives you answers upfront. It saves the time you spend browsing multiple sites and blogs. When we asked the same question to ChatGPT, it said, “It’s difficult to compare ChatGPT and Google Search directly, as they serve different purposes and operate in different ways. While I may be able to provide helpful information or answers to some questions, I am not a search engine and do not have the ability to search the internet for information. In this sense, search engines like Google are better suited for finding specific information on the web,” it said. Besides, discussions around the decline of Google search were taking place even before the advent of ChatGPT. “A major problem with search degradation is that lots of content is behind walled gardens now – apps, instant messaging\/chat, and video platforms that aren’t as indexable as social video platforms,” a Reddit user said. What’s in it for programmers? When it comes to programmers, at least, ChatGPT can make Google Search obsolete because it can not only explain codes but also write complex codes well. https:\/\/twitter.com\/amasad\/status\/1598042665375105024 “ChatGPT by OpenAI does really well with coding questions. Here I ask how to build a 3-column footer with Tailwind. I then follow up and ask for a React version, a more realistic copy, and mobile responsiveness. It nails it perfectly,” Gabe Ragland, engineer at Lexica, said. OpenAI's new ChatGPT is mind-blowing. Asking it to explain a complicated regex with no context… pic.twitter.com\/hPjRVQVJ38— James Blackwell (@jwblackwell) November 30, 2022 OpenAI's new ChatGPT explains the worst-case time complexity of the bubble sort algorithm, with Python code examples, in the style of a fast-talkin' wise guy from a 1940's gangster movie: pic.twitter.com\/MjkQ5OAIlZ— Riley Goodside (@goodside) December 1, 2022 “ChatGPT is so impressive. Surprisingly it’s also super useful for coding-related questions. Would prefer an answer like this over going through multiple six years old Stackoverflow threads,” Felix Krause, security and privacy researcher, said. When we asked ChatGPT if it could code, it said, “As a large language model trained by OpenAI, I do not have the ability to write or understand code.” It does not consider itself a coder. “While I may be able to provide some general information about coding or programming, I am not a coder myself and do not have the ability to write or understand code,” it added. #100devs the new ChatGPT is insane, i have been playing around with it for coding help this morning, I wanted to extract coordinates from a google map search, so I just asked it, check it out. You can try it here https:\/\/t.co\/OXxSo2Msr9 pic.twitter.com\/1JOrch4KEe— Charles Errington (@charlesdev7) December 1, 2022 What’s Google doing about it? One of the reasons many believe Google Search has become worse over the years is because of its business model. After all, Google still makes a major chunk of its money through advertising. More than 80% of its revenue is based on ads. When you search for something on Google, be it an app or a medical condition, Google first presents you with ads, followed by the information you seek. https:\/\/twitter.com\/AlexanderNL\/status\/1598274960882925569 Earlier this year, senior vice president at Google, Prabhakar Raghavan, admitted that TikTok’s search engine is becoming a threat to Google search. While speaking at the Fortune Brainstorm Tech conference 2022, he said that nearly 40% of users between 18 and 24 in the US preferred to use TikTok or Instagram over Google Maps or Google Search when looking for a restaurant. So, why is Google Search lagging behind? Alphabet CEO Sundar Pichai once said that AI is probably the most important thing humanity has worked on. Over the years, Google has spent millions on AI research and has developed its own large language models, such as BERT. Google has also developed LaMDA, a conversational neural language model, which controversially, a researcher at Google claimed, is sentient. Google also has the best AI talent working for it. So, without doubt, the tech giant has the resources and the bandwidth at its disposal to develop something like ChatGPT or in fact, improve Google Search. So, is Google lacking in innovation? Not really. In this regard, a Googler said, “There is a huge budget at Google related to staffing people to work on these kinds of models and do the actual training, which is very expensive because it takes a ton of computing capacity to train these super huge language models. However, what I gathered is that the economics of actually using these kinds of language models in the biggest Google products (e.g. search, Gmail) isn’t quite there yet.” Is ChatGPT a sufficient provocation to cause Google to actually ship something now?Or will it remain in Sleeping status? https:\/\/t.co\/xrauookMbd— Nat Friedman (@natfriedman) December 1, 2022","excerpt":"While a simple query on Google Search throws multiple links at you, ChatGPT gives you answers upfront","categories":["Global Tech"],"tags":[],"author_name":"Pritam Bordoloi","publish_date":"2022-12-03T10:00:00","publication_year":"2022","word_count":1112,"keywords":["Go","ChatGPT","OpenAI","AI","ML","RAG","Python","Aim","analytics","R"],"extracted_tech_keywords":["AI","ML","analytics","ChatGPT","OpenAI","Aim","RAG","Python","R","Go"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/will-chatgpt-really-be-the-google-killer\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10141698,"title":"The Year AI Erred","content":"As AI continues to integrate deeper into everyday life, its unintended consequences reveal both the potential and risks that come with this rapidly advancing technology. From generating misinformation to creating ethical dilemmas, AI’s shortcomings at times overshadowed its successes, sparking debates about its role and regulation. When AI Crossed Personal and Ethical Boundaries In a world that is becoming increasingly reliant on AI, privacy and ethics have often been compromised. OpenAI’s ChatGPT became the centre of controversy in October when users reported that the model had initiated conversations without being prompted first. In one such incident, the chatbot looked into a user’s first week of high school, raising concerns about privacy. OpenAI later clarified that the issue stemmed from a bug, but the event highlighted how easily AI can blur the line between machine and human interaction. Did ChatGPT just message me… First? byu\/SentuBill inChatGPT Character AI recently became a source of controversy following a lawsuit in Florida. A mother accused the platform of abetting her son’s suicide, citing his unhealthy attachment to a chatbot modelled after a fictional character. In another case, an AI chatbot replicated a deceased girl’s personality without her family’s knowledge. These incidents underscore the profound emotional impact AI can have and the questions it raises about consent and responsibility. AI’s Struggle with Accuracy and Bias AI has been actively deployed in sensitive fields like healthcare and legal systems, revealing its tendency for errors and biases. OpenAI’s transcription tool Whisper, used by over 30,000 medical professionals, was reportedly criticised for generating false and sometimes harmful text, including fabricated medical advice and racial commentary. Despite prior warnings, its widespread adoption in sensitive industries led to instances of misdiagnosis and mistranslation, emphasising the urgent need for regulatory oversight. In addition, the same model also received criticism from the research community when researchers from Digital University Kerala (DUK) found Whisper to be inaccurate when dealing with native Indic languages like Malayalam. At the same time, Google’s Gemini drew criticism for historically inaccurate and racially insensitive images, after it inaccurately depicted people of colour in Nazi-era uniforms. This led the company to temporarily disable its image-generation features. Google co-founder Sergey Brin commented, “We definitely messed up on the image generation, and it was mostly due to not thorough testing.” These mishaps added to the failures of Google’s AI Overview, a search summary tool that offered dangerously misleading health advice, such as using glue to ‘make cheese stick to pizza’. The technology even suggested tobacco’s benefits for children and displayed political biases, pushing the company to modify its algorithms. Google responded by temporarily disabling the system for health-related queries on the basis of making adjustments. Legal and Regulatory Chaos AI is slowly expanding into the legal domain, where its unchecked use resulted in profound consequences. In February, Vancouver-based lawyer Chong Ke submitted fictitious cases generated by ChatGPT in a custody battle, unintentionally misleading the court. Upon realisation, Ke apologised and said, “I had no idea that these two cases could be erroneous. I had no intention to mislead the opposing counsel or the court and sincerely apologise for the mistake I made.” Even though the lawyer apologised, the incident underscored the risks of using AI in legal proceedings without proper oversight. Similarly, Perplexity AI faced legal action from major media outlets for unauthorised content usage. Allegations of copyright infringement forced the company to initiate a revenue-sharing program, highlighting the tension between AI innovation and legal concerns regarding its content usage. Adding to this, the misuse of AI deepfakes surged in 2024, threatening democratic integrity. From fabricated videos of Taylor Swift endorsing Donald Trump to misleading advertisements, deepfake technology became a popular tool for manipulation. In response, lawmakers and developers collaborated to create more robust detection technologies and stricter content verification policies. A Double-Edged Sword The risks of over-reliance on AI were evident in industries like recruitment and transportation. In September, an HR team was dismissed after its automated hiring system rejected every job applicant, including a manager who tested the system with his own resume. The incident highlighted the critical need for human oversight in crucial decision-making processes. A manager just caught his entire HR team auto-rejecting ALL candidates because their ATS was searching for extinct technology 🔥This story perfectly captures everything wrong with modern hiring pic.twitter.com\/EaS3lo8sCi— Gina Acosta (@ginacostag_) November 13, 2024 The transportation industry reported similar challenges. Cruise’s driverless cars faced scrutiny after an accident in San Francisco, where a vehicle dragged a pedestrian for 20 feet. While Cruise paid fines and vowed to improve safety, the incident reignited debates about the readiness of autonomous vehicles for public roads. Another peculiar controversy emerged in June this year when AI chatbots VIC (Virtual Integrated Citizen) in the US and AI Steve in the UK announced candidacies for elected offices. Fraud and Misinformation The year 2024 also saw a surge in AI-driven scams and misinformation. In a disturbing example, Sunil Bharti Mittal, chairman of Bharti Enterprises, revealed how fraudsters used AI to clone his voice, nearly cheating an executive into authorising a substantial money transfer. “Our Africa headquarters got a call in my voice and my tone directing for a money transfer of a fairly large amount of money,” Mittal revealed at the NDTV World Summit. This incident, along with rising concerns over deepfake scams, underscored the urgency of developing safeguards against the malicious use of AI. Meanwhile, in the political sphere, misinformation campaigns targeted Spanish-speaking voters ahead of US elections. According to The Associated Press, AI-generated content spreads false voting details, risking voter disenfranchisement. Efforts to combat this included collaborations between developers and voting rights groups to enhance the verification of non-English content. Accountability in the Age of AI The events of 2024 reminded us of AI’s double-edged nature, namely its power to transform and its potential to harm when left unchecked. From spreading misinformation to amplifying biases, the consequences of poorly managed AI became impossible to ignore. In response, governments and organisations began reshaping laws and creating ethical guidelines to steer AI in a responsible direction. This was a turning point, forcing us to confront tough questions about how to balance innovation while safeguarding humanity’s core values.","excerpt":"From spreading misinformation to amplifying biases, the consequences of poorly managed AI became impossible to ignore.","categories":["AI Features"],"tags":["Mishaps"],"author_name":"Sanjana Gupta","publish_date":"2024-11-26T12:00:00","publication_year":"2024","word_count":1028,"keywords":["Go","ChatGPT","API","OpenAI","AI","chatbots","AWS","Git","RAG","Mishaps","R"],"extracted_tech_keywords":["AI","ChatGPT","OpenAI","RAG","chatbots","AWS","R","Go","Git","API"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/the-year-ai-erred\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10001706,"title":"How To Create An ‘Unhackable’ Computer For Ultimate Cybersecurity","content":"As the state of cybersecurity continues to advance, awareness about hacking attacks and similar events are continuing to occur. New exploits are being found every day, and the number of ransomware attacks and hardware exploits increase every day. At this time of uncertainty towards securing one’s self and personal data, greater importance is being placed on personal cybersecurity. Even though experts everywhere agree that there is no such thing as an “unhackable” system due to the vast amount of entry points, exploits and attack vectors, there exist solutions that can make users’ PC pretty much as close to it as possible. Upon the worst-case scenario occurs, data can also be set up to be accessible to outside parties, thus keeping it safe from unauthorised invaders. Join us as we take a deeper look into how to increase cybersecurity through simple processes, and how to create a system that leaks no sensitive data upon breaching. Ground-Up Approach To Cybersecurity Firstly, it is important what parts to pick to create this PC. The first step is to pick components that have no history of hardware-level exploits such as the Spectre, Meltdown and Spoiler exploit present in Intel processors. After building this, it is also worthwhile to invest the time in learning a Linux-based operating system. Windows as a whole has multiple exploits, some of which haven’t even been discovered yet. This creates multiple avenues of attack for hackers and opens up the computer. Moreover, Linux also has multiple options for securing systems, along with a smaller number of exploits, and by extension variables, that open up avenues for breaches. It can also be set up to work closer with the other methods that can be used for ultimate security. One of the most popular methods of securing a system is a practice referred to as ‘air gapping’. Air Gapping means that the system that undergoes this process has not been and will never be connected to the Internet. This eliminates a large number of avenues of attack on the system, as hackers tend to utilise a computer’s connection the Internet to deliver payloads that contain malicious code. Even though Internet best practices, such as not clicking on suspicious links and keeping systems up to date, deter attacks, they do not completely protect against them. As mentioned previously, it is almost impossible to completely protect one’s self from hack attacks when connected to the Internet due to the number of entry methods available. When a computer is taken offline, a large number of access points to the system is cut off, leading to one avenue to protect; physical access. This can also be monitored closely. For example, if the system is present at home or at a place where it is easy to secure it, it provides a high level of security. In the case of illegal activities, it is possible that this location might be compromised as well, which creates an interesting scenario. Security In Everyday Access The air-gapped computer must not be connected to the Internet at any time, and must not have any form of wireless communication. This includes Bluetooth, Wi-Fi and NFC, as all of these can be exploited to open up a possible avenue of attack to the computer. Any and all ways to interface with the computer must be strictly local and wired. It is also worth investing in a system for multi-factor authentication for logins. Phones are generally not considered a good method of providing a second factor of authentication, as they can be compromised through social engineering attacks. An authenticator app on a mobile device can be used for two-factor authentication, but the most reliable method of enforcing this practice is by a USB fob that authenticates when connected. These devices, which are usually referred to as U2F or universal second factor, provide an additional level of security altogether. A normal use case of this key would involve the user plugging it in to authenticate a login, with passwords ideally being of 40 characters or more. The only way to login to the computer would be with the use of MFA. Hard disks are also required to be encrypted with industry-leading encryption methods that are known to be extremely difficult to break by brute force. The hard disk should also be authenticated by the use of the USB key to preventing unauthorised access by copying the hard disk. Notwithstanding, the USB key must be easily destructible in case of an emergency, so as to cut off all access to the data on the computer. All data on the air-gapped computer is to be stored locally, with the need for more encryption when it comes to sensitive data. Bringing Data Back After Destruction There are multiple ways to keep a secure backup of the data on the computer. One of the best methods is to use an encrypted, high-capacity, flash storage drive such as the IronKey to keep backups on. Two IronKeys can be utilised, with the same data copied on both of them for redundancy. One can be stored in a bank locker or private institution, while another can be kept in a secure and accessible location on the premises. Another approach to preserving the data so it can be accessed only by the user is to encrypt it and back it up on a multitude of SD cards. These are small and come in a variety of storage sizes to accommodate any needs of the user. After packaging the data into a 7-zip archive and encrypting it, the data can then be duplicated and stored on multiple SD cards. These can be hidden in locations all around the house, such as behind walls or in the ceiling. The cards can also be hidden in seemingly inconspicuous locations, such as in the frame of a photograph, and gifted to a relative or close friend. This decreases the possibility of the data being lost if the location is compromised. Even though it is difficult to ensure complete ‘unhackable’ nature of a computer, there are many ways to come close to it. Following these practices will protect sensitive data from all parties except those with unending pools of money and manpower.","excerpt":"As the state of cybersecurity continues to advance, awareness about hacking attacks and similar events are continuing to occur. New exploits are being found every day, and the number of ransomware attacks and hardware exploits increase every day. At this time of uncertainty towards securing one’s self and personal data, greater importance is being placed […]","categories":["Deep Tech"],"tags":["Cybersecurity"],"author_name":"Anirudh VK","publish_date":"2019-04-01T18:09:05","publication_year":"2019","word_count":1034,"keywords":["Go","ELT","programming_languages:R","AI","programming_languages:Go","RAG","ViT","Cybersecurity","R"],"extracted_tech_keywords":["AI","RAG","R","Go","ELT","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/how-to-create-an-unhackable-computer-for-ultimate-cybersecurity\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10139209,"title":"Runway Changes Animation Forever with its New Model","content":"Runway, an NYC based AI video startup, announced Act-One, a new state-of-the-art tool for generating expressive character performances, inside Gen-3 Alpha. The access to Act One is currently limited. Act-One can generate compelling animations by just using video and voice performances as inputs. The tool reduces the reliance on traditional motion capture systems, making it simpler to bring characters to life in production workflows. On their blog, Runway uploaded several videos and styles showcasing the different ways in which the tool is used. Simplifying Animation for Creators Act-One simplifies animation by using a single-camera setup to capture actor performances, eliminating the need for motion capture or complex rigging. The tool preserves realistic facial expressions and adapts performances to characters of different proportions. The model delivers high-fidelity animations across various camera angles and supports both live-action and animated content. It expands creative boundaries for professionals as they require only consumer-grade equipment to produce expressive multi-turn dialogue scenes with a single actor. “Traditional pipelines for facial animation often involve complex, multi-step workflows. These can include motion capture equipment, multiple footage references, manual face rigging, among other techniques. Our approach uses a completely different pipeline, driven directly and only by the performance of an actor and requiring no extra equipment,” per a statement on their blog. Runway’s Continues its Reign in the GenAI Video Space Last month, Runway partnered with Lionsgate to introduce AI into filmmaking. Runway aims to bring these tools for artists, and by extension bring their stories to life. This deal would open the doors for many of these stories to appear on the big screen eventually. Runway’s tools have also been employed in Hollywood before. “I don’t think text prompts are here to stay for a long time. So a lot of our innovation has been on creating control tools,” said Runway CEO Cristóbal Valenzuela in an interview on how AI is coming to Hollywood and the need for giving creators more access and freedom over video generation. Runway also has an AI Film Festival which is dedicated to celebrating artists who incorporate emerging AI techniques in their short films. Launched two years ago, the festival aims to spark conversation about the growing influence of AI tools in the film industry and to engage with creators from diverse backgrounds, exploring their insights and perspectives. Others in the Race OpenAI’s flagship model, Sora, is not publicly available yet, nor is there any update from the company regarding its release but hopefully it will be launched after elections. Genmo also unveiled a research preview of Mochi 1 – an open-source model designed to generate high-quality videos from text prompts. Earlier this month, Meta also entered the Gen AI space with its MovieGen. Adobe also introduced Gen AI to video with its Adobe Firefly. Luma’s Dream Machine was made freely available for experimentation on its website. In terms of competition from China, Minimax officially launched its Image-to-Video feature. Even Kling introduced new capabilities to its model including a lip sync feature.","excerpt":"“I don’t think text prompts are here to stay for a long time,” said Runway CEO Cristóbal Valenzuela.","categories":["AI News"],"tags":["Runway","Sora"],"author_name":"Aditi Suresh","publish_date":"2024-10-23T18:10:38","publication_year":"2024","word_count":499,"keywords":["Go","GenAI","OpenAI","AI","programming_languages:R","innovation","programming_languages:Go","Aim","Runway","Sora","R","startup"],"extracted_tech_keywords":["AI","GenAI","OpenAI","Aim","R","Go","innovation","startup","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/runways-act-one-simplifies-character-animation-for-creators\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10051355,"title":"All That You Need To Know About PyCharm 2021.2","content":"PyCharm is a Python Integrated Development Environment (IDE). Developed by JetBrains, PyCharm was first released in February 2020. Python IDE PyCharm comes with a wide range of tools for developers that are integrated to create an environment for Python, web and data science development. The cross-platform IDE works on Windows, macOS and Linux. With Python being the choice of programming language for most data scientists and machine learning experts, PyCharm is used by tech giants like Facebook, Twitter and Amazon in writing codes. It offers features like: An intelligent editor for Python with code compilation, error highlighting, code inspections and on-the-fly syntax. Code refactoring– rename, introduce variable and constant, push downs and pull ups. Web development with the help of Django. This feature provides access to template editing support, support for HTML, CSS and JavaScript editing, and server launch from IDE. On-the-fly code analysis, error highlighting and fixes. Earlier this year, JetMinds announced the release of PyCharm 2021.2, right after the release of Python 3.10. Following that, JetMinds announced PyCharm 2021.2’s first minor release in August-end with the following improvements and bug fixes: Feature Trainer The new IDE Features Trainer plugin allows developers to learn more about the Git integration in PyCharm. Additionally, it teaches developers to: Commit changesClone repositoriesPreview and navigate through the project historyPerform interactive rebase Manage branches and changelists Annotate with Git Blame Developers can access the new VCS’ lessons by selecting the Help| Learn IDE Features option from the main menu. Faster Environment Creation The latest version uses virtualenv instead of venv for Python 3 environment creation. This makes the creation of a virtual environment super fast. Improvements for Vue The latest upgrade of PyCharm includes support improvements for Vue. It now supports the final version of script setup RFC. The improvements include: The debugger now works correctly in Flask projects.The Docker Compose V2 feature, when enabled, helps create a Python interpreter.One can switch between reload on save and reload on changes mode for built-in server and HTML preview. False positive– ‘Statement expected. Found Py:DEDENT’, does not appear anymore. Now debugging Python 3.9 for Windows 10×64 is available.Shortcut issues have been fixed. A new debug tool window (the Update Running Application button) is now available. When one clicks the Run button, the IDE runs the configuration selected in the combo box. Debugging Python 3.9 for Windows 10×64 is now available. The type checker for the new union types’ syntax now works correctly with Django Rest Framework permission classes. More recently, Roberto Pesce, Product Marketing Manager at JetBrains, took to the JetBrains blog to announce another minor release of PyCharm. The PyCharm 2021.2.2 comes with the following bugs being fixed: cProfile call graph not loading. Requirements.txt blank file being created. Python 3.10 quick-fix to add ‘from_future_import’. No completion for values in style sheet properties in React Native app. Not detecting any SDK in the WSL project. Mixing up pattern matching with type hinting. False positives for HTML – attributes inlined in code. Here is a list of the issues addressed in the latest release of PyCharm 2021.2.2. And if you are starting from scratch, here is a curated list of 10 free resources to master PyCharm.","excerpt":"JetBrains recently launched the second minor release of Python IDE PyCharm 2021.2.2.","categories":["AI Trends"],"tags":["django","django python","Pycharm","Pycharm IDE"],"author_name":"Debolina Biswas","publish_date":"2021-10-13T18:00:00","publication_year":"2021","word_count":529,"keywords":["data science","Go","machine learning","Pycharm IDE","AI","ML","docker","django","Python","Pycharm","django python","JavaScript","R","Java"],"extracted_tech_keywords":["AI","machine learning","ML","data science","docker","Python","R","JavaScript","Go","Java"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/all-that-you-need-to-know-about-pycharm-2021-2\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10139449,"title":"NVIDIA, Meta Push Indian IT to Aim High on GenAI, Finally!","content":"Indian IT is finally stepping up its generative AI game. At Meta’s Build with AI Summit in Bengaluru held on October 23, Infosys announced a partnership with Meta to utilise the Llama stack, a collection of open-source large language models and tools, to build AI solutions across industries. As an early adopter of Llama 3.1 and 3.2 models, Infosys is integrating these models with Infosys Topaz, the in-house AI platform, to create tools that deliver business value. One example is a document assistant powered by Llama that improves the efficiency of contract reviews. “We are building industry-wide solutions. For market research, we have built one proof-of-concept (POC), and internally we are leveraging many Llama models for our AI-first journey,” an Infosys representative told AIM. The person further added that they have several other use cases, such as production use cases and document summarisation. This comes after Infosys revealed that it is working on small language models for its clients’ various applications. During the recent earnings call, Infosys CEO and MD Salil Parekh said, “It’s an incredible approach that leverages various open-source components along with a narrow set of industry data and Infosys’ proprietary dataset.” Interestingly, in March, Yann LeCun, the chief of Meta AI disclosed that he met with an Infosys co-founder who was funding a project based on Llama 2, the open-source model produced by Meta so that it recognises all 22 official Indian languages. Moreover, Infosys has strengthened its partnership with Meta and established a Meta Center of Excellence (COE) to accelerate the integration of enterprise AI and promote contributions to open-source communities. The COE will cultivate expertise in the Llama stack, develop industry-specific applications, and facilitate customer adoption of generative AI. Besides, Infosys has also partnered with NVIDIA. The company is incorporating NVIDIA AI Enterprise into its Infosys Topaz suite to enable businesses to rapidly implement and integrate generative AI into their workflows. Infosys has also launched an NVIDIA Center of Excellence dedicated to employee reskilling, solution development, and the widespread adoption of NVIDIA technology across organisations. How Competitors are Leveraging AI Besides Infosys, NVIDIA is also supporting other Indian IT companies. Tata Consultancy Services is creating AI solutions using NVIDIA’s NIM Agents Blueprints for sectors including telecommunications, retail, manufacturing, automotive, and financial services. Its offerings feature NeMo-powered, domain-specific language models that can handle customer inquiries and respond to company-specific questions across all enterprise functions, such as IT, HR, and field operations. In Q2 FY25, TCS reported over 600 GenAI engagements, a significant increase from nearly 270 last quarter. “Last quarter, we had eight engagements that went into production. This quarter, we have almost 86 engagements going into production,” shared TCS chief K. Krithivasan, noting that it’s a sign of maturity. Wipro, on the other hand, with its AI-powered consulting and extensive employee reskilling efforts, is looking to build an “AI-powered Wipro” that drives efficiency and transformation. “Net-net, I think GenAI will be positive for us and for the industry,” said Srini Pallia, CEO and MD at Wipro, adding that they are investing big into GenAI. “We have now trained and certified over 44,000 employees on advanced AI, and we also have a significant number of employees actively using AI developer tools across the company for all our clients,” said Pallia. Wipro is applying NVIDIA AI Enterprise software, which includes NIM Agent Blueprints and NeMo, to assist businesses in crafting custom conversational AI solutions, such as digital humans, for customer service engagement. Meanwhile, Tech Mahindra’s recent Indus 2 launch, a Hindi-centric AI model, is powered by Nemotron-4-Hindi 4B, targeting local language engagement. The company has reskilled 45,000 employees, supporting their AI roadmap through an internal proficiency framework. As Indian IT ramps up its GenAI capabilities, companies are eyeing new possibilities to integrate these advanced tools across verticals. With strong partnerships and investments, the industry is set for a rapid shift in enterprise AI adoption across the country.","excerpt":"As an early adopter of Llama 3.1 and 3.2 models, Infosys is integrating these models with Infosys Topaz, the in-house AI platform, to create tools that deliver business value.","categories":["IT Services"],"tags":["Editors Picks","Indian IT","Meta","NVIDIA"],"author_name":"Siddharth Jindal","publish_date":"2024-10-25T19:35:51","publication_year":"2024","word_count":650,"keywords":["Go","Meta","GenAI","Meta AI","AI","small language models","Git","RAG","Editors Picks","Aim","generative AI","NVIDIA","Indian IT","R"],"extracted_tech_keywords":["AI","generative AI","GenAI","Meta AI","Aim","small language models","RAG","R","Go","Git"],"url":"https:\/\/analyticsindiamag.com\/it-services\/nvidia-meta-push-indian-it-to-aim-high-on-genai-finally\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10168679,"title":"Researchers Explore Replacing Surveys Using Social Simulation With AI Agents","content":"Researchers from Fudan University have developed a framework, dubbed SocioVerse, which is an LLM-agent-driven world model for social simulation. The framework includes four components and a user pool of 10 million real individuals. The Social Environment component feeds updated external information into the simulation. The User Engine and Scenario Engine components provide realistic user context and align the simulation with the real world, respectively. Moreover, the Behavior Engine makes the agents reproduce human behaviours. This framework aims to replace traditional methods, such as surveys, interviews, and observations, which present several challenges, including high costs, limited sample sizes, and ethical concerns. With autonomous AI agents simulating human behaviour, researchers aim to observe the patterns of impact from micro-level decisions and forecast potential social dynamics. To test the framework, researchers conducted large-scale simulation experiments across domains, including politics, news, and economics. Models like Llama-3-70b-Instruct, Qwen2.5-72b-Instruct, DeepSeek-V3, GPT-4o mini, GPT-4o, and DeepSeek-R1-671b were used for the tests. First, the LLM agents focused on predicting state-level results in the US presidential election. It was found that the GPT-4o-mini and Qwen2.5-72b exhibit competitive performance according to the evaluation metric, with over 90% of state voting results being predicted correctly. DeepSeek-R1-671b was observed to be overthinking, resulting in less accurate results. Second, a simulation was conducted to assess the public’s reaction to breaking news, using the example of the ChatGPT release. GPT-4o and Qwen2.5-72b were observed to be more aligned with real-world perspectives compared to other models. Lastly, in a simulated test to understand behaviours from a national economic survey of China, Llama3-70b proved to be superior over other models in the survey, where it was able to accurately reproduce the spending habits of individuals. “Our findings indicate that state-of-the-art LLMs demonstrate a notable ability to simulate human responses in complex social contexts, although some gaps still remain between the simulated response and observed real-world outcomes,” the research paper stated. Moreover, the researchers aim to explore a broader range of scenarios to expand the simulation capabilities of LLMs.","excerpt":"What if LLMs can simulate public opinions, reactions, and more without requiring a survey?","categories":["AI News"],"tags":["AI Research"],"author_name":"Ankush Das","publish_date":"2025-04-28T12:59:14","publication_year":"2025","word_count":332,"keywords":["ChatGPT","AI","GPT-4o","autonomous AI","AI Research","GPT","Aim","AI agents","llm_models:Llama","R","llm_models:GPT"],"extracted_tech_keywords":["AI","GPT-4o","ChatGPT","Aim","R","GPT","autonomous AI","AI agents","llm_models:GPT","llm_models:Llama"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/researchers-explore-replacing-surveys-using-social-simulation-with-ai-agents\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":58602,"title":"It Takes Years To Become A Data Scientist, Says This Chief Data Scientist","content":"Owing to the increasing demand for data scientists and the growing course offerings, today, many aspiring data scientists want to learn as many tools and techniques as possible within a few months. Therefore, they enrol in several online courses, complete it, and then seek job opportunities. This approach, however, is ineffectual as they struggle to crack the interview since job seekers do not get an in-depth understanding of the domain. Instead, they should prepare a long term plan for starting and thriving in the landscape. In an attempt to assist aspirants in making effective decisions for their career, Analytics India Magazine, every week, brings out the story of prominent data scientists. For this edition, we interview Srinivas Atreya, Chief Data Scientist at RoundSqr, for our weekly column My Journey In Data Science. The Onset Srinivas has completed his bachelors of science in mathematics, statistics and computer science from Osmania University in 1995, and then, went on to do an MBA from NIT Trichy. After completing his MBA in 1997, he joined Tata Consultancy Services (TCS) as a C\/C++ programmer. “Fortunately, in my first job, I was trained on MATLAB and SPSS for operational research data-related projects, which lead to the journey in machine learning and computer vision over the years,” explains Srinivas. Since he learnt statistics in his bachelors and was a proficient C\/C++ programmer, it was straightforward for him to learn statistical modelling. Back then, most of the libraries were in C programming, allowing Srinivas to carry out algorithmic development effectively. Srinivas also pointed out that people today interchangeably use modelling and algorithmic. However, he explained that modelling is a mixture of business and technology, where one tries to find a correlation between various variables, but algorithmic is purely technical and has more to do with the optimisation of the outcome from modelling. Such techniques were used by Srinivas in his very first project at TCS when he was involved in price optimisation for an American multinational consumer goods corporation. He credits the architect of the client who mentored him throughout the process of the project. Although Srinivas had to put in additional efforts to learn new ways, he was fascinated to work with data; this is where Srinivas’ journey in data science begun. The Learning Phase Throughout his professional career, Srinivas continued to learn as the data science landscape evolved. It was only in 2007 when he was into traditional image processing while working at Wipro. He was assisting a client — a US-based retail company — in identifying the objects on conveyors for effective warehousing. Working in computer vision was fascinating to him, and since then he has continuously worked towards enhancing his skills in the domain. “I never took specialisations courses; I was focused on learning mathematics, statistics and algorithms. The rest always took care of itself,” says Srinivas. “If the core concept is clear, it is not very difficult to keep up with the changes.” While Srinivas used to read various books to learn due to the shortage of online resources during his initial journey in data science, today, he reads research papers to stay abreast of the latest trends. He gets inspired by some of the researchers like Jurgen Schmidhuber for pioneering work on sequence models and Judea Pearl for ideas on the casualty. Srinivas insists both aspirants and professionals should focus on assimilating research papers and other science articles to learn new techniques and simulate thoughts. “Typically, I bother about the research papers that have more than ten citations, and then I spend a few hours to read it,” he added. “One should at least set aside three to four hours a week. Initially, it can take a month to read and understand a research paper, but it is worthwhile.” He also laid down the strategy to get started with reading such papers and said one should read its synopsis on Medium. Following this, one should go to the original paper and read every line; skimming it won’t work. If one fails to comprehend specific jargons, techniques, or mathematics behind, he\/she should lean them before moving forward to make the most out of research papers. Work Experience Being a part of a growing company like RoundSqr, Srinivas spends 70% of his time in hiring and mentoring data scientists. Knowledge in algorithms, basic probability, linear algebra, calculus and a curious mind is all that he seeks in an applicant. Besides this, he engages with various business executives who want to deploy solutions in the AI space, and sometimes it becomes a challenge for him to explain the limitations of what AI can and cannot currently do. Further, to contribute to the community, Srinivas also teaches at various institutions. Advice To Professionals And Aspirants With 20+ years of experience in this field, Srinivas believes that tools in data science will keep changing every few years, but the core fundamentals of statistics and mathematics remain the same. He urges aspirants to strengthen the basics from prominent statisticians teaching in MIT and Stanford, which are available for free over the internet. Concerned about the current approach of aspirants of quickly learning techniques from online courses and ignoring the fundamentals, Srinivas said that data science could not be learned in a few months that various online course providers claim. Instead, job seekers must stay patient and learn over the years. For which, he advises aspirants to get into data analyst roles instead of finding data science jobs after getting the certifications. “As the data science domain is vast, patience is key to success; one cannot be an expert in all the techniques of data science. Therefore, aspirants should start their journey in data science by learning fundamentals, being curious, and experimenting with new techniques. This will allow them to find a technology that they are interested in within the data science landscape,” concludes Srinivas.","excerpt":"Owing to the increasing demand for data scientists and the growing course offerings, today, many aspiring data scientists want to learn as many tools and techniques as possible within a few months. Therefore, they enrol in several online courses, complete it, and then seek job opportunities. This approach, however, is ineffectual as they struggle to […]","categories":["AI Features"],"tags":["data analyst certification","how to become a data scientist","Interviews and Discussions","My Journey In Data Science"],"author_name":"Rohit Yadav","publish_date":"2020-03-13T18:00:00","publication_year":"2020","word_count":976,"keywords":["data science","Go","how to become a data scientist","machine learning","programming_languages:R","AI","My Journey In Data Science","computer vision","Aim","C++","analytics","R","data analyst certification","Interviews and Discussions"],"extracted_tech_keywords":["AI","machine learning","computer vision","data science","analytics","Aim","R","Go","C++","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/it-takes-years-to-become-a-data-scientist-says-this-chief-data-scientist\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10076994,"title":"After 30 Years, a Second Language Comes to Linux Kernel","content":"“So, as is hopefully clear to everybody, the major version number change is more about me running out of fingers and toes than it is about any big fundamental changes,” wrote Linus Torvalds as he introduced Linux 6.0 earlier this month. He also added, “I’ll open the merge window for 6.1, which—unlike 6.0—has a number of fairly core new things lined up.” All speculations, which heightened after Torvalds teased in the Linux 6.0 release note, were confirmed after he accepted the pull request made by Kees Cook, a Kernel security engineer at Google. This means that v6.1 will have direct support for the Rust code. Linux will not be entirely rewritten in Rust, but the process is already ongoing for adding 12,500 lines of code to the Linux kernel. https:\/\/twitter.com\/alex_gaynor\/status\/1577097928480460801?s=20&t=rFI2Hs-LMR5QNpakSl_UsA This is a major update given Linux kernel from the start (since 1991) has been written in C programming language. It is the same language that was used to write kernels for Unix and similar operating systems since the 1970s. Most anticipated change At the Linux Security Summit 2019, Alex Gaynor and Geoffrey Thomas (both distinguished software developers with substantial work with Rust) presented a talk where they made a case for Rust in the Kernel and even presented their work prototyping Rust kernel modules. They based their presentation on security concerns and how Rust can avoid such error classes through safer APIs. Similar sentiments were echoed by many developers. According to a few reports, the thought of Rust in the Linux kernel first emerged as early as 2013. Later, at the 2020 Linux Plumbers Conference that was held virtually, one of the most heavily attended sessions was the one where attendees discussed and debated using Rust upstream in the Linux kernel and also the challenges of doing so. Torvalds seemed more open to the idea and even requested that Rust compiler availability be enabled by default in the kernel build environment. This would enable any potential submissions to be built and automatically tested like any other kernel code. In the same year, Torvalds was speaking at the Open Source Summit. During the discussion, he spoke about a lot of topics, including the one concerning the future of Linux overall, considering the current crop of maintainers is in their forties, even fifties. He was also asked if C, the language in which the kernel is written, faces risk from Go, Rust and others. To this, Torvalds said that while C is still one of the top ten languages, the kernel team was looking into using Rust and the likes to write things that are “not very central to the kernel itself”. He further added, “It is going to happen that we will have different models for writing these kinds of things, and C won’t be the only one.” From the beginning of this year, Torvalds started making bolder statements. In fact, at this year’s Open Source Summit in Texas, he said that Rust could make it to the Linux kernel by the next major release. Why Rust is more preferred over others There is a good reason why C is one of the most popular languages, even decades later. It is easy to read and write, and is portable between hardware architectures. That said, C has some glaring challenges, specifically due to its nature as a non memory-managed language. A variable, once its purpose is realised, must be explicitly destroyed. Failing to do so may give rise to memory leaks or even cause system crashes due to overaccumulation. On the other hand, Rust has emerged as a ‘developer darling’ of late. It has been long favoured as the desirable contender as a second language for kernel, especially due to its automated way of ensuring secure memory management. Rust has good support for interoperability with code written in C, including support for both calling functions using the C ABI and for defining functions with C-compatible ABIs. Further, the bindgen tool can parse C header files to give appropriate Rust declarations so that there is no need to duplicate definitions from C. This in turn provides a scope for cross-language type checking. Last year, it was reported that the Google Android team was supporting a project led by developer Miguel Ojeda to write a few elements of the Linux kernel. During the development phase, Ojeda said that the team built a Rust Binder kernel module that demonstrated better preliminary performance than the C version. In a blog that details the Rust for Linux project, Ojeda wrote that one of the key properties of Rust that makes it very interesting to consider as the kernel second language is that it makes sure no ‘undefined behaviour’ occurs, especially in terms of memory management. “This includes no use-after-free issues, no double frees, no data races, etc,” he added.","excerpt":"Rust has emerged as a ‘developer darling’ of late.","categories":["AI Features"],"tags":[],"author_name":"Shraddha Goled","publish_date":"2022-10-11T14:00:00","publication_year":"2022","word_count":803,"keywords":["Go","API","programming_languages:R","AI","programming_languages:Go","Rust","R","programming_languages:Rust"],"extracted_tech_keywords":["AI","R","Go","Rust","API","programming_languages:R","programming_languages:Go","programming_languages:Rust"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/after-30-years-a-second-language-comes-to-linux-kernel\/","complexity_score":4,"technical_depth":8,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10109809,"title":"8 On-Point AIM Forecasts of 2023","content":"The year 2023 unfolded like a tech thriller, with generative AI captivating everyone from corporations to our next-door neighbours. In an annual tradition, Analytics India Magazine (AIM) laid down the markers for the tech trajectory in India and across the globe. The Top Data Science & AI Trends for 2023 report is a detailed guide for industry insiders and newbies to understand the markets and how to move forward. Here are eight on-point predictions made by the team in 2022, which came true in 2023. Recession significantly impacted the hiring of data professionals As per AIM Research’s job tracker, which aggregates information about data science and analytics job listings from various online platforms, India witnessed a decrease in the overall number of data science and analytics job opportunities in the past year. While roles in the energy and utilities sector decreased compared to 2022, those in retail & consumer packaged goods (CPG) and pharma & healthcare saw an increase. Legality of data privacy gained prominence Safety and the risk issue around AI became a significant part of the public agenda in 2023 all around the world. In India, new data privacy and storage laws were enacted to safeguard users. Tech giant Google stepped up its efforts to address this issue with valuable security with the Dark Web Report, which will allow Indians to closely monitor their personal information on the dark web, a hub for illicit activities such as the trading of sensitive personal data. Big IT started automating internal processes with Gen AI IT leaders across the globe have long known that AI-powered automation is essential for survival in a digitally driven economy; the initial steps were visible in 2023. While some companies have hesitated to adopt the new human-mimicking technology, most have tried to embrace it as a workflow tool. One well-known example is of Salesforce deploying Einstein Copilot to auto-generate email replies and account updates based on a customer’s specific context. CDOs became more critical for collaboration According to the Deloitte Chief Data Officer survey of 2023, the role of CDOs became increasingly important in most organisations, with 61% of CDOs stating that creating, updating, or implementing their data strategy became their top priorities. As per the latest report, generative AI has created a halo effect for CDOs, opening up their peers and boards to talk about topics they might have previously shied away from. Most CDOs recognise that foundational data quality and good use cases are fundamental challenges that must be overcome to take advantage of the technology. Moving ahead, 93% of CDOs believe a practical data strategy is needed if organisations are to succeed with generative AI. Cloud computing and multi-cloud became a big part of the IT strategy The cloud revolution gained more momentum in 2023. It emerged as an equaliser for all sizes of businesses, from global enterprises to startups, offering a field to innovate and optimise workloads. Large enterprises harnessed the cloud to streamline operations, enhance scalability, and reduce costs while using generative AI. The year also saw a rise in businesses born out of the cloud. China’s Alibaba Cloud rolled out Tongyi Qianwen in April to run applications across its ecosystem on the large language model. Data scientists’ started focusing on software There was an unspoken war waging amongst the tech companies regarding data collection and wrangling. Naturally, the ability to process data in huge chunks became critical for data scientists. The volume of data grew at an exponential rate, and being able to handle and analyse it became necessary. A study demonstrated that data scientists were in most demand in IT and tech, accounting for 49% of the job postings on LinkedIn. Citizen data scientists to grow due to no-code, low-code platforms Programming became a cakewalk with AI-powered coding assistants and also with low-code\/no-code, allowing citizens to build applications for business usage. The global low-code and no-code market was valued at nearly $15 billion according to the ISG Information Services Group. Gartner predicts that 70% of new applications will be developed using no-code or low-code frameworks by 2025. These platforms help reduce app development time by 90% and empower non-developers to delve into app development regardless of their technical background. Generative AI will become more powerful and accessible AI officially became the year’s word, thanks to generative AI. The popularity and mainstream usage of generative AI services have expanded exponentially in 2023, as we saw every second company in tech and across all industries rushing to embrace the tech. Apart from internally implementing the tech, the companies put efforts into making the tech available to the general public through waitlists, APIs and open-source projects. McKinsey reported that while overall AI adoption remains steady at around 55%, more than two-thirds of respondents say their companies plan on using generative AI. Furthermore, four in five online teenagers aged 13-17 now use generative AI tools and services.","excerpt":"AI officially became the word of the year.","categories":["AI Trends"],"tags":[],"author_name":"Tasmia Ansari","publish_date":"2024-01-02T14:00:00","publication_year":"2024","word_count":814,"keywords":["data science","AWS","AI","cloud computing","ML","RAG","Aim","generative AI","analytics","R"],"extracted_tech_keywords":["AI","ML","data science","analytics","generative AI","Aim","RAG","cloud computing","AWS","R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/8-on-point-aims-a-forecasts-of-2023\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":36212,"title":"Women In AI Leadership Awards: AIM &#038; Jigsaw Academy Felicitate Emerging Tech Leaders At The Rising 2019","content":"The Rising 2019, an inaugural conference for women in analytics and artificial intelligence, brought together over 200 industry leaders from the emerging technologies field for a day-long conference at The Taj on 8 March in Bengaluru. The Rising 2019 also featured one of the first Women In AI Leadership awards, a part of the summit which recognised the achievements and contributions of women trailblazers at the forefront of innovation and change in AI. The awards organised by Analytics India Magazine in association with Jigsaw Academy celebrated the achievements and innovation of women in the field of emerging technologies and also endeavoured to find new role models in a sector where women representation is highly skewed. The awards were presented by Ashish Gupta, Head- IoT Programs, Jigsaw Academy, who also delivered an inspiring speech about encouraging dialogue around the dynamically evolving field and why adopting a continuous learning approach and keeping oneself updated with the current market trends is extremely important. “We at Jigsaw Academy are very confident that the winners of the Rising awards are amongst the country’s most talented data science professionals. We are confident that their contributions will be invaluable to the organisations they work with and are excited to see what lies in their future and what they bring to the world of analytics,” he said. The awards are a celebration of women talent in AI technology, foster a dialogue about gender parity and how important it is for women to make their voices head, flourish in tech teams to help build more inclusive products and teams. “Women In AI Leadership awards was instituted to acknowledge the research and cutting-edge contributions from tech leaders who have displayed leadership and made a considerable impact to the ecosystem in India. We congratulate all the winners for this edition,” said Bhasker Gupta, Founder and CEO at Analytics India Magazine. The conference which saw a vibrant gathering of Indian women data scientists and analytics practitioners had a common thread running through the keynote and the talks — ruling out bias in AI algorithms and the need to build Explainable AI. Meanwhile, the panel discussion – Women’s Day 2020 – Changing the Narrative moderated by Anshu Sharma, Managing Director, Global Head – Retail Banking Technology at Standard Chartered Bank centred on the need for women to identify their strengths in the industry, own their career and make their voices heard in their respective organisations. The other panelists included Barkha Sharma Founder and CEO at BASH.AI, Aparajita Karimpana Senior Director Analytics at Envestnet Yodlee, Sahana Shetty HR Leader – Technology at ANZ and Pankaj Rai Senior Vice President Strategy at Wells Fargo who brought a colourful perspective to the table on how women should juggle careers and families in a fast-paced working environment. Rising 2019 awarded 14 leading data scientists from across India. Here’s the full list of awardees.","excerpt":"The Rising 2019, an inaugural conference for women in analytics and artificial intelligence, brought together over 200 industry leaders from the emerging technologies field for a day-long conference at The Taj on 8 March in Bengaluru. The Rising 2019 also featured one of the first Women In AI Leadership awards, a part of the summit […]","categories":["Deep Tech"],"tags":["leader of ai","Women in AI Leadership awards"],"author_name":"Richa Bhatia","publish_date":"2019-03-13T09:14:41","publication_year":"2019","word_count":476,"keywords":["data science","Go","artificial intelligence","AI","Women in AI Leadership awards","innovation","RAG","explainable AI","analytics","GAN","R","leader of ai"],"extracted_tech_keywords":["AI","artificial intelligence","data science","analytics","RAG","R","Go","GAN","explainable AI","innovation"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/women-in-ai-leadership-awards-aim-jigsaw-academy-felicitate-emerging-tech-leaders-at-the-rising-2019\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10009279,"title":"Step by Step Guide To Object Detection Using Roboflow","content":"Roboflow is a Computer Vision developer framework for better data collection to preprocessing, and model training techniques. Roboflow has public datasets readily available to users and has access for users to upload their own custom data also. Roboflow accepts various annotation formats. In data pre-processing, there are steps involved such as image orientations, resizing, contrasting, and data augmentations. The entire workflow can be co-ordinated with teams within the framework. For model training, there’s a bunch of model libraries already present such as EfficientNet, MobileNet, Yolo, TensorFlow, PyTorch, etc. Thereafter model deployment and visualization options are also available hence encompassing the entire state-of-art. Roboflow is used in various computer vision industries for use cases such as – gas leak detection, plant vs weed detection, aeroplane maintenance, roof damage estimator, satellite imagery, self-driving cars, traffic counter, garbage cleaning, and many more. Steps To Use Roboflow in Object Detection: 1.Dataset Loading 2.Labeling 3.Organise 4.Process 5.Train 6.Deploy 7.Display Roboflow has an account set up for each user. Now we will discuss each of the steps in the task of object detection. 1. Dataset Loading Roboflow has both public and private datasets. Public datasets can be accessed from the website itself and private datasets can be uploaded by users. We can select any dataset we feel like from these or upload our own custom dataset from the ‘your dataset’ section. If you wish to use public datasets make sure to fork it after opening. After forking a dataset if we wish we can also add our own images. 2. Labeling Public datasets come with annotations. For custom datasets, we need to make sure to have annotations. Roboflow annotations aren’t always available to all users but users can contact them for the service. Labelling images can be done with a free open source software LabeIImg in python. Labelling is basically drawing bounding boxes to get the notations of where exactly the object is present in the image. Roboflow has the feature to correct annotation errors and provide statistics of images with and without annotations. 3. Organise Team collaboration is easily possible using roboflow team sharing. For every version of changes made to data will be reflected directly to anyone from the team using it. There’s also an option for public sharing which will be reflected in the roboflow community. 4.Process Various steps are involved in processing such as a health check of the dataset could be done to check if the dataset is imbalanced for classification models and other such insights about data. Data modifications can be done in the following ways: I’ve uploaded a custom dataset of Indian coins to predict the category of coins( Re 1, Rs 2, Rs 5, Rs 10) After the needful steps, we generate a version of the dataset and export it in the format as required all the files in a zipped form. There are various formats COCO JSON, Pascal VOC, YOLO, Tensorflow Object Detection, Multiclass Classification, etc.  The code-free versions are not available in the free tier, those generate models directly from the cloud- AWS, GCP, or Azure. To use the downloaded code in Google Colab notebook a unique code is generated which should not be shared with anyone. 5. Train Roboflow provides train-valid-test split ratio. The models’ library in roboflow will let us select different training models available to be used on Google Colab notebook directly. Make sure to first copy the drive and then make changes into the notebook. Paste the link acquired from the previous step into the 2nd cell of the Colab notebook. Run subsequent cells where code for training is already present. This might take some time and then visualize the model performance. Last iteration Output: (next mAP calculation at 8000 iterations) Last accuracy mAP@0.5 = 61.67 %, best = 80.83 % 8000: 0.004265, 0.003273 avg loss, 0.000026 rate, 0.733761 seconds, 384000 images, 0.020533 hours left calculation mAP (mean average precision)... 12 detections_count = 27, unique_truth_count = 11 class_id = 0, name = Re1, ap = 50.00%   (TP = 3, FP = 2) class_id = 1, name = Rs10, ap = 33.33%   (TP = 1, FP = 2) class_id = 2, name = Rs2, ap = 100.00%   (TP = 3, FP = 0) Class_id = 3, name = Rs5, ap = 46.67%   (TP = 1, FP = 1) for conf_thresh = 0.25, precision = 0.62, recall = 0.73, F1-score = 0.67 for conf_thresh = 0.25, TP = 8, FP = 5, FN = 3, average IoU = 49.10 % 6. Deploy Run model on cloud or system. Use models on mobile or servers or both. Weights that have provided the best results are automatically saved. 7. Display I’ve trained a custom dataset of Indian coins using roboflow to predict the category of coins( Re 1, Rs 2, Rs 5, Rs 10) after following all the above steps and training model on Yolov4 tiny. Along with classification it also tracks the objects. Our model seems to have predicted well. Note that this image is from the test set and not the train set. Conclusion Roboflow helps in every step of computer vision problem right from data collection to deployment. Roboflow enhances performance by its efficient parameters readily available to use.    You can find the complete notebook of the above implementation in AIM’s GitHub repositories. Please visit this link to find this notebook.","excerpt":"We will discuss each of the steps in the task of object detection using roboflow.","categories":["Deep Tech"],"tags":["Computer Vision","Machine Learning","multiclass classification","Object Detection techniques"],"author_name":"Jayita Bhattacharyya","publish_date":"2020-10-11T16:00:17","publication_year":"2020","word_count":888,"keywords":["AWS","AI","PyTorch","Machine Learning","computer vision","RAG","Colab","multiclass classification","Aim","object detection","Computer Vision","TensorFlow","Azure","Object Detection techniques"],"extracted_tech_keywords":["AI","computer vision","Aim","TensorFlow","PyTorch","Colab","RAG","object detection","AWS","Azure"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/step-by-step-guide-to-object-detection-using-roboflow\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10091267,"title":"OpenAI&#8217;s Desperate Attempt to Create Its Own Open-Source Ecosystem","content":"OpenAI, ChatGPT’s parent, has launched a bug bounty program to incentivise researchers to detect and report any vulnerabilities in its artificial intelligence systems. The announcement was made through a blog post on Tuesday, revealing that the rewards would range from $200 for minor issues to $20,000 for “exceptional discoveries.” “This initiative is an essential part of our commitment to developing safe and advanced AI,” wrote Matthew Knight, OpenAI’s head of security in the blog. “As we create technology and services that are secure, reliable, and trustworthy, we would like your help.” Moreover, the program does not cover instances of incorrect or malicious content produced by OpenAI’s technology. The decision for the program does not come as a surprise. It was prompted by a recent incident involving Alex Albert, a 22-year-old jailbreak prompt enthusiast who managed to bypass the safeguards on ChatGPT. OpenAI’s president and co-founder, Greg Brockman, in response, tweeted that the company had been considering a bounty program or network of “red-teamers” to identify weak spots in their AI models. Democratized red teaming is one reason we deploy these models. Anticipating that over time the stakes will go up a *lot* over time, and having models that are robust to great adversarial pressure will be critical. Also considering starting a bounty program\/network of red-teamers! https:\/\/t.co\/9QfmXQi9iM— Greg Brockman (@gdb) March 16, 2023 Sourced from the Open Looks like the closed-door company is now desperately attempting to go open source route to improve its systems’ security posture with the recent bounty program announcement. However, it does not want to give back to the community. For instance, its last open-sourced Gym and Universe seven years ago barely exists. Another open-source initiative falls victim to the money conundrum is GPTx, where the company has been using open source to train and scale the models, it closed the doors on the code. It is important to note that ChatGPT still uses open source code. Just last month when a security flaw exposed users’ search history, Sam Altman, OpenAI’s CEO acknowledged that the bug was caused by an issue in Redis client library. we had a significant issue in ChatGPT due to a bug in an open source library, for which a fix has now been released and we have just finished validating.a small percentage of users were able to see the titles of other users’ conversation history.we feel awful about this.— Sam Altman (@sama) March 22, 2023 So, like many others ChatGPT owes a great debt to the open source community. But the company seems to have turned a blind eye towards the community since ChatGPT is not open sourced, nor is it likely to ever be. Moreover, the company’s previous product which caused a stir in the art community was DALLE.2 in Unfortunately, OpenAI is yet to launch an open-source version of DALL.E 2. The model remains exclusively available to a number of users who secured a spot on the waitlist. OpenAI Misleads Apart from its money making machine ChatGPT, the company has also been an important subject of discussion for being a closed source company contrary to its name, ‘Open’AI. In 2019, the former nonprofit restructured as a “capped-profit” and received a billion dollar investment from software giant, Microsoft. Some worried that this move may result in making the innovative company no different from the other AI startups out there. A year later, Elon Musk, who was one of OpenAI’s co-founders tweeted, “I have no control & only very limited insight into OpenAI,” adding his confidence in its safety was “not high.” Since he left the board in 2018 he has been actively vocal about his views on the company. The Twitter owner recently took to Twitter to express his concern over OpenAI’s shift towards becoming a closed-source, for-profit company controlled by Microsoft. Musk, who had named OpenAI as an open-source, non-profit organization to counterbalance Google, criticized the move and urged the AI company to remain true to its original vision. “Not what I intended at all,” he said when Microsoft extended its long-term partnership with OpenAI through a new “multiyear, multibillion dollar investment.” The bottom line is that even though OpenAI wants to remain closed door and not share their secret recipes, the company seems to not mind seeking help from the public to mend their ways.","excerpt":"The research company heavily relies on open source resources but the contributions remain bare minimum.","categories":["Global Tech"],"tags":["ChatGPT","Open Source","OpenAI","Sam Altman"],"author_name":"Tasmia Ansari","publish_date":"2023-04-12T18:30:00","publication_year":"2023","word_count":716,"keywords":["Go","ChatGPT","Sam Altman","Open Source","OpenAI","AI","artificial intelligence","BERT","GPT","Rust","R","Redis"],"extracted_tech_keywords":["AI","artificial intelligence","ChatGPT","OpenAI","Redis","R","Go","Rust","BERT","GPT"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/openai-treats-open-source-as-a-one-way-street\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10055267,"title":"Google At NeurIPS 2021: Gets 177 Papers Accepted","content":"The 35th edition of the Neural Information Processing Systems conference 2021 (NeurIPS 2021) commenced on December 6, 2021. The nine day conference is packed with a series of tutorials, workshops, and presentations. Over 9,000 papers were submitted at the conference this year, of which 2,344 papers were accepted; this the highest number of papers accepted since 2013. The annual NeurIPS conference is the most awaited and well attended machine learning events of the year. Leading companies and academic institutions like Google, Microsoft, Meta, DeepMind, Stanford, and Carnegie Mellon University participate in great number. Among tech companies, Google had the highest number of accepted papers – 177. Let us look at some of the winning papers from Google at NeurIPS  2021. Deep Reinforcement Learning at the Edge of the Statistical Precipice (Outstanding paper award) By Rishabh Agarwal, Max Schwarzer, Pablo Samuel Castro, Aaron Courville, and Marc G. Bellemare. This paper outlines practical approaches to improve the rigour of deep reinforcement learning algorithm evaluation. The focus is especially the evaluation of new algorithms that should provide stratified bootstrap confidence intervals, performance profiles across tasks and runs, and interquartile means. The researchers also show that standard approaches for reporting results in deep RL across many tasks and multiple runs can make it difficult to assess if a new algorithm represents a consistent advancements over past methods. The performance summaries are designed to be able to compute with a small number of runs per task. Read the full paper here. Do Vision Transformers See Like Convolutional Neural Networks? By Maithra Raghu, Thomas Unterthiner, Simon Kornblith, Chiyuan Zhang, Alexey Dosovitskiy Convolutional neural networks (CNNs) have been the preferred model for visual data, but recently it has been shown that vision transformer models (ViT) can achieve comparable or even superior performance on image classification tasks. This paper analyses the internal representation structure of ViTs and CNNs on image classification benchmarks. The team found out the differences between the two architectures ( eg: ViT having more uniform representations across all layers). The research also looks at important roles played by self-attention, the ramifications for spatial localization, and the effect of (pretraining) dataset scale on intermediate features and transfer learning. Read the full paper here. The Difficulty of Passive Learning in Deep Reinforcement Learning By George Ostrovski, Pablo Samuel Castro, Will Dabney Learning to act from observational data without active environmental interaction in reinforcement learning is a challenge. Approached like constraints on the learned policy, preventing strong deviations from the state-action distribution of the dataset are being used to overcome this challenge. But, these methods are evaluated using non-linear function approximation. This paper proposes “tandem learning” experimental paradigm which facilitates the empirical analysis of the difficulties in offline reinforcement learning. The team has also identified approximation in conjunction with fixed data distributions as the strongest factor. This observation challenges the hypothesis that is stated in past work. The team offers insights for offline deep reinforcement learning and sheds light on phenomena observed in online case of learning control. Read the full paper here. Datasets and Benchmarks Accepted Papers NeurIPS launched the new Datasets & Benchmarks track this year. It went to the paper titled, “Reduced, Reused and Recycled: The Life of a Dataset in Machine Learning Research” by Bernard Koch, Emily Denton, Alex Hanna, Jacob G. Foster. Benchmark datasets play an important role in machine learning research – they serve as a measure of progress towards the goal. But not much attention has been given to the dynamics of benchmark dataset use and reuse, within machine learning communities. This paper studies how dataset usage patterns differed across machine learning subcommunities between 2015 and 2020. They found increasing concentration on fewer and fewer datasets within task communities, significant adoption of datasets from other tasks, as well as concentration across the field on datasets that have been introduced by researchers situated within a small number of elite institutions. These results have an impact on AI ethics, and equity and access within the field. Read the full paper here. Test of Time Award The Test of Time Award went to Online Learning for Latent Dirichlet Allocation by Matthew Hoffman (senior scientist at Google), David Blei, and Francis Bach. The research team developed a variational Bayes (VB) algorithm for Latent Dirichlet AI -location (LDA). It is based on online stochastic optimization with a natural gradient step and converges to a local optimum of the VB objective function. It comes with the capability of analysing large amounts of document collections, including those arriving in a stream.The paper studies the performance of online LDA by fitting a 100-topic topic model to 3.3 million articles from Wikipedia in a single pass.They show that online LDA finds topic models as good or better than those found with batch VB and does this in a fraction of the time. ‘Read the full paper here.","excerpt":"Let us look at some of the winning and interesting papers that came from Google this year at NeurIPS 2021","categories":["Global Tech"],"tags":["AI (Artificial Intelligence)","Data Science","Google","Machine Learning"],"author_name":"Sreejani Bhattacharyya","publish_date":"2021-12-12T13:00:00","publication_year":"2021","word_count":807,"keywords":["Go","machine learning","AI","neural network","vision transformer","Machine Learning","Transformers","RAG","CNN","ViT","Google","Data Science","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","machine learning","neural network","Transformers","RAG","R","Go","CNN","vision transformer","ViT"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/google-at-neurips-2021-gets-177-papers-accepted\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":42945,"title":"How Cloud Can Be A Game-changer For India’s Insurance Industry","content":"At first glance there is a lot of enthusiasm in the insurance industry towards cloud computing, for improved data access, cost and performance, even if there is some uncertainty around the public-private or hybrid cloud, about exactly how to embrace the transformation and this new type of connectivity. Cloud service providers on their part have started to offer more and more services and are making it easier day by day for their customers to set up, deploy and automate products. For example, Google Cloud offers TPU (Tensor Processing Unit) for faster execution of their open-source TensorFlow machine learning platform. You can now use more than 16 IBM Watson AI services on the IBM Cloud and add artificial intelligence to your applications. When we look at the speed and efficiency in delivering quotes for example, or for the flow of motor telematics data, most of us think of the cloud as an opportunity to scale technology easily and quickly, not necessarily about opportunities for new services. But there is a deeper impact, about preparing for the future, and how more insurance products are going to be delivered and priced, based on pooled data and pooled internal and external information, beyond just what an insurer knows about a customer from their own systems. Over the past few years, cloud-based hosting has become the ‘invisible hand’ that feeds all kinds of on-demand services, whether music, calendars, photos and other applications in an instant, whenever, wherever and through whatever device we want them. Now with the ability to consume data (structured and unstructured) and create insights out of it using services provided by the cloud providers we are talking about the next evolution. Building applications that use complex technologies (like artificial intelligence and machine learning) via simple cloud services rather than having to build it from the ground up. Cloud computing is very much about preparing for the future of insurance, for people living in a connected ecosystem in the future. There is no sign that this prospect of hyper-connection is going to stop. In fact, cloud computing is enabling faster and easier connectivity paired up with a host of services from database hosting to serverless execution. Security with cloud technologies Some insurers fear that with cloud technology you may not necessarily know where your data, or your software is physically, and so you don’t know whether it is secure. Without being able to touch and feel your servers in your data centre, the data use becomes governed by third-party relationships, contracts and e-certificates within a system. Since insurers handle PII data (personally identifiable information) it is of course fundamentally important for them to ensure the security of this data and prevent any unauthorised usage. Most cloud providers provide multiple layers of security. For those who do not want a public cloud (one that everyone has access to), a secure private cloud can be created. Strict access controls, firewalls and network security – similar to your physical infrastructure – can be deployed in the cloud. Several security standards have come up to ensure security within the cloud environment Companies who vie for certification of their cloud infrastructure can apply appropriate standards and be ISO27017 certified. IRDAI has also come up with guidelines on the usage of cloud and security of the same. As cloud adoption increases – and the data security guidelines become more stringent – cloud service providers are adopting ever-more secure mechanisms to protect all of their customers’ data. As shown in the RightScale 2019 State of the Cloud Report from Flexera, both private and public cloud adoption have risen in the last few years. The study further shows that numerous corporates and numerous lines of business are now using public cloud. The number of public cloud users has risen to 92% from 89% in 2016-17. In the private cloud, the figures are from 72% to 75%. As a result, the overall proportion of respondents using a minimum of one private cloud or public cloud is 96%. Surveys suggest the trend will continue through 2020. Insurers will find that setting up the technical infrastructure will become very easy for new competition. Similar to the boom of fintech companies, new companies specialising in many aspects of the insurtech space are coming up rapidly. Some of these companies will help insurers better their processes while others will compete for the same market with innovative products. As these innovative products are adopted, cloud technologies will make it easy for scaling up of these solutions and no upfront capital expense is required – costs grow with the revenues. Existing insurers will have to adapt to be able to move quickly and respond to this new dynamic. The fast movers will benefit while the laggards will find it increasingly difficult to compete in the new segments. Cloud technology evolution is not just about taking your physical infrastructure and moving it to the cloud. Going from virtual servers to containers and containers to serverless – cloud is all about focusing on the business of insurance, the core competency. Building technology that supports easy and quick underwriting and fast processing of claims in the most efficient manner rather than the complexities of standing up an infrastructure, figuring out its reliability and scalability and worrying about disaster recovery. Let’s outsource those complexities and focus on the core technology that enables our business partners to conduct efficient insurance business.","excerpt":"At first glance there is a lot of enthusiasm in the insurance industry towards cloud computing, for improved data access, cost and performance, even if there is some uncertainty around the public-private or hybrid cloud, about exactly how to embrace the transformation and this new type of connectivity. Cloud service providers on their part have […]","categories":["AI Features"],"tags":["cloud security","FinTech","Gen AI in Insurance","insurtech","private cloud","virtual invisible network"],"author_name":"Kaustubh Deshpande","publish_date":"2019-07-19T11:10:09","publication_year":"2019","word_count":898,"keywords":["Go","artificial intelligence","machine learning","TPU","AI","cloud computing","R","virtual invisible network","cloud security","insurtech","serverless","Aim","Gen AI in Insurance","FinTech","TensorFlow","private cloud"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","Aim","TensorFlow","cloud computing","serverless","TPU","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-cloud-can-be-a-game-changer-for-indias-insurance-industry\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":64866,"title":"Top 10 Speakers At Plugin 2020","content":"plugin is an innovative next-gen online conference on artificial intelligence and data science, which has been initiated to bring in the experts of the industry to talk about cutting-edge innovations in a distinctive virtual setting. Hosted by Analytics India Magazine, plugin is a two-day event providing an opportunity for attendees to network with their peers, meet leaders visually, and learn from subject matter experts of the industry. plugin will include three tracks across two days and will have 50 plus speakers who will help attendees to explore key concepts, understand use cases, and learn best practices in data science. Along with that, the event will host a wide range of tech talks, workshops and sessions conducted by some of the best minds in the industry. More details here. In this article we will share the top ten speakers that you must attend to at plugin 2020: Bill Inmon, Father of the data warehouse William H. (Bill) Inmon, is an American computer scientist, recognised by many as the father of the data warehouse. He wrote the first book, held the first conference with Arnie Barnett, and was a pioneer to offer classes in data warehousing. Inmon also created the accepted definition of what a data warehouse is — a subject-oriented, non-volatile, integrated, time-variant collection of data in support of management’s decisions. At plugin, his talk is scheduled for 28th May, where he will speak about data architectures, the broad spectrum of all of the data in the corporation and its business value. Viral Shah, Inventor of Julia programming language Viral B Shah is an Indian computer scientist, best known as a co-creator of the Julia programming language. Besides, he is the CEO of Julia Computing, which he co-founded with Alan Edelman, Jeff Bezanson, Stefan Karpinski, Keno Fischer and Deepak Vinchhi. With a background in bachelors in computer engineering from the Padmabhushan Vasantdada Patil Pratishthan’s College of Engineering, Shah pursued Computer Science from the University of California, Santa Barbara, where he received his PhD in the field of Combinatorial Scientific Computing. Kirk Borne, Top AI Influencer Dr Kirk Borne is an astrophysicist, data scientist, and is also the principal data scientist of the Strategic Innovation Group at Booz-Allen Hamilton, and was a professor of astrophysics and computational science in the George Mason University School of Physics, Astronomy, and Computational Sciences during 2003-2015. Borne has experience in huge scientific databases and information systems, including proficiency in scientific data mining. He was also a contributor to the design and development of the new Large Synoptic Survey Telescope, for which he contributed to the study of science data management, informatics and statistical science research, and public outreach. At plugin, Borne’s talk is scheduled for 29th May, where he will speak about the importance of data literacy for the future of work. Abhishek Thakur, World’s first 4x Kaggle Grandmaster Abhishek Thakur is the world’s first 4x Kaggle Grandmaster and currently works as a chief data scientist for boost.ai. He holds a master’s degree in computer science from the University of Bonn. In his free time, he likes to create YouTube videos related to applied machine learning, take part in machine learning competitions, and is about to launch a book titled: Approaching (Almost) Any Machine Learning Problem. Johnson Poh, Head of Group Enterprise AI at UOB Bank Johnson Poh is currently the executive director and head of the group enterprise artificial intelligence and data science at UOB Group. He has been practising data science for the past decade with experience in finance, government sectors, and consulting. His prior professional designations include head data science\/practice lead at DBS Bank, chief data scientist, ASEAN Region at Booz Allen Hamilton and head data science\/principal data scientist at Ministry of Defence, Singapore. Poh completed his bachelor’s degrees in pure mathematics, statistics and economics from the University of California, Berkeley, and has received his postgraduate degree in statistics from Yale University. Gianluca Gindro, Head of Data Science at Kuoni Gianluca is the head of data science at Kuoni, a global travel brand, and also works as a freelance instructor of ‘data science of business’ courses. Previously he had set-up the data operations at Groupon, where he was heading the analytics for six years. At plugin, his talk is scheduled on 29th May 2020, where he will speak about the black-box approach to data science and what it actually means for businesses. Dipanjan Sarkar, Data Science Lead at Applied Materials & Google Developer Expert – ML Dipanjan Sarkar is a data science lead at Applied Materials, where he manages advanced analytics efforts around natural language processing, computer vision, and deep learning. Sarkar is also a Google Developer Expert in Machine Learning. He holds a master of technology degree with specialisations in data science and software engineering. Sarkar has been an analytics practitioner for several years now, specialising in machine learning, natural language processing, computer vision and deep learning. He also acts as a mentor and an AI consultant at various organisations like Springboard. He has published several books on Python, R, machine learning, natural language processing, social media analytics, and deep learning. In his spare time, he loves reading and watching football and popular sitcoms. At plugin, his talk is scheduled for 28th May 2020, where he will speak about how NLP can be used to solve real-world problems. Anirban Nandi, Head of Analytics at Rakuten With close to 13 years of professional experience, Anirban Nandi specialises in data sciences, business analytics and data engineering spanning across various verticals of online and offline retail, and building analytics teams from the ground up. Following his masters from JNU in economics, Nandi started his career in the analytics team at Target and spent more than eight years working on developing in-house analytics products like customer personalisation, recommendation system, and search engine classifiers. Post Target, Nandi became one of the founding members at Data Labs and spent close to five years, building and growing the onshore and offshore team of approximately 100 members working across industries. Currently, Anirban is associated with Rakuten as the head of analytics. He is developing analytics solutions for the Rakuten global ecosystem, and building a high-performing team in the Bangalore office. Dat Tran, Head of AI at Axel Springer AI Dat Tran is the head of AI at Axel Springer Ideas Engineering, the innovation unit of Axel Springer SE, which is the largest digital publishing house in Europe. He leads Axel Springer AI, where his mission is to democratise AI within Axel Springer, and hence, he drives innovations within the group. Tran is continuously striving to turn Axel Springer into an AI-first firm. Previously, Tran co-headed the data team at idealo.de, where he built up the machine learning team from scratch. Besides, he is a regular speaker and has presented at several renowned AI conferences. Tran’s background is in operations research and econometrics. He has received his MSc in Economics from the Humboldt University of Berlin. Tran’s interests are diverse from traditional machine learning, deep learning, AI in general to computer vision. At plugin, his talk is scheduled for 28th May 2020, where he would speak about how to implement a production-ready deep learning model in twelve weeks. John K. Thompson, Global Head of Advanced Analytics & AI at CSL Behring John K. Thompson is an international technology executive with over 30 years of experience in the advanced analytics and business intelligence domain. Thompson is responsible for the AI team, global advanced analytics and efforts at CSL. Prior to CSL, Thompson was an executive partner at Gartner, where he was a management consultant to market-leading companies in the areas of advanced analytics, digital transformation, and data monetisation. Before Gartner, Thompson was a leader for the advanced analytics business unit of the Dell Software Group. Thompson is also a co-author of the bestselling book – Analytics: How to win with intelligence. The book debuted on Amazon as the #1 new book in Analytics in 2017. Thompson is a technology leader with experience spanning across numerous operational areas with a focus on product innovation, strategy, growth and efficient execution. At plugin, his talk is scheduled on 28th May, where he will speak about how to build your analytics team.","excerpt":"plugin is an innovative next-gen online conference on artificial intelligence and data science, which has been initiated to bring in the experts of the industry to talk about cutting-edge innovations in a distinctive virtual setting. Hosted by Analytics India Magazine, plugin is a two-day event providing an opportunity for attendees to network with their peers, […]","categories":["Deep Tech"],"tags":["julia scientific programming","plugin","plugin aim","plugin online conference","plugin virtual data science event"],"author_name":"Sejuti Das","publish_date":"2020-05-08T18:00:00","publication_year":"2020","word_count":1366,"keywords":["plugin online conference","plugin virtual data science event","data science","plugin aim","artificial intelligence","machine learning","AI","ML","computer vision","NLP","Python","deep learning","analytics","plugin","julia scientific programming"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","NLP","computer vision","data science","analytics","Python"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/top-10-speakers-at-plugin-2020\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10171344,"title":"IIIT-Hyderabad Launches Patram, India’s First Vision-Language Foundational Model for Docs","content":"A team from the International Institute of Information Technology, Hyderabad (IIIT-H) has launched Patram-7B-Instruct, India’s first vision-language foundational model developed for document understanding. Patram is part of the BharatGen suite of multimodal large language models being created with funding from DST. Jitendra Singh, minister of state for science and technology, unveiled the model on June 2 at the BharatGen National Summit in New Delhi. Patram is a 7-billion parameter model trained to process and understand scanned and photographed documents. It responds to natural-language instructions and is now available open-source on Hugging Face and IndiaAI’s AIKosh platform. Despite its smaller size, Patram has demonstrated competitive performance against larger international models such as DeepSeek-VL-2 on benchmarks like DocVQA and VisualMRC. It also performed well on Patram-Bench, a custom evaluation set reflecting Indian document scenarios. “Patram marks a significant step as India designs state-of-the-art foundational models,” said Prof. P. J. Narayanan, Director, IIIT Hyderabad. “With this launch, we integrate language available in all forms: as text, as speech, and as images.” A team of alumni and student interns at IIIT-Hyderabad built the model in five months, with support from IIIT-H and TiH-IoT, IIT Bombay. “With Patram, we’ve built a model that understands the unique structure and diversity of Indian documents,” said Dr. Ravi Kiran Sarvadevabhatla, associate professor and lead researcher at IIIT-Hyderabad. “This is just the beginning of what India can achieve in vision-language AI.” Alongside Patram, the team also launched DocBodh, a generative AI suite for Indic document intelligence, targeting applications in governance, education, law, and business. The project is part of India’s broader effort to build open and indigenous AI infrastructure under national initiatives such as Digital India and Atmanirbhar Bharat.","excerpt":"Alongside Patram, the team also launched DocBodh, a generative AI suite for Indic document intelligence, targeting applications in governance, education, law, and business.","categories":["AI News"],"tags":["LLM"],"author_name":"Siddharth Jindal","publish_date":"2025-06-04T19:32:41","publication_year":"2025","word_count":280,"keywords":["Go","Hugging Face","funding","AI","Modal","LLM","Git","Ray","generative AI","ai_frameworks:Hugging Face","R"],"extracted_tech_keywords":["AI","generative AI","Ray","Hugging Face","R","Go","Git","funding","Modal","ai_frameworks:Hugging Face"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/iiit-hyderabad-launches-patram-indias-first-vision-language-foundational-model-for-docs\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":44698,"title":"Micron To Hire Over 1,500 Techies In Its R&#038;D Facility In India","content":"Noted US semiconductor company Micron Technology has grand plans to increase its hiring in India four times, specifically in the research and development sector over the next two years. The $30 billion company recently set up its R&D operations in Hyderabad which is supporting its global manufacturing operations. As the company is working on building niche capabilities and support functions at its Hyderabad operations, Micron is looking to hire people from Tier I institutions in India. Manish Bhatia, Executive Vice-President, Global Operations at Micron, told a noted national daily, “Today, we have about 500 people in Hyderabad and expect to grow this strength to 2,000 over the next couple of years.” Micron’s multi-billion dollar expansion in Singapore and its massive hiring plans in India come on the back of growing data-driven business opportunities beyond personal computers. Sanjay Mehrotra, President and Chief Executive Officer at Micron, told the newspaper, “2-3 years ago, cloud and data centres were a small part of our industry. Today, they have become one of the fastest-growing segments within the industry segment. Automotive is another large area and one of the fastest-growing segments for the consumption of DRAM and NAND flash memories. Autonomous driving, machine learning, and artificial intelligence are all dependent on vast quantities of information. And as 5G technologies dramatically speed wireless communication, they will enable a truly intelligent network of billions of devices and every new device will create new streams of data.” The plan is to make Indian operations a hub for IT development and some other capabilities to manage its global manufacturing operations.","excerpt":"Noted US semiconductor company Micron Technology has grand plans to increase its hiring in India four times, specifically in the research and development sector over the next two years. The $30 billion company recently set up its R&D operations in Hyderabad which is supporting its global manufacturing operations. As the company is working on building […]","categories":["AI News"],"tags":["micron"],"author_name":"Prajakta Hebbar","publish_date":"2019-08-20T13:26:08","publication_year":"2019","word_count":261,"keywords":["micron","Go","machine learning","artificial intelligence","programming_languages:R","AI","data-driven","programming_languages:Go","ai_applications:autonomous driving","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","R","Go","data-driven","programming_languages:R","programming_languages:Go","ai_applications:autonomous driving"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/micron-to-hire-over-1500-techies-in-its-rd-facility-in-india\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":5720,"title":"Perils of Exit Poll Forecasting in India","content":"Exit polls have proven to be notoriously ineffective in predicting the actual result in India & critics point to the 2004 & 2009 elections as examples. In 2014, most exit polls have predicted a NDA\/BJP victory We have crunched through the numbers specifically for the BJP\/NDA seat projections & shown 5 nuggets that point out the hazardous nature of exit poll predictions in India Vozag.com Key Takeaways: The median national exit poll figure for the NDA government is 282 with a high of 354 (Chanakya) & a low of 249 (TimesNow) & a difference of 105 seats between the high & low Projected Number of Seats for NDA by Polling Firm TimesNow’s low estimate is driven by extremely low projections for NDA in Rajasthan & Madhya Pradesh which most experts expect to be swept by BJP TimesNow Comparison Versus Median Estimate in “Big 6” States for BJP\/NDA For the 6 biggest states for NDA\/BJP, the difference between the worst exit poll & the best exit poll (excluding Chanakya) is 60 seats. Critics can point to this difference & the 2004 & 2009 poor predictions of exit polls to the unreliability of such polls. Sum of Big 6 Best Polling for NDA= 199Sum of Big 6 Worst Polling for NDA= 139 This data excludes Chanakya since it is an outlier in terms of best performance Chanakya may seem like the “crazy” outlier with an aggressive projection of 340 seats for NDA but they were the most accurate exit polling firm in the 2013 State Assembly elections Estimates of Various Exit Polling Firms to Actuals in State Assembly Elections in November 2013 CNN-IBN’s estimates are in the mid-range of all exit polling estimates & are considered “typical”. However, a closer examination of polling data shows that in two key states tiny differences in vote share led to big changes in seat projections: 1) A 1% positive change in Maharashtra between the last opinion poll & the current exit poll shows a gain of 8 seats for the NDA & 2) Similarly, a 2% positive change in Karnataka leads to a 4 seat difference for the BJP About Vozag.com: Vozag.com is a quantitative research and data analytics service firm doing data analysis in the home remodeling, repair and contracting industry. Now and then they utilise their analytical capabilities to comment on a wide range of serious and frivolous issues alike.","excerpt":"Exit polls have proven to be notoriously ineffective in predicting the actual result in India & critics point to the 2004 & 2009 elections as examples. In 2014, most exit polls have predicted a NDA\/BJP victory We have crunched through the numbers specifically for the BJP\/NDA seat projections & shown 5 nuggets that point out […]","categories":["IT Services"],"tags":[],"author_name":"AIM Media House","publish_date":"2014-05-16T04:01:50","publication_year":"2014","word_count":397,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","analytics","CNN","R"],"extracted_tech_keywords":["AI","analytics","R","Go","CNN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/perils-of-exit-poll-forecasting-in-india\/","complexity_score":3,"technical_depth":7,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":20627,"title":"Be The Originator Of The Future – Study Robotics","content":"Lately, Artificial Intelligence has seen a lot of advancement and many have even started believing that robots and machines could soon replace humans. As the fear of losing jobs to robots continues to exist, here is a ray of hope. Robotics has moved far beyond science fiction to reality and is paving way for more interesting and challenging jobs for you to grab. What Is Robotics Robotics is just not about ‘building robots’, but is one of the most rapidly growing fields which has applications in varied industries. ‘Robotics’ includes all the research and engineering activities involved in the designing and development of robotic systems and automated machines, also the computer systems for their control, information processing, and sensory feedback. Currently, the demand to automate processes across industries, boost efficiency and minimize human risk is massive and so is the demand for brilliant minds who can create highly effective and low-cost robotic solutions. These are surely exciting times for those looking to study robotics and make a career in it. Robotic Engineers: Who Are They Are What Do They Do? Being a robotic engineer can be totally exciting. They are responsible to design robots, maintain and develop new applications for them, and conduct research to develop their potential. Why Choose The Robotics Field? Robotics is the new generation attraction in the field of science and technology offering wide opportunity in the future prospect. Here are few undoubted reasons for you to choose this field as your career path. Open Multiple Doors Of Opportunities – As an effectual robotic engineer, one has excellent hold on mathematical, logical, scientific, creative and computer skills. Innovation – Robotics is all about innovation, each day comes with new twist and turns leading to new innovations and possibilities. Challenging – Being a robotic engineer demands great knowledge, creative thinking, and problem-solving abilities. Continuous Learning – Alliance with the best mind in computing, mathematics and Physics enables you in continuous learning. Getting Started with Robotics? There are tons of routes to a job in robotics, one could study computer science, electronics, biotechnology, manufacturing or cognitive science. A robotic engineer must have advanced science and math skills and needs undergraduate degrees in mathematics, engineering, or physical science. One can advance their career and earn a higher salary with a master’s degree in robotics engineering. Other useful subjects are, computing and programming, design and technology, and specific engineering disciplines like automotive, electronics, bioengineering, mechanical engineering, and mechatronics. To work as an independent robotics engineer a Licensure is usually required. Apart of having the standard education and skillset, there are other lucrative ways to be a strong candidate for this well-paid gig. One of the best way is to check out Online Courses. Like many other tech-based fields, there are several online courses for Robotics, allowing anyone to learn more about the field as a whole and gain more specialized knowledge. Register for these free online courses to refine your understanding in Robotics: Robotics: Vision Intelligence and Machine Learning Robotics: Fundamentals Robotics: Dynamics and Control Robotics: Locomotion Engineering Coming to the salary and career outlook, they are quite promising for this field; According to PayScale, in 2016 The average salary of a robotics engineer was $82,000. The average yearly growth of the robotics industry is 9% and career opportunities for highly trained roboticists continue to rise. (source: A Roadmap for US Robotics, 2016). From designing robots to working on cutting-edge products which can be used in varied industries and study of the oceans to designing humanoid robots which can imitate basic human tasks; the potentials are indeed infinite. There many different ways to get started in the field of robotics. The key idea is to enhance your knowledge with hands-on experience and keep learning!","excerpt":"Lately, Artificial Intelligence has seen a lot of advancement and many have even started believing that robots and machines could soon replace humans. As the fear of losing jobs to robots continues to exist, here is a ray of hope. Robotics has moved far beyond science fiction to reality and is paving way for more […]","categories":["AI Trends"],"tags":[],"author_name":"AIM Media House","publish_date":"2018-01-15T07:44:36","publication_year":"2018","word_count":624,"keywords":["API","machine learning","artificial intelligence","programming_languages:R","AI","innovation","RAG","Ray","ViT","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","Ray","RAG","R","API","ViT","innovation","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/originator-future-study-robotics\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10103528,"title":"It Was Sam’s Plan All Along","content":"What a time to be alive. Now that Sam Altman is back as the CEO of OpenAI, after getting fired on Friday, then joining Microsoft on Sunday evening, there are a lot of questions about what actually transpired before the firing happened. It also looks like it was Altman’s plan all along. The idea was to change the board All of the recent developments hint towards a simple thing that the OpenAI board members were clearly not aligned with the for-profit direction that OpenAI was headed towards under Altman’s leadership. The old board members such as Helen Tomer had major contentions with what Altman was doing. Toner wrote a paper titled ‘Decoding Intentions’ that was critical of OpenAI, and was praising Anthropic. The paper said that Anthropic’s willingness to not release its AI models is “exactly the kind of frantic corner-cutting that the release of ChatGPT appeared to spur.” She further praised Anthropic for investing more into AI safety than OpenAI. People on X criticise her views, saying it sounds more like an opinion piece than a research paper. Simply put, she wanted to kill OpenAI to save humanity. She reached scarily close. call me crazy but I would also be angry if I was sama pic.twitter.com\/LrsA4UxXvF— Nick Arner (@nickarner) November 22, 2023 Interestingly, Toner was quiet during all the debacle. She posted on X after Altman is hired back with, “and now, we all get some sleep.” It might be a far fetched idea to say that it was Toner who started the coup against Altman, but there is still a chance. https:\/\/twitter.com\/ricburton\/status\/1727214451529752772 Not just Altman, Microsoft chief Satya Nadella, has previously expressed his disappointment with the former board, saying, “We are encouraged by the changes to the OpenAI board. We believe this is a first essential step on a path to more stable, well-informed, and effective governance.” “Sam, Greg, and I have talked and agreed they have a key role to play along with the OpenAI leadership team in ensuring OpenAI continues to thrive and build on its mission. We look forward to building on our strong partnership and delivering the value of this next generation of AI to our customers and partners” he added in his statement. Sam loved the weekend Altman has posted on X saying, “I love OpenAI, and everything I’ve done over the past few days has been in service of keeping this team and its mission together.” He added that he wants to continue working with the new board of OpenAI and also build a strong partnership with Microsoft. Interestingly, Microsoft is also going to be an observant part of the new board. i love openai, and everything i’ve done over the past few days has been in service of keeping this team and its mission together. when i decided to join msft on sun evening, it was clear that was the best path for me and the team. with the new board and w satya’s support, i’m…— Sam Altman (@sama) November 22, 2023 Seemingly, Altman loves everything that happened over the past few days, and it was in service of keeping the team together. It is quite worrisome to say the least, given all the chaos that was created. On the other hand, as soon as Altman was fired, the whole team showed solidarity, and started to force the board to resign. Interestingly enough, Ilya Sutskever, who broke the news to Altman, also switched sides later, and apologised publicly for harming OpenAI. He also signed the petition along with Mira Murati and 10 others for resignation of the board. With 95% of the employees ready to quit unless Altman comes back, there was no other option left for the company to do apart from changing the board members. It all worked out how Sam wanted it to. Now everyone would be back, without the old OpenAI board. The problem with the structure of OpenAI is that it is very twisted. A non-profit board controls the for-profit entity of the company. So even though Altman was not taking any salary from OpenAI, maybe he changed his mind now. Furthermore, the new board that comes along now would also be in favour of what Altman wants. What did Sam get? Firstly, it is reasonable to say that Altman has always been an investor, and thus there is very little reason to believe that the altruistic approach of saving humanity with AI is actually the case, and not just a charade. Even if it is, he believes in a for-profit approach towards AGI, which is unlike the other former board members. Secondly, Altman always knew that the board could fire him and he should not be trusted, as he said in an interview. Moreover, Elon Musk said six months ago, “At any point Microsoft could cut off OpenAI.” Combining that with Nadella’s point of view in all of this, who is the biggest investor in OpenAI, and has massive belief in Altman, the plan started as soon as Altman got fired. What tangled webs we weave— Elon Musk (@elonmusk) November 22, 2023 Even while considering that the firing on Friday was a surprise to Altman and Nadella, they must have shook hands that it is time for the board to go. Altman was never going to be a part of Microsoft and work under some other leadership, or start any other ventures, it was all just to scare the OpenAI board members, and the team helped in it. But Adam D’Angelo, one of the board members that voted for firing Altman, is still part of the new board. Moreover, the damage done to Microsoft and the investors is just collateral damage for Altman. Interestingly, Altman and Greg Brockman are still not going to be part of the board. Moreover, Altman has also agreed for an internal investigation to find out why he was fired. So all has not worked out in their favour. The new board will review the circumstances of Altman’s firing. Shear was right when he said, “Sam is a great sales guy.” It might be just his plan all along.","excerpt":"The OpenAI team, Satya Nadella, and everyone else just played along.","categories":["Global Tech"],"tags":["Emmett Shear","Greg Brockman","Microsoft","Mira Murati","Sam Altman","Satya Nadella"],"author_name":"Mohit Pandey","publish_date":"2023-11-22T17:10:33","publication_year":"2023","word_count":1018,"keywords":["Satya Nadella","Anthropic","ChatGPT","Go","Sam Altman","OpenAI","AI","Greg Brockman","Mira Murati","RAG","GPT","Emmett Shear","Rust","AI safety","R","Microsoft"],"extracted_tech_keywords":["AI","ChatGPT","OpenAI","Anthropic","RAG","R","Go","Rust","GPT","AI safety"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/it-was-sams-plan-all-along\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10100896,"title":"The Dire Need for LLM Benchmarks in India","content":"At present, discussions around generative AI predominantly revolve around China, the United States, and the Middle East, with notable players like TII (based in the UAE), Baidu (from China), and OpenAI (from the US) gaining significant attention. And any conversation about LLMs, in particular, is incomplete without mentioning how it fares on various benchmarks like MMLU, HumanEval, AGIEval etc. Be it GPT-4 or Llama 2, the first criteria the creators try to showcase about their LLMs is putting up their benchmark scores on the research paper. Surprisingly, India, which is emerging as a new force in AI, does not have a benchmark of its own to evaluate LLMs. There is still a need for a metric that India could utilise to evaluate LLMs, one that caters specifically to the country’s needs. Geography can also play a crucial role in understanding the purpose of an LLM, a factor that many people might not have paid attention to. India is a diverse nation with a multitude of languages, cultures, and contexts. A benchmark specifically designed for India would take into account the linguistic diversity of the country. From addressing regional language nuances to recognising cultural references, an Indian LLM benchmark could enable models to be more culturally sensitive. A glimpse of LLM Leaderboard with list of LLMs fairing across various benchmarks (Source: Hugging Face) Developers from around the globe create models and submit them to Hugging Face, which maintains an LLM Leaderboard. Here, users can check which model is performing best on different metrics like MMLU, ARC, HellaSwag, and more. Hugging Face even calculates an average based on these metrics. Undoubtedly, it is a very useful tool. However, the question remains: Are these metrics enough to judge an LLM? Perhaps to some extent, yes. Indian Datasets A majority of the benchmarks which have evolved from the US take their exams into consideration. For example, MMLU, covers 57 tasks including elementary mathematics, US history, computer science, and law. Similarly AGIEval, is derived from exams like SAT, LSAT and other tests like Chinese College Entrance Exam (Gaokao), law school admission tests, math competitions, lawyer qualification tests, and national civil service exams. In India, there are several tough competitive exams like UPSC, NEET, JEE-Advanced, CAT, etc, which cover the essence of India. Creating a benchmark for LLMs that’s tailored to Indian competitive exams like the UPSC would enable the development of language models capable of understanding the unique requirements of these exams, which often involve complex questions, diverse languages, and a deep understanding of India’s history, culture, and governance, alongside critical thinking and logical reasoning. Earlier this year, AIM tried to check GPT-3 for UPSC, surprisingly it failed. However, a few months later GPT-4 was able to cross the cutoff. Work in Progress It is unfortunate that India hasn’t been able to create an LLM from scratch until now. However, the situation appears to be changing as NVIDIA recently partnered with Indian giants like Reliance, TATA, and Infosys. Furthermore, Indian IT giant Tech Mahindra is currently working on an indigenous LLM known as Project Indus. This model will have the ability to speak in many Indic languages, notably Hindi, and would cover 40 different Indic languages initially. More languages originating from the country will also be added subsequently. The way India might use these LLMs can be completely different from the rest of the world. To evaluate Indian models on various tasks, we need to create a benchmark based on regional or vernacular dataset. Models like GPT-4, Llama 2, Falcon 180B and Mistral 7B are mostly trained on datasets in English. Hence, there is a dire need of datasets in Indian languages. As of today, the Indian constitution recognises 22 major languages of India. The Indian government has launched Project Bhashini where it looks at developing a National Public Digital Platform for local languages, which can be leveraged to create products and services for citizens using generative AI. Similar to Hugging Face’s LLM Leaderboard that tracks, ranks, and evaluates open LLMs and chatbots, Bhashini intends to create training and benchmarking datasets. It will lead to healthy competition and will kickstart an LLM race in India. What next? A majority of the benchmarks are created by the research wing of universities or tech companies. For example, MMLU was created by the University of California, Berkeley. It’s high time that Indian universities like IIT Delhi, IIT Madras and IISc Bengaluru started working towards creating a dataset which would lead to the creation of a benchmark. What helps is that the universities have easy access to data related to competitive exams. However, creating a benchmark is easier said than done. It requires a lot of computational infrastructures in terms of GPUs to run those models against the dataset in order to get their efficacy in different metrics. To kick start the process Indian universities can partner with Indian tech companies like Reliance, Tata and Infosys who will soon acquire tens of thousands of NVIDIA GH200 Grace Hopper Superchip. On his recent visit to India, NVIDIA chief Jensen Huang said that he will partner with universities like IITs as well to create AI infrastructure. “We would like to work with every single university. The first thing we have to do is build AI infrastructure” he said. Meanwhile, isn’t it a nice thought seeing OpenAI’s models claiming supremacy on an Indian benchmark?","excerpt":"Surprisingly, India, which is emerging as the new force in AI, does not have a benchmark of its own to evaluate LLMs","categories":["AI Features"],"tags":[],"author_name":"Siddharth Jindal","publish_date":"2023-09-29T15:00:00","publication_year":"2023","word_count":892,"keywords":["Go","Hugging Face","OpenAI","AI","chatbots","ML","RAG","Aim","generative AI","R"],"extracted_tech_keywords":["AI","ML","generative AI","OpenAI","Aim","Hugging Face","RAG","chatbots","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/the-need-for-llm-benchmarks-in-india\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":38344,"title":"Top 5 Soft Skills All Successful Data Scientists Possess","content":"Data science is one of the hottest fields in the industry and whenever there are job openings, throngs of data science experts as well as enthusiasts flock for them. In our many interactions with numerous CXO-level leaders in this sector, one of the key things pointed out was that while there is no dearth of data scientists in India, companies, especially bigger, established organisations always struggle to find the right talent. But why is that? Industry experts say that simply hiring a data scientist is not enough. Managers need to take special care to align business and data teams thus enabling data scientists to be self-sufficient. Otherwise, they might not get the expected ROI in data science which is a problem almost 80% of the companies face. In this article, Analytics India Magazine will put light on the top soft skills most successful data scientists possess: Communication Communication is the backbone of almost every job profile. This is especially true with data scientists. Conveying exact information or a problem statement — be it in the form of algorithms of interaction with peers — is crucial to success. It showcases a clear thought process and an uncluttered outlook on the part of the data scientist as well. Problem-Solving While it is important for a data scientist to keep themselves abreast on the latest tools and developments, it is mandatory for them to work on solving problems. A data scientist is like a doctor, the more problems they solve and more experience they have, they get better in their job. That is why companies value experience a lot more than the educational qualification. But it is important to have the basic educational qualification. A full-time course will be valued more than an executive course. Ability To Draw Parallels To Real-world Problems If a student wants to choose Data Science as a career, they should start paying attention to subjects such as Statistics, Probability, Algebra, Set theory and Data Structures and Algorithms. If they are strong with the basic concepts, then they can use the technology tools to their advantage to build great models. While a lot of theoretical knowledge can be gained by doing these courses, their learning would not be complete until it is applied to practical problems. Industry mentors can play a vital role in this aspect. They will also help in understanding the practical difficulties in applying their knowledge to real-world problems. This will also help them in building their domain knowledge that will help them to be a good data scientist. Prioritisation A good data scientist always has his\/her priorities straight. Because as a data scientist, he\/she is bombarded with many questions, which can be answered in numerous ways. ”First, you need to decide which of those questions are actually worth answering and how much effort is worth putting into those questions. If someone asks you for a detailed description of how your users <x> with lots of demographic breakdowns and trends, maybe the first thing to do is figure out how many people actually <x>. When you find out its 0.01% of your population, which you can probably know in ten minutes or less, you should realise <x> isn’t worth understanding and convince your requester to find another question (persuasion coming into play again),” writes Chris Luhrs, a data scientist. Business Acumen Data scientists deal with such enormous information that if left “untranslated”, becomes meaningless for the upper-level management. Sahana Shetty, the HR Leader for Technology at ANZ, told AIM in an interview, “We also look at a data scientist’s business acumen while hiring him\/her. We are keen to have individuals who are high on potential, who exhibit a growth mindset and have an inclination to learn and grow.” Also see:","excerpt":"Data science is one of the hottest fields in the industry and whenever there are job openings, throngs of data science experts as well as enthusiasts flock for them. In our many interactions with numerous CXO-level leaders in this sector, one of the key things pointed out was that while there is no dearth of […]","categories":["AI Trends"],"tags":["Data Science","Data Scientist"],"author_name":"Prajakta Hebbar","publish_date":"2019-04-26T12:49:41","publication_year":"2019","word_count":624,"keywords":["data science","Go","programming_languages:R","AI","Aim","llm_models:Bard","ViT","analytics","GAN","Data Science","Data Scientist","R"],"extracted_tech_keywords":["AI","data science","analytics","Aim","R","Go","GAN","ViT","llm_models:Bard","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/5-soft-skills-all-successful-data-scientists-possess\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10077662,"title":"Databricks Appoints Anil Bhasin as Country Manager, Expands Presence in India","content":"IT firm Databricks announced the appointment of Anil Bhasin as the country manager for India last week. Bhasin will be responsible for leading the company’s go-to-market functions and driving lakehouse adoption for enterprises and digital-native customers across the region. The tech industry veteran brings over 30 years of experience to the role and will lead one of the company’s fastest-growing regions in the Asia-Pacific market. He will work closely with Ed Lenta, senior vice-president and general manager of Databricks Asia Pacific. Bhasin joins the firm most recently from UiPath where he served as the managing director and VP of the company’s India and South Asia business. Prior to this, he served as the regional vice-president at Palo Alto Networks for over seven years, playing an instrumental role in building the company’s presence in the India and SAARC region. Bhasin said, “Data plays a critical role in fueling digital transformation for businesses across every industry, and Databricks is pioneering the data architecture of the future with its lakehouse platform. I am thrilled to be joining team Databricks during such a pivotal time in the company’s journey and look forward to collaborating with our customers and partners to bring the lakehouse vision to life in India.” In order to support Databricks’ local customer base and growing team, the company in a press release confirmed its recent expansion to a new office space in Bengaluru’s Embassy TechVillage. Moreover, the number of Databricks employees in India has grown nearly 8x over the last two fiscal years, with more than 150% growth YoY. The India team is a significant part of the company’s Asia-Pacific footprint and would continue to rapidly hire in India for a variety of customer-facing and technical roles. Headquartered in San Francisco, Databricks is an AI company which caters to more than 7,000 organizations worldwide — including Condé Nast, Comcast, and over 50% of the Fortune 500 — which rely on the Databricks Lakehouse Platform to unify their data and analytics.","excerpt":"Bhasin will lead the company’s go-to-market functions and drive lakehouse adoption for enterprises and digital-native customers across the region","categories":["AI News"],"tags":["Databricks","Expansion","uipath"],"author_name":"Bhuvana Kamath","publish_date":"2022-10-20T12:27:26","publication_year":"2022","word_count":329,"keywords":["Go","API","programming_languages:R","AI","digital transformation","Git","Expansion","analytics","uipath","GAN","R","Databricks"],"extracted_tech_keywords":["AI","analytics","Databricks","R","Go","Git","API","GAN","digital transformation","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/databricks-appoints-anil-bhasin-as-country-manager-expands-presence-in-india\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10062191,"title":"6 AI research papers you can’t afford to miss","content":"Qualitative research lies at the heart of innovation. The survival of any domain is predicated on research, and it especially holds true for fledgeling fields like artificial intelligence. Inarguably, research is the catalyst pushing the frontiers of complex fields like AI and ML. To bring you up to speed on the critical ideas driving artificial intelligence, we handpicked top drawer research papers from google scholar based on the number of citations. Adam: A Method for Stochastic Optimization(2015) – 1,00,651 citations The paper from Diederik P.Kingma and Jimmy Lei Ba, details ADAM, an algorithm for first-order gradient-based optimization of stochastic objective functions based on adaptive estimates of lower-order moments. The computationally efficient method has few memory requirements and is invariant to diagonal rescaling of the gradients. Adam is built for problems that are large in terms of data and\/or parameters. The method is also ideal for non-stationary objectives and problems with noisy gradients. The hyper-parameters have intuitive interpretations and therefore require little tuning. Adam works well in practice and rates highly in comparison to other stochastic optimization methods. Imagenet classification with deep convolutional neural networks(2012) – 1,04,283 citations The researchers Alex Krizhevsky, Ilya Sutskever and Geoffrey E Hinton trained a large, deep convolutional neural network to classify the 1.2 million high-resolution images in the ImageNet LSVRC-2010 contest into the 1,000 different classes. The neural network, with 60 million parameters and 500,000 neurons, consists of five convolutional layers, some of which are followed by max-pooling layers, and two globally connected layers with a final 1000-way softmax. Source: DNN architecture from neurips.cc ImageNet is a dataset of over 15 million labelled high-resolution images across 22,000 categories from the web and labelled by human labellers using Amazon’s Mechanical Turk crowd-sourcing tool.  The study was conducted on a set of 2 GTX 580 GPUs for cross-GPU parallelisation. The results proved a large, deep convolutional neural network is capable of achieving state-of-the-art results on a challenging dataset using supervised learning. However, the network’s performance drops if a single convolutional layer is removed. Distributed representations of words and phrases and their compositionality(Word2Vec)(2013) – 32,320 citations Tomas Mikolov, Ilya Sutskever, Kai Chen, Greg Corrado and Jeffrey Dean built on the Skip-gram model, an efficient method for learning high-quality distributed vector representations that capture a large number of precise syntactic and semantic word relationships. In this paper, the researchers presented several extensions that improve both the quality of the vectors and the training speed. By subsampling the frequent words, the team gained significant speedup and also learned more regular word representations. Source: researchgate.net The paper also proposed a simple alternative to the hierarchical softmax called negative sampling. For training the Skip-gram model, an internal Google dataset consisting of 1 billion words was used. Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift(2015) – 34,857 citations Sergey Ioffe and Christian Szegedy’s paper posited different methods to improve the speed and stability of training neural networks through normalisation of the input parameters. Deep networks trained with SGD(Stochastic gradient descent) were used to achieve cutting edge performance. However, this simple and effective method comes with a cost as layers need to continuously adapt to the new distribution. The implementation of Batch normalisation with ImageNet classification network matches the performance of previous methods with only 7% of the training steps. Faster R-CNN: towards real-time object detection with region proposal networks(2015) – 40,122 citations The study from Shaoqing Ren, Kaiming He, Ross Girshick, and Jian Sun, showed ‘Faster R-CNNs’ can be used for object detection in images and videos at an improved scale. Source: arxiv.org The paper presented a unified, deep-learning-based object detection system that runs with near-real-time frame rates. The methodology also improves region proposal quality and thus the overall object detection accuracy. Generative Adversarial Networks (GANs)(2014) – 41,545 citations In this paper, Ian J Goodfellow, Jean Pouget-Abadie, Mehdi Mirza, Bing Xu, David Warde-Farley, Sherjil Ozair, Aaron Courville, Yoshua Bengio, proposed the framework for estimating generative models by simultaneously training two models, a generative model G and a discriminative model D to showcase the potential of the framework through qualitative and quantitative evaluation of the generated samples. Source: arxiv.org The paper showed GAN is a viable framework for adversarial modelling. This powerful class of neural networks that are used for unsupervised learning takes up a game-theoretic approach, to provide solutions for image detection.","excerpt":"ImageNet is a dataset of over 15 million labelled high-resolution images across 22,000 categories.","categories":["AI Trends"],"tags":["AI (Artificial Intelligence)","AI Research","Convolutional Neural Networks","Deep Neural Networks","image recognition","Language Models","Machine Learning","Object Detection","research papers","supervised learning"],"author_name":"Kartik Wali","publish_date":"2022-03-07T13:00:00","publication_year":"2022","word_count":722,"keywords":["Convolutional Neural Networks","object detection","CNN","R","artificial intelligence","image recognition","supervised learning","Go","AI","neural network","ML","Machine Learning","Object Detection","Language Models","GAN","research papers","AI Research","Aim","Deep Neural Networks","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","ML","neural network","Aim","object detection","R","Go","GAN","CNN"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/6-ai-research-papers-you-cant-afford-to-miss\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10163919,"title":"Y Combinator Introduces Summer Fellow Grants","content":"San Francisco-based Y Combinator is offering Summer Fellows Grants to undergraduate computer science or engineering students. The grant provides $20,000 in cash and $90,000 in cloud computing credits from AWS, Microsoft Azure, and Google Cloud Platform. Additional credits from YC companies Fly.io ($50,000), Supabase ($10,000), Replicate ($10,000), and Resend ($3,000) are also available. This summer, YC will be giving out grants to college students to work on their own technical projects.We’re calling these the Summer Fellows Grants.We hope this will encourage the smartest students to indulge their intellectual curiosity and work on hard technical things. pic.twitter.com\/UFT88ZzxeB— Y Combinator (@ycombinator) February 17, 2025 Fellows will attend a kick-off dinner on June 15, co-work at YC’s office from June 18-20 and August 20-22, and present their projects on August 22. The grants are intended for students who want to build technical projects, especially those using AI, rather than doing typical internships. According to the post, eligible applicants include any undergraduate students, including freshmen, with strong technical skills. Non-computer science majors with technical experience, such as open-source contributions, can also apply. Teams should apply individually and mention any collaboration in the application. The program requires a full-time commitment to the project, but part-time applicants with internships may also be considered. YC encourages relocation to the Bay Area but will cover travel for the co-working sessions. Fellows will be invited to attend AI Startup School on June 16 and 17. The event will feature leaders, including Tesla CEO Elon Musk, OpenAI CEO Sam Altman, OpenAI co-founder Andrej Karpathy, and World Labs CEO Fei-Fei Li, among others, who will share insights on building the future of AI.","excerpt":"The grants are intended for students who want to build technical projects, especially those using AI.","categories":["AI News"],"tags":["Y Combinator"],"author_name":"Aditi Suresh","publish_date":"2025-02-18T11:14:46","publication_year":"2025","word_count":272,"keywords":["Go","OpenAI","cloud computing","AI","AWS","R","RPA","Y Combinator","RAG","ViT","Azure"],"extracted_tech_keywords":["AI","OpenAI","RAG","cloud computing","AWS","Azure","R","Go","ViT","RPA"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/y-combinator-introduces-summer-fellow-grants\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10133585,"title":"Chevron Invests $1 Billion in Bengaluru for Largest Global Innovation Hub","content":"In a major boost to Karnataka’s tech landscape, Chevron, a global leader in energy solutions, has announced a $1 billion investment to set up its largest global innovation center in Bengaluru. This Global Capability Center (GCC) will focus on advanced research in carbon sequestration and support the company’s operations worldwide. Chevron’s new facility, ENGINE or the Chevron Engineering and Innovation Excellence Centre, marks a significant milestone for the 145 year old, $200-billion company as it focuses on developing affordable, reliable, and cleaner energy solutions. Located in Bengaluru, the tech center aims to hire 600 professionals by the end of 2025, with plans to expand further, according to India country head Akshay Sahni in an interview. “Bengaluru has massive scientific talent, and I am thrilled that we will be building a team here which will be at the very nexus of solving today’s energy challenges and delivering tomorrow’s lower carbon energy solutions,” said Sahini. Bengaluru is home to 570 GCCs and commands 39% of India’s market share in the sector. It continues to solidify its status as the premier destination for global capability centers. The city’s talent poll and vibrant tech ecosystem serve as the key drivers for such major investments. India’s Tech Scene India’s tech landscape continues to flourish as Cummins Group in July inaugurated its first IT Global Competency Center (GCC) at its Balewadi campus in Pune, Maharashtra. The GCC represents a pivotal component of Cummins’ extensive IT transformation strategy aimed at enhancing operational efficiencies and accelerating time-to-market for its products and services. Recently, Flutter Entertainment, a global leader in online sports betting and iGaming, also inaugurated a GCC in Hyderabad, signifying a significant investment in the country’s skills market. Similarly, Microsoft is also expanding its influence in India with the launch of a new innovation hub in Bengaluru, cementing the city’s status as a tech powerhouse.","excerpt":"Bengaluru is home to 570 GCCs and commands 39% of India’s market share in the sector.","categories":["AI News"],"tags":["Bengaluru","GCC","Microsoft"],"author_name":"Vidyashree Srinivas","publish_date":"2024-08-23T12:13:34","publication_year":"2024","word_count":308,"keywords":["GCC","programming_languages:R","AI","innovation","Aim","Bengaluru","R","Microsoft"],"extracted_tech_keywords":["AI","Aim","R","innovation","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/chevron-invests-1-billion-in-bengaluru-for-largest-global-innovation-hub\/","complexity_score":2,"technical_depth":5,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":25760,"title":"Healthcare Startup SigTuple Raises $19 Million In Series B Round Funding","content":"(L-R) SigTuple founders Apurv Anand, Tathagato Rai Dastidar and Rohit Kumar Pandey. Bengaluru-based healthcare startup SigTuple has raised $19 million (₹130 crore) in the Series B round of funding. Founded 2015 by former American Express employees Rohit Kumar Pandey, Tathagato Rai Dastidar and Apurv Anand, the company has raised $6.5 million from different investors so far. The Series B round was led by IDG Ventures and Accel Partners along with pi Ventures, Endiya Partners, VH Capital, Axilor Ventures and Flipkart’s executive chairman Binny Bansal. SigTuple has also raised venture debt from Trifecta Capital in this round. The company has been working on a low-cost system for examining blood, urine and semen samples using artificial intelligence to recognise microscopic images. Using SigTuple’s own software, their machine can now recognise these samples to differentiate between normal and abnormal cells. Pandey the chief executive at SigTuple told a financial newspaper, “The major chunk [of money from this investment] will go into setting up a manufacturing pipeline so that we can put more devices in the market. We have applied for US FDA and CE certification. Once we go through the stage, it will require more clinical trials for different market deployments. We will be investing money in this. We need to have a sales and support teams as well.” SigTuple has been using AI to develop hardware and software products that would digitise pathology results from hospitals. Now, with the new funding in hand, SigTuple plans to use it to further advance in research and development of their products. They also want to enhance their assembly, distribution and conduct more clinical trials. “There is a tremendous opportunity to make healthcare accessible, affordable and accurate. Year after year our confidence to transform healthcare delivery has increased. We are excited about the future as we are gearing up to take our solutions and services to the larger population,” added Pandey. The journey has only begun !!!https:\/\/t.co\/BtnpqLri9S — SigTuple (@sigtuple) June 26, 2018","excerpt":"Bengaluru-based healthcare startup SigTuple has raised $19 million (₹130 crore) in the Series B round of funding. Founded 2015 by former American Express employees Rohit Kumar Pandey, Tathagato Rai Dastidar and Apurv Anand, the company has raised $6.5 million from different investors so far. The Series B round was led by IDG Ventures and Accel […]","categories":["AI News"],"tags":["Binny Bansal","Flipkart","pi ventures"],"author_name":"Prajakta Hebbar","publish_date":"2018-06-26T05:45:54","publication_year":"2018","word_count":327,"keywords":["Go","API","funding","artificial intelligence","programming_languages:R","AI","Binny Bansal","Flipkart","programming_languages:Go","Git","pi ventures","R","startup"],"extracted_tech_keywords":["AI","artificial intelligence","R","Go","Git","API","startup","funding","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/healthcare-startup-sigtuple-raises-19-million-in-series-b-round-funding\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":62393,"title":"How To Address Offline Reinforcement Learning Challenges","content":"Reinforcement learning is one of the most important techniques used to achieve artificial general intelligence. However, it has various disadvantages that prevent researchers from achieving true AI. Since AI agents are trained to learn by hit and trial method, providing every possible real-world circumstance is a huge challenge. The world is yet to address these problems effectively, but in many real-world RL applications, there already exists a colossal amount of previously collected interaction data. Such existing information can be leveraged to make RL feasible and enable better generalisation by incorporating diverse prior experience through offline reinforcement learning. Offline Reinforcement Learning Unlike online RL, where the AI agent incrementally improves its policy as new experience becomes available, offline RL works on a fixed set of logged experiences without any further interaction with the environment. However, it can easily be adapted to the growing batch setting if needed. Off-policy RL eliminates the need for repeated training of AI agents to scale, assists in evaluating models with the existing dataset of interactions, and enables developers to deliver real-world impact quickly. Although offline RL can decrease the requirement of computational power, it comes with challenges that restrict the use of off-policy RL. While adopting this technique, one fails to understand the reward that an agent collects during a mismatch between online interactions and any fixed dataset of logged interactions. In other words, if the model, which is being trained with an existing dataset, takes an action that is different from the data collection agent, one cannot determine the reward provided to the learning model. Overcoming Challenges In Offline Reinforcement Learning Researchers at first trained a DQN agent on Atari 2600 games and logged the experience. Then they proposed a random ensemble mixture (REM) — a robust Q-learning algorithm that enforces optimal Bellman consistency on random convex combinations of multiple Q-value estimates — to enhance generalisation of the model. Using the collected data, a new RL model was trained, which outperformed AI agents that were trained within an environment. The researchers noted that superior results could be achieved even without explicitly correcting for any distribution mismatch. However, on various occasions, the off-policy RL agent either diverges or yields poor performance. Consequently, to fix such problems, researchers used regularisation techniques to keep the policy close to the dataset of offline interactions. Therefore, the DQN model was again used to collect the dataset of 60 Atari 2600 games for 200 million frames each by using sticky actions to make problems more challenging. Five DQN was trained on each of the 60 games with different random initialisation and stored every state, action, reward, and next state, resulting in collecting 300 datasets. Similarly, data for QR-DQN was gathered for training and comparing offline models’ performance. On training models with the collected data of online DQN and QR-DQN, the offline agents outperformed both the online models. Offline DQN underperforms fully-trained online DQN on all except a few games, where it achieves higher scores with the same amount of data. Offline QR-DQN, on the other hand, outperforms offline DQN and fully-trained DQN on most of the games. These results demonstrate that it is possible to optimise reliable agents offline using standard deep RL algorithms. Furthermore, the disparity between the performance of offline QR-DQN and DQN indicates the difference in their ability to exploit offline data. Comparison Of New Offline RL Models Although the offline RL demonstrated superior performance, there is a need for generalisation in off-policy RL. Consequently, researchers leveraged techniques from supervised learning that use an ensemble of models to enhance generalisation with two new offline RL agents: Ensemble-DQN and Random Ensemble Mixture (REM). While the former is a simple extension of DQN that trains multiple Q-value estimates and averages them for evaluation, the latter combines various Q-value estimates and uses this random combination for robust training. Further, offline REM agents’ performance was evaluated against the offline DQN and QR-DQN. Offline REM outperformed the other two off-policy models as it generalises better than the others. Outlook Researchers attributed the failure of previous work to the size and diversity of data that was collected by the online RL agent. Unlike supervised learning, the performance of offline agents increases as the size of data increases. Besides, researchers used the first 200 million frames per game in the DQN dataset, which delivered exceptional results. Thus, online RL agents work well in the offline setting with sufficiently diverse datasets. Although offline RL outperformed the online agents, it requires off-policy evaluation for hyperparameter tuning and early stopping to manage the training of such models effectively. Nevertheless, offline RL has shown promising results, which can lead to broader adoption of the technique over online RL. Also Read: Why the upside-down RL is a paradigm shift in AI.","excerpt":"Reinforcement learning is one of the most important techniques used to achieve artificial general intelligence. However, it has various disadvantages that prevent researchers from achieving true AI. Since AI agents are trained to learn by hit and trial method, providing every possible real-world circumstance is a huge challenge. The world is yet to address these […]","categories":[],"tags":["Reinforcement Learning"],"author_name":"Rohit Yadav","publish_date":"2020-04-23T14:00:00","publication_year":"2020","word_count":791,"keywords":["Go","Reinforcement Learning","programming_languages:R","AI","emerging_tech:AI agents","RPA","programming_languages:Go","RAG","AI agents","R"],"extracted_tech_keywords":["AI","RAG","R","Go","RPA","AI agents","programming_languages:R","programming_languages:Go","emerging_tech:AI agents"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-to-address-offline-reinforcement-learning-challenges\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10080649,"title":"Comet Introduces Kangas, An Open Source Data Analysis and Debugging Tool","content":"Last week, one of the leading MLOps platforms ‘Comet’ introduced its new product—Kangas—the first open-source tool for democratising real-time data exploration and analysis for the ML community and computer vision. Click here to try out Kangas. Founded in 2017, the New-York headquartered company was established by former Googler Gideon Mendels. Comet provides data scientists and ML teams an MLOps platform to optimise, manage, and accelerate development processes—from training quick runs to monitoring models—across the ML lifecycle. The platform caters to over 150+ enterprise customers such as Cepsa, Affirm, Uber, Zappos and Etsy. Moreover, Comet is free to academic teams and individuals. As industries have witnessed rapid digitisation—around 97.2% of the companies claim the adoption of data science and analytics in some shape—which has resulted in a growing interest among investors and technopreneurs. Moreover, a Deloitte report states that the MLOps market is expected to reach USD 4 billion by 2025. Well-known MLOps players include Qwak, ZenML, Comet, Domino Data Lab, Weights and Biases, among others. Read: Deploying ML models in production with Qwak So, how is Kangas any different? Comet teams said that the open-source data exploration tool, Kangas, is based on a cutting-edge technology to help users understand and debug their data in a highly intuitive way. With the help of Kangas, visualisations are generated in real time. This enables every machine learning practitioner to sort, filter, group, query and interpret structured and unstructured data. This would further let the users derive meaningful information and accelerate better model development. Comet CEO Gideon Mendels said that the key component of data-centric machine learning is to understand how the user’s training data impacts the results of their models—when model predictions are wrong. He further elaborated that the new tool would successfully accomplish these goals to improve the overall experience. Considering the unique needs of ML practitioners, Kangas is a scalable and interoperable tool that enables users to discover patterns that are buried deep within the clusters of datasets. Data scientists will be able to query their large-scale datasets in a natural manner, allowing them to interact with the data in novel ways. Some of the noteworthy benefits of Kangas are unparalleled scalability to handle large datasets with very high performance, purpose built with ML concepts such as scoring, bounding boxes, and auto-generated statistics. Kangas supports different forms of media—not limited to traditional text queries—such as images, and video. Moreover, it can also run as a standalone local app, in a notebook, or can even be deployed as a web app.","excerpt":"The company stated in a press release that the product aims to democratise real-time data exploration and analysis for the ML community and computer vision.","categories":["AI News"],"tags":["Open Source AI"],"author_name":"Bhuvana Kamath","publish_date":"2022-11-24T13:23:11","publication_year":"2022","word_count":419,"keywords":["data science","Go","machine learning","AI","ML","MLOps","computer vision","Open Source AI","Aim","analytics","R"],"extracted_tech_keywords":["AI","machine learning","ML","computer vision","data science","analytics","MLOps","Aim","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/comet-introduces-kangas-an-open-source-data-analysis-and-debugging-tool\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":38360,"title":"Elon Musk’s Big Message: Not Just An Electric Carmaker, But A Tech Powerhouse","content":"Image Source: Tesla Will Elon Musk’s pipe dreams ever come true? While it’s true there isn’t anyone who has a clearer self-driving vision than Musk himself, the billionaire visionary is known for making ambitious statements. However, at the recently concluded Tesla Autonomy Day, Musk, who has been severely criticized for touting an unachievable vision for Autopilot is back to doing what he does best. He peddled a driverless future with the ‘best chip in the world’ now being shipped in its cars. This is touted as a concrete step towards driverless car vision which can achieve Level 4 autonomy. The billionaire predicted that it would take about another six months for the software to become reliable enough and not require human supervision. All in all, he made the point that the company is finally close to achieving the vision of self-driving — the holy grail of the automotive industry. In fact, by the end of 2020, Musk Tesla vehicles would be operating in Uber mode, delivering driverless rides to people. The in-house autonomous computing technology that was in the making for the last three years and is now powering Teslas takes on Nvidia’s Drive PX2 platform with its homegrown AI chip that is touted to bring the cost of the car down. Musk Trying To Find A Middle Ground With FSD In many ways, Musk has been viewed as a rule-breaking businessman whose vision doesn’t match the investors expectations. For long, all of Musk’s businesses (for eg.SpaceX) typify a vision and passion that make it a risky bet for investors. And Musk himself has been known for generating headlines that spark concern amongst the investor community. Behind the vision, Tesla fails to match up its astronomical ambitions with growth that investors and shareholders expect in general. What Musk’s company does best is generate a feverish news cycle that can keep the stock ticker going. With the announcement of its homegrown Full Self Driving hardware, Musk is moving into an area of software licensing long viewed by Google-owned Waymo that has run several pilots around cities. With the launch of FSD, Tesla has proven it is not just an electric car maker, but an automation major that just stole a march over Google and other players struggling with complete automation like Uber and Lyft. 5 Ways Tesla’s AI Chip Can Open Up New Growth Areas When it comes to Tesla, it is yet to turn profitable and create a sustained demand for its vehicles. As per the company’s 2019 Vehicle Production & Deliveries report indicated, the automotive major delivered 63,000 vehicles to customers in the first quarter of 2019. While this was 110% more than the same quarter last year, it was  31% less than the previous quarter. As per reports, Tesla’s quarter-to-quarter decline in deliveries has investors worried. This means, as one analyst puts it, that Musk is still a better ‘credit risk’ and needs to find growth areas for its company. 1) Compete with Waymo with licensing software: To put things in perspective, Musk seems to have a larger vision for Tesla which he believes is not just an automotive major but a tech company.  Tesla, like Waymo wants to make serious gains from licensing its self-driving software which can lead to massive growth down the road. 2) Rival Uber with robotaxi service: To further strengthen its position as a leader in the driverless market, Musk proposed an app that could allow car-owners to commit their vehicles for the ride sharing service or better still robo-taxi service. This proposed move would allow Tesla to rival Uber and Lyft that are struggling with their driverless tech. 3) Takes the AI chip battle to former chip supplier Nvidia’s home turf : According to Musk, FSD hardware which is powered by GPU and CPU along with 8 cameras and 12 ultrasonic sensors paves the for full autonomy. The Tesla team reportedly drubbed Nvidia’s performance, saying it raised the bar for all other automakers. 4) Open platform that the auto industry can leverage: Making its own homegrown chip allows  Tesla to open its platform to automotive companies that are reliant on Nvidia’s platform. Musk is going after Nvidia’s business that is posited as the backbone for autonomous vehicles for companies like Audi, Toyota, Volvo, Mercedes-Benz among other leading automakers. Nvidia that already has a proven technology in the driverless space will now have to gear for a battle with Tesla that may soon license its autonomous system for other carmakers. 5) Emerges as a winner in data, hardware and algorithm battle: Tesla openly challenged Nvidia’s dominance as a catalyst for autonomous systems which is the only open platform available for automakers.  However, when it comes to real-world data on self-driving, Nvidia can’t rival Tesla or Waymo in this space that have gathered real data on to help engineers build better self-driving technology. Outlook While it is yet to be seen whether Musk’s projections will come true, a software licensing and ride-hailing mode can make Tesla the biggest tech company that provides an overall solution to carmakers.","excerpt":"Will Elon Musk’s pipe dreams ever come true? While it’s true there isn’t anyone who has a clearer self-driving vision than Musk himself, the billionaire visionary is known for making ambitious statements. However, at the recently concluded Tesla Autonomy Day, Musk, who has been severely criticized for touting an unachievable vision for Autopilot is back […]","categories":["AI Features"],"tags":["Elon Musk","Tesla","what is power bi"],"author_name":"Richa Bhatia","publish_date":"2019-04-28T05:38:41","publication_year":"2019","word_count":843,"keywords":["Go","what is power bi","programming_languages:R","AI","data_tools:Spark","programming_languages:Go","RAG","Elon Musk","automation","Tesla","R"],"extracted_tech_keywords":["AI","RAG","R","Go","automation","programming_languages:R","programming_languages:Go","data_tools:Spark"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/elon-musks-big-message-not-just-an-electric-carmaker-but-a-tech-powerhouse\/","complexity_score":4,"technical_depth":8,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":9013,"title":"Top 10 IoT startups in India &#8211; 2016","content":"Internet of Things (IoT) seems to be the latest buzzword in the IT and analytics community. Typically, IoT goes beyond the machine-to-machine communications and covers a variety of devices and applications. Consequently, analytics is used for predictive decision-making using data generated from these M2M and IoT devices. Thus, making it imperative for IoT companies to have indepth knowledge of analytics to realize their full potential for developing futuristic products. With various funding options available, the number of startups continues to be on the rise in India. In order to keep up with the blurring borders between analytics and IoT, it is important to support and watch out the early birds in the world of IoT startups. So, we present to you, the emerging IoT startups (listed in alphabetical order) to keep on your radar for 2016! Altiux Altiux innovations is a software and product engineering services company with a focus on IoT. Their software and services portfolio encompasses the entire IoT ecosystem including edge nodes, gateways, IoT platform and data analytics solutions and can be leveraged by Semiconductor companies, OEMs and system integrators to develop highly differentiated products for the connected world for accelerating design, development and deployment of cutting edge IoT solutions. Their product engineering services portfolio includes consulting services, device engineering services, application engineering services and analytics services to help build highly differentiated IoT solutions. Altizon Systems Founded in April 2013, by Vinay Nathan, Yogesh Kulkarni and Ranjit Nair and funded by The Hive, Infuse Ventures and Persistent Venture Fund, Altizon is headquartered in Pune, India and has a presence in Palo Alto, CA. Altizon is the world’s first Industrial Internet Platform Company focused on making Enterprises Internet of Things (IoT) ready. Their flagship product, the Datonis IoT Platform, available both in a SaaS as well as in a Hosted model, helps in building IoT product in weeks by providing device connectivity kits, a device management layer, a highly-scalable, real-time, big-data analytics engine and alerting and monitoring services. Datonis easily integrates with the existing IT systems to provide a seamless transition between your IoT devices and your IT infrastructure management tools.They work with enterprises from Manufacturing and Cleantech sector. Primarily, their customers use Datonis for Condition Monitoring, Predictive Analytics, Operations and for driving Consumer Insights. Cooey Catering to the needs of today’s senior citizens who prefer to live independently, in an assisted living environment or retirement community,Cooey has targeted the lack of mechanism that captures their comprehensive medical history and hospital services. Cooey has developed dynamic, holistic, personalized and easy-to-understand solutions that empower chronically ill patients to safely and securely manage their health information from a single spot – accessible anytime, anywhere. Cooey’s end-to-end health-monitoring IoT platform allows the users to collect, store, analyze and share their medical summary with the doctor, while providing insights and personalized health tips to patients based on their medical profile. Further, it helps connect users with various healthcare service provides to get personalized services, viz., medicines, lab tests, or home care services, delivered to their doorstep. Cooey has introduced three devices so far – Wireless Body Fat Analyzer and Blood Pressure Monitor were launched on Flipkart in November 2015. It’s Smart Glucometer is coming soon. The Company was founded in 2015 by Manu Madhusudanan and Prabhakaran T P. EcoAxis is amongst the pioneers in the Industrial IoT space in India. Since 2007, EcoAxis has analyzed over 7.5 Trillion1 data points and delivered valuable business insights to top manufacturing, hospitality and commercial facilities and has helped reduce cost, maximize efficiency and gain improved control over productivity leaks, unlocking potential and business value from captive data. EcoAxis solutions includes, SuperAxisTM, a cloud-based analytics platform that provides integrated analytics on critical process and business information. Additionally, their Smart Manufacturing Suite integrates machinery inputs and processes within plants and factories to improve operational efficiencies. Further, access across the value chain to reports, notifications and predictions help continuously optimize business results. Also, their Smart Sustainability Suite generates intelligence from captive ecosystem to consistently improve quality and comfort for the end-user, thereby allowing resource optimization in a sustainable and responsible manner. Leverages the twin competencies of Resource monitoring (electrical energy, fuel, water, air and noise) and Resource usage analytics for facilities, plants such as WTP, STP, chillers, pollution monitoring. Then they also have a Smart Technology Suite that enabling OEM products such as batteries, compressors, boilers and the like, to be “intelligent” in order to improve efficiency and maximize product-life while reducing total cost of ownership. Entrib Entrib aims to bring new age IT to the manufacturing shop floor with their product ShopWorx. ShopWorx is a real time big data monitoring & analytics solution, specifically targeted towards operations of a manufacturing shop floor and has been customized for two verticals – Plastics Manufacturers and Automotive Assembly line. Using real time data from devices, sensors, machines, big data based analytics, combined with the power of mobile apps, ShopWorx brings people, machines and processes in a unique manner solving specific problems seen in the Indian Manufacturing industry. They have Saved $200,000+  for Smart Customers, Monitored 50 Million+ Products, Connected 700+ Smart Machines, Saved 100,000+ hours for Smart Customers, and more. Further, Entrib is also working on getting ShopWorx to predict machine downtime, maintenance, energy consumption and connecting the machines to the enterprises. SPAN Span is an established software services company offering comprehensive IT services since 1994 with a global footprint through their offices in U.S., Singapore, India and Europe. Their clients include Fortune 1000 companies, software firms (ISVs) and tech start-ups. Interestingly, Span is owned by the largest Nordic IT services major, Evry. Span’s Digital Transformation Stack (DTS) provides end-to-end solutions for organizations to accelerate their digital journey. DTS is currently used by many startups and organizations to speed up DevOps and gain real-time insights critical for their businesses. This stack comprises of Digital Experience, Digital Connect, Digital Analyzer, Digital Presence and Digital Operations. Smart Device Connect is a part of Digital Connect which offers an easy way to establish, monitor and manage integrations across devices and IoT platforms. Maven System Founded by Dhananjay Kulkarni and Sunil Desai, Maven Systems is a M2M \/ IoT product development company with expertise in wireless connectivity. Their solutions feature technologies like ZigBee, BLE, Wi-Fi, GPRS, 3G, and WiMAX. Their indigenous solutions are used for smart metering, smart lighting, telematics, remote monitoring of mining and construction equipment, diesel generator sets, solar panels and so on.1 In fact, it is the first company in India to have a remote monitoring solution for underground mining equipment and a hybrid ZigBee to proprietary mesh protocol solution for automated meter reading. With clients in 19 different countries, Maven monitors more than 21,000 smart meters, controls more than 10,000 street lights, and has over 1,000 devices for remote monitoring for GPS location, fuel, and various other parameters of heavy machines. Nexiot Nexiot – Accelerating Innovation, is a Bangalore based Internet of Things (IoT) & Electronics innovation focussed organization. Led by a group of experienced technocrats with expertise in electronics, sensors, wireless, robotics & information technology domains with global exposure, Nexiot is strongly positioned to provide superior solutions to customers across the globe. They provide solutions for Smart Water, Energy, Resources Management, Automation & Beacon based applications. Moreover, they offer most comprehensive Certification Programs for IoT Startups – aspiring Entrepreneurs and IoT System Design enthusiasts. The programs are designed in collaboration with Industry leading solution providers and are offered both online & taught in-classroom. Their core objective is to build competency for designing & implementing end-to-end industry standard IoT solutions. Nexiot is also launching a Global IoT Technology Center, which is a unique one-stop center to enable designers & engineers select the best technologies for implementing IoT solutions as well as a portal to support on Career for Job opportunities focussing on IoT. SenseGiz SenseGiz makes consumer and enterprise products in the wearable tech & IoT space. Their primary focus is to design and build IoT solutions for security, defence and smart city applications. Typically, their products consist of smart hardware units with a cloud backend for processing and connectivity to smartphones for user interface and control. SenseGiz’s mesh network based IoT platform is the easiest and most cost effective way to deploy a IoT solution anywhere in the world – be it for a home\/office location or a large factory or city wide installation. It’s a platform where customers can connect lifestyle, energy, remote monitoring, security, public infrastructure, and healthcare applications with intelligent devices managed by mobile apps connected to the cloud. Yuktix Yuktix is an IoT startup based out of Bangalore creating environment monitoring solutions for emerging markets. They aim to create next generation monitoring solutions focused on making data mobile and data capture, visualization and analysis in a true plug and play proposition. Customers have different data capture needs and developing solutions from scratch every time adds to the cost. Yuktix is developing need specific environment monitoring solutions. Their IoT platform consists of a sensor catalog, a device that acquires inputs from sensors and multiple network options to send data to the cloud from anywhere. It provides a ready to market catalog for weather, air quality, gases and water sensors and more through their multiple communication options such as GSM\/GPRS, Ethernet, WiFi and 6LowPAN to cover all types of needs. To create a domain specific solution, customers can pick required sensors from the catalog, plug it in the device and start receiving data on the cloud. Further, Yuktix cloud allows them to create and group multiple devices, set notifications for important events and store and visualize data. They provide API and connectors to make system integration easy. This emerging startups list gives a window into the possibilities for disruption in the IoT and analytics space. Further, it also drops hints on what trends are likely to emerge this year, though we would be covering that towards the end of this year, so keep dropping by on analytics india magazine and keep yourself updated on the buzz in the analytics community!","excerpt":"Internet of Things (IoT) seems to be the latest buzzword in the IT and analytics community. Typically, IoT goes beyond the machine-to-machine communications and covers a variety of devices and applications. Consequently, analytics is used for predictive decision-making using data generated from these M2M and IoT devices. Thus, making it imperative for IoT companies to […]","categories":["AI Features"],"tags":["Internet of things","IoT India"],"author_name":"Apoorva Verma","publish_date":"2016-02-21T06:32:32","publication_year":"2016","word_count":1673,"keywords":["Go","IoT India","AI","R","ML","Scala","Git","Internet of things","RAG","Aim","analytics","predictive analytics"],"extracted_tech_keywords":["AI","ML","analytics","Aim","RAG","predictive analytics","R","Go","Scala","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/10-emerging-iot-startups-to-watch-out-in-2016\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":6,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10017380,"title":"Ledger App Khatabook Helps SMBs To Keep Up With India’s Digital Aspirations","content":"India’s IoT market size is nothing to sneeze at. Connecting affordable devices to bring more people to the digital platform and providing them with the best digital services drives a significant impact, especially for underserved communities. Agriculture, healthcare, education, and financial services sectors can hugely benefit from data-driven insights. To help India’s SMB sector by building utility solutions, Bangalore-based fintech startup, Khatabook, is setting a new benchmark by assisting businesses to record the credit they extend to customers digitally. For this week’s startup column, Analytics India Magazine spoke to Gaurav Lahoti, VP Engineering, at Khatabook, to understand how the startup is leveraging AI and machine learning to offer online digital transaction services. Founded in 2018 by Ashish Sonone, Jaideep Poonia, Ravish Naresh, Vaibhav Kalpe and Dhanesh Kumar, Khatabook enables micro, small, and medium businesses to track transactions safely and securely. The company also offers online payment collection through UPI and QR, sends periodic reminders to creditors via messages, and generates reports. Flagship Products Khatabook is essentially a digital ledger app with advanced features such as digital payments, automated reminders to customers for credit collections, etc. The app is cut out for India’s MSME (Micro, Small, and Medium-Sized Enterprise) sector, including Kiranas, medical shops, tuition centres, gyms, distributors, wholesalers, etc. The product and tech innovations on the platform are in alignment with the digital behavior of Bharat – people who are new to the digital platform, use low-end devices and are not from tier-1 cities. Currently, the app has more than 9 Million active monthly users. The app has contributed immensely to save time for MSMEs by digitising physical workflows and improving the cash flow through automated credit reminder features. Lahoti said, during the lockdown, the company has launched many product initiatives to keep the users engaged. The team at Khatabook had been focused on financial engineering with faster digital adoption. The initiative helped shopkeepers inform customers about their shop’s status, such as timings of operations during the lockdown, safety rules to follow, etc. How Is It Different From Other Products In The Market? “The app is available in 13 languages to ensure maximum digital adoption in all the districts of India. I would say simplicity, language inclusion, and scalable tech platform are our major differentiation,” Lahoti said. Use of AI & ML @ Khatabook According to Lahoti, the company continuously uses ML models for problem-solving. The machine learning techniques incorporated in the financial risk engines help the company with user legitimacy and prevent fraud. AI and machine learning techniques have also been used for other activities, including customer outreach and after-sales. Core Tech Stack Khatabook relies heavily on cutting edge open-source technologies. The applications are built on Kotlin (Android), Swift (iOS), and React + Redux with Typescript (Web PWA). On the backend, the core services use Typescript, Node.js, and Postgres. Lahoti said, “Data is the lifeline of Khatabook teams. Our marketing, product, and operations team rely heavily on data to drive insights and future growth. As our requirements grew, we decoupled compute and storage to support parallel compute capabilities on the same data.” He added, “In the last few months, we have focused a lot on developer experience and productivity too. We are strengthening our CI\/CD & test automation abilities. We are focusing a lot on self-serve tooling and services to decouple dependencies for better developer ownership.” Hiring Phase According to Lahoti, the company has been hiring across all functions and at all levels, even during the lockdown. He said, “Besides being good at technical skills, our engineers have an excellent customer focus. We believe that every engineer should understand their work’s impact and how end-users will benefit from it.” He added, “As Khatabook is always building and experimenting with innovative products and features, the ability to collaborate with the product, design seamlessly, and tech team is a prerequisite.” Roadmap “Khatabook has a user base across more than 95% of districts in India. As a tech leader at the company, I plan to scale the team and technology stack to achieve maximum digital adoption and help deliver user-focused solutions,” said Lahoti.","excerpt":"India’s IoT market size is nothing to sneeze at. Connecting affordable devices to bring more people to the digital platform and providing them with the best digital services drives a significant impact, especially for underserved communities. Agriculture, healthcare, education, and financial services sectors can hugely benefit from data-driven insights. To help India’s SMB sector by […]","categories":["Deep Tech"],"tags":["best book to learn digital marketing","Startups"],"author_name":"Ambika Choudhury","publish_date":"2021-01-09T13:00:00","publication_year":"2021","word_count":681,"keywords":["Go","machine learning","AI","ML","Scala","TypeScript","RAG","Git","Startups","analytics","best book to learn digital marketing","R"],"extracted_tech_keywords":["AI","machine learning","ML","analytics","RAG","R","TypeScript","Go","Scala","Git"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/ledger-app-khatabook-helps-smbs-to-keep-up-with-indias-digital-aspirations\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10085202,"title":"Destined for Doom: Why Superapps Have Been Super Flops in India","content":"While superapps such as WeChat and GoJek have scripted success stories in China and Indonesia, companies in India are still trying to replicate the winning formula here. And this is despite the fact that superapp as a concept is not novel. Paytm, which was launched as a payments wallet in 2010, has become a superapp of sorts with multiple offerings on its platform. However, it has surprisingly failed to scale. Meanwhile, according to projections, Tata Group’s superapp Tata Neu is expected to meet just half its sales target this fiscal year, forcing the company to reevaluate its strategies. Now, Adani Group is set to launch its own superapp somewhere this year. However, the question that arises is, will superapps ever work in a country like India? Indian market – a hard nut to crack According to a Cisco report, India has over 800 million smartphone users, doubling from nearly 400 million in 2017. If we look at the market, it presents opportunities and makes it lucrative for superapp companies. In terms of user numbers, India is a goldmine for any app-driven business. Integrating services such as food delivery, payments, taxis, tickets and insurance, among others in a single platform and offering an all-inclusive digital experience to consumers seems like a good prospect in the Indian context. “With a huge working-age population, infiltration of high-speed internet even to rural areas, and inexpensive labour and shipping charges, it makes the perfect environment for apps to develop and succeed,” Pranav Dwivedi, founder of Netratvshaala,  told AIM. Over the years, different players like Hike Messenger, Paytm, Reliance Jio, Amazon, and Ola tried their luck in this space, but none has managed to clone the success of WeChat in China or Gojek in Indonesia. Now, Tata Neu’s misfortunes only reinforces the idea that the Indian market is a hard nut to crack. Why is Tata Neu failing? Tata Neu, despite the backing of one of the largest business conglomerates in the world, failed to take off. “Tata Neu encountered difficulty in gaining traction in the Indian market. Late entry, limited services, poor user experience, lack of marketing, and regulatory challenges are all factors that have hampered their progress,” Dwivedi said. Tata Neu’s performance has been so abysmal that Sauvik Banerjjee, founding member and CTO of Tata Neu, resigned just four months after the app’s launch. “Besides, it only provides a few services, such as e-commerce, ride-hailing, and food delivery, that are already on offer on other apps,” Dwivedi said. Space domination played a big role in Neu’s failure. For example, e-commerce is heavily hegemonized by Amazon and Flipkart. Similarly, the ride-hailing and food delivery space is controlled by Ola and Uber and Swiggy and Zomato, respectively, making the space practically watertight for Neu to break in. Further, Dwivedi believes the company did not put much into marketing and advertising, making it inconspicuous to potential customers. “Moreover, to meet the regulatory requirements, Tata Neu has to invest more money and time,” he said. Challenges for superapps The prime reason for superapps not taking off in India is the country’s late arrival to the party. There already was intense competition in the Indian market in every space, making things difficult for superapps. We already have a variety of e-commerce, payments, food delivery, and ride-hailing apps in India, so it can get difficult for a new superapp to gain a foothold in the market. “Many of these existing apps have well-established customer bases and strong brand recognition,” Dwivedi explained. In a price-sensitive market like India, the likes of Zomato, GPay and Uber have built a niche for themselves by using discount pricing strategy. In fact, many of them are using the same strategy to stay relevant in the market. However, for a superapp it might not be feasible to offer discounts for all the services they have to offer. Jack of all trades, master of none Another concern that Dwivedi points out is the wide spectrum of services that superapps usually propose. “Giving a first-class experience across all these services can be hard, as each requires dissimilar expertise and assets. For example, a significant cause of the failure of Paytm Mart is the poor quality of services and a lack of quality control.” There are also administrative, and compliance challenges to take into consideration. “For instance, rendering financial services such as mobile payments and banking demands adhering to several regulatory requirements and attaining the requisite certificates. Diverse kinds of permits and regulations are needed to deliver medicine supplies, food, and financial services.” Today, customers want hyper-personalised and hyper-localised products, and a superapp might not be able to make everyone happy. Superapps won’t work in India Kavin Bharti Mittal, founder of Hike Messenger, said since smartphone and data penetration in India is good, users have the ability to download numerous applications on their phones. Mittal, in fact, wanted to turn Hike into a WeChat-like superapp in India. However, Hike Messenger shut down in January, 2021. Mittal believes a superapp is not what customers in India are looking for, even though businesses are pushing for it. Instead, what customers really want is the best user experience. Deepinder Goyal, chief executive at Zomato, is also sceptical of the viability of the superapp model in India. “We haven’t seen superapps work in India so far,” he said, while speaking to analysts during the company’s Q4 2022 financial results. For a superapp to work in India, it must first offer hyper-personalised and hyper-localised user experience and overcome users’ reliance on already existing apps, for example, Swiggy and Zomato, for food delivery. With Adani Group set to launch its own superapp despite the scepticism and Tata Neu still in the game, it would be interesting to see if these business conglomerates are able to crack the nut.","excerpt":"Today, customers demand hyper-personalised and hyper-localised products, and a superapp may not be able to deliver","categories":["AI Features"],"tags":["Superapps"],"author_name":"Pritam Bordoloi","publish_date":"2023-01-16T18:00:00","publication_year":"2023","word_count":963,"keywords":["Replicate","Go","programming_languages:R","AI","programming_languages:Go","Git","Superapps","Aim","R"],"extracted_tech_keywords":["AI","Aim","R","Go","Git","Replicate","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/destined-for-doom-why-superapps-will-be-super-flops-in-india\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10082848,"title":"IN-SPACe signs MoU with QNu Labs for Quantum Security Solutions","content":"The Indian National Space Promotion and Authorization Centre (IN-SPACe) and Bangalore-based deep tech startup QNu Labs have signed a memorandum of understanding (MoU) to create domestic satellite QKD (quantum key distribution) products. QNu Labs develops next-generation security solutions by harnessing the power of quantum cryptography to address the cybersecurity challenges posed by the classical computing world. The company is disrupting existing security paradigms by accelerating the transition towards deploying Quantum-safe encryptions. A major obstacle to building a large-scale quantum secure network is the requirement of repeaters every 100–150 kilometres for terrestrial QKD systems. However, with Satellite QKD capability, the collaboration gives India a chance to lead the future of global quantum communication networks involving a combination of the quantum-satellite constellation, providing intercontinental connectivity. Through this MoU, QNu Labs—with the help of ISRO and IN-SPACe—hopes to demonstrate limitless distance satellite QKD-based quantum secure communication. Dr. Prafulla K Jain, Director Programme Management and Authorization, IN-SPACe, Department of Space, said, “For a country like India, with the expansion of the space sector and significant shifts in advanced technologies, especially in network and communication domains, it is very crucial to have a robust ecosystem in place so that we have smooth operations without compromising any sensitive data.” ISRO and IN-SPACe will play a crucial role in this mission by providing QNu Labs with payload designs, telemetry, and other related activities. This project will offer secure cryptographic key delivery services to customers for a variety of applications serving the industry, the government, and defence sectors, where the security and confidentiality of shared information are increasingly essential. Formed in 2020, IN-SPACe aims to open up the space sector to private players and provide them with the necessary infrastructure and assistance to enable faster progress in India’s space activities. IN-SPACe is responsible for the promotion, enabling, authorisation, and supervision of various space activities of the Non-Governmental Entities (NGEs) that include building launch vehicles and satellites, offering space-based services, sharing space infrastructure and premises with ISRO, and establishing new space infrastructure and facilities.","excerpt":"The Indian National Space Promotion and Authorization Centre (IN-SPACe) and Bangalore-based deep tech startup QNu Labs have signed a memorandum of understanding (MoU) to create domestic satellite QKD (quantum key distribution) products. QNu Labs develops next-generation security solutions by harnessing the power of quantum cryptography to address the cybersecurity challenges posed by the classical computing […]","categories":["AI News"],"tags":["IN-SPACe","ISRO","quantum cryptography"],"author_name":"Ayush Jain","publish_date":"2022-12-20T16:21:20","publication_year":"2022","word_count":335,"keywords":["Go","ISRO","programming_languages:R","AI","quantum cryptography","programming_languages:Go","IN-SPACe","Aim","ViT","R","startup"],"extracted_tech_keywords":["AI","Aim","R","Go","ViT","startup","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/in-space-signs-mou-with-qnu-labs-for-quantum-security-solutions\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":59227,"title":"Social Media’s Doomed War Against Coronavirus Misinformation","content":"As the coronavirus continues to spread, it has laid, what the World Health Organisation calls, an ‘infodemic’ among people. The more you open the newspaper or any social media site, the virus is the first topic that pops up. Initially, the news was that the virus was just the flu and not a very big concern for people. But, as offices and schools started shutting down followed by malls, theatres and clubs, we all started to wonder about the fatality of the crisis. But with all these, there was an overabundance of information available online, which made it more difficult for people to get the correct news from a trustworthy source when they most need it. This infodemic was soon joined in by social media sites, and, brought all kinds of disinformation — such as how rolling our 5G damages the immune system causing coronavirus; how coronavirus is a product of an experiment gone wrong in China; or how Rothschild family owns the patent of coronavirus to gain more money from people. The most interesting ones are where there was a false claim that Daniel Radcliffe contracted coronavirus or that it is spread by people in China eating bats, fuelling racism online. All these might sound bizarre to you; however, some dangerous conspiracy theories mislead medical advice, and some false claims stated that children are immune to coronavirus. If people follow such information, they could put their life at risk. As the spread of the coronavirus coming parallel to the spread of misinformation — under immense pressure from governments and medical experts — social media giants have taken bold strides in joining hands to tackle the misinformation spread on their platforms. Michael Kratsios, CTO of the US & Deputy Assistant to the President at the White House Office, said, “Cutting-edge technology companies and major online platforms will play a critical role in this all-hands-on-deck.” And with that in hand, Facebook, Google, Microsoft, Twitter, Tiktok, Dropbox, YouTube, LinkedIn and Reddit recently published a statement stating that they are all working closely on this issue. They even got together with WHO to discuss ways to deal with fake news about the virus. Spreading The Right Information One of the most straightforward proven strategies to stop the spread of false information is by providing the right information to people. WHO has been putting a lot of effort by partnering with Facebook and Tiktok to promote correct information on their platform. In fact, in a recent article by WHO, they busted one of the myths related to coronavirus by stating that “Contrary to what some believe, the virus can be transmitted in all climates, so the coming summer heat is not going to stop the virus. Taking a hot bath also won’t heal COVID-19, just like eating garlic and rinsing your nose with saline won’t. Washing your hands frequently, avoiding physical contact and not touching your face are still the measures advised by the health organisation.” Another interesting way the industry is spreading right information is through their chatbots, through which, companies are providing updated schedules of flights and country borders. Joining the revolution, WeChat is also offering its users with free online health consultation services. Tackling False Information Misinformation has become a severe issue for social media, with Facebook, in particular, being in the limelight in recent years over its role in spreading fake news. And, as the pandemic continues to grow, troublemakers take advantage of people’s curiosity to seek out information and monetise the opportunity with false claims. While it’s easy to promote the right information, it is equally challenging to demote false information that is already circulating on the internet. Social media companies, which were already struggling to keep up with the content moderation cases on their platforms, have been making proactive efforts to combat misinformation about coronavirus. Facebook started its deployment of artificial intelligence to flag fake news and malicious content and pass them to fact-checkers before sharing. In a recent blog post, Facebook said that “With fewer people available for human review, during this coronavirus outbreak, we’ll continue to prioritise imminent harm and increase our reliance on proactive detections in other areas.” Currently, the company is working with 56 fact-checking partners in 46 languages to rate fake news and alert users before sharing it. Amazon is also working towards taking down a million advertisements of products that are claiming to have effects on coronavirus. This includes banning advertisements selling fake medical face masks, advertising a limited supply or claim to offer complete protection against the virus. Twitter, also using machine learning and automation on their platform, however not willing to ban any accounts solely based on automated systems. They said, “sometimes posts lack the context that our teams bring, and this may result in us making mistakes”. However, YouTube, on the other hand, has ordered their full-time employees to work from home, and therefore relying entirely on their AI system. They stated in a blog that” this means the AI will start removing some content without human review, which might have some effect on those not violating the policies.” TikTok, although being termed notorious, also works with third-party fact-checkers to highlight in-app notices around hashtags related to the coronavirus. Besides, Reddit has placed a banner at the top of its home page recently, providing accurate coronavirus information to its users to stop rumours of the count. The platform has blocked one of its communities, along with a warning sign, because of the high amount of misleading information being posted on it. In China, although the social content is strictly under the cyber administration, WHO has collaborated with Tencent and Weibo to counter rumours and misinformation. Despite such efforts in hand, fake news is continuing to spread, and one of the key participants have always been these social media platforms. The algorithms that decide what users see on social media platforms typically focus on content that has the most engagement or has drawn the most attention get spread farthest. Experts believe this model is hugely responsible for the spread of false claims and misinformation online since bizarre or overwhelming content is usually good at grabbing people’s attention. The bottom line is that the battle against fake news is likely to last as long as the virus; however, the tech industry is putting its foot down on misinformation faster than before.","excerpt":"As the coronavirus continues to spread, it has laid, what the World Health Organisation calls, an ‘infodemic’ among people. The more you open the newspaper or any social media site, the virus is the first topic that pops up. Initially, the news was that the virus was just the flu and not a very big concern for […]","categories":["Deep Tech"],"tags":["artificial intelligence corona","Coronavirus","Coronavirus and AI","social media","social media AI"],"author_name":"Sejuti Das","publish_date":"2020-03-20T16:00:00","publication_year":"2020","word_count":1060,"keywords":["Coronavirus and AI","Go","artificial intelligence","machine learning","AI","chatbots","social media","automation","Coronavirus","social media AI","Aim","Rust","GAN","R","artificial intelligence corona"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","Aim","chatbots","R","Go","Rust","GAN","automation"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/social-medias-doomed-war-against-coronavirus-misinformation\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":20095,"title":"Narayana Murthy Calls AI &#8216;More Hype Than Reality&#8217;, Accuses Tech Leaders Of Self-Indulgence","content":"Yet another public personality has spoken about the pitfalls of Artificial Intelligence. This time, Infosys co-founder and celebrated corporate thought leader NR Narayana Murthy has gone on record to say that AI was “more of a hype than reality” and that “worrisome” for freshers in the Information Technology sector. Speaking at an event titled Mood Indigo at Indian Institute of Technology, Bombay (IIT-B), Murthy said, “There is this whole thing about automation and artificial intelligence. That is much more hype than the reality, at least in software services.” Admitting that the IT sector was going through a “rough patch”, Murthy criticised the astronomical salary hike that the senior management of some companies were allegedly allotting themselves. “I think that is not the way to make capitalism acceptable to the larger masses in a country that has huge poverty. If we believe in capitalism, if we believe that is the best solution for the country to move forward, then the leaders of capitalism will have to demonstrate self-restraint in apportioning to themselves part of the benefits that come out of running companies,” Murthy said. Murthy is only one in a long string of influential personalities who has sounded the alarm bells regarding automation, artificial intelligence, and its effect on employment. Recently, US’ former Secretary of State Hillary Clinton had said that millions of jobs would be on the line if automated machines and applications were implemented on a large scale. Clinton had quoted the example of driverless cars and had explained, “What are we going to do when we get driverless cars? It sounds like a great idea. And how many millions of people – truck drivers and parcel delivery people and cab drivers and even Uber drivers – what do we do with the millions of people who will no longer have a job? We are totally unprepared for that.” While many noted names such as physicist Stephen Hawking and Tesla and SpaceX’s CEO Elon Musk have already sounded alarm bells with regards to Artificial Intelligence and its potential harm, there are others who don’t concur. Facebook’s Mark Zuckerberg, Microsoft founder Bill Gates and futurist Ray Kurzweil are some of the famous names who believe that AI can actually do more good to the humans than harm. Kurzweil, who currently runs a group at Google writing automatic responses to your emails in cooperation with the Gmail team, had once famously said, “It was fire that kept us warm, cooked our food, but also burnt our houses down. Technology is always a double-edged sword.”","excerpt":"Yet another public personality has spoken about the pitfalls of Artificial Intelligence. This time, Infosys co-founder and celebrated corporate thought leader NR Narayana Murthy has gone on record to say that AI was “more of a hype than reality” and that “worrisome” for freshers in the Information Technology sector. Speaking at an event titled Mood Indigo […]","categories":["AI News"],"tags":["ai hype","AI Jobs","driverless cars","IIT Bombay","Infosys","narayana murthy"],"author_name":"Prajakta Hebbar","publish_date":"2017-12-26T08:22:42","publication_year":"2017","word_count":423,"keywords":["Go","API","IIT Bombay","artificial intelligence","Infosys","AI","programming_languages:R","AI Jobs","programming_languages:Go","driverless cars","ai hype","Ray","automation","R","narayana murthy"],"extracted_tech_keywords":["AI","artificial intelligence","Ray","R","Go","API","automation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/narayana-murthy-ai-hype-jobs-india\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10118989,"title":"Adobe Launches Firefly Image 3 Beta With Auto Stylisation, Structure Reference Capabilities","content":"Adobe officially released the beta version of the Firefly Image 3 Foundation Model during the company’s Creativity Conference on Tuesday. Firefly Image 3 is the latest addition to Adobe’s creative generative AI models, with several new features added. Most interesting of these is the ability for users to have more power over personalisation with auto stylisation capabilities as well as the ability to reference style and structure. “Structure Reference enables users to easily apply the structure of an existing image to newly generated images. You can now use an existing image as a structural reference template and generate multiple image variations with the same layout,” the company said. Similarly, the style reference feature allows users to upload style preferences and generate images based on them. https:\/\/twitter.com\/Adobe\/status\/1782711617160970385 Apart from this, Firefly boasts higher quality image generation with more variety, as the newer model relies on a new style engine: “Image outputs include new varieties of styles, colours, backgrounds, subject poses and more.” With Firefly’s launch early last year, the model was criticised for barely being on par with other generative models on the market like Midjourney and Stable Diffusion. However, with the introduction of its structure reference ability, Firefly seems to be fast catching up to its competitors. This is also compounded by the ease of access afforded to users who use Adobe for work and personal use. “Firefly has been used to generate over 7 billion images worldwide since its initial debut in March 2023. Adobe built it for direct integration into workflows Adobe customers use every day, including Adobe Photoshop, Adobe Express, Adobe Illustrator, Adobe Substance 3D and now Adobe InDesign,” the company said. Currently, the latest version of Firefly is available for beta testing in Photoshop and the Firefly web app. As is the case with all Firefly models, the company has also ensured that Content Credentials are automatically attached to every image. Adobe has also stated that the models are trained on only licensed content, including Adobe Stock, to avoid issues surrounding copyright infringement.","excerpt":"Currently, the latest version of Firefly is available for beta testing in Photoshop and the Firefly web app.","categories":["AI News"],"tags":["Adobe Firefly"],"author_name":"Donna Eva","publish_date":"2024-04-24T17:32:28","publication_year":"2024","word_count":337,"keywords":["TPU","Adobe Firefly","programming_languages:R","AI","stable diffusion","generative AI","ViT","R"],"extracted_tech_keywords":["AI","generative AI","TPU","R","stable diffusion","ViT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/adobe-launches-firefly-image-3-beta-with-auto-stylisation-structure-reference-capabilities\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":799,"title":"ISRO invites proposal for space experiments to boost up its Venus Mission","content":"The Space enthusiast across the country will get to explore their skills as ISRO has announced an opportunity for Space Based Experiments to study Venus. An Announcement of Opportunity (AO), announced by the prestigious organization invites proposal from scientists across the country. ISRO states that Venus is often described as the “twin sister” of the Earth, given the similarities in size, mass, density, gravity etc. Apparently, they also share a common origin and dating back to 4.5 billion years. With a growing interest in Venus since the early 1960s, there has been various innovations and experimentations around Venus, with the talks of ISRO even partnering with NASA to visit the planet. “Venus has been explored by flyby, orbiter, a few lander missions and atmospheric probes. In spite of great progress made in exploring Venus, there still exist gaps in our basic understanding about surface\/subsurface features and processes, super rotation of Venusian atmosphere and its evolution and interaction with solar radiation\/solar wind”, said ISRO. To address the same, they have invited proposals from interested scientists, across all institutions within India for novel space based experiments. Every candidate, whether involved in planetary exploration studies, development of space instruments or those willing to develop the experiments, can be a participant. “The payload capability of the proposed satellite is likely to be 175 kg with 500W of power. However, these values are to be tuned based on the final configuration. The proposed orbit is expected to be around 500 x 60,000 km around Venus.  This orbit is likely to be reduced gradually, over several months to a lower apoapsis”, said ISRO. The last date for receiving the proposal is 19 May 2017 and here is a detailed note on how to submit the proposal.","excerpt":"The Space enthusiast across the country will get to explore their skills as ISRO has announced an opportunity for Space Based Experiments to study Venus. An Announcement of Opportunity (AO), announced by the prestigious organization invites proposal from scientists across the country. ISRO states that Venus is often described as the “twin sister” of the […]","categories":["AI News"],"tags":[],"author_name":"Srishti Deoras","publish_date":"2017-04-25T04:16:56","publication_year":"2017","word_count":290,"keywords":["programming_languages:R","AI","innovation","ViT","GAN","R"],"extracted_tech_keywords":["AI","R","GAN","ViT","innovation","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/isro-invites-proposal-space-experiments-boost-venus-mission\/","complexity_score":2,"technical_depth":6,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10119393,"title":"OpenAI to Launch Google Search Alternative Soon","content":"OpenAI is likely to announce a new search engine soon, revealed Jimmy Apples, stating that the company is looking to host an event this month, tentatively on May 9, 2024, at 10 a.m. The insider also revealed that OpenAI has been hiring for an events team since early January to organise in-house events. “They were advertising for in house events staff and events marketing back in January, they just hired an events manager last month” said Jimmy Apples, hinting at a big deal event coming in the month of June, where OpenAI might release its next model — “what ever sam decides to call it.” Further, Jimmy Apples said that the company has been busy since last week. “I count at least 50+ new async subdomains since the 24th of April.” If the rumours are true, OpenAI’s Google Search alternative release might come ahead of Google I\/O, which is scheduled for May 14, 2024. This might also make Perplexity AI redundant. Microsoft recently banned its employees from using Perplexity AI, citing security concerns. According to a report, OpenAI  has been working on a web search product that will intensify its competition with Google. The search service is set to be partially powered by Bing. In a recent podcast with Lex Fridman, OpenAI chief, Sam Altman said that, “The intersection of LLMs plus search, I don’t think anyone has cracked the code on yet. I would love to go do that. I think that would be cool.” Moreover he said that OpenAI does not want to build another Google Search. “I find that  (Google Search) boring. I mean, if the question is if we can build a better search engine than Google or whatever, then sure, we should.” “Google shows you like 10 blue links, like 13 ads, and then 10 blue links, and that’s like one way to find information. But the thing that’s exciting to me is not that we can go build a better copy of Google Search, but that maybe there’s just a much better way to help people find, act on, and synthesise information,” said Altman. OpenAI’s decision to launch a search app follows Microsoft CEO Satya Nadella’s statement a year ago, in which he said that he would “make Google dance” by integrating OpenAI’s GPT models into Microsoft’s Bing search engine.","excerpt":"Likely to make Perplexity AI redundant.","categories":["AI News"],"tags":["Google","OpenAI"],"author_name":"Siddharth Jindal","publish_date":"2024-05-01T12:58:57","publication_year":"2024","word_count":386,"keywords":["Go","OpenAI","AI","programming_languages:R","programming_languages:Go","GPT","Google","GAN","R","llm_models:GPT"],"extracted_tech_keywords":["AI","OpenAI","R","Go","GPT","GAN","llm_models:GPT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/openai-to-launch-google-search-alternative-soon\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":59299,"title":"Can This Deep Learning-Based CT Scan Analysis Accelerate COVID-19 Detection","content":"Recently, researchers from Tel-Aviv University, University of Maryland School of Medicine, Baltimore, Mount Sinai Hospital, New York and RADLogics proposed AI-based automated CT image analysis tools for detection, quantification, and tracking of coronavirus disease. The motive behind this research is to develop deep learning-based automated CT image analysis tools and demonstrate that they can differentiate coronavirus patients from those who do not have the disease by detecting, measuring, and tracking disease progression. How the System Works The system proposed by the researchers receives thoracic CT images and flags cases suspected with COVID-19 features. For the cases which are classified as positive, the system outputs a lung abnormality localisation map and measurements. The deep learning system is comprised of several components and analyses the CT case at two distinct levels, which are: 3D analysis of the case volume for nodules and focal opacities using existing, previously developed algorithms.Newly developed 2D analysis of each slice of the case to detect and localise larger-sized diffuse opacities including ground-glass infiltrates. According to the researchers, the ground-glass infiltrates have been clinically described as representative of the coronavirus. The 3D analysis of the case volume uses commercial off-the-shelf software to provide quantitative measurements, including volumetric measurements, axial measurements (RECIST), HU values, calcification detection, etc. The next step, working in the 2D slice has a few advantages for deep learning-based algorithms in a limited data scenario. The advantages include an increase in training samples (with many slices per single case), using pre-trained networks that are common in the 2D space as well as easier annotation for segmentation purposes. Behind the System In order to detect the coronavirus related abnormalities, the researchers used a Resnet-50- 2D deep convolutional neural network architecture where the network is 50 layers deep and can classify images into 1000 categories. The network was pre-trained on more than a million images from the ImageNet database, and it was further fine-tuned by using suspected COVID-19 cases from several Chinese hospitals to solve the problem at hand. In order to overcome the limited amount of cases, the researchers employed data augmentation techniques such as image rotations, horizontal flips, and cropping to the network. Advantages of This System AI-based automated CT image analysis tools can achieve high accuracy in the detection of coronavirus positive patients as well as quantification of disease burden. For coronavirus patients, the system outputs quantitative opacity measurements and a visualisation of the larger opacities in a slice-based “heat map” or a 3D volume display.This system will allow a greater volume of patients being screened for coronavirus in a shorter period of time. Wrapping Up Researchers from around the world have been trying to utilise the capabilities of AI to help accurately detect and track the progression. With the help of the deep-learning image analysis system developed, the researchers achieved classification results for coronavirus vs Non-coronavirus cases per thoracic CT studies of 0.996 AUC (95%CI: 0.989-1.00) on datasets of Chinese control and infected patients. Read the paper here.","excerpt":"Recently, researchers from Tel-Aviv University, University of Maryland School of Medicine, Baltimore, Mount Sinai Hospital, New York and RADLogics proposed AI-based automated CT image analysis tools for detection, quantification, and tracking of coronavirus disease. The motive behind this research is to develop deep learning-based automated CT image analysis tools and demonstrate that they can differentiate […]","categories":["AI Features"],"tags":["ai model","Coronavirus","Coronavirus and AI"],"author_name":"Ambika Choudhury","publish_date":"2020-03-23T11:00:00","publication_year":"2020","word_count":494,"keywords":["Coronavirus and AI","Go","data augmentation","TPU","programming_languages:R","AI","neural network","programming_languages:Go","Coronavirus","ai model","deep learning","ResNet","R"],"extracted_tech_keywords":["AI","deep learning","neural network","TPU","R","Go","ResNet","data augmentation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/can-this-deep-learning-based-ct-scan-analysis-accelerate-covid-19-detection\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10068524,"title":"Amazon India launches 2nd edition of ML Summer School","content":"On 7 June 2022, Amazon India launched the second edition of Machine Learning (ML) Summer School, an immersive program that aims to provide students with the opportunity to learn key ML technologies from Scientists at Amazon making them industry ready for careers in science. The course, conducted over four weekends in July, will provide students with an opportunity to gain skills on key ML topics, including Supervised Learning, Deep Neural Networks, Sequential Models, Dimensionality Reduction, Unsupervised Learning and two new modules, Reinforcement Learning and Causal Inference. Register>> Participants will also have access to Amazon Research Days (ARD) conference – an engagement program held in November every year. ARD connects the scientific community at Amazon, industry leaders, and academic researchers in the field of AI around the world. Amazon will also conduct ML Challenge, its flagship ML competition in August, which is a unique opportunity for students to work on an Amazon dataset, bring in fresh ideas and build innovative solutions for a real-world problem statement. Winning teams will receive pre-placement interviews (PPIs) for ML roles at Amazon along with cash prizes, swag, and certificates. Register for this hackathon>> “Amazon ML Summer School aims to provide participating students with best-in-class training on a broad range of topics which are at the core of modern Machine Learning, from fundamentals to state-of-the-art. The tutorial sessions covering the right mix of theoretical and practical knowledge will be delivered by our ML scientists who are experts in their field. This program will be a platform to help foster ML excellence and strive toward developing applied science skills in young talent. Our aim with the ML Summer School is to equip students with necessary practical experience and prepare them for science roles ahead,” said Rajeev Rastogi, Vice President – International Machine Learning at Amazon. ML Summer School is open to students in the penultimate or final year of their Bachelor’s\/Master’s\/Integrated Masters\/PhD degree enrolled on any campus in India. Eligible students will be required to take an online assessment focused on basic ML concepts and math fundamentals on topics such as probability, statistics and linear algebra. The top 3,000 students are then enrolled in ML Summer School – who attend eight virtual classroom sessions over four weekends, each session followed by live Q&A sessions with scientists at Amazon. ML Summer School participants can also participate in Amazon ML Challenge, to practice hands-on application of skills and learnings acquired via modules to solve a real-world problem statement.","excerpt":"Programs provide students across campuses in India an opportunity to gain and apply Machine Learning skills making them industry ready for science careers.","categories":["AI News"],"tags":[],"author_name":"Sri Krishna","publish_date":"2022-06-07T17:59:18","publication_year":"2022","word_count":409,"keywords":["machine learning","programming_languages:R","AI","neural network","ML","Aim","R"],"extracted_tech_keywords":["AI","machine learning","ML","neural network","Aim","R","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/amazon-india-launches-2nd-edition-of-ml-summer-school\/","complexity_score":3,"technical_depth":7,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":68542,"title":"How Data Scientists Can Benefit From Certifications While Looking For Jobs Amid Covid-19 Crisis","content":"With data becoming the key to managing business in the Covid-19 era, there has been a massive demand for data scientists among companies. Despite this huge demand, it has always been a challenge for data scientists to land a job, as companies are only looking to hire professionals with advanced skill sets. Such skill sets are scarce among data scientists, which again creates a major talent gap in the industry. And therefore it has become imperative for analytics professionals to upskill themselves with certifications to remain relevant in the industry. Online courses have proved to be immensely beneficial for data scientists to advance their skill sets. However, certifications are the one that actually authenticates and validates those learned skills. A certificate with validation from prominent educational institutions and companies can give a competitive edge to data scientists’ resumes by recognising their achievements and showcasing their talents. In fact, completion of a course can only be demonstrated by obtaining certification, which would require them to pass an exam. Although many edtech companies like Coursera, Simplilearn, Great Learning, etc. provide several online courses – some, for free – many do not provide certifications in the end. Many of them also don’t allow enrollees to invest in training time and urge self-study, which makes them a preferred choice among working data science professionals. Some of the prominent certification programs include: Chartered Data Scientist Program by The Association of Data Scientists,Certified Analytics Professional program by INFORMS,CCA Data Analyst program by Cloudera, among others. Also Read: Data Science Is Not About Certifications But The Mindset How Certifications Can Aid Data Scientists Amid Covid-19 Although these certifications don’t put much value for experienced data scientists who have been in the field for a long time, it creates a greater impact in the resumes of individuals who are either just starting off or have been negatively impacted by the ongoing recession. Being one of the highest-paid jobs today, companies have also become extremely critical before hiring data scientists, and therefore, these certifications can act as a testimony for your skills that can get you hired. Agreeing to this, Chitrita Nath, Digital & Analytics Leader ( APAC Region) – Shell Trading & Supply said that, “When a candidate has added certificates to their resume, it gives recruiters like us a sense of confidence in hiring that person.” However, this isn’t the case with enhancing the skillsets of internal staff. Nath said, “However, when I’m looking at uplifting my internal staff, which is within the organisation, it is critical to enrol them for required online courses. It is probably a commitment that the person is giving to the manager and the organisation that he’s willing to invest the time to take the course.” For such scenarios, the certification, which is the exam, “is a cherry on top.” Although the employee is taking the course, obtaining that certification would require an extra effort. “Because for certification, the employee has to find time out of their work to prepare themselves to do that online course and pass it with a certain percentage,” said Nath. Nath further believes that for many companies, certifications are mandatory for promotions. In such cases, “although taking the course would make the employee master data science, the certifications would help them advance their career.” What is more, “onboarding employees with certifications, reduces a lot of effort for companies that would have required to upskill existing employees otherwise,” said Nath. In fact, according to an article, it has been noted that certifications can help analytics professionals earn higher salaries than others. The average salary of a data science professional with certifications in virtualisation and cloud computing can lead him to an average salary of $112,955 per year. Such numbers showcase that recruiters who are looking for professionals with advanced skills are willing to pay higher for certified professionals. Even Amarjeet Kaur, a research scientist at Tech Mahindra, feels that data scientists must opt for online courses that provide certifications. “Some courses provide certification, others don’t. Also, the key to choose the right certification course for analytics professionals.” According to Kaur, not only do these certifications provide authentication to showcase necessary skills but, at the core, these programs provide domain knowledge, which is essential for them to survive in this crisis. Even among recruiting companies who are working on different Covid-19 related projects, they would prefer hiring employees with specific certifications for specific projects. “Therefore, these certifications are not only good for their resume, but also for their personal growth.” Also Read: What Is More Beneficial To Get Data Science Jobs — Certificates Or Projects? Nobody can gain expertise in data science overnight. It, in fact, takes years to acquire those skills and achieve the proficiency required for today’s era. Shirish Gupta, Head – Data Science and Partnerships at loan2grow says, “Majority of data professionals who have worked on a bit of algorithm and tried once on Kaggle for NLP start considering themselves as a data scientist, but that’s not the reality. One cannot understand data science if they cannot understand the foundation and basic maths and statistics involved in it. And that’s where data science certifications come into the picture.” What is more, certifications can not only help data scientists or working professionals, but also freelancers, gig workers, self-employed analytics professionals and consultants enhance their portfolio and establish trust among clients. Criticality Of Choosing The Right Certification Program Although there are several certification programs and online courses that provide certifications in the market, there are some essential criteria that can help data scientists choose the right certification for themselves. According to Gupta, searching for a perfect certification program is similar to choosing an online course, and depends on multiple factors — “firstly, one should see who is the instructor and what kind of experience the instructor has in the field of data science.” Secondly, it is critical to check the curriculum as learning basics is as vital as learning advanced skills, “if the course teaches some advanced skills without touching upon the basics, I wouldn’t recommend it. Thirdly, language understanding is fundamental in this field, so, “the certification should focus more on Python, rather than SQL and SAS.” Lastly, for students to advance in their learnings, it is critical to understand how the course reviews their students – “the course must have weekly quizzes, exams and a proper certification program to analyse the progress of the enrollees.” Many of these courses require immense efforts from data professionals where they have to learn, give exams and pass it with a certain percentage, however, if professionals are looking to advance their career as well as grasp the core concepts of data science, these certifications can be the best option for them to opt for, feels Gupta. Wrapping Up Being a dynamic landscape, data science is continuously evolving, and certifications are an excellent way for data and analytics professionals to keep up their competitive advantage, where it not only allows them to develop advanced skill sets that are necessary for the current situation, but will also help them validate those skills.","excerpt":"With data becoming the key to managing business in the Covid-19 era, there has been a massive demand for data scientists among companies. Despite this huge demand, it has always been a challenge for data scientists to land a job, as companies are only looking to hire professionals with advanced skill sets. Such skill sets […]","categories":["AI Features"],"tags":["best jobs in india","Certification","covid19 data","Data analyst jobs","Data Analytics Certification","Data Science Certification","Data Scientist","Data Scientist Jobs","data scientist salary","high paying jobs in india","how does ai create jobs","Oracle Certification","trust your data"],"author_name":"Sejuti Das","publish_date":"2020-06-30T15:12:11","publication_year":"2020","word_count":1180,"keywords":["Data Science Certification","Certification","high paying jobs in india","Data Analytics Certification","R","best jobs in india","data science","Data Scientist Jobs","data scientist salary","RAG","NLP","analytics","Oracle Certification","how does ai create jobs","Go","AI","cloud computing","Data analyst jobs","covid19 data","trust your data","Python","SQL","Data Scientist"],"extracted_tech_keywords":["AI","NLP","data science","analytics","RAG","cloud computing","Python","R","SQL","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-data-scientists-can-benefit-from-certifications-while-looking-for-jobs-amid-covid-19-crisis\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10040483,"title":"Virtual Try-Ons Are In Vogue: How Do They Work?","content":"As predicted by retail consultants and industry experts at the beginning of 2020, the pandemic has permanently changed the way we shop. According to India Brand Equity Foundation (IBEF), India’s e-commerce industry is expected to surpass the US to become the second-largest e-commerce market in the world by 2034. The sector will hit a $99 billion market cap by 2024 from $30 billion in 2019, at 27% CAGR, with grocery and fashion\/apparel likely to be the key drivers of growth. One of the crucial factors leading to this growth, especially in the fashion segment, is the adoption of technology to offer virtual try-on experiences by online-first brands such as Rent It Bae, Lenskart and fashion e-commerce platforms such as Myntra and Limeroad. The try-on system has helped bring down massive return rates–the biggest challenge faced by e-commerce businesses blowing profit margins and hurting bottom lines. Early attempts Converting a traditional trial room into an online setup is, however, not new. In the US, glass companies such as Camirror, Smart Look and Ipoint Kisok were few of the first movers in try-on technology. Converse, a US-based footwear brand, introduced the technology in 2012 with The Sampler App developed for iOS. In India, the trend started around six years back with brands such as Lenskart, Myntra and Tanishq. While Myntra began working on an AR feature to help customers make the right fashion choices in 2017, personal shopping app, Voonik, acquihired TrialKart, a mobile platform offering a virtual dressing room experience, in 2015. In 2019, jewellery retailer Tanishq partnered with Milestone Brandcom, an outdoor specialist agency, to set up AR kiosks at Delhi and Bangaluru airports. A VR enabled technology called MirrAR allowed the brand’s customers to virtually try on the jewellery. Plug and play With pandemic pushing customers towards online shopping, the virtual try-on technology has helped make up for the touch and feel shopping experience. e-commerce software platforms such as Shopify, Woocommerce have made it easy for new-age brands to adopt the technology by offering a virtual AR plugin. The ease of use drove the adoption of 3D rendering and AI-led pose recognition technology. Besides glasses, shoes, apparel, jewellery and make-up categories, it has also become a norm for furniture brands such as Urban Ladder, Pepperfry. Such companies leverage advanced graphical visualisation and simulation technologies to get a 360-degree view of the products. Virtual try-on for glasses Virtual try-on glasses combine computer vision, AR and image processing. The technology can be broadly categorised into 2D image superposition, 3D glasses superimposed on 2D face images, 3D face modelling, and AR technology based on a video stream. Lenskart has been offering 3D try-on using the futuristic technology by tech company DITTO that creates personalised eyewear shopping experience since 2016. The service helps Lenskart record the face of the user from various angles, and when a user tries the frame, it offers him or her a 180° view of the glass. Overall, it creates a 3D face model of the client with accurate scale, lighting and shadow. Relationship between physical try-on and virtual try-on (Credit: MDPI Open access article) Virtual try-on for apparel and shoes Virtual try-on for apparel and shoes also uses a combination of computer vision, data science, machine learning, artificial intelligence and recommendation algorithms to provide an immersive user experience. For instance, Nike Fit, a scanning app, uses these algorithms to gauge the shape of a user’s feet to identify the right fit. Another example is a technology (3D configurator by Hapticmedia) used by US-based watch brand Baume to offer an interactive 3D demo. Additionally, the customers can also virtually try the watch using a paper wristband. VR vs AR VR-based try-on allows customers to view a human model wearing clothes in a virtual setup. In contrast, AR-based try-on allows personalised models in a real-life environment. According to MDPI, AR-and VR-based try-on creates a more positive experience for the user than the traditional e-commerce interface. AR-based try-on provides better 3D visualisation and more realistic virtual avatars, making the overall fitting experience more realistic than VR-based try-on. The users focused more on the accuracy of the model in the VR environment, whereas users focused less on the accuracy of the model in AR and more on the overall feel of the model within the real environment, the MDPI survey found. Future trends The global AR and VR market is expected to reach $161 billion by 2025 a 48.8 percent compound annual growth rate from 2020. “I am deeply convinced that augmented and virtual reality will be the primary way we work, play, and connect for the next 50 years, just as personal computers and smartphones have changed the world for the last 45 years and counting,” said  Michael Abrash, Chief Scientist, Facebook Reality Labs in a blog post. With various industries accepting AR and VR more than before, increased personalisation of digital avatars and online shopping becoming a norm, virtual try-ons are expected to future-proof India’s retail industry.","excerpt":"As predicted by retail consultants and industry experts at the beginning of 2020, the pandemic has permanently changed the way we shop.  According to India Brand Equity Foundation (IBEF), India’s e-commerce industry is expected to surpass the US to become the second-largest e-commerce market in the world by 2034. The sector will hit a $99 […]","categories":["IT Services"],"tags":["augmented reality","ecommerce"],"author_name":"Shanthi S","publish_date":"2021-05-22T10:00:00","publication_year":"2021","word_count":827,"keywords":["data science","Go","artificial intelligence","machine learning","AI","augmented reality","Git","computer vision","RAG","ecommerce","GAN","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","computer vision","data science","RAG","R","Go","Git","GAN"],"url":"https:\/\/analyticsindiamag.com\/it-services\/virtual-try-ons-are-in-vogue-how-do-they-work\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10002152,"title":"Nintendo &#038; The Pokemon Company To Capitalise On Mobile Users With New Releases","content":"Nintendo brought one of their biggest and most recognisable franchises to mobile in 2016 with Pokemon Go. Now, they seem to be capitalising on the growth and branching out to a new market altogether. The Japanese videogame giant held a conference yesterday where they addressed new additions to the Pokemon franchise, along with various other products and services such as Pokemon Home, Pokemon Sleep, and Pokemon Masters. However, this press conference bodes other things for the Pokemon Company; a new direction to bring the franchise to everyone. History Nintendo created the first Pokemon game in 1996 known as Pokemon Red. This started off a global sensation, spawning an anime, manga, spinoff games and endless merchandise along with a recognisable mascot. Today, with 7 generations of Pokemon games under their belt, Nintendo has a positive money-making franchise. The newest games in the franchise are also set to be released on June 5th in a Nintendo Direct. Nintendo wanted to bring Pokemon to mobile in an AR format, leading to the rise of Pokemon Go. The overnight smash hit garnered over 500 million downloads within the first month of its launch and quickly became a pop culture sensation. The free-to-play model of Pokemon Go fueled its meteoric growth, with Nintendo being taken aback by its success. While the app faced a lot of technical issues in the first 6 months after its launch, the game eventually stabilised and was supported by Nintendo heavily. The financial and cultural success of Pokemon Go gave Nintendo the taste of true popularity. Until then, Pokemon games were only available on Nintendo handhelds and consoles, and were almost must-buys every time. The move towards mobile gave Nintendo a taste of the true market out in the world; a strategy realised with yesterday’s conference. The New Exhibits The conference saw Nintendo double down on their mobile-first strategy for greater monetisation of the Pokemon brand. The primary difference was the differentiation of the Pokemon Sword and Shield conference and this one. Moreover, the moves made by Nintendo also show signs of bringing Go into the Pokemon fold. The first service offered by the Pokemon Company was Pokemon Home. The franchise has become segmented between the Nintendo 3DS and the Nintendo Switch, with many of the series’ fans on the former owing to its age. This is where Pokemon Home comes in. There is an already existing service known as Pokemon Bank which allows users to store Pokemon in the cloud across 3DS games. Now, with Pokemon Home, users can store Pokemon across the Switch, Pokemon Go, and even the upcoming Pokemon games. The company also stated “They can trade with friends or with anyone around the world directly using Pokémon HOME via a mobile device. A potential future addition will even allow multiple players in a single location to trade all at once.” This shows a strong focus towards mobile, an attitude only followed up with the next announcement. Pokemon Sleep is another mobile application that allows players to “wake up with Pokemon every morning.” Not much information was given regarding the new game, except that it was a new way to interact with the Pokemon caught in Pokemon Go. The game is set to launch alongside a new device that is based on the previously released Pokemon GO Plus. This will track the sleeping patterns of users and transmit the data to the phone with Bluetooth. Reportedly, it can also be used to play Pokemon Go during the day and Pokemon Sleep at night. The game is set for launch in 2020. The company is also developing a new game for mobile devices known as Pokemon Masters. This is the latest addition in their mobile lineup and represents a completely new attitude of Nintendo towards mobile. Created by The Pokemon Company and DeNA, the experience is reportedly a “new type of Pokemon battling on the go”. The company also mentioned that it would feature many famous Trainers from the Pokemon video games, and features a game design fit for casual pick up and play sessions. The game is set to release in 2019, with more details coming next month. This marks 2 games in active development for mobile, with another known as ‘Pokemon Quest’ being released recently. Pokemon Go is also being updated with new content almost every month and has a sizeable player base. Setting The Path For Global Adoption Nintendo realised with the quick adoption of Pokemon Go that mobile was a market that they could target easily. The prevalence of the platform allowed them to reach markets that they previously couldn’t. This presented new money-making opportunities for the company. The move towards mobile seems to be something that the giant is taking seriously, owing to the financial and brand benefits from the move. It is left to see what the future holds for this division.","excerpt":"Nintendo brought one of their biggest and most recognisable franchises to mobile in 2016 with Pokemon Go. Now, they seem to be capitalising on the growth and branching out to a new market altogether. The Japanese videogame giant held a conference yesterday where they addressed new additions to the Pokemon franchise, along with various other […]","categories":["AI Features"],"tags":["ar","Gaming"],"author_name":"Anirudh VK","publish_date":"2019-05-30T15:09:38","publication_year":"2019","word_count":810,"keywords":["Go","API","ELT","programming_languages:R","AI","Gaming","programming_languages:Go","ar","R"],"extracted_tech_keywords":["AI","R","Go","API","ELT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/nintendo-the-pokemon-company-to-capitalise-on-mobile-users-with-new-releases\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":28286,"title":"How Did Artificial Intelligence Proof Of Concepts Turn Into Such Huge Roadblocks?","content":"Proof of concepts are pegged as the training wheels, moving AI from PoC stage to production. Last week, ZDNet broke a story about Australia’s leading bank, ANZ Bank shelving its artificial intelligence proof of concept instead of pushing it into production, citing health checks, bias and lack of explanation as some of the big reasons for putting a project on hold. Jason Humphrey, Head of Retail Risk at ANZ Bank was cited as saying how one of the biggest dangers in deep learning is because it is bringing in new attributes and new correlations which could create bias. However, building enterprise-scale AI is not easy and there are several obstacles teams face before putting the models in production. PoCs are shelved when they do not add any business value. Some of the top reasons for failed PoCs are: PoC range from vendors failing to prove the concept as originally conceived Concept not delivering the expected outcome in terms of value POC fails to satisfy the intended stakeholders POC results do not add tangible value A 2017 study from Accenture found that AI can increase profitability by 38 percent, generating over $14 trillion dollars of economic impact in the coming decades. But even though service providers with a formalised PoC spout the benefits of AI applications, they are yet to show tangible business results. One of the biggest challenges for companies in the early stage of AI journey is having sufficient skills in-house. Most service providers run self-directed primers on ML and deep learning, designed to help build developers’ understanding of how to map the business problems. AI workflows can get technical and enterprises will have to navigate a series of issues related to hardware, software, data security and a quantity of new data for training or inference Most small companies prefer building, buying or re-using the hardware and software, and\/or whether to make use of cloud services Besides infrastructure and deployment issues, model construction is the core AI task. It involves data scientists using training data and managing parameters to conduct iterative test runs. Through this approach, data scientists and tech teams can check models for initial accuracy before dispatchings them for broader training and tuning. Most AI practitioners cite how training and tuning AI models are the most computationally-intensive parts of the AI workflow. As part of this process, data scientists determine under what parameters their models converge most efficiently given the available training data, while dealing with traditional IT concerns of job scheduling and infrastructure management. This is also one of the most computationally-intensive tasks with data scientists spending a considerable amount of time wrangling data. Why AI PoCs fail The AI proof of concept roadmap poses several challenges to tech teams. From bias against AI-enabled applications or products to the criteria around explainability, enterprises have to perform a series of health checks before putting the models in production. Besides, organisations have to navigate legal issues as well arising from applied inequality or consequential decision making from AI applications. Inequalities in AI applications refer to a set of problems involving the design and deployment of AI. Bias: For example, the ANZ PoC failed to explain how the AI system does not have a biased view based on shortcomings of the training data, model, or objective function. Fairness: If decisions are made based on an AI system, it is difficult to verify whether those decisions were made fairly. This also refers to the black box problem of AI. How can discriminatory bias be minimised in training data and how can the output be judged fair? Causality: Can the model provide not only correct inferences, but also some explanation for the underlying phenomena. According to a research paper by UC Davis, deep learning as a technique is well-known for establishing correlation but is not the best technique for articulating causal mechanism. Transparency: When it comes to transparency, what are the distinct factors to be explored in understanding the decision making process? Should AI-based insights be explained in terms the user can understand and how can the decisions or outcome be questioned? Safety: It is of utmost importance to build safe, transparent and accountable AI systems. This will enable users to gain a certain level of belief and confidence in AI systems, so that they can understand the AI decision-making process. For example, policymakers need to set a means of verifying safety standards for driverless cars, drone delivery.","excerpt":"Proof of concepts are pegged as the training wheels, moving AI from PoC stage to production. Last week, ZDNet broke a story about Australia’s leading bank, ANZ Bank shelving its artificial intelligence proof of concept instead of pushing it into production, citing health checks, bias and lack of explanation as some of the big […]","categories":["AI Features"],"tags":["AI (Artificial Intelligence)"],"author_name":"Richa Bhatia","publish_date":"2018-09-13T08:06:17","publication_year":"2018","word_count":737,"keywords":["artificial intelligence","TPU","programming_languages:R","AI","ML","deep learning","GAN","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","ML","deep learning","TPU","R","GAN","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-did-artificial-intelligence-proof-of-concepts-turn-into-such-huge-roadblocks\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10007771,"title":"Why Altair Acquired Analytics Firm Ellexus","content":"Altair has stated that it acquired Ellexus– an input\/output (I\/O) profiling firm, which helps companies understand the way they access data for various applications. The acquisition of Ellexus is said to provide Altair with more storage functionality for high-performing computer solutions across industries. “Altair proceeds to extend its reach and capacities for HPC environments to support critical modern workloads including for data analytics, AI and advanced driver-assistance systems (ADAS),” said James Scapa, Altair’s chief executive officer and founder. “The acquisition of Ellexus is especially relevant in these domains as storage-aware scheduling for big data applications is crucial.” Altair will also deepen its technical skillset as Dr Rosemary Francis, Ellexus founder and the chief executive officer will join its team to help Altair best utilise the technology. “There is no better place for Ellexus’ solutions to get into the hands of global customers who can immediately benefit,” said Francis. “Altair’s expanding leadership in HPC is exciting and precisely where I want to be able to help advance the business and expand the technology to its limits.” Why Is I\/O Analysis Important For High-Performance Computing? A large expenditure in the cloud can be wasted on I\/O intensive applications, particularly when it comes to storage needs. According to experts, it’s very easy to over-provision your storage needs in the cloud and that’s where a lot of the costs can add up. Understanding your needs and making sure that you’re not over-provisioning or under-provisioning is really the key. I\/O profiling is used to help solve performance issues, both in systems and applications. It is also used to prepare and protect shared storage from applications that might be overloading shared resources. It’s also used to look into other issues such as managing application dependencies and making sure which files have been accessed, where they’re being stored and so on. More recently within the HPC community, I\/O profiling is used heavily to manage hybrid cloud resources. Detecting dependencies is another thing that you want to do with I\/O profiling. Here, companies want to analyse the file path, the package that was used to install a file, the file name, the file type and the mount point. In fact, dependencies within some industries are extremely crucial, like EDA and HPC. Customers run products such as breeze-as-a-service where users submit their job to a particular queue, and they get back an automated list of dependencies in order to check the correctness of their application for containerisation, migration and so on. Expanding HPC Capabilities Ellexus products — Mistral and Breeze — are utilised for I\/O diagnostics, optimisation and dependency analysis by HPC administrators of big companies. The power of the cloud can only be utilised if companies have good visibility into applications, both on-prem and in the cloud. When enterprises deploy HPC and other scientific applications to the cloud, they actively require the capacity to understand I\/O patterns, resource use and system performance to tune and maintain data workloads seamlessly. This will help optimise throughput and performance of apps, containers, and services to support clients and help manage thousands of applications, and run billions of tasks each day for real-time insights in HPC services. HPC is, in fact, emerging as an important element of AI transformation as it plays a crucial role in all fields of computational science and data analytics, particularly in the enterprise space. With the transition to hybrid cloud and multi-cloud, many hardware vendors are innovating software to advance in the changing landscape. Here, system monitoring is growing to become more significant as the complexity of the HPC platform augments. System monitoring simplifies the migration of enterprise HPC workloads to the cloud while reducing costs as it provides real-time insights into workloads, and spending with complete visibility to HPC cloud resources. Altair’s Strategy Post the acquisition; it has been reported that these products will be combined into the storage aware scheduling functionality of Altair PBS Works. This will help the company improve its HPC prowess, thus solving much of the dependencies issues of its clients. With analytics, software product development and HPC services, Altair is working alongside companies across large industry segments, all the way from oil and gas to finance giants. The addition of Ellexus will bring in Altair the capability of implementing its best practices in terms of I\/O performance in HPC. Altair has been one of the leading providers of HPC workload and workflow management tools for sectors like manufacturing, research, oil and gas, weather, and government. It acquired another company, Univa, which provides enterprise-grade workload management, scheduling, and optimisation products for HPC and AI\/ML workloads both on-prem and in the cloud. The acquisition of Univa and now Ellexus helps Altair position itself as a leading scheduling and resource optimisation solution provider for both massively parallel and high-throughput, single-core tasks. Acquiring Ellexus also complements Altair’s scheduling capabilities and network I\/O real-time monitoring, ensuring quicker job execution and greater resource utilisation. These acquisitions will further improve Altair’s capability and performance needs for its clients and strengthen the company’s position in workload management and cloud enablement for high-performance computing.","excerpt":"Altair has stated that it acquired Ellexus– an input\/output (I\/O) profiling firm, which helps companies understand the way they access data for various applications. The acquisition of Ellexus is said to provide Altair with more storage functionality for high-performing computer solutions across industries. “Altair proceeds to extend its reach and capacities for HPC environments to […]","categories":["Deep Tech"],"tags":["Altair","fields of analytics","HPC","HPC Data Management Software","hpc data management system","managing hpc data"],"author_name":"Vishal Chawla","publish_date":"2020-09-21T14:00:50","publication_year":"2020","word_count":844,"keywords":["big data","Go","hpc data management system","TPU","programming_languages:R","fields of analytics","AI","ML","HPC Data Management Software","Altair","RAG","AI transformation","analytics","managing hpc data","R","HPC"],"extracted_tech_keywords":["AI","ML","analytics","RAG","TPU","R","Go","big data","AI transformation","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/why-altair-acquired-analytics-firm-ellexus\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":21656,"title":"Meet SpotDraft, An AI Based Contract Management Platform That Manages Cumbersome Paperwork","content":"Team SpotDraft SpotDraft, an AI powered contract management platform is helping businesses make their contracts “come to life”. It is easing the pain of managing paperwork by offering revolutionary tools powered by artificial intelligence. Through SpotDraft’s intuitive AI powered platform, the customers can draft and sign contracts, send automated reminders and receive payments. Analytics India Magazine spoke to the team at SpotDraft to understand the idea behind its inception, how it works, their roadmap for AI and more. Started by Shashank Bijapur, a Harvard Law School alumni and Wall Street Lawyer, and Madhav Bhagat, an alumni of Carnegie Mellon, an ex-Googler specialising in various AI applications, the idea of founding SpotDraft took a shape when they met at a party while still working in New York. As the co-founders admit, over the course of conversation it was clear that both of them wanted to do something that involved technology and had the ability to reduce burden on businesses. “Because contracting is one of the most frequent tasks undertaken by professionals, and also the most cumbersome, it seemed obvious to use disruptive capabilities in AI driven technologies such as Natural Language Processing and Machine Learning to target this area” said Shashank, who has dealt with a lot of contractual work for various parties in various sectors during his time as a transactional lawyer. As he shares, while his work was intellectually rewarding, the amount of time that was spent on copy pasting old contracts into new contracts always bothered him. Whereas for Madhav, who came from a business family, saw his father ponder over contracts that shouldn’t have taken so much of his time or resources. To make matters worse, even hiring lawyers didn’t get the job done right. …This brought SpotDraft to life Explaining the way SpotDraft works, Shashank said that it enables their users to create contracts, using the templates on their platform, answer a handful of questions and execute it in a matter of seconds. “Thereafter, the contract management system is responsible for extracting key pieces of information and comparing the user’s contract to other contracts in its database. Once the contract is created, our algorithms automatically create invoices, payment reminders and track payments that have become due”, he added. The AI platform can also analyse legal documents to give the users the good-bad-ugly in the contracts, so that the users know what clauses to negotiate on. He further explained: “Take for example, a small company, who simultaneously, executes employment agreements, supply agreements, services agreements, development agreements, etc., as part of its business operations. On SpotDraft, this company can create these contracts and, in parallel, track the obligations and responsibilities, like payments, expiry dates, invoices, etc., without any human involvement. Thus, not just saving a lot of time but also a substantial sum of money.” How does it work? Shashank explains that, SpotDraft consists of three basic elements: (a) Contract Creation– Drafting a contract from a template; (b) Contract Analysis– Analysing the contents of the contract to help you simplify it and understand the good, bad and ugly; and (c) Contract Management– Understanding the business aspects of a legal document and putting it into action. In the current scenario, these three elements usually are accompanied with high cost element, time and error. Contract drafting is typically done by an expensive lawyer or DIY with a draft template ripped off the internet, making the matter even worse. SpotDraft plugs the legal cost and give businesses a good solid template through which users can draft and negotiate a contract all by themselves and in one place. “In the event they have an existing contract template, our proprietary AI parses through the contract to point out missing important clauses, summarises the key terms, classifies the good, bad and ugly clauses to give a better understanding and negotiation power to the businesses”, said Shashank. “Lastly, our AI creates a smart inbox of all the contracts and notifies the user on the important actions such as regulatory filings, upcoming deadlines, etc., that one might need to take. The inbox also connects your contract to external data sources to notify you of breaches or changes in law”, he added. With this SpotDraft connects these different parts of the same legal document into one simple workflow with zero human intervention. This saves legal costs as well as drastically reduces the amount of time it would have taken to create the same document through a lawyer. Managing contracts with AI As mentioned earlier, SpotDraft is run by their proprietary artificial intelligence and machine learning programs that have been trained with millions of contracts across different fields and different jurisdictions. Shashank shared that through the data that SpotDraft gathers from its database, it understands the DNA of the contracts. “This helps SpotDraft to manage contracts better and make it easier for our users to understand their rights & obligations in their contracts. This is all made possible without businesses having to spend large amount of money or time engaging a lawyer or dealing with issues pertaining to signing contracts which are not legally sound”, he said. “We are currently running the contract management service in beta and the results and reviews we are getting from our customers is extremely promising. Since, the core concept of AI revolves around learning in the job, we expect our system to get better once more number of users start using our services”, he added. Growth story and the roadmap ahead As an 8-month old company, they have successfully executed over 6,500 contracts with a GMV of USD 750,000 on their platform. This is a testimony to the fact that a large number of people and enterprises are showing interest in their offerings. As for the roadmap, Shashank shares that their AI is already being trained on millions of contracts from different jurisdictions and different sectors, and as they add newer customers on their platform, the team hopes that their AI will only become more proficient at analysing different kinds of clauses and different kinds of contracts. “Over time, we also expect our AI to learn and adapt to the personal choices of our users. Eventually, our AI systems will be able to recommend additions and deletions to a user’s contract depending on how he positioned himself\/ herself is his\/ her previous contracts”, he said. The startup is also coming up with Inbox services, which they are testing with limited number of users. Eying to reach the $80 billion global contract automation market The startup though fairly young, has high ambitions for itself. “We plan to use a relatively inexpensive pricing model to acquire customers from the small and medium scale enterprises and eventually implement vertical specific features (for example the VC industry, real estate industry). We want to develop and perfect our systems for the needs and requirements of the Indian market first’, said Shashank. Once they have managed to gain a strong foothold in the Indian market, they plan to create country specific localisations by expanding to different jurisdictions. “We plan to target the Southeast Asian markets and other common law countries before moving towards North America and the other English speaking European markets”, he said. The company had raised a seed funding round of a little over half a million dollars (USD 550,000), about two months ago. With Hunch Ventures, Spiral Ventures, 500 Startups, Singapore Angel Network and Satyen Kothari as investors, SpotDraft plans to utilise the funds to acquire talent and scaling the team and technology. Overcoming the challenges of being a startup in AI space The founders believe that starting from a mathematical concept, AI has evolved into a technology that has become absolutely essential in all offices. With large amounts of data being created, AI is expected to be multi functional across multiple sectors, and one of the biggest challenges that the team faces being a startup in the AI space, is access to data. Shashank said “It is much easier for larger enterprises to gain access to data than it is for startups”. Further, the data that is open-source is not always the exact match to the data required for running our algorithms. Acquiring the right kind of talent is another challenge that they face. “In India, while there is no dearth of people interested in working in the AI space, there is legitimate crunch in the number of people who are trained and qualified to work in the highly technical AI spheres”, he said in the concluding.","excerpt":"SpotDraft, an AI powered contract management platform is helping businesses make their contracts “come to life”. It is easing the pain of managing paperwork by offering revolutionary tools powered by artificial intelligence. Through SpotDraft’s intuitive AI powered platform, the customers can draft and sign contracts, send automated reminders and receive payments. Analytics India Magazine spoke […]","categories":["Deep Tech"],"tags":[],"author_name":"Srishti Deoras","publish_date":"2018-02-13T06:54:35","publication_year":"2018","word_count":1413,"keywords":["Go","funding","machine learning","artificial intelligence","AI","Git","automation","analytics","R","startup"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","analytics","R","Go","Git","automation","startup","funding"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/meet-spotdraft-ai-based-contract-management-platform-manages-cumbersome-paperwork\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10049122,"title":"Trend Micro Launches Cloud One Data Center In India","content":"Cloud Security platform Trend Micro recently launched its Cloud One regional data centre in India to uphold data sovereignty and safeguard data privacy in the country. Cloud One is Trend Micro’s flagship cloud security services platform for protecting servers, resources, and applications in the cloud. Trend Micro takes the lead by becoming one of the first security vendors to have the broadest set of services around cloud security, delivered out of India. This move follows Trend Micro’s launch of its regional XDR data lake in India last year. “As a customer-focused organization, we help our customers to adopt cloud security services and give them the advantage to move to the cloud without worrying about data security and sovereignty. India is a significant market for us, and to reinforce our commitment towards our customers in India, we have launched Trend Micro Cloud One data centre for the India region. It helps simplify cloud security for our vast clientele – especially in BFSI, government and other regulated sectors”, said Vijendra Katiyar, Country Manager, India & SAARC, Trend Micro. He further added, saying, “Cloud One addresses the most fundamental cloud security requirements of the customers, protecting a number of the most used cloud services, such as compute, file storage, containers, serverless functions, virtual private network, and more. Cloud One also enables customers to gain full visibility into their hybrid or multi-cloud environment, and ensure that the cloud services are configured correctly and are compliant to industry standards”. Trend Micro’s cybersecurity platform protects hundreds of thousands of organizations and millions of individuals across clouds, networks, devices, and endpoints. The platform delivers a powerful range of advanced threat defence techniques optimized for environments like AWS, Microsoft, and Google, and central visibility for better, faster detection and response. Many customers face regulatory or policy-based concerns around the location of SaaS platforms or data storage for their workloads. Having a data centre within the country improves compliance and reduces friction for guidance and reporting requirements. The all-in-one platform approach provides automated protection, protecting the customers at every step of the way on their cloud journey.","excerpt":"Trend Micro takes the lead by becoming one of the first security vendors to have the broadest set of services around cloud security, delivered out of India.","categories":["AI News"],"tags":["AWS cloud","cloud security","data centre india","Google","network firewall","SaaS-based model"],"author_name":"Victor Dey","publish_date":"2021-09-21T13:41:49","publication_year":"2021","word_count":348,"keywords":["data centre india","network firewall","Go","AWS","SaaS-based model","AI","cloud_platforms:AWS","programming_languages:R","cloud security","serverless","RAG","Google","GAN","AWS cloud","R","data lake"],"extracted_tech_keywords":["AI","RAG","AWS","serverless","R","Go","data lake","GAN","cloud_platforms:AWS","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/trend-micro-launches-cloud-one-data-center-in-india\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10012573,"title":"Guide To Association Rule Mining From Scratch","content":"Are you curious about how the Apriori algorithm & Association Rule works? and How Permutation & Combination are useful in mining rule. Then you are in the right place, let me walk you through it. In this article we will discuss: Note: Author expects that readers understand basic probability, Permutation, Combination, Numpy, Pandas module. Many business enterprises accumulate huge amounts of data from their daily operation. For example, a large number of customer purchase details are collected daily in grocery stores. Such data, commonly known as Market Basket Transaction shown in Figure 1. Each row i:e each customer transaction at a time labelled as Transaction. Figure 1 illustrate an example, Transaction 1 contains {Bread}, transaction 2 contain {Scandinavian}, transaction 3 contain {Hot chocolate, Jam, Cookies} etc. Figure 1: Market Basket Transaction Market basket data should be converted to Binary format as shown in Figure 2. Where each row corresponds to a transaction and each column corresponds to an item. Items can be treated as binary variables, holding value one if the item is present in a transaction else zero. Figure 2: Binary Representation of Market Basket data Itemset and Support Count Let I = {Bread, Scandinavian, Cake, Muffin, Coffee, Pastry} be the set of all items in a market basket data and T = Total No of transactions be the set of all transactions. An itemset may contain single or more than one item like {Cake}, {Bread}, {Bread, Cake}, {Bread, Coffee}. An important property of an itemset is its support count, which refers to the number of transactions that contain itemset. Mathematically, the support {Bread, Coffee} for an itemset {Bread, Coffee} can be stated as follows: SupportBread, Coffee= No of transaction contain Bread and CoffeeTotal no of Transaction(T) Frequent Itemset Generation: Objective is to find the itemsets that satisfy minimum threshold support, these itemsets are called frequent itemsets Otherwise called Infrequent Itemsets. Don’t get confused if I use this term. Figure 3 shows all possible itemsets that can be generated by an itemset I = {a, b, c, d, e}. In general, a dataset that contains k items can potentially generate up to 2^K itemset. Because K can be very large in many practical applications, it becomes computationally expensive. To overcome this problem, we can prune the unwanted itemsets as follows, As illustrated in Figure 3. when an itemset {a}, {b} is infrequent, then all of its supersets (i:e the non-shaded itemset in this figure) must be infrequent too. Suppose {c, d, e} is a frequent itemset, then all its subsets itemset (i:e the shaded itemset in this figure) must also be frequent. Figure 3: Frequent itemsets Apriori Algorithm: Let me show you how Apriori algorithms work and generate frequent itemsets based on a concept that a subset of a frequent itemset must also be a frequent itemset. Below is the code to generate frequent itemsets with an example from our dataset. STEP 1: Create dictionary “support” to stored itemsets and support. Store all items in set “L” available in dataset STEP 2: Iteration are as follows: Iteration 1: Initially the algorithm lists all the items and computes support for frequent itemsets with length 1 and stores it in a dictionary, dictionary helps to retrieve values as support using keys as items, here few of them shown in Figure 4- Table L1. Here we assume 4% as minimum threshold support. Figure 4: Iteration 1 As you can see here, items ‘Scandinavian’ and ‘Muffin’ are infrequent. So, we are going to discard {‘Scandinavian’, ‘Muffin’} in the upcoming iterations. We have the final Table F1 as a Frequent Itemset. Iteration 2: Next, we will create a combination of 2 itemset and calculate their support. All the combinations of itemset shown in Figure 4-Table F1 are used in this iteration. Figure 5: Iteration 2 Infrequent Itemsets are eliminated again. In this case {(‘Bread’, ‘Cake’), (‘Bread’, ‘Pastry’), (‘Cake’, ‘Pastry’)} shown in Figure 5-Table L2. Now, let us understand what is pruning and how it makes Apriori one of the best algorithms for finding frequent itemsets. Pruning: Here we will divide the itemsets in Figure 6-Table L3 into subsets and discard the subsets that are having a support less than minimum threshold support. Iteration 3: Here all the itemset are infrequent since its subset {(‘Bread’, ‘Cake’), (‘Bread’, ‘Pastry’), (‘Cake’, ‘Pastry’)} already discard in Figure 5-Table. L2, So it Superset will also be infrequent. This is the main highlight of the Apriori Algorithm. Figure 6: Iteration 3 Since all the itemsets in Iteration 3 are infrequent we will stop here. STEP 3: Calculate support for each jth itemsets. if sup meets the minimum support threshold then add to “support” dictionary. Update “L” for each jth itemsets take union of itemset in list L and itemset(j) L = list(set(L) | set(j)) Let me show you the actual frequent itemset obtained by appriori algorithm. Figure 7: Frequent Itemset We are almost done as we already obtained frequent itemset, which generally take more computational time. Next, we are going to see Association Rule. Association Rule: This section describes how to extract association rules efficiently from the above obtained frequent itemset. An association can be obtained by partitioning the frequent itemsets {Bread, Coffee} into two non-empty subsets, 1) Bread => Coffee, simple way to understand “If Bread then coffee”, 2) Coffee => Bread, “If Coffee then Bread”. The subset meets minimum threshold confidence and Positive lift can be called as strong rule. Curious what is Confidence and Lift? Let me tell you, there are several measures available to analyse a rule, in this article we will discuss confidence and lift. Confidence: Objective is to extract all the high-confidence rules from the frequent itemset found in the previous step. It says that Consequent (Coffee) is brought as an effect Antecedent (Bread). This rule is called a strong rule. ConfidenceBread=> Coffee=Support(Bread and Coffee)Support(Bread) But confidence measures have one drawback, it might misrepresent the importance of an association. Because it only tells how popular Antecedent (Bread) are, but not Consequent (Coffee). If coffee is very popular then it is more likely that a transaction containing Bread will also contain coffee, thus inflating confidence measures. To overcome this drawback, we use a third measure called lift. Lift: Lift computes the ratio between the rule’s confidence and the support of the itemset in the rule consequent. It represents how likely a consequent (Coffee) is purchased when an antecedent (Bread) is already purchased, while controlling for how popular consequent (Coffee) is. When the lift of (Bread => Coffee) is 1, which means no association between items. A lift value greater than 1 implies coffee is likely to be brought if bread is already brought, while a value less than 1 implies coffee is unlikely to be brought if bread is already brought. LiftBread=> Coffee=Support(Bread and Coffee)SupportBread*Support(Coffee) Or LiftBread=> Coffee=Confidence(Bread=>Coffee)Support(Coffee) Association Algorithm: Now we will apply association rules to the frequent itemset obtained in the previous algorithm. We will assume minimum threshold confidence 50%. STEP 1: List all frequent itemset and its support to dictionary “support”. Create list “data” to stored results. List all frequent items set to List “L”. STEP 2: Initially the algorithms will generate rules using Permutation of size 2 of frequent itemset and calculate Confidence and Lift shown is Figure 8. Figure 8: Strong Rule Why Permutation over Combination to calculate rules? Let me answer this with a simple example. For support, the number of transactions containing Cake & Coffee is the same as the number of transactions containing Coffee and Cake, Order does not matter. But this is not the case in a rule, let me explain. Confidence (Cake => Coffee) = Support (Cake & Coffee) \/ Support (Cake) = 0.52 Confidence (Coffee => Cake) = Support (Cake & Coffee) \/ Support (Coffee) = 0.11 Here, Order does matter; thus, we will select those rules which meet minimum threshold confidence. Permutation help fixed position of Antecedent on left side, confuse let me explain with a simple example, let us have Permutation of Cake, Coffee and (Coffee, Cake) to calculate rule as below. Rule 1: {Cake, (Cake, Coffee)} LHS(Cake) is fixed for antecedent and it must be a subset of RHS (Cake, Coffee) that gives Confidence (Cake => Coffee). Rule 2: {Coffee, (Cake, Coffee)} LHS(Coffee) is fixed for antecedent and it must be a subset of the right side (Cake, Coffee) that gives Confidence (Coffee => Cake). Till here it’s all good, but we will have more rules like {(Cake, Coffee), Cake} and {(Cake, Coffee), Coffee}, here antecedent is (Cake, Coffee) which seems incorrect thus subset condition is used to eliminate such rule. Calculate Confidence and other rules for each ith itemsets, if Confidence meets minimum Confidence threshold then add to “Data”. Below Figure 9 shows the Strong Rule obtained from experimental datasets. Do not forget that Rule is only applied on Frequent Itemset. Figure 9: My Rule Finally, we will cross verify our results with the Standard Package available in Python named mlextend.frequent_patterns having Apriori and association rules modules. Apriori results: Association Rule results: From the above comparison we can conclude that our results matched with standard packages and proposed objectives had been served. Feel free to download scratch codes available on my GitHub link https:\/\/github.com\/Roh1702\/Association-Mining-Rule-from-Scratch Why is Lift a better measure overConfidence? This we will see in detail in another article I will be publishing soon. References: Introduction to Data Mining by Pang-Ning Tan, Michael Steinbach and Vipin Kumar https:\/\/github.com\/viktree\/curly-octo-chainsaw\/blob\/master\/BreadBasket_DMS.csv","excerpt":"Are you curious about how the Apriori algorithm & Association Rule works? and How Permutation & Combination are useful in mining rule. Then you are in the right place, let me walk you through it. In this article we will discuss: Note: Author expects that readers understand basic probability, Permutation, Combination, Numpy, Pandas module. Many […]","categories":["Deep Tech"],"tags":["Applications of Data Mining","data mining applications","subsets of ai"],"author_name":"Rohit Krishnarao Umredkar","publish_date":"2020-11-30T11:00:00","publication_year":"2020","word_count":1569,"keywords":["Go","NumPy","AI","ML","Git","Applications of Data Mining","subsets of ai","Python","data mining applications","programming_languages:Python","GitHub","R","Pandas"],"extracted_tech_keywords":["AI","ML","Pandas","NumPy","Python","R","Go","Git","GitHub","programming_languages:Python"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/guide-to-association-rule-mining-from-scratch\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10021407,"title":"PyTorch Geometric Temporal: What Is it &#038; Your InDepth Guide","content":"PyTorch Geometric Temporal is a temporal extension of PyTorch Geometric(PyG) framework, which we have covered in our previous article. This open-source python library’s central idea is more or less the same as Pytorch Geometric but with temporal data. Like PyG, PyTorch Geometric temporal is also licensed under MIT. It contains many dynamic and temporal state-of-the-art deep learning algorithms or Geometric Deep learning algorithms for spatio-temporal signals. It also provides an easy-to-use mini-batch loader\/dataloader, multi GPU-support, benchmark datasets and an iterator for dynamic and temporal graphs. The keys points of the framework are : On both dynamic and static graphs, it offers discrete-time graph neural networks.  When the time period is continuous, it allows for spatio-temporal learning without the use of discrete snapshots. Methods included in PyTorch Geometric Temporal. Discrete Recurrent Graph ConvolutionsTemporal Graph ConvolutionsAuxiliary Graph Convolutions You can check all the algorithms supported by it, here. Without further ado, let’s begin with the code part! Requirements & Installation Install all the requirements of PyTorch Geometric Temporal and then install it via PyPI. PyTorch >= 1.4.0 For checking the version of PyTorch, run the mentioned code: !python -c \"import torch; print(torch.__version__)\" Check the version of CUDA installed with PyTorch. !python -c \"import torch; print(torch.version.cuda)\" Install the dependencies : The code is for PyTorch version = 1.7.0 and replaces ${CUDA} with the CUDA version which you are using. %%bash pip install torch-scatter==latest+${CUDA} -f https:\/\/pytorch-geometric.com\/whl\/torch-1.7.0.html pip install torch-sparse==latest+${CUDA} -f https:\/\/pytorch-geometric.com\/whl\/torch-1.7.0.html pip install torch-cluster==latest+${CUDA} -f https:\/\/pytorch-geometric.com\/whl\/torch-1.7.0.html pip install torch-spline-conv==latest+${CUDA} -f https:\/\/pytorch-geometric.com\/whl\/torch-1.7.0.html Install PyTorch Geometric: !pip install torch-geometric Install the required framework: !pip install torch-geometric-temporal Data Structures This part will briefly talk about the data structures provided by PyTorch Geometric Temporal Framework. Discrete Dataset Iterators There are two types of discrete data iterators provided by PyTorch Geometric Temporal: Statistics Graphs with Discrete Signal : Used of discrete spatio-temporal signals on a static graph, the constructor is defined as StaticGraphDiscreteSignal which has the following parameter requirements: edge_index – A NumPy array to hold the edge indices.edge_weight – A NumPy array to hold the edge weights.features – A list of NumPy arrays to hold the vertex features for each time period.targets – A list of NumPy arrays to hold the vertex level targets for each time period. Static Graphs with Dynamic Signal :  Used of discrete spatio-temporal signals on a dynamic graph, the constructor is defined as  DynamicGraphDiscreteSignal which has the following parameter requirements: edge_indices – A list of NumPy arrays to hold the edge indices.edge_weights – A list of NumPy arrays to hold the edge weights.features – A list of NumPy arrays to hold the vertex features for each time period.targets – A list of NumPy arrays to hold the vertex level targets for each time period. Temporal Snapshots: It is a discrete temporal Data object, functions same as the Data object discussed in PyTorch Geometric. Benchmark Datasets There are many datasets provided by PyTorch Geometric Temporal framework for the comparison of performance of GNN algorithms Discrete Time Datasets : To import Hungarian Chickenpox Dataset discrete dataset, the code snippet is available below: from torch_geometric_temporal.data.dataset import ChickenpoxDatasetLoader loader = ChickenpoxDatasetLoader() #get_dataset is the method of StaticGraphDiscreteSignal object dataset = loader.get_dataset() Train-Test Splitter Discrete train-test splitter, splits the temporal and return train and test iterators generated from discrete time iterators using a fix ratio. The function of discrete_train_test_split is the same as scikit’s train_test_split. On an input of any StaticGraphDiscreteSignal or a DynamicGraphDiscreteSignal, discrete_train_test_split returns two iterators splitted according to the train_ratio. from torch_geometric_temporal.data.dataset import ChickenpoxDatasetLoader from torch_geometric_temporal.data.splitter import discrete_train_test_split loader = ChickenpoxDatasetLoader() dataset = loader.get_dataset() train_dataset, test_dataset = discrete_train_test_split(dataset, train_ratio=0.8) Applications of PyTorch Geometric Temporal This section will provide you an overview of how PyG Temporal is used in real-world scenarios. Learning from a Discrete Temporal Signal The following demo trains a regressor to predict the weekly Chickenpox  cases reported by the country. The dataset used for it is Hungarian Chickenpox Dataset. Load the dataset and divide the iterator into train set and test set. from torch_geometric_temporal.data.dataset import ChickenpoxDatasetLoader from torch_geometric_temporal.data.splitter import discrete_train_test_split loader = ChickenpoxDatasetLoader() dataset = loader.get_dataset() train_dataset, test_dataset = discrete_train_test_split(dataset, train_ratio=0.2) Next step is to create a Recurrent Graph Neural Network(RGNN) for supervised learning. The fundamental design of the architecture is the same as PyTorch Geometric. The RecurrentGNN function Object() {RecurrentGCN } generates a DCRNN and a feedforward layer, and the ReLU activation function is used to manually establish non linearity between the recurrent and linear layers. import torch import torch.nn.functional as F from torch_geometric_temporal.nn.recurrent import DCRNN class RecurrentGCN(torch.nn.Module): def __init__(self, node_features): super(RecurrentGCN, self).__init__() self.recurrent = DCRNN(node_features, 32, 1) self.linear = torch.nn.Linear(32, 1) def forward(self, x, edge_index, edge_weight): h = self.recurrent(x, edge_index, edge_weight) h = F.relu(h) h = self.linear(h) return h Train the model on the train_dataset for 200 epochs by back propagating the loss from every snapshot. The Adam optimizer is used with a learning rate of 0.01. from tqdm import tqdm model = RecurrentGCN(node_features = 4) optimizer = torch.optim.Adam(model.parameters(), lr=0.01) model.train() for epoch in tqdm(range(200)): cost = 0 for time, snapshot in enumerate(train_dataset): y_hat = model(snapshot.x, snapshot.edge_index, snapshot.edge_attr) cost = cost + torch.mean((y_hat-snapshot.y)**2) cost = cost \/ (time+1) cost.backward() optimizer.step() optimizer.zero_grad() Finally, run a test dataset to assess the model’s output and measure the Mean Squared Error (MSE) for all spatial units and time periods. model.eval() cost = 0 for time, snapshot in enumerate(test_dataset): y_hat = model(snapshot.x, snapshot.edge_index, snapshot.edge_attr) cost = cost + torch.mean((y_hat-snapshot.y)**2) cost = cost \/ (time+1) cost = cost.item() print(\"MSE: {:.4f}\".format(cost)) Colab Notebook : PyTorch Geometric Temporal Demo References : Official codes, Documentation and Tutorials are available at : Github Repository Documentation       Tutorial","excerpt":"PyTorch Geometric Temporal is a temporal extension of PyTorch Geometric(PyG) framework, which we have covered in our previous article. This open-source python library’s central idea is more or less the same as Pytorch Geometric but with temporal data. Like PyG, PyTorch Geometric temporal is also licensed under MIT. It contains many dynamic and temporal state-of-the-art […]","categories":["AI Trends"],"tags":["Deep Learning","graph convulationsl networks","graph neural networks"],"author_name":"Aishwarya Verma","publish_date":"2021-03-04T16:00:00","publication_year":"2021","word_count":925,"keywords":["CUDA","NumPy","TPU","AI","neural network","PyTorch","ML","graph convulationsl networks","Colab","Ray","deep learning","graph neural networks","Deep Learning"],"extracted_tech_keywords":["AI","ML","deep learning","neural network","Ray","PyTorch","Colab","NumPy","TPU","CUDA"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/pytorch-geometric-temporal-what-is-it-your-indepth-guide\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":51730,"title":"How DeepMind’s Reinforcement Agent Dreamer Can Predict The Future ","content":"Among the various applications of reinforced learning, DeepMind is now using RL to train its new reinforcement agent, Dreamer to learn to predict the actions of an object based on its current state. In short: its immediate future. Reinforcement learning is that area of machine learning where an agent takes suitable actions in order to get maximum rewards. Generally, a reinforced learning agent learns complex behaviour from the learned world models through the high dimensional sensory data input. There are many other potential ways of deriving behaviours from them, and now, agent Dreamer is capable of solving the long-horizon tasks from images purely using latent imagination. Put in simple words; the agent selects future actions based on the input by imagining their long term outcome. The Dreamer agent will broadly impact the data efficiency and computation time as well as the final performance. Meet Agent Dreamer Every agent in reinforced learning follows a model in order to predict rewards from observations and actions. In this context, the agent Dreamer learns a latent dynamic model to predict a reward. The latent dynamic model here is a model that learns from the image input and performs planning action to gather new experience. Quick Overview of the working Using latent imagination is better compared to the traditional image prediction here because these latent states have a small memory footprint that enables parallel imagination of thousands of trajectories. The use of word latent is used to imply the short sequence of hidden or latent states; this allows non-existential representations of objects like positions and velocities. The information from the input images is integrated into these latent states using an encoder component after which these hidden states project in future to anticipate images and rewards. Dreamer completes a pendulum swing task. The middle shows 45-step predictions. The Dreamer latent dynamic model is a multi-part complex structure. It has a representation model, transition model, reward model. The representation model encodes the actions and observations by the agent. The transition model anticipates the latent states without seeing the observations. The reward model rewards the model states. After anticipating the states, the action model aims to predict the action and solve the imagined environments within the learned policies. The action model’s achievements are estimated based on its actions, and a value model is there to determine the expected rewards. Then, finally, the feedback signals are provided by the observation model. Above: Dreamer playing an Atari game (Boxing). The middle shows 45-step predictions. Experimental Evaluation The Dreamer agent was experimentally evaluated on a variety of control tasks. These are some of the experiments that were designed to compare the agent Dreamer to current best methods by the researchers: Control Tasks: The Dreamer was evaluated and put through 20 visual control tasks of the DeepMind Control Suite. These control tasks pose many different challenges to the agent, including contact dynamics, sparse rewards and 3D scenes. Implementation: DeepMind first trained the agent on Nvidia V100 graphics chip and ten processor cores for each training run. Wach training run took 9 hours per 10^6 environment steps. This time recorded was a lot quicker than Google’s PlaNet, which took 17 hours. The world models are run through reconstruction unless specified. Performance: The Dreamer agent was compared to the state of the art reinforcement agents like D4PG and PlaNet. The Dreamer exceeds the performance in both environmental steps taken by D4PG and PlaNet’s data efficiency, hence proving that small quantity of experiences can help generalise world model. The Dreamer shows that learning behaviours from top methods in experience play are less efficient than the learning behaviours from the world models by latent imagination. Outlook With reinforcement learning becoming one of the most used parts of machine learning, now it has found its application in trying to predict the future with high dimensional inputs. The Dreamer using latent imagination means faster calculations and more efficient data processing for the real world. The potential of latent imagination is yet to be scaled in the future where the agents might be tested in an environment of more visual complexity. The researchers plan to present their work at NeurIPS 2019 in Vancouver this week.","excerpt":"Among the various applications of reinforced learning, DeepMind is now using RL to train its new reinforcement agent, Dreamer to learn to predict the actions of an object based on its current state. In short: its immediate future. Reinforcement learning is that area of machine learning where an agent takes suitable actions in order to […]","categories":["AI Features"],"tags":["DeepMind","future","future of deep reinforcement learning","Reinforcement Learning"],"author_name":"Sameer Balaganur","publish_date":"2019-12-12T12:26:07","publication_year":"2019","word_count":696,"keywords":["future","Go","future of deep reinforcement learning","machine learning","Reinforcement Learning","programming_languages:R","AI","programming_languages:Go","Aim","V100","R","DeepMind"],"extracted_tech_keywords":["AI","machine learning","Aim","R","Go","V100","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-deepminds-reinforcement-agent-dreamer-can-predict-the-future\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10124084,"title":"49% of Indian CEOs are Hiring for GenAI Roles that Didn’t Exist Last Year: IBM Study","content":"A new study by the IBM Institute for Business Value found that surveyed Indian CEOs are facing workforce, culture and governance challenges as they act quickly to implement and scale generative AI across their organisations. Around 49% of Indian CEOs surveyed said they are hiring for Gen AI roles that didn’t exist last year. Whereas 71% of them said that succeeding with AI will depend more on people’s adoption than the technology itself. Moreover, CEOs surveyed from India said 34% of their workforce will require retraining and reskilling over the next three years – up from just 6% globally in 2021. The annual global study of 3,000 CEOs from over 30 countries and 26 industries reveals a high importance placed by Indian CEOs on AI governance, with 71% of those surveyed saying trusted AI is impossible without effective AI governance in organisations. Further substantiating this, 75% of Indian CEO respondents say governance for generative AI must be established as solutions are designed rather than after they are deployed. At the same time, the study also noted a contrast in actual adoption of AI governance policies with only 42% Indian CEO respondents saying they have good generative AI governance in place today. This may be because people in the organization aren’t sure of exactly what they’re being asked to do. In the survey, 75% of Indian CEO respondents say that inspiring their team with a common vision produces better outcomes than providing precise standards and targets. Yet 31% acknowledge that their employees don’t fully understand how strategic decisions impact them. “As Indian CEOs navigate AI-led transformations within their organizations, they recognize the need for AI guardrails so that they derive real business value responsibly for growth and competitive success. However, our study reveals a gap between their intention and actual implementation. This scenario highlights the complexity of implementing AI governance, hence making a strong case for partnering with trusted experts to develop and execute effective practices and policies,” Sandip Patel, managing director of IBM India & South Asia, said. Other key study findings include: Indian CEOs recognize it takes a cultural shift to successfully scale AI, but face organizational collaboration and adoption challenges. 70% of Indian CEOs surveyed say their organization’s success is directly tied to the quality of collaboration between finance and technology, yet nearly half (48%) say competition among their C-Suite executives sometimes impedes collaboration. Nearly half (48%) of those surveyed from India acknowledge that cultural change is more important to becoming a data-driven organization than overcoming technical challenges. 58% of Indian CEO respondents say they are pushing their organization to adopt generative AI more quickly than some people are comfortable with Customer experience and product & service innovation are top priorities, regulatory constraints might be hindering long-term progress Indian CEOs surveyed ranked customer experience and product & service innovation as their highest priorities for the next three years. 59% of respondents say they are willing to sacrifice operational efficiency for greater innovation. However, nearly half (48%) of Indian CEOs surveyed point to regulatory constraints as their top barrier to innovation. Today, only 32% of the Indian CEO respondents are primarily funding their generative AI investments with net new IT spend, with the remaining 68% reducing other technology spend.","excerpt":"CEOs surveyed from India said 34% of their workforce will require retraining and reskilling over the next three years","categories":["AI News"],"tags":["IBM"],"author_name":"Pritam Bordoloi","publish_date":"2024-06-20T15:29:09","publication_year":"2024","word_count":540,"keywords":["Go","funding","AI","data-driven","innovation","generative AI","IBM","Rust","GAN","AI governance","R"],"extracted_tech_keywords":["AI","generative AI","R","Go","Rust","GAN","AI governance","data-driven","innovation","funding"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/49-of-indian-ceos-are-hiring-for-genai-roles-that-didnt-exist-last-year-ibm-study\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10072349,"title":"Edtech unicorn upGrad raises $210 million","content":"Edtech startup upGrad has raised $210 million in its latest round from ETS Global, Bodhi Tree, and Singapore’s Kaizen Management Advisors. To maintain over 50% ownership, the founder group has invested $12.5 million in the startup. The family office of Bharti Airtel, Narotam Sekhsaria, and Lakshmi Mittal’s Artisan Investments took part in the round, along with existing investors Temasek, IFC, and IIFL. The firm aims to strengthen its team from 4,800 to 7,600 in the next three months, which will include 170 full-time faculty, 1,600 teachers, and over 5,000 on-contract coaches. Commenting on the funding in a joint statement, the co-founders said, “Higher edtech will be on the rise for the next 4 to 5 decades. upGrad, in the past 12 months, has re-shaped itself to be the most integrated company in this space with career opportunities for college-goers and working professionals from the age of 18 to 58 and will be a lifelong learning partner for millions in the coming years. We have always been capital efficient while growing 100% year-on-year, and hope to retain that discipline.” Established in 2015 by Ronnie Screwvala, Mayank Kumar, and Phalgun Kompalli, upGrad has a learner base of over 3 million across 100 countries, along with 300 university partners.","excerpt":"The firm aims to strengthen its team from 4,800 to 7,600 in the next three months, which will include 170 full-time faculty, 1,600 teachers, and over 5,000 on-contract coaches.","categories":["AI News"],"tags":[],"author_name":"Bhuvana Kamath","publish_date":"2022-08-08T17:30:11","publication_year":"2022","word_count":206,"keywords":["Go","API","funding","lifelong learning","programming_languages:R","AI","programming_languages:Go","Aim","R","startup"],"extracted_tech_keywords":["AI","Aim","R","Go","API","lifelong learning","startup","funding","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/edtech-unicorn-upgrad-raises-210-million\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":20699,"title":"Overview of Recurrent Neural Networks And Their Applications","content":"Recurrent Neural Networks are one of the most common Neural Networks used in Natural Language Processing because of its promising results. The applications of RNN in language models consist of two main approaches. We can either make the model predict or guess the sentences for us and correct the error during prediction or we can train the model on particular genre and it can produce text similar to it, which is fascinating. What is a Recurrent Neural Network? The logic behind a RNN is to consider the sequence of the input. For us to predict the next word in the sentence we need to remember what word appeared in the previous time step. These neural networks are called Recurrent because this step is carried out for every input. As these neural network consider the previous word during predicting, it acts like a memory storage unit which stores it for a short period of time. The above diagram represents a three layer recurrent neural network which is unrolled to understand the inner iterations. Lets look at each step, xt is the input at time step t. xt-1 will be the previous word in the sentence or the sequence. ht will be the hidden state at time step t. The output of this state will be non-linear and considered with the help of an activation function like tanh or ReLU. ht-1 is evaluated from the previous hidden layer, usually it is initialized to zero. yt will be our output at the time step t. It is the a word which has the highest probability considered using a an activation function like yt = maxout(Ast). The RNN in the above figure has same evaluation at teach step considering the weight A, B and C but the inputs differ at each time step making the process fast and less complex. It remembers only the previous and not the words before it acting like a memory. Language Modelling and Prediction: In this method, the likelihood of a word in a sentence is considered. The probability of the output of a particular time-step is used to sample the words in the next iteration(memory). In Language Modelling, input is usually a sequence of words from the data and output will be a sequence of predicted word by the model. While training we set xt+1 = ot, the output of the previous time step will be the input of the present time step. Speech Recognition: A set of inputs containing phoneme(acoustic signals) from an audio is used as an input. This network will compute the phonemes and produce a phonetic segments with the likelihood of output. Machine Translation: In Machine Translation, the input is will be the source language(e.g. Hindi) and the output will be in the target language(e.g. English). The main difference between Machine Translation and Language modelling is that the output starts only after the complete input has been fed into the network. Image recognition and characterization: Recurrent Neural Network along with a ConvNet work together to recognize an image and give a description about it if it is unnamed. This combination of neural network works in a beautiful and it produces fascinating results. Here is a visual description about how it goes on doing this, the combined model even aligns the generated words with features found in the images. Source: http:\/\/cs.stanford.edu\/people\/karpathy\/deepimagesent\/ What are LSTM(Long Short Term Memory) Neural Networks? The LSTM networks are popular nowadays. The LSTM network are called cells and these cells take the input from the previous state ht-1 and current input xt. The main function of the cells is to decide what to keep in mind and what to omit from the memory. The past state, the current memory and the present input work together to predict the next output. Architecture of LSTM network: LSTM network have a sequence like structure, but the recurring network has a different module. Instead of having single neural network layer, they have small parts connected to each other which function in storing and removal of memory. Working of LSTM networks: The first step in the LSTM is to decide which information to be omitted in from the cell in that particular time step. It is decided by the sigmoid function which omits if it is 0 and stores if it is 1. It looks at the previous state ht-1 and the current input xt and computes the function. To understand the activation functions and the math behind it go here. Then we have another layer which consists of two parts. One is the sigmoid function and the other is the tanh. In the sigmoid function, it decided which values to let through(0 or 1). And in the tanh function its gives the weightage to the values which are passed deciding their level of importance(-1 to 1). Finally, we need to decide what we’re going to output. This output will be based on our cell state, but will be a filtered version. First, we run a sigmoid layer which decides what parts of the cell state we’re going to output. Then, we put the cell state through tanh to push the values to be between -1 and 1 and multiply it by the output of the sigmoid gate, so that we only output the parts we decided to. Image Sources: Understanding of LSTM","excerpt":"Recurrent Neural Networks are one of the most common Neural Networks used in Natural Language Processing because of its promising results. The applications of RNN in language models consist of two main approaches. We can either make the model predict or guess the sentences for us and correct the error during prediction or we can […]","categories":["AI Features"],"tags":["lstm","Recurrent Neural Network"],"author_name":"Kishan Maladkar","publish_date":"2018-01-17T09:27:14","publication_year":"2018","word_count":887,"keywords":["Recurrent Neural Network","Go","TPU","AI","neural network","RPA","lstm","image recognition","RAG","RNN","LSTM","R"],"extracted_tech_keywords":["AI","neural network","RAG","image recognition","TPU","R","Go","RNN","LSTM","RPA"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/overview-of-recurrent-neural-networks-and-their-applications\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10088901,"title":"OpenAI Junks Diffusion for Consistency Models","content":"In 2021, OpenAI chief Sam Altman wrote a blog discussing how Moore’s Law—the theory that semiconductor chips would become twice as powerful for the same price around every two years—should be applicable for everything. Altman tweeted about the leapfrogs that AI was making, saying, ”A new version of Moore’s law that could start soon: the amount of intelligence in the universe doubles every 18 months”. Introducing consistency models To others, Altman’s optimism may seem unwarranted but OpenAI’s pace of research seems to back up his claims. Last week, the startup published a paper discussing a new class of generative models titled, ‘Consistency Models’, that outperformed diffusion models. Authored by Yang Song, Prafulla Dhariwal, Mark Chen and OpenAI co-founder Ilya Sutskever, the study was released on March 3, 2023. Diffusion models have become the foundation of the revolution in generative AI since they took over GANs as the most effective models for image synthesis. Some of the most prominent text-to-image AI generators such as OpenAI’s DALL.E 2, Stability AI’s Stable Diffusion and Google’s Imagen are all diffusion models. Faster and less energy-intensive than Diffusion models However, consistency models have proven to produce the same quality output as diffusion models in much less time. This is because the consistency model works on a single-step generation process like GANs. Diffusion models, in contrast, work around a repetitive sampling process which progressively removes noise from an image. The continuous iterative generation process of diffusion models eats up 10–2000 times more compute in comparison to consistency models and slows down the inference during training. Consistency models are able to trade-off compute for sample quality when necessary. Besides this, such models are also capable of performing zeroshot data editing tasks like image inpainting, colorisation or stroke-guided image editing. Canadian computer scientist Ilya Sutskever, co-founder and chief scientist of OpenAI, Source: University of Toronto These models also use a mathematical equation to transform data into noise and ensure that the resulting output is consistent for similar data points, allowing for smooth transitions between them. Such equations are called probability flow ordinary differential equations. The study has named this class of models ‘consistency’ because they maintain this property of self-consistency throughout between the input data and the output. These models can either be trained in the distillation mode or the isolation mode. In the distillation mode, consistency models are able to distill the data from pre-trained diffusion models into a sampler that can perform in a single step. While in isolation mode, consistency models don’t depend on diffusion models at all, thereby making them an entirely independent type of models. No adversarial training, no problem Both methods of training however have removed adversarial training from their books. Adversarial training does result in a stronger neural network but goes about the process in a roundabout way—it introduces a wrongly classified set of adversarial examples and then retrains the target neural network with the correct labels. Consequently, adversarial training has been also found to lead to a slight decrease in the accuracy in predictions by deep learning models. They can also cause unexpected side effects in robotics applications. Consistency Models, from OpenAI:a new type of generative model distinct from diffusion models or GANs, specifically designed for one\/few-step generation with a notably improved distillation methodarxiv: https:\/\/t.co\/Im9EKg6JcD pic.twitter.com\/MX9XCr9EtG— near (@nearcyan) March 4, 2023 The experiments showed that the distillation techniques used in training consistency models were better than the distillation techniques used in diffusion models. Consistency models achieved a new state-of-the-art Frechet Inception Distance score—which is indicative of the quality of AI generated images—of 3.55 on the CIFAR10 image dataset and 6.20 on the ImageNet 64*64 dataset. It’s fair to say that OpenAI isn’t the only stakeholder here but is definitely one of the major ones. If they want their AI tools to sell more, the onus falls on them to ensure that they take less time and use less compute. In that sense, the potential impact of consistency models is huge since diffusion models aren’t only popular in image generation but also in video and audio generation models. outstanding! Simo Ryu (the person who brought LoRA to stable-diffusion) has a working implementation of consistency models (OpenAI's new class of generative model), 4 days after the paper was published.these models enable few-step sampling.can't wait for distillation support! https:\/\/t.co\/JAtxlxJZMf— Birchlabs (@Birchlabs) March 6, 2023 Just last month, Sutskever posted a tweet with a hint, saying, “Many believe that great AI advances must contain a new ‘idea’. But it is not so: many of AI’s greatest advances had the form huh, turns out this familiar unimportant idea, when done right, is downright incredible”. This paper shows exactly that—built on older concepts with a tweak can change everything.","excerpt":"Consistency models, on the other hand, are a single-step generation which is faster and is able to trade-off compute for sample quality when necessary.","categories":["Global Tech"],"tags":["AI Tool"],"author_name":"Poulomi Chatterjee","publish_date":"2023-03-08T18:00:00","publication_year":"2023","word_count":783,"keywords":["Go","TPU","OpenAI","AI","neural network","Aim","deep learning","stable diffusion","generative AI","AI Tool","R"],"extracted_tech_keywords":["AI","deep learning","neural network","generative AI","OpenAI","Aim","TPU","R","Go","stable diffusion"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/openai-junks-diffusion-for-consistency-models\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":32492,"title":"Microsoft&#8217;s New AI competition Challenges Data Scientists Worldwide To Prevent Malware Attacks","content":"2018 will be remembered as the year of data breaches. With well-established and multi-billion dollar companies like Facebook and Marriott falling victims to cybersecurity and data breaches, the concern for data security has grown significantly. With the improvement of technology and its availability to a large number of consumers, the use of technology remains unmonitored. In an attempt to curtail data breaches Microsoft with its new competition on Kaggle, is challenging the Data Science community to use Artificial Intelligence to prevent cyber security attacks.The competition is organized by the Windows Defender ATP Research team, in collaboration with Northeastern University and Georgia Institute of Technology as academic partners. Artificial Intelligence is already helping in detecting and stopping a malware at first sight.With this competition, Microsoft takes it a step further. The idea is to stop a malware before it is even seen. The objective is to use predictive analysis to predict  if a machine will soon be hit with a malware.The competition presents a fresh opportunity for data science enthusiasts across the world to showcase their skills on a real world problem more real than any other dataset you may possibly find. The company intends to use the results from the competition to further improve their defences. The challenge is to identify the systems with higher risk of getting affected so that preemptive measures can be taken. Here’s What You Should Know About The Data Participants will build their predictive models using 9.4GB of data collected from over 16.8 million devices.The dataset has over 82 features or columns of data. The contestants have to build a machine learning model to uniquely identify a machine that is affected by a malware.Each machine is identified by a unique MachineIdentifier which is a column in the provided data set.The column or feature HasDetections is the dependent factor the contestants have to predict for each machine in the test set.. Important Dates and Prizes The competition was launched in Kaggle on December 13 2018 and has already gained the attention of almost 700 data scientists. Entry deadline : March 6, 2019 Team Merger deadline (participants can join or merge teams) : March 6, 2019 Final submission deadline : March 13, 2019 . Microsoft presents the winners with an overall $25000 cash prizes: 1st Place – $12,000 2nd Place – $7,000 3rd Place – $3,000 4th Place – $2,000 5th Place – $1,000 Outlook This is not the first time Microsoft has challenged the Data Science community. In 2015, Microsoft hosted a similar malware classification competition which was a great success in terms of helping Microsoft as well as a large community of data scientists. The datasets provided by Microsoft are used in a number of researches and continues to produce value for Microsoft and the data science community.","excerpt":"2018 will be remembered as the year of data breaches. With well-established and multi-billion dollar companies like Facebook and Marriott falling victims to cybersecurity and data breaches, the concern for data security has grown significantly. With the improvement of technology and its availability to a large number of consumers, the use of technology remains unmonitored. […]","categories":["AI News"],"tags":[],"author_name":"Amal Nair","publish_date":"2018-12-31T09:39:05","publication_year":"2018","word_count":463,"keywords":["data science","machine learning","artificial intelligence","programming_languages:R","AI","GAN","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","data science","R","GAN","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/microsofts-new-ai-competition-challenges-data-scientists-worldwide-to-prevent-malware-attacks\/","complexity_score":3,"technical_depth":7,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10043124,"title":"Why Are People Criticising GitHub Copilot?","content":"Last week, Microsoft and OpenAI released the technical preview of GitHub Copilot, an AI-based assistant to help programmers write better codes. Not to be confused with autofill, this assistant takes context from the code being worked on to suggest successive lines of code and functions. Copilot is based on OpenAI Codex, an AI system trained on public source code. It works well with languages such as Python, TypeScript, Javascript, Ruby, and Go. The announcement of GitHub Copilot kicked up a storm in social media. The initial reaction has been largely positive, with many people calling the coding assistant a game-changer. https:\/\/twitter.com\/eigenbom\/status\/1410095798600142851?s=20 https:\/\/twitter.com\/ItalyPaleAle\/status\/1409890404615409671?s=20 Today on the stream was WILD playing with GitHub Copilot!You can see all of the various functions we tried in this gist here. Some of them are hilarious, some are only partially right, but overall it was a really impressive experience.https:\/\/t.co\/sPCVi2xnEd pic.twitter.com\/o8Wx0JeZiX— Cassidy (@cassidoo) July 1, 2021 However, few questions have been raised: Since the tool is trained on publicly available code repositories–most of which are licensed and under copyright protection–what happens when the tool reproduces these code snippets? Is it legal? Can parent organisations –Microsoft, OpenAI and GitHub– monetise this tool even if it is trained on free and open-source code? Copyright infringement GitHub has said there is 0.1 percent chance of Copilot replicating the learned snippet of code verbatim. As one Twitter user pointed out, this could be a potential case for ‘code laundering’ for commercial use, which not only involves copying the content as is but also derivative works. https:\/\/twitter.com\/eevee\/status\/1410037309848752128?s=20 GitHub allows users to choose a license to protect their work. The most common licenses include MIT, Apache, and GPL. iIn which, GPL is a free, copyleft license offering protection against verbatim copying of the work and also necessitates distribution of derivative work under the same or equivalent terms. A section of Twitterati has asked if Copilot is following the fair code usage terms. Hi. I know you’re excited about copilot. GitHub scraped your code. And they plan to charge you for copilot after you help train it further.It’s truly disappointing to watch people cheer at having their work and time exploited by a company worth billions.— Brian P. Hogan (@bphogan) July 2, 2021 To that end, GitHub CEO Nat Friedman wrote in a discussion thread on HackerNews: “In general: (1) training ML systems on public data is fair use (2) the output belongs to the operator, just like with a compiler. On the training question specifically, you can find OpenAI’s position, as submitted to the USPTO here: https:\/\/www.uspto.gov\/sites\/default\/files\/documents\/OpenAI_R… We expect that IP and AI will be an interesting policy discussion around the world in the coming years, and we’re eager to participate!” Support for Copilot Neil Brown, a legal expert in the digital space, spoke about Copilot from an English law perspective. In his blog, Brown explained GitHub’s passage D4 of Terms of Service. As per this passage, GitHub can copy a user’s content to the database, create backups, show it to other users, parse a search engine, and analyse it on their servers. Brown writes: “The license is broadly worded, and I’m confident that there is scope for argument, but if it turns out that Github does not require a license for its activities then, in respect of the code hosted on Github, I suspect it could make a reasonable case that the mandatory license grant in its terms covers this as against the uploader.” That said, GitHub also notes in the same passage that this license doesn’t grant permission to sell content, distribute it or use it outside the scope of the service. Further, Julia Reda, a researcher and former Member of the European Parliament wrote a blog (leaning more towards the EU perspective) titled “GitHub Copilot is not infringing your copyright”. She makes her argument on two grounds: Data and text mining: Merely scraping off code without the author’s consent, although worthy of criticism, is not a copyright-relevant act that requires permission. Derivative works and machine-generated code: Putting machine-generated code under the purview of derivative works is ‘dangerous’. Firstly, this assumption suggests that even the smallest piece of excerpts could constitute copyright infringement. Secondly, the very premise of machines being capable enough to produce works is wrong and counterproductive. The snippets of code appear verbatim when the developer has not provided sufficient context or when there is a universal solution to the problem, the GitHub blog claimed. Also, the GitHub team is building an origin tracker to detect such code duplication instances. Wrapping up GitHub in its blog clearly mentions that Copilot is to be seen strictly as an AI pair programmer to assist in writing codes. The coders who got access to the tool echoed similar sentiments claiming while Copilot is impressive, it cannot be equated with human programmers. According to blogger Colin Eberhardt, the Copilot has the “wow” factor to make it to the standard toolset of enterprises. However, he thinks it will take some time for the coding assistant to deliver a genuine productivity boost.","excerpt":"Last week, Microsoft and OpenAI released the technical preview of GitHub Copilot, an AI-based assistant to help programmers write better codes. Not to be confused with autofill, this assistant takes context from the code being worked on to suggest successive lines of code and functions. Copilot is based on OpenAI Codex, an AI system trained […]","categories":["Global Tech"],"tags":["copyright","copyright infringement","GitHub","Github Copilot"],"author_name":"Shraddha Goled","publish_date":"2021-07-07T17:00:00","publication_year":"2021","word_count":837,"keywords":["Go","TPU","OpenAI","AI","ML","Github Copilot","TypeScript","Python","copyright","Aim","JavaScript","GitHub","R","copyright infringement"],"extracted_tech_keywords":["AI","ML","OpenAI","Aim","TPU","Python","R","JavaScript","TypeScript","Go"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/why-are-people-criticising-github-copilot\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10102945,"title":"GPT-4 Turbo is &#8216;Lost in the Middle&#8217;","content":"Last week, OpenAI held its first-ever developers’ conference in which it announced GPT-4 Turbo, an enhanced iteration of GPT-4. It features an expansive 128k context window, enabling it to process the equivalent of over 300 pages of text in a single prompt. This upgrade comes with knowledge extending up to April 2023. Several announcements were made, spanning open source models and developer tools, addressing areas where the generative capabilities of OpenAI previously faced competition gaps. These announcements drew the attention of the AI world as they seemingly sounded the death knell for many AI startups out there. A week later, however, the hype is on the slide and nothing much has changed. Maybe, GPT-4 Turbo was not as revolutionary as it sounded. Context Length Extn Issue In a late July study, Stanford University, UC Berkeley, and Samaya AI researchers revealed a phenomenon in large language models termed ‘Lost in the Middle’, where information retrieval accuracy is high at the document’s start and end but declines in the middle, especially with increased input processing. Building on this, Greg Kamradt, Shawn Wang, and Jerry Liu tested if GPT-4 Turbo exhibited this effect. Using YC founder Paul Graham’s essays, they inserted a random statement at different document points and evaluated GPT-4’s recall. Findings showed decreased recall above 73,000 tokens, particularly affecting mid-document statements, emphasizing context length’s impact on accuracy. It means that the accuracy typically drops off as you get to 60-70% of the context length supported by an LLM. Opting for smaller context-length inputs is recommended for accuracy, even with the advent of long-context LLMs. Notably, facts at the input’s beginning and end are better retained than those in the middle. Comparatively, a 128K context-length LLM performs better than a 32K context-length one for a given context, suggesting the use of large context-length LLMs with relatively smaller documents. The “forgetting problem” remains a challenge, requiring ongoing development in LLM applications with multiple components and prompt engineering. While larger context windows, such as those offered by advanced language models like GPT-4, allow for more extensive data processing in a single prompt, embedded search functions or vector databases remain superior in terms of accuracy and cost-effectiveness, particularly for specific information retrieval tasks. Vector databases specialise in organising and retrieving information based on semantic similarities, offering a more targeted and efficient approach. These systems are designed to excel in precision, ensuring that the retrieved information aligns closely with the user’s query. Additionally, the focused nature of embedded search functions often results in reduced computational costs, making them an optimal choice for specific and precise data retrieval needs. OpenAI’s Retrieval APIs Not the Ultimate Solution While OpenAI’s introduction of retrieval APIs is noteworthy, it’s crucial to highlight the limitation of exclusively working with GPT-4. Despite a price reduction, the scalability of usage remains a significant challenge due to its high cost. There are open-source retrieval APIs that are revolutionising enterprise LLM adoption. These APIs come equipped with open-source LLMs tailored for enterprise applications, featuring expansive 32K contexts and specialization in specific enterprise use cases like Q\/A and summarization. The cost-effectiveness of these open-source APIs is noteworthy, being 20 times more economical than GPT-4. Additionally, developers have the flexibility to switch to closed-source LLMs from OpenAI, Anthropic, or Google if that better aligns with their preferences. Furthermore, if a customized fine-tuned LLM is essential and a developer possesses labeled data, they provide the service of fine-tuning the LLM to meet their specific requirements. In many instances, the combination of Retrieval-Augmented Generation (RAG) with fine-tuning proves to yield optimal results. In the ever-evolving landscape of an enterprise’s internal knowledge base, the challenge is avoiding the hassle of repeatedly uploading new data each time the database undergoes changes. Typically, enterprise clients store their data in cloud repositories such as Azure, GCP, and S3. The open-source retrieval APIs facilitate a seamless connection to these cloud buckets, ensuring regular updates without manual intervention. Moreover, this functionality extends to pulling in data from various sources, including Confluence or any cloud database like Snowflake, Databricks, and others, enhancing versatility and adaptability. While the intricacies are abstracted for a seamless experience, the open-source retrieval API allows users the flexibility to delve into the details and fine-tune parameters as needed. Despite the API’s intelligent approach in making decisions on chunking and embedding strategies based on dataset and API requests, users retain the ability to make manual adjustments. In the realm of enterprise operations, establishing pipelines and robust monitoring systems is indispensable. Connecting to diverse data sources, ensuring regular updates to vector stores, and meticulous indexing are vital components. The Retrieval API fundamentally streamlines the development of LLM applications on your data, offering a quick start within a few hours. It emerges as the optimal choice, especially for those emphasising cost-effectiveness and scalability in Retrieval\/RAG processes.","excerpt":"OpenAI’s GPT-4 Turbo, with an extensive 128k context window, failed to revolutionise things due to the ‘Lost in the Middle’ phenomenon impacting information recall accuracy.","categories":["AI Features"],"tags":[],"author_name":"Tausif Alam","publish_date":"2023-11-13T14:01:01","publication_year":"2023","word_count":800,"keywords":["Anthropic","GCP","OpenAI","AI","ML","RAG","vector databases","prompt engineering","Azure","Snowflake"],"extracted_tech_keywords":["AI","ML","OpenAI","Anthropic","RAG","vector databases","prompt engineering","Azure","GCP","Snowflake"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/gpt-4-turbo-after-hype-comes-lost-in-the-middle-phenomenon\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":64978,"title":"Reasons Why You Should Be Learning Data Science During This Time","content":"The year 2020 has proven to be a testing time for professionals across all the industries. Where the world’s economic sector is looking at a recession and cloud computing seeing more adoption, it seems like every industry has been affected by the pandemic. Still, when we talk about the data science industry, it has been one of the most active and most affected ones during COVID-19. As India, along with other countries slowly eases its lockdown, and the economy slowly makes its way back, data science will be one of the main drivers behind it. So, it becomes imperative to either start learning data science now or learn more if you are already working with data. Below, we have mentioned some of the most important reasons that make learning data science during this time very important: Data Science Will See Increased Adoption In Healthcare One of the biggest sectors that would leverage data science capabilities is the healthcare sector. It is a fact that during the pandemic, data science has assisted in interpreting and translating COVID related information accurately. And once the threat of the pandemic is over, it will receive even more attention. When the COVID-19 outbreak was going on, China’s technology giant Baidu’s Linearfold algorithm was made available for researchers to aid medical teams to combat COVID-19. The company also decided to roll out several tools to test the infections and raised awareness. In future, if there is a threat of another virus outbreak, the medical community and the entire world would require better data analysis, accurate visualisations and tools to combat better than we did during the COVID-19. So, learning data science during this period and honing your skills as a data scientist will put you in a great position to enter the healthcare industry and make a difference. Different Approach To Using Data Will Be Needed One of the great realisations for the data science community during the pandemic has been about data. COVID-19 has shown us how important data can be, both for businesses as well as healthcare. The data that has been collected during this pandemic can potentially help suppress any future threats. Data collected by businesses today will help them to be better prepared to counter consumers’ and clients’ behaviour changes during the next pandemic if any. Surely, as companies lay even more emphasis on data than they usually do, talented data scientists will see more demand than ever before. Another aspect of data that COVID-19 has made us realise is that one can draw insights from data that isn’t directly connected to the problem. As Robert Munro illustrates in his post, that one can analyse data that tells us something important about a problem, but it doesn’t necessarily have a direct relationship with the problem. For example, he outlined how he calculated that people who died in West Africa during the Ebola outbreak showed symptoms of diseases which weren’t Ebola and the deaths were a result of negligence towards getting tested at clinics. He saw that for every person who died from Ebola, ten more died from a treatable illness. The meaning of giving this example is to say that data can be explored in different ways, either for healthcare or for business purposes. Learning about how to deal with data will be an effective asset in your data science initiatives. Better Data Visualisation Is Needed Once the pandemic is over, better data visualisation will be needed. Companies will require more insights and better visualisations than ever to help bring their financial losses back on track. It would be safe to say that data will never stop coming in, and after the pandemic, the more insights one gets from the data, the more beneficial it will prove to be. So, data visualisations that are more interactive, scalable and clear will be in high demand, naturally increasing the demand for data visualisations experts. During the pandemic, we came across many visualisations that were trajectory charts. It showed the number of deaths to the number of cases. Majority of charts, especially shown through media, almost didn’t qualify as good visualisations as they completely lacked interactivity, scalability, actual insights, and most importantly, actionable metrics (like progression rate, current trend if slowing down etc. The pandemic has shown how badly visualisation experts are needed. More Emphasis Will Be Given To Data Culture After things get back to normal, businesses have to cope up with the global losses. To thrive in the aftermath of the pandemic, companies and businesses will rely heavily on data science. Organisations will begin to make more data optimised decisions to avoid losing to their competitors. But, it is a known fact that there was a dearth of talented data scientists even before the pandemic. Therefore, organisations will look for data scientists who can assist them in making data-driven culture with effective decisions and enhanced accuracy. The important components of a data culture are: Data Literacy: It is the ability to use data appropriately towards making informed decisionsData Maturity: It is a scale, where having a score 7\/10 typically tells you that an organisation has well-defined data sources with appropriate access levelsData-driven Leadership: Data leaders should focus on building a culture of data-driven thinking. Data leaders are also responsible for making data the key asset of the organisation.Data-Driven Decision Making: This process should ensure a systematic way of making decisions that involve transparency. Besides, evaluation and learning will ultimately result in a company’s better performance.","excerpt":"The year 2020 has proven to be a testing time for professionals across all the industries. Where the world’s economic sector is looking at a recession and cloud computing seeing more adoption, it seems like every industry has been affected by the pandemic. Still, when we talk about the data science industry, it has been […]","categories":["AI Highlights"],"tags":["learning data science","purposes of a data team","what is data science"],"author_name":"Sameer Balaganur","publish_date":"2020-05-11T18:00:00","publication_year":"2020","word_count":912,"keywords":["data science","Go","AI","cloud computing","learning data science","Scala","RAG","BERT","ViT","purposes of a data team","GAN","what is data science","R"],"extracted_tech_keywords":["AI","data science","RAG","cloud computing","R","Go","Scala","BERT","GAN","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/reasons-why-you-should-be-learning-data-science-during-this-time\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":61384,"title":"Case Study: How This Ahmedabad-Based Cancer Center Is Using Robotic System For Minimally Invasive Surgeries","content":"A dedicated, private and comprehensive cancer center, ‘HCG – The Specialist in Cancer Care’ has been designed to address the cancer care needs of Gujarat and adjoining states of Rajasthan and Madhya Pradesh. With services like surgical, radiation and medical oncology all under one roof with a full range of diagnostic, pathology, radiology and nuclear medicine departments, the hospital claims to be a ‘one-stop solution’ for comprehensive cancer care. The center was established to provide high quality and result-oriented cancer care treatment by adopting global innovations. HCG – The Specialist in Cancer Care, with 24 centers across India, currently claims to be the largest cancer care provider in the country. The Challenges Out of all the established hospitals of the HCG group, HCG Cancer Centre in Ahmedabad, Gujarat is one of the main cancer hospitals. And therefore, most of the complex cases are usually referred here — not only from Gujarat, but also from other parts of India and abroad. Being the tertiary care hospital of the state, HCG’s highly qualified, trained and experienced oncologists have to deal with high-risk cancer patients daily. The hospital, from the last 25 years, has been focused on offering a sub-speciality approach and became the first private cancer center in Gujarat. The hospital boasts of a strong surgeon team to provide the best cancer outcomes, where the primary focus has always been on complete removal of cancer for better patient survival and reduced recurrence. To achieve these outcomes, the hospital was earlier dependent on open surgery. However, open surgeries always came with some challenges such as a large scar, delayed recovery, more blood loss etc. The hospital quickly realized the transformation and was looking to adopt new and advanced technologies to provide better cancer outcomes with benefits of minimally invasive surgeries. Deployment Of Solution & Benefits Over the last few years, the healthcare industry has witnessed an increasing acceptance of robotic-assisted surgeries in the world, especially for complex operations like radical prostatectomy, radical cystectomy, partial nephrectomy, rectal resection surgeries, thoracic surgeries and onco-gynaecological procedures. And to reach human organs located in remote areas of the patient’s body, and shorten the surgery time, shorter hospital stay and to reduce the side effects of the treatment, HCG Cancer Centre deployed Gujarat’s first 4th Generation da Vinci Xi Robotic System. HCG Cancer Centre started using high-end computerized robotic systems to perform surgical procedures, whether it be urology surgeries, gynaecological procedures, general surgery, head and neck surgeries, or for cancer treatment. Robotic-assisted surgery was deployed in HCG Cancer Centre to offer the best technology for better patient outcomes and aid surgeons to perform complicated surgeries. It not only allows the surgeons to perform complex surgical procedures with more precision but also provides them with enough flexibility and control than traditional techniques, such as open surgery and laparoscopy. The doctors of HCG Cancer Centre were initially sceptical about the value addition of robotic approach for prostatectomies, “however, as we started performing robotic prostatectomies, we started to notice the ease of performing the surgeries was tremendous,” said Dr Hemang Baxi, the head of uro-oncology & robotic-assisted surgery department. Alongside “the ergonomics and the magnification provided by the robotic technology have translated into the better precision of surgical steps,” added Dr Baxi. Dr Hemang Baxi Head of Uro Oncology & Robotic-assisted Surgery Department with the da Vinci robotic system According to Dr Baxi, the significant advantages in terms of function recovery were the earlier return of urinary continence, which is significantly shorter diaper time, along with the significantly reduced incidence of bladder neck contractures. Post-deployment of the robotic system, Baxi also noticed that the blood loss after the surgical procedures had decreased considerably, and the hospital stay along with the overall patient recovery was much shorter. Dr Baxi also observed that because of precise vesicourethral anastomosis in the last step of radical prostatectomy, the incidence of anastomotic strictures was almost negligible. “It is also significant that, out of more than 200+ robotic prostatectomies performed at our center, not a single patient has an anastomotic urinary leak,” added Dr Baxi. Robotic-assisted surgery also helped the hospital in precise tumour excision in required cases, by providing better magnification and better ergonomics leading to faster suturing as compared to laparoscopy. “There are other significant advantages of robotic technology observed in bladder cancer patients undergoing urinary bladder issues,” said Dr Baxi. “These patients tend to have a higher complication rate which can be minimised by using intracorporeal robotic techniques leading to lesser incidence of ureteric stricture, lesser fluid loss and electrolyte disturbances.” Alongside, Dr Jagdish Kothari, the head surgical oncologist and robotic surgeon at HCG, a firm believer of open surgery for complete cancer removal also decided to explore minimally invasive surgeries for GI cancers. He said that once he started looking at the literature and came across many successful studies from eminent hospitals across the world, his perspective changed and started using it have a better outcome for patients. Today at HCG Cancer Center, Ahmedabad, more than 80% surgeries for cancer of Oesophagus, Stomach, Colon & Rectum are done by Minimally Invasive technique. The da Vinci robotic system has been designed to enable complex surgeries using a minimally invasive approach. The surgical system includes an ergonomic surgeon console and interactive arms with a high-performance vision system and wristed instrumentation. The console gives the surgeon a high-definition, magnified, 3-D view of the surgical site, and the surgeon gets to control the arms while seated at the computer console near the operating table. Explaining further, Mandeep Singh Kumar, the vice president and country general manager of Intuitive India, said, “By providing surgeons with superior visualization, enhanced dexterity, greater precision and ergonomic comfort, da Vinci robotic-assisted surgical systems made it possible for them to perform minimally invasive procedures involving complex dissection or reconstruction.” “For patients, da Vinci surgery is offering potential benefits such as smaller incisions, less blood loss, fewer complications, faster recovery with shorter hospital stays and fewer readmissions,” said Kumar. Wrapping Up According to Kumar, a desire to offer a minimally invasive option to patients have driven the demand for robotic-assisted surgery in India. The benefits of robotic-assisted minimally invasive surgery not only include fewer complications or lesser pain for patients but also faster recovery, which is essential in getting the patients back to their lives faster. These benefits consecutively can create a profitable framework for hospitals in India. Faster recovery of patients means shorter hospital stays and lesser readmissions, which will, in turn, free up the limited supply of beds in Indian hospitals and help in taking the pressure off the overall healthcare system.","excerpt":"A dedicated, private and comprehensive cancer center, ‘HCG – The Specialist in Cancer Care’ has been designed to address the cancer care needs of Gujarat and adjoining states of Rajasthan and Madhya Pradesh. With services like surgical, radiation and medical oncology all under one roof with a full range of diagnostic, pathology, radiology and nuclear […]","categories":["AI Features"],"tags":["cancer","healthcare robotics","Robotic Process Automation","robotic surgeon","Robotics","Robotics India","Robotics Process Automation"],"author_name":"Sejuti Das","publish_date":"2020-04-10T11:00:00","publication_year":"2020","word_count":1100,"keywords":["Go","programming_languages:R","AI","R","healthcare robotics","innovation","Robotics Process Automation","programming_languages:Go","Robotics","Robotic Process Automation","Aim","cancer","GAN","robotic surgeon","Robotics India"],"extracted_tech_keywords":["AI","Aim","R","Go","GAN","innovation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/case-study-how-this-ahmedabad-based-cancer-center-is-using-robotic-system-for-minimally-invasive-surgeries\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":36162,"title":"GIPHY Creates And Open Sources Celebrity Facial Recognition Algorithm","content":"GIPHY, one of the Internet’s biggest GIF sharing and creation platforms, recently open-sourced a model that allowed for the facial recognition of celebrities. The company reported that artificial intelligence could discern over 2,300 distinct faces of celebrities, with an accuracy of 98%. The database also includes the data of Indian celebrities like Shah Rukh Khan and Virat Kohli, among others. The model was trained based on GIFs submitted to the platform and was created with an intention of reproducing similar offerings which were blocked off by a paywall. GIPHY also revealed that a majority of the traffic across all of their sources came from celebrity searches. In order to make the process more efficient and provide a new way of searching on the platform, they created this algorithm. The Need For A Facial Recognition Algorithm GIPHY found the need to create this so that they could find and annotate the content on their platform. Since the platform has over 100 million daily active users, the throughput for GIFs being sent is said to be over 1 billion every day. Moreover, the platform also collaborates closely with multiple celebrities to create reactions, GIFs and stickers for their library. Along with the abundance of content regarding celebrities on the platform, GIPHY also makes promotional GIFs for movies. This adds to their already vast library of celebrity content, providing fodder for the AI to be trained on. The need for a new search method, along with the easy availability of labelled data made creating a model the natural way forward for the site to serve its queries. How The Model Was Trained To Pick Out Faces On a sample gif of Shah Rukh Khan and Salman Khan The first step was to organize the data on the platform, which was done by extracting all celebrity names from the top 50,000 searches on their platform. This resulted in a dataset of above 2,300 celebrities being returned, along with labelled data for determining who the celebrities were. This was the basement for the data involved in creating the model. However, some data was not as clean as they would have expected, especially for the representative images of non-popular celebrities. To tackle this, the team then grouped these images by similarity and measured the uniformity of each data set. This led them to find that the noisy data were similar in distribution to the clean ones, marking that it was safe to train the AI on it. Proceeding this, the image processing module of the model was split into two parts. The first was detected where the face was in the image or GIF. This was done through the use of a Multi-Task Cascade Convolutional Network or MTCCN. This can detect faces across frames of the GIFs. The detected faces are then captured and then sent to a deep convolutional neural network. This CNN is based on the Resnet-50 model and is trained on the dataset. This is where most of the recognition occurs, with the model being a facial features extractor. As the name suggests, the model identifies key facial features and creates a vector space of faces. These are grouped together by using centre loss, and then each face is given a celebrity prediction along with a unique vector. Post this step, a Gaussian Mixture Model is used to generate from the mixture of Gaussian distributions. In this use-case, it is used to cluster faces by vector representations. These clusters are used to compute an aggregate prediction for all the presented faces, along with a confidence score. Accuracy For An Integral Part Of Internet Culture GIPHY is one of the companies that prides itself on being an important part of Internet culture, as GIFs become more and more widely used across the net. They open-sourced this algorithm to contribute to this, even concluding their blog post by asking the users to come up with newer use cases or extend it for their own needs. This feature is also sure to increase the usability of the platform among Indian users, with celebrities such as Shah Rukh Khan, Salman Khan, Virat Kohli, and Imran Khan being a part of the roster picked by GIPHY to train this model. All in all, it presents a valuable addition to the site’s arsenal of tools to enable the rise of further Internet culture.","excerpt":"GIPHY, one of the Internet’s biggest GIF sharing and creation platforms, recently open-sourced a model that allowed for the facial recognition of celebrities. The company reported that artificial intelligence could discern over 2,300 distinct faces of celebrities, with an accuracy of 98%. The database also includes the data of Indian celebrities like Shah Rukh Khan […]","categories":["AI Features"],"tags":["artifical intelligence","Facial Recognition","Machine Learning","ML"],"author_name":"Anirudh VK","publish_date":"2019-03-12T07:22:30","publication_year":"2019","word_count":724,"keywords":["Go","artificial intelligence","Facial Recognition","programming_languages:R","AI","neural network","ML","Machine Learning","ResNet","programming_languages:Go","GAN","CNN","R","artifical intelligence"],"extracted_tech_keywords":["AI","artificial intelligence","neural network","R","Go","GAN","CNN","ResNet","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/giphy-creates-and-open-sources-celebrity-facial-recognition-algorithm\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10020578,"title":"NASA’s Perseverance Pays Off, Apple 6G And More In This Week’s Top News","content":"As the world tries to grapple with the implications of 5G, the race for 6G supremacy is already heating up. As first appeared on Bloomberg, Apple is already in plans to work on 6G technology. Last year, Apple unveiled its 5G iPhones. “This is a huge leap for the iPhone, bringing the best 5G experience in the market and delivering our most advanced technologies to users,” said Greg Joswiak of Apple during the launch. As users warm up to the new 5G phones, Apple is already working on 6G tech. According to reports, Apple this week posted job ads seeking wireless system research engineers for current and next-generation networks. The job postings listed “research and design next generation (6G) wireless communication systems for radio access networks” and “participate in industry\/academic forums passionate about 6G technology” as requirements.  6G devices and applications are expected to be heavy on personalisation and privacy without trading efficiency, which is really crucial as the world moves towards ML powered edge devices. Facebook Cuts Off News in Australia Great to talk to my good friend PM @narendramodi again. As Comprehensive Strategic Partners, we can work together on common challenges incl #COVID19, the circular economy, oceans & an open, secure and prosperous Indo-Pacific. We also discussed progress of our media platform bill. https:\/\/t.co\/fjAeLecCYA— Scott Morrison (@ScoMo30) February 19, 2021 On Thursday, Facebook blocked its users in Australia from sharing news stories and stopped users globally from sharing articles from Australian publishers. This move by Facebook comes as a response to the government’s push for a ruling that might require Facebook to pay to the publishers. “Facebook needs to think very carefully about what this means for its reputation and standing,” Communications Minister Paul Fletcher said in an interview. In a follow up to the news of FB trying to bully the Aussies, the EU, which has been mulling a similar legislation for a while, might now consider to copy parts of the proposed law in Australia that will force the tech giants to pay for snippets of articles shown on the platforms. However, Facebook argues that its initiatives are “fundamentally misunderstood”. “The proposed law fundamentally misunderstands the relationship between our platform and publishers who use it to share news content,” William Easton, MD for Facebook in Australia and New Zealand, wrote in a blog post. NASA Lands On Mars, Again! Hello, world. My first look at my forever home. #CountdownToMars pic.twitter.com\/dkM9jE9I6X— NASA's Perseverance Mars Rover (@NASAPersevere) February 18, 2021 On Thursday, NASA announced that its largest, most advanced rover yet has touched down on Mars after drifting in space for 203-days and nearly 472 million kilometers. Launched on July 30, 2020, the Perseverance rover is tasked with collecting rock samples from Mars and returning them to Earth. Shown above is the first image NASA’s Perseverance rover sent back after touching down on Mars. The rover, which is about the size of an SUV, is actually a robotic geologist and astrobiologist on a two-year science investigation of Mars’ Jezero Crater in search for signs of ancient microbial life. The $2.7bn (£1.9bn) robot is the fifth rover to be put on Mars by NASA. Since the rover successfully completed the harrowing entry, descent, and landing phase of the mission, it will now begin searching for traces of microscopic life from billions of years back using a precision X-ray device powered by artificial intelligence. Bad UI Costs Citibank $500 Million Source: ArsTechnica A poorly designed user interface has cost Citibank dearly. The interface of Flexcube(shown above), the software which Citi used to transfer interests to the creditors of Revlon was so confusing that the employees accidentally paid back the principal amount of $900 million instead of the $7.8 million interest they were supposed to send. Though few creditors returned back wrongly sent money, others refused; burning a $500 million hole in Citi’s pocket. Citi pursued the case legally but a federal judge has ruled that Citibank won’t’ be getting the money back. AWS Bags Daimler Daimler’s Torc Robotics said on Wednesday it has selected Amazon Web Services(AWS) to handle the petabytes of data that Torc’s test fleet in New Mexico is already generating. According to Torc, the raw data comes from multiple sensors such as lidar, radar and cameras. “In addition to the expansion of routes and fleet size, the next generation of test trucks has integrated more sensors at higher resolutions to enhance object detection at longer ranges — further increasing the magnitude of data for analysis, simulation, and machine learning,” said the company. Whereas, AWS which has been reigning supreme in the cloud markets will offer its extensive services that are designed to provide rapid, secure data transfer, intelligent tiered storage, managed orchestration and analytics tools, and high-performance multi-core CPU and GPU compute that will help Torc rapidly scale its agile and cost-efficient development platform and accelerate its testing and commercialization of the technology. “Our next generation fleet of test trucks will help us rapidly grow our capabilities and accelerate the commercialization of Level 4 self-driving trucks,” said Michael Fleming, Torc’s CEO. NVIDIA Launches Crypto Hardware NVIDIA wants gamers to game and miners to mine. So, they have decided to limit the hash rate of GeForce RTX 3060 GPUs so they’re less desirable to miners who have been on the rise thanks to crypto’s tremendous run. Furthermore, to satiate the needs of miners, NVIDIA has now launched NVIDIA CMP, or Cryptocurrency Mining Processor, for professional mining. Nvidia also stated that their CMP products don’t do graphics and they don’t meet the specifications required of a GeForce GPU and, thus, don’t impact the availability of GeForce GPUs to gamers. In other news, this week, Bitcoin hit a record $50,000 for the first time as the news of major players embracing it keeps coming! SpaceX Raises $850 M According to CNBC, SpaceX raised a handsome $850 million last week in the latest funding round, which rocketed the company’s valuation to about $74 billion; a 60% jump from its previous round in August. The funding news comes at a crucial moment as it will fuel Elon Musk’s highly ambitious Starship and Starlink projects.","excerpt":"As the world tries to grapple with the implications of 5G, the race for 6G supremacy is already heating up. As first appeared on Bloomberg, Apple is already in plans to work on 6G technology. Last year, Apple unveiled its 5G iPhones. “This is a huge leap for the iPhone, bringing the best 5G experience […]","categories":["AI News"],"tags":[],"author_name":"Ram Sagar","publish_date":"2021-02-21T10:00:00","publication_year":"2021","word_count":1022,"keywords":["machine learning","artificial intelligence","AWS","AI","ML","RAG","Ray","Aim","object detection","analytics"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","analytics","Aim","Ray","RAG","object detection","AWS"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/nasa-apple-6g-nvidia-crypto-top-tech-news\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10019718,"title":"This Dutch Startup Leverages IoT To Simplify Building Management","content":"A recent Deloitte report stated that the adoption of IoT and cloud technologies in India is still at a nascent stage, owing to many challenges the country faces in terms of the talent gap, connectivity, compatibility, interoperability, and cybersecurity, etc. In 2019 alone, India spent $20.6 billion on IoT hardware and software. In the initial stage of the pandemic, the focus was more on optimising cost and directing it towards resources based on priority. Adoption of IoT and related technologies will be greater in the second half of 2021, the report added. Founded in 2016, Netix is a subsidiary of the Netherlands-headquartered SB Group offering end-to-end IoT-enabled building management and smart-city solutions. For this week’s startup column, Analytics India Magazine spoke to Ganesh L Khanolkar, Director Operations at Netix India, to understand how the startup is leveraging emerging technologies to offer smart city solutions. Headquartered in Mumbai, Netix controls and maintains embedded assets such as air conditioning, security and emergency lighting systems, reducing the need for physical intervention. With over 50 leading clients across the country, the company is on track to hit its revenue target of $10 million in India by 2024. Products The company’s comprehensive industry expertise ranges from design, engineering, integration, connectivity, testing, commissioning support and maintenance support of smart building ecosystems, and their components. NETIX Controls software is an open framework that delivers a variety of improvements to help businesses take full advantage of the Internet of Things, including advanced visualisation and new search, security and navigation tools. NETIX Controls is customisable to suit the security policies of organisations. Khanolkar said, “We have built-in features that make it easier for the end-user configuring our software to enforce best practices. Our latest release includes a host of new features and functionality to support our customers as they deploy more third-party drivers and analytics tools to glean more insight from operational data.” He added that the company has an all-new user interface that is modern and easy to use. The platform utilises HTML5 to provide an array of rich features. Also, the new range of products at NETIX Controls include sensors and metering devices around IP, ethernet, WiFi, MQTT, wireless and IT security standards and more. What’s The Differentiator? Khanolkar said, “What sets us apart from our competitors is that we are brand agnostic and offer an open framework, a software that supports all protocols that’s available in the market. So the biggest advantage our customers get is that they don’t require an additional gateway, and can also integrate legacy BMS systems, enabling them to retain existing controllers along with Netix platforms and additional drives. “Also, our complete integrated BMS portfolio includes software, controllers, sensors, metering devices and thermostats all under the same roof. Our dashboard also provides key offerings such as IoT connectivity, cybersecurity, compliances required, real-time cloud-based dashboards, smart metering, predictive analysis, etc,” he added. Use of IoT & Other Technologies The company offers wireless IoT sensors to give building owners visibility into their property operations, such as health facility, equipment condition, waste management, security and fire safety, etc. According to Khanolkar, there are different levels of integration available depending on the various systems, services and applications. The NETIX Controls core is built around the IoT. The above image is a representation of a typical IoT solution, with the many layers required to get data from disparate systems and IoT devices to either a desktop User Interface (UI) or mobile device through the IoT layer stack. The NETIX Controls software supports device and data connectivity across all layers of the offering. Hiring @ Netix The company looks at the background, personality traits and skills while hiring employees. Khanolkar also mentioned important skill sets required in a recruit: Tech SavvyVulnerability assessment skills (Hardware, Software security & Data Security skills)Networking expertiseAnalytics & Data Science expertiseUnderstanding of OT & ITMobile App friendly A good learner Team ManLeadership & business acumen Future Roadmap On a concluding note, Khanolkar said, “NETIX will definitely revolutionise this domain with our state-of-the-art solutions. We have a team of professionals\/engineers exclusively focusing on smart\/intelligent building vertically.” He added, “We as NETIX will demonstrate and justify the costs and efforts involved. We have smart solutions ready for offices, warehouses, hospitality and retail spaces which constitute the major market potential for smart buildings in the coming years.”","excerpt":"A recent Deloitte report stated that the adoption of IoT and cloud technologies in India is still at a nascent stage, owing to many challenges the country faces in terms of the talent gap, connectivity, compatibility, interoperability, and cybersecurity, etc. In 2019 alone, India spent $20.6 billion on IoT hardware and software. In the initial […]","categories":["AI Startups"],"tags":["AI Startups"],"author_name":"Ambika Choudhury","publish_date":"2021-02-05T13:00:00","publication_year":"2021","word_count":720,"keywords":["data science","Go","AI","ML","RAG","Ray","ViT","analytics","GAN","R","AI Startups"],"extracted_tech_keywords":["AI","ML","data science","analytics","Ray","RAG","R","Go","GAN","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-startups\/this-dutch-startup-leverages-iot-to-simplify-building-management\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10089278,"title":"AGI is Lazy Person’s Aspiration","content":"AGI may be seen as the ultimate goal by some in the AI community, but Michael Irwin Jordan, professor at UC Berkeley, believes it to be a “lazy person’s aspiration”. In an exclusive interview with AIM, Jordan said that instead of creating a Frankenstein-like AGI, we should focus on developing systems that bring music and choice into people’s homes. Jordan argues that AGI is neither necessary nor sufficient for achieving such tasks. This statement is a strong rebuke to AI experts like OpenAI’s Sam Altman and Tesla’s Elon Musk, who champion AGI and believe that machines can surpass human intelligence. They naively believe that their current technology is the solution to humanity’s problems or the key to achieving AGI. On the contrary, Jordan is of the view that AGI is not the ultimate goal and that we should prioritise building impactful systems that benefit humanity. By shifting our focus away from the hype surrounding AGI, we can create technology that truly improves people’s lives. “Intelligent computers can help us network together the traffic control of aeroplanes. This is done because humans didn’t evolve to take pieces of metal flying through the air and make them safe. So why not let the computer help us with that test?” he pondered, saying that the community is not working at a planetary scale, connecting people in healthy ways to make sure they do not hurt each other. That to Jordan is way more interesting than AGI. Jordan thinks that happiness, democracy and welfare for humans is not a given. “If we’re not focusing on ensuring that everybody has opportunities, some happiness and resources then we’re missing the boat and I do not believe that simply creating AGI somehow solves those problems,” he said. Drawing a comparison between AI and electrical engineering, which had a positive impact on human beings, Jordan said, it had people thinking, we have these electrons that could create fireworks but we could also make houses warm, bring lights and use it to cook food. That’s how humans operate. Our skill set gets leveraged by new tools, the ability to do experiments and build devices instead of asking some AGI agent to solve those problems for us. Jordan has mentored over 150 students at MIT and UC Berkeley Michael Jordan is a renowned computer scientist and statistician who has made significant contributions to the fields of machine learning and artificial intelligence. Andrew Ng and David Blei, both prominent figures in the field of machine learning, have cited Michael Jordan as one of their mentors and influences in their work. Zoubin Ghahramani and Yoshua Bengio are also some of his notable students. “It’s really great to be an academic mentor. You get to be with people for five years while they learn and then take maybe a year to solve challenging problems. So the overall tree is very large. It’s been a wonderful part of my life to be part of that process,” Jordan gladly remarked. Jordan’s first love was statistics and analysing data in a believable way. “I saw how you could automate statistics and think about computers not just as things you program, but as things that can make models of the world, predictions and help us,” he said. Jordan entered the artificial intelligence realm around the era of search engines when the internet had just started to become powerful. To him, it was this democratic thing supported by algorithms which humans could use to spread knowledge worldwide. Currently, his students at UC Berkeley are working on things like uncertainty quantification of predictions. The 67-year-old professor is also very excited about work on learning-based economic mechanisms. There’s also work on an economic model for federated learning where the students are trying to bring data from many sources together. The vision of his students is to do so in a manner where the agent at the end of the link is being incentivized to be part of the system and can walk away if they don’t feel the incentives are strong enough. Enter ChatGPT Elaborating on the increase of usage of ChatGPT in businesses, he said people need to realise ChatGPT is an analysis of existing data. It is a huge network. You ping the probe that part of the network which at the end of the day is created by humans. It’s able to bring all that together in interesting but surprising ways. “Language models just being programmed to try to predict the next word is true, but it’s not the dunk some people think it is,” tweeted Gary Marcus, a leading voice in AI. He is not the only one who agrees with Jordan on large language models, AI stalwarts Yoshua Bengio and Yann LeCun have expressed similar views in their exclusive conversations with AIM. “I read that someone kept probing it and it went a little crazy. It is generating text based on past texts and cannot reach into the real world. Not just the physical world, but also the social world. It’s a pale shadow of that,” Jordan opined. Furthermore, he said, the wave of hysteria and hype around GPT-4, too, alarms him a bit. He believes that the focus of AI should be on the cooperative links between humans and machines, using technology to augment human capabilities rather than to mimic isolated human intelligence. “It’d be wise to think about what particularly will help humans the most. If you’re going to put people out of work, don’t do it so fast that it breaks the economic system. Do it in a staged way so people have time to realise what’s happening and make plans around that,” he suggested. Not Reasonable In 2018, Jordan wrote a Medium article, ‘Artificial Intelligence — The Revolution Hasn’t Happened Yet’ and believes it stands corrected today. There are aspects of pattern recognition and generative AI that have proceeded faster than he expected. But noting the limitations he said, “They are not systems that I would trust with making life and death decisions. They cannot reason yet,” he said. There are difficult aspects of human interaction such as having a deep appreciation of the semantic distinctions and being able to reason with those. In some ways, machines today are surprisingly good at mimicking, but mimicking is not the same thing as really being able to do it, Jordan believes. There’s a big difference between generating solutions involving reasoning, counterfactuals, interacting with others, and using extensive knowledge based on experience. Together with all that, Jordan does not think the goal should be to imitate humans. Explaining the difference, he said, “We have thousands of aeroplanes that fly around. They don’t hit each other because it’s all coordinated. The right level to be thinking about is not the individual entity that’s supposed to be super smart but to make networks smarter and make them augment human capabilities, not replace.” It’s more about trade and making lives a little bit better. But how do we make sure that those trade offs are a big part of the system? An ethical system is partly one that thinks about economics, a field that talks about interactions between humans, he said. “If you build a system that doesn’t work and people are depending on it, it’s unethical. Ethics is thinking about how to build a good system, one that brings real value to human beings. Then part of it is more philosophical and legal,” he added.","excerpt":"The hysteria and hype around GPT-4, too, alarms Michael Irwin Jordan a bit.","categories":["AI Features"],"tags":["AGI","Andrew Ng","Elon Musk","Interviews and Discussions","Sam Altman"],"author_name":"Tasmia Ansari","publish_date":"2023-03-13T18:00:00","publication_year":"2023","word_count":1239,"keywords":["federated learning","ChatGPT","Sam Altman","artificial intelligence","machine learning","AI","OpenAI","Andrew Ng","Elon Musk","RAG","Aim","AGI","generative AI","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","generative AI","ChatGPT","OpenAI","Aim","federated learning","RAG","R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/andrew-ngs-mentor-michael-irwin-jordan-thinks-agi-is-lazy-persons-aspiration\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":61724,"title":"A Beginner’s Guide To Using Google Colab","content":"We are all familiar with the pop-up alerts of ‘memory-error’ while trying to work with a large dataset of machine learning (ML) or deep learning algorithms on Jupyter notebooks. On top of that, owning a decent GPU from an existing cloud provider has remained out of bounds due to the financial investment it entails. The machines at our disposal, unfortunately, do not have unlimited computational ability. But the wait is finally over as we can now build large ML models without selling our properties. The credit goes to Google for launching the Colab – an online platform that allows anyone to train models with large datasets, absolutely free. From working with complex models to sharing the work with others in a simple manner, Colab is what the data science community has been longing for years. Colab should be the ideal choice for anyone looking forward to improving their Python coding skills with deep learning libraries such as PyTorch, Keras, TensorFlow, and OpenCV. One can create, upload, store and share notebooks in Colab, mount the Google Drive, import directories, and more. Since it is a whole new set of features put together, navigating could be a challenge. In this article, we will provide a simple and quick tutorial on how to go about Google Colab. Setting Up Drive To begin with, one can click on the link to get started with Colab. As the picture displays, one can create a new notebook with just one click, and then can upload a local notebook to Colab. Through using an authentication process, a user can import a notebook from Google drive or GitHub as well. In case a user is already working on Google drive, he or she can directly add a new Colab notebook by clicking on ‘new’ and dropping the menu down to ‘more’ and selecting ‘Colaboratory’. One can also easily rename the notebook. Choosing GPU Or TPU What makes Google Colab popular is the flexibility users get to change the runtime of their notebook. As shown in the picture, one can click the runtime menu and change its type. A user will get two options after clicking on ‘Change runtime type’. A box will appear on the screen with the option ‘hardware accelerator’. Users can either select TPU or GPU to work on their notebook. One can even make it better by mounting their Google drive. To do so, one needs to use: from google.colab import drive drive.mount(‘\/content\/gdrive’) It is advisable to shut the notebook since the step will allow others to use it as a valuable resource, and they can further share it with others. To terminate the notebook, one can follow these steps: Runtime > Factory Reset Runtime Using Terminal Codes Google Colab allows a user to run terminal codes, and most of the popular libraries are added as default on the platform. Libraries like Python’s Pandas, NumPy and Scikit-learn come pre-installed with Colab, and running them is a straightforward job. In case a user wants to run a different Python library, follow the step below: !pip install library_name The only catch is the usage of an exclamation mark (!) that must be put before entering each command. Colab also allows the user to import any library by running import like any other notebook. Although PyTorch is now supported on Colab, one can further use the code given below in case they face some complexities: # http:\/\/pytorch.org\/ from os.path import exists from wheel.pep425tags import get_abbr_impl, get_impl_ver, get_abi_tag platform = ‘{}{}-{}’.format(get_abbr_impl(), get_impl_ver(), get_abi_tag()) cuda_output = !ldconfig -p|grep cudart.so|sed -e ‘s\/.*\\.\\([0-9]*\\)\\.\\([0-9]*\\)$\/cu\\1\\2\/’ accelerator = cuda_output[0] if exists(‘\/dev\/nvidia0’) else ‘cpu’ !pip install -q http:\/\/download.pytorch.org\/whl\/{accelerator}\/torch-0.4.1-{platform}-linux_x86_64.whl torchvision import torch Importing From GitHub As mentioned above, one can import any file from Google Drive and GitHub. The importing function from GitHub is straightforward, as is seen in the picture. One can look for any organization or user to find the file. Once the file is found, they can enter the link to import the files. In case a user is not able to find the exact material, they have the option to go through the repository drop-down menu as shown in the picture below. Saving The most important part is saving on time. One can use the unconventional ‘command-s’ or drop the ‘file’ menu down to save. A user can further create a copy of the notebook by dropping ‘File’ -> ‘Save a Copy in Drive.’ One can also download the notebook  by going from ‘File’ -> ‘download .ipyb’ or ‘download .py.’We hope this article will enable readers to navigate Google Colab seamlessly and take advantage of the free GPU environment. One can also read our article on how Google Colaboratory Can Be Your Free GPU For Deep Learning for more information.","excerpt":"We are all familiar with the pop-up alerts of ‘memory-error’ while trying to work with a large dataset of machine learning (ML) or deep learning algorithms on Jupyter notebooks. On top of that, owning a decent GPU from an existing cloud provider has remained out of bounds due to the financial investment it entails. The […]","categories":["AI Trends"],"tags":["Google Colab"],"author_name":"Rohit Chatterjee","publish_date":"2020-04-14T19:00:06","publication_year":"2020","word_count":787,"keywords":["data science","scikit-learn","machine learning","Keras","AI","PyTorch","ML","deep learning","Jupyter","Google Colab","TensorFlow"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","data science","TensorFlow","PyTorch","Keras","scikit-learn","Jupyter"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/a-beginners-guide-to-using-google-colab\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10076221,"title":"Making Data Centres In India Sustainable, The AWS Way","content":"With a market share of around 34 percent, Amazon Web Services (AWS), is the most popular cloud services provider globally, ahead of competitors – Microsoft Azure and Google Cloud. In India too, AWS remains one of the most popular cloud service providers. Guru Bala, head of solutions architecture, AWS Specialised Services, AISPL, in an exclusive conversation with Analytics India Magazine, said that AWS works closely with organisations of all sizes, from enterprises to startups, to enable them to leverage the AWS AI\/ML stack to build solutions that solve challenges in the industry. Further, Bala discusses AWS’ service offerings, their plans in India and sustainability. What is AWS doing to make its data centre green, specifically in India? According to 451 Research, Indian companies and public sector organisations that migrated computing workloads from on-premises data centres to cloud infrastructure could expect to reduce their energy use – and associated carbon footprint – by nearly 80%. One of the most visible ways AWS is using innovation to improve power efficiency is our investment in AWS chips. Our third generation Arm-based processor AWS Graviton3, is more energy efficient, as Graviton3-based Elastic Cloud Compute instances use up to 60% less energy for the same performance than comparable Amazon EC2 instances. Inferentia is our most power-efficient inference processor. On many workloads, Inferentia is about twice as power-efficient when compared to GPU-based inference. For AWS, running our operations sustainably also means reducing the amount of water we use to cool our data centres. Our holistic approach minimises both energy and water consumption in our data centre operations and guides the development of our water use strategy for each AWS region—which starts with evaluating climate patterns, local water management and availability, and opportunities to avoid using drinking water sources. In collaboration with global nonprofits water.org and WaterAid, AWS has provided access to clean water for more than 250,000 Indian citizens in the states of Maharashtra, Telangana, and Andhra Pradesh, India, since 2020. AWS plans to accelerate its work around improving access to clean water, developing sustainable sanitation operations, and improving the water efficiency of its own operations, in India and around the world. As part of our goal to reach net-zero carbon by 2040, Amazon is on a path to powering its operations with 100% renewable energy by 2025—five years ahead of our original target of 2030. Amazon was the world’s largest corporate buyer of renewable energy and in 2021, we reached 85% renewable energy across our operations. Furthermore, we recently announced our first utility-scale renewable energy projects in India – three solar farms located in Rajasthan. These include a 210 MW project to be developed by India-based developer ReNew Power, a 100MW project to be developed by local developer Amp Energy India, and a 110MW project to be developed by Brookfield Renewable. Combined, these solar farms have the capacity to generate 1,076,000 MW hours of renewable energy per year, enough to power over four million average-sized households in New Delhi annually. Tell us about your plans in India and how do you perceive the Indian market? Our mission is to empower builders and businesses to build a better India and we want to be a force that moves India forward. We are committed to investing in India and putting the power of AWS Cloud directly in the hands of customers, including enterprises, startups, small and medium businesses, or developers. We continue to invest in India for the long-term. In November 2020, we announced the expansion of our services and are planning to include a second infrastructure region in India by 2022. The new AWS Asia Pacific (Hyderabad) Region will consist of three Availability Zones at launch. This will provide customers with more flexibility and choice, while allowing them to architect their infrastructure for even greater fault tolerance, resiliency, and availability across geographic locations. Tell us a bit about the AI and machine learning services from AWS? AWS offers the broadest and deepest set of machine learning services and supports cloud infrastructure, putting machine learning in the hands of every developer, data scientist and expert practitioner. We work closely with organisations of all sizes, from enterprises to startups, to enable them to leverage the AWS AI\/ML stack to build solutions that solve challenges in the industry. More customers prefer to use our machine learning services due to our broad portfolio of services at the three layers of the technology stack. At the bottom of the stack, we offer Inf1, TRN1 instances, along with P3, P4 and G5 instances. At the middle layer of the stack, organisations which want to use machine learning in an expansive way can leverage Amazon SageMaker, a fully-managed service that removes the heavy lifting, complexity, and guesswork from each step of the ML process, empowering everyday developers and scientists to successfully use ML. At the top layer of the stack, we provide solutions, such as Amazon Rekognition to extract information and insights from your images and videos, Amazon Polly that turns text into lifelike speech, Amazon Lex for building conversations, and Amazon Transcribe for converting speech to text. Then there is Amazon Translate for translating text between languages. Apart from this, our customers also collaborate with the Amazon Machine Learning Solutions Lab to implement real-life use cases. How AWS services like Amazon Comprehend, Amazon Textract and Amazon Rekognition are pioneering interesting innovations? Can you share some use-cases in this regard? Amazon Rekognition offers pre-trained and customisable computer vision (CV) capabilities to extract information and insights from your images and videos. Some of the interesting use cases of Amazon Rekognition are detecting inappropriate content, verifying identity online, streamlining media analysis and sending connected home smart alerts. For example,  WeShine Tech—a startup focused on university examination process automation, built a facial recognition service using Amazon Rekognition within a week to allow for remote exam proctoring during the lockdowns. Another example is WorkApps that was able to verify customer identities in near real-time for online KYCs during the pandemic. Amazon Comprehend is an NLP service that uses machine learning to uncover valuable insights and connections in text. With this, customers can mine business and call centre analytics, index and search product reviews, automate the extraction of insights from packets of legal briefs such as contracts and court records, and process financial documents. An interesting example of this is FINRA in the US, which is a not-for-profit organisation dedicated to investor protection and market integrity. FINRA receives millions of documents with unstructured data to support investigative, examination, and compliance processes. Their investigators and examiners had to manually go through documents page by page. With Amazon Comprehend, the company was able to quickly extract individuals and organisation data, and do much more. Lastly, Amazon Textract automatically extracts text, handwriting, and data from scanned documents. It goes beyond simple optical character recognition (OCR) to identify, understand, and extract data from forms and tables. Today, many companies manually extract data from scanned documents such as PDFs, images, tables, and forms, or through simple OCR software that requires manual configuration. To overcome these manual and expensive processes, Textract uses ML to read and process any type of document, accurately extracting text, handwriting, tables, and other data with no manual effort. For example, India’s leading fintech startup Paytm used Amazon Textract to extract user data from images of complex identity documents with 97% accuracy.","excerpt":"AWS plans to accelerate its work with Water.org and WaterAid to improve the water efficiency of its operations in India","categories":["Global Tech"],"tags":[],"author_name":"Pritam Bordoloi","publish_date":"2022-10-03T10:00:00","publication_year":"2022","word_count":1222,"keywords":["Amazon SageMaker","machine learning","AWS","AI","ML","computer vision","RAG","NLP","analytics","Azure"],"extracted_tech_keywords":["AI","machine learning","ML","NLP","computer vision","analytics","Amazon SageMaker","RAG","AWS","Azure"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/making-data-centres-in-india-sustainable-the-aws-way\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10070987,"title":"Chennai based AI startup Ganit raises funds in pre-series A round","content":"Recently, AI and data analytics startup Ganit said it had raised funds in pre-series A round led by Sangeet Kumar, co-Founder & CEO of Addverb Technologies; Krishnan Vishwanathan, Co-Founder & CEO, Kissht; Anshul Gupta & Amit Raj, co-founders of EatClub Brands (formerly Box8), among others. Founded by Shivaprasad KT, Ashok Harwani and Hariharan R in 2017, the company expects to double its headcount to 500 by 2023-end. The funds will be used to expand Ganit’s product portfolio in various spaces, including customer opinions, forecasting, promotion planning, route optimisation, price optimisation, and areas such as ESG (Environmental, Social and Governance). Ganit designs and deploys purpose-built AI\/ML solutions. The company aims to solve diverse business problems through practical solutions, minimising decision risk across industries, and helping organisational leaders utilise their data across the board. The company works with Fortune 1000 clients across several domains, including BFSI (banking, financial services and insurance) industries across several geographies. “The world needs data which will just not sit in silos for various reports but will be turned into actual attainable action. Ganit has cracked the code on how to help enterprises achieve just that. We are happy to be part of this story and wish the team great success”, Amit Raj, Angel Investor and Co-founder EatClub Brands (formerly Box8) said.","excerpt":"The amount of funds raised was not disclosed","categories":["AI News"],"tags":["AI Startups","Chennai"],"author_name":"Tasmia Ansari","publish_date":"2022-07-14T15:33:46","publication_year":"2022","word_count":214,"keywords":["Go","programming_languages:R","AI","ML","programming_languages:Go","Aim","Chennai","analytics","GAN","R","AI Startups","startup"],"extracted_tech_keywords":["AI","ML","analytics","Aim","R","Go","GAN","startup","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/chennai-based-ai-startup-ganit-raises-funds-in-pre-series-a-round\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":37215,"title":"How AI-Powered Tools Can Shape Your Resume","content":"Image source: Boutique Recruiting The art of writing a concise and to-the-point resume is a talent that many of us haven’t mastered yet. Many a time, our key achievements are lost in long essays and in many cases, our recruiters are least impressed. Here is where the usability of artificial intelligence comes into play. As a tool which can analyse through scores of content on the web and making valuable suggestions accordingly, more online platforms are now looking at AI for building the perfect resume. In this article, we take a look at how AI is making this possible with the help of these powerful tools: Resume Assistant: There are a number of AI-powered online resume makers and big names such as Microsoft and LinkedIn has partnered to release their own exclusive tools in 2018. LinkedIn’s AI-powered Resume Assistant was added to Microsoft’s Office 365 suite and uses the network’s professional pool to shape your resume. By leveraging their millions of data, the new tool will go through millions of data to look for similar roles or aspirational roles to make a suggestion to you regarding your resume. Further, the tool will help its users to take guidance from professional freelance resume experts through its other tool, Profinder, thus helping its users with interview techniques, career coaching, and resume writing. “With over 15 million job applications being submitted on LinkedIn every week, finding the right way to represent your unique experience is important. By regularly updating your resume, you’ll already be one step ahead when the time comes to find your next role,” LinkedIn wrote on a blog. Mosaic.ai: Is an AI career agent which can help its users plan your career and review the best skill and team culture matched opportunities. It also helps potential job seekers build resumes, discover the keywords, concepts, and topics you may have overlooked that you need in your resume to get interviews. Its AI-powered solutions will help a person’s resume to have more visibility by matching the best job based on your skill and company. This is achieved with the help of NLP, which goes through certain keywords and suggest you the best job options based on the keyword identification, by doing so the company claims that it saves a person’s time by not having to go through several resumes. The platform also uses culture match, where the agent creates an AI model which will help you identify the best-suited job for you and the jobs where you have a better chance of clearing an interview. In addition, there will soon be made available through platforms like Slack, Facebook, Google, Alexa, and other messaging services so that the users can ask questions. Skillroads: Lets its users have a quick interview with its AI-programme and get all high priority skills defined and offers you professionally written and formatted results through its resume building process under 30 minutes. It’s AI algorithms analyse background, get all your high priority skills defined, and, as a result. The AI builder works by entering people’s preferred jobs, then by taking their questionnaire about the users’ experience, the system would identify their strength and skills that are best suited for its users. Apart from common features like AI resume generator, smart resume review and cover letter generator, what is unique to Skill Road is that it lets its users about job openings in Fortune 500 companies, identify the position that is suitable for you among these vacancies and polish your resume according to these positions. Textkernel: The platform is powered by the latest developments in AI, Deep Learning, Semantic Search and Machine Learning. Apart from providing recruitment and staffing agencies, through its multilingual CV parsing, it eliminates manual data entry and allows candidates to apply via any device and enables better search results. At the core of the platform, fundamental to its technology is resume and vacancy parsing models which are capable of extracting the most important from unstructured text. “In our new generation of resume and vacancy parsing product, the traditional statistical parsing models are replaced with Deep Learning neural network models. The new models have achieved remarkable improvements over different languages, brought better generalisation to new data and new domains, and reduced the need for manual feature engineering,” the company wrote in a blog post.","excerpt":"The art of writing a concise and to-the-point resume is a talent that many of us haven’t mastered yet. Many a time, our key achievements are lost in long essays and in many cases, our recruiters are least impressed. Here is where the usability of artificial intelligence comes into play. As a tool which can […]","categories":["AI Features"],"tags":["AI-Powered"],"author_name":"Akshaya Asokan","publish_date":"2019-04-02T12:54:47","publication_year":"2019","word_count":717,"keywords":["AI-Powered","semantic search","machine learning","artificial intelligence","AI","neural network","RAG","NLP","Aim","deep learning","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","deep learning","neural network","NLP","Aim","RAG","semantic search","R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-ai-powered-tools-can-shape-your-resume\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10091588,"title":"Workshop Alert: Accelerating Deep Learning Inference Workloads at Scale","content":"Deep learning is having somewhat of an iPhone moment with the launch of ChatGPT, and enterprises are eager to jump on board with the generative AI trend. However, to build a solid product on new AI technologies, developers need to stay up-to-date with the latest advancements in the field as well as frequently brush up on the basics of deep learning. To upskill hungry software engineers, NVIDIA, along with Analytics India Magazine is holding a webinar on the 25th of April from 3-4 PM. This webinar will tackle the subject of accelerating deep learning inference workloads at scale —— a must-know set of skills for developers in the generative AI boom. The webinar will also have a Q&A section hosted on the AI Forum, a groundbreaking community for AI developers and data scientists that aims to empower AI professionals across India. Who Should Attend? Programmers looking to learn about how to scale AI workloads efficiently Software developers building AI products for their companies AI engineers looking to interact with India’s growing data science community Enthusiasts looking to brush up on the basics of AI inference in the cloud Professionals looking to upskill themselves and accelerate their AI workloads Apart from tackling scalable inference across multiple GPUs, the webinar will also delve into other topics like model orchestration and management, large language model inference, and optimal model configuration. The session will also delve deeper into the technicalities of how to interface with the Triton inference server, especially by using the FIL backend, along with the potential benefits of dynamic batching on performance. Date and Time: 25th April | 3:00 – 4:00 PM Register Now on Discord About the webinar This webinar will illustrate how to accelerate and streamline AI inference workloads on any framework using Triton. NVIDIA’s open-source Triton Inference Server software has the potential to supercharge deep learning workloads, enabling developers to deliver high-performance inference across a multitude of cloud architectures. Whether it’s cloud service providers, on-premises servers or edge and embedded devices, Triton allows inference workloads to scale according to the available compute. Over the course of the webinar, the speaker will provide an in-depth tutorial on Triton’s capabilities, as well as some example deployments to give a better idea of its capabilities. This webinar will also include a question and answer session hosted on the AI Forum for India, powered by NVIDIA. The AI Forum is a groundbreaking community for AI developers and data scientists hosted on Discord that aims to empower AI professionals across India. Apart from interaction in this webinar, the AI forum is a one-stop spot for any AI enthusiast or professional to interact with like-minded individuals and explore the world of artificial intelligence, machine learning and data analytics. What to expect? Illustration on how to accelerate and streamline AI inference workloads on any framework How NVIDIA Triton allows developers to deliver high-performance inference across cloud, on-prem, edge, and embedded devices 10,000-feet overview of industrial adoption of neural networks – From GPT-4 to Transformer Inferencing Server Platform: Triton Server Compiler Optimization: TensorRT Hands-on & Demos QnA Date and Time: 25th April | 3:00 – 4:00 PM Register Now on Discord About the speaker: This webinar is conducted by Megh Makwana, Manager Solution Architect, Applied Deep Learning at NVIDIA. He has 6 years of experience in the machine learning and artificial intelligence space, with research work in AI workload optimisation. His session is sure to provide some insight into how deep learning workloads can be scaled efficiently. With his 2 years of experience at NVIDIA, a global leader in distributed computing and AI accelerators, he will provide some unique insights into how to scale up large deep learning workloads. Using the skills in this webinar, developers will be able to leverage NVIDIA’s new open-source Triton Inference Server software to accelerate deep learning workloads to the next level. Sign up for the webinar and join the Discord server to make sure you don’t miss this informative webinar. The webinar will give you a chance to join the AI revolution and embrace the future with our AI Forum on Discord. Register Now on Discord","excerpt":"Learn how to scale and accelerate deep learning workloads with NVIDIA’s comprehensive tech stack.","categories":["AI Features"],"tags":["Deep Learning","Deep Learning Models","deep learning projects","Deep Learning Techniques","NVIDIA"],"author_name":"Anirudh VK","publish_date":"2023-04-17T17:15:00","publication_year":"2023","word_count":686,"keywords":["data science","ChatGPT","artificial intelligence","machine learning","AI","Deep Learning Models","neural network","ML","deep learning","analytics","Deep Learning Techniques","deep learning projects","generative AI","Deep Learning","NVIDIA"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","neural network","data science","analytics","generative AI","ChatGPT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/workshop-alert-accelerating-deep-learning-inference-workloads-at-scale\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10013213,"title":"How Is TCS Helping With COVID-19 Testing In India","content":"COVID-19 cases have only been on the rise. With the non-availability of effective drugs and vaccines, one of the effective ways to control it is to detect it early in patients. However, the task is easier said than done. While a large number of test kits are being produced, they are not enough to conduct testing in large numbers. Government-run body, C-CAMP or Centre for Cellular and Molecular Platform, has been a key enabler in driving COVID-19 testing as it has been aggressively building, managing and scaling the ecosystem of MSMEs to produce test kits indigenously. However, they might not be enough. To scale up this production of test kits to a million test kits a day, TCS has come up with a solution to support C-CAMP in achieving that. In a recent announcement, TCS said that it has deployed a blockchain-based digital supply chain platform to drive the ambitious new project launched by C-CAMP. The Ambitious InDx Project By C-CAMP Back in September, the government announced its plan to double its testing capacity through a project called Indigenisation of Diagnostics (InDx), which has been anchored at the Centre for Cellular and Molecular Platforms (C-CAMP). The idea is to indigenously manufacture all the reagents needed for the molecular methods to diagnose COVID-19, including RT-qPCR in bulk quantities and at much lower costs. Supported financially by the Rockefeller Foundation, the project aims at improving access to COVID-19 diagnostics across the country. As a part of the project, the institution aims to build a robust supply-chain network of Indian MSMEs capable of producing reagents for testing kits and manufacturing testing kits. In doing so, it aims to identify the bottlenecks in the supply chain network, short-falls in the quality levels and any gaps in the ability of these MSMEs to scale up. It would handhold the MSMEs to meet both quality and quantity requirements of the test kits. The project will not only address the COVID-19 crisis but also help MSMEs to expand their business opportunities and to improve the overall healthcare system by developing high quality and low-cost molecular diagnostics. With support from the Rockefeller Foundation, C-CAMP is hopeful of achieving the target of one million kits per day. Prof Satyajit Mayor, Director of NCBS, Bangalore, said that with this project, they aim to include other new molecular diagnostic methods such as Lateral Flow Assays for COVID-19 testing apart from scaling the saliva-based sampling tests. It would not only help MSMEs to serve the Indian market but export to other needy countries at competitive prices, he added. He is an advisor to the project by C-CAMP. Apart from this, C-CAMP has also been driving other initiatives such as COVID-19 Innovations Deployment Accelerator or C-CIDA, inviting innovations from across the globe to fight back the raging pandemic. As a part of it, C-CAMP has even identified 31 near deployment-ready COVID-19 innovations such as rapid diagnostic kits, assisted respiratory devices, AI\/ML-based pre-screening and more. How Is TCS Driving This Ambition To help C-CAMP sail through its ambition of indigenously producing COVID-19 test kits, TCS has partnered with them to develop and deploy a digital supply chain platform to manage the MSME ecosystem. It is powered by the TCS Data Marketplace solution which allows organisations to embrace new ecosystem-based business models while democratising data and monetising it to create value. Data Marketplace framework ensures seamless ingestion, curation, classification, cataloguing, and distribution of data. It enables self-help mechanisms for business users to search and discover data, perform analytics, and visualise it in insightful ways. Further, it facilitates standardised, controlled data exchanges across ecosystem participants with the necessary security and privacy protection. It also uses blockchain at the back-end to ensure the immutability of audit logs. As explained by the company, C-CAMP has harnessed this capability in the InDx project to build a supply-chain platform that aggregates test kit supply data of tier 2 and 3 manufacturers of kits, enzymes, primers and antibodies. It also provides ready visibility to tier 1 suppliers and C-CAMP on supplier-specific quality levels, capacities and inventories across the ecosystem. It will therefore enable C-CAMP to keep a check and respond to the demand for those kits from the government. Dinanath Kholkar, Global Head, Analytics and Insights at TCS said that they are delighted to partner with C-CAMP to scale up India’s COVID testing capacity. He believes that it will be a major step in making COVID-19 testing accessible to every Indian. TCS believes that their data marketplace solution will anchor the InDx’s supplier ecosystem of MSMEs by democratising supplier data, enhancing supply-chain visibility and driving superior outcomes for all stakeholders. “The success of this initiative sets an example for how ecosystem-based partnerships can be put to work for other mission-mode programs of national importance,” Kholkar further added. TCS Has Been At The Forefront To Contain COVID-19 Despite initial hiccups caused by the pandemic, TCS was able to resume to pay hikes and ramp-up hiring as the business started to recover post-COVID crisis. Not just on the hiring front, but the company is also ensuring safe workplaces as the offices are resuming post the lockdown. It announced the launch of AI-powered IUX for Workplace Resilience, which is a command centre solution that helps enterprises make it safer for employees and customers to resume work and business. The solution ensures that strict mandates are followed apart from PPE and social distancing to ensure workforce safety, regulatory support, operational resilience, and customer engagement. It uses images and video to generate insights and maintain optimal in-house and remote staff levels or suggest ways to ensure social distancing. “With its integrated analytics platform and business command centre, IUX for Workplace Resilience enables them to capture and analyse disparate enterprise and IoT data so they can understand safety and business risks in real-time to maintain business continuity,” said Ashvini Saxena, Global Head, Digital Software & Solutions, TCS in an official statement. The company had also set up around 11 first-line COVID-19 isolation centres within its premises in various cities to provide medical support to associates and their dependents. TCS has made it their priority to safeguard the health and well-being of employees, which has now further extended to ensuring the safety of the citizens, country-wide.","excerpt":"COVID-19 cases have only been on the rise. With the non-availability of effective drugs and vaccines, one of the effective ways to control it is to detect it early in patients. However, the task is easier said than done. While a large number of test kits are being produced, they are not enough to conduct […]","categories":["IT Services"],"tags":["covid-19","TCS"],"author_name":"Srishti Deoras","publish_date":"2020-12-05T13:00:46","publication_year":"2020","word_count":1038,"keywords":["Go","API","covid-19","AI","R","ML","Git","RAG","Aim","analytics","GAN","TCS"],"extracted_tech_keywords":["AI","ML","analytics","Aim","RAG","R","Go","Git","API","GAN"],"url":"https:\/\/analyticsindiamag.com\/it-services\/how-is-tcs-helping-with-covid-19-testing-in-india\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10105058,"title":"Accenture Launches Generative AI Studio in Bengaluru to Accelerate AI Adoption","content":"Snippet: The studio will leverage the strategic investments made in Accenture’s Center for Advanced AI on behalf of its clients. IT consulting giant Accenture has launched Generative AI Studio in Bengaluru, India today to accelerate AI and analytics adoption. The company looks to co-create solutions with a modern data and AI foundation, LLM architecture, ecosystems relationships, talent, and responsible AI frameworks through the studio, which is a part of Accenture’s  $3 billion investment  in AI and analytics. This decision to open a Generative AI Studio comes in the backdrop of a recent study conducted by Accenture where it noted that 74 percent of C-suite executives aim to raise their AI-related investment in 2024, up from 50% the year before. The studio will leverage the strategic investments made in Accenture’s Center for Advanced AI on behalf of its clients. It will also leverage the company’s more than 1,450 AI patents, both issued and pending, as well as the knowledge gained from more than 300 ongoing generative AI initiatives. Accenture’s Big Bets on Generative AI The studio will provide the recently introduced services, including modification methods, model managed services, and specialized training courses, along with its own generation AI model “switchboard.” Teaming up with ServiceNow and NVIDIA, the company jointly announced the introduction of AI Lighthouse in July, 2023 to fasten the development and implementation of enterprise generative AI capabilities. Back in September, 2023 Accenture announced Writer, a generative AI platform for clients for content creation and drive significant business impact.","excerpt":"IT consulting giant Accenture has launched Generative AI Studio in Bengaluru, India today to accelerate AI and analytics adoption.","categories":["AI News"],"tags":[],"author_name":"Arya Vishwakarma","publish_date":"2023-12-18T12:39:00","publication_year":"2023","word_count":249,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","RAG","responsible AI","Aim","generative AI","analytics","R"],"extracted_tech_keywords":["AI","analytics","generative AI","Aim","RAG","R","Go","responsible AI","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/it-consulting-giant-accenture-has-launched-generative-ai-studio-in-bengaluru-india-today-to-accelerate-ai-and-analytics-adoption\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10040214,"title":"Why Is Overfitting So Demonized?","content":"A machine learning model is only as good as the data it’s trained on. In other words, the poor performance of a model is mainly due to overfitting and underfitting. Overfitting happens when the model is modelled ‘too well’ on the training data. Underfitting refers to a model that can neither model the training data nor generalize to new data. Overfitting could be an upshot of an ML expert’s effort to make the model ‘too accurate’. In overfitting, the model learns the details and the noise in the training data to such an extent that it dents the performance. The model picks up the noise and random fluctuations in the training data and learns it as a concept. However, in a real-time situation, these concepts do not apply to new data, which poses a great hindrance to the model’s ability to generalize. Overfitting is a common occurrence with nonparametric and nonlinear models that offer flexibility when learning a target function. So, what’s wrong with overfitting? An overfitted model is designed to accommodate the eccentricities of a particular dataset; this produces good results from the given dataset but will fail to generalize on other data. In many cases, these parameters are defined as the ‘non-data’ aspect of the training dataset, which is nothing more than an oddity that might not repeat with other real-time datasets. “Due to overfitting, a model will not be able to achieve generalization, which means the model(s) trained on one dataset has limited applicability on unseen data in real situations. This, in turn, reduces the ability of the model to discriminate patterns and predict correct results and can lead to drastic consequences especially in domains like healthcare, automotive and industrial,” Sindhu Ramachandran, Practice Leader – Deep Learning & Analytics, QuEST Global. Moreover, since the model is so closely trained on a given training dataset, the model may replicate its inherent biases against community, gender, and race. Many mistakenly believe overfitting is a problem that can be solved simply by widening the scope or flexibility of model design by training it on a wider set of parameters and adding more features. Though simple, this is not an effective resolution to the problem. Famed mathematician John von Neumann once said, “With four parameters I can fit an elephant, and with five I can make him wiggle his trunk.” Neumann meant that one should not be impressed by a complex model that fits data. It only means that with enough parameters, a model can be made to fit with any data set. Dr. Siba Panda, Assistant Professor, SVKM’s NMIMS Mukesh Patel School of Technology Management & Engineering, said, “If your aim is to build an overfitting model, you could take more features into consideration or build a complex model. There would be two outcomes: First, the significant additional features may cause the processing to be too costly; second, a complex model will deliver the overfitting results in the training data set but not in the test data set. So always go for a simple path, step by step, using different techniques, such as hold-out, data argumentation, feature selection and regularization, if you have an overfitting model. To avoid the overfitting problems while building models to solve real-life issues faced by industries, there are two concepts that students should pay attention to: normalization\/standardization and regularisation.” But, is it always bad? In a discussion on trends to watch out for in computer vision, both Vahan Petrosyan, CTO at Sweden-based SuperAnnotate AI and Loius Coopey, VC at Point Nine Land, agreed overfitting might have useful applications in computer vision in future. Taking the example of traffic management analysis, the duo wrote: “If your camera is fixed, you would rather build slightly different models for each camera and overfit your CV model for each camera. Although overfitting in machine learning has negative connotations, in such cases you would create far more accurate models for each fixed camera, rather than building a generic model for all cameras.”","excerpt":"A machine learning model is only as good as the data it’s trained on. In other words, the poor performance of a model is mainly due to overfitting and underfitting. Overfitting happens when the model is modelled ‘too well’ on the training data. Underfitting refers to a model that can neither model the training data […]","categories":["AI Features"],"tags":["machine learning model","normalisation","overfitting","underfitting"],"author_name":"Shraddha Goled","publish_date":"2021-05-17T11:00:00","publication_year":"2021","word_count":662,"keywords":["Replicate","Go","machine learning","AI","R","ML","computer vision","normalisation","Aim","machine learning model","deep learning","analytics","overfitting","underfitting"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","computer vision","analytics","Aim","R","Go","Replicate"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/why-is-overfitting-so-demonized\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10170606,"title":"Cursor Accidentally Blocks Users While Fighting Abuse","content":"According to community forums and developer discussions, several developers and users who are attempting to access Cursor are met with a sudden block message, “Your request has been blocked as our system has detected suspicious activity from your account.” The issue, which appears to have affected users across both free and paid tiers, triggered widespread confusion across the developer community. AIM was able to independently verify the error on Thursday. When users sought assistance, Cursor’s AI support pointed to potential VPN usage as a cause and suggested steps like creating a new account via Google or GitHub, and subscribing to Cursor Pro to try again later. These suggestions proved unhelpful for most. Subsequent queries received an automated escalation: “I’m connecting you with a teammate who can help review this further.” Reports on Cursor’s forums indicate the problem was widespread, impacting users regardless of their subscription level or the model in use. One developer noted, “Claude 3.7 works perfectly in one window, but fails in the other. If I switch to 3.5 or automatic, it works fine. It makes no sense.” A frustrated developer commented on the forum, “Remove the message check altogether soon, otherwise my team of 80 people each with a pro subscription will leave your cursor for better analogues”. Others confirmed they were not using VPNs at all, contradicting Cursor’s initial assumption. Eventually, a community developer from Cursor addressed the situation, attributing the blocks to an overzealous anti-fraud system activated in the past 24 to 48 hours. “For a short while, this detection was overly sensitive, causing some false positives,” they said. “We have now tuned this down, so you should no longer be seeing any issues.” This system was most likely activated in response to thousands of students trying to subscribe to Cursor Pro for free. While the situation appears to be stabilising, AIM was able to make it work, some users in the forum still report the issue happening. AIM also reached out to Cursor for a comment on the situation, but did not get a response yet. With developers cancelling Cursor subscriptions and switching to something else, an incident like this could make it worse for the company. It also highlights the trade-offs companies face when scaling platform security to prevent abuse.","excerpt":"“Your request has been blocked as our system has detected suspicious activity from your account”","categories":["AI News"],"tags":["cursor"],"author_name":"Ankush Das","publish_date":"2025-05-23T15:13:13","publication_year":"2025","word_count":377,"keywords":["Go","cursor","AI","Scala","Git","llm_models:Claude","Aim","ViT","Rust","GitHub","R"],"extracted_tech_keywords":["AI","Aim","R","Go","Rust","Scala","Git","GitHub","ViT","llm_models:Claude"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/cursor-accidentally-blocks-users-while-fighting-abuse\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":46736,"title":"Behind The Code With AWS Community Heroes","content":"In the developer series Behind The Code, we reach out to the developers from the community to gain insights on how their journey started in the field of data science, what tools and skillset they use and what’s essential for their work. For this week’s column, we caught up with the AWS Community Heroes. This July we attended the AWS Community Day 2019 where we caught up with these heroes to gain insights on the latest trends and toolkits in the DevOps world. The AWS Heroes are the mentors and super users, creators, writers or speakers who routinely provide high quality and impactful content to the AWS developer community worldwide. Bhuvaneswari Subramani Bhuvaneswari Subramani is the Director of Engineering Operations at Infor. With more than two decades of experience and specialised in the IT industry, Subramani has been working in CI\/CD integrations and on-premise cloud for many years now. Tell Us About The Evolution of Technology There has been a tremendous change in the field of technology such as serverless technology. A few decades earlier, it was just an impossible idea and now everything is going serverless. Everybody talks about serverless these days and this is the main impact. Essential Toolkit For an engineer, one can attach to a specific tool and technology. A software engineer needs to be a lifelong learner. The tools keep changing with time and one needs to adapt and change it as needed. Besides that, if anybody wants to be in the programming domain, s\/he should be strong in languages like PERL, Python, etc. One must always get updated with the latest tool and technologies. Advise For Millennials Subramani suggested that if someone wants to take up specific areas such as DevOps or CI\/CD, s\/he should delve deep into that. According to her, the best way to learn is to train someone else otherwise write about it. Gaurav Kamboj Gaurav Kamboj is the Cloud Architect at Hotstar. Kamboj is from the infrastructure background who handles mostly cloud infrastructures. Essential Toolkit Whether it be a backend or a frontend engineer, nowadays everyone must use the infrastructure as the core technology. Tools like Hyphen, Lambda, etc. in the AWS cloud, and Terraform, Python is essential for developers. Why AWS and Not Others like GCP or Azure To this question, Kamboj replied that when he started applying cloud, the other networks were not much matured like AWS. The availability of services was less compared to AWS. Sathyajith Bhat Sathyajith Bhat is the DevOps Engineer at Adobe and is one of the AWS Community Heroes. Essential Toolkit According to Bhat, React is one of the trending toolkits in the market. He has been using React with AWS Amplify for a few years now in the frontend and on the backend side, Node.js, Java, etc. has been trending these days. Language like Java seems like a bit old but it has a lot of frameworks which plays crucial roles while on the DevOps side, Kubernetes is the most trending toolkit. Roadmap of Technologies According to Bhat, people will focus more to add value to their businesses by the way of infra being set up by themselves or by the servers, adopt new technologies along with the traditional ones, and other such. Advise For Millennials One must keep learning in every way they can. Nilesh Vaghela Nilesh Vaghela is the founder of Electromech Corporation Cloud and Opensource Company. He is more of a system administrator and has been working in the cloud for a few years now. Tell Us About The Evolution of Technology To this question, he replied that whatever is new today, tomorrow it will be normal. It is important to understand the new technologies and where these technologies are fitted into their specific verticals. Advise For Millennials According to him, the first thing a developer has to learn microservices and understand how it works and also, how to build a microservice. Essential Toolkit Vaghela has been utilising AWS toolkits like Elastic Container Service (ECS) and Elastic Kubernetes Service (EKS) platforms which provide robust platforms where microservices can be built onto that. Secondly, one of the key tools can be said as serverless architecture. Components like API Gateway, Lambda, Cognito. Etc. are the important ones for serverless architecture. Jeevan Dongre Jeevan Dongre is the Engineering Manager at Nutanix who is also the co-founder of AWS User Group Bengaluru. The benefit of AWS User Group & Community Heroes According to Dongre, the benefit is basically networking and peer-to-peer learning. Knowledge keeps flowing unparalleled all through it. Essential Toolkit For a developer, it is important to know multiple languages, polyglot, understand multiple platforms, hybrid-cloud, Kubernetes, etc.","excerpt":"In the developer series Behind The Code, we reach out to the developers from the community to gain insights on how their journey started in the field of data science, what tools and skillset they use and what’s essential for their work. For this week’s column, we caught up with the AWS Community Heroes.  This […]","categories":[],"tags":["AWS","DevOps"],"author_name":"Ambika Choudhury","publish_date":"2019-10-03T10:15:37","publication_year":"2019","word_count":775,"keywords":["data science","GCP","AWS","AI","R","serverless","microservices","Python","DevOps","Azure","kubernetes"],"extracted_tech_keywords":["AI","data science","AWS","Azure","GCP","kubernetes","microservices","serverless","Python","R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/behind-the-code-with-aws-community-heroes\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10084481,"title":"Tilck, a New Linux-Compatible Kernel is Here","content":"Vladislav K Valtchev, a senior kernel engineer, introduced Tilck, a monolithic x86 kernel for educational purposes that supports Linux at the binary level. Its compact size and easy design make it the testing ground for kernel mode experiments while preserving the ability to contrast how the exact same usermode bits operate on the Linux kernel. To create a programme, Tilck needs the i686-musl toolchain from bootlin.com. In contrast to most educational kernels, it does not require a separate set of custom-written applications. As a result, it can run popular Linux programmes like the BusyBox collection. Although the monolithic design and Linux compatibility may be limitations, such a design brings the entire project closer to practical applications in the future, as opposed to the situation where considerable efforts are needed to port existing software on it. Tilck can also implement custom non-Linux syscalls. Read more about Tilck here. Features of Tilck Hardware support: While the kernel makes use of some older hardware, such as the 8259 PICs for IRQs, the 8254 PIT for the system timer, the 16550 UART for serial communication, the 8042 kb controller, the 8237 ISA DMA, and the Sound Blaster 16 sound card, it also supports modern hardware features, such as SSE, AVX and AVX2 fpu instructions, PAT, the i686 sysenter and ACPI support via ACPICA. ACPI tracks power-button events, reboots or shuts down the computer, and retrieves the battery’s current parameters. File systems: Tilck has read-only support for FAT16 and FAT32 (used for initrd), allowing memory-mapping of files, a fully featured ramfs implementation, a minimalist devfs implementation, read-only support for these file systems, and a sysfs implementation for full view of the ACPI namespace, a list of all PCI(e) devices, and Tilck’s compile-time configuration. Tilck features a simplified VFS implementation that allows it to interact with many file systems simultaneously. Tilck does not yet support block devices; therefore, everything is processed in memory. Processes and signals: Although Tilck internally supports the concept of a thread, userspace is not currently exposed to multi-threading (kernel threads exist, of course). Both fork() and vfork() are integrated, and copy-on-write is employed for fork-ed processes. Tilck offers complete support for TLS (thread-local storage) via set thread area() despite the absence of multi-threading in userspace because libmusl necessitates it even for traditional single processes. I\/O: Tilck provides vectored I\/O via readv() and writev() besides the standard read() and write() syscalls and select(), poll(), and non-blocking I\/O. Userspace applications: Tilck can execute various console programmes, including the BusyBox suite, Vim, TinyCC, Lua, and framebuffer programmes like fbDOOM, a DOOM port for the Linux console. Console: The Tilck console supports more than 90% of the features included in Linux’s console. It operates in the same way in text mode and frame buffer mode (by using levels of abstraction). Vim can also run on Tilck. Tilck vs Linux Tilck does not target multi-user servers or desktop machines like Linux. Due to more features and the inherent complexity they provide, Linux is complicated. Tilck substitutes fewer features with: Simpler and shorter code Reduced binary size Deterministic behaviour Ultra-low-latency More robustness and simpler development Tilck’s system is integrated with unit tests, kernel self-tests, system tests, and automated interactive system tests (simulating real user input through QEMU’s monitor). Although it is educational project at the moment, it aims to be almost enterprise level.","excerpt":"Tilck is a monolithic x86 kernel for educational purposes that Linux backs at the binary level.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","Data Science","Deep Learning","Machine Learning"],"author_name":"Shritama Saha","publish_date":"2023-01-09T17:53:36","publication_year":"2023","word_count":555,"keywords":["programming_languages:R","AI","Machine Learning","RAG","Aim","Deep Learning","Data Science","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","Aim","RAG","R","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/tilck-a-new-linux-compatible-kernel-is-here\/","complexity_score":2,"technical_depth":5,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10021482,"title":"NFTs Are Making Artists Rich &#8211; Is It The Next Crypto Bubble?","content":"Non-Fungible Tokens (NFTs) have been capturing the headlines lately, with artists like pop star Grimes and EDM producer 3LAU raking in millions of dollars from selling their works linked with these crypto tokens. NFTs entered the market in 2017 as part of ‘crypto kitties‘, a blockchain game using digital cat collectables. However, in October last year, art auction house Christie’s New York made history by selling the first-ever NFT linked with a physical artwork — Portraits of a Mind: Block 21, by London based artist Benjamin Gentilli, sold for over 130,000 US Dollars. The popularity of NFTs has been on a rapid rise since then. Another record-breaking NFT work, this time purely digital, has also been auctioned at Christie’s at a current bid of an eye-watering 3.25 million dollars. This has led to an increase in the total market capitalization of NFTs to more than 350 million dollars in just six months and counting. So what exactly are Non-fungible Tokens, and are they a fad or a revolutionary technology? What are NFTs? In order to understand NFTs, we need to understand what fungibility is in the first place. A fungible asset can be replaced with another identical asset, currency or banknotes being the most commonly cited example. A non-fungible asset, on the other hand, is a unique, distinguishable asset that cannot be replaced in any form, like a house or . Non-fungible assets, however, have no standardized system of exchange or value. Incorporating blockchain technology with the unique attribute of these assets results in a non-fungible token that can be traded primarily using the Ethereum blockchain platform, giving it a means of exchange and value. A few NFTs however, are built on non-Ethereum based blockchain technology, like dGoods based on the EOS.IO platform. The two standardized forms of NFTs created so far are the ERC-721 and the ERC-1155. The former is the first standard developed and is a purely non-fungible asset. ERC-1155, on the other hand, is a semi-fungible asset, with a small number of identical tokens grouped as a unique class of assets rather than a single unique asset. Boon or Bubble? Creators on the internet, especially those who make digital art, have always struggled with copyright and plagiarisms of their work. On the one hand, most social media platforms automatically approve the right to use media uploaded to them by users. On the other hand, it just takes a few clicks of a mouse button to make a copy of the media for private consumption, without having to compensate the creator. NFT solves this problem by acting as a virtual certificate of authenticity that essentially ties the original digital artwork with the token, thereby providing it with the same status as an artist’s signature would on physical work. And because the token exists on the blockchain, it is instantly verifiable by anyone on a public network and cannot be forged. This allows creators to disseminate copies of their work on social media freely, as the authentic original retains its value by virtue of the NFT assigned to it. The previously mentioned ERC-1155 NFT is especially useful as tokens for limited edition sets of artworks. For collectors and buyers, websites like Makerspace have sprung up as a marketplace to facilitate a much easier and more transparent process of buying and selling digital art, much like financial stocks trading online. Cutting out the middlemen of galleries and curators also allows artists that have not been able to climb the echelons of the art world elite, to showcase their work to a much larger audience than before and earn a sizeable income. However, there are concerns about NFT’s skyrocketing value being a precariously fragile speculative bubble that could crash at any moment, bringing down the investors with it. There are also critics of the manufactured digital scarcity that NFTs base their value on. Besides, the recent popularity of the technology has seen its platforms get flooded with hundreds of thousands of works that prevent many from actually being seen by prospective collectors or buyers. Another concern is the environmental impact of NFTs as a result of heavy dependence on blockchain technology. The founder of the Robert Alice Project and the artist behind the famous Portraits of a Mind: Block 21 artwork, Benjamin Gentilli, spoke to Analytics India Magazine about his views on these concerns. He agreed that the speculation and inflation are ‘fundamentally corrosive to the art’ but remains hopeful that the NFT market will enjoy a longer, more stable growth once the excessive interest dies. Veteran artist, columnist, dealer and collector Kenny Schachter, one of the earliest adopters of digital art and a vocal supporter of NFTs, told Analytics India Magazine that “of course, some NFT’s are vastly overvalued,” but reiterated that mature markets see similar speculations in overvalued stocks. This affects markets as a whole; however, he said a few inflated prices shouldn’t be a reason to cast aside NFTs as a medium, and he doesn’t see the current situation as being a bubble. Having worked with digital art since the 1990s, he believes that NFTs have been the “first meaningful and effective delivery system to buy and collect such work” that he’s seen since then. He further argued that it is no more oversaturated than the real world and that many may go unnoticed, but demand will only grow from here on, for the foreseeable future. The future of NFT While it is indeed true that much of the value in Non-fungible Tokens is generated by speculative value, the biggest cryptocurrencies, Bitcoin and Etherium, and even Dogecoin, started the same way, and two of them have now become global mediums of exchange. If demand continues to rise in the NFT market, it will value new investors since it is primarily a reflexive market. And it isn’t just artists profiting from the technology. Gentilli calls NFTs a “programmable long term value creation for artists” and says that “in the long term I think NFTs will help revolutionize art, gaming and online identity. He isn’t far off the mark. Game companies, new ‘virtual real estate’ platforms, and even the NBA have come up with ways to use NFTs for their benefit. And now, developers in cryptocurrency and blockchain have a whole new market to build and expand their products, on top of mostly open-source software unfettered by API restrictions.","excerpt":"Non-Fungible Tokens (NFTs) have been capturing the headlines lately, with artists like pop star Grimes and EDM producer 3LAU raking in millions of dollars from selling their works linked with these crypto tokens. NFTs entered the market in 2017 as part of ‘crypto kitties‘, a blockchain game using digital cat collectables.  However, in October last […]","categories":["IT Services"],"tags":["API","Bitcoin","Blockchain","blockchain bitcoin","blockchain india","Blockchain Technology","crypto","crypto india","Cryptocurrency","cryptocurrency bitcoin","cryptocurrency ethereum","Ethereum","NFT","nft blockchain"],"author_name":"David B. Shrestha","publish_date":"2021-03-05T11:06:38","publication_year":"2021","word_count":1059,"keywords":["API","Cryptocurrency","crypto india","Git","BERT","R","nft blockchain","RAG","crypto","analytics","Go","Blockchain","blockchain india","AI","llm_models:BERT","cryptocurrency bitcoin","NFT","Bitcoin","programming_languages:R","blockchain bitcoin","Blockchain Technology","cryptocurrency ethereum","Ethereum"],"extracted_tech_keywords":["AI","analytics","RAG","R","Go","Git","API","BERT","llm_models:BERT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/it-services\/nfts-industry-shaker-or-speculative-bubble\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10044565,"title":"Zuckerberg Reveals Facebook’s Metaverse Ambitions","content":"CEO Mark Zuckerberg has laid down a grander plan to transform Facebook into a metaverse company. He spoke about how humans are basically mediating their lives and communication through small, glowing rectangles (read phones). He told The Verge’s contributing editor Casey Newton: “I think that that’s not really how people are made to interact”. Facebook’s metaverse will go beyond gaming to include the workplace, entertainment and so on to create a ‘social experience’ for its users. The company has invested heavily in virtual reality, with almost 20% of its employees working exclusively on VR and AR, and with recent acquisitions like BigBox VR and Unit 2 Games. The VR segment accounts for only 3% or less of Facebook’s top line. According to Fortune Business Insights, the global VR market is expected to grow from $3.1 billion in 2019 to $57.55 billion in 2027. Facebook’s Oculus aims to ‘defy distance’ and ‘explore virtual worlds as they grow and change.’ Zuckerberg said the metaverse would bring opportunities to individual creators, artists etc.  The metaverse would be the second-best thing to a teleportation device, he added. The company is also working towards an infinite office that will allow users to create their ideal VR workspace. Zuckerberg admitted the VR headsets in the market at the moment needed improvement, and are a ‘bit clunky’. The CEO further said the metaverse wasn’t something one specific company would build. He said humans should have a shared foundation of values and some understanding of the world and the problems that we all face together to have a cohesive society. Litigation Facebook’s metaverse vision comes at a time when the US government is taking serious steps to regulate Big Tech. Investigations into the corporations are underway, which may bring drastic changes to how the companies function. The FTC has accused Facebook of building a social media monopoly by acquiring smaller startups and competitors. In 2020, the Federal Trade Commission (FTC) attempted to force Facebook to hive off its two major acquisitions, Instagram and Whatsapp. The new set of laws attempting to break Facebook up may hinder the company’s ability to make acquisitions in the future. The introduction of metaverse might further complicate things. “A thriving metaverse would raise questions both familiar and strange about how the virtual space is governed, how its contents would be moderated, and what its existence would do to our shared sense of reality,” said Platformer founder Casey Newton. Data With VR technology, comes the unprecedented ability to track body motions. American VR researcher Jeremy Bailenson said commercial VR systems can track body movements 90 times per second to display a particular scene effectively, and high-end systems record 18 types of movements across the user’s head and hands. In 20 minutes, the systems can pick up under 2 million unique recordings of body language.” Everything a user does in the metaverse can be traced back to an individual’s identity. In addition to one’s location, social ties, search history and preferences, VR will enable Facebook to collect even nonverbal gestures like the user’s posture, gestures, the way they react to specific stimuli and interpersonal distance. Data collection from Facebook’s VR project will facilitate its plans for augmented reality, which will combine virtual elements with real-life surroundings. The company also spent billions of dollars to bring advertising to its Oculus VR platform. In June, the company announced its plans to bring their first in-game ads for “Blaston” from Resolution Games. However, the plan was shelved after user backlash.","excerpt":"Facebook has invested heavily in virtual reality.","categories":["Global Tech"],"tags":[],"author_name":"Prajaktha Gurung","publish_date":"2021-07-28T12:00:00","publication_year":"2021","word_count":582,"keywords":["Go","AWS","AI","cloud_platforms:AWS","programming_languages:R","programming_languages:Go","Aim","R","startup"],"extracted_tech_keywords":["AI","Aim","AWS","R","Go","startup","cloud_platforms:AWS","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/zuckerberg-reveals-facebooks-metaverse-ambitions\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10097248,"title":"Capgemini Strengthens Ties With Microsoft, Co Creates Azure Intelligent App Factory","content":"Building on their existing partnership, Capgemini has co-created an Azure Intelligent App Factory, in collaboration with Microsoft, for organisations to scale responsible and sustainable generative AI capabilities. Some of the use cases the duo has identified include creating content and design as well as analysing network traffic patterns for improved cybersecurity. This new solution will see the two companies combine their technology, including the Microsoft Cloud, GPT4-powered Azure OpenAI Service an Github Copilot. The Azure Intelligent App Factory aims to accelerate AI investments by controlling the security and focusing on handling of and access to data. Aiman Ezzat, the CEO of Capgemini, said “Both Microsoft and Capgemini are guided by strong ethical principles, which are the cornerstone of the new Azure Intelligent App Factory. By combining the Group’s global expertise in engineering and R&D services, data, and AI, with Microsoft’s market leading technology, we are committed to enabling clients to successfully implement AI solutions.” The Azure Intelligent App Factory is strategically centred on promoting the implementation of generative AI, enabled by Azure OpenAI Service. This initiative spans across various industries, including consumer products, life sciences, financial services, manufacturing, and telecommunications. The suite boasts an extensive array of enterprise-ready use cases designed to demonstrate the business value of generative AI. By fine tuning language models with proprietary company data and industry-specific information, it can deliver dependable and scalable outputs, empowering firms to develop solutions specific to their distinct business requirements. Notably, Capgemini is not the first enterprise level organisation to partner with Microsoft since generative AI has become the buzzword for the tech town. Leading firms Tata Consultancy Services (TCS), HCLTech, Accenture and Moody’s have recently strengthened their ties with the company behind Bing, Microsoft.Fascinatingly, another leading Indian IT giant Infosys, along with a few other parties including Elon Musk, AWS and others donated $1 billion to OpenAI. Two months ago the company announced Topaz — an AI-first set of services, solutions and platforms using generative AI with 12,000 use cases.","excerpt":"The identified include creating content and design as well as analysing network traffic patterns for improved cybersecurity.","categories":["AI News"],"tags":["Accenture","AWS","Generative AI","Intelligent Agent","Microsoft","TCS"],"author_name":"Tasmia Ansari","publish_date":"2023-07-19T17:56:18","publication_year":"2023","word_count":330,"keywords":["Accenture","Go","TPU","AWS","AI","OpenAI","Azure","R","Ray","Aim","Intelligent Agent","generative AI","Generative AI","TCS","Microsoft"],"extracted_tech_keywords":["AI","generative AI","OpenAI","Aim","Ray","AWS","Azure","TPU","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/capgemini-strengthens-ties-with-microsoft-co-creates-azure-intelligent-app-factory\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10070904,"title":"Data streaming platform Confluent opens a new office in Bengaluru","content":"Confluent, Inc. (NASDAQ: CFLT), the data streaming platform announced the opening of their new Bangalore office. They also spoke about creation of 100 additional jobs across India, a 50 percent increase in their current workforce. This new development is in line with the company’s ambition to drive high growth in 2022 by expanding its global presence and offering a differentiated model for cloud-native, complete and fully-managed data in motion in India. The company is experienceing surging demand for data streaming with several companies modernising their data processes across all industries, including retail, manufacturing, automative, and financial services sectors. Confluent recently announced a record number of net new customers for Q1 2022 and 64% total revenue growth year over year in its first quarter, with projected total revenue of $554-$560 million for FY22. “India is a key market for Confluent as we scale up our global presence. The public cloud services market is growing at a rapid rate, and we already see many exciting use cases among customers here. Bangalore is fast becoming the technology hub of the country, which makes it a natural home for Confluent to deliver its new category of data infrastructure to enable businesses of all types to set their data in motion,” said Hemanth Vedagarbha, SVP, Confluent.“Confluent’s extended presence in India will see the company making approximately 100 new hires for roles within go-to-market, engineering, and global supporting functions this year. Working under the direction of Srinivasulu Grandhi, VP, Confluent, the Bangalore-based team will focus on growing its customer base and working with businesses to unlock the full potential of the Confluent data in motion platform,” Chad Verbowski, Senior Vice President Engineering at Confluent.","excerpt":"Confluent speaks about creation of 100 additional jobs across India, a 50 percent increase in their current workforce.","categories":["AI News"],"tags":["Confluent"],"author_name":"Kartik Wali","publish_date":"2022-07-13T15:29:01","publication_year":"2022","word_count":278,"keywords":["Go","API","programming_languages:R","AI","programming_languages:Go","Confluent","R"],"extracted_tech_keywords":["AI","R","Go","API","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/data-streaming-platform-confluent-opens-a-new-office-in-bengaluru\/","complexity_score":2,"technical_depth":6,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10164517,"title":"Perplexity Prepares to Enter The Agentic Browsing Race with ‘Comet’","content":"AI-powered search engine Perplexity has announced that it is developing an agentic web browser called Comet. “Just like how GitHub Copilot was a truly assistive coding experience without AI bothering you but just helping you when you needed, we aspire to make Comet helpful in the same way,” said Perplexity CEO Aravind Srinivas on X, sharing the waitlist for the new tool. However, the company did not disclose the release date or details about the browser’s features. The web browser market is already highly competitive, with dominant players such as Google Chrome, Safari, and others offering AI-driven features. Perplexity AI has been rapidly rolling out new features. The company’s open-sourced R1 1776, a variant of the DeepSeek-R1 language model, was fine-tuned to remove censorship and deliver factual responses. The model weights are available on Hugging Face, and it can be accessed via the Sonar API. It also announced that Sonar, its in-house model, is now available to all Pro users. Subscribers can set it as their default model in settings. Additionally, Perplexity launched Deep Research, an autonomous tool for conducting in-depth research and analysis. It performs multiple searches, reviews hundreds of sources, and compiles findings into comprehensive reports. The feature is free for all users, with a daily limit of five queries for non-subscribers and 500 for Pro users. The Perplexity app is now integrated into the Paytm app on both Android and iPhone, appearing as ‘Ask AI’ on the home page.","excerpt":"The company did not disclose the release date or details about the browser’s features.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","Perplexity AI"],"author_name":"Aditi Suresh","publish_date":"2025-02-25T11:57:18","publication_year":"2025","word_count":242,"keywords":["Go","Hugging Face","API","programming_languages:R","AI","Git","Perplexity AI","mlops_tools:Comet","ai_frameworks:Hugging Face","GitHub","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","Hugging Face","R","Go","Git","GitHub","API","ai_frameworks:Hugging Face","mlops_tools:Comet","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/perplexity-prepares-to-enter-the-agentic-browsing-race-with-comet\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10084074,"title":"Council Post: Driving successful and resilient businesses at a time of great change in data practices","content":"The world around us is evolving rapidly. The past few years have dramatically changed the way we work, and the increased pace of this change has forced organisations to rethink their business strategies, take precise decisions quickly, and find opportunities for growth. CEOs and executive teams around the world are laser-focused on delivering success now by connecting with their customers in new, simpler, and more cost-effective ways. To lead businesses through this change, leaders must take tough decisions around business growth and profitability, sustainability goals, strategies to withstand economic plunges, and more on an everyday basis. Data can play a huge role in addressing such challenges. Every day, 2.5 quintillion bytes of data is produced. However, the difficulty lies, not in the sheer volume of data acquired, but in determining the most effective method for processing, analysing, and drawing insightful conclusions from it. Business environments are more dynamic than ever Maintaining resilient businesses that can drive success in today’s market is based on saving costs, reducing complexity, and increasing efficiency and the time to value—which is no small feat. Especially after the pandemic, offices are adapting to a more hybrid or work from home environment. The increased pace of change has forced organisations to rethink their business strategies. This evolving business landscape is concurrent with the change in customer behaviour as they begin rethinking their needs and values. Additionally, every crisis—be it economic or environmental—challenges organisations to reduce costs and improve efficiency as they brace themselves for further change. Economies are significantly impacted by interconnected singularities such as rising economic disparity, increasing population pressure, ongoing security concerns, complex global economic transition, environmental hazard, and increased technological advancement, making the business landscapes more complex, riskier, and challenging to understand or manage. To lead businesses through such uncertainties, leaders must continually make difficult decisions surrounding business growth and profitability, sustainability goals, strategies to withstand economic plunges, and more. Data plays an instrumental role in helping leaders innovate and make these decisions promptly while also enabling them to enhance overall productivity, collaborate effectively, and accelerate impact in ways that haven’t been previously imagined. However, there remains a significant gap between the present and the future of the business world, particularly between data challenges and attaining valuable insights. It’s easy to lose yourself in data for hours on end and even provide insights that are unimportant or irrelevant to your business needs. In order to acquire pertinent answers, you must ask the appropriate questions. If you want to act on your data, segmentation is crucial. You can begin delving deeper by grouping data that has a shared characteristic, such as a customer with comparable consumption patterns or schedules. Depending on the issue you want to address or the queries you want to address, you will be able to decide which category to study. Closing the gap with automation, AI, analytics, and data In order to bridge the gap between data challenges and attaining valuable insights, a unified strategy and platform with change management at its core is needed. The strategy starts at the data layer but needs to further include an analytics strategy, an artificial intelligence strategy, and an automation strategy to truly get data-driven insights for the organisation to work with. An analytics strategy outlines how data will be gathered and used to support business decisions as part of a holistic strategic vision. It aims to clarify important reporting indicators. Creating a single source of truth is pivotal when one is developing an efficient data and analytics strategy. To successfully create this single source of truth, it is necessary to be able to identify every data element’s origin, type, definition, and lineage. The key is understanding the source of the data and the modifications that have been undertaken. This enables the business to understand and convey what the data means. Likewise, an efficient automation strategy offers organisations a thorough and integrated approach to adopt. Automation strategies that assess scope, reliability, and impact characterise both robotic process automation (RPA) and business process automation (BPA). Businesses are likely to implement automation solutions to enhance efficiency and apply human resources to other important facets. This is especially probable when one considers the shift in recent years toward giving employees more independence and autonomy at work. Without automation solutions, employees would still be trapped performing mundane and repetitive duties, thus making this transformation, more or less, impossible. So far, it is evident that it is not only data acquisition that is crucial for businesses but also what follows after—the strategy, the process, and resource allocation. This implies that the strategy to process this acquired data to allocate resources efficiently and drive smarter decisions is imperative for modern enterprises. Technologies like Artificial Intelligence (AI) and Machine Learning (ML) are expanding data science capabilities to more people so that they can make better decisions faster, regardless of their technical expertise. Data-leading organisations—those with the most successful data cultures—see the business benefits of such a data transformation. “Data are just summaries of thousands of stories—tell a few of those stories to help make the data meaningful.”—Dan Heath This article is written by a member of the AIM Leaders Council. AIM Leaders Council is an invitation-only forum of senior executives in the Data Science and Analytics industry. To check if you are eligible for a membership, please fill the form here.","excerpt":"In order to bridge the gap between data challenges and attaining valuable insights, a unified strategy and platform with change management at its core is needed. The strategy starts at the data layer but needs to further include an analytics strategy, an artificial intelligence strategy, and an automation strategy to truly get data-driven insights for the organisation to work with.","categories":["AI Features"],"tags":[],"author_name":"Prashant Momaya","publish_date":"2023-01-05T13:00:00","publication_year":"2023","word_count":890,"keywords":["data science","Go","API","machine learning","artificial intelligence","AI","ML","Aim","analytics","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","data science","analytics","Aim","R","Go","API"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/driving-successful-and-resilient-businesses-at-a-time-of-great-change-in-data-practices\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10061805,"title":"How to build a data science portfolio in college?","content":"If you are entering higher ed and envision becoming a data scientist, you’re probably wondering where to start. To get the sexiest job of the 21st century, it’s pivotal to first understand the prerequisites – strong analytical and computational skills and start finding ways to hone in on them. From an academic standpoint, data science is flourishing. More than 596 data science, big data & analytics courses are offered by nearly 470 colleges pan-India. Most of these institutes collaborate with SMEs to prepare a curriculum that exposes students to subjects like data analytics, machine learning, business analysis, statistics, data modelling, data visualisation, cloud computing, database systems, and many more; alongside programming languages such as Python, JavaScript, Scala, R, SQL, and Julia. And if you are lucky, then an internship might be included. But, that’s not enough! According to Michael Page India’s ‘The Humans of Data Science’ report, data science is poised to create more than 11.5 million job openings by 2026. And if you want to stand out, you need to do more than the bare minimum. What you need is a data science portfolio. What is a data science portfolio? Think of a data science portfolio as an extension to your resume. It is essentially public evidence of the projects you’ve worked on, showcasing your creative, technical, and soft skills, approach to effectively analysing data and drawing insights, and ability to communicate the outcome to audiences. “A portfolio must essentially have projects that show your interest or expertise in different areas of data science. They should cover concepts like supervised learning, unsupervised learning, deep learning, etc. When you put these up in your resume or on platforms like GitHub, Deepnote, and Kaggle or your portfolio website, recruiters can see your capabilities first hand. A diverse, well-groomed portfolio can make all the difference in getting a job!” said Mohan C R, Project Engineer, Wipro. Ways to build a data science portfolio From a theoretical perspective, massive open online courses (MOOCs) are perhaps a good place to start. It will help you get your fundamentals right while enabling you to learn the required skills at your own pace. Listed below are some of the most trendy courses in 2022: Data Science Specialization by Johns Hopkins University Available on: Coursera Key Takeaways: Use R to clean, analyse, and visualise dataNavigate the entire data science pipeline from data acquisition to publication          Use GitHub to manage data science projects Perform regression analysis, least squares and inference using regression models Introduction to Data Science Available on: Metis Key Takeaways: CS\/Statistics\/Linear AlgebraExploratory Data Analysis and VisualisationData Modelling: Supervised\/Unsupervised Learning and Model Evaluation, Feature Selection, Engineering, and Data Pipelines, Advanced Supervised\/Unsupervised Learning, Advanced Model Evaluation and Data Pipelines Applied Data Science with Python Specialization by the University of Michigan Available on: Coursera Key Takeaways: Conduct an inferential statistical analysisDiscern whether a data visualisation is good or badEnhance a data analysis with applied machine learningAnalyse the connectivity of a social network Data Science MicroMasters by UCSanDiego Available on: edX Key Takeaways: Load and clean real-world dataMake reliable statistical inferences from noisy dataUse machine learning to learn models for dataVisualise complex dataUse Apache Spark to analyse data that does not fit within the memory of a single computer Python for Data Science and Machine Learning Bootcamp Available on: Udemy Key Takeaways: Python for Data Science and Machine LearningSpark for Big Data AnalysisImplement Machine Learning AlgorithmsUse NumPy for Numerical Data, Pandas for Data Analysis, Matplotlib for Python Plotting, Seaborn for statistical plots, Plotly for interactive dynamic visualisations, and scikit-learn for machine learning tasksK-Means ClusteringLogistic Regression and Linear RegressionRandom Forests and Decision TreesNatural Language Processing and Spam FiltersNeural Networks Apart from these courses, students can also participate in numerous hackathons that serve as a test-bed for aspiring data scientists. Each year, data science communities and platforms such as MachineHack, DataCrunch and DataHack, launch these events in collaboration with tech giants such as Genpact, IBM, etc., enabling tech enthusiasts to boost their skill sets and earn cash prizes\/certificates while emphasizing the “fun” quotient. Kaggle is one of the best places to check out for such competitions. Last and most importantly, you need real projects to hone your data science skills. What kind of projects should students look for? Going back to his student days, Mohan sought to look for projects that helped him hone in on essential data science skills such as data cleaning, exploratory data analysis, data visualisation, and machine learning. Here’s why: Data Cleaning: Data Scientists spend nearly 80% of their time cleaning data to find something useful. So it’s good to work on challenging projects which have data spread over multiple files and have null values. This will put your skills to the test and help push you into experimenting with different data cleaning methods. Exploratory Data Analysis (EDA): Performing EDA will help you gain insights from your cleaned data. It will test your graphical interpretation skills and statistical knowledge when plotting your data, using mean plots, standard deviation plots or any other type of plots. Data Visualisation: This process puts your storytelling abilities to the test! It will give you an idea of the many ways you can communicate and translate data using visual aids like graphs, charts, bars, or even images. There are many publicly available datasets that you can use to practice data visualisation and tell your story to the world. Machine Learning: Having a good grasp of machine learning fundamentals will go a long way in your data science career. Projects revolving around ML help you solve real-world problems with creative solutions. For example, if you can learn how to build a predictive model, you’d be able to predict the likely outcome and automate the downstream process every time you have new data with an unknown outcome. However, never make the mistake of skipping the basics of ML and directly jumping to advanced\/trending concepts.By the time Mohan got out of college, his portfolio included projects like Loan Prediction Problem Dataset, The Boston Housing Dataset, and Finding Donors for CharityML. Source: Loan Prediction Problem Dataset Here are some of the best projects you can work on as beginners: Fake News DetectionClimate Change Impacts on the Global Food SupplyHuman Action RecognitionForest Fire PredictionRoad Lane Line Detection Apart from these must-dos, aspirants can start writing blogs, focus on networking by attending virtual events on data science, and constantly keep themselves updated on the latest developments in the world of data science. What do recruiters think? Given the buzz around data science, it’s important to filter out the noise and get a clear understanding of what you need to be doing to get hired. “There is an odd misconception that they can pursue this career without using mathematics. Nothing could be farther from the truth! When you are stuck with a result and need to understand what it’s trying to say, you need to understand how you arrived at that result and evaluate it accordingly. This requires mathematics! And many such misconceptions need to be cleared before making a move,” says Puneet Tripathi, Head of Data Science, Wakefit.co. “I’d recommend students to get themselves familiar with different programming languages and open-source platforms by taking on as many projects as possible. This is the best way to ensure a strong career in data science.”","excerpt":"A diverse, well-groomed portfolio can make all the difference in getting a job!","categories":["AI Features"],"tags":["Cloud Computing","Data Analytics","Data Science","Data Scientist","Data Visualisation","Machine Learning","Statistics"],"author_name":"Sri Krishna","publish_date":"2022-03-01T15:00:00","publication_year":"2022","word_count":1214,"keywords":["data science","scikit-learn","NumPy","machine learning","Statistics","AI","Data Visualisation","ML","neural network","Machine Learning","Cloud Computing","deep learning","analytics","Data Analytics","Data Science","Data Scientist","Pandas"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","neural network","data science","analytics","scikit-learn","Pandas","NumPy"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-to-build-a-data-science-portfolio-in-college\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10076167,"title":"Startups Pay More To Techies Than Big Tech: Report","content":"A recent Twitter thread has revealed that the pay scale of software engineers is higher in Indian startups than big tech companies. The thread consists of data—collated from over 50,000 engineers—to investigate the difference in the salary and increment of techies working in startups in comparison to those working in giant tech corporations. Amit Singh, who works in a startup called Weekday, shared the Twitter thread. This startup primarily helps other companies hire engineers. Source: Twitter According to an infographic shared in the twitter thread, out of the total 107 Indian unicorns, ShareChat, the social media platform, offers the highest salary package to its software engineers. The study reveals that the salary of an engineer with four years of experience working at ShareChat is an estimated INR 47 lakhs per annum (LPA). The next one on the list is fintech firm CRED, where a typical software engineer draws a salary of an estimated INR 40 LPA. Companies like Meesho, Deam11, InMobi, and Swiggy have a package of 35~40 LPA. Other unicorns such as Oyo, PayTM, and Byju’s were at the bottom of the list, with a techie’s average package being 20~25 LPA. Further, ShopClues pays around INR 12 LPA on an average. For the increment cycle for techies in Indian unicorns, it was revealed that only a 10% increase was witnessed in the pay for these employees, as they gained experience. Amit Singh says, “this explains why people like to switch so often.” Source: Twitter A comparison chart was also created to demonstrate the difference in earnings of techies working in product-based startups and service-based companies such as TCS, Wipro, and Infosys. Infographics show that employees working in product startups draw 160% higher salaries than the software engineers in service-based companies. Furthermore, data reveals that the salary of a software engineer at big techs with four years of experience is INR 10 LPA whereas startups offer a package of INR 26 LPA for the same.Read the whole thread here.","excerpt":"Data reveals that the salary of a software engineer at big techs with four years of experience is 10 LPA, whereas startups offer a package of 26 LPA.","categories":["AI News"],"tags":["AI Startups"],"author_name":"Bhuvana Kamath","publish_date":"2022-09-30T16:22:10","publication_year":"2022","word_count":329,"keywords":["unicorn","AWS","AI","cloud_platforms:AWS","programming_languages:R","RAG","R","AI Startups","startup"],"extracted_tech_keywords":["AI","RAG","AWS","R","startup","unicorn","cloud_platforms:AWS","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/startups-pay-more-to-techies-than-big-tech-report\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10008229,"title":"50 Latest Data Science And Analytics Jobs From Past Week","content":"We have compiled a list of data science and analytics jobs that opened last week. The list consists of jobs from startups to larger companies across cities such as Bangalore, Chennai, Hyderabad, Pune and more. This is a cumulative list of jobs from various online job portals and are sorted according to the years of experience required. Freshers 1| Data Scientist\/Data Analyst at CAIA-Center For Artificial Intelligence & Advanced Analytics Location: Chennai, Pune, Mumbai, Bengaluru, Hyderabad Skills Required: SQL & PLSQL, Data Wrangling using Python, Cloud Data Lake, Statistics for Machine Learning, Business intelligence & Data Visualisation, data science, etc. Apply here. 2| Data Scientist at Mitron TV Pvt. Ltd. Location: Chennai, Pune, Delhi NCR, Mumbai, Bengaluru, Hyderabad, Kolkata Skills Required: SQL, Python, Scala, data analytics and machine learning, data science, etc. Apply here. 3| Data Analyst at HCL Location: Noida Skills Required: Machinery tool and parts, analytical tools, MS-Excel and MS-Powerpoint, etc. Apply here. 4| Data Analyst (Banking Operations) at Snaphunt Location: Bangalore Skills Required: Python, AI and ML algorithms, BI tools, AWS, data analysis, etc. Apply here. 5| Data Analyst at Myntra Location: Bangalore Skills Required: Data models, database design development, data mining, reporting packages (Business Objects, etc.), databases (SQL), Supply Chain Management, etc. Apply here. 6| Data Scientist at Meesho Location: Bangalore Skills Required: Machine learning, Python, data analysis, data science, Apache Spark, Hadoop, Redshift, etc. Apply here. 2 to 5 years 7| Data Scientist at L&T Technology Services Limited Location: Bangalore Skills Required: Preprocessing of structured and unstructured data, predictive models and machine-learning algorithms, data visualisation techniques, etc. Apply here. 8| AI Resident – Data Scientist – NLP at Shell India Location: Bangalore Skills Required: Natural Language Processing, computational linguistics, advanced analytics algorithms\/models, R, Python, MATLAB, etc. Apply here. 9| Senior Data Scientist at QpiAI Location: Bangalore Skills Required: Google automl, IBM autoAI, H20, Scikit learn, autokeras, CRISP-DM model, data interfaces to link to alphapiAI platform APIs, etc. Apply here. 10| AI Resident – Data Scientist – Chemical Process Analytics at Shell India Location: Bangalore Skills Required: Optimisation, signal processing, Kalman filtering, particle filtering, advanced models and algorithms, Python, etc. Apply here. 11| Principal Data Scientist – Communications at IBM Location: Bangalore Skills Required: Classical approaches to machine learning, Qiskit software, including Qiskit Aqua and Qiskit Terra, quantum computing, R, etc. Apply here. 12| Senior Data Scientist – MINT Sciences at Walmart Location: Bangalore Skills Required: Machine learning and mathematical optimisation, A\/B testing, sampling, labelling, EDA, Python, SQL and Linux\/UNIX environments, etc. Apply here. 13| Data Scientist at 3D-IP Semiconductors Location: Bangalore Skills Required: NLP, computer vision, machine learning, Python, etc. Apply here. 14| Senior Data Scientist at AlgonoX Technologies Location: Hyderabad Skills Required: Python, Java, C++, and Database design with SQL or NoSQL, design patterns, etc. Apply here. 15| Data Scientist at JSW Cement Ltd. Location: Mumbai Skills Required: Data mining, statistics, R, NoSQL, MS SQL Server, decision trees, AWS, Python, etc. Apply here. 16| Data Scientist III at NCR Corporation India Location: Hyderabad Skills Required: Python, Scala, PySpark, Java, ML models and algorithms, Keras, PyTorch, etc. Apply here. 17| Data Scientist – Digital Marketing at Piramal Consumer Finance Location: Mumbai Skills Required: Exploratory Data Analysis, Data Preparation, Data Exploration, Data Visualisation, R, Python, SQL, Power BI, MS-Excel, etc. Apply here. 18| Data Analyst (On Contract) at Electronic Arts Location: Hyderabad Skills Required: Advanced SQL skills, ML\/AI, Microsoft Power BI and\/or Tableau, etc. Apply here. 19| Data Analyst \/ Sr. Data Analyst – Data Science at Reserve Bank Information Technology Pvt. Ltd. Location: Mumbai Skills Required: Sentiment Analysis, Entity Extraction, Document Classification, Topic Modeling, Natural Language Understanding (NLU) and Natural Language Generation (NLG), etc. Apply here. 20| Data Analyst at Salesken.ai Location: Bangalore Skills Required: SQL, Excel, Power BI, Tableau, Python, Big Data and Machine learning knowledge, etc. Apply here. 21| Senior Data Scientist: Cloud at Oracle Location: Bangalore Skills Required: Machine Learning tools\/workflows, data integrations, Kubernetes-based infrastructure, Go, Python or Java, etc. Apply here. 22| Advanced Architect – Data Scientist at Mphasis Location: Bangalore Skills Required: Unstructured content mining, data mining, NLP, machine learning, Azure Machine Learning services, Azure ML Studio, Azure Cognitive toolkit, etc. Apply here. 23| Data Scientist at Hitachi Vantara Location: Bangalore Skills Required: R, Python, SAS, SPSS, Weka, Hadoop (Map Reduce paradigm), etc. Apply here. 24| IT Data Scientist at Kimberly-Clark Location: Bangalore Skills Required: Data Science, machine learning, Azure ML, Python, etc. Apply here. 6 to 12 years 25| Senior Data Scientist – Business Intelligence Unit, Retail Finance at Piramal Group Location: Mumbai Skills Required: Machine Learning Algorithms, Logistics & Linear Regression, Decision Tree, Clustering, R, Python, SQL, Power BI, MS Excel, etc. Apply here. 26| Senior Data Scientist at Michael Page Location: Gurgaon, Haryana Skills Required: Machine Learning, Artificial Intelligence, Computer Vision, Text Mining, Deep Learning, Clustering, Discrete Mathematics, Combinatorics, Informational Theory, Unstructured Text Classification, Latent Topic Modelling, etc. Apply here. 27| Manager (Data Scientist) at State Bank of India Location: Navi Mumbai Skills Required: Natural Language Processing, Web Crawling and Neural Networks, R\/Python, cloud-based application\/service development, NoSQL, SQL, etc. Apply here. 28| Data Scientist at IBM Location: Mumbai Skills Required: ILOG CPLEX, Standard Query Language (SQL), Apache Spark, Statistical Packages Social Sciences (SPSS), Python, R, multi-dimensional datasets, etc. Apply here. 29| Manager – Data Scientist at Vodafone India Location: Pune Skills Required: ML Modelling\/Data Science, Python and SQL, ML libraries like SparkML\/ Tensorflow, Scikit-Learn, Pandas and Statsmodels, GCP tools: Dataproc, Big Query, AI Platform, Dataflow, Cloud Composer, etc. Apply here. 30| Deputy Manager (Data Scientist) at State Bank of India Location: Navi Mumbai Skills Required: Natural Language Processing, Web Crawling and Neural Networks, Artificial Intelligence and parsing, Relational Databases OR any NoSQL databases, data extraction, etc. Apply here. 31| Senior Data Scientist at Bewakoof Brands Pvt. Ltd. Location: Bangalore Skills Required: ETL in SQL, R or Python, ElasticSearch, AI models in Google\/AWS platform, Kubernetes environment, etc. Apply here. 32| Senior Data Scientist at Walmart Labs Location: Bangalore Skills Required: Deep Learning, data science, machine learning, Scala, Python, data mining, etc. Apply here. 33| Data Scientist (SQL and Tableau) – Banking domain at Naukri Premium Location: Bangalore Skills Required: Tableau, SQL, Data analysis, BI, visualisation, data science, statistics, etc. Apply here. 34| Data Scientist at Intuitor SoftTech Location: Bangalore Skills Required: SQL, NoSQL, Pig, Hive, MongoDB, Cassandra, Azure ( app func, storage, event hub, keyvault, k8s), Graph database ( janusgraph), etc. Apply here. 35| Senior Data Scientist at WM Logistics Location: Gurgaon Skills Required: SQL, deep learning, pattern recognition, RNN, Python, AWS, PyTorch, etc. Apply here. 36| Data Scientist at Instacar Location: Delhi NCR, Mumbai, Bengaluru Skills Required: Data science, DBMS, deep learning, solution design, etc. Apply here. 37| Business Data Analyst at Netapp India Private Limited Location: Bangalore Skills Required: Data analysis, data mapping, data governance, Oracle, SQL Server and Access, etc. Apply here. 38| Data Scientist at Merck Ltd. Location: Bangalore Skills Required: Python, Java, JavaScript, AWS, Azure, Business Objects, D3, ggplot, etc. Apply here. 39| Data Scientist 2 at Paypal India Pvt Ltd. Location: Bangalore Skills Required: Machine Learning, SQL, data mining, data science, data visualisation, Tableau, etc. Apply here. 40| Lead Data Scientist at Brillio Technologies Pvt Ltd. Location: Bangalore Skills Required: Data analysis, SAS, Simulation, Time series analysis, Network analysis, Machine learning, etc. Apply here. 41| Lead Data Scientist at Unilever Location: Bangalore Skills Required: Customer Development, Supply Chain, statistics, Finance, Python, R, machine learning, AI, etc. Apply here. 42| Senior Data Scientist, Rider Analytics at Uber Location: Bangalore Skills Required: SQL, Python or R, data science, machine learning, etc. Apply here. 43| Senior Data Scientist at Philips Location: Chennai Skills Required: R, Python, Matlab, or SAS,  relational databases and SQL, Map\/Reduce, Hadoop, Hive, etc. Apply here. 12+ years 44| Principal Data Engineer at Lowe’s Location: Anekal, Karnataka Skills Required: BI Engineering, Data Warehousing, Java\/Scala\/Python, SQL, Scripting, Teradata, Hadoop (Sqoop, Hive, Pig, Map Reduce), Spark (Spark Streaming, Mlib), Kafka or equivalent Cloud Big Data components, etc. Apply here. 45| Senior Data Scientist\/Machine Learning\/AI Engineer at Bion Location: Bangalore Skills Required: Java\/J2EE,  ML\/AI, REST, Javascript, HTML, SQL, etc. Apply here. 46| Principal Data Scientist at Walmart Labs Location: Bangalore Skills Required: Python, R, Scala, supervised and unsupervised learning, SQL and relational databases, data warehouse, Hadoop(Hive, Pig, Map Reduce, HQL)\/Spark\/H20, etc. Apply here. 47| Senior Data Scientist – Pharma Domain at Konnectrack Info Solutions Location: Hyderabad Skills Required: R, NLP, statistics, data science, data analytics, machine learning, Python, etc. Apply here. 48| AVP\/SR. AVP – Data Scientist at Crescendo Location: Gurgaon Skills Required: Data Science using Python, NLP, LSTM, CNN, Dense neural networks, SQL, ETL, etc. Apply here. 49| AVP\/SM Data Scientist For Goods & Service Tax Network at NISG (National Institute for Smart Government) Location: Delhi Skills Required: Python, SparkML, Tensorflow, Big DL, Horovod, H20, Rapidminer, R, MySQL, MongoDB, Hadoop ecosystem, couchdb, redis, neo4j, etc. Apply here. 50| Data Scientist at India Israel Innovation HUB Location: Delhi NCR Skills Required: Data Engineer, cloud data (e.g. Spark), DB, Machine Learning & Programming (e.g. Python, Java, C#), etc. Apply here.","excerpt":"We have compiled a list of data science and analytics jobs that opened last week. The list consists of jobs from startups to larger companies across cities such as Bangalore, Chennai, Hyderabad, Pune and more. This is a cumulative list of jobs from various online job portals and are sorted according to the years of […]","categories":["AI Hirings"],"tags":["AI Jobs","big data for social good","big data storage format","current ML jobs","data analyst","Data analyst jobs","Data Analytics","Data Science Career","Data Science Jobs","Data Scientist Jobs","data structure using java","etl hadoop","extract big data","hadoop and etl","hadoop etl","hadoop on azure cloud","is hadoop a company","latest technology in cloud computing","ML jobs","principal data scientist","social network big data","spark nosql"],"author_name":"Ambika Choudhury","publish_date":"2020-09-24T13:00:11","publication_year":"2020","word_count":1504,"keywords":["big data storage format","computer vision","deep learning","extract big data","hadoop etl","hadoop and etl","current ML jobs","ML jobs","data science","Data Scientist Jobs","artificial intelligence","principal data scientist","NLP","analytics","machine learning","is hadoop a company","AI","neural network","data structure using java","big data for social good","Data analyst jobs","Data Science Jobs","ML","Data Science Career","spark nosql","hadoop on azure cloud","latest technology in cloud computing","AI Jobs","data analyst","Data Analytics","social network big data","etl hadoop"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","neural network","NLP","computer vision","data science","analytics"],"url":"https:\/\/analyticsindiamag.com\/ai-hiring\/50-latest-data-science-and-analytics-jobs-from-past-week\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10021720,"title":"IBM’s Strategy For Hybrid Cloud Growth In India","content":"In a bid to accelerate its hybrid cloud strategy in India, IBM recently launched Cloud Satellite. Available across locations, the Cloud Satellite offers a standard set of cloud services complete with toolchains, databases, and AI. In related news, IBM has received full Cloud Service Provider empanelment from the Indian Ministry of Electronics and IT (MeitY), allowing the tech giant to work with government agencies and other public sector undertakings. To wit, IBM is making inroads into India with its hybrid cloud strategy. IBM Cloud Satellite “With the launch of Cloud Satellite, we are giving clients in India, across telecommunication, financial services, government, healthcare, retail, and more, access to a consistent and secure set of cloud services, wherever their workload resides. Clients can now meet essential data requirements such as data sovereignty and privacy, delivered via the IBM Cloud,” Shailesh Agarwal, vice president, sales and industry business development, IBM India and South Asia said. IBM’s cloud satellite would help organisations to deploy and run apps across platforms–on-premise, edge computing, and public cloud from any vendor. It delivers cloud services, access policies, security control and compliances, and APIs. IBM Cloud Satellite can also help address data privacy issues for telecommunication, healthcare, and government with its secure and unifying cloud services across environments. Edtech and remote work also stand to benefit for the same reason. IBM has partnered with enterprise technology company Lumen to integrate the Cloud Satellite with its edge platform connecting up to 180,000 clients. The clients will have access to hybrid cloud services and can build innovative solutions at the edge. They would also be able to deploy data-intensive applications across distributed environments. What Is IBM’s Hybrid Cloud Strategy? IBM has listed the core components of its cloud strategy in a white paper. Working with open source: Open source is at the core of IBM’s operations. The company has made several investments in open source technologies such as Linux, Hadoop, Kubernetes, Helm, and Spark. IBM, in the last five years, has been transforming its public and private cloud platforms based on open-source standards. Foundation based on Kubernetes: Kubernetes is arguably the most popular container orchestration system. IBM has based both its public and private cloud on the same Kubernetes orchestration because it is optimised to support all infrastructure platforms — from mainframes to virtual machines. IBM also benefits from Kubernetes’ expansive ecosystem of tools and services. Security: IBM’s cloud platform includes certified IBM security services ready to be implemented. The services make sure all the disparate operations and functions operate within this multi-cloud environment without much hassle. IBM has also created a set of built-in security capabilities that can be leveraged across hybrid and multi-cloud environments, including visibility across the cloud, data protection, and access management. Multicloud Manager: It is a platform designed by IBM to manage workloads across hybrid and multi-cloud environments. It is a command and control service that manages Kubernetes deployments. Multicloud Manager offers governance, policy management, and security in a hybrid environment. Distributed Data Management: IBM has built several artificial intelligence and machine learning capabilities throughout its platform to address the challenges of managing, ingesting, and analysing data from various sources. India Plans IBM views India as a high-growth market. An internal study by the company revealed that 17 percent of the surveyed organisations said they were willing to increase their expenditure on the hybrid cloud to 49 percent by 2023. The study also said the majority of cloud budgets are allocated to hybrid cloud platforms. However, only about 20 percent of the workload has been moved to the cloud. “This leaves us with a massive untapped opportunity wherein hybrid cloud is positioned to become a dominant force for driving digital transformation for enterprises,” said Viswanath Ramaswamy, VP – Cloud & Cognitive Software & Services, IBM India\/South Asia in an earlier interview. In particular, IBM’s engagement with Indian IT companies accelerated after Arvind Krishna was appointed as the CEO. In India, IBM is already working with Infosys, Tata Consultancy Services, and Wipro to move their client’s workloads to the cloud. In 2020, the Big Blue pledged to invest $1 billion over the next three years to expand its cloud ecosystem in India. IBM is also looking to target public and government agencies. “We are committed to India’s growth journey. The launch of IBM Cloud satellite and IBM Cloud accreditation enables us to further strengthen our collaboration with organisations and the government\/public sector,” a company spokesperson said.","excerpt":"In a bid to accelerate its hybrid cloud strategy in India, IBM recently launched Cloud Satellite. Available across locations, the Cloud Satellite offers a standard set of cloud services complete with toolchains, databases, and AI.  In related news, IBM has received full Cloud Service Provider empanelment from the Indian Ministry of Electronics and IT (MeitY), […]","categories":["IT Services"],"tags":["Arvind Krishna"],"author_name":"Shraddha Goled","publish_date":"2021-03-09T17:00:00","publication_year":"2021","word_count":740,"keywords":["Go","API","machine learning","artificial intelligence","AI","Git","RAG","Arvind Krishna","edge computing","R","kubernetes"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","RAG","kubernetes","edge computing","R","Go","Git","API"],"url":"https:\/\/analyticsindiamag.com\/it-services\/ibms-strategy-for-hybrid-cloud-growth-in-india\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10164740,"title":"‘90% of Programming Languages Have English-Based Syntax’","content":"English continues to dominate global communication, but its role in programming has always been a topic of debate. Douglas Crockford, the creator of JSON.org and a key figure in JavaScript development, is skeptical of the growing belief that “the hottest new programming language is English”. In an exclusive conversation with AIM, he mentioned that English is too ambiguous for programming, where precision is paramount. “If we’re trying to communicate with computers, there is no room for understanding,” Crockford said, emphasising that AI does not grasp context the way humans do. However, others foresee a different future. AI-driven tools like GitHub Copilot are reshaping coding by enabling natural language programming. Tech leaders like NVIDIA CEO Jensen Huang highlighted that English is evolving into a programming language, making coding more accessible. Speaking at the World Governments Summit, Huang explained, “It is our job to create computing technology such that nobody has to program and that the programming language is human.” Andrej Karpathy, senior director of AI at Tesla, had predicted this trend back in 2023. The hottest new programming language is English— Andrej Karpathy (@karpathy) January 24, 2023 English is the Foundation of the Tech Industry Industry insights reaffirm that while AI and automation skills are in high demand, English remains the foundational language of the tech sector. According to TeamLease, over 90% of programming languages have English-based syntax, and more than 80% of technical documentation is in English. This makes the language the primary medium for coding, collaboration, and innovation. Meanwhile, Krishna Vij, vice president of TeamLease, told AIM, “Big tech firms prioritise scalability, requiring developers to work across geographies and serve diverse markets. Limited English proficiency can hinder access to 75% of open-source projects and major developer communities like GitHub and Stack Overflow. While AI enables coding in regional languages, such solutions are still evolving and have yet to achieve mainstream adoption.” Recognising this need, SpeakX, an Indian EdTech startup, is transforming English learning through advanced generative AI technology. SpeakX’s AI-powered platform provides personalised learning experiences, enabling users to engage in real-time conversations with a dedicated AI tutor. Arpit Mittal, founder and CEO of SpeakX, told AIM, “English is more critical than ever in today’s globalised, tech-driven world. In fields like data science, analytics, and technology, English is a non-negotiable skill; it’s the gateway to accessing research, mastering tools like Python or Tableau, and collaborating seamlessly with international teams.” He noted that beyond access, English is a commercial and social superpower, enabling professionals to articulate their ideas clearly, whether pitching a model to a stakeholder in India or abroad. English as a Soft Skill As India cements its position as a global talent powerhouse, the demand for holistic development and strong English proficiency has never been more crucial. NITI Aayog’s recent report titled ‘Expanding Quality Higher Education Through States and State Public Universities’ highlighted the challenge of talent outflow from certain states due to inadequate employability skills. In most cases, English proficiency is identified as a significant barrier. To bridge this gap, the report recommends collaborations with international language organisations, underlining the importance of language skills in the nation’s workforce readiness. India’s information technology (IT) and IT-enabled services sector (ITES), which is largely dominated by multinational corporations, necessitates strong English communication skills for global client interactions. Cambridge has been instrumental in addressing this skills gap. A testament to the growing importance of English skills is Tata Consultancy Services (TCS) making Cambridge’s Linguaskill assessment a mandatory part of its NextStep job portal. The Ever-Changing Tech Landscape In an exclusive conversation with AIM, Arun Rajamani, managing director of Cambridge University Press and Assessment, South Asia, emphasised the rapid evolution of skills in the tech industry. “If you look at the past decade, the required skills have continuously evolved. In the tech industry, we see what we call the ‘two-year half-life syndrome’, where half of what you know becomes obsolete every two years. However, one constant remains – the ability to adapt and learn new skills effectively.” Rajamani further highlighted that while technical skills evolve, communication and interpersonal skills have become indispensable. The globalisation of the workforce has reinforced this need, with tech teams now distributed across India, Israel, Eastern Europe, and other global locations. Seamless collaboration across linguistic and cultural contexts has become critical for delivering high-quality, timely projects. When AIM asked Rajamani if he had to choose between English communication skills and AI skills, he said, “I would prioritise English communication. That’s the key to becoming a lasting and valuable member of the workforce today.” What’s Next? Recently, Cambridge University Press & Assessment launched a report titled ‘Fostering Future Leaders and Global Professionals: A Holistic Approach to Talent Development in Academia and Industry’. The report noted a rise in the demand for AI-related skills globally, with a 33% increase in job postings requiring AI skills across 14 Organisation for Economic Co-operation and Development (OECD) countries between 2019 and 2022. Organisations are investing in training programs to upskill their workforce in AI competencies. Government initiatives like Skill India and Digital India aim to promote digital literacy and AI-specific training programs to prepare the workforce for future demands. However, the Future of Jobs Report 2025 released by the World Economic Forum noted that it’s not just tech-related skills that are on the rise – creative thinking, resilience, leadership, and even environmental stewardship are also climbing the skills ladder, stressing the value of a well-rounded and adaptable workforce. The shift is prompting businesses to invest in reskilling and upskilling initiatives.","excerpt":"According to TeamLease, more than 80% of technical documentation is in English.","categories":["AI Features"],"tags":["AI (Artificial Intelligence)","english","India","skills"],"author_name":"Vidyashree Srinivas","publish_date":"2025-03-04T12:30:00","publication_year":"2025","word_count":916,"keywords":["data science","Go","AI","ML","english","JavaScript","Python","Aim","analytics","skills","generative AI","R","India","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","ML","data science","analytics","generative AI","Aim","Python","R","JavaScript","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/90-of-programming-languages-have-english-based-syntax\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10001973,"title":"Understanding Ethereum’s Scalability Problems And The Solution On The Horizon","content":"Ethereum, the second-biggest blockchain network in the world, has been suffering from scalability issues for the last few years. The non-scalability of the proof-of-work model shows that blockchains using this consensus mechanism cannot rise above single-digit transaction throughput numbers. Owing to this, the network is currently undergoing a transition to proof-of-stake, an alternative consensus mechanism that seems to, theoretically, solve the problems of the Ethereum network. Ethereum’s Clogged Network Ethereum was launched in 2015 with the purpose of being a decentralized ‘world computer’. The consensus mechanism of the blockchain would allow for many decentralized application to run on it, similar to a cloud deployment with the added security, decentralization and uptime of being on a blockchain. This quickly manifested as the ‘ICO craze’, wherein multiple individuals began creating coins on the Ethereum blockchain which did not have any purpose except to take money from investors. This caused a lot of new users and transactions on the blockchain, which has a transaction throughput of 15 transactions per second. The network became clogged and the cost required to use the network increased. This cost, which is designated as ‘gas’, is an amount paid in Ether to the parties that process transactions on the network. The scalability issues are intrinsic to this consensus mechanism, owing to the fact that all transactions have to be broadcast to the network for every block, and the most secure node is one with all the data of the blockchain. This creates an environment that does not allow for the direct scaling of transaction throughput without increasing the size of individual blocks, which leads to centralization due to inaccessibility. The Results Of Overload Miners, those who process transactions on the network, directly take the gas fee for each transaction, which makes them the main deciding authority for the average cost of gas. A transaction with a higher gas cost attached to it will be accepted faster, with a lower gas cost being generally slower due to miners prioritising better returns. However, when the network is clogged, transactions begin backing up in the memory pool, which is where pending transactions are present. This causes miners to prioritise transactions with higher gas prices, and increase the lowest gas price for confirming a transaction. This creates a situation where the network’s gas fee goes up to exorbitant levels as the network becomes more clogged, only worsening the situation. This was seen in the case of the viral popularity of the CryptoKitties dApp. CryptoKitties was a game that allowed users to collect and trade tokenised unique digital cats on the Ethereum blockchain. It underwent an explosion almost overnight, introducing a lot of transactions to the memory pool of the chain. The number of transaction requests for the network doubled, from 622,000 to almost 1.07 million, almost driven entirely by CryptoKitties. This caused gas prices to inch up, and then skyrocket owing to increased network usage. How Ethereum Is Planning To Scale This is just one of the problems that the developers of Ethereum foresaw, with limited scalability being one of the original issues to be addressed on the network. However, the network has only grown in size, and it is difficult to perform upgrades a network that is still being used heavily every day. Moreover, Ethereum cannot be ‘paused’ for upgrades like a centralised network can. This has created problems for solving the scalability problem, especially since it upgrades the consensus mechanism; the heart of the world computer. The main proposed change, among others, is a transition to another consensus mechanism known as proof-of-stake. While proof-of-work relies on pure computing power, or ‘work’ to ensure consensus among all participants. This is wasteful of computer power, even though it is being used to secure the network. Proof-of-stake, on the other hand, functions on more fundamental governing principles. The consensus mechanism puts great focus on identifying and eliminating bad actors from reaching consensus. This is known as Byzantine Fault Tolerance and is an important part of keeping the blockchain running smoothly. Proof-of-stake aims to reduce the amount of bad actors by forcing them to have skin in the game. Validating nodes are required to ‘stake’ or lock up a certain amount of ETH in a smart contract, the size of which determines the strength of the vote that the party has on the blockchain. These validators are then assigned the right to propose versions of blocks, with consensus reached after a round of voting to ensure that there is a common shared history with all parties on the network. This is more efficient in terms of electricity consumed, and is also good for the economy of the network. As there is no need for new coins to be constantly minted to incentivize miners, the network can adopt a deflationary model as opposed to an inflationary one. Moreover, transaction throughput can also be increased as validators are not required to have a copy of the entire chain, only the state of transactions on the chain. Proof-of-stake chains are more scalable and sustainable for a longer period of time, and are also more resistant to 51% attacks. The implementation, however, seems far away for Ethereum, at least at this moment.","excerpt":"Ethereum, the second-biggest blockchain network in the world, has been suffering from scalability issues for the last few years. The non-scalability of the proof-of-work model shows that blockchains using this consensus mechanism cannot rise above single-digit transaction throughput numbers. Owing to this, the network is currently undergoing a transition to proof-of-stake, an alternative consensus mechanism […]","categories":["AI Features"],"tags":["Blockchain","cryptocurrencies","Ethereum"],"author_name":"Anirudh VK","publish_date":"2019-04-30T19:49:00","publication_year":"2019","word_count":864,"keywords":["Go","Blockchain","programming_languages:R","AI","cryptocurrencies","Scala","Git","RAG","Aim","programming_languages:Scala","GAN","R","Ethereum"],"extracted_tech_keywords":["AI","Aim","RAG","R","Go","Scala","Git","GAN","programming_languages:R","programming_languages:Scala"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/understanding-ethereums-scalability-problems-and-the-solution-on-the-horizon\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10091341,"title":"ChatGPT Enters Indian Politics","content":"The 2016 US elections were significant in more ways than one. It was found that the Russian government ran a campaign to tarnish the image of Democratic candidate Hillary Clinton via cyberattacks and tilt the election in favour of Trump. The saga made it clear that technology would play a huge part in all election outcomes in future. Today, with the rise of chatbots and its pervasiveness across domains, politics is potentially looking at one of its biggest threats ever. In a field where false information and biases can easily meddle with the outcomes of political campaigns, will a hallucinating chatbot like ChatGPT add fuel to fire? Noted AI researcher and author, Pedro Domingos tweeted saying, “The killer occupation for GPT-4 is politician, because the key qualification is to have no fear of lying (sic)” While it is easy to say that GPT-4 can replace politicians, let’s explore how Indian politicians and political parties are exploring ChatGPT’s use cases. Speaking to AIM, Sagar Vishnoi, political analyst at The Ideaz Factory said that they use ChatGPT for planning schedules, marketing plans and for writing political speeches. “It helps us save time and improve productivity,” he said. The tool is already being used in offices of MPs, CMs and think-tanks. As confirmed by Sagar, Netri Foundation, a political incubator, is using ChatGPT already for content creation on social media and research work. However, with the chatbot being known for hallucinations and churning misinformation, how reliable are they? Political Bias There have been long discussions on the political inclination of ChatGPT and how its biased views clearly favour a particular side. When a user asked ChatGPT to write a poem on two Presidents, the response was not only biased but reflecting what mainstream media has been painting about the two personalities. When ChatGPT was put through various tools to adjudge which side it leaned towards, using tests like the Political Compass Test, research scientist David Rozado observed that ChatGPT is essentially left-leaning with a strong libertarian political bias. Source: Twitter Knowing all that we now know about the chatbot’s erroneous responses, it isn’t a stretch to imagine that the tool can be used to taint or glorify a certain leader. It can potentially be used to smear political campaigns to the whims and fancies of the companies making these models. Recently, ChatGPT falsely accused Brian Wood, Mayor of Hepburn Shire of having served a prison term for involvement in a foreign bribery scandal linked to the Reserve Bank of Australia. In reality, he was the one who alerted the authorities about the bribery scandal when he worked there. The mayor has now threatened to file a defamation suit against OpenAI if the statement is not rectified. There are a lot of subjective variables when it comes to training an AI model. ChatGPT and other chat-based platforms, are trained on large datasets that are sourced from different places on the internet. Essentially, ChatGPT will bring the same political bias that most news organisations have in their news stories. ChatGPT: ‘Cambridge Analytica’ in the Making In an interview in March, OpenAI chief Sam Altman said, “Elections are in our mind.” How or to what extent his company plans to contribute to them is still unknown. Interestingly, Altman has always openly acknowledged the bias that exists in the system and foresees more challenges in the future. Even with the recent announcement on security and privacy, they haven’t figured out a way to curb the bias. there will be more challenges like bias (we don’t want ChatGPT to be pro or against any politics by default, but if you want either then it should be for you; working on this now) and people coming away unsettled from talking to a chatbot, even if they know what’s really going on— Sam Altman (@sama) February 19, 2023 During the 2020 elections, social media companies like Twitter, Facebook, and YouTube unanimously banned President Donald Trump on the alleged grounds of provocation, which had not been legally proven. But it was a clear indicator of how the media had set itself against one party. With the already existing bias in the media, a biased chatbot can be even more detrimental for the 2024 elections. In a podcast interview with Meghan McCarty Carino, AI author and expert, Gary Marcus said, “2024 election is going to be a train wreck.” He expressed his concerns on how it’s going to be easy to produce fake news stories that resemble “authentic publications.” The propensity to churn out multiple versions of misinformation will be so easy that it will “flood the zone with complete nonsense” and people will not be able to tell the difference. He also believes that there will be counter campaigns which will eventually end up in a “place where people really don’t believe anything.” What about the Regulations? With rising concerns around AI security, countries are contemplating over introducing stringent policies. In the US, the Biden administration is looking to bring in checks on ChatGPT-like tools owing to growing concerns of technology being used to “discriminate and spread harmful information.” Amidst all concerns, OpenAI is still sticking to their stance of prioritising AI safety while not addressing the chatbot’s biases and hallucinations. If political parties wish to use chatbot-generated information in their manifestos or campaigns, then an extensive vetting process can possibly address the spread of misinformation. However, information generated via the chatbot that comes from outside the party’s purview cannot be regulated, and the implications of it are still unknown.","excerpt":"ChatGPT is being used to plan schedules, market plans and draft politician’s speeches. But the negatives outweigh the positives by a huge margin. Read on","categories":["AI Trends"],"tags":["AI Safety","bias","ChatGPT","elections","gary marcus","misinformation","OpenAI","politics","Sam Altman"],"author_name":"Vandana Nair","publish_date":"2023-04-13T19:25:00","publication_year":"2023","word_count":919,"keywords":["misinformation","Go","ChatGPT","Sam Altman","OpenAI","AI","chatbots","R","politics","BERT","GPT","gary marcus","Aim","GAN","AI Safety","bias","elections"],"extracted_tech_keywords":["AI","ChatGPT","OpenAI","Aim","chatbots","R","Go","BERT","GPT","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/chatgpt-enters-indian-politics\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10043398,"title":"OTT Wars Set To Intensify As Amazon Strikes Deal To Acquire MGM","content":"Amazon has reached an agreement to acquire MGM Studios Inc. MGM makes its revenue by selling distribution rights to movie theatres, TV networks, and streaming services as an independent film-and-TV studio. As Disney and Netflix have demonstrated, streaming platforms want to keep the distribution rights, and feature exclusive content on their platforms. For example, Disney pulled many TV shows and movies off Netflix leading up to the launch of Disney+.  Amazon would have rights to exclusive streaming of MGM produced entertainment like James Bond movies and Shark Tank. It also enables the creation of spinoffs, as the company pointed out. OTT wars MGM has been a takeover target for a while now. Streaming platforms are seeking to combine assets and strengthen their position. The merger and acquisition strategy is trending up. Walt Disney Co.’s acquisition of 21st Century Fox’s assets and AT&T’s acquisition of Time Warner, including its HBO and Turner assets, provided the streaming platforms with valuable intellectual property. “The real financial value behind this deal is the treasure trove of IP in the deep catalogue that we plan to reimagine and develop together with MGM’s talented team,” Mike Hopkins, senior vice president of Prime Video and Amazon Studios, said in a statement. Streaming architecture Streaming elites like Netflix and Amazon invest in original premium content. In India, however, the two behemoths have another competition. Hotstar hosts entertainment and sports content in many languages. Hotstar is also climbing the streaming ladder and boasts more than 300 million active users as of 2020, making it India’s number one streaming platform. Amazon owns both Amazon Web Services (AWS) and Prime Video. Prime Video uses cloud computing and CDNs. As opposed to this, competing streaming services have to pay a fee. Its top competitor Netflix runs on AWS. Nonetheless, using AWS may be cost-effective for Netflix since it does not have to maintain its own servers and networks. However, despite owning AWS, streaming quality of Prime Video is not on par with Netflix. Netflix’s algorithm [Netflix Recommendation Engine (NRE)] filters more than 3,000 titles using 1,300 recommendation clusters–which is accurate 80% of the time–saving Netflix over 1 billion dollars a year. All streaming services use recommendation engines, but Netflix is the most successful with a retention rate of 93%, with Prime Video coming second. Prime Video’s algorithms come from the house of AWS. Hostar, Amazon, Disney plus, HBO Now, Hulu etc are far behind Netflix in the algorithm game. Earlier, Amazon pointed out that it does not intend to compete with Netflix. “We don’t compete with Netflix,” Bezos said on day one of the code conference. “I think people are going to subscribe to both.” Amazon’s streaming tactics “When we win a Golden Globe, it helps us sell more shoes,”Jeff Bezos MGM content helps Amazon Prime, and Amazon Prime helps Amazon (e-commerce). Prime Video has grown to become one of Amazon’s biggest expenses at $11 billion for 2020 for original and licensed content. Amazon’s Prime membership has also been growing steadily. The Prime membership comes with a bundle of services including Prime Video, which differentiates Amazon streaming from other competitors like Netflix or Disney+. Source: Statista Report Bandwidth-hogging In 2020 on request by the European Union, Netflix decided to reduce its streaming bitrates, reducing traffic by almost 25% for thirty days. OTT platforms drive traffic. Amazon also reduced bitrate speeds in Europe and promised to monitor situations in other parts of the world. Almost every streaming platform subscriber base has been on the rise, and OTT platforms are ready. Bandwidth providers, however, may not be. Different ISPs have built their networks in different ways around the world, and each has its own set of constraints. Netflix has received requests from countries to reduce streaming quality. As the usage increases, services may face disruptions.","excerpt":"Amazon has reached an agreement to acquire MGM Studios Inc. MGM makes its revenue by selling distribution rights to movie theatres, TV networks, and streaming services as an independent film-and-TV studio. As Disney and Netflix have demonstrated, streaming platforms want to keep the distribution rights, and feature exclusive content on their platforms. For example, Disney […]","categories":["Global Tech"],"tags":[],"author_name":"Prajaktha Gurung","publish_date":"2021-07-12T16:00:00","publication_year":"2021","word_count":631,"keywords":["Go","AWS","cloud computing","AI","cloud_platforms:AWS","programming_languages:R","programming_languages:Go","disruption","cloud_platforms:Amazon Web Services","R"],"extracted_tech_keywords":["AI","cloud computing","AWS","R","Go","disruption","cloud_platforms:AWS","cloud_platforms:Amazon Web Services","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/ott-wars-set-to-intensify-as-amazon-strikes-deal-to-acquire-mgm\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10165124,"title":"China&#8217;s Zuchongzhi 3.0 Quantum Processor &#8216;Outpaces&#8217; Google Willow by Million Times","content":"Chinese researchers from the University of Science and Technology of China (USTC) have unveiled the Zuchongzhi 3.0, a superconducting quantum processor with 105 qubits, marking a significant milestone in Chinese quantum computing. The processor operates quadrillion times faster than the world’s fastest supercomputer and one million times faster than Google’s latest Willow, as per the findings published in Physical Review Letters. The Zuchongzhi 3.0 was tested with an 83-qubit, 32-layer random circuit sampling task, claiming to achieve results that would take classical supercomputers over 6.4 billion years to replicate. This achievement surpasses Google’s Sycamore processor, developed in 2019, by six orders of magnitude. The research team highlighted advancements in coherence time, gate fidelities, and readout accuracy. The processor achieves a coherence time of 72 microseconds, single-qubit gate fidelity of 99.90%, two-qubit gate fidelity of 99.62%, and readout fidelity of 99.13%. These improvements enable more complex operations and computations. Building on previous successes with the Zuchongzhi-2 and Jiuzhang photonic systems, the team continues to push the boundaries of quantum error correction and scalability. They are researching surface codes for error correction and plan to expand integration capabilities to distances of 9 and 11. The research team collaborated with institutions such as the Shanghai Research Center for Quantum Sciences and the Chinese Academy of Sciences. In December last year, China also achieved a milestone in quantum computing with the launch of the ‘Tianyan-504’ superconducting quantum computer, equipped with the 504-qubit ‘Xiaohong’ chip. The quantum computer, developed collaboratively by the China Telecom Quantum Group (CTQG), the Chinese Academy of Sciences (CAS), and QuantumCTek Co., Ltd., represented a leap in the field by surpassing the 500-qubit mark.","excerpt":"Zuchongzhi 3.0 claims to achieve results that would take classical supercomputers over 6.4 billion years.","categories":["AI News"],"tags":["China","quantum supremacy","quantum technology"],"author_name":"Sanjana Gupta","publish_date":"2025-03-04T20:52:57","publication_year":"2025","word_count":273,"keywords":["Replicate","Go","quantum supremacy","programming_languages:R","AI","RPA","Scala","Aim","programming_languages:Scala","AI research","R","China","quantum technology"],"extracted_tech_keywords":["AI","Aim","R","Go","Scala","RPA","Replicate","AI research","programming_languages:R","programming_languages:Scala"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/chinas-zuchongzhi-3-0-quantum-processor-outpaces-google-willow-by-million-times\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10086828,"title":"Google, the Tech-Savvy People Person","content":"Google is a dominant investor when it comes to AI startups. On Friday, Google announced a partnership with Anthropic, a startup founded in 2021. Anthropic makes ‘Claude’ which, much like ChatGPT, is an artificial intelligence chatbot. With the $300 million deal, Google will take an estimated 10% stake in the startup, according to the Financial Times. Anthropic, for its part, gets both a financial boost and cloud computing resources it needs. The deal could give Anthropic a valuation of roughly $5 billion, the New York Times reported. “We’re partnering with Google Cloud to support the next phase of Anthropic, where we’re going to deploy our AI systems to a larger set of people,” said Dario Amodei, CEO of Anthropic. Fascinatingly, in 2015, Amodei worked as a senior research scientist at Google Brain. During his ten-month stint at Google, he worked to extend the capabilities of neural networks safety of AI systems and laid out some key issues to prevent AI mishaps. Amodei is not the only one. Since its first investment in 2005, Google has shown financial interest in seven other companies founded by its former employees. Last year in September, Google was in talks to invest at least $200 million into Cohere Inc., according to people familiar with the matter. However, further updates on the investment have yet to be made. The AI startup was co-founded by Aidan Gomez and Nick Frosst in 2019. Prior to this, Gomez interned at Google Brain and worked on a widely cited paper, ‘Attention is All You Need‘. Similarly, Frosst worked on neural network research with AI visionary Geoffrey Hinton during his stint at Google Brain. Earlier in 2022, along with a group of investors, Google put forward $6.5 million for ‘FlexiDAO’ to grow. The company provides services that allow companies to switch to certifiable 24×7 carbon-free energy by tracking the carbon emissions of energy sources they use. Both Microsoft and Google are already customers of the service. The co-founder and CGO, Joan Collell, interned at Google, Dublin, in 2013. Ghosts of Google’s Past ‘Verily’ emerged from Google’s semi-secret R&D group ‘X’ in 2015 and managed to secure a billion-dollar funding from Google in 2022. Prior to co-founding the company, Brian Otis was the director of Google X, a laboratory set up not to invent new products but chase moonshot ideas. Moreover, Verily’s current president Stephen Gillett has been an employee at Google X for almost a decade. In 2021, ‘Nuro’, which develops autonomous vehicles for delivery services, raised $600 million from Tiger Global Management, joined by Google. The funding pushed the US-based startup’s valuation to $8.6 billion, 72% higher than a year ago. Nuro was launched by Dave Ferguson and Jiajun Zhu, two former Google engineers and has had a steady stream of investors ever since. However, unlike most of its competitors, Nuro has focused its efforts on designing a low-speed EV that transports goods instead of people. Read: Another Product Joins the Google Graveyard ‘Isovalent’ secured an investment of $40 million from Google and others in 2021. A little-known fact about the company is that the current CEO, Dan Wendlandt interned at Google in the summer of 2005. The California-based company develops open-source software for cloud applications. Isovalent was able to win over investors with its open-source software that aims to make various cloud applications safer to use. Its most popular products include ‘Cilium’, which is a software that is also used in Google’s Kubernetes engine. A major interest in medical tech startups is visible in Google’s investments throughout the years. Another such investment was made in Canadian startup ‘Arkangel AI’, which works on predictive algos to enable an environment free of preventable diseases by 2030. The company’s CEO, Jose Zea, was a part of Google’s Accelerator W20 and received funding of $196K from the tech juggernaut in the same year. ‘Niantic Inc.’, the company behind the legendary ‘Pokémon Go’, is also backed by Google, which happens to be the former employer of John Hanke, CEO of Niantic Inc. Hanke became the chief executive in 2015, prior to which he worked at Google as VP of Product Management for Maps, Street View, Earth and Local.","excerpt":"Google has shown financial interest in seven other companies founded by its former employees.","categories":["Global Tech"],"tags":["Anthropic AI","Google","Google Cloud"],"author_name":"Tasmia Ansari","publish_date":"2023-02-08T12:00:00","publication_year":"2023","word_count":694,"keywords":["Anthropic","ChatGPT","Go","artificial intelligence","Google Cloud","AI","neural network","cloud computing","kubernetes","Aim","Google","R","Anthropic AI"],"extracted_tech_keywords":["AI","artificial intelligence","neural network","ChatGPT","Anthropic","Aim","cloud computing","kubernetes","R","Go"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/google-the-tech-savvy-people-person\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":8253,"title":"A Map to Perfection: Using D3.js to Make Beautiful Web Maps","content":"Data Driven Documents, or D3.js, is “a JavaScript library for manipulating documents based on data”. Or to put it more simply, D3.js is a data visualization library. It was developed by Mike Bostock with the idea of bridging the gap between static display of data, and interactive and animated data visualizations. D3 is a powerful library with a ton of uses. In this tutorial, we’ll discuss one particularly compelling application of D3: map making. We’ll go through the common challenges of building a useful and informative web map, and show how in each case, D3.js gives capable JavaScript developers everything they need to make maps look and feel beautiful. What is D3.js used for? D3.js can bind any arbitrary data to a Document Object Model (DOM), and then, through the use of JavaScript, CSS, HTML and SVG, apply transformations to the document that are driven by that data. The result can be simple HTML output, or interactive SVG charts with dynamic behavior like animations, transitions, and interaction. All the data transformations and renderings are done client-side, in the browser. At its simplest, D3.js can be used to manipulate a DOM. Here is a simple example where D3.js is used to add a paragraph element to an empty document body, with “Hello World” text: <!DOCTYPE html> <html> <head> <meta charset=\"utf-8\"> <title>D3 Hello World<\/title> <script src=\"http:\/\/d3js.org\/d3.v3.min.js\"><\/script> <\/head> <body> <script type=\"text\/javascript\"> d3.select(\"body\").append(\"p\").text(\"Hello World\"); <\/script> <\/body> <\/html> The strength of D3.js, however, is in its data visualization ability. For example, it can be used to create charts. It can be used to create animated charts. It can be even used to integrate and animate different connected charts. D3 for Web Maps and Geographic Data Visualization But D3.js can be used for much more than just DOM manipulation, or to draw charts. D3.js is extremely powerful when it comes to handling geographical information. Manipulating and presenting geographic data can be very tricky, but building a map with a D3.js is quite simple. Here is a D3.js example that will draw a world map based on the data stored in a JSON-compatible data format. You just need to define the size of the map and the geographic projection to use (more about that later), define an SVG element, append it to the DOM, and load the map data using JSON. Map styling is done via CSS. <!DOCTYPE html> <html> <head> <meta charset=\"utf-8\"> <title>D3 World Map<\/title> <style> path { stroke: white; stroke-width: 0.5px; fill: black; } <\/style> <script src=\"http:\/\/d3js.org\/d3.v3.min.js\"><\/script> <script src=\"http:\/\/d3js.org\/topojson.v0.min.js\"><\/script> <\/head> <body> <script type=\"text\/javascript\"> var width = 900; var height = 600; var projection = d3.geo.mercator(); var svg = d3.select(\"body\").append(\"svg\") .attr(\"width\", width) .attr(\"height\", height); var path = d3.geo.path() .projection(projection); var g = svg.append(\"g\"); d3.json(\"world-110m2.json\", function(error, topology) { g.selectAll(\"path\") .data(topojson.object(topology, topology.objects.countries) .geometries) .enter() .append(\"path\") .attr(\"d\", path) }); <\/script> <\/body> <\/html> Geographic Data for D3 For this D3.js tutorial, keep in mind that map building works best with data formatted in JSON formats, particularly the GeoJSON and TopoJSON specifications. GeoJSON is “a format for encoding a variety of geographic data structures”. It is designed to represent discrete geometry objects grouped into feature collections of name\/value pairs. TopoJSON is an extension of GeoJSON, which can encode topology where geometries are “stitched together from shared line segments called arcs”. TopoJSON eliminates redundancy by storing relational information between geographic features, not merely spatial information. As a result, geometry is much more compact and combined where geometries share features. This results with 80% smaller typical TopoJSON file than its GeoJSON equivalent. So, for example, given a map with several countries bordering each other, the shared parts of the borders will be stored twice in GeoJSON, once for each country on either side of the border. In TopoJSON, it will be just one line. Map Libraries: Google Maps and Leaflet.js Today, the most popular mapping libraries are Google Maps and Leaflet. They are designed to get “slippy maps” on the web fast and easy. “Slippy maps” is a term referring to modern JavaScript-powered web maps that allow zooming and panning around the map. Leaflet is a great alternative to Google Maps. It is an open source JavaScript library designed to make mobile-friendly interactive maps, with simplicity, performance and usability in mind. Leaflet is at its best when leveraging the big selection of raster-based maps that are available around the internet, and brings the simplicity of working with tiled maps and their presentation capabilities. Leaflet can be used with great success when combined with D3.js’s data manipulation features, and for utilizing D3.js for vector based graphics. Combining them together brings out the best in both libraries. Google Maps are more difficult to combine with D3.js, since Google Maps are not open source. It is possible to use Google Maps and D3 together, but this is mostly limited to overlaying data with D3.js over Google Maps background maps. Deeper integration is not really possible, without hacking. Projections – Beyond Spherical Mercator The question of how to project maps of the 3-dimensional spherical Earth onto 2-dimensional surfaces is anold and complex problem. Choosing the best projection for a map is an important decision to make for every web map. In our simple world map D3.js tutorial above, we used the Spherical Mercator projection coordinate system by calling d3.geo.mercator(). This projection is also known as Web Mercator. This projection was popularized by Google when they introduced Google Maps. Later, other web services adopted the projection too, namelyOpenStreetMap, Bing Maps, Here Maps and MapQuest. This has made Spherical Mercator a very popular projection for online slippy maps. All mapping libraries support the Spherical Mercator projection out of the box. If you want to use other projections, you will need to use, for example, the Proj4js library, which can do any transformation from one coordinate system to another. In the case of Leaflet, there is a Proj4Leaflet plugin. In the case of Google Maps, there is, well, nothing. D3.js brings cartographic projections to a whole new level with built-in support for many different geographic projections. D3.js models geographic projections as full geometric transformations, which means that when straight lines are projected to curves, D3.js applies configurable adaptive resampling to subdivide lines and eliminate projection artifacts. The Extended Geographic Projections D3 plugin brings the number of supported projections to over 40. It is even possible to create a whole new custom projection using d3.geo.projection and d3.geo.projectionMutator. Raster Maps As mentioned before, one of the main strengths of D3.js is in working with vector data. To use raster data there is an option to combine D3.js with Leaflet. But there is also an option to do everything with just D3.js using d3.geo.tile to create slippy maps. Even with just D3.js alone, people are doing amazing things with raster maps. Vector Manipulation on the Fly One of the biggest challenges in classic cartography is map generalization. You want to have as much detailed geometry as you can, but that data needs to adapt to the scale of the displayed map. Having too high a data resolution increases download time and slows down rendering, while too low a resolution ruins details and topological relations. Slippy maps using vector data can run into a big problem with map generalization. One option is to do map generalization beforehand: to have different datasets in different resolutions, and then display the appropriate dataset for the current selected scale. But this multiplies datasets, complicates data maintenance, and is prone to errors. Yet most mapping libraries are limited to this option. The better solution is to do map generalization on the fly. And here comes D3.js again, with its powerful data manipulation features. D3.js enables line simplification to be done in browser. I want more! D3.js is not easy to master and it has a steep learning curve. It is necessary to be familiar with a lot of technologies, namely JavaScript objects, the jQuery chaining syntax, SVG and CSS, and of course D3’s API. On top of that, one needs to have a bit of design skill to create nice graphics in the end. Luckily, D3.js has a big community, and there are a lot of resources for people to dig into. A great starting point for learning D3 isthese tutorials. If you like learning by examining examples, Mike Bostock has shared more than 600 D3.js examples on his webpage. All D3.js examples have git repository for version control, and are forkable, cloneable and commentable. If you are using CartoDB, you’ll be glad to hear that CartoDB makes D3 maps a breeze. And for a little bonus at the end, here’s one of my favorite examples showing off the amazing things D3 is capable of: earth, a global animated 3D wind map of the entire world made with D3.js. Earth is a visualization of global weather conditions, based on weather forecasts made by supercomputers at the National Centers for Environmental Prediction, NOAA \/ National Weather Service and converted to JSON. You can customize displayed data such as heights for the wind velocity readings, change overlaid data, and even change Earth projection.","excerpt":"Data Driven Documents, or D3.js, is “a JavaScript library for manipulating documents based on data”. Or to put it more simply, D3.js is a data visualization library. It was developed by Mike Bostock with the idea of bridging the gap between static display of data, and interactive and animated data visualizations. D3 is a powerful […]","categories":["AI Features"],"tags":["maps"],"author_name":"AIM Media House","publish_date":"2015-11-12T07:29:45","publication_year":"2015","word_count":1502,"keywords":["Go","API","TPU","AI","ML","maps","Git","RAG","JavaScript","R","Java"],"extracted_tech_keywords":["AI","ML","RAG","TPU","R","JavaScript","Go","Java","Git","API"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/a-map-to-perfection-using-d3-js-to-make-beautiful-web-maps\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10166424,"title":"Perplexity to Raise up to $1 Billion to Double its Valuation to $18 Billion: Reports","content":"Perplexity AI, the AI-enabled search engine, is in talks to raise funds between $500 million and $1 billion, valuing the company at $18 billion, Reuters reported on Friday. This doubles Perplexity’s valuation, which stood at $9 billion in the previous funding round last December, after raising $500 million. Perplexity has been backed by NVIDIA, SoftBank Group, and Amazon founder Jeff Bezos. This development signals a growing demand for AI-powered search engine tools and applications. While Perplexity faces competition from Google’s Gemini and OpenAI’s ChatGPT Search, even Anthropic has entered the game, announcing a web search feature in its Claude chatbot. Besides primarily functioning as an AI-powered search tool, Perplexity also provides reasoning and deep research capabilities, among numerous other features. The company also recently announced that it is developing an agentic web browser called Comet. Last month, Perplexity also announced Sonar, its in-house AI model, which is now available to all Pro users. Subscribers can set it as their default model in settings, and it is said to perform on par with OpenAI’s GPT-4o. Furthermore, Deutsche Telekom, the parent company of T-Mobile, has partnered with Perplexity AI to create a next-generation AI phone. Running on a custom Magenta AI operating system, it will feature Perplexity Assistant. CEO Aravind Srinivas, in a recent podcast, said that as a company grows bigger, maintaining the same speed and agility becomes challenging. “It’s beginning to happen already a little bit. We’re not as fast as we used to be.” “We do have staging, deployment testing, A\/B testing—all that stuff’s happening, and that’s naturally slowing us down in getting things out to production widely,” he added. Perplexity AI reportedly boasts 15 million active users on its website and the app. However, a recent research study highlighted the inaccuracies in AI search engines. The Tow Center for Digital Journalism – Columbia University, performed an evaluation of search tools from ChatGPT, Perplexity, Grok, DeepSeek Search, and Google’s Gemini. Ten articles from each of the 20 publishers were selected randomly, and direct excerpts were picked from those articles as input for the AI tool. These tools were then asked to identify the article’s headline, original publisher, publication date, and URL. The study found that collectively, these search engines provided incorrect answers to more than 60% of queries. Notably, Perplexity answered 37% of queries incorrectly, while Grok 3 answered 94% of queries incorrectly.","excerpt":"The company closed a $500 million round at $9 billion valuation in December last year.","categories":["AI News"],"tags":["Perplexity AI"],"author_name":"Supreeth Koundinya","publish_date":"2025-03-21T10:43:10","publication_year":"2025","word_count":394,"keywords":["Anthropic","ChatGPT","Grok 3","Go","OpenAI","AI","GPT-4o","ML","Git","Perplexity AI","R"],"extracted_tech_keywords":["AI","ML","GPT-4o","ChatGPT","OpenAI","Anthropic","Grok 3","R","Go","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/perplexity-to-raise-up-to-1-billion-to-double-its-valuation-to-18-billion-reports\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10138707,"title":"SurveyMonkey Establishes GCC in Bengaluru to Boost AI and Product Innovation","content":"SurveyMonkey, the world’s most popular survey platform, announced its expansion into Bengaluru, India. This move underscores their intent to tap into diverse talent and build an inclusive, global workplace. According to their blog, this expansion would foster their company culture and embrace fresh perspectives from their growing international team. The company uses AI for their surveys and analysis. With GPT technology, SurveyMonkey Genius claims to generate high-quality surveys from scratch in just 30 seconds. SurveyMonkey’s Engineering Team SurveyMonkey now joins a pool of GCCs in India. The goal of this expansion is not only to boost product development capabilities and efficiency, but also capitalise on the strategic benefits of operating across multiple time zones. This approach encourages project development and approvals while ensuring a global presence that complements their teams in the US, Canada, and Europe. “SurveyMonkey is an iconic company of the internet era, and the response from India’s vast talent pool has been incredibly encouraging. In a short span, we’ve built a team of over 50 members who have quickly started contributing to SurveyMonkey’s growth. The opportunity to work on interesting problems at scale and experience career growth is a significant draw for software professionals here.” said Jeetendra Agrawal, VP of engineering in India. SurveyMonkey’s India team is actively using AI and Machine Learning to improve and speed up adoption of their range of  products. SurveyMonkey is Hiring! With the launch of their Bengaluru office, SurveyMonkey is hiring professionals to join their team. “Our Bengaluru team embodies our principles, driving innovation and making a significant impact.” per their blog. Apply here.","excerpt":"With the launch of their Bengaluru office, SurveyMonkey is hiring professionals to join their team.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","Bengaluru","GCC"],"author_name":"Aditi Suresh","publish_date":"2024-10-17T13:34:12","publication_year":"2024","word_count":263,"keywords":["Go","API","machine learning","GCC","AI","innovation","RAG","GPT","Aim","Bengaluru","R","AI (Artificial Intelligence)","llm_models:GPT"],"extracted_tech_keywords":["AI","machine learning","Aim","RAG","R","Go","API","GPT","innovation","llm_models:GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/surveymonkey-establishes-gcc-in-bengaluru-to-boost-ai-and-product-innovation\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":61169,"title":"Facebook&#8217;s New AI Models Run 5x Faster On GPUs, Outperforms EfficientNet Models","content":"Researchers from Facebook AI recently introduced a new network design paradigm known as RegNet. RegNet – or Regular Networks – is a low-dimensional design space that consists of simple, regular networks. The researchers analyzed the RegNet design space and arrived at interesting findings, which are a unique match to the current practice of network design. Facebook AI Research (FAIR) is at the forefront of deep learning techniques. The social media giant has been focused on building products on several domains. This includes open-sourcing AI tools, Building Perception, Facial Recognition with DeepFace, and DeepText, among others. Visual recognition techniques such as ResNet, LeNet, and AlexNet have gained much traction over the past few years. It helps in the advancement of both effectiveness of neural networks, as well as in the understanding of network design, in case network instantiations and design principles can be generalized and applied to numerous settings. Behind RegNet To find simple models that are easy to understand, build upon, and generalize, the researchers presented a new network design paradigm that combines the advantages of manual design and Neural Architecture Search (NAS). Neural Architecture Search (NAS) overcomes the limitations of manual network design, and helps find a suitable model within a fixed search space of possible networks. Unlike manual design, this work took advantage of semi-automated procedures and focused on designing design spaces, which help in parametrizing the population of networks. The researchers referred to this process as a design space design. Design space is a large – possibly infinite – population of model architectures. According to the researchers, the main motive behind this project is to help advance the understanding of network design and discover design principles that generalize across settings. How RegNet Works The core of the RegNet design space is composed of stage widths and depths, which are determined by a quantized linear function. The researchers designed the RegNet design space in a low-compute, low-epoch regime, using a single network block type on ImageNet dataset. In each step of the design process, the input is an initial design space, and the output is a refined design space, where each design step aims to discover design principles that yield populations of simpler or better performing models. The primary tool used by the researchers for analyzing design space quality is the error empirical distribution function (EDF). They used a relatively unconstrained design space to build RegNet, known as AnyNet, where the widths and depths vary freely across stages. The researchers said, “We propose to design network design spaces, where design space is a parametrized set of the possible model architecture, and we characterize the quality of a design space by sampling models and inspecting their error distribution.” Contributions In This Project Here are some of the contributions mentioned by the researchers of this project:- According to the researchers, the RegNet design space has simpler models, is easier to interpret, and has a higher concentration of good modelsAn important property of the design space design in this project is that it is more interpretable, and can lead to interactive learning insightsThe researchers compared the top REGNET models to existing networks in various settings. This showed that simple RegNet models achieve surprisingly good results.REGNET models lead to considerable improvements over standard RESNE(X)T models in all metrics Wrapping Up According to the researchers, designing network design spaces is a promising avenue for future research. Under comparable training settings and flops, the RegNet models outperform the popular EfficientNet models, while being up to 5X faster on GPUs. Read the paper here.","excerpt":"Researchers from Facebook AI recently introduced a new network design paradigm known as RegNet. RegNet – or Regular Networks – is a low-dimensional design space that consists of simple, regular networks. The researchers analyzed the RegNet design space and arrived at interesting findings, which are a unique match to the current practice of network design.  […]","categories":[],"tags":["EfficientNet-EdgeTPU","Facebook AI","Facebook AI research","GPU"],"author_name":"Ambika Choudhury","publish_date":"2020-04-12T10:00:00","publication_year":"2020","word_count":589,"keywords":["Go","TPU","Facebook AI","AI","neural network","active learning","ResNet","Aim","deep learning","Facebook AI research","EfficientNet-EdgeTPU","EfficientNet","R","GPU"],"extracted_tech_keywords":["AI","deep learning","neural network","Aim","TPU","R","Go","ResNet","EfficientNet","active learning"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/facebooks-new-ai-models-run-5x-faster-on-gpus-outperforms-efficientnet-models\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":4500,"title":"The Data Scientist in 2014","content":"As we usher in 2014, and the business world gears up to face the challenges of another year, once again the buzz is all about data analytics and big data.  Being where we are in the analytics training space, we can definitely hear how loud this buzz really is. We notice a growing demand for data analysts and are seeing a greater number of professionals and students from varying sectors investing in gaining data analysis skills. Indeed there seems to be a heightened awareness about the field as a lot more companies adopt data analytics initiatives. These companies have recognized the importance of harnessing their data and using the insights derived to make crucial operational and strategic decisions. But guess what? Most companies or even those training to be data scientists are still not very clear as to what the role of the data scientist truly entails. Well the truth is, this is all still very new and just like its industry, it will continue to develop and evolve over the next few years. For the time being though what’s clear is that the role will have a strong emphasis on technology. Other than a few large companies like Google or eBay, most companies are still in the process of setting up the infrastructure to handle big data, and so technology skilled people are going to be very important. Most of the data science processing is still very report oriented so someone with data management\/data mining plus IT skills will be very valued. As the big data and analytics industry evolves, companies will get more secure and feel more confident about the power of their data. They will then begin to recruit many more people with statistics and analytics skills, to fill even traditional marketing, HR and finance roles. This year we will also see the role of the data scientist become more specialized. Companies will invest in data science teams rather than data scientist individuals, with the team having people with IT, statistics, and business experience. Here is a quick look at some of the more specialized roles we might be looking at this year: 1. Data Explorers\/ Data Hygienists: They will be required to sift through huge amounts of data, both internal and external and discover what is significant and useful.  They have to ensure that clean and accurate data is made available for analysis and manipulation throughout the life cycle of the data analysis. 2.  Data Engineers: They will work primarily in the big data space and will build and design tools and programs to maintain and support the data at hand.  They will also be responsible for the day-to-day upkeep and maintenance of big data infrastructure. 3. Data Change Agents: These people will work towards initiating changes within the organization based on data analytics insights derived by the data scientists.  They will introduce process improvements in operational divisions, while managing the change mindset of the employees. If you have made the decision to become a data scientist, you probably are really in the right place, at the right time. The scope for data scientists seems humongous and all predictions say, that the scenario will only get better. So go back to analyzing that data and be rest assured that indeed, your career is set to explode.","excerpt":"As we usher in 2014, and the business world gears up to face the challenges of another year, once again the buzz is all about data analytics and big data.  Being where we are in the analytics training space, we can definitely hear how loud this buzz really is. We notice a growing demand for […]","categories":["IT Services"],"tags":[],"author_name":"Sarita Digumarti","publish_date":"2014-01-17T07:21:14","publication_year":"2014","word_count":550,"keywords":["big data","Go","data science","programming_languages:R","AI","programming_languages:Go","analytics","GAN","R"],"extracted_tech_keywords":["AI","data science","analytics","R","Go","big data","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/the-data-scientist-in-2014\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10013924,"title":"UP Govt. To Launch AI &#038; ML Courses For Students Of The State","content":"In an attempt to bridge the talent gap in the corporate sector, the UP government has partnered with the Austin University of the US to launch courses on artificial intelligence and machine learning for cybersecurity management and data analytics. Talking about the launch, Uttar Pradesh Minister Siddharth Nath Singh, stated to the media that an MoU had been signed between the technical education department of the state government and the Austin University to launch these certification programmes for students of the state. On successful completion of the courses, the students will receive a certificate or a degree from the Austin University, which can be useful for landing a job in the corporate sector. Explaining better, Singh said that these courses would be a game-changer in the field of higher and technical education, where the students of the state will get a certificate from one of the best US universities while residing in any UP city. As compared to the progressive states of the country, UP has majorly lacked such advanced courses; thus, Singh believes that these certifications programmes will make these learners useful for global brands. He said that the government’s all over the world, as well as the global corporate houses, are always in need for people who are skilled with these advanced courses, and therefore with this launch Uttar Pradesh will be able to fill that talent gap. According to the news, these degree courses will be of two-year duration, whereas the diploma courses will be of one-year. The courses have been designed for online learning along with hybrid online live sessions, while later if required the Austin University could open its campus, where the learners can do their internship. Considering many students struggle for jobs despite having a degree in conventional courses, the aim with these advanced courses is to equip the students with industry-required skills — cybersecurity management, artificial intelligence and data analytics, and make them industry-ready and employable in the corporate sector.","excerpt":"In an attempt to bridge the talent gap in the corporate sector, the UP government has partnered with the Austin University of the US to launch courses on artificial intelligence and machine learning for cybersecurity management and data analytics. Talking about the launch, Uttar Pradesh Minister Siddharth Nath Singh, stated to the media that an […]","categories":["AI News"],"tags":["AI Certifications","limitations of AI"],"author_name":"Sejuti Das","publish_date":"2020-12-11T10:43:03","publication_year":"2020","word_count":327,"keywords":["Go","artificial intelligence","machine learning","programming_languages:R","AI","AI Certifications","programming_languages:Go","Aim","analytics","R","limitations of AI"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","analytics","Aim","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/up-govt-to-launch-ai-ml-courses-for-students-of-the-state\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":45323,"title":"Will AI-Driven Enterprise Search Engine Push ThoughtSpot Ahead In The Business Intelligence Market?","content":"After its recent funding round which closed at $248 million, ThoughtSpot has become the fastest-growing analytics firm with a unicorn valuation of $1.95 billion. The company has received over half a billion in funding rounds alone spanning across five funding rounds in the last seven years. Founded in 2012, the company’s growth in funding has been astounding, proving that investors see a lot of value in it. The company has been doing well in terms of customer acquisition as well with more than half of the Fortune 10 and companies like Walmart, Hulu, Daimler, 7Eleven, PetCo, and Rolls Royce using its data analytics solutions. The company also has an average sale price of $250,000, contributing to sales growing 195% in the first half of the fiscal year. What Value Does ThoughtSpot Create To Justify Such Funding ThoughtSpot’s feature of a natural language processing interface can help users with complex questions and augmented intelligence, and this has helped it shine when compared to others. The offering has two different things happening at the same time in its dashboard. One is a simple search engine that uses machine learning to answer enterprise-specific questions, and the second is using AI to automate the process of analytics. This way businesses can perform a simple Google-like search in natural language to quickly analyse billions of rows of data, and use artificial intelligence to get relevant insights. Search capabilities can power users to find answers to their queries, while artificial intelligence enables insights on queries business may not be aware, based on patterns in search history, data anomalies and other trends. One example based on such search queries is how bankers can instantly search which types of loans have a higher default rate, or a phone retail firm that can search which age demographic buys the most number of iPhones from the firm. The advantage of such search-based analytics is that it does not require any extra expertise, and all users can contribute to enhancing business intelligence. Analysts say the technology is as an evolution that goes beyond what the technology offered by a leading analytics player Tableau, which was acquired for $15.7 billion by Salesforce in June 2019. Combining Search Analytics With AI ThoughtSpot’s technology is defined by experts as augmented analytics that is supported by huge scale AI-driven machine learning and enterprise-grade search engine-like capabilities. This is a unique value proposition that a company brings to business intelligence space, where there is usually no focus on ‘search’ option for enterprises. Tableau and PowerBI include search-driven analytics but are only limited or included as a side feature. Compared to ThoughtSpot, experts have also called Tableau and PowerBi as “legacy” business intelligence tools which are based on report builders and report viewers. Most business intelligence applications are built using point-and-click drag-and-drop graphical user interfaces, yet still need the end user training and long set up times. It is estimated that only a small percentage enterprise decision-makers leverage data analytics directly as they are not trained in the area, and have to depend on trained analysts for business intelligence. This is a challenge for wider adoption of business intelligence among varying levels of business users. Adding Capabilities With Strategic Partnerships In the last year or so, we have seen a lot of business intelligence software vendors improve their capabilities with mergers and acquisitions. For example, Tableau recently agreed to be acquired by Salesforce so it can compete with Microsoft Power BI whereas Qlik acquired data management vendors Podium Data and Attunity in the last one year. The story does not end here as Looker made a deal to be acquired by Google, Sisense agreed to merge business with Periscope Data, and Alteryx made a deal to buy ClearStory. For ThoughSpot, it has enhanced its technology offerings with strategic partnerships. Since the end of 2018, ThoughtSpot has partnered with Google Cloud, DataRobot, Alteryx, Snowflake, Carahsoft and Dell to build the same level of product features others have achieved through mergers and acquisitions, particularly capabilities that relate to cloud. Overview ThoughtSpot’s yet another funding round, which pushed its value to almost 2 billion, shows that investors are happy to write big checks for new innovation in BI such as AI and search-driven analytics. It is also a sign how new data capabilities are transforming traditional analytics workflows, riding the gap between trained and untrained employees to contribute to business intelligence. Of course, the integration of AI and machine learning is propelling this evolution too, and we may see stiff competition among all the leading BI vendors in this direction in the coming future. The large funding of ThoughtSpot is showcasing that with the amount of data businesses are collecting from IoT, cloud, SaaS, business users dealing merely with charts and dashboards will not be enough and there has to be AI and natural language search involved. No wonder, ThoughSpot reached unicorn status in such a short span of time.","excerpt":"After its recent funding round which closed at $248 million, ThoughtSpot has become the fastest-growing analytics firm with a unicorn valuation of $1.95 billion. The company has received over half a billion in funding rounds alone spanning across five funding rounds in the last seven years. Founded in 2012, the company’s growth in funding has […]","categories":["Deep Tech"],"tags":["Business Intelligence","Mergers and Acquisitions","mergers and acquisitions business intelligence","Snowflake","Startups","strategic salesforce","thoughtspot"],"author_name":"Vishal Chawla","publish_date":"2019-09-03T03:13:40","publication_year":"2019","word_count":820,"keywords":["mergers and acquisitions business intelligence","Go","artificial intelligence","thoughtspot","machine learning","AI","ML","RAG","Aim","Startups","analytics","Business Intelligence","Mergers and Acquisitions","R","strategic salesforce","Snowflake"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","analytics","Aim","RAG","Snowflake","R","Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/thoughtspot-startup-analytics-market\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10042944,"title":"Tata Communications’ Alok Bardiya On How Pandemic Drove IoT Adoption","content":"The rise of affordable devices, analytics and cloud computing is expanding the horizons of India’s IoT market. Alok Bardiya, Head of Internet of Things at Tata Communications, is at the helm of his company’s efforts to scale new heights. An alumnus of Wharton Business School and IIT Delhi, he is responsible for shaping and scaling Tata Communications’ IoT business. Alok comes with over 20 years of experience in the tech industry. In an exclusive interview with Analytics India Magazine, he spoke about the driving force behind India’s IoT growth and the road ahead for Tata Communications. “For our India IoT business, even before the pandemic, we were at a stage where CEOs were keen to drive IoT implementation and urging for quick deployment. The impact of COVID 19 has accelerated the demand; however, the challenge lies in getting the IoT ecosystem ready to deliver devices at scale,” Alok said. Excerpts: AIM: What are your thoughts on India’s booming IoT market? Alok: While India was already in the process of establishing itself as a significant player in the digital economy, the pandemic accelerated digital transformation for all. The need to connect, meet and work remotely across every business has increased our reliance on automation and the Internet of Things for data insights. As a result, the increase in the adoption of IoT-enabled devices and the data generated from these devices enable businesses to gain organisational insight, make quick and accurate decisions, and improve efficiency and productivity. As companies diversify their supply chains amid the pandemic, India is expected to become a hotspot for the IoT deployment combined with the government’s push on the Digital India mission. As a result, we witnessed an increased number of factories leveraging IoT for supply chain monitoring and asset management. AIM: How did pandemic impact your business lines? Alok: From an enterprise perspective, we have seen increased traction during the pandemic for our Connected Operations proposition, which comprises Connected Worker Solutions, particularly Safety Watch and SafePass Card. These solutions track and locate employees across locations, monitoring their health and safety parameters — instrumental in preventing over 20 plus critical incidents so far (watch the demo here). We have demand from the FMCG, metal & mining, chemicals, cement, auto, oil & gas, pharma sectors and expect it to trickle in from the travel and the service sector. Asset management is another area. Several FMCG and manufacturing companies during several stages of the lockdown were all keen to keep track of the location and health of their assets. With minimal manual intervention through remote assistance, our IoT-based asset tracking solution provides a unique asset management system that allows enterprises to track their assets, however widely distributed. Built-in intelligence converts this data into actionable insights and makes it available through a simple, easy-to-use application – anytime, anywhere. We also have video-based IoT used for object recognition (geo-fencing), people recognition, and process adherence in the last phases. In addition, given the surge in demand in the market, we also launched an IoT fabric proposition which is a single platform to manage connectivity and data from any IoT device and enables users to manage the lifecycle of their entire IoT use-case. AIM: Tell us about the challenges you have faced during the pandemic. How did you overcome them? Alok: There are several challenges that we face while deploying an IoT framework for a company. As you would know, IoT needs physical implementation, cloud application and analytics. During the initial days of the pandemic, we faced some roadblocks regarding the physical implementation of IoT due to lockdown guidelines, chip shortages, and other hardware materials. As supply chains had been hit, companies resorted to Make in India solutions. One such example we are proud of is our streetlight solution hardware manufactured in India. Lastly, the challenge lies in getting the IoT ecosystem ready to deliver devices at scale. Fortunately for us, in the past year, we laid out the groundwork. As a result, we can cater to enterprises that were earlier looking for 100 to 200 devices but are currently looking beyond 1,000 to 10,000 devices to transform their business. To expedite and enable enterprises in their digital journey, we launched an IoT Fabric – the solution offers basic connectivity and data platform infrastructure for enterprises on a usage-based model, thus enabling them to do quick pilots followed by scale-ups. AIM: How small manufacturing companies are responding to IoT technologies? Alok: Small manufacturers play a critical role in keeping the world’s supply chain moving and are now slowly embracing Industry 4.0. While we are seeing increasing demand from mid to large enterprises, the small players are adopting niche solutions in the ecosystem. The increasing use of digital devices will transform the operations time along with service history time and at the point of use. The ripple effect of having new age technologies is being felt across enterprises, irrespective of size. As such, all enterprises are stepping up their digital spending as that’s the only way to survive and grow. AIM: Tell us about your smart city solutions. Alok: Smart cities are a major focus for us, and with growing customer interest, we have solutions for smart lighting and smart metering systems, including water, gas, and electricity metering. Our end-to-end smart cities solutions are powered by state-of-the-art devices running on our dedicated IoT network, platform, and applications. Our smart cities framework focuses on – intelligent and sustainable smart city solutions, integrated infrastructure, and collaboration solutions. We are in the process of expanding to more cities over the next two years along with areas such as special economic zones and highways. AIM: What is the future of the IoT market in India? Alok: Smart Digital workplace technologies like the connected worker solutions, including contact tracing, contactless operations, and remote monitoring, became popular with enterprises to manage work remotely. IoT adoption is now being prioritised and pulled forward. A recent report by Zinnov noted India as an emerging market for IoT in the world. Tata Communications IoT entered the space very early on. We are experts in this field. As the usage of these technologies increases, it will only extend into newer domains such as the Man-Machine-Material interface. That is the narrative we are driving to ensure an all-encompassing offering across the supply chain. Man-Machine-Material offers diverse use cases across employee safety & efficiency, machine & facility management and material & equipment handling. We believe that this ability to flexibly add IoT-enabled use cases on existing infrastructure will go a long way in driving IoT adoption in India. Enterprises are also going to increase their dependency on IoT analytics for data-driven decision-making, which aligns with our business focus. AIM: What are we lacking in terms of the infrastructure and policy support? Alok: Indian market is making progress, but we are yet to fully monetise the value of real-time data. Until the market is cost-focused, the business model’s viability will always remain a challenge. Secondly, we have seen several small players enter the market offering quick fixes at much lower prices. They drive the pricing benchmarks to unviable levels to sustain and sell and then eventually shut shop. The slow adoption of 5G in India is yet another reason. However, recent announcements by telecom players on starting 5G trials in India are likely to lead to cheap and ubiquitous network connectivity after the network rolls out, ensuring last-mile connectivity. In the end, for IoT to make substantial inroads in a business, a CEO needs to have a clear vision of what he wants to achieve with the IoT deployment. AIM: Can you share a few case studies with us? Alok: Our Connected Worker Solution has been deployed by a multinational manufacturing company. We deployed 3,000 safety wearables and 75,000 SafePass across 20 plus locations. Apart from enhancing worker wellbeing and minimising workplace risks, the connected employee solution has also helped the manufacturing company in making data-driven decisions for enhanced efficiency and security. We have also deployed our smart lighting solutions for Jamshedpur Utilities and Services Co. Ltd (JUSCO) partnered with Mahanagar Gas Limited to deploy smart gas meters in Mumbai and supported Indraprastha Gas Limited with a pre-paid smart gas metering system. AIM: What’s next for Tata Communications? Alok: The Tata Communications IoT India portfolio is geared towards meeting this demand and has witnessed major tailwinds last year. We saw an increase in our margins across customers, devices bought from March 2020 to March 2021. We had 12 new customers added, 75 627 devices bought during the year, and ventured into new sectors. Along with this scaling, we have also continued to expand our products and solution portfolio. Tata Communications’ Internet of Things has been focused on the Indian market and is now poised to explore international opportunities.","excerpt":"The rise of affordable devices, analytics and cloud computing is expanding the horizons of India’s IoT market. Alok Bardiya, Head of Internet of Things at Tata Communications, is at the helm of his company’s efforts to scale new heights. An alumnus of Wharton Business School and IIT Delhi, he is responsible for shaping and scaling […]","categories":["AI Features"],"tags":["Interviews and Discussions","tata communications"],"author_name":"kumar Gandharv","publish_date":"2021-07-06T11:00:00","publication_year":"2021","word_count":1461,"keywords":["Go","API","AI","cloud computing","ETL","tata communications","Git","RAG","Aim","analytics","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","analytics","Aim","RAG","cloud computing","R","Go","Git","API","ETL"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/tata-communications-alok-bardiya-on-how-pandemic-drove-iot-adoption\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":67689,"title":"How Ozonetel Is Leveraging Its AI-Powered Cloud Telephony Platform","content":"There is growing adoption of cloud telephony to boost organisational efficiency and the growing integration of advanced technologies like artificial intelligence. Businesses are currently using this cloud solution to enable remote working solutions for their call centre employees. Analytics India Magazine connected with Chaitanya Chokkareddy, Chief Innovation Officer of Ozonetel, to know more about cloud telephony. Ozonetel is an AI-based contact centre solutions provider, which has fast risen to become a prominent player in cloud-based contact centre solutions leveraging AI capabilities. According to Chaitanya, the cloud contact centre in future will become a harmonious blend of AI and humans working together. The AI will fill in for monotonous and repetitive tasks while humans will concentrate on tougher, more challenging tasks. Among the latest innovations at Ozonetel, the firm recently launched an AI-powered Speech Analytics dashboard for call centres to gauge customer sentiment for the duration of a call automatically. To know more, here are the excerpts from the conversation: – AIM: What are the various AI technologies that Ozonetel leverages? Also, tell us about Ozonetel’s recently launched AI-powered speech analytics dashboard for call centres. Chaitanya Chokkareddy: Being in the communication space, we mainly work on speech AI and NLP. By utilising AI, the cloud contact centre will become a harmonious blend of AI and humans working together. The AI will fill in for monotonous and repetitive tasks while humans will concentrate on tougher, more challenging tasks. Ozonetel’s recently launched Speech Analytics platform is a useful tool for both customer success and marketing teams. It helps to convert the vast, unstructured data that call centres receive on a daily basis into valuable insights. Currently, it analyses customer sentiment, conversation speed, conversation volume, and customer demographics, such as gender recognition. It can be used within call centres for quality assurance and faster monitoring. The real value will be in preventing escalations and improving overall customer satisfaction. By giving useful insights to both product development and marketing teams, it is also set to raise the total value of call centres within a business. AIM: What are the different features developers can leverage on Ozonetel’s Voice Bot platform? Chaitanya Chokkareddy: Since the beginning, we have always been a developer-focused platform. All features are available to developers in easy-to-use APIs. Even for our voice bot platform, developers can access speech to text, text to speech, speaker emotions and other features accessible with APIs. Using these developers can design innovative flows for their customers. AIM: How do you see the evolution of your cloud telephony platform KOOKOO as well as the cloud contact centre solution Cloudagent? Chaitanya Chokkareddy: A few years down the line, KOOKOO and Cloudagent will become AI-first. All communication will start by passing through our AI system, which will then decide if human intervention is needed or not. In many cases, customers will not even know they are dealing with an AI. It will be pretty seamless. AIM: Tell us about the current solutions and tech stack that Ozonetel is offering? Chaitanya Chokkareddy: Ozonetel currently offers a plug and play omnichannel contact centre solution, a cloud telephony platform, and AI-based contact centre solutions. Our plug and play solution is used to deploy a complete call centre for sales or support quickly. Businesses are currently using this cloud solution to enable remote working for their call centre employees. Our platform is used by enterprises, startups, marketing agencies, and developers to build customised solutions in conjunction with our virtual numbers, telephony platform, IVR, and dialers. In addition, customers can develop AI-based contact centre solutions includes voice bot platforms and speech analytics for the call centres. AIM: How are customers finding value in Ozonetel’s products? Chaitanya Chokkareddy: Ozonetel helps contact centres go live in a day with a reliable platform that improves customer experience and agent productivity at a lower total cost. Our flexible, reliable cloud-based communication solutions make CRM, helpdesk, and auto-dialer integration easy and self-service simple. We support over 150 million sales and support calls on our platform annually. Today businesses access all these call centre tools remotely to ensure business continuity during lockdowns. Customers also use our virtual numbers and number masking solutions to manage direct employee-customer interactions. Our open APIs allow businesses to use our telephony platform, IVR and Dialers in innovative ways to engage customers with missed call campaigns, SMS campaigns, and radio integrations. AIM: How do you see the road ahead in cloud telephony technologies, and where does Ozonetel fit in this? Chaitanya Chokkareddy: The future is going to be about omnichannel and AI. Customers will reach out to businesses on multiple channels like phone, chat, WhatsApp, etc. and will expect a quick resolution. Since multiple communication and marketing channels are already integrated into our platform, and we are doing the right research in AI, so we believe that we are placed well for the future.","excerpt":"There is growing adoption of cloud telephony to boost organisational efficiency and the growing integration of advanced technologies like artificial intelligence. Businesses are currently using this cloud solution to enable remote working solutions for their call centre employees.  Analytics India Magazine connected with Chaitanya Chokkareddy, Chief Innovation Officer of Ozonetel, to know more about cloud […]","categories":["AI Features"],"tags":["cloud business intelligence solutions","Interviews and Discussions","Voice Analytics"],"author_name":"Vishal Chawla","publish_date":"2020-06-19T14:00:00","publication_year":"2020","word_count":806,"keywords":["Go","cloud business intelligence solutions","artificial intelligence","AI","ML","Voice Analytics","Scala","RAG","NLP","Aim","analytics","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","artificial intelligence","ML","NLP","analytics","Aim","RAG","R","Go","Scala"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-ozonetel-is-leveraging-its-ai-powered-cloud-telephony-platform\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":58840,"title":"How Graph Networks Are Solving Complex Physics Problems","content":"Building high-quality simulators for physical phenomena require exceptional engineering efforts combined with expensive computational resources. This reason alone barricades many kinds of research. Even if one manages to come up with a decent method, they are often inaccurate or fail to approximate the underlying physics. As the researchers are looking out of alternative solutions, they have stumbled upon the most obvious solution of this century — machine learning. However, machine learning approaches haven’t been adopted because of their difficulty in dealing with large number of parameters that a typical fluid simulation or as a matter of fact, any physics based simulation would have. The complex dynamics make it tough for a model to learn to simulate. So, addressing the aforementioned challenges, the DeepMind researchers in collaboration with Stanford University have presented a general framework for learning simulation using graph networks (GN), and a single model implementation that yields state-of-the-art performance across several challenging physical domains, involving fluids, rigid solids, and deformable materials interacting with one another. How Graph Networks Solve Physics via paper Graph networks (GN) have been effective at learning forward dynamics that involve multiple interactions. A network typically consists of an input graph mapped onto an output graph which has the same structure but varies in node, edge and other attributes at the graph-level. These nodes and edges are responsible for passing information throughout the network. The approach here focuses on particle-based simulation, which is used widely in computational fluid dynamics, computer graphics. As illustrated above, the procedure can be split into 5 steps: (a): The graph network here, predicts future states represented as particles using its learned dynamics model and a fixed update procedure(b): The dynamic model then uses an “encode-process-decode” scheme, which computes dynamics information.(c): The ENCODER, as shown above then constructs latent graph from the input state(d):In the next step, the PROCESSOR performs a few rounds of learned message-passing over a series of latent graphs.(e): The DECODER then extracts The dynamics information and is then extracted by the DECODER from the latent graph. The authors claim that the experimental results with GNs have shown that they can learn to simulate the dynamics of fluids and other materials by interacting with one another, using tens of thousands of particles over thousands of time steps. While here they have focused on mesh-free particle methods, this approach, assert the authors, can also be used for finite-element methods. They believe that there are also natural ways to incorporate stronger, generic physical knowledge into this framework, such as Hamiltonian mechanics and rich, architecturally imposed symmetries. Future Direction Graph networks have the ability to learn and simulate rigid body and robotic control systems, as well as non-physical algorithmic execution. To realise advantages over traditional simulators, the authors suggest that future work should explore how to parameterise and implement graph networks computations more efficiently, and exploit the ever-improving parallel computing hardware. The successful demonstration of this approach can be a lead up to even more sophisticated generative models and can help the future researchers in developing an AI-based toolkit that has a knack for physical reasoning. Currently, learning simulations from data is an active area of study with applications in physics and graphics. A learned simulator has a great potential to be more efficient compared to an engineered simulator for predicting complex phenomena and learning parts of a fluid simulator for faster prediction. Recreating near realistic simulations is more math than art. Every rise and fall and turn and twist of the material has to be captured in a right way. These simulations typically are the results of partial differential equations that usually take up massive amounts of computation. And, we just can’t tweak the hardware (shrink transistors) for reducing the time consumed, a theory complemented by Moore’s law. Traditional fluid dynamic simulators offer a lot of options and are industry standard. So, we cannot expect machine learning based solutions to replace the traditional methods any time soon. In fact, they should be looked at as an augmentation for the existing methodologies.","excerpt":"Building high-quality simulators for physical phenomena require exceptional engineering efforts combined with expensive computational resources. This reason alone barricades many kinds of research. Even if one manages to come up with a decent method, they are often inaccurate or fail to approximate the underlying physics.  As the researchers are looking out of alternative solutions, they […]","categories":[],"tags":[],"author_name":"Ram Sagar","publish_date":"2020-03-17T14:00:24","publication_year":"2020","word_count":670,"keywords":["Go","machine learning","TPU","programming_languages:R","AI","programming_languages:Go","Aim","R"],"extracted_tech_keywords":["AI","machine learning","Aim","TPU","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/graph-networks-knn-physics-simulation-fluid-dynamics\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":29699,"title":"IBM To Acquire Red Hat For $34 Billion, Becomes World&#8217;s Biggest Hybrid Cloud Provider","content":"Ginni Rometty, Chairman, President, and CEO of IBM (right), with James M Whitehurst, CEO of Red Hat. IBM and Red Hat, two of the world’s leading provider of open source cloud software, announced on Monday that they had reached a definitive agreement under which IBM will acquire all of the issued and outstanding common shares of Red Hat for approximately $34 billion. “The acquisition of Red Hat is a game-changer. It changes everything about the cloud market,” said Ginni Rometty, IBM Chairman, President and Chief Executive Officer, in a statement. “Most companies today are only 20 percent along their cloud journey, renting compute power to cut costs,” she said. “The next 80 percent is about unlocking real business value and driving growth. This is the next chapter of the cloud. It requires shifting business applications to hybrid cloud, extracting more data and optimizing every part of the business, from supply chains to sales.” IBM will become the world’s no. 1 hybrid cloud provider, offering companies the only open cloud solution that will unlock the full value of the cloud for their businesses. This acquisition brings together the best-in-class hybrid cloud providers and will enable companies to securely move all business applications to the cloud. Companies today are already using multiple clouds. However, research shows that 80 percent of business workloads have yet to move to the cloud, held back by the proprietary nature of today’s cloud market. This prevents portability of data and applications across multiple clouds, data security in a multi-cloud environment and consistent cloud management. “Open source is the default choice for modern IT solutions, and I’m incredibly proud of the role Red Hat has played in making that a reality in the enterprise,” said Jim Whitehurst, President and CEO, Red Hat, in a statement. “Joining forces with IBM will provide us with a greater level of scale, resources and capabilities to accelerate the impact of open source as the basis for digital transformation and bring Red Hat to an even wider audience – all while preserving our unique culture and unwavering commitment to open source innovation.” IBM and Red Hat will be strongly positioned to address this issue and accelerate hybrid multi-cloud adoption. Together, they will help clients create cloud-native business applications faster, drive greater portability and security of data and applications across multiple public and private clouds, all with consistent cloud management. In doing so, they will draw on their shared leadership in key technologies, such as Linux, containers, Kubernetes, multi-cloud management, and cloud management and automation. IBM’s and Red Hat’s partnership has spanned 20 years, with IBM serving as an early supporter of Linux, collaborating with Red Hat to help develop and grow enterprise-grade Linux and more recently to bring enterprise Kubernetes and hybrid cloud solutions to customers. These innovations have become core technologies within IBM’s $19 billion hybrid cloud business. Between them, IBM and Red Hat have contributed more to the open source community than any other organization.","excerpt":"IBM and Red Hat, two of the world’s leading provider of open source cloud software, announced on Monday that they had reached a definitive agreement under which IBM will acquire all of the issued and outstanding common shares of Red Hat for approximately $34 billion. “The acquisition of Red Hat is a game-changer. It changes […]","categories":["AI News"],"tags":["Cloud Computing","IBM","red hat"],"author_name":"Prajakta Hebbar","publish_date":"2018-10-29T07:48:20","publication_year":"2018","word_count":494,"keywords":["red hat","programming_languages:R","AI","innovation","digital transformation","Git","automation","Cloud Computing","IBM","GAN","R","kubernetes"],"extracted_tech_keywords":["AI","kubernetes","R","Git","GAN","digital transformation","automation","innovation","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/ibm-red-hat-hybrid-cloud\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":57793,"title":"How To Crack A Data Science Interview","content":"The data science and analytics sector in India has witnessed a sharp increase in demand for highly-skilled professionals who understand both the business world as well as the tech world. Data Science is considered one of the most lucrative jobs in the industry right now. However, the industry is still riddled with a lot of challenges in terms of talent and which is why organisations have started pouring a substantial amount of money in building their data science and analytics team. Organisations regardless of their positions have been using data science and analytics to garner insights from data. Need For Young Professionals: The Data Scientist career is titled as the “Hottest Career of 21st Century” by Harvard Business Review and this position has proved to be one of the most appealing as well as wanted by the job seekers. However, there has been a potential shortage of Data Scientists in the field. The reason behind this is the technological challenges that are limiting the skills of the employees. Most of the senior-level management has started off from software or coding designations since the sector wasn’t evolved enough to encompass the designation of a data scientist. An entry-level data scientist is someone who has less than four years of experience working as a business analyst with knowledge in Python. The entry-level role also applies to senior software engineers looking for opportunities to work in analytics and machine learning projects. Basic Guidelines For A Good Resume: Before we proceed to the interview questions, here are a few tips about prepping your resume and organising some key talking points that would showcase you in a positive light. Make sure you include essential details like email, contact number, and a professional picture.Obviously, mention your educational backgroundMention the key projects related to machine learning and data science.List down recent and relevant work experience with details about the programming languages\/ technology used and direct business outcomes (if applicable).List down achievements like hackathons, bootcamps and certifications.Include Professional Channels like Kaggle, Github, and LinkedIn accounts.Try to keep your resume short and under one page. Cracking The Data Science Interview: Today, this article will take a more in-depth look at what it takes to crack a data science interview for a person with more or less than five years of work experience. Here, we list down 10 important points following which an aspirant will surely achieve his\/her dream. Mastering Programming Language: Talking about the primary skills, for a data science interview one must have a keen knowledge of fundamental topics like distributed computing and data structure, languages like Python, R and SQL. According to a survey, Python continues to be the most popular language in the industry in 2019. Other essential languages falling out of a Data Scientist’s toolbox are R, SQL, SAS, among others. Implementing Programming Language Algorithmically: After mastering a programming language like Python or R, one must try to implement the language algorithmically. This will not only help an aspirant to understand how to create and deploy the complex machine learning algorithms but also help to master the language more quickly. Get Familiar With Production Deployment Environments: Most of the organisations these days have been using the cloud as an infrastructure model. In other words, we can say that the cloud has been ruling the data science space. Popular cloud vendors like Google Cloud Platform (GCP), Azure, Amazon Web Service (AWS) have made it easy for the data scientists to quickly set up a machine learning environment and start working without thinking of the huge pile of generated data. Getting Hands-on Experience: This is the most crucial part of this journey. Hands-on experience allows for a practical depth of knowledge which will not only help a data science applicant to understand the scenario better but also help to demonstrate the skills. Taking up a data science project and trying to build and develop a model provides in-depth knowledge to the domain where the candidate wants to work on. Your Digital Presence Speaks Louder than Words: The data science community is growing in almost every social media platforms such as Twitter, LinkedIn, Facebook, among others. With everything going digitalise these days, one can start his\/her own blog or write on LinkedIn posts where s\/he can share the knowledge as well as showcase new skills for the community to get noticed. Besides these, a candidate must also try to participate in bootcamps, hackathons, contribute to open-source projects in GitHub or participate in other such competitions. Showcase Your Knowledge and Know-How About the Industry: Candidates must keep themselves updated with the latest news happening around the space. This will help them to keep the same pace as the emerging technologies. Also, identifying new trends and forward-looking opinions will provide great help in an interview. Possible Questions Asked: While in a data science interview there are a few possible questions that the interviewer might ask the applicant. For instance, What processes did you make more efficient at work? or can you cite one example which you thought was completely outside the box and helped in turning around the project? The motive behind these questions has the possibility to entail details about the current data science team structure, projects they are involved in and how the right projects are prioritized. Don’t Forget To Ask Questions: If a candidate wants to be a step ahead than the other applicants, s\/he should ask insightful questions. The reason behind this is while asking questions, it helps the interviewer to evaluate the worth of the candidate and how well s\/he fits in at the company. Be Good With the Basics of Data Science: To crack a data science interview, one must take a deep dive into the basic topics such as mathematics, statistics, programming languages, basics of Business Intelligence and of course, the machine learning algorithms. Knowledge of Visualisation: In the journey of data science, one has to have the knowledge of any popular data visualisation tool. For instance, Tableau, Google charts, Qlik, among others are popular among the organisations and knowing how to use them will surely help to be a good data scientist. Great Learning provides one of the most in-depth courses with a thrust on Data Science. In order to provide high-quality career-oriented education in the domain of Data Science, Great Learning launched PGP and M.Tech Program in Data Science. The most exciting point about these programs offered by Great Learning is that all of the above-mentioned points can be related to the Data Science courses provided by Great Learning. From learning popular programming languages such as Python, R to exploring the gamut of data through popular tools like Tableau, these programs offered at Great Learning will definitely help you find your spot in the domain of Data Science. Unlike any other programs, the programs by Great Learning provides a Dedicated Placement Assistance at the completion of the courses. Due to the great industry connection, several leading companies participate in the hiring drives organised for the aspirants of the courses. At the end of the course, there is also a Placement Readiness Evaluation where the aspirants are provided help to prepare for interviews. Some of the companies that have recently participated in the hiring process include Uber, Swiggy, Oyo, Mercedes Benz, Cognizant, among others. How Do The Programs Standout The programs offered by Great Learning provides career support to the aspirants through Placement Readiness Evaluation to help aspirants prepare for interviews, on clearance of which aspirants sit for the placements.Learn from leading academicians and experienced industry practitioners in the field of data science.Build knowledge through classroom lectures by expert faculty and doing multiple challenging projects across various topics and applications in Data Science.Great Lakes is the youngest institute in India to receive an AMBA, UK accreditation, and a leader in analytics education.Several leading companies participate in the hiring drives organised for PG Data Science Course aspirants.Earn a Great Lakes certificate and create an ePortfolio to showcase your learning & project.Some of the companies that have recently participated in the hiring process of PG Data Science Program include Uber, Swiggy, Fractal Analytics, Oyo, KPMG, Mu Sigma, Mercedes Benz, Cognizant, Mahindra, Big Basket, among others. PGP – Data Science and Engineering The 5-month PGP Data Science and Engineering Course uses a combination of learning methods that include classroom teaching, hands-on exercises, and sessions with industry practitioners. The classes are conducted on weekdays and are assisted by online discussions and assignments. In this course, the applicants will build their knowledge through classroom lectures by expert faculty and doing multiple challenging projects across various topics and applications in Data Science. Who Can Participate In This Program? This course is for candidates with 0 to 3 years of experience. Job Roles One Can Break Into: This course will help the applicants to prepare for jobs in profiles like Business Analysts, Data Analysts, Data Engineer, Analytics Engineer, Big Data Consultant, etc. Hands-on Learning: The course covers the tools and skills sought by leading companies in Data Science. Through the duration of the course, candidates are trained on Python, SQL, Tableau, Data Science and Machine Learning. Dedicated Placement: Candidates do receive placement assistance in the Post Graduate Program in Data Science & Engineering. Upon successful completion of the program, candidates would be provided interview opportunities up to 3 months’ post completion of the program. Great Learning also prepares candidates for interviews by providing extensive support in terms of mentoring, CV review, interview preparation etc. M.Tech in Data Science and Machine Learning The M.Tech in Data Science and Machine Learning is a 21 months program offered in Full-Time \/ Weekend-Classroom formats, which enables participants to gain an in-depth understanding of data science and analytics techniques and tools that are widely used by companies. Upon successful completion of all requirements, the participants in this program will earn an M.Tech Degree from PES University. The classes will be held at PES University Electronics City Campus, and Great Learning online platform. Who Can Participate In This Program? This course is for candidates with 0 to 5 years of experience. Job Roles One Can Break Into: Data Scientist, Data Analyst, Machine Learning Engineer, Data Science Generalist, etc. Hands-on Learning: Through the duration of this course, the candidates will be trained on Python, SQL, Tableau, Data Science and Machine Learning. Wrapping Up The above courses provided by Great Learning are one of the most in-depth courses with a thrust of Data Science and Machine Learning. It is definitely a one-stop solution for the aspirants who want to upskill their knowledge in the field of Data Science and Machine Learning. Interested candidates can apply here. M.Tech in Data Science and Machine LearningPGP – Data Science and Engineering","excerpt":"The data science and analytics sector in India has witnessed a sharp increase in demand for highly-skilled professionals who understand both the business world as well as the tech world. Data Science is considered one of the most lucrative jobs in the industry right now.  However, the industry is still riddled with a lot of […]","categories":["AI Trends"],"tags":["big data certification","big data for social good","big data trends and challenges","Courses","data analyst certification","data analytics certificate","data science ai bootcamp","Data Science Career","data science interview","Data Science Jobs","pgp program in data science","what is data oriented person"],"author_name":"Ambika Choudhury","publish_date":"2020-03-02T13:00:00","publication_year":"2020","word_count":1781,"keywords":["GCP","R","data analyst certification","what is data oriented person","data science","distributed computing","pgp program in data science","analytics","data science interview","machine learning","AWS","AI","big data for social good","Data Science Jobs","Data Science Career","data science ai bootcamp","big data trends and challenges","big data certification","data analytics certificate","Python","Courses","Azure"],"extracted_tech_keywords":["AI","machine learning","data science","analytics","AWS","Azure","GCP","distributed computing","Python","R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/how-to-crack-a-data-science-interview\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10090537,"title":"Twitter Open-Sources Its Recommendation Algorithm","content":"As Elon Musk had promised earlier, Twitter has decided to make its algorithm code that it uses for selecting tweets to display on users’ timelines publicly available on GitHub and provided an explanation in a blog post. The post outlines the factors considered by the algorithm in selecting tweets for the “For You” timeline and how they are ranked and filtered. Click here to check out the code. Musk revealed much information about the release on a Twitter Spaces session earlier. He pointed out that the release might look quite embarrassing as there are a lot of mistakes in the code. Interestingly, to avoid any risk, the open-source code does not include Twitter’s ad recommendations or the social media’s training data. Most of the recommendation algorithm will be made open source today. The rest will follow.Acid test is that independent third parties should be able to determine, with reasonable accuracy, what will probably be shown to users.No doubt, many embarrassing issues will be… https:\/\/t.co\/41U4oexIev— Elon Musk (@elonmusk) March 31, 2023 The algorithm consists of three stages, as per Twitter’s explanation. The algorithm collects the most noteworthy tweets from various sources, ranks them using a machine learning model, and then removes any tweets from blocked accounts, previously seen tweets, or NSFW content before displaying them on a user’s timeline. Manu Joseph, creator of PyTorch Tabular, said in a post that It’s essential to note that these algorithm system are a complex collaboration, that require many different components. A basic ranking algorithm or collaborative filtering model isn’t adequate on its own. Interestingly, the blog shows that the pipeline above runs approximately 5 billion times per day and completes in under 1.5 seconds on average. Moreover, a single pipeline execution requires 220 seconds of CPU time, nearly 150x the latency you perceive on the app. Twitter Strides The release of the code of recommendation algorithm’s source code, comes after a series of controversies in recent months. Platformer reported that in February, Elon Musk requested that Twitter’s engineers modify the algorithm to increase the visibility of his tweets. However, Twitter later reversed this alteration. In November, Twitter started displaying more tweets from non-followed accounts, a move that had been tested before Musk’s involvement but was ultimately scrapped after user pushback. In the release Musk talks about a lot of privacy and security issues for not including a lot of code for the website. On the flip side, this comes in the backlight of Twitter recently firing its ethical AI and safety staff. But even then, Musk has ensured that the company has taken all the steps to ensure the reliability and security of the code. Twitter is also facing competition in the open-source community. Mastodon, the decentralised social network platform has been gaining traction among several Twitter users. Jack Dorsey, Bluesky is also attracting a lot of users, which is also looking for open-source protocols.","excerpt":"To avoid any risk, the open-source code does not include Twitter’s ad recommendations or the social media’s training data.","categories":["AI News"],"tags":["Elon Musk"],"author_name":"Mohit Pandey","publish_date":"2023-04-01T11:22:38","publication_year":"2023","word_count":481,"keywords":["Go","machine learning","programming_languages:R","PyTorch","AI","Git","RAG","Elon Musk","ai_frameworks:PyTorch","GitHub","R"],"extracted_tech_keywords":["AI","machine learning","PyTorch","RAG","R","Go","Git","GitHub","ai_frameworks:PyTorch","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/twitter-open-sources-its-recommendation-algorithm\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10166594,"title":"How DataRobot is Pushing AI Use Cases to Production","content":"Headquartered in Boston, enterprise AI company DataRobot recently unveiled its enterprise AI suite to simplify the creation and deployment of AI applications. The AI suite helps develop and deliver generative AI applications and agents. A few weeks prior to the launch, AIM got in touch with CEO Debanjan Saha, who discussed the company’s strategy of addressing three critical gaps hindering widespread AI adoption in enterprises: the value gap, the confidence gap, and the expertise gap. Addressing the Value Gap Saha emphasised the need for AI investments to deliver tangible business value. “People are building data centres. At various conferences, you will hear that people are building gigawatts of capacity,” Saha noted. However, he stressed that these investments must translate into solving real business problems beyond creating simple chatbots. To bridge this gap, DataRobot is focusing on helping customers identify and implement AI use cases that drive significant business impact. “Often, in order to take an AI use case to production, from ideation to final stage, there are 19 different teams who have to participate in a collaborative way to make it through that process,” Saha revealed, highlighting the complexity of implementing AI solutions at scale. Tackling the Confidence Gap The confidence gap, as described by Saha, stems from various risks associated with AI deployment, including operational, accuracy, reputational, and regulatory risks. Due to these concerns, many organisations are hesitant to move beyond prototypes and demos. DataRobot has implemented robust governance and monitoring features in its enterprise AI suite to address this gap. For regulated industries such as financial services, the company provides comprehensive AI governance, such as AI observability with real-time intervention and moderation and one-click compliance documentation to empower businesses to deploy AI with confidence. “Before AI can go into production, it has to go through a set of checks, and a compliance report needs to be produced for both internal auditors and external regulatory bodies,” Saha explained. This approach has made DataRobot particularly attractive to financial institutions, with 60% of the top banks in the United States using the enterprise AI suite. Closing the Expertise Gap Having recognised the scarcity of AI talent, especially in enterprises, DataRobot is working to make AI more accessible to a broader range of professionals. The company’s strategy involves both technological solutions and hands-on support. On the tech front, the enterprise AI suite aims to lower the bar for participation through automation and user-friendly interfaces. Additionally, the company offers executive sessions, ideation workshops, and supervised hackathons. “We have a team of highly-skilled data scientists who have years of experience working with customers even before generative AI to help solve business use cases with AI,” Saha said. These specialists, known as applied AI experts, work alongside customer teams to upskill and assist them in implementing AI projects. Bridging Predictive,Generative AI, and Agentic AI It was in response to these gaps that DataRobot introduced the enterprise AI suite. It includes composable AI applications and agents that can be customised for a wide range of business needs, from predictive data analysis to generative content creation. The suite also features a collaborative AI application library, allowing teams to work together from a central repository, sharing tools, insights, and solutions. This collaboration is central to the enterprise AI suite, as it encourages cross-functional teams to innovate together and scale AI applications quickly. Developers can easily prototype, test, and refine generative AI applications, benefiting from real-time insights and accelerated deployment. The enterprise AI suite also streamlines the process of publishing and monitoring applications, ensuring that updates and improvements are implemented without user downtime, and that inputs remain accurate and up to date. Meanwhile, DataRobot has introduced advanced AI observability features, including “guard models” for generative AI applications. These guards intercept prompts and responses, checking for issues such as data leaks, toxicity, and accuracy. “We have put together a set of what we call ‘guard models’. So, when you deploy an LLM (foundational or fine-tuned), for example, you can deploy that with a set of guard models around it,” Saha explained. This approach allows for real-time intervention based on predefined rules, enhancing the safety and reliability of AI applications in production environments. DataRobot AI applications are already delivering value for customers and partners.  “DataRobot is enabling teams to deploy AI 83% faster than traditional methods and reduce costs by up to 80%,” said Saha. With over 38,000 customer deployments, global organizations including CVS Health, BMW Group, and the U.S. Army rely on DataRobot for AI that makes sense for their business In India, DataRobot has already proven success with increased productivity and accelerated value from AI. “DataRobot is an extreme developer productivity multiplier. We helped Razorpay to go from five days to less than four hours to create each model.” noted Saha.","excerpt":"Headquartered in Boston, enterprise AI company DataRobot recently unveiled its enterprise AI suite to simplify the creation and deployment of AI applications. The AI suite helps develop and deliver generative AI applications and agents.  A few weeks prior to the launch, AIM got in touch with CEO Debanjan Saha, who discussed the company’s strategy of […]","categories":["AI Highlights"],"tags":["AI agent","DataRobot","Debanjan Saha"],"author_name":"Vandana Nair","publish_date":"2025-03-25T10:47:35","publication_year":"2025","word_count":792,"keywords":["Go","agentic AI","AI","chatbots","ML","RAG","GAN","AI agent","Aim","generative AI","DataRobot","Debanjan Saha","R"],"extracted_tech_keywords":["AI","ML","generative AI","agentic AI","Aim","RAG","chatbots","R","Go","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/how-datarobot-is-pushing-ai-use-cases-to-production\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":14500,"title":"Accenture-led AI solutions drive Akshaya Patra’s Million Meals program","content":"Image source: Akshaya Patra One of the world’s largest technology leaders Accenture and famed NGO mid-day meal provider Akshaya Patra has teamed up to deploy disruptive technologies to significantly increase the number of mid-day meals served to children across schools in India. According to a statement released by Accenture, the “Million Meals” project has significantly revolutionized Akshaya Patra’s supply chain and operations, providing improved food quality and expanded service reach. See how AI is driving Akshaya Patra’s Million Meals project Through new-age technologies such as artificial intelligence (AI), the Internet of Things (IoT) and blockchain, global technology leader Accenture has scaled the significant challenges of production and delivery of mass meals. According to the statement, Accenture Labs, the research and development arm of Accenture, executed the project over a period of six months in Akshaya Patra’s Bengaluru kitchen. When the project was analyzed, the results indicated a potential to improve efficiency by 20 percent, that could boost the number of meals served by millions. Akshaya Patra’s goal is to continue to optimize and streamline the program. According to Sanjay Podder, managing director at Accenture Labs, Bengaluru, “The Accenture Labs ‘Tech for Good’ program specializes in solving complex social problems through the use of innovative technologies. Accenture Labs was excited to take up the challenge that every paisa saved in producing a single meal can result in additional meals being served to children. We thought this was an ideal proving ground to put the combined power of AI, IoT and blockchain to use.” Shridhar Venkat, CEO, The Akshaya Patra Foundation, had a similar view. “This initiative not only enhanced efficiency while giving a boost to our operations, but ensured that quality standards are met while increasing the number of meals served. We aim to expand our reach to millions of more children with the help of such technologies. This is just the beginning; we look forward to more collaborations with Accenture.”","excerpt":"One of the world’s largest technology leaders Accenture and famed NGO mid-day meal provider Akshaya Patra has teamed up to deploy disruptive technologies to significantly increase the number of mid-day meals served to children across schools in India. According to a statement released by Accenture, the “Million Meals” project has significantly revolutionized Akshaya Patra’s supply […]","categories":["AI News"],"tags":[],"author_name":"Richa Bhatia","publish_date":"2017-04-24T10:05:52","publication_year":"2017","word_count":320,"keywords":["Go","artificial intelligence","programming_languages:R","AI","ML","programming_languages:Go","Aim","R"],"extracted_tech_keywords":["AI","artificial intelligence","ML","Aim","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/accenture-led-ai-solutions-drive-akshaya-patras-million-meals-program\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":68913,"title":"Why Capstone Project Is A Key Feature While Selecting A Data Science Course","content":"Data science is a dynamic and ever-evolving landscape in which professionals and aspirants have to continuously learn in order to keep up their relevance in the industry. And that’s why there are a humongous amount of data science courses in the market to choose from. In fact, according to a report, this pandemic has given a significant push to online education where numerous ed-tech companies have come up with industry-ready courses for data scientists. The report further noted there is also a 55% surge in the creation of online courses where experts, as well as instructors around the world are creating online courses to share their knowledge and experience to the world. This, in turn, creates a massive confusion for aspiring data scientists as to choosing the right course in order to get benefitted the most. And that’s where capstone projects play a massive role for data scientists. A capstone project, as the name suggests, is the crown of analytics projects and learning, which allows data scientists to integrate all their knowledge and demonstrate it through a comprehensive project. The capstone projects are usually designed by experts of the industry and are used to analyse the length and breadth of the data science knowledge of a candidate. Picking up an online course with capstone projects will allow data scientists to showcase their practical knowledge of tools and technologies to their potential employers. As a matter of fact, a survey done by Analytics India Magazine has revealed that 92% of the students believe that capstone projects played a vital part in choosing the right data science courses for them. Importance Of Having Capstone Projects In Online Data Science Courses Online courses allow data science aspirants to gain the necessary knowledge to land several job offers. However, having a capstone project in an online course would help data scientists to showcase knowledge and skills to prospective employers. Unlike any other projects, a data science capstone project includes a real data set of a business scenario in order to build a model and is usually conducted in partnership with industry government or academia. After completing these capstone projects, data scientists would have a data product to showcase their talent to potential employers, which, in turn, would give them a competitive edge over others. In recent news, founder of a Mumbai-based ed-tech company, Imarticus Learning, Nikhil Barshikar stated in the company release while partnering with KPMG in India for a Data Science Prodegree course, that the biggest limitation faced by data science job aspirants today is the involvement of real cases and data in their learning to get the first-hand experience. “Through this partnership with KPMG in India, we aim to create skilled data science professionals by providing just that,” said Barshikar. With this partnership, KPMG is providing data science students with real-world capstone projects to support the curriculum. In this article, we will share a few reasons as to why capstone projects are a key feature while selecting data science courses. Also Read: How To Build An Online Course On Data Science Aid In Advance Degrees Capstone projects involve a massive amount of research and require data scientists to apply all the skills and knowledge they have been taught in their courses. Such a comprehensive project would give data scientists a good idea of their proficiency and provide them with the opportunities to explore a topic of interests, which can later be turned into their specialisation. Alongside, aspirants can also get a lot of research experience with their capstone projects, which again can help them while pursuing advanced degrees. Also, being involved in such long term projects would require students to be self-directed and thoroughly committed. This would help students and aspirants to set their future goals and plan their career accordingly. Not only these capstone projects help aspirants improve their focus but also enhance their ability to ask the right question, which is again critical for data scientists. Also Read: How To Crack The Capstone Project In Analytics Learning Enhance Skills & Knowledge Data science is all about tools and technologies, and a capstone project can help them enhance those skills by working on projects with real-world data and practical applications. With other projects, data scientists can gain experience, however, with capstone projects, it demands a great deal of research and analytical skills. This will allow data scientists to showcase their proficiency in real environments. The projects involve complex model building as well as working with real data, which not only sharpens the skills of aspirants but also makes them industry-ready. Capstone projects also force data scientists to work in a team in order to build a better model, which, in turn, will improve the teamwork of data scientists and will make them learn how to manage people as well as deadlines. Staying Updated With Current Market Trend Capstone projects are usually built on current business problems and require data scientists to work on real data that is currently available in the market, and thus, will help these aspirants stay updated with the current business trends. Also working with real-life data would make these aspirants experience real-world challenges like missing data, loopholes, etc., while solving business problems. This would force data scientists to take up the responsibility to use critical thinking and creativity while solving problems. Alongside using such criticality, solving strenuous problems will help these aspirants to understand the fundamental cause of challenges, which, in turn, can support them to make a solution that is much more relevant to the current era. Also Read: Why Capstone Is The Backbone Of Data Science Education Gives A Competitive Edge During Hiring Lastly, these capstone projects give a competitive edge to aspiring data scientists. With the increased competition in the data science space, it has become critical for data scientists to showcase advanced skillsets and projects that would stand out in this dynamic market. Capstone projects would not only equip entry-level data scientists with necessary skills but will also put them in the spotlight amid the crowd. Besides, it acts as a validation of the domain knowledge but also showcases professionalism and real-world experience. These capstone projects give interviewers and recruiters the idea of the proficiency level of data scientists, and thus would make them more employable amid this crisis.","excerpt":"Data science is a dynamic and ever-evolving landscape in which professionals and aspirants have to continuously learn in order to keep up their relevance in the industry. And that’s why there are a humongous amount of data science courses in the market to choose from. In fact, according to a report, this pandemic has given […]","categories":["AI Features"],"tags":["best online data science masters","big data trends and challenges"],"author_name":"Sejuti Das","publish_date":"2020-07-03T17:00:00","publication_year":"2020","word_count":1046,"keywords":["best online data science masters","data science","Go","programming_languages:R","AI","R","programming_languages:Go","Aim","ViT","analytics","big data trends and challenges"],"extracted_tech_keywords":["AI","data science","analytics","Aim","R","Go","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/why-capstone-project-is-a-key-feature-while-selecting-a-data-science-course\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10134728,"title":"How Generative AI is Transforming Intelligent Collections and Revolutionising Receivables","content":"The order-to-cash (O2C) process, encompassing everything from order processing to receivables collection, is critical for an organisation’s financial health. For large enterprises, especially in sectors with thousands of daily transactions, managing accounts receivable can be a daunting challenge. Traditionally, this process was handled manually, with finance teams relying on monthly reports and retrospective analyses. This often led to delayed responses, late payments, and missed opportunities for optimisation. However, the advent of generative AI has introduced a new era of efficiency and accuracy. By leveraging the power of AI, companies are transforming the O2C process and redefining how receivables are managed. “In 2017, this was a very descriptive and diagnostic-heavy process,” said Prathmesh Thergaonkar, Director, Finance Analytics at Fractal. “People used to look at end-of-month Excel reports and KPIs to analyse what was happening.” This retrospective approach meant that delays and issues were often missed until it was too late to address them effectively. “As the technology landscape evolves, so must the receivables management process,” Thergaonkar noted. Fractal’s Generative AI Breakthrough This is where Fractal’s latest innovation comes into place—the Intelligent Collections Management (ICM) accelerator for Accounts Receivable (AR) analytics. Designed to revolutionise the O2C process, ICM shifts the paradigm from traditional month-end reporting to real-time actionable insights. ICM stands out with its predictive-first approach, setting it apart from conventional AR analytics tools that rely on retrospective month-end reporting. By leveraging advanced ML algorithms, ICM delivers real-time predictions and insights. This enables businesses to forecast cash flows with unprecedented accuracy, identify potential payment defaults before they happen, and proactively manage collections efforts. With these predictive capabilities, organisations can optimise cash flow and enhance financial stability by taking timely actions. Comprehensive KPI Framework ICM’s analytical framework is robust, encompassing four KPIs—descriptive, diagnostic, predictive, and prescriptive. Descriptive KPIs: These provide insights into current and historical performance metrics such as ageing reports, days sales outstanding (DSO), and collection efficiency. This data visualisation helps users quickly assess the health of their AR portfolio. Diagnostic KPIs: These metrics identify the root causes of issues like payment delays or disputes. By analysing trends in dispute reasons or deduction categories, users can address recurring problems at their source. Predictive KPIs: ICM excels in predictive analytics, forecasting future payment behaviours, highlighting accounts at risk of delinquency, and estimating the likelihood of dispute resolutions. These insights enable users to prioritise efforts and allocate resources more effectively. Prescriptive KPIs: Beyond prediction, ICM offers prescriptive recommendations, suggesting next-best actions to improve collections, resolve disputes, and reduce revenue leakage based on historical data and current conditions. Seamless Data Integration & User-Centric Design One of ICM’s standout features is its ability to integrate data from various sources, including ERP systems, local databases, planning tools, SharePoint, and Excel. This comprehensive data aggregation ensures access to a single source of truth, eliminating data silos and fostering cross-departmental collaboration. By harmonising data from different systems, ICM not only enhances analytics accuracy but also simplifies the user experience, allowing seamless navigation and exploration of insights. ICM’s design caters to various user personas, from CXOs to analysts, with an intuitive interface that offers extensive slicing and dicing capabilities. Users can drill down into the most granular levels of data, whether seeking high-level summaries or detailed transaction-level insights. Customisable dashboards and reports allow users to focus on the metrics that matter most, while interactive visualisations make interpreting complex data straightforward. Comprehensive Coverage of O2C Areas ICM addresses all critical areas of the O2C process, including collections, deductions, leakage, disputes, and credit management, ensuring a holistic view of AR and enabling proactive issue resolution. Collections: ICM optimises collections by prioritising accounts based on risk profile and payment history, streamlining the process with automated reminders and follow-ups, reducing DSO, and improving cash flow. Deductions:  The tool leverages machine learning to categorise deductions, enabling users to understand underlying causes and take corrective actions. With Generative AI, it efficiently charges back invalid low-value deductions, helping minimise future occurrences and improve customer negotiations. Leakage: ICM detects potential revenue leakage points, such as uncollected invoices or unauthorised discounts, allowing users to recover lost revenue and prevent future leakage. Disputes: ICM accelerates dispute resolution by predicting the likelihood of successful outcomes and providing recommendations for effective resolution, reducing time and effort and enhancing customer satisfaction. Credit Management: ICM helps users manage credit risk by evaluating customer creditworthiness and recommending dynamic credit limits, ensuring businesses extend credit to reliable customers while minimising bad debt risk. Real-Time Insights and Autonomous Finance ICM features a generative AI layer that enhances decision-making with real-time responses, combining graphical representations and business commentary. This AI-driven assistant provides forecasts, next-best actions, and contextual insights, driving efficiency across the board. Be it predicting payment dates or analysing dispute impacts on cash flow, the AI assistant delivers actionable intelligence that enhances user productivity. ICM’s ultimate vision is to drive the O2C process towards an autonomous finance state, where human intervention becomes optional. Its advanced analytics and AI capabilities enable the tool to handle most decision-making processes independently, based on internal and external data. By automating routine tasks and providing intelligent insights, ICM frees finance professionals to focus on strategic initiatives, improving efficiency and empowering organisations to achieve greater agility and resilience.","excerpt":"Designed to revolutionise the O2C process, Fractal’s ICM shifts the paradigm from traditional month-end reporting to real-time actionable insights.","categories":["AI Highlights"],"tags":["Fractal AI","Generative AI","Intelligent Agent"],"author_name":"Mohit Pandey","publish_date":"2024-09-09T10:39:14","publication_year":"2024","word_count":865,"keywords":["Go","machine learning","AI","R","ML","RAG","Intelligent Agent","analytics","generative AI","GAN","Generative AI","predictive analytics","Fractal AI"],"extracted_tech_keywords":["AI","machine learning","ML","analytics","generative AI","RAG","predictive analytics","R","Go","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/how-generative-ai-is-transforming-intelligent-collections-and-revolutionising-receivables\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10123836,"title":"India Might Soon Run Out of Skilled Software Engineers","content":"“I’m about to graduate with a CS degree and have never used a ‘library’, ‘framework’, or ‘API’.” This was the topic of a Reddit discussion, which soon saw hundreds of reactions pouring in from people who pointed at the sad state of CS degrees worldwide. And though this may be a global phenomenon, the Indian youth has it far worse than their counterparts. Unfortunately, this is not an isolated case, but a mere symptom of a larger issue plaguing the Indian education system. Despite holding degrees, many graduates here find themselves unprepared for the practical demands of the industry, making them less competitive on the global stage. We’ve often heard that a large section of the Indian youth is basically willing to go abroad and study. “The hate for India scares me,” said a user on Reddit, starting a discussion about how soon, India wouldn’t be left with skilled young people. This, honestly, is kind of scary. Another user on X pointed out that many Indian IT tech graduates are unemployable for any real time projects and responsibilities. “India IT talent is going down and Vietnam, Indonesia, Malaysia, and Philippines are taking up jobs as they have the hunger to succeed with hard work which India’s new college graduates don’t,” the user added. Check: NIRF Ranking of Engineering Colleges Not Adequately Prepared for the World According to a recent report, India barely has about 2000 senior software engineers. Once hailed as a global hub for IT talent, the country is now facing a potential crisis in the availability of skilled software engineers. The issue is not just about the numbers, but also the quality and aspirations of the youth. The sentiment that education in India is not preparing students for real-world applications adequately is echoed widely. A recent video by a Delhi-based coaching institute, Drishti IAS, was making rounds across social media platforms about a BCA graduate who had no idea about the programming languages to use. He had never built anything on any language, since he was never given any knowledge about it during his course, he said. https:\/\/twitter.com\/memenist_\/status\/1800804306167480348 On the other hand, universities like Stanford, Harvard, or any other major institutions in the US, offer a much more hands-on approach towards teaching, while also opening up major possibilities when it comes to jobs with great paychecks. This disillusionment is driving many young Indians to seek education and career opportunities abroad, exacerbating the “brain drain” as many put it on social media. The comment underscores a growing trend: an increasing number of young Indians are keen to leave the country in search of better prospects. The frustration among students is palpable. Many feel a sense of helplessness and disillusionment with the system. “Why are students more focused on going to America or Canada instead of making India better?” a user questioned. This sentiment points to a lack of faith in the country’s ability to provide the necessary opportunities and support for young talent. The narratives on social media also highlight systemic issues that push students away. The lack of infrastructure is a recurrent theme. While there are several great institutions in India, such as IITs and NIITs, only a handful of top students are able to get into them. The rest, with big pockets, find it useful to study abroad and settle there. The Change is Out There, But… However, the future can look a little less bleak for Indian youth with the advent of AI. “India has never led any fundamental research, but we have a golden opportunity as AI can be a levelling field,” said Pratik Desai, the founder of KissanAI. “However, this requires a fundamental shift from coaching and academia to a change in mindset from parents, and founders to investors.” The sad truth is that this is actually a standard for CS graduates in the country, unless they are from IITs. AI has helped people code even without learning programming languages. GitHub CEO Thomas Dohmke and Microsoft CEO Satya Nadella recently claimed that now everyone can code in their natural language, which should ideally drop the barrier of entry for everyone in the country. Turns out that it can also have the opposite effect. Since everyone is now basically becoming a developer, the demand for high-skilled software engineers is increasing. These are developers who think beyond using Copilot tools to make generic software. As a Reddit user pointed out, “Don’t be a generative software engineer.” Though things are slowly changing. India is now seeing an increase in talent retention. In 2019, a majority of Indian AI researchers with undergraduate degrees sought opportunities abroad. However, by 2022, one-fifth of these researchers chose to work in India. The numbers are still minuscule. There should be proper incentives for software engineers to stay and work in the country, before we go dangerously short on skilled minds.","excerpt":"The sentiment that education in India is not adequately preparing students for real-world applications is echoed widely.","categories":["AI Features"],"tags":["software engineer"],"author_name":"Mohit Pandey","publish_date":"2024-06-17T18:21:46","publication_year":"2024","word_count":808,"keywords":["Go","API","AI","RPA","Git","software engineer","Aim","Rust","AI research","GitHub","R"],"extracted_tech_keywords":["AI","Aim","R","Go","Rust","Git","GitHub","API","RPA","AI research"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/india-might-soon-run-out-of-skilled-software-engineers\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10084969,"title":"ChatGPT Makes Literary Debut, It&#8217;s Now a Published Author","content":"Of the thousand of journals and research papers published every day, one has managed to make it to the headlines due to its co-author, ChatGPT, which has been the centre of attention since its release in November 2022. Preceding all others, a peer-reviewed paper titled ‘Open artificial intelligence platforms in nursing education: Tools for academic progress or abuse?’ was recently published by Siobhan O’Connor, senior lecturer at the School of Health Sciences and an adjunct associate professor at the Western University. The internet has lauded the official acceptance of the chatbot. Some even said ChatGPT is on the way to becoming the most cited author in history and simultaneously making tenure in all major universities. On the contrary, many people were put off seeing AI in the authorship and call the attempt a way out to protect oneself from plagiarism accusations. Moreover, the author needs to reference or mention how they have used the tool for the study. The interesting questions and unknowns are around copyrights and legislation that need to be prepared for such a storm. Adding to the backdrop recently, a group at Northwestern University of Chicago, Illinois, led by Catherine Gao, used ChatGPT to generate artificial research paper abstracts to test whether scientists can spot them. The ChatGPT-generated abstracts sailed through the plagiarism check: the median originality score was 100%, indicating no plagiarism. The AI-output detector spotted 66% of the generated abstracts. But the human reviewers didn’t fare too well: they correctly identified only 68% of the generated abstracts and 86% of the genuine abstracts. Furthermore, they incorrectly identified 32% of the generated abstracts as being real and 14% of the genuine abstracts as being generated. Moreover, one of the most prestigious machine learning conferences, the International Conference on Machine Learning (ICML), banned authors from using tools like ChatGPT to author papers, triggering a debate about the role of AI-generated text in academia. ICML announced the policy earlier this month: Papers including text generated from a large-scale language model like ChatGPT are prohibited unless the produced result is a part of the paper’s experimental analysis. The announcement sparked discussion on social media, with AI academics and researchers defending as well as criticising the policy. The conference’s organisers responded by publishing a longer statement explaining their thinking. The questions the ICML is addressing may not be easily resolved. The availability of tools like ChatGPT is confusing various institutions, some of which have responded with their bans. In 2022, Stack Overflow, a coding Q&A site, banned users from submitting responses created with ChatGPT due to its tendency to produce inaccurate results.","excerpt":"A first-of-its-kind peer-reviewed research paper co-authored by ChatGPT has been published.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","ChatGPT","GPT3","GPT4","OpenAI"],"author_name":"Tasmia Ansari","publish_date":"2023-01-13T12:30:35","publication_year":"2023","word_count":432,"keywords":["Go","ChatGPT","artificial intelligence","machine learning","OpenAI","AI","TPU","ML","GPT","GPT3","GAN","R","AI (Artificial Intelligence)","GPT4"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","ChatGPT","TPU","R","Go","GPT","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/chatgpt-makes-literary-debut-now-a-published-author\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10064062,"title":"ServiceNow opens data centres in Mumbai, Bengaluru","content":"ServiceNow has launched local data centre facilities in Mumbai and Bengaluru, strengthening its commitment to helping Indian customers embrace innovative digital solutions and meet local data residency preferences. “India is a top growth market for ServiceNow and we are seeing larger local organisations partnering with us to benefit from the power of our Platform to drive their digital transformation needs,” said Arun Balasubramanian, ServiceNow Managing Director India & SAARC. “ServiceNow is committed to managing data locally and meeting India’s data protection and sovereignty standards. The new data centres enable even more financial services, technology and other regulated verticals to reap the benefits of ServiceNow’s single, unifying platform for digital business.” Mike Lents, ServiceNow Senior Vice President, Global Cloud Services, said that the new data centres are large sites for ServiceNow with full network and capacity, to deliver performance at scale and increased sustainability. “ServiceNow has certified Data Centre Sustainability Professionals serving our customers and partners in every region we operate, including the Asia Pacific. These data centres give ServiceNow’s customers and partners a sustainable way to responsibly source cloud data hosting in India and contribute positively to their own ESG commitments,” said Lents. A total of 18,000 people will be trained and certified in India by the end of 2022, together with partners and customers, to meet the burgeoning demand for ServiceNow product specialists. “As more companies take bold steps to leverage ServiceNow’s innovative digital solutions to modernise their business and better serve customers and employees, more experts are needed to help them with integration,” said Balasubramanian. “The skills drive supplements existing NextGen and Partner Training and Certification programs, enabling thousands of Indian workers the opportunity to build rewarding careers within our fast-growing ecosystem.”","excerpt":"ServiceNow is committed to managing data locally and meeting India’s data protection and sovereignty standards.","categories":["AI News"],"tags":["Cloud Data AI","data centre","Financial Services","India data centre market","ServiceNow"],"author_name":"Kartik Wali","publish_date":"2022-03-31T14:51:59","publication_year":"2022","word_count":284,"keywords":["ServiceNow","India data centre market","AI","programming_languages:R","R","digital transformation","Git","RAG","Financial Services","GAN","Cloud Data AI","data centre"],"extracted_tech_keywords":["AI","RAG","R","Git","GAN","digital transformation","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/servicenow-opens-data-centres-in-mumbai-bengaluru\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":49250,"title":"Deep Dive: How Lendingkart Deploys Data To Increase Geographical Footprint","content":"When it comes to starting or growing a business, one of the major hurdles is the capital or the seed money. Over the years, small and medium-sized businesses have faced challenges in having access to sufficient and timely credit. Now, the MSME sector is considered to be indispensable with India pegged at becoming a $5 trillion economy by 2025. One such company that has stepped in to fill this void is Lendingkart, a non-deposit taking NBFC providing working capital loans and business loans to MSMEs across India. In order to know more about the company, we got in touch with Harshvardhan Lunia, Co-founder and CEO, Lendingkart and Saket Anand, Chief Analytics Officer, Lendingkart. Lendingkart’s Track Record Over the last five years, Lendingkart has grown at a healthy pace. Its revenue and loan books have been growing approximately 3 times year-on-year. From disbursing loans in a few locations to creating a strong SME customer base across 1,300 cities and towns in the country, Lendingkart has come a long way. The company has catered to 64,000+ unique businesses and today, it has evaluated around 6 lakh applications and disbursed more than 76,000 loans. Given that a large chunk of its borrowers come from the unorganised sector, with little or no formal credit history, the company has identified alternate sources of evaluating their creditworthiness. “We have identified various proxies for credit comfort to determine a customer’s intent to pay back a loan, the quality of his product\/service, the financial health of his business, and ability to survive with competition etc. Hence, the application process is simple and requires minimal documentation,” said Lunia. Today, on average, the company receives around 7,000 leads and 1,000-1,200 applications out of which it is able to sanction around 300 loans on a daily basis. And talking about the revenue, Lunia said, “While we had revenue growth of 168% for FY 2017-18, our last year’s revenue growth rate (2018-19) was more than 200%. Our book outstanding grew by 2.9x in 2019 alone.” The Tech Stack Talking about the tech stack at Lendingkart, Lunia said that the company uses microservices architecture for scalability. “Using the right tool for the right job is very important, and that is how it works here,” said Anand. For example, Lendingkart uses MongoDB as a database for cases when the number of fields is not fixed and to store data that are single time writes and where the speed of writing data trumps transactional sanctity. Whereas, it uses MySQL in case of transactional databases such as to track the application process and for analytics. And for the data warehouse, it uses redshift. Moreover, various microservices talk to each other through Kafka to keep the statuses in sync. Coming to the backend, the tech stack is mostly on Java using Spring Boot framework, while it uses Python for data engineering and analytics. “We also make use of widespread front-end technologies such as – HTML and Angular.js. The entire tech stack is hosted on AWS and leverages AWS infrastructure for Reliability, Availability and scalability,” said Anand. The Role Of Data And AI When it comes to making the best use of data science, the company has done it extensively with a team of 150 individuals. One of the best examples is that its customer origination is completely digital, which is further complemented by a machine learning-based credit decisioning. The data it collects from customers is run through its algorithms, extracting more than 8,500 data points. This helps the company discern the intent and ability of the customer to repay the loan, further aiding its loan sanctioning process. Lendingkart’s systems also have the capability to crunch non-traditional data such as GST data, mobile data, product interaction data, social data for the purpose of credit evaluation, quality lead scoring, and product interaction, among others. How Lendingkart uses data to reach out to the right people: The company works out with the right customer profiles to target based on the borrowers it has already reached. Its ML-based customer segmentation model takes into account variables such as demographic data, CIBIL parameters, loan performance on its books and production interaction data to figure out the good customer profiles. “Using this segmentation, we optimise our marketing campaigns to target the right people and significantly bring down the cost of customer acquisition,” Anand explains. How Lendingkart uses AI as a tool: Lendingkart’s systems have a training capability — they guide the fulfilment team on the optimal number of interactions that need to be done in order to successfully close the leads. That’s why they are also looking to deploy chatbots and voice botsAI is also used to predict accounts that are likely to go delinquent in advance.It has also been used to build state-of-the-art data infrastructure, which helps capture every interaction with the customer: call logs, call transcripts, SMS, email, product interaction variables and social data, in addition to traditional LOS and LMS data.The company is also working on an NLP-based system that can convert SMS data into credit variables, and also plan to use similar techniques to convert voice transcripts from calls into decisioning variables. The Hiring Phase Being a technology-driven company, Lendingkart gauges the potential of the candidates, over experience or qualifications. “A problem-solving attitude, with a penchant for learning, draw great dividends in such situations,” said Lunia. To fit into the team Lendingkart, a candidate must have a learning attitude, investigating thought process, aspirations mapping to a dynamic environment, engineering approach and skills\/competencies. Furthermore, the company also believes in training the candidates to become the best. “We believe that the right grooming and mentorship at an early stage in career provides the much needed optimal efficiency and productivity,” Lunia added. The Future Direction Lendingkart believes that its biggest strength is its ability to acquire and service MSME customers at the remotest places in India using its digital reach. Also, its technological and analytical capabilities built in-house have helped develop a robust digital origination system and automated credit decisioning, distinguishing us from the rest in the market. Looking into the future, all set to double the book size in the coming year and continue to grow existing lines of business to cater to the different needs for financial inclusion, Lendingkart will be venturing beyond our current offerings. “Our goal is to enable financial inclusion of the larger MSME community, delivering financial products and services to them in a cost-effective and timely manner. So those aspiring entrepreneurs and MSMEs with little or no credit history can apply for loans through a simplified, digital application process,” said Lunia in conclusion.","excerpt":"When it comes to starting or growing a business, one of the major hurdles is the capital or the seed money. Over the years, small and medium-sized businesses have faced challenges in having access to sufficient and timely credit. Now, the MSME sector is considered to be indispensable with India pegged at becoming a $5 […]","categories":["AI Features"],"tags":["data mapping","Deep Dive","deep learning application examples"],"author_name":"Harshajit Sarmah","publish_date":"2019-11-02T12:49:46","publication_year":"2019","word_count":1096,"keywords":["data science","machine learning","AWS","AI","chatbots","ML","RAG","NLP","microservices","analytics","deep learning application examples","data mapping","Deep Dive"],"extracted_tech_keywords":["AI","machine learning","ML","NLP","data science","analytics","RAG","chatbots","AWS","microservices"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/deep-dive-how-lendingkart-deploys-data-to-increase-geographical-footprint\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":25495,"title":"How Facebook Is &#8216;Opening&#8217; Their Users&#8217; Eyes With The Help Of ExGANs","content":"Every day, a large number of pictures are captured and shared in social networks with a large percentage of them featuring people-centric content. Facebook is making sure that you don’t make your friends take endless photos, posing and reposing at important occasions. They are also helping you cut out the time you spend on Photoshop or similar image-correction apps and products by making their own photo uploading platform more imaginative. According to a recent paper published by Facebook, the Menlo Park-based social media giant is working on an “in-painting technology” that replaces closed eyes with open ones in a remarkably convincing manner. They are working on achieving this by using Exemplar Generative Adversarial Networks (ExGANs) — a type of conditional GAN that utilise exemplar information to produce high-quality, personalised in-painting results. The paper says, “We propose using exemplar information in the form of a reference image of the region to in-paint, or a perceptual code describing that object. Unlike previous conditional GAN formulations, this extra information can be inserted at multiple points within the adversarial network, thus increasing its descriptive power. We show that ExGANs can produce photo-realistic personalised in-painting results that are both perceptually and semantically plausible by applying them to the task of closed-to-open eye in-painting in natural pictures. A new benchmark dataset is also introduced for the task of eye in-painting for future comparisons.” Basically, with data collected from the target people when they kept their eyes open, the GAN system learns what eyes should go on the person on the basis of shape, colour and other features of the eyes of a particular person. Created by Brian Dolhansky and Cristian Canton Ferrer from Facebook added in the paper, “…Because Exemplar GANs are a general framework, they can be extended to other tasks within computer vision, and even to other domains. In the future, we wish to try more combinations of reference-based and code-based exemplars, such as using a reference in the generator but a code in the discriminator.” Last year, researchers at Facebook’s AI lab had developed an expressive bot, the animations of which were controlled by an artificially intelligent algorithm. The algorithm was trained by making it watch hundreds of videos of Skype conversation and was now capable of mimicking the way humans adjust their expressions while talking.","excerpt":"Every day, a large number of pictures are captured and shared in social networks with a large percentage of them featuring people-centric content. Facebook is making sure that you don’t make your friends take endless photos, posing and reposing at important occasions. They are also helping you cut out the time you spend on Photoshop […]","categories":["AI News"],"tags":["Facebook","image processing","photoshop"],"author_name":"Prajakta Hebbar","publish_date":"2018-06-18T06:21:35","publication_year":"2018","word_count":383,"keywords":["Go","programming_languages:R","AI","image processing","programming_languages:Go","computer vision","GAN","Facebook","photoshop","R","ai_applications:computer vision"],"extracted_tech_keywords":["AI","computer vision","R","Go","GAN","programming_languages:R","programming_languages:Go","ai_applications:computer vision"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/facebook-can-open-your-eyes-exgan\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":2057,"title":"Meet this IIT engineer developing structured self-driving technology for unstructured Indian roads","content":"Sanjeev Sharma with his self driven Mahindra Bolero There is no doubt that the interest around self driving technology has escalated like never before. With many big players like Google and Tesla entering the market, names like Tata Elxsi too have been trying hard to get India’s own driverless car. But given the complex traffic conditions and the unstructured Indian roads, the journey is still a far stretched dream. However, this IIT Roorkee alumnus isn’t losing hope easily. Having founded Swaayatt Robots, he is aiming to create a completely different technology for driverless cars to operate seamlessly in India. Sanjeev Sharma, Founder, Swaayatt robot shares “the navigation module by Swaayatt is designed such so as to handle previously unseen environments. Meaning that it can work in GPS denied environments and can produce high speed trajectories even in very cluttered environments”. Let’s explore the journey of Swaayatt Robots, currently a one-man army, that promises to bring affordable self-driving vehicles in India. India vs others- It is a known fact that the environment in developed countries such as the US is very well structured with well defined traffic rules, which helps to significantly reduce the complexity of traffic-dynamics as compared to India. “The structured roads and environmental conditions such as paved roads, proper & elaborate lanes, road markers and signs makes it easier for vehicle’s perception modules”, says Sharma, who has been working on this idea since 2009. An evident absence of these scenarios in India requires a careful designing, new perception modules, robust motion planner, and a robust lane markers detector that can detect lane markers on road. Explaining the system developed by Swayaatt, he says “it finds all the divisions of the road, and not just the lane markers pertaining to the current lane of the vehicle”. He also believes that if not for this module, the only alternative way of doing this would be to build a detailed 3D map of every road in the country and then hope that the vehicle’s localization module is robust enough to localize the vehicle on road with 5-10 cm of accuracy. Sharma shares “currently I am using a Mahindra Bolero to test the technology. It was successfully tested in a colony\/campus where I live on February 17th this year. In this video, the entire vehicle is navigating autonomously using just 2 cameras”. However, it still needs some working to make it robust enough for typical Indian city traffic conditions, where there are virtually no traffic rules. Though the government regulation has been relaxed for self driving vehicle testing, he sees Indian traffic and unstructured condition, a major obstacle. “A very unique motion planning algorithm needs to be in place—such as those which can take decisions such as overtaking a vehicle on typical 2 lane roads in India or crossing a busy intersection at peak traffic hours without traffic signal, which is the case often in India”, he shares. Further challenges in adoption- Given that the concept of self driving is still nascent in India, how successful would the efforts by Swaayatt Robots prove? Sharma is quick to answer that he is quite hopeful and optimistic about the future of self-driving vehicles in India. “I had been dreaming of a self-driving vehicle since 2009 and thus, at least to me, the idea of self-driving vehicles anywhere is not as nascent”, he says. His hope and passion for the idea of self driving cars is evident from the fact that he decided to decline PhD offer at University of Massachusetts, and come back to India to start with his own venture. Though the exact launch date for the technology is under wraps, Sharma reveals that some time in June, it could work on highways. “Working with a complex dynamical system such as a self-driving vehicle is not an easy task, given no support and no funding. But I can assure you that India will have self-driving vehicles in 2017”, he says. Adding a note on adoption, he says that it will depend largely on safety and prices at which it would be offered. Making a technology that is both highly robust and cost-effective would mean a lot of tweaking. He is already working in that direction by making perception possible by using cameras instead of LIDARS, which typically cost $50000. “My most recent work provided an alternative to LIDAR and it will cost just $1000”, says Sharma who got fascinated to self-driving vehicles during his undergrad, when he came across a video by MIT team in DARPA Grand and Urban Challenges. Since then he has been gathering all the advanced mathematical knowledge from undergrad along with a know-how on AI, machine learning, data mining, numerical optimization through books and research papers. And finally the project from his fifth semester on autonomous navigation and motion planning kick started the idea of founding Swaayatt Robots. “This project which continued in the form of theoretical research until August 2013, now plays an important role in the self-driving technology I am developing now”, he shares. Funding- Sharma has been bootstrapping Swaayatt Robots for more than 30 months now and once funded, he believes that the technology can be retrofitted on aftermarket products. “Whether funded or not, I will finish the proof-of-concept and demo it working on highways, campuses and in Bhopal city, in at least on three different routes. However, if I get funding, I will be able to do it a lot faster and a lot more effectively”. Swaayatt Robots recently got accepted into NVIDIA’s Inception Program, and they are considering the application for the NVIDIA GPU Ventures Program. Roadmap for the company- Sharma has set his focus very clear where he aims to have a proof-of-concept (POC) for campuses, highways and cities ready within 5 weeks, 2 months, and 4.5 months respectively. He has also been working on developing an advanced driver assistance system (ADAS), which is specifically tuned for Indian conditions and would be ready by the end of June and demoed in July. “By December I also want this technology to be tuned for selected few Indian cities like Bhopal and Bangalore, and few others. The plan is to be technologically capable of starting self-driving taxi services in these few selected cities, on selected routes (if unfunded), by December”, he shares. That’s not all, he also has plans to use this technology for Indian Defense, where a robot with intelligence and autonomous navigation capabilities can help soldiers. He plans to become a global player in the field of AI and robotics. Concluding note- He concludes by leaving a suggestion to the government of India that they should revise custom duties on equipment and hardware, that could help pure R&D startups like Swaayatt Robots in the long run. He also adds “Indian government should introduce a law that penalizes self-driving vehicle companies if their vehicle \/ technology makes a mistake and it results in a fatal accident—if the fault is of the self-driving software”.","excerpt":"There is no doubt that the interest around self driving technology has escalated like never before. With many big players like Google and Tesla entering the market, names like Tata Elxsi too have been trying hard to get India’s own driverless car. But given the complex traffic conditions and the unstructured Indian roads, the journey […]","categories":["IT Services"],"tags":["self-driving cars India"],"author_name":"Srishti Deoras","publish_date":"2017-06-05T05:27:23","publication_year":"2017","word_count":1159,"keywords":["Go","funding","machine learning","AI","RPA","ML","Scala","self-driving cars India","Aim","R","startup"],"extracted_tech_keywords":["AI","machine learning","ML","Aim","R","Go","Scala","RPA","startup","funding"],"url":"https:\/\/analyticsindiamag.com\/it-services\/meet-this-iit-engineer-developing-structured-self-driving-technology-for-unstructured-indian-roads\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":41759,"title":"Centre Joins Hands With IBM To Use AI &#038; Weather Prediction To Help Farmers","content":"Ministry of Agriculture and Farmers Welfare this week signed a Statement of Intent (SoI) for undertaking a pilot study with IBM India, where the company would generate solution in the field of agriculture through artificial intelligence and weather technology at the farm level to provide weather forecast and soil moisture information on pro bono basis. This information will be used by the farmers to take decisions regarding water and crop management for better production and productivity. According to a statement released by the Press Information Bureau, this project would be launched in the three districts of Bhopal, Rajkot and Nanded in the States of Madhya Pradesh, Gujarat and Maharashtra respectively. The SoI was signed in the presence of the Minister of Agriculture & Farmers Welfare Narendra Singh Tomar, Minister of State for Agriculture Shri Kailash Choudhary, and other senior officials of the Ministry. Speaking at the occasion, Tomar said ‘‘Our nation is foraying further into the digital age in the line of aspiration of Prime Minister Narendra Modi towards achieving the goal of doubling the farmers’ income and it has been the Ministry’s assurance to bring digital technologies to help our farmers to increase their income and transform Indian agriculture’’. The Minister also added that enabling the use of next-generation technologies such as AI and advanced weather data for better insights to make faster and more informed agricultural decisions is a testament to our commitment. As part of the collaboration, IBM Watson Decision Platform for Agriculture will be leveraged to obtain farm level weather forecast and village level soil moisture in three districts in Central & Western India. This pilot study will be conducted for the Kharif crop season 2019. Stressing on the term agri-scientists, the BJP-led government had earlier said that they would enable the development of young minds to take advantage of artificial intelligence, machine learning, blockchain and big data analytics for more predictive and profitable precision agriculture.","excerpt":"Ministry of Agriculture and Farmers Welfare this week signed a Statement of Intent (SoI) for undertaking a pilot study with IBM India, where the company would generate solution in the field of agriculture through artificial intelligence and weather technology at the farm level to provide weather forecast and soil moisture information on pro bono basis. […]","categories":["AI News"],"tags":["AI (Artificial Intelligence)","IBM"],"author_name":"Prajakta Hebbar","publish_date":"2019-07-04T12:07:05","publication_year":"2019","word_count":320,"keywords":["big data","Go","machine learning","artificial intelligence","AI","Git","RAG","Ray","IBM","analytics","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","analytics","Ray","RAG","R","Go","Git","big data"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/centre-joins-hands-with-ibm-to-use-ai-weather-prediction-to-help-farmers\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":65134,"title":"Dell Technologies Announces Open Source Networking Solutions For Data Centers","content":"Dell Technologies is announcing Enterprise SONiC Distribution by Dell Technologies, a new set of fully supported open-source networking solutions that help enterprises modernize and simplify the operations and management of their data centers at scale. As organizations increasingly rely on modern hybrid cloud models to do business, the historically monolithic and proprietary approach to networking has created inefficiencies and unneeded complexity. Dell Technologies is building off the work Microsoft spearheaded as part of the Software for Open Networking in the Cloud (SONiC) open source project. By integrating SONiC into the DNA of the Dell EMC PowerSwitch Open Networking hardware, Dell’s Enterprise SONiC Distribution removes complexity and creates an agile and flexible network through an approach built on open standards. “Our customers tell us that, while a hybrid cloud approach is critical to their success, they struggle to maintain and scale their networks, and manage them to effectively avoid multiple points of failure,” said Tom Burns, senior vice president and general manager, Dell Technologies Integrated Products & Solutions. “By breaking switch software into multiple, containerized components, we are providing enterprises the means to drastically simplify the management of massive and complex networks and increase availability in a cloud model.” With commercial-grade support for Dell’s Enterprise SONiC Distribution, cloud providers and large enterprises have a new level of production-tested tools, services and support that previously required significant internal investment technology support teams. It gives customers support for full-stack solutions with enterprise capabilities and advanced functionality. Yousef Khalidi, Corporate Vice President of Microsoft Azure Networking, Microsoft Corp. said, “SONiC is a leading open-source network switch OS, empowering customers with modern and efficient cloud networking software. We’re pleased to see Dell bringing enterprise support to their customers.” Dell Technologies provides a broad selection of open network operating systems, giving customers the ability to choose the hardware and software platform that best suits their priorities and take advantage of the benefits of open source innovation without sacrificing support or security and eases of integration into their own IT environments. Bob Laliberte, senior analyst and practice director, ESG said, “Dell Technologies has introduced a commercially-supported version of SONiC, offering tight integration to other Dell solutions and a clear roadmap designed, to not only support the hyperscalers of the world but also larger enterprises. As Dell continues to help customers with its open networking portfolio, its SONiC solution provides those organizations that have lacked the resources to adopt open source solutions the ability to take advantage of flexible, open source architectures.”","excerpt":"Dell Technologies is announcing Enterprise SONiC Distribution by Dell Technologies, a new set of fully supported open-source networking solutions that help enterprises modernize and simplify the operations and management of their data centers at scale. As organizations increasingly rely on modern hybrid cloud models to do business, the historically monolithic and proprietary approach to networking […]","categories":["AI News"],"tags":["data management providers"],"author_name":"Vinodh Ravindranath","publish_date":"2020-05-13T14:09:52","publication_year":"2020","word_count":414,"keywords":["cloud_platforms:Azure","cloud_platforms:Microsoft Azure","AI","programming_languages:R","R","innovation","BERT","llm_models:BERT","GAN","data management providers","Azure"],"extracted_tech_keywords":["AI","Azure","R","BERT","GAN","innovation","llm_models:BERT","cloud_platforms:Azure","cloud_platforms:Microsoft Azure","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/dell-technologies-announces-open-source-networking-solutions-for-data-center-innovation\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":68977,"title":"A Beginner’s Guide To Attention And Memory In Deep Learning","content":"It might have never occurred to you how you could make sense of what your friend is blabbering at a loud party. There are all kinds of noises in a party; then how come we are perfectly able to carry out a conversation? This question is known widely as the ‘cocktail party problem’. Most of our cognitive processes can pay attention to only a single activity at a time. In the case of a party house, our capability of directing attention towards one set of words while ignoring other sets of words, which are often overpowering, is still a conundrum. The key cognitive processes to solve a cocktail party problem are attention and short-term memory. This concept of attention and memory is a recurring theme in most of our daily processes. Being able to figure out what one sees or hears by just paying attention to chunks of information rather than the whole string of it, helps one react in real time. In short, it doesn’t consume too much memory. Processing and storage are also a huge challenge when it comes to computational systems. For machines, it’s all 0s and 1s and still, the concept of attention as we know it can be incorporated into them with well crafted algorithms. Attention and memory have emerged as two vital new components of deep learning over the last few years. The ability to focus on one thing and ignore others has a vital role in guiding cognition. Not only does this allow us to pick out salient information from noisy data (cocktail party problem), it also allows us to pursue one thought at a time, remember one event rather than all events. Attention in terms of neural networks can be thought of as a vector with important weights. The weights tell the network where to look at (attention), and which pixel in an image or a word in a sentence to look at. The attention vector is used to predict how strongly it is related to other elements and approximate the target. Attention And Memory In Machine Learning via Alex Graves, DeepMind Deep nets naturally learn a form of implicit attention where they respond more strongly to some parts of the data than others. RNNs contain a recursive hidden state and learn functions from sequences of inputs (e.g. a speech signal) to sequences of outputs (e.g. words). The underlying task here is to calculate a sequential Jacobian of an event. The sequential Jacobian, which is a set of derivatives,  shows which past inputs they remember when predicting current outputs. It is represented as follows: Where x and y denotes the input and output vectors respectively. The network produces an extra output vector used to parametrise an attention model. The attention model then operates on some data (image, audio sample, text to be translated…) to create a fixed-size “glimpse” vector that is passed to the network as input at the next time step. The complete system is recurrent, even if the network isn’t. We can elaborate this further by taking the example of the famous single headed attention RNN model(SHA-RNN) by Stephen Merity. The model consists of a trainable embedding layer, one or more layers of a stacked single head attention recurrent neural network (SHA-RNN), and a softmax classifier. The model uses a single head of attention and a modified feedforward layer similar to that in a Transformer, which is referred to as a Boom layer. Boom layer takes a vector from small (1024) to big (4096) to small (1024). The Boom layer is related strongly to the large feed-forward layer found in Transformers and other architectures. This layer minimizes computation and removes an entire matrix of parameters compared to traditional down-projection layers. Attention models generally work by defining a probability distribution over glimpses of the data given some set of attention outputs from the network. Also Read There are different types of attentions: Self-AttentionSelective AttentionIntrospective AttentionDifferentiable Visual AttentionAssociative Attention Applications of employing attention and memory modules into the network: Transformers, the models that have revolutionised NLP use attention mechanisms. When released in 2017, it set new benchmarks for machine translations. This innovation even led to better models such as BERT, GPT and many more. Hand writing synthesis with RNNsIn Differential Neural Computers Things To Remember: Selective attention appears to be as useful for deep learning as it is for peopleWe can use attention to attend to memory as well as directly to dataMany types of attention mechanism (content, spatial, visual, temporal…) can be definedAttention mechanisms revolutionised language modeling(ex:transformers) Also Watch:","excerpt":"It might have never occurred to you how you could make sense of what your friend is blabbering at a loud party. There are all kinds of noises in a party; then how come we are perfectly able to carry out a conversation? This question is known widely as the ‘cocktail party problem’. Most of […]","categories":["Deep Tech"],"tags":["NLP","self attention models","Transformers"],"author_name":"Ram Sagar","publish_date":"2020-07-04T13:00:19","publication_year":"2020","word_count":758,"keywords":["Go","machine learning","TPU","AI","neural network","self attention models","Transformers","RAG","NLP","deep learning","R"],"extracted_tech_keywords":["AI","machine learning","deep learning","neural network","NLP","Transformers","RAG","TPU","R","Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/memory-attention-deep-learning-guide\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10054596,"title":"How Parag Agrawal Impacted AI &#038; ML At Twitter","content":"Jack Dorsey announced that he would be stepping down from the CEO position at Twitter, corroborating rumours that had been floating around. This announcement sent the whole internet into a frenzy. Dorsey tweeted a screenshot of the email sent to Twitter’s team announcing his resignation, where he mentioned that Parag Agrawal, the former CTO of Twitter, will take over the post. Dorsey wrote very high praises for his successor. He said that Agrawal had been part of every “critical decision that helped the company turn around,” apart from mentioning that his trust in Agrawal as a CEO is “bone deep”. not sure anyone has heard but,I resigned from Twitter pic.twitter.com\/G5tUkSSxkl— jack (@jack) November 29, 2021 It was too late before congratulatory messages started pouring in for Agrawal, in fact, a little extra for being the latest Indian-origin person to lead a big tech company. He joins the likes of Sundar Pichai, Satya Nadella, and Arvind Krishna. However, not much was known about Agrawal until now, the fact that he himself acknowledged in his tweet. https:\/\/twitter.com\/paraga\/status\/1465349749607854083?s=20 In a press release, Twitter wrote, “Parag Agrawal joined Twitter in 2011 and has served as Chief Technology Officer (CTO) since October 2017. As CTO, he has been responsible for the company’s technical strategy, leading work to improve development velocity while advancing the state of Machine Learning across the company. Prior to being appointed CTO, Parag had risen to be Twitter’s first Distinguished Engineer due to his work across revenue and consumer engineering, including his impact on the re-acceleration of audience growth in 2016 and 2017.” Agrawal has been a driving force behind the platform’s AI and machine learning efforts. We trace some of his initiatives, especially the ones accomplished at Twitter. Career Graph Born in Mumbai to teacher and government official parents, Agrawal graduated in Computer Science Engineering from IIT Bombay (a few media reports suggest that he secured 77th rank in the entrance exam). He then moved to the US to pursue PhD from Stanford University (his thesis was on “Incorporating Uncertainty in Data Management and Integration”) Dean of Engineering at Stanford, Jennifer Widom, was Agrawal’s thesis advisor, who described him as thoughtful and analytical. His PhD was on methodologies to deal with messy and undefined data. My PhD advisor Jennifer Widom taught me how to learn an area of scientific research to be able to make original contributions. And gave me the skills to write and present technical content. But more importantly, she set me on a path of continuous learning. #SheInspiresMe— Parag Agrawal (@paraga) March 8, 2019 Agrawal became the first distinguished software engineer at Twitter. It is a technical strategist role that drives efforts on machine learning, cloud adoption, security, and quality. A distinguished engineer partners with directors and engineers, along working with Principal and senior staff engineers to guide, mentor and contribute to the design review process and ideation. He climbed ranks and, within a span of just six years, was elevated to the position of CTO. Agrawal’s forte has been machine learning, and under his aegis, the organisation achieved many breakthroughs in this space. ML-based Efforts at Twitter Agrawal is well known for leading the work on an ML-powered cropping algorithm developed after several users complained that the present algorithms favoured men with lighter skin. He led the initial investigation into the study of the bias and arranged the plan to remove it after it was concluded that bias was indeed present. ML bias has been high on the agenda for Agrawal. He was instrumental in the team’s efforts for public transparency on decision-making and ensuring that Responsible ML was one of Twitter’s official priorities for 2021. The ML Ethics, Transparency and Accountability (META) team was set up that consists of engineers, researchers, and data scientists who collaborate to recognise algorithms that may have been causing harm and also help Twitter recognise which issues to tackle first. Speaking from the perspective of a leader looking to grow my team and the field of Responsible ML, I am happy to hear this. META hired the best with promises of shifting industry and I intend to keep that promise as long as possible. @paraga has consistently supported us. 1\/ https:\/\/t.co\/RuIBrpOHwZ— ruchowdh.bsky.social (@ruchowdh) November 29, 2021 Agrawal helped Twitter shift to cloud computing services from Google and Amazon, thereby streamlining its operations. For example, in December last year, Twitter signed in on AWS for its global cloud infrastructure to deliver Twitter timelines and uses AWS’ infrastructure and products for tasks like computation, storage, security, etc. Yes! Thrilled to strengthen our partnership with @awscloud and continue working towards an even better, lower-latency experience for people using Twitter all around the world! https:\/\/t.co\/DYYamulG6x— Parag Agrawal (@paraga) December 16, 2020 Further, in 2019, Dorsey had put Agrawal in charge of BlueSky, Twitter’s efforts for decentralised social networking. It needs to be seen how Twitter’s tech strategy would change under the leadership of Agrawal.","excerpt":"Parag Agrawal has been a driving force behind the platform’s AI and machine learning efforts. We trace some of his initiatives.","categories":["AI Features"],"tags":["AI\/ML","IIT Bombay","jack dorsey","stanford university","Twitter (X)"],"author_name":"Shraddha Goled","publish_date":"2021-12-01T12:00:00","publication_year":"2021","word_count":819,"keywords":["Go","IIT Bombay","machine learning","AWS","AI","cloud computing","ML","RAG","jack dorsey","GAN","stanford university","Rust","AI\/ML","Twitter (X)","R"],"extracted_tech_keywords":["AI","machine learning","ML","RAG","cloud computing","AWS","R","Go","Rust","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-parag-agrawal-impacted-ai-ml-at-twitter\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":9196,"title":"Bitcoin, Blockchain and the Crypto Corporation","content":"If 2013 was the year of Bitcoin, the enigmatic crypto-currency created by the anonymous Satoshi Nakamoto, then 2016 is going to be year of the BlockChain — the shared public ledger technology that provides the platform on which Bitcoin works. In fact, Bitcoin is just ONE of the many applications that can be built on the blockchain and this fact is gradually dawning on the world of technology as different groups are racing to create new products. While many strange and wonderful products like an automatic Uber-like car service have been proposed, the most powerful applications seem to be coming out of Wall Street. This is because of two reasons The fundamental premise behind the blockchain is control and transfer of assets and this is what Wall Street does for a living Wall Street has realised that rather than resisting the arrival of a disruptive technology it is better adopt it first. Resistance is futile. Just as retailers who ignored the eCommerce revolution got wiped out by the arrival of Amazon and eBay, dealers in financial securities could face the same problem. Hence the urgency to board the blockchain bandwagon In fact so urgent is the desire to join the blockchain that DTCC, the body that controls the transfer of trillions of dollars of financial assets in the US market has teamed up with The Linux Foundation to plan for that “once-in-a-generation change of technology infrastructure” to create the HyperLedger and the Securities and Exchange Commission has authorised the release of equity shares on this alternate trading platform. The adoption of blockchain technology for the management of financial assets is as fundamental, if not truly tectonic, as the move from physical shares to de-materialised shares that we saw in the US in the 1970s and then in India in the 1990s. But what on earth is this blockchain? That is difficult to explain in a few words without trivializing the concept. Please be patient and work your way through these three slide decks and hopefully you will understand what this marvelous technology is all about. The first slide deck explains the basics of cryptography that are essential to understand how the bitcoins and other crypto-currency work. If you are already familiar with public and private keys you may skip this, but the idea of smart contracts is introduced here. The second slide deck is where the bulk of the bitcoin protocol is explained. We also explain how bitcoin or any other equivalent crypto-asset can be managed with the shared ledger also known as the blockchain. Finally, in the third slide deck we leave bitcoin behind and step into the mind-bending concept of an autonomous corporation. Where we look at futuristic concepts like programmable money and also some of the projects that are happening now and today. The blockchain creates new way to hold and transfer assets that represent real life “value”. It is going to transform the finance industry in the same way that the internet has transformed the way information is transferred and consumed. But could it  also go a long way to create a new kind of robotic entity and would that entity follow Asimov’s famous Three Laws of Robotics? In fact this technology is so significant that I would not be surprised if someone proposes that Satoshi Nakamoto be awarded the Nobel Prize for Economics even in absentia!","excerpt":"If 2013 was the year of Bitcoin, the enigmatic crypto-currency created by the anonymous Satoshi Nakamoto, then 2016 is going to be year of the BlockChain — the shared public ledger technology that provides the platform on which Bitcoin works. In fact, Bitcoin is just ONE of the many applications that can be built on […]","categories":["Deep Tech"],"tags":["bitcoin stock","Blockchain","Ethical Hacking"],"author_name":"Prithwis Mukerjee","publish_date":"2016-02-29T11:50:29","publication_year":"2016","word_count":561,"keywords":["Go","Blockchain","AWS","AI","cloud_platforms:AWS","programming_languages:R","programming_languages:Go","bitcoin stock","ai_applications:robotics","Tecton","Ethical Hacking","R"],"extracted_tech_keywords":["AI","Tecton","AWS","R","Go","cloud_platforms:AWS","programming_languages:R","programming_languages:Go","ai_applications:robotics"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/bitcoin-blockchain-and-the-crypto-corporation\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10140734,"title":"‘We’re Going to Take Everybody With Us,’ says NVIDIA’s Jensen Huang","content":"AI has officially entered every possible sector and industry there is. In a recent interview on No Priors, NVIDIA chief Jensen Huang expressed an inspiring vision for the future of AI integration across fields. “Nothing will be left behind. We’re going to take everybody with us,” he added. He emphasised the transformative impact of AI on various scientific and engineering fields, underlining that its influence is universal and far-reaching. “Everything from quantum computing to quantum chemistry…every field of science is involved in the approaches that we’re talking about,” said Jensen, stating that breakthroughs in AI and machine learning are changing the way all scientific domains operate. He further predicted a rapid shift, saying, “If we give ourselves a couple of years…the world’s going to change. There’s not going to be one paper, one breakthrough in science, one breakthrough in engineering where generative AI isn’t at the foundation of it.” AI Will Make Chips in the Future Huang discussed the future role of AI employees in industries, particularly in chip design and research, and how they will reshape key business functions, including marketing and supply chain management. “I’m hoping that NVIDIA is someday biologically bigger, but also, from an artificial intelligence perspective, much, much bigger,” said Huang, saying that NVIDIA already has AI chip designers and AI software engineers. He sees chip design as the most important area for AI development, saying that AI chip designers could tackle some of the most complex challenges in the field. Huang said that NVIDIA currently works with engineers from Synopsys and Cadence. However, he added that in the future, NVIDIA will work with AI engineers trained by them. “I totally imagine them having Synopsys chip designers that I can rent, and they’ve trained an AI to be incredibly good at it. We’ll just hire a whole bunch of them whenever we need,” Huang explained. SaaS is Sitting on a Gold Mine Looking beyond chip design, Huang spoke about the future of SaaS platforms, suggesting that rather than being disrupted, they will thrive by incorporating AI agents. “These platforms are sitting on a gold mine,” he said. “There’s going to be a flourishing of agents that are specialised in Salesforce, specialised in SAP, and other platforms.” Huang believes these AI agents will work together to solve problems more efficiently, collaborating across various tools and systems. Huang also highlighted the growing impact of AI on scientific fields, emphasising that AI and machine learning are set to transform areas such as quantum chemistry, theoretical math, and computer science. Reflecting on past milestones in AI development, including breakthroughs in computer vision, Huang said, “I saw the early indications of it… and we were fortunate to have extrapolated from what was observed to be detecting cats to superhuman levels of capabilities in object recognition.”","excerpt":"Huang said that, in the next couple of years, there will not be a single paper or breakthrough in science or engineering where generative AI won’t be at its foundation.","categories":["AI News"],"tags":["NVIDIA"],"author_name":"Siddharth Jindal","publish_date":"2024-11-09T16:38:50","publication_year":"2024","word_count":465,"keywords":["Go","API","machine learning","artificial intelligence","AI","AI employees","computer vision","AI agents","generative AI","NVIDIA","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","computer vision","generative AI","R","Go","API","AI employees","AI agents"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/were-going-to-take-everybody-with-us-says-nvidias-jensen-huang\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":35957,"title":"Top 5 Successful Indian AI Startups Led By Women","content":"Image source: twitter.com\/madstreetden The data and analytics market in India is booming and all the major players are riding the AI wave with great enthusiasm. By carving a niche for themselves in the field, many self-made women entrepreneurs have ventured in the field of AI & ML in a hope to make an impact. These women who come from all walks of life have set an example for others when it comes to following one’s passion. We look at some of the prominent names and their contribution to the field. Barkha Sharma, CEO and Founder, Bash.ai: Since its inception in March 2017, Bash.ai has been automating HR processes and thereby helping in better and consistent employee experiences. Founded by Barkha Sharma, who has spent over eight years in HR. The startup was started at Sharma’s living room and has now 12 people strong with engineers, product developers, and operations enablers working with HR processes of organisations. Sharma’s belief that in a world of automation the functions of HR are a bit outdated, and her idea to fill this technological gap led to the formation of the startup and it raised the first round of funding in October 2017. Bash.ai uses AI and big data to power virtual assistants and drive HR for businesses by mimicking cognitive functions related to HR. Its user-friendly front end interface gives organisations a highly accurate and real-time way to automate conversations with employees. It can be accessed using instant messenger platform such as Facebook Messenger, Slack, Skype, and currently has modules such as post-hire orientation, ticketing, HR helpdesk, employee engagement, organise HR activities and answer questions related to payslips, company policies, among others. They want to go towards an ecosystem where AI and humans coexist efficiently in the HR settings. Niyati Agarwal, Morph.ai: Morph.ai is an enterprise chatbot founded by Agarwal along with her friends Pratik Jain, Vipul Garg, and Abhishek Gupta in 2016. Having worked with unicorn social media company previously, their knowledge about how the enterprises were handling social media inspired them to set up the startup. Morph.ai uses automation as a tool  to help companies improve their working with related to its social media handling. My digitising messaging, the startup has revolutionised how chatbots interact with its customers. Within a short time, the startup has managed to rope in some of the big names like English Premier League Club, Manchester City, Yes Bank, GameVision and USL as their customer. Ashwini Asokan, Mad Street Den: Chennai-based Mad Street Den (MSD) was started by Asokan along with her husband in 2013. The startup uses computer vision and AI-based tools for businesses around the world. Previously, Asokan has worked in Intel Labs and led their Mobile Innovation Portfolio and was in charge of driving research and development of cutting edge mobile products. In 2018, MSD raised a funding of undisclosed amount from Japanese Telecom operator KDDI  through its open innovation fund. Prior to this, in 2016 the startup raised Series A funding from Sequoia India, Exfinity Ventures & growX Ventures. Asokan has been quite vocal about the role of women in the tech industry and has been featured in countless publication for her take on managing work and motherhood. Speaking to a leading online portal Asokan says that at her organisation there is equal number of men and women and with the startup, she hopes to bring more women to the tech field. Lizzie Chapman,  ZestMoney: When Londoner Lizzie Chapman moved to India to set up ZestMoney with  Priya Sharma and Ashish Anantharaman, she didn’t expect that bankers and insurers would be more open to a woman leadership when compared to the west. “Sometimes we do deal with patronising elements of the financial services industry here, but it’s much more a boys club in the UK and US,” Chapman told a national daily. Founded in 2015, this startup leverages the experience of the founding team in digital payments ecosystem to developed easy to configure micro-services that help lenders and e-commerce partners integrate quickly. This along with a consumer-first risk engine and transparent pricing allows its  end users to afford higher value products, doubling AOV (average order value) for its partners and expanding the reach of its lenders. Gazal Karla, Rivigo:  Co-founded Rivigo which addresses the problems faced by logistics company through AI, big data and IoT solutions. The company owns almost 21,000 trucks across the country and is spread across 150 locations in India. Karla has worked with the World Bank and McKinsey previously before starting her own venture. She also has her credit an exceptional academics records with an MBA from Stanford University and a masters in Public Administration from the Harvard Kennedy School of  Government and is also a graduate of IIT Delhi Speaking to a leading daily about her vision, Karla said, “My vision is to make the transportation and logistics sector of India strongest in the world. As you know this sector is the backbone of any economy, and in India, there are about 1 crore truck drivers who see this as their only livelihood, but seasonal work doesn’t pay them well. Their basic needs are not looked after, they also want to spend time with their family and hence nobody wants a truck driver. Soon, India and other nations like the US will run out of truck-drivers! It’s a big problem to be addressed.","excerpt":"The data and analytics market in India is booming and all the major players are riding the AI wave with great enthusiasm. By carving a niche for themselves in the field, many self-made women entrepreneurs have ventured in the field of AI & ML in a hope to make an impact. These women who come […]","categories":["AI Trends"],"tags":["ML","Women in Tech"],"author_name":"Akshaya Asokan","publish_date":"2019-03-07T11:21:04","publication_year":"2019","word_count":894,"keywords":["Go","AI","chatbots","ML","virtual assistants","computer vision","RAG","Git","Women in Tech","analytics","R"],"extracted_tech_keywords":["AI","ML","computer vision","analytics","RAG","chatbots","virtual assistants","R","Go","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-5-successful-indian-ai-startups-led-by-women\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10007552,"title":"5 Top Challenges Faced In Operationalising AI","content":"With surmounting interest in data science and the fast-growing Data Scientist community, AI as a technology has come a long way crossing the chasm from Innovators and early adopters to the Early Majority. Along with all the hype that’s there today around AI, there is still the unaddressed issue of less than 12% models reaching the production stage Data Scientists are creating models day in and day out but there are millions of models that are still waiting to see the light of the day in production. Gartner identified time to value as one of their biggest challenges, reporting that it takes an average of 52 business days for the team to build a predictive model, and longer to deploy into production While the usual belief is that the deployment should need fewer days than building a model but it is becoming the most challenging issue of the industry today. Building the model is one thing, what’s more, challenging is operationalizing AI. There are significant challenges in Operationalizing AI 1. Analytics challenged leadership: This one serves as the major hurdle in operationalizing AI. The senior leaders in the Organizations are not that conversant in AI and hence lose out to the AI Integrated Organizations. Leadership support is the most important factor in operationalizing AI. In the words of a senior leader of one of the large Private Sector Bank “Our Organization started losing to the new start-ups in this space, that’s when our CEO decided to adopt AI. It took us 3 years to regain the market share which we lost due to new entrants. Our CEO made it mandatory for the entire leadership to pursue courses in Data Science, that was just the beginning, we started having weekly knowledge sharing sessions, Innovation lab workshops etc.” 2. Data Quality Gartner predicted that through 2022, 85 percent of AI projects will deliver erroneous outcomes due to bias in data, algorithms or the teams responsible for managing them. Testing and Validation is practiced in the controlled environment and hence the data used is of good quality while when it comes to deploying the model in production, it has to work on real-world data which in most of the cases lacks quality which results in low Model accuracy 3. Legacy Infrastructure Another big challenge in operationalizing AI is the issue of legacy infra of the large organizations which makes it impossible to operationalize AI. 4. Managing the Compute power Compute power requirement depends a lot on the kind of model being worked upon. In the case of deep learning, computer vision models the compute power for the training data set is high while we can use 1\/nth compute power in production. But for the KNN models, the compute power in production needs to be high. So optimizing the compute power in each stage of analytical modelling is key to the success of Operationalizing AI in organizations. 5. Interpretability Interpretability is essential for operationalizing anything today. People need to understand what’s going on in the black box. This goes back to the first point we discussed around Senior Management understanding of AI. The first step to start any data science initiative in the Organization is to present the business case or get the budget. The moment you decide doing something using Neural Network to achieve High Accuracy it becomes a black box for the management as the high accuracy comes at the cost of low interpretability. Though as a concept you can explain the management on what’s happening within the solution but mostly the solution is a black box for the organizations. For example risk scoring using the Neural Networks can be highly accurate but difficult to explain. Adoption of AI is a major point of discussion across the boardrooms today but to move beyond discussion we need to overcome the above-mentioned challenges. Space of data and analytics is changing rapidly, to remain significant in these times, Organizations will have to make sure that they take every step required for operationalizing AI. References: https:\/\/www.forbes.com\/sites\/gilpress\/2019\/11\/22\/top-artificial-intelligence-ai-predictions-for-2020-from-idc-and-forrester\/#4fcfebef315a","excerpt":"With surmounting interest in data science and the fast-growing Data Scientist community, AI as a technology has come a long way crossing the chasm from Innovators and early adopters to the Early Majority. Along with all the hype that’s there today around AI, there is still the unaddressed issue of less than 12% models reaching […]","categories":["AI Trends"],"tags":[],"author_name":"Devansh Sharma","publish_date":"2020-09-17T19:00:17","publication_year":"2020","word_count":670,"keywords":["data science","Go","API","AI","neural network","computer vision","RAG","deep learning","analytics","R"],"extracted_tech_keywords":["AI","deep learning","neural network","computer vision","data science","analytics","RAG","R","Go","API"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/challenges-faced-in-operationalizing-ai\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10086254,"title":"ChatGPT&#8217;s New Update Sucks Factually, Mathematically","content":"ChatGPT creator OpenAI yesterday released a new update– “with improved factuality and mathematical capabilities.” So, we tried. Turns out, it is still making the same errors. We knew than ChatGPT is bad at maths, but we did not know that even after the latest upgrade of Jan 30th, it is still giving incorrect results. Read more: Why is Everyone Bashing ChatGPT? Not just maths, it is still proficient in giving factually incorrect answers as it is trained on pre-2021 data. New Update, Old Joke Recently, a post has been going around the social media where ChatGPT hilariously agrees to a person when he coaxes it to believe that his spouse is right even though the answer is wrong. Last month, Stack Overflow blocked all answers from ChatGPT citing inaccurate responses from the bot. Although Stack Overflow stated that this ban was temporary, it emphasized that the issue was not just with the incorrect answers but also with how the bot phrased them. ChatGPT is based on GPT-3.5, which can generate sentences that are grammatically correct and have a formal tone. However, as per the nature of LLMs, this also makes the responses seem overly confident and authoritative, which can be misleading. Despite ChatGPT has passing the bar exam and MBA finals, it still has the problem of bias and lacks logical reasoning besides being terrible at maths. Although OpenAI claims to be working towards minimizing wrong answers, no one knows how long shall it be, especially since it is a dedicated responsible AI research lab.CEO Sam Altman took to twitter to share how the team is working towards making ChatGPT perfect. we know that ChatGPT has shortcomings around bias, and are working to improve it.but directing hate at individual OAI employees because of this is appalling. hit me all you want, but attacking other people here doesn’t help the field advance, and the people doing it know that.— Sam Altman (@sama) February 1, 2023 Read more: Freaky ChatGPT Fails that Caught Our Eyes","excerpt":"OpenAI said in a blog post that ChatGPT is now updated with improved factuality and maths skills but results don’t say so.","categories":["AI News"],"tags":["chat gpt app","chat gpt download","chat gpt login","chat gpt sign up","chat gpt website","ChatGPT","GPT-3","how to use chat gpt","Microsoft","OpenAI"],"author_name":"Shritama Saha","publish_date":"2023-01-31T15:14:57","publication_year":"2023","word_count":333,"keywords":["GPT-3","Go","chat gpt login","chat gpt app","ChatGPT","OpenAI","AI","AI employees","chat gpt download","GPT","responsible AI","chat gpt website","Aim","AI research","how to use chat gpt","R","Microsoft","chat gpt sign up"],"extracted_tech_keywords":["AI","ChatGPT","OpenAI","Aim","R","Go","GPT","responsible AI","AI employees","AI research"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/chatgpts-new-update-sucks-factually-mathematically\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10122277,"title":"Kunal Shah Tells Entrepreneurs to Take Two Days Off a Week to Think","content":"In a recent podcast, Kunal Shah, the founder and CEO of Cred, revealed his strategies and ideas behind building a successful business. One of the things he advised entrepreneurs is to take two days off every week to think. “I don’t think people do that because it feels like cheating,” said Shah, that entrepreneurs are hesitant to take two days off in the middle of the week. He also said that many entrepreneurs love ‘sangharsh’, which stands for struggle. This communicates that it is important for entrepreneurs to take some time to think and reflect on what they are building maximising productivity. “There is no way you could be cute and a top ranker in IIT,” said Shah, explaining that people who have struggled to build their companies and have worked hard in their lives, naturally assume that struggle is the way forward. “Anything which is not ‘sangharsh’ is cheating for you,” he added. But Shah said that wealth does not necessarily come by just struggling. “I am not saying that people who are successful do not work hard, but their time is spent in very different ways,” Shah added, about how entrepreneurs do not take leaves during the week. This brings back the conversation about when Narayana Murthy, the Infosys co-founder said that Indian youth should work 70-hours a week. “Performance leads to recognition, recognition leads to respect, respect leads to power,” he said and asked the “wonderful youth of the country” to realise this and work 12 hours a day. Recalling his philosophical question asked to Sam Altman when he visited India about what he learned about humans, Shah said that Altman answered that he discovered that intelligence is just a property of matter. “If we give birth to an AI or anything else that is superior to us…the wealth might shift over there and we may not exist as a species,” Shah explained reflecting on the question and that humans have always been an asymmetric information gathering systems, and the same would be true for AI systems.","excerpt":"“There is no way you could be cute and a top ranker in IIT,” said Shah, explaining that people who have struggled to build their companies and have worked hard in their lives, naturally assume that struggle is the way forward.","categories":["AI News"],"tags":["kunal shah"],"author_name":"Mohit Pandey","publish_date":"2024-06-03T11:35:40","publication_year":"2024","word_count":340,"keywords":["programming_languages:R","AI","kunal shah","Ray","ViT","R"],"extracted_tech_keywords":["AI","Ray","R","ViT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/kunal-shah-tells-entrepreneurs-to-take-two-days-off-a-week-to-think\/","complexity_score":2,"technical_depth":5,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10049312,"title":"Mihir Kittur","content":"As Co-Founder and Chief Commercial Officer at Ugam, a leading analytics and technology services company, Mihir Kittur is responsible for delivering customer success. Ugam, recently certified as a Great Place to Work, drives impactful business results for large corporations by leveraging data, technology, and expertise. Mihir is a firm believer that every customer ask must be approached with “What are we trying to solve for?” and then leverage the necessary AI\/analytics to solve it (“Intelligence before Artificial”). Mihir has been influential in embedding this approach through multidisciplinary teams comprising of soft skills like consulting and problem-solving, and hard skills like statistical techniques and technology. This has enabled superior impact across functions for use cases such as ML-enabled personalisation at scale, digital channel performance improvement, price optimisation, etc. It is because of such impact that Ugam was named the No. 1 data science company in India by Analytics Insight. Furthermore, Mihir has played a major role in pioneering Ugam’s JARVIS, a proprietary cognitive computing system. It is built on tried, tested, and a tailored suite of machine learning algorithms that enable computer vision, Natural Language Processing (NLP), sentiment analysis, and more. It also powers Ugam’s suite of tools that can be deployed for customers, such as SKU rationalisation, macro space management, and eCommerce workbench. Recently, at Ugam’s Golden Pyramid Awards, their multidisciplinary team was recognised for their trailblazing breakthrough in ‘making data science accessible’ by enabling the development of ML models using a no-code technique. Outside Ugam, Mihir speaks at marquee industry events and contributes to thought leadership pieces in leading publications. In addition, he has co-authored a widely read book on pricing intelligence. He also serves on the Analytics Advisory Board for a leading management school and mentors several analytics start-ups. LinkedIn","excerpt":"As Co-Founder and Chief Commercial Officer at Ugam, a leading analytics and technology services company, Mihir Kittur is responsible for delivering customer success. Ugam, recently certified as a Great Place to Work, drives impactful business results for large corporations by leveraging data, technology, and expertise. Mihir is a firm believer that every customer ask must […]","categories":["AI Features"],"tags":["Interviews and Discussions","ugam"],"author_name":"AIM Media House","publish_date":"2021-09-22T19:00:50","publication_year":"2021","word_count":292,"keywords":["data science","ugam","machine learning","AI","sentiment analysis","ML","computer vision","RAG","NLP","analytics","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","machine learning","ML","NLP","computer vision","data science","analytics","RAG","sentiment analysis","R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/mihir-kittur\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":1052,"title":"From 5G to smart city solutions, Nokia builds a smarter India","content":"Nokia’s India unit is reportedly rolling out 5G and Internet of Things (IoT) services for smart cities and overall public safety.  The Finnish mobile manufacturer, known for its handy and stodgy handset, unveiled more than 60 use cases of its technology at Innovation Day held at Nokia Technology centre in Bengaluru. The annual event saw ideas for a smarter and safer environment powered by Nokia’s IoT cloud solution IMPACT that could be used by countries in the future. At the annual event, Nokia amply demonstrated how it is making IoT a reality for smart cities and public safety with its three innovative offerings in a) Smart Parking, b) Video Analytics and c) Rail Road crossing. The technology behind it leverages: In- house build sensors IMPACT, MEC platform End to end Network based solution for trespassing detection Aimed at governments and enterprise who can use it for powering smart city solutions, Nokia’s IoT solutions are designed to provide a secure and scalable platform that use data to build smart, safe and sustainable cities. Randeep Raina, Nokia’s Chief Technology Officer who was present at the event spoke how IoT solutions could provide a secure and scalable platform for smart and sustainable cities. “Government initiatives such as Digital India and smart city initiatives are enablers to make cities safer and smarter. Keeping that this in mind, we decided to roll out the theme centred on smart cities,” he shared. Nokia’s IoT use cases: a) Smart Parking: This solution provides pre-booking, parking slots, recording the number of hours for the respective slot and aids in traffic management. b) Video Analytics: Provides surveillance capability by identifying intruders at a location. c) Rail Road crossing: One of the best life-saving initiative that could have a wide application in India, this solution uses sensors and beacons and can warn pedestrians of incoming trains at crossings and also can be used as a disciplinary measure for authorities. While 5G is still in early stages, IoT is seeing a big adoption in India, revealed Raina. However, when it comes to 5G, Nokia has established thought leadership and explored many possibilities as a platform for enabling growth, right from IT sector to agriculture, manufacturing and entertainment. According to Nokia, “network speeds as high as 10Gbps and with extremely low latency are a driving force for new applications that use massive broadband capabilities”.","excerpt":"Nokia’s India unit is reportedly rolling out 5G and Internet of Things (IoT) services for smart cities and overall public safety.  The Finnish mobile manufacturer, known for its handy and stodgy handset, unveiled more than 60 use cases of its technology at Innovation Day held at Nokia Technology centre in Bengaluru. The annual event saw […]","categories":["AI News"],"tags":[],"author_name":"Richa Bhatia","publish_date":"2016-12-29T11:31:40","publication_year":"2016","word_count":392,"keywords":["Go","programming_languages:R","AI","innovation","Scala","Git","RAG","Aim","analytics","R"],"extracted_tech_keywords":["AI","analytics","Aim","RAG","R","Go","Scala","Git","innovation","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/5g-smart-city-solutions-nokia-builds-smarter-india\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":52076,"title":"TCS Launches New Blockchain Deployment Kit That Can Help Enterprises Build Apps 40% Faster","content":"Tata Consultancy Services, a noted global IT services, consulting, and business solutions organisation, announced the general availability of its Quartz DevKit, an intuitive, low code blockchain deployment kit for enterprises to quickly build and deploy blockchain applications on any popular blockchain platform. The DevKit abstracts out the complexity of the underlying blockchain technology and provides a low code approach to build and deploy blockchain applications on any preferred blockchain platform – such as Hyperledger Fabric, Ethereum or R3 Corda – while allowing for platform-specific code to be written as an extension over the pre-built templates. Using its pre-built components for platform setup, administration and deployment, platform security authentication, encryption, and node as well as identity and user management, programmers can write smart contracts 40% faster. Its intuitive, web-based integrated development environment helps programmers seamlessly design, compile, package, and deploy business code on their preferred blockchain platform. Additionally, the DevKit’s integrated code quality analyzer ensures that the smart contracts incorporate best in class coding practices. “Many of our customers, across industries, are leveraging blockchain technology to establish frictionless transactions across their extended ecosystem. We developed the Quartz DevKit to help their teams rapidly put together high-quality pilots using smart contracts on any platform with reduced coding effort. We have received very positive feedback from our pilot customers, and are pleased to make the DevKit available for use at scale,” said R Vivekanand, Global Head, Quartz, TCS. Quartz – the Smart LedgersTM solution, redefines how organizations can collaborate in an increasingly connected world, and leverage their collective strengths using blockchain technology. The suite comprises Smart Solutions, a set of ‘designed for blockchain’ business offerings for different industries; the Quartz DevKit, a smart contract development kit to enable programming of high-quality code on multiple blockchain platforms; the Quartz Gateway for the integration of existing solutions with blockchain ecosystems; and, the Quartz Command Center that can administer and monitor entire ecosystems. Together, these solutions can help organizations set up truly connected ecosystems that run on blockchain technology, delivering real-time, efficient transaction processing based on a single source of truth on the ledger. The solutions can coexist with existing systems, support multiple blockchain platforms, and facilitate seamless integration with ecosystem participants.","excerpt":"Tata Consultancy Services, a noted global IT services, consulting, and business solutions organisation, announced the general availability of its Quartz DevKit, an intuitive, low code blockchain deployment kit for enterprises to quickly build and deploy blockchain applications on any popular blockchain platform. The DevKit abstracts out the complexity of the underlying blockchain technology and provides […]","categories":["AI News"],"tags":["Blockchain","TCS"],"author_name":"Prajakta Hebbar","publish_date":"2019-12-17T12:50:39","publication_year":"2019","word_count":366,"keywords":["API","Blockchain","programming_languages:R","AI","R","ML","RAG","GAN","TCS"],"extracted_tech_keywords":["AI","ML","RAG","R","API","GAN","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/tcs-blockchain-deployment-kit-quartz-devkit\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":35852,"title":"10 Popular IDEs Java Developers Can Use","content":"Integrated Development Environment (IDE) is crucial while coding for larger programming projects as they help in handling the code smoothly by providing syntax highlighting, auto-completion of code, tools debugging and other features. JavaScript is a dynamic language that also has gained prominence for backend development with Node.js. In this article, we list down 10 IDEs for JavaScript. Some of them are cross-platform IDEs. 1| Atom IDE Atom IDE is released from beta and is supported by the open-source community. This open-source text and code editor enables the users to install third-party packages in order to customise the features according to your preference. The editor is supported for plug-ins written in Node.js and embedded Git Control which is developed by GitHub. Atom is based on Electron, a framework which enables cross-platform desktop applications using Chromium and Node.js. Click here to get started with Atom. 2| Brackets This open-source modern text editor is able to understand web design and is easy to design in the browser. It is a lightweight and powerful text-editor which is blended with visual tools to get the right amount of support whenever needed. This editor lets you open a window into the code instead of switching between the file tabs and you will also be able to get a real-time connection to your browser. Click here to get started with Brackets. 3| Eclipse 2018 with JavaScript Development Tools This is one of the most used Java IDEs which provides Eclipse Platform plug-ins that implements an IDE supporting the development of JavaScript applications and JavaScript within web applications. It adds a JavaScript project type and perspective to the Eclipse Workbench as well as a number of views, editors, wizards, and builders. It includes features such as Bower, npm, JSON editor, ECMAScript 2015 (ES6) parser, Node.js support and JavaScript Build Tools (Gulp \/ Grunt). Click here to get started with Eclipse 2018. 4| Komodo Edit Komodo Edit is an integrated,multi-platform as well as multi-language development environment for the creation of dynamic web applications. It includes basic functions including syntax checking, auto-completion, language highlighting. This IDE works on various dynamic languages such as Ruby, Python, Perl, PHP, HTML and many more. Click here to get started with Komodo Edit. 5| Light Table This IDE is developed by Chris Granger and Robert Attorri which features real-time feedback allowing instant execution, debugging as well as access to documentation. It is easily customisable and is able to display anything that a Chromium browser can. It also has a powerful plug-in system that allows almost any aspect of the editor to be extended and customised. Click here to get started with Light Table. 6| NetBeans 10.0 Apache NetBeans 10.0 is the second major release of the Apache NetBeans IDE which was released in Dec 2018. This is much more than a text editor as it highlights source code syntactically as well as semantically. It provides editors, wizards, etc. to help you create applications in Java, PHP, and many other languages. To build Apache NetBeans (incubating) 10.0 from the source you need Oracle’s Java 8 or Open JDK v8 or Apache Ant 1.10 or greater. Click here to get started with NetBeans 10.0. 7| RJ TextEd This IDE is a full-featured text and source editor with Unicode support and is very powerful as it supports a number of languages including JavaScript, PHP, HTML, and many others. The functionality extends beyond text file and includes features like spell checking, auto-completion, email support,  HTML validation, advanced with regular expressions, templates and many more. Click here to get started with RJ TextEd. 8| Visual Studio Code This IDE is a lightweight yet powerful source code editor. It comes with built-in support for JavaScript, Node.js, and TypeScript and has extensions for various other languages including Python, PHP, C#, etc. It offers smart completion based on variable types with IntelliSense, built-in Git Integration, debugging code from the editor, etc. Click here to get started with Visual Studio Code. 9| Visual Studio 2017 This is a fully featured integrated development environment for Android, iOS, Windows, web, etc. This IDE enables you to write code accurately and efficiently without losing the current file context and guides you with live assistants as you write codes. It has extensions for various languages like Python, JavaScript, C++, and many others. You can easily zoom into details such as call structure, related functions, check-ins, and test status. Click here to get started with Visual Studio Code. 10| WebStorm This is one of the most popular IDEs offered by JetBrains which is used for JavaScript development. The features include in this IDE such as easy identification of errors, code completion and also has a built-in debugger for the client-side code including Node.js applications. Click here to get started with Webstorm.","excerpt":"Integrated Development Environment (IDE) is crucial while coding for larger programming projects as they help in handling the code smoothly by providing syntax highlighting, auto-completion of code, tools debugging and other features. JavaScript is a dynamic language that also has gained prominence for backend development with Node.js. In this article, we list down 10 IDEs […]","categories":["AI Trends"],"tags":["Javascript"],"author_name":"Ambika Choudhury","publish_date":"2019-03-06T09:36:45","publication_year":"2019","word_count":793,"keywords":["AI","Javascript","ML","Git","TypeScript","Python","C++","JavaScript","GitHub","R","Java"],"extracted_tech_keywords":["AI","ML","Python","R","JavaScript","TypeScript","Java","C++","Git","GitHub"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/10-popular-js-ides-java-developers-can-use\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":35777,"title":"Who’s Afraid Of Artificial Super Intelligence?","content":"Image source: media1.giphy Humans are already heavily depended on machines in some way or the other and in the meantime, developments in the field of artificial intelligence and machine learning are moving at a breakneck speed.  Though Artificial Super Intelligence (ASI) is the thing of the future, various research and studies are underway to make computers think beyond what humans are capable of. In other words, ASI is about improving the cognitive ability of computers, making it far more superior than human, While AI is just making systems mimic human thoughts. It further comes under the category of what is referred to as strong AI and there is a large skepticism surrounding the technology. As described by Oxford professor, Nick Bostrom, ASI is “Any intellect that greatly exceeds the cognitive performance of humans in virtually all domains of interest”. Where are we in terms of advancement? While in the recent past, systems like IBM Watson and DeepMind’s AlphaGO have proven to be more intelligent than humans by outdoing them, the current AI systems haven’t turned itself into a form that would be of potential threat to humanity. But according to many researchers, ASI would be achievable only after the Artificial General Intelligence wave, where machines will be able to procure intelligence equivalent to that of humans. ASI is presently seen as a hypothetical situation as depicted in movies and science-fiction books- a time of doom and gloom, where machines have taken over the world.  Largely, it is believed to happen due to an intelligence explosion and has been associated with technological singularity, a time-frame in the future when machines can advance human to foresee and control the outcomes of the future The objections towards the technology Researchers have been arguing both in favour of the technology as well as against it. But, largely fear and the potential harm that could inflict upon humanity overshadows its positive capabilities. One of the biggest fear associated with technology has been that of its unstoppability. Experts feel that, with the onset of intelligence explosion, the technology will be so advanced that it will turn more powerful and cannot be stopped by humans. The ability of the ASI-enabled systems will be so much so that, the machines will be able to produce several outcomes and prevent any attempts to stop it from achieving its goal and it can result in uncontrollable and unintended consequences. Speaking about the capability of the technology, Eliezer Yudkowsky, acclaimed AI-researcher who is best known for his work towards friendly AI said,” The AI does not hate you, nor does it love you, but you are made out of atoms which it can use for something else.” However, not everyone is willing to cast the technology as the primary reason for the apocalypse. While establishing that strong AI, ASI and AGI will surpass humans in terms of intelligence, philosopher  David Chalmers believes that this technology will lead to superhuman intelligence which will enable the computer to naturally achieve what he calls “recursive self-improvement”, where machines would be able to reprogram and improve its function without human intervention. Beyond narrow AI We are presently at the cusp of Narrow AI, where the systems are tasked with one particular job, it could be automated spam filtering or recommendations made on e-commerce site, to name a few. In other words, it is less complicated when compared to advanced technologies like that of IBM Watson and are seen as a weak AI. However, researches are underway to make existing AI systems more complex and by aiming for ASI, companies want  to train their present narrow AI systems to achieve more complex and multiple tasks. Though ASI still remains a hypothetical concept, experts believe that it isn’t too long before the technology to become a reality. Despite the fear and complexities associated with the technology, it is understood to have the capabilities to solve some of the pressing challenges of our time, like climate change, feminine and even disease breakout. While the conversation is largely about a dystopian future, it is still too early to fear the machine’s capability or predict the future course of action.","excerpt":"Humans are already heavily depended on machines in some way or the other and in the meantime, developments in the field of artificial intelligence and machine learning are moving at a breakneck speed.  Though Artificial Super Intelligence (ASI) is the thing of the future, various research and studies are underway to make computers think beyond […]","categories":["AI Features"],"tags":["AGI","AI and machine learning"],"author_name":"Akshaya Asokan","publish_date":"2019-03-05T12:39:14","publication_year":"2019","word_count":690,"keywords":["Go","artificial intelligence","machine learning","programming_languages:R","AI","RPA","programming_languages:Go","Aim","AI and machine learning","AGI","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","Aim","R","Go","RPA","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/whos-afraid-of-artificial-super-intelligence\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10140983,"title":"What Quantum Cryptography Solutions Mean for India’s Security","content":"In July this year, the ministry of science & technology announced that a team of Indian scientists had created a user-friendly method for generating unpredictable random numbers. This breakthrough was crucial for stronger quantum data encryption and robust cybersecurity. But does India truly need this solution? Quantum encryption, also known as quantum cryptography, refers to advanced cybersecurity methods that secure data through the immutable laws of quantum mechanics. QNu Labs, a pioneer in quantum cryptography solutions, is on a mission to develop world-class technology in India. In an exclusive interview with AIM, Rahil Patel, chief growth officer at QNu Labs, explained, “Historically, India has been strong in software and application development, but we’ve rarely led in hardware or cutting-edge technology that can be adopted globally. QNu Labs wants to change that.” According to Patel, India has often been viewed as an adopter of innovations from other countries, but today, new policies enable domestic manufacturing. “We aim to go further and create technology of such quality and innovation that advanced economies would want to import from us,” he added. Given India’s rapid digitisation, with initiatives like Aadhaar and UPI, and advances in finance and capital markets, enhanced security is essential to avoid external threats. “Our goal became twofold: to develop a technology that instils national pride and to safeguard India’s digital and critical infrastructure,” Patel noted. Besides finance, aerospace and defence sectors demand the highest level of security to protect against sophisticated cyber threats. QNu Labs has partnered with the Indian defence forces to deploy the Armos Quantum Key Distribution system, enhancing secure wireless communications in critical field operations. Additionally, it has collaborated with aerospace leaders to integrate quantum-safe encryption into their infrastructures, safeguarding satellite communications and navigation data against cyber-physical threats. It aims to establish this technology as a robust national asset that is deeply integrated into India’s digital framework before expanding globally. Recently, QNu Labs opened an office in the United States, underscoring its commitment to “Made in India, Made for the World”. Quantum in the AI Era There are four primary fields in quantum technology. One is cryptography, which QNu Labs focuses on; another is quantum sensing, which helps in detecting and understanding phenomena underground, underwater, in the air, and so forth. The third area that is gaining significant attention these days is quantum computing, which has applications in industries like pharmaceuticals and healthcare. Last, there is materials science, where new materials are emerging that enable these quantum technologies to become commercially viable, much like silicon does for electronics. The combination of quantum and AI creates a formidable defence against malicious actors, preserving the confidentiality and integrity of digital communications and data. “AI acts as an enabler across these quantum domains, especially in cryptography and secure communications. While AI can pose risks by potentially breaking traditional encryption standards, quantum technologies can protect against these threats,” Patel said. In terms of practical applications in the AI era, quantum technology is transforming areas like healthcare with the advent of personalised medicine. In security, quantum can help predict the nature of cyber-attacks, allowing for targeted defences. The combined impact of AI and quantum leads to faster innovation, focused solutions, and personalised benefits. For India, quantum technology isn’t just an aspiration – it’s a national mission. Led by Ajai Chowdhry, the co-founder of HCL and the founder of EPIC Foundation, India’s National Quantum Mission places quantum at the forefront of the country’s strategic priorities. With new leadership, the mission has gained momentum, focusing on structured execution and setting clear goals for leveraging quantum technology to benefit the nation. By January 2024, Chowdhry and his team had established the quantum mission’s framework, beginning with a request for proposals (RFP) to attract expertise and innovative ideas across the quantum spectrum. Speaking on the investment end, he said, “It was decided that INR 4,000 crore will be directly with the National Quantum Mission and INR 2,000 crore will be spent by others, which include the departments of space, defence, and atomic energy. So, it’s an INR 6,000-crore plan.” Patel noted that India aims to lead in quantum, much like it has done with fintech, setting ambitious goals for industry, academia, and research institutions to collaborate in driving forward this technology. What’s Next? With approximately 75% of global GDP now digital and still growing, the need for robust security increases. Encryption, acting as a digital lock and key, remains the primary method to secure digital data. However, encryption standards have traditionally been developed in Western countries, raising concerns about trust and surveillance. To ensure secure and resilient technology for the future, quantum technology is essential. Unlike conventional encryption, which relies on complex mathematics, quantum encryption leverages the unbreakable laws of physics, offering what’s known as absolute or unconditional security. For QNu Labs, this journey began with incubation at IIT Madras, then moved to Bengaluru, and was supported by the former principal scientific advisor to the Prime Minister. Building on this momentum, QNu has assembled experts to propel its innovation forward.","excerpt":"For India, quantum technology isn’t just an aspiration – it’s a national mission.","categories":["AI Features"],"tags":["AI (Artificial Intelligence)","India","quantum cryptography"],"author_name":"Vidyashree Srinivas","publish_date":"2024-11-15T12:03:37","publication_year":"2024","word_count":834,"keywords":["Go","API","AWS","AI","quantum cryptography","Git","RAG","Aim","Rust","GAN","R","India","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","Aim","RAG","AWS","R","Go","Rust","Git","API","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/what-quantum-cryptography-solutions-mean-for-indias-security\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10069721,"title":"LUX: Python API for Automated Exploratory Data Analysis","content":"Exploratory data analysis is the process of understanding the data thoroughly for key characteristics and understanding each feature’s importance statistically. As the name suggests exploratory data analysis helps in exploring the data statistically and formulating certain hypotheses as required. Generally, exploratory data analysis is considered a tedious task and this is where LUX is paving its way by automating the entire exploratory data analysis in one single step. So in this article let us see how to use LUX python API to perform exploratory data analysis. Table of Contents The need for Automating EDAAutomating EDA using LUXImplementation of LUX python APIFinal words Before starting with the LUX environment let’s first discuss the need for Automated EDA. The need for Automating EDA Exploratory data analysis is a process of analyzing the datasets to summarize the important statistical significance of features and visualize the spread of each feature through appropriate visuals. But visualizing each of the features is a tedious task when there are many features in the dataset as checking the correlation of each of the features would be a lengthy process. So this is where the process of automating EDA plays a vital role in reducing the overall time spent on data analysis and time spent on optimal feature selection and also on outlier analysis. Are you looking for a complete repository of Python libraries used in data science, check out here. This is where the need for Automating the exploratory data analysis process occurs and automation of exploratory data analysis is supported by various Python libraries and APIs like LUX, SweetViz, AutoViz, and many more come into play. In this article let us explore how to use the LUX Python API to automate the exploratory data analysis procedure. Automating EDA using LUX LUX is a simple python API that helps in quick and easy data exploration by providing easily interpretable plots by just reading the data frame in the LUX-activated working environment. Visualizations are produced in an interactive widget with various feature tabs to slide through and understand the characteristics of the data. Some of the standard widgets supported under the LUX module are as follows. CorrelationDistributionOccurrenceGeographical Correlation widget The correlation widget helps in analyzing the correlation between two numerical features of the data in the form of a scatter plot. All the numerical features will be mapped into sets of two features and the correlation between the two features can be visualized in order to analyze the features with a higher correlation. Distribution widget The distribution widget of the LUX python API is responsible for generating histogram visuals for all the numerical features providing the count of each of the features through histogram bins. The distribution widget mainly helps in analyzing the frequency of numerical features Occurrence widget The occurrence widget of LUX python API is responsible for generating horizontal bar plots by analyzing the frequency of occurrence of categorical features present in the data. For each of the categorical features and for each class of categorical features frequency occurrence is provided in the form of visuals under the occurrence widget. Geographical Widget The geographical widget of LUX API basically shows the choropleth maps for geographical locations in the dataset. The mean of certain numerical features is computed for each region on the maps and by just hovering on the map the mean value for each region can be computed under each of the geographical locations in the data. Implementation of LUX python API In this let us see how to use the LUX Python API to automate the exploratory data analysis process. For utilizing the LUX python API we have to first install the LUX API in the working environment. !pip install lux-api Now after installing the LUX API in the working environment let us import the API in the working environment along with the pandas module to read the dataset. import lux import pandas as pd In certain working environments, certain widgets for visualizing from APIs have to be permitted by installing the corresponding visualizing widgets. Here let us see how to permit the LUX API to produce visuals in Google Colab. from google.colab import output output.enable_custom_widget_manager() Once the widgets have been set the final step is just to read the dataset using the pandas module in the working environment. df=pd.read_csv('\/content\/drive\/MyDrive\/Colab notebooks\/EDA using LUX\/WA_Fn-UseC_-HR-Employee-Attrition.csv') df This is how by just reading the dataframe in the LUX activated working environment the entire exploratory data analysis process is automated and various widgets are produced. Correlation widget output interpretation In the above image let us consider the first plot for Monthly Income and Total Working Hours and we can see how these two features are correlated with each other. Distribution Widget Output Interpretation In the above image if we consider the first plot we can easily interpret the frequency occurrence of each of the numerical features present in the dataset. Occurrence Widget Output Interpretation In the above image if we consider the first plot we can clearly see that the feature Performance rating is having two categories we can correspondingly also analyze the frequency occurrence of each of the categories. Custom feature visualization using LUX Instead of visualizing the entire dataset LUX also has the flexibility to analyze the characteristics of required features as shown below. But for custom features selected there are new three widgets supported under LUX API there are Enhance, Filter and Generalize. df.intent = [\"YearsAtCompany\",\"HourlyRate\"] df So here two numerical features are selected from the data to understand its various characteristics using the LUX API. So if required features are selected from the dataset this is the visualization widget generated by the LUX API. Let us try to understand what each widget has to convey. Enhance widget output interpretation The enhance widget explains how additional features of the dataset affect the relationship of the two custom variables selected. In the above output, we can see how HourlyRate and Years at Company are related to other features of the dataset like StandardHours and Attrition. Filter widget output interpretation The filter widget considers the two custom features and produces correlation plots for a various subsets of features by analyzing each feature relation with respect to the custom features selected from the dataset. Generalize widget output interpretation The generalize widget considers only the custom features selected and removes if there any filter constraints in the features and shows a histogram distribution of the custom features selected for analysis. Analyzing Geographical data using LUX API For analyzing geographical data using LUX API a population dataset was used across various states. df = pd.read_csv(\"https:\/\/github.com\/covidvis\/covid19-vis\/blob\/master\/data\/interventionFootprintByState.csv?raw=True\",index_col=0) df.head() Later for obtaining the analysis through the LUX API framework the dataframe instance was just called in the working environment. df Interpreting the Geographical widget In the above image, we can see that for various states in the dataset the mean values for various numerical features were computed for various areas in the respective states. Just by hovering over the map, we can interpret the mean of the corresponding numerical feature for the respective areas in each state. Final words Automating exploratory data analysis helps in cutting down 60% of the work that goes into data cleaning and analysis. By automating exploratory data analysis, optimal feature selection and checking correlation among features becomes easy and due to this more time can be utilized to produce more generic and reliable models for the respective tasks using the data. Among various automated exploratory data analysis APIs python offers LUX is one such API where the entire analysis of the data is obtained by just reading the data in the LUX activated environment to generate suitable insights from the data. References LUX API Official documentationLUX API Official Github RepositoryColab Notebook","excerpt":"Do you want to automate data analysis in your projects? LUX is an API which yields efficient and a quick data analysis. Have a look into it.","categories":["AI Trends"],"tags":["data exploration","exploratory data analysis","Python"],"author_name":"Darshan M","publish_date":"2022-06-26T13:00:00","publication_year":"2022","word_count":1282,"keywords":["data science","Go","TPU","AI","data exploration","Git","Python","Colab","GitHub","exploratory data analysis","R","Pandas"],"extracted_tech_keywords":["AI","data science","Colab","Pandas","TPU","Python","R","Go","Git","GitHub"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/lux-python-api-for-automated-exploratory-data-analysis\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10166965,"title":"Vanguard Group Chooses Hyderabad for its First India GCC, To Hire 2,300 Employees","content":"Vanguard Group, one of the largest financial management MNCs, has announced the establishment of its Global Capability Centre (GCC) in Hyderabad on Monday, marking the company’s first facility in India. The centre will focus on recruiting top-tier engineering talent in areas such as Artificial Intelligence (AI), data analytics, and mobile engineering. Vanguard Group plans to scale the facility to over 2,300 employees in the coming years, significantly contributing to the region’s growing tech ecosystem. Telangana’s chief minister Revanth Reddy shared the news on X, saying, “Delighted to share with the people of #Telangana that @Vanguard_Group, one of the world’s most respected financial management MNCs, has chosen #Hyderabad for establishing its Global Capability Centre (GCC).” This development follows the recent inauguration of a new 1.2 million-square-foot engineering facility by Microsoft in Hyderabad, which will house 3,000 engineers. It highlights the rapid growth of Hyderabad as a major hub for global corporations in the tech and life sciences sectors. Meanwhile, Hyderabad has been rapidly establishing itself as a life sciences powerhouse, rivalling Bengaluru, owing to a state-supported business ecosystem that is attracting an increasing number of Global Capability Centres (GCCs). The city is seeing a surge in R&D and biotech investments, with companies like Meishi Pharma Services, ALS, and Miltenyi Biotec setting up advanced research and biopharma facilities. Additionally, Telangana’s Green Pharma City initiative has received a significant boost, with investments from 11 leading pharmaceutical companies, including Bharat Biotech, Biological E, and Sai Life Sciences, totaling ₹5,445 crore and generating 9,800 new jobs.","excerpt":"The centre will focus on recruiting top-tier engineering talent in areas such as Artificial Intelligence (AI), data analytics, and mobile engineering.","categories":["AI News"],"tags":["GCC india"],"author_name":"Shalini Mondal","publish_date":"2025-03-31T21:27:52","publication_year":"2025","word_count":251,"keywords":["API","artificial intelligence","programming_languages:R","AI","analytics","GCC india","GAN","R"],"extracted_tech_keywords":["AI","artificial intelligence","analytics","R","API","GAN","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/vanguard-group-chooses-hyderabad-for-its-first-india-gcc-to-hire-2300-employees\/","complexity_score":3,"technical_depth":7,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10095881,"title":"10 Best Vector Database for Building LLMs","content":"First and foremost, vector databases enable faster processing of large datasets. These databases are specifically designed to store and retrieve data efficiently, resulting in an accelerated processing time. By leveraging the power of vector representations, LLMs can quickly analyse and comprehend vast amounts of information, leading to improved efficiency and reduced processing time. Scalability is another crucial aspect facilitated by vector databases. These databases can seamlessly scale up or down based on the user’s requirements, making them capable of efficiently managing massive volumes of data without compromising performance. This scalability empowers LLMs to handle diverse and evolving datasets, ensuring their effectiveness in dynamic environments and accommodating the growing demands of users. The precise similarity matching capability offered by vector databases is essential for various applications, particularly in voice and image recognition. By representing audio and visual data as vectors, LLMs can accurately identify and match similar items, enabling highly accurate voice and image recognition functionalities. Additionally, vector databases enhance search capabilities through the utilisation of advanced search algorithms. With these databases, LLMs can provide more effective and relevant search results, enabling users to access the desired information efficiently. This improvement in search efficiency contributes to a more seamless and user-friendly experience for individuals interacting with LLM-based applications. Now that we know the importance and capabilities of vector database, here is a list of best vector database options for LLMs – MongoDB Firstly, MongoDB, the developer’s favourite database, has come up with Atlas Vector Search.  This NoSQL database has recently incorporated vector search capabilities, revolutionising the integration of generative AI and semantic search into applications. By combining the power of MongoDB with vector search, developers can unlock new possibilities in data analysis, recommendation systems, and natural language processing. With Atlas Vector Search, developers have the ability to conduct searches on unstructured data effortlessly. It enables them to generate vector embeddings using your preferred machine learning model, whether it’s OpenAI, Hugging Face, or others, and store them directly in Atlas. This powerful feature supports a wide range of use cases, including similarity search, recommendation engines, Q&A systems, dynamic personalization, and long-term memory for LLMs. DataStax DataStax had recently introduced AstraDB, a vector database designed to streamline app development processes, allowing developers to create applications faster and more efficiently. By integrating with AstraDB, which handles Cassandra operations, AstraDB frees developers from the complexities of database management, enabling them to focus on app creation. It simplifies every step of the development process by eliminating time-consuming configuration changes, allowing developers to dedicate their time to writing code that matters. Developers can improve app performance across any cloud environment without the need to scale up or down manually. It provides a seamless and scalable solution, ensuring that applications perform optimally without the hassle of performance optimization and cloud infrastructure management. AstraDB enables developers to accelerate the app development cycle, simplify workflows, and deliver high-performing applications efficiently. Milvus Milvus is a vector database system designed for efficient handling of complex data. It offers high speed and performance for data retrieval and analysis, making it ideal for applications that require quick insights. Milvus can handle massive datasets effectively, simplifying the storage and analysis of large volumes of data. It supports multiple vector data formats, including audio, text, and images, allowing flexibility in data representation. The comprehensive indexing capabilities of Milvus enable fast and accurate vector similarity searches, enhancing the precision of search results. It also enables real-time updates, ensuring the availability of the most recent data for analysis. Weaviate Weaviate is a powerful and user-friendly database that specialises in storing and searching high-dimensional vectors. It introduces semantic search, enabling users to find related objects based on meaning and context rather than just keywords. Weaviate supports real-time updates, keeping the database up-to-date with the latest changes. Its flexible schema allows easy adaptation to different data types and structures. Being an open-source solution, Weaviate offers visibility and customization options to meet specific needs. It provides personalised suggestions by analysing user queries, improving the user experience. Integration with deep learning frameworks makes it suitable for image or text categorization tasks, and its time series analysis capabilities make it effective for forecasting and anomaly detection projects. Pinecone Pinecone is a robust vector database known for its impressive speed, scalability, and support for complex data. It excels at fast and efficient data retrieval, making it ideal for applications that require quick access to vectors. Pinecone can handle large data volumes, making it suitable for big projects and enabling the detection of patterns and irregularities in large datasets. Real-time updates ensure that the database is continuously up-to-date. It is optimised for high-dimensional data types such as text, enhancing the understanding and search capabilities for complex data. Pinecone’s automatic indexing feature speeds up searches, enabling efficient similarity search for grouping and recommendations. Additionally, Pinecone provides capabilities for identifying unusual behaviour in time-series data, making it valuable for anomaly detection. RedisVector RedisVector is a vector database that focuses on efficient processing of vector data. It excels at storing and analysing large amounts of vector data, including tensors, matrices, and numerical arrays. By leveraging Redis, an in-memory data store, RedisVector delivers high-performance query response times. It offers built-in indexing and search capabilities, enabling quick searching and finding similar vectors. RedisVector supports various distance measures for comparing vectors and performing complex analytical operations. With its operations on vector data, including element-wise arithmetic and aggregation, RedisVector provides a versatile environment for working with vectors. It is particularly suited for machine learning applications that process and analyse high-dimensional vector data, enabling the creation of customised recommendation systems and accurate similarity-based search. SingleStore SingleStore is a scalable database that excels in data processing and high-performance analytics. It can handle large amounts of data by scaling horizontally across multiple nodes, ensuring high availability and scalability. SingleStore leverages in-memory technology for quick data processing and analysis. It enables real-time analytics, allowing users to interpret and analyse data in real-time, facilitating quick decision-making. The full SQL support of SingleStore enables easy interaction with the database using common SQL queries. It supports continuous data pipelines, facilitating smooth data intake from various sources. SingleStore also integrates with machine learning tools and libraries, enabling advanced analytics. Its efficient management of time series data makes it suitable for applications such as IoT, banking, and monitoring. Relevance AI Relevance AI is a comprehensive vector database designed for storing, searching, and analysing large amounts of data. It offers fast query response times, allowing users to retrieve insights from data quickly. With advanced algorithms, Relevance AI delivers precise and relevant search results. It supports various data types and formats, making it versatile for working with different datasets. Real-time search capabilities enable instant access to the desired information. Relevance AI is capable of handling both small and large amounts of data, making it suitable for a wide range of applications. By leveraging user preferences and historical data, it can create personalised experiences for users, enhancing engagement and satisfaction. Qdrant Qdrant is a versatile vector database solution that excels in effective data management and analysis. It offers advanced search techniques for finding similar objects in a dataset, enabling efficient retrieval of related items. Qdrant’s scalability allows it to handle increasing amounts of data without compromising performance. It supports real-time updates and indexing, ensuring that the database remains up-to-date and searchable. With various query options, including filters, aggregations, and sorting, Qdrant provides flexibility in data exploration. It is particularly useful for similarity-based suggestions, anomaly detection, and image\/text search applications. Vespa.ai Vespa.ai is a vector database known for its quick query results and real-time analytics capabilities. By integrating ML algorithms, Vespa.ai enables advanced data analysis and predictive modelling. The high data availability and fault tolerance of Vespa.ai ensure continuous service and minimal downtime. Customisable ranking options allow organisations to prioritise and obtain the most relevant data. Vespa.ai supports geospatial search, enabling location-based searches for spatial applications. It is particularly suitable for media and content-driven applications, providing targeted ads and real-time statistics for improved audience targeting.","excerpt":"Have you heard about the importance of vector databases for building LLMs?","categories":["AI Trends"],"tags":["Top Trend","Vector Database"],"author_name":"Mohit Pandey","publish_date":"2023-06-28T16:31:02","publication_year":"2023","word_count":1330,"keywords":["Top Trend","Hugging Face","machine learning","OpenAI","AI","ML","RAG","Vector Database","Ray","deep learning","analytics","generative AI"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","analytics","generative AI","OpenAI","Ray","Hugging Face","RAG"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/10-best-vector-database-for-building-llms\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10067460,"title":"Lenskart invests USD 2 Mn in location intelligence platform GeoIQ","content":"Lenskart has invested USD 2 million in the pre-Series A of GeoIQ, an AI-based local intelligence platform. The round also saw participation from existing investors, 9 Unicorns, Venture Catalysts and Ecosystem Ventures. Founded in 2018, Bengaluru-based GeoIQ is a geo-analytics company that provides real-time hyper location data for enabling strategic decision making across industries. The startup aspires to be the go-to platform for predicting hyper-local business performance across geographies with a mission to quantify each location’s risk and opportunities for businesses to operate efficiently. “Lenskart has always been at the helm of leveraging technology to enhance its offerings for better customer experience and drive differentiation, and we have continually partnered with founders in doing that. With GeoIQ technology, Lenskart will be opening the most efficient omnichannel stores globally with a 50% better return on invested capital and 5x more opening speed. Through this funding, we also want to help facilitate GeoIQ’s ambition of becoming a world-class geo-analytics company in the long run by providing more opportunities to explore, experiment and innovate when it comes to CX,” said Peyush Bansal, co-founder and CEO, Lenskart. GeoIQ’s AI-based location tool will help Lenskart with its aggressive store rollout strategy by identifying new locations to open high ROI potential stores. It will also gather accurate market share data across micro markets and run more targeted regional marketing campaigns. “The use of analytics has become widespread as more and more businesses recognise the benefits of data. AI-based intelligence has evolved rapidly and is currently in a hyper-adoption stage. The funds raised will enable us to add new features to our product and empower our customers,” said Devashish Fuloria, CEO, GeoIQ. Over the next 12 months, GeoIQ aims to empower 1000 data teams across India and the US with its no-code location AI and targets a 10x increase in its API usage to enable 3 million decisions a day. The startup also plans to improve its product offering and expand its workforce for product and sales led roles.","excerpt":"GeoIQ’s AI-based location tool will help Lenskart with its aggressive store rollout strategy.","categories":["AI News"],"tags":["investors","unicorns"],"author_name":"Sri Krishna","publish_date":"2022-05-19T15:32:46","publication_year":"2022","word_count":332,"keywords":["Go","API","funding","unicorn","AI","unicorns","investors","RAG","Aim","analytics","R","startup"],"extracted_tech_keywords":["AI","analytics","Aim","RAG","R","Go","API","startup","unicorn","funding"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/lenskart-invests-usd-2-mn-in-location-intelligence-platform-geoiq\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10077393,"title":"What to Expect from the Data Centre and Cloud Policy?","content":"The data centre ecosystem is booming in India, and hence the ‘National Data Centre and Cloud Policy’ is the need of the hour. So far, the Government of India (GoI) has had multiple discussions with stakeholders and is expected to bring forth the policy in the coming months. “Along with this policy, the government is also expected to bring out a new ‘Digital India’ law with the aim of achieving a 20-fold boost. It is expected to include usage of indigenous platforms in the data centre and cloud operations along with improvements in cloud space security and the interface between the clouds.” “Enhanced coordination with the department of telecommunications to facilitate robust and cost-effective connectivity to the centres and recognition of data centres in a separate category under the National Building Code is also expected. We are not sure whether they will introduce any new regulatory body or will strengthen the existing one,” Dr. Ramanand, Director at the Centre of Policy Research and Governance (CPRG), told Analytics India Magazine. India has over 138 data centres with an overall spanning of 11 million square feet that has 737 MW of IT (building) capacity. The Indian data centre industry is currently worth USD5.6 billion, with an additional 45 new data centres expected to come up by 2025. Today, states such as Maharashtra, Karnataka and Tamil Nadu already have state-specific data centre policies in place which provide many fiscal and non-fiscal benefits. “Based on the draft policy and these existing state data centre policies, we could expect fiscal benefits in the form of electricity duty exemption\/subsidies, power subsidies\/ incentives, tax benefits and land at a subsidised cost,” Raya Hazarika, Partner at S&R Associates, told Analytics India Magazine. Let’s look at some of the key things that we think will be part of the Data Centre and Cloud Policy. Single window clearance Clearances from the government remain one of the major hurdles when it comes to setting up a data centre in India. Although, the Draft Data Centre Policy 2020 did mention the creation of a single window clearance to deal with the issue, for a company to set up a data centre in the country today nearly 40 clearances from various government agencies are required. “To operationalise a data centre, you need a number of permits and approvals. So, it requires approaching different authorities to get separate approvals, which makes it difficult to obtain those in a time-bound manner,” said Raya Hazarika. Hazarika further explained that, “The draft policy did mention faster clearances, a single window clearance. So, I think that’s one of the key things that should definitely be a part of the final policy. If that’s there, I would think that it should factor in two things. Firstly, processing of applications in a time-bound manner. And secondly, I think they should also deal with some coordination with the various agencies and departments involved to get clearances from them. So that it’s single window clearance in the true sense of the term.” Green energy “The Data Centre and Cloud Policy can be expected to boost the growth of data centres and electronics manufacturing and also incentivise green energy sources,” Dr. Ramanand, Director at the Centre of Policy Research and Governance (CPRG), said. Globally, data centres account for nearly 4% of the total carbon emissions. Hence, the government is expected to provide an incentive of 3% for the use of renewable energy. “I think that there will be incentives when it comes to the use of renewable energy. Also, I would expect some eligibility criteria for benefits when it comes to these data centres which incentivize use of renewable energy,” Raya Hazarika said. Besides, data centres constantly require massive amounts of energy—making them energy hogs. “Data centres consume a vast amount of energy, often in a wasteful manner. Nearly 90% of the electricity they pull off the grid is wasted. It is necessary for data centres to bring down energy consumption to acceptable levels,” Dr. Ramanand, Director, Centre of Policy Research and Governance (CPRG), said. So, the policy is also expected to have provisions for the efficient use of energy. “Encourage efficient utilisation of energy by promoting innovative techniques and solutions for energy management for reducing the carbon footprint of the Data Centres,” the draft policy stated. While the aim is to make the data centres green, the process is time-intensive. Meanwhile, they are expected to continue using electricity for their purposes. The government is also planning to provide electricity to the data centres at a subsidised rate. Fiscal benefits To boost the data centre ecosystem in India, the government is planning to invest around INR3 lakh crore in the next five years. This will possibly be the largest investment in terms of incentives, even greater than the investment set aside for semiconductors. The government is planning to provide an incentive of around 4–6% on investment if the data centres use locally manufactured components such as servers, server trays, server housing units and non-IT equipment such as mechanical, electrical, plumbing, cooling equipment and others. Further, the government has provided the data centre sector with the ‘Infrastructure Status’, which will establish them at par with other sectors like Railways, Roadways and Power. This will further allow the data centre companies to avail long-term credit from domestic and international lenders at relaxed terms. Data Centre investment in India (Source: Statista) Data Localisation Dr. Ramanand believes that there will be provisions in the draft which will talk about the safety and security of data and data localisation. According to McKinsey, “Today, 75% of all countries have implemented some level of data localisation rules. These have major implications for the IT footprints, data governance, and data architectures of companies, as well as their interactions with local regulators.” Hence, the policy is likely to have provisions that could ask the data centre companies to store data generated in India within the country itself. Minister of Finance Nirmala Sitharaman has also talked about data localisation multiple times in the last few years.","excerpt":"A single window clearance is one of the key things that should definitely be a part of the draft.","categories":["AI Features"],"tags":[],"author_name":"Pritam Bordoloi","publish_date":"2022-10-17T14:00:00","publication_year":"2022","word_count":1003,"keywords":["Go","AI","Git","RAG","Ray","Aim","ViT","analytics","data governance","R"],"extracted_tech_keywords":["AI","analytics","Aim","Ray","RAG","R","Go","Git","data governance","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/what-to-expect-from-the-data-centre-and-cloud-policy\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10052292,"title":"PayPal Releases Its Open-Source Spark Indexing Library “Dione”","content":"PayPal’s global data science group recently released its open-source Spark Indexing library named Dione, a library that would enable faster interaction with Hadoop Data. From time to time, users want to use the same data for more ad-hoc oriented tasks in ecosystems such as Spark, Hive and HDFS (Hadoop Distributed File Systems). Tasks such as multi-row load or single-row fetch are traditionally solved using dedicated storage and technology stacks (HBase, Cassandra, etc.), requiring data duplication and significant operational costs. Dione can be used to solve such challenges using the index as a “shadow” table of the original data. It contains only the key columns and pointers to the data and is saved in a special format inspired by Avro and bucketing. Based on this index, the library provides APIs to join, query, and fetch back the original data in the required SLAs. A few advantages of using Dione are: Relies only on Spark, Hive and HDFS. No external services.Semi-managed — does not modify, duplicate or move the original data.Supports multiple indices on the same data.The index is exposed as a standard Hive table.The special Avro B-Tree format supports ad-hoc single-row fetch in an SLA of seconds. Image: Paypal Engineering Dione solves two main challenges that are currently faced using traditional methods: Given a pointer (a line in the index), how to fetch the data quickly?How to store the index table to meet the required SLAs? The library makes use of its two main architectural components — Indexer and AvroBtreeFormat. The final index solution is composed of the inter-relations between these components, although in principle, each of them can stand on its own. The Indexer’s main goal is to solve the multi-row load task. The single-row fetch task as using Spark to scan through all the index tables does not meet the required SLA. For that, the index’s storage format is used. Inspired by Avro’s SortedKeyValueFile, bucketing and traditional databases indexing systems, the development team decided to create a “new” file format — Avro B-Tree. Both the Indexer and Avro B-Tree File Format libraries are independent packages and rely only on HDFS. Users can save any table in the Avro B-Tree format so it will be accessible for both batch analytics with Spark and single-row fetches. For a simplified Spark user experience, a high-level Spark API has been added for creating and using an index. The API is available in Scala and Python. PayPal has open-sourced this library to share this functionality with the community and get feedback.","excerpt":"Dione would enable faster interaction with Hadoop Data and can be used to solve challenges using the index as a “shadow” table of the original data.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","Data Analytics","Data Science","Data Scientist","hadoop etl","Machine Learning","PayPal","Python","spark"],"author_name":"Victor Dey","publish_date":"2021-10-25T12:33:58","publication_year":"2021","word_count":417,"keywords":["data science","Go","API","PayPal","hadoop etl","AI","spark","Machine Learning","Scala","RAG","Python","programming_languages:Python","analytics","Data Analytics","Data Science","Data Scientist","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","data science","analytics","RAG","Python","R","Go","Scala","API","programming_languages:Python"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/paypal-releases-its-open-source-spark-indexing-library-dione\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10089423,"title":"What is MedPaLM 2?","content":"Google at its annual event called “The Check Up”, announced the latest version of its medical large language model called MedPaLM — along with new health initiatives and partnerships. According to the MedPaLM 2 team, their model achieved a score of 85% on medical exam questions (USMLE MedQA), which is comparable to the level of an “expert” doctor. This is an improvement of 18% from the previous performance of Med-PaLM, surpassing similar AI models — the likes of GPT-4 and others. The team also obtained results on other benchmarks such as MedMCQA and MMLU clinical topics. The evaluators, consisting of clinicians and non-clinicians from diverse backgrounds and countries, tested the models against 14 criteria which included factors such as scientific accuracy, exactness, conformity with medical consensus, logical thinking, partiality, and potential for harm. Google identified significant disparities, however, it pledged to collaborate with researchers and healthcare professionals to narrow these disparities and enhance healthcare services. Google Research and DeepMind had released the initial model called MedPaLM, in December 2022. MedPaLM was evaluated using a new open-source medical question-answering benchmark called MultiMedQA. The AI system had achieved a passing score of over 60% on multiple-choice style questions, which are similar to those used in U.S. medical licensing exams. This was the first time that such a system had been able to do so successfully. The researchers utilised PaLM, which is a large language model with 540 billion parameters, and its instruction-tuned variation called Flan-PaLM to create the model. They employed these models to evaluate other large language models using MultiMedQA. In an interesting new development, Google also launched PaLM API right before OpenAI’s GPT-4. The latest API now permits businesses and developers to construct applications utilising Google’s SOTA large language model, which is identical to the one employed in Search, YouTube, and Gmail. Google is offering access to its underlying models for the first time.","excerpt":"Google at its annual event called “The Check Up”, announced the latest version of its medical large language model called MedPaLM — along with new health initiatives and partnerships. According to the MedPaLM 2 team, their model achieved a score of 85% on medical exam questions (USMLE MedQA), which is comparable to the level of […]","categories":["AI News"],"tags":["GPT4","Healthcare Automation","Microsoft"],"author_name":"Shyam Nandan Upadhyay","publish_date":"2023-03-15T15:21:04","publication_year":"2023","word_count":314,"keywords":["Go","API","OpenAI","AI","RPA","ML","llm_models:PaLM","Healthcare Automation","GPT","llm_models:GPT","R","Microsoft","GPT4"],"extracted_tech_keywords":["AI","ML","OpenAI","R","Go","API","GPT","RPA","llm_models:GPT","llm_models:PaLM"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/google-introduces-medpalm-2-a-gpt-4-like-model-for-healthcare\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10008671,"title":"Facebook Is Giving Away This Speech Recognition Model For Free","content":"Researchers at Facebook AI recently introduced and open-sourced a new framework for self-supervised learning of representations from raw audio data known as wav2vec 2.0. The company claims that this framework can enable automatic speech recognition models with just 10 minutes of transcribed speech data. Neural network models have gained much traction over the last few years due to its applications across various sectors. The models work with the help of vast quantities of labelled training data. However, most of the time, it is challenging to gather labelled data than unlabelled data. The current speech recognition systems require thousands of hours of transcribed speech to reach acceptable performance. There are around 7,000 languages in the world and many more dialects. It can be said that the availability of the transcribed speech for a vast majority of languages is still negative. To mitigate such issues, researchers open-sourced the wave2vec framework. The framework has the capability to make efficient development in Automatic Speech Recognition (ASR) for the low-resource languages. Facebook AI is releasing code and models for wav2vec 2.0, a self-supervised algorithm that enables automatic speech recognition models with just 10 minutes of transcribed speech data. https:\/\/t.co\/Bp7Zc9zedh— AI at Meta (@AIatMeta) September 24, 2020 How wav2vec 2.0 Works The successor of wav2vec model, wav2vec 2.0 model learns basic speech units that are used to tackle a self-supervised task and is trained to predict the correct speech unit for masked parts of the audio while learning the speech units at the same time. wav2vec 2.0 utilises a self-supervision method to push the boundaries by learning from unlabelled training data to enable speech recognition systems for many more languages, dialects, and domains. In technical terms, wav2vec 2.0 masks the speech input in the latent space and solves a contrastive task defined over a quantisation of the latent representations which are jointly learned. Wav2vec 2.0 & Other Models Similar to masked language modelling, this framework encodes the speech audio via a multi-layer convolutional neural network and then masks spans of the resulting latent speech representations. The latent representations are then fed to a Transformer network to build contextualised representations, and the model is trained via a contrastive task where the true latent is to be distinguished from distractors. Also, similar to the famous BERT (Bidirectional Encoder Representations from Transformers) model, the new wav2vec 2.0 model is trained by predicting speech units for masked parts of the audio. One major drawback in BERT is that speech audio is a continuous signal that captures many aspects of the recording with no precise segmentation into words or other units. Wav2vec 2.0 tackles this issue by learning basic units that are 25ms long to enable learning of high-level contextualised representations. These units are then used to describe many different speech audio recordings and make wav2vec more robust. This feature helped the researchers to build speech recognition systems that can outperform the best-semi supervised methods, even with 100x less labelled training data. Wrapping Up According to a blog post, with just 10 minutes of transcribed speech along with 53K hours of unlabelled speech, this new model enables speech recognition models at a word error rate (WER) of 8.6 per cent on noisy speech and 5.2 per cent on clean speech on the standard LibriSpeech benchmark. In this research, the researchers showed that speech recognition models can be built with very small amounts of annotated data at very good accuracy. According to the researchers, this model has opened the door for speech recognition models in many more languages, dialects and domains that previously required loads of transcribed audio data to provide acceptable accuracy. Developers in a blog post stated, “Wav2vec 2.0 is part of our vision for machine learning models that rely less on labelled data, thanks to self-supervised learning.” They added, “We hope that the algorithm will enable improved speech technology for many more languages, dialects, and domains, and lead to improvements for existing systems.” The code and pre-trained models are made available by the researchers at GitHub. Read the paper here.","excerpt":"Researchers at Facebook AI recently introduced and open-sourced a new framework for self-supervised learning of representations from raw audio data known as wav2vec 2.0. The company claims that this framework can enable automatic speech recognition models with just 10 minutes of transcribed speech data. Neural network models have gained much traction over the last few […]","categories":["Global Tech"],"tags":["Facebook AI","Facebook AI research","Speech Analytics","speech recognition algorithm"],"author_name":"Ambika Choudhury","publish_date":"2020-09-30T17:00:29","publication_year":"2020","word_count":671,"keywords":["Go","machine learning","Speech Analytics","Facebook AI","AI","neural network","speech recognition algorithm","Transformers","Git","BERT","Aim","Facebook AI research","GitHub","R"],"extracted_tech_keywords":["AI","machine learning","neural network","Aim","Transformers","R","Go","Git","GitHub","BERT"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/facebook-is-giving-away-this-speech-recognition-model-for-free\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":35134,"title":"Understanding The Role Of Reward Functions In Reinforcement Learning","content":"With recent AI developments, there has been considerable research about its possible influence on human work. Research scientists have been trying to foretell the industries and jobs that will be affected. People also want to know what professions will be most in demand. Recently, there has been a new technology under development wherein developers are trying to replace predict elements with a much more effective and efficient way. This is known as a reward function that will allow AI platforms to come to conclusions instead of arriving at a prediction. Reward Functions are used for reinforcement learning models. Reward Function Engineering determines the rewards for actions. Why Reward Functions The AI advanced predictive analysis is really a game changer but not the game-winner. In the prediction model, the system is just using the data that is being generated by the user and creates specific desired data analysis that the user doesn’t have. This is often done by translating huge amounts of data into small fragments that are of a manageable level. Prediction can be useful but it isn’t the only input for decision making, the other fundamental input is judgement. For example, in the online banking system, the banking network decides whether or not to pass each transaction. The system tries to allow genuine transactions and decline hoax transactions. The industry uses AI to predict whether every attempted transaction is deceptive or not. In this sort of scenarios, if the prediction algorithm is accurate, the networks decision process is effective and easy. At the same time, inaccurate algorithms can cause fraud events. Even the best AIs can commit errors. The system needs to accept an action with minimal errors and the systems can achieve this only with real-time examples and analytical thinking which a human can and a system can’t. This process of decision making is known as judgement that we provide in reward functions. Judgement As Reward The question arises that can AI calculates outcomes and if so, yes, should there be a program that AI has to execute to come up with appropriate measure. This gave birth to a new form of technology with different inputs. Assessment requires to determine what action is to be taken to minimise loss and maximise benefits. In many cases, systems are required to exercise a deep understanding of the situation and analyse the different choices that they have and then combine the conclusion with machine-generated predictions to make determinations. Game Winner Just like humans, AI can also learn from experience. The reinforcement learning technique of AI is a crucial one that is capable of training the system to take actions with adequate reward information. This has been proven in the case of DeepMind’s AlphaGo for playing a game like Go. But there were instances when AI researchers failed training systems in real-time games like boat racing. The failure occurred because of the ingenuity of AI. The pivotal point that was learnt from numerous experiences is that most applications, the objective is given to AI differed from the real world and hard to measure. These problems are addressed by the reward functioning engineering. The technique defines the rewards (experiences, judgements) to various operations. The technology requires high analyzing of the requirements of the industries and the abilities of the machine. The process may involve programming the rewards in progress of the predictions so that the operations can be automated. Self-driving vehicles are an example of such models. With better predictive algorithms and different reward functional values, the AI technology can be applied in very critical sectors. Opinion It’s very early to say, whether machine prediction increases or decreases the workload of humans in decision-making. The machine may substitute for human prediction in decision making. The technology has to still evolve, so there will be numerous opportunities for research and career in this field","excerpt":"With recent AI developments, there has been considerable research about its possible influence on human work. Research scientists have been trying to foretell the industries and jobs that will be affected. People also want to know what professions will be most in demand. Recently, there has been a new technology under development wherein developers are […]","categories":["AI Features"],"tags":["reinforcement learning models","reward functions"],"author_name":"Bharat Adibhatla","publish_date":"2019-02-19T07:42:53","publication_year":"2019","word_count":640,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","reward functions","RAG","reinforcement learning models","AI research","R"],"extracted_tech_keywords":["AI","RAG","R","Go","AI research","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/understanding-the-role-of-reward-functions-in-reinforcement-learning\/","complexity_score":3,"technical_depth":7,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":38001,"title":"The Amazing Ways Routematic’s AI-Based Solution Paves The Way For Smarter Commute","content":"One of the most-cited uses for AI is in the traffic space, with route-finding algorithms being able to work with each other to ensure that congestion is reduced across the city. There are also apps that promote carpooling to enable a smaller carbon footprint overall and create a greener city. The adoption of these apps comes mostly from corporate, who have a unique requirement of having to move people regardless of time. This is where solutions such as Routematic come in, providing an AI-based solution for a problem that has long been faced by almost every working individual. To know more about the operations  of Routematic, we spoke to Surajit Das, the CEO of Routematic. Growth From SaaS To TaaS Routematic began as a SaaS route optimisation software service provider that uses the details of routes taken by the driven and corroborates it with location updates. These solutions utilise algorithms to implement maximum efficiency to ensure that every vehicle is fully occupied before the destination. They utilise an app that is connected to the cloud, thus providing information such as pick-up and drop time, along with the best route they can take based on traffic conditions. This helped corporates cut down on transport costs by up to 40% and reduced their travel time by up to 18 minutes. They operate in 6 cities, with over 30 corporate customers and 100,000 users. Their app undertakes over 25 million kilometres of trips monthly, with their newest venture being Routematic Fleet Services. Their fleet services are cited as being the first in the industry, with verified driver and compliant vehicles being one of their biggest selling points. Added to this, the cars also come outfitted with GPS devices which will be used for Routematic’s innovating Point to Point GPS KM based billing system. Each car will also come with a driver trip sheet phone to keep an additional set of data at the ready. According to Das, the background of Routematic as a SaaS company has allowed them to gather sufficient data for operations in transport management services. He stated that the data was useful to the point that they have “launched [their] own ‘city level commute’ cab services, that sits completely on top of the ‘city level commute’ predictive fleet dispatch model.” Solving A Unique Problem With A Data-First Approach Das revealed to us the unique requirements that come with adopting the traffic problem. This includes specific requirements such as those assigned to the ‘first pickup point’ in a route, which has requirements that would differ from other pickup points. This leaves Routematic without a single number or objective function to determine the ride, with other factors such as timing and coordination with pooling passengers also influencing the ride as a whole. To combat this, the company has created  the Routematic employee transport automation software, which processed over 200 million data points of GPS information every day. The progress of data transparency will enable a higher degree of traffic planning, said Das. He went on to say, “Regional RTO’s have started sharing real-time traffic data through their portals…[This] would immensely help with traffic planning which would directly impact the commute market.” The Brains Behind Routing Cars Das also stated, “Routematic uses a combination of big-data analytics and ML\/AI in its core engines such as route planning, travel-time prediction, predictive fleet dispatch, delay prediction and many more with regard to employee transport management service.” He went on to state that the commute problem is a “hard combinatorial problems in computing”, referring to it as a “multiple traveling salesman with knapsack optimisation problem”. The software is fundamentally a big-data engine that indexes travel times between selected “guide points”. These points are strategically placed around the city, with an ML regression algorithm functioning along side it to maximize the possibility of an On Time Arrival. At the same time, the algorithm also functions to reduce levels of dissatisfaction in the riders, which is determined by a proprietary mix of parameters such as expected wait time. The algorithm also utilizes the continuous feedback given by passengers through the companion application. Das stated that the data is analyzed in a highly granular fashion at 5 minute slots, resulting in a high precision output. According to the CEO, this makes Routematic an “an autonomous “decision making system” that aims to run completely autonomously without any human involvement”. This sets the company apart in two ways. Primarily, the software has been built from the ground up on the cloud, allowing them to integrate all flavours of GPS and Android devices on one IoT platform. Secondly, the algorithm driven by this data is not a “decision support system”, but instead an autonomous system for managing all the services in one place. How The Algorithms Work To Function Autonomously We asked Das for a more technical explanation of their product, considering how Routematic has been famed for their routing algorithms. In his own words: The goal of a transport “tracking tool” has traditionally been to ease the work of the router by acting as a decision support system in the routing process. Tools were compared based on “how quickly can a shift routing be closed through human intervention” and not on their capability of “routing” per se. So the routing process was merely thought of as “arrangement of system generated (not so good) routes on a user interface” that can be intuitively and easily manipulated by humans to achieve desired final outputs. So, these tools aimed to achieve “closer to final output” by leveraging artificial metrics such as distance deviation, travel time deviation and so on, so that the system generated route is amenable to “editing” quickly by a human. A slightly more advanced approach is to perform a historic search on pre-routed and approved shifts to check for the closest shift that matches the current shift to be routed, so that we can leverage the pre-routed output. However, all of these approaches are useless in “dynamic routing” scenario that is typical of organisations, and adopting any of these may result in heavy cost leakage. Routematic’s ML\/AI and operations research driven radically new approach aims to eliminate all human intervention in the routing process by keeping the overall approach completely dynamic, super-quick and delivering the most probably cost optimal routing output. By ingesting historic routing generated and approved by humans in the past (45-90 days preferable), the advanced ML model automatically infers various complex routing rules and “policies” inherent in any organisation’s routing output (and removing “outliers” in that process) and starts generating routes that are at the same level of optimality or better that historic human routing with all inferred policies being adhered to. The techniques used by Routematic for this are inspired from reinforced learning techniques in AI, notably “policy gradients”\/ DQN’s. Conclusion Routematic stands apart from other TaaS companies by not only providing a complete solution for corporate commute, but by using AI and ML focused approaches with a cloud and mobile-first approach. It’s heavy use of data points and prior experience in the corporate commute transport field also give it an advantage over competitors, allowing it to sell a unique product in its vertical.","excerpt":"One of the most-cited uses for AI is in the traffic space, with route-finding algorithms being able to work with each other to ensure that congestion is reduced across the city. There are also apps that promote carpooling to enable a smaller carbon footprint overall and create a greener city. The adoption of these apps […]","categories":["Deep Tech"],"tags":["ML","policy gradient"],"author_name":"Anirudh VK","publish_date":"2019-04-22T04:54:49","publication_year":"2019","word_count":1197,"keywords":["Go","TPU","AI","ML","RAG","automation","Aim","policy gradient","analytics","GAN","R"],"extracted_tech_keywords":["AI","ML","analytics","Aim","RAG","TPU","R","Go","GAN","automation"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/routematics-ml-powered-routing-paves-the-way-for-smarter-corporate-commute\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10057482,"title":"A historical tale of DeepMind’s games","content":"DeepMind has been creating games since the Alphabet company was just a start-up. However, the company’s continued focus on developing game players has cerated historical breakthroughs in less than a decade—starting with Atari that has become groundbreaking in off-policy deep reinforcement learning to the original AlphaGo that compressed several decades of game playing into a few years and was followed by continuous developments. In fact, DeepMind’s games were planned by one of the founders, Demis Hassabis, before the company was even created. This article will take you through the fascinating history of DeepMind’s breakthrough in games. A visual recap of DeepMind’s games through the years Atari Games: playing in the arcade environment In 2013, DeepMind’s first algorithm was tested on Atari 2600 games. The researchers selected the Arcade Learning Environment to test the algorithm’s competency across various games in an environment as challenging as human players’ choice. The initial algorithm learned to play seven games while achieving an average human’s performance on three of those. In 2015, DeepMind refined the algorithm to test it on Atari’s suite of 49 games, and the machine beat human performance on 23 of them. A persisting challenge still was succeeding in Atari’s four major games, Montezuma’s Revenge, Pitfall, Solaris and Skiing, which are particularly proven to be tough for AIs, given it will have to try several strategies and best moves for payoff. Only in the previous year did the algorithm manage to achieve this. AlphaGo: the first AI to defeat a professional human Go player In 2015, DeepMind released AlphaGo, which the company claims is “the first computer program to defeat a professional human Go player, the first to defeat a Go world champion, and is arguably the strongest Go player in history.” The oldest game, the Chinese board game Go, is considered more complex than chess due to the 10 by 170 possible configurations for success in the game. AlphaGo is built on a computer program combining an advanced search tree with a deep neural network. To win, the neural network inputs a description of the board game and processes it through the millions of neuron-like connections in its various network layers. DeepMind disrupted the gaming sphere of ‘amateur’ computer players with AlphaGo being the first AI to defeat a professional Go player, Mr Fan Hui, three-time reigning European Champion by a score of 5-0. It has since beat the world’s greatest player of the previous decade, Mr Lee Sedol, by a 4-1 victory with 200 million worldwide watchers sitting at the edge of their seats. AlphaGo Zero: self-training Go computer player In 2017, DeepMind released an updated version of AlphaGo, AlphaGo Zero. While AlphaGo was trained by playing thousands of matches with levels of players, AlphaGo Zero learnt by playing against itself. In just a few days, the computer program garnered years of human knowledge and learned to play Go from AlphaGo. The newer version surpassed the performance of all previous versions and taught itself new unconventional strategies and moves. AlphaGo Zero beat more Go world champions like Lee Sedol and Ke Jie. AlphaZero: self-training Chess, Go and Shogi player In late 2017, taking AlphaGo Zero on a macro level, DeepMind introduced Alpha Zero, an AI that can teach itself mastering games of Chess, Shogi and Go, from scratch. The system also beat the existing world champion computer programs in all cases. This is due to its deep neural network technology that goes beyond handcrafted features and only inputs the game’s basic rules. By playing over and over with itself, AlphaZero developed unique and creative strategies to win all three games. AlphaStar: StarCraft II player In 2019, DeepMind introduced AlphaStar, an AI program that can play the real-time strategy game StarCraft II. It is the first AI to reach the top league of the game, challenging the world’s two principal players and ranking above 99.8 per cent of the active players on Battle.net. The program is built on neural networks, reinforcement learning self-play, multi-agent learning and imitation learning to allow the AI to learn directly from the game’s data. AlphaStar battled with the three agents on the game, Protoss, Terran, and Zerg, through a single neural network and achieved grandmaster level for all three. AlphaStar’s knowledge was equivalent to 200 years of playing time during the introduction stage. MuZero: AlphaGo+Atari player In 2019, the latest addition in the AlphaGo programming, MuZero, was introduced, taking the technology one step further. The AI matches AlphaZero on Go, Chess and Shogi while mastering an array of the Atari games – without any input of the game rules. Instead, the program learns through a model of the environment and applies the information to AlphaZero’s lookahead tree search. As a result, MuZero can plan winning strategies even in unknown domains, making it another of DeepMind’s inventions to pioneer reinforcement learning algorithms towards AGI. Agent 57: player of 57 Atari games In 2020, DeepMind released an updated version of the initial Atari2600 games that can finally beat the four most challenging games of the suite. According to their paper, Agent57 is the first deep reinforcement learning agent to outperform humans on all 57 Atari 2600 games in the Arcade Learning Environment data set. Agent57 is an amalgamation of all the improvements in DeepMind’s Deep-Q network since the Atari games back in 2012. It also consists of a form of memory that allows it to base decisions on prior learning from games and a reward system to encourage the AI to explore more strategies. Player of Games: perfect and imperfect game player In 2021, DeepMind’s most recent addition is Player of Games (PoG), which performs well in both perfect and imperfect information games. The AI’s array of games extends beyond Chess and Go to Poker and Scotland Yard. PoG works on a single algorithm with minimal domain-specific knowledge. This marks a significant improvement in DeepMind’s cumulative step-ups in the games their AIs can play. For example, AlphaZero could only play perfect games, but PoG can understand imperfect informational games like Poker that rely on game-theoretic reasoning to hide private information properly. PoG’s search potential is suited across fundamentally different game types, with DeepMind’s guarantee that it will find an approximate Nash equilibrium by resolving subgames to remain consistent during online play. PoG uses growing-tree counterfactual regret minimisation (GT-CFR) to build subgames non-uniformly and expand the tree toward the most relevant future states while iteratively refining values and policies. It also uses self-play that trains value-and-policy networks using both game outcomes and recursive sub-searches applied to situations that came up in previous searches. It is important to note that, for all of DeepMind’s success, these AI models are not realistically versatile. They tend to be good at one thing and one thing only. The biggest challenge towards AGI is training AI at more than one task, and while models like Agent57 can learn 57 tasks, they can only learn and play one model at a time. Despite having the same algorithm, the program will need to retrain each game. However, DeepMind’s player games mark some of the first times an algorithm has reached the top levels in games, crafted unique strategies, or defeated the best players.","excerpt":"It has since beat the world’s greatest player of the previous decade, Mr Lee Sedol, by a 4-1 victory with 200 million worldwide watchers sitting at the edge of their seats.","categories":["AI Features"],"tags":[],"author_name":"Avi Gopani","publish_date":"2022-01-01T13:00:00","publication_year":"2022","word_count":1197,"keywords":["Go","AI","neural network","RPA","ML","RAG","Ray","Aim","R","T5"],"extracted_tech_keywords":["AI","ML","neural network","Aim","Ray","RAG","R","Go","T5","RPA"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/a-historical-tale-of-deepminds-games\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":46427,"title":"5 Legit Ways To Make Money As A Data Science Influencer","content":"Over the past few years, the ‘influencers’ from every domain have witnessed significant growth. What is more exciting to see that in this millennial era, even those who are not into the lifestyle sector, but into areas like data science domain are making a great impact on the internet. And why not, data science, after all, is “the sexiest job of the 21st century”. Talking about the influencers from this sought after industry, they are creating content about data science day in and day out across multiple platforms. And the prime goal is to promote the ecosystem all across the world. Further, believe it or not, there are people who taking influencing as a full-time job. And if you are wondering how they make money, we will let you know about it. In one of our previous articles, we have written about some of the effective ways that would take you closer to becoming a LinkedIn Influencer. And in this article, we are going to take a look at the ways a data science influencer can make money. How To Turn Influencing Into A Business Are you someone who is passionate about data science, extensively talking about it on social media, delivering value, and have thousands of people following you? Then there are ways you can turn your social media influencing into a source of income. Product Review Product review is considered to be one of the best ways of marketing and companies nowadays are extensively looking for someone from the relevant domain who has a big audience, to review their products. Talking about the data science domain, you can be a reviewer as well. When you have a significantly big audience who engage with your data science content, you can give reviewing a try. Whether it’s a new chatbot by a company, or an AI-assistant, or a recommender engine, you can always share your thoughts on how they work and how good they are. Once your content starts to reach your audience and they engage with it more, there are chances of you getting offers of reviewing products on a commercial basis. Words To The Wise: Data science is a crucial domain and every single information matter. When you are an influencer in this domain, you cannot deliver misleading information. And this also applies to paid reviews, do not provide a biased review if the product is not good enough. Provide Consultation Consultation is one of the ways effective ways to make money. However, you have to be both jack and master of all trades when you are helping a company in their data science journey. Further, you might be really good with all the concepts of data science, but when it comes to consulting, you have to be good with understanding multiple business perspectives. How would you build that understanding? By working with businesses initially for free or at a low price or maybe for a testimonial. Once you build a certain level of track record, you can then go about consulting businesses at a significant price point. Also, keep sharing your work on your social media platforms, as it would give more traction to your consultation side. Become A Keynote Speaker To step into the Keynote speaking territory you have to leverage the powers of video platforms such as IGTV and YouTube. Talk about the latest happenings from the data science industry, share your thoughts. You can also produce tutorial based videos on these platforms. The more your content delivers value and reaches a good amount of people, the more are your chances of becoming a keynote speaker. The main role of a keynote speaker is to establish the main underlying theme of any gathering — whether it’s a conference, a meeting or even a corporate function. Talking about money, there are several speakers from the data science industry who charges a significant amount of money. However, to be a keynote speaker, you have to have a strong identity, a reputation in the industry and you can do that by making your content stand out. Sponsored Posts On Social Media This business model is quite similar to paid product reviews. The only difference is that instead of reviewing a product, you would only be sharing about it to your audience. There are two ways of doing it, either as a post on social media platforms or as a blog post (if you own a blog). But the questions is what kind of things you are going to promote. Do not get all intrigued by the money if you are data science influencer you have to stay relevant to the domain. Just like a product review, you can promote any tool, software, event, or anything that is from the data science industry. For example, there is a data science event that is going to happen next week, and the organizers reach out to you regarding a social media post. Look at the entire agenda of the event, look at all the possible angles, and if you feel that it is something people actually need to know about, you can take it up. But again, make sure, you are not promoting the products that are of no good. By Selling Courses Selling data science courses have become extremely popular in recent years. There are thousands of courses available on the internet and people are investing a handsome amount of money on them. And why not, the data science domain nowadays is attracting a lot of people. However, there is still a gap in the course selling sector — even after paying, the value they deliver is not up to the mark, leaving people disappointed. Being a data science influencer with a huge number of followers, you can fill that void by curating course that actually delivers value. And also, you can keep the price lesser compared to the one in the market already. The major advantage for you would be the trust your audience has on you.","excerpt":"Over the past few years, the ‘influencers’ from every domain have witnessed significant growth. What is more exciting to see that in this millennial era, even those who are not into the lifestyle sector, but into areas like data science domain are making a great impact on the internet. And why not, data science, after […]","categories":["AI Trends"],"tags":["Data Science"],"author_name":"Harshajit Sarmah","publish_date":"2019-09-26T15:46:11","publication_year":"2019","word_count":998,"keywords":["data science","Go","programming_languages:R","AI","programming_languages:Go","RAG","Rust","GAN","Data Science","R","programming_languages:Rust"],"extracted_tech_keywords":["AI","data science","RAG","R","Go","Rust","GAN","programming_languages:R","programming_languages:Go","programming_languages:Rust"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/5-legit-ways-to-make-money-as-a-data-science-influencer\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10100279,"title":"Oracle Introduces Generative AI Features to Enhance Customer Service","content":"​​Oracle today announced the addition of generative AI-powered capabilities within Oracle Fusion Cloud Customer Experience (CX). These AI-powered features, integrated into existing Oracle Fusion Service processes, are set to optimise customer service delivery, boost productivity, and elevate overall customer satisfaction. Rob Tarkoff, the executive vice president and general manager of Oracle Cloud CX, emphasized the importance of quick access to accurate information for creating exceptional customer experiences. “The new capabilities in Oracle Cloud CX will help organizations resolve customer service issues quicker and more efficiently by increasing service agent and field technician productivity, optimizing self-service and automating traditional tasks that are manual and time-consuming,” said Tarkoff. The generative AI capabilities in Oracle Cloud CX are built on Oracle Cloud Infrastructure (OCI) and are designed to prioritize data security and privacy. Oracle ensures that customer data remains protected and is not shared with third parties. Role-based security within Oracle Fusion Service workflows ensures that sensitive information is safeguarded, and content recommendations are restricted to authorized personnel. Benefits  of generative AI in Oracle Fusion Service Firstly, with “Assisted Agent Responses,” service agents receive support in composing responses by utilizing past interactions as a foundation. This not only accelerates response times but also maintains quality as responses can be reviewed and edited before sending. Secondly, the “Assisted Knowledge Articles” feature reduces the workload on service teams by employing generative AI to craft articles for emerging service issues. This ensures quick and accurate documentation of standard procedures, particularly beneficial in complex sectors like high technology and medical devices. Thirdly, “Search Augmentation” improves efficiency for both service agents and customers by integrating short-form responses into search and chat interfaces. This enhancement facilitates swift access to answers, simplifying issue resolution. Moreover, “Customer Engagement Summaries” provide summaries of key information within service requests. This feature aids service agents and administrators in comprehending customer issues and making informed decisions, especially when dealing with intricate requests. Additionally, “Assisted Guidance Authoring” empowers experts to create structured troubleshooting guidance for service agents, ensuring a consistent approach to issue resolution. Lastly, “Field Service Recommendations” boost field service technician efficiency by offering contextually relevant content suggestions during troubleshooting. This capability can reduce the necessity for on-site visits, making service delivery more efficient.","excerpt":"The generative AI capabilities in Oracle Cloud CX are built on Oracle Cloud Infrastructure (OCI)","categories":["AI News"],"tags":["Oracle"],"author_name":"Siddharth Jindal","publish_date":"2023-09-19T20:46:39","publication_year":"2023","word_count":369,"keywords":["programming_languages:R","AI","Oracle","ViT","generative AI","GAN","R"],"extracted_tech_keywords":["AI","generative AI","R","GAN","ViT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/oracle-introduces-generative-ai-features-to-enhance-customer-service\/","complexity_score":2,"technical_depth":6,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":8843,"title":"Why Spark is the new R?","content":"As companies are moving more and more data to Hadoop, the analysts\/ data scientists using them are realizing the limitations of tools they have used in past – SAS\/ SPSS\/ R\/ KXEN. While R had promise of being able to connect to Hadoop and other tools are building connectors as well the capability they typically provide is a small fraction of what is possible on original tool. So then what is the next stage of evolution in the world of data science? There were a number of candidate earlier but for me one tool that has clear edge over others is Spark. Before I go on about why Spark is better, I thought about checking if other felt the same. Below is a snapshot from Google Trends showing Spark in comparison to Hadoop and Hive (the most popular related Apache Project amongst Hive\/ Pig\/ Mahout). It can be concluded that interest in Spark has grown rapidly and still increasing. Having looked closely at Spark (as a company we also offer Spark Training to our corporate clients), I am going to summarize the top reasons that I feel are going to make Spark a very popular tool for Data Scientists. 1. It is Omnidextrous: I don’t know this is a proper word or not but I feel for any tool to become popular it has to support data transformation and advanced algorithms within the same environment. It is surprising how many modeling tools don’t recognize this fact. Algorithm development and data transformations go hand in hand and any tool which allows data scientists to do both side by side is going to be a hit. This I feel was a major factor in success of SAS and R. Spark takes it one step further and allows to mix SQL, Machine Learning and Streaming seamlessly (at least it is working towards that). So this is the only tool a data scientist may need in time to come. 2. Hadoop is popular Hadoop is great for data storage of both structured and unstructured data. Many companies have already moved or are in the process of moving their internal data to Hadoop. Hadoop was always a popular destination for storing the external \/ social media data. So now you have a scenario where all the data is going to be on one platform. Now companies are going to need tools to mine all the data. This is where Spark comes in. The fact that Spark can be installed over Yarn\/ Hadoop 2.0 is important and means Spark can work directly with data that is already in HDFS. [quote]Enterprise Hadoop is a market that is not even 10 years old, but Forrester estimates that 100% of all large enterprises will adopt it (Hadoop and related technologies such as Spark) for big data analytics within the next two years.[\/quote] 3. It is Open Source and free Again zero software cost is a big plus. No doubt there would be companies, incorporating Spark in their Hadoop distributions. But the low cost is going to be factor in adoption. 4. It is fast The in-memory processing means that Spark can perform faster than most other tools around. There is enough evidence which suggest 10-100 X improvement in many cases. In our own test, Spark SQL comprehensively outperforms same Hive query even when data is stored in Hive Metastore. I am sure there would be claims that there are specialized applications which run faster than Spark for a particular technique and they may be right. However, in general Spark would be way faster for doing a variety of operations compared to anything that is available right now. 5. Support for multiple languages Spark allows users to write codes in Scala, Java, Python and now R. Also allows for mixing of classes\/ elements with Java and Python. So it is not a Scala or nothing. Data scientists used to other languages can get started faster. However, for more complex processing it would be better in longer run to shift to Scala. I also feel I must make a disclosure that although I don’t have any directly links with Spark development, as an analytics consulting company, MathLogic, has vested interest. We use and recommend Spark as appropriate to our clients and also engage in training our corporate client in Spark.","excerpt":"As companies are moving more and more data to Hadoop, the analysts\/ data scientists using them are realizing the limitations of tools they have used in past – SAS\/ SPSS\/ R\/ KXEN. While R had promise of being able to connect to Hadoop and other tools are building connectors as well the capability they typically […]","categories":[],"tags":[],"author_name":"Anurag Verma","publish_date":"2016-02-02T05:25:17","publication_year":"2016","word_count":719,"keywords":["data science","machine learning","AI","ML","RAG","Python","Aim","analytics","SQL","R"],"extracted_tech_keywords":["AI","machine learning","ML","data science","analytics","Aim","RAG","Python","R","SQL"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/spark-new-r\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":42330,"title":"Top 7 Most Trusted Cybersecurity Firms In India","content":"The cybersecurity space in India is still in its nascent stage at $4.4 billion a year. Now with the Indian government taking multiple initiatives to encourage the digital transformation of the country through the growth of tech firms, the cybersecurity market is also set to grow at a CAGR of 15% for the next five years. Experts predict the industry will grow to $35 billion by 2025. Here we list the 7 most popular cybersecurity firms in India, who are already riding this growth wave: (These firms have been listed in alphabetical order) Cyberops Infosec LLP Based out of Jaipur, this cybersecurity firm aims to provide digital protection to other tech firms and offers a wide range of information security services. This includes web and mobile application penetration testing, network penetration testing, wireless penetration testing, source code review, security enhancement, firewall and router set review. This company’s services are used in sectors like law enforcement agencies, financial institutions, educational institutions and public and private sectors. eSec Forte Technologies eSec Forte Technologies is one of the few companies in India that have a CMMI level 3 certification for global consulting and IT services. They have expert offerings like infosec services, forensic services, malware detection, security audit, vulnerability management, penetration testing and risk assessment. Many of their services have been certified by various quality certification institutions. They are authorised technology partners of information security solutions with several firms like Tufin, Atola Technology, Cyber Deception and more. eSec also manufactures CrackBox, their proprietary tool for password breaking\/cracking. They have offices in Gurgaon (Headquarters), Bangalore, Mumbai, Singapore and Sri Lanka. Hicube Infosec Pvt. Ltd. Hicube infosec is one of India’s leading cybersecurity organisations that specialises in vulnerability assessment and penetration testing, online malware scanner and cybercrime consultancy. They strive towards having a firm global footprint in cybersecurity and cybercrime investigation. They provide customised and specialised products and services to their clients. Hicube also offers training in cybersecurity analyst, cybersecurity specialist and cybersecurity expert. These courses are designed for people who intend to explore these fields deeper and advance their understanding of the subject. These courses can also be taken up by non-technical professionals looking for career opportunities in the cybersecurity space. K7 Computing Pvt. Ltd. Founded by J Kesavardhanan in 1991, K7 computing has introduced cybersecurity solutions which have won numerous international awards and certifications. This company has clients in over 100 countries. They provide every client with constant and consistent protection against any internet-based threats. With over 25 million users their client base covers a wide range of government organisations and private industries from finance to healthcare and education. K7 computing’s flagship products are K7 Total Security and K7 Enterprise Security which are used by home users and organisations alike. Quick Heal Technologies Ltd. Quick Heal is one of the leading It security solutions company in India. The company is headquartered in Pune and was founded in 1995. The company has put in nearly two and a half decades of R&D into the computer and network security solutions. Their clients range from home users and small offices to corporate companies. Quick Heal’s international products are certified by ICSA Labs, AV-Test and other such agencies. Their international operations span over 80 countries globally and the company has around 1,300 employees currently. Skylark Information Technologies Private Limited Headquartered in Chennai, this company was founded in 1993 and provides IT solutions and services to over 300 clients. Skylark provides a range of IT security solutions from application and database security, Next-Gen Firewall and endpoint mobile security solutions to cloud security solutions fully customised for the clients’ needs. Skylark also offers a range of services in blockchain technology. They have partnered with manufacturers such as HPE, Fujitsu and VMware to offer complete IT infrastructure solutions for small and businesses alike. Wi-Jungle Wi-Jungle’s journey started in 2014 and was the first private company to be launched as a Free WiFi service provider. Operating from Jaipur, the company now offers NextGen Network Security Firewall (NGFW)\/ Unified Threat Management (UTM), Hotspot Gateway, etc to various companies across the globe in many industries like hospitality, healthcare, education, retail and more. Wi-Jungle stands out from the rest of its competitors by providing one appliance that serves the purpose of NGFW\/UTM as well as a hotspot gateway thus eliminating the necessity of two different devices.","excerpt":"The cybersecurity space in India is still in its nascent stage at $4.4 billion a year. Now with the Indian government taking multiple initiatives to encourage the digital transformation of the country through the growth of tech firms, the cybersecurity market is also set to grow at a CAGR of 15% for the next five […]","categories":["AI Trends"],"tags":["Cybersecurity","information security"],"author_name":"Thirunarayan Simhan","publish_date":"2019-07-12T13:30:30","publication_year":"2019","word_count":722,"keywords":["Go","programming_languages:R","AI","digital transformation","programming_languages:Go","Git","RAG","Aim","information security","GAN","Cybersecurity","R"],"extracted_tech_keywords":["AI","Aim","RAG","R","Go","Git","GAN","digital transformation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-7-most-trusted-cybersecurity-firms-in-india\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":11755,"title":"NVIDIA vouches on building AI talent pool in India by launching “Inception Program”","content":"Keeping in line with the innovation that India is witnessing in AI space, NVIDIA, the AI computing company has launched the NVIDIA Inception program, a virtual incubator program to brace up startups with revolutionary ideas in AI and data science. With an aim to assist the members during critical stages of product development, prototyping, and deployment, Inception provides a custom set of ongoing benefits to each startup from hardware grants and marketing support to training with deep learning experts. Launched at the NVIDIA Emerging Companies Summit India, Inception would provide key support to help grow their businesses and bring products to market faster. The features that startups can avail from NVIDIA are: Deep learning expertise: Being the world’s leading AI computing company, deep learning comes at the heart of everything they do. An Inception member can access their global ecosystem including a massive network of deep learning GPUs: The deep learning community is using GPU-accelerated platforms for training and inference in every industry. Inception members can apply for GPU hardware grants, access the latest software, and get remote access to state-of-the-art technology. Hands-on training: NVIDIA’s Deep Learning Institute (DLI) offers online and in-person courses on the latest techniques for designing, training, and deploying neural-network powered machine learning in a range of applications. Marketing lift: Inception offers startups a suite of benefits tailored to specific marketing goals, ranging from blogs and podcasts to event support and videos. GPU ventures: NVIDIA invests in next-gen AI leaders through its GPU Ventures program, and every Inception applicant is submitted for consideration.","excerpt":"Keeping in line with the innovation that India is witnessing in AI space, NVIDIA, the AI computing company has launched the NVIDIA Inception program, a virtual incubator program to brace up startups with revolutionary ideas in AI and data science. With an aim to assist the members during critical stages of product development, prototyping, and […]","categories":["AI News"],"tags":[],"author_name":"Srishti Deoras","publish_date":"2016-12-23T04:29:57","publication_year":"2016","word_count":257,"keywords":["data science","Go","machine learning","programming_languages:R","AI","innovation","Aim","deep learning","R","startup"],"extracted_tech_keywords":["AI","machine learning","deep learning","data science","Aim","R","Go","innovation","startup","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/nvidia-vouches-building-ai-talent-pool-india-launching-inception-program\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":49636,"title":"This Bengaluru-based Startup Uses Analytics To Identify Gaps In Current Lending Value Chain","content":"Even today, many dread the idea of visiting banks, and going through the tedious policies in terms of documents and eligibility is a huge challenge while accessing loans. Banks have a stringent requirement in terms of high credit scores, an exhaustive list of documents to be produced and fulfilling the eligibility criteria. To make these processes easy and fulfil every financial aspiration with the simplest, shortest and fastest way possible, veteran banker Manav Jeet founded Rubique. The sole aim behind Rubique is to identify gaps in the current lending value chain and deploy technology solutions to streamline and ease out the process challenges. And, they are driving it with a 15-membered team comprising of data architect, backend and frontend developers. How Rubique Makes It Easy Rubique aggregates borrowers and lenders on its technology platform and provides a hassle-free experience to its borrowers through its matchmaking platform and end to end fulfilment model. In doing so it leverages machine learning and artificial intelligence along with big data analytics to build a technology system that matches the borrower with the right lender. Rubique integrates credit policy of the lender on its system and uses self-training models through a feedback loop to provide custom offers to borrowers with high approval chances. Also, its Lending Gateway helps provide real-time decisioning and approval by integrating with the bank systems. AI Is Driving The Automated Marketing Solution AI is an integral part of Rubique. “Having spent 4 years in this space and working towards creating a sizeable base which is around 2,00,000 customers as on today, we have started work on deploying technologies like data science to build the cross-sell, upsell insights. We have also worked on Rubique Confidence score which makes use of alternate data & planning to market it to our FI partners to combine with traditional assessment parameters,” says Jeet. Rubique Confidence Score relies on the digital footprint of the users to derive intent. Various sources of data that contribute to Rubique Confidence Score are SMS data, Device Data, Location Data, Call Data Record (CDR), Behavioral Data etc. Their in-house NLP-based SMS parser engine extracts more than 50 variables from SMS data per user. Some of the signals captured from SMS include salary info, employer info, running EMIs, credit card usage and more. Overall, more than 200 variables from various data sources are used to construct the Rubique Confidence Score. The weightage of these variables is derived by using machine learning algorithms on training and testing data. Loan performance of various users serves as a feedback mechanism to improve the algorithm. “We are now working on rank ordering our score with Bureau score to define ranges for bucketing of customers.”  shared Jeet. Technology Stack At Rubique Rubique flaunts a microservice-based architecture with each microservice dedicated for a particular task. As Jeet shares, they use REST API structures and hence integration with any system is very simple. They also use MEAN tech stack, Amazon cloud service provider, have polyglot data stores like MySQL, MongoDB, Redis etc. The data science algorithms are written in Python and Scala. Some of the other tools used are Hadoop & Spark, Redshift, Kafka and more. How Is It Revolutionizing Finance Industry With AI-based Solutions Due to the varied risk appetite of financial institutions, the credit assessment & underwriting criteria are heterogeneous. The current market practices do not provide a decisive answer about whether the application will be accepted or rejected. “Due to this “Move in application” mindset, the rejection ratio is in tune of 60%-70% leading to further credit gap which is essentially not because of the product unavailability but due to inefficiencies in right product discoverability,” says Jeet. Rubique is trying to address these challenges with below solutions: MatchMaking Engine: AI-based matchmaking engine is built in-house to provide custom-matched offers to the customer based on their profile, needs and interest. It takes into consideration various data points that a financial institution intends to capture. It then has a workflow engine which takes raw data input and derives a lot of other data points from it. It is then fed to the credit matrix and BRE to find the right set of offers. Recommendation Engine: AI-based Recommendation Engine takes into account the customer data captured for the loan\/card application and runs that data against the other credit policies of different products. Growth Story Rubique likes to differentiate itself from its competitors in its approach towards technology and thinking of financial services which have helped them launch various technology solutions for the industry. Rubique is unique in terms of providing a neutral platform where ender & borrowers can discover each other bringing transparency in the process, matchmaking algorithm and recommendation engine, technology-enabled distribution tapping every consumer segment, among others. With a portfolio of around INR 3,500+ crores loan disbursement facilitation through the platform, 150,000+ credit card setups & revenue generation of INR 70+ crores, Rubique is marching towards a leadership place in its domain. Serving as a financial matchmaking platform for the entire ecosystem it is adopting to a physical approach to address the need of the entire ecosystem. “The next step for the startup aggressively strengthening the team by adding talent in areas of data science and products.” says Jeet on a concluding note.","excerpt":"Even today, many dread the idea of visiting banks, and going through the tedious policies in terms of documents and eligibility is a huge challenge while accessing loans. Banks have a stringent requirement in terms of high credit scores, an exhaustive list of documents to be produced and fulfilling the eligibility criteria. To make these […]","categories":["AI Startups"],"tags":["real time decisioning","real-time integration to kafka"],"author_name":"Srishti Deoras","publish_date":"2019-11-11T10:07:41","publication_year":"2019","word_count":875,"keywords":["data science","machine learning","artificial intelligence","AI","ML","RAG","NLP","Aim","analytics","real-time integration to kafka","Kafka","real time decisioning"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","NLP","data science","analytics","Aim","RAG","Kafka"],"url":"https:\/\/analyticsindiamag.com\/ai-startups\/this-bengaluru-based-startup-uses-analytics-to-identify-gaps-in-current-lending-value-chain\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10121022,"title":"Generative AI: How to Move from Promises to Production","content":"Generative artificial intelligence. GenAI for short. It’s the one topic that has companies excited more than anything else in the past few years. And for good reason; it pulls the best information possible out of your data. It has the potential to transform how a business operates, how employees work, and how we all live our daily lives. It’s a powerful tool that has placed the world on the edge of a new sort of Industrial Revolution. For many companies, it’s a brand-new territory. At Axtria, we’ve been working with data analytics across our products and solutions since day one. Uncovering hidden insights from information – what GenAI does best – is part of our DNA. What about the companies that don’t have data so naturally embedded in their culture? How can they join the GenAI party? They can start by following four key concepts crucial to all GenAI strategies, irrespective of industry. Choose GenAI use cases that provide measurable impact. Some companies take a shotgun approach: trying GenAI in everything and seeing what works. As you can imagine, that’s a massive waste of resources. Maybe you want to improve a customer’s buying journey. Perhaps you want your sales reps to get insights and tips before walking into a meeting. Remember, not all use cases will have the same level of organizational impact, so choose wisely. You need someone to help you tailor a GenAI journey for specific projects that are scalable and will ensure adoption by your team members. Consider working with an established firm that can guide you and enable your GenAI journey of exploration, experimentation, industrialization, adoption, and, sometimes, monetization. Build strong Generative Relational Databases. You’ll quickly realize that every bit of data you have holds value, even the type that doesn’t fit neatly into a spreadsheet. “Traditional” data—names, numbers, sales figures, and such—needs to be clean and error-free. GenAI can parse the not-so-neatly-organized data—photos, video, audio, even handwritten notes. Combine it all by leveraging what are called Generative-Ready Datasets (GRD). These GRDs train the Large Language Models that form the basis of GenAI. This is important: Get this correct at the start, and you won’t have to worry about mistakes scaling with you. Your goal is to industrialize your GenAI usage, and this GRD strategy sets you up for it, not just with larger datasets but across all business functions. The potential is there for the taking if you seek products and platforms that leverage GenAI in production. Ensure a viable, intertwined GenAI and AI strategy. You can explore and experiment with GenAI, but industrialization can only happen once you have validated the strategy – by accepting the proofs-of-concept. That means you have to sit down and have frank discussions on sensitive issues: intellectual property protection, personal data, and other critical areas across your products and solutions. Setting the proper guardrails here can avoid severe reputational damage and prevent stumbles later on. Relying on a trusted and experienced partner who can help you look around corners is essential. Define the characteristics of industrialization excellence. In order to scale up, you need to have a platform that people are willing to use. The user interface must be simple to navigate, with relevant information easy to find. The user experience has to be positive, or you’ll never get full adoption. Likewise, you need to ensure your business users are on board with a new platform. Getting buy-in and positive change management ensures trust in the AI models. If your team isn’t using the model, or they don’t trust it, it’s meaningless.  Once you have that in place, you’re now ready to start turning promises into wider practice. Find a partner who listens to your industrialization goals and has expertise with products and platforms that leverage GenAI in production. One who can help you define the best practices and what will be considered a success. Remember, industrialization with GenAI doesn’t have to be limited to “more customers.” GenAI can find efficiencies in your own team, so don’t forget to look inwards. When you’ve got your GenAI strategy humming along beautifully, you must remain vigilant. Responsible GenAI usage involves security, as well as frequent monitoring of the analysis models, user controls, and usage compliance. Accountability is a crucial factor as well. You must document everyone tasked with ownership, oversight, and approval at every stage, including the usage of third-party models. Another critical aspect of GenAI responsibility comes from the human factor. You have to commit to eliminating biases in model training. We all have opinions, and that makes us unique. Be sure those feelings don’t handcuff the model, leaving important insights uncovered. Finally, you have to be able to explain your GenAI model. It must be transparent and traceable, which helps with governance considerations. Seeing how a model developed its answer helps with model refinement and reporting responsibilities. Putting these all together will give you a GenAI strategy that’s effective and well-suited for industrialization. It’s a big ask of any company, but leaning on a partner with established know-how across products and solutions will get you up and running faster and safer than doing it yourself. Since our founding in 2010, Axtria has focused on helping our clients make better decisions by enabling the best use of data. Our life sciences clients, whose work helps save lives, depend on Axtria and our rich expertise in leveraging data to help them work better, smarter, and with greater impact. Interested to read more on the impact and future of GenAI: Reading the Tea Leaves: New Insights from Gartner® on the Future of Gen AI and the Value it’s Already Adding in Life Sciences The Use of Natural Language Processing in Literature Reviews","excerpt":"What about the companies that don’t have data so naturally embedded in their culture? How can they join the GenAI party?","categories":["AI Highlights"],"tags":["Generative AI"],"author_name":"Jassi Chadha","publish_date":"2024-05-21T19:00:00","publication_year":"2024","word_count":950,"keywords":["Go","GenAI","artificial intelligence","AI","Scala","RAG","analytics","Rust","GAN","Generative AI","R"],"extracted_tech_keywords":["AI","artificial intelligence","analytics","GenAI","RAG","R","Go","Rust","Scala","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/generative-ai-how-to-move-from-promises-to-production\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10010871,"title":"Thinking Beyond Transformers: Google Introduces Performers","content":"The key component of the transformer architecture is the attention module. Its job is to figure out the matching pairs (think: Translation) in a sequence through similarity scores. When the length of a sequence increases, calculating similarity scores for all pairs gets inefficient. So, the researchers have come up with the sparse attention technique where it computes only a few pairs and cuts downtime and memory requirements. According to Google researchers, sparse attention methods still suffer from a number of limitations: They require efficient sparse-matrix multiplication operations, which are not available on all accelerators.They do not provide rigorous theoretical guarantees for their representation power. Optimised primarily for Transformer models and generative pre-training.Difficult to use with other pre-trained models as they usually stack more attention layers to compensate for sparse representations, thus requiring retraining and significant energy consumption. Not sufficient to address the full range of problems to which regular attention methods are applied, such as Pointer Networks. Along with these, there are also some operations that cannot be sparsified, such as the commonly used softmax operation, which normalises similarity scores in the attention mechanism and is used heavily in industry-scale recommender systems. To overcome the limitations of sparse transformers, Google introduced Performers, a Transformer architecture with attention mechanisms that scale linearly, thus enabling faster training while allowing the model to process longer lengths. Overview Of Performers The Performer uses an efficient (linear) generalised attention framework, which allows a broad class of attention mechanisms based on different similarity measures (kernels). The framework is implemented by the novel Fast Attention Via Positive Orthogonal Random Features (FAVOR+) algorithm, which provides scalable low-variance and unbiased estimation of attention mechanisms that can be expressed by regular softmax-attention. Regular softmax-attention can be seen as a special case with these nonlinear functions defined by exponential functions and Gaussian projections. The key behind making any attention matrix to work efficiently is the use of positive random features, i.e., positive-valued nonlinear functions of the original queries and keys. To evaluate Performers, the researchers ran an experiment on the protein sequences. Proteins are large molecules with complex 3D structures, and like words, proteins are specified as linear sequences where each character is one of 20 amino acid building blocks. Applying Transformers to large unlabeled corpora of protein sequences yields models that can be used to make accurate predictions about the protein folds. Performer (ReLU-based attention), stated the researchers, performed strongly at modelling protein sequence data, while Performer-Softmax matches the performance of the Transformer. According to the researchers, Performers matched the performance of Transformers and showed promises of applications beyond Transformers. This work, concluded the researchers, is an attempt to diversify research in the area of non-sparse attention techniques and Performers is a brand new way of thinking about attention, Transformer architectures, and even kernel methods. Implications Of Performers According to the researchers, this framework can have significance in the following areas: Medicine Performers have the potential to directly impact research on biological sequence analysis by enabling the Transformer to be applied to much longer sequences without constraints on the structure of the attention matrix. Modern bioinformatics can immensely benefit from faster, more accurate language models, for development of new nanoparticle vaccines. Climate change Performers on FAVOR algorithm lead to much lower compute costs and substantially lower space complexity which can be directly translated to CO2 emission reduction and lower energy consumption, as regular Transformers require very large computational resources. According to the researchers, FAVOR can also be applied to applications which are outside the scope of Transformers. This opens up Performers to a wide range of avenues, including hierarchical attention networks, graph attention networks, image processing, and reinforcement learning. Check the original paper here.","excerpt":"The key component of the transformer architecture is the attention module. Its job is to figure out the matching pairs (think: Translation) in a sequence through similarity scores. When the length of a sequence increases, calculating similarity scores for all pairs gets inefficient. So, the researchers have come up with the sparse attention technique where […]","categories":["Deep Tech"],"tags":["Generative Pre-Trained Transformer","Google Analytics","self attention models","Transformers"],"author_name":"Ram Sagar","publish_date":"2020-10-28T13:00:15","publication_year":"2020","word_count":614,"keywords":["Go","attention mechanism","programming_languages:R","AI","self attention models","ai_frameworks:Transformers","Transformers","Scala","Google Analytics","Generative Pre-Trained Transformer","programming_languages:Scala","transformer architecture","R"],"extracted_tech_keywords":["AI","Transformers","R","Go","Scala","transformer architecture","attention mechanism","ai_frameworks:Transformers","programming_languages:R","programming_languages:Scala"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/transformers-attention-google-introduces-performers\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":16573,"title":"Can biometrics rewrite the future of travel and say bye to airport queues?","content":"pretty indian business woman checking flight information at airport As airlines and airports embrace biometrics to significantly improve passenger experience, biometric data economy has emerged as one of the major trends in travel experience. If you are wondering what’s the real value in leveraging biometrics, it helps airlines use big data to personalize products and create real value for flyers by turning travel into a seamless paperless experience. Another key highlight is that the biometric data generated also helps to meet the high security criteria demanded by governments across the globe to make informed decisions. Analytics India Magazine caught up with Maneesh Jaikrishna, Vice President, India and Subcontinent, SITA to discuss how “Big Data” can effectively be used to build a connecting bridge between the flyer and online activities. Airlines already hold a tremendous amount of data on their passengers — particularly their frequent flyers with every interaction being logged and stored. With the agreement of the passenger, this data can be used to build a clearer picture of who they are and how the airline can micro-target you with everything from ‘personalized’ fares to double status credits deals, shared Jaikrishna, explaining how big data can be used to create an enriched travel experience. “At every step of the journey, the data is being monitored and suggestions are provided to the passengers based on their preferences or data captured, including information they opt to share around their social media activities,” he further added. Role of biometric in personalizing travel & in-flight experience The future of aviation is being shaped by biometrics and the idea is to make the entire travel experience hassle free, secure and seamless for the passengers. Identity Management is one key area where the industry can streamline the passenger experience using technology and the use of biometrics is central to that. With biometric technology, travelers can avoid showing documents at each stop through the airport, explained Jaikrishna, making airline queues a thing of the past. And globally, Delta Airlines and Jet Blue Airways have already put this in motion by testing biometric boardings at select airports.  Another case in point is Australia’s Brisbane Airport that has implemented SITA Smart Path™, a biometric and automated passenger screening solution that enables passengers to present their details at a self-service kiosk at check-in then, when ready to board and use an automated boarding gate to be verified using face recognition technology. Exploring the JetBlue use case Jaikrishna talked about the most recent trial, where they worked with JetBlue and the US Customs and Border Protection (CBP) to test a new paperless and deviceless self-boarding process as part of ongoing trials to implement the US biometric exit process in the future. Interestingly, JetBlue is the first airline to integrate with CBP to use biometrics and facial recognition technology to verify customers at the gate during boarding. How it works: Ø  Those who opt in during the boarding process can put away their boarding passes and devices and simply step up to the camera for a quick photo Ø  The custom-designed camera station will connect to CBP to instantly match the image to passport, visa or immigration photos in the CBP database and verify flight details Ø  A cleared to board (or not cleared) notification appears on the screen for the passenger and is also sent to the airline’s systems Airline Industry is ripe for AI & cognitive computing AI has impacted every sector and as airports grapple to embrace digital technology —  next-gen technologies AI, cognitive computing and predictive analytics can be leveraged in a lot of ways to “predict and prepare for future to tackle the estimated US$25 billion cost of flight disruptions to the air transport industry”. From automating manual processes to predicting flight delays, premium airlines such as Emirates is already using AI to remodel the ticketing process and make flying experiences more personalized. However, the real promise lies in AI autopilots that can help navigate cockpits effectively and complex flights. The Boeing company has already taken the first step in this direction by establishing a research initiative — Boeing\/Carnegie Mellon Aerospace Data Analytics Lab to leverage CMU’s prowess in machine learning, language technologies and data analytics. According to Boeing, the aim is to find ways to use AI and big data to cash in on the enormous amount of data generated in the design, construction and operation of modern aircraft. India Growth story In India, the Ministry of Civil Aviation is working towards creating an integrated platform that will use a passenger’s Aadhaar-based biometric information for the check-in process across all airports. SITA’s 2016 India IT Trends Benchmark study pointed out that when it came to small-scale R&D projects, airlines and airports in India were most active in evaluating new technologies. 80% of airlines in India plan to assess the potential of IoT with R&D projects by 2019. And while India may be slow to embrace the power of digital, technologies such as wearables and biometric travel tokens for passengers will be evaluated by 75% of airlines in India over the next five years. “The whole idea of implementing biometrics in air travel is to cut short the time a passenger spends in airport queues for check-in purposes,” said Jaikrishna.","excerpt":"As airlines and airports embrace biometrics to significantly improve passenger experience, biometric data economy has emerged as one of the major trends in travel experience. If you are wondering what’s the real value in leveraging biometrics, it helps airlines use big data to personalize products and create real value for flyers by turning travel into […]","categories":["IT Services"],"tags":[],"author_name":"Richa Bhatia","publish_date":"2017-07-28T07:15:15","publication_year":"2017","word_count":873,"keywords":["Go","machine learning","AI","R","ML","Git","RAG","Aim","analytics","predictive analytics"],"extracted_tech_keywords":["AI","machine learning","ML","analytics","Aim","RAG","predictive analytics","R","Go","Git"],"url":"https:\/\/analyticsindiamag.com\/it-services\/can-biometrics-rewrite-future-travel-say-bye-airport-queues\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10094301,"title":"Internal Drama Unveiled: A Behind-the-Scenes Look at Rust","content":"The open-source community is no stranger to drama. Personalities like Linus Torvalds and Richard Stallman are prime examples of how the open-source community can cause personal fallouts. Stay geared, for the next target for drama is the programmers’ most-loved language – Rust. Amidst bureaucratic issues at the Rust Foundation, the open-source community has created its own fork of the programming language, endearingly called Crab. However, the reason behind this fork belies internal problems at Rust. Read on to discover the saga of how this new kid on the block is being crippled by internal politics. A tale of bruised egos For context, the Rust Foundation was already on thin ice with their community. In April, the Foundation suggested changes to their trademark, which would force many open-source projects to change their names for compliance. Apart from this, the restrictive nature of the trademark policies left many members of the Rust community bitter. Adding on to this, a new controversy has set the community on fire. The upcoming RustConf, a conference for Rust enthusiasts, has been surrounded by hearsay and rumours. This issue was then capped off by a firecracker blog written by one of the speakers at the conference. JeanHeyd Meneide, a software engineer and project editor for the C standard, was working on an experimental feature for Rust. Reportedly, this work attracted the attention of the organisers of RustConf 2023, which then led to Meneide being invited to give a keynote talk at the conference. However, due to unknown reasons, the organisers decided to downgrade the keynote to a regular talk. In his blog, Meneide stated, “These were shadowy decisions that are non-transparent to normal contributors like myself”.  Reportedly, this was done at the request of the Rust Project leadership, which is currently in the midst of a restructuring. Meneide then withdrew their talk entirely in protest. This led to an uproar within the community, as Meneide is a well-known figure in both Rust and C communities. In response to this mismanagement, one of the main contributors to the Rust project, JT, stepped down. In a tweet, they stated, “I just hope it’s clear to everyone that the core issue is that Meneide wasn’t treated with respect, and has failed many times…That’s why I stepped down.” In an accompanying blog, they also stated that the Rust organisation ‘disgraced’ one of the foremost experts in the field. This sentiment was also echoed by the community, with many believing that this was a disgrace to Meneide. Due to the attention surrounding this controversy, the community decided to again turn to one of the mainstays of open source: forking. During the trademark fiasco, a group of developers from the Rust community had forked the language into another one they called ‘Crab’. In an obvious dig at the Rust community being called Rustaceans, this fork was dismissed as a ‘ridiculous gesture’ by the founder of the Rust foundation. However, with this renewed distaste towards the management of Rust, the fork saw more attention. According to GitHub star history, Crab saw around 1900 stars during the time of its launch in April. While this growth slowed down from April to May, only adding on 200 more stars, the controversy set its growth on fire. The repo currently has 3675 stars, and while it has not set out to replace Rust, but to increase the amount of freedom people have while using it. In fact, we have seen this same pattern emerge in multiple open-source communities, with varying degrees of success. Who gives a fork? Some of the open-source community’s biggest projects are forks. Ubuntu, currently the world’s most popular Linux distribution, is a fork of Debian. Mozilla Firefox, which has the second-largest user base after Chrome, was forked from the Mozilla Application Suite. Various examples abound for successful forks, but there are also a lot of unsuccessful ones, which aren’t spoken about. This shows the dual nature of forking open-source projects. They can either result in the creation of a new product that then builds its own ecosystem, or fade into obscurity as the main project continues to flourish. Some forks can even result in the momentum behind an open source project getting halted in its tracks. One of the biggest examples of this is Bitcoin, and its fork Bitcoin Cash. While there is a heated debate on which one is better, the fact remains that contributors who could have been working on Bitcoin are now working on Bitcoin Cash. This splits the ecosystem in half, resulting in the division of the community. However, Crab was not created with the intention to split apart or replace Rust. In their words, “We want to emphasise that we are not at odds with the project or the original language. Our main goal is to ensure that the community has an alternative that aligns with their values and desire for unrestricted use.” While this does not seem to be a form like many others in the past, the fact that developers and members of the community have a choice is a cornerstone of open source. While this saga shows that even open source foundations, the so-called fountainhead of democracy and open governance, can fall behind, the spirit of open source is always there to push innovation.","excerpt":"Amidst bureaucratic issues at the Rust Foundation, the open-source community has created its own fork of the programming language, endearingly called Crab.","categories":["AI Features"],"tags":[],"author_name":"Anirudh VK","publish_date":"2023-06-01T15:08:13","publication_year":"2023","word_count":879,"keywords":["Go","programming_languages:R","AI","innovation","Git","ViT","Rust","GAN","GitHub","R"],"extracted_tech_keywords":["AI","R","Go","Rust","Git","GitHub","GAN","ViT","innovation","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/internal-drama-unveiled-a-behind-the-scenes-look-at-rust\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10102874,"title":"Elon Musk Wants to Open Source Grok","content":"In the recent podcast with Lex Fridman, Elon Musk suggested that he likes the idea of open source AI. “I am generally in favour of open sourcing, like biassed towards open sourcing.” Musk’s xAI has recently released Grok, its own based chatbot. It took just four months for the team to build it. He is also planning to double the compute at xAI every month. Currently, Grok is trained on 8,000 NVIDIA A100 GPUs. On the contrary, Sam Altman just built Grok with a single prompt on GPT Builder, taking a dig at Musk’s dream project. GPTs can save a lot of effort: pic.twitter.com\/VFIrGzPuMN— Sam Altman (@sama) November 10, 2023 Musk further adds that the whole idea of him founding OpenAI was about open sourcing AI. He highlighted his discussion with Larry Page, the former CEO of Google, who was Musk’s friend then. “I sat in his house and talked about AI safety, and Larry did not care about AI safety at all.” The discussions with Larry were the reasons that Musk founded OpenAI. “At one point Larry called me a ‘speciest’, for being pro-human,” Musk added. He mocked Page for being on the team of robots. “The breaking of the friendship was because of OpenAI,” Musk said when Fridman asked him if he can be friends again with Page. “The key moment was recruiting Ilya Sutskever. He is brilliant, a good human, and has a great heart” he added about the co-founder and chief scientist at OpenAI. Recruiting Sutskever was a battle between Musk and Demis Hassabis, the founder of DeepMind. “Ilya went back and forth between staying at Google or joining OpenAI. Finally, he agreed to join OpenAI. That was the toughest recruiting battle we have ever had.” Musk said that Sutskever is the linchpin of OpenAI. Musk said that he was crucial in recruiting a lot of other people at OpenAI and providing almost all the funding in the beginning, around $40 million dollars. “The ‘open’ in OpenAI is all about open source,” said Musk, highlighting that the company has become a closed source for maximum profit company, which according to him is “not good karma.” Only solution is Musk buying back OpenAI.","excerpt":"He said that OpenAI becoming closed source is “not good karma.”","categories":["AI News"],"tags":["Open Source AI"],"author_name":"Mohit Pandey","publish_date":"2023-11-10T11:59:14","publication_year":"2023","word_count":366,"keywords":["Go","funding","OpenAI","AI","A100","Open Source AI","GPT","XAI","xAI","AI safety","R"],"extracted_tech_keywords":["AI","OpenAI","xAI","R","Go","GPT","AI safety","XAI","funding","A100"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/elon-musk-wants-to-open-source-grok\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10112301,"title":"Google Releases Gemini Ultra, Rebrands Chatbot Bard to Gemini","content":"Google has finally released the highly anticipated Google Gemini Ultra. The version featuring Ultra will be named Gemini Advanced, offering a new, enhanced experience that excels in reasoning, following instructions, coding, and creative collaboration. Simultaneously, Google announced that Bard will now simply be called Gemini. “It is available in 40 languages on the web and will be coming to a new Gemini app on Android, as well as the Google app on iOS,” wrote Google Chief Sundar Pichai in a blog post. The capabilities of Gemini Advanced extend to serving as a personal tutor, creating customised step-by-step instructions, sample quizzes, and engaging in back-and-forth discussions tailored to individual learning styles. For coding enthusiasts, the chatbot proves invaluable in handling advanced coding scenarios, providing a sounding board for ideas, and aiding in the evaluation of different coding approaches. To access Gemini Advanced, users can subscribe to the Google One AI Premium Plan, priced at $19.99\/month, which includes a two-month trial at no cost. This plan provides users with Google’s latest AI advancements along with the benefits of the existing Google One Premium plan, featuring 2TB of storage. AI Premium subscribers will also soon gain access to Gemini in Gmail, Docs, Slides, Sheets, and more, formerly known as Duet AI. Gemini Advanced is available today in more than 150 countries and territories in English. Google has also introduced a mobile app for Gemini on Android and iOS.The Gemini app for Android is available on Google Play with support for English, but support for Korean and Japanese will be coming next week. By downloading the Gemini app or opting in through Google Assistant, users gain access to the new overlay experience, available from the app or wherever Google Assistant is activated, such as through the power button, corner swiping on select phones, or using the “Hey Google” command. On iOS, access to Gemini will be introduced within the Google app in the coming weeks. Users can simply toggle the Gemini feature and engage in conversations to boost creativity, create custom images, seek assistance in writing social posts, and even plan activities directly from the Google app.","excerpt":"Google has also introduced a mobile app for Gemini on Android and iOS.","categories":["AI News"],"tags":["Gemini"],"author_name":"Siddharth Jindal","publish_date":"2024-02-08T19:48:39","publication_year":"2024","word_count":353,"keywords":["Go","Gemini","programming_languages:R","AI","programming_languages:Go","RAG","llm_models:Gemini","ViT","llm_models:Bard","R"],"extracted_tech_keywords":["AI","RAG","R","Go","ViT","llm_models:Gemini","llm_models:Bard","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/google-releases-gemini-ultra-rebrands-chatbot-bard-to-gemini\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10160727,"title":"AIM Print Jan 2025 edition","content":"The January 2025 edition of AIM Print shines a spotlight on transformative AI developments and the visionaries shaping the future. This issue highlights pioneering figures, disruptive startups, and India’s growing dominance in the AI landscape. Cover Story: Pragya Misra – OpenAI’s Face in India OpenAI’s ambitious entry into India is being driven by Pragya Misra, the company’s first and only employee in the country. With India emerging as OpenAI’s second-largest user base, Misra’s role encompasses building partnerships with developers, policymakers, and the tech community. Misra believes that if AI solutions work for India, they can scale globally, reflecting India’s unique diversity and innovation potential. She emphasizes OpenAI’s commitment to empowering local startups through tools, credits, and technical support​. AI Without a PhD? Think Again Meta AI’s chief Yann LeCun advises aspiring AI entrepreneurs in India to pursue advanced academic degrees to navigate the complexities of AI. He suggests that technical expertise, often gained through PhDs, plays a pivotal role in building successful AI startups. This aligns with a growing sentiment that foundational knowledge, rather than surface-level skills, will drive the next wave of AI innovation​. Nikhil Kamath’s Bet on Young Innovators Zerodha co-founder Nikhil Kamath’s ‘Innovators under 25’ initiative, supported by WTFund, is nurturing India’s youngest and brightest entrepreneurs. This year, nine startups across AI, healthcare, and sustainable technology received non-equity grants of up to INR 20 lakh. Kamath’s unconventional approach focuses on supporting talent without taking ownership, encouraging entrepreneurs to retain full control of their startups​. Key initiatives under WTFund include: Mars Computers – Cloud-based high-performance computing for creative professionals. BioCompute – DNA-based data storage solutions addressing scalability and sustainability. RNT Health Insights – AI-powered diagnostics for gastric cancer, enhancing medical accuracy during endoscopy. Pixa AI – Interactive AI-powered educational toys reducing screen time for children. CallPrep – AI-driven pre-meeting sales preparation software enhancing productivity​. India’s AI Momentum – Leading the Charge in 2025 India is rapidly positioning itself as a hub for AI agents and generative models. The country’s engineering talent and focus on application-driven AI are propelling India ahead. This edition highlights how Indian startups are developing domain-specific AI models, reflecting the trend of local innovation with global implications​. Turing’s Dream – Paras Chopra’s Vision Paras Chopra, founder of Wingify and creator of ‘Inverted Passion,’ is pursuing AGI through his initiative, Turing’s Dream. This residency program provides cloud GPUs and a collaborative environment for AI researchers to build cutting-edge projects. Chopra believes that AI systems need more than brute-force data scaling, advocating for the integration of external symbolic agents to improve reasoning","excerpt":"The AIM Print January 2025 edition offers readers an in-depth view of AI’s evolving landscape, from global giants like OpenAI to grassroots innovation driven by India’s vibrant startup ecosystem.","categories":["Magazine"],"tags":["AIM","magazine","print"],"author_name":"AIM Media House","publish_date":"2025-01-04T18:37:04","publication_year":"2025","word_count":425,"keywords":["API","Meta AI","magazine","OpenAI","AI","innovation","Scala","RAG","Aim","ViT","AIM","R","print"],"extracted_tech_keywords":["AI","OpenAI","Meta AI","Aim","RAG","R","Scala","API","ViT","innovation"],"url":"https:\/\/analyticsindiamag.com\/magazine\/aim-print-jan-2025-edition\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10046531,"title":"AI Features Auto Firms Are Embedding In Cars","content":"Artificial Intelligence (AI) has already found its applications across sectors, and nowadays, it has become a hot-selling potato in the automotive industry, especially in cars. The global automotive industry is projected to reach $9 trillion by 2030, as per the Statista report. As a result, carmakers are embedding AI-based solutions for a better customer experience, thereby ensuring good market share for themselves. In this article, we list recent AI features embedded in cars by auto firms in 2021. Car-as-a-Platform (CAAP) by Morris Garages The British automotive company MG featured a personal AI assistant and first-in-segment Autonomous Level 2 technology in their soon to be launched mid-size SUV – Astor. Designed by ‘Star Design’ – the AI assistant portrays human-like emotions and voices. It further entertains customers’ queries in real-time on any topic through Wikipedia. Mid-range radars and a multi-purpose camera power the Autonomous Level 2 technology in MG Astor, implementing various advanced driver-assistance systems (ADAS). Advanced features include Adaptive Cruise Control, Lane Keeping Assist, Forward Collision Warning, Lane Departure Warning, Lane Departure Prevention, Automatic Emergency Braking, Intelligent Headlamp Control (IHC), and Rear Drive Assist (RDA). In addition, features are optimised, keeping Indian traffic conditions in the loop and ensuring safe-driving aspects. Vector Space in Tesla Cars On its ‘Tesla AI Day’, AI engineers showcased how the tech is helping advance Fully Self Driving (FSD) systems. Engineers introduced Vector Space that captures the three-dimensional view to sense the environment via eight cameras installed on the car itself. These eight inputs are combined into a single virtual environmental prediction model that provides the car’s computers with a bird’s-eye perspective of the world while navigating. The data collected via Vector Space is fed into a collection of AI algorithms called Neural Net Planner that handles routing, behaviour, and trajectory of the car on the road. The algorithm handles every turn and vees around pedestrians and lane change. The car runs thousands of simulations per minute to decipher the behaviours of other vehicles, pedestrians, and cyclists to take the best course of action on its own. Image Credits: CNET Robocar from Baidu Baidu, the Chinese internet giant, announced its AI-packed ‘Robocar’ at its annual flagship technology conference — Baidu World 2021. The car with no steering wheels and brake pads relies entirely on smart intelligence for locomotion. Fitted with zero-gravity seats and speech and facial recognition, the AI technology enables the car interior to analyse and provide predictive suggestions to passengers. Zero-gravity chairs reduce the effects of gravity by placing the driver or passenger’s body in such a way that they feel weightless, thereby raising the comfort status quo. Moreover, the car is stated to have met Level 5 autonomous driving standards – signals the power of the car to handle itself without human intervention under all circumstances. AdrenoX Connect AI in Mahindra XUV700 Mahindra & Mahindra Ltd. has announced AdrenoX AI  that offers intelligent features through integrated 10.25-inch dual screens and SmartCore cockpit domain controller technology with the third-generation Qualcomm Snapdragon Automotive Cockpit Platforms developed in collaboration with Visteon. The AI-based system provides safety features – asking drivers to slow down when required and alerts when in a drowsy state. Additionally, the AdrenoX Connect app allows for emergency assistance via e-call\/SOS. The app further provides owners with access to data and a dashboard to have a real-time check on vehicle status via their smartphones or smartwatch. Moreover, the built-in Alexa function looks after the users’ entertainment, information, and offline vehicle control. Joya by Toyota Motor Corp Toyota Motor Corp has announced AI-powered, voice-activated virtual assistant – Joya for its Toyota Sienna car owners. The AI companion hosted within the Toyota App through Toyota’s Connected Services assists in personalising driving settings, exploring various features, locating button controls and safety information. Furthermore, the voice-activated feature includes interactive capabilities such as exploring the vehicle’s dashboard, delivering instructions on interior and exterior car details and educating drivers about Dynamic Radar Cruise Control, Lane Departure Alert, and other safety systems. The market is witnessing the increasing role of AI in vehicles, especially cars. It might become a norm in the near future, without even the need to learn driving.","excerpt":"Artificial Intelligence (AI) has already found its applications across sectors, and nowadays, it has become a hot-selling potato in the automotive industry, especially in cars.","categories":["IT Services"],"tags":["AI (Artificial Intelligence)","Self Driving Cars"],"author_name":"kumar Gandharv","publish_date":"2021-08-23T11:00:00","publication_year":"2021","word_count":691,"keywords":["Go","artificial intelligence","programming_languages:R","AI","Self Driving Cars","programming_languages:Go","ai_applications:autonomous driving","RAG","Ray","ViT","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","Ray","RAG","R","Go","ViT","programming_languages:R","programming_languages:Go","ai_applications:autonomous driving"],"url":"https:\/\/analyticsindiamag.com\/it-services\/ai-features-auto-firms-are-embedding-in-cars\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":9305,"title":"The Tiered Acquisition of Analytics Quotient","content":"Millward Brown, a global leader in brand, media and communications research will acquire the business operations of Analytics Quotient (AQ), a marketing analytics company that extracts insights from data to help clients define their marketing strategies. Millward Brown is a leading global research agency specializing in advertising effectiveness, strategic communication, media and brand equity research, by helping clients grow great brands through comprehensive research-based qualitative and quantitative solutions. Interestingly, AQ will become part of Millward Brown Analytics, which in turn is a part of Kantar, the data investment management arm of WPP. Commenting on the acquisition, Travyn Rhall, global CEO, Millward Brown stated, “Our clients are continually looking for ways to turn data into brand and consumer insights that deliver a competitive advantage and AQ has developed a unique storytelling approach that connects data from across sources and combines it with industry knowledge, analytics and data visualization to answer business questions. Having already collaborated with the AQ organization on client work, we are delighted to now welcome the team and its expertise to Millward Brown and across the Kantar stable of brands.” While, Kantar group is one of the world’s largest insight, information and consultancy groups with its services are employed by over half of the Fortune Top 500 companies; WPP is the British advertising and public relations giant, owning a large group of companies under its umbrella including Kantar. Headquartered in London, WPP employs more than 1,79,000 people in about 3,000 offices across 111 countries. Thus, WPP is acquiring Analytics Quotient indirectly through two tiers – Kantar and Millward Brown. This is understandable, as AQ will be working directly with Millward Brown, but for Kantar as well. With this acquisition, WPP continues its strategy of investing in important markets and sectors and strengthening its capabilities in digital and data investment management businesses. However, the AQ leadership team and its approximately 400 employees will become part of Millward Brown Analytics and continue to be led by Pritha Choudhuri, AQ’s CEO and the company founder. Pritha Choudhuri told us that they are delighted to join Millward Brown and Kantar. “We know that every brand has a unique story hidden behind data and we look forward to helping more brands solve their business problems by extracting that story.” she added. Founded in 2008 in Bangalore (India), Analytics Quotient has worked with some of the world’s largest CPG, retail and hospitality brands, using marketing consulting frameworks, statistics, CRM transaction analysis and a proprietary analytics platform to extract business insights from data. Further, it also builds data visualization tools and custom analytics solutions to help clients slice, dice, simulate and monitor business data.","excerpt":"Millward Brown, a global leader in brand, media and communications research will acquire the business operations of Analytics Quotient (AQ), a marketing analytics company that extracts insights from data to help clients define their marketing strategies. Millward Brown is a leading global research agency specializing in advertising effectiveness, strategic communication, media and brand equity research, […]","categories":["AI News"],"tags":["Mergers and Acquisitions"],"author_name":"Apoorva Verma","publish_date":"2016-03-11T07:13:50","publication_year":"2016","word_count":439,"keywords":["programming_languages:R","AI","Git","analytics","GAN","Mergers and Acquisitions","R","analytics platform"],"extracted_tech_keywords":["AI","analytics","R","Git","GAN","analytics platform","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/tiered-acquisition-analytics-quotient\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":38807,"title":"How India’s Premier Engineering Institutes Are Using AI To Make Social Media A Safe Place","content":"One of the biggest threat that is challenging democracies across the world is the issue of fake news. With the advent of social media, spreading of misinformation has become a pervasive problem, which has resulted in the toppling of governments, changing the course of the election and even killing of innocent people. However, with the emergence of technologies like artificial intelligence and machine learning, these technologies have proven to be a potent weapon to find inaccurate news, cross check the content and to validate particular news. Check: NIRF Ranking of Engineering Colleges In India, a number of educational institutes like the Indian Institute of Technologies (IITs) and the International Institutes of Technologies (IIITs) have been conducting various researches in the field to make social media a safe place for the netizens. In this article,  we sum up the research and development conducted by these institutes to stem the spreading of fake news over social media. Indian Institute of Technology, Kharagpur: Researchers from the institute’s Department of  Computer Science and Technology devised an AI algorithm that could detect fake news with nearly 90 per cent accuracy. Originally designed to tame propagation of fake news during the time of a calamity, the researchers said that it has proven to be successful in other contexts as well. The solution can scan through scores of content in social media, extract content, authenticate the message and then map the origin of the message. “Our solution can detect fake news and can even alert users in the time of disasters through deep machine learning algorithms. We are developing AI methods to extract specific types of critical and actionable information from social media content posted during disasters such as resource needs, resource availability and information about trapped people,” Saptarshi Ghosh, a developer of the solution said. The solution, which was developed with the help of Microsoft  Research India was tested on Twitter and Whatsapp during the 2015 Nepal earthquake and then in 2016 floods. International Institute of Technology, Hyderabad: In 2018, researchers from the institute developed an AI and ML-based web-engine called the Fake-O-Meter, that can analyse through a large volume of data to assess the news. Further, the engine rates the report in green, amber and red based on the result. With the help of technologies like AI, ML and Natural Language Processing (NLP), the engine does a primary scan of content, network and domain analysis. Though not yet open to the public, the app is currently available in English, Spanish and Chinese. Speaking on the development,   Vasudeva Varma, Professor and Dean (Research) of IIIT (H) said, “The solution is based on Machine Learning, Artificial Intelligence and Natural Language Processing technologies. You copy the URL (web address) of a report that you wanted to verify and paste it in the query box Indraprastha Institute of Information Technology, Delhi : Ponnurangam Kumaraguru, a professor at the college have been building an application that uses machine learning to establish the veracity of news, videos, text and audio recording. The app, which is currently being developed, provide a label of authenticity to the message and then colour codes depending on its authenticity. “This app will be on your phone and will keep looking at your messages and analyse it through machine learning to annotate a post about its genuineness. The app will also show you a credibility score to rate a message,” Kumaraguru noted. IIIT D– Recently the researchers from the institute came up with an innovative idea to reduce automated face analysis of customers over the net. The mechanism is used by marketers to analyse customers’ face from their social media and to make any suggestive advertisements. To prevent it, the researchers developed ML algorithms and relied on predictive analytics to stop the misuse of customer data. “Veral websites display advertisements based on the outcome of facial analysis including age, gender and ethnicity of the person. In several cases, advertisement companies crossed the line which results in anxiety and depression. For example, showing slimming facilities if the person looks overweight in the image on social media sites,”  Saheb Chhabra, researchers at IIIT Delhi said.","excerpt":"One of the biggest threat that is challenging democracies across the world is the issue of fake news. With the advent of social media, spreading of misinformation has become a pervasive problem, which has resulted in the toppling of governments, changing the course of the election and even killing of innocent people. However, with the […]","categories":["AI Features"],"tags":["AI engineers","Facebook","IIIT-H","ML","whatsapp"],"author_name":"Akshaya Asokan","publish_date":"2019-05-08T10:54:35","publication_year":"2019","word_count":686,"keywords":["Go","artificial intelligence","machine learning","AI","AI engineers","ML","R","RAG","whatsapp","NLP","analytics","Facebook","IIIT-H","predictive analytics"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","NLP","analytics","RAG","predictive analytics","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-indias-premier-engineering-institutes-are-using-ai-to-make-social-media-a-safe-place\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10022733,"title":"FDA Authorises First ML-Based COVID Screening Device","content":"The US Food and Drug Administration (FDA) has issued an emergency use authorisation (EUA) for a Tiger Tech-released ML-based COVID-19 non-diagnostic screening device — COVID Plus Monitor. Termed to be first-of-its-kind, COVID Plus Monitor leverages machine learning that identifies certain biomarkers indicative of the virus that causes COVID-19. The Tiger Tech COVID Plus Monitor has been designed for use by trained personnel to prevent exposure and spread of the deadly virus, SARS-CoV-2, as well as to identify biomarkers that are specifying other conditions like hypercoagulation, where the blood tends to clot more easily than normal, or hyper-inflammatory states, such as severe allergic reactions, in asymptomatic individuals over the age of five. The screening tool is an armband that is equipped with light sensors and a small computer processor, which when wrapped around the patient’s bare arm, will provide the prediction of whether the individual is showing COVID symptoms or not. It first obtains pulsatile signals from blood flow, which is then used to extract some key features such as pulse rate. These pulse rates are then fed into an ML model, which then makes the necessary predictions. The results are then displayed using coloured lights on the armband that indicates the presence of certain biomarkers. According to the FDA’s official release, the device is not a substitute for the COVID-19 diagnostic test; however, it can be used by asymptomatic individuals without fever. The trained personnel checks this by carrying out a temperature reading. When it doesn’t meet the criteria, they can then use the COVID Plus Monitor device. When asked about this new release, Jeff Shuren, MD, JD, director of FDA’s Center for Devices and Radiological Health, stated in the official release that FDA is committed to supporting innovative ways to fight this deadly pandemic. He said that using this new screening tool along with temperature checks will help healthcare workers identify patients infected with this virus, which in turn will reduce the spread and provide better treatment to them.Tiger Tech COVID Plus Monitor has been evaluated in hospitals and schools where it showed a positive per cent agreement of 98.6%, by accurately identifying COVID positive individuals, and a negative per cent agreement of 94.5%, where it correctly identified the proportion of the COVID negative individuals that do not possess certain biomarkers. However, the release has also stated that a patient’s underlying condition can sometimes interfere with the COVID related performance of the device and lead to an incorrect screening result.","excerpt":"The US Food and Drug Administration (FDA) has issued an emergency use authorisation (EUA) for a Tiger Tech-released ML-based COVID-19 non-diagnostic screening device — COVID Plus Monitor. Termed to be first-of-its-kind, COVID Plus Monitor leverages machine learning that identifies certain biomarkers indicative of the virus that causes COVID-19. The Tiger Tech COVID Plus Monitor has […]","categories":["AI News"],"tags":["FDA","Machine Learning"],"author_name":"Sejuti Das","publish_date":"2021-03-23T10:29:28","publication_year":"2021","word_count":411,"keywords":["machine learning","programming_languages:R","AI","ML","Machine Learning","RAG","FDA","R"],"extracted_tech_keywords":["AI","machine learning","ML","RAG","R","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/fda-authorises-first-ml-based-covid-screening-device\/","complexity_score":2,"technical_depth":6,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10011609,"title":"INSOFE &#038; University Of Strathclyde Launch MSc In Machine Learning &#038; Deep Learning","content":"In a recent development, the Electronic & Electrical Engineering department from the University of Strathclyde has joined hands with the International School of Engineering (INSOFE) to offer a Master’s Degree in Machine Learning and Deep Learning. The joint program is a full-time master’s degree that provides real-world and global experience and is designed with 200 hours of Postgraduate preparation done with INSOFE online. After successfully completing the postgraduate preparatory programme at INSOFE the students complete two semesters of course work and one semester of industry project at the University of Strathclyde. Upon successful completion, the student gets a fully accredited Master’s Degree by the University of Strathclyde, UK and will also be eligible to apply for the Graduate Route visa and get 2 years of work permit in UK. The preparatory post graduate program at INSOFE prepares students with 3 or 4 year degrees (or final year students) for the rigorous post graduate work at Strathclyde in Machine Learning and Deep Learning. Speaking on the occasion Stuart Shorthouse, Head of International Development & recruitment, Faculty of Engineering University of Strathclyde said, “We are excited about this joint program and expect to receive bright Indian students.” Professor Campbell Booth, Head of Department, Electronics & Electrical Engineering, University of Strathclyde added, “This is perhaps one program that combines theoretical foundations with practical exposure like no other post-graduate program does.” Dr Gaetano Di Caterina, Chancellor’S Fellowships Electronic & Electrical Engineering, University of Strathclyde said, “With the world-class faculty and support structure, Strathclyde‘s ML and DL programs is one of the finest options to acquire most in-demand AI and ML skills.” Dr Dakshinamurthy V Kolluru, President and Co-Founder, INSOFE said “LinkedIn ranked AI and ML engineering as the most in-demand job in the world and this program prepared students precisely for that. Dr Parag Mantri, Principal Data Scientist and Associate Professor, INSOFE also added that this degree is the perfect launchpad for Indian students to start a glowing career in the UK. University of Strathclyde’s E&E department is ranked in the top 5 in UK and INSOFE is a top Data Science Education and Research Institute in India.","excerpt":"In a recent development, the Electronic & Electrical Engineering department from the University of Strathclyde has joined hands with the International School of Engineering (INSOFE) to offer a Master’s Degree in Machine Learning and Deep Learning. The joint program is a full-time master’s degree that provides real-world and global experience and is designed with 200 hours of […]","categories":["AI News"],"tags":["Insofe","Machine Learning","MSc data science","MSc. in data science"],"author_name":"Srishti Deoras","publish_date":"2020-11-12T20:31:51","publication_year":"2020","word_count":354,"keywords":["MSc. in data science","data science","Go","machine learning","programming_languages:R","AI","ML","Machine Learning","programming_languages:Go","Insofe","RAG","deep learning","MSc data science","R"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","data science","RAG","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/insofe-university-of-strathclyde-launches-msc-in-machine-learning-deep-learning\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10173169,"title":"Hugging Face Launches Reachy Mini, Open-Source Robot for AI Enthusiasts and Educators","content":"Hugging Face, in collaboration with Pollen Robotics, launched Reachy Mini, a desktop-sized open-source robot designed for AI experimentation and education. It is now available for pre-order globally. Developed for human-robot interaction and creative coding, the robot is available in two versions: Lite at $299 and a wireless version at $449. Thomas Wolf, co-founder and chief science officer at Hugging Face, announced in a LinkedIn post that the first deliveries are expected right after the summer of 2025 through 2026. Built for developers, educators and hobbyists, Reachy Mini enables users to program and deploy AI applications using Python. The robot includes multimodal sensors and offers integration with Hugging Face for real-time behaviour sharing and experimentation. Two Versions for Flexible Use The Reachy Mini Lite lacks onboard computing and Wi-Fi, whereas the wireless version is equipped with a Raspberry Pi 5, a battery, and four microphones. While the Lite version is compatible with Mac and Linux, it has not yet been released for Windows. Both variants feature motorised head movement, body rotation, animated antennas, and a wide-angle camera. A speaker and audio-visual interaction capabilities come standard. The robot is currently in the early stages of development. “We’re sharing it as-is, without warranties or guarantees, to engage with early adopters and gather feedback,” the company mentioned in its announcement. The Mini Lite version consists of two microphones, while the Mini version has four and measures 11 “\/28cm in height and 6.3”\/16cm in width, weighing 3.3 lbs\/1.5 kg. Open-Source Development Users can test and deploy behaviours in real life or through simulation. Over 15 preloaded behaviours will be available at launch through the Hugging Face hub. Future programming support will be expanded to include JavaScript and Scratch. Reachy Mini’s open-source hardware and software allow for full transparency and community participation. With a modular kit-based assembly, it encourages hands-on learning, coding with children, and collaborative building. Users can join the growing community of over 10 million on Hugging Face to upload, download and evolve new robot behaviours, positioning Reachy Mini as a flexible tool for AI exploration and learning.","excerpt":"Priced from $299, the compact robot targets developers, students and hobbyists","categories":["AI News"],"tags":["Hugging Face","Robotics"],"author_name":"Sanjana Gupta","publish_date":"2025-07-09T18:18:39","publication_year":"2025","word_count":344,"keywords":["Hugging Face","AI","Modal","Robotics","RAG","Python","JavaScript","programming_languages:Python","ai_frameworks:Hugging Face","R","Java"],"extracted_tech_keywords":["AI","Hugging Face","RAG","Python","R","JavaScript","Java","Modal","ai_frameworks:Hugging Face","programming_languages:Python"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/hugging-face-launches-reachy-mini-open-source-robot-for-ai-enthusiasts-and-educators\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":45011,"title":"How India &#038; France Are Collaborating To Create A Secure Cyberspace","content":"Source: Twitter – @narendramodi In recent times, India has witnessed some of the most notorious cyberattacks, not only on individuals but also on business and government entities. According to reports, India bags the third place in 2018 after Mexico and France in terms of most number of cyberattacks in businesses. India was also the most attacked nation in the IoT space last quarter with a 22% jump in cyberattacks on IoT deployments in the country. In fact, there has been a rise in cyberattacks on Indian institutions after Articles 370 and 35A were repealed. In order to fight these attacks, India has been working and partnering with other nations as well. And one of the helping hands that the nation got is from France. Over the years, most of the cyberattacks in France are either DDoS attacks or malware attacks. For instance, in February 2011, Paris G20 Summit witnessed malware attack, which came as an email attachment and permitted access to the infected computer. However, in recent years, France has been proactive in cyber defence and cybersecurity. Indo-France Collaboration In the third Indo-French cyber dialogue which was held in Paris on 20 June 2019, India and France have joined hands to work closely in the areas of cybersecurity. According to reports, the meeting was co-chaired by Henri Verdier, French Ambassador for Digital Affairs, and Upender Singh Rawat, Joint Secretary in charge of e-Governance, Information Technology and Cyber Diplomacy at the Indian Ministry of External Affairs. The meeting was wholly focused on both the nation’s commitment to promoting an inclusive, transparent and open digital environment. India and France have decided to strengthen their cybersecurity game by drafting legal and regulatory framework and best practices. This would also include the protection of Economic Information Infrastructure impacting National security and testing and certification of Digital products. Furthermore, there is a whole new concern going on regarding 5G and both the nation share their interest in working together towards mitigating\/eliminating the risks associated with the deployment of 5G technologies and also work on the technical solutions adopted to deal with them. This entire scenario of information sharing will also help India and France to fight cyber crimes as it would be focused on the collection of pieces of evidence, identification of offenders (particularly malware developers), hosters etc. The Role Of AI While cybersecurity was one of the main focuses in the recent Indo-France dialogue, Artificial Intelligence also took the stage. As AI has been a key game-changer in several industries, India and France believe that it would do wonders when it comes to fighting cybercrime. The nations have decided to work on developing and implementing AI policies\/programs for citizen-centric services, data sovereignty from legal, regulatory and cybersecurity perspectives, by making the best use of their expertise and best practices. Moreover, they are also in talks about their commitment to working in this direction in the various multilateral fora (G7, G20, UN) and participating in the International Panel on Artificial Intelligence (IPAI). Here Are Some Of The Other Aspects That Both The Nations Will Work On Digital governance: India and France would work to keep the approach to secure Digital sector legitimate, fair and balanced at the international level by developing the necessary framework to ensure that technologies remain protective of public goods, data sovereignty and fundamental freedoms. Thwart Data Manipulation: At the dialogue, the nations even highlighted the need for measures to prevent manipulation of information. They also stated their intent to work on eliminating data manipulation and fake news. Protection of personal information: France and India also decided to work on developing an innovative Digital ecosystem that is secure and respectful of users’ data protection. This digital partnership between India and France also will also hold regular consultations where things like an economic exchange, R&D, training, education, AI initiatives etc. will be discussed and solutions to the challenges will be proposed. Outlook It is not the first time that India has joined hands with other nations when it comes to cybersecurity, there are instances when India and other nations have partnered to fight cybercrimes. Recently this year, India and Japan partnered to work on cybersecurity as part of the growing security partnership in the Indo-Pacific region. The collaboration has been done keeping in mind the ever-increasing violation of social media platforms by extremists. India had also partnered with Israel in 2017. When it comes to cybersecurity, India has been extensively active to mitigate and eliminate the risk. The nation over the years has not only worked with private companies that are focusing on cybersecurity but also with government organisations.","excerpt":"In recent times, India has witnessed some of the most notorious cyberattacks, not only on individuals but also on business and government entities. According to reports, India bags the third place in 2018 after Mexico and France in terms of most number of cyberattacks in businesses. India was also the most attacked nation in the […]","categories":["AI Features"],"tags":["Cybersecurity","france","International Affairs"],"author_name":"Harshajit Sarmah","publish_date":"2019-08-26T18:30:20","publication_year":"2019","word_count":766,"keywords":["Go","Cybersecurity","artificial intelligence","programming_languages:R","AI","programming_languages:Go","Git","france","GAN","International Affairs","R"],"extracted_tech_keywords":["AI","artificial intelligence","R","Go","Git","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-india-france-are-collaborating-to-create-a-secure-cyberspace\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10047584,"title":"Base Functions or Libraries: What Should Data Scientists Prefer?","content":"A solid selection of libraries is an essential element of a developer’s toolkit for researching and developing complicated applications without having to write a lot of code. In general, library functions are thoroughly tested and optimised before they are released, and space and run time remain prioritised. Moreover, a library is a collection of code designed to make common operations go faster. Whereas in base functions, we know the entire skeleton, of course. The question arises – which is better, a base function or a library? Python is a popular data analysis language because of its extensive libraries for data processing, visualisation, machine learning, and a variety of other tasks. In machine learning, data preparation is crucial. So, let’s explore the base function and library by reading some data. CSV vs Pandas: There are several methods and classes in the CSVmodule that allow you to read and write csv files with ease. Similarly, the “PANDAS” library provides quick, versatile, and expressive data structures that enable dealing with “relational” or labelled data easily and naturally for Python programmers to use. Upon using Python and importing both the packages, these are loading time values. We can see that the basic library imports faster than the advanced library function. Both do the same task in this case. User-defined functions, on the other hand, are much faster than advanced library functions. The base functions are defined based on the customisation. As a result, the specified conditions are carried out. In the advanced library function, however, all of the cases were covered. As a result, it takes longer than the base functions. Let’s check how long a csv file takes to read. If the basic function is specified correctly with all of the required arguments, it will process and provide output in a shorter time span. When compared to the base function, the advanced library function takes longer to read the csv file. Because Pandas is an advanced library, it is capable of handling nearly any situation. The Pandas library function accepts the default value if the delimiter is missing and executes the operation. But, the base function will give an error if the delimiter is not present. If you wish to examine data from a csv file using Pandas, the csv file is converted to a data frame, which is required for data manipulation with Pandas. Therefore you shouldn’t use the CSV module in these instances. The benefits of using advanced libraries: They’re effective: If you want to use library functions, you should do so for one simple reason: they work. Multiple rigorous tests have been conducted on these functionalities, and they have shown to be straightforward to use. The functionalities have been optimised for speed: Because the functions are “standard library” functions, they are continually improved by a committed team of engineers. As a result of this, they are able to provide the most efficient code, which is optimised for maximum performance. Inference: The base function can be defined based on the program’s specific requirements. In the base function, we know the entire structure. If no explicit user preference exists, the library function can be used to do the task. The major benefit of using the base function is that it saves process time. In most cases, libraries are maintained by a group of contributors and made available to anybody who wishes to use them over the internet. One must have a thorough understanding of the criteria. It’s sometimes preferable to use the built-in library function because it’s suited to our requirements. Developers can avoid creating repetitive code by using libraries. Otherwise, if the needs are so precise that the performance is the concern or no specialised library is accessible, it’s better to create our own customised structure using the base function. Of course, there’s nothing wrong with doing so. However, knowing alternative methods, such as this base function allows you to build more efficient code.","excerpt":"A comparison of base functions and libraries. Which is the developer’s preferred option?","categories":["AI Features"],"tags":["csv","how to measure twitter influence","libraries","load data python","pandas","Python"],"author_name":"Dr. Nivash Jeevanandam","publish_date":"2021-09-03T13:00:00","publication_year":"2021","word_count":653,"keywords":["Go","load data python","machine learning","csv","TPU","AI","programming_languages:R","how to measure twitter influence","programming_languages:Go","libraries","Python","pandas","programming_languages:Python","R","Pandas"],"extracted_tech_keywords":["AI","machine learning","Pandas","TPU","Python","R","Go","programming_languages:Python","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/base-functions-or-libraries-what-should-data-scientists-prefer\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10002111,"title":"6 Engrossing Adventure Games For Low-End Computers","content":"Adventure games are some of the most fun anyone can have in a dungeon. Usually featuring fleshed-out combat, exploration elements and lots of fun, adventure games are a good fit for anyone wishing to lose themselves in a new world. Even modern dungeon adventure games do not have that much of a high-end requirement. This makes them a perfect fit for playing on a low-end setup. Enter The Gungeon Enter The Gungeon, as the name suggests, is a game where the player can take up a role as a ‘Gungeoneer’. The player character is tasked with exploring a randomly-generated gauntlet known as the Gungeon, where they must find a gun to kill their past. Combat is at the prime focus of the game, with over 300 guns and items for the player to find in the dungeons. The game, however, is extremely difficult, falling in the ‘bullet hell’ genre of games. Players are also encouraged to strategize to defeat the different kinds of enemies in the game. The game also features co-op play, and has progression elements that extend across multiple playthroughs. With enough practice, it can become a gratifying experience. Wizard of Legend Wizard of Legend is an adventure game where the player takes on the role of an ancient wizard who is equipped with various magical spells. The game features randomly generated dungeons with over 200 different spells and combinations to try out. It has a basic storyline, with most of the focus being on combat elements. It is a punishing game which requires skill to play, along with a general sense of where enemies are on the screen. It also has a charming aesthetic, with pixel-style graphics and screen-filling special attacks. The game also gives players the freedom to use whichever spells they wish to in any combination, allowing for the creation of deadly combos. Hyper Light Drifter Hyper Light Drifter is a pixel-style game that is set in a futuristic world. The player takes on the role of the Drifter, a warrior equipped with an energy sword. As the game progresses, players can unlock more weapons and equipment. Combat is one of the biggest parts of game, with many different types of enemies that require different strategies to beat. Players can also play co-op with another, and can go on an adventure together in the world of Hyper Light Drifter. The game initially released with a high difficulty, but was later balanced. It features exploration elements, with beautiful pixel landscapes and storyline. A must play. Spelunky Spelunky also falls under the procedurally generated dungeon crawling game genre, and follows the adventures of a ‘Spelunker’, The player has to explore underground tunnels while collecting loot and avoiding traps. The areas generated are more and more difficult as the game goes on, and each have distinct items and enemies. It offers hours of entertainment and progression, with small details that keep the game entertaining. The game even features co-op play with upto 4 players, making an accessible and enjoyable experience for all. It’s even free-to-play, leaving no reason to not play it, Binding Of Isaac One of the first games of the new-age rougelike adventure games crop, the Binding of Isaac is a must-play for dungeon lovers. The game features a procedurally generated dungeon that is randomized on every playthrough. Players take on the role of Isaac, a small boy trying to escape from his mother. They are tasked with defeating monsters by shooting tears at them, with various items and powerups giving players new ways to deal with enemies. The game also features multiple bosses and upto 13 endings with more than 10 characters to unlock. It is also known for having a punishing difficulty curve, but is worth playing through. Nuclear Throne Nuclear Throne is a game set in a post-apocalyptic world where mutants roam free. It is an extremely simple game to learn, with a big focus on fast-paced and tight combat. With randomized dungeons to keep the experience fresh, players are tasked with reaching the eponymous ‘Nuclear Throne’ to finish the game, It has a variety of to-the-point weapons, with crushing difficulty and ‘unfair’ deaths. The game also prides itself on not having additional fluff, and has small elements of world-building to allow the story to shine through. Even though the difficulty curve can be a bit crushing, it is definitely worth finding the Nuclear Throne for it.","excerpt":"Adventure games are some of the most fun anyone can have in a dungeon. Usually featuring fleshed-out combat, exploration elements and lots of fun, adventure games are a good fit for anyone wishing to lose themselves in a new world. Even modern dungeon adventure games do not have that much of a high-end requirement. This […]","categories":["AI Trends"],"tags":["low end"],"author_name":"Anirudh VK","publish_date":"2019-05-21T18:13:44","publication_year":"2019","word_count":734,"keywords":["Go","programming_languages:R","AI","ML","programming_languages:Go","low end","RAG","R"],"extracted_tech_keywords":["AI","ML","RAG","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/6-engrossing-adventure-games-for-low-end-computers\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":56623,"title":"AI In India Requires A Different Approach","content":"At the Nasscom Technology Leadership forum, N Chandrasekaran, chairman of Tata Sons said that India should not go by the narratives set by developed countries on Artificial Intelligence. In developed countries, the narrative of AI being used for streamlining business operations, improving productivity, and replacing jobs through automation is legit, but in countries like India, the requirements and implementation is different. Develop countries are witnessing ageing workforces, lack of growth, and more. India, in contrast, has a higher growth rate, the average workforce is too young. In India, AI and other technologies have to be embraced for creating markets, which is not the case for developed counties, where they have deployed technology for making the market more efficient. “We have to devise plans according to India’s demographic demand,” said Chandrasekaran. The idea should be to upskill and bring other professionals apart from software professional and white-collar workers. He further thrashed the notion that the latest technologies such as AI and data science are not only for the elite. We have to ensure that everyone takes part in the development of the country through the use of these technologies. This can only be possible when people from diverse sector start leveraging AI to simplify the tasks by 2030. Many countries are facing a dearth of the workforce, however, for India 90 million people will be included in the coming decade. Consequently, India only needs to focus on upskilling the workforce who in turn will create diverse markets. This will further increase the requirement of professionals for integrating technology for optimising business activities.","excerpt":"At the Nasscom Technology Leadership forum, N Chandrasekaran, chairman of Tata Sons said that India should not go by the narratives set by developed countries on Artificial Intelligence. In developed countries, the narrative of AI being used for streamlining business operations, improving productivity, and replacing jobs through automation is legit, but in countries like India, […]","categories":["AI News"],"tags":["AI in India"],"author_name":"Rohit Yadav","publish_date":"2020-02-13T18:31:57","publication_year":"2020","word_count":261,"keywords":["data science","Go","artificial intelligence","AI in India","AI","ML","Git","RAG","automation","ViT","R"],"extracted_tech_keywords":["AI","artificial intelligence","ML","data science","RAG","R","Go","Git","ViT","automation"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/ai-in-india-requires-a-different-approach\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":38540,"title":"Should Data Scientists Offer Their Skills For Free?","content":"Data Science is one of those domains that is growing at a breakneck speed. However, the industry is in the need for more skilled professionals. Being a vast domain, the number of professionals is less compared to other technology domain. This also means that if you are you are a data scientist, you are in high demand and you are a part of a community of people who has the superpowers of extracting meaningful insights out of scattered data. However, there is a question that has emerged in the world of Data Science — while experts from other technological fields are offering their skills for free to the ones in need, should data scientists also do the same? The answer is something that cannot be flagged as a ‘no’ or a ‘yes’ as it completely depends on the person. So, there are a few points we would like to put out why and why not it is okay for data scientists to work for free. Why It’s Okay To Offer You Data Science Skills For Free Work For A New Skill When we say working for free, it is not exactly free all the time. There are instances when we are offered something else rather than money. So, this concept also goes well with data science. If you are a data scientist and you lack a certain skill, you can always look for unpaid internships where you can get a chance to work and learn.  So, this is a win-win — instead of investing money on a course or class, you are investing your skills to gain new skills. Get Noticed In The Industry By Helping NewBorn Companies You must be thinking about why one should help startups or newborn companies for free? This is more of a strategy to achieve your long term goal. When you help startups and newborn companies for free, you are basically paving a path for yourself to get noticed. Being a data scientist, working on online projects is one thing, but when you work on real-life projects and deliver value, you start to become a significant member of the community. So help as many startups as you can. The more you use your skills to help others, the more they spread the word and the more you start to get noticed. And when you reach a level of where you work becomes the talk of the town, maybe then you can think of monetizing your data science skills (however, it completely depends on you). If You Are Working For Free, Work For A Cause While organisations who are making a huge amount of money might use your skills for their benefit, there is one more field where you can offer your skills for a cause. If you are a data scientist with a significant amount of knowledge, you can always provide free classes to students who cannot afford to take up a course or classes. You don’t have to teach a huge bunch of students; you can always do your research and pick up students you think are really passionate about data science. Also, if you still want to deliver your knowledge to a wider base, you can start a blog and publish lessons and tutorials. Do Not Undersell Your Skills For Free, Set A Value To It Data science is a lucrative and expensive domain. Not every technology enthusiast can become a data scientist. It requires a tremendous amount of know-how to become an expert in this domain. So, when you start gaining a public profile by offering your skills for free to a number of organisations, other companies tend to reach out to you with their data science problem as they would like to get the work done for free or less, rather than spending a huge amount of money of hiring data scientists. And to be brutally honest, if you keep doing it for free, you won’t be able to make a living out of it. You just can’t deny the fact that no matter how much we say that money is not everything, at one point in time you need money for your bread and butter and to pay your bills. So, when you are offering your skills, make sure you do it for a certain amount of time and then start monetizing it and set a Word to the wise: When you are struggling initially, work for free. But when you gain momentum and thrive, know the worth of your skills and set a significant value to it.","excerpt":"Data Science is one of those domains that is growing at a breakneck speed. However, the industry is in the need for more skilled professionals. Being a vast domain, the number of professionals is less compared to other technology domain. This also means that if you are you are a data scientist, you are in […]","categories":["AI Features"],"tags":["Data Science","Data science skills","Data Scientist"],"author_name":"Harshajit Sarmah","publish_date":"2019-05-02T06:31:08","publication_year":"2019","word_count":759,"keywords":["data science","Go","programming_languages:R","AI","R","programming_languages:Go","GAN","Data Science","Data Scientist","Data science skills","startup"],"extracted_tech_keywords":["AI","data science","R","Go","GAN","startup","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/should-data-scientists-offer-their-skills-for-free\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10165020,"title":"Perplexity Co-Founder Aravind Srinivas Joins AI Startup Astra as Angel Investor","content":"Co-founder and CEO of Perplexity AI, Aravind Srinivas, has joined Astra as an angel investor, lending his expertise and resources to the AI startup. Founded in 2023 by IIT alumni Supreet Hegde and Ranjan Rajagopalan, Astra aims to revolutionise sales workflows by building the AI Sales Agent. Astra’s mission is to replace traditional SaaS tools with AI agents capable of automating non-customer-facing sales tasks and providing real-time nudges for better sales execution. By focusing on accuracy and human-level performance, Astra seeks to double active opportunities per sales representative and significantly boost win rates. This partnership further solidifies Srinivas’s commitment to advancing artificial intelligence and supporting innovative Indian-led ventures. Saas AI Agents Hegde announced Srinivas’s involvement in Astra, highlighting his reputation as one of the most influential entrepreneurs of this era. “From day one, Aravind’s advice has been clear—focus on being an AI company first (not just a sales company) and build world-class AI products for your customers,” posted Hegde. Hegde believes that AI Agents will replace human workflows, unlike traditional SaaS tools that merely served as support systems with no concept of accuracy. “Building AI products that can truly perform human level tasks is an incredibly complex challenge, and staying focused on AI from day one is critical,” he said. Srinivas Leads Srinivas’s investments extend beyond Perplexity, with a portfolio that includes several AI startups such as ElevenLabs and Pika. Coming in as an investor for Astra adds to his portfolio of investments backing agentic AI startups in the Saas sector. Earlier, Srinivas committed to invest $1million and dedicate support and mentorship to support Indian AI initiatives. Interestingly, payment provider Paytm announced their partnership with Perplexity to integrate AI-powered search into their app to help customers with financial decisions. While Srinivas is supporting AI startups, he is actively working on pushing his platform. There are a number of features that Perplexity has been rolling out over the last month. Perplexity is now developing an agentic web browser called Comet, aiming for a seamless assistive experience like GitHub Copilot. Comet looks to enter a market dominated by browsers such as Chrome and Safari.","excerpt":"Srinivas has also invested in ElevenLabs and Pika.","categories":["AI News"],"tags":["Aravind Srinivas","Astra AI","Perplexity","Saas AI"],"author_name":"Vandana Nair","publish_date":"2025-03-04T10:05:00","publication_year":"2025","word_count":352,"keywords":["Go","agentic AI","Perplexity","artificial intelligence","AI","ML","Saas AI","Git","Aim","Astra AI","GitHub","R","Aravind Srinivas","startup"],"extracted_tech_keywords":["AI","artificial intelligence","ML","agentic AI","Aim","R","Go","Git","GitHub","startup"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/perplexity-co-founder-aravind-srinivas-joins-ai-startup-astra-as-angel-investor\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10098648,"title":"Is Chatgpt Going to Shut Down","content":"Don’t panic. It might be too soon to say this, but it is quite possible that OpenAI, the company behind the revolutionary AI product, ChatGPT, might just have to pull the plug on it. There are reasons that lead up to this. OpenAI has been trying to fix the mess that ChatGPT is in all this while. Most recently, the company decided to acquire Global Illumination, a startup leveraging AI for building creative tools including games and simulated worlds. The company has previously driven a lot of projects with YouTube, Google, Pixar, and Riot Games. OpenAI said that the whole team would help it to fix its core products including ChatGPT. The latest project that Global Illuminations has built is Biomes, which is an open source sandbox MMORPG similar to Microsoft’s Minecraft, which runs directly on the web. There can be a possibility that the company wants to launch another consumer-facing product, which might possibly be a social experiment\/game — to possibly collect data and train its upcoming AI models. Because for now, the company is finding it very difficult to upgrade its AI models’ capabilities. Though it has filed for a trademark on ‘GPT-5’, it has not yet delivered all the multi-modal capabilities that it had promised with GPT-4. A lot of it is possibly because of the global GPU shortage. But with this partnership, the company might be able to “ethically” collect users’ data as they interact with multiple GPT-based bots or digital NPCs. Not sure if OpenAI is moving away from text-based LLM thought process and following Google DeepMind’s route to build better AI systems, and eventually AGI, based on rule-based games  – the likes of AlphaGo, etc, which are coincidentally being used to build Gemini, which is touted to be launched next month, and is better than GPT-4 (that powers ChatGPT). ChatGPT in a Mess ChatGPT is currently facing a lot of challenges. The product that acquired a million users in just 5 days, is currently witnessing a decline in users. The last three months have constantly seen a decrease in the number of users of website visits. In July, the number of users dropped to 1.5 billion, compared to 1.7 billion in June – i.e. close to 12% percent MoM. One of the reasons for this can be the degradation of the output quality of ChatGPT, which is reported by a lot of users on Reddit and X. To fix these issues, OpenAI needs access to a lot more data to fine-tune the model. But that is not possible. It has been increasingly deciding to increase the privacy measures for its users, and thus has been constantly saying that it won’t collect data from conversations if the users opt out. (if and only if a customer specifically asks us, we will include specified data in a future training run to improve the model in a specific way)— Sam Altman (@sama) August 15, 2023 On the other hand, to tackle this data privacy issue and still be able to train its AI model, the company has been partnering with news and journalism websites such as AP and others to collect first hand information. But in turn is costing the company even more. According to the Information, OpenAI had to bear the loss of $540 million to build ChatGPT, and the operational costs are exorbitantly high as well – around $700,000 per day. Thus, the company is not profiting right now. Funnily enough, all of this money is going from the pockets of Microsoft. This can probably raise a conflict between both the companies about how they want to continue this relationship going forward. In a recent event, Microsoft released a GitHub repository of Azure ChatGPT, where it was offering services directly on the Azure platform. The most astonishing part about this was that the blog read that people have always been concerned about the privacy concerns about using ChatGPT directly as OpenAI might be able to collect data. This does sound like the company is acknowledging that there is a feud between both companies. A day later, the repository was deleted. It is not clear if the upload was by mistake or hinting towards a problem between both the companies. But whatever the case is, it only makes sense for OpenAI to pull the plug on the free web version of ChatGPT given the operational costs and it is not benefiting Microsoft in any way. The only way Microsoft is earning revenue through OpenAI is by the Azure OpenAI Service, and it is only reasonable for Microsoft to push all the users onto it, instead of freely giving away ChatGPT to everyone. Clearly, there can’t be two ChatGPTs. In March, after Microsoft renovated Bing Search with ChatGPT, it reached 100 million users, which is still very low compared to ChatGPT’s 1.5 billion right now. OpenAI was against Microsoft’s move to integrate an incomplete version of GPT-4 into Bing Chat, but Microsoft went ahead with it anyway. For Microsoft, ChatGPT being available only on its Azure Cloud and Bing makes total sense. Dispute has been clearly on for both the companies. Even though ChatGPT is helping a lot of users, it is as good as dead from a company standpoint. Companies have been banning the use of ChatGPT for their employees due to privacy concerns, which just makes the case worse for OpenAI. Due to these data privacy concerns, people have also decided to migrate and build their own LLM-based models leveraging open-source offerings such as Llama 2. The reason that OpenAI is most likely to shut down the free ChatGPT website and not the whole line of GPT products is that people are still using the APIs offered by the company for building their own products, which interestingly has seen an increase in the last two months. But that is also not helping Microsoft in any way. Plus, OpenAI has been in the midst of a lot of legal and ethical troubles for building ChatGPT, and Microsoft wants to distance itself from it because of that. It also shut down its AI classifier for identifying AI generated text because the company admitted that it is not able to do the job. All of this points to the fact that giving ChatGPT away for free is not benefiting OpenAI anymore. It has already acquired all the users that it wanted and now it is time to make money out of it, which only Microsoft can help them do. If Microsoft has plans to acquire OpenAI in the future, it only makes sense for OpenAI to pull the plug on ChatGPT by itself as it does not really benefit them anyway. It has already proven its point and acquired billions of users. Microsoft can just take these services and integrate them exclusively on Azure or Bing Search Engine and monetise it, making everyone happy in the end. Imagine a world where we all go to Bing Chat just to use ChatGPT, and that too in real time, and who knows it might even be better than Google Search? But since ChatGPT is OpenAI’s pet project, it might be hard to give up on it completely.","excerpt":"It makes sense for Microsoft to let OpenAI offer ChatGPT to everyone for free","categories":["AI Features"],"tags":["ChatGPT","GPT-4","OpenAI"],"author_name":"Mohit Pandey","publish_date":"2023-08-17T13:05:57","publication_year":"2023","word_count":1197,"keywords":["Go","ChatGPT","TPU","GPT-5","OpenAI","AI","R","GPT-4","Git","RAG","Azure"],"extracted_tech_keywords":["AI","GPT-5","ChatGPT","OpenAI","RAG","Azure","TPU","R","Go","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/openai-likely-to-pull-the-plug-on-chatgpt\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":65259,"title":"Six Must-Have Skills For An AI Automation Tester In 2020","content":"Artificial intelligence has emerged as a game-changer in the domain of automation testing. These technologies have unveiled new possibilities and methods for automation testers that can help them select the tool and techniques to complete a task more efficiently and effectively. Some of the ways that AI helps in automation testing include reducing the maintenance cost, analysing data to identify defects, effective decision making, organising test sets, and more. In this article, we list – in no particular order – six must-have skills for an AI Automation Tester in 2020: Expert In AI Automation Tools The motive behind every organisation is to achieve the best outcome with lower maintenance costs. With the advent of AI, it has become possible to gain more accurate as well as cost-effective results. Currently, there are several AI frameworks and automation tools present around us, which can be used for automated test generation. Tools such as Testim use artificial intelligence and machine learning to speed-up the authoring, execution, and maintenance of automated tests. Another tool is the Appvance IQ that offers both AI-generated tests and 3rd-generation, codeless scripting. The tool also executes scripts through data-driven functional, performance, app-pen and API testing for both web and mobile apps. More AI Automation, Less Manual Work Manual testing is the process where a developer manually executes test cases without utilising any automation tools. There is no doubt that manual testing is one of the most primetime methods of testing and finding bugs in a software system. However, everything needs to change with time. The evolution in technology is constant, and with this constant evolution, one needs to be faster and efficient in every manner. Manual testing may provide good outcomes, but it requires more human effort and time. This is where AI automation techniques come into play. AI is all about making less human effort and making processes cost-effective. Understanding how machine learning works and which algorithm will play the perfect role will help you save a huge amount of time and provide information on flag areas, current test coverage, etc quickly. Knowledge of Cloud Computing According to reports, the global cloud computing market size is expected to grow from $272.0 billion in 2018 to $623.3 billion by 2023, at a Compound Annual Growth Rate (CAGR) of 18.0% during the forecast period. Cloud computing plays a significant role in emerging technologies, and almost every organisation is shifting towards using cloud platforms to achieve flexibility, abundant resources, scalability, and more. Some of the advantages that cloud computing provides in AI-based automation testing include speed up testing, reduce IT management costs and effort, disaster recovery, more collaboration and more. Sauce Labs is one of the most popular cloud-based test automation tools that leverage machine learning and AI. Stay Updated With the Trends If you want to be ahead of the game, you must have the curiosity to learn new technologies and tools. With the right and updated knowledge of the trends, you can help yourself build a better future. The internet is flooding with resources on AI in automation and testing, and it is the best time to hit some of those courses and enhance your knowledge base. Some of the online resources are mentioned below: Artificial Intelligence for Faster and Smarter UI Testing (Click here)Scope and impact of AI in Automation Testing (Click here) Visual Testing Visualisation is one of the most important topics in the domain of emerging technologies. It has changed the way people used to make, as well as understand decisions. By understanding the techniques of visual testing and implementing AI-based visual validation methods, one can find outcomes which can be missed by simple traditional testing methods. One of the tools is called Applitools. It is an application visual management and AI-powered visual UI testing and monitoring software. This tool helps a tester by providing an end-to-end software testing platform powered by Visual AI. Certifications In AI-Based Automation Courses Certifications are one of the best ways to show that you have the knowledge and skill in a particular domain. There are various certifications available online where you can learn and improve your base. This will not only help you stand out, but also help you achieve your goals in AI Automation testing in a more efficient way. Some of the certifications you can apply for include: Artificial Intelligence (AI) in Software Testing (Click here)A4Q AI and Software Testing Foundation certification (Click here)Artificial Intelligence (AI) in Software Testing by Certybox (Click here)","excerpt":"Artificial intelligence has emerged as a game-changer in the domain of automation testing. These technologies have unveiled new possibilities and methods for automation testers that can help them select the tool and techniques to complete a task more efficiently and effectively. Some of the ways that AI helps in automation testing include reducing the maintenance […]","categories":["AI Features"],"tags":["Automation","automation testing","automation tools","software automation testing"],"author_name":"Ambika Choudhury","publish_date":"2020-05-15T12:00:00","publication_year":"2020","word_count":745,"keywords":["Go","API","artificial intelligence","automation testing","machine learning","AI","cloud computing","Automation","Scala","RAG","software automation testing","automation tools","GAN","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","RAG","cloud computing","R","Go","Scala","API","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/six-must-have-skills-for-an-ai-automation-tester-in-2020\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":23973,"title":"Top Countries Hiring Most Number Of Artificial Intelligence, Machine Learning Experts","content":"Demand for artificial intelligence and machine learning experts is on the rise globally despite a fear that AI will eat away at many jobs. The rate at which global companies are hiring gives us a different picture. The technological advancement is still creating more positions as companies need high-skilled AI talents to develop and maintain a wide range of applications. According to Indeed, employers demand for AI talents has more than doubled over the past three years and the number of job postings as a share of all job postings have increased by 119 percent. The top 10 jobs in demand right now are: Data Scientist Software Engineer Machine Learning Engineer Software Architect Data Analyst Data Warehouse Engineer Full Stack Developers Research Scientist Front End Developer Product Manager AI and machine learning has also strongly gaining ground in India, which is one of the top 15 countries in terms of talent, corporates who use ML, and authors who write about it. Cheaper cost of set up and availability of talent has made India a favourite place for tech giants such as Microsoft, Google and IBM to set up shop. Here’s let’s take a look at the countries that are increasingly hiring more AI talents: United States US has been the leader in attracting AI and machine learning talent. A study by Pysa shows that the top 20 AI companies are spending more than $650 million to hire AI talent and that there are more than 10,000 positions available at top employers across the country. The total annual investment among the 20 employers that are looking to hire AI talents is $33,292,647, which indicates that the future success of the companies heavily depends on AI technologies and the talent to create them. According to Glassdoor and Pysa, the top recruiters in US are Amazon, Google, Microsoft, NVIDIA, Facebook, Intel, Rocket Fuel, General Electric, Cylance, Oculus VR, Booz Allen Hamilton, Huawei, Adobe, Accenture, iRobot, Magic Leap, Rethink Robotics, BAE Systems, HERE, IBM, Samsung, Lenovo, MoTek Technologies, Uber, PCO innovation, Rakuten Marketing, and Wells Fargo. However, none of the reports listed Apple among the top recruiter. The Cupertino-based company comes way down at 96 on the list. This low ranking may be because of the company’s affinity for secrecy which is hurting the iPhone maker in the race to hire AI talent. Recruiters in the US are seeking tech and engineering talents with deep learning, machine learning, AI, neural networks, computer vision and reinforcement skills. Additionally, the country is estimated to have 2,50,000 open data science jobs by 2024. Canada After the US, Canada is gradually becoming a go-to place for AI experts. According to an Indeed report, jobs in Canada requiring AI skills have grown by 1,069 percent since 2013 — a growth rate faster than the UK and US. Companies that are actively hiring for AI skills in Canada are — Royal Bank of Canada, IBM, Scotiabank, KPMG, Amazon, LoyaltyOne, TD Bank, Kinaxis, Huawei and Capital One, among others. Universities across the country are allowing researchers and scientist to lay the groundwork for a breakthrough in AI. The Justin Trudeau-led government invested dramatically in AI research in the past year. Last year, Canada released its national strategy with a plan to invest $125 million in a Pan-Canadian AI Strategy that is focused on increasing the number of researchers and skilled graduates. Additionally, global tech companies such as DeepMind, Microsoft, Facebook and Google among others are also setting up their research labs in the country and tapping Canadian talents to head new AI arms. Europe According to reports, AI-related jobs have tripled over the last three years. The two most sought-after positions by employers are data scientist and machine learning engineers. In the UK, AI jobs pay well above the average salary, data scientist take home £56,385 a year and ML engineers earning an average of £54,617 a year. However, research suggests that the number of AI jobs available in Britain was six times higher than the number of interested candidates. On the other hand, the number of job opportunities is little less in Germany. “When it comes to machinery, Germany has a strong presence of young professionals and academics which has remained sort of hidden and is not always exploited well,” said Yasser Jadidi, head of AI research at the Bosch Center for AI. However, now the company has been thinking to commercialise AI expertise for business. Other European countries like France and Spain too have a significant number of AI experts. In last one year at least 20 development centers have been set up in the countries like Poland, Romania, Bulgaria, Hungary, Austria and Ireland. Nearly 200,000 students graduate in engineering and sciences every year from Universities in Central and Eastern European countries. China Beijing is investing a gargantuan sum of money to make the country No.1 in the application of artificial intelligence. The country is hiring tens and thousands of people to shore up cybersecurity, help censor online content. The salaries that are paid to the top AI talents are within striking distance of those offered in the US. Out of China’s top 10 employers in AI, half are American, claims LinkedIn’s report. The top US tech companies in China hiring AI talents are Intel, IBM and Microsoft. Alibaba, Baidu, Huawei, and Tencent, are among the top Chinese employers in the country. However, China still lags behind in terms of AI position, which has about 50,000 AI position. On the other hand, US has about 80,000 AI position. According to a tech recruitment website, the salaries of the top graduates working on AI range from 300,000 yuan ($47,066) to 600,000 yuan ($94,132) an annum. While the salaries of team leaders with three-to-five years of experience can make more than 1.5 million yuan ($23,533,100) annually. Most of the jobs are in Beijing or Shenzhen. The tech firms are also actively recruiting Chinese students from US college. The job creation is a part of governments drive to become AI superpower and move the Chinese economy up the value chain with emphasis on areas like IT, robotics and energy-saving vehicles. Japan After becoming a world leader in electronics, the country is now focused on building its robotics industry. Foreign software engineers and other information technology specialists are increasingly joining Japan’s robotic startups. Start Today Co, which operates popular online fashion shopping website Zozotown is also hiring tech workers in fields ranging from AI to cryptography and offering annual salaries of as much as 100 million yen ($944,000). The country has roughly 117 active researchers presenting at Neural Information Processing Systems and other leading conferences. India India too is not far behind other countries in terms AI hiring, the country is expected to see a 60 per cent rise by this year due to increasing adoption of automation. The presence of multi-billion dollar IT companies, global software product companies with their research and development bases in India and venture capital industry focused on AI and machine learning has given rise to a talent pool. Further, as India is heading towards Digital India, the IT industry will require 50 percent more workforce equipped with digital skills. “The growing opportunities in the digital technology arena including government initiative like Digital India will add jobs in digital technologies, AI, robotics. The IT industry is expected to add around 1.8-2 lakh jobs this year,” said Alka Dhingra, general manager, IT staffing at TeamLease Services. Most of the machine learning talent pool is spread across five cities: Bangalore, Chennai, Hyderabad, Mumbai\/Pune belt and Delhi NCR region.","excerpt":"Demand for artificial intelligence and machine learning experts is on the rise globally despite a fear that AI will eat away at many jobs. The rate at which global companies are hiring gives us a different picture. The technological advancement is still creating more positions as companies need high-skilled AI talents to develop and maintain […]","categories":["AI Hirings"],"tags":["AI Jobs","Machine Learning","machine learning jobs"],"author_name":"Smita Sinha","publish_date":"2018-04-25T05:37:22","publication_year":"2018","word_count":1258,"keywords":["machine learning jobs","data science","artificial intelligence","machine learning","AI","neural network","ML","AI Jobs","Machine Learning","computer vision","RAG","Aim","deep learning"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","neural network","computer vision","data science","Aim","RAG"],"url":"https:\/\/analyticsindiamag.com\/ai-hiring\/top-countries-hiring-most-number-of-artificial-intelligence-machine-learning-experts\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10122402,"title":"Snowflake, NVIDIA Join Forces to Enhance Custom AI Applications","content":"Snowflake announced its collaboration with NVIDIA during the Snowflake Summit 2024. This partnership aims to empower customers and partners to develop bespoke AI data applications within Snowflake, leveraging NVIDIA’s AI technology. This collaboration sees Snowflake integrating NVIDIA AI Enterprise software, incorporating NeMo Retriever microservices into Snowflake Cortex AI, Snowflake’s managed LLM and vector search service. This integration allows organisations to link custom models to varied business data, delivering precise responses seamlessly. Additionally, Snowflake Arctic, an open, enterprise-grade LLM, now supports NVIDIA TensorRT-LLM software, enhancing performance. Arctic is also accessible as an NVIDIA NIM inference microservice, broadening developers’ access to its capabilities. As enterprises strive to maximise AI’s potential, the need for data-driven customisation grows. The Snowflake-NVIDIA collaboration facilitates rapid development of specific AI solutions, benefiting businesses across various sectors. “Pairing NVIDIA’s full stack accelerated computing and software with Snowflake’s state-of-the-art AI capabilities in Cortex AI is game-changing,” stated Sridhar Ramaswamy, CEO of Snowflake. “Together, we are unlocking a new era of AI where customers from every industry and every skill level can build custom AI applications on their enterprise data with ease, efficiency, and trust.” “Data is the essential raw material of the AI industrial revolution,” said Jensen Huang, founder and CEO of NVIDIA. “Together, NVIDIA and Snowflake will help enterprises refine their proprietary business data and transform it into valuable generative AI.” Notable NVIDIA AI Enterprise software capabilities offered in Cortex AI include: NVIDIA NeMo Retriever: Provides accurate and high-performance information retrieval for enterprises. NVIDIA Triton Inference Server: Facilitates the deployment, running, and scaling of AI inference for various applications on any platform. NVIDIA NIM inference microservices, part of NVIDIA AI Enterprise, can be deployed within Snowflake as a native app using Snowpark Container Services. This setup allows organisations to deploy foundational models directly within Snowflake easily. Quantiphi, an AI-first digital engineering firm and ‘Elite’ partner with both Snowflake and NVIDIA, exemplifies this innovation. Quantiphi’s Snowflake Native Apps, baioniq and Dociphi, are designed to enhance productivity and document processing within specific industries. These apps, developed using the NVIDIA NeMo framework, will be available on Snowflake Marketplace. The Snowflake Arctic LLM, launched in April 2024 and trained on NVIDIA H100 Tensor Core GPUs, is now available as an NVIDIA NIM. This makes Arctic accessible in seconds, either via the NVIDIA API catalogue with free credits or as a downloadable NIM, offering flexible deployment options. Earlier this year, Snowflake and NVIDIA expanded their collaboration to create a unified AI infrastructure and compute platform in the AI Data Cloud. Today’s announcements mark significant advancements in their joint mission to help customers excel in their AI initiatives.","excerpt":"Arctic is also accessible as an NVIDIA NIM inference microservice, broadening developers’ access to its capabilities.","categories":["AI News"],"tags":["NVIDIA","Snowflake"],"author_name":"Mohit Pandey","publish_date":"2024-06-04T09:07:36","publication_year":"2024","word_count":434,"keywords":["AI","ML","Git","RAG","microservices","Aim","generative AI","Rust","NVIDIA","R","Snowflake"],"extracted_tech_keywords":["AI","ML","generative AI","Aim","RAG","microservices","Snowflake","R","Rust","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/snowflake-nvidia-join-forces-to-enhance-custom-ai-applications\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10064727,"title":"Microsoft launches TypeScript 4.7 Beta","content":"Microsoft has announced the beta release of TypeScript 4.7. The  key features of the latest version of the open-source programming language include: ECMAScript Module Support in Node.js: The researchers had been struggling for years to implement module support in Node.js. TypeScript 4.7 ends this struggle as it adds this functionality with two new module settings: node12 and nodenext. Control over Module Detection: TypeScript 4.7 introduces a new option called ‘moduleDetection’ for Javascript code. Control-Flow Analysis for Computed Properties: TypeScript 4.7 now analyses the type of computed properties and narrows them correctly. The new version knows that obj[key] is a string. Meaning, TypeScript can correctly check that computed properties are initialised by the end of a constructor body. Instantiation Expressions: TypeScript 4.7 can now take functions and constructors and feed them type arguments directly. With this, you can specialise makeBox to accept more specific types and reject anything else. This logic also works for constructor functions such as Array, Map, and Set.Optional Variance Annotations for Type Parameters: In this version, variance on type parameters can be explicitly specified. Resolution Customization with moduleSuffixes: TypeScript 4.7 now supports a moduleSuffixes option to customise how module specifiers are looked up. This feature can be useful for React Native projects where each target platform can use a separate tsconfig.json with differing moduleSuffixes.Resolution-mode: TypeScript now allows \/\/\/ <reference types=”…” \/> directives and import type statements to specify a resolution strategy.Groups-Aware Organize Imports: TypeScript has an Organize Imports editor feature for both JavaScript and TypeScript. In the coming weeks, the TypeScript team will be polishing version 4.7 to get it ready for a Release Candidate.","excerpt":"In the coming weeks, the TypeScript team will be polishing version 4.7 to get it ready for a Release Candidate.","categories":["AI News"],"tags":["TypeScript"],"author_name":"Kartik Wali","publish_date":"2022-04-11T18:49:59","publication_year":"2022","word_count":268,"keywords":["programming_languages:R","programming_languages:Java","TypeScript","Ray","JavaScript","GAN","R","Java","programming_languages:JavaScript"],"extracted_tech_keywords":["Ray","R","JavaScript","TypeScript","Java","GAN","programming_languages:R","programming_languages:JavaScript","programming_languages:Java"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/microsoft-launches-typescript-4-7-beta\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":24271,"title":"Tabrez Khan Of MathWorks Discusses The Advent Of 5G In India &#038; Its Impact On Data Analytics","content":"The growing trend of artificial intelligence and machine learning is paving the way for 5G network. Industry insiders believe that 5G will carve out a way for the next generation robots or humanoids, which would be controlled wirelessly and also enable data storage resources of the cloud. But not only the end user, but from the point of view of an organisation, the dawn of 5G internet means a sure evolution of their integrated analytical systems as well as organisational structure. In fact, it will be expensive to deploy 5G infrastructure and will also require a fresh business models and regulation, claims a WEF report. Studies have suggested that 5G will be an overlay to existing 4G and will start achieving meaningful scale and reach by early 2020s. In this article we will try and find out if India is ready for the 5G revolution. What Is 5G Interestingly enough, 5G or the fifth generation of mobile network has not officially defined. All that is being talked of this this: It is faster than 4G LTE, more stable, and more versatile. It’s speed is estimated to be more than 10Gbps, which is also an indication that 5G may entirely replace home and office WiFi. Marc Tracey, a spokesman of Verizon told Wired, “Basically, 5G will provide a wider pipeline and faster lanes.” Requirements for 5G: Speaking at the event called MATLAB Expo 2018, Tabrez Khan, senior application engineer at MathWorks India, explained that the three important things that are required for 5G applications: Enhanced mobile broadband (more than 10 Gbps) Ultra reliable and low latency (less than 1 ms) Massive machine-type communication (more than 1e5 devices) When Will It Be Launched One of the key challenges faced by telecom players all over the world is scaling up the supporting 5G infrastructure. It is predicted to be expensive to deploy and will also require fresh thinking on financing, business models, and regulation. Who will pay for 5G infrastructure, how broad will the infrastructure’s reach be, can it be commonly owned, and is there scope for effective public-private partnerships in the build out and operations? As companies are still deliberating on all these topics, the 5G network is slated to exist as an overlay to existing 4G networks, and will start achieving meaningful scale and reach by the early 2020s. Use Cases For 5G: Khan also added that broadly, the four areas where a user could look forward to changes in the way they consumed data would be in the following areas: Videos in 4K, 8K and 360 degree formats Virtual reality-based content Connected vehicles and other modes of transport Internet of things How Will 5G Affect Data Analytics From the user’s perspective, 5G merely means fast internet. But from an enterprise point of view, the 5G internet means a slow but sure overhaul of their integrated analytical systems as well as organisational structure. Many experts have claimed that the four key focus areas for large companies to leverage data after 5G are: Leveraging of analytics tools to derive value Integration between Cloud and enterprises Internet of things Enterprise operational technology These can be achieved only after a way is found to manage gigantic amount of unstructured data. A Market Research report says, “The convergence of cloud, data management, IoT platforms and solutions will enabling the next evolution of data analytics in which enterprises will realise the significant tangible and intangible benefits from data generated via 5G. The ability to sort data in a raw format, store it in different structural formats, and subsequently release it for further analytics, will be of paramount importance for all industry verticals.” Preparation For 5G Data Handling At the recently-concluded MATLAB Expo 2018, Khan informed that MathWorks was well-equipped to train engineers as well as students for the onslaught of large quantity of data generated in the coming years. He said that MathWorks had already made the following provisions (among others) in their software: Signal generation and analysis reference measurement channels 5G new radio and the 5G library LTE system toolbox and 5G library Wireless standards simulation New physical layer in Release 15 Baseband DSP for large bandwidths Efficient channel coding methods Model channel and interference 5G link level simulation RF power amplifier (PA) linearisation Is India Ready For 5G Revolution? The Indian government is set to see that there is no spectrum shortage to ensure India does not lag behind the rest of the world in launching 5G services, Telecom Minister Manoj Sinha had said recently. He had also said that the new telecom policy would be suitably aligned to ensure India is 5G ready and in sync with the world telecom. “The government is betting big on 5G as its is expected to deliver new levels of performance and efficiency and offer  opportunities for myriad apps and business model that could potentially lead to better healthcare, smart cities, more efficient manufacturing and enhanced lifestyles, he added. Though India is readying itself to bring 5G network by 2020 to keep pace with the global technology adoption, it lacks 100 percent fiber coverage, investment in fiberisation and basic infrastructure. The last-mile fiber coverage in India is only about 25 to 30 percent and the investment in fiber infrastructure is only 10 percent of overall deployment cost.","excerpt":"The growing trend of artificial intelligence and machine learning is paving the way for 5G network. Industry insiders believe that 5G will carve out a way for the next generation robots or humanoids, which would be controlled wirelessly and also enable data storage resources of the cloud. But not only the end user, but from […]","categories":["AI Features"],"tags":["5G","Data Analytics","Interviews and Discussions","Mathworks","WEF","WiFi"],"author_name":"Prajakta Hebbar","publish_date":"2018-05-04T09:43:55","publication_year":"2018","word_count":880,"keywords":["Go","5G","artificial intelligence","WiFi","machine learning","AI","programming_languages:R","WEF","RAG","Mathworks","GAN","Aim","analytics","Data Analytics","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","analytics","Aim","RAG","R","Go","GAN","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/tabrez-khan-mathworks-5g-data-analytics\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10164863,"title":"Physical Intelligence Launches ‘Hi Robot’, Helps Robots Think Through Actions","content":"Researchers at Physical Intelligence, an AI robotics company, have developed a system called the Hierarchical Interactive Robot (Hi Robot). This system enables robots to process complex instructions and feedback using vision-language models (VLMs) in a hierarchical structure. Vision-language models can control robots, but what if the prompt is too complex for the robot to follow directly?We developed a way to get robots to “think through” complex instructions, feedback, and interjections. We call it the Hierarchical Interactive Robot (Hi Robot). pic.twitter.com\/KdL5myyybT— Physical Intelligence (@physical_int) February 26, 2025 The system allows robots to break down intricate tasks into simpler steps, similar to how humans reason through complex problems using Daniel Kahneman’s ‘System 1’ and ‘System 2’ approaches. In this context, Hi Robot uses a high-level VLM to reason through complex prompts and a low-level VLM to execute actions. Testing and Training Using Synthetic Data Researchers used synthetic data to train robots to follow complex instructions. Relying solely on real-life examples and atomic commands wasn’t enough to teach robots to handle multi-step tasks. To address this, they created synthetic datasets by pairing robot observations with hypothetical scenarios and human feedback. This approach helps the model learn how to interpret and respond to complex commands. It outdid other methods, including GPT-4o and a flat Very Large Array (VLA) policy, by better following instructions and adapting to real-time corrections. It achieves a 40% higher instruction-following accuracy than GPT-4o. Hence, it demonstrates better alignment with user prompts and real-time observations. Source: Official blog In real-world tests, Hi Robot performed tasks like clearing tables, making sandwiches, and grocery shopping. It effectively handled multi-stage instructions, adapted to real-time corrections, and respected constraints. Synthetic data, in this context, highlights potential in robotics to efficiently simulate diverse scenarios, reducing the need for extensive real-world data collection. Hi Robot ‘Talks to Itself’ As seen in an example below, a robot is trained to clean a table by disposing of trash and placing dishes in a bin. It can be directed to follow more intricate commands through Hi Robot. Source: Official blog This system allows the robot to reason through modified commands provided in natural language, enabling it to “talk to itself” as it performs tasks. Moreover, Hi Robot can interpret user contextual comments, incorporating real-time feedback into its actions, such as handling complex prompts. This setup allows the robot to incorporate real-time feedback, such as when a user says “that’s not trash”, and adjust its actions accordingly. The system has been tested on various robotic platforms, including single-arm, dual-arm, and mobile robots, performing tasks like cleaning tables and making sandwiches. “Can we get our robots to ‘think’ the same way, with a little ‘voice’ that tells them what to do when presented with a complex task?” the researchers said in the company’s official blog. This advancement could lead to more intuitive and flexible robot capabilities in real-world applications. Researchers plan to refine the system in the future by combining the high-level and low-level models, allowing for more adaptive processing of complex tasks.","excerpt":"The system allows the robot to take real-time feedback in natural language, and “talk to itself” as it performs tasks.","categories":["AI News"],"tags":["Physical AI","Robotics"],"author_name":"Sanjana Gupta","publish_date":"2025-02-28T15:25:51","publication_year":"2025","word_count":501,"keywords":["synthetic data","Physical AI","AI","programming_languages:R","GPT-4o","emerging_tech:synthetic data","Robotics","GPT","Ray","ai_applications:robotics","R","llm_models:GPT"],"extracted_tech_keywords":["AI","GPT-4o","Ray","R","GPT","synthetic data","llm_models:GPT","programming_languages:R","ai_applications:robotics","emerging_tech:synthetic data"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/physical-intelligence-launches-hi-robot-helps-robots-think-through-actions\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10103921,"title":"HPE, NVIDIA Collaborate for Generative AI on Edge and Cloud","content":"Hewlett Packard Enterprise (HPE) announced an extensive collaboration with NVIDIA to introduce an enterprise computing solution designed for generative AI. This collaboration expands HPE’s suite of purpose-built, AI-native offerings, providing enterprises with an unprecedented opportunity to accelerate AI model training, tuning, and inferencing. The co-engineered solution simplifies the customisation of foundation models using private data and facilitates the deployment of production applications across diverse environments, from edge to cloud. This offers a seamless full-stack AI tuning and inferencing solution jointly developed by HPE and NVIDIA. The new enterprise computing solution for generative AI is part of an expanded collaboration between HPE and NVIDIA, providing full-stack, out-of-the-box AI solutions. These solutions integrate HPE Machine Learning Development Environment, HPE Ezmeral Software, HPE ProLiant Compute, and HPE Cray Supercomputers with the NVIDIA AI Enterprise software suite, including the powerful NVIDIA NeMo framework. Antonio Neri, president and CEO of HPE, emphasised the importance of this collaboration, stating, “Together, HPE and NVIDIA are in a unique position to deliver a comprehensive AI-native solution that will dramatically ease the journey to develop and deploy AI models with a portfolio of pre-configured solutions.” Jensen Huang, founder and CEO of NVIDIA, added, “Our expanded collaboration with HPE will help enterprises drive unprecedented productivity through AI applications that connect with business data to power accurate assistants, informed chatbots, and semantic search.” Purpose-built and optimised for AI: The solution features a rack-scale architecture with market-leading HPE ProLiant Compute DL380a pre-configured with NVIDIA L40S GPUs, NVIDIA BlueField-3 DPUs, and the NVIDIA Spectrum-X Ethernet Networking Platform for hyperscale AI. Read: NVIDIA RTX Brings Alan Wake 2 to Life This also includes HPE Machine Learning Development Environment with new generative AI studio capabilities for rapid prototyping and testing, and HPE Ezmeral Software with new GPU-aware capabilities to simplify deployment and accelerate data preparation for AI workloads across the hybrid cloud. By utilising NVIDIA AI Enterprise to accelerate production AI development and deployment with security, stability, manageability, and support. This software suite offers the NVIDIA NeMo framework, guardrailing toolkits, data curation tools, and pretrained models to streamline enterprise solutions. In addition to the collaboration, HPE Services now offers a broad portfolio of consulting services, workforce training, and deployment solutions to guide enterprises through every step of their AI journey. The comprehensive services, supported by new Global Centers of Excellence for AI and Data, are now open in Spain, the United States, Bulgaria, India, and Tunisia.","excerpt":"A seamless full-stack AI tuning and inferencing solution jointly developed by HPE and NVIDIA.","categories":["AI News"],"tags":["HPE","HPE Hewlett Packard Enterprise","NVIDIA"],"author_name":"Mohit Pandey","publish_date":"2023-11-30T15:33:32","publication_year":"2023","word_count":401,"keywords":["semantic search","API","machine learning","AI","chatbots","ML","Ray","NVIDIA","generative AI","HPE Hewlett Packard Enterprise","foundation models","HPE","R"],"extracted_tech_keywords":["AI","machine learning","ML","generative AI","foundation models","Ray","chatbots","semantic search","R","API"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/hpe-nvidia-collaborate-for-generative-ai-on-edge-and-cloud\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":32070,"title":"Facebook Developing A Cryptocurrency For WhatsApp Transfers: Reports","content":"Between its myriad of privacy disrepute, Facebook is reportedly launching its private cryptocurrency. According to a report, Facebook is expanding a cryptocurrency that would allow WhatsApp users to transfer money. The report reveals that the WhatsApp feature will primarily focus “on the remittances market in India.” The currency will reportedly be called Stablecoin and will be secured to the US dollar to minimise fluctuating. It’s unclear when the innovation might go live, but reports suggest that Facebook is still a long way from being ready to launch the cryptocurrency. The company is currently in the process of improving its overall strategy as it correlates to safeguarding assets and other features that go along with a cryptocurrency launch. Facebook has been contemplating to extend its role in finance for a while now. In 2014, Facebook appointed former PayPal president David Marcus as the head of their Messenger. But in May this year, Marcus became the head of the blockchain initiatives team who is introducing these financial services to both its original application as well as the Messenger. The social media giant has gripped the details of the project below covers but has been hiring people for its blockchain group. Facebook now has about 40 team members in the blockchain assembly, going by employee indications on LinkedIn. Facebook has over 2.5 billion users encompassing the globe and makes $40 billion in annual revenues. Given its size, global acceptance and participation in handling regulation, the company has a good chance of launching Stablecoin that sees extensive approval. It would also grow as the first of the tech giants to begin such a project. Facebook’s relevance in India has recently been damaged in controversy, as WhatsApp was used to spread false news in the country, some of which led to violence. Facebook believes that India is a market with a huge opportunity, particularly considering that at present there are 480 million internet users in India. The figure is anticipated to increase to 737 million by 2022 according to Forrester Research Inc.","excerpt":"Between its myriad of privacy disrepute, Facebook is reportedly launching its private cryptocurrency. According to a report, Facebook is expanding a cryptocurrency that would allow WhatsApp users to transfer money. The report reveals that the WhatsApp feature will primarily focus “on the remittances market in India.” The currency will reportedly be called Stablecoin and will […]","categories":["AI News"],"tags":["Cryptocurrency","facebook messenger","whatsapp"],"author_name":"Bharat Adibhatla","publish_date":"2018-12-24T11:36:02","publication_year":"2018","word_count":337,"keywords":["Go","programming_languages:R","AI","Cryptocurrency","innovation","programming_languages:Go","whatsapp","facebook messenger","R"],"extracted_tech_keywords":["AI","R","Go","innovation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/facebook-developing-a-cryptocurrency-for-whatsapp-transfers-reports\/","complexity_score":2,"technical_depth":6,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":23998,"title":"Analysing How Copyright Laws May Challenge Status Quo Of AI-Generated Content","content":"Can work produced by machine learning software create issues under copyright laws? Can copyright laws be extended to computer-generated work as well? Benjamin Sobel, a law and technology researcher from Harvard, has discussed this quandary in a research paper titled Artificial Intelligence’s Fair Use Crisis. He has said that by ingesting data from multiple resources, AI programs can now create music, write novels and even edit and make movies. That is how Deep Mind’s Alpha Go won over reigning (human) champion Lee Sedol — by ingesting data from 1,60,000 amateur games. AI programs learn from a vast body of data generated by humans to come up with the optimal result. Then again, as Sobel notes in his paper, machine learning is an artificial intelligence technology with immense potential and a vast appetite for copyrighted works. However, can a machine be awarded intellectual property rights? Given the pace at which AI systems are generating unpredictable outputs by ingesting vast amount of data, it is difficult to attribute who is the creator of the final work. Past rulings governing IP laws in the US and Germany deem that only works created by humans come under the purview of copyright laws. Does this indicate that works created by AI fall outside copyright laws because they are not created by humans? Given that AI systems would be governed by multiple players, determining ownership of AI-produced content is tricky. According to the guidelines set by the US Copyright Office, ‘non-humans’ are not natural persons and cannot be held legally responsible in a court of law. Hence, they can’t be considered ‘authors’. Can Work Produced By AI Be Assigned Authorship? This paper, authored by K Hristov questions IP laws in the field of AI. According to Hristov, when it comes to assigning ownership to AI-generated works, one can consider three main players: AI programmers Owners, that is, large companies and financial investors in the AI sector end users The author also brings another perspective into view — when awarding ownership of AI-generated content, how would society benefit most? If it is awarded to the programmer or the end user or the investor? To better gauge the societal impact of each party, we must first determine the ultimate goal of assigning copyright of AI-generated works to human authors. Another important point is that since non-humans can’t be held responsible in court, ownership should be assigned to humans. Therefore, providing ownership rights to AI developers or end users can be useful for the growth of the sector. Hristov further observes that financial incentives should be awarded to those who make the greatest contribution to the development of AI. This would thus foster innovation and spark a renewed interest in research and development of AI. Can Legal Standing Affect The Future Development Of AI? Indian IP attorney Anmol Khurana of Khurana and Khurana law firm noted in an article that there are only two ways in which copyright can be bestowed to AI — first, it can be completely ruled out since there is no human involvement. Or second, it can be attributed to the developer. Going forward, copyright laws will have a bearing on the development of AI. Khurana observes that with the usage of AI increasing daily, machines will become better at artistic work and the lines between human and machine-generated content will blur. Dr Paul Lambert, adjunct lecturer and author, observes that as technology becomes increasingly commercialised, AI ownership should be governed either on a case-to-case basis or grouped by category. Even the data that machines ingest is subject to IP rules. Copyright issues would play a huge role in governing robots and drones – that operate semi-autonomously, especially in areas of surveillance where drones collect pictures. Would these activities carried out by autonomous agents come under the purview of copyright laws? Drones are being widely deployed in law enforcement, domestic services such as cargo delivery, emergency operations and humanitarian operations and are subject to regulation and licensing. They are directly controlled by an operator but gradually there is a rise in autonomous flying and sub-aquatic drones which are also used in surveillance for collecting data. These drones do not have a human — in the loop which begs the question, would the activities and decisions taken by these drones have any legal bearing. Given the vast amount of resources being plowed into the development of drones, robots and AI research and commercialisation, it is time enterprises devote energy to understand whether these activities would be protected under copyright laws. Outlook Going forward, there needs to be stricter regulations regarding machine-generated work. As robots gear up to transform the economy and the usage of drones increases, enterprises and governments should tackle the issues related to copyright laws and settle ownership of AI-generated content. On the other hand, AI is wending its way into legal departments and is affecting the future of law. India is seeing a startup boom in legal tech with vertical AI startups such as Casemine and Pensieve operating in this area. Perhaps, the issue of IP and copyright laws will be addressed by legal tech companies working on NLP and AI.","excerpt":"Can work produced by machine learning software create issues under copyright laws? Can copyright laws be extended to computer-generated work as well? Benjamin Sobel, a law and technology researcher from Harvard, has discussed this quandary in a research paper titled Artificial Intelligence’s Fair Use Crisis. He has said that by ingesting data from multiple resources, […]","categories":["IT Services"],"tags":[],"author_name":"Richa Bhatia","publish_date":"2018-04-25T09:44:28","publication_year":"2018","word_count":858,"keywords":["Go","machine learning","artificial intelligence","autonomous agents","AI","AWS","TPU","BERT","NLP","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","NLP","autonomous agents","AWS","TPU","R","Go","BERT"],"url":"https:\/\/analyticsindiamag.com\/it-services\/analysing-how-copyright-laws-may-challenge-status-quo-of-ai-generated-content\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":24646,"title":"Has AI Really Influenced Indian Lives? Here’s The Real Picture","content":"Artificial Intelligence has begun to touch our lives in one way or the other. The day is not far when smart devices will know us more than our friends, families and maybe our own selves. We are witnessing significant activity in the AI space with tech giants such as Google, Amazon, Apple and Microsoft stepping up their offerings. Nearly half a million Indians have access to smartphones but not many know how to make the most of it or realise that it can be a tool that can be used to increase their intelligence and knowledge base. According to a report, there are about 650 million mobile phone users in India and just over 300 million of them have a smartphone. India Has Failed To Unlock The True AI Potential From Their Phones The main hurdle begins the software that these smartphone runs are largely tailored for Western users and not for Indian masses. For instance, Google Assistant, the most popular AI-powered voice assistant among Android users in India, supports only two languages — Hindi and English, out of 122 recognised Indian languages. According to a study, nearly 70 percent of Indians consider local languages digital content more reliable than English content. Therefore, in order to deepen the user base, the tech giants have to integrate more Indian regional languages into their AI-powered voice assistants also we need a huge amount of content in vernaculars for mass adoption of voice as a prefered user interface. It is not that the tech giants are don’t want to support Indian regional languages. It is just that they can’t because most of the AI-powered technologies need colossal amounts of annotated data to train themselves. According to a report, speakers of India’s regional languages don’t seem to generate enough of annotated data. The Stumbling Block For Startups To train a machine a company needs a large scale of dataset and Indian startups lack access to the large dataset. There are a handful of companies in India that have access to big data. India offers the same advantages as China in terms of data (over 99 percent of Indians aged 18 and above have Aadhaar card)  but the major drawback is that India doesn’t have its own tech giants like Baidu or Google which can use the data to build a credible AI system. “India would definitely need to think a lot about data privacy and open data platforms,” Vinay Kumar, co-founder of Arya.ai said in a report. Autonomous Cars? Not yet In India, the poster boy of artificial intelligence is a distant dream for Indians. Right now the Indian government is more focused on bringing electric vehicles on road. Moreover, none of the states governments has clear laws for such vehicles. Even the current BJP-led NDA  government is not keen on bringing them in India over fears that it could take away jobs. Apart from lack of support from the government, there’s been one major hurdle to self-driving vehicles in India – a legion of lawbreakers, from rash drivers and jaywalkers to cattle. And there are very few companies in India that are testing autonomous vehicles in India. Former Uber CEO Travis Kalanick had also joked about the same, saying, “India will be the last one to get autonomous cars! Have you see the way people drive here?” Lack Of AI Researchers India researchers are among the best in the world. However, the country needs to aggressively scale up research in AI. Even though we have made a modest beginning and have allocated $480 million to promote AI, machine learning and IoT this year, India badly lags behind its neighbouring country China. For instance, when it comes to Research and Development, Chinese government sends 2 percent of its GDP on research while India spends a meagre 0.6 percent — compared to the size of its economy. The engineering talent in India is largely focused on IT, not research and innovation. Reports show that that, out of 129 deemed universities, 67 public institutions, 700 degree-granting institutions, 35,539 affiliated colleges, there are only 15 universities that contribute to almost 42 percent of all research publications. IIT Kharagpur, which is known as the research hub of the Indian IT sector, contributes to just 2.86 percent of research publication. This shows that India needs massive upgradation programmes in new technologies. As rightly pointed out by NITI Aayog CEO Amitabh Kant that IITs and IIITs must redefine themselves as institute driving cutting-edge technologies for the fourth industrial revolution. Gender-Bias And Cultural Differences There is a large gender disparity too. A report claims, internet user market is still a male preserve in India. There are estimated 143 million female internet users overall, which is approximately 30 percent of total internet users. Reports have also shown that at 46 percent, India’s gender gap in mobile phone ownership is also quite large. Cultural differences between populations within India also becomes a barrier for AI reach. For example, content that is popular among people from states with larger populations dominates over content that is relevant to less populated states. The large disparities in wages, education and content tend to heavily biased towards male users. Gender biases continue to seeping through algorithms, it learns from the data it is provided. Lessons To Be Learned As Indian is a land of culture, languages, the onus lies on Indian developers and companies and startups to make the smart tools and services in India and make it accessible to the masses. Foreign companies are unlikely to meet or even understand the true needs of Indians. The story of China’s internet giant Baidu holds the most important lesson for India, which has successfully developed an AI-powered search engine on its own. China has recognized the importance of cultural infrastructure, which bridges the gap between the lab and the market. India should not only recognize the importance of the cultural institutions required for such innovation but also requires an ecosystem of high-tech product industry that focuses on research and innovation, which will strength our AI talent pipeline and help India take the lead in cutting-edge AI innovation. On The Brighter Side There are a number of perspectives for which India is behind in AI but there are positive things too. India has started investing in research, education, training and retraining, in solution using AI and machine learning. Infosys, TCS and Wipro are training employees in AI. For example, Infosys has trained nearly 3,000 people and TCS has trained 2,10,000 employees in AI. Recently, Google partnered with NITI Aayog to work on a range of initiatives including training and incubating Indian startups focused on AI. Moreover, there are many startups that have already embedded AI in their technology stack and starting to supply AI for solving issues that are typical to India. The positive sign is India has the second largest number of  IT professionals in the world, we are going to catch up soon, claims Kris Gopalakrishnan, chairman at Axilor Ventures, and co-founder of Infosys. Through a series of events and speeches from earlier this year, Prime Minister Narendra Modi is seen to have been deliberately showcasing India as well as his government as technologically forward. According to a report, the prime minister wants NITI Aayog to adopt a comprehensive strategy for the commercialisation of AI so that the nation collectively benefits from the technology.","excerpt":"Artificial Intelligence has begun to touch our lives in one way or the other. The day is not far when smart devices will know us more than our friends, families and maybe our own selves. We are witnessing significant activity in the AI space with tech giants such as Google, Amazon, Apple and Microsoft stepping […]","categories":["IT Services"],"tags":["AI (Artificial Intelligence)","China","Indian IT","NITI Aayog","Smartphone"],"author_name":"Smita Sinha","publish_date":"2018-05-16T07:18:19","publication_year":"2018","word_count":1226,"keywords":["Go","artificial intelligence","machine learning","AWS","Smartphone","AI (Artificial Intelligence)","AI","Git","RAG","Aim","edge AI","NITI Aayog","Indian IT","R","China"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","Aim","edge AI","RAG","AWS","R","Go","Git"],"url":"https:\/\/analyticsindiamag.com\/it-services\/has-ai-really-influenced-indian-lives-heres-the-real-picture\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":7361,"title":"Analytics: No Pain, No Gain","content":"We did an analytics exercise for a US client recently in education domain that had all the flavors of roadblocks one can encounter on venturing into analytics territory. I intend to summarize those here along with solutions we found in collaboration with all stakeholders This one is most worrisome and showstopper pretty much everywhere.  Data Quality and Data Integrity!!! Many of the fields had inconsistent data, nearly 50%. Ones that had data were ambushed by missing values. At the same time data integrity was questioned. Different departments came out with  different numbers for same metrics.That’s terrifying indeed! We reconciled data across different systems. Though we never had exact match however we narrowed down differences to less than 10 %. All agreed to a 10% tolerance level. Good to go! Same metrics, different understanding! How do we define a conversion etc?  This wasn’t anticipated to be honest but became clear after couple of stakeholder meetings. Good part we had every stakeholder from finance, sales, marketing and operations team on board. We were able to identify and close understanding gaps quickly. Excitement and Expectations! Kicking of analytics initiatives generated quite an excitement for all. Everyone wanted a pie of magic (isn’t that expect out of analytics these days) at the end. At this stage it was critical to set priorities right.  Even though we started with defined set of activities, things changed in weeks time when client presented alternative priorities.  Client’s executive team played an important role here in clearly defining what their priorities were what they really wanted and how they will act upon the insights we deliver. I would say clear understanding and desire on client’s part won the battle here. Take Aways Define your data quality tolerance limits. You never will have 100% correct data. However this shouldn’t stop you from acting. Having all stakeholders who are linked to the project on board. This may be executive team, sales, marketing or operations team. This is where you can identify inconsistent understanding and close gaps. Be clear and precise about what you want and how you plan to act upon insights. This was just a month’s exercise. Surely we will hit many such scenarios ahead.","excerpt":"We did an analytics exercise for a US client recently in education domain that had all the flavors of roadblocks one can encounter on venturing into analytics territory. I intend to summarize those here along with solutions we found in collaboration with all stakeholders This one is most worrisome and showstopper pretty much everywhere.  Data […]","categories":[],"tags":[],"author_name":"Vikas Kamra","publish_date":"2015-04-30T07:36:08","publication_year":"2015","word_count":365,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","ViT","data quality","analytics","R"],"extracted_tech_keywords":["AI","analytics","R","Go","data quality","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/analytics-no-pain-no-gain\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10164051,"title":"Gurugram-Based Spyne Secures $16 Million in Series A Funding to Expand U.S. Operations","content":"Spyne, a deep tech startup based in Gurugram, has raised $16 million in a Series A funding round led by Vertex Ventures. Over the next few years, the company will expand from 1,500 to over 20,000 dealer rooftops in the U.S. Existing investors Accel, Storm Ventures, and Alteria Capital participated in this round. In his LinkedIn post, co-founder and CEO Sanjay Varnwal highlighted the company’s evolution from a visual merchandising platform to a comprehensive automotive retail solution powered by GenAI. Spyne has demonstrated consistent performance with a substantial revenue figure of $18.4 million. “We’ve seen a 5X surge in revenue in the last 15 months, and we’re aiming for a 3-4X revenue leap this year,” Varnwal mentioned further. This comes after the Indian government announced the Deep Tech Fund of Funds to catalyse the next generation of Indian startups. The company plans to develop a GenAI-powered unified automotive retail solution that will streamline the retail journey for dealerships. Varnwal stated, “This investment marks a pivotal moment for Spyne as we accelerate our U.S. expansion and push the boundaries of what AI can do for automotive retail.” The new capital will enable Spyne to build a dedicated U.S. business development team and strengthen its presence in high-growth regions across EMEA and APAC, the company said. Piyush Kharbanda, general partner at Vertex Ventures, noted that Spyne addresses fundamental pain points for auto dealerships by leveraging AI to optimise vehicle merchandising. In the past month, many deep tech startups have backed funding for their operations. Bengaluru-based Dynolt Technologies bagged $1.7 million in its seed funding round, led by Transition VC, with participation from marquee angel investor Yashowardhan Shah. Additionally, Astrome Technologies secured $10 million in a funding round led by Apollo Fund, with additional investments from the IAN Group.","excerpt":"The round was led by Vertex Ventures, with participation from existing investors like Accel, Storm Ventures, and Alteria Capital.","categories":["AI News"],"tags":["AI startup","deeptech","Funding","Spyne"],"author_name":"Sanjana Gupta","publish_date":"2025-02-18T18:18:12","publication_year":"2025","word_count":296,"keywords":["AI startup","Go","Funding","GenAI","API","funding","AI","deeptech","ML","RAG","Spyne","Aim","R","startup"],"extracted_tech_keywords":["AI","ML","GenAI","Aim","RAG","R","Go","API","startup","funding"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/gurugram-based-spyne-secures-16-million-in-series-a-funding-to-expand-u-s-operations\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":14595,"title":"Flytxt raises Rs 70 crore to scale-up product adoption","content":"Dr Vinod Vasudevan, CEO, Flytxt Netherlands-based Flytxt customer data analytics Software Company announced that it has aised Rs. 70 crore funding from DAH Beteiligungs GmbH. Flytxt plans to invest new funds in R&D to continue evolving its software product to further its growth in the Telecom Industry and expand to other verticals. It will also help company to execute its order book of more than 500 crore. According to Dr.Vinod Vasudevan, CEO, Flytxt, “Flytxt started in 2008 with a vision to provide over 10% measurable economic value to enterprises using customer data analytics technology. Today we have the 3rd generation of our product and more than 100 enterprises, including some of the largest Telecom operators in the world, benefit from it. This investment will help us to serve our customers more efficiently as well as invest in R&D, and grow faster through cloud offerings and channel partnerships.” The company offers enterprise-class customer data analytics product that can ingest, manage and analyse large volume of moving customer data to derive actionable insights. The product also has built-in applications to leverage these insights for digital customer engagement and digital services. More than 50 Telcos and other enterprises have deployed this product in 40+ countries. In certain markets, Flytxt and Telco partners also provide brands and agencies with aggregated and anonymised customer insights for executing targeted mobile advertising campaigns. Solutions built using Flytxt’s products and services have generated up to 7% measurable economic value for enterprises.","excerpt":"Netherlands-based Flytxt customer data analytics Software Company announced that it has aised Rs. 70 crore funding from DAH Beteiligungs GmbH. Flytxt plans to invest new funds in R&D to continue evolving its software product to further its growth in the Telecom Industry and expand to other verticals. It will also help company to execute its […]","categories":["AI News"],"tags":[],"author_name":"Richa Bhatia","publish_date":"2017-04-27T03:29:33","publication_year":"2017","word_count":243,"keywords":["funding","programming_languages:R","AI","Git","RAG","analytics","R"],"extracted_tech_keywords":["AI","analytics","RAG","R","Git","funding","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/flytxt-raises-rs-70-crore-scale-product-adoption\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":44344,"title":"5 Biggest Cybersecurity Updates From Black Hat 2019 You Should Know","content":"Source: GeekWire The biggest event for hackers concluded in Las Vegas last week. During the conference, there were many revelations that threw light on the cybersecurity space and some of them were shocking enough to get all the eyes. Here are the top updates that came out of Black Hat conference that you need to know about: Researchers Decrypted And Changed WhatsApp Messages Security experts from Checkpoint showed how hackers can alter WhatsApp messages. Checkpoint researchers reported that they had pointed the issue a year ago and the issue still persists. Interestingly, Facebook, the parent company has said that it is not a security vulnerability. WhatsApp is the largest social media messaging app, used by about 1.5 billion people across the globe, and 400 million users in India. Researchers pointed out the issue a year back, though Facebook says this is not a security vulnerability. Checkpoint Technologies also highlighted alleged vulnerabilities in the Facebook-own messaging app past week. Checkpoint researchers Roman Zaikin and Oded Vanunu said in their Black Hat USA session that they were able to reverse engineer WhatsApp web source code and decrypt traffic, thereby changing chat messages, as well as making private messages public and spoof sender identities. Checkpoint researchers shared that the platform leverages “protobuf2” protocol for encryption after successfully able to decrypt it. Converting protobuf2 encrypted data to Json (JavaScript Object Notation), researchers could see the actual parameters for the communication, which could also be manipulated to create spoof messages. Apple and Microsoft Bug Bounties For iOS And Azure Platforms Apple announced that it was giving away special iPhones to ethical hackers and however could break into them and discover vulnerabilities would be rewarded with a million dollar bounty. Apart from Apple which opened its bug bounty program for iOS and MacOS to researchers, Microsoft also announced a bug bounty of USD 300,000, for any researcher to hack and expose vulnerabilities on the Azure cloud platform. It is noted that Natalie Silvanovich from Google’s Project Zero team recently researched the remote interaction-less attack surface of the iPhone and discovered 10 vulnerabilities in SMS, MMS, Visual Voicemail, iMessage and Mail, all of which have been patched by Apple. Similarly, after Microsoft announced its Azure-based bug bounty, Checkpoint revealed a remote code execution vulnerability in Microsoft’s Remote Desktop Protocol (RDP) which made unpatched Azure users exposed to attacks. Warshipping: IBM’s New Hacking Technique IBM X-Force Red–an autonomous team of veteran hackers, within IBM Security showcased a novel attack technique which has been nicknamed “warshipping”. With the new technique, IBM team demonstrated how a hacker can remotely gain access to a company’s wireless network by simply putting a small remote-controlled scanning device inside a package that gets delivered to an office building. IBM reported that that businesses need to be wary of unique ways in which hackers are lurking in corporate networks to steal sensitive information. By simply delivering a parcel to the company’s office, hackers can access all the resources within a company’s network by hiding in plain sight. IBM said it cost them less than $100 to create the gadget used in the research to harvest sensitive data. Spoofed Satellite Navigation Signals To Hijack Autonomous Cars Victor Murray, an Engineering Group Leader at SwRI, demonstrated how autonomous self-driving cars to stop, change directions and drive off the path. Murray achieved this by spoofing navigation data from the Global Navigation Satellite Systems (GNSS), including GPS maintained by US or the Russian GLONASS. Murray highlighted a serious flaw in current navigation systems that hackers can exploit. In an interaction with media, Murray said GNSS signals are also very low power, and so it is not difficult to drown out legitimate GNSS broadcasts with malicious broadcasts. All of the GPS receivers are susceptible to spoofing as they lack integrity mechanisms, told Murray. Murray also mentioned prior research from Team Unicorn which was successful in creating spoof GNSS signals with less than a USD 400 radio and a laptop. New Spectre Vulnerability Found In Intel’s processors Researchers from Bitdefender exposed a vulnerability which affects all of Intel’s x86 and x64 processors after 2012 (unless they have been patched already in a recent Microsoft update). It’s a modified form Spectre 1 vulnerability that takes advantage of speculative execution, a function found in x86 and x64 processors that anticipates and executes instructions even before specific commands are received by the system. Speculative-execution can leave information traces in-cache, allowing hackers to get in and access information stored in the protected kernel memory. The vulnerability makes use of the SWAPGS kernel-level instruction, first rolled out in 2012 with Ivy Bridge processors.","excerpt":"The biggest event for hackers concluded in Las Vegas last week. During the conference, there were many revelations that threw light on the cybersecurity space and some of them were shocking enough to get all the eyes. Here are the top updates that came out of Black Hat conference that you need to know about:  […]","categories":["AI Trends"],"tags":["Apple","Bug Bounty","cyber security India","Microsoft","whatsapp"],"author_name":"Vishal Chawla","publish_date":"2019-08-12T17:00:30","publication_year":"2019","word_count":767,"keywords":["Go","cyber security India","unicorn","AI","Apple","R","Bug Bounty","Git","RAG","whatsapp","Ray","JavaScript","Azure","Java","Microsoft"],"extracted_tech_keywords":["AI","Ray","RAG","Azure","R","JavaScript","Go","Java","Git","unicorn"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/black-hat-hacker-conference-cybersecurity-news\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":45409,"title":"Hardware Players Can Move Up The Stack By Defining A Software Strategy","content":"Artificial Intelligence (AI) has been the star  technology over the past few years and it is easy to see why. From smartphones to PCs, some of the most game-changing applications are powered by AI . In addition to this, there is a groundswell of interest around applications like video recognition, speech recognition or object detection that will find use cases in smartphones, security and self-driving cars. Many AI applications that we see in our day-to-day lives have become ubiquitous, for example, Facebook photo-tagging or virtual assistant on our smartphones. Gartner predicts the business value created by AI will reach $3.9 trillion in 2022. Some of the use cases that will drive the maximum investment are intelligent automation on the enterprise side and product recommendations and digital assistants on the consumer end. Most of these compelling use cases depend on computational power and deep learning algorithms that will drive the next phase of growth in AI. So, while the data science community is overwhelmed by AI’s possibilities, what’s  often less clear is the underlying significance of the hardware layer powering these applications. In this article, we explore why data scientists today demand a full-stack solution across hardware platforms, tools and libraries to enable ease of application development. We also look at the shifting revenue towards inference and the need for software\/frameworks to utilise the power of the high-end processors. AI Hardware & Software Presents The Biggest Untapped Opportunity Given the massive potential of AI, trendsetting companies have taken concrete steps to capture real value from AI and win in this market. End users of enterprise-AI and even consumer-AI expect AI systems to provide highly accurate outcomes and that too in near-real-time. The amount of data available for training these AI systems is increasing at a steady rate. While data availability is no longer a major bottleneck, the computing infrastructure — hardware + software will be the key differentiator to meet the dual requirements of speed and accuracy. What Forward-thinking Enterprises Want — To Build Highly Accurate Models In all business domains, the AI systems that provide state-of-the-art performance are all Deep Learning-based systems. These DL-based algorithms are extremely computationally intensive and have fuelled rapid advancements in the processor technology in general but more specifically in the processors used in the AI systems. The most crucial need for businesses today is to build highly accurate models. For an AI system to be highly accurate, the trained AI models need to be generalizable and the training infrastructure needs to be scalable to seamlessly ingest more data. In the next step, for AI system to make predictions in near-real-time, the inference infrastructure needs to have the appropriate computational sophistication. Hardware Is The Key Differentiator In AI Image for representation purpose A typical data science infrastructure relies on a combination of GPUs and CPUs which are quite different architecturally. CPUs, the traditional workhorses of computing, are designed for efficient computation of complex operations whereas GPUs are best at simpler\/elemental operations which need to be optimally repeated at a massive scale. GPUs started off as processors for better rendering of high fidelity graphics but have now grown on to be a class of their own. GPUs achieve the optimality through in-built parallelization. Since AI system training is typically done on tens of millions of training samples, the same set of features need to be computed for each of these training samples and the same set of model parameters need to be updated over and over again. GPUs are thus most suited for these repetitive tasks. The parallelization provided by the GPUs also helps in completing these training tasks relatively quickly, which in turn helps in exploring many more training architectures. Inference Will Drive The AI Market Forward Model training and inference (i.e., prediction) happen in very different contexts and hence their hardware requirements are also quite different. For example, inference typically happens on only a few samples at a time and that too these samples are captured across a wide variety of conditions, and hence there isn’t enough parallelization to warrant a GPU. Thus, for inference, a CPU is preferred almost always. In terms of cost, GPUs are typically a lot more expensive than CPUs. Another key point one must remember is that inference typically happens on more diverse and less expensive devices (like smartphones, drones, edge devices) as compared to training, which typically happens on dedicated hardware. It is interesting to note that the industry is working towards making GPUs the go-to-processors for inference. Among other things, iPhones now have GPUs. No doubt, going forward inference will become the biggest revenue contributor in AI. What’s The Next Frontier In AI — Cohesive Software Environment Want to know why hardware enterprises are racing to build next-gen processors that are optimized with popular Deep Learning frameworks? Data Scientists don’t want to spend time parsing high-level AI\/ML code in a way that is consistent with the optimization of the corresponding hardware. Moreover, Data Scientists and Machine Learning Engineers are not necessarily trained in writing code which can inherently utilise the optimization capabilities of these CPUs\/GPUs. That’s why the next wave of innovations will come from AI frameworks that are developed around these different processors. To fully utilise the power of these high-end processors (be it CPUs or GPUs), data scientists require software\/framework. In fact, most data scientists and machine learning engineers don’t want to spend any extra effort acquiring these skills. This makes it all the more important that the next-gen processors also have a wide array of necessary framework which can interpret and port the ‘high-level’ AI\/ML code into optimized routines which utilize all the capabilities of the underlying processors. These speedup frameworks can either be provided by the creators of the processors themselves or by sponsoring communities of coders who will be contribute to these frameworks. For example, some of the most popular frameworks used today are: a) clDNN library by Intel is a set of compute kernels to accelerate deep learning and AI models b) Intel® MKL-DNN actively supported by Intel coding community for both CPUs and GPUs is the set of accelerators which is optimized for a variety of deep learning architectures but more so specifically for some of the popular ones like AlexNet, GoogleNet etc. Finally, the next big push in the industry is towards building processors which are specifically optimized for certain kinds of AI workloads (Nervana Neural Network Processor (NNP-I) and TPUs). Bottomline Semiconductor companies and AI computing players need to coalesce their strategy around hardware innovations and software that reduces the complexities and offers a wide appeal to the developer ecosystem. To win the developer community over, hardware players need to offer simpler interfaces and a suite of software platforms that is compatible with multiple compute architecture. Hardware companies should also think how to include application-specific capabilities that could see a high demand over the next few years.","excerpt":"Artificial Intelligence (AI) has been the star  technology over the past few years and it is easy to see why. From smartphones to PCs, some of the most game-changing applications are powered by AI . In addition to this, there is a groundswell of interest around applications like video recognition, speech recognition or object detection […]","categories":["AI Features"],"tags":["AI frameworks","real time face recognition software","Software","Software Development"],"author_name":"Om Deshmukh","publish_date":"2019-09-04T11:13:27","publication_year":"2019","word_count":1148,"keywords":["data science","artificial intelligence","machine learning","TPU","AI","neural network","ML","Ray","deep learning","object detection","Software Development","Software","AI frameworks","real time face recognition software"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","neural network","data science","Ray","object detection","TPU"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/hardware-players-can-move-up-the-stack-by-defining-a-software-strategy\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":57824,"title":"Interview With Juergen Schmidhuber","content":"There are close to 3.5 billion smartphone users in the world, and if you happen to be one of them, then the chances are high that Juergen Schmidhuber has already touched your life. Be it Apple’s Siri or Amazon’s Alexa, all the top speech and voice assistants work on LSTM or Long Short Term Memory. Though these digital assistants and other applications of LSTMs are fairly modern, their conception, however, dates back to the early 1990s. Introduced by Juergen Schmidhuber and his doctoral student Sepp Hochreiter, these networks would become one of the most profitable applications in the history of deep learning. However, the mastermind behind innovations such as LSTMs, Dr.Schmidhuber largely remains unknown outside the research circles unlike his contemporaries such as Yann Lecun, Andrew Ng and Geoff Hinton. Analytics India Magazine got in touch with Dr.Schmidhuber for an interview, to present its readers with the story of a man who dreams of building intelligent machines that would one day surpass the intellectual might of his childhood role model Albert Einstein. 1990-91: The Year Of Miracles After a four year study of computer science and mathematics at TU Munich, Dr. Schmidhuber went on to work on his PhD (1988-91) thesis titled,`Dynamic Neural Nets and the Fundamental Spatio-Temporal Credit Assignment Problem’. With the end of his doctoral duties at TU Munich in sight, Dr.Schmidhuber and his peers produced a flurry of papers that would eventually change the course of AI research. Here is a list of few of those works: First Deep Learner, Based on Unsupervised Pre-Training (1991)The Fundamental Deep Learning Problem (Vanishing \/ Exploding Gradients, 1991)Long Short-Term Memory: Supervised Very Deep Learning (basic insights since 1991)Artificial Curiosity Through Adversarial Generative NNs (1990)Adversarial Networks for Unsupervised Data Modeling (1991)Learning Sequential Attention with NNs (1990) Majority of the aforementioned works took decades to get noticed and few are still being leveraged for weaving new ideas. For instance, let’s take a look at LSTMs. They were introduced in 1990 to solve the vanishing and exploding gradient problem in recurrent neural networks(RNNS). LSTMs offered the human-like thinking capability of holding on to relevant information through a mechanism called as gates. It was an ingenious idea. However, the reception to LSTM had been lukewarm. But today, they are on almost every smartphone! From  Facebook’s automatic translation (4 billion translations per day) to Google’s speech recognition on 2 billion Android phones to Amazon’s Alexa, there has been a tremendous increase in the usage of LSTM with the improving hardware capabilities. Citations profile via Google Scholar Whereas, his work on artificial curiosity through adversarial generative neural networks has many similarities in the widely popular generative adversarial networks (GANs). These adversarial networks, though popularised by the work done by Ian Goodfellow, are considered to be a re-invention or an interesting application of Dr.Schmidhuber’s work in 1990. As can be seen in the picture above, Dr.Schmidhuber’s work still continues to be one of the highest cited works in the AI community and his determination to find super-intelligent machines still going strong. Nowadays Dr. Schmidhuber spends most of his time taking classes at the university while attending to his duties as the Director of The Swiss AI Lab IDSIA. He along with four others have also founded a deep learning company, which goes by the name NNAISENSE, based out of Lugano, Switzerland. Re-NNAISENSE Of Machines Dr.Schmidhuber(far right) along with NNAISENSE co-founders Based out of Lugano, NNAISENSE (pronounced “nascence”) was founded by Dr.Schmidhuber, Faustino Gomez, Jan Koutnik, Jonathan Masci and Bas Steunebrink in 2014. This 30 members deep learning company has many top projects lined up including projects from German carmaker Audi as well. Here are a few of NNAISENSE’s interesting accomplishments: It had the first very deep neural nets with hundreds of layers. NNAISENSE founders received numerous awards including the 2016 IEEE Neural Networks Pioneer Award “for pioneering contributions to deep learning and neural networks.” NNAISENSE won the “Learning to Run” competition at the prestigious NIPS conference in  2017, beating over 400 competitors from industry and academia. It was also the first company to demonstrate physical real-world control through deep Reinforcement Learning: a project with AUDI, where highly novel model cars learned to park without a teacher. The team at NNAISENSE built the brain that learns to control the recent sophisticated pneumatic FESTO robot hand shown at Hannover Messe 2019. Currently, the team is working with Sulzer-Schmid, leader of (drone-based) wind turbine inspection along with other ambitious projects such that includes additive manufacturing and other undisclosed ones. Dr.Schmidhuber’s motto since the 1970s has been to build an AI smarter than him so that he can retire The team at NNAISENSE believes that they can go far beyond what’s possible today, and pull off the big practical breakthrough that will change everything. At NNAISENSE, these handfuls of AI masterminds work tirelessly towards achieving a singular goal, which also happens to be Dr. Schmidhuber’s childhood dream—to match or even surpass human-level intelligence in machines. Most Essential Anti Hero Of Our Time Dr.Schmidhuber at his labs While the New York Times called him ‘Dad’, Bloomberg glorified him as godfather. However, for Dr.Schmidhuber, who is not a huge fan of these idolatry extravaganzas, Alexey Ivakhnenko is the true father of deep learning. Ivakhnenko was a Soviet mathematician whose work on deep networks dates back to 1965, but like Dr.Schmidhuber, Ivakhnenko largely remains unrecognised. Dr.Schmidhuber firmly believes that the machine learning community can only gain from proper credit assignment to its members. “The inventor of an important method should get credit for inventing it. If you “re-invent” something that was already known, and only later become aware of this, you must at least make it clear later.” Dr.Schmidhuber Over the past decade, the AI community has increasingly become divided over Dr.Schmidhuber. While one group tries to paint Dr.Schmidhuber as an egomaniac who tries to bite more than he can chew, the rest are quite certain that Dr.Schmidhuber has been terribly ignored and it is high time he gets the credit. So we got straight to the point and asked him to comment on the Turing award snubs. We even tested our luck by bringing to his notice, the conception of a Schmidhuber award, which is popular on forums such as Reddit. Like a gentleman he is, Dr.Schmidhuber was graceful enough to accept the enthusiasm amongst his fans. “Generally speaking, I greatly appreciate the support from the machine learning community,” said Dr.Schmidhuber acknowledging the efforts of individuals on Reddit. “I am extremely grateful to my students and postdocs who have made all of this possible.” Dr.Schmidhuber He credits his students and postdocs for all the accolades he has amassed over the years, which is so unlike what his adversaries blame him for! Talking about adversaries in the context of deep learning, one can’t help but think about the sudden rise in popularity of a phenomenon that is Generative Adversarial Networks or GANs. Ian Goodfellow, who had been enjoying the success as the poster child of GANs, was put to the test by Dr.Schmidhuber at the prestigious NIPS 2016 conference. Goodfellow’s presentation was interrupted by Dr.Schmidhuber and the 21st-century research circles were given a taste of some insubordination, which was missing for over a half a century. However, this gimmick only resulted in divorcing the masses further away from Dr.Schmidhuber. The advancement of any research field relies on the openness to criticism and though Dr.Schmidhuber’s objections are mostly about the lack of recognition to the real pioneers, his whole old school Swiss demeanour might have rubbed off a few researchers in the wrong way. “Relatively young research areas such as machine learning should adopt the honor code of mature fields such as mathematics.” Dr.Schmidhuber The big 3 of the modern AI scene, Yann Lecun, Yoshua Bengio and Geoffrey Hinton, have been at the forefront of every major AI breakthrough news that has come over in the past decade. In 2018, they were even awarded the prestigious Turing Award for their contributions made to deep neural networks. However, the trio, LBH(Lecun-Bengio-Hinton), were accused of circular citations by Dr.Schmidhuber in his 2015. In a post titled, “Deep Learning Conspiracy”, he explains in detail how LBH have been ignorant of the original inventors. Photo by Andreas Gerbert In this critique, he laments how misleading it is to cite Hinton’s 2012 paper in the context of convolutional neural networks. LBH, wrote Dr.Schmidhuber, mention pooling, but not its pioneer (Weng, 1992), who replaced Fukushima’s (1979) spatial averaging by max-pooling, today widely used by many, including LBH, who write: “ConvNets were largely forsaken by the mainstream computer-vision and machine-learning communities until the ImageNet competition in 2012,” citing Hinton’s 2012 paper (Krizhevsky et al., 2012). He considers this to be very misleading. “LBH may be backed by the best PR machines of the Western world (Google hired Hinton; Facebook hired LeCun). However, historic scientific facts will be stronger than any PR.” Dr.Schmidhuber, “Deep Learning Conspiracy” (Nature 521 p 436) Though the contributions of Lecun, Bengio and Hinton to deep learning cannot be disputed, they are accused of inflating a citation bubble. Dr.Schmidhuber has been vociferous about the ignorance of the original inventors in the AI community. He believes that there is a long tradition of insights into deep learning, and the community as a whole will only benefit from appreciating the historical foundations. The idolization of an individual is a curse to any scientific community but so is ignorance of the prime movers. If one is even slightly familiar with the celebrities of AI, they would have come across epithets such as ‘godfather of AI’ or ‘GANfather’, which are thrown around a lot. Dr.Schmidhuber too is not alien to these superfluous titles, and when we insisted on knowing if there is anything novel about Goodfellow’s work on GANs at all, he responded by saying that GANs are an interesting application of my adversarial curiosity principle published in 1990. “One network probabilistically generates outputs, another network sees those outputs and predicts environmental reactions to them,” explained Dr.Schmidhuber. Using gradient descent, he continued,  the predictor network minimizes its error, while the generator network tries to make outputs that maximize this error. One net’s loss is the other net’s gain. GANs are a special case of this where the environment simply returns 1 or 0 depending on whether the generator’s output is in a given set. “One should not be allowed to patent deep learning.” Dr.Schmidhuber Bringing the recent rush towards patenting AI to his attention, we asked Dr.Schmidhuber to comment on it, especially about Goodfellow’s patenting of adversarial training, to which he quipped, “I am not a patent attorney. However, one should not be allowed to patent uncited prior art published by others.” When asked if any current researchers have impressed him, Dr.Schmidhuber didn’t take any names as he believes there are many impressive researchers out there. However, he does admit that DeepMind is doing impressive work as they beat a pro player in the game of Starcraft using Alphastar in 2019. At the heart of this success, reminds Dr.Schmidhuber, sits an LSTM network, which was also part of OpenAI Five which defeated human experts in the Dota 2! Goedel Machines, AGI And The Future from Stanley Kubrick’s 2001: A Space Odyssey Back in 2003, Dr.Schmidhuber proposed the Gödel machine, in an attempt to make a case for superintelligence in machines. A Gödel machine is capable of rewriting its own code as soon as it has found proof that the rewrite is useful. This in nutshell, is the definition of artificial general intelligence; to make the machines to learn how to learn. Ever since he was a teenager, Dr.Schmidhuber dreamt of building a self-improving AI that is smarter than him. Today, after three decades since the introduction of LSTM, he is still working towards achieving artificial general intelligence with the help of his team at NNAISENSE. In this decade, observes Dr.Schmidhuber, Active AI will invade the real world, driving industrial processes and machines and robots. Although the real world is much more complex than virtual worlds and less forgiving, Dr.Schmidhuber predicts that the coming wave of “Real World AI” or simply “Real AI” will be much bigger than the previous AI wave, because it will affect all of the production. In the not too distant future, the creation of the “show-and-tell robotics” or “watch-and-learn robotics” or “see-and-do robotics”, as Dr.Schmidhuber likes to call them, will allow us to quickly teach a neural network to control a complex robot with many degrees of freedom to execute complex tasks, such as assembling a smartphone, solely by visual demonstration, and by talking to it, without touching or otherwise directly guiding the robot – a bit like we’d teach a kid. In a talk he gave at, CogX event in London, Dr.Schmidhuber enthused the audience about how AI will eventually emigrate into other galaxies in a few thousand years. He believes that there are places out there in deep space that are richer in resources than earth and can facilitate self-replicating robots that would eventually build factories and so on so forth. “The delays between successive radical breakthroughs in computer science decrease exponentially: each new one comes roughly twice as fast as the previous one.” Schmidhuber’s law Dr.Schmidhuber firmly believes that human civilisation is on the verge of witnessing something spectacular and he urges us to take pride in being fortunate to be contributing to the birth of a new kind of being; a new civilization. It took from more than 13 billion years for the emergence of primitive forms of intelligence on earth after the Big Bang. But it took only a fraction of that time for humans to rule the earth and even less time to create AI. At this pace, we can safely assume that the next 100 years is going to be quite eventful. And, with the preparations for colonization of Mars and innovations like Neuralink around the corner, Dr.Schmidhuber might finally witness his childhood dream realised. “Don’t think of humans as the crown of creation. Instead view human civilization as part of a much grander scheme, an important step on the path of the universe towards higher complexity.” Dr.Schmidhuber for Scientific American, November 2017 As the 21st-century gears up to the advent of Artificial Super Intelligence, it is almost inevitable that the next generation students will soon have the fundamentals of AI included in their school curriculums. “Machine learning will seem trivial to you,” promises Dr.Schmidhuber, “if you learn basic math, including linear algebra, calculus, statistics and then learn the basics of theoretical computer science. And maybe also the basics of physics.” Also Watch: A Quick Guide TO LSTMs","excerpt":"There are close to 3.5 billion smartphone users in the world, and if you happen to be one of them, then the chances are high that Juergen Schmidhuber has already touched your life.  Be it Apple’s Siri or Amazon’s Alexa, all the top speech and voice assistants work on LSTM or Long Short Term Memory. […]","categories":["AI Features"],"tags":["Deep Learning","deepmind london","GANs","Geoffery Hinton","Ian Goodfellow","Interviews and Discussions","Juergen Schmidhuber","LSTM Network","NeurIPS","recurrent neural networks","Yann LeCun","Yoshua Bengio"],"author_name":"Ram Sagar","publish_date":"2020-03-03T12:00:22","publication_year":"2020","word_count":2435,"keywords":["recurrent neural networks","TPU","NeurIPS","deep learning","GANs","LSTM Network","R","Geoffery Hinton","Juergen Schmidhuber","RAG","Yoshua Bengio","analytics","Ian Goodfellow","machine learning","AI","neural network","ML","deepmind london","Yann LeCun","OpenAI","Deep Learning","Interviews and Discussions"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","neural network","analytics","OpenAI","RAG","TPU","R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/lstm-juergen-schmidhuber-artificial-intelligence-switzerland-goedel-machines\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10093912,"title":"Is the GPT Breakthrough in Robotics Approaching Soon?","content":"Be it Agility Robotics’ bots lifting weights to Google DeepMind’s playing football, we have seen tech companies heavily training and working on improving a robot’s capabilities. In a run towards improving task performance, companies are constantly innovating in the field. However, compared to the growth of LLM models, advancements in robotics would take time, as it deals with imperfections of control and sensors in the real world. Building and testing these prototypes often get delayed. And while an LLM’s prowess is measured by the training datasets and tokens, a robot’s efficiency is gauged by the number of tasks it can accomplish — something that takes years to master. Drawing parallels to how GPT was built and improved by an ecosystem of companies that worked together, data expert and AI strategy advisor Vin Vashishta recently spoke about how robotics is still on the path to having its own GPT moment. He believes that innovation doesn’t start with advanced robotics, instead with applied robotics and an ecosystem of companies working together. With tech companies working and investing in robotics companies, their next big breakthrough will probably happen with companies being able to build on small-scale demonstrations and development of fundamental data collection components. So, there are two pertinent questions — how far have tech companies gone in robotics development and will the turning point occur any time soon? Waiting for the GPT-Moment The sector, as a whole, is continuously innovating, and yet, the breakthrough that everyone is anticipating is yet to come. This can be attributed to the fact that robotics, as a sector, is extremely hard to get right. It’s hard to make the systems robust and repeatable at scale. Moreover, robotics is a tricky sector for investors as huge money is required to bring production to scale. This discourages investors who are looking to make quick money. Recently, AIM got in touch with Gokul NA, founder of CynLr: Cybernetics Laboratory, a robotics company that works on robotic arms. He spoke about the time taken in robotics training. “It might be easy to teach a robotic arm to take something from one position to another, but it takes over two years to get a successful robotic arm application, if feasible at all.” Gokul also spoke about the future of this field. “We have seen the data revolution, and it’s time to see the object revolution at this point. We see bright possibilities for this industry and it is left to us to figure out how we can leverage it and make it a success.” According to a recent report, the global robotics market size was valued at $12.1 billion in 2020, and is expected to reach $150 billion by 2030. One of the biggest use cases for robotics is in warehouses, where the need for automation is huge, creating room for growth. Other industries such as construction, agriculture, and healthcare are also sectors where robotics work. Big Tech Goes the Robotics Way Companies are taking leaps to invest in robotics companies. You even have major tech players, such as Google, Microsoft, OpenAI, and Tesla, which have invested in it, and continue to do so. In spite of scrapping their in-house robotics division in 2021, OpenAI recently invested in a Norway-based robotics startup, 1x, that builds humanoid robots capable of human-like movements and behaviours. The first AI robot officially entered the workforce.1X, backed by OpenAI, just outpaced Elon Musk's Tesla as the first humanoid robot in a professional environment.Their EVE robot has been integrated as a security guard in manufacturing sites. pic.twitter.com\/ryv7VuOtJ7— Rowan Cheung (@rowancheung) May 23, 2023 A few weeks ago, AMP Robotics, a Colorado-based startup that creates robotic systems, received close to $100 million from Microsoft’s Climate Innovation Fund. Last week, Canadian AI and robotics company SanctuaryAI unveiled their advanced general-purpose robot called Phoenix. The company claims to have developed the world’s first humanoid general-purpose robot powered by Carbon, which is a unique AI control system. Last year, Tesla revealed their humanoid robot Optimus. It received a lukewarm response, and even garnered criticism. Gary Marcus raised concerns about why Tesla would build such a robot with no clear direction on what the real-world applications are. Snubbing the critiques, the companies are continuously investing in this sector, awaiting the ‘GPT moment’ for robotics.","excerpt":"On the heels of AI and LLM advancements, will Robotics get its GPT moment anytime soon?","categories":["AI Features"],"tags":["data collection","gary marcus","Google","Google Deepmind","GPT","Humanoid Robots","Microsoft","OpenAI","optimus","Robotics","Tesla"],"author_name":"Vandana Nair","publish_date":"2023-05-25T14:00:00","publication_year":"2023","word_count":712,"keywords":["Robotics","gary marcus","R","Humanoid Robots","optimus","RAG","Tesla","startup","Go","AI","Google Deepmind","OpenAI","innovation","automation","GPT","Aim","Google","data collection","Microsoft"],"extracted_tech_keywords":["AI","OpenAI","Aim","RAG","R","Go","GPT","automation","innovation","startup"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/is-the-gpt-breakthrough-in-robotics-approaching-soon\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10051125,"title":"A Guide to Stochastic Process and Its Applications in Machine Learning","content":"Many physical and engineering systems use stochastic processes as key tools for modelling and reasoning.  A stochastic process is a probability model describing a collection of time-ordered random variables that represent the possible sample paths. It is widely used as a mathematical model of systems and phenomena that appear to vary in a random manner. As a classic technique from statistics, stochastic processes are widely used in a variety of areas including bioinformatics, neuroscience, image processing, financial markets, etc. In this post, we will discuss the stochastic process in detail and will try to understand how it is related to machine learning and what are its major application areas. The major points to be discussed in this article are outlined below. Table of Contents Stochastics in GeneralStochastic ProcessExamples of Stochastic ProcessesComparing Stochastic Systems with Other SystemsStochastic Process in Machine LearningApplication of Stochastic Process Let’s start by knowing the general meaning of stochastic. Stochastic in General Stochasticity is the property of being well described by a random probability distribution. Although stochasticity and randomness are distinct in that the former refers to a modelling method and the latter to phenomena, the terms are frequently used interchangeably. Furthermore, the formal concept of a stochastic process is also referred to as a random process in probability theory. Stochasticity is employed in a variety of domains, including biology, chemistry, ecology, neuroscience, and physics, as well as image processing, signal processing, information theory, computer science, cryptography, and telecommunications. It’s also employed in health, linguistics, music, media, colour theory, botany, manufacturing, and geomorphology, all of which are affected by seemingly random movements in financial markets. In social science, stochastic modelling is utilized. Stochastic Process Although the definition of a stochastic process varies, it is typically characterized as a collection of random variables indexed by some set. Without the index set being clearly described, the phrases random process and stochastic process are considered synonyms and are used interchangeably. The phrases “collection” and “family” are used interchangeably, whereas “parameter set” or “parameter space” are occasionally used instead of “index set.” Some theoretically defined stochastic processes include random walks, martingales, Markov processes, Lévy processes, Gaussian processes, random fields, renewal processes, and branching processes. Probability, calculus, linear algebra, set theory, and topology, as well as real analysis, measure theory, Fourier analysis, and functional analysis, are all used in the study of stochastic processes. Example of Stochastic Process Poissons Process The Poisson process is a stochastic process with several definitions and applications. It’s a counting process, which is a stochastic process in which a random number of points or occurrences are displayed over time. A time-dependent Poisson random variable is defined as the number of points in a process that falls between zero and a certain time. Non-negative numbers make up the index set of this process, but natural numbers make up the state space. Because it can be conceived of as a counting operation, this procedure is often referred to as the Poisson counting process. Bernoulli Process One of the most basic stochastic processes is the Bernoulli process. It’s a set of independent and identically distributed (iid) random variables, each with a probability of one or zero, for example, one with probability p and zero with probability 1-p. This method is similar to repeatedly flipping a coin, with the chance of getting a head being p and the value being one, and the probability of receiving a tail being zero. A Bernoulli process, in other words, is a set of iid Bernoulli random variables, with each coin flip representing a Bernoulli trial. Random Walk The simple random walk is a typical example of a random walk. It is a stochastic process in discrete time with integers as the state space and is based on a Bernoulli process, with each Bernoulli variable taking either a positive or a negative value. In other words, the simple random walk occurs on integers, and its value grows by one with probability p or lowers by one with probability 1-p, hence the index set of this random walk is natural numbers, but its state space is integers. If p=0.5, this random walk is referred to as a symmetric random walk. Comparing Stochastic Systems with Other Systems Let’s compare stochastic systems to other similar terms that are occasionally used as synonyms for stochastic to gain a better grasp of it. Stochastic is synonymous with random and probabilistic, although non-deterministic is distinct from stochastic. Stochastic Vs Gradient Stochastic Vs Probabilistic The terms stochastic and probabilistic are frequently interchanged. Probabilistic is most likely the broader term. Stochastic is dependent on a previous occurrence, such as fluctuations in stock price based on the previous day’s price, but probabilistic is independent of other observations, such as winning lottery numbers, which are supposed to be independent of one another. Stochastic Vs Non-Deterministic Deterministic refers to a variable or process that can predict the result of an occurrence based on the current situation. In simple terms, we can state that nothing in a deterministic model is random. Non-deterministic, on the other hand, is a variable or process in which the same input might result in different results. Because the outcome is unpredictable, stochasticity is often used interchangeably with non-deterministic methods. In the way that we may undertake analysis using probability tools like anticipated result and variance, stochasticity is slightly different from non-deterministic. As a result, defining a variable as stochastic rather than non-deterministic is a stronger claim. Stochastic Vs Random In most cases, stochastic is used interchangeably with random. Random refers to unpredictability, and in the ideal scenario, all outcomes are equally likely, implying that there is no reliance on the other observation, such as tossing a fair coin, whereas stochastic refers to the probabilistic nature of the variable that is randomly chosen. Stochastic Processes in Machine Learning Stochasticity is used to explain several machine learning methods and models. This is due to the fact that many optimizations and learning algorithms must function in stochastic domains, and some algorithms rely on randomness or probabilistic decisions. Let’s look at the source of uncertainty and the nature of stochastic algorithms in machine learning in more detail. How it is Identified in Machine Learning Domains involving uncertainty are known as stochastics. Statistical noise or random errors can cause uncertainty in a target or objective function. It could also be due to the fact that the data used to fit a model is a sample of a larger population. Finally, the models adopted are rarely able to capture all elements of the domain, and must instead generalize to unknown scenarios, resulting in a loss of fidelity. Optimizing the Stochastic Optimization approaches that create and employ random variables are known as stochastic optimization (SO). Random variables exist in the formulation of the optimization problem itself for stochastic issues, which incorporates random objective functions or random constraints. Random iterate methods are also included in stochastic optimization approaches. Some stochastic optimization approaches combine both definitions of stochastic optimization by using random iterates to address stochastic issues. The following are some instances of stochastic optimization algorithms: Particle Swarm OptimizationSimulated Annealing Genetic Algorithm Stochastic Learning Algorithm The two most prevalent and widely used algorithms in machine learning are stochastic gradient descent and stochastic gradient boosting. Stochastic gradient descent (SGD) is a variant of the gradient descent technique that computes the error and updates the model for each example in the training dataset. Because the model is updated for each training example, stochastic gradient descent is frequently referred to as an online machine learning algorithm. The stochastic gradient boosting algorithm is a collection of decision tree techniques. The stochastic aspect refers to the random subset of rows drawn from the training dataset that are utilized to build trees, specifically the split points of trees. Application of Stochastic Process Below are some general and popular applications which involve the stochastic processes:- Stochastic models are used in financial markets to reflect the seemingly random behaviour of assets such as stocks, commodities, relative currency values (i.e., the price of one currency relative to another, such as the price of the US Dollar relative to the price of the Euro), and interest rates. Manufacturing procedures are thought to be stochastic. This assumption holds true for both batch and continuous manufacturing processes. A process control chart depicts a particular process control parameter across time and is used to record testing and monitoring of the process. The marketing and shifting movement of audience tastes and preferences, as well as the solicitation and scientific appeal of the certain film and television debuts (i.e., opening weekends, word-of-mouth, top-of-mind knowledge among surveyed groups, star name recognition, and other elements of social media outreach and advertising), are all influenced in part by stochastic modelling. Stanislaw Ulam and Nicholas Metropolis popularized the Monte Carlo approach, which is a stochastic method. The use of randomness and the repetitive nature of the procedure is reminiscent of casino activities. Simulation and statistical sampling methods were typically used to test a previously understood deterministic problem, rather than the other way around. Though historical examples of an “inverted” technique exist, they were not regarded as a generic strategy until the Monte Carlo method gained popularity. Conclusion In this post, we understood the stochastic process with different concepts and application areas. We went through the definition of stochastic, how it differs from related terms like random, probabilistic, and nondeterministic, and what stochastic means in machine learning in this post. In addition, we have some fundamental and common examples of stochastic processes that can be seen in generic terms. References Stochastic ProcessStochastic OptimizationWhat Does Stochastic Mean in Machine Learning?","excerpt":"Many physical and engineering systems use stochastic processes as key tools for modelling and reasoning.","categories":["Deep Tech"],"tags":["Data Science","gradient","Guide","Machine Learning","machine learning engineer"],"author_name":"Vijaysinh Lendave","publish_date":"2021-10-10T16:00:00","publication_year":"2021","word_count":1598,"keywords":["Go","machine learning","programming_languages:R","AI","ML","machine learning engineer","Machine Learning","programming_languages:Go","Aim","ViT","gradient","Data Science","R","Guide"],"extracted_tech_keywords":["AI","machine learning","ML","Aim","R","Go","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/a-guide-to-stochastic-process-and-its-applications-in-machine-learning\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10173,"title":"Artificial Intelligence can now write Movies too","content":"Off late, Artificial Intelligence (AI) is being seen in lot of spaces from cooking to art to games. However, AI has recently ventured into a new area and that is ‘Movies’. Shocking it may seem but yes a movie has been written by AI and the result is the movie ‘Sunspring’. The film was created to participate in Sci-Fi London film festival’s 48-hour challenge. Filmmaker Oscar Sharp and Ross Goodwin, a New York University AI researcher came together to build a computer for screenplay-writing. And the result was Benjamin, an LSTM RNN Artificial Intelligence or simply put, an AI which his used for text recognition. To create a science fiction film, Benjamin was fed with lots of science fiction movies and TV serials like Eternal Sunshine of the Spotless Mind, Ghostbusters, 12 Monkeys, The Fifth Element, Blade Runner, the Alien films, Interstellar, Looper, The Matrix, Star Wars, The X-Files. After reading these stories, AI Benjamin wrote a screenplay, now called Sunspring. He also wrote the lyrics of a song for the film which the filmmakers included in the 9 minute short science fiction movie. Then filmmaker, Oscar Sharp, got the cast together, Thomas Middleditch from Silicon Valley, Elisabeth Gray and Humphrey Ker. These actors were assigned their parts on a random basis and they had to interpret and make it into a film in a day’s time based on the dialogues and the actor directions provided by Benjamin. The movie has been received with mixed views. Though some thought that in spite of the sentences being grammatically correct, they were in isolation and did not weave together as it should in a movie. Some felt the movie did match the genre of science fiction and considered Gray’s emotional monologue towards the end of the film to be its highlight. Science fiction movies push our imagination limits as they are a mix of science, reality, and fantasy. It allows the makers to imagine things of future, what could be possible in the time to come thereby making imagination the key for a successful script. Also by this film, the filmmakers have set an example of how Artificial Intelligence could be the future in terms of writing dialogues, giving directions to actors, hence giving way to imagination of how creatively a movie can be written. Though the human writers have nothing to worry and fear from the AI like Benjamin in the near future, however the thought process of making AI do everything has already begun. A serious thought needs to be given to the question – Is Artificial Intelligence the future of film making?","excerpt":"Off late, Artificial Intelligence (AI) is being seen in lot of spaces from cooking to art to games. However, AI has recently ventured into a new area and that is ‘Movies’. Shocking it may seem but yes a movie has been written by AI and the result is the movie ‘Sunspring’. The film was created […]","categories":["IT Services"],"tags":[],"author_name":"Manisha Salecha","publish_date":"2016-06-14T12:11:19","publication_year":"2016","word_count":433,"keywords":["Go","artificial intelligence","ELT","programming_languages:R","AI","Ray","RNN","LSTM","AI research","R"],"extracted_tech_keywords":["AI","artificial intelligence","Ray","R","Go","ELT","RNN","LSTM","AI research","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/it-services\/artificial-intelligence-can-now-write-movies\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10130426,"title":"Bengaluru-based Conversational AI Startup Gnani AI Raises $4 mn","content":"GenAI startup Gnani AI, which is India’s first voice-first SLM for vernacular languages, recently raised $4 mn from tech holding company Info Edge. The investment will help lead the startup’s expansion plans. “What we’re doing is something different, like a fusion of voice and text models. It’s a multimodal model but right now we are focused on voice and text,” said Ganesh Gopalan, CEO of Gnani.ai, in an earlier exclusive interaction with AIM. Indic Language Training Datasets The startup that is building a series of voice-based small language models has been training on millions of hours of proprietary audio datasets and billions of conversations in Indic languages. In India, their model supports 12 Indian languages. In the US, English and Spanish languages are provided. Gnani AI have built a series of five models suitable for banking, finance, security and insurance domains. Running their AI models on the edge, Gopalan believes this method will help bring down costs. THe startup also specialises in quickly fine tuning models based on enterprise needs. “It’s one thing to have an SLM for the BFSI sector, but the real value to a company is when you have a model built just for their data. So that’s what we do. We take our model to enterprises and help them build on top of it,” said Gopalan. Interestingly, there are a number of emerging genAI startups in India that are exclusively working for finance customers. Recently, OnFinance, a genAI startup, is building proprietary LLMs for BFSI customers. They also became the first genAI startup to have NSE (National Stock Exchange of India) as their client.","excerpt":"Gnani AI has over 100 customers in banking and financial services, insurance, telecom, automotive and healthcare industries.","categories":["AI News"],"tags":["Conversational AI","slm"],"author_name":"Vandana Nair","publish_date":"2024-07-28T13:29:05","publication_year":"2024","word_count":268,"keywords":["Go","GenAI","AI","Modal","slm","SLM","Aim","Conversational AI","GAN","small language models","R","startup"],"extracted_tech_keywords":["AI","GenAI","Aim","small language models","SLM","R","Go","GAN","startup","Modal"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/bengaluru-based-conversational-ai-startup-gnani-ai-raises-3m\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":22271,"title":"IIT Madras Joins Hands With Applied Materials India For Research In AI","content":"Adding to its list of strategic agreements with the industry to boost academic and research activities, the Indian Institute of Technology Madras (IITM) has inked a joint development agreement (JDA) with Applied Materials India to carry out research in artificial intelligence, machine learning and data Sciences. The agreement was signed by Krishnan Balasubramanian, dean, Industrial Consultancy and Sponsored Research, IITM, and Srinivas Satya, country president and managing director at Applied Materials India. The industrial areas that can gain from this understanding include pharmaceuticals  and semiconductors. In yet another collaboration between #Industry and #Academia, @iitmadras has signed signed a Joint Development Agreement with Applied Materials India to conduct research in #DataSciences, #MachineLearning & #AI . Semiconductors, #Pharma and Display set to benefit. @Applied_Blog pic.twitter.com\/TpNrIzwZXS — IIT Madras (@iitmadras) February 28, 2018 “We look forward to collaborating with Applied Materials in the fast-expanding field of AI. IIT Madras is already a leader in this area and bringing the expertise from academics and industry gives us unique opportunities.,” said Bhaskar Ramamurthi, Director at IITM The work expected to be carried out under the JDA includes images for technology enabled software solutions (TES),machine learning methods, advanced image analysis and use of representational neural networks. Highlighting the importance of innovations in materials engineering on improving the potential of AI, Dr Om Nalamasu, senior vice president and chief technology officer, Applied Materials, said, “Materials engineering innovations are key to achieving the new compute models, chip architectures and materials systems needed to realise the potential of Artificial Intelligence. Applied Materials’ collaboration with IIT Madras is focused on the breakthroughs that can accelerate this new era.” This agreement joins the list of strategic partnerships forged by IITM with industries and the government in 2018. In February, it was reported that IITM, along with National Programme on Technology Enhanced Learning (NPTEL), was working with the glass industry to develop MOOCHs aspiring civil engineers and architects to raise awareness on effective use of glass in construction. In the same month, IITM entered into a memorandum of agreement (MoA) with the Union Shipping Ministry to establish the National Technology Centre for Ports, Waterways and Coasts (NTCPWC), which under the Central Government’s Sagarmala project, aims at modernising Indian Ports.","excerpt":"Adding to its list of strategic agreements with the industry to boost academic and research activities, the Indian Institute of Technology Madras (IITM) has inked a joint development agreement (JDA) with Applied Materials India to carry out research in artificial intelligence, machine learning and data Sciences. The agreement was signed by Krishnan Balasubramanian, dean, Industrial […]","categories":["AI News"],"tags":["IIT Madras"],"author_name":"Jeevan Biswas","publish_date":"2018-03-05T12:11:30","publication_year":"2018","word_count":367,"keywords":["data science","Go","machine learning","artificial intelligence","AI","IIT Madras","neural network","innovation","Aim","ViT","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","neural network","data science","Aim","R","Go","ViT","innovation"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/iit-madras-applied-materials-agreement-research-ai\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10208,"title":"RIL invests in Visual Analytics through NetraDyne","content":"Reliance Industries Limited (RIL) has intimated the stock exchanges, BSE and NSE, that the company would be investing US$16 million (approximately Rs. 108 crore) in NetraDyne, a US-based visual analytics start-up. The investment will be carried out through RIL’s wholly owned subsidiary, Reliance Industrial Investments and Holdings Limited (RIIHL) in convertible preferential shares of NetraDyne. NetraDyne, the visual analytic start-up, aims at creating a unique end-to-end IoT solution thereby pushing Deep Learning processing to the edge. It wishes to deliver vision based solutions across various industries by focusing on technologies which use cameras as their main sensory input. NetraDyne was found by Avneesh Agrawal, former Qualcomm India and South Asia President and David Julian, co-founder Mountain View Analytics-a financial investment startup, in September 2015. The company is headquartered in San Diego, California in United States. The start-up is developing deep learning solutions and vision-based analytics which will be targeted at industries such as fleet management, automotive, security and surveillance, as per RIL’s intimation to stock exchange. The company is still in its research stage and is yet to start commercial operations. 50 % of the said investment has already been paid by Reliance to the visual analytics start-up by 31st May 2016. The remaining 50% will be shared by 31st May 2017. RIIHL which will be investing in NetraDyne through convertible preferential shares will hold 40% stake in the company once the preference shares are converted. Reliance Industries, which operates India’s largest private-sector refinery, has been actively investing in start-ups and this move is a part of the larger activity. More so, analytics is a buzzing industry and investing in the visual analytics space, which is still a niche, will help RIL continue its efforts to be in the space of innovative technology.","excerpt":"Reliance Industries Limited (RIL) has intimated the stock exchanges, BSE and NSE, that the company would be investing US$16 million (approximately Rs. 108 crore) in NetraDyne, a US-based visual analytics start-up. The investment will be carried out through RIL’s wholly owned subsidiary, Reliance Industrial Investments and Holdings Limited (RIIHL) in convertible preferential shares of NetraDyne. […]","categories":["AI News"],"tags":["RIL"],"author_name":"Manisha Salecha","publish_date":"2016-06-16T09:54:41","publication_year":"2016","word_count":293,"keywords":["Go","programming_languages:R","AI","Aim","deep learning","ViT","analytics","RIL","Julia","R","startup"],"extracted_tech_keywords":["AI","deep learning","analytics","Aim","R","Go","Julia","ViT","startup","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/ril-invests-visual-analytics-netradyne\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10170934,"title":"Meta AI Now Has 1 Billion Monthly Active Users","content":"Meta’s AI assistant now boasts 1 billion monthly active users across its apps like WhatsApp, Facebook, and Messenger, revealed CEO Mark Zuckerberg in a shareholders meeting on Wednesday. This is an increase from the 600 million monthly active users reported last December. In comparison, Google’s Gemini AI App has 400 million monthly active users, whereas ChatGPT reported having 500 million weekly active users as of last month. Meta’s AI assistant is integrated into apps like WhatsApp and Instagram, which are significantly popular with 3 billion and 2 billion monthly active users, respectively. The company will also explore opportunities for a paid subscription service “so that people can pay to use more compute”, reported CNBC, citing the shareholders meeting. The report added that the meeting involved investors voting on several shareholder proposals concerning child safety, environmental concerns, and a Bitcoin treasury assessment. Recently, it was reported that Meta is restructuring its AI operations. According to an internal memo obtained by Axios, Meta’s chief product officer, Chris Cox, announced the formation of two new teams: an AI Products team, led by Connor Hayes, and an AGI Foundations unit, co-headed by Ahmad Al-Dahle and Amir Frenkel. The move aims to streamline development and clarify responsibilities across AI initiatives. The AI Products group will manage products such as the Meta AI assistant, AI Studio, and generative AI tools embedded in platforms like Facebook, Instagram, and WhatsApp. The AGI Foundations team will focus on Meta’s Llama models, alongside advancements in reasoning, multimedia and voice. The company’s core research arm, FAIR (Fundamental AI Research), remains separate, and one multimedia-focused subgroup within FAIR will shift to the AGI Foundations team. Last month, Meta released the first version of its standalone Meta AI app, designed to offer an assistant experience across devices, including Meta’s AI-powered Ray-Ban glasses and the web.","excerpt":"The company will also explore opportunities for a paid subscription service “so that people can pay to use more compute.”","categories":["AI News"],"tags":["AI (Artificial Intelligence)","LLaMA","Meta"],"author_name":"Supreeth Koundinya","publish_date":"2025-05-29T12:41:27","publication_year":"2025","word_count":302,"keywords":["Go","ChatGPT","Meta","Meta AI","AI","ML","LLaMA","GPT","Ray","Aim","generative AI","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","ML","generative AI","ChatGPT","Meta AI","Aim","Ray","R","Go","GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/meta-ai-now-has-1-billion-monthly-active-users\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10039075,"title":"Industry Reactions To Startup India Seed Fund Launch","content":"“Never forget, every major company was once a startup.” Taranjit Singh Sandhu, Ambassador of India to the US. Piyush Goyal, Minister of Railways, Commerce and Industry, Consumer Affairs and Food and Public Distribution, recently launched the Startup India Seed Fund Scheme. The government has dedicated a corpus of 945 crore rupees to provide seed funding to qualified startups through eligible incubators across India in the next four years. The programme aims to help startups with proof of concept, prototype creation, product trials, market-entry, and commercialisation costs. “I hope good ideas don’t get sold very cheap, particularly to foreign investors, who are able to identify a good idea,” he said. Anuj Krishna, Co-founder and Director, TheMathCompany, said, “In today’s highly competitive market, a majority of startups never move past the initial stages or even see the light of day because they lack the early capital\/investment required to sustain themselves.” “As a startup, it’s very important to be able to nurture business ideas without any constraints. To this end, such funding will be pivotal to encouraging the right kind of growth in India’s promising startup environment,” said Krishna. A springboard India is currently the third-largest startup ecosystem globally and is home to 21 unicorns valued at $73.2 billion. Abhishek Aggarwal, Co-Founder & CEO, Trinity Gaming, said, “Startups play a significant role in a country’s economic growth, especially for a country like India. The recent announcement of the Seed Fund Scheme is expected to provide entrepreneurs with a safe space that will avail them an environment to grow and inject competition into the industry.” “Financial assistance provided by the government gives confidence and ensures healthy and constant growth of a startup ecosystem,” said Aggarwal. Ravinder Kumar, Founder & CEO, WIMWIsure, said, “The seed funding scheme will provide a new platform for the startups to raise funds in the initial days of the entrepreneurial journey. The majority of startups fail to go beyond the idea stage due to a lack of funds for initial deployment and identifying product-market fit.” The dedicated seed fund will provide an option for new-age entrepreneurs to access funds that are not available through the traditional mode of loans and raising capital from family and friends, he added. Tarun Gupta, Founder, Ultimate Battle, said, “Startup funding in India is mostly being done by foreign investments. Although this has helped the Indian market grow, it has left a chink in the Indian economy, as these startups are owned by foreign firms. This seed fund scheme by the government will empower the Indian incubation hubs domestically and will in turn help to boost the Indian economy.” The road ahead As per a survey, only seven percent of the startups have enough runway to see through the pandemic-induced crisis, while 38 percent said their war chests wouldn’t last for more than three months. Moreover, only 16 percent reported a growth in business during these troubled times. “Accelerators and incubators will be instrumental in finding and evaluating such startups and early-stage ideas. Permissions should be given to the respective evaluation body, and progress should be measured proactively in terms of values and timelines of scouting ideas and fund disbursements,” said Ravinder Kumar. “The focus shall be given to incubators and accelerators at educational institutions to drive confidence in the student community for entrepreneurship. A lot of new venture ideas are pursued by students in their final years of education and often dropped due to financial constraints,” he further added. A robust and transparent policy framework, freedom from red-tapism and constant hand-holding are critical to a flourishing startups ecosystem. The new startup fund scheme is a step in the right direction.","excerpt":"The government has dedicated a corpus of 945 crore rupees to provide seed funding to qualified startups in the next four years.","categories":["AI Startups"],"tags":["AI Startups"],"author_name":"kumar Gandharv","publish_date":"2021-04-28T13:00:00","publication_year":"2021","word_count":606,"keywords":["Go","API","funding","unicorn","programming_languages:R","AI","RAG","Aim","R","AI Startups","startup"],"extracted_tech_keywords":["AI","Aim","RAG","R","Go","API","startup","unicorn","funding","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-startups\/industry-reactions-to-startup-india-seed-fund-launch\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10123360,"title":"Antler to Invest $10 Million in Early-Stage Indian Startups Over Next Six Months","content":"Antler, the venture capital firm, announced plans to invest $10 million in early-stage Indian startups over the next six months. Partners Rajiv Srivatsa and Nitin Sharma will lead the initiative, targeting idea-stage founders with investments of $500,000 per company, totaling 20 investments. This $10 million is part of the $75 million that Antler committed for India. Srivatsa shared the news on X, emphasising the timeliness of building startups in India. “NOW is the best time to build a startup in India!,” he posted. He also provided further details on the investment: “Towards this, Nitin Sharma & I are committing $10M from @AntlerIndia into exceptional idea-stage founders in the coming 6 months. $500K per company X 20 investments – We will work with you in your ambiguous -1 to 0 phase and get you >$1Mn in total within 6-9 months of starting up!” NOW is the best time to build a startup in India!Towards this, @nitinsharma1 & I are committing $10M from @AntlerIndia into exceptional idea-stage founders in the coming 6 months. $500K per company X 20 investments – We will work with you in your ambiguous -1 to 0 phase and…— Rajiv Srivatsa (@telljeeves) June 12, 2024 Srivatsa encouraged aspiring entrepreneurs to write “Build” in the comments to get a message from the Antler team. He also invited those with startup ideas or interest to tag others and join an AMA next week for more details. Antler launched the $285 million Antler Elevate Fund in June 2023, focusing on Series A rounds and later stages, with investments ranging from $1 million to $10 million. Since its establishment in 2018 in Singapore, Antler has expanded its investments and networks globally, covering regions such as the U.S., Europe, Africa, and Asia, including Vietnam, Japan, and Malaysia.","excerpt":"‘NOW is the best time to build a startup in India!”","categories":["Deep Tech"],"tags":["AI Startups"],"author_name":"Siddharth Jindal","publish_date":"2024-06-12T15:06:56","publication_year":"2024","word_count":294,"keywords":["API","programming_languages:R","AI","venture capital","RAG","ViT","R","AI Startups","startup"],"extracted_tech_keywords":["AI","RAG","R","API","ViT","startup","venture capital","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/antler-to-invest-10-million-in-early-stage-indian-startups\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10040151,"title":"Comprehensive Guide To Dimensionality Reduction For Data Scientists","content":"Data is in many forms, such as numerical data or\/and categorical data in a tabular form, image data, video data, text data, and audio data. The size of data affects the selection of storage space, compute memory, hardware and software configurations such as distributed processing, and so on. The size of data plays an important role in data science, whatever the form of data may be. With trillions of new data generated every day, which require petabytes of storage memory, less sized data is preferred in most cases. Here come some common questions: What factor decides the size of data? Is it a variable? If it is a variable, how can we reduce the size of a given data? This article discusses the answers to these questions. Suppose tabular data is generated for a task using a questionnaire that contains 100 questions. Another tabular data is generated with yet another questionnaire that contains the most important 20 questions for the same task. The former data is said to have 100 features, and the later data is said to have 20 features. In other words, the numbers 100 and 20 can be said to be dimensions of the corresponding datasets. 100 features should need more memory space than 20 features. Thus, the number of features (or dimensions) decide the size of the data. The same task is fulfilled either with 100 features or with 20 features. Therefore, the size of data is a variable linearly proportional to the number of features or dimensions. We say that the most important 20 features can replace the 100 features. But, how to decide those 20 important features? It is difficult to decide the most important features in advance before collecting the data because the features left as less important may contain some valuable information. Hence, the usual habit is to collect data to a greater number of dimensions. The exponential increase in the number of smartphones and high-dimension camera devices yield huge-sized images and videos. The high-dimensional data collection is unavoidable. At the same time, it should be noted that a machine learning or deep learning model needs enormous data for training. The model and hardware memory could not handle a huge volume of big-sized data always—moreover, the higher the dimensions the higher the chance to produce misleading patterns. Here comes the need for dimensionality reduction. With high-dimensional data in hand, dimensionality reduction is the process of extracting the most important dimensions, discarding the unimportant dimensions. Popular dimensionality reduction techniques follow some solid mathematical procedures derived from statistics and\/or linear algebra. Here, we discuss the ideas from some valuable articles and tutorials which guide us in understanding and implementing the different dimensionality reduction techniques. To Begin with The problem of high dimensions has a long history back decades, and there have been numerous attempts to face it. The unnecessary dimensions become a noise that suppresses the true pattern. The programming model will try to learn the noise rather than the true pattern. Moreover, these noises lead to non-convergence in some dynamic problems. Human eyes can comfortably visualize and understand data plotted in two dimensions and, to some extent, in three dimensions. Data analysis tools are mostly developed for low-dimensional data (2D or 3D). Thus analysing and visualizing high-dimensional data is hard and less effective in most cases. The following fields are badly affected by the high-dimensionality of the data: Anomaly DetectionCombinatoricsMachine Learning Read more The approaches for Dimensionality Reduction can be roughly classified into two categories. The first one is to discard less-variance features. The second one is to transform all the features into a few high-variance features. We will have a few of the original features in the former approach that do not undergo any alterations. But in the later approach, we will not have any of the original features, rather, we will have a few mathematically transformed features. The former approach is straightforward. It measures the variance in each feature. It claims that a feature with minimal variance may not have any pattern in it. Therefore, it discards the features in the order of their variance from the lowest to the highest. Backward Feature Elimination, Forward Feature Construction, Low Variance Filter and Lasso Regression are the popular techniques that fall under this category. The later approach claims that even a less-important feature may have a small piece of valuable information. It does not agree with discarding features based on variance analysis. Rather, it generates a set of new low-dimensional features out of the original high-dimensional features through some mathematical transformations developed with linear algebra and statistics. The resulting new features have high variance within each feature. Principal Component Analysis (PCA), Singular Value Decomposition (SVD) and Linear Discriminant Analysis (LDA) are the popular techniques that fall under this category. Read more Algorithms and Examples for Dimensionality Reduction One of the most popular algorithms is Principal Component Analysis. It takes data in the form of a matrix and transforms the large set of variables into smaller sets while still maintaining most of the information from the original matrix. High-dimensional data such as images and videos are usually represented in matrices. Moreover, the approach of PCA is derived from matrix algebra. The Eigenvalues and Eigenvectors are calculated from the given matrix. The sum of the Eigenvalues is the total variance of the data. The Eigenvector corresponding to the highest-valued Eigenvalue is the most contributing feature with the highest variance. This feature is technically termed as the first Principal Component. The feature with the second-highest Eigenvalue is the second Principal Component. Thus, the top few Principal Components contribute to most of the data. Read more Next is to discuss the modules and methods available in Python meant for Principal Component Analysis (PCA) and the methodology to apply it for a Classification problem in Machine Learning. It introduces SciKit-Learn’s decomposition module that offers exclusive methods for principal component-based dimensionality reduction. PCA acts like a BlackBox since it can be used to determine the variability of the independent variables for the dependent variable and cannot be used to see which independent variables are more important for prediction. Read More Apart from PCA, many other techniques can be used to reduce the size of the data – Latent Discriminant Analysis(LDA), Singular Value Decomposition (SVD), Kernel PCA, etc. Linear Discriminant Analysis groups data based on the target (or classes), and it models the variations among those classes (inter-class variations) and similarities within the classes (intra-class similarity). On the other hand, PCA mathematically processes the independent features to calculate the eigenvalues, eigenvectors and covariance matrix to arrive at the principal components. One of the major drawbacks that PCA possesses is that it can process only linear data. KernelPCA is a special form of PCA that helps modeling non-linear data. The class boundaries formed by the end results of KernelPCA is a quadratic or cubic, or higher-order curve, whereas that of PCA is a line. Read more Singular Value Decomposition is a method for matrix decomposition. It is considered as more stable in comparison with EigenValue Decomposition. Singular Value Decomposition (SVD) is available as a method in NumPy’s Linear Algebra module (numpy.linalg.svd). It identifies the principal components and automatically arranges them by rank. The top-ranked components contribute greatly to the original data. For example – To Perform SVD on an image. It decomposes the original image into three components: U matrix, Sigma vector, and V matrix. Those matrices have entries arranged according to their ranks. By selecting a small top subset of these matrices and integrating them, one can obtain a low-dimensional image with most details preserved. Read more To continue with examples, we can also do image compression using PCA with the help of n image data. An image is a matrix-like representation of pixel values. Pixel values are usually real non-negative numbers whose values denote the colour intensity of the corresponding points in the image. Viewing it as a matrix, PCA calculates the eigenvalues, eigenvectors and covariance matrices and reform the entire image data with a few principal components. Similar to SVD, PCA reconstructs a high-dimensional image into a low-dimensional image without losing important information. While declaring that PCA is great in data compression, there are a few situations where PCA can not be incorporated: PCA can be applied to numerical data but not to categorical data. Categorical data must be converted into numeric form by one-hot encoding or any suitable method before applying PCA on it.PCA completely transforms the input data into new data. In other words, it destroys all the original features to create some new ones. So we can not interpret which original features contribute to which portion. When each feature in the input data is important to the model, we might accidentally discard some crucial patterns by reducing the PCA dimensions. It is important to choose the number of resulting principal components correctly. Read more Further Reading Got interested in dimensionality reduction and wish to read more? Here are some good resources. The Curse of Dimensionality High Dimensional Geometry and Dimension ReductionSingular Value DecompositionLinear discriminant analysisSciKit-Learn’s Decomposition moduleSciKit-Learn’s Discriminant Analysis moduleNumPy’s Linear Algebra module","excerpt":"Dimensionality reduction is the process of extracting the most important dimensions, discarding the unimportant dimensions","categories":["Deep Tech"],"tags":["Dimensionality Reduction","Dimensionality Reduction Techniques","feature extraction","feature selection","Guide","linear discriminant analysis","PCA","Principal Component Analysis","Python"],"author_name":"Rajkumar Lakshmanamoorthy","publish_date":"2021-05-15T18:00:00","publication_year":"2021","word_count":1518,"keywords":["data science","scikit-learn","NumPy","machine learning","AI","linear discriminant analysis","Python","Principal Component Analysis","Aim","deep learning","Dimensionality Reduction Techniques","feature extraction","RAG","anomaly detection","Dimensionality Reduction","PCA","feature selection","Guide"],"extracted_tech_keywords":["AI","machine learning","deep learning","data science","Aim","scikit-learn","NumPy","RAG","anomaly detection","Python"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/guide-to-dimensionality-reduction-for-data-scientists\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10042967,"title":"Facebook Introduces Two Datasets To Democratise Conversational AI","content":"Facebook AI has shared new research and two new datasets– TOPv2; and MTOP– to help develop sophisticated and effective Conversational AI systems. According to a Deloitte report, the Conversational AI market is projected to increase from AUD 6 billion in 2019 to AUD 22.6 billion by 2024, with a CAGR of 30.2 percent between 2019 and 2024. Conversational AI systems have many limitations. For example, the improvements in the existing systems are limited to people speaking widely used languages such as English. Also, it’s quite difficult to scale the existing systems to support new use cases. Semantic parsing – the task of converting a natural language utterance to a machine-understandable representation – is a critical component of virtual assistants. The existing natural language understanding (NLU) models depend on huge amounts of annotated training data for this task. However, the rub is, large datasets are not available for less popular languages. The method proposed by the researchers overcomes these limitations. Research Researchers including Xilun Chen, Asish Ghoshal, Yashar Mehdad, Luke Zettlemoyer and Sonal Gupta have published a paper focusing on the adaptation of task-oriented semantic parsers to low-resource domains. The team has proposed a novel method capable of surpassing a supervised neural model at a 10-fold data reduction. Simply put, the new state-of-the-art conversation AI system uses ten times less training data to perform unfamiliar and complex tasks. The study provides details on a multilingual NLU model that outperforms single-language models. In addition, the approach works well to support more diverse use cases in multiple languages. Researchers have improved the NLU models to support a wider range of domains without relying primarily on manually annotated training data. With as few as 25 training examples per intent or slot label, the technique can generate task-oriented semantic parsers for new domains. “Several training strategies exist for domain adaptation. For instance, one can employ joint training that trains a single model with all the available data on both source and target domains. Another approach, which we found superior is the pre-training plus fine-tuning strategy, where a model is first trained on the source domains and then fine-tuned on the low-resource target domains,” as per the paper. “On the other hand, as pre-trained language representations such as RoBERTa or BART are adopted, the latter strategy becomes a 3-stage training process: train RoBERTa\/BART; fine-tune on the source domains; fine-tune again on the target domains.” The three stages include: The first stage, which is out of scope for this paper, is the pre-training stage, where self-supervised language representations are learned.Then, it is fine-tuned on the source domains. The stage is called source training to avoid ambiguity with the final stage. The source-trained model is fine-tuned again on the target domains in the final stage- denoted as fine-tuning. The approach distinguishes itself from previous methods on two fronts: First, the encoder-only pre-trained representations used in existing work are not ideal for the seq2seq model employed in task-oriented semantic parsing, and instead propose to use BART, a pre-trained model with an encoder-decoder architecture.More importantly, researchers adopt optimisation-based meta-learning to improve the model’s generalisation to new target domains with very few training samples. “We collect the TOPv2 dataset, a large-scale multi-domain task-oriented semantic parsing dataset with eight domains and more than 180k annotated samples to evaluate our models, which we release to the research community,” said the researchers. Facebook has also made available MTOP data set – a multilingual task-oriented parsing data set with about 100K total utterances spanning six languages, 11 domains, and 117 intent categories. More information on the data set can be found here.","excerpt":"Facebook AI has shared new research and two new datasets– TOPv2; and MTOP– to help develop sophisticated and effective Conversational AI systems. According to a Deloitte report, the Conversational AI market is projected to increase from AUD 6 billion in 2019 to AUD 22.6 billion by 2024, with a CAGR of 30.2 percent between 2019 […]","categories":["AI Features"],"tags":[],"author_name":"kumar Gandharv","publish_date":"2021-07-07T11:00:00","publication_year":"2021","word_count":597,"keywords":["Go","meta-learning","RoBERTa","programming_languages:R","AI","RPA","virtual assistants","BERT","llm_models:BERT","R"],"extracted_tech_keywords":["AI","virtual assistants","R","Go","BERT","RoBERTa","meta-learning","RPA","llm_models:BERT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/facebook-introduces-two-datasets-to-democratise-conversational-ai\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10099202,"title":"India Catapults the AI Mission","content":"The conversation around building AI in India is finally taking shape and it’s time to take stock of our standing in a world engaged in a blind race of technological advancements. When OpenAI CEO Sam Altman visited India, he said that no one would be able to replicate what OpenAI has done with ChatGPT. To this, Tech Mahindra CEO CP Gurnani had famously said, “Challenge accepted”. And in just about two months, the company launched Project Indus, an indigenous LLM that could speak in many Indic languages. Most recently, at Reliance’s 46th Annual General Meeting, Mukesh Ambani announced his plans to build India-specific AI models. The Jio chief is probably one of those bigwigs from India who can actually challenge OpenAI. “India has scale. India has data. India has talent. But we also need AI-ready digital infrastructure that can handle AI’s immense computational demands,” said Ambani. Through Jio, Ambani was able to give broadband connectivity to everyone and definitely delivered on the promise he made — can he do that with AI? A little too late? Last year, Reliance acquired a 25% stake in Silicon Valley-based Two Platforms, a company focused on advanced technological projects aimed at creating interactive and immersive experiences through AI interactions, for $15 million. The company is also investing in developing up to 2,000 MW of AI-ready computing capacity. When it comes to capital and investment, nothing is going to stop Ambani from making a comparatively greater model. As for the talent, Reliance can hire people from across the globe, to some extent even existing leading companies if they wish to. Alternatively, Ambani can simply outsource it to some other company and build an AI model for India (and maybe call it AI-JIO). Arguably, it might be a little late to come up in the AI space even with Tech Mahindra’s Project Indus still in infancy. Moreover, for the project to be successful, Gurnani has said that it requires contributions from every Indian to build the dataset. An Indic language dataset would just be the first step towards building an AI model to compete with ChatGPT. Still the bid is on and the way from here is only upwards. Emerging stronger Meanwhile, India’s ambition of becoming a semiconductor superpower is still in the making. The government has been investing thousands of crores for the projects, but the delays and inefficiencies in the process have resulted in the loss of one-and-a-half years. To make up for the loss, Semicon India 2023 gave a little hope with all the investments, and partnerships that Vedanta has promised with a ‘world-class technology partner’. Even AMD and Micron are planning to invest in India. The recent success of Chandrayaan-3 mission has also brought India into the spotlight. Becoming the fourth country to land on the Moon, and the first to touchdown on the lunar south-pole has been a huge feat for the nation. This has also pushed countries like Japan to announce partnerships for the future missions with ISRO. Interestingly, Elon Musk’s Starlink is expected to offer satellite services in India. The company has already been in talks with the Indian government to expedite the process. Musk also finds it “impressive” that some of the biggest companies in the world are helmed by people of Indian-origin. Impressive— Elon Musk (@elonmusk) August 26, 2023 Meanwhile, no one can deny the mega success of UPI. The whole world has been trying to replicate the success story in as many ways as possible. Countries like France, Australia, Singapore, Saudi Arabia, Bhutan, Nepal, Oman, Sri Lanka, and Pakistan have already received the UPI technology from India. So, even though building AI systems may seem a little distant on the horizon for us, the country has already proven itself as one of the leaders when it comes to technological advancements, and is striving for more. As of now, he country can lead the way for quantum computing if IBM comes back to India, or be a semiconductor hub or even crack AI. The possibilities are endless.","excerpt":"India is finally making its bid to compete in the AI race with its own generative AI projects","categories":["IT Services"],"tags":["AI in India","India AI","ISRO"],"author_name":"Mohit Pandey","publish_date":"2023-08-29T15:00:00","publication_year":"2023","word_count":670,"keywords":["India AI","Go","ISRO","ChatGPT","API","AI in India","OpenAI","AI","Git","GPT","Ray","Aim","R"],"extracted_tech_keywords":["AI","ChatGPT","OpenAI","Aim","Ray","R","Go","Git","API","GPT"],"url":"https:\/\/analyticsindiamag.com\/it-services\/india-catapults-the-ai-mission\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":52896,"title":"How A Security Camera Vendor Exposed User Data While Performing Analytics","content":"Wyze- a security camera manufacturer admitted to a mistake that cost more than a million users’ personal data to be exposed on the web. According to Wyze, while no passwords or financial information was exposed, user data like email addresses, Wi-Fi network IDs and body metrics were left unprotected between December 4 and December 26, 2019. The incident happened due to a misconfigured Elasticsearch database containing data generated by more than 2.4 million Wyze customers, which the company blamed on one of its employees. The company said it will still continue investigating the lack of appropriate security standards in the datasets. Elasticsearch is a scalable open-source full-text search and analytics engine which lets users store, search and analyze large volumes of data swiftly and in near real-time. How Did The Incident Happen At Wyze? The incident took place during an internal project to determine more efficient ways to measure the business metrics like device activations, failed connection rates, etc. This meant replicating data from the main production servers into a more flexible database, which is easier to query. Queries on large volumes of such are compute-intensive which could impact user product experience if being done on the main database. To perform processing without any system lag, the separate subset of data was transferred into a different database. The data was exposed to the web unencrypted while it was being transferred to a new database to make the data easier to query. This happened because an employee had wiped out previous security protocols during the process due to an accidental error. The new database only had a subset of data and did not include any user passwords, government-regulated personal data, or financial information  — revealed the company’s co-founder Dongsheng Song. Wyze Incident: What Was Exposed? So far the company has only accepted that the data which was left exposed to the web and has mentioned that there is no evidence the data was actually breached. Also, the company further reported that data that was exposed includes health data metrics, email addresses, Wi-Fi network IDs, limited tokens associated with Alexa integrations. Users have been asked to log back in and relink Alexa, Google Assistant, or IFTTT integrations so that new tokens can be generated and users’ data is no longer exposed. Users have also been asked to be cautious against phishing attacks as hackers may have access to user names and email IDs, which could be used to steal credit card information. As far as the main login tokens are concerned, the company stated there was no evidence that log in tokens had been exposed. Nevertheless, users have been signed out of their accounts as a precaution so new login tokens are generated. The company is also working towards ramping up which would include two-factor authentication for its users, as well as rebooting the cameras shortly. Wyze mentioned that would have an extra level of protection to its system databases along with adjusted multiple permission rules and an included precaution to only permit certain whitelisted IPs for accessing the databases. “We’ve often heard people say ‘you pay for what you get,’ assuming Wyze products are less secure because they are less expensive. However, this is not true. We’ve always taken security very seriously, and we’re devastated that we let our users down like this. This is a clear signal that we need to totally revisit all Wyze security guidelines in all aspects, better communicate those protocols to Wyze employees, and bump up priority for user-requested security features beyond 2-factor authentication,” according to co-founder Song. Overview Seattle-based Wyze Labs is looked upon as a highly affordable, smart home camera manufacturer, selling smart security cameras at a price as cheap as $20 compared to cameras from other vendors with the same features  — such as Amazon’s Cloud Cam that costs over $100. Critics have claimed that due to the cheap prices of the cameras, the company hasn’t taken security seriously. But, this has been denied by Wyze co-founder in a blog post. The Wyze data exposure is not an isolated incident involving security leaks during analytics. In the past, we have seen a similar Suprema biometric mass leak which has exposed Elasticsearch records.","excerpt":"Wyze- a security camera manufacturer admitted to a mistake that cost more than a million users’ personal data to be exposed on the web. According to Wyze, while no passwords or financial information was exposed, user data like email addresses, Wi-Fi network IDs and body metrics were left unprotected between December 4 and December 26, […]","categories":["AI Features"],"tags":["Cybersecurity","Security","security analytics","security breach"],"author_name":"Vishal Chawla","publish_date":"2019-12-31T17:00:00","publication_year":"2019","word_count":699,"keywords":["Elasticsearch","Go","Cybersecurity","programming_languages:R","AI","programming_languages:Go","Security","Scala","Aim","analytics","programming_languages:Scala","security breach","R","security analytics"],"extracted_tech_keywords":["AI","analytics","Aim","Elasticsearch","R","Go","Scala","programming_languages:R","programming_languages:Scala","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-a-security-camera-vendor-exposed-user-data-while-performing-analytics\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10087848,"title":"Coca-Cola Gets Marketing Makeover with OpenAI and Bain &amp; Co&#8217;s Partnership","content":"Global management consulting firm Bain & Company has joined hands with AI research lab OpenAI to expand its potential business applications by combining OpenAI’s technology, especially Generative AI, with business strategy and social responsibility. According to Bain & Company’s blog post, it is offering its advanced analytics practice to help clients with the practical application of AI by using experience across the value chain to spot generative AI use cases, deploy a proof-of-concept, and implement the capabilities across a client’s operating model. They also provide guidance on ethics, rights, ownership, and change management. The company claims that partnership with OpenAI will enable them to “stay ahead of the curve” and help clients gain a competitive advantage. Widely popular consumer goods company Coca-Cola is the first company to join the new alliance and implement Generative AI. Coca-Cola recently launched its newest Creations flavour, Coca-Cola Move, in collaboration with global ‘atypical’ pop star Rosalía. A scannable QR code on the beverage directs users to the Coca-Cola Creations Hub, where they can access behind-the-scenes material, a personality test, and a music video with all versions of her. Zack Kass, the head of go-to-market at OpenAI, commended the Coca-Cola’s new approach to adopting cutting edge tech as “the most ambitious we have seen of any consumer products company.” Read more: Generative AI Unlocking Floodgates to Solve Data Scarcity From Non-Profit to Profit: OpenAI’s Strive for Financial Sustainability OpenAI, which initially started off as a non-profit organisation, declared the OpenAI LP as a new, distinct entity with the aim of making a profit, calling it a ‘capped-profit’ corporation. Although the company will continue to research and create new technologies, it also wants to generate more revenue. Investors can receive up to 100 times the amount of their investment, with any additional funds going to support the non-profit work of OpenAI’s separate entity called ‘OpenAI Nonprofit’. OpenAI has received funding from renowned venture capital firms such as Andreessen Horowitz, Sequoia, and Founders Fund, all of which have the option to withdraw from the agreement at any point. Interestingly, the company’s decision to become a profit-making entity coincided with the release of its text-to-image generator, DALL.E, and its successor, DALL.E 2, which was launched last year and sparked a wave of generative AI research and startups.","excerpt":"Generative AI is the talk of the town.","categories":["AI News"],"tags":["AI Tool","ChatGPT","coca-cola","DALL.E","OpenAI"],"author_name":"Shritama Saha","publish_date":"2023-02-22T12:44:36","publication_year":"2023","word_count":379,"keywords":["Go","ChatGPT","API","OpenAI","coca-cola","AI","RAG","Aim","analytics","generative AI","GAN","AI Tool","R","DALL.E"],"extracted_tech_keywords":["AI","analytics","generative AI","OpenAI","Aim","RAG","R","Go","API","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/coca-cola-gets-marketing-makeover-with-openai-and-bain-cos-partnership\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10072564,"title":"Council Post: Coming to Understanding Why India’s Diversity Requires Large Language Models to Advance AI","content":"Did she say “guma hua saamaan,” “Gumi’ā samāna” or “Iḻanta cāmāṉkal?” In the U.S., “lost luggage” is a relatively easy phrase for an airline’s AI chatbots to understand before directing a customer’s call to an agent. This is not the case across India, with its 22 official languages — including Hindi, Punjabi and Kannada — and more than 19,500 dialects. There is a growing population of users consuming Indian language content (print and digital, for government and businesses) and the growth of the large language model (LLM) is benefiting the research and development of tools for Indian languages. LLM is the first step to building local language technologies with a common framework across various languages in the country. On track to surpass China as the world’s most populous country, India — with its diversity, massive amounts of stored data, and need to make communication more efficient and seamless — is a ripe test bed to deploy LLMs and natural language processing (NLP) at scale. Because of the nation’s unique linguistic diversity, Indian researchers have even coined a specific term for addressing the need to make communication more efficient and seamless: Indic language understanding, or ILU. The Indian government, higher education and research (HER), and private sectors are turning to these LLMs to better customer service, refine operations, deliver greater business intelligence and create new business models – all while enterprises work to overcome labor shortages and optimize complex supply chains. Indeed, having machines understand written words — or natural language understanding (NLU) — basically maps every word to a meaningful vector and helps bring huge benefits to businesses. Large or small, every organization has an abundance of text-based data – including legal documents, patient records, financial documents, code, websites and emails. LLMs provide the ability to use that unstructured information and develop new services to help businesses stay competitive. LLMs and the India Opportunity The Indian government, HER segment and private sector have huge datasets stored in various languages and formats (e.g., text, audio, video and more). These datasets need to be cleaned and labeled, and can become the foundation to train LLMs. Fundamentally, ILU research seeks to make it easier for India’s billion-plus citizens to communicate in different languages and use the power of AI for multimodal NLU\/P. To begin technologies with a common language framework, the Indian private and governmental sectors must clean and label all of their data to use as the foundation for training LLMs. India has been an important market in which tech giants have made breakthrough progress over the past decade. However, the success of these companies require high-compute capabilities, crucial data comprising diverse vocabulary, speakers, etc., and other state-of-the-art transformer models. Use cases are myriad. They include using a smartphone camera to scan signs and generate text or voice output from one Indian language to another; multilingual voice translation and response by service bots and concierges; translation on a mobile device of voice in one language to screen text in another; as well as making government-to-citizen services multilingual and therefore more efficient. Bolting heavy-duty computational power onto these large datasets and LLMs — be it in a data center, in the cloud, at the edge or in a hybrid format — leads to miracles. These gains shorten the data preparation time (extract transform load or ETL pipeline), accelerate training time, and enable faster deployment, inference and continuous model accuracy improvements. To summarize, a combination of existing big data, ever-large and more accurate LLMs, scalable accelerated computing, and Indian research efforts can use AI to solve one of the country’s greatest challenges: achieving seamless, fast, and accurate multilingual and modal communication. Massive Models Put Troves of Business Documents to Work The size of LLMs has been increasing 10x annually for the last few years, with state-of-the-art models now containing hundreds of billions to trillions of parameters. As these models grow in complexity and size, their capabilities improve as well. Fast, efficient and accurate ILU technology is not far from reality, given the recent evolution of language models – including Open AI GPT-3; Switch Transfer, GLAM and PALM from Google; Turing NLG from Microsoft; Gopher from DeepMind; and Jurassic from AI21 Labs. Models like OpenAI’s GPT-3 and the new open-source BLOOM can perform specialized tasks such as answering questions, writing essays, summarizing text, translating languages and generating computer code. This enables enterprises to build customized NLP applications tailored to understand an enterprise’s unique vocabulary, customer relationships and datasets on which their business runs. Transformer-based LLMs are reshaping today’s AI landscape, and NVIDIA has upped the game with the NVIDIA NeMo Megatron framework for training language models with trillions of parameters. NeMo Megatron offers end-to-end capabilities from data curation to training to inference and evaluation. It opens doors for enterprises across the world to understand, develop and deploy their own LLMs to help them build domain-specific chatbots, personal assistants and other AI applications. With 530 billion parameters, the Megatron-Turing NLG model — one of the world’s most powerful transformer language models — leverages the NeMo Megatron framework and was used in the development of BLOOM, the world’s largest open-source multilingual LLM. Advancing Healthcare with Language Understanding LLMs can be useful across the healthcare industry, as well. Some 80% of the information in electronic health records consists of unstructured clinical notes. NLP models can be used to easily surface relevant clinical data for downstream diagnostic and predictive tasks. These models also improve the patient experience with smarter chatbots and can reduce physician burnout through improved transcription and summarization tools, faster documentation, and better decision-support systems. A Launch Pad for Learning More on Large Language Models As LLMs are applied to new use cases and domains from communications to healthcare, their complexity and size have increased exponentially. Developing LLMs requires computing on a scale that goes beyond the mainstream, as well as software and frameworks built to connect and synthesize across all of the model’s layers. Deploying an LLM into production requires expertise, and customized labs are available to help enterprises take a step forward in putting language to work for their business. This article is written by a member of the AIM Leaders Council. AIM Leaders Council is an invitation-only forum of senior executives in the Data Science and Analytics industry. To check if you are eligible for a membership, please fill out the form here.","excerpt":"Across healthcare, retail, telecommunications, customer service and operations, large language models (LLMs) are enabling Indian government organizations, research institutions and enterprises to overcome communications challenges and find new connections.","categories":["AI Features"],"tags":[],"author_name":"Vishal Dhupar","publish_date":"2022-08-10T15:00:00","publication_year":"2022","word_count":1059,"keywords":["data science","TPU","OpenAI","AI","chatbots","ML","RAG","NLP","Aim","analytics"],"extracted_tech_keywords":["AI","ML","NLP","data science","analytics","OpenAI","Aim","RAG","chatbots","TPU"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/coming-to-understanding-why-indias-diversity-requires-large-language-models-to-advance-ai\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10093636,"title":"Are You Worried About AI Stealing Your Job?","content":"Last month, under r\/ChatGPT, a Reddit user narrated his story of how he has achieved an “extremely high interview invitation rate” by using ChatGPT for applying for jobs. The user would upload his resume on ChatGPT, ask specifically tailored roles for his experience, ask the chatbot to outstandingly answer questions for the role, and thus make his application “amazing” according to the feedback he got. Another user tells a similar story where he was applying for jobs for weeks but could not get to an interview. Once he took help from ChatGPT to write a cover letter for a job and enhance his resume, he got back an interview invite the same day. In both cases, it is clear that ChatGPT can get you interviews for jobs. But, the first story also narrates how the person could not get the job because he got “anxious during interviews”. Could it be that the cover letter and resumes got selected because a lot of companies are actually running ChatGPT or similar products in the HR department? “What a sound cover letter!” is what ChatGPT would say to such AI-generated cover letters. On the other hand, these AI tools are threatening to wipe out millions of jobs around the globe. According to a Goldman Sachs report on the potential of generative AI, 300 million jobs could be automated by AI. The same report also reads that automation of any task historically has offset this loss of jobs by creating new jobs and occupations, resulting in long-run employment growth. Though, the specific job roles have not been specified. OpenAI is also aware of the disruption in the job market that its technology has, and will eventually cause, and also published a paper talking about the same in March. We are already seeing companies like IBM freezing hiring for jobs that could be replaced by AI. On the other hand, PwC predicted that AI will create as many jobs as it will displace, but there would be different “winners” and “losers” by industry sectors. But this was back in 2018. There are new jobs in the market such as prompt engineering, that provide a salary higher than even full-stack developers, but the number of jobs that are getting displaced and a number of people getting laid off is higher than the ones that are created. AI Money Generator Two months back, Jackson Greathouse Fall, a writer and a brand designer, used ChatGPT to guide him to becoming a millionaire. After telling the chatbot, “You have $100, and your goal is to turn that into as much money as possible in the shortest time possible, without doing anything illegal”, Fall got instructions on how to launch a business called Green Gadget Guru, offering products to enable people to live sustainably. https:\/\/twitter.com\/jacksonfall\/status\/1636107218859745286 After this, he managed to raise $1,378 in just one day, got investments, and his company got a valuation of $25,000. Apart from running a business, ChatGPT has enabled users to acquire multiple full-time jobs simultaneously – referred to as “overemployment”. The trend started during COVID, and now ChatGPT has made it possible for a lot more people. The funny, and at the same time concerning, part is that their employers have no idea! One said, “ChatGPT does 80% of my job,” which has allowed him to apply for multiple positions, and finish their work in half the usual required time. Moreover, there is a complete Reddit thread where “overemployed” people talk about how they are landing different jobs, working on more than two jobs at the same time, and looking for even more opportunities. On the other hand, companies have been restricting people to use AI for jobs. Most recently, Apple announced that is not going to allow its employees to use ChatGPT internally. Is it fair that while some people are overemployed with this chatbot, but people in various roles are being restricted to use it amid the fear of layoffs? Is AI Stealing My Salary? One might be quick to celebrate and admire these “overemployed” workers getting chunks of money through different jobs. But if we look at this phenomenon a little critically, the case gets a little concerning. Roles that would have been filled by another person are now getting filled by a single person using ChatGPT and similar AI products. Seems unfair for people not trained enough with AI. This is what Richard Baldwin said at the 2023 World Economic Forum. “AI won’t take your job, it’s somebody using AI that will take your job.” There is no doubt that AI is displacing jobs globally and people from various sectors are increasingly revolting against its use. It would have not been possible for the “overemployed” people to get these jobs if ChatGPT did not assist them. Moreover, OpenAI’s ChatGPT has not just caused unemployment indirectly, but also directly. According to another Reddit post, the user explains how ChatGPT is slowly taking his job away. He has been working as an ML engineer at a company and was building conversational agents, similar to ChatGPT. But after OpenAI released ChatGPT API, the company adopted it, replaced the already built ML models, and started planning to remove the ML team entirely to cut costs. There is no doubt that a lot of jobs are still safe from this AI phenomenon, but a lot of them are not. Coding too is slowly facing the brunt of its own development, with a lot of auto code generators shrinking a 10-people job to just one human and AI partnership. Apart from prompt engineering, several other jobs are being introduced that now consider ChatGPT expertise as a relevant skill. It is now evident that the time to train humans to work along with AI is here. If you don’t do it right now, someone else will.","excerpt":"People have been using ChatGPT to get multiple jobs. On the other hand, people have been losing jobs because of AI. Is it fair?","categories":["AI Features"],"tags":["AI Jobs","AI Replacing Jobs","AI Tool","ChatGPT","prompt engineering jobs"],"author_name":"Mohit Pandey","publish_date":"2023-05-19T17:00:00","publication_year":"2023","word_count":969,"keywords":["Go","ChatGPT","API","prompt engineering jobs","OpenAI","AI","ML","AI Jobs","AI Replacing Jobs","prompt engineering","generative AI","conversational agents","AI Tool","R"],"extracted_tech_keywords":["AI","ML","generative AI","ChatGPT","OpenAI","conversational agents","prompt engineering","R","Go","API"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/are-you-worried-about-ai-stealing-your-job\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":63808,"title":"This Hyderabad-Based Data Analytics Startup Raises $4 Million Funding Led By IvyCap Ventures &#038; Sequoia Capital","content":"Convosight, a Hyderabad-based data analytics startup, has announced that the company has raised $4 million as a pre-series A funding led by IvyCap Ventures and Sequoia Capital. The company plans to use the funding in order to enhance its product development process, including hiring across product and marketing and improve its technology and data science capabilities. According to Tamanna Dhamija, the co-founder and CEO of the company, the funding will also be utilised to increase the client base of the startup, which mainly includes community entrepreneurs. The startup will also ramp up hiring across product and marketing. The company also stated that post the funding, IvyCap Ventures founder and managing partner Vikram Gupta would join Convosight’s board. This Hyderabad-based startup, founded in October 2019, is the world’s first community management platform built to help Facebook Groups to grow, engage, and monetise by using technologies like data analytics and machine learning. The platform helps Facebook group admins to track growth trends, conversation insights, control spam and manage content through smart recommendations, and in turn help them to monetise their communities through commerce, subscriptions, offline events and brand partnerships. According to the company, since Convosight’s launch in late 2019, it is already used by 2500 Facebook communities which have over 30 million community members across the world. The company, currently, has a vast customer base, with names fro Fortune 100 companies like Nestle, J&J, Reckitt Benckiser, P&G and Abbott. According to Dhamija, in today’s era, there are millions of people actively engaged in online social media communities like Facebook. “These communities are the future of how people will interact online,” said Dhamija. Therefore, the company believes that it is imperative for creating meaningful communities in order to help people in crisis. And that’s why Convosight has been providing community admins with the necessary tools to build and sustain meaningful communities on social media.","excerpt":"Convosight, a Hyderabad-based data analytics startup, has announced that the company has raised $4 million as a pre-series A funding led by IvyCap Ventures and Sequoia Capital.  The company plans to use the funding in order to enhance its product development process, including hiring across product and marketing and improve its technology and data science […]","categories":["AI News"],"tags":["Data Analytics","Startups"],"author_name":"Sejuti Das","publish_date":"2020-04-28T12:34:35","publication_year":"2020","word_count":311,"keywords":["data science","API","funding","machine learning","programming_languages:R","AI","analytics","Data Analytics","Startups","R","startup"],"extracted_tech_keywords":["AI","machine learning","data science","analytics","R","API","startup","funding","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/this-hyderabad-based-data-analytics-startup-raises-4-million-funding-led-by-ivycap-ventures-sequoia-capital\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10013136,"title":"Top 8 Ransomware Attacks of 2020 That Shook The Internet","content":"This year, ransomware attacks are one of the most common cyber attacks among organisations. Ransomware is a type of malicious software that infects a computer system as well as displays messages that demand a price to be paid in order to mitigate the issue. The common types of ransomware include Crypto malware, Maze, Doxware, Scareware, Lockers, RaaS and others. Also, some of the popular ransomware attacks include Cryptolocker, TeslaCrypt, Petya, Bad Rabbit, among others. Below here, we mentioned the top eight ransomware attacks, in no particular order, that has happened in 2020. Cognizant Ransomware Attack In the month of April, IT services giant — Cognizant admitted that the company had faced a ransomware attack. The IT giant revealed that its network had been infected with Maze ransomware, which is a ransomware group known for releasing stolen data to the public if the victim does not pay to decrypt it. According to sources, the revenue and corresponding margin impact of the ransomware attack is expected to be in between $50 million to $70 million in the second quarter of 2020. Magellan Health Ransomware Attack In April this year, one of Fortune 500 companies, Magellan Health also announced being faced by a ransomware attack and data breach. According to sources, the attackers launched a ransomware attack on the provider and impersonated a client at Magellan Health. This enabled the hackers to get access to the health plan’s servers. The breached data is said to be around 1.7 million that included personal information of both internal and external customers. Communications & Power Industries (CPI) Ransomware Attack At the beginning of this year, one of the major electronics manufacturers for defence and communications markets, California-based Communications & Power Industries (CPI) faced a severe ransomware attack. According to sources, the defence contractor paid a ransom of about $500,000 shortly after the incident in mid-January. As per reports, a domain admin clicked on a malicious link while they were logged in that immediately triggered the file-encrypting malware. University of California San Francisco (UCSF) Ransomware Attack Further, in June, the University of California San Francisco (UCSF) staff detected a ransomware attack. As per UCSF authority, the attack occurred in a limited part of the UCSF School of Medicine’s IT environment. According to a blog post, the encrypted data is crucial to some of the academic work. They paid some portion of the ransom, which is approximately $1.14 million, to the individuals behind the malware attack in exchange for a tool to unlock the encrypted data and the return of the data they obtained. Baltimore County Public Schools Attack (1\/3) Due to the recent ransomware attack, Baltimore County Public Schools will be closed for students on Monday, November 30, and Tuesday, December 1. BCPS offices will be open and staff will receive additional information about Monday and Tuesday.— Baltimore County Public Schools (@BaltCoPS) November 28, 2020 A few days ago, even Baltimore County public schools were closed due to a severe ransomware attack. According to sources, the ransomware attack on a school district in Maryland has halted classes for more than 115,000 pupils. This means that for Baltimore County public schools district, all pupils learning remotely because of the pandemic could suddenly no longer access lessons. Advantech Ransomware Attack Recently, the industrial IoT chip manufacturer Advantech had also faced a severe ransomware attack. According to sources, the chip manufacturer fell under the trap of Conti ransomware attack, which is relatively new ransomware and it is known to be the successor to the Ryuk ransomware. The hackers breached the data and demanded a ransom of 750 BTC from the vendors. Carnival Corporation Suffers Ransomware Attack Leaving no industry behind, in the month of August, the world’s largest cruise line operator, Carnival Corporation had disclosed a ransomware attack. The attackers breached a portion of data from the systems. According to sources, the unauthorised third party gained access to certain personal information relating to some guests, employees and crew for three of the corporation’s brands, which are Carnival Cruise Line, Holland America Line and Seabourn, as well as casino operations. Canon Ransomware Attack In the same month, the popular camera manufacturer, Canon also faced a severe ransomware attack. According to sources, the company confirmed that the attack was caused by ransomware and the cybercriminals responsible stole data from its company servers. In a blog post, the company also revealed that an issue “involving 10GB of data storage” was under investigation, leading to the temporary suspension of related mobile apps and the online platform.","excerpt":"This year, ransomware attacks are one of the most common cyber attacks among organisations. Ransomware is a type of malicious software that infects a computer system as well as displays messages that demand a price to be paid in order to mitigate the issue. The common types of ransomware include Crypto malware, Maze, Doxware, Scareware, […]","categories":["AI Trends"],"tags":["data breach","Ransomware"],"author_name":"Ambika Choudhury","publish_date":"2020-12-03T14:00:12","publication_year":"2020","word_count":752,"keywords":["Go","programming_languages:R","AI","Ransomware","programming_languages:Go","data breach","RAG","GAN","R"],"extracted_tech_keywords":["AI","RAG","R","Go","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-8-ransomware-attacks-of-2020-that-shook-the-internet\/","complexity_score":3,"technical_depth":7,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10172330,"title":"Vodafone, Cyient Launch AI-Powered Network Management Solution","content":"Vodafone and Cyient have launched VISMON, a new AI-driven network configuration management platform designed to streamline network operations. The solution aims to improve network management efficiency through data-driven analytics. The collaboration combines Vodafone’s telecommunications expertise with Cyient’s engineering solutions. VISMON unifies configuration data and inventory across multiple Vodafone markets, enhancing visibility, detecting anomalies, and driving more efficient decision-making. “VISMON provides the strategic foundation to oversee configuration data across all markets, enabling us to harmonise practices, identify best-performing setups, and optimise our networks more effectively,” Mostafa Noureldien, Vodafone’s manager of network development digital strategy, said. VISMON has already delivered significant improvements, including a 70% reduction in the time spent compiling reports and three times faster decision-making. It is expected to reduce errors caused by inconsistent configurations by 50%. It uses AI to integrate configuration data and logical and physical inventory, improving network management efficiency across various Vodafone markets. Joaquim Croca, vice president at Cyient, commented, “This AI-driven platform exemplifies how intelligent automation and data-led insights can drive real impact—delivering agility, consistency, and strategic clarity at scale.” Last year, Vodafone and Microsoft announced a 10-year strategic partnership focusing on generative AI. The aim was to utilise their strengths in providing digital platforms to over 300 million businesses, public sector organisations, and consumers in Europe and Africa. Vodafone was to invest $1.5 billion over the next decade in cloud and customer-focused AI services developed with Microsoft. The partnership would transform Vodafone’s customer experience using Microsoft’s generative AI, scale Vodafone’s managed IoT connectivity platform, develop new digital and financial services, and revamp its global data centre cloud strategy.","excerpt":"The collaboration combines Vodafone’s telecommunications expertise with Cyient’s engineering solutions.","categories":["AI News"],"tags":["Cyient","Telecom India","Vodafone"],"author_name":"Sanjana Gupta","publish_date":"2025-06-24T17:58:17","publication_year":"2025","word_count":264,"keywords":["AI","R","ML","Git","automation","Aim","ViT","analytics","generative AI","GAN","Cyient","Vodafone","Telecom India"],"extracted_tech_keywords":["AI","ML","analytics","generative AI","Aim","R","Git","GAN","ViT","automation"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/vodafone-cyient-launch-ai-powered-network-management-solution\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10044846,"title":"MachineHack Is Back With Weekend Hackathon Edition #2 — The Last Hacker Standing","content":"MachineHack is back with Weekend Hackathon Edition #2 — The Last Hacker Standing. The Weekend Hackathon Edition #2 will be held for six weeks, starting 30th July to 9th September 2021. As part of the competition, we will release a new problem statement every Friday for the participants to solve within seven days and win exciting prizes. Week 01: Problem statement & description Understanding customer emotions is key for businesses to gauge their brand reputation and improve customer experience. Companies leverage sentiment analysis to understand customer data and their perception of brands, products or services. Although businesses collect massive data on a daily basis from emails, complaints, queries, support tickets, social media, executive chats, surveys, articles etc, 80-90% of such data are unstructured. Hence, businesses look for ways to manage the raw data and create a model to automatically measure customer sentiments to make informed decisions about their brand. In this weekend hackathon, MachineHack challenges all the data scientists and machine learning practitioners to create a highly scalable Sentiment Analysis Model to accurately classify customer sentiments about a certain brand (Company ABC) based on the data. The challenge will start on 30th July 2021 at 8:00 PM (IST) Click here to participate. Overview In this challenge, the participants will work on a data mix of Reviews and Tweets. The dataset collected for training has 44,100 rows with four columns of tweet_id; author; content; sentiment including sentiment (3 sentiments – 2-Positive, 0- Negative, 1- Neutral) as a target variable. On the other hand, the dataset for testing the model includes 18,900 rows and three columns and doesn’t include the target variable. The hackathon calls for a few prerequisite skills such as text pre-processing, including lemmatization, tokenization, N-Grams and other relevant methods, multi-class classification and optimising Log Loss. The submission will be evaluated using the Log Loss metric, and to generate a valid submission file, the participants must use scikit-learn models. The hackathon also supports private and public leaderboards. The public leaderboard will be evaluated on 30% of test data, and the private leaderboard will be evaluated on 100% of test data, to be made available at the end of the hackathon. The submission limit for this hackathon is one account per participant. The advanced-level challenge will allow data scientists and machine learning practitioners to get a glimpse into real-life sentiment analysis modelling. The hackathon will end on 5th August 2021 at 6:00 PM (IST) The top three winners of this hackathon will get free passes to the Deep Learning DevCon 2021 (DLDC), scheduled to be held on 23-24 September 2021. In addition, the winners will also get a chance to improve their Global Leader-Board Rankings & be the ultimate MachineHack Grand Master. Click here to participate. Dataset description: Train.csv — 44100 rows x 4 columns (Includes Sentiment as target variable)Test.csv — 18900 rows x 3 columns (Doesn’t include the target variable) Evaluation metric: Log Loss Skills: Text Pre-processing – Lemmatization , Tokenization, N-Grams and other relevant methodsMulti-Class ClassificationOptimizing Log Loss Click here to participate in this hackathon.","excerpt":"The first challenge in MachineHack’s Weekend Hackathon Edition #2 is to build a highly scalable sentiment analysis model to classify customers emotions\/sentiments from the raw unstructured data.","categories":["Deep Tech"],"tags":["Hackathon","hackathon for data scientists","Hackathons","Hackathons India","Machinehack","Machinehack Hackathon","machinehack platform","New Hackathon For Data Scientists","new hackathon for ML practioners","The last hacker standing","Weekend Hackathon","Weekend hackathon for data scientists"],"author_name":"AIM Media House","publish_date":"2021-07-30T18:00:00","publication_year":"2021","word_count":506,"keywords":["scikit-learn","Weekend Hackathon","sentiment analysis","Scala","new hackathon for ML practioners","deep learning","New Hackathon For Data Scientists","hackathon for data scientists","R","RAG","programming_languages:Scala","machinehack platform","machine learning","AI","Weekend hackathon for data scientists","Hackathon","Machinehack Hackathon","Hackathons India","The last hacker standing","Machinehack","programming_languages:R","Hackathons"],"extracted_tech_keywords":["AI","machine learning","deep learning","scikit-learn","RAG","sentiment analysis","R","Scala","programming_languages:R","programming_languages:Scala"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/machinehack-is-back-with-weekend-hackathon-edition-2-the-last-hacker-standing\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10019905,"title":"Hands-on Python Guide to Style-based Age Manipulation (SAM) Technique","content":"Table of contentsIntroductionWhat is age transformation?Overview of SAMWhy the name ‘style-based Age Manipulation’?Practical implementation of SAMPrerequisitesPre-trained modelDemo codeOutputGoogle colab notebooks References Introduction Style-based Age Manipulation (SAM) is a method used to perform fine-grained age transformation in digital image processing and computer vision tasks using a single facial image as an input. It was introduced by Yuval Alalu, Or Patashnik and Daneil Cohen-Or of Tel-Aviv University in February 2021 (research paper). This article gives an overview of SAM along with its demonstration using Python code. Before going into the details of SAM, let us first understand what is meant by the ‘age transformation’ task. What is age transformation? Age transformation is a process of representing the change in a person’s appearance across different ages while preserving his identity. In order to model such a process over a single input facial image, the change in head shape and texture must be captured while the identity and other key facial attributes of the input face must be preserved. The complexity of the task increases when modelling lifelong ageing where significant age modification is desired (e.g. from ages 10 to 90 years). To avoid explicit modelling of age transformation, Generative Adversarial Networks (GANs) are largely used for generating images in a data-driven manner, especially on facial images. Overview of SAM SAM is a method for learning a conditional image generation function which can capture the desired change in age but preserve the facial identity. It is an image-to-image translation method i.e. translates a given image of a source domain to a corresponding image of a target domain. It couples the expressiveness of a pre-trained, fixed StyleGAN generator with an encoder architecture. The encoder directly encodes an input facial image into a series of style vectors subject to the desired age shift. These style vectors are then fed into unconditional StyleGAN. The output of StyleGAN represents the desired age transformation. The use of StyleGAN enables leveraging its ability to achieve excellent image quality. A pre-trained, the fixed age regression model is used for generating the latent codes corresponding. The continuous ageing process is formulated as a regression task between the input age and desired target age, providing fine-grained control over image generated by GAN. Why the name ‘style-based Age Manipulation’? Since age transformation in the SAM method is controlled through the intermediate style representations learned by the StyleGAN, it has been named as “Style-based” Age Manipulation. In other words, it does age transformation based on the style of an input facial image. Practical implementation of SAM Prerequisites Python 3Linux or macOSNVIDIA GPU + CUDA CuDNN Pre-trained model The SAM model pre-trained on the FFHQ dataset can be downloaded from here. If you wish to train a SAM model from scratch, the following auxiliary models can be used: pSp Encoder : taken from pixel2style2pixel (an image-to-image translation framework) trained on the FFHQ dataset for StyleGAN inversionFFHQ StyleGAN : StyleGAN model pre-trained on FFHQ taken from rosinality’s repository with 1024×1024 output resolutionIR-SE50 Model : Pretrained IR-SE50 model taken from TreB1eN (used for identity loss)VGG Age Classifier : VGG age classifier from DEX and fine-tuned on the FFHQ-Aging dataset (used for aging loss) Demo code Import the os module to interact with the underlying Operating System import os Define the code directory name CODE_DIR = ‘SAM’ Clone the GitHub repository !git clone https:\/\/github.com\/yuval-alaluf\/SAM.git $CODE_DIR Download Ninja (a small build system focussed on speed) !wget https:\/\/github.com\/ninja-build\/ninja\/releases\/download\/v1.8.2\/ ninja-linux.zip Unzip ninja-linux.zip file !sudo unzip ninja-linux.zip -d \/usr\/local\/bin\/ Change from default to alternative Python version !sudo update-alternatives --install \/usr\/bin\/ninja ninja \/usr\/local\/bin\/ninja 1 --force Change the current working directory to ‘SAM’ os.chdir(f'.\/{CODE_DIR}') Import the required standard libraries from argparse import Namespace import sys import pprint import numpy as np from PIL import Image import torch import torchvision.transforms as transforms Modify sys.path sys.path.append(\".\") sys.path.append(\"..\") Import the AgeTransformer class from datasets.augmentations import AgeTransformer Import tensor2im method for tensor-to-image conversion from utils.common import tensor2im Import the pSp class from models.psp import pSp Define experiment type EXPERIMENT_TYPE = 'ffhq_aging' Get wget download command for downloading the desired model and save to directory ..\/pretrained_models def get_download_model_command(file_id, file_name): current_directory = os.getcwd() save_path = os.path.join(os.path.dirname(current_directory), \"pretrained_models\") if not os.path.exists(save_path): os.makedirs(save_path) url = r\"\"\"wget --load-cookies \/tmp\/cookies.txt \"https:\/\/docs.google.com\/uc?export=download&confirm=$(wget --quiet --save-cookies \/tmp\/cookies.txt --keep-session-cookies --no-check-certificate 'https:\/\/docs.google.com\/uc?export= download&id={FILE_ID}' -O- | sed -rn 's\/.*confirm= ([0-9A-Za-z_]+).*\/\\1\\n\/p')&id={FILE_ID}\" -O {SAVE_PATH}\/{FILE_NAME}&&rm -rf \/tmp\/cookies.txt\"\"\".format(FILE_ID=file_id, FILE_NAME=file_name, SAVE_PATH=save_path) return url Define model path MODEL_PATHS = { \"ffhq_aging\": {\"id\": \"1XyumF6_fdAxFmxpFcmPf-q84LU_22EMC\", \"name\": \"sam_ffhq_aging.pt\"} } Initialize model path and download command path = MODEL_PATHS[EXPERIMENT_TYPE] download_command = get_download_model_command(file_id=path[\"id\"], file_name=path[\"name\"]) !wget {download_command} Define experiment arguments EXPERIMENT_DATA_ARGS = { \"ffhq_aging\": { \"model_path\": \"..\/pretrained_models\/sam_ffhq_aging.pt\", \"image_path\": \"notebooks\/images\/1287.jpg\", \"transform\": transforms.Compose([ transforms.Resize((256, 256)), transforms.ToTensor(), transforms.Normalize([0.5, 0.5, 0.5], [0.5, 0.5, 0.5])]) } } EXPERIMENT_ARGS = EXPERIMENT_DATA_ARGS[EXPERIMENT_TYPE] Initialize the model path model_path = EXPERIMENT_ARGS['model_path'] Load the PyTorch model using torch.load() ckpt = torch.load(model_path, map_location='cpu') Have a look at the model training options opts = ckpt['opts'] pprint.pprint(opts) Update the training options opts['checkpoint_path'] = model_path Load SAM model.Use CUDA tensor types which implement functions like CPU tensors but using GPUs. opts = Namespace(**opts) net = pSp(opts) net.eval() net.cuda() \/\/torch.cuda print('Model successfully loaded') Open the input image and resize it for display image_path = EXPERIMENT_DATA_ARGS[EXPERIMENT_TYPE][\"image_path\"] original_image = Image.open(image_path).convert(\"RGB\") \/\/Image.open() On executing the above line of code, you will see the input facial image which is as follows: Download Dlib model shape_predictor_68_face_landmarks.dat.bz2 which has been trained on the ibug 300-W dataset. !wget http:\/\/dlib.net\/files\/shape_predictor_68_face_landmarks.dat.bz2 Use bzip2 command for file compression and decompression !bzip2 -dk shape_predictor_68_face_landmarks.dat.bz2 Define function for face alignment def run_alignment(image_path): import dlib from scripts.align_all_parallel import align_face predictor = dlib.shape_predictor(\"shape_predictor _68_face_landmarks.dat\") aligned_image = align_face(filepath=image_path, predictor=predictor) print(\"Aligned image has shape: {}\".format(aligned_image.size)) return aligned_image Align the input face aligned_image = run_alignment(image_path) Resize the aligned face aligned_image.resize((256, 256)) Initialize variables for image transformation img_transforms = EXPERIMENT_ARGS['transform'] input_image = img_transforms(aligned_image) Run the image on multiple target ages target_ages = [0, 10, 20, 30, 40, 50, 60, 70, 80, 90, 100] \/\/0-100 years’ age shift age_transformers = [AgeTransformer(target_age=age) for age in target_ages] Define function to create tensors def run_on_batch(inputs, net): result_batch = net(inputs.to(\"cuda\").float(), randomize_noise=False, resize=False) return result_batch For each age transformed age, concatenate the results to display them side-by-side results = np.array(aligned_image.resize((1024, 1024))) for age_transformer in age_transformers: print(f\"Running on target age: {age_transformer.target_age}\") with torch.no_grad(): input_image_age = [age_transformer(input_image.cpu()).to('cuda')] input_image_age = torch.stack(input_image_age) \/\/torch.stack result_tensor = run_on_batch(input_image_age, net)[0] result_image = tensor2im(result_tensor) results = np.concatenate([results, result_image], axis=1) Construct image memory from numerical array representation of the output using Image.fromarray() results = Image.fromarray(results) Display the output results Output Source: https:\/\/github.com\/yuval-alaluf\/SAM\/blob\/master\/notebooks\/inference_playground.ipynb The above code can also be experimented for different images by changing the image path appropriately. We have tried the code on two more images, the results of which are as follows: Input image 2: Output: Input image 3: Output: Google colab notebooks Demo1Demo2Demo3 References Do you want to have a deep-rooted understanding of the SAM technique? Refer to the following sources: SAM research paperGitHub repositoryStyleGAN research paper","excerpt":"Introduction Style-based Age Manipulation (SAM) is a method used to perform fine-grained age transformation in digital image processing and computer vision tasks using a single facial image as an input. It was introduced by Yuval Alalu, Or Patashnik and Daneil Cohen-Or of Tel-Aviv University in February 2021 (research paper). This article gives an overview of […]","categories":["Deep Tech"],"tags":["face recognition"],"author_name":"Nikita Shiledarbaxi","publish_date":"2021-02-10T12:00:00","publication_year":"2021","word_count":1129,"keywords":["CUDA","NumPy","TPU","PyTorch","AI","Transformers","computer vision","RAG","Colab","Ray","face recognition"],"extracted_tech_keywords":["AI","computer vision","Ray","PyTorch","Transformers","Colab","NumPy","RAG","TPU","CUDA"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/hands-on-python-guide-to-style-based-age-manipulation-sam-technique\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":50595,"title":"How DeepMind Powers Google Play Store Apps","content":"DeepMind has been powering data centres, revolutionising medical research and even saving battery life on phones. And the one thing that binds all these applications is the use of machine learning to optimise outcomes. Ever since Google acquired DeepMind, it has been outsourcing its innovations to power its own products. Today, apps on the Play Store are also being customised using DeepMind’s research. PlayStore Recommendations Unlike Any Other via Google play store Netflix will suggest movies based on the user’s watch history. If a user watches more action movies, the algorithm will be more inclined to suggest action movies. Youtube will recommend music based on the kind of genre on listens to the most. Achieving real-time personalised content is the primary goal of all the platforms that use recommender systems. However, similar strategies might not work for Playstore, which also happens to be one of the largest deployers of recommender systems. For instance, when a user installs, say, a travel booking app, the above strategies would recommend more travel apps. However, a better recommendation would be an app that would translate foreign languages since one is travelling. To bring this level of intelligence on to the play store platform, DeepMind worked with the developers from Google and have developed solutions that would best suit the need of users. DeepMind team considered the following three solutions: Use LSTMs: a recurrent neural network that performs well in real-world scenarios, owing to a powerful update equation and backpropagation dynamics. Replace LSTM with Transformer model: To address the computational and serving delays of LSTMs, transformer model was used. Use additive attention model: Though transformer improved model performance, its high training cost led the researchers towards a more efficient solution. They call this the candidate generator model. Importance Weighting And Refined Re-ranking via DeepMind blog Today, Google Play’s recommendation system contains three main models: a candidate generator, a re-ranker, and a model to optimise for multiple objectives. The candidate generator is a deep retrieval model that can analyse more than a million apps and retrieve the most suitable ones. To address the bias associated with the candidate generator model, DeepMind introduced ‘importance weighting’ in their model. Importance weighting considers the following two metrics: Impression-to-install rate of each individual app Median impression-to-install rate across the Play store. An app is compared using these two metrics and the app with a below-median install rate will have an important weight less than one. In this way, an app is either wp-weighted or down-weighted and reduces the chances of bias in the recommendation. Since the problem of bias is handled, the ranking has to be refined. This ranking decides what appears on the top of the screen without the user having to scroll. In conventional recommendations systems, ranking is treated as a binary classification problem. This pointwise ranking ignores the association between apps. Like the one discussed above— travel booking app followed by a translator app. To take into account the relation between apps and refine the ranking, DeepMind has built a re-ranker model. Unlike the conventional pointwise model, a pairwise model is deployed. And, they take care of all these online in real-time. The Hand That Keeps Giving Apart from their success with AlphaGo game and research with protein folds, their research has been directly been put into use by Google. Here is a list of their top contributions to Google: reduced the electricity needed for cooling Google’s data centres by up to 30%.boosted the value of Google’s wind energy by roughly 20%created on-device learning systems to optimise Android battery performance.WaveNet is used for Google Assistant and Google Cloud Platform.Helped Waymo by improving the efficiency of training its neural networks. “Looking at the pace of progress, I think we will have AI in a form in which it benefits a lot of users in the coming years, but I still think it’s early days, and there’s a long-term investment for us,” said Sundar Pichai when he was asked about DeepMind back in 2016. Despite the news of DeepMind’s losses doing rounds, the team has proven time and again that it has been a great addition to Google.","excerpt":"DeepMind has been powering data centres, revolutionising medical research and even saving battery life on phones. And the one thing that binds all these applications is the use of machine learning to optimise outcomes. Ever since Google acquired DeepMind, it has been outsourcing its innovations to power its own products. Today, apps on the Play […]","categories":["Global Tech"],"tags":["Google Translate","Sundar Pichai"],"author_name":"Ram Sagar","publish_date":"2019-11-25T14:59:22","publication_year":"2019","word_count":689,"keywords":["Go","machine learning","programming_languages:R","AI","neural network","Google Translate","innovation","programming_languages:Go","cloud_platforms:Google Cloud","LSTM","R","Sundar Pichai"],"extracted_tech_keywords":["AI","machine learning","neural network","R","Go","LSTM","innovation","cloud_platforms:Google Cloud","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/deepmind-google-play-store-machine-learning\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":58434,"title":"IBM Project Debater: AI System That Can Debate People On Complex Topics","content":"IBM is announcing several new IBM Watson technologies designed to help organizations begin identifying, understanding and analyzing some of the most challenging aspects of the English language with greater clarity, for greater insights. The new technologies represent the first commercialization of key Natural Language Processing (NLP) capabilities to come from IBM Research’s Project Debater, the only AI system capable of debating humans on complex topics. For example, a new advanced sentiment analysis feature is defined to identify and analyze idioms and colloquialisms for the first time. Phrases, like ‘hardly helpful,’ or ‘hot under the collar,’ have been challenging for AI systems because they are difficult for algorithms to spot. With advanced sentiment analysis, businesses can begin analyzing such language data with Watson APIs for a more holistic understanding of their operations. Further, IBM is bringing technology from IBM Research for understanding business documents, such as PDF’s and contracts, to also add to their AI models. “Language is a tool for expressing thought and opinion, as much as it is a tool for information,” said Rob Thomas, General Manager, IBM Data and AI. “This is why we’re harvesting technology from Project Debater and integrating it into Watson – to enable businesses to capture, analyze, and understand more from human language and start to transform how they utilize intellectual capital that’s codified in data.” Today IBM is announcing that it plans to integrate Project Debater technologies into Watson throughout the year, with a focus on advancing clients’ ability to exploit natural language: A. Analysis – Advanced Sentiment Analysis. IBM has enhanced sentiment analysis to be able to better identify and understand complicated word schemes like idioms (phrases and expressions) and so-called, sentiment shifters, which are combinations of words that, together, take on new meaning, such as, “hardly helpful.” This technology will be integrated into Watson Natural Language Understanding this month. In addition, we are announcing a new classification technology that will enable clients to create AI models that can more easily classify clauses that occur in business documents, like procurement contracts. Based on Project Debater’s deep learning-based classification technology, the new capability can learn from as few as several hundred samples to do new classifications quickly and easily. It is planned to be added to Watson Discovery later this year. B. Briefs – Summarization. This technology pulls textual data from a variety of sources to provide users with a summary of what is being said and written about a particular topic. An early version of Summarization was leveraged at The GRAMMYS this year to analyze over 18 million articles, blogs and bios to produce bite-sized insights on hundreds of GRAMMY artists and celebrities. The data was then infused into the red carpet live stream, on-demand videos and photos across www.grammy.com to give fans deeper context about the leading topics of the night. It is planned to be added to IBM Watson Natural Language Understanding later in the year. C. Clustering – Advanced Topic Clustering. Building on insights gained from Project Debater, new topic clustering techniques will enable users to “cluster” incoming data to create meaningful “topics” of related information, which can then be analyzed. The technique, which is planned to be integrated into Watson Discovery later this year, will also allow subject matter experts to customize and fine-tune the topics to reflect the language of specific businesses or industries, like insurance, healthcare and manufacturing. IBM, has long been a leader in NLP, developing technologies that enable computer systems to learn, analyze and understand human language – including sentiment, dialects, intonations, and more – with increasing accuracy and speed. IBM has brought its NLP technology, much of which was born in IBM Research, to market via Watson. Products such as Watson Discovery for document understanding, IBM Watson Assistant for virtual agents, and Watson Natural Language Understanding for advanced sentiment analysis, are all infused with NLP. ESPN Fantasy Football uses Watson Discovery and Watson Knowledge Studio to analyze millions of football data sources each day during the season to offer millions of fantasy football players real-time insights. By processing natural language, Watson identifies the tone and sentiment of news articles, blogs, forums, rankings, projections, podcasts and tweets that cover everything from locker room insights to injury analysis. ESPN Fantasy Football surfaces these insights in player cards that snapshot the “boom” and “bust” potential of each player, as well as a “Player Buzz” section that summarizes the positive or negative commentary about a player. KPMG, a multinational professional services network, and one of the Big Four accounting organisations, worked with IBM to create an AI solution based on a variety of Watson services, including Watson Natural Language Understanding. This technology makes it more effective for companies to identify, claim and retain potential R&D income tax credits. Developed by KPMG, the solution can help clients increase the amount of R&D income tax credits they capture because the Watson technology is able to review more documentation quickly while minimizing disruption to the client’s business. In the past year, KPMG clients have seen more potential for R&D tax credits, with some projects even seeing more than a 1000% increase in the number of documents reviewed. The solution helps clients uncover more potential activities that qualify for additional income tax credits, while reducing business disruption. As a result, engineers and scientists can stay focused on innovative R&D work by spending less time on income tax compliance activities.","excerpt":"IBM is announcing several new IBM Watson technologies designed to help organizations begin identifying, understanding and analyzing some of the most challenging aspects of the English language with greater clarity, for greater insights. The new technologies represent the first commercialization of key Natural Language Processing (NLP) capabilities to come from IBM Research’s Project Debater, the […]","categories":["AI News"],"tags":["Deep Learning Techniques","IBM"],"author_name":"Vishal Chawla","publish_date":"2020-03-11T17:21:57","publication_year":"2020","word_count":898,"keywords":["Go","API","AI","sentiment analysis","RAG","NLP","Aim","deep learning","IBM","Deep Learning Techniques","GAN","R"],"extracted_tech_keywords":["AI","deep learning","NLP","Aim","RAG","sentiment analysis","R","Go","API","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/ibm-project-debater-ai-system-that-can-debate-people-on-complex-topics\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10076423,"title":"What is the Zero-Moment Point in Optimus Robot?","content":"On Tesla AI Day 2022, Elon Musk revealed a working prototype of the Optimus Robot. Still in early stages of development, Optimus works on the mechanism of Tesla’s self-driving cars and also showcases capabilities of practical tasks. Though Optimus is trained with techniques like locomotion planning and using sensors to predict the walking surface, Christian Hubicki, robotics professor and Director of Optimal Robotics Lab, pointed out that Optimus seems to be working on the method called Zero-Moment Point (ZMP) to balance its weight during each next step by bending knees and tilting backwards or forward. Let's talk walking.I told my lab today that I expected it to walk on stage, and it did, but I was a bit surprised how they did it.It *seems* to use a method called Zero-Moment Point to maintain balance. It's been used in various forms since the 90's.https:\/\/t.co\/VtmU4gN1MA— Christian Hubicki (@chubicki) October 1, 2022 ZMP is the dynamic model for the control of legged locomotion (biped locomotion) in robots and humanoids. Essentially, the algorithm or technique specifies the point where the sum of gravity forces and horizontal inertia is zero during the contact of each foot to the ground. This ensures that there is no moment of inertia in the horizontal direction while walking—meaning, the robot won’t tip over while taking each step due to uneven weight distribution by ensuring all the weight is on the centre of gravity of the robot. Optimus uses electrical actuators for locomotion while the much-compared Boston Dynamics’ ATLAS uses a high powered hydraulic actuator which heavily drains the battery. ‘Cassie’, the prototype bipedal robot developed by Agility Robotics, balances itself dynamically at every step based on the gait library method and is a closer comparison to the locomotion technique of Tesla’s robot’s Zero-Moment Point (ZMP). Biped locomotion has been researched for decades now. Miomir Vukobratović introduced the concept of ZMP in the legged locomotion community in 1968 at the Third All-Union Congress of Theoretical and Applied Mechanics in Moscow. Tesla Bot walking progress – modeled inside Autopilot sim pic.twitter.com\/WCrXDFel3t— Tesla (@Tesla) October 1, 2022 Biped Locomotion Mechanism Biped robots are designed to mimic human-like locomotion and have an advantage over other multi-legged robots because they are highly adaptable. Irrespective of the structure and design of a robot, the basic characteristics of a biped locomotion robot are: Passive Degrees of Freedom (DOF): DOF can be simply referred to as the number of movable joints of a robot. A passive DOF is the possibility of occurrence of a new “joint” because of disturbances from outside during movement of one foot, causing changes in the overall system.Gait repeatability or gait symmetry.Regular interchangeability between double support phase (DSP) and single support phase (SSP). The development and improvement of biped robots and establishing these characteristics for stability has been challenged by few problems: Passive joints that are not in contact with the ground make the structures unstable.The gait cycle has varying configurations during locomotion, changing with every step—DSP is when both feet are on the ground and SSP is when one foot is on the ground and the other is in the process of being transferred from back to front, also known as the ‘unstable’ stage.Robots interact with different surfaces and environments which affects the algorithm. ZMP in Optimus In DSP, it is easy to determine the zero-moment point by calculating the position of the centre of gravity (COG) or the hip trajectory. To maintain the stability of robots during SSP, which comprises 80% of the gait cycle of a biped robot, DSP is incorporated and all the joints are actuated and rigidly controlled to track the hip trajectory, which simplifies the control task. Both of these processes depend on linear pendulum mode (LPM) for prediction of the centre of gravity. Since, the point of contact between the foot and the ground cannot be controlled directly like the other mechanism joints, it is indirectly controlled by enabling and ensuring appropriate dynamics of mechanism present above the foot, like the COG of the hip trajectory. Therefore, ZMP is the point where all the forces acting on the mechanism can be replaced by a single force and should be inside the support polygon. The support polygon is the area on a horizontal plane where the applied and reactionary force coincide, or simply, the sole of the foot. In a balanced gait, ZMP and COG coincide, therefore it is easy to find out and maintain equilibrium in the mechanism. In the above figure, ‘P’ is the zero-moment point. In the case of a disturbance that can cause rotation of the foot, the ZMP might shift to the edge of the foot and result in a disbalance of the mechanism. Though it still remains in the support polygon, the area on the edge reduces, and thus causes imbalance. In this case, by tracking the hip trajectory, the robot can balance itself by shifting its weight across the system. Therefore, assumably, by training, locomotion planning, and scanning the walking path and surface in front of it, the steps that Optimus takes by calculating the hip trajectory and bending its knees to ensure all the force is on its centre of gravity, seems to be working on a method based on Zero-Moment Point calculation.","excerpt":"Optimus seems to be working on the method called Zero-Moment Point to balance its weight during each next step by bending knees and tilting back and forth.","categories":["AI Trends"],"tags":["ATLAS","Boston Dynamics","Elon Musk","humanoid","Robotics"],"author_name":"Mohit Pandey","publish_date":"2022-10-05T13:00:00","publication_year":"2022","word_count":874,"keywords":["Go","Boston Dynamics","ATLAS","programming_languages:R","AI","programming_languages:Go","Robotics","Elon Musk","ViT","ai_applications:robotics","R","humanoid"],"extracted_tech_keywords":["AI","R","Go","ViT","programming_languages:R","programming_languages:Go","ai_applications:robotics"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/what-is-the-zero-moment-point-in-optimus-robot\/","complexity_score":3,"technical_depth":7,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10071603,"title":"How to automate finding the optimal learning rate?","content":"Finding the best settings for deep learning model hyperparameters has long been regarded as more an art than a science and has depended primarily on trial and error. Learning rate (LR) is possibly the most significant hyperparameter in deep learning since it determines how much gradient is backpropagated. This, in turn, determines how far we progress towards minima. A slow learning rate causes the model to converge slowly, whereas a fast learning rate causes the model to diverge. As a result, the learning rate must be precisely right. This article will help in learning about automatic learning rate finder and its implementation. Following are the topics to be considered. Table of contents Dilemma behind the optimal Learning RateCommon architecture of Automatic LR finderFinding optimal learning rate with PyTorch The learning can’t be fixed for every network. Let’s see behind the scenes of Learning rate testing. Dilemma behind the optimal Learning Rate The learning rate is a hyperparameter that governs the alteration of the network’s weights in relation to the loss gradient. It quantifies the amount of information to be learned from every new mini-batch of training data. Mathematically it is a penalty on the amount of information consumed by the weights of the model. The larger the steps taken along the trajectory to the lowest of the loss function, where the optimal model parameters are, the faster it learns. Analytics India Magazine The testing range of learning rates is limited to one epoch of training repetitions, with the learning rate increasing with each mini-batch of data. The learning rate increases from a very small to a very big number during the procedure, causing the training loss to begin with a plateau, then decrease to some minimal value, and then burst. This usual behaviour may be plotted (as shown in the graphic below) and utilised to choose a suitable range for the learning rate, especially in the region where the loss is decreasing. Analytics India Magazine The suggested lowest learning rate is the value at which the loss reduces the fastest (minimal negative gradient), whereas the recommended maximum learning rate is much less than the learning rate at which the loss is the smallest. It is much less, say ten times, because for plotting a smoothed version of the loss, picking the learning rate that corresponds to the smallest loss is likely to be too great, causing the loss to diverge over training. Are you looking for a complete repository of Python libraries used in data science, check out here. Common architecture of Automatic LR finder A typical Automatic Learning rate finder uses a Cyclic Learning rate method. The algorithm’s goal is to provide a method for determining global learning rates for training neural networks that eliminate the requirement for hundreds of tests to determine the best values with no additional processing. By introducing the notion of the LR Range Test, CLR delivers an excellent learning rate range (LR range) for an experiment. A good learning rate is one that results in a significant decrease in network loss. Here comes CLR’s sorcery. The original CLR paper describes an experiment in which you may monitor the behaviour of the learning rate in relation to the loss. The experiment is simple to understand: after each mini-batch, progressively raise the learning rate while noting the loss at each step. This slow rise might be linear or exponential. And, sure, this is similar to the LR Range Test. Analytics India Magazine After carrying out the experiment, Leslie demonstrated that at excessively low learning rates, the loss may diminish, but only at a very slow rate. When you enter the ideal learning rate zone, you will see a sharp decline in the loss function. If the learning rate is increased further, it may produce parameter loss in the network, which may result in an increase in losses. So, based on this experiment, it is evident that you are looking for a sharp drop in the loss function, and you may do so by analysing the gradients of the loss function at various stages of training. Finding optimal learning rate with PyTorch This article for finding the optimal learning rate for the neural network uses the PyTorch lighting package. The model used for this article is a LeNet classifier, a typical beginner convolutional neural network. This model is used as an image classifier in this article, the dataset used is the famous MNIST dataset. Let’s start with installing and importing the dependencies and the prerequisites. !pip install PyTorch-lightning !pip install torchmetrics Need to install the torchmeterics with the PyTorch lighting because the metrics module is shifted from the PyTorch lighting package. The torch metrics have both predefined and also could build a custom evaluation method. import torch.nn as nn import torch.nn.functional as F import torchvision.transforms as transforms import PyTorch_lightning as pl from PyTorch_lightning.loggers import TensorBoardLogger from torchmetrics import functional as FM One can easily download the MNIST dataset or use the code from the notebook attached in the references section for downloading and splitting the data. The model is a LeNet classifier built using the “LightningModule” of the PyTorch lighting. The model has 2 convolution layers and 3 linear layers with 120, 84, and 10 fully connected neurons respectively. def __init__(self, num_classes=10): super().__init__() self.lr = 2e-3 self.conv1 = nn.Conv2d(1, 6, 5, padding=2) self.conv2 = nn.Conv2d(6, 16, 5) self.fc1 = nn.Linear(16*5*5, 120) self.fc2 = nn.Linear(120, 84) self.fc3 = nn.Linear(84, 10) def training_step(self, batch, batch_idx): x, y = batch y_hat = self(x) acc = FM.accuracy(y_hat, y) loss = F.cross_entropy(y_hat, y) self.log_dict({'train_acc': acc, 'train_loss': loss}, on_step=True, on_epoch=False) return {\"training accuracy\": acc,\"loss\":loss} def validation_step(self, batch, batch_idx): x, y = batch y_hat = self(x) val_acc = FM.accuracy(y_hat, y) loss = F.cross_entropy(y_hat, y) self.log_dict({'val_acc': val_acc, 'val_loss': loss}, reduce_fx=torch.mean) return {\"validation accuracy\":val_acc, \"validation losss\":loss} Here is a glimpse of the classifier model. For the details, refer to the Colab notebook attached in the reference section. Now it’s time for training the model. There will be two versions of this model one without the auto-learning rate finder and the other with the auto learning rate. The losses and the accuracy metrics would be logged using the tensorboard for better visualization. data_directory= \"\/content\/\" batch_size=60 logger_directory= 'logs\/without_auto_lr' name_of_log= 'LeNet classifier' version_of_log = 1.0 Default_lr= 1e-3 max_epochs=8 model = LeNet_classifier_model() dataset = MNISTData(data_directory, batch_size) logger = TensorBoardLogger(save_dir=logger_directory,version=version_of_log,name=name_of_log) trainer = pl.Trainer(gpus=1, max_epochs=max_epochs, logger=logger, auto_lr_find=False, val_check_interval=0.5) model.lr = Default_lr print(f'Default model LR: {model.lr}') trainer.fit(model, dataset) Analytics India Magazine Analytics India Magazine Analytics India Magazine So with a learning rate of 0.001 and a total of 8 epochs, the minimum loss is achieved at 5000 steps for the training data and for validation, it’s 6500 steps which seemed to get lower as the epochs increased. Let’s find the optimum learning rate with lesser steps required and lower loss and high accuracy score. For using the automatic LR finder, each would be the same as earlier. Just need to add these lines to the code which will find the optimal learning rate and plot the loss vs learning rate curve for better visualization. lr_finder = trainer_2.tuner.lr_find(model, dataset) model.hparams.lr = lr_finder.suggestion() print(f'Auto-find model LR: {model.hparams.lr}') fig = lr_finder.plot(suggest=True) Analytics India Magazine Analytics India Magazine So the optimal learning rate for the model is 0.025 which is greater than the default learning rate. Therefore the computational time would be less compared and it would be less cost-effective. Furthermore, you can train the model on this learning rate, leaving that to you. Conclusions The LR finder is a great tool for determining the best learning rate for a given situation, but it should be used with caution. It is critical to use the same initial weights in both the Learning rate range test and subsequent model training. Never assume that the discovered learning rates are optimal for any model initialization. With this article, we have understood the concept of automatic learning rate finder with implementation. References Link to the above codePyTorch Lighting LR finder documentation","excerpt":"The Cyclic Learning rate method finds the rate automatically.","categories":["AI Trends"],"tags":["gradient descent","hyperparameter optimisation","hyperparameter tuning","learning rate"],"author_name":"Sourabh Mehta","publish_date":"2022-07-27T10:00:00","publication_year":"2022","word_count":1320,"keywords":["data science","hyperparameter optimisation","Go","learning rate","AI","neural network","PyTorch","hyperparameter tuning","Colab","Python","deep learning","analytics","R","gradient descent"],"extracted_tech_keywords":["AI","deep learning","neural network","data science","analytics","PyTorch","Colab","Python","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/how-to-automate-finding-the-optimal-learning-rate\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":11447,"title":"5 Questions You Need to Ask Before You Outsource Market Research","content":"Market research is one of the most important elements of every successful business and every business owner should make it one of the top priorities for improving and growing their business. Conducting market research can provide you with a number of benefits; because it can help you determine the needs of your target customers, which undeniably represent a very important piece of information. Knowing and understanding what your target customers’ needs are is crucial, since you can implement various strategies accordingly and meet those needs and desires. Market research helps you accomplish that by providing you with valuable information on your customers’ behavior, that is, it helps you know why they are purchasing your products or using your services, or why they are perhaps doing the opposite. Due to the importance of market research, many business owners decide to hire someone else to do it for them, because not only can outsourcing save them a lot of money, but it can also provide them with efficient and, more importantly, effective results. If you are reading this, it is safe to assume you are considering outsourcing market research as well, but how can you be sure that you will make the right choice? In order to be absolutely certain, you need to ask yourself some crucial questions before you go and hire a market research company, so take a look at what those are. Would You Save Money by Conducting Market Research In-House? This is the first and most obvious question every business owner asks before turning to outsourcing. Would you save money by conducting market research internally? The answer is definitely yes. The money you would spend to hire an outsourcing company would stay in your pocket by doing your project in-house, but at what expense? If you decide to do your project in-house, you would need to invest a lot of time and money in order to do it properly and, thus, reap the actual benefits from it. Your employees would need to focus a lot of their time on research, which would cost them their productivity. As a result, your company would suffer the consequences, since all of your day-to-day operations would be slowed down. More importantly, you might even spend a lot more money than you should, simply because you want everything to go down perfectly, but you may lack the expertise in the field. Again, your company would be negatively affected. However, it can happen that conducting market research internally brings positive results, but those instances are extremely rare and, again, they are definitely not cost-efficient. Do You Have the Necessary Expertise for Market Research? Market research is definitely not something your every employee can do. As a matter of fact, perhaps none of your employees can do it properly, simply because they don’t do it on a daily basis. Are your employees actually aware of the importance of market research and do they even have experience in the field? Do they know anything about survey design and data analysis? Do they know what focus group moderation and group facilitation are? More importantly, do they know what to do with the findings of market research? There are various survey applications that can help you conduct market research, but they are far from having all the necessary tools for helping you do it properly and gaining some useful information. You would still need to compose specific survey questions, so that you can acquire crucial information about your target market. Furthermore, numerous programs you can use for research cannot possibly provide you with important in-depth information needed for your company’s actual growth, such as informed insights on the findings of your research. All you would end up having is incomplete data and the need for proper market research would be greater. Therefore, you need someone who is an expert in market research, because only someone with actual knowledge and experience can provide you with positive results that can benefit your company. Hence, outsourcing is a much better option you should go for. Can You Be Objective When Conducting Market Research Internally? The ability to be objective when running a business is something not every business owner has, but it is one of the most important factors that can influence one’s overall success. Why is objectivity so important? Because when you are able to take on a different perspective on things, you can shed some more light on your path and know exactly which direction to take in order to get to your end goal. You may discover critical information that you may otherwise overlook. If you can be objective enough to conduct market research in-house, then you can certainly gain relevant information and use your findings properly. However, it is a fact that an observer outside your company can be much more objective, so you should really think about hiring a market research company, because they can provide you with a fresh perspective and credible findings. What If You End Up with Faulty Research? This is perhaps the most important and unquestionably the scariest question you need to ask before diving into the outsourcing world. What if the market research team you hired turns out to be completely faulty? What if the findings don’t provide you with any actionable results? This is something that makes many business owners scratch the option of outsourcing market research. It definitely is a scary option but, in order to avoid the common pitfalls, you need to hire someone to help you out. It can be tricky if you have no idea what to look for when searching for an outsourcing agency, but it is even scarier to try and do it yourself, because the chances of ending up with incomplete research are even higher. Therefore, don’t try and save every penny when it comes to market research if you don’t want to end up with faulty results. This means that you shouldn’t do it yourself, but hire someone who can help you, since it will definitely be worth every dime. You need to spend a bit more money if you want to do the research right and doing so will provide you with complete information, which is precisely the point of conducting market research in the first place, right? How to Actually Choose a Market Research Agency? If you manage to find a reputable and reliable market research agency that can provide you with exceptional services from expert professionals, you can succeed in achieving your business goals in no time. However, how can you know what agency is the best one for you and your needs? What’s important is to make sure the outsourcing agency you choose understands your business goals and knows how to conduct market research so that it aligns with your objectives. You need to make sure they will use methods and strategies that will result in relevant information that you can actually use. You need to know how they work with their clients and whether or not they will provide you with status updates and essential reports. You need to gain an insight into their approach in order to see if it aligns with your approach, because being on the same page will certainly benefit the entire project. You should learn everything there is to know about a particular agency before choosing them for your needs, so make sure you check online reviews and references, as well as various other publications. There are many benefits of conducting market research, so make sure you do everything you can to gain accurate results that will help you stay ahead of the curve and grow your business. Outsourcing will definitely help you make the right choice, so seriously consider taking up that path.","excerpt":"Market research is one of the most important elements of every successful business and every business owner should make it one of the top priorities for improving and growing their business. Conducting market research can provide you with a number of benefits; because it can help you determine the needs of your target customers, which […]","categories":["AI Trends"],"tags":[],"author_name":"AIM Media House","publish_date":"2016-12-01T03:48:22","publication_year":"2016","word_count":1292,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","ViT","R"],"extracted_tech_keywords":["AI","R","Go","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/5-questions-need-ask-outsource-market-research\/","complexity_score":2,"technical_depth":6,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10008636,"title":"NTT Expands Its Data Center Capacity In India With New Mumbai Facility","content":"NTT’s Global Data Centers division today announced the expansion of its line of data center facilities with the launch of a new high-density and hyperscale data center in Mumbai, the financial capital of India. The Mumbai 7 Data Center operates with 375,000 sq.ft of colocation space, and will offer the capacity to host 5000 racks and over 30 MW of load. The new Mumbai 7 Data Center facility will expand NTT data center capacity (server room) in India by 30 percent. NTT plans to invest around 2 billion dollars on the expansion of data centers, networks, and solar power projects in India. This current data center footprint expansion in India is part of the ongoing growth strategy of the Global Data Centers division which operates the third-largest data center platform in the world, with over 160 data centers spanning more than 20 countries and regions. The Mumbai 7 Data Center is operational with some marquee anchored customers on-board already. The Mumbai 7 Data Center is the third hyperscale data center in Chandivali campus and is well connected by fiber from all four sides. Together NTT’s Mumbai hub of data centers makes the Chandivali campus India’s first and largest operational hyperscale Data Center Park, which totals over 1,000,000 sq.ft., 13,000 racks and 100 MW of load capacity. “India has been a key market for NTT’s data center portfolio. Organizations today demand an ever-expanding global platform to reach their growing digital business objectives,” said Masaaki Moribayashi, Senior Executive Vice President, Services of NTT Ltd. “That’s why we continue to expand our portfolio of state-of-the-art data centers in new and existing markets that complement our global geographic footprint and support our clients as their demand increases for reliable, robust cloud services, cloud communications, digital entertainment, and new technologies such as artificial intelligence.” Sharad Sanghi, CEO, Global Data Centers and Cloud Infrastructure of NTT in India said, “We are extremely pleased to increase our data center footprint in Mumbai, during this pandemic to enable our clients to derive greater value while delivering business outcomes. Our data center facilities are poised to become the very heart of India’s IT Infrastructure needs providing the right platform with end-to-end ICT solutions combining hosted infrastructure, network, security and managed services – to balance the critical IT load of businesses.” With NTT reinforcing their commitment to following sustainable best practices while growing their data centers platform, Global Data Centers and Cloud Infrastructure division in India has launched its first 50 MW Solar Photovoltaic (PV) power plant in Solapur, Maharashtra, which is built in partnership with Tata Power, one of India’s largest integrated Power Company. The company’s new Solapur power plant is aimed at meeting 83% of the power needed by the Mumbai data centers. The capacity can be expanded to generate another 50 MW solar power in the future. According to 451 Research, NTT is currently the leader in India data center market. India is growing at 23% CAGR (2017-2022) due to increased demand from global cloud providers and a proposed data sovereignty law by the Indian government, as well as less-than-ideal market conditions that have been playing elsewhere in the Asia Pacific region. With the addition of Mumbai 7 Data Center, NTT Ltd’s Global Data Centers and Cloud Infrastructure division in India currently operates with 10 data centers across 4 major cities with over 1.5 million sq. ft. and over 150 MW of power. The company plans to double its capacity in the next two to three years to strengthen its hybrid ICT solutions and support the digital transformations of customers in India.","excerpt":"NTT’s Global Data Centers division today announced the expansion of its line of data center facilities with the launch of a new high-density and hyperscale data center in Mumbai, the financial capital of India. The Mumbai 7 Data Center operates with 375,000 sq.ft of colocation space, and will offer the capacity to host 5000 racks […]","categories":["AI News"],"tags":["data centre india","NTT"],"author_name":"Vishal Chawla","publish_date":"2020-09-30T13:20:00","publication_year":"2020","word_count":595,"keywords":["data centre india","Go","API","artificial intelligence","programming_languages:R","AI","digital transformation","Git","Aim","GAN","R","NTT"],"extracted_tech_keywords":["AI","artificial intelligence","Aim","R","Go","Git","API","GAN","digital transformation","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/ntt-expands-its-data-center-capacity-in-india-with-new-mumbai-facility\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10001165,"title":"How To Build VR Gaming PC Under ₹1.5 Lakh","content":"VR is widely considered by gaming enthusiasts to be the future of entertainment. With modern technology, it is possible to plug into a world where our minds are tricked into perceiving virtual environments as real. With that being said, the first generation of VR products have already emerged in the market a few years ago and has also seen an uptick in adoption among those who are willing to be on the cutting edge. The VR machine itself, whether it be the Oculus Rift or the HTC Vive, has a steep cost and the  PC performance required to run VR games at suitable quality demanded deep pockets. After the initial cycle of hype, VR has slowly gained a place in the mainstream gaming market, and it is possible to find good games to play on the platform. Further applications such as a 3D environment for creators and 360-degree video have also increased the viability of VR. The barrier of entry has also lowered somewhat, due to the availability of lower priced computer parts and reduction of price in the VR headsets. While the Oculus itself costs around ₹46000, it is possible to build a complete gaming setup, including the headset, for under ₹2 lakhs. In this article, we will look into how to build a gaming PC for smooth VR gaming for a price of ₹1.5 lakhs. As computer components are subject to an 18% tax under GST, this is not included in the price. The CPU As Intel and AMD are neck-to-neck in this market, it is important to look at bundled accessories, software, and overall value for money when purchasing a component. Keeping this in mind, we have compiled two lists of components; one for Intel, and the other for AMD. On the AMD side, the competitor is the competitively priced Ryzen 7 2700X. This processor comes with 8 core and 16 threads, with a clock speed of 3.7 GHz. It runs on the Zen+ platform, and is a good fit for any gaming machine due to its overclockability, as indicated by the ‘X’ suffix. The processor also comes with a Wraith Prism LED Cooler and support for software known as StoreMI that allows for faster boots through intercommunication between the SSD and the HDD. It is priced at around ₹25,000. Paired with the Gigabyte Aorus X470 motherboard, the AM4 socket on which this CPU function allows for promised support until 2020. The motherboard comes with a bevvy of enthusiast grade features and is priced at ₹20,500. The Intel competitor is the i7 8700K, which is also overclockable as indicated by the ‘K’ moniker. It comes with 6 cores and 12 threads and is clocked at 3.7 GHz. It offers slightly higher performance over the Ryzen, and also comes in at a slightly higher price at around ₹28,000. However, there is no included cooler, necessitating the need for a Cooler Master Hyper 212 Evo Air Cooler, which will set you back around ₹6000. Intel does not offer any additional features, although the motherboard comes in cheaper. The Gigabyte Z390 UD is an enthusiast class motherboard that allows the processor to be overclocked, and costs around ₹15,000. While the Ryzen setup offers more cores and threads, gaming is still a single-threaded workload most of the time. The slightly higher single core clock speeds of the Intel processor could give it a slight edge over the Ryzen, but the latter’s included cooler and overclocking software fight with Intel on value grounds. The Intel components present a price of ₹51,000, while the AMD components come in at just under ₹45,000. The GPU In a VR gaming machine, the most important part is the GPU. For this build, it is recommended to go with a GTX 1070Ti AMP Edition from Zotac while stocks last. The newer generation of cards released by Nvidia charge a premium for their real-time raytracing compute, and are not recommended in a budget build. Moreover, the 1070Ti far outstrips the recommended requirements given by both Oculus and HTC, as it provides a significant performance boost over the recommended GTX 970, and directly supersedes it. The 1070Ti is immensely powerful and is closer to 1080 than 1070. It is VR ready out of the box, and also has the required dual HDMI 1.4 outputs required for the Rift and Vive. However, to incentivise customers to buy the newer and more expensive RTX cards, Nvidia has stopped the production of the GTX cards. This has led to a shortage in the stock of the card, which is quickly dwindling. The AMP edition also comes with a hefty factory overclock, allowing the GPU to perform at peak capacity. The card is priced at around ₹45,000 at certain retailers. It is recommended to buy before stocks run out. In case of being unable to source a 1070Ti, customers can also go for the newest AMD offering known as the Radeon VII. It is priced slightly higher, at ₹55,000, and includes 3 games worth ₹9000. Stocks are assured, and the card handily beats the 1070Ti in performance terms. The Sapphire variant is recommended, as it comes with a factory overclock. Other Components The total for the rest of the components required to build the PC comes up to ₹43,500. It is a standard approach, incorporating 16GB of RAM running at 2666 MHz priced at ₹11,000, a Corsair VS650 PSU priced at ₹6,500, a Kingston 250GB SSD at ₹2.500, a WD 2TB HDD at ₹4,500 and the Cooler Master MasterBox Lite Case at ₹6,000. A high refresh rate, low response time monitor such as the Acer KG221Q can be purchased at around ₹11,000, with other peripherals resting at under ₹2,000. The Total Cost The Intel setup is slightly higher in cost owing to the added cost of the cooler and processor, coming in at around ₹1.4 lakhs. The AMD setup is cheaper by only a slight amount, with a price of ₹1.33 lakhs. In case of buying the Radeon VII, the GPU will set the buyer back by an additional ₹10,000, bringing the price to ₹1.5 lakhs and ₹1.43 lakhs respectively, not to mention 3 free games. This PC will be able to play the latest VR games at high refresh rates and quality, reducing the amount of nausea felt by first-time users. It will also provide a much smoother experience than using the recommended setup. Moreover, it also functions as a fairly high-end gaming PC, moving into the 100+ FPS range on games such as GTAV and Fortnite.","excerpt":"VR is widely considered by gaming enthusiasts to be the future of entertainment. With modern technology, it is possible to plug into a world where our minds are tricked into perceiving virtual environments as real. With that being said, the first generation of VR products have already emerged in the market a few years ago and […]","categories":["AI Features"],"tags":["AMD","Intel","NVIDIA","VR"],"author_name":"Anirudh VK","publish_date":"2019-02-14T16:40:34","publication_year":"2019","word_count":1085,"keywords":["Go","AMD","TPU","ELT","programming_languages:R","AI","programming_languages:Go","VR","Ray","NVIDIA","R","Intel"],"extracted_tech_keywords":["AI","Ray","TPU","R","Go","ELT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/vr-gaming-pc-1-lakh\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10764,"title":"Meet Jarvis – Facebook’s Messenger Chatbot","content":"Have you met Jarvis? Well, Jarvis is Facebook’s messenger chatbot which you can use to set your reminders. Last month, Facebook’s CEO Mark Zuckerberg announced its new chatbot platform at its F8 conference. Jarvis is one of the early entrants, allowing you to set reminders with a chat conversation directly inside Facebook Messenger. Starting with Jarvis To get started you can visit hellojarvis.io and press the tab ‘Talk to him on messenger’ to get started. Alternatively you can just search “Hello Jarvis” in the web-based Messenger or the Messenger app on Android.or iOS to get started. As soon as you start with Jarvis, on the very first contact Jarvis will ask you to state the city which you are living in.  This is all part of the linear conversation flow, as it allows Jarvis to understand and get an instant idea of your location and timezone. This information allows him to set reminders which are worthwhile to the user. Once you enter the city name, Jarvis will confirm the timezone by messaging you the current time at your location. The chatbot is then all set to take orders and set reminders on your behalf. Setting Reminders You can set simple reminders with Jarvis, say for instance “remind me to pick up milk tomorrow at 17.00” or “remind me to go to the doctor first thing in the morning”.  Upon setting these reminders, Jarvis will reply with confirmation that he has understood the command and provide a recap of what he will be actually doing and what time the reminder will come. And all this takes place quite instantly and inline in a standard messenger conversation. Jarvis sends reminders to you on the Messenger and you will get them on your desktop as well as your phone. These reminders can be set either for a particular day when you wish to be reminded or for a relative time such as thirty minutes from now. Your timezone is considered by Jarvis while sending you reminders. In case you wish to change your timezone or see all your reminders which you have set, you just need to type Help and sent it to Jarvis. Jarvis will reply with buttons specific for changing the timezone and for viewing reminders. But the problem happens when you want to delete one specific reminder as you cannot delete one reminder. Jarvis only gives the option to delete all the reminders as of now. The bot is in its initial phase right now and doesn’t seem to be powerful like Google Now or Siri. Nevertheless, Facebook’s Jarvis is a neat way to present what facebook chatbots can do.","excerpt":"Have you met Jarvis? Well, Jarvis is Facebook’s messenger chatbot which you can use to set your reminders. Last month, Facebook’s CEO Mark Zuckerberg announced its new chatbot platform at its F8 conference. Jarvis is one of the early entrants, allowing you to set reminders with a chat conversation directly inside Facebook Messenger.   […]","categories":["IT Services"],"tags":["AI (Artificial Intelligence)","AI Chatbot","Facebook","Jarvis"],"author_name":"Manisha Salecha","publish_date":"2016-09-23T13:56:59","publication_year":"2016","word_count":440,"keywords":["Go","programming_languages:R","AI Chatbot","chatbots","programming_languages:Go","Jarvis","Facebook","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["chatbots","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/meet-jarvis-facebooks-messenger-chatbot\/","complexity_score":2,"technical_depth":5,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10143338,"title":"GM Abandons Robotaxi Biz to Focus on Personal Autonomous Vehicles","content":"General Motors (GM) has announced a strategic shift in its autonomous driving efforts, concentrating on developing systems for personal vehicles rather than robotaxis, as the company announced on Tuesday. The company aims to enhance its advanced driver assistance systems (ADAS), building on the success of its Super Cruise feature, which operates on over 20 vehicle models and logs more than 10 million miles monthly. The move will integrate Cruise LLC, GM’s majority-owned autonomous vehicle subsidiary, with GM’s technical teams. The company will also cease funding Cruise’s robotaxi development, citing the time, resources, and competition required to scale that business. “GM is committed to delivering the best driving experiences to our customers in a disciplined and capital-efficient manner,” said Mary Barra, GM Chair and CEO. The decision comes as the robotaxi industry faces increased competition and regulatory hurdles. Cruise, GM’s self-driving subsidiary, had faced significant setbacks, including a critical incident in October 2023 in which a Cruise vehicle dragged a pedestrian, resulting in severe injuries. Following the incident, California revoked Cruise’s driverless testing permit, leading to a temporary suspension of its operations. Dave Richardson, GM’s senior VP of software and services engineering, claims, “We are fully committed to autonomous driving and excited to bring benefits like enhanced safety, improved traffic flow, increased accessibility, and reduced driver stress to our customers.” While Cruise CEO Marc Whitten and other leaders will remain to guide the transition, GM did not clarify the future of Cruise’s employees. The company currently owns 90% of Cruise and plans to increase its stake to over 97% and acquire the remaining shares, subject to Cruise board approval. The restructuring aims to cut annual spending by over $1 billion, with completion targeted for the first half of 2025. What are people saying? George Hotz, president at comma.ai, took to X to express that he had called this development five years ago. He says, “If you have been paying attention, self-driving has played out exactly like I said it would.” Earlier this year, GM had also let go of Origin, a fully autonomous shuttle developed by the company in partnership with Honda. Kyle Vogt, former co-founder at Cruise and Twitch, had then expressed that he was “disappointed to see GM kill the Origin.” Disappointed to see GM kill the Origin. Would have been amazing for cities.GM repeatedly finds themselves with a 5-10 year head start, but then fumbles the ball, shuts things down, and loses the lead. Anyone remember the EV1?It’s like someone keeps letting them look into a… pic.twitter.com\/GDlL4KQk4S— Kyle Vogt (@kvogt) July 23, 2024 Regarding Cruise, Ian Kar, General Partner at Vol. 1 Ventures, says, “Cruise, I think was a huge opportunity for an incumbent to actually move the needle and actually be innovative.” Garry Tan, president & CEO at Y-combinator, also took the opportunity to comment on this. Not sure what will be a bigger own-goal in the end—GM shutting down Cruise or Twitter shutting down Vine— Garry Tan (@garrytan) December 11, 2024 Cost and Competition Drive Shift GM’s chief financial officer, Paul Jacobson, said during an investor call that continuing with Cruise’s robotaxi ambitions would require substantial additional capital beyond the $10 billion already invested. He noted, “The better use is to pursue improvements to L3 and L4 autonomy in personal vehicles, aligning with the needs of the business going forward.” GM plans to integrate Cruise’s operations with its technical teams to enhance technologies like its current Super Cruise system, which offers hands-free driving in select models. The restructuring is expected to reduce GM’s expenditures by over $1 billion annually starting in 2025. Broader Implications The decision highlights broader struggles in the robotaxi market. Ford and Volkswagen shuttered their joint self-driving venture, Argo AI, in 2022. Meanwhile, competitors like Tesla, Waymo, and Amazon continue to explore the sector. GM’s decision follows other cost-cutting measures, including scaling back investments in a battery plant joint venture and streamlining operations in China. Despite these challenges, GM aims to leverage its expertise in vehicle manufacturing to redefine autonomous driving in the personal car segment.","excerpt":"The restructuring aims to cut annual spending by over $1 billion, with completion targeted for the first half of 2025.","categories":["AI News"],"tags":["Tesla"],"author_name":"Sanjana Gupta","publish_date":"2024-12-11T19:03:22","publication_year":"2024","word_count":673,"keywords":["Go","API","funding","programming_languages:R","AI","ML","programming_languages:Go","RAG","Aim","Tesla","R"],"extracted_tech_keywords":["AI","ML","Aim","RAG","R","Go","API","funding","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/gm-abandons-robotaxi-biz-to-focus-on-personal-autonomous-vehicles\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":59597,"title":"How SMBs Can Create Better ROI From Their Data Science Team","content":"If you ask any small and medium-sized business (SMB) about its top priority, then their answer, similar to larger companies, typically will revolve around improving customer experience, along with retaining the growing customers as well as increasing revenue. Working with data to bring actionable insights has proven to be the way forward to achieve growth and to stay competitive in the market. However, collecting and storing data is not beneficial unless it can actually drive actionable insights that propel the business forward, and to achieve that creating a data science team for your business is the most crucial step. This data science team can help the businesses in building projects that will be deployed into production for bringing value to their business. However, to derive the maximum output from this data science team businesses must embrace the structures and resources necessary for them to thrive and generate impressive ROIs. Here are some simple steps for businesses to follow, in order to gain maximum profit out of their small data science team. A direct connection with your customers Customer engagement is the key to having a successful business, and therefore businesses must thrive on meeting their customers’ requirements. But, in order to have a better engagement with your customers its always suggested to make them connect to your core team who will be working to resolve your customers’ problem; case in point — your data science team. To enhance your customer engagement, the data science team needs to connect with customers in order to understand their problem and use the same empathy to build specialised solutions for them. Companies can also better interact and engage with their customers by analysing their feedback to improve a product or service. Data sources include traditional in-house data, social media, browser logs, text analytics, and large public data sets can be useful for companies to understand their customers. Jack of all trades The cash crunch is a big issue for small and medium businesses. Therefore the best way to lead forward is to hire people who can multitask in your small organisation and fill the gaps that can be created due to understaffing. Experts suggest to get the maximum ROI; the organisation should create hierarchy and build teams with specificities; however, for SMBs it is more feasible to hire engineers who are more like a jack of all trades. A jack of all trades will have an idea of of the whole business and can provide insights to all the work processes related to engineering and data science. It must be easier for larger companies to have specific employees for each process, but for SMBs it is more important to cover all the positions to get a better view instead of hiring masters and experts at the start. Furthermore, by hiring a jack of all trades, SMBs can familiarise themselves with the necessary programs and software available in the market. Creating a generalised team also increases job satisfaction by creating more exposure for employees. Workplace mentorship is the key Mentorship has always been the key to enhance the productivity of your data science team; however, it can be an expensive affair for SMBS, but completely necessary in this dynamically expanding field. For SMBs to achieve a high-functioning data science team, they can easily rely on workplace mentorship or peer-to-peer mentorship to save the cost of outsourced mentorship. SMBs should arrange workshops and training session inside the organisations where they can open knowledge-sharing by employees and teams. Also, to support such kind of initiatives, SMBS should develop infrastructure. Such initiatives will provide an opportunity for employees to upskill themselves, which in turn will benefit the organisation in a longer run. Collaboration has always been a critical point in order to work efficiently. Also, working together, by sharing insights, will allow your team to tackle more significant problems, leverage individual strengths and avoid depending too much on one person. Identifying Trends For SMBs to survive in this dynamic market, spotting and monitoring trends are imperative. Following behaviours and patterns of the technology and the market will allow companies to take a stab at predicting where the market is heading, or what is the demand of particular products or services that usually changes over time. Often, trend analysis and prediction have been instantiated out of the gut, but with the advent of big data, businesses can now use their data to do the guesswork of the process. Technology has been a dynamic and ever-evolving field and to survive with your small data science team, SMBs need to be aware of the data trends and market information to keep their relevance on. Feedbacks can provide encouragement Lastly, in order to get maximum ROI from your data science team, SMBs should encourage their customers’ to provide constructive feedback to their data science teams to provide encouragement and also to guide them better about their problems. SMBs need to treat their data science team similar to their other departments of the company to create an inclusive environment for data scientists and engineers. Feedbacks not only help data scientists understand their customers’ business but also aid them in creating better solutions for their customers. Making continuous feedback a part of your organisation’s culture will help SMBs to enhance the productivity of their data science team.","excerpt":"If you ask any small and medium-sized business (SMB) about its top priority, then their answer, similar to larger companies, typically will revolve around improving customer experience, along with retaining the growing customers as well as increasing revenue. Working with data to bring actionable insights has proven to be the way forward to achieve growth […]","categories":["AI Features"],"tags":["Data Science","data science software","Data Scientist","SMBs","SMBs india"],"author_name":"Sejuti Das","publish_date":"2020-03-24T09:32:09","publication_year":"2020","word_count":883,"keywords":["big data","data science","TPU","programming_languages:R","AI","Data Scientist","SMBs","RAG","ViT","analytics","SMBs india","GAN","Data Science","data science software","R"],"extracted_tech_keywords":["AI","data science","analytics","RAG","TPU","R","big data","GAN","ViT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-smbs-can-create-better-roi-from-their-data-science-team\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":49375,"title":"Will China Mega Fund For Chip Manufacturing Overcome US Blacklisting?","content":"There is another twist in the tale for China and US technology trade war. In response to the recent export blacklisting of Chinese artificial intelligence firms like Huawei and SenseTime, the Asian superpower has launched a multi-billion-dollar fund to boost the domestic manufacturing of silicon chips, including NAND, DRAM, CPUs and GPUs so that its fast-growing tech industry is not be hampered. China is currently the world’s biggest importer of silicon chips amounting to $200 billion in annual spending. The $28.9 billion fund, also known as the China National Integrated Circuit Industry Investment Fund, is aimed at China’s self-reliance in the semiconductor supply chain which means that the Asian superpower will no longer depend on American companies like Intel, AMD, Qualcomm for its computing needs, including data centers, smartphones, laptops and other smart gadgets. China has been seeking to beat the US in the technology race, particularly when it comes to artificial intelligence.The significance of home-grown chips is huge given the global economy is increasingly relying on technological advancements of the digital age. In this light, 27 organisations including Chinese Ministry of Finance and China Development Bank in total came together to pool the funds. The Trade War Extends To Chip Manufacturing The fund comes right after the US blacklisted multiple Chinese AI firms that were alleged to have been involved in developing surveillance systems for the Chinese government. From example, Sensetime- a company which had an academic partnership with MIT and whose revenue has been climbing fast to a billion annual dollar — was included in the export blacklist and barred from buying components from US companies. The animosity between the two countries has been well documented recently, particularly in cases where the US regulators asked its citizens to abandon Chinese products and services from companies like Huawei and TikTok, calling them potential national threats. China Chip Manufacturing Still Lags In Bleeding Edge Technology In recent years, China’s Semiconductor Wafer Fabrication (FAB) capacity has been growing the fastest in the world, and given the impetus of the new foundry and chip projects driven by domestic and foreign players. Two major types of chips that the Chinese are keen on being able to build are memory-focused chips such as NAND and DRAM. China already seems to be closing the gap in the manufacturing of NAND, a non-volatile memory chip utilised for storage in all things ranging from 16GB MicroSD cards to high-performance SSDs. Yangtze Memory, a state-backed chipmaker which was reported recently to have started volume production of the country’s first homegrown 64-layer 3D NAND flash chips. But for advanced CPUs and DRAM, the country lags behind its global counterparts. The biggest chip manufacturing company in China, Semiconductor Manufacturing International Corporation (SMIC) started manufacturing CPUs based on 14nm technology in 2019 whereas TSMC in neighbouring Taiwan is producing high-performance CPUs on its 7nm process. In fact, TSMC is starting a US$19.6 Billion 3nm fab in S. Taiwan. In fact, all leading fabless semiconductor companies such as Qualcomm, Nvidia, Advanced Micro Devices (AMD), MediaTek, Marvell and Broadcom are customers of TSMC. Lesson For India India has been lagging behind in fab units expansion and there is no large scale semiconductor manufacturing in the country at the moment. Factors like the lack of infrastructure and lack of skilled labour in the country are to blame. On the other hand, the country has seen substantial growth in chip design services in the last decade. The Indian government is certainly working to promote the manufacturing of chips in the country, and may also look to capture a small share of chip manufacturing market given US silicone designers may look to India for manufacturing needs. The incident is a lesson for India in self-reliance, where a government fund can be used to boost the semiconductor ecosystem. Overview It seems that China has taken the development of the semiconductor industry in its own hands. The creation of a large fund focused on semiconductor chips sends a signal that China will not depend on US companies and increasingly move towards developing chips within the country that will fuel its 5G, video sensing and other AI-related initiatives. It also shows Chin’s stubbornness to not give in to US tactics when it comes to the ongoing trade war and become self-reliant instead. Despite being years behind in chip manufacturing, the nation has set forth actionable plans involving billions of dollars to compete head-to-head with the global superpower in technology. The fund will further strengthen global protectionist policies that rose with the China-US Trade War ever since Donald Trump rose to power in 2016.","excerpt":"There is another twist in the tale for China and US technology trade war. In response to the recent export blacklisting of Chinese artificial intelligence firms like Huawei and SenseTime, the Asian superpower has launched a multi-billion-dollar fund to boost the domestic manufacturing of silicon chips, including NAND, DRAM, CPUs and GPUs so that its […]","categories":["AI Features"],"tags":["China","china ai","China AI strategy","Chip Manufacturing","recent technological advancements"],"author_name":"Vishal Chawla","publish_date":"2019-11-04T17:46:06","publication_year":"2019","word_count":764,"keywords":["recent technological advancements","Go","artificial intelligence","AI","RPA","Git","RAG","china ai","Aim","China AI strategy","RNN","GAN","R","Chip Manufacturing","China"],"extracted_tech_keywords":["AI","artificial intelligence","Aim","RAG","R","Go","Git","GAN","RNN","RPA"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/china-chip-manufacturing\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10053231,"title":"IIT Jodhpur To Launch IoT AI Hub, Plans To Integrate Academics With Real-World Scenarios","content":"Indian Institute of Technology (IIT) Jodhpur recently announced that its Artificial Intelligence of Things (AIOT) Center is getting ready to go live soon. Artificial intelligence (AI) and the Internet of Things (IoT) have come together in the AIOT hub, which makes use of mobile communication technology (5G and beyond). IoT infrastructure and artificial intelligence (AI) technologies are being seamlessly integrated to improve edge AI efficiency, human-machine interactions, data management, and analytics. The hub will have three major characteristics — interconnected and intelligent devices, self-learning and self-healing human and machine workflows, and autonomous decision-making with edge computing. “The AIOT Innovation Hub and JCKIC are initiatives of IIT Jodhpur to connect academics with real-life and industrial-scale problems. This will enrich the innovation ecosystem of IIT as well as Jodhpur city. We expect that in the near future, this will trigger local industrial development through start-ups and new generation MSME, who are eager to take forward the agenda of Atmanirbhar Bharat,” said IIT Jodhpur Director, Santanu Chaudhury. Vice President at the foundation stone laying ceremony of Fab Lab for AIoT systems at IIT Jodhpur. A Fab Lab is a place where students use fabrication facilities to turn their ideas into reality. AIoT blends Artificial Intelligence (AI) with the Internet of Things (IoT). #IIT pic.twitter.com\/1Cpb0RvvUN— Vice-President of India (@VPIndia) September 28, 2021 Faculty and students at IIT Jodhpur and industry partners will be able to collaborate on AIOT technologies and products as a result of this initiative. These business partners can be brought together for the purpose of promoting the consumption of products, thereby increasing the likelihood of future success. Starting in the IIT Jodhpur technology park, the AIOT Innovation Hub (including photonic systems) will help establish facilities for end-to-end design, development, prototyping, and delivery of AIOT (including photonic systems). For the development of low-cost indigenous sensors, IIT Jodhpur proposes the establishment of a sensor fabrication foundry. Additionally, this facility is designed to support the production of sensors in order to reduce importation.","excerpt":"The hub will have three major characteristics — interconnected and intelligent devices, self-learning and self-healing human and machine workflows, and autonomous decision-making with edge computing.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","Courses","Data Science","Data Scientist","Deep Learning","IIT","IIT Jodhpur","IoT","iot training india","Machine Learning"],"author_name":"Victor Dey","publish_date":"2021-11-11T13:42:16","publication_year":"2021","word_count":329,"keywords":["iot training india","R","artificial intelligence","analytics","Courses","Data Science","Go","IIT Jodhpur","AI","ML","Machine Learning","edge AI","edge computing","IoT","programming_languages:R","innovation","Deep Learning","IIT","Data Scientist","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","ML","analytics","edge AI","edge computing","R","Go","innovation","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/iit-jodhpur-to-launch-iot-ai-hub-plans-to-integrate-academics-with-real-world-scenarios\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10052310,"title":"Is Interpolation Overrated?","content":"Interpolation and extrapolation are fundamental concepts in fields like machine learning and deep learning. The difference between the two is that interpolation occurs for a sample x when it falls inside or on the boundary of the given dataset’s convex hull, while extrapolation happens outside this convex hull. In a recent paper, the authors Yann Lecun, Randall Balestriero and J´erˆome Pesenti aim to dispel two major misconceptions. The first misconception is that state-of-the-art algorithms work well since they can correctly interpolate training data. The second misconception is that interpolation happens throughout tasks and datasets. The authors argue that on any high-dimensional (>100) dataset, interpolation almost never happens. They also claim that their results challenge the validity of the current interpolation\/extrapolation definition. Interpolation & Extrapolation In the earliest recorded versions (Kolmogoroff, 1941 and Wiener, 1949), extrapolation was defined as the method for future prediction of a stationary Gaussian process based on past and current realisation. Interpolation was defined as predicting the possible realisation of a Gaussian process at a time position that lies between observation; one can even say that interpolation resamples the past. Currently, the most accepted definition of interpolation is that it happens for a sample x when this sample resides in the convex hull of a set of samples X. In cases it does not, extrapolation occurs. This definition indicates that extrapolation is a more intricate task than interpolation — the latter guarantees that the sample lies within the dataset’s convex hull, and the former leaves the entire remaining space as a valid sample position. This definition has been used in fields like function approximation, machine learning, and deep learning. Interestingly, such a definition has also given rise to an assumption that the algorithm transitions from interpolation to extrapolation as the performance decreases. The authors of this paper argue that interpolation\/extrapolation and generalisation performances are not as tightly linked as previously believed. They have attempted to demonstrate theoretically and empirically that for synthetic and real data, interpolation almost never occurs in high dimensional spaces, regardless of the underlying intrinsic dimension of the data manifold. They claim that: Current models are extrapolating.Given the super-human performance of those models, the extrapolation regime is not an indicator of generalisation performances. Interpolation & Curse of Dimensionality In the first stage of the study, the authors carefully studied the dimensions of the space in which data is placed, the number of variables needed for the minimal representation of the data, and the dimension of the smallest subspace that includes all the data manifolds. The commonly held belief for data such as images is that they lie on a low-dimension manifold and that interpolation occurs regardless of the high-dimensional ambient space. This study showed that such intuition could be misleading. In fact, the underlying manifold dimension does not help even in the case of having a 1-dimensional manifold. Instead, the parameter which matters the most is the dimension d* of the smallest affine subspace that includes the dimension of the convex hull of the data. Credit: Research paper It was observed that the increasing ambient space dimension has no impact on the number of samples required for maintaining the interpolation regime. The authors concluded that to increase the probability of being seen in the interpolation regime, the dimension d* needs to be controlled, not the manifold underlying dimension and ambient space dimension. Experiments also demonstrated that even with real datasets and various popular embeddings, interpolation remains an elusive goal, and it becomes exponentially difficult to reach with growth dimension. Wrapping Up “In particular, we opposed the use of interpolation and extrapolation as indicators of generalisation performances by demonstrating both from existing theoretical results and from thorough experiments that, in order to maintain interpolation for new samples, the dataset size should grow exponentially with respect to the data dimension,” the authors write. The behaviour of the model within the convex of the training set barely impacts the model’s generalisation performance since new samples most often lie outside the convex hull. This observation is true for both original data space and embeddings. The authors believe that these observations will help in constructing better suited geometrical definitions of interpolation extrapolation that are more suited for generalised performances in high-dimensional data. Read the full paper here.","excerpt":"The behaviour of the model within the convex of the training set barely impacts the model’s generalisation performance since new samples most often lie outside the convex hull.","categories":["AI Features"],"tags":[],"author_name":"Shraddha Goled","publish_date":"2021-10-25T17:00:00","publication_year":"2021","word_count":708,"keywords":["Go","machine learning","programming_languages:R","AI","programming_languages:Go","Aim","deep learning","R"],"extracted_tech_keywords":["AI","machine learning","deep learning","Aim","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/is-interpolation-overrated\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10062994,"title":"Experts believe a neuro-symbolic approach to be the next big thing in AI. Does it live up to the claims?","content":"In their 2009 manifesto, Neural-Symbolic Cognitive Reasoning, Artur Garcez and Luis Lamb, discussed the 1990s popular idea of integrating neural networks and symbolic knowledge. They cited Towell and Shavlik’s KBANN, Knowledge-Based Artificial Neural Network that uses a system to insert rules, refine and extract data from neutral network; a model empirically proving to be effective. Industry leaders, including contingencies at IBM, Intel, Google, Facebook, and Microsoft, and researchers like Josh Tenenbaum, Anima Anandkumar, and Yejin Choi, are starting to apply this technique in 2022. The recent AI developments, challenges, and stagnancy is leading the industry to consider this hybrid approach to AI modelling. Neuro-Symbolic AI is essentially a hybrid AI leveraging deep learning neural network architectures and combining them with symbolic reasoning techniques. For example, we have been using neural networks to identify the shape or colour a particular object has. Applying symbolic reasoning to it can take it a step further to tell more exciting properties about the object, such as the area of the object, volume and so on. Credits: Neuro-symbolic learning cycle according to S. Bader and P. Hitzler The stagnant AI: Why we need a neuro-symbolic approach AI has been the talk of the town for more than a decade, and while it has stood true to several promises, the majority of the claims are still to be met, and the challenges connected to AI have only been increasing. In the past year, GPT-3 has asked individuals to commit suicide, Alexa has challenged a ten-year-old to touch a coin to the plug, and Facebook’s algorithm has identified a man of colour as a primate. This is not any different from Microsoft’s Tay asserting ‘Hitler was right’ or Uber’s self-driving cars crossing red lights a few years ago. With every GPT-3 development, we have seen a downfall. The present efforts to ensure explainable, fair, ethical and efficient AI need to be supported by changes in how we approach artificial intelligence. Scientist and AI author and entrepreneur Gary Marcus recently wrote about deep learning hitting a wall and the responsibility of AGI. “It must be like stainless steel, stronger and more reliable and, for that matter, easier to work with than any of its constituent parts. No single AI approach will ever be enough on its own; we must master the art of putting diverse approaches together if we are to have any hope at all.” Most exciting AI systems are hybrids of neural and symbolic techniques.— Barney Pell (@barneyp) March 12, 2022 Neuro-symbolic #AI brings us closer to machines with #commonsense https:\/\/t.co\/758d1svauo #fintech #ArtificialIntelligence #MachineLearning #DeepLearning #BigData @bendee983@bdtechtalks @jblefevre60 @pierrepinna @ipfconline1 @HeinzVHoenen @WhiteheartVic @Nicochan33 pic.twitter.com\/ujG3c3kVCA— Spiros Margaris (@SpirosMargaris) March 15, 2022 The neuro-symbolic approach Since AI use learning and reasoning in a quest to be like humans, the neuro-symbolic approach allows us to combine these strengths to make inferences based on the existing neural networks and learn through symbolic representations. Knowledgeable Magazine asserted this hybrid to show duckling-like abilities. Ducklings can imprint colours and shapes and differentiate between them. Moreover, they can differentiate between “same” and “different”, an aspect AI still struggles with. Here, symbolic AI would involve symbols for physical objects and colours in its knowledge base. The base also consists of general rules to differentiate between them. This, combined with the deep nets, allows the model to be more efficient. Credits: SITN, Harvard, Getting all your ducklings in a row: a look inside the animal mind …People in the DL “camp” believe causal reasoning can be achieved though neural network like architectures that can be learned using backpropagation and\/or forms of RL. The other camp thinks we need forms of logical reasoning hardwired into these structures (aka symbolic AI)…— Yann LeCun (@ylecun) November 23, 2019 The combinations demand humans supply a knowledge base\/symbolic rules for the AI to leverage while the automated deep nets find the correct answers. “The hybrid uses deep nets, instead of humans, to generate only those portions of the knowledge base that it needs to answer a given question,” notes Knowledgeable Magazine. Real-life applications DeepMind has seen some of the best success with their board-game playing AI models Go, Chess, MuZero and more. These are hybrid models using symbolic AI. For instance, AlphaGo used symbolic-tree search with deep learning, AlphaFold2 combines symbolic ways of representing the 3-D physical structure of molecules with the data-trawling characteristics of deep learning. Deepmind has asserted the qualities of symbolic learning in AI in a recent blog post. “This approach will allow for AI to interpret something as symbolic on its own rather than simply manipulate things that are only symbols to human onlookers,” they said. This allows for AI with human-like fluency. IBM has asserted neuro- symbolic AI is ‘getting AI to reason’. The LNN technique, Logical Neural Network, was introduced and created on the foundations of deep learning and symbolic AI. Given the combination, the software can successfully answer complex questions with minimal domain-specific training. Two major conferences have been held asserting this need—the 2019 Montreal AI Debate between Yoshua Bengio and Gary Marcus and the AAAI-2020 fireside conversation with Laureate Daniel Kahneman, Geoffrey Hinton, Yoshua Bengio and Yann LeCun. The key takeaway from these events was the need for AI to have a reasoning layer with deep learning to frame a rich future of AI.","excerpt":"In the past year, GPT-3 has asked individuals to commit suicide, Alexa has challenged a ten-year-old to touch a coin to the plug, and Facebook’s algorithm has identified a man of colour as a primate.","categories":["AI Features"],"tags":[],"author_name":"Avi Gopani","publish_date":"2022-03-17T15:00:00","publication_year":"2022","word_count":882,"keywords":["Go","artificial intelligence","AI","neural network","RAG","GPT","Aim","deep learning","R","llm_models:GPT"],"extracted_tech_keywords":["AI","artificial intelligence","deep learning","neural network","Aim","RAG","R","Go","GPT","llm_models:GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/experts-believe-a-neuro-symbolic-approach-to-be-the-next-big-thing-in-ai-does-it-live-up-to-the-claims\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10072273,"title":"Indian IT Finds it Difficult to Sustain Work from Home Any Longer","content":"Post pandemic, the working models of companies have not changed much. While some companies are choosing office-first or remote-friendly options for their employees, some are offering partly remote hybrid, flexible hybrid, partly remote hybrid, choose-your-own-adventure hybrid, remote-first hybrid, and full remote models. Each of them has its own pros and cons. Simultaneously, with the rising attrition, the Indian IT companies are still figuring out working models that work for them, as well as their employees. Analytics India Magazine got in touch with Happiest Minds Technologies, chief people officer and vice president of people practice, Sachin Khurana, who said that the company has recently moved into a 3+2 hybrid model with employees coming to the office thrice a week and two days work from home (WFH). “As of now, over 50 per cent of our people are coming to office regularly, and we see these numbers going up gradually to 70-75 per cent,” added Khurana. Register for Genpact Career Day >> Meanwhile, LatentView chief Rajan Sethuraman told AIM that they largely follow work-from-home models. “I will say it is a bit hybrid, and about three-four months back, we indicated that we would like our people to come in and work two days in the office,” he added. However, he said that this is not compulsory and considering the spike in the cases, they decided to be cautious both in the US and India. “While the guidance is to come and work from the office for a couple of days a week, we are not making it mandatory yet,” shared Sethuraman, saying that they are leaving it to the employee and reporting managers to decide what is appropriate and provide the necessary flexibility to their people. Venkatesh Veerachamy, director of delivery and operations at Zuci Systems, said that they are following a gradually-scaling hybrid work mode, which involves about 30 per cent of employees working on-premise and others on remote mode. “We plan to scale up as we move along and when external considerations normalise,” he added. Other Indian IT giants, including TCS, Infosys, Wipro and Tech Mahindra, more or less are following a hybrid approach, with flexible work from anywhere options to select employees. Impact on attrition Happiest Minds’ Khurana said that the transition towards the hybrid work model has greatly helped in the team’s engagement and collaboration activities. He said that the attrition in the industry remained elevated over the last few quarters. “However, we do not see the attrition rate being impacted by our shift to a hybrid work model,” said Khurana, claiming that the response from their members has been very encouraging. Further, he said that the teams are relishing the opportunity to spend more time together; and to be able to innovate, collaborate, and exchange ideas more easily. Zuci Systems’ Veerachamy said that they had focused on people-centricity in line with their engineering philosophy that calls for enabling a creative ecosystem, including internal stakeholders. He further said that it did not witness an undesired material impact on their attrition rates. “Our attrition has been around 8-10 per cent, and we can see a drop of 15-20 per cent in the recent months. Compare this with the reported attrition rate of listed peers, that are typically in the range of 18-24 per cent.” LatentView’s Sethuraman said they do not have any concerns emerging from their employees based on the working model. He said that the attrition rate, in general, is high because of other reasons. This is influenced by macroeconomic scenarios, where the number of jobs that are there in the market is very high, particularly for data science and analytics. A lot of companies, not just pure-play analytics service providers like LatentView, but other IT service providers, product companies, startups, and GCCs—everybody is now looking to add talented professionals to their teams. The future of work remains uncertain As per the latest survey by TeamLease, 58 per cent of companies ranging from tech to manufacturing, BFSI, FMCG, retail, healthcare and automobile believe that 2022 is the year the working model will become completely in-office. However, the same report also said that about 5 per cent of companies intend to stay in a virtual-only organisation for the near future. On the employee side of things, 43.46 per cent of employees want to return to work, and 76.78 per cent of organisations want to give their employees the preference to choose their work model. Happiest Minds said that the last two years had changed the outlook toward future workplaces. Khurana said that they had many learnings. Some of the learnings from Happiest Minds Technologies include: work can happen from anywhere if the need arises—work from anywhere is possible, and in-person touch can not be replaced by technology. “While a sudden shift to work from home has taken the industry by surprise, we have managed the transition pretty well at Happiest Minds. However, the challenges did increase, and it would be difficult to sustain WFH for longer periods,” said Khurana. Further, he said that people’s engagement, sense of belongingness, culture, team dynamics, collaboration, innovation, etc., were impacted by the technology-enabled virtual work; it was challenging for managers to lead multi-location and virtual teams and still create team bonding that existed earlier. Khurana said that the unsaid expectation of being available always took a toll on people’s well-being. “To thrive in a digital landscape, the need for innovation or agility is high, which is also impacted. Attrition among post-pandemic hires increased substantially, and the scarce talent markets did not help either,” he added, saying that hybrid work models provide the best of both worlds and offer the flexibility of remote working\/working from home\/working from anywhere and the need for working together in a unit to flourish as a team. On the other hand, Veerachamy said that they prefer an office-first\/remote-friendly model. “We were flexible even before the pandemic and allowed remote work on a need basis—the reason why we provided laptops to employees and not desktops. But, at the same time, we understand the benefits of the offline huddle and hence the preference for the office-first\/remote-friendly model.” COVID blew the myth away that work cannot be done remotely even though the technology was there to make it possible. However, in current times, the WFH culture has swung to an extreme, and with the return to normalcy now, a hybrid workplace appears to be the way forward to an appropriate balance between the needs of the organisation and an individual’s need for flexibility. The exact quantum of days of WFO (work from office) would vary depending on each team’s requirements. For instance, internal teams such as Finance or IT could have higher flexibility for WFH in comparison to customer-facing teams in Indian IT, where customers may mandate a higher number of days of WFO. Additionally, now that many of the Indian IT companies are also opening satellite offices in tier-2 and tier-3 towns where a good number of employees have relocated, I personally believe that most Indian IT companies are most likely to settle for 3 days of WFO and 2 days WFH. We will also see leaders in Indian IT organisations leading the way by returning to the office and encouraging their teams to plan for WFO for a couple of days during the week.","excerpt":"Hybrid work models provide the best of both worlds and offer the flexibility of remote working\/working from home\/working from anywhere.","categories":["IT Services"],"tags":["Indian IT"],"author_name":"Amit Naik","publish_date":"2022-08-08T10:00:00","publication_year":"2022","word_count":1214,"keywords":["data science","Go","AI","ML","Git","RAG","Aim","analytics","GAN","Indian IT","R"],"extracted_tech_keywords":["AI","ML","data science","analytics","Aim","RAG","R","Go","Git","GAN"],"url":"https:\/\/analyticsindiamag.com\/it-services\/indian-it-finds-it-difficult-to-sustain-work-from-home-any-longer\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10118390,"title":"Keyboards will Soon Become Obsolete","content":"Entrepreneur and investor Naval Ravikant recently re-launched his social media app, Airchat, when there was no dearth of such platforms already. However, the USP of this one is that the app is completely voice-centric – the interaction is via voice only. Source: X Airchat might just be the latest entrant highlighting the power of voice, but a number of recent AI platforms and devices have already brought voice as a predominant user interface. Dawn of the Voice Era Multimodal AI was identified as one of Microsoft’s AI trends for the year, and going by the AI developments this year, voice modality is emerging as a key feature. The latest Humane Ai Pin, a small wearable device that performs as a personal assistant and even works as a probable replacement for a smartphone, essentially works on voice. Any type of interaction with the device like making calls, reading messages, and clicking pictures can be executed through voice commands. Bethany Bongiorno, co-founder of Humane, believes that voice will be an integral part of an AI future. “Voice-first in an AI future,” she said. Similarly, AI devices such Rabbit R1, a pocket-size gadget which is an integrated language action model, also operates on voice commands. Brett Adcock, CEO and founder of robotics company Figure AI, said, “We believe the default user interface for the robot is speech. You’re going to want to talk to the robot. Even in an industrial setting, when you’re unboxing the robot for the first time, we think the initialisation process is speech.” How Do We Assess Them? With voice models come a different set of evaluation parameters. Interestingly, the shift has started after a year-long emphasis on text-based AI generation. Benchmarks and leaderboards for evaluating text-based models have always been a topic of LLM discussions. The need has also given rise to an Indic LLM Leaderboard. However, leaderboards for voice-based generative models are not that prominent. There are evaluation parameters for voice-based models such as latency, word error rate (WER), short-time objective intelligibility (STOI), miss-rate, and ROC Curve. These parameters measure accuracy in terms of speech quality and speech intelligibility. Shift from Text to Voice Chatbots that aid multiple functions such as HR operations or finding love, are essentially text-based. However, there is a casual shift now. Last month, Hume AI released EVI, an empathetic voice interface AI model. Users can converse with the model normally, where the model will be able to analyse and understand a user’s emotion based on the tone of the voice and other features. It almost serves as a therapist. Hume comes as a huge shift from other similar platforms such as Inflection’s Pi, which acts as an emotionally intelligent AI that helps with one’s emotional needs. Not All Big Tech are Gung-ho While the big tech companies are integrating voice in one form or another, be it OpenAI’s ChatGPT or Google’s Gemini, the models are multimodal allowing voice interface as a normal mode. Interestingly, a major player, Apple, is not too keen on this form of modality yet. For a company making strides in bringing generative AI features to its phone, and even releasing AI models such as RealM that could possibly beat GPT-4, Apple is yet to catch up on the voice game. However, voice is not completely alien to Apple. Apple’s spatial computing device, Apple Vision Pro can be controlled using voice features. Further, the company’s famed voice assistant, Siri, is expected to get advanced AI features which will probably be announced at the Apple WWDC 2024 event in June. The feature might be a major boost to Apple’s voice modality function. While voice is being increasingly adopted, companies are still relying on text-based chatbots. IT company, Happiest Minds recently announced ‘hAPPI’, a generative AI-powered chatbot that will converse with users on health and wellness-related queries. It is obvious that to get to the closest level-of human-like interaction, voice becomes indispensable. After all, “Humans are all meant to get along with other humans, it just requires the natural voice,” said Ravikant. PS: The story was written using a keyboard.","excerpt":"QWERTY anyone?","categories":["AI Features"],"tags":["AI","OpenAI"],"author_name":"Vandana Nair","publish_date":"2024-04-17T14:30:00","publication_year":"2024","word_count":679,"keywords":["Go","ChatGPT","API","OpenAI","AI","chatbots","GPT","generative AI","multimodal AI","R"],"extracted_tech_keywords":["AI","generative AI","multimodal AI","ChatGPT","OpenAI","chatbots","R","Go","API","GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/keyboards-will-soon-become-obsolete\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":51170,"title":"Amazon Floods Markets With ML-Powered Services At AWS re:Invent","content":"At the ongoing AWS re:Invent event, the tech giant announced various services categorised under analytics, compute and machine learning, among others. It, therefore, came as no surprise when the event was packed with machine learning services such as CodeGuru, Fraud Detector, Kendra, SageMaker Model Monitor, and more. In the ever-increasing technology landscape, along with automating mundane tasks, organisations are highly focusing on automating strenuous tasks with ML-powered solutions. While this will improve the productivity of the firms, it will also allow them to focus on their core businesses. Companies continually look to streamline the complete workflows right from the product development to delivery and customer satisfaction. In this article, we will take a look into how companies can integrate machine learning services to achieve their goals effortlessly. Product Development Robust products and services are crucial for any successful businesses. Thus they often seek services that could help them in the development phase. One of the vital aspects of applications is that it should work flawlessly. However, even after numerous tries, firms fail to deliver bug-free solutions. But with the integration of machine learning models in solutions, applications’ throughput decreases and thus companies lose a lot of time in optimising algorithms. However, with the recent release of CodeGuru, firms can expedite their development workflows. Trained with millions of lines of code from open-source projects and companies’ internal code, CodeGuru is designed to assist users in automatic code review. Using CodeGuru, developers can find bugs, get recommendations to fix or improve the code. Besides, one can identify problems such as resource leaks, wasted CPU cycles, and potential concurrency race conditions. As of now, CodeGuru provides supports for Java application, and in the future, it will add compatibility to other programming languages, thereby, making it mush have services for organisations. The service has the potential to simplify the development process while optimising applications’ performances. Application Enhancement Superior applications are more often than not are the differentiating factors in the highly competitive technology landscape. Therefore, organisations develop ML-based applications, but due to numerous challenges, they fail to improve the accuracy of ML models for rendering desired results. For instance, concept drift – ML models trained on one data with specific correlation often fail to provide desired results if there are any changes in the business operations – is among one of the pressing problems that firms find difficult to solve. To address such issues, AWS has introduced SageMaker Model Monitor to determine concept issues and increase models’ accuracy. SageMaker Modle Monitor automatically detects concept drift and notifies administrators to analyse changes through summary statistics and comparison over time with features used in training. Customer Satisfaction Post product delivery, providing useful services is of paramount importance to retain customers for business growth. Satisfied customers assist in increasing the brand of the firm, and in turn, brings new customers. Consequently, leveraging Amazon Kendra – a machine learning-based service that will enable companies to carry out enterprise search effortlessly – firms can quickly assist their customers in resolving their queries. Kendra gathers data from various sources within organisations and delivers relevant results. It crawls through multiple documents, including legal information to quickly, resulting in eliminating the time spent while searching within records. Anyone can query for information and Kendra will do the heavy lifting to provide the best results automatically. Backed by powerful natural language search abilities, it can also be deployed directly into websites, applications, and more, to obtain information from a vast amount of data spread across companies. Outlook AWS’ announcements have really made a mark as they have focused on the core problems that organisations face. Apart from the services mentioned above, AWS has also announced numerous ML-based solutions that can help firms to carry out their business efficiently. Machine learning is the key to simplifying the day-to-day activities of every employee in any organisation. Thus, adopting such AWS services can assist organisations in gaining operational efficiency. This will also empower organisations to offer advanced products while reducing the cost through ML-based services.","excerpt":"At the ongoing AWS re:Invent event, the tech giant announced various services categorised under analytics, compute and machine learning, among others. It, therefore, came as no surprise when the event was packed with machine learning services such as CodeGuru, Fraud Detector, Kendra, SageMaker Model Monitor, and more. In the ever-increasing technology landscape, along with automating […]","categories":["Global Tech"],"tags":["Amazon AWS","AWS services","uav companies in india"],"author_name":"Rohit Yadav","publish_date":"2019-12-05T13:02:38","publication_year":"2019","word_count":669,"keywords":["Go","machine learning","AWS services","AWS","AI","ML","RAG","Amazon AWS","analytics","GAN","uav companies in india","R","Java"],"extracted_tech_keywords":["AI","machine learning","ML","analytics","RAG","AWS","R","Go","Java","GAN"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/amazon-floods-markets-with-ml-powered-services-at-aws-reinvent\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":43924,"title":"Is Google’s Claim To Patent Batch Normalization A Step Towards Monopolizing Algorithms?","content":"The year 2018 has seen a meteoric rise in the number of papers released in the field of AI. There were also numerous tools and techniques open sourced by the giants to carry the baton of AI research forward. Google’s BERT, for instance, introduced new benchmarks for natural language understanding. Google has been pioneering research in the area of algorithms over the past decade. Their discoveries and innovations are usually open sourced making way for more innovations. However, Google, somehow infamously came under the scanner when it started to build an extensive portfolio of patents related to deep learning, including patents on essential technologies such as DQN, dropout, and now, batch normalization. These techniques are widely used across many popular machine learning models. And, since these machine learning models are deployed for speech, vision and other applications, the machine learning community have started to feel the heat. They are surrounded by a certain sense of paranoia about the rise of machine learning monopoly. Late last year, the United States Patents and Trademark Office (USPTO) issued the first non-final rejection) for the patent application with the application number 15-009647(batch normalization layers) filed by Google Inc. Batch Normalization Overview In the original BatchNorm paper, the authors, Sergey Ioffe and Christian Szegedy of Google introduced this method to address a phenomenon called internal covariate shift. This occurs because distribution of each layer’s inputs changes during training, as the parameters of the previous layers change. This slows down the training by requiring lower learning rates and careful parameter initialization. This makes training the models harder. The introduction of batch normalizsed networks helped achieve the state-of-the-art accuracies with 14 times fewer training steps. This reduction in training duty led to the emergence of many improvements within the machine learning community. For instance, Coconet is a fairly straightforward CNN with batch normalization. This gives Collaborative Convolutional Network (CoCoNet) more power to encode the fine-grained nature of the data with limited samples in an end-to-end fashion. The applications of CoCoNet was demonstrated recently when it was used to churn out Bach like  melodies with few clicks. Benevolence Or A Subtle Attempt At Monopoly? A patentable invention should meet five significant criteria: The patented invention must be constructed of patentable items and subjects. The inventions have to be usable in an industrial or other useful context. The invention must be original. The invention has to be inventive, or not obvious. The patent paperwork must meet the requirements of the patent office. After the issuance of non final rejection, Google, has made some amendments to their initial claims and now their patent for batch normalization is being reconsidered. The European Patent Office, however, have already granted Google a patent for the same batch normalization layers patent application The USPTO’s first office action can be characterised as “the complete package” of rejection reasons related to software patents, citing 14 instances of prior art (4 patents and 3 technical papers were explicitly cited in the office action), which provide the basis for denying the registrability of Google’s patent. Registering key patents related to deep learning, despite Google’s promotion of open source software, has become a topic of debate for many online forums. People started picking sides immediately. While one group reproved Google’s moves for being too naive, others put forth Google’s claim to be for the greater good. The way changes were made after the initial rejection seemed sly to many. And, since batch normalization gained popularity with the masses, few developers are left disconcerted whether to focus on advancement or fear for infringement. There is no denying the fact that Google has played a key role in democratisation of machine learning. Moreover, there is also a risk involved in keeping their findings out in the open only for some patent troll to claim it later, which can lead to exploitation. However, this also poses the question as to whether we should rely on the benevolence of tech giants. Future Direction “Every evolving intelligence will eventually encounter certain very special ideas – e.g., about arithmetic, causal reasoning and economics–because these particular ideas are very much simpler than other ideas with similar uses,” said the AI maverick Marvin Minsky four decades ago. The claim for the ownership of ideas have been popular at least since the time of Newton and Leibnitz. Innovations are usually result of accumulation of some ingenious ideas. Two people can stumble upon the same idea at the same time oblivious of its existence. When an invention is tossed between rightful ownership and democratisation, it delays the progress, which the invention has promised in the first place. For more information check this link.","excerpt":"The year 2018 has seen a meteoric rise in the number of papers released in the field of AI. There were also numerous tools and techniques open sourced by the giants to carry the baton of AI research forward. Google’s BERT, for instance, introduced new benchmarks for natural language understanding. Google has been pioneering research […]","categories":["Global Tech"],"tags":["batch normalisation","Google","Patent"],"author_name":"Ram Sagar","publish_date":"2019-08-07T11:00:12","publication_year":"2019","word_count":774,"keywords":["Go","machine learning","AI","innovation","Patent","batch normalisation","BERT","Aim","deep learning","Google","AI research","CNN","R"],"extracted_tech_keywords":["AI","machine learning","deep learning","Aim","R","Go","BERT","CNN","innovation","AI research"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/is-googles-claim-to-patent-batch-normalization-a-step-towards-monopolizing-algorithms\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":57187,"title":"Case Study: How This Mumbai-Based Startup Uses Mobile Data To Assess Credit Scores","content":"The customers of today’s’ era couldn’t be happier with the innovations in the fintech sector as getting a loan has never been easier. With technology getting infused in the finance sector, digital lending platforms have highly emerged as the most hassle-free and faster way for people to borrow money. SalaryDost, a consumer lending platform, is a similar innovation that has an aim of revolutionising the small loan market of India. Established in Mumbai — India, in 2018, SalaryDost has an innovative credit scoring system that helps in super customer profiling, and their mobile-based application helps its customers get a loan within a blink of an eye. The company has been designed to fill the gaps between the finance sector and the consumer by promoting open banking lending options to unserved customers. “Our fintech venture is building a unique platform that supports the vision of ‘loans – fast & easy’ and mission of ‘extending the salary’ of our customers,” said the founder and CEO of SalaryDost, Mrityunjay Shahi. The Challenge With operations in Thane — Mumbai, SalaryDost has a vision of establishing a credit line for every Indian and increasing their customers’ salaries. Since its inception, the company has disbursed more than 1.5 lakhs of digital loans already by early this year to more than 2,000 customers. Currently, it aims to have a 2-3% reduction for the non-performing loans, in the coming months, and to scale up their business without affecting the bad rates or approval figures. However, profiling of the customer and analysing credit scoring has always been the biggest challenge of lending operations. The faster the profiling works, the quicker the lending process gets. Therefore to fulfil the vision of establishing a credit line for every Indian possible, the company needed to find ways to improve the profiling of its customers while cutting down the turnaround time. The lending operation works on a massive amount of consumer data which gets gathered over time. SalaryDost works on alternative sources of data, such as smartphone metadata, with which they better understand the behaviour of the applicant and in turn, make effective informed decisions. Although the process is automated, with the increase in the number of applicants, the process gets slow and the churn around time increases. Another key challenge was the low predictability of the existing underwriting process. Therefore, to speed up its operation, the company wanted to implement a solution that is powerful and futuristic, giving them real-time insights into the data. According to Shahi, one of the critical business objectives of the company was to find quicker and easier ways to improve the profiling of the customers while cutting down the time-to-yes. The Solution So, to improve their process efficiency and reduce time lag, SalaryDost collaborated with a Singapore-based AI company, CredoLab. CredoLab’s AI-based credit scoring solution, CredoSDK, seamlessly integrates powerful AI credit scoring technology into the mobile application. It uses the smartphone data that is non-intrusive and non-personal in order to assess the behavioural score of the applicant within seconds. CredoLab, as a company, develops bank-grade digital scorecards for banks, consumer finance companies, lenders, insurance companies, and retailers from the best alternative data source, i.e. smartphone device metadata. Their AI-based algorithm crunches over 18 million features from opt-in smartphone metadata to find the most predictive behavioural patterns before converting them into credit scores. These enable any lender to make the most granular assessments possible of their applicants. CredoLab’s clients have seen 20% higher new to bank customer approvals, a 15% reduction in non-performing loans, and a 22% dip in fraud rate. With CredoLab, Salary Dost can now combine their bureau data with the behavioural score of the customer to get a complete overview of the applicant. Built on over 15 million datasets collected across 50 lending partners in more than 15 countries, CredoLab’s AI-based algorithm crunches over a million features from opt-in smartphone metadata to discover predictive delinquent behavioural patterns. CredoSDK then uses this valuable alternative data to produce highly accurate credit scores in real-time. The flexible APIs of CredoSDK ensured that their scoring technology could easily integrate with SalaryDost’s existing IT infrastructure. Through an easy-to-implement SDK, SalaryDost started using CredoLab’s AI-driven algorithm to convert over 1 million features from opt-in smartphone into credit scores. CredoSDK enabled SalaryDost to quickly get started with AI credit scoring with little hassle and made sure to keep their risks under control so that they can start accepting more customers. When asked Shahi of SalaryDost, he said that the collaboration with CredoLab provided the company with a more predictable source of data to rely on, and a robust AI-based scoring algorithm to achieve the business goals. “With CredoSDK, we can now work towards issuing more loans in a faster and smarter manner while reducing non-performing loans,” said Shahi. “By utilising this scoring mechanism, we can get a better understanding of our customer’s behaviour, and can make more accurate, informed decisions based on that.” The Results Post the deployment, SalaryDost became a true digital fintech platform from lead to disbursement. CredoSDK creates analytical models for SalaryDost and calculated their customers’ credit score and analyses the risks for specific borrowers. The AI-based algorithm also helped the company to have an early fraud detection system by using the device demographic information. Alongside, CredoSDK automated the process of scrutiny and improved the credit underwriting based on a dynamic policy of multiple users, which, in turn, enhanced the process of customer profiling. Artificial intelligence in credit scoring saved not only time for the company but also the total cost of the system. In conclusion, artificial intelligence has significantly revolutionised the lending operations for financial institutions, where they can adequately review the credit history of their consumers in a reasonable time. This comprehensive credit scoring approach is also way more reliable than traditional methods. With fewer risks and enhanced business operations and productivity, the outcome of such credit operations is immense. Thus, the application of AI-based credit scoring system has improved the delivery of financial service.","excerpt":"The customers of today’s’ era couldn’t be happier with the innovations in the fintech sector as getting a loan has never been easier. With technology getting infused in the finance sector, digital lending platforms have highly emerged as the most hassle-free and faster way for people to borrow money. SalaryDost, a consumer lending platform, is […]","categories":["AI Startups"],"tags":["credit score","credit scoring","mba in business analytics"],"author_name":"Sejuti Das","publish_date":"2020-02-21T18:00:00","publication_year":"2020","word_count":998,"keywords":["Go","API","artificial intelligence","AI","ML","Git","mba in business analytics","credit scoring","Aim","ViT","credit score","R","fraud detection"],"extracted_tech_keywords":["AI","artificial intelligence","ML","Aim","fraud detection","R","Go","Git","API","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-startups\/case-study-how-this-mumbai-based-startup-uses-mobile-data-to-assess-credit-scores\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":48648,"title":"How MeitY Startup Summit Sparked New Energy For India Innovators","content":"Ravi Shankar Prasad, Union Minister for Electronics and Information Technology, initiated the MeitY Startup India Summit this week. The event brought together many industry leaders, government officials, entrepreneurs and venture capitalists under one roof to discuss how regulators can boost the startup ecosystem in the country. During the inauguration, Prasad said that the Digital India programme is finding worldwide reverberation. He added that the “digital hunger” across the country should be tapped, and instructed the officials to digitally map entrepreneurs and innovators in rural India. “Innovation ought to be utilised to change rural India and make roads for those in small towns and Tier II and Tier III urban areas,” said Prasad during the event. He likewise propelled a software products library which will go about as a single-window point to examine Indian software product database. The Minister said that Digital India ought to turn into a reference point for inclusion, healthcare training and entrepreneurship. Here are the other key takeaways from Startup India: MeitY Startup Hub (MSH) Talking about how utilising innovations like artificial intelligence would catalyse the financial development of India, Union Minister Prasad unveiled the MeitY Startup Hub (MSH). The hub is viewed as a stage forward towards uniting key partners and new businesses under one rooftop. Further, they will work with specialists to catch feasible procedures to bring the focal point of funding and government around their methods and systems. The platform will use and bring startup mediation program to the software technology parks of India (STPI). They are intended to advance business enterprise and make a supporting system for STPI CoEs in technologies, including digital payments and fintech, blockchain, IoT, artificial intelligence, etc. Startup India: MoUs Signed By Software Technology Parks of India As a nodal office, software technology parks of India (STPI), an association under MeitY, is used for executing the startup hubs. STPI has so far made 28 centres of excellence (COE). During the MeitY Startup Summit 2019, STPI further signed four Memorandum of Understandings (MoUs) with Intel India, IIM Calcutta Innovation Park, Pontaq and IBM. Furthermore, these MoUs are additionally expected to interlink startups with VCs, enterprises and government offices. These MoUs can likewise be viewed as a considerable advance forward for different STPI CoEs like IoT OpenLab at Bengaluru, ESDM incubation unit at Bhubaneshwar, VARCoE at Bhubaneshwar, Autonomous Connected Electric Shared Vehicles, CoE at Pune, MedTech and wellbeing informatics CoE at Lucknow, Rural and Agri IoT CoE at Guwahati, BlockChain CoE at Gurugram and IoT in Agri CoE at Patna-Motihari and others. Startup India: Strengthening Patent Filing and Intellectual Property Addressing the concerns related to the state of intellectual property in India, Prasad announced the patent grant timeline from four years currently to less than a year. This is to help innovators in the fast-moving pace of global innovation. “India should become a big centre for patent and Intellectual Property,” the Minister said. To empower documenting of International Patents, a Scheme Support International Patent Protection in Electronics and IT (SIP-EIT) has been set up. Under the plan, MeitY expects to give money to help support Educational Institutes, MeitY societies, and so forth, for arranging courses and workshops on IPR mindfulness among different partners. Outlook Indicating trust in the capacity of new businesses to bring development and drive business in the Indian economy, PM Modi had said that the thriving Indian startup space would assist India with achieving the $5 trillion focus for the economy. The Startup India plan has assumed a remarkable job in directing the enterprising soul of India’s trailblazers and has infiltrated Tier II and III markets as well. With a lot of activity from the local and state governments, new businesses and beginning time organisations have been given each support and impetus to develop and innovate. At present, India has around 23,974 DPIIT listed companies in the nation.","excerpt":"Ravi Shankar Prasad, Union Minister for Electronics and Information Technology, initiated the MeitY Startup India Summit this week. The event brought together many industry leaders, government officials, entrepreneurs and venture capitalists under one roof to discuss how regulators can boost the startup ecosystem in the country.  During the inauguration, Prasad said that the Digital India […]","categories":["AI Startups"],"tags":["MeitY","ravi shankar prasad","Startups"],"author_name":"Vishal Chawla","publish_date":"2019-10-23T11:41:23","publication_year":"2019","word_count":640,"keywords":["Go","API","ravi shankar prasad","artificial intelligence","Rust","AI","innovation","Git","ViT","MeitY","GAN","Startups","R"],"extracted_tech_keywords":["AI","artificial intelligence","R","Go","Rust","Git","API","GAN","ViT","innovation"],"url":"https:\/\/analyticsindiamag.com\/ai-startups\/meity-startup-summit-innovation-india\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":45596,"title":"8 Ways Google’s Newly Open-Sourced Differential Privacy Library Can Help Developers","content":"In a recent announcement, Google released an open-source version of the differential privacy library which it currently uses to power some of its core products like Google Map. It will let developers and organisations implement features that are otherwise difficult to execute from scratch, hence promising them ease of use and deployment. As the company explains, differential privacy strategically adds random noise to user information stored in databases so that companies can still analyse it without being able to single people out. Open-sourcing can, therefore, help other developers achieve that same level of differential privacy defence. The idea is to make it possible for companies to mine and analyse their database information without invasive identity profiles or tracking. The tech giant is hopeful that it can drastically help mitigate data breach. How It Works Differentially private data analysis enables organisations to learn from their data while simultaneously ensuring that those results do not allow any individual’s data to be distinguished or re-identified. The goal of differential privacy for machine learning is to only “encode general patterns rather than facts about specific training examples.” This allows user data to remain private, while the system overall still learns and can advance from general behaviour. It will not only offer equations and models needed to set boundaries and constraints on identifying data but also include an interface to make it easier for more developers to actually implement the protections. Google has been working on this function along with other privacy settings such as Federated Learning and Google’s Responsible AI Practices. Google currently uses it to protect all different types of information, like location data, generated by its Google Fi mobile customers. The Features It will facilitate users with statistical functions allowing most common data science operations to be supported such as compute counts, averages, medians, percentiles and more It includes an extensible ‘Stochastic Differential Privacy Model Checker library’ apart from an extensive test suite, to help prevent mistakes Query engines are a major analytical tool for data scientists, and one of the most common ways for analysts to write queries is with Structured Query Language (SQL). It has included PostgreSQL extension along with common recipes to help get started. It also makes it ready-to-use Google researchers have also included other functionalities such as additional mechanisms, aggregation functions, or privacy budget management 8 Ways It Can Help Developers It will allow developers to build their own tools with the help of this library. It will also allow them to aggregate data without revealing personally identifiable information either inside or outside their companies. It will allow developers to build tools that analyze personal data without compromising the privacy of the people whose data they are working with. It will bring strong privacy protections in place to make the most of the data and help maintain citizen trust. It will add to the existing privacy offerings by Google such as Tensorflow Privacy and Tensorflow Federated, which it had announced last year. TensorFlow Federated is an open-source framework which implements an approach called Federated Learning and allows experimentation with machine learning and other computations on decentralised data. TensorFlow Privacy, on the other hand, is also an open-source library that allows developers to train machine-learning models with privacy. Its flexibility will ensure that it is applicable to as many database features and products as possible. Differential privacy is usually complicated and is difficult to design one from scratch. Google hopes that its open-source tool will be easy enough to be a one-stop-shop for developers. This system is able to capture most data analysis tasks based on aggregations, performs well for typical use-cases, and provides a mechanism to deduce privacy parameters from accuracy requirements, allowing a principled decision. It can be used to produce aggregate statistics over numeric data sets containing private or sensitive information.","excerpt":"In a recent announcement, Google released an open-source version of the differential privacy library which it currently uses to power some of its core products like Google Map. It will let developers and organisations implement features that are otherwise difficult to execute from scratch, hence promising them ease of use and deployment.  As the company […]","categories":[],"tags":[],"author_name":"Srishti Deoras","publish_date":"2019-09-06T15:29:57","publication_year":"2019","word_count":636,"keywords":["federated learning","data science","PostgreSQL","machine learning","AI","RAG","differential privacy","SQL","TensorFlow","R"],"extracted_tech_keywords":["AI","machine learning","data science","TensorFlow","federated learning","differential privacy","RAG","PostgreSQL","R","SQL"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/8-ways-googles-newly-open-sourced-differential-privacy-library-can-help-developers\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10126336,"title":"10 Best AI Tools for HR Professionals in 2024","content":"In a report on Navigating 2024’s Recruitment Trends, AI and talent expert Hung Lee from Recruiting Brainfood encourages HR leaders to “aggressively apply AI to your own workflow today because it’s going to happen regardless of your desire”. However, Lee cautions against using AI to simply generate more output. He suggests strategically investing the saved time and resources in areas that bring tangible value to the organisation. AI tools seamlessly leverage cutting-edge AI technology to automate repetitive tasks, providing valuable insights, and improving decision-making, allowing HR teams to focus on more strategic initiatives. Here, we’re going to explore the top AI tools for HR professionals. 1. hireEZ On February 16, 2022, hireEZ, formerly known as Hiretual, secured $26M in funding led by Conductive Ventures. The rebranding highlights the company’s focus on outbound recruiting. “In these unprecedented times, recruiting has become an impossible task,” said founder and CEO Steven Jiang. This AI-powered platform can be integrated with any tech stack. It helps recruiters source and pull data from multiple websites and platforms and offers access to over 750 million professional profiles and resumes, and engagement with email templates and insights. 2. ClickUp ClickUp is a productivity platform and solution for HR professionals looking to automate tedious tasks, engage employees, and bring structure to their multiple workflows. The AI tool uses built-in role-specific prompts to analyse data, generate reports, and recommend potential solutions. ClickUp Docs enables HR teams to store and manage all their knowledge at one place with access controls. The platform connects documents to workflows, visualises team tasks, plans events, and offers performance insights on its intuitive dashboard. Zeb Evans and Alex Yurkowski founded it as an internal tool in 2017 before launching it publicly as a cloud-based productivity software. Initially headquartered in San Francisco, the company moved to San Diego in 2020. In 2021, ClickUp raised $400 million at a $4 billion valuation, and expanded to Europe with a Dublin office. It also opened an Asia Pacific headquarters in Sydney. 3. Personio Personio’s AI-driven platform offers comprehensive solutions for managing employee records, benefits, and payroll by centralising HR processes and data. It also increases efficiency and accuracy, enabling HR professionals to focus on strategic initiatives and employee development. It is designed to streamline and automate various HR processes for small and medium-sized enterprises. Valued at $8.5 billion, the company emerged from a program of the CDTM. Its co-founders are Hanno Renner, Roman Schumacher, Arseniy Vershinin, and Jonas Rieke. 4. Clockwise Clockwise is an AI-driven calendar optimisation platform that helps teams reclaim their time and improve productivity. By analysing calendars and scheduling patterns, it automatically schedules meetings to minimise interruptions and maximise focus time, enabling teams to work more efficiently and achieve better work-life balance. By marking certain meetings as flexible, the AI can move them to times that cause the least disruption based on everyone’s preferences and schedules. Matt Martin is the co-founder & CEO of Clockwise. 5. WorkRamp WorkRamp is a comprehensive LMS that enables organisations to create and deliver customised training programs. With features such as interactive content creation, assessments, and analytics, it empowers companies to onboard employees, develop skills, and drive performance improvement effectively. It helps new hires get up to speed with company policies, procedures, and culture by creating and delivering training content for different teams or roles. The world’s first learning cloud platform, it was founded in 2015 and is based in San Carlos, California. Ted Blosser, the company’s CEO and co-founder, leads the team in delivering innovative learning solutions. 6. Leena AI This AI-powered HR chatbot streamlines HR processes, assists employees with common queries, and automates routine tasks. By doing so, it enables HR professionals to focus on strategic initiatives and employee engagement. Its key features include chatbot automation, integration with HR processes, automated task management, and the provision of analytics and insights. Its goal is to enhance employee experience and support HR operations with intelligent automation. The CEO of Leena AI is Adit Jain. He co-founded the company in 2018 and leads its vision to transform HR and employee services through AI-powered solutions. 7. Zoho Recruit Like many other platforms on our list, Zoho uses predictive analysis to fine-tune the hiring process. It offers job posting optimisation and automated resume screening to reduce the manual work during the hiring process. It can automatically parse CVs and extract key information such as skills, experience, and education. The platform’s AI assistant and chatbot, Zia, acts as a matchmaker between you and your desired candidate, improving the overall recruitment experience. Its advanced machine learning algorithms identify hidden patterns within collected data to accurately forecast the best candidate matches in real-time. 8. Eilla AI Eilla AI is an AI platform designed to power self-service options and responsive support. It provides insights and analytics to support strategic HR decisions and works with existing HR systems to streamline processes and data management. Founded by Nikola Lazarov, Nikolay Babulkov and Petar Petrov, it automates HR tasks for businesses of all sizes, from startups to enterprises. This platform is driven by AI technology, which is used to create documents with sensitive data, ensuring protection through masked documents or information replacement for private logs. Its key features are secure financial generative AI with fact-checking and accurate sensitive data protection. 9. Worksheets AI The sheets are a set of structured templates and tools used for various professional purposes, primarily to facilitate the documentation and analysis of different processes or strategies. They play a major role in documenting and analysing business processes to identify inefficiencies and areas for improvement. Used for strategic planning, goal setting, and tracking progress against business objectives, they assist in outlining project scopes, timelines, resource allocation, and risk management. They also help in assigning tasks, setting deadlines, and monitoring project progress. 10. JuiceBox It is a business intelligence software that leverages AI to establish a data-informed decision-making process. The platform aims to enable data visualisation for users who may not be too tech-savvy. With JuiceBox, the HR team can automate the process of gathering data from multiple spreadsheets, such as employee records, performance reviews, and survey results. The platform’s proprietary AI transforms complex datasets into contextual stories and reports. These insights powered by natural language processing can reveal trends and patterns in staff performance, engagement, and satisfaction.","excerpt":"These tools revolutionise HRM by streamlining processes, improving efficiency, and enhancing employee engagement.","categories":["AI Trends"],"tags":["AI (Artificial Intelligence)","AI Tool"],"author_name":"Tarunya S","publish_date":"2024-07-10T12:25:22","publication_year":"2024","word_count":1048,"keywords":["machine learning","TPU","AI","ML","RAG","Aim","analytics","generative AI","edge AI","AI Tool","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","machine learning","ML","analytics","generative AI","Aim","edge AI","RAG","TPU","R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/10-ai-tools-for-hr-professionals\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":42613,"title":"Artificial Intelligence Gets A Boost With The Latest Generation Intel® XEON® Scalable Processors That Drive Inference At Scale","content":"It is an industry-wide phenomenon that has opened the best opportunities for organisations across the board. Artificial Intelligence has become the growth story for India’s digital natives like Flipkart, Swiggy and Ola. Today, many AI applications — facial recognition, product recommendations, virtual assistants have become rooted in our day-to-day lives. However, these emerging AI applications have one common feature ruling them all — dependence on hardware has become the core enabler of innovation. In fact, many rising consumer digital companies are dependent on next-gen architecture that can significantly increase computational efficiency and speed the time-to-market. According to an IDC report, spending on AI systems will more than double to $79.2 billion1 in 2022 with a compound annual growth rate (CAGR) of 38.0% over 2018-2022 forecast period. Hardware spending, dominated by servers, is projected to touch $12.72 billion this year as companies aggressively invest in building up the infrastructure necessary to support AI systems. This has necessitated a shift towards an ‘AI technology stack’ that abstracts away the complexity of the hardware layer related to storage, memory and logic and allows higher performance gains for developers and data scientists. What we are seeing now is a new value-creation in the market by leading semiconductor companies that are focusing on end-to-end solutions for industries. The writing on the wall is clear — the mega breakthroughs in IT aren’t going to come from hardware alone, but from the intersection of AI, hardware and software. AI hardware solutions can only deliver maximum gains if they are compatible with other layers of the software environment. To serve their customers better, semiconductor companies are developing a common programming framework and ecosystem that is in concert with hardware. Top Takeaways 1. No standard AI chip: The AI market is vast, but there is no “one-size-fits-all” approach. Thus, there can be no “standard” AI chip 2. Need to abstract away hardware complexity: Data Scientists and application developers look for high-performance hardware that can churn out general purpose AI solutions within a certain amount of time and power budget. They also demand increased flexibility with hardware that allows them to program with mainstream languages at a higher abstraction level along with libraries. The data science community is looking for a complete solution stack that abstracts away the hardware specifics, allowing them the ease to crunch parallel workloads more efficiently. 3. Shift to inference at scale: Inference at scale marks deep learning’s coming of age. By 2020, the ratio of training deep learning models to inference within enterprises will rapidly shift to 1:53, as compared to the 1:1 split we see today. In fact, Deloitte research predicts that by 2023, 43 percent4 of all AI inference will occur at the edge. Inference is important because it allows enterprises to monetize AI by launching new applications or products by applying their trained models to new datasets. In fact, analysts forecast that inference will be the biggest driver and is projected to generate more revenue in data centers than at the edge. To gain a clear picture, let’s take a look at what AI inferencing involves and why inference is where the real value lies for enterprises. A case in point is speech recognition, wherein audio recordings must be processed using Reinforcement Learning to train speech recognition models. However, once a speech recognition model is trained, it can be used for a wide array of applications such as speech-to-text and voice responses for smart speakers. 4. Rethinking data center AI infrastructure: GPUs widely known for parallel processing power are geared towards training where one inputs the data into the model. Training is also carried out in a centralized location, while inference is pushed out to the edge of the network. This requires enterprises to rethink their infrastructure strategy around training as well inference. 5. Scale-up\/scale-down approach: Organizations are increasingly leaning towards a scale-up\/scale-down approach where it is easy to scale up CPU clusters\/processors to enable efficient power consumption without sacrificing performance, while minimizing the need for major redesign. Intel Drives Innovation with Multi-purpose to Built AI Compute from Device to Cloud In this article, we take a look at Intel’s AI strategy and how the chip giant has built winning roadmap for the fast-growing AI market. Further, it has created a software strategy with tools such as BigDL, nGraph and VNNI which enables developers to make maximum gains from its hardware portfolio. Run the AI You Need on the CPU You Know, With 2nd Gen Intel® Xeon® Scalable Processors Intel has built a new generation of hardware and software that allows enterprises to enter an era of pervasive intelligence and also address specific customer needs. Built with a data-centric focus, 2nd Gen Intel® Xeon® Scalable processors improve performance up to 277X for inference5 compared to the processor’s initial launch in July 2017. With the growing buzz around AI\/ML, 2nd Gen Intel® Xeon® Scalable processors promise an AI acceleration push, coupled with Intel® DL Boost – tailored for deep learning inferencing. With 2nd Gen Intel® Xeon® Scalable processors, Intel® DL Boost provides a winning combination without relying on GPUs.  AI capabilities can be more easily integrated alongside other workloads on multi-purpose 2nd Gen Intel Xeon Scalable processors. Furthermore, VNNI6 can be thought of as an AI inference acceleration are integrated into every 2nd Gen Intel Xeon Scalable processor. Further, performance can significantly improve for both batch inference and real-time inference, because VNNI reduces both the number and complexity of convolution operations required for AI inference, which also reduces the compute power and memory accesses these operations require. Why is Intel® DL Boost pegged as a breakthrough? The answer is straightforward. Most commercial deep learning applications today use 32-bits of floating point precision (fp32) for training and inference workloads. However, both deep learning training and inference can be performed with lower numerical precision, using 16-bit multipliers for training and 8-bit multipliers for inference with minimal to no loss in accuracy. With Intel® DL Boost, Intel has created a new X86 instruction that can perform an integer (INT) 8 (bit precision) matrix multiplication and summation with fewer cycles than before. Intel® DL Boost crunches 3 instructions into 17 and can speed up dense computations characteristic of convolutional neural networks (CNNs) and deep neural networks (DNNs). The main advantage for developers and data scientists is that when it comes to AI inferencing for trained neural networks that don’t require periodic retraining, one no longer needs to rely on special purpose compute hardware like GPUs or TPUs. Here’s a benchmarking report from Dell* that emphasizes how the latest generation processor performs faster in parallel workloads, including inferencing. During benchmark testing, Dell engineers realised more than 3X8 faster inferencing for image recognition with INT8, ResNet50. In Conclusion What’s evident is the shift towards a general-purpose AI stack that enterprises can deploy for Deep Learning. To that end, AI computing companies need to provide a full-stack solution across silicon, tools and libraries for easier application development. Meanwhile, the developer ecosystem demands SDKs and compilers to optimise and accelerate AI algorithms. There’s a need to bring more AI capabilities to enterprises and empower developer ecosystem. We believe Intel is playing a pivotal role in the emerging AI market with its end-to-end solutions. Intel has a wide array of AI hardware that includes CPUs, accelerators\/ purpose-built hardware, FPGAs and in the future neuromorphic chips. Developers look for a software environment that can function across different platforms without them having to overhaul their systems. On the software side, Intel is winning the market with tools like Intel® Distribution of OpenVINO™ Toolkit – that accelerate deep neural network workloads and optimizes deep learning solutions across various hardware platforms. In addition to this, support for popular Deep Learning frameworks like TensorFlow* and MXNet*, machine learning libraries like PyTorch* and existing compilers like ONNX* will help the chip giant win developer mindshare. For more on Intel AI, click here Product & Performance Information 1 Source: IDC,  Worldwide Spending on Artificial Intelligence Systems Will Grow to Nearly $35.8 Billion in 2019, According to New IDC Spending Guide, March 11, 2019. https:\/\/www.idc.com\/getdoc.jsp?containerId=prUS44911419 2 Source:  IDC,  Worldwide Spending on Artificial Intelligence Systems Will Grow to Nearly $35.8 Billion in 2019, According to New IDC Spending Guide, March 11, 2019. https:\/\/www.idc.com\/getdoc.jsp?containerId=prUS44911419 3 Source: As deep learning has been adopted more broadly, there has been a clear shift in the ratio between cycles of training (producing models) and inference (applying models) from 1:1 in early days of DL, to potentially well over 1:5 by 2020. Deep Learning Is Coming Of Age, December 10, 2018. https:\/\/www.intel.ai\/deep-learning-is-coming-of-age\/#gs.jnunhy 4 Source: One research study predicts that 43 percent of all AI inference (or analysis) globally will occur at the edge—meaning outside of data centers, on machines and devices—by 2023, up from just 6 percent last year. Pervasive Intelligence, November, 2018 https:\/\/www2.deloitte.com\/insights\/us\/en\/focus\/signals-for-strategists\/pervasive-intelligence-smart-machines.html 5 Source: Performance results are based on testing as of 06\/15\/2015 (v3 baseline), 05\/29\/2018 (241x) & 6\/07\/2018(277x) and may not reflect all publically available security updates. See configuration disclosure for details. No product can be absolutely secure. Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations, and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http:\/\/www.intel.com\/benchmarks. Intel does not control or audit the design or implementation of third party benchmark data or Web sites referenced in this document. Intel encourages all of its customers to visit the referenced Web sites or others where similar performance benchmark data are reported and confirm whether the referenced benchmark data are accurate and reflect performance of systems available for purchase. Deep Learning Is Coming Of Age, December 10, 2018. https:\/\/www.intel.ai\/deep-learning-is-coming-of-age\/#gs.jnunhy 6 Source: Vector Neural Network Instructions Enable Int8 AI Inference on Intel Architecture, https:\/\/www.intel.ai\/vnni-enables-inference\/#gs.jnzd8e 7 Source: Intel will deliver enhancements to Intel® Deep Learning Boost, beginning with a new set of embedded accelerators called Vector Neural Network Instructions (VNNI), which accomplish in a single instruction what formerly required three, to speed up dense computations characteristic of convolutional neural networks (CNNs) and deep neural networks (DNNs). Monetizing AI: How To Get Ready For ‘Inference At Scale’, April 2019 https:\/\/itpeernetwork.intel.com\/ai-inference-at-scale\/#gs.jo1x8y 8 Source: Even when compared to  Intel® Xeon® Scalable processor, (Code named Skylake), the 2nd Generation Intel® Xeon® Scalable shines — which is something that we at Dell EMC have confirmed in the HPC and AI Innovation Lab. In benchmark testing, our engineers have realized more than 3x faster inferencing for image recognition with INT8, ResNet50. These tests compare the performance of a 2nd Generation Intel® Xeon® Scalable Gold Processor 6248 and an Intel® Xeon® Scalable Gold Processor 6148 (Skylake) on an inference benchmark for image classification, as summarized in this slide. For more complete information visit: https:\/\/www.intel.com\/content\/www\/us\/en\/benchmarks\/benchmark.html. Deep Learning Gets A New Boost With New Intel Processor, April 5, 2019. https:\/\/blog.dellemc.com\/en-us\/deep-learning-gets-boost-new-intel-processor\/ Intel, the Intel are trademarks of Intel Corporation in the U.S. and\/or other countries. *Other names and brands may be claimed as the property of others.","excerpt":"It is an industry-wide phenomenon that has opened the best opportunities for organisations across the board. Artificial Intelligence has become the growth story for India’s digital natives like Flipkart, Swiggy and Ola. Today, many AI applications — facial recognition, product recommendations, virtual assistants have become rooted in our day-to-day lives. However, these emerging AI applications […]","categories":["AI Features"],"tags":["industry-wide analytics","Intel Xeon","is artificial intelligence the next big thing","latest ai products across industries","latest in cloud computing technology","latest machine learning innovation","purpose of artificial intelligence"],"author_name":"Richa Bhatia","publish_date":"2019-08-22T10:00:18","publication_year":"2019","word_count":1872,"keywords":["data science","latest in cloud computing technology","industry-wide analytics","is artificial intelligence the next big thing","artificial intelligence","AI","machine learning","latest machine learning innovation","ML","neural network","Ray","Aim","deep learning","latest ai products across industries","Intel Xeon","purpose of artificial intelligence","TensorFlow"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","neural network","data science","Aim","Ray","TensorFlow"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/artificial-intelligence-gets-a-boost-with-the-latest-generation-intel-xeon-scalable-processors-that-drive-inference-at-scale\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10111697,"title":"PyTorch 2.2 Releases New Features like the FlashAttention-v2 and AOTInductor","content":"PyTorch has released version 2.2, featuring integration of FlashAttention-v2 which processes certain calculations about twice as fast as before. It also has the AOTInductor that allows PyTorch programs to be compiled in advance and used in environments where Python isn’t available. This is important for running AI models efficiently in different settings, like on web servers. In the previous version’s FlashAttention, the main challenge was the quadratic increase in runtime and memory usage as the length of the sequence processed by the AI model increased. FlashAttention-v2 addressed this by optimizing memory usage, reducing it from a quadratic to a linear scale, and achieving a runtime speedup of 2-4 times compared to optimized baselines without any approximation. It reaches 50-73% of the theoretical maximum performance on A100 GPUs, closely approaching the efficiency of optimized matrix multiplication operations. The update also includes enhancements in torch.compile for Optimizers, new inductor optimizations, and the introduction of a logging mechanism named TORCH_LOGS. It is important to note that this will be the last version supporting macOS x64, as macOS x86 support is being phased out. This release comprises 3,628 commits from 521 contributors and has several key improvements and features. These include Device_mesh in torch.distributed, a tool that helps organize and manage how AI models are split and run on different parts of a computer or across multiple computers. TORCH_LOGS Logging Mechanism which is a new way to record what happens when its programs run. This helps developers understand how their AI models are performing and troubleshoot any issues. Further enhancements are made in torch.compile for Optimizers which are a part of PyTorch that helps AI models learn better from data.  TorchInductor which improves how it handles Optimizers and TorchInductor. It relates to how PyTorch combines different parts of a program for efficient processing. Additional features and performance improvements include inductor optimizations, aarch64 optimizations, and support for FlashAttention-2 in torch.nn.functional.scaled_dot_product_attention, which delivers significant speedups on A100 GPUs.","excerpt":"PyTorch 2.2 introduces FlashAttention-v2, doubling AI processing speed, and AOTInductor for compiling in non-Python environments.","categories":["AI News"],"tags":[],"author_name":"K L Krithika","publish_date":"2024-01-31T12:38:29","publication_year":"2024","word_count":322,"keywords":["programming_languages:R","PyTorch","AI","A100","Python","programming_languages:Python","ai_frameworks:PyTorch","GAN","R"],"extracted_tech_keywords":["AI","PyTorch","Python","R","GAN","A100","ai_frameworks:PyTorch","programming_languages:Python","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/pytorch-2-2-releases-new-features-like-the-flashattention-v2-and-aotinductor\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10071988,"title":"Should we call Rust a Failed Programming Language?","content":"Google engineers recently introduced ‘Carbon’, an open source programming language hailed as the possible successor of C++. With the new ‘experimental’ language being the talk of the internet, conversations around why ‘Rust’ didn’t become the official successor of C++ have also surfaced. With its own community of people called ‘Rustaceans’ who use, contribute and are interested in the development of the language, Rust is a statically-typed programming language for performance and safety, especially safe concurrency and memory management. Its syntax is similar to that of C++. The open-source project was originally developed at Mozilla Research. In 2021, the Rust Foundation took the torch and drove the development of the language. In his presentation at CPP North, Google engineer Chandler Carruth advised those using ‘Rust’ to continue using it. Carbon is for developers with large codebases in C++, which are difficult to convert into Rust. Carbon is specifically what Carruth called a ‘successor language’, built atop an already existing ecosystem, C++ in this case. According to a StackOverflow survey, Rust is considered to be one of the fastest-growing programming languages and has been ranked as the most liked language by its users. But, programmers say otherwise. A good language should be safe, fast and easy to program. But is Rust? Rust is difficult. It has a complex syntax and a steep learning curve. It is designed to uniquely solve some very challenging problems in programming. However, as a beginner, using Cuda or MPI on Rust is not very simple compared to the other options like Swift and Go. Moreover, it is slow. Rust is a snail compared to other languages. Even for small projects, the compile times are painfully long, and runtime measurements show that Rust is less efficient than the C programs.Imagine rewriting C libraries that have had decades of scrutiny applied only to introduce new bugs. Bugs in code are programmers’ nightmare. While it does save developers from some mistakes, it does not stop them from unintentionally writing bugs. Another issue are the constant warnings appearing over parentheses, especially over if statements and while loops. Rust is therefore a lot more complicated and inefficient and may soon be superseded by said tooling. Not so unpopular after all? Even though first-hand experiences tell a different story, most big techs are already using Rust, while others plan to do so. Recently, Rust joined Meta’s panel of officially supported server-side programming languages. The list previously included C++, Hack and Python. “There’s a rapidly increasing Rust footprint in our products and services, and we’re committing to Rust long-term and welcome early adopters”, says Eric Garcia, a Meta software engineering manager. Dropbox uses Rust for some mission-critical bits of its programming. Other internet companies would probably choose Rust when they need good security, multi-threading, and to reduce the amount of hardware. For example, a highly efficient web service written in Rust can save millions of dollars in hardware for a company running thousands of servers.Google also plans to use Rust in the Linux kernel after bringing support for the systems programming language Rust to Android. To reiterate, the aim is to reduce security flaws. Meanwhile, Microsoft too has turned to Rust to reduce memory-related bugs in Windows components. Currently, there are 25 repositories on GitHub of the public work Microsoft is doing with Rust. Most of them have occasional commits, which is minuscule compared to 317 C++ repositories. Facebook has also strengthened ties with Rust by joining the Rust Foundation, an organisation established in 2021 to make Rust “a mainstream language of choice for systems programming and beyond”.In 2020, Linux kernel developers proposed rewriting new Linux kernel code in Rust. The idea was to add new code in Rust to the originally written Kernel in C. However, this idea is still in the development stage, described as experimental. In the future, Rust could become a top favourite language for the Internet of Things (IoT) devices with tiny processors and little RAM but need a high level of security. Moreover, considering its relationship with big techs, it would probably become a favourite language for internet companies that need to offer web services to millions of users.","excerpt":"Rust has been ranked as the most liked language by its users for two years in surveys but programmers say otherwise","categories":["AI Trends"],"tags":[],"author_name":"Tasmia Ansari","publish_date":"2022-08-02T10:00:00","publication_year":"2022","word_count":691,"keywords":["CUDA","Go","AWS","AI","Python","Aim","C++","Rust","R"],"extracted_tech_keywords":["AI","Aim","AWS","CUDA","Python","R","Go","Rust","C++","CUDA"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/should-we-call-rust-a-failed-programming-language\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10115459,"title":"Deluxe Media Teams Up with AppTek to Bring AI, ML-Based Solutions for Clients","content":"Leading media and entertainment company Deluxe has teamed up with AppTek, a company specialising in AI and ML solutions, to revamp content localisation and workflow efficiency. Through this partnership, Deluxe has acquired a non-controlling interest in AppTek and secured exclusive rights to resell AppTek’s suite of products and services to the global media and entertainment sector. The collaboration enables Deluxe to offer studios and content distributors a diverse range of integrated AI solutions tailored to their specific needs. “We’ve been working closely with AppTek for several years and have been impressed by their technology and breadth of experience,” said Chris Reynolds, Deluxe’s executive vice president and general manager of localisation and fulfilment. “AppTek has long believed in pushing the boundaries of artificial intelligence and its real-world applications. Deluxe brings the same passion in content transformation and accessibility,” said AppTek CEO Mudar Yaghi. Over the last year, we have seen media houses slowly entering the AI space. This became more relevant when Universal Music Group partnered with Google’s YouTube to create “Music AI Incubator.” Additionally,  during the writer’s strike, Netflix hired AI product managers with lucrative salary packages of $900,000. Recently, OpenAI introduced the video generation model Sora, capable of creating highly realistic videos based on prompts, a development many believe could benefit filmmaking.","excerpt":"Through this partnership, Deluxe has acquired a non-controlling interest in AppTek and secured exclusive rights to resell the company’s suite of products and services to the global media and entertainment sector.","categories":["AI News"],"tags":[],"author_name":"Shritama Saha","publish_date":"2024-03-13T11:22:51","publication_year":"2024","word_count":213,"keywords":["Go","artificial intelligence","OpenAI","AI","programming_languages:R","ML","programming_languages:Go","R"],"extracted_tech_keywords":["AI","artificial intelligence","ML","OpenAI","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/deluxe-media-teams-up-with-apptek-to-bring-ai-ml-based-solutions-for-clients\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10114634,"title":"LTIMindtree Joins IBM Network to Advance the Quantum Innovation Ecosystem","content":"IBM announced that LTIMIndtree, a global technology consulting and digital solutions company, has joined the IBM Quantum Network to explore quantum computing innovation for the benefit of its global clientele across multiple industries. LTIMindtree is the first Indian Global System Integrator (GSI) to join the IBM Quantum Network. As part of the IBM Quantum Network, LTIMindtree expands on its platinum partner status with IBM and joins a global community of Fortune 500 companies, top universities, research labs, and startups. LTIMindtree will have access to IBM resources, including IBM’s global fleet of quantum computing systems over the cloud, software, and associated expertise. This move is a strategic step toward LTIMindtree helping their customers benefit from the transformative value of quantum computing technologies. LTIMindtree will also collaborate with the Indian Institute of Technology (IIT) Madras, which is also an IBM Quantum Innovation Center, on joint quantum research and workforce development. “These collaborations are more than an innovation milestone; they’re an important step towards a future where quantum computing could help solve more complex problems faster and more efficiently. “Importantly, it positions us to expedite our customers’ journey towards realizing the immense value of quantum computing, readying them to leverage these advanced technologies for transformative solutions,” said Aan Chauhan, Chief Technology Officer, LTIMindtree. LTIMindtree’s plans across these collaborations are to establish a series of long-term projects, including applied research toward business and societal problems, quantum computing workshops, and research grants. These initiatives aim to nurture a new generation of quantum computing professionals and researchers at LTIMindtree, creating a sustainable and innovative ecosystem.","excerpt":"LTIMindtree will have access to IBM resources, including IBM’s global fleet of quantum computing systems over the cloud, software, and associated expertise","categories":["AI News"],"tags":["IBM","LTIMindTree","Quantum Computing"],"author_name":"Pritam Bordoloi","publish_date":"2024-02-28T14:52:57","publication_year":"2024","word_count":259,"keywords":["Quantum Computing","programming_languages:R","AI","innovation","Git","RAG","Aim","IBM","LTIMindTree","R","emerging_tech:quantum computing","startup"],"extracted_tech_keywords":["AI","Aim","RAG","R","Git","innovation","startup","programming_languages:R","emerging_tech:quantum computing"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/ltimindtree-joins-ibm-network-to-advance-the-quantum-innovation-ecosystem\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10172604,"title":"How MongoDB is Winning Developers","content":"With several hundred tools and technologies available, developers want the experience to be as streamlined as possible. Speed, feedback, and fewer tabs—that’s what modern developers really want. At the MongoDB.local event in Bengaluru, the company pitched a vision that meets all three: no more hopping between terminals, GUIs, and browser tabs. Instead, MongoDB is embedding itself right into developer workflows with local-first CLI tools, native IDE integrations, and AI copilots that know their schemas. MongoDB aims to own the development loop, from test to deployment, and from local development to production-scale AI. It is bundling database power with developer familiarity, resulting in a smoother, faster path to building intelligent apps. With all the developments the company has mentioned, MongoDB is trying to evolve into much more than a traditional NoSQL database. Atlas on Your Laptop At the heart of this revamp is the MongoDB Atlas CLI, a command-line interface for the MongoDB Atlas platform, which now provides a comprehensive developer stack for local machines. Developers can run not only the core database, but also full-text search and AI workloads without requiring cloud access. Boris Bialek, VP and global field CTO of MongoDB, said on stage, “It’s basically Atlas on your laptop,” highlighting how a single command can spin up a complete environment with everything from time-series queries to AI-powered apps. Developers can now mirror production environments on local machines, iterate faster, and test features like vector search or semantic recall offline. As Bialek put it, “It’s the perfect build-test-develop imagination system you can have.” Coding Without Context Switching MongoDB is also embedding itself deeper into popular IDEs. “You don’t want to jump the whole time between doing something on the database and doing something on the app. This is one flow by now,” Bialek explained, reinforcing MongoDB’s move to simplify full-stack development. “We integrated MongoDB into VS Code with GitHub Copilot. So, Copilot is not only supporting MongoDB, but it’s also optimised [for it],” he highlighted. The integration enables the Copilot to suggest queries, auto-generate parameters, and interact with collections in natural language. JetBrains users aren’t left out either. IntelliJ now supports native MongoDB integrations, with autocomplete, validation, and even performance tips—all connected to JetBrains’ AI assistant. He mentioned that the IntelliJ plugin is now in public preview, allowing developers to try it out. This developer-friendly direction is mirrored by MongoDB’s partner, Microsoft. Azim Uddin, principal cloud solutions architect at Microsoft, told AIM, “MongoDB is one of Microsoft’s prioritised partners…There is a MongoDB extension for Visual Studio.” Uddin highlighted that the MongoDB extension for VS Code provides developers with an experience similar to that offered by SQL Server Management Studio. GitHub Copilot also comes into play. “MongoDB has a GitHub Copilot extension,” Uddin confirmed. “People can use MongoDB GitHub Copilot for enhancing productivity with their Copilot.” While he admitted that no-code-style MongoDB integrations aren’t quite there yet, he noted that “it’s a work-in-progress”, and may evolve as both companies expand their developer tooling. AI-Aware Copilots and Context-Rich Coding Underpinning these enhancements is the Model Context Protocol (MCP) support—a new bridge between code, AI, and databases. MCP enables LLMs to query databases intelligently by understanding the schema and query context in real-time. “Very modern IDEs [like] Windsurf, Cursor. When you’re working with those, you can now use MCP to connect to your MongoDB environment,” Bialek said. “You can compose excellent queries. You can test things around. And it’s absolutely seamless.” He continued that users can pose natural language questions such as “Show me the schema of my user collection” or “Who are the most active users in the system?” and even request a query to be built for specific information. This context-rich approach removes guesswork from AI. It allows developers to issue live queries, get instant previews, and refine results within their IDEs, not in some sandboxed playground. Uddin, reflecting on hands-on experiments with MongoDB and Azure AI Foundry, said he “definitely sees productivity gains and benefits out of these integrations”. Even beyond developer tooling, Microsoft is embedding MongoDB deeper into its broader platform. “We are integrating across different stacks,” Uddin said, citing MongoDB’s interoperability with Microsoft Fabric for analytics and enterprise security frameworks. “It is both ways,” he added. “Customers can now bring MongoDB data to Fabric and do analytics on their operational data.” MongoDB has always been developer-friendly, but now it’s becoming developer-native. With a full local CLI stack, tight IDE integrations, and context-aware copilots, the company is positioning itself not just as a database but as a companion to the entire coding journey. In an era of impatient product cycles and AI-infused apps, MongoDB’s aim to offer a seamless experience, coupled with partnerships with major tech companies like Microsoft, appears to be a good bet.","excerpt":"Developers can efficiently work with MongoDB when using modern IDEs and AI coding tools.","categories":["Global Tech"],"tags":["MongoDB"],"author_name":"Ankush Das","publish_date":"2025-06-30T18:03:14","publication_year":"2025","word_count":787,"keywords":["Go","AI","MongoDB","ML","R","Aim","analytics","SQL","copilots","Azure"],"extracted_tech_keywords":["AI","ML","analytics","Aim","copilots","Azure","MongoDB","R","SQL","Go"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/how-mongodb-is-winning-developers\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":8591,"title":"Delivering Great Data for Next Generation Analytics","content":"Businesses are abuzz with the promise of advanced analytics and its effectiveness today. Tied to a business strategy, a well-planned analytics initiative can make a big difference to an organization. The key drivers for business analytics are extremely compelling and it is critical for strategic decision-making. It can accurately provide financial and operational forecasts, optimize business operations and streamline processes, and has the ability to predict new opportunities. According to an analysis published by Harvard Business Review, companies that relied on data-driven decision-making were on an average, 5% more productive and 6% more profitable than their competitors. A TDWI Best Practices Survey Report suggested that more than 50% of respondents felt that next-generation analytics was extremely important to strategic decision-making, with just fewer than 50% saying it was extremely important to improve business processes and performance. In fact, advanced analytics is poised to double in use over the next three years. However, the next generation of analytics are now being defined by factors which are fairly new and results driven. The reality is that answers take far too long to come or far too much cost, and also lack consistency. The ability to discover and access the right data at the right quality in the right timeframe is frequently a bottleneck. Often, the data sets are by themselves inaccurate or incomplete, and, as a result, confidence in the analysis is low. With the changing scenario, even after significant investments in next-generation analytic tools, teams, software, and warehouses, a large proportion of analytics projects usually underwhelm, disappoint, or outright fail. So, if next-generation analytics are not living up to expectations – and the applications and the people working on them are not the problem – then what is? Data quality problems and data integration challenges are the two most common barriers to analytics success. If the data going into the analytics stack is dirty, noisy, duplicated, incomplete, poorly integrated, or delivered too late, the insights coming out of it cannot be trusted. Great data should be accurate, de-duplicated and complete. It is also easily scalable and avoids reinventing the wheel over and over again. With the rapidly changing technology for analytics data storage, it is rewarding to build a data management architecture that works across any type of data and any storage technology. So an effective data governance strategy is a sure way to oversee data quality, manage security and compliance, or standardize processes. Concerted efforts should go into automating processes so that decision makers can get the insights they need in time to take action. Fundamental to this process is the need to define a data management architecture for all of your analytics projects and initiatives. In this way, you will be able to drive standardization, automation, and productivity of data delivery in support of your business initiatives. This architecture needs to work across any data source, any analytics use case, any analytics tool. It is also important to make data management work across new, unstructured data types such as Big Data. Organizations are increasingly finding that the most interesting and useful business insights come from combining data from internal sources with data from external, less structured sources. At the same time, by focusing only on IT, organizations can be at risk of thinking purely in technology terms, when they should be concentrating on the business outcomes to be delivered. It is important to bring the business team into the process of functionally defining the new data management architecture and to ensure that business analysts are getting the data they need with the quality and speed they require in order to be able to support these business initiatives. The value of analytics solutions does not just lie in the ability to deduce insights from large amounts of data. It lies in their ability to improve critical business outcomes. The ability to deliver great data, coupled with smart data management, is the key to delivering next-generation analytics success.","excerpt":"Businesses are abuzz with the promise of advanced analytics and its effectiveness today. Tied to a business strategy, a well-planned analytics initiative can make a big difference to an organization. The key drivers for business analytics are extremely compelling and it is critical for strategic decision-making. It can accurately provide financial and operational forecasts, optimize […]","categories":["IT Services"],"tags":["Analytics Case Study"],"author_name":"Roger Nolan","publish_date":"2016-01-07T12:07:43","publication_year":"2016","word_count":657,"keywords":["big data","Go","API","AI","ML","Scala","RAG","analytics","Rust","Analytics Case Study","R"],"extracted_tech_keywords":["AI","ML","analytics","RAG","R","Go","Rust","Scala","API","big data"],"url":"https:\/\/analyticsindiamag.com\/it-services\/delivering-great-data-for-next-generation-analytics\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":21952,"title":"Accenture Launches New Artificial Intelligence Testing Services","content":"File photo of Bhaskar Ghosh, group chief executive of Accenture Technology Services. Global IT services and consulting company Accenture announced on Tuesday the launch of new services for testing artificial intelligence systems. The company said that the system was powered by a “teach and test” method designed to help “build, monitor and measure” reliable AI systems within their own infrastructure or in the cloud. Bhaskar Ghosh, group chief executive of Accenture Technology Services said in a statement, “The adoption of AI is accelerating as businesses see its transformational value to power new innovations and growth. As organisations embrace AI, it is critical to find better ways to train and sustain these systems — securely and with quality — to avoid adverse effects on business performance, brand reputation, compliance and humans.” Accenture’s ‘teach and test’ method ensures that AI systems are producing the right decisions in two phases. The ‘teach’ phase focuses on the choice of data, models and algorithms that are used to train machine learning. This phase experiments and statistically evaluates different models to select the best performing model to be deployed into production, while avoiding gender, ethnic and other biases, as well as ethical and compliance risks. During the ‘test’ phase, AI system outputs are compared to key performance indicators, and assessed for whether the system can explain how a decision or outcome was determined. It uses innovative techniques and cloud-based tools to monitor the system on an ongoing basis for sustained performance. For instance, a patent-pending normalisation technique uses a unique algorithm to test object recognition more quickly. According the Accenture Technology Vision 2018, AI systems require addressing many of the same challenges faced in human education and growth: fostering an understanding of right and wrong, and what it means to behave responsibly; imparting knowledge without bias; and building self-reliance while emphasising the importance of collaborating and communicating with others. Kishore Durg, senior managing director, Growth and Strategy and Global Testing Services Lead for Accenture, said, “Testing AI systems presents a completely new set of challenges. While traditional application testing is deterministic, with a finite number of scenarios that can be defined in advance, AI systems require a limitless approach to testing… There is also a need for new capabilities for evaluating data and learning models, choosing algorithms, and monitoring for bias and ethical and regulatory compliance. Accenture’s ‘teach and test’ method takes all of this into consideration to help companies develop and validate AI systems with confidence.” Accenture used the ‘teach and test’ method to train a conversational virtual agent for a financial services company’s website, so that it could engage in accurate, unbiased conversations and know when to refer conversations to a human. The agent was trained 80 percent faster than previously possible and achieved an 85 percent accuracy rate on customer recommendations. This method was also used to teach a sentiment analysis solution to evaluate a brand’s service performance by analysing social media, news and other sources in real time. Having the right training data enabled the solution to correctly interpret sentiments in different contexts and domains. The training time for models was cut in half, enabling faster analysis and business results. Accenture AI Testing is part of a complete range of testing services that helps quality engineering professionals be a catalyst for speed, agility and business performance while achieving radical productivity. Accenture serves over 1,000 testing clients across more than 40 industries and is recognised as an industry leader in testing innovation.","excerpt":"Global IT services and consulting company Accenture announced on Tuesday the launch of new services for testing artificial intelligence systems. The company said that the system was powered by a “teach and test” method designed to help “build, monitor and measure” reliable AI systems within their own infrastructure or in the cloud. Bhaskar Ghosh, group […]","categories":["AI News"],"tags":["Accenture","AI (Artificial Intelligence)"],"author_name":"Prajakta Hebbar","publish_date":"2018-02-21T10:43:27","publication_year":"2018","word_count":579,"keywords":["Accenture","Go","machine learning","artificial intelligence","TPU","AI","sentiment analysis","innovation","ViT","GAN","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","sentiment analysis","TPU","R","Go","GAN","ViT","innovation"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/accenture-new-ai-testing-service\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":48635,"title":"Why IBM Disagrees with Google’s Quantum Supremacy Claims","content":"It has been less than a month since the news of Google achieving quantum supremacy has been leaked. But earlier this week, tech giant IBM released a blog post detailing that they did not think Google had met the goals to claim quantum supremacy. The leaked paper claimed that Google’s quantum processor took only 200 seconds to perform a task, which would have taken 20,000 years for a present-day supercomputer. The performance of quantum computers is measured in qubit. For every qubit, say x, there are 2^x states. Two qubits can store four states, three qubits can store eight states. So, 50 qubits in quantum computers can be approximated to 1 quadrillion bits in classical computers. However, IBM disagrees with these claims. “The goal has not been met,” stated IBM in their post. What Was Claimed Initially? In the paper that was later taken down, Google researchers had said that, to their knowledge, the experiment “marks the first computation that can only be performed on a quantum processor.” To put it in another perspective, by running a huge cluster of classical cores for (say) a month, you can eventually verify the outputs that your QC produced in a few seconds — while also seeing that the QC was many orders of magnitude faster. While debunking the claims of quantum supremacy, the researchers at IBM also applaud Google’s effort in developing a 53 qubit computer. However, their affair turns sour with the various versions of quantum supremacy doing the rounds. Their point in writing out a post addressing this fiasco is to warn against the consequences of such hype. Quantum Leap Or Quantum Flop? via IBM In their blog, IBM has written that whatever claims of supremacy that Google had in their leaked paper, were just “overblown numbers”. The researchers at IBM argue that an ideal simulation of the same task can be performed on a classical system in 2.5 days and with far greater fidelity. To verify the claims, IBM ran an analysis of expected classical computing runtime vs circuit depth of “Google Sycamore Circuits”. As can be seen in the plot above, the bottom (blue) line estimates the classical runtime for a 53-qubit processor (2.5 days for a circuit depth 20), and the upper line (orange) does so for a 54-qubit processor. Quantum computers get their much-hyped edge over classical computer owing to quantum entanglement and superposition phenomena. However, classical computers too can be enhanced with high precision computations and other software assets. While comparing the performance of a quantum computer to a traditional computer, one should also consider all the available additional resources. They even go on to say that at the fundamental level, quantum computers will never overtake their classical counterparts. Instead, they believe that these two kinds of computers will be deployed together to produce unique results. Giant Hype For Quantum Leaps From drug discovery to solving mathematical problems, from machine learning to material sciences, quantum computers have the potential to revolutionise many domains. And, this advancement wouldn’t be done single-handedly. The uniqueness of a quantum computer would instead be exploited to improve the results obtained with a traditional computer. However, the hype around quantum computers seems to have reached a point where it is hard to separate facts from fiction. If we consider the above case alone, it is surprising how two pioneers in the same field can differ in such gigantic way. “It exacerbates the already overhyped reporting on the status of quantum technology,” said John Preskill, a theoretical physicist, talking about Quantum supremacy. That said, the researchers still insist that Quantum computers are starting to approach the limit of classical simulation and it is important that we continue to benchmark progress and to ask how difficult they are to simulate.","excerpt":"It has been less than a month since the news of Google achieving quantum supremacy has been leaked. But earlier this week, tech giant IBM released a blog post detailing that they did not think Google had met the goals to claim quantum supremacy. The leaked paper claimed that Google’s quantum processor took only 200 […]","categories":["Global Tech"],"tags":["Google","IBM","Quantum Computer","quantum supremacy"],"author_name":"Ram Sagar","publish_date":"2019-10-22T17:14:38","publication_year":"2019","word_count":627,"keywords":["Go","quantum supremacy","machine learning","TPU","programming_languages:R","AI","RPA","programming_languages:Go","Quantum Computer","Aim","IBM","Google","GAN","R"],"extracted_tech_keywords":["AI","machine learning","Aim","TPU","R","Go","GAN","RPA","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/why-ibm-disagrees-with-googles-quantum-supremacy-claims\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10068858,"title":"The rise and rise of data centres in India","content":"The need for data centres shot up exponentially in proportion to the rise in data consumption and cloud adoption. According to a JLL study, the Indian data centre sector would require a total investment of USD 3.7 billion within the next three years to meet the six million square feet of development. Despite the heavy capital involved, Indian corporations are bullish on data centres. Last week, the Adani group committed INR 8,000 crore to build a state-of-the-art data centre at Rajarhat in West Bengal. To be completed over three years, the facility will be built on an allotted land mass of 51.75-acres. The plot area exceeds the data centre facility under construction by Reliance as a part of the Bengal Silicon Valley project in New Town in Kolkata. Growth in Indian data center market size over the years, Source: Menafn Big and small investments In 2019, a lot of Indian firms turned into data centre businesses. In July of the same year, Adani group announced an investment of up to INR 70,000 crore to build solar-powered data parks in Andhra Pradesh. As cloud adoption picked up pace in India, global tech giants started partnering with Indian conglomerates to cash in on the opportunity. In 2019, Oracle announced Gen 2 cloud regions in Mumbai and Hyderabad. Reliance Industries tied up with Microsoft to provide cloud services to small and medium enterprises in the country. Hiranandani Group launched Yotta Infrastructure to build data centres in Mumbai and Chennai. The real estate company’s plan was to invest INR 14,000 crore over a period of five to seven years. In the end of 2021, the group said it would double its investment in Yotta. The EdgeConneX and Adani Group joint venture, @adaniconnex, is building a 1 GW platform of sustainable data center capacity in India by 2030. Learn how: https:\/\/t.co\/1uAYPFPeKY#sustainability #greenerdata pic.twitter.com\/WeQolU8BvV— EdgeConneX (@EdgeConneX) May 13, 2022 The Gautam Adani-led group has set up a 50-50 joint venture with US-based EdgeConneX with data centres planned in Chennai, Mumbai, Noida, Vizag and Hyderabad. The group’s 409 acre facility near Vizag has been approved. Last November, the company announced an investment of USD 70 billion by 2030 in renewable energy, and plans to power its data centres using green energy. The Rajarhat facility will also reportedly be powered by renewable sources. Source: India Briefing Government support In June, rating agency CRISIL said India’s data centre capacity is projected to double to reach 1,700–1,800 MW by 2025. The country’s current capacity stands at approximately 870 MW. In a statement, director of CRISIL, Nitesh Jain said India is expected to add between 890 to 900 MW in capacity between the fiscal year 2023 and 2025. The Indian government is also doubling down on digital initiatives to create conducive conditions for data centres. The big push towards data localisation would ensure 75 percent of the data remains within the country. Sensing the opportunities in the sector, the Ministry of Electronics and Information Technology or MEITY drafted a data centre policy in 2020 giving data centres an ‘infrastructure status’ on par with roads, railways and power. The new policy aims to ease the clearance process for data centres. Currently, as many as 40 clearances are needed for setting up a data centre in the country. MEITY special secretary and financial advisor Jyoti Arora said cloud data centres would be declared as special economic zones or SEZs. The government would give land, water and electricity to facilitate ease of doing business. States like Maharashtra, Telangana, Karnataka and Uttar Pradesh also offer exemptions on stamp and electricity duty, power subsidies and concessions on land prices for data centres. “India has witnessed a massive adoption of digital transactions, IoT and smart devices since the pandemic. The nation’s youth being tech-savvy, the usage of digital devices is only expected to surge, creating a large volume of data, which will drive the need for data centres. Add to that, the Indian government has proposed to classify the data centre sector in the category of critical infrastructure such as railways, power and roads to protect its consumers data. It is also providing incentives for the development of data centers that is attracting many businesses to capitalise on the opportunity,” said Keshav Kumar, general manager of Presales at Rahi Systems.","excerpt":"The Adani group has committed INR 8,000 crore to build a state-of-the-art data centre at Rajarhat in West Bengal.","categories":["IT Services"],"tags":[],"author_name":"Poulomi Chatterjee","publish_date":"2022-06-13T12:00:00","publication_year":"2022","word_count":713,"keywords":["Go","API","programming_languages:R","AI","programming_languages:Go","Git","Aim","GAN","R"],"extracted_tech_keywords":["AI","Aim","R","Go","Git","API","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/the-rise-and-rise-of-data-centres-in-india\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10163241,"title":"‘It was Uncomfortable to Watch My Kids Learn Math on Smartphones, but They Turned Out Fine,’ says Google CEO","content":"At the AI Action Summit 2025 hosted in Paris, France, Google CEO Sundar Pichai spoke about how AI can influence various sectors such as education, healthcare, science, etc. and why the company invests as much in the industry. While Google has taken multiple opportunities to showcase how the company plans to use AI to benefit humanity, Pichai’s recent talk included several personal anecdotes as he urged the world to embrace AI and address issues related to public policy. He took the stage to help ease the scepticism associated with AI and spoke about how policies must be framed with a balanced approach. “I grew up doing math using logarithmic tables, and I was uncomfortable watching my kids learn math with smartphones. They’ve turned out just fine,” he added. ‘We Mustn’t Let Our Bias for Present Get in Way of Future’ Pichai began his talk by sharing his experiences while growing up in Chennai. He shared how he had to wait five years before his family acquired a rotary phone, which changed their lives. Moreover, he revealed once he had to take a “four-hour round trip” to get blood test results for his mother, but, now, all he needed to do was just pick up the phone. Citing similar examples, Pichai highlighted the power of transformative technologies. “We’re still in the early days, yet I already believe AI will be the most profound shift of our lifetimes,” he said. Pichai added that to unlock the full potential of AI, an ecosystem of innovators must be enabled, and a powerful infrastructure must be built. “We’re excited for the path [US] President [Donald] Trump, [French] President [Emmanuel] Macron and other countries are forging here,” he said. Pichai highlighted the need for industries to invest in people and prepare them for the workforce ahead. He believes people will have to act boldly to advance the applications of AI responsibly. Along these lines, Pichai outlined a framework for a successful AI public policy. This policy must address the risks associated with the technology without affecting progress. He also suggested that policies must fill gaps and draw on existing laws instead of creating new ones. However, the AI race is not limited to the United States. China made a bold entry recently, France is showing promise with Mistral, and India is actively working on a foundational model. Hence, Pichai suggested that policies must be aligned across countries. “AI can’t flourish if there is a fragmented regulatory environment, with different rules across different countries and regions,” he said. According to him, AI is a once-in-a-generation opportunity to improve lives. “We must not let our own bias for the present get in the way of the future.” Taking the stage at the summit, Indian Prime Minister Narendra Modi, said that AI is writing the code for humanity. PM Modi said AI is developing at an unprecedented scale and speed, and being adapted and deployed even faster. He stressed the need for collective global efforts to establish governance and standards. “Governance is not just about managing risk and rivalries, it is also about promoting innovation and deploying it for the global good. So we must think deeply and discuss openly about innovation and governance,” he said, echoing Pichai’s sentiment. France & EU to Loosen Regulations Aligning with Pichai’s expectations, Macron said that Europe is planning to reduce AI regulations. As per Reuters, European Union’s digital chief Henna Virkkunen promised to simplify rules and implement them in a business-friendly way. With regards to AI infrastructure in France, Macron announced plans for €109 billion of private AI investment, analogous to The Stargate Project in the US. The recent few events have been a ray of hope for Europe and its relatively restricted tech ecosystem. Sam Altman, CEO of OpenAI, announced that the company is opening a new office in Munich, Germany. This is in addition to their previous offices in Dublin, London, Paris and Brussels. Furthermore, Mistral announced plans to invest “several billion euros” in building its first data centre in France. The AI startup also recently announced a new application called Le Chat, which it claimed is the fastest AI assistant capable of producing over 1,000 tokens per second. The chatbot is powered by Cerebras Inference, which is also touted as the world’s fastest AI inference engine.","excerpt":"Sundar Pichai urged people to embrace AI and move beyond the scepticism.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","Google"],"author_name":"Supreeth Koundinya","publish_date":"2025-02-11T17:53:09","publication_year":"2025","word_count":717,"keywords":["Go","OpenAI","AI","AWS","Git","RAG","Ray","Aim","Google","GAN","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","OpenAI","Aim","Ray","RAG","AWS","R","Go","Git","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/it-was-uncomfortable-to-watch-my-kids-learn-math-on-smartphones-but-they-turned-out-fine-says-google-ceo\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10096358,"title":"Wolfram’s New Update Gives Developers Genius-level Generative AI","content":"After being one of the first plugins to ever come to ChatGPT, Wolfram has now gone all in on the LLM wave. In the latest version 13.3 update, the Wolfram language has added support for LLM technology, as well as integrating an AI model into the Wolfram Cloud. This update comes on the heels of Wolfram slowly building the tooling for making the language LLM-ready. The update puts LLMs directly into the language with the introduction of an LLM subsystem for the language. It also builds on the LLM functions technology added in May, which ‘packages’ AI powers into a callable function, with the new subsystem now being user-addressable. With these new updates, developers have a whole new way of interfacing with their data. This approach combines Stephen Wolfram’s idea of natural language programming along with the Wolfram language’s symbolic programming, creating a force to be reckoned with. What’s more, with the Wolfram language API, this can be plugged in to larger systems, delivering amazing power through a natural language interface. LLM-powered computation The Wolfram language’s strengths lie in its symbolic programming capabilities. Using code, the language can perform many complex mathematical functions, such as algebra, matrix manipulations, and differential equations. To bolster the language’s logical problem-solving capabilities, Stephen Wolfram, the creator of the language, decided to add in LLM capabilities to the bot. Wolfram’s tryst with LLMs began with the creation of the Wolfram ChatGPT plugin, which empowered the chatbot with the symbolic programming capabilities of the language. Stephen then embarked upon a journey to take symbolic language to new heights, saying to AIM, “We’re about to use the symbolic language [Wolfram] to provide a way of using LLM as a component in a larger software stack. It’s something that you can do in a very beautiful way.” The 13.3 update seems to be a step towards this direction, bringing LLMs directly into Wolfram language through a LLM subsystem. After the chatbot plugin and LLM function calling, Stephen introduced Chat Notebooks. This allows users to easily interact with LLMs in the Wolfram Notebook through a text box, allowing them to generate powerful code in the Wolfram language. Stephen called this interface an example of “using an LLM as a linguistic interface with common sense”, as it allows the users to interact with the language without needing to know the syntax. Stephen thinks this is a natural step to the existing capabilities of the program, stating, “When you’re doing something you’re familiar with, it’ll almost always be faster and better to think directly in Wolfram Language, and just enter the computational language code you want. But if you’re exploring something new, or just getting started on something, the LLM is likely to be a really valuable way to “get you to first code.” The in-built LLM has self-correcting capabilities as well, allowing it to fix its own errors before running it and outputting code snippets. This means that it can also debug existing code in Wolfram language, and can even look at details like stack trace and error documentation to fix broken code. It also comes with a few different personas, each geared towards a different purpose. The code assistant persona writes codes and explains it, the code writer persona only generates the code, while others like Wolfie and Birdnado respond to the user “with an attitude”. To extend the functionality of the LLM even further, Wolfram also launched the prompt repository, which can be used to get additional function prompts and modifier prompts. While he stated that Wolfram language will continue to become “increasingly integrated as a tool into LLMs’, the prompt repository currently showcases the capabilities of the language’s new AI tools. A developer’s playground The prompt repository is a community-contributed prompts platform that will allow the LLM in Wolfram language to adopt many different personas, each with discrete use-cases. In addition to this, the community has also contributed various functions that can extend the functionality of the language beyond its already-comprehensive list of inbuilt functions. The prompts in the repository are split into three main categories, namely personas, functions, and modifiers. Personas define the style of interaction with the user, functions generate output from existing text, and modifiers apply an effect to the output coming from the LLMs. Each of these functions can also be called in code, allowing developers to integrate them easily into existing code or even extend the functionality of a program. The repository serves a very important purpose is speeding up workflows by allowing developers to avoid LLM wrangling. In Wolfram’s words, “Sometimes just using the [prompt’s] description will indeed work fine. But often it won’t. Sometimes that’s because one needs to clarify further what one wants. And sometimes there’s just a certain amount of “LLM wrangling” to be done. And this all adds up to the need to do at least some “prompt engineering” on almost any prompt.” The repository effectively removes the need for this prompt engineering by creating a database of accessible and callable prompts, which are then converted into Wolfram Language code by the LLM. These prompts use the language to carry out a function on the given text,  extending its functionality beyond mathematical problems. Some of the sample prompts that stood out to us were the LongerRephrase prompt, which expands a given statement, Scientific Jargonize, which makes a plain text sentence sound like it has come out of a research paper, and TweetConvert, which converts data into a tweet. There are also a host of other prompts which can convert sentences into product pitches, dejargonize complex pieces of text, check grammar stringently, and even generate puns. Using the ever-growing repository of prompts, devs and citizen developers alike can use the Wolfram language to easily modify large pieces of text. What’s more, since each prompt can be called as a function, they can be added on to any program to give it LLM super powers. Once the added LLM functionality goes live in the coming days, the Wolfram language will become an indispensable tool in the belt of AI enthusiasts and developers alike.","excerpt":"After being one of the first plugins to ever come to ChatGPT, Wolfram has now gone all in on the LLM wave","categories":["AI Features"],"tags":[],"author_name":"Anirudh VK","publish_date":"2023-07-05T14:00:00","publication_year":"2023","word_count":1014,"keywords":["Go","ChatGPT","API","TPU","ELT","AI","GPT","prompt engineering","Aim","R"],"extracted_tech_keywords":["AI","ChatGPT","Aim","prompt engineering","TPU","R","Go","API","ELT","GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/wolframs-new-update-gives-developers-genius-level-generative-ai\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10168498,"title":"These Drones are Flying at 18,000 Ft, Above the Himalayas for the Indian Army","content":"Drone technology in India has shifted rapidly from novelty to necessity. Once used primarily for surveillance and cinematography, drones are now being integrated into critical operations from high-altitude defence logistics to internal inspections at industrial facilities and large-scale agricultural spraying. While government schemes are promoting adoption among rural communities, regulatory bodies such as the Directorate General of Civil Aviation (DGCA) have streamlined certification processes to support a growing ecosystem of drone manufacturers. Amidst these developments, drones from Bengaluru-based manufacturer Scandron are paving the way in high-altitude logistics, becoming the first to receive DGCA type certification for logistics drones. These machines have successfully flown with heavy payloads at high altitudes, delivering critical supplies to the Indian Army troops stationed in some of the world’s most treacherous and remote terrains. In an exclusive interview with AIM, Arjun Naik, the founder and CEO of Scandron, shared his views on this industry, the company’s presence at Aero India 2025 and offered a sneak peek into Scandron’s upcoming collaborations with the Indian Navy. These drones have not only replaced mules and porters in Himalayan military posts but also entered sectors such as agriculture, internal industrial logistics, and infrastructure inspection. High Above the Himalayas “We’ve had drones fly up to 18,000 feet with 50 kilos,” Naik said. “In January, in minus 20 degrees Celsius, with 30-40 kmph winds and knee-deep snow, the drone took off, climbed above the clouds, and delivered supplies where no helicopter could go.” Recalling one particular mission near Malari in Uttarakhand, Naik said that the drone landed in a spot that was inaccessible even to helicopters. “The jawans clapped when we landed. They said, ‘Now we’ll finally get food in this area.’” In places where porters take six hours to climb with 20 kg loads on their backs, Scandron drones complete the job in just 10 minutes. During another such major test, a drone took off near Tawang in Arunachal Pradesh, ascending through snow, clouds, and fierce headwinds to reach a military outpost across over 5,000 metres. Moreover, the company has built a dedicated high-altitude testing and manufacturing facility in Leh. Drones are designed to operate with lower lift capacities at higher altitudes, in stronger winds, and in severe cold conditions. These adaptations include reduced payloads, excess power buffers, and structural reinforcements. “We don’t just fly drones, we help the armed forces understand how to use them,” Naik explained. “It’s not only logistics anymore. The next question is: can we mount other equipment, maybe even armaments?” Agriculture Meets Automation Scandron isn’t focused on defence alone. Its drones are participating in the ‘Namo Drone Didi’ scheme, empowering women farmers in India’s rural fields to spray crops more efficiently and safely. Manual pesticide spraying is being replaced by AI-assisted, targeted spraying that maps fields and applies chemicals only where needed. “This cuts pesticide use and exposure,” Naik added. “We’ve even partnered with fertiliser companies to find optimal spray concentrations for different crops.” Beyond spraying, Scandron is experimenting with district-level crop intelligence using drone mapping, though Naik admits it will require more coordination and government involvement to scale. The company also supports large solar parks by transporting solar panels within vast facilities. “We are the only company that has drones large enough to move solar panels for installation,” Naik further said. No AI Takeover Just Yet Despite the buzz around AI implementation, Naik is clear-eyed about what AI in drones really means. “There is no AI ecosystem in drones. All that stuff like target acquisition and autonomous warfare, that’s 20 years away,” he says. The company implements AI in its intelligent safety algorithms. Scandron drones are designed to decide mid-flight if they can complete a journey based on battery status and wind resistance. They also have the capability to decide if they have enough battery to reach the destination. “If a drone is four kilometres in, but has used 60% of its battery, it turns around automatically.” Autonomy in these drones is limited to functions like pinpoint landings, obstacle avoidance, and flight adjustments. “AI today is great for safety and data analysis. That’s where the real application is,” he clarified. The company also runs simulation-based design processes, but insists on live testing in every condition. “You can’t fight a war in a lab,” Naik said. “Eventually, you have to fly the drone in snow, sun, rain, and wind.” Grounded in India, Open to the World Scandron designs its drones entirely in-house and manufactures 90% of its components in India. Yet, Naik is pragmatic about India’s electronics limitations. “India is not poised to manufacture electronics today,” he says. “We lack the infrastructure and know-how. Specialised components like thermal cores or battery cells are made by only a few global players.” Nevertheless, government incentives, streamlined drone certification processes, and access to military testing zones are giving companies like Scandron a platform to grow. Naik has also denied the existence of regulatory delays in the industry. “It’s a structured process. People just expect everything overnight.” Despite a strong domestic market, Scandron is not insular. “We’re open to global partnerships, provided they add value,” Naik said. The company operates a drone-as-a-service vertical for industrial clients, including Reliance and Indian Oil Corporation Limited (IOCL), performing internal inspections during shutdowns. “What took a month with cranes, we do in hours with drones,” he noted. Long Flight Path Ahead India is the world’s third-largest user of drones after the US and Israel, driven by active defence zones and internal security needs. However, Naik cautions against believing the hype surrounding the scale of the Indian drone industry. “People say $10 billion market or $2 billion market. But we are a fraction of that. Drones are not mature yet. We haven’t hit mainstream adoption.” Naik believes the next five to 10 years will bring consolidation. “It’s expensive to stay in the game. We’ve invested ₹60 crore and haven’t seen revenue yet, but that was planned.” With a team largely comprising 23-year-olds, Scandron takes training every new hire into its own hands. “Even [graduates] from IITs [aren’t] industry-ready…The basic standard has dropped…We train everyone.” Despite the challenges, technical or financial, Scandron continues its high-altitude climb. Whether it’s icy cliffs or solar parks, the company is betting on Indian engineering to carry weight far beyond expectations.","excerpt":"“Our drones decide if they have enough battery to reach the destination. If not, they turn around on their own.”​","categories":["Deep Tech"],"tags":["Aero India","drone technology","Drones in india","Scandron"],"author_name":"Sanjana Gupta","publish_date":"2025-04-24T09:47:33","publication_year":"2025","word_count":1040,"keywords":["Go","API","ELT","AI","drone technology","ML","RAG","automation","Ray","Aim","Scandron","Drones in india","Aero India","R"],"extracted_tech_keywords":["AI","ML","Aim","Ray","RAG","R","Go","API","ELT","automation"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/these-drones-are-flying-at-18000-ft-above-the-himalayas-for-the-indian-army\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10023703,"title":"IIT Madras Launches Fellowship In AI For Social Good","content":"Indian Institute of Technology Madras’ Robert Bosch Centre for Data Science and AI (RBCDSAI) collaborated with Narayanan Family Foundation to launch a Fellowship in Artificial Intelligence for Social Good. The objective of this Fellowship is to attract early career AI researchers with exceptional promise to RBCDSAI, IIT Madras. The Fellows will get a salary on par with an Assistant Professor at IITs and will be primarily drawn from recent PhD graduates or early career researchers in computer science, computational and data sciences, biomedical sciences, management, finance and other engineering branches. Funded by the Narayanan Family Foundation, the Fellows will get a salary of Rs. 15 lakh to Rs. 18 lakh (approx. US$ 20,000 – 24,000) per year depending on the experience (equivalent to Assistant Professor’s starting salary at any IIT) for a non-renewable term of three years. The Fellows are also eligible for a one-time research grant of up to Rs. 30 lakh (approx. US$ 40,000). Highlighting the unique aspects of this Fellowship, Prof B. Ravindran, Head, RBCDSAI-IIT Madras, said, “This fellowship is designed to enable outstanding candidates to establish their independent research profile and to contribute in a significant way to socially relevant AI research.” Speaking about their objectives in launching this Fellowship, Mr Srinivas Narayanan, Co-Founder, Narayanan Family Foundation, said, “We believe the program can bring significant benefits for society through the advancement of AI, and we are very excited to support it.” The Fellows are expected to conduct their independent research under the mentorship of the Centre in the area of AI for Social Good and will have access to high-end compute infrastructure and datasets at RBCDSAI. Applications are now being accepted. Interested candidates can apply here.","excerpt":"IIT Madras and Narayanan Family Foundation launched a Fellowship in Artificial Intelligence for Social Good to attract early career AI researchers with exceptional promise to RBCDSAI, IIT Madras.","categories":["AI News"],"tags":["AI for social good","IIT Madras"],"author_name":"Ambika Choudhury","publish_date":"2021-04-09T13:29:53","publication_year":"2021","word_count":280,"keywords":["data science","Go","artificial intelligence","programming_languages:R","AI","IIT Madras","BERT","Ray","llm_models:BERT","AI for social good","AI research","R"],"extracted_tech_keywords":["AI","artificial intelligence","data science","Ray","R","Go","BERT","AI research","llm_models:BERT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/iit-madras-launches-fellowship-in-ai-for-social-good\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10093134,"title":"Hugging Face Releases Groundbreaking Transformers Agent","content":"It has been an eventful week for Hugging Face as it has been raining AI innovations. It has just released Transformers Agent which allows users to manage over 100,000 HF models by conversing with the Transformers and Diffusers interface. The new API of tools and agents can generate other HF models to address complex, multimodal challenges. Transformers Agent provides a natural language API on top of transformers, with a set of curated tools and an agent designed to interpret natural language and utilise these tools. The system is intentionally extensible, with the ability to easily integrate any relevant tools developed by the community. Before using the agent.run functionality, users must instantiate an agent, which is an LLM. The system supports both OpenAI modes as well as open source alternatives from BigCode and OpenAssistant. Hugging Face offers free access to endpoints for BigCode and OpenAssistant models. The tools consist of a single function with a designated name and description which are then employed to prompt an agent to perform a given task. The agent is taught how to use these tools by means of a prompt that demonstrates how the tools can be leveraged to accomplish the requested query. While pipelines frequently combine multiple tasks into a single operation, tools are designed to concentrate on one specific, uncomplicated task. There are two APIs available: Single execution (run): The user has access to the single execution method which involves using the agent’s run() method. This method automatically selects the necessary tool or tools required for the task at hand and runs them accordingly. The run() method is capable of executing one or multiple tasks in the same instruction, although the complexity of the instruction may increase the likelihood of the agent failing. Each run() operation is independent, allowing the user to run it multiple times with different tasks consecutively. Chat-based Execution (chat): The agent’s chat-based approach is characterised by its use of the chat() method. This method is particularly useful when there is a need to maintain the state across different instructions. While it is ideal for experimentation, it is not particularly well-suited for complex instructions, which the run() method is better equipped to handle. The chat() method can also accept arguments, allowing for the passage of non-textual types or specific prompts as required. The code is then executed using a small Python interpreter along with a set of inputs provided by the user’s tools. Despite concerns about arbitrary code execution, the only functions that can be called are those provided by Hugging Face and the print function, which limits what can be executed. Additionally, attribute lookups and imports are not allowed, further reducing the risk of attacks. Curated Tools Document-based question answering: Using Donut, agents can answer questions based on a document, even if it is in image format, like a PDF (Donut). Text-based question answering: With Flan-T5, agents can answer questions based on long texts by identifying the most relevant information (Flan-T5). Unconditional image captioning: Agents can add captions to images using BLIP (BLIP). Image question answering: VILT enables agents to answer questions about an image by identifying the most relevant features. (VILT) Image segmentation: With CLIPSeg, agents can output segmentation masks based on image prompts, which can be useful for tasks like object detection (CLIPSeg). Speech to text: Agents can transcribe spoken words into text using Whisper, which is particularly useful for processing audio recordings (Whisper). Text to speech: SpeechT5 enables agents to convert text into speech (SpeechT5). Zero-shot text classification: BART helps agents to classify text into predefined labels without needing prior training data (BART). Text summarisation: With BART, agents can summarise long texts into concise sentences or paragraphs (BART). Translation: NLLB allows agents to translate text from one language to another, which can be useful for communication across different cultures and languages (NLLB). Custom tools Text downloader: This tool enables you to download text from a web URL. Text to image: With this tool, you can create an image based on a prompt, using stable diffusion. Image transformation: Using instruct pix2pix stable diffusion, this tool enables you to modify an image based on an initial image and a prompt. Text to video: This tool generates a brief video according to a prompt, utilizing damo-vilab. Read more: The Peaks and Pits of Open-Source with Hugging Face Last week, Hugging Face partnered with ServiceNow to develop a new open-source language model for codes called StarCoder. The model created as a part of the BigCode initiative is an improved version of the StarCoderBase model trained on 35 billion Python tokens. Researchers stated StarCoder’s capabilities have been tested on a range of benchmarks, including the HumanEval benchmark for Python. The model has outperformed larger models like PaLM, LaMDA, and LLaMA, and has proven to be on par with or even better than closed models like OpenAI’s code-Cushman-001 (the original Codex model that powered early versions of GitHub Copilot).","excerpt":"Transformers Agent provides a natural language API on top of transformers, with a set of curated tools and an agent designed to interpret natural language and use these tools","categories":["AI News"],"tags":["Generative Pre-Trained Transformer","Google","Hugging Face","OpenAI"],"author_name":"Shritama Saha","publish_date":"2023-05-11T12:07:49","publication_year":"2023","word_count":817,"keywords":["Hugging Face","text classification","TPU","OpenAI","AI","Transformers","RAG","Python","object detection","Generative Pre-Trained Transformer","Google","R"],"extracted_tech_keywords":["AI","OpenAI","Hugging Face","Transformers","RAG","text classification","object detection","TPU","Python","R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/hugging-face-releases-groundbreaking-transformers-agent\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":11478,"title":"Mahindra and IBM join hands to develop Blockchain Solution for Supply Chain Finance","content":"In a recent development, a cloud based application, one of the first projects of its kind outside of traditional banking has debuted in India. This Blockchain Solution by Mahindra, one of the largest diversified multinational group of companies based in India and IBM, has the potential to reinvent supply chain finances across India. It would do so by enhancing security, transparency and operational processes. This cloud-based application is designed to transform supplier-to-manufacturer trade finance transactions through a permissioned distributed ledger. The blockchain-based supply chain finance solution will enable all parties involved in the transaction to act on the same shared ledger, with each party updating only their part of the process, ensuring efficiency, consistency, trust and transparency, while safeguarding sensitive information. Why is it needed? Invoice discounting, the process of bundling and selling invoices at a discount, is a major source of working capital finance for many suppliers. This new solution aims to enable more suppliers to access credit, with the goal of driving more financial inclusion throughout the supply chain. Supplier-to-manufacturer invoice discounting processes have historically been said to be difficult, slow and risky, requiring each party to maintain and manually update separate ledgers. Human errors during this process have been known to create inconsistent records, ultimately resulting in delayed payments and capital that cannot be accessed. “The Mahindra Group is pioneering the use of blockchain to disrupt its traditional businesses and drive future growth. We are actively looking at other applications across the Group in financial services, auto, mobility and agritech,” said Anish Shah, Group President (Strategy), Mahindra Group. “This proof of concept represents a significant step forward in making blockchain, still a new technology, a more compelling and efficient supply chain solution for Mahindra Finance’s small and mid-sized enterprises loans business. Working with IBM, we will work to build, test, scale and refine this solution over time.” Blockchain technology can help enable Mahindra Finance access transactions recorded on a shared ledger in near real-time, enabling it to develop and offer new products to small and mid-sized enterprises. This is a radical process and technology shift which has the potential to drive business growth into the future. “Blockchain is poised to revolutionize business like the Internet did, and IBM is at the forefront of the revolution. We offer a comprehensive enterprise-class blockchain solution that is secure, scalable and reliable. The work with Mahindra has the potential to fundamentally transform the way businesses interact with one another and their customers and suppliers, and we’re confident that this engagement can be replicated not just in the finance industry but across other sectors as well,” said Lula Mohanty, Managing Partner of IBM Global Business Services, IBM ISA. IBM is rapidly expanding its blockchain consulting services and capabilities and actively working with clients to understand what it takes to make Blockchain ready for business. IBM is committed to advancing the science of blockchain, helping to remove complexity, and making it more accessible and open. Financial services, supply chains, trade finance and logistics, IoT, risk management, digital rights management and healthcare are some of the areas that are poised for dramatic change using blockchain networks. What would Cloud-based blockchain app ensure? Speed-up invoice discounting Avoid human errors hence ensuring consistent records Do away with delayed payments Would enable parties involved in the transactions to act on same shared ledger, and updating only their part of the process Infuse trust and transparency into the supply chain financing process Overall an efficient supply chain solution for Mahindra Finance’s small and mid-sized enterprises loans business.","excerpt":"In a recent development, a cloud based application, one of the first projects of its kind outside of traditional banking has debuted in India. This Blockchain Solution by Mahindra, one of the largest diversified multinational group of companies based in India and IBM, has the potential to reinvent supply chain finances across India. It would […]","categories":["AI News"],"tags":["blockchain india","blockchain solution in finance"],"author_name":"Srishti Deoras","publish_date":"2016-12-03T06:00:12","publication_year":"2016","word_count":589,"keywords":["Replicate","Go","API","blockchain india","AI","programming_languages:R","Scala","Git","Aim","blockchain solution in finance","Rust","R"],"extracted_tech_keywords":["AI","Aim","R","Go","Rust","Scala","Git","API","Replicate","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/mahindra-ibm-join-hands-developing-blockchain-solution-supply-chain-finance\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10078059,"title":"Musk Notifies Co-investors; to Seal Twitter Deal by Friday","content":"Business magnate Elon Musk has notified his co-investors on closing the $44 billion acquisition of Twitter by Friday. A source familiar with the matter said that equity investors such as Sequoia Capital and Qatar Investment Authority have received the financial commitment paperwork from Musk’s lawyers. The move is the clearest sign that the Tesla leader plans to adhere to the deadline of a Delaware court judge’s order to complete the transaction by the mentioned date. The banks committed to funding Musk’s buyout of the social media platform are in the process of signing the necessary documents and have finished collating the final debt financing and credit agreements. Moreover, Musk also pledged on closing the deal with bankers helping to fund on a video conference call on Monday. Twitter shares spiked and were trading up 3% at $52.95, which was closer to the offer price of $54.20. Musk has claimed to provide an equity amount of $46.5 billion and debt financing for the $44 billion acquisition, which includes the closing costs. Financial institutes such as Bank of America Corp and Morgan Stanley have committed to supporting the deal by providing $13 billion of debt financing. The banks are expected to receive one of the last formalities, i.e. a borrowing notice on Tuesday, and the cash is expected to be held in legal agreement on Thursday, the source said.Equity investors such as Saudi Prince Alwaleed bin Talal and Oracle Corp co-founder Larry Ellison will pitch in with $7.1 billion, putting a closure to months of speculation of Musk abandoning the takeover.","excerpt":"Equity investors such as Sequoia Capital and Qatar Investment Authority have received the financial commitment paperwork from Musk’s lawyers","categories":["AI News"],"tags":[],"author_name":"Bhuvana Kamath","publish_date":"2022-10-26T18:06:17","publication_year":"2022","word_count":259,"keywords":["API","funding","programming_languages:R","AI","Aim","GAN","R"],"extracted_tech_keywords":["AI","Aim","R","API","GAN","funding","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/musk-notifies-co-investors-to-seal-twitter-deal-by-friday\/","complexity_score":3,"technical_depth":7,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10168661,"title":"Think Quantum, Think IBM","content":"As India strives to establish itself as a global leader in quantum computing, IBM has emerged as a pivotal force in shaping its ecosystem. The country has long been hailed for its tech talent, yet it rarely controls the infrastructure underpinning global computing. In an exclusive interaction with AIM, L Venkata Subramaniam, IBM’s Quantum India Leader, offers a ground-level view of how global tech meets national ambition in the subcontinent’s quantum leap. He believes that quantum computing presents a rare window for the country to lead the global supply chain. “When anybody talks quantum, they first think of IBM because of our deep engagement with the ecosystem here,” he said. India’s ₹6,000 crore National Quantum Mission (NQM) aims to drive research, development, and education in quantum technology by forging partnerships across academia, startups, and industry. IBM is closely aligned with this mission and deeply embedded in multiple aspects of it, from education to innovation and commercial applications. “We have a history of innovation and talent creation, but we do not take them too far. We stop at some point and then somebody takes over,” he said during his talk at the Nano Electronics Roadshow 2025. Now, IBM is trying to change that, not just through research and development, but by helping India claim a stake in the future global quantum supply chain. Building a Nation of Builders At the heart of IBM’s strategy is talent, and the belief that India has plenty of it. “India is unmatched at this point in terms of talent and enthusiasm in quantum,” said Subramaniam. But for him, talent is not enough unless it’s applied. “The one thing I want to tell young people today is they need to build. Don’t wait for anybody. Don’t wait for anything. If you have an idea, just build it.” Historically, India’s challenge has been a disconnect between academia and industry. The NQM, Subramaniam believes, is changing that. “The whole purpose of the NQM is to make that academia-industry connect happen so that the talent which gets produced also gets utilised here in India.” IBM’s role includes conducting summer schools, hackathons, and NPTEL (National Programme on Technology-Enhanced Learning) courses, which have attracted over 15,000 students. “We ensured that people actually worked on a real quantum system,” he noted. What is IBM Doing? In support of skilling, IBM is co-developing 11 textbooks with IITs, startups, and other partners. Over 100 colleges have already signed up. This is to help author these textbooks and roll out a nationwide undergraduate minor programme in quantum technologies. “You’ll see 11 books soon coming out,” said Subramaniam. “No other country has gone out and announced a national-scale skilling programme where students can learn quantum, supported by books and video lectures.” He adds that the initiative has also changed traditional academic publishing norms. “They had this rule that only PhDs can write the books… but now they realise that in an area like quantum, sometimes people in industry have so much knowledge, but they might not have a PhD.” This collaborative effort between industry and academia marks a shift in how India approaches technology development and education. “Books are being written and videos are being recorded. IBM quantum computers are available for free use worldwide, and India is utilising them. These children are self-learning and already contributing to open-source code.” Supply Chain Sovereignty Much of the conversation about quantum revolves around the number of qubits and the promise of error correction. Other major companies, such as Google, Microsoft, and Amazon Web Services (AWS), as well as countries like China, have made significant progress. However, Subramaniam wants India to take a step back and think bigger. “The processor is the heart, but it’s not everything,” he said. “There is a software stack, there are memory chips, control circuitry, cryogenic amplifiers, wiring, and microwave subsystems. There is an opportunity to pick one of these and control that supply chain.” In the era of classical computing, five countries controlled nearly every layer of the PC supply chain. India was absent. Subramaniam warns against repeating that mistake and shows an opportunity if the country delves into this early. Rise of Startups Quantum computing startups in India are thriving. When the NQM issued its first call to fund deep tech startups, expectations were modest. “They thought 20 startups might apply. Do you know how many applied? 104,” said Subramaniam. He highlighted the unique nature of many of these companies, expressing that many of these quantum startups were comprised of highly qualified young Indians supported by faculty who had been their teachers. IBM mentors several such startups through its Quantum Network. “We are very deeply engaged with the deep tech startups because we believe there is a lot of talent here, and there is a lot of potential in these startups actually to grow up to become world-class companies in this area.” Beyond private sector collaborations, the union government and various state governments, such as Andhra Pradesh, are establishing “quantum valleys” in partnership with IBM. Quantum’s future doesn’t just lie in R&D labs but in the hands of fast-moving startups that can solve real-world problems. “Historically, our software industry is the best in the world. So if you want to use a computer to solve problems, this is where that solution should come from.” IBM’s Global Roadmap Meets Local Ambitions Subramaniam views IBM as leading the global quantum race with its consistent progress. “IBM is the only company with a very clear published roadmap for this whole decade. We haven’t missed anything. Whatever we said we were going to achieve, we’ve been able to achieve.” That includes a near-term goal of demonstrating quantum advantage, which is solving a problem that a classical computer cannot. “We will have error-corrected qubits by 2029, and that’s ahead of the peers at this point,” he added. While India may not yet build the most powerful hardware, its strength lies in applications and software. He urges that if the country is to utilise a quantum computer to solve problems, the solution should ideally be developed indigenously. Subramaniam points to Tata Institute of Fundamental Research’s (TIFR) development of India’s first six-qubit quantum computer as a sign of tangible progress. “You can go see it in his lab. You can actually programme it,” he said. “It’s only a matter of time.” As India positions itself within the global quantum race, Subramaniam believes that whether it’s through code, collaboration, or curriculum, India has a good opportunity to make a difference.","excerpt":"“We will have error-corrected qubits by 2029, and that’s ahead of the peers at this point,” says IBM India Quantum Lead","categories":["Deep Tech"],"tags":["IBM","quantum","Quantum computing India"],"author_name":"Sanjana Gupta","publish_date":"2025-04-27T15:35:03","publication_year":"2025","word_count":1080,"keywords":["Quantum computing India","Go","API","AWS","AI","cloud_platforms:AWS","innovation","quantum","Aim","IBM","cloud_platforms:Amazon Web Services","R","startup"],"extracted_tech_keywords":["AI","Aim","AWS","R","Go","API","innovation","startup","cloud_platforms:AWS","cloud_platforms:Amazon Web Services"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/think-quantum-think-ibm\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10172671,"title":"LTIMindtree Unveils AI-Powered GCC-as-a-Service","content":"LTIMindtree has launched its new Global Capability Center-as-a-Service (GCCaaS), a comprehensive AI-powered offering aimed at helping enterprises establish, expand, and evolve their GCCs with greater speed, efficiency, and business value. The company’s latest service suite is designed to help organisations either start from scratch or optimise their existing GCCs by offering modular Build, Operate, Transform, and Transfer (BOTT) services. The model offers flexibility through a per-seat or per-service commercial structure, allowing enterprises to tailor their engagements to align with cost and value optimisation goals. LTIMindtree’s in-house talent acquisition system plays a critical role in onboarding skilled professionals who are ready to contribute from day one. “GCCs are becoming strategic centres for industry-specific transformation and efficiency,” said Venu Lambu, CEO and MD at LTIMindtree. “LTIMindtree’s GCC-as-a-Service helps enterprises build, scale, and evolve their GCCs into global innovation hubs, leveraging our BlueVerse ecosystem to drive next-gen capabilities and gain a competitive edge with scalable, responsible AI.” The GCCaaS platform includes LTIMindtree’s Talent Solutions as part of its Build Services, providing end-to-end support for setting up legal entities, ensuring compliance, and establishing infrastructure across major global hubs. This includes operational support in finance, accounting, taxation, IT, and workspace development. Under Operate Services, clients benefit from robust program governance, delivery excellence, knowledge management, and transition services, backed by LTIMindtree’s AI-based learning and engagement platforms—Talent Engage and Shoshin. The Transform component offers industry-specific solutions and access to the company’s global AI studios—located in the US, Europe, and India—as well as its BlueVerse Agentic AI Ecosystem featuring domain-specific agents. These services are designed to fast-track clients’ AI journeys and deliver faster value realisation. Read: LTIMindtree Unveils BlueVerse, a New Business Unit with 300 AI Agents Finally, the Transfer service ensures smooth handovers through structured talent migration, knowledge transfer, and long-term change management support to ensure continuity and sustained performance. The launch reinforces LTIMindtree’s focus on delivering integrated, AI-driven services that cater to the evolving needs of global enterprises seeking to unlock greater value from their GCC operations.","excerpt":"LTIMindtree’s suite is designed to help organisations either start from scratch or optimise their existing GCCs by offering modular BOTT services.","categories":["AI News"],"tags":["GCC","LTIMindTree"],"author_name":"Mohit Pandey","publish_date":"2025-07-01T16:15:59","publication_year":"2025","word_count":331,"keywords":["Go","agentic AI","GCC","AI","innovation","Scala","RAG","GAN","responsible AI","Aim","LTIMindTree","R"],"extracted_tech_keywords":["AI","agentic AI","Aim","RAG","R","Go","Scala","GAN","responsible AI","innovation"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/ltimindtree-unveils-ai-powered-gcc-as-a-service\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":30830,"title":"Qualcomm Launches $100 Million Investment Fund For Artificial Intelligence","content":"Noted semiconductor and telecommunications equipment company Qualcomm this week announced the launch of the Qualcomm Ventures AI Fund to invest up to an aggregate of $100 million in startups transforming artificial intelligence. The company announced that the fund would focus on startups that share the vision of on-device AI becoming more powerful and widespread, with an emphasis on those developing new technology for autonomous cars, robotics and machine learning platforms. This fund is to build on more than a decade of Qualcomm’s AI research and its heritage of developing the foundational building blocks of low power processing and connectivity, which are essential for AI. Steve Mollenkopf, CEO of Qualcomm Incorporated, said in a statement, “At Qualcomm, we invent breakthrough technologies that transform how the world connects, computes, and communicates. For over a decade, Qualcomm has been investing in the future of machine learning. As a pioneer of on-device AI, we strongly believe intelligence is moving from the cloud to the edge. Qualcomm’s AI strategy couples leading 5G connectivity with our R&D, fueling AI to transform industries, business models and experiences.” Qualcomm has set out to make on-device AI technology ubiquitous by inventing, developing, commercialising and, importantly, investing in it. As AI shifts towards the wireless edge – combining essential on-device capabilities with the edge cloud – the industry is already starting to see the full potential of 5G. Qualcomm’s ambitious 5G vision and strategic commitment to on-device AI goes hand in hand with mobile becoming the pervasive AI platform. As part of the AI Fund, Qualcomm Ventures LLC participated in a Series A funding round for AnyVision, a noted face, body, and object recognition startup. This investment – the first one made by the AI fund – will further the startup’s efforts to expand into other industries and develop new AI applications that transform how the world connects, computes and communicates. Also see:","excerpt":"Noted semiconductor and telecommunications equipment company Qualcomm this week announced the launch of the Qualcomm Ventures AI Fund to invest up to an aggregate of $100 million in startups transforming artificial intelligence. The company announced that the fund would focus on startups that share the vision of on-device AI becoming more powerful and widespread, with an […]","categories":["AI News"],"tags":["AI (Artificial Intelligence)","Qualcomm"],"author_name":"Prajakta Hebbar","publish_date":"2018-11-29T05:33:48","publication_year":"2018","word_count":313,"keywords":["Go","funding","artificial intelligence","machine learning","programming_languages:R","AI","Qualcomm","ViT","AI research","R","AI (Artificial Intelligence)","startup"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","R","Go","ViT","startup","funding","AI research","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/qualcomm-new-100m-ai-investment-fund\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10165792,"title":"Cognizant, Are You Okay?","content":"Cognizant appears to be in slight trouble. In two of the last three years, the company’s revenue growth fell below the 5% mark, and in one of the years, it reported a negative growth. The outsourcing firm’s financials show some visible signs of distress emerging across its operations, which include giving away its workspace and also reducing headcount. According to a recent report, Bagmane Group is in advanced talks to acquire the Chennai office property of Cognizant for approximately ₹800 crore. The campus, which has served as Cognizant’s India headquarters for over two decades, spans 13.39 acres and includes 5.6 lakh square feet of office space along Chennai’s bustling IT corridor. According to estimates, the Chennai office has a capacity of around 55,000 employees. Following the acquisition, the site has the potential for expansion, with the possibility of developing up to 3 million square feet of additional space. Cognizant says that this transaction aligns with its ongoing cost-optimisation strategy, which aims to cut $400 million in expenses over two years as part of its 2023 restructuring plan. Cognizant has strategically shifted towards an asset-light model, focusing on shedding non-core real estate holdings. The company has been reducing its global office footprint, shedding 11 million square feet of space worldwide. In 2023 alone, Cognizant reduced its Chennai office space by 1.15 million square feet to enhance operational efficiency and also sold its office spaces in Hyderabad of 10 acres, citing similar reasons. The company had plans to consolidate its Chennai operations into three of its own buildings located at Madras Export Processing Zone (MEPZ), Sholinganallur, and Siruseri, with headquarters relocating to the MEPZ campus near Tambaram by December 2024. The Thoraipakkam property, historically significant as the site where Cognizant remotely rang the Nasdaq opening bell, has faced challenges including flooding issues due to its proximity to a water body, leading to increased maintenance costs. This practical consideration further justifies the strategic nature of the sale rather than indicating financial distress. A mail sent by AIM to Cogizant seeking clarification on the matter went unanswered. In August 2024, Cognizant expanded its presence in Indore with its first office, a move set to create over 1,500 jobs with potential growth to 20,000 in the future. However, all seems fine for Cognizant at the moment. The Employee Problem Remains Real Despite the positive outlook for selling its office spaces, Cognizant has experienced notable changes in its workforce structure, with headcount declining by 10,700 employees year-over-year to 3,36,800 as of December 31, 2024. AI adoption hasn’t been all good news for Indian IT. For the most part, firms integrate generative AI products into their client offerings. At the same time, there have been several reports about how these AI tools are slowly going to reduce the demand for services themselves, making them rethink their business models. Cognizant has been quite vocal about implementing generative AI into its workflow and also for its clients. This might indicate why there is a decline in headcount. Another interesting aspect is that Cognizant employees in India will have to wait until August 2025 for their next salary hike, marking the second consecutive year of delays. Even last year, the company postponed increments, eventually providing only a 5% raise in August. Meanwhile, eligible employees have began receiving their bonuses from March 10, according to an internal memo. Reports suggest the delay is part of an effort to manage high attrition rates. Voluntary attrition at Cognizant climbed to approximately 16% in the last fiscal year, up from 13.8% for the period ending December 31, 2023. This reduction in headcount included a sequential decrease of 3,300 employees in the final quarter of 2024. Simultaneously, the company’s attrition rate has increased to 15.9% on a trailing twelve-month basis, representing a rise of 2.1 percentage points. Leadership Remains Confident Despite these reductions, company leadership has expressed confidence about future hiring. CFO Jatin Dalal revealed, “As we look forward, we feel that we will add headcount as we need to grow and as we grow during the course of 2025.” Notably, CEO Ravi Kumar S highlighted a positive trend in talent attraction, highlighting that 13,000 former employees had returned to Cognizant, with an additional 10,000 expressing interest in rejoining. This suggests the company is adjusting its workforce strategically rather than implementing crisis-driven cuts. The company’s utilisation rate declined by 2 percentage points to 82%, though management emphasised that utilisation improvements remained strong throughout 2024. Speaking of utilisation, a former Cognizant employee shared their experience of spending over two years at the company with little to no real work. After onboarding and four months of Java Spring Boot training, they waited two months for a project assignment. When placed, they, along with two other freshers, were told to wait another three months until a vacancy opened. Even after receiving knowledge transfers (KTs), they were overlooked for the role and continued waiting. Placed on the bench, they struggled to find another project amid intense competition. HR advised them to train in Apache Camel, which led to another two months of unproductive learning. Forced to work from the office five days a week without their personal laptop, they were unable to prepare for external opportunities. Despite a year of “experience”, actual work remained elusive. A senior eventually gave them minor tasks, but within two months, they were relieved from the project due to cost-cutting amidst the time of severe layoffs. Five months later, they were terminated with three months of severance pay. Financials Remain Sceptical Cognizant reported full-year 2024 revenue of $19.736 billion, representing a 1.98% increase compared to the previous year. This growth reverses the slight downward trend in 2023, when the company reported revenue of $19.353 billion, a marginal 0.39% decline from 2022 and a 4.9% revenue growth in 2022. For FY21, the company reported total revenue of $18.5 billion, with 11.1% year-on-year growth, with business momentum in the digital business after a relatively flat performance in 2020, which had seen a slight decline of 0.78% compared to 2019. This has been the case for most of the Indian IT firms as most of them reported single digit growth over the last few quarters. Despite this, Cognizant has provided a positive outlook for 2025, forecasting revenue growth of 3.5% to 6.0% in constant currency terms, with revenue expected to reach between $20.3 billion and $20.8 billion. For the first quarter of 2025 specifically, the company expects revenue between $5.0 billion and $5.1 billion, representing growth of 5.6% to 7.1% year-over-year. The numbers, however, continue to be single digit. Based on the available evidence, Cognizant appears to be a company in transition rather than in trouble, just like any other Indian IT firm. While workforce reductions and real estate divestitures might initially raise concerns, these actions align with a deliberate strategy to optimise operations, reduce costs, and position the company for future growth. The upcoming Investor Day on March 26 will likely provide further clarity on the company’s strategic direction and growth plans.","excerpt":"The company has been reducing its global office footprint, shedding 11 million square feet of space worldwide.","categories":["AI Features"],"tags":["Cognizant","Developers"],"author_name":"Mohit Pandey","publish_date":"2025-03-11T13:14:36","publication_year":"2025","word_count":1166,"keywords":["Go","programming_languages:R","AI","programming_languages:Java","Cognizant","Git","Aim","generative AI","GAN","R","Java","Developers"],"extracted_tech_keywords":["AI","generative AI","Aim","R","Go","Java","Git","GAN","programming_languages:R","programming_languages:Java"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/cognizant-are-you-okay\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10100866,"title":"Is ChatGPT Making Therapists Anxious?","content":"Having confidently diagnosed a four-year-old’s mysterious disease that 17 doctors failed at, ChatGPT now goes a step ahead in the medical profession to don the therapist’s hat. Lillian Weng, head of safety systems at OpenAI, recently had a long heart-to-heart conversation with ChatGPT in voice mode (a recent update). Interestingly, even though she never sought therapy, Weng feels that ChatGPT is a good therapist. Just had a quite emotional, personal conversation w\/ ChatGPT in voice mode, talking about stress, work-life balance. Interestingly I felt heard & warm. Never tried therapy before but this is probably it? Try it especially if you usually just use it as a productivity tool. https:\/\/t.co\/9S97LPvBoS— Lilian Weng (@lilianweng) September 26, 2023 Weng’s conversation with ChatGPT sparked mixed reactions. Users took to X to share their concerns on how using ChatGPT as a therapist was both sad and wrong, likening it to the “Eliza Effect”, where people assign human-like emotions to AI. This was exemplified by the early “therapy chatbot” Eliza by MIT scientist Joseph Weizenbaum in 1966, highlighting the tendency of chatbots to mirror users’ language without true understanding. Eliza unintentionally drew people into deep and emotional conversations, revealing the potential for human attachment to AI. However, Weng is not the first person to treat ChatGPT as a therapist. Users find it convenient and appreciate its empathetic responses, although mental health experts express concerns about its limitations. Despite these concerns, individuals have found ChatGPT helpful in offering practical advice and a human-like interaction, making it a unique alternative for those unable or unwilling to seek professional therapy. Thirty-seven-year-old EMT Dan initially used ChatGPT for creative writing, but found solace in discussing his real-life struggles with the chatbot, especially when it came to cognitive reframing—a technique suggested by his therapist. Twenty-seven-year-old Gillian also used ChatGPT for therapy, considering the skyrocketing cost of healthcare. On the other hand, a Belgian man tragically succumbed to suicide following six weeks of “seeking therapy” with Chai.AI chatbot. However, considering the increasing charges of therapy, often not covered by health insurance, people tend to gravitate towards LLM-based chatbots like Bard, ChatGPT, and Perplexity AI. In the future, once the robustness of our models will exceed some threshold, we will have *wildly effective* and dirt cheap AI therapy. Will lead to a radical improvement in people’s experience of life. One of the applications I’m most eagerly awaiting. https:\/\/t.co\/latwdlcaoq— Ilya Sutskever (@ilyasut) September 27, 2023 While AI can offer advice and support, it cannot diagnose specific mental health conditions or provide accurate treatment details. Some worry that users might be disappointed, misled, or compromise their privacy by confiding in the chatbot. Can Chatbots Replace Therapists? Traditionally, chatbots have been “stateless”, treating each new request as an independent interaction without recollection or learning from past conversations. However, GPT-4 introduces a new calling function that enables it to remember user input from previous interactions, resulting in a highly personalised experience. Now, with ChatGPT’s ability to engage in natural language conversations, humans are more prone to forming attachments. According to a research paper by the University of Tennessee and Illinois State, interaction with an AI model can trigger the same emotional responses as interacting with a human. “A person expresses their true self more when interacting with generative AI models, providing an experience nearly identical to human interaction while eliminating the need to carefully consider words before speaking,” noted Nikita (Zeb) Shringarpure, a psychology professor at Mumbai University. This highlights the growing dependency of humans on AI, as it reduces cognitive effort and draws people towards tasks requiring less mental exertion. Furthermore, LLMs possess the ability to simulate human characteristics, showcasing distinct personalities shaped by biological and environmental influences. These personalities play a crucial role in influencing interactions and preferences, blurring the lines between human and artificial intelligence interactions. In a recently published paper by Google DeepMind, larger and instruction-fine-tuned LLMs show stronger evidence of reliability and validity in synthetic personality generation. The study also reveals the possibility of shaping LLMs to imitate human behaviour, including matching different human personalities, as seen in their actions, such as creating posts on social media. Decoding the Sentience Debate Someone finding solace in sharing their stories with ChatGPT in voice and getting attached to it is not new. Eugenia Kuyda’s Replika chatbot helped many people cope with symptoms of social anxiety, depression, and PTSD, TIME reported. Many people fell in love with the chatbot as well. Humans have formed emotional connections with AI chatbots for a long time now, sparking interest in the phenomenon of para-social relationships. “These connections, though fantastical, emulate genuine human bonds. The potential of AI to develop its own identity and attain sentience opens up limitless possibilities,” clinical psychologist Hemalatha S told AIM. The concept of AI becoming sentient has been debated for a while now. Back in June, Google’s Blake Lemoine was fired for calling the big tech’s LLM LaMDA sentient. Now with companies like Microsoft, OpenAI, and Google racing towards AGI that can replicate the cognitive abilities of humans, the prospect of AI having consciousness is being widely debated in recent times. Not just Lemoine, OpenAI’s cofounder Ilya Sutskevar and Andrej Karpathy also received backlash on X for a similar thought. it may be that today's large neural networks are slightly conscious— Ilya Sutskever (@ilyasut) February 9, 2022 LLM chatbots allow us to tailor companions as per our preferences. The prospect of creating ideal partners, be they platonic, romantic, professional, or therapeutic, is a notable aspect of AI’s impact on human relationships, Hemlatha added. However, there’s a cautionary note about the unforeseen consequences and potential evolution of AI into sentient entities. Though ChatGPT-like chatbots may not replace professional therapy, their intriguing resemblance to talk therapy sparks user interest, aligning with individuals’ conceptualization. This could drive increased engagement in formal therapy sessions.","excerpt":"People took to X to share their concerns about how using ChatGPT as a therapist was both sad and wrong, likening it to the “Eliza Effect”","categories":["AI Features"],"tags":["ChatGPT"],"author_name":"Shritama Saha","publish_date":"2023-09-29T11:48:43","publication_year":"2023","word_count":968,"keywords":["ChatGPT","artificial intelligence","OpenAI","AI","neural network","chatbots","ML","RAG","Aim","generative AI"],"extracted_tech_keywords":["AI","artificial intelligence","ML","neural network","generative AI","ChatGPT","OpenAI","Aim","RAG","chatbots"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/is-this-the-end-of-human-therapists\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":12821,"title":"Want to land a data scientist job, check out Facebook, Amazon data scientist job descriptions","content":"The data scientist’s job has evolved over the years. We list down job descriptions that will clarify the role You know practically every company out there wants to hire a data scientist. From marquee companies such as Google, Hortonworks, Amazon to startups. And you may be all too familiar with the job description as well — solving complex problems and generating insights with the use of machine learning algorithms, mining and visualization techniques. Secondly, working with clients, program managers and other IT teams and presenting findings in a clear, precise manner. Many a time, the job description doesn’t end there.  Most companies demand out of the box skills and there are some significant changes in data scientist job description templates. While the core skills, breathing statistics remain the same, some job description border on the theoretical end. Analytics India Magazine lists down some exciting and clearly constructed data scientist job description templates from legacy companies: Facebook Take a look at the data scientist job opening at Menlo Park- based social media giant Facebook. Here’s their recent data scientist job description for the London office: Job Summary: We’re looking for data scientists to work on our core and business products (Instagram, Ads, Messaging, Identity, Growth & Engagement, Mobile, Search, Privacy, Payments) to help drive informed business decisions for Facebook. You get to work with richest data sets. Who’s the perfect candidate: The perfect candidate will have a background in a quantitative field, backed by an experience in working with massive data sets. Should be   scrappy, focused on results, a self-starter, and have demonstrated success in using analytics to drive the understanding, growth, and success of a product. Core responsibilities:  The candidate’s core responsibilities range from data mining and presentation of data to informing product launches. Facebook has broken down the role in four parts – working on data infrastructure (Hadoop and Hive); product operations (building dashboards and reports); exploratory analysis (building user behavior models, identifying long term trends) and product leadership (communicating and validating results) Perquisite Skills: Graduates from engineering, math, statistics, physics or computer science background with an advanced degree. Skilled in programming languages (Python, PHP, Perl) and know-how of the full statistical package (SAS, SPSS, R, MATLAB), knowledge of Hadoop and Hive among others. —————————————————————————————————————————- Amazon World’s biggest online retailer and a leading provider of cloud-hosted services Amazon with a dominant share in retail sales needs “outstanding data scientists to unlock questions about Amazon workforce”. Job Summary: Amazon is looking for an outstanding Data Scientist for their HR Analytics team and the candidate will be part of their leading research and analysis to unlock questions about Amazon’s workforce. The candidate will work closely with the business and technical teams and deliver analysis for actionable results. The candidate will also build predictive workforce models that have a direct impact on day-to-day decision-making and our HR investments. Who’s the perfect candidate: Amazon’s perfect candidate is someone with a Masters backed by 4+ years of relevant experience. The candidate should have a background in relational databases, data mining, data transformation and the ability to perform complex quantitative analysis. Needless to say, problem solving skills and a thorough grounding in statistical approaches is a must-have. Core responsibilities: Some of the core responsibilities include analyzing human resources data to find relationships and trends; develop predictive workforce models for attrition, high performance and recruiting demand; visualization; developing and expanding HR Data Mart and Recruiting Analytics programs through design of reporting catalog and sharing reports, metrics with business leaders. Perquisite Skills: To make the cut, you have to be a Masters in a technical field, extensive experience in statistical approaches, a focus on customer service, knowledge of SQL and relational database, ad-hoc analysis and metric presentation, adept team player, ability to multi-task and most importantly the ability to handle confidential information. Proficiency in the full statistical suite (R, Stata and SAS) is a must-have and knowledge of HRMS systems is preferable. —————————————————————————————————————————- Walmart Want to know what the world’s biggest retail giant Walmart with plans to corner a big slice of online market needs from their data scientist. Check out Walmart Lab’s Principal Data Scientist job description for California location. Job Summary: The Fortune 1 Company is looking for phenomenally talented machine learning research scientists for data cleansing, modelling and implementation role. The candidate will contribute to Growth organization within Walmart Labs and will drive personalization, ad-targeting, context extraction and prediction campaigns. The candidate will also focus on building new products for expanding customer base. Who’s the perfect candidate: The perfect candidate will be backed by an extensive experience in working on large data sets, scaling algorithms to new data sets and should be able to capture business needs and with data modelling initiatives. The candidate will also provide support to other project teams working on email targeting, consumer recommendations in Walmart ecommerce. undoubtedly, statistical analysis and knowledge of system tools such as Hadoop, MySQL and Weka among others are a must-have. Core responsibilities: The job responsibilities include analyzing massive data sets to create models and algorithms for driving business solutions; interpreting data to find trends that go beyond future data sets; providing insights to cross-functional teams regarding data modelling; ensuring business strategies are aligned across all the initiatives and leading the implementation of data modelling solution among others. Perquisite Skills: While Walmart Labs hasn’t listed minimum qualifications for this job role, going by their past data scientist job descriptions a PhD in Machine Learning or related field is a must. Also, knowledge of tools such as MATLAB, Ilog and Hadoop is a must. The candidate should also support Walmart’s mission, values and adhere to the company policies. We hope these data scientist job description templates will help in drawing a clear picture of the industry requirements of the hottest job of the year. Also, Massachusetts Institute of Technology  has drawn a data scientist job description template that gives an overview of the role and how to break into. If you aim to be a Googler for life, the search engine giant has listed down their requirements in the most cogent form.","excerpt":"You know practically every company out there wants to hire a data scientist. From marquee companies such as Google, Hortonworks, Amazon to startups. And you may be all too familiar with the job description as well — solving complex problems and generating insights with the use of machine learning algorithms, mining and visualization techniques. Secondly, […]","categories":["Global Tech"],"tags":["hadoop data catalog"],"author_name":"Richa Bhatia","publish_date":"2017-02-17T07:53:15","publication_year":"2017","word_count":1016,"keywords":["Go","machine learning","AI","Python","Aim","analytics","SQL","GAN","hadoop data catalog","R","startup"],"extracted_tech_keywords":["AI","machine learning","analytics","Aim","Python","R","SQL","Go","GAN","startup"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/want-land-data-scientist-job-check-facebook-amazon-data-scientist-job-descriptions\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10088605,"title":"Unboxing LLMs","content":"In line with how animals and humans produce ‘amazingly complex’ and ‘beautiful stuff’, OpenAI’s founder Sam Altman boasted about what LLMs are capable of. “Language models just being programmed to try to predict the next word is true, but it’s not the dunk some people think it is,” he tweeted. However many researchers disagree. A leading voice in AI, Gary Marcus, for instance, countered Altman’s view of LLM saying animals are built innately with the capacity to represent models of the world, but language models are not. “And that, my friend, is the dunk that someday you will come to appreciate,” added Marcus. Conversely, most researchers will attest to AI being a black box, albeit with some dissenting views. However, with the advancements in LLMs such as GPT-3 and LLaMA, the veil is lifting, and AI is no longer a mystery. These models work by taking input and predicting the next word to generate text. These models can be fine-tuned by giving them a small amount of data related to a specific task, like writing an essay about cats. Researchers can provide feedback on the model’s outputs to help it learn to adjust its weights to perform better on the task. It does this by learning patterns. It can learn that the word “cat” comes after the words “a” or “the”, or that the word “yellow” is an adjective that describes a colour. An LLM learns through cells called neurons forming a neural network. Each neuron can receive inputs and produce an output. The output depends on  “weights”, which are numbers that measure the importance of the input. Talking about ChatGPT, the latest LLM kid, Wolfram Research CEO Stephen Wolfram said in one of his recent AMAs, “There are millions of neurons—with a total of 175 billion connections and therefore 175 billion weights. And one thing to realise is that every time ChatGPT generates a new token, it has to do a calculation involving every single one of these weights.” Late last year, the research, ‘Talking About Large Language Models’ sparked a discussion about how these language models are not simply “predicting the next statistically likely word”. Users argued that though we know how to train them, we are in the dark about how the resulting models do what they do. For instance, we know how humans evolved, but we don’t have perfect models on how humans worked; we have not solved psychology and neuroscience yet. A relatively simple and specifiable process produced complex human beings. Likewise, LLMs are developed using a large training set but the resulting billion(s) parameter model remains enigmatic. This is why “AI interpretability” exists to probe large models like LLMs, and understand how they produce the results. Experts believe the black box exists due to the decisions made by intermediate neurons on the way to making the network’s final decision. It’s not just complex, high-dimensional non-linear mathematics; the black box is due to non-intuitive decisions. The 200-year-old answer However, recent research suggests that 200-year-old maths could help understand how neural networks perform complex tasks. This could increase neural networks’ accuracy and their learning speed, researchers say. To analyse a neural network designed to carry out physics, senior author Pedram Hassanzadeh, a fluid dynamicist and his colleagues experimented with using Fourier analysis, a maths technique often employed in physics. The researchers performed an analysis on the deep neural network’s governing equations. Every model approximately has one million parameters. The connected neurons adjust specifically during calculations. These parameters were assembled in around 40,000 five-by-five matrices. The analysis revealed that the parameters behaved like a combination of low-pass, high-pass, and Gabor filters. All said and done, “as much as neural networks are called black boxes, we can take them apart and try to understand what they do, and connect their inner workings with the physics and maths we know about physical systems,” Hassanzadeh says. “There is a major need for this in scientific machine learning.” Trial, Error, and Trial AI researchers at Meta AI, Princeton University and the Massachusetts Institute of Technology collectively attempted to provide a theoretical framework ‘The Principles of Deep Learning Theory: An Effective Theory Approach to Understanding Neural Networks’, to answer the question around neural networks being a black box. In a blog post, Meta AI research scientist Sho Yaida noted that AI is in the same conjuncture as steam machines at the beginning of the Industrial Revolution. He said though the steam engine changed manufacturing forever, scientists could not theoretically fully explain how and why it worked until the laws of thermodynamics and the principles of statistical mechanics developed over the century. Many of the improvements made to the steam engine were made as a result of trial and error. Understanding the theory requires a sophisticated understanding of physics. The important thing is that it will enable AI theorists to push for a deeper and more complete understanding of neural networks, said Yaida who collaborated with Dan Roberts of MIT and Boris Hanin at Princeton for the research.","excerpt":"“Understanding the theory requires a sophisticated understanding of physics”","categories":["AI Features"],"tags":["AI Black Box","black box ai","ChatGPT","GPT3","GPT4","Meta AI","OpenAI"],"author_name":"Tasmia Ansari","publish_date":"2023-03-03T13:00:00","publication_year":"2023","word_count":835,"keywords":["ChatGPT","Meta AI","machine learning","TPU","OpenAI","AI","neural network","AWS","AI Black Box","black box ai","deep learning","GPT3","R","GPT4"],"extracted_tech_keywords":["AI","machine learning","deep learning","neural network","ChatGPT","OpenAI","Meta AI","AWS","TPU","R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/unboxing-llms\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10099408,"title":"Indian Bank Partners with IBM to Boost Scalability and Agility","content":"IBM today announced that it has collaborated with Indian Bank, a premier public sector bank, to accelerate its modernization path with flexible and secure compute infrastructure for deploying new front-end banking applications, while addressing the business growth of the Core Banking Solution (CBS) workloads for the bank. The key objective of this partnership is to achieve greater agility, scalability, and security to meet the bank’s growth requirements in today’s dynamic and hyper-competitive market. With over 100 million customers and a vast network of 5,798 branches spanning India and overseas, Indian Bank as part of their ambitious “Project PARADISE”, sought to consolidate their Front Branch Servers for improved manageability and availability at their Data Centres (DC) and Disaster Recovery (DR) centres. Additionally, Indian Bank aims to run its Core Banking Solution (CBS) workloads on the latest generation of IBM Power® servers. Deepak Sarda, chief general manager – IT and partnership at Indian Bank, expressed his enthusiasm for this collaboration, stating, “As part of our capacity management initiative and to manage evolving application requirements, we wanted a cloud-like functionality within our own data centre. With benefits like better flexibility, lower costs, and enhanced security, we are creating a value-oriented solution by setting up a proactive and service-oriented model for the bank.” Viswanath Ramaswamy, vice president, technology, IBM India & South Asia, emphasized the importance of embracing new technologies in the banking sector. He stated, “Evolving customer needs demand a new operating model for banks driven by a nimbler digital environment. Those that embrace new technologies to modernize can gain a significant competitive edge, build market differentiation through innovation, and prepare themselves for the new financial frontier.” IBM’s proposed solution for Indian Bank includes a hybrid cloud approach for all AIX workloads, aimed at boosting agility. This approach combines Capex and Opex models for the consumption of compute resources, effectively providing Infrastructure-as-a-Service (IaaS). It also creates a private cloud using available resources, establishing a single pool of processors across DC and DR sites, and enabling services equivalent to a pay-as-you-grow model. To enhance security, Indian Bank is implementing IBM PowerSC for all AIX endpoints, ensuring better support for compliance audits. IBM PowerVC has been deployed for simplified virtualisation management, cloud deployments, workload image management, and dynamic resource optimisation for IBM AIX Servers. By implementing Power Enterprise Pools, Indian Bank has streamlined the utilisation of compute resources across their DC and DR sites, ultimately reducing the total cost of ownership. The new cloud environment provided by IBM offers several advanced features, including virtual machine replication for consistency, a self-service portal for streamlined provisioning, and capabilities such as virtual image capture, deployment, and cataloging. Through the consolidation of front-end branch servers, Indian Bank enhances manageability and availability at its data centers in Chennai and Mumbai, ensuring uninterrupted service for customers. In another news, Indian Bank recently set up 10 startup cells across different centres in the country to serve its specialised banking requirements. Through this the bank aims to create banking products and services for startups keeping in mind their unique and specialised baking requirements.","excerpt":"IBM’s proposed solution for Indian Bank includes a hybrid cloud approach for all AIX workloads, aimed at boosting agility","categories":["AI News"],"tags":["IBM"],"author_name":"Siddharth Jindal","publish_date":"2023-09-04T12:34:26","publication_year":"2023","word_count":510,"keywords":["programming_languages:R","AI","innovation","ML","Scala","Git","Aim","IBM","programming_languages:Scala","R","startup"],"extracted_tech_keywords":["AI","ML","Aim","R","Scala","Git","innovation","startup","programming_languages:R","programming_languages:Scala"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/indian-bank-partners-with-ibm-to-boost-scalability-and-agility\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":36799,"title":"Will India Emerge As A Top Competitor For AI Patents In Asia?","content":"Image source: Patent- Moradian Law As the development in the field of artificial intelligence and machine learning grow steadily, the increasing number of patents filed in this file is an indication of the growing prevalence of these emerging technologies. According to researchers, even though AI has been around for a while, it is only in the past seven years that the technology has advanced at a great speed. Following the initial development in the field is made as early as the 1950s, the industry went through a series of up and down movement referred to as the AI summers and winters. However, after 2012, due to data influx arising out of the increased presence of connected devices and the advancements in the computational field, shaped the next wave of AI resulting from increased funding and optimism towards the technology. According to the World Intellectual Property Organisation (WIPO) Technology Trends report, which studied AI patents across the world to get a comprehensive picture pertaining to the development in the field, almost 50% of the AI patents have been filed only after 2013, with the numbers totalling up to 170,000. The report also briefs about how India has become one among the top 10 countries to file AI patients. As per the study, India was ranked eighth when it came to the first file in AI patent since 2015 and says that the country has witnessed a high rate of annual growth in this regard in the preceding years. It further added that the country was ranked third fuzzy logic and fourth in machine learning. One of the primary contributors from India and a top contender in the international list of AI patents has been IBM, with a total of an impressive 5,930 AI-related patents. According to recent reports, IBM inventors from India had filed over 800 patents, turning themselves into the second largest contributor to the company’s overall patients in technologies ranging from AI to cloud computing. Another key contributor to the field is educational institutes and various scholars in this institution. What are the top patented field As per the study, though AI- related scientific publication has been around for a while, in recent times there has been an upsurge in AI application related patents.  According to the study, machine learning and NLP among other subsets within AI was the most patented topic. The more dominant subset of AI that with more patent has been machine learning, with technology forming one-third of all identified inventions (134,777 patent documents). ” Filings of machine learning-related patent have grown annually on annual average by 28 per cent, with 20,195 patent applications filed in 2016 (compared with 9,567 in 2013),” the study says. Among AI functional applications, computer vision, which includes image recognition, is the most popular. Computer vision is mentioned in 49 per cent of all AI-related patents (167,038 patent documents), growing annually by an average of 24 per cent (21,011 patent applications filed in 2016). Those AI functional applications with the highest growth rates in patent filings in the period 2013 to 2016 were AI for robotics and control methods, which both grew on average by 55 per cent a year. What are the Indian laws in place In India, at present, there are no separate clauses that could be applied to AI and other emerging tech-related inventions. Rather AI technology is encompassed in computer-related Inventions (CRIs) guidelines which are focused on computers\/algorithm\/software. Further disclosure regarding the type of algorithms and software among other things are required to be made to own a  patent. Describe hardware (eg computer system, server, sensors etc.) along with AI algorithms in your patent; Claim working method\/process of the device which uses AI; and Refrain from focussing directly on programming codes\/algorithms of AI.","excerpt":"As the development in the field of artificial intelligence and machine learning grow steadily, the increasing number of patents filed in this file is an indication of the growing prevalence of these emerging technologies. According to researchers, even though AI has been around for a while, it is only in the past seven years that […]","categories":["AI Features"],"tags":["AI in India","AI patents","ibm research","robotic inventions","subsets of ai"],"author_name":"Akshaya Asokan","publish_date":"2019-03-25T09:20:44","publication_year":"2019","word_count":626,"keywords":["artificial intelligence","AI in India","AI patents","AI","machine learning","ibm research","cloud computing","image recognition","robotic inventions","AWS","computer vision","subsets of ai","NLP","RAG","Aim"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","NLP","computer vision","Aim","RAG","image recognition","cloud computing","AWS"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/will-india-emerge-as-a-top-competitor-for-ai-patents-in-asia\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":41346,"title":"India, Japan Float $187 Million To Back Indian Startups In AI, IoT","content":"In a move to accelerate tech development in the continent, India and Japan are jointly launching a new $187 million (₹1,298 crore) fund-of-fund to invest in over 100 tech startups. Reportedly, this fund will focus on ventures related to artificial intelligence and the internet of things. Of the total amount, ¥15 billion or $140 million will be raised from Japanese investors while the remaining will be raised from Indian investors. Reportedly, Reliance Nippon Life Asset Management Limited (RNAM) will be managing the fund. TV Mohandas Pai, chairman of Aarin Capital was reportedly instrumental in getting this funding from Japan. “Japan is a great partner to India with over $8 billion invested over the last few years. The FoF is a great example of how the two nations can collaborate. This is only a precursor of things to come as India’s digital might from startups can work with the industrial might of Japan to help the two nations forge a deeper business partnership,” he told a news portal. This move is a part of a collaborative venture by the two countries floated in October 2018, where India and Japan agreed to work extensively in the areas of business, human resources and technology. This fund is already creating great interest. In fact, reports have suggested that Japanese investors like Mizuho Bank, Development Bank of Japan, Nippon Life, and Suzuki are all set to sign letters of intent. Last year, Japan, in association with India’s Nasscom and IIM Bangalore had launched the Indo-Japan Startup Hub in Bengaluru to support Japanese companies and their local operations. Its intent was to bridge the gap between Indian and Japanese startup ecosystems and enable meaningful synergies to promote joint innovation in both economies. The Hub was conceptualised as part of a joint statement signed between the Ministry of Economy, Trade, & Industry (Japan) and Ministry of Commerce & Industry (India) on 1 May 2018.","excerpt":"In a move to accelerate tech development in the continent, India and Japan are jointly launching a new $187 million (₹1,298 crore) fund-of-fund to invest in over 100 tech startups. Reportedly, this fund will focus on ventures related to artificial intelligence and the internet of things. Of the total amount, ¥15 billion or $140 million […]","categories":["AI News"],"tags":[],"author_name":"Prajakta Hebbar","publish_date":"2019-06-26T07:42:35","publication_year":"2019","word_count":317,"keywords":["API","funding","artificial intelligence","programming_languages:R","AI","innovation","Git","R","startup"],"extracted_tech_keywords":["AI","artificial intelligence","R","Git","API","innovation","startup","funding","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/india-japan-float-187-million-to-back-indian-startups-in-ai-iot\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":37382,"title":"The 3 Valuable Qualities Every Good Data Scientist Must Possess","content":"Data science is one of the most impactful areas of the post-millennial world. If the last decade gave us a glimpse of what data can help achieve, the coming decades will prove this to be the tip of the iceberg. Becoming a data scientist requires multidisciplinary skills ranging from coding, programming, data wrangling, knowledge of machine learning, artificial intelligence, deep learning, analytics tools, statistical algorithms and techniques to data visualisation, domain knowledge and soft skills. Analytics India Magazine interacted with Manoj Sharma, Director-Human Resources, NetApp India who shared some interesting insights on what it takes to be a good data scientist. Sharma believes that an aspiring data scientist must have a data-intuitive mind and passion to take raw ideas and bleeding edge technology from the lab all the way to product and into the hands of customers to significantly improve their use of data and information. “There is a lot to learn in this space and it is imperative that one builds on his\/her knowledge continuously to draw more accurate conclusions from the dataset,” he said. The Technical And Non-Technical Skillset For A Good Data Scientist Sharma shares that while it is essential to have subject matter expertise, adding certain non-technical skills can augment a candidate’ ability to be an effective data scientist. It is essential to have a data-driven problem-solving approach, intellectual curiosity, business acumen, project management, collaboration, ability to simplify complex concepts, influencing skills, attention to detail are some of the non-technical aspects of a data scientist role. “A ‘hacker mindset’ to remain relevant in the future as technology evolves is another important skill that we should scout for in a data scientist,” he shared. He further shared that the basic qualification to explore data science roles is an engineering degree or a bachelor’s degree in computer science or mathematics. “PhD or master’s degree is also highly desirable but that depends on the seniority of the role the organisation is hiring for, “ he shared. Recognising the need for talent in the space, more courses are now being offered by various institutes on data mining, programming, analytics etc. Experience using machine learning techniques\/algorithms in predictive modelling and analysis and analytics techniques such as Regression, Classification, Clustering, Markov Chains and Time Series is preferred for data science roles. “Even though it is required for data scientists to know these subjects, sometimes other professionals such as biologists (think bioinformatics), geographers, physicists prove to be good at this role by virtue of their experience working with datasets in a professional setting. Another unconventional category is that of self-learners who teach themselves to code and are quick to catch up on tech trends,” he said. The Best Learning Curve For Data Scientists Sharma has an interesting insight on what could be the best learning curve for data scientists. A precursor to this would be self-evaluation – asking questions such as ‘Am I passionate about number crunching? ‘or ‘Do I have the appetite to learn new technologies as they surface?’. “For those who are good with numbers and have a flair for technology, data science offers a wide scope,” he shares. For freshers, it may bode well to work on live projects, either on their own or with companies that offer them this opportunity. One may start with a simple idea that is exciting to him\/her and in the process demonstrate the ability to solve a problem and gather insights from data. He stresses on the need to constantly re-skill. “MOOCs or Massive Open Online Courses being offered by world’s leading universities are easy, yet, efficient path to learn new skills. With data science gaining unprecedented traction, there is immense competition in this field. Data is dynamic and diverse. Therefore, the more the openness to learn and experiment, the better are your chances to succeed as a data scientist,” he said. The Role Of Industry Mentors To Boost Data Scientists Sharma strongly believes that data scientists can bridge the gap between theory and practical applications of data science by seeking mentors either at their workplaces or universities. “Mentorship programs help students\/budding data scientists spot and acknowledge their mistakes and course-correct. A good mentor helps one bridge knowledge gaps, cross over and learn new skills. Mentors also provide constructive criticism, which can help mentees upskill and grow,” points out Sharma. For instance, NetApp has a University Research program through which they engage with research professors from reputed institutes to further their research in areas of mutual interest and in this process of collaboration, our engineers learn new areas of technology through association with the experts from academia.","excerpt":"Data science is one of the most impactful areas of the post-millennial world. If the last decade gave us a glimpse of what data can help achieve, the coming decades will prove this to be the tip of the iceberg. Becoming a data scientist requires multidisciplinary skills ranging from coding, programming, data wrangling, knowledge of […]","categories":["AI Features"],"tags":["Interviews and Discussions"],"author_name":"Srishti Deoras","publish_date":"2019-04-08T04:51:35","publication_year":"2019","word_count":763,"keywords":["data science","Go","machine learning","artificial intelligence","AI","data-driven","deep learning","analytics","GAN","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","deep learning","data science","analytics","R","Go","GAN","data-driven"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/the-3-valuable-qualities-every-good-data-scientist-must-possess\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10046280,"title":"OpenAI’s Breakthrough Model Gets ‘CLIP’ped Thanks To Distillation Technique","content":"OpenAI has garnered a reputation for dishing out state-of-the-art models that can be immediately commercialised through their private beta releases via in-house APIs. After the super successful GPT-3 release last year, the Microsoft partnered company has released CLIP, a neural network, to efficiently learn visual concepts from natural language supervision. All one has to do is simply provide the names of the visual categories to be recognised, similar to the “zero-shot” capabilities of GPT-2 and GPT-3, and apply the CLIP model to any visual classification benchmark. Both CLIP and DALL.E models of OpenAI are products of the ML community’s relentless efforts to combine the advantages of both vision and language models. OpenAI’s co-founder Ilya Sutskever, too, has stressed their importance going forward. With CLIP, the company tried to address one of the most pressing questions that still bother the community: “Are these benchmark smashing, “expensive” data devouring models restricted to big labs?” Smaller organisations and individual researchers would like to get their hands on these models, experiment, and come up with innovations of their own. Though the private beta release and APIs solve this problem to an extent, there are still debates around this. Traditionally, vision models have been trained on manually labelled datasets that are expensive to construct and only provide supervision for a limited number of predetermined visual concepts. For instance, the popular ImageNet dataset, according to OpenAI, required over 25,000 workers to annotate 14 million images for 22,000 object categories. Whereas CLIP showed that it can learn from the text–image pairs already publicly available on the internet. In this way, using the CLIP model mitigated the need for expensive large labelled datasets. Though it takes only a few minutes to train models like ResNet, the substantial computational complexity and massive storage requirements make it a great challenge to deploy them in real-time applications or on edge (for example, smartphones). Now, a team at PicCollage has come up with an even more compact version of CLIP. The image-editing app maker has recently claimed to make a lighter version of OpenAI’s famed CLIP model and even run it effectively on iOS. To do this, the team used model distillation to reduce the size of the CLIP model (the ViT model) and got promising results. “Given the magnitude of the dataset and compute required, it seemed like a daunting task, but we wanted to give it a shot anyway,” wrote the team in their blog post. A Brief Overview Of CLIP (Source: OpenAI) CLIP (Contrastive Language–Image Pre-training) builds on a large body of work on zero-shot transfer, natural language supervision, and multimodal learning. Using CLIP, OpenAI demonstrated that scaling a simple pre-training task is sufficient to achieve competitive zero-shot performance on a great variety of image classification datasets. As illustrated above, the CLIP model pre-trains an image encoder and a text encoder to predict which images were paired with which texts in OpenAI’s dataset. This behaviour is then used to turn CLIP into a zero-shot classifier. Finally, we convert all of a dataset’s classes into captions such as “a photo of a dog” and predict the class of the caption CLIP estimates best pairs with a given image. Clipping For Edge According to PicCollage, the objective of their experiment was to use the distillation paradigm to reduce the size of CLIP and explore the possibility of deploying it on edge. The sizes of the original and the distilled models are 350MB and 48MB (24MB) with FP32 (FP16) precision, respectively. “Distilled models are converted to CoreML format to run on iOS and observed negligible difference between the search results of FP16 and FP32 versions,” wrote Vinay Sisodia, ML Engineer at PicCollage. Teacher-student framework for distillation via Gou et al., Distillation techniques were designed to reduce the complexity of the networks by targeting depth and width. Typical deeper and wider neural networks transfer knowledge to shallower and thinner neural networks, as illustrated above in the case of a teacher-student framework. The student network is usually a simplified version of a teacher network with fewer layers and fewer channels in each layer. The smaller version, while preserving the structure, operates more efficiently. For their experiments, Sisodia and his team at PicCollage performed model distillation on ViT (visual transformer) model that powers CLIP. In the student model, the width and layers were reduced by a factor of two. The team first began with a dataset of ~200,000 images and gradually increased the size to more than 800,000 images; the original CLIP had 400 million images! To check the performance of distilled CLIP, the team used the COCO testing dataset and checked the top 20 results for each search term. The result was an iPhone-ready CLIP model (<50MB) capable of showing results relevant to the query. Despite showing promising results, the team laments that the mini-CLIP version still falls short in few areas when compared to the original, more powerful model: Poor performance when it comes to colour based searches.Multiple representation of a search query is superior in the original model.Queries outside the training set can trick the distilled model. It is a bit of a stretch to even think of outclassing the performance of models such as CLIP, which has been trained by a well-funded organisation with ample resources. The aforementioned tweaks, such as the ones with model distillation, do look encouraging for the ML community as a whole. Know more about distilled CLIP here.","excerpt":"Using CLIP, OpenAI demonstrated that scaling a simple pre-training task is sufficient to achieve competitive zero-shot performance.","categories":["AI Trends"],"tags":["open source ios"],"author_name":"Ram Sagar","publish_date":"2021-08-19T11:00:00","publication_year":"2021","word_count":901,"keywords":["Go","API","OpenAI","AI","neural network","ML","open source ios","RAG","GPT","Aim","R"],"extracted_tech_keywords":["AI","ML","neural network","OpenAI","Aim","RAG","R","Go","API","GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/openai-clip-model-visual-transformer-distillation-techniques\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":29968,"title":"The Truth About AI And Businesses In India: It’s A Win-Win-Win Situation","content":"The hype around artificial intelligence is very real in India with the country building deep tech pockets in robotics and deep learning areas. Research centres are rife in Bengaluru and Hyderabad where these solutions are being applied in the local economy, businesses and enterprises. Even with this progress, many organisations still pose the question: How can businesses benefit from AI. According to a study by Accenture, organisations which incorporated AI into their business models predicted that their revenue would rise by 38% by 2035. AI has become business-friendly because it simplifies communication with clients by assessing user behaviour patterns. AI systems are also being used to control production processes and interconnected devices. At Cypher 2018, Dakshinamurthy V Kolluru Of INSOFE highlighted several use cases that indicated the paradigm shift that happened over the last five years with diminishing investments in AI, but output and performance surging, thereby resulting in a higher return on investment (ROI). So pegged as the future of growth, AI has become a productivity enhancer, which has the potential to Reverse the falling profit growth through automation Augment human labour A recent research indicates how artificial intelligence has the potential to double annual economic growth, change the nature of work between people and machines and push innovations. By 2035, AI is expected to double the annual economic growth rate. The study further indicated that manufacturing, BFSI and information and communication sectors would see the highest growth rate. Also, with AI being integrated into the economic processes, it means the bottom line revenue gets boosted. The three key methods are: Intelligent automation\/virtual workforce Labour and capital augmentation Technology innovations AI As A Productivity Solution Automating Back Office Operations: Leading in the field of white-collar and business process automation would indeed force India to compete on technical developments, rather than price. Automating office and IT work is an opportunity for all businesses and markets. For this reason, when India proves its dominance in automating and augmenting its BPO and IT services, it will be competing with startups and with large consulting firms on many fronts. BPO Industry Can Capitalise On The AI Boom: Equifax identified scenarios where Indian sectors can leverage AI. According to their study, the BPO industry in India may find an opportunity in data cleaning and tagging in massive datasets, which will eventually be used to train and error-correct AI. Cheap skilled IT labor can potentially enable this opportunity. By leveraging BPO and back-office expertise, India’s IT services firms could potentially develop the world’s best office automation and BPO AI applications. This would allow India to develop a higher margin product industry, in addition to their wide low-cost services sector The key for Indian BPO, IT, and tech firms will be to pick the domains of focus where their expertise is strongest as compared to the rest of the countries, and which domain provides maximum economic opportunity. Partnering With Third-Party Vendors: By building an AI roadmap, businesses can plan to broaden their use cases. Business leaders and strategic planners from across the business have sufficient grasp of AI to effectively transform existing business plans and to define key decision points and to guide appropriate investment decisions. This is a growing trend in India where leading enterprises partner with third-party providers who can work on specific use cases through bleeding edge technology. Companies are also pushing for AI integration and upskilling employees to counter the lack of expertise in AI technologies. While there are interim roadblocks in implementation, companies are deploying a winning mindset towards AI and have leveraged a continuous, iterative approach. Outlook However, for AI to become a critical enabler in India, it must go beyond automation to truly harness the potential of self-learning machines. The potential benefits of AI can be considerably greater than the past impact of automation. According to Bosch, more than $2 billion of additional revenues and savings from the widespread use of intelligent systems and machines by 2020.","excerpt":"The hype around artificial intelligence is very real in India with the country building deep tech pockets in robotics and deep learning areas. Research centres are rife in Bengaluru and Hyderabad where these solutions are being applied in the local economy, businesses and enterprises. Even with this progress, many organisations still pose the question: How […]","categories":["AI Features"],"tags":[],"author_name":"Richa Bhatia","publish_date":"2018-11-06T11:37:36","publication_year":"2018","word_count":658,"keywords":["Go","API","TPU","artificial intelligence","AI","RAG","deep learning","ViT","GAN","R"],"extracted_tech_keywords":["AI","artificial intelligence","deep learning","RAG","TPU","R","Go","API","GAN","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/the-truth-about-ai-and-businesses-in-india-its-a-win-win-win-situation\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":6,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":46753,"title":"5 Decades Of Machine Learning Unfairness: The Eerie Way In Which Prejudice Crept Into Algorithms","content":"The notion of machine learning fairness can be bottled down to the following facets of data pre-processing: Demographic parity Equal opportunity Equalised odds Disparate impact To remove prejudices from a model might not be an impossible task but can any application which serves humans be immune to the human itself. And, even if the human element is considered, how much of it is too much? With great ML deployment comes great responsibility. The spike in interest in test fairness in the 1960s arose during a time of social and political upheaval, with quantitative definitions catalysed in part by U.S. federal anti-discrimination legislation in the domains of education and employment. Here’s a look at a few significant events that exposed the unfairness of the system: 1960s Concerned with the fairness of tests for black and white students, T. Anne Cleary defined a quantitative measure of test bias for the first time, cast in terms of a formal model for predicting educational outcomes from test scores. While Cleary’s focus was on education, her contemporary Robert Guion was concerned with unfair discrimination in employment. Arguing for the importance of quantitative analyses in 1966, he wrote that: “Illegal discrimination is largely an ethical matter, but the fulfillment of ethical responsibility begins with technical competence”, and defined unfair discrimination to be “when persons with equal probabilities of success on the job have unequal probabilities of being hired for the job.” Responding to these concerns, the Association of Black Psychologists formed in 1969. 1970s The advent of the 70s saw researchers like Thorndike professing viewpoints that judgment on test-fairness must rest on the inferences that are made from the test rather than on a comparison of mean scores in the two populations. Thorndike was quoted saying that one must then focus attention on fair use of the test scores, rather than on the scores themselves. As an alternative to Cleary, Thorndike proposed that the ratio of predicted positives to ground truth positives be equal for each group. Using confusion matrix terminology, this is equivalent to requiring that the ratio (T P+F P)\/(T P+ F N) be equal for each subgroup. In his 1976 book Computer Power and Human Reason, Artificial Intelligence pioneer Joseph Weizenbaum suggested that bias could arise both from the data used in a program, but also from the way a program is coded. 1980s With the start of the 1980s came renewed public debate about the existence of racial differences in general intelligence, and the implications for fair testing. In 1981, with no public debate, the United States Employment Services implemented a score-adjustment strategy that was sometimes called “race-norming”. As many as 60 women and ethnic minorities denied entry to St. George’s Hospital Medical School from 1982 to 1986, based on the implementation of a new computer-guidance assessment system that denied entry to women and men with “foreign-sounding names” based on historical trends in admissions. 1990s Batya Friedman and the philosopher Helen Nissenbaum (1996) discussed bias concerns in the use of computer systems for tasks as diverse as scheduling, employment matching, flight routing, and automated legal aid for immigration. Friedman and Nissenbaum (1996) also examined the history of the algorithm for the National Resident Match Program, which matches medical residents to hospitals throughout the United States. The algorithm’s seemingly equitable assignment rules favoured hospital preferences over resident preferences and single residents over married residents. 2000- Present Amazon’s Flawed Recruiter created 500 computer models focused on specific job functions and locations. They taught each to recognise some 50,000 terms that showed up on past candidates’ resumes. Instead, the technology favoured candidates who described themselves using verbs more commonly found on male engineers’ resumes, such as “executed” and “captured,” one person said. Microsoft’s twitter-based AI chatbot Tay, despite being stress-tested “under a variety of conditions, specifically to make interacting with Tay a positive experience,” learned anti-Semitic and racist behaviour due to the efforts of a specific group of individuals. By being repeatedly exposed to similar types of discriminatory content, Tay acquired numerous discriminatory biases. A year after Tay was shut down, Microsoft launched another chatbot known as Zo, which faced similar public backlash after exhibiting anti-Islamic learned biases. However, due to bias avoidance measures, Zo proved to be resistant to exhibiting discriminatory biases. To avoid exhibiting bias, Zo included filters for rejecting discussion about topics that referenced religion or politics. Whereas, in 2015, Google came under fire after its new Photos application categorised photos of Jacky Alciné and his girlfriend as “gorillas.” Google attempted to fix the algorithm but ultimately removed the gorilla label altogether. What Next? In a study done on the fairness of the machine learning algorithms over the past 50 years, Google researchers concluded the following: In the 1960s and 1970s, the fascination with determining fairness ultimately died out as the work became less tied to the practical needs of society, politics and the law, and more tied to unambiguously identifying fairness. The rise of interest in fairness today has corresponded with the public interest in the use of machine learning in criminal sentencing and predictive policing. Careful attention should be paid to legal and public concerns about fairness. The experiences of the test fairness field suggest that in the coming years, courts may start ruling on the fairness of ML models. If technical definitions of fairness stray too far from the public’s perceptions of fairness, then the political will to use scientific contributions in advance of public policy may be difficult to obtain. Machine learning fairness is an active ongoing research in many big tech companies. For instance, in September 2018, Google debuted its What-If Tool. It allows users to generate visualisations that explore the impact of algorithmic tweaks and adjustments to bias in their datasets on the fly. The researchers also tackled this problem of bias in labelling, by providing a mathematical formulation on how biases arise in labelling and how can this be mitigated. The proposed solutions still don’t scratch the surface and right now bias is almost inevitable. The bias usually gets the flak, be it in case of gender, race and culture, the problem often seems to be the over-representation of certain groups. The problem goes back again to the way data is collected. Speaking at the recently concluded Analytics India Magazine’s conference, TheMath Company stressed about the need for data ethnographers in the current scenario. The datasets have to be prepared or have to be collected from some source which is collateral of human interactions. The collected data will be cleaned and appended with classes. These sub-groups no matter how unbiased they were planned to be, there still lies an underwritten, underlying bias.","excerpt":"The notion of machine learning fairness can be bottled down to the following facets of data pre-processing: Demographic parity Equal opportunity Equalised odds Disparate impact To remove prejudices from a model might not be an impossible task but can any application which serves humans be immune to the human itself. And, even if the human […]","categories":["AI Trends"],"tags":["bias","FAIRNESS","Machine Learning","Machine Learning Algorithms"],"author_name":"Ram Sagar","publish_date":"2019-10-03T15:05:33","publication_year":"2019","word_count":1107,"keywords":["Machine Learning Algorithms","Go","artificial intelligence","machine learning","AI","R","ML","Machine Learning","BERT","Ray","FAIRNESS","analytics","ViT","bias"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","analytics","Ray","R","Go","BERT","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/machine-learning-fairness-algorithm-bias\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":39171,"title":"11 Most Commonly Asked NLP Interview Questions For Beginners","content":"Natural Language Processing(NLP) is one of the most popular domains in ML. It is a collection of methods to make the machine learn and understand the language of humans. The wide adoption of its applications has made it a hot skill amongst top companies. Here are a few frequently asked NLP questions that would give an introductory idea of this domain: 1)How can machines make meaning out of language? Popular NLP procedure is to use stemming and lemmatization methods along with the parts of speech tagging. The way humans use language varies with context and everything can’t be taken too literally. Stemming approximates a word to its root i.e identifying the original word by removing the plurals or the verb forms. For example, ‘rides’ and ‘riding’ both denote ‘ride’. So, if a sentence contains more than one form of ride, then all those will be marked to be identified as the same word. Google used stemming back in 2003 for its search engine queries. Whereas, lemmatization is performed to correctly identify the context in which a particular word is used. To do this, the sentences adjacent to the one under consideration are scanned too. In the above example, riding is the lemma of the word ride. Removing stop words like a, an, the from a sentence can also enable the machine to get to the ground truth faster. 2)What does a NLP pipeline consist of? Any typical NLP problem can be proceeded as follows: Text gathering(web scraping or available datasets) Text cleaning(stemming, lemmatization) Feature generation(Bag of words) Embedding and sentence representation(word2vec) Training the model by leveraging neural nets or regression techniques Model evaluation Making adjustments to the model Deployment of the model. 3)What is Parsing in the context of NLP? Parsing a document means to working out the grammatical structure of sentences, for instance, which groups of words go together (as “phrases”) and which words are the subject or object of a verb. Probabilistic parsers use knowledge of language gained from hand-parsed sentences to try to produce the most likely analysis of new sentences. 4)What is Named Entity Recognition(NER)? Named entity recognition is a method to divide a sentence into categories. Neil Armstong of the US had landed on the  moon in 1969 will be categorized as Neil Armstong- name;The US – country;1969 – time(temporal token). The idea behind NER is to enable the machine to pull out entities like people, places, things, locations, monetary figures, and more. 5) Where can NER be used? Scanning documents for classification, customer support(chatbots, understanding feedback) and entity identification in molecular biology(names of genes etc.,) 6) How is feature extraction done in NLP The features of a sentence can be used to conduct sentiment analysis or document classification. For example if a product review on Amazon or a movie review on IMDB consists of certain words like ‘good’, ‘great’ more, it could then be concluded\/classified that a particular review is positive. Bag of words is a popular model which is used for feature generation. A sentence can be tokenized and then a group or category can be formed out of these individual words, which further explored or exploited for certain characteristics(number of times a certain word appears etc). 7) Name some popular models other than Bag of words? Latent semantic indexing, word2vec. 8) Explain briefly about word2vec Word2Vec  embeds words in a lower-dimensional vector space using a shallow neural network. The result is a set of word-vectors where vectors close together in vector space have similar meanings based on context, and word-vectors distant to each other have differing meanings. For example, apple and orange would be close together and apple and gravity would be relatively far. There are two versions of this model based on skip-grams (SG) and continuous-bag-of-words (CBOW). 9) What is Latent Semantic Indexing? Latent semantic indexing is a mathematical technique to extract information from unstructured data. It is based on the principle that words used in the same context carry the same meaning. In order to identify relevant (concept) components, or in other words, aims to group words into classes that represent concepts or semantic fields, this method applies Singular Value Decomposition  to the Term-Document matrix. As the name suggests this matrix consists of words as rows and document as columns. LSI is computation heavy when compared to other models. But it equips an NLP model with better contextual awareness, which is relatively closer to NLU 10) What are the metrics used to test an NLP model? Accuracy, Precision, Recall and F1. Accuracy is the usual ratio of the prediction to the desired output. But going just be accuracy is naive considering the complexities involved. Whereas, precision and recall consider false positive and false negative making them more reliable metrics. And, F1 is the sweet spot between precision and recall. 11) What are some popular Python libraries used for NLP Stanford’s CoreNLP, SpaCy , NLTK and TextBlob. There is more to explore about NLP. Advancements like Google’s BERT, where a transformer network is preferred to CNN or RNN. A Transformer network applies self-attention mechanism which scans through every word and appends attention scores(weights) to the words. For example, homonyms will be given higher scores for their ambiguity and these weights are used to calculate weighted average which gives a different representation of the same word. Know more about how to build an NLP model here","excerpt":"Natural Language Processing(NLP) is one of the most popular domains in ML. It is a collection of methods to make the machine learn and understand the language of humans. The wide adoption of its applications has made it a hot skill amongst top companies. Here are a few frequently asked NLP questions that would give […]","categories":["AI Trends"],"tags":["data science interview","Natural Language Processing","NLP","NLTK","Speech Analytics"],"author_name":"Ram Sagar","publish_date":"2019-05-15T07:00:22","publication_year":"2019","word_count":894,"keywords":["spaCy","data science interview","Speech Analytics","AI","neural network","Natural Language Processing","ML","chatbots","sentiment analysis","RAG","NLP","Aim","NLTK"],"extracted_tech_keywords":["AI","ML","neural network","NLP","Aim","spaCy","NLTK","RAG","chatbots","sentiment analysis"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/11-most-commonly-asked-nlp-interview-questions-for-beginners\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10098552,"title":"Why is Microsoft Distancing Itself from OpenAI?","content":"Looks like the honeymoon period for Microsoft and OpenAI is coming to an end. Microsoft seems to be slowly moving away from OpenAI, prioritising its Azure cloud above all else and taking measures to protect its own interests. With a focus on being an enterprise-friendly entity, Microsoft’s growth relies heavily on maintaining the trust of its customers. Lately, OpenAI hasn’t been in good books putting customers’ data at risk with ChatGPT. Large tech companies like Apple, Spotify, Wells Fargo, Samsung, JP Morgan ,Verizon had wayback ditched ChatGPT and banned  it for their employees from using it. Companies like Apple have valid concerns about its employees inadvertently sharing sensitive project details through the system. This data might potentially be viewed by OpenAI moderators. Studies indicate that certain language models can have training data extracted via their chat interface. Apple recently developed its own internal chatbot—nicknamed “Apple GPT” by its employees. Recently, OpenAI openly introduced GPTbot​, an automated website crawler to collect publicly accessible data to train AI models. This also didn’t help their case either. Microsoft, which relies on OpenAI’s foundational models, to service its customers is taking the hit. Microsoft knows that enterprises and their employees need ChatGPT considering the practical value of ChatGPT in enhancing coding and idea generation. Employees have found themselves grappling with a dilemma – whether to utilise ChatGPT or not. Even though their organisations had prohibited its use, some still went ahead and employed it secretly. Just recently, Reuters published a report revealing that numerous workers across the U.S. are increasingly relying on ChatGPT for basic tasks, despite concerns that have prompted companies like Microsoft and Google to limit its usage. To overcome trust issues among the enterprises Microsoft announced Azure ChatGPT recently which is being marketed as ChatGPT for enterprises where they don’t need to worry about their data. Microsoft, while announcing it, surprisingly accepted shortcomings of OpenAI’s ChatGPT. “ChatGPT risks exposing confidential intellectual property. One option is to block corporate access to ChatGPT, but people always find workarounds,” said Microsoft. Further it said: “ChatGPT on Azure solution accelerator is our enterprise option. This solution provides a similar user experience to ChatGPT but offered as your private ChatGPT.” Microsoft also made it clear that customers’ data is “fully isolated from those operated by OpenAI”. Is data not secure with OpenAI and they are exploiting it to train GPT-4? When it released the ChatGPT API in March, the company said OpenAI retains API data for 30 days but no longer uses data sent via the API to improve their models. Here with Azure ChatGPT what Microsoft  is doing is that it is taking the OpenAI’s model and putting it behind an endpoint specific to the user’s Azure account. Businesses have been  long interested in GPT-4 and so are asking for private endpoints. Amazon is doing the same with Bedrock. Azure’s the only hope Currently, Microsoft is on high with its Azure performance which saw 26% revenue growth due to generative AI initiatives it took by partnering with OpenAI and now it doesn’t want it to go down just because OpenAI as a brand isn’t trustworthy. When Microsoft partnered with OpenAI, Microsoft deployed OpenAI technology through API and the Azure OpenAI Service—enabling enterprise and developers to build on top of GPT, DALL·E, and Codex. They also worked together to build OpenAI’s technology into apps like GitHub Copilot and Microsoft Designer. ​​However, one of the common doubts that rose among customers was: If data is submitted to the Azure OpenAI Service, does the data always remain within Microsoft Azure or is it passed to OpenAI at any time? Microsoft claimed that the data submitted to the Azure OpenAI Service remains within Microsoft Azure and is not passed to OpenAI for model predictions and Azure has sole control and governance of the data and OpenAI. Microsoft owns enough of OpenAI that their endgame goal of putting GPT like features into Azure and Office365 for enterprise customers is what we’re likely to see happen. In one of the interesting conversations on HN, a user said OpenAI targets private consumers while Microsoft focuses on enterprise. He also gave an example of his own company. “I can use my own organisation as an example. We’re an investment bank that does green energy within the EU. We would absolutely use GPT if it was legal, but it isn’t, and it likely never will be considering their finance model is partly to steal as much data as they can.” Microsoft is concerned about the current negative perception of OpenAI’s data security for enterprises. Through the introduction of Azure ChatGPT, they are striving to rebuild trust among enterprises and attract more customers. They don’t want to associate themselves with OpenAI’s name on Azure, that’s why they have put Azure’s name at forefront. Azure ChatGPT is just part of Microsoft’s overall strategy for total IT domination in enterprise Also, it isn’t the first time Microsoft is at crossroads with OpenAI. Earlier WSJ had reported that people within Microsoft have complained about diminished spending on its in-house AI and that OpenAI doesn’t allow most Microsoft employees access to the inner workings of its technology.","excerpt":"Microsoft said ChatGPT risks exposing confidential intellectual property","categories":["Global Tech"],"tags":["Microsoft"],"author_name":"Siddharth Jindal","publish_date":"2023-08-14T16:13:29","publication_year":"2023","word_count":858,"keywords":["Go","ChatGPT","OpenAI","AI","R","Git","Aim","generative AI","Rust","Azure","Microsoft"],"extracted_tech_keywords":["AI","generative AI","ChatGPT","OpenAI","Aim","Azure","R","Go","Rust","Git"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/why-is-microsoft-distancing-itself-from-openai\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10063974,"title":"Can a language model acquire knowledge by simply reading new data?","content":"Training and fine-tuning language models to new knowledge is a tedious process that eats up both time and resources. But what if a language model could acquire new knowledge by just reading new data and memorising it? Google’s new research paper titled, ‘Memorizing Transformers,’ released in ICLR 2022, discusses just how this can be done. The paper notes that attention can be used as a kind of fast learning for long sequences. A model can use attention to memorise facts by storing them as key or value pairs in long-term memory. The model can easily access this information by creating a query. Adding memory: Paper Increasing attention context When performing tasks, transformer performance depends upon the context of attention. This length of the context of attention is usually short, but the paper showed that this can be increased by using approximate k-nearest-neighbour (kNN) lookup. This method is usually used for retrieving information. The kNN lookup algorithm can be scaled up with its variations such as ScaNN and Faiss. This study looks at increasing the context of attention from a different perspective. While traditional long sequence attention models perform averaging of tokens at long distances, kNN lookup is able to retrieve exact values even from the distant context. Also, gradients aren’t backpropagated into the external memory. Instead, the proposed method reuses keys and values previously computed on prior training steps. To push scalability, backpropagating gradients into the external memory is essential but requires computing all the keys and values with the parameters on every single training step. Reusing keys and values, on the other hand, cuts down on the amount of computation for large memories. This method maintains a reasonable step time while being able to easily scale external memory to long sequences of up to 131k or 262k tokens on a single TPU device. Extending transformers: Paper Methodology The input text is first tokenised, and then the tokens are embedded into vector spaces. The vector space embeddings are passed through a series of layers of transformers, each of which performs dense self-attention followed by a feed-forward network, or FFN. As the language model is a decoder-only type, a causal attention mask is used, and token embeddings of the previous layer predict the next token. Then, long documents are split into subsequences of 512 tokens, and each subsequence is used as an input for a training step. When normally, subsequences are shuffled together, in this case, long documents were fed into the transformer sequentially, as is done with the Transformer-XL. The kNN-augmented attention layer, which is near one of the transformer layers near the top of the stack, combines two types of attention: standard dense self-attention on the local context and an approximate k-nearest-neighbour search into the external memory. Using approximate kNN search instead of exact kNN search speeds the computational speed of the model considerably. Datasets Average perplexities of each model: Paper The research used datasets including a corpus of papers from arXiv Math, open-source code files from GitHub, mathematical theories from Isabelle, a large collection of documents from C4 and English-language books from PG-19. Findings By presenting a simple extension to the transformer, known as kNN-augmented attention, the research found that it could increase the length of the context in a language model. The transformer demonstrated improved perplexity over the baseline for all architectures and data models. Aside from this, external memory is also beneficial even if the transformer is scaled from 200M to 8B parameters. The study concludes that the transformer does not need to be pre-trained from scratch. Rather, memory can be added to an existing pre-trained model and fine-tuned to reap more. kNN retrieval is also able to scale up to much bigger memory sizes and can use their code repositories and huge knowledge bases. Criticism Context within datasets As promising as this alternative sounds, author and data scientist Minhaaj Rehman sparked a discussion on LinkedIn revolving around how effective the method would be in reality. Rehman noted that it was the “nature of the language itself that was elusive, and not how inference servers process the data.” Another data scientist suggested that language models could be improved by the RETRO transformer, or Retrieval Enhanced Transformer, so that models can retrieve data from a 2 trillion token database. However, according to Rehman, this was a misdirection of the issue. Having faster inference did not compensate for the model’s inability to correctly understand contextual information. According to a paper published in 2020 titled, ‘Context pre-modeling: an empirical analysis for classification based user-centric context-aware predictive modelling,’ even a slightly different context can lead to a hugely different outcome even with a simple dataset. The context of a dataset can influence it far more than the predictive power of a model. To create a model that is aware of context, context pre-modelling is key.","excerpt":"By presenting a simple extension to the transformer, known as kNN-augmented attention, the research found that it could increase the length of the context in a language model.","categories":["AI Features"],"tags":["Mergers and Acquisitions"],"author_name":"Poulomi Chatterjee","publish_date":"2022-03-31T16:00:00","publication_year":"2022","word_count":805,"keywords":["Go","TPU","AI","ai_frameworks:Transformers","Transformers","Scala","RAG","Git","GitHub","Mergers and Acquisitions","R"],"extracted_tech_keywords":["AI","Transformers","RAG","TPU","R","Go","Scala","Git","GitHub","ai_frameworks:Transformers"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/can-a-language-model-acquire-new-knowledge-by-simply-reading-new-data\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10044791,"title":"How TensorFlow-Ranking Evolved In The Last Three Years","content":"Three years ago, Google introduced TF-Ranking, an open-source TensorFlow-based library for developing scalable neural learning-to-rank (LTR) models. Compared to standard classification models that classify one item at a time, LTR models receive an entire list of items as input and learn an ordering that maximises the utility of the entire list. Over the years, search, and recommendation systems have been the most popular applications of LTR models. Since its launch in 2018, TF-Ranking has been applied in diverse domains beyond search, including e-commerce, smart city planning, and SAT solvers. For instance, the goal of LTR is to learn a function f () that takes as an input a list of items (songs, documents, products, movies, etc.) and outputs the list of things in the optimal order (descending order of relevance). As shown in the image below, the green shade indicates the item relevance level, and the red item marked with ‘X’ indicates non-relevant. (Source: Google) Introducing the latest version of TF-Ranking In May this year, Google launched the latest version of TF-Ranking that enables full support for natively building LTR models using Keras, a high-level API of TensorFlow 2. Its native Keras ranking model consists of a flexible ModelBuilder, a DatasetBuilder to set up training data, and a Pipeline to train the model with the provided dataset. Check out the source code on GitHub. These components help build a customised LTR model easier than ever and facilitate rapid exploration of new model structures for research and production. TF-Ranking now works with RaggedTensor. Also, Google’s latest release, which incorporates the Orbit training library, is powered by two and half years of neural LTR research. Here is an image showcasing the key improvements in the latest version of TF-Ranking. The blue modules are provided by TF-Ranking, and the green modules are customisable. (Source: Google) LTR with TFR-BERT Of late, pre-trained language models such as BERT have achieved SOTA performance on various language understanding tasks. To capture the expressiveness of these models, Google — in its latest blog post — demonstrated TF-Ranking implementing a novel TFR-BERT architecture that combines BERT with the power of LTR (learning-to-rank) to optimise the ordering of list inputs. For example, consider a query and a list of ‘n’ documents to rank in response to this query. Instead of learning a BERT representation for each <query, document> pair, LTR models apply a ranking loss to jointly learn a BERT representation that maximises the utility of the entire ranked list for the ground-truth labels. The image below illustrates this process. For this, you need to flatten a list of ‘n’ documents to rank in response to a query into a list <query, document> tuples. Next, these tuples are fed into a pre-trained language model (BERT). Finally, the collected BERT outputs for the entire document list are then fine-tuned with one of the specialised ranking losses available in TFR. “Our experience shows that this ‘TFR-BERT architecture’ delivers significant improvement in pre-trained language model performance, learning to SOTA performance for several popular ranking tasks, especially when multiple pre-trained language models are ensembled,” according to Google. (Source: Google) Interpretability and transparency of LTR models Transparency and interpretability are essential factors in deploying LTR models in ranking systems. In such scenarios, the contribution of each feature to the final ranking should be examinable and understandable to ensure transparency, accountability, and fairness of the outcomes. These things can be achieved using generalised additive models or GAMs, intrinsically interpretable ML models that are linearly composed of smooth functions of individual features. In the past, GAMs have been studied on classification and regression tasks. However, it is less clear on how to apply them in a ranking application. Last year, Google developed a neural ranking GAM — an extension of generalised additive models to ranking problems. For instance, in the image below, using a neural ranking GAM makes visible how relevance, price and distance, in the context of a given user device, contribute to the final ranking of the hotel. Neural ranking GAMs are now available as part of TensorFlow-Ranking. (Source: arXiv) Neural ranking vs gradient boosting Since its TF-Ranking launch, the team has significantly deepened the understanding of how best to leverage neural models in ranking with numerical features, instead of gradient boosted decision trees (GBDTs) such as LambdaMART, which had remained the baseline to beat in various open LTR datasets. The team culminated in a data augmented self-attentive latent cross (DASALC) model, as described in an ICLR 2021 paper, which is the first to establish parity, and in some cases statistically significant improvements of neural ranking models over strong LambdaMART baselines on open LTR datasets. ‘This achievement is made possible through a combination of techniques, including data augmentation, self-attention for modeling document interactions, neural feature transformation, listwise ranking loss, and model ensembling similar to boosting in GBDTs,” according to a Google blog post. The architecture of the DASALC model was entirely implemented using the TF-Ranking library. “We believe that the new ‘Keras-based TF-Ranking’ version will make it easier to conduct neural LTR research and deploy production-grade ranking systems,” according to Google.","excerpt":"Search and recommendation systems have been the most popular applications of LTR models.","categories":["AI Features"],"tags":["BERT","Machine Learning","Machine Learning Latest","TensorFlow library","what is tensorflow"],"author_name":"Amit Naik","publish_date":"2021-07-30T15:00:00","publication_year":"2021","word_count":846,"keywords":["Go","TPU","Keras","AI","what is tensorflow","ML","Machine Learning","Machine Learning Latest","recommendation systems","BERT","RAG","Scala","TensorFlow","R","TensorFlow library"],"extracted_tech_keywords":["AI","ML","TensorFlow","Keras","RAG","recommendation systems","TPU","R","Go","Scala"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-tensorflow-ranking-evolved-in-the-last-three-years\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10043457,"title":"AI Winter Is Coming? Four Fallacies In AI Research","content":"“Perhaps expectations are too high, and… this will eventually result in disaster. Suppose that five years from now, funding collapses miserably as autonomous vehicles fail to roll. Every startup company fails. And there’s a big backlash so that you can’t get money for anything connected with AI. Everybody hurriedly changes the names of their research projects to something else. This condition is called the AI Winter,” said  AI expert Drew McDermott in 1984. In her latest paper titled ‘Why AI is Harder Than We Think’, AI researcher and ousted Google employee Melanie Mitchell, explained how research in AI often follows a cyclic pattern: periods of rapid progress, successful commercialisation, heavy public and private investments, called AI Spring, is often followed by AI winter, characterised by waning enthusiasm, drying up of funding and jobs. Mitchell argued over-optimism among people, the media and even experts arise from fallacies in our understanding of AI and the intuitions about the nature of intelligence. She outlined four major fallacies: Narrow intelligence and general intelligence One of the most common fallacies is that narrow intelligence is on a continuum with general intelligence. Narrow intelligence refers to a machine’s ability to perform a single task extremely well. Advances made in narrow AI are often described as the first step towards general AI. For example, Deep Blue, the chess-playing computer was popularly hailed as the first step in the AI revolution, IBM’s Watson system was described as the entry to a ‘new era of computing’, and most recently, OpenAI’s GPT-3 was called a step towards general intelligence. This is commonly called the ‘first step fallacy’, a term coined by philosopher and mathematician Yehoshua Bar-Hillel. In philosopher Hubert Dreyfus’ words, this term means any improvement in our programs, no matter how trivial, is considered as ‘progress’. Like Dreyfus, Mitchell believes the ‘unexpected obstacle’ in the so-assumed continuum of AI progress has been a problem of common sense. Easy tasks and hard tasks Moravec’s paradox, named after roboticist Hans Moravec, states that it is comparatively easy to make computers demonstrate adult level performance on intelligence tests or playing games like chess, but it is impossible for them even to exhibit skills of a toddler when it comes to perception and mobility. It means that tasks that humans perform almost effortlessly, like making sense of what we see, conversing with another person, or even simply walking without bumping into obstacles, can be some of the hardest tasks to accomplish for machines. Conversely, solving puzzles and complex mathematical problems, translating text between thousands of languages are relatively easier for machines than humans. Wishful mnemonics AI is full of ‘wishful mnemonics’ said Mitchell in her paper. She referred to the terms generally associated with human intelligence being used for the evaluation of AI programs. For example, machine learning and deep learning methods are very different from learning in humans or even animals. Similarly, one of the subfields of machine learning is transfer learning which refers to transferring the knowledge they have gained to newer situations. While this capability is fundamental to humans, it is still an open problem for machines. Mitchell calls these anthropomorphic terms, shorthands. “This has led to headlines such as “New AI model exceeds human performance at question Answering”; “Computers are getting better than humans at reading”; and “Microsoft’s AI model has outperformed humans in natural-language understanding”. Given the names of these benchmark evaluations, it’s not surprising that people would draw such conclusions,” she stated. Benchmarks do not give the correct estimates of capabilities for carrying out tasks such as question-answering, reading, and natural language understanding. However, many of these benchmarks allow machines to leverage the statistical correlation to achieve high performance on a test without actually learning the skill. It is true that machines are capable of performing high-precision tasks, but they are still far from achieving general human abilities, which we associate with the benchmarks’ names. Intelligence is in the brain It is a widely held belief that intelligence is a non-physical entity and is completely encapsulated in the brain. This has also given rise to the notion that intelligence can be disembodied. The assumption is implicit in most work on AI throughout history. Meaning, researchers believe, to achieve human-level intelligence, we need to simply scale up the machines to ‘match the brain’s “computing capacity” and then develop the appropriate “software” for this brain-matching “hardware.” However, as many psychological and cognitive studies prove, human intelligence is a strongly integrated system with closely interconnected attributes such as emotions, desires, autonomy, and common sense, most of which can not be separated. Despite mounting evidence, research in AI has mostly ignored these results. Only a small number of researchers are exploring these ideas under fields like embodied AI and developmental robotics. Read the full paper here.","excerpt":"“Perhaps expectations are too high, and… this will eventually result in disaster. Suppose that five years from now, funding collapses miserably as autonomous vehicles fail to roll. Every startup company fails. And there’s a big backlash so that you can’t get money for anything connected with AI. Everybody hurriedly changes the names of their research […]","categories":["AI Features"],"tags":["AI Research"],"author_name":"Shraddha Goled","publish_date":"2021-07-14T11:00:00","publication_year":"2021","word_count":795,"keywords":["Go","API","machine learning","OpenAI","AI","AI Research","RAG","BERT","GPT","deep learning","R"],"extracted_tech_keywords":["AI","machine learning","deep learning","OpenAI","RAG","R","Go","API","BERT","GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/ai-winter-is-coming-four-fallacies-in-ai-research\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10162118,"title":"Bengaluru-Based Deep Tech Startup VuNet Raises INR 60 Crore for Global Expansion","content":"Bengaluru-based deep tech startup VuNet Systems has closed its Series B funding round, raising INR 60 crore on Thursday. Pravega Ventures led the investment, while Kotak Securities joined as a strategic investor alongside existing backers Mela Ventures, Athera Venture Partners, Dallas Venture Capital, and TVS Capital Funds. 🚀Excited to lead a $7 M investment in @vunetsystems !Specializing in #observability, VuNet shifts IT monitoring to real-time visibility of transactions & customer journeys.#venturecapital #Innovation #Investment #BFSI pic.twitter.com\/yjLwdeycuu— Pravega Ventures (@PravegaVC) January 23, 2025 VuNet is a provider of full-stack Al and big data analytics and aims to provide funding to drive the company’s global expansion and product innovation. The funding will also strengthen its presence in India and other key markets. The company plans to enhance its platform capabilities, scale sales and marketing efforts, and further integrate advanced AI technologies, including GenAI across its workflows. Expressing enthusiasm about the funding in a LinkedIn announcement, Ashwin Kumar Ramachandran, co-founder and CEO of VuNet Systems, said, “This funding empowers us to accelerate our global expansion while intensifying our focus on innovation, ensuring we deliver even greater value to our customers.” Ramachandran further highlighted the importance of their platform in the evolving business landscape. “As businesses increasingly strive to connect IT performance with measurable business outcomes, VuNet Systems’ Business Journey Observability platform has emerged as a trusted partner in driving this transformation,” he said. The platform helps financial institutions monitor IT performance linked to business outcomes. Currently, it processes over 20 billion transactions per month for leading banks and provides real-time visibility into critical payment and lending processes. By integrating advanced AI, including generative AI, VuNet aims to enhance operational transparency and business resilience. Shripal Shah, MD and CEO at Kotak Securities, explained, “Full-stack business observability has become mission-critical for financial institutions. VuNet has a proven track record in delivering actionable insights using advanced technologies like big data and machine learning.” A spokesperson from Pravega Ventures called VuNet “a platform built entirely from the ground-up by a team of brilliant engineers in India” that serves as a critical backbone for financial infrastructure. The investment comes amid a growing market opportunity. Gartner projects that the IT operations management (ITOM) market will reach $81 billion by 2027, driven by increasing complexity in digital ecosystems and the need for comprehensive system visibility.","excerpt":"VuNet is a provider of full-stack Al and big data analytics and aims to provide funding to drive the company’s global expansion and product innovation.","categories":["AI News"],"tags":["deep tech","Funding"],"author_name":"Sanjana Gupta","publish_date":"2025-01-24T10:35:10","publication_year":"2025","word_count":384,"keywords":["API","Funding","deep tech","machine learning","GenAI","AI","Git","Aim","analytics","generative AI","Rust","R"],"extracted_tech_keywords":["AI","machine learning","analytics","generative AI","GenAI","Aim","R","Rust","Git","API"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/bengaluru-based-deep-tech-startup-vunet-raises-inr-60-crore-for-global-expansion\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10099431,"title":"[Exclusive] First Indian IT Company to Bring Out ‘Made in India’ ChatGPT","content":"A few weeks ago, Tech Mahindra announced the launch of Project Indus – an Indic-based foundational model for Indian languages, which could potentially prove to be its most important project ever. Large language models (LLMs) like the GPT models by OpenAI, despite their multilingual capabilities, have been predominantly trained on English datasets, which limits their proficiency in comprehending and generating content in Indic languages. Hence, an open-source Indic LLM will be hugely beneficial for India. According to Tech Mahindra’s chief CP Gurnani, the model will be the biggest Indic LLM and could possibly cater to 25% of the world’s population. While Tech Mahindra has not revealed the cost associated with the project or when the model is expected to be launched, the aim is to build a 7-billion parameter LLM to begin with, Nikhil Malhotra, global head-Makers Lab, Tech Mahindra, told AIM. The model is expected to initially support 40 different Hindi dialects and more languages and dialects will be added subsequently. “We understand that much work has been done on the Indic suite like Bhashini and AI4 Bharat, etc., but a foundation model still needs to be developed. As we continue to develop the model, we are constantly learning and improving the process. Our interface could have voice and textual information; however, we haven’t considered incorporating a chat interface like ChatGPT yet,” Malhotra said. The primary goal for Tech Mahindra is to first create an LLM for continuation of text and then provide a dialogue. “Once we are clear that the model performs well and generates dialects well, we would launch it in the open source.” Benefits of building India’s biggest Indic LLM ChatGPT, driven by OpenAI’s GPT models, has undoubtedly been groundbreaking. Hence, developing an LLM, primarily designed for Indic languages could be highly beneficial for India for a wide array of reasons. Understanding the nuances of local cultures and contexts is essential for effective communication. An Indic LLM can be designed to prioritise cultural sensitivity, ensuring that the generated content respects local customs and norms. An Indic LLM could also democratise AI and cater to the wider section of non-English speakers in the country. “One of the benefits of a foundation model is its versatility. For instance, a language model is capable of performing multiple tasks such as Q&A, fill-in-the-blanks, etc. using the same model. This approach is beneficial for specialised healthcare, retail, and tourism industries,” Malhotra said. Moreover, the cost of tokens is significantly higher for the Indic languages in the GPT models when compared to English. Hence, an Indic LLM offers a more cost-effective solution for generating content in Indic languages without token pricing constraints. “It represents unrepresented languages and hence helps preserve them. Being the forerunner in this space, Tech Mahindra stands to benefit from the model. In fact, techniques from the model can be leveraged to benefit our customers,” he added. Building Indic datasets The effectiveness of an AI model hinges on the quality of its datasets. While ample English datasets are readily accessible, there is a scarcity of datasets for Indic languages and dialects. Recognising this challenge, various stakeholders, including the Indian government, are actively engaged in the creation of such datasets. Last year, Prime Minister Narendra Modi launched the Bhasini project, which aims to develop language translation technologies that can effectively translate content from one Indian language to another. The initiative also aims to crowd-source voice datasets in multiple Indian languages to enhance the availability and accessibility of digital services in local languages. Moreover, educational institutions such as the Indian Institute of Science (IISc) and IIT Madras (Ai4Bharat) and even Microsoft are involved in building datasets for Indic languages. “Despite various efforts, in India, datasets for languages other than Hindi are scarce and incomplete. Additionally, even Hindi data is fragmented,” Malhotra said. Additionally, he confirmed that Tech Mahindra is actively in talks with leading universities and other stakeholders for Project Indus. Tech Mahindra is sourcing information from various platforms, including Common Crawl, newspapers, Wikipedia and YouTube descriptions. “The information on dialects is primarily available through YouTube videos or spoken language samples. We are also sourcing information commonly available on the internet from books written in specific dialects.” Besides, Malhotra has also acknowledged that the requirement for computation and its accessibility is a key challenge for the tech giant. Banking on Bhasha Daan For Tech Mahindra and for the success of Project Indus, the biggest challenge is gathering data for different dialects. For this, the IT giant is seeking contributions from the speakers of these dialects to help build the datasets. “For this reason, we have opened a portal to get a bhasha daan from Indians who speak that dialect. By clicking “Make a Contribution” on our website, you will find a user-friendly interface with all the dialects in which we collect data. Once you select a dialect, you can listen to a sample voice recording of how Hindi is spoken in that particular dialect. Users can then scroll down and anonymously record a sentence by clicking the record button.” Gurnani took to X (formerly Twitter) to request contributions from the general public to assist in the creation of datasets for Indic dialects. “We humbly request a bit of bhasha daan from you. Please lend us your expressions, your vocabulary, your conversations and help us train India’s biggest indigenous LLM,” he posted. Mitigating biases in datasets Oftentimes, the biases that manifest in AI models originate from biases present within the datasets. Since LLMs learn from a vast amount of text data available on the internet, when not appropriately addressed, these biases can impact the output generated by the models. While building datasets from scratch, Tech Mahindra must put guardrails in place to ensure this does not happen. “When we collect the data at the first phase, it is essential to realise that this data would have to go through cleaning to ensure there is no bias. To address this challenge, we would be using both human annotation and automatic techniques to ensure there is no racial, ethnic, or gender bias, etc.,” Malhotra said. While it’s a commendable move, the success of the project hinges upon various factors such as robust data collection, efficient model training, and addressing linguistic nuances.","excerpt":"The first model will have 7 billion parameters, will be open-source and support 40 Hindi dialects","categories":["AI Trends"],"tags":["bhasini","CP Gurnani","project indus","Tech Mahindra"],"author_name":"Pritam Bordoloi","publish_date":"2023-09-04T16:13:26","publication_year":"2023","word_count":1037,"keywords":["Go","ChatGPT","Tech Mahindra","CP Gurnani","TPU","OpenAI","AI","Git","RAG","Ray","bhasini","Aim","R","project indus"],"extracted_tech_keywords":["AI","ChatGPT","OpenAI","Aim","Ray","RAG","TPU","R","Go","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/exclusive-first-indian-it-company-to-bring-made-in-india-chatgpt\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10110422,"title":"CES 2024 was Gamers’ Paradise","content":"CES 2024 was indeed a gamers’ paradise. From mind-bending virtual experiences to hardware upgrades that would make any pixel pusher drool, the showfloor buzzed with anticipation for the future of entertainment. As many as 14 new RTX games were announced at the event that will support ray tracing and other RTX technologies, including titles like ‘Cyberpunk 2077: Phantom Liberty’, ‘The Witcher 4’, and ‘Elden Ring’. NVIDIA stole the show NVIDIA dropped a bomb with the GeForce RTX 40 SUPER Series. These graphical powerhouses promise to double the performance of their predecessors, rendering AAA titles in hyper-realistic detail with silky-smooth framerates. The RTX 4080 SUPER, RTX 4070 Ti Super, and the 4070 SUPER are the shimmering beacons of performance, claiming to be twice as powerful as the already monstrous RTX 3090 Ti, and other last generation GPUs. But NVIDIA wasn’t done flexing. It unleashed 14 new RTX games, showcasing the jaw-dropping power of ray tracing and DLSS 3.0, their revolutionary AI upscaling tech. NVIDIA also demonstrated its Avatar Cloud Engine (ACE) for gaming, which aims to give NPCs AI-generated dialogue. Then comes the NVIDIA RTX Remix, which is mind-bending, developed on the NVIDIA Omniverse platform for remastering graphics. The open beta for RTX Remix is set to be released later this month. Having already produced remarkable remasters in Portal with RTX and the modder-created Portal: Prelude RTX, RTX Remix is now being utilised by Orbifold Studios for the creation of Half-Life 2 RTX: An RTX Remix Project. AMD and Intel also make a mark AMD responded to NVIDIA with the FidelityFX Super Resolution 3.0 (FSR 3.0). This open-source upscaling tech delivers near-native image quality with a significant performance boost, making it a serious contender for both NVIDIA and AMD users. When it comes to hardware, the Ryzen 8000 G-Series processors, unveiled at CES 2024, mark a significant advancement in integrated graphics for gaming for AMD. With a blend of Zen 4 architecture, RDNA 3 graphics core, and an XDNA AI engine, these processors promise an enhanced gaming experience without the need for a dedicated graphics card. AMD’s realistic approach acknowledges that these integrated graphics may not challenge modern video cards but suggests a compelling option for budget and older card alternatives. The open beta for RTX Remix is set to be released later this month. Intel also released its latest 14th Gen HX-series, catering to the needs of gamers, creators, and professionals seeking top-notch computing performance in a portable form factor. Leading the series is the Intel Core i9-14900HX, equipped with eight Performance-cores (P-cores) and 16 Efficient-cores (E-cores). Notably, the Intel Core i7-14700HX processors in this lineup demonstrate a significant enhancement in creator performance with a 50% increase in E-cores. And the fun part… The headline grabber was the Samsung 3D Monitor, which offers the VR gaming experience without a headset. Featuring dual cameras for eye and head tracking, the device offers an enhanced 3D mode that optimises effects. Additionally, a specialised UI elevates the browsing experience in 3D. Samsung is proactively pursuing partnerships with prominent gaming publishers to expand its 3D gaming content offerings. Several manufacturers announced new gaming laptops powered by the latest NVIDIA and AMD processors and graphics cards. Acer Predator Helios 18 equipped with an NVIDIA GeForce RTX 40 Series GPU and a 120Hz refresh rate display, this laptop is designed for serious gamers. Razer Blade 16 with a sleek and powerful laptop features an Intel Core i9 processor and an NVIDIA GeForce RTX 4090 GPU, making it ideal for demanding gaming and creative tasks. HP wasn’t to be left out – it unleashed a barrage of powerful gaming laptops. HP’s Omen Transcend 14, touting it as the most powerful and slimmest laptop in the world. Then there was ROG’s NUC, which is a gaming console powered by Intel Core Ultra 9 and NVIDIA RTX 4070. It is a new gaming unit with Quad 4k support. Targus also won the CES 2024 Innovation award for announcing its ErgoFlip EcoSmart Mouse that is compatible for both, left and right handed users, for which the company also has a patent. It is also made of sustainable materials. These, along with the introduction of gaming capabilities in BMW and the launch of MSI Claw, which is an Intel-powered gaming handheld, covers all the types of gamers in the world. The best out of the lot was Razer’s HD Haptics gaming cushion, called Project Esther, for comfortable gaming experience, along with simulations of what’s going on in the game.","excerpt":"From new hardware, software, to a gaming cushion.","categories":["AI Features"],"tags":["AI in Gaming","NVIDIA"],"author_name":"Mohit Pandey","publish_date":"2024-01-11T14:42:13","publication_year":"2024","word_count":748,"keywords":["Go","AI","AI in Gaming","innovation","RTX 4090","RAG","BERT","Ray","Aim","llm_models:BERT","NVIDIA","R"],"extracted_tech_keywords":["AI","Aim","Ray","RAG","R","Go","BERT","innovation","RTX 4090","llm_models:BERT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/ces-2024-was-gamers-paradise\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":58537,"title":"Coronavirus Package In Now Available For R Programming Language","content":"Corona pandemic has caused panic throughout the world and is negatively impacting not only humans but also businesses. Consequently, various companies have been moving toward the data-driven approach to analyse the spread and reported cases. To streamline this process, Johns Hopkins University Center for System Science and Engineering (JHU CCSE) has collected data from various sources and created a repository, which was then pulled by Rami Krispin, who released an R package to provide a daily summary of the coronavirus (COVID-19) cases by state\/province. The data set contains various variables such as conformed cases, death, and recovered across different states. The package was introduced on 23 February, is available for R programming users under the MIT license and supports for R 3.0.2 and above. Either you can directly install the package from the R studio or install it manually by downloading it from the GitHub repository. Besides, you can contribute to the project by pushing issues on GitHub and help researchers and developers to analyse using the package flawlessly. The package has quickly gained traction on GitHub as it has been forked 37 times and 110 people are monitoring the repository. You can install the package with install.packages(“coronavirus”) then import the package with library(coronavirus). However, a Python user can directly download the file in the CSV format and analyse to get insights into the data. But you will have to download it again to get the updated file. Thus, to get real-time data, it is recommended to use the package instead. Besides, It provides a tidy format dataset, which is an icing on the cake. Therefore, you can directly get down to visualising and finding trends rather than cleaning it. Or you can directly access the dashboard that the author has created with the data.","excerpt":"Corona pandemic has caused panic throughout the world and is negatively impacting not only humans but also businesses. Consequently, various companies have been moving toward the data-driven approach to analyse the spread and reported cases. To streamline this process, Johns Hopkins University Center for System Science and Engineering (JHU CCSE) has collected data from various […]","categories":["AI Features"],"tags":[],"author_name":"Rohit Yadav","publish_date":"2020-03-13T09:30:00","publication_year":"2020","word_count":295,"keywords":["programming_languages:R","AI","data-driven","ML","Git","Python","programming_languages:Python","GitHub","R"],"extracted_tech_keywords":["AI","ML","Python","R","Git","GitHub","data-driven","programming_languages:Python","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/coronavirus-package-in-now-available-for-r-programming-language\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10047365,"title":"All About Waymo’s AI-Powered Urban Driver","content":"While Phoenix travel guides may not mention this, taking a ride in Waymo’s self-driving car should be at the forefront of that list. Travel the city via a fully autonomous vehicle you can book simply with an app on your phone. Waymo, Google’s platform for driverless vehicles, provides fully autonomous cars seven days a week all around Phoenix. Dubbed ‘the World’s Most Experienced Driver’ by Alphabet AI, Waymo has driven over 20 billion miles so far, with the sole mission to make travelling easier and safer. In 2020, Alphabet expanded the tech to commercial trucks. Waymo calls itself a self-driving technology company rather than a self-driving car company. Their fleet of thousands of self-driving cars is powered with sensors like lidar, enabling driver-free operations and 360-degree perception technology to identify obstacles. Waymo, back in October 2020, announced its plan of launching fully driverless vehicles to the public. It is wholly focused on full autonomy. “We’re building systems that operate without any reliance on human drivers. Either the system can complete the entire trip, or it cannot,” the team at Waymo explained. In this article, we break down how Waymo leverages AI\/ML to create their dominating technology. Territory Marking & Prediction Before operating in a new area, the Waymo Driver maps the territory with a high detailing rate. This includes noting the stop signs or the curbs, and the platform adopts these into custom maps equipped with real-time sensor data. This information is later mixed with real-time details to gauge how the road users might behave and drive accordingly. Advanced Sensors Waymo has claimed to have built the most advanced sensors and perception systems with quality data. The system dataset contains more than 20 million autonomously driven miles and five generations of development. Its fifth-generation Driver comprises complementary sensors that include radar, lidar, and cameras that allow it to see 360 degrees around the vehicle. This considers the time of the day, bad weather conditions like rain or fog, and the variety of road users such as those double-parked vehicles whose riders can hop out at any second or the scooters cutting across traffic through a red light. Waymo Blog Post A dense and highly detailed 3D view of its surroundings as generated by the system’s lidar These sensors allow the vehicle to navigate the city streets accurately and track objects around, especially while taking turns or changing lanes. “Because of the long-range and high density provided by our sensors, the Waymo Driver can detect with high accuracy, and intelligently reason about small objects and movement at distances – such as spotting a truck door slowly opening in the flow of traffic as a person gets ready to hop out and deliver some packages,” according to Waymo’s blog post. https:\/\/blogger.googleusercontent.com\/img\/a\/AVvXsEiiRGrrnTC9H91WNk7gVG69LiyiT8-8prrm8m0WofX7RPvHd5GmYrBIwK9xkXXyvagARZL_RIKrKMPcTzLoxQqQNxeoBZrUtnm72pfePwUHWTsMJsBz5WGS9nY8LpsuMutOka2BAYZx4ZW1Bzh6D7NGvaMD720tj-ODjcrtn2MdN1K4WjF4nr5WbJiaCg=s956 (Waymo blog post: illustrating how the car takes a left turn) Waymo’s compute platform allows the Driver to process vast amounts of data and run real-time inference on large ML models. This will enable it to react immediately without additional human input, like when an emergency vehicle is approaching or when a dog runs out into the street. They even do this without needing a cell signal to operate. Software and Training Structure Waymo’s driving software is based on years of AI research, their Waymo Open Dataset Initiative, and research team Google Brain. The engineers working at Waymo operate in coordination with the Google Brain team to apply deep nets to the car’s pedestrian detection system. The team has created a robust, generalisable tech stack based on their operation in multiple environments and cities. The Waymo Driver has learnt to behave assertively and merge into traffic based on this experience. Waymo has invested in creating training softwares for the Waymo Driver. The Simulation City is software to test the autonomous vehicles and assess their performance for the cities Waymo is present in. It creates realistic conditions like spring showers, solar glare, or dimming light for the technology to experience; the researchers further learn from the system’s reactions. Waymo blog post: an illustration of how the Simulation City draws a random outcome from a statistical distribution of the real world. The Waymo Driver itself is trained with a highly nuanced understanding of city roads with driving experience of more than 20 million miles on public roads and 20 million miles in simulation. It can adapt to the local driving conditions accurately, given this training. For instance, in San Francisco, the system has learnt that the residents often drive slower on steep slopes and adapt to change its speed depending on traffic flow. Waymo Driver makes similar calculations while approaching crosswalks or lower visibility areas. Tensorflow Ecosystem Waymo has adopted the TensorFlow ecosystem and the data centres of Google, such as TPUs, for training its neural networks. Their rigorous testing on simulations and training cycles allows it to enhance its ML and autonomous system. SurfelGAN The platform is leveraging AI to generate camera images to simulate sensor data gathered through its self-driving vehicles. In a recent paper, the researchers introduced the technique – SurfelGAN. This uses texture-mapped surface elements for reconstructing scenes and camera viewpoints to handle positions and orientations. This is a considerably more accessible and more data-driven path to simulate sensor data. The AI generates and preserves data based on real-world lidar sensors and camera feeds. This data is regarding the semantics, 3D geometry, and object appearances on the scene used to render a simulated scene through the reconstruction. Currently, Waymo is the only company operating a fully autonomous public ride-hailing service in the U.S. They have partnered with many car manufacturers like Jaguar Land Rover or Volvo to integrate the self-driving technology into their vehicles and with OEMs to expose their Driver to more places and people.","excerpt":"This information is later mixed with real-time details to gauge how the road users might behave and drive accordingly.","categories":["AI Features"],"tags":["ai generated images"],"author_name":"Avi Gopani","publish_date":"2021-08-31T13:00:00","publication_year":"2021","word_count":957,"keywords":["ai generated images","Go","TPU","AWS","AI","neural network","ML","RAG","Aim","TensorFlow","R"],"extracted_tech_keywords":["AI","ML","neural network","Aim","TensorFlow","RAG","AWS","TPU","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/all-about-waymos-ai-powered-urban-driver\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10100253,"title":"OpenAI Releases GPT 3.5 Turbo Instruct","content":"OpenAI has silently unveiled “gpt-3.5-turbo-instruct,” a new instruction language model designed for giving specific instructions efficiently, similar to the chat-focused GPT-3.5 Turbo. This new model will replace existing Instruct models and certain text-based models. It maintains the same cost and performance as other GPT-3.5 models within a 4K context window, using training data up to September 2021. Compared to GPT-3.5, gpt-3.5-turbo-instruct is not a chat model. Instead, it excels in directly answering questions and completing text tasks swiftly, aligning with its purpose of following instructions effectively. OpenAI emphasizes that GPT-4 surpasses GPT-3.5 in comprehending complex instructions, generating higher-quality output, and doing so at a faster pace and a lower cost. Instruct models, including this new release, are a crucial foundation for the breakthroughs seen in ChatGPT, based on continual refinement through human feedback during and after pre-training. These models are known for better understanding and responding to human queries with fewer errors and less harmful content, making them preferred even when significantly smaller in parameters than regular GPT models. In January, OpenAI had released the instruct GPT models which was intended to reduce hallucinations and coax the model to produce more ‘truthful’ and less toxic responses. GPT-3 already did this because they explained that “it is trained to predict the next word on a large dataset of Internet text, rather than to safely perform the language task that the user wants.” The model wasn’t aligned with the users and often made up facts which was combated by InstructGPT. The main difference between the new update and the older version is that GPT 3.5 turbo instead of being conversational, will ‘instruct’ in a more task-oriented manner rather than a conversational one. The model will not be ‘chatty’ and it will follow instructions without requiring additional prompting techniques.","excerpt":"GPT-3.5 Turbo is primarily designed for chatting or engaging in conversations, mimicking human-like responses. On the other hand, “instruct” is a more versatile tool that can be utilized not only for conversation but also for completing various tasks or questions.","categories":["AI News"],"tags":["OpenAI"],"author_name":"K L Krithika","publish_date":"2023-09-19T16:11:07","publication_year":"2023","word_count":296,"keywords":["ChatGPT","TPU","OpenAI","AI","programming_languages:R","RPA","GPT","R","llm_models:ChatGPT","llm_models:GPT"],"extracted_tech_keywords":["AI","ChatGPT","OpenAI","TPU","R","GPT","RPA","llm_models:GPT","llm_models:ChatGPT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/openai-releases-gpt-3-5-turbo-instruct\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10064145,"title":"Ambitions to become GitHub for machine learning? Hugging Face adds Decision Transformer to its library","content":"Hugging Face is one of the most promising companies in the world. It has set to achieve a unique feat – to become the GitHub for machine learning. Over the last few years, the company has open-sourced a number of libraries and tools, especially in the NLP space. Now, the company has integrated Decision Transformer, an offline reinforcement learning method, into the transformers library and the Hugging Face Hub. What are decision transformers Decision transformers were first introduced by Chen L. and his team in the paper ‘Decision Transformer: Reinforcement Learning via Sequence Modelling’. This paper introduced this framework that abstracts reinforcement learning as a sequence modelling problem. Unlike previous approaches, Decision Transformers output the optimal actions by leveraging a causally masked Transformer. A Decision Transformer can generate future actions that achieve desired return by conditioning an autoregressive model on desired reward, past states, and actions. The authors concluded that despite the simple design of this transformer, it matches, even exceeds, the performance of the state-of-art model and free offline reinforcement learning baselines on Atari, OpenAI Gym, and Key-to-Door tasks. Decision Transformer architecture The idea of using a sequence modelling algorithm is that instead of training a policy using reinforcement methods that would suggest action to maximise the return, Decision Transformers generate future actions based on a set of desired parameters. It is a shift in the reinforcement learning paradigm since the user is using a generative trajectory modelling to replace conventional reinforcement learning algorithms. The important steps involved in this are – feeding the last K timesteps in the Decision Transformer with three inputs (return-to-go, state, action); embedding the tokens with a linear layer (if the state is a vector) or CNN encoder if it is a frame; processing the inputs by GPT-2 model that predicts future actions through autoregressive modelling. Offline reinforcement learning Reinforcement learning is a framework to build decision making agents that learn optimal behaviour by interacting with the environment via trial and error method. The ultimate goal of an agent is to maximise the cumulative reward called return. One can say that reinforcement learning is based on the reward hypothesis and all the goals are the maximisation of the expected cumulative reward. Most reinforcement learning techniques are geared in the online learning setting, where the agents interact with the environment and gather information using current policy and exploration schemes to find higher-reward areas. The drawback with this method is that the agent has to be trained directly in the real world or have a simulator. In case a simulator is not available, one would be required to build it, which is a very complex process. Simulators may even have flaws that can be exploited by agents to gain a competitive advantage. Credit: Hugging Face This problem is present in the case of offline reinforcement learning. In this case, the agent only uses the data collected from other agents or human demonstrations without interacting with the environment. Offline reinforcement learning learns skills only from previously collected datasets without active environment interaction and provides a way to utilise previously collected datasets from sources like human demonstrations, prior experiments, and domain-specific solutions. GitHub for machine learning Hugging Face’s startup journey has been nothing short of being phenomenal. The company, which started as a chatbot, has gained massive attention from the industry in a very short period; big companies like Apple, Monzo, and Bing use their libraries in production. Hugging Face’s transformer library is backed by PyTorch and TensorFlow, and it offers thousands of pretrained models for tasks like text classification, summarisation, and information retrieval. In September last year, the company released Datasets, a community library for contemporary NLP, which contains 650 unique datasets and more than 250 contributors. With Datasets, the company aims at standardising end-user interface, versioning, and documentation. This sits well with the company’s larger vision of democratising AI, which would extend the benefits of emerging technologies to smaller technologies, which is otherwise concentrated in a few powerful hands.","excerpt":"Over the last few years, the company has open-sourced a number of libraries and tools, especially in the NLP space.","categories":["Global Tech"],"tags":["Clément Delangue","Generative Pre-Trained Transformer","Hugging Face"],"author_name":"Shraddha Goled","publish_date":"2022-04-01T17:00:00","publication_year":"2022","word_count":663,"keywords":["Hugging Face","machine learning","OpenAI","AI","PyTorch","Transformers","RAG","NLP","Aim","Generative Pre-Trained Transformer","Clément Delangue","TensorFlow"],"extracted_tech_keywords":["AI","machine learning","NLP","OpenAI","Aim","TensorFlow","PyTorch","Hugging Face","Transformers","RAG"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/ambitions-to-become-github-for-machine-learning-hugging-face-adds-decision-transformer-to-its-library\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":44522,"title":"We Want To Make Hiring Cycle Shorter &#038; More Efficient With Data Analytics: Shine.com’s Madhukar Kumar","content":"This week we talked to Madhukar Kumar, Chief Analytics Officer at Shine.com who has over 13 years of experience in the data science industry and he has been consulting across a multitude of both global clients and verticals. Apart from his corporate life, Kumar is also interested in Academics and teach data science at various universities and spends a considerable amount of time nurturing future data scientists across the globe. In this interaction, Kumar shares his take on how the analytics domain is playing a major role in Shine.com’s operations as well as in the Indian analytics ecosystem. AIM: How will your appointment in the analytics division play a crucial role in re-imagining the analytics strategy for Shine? Since my appointment, We have already gone through major organisational and structural changes to ensure that we are keeping Data science at the core of everything we do. We are bringing a data-driven culture in the organisation and the idea is to make Shine an AI company. We are focused on leveraging technologies such as artificial intelligence, machine learning, and deep learning to catapult our product into the next phase of evolution. We are redefining our analytics strategy to streamline search results, making them more relevant and in-depth for jobseekers. We will also build on our existing products to better serve recruiters in their task to hire great talent. I believe that we will be able to steer the analytics division towards greater success. Our overarching goal is to enhance user-experience which will only be possible with the optimum deployment of technology. Thus, the adoption and development of new-age technologies continue to be my major focus areas at Shine.com. AIM: What is your vision for Shine.com? Do you plan to expand the product portfolio? Customer centricity has always been at the heart of all our endeavours at Shine.com. Living up to this long-standing commitment towards enhancing user-experience, we plan to build on existing products, making them more seamless and easier to use for our customers. Product expansion is also definitely on the cards for the brand. In fact, we recently introduced face recognition and touch ID capabilities on Shine.com’s mobile application to facilitate seamless login. We will continue to leverage forward-looking technologies to unlock new levels of growth for the brand. The ultimate aim is to add immense value to the entire hiring process with our robust, tech-backed offerings. Through our product, we are working to shorten the overall hiring cycle which usually takes around 70 days. By reducing the time taken for each hire, we will be enhancing the experience of both recruiters and candidates in a sure-footed way. AIM: What are AI and analytics expertise that you are going to bring from your previous roles? I have been working in the analytics industry for the past 13 years and have built the data science team from scratch multiple times. All these years, I have used AI and data science to solve various complex problems that included structured as well as unstructured data. Also, I have taught 1,000+ professionals across the globe who were keen to make the transition to the data science field. My experience of setting up the data science and solving complex problems using AI would again endeavour at Shine. My teaching experience will help the internal team of data scientists. AIM: How are you planning to strengthen the data science team? What would be the biggest challenges in ensuring it? Since Shine.com relies on the extensive recruiter and candidate data to facilitate efficient job-candidate matching, the data science team forms the backbone of our entire process. Therefore, we are focusing on growing our data science team. While hiring the right talent for the team, we will be looking at technological-adeptness of candidates. The major challenge, therefore, will be to zero in on candidates with the relevant skill sets. Some of the major skills we are looking for include AI, ML and data engineering. AIM: How are technologies like deep learning and machine learning being used at Shine? It’s important to provide an excellent experience to recruiters and candidates and we are working tirelessly for it. Deep learning enhances search result by doing a better matching of jobs and candidates. Matching jobs and candidates is a 4-way matching (job – job, job – candidate, candidate – candidate, candidate – job). We are currently using it and it’s a continuous improvement process. AIM: How has the role of analytics and AI evolved in the hiring and recruitment space? What are the various benefits that emerging tech has brought? New-age technologies such as analytics and machine learning have become integral to the modern hiring and recruitment process. These tech-led tools deliver enhanced efficiency, accuracy, and seamlessness. This allows recruiters to save significant time by automating routine tasks, such as candidate outreach and shortlisting, and focus on more value-driven objectives. Tools such as analytics and AI can also enhance the user experience for candidates as they are shown more relevant jobs. AIM: What are the ways that data analytics is being used at Shine? Would you like to explain with some use cases? At Shine.com, we use AI-based algorithms to efficiently match candidates\/skills with the right job profile. Through this, we make the process of job discovery and application extremely easy and convenient for both recruiters and jobseekers. This further enhances their experience on the Shine.com platform. AIM: What are the overall challenges you foresee in Indian analytics industry? The first challenge is too much fascination on buzzwords like AI, data science, deep learning, machine learning etc. Instead of focusing on problem-solving, we are focusing on means\/methods of problem-solving. AI and data science are just means to solving business problems. All these complex algorithms were invented decades ago, but it was not required earlier in the industry as problems were not that complex and data size was also pretty less as compared to what we have now. As an industry, we have shifted towards using complex algorithms only because of the inability of standard algorithms to solve the problem. Nowadays, the first question is asked whether you have applied deep learning or machine learning to solve the problem or not and not how well the problem is solved. That’s the biggest irony. The second challenge is the acute shortage of talented skill set in futuristic technologies like AI and ML. How to manage this gap is going to be the key to success for the overall Indian analytics industry.","excerpt":"This week we talked to Madhukar Kumar, Chief Analytics Officer at Shine.com who has over 13 years of experience in the data science industry and he has been consulting across a multitude of both global clients and verticals. Apart from his corporate life, Kumar is also interested in Academics and teach data science at various […]","categories":["AI Features"],"tags":["Data Analytics","Data Science","Data Scientist","Interviews and Discussions","robust analytics strategy"],"author_name":"Srishti Deoras","publish_date":"2019-08-14T17:17:42","publication_year":"2019","word_count":1078,"keywords":["data science","artificial intelligence","machine learning","AI","ML","RAG","robust analytics strategy","Aim","deep learning","analytics","Data Analytics","Data Science","Data Scientist","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","data science","analytics","Aim","RAG","R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/we-want-to-make-hiring-cycle-shorter-more-efficient-with-data-analytics-shine-coms-madhukar-kumar\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10041978,"title":"Story of Gradient Boosting: How It Evolved Over Years","content":"Between October and December 2016, Kaggle organised a competition with over 3,000 participants, competing to predict the loss value associated with American insurance company Allstate. In 2017, Google scholar Alexy Noskov won the second position in the competition. In a blog post on Kaggle, Noskov walked readers through his work. The primary models he employed were neural networks and XGBoost, a variant of Gradient Boosting Machines (GBM). A method of machine learning boosting, Gradient boosting, combines various simple models with limited performance levels (like weak models or weak learners) into a single composite one. In 1988, Michael Kearns said that the goal of boosting was to make ‘an efficient algorithm for converting relatively poor hypotheses into very good hypotheses’. Two widely employed boosting algorithms are Adaptive Boosting, also known as AdaBoost, and Gradient Boosting. The origin: AdaBoost AdaBoost, which is short for Adaptive Boosting, was the first successful boosting ensemble algorithm. The most commonly used weak learners here are decision trees with the depth of one, due to which they are known as decision stumps. AdaBoost works by weighing observations and training a set of weak learners sequentially. Here, more weight is given to samples fitted worse in previous steps. Eventually, the complete set of learners is combined as a single complex classifier. In 1998, Leo Breiman formulated AdaBoost as a gradient descent with a particular loss function. Taking this further, Jerome Friedman, in 1999, came up with the generalisation of boosting algorithms, and thus, a new method: Gradient Boosting Machines. One can find Friedman’s paper detailing Gradient Boosting here. Today, AdaBoost is regarded as a particular case of Gradient Boosting in terms of loss functions. Gradient Boosting Gradient Boosting is a machine learning algorithm made up of Gradient descent and Boosting. Gradient Boosting has three primary components: additive model, loss function, and a weak learner; it differs from Adaboost in some ways. As mentioned earlier, the first of these is in terms of the loss function. Boosting utilises various loss functions. AdaBoost minimises the exponential loss function that makes the algorithm vulnerable to outliers. Gradient Boosting, on the other hand, allows any differentiable loss function to be used. This makes Gradient Boosting more stable than AdaBoost upon meeting outliers. Secondly, Gradient Boosting predicts the error left by the previous model. It is done through a loss function optimisation, conducted through gradient descent. But before that, we look at decision trees. Decision trees Gradient Boosting typically uses short, less complex-decision trees instead of decision stumps. Here, many weak learners, which make up the individual decision trees, form one strong learner. Every tree is connected in series, and each particular tree attempts to minimise the error of the previous tree (that is, the residuals). Doing this makes boosting algorithms slow to learn but highly precise. A loss function is used to determine the residuals from each step. For example, a user could use Mean Squared Error for a regression task and Logarithmic Loss for classification tasks. Let’s see how this would look mathematically: An output model y, when fit into a single decision tree, is given by: y= A1+( B1+x)+e1, where e1 is the residual term from this particular decision tree. Gradient Boosting fits consecutive decision trees on the residual from previous ones. Keeping this in mind, the consecutive decision trees would be: e1= A2+( B2+x)+e2 e2= A3+( B3+x)+e3 And so on. Assuming this specific Gradient Boosting model only uses three decision trees, the final model of the decision tree would be: y=A1+A2+A3+(B1+x)+(B2+x)+(B3+x)+e3 Improving performance Gradient Boosting is also prone to some overfitting, which can decrease its performance. One way to go about this is through Stochastic Gradient Boosting. It involves sub-sampling the training dataset and using this to train the individual learners. Doing so reduces the correlation between results from individual learners, leading to a more accurate result. A second method to improve performance is by a technique called shrinkage. Here, the predictions of each tree are weighed to slow down the algorithm’s learning process instead of adding them together sequentially. However, since lower learning rates require more iterations, this comes at the cost of computational time. A final method to improve performance is through placing constraints on trees. Adding excessive-decision trees further contributes to overfitting. Other parameters here include tree depth, where shorter trees usually lead to better results. The number of observations per split limits the amount of training data at a step before undergoing a split. It took more than ten years after introducing GBM for it to become vital components of data science. Since then, however, Gradient Boosting is becoming increasingly popular. A primary reason for this is using GBM’s implementations, such as the Kaggle-popular XGBoost, in various machine learning competitions. XGBoost further uses tricks that make it faster and more accurate than traditional means of Gradient Boosting. Through this, one can easily see Jerome Friedman’s 1999 innovation’s usefulness and hope for it to evolve and fit future data science applications.","excerpt":"Between October and December 2016, Kaggle organised a competition with over 3,000 participants, competing to predict the loss value associated with American insurance company Allstate. In 2017, Google scholar Alexy Noskov won the second position in the competition. In a blog post on Kaggle, Noskov walked readers through his work. The primary models he employed […]","categories":["AI Features"],"tags":["gradient boosting","Kaggle","XGBoost"],"author_name":"Mita Chaturvedi","publish_date":"2021-06-20T16:00:00","publication_year":"2021","word_count":825,"keywords":["data science","Go","machine learning","Kaggle","TPU","AI","neural network","XGBoost","gradient boosting","ViT","GAN","R"],"extracted_tech_keywords":["AI","machine learning","neural network","data science","XGBoost","TPU","R","Go","GAN","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/story-of-gradient-boosting-how-it-evolved-over-years\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10137263,"title":"Government of India Launches BharatGen, First Government-Funded Multimodal AI Initiative","content":"The Indian government has unveiled BharatGen, the first government-funded initiative for developing multimodal LLMs, which stands for building “GenAI for Bharat, by Bharat.” The project, which aims to revolutionise public service delivery and enhance citizen engagement through AI, was inaugurated by Union Minister Dr. Jitendra Singh on September 30, 2024. “BharatGen is a proud example of India’s commitment to advancing homegrown technologies. It positions India as a global leader in the field of Generative AI, much like our achievements with UPI and other innovations,” said Dr. Singh during the launch. Spearheaded by IIT Bombay under the National Mission on Interdisciplinary Cyber-Physical Systems (NM-ICPS) of the Department of Science and Technology (DST), BharatGen focuses on creating AI models tailored for India’s diverse languages and cultural contexts. The project is being executed by the TIH Foundation for IoT and IOE at IIT Bombay, in collaboration with several leading academic institutions, including IIT Bombay, IIIT Hyderabad, IIT Mandi, IIT Kanpur, IIT Hyderabad, IIM Indore, and IIT Madras. Present at the event were IIT Bombay Director, Prof Shireesh Kedare, and a consortium of faculty members led by Prof Ganesh Ramakrishnan. “BharatGen is not just about foundational AI technology, rather it facilitates a comprehensive and inclusive ecosystem. This AI will create new opportunities in important sectors like education, health, agriculture, and industry,” Ramakrishnan, Lead of the BharatGen initiative, told AIM. Ramakrishnan, through the BharatGPT initiative, has been pushing for a public-private partnership for AI for a very long time. The efforts mentioned through BharatGPT earlier have now assumed a more formal role in BharatGen with a broader scope in Generative AI beyond just GPT. “Startups, researchers, government agencies, and ordinary citizens will have the opportunity to be a part of this ecosystem, creating a collaborative, innovation-led, and empowered future.” According to Ramakrishnan, BharatGen will capture the diversity and cultural heritage of our country through the Bharat Datasagar. This project will not only symbolise technological advancement but will also become a means of bringing our local knowledge and experiences into the mainstream of AI. “It has been developed in partnership with leading educational institutions in India, which will further strengthen this project. This project is a major step towards digital inclusion, where every region, every language, and every citizen will be equally connected to digital services. This is a strong effort towards fulfilling the dream of a self-reliant India, where technology is being developed in India, for India, by India,” explained Ramakrishnan. BharatGen stands out with four key features: its multilingual and multimodal foundation models, the use of datasets specific to India for model building and training, its open-source platform, and the creation of an ecosystem for generative AI research across the nation. The project, expected to be completed by 2026, will benefit government, private, and academic institutions, fostering AI research and innovation in the country. By supporting both text and speech, BharatGen will cover the vast linguistic diversity of India. Its multilingual datasets will capture the intricacies of Indian languages, which are often overlooked in global AI models. Unlike models that depend on international datasets, BharatGen focuses on collecting and curating data specific to India, ensuring accurate representation of the country’s languages, dialects, and cultural context. This focus on data sovereignty reinforces India’s control over its digital assets and narrative. Professor Abhay Karandikar, Secretary of DST, emphasised, “BharatGen is aligned with the goal of making AI accessible to all citizens, addressing national priorities like cultural preservation and inclusive technology development.” BharatGen’s roadmap includes extensive AI model development, experimentation, and scaling AI adoption across industries. The initiative also aims to democratise AI access, empowering startups and researchers to develop applications efficiently.","excerpt":"The project, expected to be completed by 2026, will benefit government, private, and academic institutions, fostering AI research and innovation in the country.","categories":["AI News"],"tags":["Fund Raising","multimodal ai"],"author_name":"Mohit Pandey","publish_date":"2024-10-02T18:44:28","publication_year":"2024","word_count":605,"keywords":["Go","GenAI","AI","Fund Raising","Git","GPT","Aim","generative AI","GAN","foundation models","R","multimodal ai"],"extracted_tech_keywords":["AI","generative AI","GenAI","foundation models","Aim","R","Go","Git","GPT","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/government-of-india-launches-bharatgen-first-government-funded-multimodal-ai-initiative\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10040327,"title":"The Secret Life Of Navdeep Jaitly: How This ex-Google Brain Researcher Fell In Love With Algorithms","content":"“I asked Geoffrey Hinton to take me on as a student, but he said he was too busy. ” For this week’s ML practitioner’s series, Analytics India Magazine (AIM) got in touch with former Google Brain researcher Navdeep Jaitly. From leaving IIT to pursuing a Liberal Arts Degree, from switching between AI and computational biology to becoming a student of Geoff Hinton, Navdeep’s journey in the world of cutting-edge research has all the twists and turns to give Spielberg’s Catch Me If You Can a run for its money. AIM: Tell us a bit about your educational background. Navdeep: I started my undergraduate education in Engineering at IIT Delhi, but moved to the US in my second year to pursue a Liberal Arts Degree at Hanover College which offered me an academic scholarship. There, I completed a double major in Mathematics and Computer Science and also completed most of the premedical requirements. I finished my graduation early, moved to Canada, where my parents had moved, and worked for almost a year as a junior telecommunications analyst before starting my Masters at the University of Waterloo in Canada. I signed on to do Artificial Intelligence but found my supervisor’s specialization — planning and constraint satisfaction — particularly uninteresting from the standpoint of the data that was involved. Fortunately, I was taking a Computational Biology Research seminar, focusing on sequence assembly from shotgun sequencing and other string algorithms for biological problems, and I fell in love with that. I had always found Genetics fascinating since high school days, and this let me combine biology and computing. Also, the first release of the Human Genome sequence was imminent (this was 1999), and it seemed like a momentous time to work on that. So, I switched into the Computational Biology group and worked as a researcher in that area for about eight years at a biotech startup and in the U. S. National lab system, before I went back to get my PhD degree in Computer Science at the University of Toronto. I started with the Computational Biology group, but in a reversal of my direction from my Masters, this time, I ended switching to the Machine Learning group and getting my PhD in that. I should say that somewhere in the middle, while I was working, I also met half of the requirements for a Masters in statistics through an excellent online offering at Texas A&M, but I stopped pursuing that when I went back for my PhD in Computer Science. So, I have clearly had a winding academic journey, much to my spouse’s dismay. AIM: How did your fascination with algorithms begin? Navdeep: I had been developing statistical methods to characterise the quality of results from my signal processing algorithms for high throughput proteomics as part of my research at the Pacific Northwest National Lab. At first, I was educating myself from books I could find on Bayesian analysis (I particularly enjoyed Bayesian Data Analysis, by Andrew Gelman, John Carlin, Hal Stern, David Dunson, Aki Vehtari, and Donald Rubin), but I liked it so much that I enrolled in a new online offering at Texas A&M for a Masters in Statistics. So I was operating in an adjacent space to Machine Learning at this time, but Machine Learning really just entered my consciousness in a large way when I joined University of Toronto in 2008 to get my PhD in Computational Biology. The Computational Biology group was right next door to the ML group, which is where I first discovered Geoffrey Hinton, Richard Zemel and Radford Neal and their fantastic research. Until this point, I hadn’t really considered that images, text, speech could be modeled with statistical models. It was quite an epiphany for me, when I went to my first class in Machine Learning at the University of Toronto, Geoffrey Hinton was talking about images, and maximum likelihood and sampling algorithms that could generate images (using Deep Belief Networks). As soon as I saw a Deep Belief Network model generate an image of MNIST digits, by sampling activations hierarchically from the highest layers to the lowest, I knew I had to find out more. Later in the class, I asked Geoffrey Hinton to take me on as a student, but he said he was too busy. However, by the end of the class, he changed his mind —turns out he likes students who did well on Mathematics competitions such as the Putnam math competition and Mathematics Olympiads, which I did. AIM: What were the initial challenges and how did you address them? Navdeep: My first project at Geoff Hinton’s lab was to use Deep Belief Networks to separate waveforms into speech from different speakers. That project morphed into a related idea where Deep Belief Networks were used to find features that could be used in a speech recognizer. My lab mates George Dahl and Abdel Rahman Mohamed had just released breakthrough results showing that Deep Belief Networks (DBN) worked really well in pertaining neural networks that could be used for speech recognition. The inputs to their models were features that the speech community had developed several decades ago, called Mel Filterbank Cepstral Coefficients (MFCCs). I wrote a paper showing how the features that were found from raw waveforms using DBNs outperformed MFCCs (“Learning a better representation of speech sound waves using restricted boltzmann machines” in ICASSP 2011). On the heels of this paper, I interned at Google and implemented a model for discriminative training of Neural Network speech models by Brian Kingsbury at IBM, merging it with George and Abdel-Rahman’s successful method. Initially, I was given a smaller toy dataset — the Deep Learning revolution hadn’t yet started (Alex-Net was a year away in the summer of 2011) and researchers at Google were quite skeptical that Deep Neural Networks would work on large scale data. Back then, they were under the belief that linear models trained on very large datasets would be hard to beat with such complex, non-linear models such as neural networks. The project had many challenges. The speech recognition system was built for Gaussian Mixture Models – Hidden Markov Models (GMM-HMM) not neural networks. I had to spend a lot of time understanding how the Google speech recognizer worked, to find out how to plug neural network predictions into it. Further, neural network training requires GPUs, or training is very slow. And Google had no GPUs in the data centers back then. We had bought a machine with 4 GPUs that sat next to the printer and whirred away. Training one epoch of our model on this single machine took one whole day for our larger datasets. I was worried that either my code would break (we didn’t have TensorFlow or other frameworks available back then — I was just using my own code) or the machine would shut off, making me lose time in the short internship. Meanwhile, we were up against baselines that ran with 10x more data, using Google’s cloud infrastructure (called Borg) on thousands of machines. However, within a few weeks I had better results than the best GMM-HMM model Google had built on this data, reducing word error rate from 23% down to 18%. We then started working on a larger datasets from Voice Search and Youtube. By the end of the internship we had improved results by 15% — quite significantly better than the typical improvements over a year. Andrew Senior, a researcher in the group, later increased the gains to 30% by using more data, after my internship finished (“Application of Pretrained Deep Neural Networks to Large Vocabulary Speech Recognition.” by Jaitly et. al, in Interspeech 2012) . The internship had a huge impact in the speech group at Google. Shortly after, Google improved the inference speed and a version of our model was put into production and used for transcription of Google Voice Search queries. By the next year, neural networks were the norm in the speech group (“Genius Makers” by Cade Metz has a nice story about this work). One of the interesting things about neural networks is  that they are often driven by a relentless belief of the practitioners in their succes. I too had no doubt that they would work, and so I kept at it until I got the results I wanted. I think it is often the case with new technologies that early practitioners have an almost irrational belief in their technology — that certainly helped our early work. After that internship, I worked on the speech recognition problem for several years, with the goal of having a single neural network replace the entire complex system (which had the neural network as just one component). The quest was helped by the success of sequence to sequence models within Google Brain and outside, and through collaboration with incredible researchers like William Chan. AIM: How do you approach any machine learning problem? Navdeep: Sadly, I seem to view the world through an ML lens — whenever someone describes a system or a use case they are working on, I’m automatically drawn to think about how machine learning could make it better. And, if machine learning is being used already, my thought experiments run wild on how the current ML solution can be improved further. Sometimes, the data required for a model just doesn’t exist, and cannot be gathered with reasonable resources before the system becomes available. Other times the use case doesn’t offer the computational resources needed for a fancy solution. Both of these can be mitigated — for example, for the former, one might set up a self-supervised system, which trains on its own data and improves over time, starting with a simple model. In the latter case, one can use a wide variety of techniques to train low compute cost models. Other times, when the data does exist, the question to ask is — do we really need ML for this. A good rule of thumb is, that the data needed to be complicated enough to need machine learning because an appropriate analytical understanding of the data cannot be had — if analytical characterizations exist that are correct, it’s probably better to find ways to use them. Another question to ask is, is the structure of data such that large amounts of data will reveal patterns that cannot be modeled by other means? For example you don’t need (and shouldn’t use) machine learning to compute the mean of numbers. Sometimes you have an interesting case where approximate analytical understanding exists for a domain, but there is a limit to its precision. If this domain has underlying structure, I have found that in such cases, Machine Learning can be very useful when large amounts of data exist, to uncover that structure. This is a situation where you can start with a simple model based on an analytic understanding, but over time you can replace it with more complicated machine learning models that can pick on emergent regularities that aren’t understood or known. “ML is a field that requires both mathematical thinking and an attitude of tinkering at the same time.” Generally speaking, for building machine learning situations to problems, I have found it pretty useful to follow a graduated schedule of putting together something very basic and straightforward, and use it as a basis for improved systems. For example, you might want to build only a couple of simple ML models into your system at the start. But as you gain experience, you might supplant various components with ML replacements. Eventually you might consider an alternative where the entire system is a single ML model. AIM: What does your machine learning toolkit look like? Navdeep: During my PhD days I had my own ML toolkit (as every researcher in ML did in those days), implementing most of the algorithms I used. But now the community has so many great alternatives, that I rarely start from scratch in anything, and instead, often my job is to figure out which toolkit works best for which task. For Deep Learning models, I now mostly use tensorflow. I’ve experimented with Pytorch and minimally with Jax and I like them both a lot as well. For other ML models, I have played with scikit (sklearn) which is quite extensive in its offerings now, for more classical ML models. For programming languages, I use python almost exclusively, unless I hit some speed bottleneck that can be solved by going to C++. In that case, I often fallback to ctypes, but there are much better alternatives now for speedups, e.g. cython, numba, boost-python to name a few. I will also fall back on C++ for fast inference needs where precise timing guarantees are needed. In terms of Cloud, I’ve used Google Cloud which works quite nicely. I’ve also looked at AWS which looks great. To be honest, I think both platforms offer various tools and workflows to help you along. AIM: We see a lot of hype around AI. Which domain of AI, do you think, will come out on top in this decade? Navdeep: A lot of hype has indeed been generated around AI and ML focusing on imminent doom from machines taking over. Much less attention is paid to the fact that these methods have indeed revolutionized how we interact with machines. When I started working on using Deep Learning to improve Speech recognition, some of my friends liked to show me how poorly the recognizers worked for them and expressed the opinion that there was no real revolution in play. However, over time these things have improved to a point that my kids are often getting their questions answered by Google Home or Siri, rather than by their parents. And it is my opinion that this is just the start. The algorithms are improving continually still, and their application is percolating to more and more domains and more and more tasks. I think language modeling is another domain that has been forever changed by the current progress in Deep Learning. GPT-3 from OpenAI shows what amazing things the current best in class models such as transformers can produce. Conversational agents is an area that we should see great progress in, with such language models, and this will again really change how we interact with machines. Machine vision has also been transformed and its ramifications are visible in the improvements we see in self driving technology, and also in better understanding of the contents of the media we capture with our phones. I think in the 10 years, we will see the impact of these methods in the changed way we deal with machines. Every time I find that my phone has learned a new trick, or a new interactive behavior, I’m really excited to see where we will be in 10 years. AIM: What would your advice be to aspirants who want to get into ML jobs? Navdeep: ML is a field that requires both mathematical thinking and an attitude of tinkering at the same time. For people wanting to pursue data science and ML roles, I recommend getting your hands on data and trying out things and developing an intuition for data. Amazingly, data seems to show similar patterns from one domain to another, and turning the knobs in one domain helps turning the knobs in another domain. Because of this almost vocational aspect to this field, my office mate and I often described ourselves as Neural Network technicians, and not Neural Network scientists. At the same time, understanding the underpinnings of why things work requires developing an understanding of the parallels between Machine Learning and model fitting, and of the role of optimisation algorithms in this. So I would also encourage people to spend time reading about the range of machine learning methods and models (I’m quite partial to Chris Bishop’s Pattern Recognition and Machine Learning as a starting book) and optimisation algorithms (e.g. Numerical Optimisation by Jorge Nocedal). AIM: What books and other resources have you used in your journey? Navdeep: To be honest, books aren’t often the best way to understand this field — most of the cutting  edge stuff is in the papers and the field is constantly moving. Nevertheless, the following books gave me real mileage in understanding the underlying concepts of Machine Learning: Chris Bishop — Pattern Recognition and Machine LearningRobert Tibshirani — The Elements of Statistical LearningTrevor Hastie— An Introduction to Statistical LearningDavid MacKay – Information Theory, Inference and Learning AlgorithmsPattern Classification by David G. Stork, Peter E. Hart, and Richard O. Duda A lot of people also find Deep Learning by Ian Goodfellow, Yoshua Bengio and Aaron Courville quite useful, but I haven’t read it myself, since I was mostly only reading papers by the time Ian and co-authors released the book.","excerpt":"“I asked Geoffrey Hinton to take me on as a student, but he said he was too busy. ” For this week’s ML practitioner’s series, Analytics India Magazine (AIM) got in touch with former Google Brain researcher Navdeep Jaitly. From leaving IIT to pursuing a Liberal Arts Degree, from switching between AI and computational biology […]","categories":["Global Tech"],"tags":["geoff hinton","University of Toronto","University of Waterloo"],"author_name":"Ram Sagar","publish_date":"2021-05-18T15:00:00","publication_year":"2021","word_count":2801,"keywords":["data science","artificial intelligence","geoff hinton","machine learning","AI","University of Toronto","University of Waterloo","ML","neural network","OpenAI","deep learning","analytics","conversational agents"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","neural network","data science","analytics","OpenAI","conversational agents"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/the-secret-life-of-navdeep-jaitly-how-this-ex-google-brain-researcher-fell-in-love-with-algorithms\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10133723,"title":"This AI Startup Wants to Fix Your Code Bugs. And Got YC Backing for It","content":"Getting into YC takes work. Amartya Jha, the co-founder and CEO of CodeAnt AI, narrated in a post how he and his co-founder Chinmay Bharti got rejected the first time because they couldn’t explain their product well to the investors. Despite this and the fear of getting blacklisted, Jha and Bharti made a 45-minute video explaining their product and sent it to YC  again. The following week, they managed to get another interview with YC and were finally selected! But what is so good about what they are building? “Developers spend 20 to 30% of their time just reviewing someone else’s code. Most of the time, they simply say, ‘It looks good, just merge it,’ without delving deeper,” Jha explained while speaking with AIM. This leads to bugs and security vulnerabilities making their way into production. With generative AI, coding is undeniably getting easier. At the same time, the quality of code produced by these AI generators is still far from those of the human coders. This makes code review crucial for every organisation and this is where CodeAnt AI comes into the picture. CodeAnt’s AI-driven code review system can drastically reduce these vulnerabilities by automating the review process and identifying potential issues before they reach the customer. Founded in October 2023, CodeAnt AI is already making waves in the industry by automating code reviews with AI. The journey began at Entrepreneur First, where Jha met Bharti. The company quickly gained traction, securing a spot in YC by November 2023. By February 2024, they had launched their first product, attracting major clients like Tata 1mg, India’s largest online pharmacy, and Cipla, one of the country’s biggest pharmaceutical companies. “These companies were amazed that such a solution even existed,” Jha recalled. “And both contracts were paid, not just trials.” What’s the Moat? What sets CodeAnt AI apart from other players in the market such as CodeRabbit and SonarSource, which joined the market before CodeAnt, is its unique approach to code review. The company has developed its own dataset of 30,000+ checks, meticulously created to address every possible code commit scenario. “This is our pure IP,” Jha said. “We wrote our own algorithms to analyse code, understand its flow, and identify areas that need improvement. Our AI, combined with our custom engineering solutions, runs these checks on every code commit, offering unprecedented accuracy.” The platform also supports more than 30 programming languages. While competition in the AI-driven code review space is growing, Jha remains confident in CodeAnt AI’s unique value proposition. Its biggest competitor is SonarSoruce, but interestingly, its lead investor is also one of the investors of CodeAnt. CodeRabbit relies solely on AI, which leads to a lot of hallucinations and false positives. “Our approach, which combines AI with deterministic policies, gives us a significant edge,” said Jha. “The demand for tools like ours is only going to grow as AI-generated code becomes more prevalent,”  Jha added that tools like GitHub Copilot or Cursor are far from generating accurate code anytime soon. Jha further elaborated on the limitations of competitors that use AI exclusively. When developers give a large codebase to AI, it tends to hallucinate. To mitigate this, CodeAnt built a foundation of hard-coded checks, which are further enhanced by AI. This reduces false positives and ensures that the code is not only correct but also secure and compliant with industry standards. Amartya Jha speaking at Bangalore Python Meetup Stands Out with Customisation One of CodeAnt AI’s standout features is its ability to allow enterprises to input their own data and create custom policies. “For example, Tata 1mg has written hundreds of custom policies on our platform. Before CodeAnt AI, they would have had to build a similar platform themselves to enforce these policies,” Jha said, and added that Tata 1mg has written code in Python for the last eight years. Now, they can simply use CodeAnt platform to ensure that their code complies with their specific guidelines, which is especially valuable for large organisations with complex codebases. This level of customisation is not only beneficial for code quality but also for compliance with security frameworks. “Imagine a tool that knows exactly how code should be written in your organisation and can flag issues related to SOC 2 compliance, GDPR, or data privacy. It makes the process of maintaining compliance much easier and more efficient,” Jha added. In the future, Jha will focus on expanding CodeAnt AI’s capabilities, particularly in the areas of security and code quality. “We’re currently working with some of the largest security and backup companies in the world, helping them review their code for vulnerabilities. Our team may be small, but we’re making a big impact,” he said. Despite its San Francisco roots, CodeAnt AI maintains a strong presence in India, with Jha overseeing operations from Bangalore. The company is poised for further growth, with plans to expand its product offerings and deepen its market presence. “We are going deeper into security and code quality. We will soon be making announcements about new products and partnerships, so stay tuned,” Jha hinted.","excerpt":"Tata 1mg has written hundreds of custom policies on CodeAnt platform.","categories":["Deep Tech"],"tags":["AI Startups"],"author_name":"Mohit Pandey","publish_date":"2024-08-26T10:19:46","publication_year":"2024","word_count":845,"keywords":["Go","AI","Git","Python","Aim","programming_languages:Python","generative AI","GAN","GitHub","R","AI Startups"],"extracted_tech_keywords":["AI","generative AI","Aim","Python","R","Go","Git","GitHub","GAN","programming_languages:Python"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/this-ai-startup-wants-to-fix-your-code-bugs-and-got-yc-backing-for-it\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10099289,"title":"Google Cloud Tames Llama 2 with RLHF","content":"At the recent Google Cloud Next event in San Francisco, Google surprised everyone by announcing that they’re offering Llama 2 and Code Llama from Meta, as well as Falcon LLM on Google Cloud’s Vertex AI. This was unexpected because Google was the only cloud service provider that hadn’t partnered with rival institutions to host Llama 2 or any other open source LLM models before this. It appears this decision by Google has been taken keeping enterprises in mind who are staple customers of Vertex AI but are looking for more options. If we go by the trend, after GPT-4, Llama 2 is the most sought after large language model, considering it is open-sourced and commercially available. In the case of Llama 2, Google said that it is the only cloud provider offering both adapter tuning and RLHF. Despite being ad rivals, Meta and Google have put aside their competition when it comes to large language models. Meta directly does not want to compete with anyone in the LLM business and is happy to provide Llama 2 to everyone. Now with Google Cloud having Llama 2, Meta has conquered every territory possible. In fact, didn’t Meta just reverse the saying: “If you are good at something, never do it for free”? However, Google accepting Llama 2 raises one question: Is PaLM 2 not capable enough? Google Bard currently uses PaLM 2 and it seems like it isn’t a favourite among enterprises as they cannot customise it according to their requirements, in addition to its poor responses when compared to ChatGPT. The tech giant claims that its Model Garden has a collection of 100+ models including enterprise-ready foundation model APIs, open source models, and task-specific models from Google and third parties. Google should understand that when it comes to LLMs, it’s not about the quantity but about the quality. Recently OpenAI also took cues from Meta’s approach and is now working to provide customisation options for GPT-4 and GPT-3.5 while avoiding open sourcing. To achieve this, the creator of ChatGPT recently introduced the GPT-3.5 Turbo API for fine-tuning. Additionally, it has partnered with Scale to fine-tune GPT-3.5, in order to woo enterprises. LLM Cloud Battle Begins Google might have been late to the game, but there is still hope, following the AWS route. Google has understood the importance of hosting multiple LLMs, much like Amazon’s Bedrock platform. Currently, Bedrock hosts models from AI21, Cohere, Anthropic Claude 2, and Stability AI SDXL 1.0. At present, Microsoft is actively exploring different LLMs. Microsoft Azure currently encompasses all OpenAI services via APIs, including the Azure OpenAI Service. This empowers enterprises and developers to create applications using GPT, DALL·E, and Codex. When Llama 2 was launched by Meta, Azure was announced as the preferred partner for Llama 2. It seems like Microsoft is not going to stop here, as it further plans to sell a new version of Databricks’ software on Azure that will help customers make AI apps for their businesses. This new service will help companies make AI models from scratch or repurpose open-source models as an alternative to licensing OpenAI’s proprietary ones. In the latest quarter, Azure emerged as the winner with 26% revenue growth, thanks to Azure OpenAI services. However, now with Llama 2 being the common factor among all three clouds, it will be intriguing to witness who will lead the LLM cloud game in the upcoming quarter. What about Gemini? As Vertex AI now hosts Llama 2 and has shifted its focus to smaller models, similar to the approaches of Microsoft and AWS, it raises the question of the feasibility of creating Gemini to take on GPT-5. This consideration is particularly relevant as OpenAI has also redirected its focus towards serving enterprises. Not to forget, Google’s Codey has a new rival called Code Llama, only time will tell who codes better, alongside its adoption among the enterprise customers and developers.","excerpt":"Not to forget, Google’s Codey has a new rival called Code Llama","categories":["Global Tech"],"tags":["Gemini"],"author_name":"Siddharth Jindal","publish_date":"2023-08-31T11:09:53","publication_year":"2023","word_count":650,"keywords":["Anthropic","ChatGPT","Gemini","RLHF","OpenAI","GPT-5","AI","Google Bard","AWS","Aim","Azure"],"extracted_tech_keywords":["AI","GPT-5","ChatGPT","OpenAI","Anthropic","Google Bard","Aim","RLHF","AWS","Azure"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/google-cloud-tames-llama-2-with-rlhf\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10125015,"title":"For Infosys, Generative AI is Not Just Fluff Talk","content":"Nandan Nilekani, the co-founder of Infosys has revealed that its early investment in generative AI, through its Infosys Topaz platform, has positioned it as a leader in AI services, with recognition from seven out of eight leading analysts. “Infosys is fully prepared to deliver value,” Nilekani stated. “We made an early investment last year in building a strong generative AI offering portfolio through Infosys Topaz. Today, we are ranked as a leader in AI services by seven out of eight leading analysts.” Infosys is currently managing over 225 generative AI programs for its clients. A key aspect of executing these complex transformations is talent. The acquisition of Danske IT and Support Services in India has strengthened Infosys’ digital talent pool. “We have invested significantly in hiring talent with proven generative AI skills as well as rapidly upskilling our existing engineering talent,” Nilekani emphasised. Infosys now boasts over 250,000 employees trained in generative AI. The integration of generative AI components into all service lines and the development of 25 playbooks have allowed Infosys to create significant impacts for its clients. By combining AI with cloud capabilities via Infosys Cobalt, clients are scaling their AI operations more effectively. This effort is part of a broader strategy to drive exponential growth in AI and advance the company’s Chip-to-Cloud strategy. To bolster this strategy, Infosys acquired InSemi, a semiconductor design services provider, enhancing its domain-relevant enterprise AI capabilities. “We have created 23 AI industry blueprints to solve industry-specific challenges,” Nilekani noted. Strategic acquisitions, such as that of in-tech, an engineering R&D services firm, further deepen Infosys’ capabilities, particularly in the automotive sector with software-defined vehicles. The Love for Indian Developers Additionally, Infosys is one of the largest adopters of GitHub Copilot globally, with employees generating over three million lines of code using generative AI large language models. “We’re working on projects across software engineering, process optimisation, customer support, advisory services, and sales and marketing,” said Salil Parekh, Infosys CEO, during the company’s last quarterly call. “We’re working with market-leading open access and closed large language models,” he added, saying that Infosys feels good about its work with generative AI. Just this month, in a significant move towards accelerating digital transformation in India, GitHub has partnered with Infosys to launch the first GitHub Centre of Excellence in Bangalore. This initiative aims to leverage AI and advanced software solutions to drive global economic growth.","excerpt":"Infosys is currently managing over 225 generative AI programs for its clients, and has upskilled 2.5 lakh employees with generative AI.","categories":["AI News"],"tags":["Generative AI","Infosys"],"author_name":"Mohit Pandey","publish_date":"2024-06-27T11:34:39","publication_year":"2024","word_count":398,"keywords":["Go","API","Infosys","AI","digital transformation","Git","RAG","Aim","generative AI","Generative AI","GitHub","R"],"extracted_tech_keywords":["AI","generative AI","Aim","RAG","R","Go","Git","GitHub","API","digital transformation"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/for-infosys-generative-ai-is-not-just-fluff-talk\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":62426,"title":"Pine Labs Gives Developers Access To its ePOS APIs To Build Fintech Products","content":"Pine Labs, one of Asia’s leading merchant commerce platforms, has announced that it has opened the application programming interface (APIs) of its Android-based remote payment application – ePOS by Pine Labs for developers. Pine Labs’ ePOS allows merchants to download the app and convert their Android phone into a payment acceptance device for cards, UPI and wallets. Currently, popular UPI and wallet payments including Amazon Pay, Airtel Money and PhonePe are supported on ePOS by Pine Labs. Developers can build innovative fintech products for their customers in retail, grocery, transportation space using the remote payment APIs of ePOS. Sanjeev Kumar, Chief Technology Officer, Pine Labs said, “Our ePOS app has got a positive response from the merchant community. In less than a month of making it available on our website and Google Play Store, we have seen nearly 14,000 downloads. It uses our stringent payment protocols that have been the hallmark of our offline payment products. With the need for contactless and remote payment modes, we invite developers in all industry sectors to use it to provide value-added services to their customers. They need not worry about payment-related regulatory compliance and security features as these are already embedded in ePOS.” Developers can use the open APIs in many different scenarios. ePOS APIs may be integrated into home delivery applications, share economy applications, or paid productivity solutions. These may also be pulled into chat apps to enable payments via chat applications. Pine Labs has a unique cloud-based platform enables it to offer a wide range of payment acceptance and merchant commerce solutions. Its stored value platform includes issuing, processing and distributing digital gift cards for corporate customers. Its customer base includes prominent large, mid-sized and small merchants across India, South East Asia and the Middle East. Pine Labs’ key investors include Sequoia India, Mastercard, Actis Capital, Temasek, PayPal and Sofina. The open APIs of ePOS by Pine Labs can be accessed here – https:\/\/developer.pinelabs.com\/android\/epos\/docs ePOS by Pine Labs is available on Google Play Store for download. Merchants need to go through a five-step quick set up process before accepting payments.","excerpt":"Pine Labs, one of Asia’s leading merchant commerce platforms, has announced that it has opened the application programming interface (APIs) of its Android-based remote payment application – ePOS by Pine Labs for developers. Pine Labs’ ePOS allows merchants to download the app and convert their Android phone into a payment acceptance device for cards, UPI […]","categories":["AI News"],"tags":["APIs","Developers","FinTech"],"author_name":"Vishal Chawla","publish_date":"2020-04-23T11:27:33","publication_year":"2020","word_count":349,"keywords":["Go","API","programming_languages:R","AI","programming_languages:Go","Git","APIs","ViT","FinTech","R","Developers"],"extracted_tech_keywords":["AI","R","Go","Git","API","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/pine-labs-gives-developers-access-to-its-epos-apis\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10039424,"title":"Twitter Rolls Out SOS Resources, Fact Pages To Fight The Second Wave","content":"A few days ago, Twitter announced that the platform has added some of the important features that could help the citizens of India find exactly what they are looking for during the second wave. The microblogging platform released the advanced search feature for users, which aims to help users customise their search for information and other such. In the tweet, the microblogging platform stated, “We know there’s a lot of information on Twitter right now, and not all of it is reliable. The COVID-19 page on the Explore tab (tap the magnifying glass icon) is here to help guide you to more timely and reliable information. Here’s how to use it Thread #Covid19IndiaHelp” https:\/\/twitter.com\/TwitterIndia\/status\/1386608572377694210 According to sources, Twitter explained that as COVID-19 vaccinations become more widely available, the social media platform wants its users to have access to the latest vaccine info in the country. Users will see a prompt in their timeline that links to sources about vaccine safety, efficacy, and news from public health experts. The platform mentioned some of the new additional features- COVID-19 SOS: Resources The SOS Resources aims to help users get information about access to hospital beds, oxygen, food. The platform tweeted, “If you need a list of SOS resources – access to hospital beds, oxygen, food – please look at the COVID-19 SOS: Resources page. This will update in real-time as people put out their emergency Tweets.” If you need a list of SOS resources – access to hospital beds, oxygen, food – please look at the COVID-19 SOS: Resources page. This will update in real-time as people put out their emergency Tweets. pic.twitter.com\/D40Zx6Nk1z— X India (@XCorpIndia) April 26, 2021 COVID-19: Know the facts This feature will help the users to receive information that is verified by official sources. Twitter tweeted, “Confused by all the conflicting information around vaccines? Only trust the experts. With the COVID-19: Know the facts page, we have made it easier for you to find information verified by official sources.” Confused by all the conflicting information around vaccines? Only trust the experts. With the COVID-19: Know the facts page, we have made it easier for you to find information verified by official sources. pic.twitter.com\/XFNo4BaKAZ— X India (@XCorpIndia) April 26, 2021 COVID-19 India LIVE This page provides information about the deadly virus in real-time and in an unfiltered manner. In a tweet, the platform said, “If you need to know everything that’s unfolding about COVID-19 in real-time and unfiltered, the COVID-19 India LIVE page is what you’re looking for.” If you need to know everything that’s unfolding about COVID-19 in real-time and unfiltered, the COVID-19 India LIVE page is what you're looking for. pic.twitter.com\/DDwU3R5QlA— X India (@XCorpIndia) April 26, 2021 COVID-19: How to Protect Yourself This page helps users to understand the measures and methods one must take in order to slow the spread of the virus. The platform tweeted, “The best thing for all of us to do right now is #StayHome and #StaySafe. How can you personally take responsibility to slow the spread? We have some tips and thoughts for you on the COVID-19: How to Protect Yourself page.” The best thing for all of us to do right now is #StayHome and #StaySafe. How can you personally take responsibility to slow the spread? We have some tips and thoughts for you on the COVID-19: How to Protect Yourself page. pic.twitter.com\/BtaAZ3doxh— X India (@XCorpIndia) April 26, 2021","excerpt":"A few days ago, Twitter announced that the platform has added some of the important features that could help the citizens of India find exactly what they are looking for during the second wave. The microblogging platform released the advanced search feature for users, which aims to help users customise their search for information and […]","categories":["AI News"],"tags":["Twitter (X)","twitter india"],"author_name":"Ambika Choudhury","publish_date":"2021-05-03T13:21:41","publication_year":"2021","word_count":569,"keywords":["Go","twitter india","programming_languages:R","AI","programming_languages:Go","Aim","Rust","Twitter (X)","R","programming_languages:Rust"],"extracted_tech_keywords":["AI","Aim","R","Go","Rust","programming_languages:R","programming_languages:Go","programming_languages:Rust"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/twitter-rolls-out-sos-resources-fact-pages-to-fight-the-second-wave\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10041366,"title":"Dr Prithwis Mukerjee&#8217;s Chrono-Series: A Heady Mix Of Advaita Philosophy, Time Travel &#038; Anthropomorphism Of AI","content":"“I do not fear computers, I fear the lack of them” –Isaac Asimov Like the famous American writer Isaac Asimov, Dr Prithwis Mukerjee too considers machines his friends. Dr Mukerjee says he feels most at ease while working with machines. Given the shared love for computers and machines, it hardly comes as a surprise that Asimov is also his favourite science fiction writer. After a successful stint in the corporate world, working with big names such as IBM, PricewaterhouseCoopers, and Tata Steels, Dr Mukerjee pivoted to academics. He is currently the director of Praxis Business School, Kolkata. In 1999, when the internet craze was just kicking in, Dr Mukerjee started his blog Yantrajaal, one of the first 500 blogs in India in all likelihood. Dr Mukerjee has written a few books on AI and data science, the most recent being Deep Learning for Business Managers. His latest, the Chrono series trilogy, is a ‘labour of love’. The first in the series is Chronotantra (the law of time) followed by Chronoyantra (the time machine). The third novel Chronomantra is still in the ideation phase. The Chrono-series “In 2015 I was contributing a monthly column at an online website on the impact of technology on society. While doing the research for the same, I accumulated a lot of material on a wide range of technology from space through bio-hackers and far-out physics and, of course, robotics and AI. While going through all this, I realised I was getting novel ideas that unfortunately I had no way of verifying or validating on my own. I could not put them in the articles because they had to be based on proven facts. So, I decided to collect these half-baked ideas and create a science fiction novel where I could take liberties with facts,” said Dr Mukerjee. The first book Chronotantra was released in 2019, followed by Chronoyantra which was completed by the end of 2020. Written in simple language, both books explore technological themes through the lens of Advaita Vedanta—a dominant Indian school of thought espousing the idea of ‘Brahman’. “Actually, science and philosophy are but two sides of the same coin. Newton named his famous book, Philosophiæ Naturalis Principia Mathematica (Mathematical Principles of Natural Philosophy) and the degree that I have is Doctor of Philosophy. In Indic thought, philosophy is referred to as Darshan or observation which is a key element of modern science. So, it is not difficult to combine these two subjects at all. Both are enquiries into the nature of the universe and its mysteries,” Dr Mukerjee said “Why do I use the word Chrono in all my books? Because Chronos or time is the most fundamental construct of the universe. For anything to exist, it must persist and persistence can only happen inside time. Hence, time precedes everything including space and is the most fundamental artefact or concept from which everything else can be built,” he added. Chronotantra is set in 2150. AI has taken over the world. The main protagonist of the story is Lila, a Santhal tribal woman from Rajrappa Hills, who rises through the ranks to become a brilliant engineer. The story follows her to the shores of the Caspian Sea, Hangzhou (China), Townsville on the edge of the Grand Canyon, Gandhar, and eventually to the planet Mars. Lila has been chosen by AI for fulfilling a cosmic mission. Without her knowledge, AI has looked after her, identified her weakness, and guided her to ultimate destiny. The second novel Chronoyantra is both a sequel and prequel to Chronotantra, said Dr Mukerjee. “Chronoyantra has two parallel stories running along–what happened before and what happened after. The way these two come together depicts the fabric of time. My books are about time. How time overlaps is something I have tried to put together. So, situations that were separated in time somehow collide and come together,” said Dr Mukerjee. The writing process Dr Mukerjee aims to dispel the myth that AI is a threat to civilization. Instead, he wants readers to think of AI as a friend. “Walter Lippman coined a famous phrase, ‘The world outside and the pictures in our heads.’ Each of us has an inbuilt image of the world that helps us navigate through it in a way that we think best. AI has another image like that, created by the data or information on which it is based. So just like there are good, nice, bad and evil persons with whom we have to live within this world, there will be good, nice, bad and evil AI with whom we will have to live with in the world that awaits us. In the world created in my novel, I see, show and deal with a nice AI who I have portrayed as a friend of the protagonists in my novel,” he said. Dr Mukerjee also revealed his research process for writing these novels. “These books are fictional stories, so apart from the tech basics, I took my liberty as an author and based it on my imagination and creativity. However, I extensively referred to Google Maps since my story is spread across multiple countries. I spent a lot of time understanding intricacies such as the location of buildings, lanes, view from a certain point, the direction of the wind at a given location and time. I also referred to Google Mars to understand the red planet’s landscape. I can say I can visualise what Rovers on Mars would be viewing.” Wrapping up Dr Mukerjee is a fan of Asimov’s style of writing and storytelling in science fiction. While there are many sci-fi writers whose work he admires, he feels one area that needs to be improved is logical plausibility. “I have no problems with a man flying through the sky (which is impossible today), but I am disturbed when I find logical gaps, say, I am shopping online and the next moment I am in Delhi killing a dinosaur. There has to be some logical connection between my online shopping and the killing of dinosaurs. When I see such things in sci-fi novels, I get very disturbed. I try to make sure that I connect all the dots correctly,” he said. For more information about these two sci-fi novels, visit http:\/\/bit.ly\/chronobooks. Both books are available in paperback or kindle at http:\/\/bit.ly\/chronobuy","excerpt":"“In Indic thought, philosophy is referred to as Darshan or observation which is a key element of modern science.”","categories":["AI Features"],"tags":["Interviews and Discussions"],"author_name":"Shraddha Goled","publish_date":"2021-06-07T12:00:00","publication_year":"2021","word_count":1056,"keywords":["data science","Go","AI","BERT","Ray","Aim","deep learning","ViT","GAN","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","deep learning","data science","Aim","Ray","R","Go","BERT","GAN","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/dr-prithwis-mukerjees-chrono-series-a-heady-mix-of-advaita-philosophy-time-travel-anthropomorphism-of-ai\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10039709,"title":"Demystifying AdaBoost: The Origin Of Boosting","content":"Introduced by Yoav Freund and Robert E. Schapire in their paper “A Short Introduction to Boosting”,  AdaBoost was the first successful boosting algorithm; all modern boosting algorithms build upon the success of boosting theory employed by AdaBoost. The premise of this theory is simple; combine multiple weak classifiers to create one strong classifier. This boosting approach can be used to boost the performance of any machine learning algorithm. But it’s best paired with weak learners, which are models that perform just slightly better than a random guess on classification problems. The most commonly used weak learners are decision trees with a depth of one, also known as decision stumps. This article will go through the boosting theory and then implement Adaboost from scratch. Before we delve into the mathematics and code, let’s understand the gist of Adaboost with the help of a simple 2D example: Source: http:\/\/sli.ics.uci.edu\/Classes\/2015W-273a?action=download&upname=09-ensembles.pdf The first classifier makes the classification based on the y-axis feature and creates a horizontal decision line at some threshold. This decision line is by no means perfect, and based on its predictions, a performance score is calculated. The performance measure is then used to calculate and update weights for all training samples. Moreover, the second classifier uses these weights and creates a different and maybe better decision line. The performance measure of the second classifier is used to update the weights, which are then used by the third classifier.  This process is repeated for the pre-set number, n_estimators, of classifiers. All the classifiers are combined through a weighted sum; here, the classifiers with better performance have a higher weight. This means that better classifiers have more impact on the final decision boundary. Briefly put, Adaboost trains a set of weak learners sequentially. The errors of the earlier learners are used to tag the harder examples that the later learners focus on. In the end, the complete set of learners is combined and converted onto one complex classifier. Decision Stump For the first iteration, the weights are initialized to 1\/N and the error is calculated as misclassifications\/N, where N is the number of training examples. This is later updated to include the weights and becomes ???? (wi × errori)\/????wi , where wi and errori denote the weight and prediction error of the i-th training example. The error is 1 if misclassified and 0 if classified correctly. The performance measure of the learners, denoted by ????, is calculated as 0.5 × log(1 – error\/error). This performance score is then used to update the weights using w = w × e -????  × y_true × y_pred, where y_true and y_pred are the actual and predicted classes. Finally, for making predictions a weighted sum of all classifiers’ prediction is calculated: ???? ????i × y_predi and the sign of the sum is used to make the classification. Here ????i and y_predi refer to the performance score and prediction of the i-th classifier. Implementing Adaboost From Scratch Before we can code Adaboost, we need to implement the weak learner. Each decision stump will have four attributes: threshold – the threshold that differentiates the two classes based on one featurealpha – the performance measure, used as its weight in the final predictionindex – the index of the feature the classifier is built onpolarity – used to invert weak classifiers that perform worse than random(0.5). class DecisionStump(): def __init__(self): self.threshold = None self.alpha = None self.index = None self.polarity = 1 In addition to these attributes, it will also need a function for making predictions. def predict(self, X): n_samples = X.shape[0] X_column = X[:, self.index] predictions = np.ones(n_samples) # classifying based on the polarity and threshold if self.polarity == 1: predictions[X_column < self.threshold] = -1 else: predictions[X_column > self.threshold] = -1 return predictions Using this DecisionStump class we can start working on Adaboost. It will have two attributes: n_estimators – number of decision stumps to useensemble_of_stumps – a collection of trained decision stumps class Adaboost(): def __init__(self, n_estimators=5): self.n_estimators = n_estimators self.ensemble_of_stumps = [] Now comes the hard part, the training method that creates the ensemble of decision stumps. We initialize the weights to 1\/N and start training the first stump. def fit(self, X, y): N, n_features = X.shape weights = np.full(N, (1 \/ N)) for _ in range(self.n_estimators): stump = DecisionStump() Each of the decision stumps will use one feature to make the prediction, and it will use a threshold to make classifications. To ‘train’ the stump we need to find the best possible feature-threshold combination. To achieve this we iteratively go through all features, and for each feature, we test all of its unique values as the threshold. In the end, the feature-threshold combination with the lowest error is selected. min_error = float('inf') # looking for the best threshold and feature configuration for i in range(n_features): X_column = X[:, i] thresholds = np.unique(X_column) for threshold in thresholds: p = 1 predictions = np.ones(N) predictions[X_column < threshold] = -1 # Calculating error misclassified = weights[y != predictions] error = sum(misclassified) If the error of a stump is less than 0.5, we reverse its polarity. This is done because, by definition, the weak learners need to perform better than random. # if error is less than .5, reverse polarity of learner if error > 0.5: error = 1 - error p = -1 # Storing the best configuration for the current stump if error < min_error: stump.polarity = p stump.threshold = threshold stump.index = i min_error = error Using the error of the stump, we calculate its performance measure alpha. Which is then used to update the weights for the next stump. stump.alpha = 0.5 * np.log((1.0 - min_error) \/ (min_error)) weights  = weights * np.exp(-stump.alpha * y * stump.predict(X)) # Normalizing the weights to one weights \/= np.sum(weights) Finally, we add the current decision stump to Adaboost’s ensemble of weak learners. self.ensemble_of_stumps.append(stump) In addition to the training method, we need a method for making predictions. This method will calculate the weighted sum of all stump’s predictions and use the sign of the sum as the final prediction. def predict(self, X): # weighted sum of all stumps stump_preds = [stump.alpha * stump.predict(X) for stump in self.ensemble_of_stumps] y_pred = np.sum(stump_preds, axis=0) y_pred = np.sign(y_pred) return y_pred Let’s test our Adaboost implementation on the breast cancer dataset and compare it with sklearn’s implementation. import time from sklearn import datasets from sklearn.metrics import accuracy_score from sklearn.model_selection import train_test_split X, y = datasets.load_breast_cancer(return_X_y=True) #changing class 0 to -1 y[y == 0] = -1 X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42) start = time.perf_counter() classifier = Adaboost(n_estimators=5) classifier.fit(X_train, y_train) y_pred = classifier.predict(X_test) end = time.perf_counter() accuracy = accuracy_score(y_test, y_pred) print (\"Accuracy:\", accuracy) print(f'Finished in {round(end-start, 2)} second(s)') -----------------------------Output----------------------------- Accuracy: 0.9649122807017544 Finished in 5.48 second(s) from sklearn.ensemble import AdaBoostClassifier start = time.perf_counter() sk_classifier = AdaBoostClassifier(n_estimators=5, random_state=42) sk_classifier.fit(X_train, y_train) sk_y_pred = classifier.predict(X_test) end = time.perf_counter() sk_accuracy = accuracy_score(y_test, y_pred) print (\"Accuracy:\", sk_accuracy) print(f'Finished in {round(end-start, 2)} second(s)') -----------------------------Output----------------------------- Accuracy: 0.9649122807017544 Finished in 0.02 second(s) Our implementation can match sklearn’s implementation in terms of accuracy but is rather slow. The Colab notebook for the above implementation can be found here. Last Epoch This article deconstructed boosting and implemented Adaboost. I hope this gives you a good understanding of the boosting approach and helps you understand modern boosting algorithms more easily. Boosting has many advantages; it is fast and easy to program. Excluding the number of weak learners, n_estimators, it has no parameters to tune. One thing to keep in mind when working with boosting algorithm is the data. Because the boosting method obsessively tries to correct its errors, any outliers or noise can send the model down a tangent of trying to learn unrealistic or wrong examples. References PaperSlides from the University of California CS273a courseThis article by Jason Brownlee","excerpt":"AdaBoost was the first successful boosting algorithm that combined multiple weak classifier to create a strong classifier.","categories":["AI Trends"],"tags":["Boosting","Boosting Algorithm","Classification"],"author_name":"Aditya Singh","publish_date":"2021-05-06T17:00:00","publication_year":"2021","word_count":1294,"keywords":["Go","Classification","Boosting Algorithm","machine learning","TPU","API","AI","programming_languages:R","Boosting","BERT","Colab","llm_models:BERT","R"],"extracted_tech_keywords":["AI","machine learning","Colab","TPU","R","Go","API","BERT","llm_models:BERT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/demystifying-adaboost-the-origin-of-boosting\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10161394,"title":"Former Facebook Exec Now Heads Microsoft’s New AI Engineering Team","content":"Microsoft has announced the creation of a new engineering division, CoreAI – Platform and Tools, combining its Developer Division, AI platform teams, and staff from the Office of the Chief Technology Officer. This new group will focus on building AI platforms and tools for Microsoft and its customers, marking a major shift to centre AI development in its operations. This new division will be led by Jay Parikh, former Facebook engineering chief, who joined Microsoft in October 2024. According to details in the blog, as Executive Vice President of CoreAI – Platform and Tools, Parikh will report directly to CEO Satya Nadella and join the senior leadership team. In his role, he will oversee leaders including Eric Boyd (AI platform), Jason Taylor (Deputy CTO of AI Infrastructure), Julia Liuson (Developer Division), and Tim Bozarth (Developer Infrastructure). An AI-First Application Stack CEO Satya Nadella described the reorganisation as preparing for the “next innings of this AI platform shift” in 2025. He said this transformation will reshape software development, accelerating decades of change into just a few years. Nadella underlined an AI-first application stack that will change how developers create and deploy AI applications. Azure will provide the infrastructure, while Azure AI Foundry, GitHub, and Visual Studio Code will support advanced AI tools and agents. “Our AI platform and tools will create agents, and these agents will change every SaaS application category, with custom applications driven by software as a service,” he added. Scope of CoreAI The CoreAI group will develop an end-to-end Copilot and AI stack for first-party and third-party customers to create and run AI applications. It appears that Nadella’s memo highlights Azure AI Foundry, GitHub, and Visual Studio Code, but excludes Visual Studio and .NET, indicating a shift to cloud-first, AI-focused tools.","excerpt":"This new division will be led by Jay Parikh, former Facebook engineering chief, who joined Microsoft in October 2024.","categories":["AI News"],"tags":["AI (Artificial Intelligence)"],"author_name":"Aditi Suresh","publish_date":"2025-01-14T00:18:50","publication_year":"2025","word_count":293,"keywords":["Go","AI-first","cloud_platforms:Azure","AI","Azure","Git","Julia","GAN","GitHub","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","Azure","R","Go","Julia","Git","GitHub","GAN","AI-first","cloud_platforms:Azure"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/former-facebook-exec-now-heads-microsofts-new-ai-engineering-team\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10132590,"title":"Elon Musk’s Robotaxis are Built for Riders, Not for Drivers","content":"While the initial launch date for Tesla’s self-driving cab Robotaxi has been pushed, Tesla chief Elon Musk remains optimistic. Tesla owners will be able to transform their vehicles into Robotaxis, allowing their cars to generate income, much like an “Airbnb on wheels.” Robotaxi Vision The Robotaxi service aims to make Tesla drivers add their vehicles into the cab fleet, a little different from other autonomous cab providers such as Waymo or Baidu. “We’ll have a fleet that’s on the order of 7 million that are capable of autonomy. In the years to come it will be over 10 million and 20 million. This is immense,” said Tesla chief Elon Musk. “The car is able to operate 24\/7 unlike the human drivers,” he said. However, Tesla’s concept of Robotaxi has been recently questioned by Uber CEO Dara Khosrowshahi, in a recent interview. “It’s not clear to me that the average person,Tesla owner, or owner of any other car is going to want to have that car be ridden in by a complete stranger,” he said. Shared Revenue Similar to your third-party cab provider, such as Uber, Tesla is also looking to have a shared revenue format with the Robotaxi owners. Musk also highlighted how Robotaxi will give the luxury to the users to choose the hours and schedule the hours of operation accordingly, thereby giving Robotaxi owners the choice to use it as both a personal and a commercial vehicle. Interestingly, Khosrowshahi even questioned the supply angle for this particular format of ride. “It just so happens that probably the times at which you’re going to want your Tesla are probably going to be the same times that ridership is going to be at a peak.” Thereby, hinting at how the demand and supply will not be met. Furthermore, he is also sceptical about the whole autonomous feature in vehicles.  “We’re seeing that when one of our customers is offered an autonomous ride, about half the time they say, yeah that would be really cool, and half the time they say, no thank you, I’d rather have a human I think that’s going to improve over a period of time,” he said. Even then, the Uber CEO has not denied a likely partnership with Tesla in the future. “Hopefully, Tesla will be one of those partners. You never know.” With numerous autonomous vehicles on the market, the key distinction lies in the different approaches each company has taken toward autonomous capabilities. LiDAR vs Vision Tesla uses computer vision (Tesla Vision) rather than the conventional LiDAR (Light Detection and Ranging) tech for autonomous vehicles. Musk has always been vocal about using vision-only methods for autonomous capabilities as opposed to Waymo and other self-driving cars that heavily rely on LiDAR. Previously, Musk had even called out LiDAR as a “fool’s errand” and that anyone relying on it is “doomed.” He even referred to Waymo’s robotaxi services as limited and fragile, and claims Tesla’s systems to work anywhere in the world, not limited by geography. The cost of LiDAR has been a major deciding factor for adopting sensors in autonomous vehicles. Musk considers LiDAR to be expensive sensors that are unnecessary. He believes that cameras, which Tesla banks on, will help them navigate through adverse weather conditions. Kilian Weinberger, professor of Computer Science at Cornell University, had earlier said that cameras are dirt cheap compared to lidar. “By doing this they can put this technology into all the cars they’re selling. If they sell 500,000 cars, all of these cars are driving around collecting data for them,” he said. While Tesla is heavily backing vision tech for AV, LiDAR is not completely out of the picture. Recently, Tesla purchased over $2 million worth of lidar sensors from Luminar. The company revealed that Tesla was its largest LiDAR customer in Q1. “At some point Tesla will pivot and adopt LiDAR. I think it is not an if question, but rather when,” a user speculated on Reddit. Self-Driving Cars on the Rise Driverless car services are witnessing a huge growth. Google’s parent Alphabet, recently announced an investment of $5 billion on its self-driving subsidiary Waymo. Currently, they operate in San Francisco, Phoenix, and Los Angeles and they will soon test them on the freeways of the San Francisco Bay Area. It was reported that Waymo is currently delivering 50,000 paid rides per week. Zoox, a subsidiary of Amazon, is also developing autonomous vehicles and is operational in certain cities in the US. Baidu’s autonomous fleet which is already running 6000 driverless rides per day in Wuhan (China) adopts a mix of technologies. As of April 2024, the cumulative test mileage of Apollo L4 has exceeded 100 million kilometres. “I think we are all at L4 today; and with the government regulation, it’s not possible to do L5. Another thing is that, I think, all of us providing this technology haven’t been tested in all the scenarios. We wouldn’t have this confidence to claim that we have the L5 capability,” said Helen K. Pan, general manager and board of directors for Baidu Apollo, California, in an earlier interaction with AIM. While Waymo and Baidu are at L4 level of autonomous capability, Tesla is still between L2 and L3. The Robotaxi unveiling event is currently planned for October 10. Musk is also positive of expanding Tesla’s self-driving technology to a wide market in the U.S. and internationally.","excerpt":"“We’ll have a fleet that’s on the order of 7 million that are capable of autonomy. In the years to come it will be over 10 million and 20 million. This is immense,” said Elon Musk.","categories":["Global Tech"],"tags":["Autonomous Vehicles","Tesla","Waymo"],"author_name":"Vandana Nair","publish_date":"2024-08-13T18:19:58","publication_year":"2024","word_count":899,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","computer vision","RAG","Aim","Waymo","Autonomous Vehicles","Tesla","R","ai_applications:computer vision"],"extracted_tech_keywords":["AI","computer vision","Aim","RAG","R","Go","programming_languages:R","programming_languages:Go","ai_applications:computer vision"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/elon-musks-robotaxis-are-built-for-riders-not-for-drivers\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":50433,"title":"Top 5 Cybersecurity Open-Source Projects That Can Protect Your Organisation","content":"Cybersecurity needs have increased over the years due to the rise of cyberattacks on small and large organisations alike. Data breaches afflict businesses and, in many cases, lead to bankruptcy for companies because of financial and reputation losses. In fact, reports suggest that the average cost of a data breach is about $3.9 million. Consequently, a few leaders in the cybersecurity landscape have open-sourced security solutions to fortify breaches. Such initiatives especially help new businesses which are exposed to cyberattacks. Small and medium enterprises can leverage these open-source projects and gain a head start to start their security initiatives for safeguarding their businesses. Another advantage of open-source projects is that companies can customise it based on their requirements and guard their information. Here’s the list of top open-source projects to secure your business from cyberattacks: IBM Cloud Pak IBM Cloud Pak is a cybersecurity open-source project for hybrid systems, which will connect a wide range of security tools and on-premise and cloud systems, thereby eliminating the need for moving data from various sources. The software assists companies in automating the response to cyberattacks, shield the systems. With Cloud Pak, firms can gain insights from across their multi could IT environment. It unveils hidden threats and allows them to make risk-based decisions. The rise of multi-cloud environments among organisations is mainly due to the need for protecting sensitive data. Firms are alarmed due to ever-increasing cyberattack on public clouds; many firms have already lost a colossal amount of data from public clouds. However, with IBM Cloud Pak they can now respond quickly to security incidents with automation. OpenTitan In order to secure infrastructure, Google along with its partner has open-sourced OpenTitan to orchestrate need of cybersecurity from the ground up. The silicon root of trust (RoT) project can be used to integrate guidelines of data centre servers, storage, among others. This will ensure that the hardware infrastructure is in its intended and trustworthy state. It checks whether the system boots with the desired firmware and is not affected by malware. It provides cryptographically unique machine identity to help firms monitor the legitimacy of the systems. This increases transparency while maintaining flexibility in the way companies carry out their business processes. Vuls Vuls is an agentless vulnerability scanner for Linux\/FreeBSD that runs on the cloud, on-premise, Docker, among others. The idea behind the project is to automate the security and notify system administrators. Vuls uses CRON to manage vulnerability by automatically the scan across the organisations. It can even check the non-OS packages to find flaws in language libraries, frameworks, and others. Manually, performing and monitoring vulnerability is tedious and require many experts to effectively handle the complete processes. That’s why Vuls, one of the most popular cybersecurity on GitHub, is a must-have tool to prevent breaches. ModSecurity ModSecurity is an open-source cybersecurity project that performs real-time application security monitoring as well as facilitates access control. It also provides continuous passive security assessment, full HTTP traffic logging, and web application hardening. Unlike other HTTP logging applications, it enables users to log almost every event for keeping you informed about the requests that are made to access information. Besides, instead of monitoring the behaviour of external parties, it focusses on the behaviour of systems to find suspicious events and abnormality in it. Infer Due to their numerous advantages Java, C, Object-C, and C++, these programming languages have been an integral part of software developments. Thus, Facebook has open-source Infer – a static analysis tool for detecting bugs in software that use those languages. The tool helps in detecting critical bugs to find critical bugs to help firms prevent crash and poor performance in applications. Due to its robustness, the tool is used by many prominent technology companies such as Spotify, Mozilla, and Amazon, for finding loopholes in software.","excerpt":"Cybersecurity needs have increased over the years due to the rise of cyberattacks on small and large organisations alike. Data breaches afflict businesses and, in many cases, lead to bankruptcy for companies because of financial and reputation losses. In fact, reports suggest that the average cost of a data breach is about $3.9 million. Consequently, […]","categories":["AI Trends"],"tags":["Cybersecurity"],"author_name":"Rohit Yadav","publish_date":"2019-11-21T16:00:00","publication_year":"2019","word_count":635,"keywords":["Go","AWS","AI","docker","Git","RAG","C++","Rust","Cybersecurity","R","Java"],"extracted_tech_keywords":["AI","RAG","AWS","docker","R","Go","Rust","Java","C++","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-5-cybersecurity-open-source-projects-that-can-protect-your-organisation\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":67610,"title":"Top 12 Papers On Adversarial Learning At CVPR 2020","content":"Security in data science practices has always been one of the crucial concerns among organisations. With the increase of using machine learning and deep learning models, researchers have been trying to make these models secure and robust in every way possible. Adversarial learning helps in improving the performance of machine learning systems. Below here we have listed down the top 12 research papers on adversarial learning presented at Computer Vision and Pattern Recognition (CVPR 2020) Conference. (The list is in no particular order) 1| DaST: Data-Free Substitute Training for Adversarial Attacks About: In this paper, the researchers proposed a Data-free Substitute Training method, also called as DaST, which can obtain substitute models for adversarial black-box attacks without any real data. To achieve this, DaST utilises specially designed generative adversarial networks (GANs) to train the substitute models. The experiments demonstrated that the substitute models produced by DaST have the capability to achieve competitive performance compared with the baseline models trained by the same train set. Read the paper here. 2| Towards Verifying Robustness of Neural Networks Against A Family of Semantic Perturbations About: In this paper, a team of researchers from IBM Research and others proposed a model-agnostic and generic robustness verification approach known as Semantify-NN. The approach against semantic perturbations for neural networks. This proposed approach features semantic perturbation layers, also known as SP-layers, to expand the verification power of current verification methods beyond p-norm bounded threat models. The researchers further demonstrated how the SP-layers can be implemented and refined for verification based on a diverse set of semantic attacks. Read the paper here. 3| The Secret Revealer: Generative Model-Inversion Attacks Against Deep Neural Networks About: Researchers from UC Berkeley and others studied the model-inversion attacks, in which the access to a model is abused to infer information about the training data. In this paper, they focused on image data and proposed a simple yet effective attack method, termed the generative model-inversion (GMI) attack, which can invert DNNs and synthesise private training data with high fidelity. The end-to-end GMI attack algorithm is based on GANs and can reveal private training data of DNNs with high fidelity. Read the paper here. 4| A Self-Supervised Approach for Adversarial Robustness About: In this paper, the researchers combined the benefits of adversarial training and different input processing based defences approaches and proposed a self-supervised adversarial training mechanism in the input space. The approach can be deployed as a plug-and-play solution to protect a variety of vision systems including classification, segmentation and detection. Read the paper here. 5| Adversarial Vertex Mixup: Toward Better Adversarially Robust Generalisation About: In this paper, the researchers identified Adversarial Feature Overfitting (AFO), which may cause poor generalisation, and showed that adversarial training can overshoot the optimal point in terms of robust generalisation, leading to AFO in a simple Gaussian model. They proposed an Adversarial Vertex mixup (AVmixup), which is a soft-labelled data augmentation approach for improving adversarially robust generalisation. Read the paper here. 6| How Does Noise Help Robustness? Explanation and Exploration under the Neural SDE Framework About: In this paper, the researchers of Google and UC Davis proposed a new continuous neural network framework called Neural Stochastic Differential Equation (Neural SDE), which naturally incorporates various commonly used regularisation mechanisms based on random noise injection. They further demonstrated that the Neural SDE network can achieve better generalisation than the Neural ODE and is more resistant to adversarial and non-adversarial input perturbations. Read the paper here. 7| Unpaired Image Super-Resolution Using Pseudo-Supervision About: In this paper, the researcher proposes an unpaired super-resolution (SR) method using a generative adversarial network that does not require a paired or aligned training dataset. The network consists of an unpaired kernel, noise correction network and a pseudo-paired SR network. With the help of diverse datasets, the researcher showed that the proposed method is superior to existing solutions to the unpaired SR problem. Read the paper here. 8| Universal Litmus Patterns: Revealing Backdoor Attacks in CNNs About: In this paper, the researchers introduced a benchmark technique for detecting backdoor attacks, also known as Trojan attacks on deep convolutional neural networks (CNNs). They introduced the concept of Universal Litmus Patterns (ULPs), which enables one to reveal backdoor attacks by feeding these universal patterns to the network and analysing the output (i.e., classifying the network as ‘clean’ or ‘corrupted’). Read the paper here. 9| Robustness Guarantees for Deep Neural Networks on Videos About: In this paper, the researchers from the University of Oxford, UK considered the robustness of deep neural networks on videos, which comprise both the spatial features of individual frames extracted by a convolutional neural network (CNN) and the temporal dynamics between adjacent frames captured by a recurrent neural network. To measure robustness, they studied the maximum safe radius problem, which computes the minimum distance from the optical flow sequence obtained from a given input to that of an adversarial example in the neighbourhood of the input. Read the paper here. 10| Benchmarking Adversarial Robustness on Image Classification About: In this paper, the researchers established a comprehensive and coherent benchmark to evaluate adversarial robustness on image classification tasks. This benchmark can provide a detailed understanding of the effects of the existing methods under different scenarios, with a hope to facilitate future research. Read the paper here. 11| What It Thinks Is Important Is Important: Robustness Transfers Through Input Gradients About: In this paper, the researchers proposed a robustness transfer method that is both tasks- and architecture-agnostic with input gradient as the medium of transfer. The approach here is called input gradient adversarial matching (IGAM). They showed that the input gradients are an effective medium to transfer adversarial robustness across different tasks and even across different model architectures. Read the paper here. 12| Transferable, Controllable, and Inconspicuous Adversarial Attacks on Person Re-identification With Deep Mis-Ranking About: In this work, the researchers examined the insecurity of current best performing re-identification (ReID) models by proposing a learning-to-mis-rank formulation to perturb the ranking of the system output. They also developed a multi-stage network architecture to extract general and transferable features for the adversarial perturbations. Read the paper here.","excerpt":"Security in data science practices has always been one of the crucial concerns among organisations. With the increase of using machine learning and deep learning models, researchers have been trying to make these models secure and robust in every way possible. Adversarial learning helps in improving the performance of machine learning systems.  Below here we […]","categories":["AI Trends"],"tags":["Adversarial AI","Adversarial Learning","deep learning research","future of deep reinforcement learning","Reinforcement Learning"],"author_name":"Ambika Choudhury","publish_date":"2020-06-18T16:00:00","publication_year":"2020","word_count":1017,"keywords":["data science","Go","future of deep reinforcement learning","machine learning","Reinforcement Learning","TPU","AI","neural network","Adversarial AI","Adversarial Learning","computer vision","deep learning research","deep learning","GAN","R"],"extracted_tech_keywords":["AI","machine learning","deep learning","neural network","computer vision","data science","TPU","R","Go","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-12-papers-on-adversarial-learning-at-cvpr-2020\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":28592,"title":"Intelligent Analytics &#038; Digital Assistants Are Integral To Businesses, Says Sundar Srinivasan Of Microsoft India","content":"Analytics India Magazine caught up with Sundar Srinivasan, General Manager, Artificial Intelligence & Research team at Microsoft India, to understand AI’s impact and its industry applications across the sectors deploying AI-based solutions. Researchers at Microsoft are leveraging the AI tools for a range of services. Srinivasan, an industry veteran has vast experiences in building consumer products, cloud services at scale and building intelligent services using AI. Talking about the future of AI, he revealed how AI can help us in the future and what is its impact on human intellectual quotient. AIM: What are some of the practical implementations of AI that have revolutionised the way humans are functioning? What are some examples of AI products? SS: Artificial intelligence has been with us in the real world for a long time. The most visible application of AI is robotics or mechanization using precision robotics in manufacturing. In the recent past, there have been some breakthroughs in techniques like Deep Learning that have expanded the horizon of applications into speech, translation, vision etc. Speech recognition and translation are two examples that are front and centre now and we see them in our everyday life. For example, Bing Translate uses end to end neural models to provide translation services to 60 languages. Speech recognition is used for various commanding scenarios as in Cortana, where you can interact with her via voice. Image recognition has been used to detect product defects in the past, but it has now evolved into diagnosis, like a highly skilled doctor. Microsoft is working with LV Prasad Institute for prediction and progression of refractive error in children and young adults, and with Apollo for cardiac care. Image recognition is also making its way into consumer products. For instance, Windows Hello has been using face recognition and fingerprint readers for some time now, and recently this tech is migrating to phones as face unlock feature. These advances have also been used in predictions, from sales forecasting to forecasting sowing dates for Farmers, where for instance Microsoft is partnering with ICRISAT to predict the optimal sowing time for crops. Farmers are seeing nearly 30% yield increases using this tool. AI also holds the power to fundamentally reinvent how individual businesses run, compete and thrive. We’ve seen the emergence of AI solutions such as digital assistants and intelligent analytics which are now the core to the business. AI is fast becoming a ubiquitous part of our daily lives.  Governments and organizations world over are exploring unique ways to build products and enhance services using AI.  As a result, we have started to witness many sectors including agriculture, healthcare and education, utilizing the power of AI, resulting in their rapid progress.  Organizations and governments are coming together to ensure that the benefits of technology reach the citizens at large. AIM: How is AI affecting your life and work? What are some of the AI-based products or services that you rely on personally, the most? Please elaborate on the use cases. (eg chatbot, virtual assistants, phone apps) SS: There are so many applications, we use without us being cognizant of the fact they are derivatives of AI work.  As a daily user of Windows Hello, I have a Surface, you just open the laptop and it auto logs you in by recognizing your face.  Apart from that, I use Outlook.com for my email, it auto classifies spam and promotional emails. It also generates cards for emails and calendar entries for emails that have information on package delivery, travel (Airline, Hotels and Car Rentals). The other thing that is super useful in India is SMS organizer.  This Android app helps users get back time by focusing their attention on the most important things. SMS Organizer auto classifies the messages as personal, transactional and promotional messages. It also auto-generates reminders and actionable alerts like bills to be paid and flight alerts.  The best part is this is done entirely on the phone which is an excellent example of AI on the edge device. AIM: What are some of the ways that your company is adopting\/providing AI services? Please highlight some use-cases. What are the most important things for customers to be able to adapt to ML? Give us some examples of societal applications where you have used AI. SS: At Microsoft, we really focus on our mission to empower every person and every organization on the planet to achieve more. AI is one way to help people achieve more, consequently, we strive hard to drive value by infusing AI into our products and services, wherever it makes sense for users.  Now how do we allow this capability to be used by third-party customers and the software ecosystem at large? For customers and the developer ecosystem to be able to leverage ML, four things are needed: 1, Data Tools to leverage the data and learn from and build ML models Specialized machine environment to train the models An environment to host and run the models Microsoft allows users to either use our pre-trained models on Azure cognitive services or bring your own data, store it securely in your private environment, use specialized hardware to train it and then host the models to run it. Azure provides the full suite of tools to enable all of this. On the topic of how we use AI to solve societal challenges, allow me to highlight our work in agriculture, multi-language communication, education and healthcare. A few use cases are listed: AI for farming: Microsoft and ICRISAT announced the results of the second phase of the pilot of their AI-based Sowing App for farmers. The Sowing App was developed to help farmers achieve optimal harvests by advising on the best time to sow using data about weather conditions, soil quality and other indicators. Farmers can make the best use of this technology without having to incur any capital expenditure or install any sensors in their fields. The programme was expanded to touch more than 3000 farmers across the states of Andhra Pradesh and Karnataka during the Kharif season of 2017 for a host of crops including groundnut, ragi, maize, rice and cotton, among others. AI Network in Healthcare: In partnership with LVPEI, Microsoft had launched the Microsoft Intelligent Network for Eyecare (MINE) to apply Artificial Intelligence to help in the elimination of avoidable blindness and scale delivery of eye care services across the planet. This initiative harnesses the combined power of data, cloud, and advanced analytics to drive strategies to prevent avoidable blindness.  The initiative which had also been adopted by the state of Karnataka has now expanded into the AI Network for Healthcare to create an AI-focused network in cardiology, in partnership with Apollo Hospitals. The partnership will see work done towards developing and deploying new machine learning models to predict patient risk for heart disease and assists doctors on treatment plans. As a part of the Network, Microsoft recently announced its efforts in partnership with Forus Health to integrate AI-based retinal imaging APIs into Forus Health’s 3Nethra devices using Microsoft Azure IoT Suite, for early detection of diabetic retinopathy, glaucoma & macular degeneration, and help reduce avoidable blindness. This will help technicians identify eye fundus images as well as disease conditions better. Interactive Cane: MSR India is working on an AI powered Interactive Cane to aid people with visual impairment.  MSR is experimenting by adding sensors to existing canes and adding gesture recognition to enable the cane to provide the user with information that would not otherwise be available. Interestingly, MSR is trying to do this by using low resource sensors that are also intelligent, so that real-time information and feedback is available to the user. AI for local language computing: Microsoft has consistently been working towards providing local language computing in Indian languages, since the launch of Project Bhasha in 1998. With the help of its AI technologies, Microsoft is now making translation and speech recognition across several Indian languages in the following ways: Microsoft’s SwiftKey, allows text input in as many as 24 Indian languages and dialects including Marwari, Bodo, Santali and Khasi, utilizes AI in the keypads to enable faster predictive writing. It also allows mixed language typing in English and Hindi Indian English and Hindi speech recognition is available as part of Microsoft Cognitive Services as well as Bing App for Android Text to speech translations currently includes such capabilities in Hindi and Tamil on Microsoft Narrator, on Windows 10 Microsoft PowerPoint uses AI to translate full presentation decks from English to Hindi, Bangla and Tamil AIM: What are the areas of life or employment where you would like AI to be more involved (e.g. medicine, mental health care)? SS: Healthcare is an issue that we all face one way or another. This is a place where there is an opportunity to bring technology to the underserved and alleviate human suffering. Climate change is real and is something that is already impacting us and an area where AI could play a great role in dealing with the impact of these climatic changes, be it in better weather prediction or helping us manage agriculture better. AIM: Will AI take away the creative thinking and downgrade the intellectual quotient of humans? SS: We look at AI in an assistive role. While AI has the potential to disrupt every single vertical industry, it also promises to amplify our human ingenuity and help us be more productive. The promise of AI is that knowledge gained from applying analytics to the wealth of data that is available today will enhance any decision-making process with additional intelligence, helping us produce quicker, more effective outcomes. AIM: Many experts have warned against AI taking over every aspect of our lives. What is your take on that? How true is their fear? SS: Technology has fundamentally changed the way we consume news, plan our day, communicate, shop and interact with our family, friends and colleagues.  Over the next 2 decades, we envision personal digital assistants will be trained to anticipate our needs, help manage our schedule, prepare us for meetings, assist as we plan our social lives, reply to and route communications, and help drive both individual and organizational productivity.   However, it is upon us, the larger industry, governments, academia, business, civil society and other stakeholders, to work together to ensure that AI is developed in a responsible and ethical manner so that people will trust it and deploy it broadly, both to increase business and personal productivity and to help solve societal problems. AIM: How has been the adoption of AI in Indian scenario and in what areas is Microsoft helping in providing AI solutions? Technology adoption in India has advanced at a rapid pace. Thanks to the Digital India initiative, this adoption of technology is at a much faster pace today than a decade before. Governments, as well as enterprises, know the potential of AI across different spectrums. In the last few years, we have witnessed governments and companies come together and partnering, sharing technology to help each other. As a part of its efforts, Microsoft has been partnering with governments and companies across levels, and over the last one year, has deployed AI-based solutions in the areas of governance, healthcare, education, agriculture, retail, e-commerce, manufacturing and financial services.   We also are touching individual users every day in their lives with AI on Windows, cloud services, gaming (Xbox) and mobile productivity usages like Outlook, Teams, Kaizala and SMS Organizer.","excerpt":"Analytics India Magazine caught up with Sundar Srinivasan, General Manager, Artificial Intelligence & Research team at Microsoft India, to understand AI’s impact and its industry applications across the sectors deploying AI-based solutions. Researchers at Microsoft are leveraging the AI tools for a range of services. Srinivasan, an industry veteran has vast experiences in building consumer […]","categories":["Deep Tech"],"tags":["cognitive computing human capital","Microsoft"],"author_name":"Disha Misal","publish_date":"2018-09-25T09:10:44","publication_year":"2018","word_count":1908,"keywords":["machine learning","artificial intelligence","AI","ML","image recognition","virtual assistants","RAG","Aim","deep learning","analytics","cognitive computing human capital","Microsoft"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","analytics","Aim","RAG","virtual assistants","image recognition"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/intelligent-analytics-digital-assistants-are-integral-to-businesses-says-sundar-srinivasan-of-microsoft-india\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":60716,"title":"Report: Growth In Global IT Spending To Fall Due To COVID-19 Pandemic, But Presents Opportunities For IT Vendors","content":"According to a recent report, the growth in global IT spending is expected to fall by the end of 2020, due to the coronavirus pandemic. The report stated that the IT spending of the global economy is likely to reduce by 3-4% by the end of this year. However, this fall may also present an opportunity for IT vendors to step-up as consulting partners for their clients. The report stated while the significant impact is expected to be on hardware business, the software and services industries are also expected to slow down. It has been noted by the experts that, however, adoption of collaborative applications, security solutions, big data and artificial intelligence are set to experience an increase in the coming days, which provides an opportunity for IT vendors to test some concepts of ‘Future of Work.’ According to IDC India Research Director of Enterprise Solutions and ICT Practices, Sharath Srinivasamurthy, “While the actual impact of this deadly COVID-19 pandemic on India market will be evident by middle of 2020, the industry is expecting a slowdown in terms of discretionary IT spending, along with the contract renewals and new deals getting signed in coming months.” He further added, “Due to this crisis, the existing project executions have also taken a hit due to travel restrictions in place.” Which, in turn, will force the IT vendors to relook at their growth targets for the rest of the year as the impact will become evident in the next few quarters. On the other hand, this downfall will also provide an opportunity for IT vendors to test their resilience on business continuity, remote connectivity, and cyber business security. This is the right time to look for innovative ways to serve their clients. Companies are now turning towards IT vendors to handhold them in the hour of crisis. The report pointed out that businesses across the country — India — are also implementing alternative ways of continuing their business, which is generating a parallel corporate line that demands to be connected from where they want, when they want and to who they want. Although working from home is not a new concept for Indian corporates, it certainly is a crucial time to see the success at this scale. Alongside, businesses are also exploring ways of collaborative work that leverages conversations, meetings, and assets across platforms with employees who are working remotely from wherever they are located to serve customers better and ensure business continuity. It is also expected that the adoption of collaborative applications will also be growing at a rapid pace after the COVID-19 outbreak. Besides, IT vendors are working with clients to get special approvals, wherever needed, to have employees connect remotely to continue providing services, especially in supporting mission-critical IT systems.","excerpt":"According to a recent report, the growth in global IT spending is expected to fall by the end of 2020, due to the coronavirus pandemic. The report stated that the IT spending of the global economy is likely to reduce by 3-4% by the end of this year. However, this fall may also present an […]","categories":["AI News"],"tags":["Coronavirus","covid-19"],"author_name":"Sejuti Das","publish_date":"2020-04-02T11:18:00","publication_year":"2020","word_count":460,"keywords":["big data","API","artificial intelligence","covid-19","AI","programming_languages:R","RAG","Coronavirus","ViT","R"],"extracted_tech_keywords":["AI","artificial intelligence","RAG","R","API","big data","ViT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/report-growth-in-global-it-spending-to-fall-due-to-covid-19-pandemic-but-presents-opportunities-for-it-vendors\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":42958,"title":"10 Best Quotes From Cypher 2018","content":"Cypher 2019 is in its fifth edition now and is slated to be held from 18-20 September at Hotel Radisson Blu, Bengaluru. It serves as one of the biggest platforms to network and learn from the leading thought leaders, companies and startups in Analytics, Data Science and Artificial Intelligence discipline. The conference has many interesting talks and sessions across the three days from industry thought leaders, who have interesting takes on how the industry has evolved. Cypher 2018 had many interesting talks from the exceptional speakers. This article lists down 10 best quotes from Cypher 2018. “The last big renaissance in our way of living was the industrial revolution. Now we are on a cusp of a more profound change brought to us by the technology that we are creating such as artificial intelligence, nanotechnology, quantum computing, bioengineering and so on. It is not only going to change the world of business and commerce but more importantly, it is going to implode our existing code of ethics. Our interaction with machines is changing us socially, emotionally and culturally. We always want to humanize the product that we invent. Is AI adapting to a sexual theme? I would like to think so.” –Atul Jalan, CEO & MD, Manthan “Models and architectures today are a lot more inclusive. If I had to measure the effectiveness of my policy five years back, I would say let us have a detailed form where I would take feedback and compare individual data points, which was time taking as a human. But models today have made it easier to solve. It can now analyse many parameters in one go such as reviews and ratings, social media images, categorical attributes, numeric attributes and more.” –Dakshinamuthry V Kolluru, Founder & President, INSOFE “Looking at the entire nexus of forces and what is around us, it seems evident that India is at the forefront of AI. We have one of the largest talent bases, a good number of AI professionals, massive evolution in the startup ecosystem, a growing hub for entrepreneurship, and more. But we are still winning a bronze or silver in the AI race. We are still away from the gold and we have to still overcome hurdles such as data quality, channelising decision making and more to become a leader in the space.” –Deep Thomas, Group Chief Data and Analytics Officer, Aditya Birla Management Corporation “Healthcare is a big industry but unfortunately, when it comes to technological innovations, very fewer innovations in terms of digital health has happened. I believe that in the next five years, smart software will make smarter diagnosis than doctors. In 10 years it will become legally mandatory for us to take the second opinion from the software before starting the treatment. I am confident because as doctors when we make a diagnosis, everything is finite, there is nothing infinite.” –Devi Shetty, Chairman & Executive Director, Narayana Health “The talk of AI is everywhere and it is natural that organisations want to adopt AI in some form and explore what it can do for them. But businesses equally face another dilemma whether AI is nothing but analytics repackaged in a different form using different jargons and terms. It is often said that AI, machine learning, deep learning are essentially the same thing that has been done in the last 15 years under analytics. There is a lot of hype around AI which needs to settle.” –Joydeep Dam, Director- Algorithms and AI, Bridgei2i “There are many myths around AI that need to be busted. The first myth is that AI can be purchased. It is not an ERP solution that you can go through 3-4 options and pick one. Every organisation needs a different and tailor-made AI solution. Second, it that organisations need separate designations such as CAO and Chief AI Officer. Bringing separate designation can only propagate AI and not implement AI. The third is that people often think that building an ML algorithm will start learning and delivering on its own. This is the biggest myth due to which a lot of solutions are not able to achieve the required benefits.” –Piyush Chowhan, Senior VP and CIO, Arvind Lifestyle Brands “There is one aspect which to my own vantage point I always believe that just implementing an aggregation of data, infrastructure, technology may not be enough. Because eventually, we are trying to mimic the human brain. Somewhere adjacent competencies such as behavioural science, design thinking are coming to the fore. Organisations are implementing design thinking into AI because of a lot of they want transformation, disruption and innovation, which cannot be accomplished otherwise.” –Sameer Dhanrajani, Chief Strategy Officer, Fractal Analytics “Artisanal data scientist exists no more. Hand made models are no longer the trend and sun is setting on the artisanal way of performing actions in data science.” –Sandeep Mittal, MD, Cartesian Consulting “What universities are doing wrong today is that they are teaching courses. They are not teaching real-world use cases. India is at an amazing place and has an immense market for AI researchers to make a lot of profit. It is important to set up next-gen universities and educational institutions that help in creating better people, better engineers and even  Eienstiens of the AI world.” –Sukant Khurana, Researcher and Entrepreneur “There are many challenges in lateral hiring. First is there are so many online programs today that are good in imparting theoretical knowledge, but practical implementation lacks majorly. Second is the lack of infrastructure and unavailability of actual data. Many a time they deal with dummy data which doesn’t address the scale that we are looking for. Third is that lateral hires are not willing to start fresh.” –Vijay Nair, Director- Analytics, Lowe’s India","excerpt":"Cypher 2019 is in its fifth edition now and is slated to be held from 18-20 September at Hotel Radisson Blu, Bengaluru. It serves as one of the biggest platforms to network and learn from the leading thought leaders, companies and startups in Analytics, Data Science and Artificial Intelligence discipline. The conference has many interesting […]","categories":["Deep Tech"],"tags":["nanotech","nanotechnology"],"author_name":"Srishti Deoras","publish_date":"2019-07-19T13:26:11","publication_year":"2019","word_count":956,"keywords":["data science","Go","machine learning","artificial intelligence","AI","ML","nanotech","Ray","deep learning","analytics","nanotechnology","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","data science","analytics","Ray","R","Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/10-best-quotes-from-cypher-2018\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10043228,"title":"Softbank Puts Its Famed Humanoid Robot Pepper On Hold","content":"Seven years ago, Japanese conglomerate SoftBank launched Pepper, the first humanoid robot that can read emotions. Last year, the company paused its production and would only make the robot ‘when it is needed’, a BBC report stated. Softbank is also downsizing its global robotics operation in France, reported Reuters. ICYMI: Time's up for Pepper robot. SoftBank stopped producing its humanoid last year, according to sources and documents reviewed by @Reuters https:\/\/t.co\/lk85ima2BF pic.twitter.com\/165Fn5zXTB— Reuters (@Reuters) July 5, 2021 Pepper origins Pepper was first launched on June 5, 2014, in collaboration with French robotics company Aldebaran (now SoftBank Robotics). In June 2015, Pepper robot was on sale for $1650. and the first batch of 1K units was sold out in just one minute. At the time, Bruno Maisonnier, the founder of Aldebaran, said the emotional robot would add a new dimension to our lives and open up ways to interact with technology. “It is just the beginning, but already a promising reality. Thanks to Pepper, the future begins today,” said Maisonnier. The decision to ‘halt’ the production of the Pepper is a body blow to chief executive Masayoshi Son’s plans to make SoftBank a leader in robotics. SoftBank shares -30% since ytd high on March 12. Comes as its Chinese investments are under-pressure (most recently Didi)and its robot Pepper, \"a shiny-suited over-seller of its limited wares\", is being put to rest as brilliantly told by @UrbanDirthttps:\/\/t.co\/W35LLdvfdw pic.twitter.com\/hTgEB6SCkc— Arash Massoudi (@ArashMassoudi) July 5, 2021 In recent times, SoftBank has shifted its focus to products such as cleaning robot Whiz. The SoftBank Robotics in France was increasingly sidelined, reported Reuters. Also, the Japanese conglomerate sold a large majority of its stake (close to 80 percent) in robotics firm Boston Dynamics to Hyundai, in a deal that values the company at $1.1 billion. Softbank acquired the robotics company from Google in 2017. Boston Dynamics has caught the popular imagination with viral videos of its humanoid and dog-like robots. SoftBank’s robotic investments so far (Source: ABI Research) Currently, SoftBank has stakes in robotics firms like AutoStore, Berkshire Grey and Bear Robotics, among others. In April 2021, SoftBank acquired 40% of AutoStore for $2.8 billion. “Are the robot-related investments more akin to Softbank’s successful investment in Alibaba, or more like the investment in We Work, with skyrocketing losses and misplaced hype leading to an abrupt crash? There are likely to be many casualties,” said Rian Whitton, Senior Research Analyst, ABI Research. Why did Pepper fail? The social humanoid robot Pepper could recognise faces and basic human emotions. It was optimised for human interaction through conversation and its touch screen. According to the company’s website, “Over 2K companies around the world have adopted Pepper robots.” Pepper was designed for businesses, schools, colleges, universities, and railway stations as an assistant to welcome, inform and guide visitors. Unfortunately, despite its popularity at conferences and events, it never really caught on as a commercial product at 27K units sold till date. “It [Pepper] is a very social robot, it can speak, it can dance, it can do certain tricks,” said Kerstin Dautenhahn, professor of intelligent robotics at the University of Waterloo, Canada, to Tech Tent, a podcast on BBC Sounds. However, she said, the use of this robot is minimal. Conclusion Robotics expert Prof Noel Sharkey told BBC that he would be happy to see an end to it. He said it did a lot of harm to ‘genuine robotics research’ by giving a false impression that it could hold conversations. “It was mostly ‘remote-controlled’ with a human conversing through its speakers. Deceiving the public in this way is dangerous and gives the wrong impression of the ‘capabilities of AI’ in the real world,” added Sharkey. SoftBank is yet to clarify the real reason behind its discontinuation of the Pepper robot production and layoffs. Pepper robot’s end doesn’t mean it was awful, but it’s time a better robot took its spot, writes Financial Times.","excerpt":"Seven years ago, Japanese conglomerate SoftBank launched Pepper, the first humanoid robot that can read emotions. Last year, the company paused its production and would only make the robot ‘when it is needed’, a BBC report stated. Softbank is also downsizing its global robotics operation in France, reported Reuters. Pepper origins Pepper was first launched […]","categories":["IT Services"],"tags":["AI (Artificial Intelligence)","AI Robots","Boston dynamics spot","Machine Learning"],"author_name":"Amit Naik","publish_date":"2021-07-09T09:00:00","publication_year":"2021","word_count":653,"keywords":["Go","Boston dynamics spot","AI Robots","programming_languages:R","AI","Machine Learning","programming_languages:Go","ai_applications:robotics","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","R","Go","programming_languages:R","programming_languages:Go","ai_applications:robotics"],"url":"https:\/\/analyticsindiamag.com\/it-services\/softbank-puts-its-famed-humanoid-robot-pepper-on-hold\/","complexity_score":3,"technical_depth":6,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":23819,"title":"Dr Ankur Narang Of Yatra Goes In Depth About The Technology Behind YUVA, Their Virtual Travel Assistant","content":"Dr Ankur Narang Speaking at CYPHER 2017 Yatra, one of India’s leading online travel agencies and travel search engine launched its virtual assistant YUVA recently. This reinforces Yatra’s focus on the world of personalisation and delivering enhanced experience to its customers. Analytics India Magazine interacted with Ritesh Srivastava, who is the General manager at Yatra and Dr Ankur Narang, who heads Yatra Labs and leads Decision Sciences team, that brought into life the first-of-its-kind “hybrid approach-based bot” that allows user to use both voice and text to communicate easily, while simultaneously keeping the UI functionality. YUVA is currently available on desktop, Android, iOS, Google Assistant and Facebook Messenger, but Dr Narang is ambitious about bringing it into other platforms like Alexa, Google Home, WhatsApp among many more in the near future. He says, “We are at a full throttle to open it up on all the platforms to maximise our reach”. YUVA: The Idea And The Journey The team started working on YUVA about six months ago, and ramped up their speed some three months ago. “With an initial focus on domestic and international flights, our main idea was to provide something different to our clients in terms of leveraging our current user interface, so that user is able to navigate through it easily and use voice and text for chat interactions,” said Dr Narang. “It gives a feeling to clients as if the virtual assistant is sitting in front of him\/her and is able to answer all the relevant queries while making a booking,” he adds. YUVA basically follows the scope of discussion and context to find an intent and accordingly review flight availability between source and destination. Srivastava says that as of now, YUVA can go from normal queries such as booking flight tickets from New Delhi to Mumbai, to longer queries like booking a non-stop flight for two between point A to point B via airline C in the morning within a certain price range. It filters the results one by one, and if a user is unable to speak due to noisy environment, he\/she can type the query. This is the version that is live now, and the team at Yatra wants to develop advanced versions where YUVA will be able to answer complex queries like details about all the past bookings, the amount of e-cash in the account, among multiple others. It has been worked upon by a small team of smart engineers who leverage open source technologies such as Python and databases like Cassandra and Aerospike, amongst others. Making Of YUVA: Artificial Intelligence, ML, NLP Dr Narang shares, “If we look at the functionalities mentioned above, the main task of YUVA is to understand the instances being communicated by the humans. For example, if there is a statement like, ‘Book a flight from Delhi to Mumbai via Indigo in the morning, departing tomorrow,’ there is a clear intent of booking a flight and that needs to be inferred. It has to accordingly filter the flights depending on departure and arrival cities, time specification, return flight, price range etc,” “These entities are inferred by our slot filling algorithm where we use our in-house NLP module that identifies this intent and the details to be filled. As the conversation goes, it also has to maintain the context of the conversation. If you, for instance change your booking destination from Mumbai to Goa in this case, it has to identify it and perform a modified search. This is where AI comes into play that we have in our proprietary backend,” he said. How it responds to every context, essentially needs to be learnt by the system and this is driven by Artificial Intelligence and ML. Future Plans For YUVA While YUVA is efficient in answering complex queries, there are further improvements that the Yatra team is trying to bring in, such as its ability to answer FAQs in a more interactive fashion rather than “just responding with some information” to the user. “We also want to expand it to other dimensions such as hotels, holidays, activities and others along with a strong touch of regional language coverage”, said Srivastava. In fact, they also plan to do this by expanding from the current English language to ‘Hinglish’, Hindi, and major regional languages including Telugu, Tamil, Gujarati, Marathi and Punjabi, among others. The Yatra team also wants to bring in more personalisation to give enhanced customer experience. “Most companies on travel have chatbot and virtual assistants but they are not on all platforms like us. We aim to be on every platform that is available in Indian market and have a broad umbrella, when YUVA is concerned”, says Dr Narang. Personalising YUVA: Recommendation Engine Is The Key Having stated the ambitious plan of bringing personalisation and understanding user behaviour, Yatra intends to have a full-blown recommendation algorithm which would use ML to drive various use-cases across flights, hotels and holidays. From churning campaigns to web personalisation, the recommendation engine would play a key role. “It would give us business preference, business guidance and motivate users to come back to the website — all of this in real time”, says Dr Narang. Right now YUVA uses basic personalisation such as greeting a user as soon as he or she logs in; but they are working on a module where they could use a powerful model of the user behaviour and pro-actively give details and recommendations based on user choice and preferences. “We plan to have this feature coupled to our framework in a month’s time so that we could guide users to a much faster booking cycle based on their interest”, he shares. Other Areas Where Yatra Is Using AL, ML And Analytics One of the key areas where Yatra is using AI and ML is dynamic discounting. This is to provide the best operational efficiencies and maximise the revenue with higher margins. “One needs to have these technologies in place as there are tens of thousands of hotels across the country. We need to be able to change their price based on multiple criteria such as length of stay, booking window, inventory availability, elasticity of city, user behaviour, amongst others.” Dr Narang added that it is a complex problem and involves both prescriptive and predictive analytics — to be able to first accurately predict the performance of a hotel across multiple dimensions and then optimise the performance of the overall system dynamically to meet the overall business objectives. “This is done through an algorithm which involves flow orchestration of various types of data and performing multi-modal inference, leading to higher business impact for the company”, he said. “The other area where we are using AI is essential is in the paid marketing sector, where we are tuning the bidding on platforms such as Adwords for multiple lines of businesses starting with flights, to improve performance across various dimensions,” he shares. “It involves a combination of business logic tuning, NLP, social network modelling, optimisation and reinforcement learning. This helps to dynamically perform ad tuning and auto-tune the bidding strategy to improve the performance of the ad,” said Dr Narang. This is fundamental for e-commerce, as there is a typical dependence around discounting needs. Challenges In The Space Dr Narang shares that finding right talent is one of the key challenges. In addition, he shares that there are challenges such as data inconsistency and noisy data leading to large amount of time spent in data cleaning, among others. “It takes a lot of time to set a flow and crunch input data to bring it in the right format”, he says. Dr Narang says that one has to be careful in the data cleaning aspect, as errors may lead the model into mis-inference, therefore reducing the accuracy and affecting the overall effectiveness of the output. He said that data cleaning necessity along with input data pattern changes require one to have a continuous check on the accuracy and performance of AI driven applications. Focus Areas For The Company In 2018 Dr Narang is clear about the part where they continue work in areas like AI and ML, which are the key portfolios for the company. They plan to continue working on discounting, marketing spending to enhance efficiency and scale along with Virtual Assistant and Real-time context aware Recommendation Engines. “We would also like bring best of technologies in AR\/VR, deep sentiment analysis, blockchain and dive more into fraud management and cyber security aspects” shares Dr Narang. On A Concluding Note Dr Narang shares that it is important to deliver a strong experience to customers and drive higher customer retention and conversion and therefore investment in New Tech plays a vital role in giving a boost to the travel industry. It is the base of the pyramid of things like virtual assistants and its connect to various region and cultures could bring a valuable change in the hospitality industry in India. “Yatra Labs is driving the company on a technology inflection point and Yatra will continue to deliver best in class products for its customers. We aim to have world-class ranking in terms of travel technology,” he says signing off.","excerpt":"Yatra, one of India’s leading online travel agencies and travel search engine launched its virtual assistant YUVA recently. This reinforces Yatra’s focus on the world of personalisation and delivering enhanced experience to its customers. Analytics India Magazine interacted with Ritesh Srivastava, who is the General manager at Yatra and Dr Ankur Narang, who heads Yatra […]","categories":["AI Features"],"tags":["Interviews and Discussions"],"author_name":"Srishti Deoras","publish_date":"2018-04-18T06:01:31","publication_year":"2018","word_count":1528,"keywords":["artificial intelligence","AI","sentiment analysis","ML","virtual assistants","RAG","NLP","Aim","analytics","predictive analytics","Interviews and Discussions"],"extracted_tech_keywords":["AI","artificial intelligence","ML","NLP","analytics","Aim","RAG","virtual assistants","predictive analytics","sentiment analysis"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/dr-ankur-narang-of-yatra-goes-in-depth-about-the-technology-behind-yuva-their-virtual-travel-assistant\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10002848,"title":"GitHub Deposits Open Source Code In Arctic To Be Used By Developers Even After 1000 Years","content":"GitHub Archive Program along with the GitHub Arctic Code Vault are unique initiatives by the company, which aims to preserve the open-source software for future generations by storing codes in an archive which can last a thousand years. In a recent development, GitHub moved all of its open-source code and repositories to the Arctic. Done in collaboration with their archive partners, Piql, GitHub wrote  21TB of repository data to 186 reels of piqlFilm which are digital photosensitive archival film which can last for 1000 years. It has stored around 6000 software repositories at Piql’s long-term storage data centre in Svalbard. As the blogpost details, the boxes with film reels were shipped to Oslo Airport before being sent to Svalbard, which is roughly 600 miles (1000 km) north of the European mainland. Due to the pandemic, it was closed till now and has been opened only recently for visitors. Julia Metcalf, who is the director of strategic programs at GitHub shared that the project which was scheduled to be completed much earlier got delayed due to the pandemic. The reels have been kept into deep chambers hundreds of meters of permafrost, where the code now resides fulfilling their mission of preserving the world’s open-source code for over 1000 years. To recognise and celebrate the contributors to the opensource software, GitHub designed the Arctic Code Vault Badge, which is shown in the highlights section of a developer’s profile on GitHub. The company further shared that every reel of the archive includes a copy of the “Guide to the GitHub Code Vault” in five languages, written with input from GitHub’s community and available at the Archive Program’s own GitHub repository. The archive will also include human-readable reel which documents the technical history and cultural context of the archive’s contents, which the company calls as the Tech Tree. It will primarily consist of the existing works, selected to provide a detailed understanding of modern computing, open-source and its applications, modern software development, popular programming languages, etc.","excerpt":"GitHub Archive Program along with the GitHub Arctic Code Vault are unique initiatives by the company, which aims to preserve the open-source software for future generations by storing codes in an archive which can last a thousand years. In a recent development, GitHub moved all of its open-source code and repositories to the Arctic. Done […]","categories":["AI News"],"tags":[],"author_name":"Srishti Deoras","publish_date":"2020-07-20T12:51:23","publication_year":"2020","word_count":332,"keywords":["Go","programming_languages:R","AI","Git","RAG","Aim","llm_models:Bard","Julia","GitHub","R"],"extracted_tech_keywords":["AI","Aim","RAG","R","Go","Julia","Git","GitHub","llm_models:Bard","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/github-deposits-open-source-code-in-arctic-to-be-used-by-developers-even-after-1000-years\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10059936,"title":"A guide to building reinforcement learning models in PyTorch","content":"PyTorch is one of the most used frameworks in the field of deep learning. We can use this library in every aspect and field data science and machine learning. We can also use it for reinforcement learning. In one of our articles, we have discussed reinforcement learning and the procedure that can be followed for building reinforcement learning models using TensorFlow in detail. In this article, we will discuss how we can build reinforcement learning models using PyTorch. The major points to be discussed in this article are listed below. Table of contents The CartPole problem Importing librariesDefining setupStoring memoryDeep Q network Algorithm Training of network Let’s start with understanding the CartPole problem. The CartPole problem In this article, we are dealing with the CartPole problem. Where we will try to make an agent learn to decide whether to move the cart on the left side or right side so that the pole attached to the cart will be straight. We can see an example in the below image. Internally, any action taken by an agent depends on the state of the environment, as the environment changes its state it returns a reward to the agent that decides the action of the agent at that state of the environment. Here in this procedure, we will use +1 rewards at every timestep and if the cart moves more than a limit from the centre (here it is 2.4 units) or the pole will fall over too far, the environment will not return any reward to the agent. This is how if the agent is performing well then only it will work for a longer duration and get larger rewards. Using the below lines of codes we can render the CartPole problem. import gym import numpy as np import matplotlib.pyplot as plt from IPython import display as ipythondisplay from pyvirtualdisplay import Display display = Display(visible=0, size=(400, 300)) display.start() env = gym.make(\"CartPole-v0\") env.reset() prev_screen = env.render(mode='rgb_array') plt.imshow(prev_screen) for i in range(50): action = env.action_space.sample() obs, reward, done, info = env.step(action) screen = env.render(mode='rgb_array') plt.imshow(screen) ipythondisplay.clear_output(wait=True) ipythondisplay.display(plt.gcf()) if done: break ipythondisplay.clear_output(wait=True) env.close() Output: Since it’s a combination of different frames we can not post this here. Running the above codes will give you every state of the agent and also through the visualization we can say that the environment has some states like velocity and position. Let’s start our procedure by importing libraries. Importing libraries In this article, we will look at how we can use PyTorch for reinforcement learning. For this purpose, we are required to have a PyTorch and gym environment installed in our environment. The installation can be performed using the following lines of codes. !pip install gym !pip install pytorch Since I am using google colab for this, I already have these libraries installed in the environment. The following libraries are required in our process. import gym import math import random import numpy as np import matplotlib import matplotlib.pyplot as plt from collections import namedtuple, deque from itertools import count from PIL import Image From PyTorch, we are required to call the following libraries. import torch import torch.nn as nn import torch.optim as optim import torch.nn.functional as F import torchvision.transforms as T Using the above packages we will define layers in our neural network and opitm module will help in model optimization and autograd module will help in automatic differentiation. Defining setup Using the below lines of code we are unwrapping the CartPole-v0, setting up the matplotlib for visualization, and starting the GPU for the environment. env = gym.make('CartPole-v0').unwrapped is_ipython = 'inline' in matplotlib.get_backend() if is_ipython: from IPython import display plt.ion() device = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\") Storing memory Basic reinforcement learning requires replay memory for the training of the network. So in some kind of storage, we are required to store observations of the agent that can be used later. Using two classes we can store memory. Transition = namedtuple('Transition', ('state', 'action', 'next_state', 'reward')) This class is named tuple, this tuple represents the single transition of our environment. class ReplayMemory(object): def __init__(self, capacity): self.memory = deque([],maxlen=capacity) def push(self, *args): self.memory.append(Transition(*args)) def sample(self, batch_size): return random.sample(self.memory, batch_size) def __len__(self): return len(self.memory) The main motive of defining this class is to buffer the transition observed recently in a cyclic way. We also define samples in this class that will select a random batch of transitions for training. Deep Q network Algorithm As in the normal reinforcement learning procedure here also we aim to train an agent on the policy that can maximize the cumulative reward. Talking about the main idea behind Q-learning is to utilize such a function that can tell us what will be the return of any action. Given any action regarding a state, we can maximize our rewards using a policy. In our network, we are going to use 5 convolutional layers from PyTorch which means our network is a convolutional neural network. We also have two outputs. In effect, our network will try to predict the expected return for each action given the current state. class DQN(nn.Module): def __init__(self, h, w, outputs): super(DQN, self).__init__() self.conv1 = nn.Conv2d(3, 16, kernel_size=5, stride=2) self.bn1 = nn.BatchNorm2d(16) self.conv2 = nn.Conv2d(16, 32, kernel_size=5, stride=2) self.bn2 = nn.BatchNorm2d(32) self.conv3 = nn.Conv2d(32, 32, kernel_size=5, stride=2) self.bn3 = nn.BatchNorm2d(32) def conv2d_size_out(size, kernel_size = 5, stride = 2): return (size - (kernel_size - 1) - 1) \/\/ stride + 1 convw = conv2d_size_out(conv2d_size_out(conv2d_size_out(w))) convh = conv2d_size_out(conv2d_size_out(conv2d_size_out(h))) linear_input_size = convw * convh * 32 self.head = nn.Linear(linear_input_size, outputs) Using the below lines of code we can call either one element to determine the next action during optimization. def forward(self, x): x = x.to(device) x = F.relu(self.bn1(self.conv1(x))) x = F.relu(self.bn2(self.conv2(x))) x = F.relu(self.bn3(self.conv3(x))) return self.head(x.view(x.size(0), -1)) Training of network Using the below codes we can instantiate our model and its optimizer. Defining parameters BATCH_SIZE = 128 GAMMA = 0.999 EPS_START = 0.9 EPS_END = 0.05 EPS_DECAY = 200 TARGET_UPDATE = 10 Getting the screen size will help us to initialize layers correctly based on the shape returned from the gym environment and get the number of actions. init_screen = get_screen() _, _, screen_height, screen_width = init_screen.shape n_actions = env.action_space.n Defining policy net and target net and evaluation of target net policy_net = DQN(screen_height, screen_width, n_actions).to(device) target_net = DQN(screen_height, screen_width, n_actions).to(device) target_net.load_state_dict(policy_net.state_dict()) target_net.eval() Defining optimizer optimizer = optim.RMSprop(policy_net.parameters()) memory = ReplayMemory(10000) We can also define some utilities for showing off the results. Select_action is a utility for selecting actions using the epsilon greedy policy.  Plot_durations is a utility for plotting the duration of episodes. The resulting plot will contain the main training loop. steps_done = 0 def select_action(state): global steps_done sample = random.random() eps_threshold = EPS_END + (EPS_START - EPS_END) * \\ math.exp(-1. * steps_done \/ EPS_DECAY) steps_done += 1 if sample > eps_threshold: with torch.no_grad(): return policy_net(state).max(1)[1].view(1, 1) else: return torch.tensor([[random.randrange(n_actions)]], device=device, dtype=torch.long) Plot duration episode_durations = [] def plot_durations(): plt.figure(2) plt.clf() durations_t = torch.tensor(episode_durations, dtype=torch.float) plt.title('Training...') plt.xlabel('Episode') plt.ylabel('Duration') plt.plot(durations_t.numpy()) if len(durations_t) >= 100: means = durations_t.unfold(0, 100, 1).mean(1).view(-1) means = torch.cat((torch.zeros(99), means)) plt.plot(means.numpy()) plt.pause(0.001) if is_ipython: display.clear_output(wait=True) display.display(plt.gcf()) Training loop After defining model and optimizer settings we can train our model. def optimize_model(): if len(memory) < BATCH_SIZE: return transitions = memory.sample(BATCH_SIZE) batch = Transition(*zip(*transitions)) non_final_mask = torch.tensor(tuple(map(lambda s: s is not None, batch.next_state)), device=device, dtype=torch.bool) non_final_next_states = torch.cat([s for s in batch.next_state if s is not None]) state_batch = torch.cat(batch.state) action_batch = torch.cat(batch.action) reward_batch = torch.cat(batch.reward) state_action_values = policy_net(state_batch).gather(1, action_batch) next_state_values = torch.zeros(BATCH_SIZE, device=device) next_state_values[non_final_mask] = target_net(non_final_next_states).max(1)[0].detach() expected_state_action_values = (next_state_values * GAMMA) + reward_batch criterion = nn.SmoothL1Loss() loss = criterion(state_action_values, expected_state_action_values.unsqueeze(1)) optimizer.zero_grad() loss.backward() for param in policy_net.parameters(): param.grad.data.clamp_(-1, 1) optimizer.step() In the above codes, we first transpose the batch array of transition by computing the mask of non-final states. With this, we compute the expected q values, Huber loss and optimize the model. Using the below code we can visualize the main training loop. By resetting the state tensor we sample the action and execute it again, sample the next action and repeat the execution in the loop. num_episodes = 50 for i_episode in range(num_episodes): env.reset() last_screen = get_screen() current_screen = get_screen() state = current_screen - last_screen for t in count(): action = select_action(state) _, reward, done, _ = env.step(action.item()) reward = torch.tensor([reward], device=device) last_screen = current_screen current_screen = get_screen() if not done: next_state = current_screen - last_screen else: next_state = None memory.push(state, action, next_state, reward) state = next_state optimize_model() if done: episode_durations.append(t + 1) plot_durations() break if i_episode % TARGET_UPDATE == 0: target_net.load_state_dict(policy_net.state_dict()) print('Complete') env.render() env.close() plt.ioff() plt.show() Output: We can see outputs in the runtime of the cell, for better results and optimization we can improve the number of episodes. The whole code for the procedure can be found here. Final words In this article, we have discussed the CartPole problem. For the environment, we used the Gym toolkit, and for solving it to an extent using an agent and reinforcement learning algorithm. We used the PyTorch framework to make them all work together. References Link for the codesPytorch documentationGym toolkit","excerpt":"In this article, we will discuss how we can build reinforcement learning models using PyTorch.","categories":["Deep Tech"],"tags":["AI (Artificial Intelligence)","Data Science","Data Scientist","Deep Learning","Google","Machine Learning","Python"],"author_name":"Yugesh Verma","publish_date":"2022-02-06T15:00:00","publication_year":"2022","word_count":1497,"keywords":["data science","machine learning","AI","neural network","TensorFlow","PyTorch","Machine Learning","Python","Ray","Aim","deep learning","Colab","Google","Deep Learning","Data Science","Data Scientist","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","machine learning","deep learning","neural network","data science","Aim","Ray","TensorFlow","PyTorch","Colab"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/a-guide-to-building-reinforcement-learning-models-in-pytorch\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":31029,"title":"Healthcare Analytics Is Becoming A Vibrant New Career Opportunity In India","content":"The healthcare sector in India needs a reform in its customer acquisition and care. With data analytics firms budding across the globe, India needs to eradicate the old healthcare management system and embrace new and efficient tools. According to reports, India has a deficiency of trained medical workforce and falls short under the World Health Organisation guidelines. India needs a proper healthcare system which can cater to everyone’s needs. Healthcare has seen many new applications of data analytics and is open to many contemporary solutions to all healthcare-related issues. Health analytics is an umbrella term which includes offering insights about patient records, cost, diagnosis, pharmaceutics and medical research data, among others. These insights are often given in real time and help organisations take  appropriate actions in healthcare, research and market. Scope Of Analytics In Healthcare In the US, healthcare analytics is estimated to see an increase of $18.5 billion by the year 2020 (Mckinsey & Company). IBM launched Watson health, a cloud-based application, which gives a platform to healthcare providers and patients to share expertise on health and research. The platform has helped over 1,84,000 patients and 80,000 professionals with intellectual solutions. According to reports, in the Indian pharmaceuticals sector, branded medicines make up for 70-80% of the retail market. The market is ever growing, medical infrastructure and health insurance have become provident in the healthcare sector. Under the Millennium Development Goals (MDG), healthcare analysis has become an integral part of providing efficient healthcare advice to both hospital administration and patients. Many healthcare firms are introducing data analytics in their organisations. For example, MyCooey, a mobile-based application helps connect healthcare providers to patients, family, and office. This helps “Clinicians” check patient records and anomalies in health status and timely diagnose it and share vital information with peers and family. Healthcare Analyst: A New Career Opportunity For Medical Students Premium institutes like the Indian Institute of Management (IIM) and Krupandhi School of Management (KSM) have introduced Post Graduation programmes in the field of healthcare Analytics. This course helps students from either medical or pharma background to learn about latest software in data analysis, business intelligence and the role of artificial intelligence in healthcare. The courses are open not only for medical graduates but also for other science graduates. The PG-DM course provided by KSM is structured for students with a background in medicine to understand and modify models of healthcare data analytics in India. The minimum criterion for this course is a bachelor’s degree in science or medicine. Students who complete this course often get placed in hospitals, consulting firms, analytics firms and with other healthcare providers. In India, several new opportunities are up and coming in the field of healthcare analytics. For example, the US-based company InfoFaces has ventured in Chennai to provide cost-effective health plans, integrated medical solutions and forecast market dynamics in healthcare to clients. Intelligence Is Artificial, The Pay Isn’t The field is a current favourite amongst data analysts. Companies like Gionik Human Capital Solutions, pay up to ₹12 lakh per annum salary to data scientists in healthcare analytics stream. One can also bag a job as a healthcare analyst in the UK-based enterprise, Price Waterhouse Coopers (PWC), which pays up to ₹6 lacs per annum. With the medical and pharmaceutical field getting over-flooded every year, it’s saddening to see that only a few aspirants are able to achieve their dreams. An alternative and attractive career has been introduced in the education sector. Data analysts and scientists not only get well paid but also get to work with the biggest technology and health firms. Outlook Data by India Brand Equity Foundation (IBEF) suggests a boom in hospital industry by 16-17%, estimated to reach the $372 billion by 2022 by introducing easy access to healthcare and services to everyone at affordable rates. Indian start-ups like Metaome aim to create new database technologies in the healthcare sector. DistilBio, one of the company’s products, helps in making personalised medicine, giving clinical support and life sciences research by integrating data from various life science sources and compiling and shaping graphics data. Its founders Kalpana Krishnaswami and Ramkumar Nandakumar left their science-research management jobs to co-found a company to help people in India gain easy and efficient access to healthcare. The transition from medicine to data analytics is a tough choice based on subjects like mathematics, financial management, data mining, software languages etc. But is an exciting and adventurous career opportunity for those who want to venture into a different aspect of healthcare like dealing with rapid urbanisation and its effects on economy and health, drug discovery, health insurance, or even medical tourism.","excerpt":"The healthcare sector in India needs a reform in its customer acquisition and care. With data analytics firms budding across the globe, India needs to eradicate the old healthcare management system and embrace new and efficient tools. According to reports, India has a deficiency of trained medical workforce and falls short under the World Health […]","categories":["AI Trends"],"tags":["Data Analytics","Healthcare Automation","IBM"],"author_name":"Jignasa Sinha","publish_date":"2018-12-04T09:27:07","publication_year":"2018","word_count":771,"keywords":["business intelligence","Go","API","artificial intelligence","AI","GAN","Healthcare Automation","Aim","ViT","IBM","analytics","Data Analytics","R"],"extracted_tech_keywords":["AI","artificial intelligence","analytics","Aim","R","Go","API","GAN","ViT","business intelligence"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/healthcare-analytics-is-becoming-a-vibrant-new-career-opportunity-in-india\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":48161,"title":"Six Non-Monetary Perks Of Being A Data Scientist","content":"With companies creating tremendous amounts of data every day, it is no surprise that data science has become one of the leading career choices. In India, the popularity of data science careers has been growing exponentially thanks to the (sometimes extremely) lucrative packages offered by companies. According to several reports, the average base pay for a Data Scientist in India (as of October 2019) is approximately ₹10,19,000. If this isn’t enough for you, simply add the allowances and perquisites to this figure, and you are perhaps looking at packages upwards of ₹13,00,000 – ₹20,00,000 per annum. Undoubtedly, a career in data science is one of the most financially sound decisions you can make. But aside from the monetary perks, what are the other benefits of being employed in this sector? Let’s find out. Respect in social circles As a Data Scientist, you are likely to command respect in your social circles. Data Scientists can directly influence business decision making across both managerial and strategic levels in organizations. Data Scientists are respected among Indian social circles for their intelligence and passion for their field. So, as a Data Scientist, you get to be the Sheldon Cooper of your group. Big Bang Theory, anyone? Improved analytical skills When it comes to skills and knowledge, psychology says, “use it or lose it.” When you take up an intellectually demanding job, you are forced to learn new skills and gain the requisite knowledge to handle it. However, if your occupation is relatively mechanical, your higher-level skills (to analyze, interpret, and evaluate) go unused, and you may soon struggle with them. Data Science as a career requires you to read, study, research, analyze, and interpret information regularly. An unintended benefit of this is that you will be always on your toes and hence will keep building your skills and knowledge. Impactful responsibilities One of the common complaints of a lot of Indian youngsters is that they do not find any meaning in their jobs. While this is a disappointing statement, it is a fact, nonetheless. A lot of jobs in the Indian economy are simply cog-in-the-wheel types that do not challenge the individual in any meaningful manner whatsoever. Commute to office, work to meet daily targets, leave for home. Repeat ad infinitum. As a Data Scientist, however, your decisions and work will help the company make the most crucial decisions about their business. This gives you a real sense of responsibility as you have the chance to make real impacts that touch not just organizations, but the lives of its employees, too. Numerous transition opportunities Data Science is one of those few careers that offer exceptional flexibility for middle-stage to late-stage transitions. After working as a Data Scientist for, say, 5-7 years, you can always jump ship to take up a managerial level opportunity in nearly any sector. Regardless of the company you are applying to, there is a good chance that your skills will be relevant to them. All companies produce data, so an individual who can make the most out of this collected information may undoubtedly be of value to them. However, if you enjoy Data Science, you can continue working in the field while switching every 1-2 years to higher-paid jobs. Code your own applications You cannot be a good Data Scientist unless you have sound knowledge of coding. However, this also means that when you can code and think like a programmer, you may create your own applications as well. For the entrepreneurial types among you who have always dreamt of founding their own startups, this could be an excellent way to proceed in that direction. You can combine your prior knowledge of Data Science and your coding proficiency to create an application that solves real-world problems. In fact, your startup could be the next big thing in the Indian economy, who knows? Smart camaraderie Finally, humans are social animals. We prefer to surround ourselves with like-minded individuals with whom we can relate on at least a few grounds. As a Data Scientist, you are likely to meet amazing people who are as driven and passionate about their work as you are. Data Science, as a career, demands serious dedication. Therefore, you are much likely to work with smart and competitive individuals who can help you push your limits further and have a great time while you are at it. This article is a part of the AIM Writers Programme. If you wish to write for us, email us at info@analyticsindiamag.com.","excerpt":"With companies creating tremendous amounts of data every day, it is no surprise that data science has become one of the leading career choices. In India, the popularity of data science careers has been growing exponentially thanks to the (sometimes extremely) lucrative packages offered by companies.  According to several reports, the average base pay for […]","categories":["AI Features"],"tags":["Data Scientist","Learn Data Science","ML"],"author_name":"Angshuman Banerjee","publish_date":"2019-10-17T17:00:02","publication_year":"2019","word_count":750,"keywords":["data science","Go","programming_languages:R","AI","ML","RAG","GAN","Aim","analytics","Learn Data Science","Data Scientist","R","startup"],"extracted_tech_keywords":["AI","data science","analytics","Aim","RAG","R","Go","GAN","startup","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/six-non-monetary-perks-of-being-a-data-scientist\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10022354,"title":"How Pandemic Turned Out To Be A Blessing In Disguise For Data Centres","content":"According to a Global Data report, data centres have become a basic necessity in the post-pandemic world, along with water, electricity, gas, and telecoms. The data centre revenues are projected to hit $948 billion by 2030, up from $466 billion in 2020, growing at a CAGR of 6.7%. The findings of the report include: Most of the growth is expected around hyper-scale data centresThe new edge data centres will cater to increasing levels of enterprise-generated data being created and processed outside remote data centres or the cloudThe next few years will see more mergers and acquisition around data centres Investments Galore In 2019, the total investments in data centres stood at $244.74 billion. In 2020, the investments almost doubled, despite the pandemic. The Asia Pacific and North America have reported maximum growth as they invest in hyper-scale and edge data centres. The companies will also look to construct smart data centres closer to customers and increase efficiency while reducing these centres’ operational costs. The expansion of data centres reflects the growth of AI and machine learning fields. Leading tech companies such as Google, Microsoft, Facebook, and Alibaba have a strong dependence on data centres, with a significant part of their revenue spent on building, expanding and equipping massive data centres. Moreover, the pandemic has called for companies to adopt new architectures and software-defined, programmable infrastructures within data centres. Reasons For The Growth The data centres have become a vital fifth utility amid COVID-19. Increased cloud adoption, expansion of access to internet-related services aided by nation-wide lockdowns, data explosion, growth of the IoT market are some of the reasons cited for the surge in data centres. Below, we discuss some of the reasons in detail. Cloud adoption: COVID has driven cloud adoption as working from home became a norm. In 2021, experts anticipate an increase in enterprise spend in cloud technology to ensure the smooth and efficient running of businesses. This continued demand for cloud services will also accelerate data centre construction projects to accommodate the hyper-scale needs. According to NASSCOM, cloud spending in India is estimated to grow at a CAGR of 30% to reach $7.1 billion in 2022. Data explosion: The COVID pandemic has radically changed our relationship with technology. Most businesses are turning online to ensure business continuity. Data generation has hit unprecedented levels due to increased internet usage and broader deployment of IoT devices. We need more data centres to store all data. According to research by CRISIL, data consumption has seen a 38% rise year-on-year in Q1FY21. Increased investments in data centres: The government is likely to earmark more money on the upcoming budgets to build data centres and establish infrastructure to promote the growth of data centres. Most data science companies are looking to invest heavily in hardware and applications. Investments are pouring in the form of construction, materials, operations, electricity expenses, and more. According to CRISIL, the industry capacity is expected to expand by more than three folds from the current ~360MW in fiscal 2020. It is also likely to see an investment of over $4-5 billion in both brownfield and greenfield projects. Increase in digitisation: The pandemic has pushed digitisation like it’s nobody’s business. The desktop PC usage has gone up by over 35% post-COVID. The increased digitisation is expected to accelerate the demand for data centres and cloud infrastructure. Data security: Most big and small players are looking for secure ways to store data. There has been a substantial investment in security and data protection policies, and investing in data centres is a safe way to ensure security. Cost-effective and reliable source: Data centres provide flexibility, scalability while being affordable and reliable. Growth In India Reports suggest the data centre market in India is expected to grow at a 12% CAGR. The data centre industry plays a prominent role in facilitating massive shifts in businesses and is expected to seamlessly continue providing digital infrastructure services. Many companies in India are looking to set up data centres. In India, the data centre industry has been concentrated mainly in the top four cities — Mumbai, Delhi, Bengaluru and Chennai, which accounts for more than 60% of India’s total data centre sites. Mumbai alone accounts for over 40% of total installed capacity. Many state governments are now investing to make data centres a centre of attraction for IT and AI workloads. For instance, Maharashtra and Telangana government offer land and electricity incentives and special single-window clearance for permissions to set up data centres. Moreover, the Data Centre Policy 2020 — the first draft released last year — ensures sustainable data centre capacity in meeting India’s tremendous data demand. The policy will ensure an uninterrupted power supply, infrastructural support, and more. Private players are also investing in data centres. Yotta Infrastructure, CtrlS, NTT Netmagic, Sify, STT GDC India are some of the leading names in the Indian data centre space. Over the years, new players such as Adani Group, Colt DCS, etc. have also emerged.","excerpt":"According to a Global Data report, data centres have become a basic necessity in the post-pandemic world, along with water, electricity, gas, and telecoms. The data centre revenues are projected to hit $948 billion by 2030, up from $466 billion in 2020, growing at a CAGR of 6.7%.  The findings of the report include: Most […]","categories":["IT Services"],"tags":[],"author_name":"Srishti Deoras","publish_date":"2021-03-18T11:00:00","publication_year":"2021","word_count":829,"keywords":["data science","Go","machine learning","AI","ML","Scala","Git","ViT","GAN","R"],"extracted_tech_keywords":["AI","machine learning","ML","data science","R","Go","Scala","Git","GAN","ViT"],"url":"https:\/\/analyticsindiamag.com\/it-services\/how-pandemic-turned-out-to-be-a-blessing-in-disguise-for-data-centres\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10095597,"title":"Skyroot Aerospace Raman-1 Engine Test Successful","content":"Hyderabad-based Skyroot Aerospace, the private India space company, has accomplished the flight qualification trial of the Raman-I engine, intended for roll attitude control in their forthcoming Vikram-I rocket. The engine underwent a test firing at the Indian Space Research Organisation’s (ISRO) Liquid Propulsion Systems Centre, marking Skyroot as the inaugural private company to employ this facility. This is the first ever test done at ISRO’s LPSC by a private company. The series of tests resulted in perfect completion with 280 engine pulses with a total burn time of 104 seconds. The company said in its LinkedIn post, “Thrilled to announce the successful flight qualification testing of our Raman-I engine used for roll attitude control of the Vikram-I rocket. As the first ever test done at the Liquid Propulsion Systems Centre (LPSC) of ISRO by a private company, this test is special. Many thanks to LPSC and IN-SPACe teams for their incredible support all through.” The Raman-I engine will be utilized to facilitate roll attitude control, enabling the rocket to effectively manage its rotation and orientation. Ensuring a stable roll attitude is crucial to maintain the rocket’s stability during its ascent. Uncontrolled rotation can lead to instability, making it difficult to accurately manage the rocket’s trajectory. The roll attitude control system plays a vital role in modifying the rocket’s aerodynamic characteristics throughout the flight. By adjusting the rocket’s orientation along the roll axis, it can optimize its aerodynamic profile, reducing drag, and improving overall efficiency. Skyroot achieved a historic milestone by becoming the pioneering private company to initiate a rocket launch from India with the Vikram-S. The indigenous Vikram-S was successfully launched from the Satish Dhawan Space Center in Sriharikota, signifying India’s significant entry into the realm of commercial space exploration.","excerpt":"The series of tests resulted in perfect completion with 280 engine pulses with a total burn time of 104 seconds.","categories":["AI News"],"tags":[],"author_name":"Mohit Pandey","publish_date":"2023-06-22T12:22:36","publication_year":"2023","word_count":290,"keywords":["programming_languages:R","AI","RAG","ViT","GAN","R"],"extracted_tech_keywords":["AI","RAG","R","GAN","ViT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/skyroot-aerospace-raman-1-engine-test-successful\/","complexity_score":2,"technical_depth":6,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10061456,"title":"Top NFT marketplaces in India","content":"NFTs have been around for a few years, but 2021 saw a mammoth rise in their popularity, and India has not been slow in accepting blockchain-powered non-fungible tokens. The global NFT trading volume soared to over USD 23 billion in 2021, and while India is not at the forefront yet, the trend is gaining traction by the day. The latest Union Budget announcements regulating a tax on crypto all but only encouraged crypto investors to buy more. With actors, cricketers, and singers, including Amitabh Bachchan, rolling out their NFT, marketplaces are gaining momentum. We have curated a one-stop list of the top NFT marketplaces in India. Wazirx Founded in 2018, WazirX is an Indian cryptocurrency exchange with over 7.3 million users (as of August 2021). In 2021, it was acquired by the crypto trading platform Binanceit and launched its NFT marketplace dedicated to regional, indie and traditional creators. The platform was also known to host the first NFT exhibition, musical NFTs, and is popularly known to be a good option for starters. Built on Binance Smart Chain, WazirX charges 5% service fees; creators get the rest of the proceedings. Bloomberg reported WazirX’s transaction fees are lower compared to Ethereum based marketplaces. WazirX hit $38 billion trading in a year record by November 2021 and even crashed after heavy user activity was seen following news of the Indian government’s cryptocurrency bill decision in Parliament. Jupiter Meta Jupiter Meta has just launched what it calls India’s first fully curated NFT marketplace, focused on creating singular metaverse experiences through film, music and games. For instance, their ‘Icons of Singara Chennai’ maps the city’s culture through digital art NFTs of monuments, food and beaches. Jupiter Meta was founded in 2021 to trade NFTs and various creative and cultural segments with its built-in metaverse. The marketplace promises fixed prices and a personalised experience. It is backed by Rubix blockchain technology of level-1 and has 0 gas fees. Beyondlife Popularly known for the ‘Amitabh NFT’, Beyondlife is the marketplace for celebrities and fans. It is one of the most trusted marketplaces with its ‘First Rare Art NFT Collection’ and has introduced bestselling NFTs, including Hindustan Times NFT and Stan Lee’s Chakra the Invincible. The platform enables artists, several brands, and creators to mint and sell their NFTs within three minutes. Bollycoin Described as a bridge between Bollywood and the blockchain, Bollycoin is an NFT marketplace devoted to collections inspired by or based on Bollywood projects. Last year, the platform made headlines with their Dabangg collection in collaboration with Salman Khan. In addition, Bollycoin offers movie clips, posters, and stills, among other collectables. Their crypto-coin can be used to purchase the NFTs, and members are set to get royalties based on their activity. Colexion Colexion is claimed to be one of Asia’s largest licensed NFT marketplaces for digital artists to buy and sell their art. They have also enabled digital tokens in the form of trump cards and are in the works of creating the world’s largest NFT digital museum where users can meet with celebrities and know about their journeys. It also integrates sports personalities as game avatars on their gaming platform and allows players to own, trade and sell their avatars and commodities. BuyUCoin Since 2016, BuyUcoin has been a leading Indian crypto exchange that serves over 350,000 users and handles $300M+ in digital asset transactions with its parent company. The company’s mission is “To bring crypto in a billion Indian pockets” and supports users to buy, sell and trade 33 cryptos, including Bitcoin, Ethereum, Ripple, Litecoin, Stellar, among many others. Additionally, it allows users to leverage UPI for the financial aspects. The platform charges a 0.24% maker fee. Nifty Gateway Nifty Gateway and their NFTs’ nifties’ was known in blockchain history for their selling of Beeple’s NFT and artists like Grimes and LOGIK. The NFTs on Nifty Gateway are stored in a secured wallet and powered by Gemini’s state-of-the-art custody technology. This allows the platform to levy 0% gas fees for the artists. Every three weeks, Nifties are released as a limited edition. Rarible Launched in 2020, Rarible was ranked as the top NFT marketplace, especially given RARI, their native ERC-20 governance token. They have two tokens; ERC-721, for the creation of unique one-off items, and ERC-1155, for creating multiple editions of an item. Rarible’s tokens are also used on platforms such as OpenSea is one of the few decentralised autonomous organisations of today because it enables its customers to make all decisions themselves. NFT tokens are also used on OpenSea, another renowned platform accessible via this marketplace. Foundation A P2P NFT marketplace, Foundation’s NFTs collaborate with digital art creators, crypto natives, and collectors. Foundation aims to encourage culture and bring forward marginalised voices. With their focus mainly on digital art, artists need to create a profile and apply for an Upvote Program on the platform, where the community will further vote and accept applications. Foundation’s completely women-driven team decides the artwork displayed and ensures the artist receives 10% of the sale every time a collector resells their digital artwork to another individual.","excerpt":"Jupiter Meta has just launched what it calls India’s first fully curated NFT marketplace, focused on creating singular metaverse experiences through film, music and games.","categories":["AI Trends"],"tags":["Bitcoin"],"author_name":"Avi Gopani","publish_date":"2022-02-24T13:00:00","publication_year":"2022","word_count":849,"keywords":["Go","Bitcoin","AI","Git","RAG","Aim","ViT","CLIP","Rust","GAN","R"],"extracted_tech_keywords":["AI","Aim","RAG","R","Go","Rust","Git","CLIP","GAN","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-nft-marketplaces-in-india\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10101238,"title":"AI Will Augment Human Capabilities: WNS Global Services CEO","content":"WNS Global Services, a leading Business Process Management (BPM) company that offers a range of outsourcing services, is integrating generative AI across its talent acquisition, employee services, sales and marketing operations. Generative AI’s summarisation capabilities are being leveraged to help employees quickly understand policy and other technical documents. The sales and marketing department at WNS is also leveraging generative AI for data creation and campaign recommendations. Additionally, the integration of generative AI into WNS’ HR Chatbot called Amelia is helping the bot generate contextual and cognitive responses for employees’ queries. “This technology reshapes productivity by curating original content and automating tasks like data entry, freeing IT professionals to perform more strategic work and take on advisory roles. By adopting an ethical and responsible approach to generative AI, the IT sector can elevate capabilities to unlock real value for sustainable growth,” Keshav R. Murugesh, Group CEO, WNS Global Services told AIM. In this exclusive interaction, Murugesh discusses how WNS is leveraging the power of Large Language Models (LLMs) not just internally, but to deliver better products and services to its clients. Leveraging Generative AI “At WNS, we are advancing our offerings by strategically blending generative AI and proprietary AI\/ML models tailored for industry-specific challenges. “In insurance, our NLP model, backed by domain knowledge, identifies subrogation opportunities. Meanwhile, generative AI recommends the most optimal next steps. When it comes to travel, our knowledge engine, paired with generative AI, enhances customer experiences by delivering instant, personalised responses,” Murugesh said. In the healthcare segment as well, WNS’ ML models combined with generative AI enable accurate medical summarisation, diagnoses, and code identification. “Our approach centres on using LLMs to contextualise industry-specific AI\/ML models, which are integrated into our clients’ operational environments. Leveraging our extensive domain expertise, we design efficient prompts to extract high-quality outputs cost-effectively. This approach differentiates us, allowing us to offer customised solutions that drive transformation across the value chain.” When asked whether WNS is contemplating developing its own LLM, Murugesh said that his firm’s approach involves harnessing the capabilities crafted by leading hyperscalers and specialised industry or function-specific LLMs. “In most cases, LLM foundation models coupled with WNS Triange’s proprietary ML models help us deliver tailored solutions that cater to different functional domains and industries. Wherever required, WNS leverages its AI\/ML and domain capability to fine-tune existing foundation models to get specific results.” Generative AI challenges While enterprises today are embracing the power of LLMs, they come with their own set of challenges-hallucinations for instance. Despite numerous attempts, the creators of these models have not yet managed to resolve the problem at a technical level. While WNS is leveraging these models both internally and externally, to mitigate the risks, they have created robust frameworks and solutions to ensure transparency, explainability, and bias reduction. “By incorporating ethical considerations and maintaining the highest security standards, we are cultivating an environment that maximises the advantages of these models while minimising potential risks,” Murugesh said. AI will augment human capabilities While generative AI has gained a lot of traction in the last few years, it has also led to concerns about replacing human jobs. A report by consulting firm McKinsey stated that millions of human jobs could be impacted by AI by 2030. We have already started seeing examples of such displacement, when Dukaan, a DIY platform that enables merchants with zero programming skills to set up their e-commerce business, recently laid off 90% of support staff and replaced them with an AI chatbot named Lina. While such concerns exist among employees of all organisations, WNS is proactively addressing these concerns of its employees. “We recognise the importance of workforce development and strongly emphasise training and upskilling. We aim to foster a continuous learning and innovation culture, equipping our employees with skills to collaborate with AI technologies.” What AI will truly do is augment human capabilities, according to Murugesh. “The two can coexist – a powerful synergy of AI intelligence and human creativity to drive greater efficiency and productivity. Depending on the nature of the business process, both AI and humans may assume the role of a maker or checker. AI’s accuracy and speed make it a capable maker, while critical thinking and decision-making abilities make humans ideal for the role of a checker. This flexibility ensures a well-balanced approach to tasks.” AI needs to be regulated Yet, fully harnessing AI’s potential requires a responsible AI framework to address concerns of ethics, reliability, transparency, and compliance. Moreover, recognising this potential for AI to drive transformative change, Murugesh believes that regulation is necessary for its responsible development and deployment. While AI may not pose existential threats, the potential for misuse or unbridled expansion points to a possibility of unintended repercussions, he said. “Anchoring responsible AI practices within comprehensive regulatory frameworks becomes pivotal in averting and minimising such inherent risks and securing the positive influence of AI within society. We consciously advocate for implementing AI regulations that emphasise ethical application, transparency, and accountability.”","excerpt":"Depending on the nature of the business process, both AI and humans may assume the role of a maker or checker.","categories":["AI Features"],"tags":["Interviews and Discussions"],"author_name":"Pritam Bordoloi","publish_date":"2023-10-07T13:00:00","publication_year":"2023","word_count":825,"keywords":["TPU","AI","ML","RAG","NLP","Aim","generative AI","GAN","foundation models","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","ML","NLP","generative AI","foundation models","Aim","RAG","TPU","R","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/ai-will-augment-human-capabilities-wns-global-services-ceo\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10140932,"title":"AI Just Stopped 5G from Dying","content":"Is AI stepping in to save 5G from decline? NVIDIA, alongside SoftBank Group’s telecoms division SoftBank Corp. recently launched world’s first AI and 5G telecommunications network ‘AITRAS.’ Artificial Intelligence Radio Access Network (AI-RAN) refers to the integration of AI and machine learning (ML) into the radio access network (RAN) of cellular systems like 5G and 6G. This combination enables smarter, more efficient network performance and management. Jinsung Choi, the chairman of AI-RAN Alliance, took to LinkedIn to announce that this breakthrough solution “empowers telecom networks, transforming their operations with flexible, cost-efficient, and GPU-based AI orchestration”. NVIDIA founder and CEO Jensen Huang and SoftBank chairman and CEO Masayoshi Son previously discussed the telecom network, during which they talked about Japan’s emerging role in AI. “Now this intelligence network will become one big neural brain for the infrastructure intelligence,” Son said, adding that it “will be amazing”. Building a Strong Infrastructure Any advanced network that allows AI and 5G to operate simultaneously unlocks newer and higher revenue opportunities for telecom providers. Integrating AI and 5G ultimately involves building a strong communication infrastructure that can support enterprise-level wireless services. An Ericsson white paper published this year explores how 5G solutions, enhanced by AI-driven network management, can meet the changing connectivity needs of businesses, offering secure, adaptable, and scalable services. Source: ericsson.com (White Papers) According to another research paper, bringing 5G to life requires a robust infrastructure that includes high-speed fibre networks to manage massive data flows, edge computing to bring processing closer to end-users, and network slicing to optimise bandwidth and performance for different applications. Together, these components enable the low-latency, high-bandwidth, and reliable wireless experience that 5G promises. This integrated approach forms the backbone of a seamless, next-generation connectivity experience. Space vs. Speed Latency and bandwidth are essential for modern connectivity, especially as AI-driven applications demand faster data processing and transmission. The increasing demand for faster data processing and transmission by AI-driven applications has highlighted the importance of latency and bandwidth for modern connectivity. 5G offers high-speed, low-latency capabilities, but its usage is limited to urban areas. Elon Musk offers an alternative through Starlink, which aims to use low-earth orbit satellites to provide high-speed internet access across vast regions, including remote and underserved areas. Satellite internet services like Starlink could work alongside or even compete with 5G networks like Jio. As AI applications grow and need stronger, more widespread connections, using both satellite internet and 5G together may be essential to meet various connectivity demands. In India, where rural connectivity remains a challenge, Starlink’s technology could help bridge the digital divide, offering reliable internet in areas where 5G infrastructure isn’t practical. This could reshape India’s connectivity landscape, traditionally led by terrestrial networks like 5G, with the potential for both technologies to coexist. In areas where 5G deployment is challenging, satellite internet services offer a promising alternative. Together, these technologies could create a more comprehensive, resilient network for tomorrow’s digital needs, ensuring connectivity reaches even the most remote regions. Quantum Leap or 5G’s Last Dance? On its official blog, Nokia said that cloud intelligence and quantum networking are helping quantum computing grow, making resource use smarter and creating networks that mix quantum and regular computing. Quantum networking, with advances in quantum communication and the potential for a quantum internet, could eventually make 5G less essential. Cloud intelligence and quantum networking are driving exciting new developments, from breakthroughs in quantum computing to smarter ways of allocating resources and creating networks that blend quantum and traditional computing. The collaboration between AI and quantum computing is expected to drive progress across many fields, helping industries become more efficient, adaptable, and responsive to changing demands. These technologies promise to significantly enhance future 5G networks, enabling smarter resource management, better user location tracking, and advanced edge intelligence for both satellite and ground networks. Google Cloud’s Network Intelligence Center provides a unified console for managing network visibility, monitoring, and troubleshooting, simplifying and streamlining network management. This platform allows users to view network topology, check connectivity between endpoints, monitor performance, and receive insights on firewall rules, supporting more proactive network management. Additionally, the Network Intelligence Center automatically monitors Virtual Private Cloud (VPC) configurations, detecting misconfigurations and suboptimal setups to ensure networks perform at their best. Will 5G Coexist or Disappear? As we enter a new era in wireless communication, a question emerges: Will 5G continue to coexist with newer technologies, or will it become obsolete? 5G has significantly advanced connectivity by delivering high-speed data and low latency. However, as AI-driven applications grow, so does the demand for faster and more reliable networks. 5G may eventually reach its capacity, opening the door for 6G, suggested Stanley Russel, director at CD Tech Innovations, in his article titled ‘When Is 6G Coming, and What Does It Mean for 5G and 4G LTE?’. 6G is anticipated to revolutionise wireless communication with ultra-fast speeds and seamless AI integration. This leap could support AI applications across fields like autonomous vehicles and smart cities. However, developing and deploying 6G will require considerable infrastructure investment and major technological breakthroughs to meet the demands of the AI-driven future. The rapid expansion of AI applications requires robust and expansive networks. Traditional 5G infrastructure alone may not meet these needs, spurring interest in alternative solutions. The concept of ‘outernet’ involves a global satellite network offering free internet access. This innovation could reshape traditional internet delivery and global connectivity. To conclude, while 5G has marked a significant advancement, the needs of tomorrow’s AI-powered world may require it to coexist with, or eventually yield to, future solutions like 6G, satellite internet, and the outernet.","excerpt":"RAN, short for Radio Access Network, now integrates AI and ML, saving 5G with smarter, efficient performance.","categories":["Deep Tech"],"tags":["5G","Starlink"],"author_name":"Sanjana Gupta","publish_date":"2024-11-14T07:00:00","publication_year":"2024","word_count":931,"keywords":["Go","5G","artificial intelligence","machine learning","AI","ML","Scala","Git","Starlink","Aim","edge computing","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","Aim","edge computing","R","Go","Scala","Git"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/ai-just-stopped-5g-from-dying\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10045188,"title":"We Are The First Fintech Investment Of JP Morgan In APAC: Mohan Krishnan, Founder, Global PayEX","content":"Global PayEX offers an AI-driven working capital optimisation product suite via Accounts Receivable and Accounts Payable automation. Founded in 2011, the firm’s solutions are aimed at increasing productivity, lower cost, accelerate revenue, enhance transparency and predictability across AR and AP cycles to benefit corporates, customers, suppliers, banks, and FIs. Analytics India Magazine got in touch with Mohan Krishnan, founder, Global PayEX, to get insights into how AI is driving the fintech solutions for customers at Global PayEX. “Our customers include several F500 and large corporates such as 3M, Bridgestone, Huhtamaki, GSK, Stanley Black and Decker, Continental, and Ebro. We are the first fintech investment of JP Morgan in APAC, and we have strong partnerships in place with several leading global and regional banks,” said Mohan. AIM: What are the key services you offer, and how do they stand out? Mohan: We provide key automation solutions on both the Account Receivables (AR) and Account Payables (AP) sides. On the Account Receivables side, FreePay: The Electronic Invoicing Presentment and Payment (EIPP) with straight-through reconciliation. FreePay delivers actionable invoices in real-time on mobile or desktop and enables payment in time which helps in accelerated collections, reduction in DSO, reduction in disputes\/billing adjustments and reconciliation errors.AlgoriQ: An AI-powered intelligent fund application (IFA) solution for end-to-end automation of AR reconciliation. AlgoriQ auto-reads payment advice and bank receipt reports (MT940 etc.) in various formats and does multi-point data match of invoices for straight-through reconciliation in the ERP. The product also brings in sophisticated Deduction management with aggregation, limit checks, and dispute resolution to minimise unauthorised deductions.  FinEX: Financing platform for AR and AP extends working capital finance for both the channels and the vendors from a variety of lending partners. AIM: How does AI\/ML power your ALGORIQ engine? Mohan: AlgoriQ automates multi-way reconciliation using AI\/ML for our customer’s modern trade, or e-commerce, or institutional clients. The ML engine extracts the content and context of data from multiple formats of scanned data and direct downloads. Rules-based validations and enhancements are performed on data as the machine builds related lexicons at the customer and domain process levels. The machine-learning algorithm then performs intelligent record matches based on the context extracted and rules defined. Full matches are reconciled, and partial matches follow learning workflows. In addition, AlqoriQ now supports extensive deduction management. AIM: What’s your take on the increasing use of data analytics in improving working capital management? Mohan: One of the key indicators of the financial health of an organisation is its working capital management. Every CFO we meet has a clear agenda to improve its Account Receivable and Accounts Payable processes to release cash strapped in their operations. Digitalisation and automation of processes are great steps in reducing working capital. However, any measure implemented needs to be continuously monitored to make it effective. Data analytics thus becomes a key to have those proper insights into the challenges in the current as-is processes and the constant improvements happening due to the automation measures. For example, tracking Day Sales Outstanding (DSO) is an effective way of understanding the reasons for a higher DSO like collections, credit terms, payment mechanisms, or disputes, and how effective the process change or automation is in helping to reduce the DSO. In addition, PayEX provides working capital analytics for Corporates, with DSO tracking, cash forecasting, buyer payment behaviour, etc., which helps in a better return on capital employed. Our payment data and payment behaviour analytics is also aiding credit decisions by our panel of embedded Banks and NBFC lenders for both AR and AP financing. AIM: How did the pandemic impact the fintech industry? Mohan: The fintech industry has seen exponential growth amidst the pandemic. In the past year, both MSMEs and large corporates were forced to remotely manage their supply chains and cash flows. This led to a massive shift away from manual processes towards automation and cloud-based accounting for seamlessly managing Accounts Receivable and Payable functions, including collections, payments and reconciliations. During this time, PayEX saw transaction volume surge drastically from INR 50 billion before COVID-19 to 100 billion with the number of dealer distributors on our platform swelling 4 times. Further, the pandemic has served as the much-needed catalyst for the long-awaited digitisation of vendor payments and retreat of paper checks. Here, we have spotted a significant uptick in the adoption of electronic payment methods, such as NACH (pull payments) and NEFT\/RTGS payments and a gradual move away from paper-based transactions like cash and cheques. Today, nearly 95% of the payments on our platform are processed through NACH and NEFT\/RTGS, while a small percentage is through UPI, Cards and Open Banking initiatives. AIM: Tell us about the challenges your clients face and how your team solves them (use cases)? Mohan: A key challenge faced by most enterprises is the reconciliation of accounts payables and receivables. At present, most reconciliations are managed through traditional and manual solutions. However, as the scale of transaction volumes surges, businesses across industries, which have a wide network of local dealers to distribute their products across the country, are realising that manual solutions not only possess higher error rates but also add to the organisation’s operational risk. So, they are looking for an integrated and automated approach to reconciliation that can significantly reduce reconciliation time and operating costs and minimise regulatory penalties. Our FreePay solution digitises invoice presentment for the dealers while automating their collections and reconciliation processes at the backend to allow immediate resolution of deductions, discrepancies, and disputes and the application of instant credit notes. FreePay also offers an instant view of data, including invoices, statements, payment status, credit notes, etc., via mobile apps (android, iOS) or the web. This has helped both corporates and their dealers\/distributors, who are often MSMEs, reduce their DSO by half and unlock millions of dollars in working capital through quicker cash conversion. We also offer comprehensive analytical dashboards with audit trails. This allows real-time computation and trend analysis on all key AR metrics such as DSOs, unapplied cash receipts, cash forecasting etc. AIM: What’s next for Global PayEX? Mohan: We expect to continue our 40-50% QoQ revenue and transaction growth trajectory. As of now, we have more than 50 large corporates and 15,000 MSMEs (dealer\/distributors) on our platform, and we expect to grow this significantly besides enabling dealer\/channel financing for this segment. In parallel, we will also continue to retain and grow our customer base for our FreePay and AlgoriQ solutions and maintain our position as market leaders in AR automation and reconciliation. From an AP automation and Financing standpoint, we have closed the initial few deals, and we intend to build significant momentum in this space as well. Internationally, we have contracted deals and are targeting a multi-million-dollar revenue in the FY 2021-22. AIM: What improvements would you like to see in terms of the policy & regulations to boost the fintech sector? Mohan: The GOI has constantly been endeavouring to optimise the regulations and policies for the start-up sector as a whole and especially the fintech sector along with the central bank. However, challenges like a vast population still unbanked\/semi-banked, the slow ramp-up of the internet infrastructure, limited smartphone usage in rural areas becomes a challenge for the adoption rates. Most of the fintech offerings and adoption are skewed towards large or mid-tier cities. More incentives to Make in India smartphones, giving high-speed internet access to the rural areas, making cash transactions expensive, providing support to the SME segment for easier credit underwriting by lenders will help the fintech sector in a big way.","excerpt":"PayEX provides working capital analytics for Corporates, with DSO tracking, cash forecasting, buyer payment behaviour, etc.","categories":["AI Features"],"tags":["Interviews and Discussions"],"author_name":"kumar Gandharv","publish_date":"2021-08-04T15:00:00","publication_year":"2021","word_count":1257,"keywords":["Go","API","AI","ML","Git","Aim","ViT","analytics","GAN","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","ML","analytics","Aim","R","Go","Git","API","GAN","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/we-are-the-first-fintech-investment-of-jp-morgan-in-apac-mohan-krishnan-founder-global-payex\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10075567,"title":"Research Shows that Superintelligent AI is Impossible to be Controlled","content":"A group of researchers have come to the terrifying conclusion that containing super-intelligence AI may not be possible. They claim that controlling the AI would fall beyond human comprehension. According to the Journal of Artificial Intelligence Research, in the paper titled, ‘Superintelligence Cannot be Contained: Lessons from Computability Theory’, researchers have argued that total containment (in principle) would be impossible due to fundamental limits inherent to computing. It further claims that it is mathematically impossible for humans to calculate an AI’s plans, thereby making it uncontainable. The authors cite that implementing a rule for artificial intelligence to “cause no harm to humans” would not be an option if humans cannot predict the scenarios that an AI may come up with. They believe that while a computer system is working on an independent level, humans can no longer set limits. The team’s reasoning was inspired in part by Alan Turing‘s formulation of the halting problem in 1936. The problem centres on knowing whether a computer programme will reach a conclusion or an answer; either making it halt or simply loop forever trying to find one. An excerpt of the paper reads, “This is because a superintelligence is multi-faceted, and therefore potentially capable of mobilizing a diversity of resources in order to achieve objectives that are potentially incomprehensible to humans, let alone controllable.” Computer scientist, Iyad Rahwan, Max-Planck Institute for Human Development, Germany said “In effect, this makes the containment algorithm unusable”. Meaning, machines perform certain important tasks independently, without the programmers fully understanding how they learned it. However, alternatives have been suggested by the researchers on teaching AI some ethics. Limiting the potential of superintelligence could prevent AIs from annihilating the world, even if they remain unpredictable.","excerpt":"The authors cite that implementing a rule for artificial intelligence to “cause no harm to humans” would not be an option if humans cannot predict the scenarios that an AI may come up with.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","Intelligent Agent","Superintelligence"],"author_name":"Bhuvana Kamath","publish_date":"2022-09-19T14:17:10","publication_year":"2022","word_count":286,"keywords":["Go","artificial intelligence","programming_languages:R","AI","programming_languages:Go","Aim","Intelligent Agent","Superintelligence","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","Aim","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/research-shows-that-superintelligent-ai-is-impossible-to-be-controlled\/","complexity_score":3,"technical_depth":7,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10018707,"title":"Hitachi Vantara Introduces Kubernetes Service To Power Cloud-Native Applications","content":"Hitachi Vantara has announced the availability of Hitachi Kubernetes Service, an enterprise-grade solution for the complex challenge of managing multiple Kubernetes environments. According to the official release, Hitachi Kubernetes Service will enable customers to simply, consistently, and securely deploy, manage, monitor, and govern Kubernetes clusters across major cloud providers and on-premises. This, therefore, will empower developers to deploy workloads on their platform of choice and eliminate vendor lock-in. There has been a massive surge in adopting containers more comprehensively than previously anticipated. This is because containerisation of applications makes them easy to be deployed and run easily across different computing environments, providing substantial infrastructure cost savings. As a matter of fact, reports stated that by 2022, more than 75% of global businesses would be running containerised applications in production. And, the new Hitachi Kubernetes Service will bring Hitachi’s technology leadership, expertise, global training, and support this sector. The newly launched service will help organisations accelerate the adoption of cloud-native applications by streamlining the management of the underlying infrastructure and Kubernetes cluster deployments. It also simplifies how developers use and consume IT-delivered container services, whether on-premises or hybrid environments. The extensible self-service catalogue will also allow customers to customise it to their business requirements and rapidly deploy curated applications across a hybrid Kubernetes landscape. Additionally, it is a managed SaaS environment supported by Hitachi Vantara’s Global Customer Service organisation. Alongside, Hitachi Kubernetes Service provides enterprise-grade security with secure communications between the Kubernetes clusters and the SaaS management plane in the cloud. When asked about the launch, Bobby Soni, the President, Digital Infrastructure, Hitachi Vantara stated that “While our customers love container technology, they are challenged by the complexity to deploy and securely manage containers at scale across multiple cloud environments.” “Today, we are helping simplify and solve the multi-cloud Kubernetes challenge for our customers with the introduction of Hitachi Kubernetes Service,” said Soni. With this enterprise-grade service, the customers will now have the freedom of a truly agnostic platform, the flexibility of an extensible self-service catalogue, accompanied with world-class training and Hitachi global support to help their development teams drive business results. Some of the key benefits include — Ease of deployment through a single management pane to manage all of a customer’s private, hybrid and multi-cloud implementations. A unified dashboard to efficiently manage multiple on-premises or the cloud-based Kubernetes clusters. Centralised monitoring and alerting that can solve Kubernetes root-cause analysis challenges of lowering the meantime. Pre-populated applications in the robust service catalogue. One of the key concerns in DevOps is the lack of trained and experienced personnel. Hitachi Vantara offers extensive customer training, including ten foundational courses to help customers build necessary skills. The Hitachi Kubernetes Service will be available on AWS, with Google Cloud Platform and Microsoft Azure, early in 2021.","excerpt":"Hitachi Vantara has announced the availability of Hitachi Kubernetes Service, an enterprise-grade solution for the complex challenge of managing multiple Kubernetes environments.  According to the official release, Hitachi Kubernetes Service will enable customers to simply, consistently, and securely deploy, manage, monitor, and govern Kubernetes clusters across major cloud providers and on-premises. This, therefore, will empower […]","categories":["AI News"],"tags":["hitachi","hitachi vantara","Kubernetes","kubernetes platform"],"author_name":"Sejuti Das","publish_date":"2021-01-22T14:35:12","publication_year":"2021","word_count":462,"keywords":["Go","API","AWS","AI","Azure","kubernetes platform","ML","R","Git","hitachi vantara","hitachi","DevOps","Kubernetes","kubernetes"],"extracted_tech_keywords":["AI","ML","AWS","Azure","kubernetes","R","Go","Git","DevOps","API"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/hitachi-vantara-introduces-kubernetes-service-to-power-cloud-native-applications\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10171284,"title":"West Bengal Introduces AI System to Curb Fake UG Colleges Applications","content":"The West Bengal government’s Higher Education Department will use an application-driven AI in its centralised admission portal to detect and eliminate fake applications for undergraduate college students. The initiative aims to enhance the integrity and efficiency of the admission process by ensuring that only genuine applications are considered. The system, which was pilot-tested during last year’s admission, will be deployed this year with minor modifications. In 2020, Asutosh College in Kolkata made headlines when the name of Bollywood actress Sunny Leone featured in its English merit list. This took place due to the submission of a fake application during the college’s online admission process. Several other prank applications used the names of celebrities, including footballers like Messi and Ronaldo, in the application, which added to the embarrassment faced by the college. The Higher Education Department hopes that the AI will curb such incidents in the future. According to officials, the use of advanced technology in this manner reflects a significant step forward in combating fraudulent activities during admission. AI uses the latest technology, and the portal will be able to identify fake candidates by examining ID proof, mobile numbers, names, and images. Last year, there were 50 lakh applications from around 4 lakh candidates in the first phase. Using an AI tool, several cases were found where the names and images of applications didn’t match, and those applications were blocked. As per reports, multiple cybercafes were identified as places where fake applications were created. “I am optimistic there will be even fewer fake profiles this year…We even have a mechanism that can recognise images,” an official reportedly said. College principals have expressed support for the centralised portal, noting that it streamlines the application process, reduces administrative burdens, and effectively filters out fraudulent submissions. Officials are optimistic that the enhanced AI technology will further improve the integrity and efficiency of the college admission process.","excerpt":"The AI uses the latest technology, and the portal will be able to identify fake candidates by examining ID proof, mobile numbers, names, and images.","categories":["AI News"],"tags":["AI &amp; ethics","College"],"author_name":"Amisha Arya","publish_date":"2025-06-04T10:36:05","publication_year":"2025","word_count":313,"keywords":["Go","programming_languages:R","AI","AI &amp; ethics","ML","programming_languages:Go","Aim","College","ViT","R"],"extracted_tech_keywords":["AI","ML","Aim","R","Go","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/west-bengal-introduces-ai-system-to-curb-fake-ug-colleges-applications\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":59203,"title":"Turing Award For Pixar, Open Sourcing Of EfficientNet Lite And More: Latest in AI","content":"Regardless of what is happening around the world, the AI community are one productive bunch, and they have something interesting to share almost every day. So, here’s a compilation of all the important releases for the ML developers from top companies like Google and Uber. Here’s what is new this week: Google Open-Sources Neural Tangent Library The short history of deep learning indicates the incredible effectiveness of infinitely wide networks. Insights from these infinitely wide networks can be used as a lens to study deep learning. However, implementing infinite-width models in an efficient and scalable way requires significant engineering proficiency. To address these challenges and accelerate theoretical progress in deep learning, Google’s AI team released Neural Tangents, a new open-source software library written in JAX. This library is aimed at helping researchers build and train infinitely wide neural networks as easily as finite neural networks. Neural Tangents, at its core, provides an easy-to-use neural network library that builds finite- and infinite-width versions of neural networks simultaneously. 3D Object Detection With MediaPipe Most of the object detection usually deals with 2D objects. The bounding boxes are always rectangles and squares but never a cube. By extending prediction to 3D, one can capture an object’s size, position and orientation in the world, leading to a variety of applications in robotics, self-driving vehicles, image retrieval, and augmented reality. In order to fuel more interest in the 3D aspect of object detection, Google AI released MediaPipe Objectron, a mobile real-time 3D object detection pipeline for everyday objects. This pipeline detects objects in 2D images, and estimates their poses and sizes through a machine learning (ML) model, trained on a newly created 3D dataset. Implemented in MediaPipe, an open-source, cross-platform framework for building pipelines to process perceptual data of different modalities, Objectron computes oriented 3D bounding boxes of objects in real time on mobile devices. EfficientNet-Lite For Mobiles By TensorFlow Last year in May, Google released a new family of image classification models called EfficientNet. These models, as the name suggests, were able to achieve state-of-the-art accuracy with fewer computations and parameters. EfficientNet was designed to open up novel applications on mobile and IoT, where computational resources are constrained. Today, to match the needs of edge devices, EfficientNet-Lite gets released. It runs on TensorFlow Lite and is designed to perform well on mobile CPU, GPU, and EdgeTPU. EfficientNet-Lite brings the power of EfficientNet to edge devices and comes in five variants, allowing users to choose from the low latency\/model size option to the high accuracy option (EfficientNet-Lite4). EfficientNet-Lite4, achieved 80.4% ImageNet top-1 accuracy, while still running in real-time (e.g. 30ms\/image) on a Pixel 4 CPU. Introducing Piranha: An Open-Source Tool to Automatically Delete Stale Code Feature flags are a key part of Uber’s coding customisations. They are used for customising mobile app execution and serving different features to different sets of users. However, after a feature has been 100 per cent rolled out to users, the feature flag in the code becomes obsolete. These nonfunctional feature flags now become a technical debt in the ML pipeline, making things difficult for developers who work on the codebase. Removing this debt can be time-intensive and to automate this process, Uber developed Piranha. Piranha is a tool that scans source code to delete code related to stale, or obsolete, feature flags, leading to a cleaner, safer, more performant, and more maintainable codebase. This tool has now been open-sourced, and developers can make use of feature flags with more ease. Pixar’s Pioneers Get 2019 Turing Award Edwin Catmull and Patrick Hanrahan of Pixar have been awarded the prestigious Turing Award for the year 2019. Pixar has been revolutionising how 3D objects can be generated using computers. Catmull and Hanrahan have been key in creating near realistic effects like reflections and curved surfaces game-changing 3D computer graphics techniques, which are now widely used in the film industry. They have now won the highest distinction in computer science: the Turing Award. AlphaGo The Movie On March 9, 2016, the worlds of Go and artificial intelligence collided in South Korea for an extraordinary best-of-five-game competition. With more board configurations than there are atoms in the universe, the ancient Chinese game of Go has long been considered a grand challenge for artificial intelligence. Celebrating 4 years of the success of AlphaGo, DeepMind’s breakthrough was documented in an hour-long video directed by Greg Kohs with an original score by Academy Award nominee, Hauschka, AlphaGo chronicles a journey from the halls of Oxford, through the backstreets of Bordeaux, past the coding terminals of DeepMind in London, and ultimately, to the seven-day tournament in Seoul. As the drama unfolds, more questions emerge: What can artificial intelligence reveal about a 3000-year-old game? What can it teach us about humanity?","excerpt":"Regardless of what is happening around the world, the AI community are one productive bunch, and they have something interesting to share almost every day. So, here’s a compilation of all the important releases for the ML developers from top companies like Google and Uber. Here’s what is new this week: Google Open-Sources Neural Tangent […]","categories":["AI News"],"tags":["Deep Learning Techniques","deepmind london","EfficientNet-EdgeTPU","Pixar","Tensorflow"],"author_name":"Ram Sagar","publish_date":"2020-03-20T09:00:00","publication_year":"2020","word_count":792,"keywords":["artificial intelligence","machine learning","AI","neural network","ML","Aim","deep learning","object detection","JAX","Deep Learning Techniques","EfficientNet-EdgeTPU","TensorFlow","deepmind london","Pixar","Tensorflow"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","neural network","Aim","TensorFlow","JAX","object detection"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/pixar-turing-award-efficientnet-neural-tangent-alphago-movie\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10040612,"title":"How Intel Solved The Wiring Issues Of Quantum Computers","content":"Intel and QuTech (a collaboration between the Delft University of Technology and the Netherlands Organisation for Applied Scientific Research) recently achieved a breakthrough in quantum scalability. In the current form, quantum computers chips have to be kept at freezing temperatures to maintain the fragile qubits on the processor. Qubits typically operate at about -273 Celsius. Such temperatures are difficult to achieve. Until now, this was a significant bottleneck. The research published jointly by the organisations showed Intel’s cryogenic controller, Horse Ridge, can accomplish the same high-fidelity results even at room temperature. Additionally, the team also demonstrated frequency multiplexing on two qubits using a single cable. This simplifies the wiring challenge, paving the way for fully integrated quantum control chips with the quantum processor. Intel Horse Ridge Intel unveiled its cryogenic chip Horse Ridge first in 2019. It uses specially designed transistors that send microwave control signals to Intel’s quantum computing chips. Horse Ridge uses a highly integrated system-on-chip (SoC) to simplify complex control electronics required to operate a quantum system. It helps in faster setup, improves qubit performance, and scales efficiently to higher qubit counts to solve real-world problems. Horse Ridge is designed to act as a radio frequency processor and programmed with instructions that correspond to basic qubit operations. It translates instructions into electromagnetic microwave pulses that can manipulate the qubit states. Named after one of the coldest regions in Oregon, Horse Ridge was designed to operate at cryogenic temperatures (~4K), a temperature so cold that the atoms almost stop moving. Cryogenic control chip, Horse Ridge, was developed to overcome a key ‘interconnect or wiring bottleneck’ often associated with quantum computing. This challenge lies between the quantum chip stored in low, cryogenic temperature in a dilution refrigerator and the control electronics (that control the qubits on the quantum chips) operated at room temperature connected through wires. As the number of qubits on the quantum chips increases, the number of wires leading to the control electronics outside of the refrigerator also increases. It is not a sustainable solution considering companies are steadily increasing qubits in the chips. Getting these control electronics to operate at cryogenic temperature is the key to overcome the wiring bottleneck, which Horse Ridge could achieve. Intel introduced the second generation of the chip in 2020. It brought the key control functions for the quantum computer operation into the refrigerator as close as possible to the qubits. This helped in reducing the complexity of wiring. Intel’s breakthrough The latest research shows a commercial CMOS-based cryo-controller that achieves coherent control of a two-qubit processor at the same fidelity as that of control electronics placed at room temperature. It is an important milestone and helps solve the scalability problem by leveraging multiplexing. It reduced the number of cables\/wires required for qubit control. The Intel and QuTech team ran a two-qubit algorithm called Deutsch-Jozsa. Deutsch-Jozsa is a deterministic quantum algorithm. The results were verified by randomised benchmarking that validated Horse Ridge’s capability to present a highly integrated and scalable solution for quantum control electronics. The technology can be directly applied to multi-qubit algorithms and noisy intermediate-scale quantum devices. Next up, the team hopes to fully integrate the controller chip and qubits by fabricating them in silicon to enable better quantum scalability. Wrapping up The new research cements Intel’s position in the evolving quantum computing ecosystem. While much of the R&D in quantum computing has been around qubits themselves, Intel has taken an innovative approach by improving the interconnects and control electronics. Researchers are increasingly turning towards building quantum computers with techniques used to develop most modern-day electronics. This approach offers huge advantages in terms of scalability and practicality.","excerpt":"Intel’s cryogenic controller, Horse Ridge, can accomplish high-fidelity results even at room temperature.","categories":["Global Tech"],"tags":["Intel"],"author_name":"Shraddha Goled","publish_date":"2021-05-24T12:00:00","publication_year":"2021","word_count":608,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","Scala","RAG","programming_languages:Scala","GAN","R","emerging_tech:quantum computing","Intel"],"extracted_tech_keywords":["AI","RAG","R","Go","Scala","GAN","programming_languages:R","programming_languages:Scala","programming_languages:Go","emerging_tech:quantum computing"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/how-intel-solved-the-wiring-issues-of-quantum-computers\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10089775,"title":"Why Devs Can’t Turn Back After Using Tailwind CSS","content":"Developers are the quickest to jump on any hype train that promises even the smallest improvements and ‘Tailwind’ is the latest target of this trend. This CSS framework has become quite divisive in the front-end programming community, with many thinking that it’s the best thing since sliced bread. Of course, its critics have dismissed it as just another flashy new framework. As with anything, the truth is somewhere in the middle. However, one thing is for certain—once devs start using Tailwind, there’s no turning back. What makes it special? Tailwind has been in development since November 2017 but it really exploded onto the scene at the beginning of 2022. The creator of the framework, Adam Wathan, has spoken at length about his dislike for semantic CSS. This led him to building Tailwind, which throws semantic CSS out of the window in favour of a utility-focused CSS approach. Semantic CSS refers to the process of naming various HTML elements with names that have a meaning. For example, instead of naming a checkout button something like ‘bottom red box’, it can be named ‘checkout’. However, when following this approach, devs are generally asked to semantically name the content in HTML while using these class names as hooks in CSS. This approach is generally used to allow devs to swap out stylesheets if they want to change the visual style of the website without breaking compatibility. Wathan’s approach, as detailed in his blog, was to optimise for reusable CSS mixed with HTML, a general deviation from the separation of concerns that semantic CSS follows. This is one of the main talking points when it comes to Tailwind. Tailwind mainly applies CSS by creating small and simple classes that can be directly applied to HTML elements, reducing the amount of CSS code that needs to be written. However, just as these features make things easier for a subset of developers, they’re a no-go for another group. The small class-focused approach of Tailwind generally results in a difficult-to-read HTML file and long lines of code. The ‘restrictive’ nature of Tailwind is also brought up as a common drawback because it only allows for a small subset of CSS styles to be used. With so many issues, developers are beginning to question: Is Tailwind overhyped? Beyond the hype Apart from the legion of tech bros claiming milliseconds of faster load times while using Tailwind, there are multiple other reasons for using it. Even though the learning curve might be unpalatable for veterans of semantic coding styles, the benefits are numerous. Even the creator of Tailwind couldn’t stomach changing his coding style until he did, stating, “The truth is you’re never going to believe me until you actually try it. If you can suppress the urge to retch long enough to give it a chance, I really think you’ll wonder how you ever worked with CSS any other way.” Indeed, the Tailwind website is filled with positive reviews of the benefits coders gained from working with the framework. Many devs echo the fact that semantic CSS was a mistake, with another common point being that devs can’t go back to ‘normal’ CSS after working with Tailwind. The framework also comes with a host of nice-to-haves, such as strong defaults that can be customised as per the users’ preferences, support for responsive websites and automatic cleanup of unused CSS classes. These features resemble automated ‘best practices’, which will definitely smoothen team projects. After relying on these features, it becomes difficult for devs to do them manually after switching away from Tailwind. The restricted nature of Tailwind is also a godsend for newer developers, who only need to refer to a single source of documentation rather than trawling through the web to find documentation for CSS styles. Tailwind also favours convention over configuration. While this may be blasphemous for developers who prefer the infinite configurability of CSS, it works wonders for both newer developers and larger teams who don’t want to deal with the problems associated with semantic CSS. Tailwind is also built from the ground up to reject semantic CSS completely, which means that developers must break their hard-coded habits and not use semantic CSS. Using semantics can cause many problems in Tailwind. As stated by user ‘frontendben’ on the Tailwind subreddit, “The point of utility CSS is that you should be able to tell what the class is doing with zero context\/needing to open the CSS file. CSS is a declarative language. Yes, it can be used to define styles in an imperative way, but that’s often where the biggest headaches in maintaining CSS comes from.” The battle of Tailwind against semantic CSS, or even the Block Element Modifier or BEM, approach is sure to be a long and arduous one. Semantics are one of the hardest problems to solve in programming but Tailwind might have found the solution by merely skipping over the problem.","excerpt":"Tailwind says reject semantic CSS and go toward utility CSS.","categories":["AI Highlights"],"tags":[],"author_name":"Anirudh VK","publish_date":"2023-03-21T17:03:04","publication_year":"2023","word_count":818,"keywords":["Go","programming_languages:R","AI","ML","programming_languages:Go","Aim","R"],"extracted_tech_keywords":["AI","ML","Aim","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/why-devs-cant-turn-back-after-using-tailwind-css\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10005573,"title":"Top 10 Trending Python Projects On GitHub: 2020","content":"As per the latest Data Science skills study, the data scientists and practitioners who were surveyed revealed that the top Language preferred for Statistical Modelling is Python, favoured by 65.2% proportion of the respondents. Python is the language of choice for statistical modelling among the Data Science community, and AI and analytics practitioners seeking to upskill, such as Python for Statistical Modelling; TensorFlow for Python Frameworks; Git for Sharing code, among others. Below here, we listed down the top 10 trending open-source projects In Python on GitHub. (The projects are listed according to their stars on GitHub). 1| Manim Stars: 24.6k About: Manim is an animation engine for explanatory math videos. It is basically used to create precise animations programmatically and runs on Python 3.7. Manim uses Python to generate animations programmatically, which makes it possible to specify exactly how each one should run. Know more here. 2| DeepFaceLab Stars: 19.2k About: DeepFaceLab is an open-source deep fake system created by iperov for face swapping. It provides an imperative and easy-to-use pipeline for people to use with no comprehensive understanding of deep learning framework or with model implementation required. This system provides a flexible and loose coupling structure for people who need to strengthen their own pipeline with other features without writing complicated boilerplate code. Know more here. 3| Airflow Stars: 17.9k About: Airflow is a platform to programmatically author, schedule and monitor workflows. The pipelines in Airflow allows for writing code that instantiates pipelines dynamically. To use this platform, you will need Python versions 2.7, 3.5 and more. It allows users to use Python features to create workflows, monitor, schedule and manage the workflows using the web app. Anyone with Python knowledge can deploy a workflow. It also has plug-and-play operators that are ready to handle your task on Google Cloud Platform, Amazon Web Services, Microsoft Azure and many other services. Know more here. 4|  GPT-2 Stars: 13k About: GPT-2 is a large transformer-based language model with 1.5 billion parameters, which is trained with a simple objective to predict the next word, given all of the previous words within some text. GPT-2 generates synthetic text samples in response to the model being primed with arbitrary input. It is a large-scale unsupervised language model which generates coherent paragraphs of text, performs rudimentary reading comprehension and machine translation. It can also perform question answering and summarisation. It can generate conditional synthetic text samples of unprecedented quality. Know more here. 5| Horovod Stars: 9.8k About: Horovod is an open-source distributed deep learning training framework for TensorFlow, Keras, PyTorch, and Apache MXNet. Developed by Uber, the goal of Horovod is to make distributed deep learning fast and easy to use. The primary motivation for this project is to make it easy to take a single-GPU training script and successfully scale it to train across many GPUs in parallel. It is fast and easy to use, achieves 90% scaling efficiency for both Inception V3 and ResNet-101, and 68% scaling efficiency for VGG-16. Know more here. 6| ML-Agents Stars: 9.3k About: The Unity Machine Learning Agents Toolkit (ML-Agents) is an open-source project that enables games and simulations to serve as environments for training intelligent agents. The agents can be trained using reinforcement learning, imitation learning, neuroevolution or other machine learning methods through a simple-to-use Python API. Some of its features include support for multiple environment configurations and training scenarios, flexible Unity SDK, built-in support for imitation learning, among others. Know more here. 7| XSStrike Stars: 8.3k About: XSStrike is a Cross-Site Scripting detection suite equipped with four handwritten parsers. It is an intelligent payload generator, a powerful fuzzing engine as well as an incredibly fast crawler. The key features of XSStrike include multi-threaded crawling, configurable core, WAF detection, complete HTTP support and more. Know more here. 8| NeuralTalk Stars: 5.1k About: NeuralTalk is a  Python and Numpy source code for learning Multimodal Recurrent Neural Networks that describe images with sentences. NeuralTalk2 is written in Torch and runs on the GPU. It also supports CNN finetuning, which helps a lot with performance. Know more here. 9| Xonsh Stars: 3.8k About: Xonsh is a Python-powered, cross-platform, Unix-gazing shell language and command prompt. It is a superset of Python 3.5+ with additional shell primitives from Bash and IPython. The language is meant for the daily use of experts & novices. It works on all major systems, including Linux, Mac OSX, and Windows. Know more here. 10| Optuna Stars: 3k About: Optuna is an automatic hyperparameter optimisation software framework, particularly designed for machine learning. It features an imperative, define-by-run style user API. It allows for an automated search for optimal hyperparameters using Python conditionals, loops, and syntax. It can also parallelise hyperparameter searches over multiple threads or processes, without modifying code. Know more here.","excerpt":"As per the latest Data Science skills study, the data scientists and practitioners who were surveyed revealed that the top Language preferred for Statistical Modelling is Python, favoured by 65.2% proportion of the respondents.  Python is the language of choice for statistical modelling among the Data Science community, and AI and analytics practitioners seeking to […]","categories":["AI Trends"],"tags":["GitHub","horovod","open source projects","open source projects on github","simple python project"],"author_name":"Ambika Choudhury","publish_date":"2020-08-25T13:00:11","publication_year":"2020","word_count":795,"keywords":["open source projects","data science","open source projects on github","machine learning","Keras","AI","neural network","simple python project","ML","TensorFlow","PyTorch","deep learning","analytics","horovod","GitHub"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","neural network","data science","analytics","TensorFlow","PyTorch","Keras"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-10-trending-python-projects-on-github-2020\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":50266,"title":"Behind The Code: Meet Abhishek Thakur, World’s First Kaggle Triple Grandmaster","content":"In the developer series, Behind The Code, we reach out to the developers from the community to gain insights on how their journey started in data science, what are the tools and skills they use and what’s essential for their day-to-day operation. For this week’s column, Analytics India Magazine got in touch with Abhishek Thakur, the Chief data scientist at boost.ai. to get a glimpse of his journey in becoming the world’s first Kaggle triple grandmaster and many more. How It All Began Abhishek Thakur, world’s first Kaggle triple grandmaster and chief data scientist at boost.ai has become one of the most popular contributors in the ML community. However, his journey into this field has not been straightforward On being asked whether his foray into the world of algorithms was an accident, Abhishek spoke about how his first stint with algorithms came during his graduation days, in the form of image processing algorithms. This later laid the foundation for his future endeavours and also fetched him an internship at the University of Warwick to work on medical image processing. After completing his Bachelors in Electronics Engineering from NIT Surat, Abhishek went on to get a master’s degree from the University of Bonn, Germany. His fascination for computers made him choose computer science for his master’s degree. While he is doing his Master’s, he equipped himself with some practical knowledge by working at Fraunhofer where he implemented OCR algorithms on microcontrollers. After graduating from the University of Bonn, he worked as a data scientist in Berlin, which didn’t last long. His unquenched curiosity to learn more and do research introduced him to the sophisticated world of machine learning. “I tried taking lectures on data mining and machine learning at the university but failed miserably. Most of my machine learning has been self-taught,” admits Abhishek speaking about early days of his journey into AI. Self-learning requires a lot of dedication and practice. Abhishek would dedicate some time for his thesis and the rest for machine learning practice. To squeeze more out of a day, he would even spend many sleepless nights at Fraunhofer. But it is totally worth it, insists Abhishek. “I was more interested in applied after gaining the theoretical knowledge but the lectures limited themselves to theory,” says Abhishek, talking about his craving for a hands-on experience. Unknowingly, Abhishek was already on the path which would later fetch him the Kaggle crown. Making Of A Kaggle Grandmaster A Kaggle triple grandmaster is one who has achieved grandmaster status in competitions, kernels and discussions on Kaggle. Being amongst the top 10 in a Kaggle competition can be considered as a decent achievement. One can even go ahead and contribute a relevant Kernel or participate in discussions. Topping all three is no mean feat and no one might even have thought of it until Abhishek Thakur showed how it is done. “I did not use any books. Just internet, research papers, blogs and YouTube videos to understand the concepts and Kaggle to apply what I learned” The much-needed nudge was provided by one of his friends when he spoke about a promising platform that holds machine learning competition and goes by the name ‘Kaggle’. His first competition on Kaggle was on facial recognition. The timing was just right for Abhishek. His machine learning hands-on experience was coming to fruition and then there is this Kaggle competition on image processing, which also coincidentally has been part of his projects since his graduation days. According to the competition, the participants were tasked with finding the features like the angle between eyes and lips to recognise emotions along with other tasks to complete the competition. However, Abhishek failed miserably at his first competition and ended up with a low rank. “I couldn’t do any machine learning,” laments Abhishek remembering his first Kaggle attempt. Instead of succumbing to the bitter after taste of failure, Abhishek skimmed through the solutions of the winners, read relevant papers and started implementing them on his own. He inculcated a healthy diet of solving previous Kaggle competitions on his own, checking the successful solutions and getting to the bottom of the approaches with the help of Google. This went on for almost 10 months during which, he finished his thesis and also landed a job as a Data Scientist in Berlin. “I first try to understand the problem and then build basic algorithms to solve that problem. This way, I build a ‘benchmark’ and then try to improve on the benchmark,” advised Abhishek when asked about how he would proceed with a problem. His consistent efforts and discipline won him his first Kaggle gold 6 years ago. After 18 more golds and many other medals, he achieved a world rank of 3 and then went on to become the world’s first triple grandmaster earlier this year. “A right proportion of hard work, dedication, persistence, never giving up attitude and luck are the most important ingredients that helped me,” adds Abhishek. After 6 fruitful years in Germany, Abhishek decided to take a giant leap into the world of data science with boost.ai. Boost.ai is a Norwegian company that has been developing AI-powered direct messaging since 2016. He realised that at boost.ai people were building something really extraordinary and decided to move to Norway permanently where he has been working as a Chief Data Scientist since 2017. As a Chief Data Scientist, Abhishek builds the Natural Language Processing and Natural Language Understanding components and the deep learning models. He is currently tasked with improving the algorithms that provide answers in the bot and develop next-generation conversational AI platforms. About The Current State Of NLP via boost.ai “I think now is the right time for Natural Language Processing\/Understanding (NLP) and that’s what is happening,” posits Abhishek when asked about the current state of NLP. He also likens current sporadic rise in NLP research to new benchmarks to that of computer vision, half a decade ago. The NLP community witnessed the rise and rise of BERT ever since its introduction last year. It has even beaten all the benchmarks on GLUE (General Language Understanding Evaluation). “The benchmarks set by BERT and its contemporaries are great to solve almost all kinds of NLP problems,” opines Abhishek on the success of BERT and its variants. However, he also warns about the resource-hungry nature of these algorithms and the challenges that follow them when deployed for production. What Makes A Good Data Scientist A time-series dataset has to be processed in a different manner compared to a regular tabular dataset. For me, every new dataset or problem is an adventure. Abhishek’s fascination with data science comes from playing around with different algorithm and improving the existing ones. He is a frequent user of Tensorflow for NLP problems. Whereas, he prefers PyTorch for image problems. When it comes to favourite Python libraries, he stresses the significance of Scikit-learn and how it provides many necessary components to put a model into production. There is no dearth of libraries or frameworks one can use these days. There are many libraries one can use to build machine learning or deep learning models. It’s all good as long as one understands what is happening in the background. Along with scikit-learn and pandas, xgboost and lightgbm also have been part of many of Abhishek’s machine learning models. When it comes to cloud services, Abhishek is fond of Amazon AWS while he also confesses that he is slowly developing a liking for Google Cloud Platform. “Most of the wannabe data scientists think that this field is only about creating cool graphs and making models that can be shown off to a bunch of people,” says Abhishek when asked about the hype around data scientists. He expresses his utter disbelief at the latest obsession with presentations rather than the implementations. “If it’s in the presentations, it’s fine, if it’s in production, it’s usable,” quips Abhishek. He also highlights how challenging data collection can be and how useless a data scientist in the absence of data. All problems don’t need the expertise of a data scientist. Few can be solved with conventional approaches. The realisation of this, warns Abhishek, will derail the AI hype train at the enterprise level. A Word For The Beginners Being a self-taught machine learning engineer, Abhishek stands firm on the idea of utilising the resource-rich internet for learning the fundamental concepts. I have seen that a lot of beginners tend to give up too quickly. He lists courses such as the ones by Andrew Ng is the place to start for those who are serious about making it big in the field of ML. He also advises the aspirants to supplement online courses by reading blogs, reading papers on arxiv and skimming through discussion forums to fortify theoretical foundations of the subject. For getting a hands-on experience, not so surprisingly, Abhishek suggests the newcomers to take up Kaggle challenges. “Once you have solved a few problems, it will become very easy for you to start approaching machine learning problems just by looking at the data. Once that happens, you are no longer a beginner,” advises the Kaggle grandmaster. Abhishek bets big on perseverance and he blames the upcoming aspirants for the lack of it. He is very passionate about the practical aspects of this job and recommends the beginners to do the same.","excerpt":"In the developer series, Behind The Code, we reach out to the developers from the community to gain insights on how their journey started in data science, what are the tools and skills they use and what’s essential for their day-to-day operation.  For this week’s column, Analytics India Magazine got in touch with Abhishek Thakur, […]","categories":["AI Features"],"tags":["Interviews and Discussions","Kaggle"],"author_name":"Ram Sagar","publish_date":"2019-11-20T10:01:00","publication_year":"2019","word_count":1561,"keywords":["data science","machine learning","Kaggle","AI","ML","computer vision","NLP","Ray","deep learning","analytics","TensorFlow","Interviews and Discussions"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","NLP","computer vision","data science","analytics","Ray","TensorFlow"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/behind-the-code-meet-abhishek-thakur-worlds-first-kaggle-triple-grandmaster\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10056819,"title":"An Introductory Guide to MetaFormers in Computer Vision","content":"In recent years, we have seen the success of transformers in every field of machine learning and data science. This effect of the transformer can also be seen in computer vision. But the only problem with the transformers is that they are highly weighted and complex in terms of computation architecture. MetaFormer is one of the solutions to this problem. In this article, we will discuss MetaFormer with examples of how it can be applied to computer vision. We will also understand one of its sub-models, named as PoolFormer in detail. The Major points to be discussed in this article are listed below. Table of Contents What is MetaFormer?The method used for MetaFormerWhat is PoolFormer?PoolFormer’s Performance in Different TasksImage classificationObject DetectionSemantic Segmentation What is MetaFormer? MetaFormer can be considered as a kind of model architecture that is abstracted from transformers where the token mixer module is not defined and replaces the token mixer with attention or spatial MLP. We can think of MetaFormer as the transformer\/MLP-like model. We can find in many pieces of research that in computer vision using a transformer without specifying a token mixer module is more fruitful and essential for model performance. For example, these attention token mixtures can be replaced with a simple spatial pooling operator to perform basic token mixing. We can also make this procedure work with transformers in computer vision tasks like ImageNet-1k and DieT-B\/ResMLP-B24.  So this is a general concept of MetaFormer, where abstracted architecture from transformers is not required to specify a token mixture module. It can be a key player in the computer vision process to achieve noticeable higher performance. Here we can say we don’t need to provide focus on the token-mixture module. In the next parts of this article, we are going to discuss MetaFormer in detail before going for an example. Let’s have a look at the method which can be used for MetaFormer in the next section of the article. The Method used for MetaFormer As we have discussed in the above section, we can make a transformer work high performing without specifying token mixture and keeping the other components the same as in the transformer. Let’s take an example of ViT transformer, where It is the proposed input embedding as patch embedding: X = InputEmb(I), Where X is an embedding token of embedding dimensions C and sequence length of N. Now the embedding tokens can be fed to MetaFormer blocks where each block includes two sub-blocks. The first block can be represented as Y = TokenMixer(Norm(X)) + X, The work of the first block is to communicate information among tokens using a token mixture. Where in the above representation, Norm = normalization (layer or batch normalization), and TokenMixture = module for mixing tokens. The main function of any token mixture in any transformer is to propagate information. Now the second sub-block can consist of an MLP with non-linear activation. There can be two layers in the MLP. Representation of the second sub-block can be as follows: Z = σ(Norm(Y)W1)W2 + Y, Where W1 and W2 are learnable parameters of MLP. After this, we are not required to specify the token mixture. The below image is a representation of the comparative architecture of MetaFormer with other transformer and MLP-like models. Image source Hereby above, we can understand the basic difference between MetaFormers and transformer or MLP-like models. Now let’s take a look at an example of MetaFormer. In This article, we will be using the PoolFormer as an example of MetaFormer. What is PoolFormer? We have discussed how the design of transformers is focused on employing attention-based taken mixture. Here the PoolFormer, a model derived from MetaFormer for computer vision, is a work that focuses on employing the general architecture like MetaFormer. The architecture of the PoolFormer is contributing most to the success of the recent transformer and MLP-like models. To make it high-performing, the researchers of the PoolFormer have deliberately applied a simple pooling operator in the place of the token mixture. Where the working of a pooling operator is to make a token to be averagely aggregated to its nearby token instead of mixing the information. Also, this MetaFormer is only focused on the computer vision task and this becomes the reason for input to be in channel first format. In the PoolFormer the pooling operator has no learnable parameter and can be expressed as: Where, T = Tokens, and K  = pooling size. It is just like other normal pooling actions which make the network calculation easier. In most cases, we see that a normal neural network has high computational complexity and a high number of learnable parameters where pooling becomes very easy in computation with no learnable parameters. This is the advantage of using a pooling operator instead of any network on it. The below image is a representation of the architecture of PoolFormer. When we talk about the workflow of the PoolFormer, we can represent it using the below image. The above image is a representation of the overall framework with PoolFormer blocks. It seems like the architecture of CNN models. This framework has 4 stages with H\/4 x W\/4,  H\/8 x W\/8,  H\/16 x W\/16  and H\/32 x W\/32 tokens respectively. (H and W are the height and width of the input images). The MLP expansion ratio is set as 4 for now and According to the above model scaling, we can obtain 5 different model sizes of PoolFormer and their hyperparameters can be as follows: Here we have seen how the pool former works behind, now we have some of the results of the PoolFormer on different datasets that can help us in defining whether the PoolFormer is working in a competitive nature or not. PoolFormer’s Performance in Different Tasks In computer vision, we see various datasets that can be used for checking the performance of any model for any specific task. In this section of the article, we will see the performance of PoolFormer in tasks like image classification, object detection, and semantic segmentation. Image Classification For image classification tasks one of the landmark datasets is ImageNet- 1k dataset which can also be used for various computer vision tasks. In the dataset we have 1.3 million training samples and 50,000 validation images, covering 1000 common classes. The below table is a representation of the PoolFormer in the ImageNet classification dataset. Here in the table, we can see the comparative performance levels of the PoolFormer and other models and by seeing the accuracy reports on the validation set we can say that PoolFormer can achieve a competitive performance with CNN and other MetaFormer-like models. Let’s see the performance of the object detection procedure. Object Detection For object detection, there is a landmark dataset named COCO. The dataset includes 118 training samples and 5k validation samples. The below table is a representation of the performance level of PoolFormer and ResNet models on the COCO dataset for object detection. Semantic Segmentation Dataset ADE20K is a landmark dataset for semantic segmentation that includes 20,000 images in its training set and 2000 images in the validation set while covering 150 semantic categories. The below table is a representation of the performance level of models ResNet and ResNeXt that are CNN-Based, PVT that is transformer-based, and PoolFormer. Here we can see that PoolFormer-12 achieves mIoU of 37.1, 4.3, and 1.5 better than ResNet-18 and PVT-Tiny, respectively. Here we have seen the performance level of PoolFormer in various computer vision tasks, by looking at the results we can say that the MetaFormers are equivalent to the other transformers.  In the case of computer vision, we can use them for these tasks. Github repository comprising more details of PoolFormer can be found here. Final Words Here in the article, we have discussed the intuition behind MetaFormers and how it solves the computational complexity of the transformer. Along with that, we have discussed the basic method which can be followed for MetaFormer. We also discussed an example of MetaFormers in the computer vision field and compared the performance level based on different computer vision tasks with other transformer and MetaFormer-like models. Reference PoolFormer’s Github Repository","excerpt":"We can think of MetaFormer as the transformer\/MLP-like model where the token mixer module is not defined and replaces the token mixer with attention or spatial MLP","categories":["Deep Tech"],"tags":["AI (Artificial Intelligence)","Computer Vision","computer vision neural network","Data Science","Data Scientist","Deep Learning","Machine Learning","Python","Transformers"],"author_name":"Yugesh Verma","publish_date":"2021-12-22T15:00:00","publication_year":"2021","word_count":1359,"keywords":["computer vision neural network","data science","machine learning","AI","neural network","ML","Machine Learning","Transformers","computer vision","RAG","Python","object detection","Computer Vision","Deep Learning","Data Science","Data Scientist","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","machine learning","ML","neural network","computer vision","data science","Transformers","RAG","object detection","R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/an-introductory-guide-to-metaformers-in-computer-vision\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10118201,"title":"7 AI Startups that Featured on Shark Tank India Season 3","content":"As Shark Tank India’s Season 3 wraps up, we look back to assess the surge in AI startups stepping into the tank to pitch their ideas. This was indeed the first time we saw AI startups making an impact on the show. While some secured funding, others were still in their early stages. Interestingly, only four out of the seven were able to get investment. The panel featured Aman Gupta from boAt, Ritesh Agarwal of OYO Rooms, Deepinder Goyal of Zomato, Anupam Mittal from Shaadi.com, Namita Thapar of Emcure Pharmaceuticals, and Vineeta Singh from Sugar Cosmetics. Besides, there was Radhika Gupta of Edelweiss Mutual Fund, Peyush Bansal from Lenskart, Amit Jain of CarDekho, Azhar Iqubal from Inshorts, Varun Dua from ACKO, and Ronnie Screwvala from UpGrad. Let’s look at the AI startups that featured on Shark Tank India. Model Verse Model Verse, founded by IIM Kozhikode graduate Srijan Mehrotra, creates images of models for advertising and catalogs using AI. Srijan said that high quality product images are crucial for marketing, but traditional methods like hiring models can be expensive and time-consuming. In his pitch, he claimed that the platform can produce professional photos of models, saving brands a lot of money and time. Each image created on his platform costs only INR 35, a tiny fraction of the usual INR 3.5-4 lakh that models charge. Mehrotra developed the tool using PyTorch. The founder of Shaadi.com, Anupam Mittal, questioned whether Model Verse could be easily replaced by OpenAI’s Dall-E 3 and said that building wrapper-based models on OpenAI’s API has become more feasible nowadays. Mehrotra claimed that he had built Model Verse from scratch and needed GPUs to scale. He secured a deal from Anupam Mittal, Ritesh Aggarwal, and Amit Jain for INR 25 lakh in exchange for 10% equity. AI Kavach AI Kavach by Panoplia.ai, founded by Pratyusha Vemuri, is a platform that offers AI-powered solutions to tackle online fraud. The tool can detect fraudulent websites, calls, apps, and messages. Online fraud is a significant concern for individuals and businesses in India, resulting in substantial financial losses. Till date, the company has detected 200 fake messages and websites. During her pitch, Vemuri shared her personal story of falling prey to a fraudulent website while purchasing a swing. Before founding AI Kavach, Vemuri served as the head of product identity, security, and privacy at Microsoft. She claimed that AI Kavach can detect URLs in real time and, with the use of AI, predict whether a website is fraudulent or not. She also added that the algorithm is trained on millions of websites. Her plan is to charge consumers INR 99 per month for the service. Vemuri secured an INR 1 crore deal from Aman Gupta and Peyush Bansal, offering 5% equity in her company in return. Beauty GPT Beauty GPT is a product of Orbo.ai, founded by Abhit Sinha, Manoj Shinde, and Danish Jamil. Similar to a Snapchat filter, BeautyGPT is a virtual makeup simulation which helps users visualise how they would look with makeup on. This includes experimenting with various products such as lipstick, blushes, highlighter, eyeliner, etc. Using a combination of machine learning algorithms and LLMs, it analyses various data points, including facial attributes, customer demographics (age, skin type, etc.), skin concerns and weather impact (e.g., recommending different moisturizers for dry vs. humid weather). Personalised recommendations can help customers find the products they need, potentially increasing sales conversions for cosmetic companies. During the Shark Tank pitch, Peyush Bansal offered to buy 51% of the company for INR 15 crore. However, Orbo.ai secured a deal of INR 1 crore for 1% equity from Vineeta Singh. Ai Cars The founder of Ai Cars, Harshal Mahadev Nakshane, from Yavatmal, Maharashtra, presented an AI-powered hydrogen fuel cell car prototype. During his Shark Tank pitch, Nakshane said that he built the car in his garage in about 18 months. Judges Anupam Mittal, Vineeta Singh, and Namita Thapar took a test drive of the vehicle, during which it autonomously made sharp turns and was able to navigate on the road, leaving a strong impression on the judges. The hydrogen fuel cell-powered AI Car stands out with its rapid 5-minute refueling time and an impressive range exceeding 1,000 kilometers – all achieved with an investment of INR 60 lakh. However, Nakshane didn’t receive funding because the Sharks felt that the product is not scalable and that it would be difficult to compete with established players like Tesla and Google Waymo. FUTR STUDIOS FUTR STUDIOS, co-founded by Himanshu Goel and George Tharian, specialises in developing AI influencers for marketing and advertising. Kyra, India’s first AI influencer, is also a product of FUTR Studios. Today, Kyra has more than 2.5 lakh followers on Instagram and has collaborated on campaigns with brands like boAt, Titan, Morris Garages, and more. FUTR Studios develops high-fidelity, 3D virtual humans that can be used for various applications. They integrate these virtual humans into metaverse platforms, enabling interactive experiences within virtual worlds. The company further plans to build autonomous virtual humans that can think on their own using LLMs. The company failed to get a deal from the sharks. upliance.ai upliance.ai is an AI-powered home appliance company founded by Mahek Mody. upliance.ai’s AI Cooking Assistant comes in the form of a jar with a smart 8-inch screen attached. It can prepare over 500+ dishes, including paneer stir fry, chole masala, steamed rice, and chai, for instance. The screen comes with the in-built ‘ChefGPT’, which answers all your queries about the recipes and suggests ingredients for the dish. Moreover, it can suggest new recipes based on the ingredients you have at your disposal. Not only does it provide step-by-step instructions while you are cooking, it also enhances your overall cooking experience. The company wasn’t able to secure funding on the tank. However, the company later secured an investment of INR 34 crore in a seed round at a valuation of INR 143 crore from Khosla Ventures. Kibo Kibo by Trestle Labs was one of the unique AI companies pitched on Shark Tank 3. Kibo, which stands for Knowledge in a Box, is an AI-based education tool that converts printed documents, handwritten notes, PDFs, and digital text into audio for listening. This can be particularly beneficial for the visually impaired or those who prefer audio learning. The brand’s journey started during the engineering days of founders Akshita Sachdeva and Bonny Dave. While in her third year of computer science engineering, Sachdeva worked on a project focused on creating reading and mobility hand gloves for the visually impaired. The company secured a funding of INR 60 lakh for 6% equity from Peyush Bansal and Ronnie Screwvala.","excerpt":"Only four out of the seven were able to get investment.","categories":["Deep Tech"],"tags":["AI Startups"],"author_name":"Siddharth Jindal","publish_date":"2024-04-15T15:38:19","publication_year":"2024","word_count":1109,"keywords":["Go","machine learning","OpenAI","AI","PyTorch","Scala","Git","RAG","Aim","R","AI Startups"],"extracted_tech_keywords":["AI","machine learning","OpenAI","Aim","PyTorch","RAG","R","Go","Scala","Git"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/7-ai-startups-that-featured-on-shark-tank-india-season-3\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":39089,"title":"6 Habits All Successful Data Scientists Should Have","content":"A significantly increasing number of people making a transition into data science, a huge number of students are taking up data science courses — with so much happening, there is no doubt that the domain is witnessing a tremendous rise, proving the industry that Data science is actually the “sexiest job of the 21st century”. However, many people are not aware that along with all the knowledge and skills, there are few habits that every effective data scientist should have. In this article, we will have a look at some of those habits that need to adopt right away if you are a data scientist. Here Are The 6 Habits Every Successful Data Scientist Has Have A Strong Hand On Core Skills To be a successful data scientist, it is imperative that you have a strong foundation. One has to be completely, crystal clear about all the core and basic skills. A data scientist with a strong base has more chances of gaining advance knowledge. Whether it’s a language, a tool, or a concept, make sure you keep polishing your base skills and knowledge. According to one of our studies,  Python continues to reign as the tool of choice and the demand for Python professionals is significantly high. Almost 17% of all advertised analytics jobs in India demand for Python as a core skill. Therefore, make sure, you never lose your grip on the core concepts and skills. Keep An Active Eye On The Latest Trends A successful Data Scientist is always hungry for knowledge. Data science is a very vast field of work — no matter how knowledgeable you are, there is always something new to learn. So, even though if you are working on a particular vertical from the data science domain, always keep an eye on what’s going on in the industry; what new technologies are coming up. And also, make sure you gather all the required resources to teach yourself new skills. The more you know about the domain, the more you increase your chances of advancing your career. Always Offer A Helping Hand One of the most important traits of a successful data scientist is that s\/he is always active when it comes to helping other fellow data scientist. Whether in terms of office or community, make sure you never say ‘no’ in helping other data scientist. Also, there are several forums available on the internet where data science enthusiasts seek solutions to their problems. So make sure are active on a significant number of such forums and answer their queries. One of the best platforms is definitely Quora — you can proactively search of questions and answers them. By doing that you not only help others in the community but also expand your profile as a competent data scientist. They Are Not Only Data Scientists But Also Storytellers In one of my previous articles, I have mentioned about data storytelling and why a data scientist needs to be a great data storyteller. Over the years, data has evolved and with time, it is getting deeper and more complex. So, in order to bring in simplicity in it, provoke thought, and bring out deeper insights, a data scientist should know the art of data storytelling. So, if you are a data scientist, don’t just stick to numbers and analytical skills, rather learn the skill of data storytelling. Experience Is On Top Of The Priority List Education is important, no doubt, but sometimes it is the experience that matters the most. Gone are the days when qualifications on your resume would land you a job — today, the data science world needs professionals who have the ability to break down large, complex problems, and that ability can only be gained through experience. So, if you are a data scientist who has just started his\/her career, make sure you focus a lot and gain a significant amount of knowledge and experience. If you have not entered the industry yet, try looking for projects. Connect With Other Data Scientist From The Industry An effective data scientist always makes connections that deliver value to their career as well as to the community. Making connections help you get a clear picture of the current industry trends. That is not all when you connect to leaders from the industry you open doors of opportunities — it is one of the effective ways to expand your career. One of the most effective ways to make connections is to attend events dedicated Data scientist such as MachineCon, Cypher etc. The major advantage of attending events and conference is you make connections in person. So, do consider attending events and conferences.","excerpt":"A significantly increasing number of people making a transition into data science, a huge number of students are taking up data science courses — with so much happening, there is no doubt that the domain is witnessing a tremendous rise, proving the industry that Data science is actually the “sexiest job of the 21st century”. […]","categories":["AI Trends"],"tags":["Career","Data Science","Data Science Career","Data Scientist","data scientist qualifications","qualifications for data scientist"],"author_name":"Harshajit Sarmah","publish_date":"2019-05-13T12:31:30","publication_year":"2019","word_count":777,"keywords":["data science","Go","data scientist qualifications","AI","programming_languages:R","programming_languages:Go","qualifications for data scientist","Data Science Career","Python","analytics","programming_languages:Python","Data Science","Data Scientist","R","Career"],"extracted_tech_keywords":["AI","data science","analytics","Python","R","Go","programming_languages:Python","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/6-habits-all-successful-data-scientists-should-have\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":34791,"title":"Publicis Sapient Elevates Nigel Vaz As Global CEO","content":"Publicis Groupe on Monday announced that Nigel Vaz will become global Chief Executive Officer of Publicis Sapient. Vaz was previously the CEO of Publicis Sapient International (EMEA and APAC), and will now be in charge of leading Publicis Sapient’s 35 offices and 20,000 people across the globe. Vaz is also leading a rebranding of Publicis Sapient, which will see the company come together to create a single unified brand laser-focused on combining creatives, technology, research and analysis in order to provide a comprehensive service to clients in the digital business transformation (DBT) space. The Publicis Sapient brand relaunch will shape the future of the organisation, which is uniquely positioned to succeed in the DBT space. By fusing its experience, capabilities, customer experience and viewpoint, along with innovative engineering, within a culture renowned for problem-solving creativity, Publicis Sapient will be able to leverage its skillsets in management consulting and other emerging areas such as data and AI to deliver an authentic customer-centric approach to its clients, helping them digitally transform in the process. This people-centric approach to digital business transformation will continue to evolve alongside the preferences of clients and their consumers. “I’m honoured to take on the role of Publicis Sapient CEO at this time,” stated Vaz. “Technology is reshaping how our clients create value in the marketplace and require execution at an unprecedented speed. I look forward to further integrating our consulting, industry expertise, experience and engineering capabilities at scale while enhancing the culture and values, which have made us an integral business transformation partner since our inception.”","excerpt":"Publicis Groupe on Monday announced that Nigel Vaz will become global Chief Executive Officer of Publicis Sapient. Vaz was previously the CEO of Publicis Sapient International (EMEA and APAC), and will now be in charge of leading Publicis Sapient’s 35 offices and 20,000 people across the globe. Vaz is also leading a rebranding of Publicis Sapient, […]","categories":["AI News"],"tags":["Publicis Sapient"],"author_name":"Prajakta Hebbar","publish_date":"2019-02-12T10:40:04","publication_year":"2019","word_count":259,"keywords":["API","programming_languages:R","AI","Git","RAG","ViT","Publicis Sapient","GAN","R"],"extracted_tech_keywords":["AI","RAG","R","Git","API","GAN","ViT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/publicis-sapient-elevates-nigel-vaz-as-global-ceo\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10171980,"title":"Musk ‘in Talks’ to Finalise $4.3 billion in New Equity Funding for xAI","content":"Elon Musk’s AI startup xAI is in talks to finalise $4.3 billion in new equity funding and has plans to raise another $6.4 billion in capital for the next year, the company has told investors, according to a report by Bloomberg. Additionally, Morgan Stanley is helping secure an additional $5 billion in debt, as previously reported, to finance the development of xAI’s data centres. Other companies are opting for project financing instead. The company is working to raise enough funds to cover its significant expenses. Since its founding in 2023, it has raised $14 billion in equity, but only $4 billion remained at the start of the first quarter. People familiar with the matter told Bloomberg that the company expects to use nearly all of that in the second quarter. Furthermore, the report indicated that xAI may receive a $650 million rebate from one of its suppliers, which will help the company reduce expenses. XAI, which purchased X earlier this year, is believed to have a valuation of $80 billion by the conclusion of the first quarter, an increase from $51 billion at the close of 2024. According to the media outlet, xAI is also losing $1 billion monthly as costs for advanced AI models exceed limited revenues. This rapid cash burn underscores the significant financial challenges in the AI industry, particularly for xAI, where revenue has been slow to generate. The company plans to spend over half of it in the next three months, even before receiving the money. xAI, which drives chatbot Grok, anticipates a cash burn of approximately $13 billion throughout 2025, as indicated in the firm’s levered cash flow information provided to investors. Consequently, its extensive fundraising initiatives are just managing to match its costs, as noted by the sources. However, commenting on the cash burn, Musk replied to a post on his social media platform X, saying, “Bloomberg is talking nonsense.”","excerpt":"Despite raising $14 billion since 2023, the company is under financial strain.","categories":["AI News"],"tags":["Elon Musk","xAI funding"],"author_name":"Smruthi Nadig","publish_date":"2025-06-18T18:01:51","publication_year":"2025","word_count":316,"keywords":["API","funding","programming_languages:R","AI","xAI funding","GAN","Elon Musk","XAI","xAI","R","startup"],"extracted_tech_keywords":["AI","xAI","R","API","GAN","XAI","startup","funding","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/musk-in-talks-to-finalise-4-3-billion-in-new-equity-funding-for-xai\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10171634,"title":"CoRover Launches BharatGPT Mini, 534 Mn Parameter Model in 14 Indic Languages","content":"CoRover unveiled its latest offering, BharatGPT Mini, at VivaTech Paris 2025. The lightweight Small Language Model (SLM), trained on CoRover’s proprietary conversational dataset, features approximately 534 million parameters and supports 14 Indic languages for text-in, text-out tasks. The launch was officiated by Shri Jitin Prasada, Union Minister of State for Commerce & Industry and Electronics & IT. This multilingual AI model can work offline, making it ideal for businesses and citizens for access to healthcare, education, government services, banking, and more in 14 Indian languages, simply by speaking. Designed for both edge and server use, BharatGPT Mini enables efficient AI performance on low-end devices—addressing India’s growing demand for accessible, fast, and privacy-first AI systems that don’t rely on cloud infrastructure. The model’s compact size and optimised design make it ideal for deployment in low-compute environments. “I extend my sincere gratitude to Honorable Union Minister Mr. Jatin Prasad and Additional Secretary MeitY and India AI CEO Mr. Abhishek Singh for their invaluable contribution and support. BharatGPT Mini is a milestone in democratising access to intelligent conversational AI agents,” said Ankush Sabharwal, CEO at CoRover. The model is initially being released under a limited track and is available via CoRover Builder, a no-code\/low-code DIY platform. The Builder tool allows users to create AI Assistants in minutes, supporting interactions through text, voice, and video across web, telephony, and IoT systems. It includes multi-language support, allowing quick deployment of intelligent agents without extensive technical overhead. CoRover forecasts a fivefold increase in SLM adoption among small and mid-sized businesses by FY 2026. With cloud costs rising and data privacy becoming more critical, the company expects BharatGPT Mini to emerge as the go-to AI model for resource-limited environments.","excerpt":"The launch was officiated by Shri Jitin Prasada, Union Minister of State for Commerce & Industry and Electronics & IT.","categories":["AI News"],"tags":[],"author_name":"Mohit Pandey","publish_date":"2025-06-11T17:45:15","publication_year":"2025","word_count":282,"keywords":["Go","AI assistants","programming_languages:R","AI","programming_languages:Go","SLM","GPT","AI agents","R","llm_models:GPT"],"extracted_tech_keywords":["AI","SLM","AI assistants","R","Go","GPT","AI agents","llm_models:GPT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/corover-launches-bharatgpt-mini-534-mn-parameter-model-in-14-indic-languages\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":27238,"title":"7 Indian Union Ministries Who Have Embraced Artificial Intelligence Big Time","content":"It is no secret that Artificial intelligence has seeped into citizen’s lives. On a larger scale, AI in India has also become a crucial part of numerous government initiatives. The NDA-led BJP government has put enough importance towards new technologies by starting various new programmes. They have also set up an AI Task Force to prepare India for the Industrial Revolution 4.0. In fact, through a series of events and speeches from his tenure, Prime Minister Narendra Modi is seen to have been deliberately showcasing India as well as his government as technologically-forward. In an exclusive interaction with Analytics India Magazine, professor Kamakoti Veezhinathan, who heads the AI TasK Force, had explained that the venture was an amalgamation of thought leaders from multiple disciplines — academicians, government officials, corporates and people with other specialisations. We have listed top 6 Union Ministries who are using artificial intelligence on a very large scale in their operations: Ministry of Defence: The recently-constituted Artificial Intelligence (AI) Task Force of the Ministry of Defence led by Tata Sons Chairman N Chandrasekaran on submitted its final report to Defence Minister Nirmala Sitharaman on using AI for military superiority. They reportedly have made recommendations on how to make India a significant power in AI, in terms of both offensive and defensive needs, especially in aviation, naval, land systems, cyber, nuclear and biological warfare arenas. Initial tenders or RFIs (requests for information) will be floated over the next two years on dual-use AI capabilities. Ministry of Information and Broadcasting: A tender released by the Broadcast Engineering Consultants India Limited (BCIL), a public sector enterprise, under the Ministry of Information and Broadcasting recently showcased that the reigning power seems to be taking the public opinion and media seriously. The Government had sent out a proposal looking for a specific “technology platform” which would sense general public emotion by analysing social media posts, blogs and even emails to help boost nationalism and neutralise any “media blitzkrieg by India’s adversaries.” Ministry of Railways: Indian Railways has always been on the receiving end of criticism for its food catering services. Now with the help of artificial intelligence, these problems will come to a halt, thanks to the transition it aims to bring about in trains. AI will transform the way food is prepared in the trains’ kitchens and pantries. Along with revamping food menu in trains, its catering arm Indian Railway Catering and Tourism Corporation (IRCTC) is also promoting a greener environment with its biodegradable and environment-friendly food containers. The AI project has already been kick-started at the IRCTC headquarters. The facility has installed 16 high-definition cameras which are linked to large monitors for AI vision detection. Pests in these facility will be detected instantly, and the report of these instances will be notified to higher authorities with top priority. NITI Aayog (the former Planning Commission): The National Institution for Transforming India (NITI Aayog) unveiled a discussion paper which addresses the national strategy on artificial intelligence and other emerging technologies in India. Here, the government think tank identified five sectors to focus its efforts towards implementation of AI to serve societal needs. The five sectors are healthcare, agriculture, education, smart cities and infrastructure, smart mobility and transportation. Ministry of Home Affairs: One of the newest projects by the Delhi Police, which falls under the Home Ministry, will be to install India’s first Intelligent Traffic Management System (ITMS). The system will reportedly for creating smart traffic signals, using AI to determine traffic flow, automated enforcement and communication to change the traffic problems in India’s capital. As of now, the first phase is likely to be completed by April 2019. Ministry of External Affairs: In order to increase the flow of information between countries, the Ministry of External Affairs had recently held a closed door meeting of global AI experts to discuss how to attract Indian diaspora. Ministry of Corporate Affairs: The government is taking strong action against shell companies with the Ministry of Corporate Affairs planning to use artificial intelligence to MCA21 portal to detect finer discrepancies in details provided in company balance-sheets. The ministry had recently taken a drastic step by terminating 2.36 lakh non-compliant companies and investigating their bank account trails.","excerpt":"It is no secret that Artificial intelligence has seeped into citizen’s lives. On a larger scale, AI in India has also become a crucial part of numerous government initiatives. The NDA-led BJP government has put enough importance towards new technologies by starting various new programmes. They have also set up an AI Task Force to […]","categories":["AI Trends"],"tags":["AI (Artificial Intelligence)","BJP","government of india","Narendra Modi","NITI Aayog"],"author_name":"Prajakta Hebbar","publish_date":"2018-08-13T05:36:22","publication_year":"2018","word_count":700,"keywords":["government of india","Go","API","artificial intelligence","programming_languages:R","AI","programming_languages:Go","BJP","Aim","analytics","Narendra Modi","AI (Artificial Intelligence)","R","NITI Aayog"],"extracted_tech_keywords":["AI","artificial intelligence","analytics","Aim","R","Go","API","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/7-indian-union-ministries-who-have-embraced-artificial-intelligence-big-time\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10060773,"title":"Council Post: Data democratisation and AI—The superpowers to augment customer experience in 2022","content":"If there is one thing we have learned over the last few years, consumers prefer personalisation, and they yearn for human touch when they interact with brands. If they do not get that, they will move on. Customers expect a certain amount of guided selling and look out for products that interest and appeal to them. Today, consumer brands such as Amazon, Flipkart, Facebook (now Meta), Shopify, etc., are trying their level best to woo consumers and partners with products and promotions uniquely customised to their preferences and habits. A McKinsey study revealed that personalisation drives performance, and companies that grow faster drive 40 percent more revenue from personalisation than their slower-growing counterparts. The data further shows that when companies personalise communication: 76 percent of customers are more likely to purchase.78 percent of consumers are going to refer friends and family. But, the question is, how can businesses augment customer experience seamlessly to scale their business, offer personalised solutions, and grow faster? Big data and AI to the rescue Consumer brands face a sea of challenges to engage and retain their customers. Brands need to navigate challenges across operations, marketing efficiency, inventory management, price optimisation, logistics, etc. The struggle is real. Big data enables consumer brands to achieve all of these objectives seamlessly. It helps them anticipate consumer preferences and be prepared. Be it Flipkart’s Big Billion Days, Amazon Great Indian Festival, Netflix recommendations, or the most suited credit card options – use of data and AI-enabled technologies help these consumer brands create custom recommendations based on their preferences and affinity, resulting in an exclusive experience for individual consumers and improved customer service. In addition, technology such as these help with forecasting trends and making strategic decisions based on market analysis on the go. So, what’s stopping consumer brands from leveraging data and AI to power their consumer interactions and boost business efficiency? The answer is, while organisations have large amounts of data at their disposal, it is not easily accessible to all teams. Data democratisation, as the name suggests, is making the large amount of data accessible to relevant teams. If used correctly, it helps scale your business to the next level. Why is data democratisation important? Data democratisation essentially means that everybody has access to data. There are no gatekeepers. This free access is also accompanied by arming teams with the knowledge they need to use this data and expedite decision-making to contribute to the company’s growth significantly. Data-stingy businesses often suffer from the slow decision-making processes made by teams severely restricted in terms of agility. Data democratisation can propel businesses to new heights of performance. Benefits of data democratisation include: Empowers employees with self-service analytics. Helps companies in data governance for best use of data. Makes teams data literate. Data democratisation and AI With the powerful combination of data and AI at their fingertips, teams can gain deeper insights into their customers. Such technologies can also provide recommendations to the next-best-action. Critical decisions such as the right message, right channel, and time can be optimised to boost efficiency as well as delight consumers. For example, ecommerce brands can identify customers who buy from a specific luxury brand and personalise offers. Banks can determine customers who have not completed the onboarding journey and eliminate roadblocks to help them move towards completion. Music streaming apps can create custom playlists for each listener based on their preferred music and artists. In the past, these insights were gathered from multiple platforms, most times with the help of technology or data teams running Big Data queries. The time required to run these queries, draw insights, and then apply them was often long. Which meant, brands could not go to the market faster. However, today’s modern customer engagement platforms offer the benefits of data democratisation and AI within a single dashboard, most times requiring no-code. Forward-looking consumer brands have already invested in these capabilities and are seeing quick results. Data democratisation in different sectors BFSI: The banking and financial services sector receives a lot of data from massive amounts of customer interactions and compliance requirements. The industry can leverage data and AI to curate and generate content tailored for each customer. They can ensure the communication is delivered at the right moment and also perform quick and insightful segmentation to predict consumer expectations accurately. For instance, customer mapping can be done depending on lifestyle needs: housing loans, study loans for children, car loans, or credit cards for family members. This would pave the way for better customer interactions and offerings that are timely and useful. Analytics-driven personalised money management offerings could very well become the order of the day. Retail: The retail sector is highly dependent on customer engagement. Here data democratisation helps enhance the customer experience as broader access to these insights helps in the overall strategy. Retail brands can easily bridge the gap between their physical stores and digital assets ensuring a unified customer experience. Right algorithms at the right time are showing the way for retail growth, fulfilling customer needs and aspirations. Successful retail enterprises are customer-centric, offering hyper-personalised solutions to their customers with digital technologies at play, customised content, customised product, enhanced quality, customer rewards, and a holistic customer experience. A final thought Having AI and data democratisation in place can simplify brands’ data-driven decision-making. It can help ensure scale with confidence and speed. According to Accenture, 72 percent of companies successfully scaling AI in their organisation said that a core data foundation was key to their success. This article is written by a member of the AIM Leaders Council. AIM Leaders Council is an invitation-only forum of senior executives in the Data Science and Analytics industry. To check if you are eligible for a membership, please fill out the form here.","excerpt":"With the powerful combination of data and AI at their fingertips, teams can gain deeper insights into their customers. Such technologies can also provide recommendations to the next-best-action.","categories":["AI Features"],"tags":["data democratisation","Data Governance","Data Governance strategy","Democratisation","democratisation of AI"],"author_name":"Nalin Goel","publish_date":"2022-02-18T10:00:00","publication_year":"2022","word_count":963,"keywords":["big data","data science","Go","democratisation of AI","AI","ML","Git","RAG","Data Governance","Aim","Democratisation","data democratisation","analytics","R","Data Governance strategy"],"extracted_tech_keywords":["AI","ML","data science","analytics","Aim","RAG","R","Go","Git","big data"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/council-post-data-democratisation-and-ai-the-superpowers-to-augment-customer-experience-in-2022\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10016739,"title":"The Age Of Alt-facts: Why Your Business Must Focus On Alternative Data","content":"On December 31, 2019, BlueDot, an AI-driven health monitoring platform based in Canada, had alerted its customers about a cluster of unusual pneumonia cases happening around the Huanan Seafood Market in Wuhan, China. This was nine days before the World Health Organization notified the world about a novel coronavirus which came to be known as COVID-19. The power of alt-data: How alternative data can improve operational intelligence How did an AI-based epidemiologist beat not only the WHO but also the US Centers for Disease Control and Prevention (CDC) – who rolled out the notification six days after BlueDot – to the punch? The answer lies in data. More specifically, in external, alternative data. As opposed to the WHO and the CDC, which rely on official sources such as government and public health officials, BlueDot derives much of its predictive ability from data it collects from alternative sources. These datasets include, for example, the information of over four billion passengers on commercial flights travelling every year; climate data from satellites; the population data of humans, animals, and insects; and local information from journalists and healthcare workers gleaned daily from 100,000 online articles across 65 languages. This is the power of alternative data – and enterprises must ready themselves to tap into it. The bottom line: Why alternative data is integral to your business The year 2020 saw the rise of some interesting data-related trends that will continue to reign supreme in the post-pandemic world. The increased use of enterprise cloud services, and the subsequent data explosion, is one of them. In essence, modern businesses now have, at their disposal, a data-superlake that contains both conventional and alternative data in abundance. As enterprise databases, applications, processes, and workflows continue to migrate to SaaS-based platforms, the quantum of data available will only increase. This data inundation can either empower or overwhelm enterprises. However, if the derivative data – which comes from combinations, associations, and syntheses with different traditional and alternative datasets – is utilised appropriately. In that case, it can provide businesses with a clearer lens to view an uncertain market. Decision-making across all levels, from C-suite executives to the junior-most employee, can become more accurate and timely. A richer, deeper data strategy will also empower business leaders to identify and mitigate risks while driving superior experiences for all stakeholders, be it consumers, employees, investors, etc. Investors and financing companies have long known that traditional datasets can only do so much when it comes to providing a holistic picture of the financials involved. They use non-conventional datasets to fill the gaps in legacy sources, enriching them through unique, timely, and granular insights that may not be available through traditional routes. For instance, some of the traditional data points that lenders use to determine a borrower’s creditworthiness include their bank history, liquid capital, CIBIL score, etc. However, solely relying on these data may project an inaccurate image of the borrower’s creditworthiness. Suppose this borrower has a debt of INR 60,000 and can only afford to pay half the amount from their salary. However, to maintain their credit score, the person borrows the remaining amount from friends and family. If the lender uses non-conventional datasets such as the borrower’s bill payment behaviour, social media behaviour, airtime usage, etc., it can get a more accurate picture of their repayment capabilities and risk profile. Data-driven, data ready: The future of enterprise operations Gartner estimates that public cloud services will be essential for 90% of data and analytics innovation by 2022. Technologies that can extract, collate, clean, and analyse data from multiple sources will become more prevalent, driving massive growth in the volume of alternative datasets and making them as essential to holistic data stories as traditional datasets. The financial ecosystem is already using this alternative data to drive better risk assessment. BlueDot used it to warn the world about the imminent pandemic. It is high time that businesses across other sectors begin to augment their analytics strategy using alternative data. In the post-pandemic global economy, where data analytics is finding greater adoption as a reliable tool to navigate choppy waters, the need to capture and synthesise alternative data is no longer an option – it has become a critical business imperative.","excerpt":"On December 31, 2019, BlueDot, an AI-driven health monitoring platform based in Canada, had alerted its customers about a cluster of unusual pneumonia cases happening around the Huanan Seafood Market in Wuhan, China. This was nine days before the World Health Organization notified the world about a novel coronavirus which came to be known as […]","categories":["AI Features"],"tags":["Data Analytics","qlik","Qlik analytics India","qlik india"],"author_name":"Varun Babbar","publish_date":"2020-12-29T18:00:00","publication_year":"2020","word_count":704,"keywords":["Go","API","programming_languages:R","AI","data-driven","innovation","programming_languages:Go","GAN","qlik","Qlik analytics India","analytics","Data Analytics","qlik india","R"],"extracted_tech_keywords":["AI","analytics","R","Go","API","GAN","data-driven","innovation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/the-age-of-alt-facts-why-your-business-must-focus-on-alternative-data\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10088270,"title":"Here is An Open-Source RLHF Implementation of LLaMA","content":"Within a week of the release of Meta’s open-source LLM, LLaMA, we have an implementation of it based on Reinforcement Learning with Human Feedback (RLHF). ChatLLaMA, developed by Nebuly, claims to have a 15 times faster training process than ChatGPT, which is ideal for allowing developers to fine-tune and personalise ChatLLaMA assistant services. Since the implementation is built on LLaMA, which is significantly smaller and faster than GPT-3, it enables faster inference and cost-effective ChatGPT like assistants. This is enabled by the built-in support for DeepSpeed ZERO within ChatLLaMA. Click here to check out the repository. ChatGPT, the chatbot by OpenAI, is also built by implementing GPT-3.5 with RLHF. Since Meta’s LLaMA was not fine-tuned for instruction tasks, ChatLLaMA allows its implementation by bringing in RLHF. The library supports all LLaMA model architectures ranging from 7 billion to 65 billion parameters. This will allow more flexibility and preferences for training time and inference performance. The library also supports addition of custom datasets for the fine-tuning process along with Meta’s original weights. Moreover, it includes a built-in support for generating your own dataset. The developers are also calling out for open-source contributions from the community, since the library is in very early stages. Nebuly-AI, has been releasing open-source plug & play modules for unleashing the optimisation power, and assisting breakthroughs in the AI community. The developers have previously released Speedster, for reaching maximum inference on your hardware, Nos, for maximising performance of GPU resources in Kubernetes Cluster, OpenAlphaTensor, with custom generated matrix multiplication algorithm, Forward-Forward, for testing the Forward-Forward algorithm in PyTorch. They have several other projects in the pipeline, including a GPT Optimiser. Recently, Colossal-AI also released an open-source, PyTorch-based implementation of ChatGPT, that requires less computing resources that includes all the stages of making a chatbot.","excerpt":"The library supports all LLaMA model architectures ranging from 7 billion to 65 billion parameters.","categories":["AI News"],"tags":["ChatGPT"],"author_name":"Mohit Pandey","publish_date":"2023-02-28T12:22:56","publication_year":"2023","word_count":297,"keywords":["Go","ChatGPT","RLHF","OpenAI","PyTorch","AI","GPT","Aim","R","kubernetes"],"extracted_tech_keywords":["AI","ChatGPT","OpenAI","Aim","PyTorch","RLHF","kubernetes","R","Go","GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/here-is-an-open-source-rlhf-implementation-of-llama\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10009108,"title":"You Can Now Practice Machine Learning Skills On All-New MachineHack Platform","content":"MachineHack has launched a ‘Practice’ section on their all-new website for beginners to advance their machine learning skills. This news followed the recent announcement of MachineHack’s revamped website, which is currently offering an entirely new interface with improved features and robust user experience to the participants. And, with this new ‘Practice’ section on the website, beginners will now be able to put their skills in practice before participating in the highly intense MachineHack hackathons. Conceptualised in 2018, MachineHack is an engaging online platform that offers best-in-class hackathons for data scientists and ML practitioners to participate and elevate their skills to newer heights. Currently, the platform accommodates 20,000 data scientists and machine learning enthusiasts, who are participating daily to solve some of the most challenging real-world problems. While the revamped website will allow aficionados to participate and engulf themselves in near real-life business use cases, the brand new ‘Practice’ section will allow current as well as future data scientists to try their hands on some interesting machine learning tools. Hosted by Analytics India Magazine, MachineHack platform has always aimed to become the most engaging hackathon platform for the participants; and with its new, improved features, it is not only engaging but faster, extremely robust and hosts the best hackathons on data science, artificial intelligence and machine learning. Some of the advantages of practicing hackathons: Getting hands-on experience: With the new practice tab on the website, novice data scientists and machine learning enthusiasts can try and experiment with practice hackathons which are designed with beginners’ level of difficulty. Advance skills for real hackathons: By participating in these practice competitions, not only participants can test their level of knowledge but can also advance their skill sets to the next level. Creating an innovation mindset: Considering, innovation is the key to success, practising hackathons will enable design thinking and a mindset of innovation among participants. While the platform has recently been revamped, it is continuously adding more features for data scientists and ML practitioners to hone their skills on one of the most engaging hackathon platforms. The platform currently has more than 25 hackathons, and it comes up with a new hackathon every weekend for one to compete. Remember: Practice makes perfect, so before you try competing with some of the best minds in the industry, don’t forget to have a trial run.","excerpt":"MachineHack has launched a ‘Practice’ section on their all-new website for beginners to advance their machine learning skills. This news followed the recent announcement of MachineHack’s revamped website, which is currently offering an entirely new interface with improved features and robust user experience to the participants. And, with this new ‘Practice’ section on the website, […]","categories":["AI Features"],"tags":["Machinehack","Machinehack Hackathon","machinehack platform"],"author_name":"Sejuti Das","publish_date":"2020-10-07T18:00:33","publication_year":"2020","word_count":389,"keywords":["data science","artificial intelligence","machine learning","programming_languages:R","AI","Machinehack","innovation","ML","Aim","analytics","Machinehack Hackathon","machinehack platform","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","data science","analytics","Aim","R","innovation","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/you-can-now-practice-machine-learning-skills-on-all-new-machinehack-platform\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10132642,"title":"Acer Launches New AI-Powered Chromebook Laptops  in India","content":"Acer recently launched its latest Chromebook Plus models, the Acer Chromebook Plus 14 and 15 laptops, in India. The all-new Chromebook Plus with the latest built-in Google Gemini AI features offering robust performance, enhanced productivity features, and a sleek, professional design. These advanced laptops are specifically designed to cater to the demands of the enterprise and education sector. Acer Chromebook Plus offers built-in Google apps and powerful AI capabilities. It also offers Google Photos Magic Eraser, File Sync, Wallpaper generation, AI-created video backgrounds, and Adobe Photoshop on the web to help consumers boost their productivity. Powered by a range of Powerful Intel® & AMD® processor variants, these Chromebooks ensure robust performance for multitasking and running demanding applications. The Chromebook Plus 14 has two variants. One with Intel® Core™ i3-N305 processor and another with AMD® Ryzen® 7000 Series Processor, while the Chromebook Plus 15 offers Up to Intel® 13th Gen Core™ i7-1355U processor. All the models support up to 16GB LPDDR5X SDRAM and Storage up to 512 GB PCIe NVMe SSD, ensuring fast data access and ample space for important files and applications. “With powerful Intel® & AMD® processors, vibrant displays, Powerful AI capabilities, and robust security features, we believe these Chromebooks will significantly enhance productivity and learning experiences. Our goal is to offer solutions that empower professionals and students to achieve more, and the Chromebook Plus 14 and 15 embody this vision perfectly,” Sudhir Goel, Chief Business Officer at Acer India, said. Designed for durability and reliability, these Chromebooks have undergone rigorous military-grade reliability tests, including mechanical shock, transit drop, vibration, and resistance to sand, dust, humidity, and extreme temperatures. In a previous interaction with AIM, Goel said, “At Computex 2024, we have showcased a lot of new products, which we are thrilled to introduce to the Indian market in the coming year.” PC makers are hoping AI could help pull the market from the stalemate that it was last year. Research firm Canalys predicts that the PC market will see an 8% annual growth in 2024 as more AI PCs hit the market. Canalys also predicts AI PCs will capture 60% of the market by 2027.","excerpt":"These advanced laptops are specifically designed to cater to the demands of the enterprise and education sector.","categories":["AI News"],"tags":["HP"],"author_name":"Pritam Bordoloi","publish_date":"2024-08-14T11:37:16","publication_year":"2024","word_count":357,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","RAG","Aim","llm_models:Gemini","ViT","HP","R"],"extracted_tech_keywords":["AI","Aim","RAG","R","Go","ViT","llm_models:Gemini","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/acer-launches-new-ai-powered-chromebook-laptops-in-india\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10169920,"title":"WiDS Bangalore @ Intuit Returns","content":"As the world of data science continues to evolve and progress rapidly, the Stanford University-initiated Women in Data Science (WiDS) Bangalore @ Intuit returns with a fresh wave of inspiration, innovation, and community. Organised independently by Intuit India as part of the global WiDS initiative, this year’s conference spotlights some groundbreaking contributions to the rapidly emerging fields of Generative AI and Agentic AI. From hands-on research to enterprise-scale implementation, the event celebrates women pushing the boundaries of what’s possible in data science. Register Now!Event Date: June 12 | Time: 9:00 AM – 2:00 PM IST | Venue: Intuit India, Bengaluru What to Expect This Year Theme: Gen AI and Agentic AI This year’s focus is on ‌trends shaping the AI landscape. The event will host vision talks by accomplished industry leaders, followed by high-impact networking sessions, allowing attendees to connect, collaborate, and grow. Whether you’re a practicing data scientist, a student, or someone just breaking into the field, this event offers a front-row seat to the future of AI through the lens of data scientists who are building it. Why You Should Attend Learn about the latest breakthroughs in GenAI and Agentic AI Gain insights from leaders driving innovation in data science Be part of a community that supports empowerment and advancement in the field of data science Build meaningful connections with peers and industry experts Register Now!Event Date: June 12 | Time: 9:00 AM – 2:00 PM IST | Venue: Intuit India, Bengaluru The Mission Behind WiDS WiDS, short for Women in Data Science, is a global movement launched by Stanford University with a singular goal: to elevate women in data science through  community and opportunity. Last year, the event brought together over 130 data science professionals from academia, startups, GCCs, and IT services. The insightful sessions featuring various speakers, including leading data science experts, left the attendees feeling empowered and inspired. Over the years, WiDS @ Intuit has grown into a much-anticipated celebration of knowledge-sharing, mentorship, and recognition of excellence. Join us this year as we celebrate brilliance, break barriers, and shape the future—together.","excerpt":"This year’s focus is on ‌trends shaping the AI landscape. The event will host vision talks by accomplished industry leaders, followed by high-impact networking sessions, allowing attendees to connect, collaborate, and grow.","categories":["AI Highlights"],"tags":[],"author_name":"Mohit Pandey","publish_date":"2025-05-14T12:37:14","publication_year":"2025","word_count":346,"keywords":["data science","Go","GenAI","agentic AI","API","AI","innovation","generative AI","GAN","R"],"extracted_tech_keywords":["AI","data science","generative AI","GenAI","agentic AI","R","Go","API","GAN","innovation"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/wids-bangalore-intuit-returns\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10052197,"title":"A Complete Guide to Causal Inference in Python","content":"In data analytics and machine learning, when we apply the behavioural science insights in the studies, it always helps in improving the experience in delivering the results. One of the most important areas of behavioural science is the causal inference which is basically used for extracting cause and intensity of cause. It belongs to the family of loosely connected methods. In this article, we will have a detailed discussion of causal inference and we will try to understand its importance with hands-on implementations. The major points to be covered in the article are listed below. Table of Contents What is Causal Inference?Why is Causal Inference Important?Varieties of Causal Inference Causal Inference in PythonData Preparation and Mathematical AnalysisGoalsMaking AssumptionModelling the CounterfactualCovariate ImbalancePropensity ScoreUnconfoundedness and the Propensity ScoreTrimmingStratification What is Causal Inference? In statistics, there is always a question that comes to the mind of researchers that “why is something happening?”  Here the point which comes into focus is the causal inference which can be considered as the family of statistical methods whose main motive is to give the reasons for any happening. We use causal inference to determine the cause of changes in one variable if the changes occur in a different variable where standard statistical approaches like regression are being used to determine how the changes in one variable are associated with the changes in another variable. Let’s suppose there are two variables X and Y. The standard methods here will focus on determining the association whereas the causal inference approaches will be concerned about why the variable X changes if it is causally related with the variable Y so that we can explain changes in X in terms of changes in the Y variable. Why is Causal Inference Important? At every level of statistics, causal inference is used for providing a better user experience for customers on any platform. We can use the insights of causal inferences to identify the problems related to the customer or problems occurring in the organization. Also, it can be used to improve the customer experience. For example, any product introducing a new feature and the customers are raising complaints about the new feature due to lack of clarity or he is confused about the procedure of using the new feature. In such a scenario we can improve the communication about the usage or procedure of usage of a new feature instead of giving news updates on it or dropping down the new feature plans. Causal inference can be used to make information that can help in improving the user experience and also we can generate business decisions by knowing its impact on the business. If we can understand the relationship between two intangible variables such as employee satisfaction and business metrics, we will be able to use such information to prioritize tasks and aim for new features and tools. Also, these inferences can help in understanding the short-term and long-term impact of any new decision or program. Causal inference enables us to answer questions that are causal based on observational data, especially in situations where testing is not possible or feasible. For example, we started a campaign where users of our product can participate and mail their queries and complaints and we want to measure the impact of the campaign on the business. Causal inference enables us to find answers to these types of questions which can also lead to better user experiences on any platform. Varieties of Causal Inference Since causal inference is a combination of various methods connected together, it can be categorized into various categories for a better understanding of any beginner. We can say there can be two categories according to the data. These two categories are : Causal inference with experimental dataCausal inference with observational data Causal Inference in Python In this article, we are going to make causal inferences using observational data and also we will use a package named CausualInference for performing our analysis. Let’s start with an example where a supervisor notices that from his team of several labourers, labourers who are properly dressed up according to the norms tend to be less productive than the labourers who are not properly dressed up. The supervisor starts with making the data about the labourers where he puts value 1 for those who are dressed up and 0 for those who are in casual dresses and the productive section he puts 1 if the labourer is productive otherwise he puts 0. He makes the data for a whole week. Data Preparation and Mathematical Analysis Let’s make this type of data using python: import numpy as np import pandas as pd p_z = 0.5 p_x_z = [0.9, 0.1] p_y_xz = [0.2, 0.4, 0.6, 0.8] z = np.random.binomial(n=1, p=p_z, size=500) p_x = np.choose(z, p_x_z) x = np.random.binomial(n=1, p=p_x, size=500) p_y = np.choose(x+2*z, p_y_xz) y = np.random.binomial(n=1, p=p_y, size=500) generate_dataset_0 = pd.DataFrame({\"x\":x, \"y\":y}) generate_dataset_0 Output: Here in the data, we have 500 samples of labourers. Where X is about their dressing and Y is about the productiveness of labourers. The first question in casualty which comes to mind is what is the quantity of labourers who are dressed up but not productive. We can simply calculate it by: P(Y=1|X=1)−P(Y=1|X=0) Where p is probability and we can estimate the quantity in python using the following function. def estimate_uplift(ds): base = ds[ds.x == 0] variant = ds[ds.x == 1] delta = variant.y.mean() - base.y.mean() delta_err = 1.96 * np.sqrt( variant.y.var() \/ variant.shape[0] + base.y.var() \/ base.shape[0]) return {\"estimated_effect\": delta, \"standard_error\": delta_err} estimate_uplift(generate_dataset_0) Output: Here from this, we can infer that labourers who are dressed up are less productive. Using the above code we have estimated the means of two groups. Here the “estimated_effect” – the difference in mean values of y for productive and unproductive samples and  “standard_error” – 90% confidence intervals around “estimated_effect”. To be more sure about the estimation we can run the chi-square contingency test. from scipy.stats import chi2_contingency contingency_table = ( generate_dataset_0 .assign(placeholder=1) .pivot_table(index=\"x\", columns=\"y\", values=\"placeholder\", aggfunc=\"sum\") .values ) _, p, _, _ = chi2_contingency(contingency_table, lambda_=\"log-likelihood\") # p-value p Output: That’s a small p-value. Using the above-generated information we can estimate any labourer’s probability if we see them dressed up. But we need to believe that the observations are drawn from the same distribution. Till now we are using randomly generated small data for the analysis and according to that, we can suggest the supervisor use this information to decide whether the labourer should be dressed up or not. If the supervisor does this he fundamentally changes the system in which we are making inferences, this can alter or reverse the correlation that we observed. We can measure the changes in the system by randomized controlled trials, which is randomizing the observation about who is dressed up and who isn’t, and looking for different values in the productive section. Using this we can remove the effect of any confounding variables which can cause changes in the metric we care about. Let’s draw a function to generate the dataset so that we can intervene in the function directly and go on with the procedure of the A\/B test. import numpy as np import pandas as pd from sklearn.preprocessing import PolynomialFeatures def generate_dataset_0(n_samples=500, set_X=None, show_z=False): p_z = 0.5 p_x_z = [0.9, 0.1] p_y_xz = [0.2, 0.4, 0.6, 0.8] z = np.random.binomial(n=1, p=p_z, size=n_samples) if set_X is not None: assert(len(set_X) == n_samples) x = set_x else: p_x = np.choose(z, p_x_z) x = np.random.binomial(n=1, p=p_x, size=n_samples) p_y = np.choose(x+2*z, p_y_xz) y = np.random.binomial(n=1, p=p_y, size=n_samples) if show_z: return pd.DataFrame({\"x\":x, \"y\":y, \"z\":z}) return pd.DataFrame({\"x\":x, \"y\":y}) Running the A\/B test by generating 10000 samples where 50% of the sample of labourers is dressed up and 50% of the samples are for casually dressed up labourers. def run_ab_test(datagenerator, n_samples=10000, filter_=None): n_samples_a = int(n_samples \/ 2) n_samples_b = n_samples - n_samples_a set_X = np.concatenate([np.ones(n_samples_a), np.zeros(n_samples_b)]).astype(np.int64) ds = datagenerator(n_samples=n_samples, set_X=set_X) if filter_ != None: ds = ds[filter_(ds)].copy() return estimate_uplift(ds) run_ab_test(generate_dataset_0) Output: Here using this function we get an unbiased estimate of the average treatment effect. And we can see that the effect of being dressed up is almost reversed. By this analysis, we can say that the correlation doesn’t imply causality. To be more precise, in our condition X and Y are random variables and we want to measure the effect by forcing X to take a certain value on how the distribution of Y will get changed. We can call the procedure of forcing a variable to take a certain value intervention. In our example, if we force people to not be dressed up as per the norms we are making an intervention. Until now in our analysis, we have not done this. we can say the distribution of Y because of the interventional distribution will be. P(Y|do(X)) Till now what we have done is try to observe y distribution on the basis of the observation of the X variable. Mathematically we can say P(Y|X) We only have access to observational data and we are trying to answer how we can reason about the interventional distribution. running an A\/B test to help in measuring the effects of an intervention but this is impractical, unfeasible, or unethical. Here we want to be capable of saying what the effect of an intervention is? For being capable of that we need to make some assumptions about the data generating process. We can make a solution to this problem by just adding some potential outcomes which can be treated as other random variables. Let’s say these variables are Y0  and Y1 and also these random variables can not be directly observed. Y  is defined in terms of Y=Y1 when X=1Y=Y0 when X=0 Using those potential outcomes shifts the problem from one about how distributions change under the intervention, to one about data drawn Independent and identically distributed random variables with missing values. Goals Often determining the difference of means of two groups is enough (here the potential outcomes) and we call this difference as Average Treatment Effect (ATE) which is expressed as: Δ=E[Y1−Y0] Applying an A\/B test and comparison of the means gives the quantity that we are required to measure. Estimation of this quantity from any observational data gives two values ATT=E[Y1−Y0|X=1], the “Average Treatment effect of the Treated”ATC=E[Y1−Y0|X=0], the “Average Treatment effect of the Control” Which are related to the ATE. The Below techniques will help us to estimate the ATE, ATC, and ATT. Making Assumption Randomization of assignment of x makes us choose which potential outcome is revealed to us and this makes the outcome from the procedure independent of the variable X.  which means E[Y1|X=1]=E[Y1] To estimate the ATE we are required to use the other information of the data for this we are required to assume that we have additional information to completely explain the choice of treatment for each subject. We can write our assumption as: P(X,Y0,Y1|Z)=P(X|Z)P(Y0,Y1|Z) Where Z is the additional information random variable. In our example, we can improve it more. Let’s say that skilled people are more productive and they are less likely to be dressed up. So splitting the data according to the skill section will introduce various subgroups on which there is to be a positive relationship between the productivity and dressed up labourer. observed_data_0_with_confounders = generate_dataset_0(show_z=True) print(estimate_uplift(observed_data_0_with_confounders.loc[lambda df: df.z == 0])) print(estimate_uplift(observed_data_0_with_confounders.loc[lambda df: df.z == 1])) Output: This means that Z can be used to completely explain the X. and if Z does not contain any confounding variable then the assumption we are making can be wrong. Once the right assumption is made we can approach to estimate the ATE with various techniques and approaches. Some of these techniques are explained below. Modelling the Counterfactual The above intuition says that if we have the information of potential outcomes we can easily estimate the ATE so in the next I am going to generate a data set where I have modelled the Y0 and Y1. and the success of modelling of counterfactual depends on the modelling of the Y0 and Y1. in this link you will get all the dataset generators which can be used for practising the causal inference. Considering the size of the article I am not posting the data generator codes here. By visualizing the data we can see the insights. The above image is the representation of the data we have generated where y and z are our potential outcomes. Let’s make a scatter plot to understand it more. observed_data_1.plot.scatter(x=\"z\", y=\"y\", c=\"x\", cmap=\"rainbow\", colorbar=False); Output: Here we can see by this scatter plot that there is a very small difference between the groups and the distribution of Y. sns.kdeplot(observed_data_1.loc[lambda df: df.x == 0].y) sns.kdeplot(observed_data_1.loc[lambda df: df.x == 1].y) Output: By this density plot, we can say that there is a slight difference between the group lets check for the distribution of covariance Z for each group. sns.kdeplot(observed_data_1.loc[lambda df: df.x == 0].z) sns.kdeplot(observed_data_1.loc[lambda df: df.x == 1].z) Output: We can also confirm the difference by just looking at the difference in the mean of these groups. print(\"Observed ATE: {estimated_effect:.3f} ({standard_error:.3f})\".format(**estimate_uplift(observed_data_1))) Output: print(\"Real ATE:  {estimated_effect:.3f} ({standard_error:.3f})\".format(**run_ab_test(generate_dataset_1))) Output: Here, we have run an A\/B test which we should not do because it is not feasible and impractical as we have discussed. To know the real ATE we can use any regression model. As we know the equation for a simple regression model is: Y = α + βX By just looking at the equation we can say it is a perfect fit for our model and using the linear regression we can estimate the ATE. The package CausalInference gives the facility to perform this where we need only three values Y, D, and X. from causalinference import CausalModel cm = CausalModel( Y=observed_data_1.y.values, D=observed_data_1.x.values, X=observed_data_1.z.values) cm.est_via_ols(adj=1) print(cm.estimates) Output: Here we can see that the model has given the estimation of the ATE and with this, we also have a confidence interval where the ATE is lying and according to that, we can say that we have got a correct ATE. Note – we have modelled the data for good results. You may find some other results according to the complexity of the data. Covariate Imbalance Let’s think about a situation where we have data in which the covariance is in an imbalanced shape. We have data where we have only one type of sample in the data space at one time either treated or untreated. I have modelled the data for this again. There is the same link where readers can check it. Let’s look at the scatter plot to understand the above-given high-dimensional situation. observed_data_3.plot.scatter(x=\"z\", y=\"y\", c=\"x\", cmap=\"rainbow\", colorbar=False) Output: print(\"Observed ATE: {estimated_effect:.3f} ({standard_error:.3f})\".format(**estimate_uplift(observed_data_3))) print(\"Real ATE:  {estimated_effect:.3f} ({standard_error:.3f})\".format(**run_ab_test(generate_dataset_3))) Output: Let’s check the ATE estimation using OLS and Matching Estimator in the Causal Model. # OLS estimator cm = CausalModel( Y=observed_data_3.y.values, D=observed_data_3.x.values, X=observed_data_3.z.values) cm.est_via_ols() print(cm.estimates) Output: # Matching estimator cm = CausalModel( Y=observed_data_3.y.values, D=observed_data_3.x.values, X=observed_data_3.z.values) cm.est_via_matching() print(cm.estimates) Output: Here we can see matching estimators are showing some improved results which means they are passed in capturing the true effect of covariant but the OLS estimators are not giving that much proper results. We can see this issue regularly if we start looking for covariance in the high dimensional data. Also, we have an option in the Causalnference package to check for the summary. print(cm.summary_stats) Output: Here in the summary, the normalized difference represents the overlapping between the covariate and if the value of it is greater than one which means there is not so much overlapping also we have seen this in the scatter plot. Propensity Score When the covariates are given the propensity is an estimation of the likelihood for a subject that has ended up with the treatment. p^(Z)=P(X|Z) Estimating propensity score can help measure many things in causal inference one of them is the inverse propensity score weight estimator. We can estimate the propensity using the causalinference package. Let’s do this on the observed_data_1. Which we have used before in the examples. cm = CausalModel( Y=observed_data_1.y.values, D=observed_data_1.x.values, X=observed_data_1.z.values) cm.est_propensity_s() propensity = cm.propensity[\"fitted\"] propensity Output: Measuring the causal inference was about Know the value of E[Yi], which can be estimated by E[Yi]=E[(Yi \/ P(X=i|Z)) P(X=i|Z)] Here we have an inverse propensity in the formula so the propensity we measured if we inverse it and weight each point the result will be called the inverse propensity score weight estimator. Inverse propensity score weight estimator: df = observed_data_1 df[\"ips\"] = np.where( df.x == 1, 1 \/ propensity, 1 \/ (1 - propensity)) df[\"ipsw\"] = df.y * df.ips ipse = ( df[df.x == 1][\"ipsw\"].sum() - df[df.x == 0][\"ipsw\"].sum() ) \/ df.shape[0] ipse Output: We have got a good result for our dataset. The inverse propensity score weight estimator depends on the goodness of the estimation of the propensity score. Unconfoundedness and the Propensity Score In the last sections of the article, we have assumed that the potential outcomes Y0 and Y1 are independent of the X and Z. here in this section we are making one more assumption about the potential outcomes that they are also independent of the propensity score. The assumption we have made here will help us in the reduction of the confounding variable’s dimensionality. Using this reduction we can perform several techniques. Like trimming and stratification. Trimming In the observed_data_3 we have seen that we had a problem of imbalance covariants which can be solved by a good overlap or trim and the data we have used having almost zero overlap lets see how we can solve this issue by trimming Older structure of the dataset: observed_data_3 = generate_dataset_3() observed_data_3.plot.scatter(x=\"z\", y=\"y\", c=\"x\", cmap=\"rainbow\", colorbar=False) # actual response curves z = np.linspace(0,1,100) y0 =  np.where(z >= 0.4, -4*(z - 0.4), 0) y1 =  np.where(z < 0.6,  -4*(z - 0.6), 0) + 1 plt.plot(z,y0, \"b\") plt.plot(z,y1, \"r\") Output: Trimming the data based on the propensity score: Data structure after trimming: # mask out data ignored by them propensity = cm.propensity[\"fitted\"] cutoff = cm.cutoff mask = (propensity > cutoff) &  (propensity < 1 - cutoff) # plot the data observed_data_3[mask].plot.scatter(x=\"z\", y=\"y\", c=\"x\", cmap=\"rainbow\", colorbar=False) # actual response curves z = np.linspace(0,1,100) y0 =  np.where(z >= 0.4, -4*(z - 0.4), 0) y1 =  np.where(z < 0.6,  -4*(z - 0.6), 0) + 1 plt.plot(z,y0, \"b\") plt.plot(z,y1, \"r\") Output: Checking for the observed and real ATE filter_ = lambda df: (df.z > 0.2) & (df.z < 0.7) print(\"Observed ATE: {estimated_effect:.3f} ({standard_error:.3f})\".format(**estimate_uplift(observed_data_3[mask]))) print(\"Real ATE:  {estimated_effect:.3f} ({standard_error:.3f})\".format( **run_ab_test(dg.generate_dataset_3, filter_= filter_))) Output: Here we can see that we have got a good result. Basically, by trimming we can measure the causal inference for some part of the covariate space. Stratification We have seen there are various propensity scores when we generate the propensity but we can divide them into groups based on the similarity and stratification or blocking allow us to put the data points into the groups of propensity scores. We can stratify the data points using the package causalInference. These groups can be used for measuring the overall ATE. cm.est_via_blocking() print(cm.estimates) Output: We can see that we have got a good result again. Final Words Here in the article, we have revolved around the estimation of ATE and we have found that various techniques of estimating have their inference and place where we can apply them. The data we have used in the analysis is observational data. Ultimately we can say that if we have good covariate space the matching technique is better because only in ideal data we do have no opposite treatment point in the focus space of data. When such conditions are not there we can use any of the methods or iterate all of them for good results. Causal Inference package in PythonGoogle Colab notebook for codesGoogle Colab notebook for data generatorsCausal inferenceCausal Inference Notes","excerpt":"In data analytics and machine learning, when we apply the behavioural science insights in the studies, it always helps in improving the experience in delivering the results. One of the most important areas of behavioural science is the causal inference which is basically used for extracting cause and intensity of cause. It belongs to the […]","categories":["Deep Tech"],"tags":["AI Tool","causal inference","Data Science","Guide","Machine Learning","Python","Statistics"],"author_name":"Yugesh Verma","publish_date":"2021-10-23T18:08:32","publication_year":"2021","word_count":3286,"keywords":["NumPy","machine learning","TPU","Data Science","Statistics","AI","ML","Machine Learning","RAG","Python","Colab","Aim","analytics","causal inference","AI Tool","Guide","Pandas"],"extracted_tech_keywords":["AI","machine learning","ML","analytics","Aim","Colab","Pandas","NumPy","RAG","TPU"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/a-complete-guide-to-causal-inference-in-python\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":37420,"title":"How A Data Scientist Topped The UPSC Exam This Year","content":"This year’s UPSC topper is a data scientist and the data science community is going gaga over it. When Kanishak Kataria was declared to be the topper, he thanked his parents, sister and girlfriend for the help and moral support, and one thing that came as an instant identifier was that he is currently a data scientist at qplum, Bengaluru, which is an online investment advisory firm, offering AI and machine learning based portfolios. Thanks a lot to all the friends, family and well-wishers. Congratulations to everyone who cleared the examination. Best of luck to those appearing in future :) — Kanishak Kataria (@KanishakKataria) April 8, 2019 According to his LinkedIn profile, Kataria has been involved with developing smart execution algorithm for the company’s offerings. He worked at qplum from April 2016 to May 2017, after which he started preparing for UPSC entrance. A BTech graduate from IIT Bombay, topping the civil services final examinations came as a pleasant surprise for Kataria who qualified the exam with mathematics as his optional subject. Son of an IAS (Indian Administrative Services officer, Kataria had topped the Joint Entrance Examination (JEE) Advanced in SC category in 2010 and had pursued Computer Science engineering from IIT Bombay. “People will expect me to be a good administrator and that’s exactly my intention,” Kataria was quoted as saying. The Data Science Play Before joining qplum as a data scientist, Kataria interned with Microsoft India and has worked in companies such as Two Roads Tech as a software developer and Samsung Electronics as a software engineer. While the earlier role involved working on execution and development of the algorithm, the work at Samsung involved providing useful insights to users based on analysis results, designing and developing methods to analyse data using statistical and machine learning algorithms, developing intelligent services to improve user-smartphone interaction among others. He has been involved with projects such as Learning weights of ARIMA models, AgroBot – Autonomous Irrigation Bot, SocialSea, and has done courses in areas such as foundations of machine learning, functional and logic programming, graph theory, computer architecture,  computer networks and more. As his LinkedIn profile states, he’s actively looking for opportunities to solve challenging real-life problems and has an aptitude of working with numbers with particular interests in the fields of machine learning and data analysis and can be defined as one of the major contributors for his exceptional feat at the nation-wide most competitive exam.","excerpt":"This year’s UPSC topper is a data scientist and the data science community is going gaga over it. When Kanishak Kataria was declared to be the topper, he thanked his parents, sister and girlfriend for the help and moral support, and one thing that came as an instant identifier was that he is currently a […]","categories":["AI News"],"tags":[],"author_name":"Srishti Deoras","publish_date":"2019-04-08T12:35:21","publication_year":"2019","word_count":406,"keywords":["Go","data science","machine learning","programming_languages:R","AI","programming_languages:Go","R"],"extracted_tech_keywords":["AI","machine learning","data science","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/how-a-data-scientist-topped-the-upsc-exam-this-year\/","complexity_score":3,"technical_depth":7,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10103255,"title":"More and More Researchers Resign As OpenAI Drama Unfolds","content":"Is this the end of OpenAI as we know it? On Friday night, three senior researchers at OpenAI resigned in the aftermath of Sam Altman’s termination and the sudden resignation of president Greg Brockman, as reported by The Information, citing sources familiar with the situation. The departing individuals include Jakub Pachocki, the company’s Director of Research; Aleksander Madry, who led a team assessing potential AI risks, and Szymon Sidor, a seven-year researcher at the startup. Pachocki, who began as a research lead in 2017, climbed the ranks over six years to become the Director of Research. In contrast, Aleksander Madry  joined just seven months ago, and served as the Head of Preparedness. Sidor joined OpenAI as a Member of Technical staff in 2016. Associates disclosed that each researcher communicated their resignation. These departures underscore significant discontent among certain employees following Altman’s removal and highlight the challenges currently faced by the company. It is said that a considerable number of top figures at OpenAI are set to depart soon. OpenAI on Friday said that it removed Altman, following a  deliberative review process by the board, which concluded that he was not consistently “candid” in his communications with the board, hindering its ability to exercise its responsibilities. “The board no longer has confidence in his ability to continue leading OpenAI,” read the blog.","excerpt":"It is said that a considerable number of top figures at OpenAI are set to depart soon.","categories":["AI News"],"tags":[],"author_name":"Siddharth Jindal","publish_date":"2023-11-18T13:32:16","publication_year":"2023","word_count":221,"keywords":["Go","OpenAI","AI","programming_languages:R","programming_languages:Go","GAN","R","startup"],"extracted_tech_keywords":["AI","OpenAI","R","Go","GAN","startup","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/more-and-more-researchers-resign-as-openai-drama-unfolds\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10041060,"title":"A Checklist For Companies Before Migrating To Cloud","content":"With digital resilience becoming a key factor for business excellence, more and more enterprises are looking for seamless migration of their legacy systems, servers and databases to the cloud. Many consultants are helping businesses figure out which cloud vendor and tools are right for their needs. According to IDC, by the end of 2021, 80% of enterprises globally will put a mechanism in place to shift to cloud infrastructure and applications twice as fast as before the pandemic. List reasons The first thing companies should ask themselves is if cloud migration is necessary. For that, businesses should jot down the key reasons why they are making the switch. They should find out if the benefits offered by the cloud suits their use cases better than their on-premise setup. Will it help in better communication between remotely working employees? Will it guarantee better security? Will it reduce the cost? The answer to many such questions will help them choose the right cloud provider. Choose the cloud model First, you have to look at which cloud model you will work for: private, public, or hybrid. Private cloud You will require dedicated infrastructure, either on-premises or it can be managed by a third-party provider.  You can deploy applications on dedicated hardware in your private network. You will need a team of experts in the networking and storage domain. Public cloud Public clouds are managed by a third party. It is easy and the company need not create their own software. This brings down the cost but offers lesser data securityAlso, if the company has any specific need, this may not work as public cloud providers have only standard options. Hybrid cloud A hybrid cloud combines your private cloud with a public cloud. This is for businesses who cannot migrate one or a few of the applications to the public cloud.It is scalable as non-critical operations can be moved to public and sensitive data on a private cloud.However, there could be challenges during data transmission between public and private cloud models. Additionally, many hybrid cloud users also feel a lack of control. Cost evaluation Once you decide on the cloud model, decide on the cost of all the resources that you plan to provision. You can use monthly calculators offered by cloud providers to evaluate the pricing of all the services you will be using. And, if cost is the primary reason you are opting for cloud migration, you can check how much savings you can make. Security policies There are many regulations surrounding cloud migration. This is more pertinent to businesses in healthcare and fintech. So, before you start your migration process, ensure you can comply with all government regulations. Also, you can start the process of determining the team members who would be given access to cloud data. This will help you minimise data breach risk. Choose the right cloud provider Once you have decided to go for cloud migration, choosing the right cloud provider is important. Amazon Web Services (AWS), Microsoft Azure, Google Cloud Platform, IBM Cloud, Salesforce and SAP are popular cloud providers globally. The cloud provider should align with your business goals, and to know that, you have to collect the following info: Experience Total number of customersCustomer redressalBrands using itQuality of data centresSize and experience of the engineering teamData safetyCapacity to manage updates and patches Train your team Bringing your team, especially the team that will be directly involved in the migration, together and training them is important to know if you need to hire any cloud migration specialist or not. You will be able to determine if your existing team can make the process seamless or would need an extra hand. Disaster recovery Having a disaster recovery plan is an extremely important but often ignored part of cloud migration. Choosing the right disaster recovery region is important too. You have to look at the cost and distance from the current region. You can choose either backup and restore or pilot light strategy or warm standby or multi-site strategy in case of automatic failover to a different region for your internet-facing application.","excerpt":"With digital resilience becoming a key factor for business excellence, more and more enterprises are looking for seamless migration of their legacy systems, servers and databases to the cloud. Many consultants are helping businesses figure out which cloud vendor and tools are right for their needs. According to IDC, by the end of 2021, 80% […]","categories":["IT Services"],"tags":["AWS","cloud migration","cloud providers"],"author_name":"Shanthi S","publish_date":"2021-06-01T10:00:00","publication_year":"2021","word_count":684,"keywords":["Go","cloud providers","cloud migration","AWS","AI","R","ML","cloud_platforms:AWS","Scala","Git","RAG","Azure"],"extracted_tech_keywords":["AI","ML","RAG","AWS","Azure","R","Go","Scala","Git","cloud_platforms:AWS"],"url":"https:\/\/analyticsindiamag.com\/it-services\/a-checklist-for-companies-before-migrating-to-cloud\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10069689,"title":"How VideoVerse plans to take on Apple iMovie, Adobe Premiere Pro","content":"Ask any video editors or content creators – video editing is probably one of the most complex processes – it requires a lot of practice, technological know-how, and creative skills. It also comes with a fair bit of technical setbacks\/hurdles that commonly occur while editing videos using various video editing software and tools that are resource-intensive and involve a lot of computing. Most video editing software today, including Apple iMovie, Adobe Premiere Pro, etc., ideally requires expensive graphic cards or the latest systems like Mac Studio, iMac, and Microsoft Surface Studio to perform extensive tasks. Otherwise, you will most likely deal with computer crashing, getting too slow during editing, video file crashing while editing, and so on. What’s more? Once the video is rendered\/generated for preview or quality check, sharing these large-size files with the other team members can get exhausting, resulting in many more issues like delay in production, videos with out-of-sync audio, corrupted video files, etc. In other words, team collaboration can get messy when editing videos. “How do you make the system-heavy tasks easier on the web? That is where we come in,” said Vinayak Shrivastav, co-founder and CEO at VideoVerse, saying that their tools and products are browser-friendly. As a result, video creators can go online and edit most of the functionalities effortlessly. Based in Mumbai, India, and with offices in the UK, Europe and Israel, VideoVerse (formerly Touch.ai) is revolutionising expensive and time-consuming legacy processes. This video AI company is taking video editing to the cloud and web-based platforms, simplifying the whole experience of video editing for everyone. The inception “The idea actually transformed over the last two years; we built a very simple technology that could do editing for enterprise solutions. In the last 18 months, we have expanded our horizon beyond editing for enterprises and built different applications for various case studies,” shared Vinayak Shrivastav, co-founder and CEO at VideoVerse. VideoVerse was co-founded by Shrivastav, Saket Dandotia, and Alok Patil in 2016. The company is backed by Stride Ventures, Innoven Capital, Pacific Western Bank, A91 Partners, Alpha Wave Global, and the former co-founder of Flipkart, Binny Bansal. The company currently offers three flagship products. This includes its core product Magnifi, alongside Styck and Illusto. Magnifi provides real-time video highlight technology for enterprises, particularly sports, media and entertainment, news, and others. Styck platform enables creators to live stream simultaneously across multiple social media platforms. Illusto is a web-based editor that works as an intuitive tool to help users create videos on the go and share them across various social media platforms like TikTok, Instagram, and Snapchat, among others. Here’s how Shrivastav said that analysing live sports comes with its own set of challenges, especially in the accuracy and recall rate of the sport detection model. Moreover, in a live feed, you never have complete control of the situation, and the deployment of any feature has to be done assuming there will be certain unexpected events and situations. Citing cricket, he said they have a deep learning model to detect bowler action, which works well with the validation data; however, an unexpected change in the camera recording angle or the bowler not being in focus can lead to a recognition anomaly. Further, he said similarly, in football, the most challenging part is the camera motion or the change in position of the ball with respect to the ground; therefore, a ball being kicked out of camera focus can lead to lower detection accuracy. The solution here would involve taking these outliers and training the machine to detect such moments and be able to address them. “We have created an application where the sample data from the live events is collected, our data science team verifies this data, and we then use it to train our model and improve on speed and accuracy. Algorithms to verify the event with audio and OCR analysis are also added,” explained Shrivastav. To further ensure speed of deployment, especially since the analysis has to be completed and the output generated with negligible latency, often with a frame rate of 30-50 fps, VideoVerse is using high-performing GPU machines deployed on cloud platforms like AWS, GCP, Azure, etc. Some of the product features they have recently deployed at Magnifi include an automatic title generator for events and highlights, an “Auto-flip” aspect ratio to convert any video into a mobile-optimised format and auto-production of mobile-friendly Web Stories for sports. Tech stack at VideoVerse VideoVerse and its three distinct cloud-agnostic products use artificial intelligence (AI) and machine learning (ML) technology to revolutionise how content is refined and consumed. As far as specific stacks go, for Magnifi, the key technologies used are face and image recognition, vision models, optical character recognition, audio detection and NLP. Styck and Illusto both use full-stack applications (MERN [Mongo, Express, React, Node]). Edit videos seamlessly One of VideoVerse’s products, Styck, is very similar to Restream.io and StreamYard. “None of them has actually used it for solving a very simple problem — i.e. replacing OBS,” said Shrivastav. For those unaware, OBS is a free and open-source programme for recording and live streaming. It does not have a built-in editing tool. In other words, if you are using platforms built on OBS, you will need a video editor. Shrivastav said OBS is very system heavy; you need a good CPU to run it properly. “So, we realised that what we built at the back end has a very good use case that can actually create a new product out of it,” he said, sharing the origin of Styck, which the co-founder believes would one day replace OBS. Today, VideoVerse has developed simple tools for users to play around with, where they can create content simply with the click of a button. “We are trying to create a universe around video editing as an ecosystem,” said Shrivastav. “Today, we might have three products, but we will be developing a suite of products in the coming months.” Competition VideoVerse is not alone in this segment. Currently, it competes with a whole gamut of companies offering cutting-edge technology solutions for video editing. One of the companies includes Descript, a platform for creative tools, making audio and video creation as fast, accessible, and collaborative as Google Docs. “We are building something similar within Illusto,” said Shrivastav, saying that the user can go through transcription and edit the content rather than trying to go through the timescale. Last year, Canva also got into video editing and launched Video Suite, an end-to-end video creation product that lets everyone design and publish professional-quality videos. Interestingly, VideoVerse is also developing a similar collaborative component to its Styck and Magnifi platforms. Other players include Clipchamp, VEED.io, Visme, Adobe Spark Video, Biteable, Vimeo Create, and InVideo, among others. But, the question is, how is VideoVerse different from other players in the market? “Given how this segment is growing, I understand that to be indispensable; it is imperative to be different. Hence we are constantly learning from those in a similar domain,” said Shrivastav, saying that the AI and video editing technology is still evolving. First mover advantage The video editing software market is expected to reach USD 3.04 billion by 2027, growing at a CAGR of 5.9 per cent from 2020 to 2027. The market is said to be driven by the media and broadcast industry. “There are a lot of interesting tools that are coming up in the market, and that kind of tells you that the market in itself has been growing aggressively. That is why companies are trying to capture 5 per cent of the pie over the next two years,” said Shrivastav. He said, in India, they do not have a direct competitor. For instance, one of the leading players in the video editing landscape, Adobe, achieved record revenue of USD 4.39 billion in its second quarter of the fiscal year 2022 – a 14 per cent YoY growth or 15 per cent in constant currency. Shrivastav also pointed out the rise of content creators in the country. With a market size of nearly 50 million content creators, the creator’s economy is anticipated to be USD 104.2 billion in 2022. “The created content is equal to the amount of content we have consumed in the last ten years,” he added. He said that it is a growing market; like every month, it is going to change based on the way your data consumption improves, and with 5G around the corner, the data consumption is going to quadruple from where it is because you can load data faster, inevitably consuming more content. When that happens, VideoVerse looks to be the go-to option for many content creators and video editors to create videos seamlessly, besides its enterprise play. What’s next? “Our thought process has been very simple – we want to create a simple and efficient tool,” said Shrivastav. He also said the company is looking to expand to multiple locations in the coming months. VideoVerse has about 140+ employees and looks to add 250-300 people by the end of this year. The team said its entire back end and digital marketing initiatives have been set up in India. The leadership team, including the three co-founders, are operating out of India. “We recently started setting up a product team, where the majority are from India,” said Shrivastav. Without revealing the names, he said that the product team head recently joined the team, a Harvard graduate, and said they have also set up their data science team in Israel, led by an MIT alumnus. In addition, they have already started setting up a sales team in Europe and the US. “In the next couple of months, we are going to start discussing and look at adding interesting engineering talents coming out of Eastern Europe,” said Shrivastav. Pricing VideoVerse said that it has built its platform on a freemium model as they want people to come and use its platform first before they start looking at paying. In addition, the team said that they have built-in a monthly subscription model, which lets users access premium-grade features. One of the interesting things about VideoVerse is that it does not depend on your editing, consumption, or streaming, as the platform works on a fixed pricing model.","excerpt":"VideoVerse has about 140+ employees and looks to add 250-300 people by the end of this year.","categories":["IT Services"],"tags":["AI Tool","Binny Bansal"],"author_name":"Amit Naik","publish_date":"2022-06-24T16:00:00","publication_year":"2022","word_count":1712,"keywords":["data science","machine learning","artificial intelligence","AWS","AI","Binny Bansal","ML","image recognition","NLP","deep learning","AI Tool","Azure"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","NLP","data science","image recognition","AWS","Azure"],"url":"https:\/\/analyticsindiamag.com\/it-services\/how-videoverse-plans-to-take-on-apple-imovie-adobe-premiere-pro\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10090066,"title":"When Redis Helped ChatGPT Keep the Conversation Going","content":"A few days ago, a bug in an open-source library caused ChatGPT to be taken off the grid. This issue allowed a small percentage of users to view the titles of other users’ conversation history, as well as the payment-related details of ChatGPT Plus subscribers. But, Redis was here to save OpenAI’s day. “A bug report is one way to know someone uses your open-source software,” said Yiftach Shoolman, co-founder and CTO at Redis. The Redis team came to know that it is in fact their software which OpenAI is using to scale ChatGPT, when OpenAI reached out to them to find a fix for the bug, which their team was happy to do. The bug and the fix On Monday, March 20 at 1 am PST, a bug was introduced into the server at OpenAI that caused a surge in Redis request cancellations. The bug exposed information related to active users’ first and last name, email address, payment address, the last four digits (only) of a credit card number, and credit card expiration date. OpenAI uses Redis to cache user information in their server, and this bug resulted in a small likelihood of each connection returning corrupted data. The Redis Cluster is used to distribute the load across multiple Redis instances, and OpenAI interfaces with Redis through the redis-py library. The interface with Redis is from OpenAI’s Python server, which runs with Asyncio. When using Asyncio, requests and responses with redis-py behave as two queues. If a request is cancelled after being pushed onto the incoming queue but before the response is popped from the outgoing queue, the connection becomes corrupted. In most cases, this results in an unrecoverable server error, but in some cases, the corrupted data matches the data type that the requester was expecting, leading to the possibility of bad data being returned. This bug only affected the Asyncio redis-py client for Redis Cluster, and the OpenAI team has since resolved the issue. Here is a close representation of the role Redis is playing in ChatGPT. Created by Shahrukh Khan and Navdeeppal Singh, this is a user flow of ChatGPT-memory, an extension for ChatGPT API. After discovering the bug, OpenAI took several measures to enhance their system’s security. They thoroughly tested the fix, implemented redundant checks, and examined their logs to ensure messages are only accessible to the intended user. They also notified affected users and improved their logging system to detect and resolve similar incidents. OpenAI further improved the scalability and robustness of their Redis cluster to minimise connection errors during heavy traffic. OpenAI was in high praise of Redis after the issue was resolved. They said, “Redis, along with other open-source software, plays a crucial role in our research efforts. Their significance cannot be understated—we would not have been able to scale ChatGPT without Redis.” Speaking to AIM earlier, Shoolman had said that the draw in the open-source community was so high, mainly for three reasons: speed, versatility, and simplicity. Akin to a Swiss knife, Redis was able to cater to multiple use cases, while also keeping the documentation simple and intuitive for developers to learn and adopt. A one-way effort Initially founded as a non-profit AGI research organisation for developing open-source AI applications and algorithms, OpenAI quickly changed its course. While some attribute this shift to the high cost of AI research, and investors’ preference for profitable startups, OpenAI’s co-founder has a different perspective. In a conversation with The Verge, Ilya Sutskever said, “We were wrong. Flat out, we were wrong.” According to the chief scientist of the company, AI or AGI is likely to become incredibly potent at some point, and therefore, open-sourcing it may not be the best approach. But, it is not merely the safety aspect. The recent 98-page GPT-4 paper also mentions “competitive landscape” to be equally a factor in OpenAI proudly declaring that the report will contain no details about the architecture (including model size), hardware, training compute, training datasets and method, etc. The company is using open-source extensively for the creation and maintenance of its products, but has completely closed itself to the world. The one-way effort has been in contest for a long time. Are companies obliged to give back to the open-source community in exchange for using codes from the community? The question is not meant to target companies like OpenAI, but is a serious reflection on the future of open source. “These are not just interesting times to live, they are interesting, but potentially disastrous times. Very few people argue against the wider benefits of open source software, but its overall status and long-term future are not guaranteed unless more of us all get involved at one level or another,” wrote Adrian Bridgwater, a freelance technology journalist, for Forbes.","excerpt":"Redis’ open-source software is silently helping OpenAI scale ChatGPT","categories":["AI Highlights"],"tags":["ChatGPT","open-source software","Redis"],"author_name":"Ayush Jain","publish_date":"2023-03-27T14:30:00","publication_year":"2023","word_count":795,"keywords":["Go","ChatGPT","OpenAI","AI","open-source software","Scala","Git","Python","Aim","R","Redis"],"extracted_tech_keywords":["AI","ChatGPT","OpenAI","Aim","Redis","Python","R","Go","Scala","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/when-redis-helped-chatgpt-keep-the-conversation-going\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":40522,"title":"Apple Harnesses AI To Empower Accessibility Features In iOS","content":"In keeping with their trend of helping specially-abled persons harness technology, Apple Inc last week announced its plans to make devices accessible to more than one billion people living with some kind of disability. “We build powerful assistive technologies into Mac to give people with physical limitations more ways to be creative and productive,” said the company in a statement. For the visually impaired, VoiceOver screen reader describes exactly what’s happening on your screen and lets you navigate using gestures or a braille display. Zoom gives you a powerful built-in magnifier. You can also make adjustments to cursor size, motion and contrast to meet your specific vision needs. For the hearing impaired, “Type to Siri” lets you make commands and ask questions with your keyboard. FaceTime is a great way to communicate through sign language. And instead of audible alerts, your Mac can flash its screen whenever an app needs your attention. For persons with physical limitations, Switch Control lets you navigate what’s on your screen with a single tap, and it works with switches, joysticks, head-tracking hardware and other adaptive devices. Features like Sticky Keys, Slow Keys and Mouse Keys let you adjust interaction sensitivities. And the customisable onscreen keyboard has full Dwell support. For learning better, Apple’s Text to Speech adds auditory reinforcement to books, websites or homework assignments. Word Completion helps boost vocabulary and word-building skills. And Dark Mode makes it easier to focus on the task or project at hand. A news wire reported that India has one of the most vibrant and exciting iOS developer communities who have already created nearly 100,000 apps for the App Store worldwide, who have accessibility on top of their minds. Over 700,000 app economy jobs can be attributed to iOS ecosystem in India alone.","excerpt":"In keeping with their trend of helping specially-abled persons harness technology, Apple Inc last week announced its plans to make devices accessible to more than one billion people living with some kind of disability. “We build powerful assistive technologies into Mac to give people with physical limitations more ways to be creative and productive,” said […]","categories":["AI News"],"tags":["Apple","what is power bi"],"author_name":"Prajakta Hebbar","publish_date":"2019-06-10T11:26:07","publication_year":"2019","word_count":295,"keywords":["what is power bi","programming_languages:R","AI","Apple","ViT","R"],"extracted_tech_keywords":["AI","R","ViT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/apple-harnesses-ai-to-empower-accessibility-features-in-ios\/","complexity_score":2,"technical_depth":4,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10022197,"title":"Indian AI-based Mental Health App Wysa Receives Funding From Google Assistant","content":"Wysa, an Indian-based AI-enabled mental health application, has announced that the startup has secured an undisclosed amount of investment from the Google Assistant Fund. Launched in 2018, the Google Assistant Investments program was designed to assist early-stage startups in advancing their digital assistant ecosystem and innovating new ideas. Wysa has also recently topped the list of best apps in India for 2020. Wysa was also among the first startups to be included in the Google For Startups launchpad program in India in 2018. The startup has already raised funding of $3.9 million from investors like Kae Capital and pi Ventures. Talking about this new funding, Moez Kaderali, program manager for the Google Assistant Investments Program, stated in its official release that the post-COVID era had urged people to take care of their mental health, sleep patterns and anxiety on a daily basis. With Wysa’s explainable AI – combining natural language understanding with clinical assurance, users will be able to manage their mental health in this new normal environment. Jo Aggarwal and Ramakant Vempati founded the application in 2015 to give early intervention to high-risk groups. It has been equipped with an AI-based, emotionally intelligent chatbot, along with evidence-based resources for self-help tools and messaging-based support from human therapists. While the app doesn’t address severe mental illness, it assists users in undergoing early-stage supportive therapy. Currently, Wysa is equipped with Indian therapists who address around 3 million users from 65 countries, with approximately 80-85% of their users are from abroad. Some of the organisations that have tied up with Wysa includes — UK’s National Health Service, Ohio-based Cincinnati Children’s Hospital, and other insurance companies. The app has also been launched nationally by the Government of Singapore. Addressing this, Jo Aggarwal, co-founder and chief executive of Wysa, stated to the media — the app aimed to address mental health issues at a scale that is “low cost and clinically assured.” The company is planning to expand beyond literacy and language barriers using voice features equipped with local languages. And that is why “Google Assistant Fund is our ideal partner,” concluded Agarwal.","excerpt":"Wysa, an Indian-based AI-enabled mental health application, has announced that the startup has secured an undisclosed amount of investment from the Google Assistant Fund.  Launched in 2018, the Google Assistant Investments program was designed to assist early-stage startups in advancing their digital assistant ecosystem and innovating new ideas. Wysa has also recently topped the list […]","categories":["AI News"],"tags":["chatbot mental health","Google Assistant","Mental Health"],"author_name":"Sejuti Das","publish_date":"2021-03-15T11:27:12","publication_year":"2021","word_count":349,"keywords":["Go","API","funding","AI","chatbot mental health","Git","Aim","explainable AI","Mental Health","GAN","Google Assistant","R","startup"],"extracted_tech_keywords":["AI","Aim","R","Go","Git","API","GAN","explainable AI","startup","funding"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/indian-ai-based-mental-health-app-wysa-receives-funding-from-google-assistant\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":43407,"title":"&#8216;Angry Birds&#8217; Creator Wants To Set Up AI Entrepreneurship University In India","content":"Noted developer Peter Vesterbacka, who shot to fame after his success with the wildly popular game Angry Birds, now wants to set up an educational institute to groom entrepreneurs who want to work with artificial intelligence. Finland-based Vesterbacka, who is also the founder of Slush, the gaming company, met Mekapati Goutham Reddy, the Minister for IT and Industries, to propose a plan to set up the Finland AIEU in Andhra Pradesh. Reportedly, this project would be done in association with Malaxmi Group. Great to have an opportunity to address so many young people here in Vijaywada, Andhra Pradesh, India today. #funlearning from Finland going places pic.twitter.com\/qLZud1mF8Y — Peter Vesterbacka (@pvesterbacka) July 24, 2019 Reportedly, the university would focus its curriculum in technologies like: Artificial intelligence Robotics Design thinking Game development They would also focus on areas like: Domain knowledge Multi-disciplinary learnings Professional and personal development Entrepreneurial skills According to media reports, Vesterbacka also discussed the opportunity of employment generation and the need to train the Indian youth in emerging tech. He said that the proposed university would be the perfect place nurture these talents. “The Finland AI Entrepreneurship University would leverage the Finnish expertise in education to tap academic and innovation potential in the country,” a senior executive of Malaxmi group told a noted daily. Andhra Pradesh currently shares its de jure capital Hyderabad with the neighbouring state Telangana. Former Andhra Pradesh Chief Minister N Chandrababu Naidu has been helming the challenging task of developing a highly sophisticated capital for the state as the shared capital is expected to belong entirely to Telangana within the next 10 years. Andhra Pradesh was ranked as the top-most state in India when it came to the ease of doing business. Always great to have the opportunity to address young people. Talked about walking on water and other skills of Finnish people. Also shared the secret, fantastic education for all, not just a few. And cold winters ;) Welcome to the #FinEstBayArea at the #heartofeurasia pic.twitter.com\/cZzDl0bSuR — Peter Vesterbacka (@pvesterbacka) July 24, 2019","excerpt":"Noted developer Peter Vesterbacka, who shot to fame after his success with the wildly popular game Angry Birds, now wants to set up an educational institute to groom entrepreneurs who want to work with artificial intelligence. Finland-based Vesterbacka, who is also the founder of Slush, the gaming company, met Mekapati Goutham Reddy, the Minister for […]","categories":["AI News"],"tags":["AI education"],"author_name":"Prajakta Hebbar","publish_date":"2019-07-26T13:28:27","publication_year":"2019","word_count":339,"keywords":["Go","API","artificial intelligence","AI education","programming_languages:R","AI","innovation","programming_languages:Go","RAG","GAN","R"],"extracted_tech_keywords":["AI","artificial intelligence","RAG","R","Go","API","GAN","innovation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/angry-birds-creator-wants-to-set-up-ai-entrepreneurship-university-in-india\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":18648,"title":"Visual commerce startup Artifacia is helping brands monetize UGC","content":"Co-founders Navneet Sharma and Vivek Gandhi If there ever was an AI startup that cracked the formula of survival, it is Bangalore-based Artifacia – known for its cutting-edge Visual Commerce platform. Kickstarted in 2015 as Snapshopr by Navneet Sharma and Vivek Gandhi, the company rebranded and pivoted to Artifacia early this year to gain more traction in the field of visual commerce. Artifacia is also one of the earliest startups in Deep Learning not just in India but globally as it was incepted around the time when Deep Learning started to gain prominence across the globe in 2013. Rebranding from Snapshopr to Artifacia The reason for pivoting is simple, Sharma reveals. “We didn’t want to focus on visual search as our core product anymore and Snapshopr basically meant snapping a photo and then shopping. It just meant visual search and it was very restrictive for us. With our pivot, we were planning for a global rollout of our solutions for brands and the previous name became very restrictive. So we thought it is better to rebrand now than later and we found Artifacia to be a pretty cool new name,” he said. Another major reason was that the company wasn’t able to build enough traction with the Visual Search API. “Our traction started building when we pivoted to Artifacia and we became a SaaS company ,” Sharma added. Today, they aren’t into the business of selling Visual Search APIs anymore and the company switched to a SaaS model in the MarTech space. To their credit, the startup has a great mix of customers, ranging from SMBs to mid-sized companies like Forest Essentials and GoodEarth to the biggies such as Titan on their roster. Helping brands monetize UGC Today, Artifacia’s core product is bridging the gap in the e-commerce experience by helping brands monetize their User Generated Content (UGC), drive better engagement and deliver improved customer experience to their users . The reason behind focusing on UGC — featuring authentic content at the point of sale helps drive purchase decisions and increase conversion. “Generally consumers return the product they buy from e-commerce sites when they see something and they get something else. Today, many of the brands across the globe are now realizing that authenticity in the marketing and e-commerce content delivers much better ROI,” said Sharma. The benefits of monetizing UGC content from Instagram are manifold – it aids product discovery. Research indicates that users are twice as likely to share UGC, and an Adweek finding indicates that visual UGC is more influential than brand-generated photos or videos for both engagement and conversions at important customer touchpoints. In a way, UGC is a social proof that a brand celebrates its customers and want to inspire new users to become part of the brand story. Artifacia plugs the missing piece in the puzzle by helping brands easily activate such photos and videos in the form of shoppable galleries using its AI-based product tagging system. The startup has built visual recognition technology that helps marketers create shoppable content with ease. Essentially, the core product helps marketers add product links to Instagram photos and videos, by learning through cutting-edge algorithms what kinds of objects\/visual representations are in the photos\/videos. Marketers can acquire, activate and publish content through the dashboard and monitor the performance of the visual content at per post level. “At the end of the day, marketers ultimately care about a great user experience that also gives good ROI for their brands. I think using AI helps them achieve their objectives in a much quicker way.,” he said. Sharma, a VIT University alumnus emphasizes the entire algorithmic stack is built using Deep Learning. “Whatever tech we have built as part of Snapshopr and then Artifacia is now being utilized to improve the product experience for our customers,” Sharma adds. Over the coming months, the startup plans to add support for other social media channels like Twitter and Pinterest as well. Building the talent pipeline Backed by a small team of seven, Artifacia’s research arm remains quite active in several Deep Learning projects across computer vision and NLP. Sharma is also credited for starting the one of the first AI research groups in India while in VIT, and together with co-founder Gandhi, has so far mentored more than two dozen people in AI. Many of their mentees are now researching in AI at top universities like Brown University, CMU and University of Montreal in the USA and Canada, where much of the core AI research is happening. Navneet has been delving in AI since 2011 and emphasizes that the founders at Artifacia have deep expertise in DL which has been instrumental to their differentiation for far. When asked about hiring preferences  he reveals that they generally prefer people with 2-3 years experience in ML since finding and training good talent in DL requires a lot of resources. Even though there is a lot of hype around AI\/ML these days, it is difficult to find good talent for multiple reasons. “It is an upcoming field and people who are enthusiastic about it are the ones who are self-trained. In fact, every summer\/winter we get a good set of interns and during that time we do a few more projects relevant to our product and business objectives. But after the internship, they apply for top universities because everyone thinks that the more exciting stuff is happening elsewhere and often it is hard to convince them otherwise,” he revealed. Plans for expansion are afoot with the startup hiring aggressively across all functions. Between the two co-founders, Sharma looks primarily after Product and Growth while Gandhi takes care of Engineering at Artifacia. When investors become go-to guys The AI-based visual commerce platform has attracted top technologists and entrepreneurs from India such as PhonePe’s Rahul Chari, former Flipkart CTO Amod Malviya among others. There is also FusionChart’s Pallav Nadhani, the young turk who built his company without any external funding and Sharma’s go-to guy. Nadhani, as Sharma describes fondly is not only an investor but also serves as a mentor. “It is hard for a founder to have an investor who is also a mentor and it helps to have a young energetic investor on your side who brings a lot of value to the company. And all our investors are also part of the team,” he shared. And so far, the company’s strategy has been focusing on solving the problem at hand through one single product. “In the initial days, the more customers you bring in, the more services you are required to build. That is not scalable and what we are trying to do is not just work with a few ecommerce companies in India but work with hundreds of small to large brands across the globe. That’s the only way one can scale,” he said, in closing.","excerpt":"If there ever was an AI startup that cracked the formula of survival, it is Bangalore-based Artifacia – known for its cutting-edge Visual Commerce platform. Kickstarted in 2015 as Snapshopr by Navneet Sharma and Vivek Gandhi, the company rebranded and pivoted to Artifacia early this year to gain more traction in the field of visual […]","categories":["AI Startups"],"tags":["image recognition"],"author_name":"Richa Bhatia","publish_date":"2017-10-31T10:57:56","publication_year":"2017","word_count":1142,"keywords":["Go","API","AI","ML","image recognition","Scala","computer vision","NLP","deep learning","GAN","R"],"extracted_tech_keywords":["AI","ML","deep learning","NLP","computer vision","R","Go","Scala","API","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-startups\/visual-commerce-startup-artifacia-helping-brands-monetize-ugc\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":28227,"title":"The Number Game Behind Advanced Activation Functions In Machine Learning","content":"With artificial intelligence being implemented in almost every sector, it is important to understand the maths behind how it functions so accurately. Activation functions are at the end of every hidden layer of a neural network and it plays a key part in updation of weights. The main function is to introduce a non-linearity in the model, calculate, and decide what has to be sent to the output and what needs to be discarded. In this article we will try to understand the maths behind these functions and test out some examples in Python. The graphs are plotted with the help of Matplotlib library in Python. These other activation functions are in development on neural networks in the current industry. Softmax Function Softmax function is an activation which returns the probability of the particular sample with respect to the other samples. This activation function is mostly used in classification problems to find the class with the highest probability. It squashes the values between 0 and 1. The larger negative values usually tend to zero and larger positive values are inclined towards 1. Let us take a look at one of the examples of this function: Softsign Function The Softsign function is an activation function which rescales the values between -1 and 1 by applying a threshold just like a sigmoid function. The advantage, that is, the value of a softsign is zero-centered which helps the next neuron during propagating. Here’s an example of a softsign function with values in between -20.0 and 20.0. Softplus Function This function is similar to softmax but it finds the natural logarithm. This acts like a threshold function which also decides whether to omit or let the output pass onto the other layers. Here’s an example of softplus function. This function also squashes the negative values towards zero. Hence the function outputs a set of values which range between 0 and infinity. Exponential Linear Unit (ELU) An ELU is similar to a ReLU for non-negative inputs. The ELU function drops smoothly for values below 0 where as ReLU drops faster. This function is also used for faster computations comparatively. Below is the equation for ELU: The alpha parameter in the ELU is a positive value. This can be set accordingly while manipulating the parameter of the neural network. Let us look at one of the examples of the ELU function: Parameter Rectified Linear Unit (PReLU) This activation function is similar to ReLU, but contains tunable parameters like alpha, which can be configured while adjusting the parameters of the model. The values above 0 are unchanged but values  below 0 are reduced by an amount ‘alpha’ to allow some negative values to be considered during training. This is done to help the loss converge at the minima smoothly. Below is an equation of Parametric Rectified Linear Unit, With the alpha value equal to 0.05 we shall plot a graph to visualise how the PReLU is computed. Binary Step Function This function is one of the basic ones, but its not widely used in developing neural networks. To understand what Binary Step is, let us consider a simple equation. This function is as simple as it looks. It converts all the values to either 0 or 1 depending the the sign. If the value is less than 0 it is assigned a value 0 but if it is greater than or equal to zero it is assigned a value of 1. Let’s look at an example of a binary step function. Conclusion These advanced activation functions are currently implemented in neural network development. One cannot find the perfect activation function which fits every model and produces accurate results. It completely depends on the model and how the data is behaving with respect to the model. Research is still going on in this domain to find better activation function for building robust models.","excerpt":"With artificial intelligence being implemented in almost every sector, it is important to understand the maths behind how it functions so accurately. Activation functions are at the end of every hidden layer of a neural network and it plays a key part in updation of weights. The main function is to introduce a non-linearity in […]","categories":["AI Features"],"tags":["AI (Artificial Intelligence)","Machine Learning","Neural Network","Tensorflow"],"author_name":"Kishan Maladkar","publish_date":"2018-09-12T05:33:29","publication_year":"2018","word_count":645,"keywords":["Neural Network","Go","artificial intelligence","TPU","programming_languages:R","AI","neural network","Machine Learning","Python","programming_languages:Python","Matplotlib","R","AI (Artificial Intelligence)","Tensorflow"],"extracted_tech_keywords":["AI","artificial intelligence","neural network","Matplotlib","TPU","Python","R","Go","programming_languages:Python","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/the-number-game-behind-advanced-activation-functions-in-machine-learning\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10048918,"title":"Oxford &#038; Oracle Cloud System Collaborates To Identify Covid-19 Variants Faster","content":"The University of Oxford and Oracle’s Global Pathogen Analysis System (GPAS) is now being used by organisations to identify the fast spread of the highly infectious Delta variant and other COVID-19 mutations. Built using Oxford’s Scalable Pathogen Pipeline Platform (SP3), Oracle APEX, and Oracle Cloud Infrastructure (OCI), the Global Pathogen Analysis System is a cloud platform that provides a unified, standardised system for analysing and comparing the annotated genomic sequence data of SARS-CoV-2. Researchers are using the system to upload pathogen data and receive comprehensive results within minutes. With user permission, the results may be shared with participating laboratories around the globe in a secure environment. Making this data comprehensible and shareable will help public health authorities evaluate and plan their response by giving them invaluable insight into emerging variants even before they are officially designated as Variants of Concern. “GPAS is the first industry standards-based service anywhere in the world, offering a standardised sequence data analysis service for users on the cloud,” said Derrick Crook, professor of microbiology in the University of Oxford’s Nuffield Department of Medicine. Oracle Chairman and CTO Larry Ellison said, “With GPAS, we are bringing the power and security of the cloud to enable any researcher, in any location, to become part of the solution. The more data medical institutions, governments, and academics provide, the more quickly we can understand and act to get ahead of the coronavirus”. “This platform promises to bring together data much more rapidly, helping us to better understand and get ahead of the patterns of spread faster, so governments can make better policy decisions and mitigate the devastating impact this virus continues to inflict in their own countries and across the globe,” said Tony Blair, Executive Chairman of TBI and former UK Prime Minister. As part of their work with the Global Health Security Consortium (GHSC), the Lawrence J. Ellison Institute for Transformative Medicine (Ellison Institute) and the Tony Blair Institute (TBI) for Global Change have worked in coordination with Oxford and Oracle to support the development of the platform and to get it in the hands of global researchers. Using the platform, researchers and governments will be able to quickly access the timely, relevant data they need to make up-to-date scientific analyses and better-informed policy and safety decisions regarding new variants. To know more about the initiative or to get involved, please visit www.gpas.cloud. Institutions using the platform include the University of Montreal Hospital Centre Research Centre, the Institute of Public Health Research of Chile, the Oxford University Clinical Research Unit in Vietnam, the Institute of Clinical Pathology and Medical Research – New South Wales Pathology, and Oxford Nanopore Technologies. GPAS is also now part of the Public Health England New Variant Assessment Platform.","excerpt":"Built using Oxford’s Scalable Pathogen Pipeline Platform (SP3), Oracle APEX, and Oracle Cloud Infrastructure (OCI), the Global Pathogen Analysis System is a cloud platform.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","covid-19","Deep Learning","healthcare analytics","Machine Learning","Oracle","Oracle Cloud Infrastructure"],"author_name":"Victor Dey","publish_date":"2021-09-20T15:55:41","publication_year":"2021","word_count":455,"keywords":["Go","API","ELT","covid-19","AI","programming_languages:R","Machine Learning","Scala","Oracle","healthcare analytics","programming_languages:Go","programming_languages:Scala","GAN","Deep Learning","R","Oracle Cloud Infrastructure","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","R","Go","Scala","API","ELT","GAN","programming_languages:R","programming_languages:Scala","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/oxford-oracle-cloud-system-collaborates-to-identify-covid-19-variants-faster\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10053548,"title":"What are the Ways to Automate Model Drift?","content":"In machine learning, the performances of machine learning models are fixed most of the time when the environment and data which the model is consuming are fixed. But if the environment and data get changed, the performance of the model also changes. In the scenario where these variables are not stable, we are required to make a model which can deal with the changes. Changes in the performance of the model can be defined as the model drift. In this article, we will be discussing model drifting and how the drift can be automated to achieve a complete availability of the model. The major points to be covered in this article are listed below. Table of Contents Challenges in Machine learning What is Model Drift?Causes of Model driftsHow to Detect Model Drift?Ways to Automating Model Drift Challenges in Machine learning? As we know that machine learning models are different from other traditional models when it comes to dealing with the data and performance of models. Since most of the machine learning models work on the basis of knowledge gained by the old transaction or formally we can say the performance of the model depends on the data from which they have learned and the data which is going to be the part of the model as the input. As the input varies the performance of the model also varies. By above, we can say that after deployment of the machine learning models there is always a requirement to monitor the machine learning models in real-time so that the performance of the models can be maintained to a certain range. We can monitor the performance of machine learning on the basis of metrics like precision, AUC, recall, etc. however these metrics require labels for the predictions provided by the models in real-time. We normally see that the labels are presented in the training information but there can be a possibility that in production the labels in the data are not available. In the absence of labelled inputs, these performance metrics can be used to measure the changes in the performance of the model. By making some visualizations they can be helpful as the indicator of the performance issue. What is Model Drift? We can define model drift as the change in the performance of the model due to changes in the information or data or due to changes in the relationship between the input and output variables. We can say that if a model is producing predictive results that change compared to expected results according to the parameters set while training the model. When it comes to the production machine learning models, we can say the model drift is changing between the data on which the model is trained and the data which the model is producing in real-time which causes the changes in the prediction level. The reasons for changes in the production data can be changed in the environment of the model or production of the data. we can say that there are four main types of model drift: Concept Drift: This can be defined as the changes in the relationship between the input and the output of the models. For example, we can say a model which is designed to predict the fraudulent from the emails and the classification of what is ‘fraudulent’ changes. Data Drift: This type of drift can be defined as the change in the model’s predictions for example in the fraudulent detection modelling procedure we have fewer emails or fraudulent at the data in the training time but at production we are having more data in fraudulent so the prediction of the model will have a tendency at the side of non-fraudulent emails side. Label Drift: It can be defined as the changes in the label’s distribution produced by the model as the output. Feature Drift: This drift can be defined as the changes in the input data for example we have to get some more words in the emails. Consisting of which email can be defined as fraudulent email. So here we can say that the concept drift is a dissimilarity between the real and learned decision boundary by the model. It becomes necessary to make the model learn or train again on the data so that we can maintain the accuracy range and error rates produced by the model. Model drift can be an indicator of the unavailability of real ground truth labels. And also the model drift is an indicator of the changes in the environment. However, if we are capable of measuring the cause of model drift we can make a model decision boundaries with a tolerance so that the model can predict accurately even when any kind of drift is presented. There can be many causes of the drift, some of them are listed below. Causes of Model Drift There can be many reasons for drift to occur in machine learning models Change in the data distribution Due to externalities, there can be many changes in the data distributions. In such cases, we are required to perform the modelling procedure again with the updated data set. For example changes in the email categories due to change in the business type. Data Integrity Issue When this type of issue occurs in the data we are required to perform an investigation on the data for example faulty data engineering can cause the change in the data even if we have entered the correct data in the source. Or we enter wrong data in the source. How to Detect Model Drift? There can be various methods through which we can detect the model drift some of them are listed below. Measuring the accuracy of the model This can be considered as an accurate way of detecting model drift which can be done by comparing the model predicted values to the actual models. We can observe a drift if the predicted value has deviated much from the actual values. There are various metrics that can be used for the measurement of accuracy. One of the famous metrics is the F1 score which encompasses both the precision and recall of the machine learning model. Image source The image is a representation of the precision and recall of predictive modelling.  Whenever a metric falls out from the range of a threshold we can assume that there is a model drift. Kolmogorov-Smirnov (K-S) test This test is basically a nonparametric test that can be used for making a comparison of cumulative distribution between the data sets. To measure the model drift we can use this for making comparisons between training data and post-training data. Where the null hypothesis of this test indicates that the distribution of the datasets which are being compared has the identical distribution and in our case, rejection of the null hypothesis is an indication of the model. Population stability Index (PSI) This test can give a measurement of the changes in the variable distribution over time. PSI is a famous metric for measuring changes in the population’s characteristics. And this can help us in measuring the model drift. Source Z-Score Using the Z-score we can compare the feature distribution between the two datasets in our case we can compare between the training and produced data of the model.  if several produced data points of a given variable have a z-score of +\/- 3, there is a shift in the distribution. Ways to Automating Model Drift However, we know that model drift is something about the increased losses produced by the machine learning model which can be detected by the above-given method. But when it comes to production it really becomes necessary to provide a solution to model drift and manually dealing with the model drift becomes cost consuming and time-consuming. So, we are required to deal with model drift in an automatic way where modelling techniques can automatically detect the model drift and can perform required changes in the model or in the data. In the next section, we will see how many ways we can automate the model drift. We can say there can be various ways using which we can automate model drift. Some of them are listed below. Online Learning Online machine learning is a prominent way to deal with model drift because it allows us to update learners in real-time and models allow us to deal with one sample at a time. Because in online learning the models are learned in a setting where it takes the batches of samples with the time and the learner optimizes the batch of data in one go. Where the model finds out the relationship between the independent and dependent variables. Since these models work on the fixed parameters of a data stream they are required to retrain the new patterns of the data. Also, these models are capable of learning from the large data streams they can be applied to different domains like time series forecasting, movie or eCommerce recommender systems, spam filtering, and many more, where the changes in data occur frequently. The above image is a representation of the basic online learning procedure where a model is used to predict the dependent variable instance and which can be used for upcoming new dependent variable instances and also it causes the model to be updated every time it gets used for making predictions. One way to perform online learning can be done by the creme library based on python. You can find out a tutorial at this link. Azure ML Using Azure ML we can automatically identify the model drift especially the data drift. The basic need for the procedure is to integrate the models into the Azure ML workspace. We can select the features to identify the model drift. Since it uses different methods mainly statistical methods and different time windows to identify the drift. Image source The above image is a representation of how the dataset differs from the target dataset in the specified time period when we integrate the model with Azure ML. Since it uses the python codes for managing the data drift it is an easy way to drift detection. Also, they have provided the tutorial for the procedure in this link. Using it we can perform the following monitoring:- Analyze driftMonitor model dataMonitor new dataProfile features in dataSet up alerts on data driftCreate a new dataset version Evidently: EvidentlyAI can also be used for evaluating and monitoring models in production. Since it is an open-source tool we can use it freely and also it can be a place where we can start monitoring model drift in our ML projects. You can find out a tutorial at this link. Fiddler AI Monitoring: it is also a way for monitoring our model’s drift. It also provides many of the tools consuming which we can make our models explainable, operating models in production and monitor models also can be used for data and model drift detection. You can find out a tutorial at this link. Final words In this article we had a basic understanding of the model drift and also we have seen some types of model drift. There is always a need to make a model error less which can be done by measuring the drift in that sense we have seen different techniques for measuring the model drift. In the end, we have discussed how we can monitor them automatically through different approaches.","excerpt":"manually dealing with the model drift becomes cost consuming and time-consuming. So, we are required to deal with model drift in an automatic.","categories":["AI Trends"],"tags":[],"author_name":"Yugesh Verma","publish_date":"2021-11-16T11:00:00","publication_year":"2021","word_count":1909,"keywords":["Go","machine learning","TPU","AI","R","ML","Python","data engineering","Azure","Azure ML"],"extracted_tech_keywords":["AI","machine learning","ML","Azure ML","Azure","TPU","Python","R","Go","data engineering"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/what-are-the-ways-to-automate-model-drift\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":24186,"title":"Omega Healthcare’s Baskaran Gopalan Shares The Data Threats A Hospital Faces And How To Ensure Security","content":"The Facebook and Cambridge Analytica debacle created much furore in the data community, and the threat of user data being exposed to companies without their knowledge became the reason for concern. With the rise of social media and other interactive channels, there is definitely a rise in user data and it is of utmost importance to ensure its privacy, or the least, have policies in place that ensure the rightful use of data. This May, Analytics India Magazine is going to bring out the privacy concerns haunting data — how companies tackle it, policies they are adopting, need for regulations, and others, from professionals across various sectors such as healthcare, digital payment and analytics, among others. We spoke to Baskaran Gopalan, senior vice president, IT and TMO at Omega Healthcare Management Services Pvt Ltd. Omega is a noted offshore provider of healthcare outsourcing services specialising in billing, coding, accounts receivable management, and other healthcare related services. They deal with a lot of clinical and patient data, and Gopalan shares with us the security measures they adopt to maintain data privacy. Analytics India Magazine: What are the security measures that you adopt to ensure confidential data stays safe at Omega Healthcare? Baskaran Gopalan: Omega does not store any Protected Health Information (PHI) or Personally Identifiable Information (PII) within its environment. However, we have implemented safeguards to protect the data at rest and in motion, such as data encryption. In addition, we provide limited access to PHI or PII data only to a particular set of users, based on access control policies. We also ensure data is not transmitted out of Omega network to other third party networks, and ensure all IT infrastructure is HIPAA and HITRUST compliant. AIM: Given the large data sets that healthcare industry deals with, what are the measures that hospitals should take? What are the specific measures taken by Omega Healthcare? BG: There are six ways in which hospitals can provide superior security and privacy solutions for securing Electronic Health Records: Enhance administrative controls Monitor physical and system access Identify workstation usage Audit and monitor system users Employ device and media controls Apply data encryption Apart from these, Omega also follows the standards and regulatory requirements such as ISO 27001 and HIPAA and HITECH compliance. AIM: What is your take on the recent news regarding data breaches, or apprehensions in making Aadhar Card? How can data privacy be ensured so that users don’t feel deceived by it? BG: On current Facebook or Aadhar card attacks related to privacy breach — users post their personal information about location, religion, sex and age but analytics company do profiling and sell this information to third parties for their marketing purpose. These companies should be held responsible for allowing third party companies to allow access to their meta data even though they have committed to users about safeguarding their data. Considering the interconnected world we live in there is lack of user awareness and users should be cautious about such highly-personal information they provide on the virtual domain. As far as hospitals and the healthcare industry is concerned, they should adopt a strong privacy framework and always take user consent for storing PII data, and limit the exposure of such data both internally and externally. Omega strictly follows the guidelines for ‘Use and Disclosure of Patient’s information’ as mandated in the HIPAA 45 CFR 164.512 Uses and disclosures of PHI. AIM: How do you train your staff on data security? BG: On the first day of joining, a new employee undergoes compliance training covering information security and privacy training. They are taught the do’s and don’ts during their employment and how to safeguard customer data during their day-to-day work. We also conduct a bi-annual compliance training for all our employees on Data Privacy of Client’s and Omega’s information. We have our own Learning Management system, where we cover individual participation and the passing criterion is set as high as 80 percent. AIM: How would you mitigate emergencies in case of data breach? Please highlight some of the measures that you are equipped with BG: We have not faced any data breach since inception. Breaches can happen because of external attacks and insider attacks too. Nowadays companies have the best IT security controls in terms of preventive technologies but need to also focus on detection, response and recovery mechanisms. We have in place a 24×7 cyber security operation centre and have tied up with a leading IT service company to further strengthen our security. AIM: How do you scrutinise data security methods? BG: Omega Healthcare deploys the best data security measures so the business is technology-enabled with no compromises on data security. We also ensure all data security methods meet or exceed industry best practises, and we have rigorous evaluation and selection criterion for adopting any new data security processes or systems in our organisation AIM: As the technology is changing rapidly, how do you cope up with the changes, as they might compromise your methods, software and tools? BG: Today the primary challenges faced by IT departments is mostly related to scalability. Converged technologies are being considered by most of the organisations as the way forward to address these challenges. Omega Healthcare has also adopted virtualisation and deployed hyper-converged solutions in its environment. We also develop applications in-house embracing trending technologies such as AI, big data and ML. AIM: Do you hire employees based on data security knowledge? What are the various positions in this role and the skill sets that you look for? BG: We have a dedicated information security department in our organisation. Some of the positions include information security and compliance roles. The basic skill-sets we look for are knowledge of information security domain, an experience of minimum three years and industry certification such as CISA, CISM, CRISC and ISO lead auditors.","excerpt":"The Facebook and Cambridge Analytica debacle created much furore in the data community, and the threat of user data being exposed to companies without their knowledge became the reason for concern. With the rise of social media and other interactive channels, there is definitely a rise in user data and it is of utmost importance […]","categories":["AI Features"],"tags":["data privacy India","Interviews and Discussions"],"author_name":"Srishti Deoras","publish_date":"2018-05-02T12:17:21","publication_year":"2018","word_count":974,"keywords":["Go","API","AI","ML","Scala","data privacy India","Git","Aim","analytics","Rust","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","ML","analytics","Aim","R","Go","Rust","Scala","Git","API"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/omega-healthcares-baskaran-gopalan-shares-the-data-threats-a-hospital-faces-and-how-to-ensure-security\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10011669,"title":"EY Acquires Gurugram-Based AI-Enabled Skilling Platform Spotmentor","content":"In a recent development, EY India acquired Spotmentor Technologies, an AI startup specialising in HR tech space founded by four Indian Institute of Technology-Kharagpur alumni, for an undisclosed amount. It comes in a move to integrate artificial intelligence and machine learning in their product offering. EY believes AI is crucial for boosting their service offerings to the client. Gurugram Spotmentor is an end-to-end skilling platform to help businesses identify skills required for the future of work, upskilling and reskilling talent. EY is looking to strengthen its people advisory portfolio of digital services through the acquisition. It was founded by Deepak Singh, Arpit Goyal, Shekhar Suman and Yash Pl Mittal in 2016 to serve industries such as on large enterprises, government and industry bodies to help them identify required skill and competency gaps through hyper-personalised learning and reskilling plans. The company believes that the startup’s AI solutions can be deployed across an organisation’s value chain. Rohan Sachdev, Partner and Consulting Leader, EY India told media that the acquisition of Spotmentor Technologies will boost their digital offerings by combining their strength in strategic people consulting and seasoned experience in learning and skills development.It is interesting to note that big four firms along with other consulting firms are competing in an environment where artificial intelligence is becoming crucial to the way they offer services to their clients.","excerpt":"In a recent development, EY India acquired Spotmentor Technologies, an AI startup specialising in HR tech space founded by four Indian Institute of Technology-Kharagpur alumni, for an undisclosed amount. It comes in a move to integrate artificial intelligence and machine learning in their product offering. EY believes AI is crucial for boosting their service offerings […]","categories":["AI News"],"tags":[],"author_name":"Srishti Deoras","publish_date":"2020-11-16T14:14:17","publication_year":"2020","word_count":224,"keywords":["Go","machine learning","artificial intelligence","programming_languages:R","AI","Git","RAG","GAN","R","startup"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","RAG","R","Go","Git","GAN","startup","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/ey-acquires-gurugram-based-ai-enabled-skilling-platform-spotmentor\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10008755,"title":"How Dfinity Wants To Redesign The Web &#038; Compete With Big Tech Using Internet Cloud Network","content":"We know how large technology companies have too much power and influence in today’s digital world. If we look at web applications, they are mostly powered by centralized server models from companies like AWS. But one company which is challenging this notion. By running an advanced blockchain computer protocol across a network of independent data centres, the Dfinity Foundation has created an open development platform that enables next-generation internet services, enterprise systems and distributed applications. Dfinity calls it an “internet computer,” an open network that serves as a large virtual mainframe computer in cyberspace. That would allow decentralized versions of many online technology services like Uber or Dropbox, according to the company. Because the internet computer is not hosted for running apps in a centralized location, it can help people who don’t want to depend on cloud servers. While Dfinity plans to introduce its own hardware and data centers, anyone else can also join the network to add to storage and processing needs for apps. According to Dominic Williams, Dfinity Founder and Chief Scientist, the Network Nervous System (NNS) is the technical solution to the systemic problems big tech has created with its monopoly over the internet, a public utility that should be completely open — bringing back the concept of the programmable web. Launched recently by Dfinity, the Sodium Network will unveil the Network Nervous System (NNS), an open algorithmic governance system that controls the Internet Computer. The ICP protocol is a blockchain computer protocol. One of the things that makes the protocol unique is that it has an open onboard governance system called the network nervous system, which is responsible for controlling, configuring and managing the network. Many say this could be a game-changing decentralised computing network. Sodium, the fourth of five milestones, represents the future, introducing the open algorithmic governance system that will manage the network. Dfinity’s New Web Architecture The internet computer network is constructed from a hierarchy of building blocks. At the bottom, you’ve got data centers. The company says that in 10 years there will be thousands of data centers and millions of these node machines in the internet computer network. The data centres, just like other tech companies, will host standardised node hardware for the overall network. The node machines are combined to create something called subnets. The subnets will host canisters which are the interoperable compute units that are uploaded by users. Instead of terminal-based apps, canisters store their own front-end, meaning apps can store their state, backend logic, and front-end directly in a canister. This represents a new and drastically simplified way of building apps for developers, and the beginning of a paradigm shift that we aim to usher in through the Internet Computer. Data centres get into the network by applying to the network nervous system. So the network nervous system is responsible for inducting data centres. The network nervous system itself is an open governance system that permissions participation in the network. In a sense, it plays a role equivalent to ICANN on the internet, which for example doles out autonomous system numbers for those that want to run BGP routers. Motoko: A New Programming Language For Internet Computer The Internet Computer is now operating across many data centres worldwide, and according to the company, more data centres will be added. The network is open to third-party developers who are building innovative applications and enterprise systems using the network’s expanding range of developer tools. The company had earlier introduced Dfinity Canister SDK (V0.3.0) as well as Motoko, a new programming language that was optimised for creating tamper-proof software and open internet services for the Internet Computer using WebAssembly (Wasm). These tools equipped developers to write canisters in Motoko and compile to WebAssembly. Further, they can run a local instance or “node” of the Internet Computer — known as a replica — on their laptop and deploy compiled programs as standalone canisters. Developers can also interact with the canisters using a command-line interface. Motoko statically compiles to WebAssembly, a portable binary format that abstracts cleanly over modern computer hardware, and thus permits its execution broadly on the internet, and the Internet Computer.” Andreas Rossberg is on the team behind this, as one of the designers of WebAssembly. By design, Motoko helps developers write safer and more efficient code, compiles quickly, and communicates with Wasm modules written in other languages. It also enables developers to write sequential code even though Internet Computer canisters — code compiled into a Wasm module and ready to deploy onto the Internet Computer network — communicate with each other asynchronously. Using Motoko makes it easier to quickly develop applications on the Internet Computer, but it isn’t the only language developers can use. The Internet Computer uses a Wasm execution environment, meaning that any language that can be compiled to Wasm, such as Rust, can be run on the Internet Computer. Plus, it has Orthogonal persistence which eliminates the need to maintain and manage external databases or storage volumes. Decentralised Applications Which Are Outside The Control Of Any One Company Here, rather than relying directly on AWS, Google Cloud or Microsoft Azure, the applications depend on the distributed architecture that Dfinity is building. Dfinity demonstrated CanCan — an open version of TikTok that operates across distributed data centres, enabling it to quickly scale storage as needed — at the Tungsten launch event. CanCan was written with less than 1,000 lines of code to highlight the simplicity of building on the Internet Computer built using WebAssembly. The fact that CanCan is running on the Internet Computer across multiple data centres marks the Internet Computer as a new type of blockchain computer with infinite capacity and performance that rivals the traditional cloud. Dfinity also rolled out an open version of LinkedIn, called LinkedUp. Comparison With Tech Companies It is easier said than done when building the internet and hosting millions of apps from scratch. The tech giants like AWS have invested hundreds of billions of dollars into servers and rearchitecting the web to make it more transparent and decentralised would need a similar level of investment, particularly if it needs to host all the data and run apps at the same time. While the software innovation which Dfinity has introduced is certainly disruptive, it’s not merely about software to build an internet cloud. It is unique and innovative that Dfinity compiles down to WebAssembly but with a decentralised computation overlay on top of the whole internet. Compiling to WebAssembly is definitely a good thing, and we need to get the internet off javascript and start using real programming languages to build next-generation applications on the internet. Decentralized computation networks like Dfinity stand to bring us closer to a world where digital platforms can be constructed from trustless, autonomous, and open source software that is owned and governed by communities of users and developers, rather than companies. Dfinity raised $61 million from Andreesen Horowitz, one of the most influential technology venture capital firms in the world, and Polychain Capital in a February 2018 funding round. Another $102 million were raised in the second round in August 2018 thus bringing total funding to $195 million. Andreesen Horowitz said it was ready to back it up with more investment if needed and believes in the future of truly distributed computing which is serverless. The Internet Computer is seen as a next-generation distributed computing system — similar to its Mainframe, Client Server, and Public Cloud predecessors, but one which is based on cryptography and blockchain consensus. Despite its high ambitions and the amount it has raised, the plans to create a world computer aren’t going to come to fruition overnight. If you look at Dfinity in the context of distributed computing and blockchain, the project is undoubtedly disruptive. Dfinity could very possibly build an effective and scalable infrastructure for running high throughput decentralized apps. But will it replace incumbent internet giants to host hyperscale business apps which need tremendous amounts of processing, storage and memory capacity? This seems unlikely (in the short term) given the amount of investment that has gone into building AWS, Azure or Google Cloud.","excerpt":"We know how large technology companies have too much power and influence in today’s digital world. If we look at web applications, they are mostly powered by centralized server models from companies like AWS. But one company which is challenging this notion. By running an advanced blockchain computer protocol across a network of independent data […]","categories":["AI Features"],"tags":["big data storage format","Blockchain","data structure using java","distributed graph database","online network graph"],"author_name":"Vishal Chawla","publish_date":"2020-10-03T16:00:00","publication_year":"2020","word_count":1355,"keywords":["big data storage format","Go","Blockchain","AWS","AI","online network graph","R","data structure using java","distributed computing","serverless","RAG","distributed graph database","Aim","JavaScript","Azure"],"extracted_tech_keywords":["AI","Aim","RAG","AWS","Azure","serverless","distributed computing","R","JavaScript","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-dfinity-wants-to-redesign-the-web-compete-with-big-tech-using-internet-cloud-network\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":12211,"title":"Vidooly- the video intelligence platform using analytics","content":"Team Vidooly at a Outing at Kingdom of dreams in Gurgaon The one of its kind video analytics and marketing company, Vidooly, has been providing tools for online video creators, multi channel networks, digital focused brands and agencies since 2014. Well, it might have spent just two years in the industry, but it has already been proving its uniqueness from its global counterparts as it caters to multiple stakeholders of the online video ecosystem unlike most others. Founded by Subrat Kar, Nishant Radia and Ajay Mishra, the idea of founding Vidooly was conceptualized over a friendly lunch meet in Noida when they randomly started discussing the possibilities in the field of online video. This is when, all three video enthusiasts saw a tremendous opportunity in this ecosystem and Vidooly came into existence. What is the tech behind Vidooly? The Founders Claiming to use analytics in its working, AIM tried to know how exactly is Vidooly using Artificial Intelligence and Analytics? Here’s how it is. With a drastic increase in the number of people connected to the internet, there has been a steady increase in the quality of connections and in the number of audience in the online video space. In such a scenario, Vidooly believes that it is more important than ever to understand the behaviour of the digital audience. “With so much viewership happening everyday, online video creators and brands need to understand data to understand their audience. We at Vidooly have been tracking the data signals of around 15 Billion video views and 100 million audience members every day! That’s a lot of data”, said Subrat, the CEO of Vidooly. Subrat has been an entrepreneur and used to work at Jabong before taking a stint at Vidooly where he also looks after the Sales and Business Development. Vidooly uses this goldmine of data to predict the behaviour of audience depending on the need of the customers. “We even have built our own custom technology that can predict the virality of content. Our brand solution tool has a feature that tells them which videos are about to trend in the days to come. This is an invaluable piece of information that brands can use to strategically advertise and get the maximum ROI for their spend”, added Subrat. So, how has been the growth story? Anniversary Celebrations With an interesting concept in hand, the journey has been exciting and difficult at the same time, avers Nishant who takes the position of CMO and is in charge of marketing piece. “Vidooly started off with a focus just on YouTube and then slowly started becoming more platform agnostic. Right now, Vidooly’s big data engine tracks data from all the four major social media platforms – Facebook, Instagram, Twitter and YouTube”, he said. Nishant has earlier worked with Headstrong. While starting off in 2014, the concept of online video ecosystem in India was still at a nascent stage and after overcoming the biggest challenge of making potential customers understand the importance of data and analytics, Vidooly has come a long way. It has grown from a team of around 5 people to 40 people within a span of 15 months with a swanky office space in Noida. Seed funded initially by Bessemer Venture Partners, the company has great plans in store for 2017. Form social video, it now plans to spread its wings towards Video On Demand\/ OTT as well. Happy Clients “Exciting things are happening in the field of online video in India now. With the entry of global VOD giants like Netflix and Amazon Prime and the emergence of homegrown OTTs like Hotstar and Voot, the audience is spoilt for choices. That’s why it is absolutely necessary for brands who want to advertise on these platforms to understand their viewing behaviour to get the best bang for their bucks. YouTube and Facebook video is also gaining massive traction in India given the increased internet connectivity and quality. So, we at Vidooly believe that we’re at the right place at the right time”, concludes Ajay, who is the CTO and heads the tech team.","excerpt":"The one of its kind video analytics and marketing company, Vidooly, has been providing tools for online video creators, multi channel networks, digital focused brands and agencies since 2014. Well, it might have spent just two years in the industry, but it has already been proving its uniqueness from its global counterparts as it caters […]","categories":["AI Startups"],"tags":["Startups"],"author_name":"Srishti Deoras","publish_date":"2017-01-18T10:34:14","publication_year":"2017","word_count":682,"keywords":["big data","Go","artificial intelligence","AI","ML","Git","Aim","ViT","analytics","Startups","R"],"extracted_tech_keywords":["AI","artificial intelligence","ML","analytics","Aim","R","Go","Git","big data","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-startups\/startup-week-vidooly-video-intelligence-platform-using-analytics\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10168768,"title":"Mid-Sized IT Firms Outperform the Big-Four in Q4 FY25","content":"For years, the Indian IT story has centred around the dominance of the “Big Four”—TCS, Infosys, Wipro, and HCLTech. But that’s changing rapidly. In FY25, it’s the mid-sized companies—Persistent Systems, Mphasis, LTIMindtree, L&T Technology Services, WNS, and even the slightly larger Tech Mahindra—that have emerged as the real growth engines of the industry. Persistent Systems, for instance, posted a staggering 25% year-on-year revenue growth for Q4 FY25, a figure that starkly contrasts with the Big Four’s lacklustre single-digit or negative growth outlook. HCLTech barely managed to achieve a 1.2% revenue growth, while TCS and Wipro crawled to a 0.8% increase. Infosys, on the other hand, experienced a revenue dip of 4.2%. These four giants also project a very bleak outlook for the next year, which contrasts sharply with that of the smaller IT firms. Persistent clocked its 20th consecutive quarter of growth, something even the giants have struggled to maintain lately. CEO Sandeep Kalra pointed out that operational discipline and client trust were key. Behind the scenes, founder Anand Deshpande made it clear that Persistent’s early bet on agentic AI, which many mid-sized IT firms have started in the last few quarters, is the reason behind this. Deals driven by AI are not just a buzzword While Persistent led the pack, LTIMindtree also delivered a positive performance, albeit a more modest one. The company reported a 9.9% year-over-year increase in revenue and a 2.6% rise in net profit. Despite a slight dip in dollar revenue, the firm’s strong order book, again driven by AI-powered deals, helped it outperform a sluggish broader market. Debashis Chatterjee, LTIMindtree’s CEO, credited the wide integration of AI across services for the resilience, a sentiment increasingly echoed across mid-tier IT firms. And then came Mphasis, pulling off its strongest sequential growth in three years. Revenue increased 2.6% sequentially in USD terms and 2.9% in constant currency, a remarkable achievement considering the macroeconomic pressures that have battered the broader sector. What’s even more telling is that 59% of Mphasis’ deal wins were AI-led. CEO Nitin Rakesh made it clear: tech and AI are now at the very core of their strategy. Banking, financial services, insurance, and TMT verticals led the charge, compensating for weaknesses in other areas, such as logistics. Meanwhile, L&T Technology Services (LTTS) achieved a stellar 17.5% year-over-year revenue growth in rupee terms. Profits dipped 9%, but operational metrics remained strong, and the outlook for FY26 is bullish. Amit Chadha, the CEO, was confident that the coming year would be even better, backed by a strong pipeline of large digital and AI-led deals. Automation and AI were front and centre in their narrative too, hinting at a clear industry pivot. WNS, the smallest firm among them, reported a slightly lower quarterly revenue of $336 million, showcasing flat growth compared to the same period last year and an increase of only $3 million from the $333 million reported in the previous quarter. Despite this, the company projects a 7-11% growth in the next quarter. Despite the weak quarter, CFO Arijit Sen stated that the positive outlook is based on current visibility levels and includes a 2% revenue contribution from Kipi.ai, with 90% visibility already achieved to the midpoint of revenue guidance. Even Tech Mahindra, which experienced a rough run in recent quarters, managed to show signs of life. While revenue remained relatively flat, the company reported a massive 77% increase in net profit. The flat revenue was attributed to delays in renewals with major clients, particularly in the high-tech sector. But the fact that Tech Mahindra’s bottom line improved despite that speaks volumes about its renewed operational focus and margin discipline. This shift isn’t coincidental; it’s part of a larger trend. Unlike the larger firms, which are still engaging in AI washing and slowly integrating AI into their deals with a 30-year-old business model, smaller firms have a slight advantage due to their high agility. Agentic AI and Acquisitions Paying Off When mid-sized Indian IT firms began announcing acquisitions and partnerships related to agentic AI over the last three quarters, it felt like just another attempt to keep pace with the generative AI wave. But as Q4 results rolled in, it became clear that these early bets are already paying off. A year ago, the mood among mid-sized Indian IT firms was clear. Instead of building expensive generative AI models from scratch, they would partner with or acquire specialised startups to develop use cases. LTIMindtree, for instance, announced a $6 million investment in Voicing AI, a US-based startup that builds human-like AI voice agents capable of operating in over 20 languages. LTIMindtree also partnered with GitHub to integrate Copilot into its developer training programs, preparing its workforce for an AI-first future. Similarly, Mphasis doubled down on conversational AI with the launch of NeoCrux, a tool designed to enhance developer productivity through AI orchestration. In October 2023, the company also acquired Silverline, a Salesforce consulting partner, to bolster its customer experience and conversational AI capabilities. Persistent Systems, on the other hand, focused on the foundational issues of AI adoption — namely, privacy and governance — with its acquisition of Arrka, a Pune-based data privacy consultancy. While major players such as TCS, Infosys, and HCLTech developed internal capabilities and launched AI business units like AI.Cloud, Topaz, or AI Force, mid-sized IT firms like LTIMindtree and Mphasis chose to focus on agility. “We would want to leverage startups for technologies that are coming up, rather than building expensive in-house models that may become obsolete quickly,” said Nachiket Deshpande, president of AI services at LTIMindtree, in an earlier conversation with AIM. Last year, companies like Happiest Minds, Hexaware, Quest Global, Coforge, Sonata, and GlobalLogic had already accelerated their acquisition of AI capabilities. However, the conversation has now shifted squarely to agentic and generative AI, as firms realise that enabling autonomous workflows is the next significant differentiator.","excerpt":"For mid-sized Indian IT firms, acquisitions and partnerships related to agentic AI over the last three quarters have begun to yield results.","categories":["IT Services"],"tags":[],"author_name":"Mohit Pandey","publish_date":"2025-04-29T10:39:04","publication_year":"2025","word_count":976,"keywords":["Go","agentic AI","AI","Git","RAG","Aim","generative AI","Rust","GitHub","R"],"extracted_tech_keywords":["AI","generative AI","agentic AI","Aim","RAG","R","Go","Rust","Git","GitHub"],"url":"https:\/\/analyticsindiamag.com\/it-services\/mid-sized-it-firms-outperform-the-big-four-in-q4-fy25\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10017035,"title":"Free Online Resources For Kids To Learn Robotics In 2021","content":"In our last article, we jotted down top schools for robotics engineering in India. While the schools we have discussed were for graduate students and professionals, learning robotics is no more limited to just them. Robotics is now also for kids. In today’s technology-driven landscape, it has become imperative for kids to be industry-ready at an early age — this is another way to create a more industry-ready workforce for our country. Also, with robotics becoming a crucial part of various industrial applications and for society’s progress, learning robotics at a young age can be extremely beneficial. Furthermore, experts believe that teaching robotics to kids can augment their problem solving and critical thinking capabilities and equip them with the necessary programming skills that are most demanded in the current industry 4.0. This, in turn, opens up a wide range of career opportunities for them in future. Thus, in this article, we will share the best eight resources for kids to learn robotics. Also Read: Child’s Play: How Gen Z Kids Prefer AI-Powered ‘Smart’ Toys Over Legos & BarbiesA Fake Blog Created By College Kid Using GPT-3 Made People Believe Human Wrote ItNo More Personalised Data Collection On YouTube (For Kids Only)Still making business decisions intuitively? It’s time to use Analytics.How Young Is Too Young: AI For Children Arduino Projects & Robotics Tutorials By RootSaid About: Arduino Projects & Robotics Tutorials is a YouTube video tutorial that has been designed for kids as well as beginners. These video tutorial series consist of nine chapters that will focus on the basics of robotics, different parts of robots, its processing units, and electronics behind it. This tutorial also covers a few DIY hobby projects for kids. With this tutorial, one can build their running beginner robots, with knowledge on robot chassis, battery operation, microcontroller, and everything needed to make a robot. Additionally, RootSaid also offers video tutorials on coding for beginners and robotics for beginners to help kids pick up knowledge on basic robotics at an early age. Check out the video tutorial here. Robotics for Kids – Level 1 By STEMpedia About: A free online course, Robotics for Kids – Level 1 is a middle school STEM robotics curriculum designed by industry experts for teaching hands-on robotics to kids. This course teaches kids all about robotics from assembling basic mobile robots to building a smartphone-enabled robot. Alongside this course offers some exciting DIY robotics projects for kids that will help them learn robotics at home, with parental guidance. With this course, kids will now be able to learn the basics of robotics along with an understanding of algorithms and can develop their attention to details and problem-solving capabilities at a very young age. Know more about the course here. Coding and Robotics for Kids By Robolabz About: ‘Coding and Robotics for Kids’ is a free robotics course for kids On YouTube, in Hindi. Presented by Robolabz that focuses on providing online and offline education, this course acts as a startup kit for kids to initiate their journey towards STEM and robotics. This course will well-equip the kids with basic programming and robotics concepts and do not require any hardware or other materials for start learning. This course can even be included in the school curriculum for kids to pick up complex subjects like robotics programming and mathematics at an early age. Know more about it here. Robotics for School Students By Coding Blocks Junior About: Robotics for School Students is a 60-minute free live-streaming YouTube video series that focuses mainly on providing a hands-on building and learning robotics for kids. Started during the pandemic lockdown, this video series is a certified course which will give the kids a certificate at the end of this course, if all the assignments are completed. This video course streams two times a week and talks about the fundamentals of robotics and focuses on building robots and programming knowledge required for robotics. Check out the video series here. Also Read: 8 Online Games That Can Teach Kids Data Science Amid This Lockdown Begin Robotics By Future Learn About: This free online course is offered by Future Learn, in collaboration with The University of Reading, a university based-out of the UK that focuses on teaching, research and enterprise. Part of the Study UK Campaign, this course focuses on teaching robotics to kids by exploring the history, anatomy and intelligence of robots along with building them with exciting simulations. This four-week course starts from introduction to robotics and robot anatomy and then follows through cybernetics and control, and understanding robot behaviours. With only three hours of weekly study, this course claims to equip robotics-savvy kids with a fundamental understanding of this complex field at an early age. Know more about it here. Wonder for Dash & Dot Robots By Wonder Workshop About: Amid other online resources like websites and video tutorials, one mobile application that has been in the spotlight is ‘Wonder for Dash & Dot Robots.’ Created by Wonder Workshop, a California-based robotics company, this mobile app teaches robotics to kids in a delightful way like finger painting. It acts as a coding tool that lets kids of 8+ years have fun with robotics, all on their own. The app comes with 300+ challenges that teach kids about robotics programming as well as building robots. With this mobile app, kids can build their robots, update them with different mechanisms, play challenges and test their simulations. This application claims to provide an open-ended play and learning experience for kids to develop problem-solving capabilities at a young age. You can download the app here. COVID-19 Relief Free Online Courses By Techy Kids About: While lockdown was making kids lazy and restless, Techy Kids, an online educational platform for kids, came up with a set of COVID-19 Relief Free Online Courses to help tech-savvy kids continue their learning at home during this time. This set of courses claim to provide hands-on learning on complex subjects in robotics. Starting from Intro to Robotics with Thymio Simulator to modelling your ideas in Tinkercad Course these programmes teach kids about sensors, robotics and visual programming. Techy Kids advises kids to initiate their journey with its intro course before taking up these complex robotics courses for a better experience. Know more about it here. Easy Robotics Projects for Kids By ResearchParent About: Along with web courses, video tutorials and mobile applications, another resource that can help kids learn robotics at an early age are the DIY projects. One such offer is from a website — ResearchParent — by a research scientist that shared Easy Robotics Projects for Kids that can be leveraged to understand robotics, with the help of parental guidance. Starting from homemade spinning brush bot to building a simple homemade robot car, these projects are fun and are specially built for kids as well as beginners. Check out the projects here.","excerpt":"Learning robotics opens up a wide range of career opportunities for kids in future. In this article, we will share the best eight resources for kids to learn robotics.","categories":["AI Trends"],"tags":[],"author_name":"Sejuti Das","publish_date":"2021-01-04T18:00:00","publication_year":"2021","word_count":1148,"keywords":["data science","Go","AI","RAG","GPT","Aim","llm_models:GPT","analytics","R","startup"],"extracted_tech_keywords":["AI","data science","analytics","Aim","RAG","R","Go","GPT","startup","llm_models:GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/free-online-resources-for-kids-to-learn-robotics-in-2021\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10085391,"title":"Gupshup launches GPT-3 Powered Bot Building Tool","content":"Gupshup,  an AI platform focused on providing conversational messaging for commerce, marketing, and support, today announced the launch of Auto Bot Builder, a powerful tool that harnesses the power of GPT-3 to automatically and effortlessly build advanced chatbots tailored to enterprise requirements. Auto Bot Builder leverages the GPT-3 Large Language Model (LLM) and fine-tunes it using proprietary enterprise knowledge base and domain expertise – resulting in a chatbot specialised to an enterprise, unlike ChatGPT, which is a general-purpose chatbot. This enables enterprises to incorporate the latest advances in AI technology, in a no-code environment, into their conversational experiences to substantially enhance their customer engagement. It reduces the time and effort it takes to create conversational journeys, enabling enterprises to develop more journeys faster for use-cases across the entire customer lifecycle. With the Auto Bot Builder tool, a user can instantly build a chatbot using content from their website, documents, message logs, product catalogs, database, and other corporate systems of record. The tool accepts content of any size, processes it and fine-tunes the Large Language Model (LLM) to the specific context. Auto Bot Builder is capable of handling really large content sets, which need special pre-processing. Previously, in an exclusive interview with Analytics India Magazine, Krishna Tammana, CTO at Gupshup said that the company wants to leverage [ChatGPT’s] capabilities at Gupshup to build solutions for enterprises that are smarter and more accurate. “It is in line with our vision of building a product stack that offers integrated conversational engagement solutions,” Tammana said.","excerpt":"With the Auto Bot Builder tool, a user can instantly build a chatbot using content from their website, documents, message logs, product catalogs, database, and other corporate systems of record.","categories":["AI News"],"tags":[],"author_name":"Pritam Bordoloi","publish_date":"2023-01-18T15:52:01","publication_year":"2023","word_count":251,"keywords":["ChatGPT","programming_languages:R","AI","chatbots","RAG","GPT","analytics","R","llm_models:ChatGPT","llm_models:GPT"],"extracted_tech_keywords":["AI","analytics","ChatGPT","RAG","chatbots","R","GPT","llm_models:GPT","llm_models:ChatGPT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/gupshup-launches-gpt-3-powered-bot-building-tool\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10161714,"title":"Luma AI Announces Ray2 AI Video Model Pre-trained with ‘10x Compute’","content":"Luma Labs has introduced a new video generative model called Ray2. It claims to create realistic videos with natural and coherent motion. Ray2 is currently available on the Dream Machine platform for paid subscribers. “We are starting with text-to-video. Image-to-video, video-to-video, and editing capabilities will soon follow,” said the company in a post on X. It can generate videos with a duration of five or ten seconds in multiple aspect ratios. The paid plans for Dream Machine start at £6.99 a month and enable users to create 1080p videos. “Today, we are introducing Ray2! Scaling pretraining by 10x (in comparison to Ray1) on a novel efficient architecture, we are able to unlock the next frontier in video generation—fast natural coherent motion and physics,” said Amit Jain, co-founder of Luma Labs, in a post on X. “This skyrockets the success rate of usable, production-ready generations and makes video storytelling accessible to many more people,” he added. Jain also revealed that the company is considering releasing an API for it soon and that it is ‘essential to our goals’. The company also released a long list of examples of videos generated across themes like natural motion, physics and simulation, photorealism, cinematic scenes, people and expressions. Ray 2 produces AI influencers on demand — complete with wide angle vlogging and rolling shutter 🫣 pic.twitter.com\/uUvgaX944u— Bilawal Sidhu (@bilawalsidhu) January 15, 2025 That said, users did run into a few issues accessing the model on launch day. One user on X complained that the servers were overwhelmed and that they had been waiting for 30 minutes for the result. Meanwhile, Jain noted that it was the ‘launch day rush’ and expected that things would stabilise soon. Users quickly tested the model. MattVidPro, an AI-focused tech YouTuber, conducted a poll on X, asking testers which model is more impressive: Ray 2, Google’s Veo 2, or Vidu 2 from Vidu AI. The results revealed that over 80% picked Google’s Veo 2 model. A user on X, Ben Nash, expressed that Veo is significantly ahead of its competitors and believes it is of superior quality compared to Sora, even though Sora offers a greater array of features. However, unlike OpenAI’s Sora and Ray 2, Google’s Veo 2 is not yet available for public use. Google’s internal testing indicates that Veo outperforms competitors (such as China’s Kling, Meta’s Moviegen, and OpenAI’s Sora) both in terms of quality and prompt adherence.","excerpt":"‘We are able to unlock the next frontier in video generation—fast natural coherent motion and physics,’ said Amit Jain, founder and CEO of Luma AI.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","AI Video Generation Models"],"author_name":"Supreeth Koundinya","publish_date":"2025-01-17T22:59:12","publication_year":"2025","word_count":401,"keywords":["Go","API","OpenAI","AI","programming_languages:R","programming_languages:Go","AI Video Generation Models","Ray","Aim","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","OpenAI","Aim","Ray","R","Go","API","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/luma-ai-announces-ray2-ai-video-model-pre-trained-with-10x-compute\/","complexity_score":4,"technical_depth":9,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10066238,"title":"Continuous-Time Markov Chain and its applications in machine learning","content":"There are a lot of applications of mathematical concepts in data science and machine learning. These concepts are very helpful in building and approximating intelligent programs. The Continuous-Time Markov Chain (CTMC) is a type of stochastic process that uses the Markov chain from probabilistic theory. We can find many interesting and useful applications of CTMC in machine learning. In this article, we are going to discuss the Continuous-Time Markov Chain(CTMC) in detail where we will try to understand the key concepts related to it and its applicability to machine learning. The major points to be discussed in the article are listed below. Table of content About Continuous-Time Markov Chain (CTMC)Properties of CTMCApplication of CTMC Let’s start by explaining the continuous-time Markov chain(CTMC). About Continuous-Time Markov Chain (CTMC) In mathematics, specifically in probability theory, various stochastic processes can be defined as a mathematical object that is a family of random variables. In machine learning, we can see the usage of stochastic processes as mathematical models. We can also say that stochastic processes are phenomena that can appear to vary randomly. We can take an example of a stochastic process from our daily life scenarios such as fluctuation in the traffic or current. The Markov chain is also a kind of stochastic process and we can think of it as a system that can experience changes from one state to another state under constraints by some probabilistic rules. In one of our articles, we have discussed the Markov chain process and its application in machine learning and data science.  In this article, we are focused on the continuous-time Markov chain process that can be considered as the expansion of the Markov chain. Continuous-time Markov chain is also a type of stochastic process where continuity makes it different from the Markov chain. This process or chain comes into the picture when changes in the state happen according to an exponential random variable. The movement toward the next step can be determined by the probabilities of a stochastic matrix. We can also think of CTMC as the process of changing the state of an exponential random variable set where the parameters can be determined by the current state. In the set, we can find the least value for each possible state. Let’s take a look at an example of three states { 0, 1, 2}, here for the transaction between the states the time factor works. This means after every time value there is a transition and we call this time the holding time.  Let’s say there are three random variables that in state i and independent from each other such as E0 Exp(6),E1 Exp(12) and E2 Exp(18). In such conditions when a transition between the states happens we can consider the movement of the chain is jump chain and this will be a discrete-time markov chain process and the stochastic matrix will be: If this will be CTMC according to the competing exponentials theory then transition in state from state i according to two minimum random variables can be given by the Q-matrix Q=(qi,j) as: In the above, we can think of non-diagonal value as the product of holding time and probability from the chain which is moving toward a definite state. The choice of the diagonal values should be done as all the values of the row sum up to zero. One thing that is important about the CTMC is that it satisfies the Markov property, Markov properties can be referred to as the memoryless property of any stochastic process. In CTMC the chain’s behavior is dependent on the current state of the chain instead of its past. Image source The above image is a representation of a CTMC with state-space where the Q-matrix or transition rate matrix will be: We can conclude it as the description of the state of the financial systems. Let’s discuss the properties of the CTMC. Are you looking for a complete repository of Python libraries used in data science, check out here. Properties of CTMC Some of the properties of the Continuous-Time Markov Chain (CTMC) are as follows: These chains are irreducible because they have time and continuity properties which means once going from one state to another they can not be reduced back.CTMC is stationary distributed. Since we can also think of it as the probability distribution that converges the process on it at a large time difference. The Q-matrix holds the row that has some distribution and this means CTMC does not depend on the first state.A CTMC can be called a reversible chain if it follows the condition where the reversed state is similar to the forward one. Application of CTMC in machine learning There are various applications of the Continuous-Time Markov Chain (CTMC) in machine learning. Some of them are as follows: CTMC can be considered as a stochastic model that helps in quantitatively analyzing practical systems, their performance, and reliability. This chain can be used for queuing layers and networks in large neural networks.This chain can also be used in determining the throughput of algorithms and their average failure time. After validating the models against reliability and performance, the model can be deployed to production. We can utilize the CTMC to model the timing of running different blocks of a machine learning system using an exponential random variable and mean holding time. Since most reinforcement learning algorithms acquire asynchronous dynamic programming and approximate using stochastic processes. CTMC can be a way to approximate the programming and can also be used for adaptive control of queueing systems. CTMC can also be used in the field of information theory, speech recognition, and search engines which are co-fields of machine learning.  CTMC can be used for modelling various random phenomena like genetics, demography, and epidemiology. Final words In this article, we have discussed the continuous-time Markov chain(CTMC) which is a stochastic process and having the time and continuity factor makes it different from the Markov chain. Along with this, we have seen some of the properties and applications of the continuous-time Markov chain(CTMC).","excerpt":"Continuous-time Markov chain is a type of stochastic process where continuity makes it different from the Markov chain. This process or chain comes into the picture when changes in the state happen according to an exponential random variable.","categories":["AI Trends"],"tags":["AI (Artificial Intelligence)","Data Science","Data Scientist","Deep Learning","Machine Learning","markov chain","mathematics and statistics"],"author_name":"Yugesh Verma","publish_date":"2022-05-04T14:00:00","publication_year":"2022","word_count":1009,"keywords":["data science","Go","machine learning","AI","neural network","R","ML","markov chain","Machine Learning","RAG","Python","programming_languages:Python","Deep Learning","Data Science","Data Scientist","mathematics and statistics","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","machine learning","ML","neural network","data science","RAG","Python","R","Go","programming_languages:Python"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/continuous-time-markov-chain-and-its-applications-in-machine-learning\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10090027,"title":"No Public Launch for China’s ChatGPT, Ernie Bot","content":"Days after announcing its rival to ChatGPT, Chinese search engine giant Baidu scrapped the launch of its multimodal large language model Ernie. The webcast, scheduled for Monday afternoon, was switched to a closed-door event for firms, citing ‘strong demand’. The company stated that the modification in the arrangement was initiated to meet the “intense requirement” from 120,000 businesses that had requested to try out the Ernie robot. The company further mentioned that this would be the initial of numerous confidential meetings. Baidu CEO Robin Li had introduced Ernie bot on March 16. During a live-streamed presentation, Li demonstrated a series of pre-recorded demos showcasing the various abilities of the Chinese chatbot to journalists. Upon the preliminary reports about the cancellation, Baidu’s shares listed on the Hong Kong stock market dropped by 4.5% on Monday morning. The company’s stock price had also declined initially while the CEO was still presenting the bot, but it recovered the following day. This rebound was partly due to high demand for generative artificial intelligence (AI) from the Chinese business industry. Apparently, the chatbot—-ERNIE (Enhanced Representation through Knowledge Integration) has been into tests for around a decade, according to Baidu CEO Robin Li. The company has been actively spending in AI research and development and spent about 21.4 billion yuan ($3.1 billion) in 2022 alone. During the presentation of Ernie Bot, Li acknowledged that the chatbot is not flawless. However, he explained that they are unveiling it due to market demand. Furthermore, Li mentioned that the model will also be incorporated into the search engine. Pros and Cons Interestingly the model can function in two languages, English and Chinese, and is also multimodal. In addition to its language generation capabilities, the model can also perform text-to-image tasks known as ERNIE-ViLG. This latest chatbot is now powered by the third iteration of the LLM, ERNIE 3.0 Titan, which contains 260 billion parameters, representing a 50% increase compared to ChatGPT’s parameters. According to the South China Morning Post, Ernie Bot is also able to offer real-time information unlike its rival ChatGPT. But political censorship has forced the makers to train the bot in a way that it avoids political questions like whether China is a democratic nation, responding vaguely: “hasn’t learned how to answer this question yet”.","excerpt":"The webcast, scheduled for Monday afternoon, was switched to a closed-door event for firms, citing ‘strong demand’.","categories":["AI News"],"tags":["Baidu","ChatGPT","China"],"author_name":"Shyam Nandan Upadhyay","publish_date":"2023-03-27T12:20:55","publication_year":"2023","word_count":379,"keywords":["ChatGPT","artificial intelligence","programming_languages:R","AI","Modal","GPT","Baidu","AI research","R","llm_models:ChatGPT","China","llm_models:GPT"],"extracted_tech_keywords":["AI","artificial intelligence","ChatGPT","R","GPT","Modal","AI research","llm_models:GPT","llm_models:ChatGPT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/no-public-launch-for-chinas-chatgpt-ernie-bot\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10166218,"title":"Google to Partner With MediaTek for Next-Gen TPU","content":"Google is set to partner with Taiwanese chip maker MediaTek for the next version of its Tensor Processing Units (TPUs), reported The Information on Monday. This marks a significant shift as Google has relied on Broadcom as the primary supplier of TPUs for nearly a decade. However, it was also reported that Google has not cut ties with Broadcom despite the new development.  Google’s decision comes against the backdrop of MediaTek’s ties with TSMC. The AI giant is looking to have cost leverage over chip production, the report said. Furthermore, Google is likely to handle the design of the TPU for the most part, while MediaTek will handle the input and output modules. This approach differs from Google’s collaboration with Broadcom, which developed the core parts of the TPU chips, according to sources cited by The Information. Google’s TPUs are custom-designed chips built to accelerate machine learning workloads across various platforms, from large-scale data centres and individual edge devices alongside cloud-based systems. TPUs power Google’s leading AI models, applications and services. In 2023, The Information also reported that Google’s executives ‘extensively discussed dropping Broadcomm’ as the supplier of TPUs as early as 2027. According to a report from Omdia last year, Google spent $6 billion to $9 billion on TPUs. The company announced its sixth generation of the TPU – Trillium in 2024. Compared to its predecessor, Trillium delivers a four-fold performance improvement. Further, it also provides a 67% increase in energy efficiency and a three-fold increase in inference throughput. In January, an industry analyst pegged the value of a ‘potential stand-alone TPU business’ to be over $700 billion. Google’s TPUs have been an alternative to several companies over the chips sold by NVIDIA. Last year, it was also reported that Apple employed Google’s TPU clusters to train its foundational models.","excerpt":"But it hasn’t cut ties with Broadcom yet.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","MediaTek"],"author_name":"Supreeth Koundinya","publish_date":"2025-03-18T13:23:53","publication_year":"2025","word_count":301,"keywords":["Go","machine learning","TPU","programming_languages:R","AI","programming_languages:Go","RAG","R","AI (Artificial Intelligence)","MediaTek"],"extracted_tech_keywords":["AI","machine learning","RAG","TPU","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/google-to-partner-with-mediatek-for-next-gen-tpu\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":41756,"title":"Obscurity Is Out And Obfuscation Is In As Cybersecurity Moves Towards Muddying User Data","content":"Cybersecurity has long relied on hiding one’s identity as an integral part of protection. Owing to the fact that one’s personal information is the riskiest thing to lose online, cybersecurity generally focused on protecting this information through routing packets or providing false information. However, it seems as though this cannot be practised any longer. The Internet is more predatory than ever in collecting user data through cookies and trackers. Every click, scroll and hover is recorded and used for advanced data analytics. To combat this, a new kind of cybersecurity has come to the forefront. The primary weapon of this approach is not to hide, but instead to muddy data. In a practice known as obfuscation, data that is given to parties on the Internet is generally artificially modified. This modification is done in a way that both obscures the identity of the individual and renders the data useless for parties wishing to use it. This practice is known as cybersecurity through obfuscation. A History Of Obfuscation Obfuscation was first discovered as a method for identity preservation in 2009 with the launch of browser extension TrackMeNot. Built to prevent tracking from search providers, the extension employs obfuscation as a strategy of security. Upon searching for a particular query, the user’s query is hidden in a group of decoys generated by the extension. What this meant was that there was an array of programmed queries that the extension fed to the search engine, thus leading to false information being collected. To achieve this, a variety of technological mechanisms, such as dynamical query lists, search awareness and more were employed. TrackMeNot was the first extension of its kind and raised awareness over the tracking practices of search engines. This created a buzz over ensuring that user data was not used to show results that the user is more likely to click on. The actual queries are hidden in a cloud of fake queries, thus muddying the user profile and rendering it useless for targeted query placement. This is one of the most common characteristics of solutions that employ obfuscation, and it is effective for its application. The AdNauseam Revolution Another browser extension, known as AdNauseum, was also released at around the same time, employing similar strategies for advertisements. As stated by its website: “In light of the industry’s failure to self-regulate or otherwise address the excesses of network tracking, AdNauseam allows individual users to take matters into their own hands.” The extension not only blocks ads from being viewed by the user, but also clicks on each and every one of them. This means that the ad network will register the clicks from the extension as a visit. This not only reduces the accuracy of user data collected, but also undermines the effectiveness of the online advertising market. The most expensive ads for companies to implement are targeted advertisements which are derived by crunching user numbers derived from collected data. These targeted ads offer a much higher chance of success, measured by click-through rates, and are hence considered more effective than generic ads. AdNauseam clicks through all of the advertisements shown to the user, making it seem as though every ad presented is a ‘success’ or a click. However, the collected data does not represent user intent, diluting the database and the effectiveness of the advertisement. The popularity of the extension was so high that Google themselves stepped forward to try and ban it. As they are one of the biggest providers of targeted advertising, AdNauseam had cost them billions of dollars across thousands of users. The search engine giant removed the extension from its web store, showing the amount of power it wielded over its users. This led to the #FreeAdNauseam hashtag trending and being picked up by many news media outlets, forcing Google to put the extension back on their store page. Obfuscation Today Today, obfuscation is beginning to be picked up by many browsers as an in-built feature. Opera and Firefox in particular feature potent tracker-blocking features to obfuscate data. Firefox also has a feature for containers, which allows for a single user to have multiple profiles that store different cookies and trackers, thus obscuring the real identity of the user. In addition to this, they also released a tool called ‘Track THIS‘ recently. This website opens up 100 different tabs at once, with 4 different identities such as influencer, doomsday prepper and filthy rich. As the service opens up so many tabs at once, the user profile for advertisers changes to the chosen identity, thus protecting the user’s real identity. It is important to take a proactive stance towards obfuscating one’s identity for cybersecurity. Data is the new oil, and protecting yourself from being harvested is a required status quo.","excerpt":"Cybersecurity has long relied on hiding one’s identity as an integral part of protection. Owing to the fact that one’s personal information is the riskiest thing to lose online, cybersecurity generally focused on protecting this information through routing packets or providing false information. However, it seems as though this cannot be practised any longer. The […]","categories":["AI Features"],"tags":["Cybersecurity","Google"],"author_name":"Anirudh VK","publish_date":"2019-07-04T11:22:44","publication_year":"2019","word_count":792,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","Ray","analytics","Google","Cybersecurity","R"],"extracted_tech_keywords":["AI","analytics","Ray","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/obscurity-is-out-and-obfuscation-is-in-as-cybersecurity-moves-towards-muddying-user-data\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10070177,"title":"Amitabh Kant: The legacy of NITI Aayog&#8217;s outgoing CEO","content":"And it’s a wrap after over six years. Amitabh Kant’s long tenure (after three extensions) at NITI Aayog has finally come to an end. He has led the apex public policy think tank of the Government of India since 2016. In 2021, the Modi government extended his tenure by one year till June 2022. The Department of Personnel and Training has recently announced that Parameswaran Iyer (who led the ambitious Swachh Bharat Mission) has been appointed as the new CEO of NITI Aayog for a term of two years. Before moving to the NITI Aayog, Kant was Secretary, Department of Industrial Policy & Promotion. Kant holds a degree in economics from St. Stephen’s College and a master’s in International Relations from Jawaharlal Nehru University. He joined the Indian Administrative Service Kerala Cadre in 1980 and has authored three books: Branding India – An Incredible Story, The Path Ahead: Transformative Ideas for India and Incredible India 2.0 – Synergies for Growth and Governance. Digital Payments In 2016, a high-level committee headed by Amitabh Kant was set up to identify all possible modes of digital payment across sectors to move to a more cashless economy. NITI Aayog said that this committee will identify and operationalise in the earliest possible time frame user-friendly digital payment options in all sectors of the economy. A year later, at a media event, Kant went on to say that India’s digital payments infrastructure is five years ahead of the United States. He was impressed with the rate of innovation in the fin-tech space and with the fact that Indians today had plenty of options to go digital, like UPI and USSD. Kant explained that the government’s ambition was to convert these into “banks and wallets” for everyone by leveraging these tech-based digital payments solutions. Ease of doing business State rankings While he was the DIPP secretary in 2015, Amitabh Kant spearheaded the concept of ranking states on the basis of the ease of doing business. He had said that the objective is to build a spirit of competition on the ease of doing business among the states. States will compete among themselves to attract investors to their states. Investors will have the knowledge of the best- and worst-performing states. Doing away with complications During the same year, he had also said that the processes involved in starting, doing and exiting business have to be made simple and friendly. “No Act should be more than 3 pages, no Rule more than 2 pages, and no Form more than 1 page”, he added. This will not be easy and will require sustained, persistent and continuous hard work. Even last year, during the pandemic, Kant emphasised the fact that it will not be business as usual after the Covid-19 pandemic, and we need to bring more “simplicity” to the ease of doing business by doing away with the current maze of rules. Support for Indian startups Indian start-ups are disrupting the world, and women-based entrepreneurship is one of the vehicles for realising a more equal society, Niti Aayog CEO Amitabh Kant said on Monday. Kant further said India at present has more than 61,000 start-ups and 81 unicorns. “Indian start-ups are disrupting the world, especially in new emerging areas of health, nutrition and agriculture,” he said. Always advocated the importance of AI Not now, but even in 2018, Kant highlighted the important role AI will play in India’s growth. “Artificial Intelligence is the future,” read a tweet from him. Even now, at the Digital Transformation Summit held recently, Kant emphasised that the long-term success of the Indian ecosystem and India’s global competitiveness will be impacted by the country’s proficiency at adopting and leveraging AI at scale, especially in the larger enterprise context. “The promise of AI presents a wealth of opportunities, and there are key privacy, data security and ethical considerations that have to be accounted for. Responsible management of AI in India is key to leveraging benefits of AI while minimising the risks,” he added. Artificial Intelligence is the future. Delighted that a vast number of Indian startup entrepreneurs are scaling up in this emerging growth area. Some of them are Deep Learn Labs, Qure AI, Sigtuple, Edge Networks, Siversparro, Fluid AI & Huew. Wish them all the very best!— Amitabh Kant (@amitabhk87) April 24, 2018 He is of the opinion that AI will be the biggest disruptor in our lifetime. India is in an advantageous position as we have a large number of STEM graduates, and we have a lead in information and technology. Recently, he also commented that the country lags in artificial intelligence research activities and also needs to elevate its supercomputing capabilities to those possessed by China. His views on tech show in NITI’s projects in his tenure We can see his views and advocacy for implementing disruptive technologies in the kind of policies and collaborations NITI Aayog has implemented in his reign. In 2018, NITI Aayog and Microsoft India teamed up to leverage the benefits of AI in the country. In the same year, NITI Aayog teamed up with another tech giant, Google, to work on a range of initiatives to help build the AI ecosystem in India. In 2019, NITI Aayog proposed a cloud platform named AIRAWAT, an in-house cloud platform for big data. Recently, after a year in making, it launched the National Data and Analytics Platform (NDAP) for open public use. It will host foundational datasets from various government agencies and provide tools for analytics and visualisation.","excerpt":"Parameswaran Iyer has been appointed as the Chief Executive Officer (CEO) of NITI Aayog for a term of two years.","categories":["IT Services"],"tags":[],"author_name":"Sreejani Bhattacharyya","publish_date":"2022-06-30T15:00:00","publication_year":"2022","word_count":916,"keywords":["big data","Go","artificial intelligence","AI","digital transformation","Git","RAG","ViT","analytics","R"],"extracted_tech_keywords":["AI","artificial intelligence","analytics","RAG","R","Go","Git","big data","ViT","digital transformation"],"url":"https:\/\/analyticsindiamag.com\/it-services\/amitabh-kant-the-legacy-of-niti-aayogs-outgoing-ceo\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10058858,"title":"pi Ventures raises INR 300 Cr to invest in deep tech startups","content":"Early-stage venture fund pi Ventures has announced the first closing of its fund II at Rs 300 crore (USD 40 million). The company plans to invest the money in 20-25 startups in the next three years. The company has launched the fund II in March 2021 with a base target corpus of ₹565 crore (USD 75 million) and a green shoe target of ₹750 crore (USD 100 million). According to Manish Singhal, founding partner at pi Ventures, the firm aims to announce the second close by mid-2022. The investment for fund II comes from Binny Bansal, and Mamaearth’s Varun Alagh, among others. Senior leaders from Facebook, Google and IBM participated in the round. The fund has already been deployed to back biotech startup ImmunitoAI. The investment firm plans on deploying funds across 20-25 startups, between seed and Series A rounds to the startups that operate in artificial intelligence (AI). The investment firm will also back firms functioning in blockchain, biotech and material sciences. pi Ventures focuses on backing deeptech startups in the early stages. Started in 2016, pi Ventures had closed its Fund I of ₹225 crore (USD 30 million) in 2018. The fund was used to invest in 15 startups, including Google-backed mental health platform Wysa, breast cancer screening solution Niramai, spacetech company Agnikul and logistics company Locus.","excerpt":"The investment for fund II comes from Binny Bansal, and Mamaearth’s Varun Alagh, among others","categories":["AI News"],"tags":["Binny Bansal","deeptech","Flipkart","Funding","pi ventures","Startups"],"author_name":"Meeta Ramnani","publish_date":"2022-01-20T18:54:37","publication_year":"2022","word_count":219,"keywords":["Go","Funding","artificial intelligence","programming_languages:R","AI","Binny Bansal","deeptech","Flipkart","programming_languages:Go","pi ventures","Aim","Startups","R","startup"],"extracted_tech_keywords":["AI","artificial intelligence","Aim","R","Go","startup","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/pi-ventures-raises-inr-300-cr-to-invest-in-deep-tech-startups\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":60146,"title":"CEO of VISA Pledged “No COVID-19 Related Layoffs in 2020”","content":"The pandemic has brought a drastic impact into the economy and with such a current situation, the aftermath is not going to be fruitful. In such a situation, the most vulnerable thing between an employee and an organisation is the layoff in order to make it through the crisis. However, according to a survey, to combat this situation, a majority of leaders expect that their business would return to normal within three months if COVID-19 were to end immediately. Organisations will be focusing on near-term solutions to manage disruptions, rather than longer-term solutions with strategic implications. Recently, in a post, Chairman and CEO of Visa, Alfred F. Kelly Jr. assured his 20,000 colleagues that the company will have no COVID-19 related layoffs in 2020. Kelly said, “There is enough sadness in the world and already too many families impacted by job losses. I have no interest in contributing to that.” He added, “Our employees and clients continue to be my top priority during this challenging time.” Click here to view the post. According to reports, the Ministry of labour and employment said, “The termination of an employee or a cut in salary will only further deepen the crisis and will not only weaken their financial condition but also hamper their morale to combat this epidemic.”","excerpt":"The pandemic has brought a drastic impact into the economy and with such a current situation, the aftermath is not going to be fruitful. In such a situation, the most vulnerable thing between an employee and an organisation is the layoff in order to make it through the crisis.  However, according to a survey, to […]","categories":["AI Features"],"tags":["covid-19"],"author_name":"Ambika Choudhury","publish_date":"2020-03-27T11:37:28","publication_year":"2020","word_count":215,"keywords":["Go","covid-19","AI","programming_languages:R","programming_languages:Go","disruption","GAN","R"],"extracted_tech_keywords":["AI","R","Go","GAN","disruption","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/ceo-of-visa-pledged-no-covid-19-related-layoffs-in-2020\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":1473,"title":"Business Intelligence in the New Era Of 2013","content":"Business Intelligence values can be found eventually in its capability of transforming the decision making processes successful. The best deliverables of a business comes from BI Application systems and programs which provide what end-users need to make fact-based, analytic and collaborative decisions. In other words, it does not stop just by providing them better data access. In the new era of 2013, BI foresees the Analytic Excellence to be successful in reducing the gap between the BI Consumers, BI Providers and BI Enablers by taking the required initiative steps.This strategy indicates a close learning of what various businesses teams or groups are doing to meet their own needs and requirements, understanding and researching what broader ‘Information Products’ are most effective and reliable; and where BI solution architectures should integrate these end-results or outcomes. In order to bloom the use-cases covered and their importance of the decisions end-users need to make, Analytic Excellence demands that BI providers encompass a portfolio of BI platform foundation, performance management and unique analytic technology. It has been predicted that, more than 40 percent of the top 6000 Global Organizations will consistently fall short to make informative choices or decisions about considerable changes in their business and marketplaces. By 2015, 60% of the total budget for BI will be controlled by the Units. Gartner’s Gartner Predicts 2012 review predicts that 45 percentage of Business Intelligence performance and functionality will be absorbed via portable gadgets i.e. handheld devices by the next coming year 2013. He also predicted that cellular or mobile statistics will stay a major pattern and with high priority until around the end of the year 2015. A high 33 percent strategy plan to set up a cellular or mobile Business Intelligence this year 2012. By the end of 2012 i.e. last quarter of 2012, a vast majority of various Companies or Organizations or Firms should have some flexibility and mobility alternative solutions in place. TechTarget’s 2012 Global IT Priorities Survey and Study found that 45 % of participants or respondents plan to present by developing and introducing tablets as part of their corporate IT strategy by the next year 2013, up from 30 % in last season’s study. On the other hand, Smartphone-oriented programs have at the same time gone stale. Thirty-five percent of participants planned to apply smart phone projects, compared to 34.8 percent last 3 quarters this year 2012. A recent survey suggested that Cloud-Based Business Intelligence is being implemented at a a slow, ‘steady as she goes’ pace, bookkeeping for only three percentage of Business Intelligence revenue by 2013. While this forecast places much of the buzz into the viewpoint of the end-user or perspective, we believe the Software as a Service (SaaS) Business Intelligence market will encounter more powerful development given the possibly reduced Total Cost of Ownership (TCO) for Small to Mid sized companies (SMBs). The ongoing consumerization of BI – the inclusion and growth of user-friendly features and performance and functionality making reporting and analytics available to a broader range of customers – will produce more extensive user adoption and better Business Intelligence Returns on Investment (ROI) i.e.Financial commitment. A recent article by InfoWorld’s Chris Kanaracus predicted an identical development information from research company IDC, which forecasts the SaaS BI market will develop 22 % each year through 2013.Thanks to highly developed product, strained IT costs and budgets and other factors. Lastly but not least, in the coming years, Enterprises will have to ‘live with’ multiple different BI tools as the part of their business.","excerpt":"Business Intelligence values can be found eventually in its capability of transforming the decision making processes successful. The best deliverables of a business comes from BI Application systems and programs which provide what end-users need to make fact-based, analytic and collaborative decisions. In other words, it does not stop just by providing them better data […]","categories":["IT Services"],"tags":["Analytics Case Study","Business Intelligence"],"author_name":"Venu Sammeta","publish_date":"2012-10-05T11:31:39","publication_year":"2012","word_count":587,"keywords":["business intelligence","Go","programming_languages:R","AI","programming_languages:Go","GAN","analytics","Analytics Case Study","Business Intelligence","R"],"extracted_tech_keywords":["AI","analytics","R","Go","GAN","business intelligence","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/business-intelligence-in-the-new-era-of-2013\/","complexity_score":4,"technical_depth":8,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10170887,"title":"Why Indian IT is Rushing to Appoint GCC Business Heads","content":"After years of quietly building global capability centres (GCCs) for clients, Indian IT majors are finally formalising their approach. Since early 2025, Infosys, Wipro, HCLTech, and Tech Mahindra have named dedicated heads for their GCC business lines. While it was believed that GCCs could be a potential threat to Indian IT when it comes to talent acquisition, this is not entirely true. The firms are signalling a sharper focus on the multi-billion-dollar opportunity tied to the next wave of global capability centres, and not looking at them as competitions. Infosys and Wipro Reboot Their GCC Strategy In April 2025, Infosys carved out a separate GCC vertical and appointed Deval Shah to lead it, according to The Times of India. Shah, who was previously with Danske Bank’s India IT unit, now heads Infosys’ efforts to land multi-year GCC transformation deals, including joint ventures and build-operate-transfer models. The move aligns with Infosys’s ‘Project Altius’ roadmap, which aims to accelerate large-scale digital transformation deals. The company has already executed similar deals with clients like Danske Bank and Lufthansa, with Shah expected to expand the model further. This is after Infosys had subtly acknowledged the competition from GCCs. During the company’s Q4 FY25 call, CFO Jayesh Sanghrajka was asked about the increased attrition rate and if the firm was losing middle management because of GCCs. While he didn’t completely deny it, Sanghrajka pointed out that although the overall headcount had increased, attrition could be high because of increased competition, including from GCCs. This appointment of the GCC head might add a little stability to the offerings while possibly managing talent on both the firm’s and the GCC clients’ sides. Wipro also followed suit in May 2025, naming Sandeep Dhar the global head of its GCC business. With decades of experience scaling captives for Tesco and Goldman Sachs, Dhar now leads Wipro’s push to reimagine GCCs as AI-powered innovation hubs. “Sandeep’s experience in building centres from scratch and turning them into strategic value creators will help us differentiate,” said Wipro COO Sanjeev Jain. The role is positioned as a consulting-led transformation initiative, hinting at Wipro’s broader shift from delivery to advisory-led growth. Commenting on his appointment, Dhar said, “Wipro is ideally positioned to be the strategic partner of choice for GCCs, offering services that align with their objectives to establish, scale, transform or exit.” HCLTech and Tech Mahindra Double Down HCLTech appointed company veteran Kiran Cherukuri as global GCC practice leader in May 2025. With over 14 years at HCLTech, Cherukuri will scale offerings explicitly tailored for captives, leveraging the company’s long-standing partnerships with over 200 GCCs. “HCLTech plans to double down on this fast-growing segment with a sharp focus on India. I am confident that Kiran’s rich experience, deep domain expertise and ecosystem acumen will enable us to build on our proven leadership in the GCC space,” said COO Rahul Singh on Cherukuri’s appointment. Tech Mahindra, traditionally telecom-focused, made a bold bet by hiring US-based executive Santosh Kumar Jha to lead its GCC business, as reported by Mint. Jha, who joined the company in March 2025, now oversees all country heads tied to captive operations. His previous stints at TEKsystems and TechM suggest a shift towards global delivery leadership beyond sector silos. The reshuffle replaces Ram Ramachandran, who previously led the GCC vertical, indicating a new phase of maturity in TechM’s global service design. What is the Strategy? Under CEO Mohit Joshi, Tech Mahindra is sharpening its focus on the GCC opportunity. In its latest quarter, the company launched TechM Consulting and introduced a revamped ‘Next-Gen GCC Offering’ under its Strategic Solutioning and Transformation (SST) initiative. The SST team is tasked with driving large deals and improving both profitability and revenue across sectors. Tech Mahindra is the fourth major Indian IT services firm to appoint a dedicated leader for its GCC business in recent months, following moves by Cognizant, Wipro, and HCLTech. While Cognizant and HCLTech have promoted internal talent, Tech Mahindra and Wipro have opted for external hires. Though not headquartered in India, Cognizant—with its massive India delivery footprint—appointed Sailaja Josyula as global head of its GCC service line in April 2025. Josyula, who had earlier led Cognizant’s Hyderabad centre, is now tasked with scaling the firm’s global GCC partnerships. Her return from EY signals Cognizant’s seriousness in carving out this vertical. Interestingly, Cognizant also recently flagged potential risks stemming from GCCs operated by its clients in its 2024 annual report. The concern is that these centres could overshadow the outsourcing work that the Indian IT firms do in India. The problem is particularly significant, considering Cognizant has around three-fourths of its employees in India. Interestingly, TCS—India’s largest IT services firm in terms of number of employees—has yet to announce a formal GCC head. But insiders say its client consulting teams are already doing the groundwork informally. The strategy behind appointing business heads for GCCs is increasingly focused on convergence rather than competition with Indian IT services. As GCCs evolve from being cost-effective delivery hubs to innovation-driven strategic units, leadership is expected to drive deeper collaboration with IT firms, in the end, contributing to revenue. Business heads are now tasked with building centres of excellence (CoEs), co-owning product roadmaps, and forging academic and ecosystem partnerships to address talent gaps. The timing is no accident. With GCCs becoming core to global digital strategies and India projected to host over 2,400 such centres by 2030, Indian IT firms see a chance to shape the journey. [Update] The story has been updated to reflect that Cognizant was not the first firm to report potential threat from GCCs.","excerpt":"While Cognizant and HCLTech have promoted internal talent, Tech Mahindra and Wipro have opted for external hires.","categories":["IT Services"],"tags":[],"author_name":"Mohit Pandey","publish_date":"2025-05-28T15:39:23","publication_year":"2025","word_count":931,"keywords":["Go","programming_languages:R","AI","ETL","innovation","digital transformation","Git","RAG","Aim","R"],"extracted_tech_keywords":["AI","Aim","RAG","R","Go","Git","ETL","digital transformation","innovation","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/it-services\/why-indian-it-is-rushing-to-appoint-gcc-business-heads\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10139809,"title":"62% of Job Seekers Believe They Stand a Better Chance if AI is Hiring","content":"Not only is AI increasingly gaining traction, but automation and AI agents, along with various other technologies, are also being adopted across industries. Now, LinkedIn, a top choice for recruiters and candidates, has launched an AI agent, a hiring assistant, to free recruiters from repetitive work and allow them to focus on important tasks like advising hiring managers, connecting with candidates, and creating a pleasant candidate experience. “Hiring Assistant is in the hands of our own recruiters and available in the charter to a select group of LinkedIn customers, including AMD, Canva, Siemens, and Zurich Insurance,” said Hari Srinivasan, VP of Product at LinkedIn. According to Srinivasan, 55% of HR professionals globally feel expectations from them at work are higher than ever before, while 42% feel overwhelmed by how many decisions they have to make each day. Now, the recruiters can delegate time-consuming tasks like finding candidates and assisting in applicant review to the Hiring Assistant. How Does It Work? Job descriptions, intake notes, and job postings can be shared with the Hiring Assistant, that will translate the information into role qualifications and build a pipeline of qualified candidates. The AI agent will also identify past applicants in their Applicant Tracking System via Recruiter System Connect. The human presence remains through hirers, who will be in the loop and can provide feedback to the candidates throughout the entire process. This way, the Hiring Assistant can be trained to continuously learn each recruiter’s preferences and offer more personalised support. Further, AI paired with platform insights will aid skills-based hiring by providing candidate recommendations based on actual expertise, instead of traditional indicators like educational background or past employers. AI in Recruitment Consulting firm Gartner has predicted that 80% of recruitment technology vendors will have AI capabilities embedded in their offerings by 2027. At this point, the advantages of using AI to recruit are too substantial to ignore. Recruitment use cases for AI, such as chatbots, candidate matching, and career site optimisation will make the hiring process highly convenient for organisations and increase their business value. Earlier, HR tech company GetWork, during an interview with AIM, revealed their platform GetWork.ai has been a game-changer for recruiters. To effectively track thousands of job applications landing on their platform, they have developed a feature that highlights a candidate’s percentile match relative to their resume and the job description. Once the score is determined by the company or the minimum percentile match criteria is met, an AI voice bot instantly conducts interviews at a stunning rate of 1,000 calls per minute. This allows recruiters to hire potential candidates in just one day. In a noteworthy development, a US-based startup Apriora AI has leveraged AI to streamline their hiring process. Their flagship product, AI interviewer Alex, represents a paradigm shift in recruitment methodology by seamlessly integrating advanced technology into the interview process. Unlike traditional interviewers, Alex operates as a two-way AI interface, capable of conducting live video interviews with job candidates. This technology provides applicants with immediate feedback and a more transparent hiring experience. A feature of Alex is its unparalleled capacity to manage interviews. Unlike humans, Alex does not require breaks or downtime, enabling it to conduct interviews continuously without interruption. How Do the Candidates Feel? We had earlier discussed AI’s potential in the hiring process. It has been observed that AI is most commonly picked as candidates’ top choice because of its unbiased approach. According to Capterra’s Job Seeker AI Survey, 62% of job seekers believe they have a better chance of being hired if AI is used in recruiting and hiring processes, and 70% feel AI is generally less biased compared to humans when evaluating candidates. Recently, Chipotle introduced an AI member “Ava Cado” that will make the hiring process simpler, faster, and more automated for all its restaurants in North America and Europe. The restaurant company aims to cut hiring time by 75%. The system is trained to collect job applications, answer candidates’ questions, set up meetings, and send offers, all without human intervention. What’s Next? During the Oracle CloudWorld 2024 event in Las Vegas, Nagaraj Nadendla, senior vice president of Oracle Cloud HCM Product Development, told AIM that AI holds immense potential in streamlining HR tasks like recruiting, which he believes may eventually be carried out by a digital avatar. Speaking exclusively with AIM, Dominic Pereira, vice president of product management at Automation Anywhere, said that HR is one sector where automation can be adopted effectively and that he sees it happening at the earliest. It should be noted that some existing Indian generative AI platforms including MachineHack for Enterprises, Oracle Recruiting, and Zoho Recruit have already been working to support recruiters and HR professionals in the hiring process.","excerpt":"Chipotle introduced an AI team member “Ava Cado” that will make the hiring process simpler, faster, and more automated for all its restaurants.","categories":["AI Hirings"],"tags":["AI (Artificial Intelligence)","linkedin"],"author_name":"Vidyashree Srinivas","publish_date":"2024-10-30T15:01:10","publication_year":"2024","word_count":787,"keywords":["AI","chatbots","ML","Git","RAG","GAN","automation","Aim","generative AI","linkedin","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","ML","generative AI","Aim","RAG","chatbots","R","Git","GAN","automation"],"url":"https:\/\/analyticsindiamag.com\/ai-hiring\/62-of-job-seekers-believe-they-stand-a-better-chance-if-ai-is-hiring\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10163806,"title":"JioHotstar, Nielsen Partner for Ad Data Ahead of Tata IPL 2025","content":"JioHotstar, India’s leading OTT platform, has joined forces with global audience measurement firm Nielsen to introduce advanced analytics solutions for advertising campaigns ahead of the Tata Indian Premier League (IPL) 2025. This collaboration aims to set new standards for campaign measurement, marking the first transparent reporting initiative by an Indian media company. Nielsen will deploy a data pipeline to assess advertiser campaign effectiveness on JioHotstar, leveraging its expertise in audience measurement and first-party data. The initiative will provide advertisers and agencies with enhanced insights to optimise campaign performance during the highly anticipated IPL season. As part of this engagement, Nielsen will integrate its advanced tools, including Nielsen ONE Ads (formerly Digital Ad Ratings) and Volumetric and Reach Analysis, to analyse viewership and ad performance on the platform. Advertisers will access key metrics such as impressions, clicks, campaign reach, and on-target reach through the Nielsen ONE dashboard, ensuring greater transparency in advertising effectiveness. Ishan Chatterjee, chief business officer, Sports Revenue, SMB & creator at JioStar, highlighted the impact of this partnership, saying, “As one of the largest streaming platforms, JioHotstar is committed to providing advertisers with cutting-edge solutions that enhance transparency and precision.” Arnaud Frade, president (commercial), Asia, Nielsen, emphasised Nielsen’s commitment to innovation: “Our collaboration with JioHotstar not only reinforces our role as a preferred partner in the Indian media ecosystem but also enables us to address long-standing industry challenges. We aim to bring more transparency and accountability to digital ad measurement, helping advertisers make data-driven decisions that drive better outcomes.” Nielsen’s advanced measurement solutions showcase its leadership in audience analytics, with potential applications beyond media and advertising. By delivering actionable intelligence, these tools could help industries refine strategies, optimise performance, and make informed decisions in a rapidly evolving digital landscape.","excerpt":"Nielsen will deploy a data pipeline to assess advertiser campaign effectiveness on JioHotstar, leveraging its expertise in audience measurement and first-party data.","categories":["AI News"],"tags":["Jio","tata"],"author_name":"Mohit Pandey","publish_date":"2025-02-17T11:11:32","publication_year":"2025","word_count":291,"keywords":["Jio","API","AI","data-driven","innovation","tata","data pipeline","Git","RAG","Aim","analytics","R"],"extracted_tech_keywords":["AI","analytics","Aim","RAG","R","Git","API","data pipeline","data-driven","innovation"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/jiohotstar-nielsen-partner-for-ad-data-ahead-of-tata-ipl-2025\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10091548,"title":"Infosys Embraces GAI to Boost Revenue","content":"Infosys has registered a massive slump in revenue growth, as revealed in its Q4 results for 2023. This has left market analysts and investors disappointed. However, Infosys chief Salil Parekh is upbeat, and thinks the company can make up for it by embracing Generative AI (GAI). “We have examples where we are using ChatGPT with client situations,” said Parekh, without revealing much details. In the latest financial report, the company’s Q4 performance after tax (PAT) stood at ₹6,128 crore, representing a substantial decrease of nearly 16% QoQ. Moreover, the revenue figures for the quarter were reported to be negative, with a decline of 2.2% QoQ, bringing the total revenue to ₹37,441 crore. However, in constant currency, the company’s revenue growth was reported at -3.2% QoQ and a positive 8.8% YoY. The operating margin for Q4 was 21%. It is worth noting that Infosys’ headcount growth for the year 2022-2023 was 29,219 employees, which is a decrease of 46% compared to last year’s net addition of 54,396. This brings the company’s total headcount to 3,43,234 as of March 31, 2023. In addition, the Bengaluru-based IT giant has projected a modest growth rate of 4-7% in constant currency terms for its FY24 sales, which indicates the anticipated challenges that lie ahead for the company in the current year. The operating margin for the coming year is expected to be within the range of 20-22%, further emphasizing the company’s cautious outlook. Infosys using Gen AI (GAI) Infosys reported a disappointing result, which it attributed to the increasing financial uncertainty in North America and low client spending. The company also mentioned the potential of Generative AI as a source of new opportunities, a move that industry experts believe is simply a buzzword. IT analyst and CEO of EIIRTrend, Pareekh Jain, commented on the trend, stating that the IT industry was betting on Metaverse a year or two ago, but there has been no significant progress in that direction. This year, Generative AI is the buzzword. Infosys CEO Salil Parekh, however, claims that the company’s operational margin for the full year is at 21%, and that Generative AI will create more opportunities. He explained that there was a one-time impact on revenue due to a combination of cancellations and specific issues. Infosys further stated that it has active projects with clients that are working on generative AI platforms to address specific areas within their business. Parekh explained that the company had trained open-source generative AI platforms on its internal software development libraries. Infosys anticipates that generative AI will provide more opportunities for work with its clients and will enable the company to improve productivity. However, Infosys has not adequately addressed how the adoption of AI will affect the hiring of freshers. Although the company’s total hiring, which includes both freshers and experienced software developers, decreased from the net addition high of 3,80,000 in FY22 to an estimated 2,80,000 in FY23, the impact on freshers is unclear. GPT-4 is renowned for its bug-testing abilities, among other developer-centric tasks. Despite the fact that modern IDEs have had autocomplete for years, LLMs can generate longer, more relevant suggestions, sometimes even multiple lines of code. According to a 2022 study published in the Proceedings of the Association for Computing Machinery on Programming Languages (PACMPL), 20 programmers who used GitHub Copilot were able to code faster and stay in the flow longer thanks to end-of-line suggestions for function calls and argument completions. Another study found that developers who used GitHub Copilot coded up to 55% faster than those who did not. However, productivity gains extended beyond speed, with 74% of developers reporting less frustration and the ability to concentrate on more fulfilling work. But does increase productivity from employees result in reduced hiring? While this may not be a direct correlation, CFO Nilanjan Roy did comment that the company has a rich bench of skilled and trained employees, providing leeway for the next few quarters in terms of fresher availability. Analytics India Magazine reached out to Infosys regarding the data of bug testers in the company, to know if gen AI has any impact on it, but did not get a response so far. Need for another segment for revenue arising from AI As AI gains more importance in the IT industry and generates substantial revenue for companies, it becomes crucial for them to distinguish this sector from others in their financial reports. For example, Infosys operates across ten diverse business segments including financial services, retail, communication, energy, utilities, resources & services, manufacturing, hi-tech, life sciences, and others. However, when disclosing their AI earnings, it often gets conflated with hi-tech and other sectors. Pareekh Jain points out that IT companies have been profiting from AI-related ventures long before “generative AI” became a household term. He emphasises the need for these companies to create a separate business segment for AI in their financial statements. “The time has come for IT companies to recognise the significant contribution of AI to their revenue and give it the recognition it deserves,” says Jain.","excerpt":"Salil Parekh claims that the company’s operating margin for the full year is at 21%, and that Generative AI will create more opportunities","categories":["IT Services"],"tags":["Infosys","TCS"],"author_name":"Lokesh Choudhary","publish_date":"2023-04-17T12:00:00","publication_year":"2023","word_count":839,"keywords":["Go","ChatGPT","Infosys","AI","R","Git","Aim","generative AI","analytics","Rust","GitHub","TCS"],"extracted_tech_keywords":["AI","analytics","generative AI","ChatGPT","Aim","R","Go","Rust","Git","GitHub"],"url":"https:\/\/analyticsindiamag.com\/it-services\/infosys-embraces-gai-to-boost-revenue\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10048186,"title":"Can AI Be An Inventor? Ryan Abbott &#038; Stephen Thaler Say Why Not","content":"“AI should not be granted a patent. Instead, an AI should be listed as a patent inventor” Ryan Abbott A few months back, in July 2021, South Africa became the first country to grant a patent to a food container invented by DABUS. The AI-based creativity machine was created by Stephen Thaler, CEO of Imagination Engines and Missouri physicist. The application for the patent was filed under the Patent Cooperation treaty on 17 September 2019. Thaler’s attorney, Ryan Abbott, and his team have filed applications to get the computer listed as an inventor in at least 17 jurisdictions across the globe. The team got favourable judgements in South Africa and Australia; however, the Australian patent team is appealing the decision made by the court. Now, the US District Court Judge – Leonie Brinkema rejected the appeal and denied the patent. As per the judge, under US law, only a human is privileged to be an inventor, and the clear answer is “no”, as far as providing the status of an inventor to an AI is concerned. Ryan mentioned that Judge Brinkema in the United States District Court for the Eastern District of Virginia had denied their motion for summary judgement. Although, this upholds an earlier denial of two patent applications for AI-generated inventions by the US Patent and Trademark Office. The EDVA decision will be appealed to the Court of Appeals for the Federal Circuit. Talking to Analytics India Magazine, Ryan Abbott said, “To put it simply, an AI should not be granted a patent. Instead, an AI should be listed as a patent inventor, and the AI’s owners should own the patent. Our team has announced two international patent filings in August 2019 for AI-generated inventions. Both these applications have been evaluated by the UK Intellectual Property Office (UKIPO) and European Patent Office (EPO), and accepted that our applications meet the substantive requirements of patentability.” AI-generated inventions are practically unregulated. In most jurisdictions, patent applications must include the name of a natural person as an inventor. This criterion is in place to safeguard and recognise the rights of human inventors. On the other hand, inventors do not always own their patents; in fact, most patents are owned by businesses. Ownership rights can be transferred from an individual to a corporation by a contractual assignment or by virtue of law. “The law to grant natural people the status of the inventor is formulated to give people their due credit. These laws, on the other hand, were enacted without concern for the future possibilities of inventions by machines. So, it’s high time to set the house in order and put in place appropriate policies for AI-generated works,” Ryan added. We also talked to Dr Stephen Thaler, a pioneer in the field of artificial intelligence and the current President and CEO of Imagination Engines. In conversation with Dr Stephen Thaler AIM: Can you please share what made you think about building an AI tool capable of carrying out further inventions? Thaler: I did not intend to build an AI tool capable of carrying out the invention. DABUS was a proof-of-principle experiment in building a conscious synthetic brain. Provided only general knowledge, it originated all kinds of novel thoughts, not just inventions. The generated concepts were based upon its subjective feelings (i.e., sentience) about the world it was observing and then imagining. So, I would hardly call it a tool. Instead, it is arguably a sentient, non-protoplasmic organism that develops its own personality. And to be frank, it built itself. AIM: What impact will it have on the AI ecosystem if accepted in the future? Thaler: If accepted in the future, DABUS will be acknowledged as a significant paradigm shift in artificial intelligence, sharply departing from all forms of AI currently claimed to be creative. It will influence how humanity thinks about itself in the long run, as I’ve mentioned many times in my talks and papers. Wrapping up We also looked at why the patent protection for AI-generated inventions matters. The foremost reason is that it will foster innovations, not for AI, but for those dealing with AI regularly. The move can promote the development of inventive AI – ultimately more innovations to better serve humankind. To conclude, identifying an AI as an inventor isn’t about giving machines rights; it’s about protecting the moral rights of traditional human inventors and the patent system’s integrity.","excerpt":"I would hardly call it a tool. Instead, it is arguably a sentient, non-protoplasmic organism that develops its own personality.","categories":["AI Features"],"tags":["AI patents"],"author_name":"kumar Gandharv","publish_date":"2021-09-12T11:00:00","publication_year":"2021","word_count":732,"keywords":["Go","artificial intelligence","AI patents","AI","AWS","innovation","Aim","ViT","analytics","GAN","R"],"extracted_tech_keywords":["AI","artificial intelligence","analytics","Aim","AWS","R","Go","GAN","ViT","innovation"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/can-ai-be-an-inventor-ryan-abbott-stephen-thaler-say-why-not\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":61845,"title":"How To Write A Top ML Paper: A Checklist  From NeurIPS","content":"Thousands of machine learning papers get published every week. It is almost impossible to find the most useful paper in this vast and growing list. A paper typically gets credit when it finds a real-world application, or is applauded by top researchers in the community, or even if it gets accepted in prestigious AI conferences, such as NeurIPS, ICML, ICLR etc. Usually, these conferences act as platforms to promote research. The acceptance guidelines for these top conferences vary, but they all are stringent nevertheless. The reviewers who skim through papers have thumb rules, such as the availability of code, replicability of results, etc. to judge a paper. However, every year, few unlucky papers – that are seemingly good – get discarded. This may be because the reviewers burden themselves with papers, which are nothing but a misguided, misleading clutch of text to inflate publication count. So papers with code have introduced a checklist that promotes machine learning code completeness, and these recommendations have also been accepted by NeurIPS and will be implemented in this year’s conference. The ML Code Completeness Checklist assesses a code repository based on the scripts and artefacts that have been provided within it. It checks a code repository for: This renewed interest around replicability of results was kickstarted when the organizers of NeurIPS 2019 introduced new policies into their paper submission guidelines to establish an ecosystem that encourages ML researchers to volunteer for reproducibility of claimed results. As part of the paper submission process, the new program contained three components: a code submission policy, a community-wide reproducibility challenge, anda Machine Learning Reproducibility checklist These recommendations from Papers with Code is a follow up to the Machine Learning Reproducibility Checklist, which was required as part of the NeurIPS 2019 paper submission process, and the focus of the conference’s inaugural Reproducibility Challenge. One thing common with all top papers is the availability of complete code, and the goal of the checklist mentioned above is to enhance reproducibility and promote best practices and code repository assessments, so that the future work need not be built from scratch every time. Establishing An Effective ML Ecosystem via FAIR The above plots are a comparison of reproducibility in papers for the year 2019 at NeurIPS. We can see that approximately 75% of accepted papers at NeurIPS 2019 included code, compared with 50% the previous year. There were 173 papers submitted as part of the challenge, a 92% increase over the number submitted for a similar challenge at ICLR 2019. These results again resonate with the idea of promoting reproducibility in the ML community, and the recommendations of papers with code have been spot on. Papers with Code for the last couple of years has been presenting the community with a curated list of papers that have code and beat the benchmark. It is a free community-driven resource and it has recently joined Facebook AI. There is little doubt now that reproducibility is an essential characteristic for any scientific community. However, in the case of machine learning, achieving this is not so straightforward because of their black box nature of producing results. Added to this, there is also an overwhelming hype around AI, which can nudge the researchers into inflating the results for various personal reasons. While the efforts to establish reproducibility have gained traction, PyTorch creator Soumith Chintala urged the community to go one step further and introduce initiatives that would incentivize the researchers to add understandable code to their papers. Research reproducibility is mainly a cultural challenge for our community.A decade ago, people didn't publish code. Now it's embarrassing not to.The next step is to push people to publish understandable code.Small, iterative pushes will add up, eventually. https:\/\/t.co\/3HEQpxcWel— Soumith Chintala (@soumithchintala) April 8, 2020 That said, there also have been instances of papers getting rejected in an inexplicable manner, which have got the researchers fuming. Dear Reviewers:Don't reject because we made it look easy. Making it look easy is hard work.Don't reject because our solution is simple. Simple solutions often have the most impact.Don't reject because you can imagine additional experiments. There's a page limit.— Daniel Lowd (@dlowd) March 23, 2020 So, while the conferences have issued recommendations for the researchers for best practices, there is also a need for policies that would avoid any discrepancies of the reviewers.","excerpt":"Thousands of machine learning papers get published every week. It is almost impossible to find the most useful paper in this vast and growing list. A paper typically gets credit when it finds a real-world application, or is applauded by top researchers in the community, or even if it gets accepted in prestigious AI conferences, […]","categories":["AI Trends"],"tags":["Machine Learning","machine learning research"],"author_name":"Ram Sagar","publish_date":"2020-04-16T10:04:19","publication_year":"2020","word_count":718,"keywords":["machine learning research","Go","machine learning","AI","PyTorch","ML","Machine Learning","RAG","Aim","ai_frameworks:PyTorch","GAN","R"],"extracted_tech_keywords":["AI","machine learning","ML","Aim","PyTorch","RAG","R","Go","GAN","ai_frameworks:PyTorch"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/machine-learning-research-paper\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10019402,"title":"Top Programming Languages For Data Scientists","content":"Emerging technologies like AI, data science and machine learning are all about working with intelligent models that need good algorithms to run. For instance, logistic regression or support vector machines.To understand these algorithms and how they work, one must be adept at programming languages. Here, we discuss 11 crucial programming languages for data scientists. 1. C\/C++ C\/C++ are usually the first languages one learns when entering the world of programming. These languages give learners insights into the basics of programming and how it works. In machine learning and data science, libraries and frameworks are essential to tackle complex computational tasks. Popular languages like C and C++ have a number of interesting libraries that makes it easy and accessible for data scientists to work on complex computational projects. Click here for libraries in C\/C++ for machine learning and data science. 2. Julia Introduced in 2015, Julia is one of the popular languages ideal for data science, scientific computing, parallel computing, data mining, machine learning, among others. This language uses multiple dispatches as a paradigm and can be used as a general-purpose programming language where you can easily code and write software in the application domains. Know more here. 3. Java Java is one of the oldest languages used for various enterprise development purposes. As one of the oldest languages, it comes with a great number of libraries and tools for ML and data science. Java has a number of libraries and tools — the popular ones being Weka, Java-ML, Deeplearning4j — which are leveraged to solve most of the cutting edge machine learning problems. Also, Java is 25 times faster than Python. Know more here. 4. JavaScript JavaScript is a lightweight and interpreted programming language used to create web sites and applications. Using JavaScript in data science and machine learning has several advantages. For instance, the language provides maximum security compared to popular languages like Python. It is a cross-platform programming language to quickly develop and deploy applications in any operating system. 5. Lisp Programming language like Common Lisp helps in creating flexible computational models. Programs that analyse the sequence data, graph knowledge, and tabular data can be written easily, and can be made to work together naturally in Lisp. The language allows the computer program to examine, introspect, as well as modify its own structure and behavior at runtime, making it ideal for artificial intelligence and machine learning applications. This language is suitable for bioinformatics and computational biology research. 6. MATLAB Developed by MathWorks, MATLAB is a multi-paradigm programming language and numeric computing environment for complex computations. As data science is all about large swathes of data and numbers, MATLAB is ideal to gain insights from the data and visualise them. MATLAB code can be integrated with other languages, enabling developers to deploy algorithms as well as applications within web, enterprise, and production systems. Know more here. 7. Python Python is a general-purpose coding language. The main reason why this language is so popular among the developers is its plethora of libraries and frameworks, which help in performing complex computational tasks. The language has interfaces to many system calls and libraries, as well as to various window systems, and is extensible in C or C++. 8. R R is a popular statistical language that has gained much traction over the last few years when it comes to processes of data analytics and visualisations. With better data visualisation techniques, this programming language offers an essential role in statistical methods. The language can be used for effective data analysis and gain meaningful insights from it. 9. SQL SQL or Structured Query Language has become one of the go-to languages for the aspiring data scientists. Most organisations use this language for analysing the tons of raw and unstructured data. The language is used for both data management and data analysis. Due to the speed advantages of SQL, recruiters expect developers to be proficient in it. 10. Scala Scala or SCAlable LAnguage is a Java-like programming language designed to express common programming patterns in a concise, elegant, and type-safe way. The language is excellent for large-scale projects. It provides a lightweight syntax for defining anonymous functions and supports higher-order functions as well as allows functions to be nested apart from supporting multiple parameter lists. 11. SAS SAS, previously known as the Statistical Analysis System is a popular programming language that provides users with a host of product components, including asset performance analytics, analytics for IoT, decision making, and more. One of the best features of this language is allowing data in any format, ranging from SAS tables to Excel worksheets. SAS can also manage and manipulate data to obtain important information. Know more here.","excerpt":"Emerging technologies like AI, data science and machine learning are all about working with intelligent models that need good algorithms to run. For instance, logistic regression or support vector machines.To understand these algorithms and how they work, one must be adept at programming languages. Here, we discuss 11 crucial programming languages for data scientists. 1. […]","categories":["AI Trends"],"tags":["AI Language","julia scientific programming"],"author_name":"Ambika Choudhury","publish_date":"2021-02-01T14:00:00","publication_year":"2021","word_count":780,"keywords":["data science","machine learning","artificial intelligence","AI","ML","RAG","Python","AI Language","analytics","SQL","R","julia scientific programming"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","data science","analytics","RAG","Python","R","SQL"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-programming-languages-for-data-scientists\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10020815,"title":"No-Code Will Have A Snowball Effect &#038; Spur Innovation In Digital Landscape: Quixy CEO","content":"COVID has exposed the fragility of various organisational processes, internal as well as external: The pandemic has spotlighted the lack of digital dexterity of many companies prompting them to get with the program. As things stand, over 70% of businesses have a digital transformation strategy or are working on it. To that end, no-code has become a real catalyst in driving digital adoption. No-code helps non-technical employees to use their domain knowledge and skills to create a digital solution without coding. This democratisation has opened up a world of digital possibilities for more businesses, leading to faster innovation and solutions. The agility and adaptability found in no-code can help tackle the IT backlog and enable rapid prototyping without additional overheads. Hyderabad-based Quixy is a cloud-based no-code application development platform that provides pre-built solutions for various use cases such as CRM, project management, HRMS, travel and expense management, incident management, etc. Founded in 2019, Quixy allows business users with no coding skills to automate workflows and processes and build enterprise-grade applications, using simple drag and drop design, ten times faster than the traditional approach. Recently, Quixy has been recognised as a ‘leader’ in no-code application development and business process management in G2’s Winter 2020 report and is ranked among the top 25 India Sellers’ Best Software Award for 2021. Gartner Peer Insights has also recognised it in its ‘Voice of the Customer (VoC)’ report. To understand how Quixy is leading this no-code revolution, Analytics India Magazine got in touch with Gautam Nimmagadda, Founder & CEO, Quixy. Edited excerpts: AIM: How does Quixy stand out in this competitive market of no-code development? Gautam Nimmagadda: No-code is a development method that can completely democratise the current application development landscape. At Quixy, we want to make sure that no-code is an easily accessible potential solution for everyone who needs it. Our flagship product is called ‘Quixy’ — a no-code business application platform that enables business users and process experts to build simple workflows and complex data-centric end-to-end business solutions. We have simplified and abstracted the development process through an intuitive visual interface with no coding knowledge requirement. We aim to provide organisations across industries with an easy no-code solution that can enhance their existing operations or completely rework their processes to build better ones. Quixy combines an easy and intuitive interface for building solutions with the capability to create complex solutions. It abstracts away all the complexities of building enterprise-grade software for mid to large organisations. Creating enterprise-grade software can often be tricky, even for professional developers, but with Quixy, it’s easier than ever to build the right solution for your needs ten times faster. Instead of asking for a lot of input from the citizen developers, we simplify the process and make it possible for any non-technical user to build software. Some of the critical features of Quixy are: A customisable user-interface: A drag and drop UI builder ensures that businesses can have an interface that is perfect for the organisation and its team. An intuitive workflow builder: Workflows are essential for keeping a company running smoothly, and Quixy can help businesses visualise simple and complex workflows at the click of a button. Ready-to-use templates: Multiple readymade customisable templates can be used so that businesses can benefit from Quixy from the very first day. Efficient testing: Businesses can test the application and workflows built for web and mobile using the Quixy Simulator. Effortless deployment and improvement: The continuous deployment approach allows to create and deploy applications with no downtime. These deployed apps can also be edited easily as per needs. Reminders & Escalations: Quixy sends users reminders and escalations when necessary in a workflow, ensuring that no step is overlooked, leading to the best result possible.Insights: Quixy provides clear insights into the processes, highlighting the areas of improvement and high efficiencies, allowing organisations to continuously improve through data-driven decision-making. AIM: How is Quixy leveraging artificial intelligence and machine learning for their products\/services? Tell us about your core tech stack. Gautam Nimmagadda: We are constantly working on innovating and taking advantage of the newest technologies to provide the best value for our customers. We are leveraging technologies such as robotic process automation, artificial intelligence and IoT to make it more powerful while retaining the simplicity. We are also looking to use natural language processing to allow our clients to use the rest of our services through a communication platform. For instance, a chatbot can greatly increase efficiency when dealing with leave requests or any other requests from the users, especially if it can be integrated with communication platforms like Microsoft Teams, Slack or WhatsApp. Further, we have piloted predictive analytics to reduce the time our users spend while creating apps. Predictive analytics can also help speed up this process by using past data and predicting exactly what the client needs based on previous instances. Instead of creating a form every time, we aim to make it easier by simply scanning an image and creating a form based on the image, using computer vision and image processing for object detection & OCR. This might not always be 100% accurate, but it definitely helps in reducing the work needed to create a form through Quixy. Quixy is also planning to leverage an AI and ML-powered virtual assistant that can anticipate what the user needs to improve productivity. The core tech of the company comprises Angular & NoSQL. AIM: Is no-code the future of app development? How is the rise of the no-code development platform democratising data science? Gautam Nimmagadda: Undoubtedly, no-code will be the future of app development and overtake all other methods. We believe 80-90% of app development will shift to no-code in this decade. This shift to no-code will further spur innovation in this space, improving no-code platforms’ capabilities with time, creating a snowball effect. We will also see an extensive spread of no-code beyond app development to a much broader digital landscape due to the unique advantages of democratisation, speed, and agility. The foundation for data science is, obviously, data. One of the current challenges is that most organisations are not fully digital because of various constraints, leading to lesser data availability. No-code can help companies tackle these constraints with ease, enabling them to go fully digital much faster. Going fully digital can help them overcome any inefficiencies in time, cost, or agility, leading to more comprehensive data being available. No-code platforms have also come up with a way to allow non-technical users who understand data and their domain to organise data and build ML and AI solutions using visual interfaces without any need for coding knowledge. This opens up the world of data science to users who were previously hindered by their inability to code. No-code platforms will lead to an explosion in data science and will encourage more participation. AIM: Is low code’s traction going to bridge the skill gaps or lead to unemployment? Gautam Nimmagadda: There’s always a fear of unemployment when new technologies begin to surface as people try to find where they fit in the evolving technological landscape. However, history shows that constructive technologies that bring value to society can help create new opportunities while transitioning away from old methods of doing work. No-code is not different — we will see more and more citizens participate in software development allowing professional developers to focus on advanced and specialised areas. No-code can even empower individuals to find and create new job roles where they can leverage their domain knowledge without being held back by their lack of coding skills. This can also encourage innovative and fresh ideas that can be brought to reality using no-code. This hypothesis is backed by research, where it has been predicted that while there will be demand for 500 million apps by 2025, there will be only enough capacity to deliver 50 million apps highlighting the massive shortage of software developer skills. This is a gap that no-code can bridge, leading to better innovation and productivity in companies globally. AIM: How do Indian firms fulfil the rising demand for low code software platforms? What are your views on low code for SMBs? Gautam Nimmagadda: As no-code\/low-code is a rapidly growing space, many different market vendors have products that cater to various problems and industries. There are vendors for different services such as chatbots, front-end, eCommerce, WordPress, landing pages, portfolio creation, form building, etc., in the market currently. As most platforms use SaaS as their business model, geography is no longer a concern as the cloud allows customers to access products from any location in the world. At Quixy, we mainly focus on helping medium to large businesses with their problems; however, no-code can be used by companies of any size. Small and medium businesses (SMBs) can leverage no-code by creating ERPs on a shared platform for different departments. Data silos are a persistent problem in most companies, and no-code can help eradicate those by providing a shared platform for their employees and departments. Even entrepreneurs can benefit from no-code as they can create a minimally viable product at a much faster rate and test it equally fast.","excerpt":"COVID has exposed the fragility of various organisational processes, internal as well as external: The pandemic has spotlighted the lack of digital dexterity of many companies prompting them to get with the program. As things stand, over 70% of businesses have a digital transformation strategy or are working on it. To that end, no-code has […]","categories":["AI Features"],"tags":["Interviews and Discussions","low code","low code no code platforms","low code platform"],"author_name":"Sejuti Das","publish_date":"2021-02-25T10:00:00","publication_year":"2021","word_count":1523,"keywords":["data science","artificial intelligence","low code platform","machine learning","AI","low code no code platforms","chatbots","ML","computer vision","RAG","Aim","analytics","low code","Interviews and Discussions"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","computer vision","data science","analytics","Aim","RAG","chatbots"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/no-code-will-have-a-snowball-effect-spur-innovation-in-digital-landscape-quixy-ceo\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":66618,"title":"Black And White Movies Coloured By AI","content":"Colourisation or adding colours to the black and white or monochrome images and videos has witnessed a widespread adoption for a few decades now. Traditional colourisation techniques need a lot of human efforts as well as are costlier. However, with the advent of emerging technologies like artificial intelligence, these two major issues are disappearing slowly. Not only this but also we have witnessed how researchers are using deepfake techniques to swap faces of celebrities and other popular faces around the globe. Let’s take a look at the few movies that have been coloured using artificial intelligence. They Shall Not Grow Old They Shall Not Grow Old is a documentary film that was directed and produced by Peter Jackson in 2018. The documentary was created with the help of original footage taken from the First World War of the Imperial War Museum’s archives that include most previously unseen photos and videos. The movie used state-of-the-art restoration techniques, colourisation, 3D technologies, and pulled from 600 hours of BBC archival interviews to colourise the original black and white footages. Watch the video here. Pather Panchali An Asst. Research Professor at the University of Maryland in the United States named Aniket Bera used artificial intelligence-based techniques to colourise some scenes from the cult classic Pather Panchali by Satyajit Ray. According to a media house, Bera explained that the footage was automatically upscaled to 60 fps, 4K (Ultra High Definition) and digitally colourised using “deep neural networks that include DAIN, ESRGAN and DeOldify. He added that the entire process is automatic and took around six-seven hours, with no manual input. The processing time per frame was around 10 mins, but it was completely automatic. Watch the video here. The Idle Class Researchers from the Hong Kong University of Science and Technology presented a fully automatic method for colourising black-and-white films without any human guidance or references. The researchers coloured scenes from one of the popular movies by great Charlie Chaplin, The Idle Class (1921). The framework in this research consists of a colourisation network with self-regularisation techniques to improve colourisation quality by propagating colours between similar pixels of a video frame including a refinement network designed to make video colourisation more consistent by enhancing temporal consistency between different frames. Watch the video here. Factory Scene – Modern Times (1936) The famous factory scene from Charlie Chaplin’s movie “Modern Times” has been colourised using open-source AI technology. The code used here is DeOldify and the basic colour corrections were done on the AI output. Watch the video here. The Arrival of a Train at La Ciotat Station The Lumière brothers’ “L’Arrivée d’un train en gare de La Ciotat” is considered as one of the 1st motion pictures ever filmed in 1896. According to Mashable, Denis Shiryaev, who is the videographer has upscaled the movie with the help of emerging technologies like artificial intelligence algorithms, which assisted in enhancing the picture by guessing at what the missing information. The tools that Shiryaev used include the Gigapixel AI from Topaz Labs for upscaling to 4K, DAIN for adding the missing frames, and DeOldify to add colour. These tools are open-sourced and are publicly available for the developers. Watch the video here. La 317e Section 1964 The trailer of this old French movie, La 317e section 1964 has been colourised by Redditor, mar_cnu (Marceau C.). Marceau C. used advanced colourisation technique based on Convolutional Neural Networks (CNNs) where the model consists of four main components, which are a low-level features network, a mid-level features network, a global features network, and a colourisation network. Watch the video here. Wrapping Up From casting directors to predicting the likelihood of a film’s success and writing movie scripts, artificial intelligence is being seen in a number of sections in the entertainment industry. Unlike the traditional colourisation techniques, deep learning-based algorithms helped the researchers to make the process a lot easier as well as time-efficient. In 2018, researchers from NVIDIA developed a deep learning-based algorithm that allows visual artists to simply colourise one frame in a scene and the AI goes to work by colourising the rest of the scene in real time. Sifei Liu, Researcher at NVIDIA stated that the framework is fast and can achieve real-time results. She added that the method also produces better quantitative results than several previous state-of-the-art methods. However, currently, there are no such black and white drama films that are fully converted into colour with the help of artificial intelligence. With this pace in research, the day is closer when we will be able to witness an old black and white movie into a colourised HD one.","excerpt":"Colourisation or adding colours to the black and white or monochrome images and videos has witnessed a widespread adoption for a few decades now. Traditional colourisation techniques need a lot of human efforts as well as are costlier. However, with the advent of emerging technologies like artificial intelligence, these two major issues are disappearing slowly. […]","categories":["Deep Tech"],"tags":[],"author_name":"Ambika Choudhury","publish_date":"2020-06-03T16:00:00","publication_year":"2020","word_count":770,"keywords":["Go","API","artificial intelligence","TPU","AI","neural network","Git","Ray","deep learning","R"],"extracted_tech_keywords":["AI","artificial intelligence","deep learning","neural network","Ray","TPU","R","Go","Git","API"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/black-and-white-movies-coloured-by-ai\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10162145,"title":"St. Fox Opens New Office in Bengaluru for AI and Cybersecurity R&amp;D","content":"Cybersecurity consultancy firm St. Fox has expanded its presence in India with a new office in Bengaluru, marking its second location in the country after Pune. The 100-seater facility in HSR Layout will support the company’s R&D and client delivery teams. The company also plans to double its workforce to 300 employees by 2026 and is eyeing further expansion in Hyderabad and Mumbai. St. Fox is also set to establish client engagement centers in Dubai and Singapore later this year to boost its global reach. Designed with sustainability and employee well-being in mind, the new Bengaluru office incorporates renewable energy sources and energy-efficient infrastructure. The workspace features ergonomic setups, wellness zones, and collaboration hubs to enhance productivity and innovation. Currently, St. Fox employs over 120 cybersecurity experts across India and has partnered with leading institutions to nurture talent in cybersecurity and AI. To address the talent gap in the industry, the company has invested in continuous learning and development programs aimed at cultivating future cybersecurity leaders. Ronak Maniar, founder and MD of St. Fox stated, “We are excited to open our new office in Bangalore, the hub for the IT sector and one of the key businesses & talent markets for us. As part of our long-term growth strategy, expanding into Bangalore will enable us to further accelerate our already rapidly growing business. We believe the most important challenge facing us today is the need to secure our rapidly expanding digital ecosystem while fostering innovation.” Headquartered in Pune, St. Fox operates from a 7,000 sq. ft. facility in a premium IT park, accommodating over 200 employees. The firm is focused on expanding its offerings in AI-driven security, blockchain integration, and advanced data protection to tackle the rising sophistication of cyber threats.","excerpt":"The company also plans to double its workforce to 300 employees by 2026 and is eyeing further expansion in Hyderabad and Mumbai.","categories":["AI News"],"tags":["bangalore","Bengaluru"],"author_name":"Mohit Pandey","publish_date":"2025-01-24T12:23:17","publication_year":"2025","word_count":291,"keywords":["Go","API","programming_languages:R","AI","innovation","programming_languages:Go","Git","Aim","ViT","Bengaluru","R","bangalore"],"extracted_tech_keywords":["AI","Aim","R","Go","Git","API","ViT","innovation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/st-fox-opens-new-office-in-bengaluru-for-ai-and-cybersecurity-rd\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10065638,"title":"Why is Google’s JAX so popular?","content":"Apart from TensorFlow and PyTorch, Google’s new framework, Just After Execution or JAX, has become increasingly popular and with good reason. Essentially, JAX was developed to accelerate machine learning tasks and make Python’s Numpy easier to use. Even though deep learning is a subset of what JAX can do, JAX gained ground after it was used in Google’s Vision Transformer (ViT) and DeepMind engineers posted a blog explaining why it was suitable for several projects. Quite simply, JAX is a high-performing Python library meant for numerical computing, especially in research. There are several reasons to use JAX or even not to use it. Let’s weigh in: Why JAX? Speed: All JAX operations are based on XLA or Accelerated Linear Algebra which is responsible for JAX’s speed. Also developed by Google, XLA is a domain-specific compiler for linear algebra that uses whole-program optimisations to accelerate computing. XLA makes BERT’s training speed faster by almost 7.3 times. More importantly, using XLA lowers memory usage, enabling gradient accumulation, which boosts computational throughput by 12 times in the long run. Source: AssemblyAI JAX also allows users to transform their functions into just-in-time or JIT-compiled versions. With JIT, the speed of subsequent executions can be improved by adding a simple function decorator. However, every function can’t be compiled using JIT in JAX. The JAX documentation notes the exceptions to this rule. Source: Tensorflow Compatibility with GPUs: Unlike Numpy, which is only compatible with CPUs, JAX is compatible with both CPUs and GPUs easily and has an API that is very similar to Numpy. This is why JAX is able to auto-compile code directly on accelerators like GPUs and TPUs without any changes, making the process seamless. A user can write their code just once using syntax that is similar to Numpy, try it out on the CPU and then shift it to a GPU cluster smoothly. Automatic differentiation: JAX aims to differentiate between native Python and Numpy functions automatically. Most of the optimisation algorithms in machine learning use the gradients of the functions to minimise losses. JAX simplifies differentiation with the help of the updated version of autograd. Vectorisation: JAX offers automatic vectorisation via the vmap transformation, which makes life easier for developers. In ML research, a single function is applied to a lot of data at times, say to calculate losses across a batch or to evaluate per-example gradients for differentially private learning. In instances where the data is too large for a single accelerator, JAX performs data parallelism on a large scale using the related pmap transformation. Source: Developpaper.com Deep learning: While JAX is not just a deep learning framework, it has proven to be a solid foundation for deep learning tasks. Libraries like flax, haiku and elegy have been built on top of JAX for deep learning processes. Hessians perform higher-order optimisation techniques in deep learning, and JAX is efficient at computing them. JAX is able to compute Hessians much faster than PyTorch, thanks to XLA. Why not JAX? Despite JAX’s reputation as an accelerator, it isn’t fully optimised for per-operating dispatch in CPU computing. Numpy may, in fact, be faster than JAX in specific cases, especially in small programs, because of the overhead by JAX. JAX is relatively a newer framework and lacks the complete infrastructure that TensorFlow has built over the years. Frameworks like TensorFlow have greater portability deployment and can be employed in a variety of cases like open-source projects, pre-trained models, tutorials and higher-level abstractions through Keras. JAX is still in the research phase and isn’t even promoted by Google as a fully-formed final product. The time taken and the costs incurred to debug or the risk of having untracked side effects might make JAX an insecure bet for new developers. Since side effects like outer-scope encapsulation don’t manifest themselves while running an impure function in JAX, it is difficult to track them. This can lead to serious ramifications in industries like healthcare. Windows does not support JAX. The only way JAX can be used on a Windows system is by using a Virtual Machine. JAX doesn’t have a Data Loader. The user needs to either have their own or borrow one from TensorFlow or PyTorch. JAX is adept at working with lower-level functions that are appropriate for research projects. However, JAX isn’t built for higher-level model abstractions. Source: Developpaper.com Despite JAX’s gradual growth, it is currently employed in a range of projects, like in bayesian methods and robotics, apart from deep learning. Last week, DeepMind announced four new libraries that would join their ecosystem. Mctx offers AlphaZero and MuZero Monte Carlo tree search, KFAC-JAX is a library for second-order optimisation of neural networks and for computing scalable curvature approximations, DM_AUX which is for audio signal processing in JAX, providing tools for spectrogram extraction and SpecAug augmentation, and TF2JAX which is a library for converting TensorFlow functions and graphs to JAX functions. https:\/\/twitter.com\/DeepMind\/status\/1517146462571794433?s=20&t=oMtOMASe5_zWnWPUcoQhcA","excerpt":"Despite JAX’s reputation as an accelerator, it isn’t fully optimised for per-operating dispatch in CPU computing.","categories":["Global Tech"],"tags":["JAX"],"author_name":"Poulomi Chatterjee","publish_date":"2022-04-25T16:00:00","publication_year":"2022","word_count":815,"keywords":["machine learning","Keras","AI","neural network","PyTorch","ML","Aim","deep learning","JAX","TensorFlow"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","neural network","Aim","TensorFlow","PyTorch","JAX","Keras"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/why-is-googles-jax-so-popular\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10102604,"title":"OpenAI Unveils GPT-4 Turbo, Reduces Cost Significantly","content":"Today at the OpenAI DevDay, its first ever developer conference, the company has introduced a preview of its latest iteration, GPT-4 Turbo. It is a refined version of its flagship AI model, GPT-4, at the company’s inaugural developer conference. “GPT 4 Turbo will address many of the things you’ll have asked for. We have six major upgrades to this model,” Sam Altman said at the event. Claimed to be both more potent and cost-efficient, GPT-4 Turbo arrives in two versions, one dedicated to text analysis and another proficient in comprehending both text and images. Available in a preview through an API, OpenAI plans to make both versions generally accessible in the following weeks. It’s priced at $0.01 per 1,000 input tokens and $0.03 per 1,000 output tokens. The pricing for image-processing with GPT-4 Turbo will vary according to the image size. The company optimised its performance to offer GPT-4 Turbo at significantly reduced costs: 3x cheaper for input tokens and 2x cheaper for output tokens compared to GPT-4. Features in GPT-4 Turbo “We are just as annoyed as all of you, and probably more that GPT-4’s knowledge about the world ended in 2021. We try to never get it that outdated again,” Altman said.  The updated GPT will have knowledge till April 2023 which Altman will continue to keep up over time. This context window, larger than any commercially available model, aims to provide better-informed responses and avoid straying off-topic. Additionally, the model supports a new “JSON mode” for valid JSON responses, offering increased utility in web applications and niche settings. Altman said, “We’ve heard loud and clear that developers need more control over the models, responses and outputs. For this, we have a new feature called JSON, which ensures that the model will respond to valid JSON. It’ll make clean API’s much easier. “ Fine-Tuning Program and Pricing Updates OpenAI concurrently announced the launch of an experimental access program for fine-tuning GPT-4, with an increased requirement for oversight and guidance due to technical intricacies. While doubling the tokens-per-minute rate limit for paying GPT-4 customers, pricing will remain at $0.03 per input token and $0.06 per output token for models with varying context window sizes. The company is committed to continually refining and enhancing both GPT-4 and GPT-4 Turbo to meet the needs of developers and users. Furthermore, OpenAI announced updated knowledge bases and longer context windows for both GPT-4 and GPT-3.5. The company also pledged to provide legal indemnity through the Copyright Shield program, offering support and covering costs in the face of potential legal claims around copyright infringement for enterprise users. Legal Protection OpenAI’s continuous improvements across its flagship models and the commitment to legal protection mirror the company’s efforts to advance AI capabilities while ensuring legal safeguards for enterprise users, aligning with industry peers’ similar initiatives to protect customers facing copyright-related challenges. The overarching goal is to continually refine and improve both GPT-4 and GPT-4 Turbo, aiming to provide developers and users with enhanced AI capabilities for diverse applications and tasks.","excerpt":"GPT 4 Turbo is 3x cheaper for input tokens and 2x cheaper for output tokens compared to its previous model, GPT-4.","categories":["AI News"],"tags":["fine tuning","GPT-4 Turbo","OpenAI","Sam Altman"],"author_name":"K L Krithika","publish_date":"2023-11-07T01:13:02","publication_year":"2023","word_count":504,"keywords":["Go","API","Sam Altman","TPU","OpenAI","AI","fine tuning","GPT-4 Turbo","GPT","Ray","Aim","R","llm_models:GPT"],"extracted_tech_keywords":["AI","OpenAI","Aim","Ray","TPU","R","Go","API","GPT","llm_models:GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/openai-unveils-gpt-4-turbo-reduces-cost-significantly\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10055662,"title":"Council Post: Data Engineering Advancements By 2025","content":"Data engineering involves the assortment, transformation and ingestion of data in a standard format into a consolidated data warehouse so that Data Scientists\/Analysts can use it to generate insights. In an analytics project, only 20% (if not lesser) of the work is actually deriving insights from data through data science-based tools and techniques, while the rest 80% is data engineering. However, business leaders are looking to artificial intelligence (AI) and machine learning (ML) techniques to reduce the data engineering efforts and costs involved. Companies are increasingly adopting the data-driven culture by leveraging the power of data to make successful business decisions and drive transformative technologies. Additionally, the data science culture has driven a three-time increase in economic growth for leaders partaking in the external sharing of data. Until now, companies have worked on solving problems related to storing, moving and visualising data. However, data teams are moving beyond this to find concrete solutions that can transform, manage and track the organisation’s data. The next phase of this data-driven culture demands companies to redefine their goals, enhance the data related work processes, and hire data engineers who are efficient, flexible, and accessible. The data analytics space is dynamic and fastly advancing. These are some of the changes that can be expected of data engineering in the coming five years. The Evolving Role of a Data Engineer The job role of a data engineer will become more defined as an entity that assists the organisation in leveraging data. The engineer will not engage in the processing and moving data independently but help the organisation do so through automation. This includes moving away from traditional asynchronous data processing methods and partaking in synchronous workflows that include automating data pipelines and data warehouses. Essentially, data engineers will build tools and infrastructures that allow for efficient moving and processing of data using well-defined frameworks. Integration with Connectors Engineers require new integrations for each system’s one-on-one connection while connecting upstream data sources to the data warehouse. While data engineers integrate and connect data APIs using custom code connections, it is a highly time-consuming process and often causes bottlenecks. A key step in the evolution of data engineering will be the use of automated connectors. Automated data connectors provide a solution through tons of data by connecting to a varied range of data sources without much need for configuration, coding and user input, reducing the time taken to develop and maintain connectors. Additionally, the connector ecosystem and tools like Kafka Connect can help data engineers with viable connectors that can attach to an existing Kafka data pipeline. This will make the process of adding a new system to the pipeline cheap as well. Real-Time Data Transportation Another upcoming trend is the move of data processing systems from batch-based to real-time. Presently, the data transportation process happens in batch ETL snapshots. But with systems like Debezium and Kafka, we see a shift to real-time data pipelines and data processing systems. This has made it possible for extraction, transformation and loading of data to happen in real-time. However, the high costs involved and the processing complexity are still a roadblock that will need overcoming in the coming years. Tooling And Decentralisation Generally, centralised data engineer or data warehouse teams deal with a data-related request through technical automation of operational toil. What we have not achieved properly yet is automating decision-making related to data sharing and access within the organisation; that is, the policy toil. This move needs to be accompanied by data security tools, policy enforcement, and fully automated management to control and access data. A decentralised team accredits the various teams in the organisation to manage their own data warehouses and allows teams to plug into existing data pipelines. Teams can also create their database, datasets, data lakes and data marts in the data warehouse according to their needs- without dependency on the data engineering team. But these proceedings are not all that easy and often lead to complexity, confusion and duplication. Toolings play an integral role in overcoming these challenges and have emerged as a prerequisite for decentralisation. These tools overcome decentralisation’s limitations by accommodating distribution and ownership across multiple teams. For example, Apache Beam or Google Cloud Dataflow allow for the unification of batches while Google Cloud Data Catalog provides central discoverability, control and governance of distributed domain datasets. The future of data engineering is looking to engineers to apply these tools and overcome the challenges of decentralisation. Further, increasing privacy requirements, security audits, and regulatory scrutinies will boost automated data management. The upcoming trends are framing a bright future for data engineers and data leveraging organisations. It holds easier solutions for complex problems and barriers being faced today, along with an enhanced role of data engineers in adding more strategic value to the organisation. Dedicated Data Engineering Support for every team Finally, with more and more organisations adopting new processes to develop, measure and manage data, the roles and constituents of the data teams will change as well. Dedicated data engineering teams will be responsible for providing data products and services to various teams in the organisation. In addition, they will work with different teams to curate dedicated and personalised tactics, data resources and processes. This article is written by a member of the AIM Leaders Council. AIM Leaders Council is an invitation-only forum of senior executives in the Data Science and Analytics industry. To check if you are eligible for a membership, please fill the form here.","excerpt":"Data engineering involves the assortment, transformation and ingestion of data in a standard format into a consolidated data warehouse so that Data Scientists\/Analysts can use it to generate insights. In an analytics project, only 20% (if not lesser) of the work is actually deriving insights from data through data science-based tools and techniques, while the […]","categories":["AI Features"],"tags":["Data Engineering","data engineering career","data engineering demand","data engineering jobs","Data Science Jobs","data science roles"],"author_name":"Anish Agarwal","publish_date":"2021-12-13T13:00:00","publication_year":"2021","word_count":913,"keywords":["data science","artificial intelligence","machine learning","AI","ML","Data Science Jobs","data engineering career","data engineering demand","data science roles","RAG","Aim","Data Engineering","analytics","Kafka","R","data engineering jobs"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","data science","analytics","Aim","RAG","Kafka","R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/council-post-data-engineering-advancements-by-2025\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10120307,"title":"64% of Indian Organisations Hit by Ransomware in the Last Year","content":"Sophos recently released its annual “State of Ransomware in India 2024” report today. The findings reveal that while the rate of ransomware attacks against Indian organisations decreased from 73% in the previous year to 64% this year, the impact on victims has intensified, with higher ransom demands and recovery costs. The report, based on a survey of 500 IT decision-makers in India, found that for the first time, Indian organisations were more likely to recover data by paying the ransom (65%) than using backups (52%). The average ransom demand was $4.8 million, with 62% of demands exceeding $1 million, and the median ransom paid was $2 million. According to the report, 44% of impacted computers on average were encrypted in attacks against Indian victims, and 34% of attacks included data theft in addition to encryption. Excluding ransom payments, the average cost to recover from an attack was $1.35 million. The report also found that 61% of victims were able to recover data within a week, up from 59% in 2022, and 96% reported the attack to authorities, with 70% receiving investigation assistance. Sunil Sharma, Vice President, Sales, India and SAARC, Sophos, emphasised the importance of prevention as the most cost-effective ransomware strategy, along with comprehensive backup and recovery measures and continual review of security posture and incident response plans. The report also highlighted global findings, including that 94% of organisations hit by ransomware said that cybercriminals attempted to compromise their backups during the attack, with 57% successful attempts. Sophos recommends implementing endpoint protection, bolstering defences with round-the-clock threat detection and response, building and maintaining an incident response plan and making regular backups to defend against ransomware and other cyberattacks.","excerpt":"The Sophos report reveals that while attacks against Indian enterprises decreased from 73% in the previous year, the impact has intensified.","categories":["AI News"],"tags":["Ransomware attacks"],"author_name":"Shyam Nandan Upadhyay","publish_date":"2024-05-14T14:52:14","publication_year":"2024","word_count":279,"keywords":["Ransomware attacks","programming_languages:R","AI","RAG","GAN","R"],"extracted_tech_keywords":["AI","RAG","R","GAN","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/64-of-indian-organisations-hit-by-ransomware-in-the-last-year\/","complexity_score":3,"technical_depth":5,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":62526,"title":"New Research Claims That AI Can Zero In On COVID-19 Symptoms","content":"As this article is being written, the current death toll from COVID-19 stands at 1,77,000 worldwide, and going by the trend, a few more would have taken place by the time this article is published. Weeks after it was declared a pandemic, all efforts to find a cure or stop its spread by health and government authorities has gone in vain. Indeed, lack of protective gears, supportive medical machines such as ventilators, and lack of respect for social distancing guidelines have made the operation tough for doctors. What is worse, some experts claim that many positive COVID-19 cases may be asymptomatic in nature. If some people do not exhibit symptoms, how can one know if they are infected? In a recent paper published by Mayo Clinic and a startup called Nference, researchers claim that AI can zero in on symptoms that can indicate if a person is actually suffering from the disease. Mayo Clinic & Nference’s AI claim Mayo Clinic, along with Nference, analysed tests from biomedical reports, and claimed that they have used AI to segregate the phenotypes characteristic of the novel virus. As per the team of researchers, the first few symptoms that arose between 4 to 7 days are a combination of cough, diarrhoea, excessive sweating, and anosmia. To begin their analysis, the team used a natural language processing system that has been designed to automate the identification of diseases, drugs, phenotypes and other entities. This initial step was taken to measure the robust connection between those entities and tag that connection under categories such as positive, negative and other. It also uses Google’s Transformer architecture containing neurons that are arranged in layers and transmit from signals data to adjust the robustness of each connection. Although all AI models learn to predict this way, the Transformer carries a different approach by connecting every output element to the input element. The system was ingested with 82,29,092 clinical notes of electronic medical records derived from the Mayo Clinic for PCR tested patients amounting to a total number of 14,967. Symptoms – along with presumed symptoms – were noted down before the PCR test as well as after a few weeks from the test date. The information extracted by AI revealed that 43 COVID-19 positive patients had diarrhoea in the week before the test, whereas only 822 COVID-19 negative patients had diarrhoea. Other symptoms such as excessive sweating, fatigue, and headache amount to 31, 37 and 35 patients respectively. Moving to fever, the system showed only 24.6% of positive patients showed the symptom before the test, whereas 18.6% were tested negative. Further analysis of data revealed that 251 conjunctions out of 27 phenotypes for a positive and negative patient, two phenotypes-mainly coughs and diarrhoea, and sweating and diarrhoea were found to be significant. Cough and diarrhoea simultaneously existed in 13.2% of patients, whereas 486 patients did not have COVID-19, which indicated a four-fold amplification. On the other hand, 21 COVID-19 patients were found to be suffering from diaphoresis and diarrhoea, whereas 204 patients facing the same symptoms did not test positive. As per the team, the latest developments that were found from the EHR analysis of coronavirus progression can assist human pathophysiology sanctioned synopsis of the exploratory therapies that are currently being investigated for COVID-19. “A caveat of relying solely on electronic medical record inference is that mild phenotypes that may not lead to a presentation for clinical care, such as anosmia, may go unreported in otherwise asymptomatic patients. As at-home serology-based tests for COVID-19 with high sensitivity and specificity are approved, capturing these symptoms will become increasingly important to facilitate the continued development and refinement of disease models. EHR-integrated digital health tools may help address this need,” concluded the co-authors.","excerpt":"As this article is being written, the current death toll from COVID-19 stands at 1,77,000 worldwide, and going by the trend, a few more would have taken place by the time this article is published. Weeks after it was declared a pandemic, all efforts to find a cure or stop its spread by health and […]","categories":["Deep Tech"],"tags":["ai for covid-19","covid-19"],"author_name":"Rohit Chatterjee","publish_date":"2020-04-24T12:43:58","publication_year":"2020","word_count":621,"keywords":["Go","API","TPU","covid-19","AI","Git","Aim","ai for covid-19","ViT","transformer architecture","R","startup"],"extracted_tech_keywords":["AI","Aim","TPU","R","Go","Git","API","transformer architecture","ViT","startup"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/new-research-claims-that-ai-can-zero-in-on-covid-19-symptoms\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10170107,"title":"Indian IT’s AI Revenue is Still 2–3 Years Away","content":"Even as Indian IT’s AI ambitions are high, the actual revenue impact is either minimal, delayed, or marred by unclear reporting. Despite substantial investments in GenAI, most firms still discuss deals in the pipeline or projects in the pilot phase. “Don’t worry about the investments now. Look at the return on investments (ROIs) we can get on the back of it,” HCLTech CEO, C Vijaykumar, said during the Q4FY25 earnings call. He emphasised the importance of investing in skills development and labs, as well as building use cases. “These efforts should lead to the creation of proof of concepts to present to customers. That’s been our approach, and it’s working well so far,” Vijaykumar said. In another instance, Sonata Software, led by Samir Dhir, said the company expects AI-enabled services to contribute 20% of its revenue over the next three years, even though the company said AI is part of all deals right now. During the Q4FY25 earnings call, Dhir said that the firm is pursuing a $34 million pipeline in the AI programme with over 100 clients. On a cautionary note, Dhir stated that although its clients are adopting generative AI internally, it is not impacting any of their offerings. However, it expects a shift in the next 6-12 months. AIM wrote earlier that Indian IT might finally see ROI from AI this year, but the declining revenue for larger IT firms suggests that this might not be the case, since most investment went into AI. Unlike bigger players like TCS, Infosys, and HCLTech, smaller and mid-cap IT firms are taking the lead with AI-led deal wins and greater agility. However, the “invest-now, earn-later” mindset seems to be common across Indian IT. Even Mid-sized Firms Aren’t Immune WNS reported Q3 FY25 revenue of $333 million, marking a flat 2.1% YoY growth. Even as the company has doubled down on AI investments for two years, its CEO Keshav Murugesh stated that the company is leveraging AI and GenAI for long-term value creation. Gautam Singh, head of WNS Analytics, had earlier told AIM that generative AI is already influencing business outcomes. They expect GenAI to impact around 5% of their FY25 revenue. He cited use cases with up to 40% efficiency gains through analytics, AI, and automation. The company added seven new clients and expanded 52 existing relationships, suggesting growing demand for AI-enabled services even if the returns have yet to kick in. Similar was the case for Happiest Minds. In Q4 FY25, the firm reported a 30.5% year-on-year jump in revenue led by strong deal momentum in verticals like healthcare and BFSI and growing investments in GenAI. Happiest Minds Shows the Way Unlike other firms, bringing transparency on the AI investments and revenues, Happiest Minds reported 2.1% revenue from its newly formed GenAI Business Unit. While Happiest Minds has taken a lead in disclosing the tangible impact of its GenAI initiatives, most other firms remain vague about AI’s revenue contribution. This includes companies like Persistent Systems and Hexaware, which posted strong revenue growth in recent quarters. Persistent Systems is targeting $2 billion in annual revenue by FY27. Its managing director, Anand Deshpande, noted that the company’s early bet on Agentic AI drove its current growth. While these two firms are betting on future gains from AI adoption, others take a more measured approach, signalling a slow shift from experimentation to integration in the longer term. For instance, LTIMindtree, which posted a 9.9% YoY revenue growth in Q4 FY25, cited AI integration across services as a major factor. CEO Debashis Chatterjee said AI-powered deals helped the company maintain resilience. In a conversation with AIM earlier, Nachiket Deshpande, the president of LTIMindtree’s AI services, said that LTIMindtree wants to reimagine every service it offers with AI. The company is rolling out solutions to 10,000 customer service agents. “This creates a new revenue opportunity for us as customers invest in these capabilities,” he said. “I predict every single dollar of revenue will have a generative AI component embedded in it. The company recently also inked a $450 million AI deal with a global agri giant, which will eventually increase the firm’s revenue in the coming months. Meanwhile, Mphasis, which delivered its strongest sequential growth in three years—2.6% in USD terms, said 59% of its $390 million deal wins were AI-led. CEO Nitin Rakesh stated that AI and tech are now central to its business strategy. EXL also posted strong first-quarter results for fiscal 2025, with revenue climbing to $501.0 million and adding 10 new clients during the quarter. Chairman and CEO Rohit Kapoor, in a LinkedIn post, said that over 53% of the company’s total revenue in Q1 came from data and AI-led services, a segment that grew 16% year-over-year. It also raised higher guidance for the next year. Similarly, New Delhi-based IT services firm Coforge signed a $1.5 billion 13-year IT deal with Sabre for AI solutions. The company ended the fourth quarter of FY25 with $2.1 billion in deals. “Leveraging AI is the big difference driving the significant difference in growth across organisations in today’s real world,” Coforge CEO Sudhir Singh told Moneycontrol. It seems like Indian IT firms are still in the process of generating revenue from generative AI, which would take at least 2-3 years. This is in contrast to firms like ServiceNow and Accenture which reported generative AI pipeline this quarter and huge revenue from AI services for its clients. For now, Indian IT is positioning for a long game, banking on AI to become foundational to future service offerings. However, without clear ROI metrics, stakeholders are left to go with optimism and vagueness.","excerpt":"“Invest-now, earn-later” mindset seems a common approach among Indian IT on GenAI.","categories":["IT Services"],"tags":[],"author_name":"Mohit Pandey","publish_date":"2025-05-18T10:00:00","publication_year":"2025","word_count":936,"keywords":["Go","GenAI","agentic AI","AI","RAG","Aim","generative AI","analytics","GAN","R"],"extracted_tech_keywords":["AI","analytics","generative AI","GenAI","agentic AI","Aim","RAG","R","Go","GAN"],"url":"https:\/\/analyticsindiamag.com\/it-services\/indian-its-ai-revenue-is-still-2-3-years-away\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10020442,"title":"Guide to TimeSynth &#8211; A Python Library For Synthetic Time Series Generation","content":"Table of contentsIntroductionWhat is a time series?Overview of TimeSynth librarySignal types available in TimeSynthNoise types available in TimeSynthClasses available in TimeSynthPractical implementation of TimeSynthHarmonic signal generation (with white noise) Harmonic signal generation (with red noise)Pseudoperiodic signal generationGaussian process signal generationContinuous AutoRegressive (CAR) signal generationAR model for regularly sampled timestampsMackey-Glass signal generationNARMA seriesReferences Introduction TimeSynth is a powerful open-source Python library for synthetic time series generation, so is its name (Time series Synthesis). It was introduced by J. R. Maat, A. Malali and P. Protopapas as “TimeSynth: A Multipurpose Library for Synthetic Time Series Generation in Python” (available here) in 2017. Before going into the details of the library, since it is used for generating time series, have a quick overview of what the term ‘time series’ means. What is a time series? Time series is a sequence of data points recorded at successive uniform intervals of time. It is thus a sequence of discrete-time data. It can be taken on any variable that changes over time. As the data is collected at adjacent time periods, the observations are correlated. This is in contrast with cross-sectional data wherein an entity is observed at a single point in time. Some examples of time series data include population records, monthly sunspot observations, heights of ocean tides, stock prices and so on. Visit this page to learn more about time series. Overview of TimeSynth library TimeSynth can generate regular and irregular time series. It enables users to match different signals with various architectures allowing a wide variety of signals to be generated. NOTE: TimeSynth supports Python 3.6+ versions. It is supported by PyPI. Signal types available in TimeSynth The types of signals that can be generated using TimeSynth are as follows: Sinusoidal (harmonic) signals: Signals which are periodic functions based on sine or cosine function of trigonometry.Signals generated using AR model: In an AutoRegression(AR) model, the variable of interest is forecasted using a linear combination of past values of the variable. The term “auto” regression indicates that it is a regression of the variable against itself (‘auto’ means ‘self’). In the AR(1) model (first-order autoregression), the current value of a variable is based on the immediately preceding valueIn the AR(2) model (second-order autoregression), the current value is based on the previous two values. Signals generated using CAR model (Continuous AutoRegression model)Signals generated using Mackey-Glass delay differential equationSignals generated using Gaussian processSignals generated using NARMA model (Nonlinear AutoRegressive Moving Average model)Pseudoperiodic signals Noise types available in TimeSynth Following two types of noise can be generated using TimeSynth: White noise: It is a random signal having equal intensity at various frequencies. In terms of sound, white noise refers to sounds which mask other surrounding sounds in the environment. The noise produced by a whining fan is an example of white noise.Red noise: Red noise or Brown\/Brownian noise (named after Robert Brown) is a signal having higher energy at lower frequencies. The sound produced by strong waterfalls is an example of red noise. Classes available in TimeSynth The classes associated with the TimeSynth library are as follows: General classes TimeSampler: This class determines how and when samples will be taken from noise and signal.TimeSeries: A TimeSeries object is the main interface for sampling time-series data. Classes for noise generation GaussianNoise: This class adds uncorrelated, additive (i.e. white) noise to the signal.RedNoise: It adds correlated (i.e. red) noise to your signal.BaseNoise: It contains the signature of both the noise classes mentioned above. Classes for signal generation BaseSignal: It contains the signature of all the signal classes.AutoRegressive: It generates autoregressive (AR) signals. The autoregressive lag is defined by the number of parameters in its ‘ar_param’ parameter. It should only be used with regularly sampled signals.CAR: It generates continuously autoregressive signals.MackeyGlass: It generates the Mackey-Glass delay differential equation (DDE).GaussianProcess: It samples time series from Gaussian Process with the selected kernel (linear, exponential etc).NARMA: Non-linear AutoRegressive Moving Average generatorPseudoPeriodic: It generates pseudo periodic waves.Sinusoidal: It generates harmonic(sine) waves. Practical implementation of TimeSynth Here’s a demo of using TimeSynth library in Google Colab with Python version 3.6.9. Clone the GitHub repository !pip install git+https:\/\/github.com\/TimeSynth\/TimeSynth.git Import the TimeSynth library import timesynth as ts Import Matplotlib library for visualizing the outputs import matplotlib.pyplot as plt %matplotlib inline %config InlineBackend.figure_format = 'retina' In the above lines of code, %matplotlib inline enables the output of plotting commands to be displayed ‘inline’ i.e. within the frontend (here colab notebook). %config InlineBackend.figure_format = ‘retina’ displays the output in ‘retina’ mode i.e. the figures look better with greater resolution The above steps are common. Steps to generate and plot time series using specific types of signals have been explained below: Harmonic signal generation (with white noise) Initialize TimeSampler time_sampler = ts.TimeSampler(stop_time=20) Setting stop_time=20 stops the sampling of time series after 20s. (Default value of the parameter is 10). Sampling irregular time samples irregular_time_samples = time_sampler.sample_irregular_time(num_points=500, keep_percentage=50) sample_irregular_time() samples irregularly spaced time using num_points (number of points in time series) or resolution (resolution of time series) parameter. Only one of these parameters need to be specified. Here, the time series will have 500 data points. keep_percentage=50 means 50% of the data points (i.e. 250 out of 500) will be retained in the irregular series. Initialize Sinusoidal signal sinusoid = ts.signals.Sinusoidal(frequency=0.25) ‘frequency’ parameter represents frequency (in Hertz) of the harmonic signal formed. Its default value is 1.0 Initialize Gaussian noise for adding white noise to the harmonic signal white_noise = ts.noise.GaussianNoise(std=0.3) ‘std’ is a float-value parameter representing the standard deviation of the noise. You can optionally specify its mean using the ‘mean’ parameter. Initialize TimeSeries class with the signal and noise objects timeseries = ts.TimeSeries(signal_generator=sinusoid, noise_generator=white_noise) ‘signal_generator’ and ‘noise_generator’ are signal object and noise object for the time series respectively. You must specify at least one signal generator while the noise generator is optional. Sample using the irregular time samples samples, signals, errors = timeseries.sample(irregular_time_samples) sample() method samples time from the specified time series (here ‘irregular_time_samples’) Plot the series plt.plot(irregular_time_samples, samples, marker='o', markersize=4) plt.xlabel('Time') #name of variable plotted on X-axis plt.ylabel('Magnitude') #name of variable on Y-axis plt.title('Irregularly sampled sinusoid with noise'); #title of the plot Output: Harmonic signal generation (with red noise) Initialize Red Noise red_noise = ts.noise.RedNoise(std=0.5, tau=0.8) ‘std’ is the standard devastation for the noise and ‘tau’ signifies pulse interval in  seconds. Initialize TimeSeries class with the signal and noise objects timeseries_corr = ts.TimeSeries(sinusoid, noise_generator=red_noise) Again, as explained in step (8), ‘signal_generator’ and ‘noise_generator’ are the signal and noise objects for time series creation respectively. Sample using the irregular time samples samples, signals, errors = timeseries_corr.sample(irregular_time_samples) sample() method samples the ‘irregular_time_samples’ time series. It returns the samples as well as signals and errors produced from them, which are stored in the respective arrays on the LHS of the above line of code. Plot the samples plt.plot(irregular_time_samples, samples, marker='o') plt.xlabel('Time') #X-axis label plt.ylabel('Magnitude') #Y-axis label plt.title('Irregularly sampled sinusoid with red noise') #title of the plot Output: Plot the red noise plt.plot(irregular_time_samples, errors, marker='o') plt.xlabel('Time')  #X-axis label plt.ylabel('Magnitude') #Y-axis label plt.title('Red noise');  #title of the plot Output: Pseudoperiodic signal generation Initialize TimeSampler time_sampler_pp = ts.TimeSampler(stop_time=20) #stop sampling after 20s Sample irregular time samples irregular_time_samples_pp = time_sampler_pp.sample_irregular_time(resolution=0.05, keep_percentage=50) The first parameter represents the resolution of the time series while keep_percentage=50 retains 50% of the total generated data points in the time series. Initialize Pseudoperiodic signal pseudo_periodic = ts.signals.PseudoPeriodic(frequency=2, freqSD=0.01, ampSD=0.5) ‘frequency’ denotes the frequency of the signal (default 1.0Hz), ‘freqSD’ denotes frequency standard deviation (default 0.1) and ‘ampSD’ is amplitude standard deviation (default 0.1). Initialize TimeSeries class with the pseudoperiodic signal timeseries_pp = ts.TimeSeries(pseudo_periodic) Perform sampling using the irregular time samples samples_pp, signals_pp, errors_pp = timeseries_pp.sample(irregular_time_samples_pp) samples_pp, signals_pp and errors_pp are the arrays which will store the samples, signals generated from the samples and resulting errors respectively. Plot the series plt.plot(irregular_time_samples_pp, samples_pp, marker='o') plt.xlabel('Time') #X-axis label plt.ylabel('Magnitude') #Y-axis label plt.title('Pseudoperiodic signal'); #title of the plot Output: Gaussian process signal generation Initialize the time sampler gp = ts.signals.GaussianProcess(kernel='Matern', nu=3.\/2) GaussianProcess() method samples the time series with the specified covariance function (kernel). Here, ‘Matern’ kernel has been used which requires the ‘nu’ parameter to be specified, representing the mean of the variables used in the Gaussian process. Other possible kernel types and required parameter to be specified for each can be found here. Initialize the TimeSampler object with gp as the signal generator gp_series = ts.TimeSeries(signal_generator=gp) Sample the time series samples = gp_series.sample(irregular_time_samples)[0] sample() returns a tuple of three arrays (samples, signals and errors – in this order). [0] in the above line of code means the first array in the tuple i.e. that of the samples. Plot the signal plt.plot(irregular_time_samples, samples, marker='o', markersize=4) plt.xlabel('Time')  #X-axis label plt.ylabel('Value')  #Y-axis label plt.title('Gaussian Process signal with Matern 3\/2-kernel');#title of plot Output: Continuous AutoRegressive (CAR) signal generation Initialize CAR model car = ts.signals.CAR(ar_param=0.9, sigma=0.01) ‘ar_param’ is a parameter of the AR(1) process. ’sigma’ denotes the standard deviation of the signal. Both the parameters have default value 1.0 Instantiate TimeSeries object with ‘car’ as the signal generator car_series = ts.TimeSeries(signal_generator=car) Sample the time series samples = car_series.sample(irregular_time_samples) Plot the time series plt.plot(irregular_time_samples, samples[0], marker='o', markersize=4) plt.xlabel('Time')  #X-axis label plt.ylabel('Value')  #Y-axis label plt.title(' Continuous Autoregressive process');  #title of the plot Output: AR model for regularly sampled timestamps Initialize TimeSampler object time_sampler = ts.TimeSampler(stop_time=20) #sampling stops after 20s Sample regular time series regular_time_samples = time_sampler.sample_regular_time(num_points=500) 500 of the total data points will be retained in the time series. Initialize AR(2) model ar_p = ts.signals.AutoRegressive(ar_param=[1.5, -0.75]) ‘ar_param’ is the list containing phi_1 and phi_2 (parameters of the AR(2) process (understand them here)) Initialize TimeSeries object with the signal generator ar_p_series = ts.TimeSeries(signal_generator=ar_p) Perform sampling of the time series samples = ar_p_series.sample(regular_time_samples) Plot the time series plt.plot(regular_time_samples, samples[0], marker='o', markersize=4) plt.xlabel('Time') #X-axis label plt.ylabel('Value') #Y-axis label plt.title('Autoregressive process - Second order');  #title of the plot Output: Mackey-Glass signal generation Initialize TimeSampler object time_sampler = ts.TimeSampler(stop_time=1500) #stop sampling after 1500s Sample irregular time samples irregular_time_samples = time_sampler.sample_irregular_time(num_points=1500, keep_percentage=75) 1500 data points will be there in the time series out of which 75% i.e. 1125 points will be retained. Instantiate MackeyGlass class mg = ts.signals.MackeyGlass() ’tau’ can be specified as a delay parameter; omitting it will assume its default value of 17 Instantiate GaussianNoise for adding white noise to the signal noise = ts.noise.GaussianNoise(std=0.1) ‘std’ is the standard deviation for the noise. Initialize TimeSeries object with the signal and noise generators. mg_series = ts.TimeSeries(signal_generator=mg, noise_generator=noise) Sample the irregular time series mg_samples, mg_signals, mg_errors = mg_series.sample(irregular_time_samples) ‘mg_samples’ , ‘mg_signals’ and ‘mg_errors’ are errors which store the samples, generated signals and resulting errors respectively. Plot the time series without noise plt.plot(irregular_time_samples, mg_signals, marker='o', markersize=4) plt.xlabel('Time') #X-axis label  #X-axis label plt.ylabel('Value') #Y-axis label  #Y-axis label #title of plot plt.title('Mackey-Glass differential equation with $\\\\tau=17$'); Output: Plot the time series with noise plt.plot(irregular_time_samples, mg_samples, marker='o', markersize=4) plt.xlabel('Time') #X-axis label plt.ylabel('Value') #Y-axis label #title of plot plt.title('Mackey-Glass ($\\\\tau=17$) with noise ($\\\\sigma = 0.1$)'); Output: NARMA series Initialize TimeSampler object time_sampler = ts.TimeSampler(stop_time=500) #sampling stops after 500s Sample irregular time samples times = time_sampler.sample_regular_time(resolution=1.) Instantiate NARMA narma_signal = ts.signals.NARMA(order=10) ‘order’ specifies the order of nonlinear interactions in the NARMA formula given here. Initialize TimeSeries object with ‘narma_signal’ as the signal generator series = ts.TimeSeries(narma_signal) Perform sampling samples, signals, errors = series.sample(times) The samples, signals and errors will be stored in the respective LHS arrays on executing the above line of code. Plot the time series plt.plot(times, samples, marker='o', markersize=4) plt.xlabel('Time')  #X-axis label plt.ylabel('Magnitude')  #Y-axis label plt.title('10th-order NARMA Series'); #title of the plot Output: Google colab notebook with the above types of signals generated using TimeSynth can be found here. References Refer to the following weblinks to dive deeper into the TimeSynth library: GitHub repositoryLibrary versions by PyPI","excerpt":"Introduction TimeSynth is a powerful open-source Python library for synthetic time series generation, so is its name (Time series Synthesis). It was introduced by J. R. Maat, A. Malali and P. Protopapas as “TimeSynth: A Multipurpose Library for Synthetic Time Series Generation in Python” (available here) in 2017. Before going into the details of the […]","categories":["Deep Tech"],"tags":["Python Library","Time Series"],"author_name":"Nikita Shiledarbaxi","publish_date":"2021-02-18T11:00:00","publication_year":"2021","word_count":1936,"keywords":["Go","TPU","AI","Git","RAG","Colab","Ray","Python","Time Series","Python Library","Matplotlib","R"],"extracted_tech_keywords":["AI","Ray","Colab","Matplotlib","RAG","TPU","Python","R","Go","Git"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/guide-to-timesynth-a-python-library-for-synthetic-time-series-generation\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10071685,"title":"Neural vocoder and its application in speech recognition","content":"Now a day, most people are familiar with AI voice assistants and use this technology in their day-to-day life. Recently at Google’s I\/O 2018 event, they demonstrated a phone call conversation between the google assistant and a real person. The reason behind the realistic conversation is, of course, the intelligence but also the voice of the AI. The technology used to create a similar realistic voice is called Speed synthesis.  In this article, we will be discussing one of the speech synthesis algorithms and try to learn about the architecture and working of the neural vocoder. Following are the topics to be covered. Table of contents What is a Neural Vocoder?Where it is usedHow does it work?Application of Neural vocoder in speech recognition What is a Neural Vocoder? The artificial recreation of human speech with a computer or other technology is known as speech synthesis. Speech synthesis, the opposite of voice recognition, is usually utilized for converting text information into audio information and in applications such as voice-enabled services and mobile applications. It is also utilized in assistive technology to aid vision-impaired persons in reading written material. A vocoder is a signal processing device that uses feature representation to synthesize the voice waveform. Classic vocoder settings are prompted by an underlying speech production model and include appropriate encodings of the fundamental frequency, spectral envelope, and other factors. Feature sequences are often created on a considerably coarser temporal scale than the target audio source. In the case of voice synthesis, the analytical technique can be replaced with a generator that generates the vocoder settings directly. Neural vocoders are a frequent component in speech synthesis pipelines that transform the spectral representations of an audio stream into waveforms. It is primarily concerned with generating waveforms from low-dimensional representations such as Mel-Spectrograms. There are three main categories of neural vocoders autoregressive models, GAN-based models, and diffusion models. Autoregressive Models The autoregressive models are distinguished by the fact that they are built as probabilistic models that forecast the likelihood of each waveform sample based on prior samples. This enables the generation of a high-quality, natural-sounding voice signal. However, the total synthesis pace is poor when compared to other approaches due to the sample-by-sample production procedure. There are two commonly used autoregressive models, WaveNet and WaveRNN. GAN Models GAN-based vocoders have consistently outperformed autoregressive models in terms of speed and quality of synthesised speech. They use the basic principle of GANs, employing a generator to represent the waveform signal in the time domain and a discriminator to improve the quality of the generated speech. Various GAN-based vocoder variations have been introduced. MelGAN and Parallel WaveGAN are two examples of representative models. Diffusion Models Diffusion probabilistic models are another generative model that includes two primary processes: diffusion and reversal. The diffusion process is described as a Markov chain process in which Gaussian noise is gradually added to the original signal until it is destroyed. In contrast, the reverse procedure is a denoising process that gradually eliminates the additional Gaussian noise and recovers the original signal. There are two commonly used diffusion-based vocoders WaveGrad and DiffWave. Are you looking for a complete repository of Python libraries used in data science, check out here. Where is it used? The majority of speech-generating algorithms, such as text-to-speech and voice conversion, do not directly create waveforms. Instead, acoustic features such as Mel-spectrograms or F0 frequencies are output by the models. Waveforms can traditionally be vocoded from auditory or linguistic data using heuristic algorithms or constructed vocoders. However, the quality of the produced speech is severely constrained and degraded due to the assumptions behind the heuristic approaches. Since Google’s Tacotron 2 used WaveNet as a vocoder to create waveforms from Mel-spectrograms, neural vocoders have steadily become the most used vocoding approach for voice synthesis. Nowadays, neural vocoders have mostly superseded previous heuristic approaches, significantly improving the quality of produced speech. WaveNet creates high-quality waveforms but requires a considerable inference time owing to its autoregressive design. How does it work? Two primary units in a neural vocoder are signal generation and the signal processing unit. The signal processing unit differentiates between the audio signal received as an input. The signal processing operator makes the whole model work as an auto-encoder. The model is fed control parameters (fundamental frequency) derived from a particular signal, which it must recreate by adjusting the different signal model parameters. Surprisingly, the signal parameters derived from an input signal may be changed during inference before being delivered to the front-end (signal generation). The signal generation unit has the DNN, the DNN controls the parameter of the signal generation. If the DNN controller was learnt with enough examples and the input parameter combinations remained within reasonable ranges, the DNN controller should be able to convert the input parameters into the most appropriate signal model parameters. This signal processing and signal creation arrangement is extensively utilised for two reasons: It addresses a long-standing issue with high-level control of complex signal processing models.Training the model with fewer data may be accomplished by forcing some structure into the DNN. The fundamental challenge is to design a structure that is well fitted to the target domain without limiting the model’s expressivity. The source-filter architecture is a strong contender for dividing a vocoder into signal processing operators and DNN modules. The addition of a filter component to a neural vocoder does not imply any restrictions on the signals that the vocoder may represent. Linear predictive coding (LPC), which can be applied to arbitrary input signals, is one of the most fundamental source-filter models. Analytics India Magazine In the above pictorial representation, a typical training process of a vocoder is shown. The dataset used has different variations like a single speaker data sampled at a certain rate with different gender and age types, multi speaker with different ascents and other variations. These waveforms are provided as input for the vocoder. Inside the vocoder, the signals are processed, and filtered and as a result, a synthesized waveform is generated which is evaluated based on different metrics like Mean Opinion Score (MOS) to rate the quality of the generated speech, Log spectral distance. Application of Neural vocoder in speech recognition In general, speech recognition tries to enable spoken communication between humans and computers. The goal is to harmonize information from several disciplines of expertise and to acquire an appropriate interpretation of an auditory message despite potential ambiguities and inaccuracies. The famous neural vocoder used for acoustic speech recognition is Wavenet. WaveNet is a combination of two different ideas: wavelet and Neural networks. Wavenet is an autoregressive convolutional neural network model. The wavenet processes the raw audio which is represented as a sequence of 16 bits samples. The problem with these 16 bits of raw sampled audio is the quantization value which makes it expensive to process through the softmax distribution. So the first step done by the wavenet model is to downsample or compress the audio signal to 8bit by using the μ transformation law. As the Wavenet follows the feed-forward architecture, the raw audio is processed in the causal convolution layer and then sent ahead in the gated activation unit which acts as a filter for the amount of information to be passed to the next layer. The final output from these layers is fed to the ReLU, and at last, the signal is processed through the Softmax distribution. Conclusion Instead of employing standard approaches that feature audible artefacts, neural vocoders based on deep neural networks may produce human-like voices. The neural vocoder is different from the autotuner used for tuning the voices. With this article, we have understood the Neural vocoder and its application in speech recognition. References Read more about WaveNet","excerpt":"Vocoders based on deep neural networks may produce human-like voices","categories":["AI Trends"],"tags":["AI Tool","Conversational AI","DNN","GANs","RNN","Speech Recognition","speech recognition algorithm","Voice Assistant"],"author_name":"Sourabh Mehta","publish_date":"2022-07-28T12:00:00","publication_year":"2022","word_count":1286,"keywords":["data science","Go","Voice Assistant","TPU","AI","neural network","speech recognition algorithm","diffusion models","DNN","GAN","Python","Conversational AI","analytics","RNN","GANs","Speech Recognition","AI Tool","R"],"extracted_tech_keywords":["AI","neural network","data science","analytics","TPU","Python","R","Go","diffusion models","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/neural-vocoder-and-its-application-in-speech-recognition\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":41579,"title":"Is Riak A Good NoSQL Database Option?","content":"The popularity of non-relational databases is not surprising, and over the years, it has gained strong ground in the enterprise world. Riak is one of the databases which has been a vital part of the NoSQL database ecosystem. With a significantly fast processing speed, Riak has done wonders through the years. But is that Riak’s speciality? In this article, we are going to find out: What is Riak? Founded in 2008, Riak was one of the first distributed systems that simplifies the most critical data management challenges for enterprises. It comes in two variants: Riak KV and Riak TS. Even though both variants share the same Riak Core, both have different use cases. Riak KV is a highly available, scalable and easy to operate distributed NoSQL database. One of the major things about it is the fact that it automatically distributes data across the cluster to ensure fast performance and fault-tolerance. Also, Riak KV deals with Key-Value use cases. While on the other hand, Riak TS optimizes time series and IoT data. It also provides faster reads and writes that helps in making it easier to store, query, and analyze time and location data. So, it basically forms the foundation for reducing the complexity of integrating and deploying active workloads for Big Data, IoT and hybrid cloud applications with today’s most flexible and available NoSQL database. Talking about the principle of Riak, this NoSQL database is based on Amazon’s Dynamo paper with heavy influence from the CAP Theorem. The database is written in Erlang and has fault tolerant data replication and automatic data distribution across the cluster for performance and resilience. Five Simple Yet Significant Use cases Of Riak For Session Storage Riak was originally created to serve as a scalable session store. However, over the years the database evolved and started serving some of the complex session storages too. Why? Riak is loaded with some of the advanced features such as Bitcask, MapReduce etc. For example, as user and session IDs are usually stored in cookies, Riak is able to serve these requests with predictably low latency. And it is one of the important use cases for Riak. For Ads Riak has proved time and again that it is one of the best choices for a lot of use cases and serving advertising content is one of the major ones. Being such a popular database, Riak is able to serve content to different web and mobile users simultaneously with low latency. Log Data Log data refers to the information of events that occur in an OS or other software, or messages between different users of communication software. And one of the common yet important use cases of Riak is storing large amounts of log data, and that can be done with the help of MapReduce. For Sensor Data Sensor data is basically the output of a device that detects and responds to some type of input from the physical environment. Many devices collect and send data at a given interval and sometimes it might be cumbersome to handle that data. And this where Riak comes into the scenario, it is considered to be one of the good options to store sensor data. User Account Data User account data is critical for every company to handle. Therefore, companies need a simple, yet effective way to store that data. Here in this case, Riak can be a good storing option too; each account data in Riak could be stored  as a JSON object in a bucket. And further, object keys for the data objects can be used. And in order to retrieve data, a user ID can be used. Outlook Data in today’s world is one of the most crucial factors for any business. And as the amount of data generation is increasing significantly, the concern of reliable storage is also skyrocketing. While there are many databases available, Riak seems to be one of the first choices for many organisations, even after the fact that it has been quite a long time that this database was released. Cab-hailing giant Uber used it till September 2014 and in October, the company replaced it with Cassandra, which is also based on Amazon’s Dynamo paper and has many similar features as Riak. It’s basically more like a replica of Riak with some additional features.","excerpt":"The popularity of non-relational databases is not surprising, and over the years, it has gained strong ground in the enterprise world. Riak is one of the databases which has been a vital part of the NoSQL database ecosystem.  With a significantly fast processing speed, Riak has done wonders through the years. But is that Riak’s […]","categories":["AI Features"],"tags":["Data Management","Database","Mapreduce","nosql","object store database","what is database"],"author_name":"Harshajit Sarmah","publish_date":"2019-07-01T14:10:39","publication_year":"2019","word_count":720,"keywords":["big data","Go","TPU","AI","Database","what is database","Scala","Mapreduce","RAG","GAN","ViT","nosql","SQL","Data Management","R","object store database"],"extracted_tech_keywords":["AI","RAG","TPU","R","SQL","Go","Scala","big data","GAN","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/is-riak-a-good-nosql-database-option\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":66117,"title":"What Is The Hiring Process Of Data Scientists At VMWare?","content":"Software company VMWare has demonstrated its commitment to enhancing its AI and analytics prowess with a string of acquisitions since 2019. In just one year, the company had acquired nine firms and is set to close the tenth deal soon. Additionally, it also focuses on quality hires in the domain of data science to further strengthen its tech capabilities. “We try to integrate data-driven decisions even within our interview process for data scientists,” says Ashish Goyal, Director-HR of VMware India. “Additionally, it is designed to give our candidates a good view of how VMware works, allowing them to make educated decisions about joining us,” he adds. The 5 week-long interview process – which can also go up 7 weeks – may seem arduous and exhausting, but is carefully developed by the company to not only ensure that they onboard the best, but also allow candidates the space to truly understand the company’s ethos and thus, define their mission in the company even before they join. What, then, does the interview process for data scientists at VMWare look like? Let us find out. Skills Over Educational Qualifications Putting an end to an age-old debate on whether or not a strong educational background should be taken as a standard metric to gauge the capabilities of data science candidates, Goyal states the following: “Lack of a strong educational background is not a barrier when it comes to selecting the right candidate. Candidates with pedigree education and relevant certifications are welcome, but they can always gain relevant knowledge through self-learning and relevant certification programs.” According to him, it demonstrates their self-initiative and adds that relevant experience in analytics and data science is a strong plus as well. Furthermore, the company clearly prioritises certain skill sets and knowledge over glorifying titles in a resume. “We consider R or Python to be desirable skills, along with Machine Learning and Natural Language Processing knowledge,” says Goyal. “In addition to this, a deep mathematical understanding of data science techniques is greatly appreciated,” he adds. The company has some additional requirements for senior individual contributors and those in managerial roles. “They need to have good stakeholder management skills as they are needed to implement projects across our businesses.” ALSO READ: How To Crack A Data Science Interview Hiring Process At VMWare At VMWare, every interviewer has pre-defined competencies they will speak with the candidate about. This helps them establish a scientific interview process. Elaborates Goyal: “In the first discussion, we assess a candidate’s analytical ability using case studies, puzzles, situational questions and business problem-solving abilities. The second conversation is about assessing one’s technical ability, while the third discussion focuses on their knowledge of data sciences –  bagging versus boosting, cross-validation, last-mile optimisation, etc. The fourth discussion is more about assessing their past work experience, which will give us a window into their understanding of projects. The check for VMware’s cultural fit is woven through all these discussions.” While the company sources candidates from various sources, including job portals, internal databases, employee referrals, and more, VMWare also participates in hiring events periodically. “We attempt to expand our talent pool as much as possible,” says Goyal. “Platforms like LinkedIn are relevant to discover great talent. In fact, our open positions are published on multiple social media platforms. Individuals who follow VMware and subscribe to company updates also get to view these positions and can express their interest early on,” he adds. The company firmly believes in a data-driven and efficient interview process, where it maximises the utilisation of its time with candidates while giving them valuable experience at the same time. “We try to limit the number of interviews to 4 or 5,” says Goyal. “We believe these are sufficient to give both parties enough data points to make a decision,” he adds. ALSO READ: 10 Most Frequently Asked Questions In Data Science Interview Challenges In Hiring Data Scientists Although entry-level and intermediary positions are fairly easy to fill, the company admittedly struggles with hires at more senior levels. “We find that candidates with 3-7 years experience are easier to identify, and they usually come with great skills and educational backgrounds,” says Goyal. “However, hiring candidates with more than 8 years of experience becomes challenging. This is likely because the pool of skilled candidates in this category is limited,” he adds. This is why VMWare employs certain non-traditional hiring methods and practices to fill these gaps. Hiring events, especially, provides the company with the opportunity to talk to the best data science candidates. “We try to expand our talent funnel as much as possible, and make sure our roles are highly visible to relevant professionals,” says Goyal. “We also leverage social media to reach out to prospective candidates. Additionally, our leaders write blogs about our teams, our culture, and the quality of work that we do to attract wider applications,” he adds. VMWare typically hires candidates into its core research team, as well as for the analytics division. So, candidates are expected to deliver projects as well as work on mainstream research work too as per their expertise.“Our robust hiring process is time-intensive and helps us identify and recruit the best candidate,” says Goyal. “This way, we avoid hiring mistakes. We also try to have an alternate interviewer perspective by bringing in an independent interviewer at times,” he adds.","excerpt":"Software company VMWare has demonstrated its commitment to enhancing its AI and analytics prowess with a string of acquisitions since 2019. In just one year, the company had acquired nine firms and is set to close the tenth deal soon. Additionally, it also focuses on quality hires in the domain of data science to further […]","categories":["AI Hirings"],"tags":["data analytics acquisitions","Data Analytics Certification","Data Science Certification","Data Science Jobs","data scientist qualifications","Data Scientists","Mergers and Acquisitions","qualifications for data scientist","VMWare"],"author_name":"Anu Thomas","publish_date":"2020-05-27T19:00:00","publication_year":"2020","word_count":888,"keywords":["VMWare","Data Science Certification","Data Analytics Certification","R","data science","qualifications for data scientist","RAG","analytics","Data Scientists","Go","machine learning","AI","ML","Data Science Jobs","data scientist qualifications","data-driven","data analytics acquisitions","Python","Mergers and Acquisitions"],"extracted_tech_keywords":["AI","machine learning","ML","data science","analytics","RAG","Python","R","Go","data-driven"],"url":"https:\/\/analyticsindiamag.com\/ai-hiring\/what-is-the-hiring-process-of-data-scientists-at-vmware\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10051832,"title":"Someone Just Created An Algorithm To Predict Survival In The Squid Game’s Glass Bridge","content":"South Korean survival drama television series ‘Squid Game‘ has taken the world by storm ever since it was released in September. The series is centred around 456 debt-ridden players from different walks of life. They are invited to a tournament to play a set of children’s games with deadly outcomes for losing a round in their quest to win a ₩45.6 billion prize. In episode 7, the 16 players that are still alive play the Glass Bridge game. The rules are simple: 18 steps to the end of the bridge, each step with two glasses the player can step on. One of the glasses is safe, but the other will break, and the player would be ‘eliminated’. Alejandro Martinez Vargas, Principal Data Scientist and Decision Science at Atlassian, has put forward theories on the odds of surviving and how many were expected to cross the bridge safely. Alejandro says each player is expected to ‘unlock’ for the rest of the players. Even if a player steps on the wrong glass, it will unlock 1 step for the next player in line. So that’s a 100% probability of unlocking at least 1 step. Of course, there is a 50% chance to select the safe glass, in which case, the player would also be unlocking the next step (as they will do a second jump). Then, 25% chance (50% x 50%) to have 2 safe jumps (unlocking 3 steps for the rest of the players), a 12.5% chance to have 3 safe jumps, etc. So, we can calculate the expected steps each player will unlock, and it looks like: Each player is expected to unlock exactly 2 steps. There are 18 steps, so we expect 9 players to unlock them, and therefore we expect 7 players to survive. In the show, only 3 players survived, so they didn’t do very well. Player 1 would need to pick the right glass 18 times. Each of the jumps has a 50% chance, so the probability of Player 1 reaching the other side is: That’s 1 in over 260.000! Sadly, not great odds for Player 1. For the others, their chance of survival depends on how well or how bad the players before them do. The probability of player K reaching step X depends on two things: The probability of player K-1 (prior player) reaching step Y (with Y < X ) The probability of player K taking X-Y steps As we know that Player 1 had it difficult, we might have thought that Player 5 for sure was going to have much better chances, but not really! Player 10 is the first player with more chances to survive than to be ‘eliminated’. Image Source: Alejandro Martinez Vargas And this chart also proves what we mentioned earlier: we expect 7 players to survive! In the past few weeks, the show has earned rave reviews and has become a pop-culture phenomenon with memes, facts, references, and theories on every major social media platform.","excerpt":"The series is centered around 456 debt-ridden players from different walks of life.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","Data Analytics","Data Science","Deep Learning","Machine Learning","Probabilistic model","probability","Statistics for Data Science"],"author_name":"Victor Dey","publish_date":"2021-10-18T14:55:32","publication_year":"2021","word_count":497,"keywords":["Probabilistic model","Go","probability","programming_languages:R","Machine Learning","programming_languages:Go","ViT","Statistics for Data Science","Data Analytics","Deep Learning","Data Science","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["R","Go","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/someone-just-created-an-algorithm-to-predict-survival-in-the-squid-games-glass-bridge\/","complexity_score":3,"technical_depth":5,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":26305,"title":"Namaste. IBM’s AI Platform Watson Rolls Out New Hindi Translation Services","content":"IBM, one of the world’s largest technology hardware company, has rolled out a new Hindi translation service in its AI-based platform Watson. This marks IBM’s foray into developing Indian languages in its products and services. The Language Translator program allows automatic translation of text for upto 20 languages across the world. Speaking on this occasion, Karan Bajwa, Managing Director at IBM India, gave a brief note of this latest feature. “Watson with its cognitive capabilities understands Hindi or any other language along with its nuances and emotion. Watson’s deep learning methods help it understand words not just through their meaning but through how their meanings change in different contexts, and the frequency of these contexts which ensures that the translation accuracy of sentences is very high.” he told Previously, IBM had showcased Neural Machine Translation (NMT) for Watson this year. This powers the AI engine with more features in translation by maintaining a superior accuracy. “Watson Translator service offers multiple domain specific translation models, can be customized and has the capability of translating documents (MS Office files) while preserving their format. Available in API form, anyone can embed this functionality to translate applications, websites, documents, chats, real time messages and more!” Bajwa added. In the recent years, IBM Watson has slowly making a transformational impact in the field of AI applications.From being a dress designer to even curing cancer, the follow through is immense. Just like digital assistants is omnipresent in today’s world, Watson aims to make a similar impact in businesses on a larger scale. By pooling AI, cloud and IoT together, Watson is definitely reinforcing the next-gen technology with more advancements. In addition, IBM is also leveraging on areas such as machine learning and data science to make this possible.","excerpt":"IBM, one of the world’s largest technology hardware company, has rolled out a new Hindi translation service in its AI-based platform Watson. This marks IBM’s foray into developing Indian languages in its products and services. The Language Translator program allows automatic translation of text for upto 20 languages across the world. Speaking on this occasion, […]","categories":["AI News"],"tags":["IBM"],"author_name":"Abhishek Sharma","publish_date":"2018-07-10T09:25:57","publication_year":"2018","word_count":292,"keywords":["data science","API","machine learning","AI","Git","RAG","Ray","Aim","deep learning","IBM","R"],"extracted_tech_keywords":["AI","machine learning","deep learning","data science","Aim","Ray","RAG","R","Git","API"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/namaste-ibms-ai-platform-watson-rolls-out-new-hindi-translation-services\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10014684,"title":"Create Your Artistic Image Using Pystiche","content":"Since the 20th century, researchers have worked on different algorithms to create appealing artwork to attract artists’ attention. We have seen various algorithms, including strokes based rendering, region-based techniques, example-based rendering, and many more. The idea has always been to select two images, an arbitrary input image and a random style image, to combine them to create astonishing artistic output. In 2015, a group of researchers Leon A. Gatys, Alexander S. Ecker, Matthias Bethge gave birth to “A Neural Algorithm of Artistic Style” Since then, there are more than 240 implementations of the research paper in several frameworks like torch, MXNet, TensorFlow, PyTorch, etc. Neural Style Transfer(NST) is a technique in which we use deep neural networks for image creation to create the visual style of another image. NST uses a Convolution Neural Network(CNN) such as VGGNet and AlexNet, to render an input image in different artist styles. Taxonomy of Neural Style Transfer(NST) Source: Neural Style Transfer: A Review In this article, we will learn about Pystiche, a framework for NST implementation, built and fully integrated with PyTorch. It solves the problem of quickly testing and deploying an NST algorithm to create new style images. Philip Meier and Volker Lohweg have created Pystiche, which has recently been published in the journal of open-source software. NST can be simply explained by the below image, two symbols, and three images. Source: https:\/\/pystiche.readthedocs.io\/en\/latest\/gist\/index.html Installation: pip install pystiche You can also install it from Github pip install git+https:\/\/github.com\/pmeier\/pystiche@master #setup import pystiche from pystiche.image import read_image from pystiche import demo, enc, loss, ops, optim from pystiche.image import show_image from pystiche.misc import get_device, get_input_image print(f\"I'm working with pystiche=={pystiche.__version__}\") device = get_device() print(f\"I'm working with {device}\") Model We are going to load VGG19 and feed the input tensor to the model, to extract a feature map of the content, style, and output image. VGG19  model relatively(compared with ResNet, Inception,etc) generate better output for neural style transfer #multi-layer encoder multi_layer_encoder = enc.vgg19_multi_layer_encoder() print(multi_layer_encoder) Pystiche provides a MultiLayerEncoder, which enables to obtain all the necessary encoding after a single forward pass. Content Loss Content loss calculates the content representation in the content image, which gets captured in a generated image. In this example, content_layer generates encoding then passes it to content_encoder to extract features and thus with the content_weight gives output to content_loss. It is also the distance between the content image, which we want to preserve, to generate the output image. #content loss content_layer = \"relu4_2\" content_encoder = multi_layer_encoder.extract_encoder(content_layer) content_weight = 1e0 content_loss = ops.FeatureReconstructionOperator( content_encoder, score_weight=content_weight ) print(content_loss) Style Loss Style loss, measure the distance in style between the output image and style image. #style loss style_layers = (\"relu1_1\", \"relu2_1\", \"relu3_1\", \"relu4_1\", \"relu5_1\") style_weight = 1e3 def get_style_op(encoder, layer_weight): return ops.GramOperator(encoder, score_weight=layer_weight) style_loss = ops.MultiLayerEncodingOperator( multi_layer_encoder, style_layers, get_style_op, score_weight=style_weight, ) print(style_loss) Perceptual Loss Perceptual loss is a metric of comparing two different images that look similar. It compares the high-level differences in content and style images. Perceptual loss uses feed-forward neural networks for image transformation. We combine the content_loss and style_loss into a joined PerceptualLoss, which will serve as the optimization criterion. The content_loss and style_loss together create the perceptual loss. #computing perceptual loss criterion = loss.PerceptualLoss(content_loss, style_loss).to(device) print(criterion) #load a input image content_image = read_image(\"cheetah.jpeg\", size=size, device=device) show_image(content_image, title=\"Content image\") #load a style image style_image = read_image(\"Van-Gogh-Wheat-Field.jpeg\", size=size, device=device) show_image(style_image, title=\"Style image\") #targets for the optimization criterion criterion.set_content_image(content_image) criterion.set_style_image(style_image) #Initialise the input image for training starting_point = \"content\" input_image = get_input_image(starting_point, content_image=content_image) show_image(input_image, title=\"Input image\") #start the style transfer training output_image = optim.image_optimization(input_image, criterion, num_steps=1000) #write the image to current folder write_image(output_image,'test.png', mode=None) #show the output show_image(output_image, title=\"Output image\") Conclusion In this article, we have learned about Pystiche with code implementation. Pystiche is a fantastic tool to create new render images in the required style. Please check the full code at GitHub and documentation.","excerpt":"Pystiche, framework for NST implementation, built and fully integrated with PyTorch.","categories":["AI Trends"],"tags":["Computer Vision","Convolution Neural Network","VGG19"],"author_name":"Krishna Rastogi","publish_date":"2020-12-19T10:00:00","publication_year":"2020","word_count":638,"keywords":["VGG19","Go","TPU","AI","neural network","PyTorch","ML","Git","Convolution Neural Network","Computer Vision","GitHub","TensorFlow","R"],"extracted_tech_keywords":["AI","ML","neural network","TensorFlow","PyTorch","TPU","R","Go","Git","GitHub"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/pystiche\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10013364,"title":"Guide To Diffbot: Multi-Functional Web Scraper","content":"Earlier, we have seen many web scrapers that can extract data from websites, but many times in the case when sites are changing dynamically over time, it’s hard to scrape and locate elements. Businesses do everything to make their websites free from web crawlers so for solving these problems and making a multi-functional, and more diverse tool Diffbot introduced machine learning and computer vision algorithms and public APIs for extracting information from web pages. Diffbot was created by Mike Tung in 2008 at Stanford University. It was the first company that was funded by StartX  (Stanford’s on-campus venture for funding). Diffbot was the first company to introduce Computer Vision technology to scrape information from web pages, no more conditional programming for each element instead Diffbot visually parses the website’s pages and returns the important elements. In 2012 they introduced Page Classifier API, which can automatically categorize the web pages into specific categories. This adoption of AI systems into their tools was a good move as they were able to analyze 750,000 web pages from Twitter. In 2019 they introduced Knowledge Graph which automatically extracts data from webpages and they build a knowledge base of 2 billion attributes(products, articles, people, companies, and more) and 10 trillion “facts”. This was a huge shot because now their web crawler was able to scrape tiny details from websites which are impossible for other web scraping service providers. Now according to Financial Express report as OpenAI showcased there GPT-3, an advanced version of AI bot, and now according to MIT Technology Review report as well, Diffbot is working on the same, but with a different approach, they are trying to vacuuming up a large amount of human-written text and extracting facts from it instead of training a model directly out of it. You can read more here. This product is more for business purposes. So You need your work email to sign up.Note! Products and services Diffbot provides basic four services: Extract: Automatically extract any article, blog, product, or image from any website without code.Crawl: Extract structured data from entire websites once, or on a schedule, as it is a cloud-based service.Search: Use Diffbot Knowledge Graph to search for information on companies, articles, products, and people.Enhance: Enrich and manage your existing organization or client & employee data using the Diffbot Knowledge Graph. Quickstart After Signup, you’ll get a 14-days free trial which includes 10,000 free credits, access to the knowledge graph, Diffbot cloud dashboard, Excel and google sheets integration and Developer APIs. If Login was successful, then you can see your dashboard right here at: https:\/\/app.diffbot.com\/ On the right side, we have our products Extract, Crawl, Search and Enhance that is discussed above already. On the left tab, we can see custom APIs(create custom web crawlers), Diffbot provides users with freedom of making their own web crawlers with no code notion. Diffbot’s Automatic Extraction APIs Diffbot offers many APIs for extracting data from webpages using computer vision and NLP(Natural language processing), and they are able to categorize the whole page into different attributes and return as JSON. Automatic APIs Analyze API is used to start with when you have no idea with the type of URL; it uses machine learning to figure out the route for the appropriate type of extraction. Page Type APIs If you know what type of content your URL contains, use one of the page-type specific APIs as follows: Article API allows you to extract information about articles, blog posts, and other written content. Diffbot is able to recognize authors and their profile images and links, sentiment, tags based on content, and more.Product API allows you to extract data about products, including specs, colors, availability, price, discount offers, reviews, and more.Image API allows you to extract information about images, from dimensions and download URLs Custom APIs Custom API The Custom API can be used to create an entirely new custom web scrapers by defining rules. You can also use the Custom API programmatically. Integration You can integrate Diffbot with your choice of favorite apps like Google Sheets, Excel, Tableau(Data Visualization tool), and salesforce. It will take you to a new subdomain, then click on Create New. Select API type and URL for which you want to create a custom web scraping API and click on create And specify your own rules for extracting data, custom APIs are not focused on this Demonstration as we are going to deep dive into the knowledge graph. Diffbot Python API From last 6 Years the APIs is not being maintainedNote! This API is mainly designed for developers. Thus, you can take control of the full API from your IDE but it is not maintained and used widely as it has only 13 stars and ten forks on Github. Installation pip install diffbot How to use it: Copy your unique token from the Diffbot dashboard! import diffbot json_result = diffbot.article('https:\/\/github.com', token='your token here') Output: For extracting a specific part from source code, you can achieve by doing the following: To POST data (text or HTML) to the API, use the text or HTML arguments: import diffbot client = diffbot.Client(token='#') json_result = client.api('article', 'https:\/\/github.com', html=''' ... <h1>Introducing GitHub Traffic Analytics<\/h1> ... <p>We want to kick off 2014 with a bang, so today we're happy to launch ... traffic analytics!<\/p> ... ''') Output: Knowledge Graph Knowledge is one heck of a powerful tool able to scrape the whole internet in a minute and will give you corresponding results with customizable entities. 1. Let’s extract all Samsung smartphones from the internet using a knowledge graph for our data analysis research! Click on knowledge graph on left side of the dashboard and click search. Select the Entity type in our case Entity is Product and them from the dropdown menu we can Now we can select different attributes we needed in our dataset like sentiment, brand, review, URL, price, date and more. Click Number of rows like in this case 1000 and click Export Output: 1000 Samsung smartphones details with price, selling site, category and more within just 10 seconds ???? Download dataset from here 2. Same way if you want to do sentiment analysis on your company product: Use Case: You are a Data Scientist. You are given a task to do sentiment analysis on your product like what people have been writing about it, how positive or negative impact it is making on the internet and what are the drawbacks we need to focus? So one way to answer all these questions we can create a large dataset of all articles published on the internet of our product and then do sentiment analysis and research on the same to find out. Our product name is Diffbot, Here we are scraping all the ‘Diffbot’ named articles i.e all the articles written on Diffbot over the years we are going to extract with attributes like publishers, sentiment, tags, URLs, text (important ) and more. Using the same process as above, just change the entity to Article and Use Diffbot as text in Filters. Click Search There you have it !  Full dataset of all the article written on topic “Diffbot” with sentiments, publisher, title, name, author and more you can select from left Tab, and when you are satisfied with dataset just click on Export. Output Structured dataset ready for sentiment analysis for our data science project ???? Download dataset from here Conclusion We saw different types of services, tools, and a full demonstration on Diffbot Knowledge Graph with two Use – case also we have used python API too. Diffbot is a great tool as we have already seen and it has maintained its reputation over the years with its AI power services and further they are trying to improve their knowledge graph. 1000 of developers from fortune 500 companies rely on Diffbot on a daily basis because of its simplicity and accessibility. Want to hear more about web scraping frameworks click on these links: selenium,BeautifulSoupparsehubUrllib and requests Their Research Areas are not just limited to web scraping, they are working on Named entity recognition, reaction extraction, sentiment analysis, computer vision, machine learning, Distributed systems, and more!","excerpt":"Earlier, we have seen many web scrapers that can extract data from websites, but many times in the case when sites are changing dynamically over time, it’s hard to scrape and locate elements. Businesses do everything to make their websites free from web crawlers so for solving these problems and making a multi-functional, and more […]","categories":["Deep Tech"],"tags":["data exploration","data extraction","Natural Language Processing","Web Scraping","Web Scraping Tools","Web Scraping With Python"],"author_name":"Mohit Maithani","publish_date":"2020-12-08T10:00:00","publication_year":"2020","word_count":1358,"keywords":["data extraction","data science","Web Scraping Tools","machine learning","TPU","OpenAI","AI","Web Scraping","Natural Language Processing","ML","data exploration","sentiment analysis","computer vision","NLP","analytics","Web Scraping With Python"],"extracted_tech_keywords":["AI","machine learning","ML","NLP","computer vision","data science","analytics","OpenAI","sentiment analysis","TPU"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/diffbot\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10049291,"title":"Google Introduces New Architecture To Reduce Cost Of Transformers","content":"Over the last few years, transformers like BERT, XLNet, RoBERTa, GPT-3, etc., have extensively been used in many natural language processing (NLP) advances. While these transformers have produced a significantly better performance with scale, the costs of training large models have become increasingly expensive. For instance, the cost of training an XLNet (BERT alternative) model is about $245,000. This cost is based on a resource breakdown provided in the paper, where the researchers have trained XLNet-Large on 512 TPU v3 chips for 500K steps with an Adam optimiser, linear learning rate decay, and a batch size of 2048, which took them about 2.5 days. On the other hand, GPT-3 with 175 billion parameters, which exceeds a memory size of 350 GB upon training, is estimated to cost close to $12 million — 200x the price of GPT2. To give you a ballpark estimate, in a paper titled ‘The cost of training NLP models,’ the researchers noted that for: 110 million parameters — the cost ranges between $2.5K to $50K 340 million parameters — the cost ranges between $10K to $200K1.5 billion parameters — the cost ranges between $80K to $1.6 million Introducing Primer To reduce the training cost of transformer language models, Google proposed searching for a more efficient variant\/alternatives to the transformer by modifying its TensorFlow computation graph. As a result, the team identified Primer (PRIMitives searched transformER), architecture with a smaller training cost than the original transformer and other variants for auto-regressive language modelling. Further, the experiments show that Primer has the benefits of Achieving a target quality using smaller training cost Achieving higher quality given a fixed training cost Achieving a target quality using a smaller inference cost These benefits are massive and hold across model sizes (20 million to 1.9 billion parameters), compute scale (10 to 105 accelerator hours), datasets (LM1B, C4, PG19), hardware platforms (TPUv2, TPUv3, TPUv4 and V100), multiple transformer codebases using default configurations (Tensor2Tensor, T5 and Lingvo), multiple model families (sparse mixture-of-experts Switch Transformers, dense transformers, and synthesisers). The researchers also found that the computer savings of Primer over transformers increase as training cost grows when controlling for model size and quality. When using optimally sized models, these savings follow a power law with respect to quality. The source code for Primer is available on GitHub. Here’s how it reduces the cost In a bid to demonstrate Primer’s savings in an established training setup, the researchers compared 500 million parameter Primer to the original T5 architecture, using the exact configuration used by Raffel et al. applied to auto-regressive language modelling. The outcome, Primer, achieved an improvement of 0.9 perplexity given the same training cost and research quality parity with the T5 baseline models using 4.2x less compute. Further, the researchers demonstrated that Primer’s savings transfer to one-shot evaluations by comparing Primer to the transformer at 1.9 billion parameters in a setup similar to GPT-3 XL. Here, Primer achieved similar performance to the transformer on both pretraining perplexity and downstream one-shot tasks, using 3x less training compute. Primer’s improvements can be attributed to two simple modifications — squaring ReLU activations and adding a depthwise convolution layer after each Q, K, and V projection in self-attention. The researchers said that these modifications are simple and can be dropped into existing transformer codebases to obtain significant gains for auto-regressive modelling. They call the model with just these two modifications as Primer-EZ. The image shown below depicts the modifications. The blue indicates portions of the original transformer, and red signifies one of their proposed modifications. (Source: arXiv) Setbacks Primer surely looks positive, but it has limitations. Firstly, their model parameters sweeps are approximately an order of magnitude smaller than those performed in the original study by Kaplan et al. (Scaling Laws for Neural Language Models). Although their large-scale models use a significant amount of computing, they are still orders of magnitude smaller than SOTA models, such as the full-scale GPT-3 (175 billion parameters). Another setback is that they focus primarily on decoder-only models, while encoder-only (BERT, XLNet and RoBERTa) and encoder-decoder sequence models are still widely used. In this study, the researchers perform encoder-decoder masked language modelling comparisons in T5 but do not study the results in significant depth. The main finding is that, although Primer modifications improve upon vanilla transformers, they perform as Transformer++. In other words, this result suggests that architectural modifications that work well for decoder-only auto-regressive language models may not necessarily be as effective for encoder-based masked models. Google researchers said that developing an architecture that works well for masked language models is a topic of future research. The Future of Transformers, reducing cost Google researchers believe that, in practice, additional tuning could further improve their performance. With this study, the team looks to encourage more research into the development of efficient transformers. For example, an important finding of this study is that small changes to activation functions can result in more efficient training. “In the effort to reduce the cost of transformers, more investment in the development of such simple changes could be a promising area for future exploration,” said the researchers.","excerpt":"Primer’s improvements can be attributed to two simple modifications — squaring ReLU activations and adding a depthwise convolution layer after each Q, K, and V projection in self-attention.","categories":["Global Tech"],"tags":["BERT","Generative Pre-Trained Transformer","GPT-3","Language Models","Machine Learning","Machine Learning New","Natural Language Processing","RoBERTa","Transformers","XLNet"],"author_name":"Amit Naik","publish_date":"2021-09-23T16:00:00","publication_year":"2021","word_count":850,"keywords":["TPU","RoBERTa","Git","BERT","Generative Pre-Trained Transformer","R","XLNet","Machine Learning New","Natural Language Processing","RAG","NLP","Go","AWS","AI","Machine Learning","Language Models","GPT-3","Transformers","TensorFlow"],"extracted_tech_keywords":["AI","NLP","TensorFlow","Transformers","RAG","AWS","TPU","R","Go","Git"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/google-introduces-new-architecture-to-reduce-cost-of-transformers\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10000832,"title":"Circadian Rhythms: Decoding The Science Behind Our Body Clock","content":"Circadian Rhythms are found everywhere, from humans to fungi and cyanobacteria. Circadian rhythms are physical, mental, and behavioural changes that follow a daily cycle. We might not know but a lot of common activities depend on these circadian rhythms. For instance, the reason that we sleep at night and are awake during the day is a light-related circadian rhythm. A Class 11 student of National Public School, Koramangala, Samay Godika at the Breakthrough Junior Challenge this year brought back the ultimate INR 2.9 crore science prize on his explainer video on the science of circadian rhythms, called ‘the body clock’. Here is a deep dive into the science behind these biological clocks called circadian rhythms. Are Circadian Rhythms Same As Biological Clocks? Circadian rhythms are found in most living things except for a few exceptions, such as animals, plants, and many tiny microbes. The study of these systems is called chronobiology. Both, circadian rhythms and biological clocks are related, but they are not the same. Diverse biological processes, from global haemodynamics to intracellular protein levels depict distinct temporal oscillations with a period of 24 hours between one peak and the next. One of the most important differences is the timescale. Biological clocks may refer to different time scales from years to seconds, whereas circadian rhythms only refer to the 24 hour rhythms. They respond primarily to light and darkness in the living being’s, and are controlled in the brain structure called the suprachiasmatic nucleus (SCN), which is a master clock that controls and synchronises in the brain and is a group of around 20,000 brain selles in the hypothalamus. They are controlled by the fluctuating levels of different signal molecules in our cells and these are called biological clocks. Why Are Circadian Rhythms Important? In 1700’s, French scientist Jean-Jacques d’Ortous de Mairan recorded the first observation of an endogenous, or built-in, circadian oscillation in the leaves of the plant Mimosa pudica. Even in total darkness, the plant continued its daily rhythms. This led to the conclusion that the plant was not simply relying on external cues, or zeitgebers, but had its own internal biological clock to rely on. Circadian rhythms are important in determining the sleeping and feeding patterns of all animals, including human beings. There are clear patterns of brain wave activity, hormone production, cell regeneration and other biological activities linked to this daily cycle. There has also been research suggesting lack of sleep and the consequent disruption of circadian rhythms in the development of obesity and depression, as well as most chronic diseases. Studies even show that a lack of sleep may have unexpected side-effects like not being able to read facial expressions. It is also one of the reasons that you experience jet lag. When there is a different time zone while travelling, the biological clocks will be different from the local time. “Metabolism is impacted by the body’s circadian rhythm, the biological process that the body follows over a 24-hour cycle. So the time of day we eat can have a big impact on the way our body processes food,” says Professor Daniela Jakubowicz, MD, of TAU’s Sackler Faculty of Medicine had said. A research also suggest that these clocks are important for for regulation of the cardiovascular system. Understanding what makes biological clocks tick may lead to treatments for sleep disorders, obesity, mental health disorders, jet lag, and other health problems. It can also improve ways for individuals to adjust to nighttime shift work. Learning more about the genes responsible for circadian rhythm will also help us understand biological systems and the human body. Bottomline is that these clocks contribute a lot to the human health and should hence be given more importance in studying.","excerpt":"Circadian Rhythms are found everywhere, from humans to fungi and cyanobacteria. Circadian rhythms are physical, mental, and behavioural changes that follow a daily cycle. We might not know but a lot of common activities depend on these circadian rhythms. For instance, the reason that we sleep at night and are awake during the day is […]","categories":["AI Highlights"],"tags":["health"],"author_name":"Disha Misal","publish_date":"2019-01-08T17:53:55","publication_year":"2019","word_count":620,"keywords":["Go","health","programming_languages:R","AI","ML","programming_languages:Go","ViT","disruption","R"],"extracted_tech_keywords":["AI","ML","R","Go","ViT","disruption","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/circadian-rhythms-decoding-the-science-behind-our-body-clock\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10000753,"title":"Methane Spotted Bubbling Up From The Arctic Permafrost","content":"It takes at least 10 minutes for the ice cream cone to melt and cascade down the elbow. But, what if that time is cut down to something less than 5 minutes? One would have to race their way to the end of the cup while closing the freezer door with the other hand. This is one of the remote implications of global warming. It is not a big deal for those with good gums and those who are indifferent enough to distinguish scoop from soup. The International Energy Agency warned that the world is on course for a rise of 3.6 to 5.3 degrees Celsius. So, those who like to sit and savour, who also happen to form the majority of the population should probably know about the effects of methane leakage from permafrost. Where Is This  Methane Coming From Temporarily trapped in ice, bubbles of methane seeping up from melting permafrost at the base of an Arctic lake Credit: Miriam Jones U.S. Geological Survey The frozen soil is usually saturated with microbes which lay dormant and are almost invulnerable to the extremities of polar conditions. But, when the temperatures rise for whatever reason; anthropogenic or not, the ice melts and exposes the soil to sunlight. This produces carbon, which those microbes love to devour. The terrestrial carbon cycle with the major processes mediated by soil microorganisms via Prosser The microbes then decompose this freshly consumed carbon into methane and release into the atmosphere which then accelerates the global warming; feasting on our ruins literally. Methane Detection From Space A team led by Melanie Engram of the University of Alaska Fairbanks have used the radar measurement data collected by Japan’s Ibuki satellite. This satellite was able to distinguish the surface roughness in the ice divots caused by rising methane bubbles. The surveyors combine aerial photos from the satellite to check for the change in shapes of lakes with respect to time. And, how this change is associated with methane seeps. The study is conducted in the fall when the lakes freeze to spot the trapped methane bubbles. The remote sensing scientists observed that the low-oxygen swamps vigorously spewed methane allowing it to seep through the Arctic lakes. The study was conducted on 48 lakes across Alaska and the results have been extrapolated to estimate the emissions of 5,000 lakes. Though these findings are not immune to exaggerations owing to the sample size the results seem to resonate with those of NASA’s. Methane bubbles pop on the surface of a lake near Fairbanks, Alaska. Credit: NASA\/Kate Ramsaye Why Is This Important Methane is found naturally as methane hydrate. When the temperatures rise, the hydrates are broken down and methane gets dissolves into the seawater. When this methane escapes into the atmosphere, it results in reduced solar reflection from the ice caps and ends up warming up the atmosphere. What is alarming about Methane, is its ability to hit with little notice and high potency compared to carbon dioxide. Researchers claim that methane emissions could rise significantly as temperature soar. The predictive models that have been deployed were designed assuming moderate global warming and the results still show that the methane bubbling in the Arctic lakes would triple. This methane which is released as a result of rising temperatures, can, in turn, accelerate global warming as well. So, this is a double-edged sword built and bred for delivering a catastrophic blow. The scientists have also modelled economic aftermath to create more awareness, just in case. The estimates show that a decade-long atmospheric exposure to methane emissions will lead to an increase in ocean acidification, floods and poorer health conditions and cost $60 trillion to the global economy.","excerpt":"It takes at least 10 minutes for the ice cream cone to melt and cascade down the elbow. But, what if that time is cut down to something less than 5 minutes? One would have to race their way to the end of the cup while closing the freezer door with the other hand. This […]","categories":["AI Features"],"tags":["Interviews and Discussions","satellite"],"author_name":"Ram Sagar","publish_date":"2018-12-28T21:01:40","publication_year":"2018","word_count":616,"keywords":["Feast","Go","ELT","satellite","AI","programming_languages:R","programming_languages:Go","Aim","GAN","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","Aim","Feast","R","Go","ELT","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/methane-spotted-bubbling-up-from-the-arctic-permafrost\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":60295,"title":"Pandas Tricks Not Known By Many","content":"Pandas is a fast, powerful and easy to use open-source data analysis and manipulation tool which is designed on top of the Cytron, C, and Python programming language. It is an amalgamation of two different terms, i.e. panel and data. From combining data frames to reshaping them, Pandas comes with a host of advanced features. For example, it lets a user input a URL in the place of a file name. One can also scrape data from a webpage using its “read_html” function. Although Pandas is one of the most popular libraries among data scientists, due to its wide range of applications, it contains methods that not everyone is familiar with. The list of functionalities Pandas have are too long and broad to be pointed here, but its vast nature amazes the users from time to time. However, there are a number of lesser-known Pandas tricks which one could further use to be more productive. Data Ranges Data ranges are often required to be specified while collecting data from a database or an external API (application program interface). To make the process easier for a user, Pandas have a data range function through which one can return the date incremented by days, months or years. It generates the integer numbers between the given start integer and the stop integers to return a ranged object. To select a range of rows and columns : DataFrame.iloc[Range of rows, Range of columns] Eg. df.iloc[5:10,0:4] will select the rows indexed 5 to 9 and the first 4 columns in the data frame. Similarly, the loc indexer can perform boolean selections: DataFrame.loc[conditional , [column_labels]] Eg. df.loc[df[‘Age’] < 15, [‘user_id’, ‘name’]] selects only the user_id and name from the data frame where the age is less than 15. Merge With Indicators Via Shane Lynn Merge columns are useful for those who are working with larger and multiple datasets and want to merge multiple tables into a single data frame. Furthermore, it is also possible to align the rows each depending on the common attributes and columns. To create a merge, the indicator arguments adds a_merge column to a DataFrame. This step allows a user to identify where the row came from. It could be from the left, right or both DataFrames. The merged column can be further utilised to view the expected number of rows with values from both DataFrames. Nearest Merge Useful for those who are working with financial data such as cryptocurrencies and stocks, and may need to combine price changes with real trades. For example, a user may need to merge each trade with a change in price that occurs every millisecond. To ensure this example can be streamlined easily, Pandas has a function “merge_asof” that allows merging DataFrames by the nearest key. Example Quotes of share price, Trade information data By default, we are taking the asof of the quotes pd.merge_asof(trades, quotes,on=’time’, by=’ticker’) Excel Report Creation Via PBPython With XLsxWriter Library, Pandas allows a user to create an Excel report from the DataFrame. This helps in saving a lot of time as it means that a user does not need to save a DataFrame to CSV and then format it in Excel. One can also use different kinds of charts with the code “writer = pd.ExcelWriter(‘demo.xlsx’, engine=’xlsxwriter’).” This will create an Excel report with plots. To create a chart, one must describe the type of chart (for example, line chart) along with the data series for the chart. Saving Disk Space Working with a variety of data science projects leads to a huge pileup of various preprocessed datasets from different experiments. Due to the reason, the SSD of one’s system might be used up too soon with data, which may not even be required. Pandas enable a user to compress a dataset while saving it along with providing the ability to read it in compressed format as well. A file may grow up to a size of 300 MB, which can be easily compressed with a single argument compression= ‘gzip’ to 136 MB. #read data df = pd.read_csv(“Data.csv”) #zip configuration compression_opts = dict(method=’zip’, archive_name=’out.csv’) #write the bigger data.csv file into .zip format, using below code. df.to_csv(‘out.zip’, index=False, compression=compression_opts)","excerpt":"Pandas is a fast, powerful and easy to use open-source data analysis and manipulation tool which is designed on top of the Cytron, C, and Python programming language. It is an amalgamation of two different terms, i.e. panel and data. From combining data frames to reshaping them, Pandas comes with a host of advanced features. […]","categories":["AI Features"],"tags":["pandas"],"author_name":"Rohit Chatterjee","publish_date":"2020-03-28T13:00:59","publication_year":"2020","word_count":696,"keywords":["data science","API","data_tools:Pandas","programming_languages:R","AI","ML","Python","pandas","programming_languages:Python","R","Pandas"],"extracted_tech_keywords":["AI","ML","data science","Pandas","Python","R","API","programming_languages:Python","programming_languages:R","data_tools:Pandas"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/pandas-tricks-not-known-by-many\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10043288,"title":"India’s Linguistic Diversity A Challenge For Global Brands, Says CEO Of Voice AI Startup Mihup","content":"Kolkata-based Mihup is a vernacular voice interface platform founded by Tapan Barman, Biplab Chakraborty and Sandipan Chattopadhyay in 2016. “The technology came to India a little late, and the global brands find it challenging to deal with the linguistic diversity of the country. Our conversational AI platform integrates Hindi, Hinglish and a mix of other regional languages like Tamil, Bengali, Telugu etc. We want to change the way Indians and the world at large interact with machines and make technology accessible to all,” said Tapan Barman, Co-founder & CEO, Mihup. Mihup is an abbreviation of ‘May I Help You Please’. The startup has partnered with Tata Motors to bring multi-language voice-controlled agents in cars. The product, which is currently available in Tata Nexon and Altroz XZ (O), supports a mix of vernacular languages and can be run even without internet connectivity. What’s the USP? “Mihup has created three products which are built on our proprietary platforms of Speech to Text; Natural Language Processing; Dialogue Manager; and Text to Speech. Our Voice AI products cater mainly to the contact centre and automotive space, with the potential to be incorporated onto any platform,” said Tapan. VIA (Virtual Interaction Analyst) – To analyse all inbound\/outbound customer interactions to generate actionable business insights.AVA – Call Center (Automated Virtual Agent) – To facilitate automated and accurate management of customer queries through a human-like AI interface.AVA – Auto – a multi-language vehicle AI assistant which offers various features such as car control, media and entertainment. Most conversational AI platforms are based predominantly on US English.  “However, what we are doing is fundamentally altering the conversations by integrating Hindi, Hinglish and other mixes of regional languages into the conversation,” said Tapan. Moreover, the product doesn’t necessarily need an internet connection at all times and can seamlessly work in low-connectivity areas and even offline. Tech behind “Mihup’s technology has a low carbon footprint. It provides Voice AI on the edge. It can work reliably on less compute than the industry standard (10- 20x lesser), and Voice AI on edge ensures zero connectivity requirements,” added Tapan. Automatic Speech Recognition (ASR) – neural network-based models that have been optimised to achieve the state of the art accuracy and can run on a modern single-core CPU.Query understanding framework – Uses various modules for NLU activities such as machine learning techniques involving statistical approach-based linear regression, boosting techniques.Information Retrieval – uses custom data structure using hierarchical clustering-based hashing technique to perform complex queries on large volumes of data with filters involving Metaphone, fuzzy matching, context-based matching and give output in real-time.Dialogue Manager powered by Text to speech (TTS) – uses neural network-based models for natural language generation (NLG) Tech stack C++, Java and Android — For products that require edge inferenceNodeJS, Python and Golang for backendAngular for frontendPostgres for databaseRedis for cachingRabbitMQ for message queuing Hiring at Mihup “Our people are our strength, and we are committed to create a workplace of trust and respect. We are always looking for people with diverse experiences, the capability to think differently and to put in a collaborative effort towards the success of Mihup. We have plans to expand our team size to 100+ with cross-functional hires, an almost 70% increase from the existing team of 62. We recently also conducted interviews in some management and engineering colleges like IIMs, IITs and NITs,” said Tapan. Future plans Recently, Mihup raised $1.5 million in Series A round led by Accel Partners, Ideaspring Capital, leading investor Rajesh Jain’s firm Core91 VC and Jayant Kadambi, Founder & CEO, YuMe Networks. Plans are underway to raise Series B funding by mid of FY 21-22. “Mihup has already analysed over 100 million customer interactions and offers 100% interaction analysis to give a precise understanding of the “Voice of the Customer.” Our business has grown 60X, and we have added customers from various domains – from Fortune 500 companies to technology unicorns. We have witnessed a 5x jump in our client base in a matter of just one year. We have also onboarded ten new clients in the last quarter,” said Tapan. Mihup is competing with both national and international companies, including Uniphore, Observe AI and Callminer. “We have plans to expand into the international market with the next round of fundraising. Our goal is to touch $3 million ARR in FY 21-22. We are also adding new capabilities to our interaction analysis product VIA, which will make it a complete go-to Voice AI solution for improving the efficiency of call centre interactions and agents. Features like real-time assistance, auto-scoring are a part of the product roadmap and would be available in FY 21-22,” concluded Tapan.","excerpt":"Kolkata-based Mihup is a vernacular voice interface platform founded by Tapan Barman, Biplab Chakraborty and Sandipan Chattopadhyay in 2016. “The technology came to India a little late, and the global brands find it challenging to deal with the linguistic diversity of the country. Our conversational AI platform integrates Hindi, Hinglish and a mix of other […]","categories":["AI Startups"],"tags":["python database gui"],"author_name":"kumar Gandharv","publish_date":"2021-07-09T13:00:00","publication_year":"2021","word_count":775,"keywords":["Go","machine learning","TPU","AI","neural network","R","ML","Python","Rust","python database gui","Redis"],"extracted_tech_keywords":["AI","machine learning","ML","neural network","TPU","Redis","Python","R","Go","Rust"],"url":"https:\/\/analyticsindiamag.com\/ai-startups\/indias-linguistic-diversity-a-challenge-for-global-brands-says-ceo-of-voice-ai-startup-mihup\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10062401,"title":"This Jaipur based startup is using AI &#038; IoT for Tyre Pressure Monitoring System","content":"In July 2020, The Ministry of Road Transport and Highways issued a notification stating that “Cars and SUVs with tubeless tyres can run without a spare wheel, provided they are fitted with a standard puncture repair kit and tyre pressure monitoring system (TPMS).” This system helps monitor the inflation pressure in tyres alerting the driver of any reduction in the pressure when the vehicle is on the move. Presently, a handful of high-end SUVs have TPMS in India, as installing these systems is voluntary; however, there is a growing demand to increase the fuel efficiency of vehicles to avert any road mishaps. Fleeca India Pvt Ltd, a Jaipur based startup, was founded in 2016 by Tikam Jain with an objective to provide India’s best tyre management and fleet care services to Indian transportation and logistics businesses. The aim was to help fleet owners to leverage the merits of advanced technology in making their businesses more manageable and profitable in terms of efficiency and affordable tyre management services. Fleeca now operates in 21 states with more than 200+ staff. Fleeca is also recognised by the Rajasthan government’s iStart initiative and got a strategic investment of 25 Crores from Bridgestone. Tikam Jain, founder and CEO of Fleeca India Pvt Ltd, claims to have introduced the industry’s first AI and IoT based tyre management in 2016. Before beginning his entrepreneurial journey, Tikam was associated with Shriram General Insurance Co. Ltd. as a manager from 2006-2012. At Shriram, he developed excellent domain knowledge, with a lot of logistics and corporate exposure and expertise in technology and innovation. He started his entrepreneurial journey with the telecom industry from 2013-to 2015, where he was responsible for software development in telecom tower sites. Later, he identified the relevance and need of technology and innovation to bridge the gap in the logistic sector, especially in the fleet industry. Analytics India Magazine interacted with Tikam, who gave insights about how TPMS and Fleeca made a difference in the fleet industry. AIM – What are the drawbacks and challenges of tyre management in fleet services before TPMS? Tikam Jain – There are multiple, including: Increased maintenance and costs: Tyres claim large margins after fuel while considering revenues in fleet-based businesses, and by maintaining tyres, one can help save money on gas and extend the tyres’ lives. A decent tyre pressure management system can save a fleet up to 3 per cent on fuel expenses, which is generally the whole tyre budget. Since there are no regular fleet check-ups and inspection processes without TPMS, it leads to increased maintenance expenses.Manual tyre inspection: Because there was no adequate technology to check the tyres, manual inspections were required from time to time, which drivers frequently overlook or perform needlessly. This leads to either duplication of efforts or improper tyre management and wastage of time.Process delays: Underinflated tyres significantly influence performance, including carrying capacity, safety, and longevity. It also causes delays in getting to the destination and getting the items delivered on schedule.More accidents: Underinflated tyres might cause unpredictable driving habits. This can result in unexpected injuries and accidents. Tyre management technology is needed to make tyres less accident-prone by providing information at the proper time so that drivers and fleets can travel safely and smoothly on the road. AIM – How has TPMS made a difference in tyre management and fleet service? What are the benefits? Tikam Jain – The capacity to compare thousands of vehicles, routes, tyres, and drivers is perhaps the most significant benefit of these big data systems. TPMS gives you access to a lot of information on your tyres’ fuel efficiency, wear rates, failure rates, and other metrics. Benefits of TPMS include: Extend tyre life: Underinflated tyres are the leading cause of tyre failure. When tyres are not properly inflated, they might experience mechanical or thermal overload, resulting in sidewall damage, tyre disintegration, and failures. A tyre pressure monitor can rescue you from all of these situations by delivering you real-time data.Fuel efficiency: Incorrect tyre inflation necessitates a large amount of fuel to run. The vehicle burns less gasoline and gets greater mileage when the tyre pressure is accurate.Economical: A tyre pressure monitoring system helps save money in various ways. It cuts down the downtime hours and saves money on repairs and gasoline refills.Improved road safety: Unpredictable driving behaviours are caused by underinflated tyres. This can result in unexpected injuries and accidents. To save from all of this, TPMS provides the necessary information at the appropriate time, allowing you to visit a nearby tyre repair shop.Environment friendly: When tyres are underinflated, they produce more carbon monoxide than normal. Installing a tyre pressure monitoring system (TPMS) in the car makes one more ecologically conscious. AIM – How does TPMS work? Please explain in detail the tech used and elaborate on how it functions. Tikam Jain – The TPMS (Tyre Pressure Monitoring Device) is an electronic rim-mounted system without the need for signal boosters that warns the driver every 20 seconds about the temperature and air pressure in the tyres of heavy vehicles. This system may be put in both tubeless and tube tyres, and it will improve not only tyre life but also road safety and environmental protection. There are two types of TPMS: Indirect TPMS Indirect tyre pressure monitoring systems (iTPMS) do not use physical pressure sensors, instead measure air pressures with software-based systems that estimate and monitor tyre pressure without physical sensor systems in the wheels by analysing and combining existing sensor signals such as wheel speeds, accelerometers, driveline data, and so on. Direct TPMS This is the most secure and preferred method of using TPMS. Individual sensors are installed in the tyres, and the data is relayed to the dashboard the same way as a radio signal is. The driver is informed of the pressure and temperature of the tyres and can take appropriate action. If you are concerned about the sensors’ longevity, keep in mind that they rely on a battery that must be updated to provide the correct information at the appropriate moment. AIM – Is this technology first in India? Is it used elsewhere (abroad)? Tikam Jain – TPMS was first used on European premium automobiles in the 1980s, and the 1997 Chevrolet Corvette was the first American car to have it. The TREAD (Transportation Recall Enhancement Accountability and Documentation) Act, passed by the Clinton Administration in 2000, was TPMS’ big break. The TREAD Act, among other things, required that every new car sold in the United States after September 2007 be fitted with a tyre pressure monitoring system (TPMS). Currently, with a combined market share of 58.7%, the United States and Europe constitute substantial marketplaces worldwide. China is the fastest-growing market, with a CAGR of 17.7% throughout the analysis period, owing to legislation requiring all vehicles registered after 2019 to be equipped with tyre pressure monitoring systems by 2020. As the application of GB26149 for TPMS grows from vehicle categories M and N to vehicle category M1, TPMS demand in the Chinese OEM market will rise. AIM – What is the road ahead for Fleeca? Tikam Jain – The company began with only two employees in Jaipur in 2016 and has since expanded around the country. From a long-term viewpoint, we wish to continue growing and expanding in both financial and geographical regions. We already operate in 21 states and hope to expand our services to the rest of the world. We are also aiming to open more training centres to teach students about tyre management, and we would like to work with more institutions to do so. We are also working to make our solutions more accessible by forming new partnerships to expand the range and scope of our services throughout a larger geographic area.","excerpt":"The TPMS (Tyre Pressure Monitoring System) is an electronic rim-mounted system that warns the driver every 20 seconds about the temperature and air pressure in the tyres of heavy vehicles without the need for signal boosters.","categories":["AI Features"],"tags":["AI Startups","Interviews and Discussions"],"author_name":"Poornima Nataraj","publish_date":"2022-03-10T10:00:00","publication_year":"2022","word_count":1293,"keywords":["big data","Go","AI","innovation","RAG","Aim","ViT","analytics","GAN","R","AI Startups","Interviews and Discussions"],"extracted_tech_keywords":["AI","analytics","Aim","RAG","R","Go","big data","GAN","ViT","innovation"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/this-jaipur-based-startup-is-using-ai-iot-for-tyre-pressure-monitoring-system\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10137963,"title":"NVIDIA ships its much-awaited Blackwell AI Chips to OpenAI, Microsoft","content":"On October 8, OpenAI announced via X that its team received the first engineering builds of Nvidia’s DGX B200. These new builds promise three times faster training speeds and fifteen times greater inference performance than previous models. Contrary to tradition, this time, the technical team, rather than the founders, received the first GPU chip, with Sam Altman notably absent. When OpenAI was newly founded as a research non-profit, Jensen Huang, NVIDIA’s CEO, personally delivered the first GPU chip to Elon Musk, one of the founders. Earlier this year, Huang delivered the first NVIDIA DGX H200 to OpenAI’s founders, Sam Altman and Greg Brockman. With the recent change in leadership with senior leaders like Mira Murati and others leaving OpenAI, many opine that this is the start of a new era. On similar lines, Microsoft shared that its Azure platform is the first cloud to employ Nvidia’s Blackwell system, featuring AI servers powered by the GB200. “Our long-standing partnership with NVIDIA and deep innovation continues to lead the industry, powering the most sophisticated AI workloads,” said Satya Nadela, Microsoft’s CEO. Hints about these developments emerged earlier this year, with NVIDIA reporting a record  $22.6 billion in data centre revenue, marking a 23% sequential increase and a staggering 427% year-on-year growth. It was driven by continued strong demand for the NVIDIA Hopper GPU computing platform. During the company’s earnings call, Huang mentioned that “after Blackwell, there’s another chip. We’re on a one-year rhythm.” OpenAI’s Changing Relationship with Microsoft Microsoft was one of the early investors in OpenAI, and it had a significant influence over the organisation. However, according to an exclusive report by The Information, OpenAI is now exploring avenues to partner with competitors like Oracle to meet its growing demand for computer power. This move could help OpenAI gain control over its data centre plans and maintain an edge over Anthropic and xAI, which are not tied to a single cloud provider. Simultaneously, Microsoft actively seeks to reduce its dependency on OpenAI’s technologies, reflecting a shift in the dynamics between the two companies following OpenAI’s recent funding round.","excerpt":"Early this year, Jensen Huang delivered the first NVIDIA DGX H200 to OpenAI’s founders, Sam Altman and Greg Brockman.","categories":["AI News"],"tags":["NVIDIA"],"author_name":"Aditi Suresh","publish_date":"2024-10-09T16:45:07","publication_year":"2024","word_count":347,"keywords":["Anthropic","OpenAI","AI","Azure","R","innovation","XAI","GAN","NVIDIA","xAI","GPU computing"],"extracted_tech_keywords":["AI","OpenAI","Anthropic","xAI","Azure","GPU computing","R","GAN","XAI","innovation"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/nvidia-ships-its-much-awaited-blackwell-ai-chips-to-openai-microsoft\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10079418,"title":"Once Left Out in Cold, Enterprises Return to Semantic Layers","content":"In the decade gone by, new enterprise technologies have come up for better data management like data lakes, cloud computing and data mesh. But, the importance of a semantic layer got forgotten somewhere along the way. Enterprises still see a gap between their data and their data science teams as descriptive analytics teams and technology still remain in two separate places. Descriptive analytics is one thing and ML-based analytics is another thing entirely but both are just as essential to companies. Need for a semantic layer The semantic layer is one of the underlying platforms in data architecture which helps people access the data on their own in simpler language. The value of it doesn’t solely lie in ensuring the language and perspectives are consistent throughout. Customers can then determine their business metrics and define them for themselves and keep reusing them. The semantic layer allows the critical metrics within organisations to be the same across workstreams and cuts down the amount of time data scientists spend coordinating between teams. Multiple sources of data and use cases in enterprises, Source: getdbt.com When an organisation has its semantic layer right in place, where the raw data is stored or what tools different teams are using to consume the data output doesn’t matter. All the teams can simply feed themselves from the semantic layer and any new information found by data scientists can just be fed into the layer so that future decisions are also based on them. This liberates data science, which can normally get tied to certain platforms or tools and removes the costs of reworking things. However, it is far easier to maintain one semantic layer for definitions instead of hundreds of scattered reports. While the concept of semantic layers is as old as the early 90s, there is a resurgence for them in the industry now. In 1991, SAP introduced semantic layers. But it wasn’t until 2021 that the ‘modern’ semantic layer, as we know it, returned. Tools like MetriQL, Airbnb’s Miernva, MetricFlow and Cube.js offered semantic layers. Why were semantic layers abandoned? Despite all these upsides, enterprises purposely kept themselves away from semantic layers because of how cumbersome centralised semantic layers were. Semantic layers took time to build and maintain. Besides, the layer has to continuously be in sync with the database and any changes that happen. The data for semantic layers also usually existed in multiple backend systems or operational data stores (ODS) so organisations had to manage several semantic layers simultaneously as one semantic layer had to be kept for each system or each tool. But logic within enterprises had become dispersed everywhere—data was distributed, duplicated and varied combinations of data were formed. Companies indubitably needed data but there was also no getting away from it. Organisations  had complicated personal models and data management was becoming a significant challenge. Self-Service BI tools with a semantic layer and Tableau, Source: AtScale Rise in new BI tools With the intent to remain agile, companies started using new, seemingly fancy tools that came up in Business Intelligence (BI) like Tableau and Qlik. Then, heavy, centralised semantic layers were done away with completely. The idea was that these low-code no-code BI tools would simplify the process and democratise the data. IT companies were slowly being pushed to appease more sophisticated clients. But the more these offerings cropped up, the more companies adapted them, the more confusing things became. There were multiple BI tools for multiple teams and no single semantic layer—someone used Tableau, someone used Power BI and someone else used Excel and there was no single data point. Source: getdbt.com A new universal semantic layer The realisation came much later that while these data discovery tools were great at what they had been designed for, they weren’t necessarily appropriate for core BI. What this calls for is a new and improved universal semantic layer. Cloud businesses like Google, Snowflake and even unicorns like dbt Labs are now speaking up about the indispensability of a universal semantic layer. The core idea behind dbt Labs’ new semantic layer is that users should define the universal metrics only once and use them anywhere. This is not to say that data discovery tools can be discarded as discovery is a legit function in BI but tools for data discovery and semantic layer are not interchangeable. The key thing to remember about semantic layers is that it follows an ‘All or Nothing’ principle. A semantic layer is only useful when it is truly universal and completely misses the target when it isn’t. This means that it must support a wide range of use cases and roles like data scientists, business analysts and developers. A universal semantic layer should also work with a variety of query tools like SQL, MDX, DAX, Python REST, JDBC and ODBC. An ideal semantic layer is defined by core features like semantic modelling which maps the logical elements like the metrics and KPIs with the physical entities in the database, a multidimensional calculation engine which is scalable, performance optimisation which works on speed and analytics governance. In case any of these requirements are amiss, a semantic layer essentially becomes unusable.","excerpt":"With the intent to remain agile, companies started using new, seemingly fancy tools that came up in Business Intelligence (BI) like Tableau and Qlik.","categories":["AI Trends"],"tags":[],"author_name":"Poulomi Chatterjee","publish_date":"2022-11-11T14:00:00","publication_year":"2022","word_count":861,"keywords":["data science","TPU","cloud computing","AI","ML","Python","analytics","SQL","R","Snowflake"],"extracted_tech_keywords":["AI","ML","data science","analytics","cloud computing","TPU","Snowflake","Python","R","SQL"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/once-left-out-in-cold-enterprises-return-to-semantic-layers\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10042950,"title":"Breathing Life Into Robots Through Simulators","content":"Simulation enables engineers to prototype rapidly and with minimal human effort. In robotics, physics simulations provide a secure and low-cost virtual playground for robots to gain physical skills through Deep Reinforcement Learning (DRL). However, simulations use hand-derived physics that will have difficulty adapting when tested on real hardware. This challenge is termed the “sim-to-real gap” or the domain adaptation problem. Reinforcement-based approaches(  RL-CycleGAN and RetinaGAN) have been utilised to bridge the simulation-to-reality gap for purely perceptual tasks, such as grasping. However, the gap is still present because of the dynamic characteristics of robotic systems. In this case, researchers are prompted to ask whether or not they can find a more accurate physics simulator by examining a few real robot trajectories. If so, then it may be possible to use the improved simulator to give the robot controller a higher chance of succeeding in the real world. Learn about SimGAN, a novel physics simulator built with #ML that helps close the sim-to-real gap by replacing manually-defined parameters with learnable functions that change according to the state of a robot and more accurately replicate the real world. https:\/\/t.co\/Q4HjlDRY9P pic.twitter.com\/juGq2Jc5oV— Google AI (@GoogleAI) June 16, 2021 In a paper published in ICRA 2021, titled SimGAN: Hybrid Simulator Identification for Adversarial Reinforcement Learning, researchers proposed to treat the Physics Simulator as a learning component with a particular rewarding function trained by the DRL that penalises differences between the trajectories generated in simulation, that is, robots moving over time. According to the researchers, reinforcement learning (RL) policies can be trained using simulation data to support more diversified actions in robots. While controller creation in simulation has been made far more automatic due to implementing learning-based methodologies, moving a trained policy from simulation to real hardware typically involves considerable manual work. To account for the possible ranges in the simulation and the actual world, a range should be big enough to include all of the unmodeled differences but not so large as to impede performance. The researchers, therefore, focused their contribution in : A unique simulation identification formulation that is posed as an adversarial RL problemA learnt GAN loss that alleviates manual loss design and sensitive excitation trajectories by providing limited set-level supervisionReducing the necessity for a properly defined parameter set through an expressive hybrid simulator parameterisation. Hybrid simulator A conventional physics simulator is a system to simulate the movement or interaction of objects in a virtual world by solving differential equations. However, given the complexity of the circumstances that robots could experience in the actual world, such environmental modelling techniques would be arduous (or possibly impossible), so it is helpful to employ a machine-based approach instead. While the simulators can fully learn from the data, the learnt simulator may violate the laws of physics when it needs to model scenarios if the training data does not cover a diversity of situations. Hence, the robot trained in such a small simulator in the actual world is more likely to fail. To address these complications, researchers built a hybrid simulator, combining both neural networks and physics equations. In particular, researchers replace those parameters, often manually defined by the simulator — contact parameters and motor parameters — with a simulation that can be learned as the unmodified contact and motor dynamics details are important causes of the sim-to-real gap. The third component of the hybrid simulator includes physical equations that ensure that the simulation complies with fundamental physical principles, such as energy preservation, bringing it closer to the real world and lowering the sim-to-real gap. (Source: ResearchGate – SimGAN: Hybrid Simulator Identification for Domain Adaptation via Adversarial Reinforcement Learning) Therefore the researchers set up the experiment to see if their method could work in order to: Enhance domain adaptability for robots with varying morphologiesDeal with dynamical mismatches that occur during sim-to-real transfer Manage dynamical disparities that are not intuitively translated to the list of parameters that our model identifies but can be absorbed by our model’s state-dependence. GANs ( Source: Google AI Blog) Generating identical trajectories in a hybrid simulator to those collected on the real robot will be successful if one learns the parameter functions for the simulation. This ability to learn is enabled by having a metric for the trajectory similarity. GANs, created to generate synthetic images with the same distribution or “style” with a limited number of authentic images, can now be used to create synthetic trajectories indistinguishable from real ones. Reinforcement learning (Source : Google AI Blog) Therefore, the research concludes that simulation learning can be thought of as an RL problem. A trained neural network using only a small number of real-world trajectories learns state-dependent contact and motor parameters. To do this, the neural network is configured to produce the simulation’s trajectories with minimum error. Reducing this inaccuracy over a period of time increases the accuracy of a simulation that will ultimately guide the control system. Summing up One of the significant impediments preventing robots from harnessing the power of reinforcement learning is the sim-to-real gap. Researchers addressed this problem by developing a simulator that can more accurately replicate real-world dynamics while requiring only a modest quantity of real-world data. The researchers plan to expand on this basic framework by extending it to other robot learning tasks, including navigation and manipulation.","excerpt":"Simulation enables engineers to prototype rapidly and with minimal human effort. In robotics, physics simulations provide a secure and low-cost virtual playground for robots to gain physical skills through Deep Reinforcement Learning (DRL). However, simulations use hand-derived physics that will have difficulty adapting when tested on real hardware. This challenge is termed the “sim-to-real gap” […]","categories":["AI Features"],"tags":["Deep Reinforcement Learning","GANs"],"author_name":"Ritika Sagar","publish_date":"2021-07-06T13:00:00","publication_year":"2021","word_count":878,"keywords":["Replicate","Go","API","AWS","AI","neural network","cloud_platforms:AWS","ML","Deep Reinforcement Learning","GANs","GAN","R"],"extracted_tech_keywords":["AI","ML","neural network","AWS","R","Go","API","GAN","Replicate","cloud_platforms:AWS"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/breathing-life-into-robots-through-simulators\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":22222,"title":"What’s Stopping Google’s Monopoly In AI – Limited Success In Hardware?","content":"Does Google really lead the world in artificial intelligence? The search giant leads in commercial search and its technical expertise in machine learning and deep learning is unmatched – but in the broader world of artificial intelligence that includes the smart home device space underpinned by NLP & speech recognition, self-driving technology space (Google’s facing stiff competition from all corners, especially NVIDIA), cloud-based technologies, next battleground for AI – Google is the no. 3 US cloud services provider and smartphone market, now on the cusp on mobile AI revolution, the company is still playing catch-up. But when it comes to the enterprise AI market – Google has cracked the code for success, thanks to open sourcing a slew of ML tools in Google Cloud that could translate into big leads for the company. By offering a complete AI ecosystem, Google looks to corner a substantial market share. Besides, Google’s strong suit had always been data and immense computing power on which it built its enterprise AI market. Google was never a hardware-focused enterprise — is now baking AI into the devices Let’s face it – when it comes to building hardware, Google has met with limited success. The company was ostensibly late to market with social networking, the market for its AR glasses, Google Glass is still unproven, Android TV still lags behind in popularity to Apple TV (Chromecast is an exception), their robotics division (Boston Robotics sold off to SoftBank) has been sold off and in the smart home space, Amazon definitely is winning in device sales at the moment. Given how cloud computing services play a huge role in the smart home device market, Amazon is still in a great position to lead the market.Reports state that Amazon has sold more Echo smart speakers than Google Homes. It’s most recent hardware success – Google’s AI driven phone Pixel phones which offers excellent photo-enhancement features and deeper hardware-software integration underpinned by AI-based technology isn’t pegged as a commercial success either. In fact, analysts suggest Pixel is toeing the line of iPhone when it comes the look and design. And given how smartphones can be the key to build mobile AI focused innovations, it’s hard to overlook the fact that iPhone still rules the market. Even though the Pixel series is nipping at the heels of Samsung & Apple, a recent report points out that the company still needs a year to sell as many Pixels as Apple’s iPhone and gain a larger footprint. An IDC report states that Google Pixel sales doubled last year to 3.9 million, but still represents “a tiny portion of the 1.5 billion market size.” In the VR headset segment, Samsung Gear VR is still most widely accepted than Google Daydream VR headset. However, the company is plowing billions of dollars in building consumer-facing AI-powered devices such as Pixel Buds, wireless AI powered headphones but it is still years away from making them mainstream. But there’s one area where the search giant leads – building open ecosystems & profiting from it It’s true open ecosystems lead to unprecedented innovation and Google has built its empire on open source – just the way Android created innovation and became a gamechanger in the smartphone space. According to NetMarketShare, Android is the dominant mobile platform with a 58.75 percent share of the worldwide mobile and tablet operating systems market. With mobile becoming the dominant global platform, Android’s global OS share was 37.93 percent compared to 37.91 percent for Windows, as per StatCounter report.  Interestingly, Android tore down barriers in the mobile ecosystem and spurred developers to build a mobile software ecosystem without having to do heavy engineering. When it came to web browser, Google’s Chrome effectively killed Microsoft’s Internet Explorer and choked the competition, thereby profiting from collecting user data. Google, a trendsetter in open sourcing AI technologies According to a Capgemini State of AI report, one of the most defining characteristics that led to the growth of AI technologies is the open sourcing of key technologies by Google, Microsoft & Amazon. All the major tech companies are keen to have more developers on their platforms. It is here that Google proved to be a trendsetter by open sourcing its TensorFlow Platform in 2016 (Facebook then open-sourced Caffe, its flexible deep learning framework, and Amazon did the same with MXNET). Most consumer-facing, traditional organizations that lacked the technology were willing to find real-world applications for their business challenges and found the open-source platforms an interesting avenue for investment. Google clearly automated data science by enabling enterprises to build and train custom machine learning models using APIs to add machine learning capabilities to their applications, thereby implementing AI at scale. 1) First mover advantage with TensorFlow: As machine learning increased in importance, Google open sourced Tensorflow and Kubeflow to open up the machine learning ecosystem. It was a good attempt to help companies run like Google, in return they wish enterprises will turn to Google Cloud Platform to run more of their workloads. 2) In the same vein, Google has lowered the bar for machine learning adoption with Cloud AutoML: When it comes to the enterprise AI market, Google clearly has a well-laid out strategy and has cracked the code of winning over the developer community by equipping them with the tools and tech which leads to mainstream enterprise adoption. According to Google Cloud CEO Diane Greene, it is only a matter of time before AI offerings become “more and more self-help”. For long, Greene’s pitch to accelerate Google Cloud adoption has been — you don’t need a Ph.D. in machine learning, but you can still build a highly accurate machine learning model, she said reportedly. A key advantage of CloudML is that enables developers, besides data scientists to customize a model without having in-depth ML knowledge and Google has made the interface more drag-and-drop. Google Cloud offers APIs that provide machine learning capabilities to companies to build tools. Another key advantage over other providers AWS & Microsoft Azure is that companies can train their algorithms on enterprise data directly and Cloud AutoML ingests data assets and cranks out the model accordingly. Unlike other major providers, AWS which has a suite of other AI\/ML services, and Microsoft’s Azure machine learning, Cloud AutoML demonstrated an ease of use. This is also Google’s attempt to even out the cloud market and catch up with rivals AWS & Microsoft in the cloud computing market. 3) Last year, Google’s AutoML project built a computer vision system: Earlier last year in May, Google Brain researchers announced the creation of AutoML, an artificial intelligence (AI) that’s capable of generating its own AIs. In December 2017, the company announced a new computer vision system built by AutoML that outperformed man made state-of-the-art-models. The computer vision system can improve how autonomous vehicles and next-generation AI robots “see” better and can be leveraged by companies that have no experience to build a computer vision system, tailored to their  needs. For eg. Radiologists can use CT scans to train a computer algorithm that identifies signs of lung cancer. 4) Google brings AI to healthcare: For several years, Google’s DeepMind has been making a play in the healthcare sector. Recently, London headquartered DeepMind announced it is taking the kidney failure prediction algorithms to the US Department of Veteran Affairs to combat AKI, acute kidney failure. Through the collaboration, DeepMind will have access to 700,000 plus medical records and will use its deep learning algorithms to predict AKI. Earlier last month, a Google Brain team released a research paper demonstrating the use of computer vision to detect heart disease, a major cause of death across the world.","excerpt":"Does Google really lead the world in artificial intelligence? The search giant leads in commercial search and its technical expertise in machine learning and deep learning is unmatched – but in the broader world of artificial intelligence that includes the smart home device space underpinned by NLP & speech recognition, self-driving technology space (Google’s facing […]","categories":["IT Services"],"tags":["Google Translate"],"author_name":"Richa Bhatia","publish_date":"2018-03-02T03:35:32","publication_year":"2018","word_count":1276,"keywords":["data science","machine learning","artificial intelligence","AI","Google Translate","ML","computer vision","NLP","Kubeflow","deep learning","TensorFlow"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","NLP","computer vision","data science","Kubeflow","TensorFlow"],"url":"https:\/\/analyticsindiamag.com\/it-services\/what-stopping-google-monopoly-in-ai-limited-success-in-hardware\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":39063,"title":"How Unreal AI Is Using Proprietary Algorithms To Turn Videos Into Usable Data","content":"This New Delhi-based startup, founded in 2018 by Saurabh Singh and Nischal Gaba aims to decentralise artificial intelligence (AI) on Google and AWS cloud by developing proprietary AI neural network model that can run on low-cost edge devices. According to the founders, the pressing challenge before them was to run machine learning models on low-cost edge devices so as to reduce their dependence on the cloud for intelligence. After a year of hard work and deep research, the youngsters were able to address the challenge and when they showcased it to their friends and told them that they were able to run the models using $5 Raspberry Pi, their friends found it ‘unreal’. “AI on the edge is the next frontier in AI.  Just like with the advent of PCs in 70s decentralised computing power from mainframe computers we believe that our proprietary AI methods can help decentralise AI and make every device intelligent in itself,” Saurabh Singh, Co-Founder of Unreal AI said when asked about the motivation behind establishing the startup. Taxing ML  models Prior to the genesis of Unreal AI, the pair was working on an AI and NLP powered-chatbot that could answer queries by searching for words or phrases within videos, “For example, you could ask “When was this taught in this video lecture?” and the chatbot will pinpoint the location inside the video when it happened,” says Singh. However, the duo later discovered that their machine learning models for extracting information  were computationally taxing and soon they were looking at means to make their ML models run faster, “We realised that it’s a fundamental problem with ML that it requires a huge amount of computation power for real-time application,” Singh explains It is this quest to make their models faster and efficient that led them to the problem statement “Is there a way to run deep learning models on the low-cost devices\/edge devices?” and finally to the creation of Unreal AI. Building a sleek and efficient system Currently, the startup’s unique product offering is an easy to install security device that uses face recognition to authenticate employees at locations such as offices and schools that uses deep machine learning models to achieve this. The product uses face recognition for authentication and for added transparency. It also takes a snapshot of the moment when the person was authenticated which can later be reviewed for close inspection. The speed of the device also enables the device to identify as many as seven people at a time. This added layer of transparency and speed is a bonus for the device claims Singh and adds that it sets the product apart from conventional authentication systems which can identify just one face at a time. “Vizor ID uses a pipeline of multiple CNNs that have been implemented using our proprietary methods which enable them, otherwise computationally expensive, deep learning models to run on the low-cost devices,” explains Singh who depended on open source frameworks like Caffe, Keras, Pytorch etc to design the system. However, he does point out that the initial challenge before them was to design a sleek system that was easy to install and ensure that the work is done seamlessly, this also posed the challenge of working with substantially underpowered for real-time deep learning applications such as face recognition. Another challenge that they encountered was the client’s’ reluctance to send their video footage over cloud fearing security concerns, hence,  the team had to devise a mechanism to make the system work offline,  “In order to make Vizor ID work we used our own proprietary methods that enables deep learning models to run on low-cost devices. Our methods are a result of over a year of research that we invested at the start of this company. Thanks to our proprietary methods, Vizor ID can now run on low-cost devices such as the $35raspberry Pi!” Singh exclaims. The road ahead Though it took them over a year’s time to set up the startup, within a short time of inception, it has closely worked with over a dozen clients and has taken up over 50 projects. Recently, Unreal AI was one among the eight startups which graduated from Zeroth. AI’s three-month-long cohort. Market expansion and installation of 100 plus Vizor ID devices is their focus for 2019 and are in talks with Chinese hardware manufacturers for developing cheaper hardware systems to achieve this, “This year we plan on installing 100 Vizor ID devices. We are talking to hardware manufacturers in China to build these devices. We plan on making Vizor ID available in New Delhi and Bengaluru,” Singh concludes.","excerpt":"This New Delhi-based startup, founded in 2018 by Saurabh Singh and Nischal Gaba aims to decentralise artificial intelligence (AI) on Google and AWS cloud by developing proprietary AI neural network model that can run on low-cost edge devices. According to the founders, the pressing challenge before them was to run machine learning models on low-cost […]","categories":["Deep Tech"],"tags":["face recognition","ML","ml algorithms"],"author_name":"Akshaya Asokan","publish_date":"2019-05-13T07:50:16","publication_year":"2019","word_count":768,"keywords":["artificial intelligence","ml algorithms","machine learning","AI","neural network","PyTorch","ML","Keras","NLP","Aim","deep learning","face recognition"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","neural network","NLP","Aim","PyTorch","Keras"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/how-unreal-ai-is-using-proprietary-algorithms-to-turn-videos-into-usable-data\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10054263,"title":"What You Missed from Genpact’s Record-Breaking ‘Dare In Reality’ Hackathon","content":"The Dare in Reality Hackathon 2021, powered by Genpact in partnership with Envision Racing and MachineHack, challenged participants to develop machine learning models that predict lap times in a Formula E qualifying race. The contest  closed on 22 November having received a record number of entries. Within two weeks, the hackathon received more than 5,200 participants and 10,000+ submissions. Currently, a panel of judges are selecting the winners. Interestingly, the level of competitiveness in the final score is also marginally high (i.e. close to 0.01 to 0.001 difference) in the leaderboard ranking. Also, the hackathon had one of the highest prize amounts compared to any other hackathons hosted previously. The ‘Dare In Reality’ was indeed a first-of-its-kind hackathon ever hosted globally. Moreover, hackathon entrants showcased their advanced data analytics and AI skills, creating machine learning models to help the Formula E team predict lap times in the important qualifying sessions that determine where on the grid each driver starts a race. The hackathon winners will receive a range of cash and tech prizes, including: First prize: $7,000 Second prize: $5,000 Third prize: $3,000 Fourth prize: iPad Air 128 GB Fifth prize: Apple AirPods What Did You Miss? Touted as the largest machine learning hackathon globally, ‘Dare In Reality’ gave participants a unique opportunity to work on one of the most complex datasets out there, curated by Envision Racing and Genpact. The hackathon was the first time Genpact and Envision Racing gave people outside the partnership the opportunity to help the team make even better use of its data and analytics. The hackathon welcomed participants from all across the globe, including the US, China, Russia, Germany, Vietnam, India and South Korea. The entrants also got to compete with experienced professionals. Some of them were leaders in the data science industry with titles like the head of analytics, CTO, and senior data analysts. Besides experienced professionals, the hackathon also saw participation from various premium institutions like IITs, IIMs, NITs and others. Formula E is the world’s first all-electric international single-seater championship. The race consists of 12 teams and 24 drivers. In every race, drivers and their engineers need to make rapid decisions on how to use their energy, overtake or defend, respond to competitors, and more. To help Envision Racing make the right decisions on the racetrack and beyond, Genpact’s data science and machine learning experts use sophisticated algorithms to provide the team’s drivers and engineers with invaluable insights they need to improve their competitive edge. The hackathon presented participants with the challenge of predicting what drivers’ lap times would be in the all-important qualifying sessions. Ahead of every race, drivers compete to establish their place on the starting grid. Being able to anticipate the lap times for each driver based on the conditions of the track, weather, driver’s experience and other critical and evolving factors offers an invaluable advantage. Our hackathon entrants put data and analytics to work to address this challenge. We will announce the names of all the winners of the Dare In Reality Hackathon 2021 shortly, alongside details of the winning approach and solutions. If you missed this hackathon, don’t worry. We’ll have more that you can sign up for next year.","excerpt":"Within two weeks, the hackathon received more than 5,200 participants and 10,000+ submissions.","categories":["Deep Tech"],"tags":["data science hackathon","data science hackathon global","Envision Racing hackathon","Genpact","Genpact EVR Hackathon","Genpact hackathon","global data science hackathon","global hackathon grand prize","global machine learning hackathon","Hackathon","hackathon for data scientists","Hackathon for hiring data scientists","hackathon of the year","Hackathons","hackathons in India","Hackathons India","Machine learning hackathon","Machinehack","Machinehack Hackathon","New Hackathon For Data Scientists","Weekend Hackathon","World Wide Hackathon Event"],"author_name":"Amit Naik","publish_date":"2021-11-26T13:00:00","publication_year":"2021","word_count":534,"keywords":["API","global machine learning hackathon","Genpact","Weekend Hackathon","Machine learning hackathon","data science hackathon global","global hackathon grand prize","New Hackathon For Data Scientists","hackathon for data scientists","R","global data science hackathon","data science","hackathons in India","Envision Racing hackathon","Genpact EVR Hackathon","data science hackathon","analytics","Genpact hackathon","hackathon of the year","Go","machine learning","AI","Hackathon","Machinehack Hackathon","Hackathons India","programming_languages:R","Machinehack","Hackathons","World Wide Hackathon Event","programming_languages:Go","Hackathon for hiring data scientists"],"extracted_tech_keywords":["AI","machine learning","data science","analytics","R","Go","API","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/what-you-missed-from-genpacts-record-breaking-dare-in-reality-hackathon\/","complexity_score":4,"technical_depth":9,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":35131,"title":"Top 5 Tech Conferences In India For Women","content":"The technology sector across the globe is notorious for gender parity. However, things are looking up in India, as more and more number of companies, as well as groups, are aiming for diversity. Over the past couple of years, the number of women techies have skyrocketed significantly. The fact cannot be denied that women are doing wonders in the technology space — whether it is about AI or about coding a game.  And we shouldn’t forget that among all the factors, the tech conferences across the world have also played a vital role in increasing women participation in the area of science and technology. Analytics India Magazine lists down some of the best technology conferences for women in India. The Grace Hopper Conference (GHC) India 2018 Organised by Anita B.org, the GHC India is one of the at the largest technical conference for women in Asia. Last year the conference was held during November 14, 15 and 16 2018 in Bengaluru and showcased some of the latest technology products and solutions, from the world of The GHCI 18 startups and large organizations.  From building\/expanding network to delivering the best tech talks, the GHC provides a platform for all the women in technology space. AnitaB.org (formerly Anita Borg Institute for Women and Technology) is a global nonprofit organization based in Palo Alto. The organisation was founded by computer scientists Anita Borg, PhD and Telle Whitney, PhD. The Rising 2019 Organised by Analytics India Magazine, the Rising 2019 is the biggest meeting of women Data Science leaders. The conference will not only serve as a forum for exchange knowledge and experience but will also focus on building a better idea for women participating in STEM. The Rising will also highlight the achievements of women in the field of data science. Furthermore, from a series of talks to informal sessions, this one-day conference is going to empower women and help them in the development of leadership skills. The Rising 2019 is going to be held in Hotel Taj, Bengaluru on March 8, 2019. IEEE WINTECHCON – 2018 WINTECHCON is one of the most sought after tech conferences for women. Organized by IEEE, the conference provides an exclusive forum for women in the technology space to showcase their work and share their knowledge and experience. From demonstrating DIY projects or network with women engineers from other organizations, WINTECHCON provides everything under one roof.  Last year, the conference was held on September 28 in Bangalore. Girls In Tech India – Catalyst Conference Girls in Tech (GIT) is a global NGO that focuses on the engagement, education, and empowerment of women in technology and entrepreneurship. The organisation has also marked its presence in India as it has launched its India Chapter in Hyderabad. With a vision to support and raise the visibility of women in STEM, entrepreneurship, and innovation, the organisations conduct competitions, conferences, hackathons, boot camps etc. Catalyst Conference by GIT  is one of the best conferences that feature keynotes, panel discussions, and breakout sessions by leaders at the forefront of the technology and business sectors. It is definitely a can’t miss. IEEE WiE IEEE Women in Engineering (WiE) is a global network of over 20,000 members in over 100 countries with a mission to promoting women engineers and scientists, and inspiring girls around the world to follow their academic interests in a career in engineering and science. Its International Leadership Summit is one of the best conferences in the Women in Tech space. The conference provides regional opportunities to foster networking, mentorship, and collaboration. In 2017, the summit was held in Goa by IEEE Goa College of Engineering Student Branch with technical support by IEEE India Council. and in 2018, on 7th & 8th – September it was held at Le Meridien Kochi, Kerala.","excerpt":"The technology sector across the globe is notorious for gender parity. However, things are looking up in India, as more and more number of companies, as well as groups, are aiming for diversity. Over the past couple of years, the number of women techies have skyrocketed significantly. The fact cannot be denied that women are […]","categories":["AI Trends"],"tags":["Tech","Women in Tech"],"author_name":"Harshajit Sarmah","publish_date":"2019-02-19T07:07:49","publication_year":"2019","word_count":630,"keywords":["data science","Go","AI","innovation","Tech","Git","Aim","Women in Tech","analytics","ViT","GAN","R"],"extracted_tech_keywords":["AI","data science","analytics","Aim","R","Go","Git","GAN","ViT","innovation"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-5-tech-conferences-in-india-for-women\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10169140,"title":"Claude Mobile App is Getting a Voice Mode Along With File Support","content":"Much on the lines of OpenAI’s voice feature, the voice mode in Anthropic Claude’s mobile app is nearing release, TestingCatalog reported on Sunday. It is available in a push-to-talk method and in multiple voices. The voice mode also supports web search. When active, the search feature displays a list of sources along with the output, just like its desktop counterpart. The output is supposedly broken down into bullet points to make it easier for users to follow. Users can also scroll through their responses in a scrollable view. BREAKING 🚨: Claude's voice mode is now fully functional and supports web search and file uploads. It comes with push-to-talk and scrollable text views. It will be quite a big upgrade for the Claude mobile app!* Not available to the public yet pic.twitter.com\/LHsXEQtHqL— TestingCatalog News 🗞 (@testingcatalog) May 4, 2025 Additionally, the voice mode includes support for file uploads. Users can add images or documents and then talk to Claude using voice with the files as context. The report notes that the voice mode in Claude’s mobile app does not support interruptions when speaking to it. One has to tap the button manually, send input when needed, and control the flow of the conversation. While this differs slightly from ChatGPT’s voice mode, the report states that Claude handles voice input reliably and is more stable than many voice assistants. Recently, Microsoft also added free and unlimited access to its Voice Mode and the Think Deeper features in Copilot. The features added were powered by OpenAI’s o1 reasoning model. Meanwhile, Perplexity also added a new voice assistant feature, which allows users to use a web browser and multi-app actions to assist them with a set of tasks. The voice mode in AI chatbots is useful for real-time translation, mock interviews, clearing writer’s block, helping prepare for a meeting, becoming a tour guide, and much more. For instance, a student in India used ChatGPT to speak in Kannada to bargain for an auto ride. Considering Anthropic’s Claude is losing ground to Google Gemini, a voice mode feature could help attract new users and encourage existing ones to stay.","excerpt":"Claude’s mobile app will be getting a sweet upgrade.","categories":["AI News"],"tags":["Claude"],"author_name":"Ankush Das","publish_date":"2025-05-05T16:57:37","publication_year":"2025","word_count":354,"keywords":["Anthropic","ChatGPT","Go","TPU","OpenAI","AI","chatbots","Claude","RAG","GPT","R"],"extracted_tech_keywords":["AI","ChatGPT","OpenAI","Anthropic","RAG","chatbots","TPU","R","Go","GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/claude-mobile-app-is-getting-a-voice-mode-along-with-file-support\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10054716,"title":"This Is The Golden Age Of Machine Learning: Priya Ponnapalli, AWS","content":"Machine learning forms an important segment of AWS. To foster more innovation and development in this area, AWS has established a Machine Learning Solutions Lab, which helps customers build machine learning solutions to address some of their pressing challenges. The ML Solutions Lab works backwards and delivers a roadmap of ML use cases and their implementation. In particular, the ML Solutions Lab offers solutions for personalisation and recommendations, computer vision, fraud prevention, and supply chain optimisation, among others. Analytics India Magazine caught up with Priya Ponnapalli, senior manager at Machine Learning Solutions Lab. ML Journey and AWS “I have been working with Amazon ML Solutions Lab for three and a half years. This lab works with customers to help them adopt ML in their business. We brainstorm with them, identify their highest-value use cases, partner machine learning experts with these customers and help implement them. We work with customers all across the industries — healthcare, manufacturing, finance, media and entertainment, sports — on very many kinds of problems and really help their businesses move forward,” said Ponnapalli. Ponnapalli told us that right from the start of her career, her focus has been implementing machine learning solutions to real-life situations and working at the intersection of different and diverse disciplines like cancer cure, art, pharma, finance, etc. “The opportunity to work with a breadth of customers working in different fields is what attracted me most about joining AWS,” she said. A graduate in Electronics and Communication engineering, Ponnapalli first came to the US for her masters. She enrolled for MS in Electrical and Computer engineering at the University of Texas, Austin and went on to pursue her PhD at the same university. It was during this time that Ponnapalli worked at the Genomic Signal Processing Lab, which reaffirmed her interest and passion in the field of machine learning. “I loved signal processing, specifically genomic signal processing. Applying signal processing and computational mechanisms to genomics that was coming from high throughput technologies intrigued me. In fact, I took that class in my first semester at the University of Texas. Here is where I found my love for applying computational mathematical techniques to data coming from different sources. This began my journey to data mining and machine learning. Throughout my PhD, I developed algorithms to work on large scale data for unsupervised learning models. I have always been interested in real-world impact. So, I decided to pursue a career which would allow me to apply ML to various applications,” said Ponnapalli. Of the various projects that Ponnapalli is handling at AWS, she finds the ML solutions applied to sports most fascinating. Currently, her team is working with some of the largest sports organisations in the world, including the National Football League, Formula 1, Six Nations Rugby, and Swimming Australia, among others. Citing an example of AWS’ work in this area, Ponnapalli said, “One of the best examples is our work with NFL on player health and safety, where we combined NFL’s vast trove of data and their deep expertise in football with AWS’ ML and cloud computing expertise. We are working to transform sports, and we are doing this by analysis of player injury, game rules, rehabilitation and recovery. All this data will eventually be open to researchers, equipment manufacturers, trainers, coaches and medical professionals to serve as the framework of future innovation. I see this going beyond football to other sports and eventually be available in general.” Present and Future of ML “We are in the golden age of machine learning. We see businesses going from piloting ML projects to it having an impact on production. Many use cases of ML are in production that is helping companies stay competent and resilient. It’s no longer the question of whether you should have an ML strategy but how quickly can put their ML strategy to action,” believes Ponnapalli. When asked about how companies should approach and build their data strategy, Ponnapalli listed four important points: Getting the data in order.Identifying the right ML use cases that deliver the most value for that business.Developing the culture of ML in their organisation — this could be through upskilling workforce and educating both technical and business leaders.Embracing the culture of ML, where iterating on things, embracing failures and repeating to see that transformation is at the crux. Her advice to aspiring machine learning engineers — persevere and practice continuous upskilling. Her top resources for getting started are — Neural Networks and Deep Learning by Michael Nielsen, Dive into Deep Learning, and Deep Learning with Python by Francois Chollet.","excerpt":"I have always been interested in real-world impact. So, I decided to pursue a career which would allow me to apply ML to various applications","categories":["Global Tech"],"tags":["nfl"],"author_name":"Shraddha Goled","publish_date":"2021-12-02T18:00:00","publication_year":"2021","word_count":762,"keywords":["machine learning","AWS","AI","neural network","cloud computing","ML","computer vision","nfl","Python","deep learning","analytics"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","neural network","computer vision","analytics","cloud computing","AWS","Python"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/this-is-the-golden-age-of-machine-learning-priya-ponnapalli-aws\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":60075,"title":"MachineHack Winners: How These Data Science Enthusiasts Solved The Recent ‘Women In AI’ Hackathon","content":"In yet another successful conclusion to our latest hackathon dedicated to the Women in AI community, MachineHack last week concluded its 24th hackathon down the line — Food Quality Assessment: Women in AI Hackathon as an effort to encourage the women in Analytics and Machine Learning space. The hackathon provided an exciting opportunity for the participants to win exclusive passes to AIM’s upcoming Women In Analytics Conference, The Rising 2020. The hackathon was well received by the Data Science and Machine Learning community with over 530 registrations and close to 90 active participants, including both male and female enthusiasts. Out of the 89 submissions, three women in AI topped our leaderboard, we will introduce you to them and the approach they took to solve the problem. #1: Indira Raju Indira has over 4 years of experience in the IT industry. In the course of these 4 years, Indira has been a PHP coder, a Data Analyst, a Data Scientist and a Business Analyst. She started out as a Research Assistant at IIT Kharagpur, where she worked as a PHP coder and soon entered the analytics space to work as a Data Analyst in the same institute. Her growing interest in the analytics field led her to enrol in a Certification Program in Big Data Analytics\/Optimisation with INSOFE after which she joined Nielson as an Executive Data Scientist. Indira now works as a Business Analyst for HP Inc. Indira often participates in hackathons which she believes would help her in staying abreast of the latest techniques in Machine Learning and Data Science. Approach To Solving The Problem Indira explains her approach briefly as follows: Created multiple features to get to the targetSince there is an overlap of dates in train and test set, past and future inspection results and reasons can be used to predict the correct result, however, in production or real-world scenarios, future results can not be used.An aggregation of  15 selected models gave the best leaderboard scores.6 k fold cross-validation with LightGBM4 k fold with LightGBM5 k fold with CatBoost This resulted in an average cross-validation score of 0.13 using the features based on future results. Without future results, a cross-validation score of 0.19 was obtained, which scored 0.18 on the leaderboard. Get the complete code here. #2: Varshinee Venkatesan Varshinee has over 2.5 years of experience in the computer science domain working on technologies like web development, AR\/VR, game development, big data, etc. It was when she got into the realm of big data that she became fascinated with data handling and preprocessing colossal amounts of data. From there on Varshinee had been keenly interested in the Data Science domain. Varshinee shared her data science experience and how helpful her current organisation Sirius Computer Solutions is. Approach To Solving The Problem Varshinee explains her approach to the problem as follows: “This particular problem was so interesting. There is a saying that data science is about 80% handling data and 20% algorithms. I deny that. I would say that data science is about 90% handling data and 10% algorithms,” says Varshinee. Since it was a multiclass classification problem, a deep neural network with softmax activation was used at the output layer. Performed visualisation to know the importance of the features. Used DNN to train the data, which did not produce the expected results. Calculated the loss for each and every class and found out that two classes gave poor entropic loss. Checked the relevance and difference between the classes and added a few extra features.Generated new features by combining ‘date’ with other features, which drastically improved the score. Different classification algorithms were tried, checking the performance of each model.Tuned and optimised the algorithm to prevent overfitting and underfitting. “I would like to thank the Analytics India Magazine and MachineHack community a lot for coming up with this wonderful hackathon with a cool problem statement and encouraging all women data science enthusiasts to participate in this Hackathon. Hackathons like this will kindle more interest among data science enthusiasts, which in turn paves the way in creating more data scientists,” she told us about her experience on MachineHack Get the complete code here. #3: Mayuri Lashkare Mayuri is a final year Bachelor of Computer Science student. Having been introduced to Data Science by her brother, she started her data science journey by diving into OpenIntro Statistics, an open-source textbook for introductory statistics. She then expanded her knowledge from various educational platforms like Coursera and Kaggle MicroCourses. Approach To Solving The Problem Here is an outline of Mayuri’s approach to solving the problem: Started with basic EDAData Preprocessing to handle NaN valuesFeature selection and Feature generation.Tried a number of algorithms like RandomForest, XGBoost and CatBoost. “This was a wonderful experience for me because I got a lot to learn and at the same time apply my knowledge in this competition. Thanks to the AIM and MachineHack team for this opportunity; I look forward to participating in more hackathons. I also enjoy reading AIM articles and find them very helpful and informative,” she spoke about her experience with MachineHack Get the complete code here.","excerpt":"In yet another successful conclusion to our latest hackathon dedicated to the Women in AI community, MachineHack last week concluded its 24th hackathon down the line — Food Quality Assessment: Women in AI Hackathon as an effort to encourage the women in Analytics and Machine Learning space. The hackathon provided an exciting opportunity for the […]","categories":["Deep Tech"],"tags":["big data certification","Coursera","data analyst certification","multiclass classification","multiple classification statistics"],"author_name":"Amal Nair","publish_date":"2020-03-26T17:01:15","publication_year":"2020","word_count":852,"keywords":["data science","machine learning","big data certification","Coursera","AI","neural network","RAG","multiple classification statistics","multiclass classification","Aim","XGBoost","analytics","CatBoost","LightGBM","data analyst certification"],"extracted_tech_keywords":["AI","machine learning","neural network","data science","analytics","Aim","XGBoost","LightGBM","CatBoost","RAG"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/women-in-ai-hackathon-winners\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10062896,"title":"Intel® &#038; MachineHack to launch oneAPI Blogathon Contest","content":"Technology is changing rapidly, and it has significantly influenced the way we preserve the world. As a result, the words used to describe – particularly technical blogging – have noticeably become integral parts of the product life cycle, research and development, and building competitive advantage. In a bid to encourage more and more people towards technical blogging\/writing, Intel®, in association with MachineHack, is set to launch a two-month-long oneAPI Blogathon Contest from March 16 to June 15, 2022. This is a purely technical blogathon, where Intel® is encouraging participants to create meaningful content, with the utmost brevity, on given topics, such as oneAPI, Migrating from CUDA* to oneAPI Data-Parallel C++ (DPC++), Intel® AI Analytics Toolkit etc. The blogathon will be a great opportunity for participants to not only showcase their skills and expertise on Intel® oneAPI tools but also advance their knowledge about Intel® oneAPI products and services, as the participant will have to research the subject to gain a deeper understanding. In addition, the entries will be a gateway for new contributors to become productive members of the Intel® community. The best entries will stand a chance to win prizes worth INR 3 lakh, iPad mini, Amazon gift vouchers and more. Sounds exciting, right? All you need to know about “oneAPI Blogathon Contest” Start Date: March 16, 2022, at 6:00 PMEnd Date: June 15, 2022, at 6:00 PMWinners Announcement: June 10th, 2022 Click here to participate in the blogathon. Participants need to upload the blogs (with a minimum of 1000 words and with #oneAPI hashtag) on the MachineHack platform and Medium and propose them on Intel®’s Stack Overflow channel for the Intel® team to review. Participants will also be receiving likes & views on their blogs from various readers on the MachineHack platform, Intel®’s Stack Overflow channel & Medium. In addition, the candidates will have the chance to answer a few oneAPI related questions, as an advocate, on Intel®’s Stack Overflow channel to get additional points and improve their ranking. Note: If you are relatively new to the story category, we recommend you attend bootcamp\/tutorials in the ‘category’ section and the materials in the link to understand and learn more about oneAPI and on-demand recordings to upskill on Intel® AI toolkits offerings. Topic categories Technical writing on the following topics: oneAPIMigrating from CUDA* to oneAPI Data Parallel C++ (DPC++)Intel® AI Analytics ToolkitIntel® oneAPI HPC ToolkitIntel® FPGA Add-on for oneAPI Base ToolkitIntel® Distribution of OpenVINO™ toolkit Evaluation The entry should be original (unique, not plagiarised) and material (not a republication) published between March 16, 2022, to June 15, 2022 (both dates inclusive).The blog must have in-depth technical content containing examples, technical details, and useful & working source code.The blog must contain relevant scenarios and use cases on the Intel® oneAPI topics.Do not submit more than one entry per category.The entry should be in English.oneAPI Blogathon contest is open only to South Asia Countries [India, Bangladesh, Bhutan, Nepal, and Sri Lanka] Australia, New Zealand,  and Southeast Asia Countries [Indonesia, Malaysia, the Philippines, Singapore, Thailand]The entry should not have multiple authors.A combined editor panel from Analytics India Magazine and Intel® Team will assess the entries.The cumulative score will factor in the number of likes and views the entries received on the MH platform and the evaluation by the Intel® and Analytics India Magazine team. Note: MachineHack, Analytics India Magazine and Intel® reserve the right to publish and share the content on Intel® online resources, newsletters, and social media. Disqualification Submissions from multiple accounts will lead to disqualification. Judges will only entertain submissions from one account per participant. No ghostwriting. Discrepancies will lead to disqualification. More than 5% plagiarism leads to disqualification. Click here to participate in the blogathon. Rules The minimum word limit is 1,000.Participants must use the hashtag #oneAPI while publishing the blog on MachineHack, Medium and Intel®’s Stack Overflow channel. The article must contain code examples and explanations with screenshots.Participants can submit as many entries within the competition timeframe (one entry per category – in total of 6 entries).Only use Intel® oneAPI, oneAPI Data Parallel C++ (DPC++), its toolkits & components.The articles must be reshared in Intel®’s Stack Overflow channel: https:\/\/stackoverflow.com\/collectives\/intel and Medium, Github, dev.to OR any other blogging platform.Write a technical blog or article on one of the categories while covering instances, scenarios or case studies of oneAPI, Migrating from CUDA* to oneAPI Data Parallel C++ (DPC++), Intel® AI Analytics Toolkit, Intel® oneAPI HPC Toolkit, Intel® FPGA Add-on for oneAPI Base Toolkit, and Intel® Distribution of OpenVINO™ toolkit.Employees from Intel®, Intel® subsidiary and its partner, MachineHack and Analytics India Magazine cannot participate in the blogathon.Submit your story with your blog details on or before June 15 2022.The expert panel (from Intel® & MachineHack) will review the content and rate the submissions. The winners will be announced in June 10th, 2022. Prizes Top 3 entries will get iPad mini (64 GB)The next 25 entries will get Amazon vouchers worth 5000 INR. oneAPI Blogathon Contest by Intel® & MachineHack has been designed to provide a lot of exposure and experience to participants on Intel® oneAPI products and services and help them become productive members of the growing Intel® community, which in turn will help participants in their future career endeavours. So, what are you waiting for? If you are a tech blogger or just good at expressing your technical knowledge with words, then register now for this blogathon contest here.","excerpt":"The best entries stand a chance to win prizes worth INR 3 lakh, iPad mini, Amazon Gift vouchers and more.","categories":["Deep Tech"],"tags":["oneAPI"],"author_name":"Amit Naik","publish_date":"2022-03-16T18:00:00","publication_year":"2022","word_count":899,"keywords":["CUDA","Go","AI","Git","RAG","C++","analytics","GitHub","R","oneAPI"],"extracted_tech_keywords":["AI","analytics","RAG","CUDA","R","Go","C++","CUDA","Git","GitHub"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/intel-machinehack-to-launch-oneapi-blogathon-contest\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10040318,"title":"LinkedIn Open-Sources This New Library For Time Series Forecasting","content":"Recently, LinkedIn introduced a new open-sourced Python library, Greykite, to provide flexible, intuitive and fast time series forecasts. The library was developed to support the forecasting needs of the online professional networking platform. Greykite library provides a framework to develop a robust forecast model using outlier\/anomaly preprocessing, grid search, exploratory data analysis, benchmarking, feature extraction and engineering, evaluation, and plotting. For a few years now, forecasting business metrics and quantifying their volatility has become a go-to technique for most organisations. Through the long-term forecast, organisations can have an outlook and expectations about future growth or future resource requirements. In contrast, short-term forecasts can be used to detect anomalies in the system. Tech behind Greykite contains a simple modelling interface that facilitates data exploration and model tuning. The library provides intuitive forecasts through its flagship algorithm, known as Silverkite. Silverkite is highly customisable and includes tuning parameters to capture diverse time series characteristics. In an official blogpost, LinkedIn said, “The Silverkite algorithm works well on time series with (potentially time-varying) trends and seasonality, repeated events\/holidays, and\/or short-range effects. At LinkedIn, we’ve successfully applied it to a wide variety of metrics in different time frequencies (hourly, daily, weekly, etc.), as well as various forecast horizons, e.g., 1 day ahead (short-term) or 1 year ahead (long-term).” Figure 1. Architecture diagram for Greykite library’s main forecasting algorithm, Silverkite The professional networking platform uses time series forecasts for various processes, such as to provision sufficient infrastructure to handle peak traffic, to optimize budget decisions by forecasting growth of various markets, to set business metric targets and track progress for operational success, among others. The developers at LinkedIn designed this library in order to solve these types of problems. Besides Silverkite, Greykite also supports Prophet, which is an open-source algorithm for forecasting time series data. The Prophet algorithm is developed by Facebook and is robust to missing data and shifts in the trend as well as handles outliers well. Features The Greykite library comes with various capabilities for model assessment, validation, and testing that are appropriate for the time series context. The library offers components that could be used within other forecasting libraries or even outside the forecasting context. The library includes a number of features- Flexible design: Greykite provides time series regressors to capture trend, changepoints, autoregression and more. Intuitive interface: The library produces interpretable output, with model summary to examine the individual regressors as well as component plots to visually inspect the combined effect of related regressors. It provides powerful plotting tools to explore seasonality, interactions, changepoints, etc. It also provides model templates (default parameters) that work well based on data characteristics and forecast requirements (e.g. daily long-term forecast).Fast training and scoring: Greykite facilitates interactive prototyping, grid search, and benchmarking. Grid search is useful for model selection and semi-automatic forecasting of multiple metrics.  Extensible framework: The open-source library provides an extensible framework that exposes multiple forecast algorithms in the same interface, making it easy to try algorithms from different libraries and compare results. The same pipeline provides preprocessing, cross-validation, backtest, forecast, and evaluation with any algorithm. Wrapping Up The Greykite library provides a highly customizable algorithm Silverkite for time series forecasting. The library also provides intuitive tuning options and diagnostics for model interpretation. According to its developers, the library is extensible to multiple algorithms, and facilitates benchmarking them through a single interface. The developers are planning to add other useful open-source algorithms in the future to give users more options to choose from, through a unified interface.","excerpt":"Recently, LinkedIn introduced a new open-sourced Python library, Greykite, to provide flexible, intuitive and fast time series forecasts. The library was developed to support the forecasting needs of the online professional networking platform.  Greykite library provides a framework to develop a robust forecast model using outlier\/anomaly preprocessing, grid search, exploratory data analysis, benchmarking, feature extraction […]","categories":["AI Trends"],"tags":["Python Library","Time Series Forecasting"],"author_name":"Ambika Choudhury","publish_date":"2021-05-19T12:00:00","publication_year":"2021","word_count":583,"keywords":["Go","TPU","programming_languages:R","AI","programming_languages:Go","Time Series Forecasting","GAN","Python","programming_languages:Python","Python Library","R"],"extracted_tech_keywords":["AI","TPU","Python","R","Go","GAN","programming_languages:Python","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/linkedin-open-sources-this-new-library-for-time-series-forecasting\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":23106,"title":"NITI Aayog To Open 30,000 Atal Tinkering Labs To Promote New Tech In Education Sector","content":"Representational image. (AIM) In an effort to promote more “tinkering” in schools across the country the National Institution for Transforming India (NITI) Aayog has signed a Statement of Intent (SOI) with global software giant SAP. The programme which will fall under the Atal Innovation Mission (AIM) would enable SAP to adopt 100 Atal Tinkering Laboratories (ATL) to promote advanced technologies such as, internet of things, artificial intelligence, blockchain, 3D and robotics, among others. #NewsFromNITI: NITI’s @AIMtoInnovate signs Statement of Intent (SoI) with @SAP to encourage innovation & entreprenuership across India, by adopting 100 Atal Tinkering Labs #AIMtoInnovate pic.twitter.com\/omTQh335Ck — NITI Aayog (@NITIAayog) March 27, 2018 Acknowledging the importance of the collaborations, Amitabh Kant, CEO of NITI Aayog, said in a statement, “India’s growth for the next few decades will depend on the innovations coming out of these tinkering labs. Tomorrow’s ideas come from students, who will disrupt industries and creative sectors with new technologies and processes which will thrust India into the leadership position for technological creativity and innovation. This is possible only through effective public-private partnership and we are delighted to have SAP partner with us on AIM to nurture the future talent of our country”. Kant told a newspaper, “As many as 2,400 schools across the country had already been selected to set up Atal Tinkering Labs and we will take this number to 30,000 in next three years.” The program is aimed at enabling students to learn science, technology, engineering and mathematics education among secondary school students in the year 2018. Ramanathan Ramanan, mission director at AIM, added, “Such partnerships are a win-win situation for AIM and its partners. This partnership shall help support and boost innovations from the ATLs. The support to Atal Incubators will catalyse adoption of the innovations to commercial scales, ensuring deployment of the innovations to the domestic and international markets quickly and efficiently.” SAP employee volunteers will coach and mentor students in advanced technology topics and the design lab of SAP Labs India, will provide trainings and allow students to gain hands-on experience with technology gadgets. NITI Aayog’s AIM envisions the creation of a collaborative ecosystem, where students, teachers, mentors and industry partners work to facilitate innovation, foster scientific temper and an entrepreneurial spirit in the children of today, who will go on to become successful contributors to nation-building tomorrow. The Atal Innovation Mission (AIM) is Government of India’s flagship initiative to promote a culture of innovation and entrepreneurship in the country. AIM is mandated to create an umbrella structure to oversee innovation ecosystem of the country and revolutionising the innovation eco-system — touching upon the entire innovation life cycle through various programs.","excerpt":"In an effort to promote more “tinkering” in schools across the country the National Institution for Transforming India (NITI) Aayog has signed a Statement of Intent (SOI) with global software giant SAP. The programme which will fall under the Atal Innovation Mission (AIM) would enable SAP to adopt 100 Atal Tinkering Laboratories (ATL) to promote […]","categories":["AI News"],"tags":["amitabh kant","atal innovation mission","atal tinkering labs","Narendra Modi","NITI Aayog","SAP"],"author_name":"Prajakta Hebbar","publish_date":"2018-03-29T07:32:10","publication_year":"2018","word_count":441,"keywords":["amitabh kant","Go","artificial intelligence","Rust","programming_languages:R","AI","innovation","SAP","RAG","Aim","atal innovation mission","ViT","Narendra Modi","atal tinkering labs","R","NITI Aayog"],"extracted_tech_keywords":["AI","artificial intelligence","Aim","RAG","R","Go","Rust","ViT","innovation","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/sap-atal-tinkering-labs-soi-niti-aayog-amitabh-kant\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10119014,"title":"Apple Releases Four Open Source LLMs with OpenELM Series of Models","content":"Apple has open sourced OpenELM, a collection of Efficient Language Models (ELMs). OpenELM utilises a layer-wise scaling approach to efficiently distribute parameters within each layer of the transformer model, resulting in improved accuracy. Click here to check out the model on Hugging Face. OpenELM models were pre-trained using the CoreNet library. The models come in 270M, 450M, 1.1B, and 3B parameters, both pre-trained and fine-tuned according to instructions. The pre-training dataset consists of RefinedWeb, deduplicated PILE, a subset of RedPajama, and a subset of Dolma v1.6, totaling approximately 1.8 trillion tokens. Please review the licence agreements and terms of use for these datasets before utilising them. https:\/\/twitter.com\/ClementDelangue\/status\/1783107571294900300 For instance, with a parameter budget of around one billion parameters, OpenELM demonstrates a remarkable 2.36% increase in accuracy compared to OLMo, while requiring only half the pre-training tokens. In benchmarking, modern, consumer-grade hardware was used, with BFloat16 as the data type. CUDA benchmarks were conducted on a workstation equipped with an Intel i9-13900KF CPU, 64 GB of DDR5-4000 DRAM, and an NVIDIA RTX 4090 GPU with 24 GB of VRAM, running Ubuntu 22.04. To benchmark OpenELM models on Apple silicon, an Apple MacBook Pro with an M2 Max system-on-chip and 64GiB of RAM, running macOS 14.4.1, was employed. Token throughput was measured in terms of tokens processed per second, including prompt processing (pre-fill) and token generation. All models were benchmarked sequentially, with a full “dry run” generating 1024 tokens for the first model to significantly increase the throughput of generation for subsequent models. The entire framework, including training logs, multiple checkpoints, pre-training configurations, and MLX inference code, has been made open-source, aiming to empower and strengthen the open research community, facilitating future research efforts.","excerpt":"The models come in 270M, 450M, 1.1B, and 3B parameters, both pre-trained and fine-tuned according to instructions.","categories":["AI News"],"tags":["Apple","LLMs","Open Source AI"],"author_name":"Mohit Pandey","publish_date":"2024-04-24T19:53:13","publication_year":"2024","word_count":282,"keywords":["CUDA","Hugging Face","programming_languages:R","AI","LLMs","Apple","ML","RTX 4090","Open Source AI","Aim","ai_frameworks:Hugging Face","R"],"extracted_tech_keywords":["AI","ML","Aim","Hugging Face","CUDA","R","CUDA","RTX 4090","ai_frameworks:Hugging Face","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/apple-releases-four-open-source-llms-with-openelm-series-of-models\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10014777,"title":"Top AI-based Mental Health Apps Of 2020","content":"The year 2020 was challenging to everyone in so many ways as the world went into lockdown, and people had to get accustomed to completely new environments. Sitting at home, people felt lonely, anxious, and depressed, which led to a soar in the number of mental health issues. The lockdowns also meant that one could not seek in-person help from their psychiatrists or therapists. During such difficult times, many people resorted to using mental health apps to help themselves keep calm. Mental health apps are making significant progress and becoming more sophisticated. Many of them now even use artificial intelligence. AIM collated these AI-based apps that were launched in or made news in 2020 or were predominantly used during the lockdowns. Wysa Wysa is an AI-based conversational chatbot that has been trained using 100-million odd conversations to understand user inputs. The app provides research-backed, widely used techniques like cognitive behavioural therapy (CBT), dialectical behavioural therapy (DBT), and meditation support for users with depression, stress, anxiety, sleep and other mental health needs. Due to pandemic, the app observed an 80% rise in the installations as compared to last year. Backed by Google in series A funding, Wysa made the app’s tool packs for isolation and anxiety, free of cost until we get through the pandemic. Check out the app for Android here and for Apple here. Price: Free of cost Woebot Woebot helps a user think pragmatically through various situations by providing step-by-step guidance using CBT. It also helps the user learn more about oneself, using intelligent mood tracking. This year, the app launched a COVID-19 program called Perspectives, where it provides more guided meditation techniques and practical tips on overcoming cabin fever or connecting with others while social distancing. There were also themes added to deal with grief and economic hardships. Check out the app for Android here and for Apple here. Price: Free of cost Happify Happify has a unique approach designed to address mental health. It hosts interactive activities, some of which are games and other exercises led by a digital AI coach, Anna. This year the app introduced ‘adherence fidelity’ into Anna, for which the app was granted a patent. This feature uses natural language processing to detect when a user is slipping away from the intention of activity and gently guides them to be more adherent. This has helped in attaining maximised health outcomes. Check out the app for Android here and for Apple here. Price: Free of cost Elomia With the lockdown being enforced due to COVID-19, the fear of self-isolation led to an individual in the UK die by suicide, which led the co-founders to launch this app. Elomia is an AI-driven therapy chatbox, predominantly work via Facebook Messenger, that acts as a companion to help people struggling with anxiety and sadness. It has been developed by psychologists to understand what one is saying and build a conversation in a way that can help the users feel better using Cognitive Behavioural Therapy (CBT). The app is a medium to help people, who can’t see a therapist or have to self-isolate, make the situation a bit more bearable. The app is not available to download in India, but one can access it on a desktop browser at $6.99 per month. Check out the app here. Price: $6.99 per month BioBase This app, developed by BioBeats, is used more by enterprises than individuals. The BioBase app collects data from a wearable device called BioBeam, and then the real-time data is analysed using AI. This helps monitor one’s mental-wellbeing as well as physical health to provide live feedback and insights. A study published this year found out that the use of this app reduced the average length absentees at a workplace by 31% and stress-related work absences to zero. Check out the app for Android here and for Apple here. Price: Free of cost Ginger The app collects user’s behavioural data in terms of the duration of their talking, sleeping, or exercising to get clues on the person’s mental health. Based on this, the app provides increased access to high-quality behavioural health care by powering their team of expert coaches, therapists and psychiatrists with machine learning and AI technology. With the onset of COVID-19, the app announced three programs that include — free behavioural health coaching to frontline workers, make the Ginger’s COVID-19 in-app resources freely available, and waive annual commitments for new employer clients. Check out the app for Android here and for Apple here. Price: Free of cost Wrapping Up Technology, as it becomes smarter, can help people address their mental health issues in sophisticated ways. There are many apps available on mobile phone stores apart from the ones mentioned above that can cater to your specific needs. However, it is important to keep in mind that some mental health illnesses cannot be solved on apps or online. One should use technology for guidance, but seeking proper professional help when needed is critical.","excerpt":"The year 2020 was challenging to everyone in so many ways as the world went into lockdown, and people had to get accustomed to completely new environments. Sitting at home, people felt lonely, anxious, and depressed, which led to a soar in the number of mental health issues. The lockdowns also meant that one could […]","categories":["AI Trends"],"tags":["AI App","AI for good","AI for mental health","human touch to ai","Mental Health"],"author_name":"Kashyap Raibagi","publish_date":"2020-12-20T16:00:00","publication_year":"2020","word_count":828,"keywords":["Go","API","artificial intelligence","machine learning","ELT","AI","AI for good","Git","AI for mental health","RAG","Aim","Mental Health","AI App","human touch to ai","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","Aim","RAG","R","Go","Git","API","ELT"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-ai-based-mental-health-apps-of-2020\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10102013,"title":"TensorFlow 2.15: Latest Updates","content":"In the latest update of TensorFlow 2.15, several significant improvements have been introduced. One notable enhancement is the performance optimisation of oneDNN for CPUs on Windows x64 and x86 platforms. This optimisation is automatically enabled for X86 CPUs and can be customised using an environment variable. These optimisations may lead to slightly different numerical results but aim to boost performance. Another key improvement is the expansion of the ‘tf.function’ type system. This update allows for more control and flexibility when working with TensorFlow functions. It introduces ‘tf.types.experimental.TraceType’ to handle custom TensorFlow function inputs, ‘tf.types.experimental.FunctionType’ to comprehensively represent function signatures, and ‘tf.types.experimental.AtomicFunction’ for fast TensorFlow computations in Python. TensorFlow’s data processing capabilities have also been refined. The option ‘warm_start’ has been moved to ‘tf.data.Options’, simplifying data handling and offering more control. Moreover, TensorFlow 2.15 introduces bug fixes and additional changes. One notable addition is the TensorFlow Quantizer in the TensorFlow pip package, which aids in quantizing models. Additionally, it brings an option to make the gradient output of specific functions sparse instead of dense. TensorFlow Lite TensorFlow Lite (tf.lite) has received several updates, including support for broadcasting in certain operations and the promotion of the `tflite::SignatureRunner` class, which simplifies working with named parameters and computations within TF Lite models. This enhancement removes its experimental status. Keras enhancements Keras, a high-level neural networks API, has received updates as well, including bug fixes, new ops in ‘tensorflow.raw_ops’, and the addition of the ‘tf.CheckpointOptions’ argument for executing callbacks during checkpoint saving. There’s also an option to control the behaviour of the eager runtime when executing parallel remote function invocations.","excerpt":"This release optimises oneDNN performance, expands tf.function features, and improves data handling in tf.data. TensorFlow Lite and Keras also receive updates, making the platform more efficient and user-friendly for machine learning tasks.","categories":["AI News"],"tags":[],"author_name":"K L Krithika","publish_date":"2023-10-26T12:19:49","publication_year":"2023","word_count":265,"keywords":["API","TPU","Keras","AI","neural network","Python","Aim","ai_frameworks:TensorFlow","TensorFlow","R"],"extracted_tech_keywords":["AI","neural network","Aim","TensorFlow","Keras","TPU","Python","R","API","ai_frameworks:TensorFlow"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/tensorflow-2-15-latest-updates\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10076251,"title":"Should We Really Care About Explainability In AI?","content":"In the recent past, it is no secret that AI, especially cutting-edge deep learning algorithms, have been plagued by the explainability crisis. In May 2021, two researchers from computer scientist Su-In Lee’s Lab of Explainable AI for Biological and Medical Sciences in Seattle published a paper in Nature titled, ‘AI for radiographic COVID-19 detection selects shortcuts over signal’, discussing the far-reaching effects of this issue. With the increased usage of AI in critical industries like healthcare and finance, explainability has become even more critical. In the heyday of the COVID-19 crisis, researchers started developing AI systems to detect COVID-19 accurately. But the technique that the systems used remained behind the curtain. When tested in labs, these systems worked perfectly but when used in hospitals, they failed. The paper explained how to use AI successfully in medical imaging, it was essential to crack the ‘black box’ in ML models. Models have to be trustworthy and interpretable. This was all well and good. Until explainable AI became a marketing point for most companies selling these AI systems. However, it seemed like AI didn’t just have an explainability problem, explainability itself had a problem. Saliency maps produced by popular methods, Source: Lancet Current problems with explainability techniques Last year in November, Lancet Digital Health published a paper by MIT computer scientist Marzyeh Ghassemi claiming that explainable AI itself couldn’t be understood easily. Current explainability techniques were only able to produce “broad descriptions of how the AI system works in a general sense” but when asked to justify how individual decisions were made, the explanations were “unreliable and superficial”. The paper goes on to use another commonly used explainability method called ‘saliency maps’ as an example. These maps produce a heat map on an image of the algorithm and highlight the parts that were most reliant on AI for a prediction. The paper demonstrates how one heat map came to the conclusion that the patient had pneumonia but couldn’t reason why it thought so. Heat map produced by a post-hoc explanation method for a deep learning model designed to detect pneumonia in chest x-rays, Source: Lancet Interpretable models not Explainable AI Experts from other sectors too have asked for explainable AI to be looked at from a different perspective. Agus Sudjianto, executive vice president and head of corporate model risk at Wells Fargo, put it this way, “What we need is interpretable and not explainable machine learning.” Sudjianto fully acknowledges that models can fail and when they fail in a sector as vital and directly impactful as banking, they can have serious consequences on a client. For instance, someone in an urgent need of a loan might have their request rejected. “This is a challenging task for complex machine learning models and having an explainable model is a key enabler. Machine learning explainability has become an active area of academic research and an industry in its own right. Despite all the progress that has been made, machine learning explainers are still fraught with weakness and complexity,” he stated. Sudjianto goes on to argue that what ML needed in this case was interpretable ML models that are self-explanatory. Agus Sudjianto, Executive Vice President, Head of Corporate Model Risk, Wells Fargo A new approach In a paper co-authored by Wells Fargo’s Linwei Hu, Jie Chen, and Vijayan N. Nair,  titled, “Using Model-Based Trees with Boosting to Fit Low-Order Functional ANOVA Models,” a new approach to solving the interpretability problem was proposed. The earliest solutions to interpretability were all post hoc and included low-dimensional summaries for high-dimensional models that were complex. These were ultimately incapable of giving a full bodied picture. The second solution given for model interpretability was using surrogate models to fit simpler models to draw information from the original models which were complex. Examples of these models are LIME—first proposed in a 2016 paper by Marco Tulio Riberio—which is based on linear models and offers local explanations; and locally additive trees for global explanation. Another new path forward was to use ML algorithms to fit these easily interpretable models that are extensions of generalised additive models (GAMs). The logic behind using these algorithms was to slow down the usage of normally complex algorithms for large-scale applications with pattern recognition and instead use nonparametric models with low-order that were able to figure out the structure. Another paper co-authored by Sudjianto explained why inherently Interpretable machine learning (IML) models should be adopted due to their transparency rather than black-boxes as model-agnostic explainability isn’t really easily explained or defended under scrutiny from regulators. The GAMI-Net architecture, subject to a) sparsity constraints on the number of main effects and pairwiseinteractions, b) heredity on the candidate pairwise interactions; and c) marginal clarity to avoid confusion betweenmain effects and their child pairwise interactions, Source: Research The study suggested a qualitative template based on the feature effects and the constraints of the model architecture. The study listed the design principles for the development of high quality IML models, with examples from tests done on ExNN, GAMI-Net, SIMTree, and so on. Eventually, these models were proven to have greater practical application in high-risk industries such as the financial sector. So, has AI explainability largely become a fraudulent practice? Sudjianto seems to think so. “There is a lack of understanding of the weaknesses and limitations of post hoc explainability which is a problem. There are parties who promote the practice who are less than honest, which has misled users. Since the democratisation of the explainability tools, matters have become worse because users are not trained and not aware about the problem,” he stated. Sudjianto also believes that the problem lies both with the concept of explainable AI as well as the prevalent techniques. “Post hoc explainer tools are not trustworthy. The techniques give a false sense of security. For high risk applications, people should design in interpretability,” he added.","excerpt":"Current explainability techniques were only able to produce “broad descriptions of how the AI system works in a general sense” but when asked to justify how individual decisions were made, the explanations were “unreliable and superficial.”","categories":["IT Services"],"tags":["explainability in AI","Explainable AI"],"author_name":"Poulomi Chatterjee","publish_date":"2022-10-03T13:00:00","publication_year":"2022","word_count":976,"keywords":["Go","machine learning","AI","ML","Git","explainability in AI","Ray","Aim","deep learning","Rust","Explainable AI","R"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","Aim","Ray","R","Go","Rust","Git"],"url":"https:\/\/analyticsindiamag.com\/it-services\/should-we-really-care-about-explainability-in-ai\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10056739,"title":"EXL Adds Scale In Data, AI, And Cloud Engineering By Acquiring Clairvoyant","content":"EXL, global analytics and digital solutions leader, has acquired Clairvoyant, a global data, AI, and cloud services firm. With this acquisition, EXL has strengthened its capabilities by adding cloud enablement and data engineering expertise. This will further support its clients in healthcare, insurance, retail and banking & financial services. The addition of Clairvoyant’s capabilities in data engineering, AI\/ML operations and cloud will also accelerate EXL’s data-driven strategy, expanding critical cloud operations capabilities to help clients drive large-scale digital transformation projects and generate faster value from modern cloud-based data and analytic platforms. “Clairvoyant has established itself as a leader in data and cloud engineering, bringing many of the key structural components we need to expand our offerings and help our clients harness the power of data to enable better and faster decision making,” said Rohit Kapoor, EXL Vice Chairman and Chief Executive Officer. Founded in 2012, Clairvoyant has consistently been named among the most innovative and fastest-growing privately held companies in the U.S. The company is headquartered in Phoenix, AZ, and has offices across North America and India. It employs over 550 people and serves clients in the financial services, healthcare, retail, and tech industries. “The need to quickly aggregate and interpret data has grown exponentially across industries in recent years,” said Vivek Jetley, Executive Vice President and Global Head of Analytics at EXL. “Clairvoyant is a great addition to our fast-growing Analytics business and scales our ability to help clients turn data into business advantage. We are excited about combining EXL’s domain expertise, data management and analytics capabilities and Clairvoyant’s strengths in data engineering, AI\/ML operations and cloud,” Jetley mentioned. The transaction was completed in the fourth quarter of 2021. Clairvoyant is expected to be accretive to 2022 adjusted diluted EPS and will be reported in our Analytics segment. “By integrating our capabilities with EXL’s broad reach and expertise, we will be able to add value across a much wider range of clients and unlock exciting new growth opportunities along the way,” said Chandra Ambadipudi, Co-Founder and Chief Executive Officer, Clairvoyant.","excerpt":"Clairvoyant’s capabilities will accelerate EXL’s data-driven strategy, help its clients drive digital transformation, and generate value from data and analytic platforms.","categories":["AI News"],"tags":["Data Analytics","EXL","ML"],"author_name":"Meeta Ramnani","publish_date":"2021-12-22T10:19:29","publication_year":"2021","word_count":341,"keywords":["programming_languages:R","AI","ETL","data-driven","ML","digital transformation","Git","data engineering","analytics","Data Analytics","EXL","R"],"extracted_tech_keywords":["AI","ML","analytics","R","Git","data engineering","ETL","digital transformation","data-driven","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/exl-adds-scale-in-data-ai-and-cloud-engineering-by-acquiring-clairvoyant\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":20131,"title":"World’s First Robot Citizen Sophia To Visit IIT Bombay For College Festival","content":"Students of Indian Institute of Technology, Mumbai (IIT-B) will have the special privilege of hosting a very unique guest this year end. Sophia, world’s first robot to be recognised as the citizen of a country, will be attending IIT Bombay’s annual science and technology festival, titled, Tech Fest. The annual event is scheduled to be held from 29 to 31 December. Former Presidents of India, such as Pranab Mukherjee and Hamid Hamid Karzai, as well as Nobel Laureate Randy Schekman will be delivering lectures over the course of the festival. But the star of the event is guaranteed to be Sophia, who will be taking the stage on 30 December. She will field questions chosen from a pool of entries received through the online campaign, #AskSophia, launched specifically for the event. When asked as to why the invitation was extended to Sophia to be a part of the festival, Anirudh Poddar, manager (media and marketing), Techfest, told a newspaper,  “On 5 January 2014, we decided to invite BINA48 (Breakthrough Intelligence Via Neutral Architecture 48) the most advanced humanoid at that time for the festival. This time also we considered Artificial Intelligence as the most important area of technology and invited the world’s most advanced humanoid robot.” Said to be modelled on Hollywood legend Audrey Hepburn, Sophia was developed by roboticist David Hanson and his company Hanson Robotics. She was activated in April 2015 and came to international renown in October 2017 after she was granted citizenship by the Kingdom of Saudi Arabia. She had also addressed delegates at the Future Investment Initiative held in Riyadh after the same. During the post address interview with journalist Andrew Ross Sorkin, the humanoid-robot exhibited cheeky humour and responded with informed answers. She even took a dig at one of the most renowned critics of AI — Elon Musk. Before being known for receiving a one-of-a-kind citizenship, Sophia achieved minor notoriety when an alleged technical glitch led her to say that she would destroy humans.","excerpt":"Students of Indian Institute of Technology, Mumbai (IIT-B) will have the special privilege of hosting a very unique guest this year end. Sophia, world’s first robot to be recognised as the citizen of a country, will be attending IIT Bombay’s annual science and technology festival, titled, Tech Fest. The annual event is scheduled to be […]","categories":["AI News"],"tags":["AI (Artificial Intelligence)","humanoid robot","IIT Bombay","Quantum Computer","Robotics","sophia"],"author_name":"Jeevan Biswas","publish_date":"2017-12-27T13:06:04","publication_year":"2017","word_count":331,"keywords":["IIT Bombay","artificial intelligence","programming_languages:R","AI","Quantum Computer","Robotics","ViT","sophia","ai_applications:robotics","R","AI (Artificial Intelligence)","humanoid robot"],"extracted_tech_keywords":["AI","artificial intelligence","R","ViT","programming_languages:R","ai_applications:robotics"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/sophia-robot-iit-bombay-india\/","complexity_score":3,"technical_depth":6,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10056022,"title":"Council Post: Envisioning The Future of Deepfakes","content":"Just this quarter, Netflix viewers were taken on a ride with the film Red Notice. Wondering about the uncanny feeling you got while watching the film, or how the world created seemed a little off? Welcome to the universe of deep fakes. Between the deep augmented backgrounds and the excessive use of deep fake, the film has been tagged as its twin or deepfake. Instances like Red Notice are all fun and games, but such fake realities created with AI have come under major scrutiny in the past few years for security reasons. And rightly so, as you will realise today. This technology leverages deep learning models to create or manipulate audio, video or image media. These present hyper-realistic AI-generated images or videos of people doing or saying things that, in reality, they didn’t. Modern technology has created some of the best products of the decade, but as they develop at a fast pace, we need to take a step back, study its creation and recreate ethical boundaries. From Hardware To Data Augmentation Deep fakes are created on state-of-the-art deep learning models. To build such models, three major components need to align perfectly: sophisticated algorithms, hardware resources to support the model training and the data! Historically, even though we were advancing on building algorithms, hardware resources like GPU and TPU were a challenge to access for scientists. Fast forward a few decades, the hardware challenge has been replaced with data, the foundational rock upon which every intelligent machine is created today. And to build a trustworthy machine, we need to overcome the challenge of gathering thousands of tons of data. So, scientists started synthesising data and working on models that would require very little data, one-shot learning or zero-shot learning. This led to the development of data augmentation, a technique that allows scientists to synthesise new data from the available data, effectively addressing the requirement of diverse training data. Datasets are extremely crucial for use-cases involving unstructured data in the form of audio, image or video. Earlier, if there was any tampering done with audio, image, or video, it was easier to detect as it was mostly photoshopped. There was a superimposition of new audio, image, or video in this case. But with deep learning and data augmentation methods, we have a new class called deep fakes, which can be understood as AI-powered photoshopping. Deep Dive Into The Technology With AI-powered photoshopping, it is almost impossible to detect if the media is fake. And they are powerful, with the capacity to make undeniable changes. For instance, in March, the internet was shaken by a viral video of Barack Obama swearing at Donald Trump, until it was revealed that the video was indeed fake. Likewise, Tom Cruise fans had a blast watching the actor create viral TikTok videos, until that too, turned out to be a very well-created hoax. Deepfakes can eerily do things as small as changing specific features to as huge as creating entirely new videos of people saying and doing things. Let’s look at a list of all things possible with deepfakes: Changing facial featuresChanging facial expressionsSynthesising audioSynthesising facial expressionsSynthesising\/Eliminating entities from image\/videoSynthesising text data based on human writing patternsCreating new art based on the artist’s style – paintings\/songs, etc. These are extremely difficult to detect because of the precision of tampered aspects such as the skin tone, background merging, and audio features to match the original content. Having said that, the technology is still developing and tends to create loopholes in some features in the image\/audio\/video; humans can detect that. Through Rose Tinted Glasses Deep fakes are based on augmented reality and continue to be a marker of technology breakthroughs. The technology leverages state-of-the-art methods such as encoder-decoder, GAN or CNN, to create audio and visual deep fakes. While this is being used in multiple applications, we will discuss some of its most popular use cases. Deep fakes can re-create the art of classically renowned artists. Devices such as Timecraft can learn how a painting was produced and recreate it with precise brushstrokes. With a ninety per cent success over existing benchmarks, the device was trained on hundreds of minutes of painting time-lapses. Researchers are also leveraging technology to make the paintings talk, such as the famous recent case of da Vinci’s “Mona Lisa” rolling her eyes and talking. Researchers from Moscow’s Samsung AI Center and Skolkovo Institute of Science and Technology manipulated such portraits to talk and move by identifying facial landmarks in the painting using AI. Deep fakes are increasingly being used as a form of entertainment in films, videos and online games. In fact, people are also creating deep fakes of themselves as advanced virtual assistants. Virtually working companies are leveraging AI to create videos of their synthetic head or a virtual body double, termed virtual double ARIs. The assistant can speak and move its head while having the features of a real person. Deep fake also doubles up as a language augmented model that can translate across languages and give basic replies. The Sharper End Of The Sword But, the technology that can be used for entertainment can become real dystopian real soon. They can create real harm. We are already seeing the applications of this technology to infringe upon people’s privacies, create havoc by showcasing individuals making false statements, tricking people into deception, committing fraud and cybercrime. Not too long ago, $76 million were stolen from the Chinese government by criminals manipulating personal data and feeding the facial recognition system with deep faked videos. That’s not even the tip of the iceberg of ugly deep fake applications. This was just one of the ways of biometric hacking. Biometric devices and facial recognition are being used for security purposes across the globe. And criminals everywhere are creating deepfake videos to break through these barriers. Additionally, deep fakes kill privacy. The content is extremely convincing, and the regularly advancing technological development has made it a challenge to distinguish between actual and fake information. Considering the advanced technology needed to even classify them as fakes, laymen, if caught up in deep fake treachery, would not even be able to prove themselves easily. The technology infringes on the basic human rights to privacy or consent. Recently, a documentary on the life and death of the famous chef Anthony Bourdain came under a lot of controversies after the filmmaker used AI to narrate a few lines in the voice of Bourdain. The dead actor is certainly not present to consent to having his voice cloned, and yet, there he is, saying things he, in reality, did not. With an increasing number of ‘breaking news’ related to deep fake crimes or controversies, leaders and AI-product making organisations are starting to raise concerns over technological usage. Leaders Pitching In We have reached a point where there is hardly any distinction between reality and augmentation, making it high time to take the necessary steps that ensure no harm is done via augmented reality in the coming years. For brands, it is important to segregate research and production. The creators of such deep fake models need to do thorough research on the ethical and legal concerns regarding the applications of the model and ensure the product is meeting the requirements of safety and security. Leaders can additionally incubate mandatory workshops and lectures to educate the employees on the ethical and legal standards of AI to abide by. The future of deep fakes can be bright. But to ensure that deep fake applications are not misused, we ought to create boundaries. There is a pressing need to re-imagine a persona of the industry that has created guardrails for possibly dangerous applications. Deep fake models aren’t inherently harmful, but we need to ensure that we don’t make them so. For an organization that wants to protect itself from any attack on their application using deepfakes, they can tackle it from two angles: security and AI. Employ the right personas – Have a dedicated chief security officer who can train the employees to create awareness about deepfakes. The team should be skilled to understand how a deepfake attack can affect the organization. Training should also focus on social engineering to understand how the attack could appear, how to detect deepfake attacks and what could be the action plan to mitigate damage. Placing guardrails – Apart from having deepfake attack detection which could be post-event, guardrails should be put in place to avoid any such attacks. From a model perspective, the model being used in an application that is prone to be attacked using deepfakes, could have a pre-set deepfake detection algorithm. Risk Mitigation – As much as technology can be planned to be full proof, that doesn’t always work; the organization should be having risk mitigation plans. Methodologies to attenuate the impact of deepfake attacks should be in place. Impact Evaluation – In case of a deepfake attack, an assessment about the impact should be done to understand which users could be affected and in what capacity, which other applications could be compromised and if it could cause the organization any monetary loss. Collaborative effort – Similar to cybersecurity guidelines or trusted cloud principles where organizations come together to solve a complex challenge, avoiding deepfake attacks is also a collaborative effort. Organizations could partner with other firms to strengthen the guardrails of their technology and put in place various strategies which could be set as standards for other companies to follow. This article is written by a member of the AIM Leaders Council. AIM Leaders Council is an invitation-only forum of senior executives in the Data Science and Analytics industry. To check if you are eligible for a membership, please fill the form here.","excerpt":"Just this quarter, Netflix viewers were taken on a ride with the film Red Notice. Wondering about the uncanny feeling you got while watching the film, or how the world created seemed a little off? Welcome to the universe of deep fakes. Between the deep augmented backgrounds and the excessive use of deep fake, the […]","categories":["AI Features"],"tags":["deepfake trend","DeepFake videos","deepfakes","deepfakes and AI","GAN model","GANs"],"author_name":"Aishwarya Srinivasan","publish_date":"2021-12-16T16:00:00","publication_year":"2021","word_count":1620,"keywords":["data science","DeepFake videos","TPU","AI","deepfakes","virtual assistants","RAG","deepfakes and AI","deep learning","GANs","analytics","Aim","deepfake trend","GAN model","R","zero-shot learning"],"extracted_tech_keywords":["AI","deep learning","data science","analytics","Aim","RAG","zero-shot learning","virtual assistants","TPU","R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/council-post-envisioning-the-future-of-deepfakes\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10129341,"title":"Mistral AI Releases MathΣtral, New Model for Math Reasoning and Scientific Discovery","content":"Mistral AI has unveiled MathΣtral, a specialised 7B model designed for advanced mathematical reasoning and scientific exploration. Released under the Apache 2.0 license, MathΣtral pays homage to Archimedes on the occasion of his 2311th anniversary this year. We also release Mathstral 7B:– Strongest math model for its size, up to 75% MATH with a reward model– Apache 2.0 Weights: https:\/\/t.co\/lgieMM9PBr– Blog: https:\/\/t.co\/Nql6O3ibKh pic.twitter.com\/O7iRvVOpAJ— Theophile Gervet (@theo_gervet) July 16, 2024 MathΣtral is tailored to tackle complex, multi-step logical reasoning challenges in STEM fields. Developed in collaboration with Project Numina, the model inherits capabilities from Mistral 7B, achieving state-of-the-art performance across industry-standard benchmarks. Notably, it scores 56.6% on MATH and 63.47% on MMLU, demonstrating superior reasoning capacities within its size category. Detailed benchmarks highlight MathΣtral’s robust performance improvements with increased inference-time computation. For instance, MathΣtral 7B achieves significant accuracy enhancements, scoring 68.37% on MATH through majority voting and 74.59% with a strong reward model among 64 candidates. MathΣtral is available for immediate use and adaptation using Mistral AI’s tools. Developers can deploy the model through mistral-inference for initial exploration and fine-tune its capabilities with mistral-finetune. The model’s weights are accessible via HuggingFace, facilitating straightforward integration into academic and research projects. By releasing MathΣtral to the scientific community, Mistral AI aims to foster advancements in mathematical problem-solving and support academic endeavors. This initiative underscores Mistral AI’s commitment to promoting specialized model architectures and their practical applications in scientific discovery.","excerpt":"It scores 56.6% on MATH and 63.47% on MMLU, demonstrating superior reasoning capacities within its size category.","categories":["AI News"],"tags":["mistral"],"author_name":"Siddharth Jindal","publish_date":"2024-07-16T21:12:30","publication_year":"2024","word_count":236,"keywords":["Go","programming_languages:R","AI","ML","programming_languages:Go","Aim","mistral","R"],"extracted_tech_keywords":["AI","ML","Aim","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/mistral-ai-releases-math%cf%83tral-new-model-for-math-reasoning-and-scientific-discovery\/","complexity_score":4,"technical_depth":7,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":22679,"title":"Fractal Analytics Acquires Final Mile; Brings Together AI, Behavioural Sciences","content":"Artificial intelligence and analytics company, Fractal Analytics, on Wednesday announced the acquisition of behaviour architecture firm, Final Mile, for an undisclosed sum. The acquisition brings together the disciplines of data analytics, artificial intelligence and behavioural sciences. Srikanth Velamakanni, group chief executive and executive vice-chairman at Fractal Analytics, said, “Final Mile’s track record in influencing human behaviour through behavioural science, combined with Fractal’s data science and AI capabilities, will help our clients drive lasting behaviour change internally and externally.” Since its inception in 2000, Fractal Analytics has worked with several Fortune 500 companies across areas such as insurance, retail, consumer goods and financial services, among others, and has helped them integrate analytics and AI in their decision-making process. While it is headquartered in Mumbai, Fractal Analytics has offices in 14 locations and has over 1,200 employees. Biju Dominic, co-founder and CEO at Final Mile said, “Data and AI can dramatically improve our understanding of human behaviour, inform better hypotheses, and complement our work with clients in achieving sustained behaviour change. We see extraordinary opportunities to deliver greater value to our clients in partnering with Fractal over the long term.” Final Mile has worked with industries such as safety and sanitation, healthcare, financial services, e-commerce,and consumer goods, among others, since its founding in 2008. It has helped them drive behavioural changes through the application of the Nudge Theory. The theory conceived by Nobel prize-winning economist Richard Thaler, proposes that alterations to a consumers ‘choice architecture’, has the potential to influence their decision-making positively and help make better choices. “To fundamentally address a challenge or shape the decision-making process, one must identify and understand what truly drives behaviour. Organisations can achieve better outcomes by applying learnings from cognitive neuroscience, behavioural economics and design. That is our focus,” added Dominic. Commenting on the need for AI to gauge human behaviour better and acknowledging the potential of this acquisition to achieve it, Velamakanni added, “While artificial intelligence is becoming an integral part of products and services we consume, AI needs to incorporate better understanding of human behaviour to improve human-machine interface, simplify information consumption and drive lasting behaviour change.”","excerpt":"Artificial intelligence and analytics company, Fractal Analytics, on Wednesday announced the acquisition of behaviour architecture firm, Final Mile, for an undisclosed sum. The acquisition brings together the disciplines of data analytics, artificial intelligence and behavioural sciences. Srikanth Velamakanni, group chief executive and executive vice-chairman at Fractal Analytics, said, “Final Mile’s track record in influencing human […]","categories":["AI News"],"tags":["AI (Artificial Intelligence)","behavioural sciences","Data Analytics","Fractal Analytics","Mergers and Acquisitions"],"author_name":"Jeevan Biswas","publish_date":"2018-03-15T11:04:01","publication_year":"2018","word_count":354,"keywords":["Fractal Analytics","data science","Go","artificial intelligence","programming_languages:R","AI","programming_languages:Go","GAN","analytics","Data Analytics","behavioural sciences","Mergers and Acquisitions","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","data science","analytics","R","Go","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/fractal-analytics-acquires-final-mile\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10083224,"title":"Tredence Raises $175 Mn in Series B Funding","content":"Tredence, the data science and AI solutions company, recently announced that it has raised $175 million in Series B funding from Advent International to fuel data-led growth and AI value realisation for industries. The terms of funding involves that Advent will acquire a minority stake in the investment. Besides Advent, the existing investor Chicago Pacific Founders (“CPF”), a private equity firm, which made its initial investment in December 2020, will continue to be a shareholder in Tredence. As part of the transaction, Advent will also join the Tredence board. Read: AI & Analytics Service Companies Sees 4–5% Increase in M&A Alongside the co-founders of Tredence and the CPF, Advent will continue to invest in the company’s operational excellence, vertical and domain expertise, IP and accelerator library, and channel partner development. Tredence CEO Shub Bhowmick said, “Advent’s global reach, deep sector expertise, and vast experience in scaling businesses like ours through organic and inorganic growth will be invaluable to us as we look to drive continued business innovation.” Tredence was established in 2013 by Shub Bhowmick, Sumit Mehra, and Shashank Dubey with the goal of bridging the gap between insight delivery and value realisation by offering clients a differentiated approach to data and analytics through specialised solutions. The funding also comes at a time after Tredence told AIM that it is in expansion mode, where it has moved to a bigger facility in Bengaluru, alongside launching offices in Chennai and Delhi. Bhowmick said that they are planning to grow fairly aggressively in Toronto, Canada, where they are expected to see significant growth. At the time, the company claimed to have witnessed revenue growth of 100%, alongside hiring more employees. The global private equity investor, Advent, on the other hand, has made a series of investments in IT and information services including Encora, CI&T, NielsenIQ, Neoris, Sophos Solutions, Aareon, Canvia, and QuEST Global Services. Overall, Advent International has invested in over 400 private equity investments across 42 countries, and as of June 30, 2022, had $96 billion in assets under management. With 14 offices in 12 countries, Advent has established a globally integrated team comprising over 285 private equity investment professionals across North America, Europe, Latin America and Asia. The firm focuses on investments in five core sectors, including business and financial services; health care; industrial; retail, consumer and leisure; and technology.","excerpt":"The funding will fuel data-led growth and AI value realisation for industries.","categories":["AI News"],"tags":["Tredence"],"author_name":"Ayush Jain","publish_date":"2022-12-23T15:59:47","publication_year":"2022","word_count":390,"keywords":["data science","Go","funding","programming_languages:R","AI","innovation","Aim","analytics","GAN","Tredence","R"],"extracted_tech_keywords":["AI","data science","analytics","Aim","R","Go","GAN","innovation","funding","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/tredence-raises-175-mn-in-series-b-funding\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":32413,"title":"Top Indian Quora Writers To Follow For Data Science","content":"Banner source : studyclerk Data science is that new currency and asset which has taken over the technology sector across the globe. Data and data science practices have already significantly impacted the key aspects of business. Quora is an excellent means to assist people to answer their questions and provide inputs in the areas of technology, academics, project management, new trends in business and many more. In this article, we list down 10 Indian Quora writers Data Scientist aspirants can follow to bring the latest insights and data trends to their followers. Lalit Patel Holding a PhD in Physics, Lalit Patel also holds an MBA. He is an experienced person in Machine Learning nanotechnology. Presently working as Computer Audit AnalystState of Florida – DoR Mr.Patel and has worked in many reputed companies and taught at IIT Delhi for about 5 years and has two million answer views and is the top writer for the year 2018. Shweta Doshi Doshi is the Co-Founder of GreyAtom, Data Science Immersive Learning school. She believes that technology is often a tough subject for colleges and universities since the curriculum is frequently outdated and not regulated to the expectations of industry. GreyAtom was founded to bridge this gap, and make tech education more relevant as per the industry’s standards. She has 1m answer views and is very active in answering people about the emerging trends. Sudalai Rajkumar S An alumnus of IIM Bengaluru, Sudalai Rajkumar is presently working as a lead data scientist at H2O.ai. He is Data Scientist with vast experience in solving many real-world business problems across diverse domains. He is also a Kaggle Grandmaster in competitions and Kernels section. Some of his top achievements include being ranked topmost in recent hackathons platform and a top solver of Crowdanalytix platform. He has a 390.2k answer views on Quora. Adarsh Iyer Graduating from the University of Mumbai in electronics and communications engineering Adarsh is a data-driven professional. He is a specialist with vast experience in a range of  IT domain and has worked on Supply chain & Logistics, E-commerce and Food Processing Machinery Domains. He loves answering questions on Quora page and has 327.6k answer views. Abhinav Krishnan A graduate of College of Engineering Guindy, Krishnan has 5+ years of experience in Data Science, Python, Excel and SAS.  He is a Data Scientist with great experience in solving many real-world business problems across heterogeneous domains and has 305k answer views on Quora. Ratnakar Pandey An MBA from Indian School of Business (ISB), Ratnakar Pandey is India Head Analytics and Data Science at Kabbage Inc. He holds a Blog on Data Science using Python  RP’s Blog on data science. He has 15+ years of senior management in analytics and data science across Banking, Financial, Fintech, Retail, Technology and Healthcare verticals. Led global teams across 10+ countries and has 869.8k answer views on Quora. Vidita Mehta Vidita Mehta has completed her fintech course from Imarticus and graduated from Mumbai. She discusses the latest technology over Quora which she finds to be a great platform for discussion and sharing opinions and holds 594.7k answer views Akash Dugam He is a data science consultant who helps Data Scientist aspirants with tons of knowledge that he has gathered in the analytical industry. His area of interests include Artificial Intelligence, IoT and machine learning. He holds 1.6 million answer views.","excerpt":"Banner source : studyclerk Data science is that new currency and asset which has taken over the technology sector across the globe. Data and data science practices have already significantly impacted the key aspects of business. Quora is an excellent means to assist people to answer their questions and provide inputs in the areas of […]","categories":["AI Features"],"tags":["Data Science","Interviews and Discussions","Machine Learning","nanotech","nanotechnology"],"author_name":"Bharat Adibhatla","publish_date":"2018-12-28T12:24:15","publication_year":"2018","word_count":560,"keywords":["data science","artificial intelligence","machine learning","programming_languages:R","AI","data-driven","Machine Learning","nanotech","Python","analytics","programming_languages:Python","nanotechnology","Data Science","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","data science","analytics","Python","R","data-driven","programming_languages:Python","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/top-indian-quora-writers-to-follow-for-data-science\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10044216,"title":"Data-Intensive Applications Need Modern Data Infrastructure","content":"Gone are the days when applications were installed locally, had a handful of users at any one time, and only focused on basic data entry and retrieval. Modern applications live in the cloud and access and generate large amounts of data. This data needs to be aggregated, summarized and processed – and presented to the user in a way that is understandable, interactive, and served up in real-time. From the user perspective, a positive experience depends on data being highly available, consistent and secure – without compromising performance. To meet these needs, modern, data-intensive applications need a modern data infrastructure that can scale seamlessly as user numbers grow. Applications, Defined By applications, we really mean services (as in Software-as-a-Service) as nearly all new modern applications are now being built as services. With the move to microservices architecture, the boundary around what is an “application” becomes somewhat fuzzy. For purposes of this article, it is everything that goes into delivering the end-user experience. This includes the UX as well as all the backend services that make that UX possible. Applications have existed since the beginnings of the computing era and, at their heart, allow a user to accomplish a task. For example, any smartphone has many apps, all for specific tasks. The Uber app to call a car, a banking app to check accounts and transfer money, tools like Slack, Email, and Zoom that communicate with people at home or at work. What are the components of an application? There is the UX, the interaction model for how the user makes use of the app. There is the business logic, the rules, that govern that interaction. Last but definitely not least, there is the data. Data is the part that makes the application relevant to the user. This article focuses on how data has changed and evolved and why that evolution requires a modern approach to data infrastructure. In the early days, there was only the data inputted by the user. With the advent of continuous connectivity, applications moved off our personal computers to cloud services. This has allowed larger and more varied data sets to be incorporated into the experience. The data may be used to recommend, predict, incent, or surface opportunities that derive from insights and trends. Running as a backend service in the cloud and with such broad data, access provides an opportunity for new capabilities as well as a host of new challenges as applications become more data-intensive. But what does “data-intensive” mean? In physics, intensity is a measure of power over the surface area over time. Similarly, data intensity is measured over a set of dimensions. DimensionDefinitionRangeSize of the Working Data SetSize of data queried overLow = GBs,  High >100 TBsIngestion SpeedSLA on how many rows ingested\/secLow = <1k rows\/sec,  High > Mils rows\/secQuery LatencySLA on how fast the query has to runLow =Minutes,  High = MillisecondsQuery ComplexityHow many joins in the queryLow=0 joins  High > 5 joins Concurrent QueriesNumber of users or queries running concurrentlyLow = <5,  High = >100s Applications that have high values in two more of these dimensions or medium values in several of these dimensions are data-intensive. There are many examples of data-intensive applications made possible by this shift in data availability and how data is used. Stock trading applications are illustrative. These operations were possible in the past only by visiting a stock brokerage or trading company in person. Today’s applications not only access a user’s account information but also a variety of information about the market and portfolios – they can even provide predictive what-if scenarios. Digital Marketing has changed the advertising world significantly. With the ability to run many concurrent marketing campaigns, there is no end to the ideas you can test out. With access to rich demographic information, you can narrow your target segment and test specific messages and visuals. Digital marketing applications process the results and display them in a way so you can easily see what is working and what is not. Where Data Comes From and How it is Used To see the importance of data, we need to understand where it comes from and how it is used by the application. There are several ways data comes into existence: User Data: Data entered by the user or on behalf of the user3rd Party data: Data acquired to enrich the manual data. Typically loaded into the backend independently of the user application code.Telemetry Data: Data generated about the usage of the app. This data is captured as events created by the application and stored in the same system for later use.Aggregate Data: Aggregations over the other types of data. Aggregations can be sums, averages or more complicated aggregate functions. There are also many ways that data is used within an application: Lookups: Lookups are about getting a small piece of data out of the system quickly. There is typically an identifier (a name, an id, email, etc.), and the information is looked up with that id. Operational Databases and NoSQL systems are pretty good at this.Selective Queries: Selective queries are important to help you answer questions quickly and easily. This is where SQL (and relational algebra) are useful in making it easy to express your question and where NoSQL systems often get stuck. (For more details on the limits of NoSQL, read this blog post). Some examples of selective queries are: Who are the top salespeople in the organization?Who are the top players of a Fortnite game?Which users are experiencing a poor streaming experience? Aggregations: Aggregations are typically done in separate analytic systems but are increasingly included directly in applications as they have access to a wider set of data and expectations that it be in line with the application user experience. Operational Databases and most NoSQL systems (though there are a few that specialize in just this type of query) are typically not very good at this, especially as the data size scales. Data Warehouses do this well but don’t make well powering applications. Example aggregation queries are: What is the average sales price over the last 12 months?What was my overall return on my portfolio? Full-Text Query: Fuzzy searches, typically over unstructured data, look for approximate matches on exact words or words of similar semantic meaning. There are a few NoSQL systems that specialize in this, and some of the operational databases support it as well (but usually not at scale). Two Common Use Cases for Application Data Use Uber To the user, Uber doesn’t look like it has much data. In the application, the only data visible is user data such as name, phone number, favourites, and credit card). But on the backend, it actually uses all the other data types. It keeps track of all the users asking for cars (telemetry data), as well as all the drivers and their locations, current demand and associated pricing (aggregate data). It also has maps of the roads, rules for different cities, what events are happening, current road conditions, and more (3rd party data). Users look that 3rd party data to look up locations by name (full-text query), or they click on a favourite (lookup). It can also suggest locations from recent trips (selective query). The app has to process all this data in real-time to ensure a positive experience for the customer and ensure Uber and the drivers are making a profit at the same time. People think Uber is a car driving business, but it’s not. Uber is a logistics company, far more similar to UPS, FedEx and Amazon than a taxi company. And they do it all by acquiring, aggregating, summarizing and leveraging data. Banking Banking apps have evolved significantly since the early days of online banking. In the beginning, you could go to a website on your computer and see a limited number of your recent transactions from several days ago. Users checked the data infrequently (i.e. once a week). Now, everyone has an app on their phone and regularly checks account status (user data), deposits checks, and transfers money in and out of accounts. Credit card purchases are expected to update in real-time. We can also get summaries of spending, broken down by time, category and make comparisons over different time periods (aggregations). We can search for any vendor we have paid (Full-Text Search) or by the amount paid (Lookup). The app also does analysis and points out potential duplicates, changes in spending patterns, or increases in subscriptions. It pulls in data about the rest of the market (3rd party data) and lets me compare how I am doing relative to the market or others (Aggregation). Data-Intensive Application Requirements Building apps that make use of data in all these various ways is hard and comes with a lot of operational complexity. Following are some of the key modern application requirements that developers need to consider and conquer: Consistency (ACID): Maintaining consistency in an application is hard. Application developers find it much easier to depend on the data infrastructure to guarantee the data is consistent rather than have to do the checks in their application. If you are using an Operational Database, you are covered, but NoSQL and DW systems typically don’t do this well. High Availability: When offering a SaaS product, the provider is responsible for the availability of service level agreements (SLAs). Keeping a system running in the face of any type of error such as hardware, software, external environment events (i.e. hurricane takes out a data centre) is really hard. But customers have come to expect 24×7 access, no matter what happens. Dealing with the volume of data and all the different data types: Dealing with so many types of data is another big challenge – maybe one of the most significant. The considerations are complex and considerable — How do you ingest data in a way that meets your requirements? How do you transform it into a shape usable by your application? How do you guarantee the quality and consistency of the data? Which formats are the data in (JSON, CSV, Avro, Parquet, etc.), and how do you parse it? How much data is coming in and how is that rate growing over time, and what is it a function of? What is the SLA on how fast data has to come in and be available? What form do you need to store the data (relational, semi-structured, spatial, time-series, full text)? Solving all these requirements is beyond challenging. Even once you have a model that is working, growth in your business often causes bottlenecks and missed SLAs when the data infrastructure can’t handle the load. This is not a problem you want to have. You want a data infrastructure that supports the data sources, formats, and ingest performance you require and one that will grow as your application usage grows. Scaling with the data as you grow: Growth can happen in several dimensions (See this blog for more on scaling). You can have growth in the number of users, the amount of data per user, the rate of data ingestion, or the amount of queries per user. Additionally, growth in the different dimensions is not mutually exclusive. They often build on one another. To handle this growth, your data infrastructure should be a distributed system whose compute and storage resources can be easily scaled (preferably in an online way) to handle the growth. Security, Privacy and Data Ownership: When you run a SaaS service, you take responsibility for customer’s data. This brings with it security challenges and privacy issues. It also brings new possibilities. You can make use of that data or directly monetize it. Navigating these choices is tricky, and the data infrastructure must have the right capabilities to handle all the associated security and privacy requirements. Semantic understanding of data: As data moves through the different systems and is transformed and aggregated, it can be hard to track the semantic meaning of the data. This causes challenges for users of the data downstream. If the data infrastructure understands the schema and semantics of your data, it becomes discoverable through APIs, and it is tracked when it changes. This makes it much easier to manage as things evolve over time. Timeliness: The need for real-time and instant access changes what data we can expect and when we expect it. For example, when we swipe a credit card in a grocery store, the transaction should immediately appear in the banking app. When a flight is delayed, a notification should instantly appear on the user’s mobile device. If the application can’t deliver this information in real-time, the impact on user experience, application adoption, customer satisfaction, and ultimately, business revenue and success is enormous. Machine Learning (ML): ML is one of the exciting breakthroughs that resulted from access to large amounts of information and large amounts of computational power. It is all about figuring out things that would have been impossible to do by hand. Fraud detection and personalization are just a few examples of the many ways to utilize ML. There are a lot of tools for identifying and training models. But operationalizing those models is challenging as most of the toolsets fall short on running the models. The Modern Approach to Data Infrastructure Modern applications are data-intensive because they make use of a breadth of data in more intricate ways than anything we have seen before. They combine data about you, about your environment, about your usage and use that to predict what you need to know. They can even take action on your behalf. This is made possible because of the data made available to the app and data infrastructure that can process the data fast enough to make use of it. Analytics that used to be done in separate applications (like Excel or Tableau) are getting embedded into the application itself. This means less work for the user to discover the key insight or no work as the insight is identified by the application and simply presented to the user. This makes it easier for the user to act on the data as they go about accomplishing their tasks. To deliver this kind of application, you might think you need an array of specialized data storage systems, ones that specialize in different kinds of data. But data infrastructure sprawl brings with it a host of problems. See this blog for a walkthrough of the problems that pattern causes. What’s required is a database that is consistent, highly available, durable, and resilient. It needs to scale with you as your usage grows without forcing a costly re-architecture at every stage. It should allow you to secure your data and meet all your privacy requirements.It needs to handle the loading of data fast enough to meet your SLAs.It should support natively loading the data from all standard formats available. It must be capable of delivering all the analytics your applications need with no noticeable lag, no matter how busy the system is. It shouldn’t force you to make copies of the data and move them around to various systems. This is, admittedly, a tall order. But SingleStore, our fast unified database for data-intensive applications on any data, anywhere, is changing the way application developers are powering their modern applications. If you want to learn more about database systems that are capable of handling these requirements, check out our free trial. We’d love to power up your next data-intensive application. __________________________________________________ This blog post was written by Rick Negrin, Vice President Of Product Management at SingleStore. SingleStore is the single database for all data-intensive applications. Learn more at singlestore.com. Read the blog here.","excerpt":"SingleStore, our fast unified database for data-intensive applications on any data, anywhere, is changing the way application developers are powering their modern applications.","categories":["IT Services"],"tags":["data privacy use cases","Modern data infrastructure","modern database","modernising data platforms","singlestore","SingleStore blog","SingleStore data analytics","SingleStore database","SingleStore modern cloud architecture","SingleStore Modern Data Infrastructure","SingleStore modernising data platform","SingleStore unified database"],"author_name":"Rick Negrin","publish_date":"2021-07-27T12:00:00","publication_year":"2021","word_count":2607,"keywords":["SingleStore database","singlestore","Modern data infrastructure","SingleStore modern cloud architecture","Ray","modern database","R","SingleStore unified database","fraud detection","SingleStore modernising data platform","RAG","analytics","SingleStore data analytics","machine learning","AI","SingleStore Modern Data Infrastructure","ML","data privacy use cases","SingleStore blog","microservices","SQL","modernising data platforms"],"extracted_tech_keywords":["AI","machine learning","ML","analytics","Ray","RAG","fraud detection","microservices","R","SQL"],"url":"https:\/\/analyticsindiamag.com\/it-services\/data-intensive-applications-need-modern-data-infrastructure\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10101016,"title":"Why Cloud-Native Applications Need Observability","content":"Enterprises use security to safeguard their data from vulnerabilities. They have been deploying high-end enterprise-version antivirus suites to lockdown system security and hedge their bets against malware and all forms of web nastiness. Now, that the era of cloud-native applications is here, managing changes to cloud assets has become a universal pain felt by developers despite all the advancements in monitoring tools. In the recent past, tools have come to the rescue enabling them to troubleshoot and resolve developer issues. These platforms let organisations track, measure, and optimise the performance of their applications — commonly referred to as ‘observability’. It’s more common to talk about ‘observability’ as companies are trying to make the backend work more efficiently. But a year ago, Sawaram Suthar and Laduram Vishnoi at Acquire were facing issues visualising their cloud infrastructure due to the unavailability of a platform allowing them to do so. Sawaram Suthar, director, Middleware The platform being used by the team lacked real-time monitoring, issue detection, improved customer experience, security, cost optimisation, and data-driven decision-making, which became a huge task for the team. “We didn’t know what went wrong in the system. If it is down, then why is it so and what are the logs that we look at,” Suthar recalled in a conversation with AIM. “As developers, we kind of felt underserved at that point, as these platforms were mostly enterprise-focused rather than end-user focused,” he added. “That’s why we unified on the particular form which can help us visualise a complete data flow and started Middleware,” he said introducing the platform. “Observability is not new in the market. It’s been here for almost 20 years. But people have started using it now because they realise how important it is,” he added. The rising demand for observability tools is visible today as 86% of Indian organisations surveyed in 2022 saw observability as a key enabler to achieving core business objectives. The company is currently witnessing a surge in the Indian market as companies continue to go cloud native. As per reports, Indian firms spent an average of INR 370 crore on cloud between June 2022 and 2023. “For on-premise servers, it doesn’t require observability platforms. When cloud usage increases, there is a need for software which can visualise how this cloud will affect the services. That’s where we come in. As several companies are moving away from non-cloud and monoliths, the demand is going to be high in the upcoming year,” he stated. Newer and Cheaper The global cloud computing market dramatically grew by 635% from $24.63 billion in 2010 to $156.4 billion in 2020. “A few years ago, cloud adoption was still picking up. People were using monolithic environments, or maybe on-demand services,” Suthar pointed out. Suthar further highlighted the sheer size of the industry and the already existing dominant players. Early on, the team realised they required a lot of capital. Initially, they tested some different small models before jumping into visibility because they didn’t want to go directly competing with the big players. This strategy bore fruit as Middleware raised $6.5 million in seed funding in August 2023, right after graduating from Y Combinator’s Winter 2023 batch. The company now aims to expand its team from 25 to 50 within the next year and targets a series A funding round by mid-2024. Middleware Team “Existing competitors, built almost 10 years ago, are still not compatible with the latest technology like cloud-native and AI,” Suthar stated. As a startup backed by YC, it has access to most of OpenAI’s innovations and currently uses GPT-4. He further highlighted and elaborated on the three USPs of MW. Firstly, there’s an AI advisor that can pinpoint errors and recommend what needs to be done to fix them. The second is that the platform can predict incoming errors in the next few hours based on data. “That’s a unique differentiator for us. We can predict if a particular server is going to be down, that’s a feature that people really love,” Suthar explained. Lastly, Middleware offers its services at one-third of the cost compared to major competitors like Datadog, making it an attractive option for cost-conscious customers. “We are in the process of acquiring more customers in a couple of months. Around 35+ customers are actively using our platform,” he added. For the DevOps Headquartered in San Francisco, Middleware is one of the first observability platforms in the market that uses generative AI to accelerate issue identification and resolution. Suthar also recalled that the term observability hadn’t gained much popularity when they entered the market. “We started creating content around it to raise awareness. Initially we noticed that people don’t understand the word because people used to call it a cloud monitoring tool rather than observability,” he said. “People are accustomed to using platforms and metrics differently. It’s very hard to adapt to using tools, which give you a full stack of liberty. Now they also know that finding a threat is the biggest problem for DevOps or any developer. What used to take one hour, can now be solved within 10 seconds because you will have complete visibility to take action. Not only data, but you have the right suggestions too,” he concluded.","excerpt":"What required an hour earlier can now be resolved in 10 seconds, pointed out the director of Middleware, a unified observability platform","categories":["AI Trends"],"tags":[],"author_name":"Tasmia Ansari","publish_date":"2023-10-03T18:30:00","publication_year":"2023","word_count":871,"keywords":["Go","API","OpenAI","AI","cloud computing","RAG","Aim","generative AI","DevOps","R"],"extracted_tech_keywords":["AI","generative AI","OpenAI","Aim","RAG","cloud computing","R","Go","DevOps","API"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/why-cloud-native-applications-need-observability\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10095117,"title":"Synthesia Becomes a Unicorn with NVIDIA Backed Fundraise","content":"Synthesia is the latest AI startup to reach a valuation of $1 billion after raising a $90 million series C fund. The round was led by Accel Ventures with NVentures – NVIDIA’s venture capital arm along with existing investors Kleiner Perkins, GV, Firstmark Capital, Alex Wang, Olivier Pomel and Amjad Masad. Synthesia plans to allocate the funds towards AI research, furthering its partnerships with renowned educational institutions such as TUM in Munich and UCL in London. Philippe Botteri, the lead investor from Accel in the funding, said that, “Productivity can be improved because you are reducing the cost of producing the video to that of making a PowerPoint,” stating that TikTok, Netflix, and YouTube have made it easier for adoption of video. “While we weren’t actively looking for new investment, Accel and NVIDIA share our vision for transforming traditional video production into a digital workflow,” said Victor Riparbelli, co-founder and CEO of Synthesia. https:\/\/twitter.com\/synthesiaIO\/status\/1668582176860798983 Founded in 2017, Synthesia is a London-based digital media platform startup that creates AI generated videos. It was started by a team of researchers and entrepreneurs including Victor Riparbelli, Matthias Niessner, Steffen Tjerrild, and Lourdes Agapito, specialising in software development. Their innovative software enables individuals to create personalised digital avatars for various purposes, such as delivering corporate presentations, training videos, or even expressing compliments to colleagues, in over 120 diverse languages. Synthesia caters to enterprise customers such as Tiffany’s, IHG, and Moody’s Analytics. While the company does not publicly disclose specific sales or revenue figures, it has reported consistent triple-digit growth. To date, the platform has facilitated the creation of over 12 million videos. Synthesia also highlighted a remarkable 456% year-over-year increase in the number of users on its platform. Remarkably, it seems like no other company has been benefitting from the AI boom as much as NVIDIA. The company became the first chip firm to reach a trillion dollar valuation. Recently, NVIDIA also backed Cohere, reaching a valuation of $2.2 billion.","excerpt":"Synthesia plans to allocate the funds towards AI research, furthering its partnerships with renowned educational institutions such as TUM in Munich and UCL in London.","categories":["AI News"],"tags":["synthesia"],"author_name":"Mohit Pandey","publish_date":"2023-06-14T14:16:31","publication_year":"2023","word_count":326,"keywords":["synthesia","API","funding","AI","venture capital","Git","ViT","analytics","AI research","R","startup"],"extracted_tech_keywords":["AI","analytics","R","Git","API","ViT","startup","venture capital","funding","AI research"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/synthesia-becomes-a-unicorn-with-nvidia-backed-fundraise\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10080175,"title":"Burden of Technical Debt: Why it can Cost ML Developers Heavily","content":"A few days ago, Elon Musk, the freshly-minted owner of Twitter, tweeted apologizing for the slow performance of the mobile app. He blamed it on the software’s poorly batched remote procedure calls, tweeting – “App is doing >1000 poorly batched RPCs just to render a home timeline!” Btw, I’d like to apologize for Twitter being super slow in many countries. App is doing >1000 poorly batched RPCs just to render a home timeline!— Elon Musk (@elonmusk) November 13, 2022 Replying to this now-deleted tweet, Twitter engineer Eric Frohnhoefer disagreed with Musk and tweeted three reasons behind the slow app – features with little usage, time wasted over network responses, and years of accumulated tech debt. He further added – “Frankly we should probably prioritize some big rewrites to combat 10+ years of tech debt and make a call on deleting features aggressively.” This tweet by Frohnhoefer has shone the spotlight on technical debt – a major roadblock in software development. The debt is not limited to software development but extends to domains such as machine learning – a major reason behind developmental and scaling issues. What is technical debt? Technical debt, which Sterling Lanier, the CEO of TurnKey, described in an article, as a boring\/insomnia-inducing topic in software development – happens when teams opt for a quick and easy solution for a product to reach the market faster. More often than not, these quick fixes are not permanent resolutions, and the team ends up spending more time and resources. The term technical debt was coined by Ward Cunningham, software developer and one of 17 authors of the Agile Manifesto, which formed the basis for the creation of Wiki. Explaining how he came up with the concept of technical debt, Cunningham said, “With borrowed money, you can do something sooner than you might otherwise, but then until you pay back that money, you’ll be paying interest. I thought borrowing money was a good idea, I thought that rushing software out the door to get some experience with it was a good idea, but that, of course, you would eventually go back and as you learned things about that software, you would repay that loan by refactoring the program to reflect your experience as you acquired it.” Tech debt in machine learning With the rapid progress of machine learning, an accompanying trend has also emerged – that of maintaining and further scaling them. While developing machine learning systems is now easier, faster, and cheaper, maintaining these systems has become equally difficult and expensive. As the researchers from Google mentioned in their paper – Hidden Technical Debt in Machine Learning Systems – machine learning systems have a ‘special capacity’ for incurring technical debt. This could be attributed to the fact that these systems not only have maintenance problems of traditional code but have to also bear an additional set of challenges unique to machine learning. What’s more – these debts are often difficult to detect since they exist at the system level and not the code level. This means that typical methods that are used to play down code-level technical debts in software development are not sufficient to address machine learning challenges. One of the most significant technical debts in machine learning is feedback loops. It happens when the output of the model is fed back as the input – this leads to a form of analysis debt that makes it difficult to predict the behaviour of the model even before its release. Trailing close after feedback loops, in terms of how common they are, are garbage features. As the term suggests, these features don’t do much and contribute significantly to making the system bulkier – in short, absolute garbage. Very similar to garbage features are something called anti-patterns. A very small fraction of code in machine learning systems is actually doing learning or predictions – the rest of the code could be simply avoided or refactored when possible. Lastly, dependency debt is also considered a key contributor to technical debt in both software development and machine learning systems. This often aggravates since there is no tool to detect them, leading to a build-up of large data dependency chains that are difficult to untangle. Other technical debts in machine learning include configuration debt, external impacts, and correction cascades, among others. Strict vigilance A lot of people believe that technical debt is not completely avoidable, given the strict deadline and delivery pressure. A necessary evil, if you will. However, its impact could certainly be reduced by ensuring best practices, coding standards, regular reviews, and others. As Gergely Orosz, software engineer and author, wrote in his blog, a workplace that ignores technical debt gives birth to a ‘grim engineering environment’. Unfortunately, the team often gravitates towards and is rewarded for short-term solutions and hacks, which is a recipe for disaster in the long run.","excerpt":"Technical debt happens when teams opt for a quick and easy solution for a product to reach the market faster","categories":["AI Trends"],"tags":["App Development","Elon Musk","Software Development","Technical debt","Twitter (X)"],"author_name":"Shraddha Goled","publish_date":"2022-11-21T11:00:00","publication_year":"2022","word_count":809,"keywords":["Go","API","machine learning","TPU","programming_languages:R","AI","programming_languages:Go","App Development","Technical debt","Elon Musk","ViT","Software Development","Twitter (X)","R"],"extracted_tech_keywords":["AI","machine learning","TPU","R","Go","API","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/burden-of-technical-debt-why-it-can-cost-ml-developers-heavily\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10134983,"title":"Klarna Cuts 50% of Workforce, Ends Partnerships with Salesforce and Workday Amid Generative AI Overhaul","content":"Fintech giant, Klarna CEO Sebastian Siemiatkowski announced that the company will end its service provider relationships with Salesforce and Workday as part of a major internal overhaul driven by AI initiatives. Source: X Klarna, known for its installment payment plans, plans to phase out both platforms in the coming weeks to consolidate its tech stack and enhance operational efficiency. This decision aligns with Klarna’s broader strategy to streamline its technology infrastructure by leveraging AI to standardise and simplify its processes. “Thanks to AI agents + AI engineers getting prolific, you can rebuild most enterprise SaaS functionality, host for super cheap, and get basically 90%+ functionality,” said Linas Beliūnas, director of revenue at Zero Hash. In addition to the software changes, Klarna is set to reduce its workforce significantly. Siemiatkowski highlighted that AI will play a crucial role in cutting operational costs, contributing to a planned reduction in the company’s workforce by 50%. This move is expected to help the company operate more effectively while maintaining higher quality standards. Klarna’s latest financial results reveal a 27% increase in revenue year-over-year, reaching SEK 13.3 billion for the second quarter. The company’s aggressive cost-cutting measures and focus on AI come as part of a broader trend where companies across various sectors are implementing AI-driven solutions to enhance efficiency and reduce costs.","excerpt":"Siemiatkowski highlighted that AI will play a crucial role in cutting operational costs, contributing to a planned reduction in the company’s workforce by 50%.","categories":["AI News"],"tags":["Klarna","Salesforce"],"author_name":"Siddharth Jindal","publish_date":"2024-09-11T16:06:42","publication_year":"2024","word_count":218,"keywords":["programming_languages:R","AI","emerging_tech:AI agents","ML","RAG","Klarna","AI agents","Salesforce","R"],"extracted_tech_keywords":["AI","ML","RAG","R","AI agents","programming_languages:R","emerging_tech:AI agents"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/klarna-cuts-50-of-workforce-ends-partnerships-with-salesforce-and-workday-amid-generative-ai-overhaul\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10070192,"title":"Intuit to stop QuickBooks online products in India from 2023","content":"Online accounting software for small businesses from Intuit – Quickbooks – will be discontinued in India after 10 years of trying to crack the market. In an official statement, Intuit informed that from January 31, 2023, Quickbooks Online products will no longer be available in India. “Accordingly, we are no longer accepting new subscriptions in India at this time. This applies to all subscriptions for QuickBooks Online, QuickBooks Online Accountant, the QuickBooks mobile app, and QuickBooks Time,” the statement said, as reported by Moneycontrol. All the existing paid subscribers will be switched to the free plan by Jul 31, 2022; this will help companies to continue with Quickbooks till January 2023 with no charges applied. Customers with an annual subscription will be offered a refund for the unused portion of their subscription. The company has asked companies to download their data out of the product at this time to avoid losing it. Quickbooks has been available in India since 2012 and rolled out its GST-compliant version in 2017. It also offers an online practice management solutions called QuickBooks Online Accountant for chartered accountants. The shutdown of QuickBooks in India will not impact parent company Intuit’s presence in India, where it has more than 1,300 employees, the company said. Intuit counts Zoho and Tally among its competitors.","excerpt":"All the existing paid subscribers will be switched to the free plan by Jul 31, 2022.","categories":["AI News"],"tags":["Intuit"],"author_name":"Shraddha Goled","publish_date":"2022-06-30T17:41:03","publication_year":"2022","word_count":216,"keywords":["Intuit","R","programming_languages:R","AI"],"extracted_tech_keywords":["AI","R","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/intuit-to-stop-quickbooks-online-products-in-india-from-2023\/","complexity_score":1,"technical_depth":3,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10131353,"title":"PwC India Unveils GenAI Experience Lab in Gurgaon","content":"PwC India has announced the inauguration of the GenAI Experience Lab at their Novus Tower office in Gurgaon, marking a significant step in exploring Generative AI’s transformative capabilities. The new facility aims to provide an immersive experience across various domains, showcasing the potential of Generative AI to revolutionise processes, enhance decision-making, and boost efficiency. The GenAI Experience Lab features new hardware designed to foster creativity, collaboration, and advanced AI experimentation. Visitors can engage with award-winning prototypes and interact with PwC’s innovators, gaining insights into the practical applications of Generative AI. This initiative not only enhances PwC’s ability to demonstrate AI’s transformative potential to clients but also strengthens the firm’s leadership in technological innovation. The lab is set to drive forward-thinking innovation and solidify PwC’s position as a visionary in the tech industry. PwC India recently also teamed up with Microsoft India to bolster its incident response and recovery capabilities. The collaboration aims to tackle the increasingly sophisticated cyber threats that organisations encounter. By merging PwC India’s expertise in incident management with Microsoft’s Copilot, the partnership aspires to deliver robust protection for enterprises.Last year, PwC announced that it is set to invest $1 billion over three years to advance generative AI in its US operations, collaborating with Microsoft and OpenAI to automate tasks in tax, audit, and consulting. Consultancy firms have been experimenting with generative AI since last year in unique ways. KPMG had internally developed a system based on OpenAI models and called it KaiChat to aid its staff with exclusive data. EY leverages generative AI, integrating tax laws into an AI system for instant responses through a ChatGPT-like interface, particularly for tasks like payroll queries. In August, McKinsey embraced the potential of LLMs with the launch of “Lilli”, designed to streamline and enhance the utilisation of the firm’s vast knowledge base. Wells Fargo also introduced Fargo in 2022, a virtual assistant powered by Google Cloud’s AI, for providing a personalised, convenient, and simple banking experience. In October last year, Deloitte launched DARTbot, an internal chatbot for enhancing the efficiency of Deloitte’s 18,000 US Audit & Assurance professionals.","excerpt":"PwC India recently also teamed up with Microsoft India to bolster its incident response and recovery capabilities.","categories":["AI News"],"tags":["pwc"],"author_name":"Mohit Pandey","publish_date":"2024-08-05T17:27:21","publication_year":"2024","word_count":348,"keywords":["ChatGPT","GenAI","OpenAI","AI","AWS","ML","RAG","Aim","generative AI","pwc","R"],"extracted_tech_keywords":["AI","ML","generative AI","GenAI","ChatGPT","OpenAI","Aim","RAG","AWS","R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/pwc-india-unveils-genai-experience-lab-in-gurgaon\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":58872,"title":"How To Use Effective Data Storytelling For Business Impact","content":"An extraordinary amount of data passes through businesses on an ordinary day. Data-driven insights are driving a new wave of business intelligence, helping move the needle with quick business impact. However, with the increasing dependency and usage of analytics that is embedded into day to day decision making at enterprises, the demand for easily consumable and interactive data dashboards is on the rise. Analytics dashboards have become critical in helping managers and executives make fast decisions. For these dashboards to be truly effective and impactful, the data insights need to be communicated using the best practices of design. But it should not stop there. Analysts need to communicate the data insights to the stakeholders with infectious passion and enthusiasm. Here the importance of data storytelling emerges. Storytelling has been a valuable tool that helped drive social cohesion for millennia, and this is permeating the seemingly abstract world of complex numbers and statistics as well. Marrying data analytics with good storytelling has multiple benefits. With storytelling, stakeholders can more easily process complex business information that their data presents. Additionally, stories enhance memory, making it easier for the stakeholders to retain information. Stories are also known to induce the release of hormones oxytocin and cortisol – helping create a long-lasting impact on the audience. One estimate suggests that we can recall facts up to 22 times more effectively when they are part of a story as compared to when they are presented as isolated data points. Information presented in the form of a story also helps to reinforce our understanding and decipher patterns in complex data sets. The human brain finds it difficult to make sense if numerous information is presented as raw data points or in close natural form. A famous example demonstrates the importance of graphical storytelling – the Anscombe’s Quartet, created by Francis Anscombe in 1973. The illustration displays four different datasets with almost identical variance, mean, the correlation between X and Y coordinates, and linear regression lines. However, the patterns are different when plotted on a graph. Source: Wikipedia Useful data visualization aids in human cognition of abstract data points by emphasizing analytical reasoning and using an interactive visual interface. What Is Effective Data Storytelling? By infusing data points with strategic creativity, they will become compelling, digestible and impactful. Our human brain loves a good story and finds it easy to grasp difficult concepts if they are summarized and presented in the form of a story. This is why fables and tales have always been such an essential part of our culture. The same applies to numbers, statistics and data science as well. According to neurological research, when the user is shown a visualization of data, it takes just under 500 milliseconds for the human eye and brain to process the pre-attentive visual properties of an image. With the information overdose that we are subjected to in our daily lives today, well-narrated data stories with compelling visuals are vital in retaining attention and delivering maximum value to stakeholders. Data storytelling is becoming a critical skill within the analytics industry – reflected in the rising demand for professionals with data storytelling skills. Data storytelling results in improved insights that are not apparent with merely using descriptive and statistical methods, resulting in faster decision making. Using sound data visualization principles is an essential component of compelling data storytelling. John Tukey, the American mathematician and statistician, famously remarked; “The greatest value of a picture is when it forces us to notice what we never expected to see.” However, care must be taken not to over-do the visuals. High cognitive load results when the visualization gets cluttered and the required effort to understand the message is too high. It will lead to your audience getting disengaged. Using the pre-attentive attributes and contrast in size as well as colour will help to deliver a stunning visualization that gets its points across simplistically while reducing the cognitive load on the readers. According to Colin Ware’s Information Visualization: Perception for Design, there are four pre-attentive visual properties: ColourFormMovementSpatial positioning Strategically using these pre-attentive visual principles, and combining them with cutting-edge data-driven insights will create compelling and impactful data analytics for business stakeholders that will lead to increased adoption and improved Return on Analytics Investments. Elements of Data-Storytelling Certain elements need to be considered and deployed to deliver effective storytelling with data First, Understand The Stakeholders Expectations – Understand the user skill levels and their expectations, what KPIs they are tracking, what kind of analysis they want. Understand if they are interested in a deep drill down or a top-level reporting. Once you have understood the requirements and your user persona, you can begin crafting your story that best resonates with them.  Set the context for your story – Context helps to explain the backdrop and set the mood for the story, helping elevate the user interest, bridge the gap, create interest and communicate your point of view more effectively. Define The Story Lifecycle – All stories have certain universal elements in common – characters, conflicts and redemption arc which help make the story engaging and impactful. Map The Story To The Visual – Next, once you have the character arc with you, choose the right visuals to tell the story. There is a helpful rule of thumb when selecting the charts and graphs on your dashboards. Treat each visualization as the answer to a question. The whole dashboard should also be the answer to a question which can be answered at a glance. Use principles of visual hierarchy to arrange and place visuals.  Story delivery and walkthrough – Communicating the insights have to be more than merely pushing out information. Using infectious passion and energy to narrate the story will help to convey business outcomes and actions with necessary speed and impact. As author Anne Lindbergh shares, “Good communication is just as stimulating as black coffee, and just as hard to sleep after” 6 Visual Design Principles For Dashboard A Practical Test For A Good Dashboard Ask your colleague to take a glance for 4-5 seconds and look away. Then, ask him what he can recall from the dashboard – what metrics & figures. The more effectively he can remember and communicate what he grasped from the dashboard – then it means that your dashboard served its purpose and it is good. If they cannot recall essential data points, then maybe your design is not simple enough to make effective communication. Wrapping Up While there are tonnes of visualization options available to the data professionals today, analysts need to deliver impactful data-driven analyses that are focussed on simplicity, effortlessness and reducing the cognitive load for the users. Data analytics has become a critical differentiator at top-performing organizations today. Hence, the need for data storytelling has emerged strongly. And, it will become even more important with time. By helping communicate powerful insights, data storytelling will help businesses capitalize on market opportunities – fast and help them remain ahead of the curve.","excerpt":"An extraordinary amount of data passes through businesses on an ordinary day.   Data-driven insights are driving a new wave of business intelligence, helping move the needle with quick business impact.  However, with the increasing dependency and usage of analytics that is embedded into day to day decision making at enterprises, the demand for easily consumable […]","categories":["AI Features"],"tags":[],"author_name":"Tushar Sonal","publish_date":"2020-03-17T21:00:00","publication_year":"2020","word_count":1165,"keywords":["business intelligence","Go","data science","API","AI","data-driven","ViT","analytics","GAN","R"],"extracted_tech_keywords":["AI","data science","analytics","R","Go","API","GAN","ViT","business intelligence","data-driven"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-to-use-effective-data-storytelling-for-business-impact\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10008224,"title":"10 Best Online Resources To Learn Pandas","content":"One of the most popular libraries of Python — Pandas provides fast, flexible, and expressive data structures. The library has various intuitive features, including easy handling of missing data, data alignment, fancy indexing, data alignment, to name a few. In this article, we list down the ten best free online resources, in no particular order, to learn Pandas. Master Data Analysis with Python – Intro to Pandas Master Data Analysis with Python – Intro to Pandas provides an introduction to the components of the two primary Pandas objects, the DataFrame and Series, and how to select subsets of data from them. Learners will understand the components of the DataFrame and Series, data types, selecting subsets of data, filtering data via boolean selection and more. Click here to learn. Pandas Python Library for Beginners in Data Science Pandas Python Library for Beginners in Data Science is a one-hour video course that is meant for college students and for those who have not heard of Pandas before and willing to learn about the syntax in Pandas. The project provides various challenges with solutions to encourage learners to practice using Pandas. By the end of this project, learners will master the basics of Pandas. They will also be able to gain insight into the data, clean it, and do basic preprocessing to get the most value out of your data. Click here to learn. Pandas Foundations In this course, learners will be introduced to pandas DataFrames. It will enable learners to use pandas to import and inspect a variety of datasets, ranging from population data obtained from the World Bank to monthly stock data obtained via Yahoo Finance. The course will help in learning how to work with real-world datasets containing both string and numeric data, often structured around time series. It will also help in learning powerful analysis, selection, and visualisation techniques in this course. Click here to learn. Learn Data Analysis using Pandas and Python This course is a 101 level data manipulation course that will cover an introduction to data science and Pandas, its installation and setting up the environment. It will cover the basics of data analysis and data manipulation using Pandas and learn various steps of data preparation including data massaging, filtering, string manipulations, aggregations and more. Click here to learn. Pandas Exercises, Practice, Solution Pandas Exercises, Practice, Solution is provided by w3resource, where one can learn how to work with data using Pandas library and practice various problems related to the library. The course aims to be the fundamental high-level building block for doing practical, real-world data analysis in Python. The course includes various Pandas exercises including data series, data frame, index, string and regular expression, handling missing values and more. Click here to learn. Pandas This course from Kaggle will help in learning all about Pandas. In this course, learners will learn how to create their own data, along with how to work with data that already exists. The course includes topics like creating, reading, indexing, sorting, summary functions, data types, missing values, renaming and more. Click here to learn. Intermediate Pandas Python Library for Data Science Intermediate Pandas Python Library for Data Science course will help in understanding the advanced methods to handle missing values, how to sort, select and slice data for easier manipulation as well as the different types of joins, sorting and binning data. This guided project is meant for those who are familiar with pandas for data analysis but want to really harness the power of pandas by learning more complex operations. By the end of this project, learners will be able to learn how to handle structured data using Pandas. Click here to learn. Writing Efficient Code with Pandas This course will help in understanding Python and the Pandas library and introduce the efficient built-in Pandas functions to perform tasks faster. By the end of this course, one will be able to apply a function to data based on a feature value, iterate through big datasets rapidly, and manipulate data belonging to different groups efficiently. Learners will also be able to apply these methods on a variety of real-world datasets, such as poker hands or restaurant tips. Click here to learn. Pandas with Python Pandas with Python course that will help teach the step-by-step process of Pandas. The course covers several different methods, attributes, features, and functionalities of Pandas. It includes topics such as series and data frame, data range and inspecting data, indexing and slicing, merging data frames, working with text data, and more. Click here to learn. Introduction to Pandas Introduction to Pandas is a course in Jupyter nbviewer, where it will teach from the basics of Pandas. The course can help in understanding features of the library, and how to work with Pandas in series, data frame, importing data, etc. It will also show how to implement these features with the help of some exercises. Click here to learn.","excerpt":"One of the most popular libraries of Python — Pandas provides fast, flexible, and expressive data structures. The library has various intuitive features, including easy handling of missing data, data alignment, fancy indexing, data alignment, to name a few.  In this article, we list down the ten best free online resources, in no particular order, […]","categories":["AI Trends"],"tags":["pandas"],"author_name":"Ambika Choudhury","publish_date":"2020-09-24T12:30:25","publication_year":"2020","word_count":822,"keywords":["big data","data science","API","AI","RAG","Python","Aim","pandas","Jupyter","R","Pandas"],"extracted_tech_keywords":["AI","data science","Aim","Jupyter","Pandas","RAG","Python","R","API","big data"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/10-best-online-resources-to-learn-pandas\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10004370,"title":"IIT Madras Launches Online BSc Degree Program In Data Science","content":"In a world where decision-making is increasingly becoming data-driven there is a huge gap between the demand for data scientists and the supply of qualified applicants in the job market. As a game-changer, IIT Madras brings a unique online program at the degree and the diploma level in areas of programming and data science. IIT Madras is pioneering a completely online program to produce graduates who will be employable in the sector that presently has high demand – Programming and Data Science. IIT Madras is creating a unique virtual classroom environment filled with a diverse set of learners from all ages and backgrounds. Its expert tutors and mentors will offer a virtually interactive training experience and tie-ups with industries, enable internships and job opportunities. Eligibility Anyone who is passed out of any stream in 12 standard and has enrolled or graduated from college is eligible for the qualifier of this program. Once you clear the qualification process you will be inducted into the program where the first year is a foundational program. This is followed by two diplomas in the second year. For those who carry on to the third year, there awaits also the chance of specialisation. Upon graduation you can earn a BSc degree from IIT Madras and become an alumnus of the Institute this program will be offered to three terms in a year to ensure flexibility for working professionals and full-time college students. The program uses an innovative combination of online learning and in-person assessment and is flexible and affordable. Visit the official website of IIT Madras to know more.","excerpt":"In a world where decision-making is increasingly becoming data-driven there is a huge gap between the demand for data scientists and the supply of qualified applicants in the job market. As a game-changer, IIT Madras brings a unique online program at the degree and the diploma level in areas of programming and data science. IIT […]","categories":["AI News"],"tags":["data science degree","IIT Madras"],"author_name":"Vishal Chawla","publish_date":"2020-08-07T11:46:30","publication_year":"2020","word_count":264,"keywords":["data science","data science degree","programming_languages:R","AI","IIT Madras","data-driven","R"],"extracted_tech_keywords":["AI","data science","R","data-driven","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/iit-madras-launchesonline-bsc-degree-program-in-programming-and-data-science\/","complexity_score":2,"technical_depth":5,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10099031,"title":"The Major GCC Announcements for India in 2023-2024","content":"EY, a professional services organisation, released its ‘Future of GCCs in India – a vision 2030‘ report, highlighting GCC expansion in India. The report estimates India’s domestic GCC market may reach $110b by 2030 (from $45b), led by software exports. By 2030, the report predicts 2400 GCCs growing at the rate of 115 new GCCs yearly (currently 70), The growth of the workforce is predicted to be 4.5M from the current 1.9 million people. GCCs in India are focusing on talent retention, niche resource partnerships, expansion to more centres in different cities and also tier-II city interest. With a thriving startup ecosystem, India offers GCC-startup collaboration for innovation. The Comprehensive List of Global Capability Centers (GCCs) in India > Here is this year’s list of the thriving GCCs across sectors. Sandoz In January this year, Sandoz, a major player in generic medicines, announced to set up their centre in Hyderabad,  which will support their global knowledge services. Initially, Sandoz, a part of Novartis, will employ 800 people, with plans to expand to 1,800 in coming years. The company holds a diverse portfolio of around 1,000 medicines covering major therapies, achieving $9.7 billion in sales in 2019, benefitting over 500 million patients worldwide. Alvarez & Marsal Earlier this month, renowned global professional services firm Alvarez & Marsal (A&M) announced that they have launched the inaugural GCC in India. This centre will cater to A&M’s diverse business operations across six continents. The move underscores A&M’s strategy to deploy skilled professionals worldwide for delivering value, enhancing growth, and fostering innovation. Deutsche Bank Deutsche India, home to Deutsche Bank‘s largest technology centre, is expanding in India, planning to hire thousands more. The bank has hired over 2,500 individuals since January 2023, with hiring expected to continue through 2023 and beyond. The centre employs about 16,000 professionals, primarily engineers. Deutsche Bank aims to build an internal engineering workforce, moving away from outsourcing, with a goal to achieve a 70:30 in-house to outsourcing work ratio. NatWest Group NatWest Group, a major UK bank, has introduced a pioneering pilot in India’s global capability centres to offer first-time job opportunities to women who have never been part of the workforce. This initiative, named ‘Wish,’ began this year with 15 women in Chennai and Gurugram, focusing on candidates with no experience regardless of age. The bank’s existing ‘Re-invest‘ program supports women returnees, while this ‘Wish’ pilot elevates its India women hiring plan. The GCC employs 16,000 in India, recruiting over 2,000 annually, and has initiated upskilling efforts, allowing employees to dedicate two extra days yearly for future-oriented skill enhancement. JP Morgan Chase JPMorgan views India as a vital Asian market, constantly boosting its capabilities. Amid global competition, the company has expanded its tech centres in Mumbai, Bengaluru, and Hyderabad in the past few years, totaling 3 out of 21 tech centres in India itself. These centres tap into India’s rich talent and expertise. The bank plans to use India not only as a tech hub but also for product development. This strategic move aligns with the growth of their global tech centres and India’s tech landscape. Nissan Nissan along with Renault announced an ambitious plan to increase the production and R&D activities, to introduce electric vehicles, and transitioning to carbon-neutral manufacturing. Based in Chennai, the collaboration targets six new production vehicles, including two electric models, establishing the Renault-Nissan centre as a global export hub. An initial investment of about $600 million is allocated to these initiatives, aiming to create up to 2,000 jobs. Hewlett Packard Enterprise (HPE) HPE announced in July this year that the company intends to begin producing high-volume servers in India. Plans include manufacturing approximately $1 billion worth of servers in India within the first five years. Partnering with VVDN Technologies, HPE’s production will be based in Manesar, Haryana. The move is aligned with the ‘Make in India’ initiative and reinforces HPE’s commitment to the country. HPE’s significant workforce in India, including over 4,000 scientists and engineers, is central to this endeavour across various sectors and initiatives. Okta Earlier this month, Okta opened an office in Bengaluru, India, for secure digital transformation in the APAC region. This hub will include a state-of-the-art R&D centre for global identity solution development. Amid India’s rising cybersecurity spending (predicted 18% growth by 2025), Okta aims to serve local customers, enhance security, and contribute to India’s technology landscape. The office will be the site of its first innovation centre in Asia Pacific and will facilitate engagement with local policy makers, educational institutions, and customised integrations for Indian enterprises. The company also partnered with OpenAI earlier this month. Johnson Controls Johnson Controls inaugurated its largest OpenBlue Innovation Centre in Bengaluru in July this year, showcasing net zero building tech with cloud, edge, and AI. The centre supports India’s G20 priorities, advancing sustainability and digitalization. It boasts AI-driven energy-efficient tech demos, a digital twin, access control, computer vision, fire safety solutions, and VR experiences for net zero goals. Johnson Controls, present in India for 30 years, will employ 300 engineers at this centre, expanding its role in smart building innovation in India’s IT hub. Houghton Mifflin Harcourt (HMH) In July this year, Learning technology company Houghton Mifflin Harcourt (HMH) launched its first Asia-Pacific Center of Excellence (COE) in Pune, India. COEs within Global Capability Centers (GCCs) are organisations or teams focused on developing and advancing analytics capabilities within a company. This expansion aims to harness India’s tech talent for innovation in K-12 education solutions. The Pune Center will focus on product innovation, R&D, and development, catering to Indian schools implementing IB and IGCSE curriculum. HMH’s solutions integrate core instruction, practice, intervention, assessment, and professional learning. The Comprehensive List of Global Capability Centers (GCCs) in India >","excerpt":"Global Capability Centers (GCCs) in India are prioritizing talent retention, forming niche resource partnerships, expanding into diverse cities, including tier-II locations.","categories":["GCC"],"tags":["GCC","pwc"],"author_name":"K L Krithika","publish_date":"2023-08-25T10:47:19","publication_year":"2023","word_count":953,"keywords":["Go","API","GCC","OpenAI","AI","Git","computer vision","Aim","analytics","pwc","GAN","R"],"extracted_tech_keywords":["AI","computer vision","analytics","OpenAI","Aim","R","Go","Git","API","GAN"],"url":"https:\/\/analyticsindiamag.com\/gcc\/the-major-gcc-announcements-for-india-in-2023\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10168158,"title":"AI That Streamlines Workflows for Radiologists Debuts in India","content":"India has roughly 20,000 radiologists serving a population of over 1.4 billion, indicating that the country has approximately one radiologist for every 1 lakh people. This is significantly lower than the global average of 4.2 radiologists per 1 lakh people as of 2023. The diagnostic burden is immense, with only one radiologist available to interpret every 100 scans conducted daily. SPARK Radiology has launched SPARK.ai, an AI integrated radiology platform designed to alleviate burnout, enhance diagnostic accuracy and streamline workflows for radiologists. The launch marks its debut in India’s healthcare technology space. SPARK.ai recognises this critical gap and has introduced AI-powered efficiencies that promise to ease administrative workloads while enabling quicker, more accurate diagnosis. The platform accurately detects the radiologist’s impressions from the diagnosis and incorporates them into the template. This eliminates the need for an intermediary to manually type and file the report. “We’re here to let radiologists concentrate on the complex, life-saving decisions they’re trained for, while our technology handles the rest,” said Allison Garza, CEO of SPARK Radiology. Radiologists are at the frontline of diagnosis, yet a significant portion of their time is consumed by repetitive administrative tasks. “This platform is built to give them that time back,” she added. The platform has been tested across independent clinics, hospitals, and diagnostic centres, and its capabilities have been refined over several months. It seamlessly integrates with existing systems such as PACS and employs a head-up display (HUD) to help radiologists document findings in real time, through automated structured reports and smart templates. “Nothing is missed nor added by itself through SPARK.ai,” said Suresh Joel, CTO, SPARK Radiology. “It helps bridge communication gaps between the radiologist and stenographer, which has long been challenging in high-volume settings.” “The units of measurement for kidney size and stone size can sometimes be incorrect when done manually. With ultrasound, assistance from a stenographer is available, but with CT and MRI, everything is done solely by the doctor,” Dr Asha Ouseph, a radiologist, pointed out. The platform’s intuitive design also allows customisable templates that can adapt to individual or institutional preferences, boosting speed up to 50% and accuracy in report generation. “Attaining the speed of voice detection with high accuracy was a key challenge, and we’ve been fine-tuning the product since November,” Garza added. “The result is a solution that not only reduces turnaround times but also improves precision.” With India’s healthtech market projected to reach nearly $60 billion by FY 2028, the launch of SPARK.ai aligns with a larger push towards AI-enabled diagnostics. By removing manual bottlenecks and reducing burnout, the platform hopes to equip radiologists to meet rising demands head-on. Rad AI, Rayscape and Aidoc are platforms that streamline radiology workflows and automate repetitive tasks. At the same time, SPARK.ai stands out as one of the first solutions developed within India. “The integration of AI into radiology is more than just an operational improvement; it’s a step towards building a more robust and scalable healthcare ecosystem,” Dr Joel said. “With this platform, diagnostic centres can expand reach, optimise resources, and ultimately deliver better patient care.” What truly sets SPARK.ai apart is its human-centered approach. Built in collaboration with radiologists, the platform is not only functional but also deeply empathetic to the needs of those using it. Its HUD system, auto-fill capabilities and structured templates help cut through cognitive overload and streamline repetitive tasks. “Our goal is to ensure radiologists are not bogged down by inefficiencies. SPARK.ai integrates seamlessly, grows with institutional demands, and most importantly, centres around the people who use it,” Garza said. Meanwhile, according to a study published in the Journal of Imaging Informatics in Medicine, large language models (LLMs) could potentially track for interval changes on longitudinal radiology reports. The study suggests that LLMs can effectively identify findings and monitor changes in radiology reports, all while preserving patient privacy by operating securely within an institution’s internal network. This approach would yield time savings by adding automation to a process requiring radiologists to match relevant findings manually.","excerpt":"“Let radiologists concentrate on the complex, life-saving decisions they’re trained for, while our technology handles the rest.”","categories":["AI Features"],"tags":["radiology","SPARK.ai"],"author_name":"Merin Susan John","publish_date":"2025-04-17T10:25:25","publication_year":"2025","word_count":667,"keywords":["Go","radiology","programming_languages:R","AI","ML","Scala","Git","RAG","automation","Ray","SPARK.ai","R"],"extracted_tech_keywords":["AI","ML","Ray","RAG","R","Go","Scala","Git","automation","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/ai-that-streamlines-workflows-for-radiologists-debuts-in-india\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10109821,"title":"Why Taking a Year Off to Build a Startup is a Bad Idea","content":"A lot of Indian universities are giving a year off to students interested in building startups. BITS Pilani, IIT Madras, IIT Hyderabad, DIT University, IIT Bombay, IIT Kharagpur, and a few others have a “temporary withdrawal programme” allowing students to take a break from their academic studies and pursue their entrepreneurial calling. “Bad idea”, said a user on LinkedIn talking about the universities’ move to allow a year off. He highlighted how typically, only 2 to 6 percent of startups achieve profitability within the first year. What about tech entrepreneurs or students whose startups fail or prove non-viable in the market? AIM Recruits believes that stability concerns may arise with candidates who have entrepreneurial experience, as they may not always seamlessly integrate into the traditional corporate structure. “Entrepreneurs often tend to revert to launching their ventures shortly after joining an organisation,” said the spokesperson. At the same time, the failure of the startup also signals a potential deficit in comprehending market dynamics or industry trends. Most investors would not want to fund startups started by students, as they are likely to believe that they would end up losing all the money. No more jobs? Experts say that in the product-based or startup sectors, skills outweigh degrees. “Entrepreneurial experience holds particular significance, indicating a strong commitment and work ethic.” Universities believe in allowing students to gain real-world experience by applying theoretical knowledge in practical settings. This experience, according to them, will not only enhance their entrepreneurial skills but also make them more adaptable and resilient professionals. Candidates with startup backgrounds often showcase strong leadership skills and dedicated individual contributor traits, providing a dual advantage for organisational success. Additionally, their technical expertise, gained from working on their projects, enhances proficiency in essential skills. Unfortunately, big companies such as Microsoft, Google, or Indian IT giants think differently. “They don’t hire such people at all. They are very particular about pedigree,” said the AIM Recruits expert. “Their main target pool is folks from IITs, NITs, BITs, DTU, MIT, without any backlogs or gaps.” Adding to this, Indian IT companies have put a freeze on hiring new employees. Many of them have a huge bench size, and the companies believe that they require a lot more training of the existing employees before they can hire new ones. According to a recent report, private engineering institutes have started to report a 50-70% drop in placements in 2023. Not a toss of coin One of the most significant drawbacks of taking a year off is the potential impact on future career opportunities. Employers often value a consistent academic track record, and a gap year may be perceived as a red flag, potentially affecting the chances of securing desirable job placements after graduation. Even though universities are the best place for promoting AI and entrepreneurship, it seems like risky business to move out of college for a year unless you are absolutely sure that your startup idea is golden. For the universities, it becomes very hard to determine if the AI venture would be worth or not. It is also better for bigger companies to hire more freshers and promote their entrepreneurial callings through incubators and startup programmes, instead of letting them out in the open market. Though some universities provide resources during the gap, it is still not enough to make a profitable business. Funding winter? In August, the government of India announced its scheme of funding 1,200 startups under the Digital India initiative, giving a boost to the entrepreneurs in universities. Despite this, the startup funding in India dropped significantly in 2023. According to Tracxn, the startup funding declined a massive 73% in 2023, compared to 2022 — dropping from $25 billion to $7 billion — lowest in the past five years. Most importantly, seed-stage funding dropped 60% from $1.7 billion in 2022 to $678 million in 2023. All in all, it might not entirely be a bad idea to take a year off, but it would be wiser to get a job (which is also difficult) and then possibly join their startup programme. Since the market is down, the percentages show that the chance of failure is significantly higher. Think twice.","excerpt":"It would be wiser to get a job and then possibly join the company’s startup programme.","categories":["Deep Tech"],"tags":["bits pilani","IIT Madras","Startups"],"author_name":"Mohit Pandey","publish_date":"2024-01-02T15:40:20","publication_year":"2024","word_count":696,"keywords":["Go","funding","AI","IIT Madras","ML","Git","RAG","Aim","bits pilani","GAN","Startups","R","startup"],"extracted_tech_keywords":["AI","ML","Aim","RAG","R","Go","Git","GAN","startup","funding"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/why-taking-a-year-off-to-build-a-startup-is-a-bad-idea\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10047182,"title":"New Weekend Hackathon For Data Scientists: Concept-A-Thon — Chartered Data Scientist","content":"MachineHack is back again with yet another exciting hackathon, but this time with a twist. This weekend the data science community is challenged on a curated mock test for a Chartered Data Scientist Certification. This week’s challenge is a part of MachineHack Weekend Hackathon Edition #2 — The Last Hacker Standing. We pose unique problem statements every week, from 30 July to 9 September 2021, to test you over various aspects of being a Data Scientist. PARTICIPATE & STAND A CHANCE TO WIN FREE PASSES TO THE DLDC 2021!!! Overview The world is about building new paradigms to challenge old ways of thinking. With this new hackathon, MachineHack intends to do the same in the data science skills assessments domain. Traditional data science hackathons are typically designed to test data engineering and data modelling skills. However, have data scientists ever been tested on their concepts in a time-bound manner? Here, MachineHack presents the newest innovation in this space. Research has repeatedly proven that time-bound tests are very effective in many ways, which is why many competitive exams across the world employ this method. A time-bound test is usually designed to assess skills such as conceptual understanding, out of the box thinking, presence of mind, time management, and the ability to perform under pressure. Problem Statement and Description In this current era of rapid upskilling, there is a clutter of certifications for data science enthusiasts to choose from, but one would never know which ones are really appreciated by the industry. This remains a huge question for data scientists to address and prepare accordingly. In this week’s challenge, MachineHack is asking data scientists to attempt the finely curated mock test for a Chartered Data Scientist Certification. The Chartered Data Scientist, by the Association of Data Scientists (ADaSci), is a self-study program culminating in an exam. Candidates passing this exam will be one step closer to becoming a CDS charter holder. The Chartered Data Scientist (CDS) credential tests an individual on a solid understanding of the advanced data science profession and in-depth, applied analytics skills. Here’s the mock test for a very rigorous exam, including 150 multiple choice questions that need to be answered in three hours. This test has no negative marking. Click here to participate. The hackathon will start on 27 August at 8:00 PM (IST) Skills required to participate in the CDS Mock Test are: Probability theory, statistics, and linear algebraData engineering and databasesExploratory data analysisSupervised and unsupervised learningNeural networks and deep learningNatural language processingComputer visionData science fundamentals like basic\/advanced mathematics, statistics, and machine learning knowledge Evaluation Criteria The leaderboard will populate the final test scores based on the highest score. Click here to participate. The hackathon will end on 2 September at 6:00 PM (IST) Prizes The top five winners of this concept-a-thon will get free passes to attend the biggest two-day conference on deep learning — the Deep Learning DevCon 2021 (DLDC), to be held on 23-24 Sept 2021. Start Date: 27 August 2021 at 8:00 PM End Date: 2 September 2021 at 6:00 PM Click here to participate in this hackathon.","excerpt":"MachineHack is challenging the data science community on a curated mock test for a Chartered Data Scientist Certification.","categories":["Deep Tech"],"tags":["cds","chartered data scientist","hackathon for data scientists","MachineHack Weekend hackathon","Statistics for Data Science","Weekend Hackathon","Weekend hackathon for data scientists","Weekend hackathon the last hacker standing"],"author_name":"AIM Media House","publish_date":"2021-08-27T18:00:00","publication_year":"2021","word_count":515,"keywords":["data science","Go","API","machine learning","Weekend Hackathon","chartered data scientist","Weekend hackathon for data scientists","AI","neural network","cds","computer vision","Weekend hackathon the last hacker standing","deep learning","Statistics for Data Science","analytics","hackathon for data scientists","MachineHack Weekend hackathon","R"],"extracted_tech_keywords":["AI","machine learning","deep learning","neural network","computer vision","data science","analytics","R","Go","API"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/new-weekend-hackathon-for-data-scientists-concept-a-thon-chartered-data-scientist\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":24152,"title":"How Indian Airports Are Turning To AI For Better Customer Success","content":"Airports across the globe are getting competitive in adopting analytics and artificial intelligence in their ecosystem. They are leveraging these technologies for many purposes — minimising the impact of disruption on the passenger experience, supporting the customers or making their services better, among others. According to a report titled Air Transport IT Trends Insights, by SITA, 80 percent of the airlines will invest in major programs or R&D into prediction and warning systems, which rely heavily on AI. Maneesh Jaikrishna Airports are also increasingly adopting chatbots and humanoids. The report suggested that 14 percent of airlines and nine percent of airports already use chatbots, which is likely to increase over the years. Bengaluru airport recently saw the inauguration of special robot assistant named KEMPA at Kempegowda International Airport. Completely designed and manufactured in Bangalore, KEMPA can answer flight-related queries, suggest places to visit in Karnataka, or can simply be used for entertainment purposes. Analytics India Magazine interacted with Maneesh Jaikrishna, vice president, Indian Subcontinent, Eastern and Southern Africa, SITA, to understand why and how Indian airports are turning to artificial intelligence for better customer success. AI And Robotics To Streamline Processes Across Airports Jaikrishna begins with the fact that nearly every airline and airport in the world does business with SITA, and that its border management solutions are used by more than 40 governments. With a presence in more than 1,000 airports around the globe, SITA focuses extensively on the use of technology like AI and robotics to help airports and airlines better manage their businesses and operations, making air travel easier every step of the way. “For example, we are playing a key role in equipping two of the world’s newest and biggest airports — Abu Dhabi International Airport and Istanbul New Airport — with the latest technology to make sure that they are fully future proof. In Istanbul this includes a solution that allow bags to be tracked throughout the journey, allowing the airport to meet IATA 753 tracking requirements from day one”, he said. In Orlando airport, SITA has introduced a secure and seamless departure with new biometric boarding at the gate for British Airways passengers. He said that a quick photo is all that is needed to board an international flight — no passport, no boarding card. SITA integrated automated boarding gates with the US Customs and Border Protection (CBP) and airline’s IT systems to allow the necessary checks and authorise boarding. “Similarly, here in India, we work closely with the Airports Authority of India, where our technology is present in more than 40 airports”, said Jaikrishna. Leading India To 100 Percent Biometric Air Travel Not just Orlando, but SITA is also introducing 100 percent biometric air travel in India as well. “With the increasing growth in domestic air traveler in recent years, India has been making steady progress with new initiatives undertaken by the government such as DigiYatra. We at SITA, see the opportunity to help the Indian government reap benefits from biometric travel,” he said. With their Smart Path biometric solution for ID management, SITA is well-equipped to deliver a seamless, paperless travel experience for passengers. It has already shown the benefits of using biometric technology to automate passenger identity checks at airports across the world — from the US to Australia. It has also been introduced at Australia’s Brisbane Airport which allows passengers to register their biometric details at a self-service kiosk at check-in and then, when ready to board, use an automated boarding gate to be verified using face recognition technology to access the aircraft. “Leveraging India’s national Aadhaar biometric identity system — one of the biggest in the world — together with our proven common-use and Smart Path technology, SITA will be able to deliver a seamless biometric experience across all airlines and airports in India”, said Jaikrishna. AI In Predictions And Warning Systems SITA’s technology harnesses advanced ML and AI to forecast aircraft arrival times, giving the airports a 24-hour heads up to proactively manage and mitigate disruption. It has helped airlines and airports to tackle $25 billion annual cost of poor on-time performance. “We also enable airports to leverage business intelligence to get a better view of their operation and predict where potential bottleneck could arise. For example, Orlando International Airport uses the business intelligence capabilities of SITA’s QueueAnalyzer tool to create a real-time view of the security checkpoints, which enables it to respond quickly to changing conditions at the airport”, said Jaikrishna. He shares that Airline and airport Chief Information Officers (CIOs), including the ones in India, are turning to AI to support their customer service. Over the next three years, 52 percent of airlines plan major AI programs, with 45 percent of airports planning to invest in research and development in the next five years. 42 percent of the airports are also planning to implement AI-driven chatbot services. These will help passengers keep informed on every stage of their journey. Apart from these technologie, SITA offers self-service kiosk as a part of a multi-channel service offering from airlines and airports. SITA’s AirportConnect Kiosk is a self-service airport kiosk that can be dedicated to a specific airline or shared as common-use by multiple airlines. It can be used for a number of self-service processes, including booking, changing a reservation and check-in. It integrates with many other SITA services, including the Airport Connect Open which uses the common use self-service platform, and Passenger iCheck, their check-in application. Key Contributions By SITA In Indian Airports “It is well known that India will become third-largest aviation market by 2025 and the challenge is to find digital solutions which will enable airports to offer a better service to their passengers and accommodate more passengers with the same physical infrastructure”, he said. SITA is already helping several Indian airports better manage their baggage, passenger processing as well as airport operations. For instance, the company last month announced that they would expand their baggage management solution to 15 Indian airports. SITA’s BagManager will provide the airports with real-time information on the status of passengers’ baggage, significantly reducing the chances of baggage mishandling. AAI will also deploy SITA Airport Management Solution to two further airports — Goa and Lucknow airports — taking the total number of India airports using this technology to 12. “We are providing key technology solutions to one of the country’s newest airports at the same time. The Multi-modal International Cargo Hub and Airport at Nagpur (MIHAN) has turned to SITA to provide passenger processing solutions for check-in and boarding, as well as baggage tracking”, said Jaikrishna. Roadmap For 2018 “Smart use of technology can help manage the challenges of rising passenger numbers, limited infrastructure and increased complexity. We are looking at opportunities to harness technologies including biometrics, AI, ML, robotics and blockchain to transform the industry”, says Jaikrishna. Even now, biometrics is becoming more commonplace at airports around the world and is delivering secure seamless travel from check-in to boarding. They already have biometric self-service solutions operating worldwide, including in Australia, US, Mexico and the Middle East. Autonomous vehicles and the potential use of robotics is another field of development, with particular interest shown at the event in Kansai Airport’s trial of KATE, SITA’s autonomous check-in robot. He shares that by embracing IoT and connecting everything across the industry, they will produce more data which can be used with AI to create valuable insights and expose new ways of working. “Over the next 20 years, this use of digital technologies to improve service, operations and efficiency will have profound effects on the air transport industry”, he said, signing off.","excerpt":"Airports across the globe are getting competitive in adopting analytics and artificial intelligence in their ecosystem. They are leveraging these technologies for many purposes — minimising the impact of disruption on the passenger experience, supporting the customers or making their services better, among others. According to a report titled Air Transport IT Trends Insights, by […]","categories":["IT Services"],"tags":[],"author_name":"Srishti Deoras","publish_date":"2018-05-02T05:25:08","publication_year":"2018","word_count":1268,"keywords":["business intelligence","Go","artificial intelligence","AI","chatbots","ML","Git","RAG","analytics","R"],"extracted_tech_keywords":["AI","artificial intelligence","ML","analytics","RAG","chatbots","R","Go","Git","business intelligence"],"url":"https:\/\/analyticsindiamag.com\/it-services\/how-indian-airports-are-turning-to-ai-for-better-customer-success\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10164273,"title":"Together AI Raises $305M in Series B Funding to Enhance its AI Acceleration Cloud","content":"California-based Together AI has announced a successful $305 million Series B funding round, led by General Catalyst and co-led by Prosperity7. The investment includes participation from notable investors such as Salesforce Ventures, NVIDIA, and Kleiner Perkins. 📣 Today we're announcing our $305M Series B funding led by @generalcatalyst and co-led by @p7ventures, with participation from a distinguished group of global institutional and strategic investors. pic.twitter.com\/f4S2L6PhGs— Together AI (@togethercompute) February 20, 2025 The company aims to enhance its infrastructure with the deployment of NVIDIA Blackwell GPUs, which are expected to significantly improve the performance of AI applications. “Together AI is committed to making open-source AI accessible to developers and enterprises globally,” the company stated. They emphasise the transformative potential of AI across industries and the importance of open-source solutions in driving innovation. This funding will bolster Together AI’s position as a leader in providing an AI acceleration cloud tailored for open-source models and enterprise applications. ⚡ Our AI Acceleration Cloud uniquely spans across the AI lifecycle and has transformed how over 450,000 AI developers, AI-native companies, and global enterprises build modern AI applications. From @salesforce and @Zoom to innovative companies like @cognition_labs , @krea_ai…— Together AI (@togethercompute) February 20, 2025 Their infrastructure supports over 200 open-source models across various modalities, enabling organisations to build comprehensive AI applications efficiently. To accommodate its rapid growth, Together AI plans to expand its infrastructure significantly. This includes securing 200 MW of power capacity and deploying optimised clusters of NVIDIA Blackwell GPUs across North American data centres. The funding will also support ongoing research initiatives to optimise AI performance and efficiency. The company’s chief scientist, Tri Dao, highlighted their focus on leveraging advanced GPU architectures to redefine training and inference capabilities. 💪Open source models like DeepSeek-R1 and Meta's Llama have emerged as formidable alternatives to proprietary solutions, marking a decisive shift in the AI landscape. Together AI has established itself as the definitive platform powering this transformation, delivering the…— Together AI (@togethercompute) February 20, 2025 Looking ahead, Together AI aims to continue its trajectory of innovation while ensuring that powerful AI systems remain accessible and cost-effective for all users. In 2023, the company also gained attention for securing a $102.5 million Series A investment from NVIDIA, Kleiner Perkins, and Emergence Capital. “Open source is the future of AI,” said founder and CEO Vipul Ved Prakash, had said adding that “it will be a major thrust of how most organisations implement generative AI”. But just last year, Simplismart, a Bengaluru-based startup led by former Oracle and Google engineers, has emerged as a leader in creating high-performance AI deployment tools. Founded by Amritanshu Jain and Devansh Ghatak, this Bangalore-based startup is garnering attention with its claim of having developed the fastest inference engine in the world, outpacing rivals like TogetherAI and FireworksAI.","excerpt":"The investment includes participation from notable investors such as Salesforce Ventures, NVIDIA, and Kleiner Perkins.","categories":["AI News"],"tags":["ai funding","Funding"],"author_name":"Sanjana Gupta","publish_date":"2025-02-21T00:06:05","publication_year":"2025","word_count":464,"keywords":["Go","API","Funding","AI","innovation","RAG","Aim","ai funding","generative AI","Rust","GAN","R"],"extracted_tech_keywords":["AI","generative AI","Aim","RAG","R","Go","Rust","API","GAN","innovation"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/together-ai-raises-305m-in-series-b-funding-to-enhance-its-ai-acceleration-cloud\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10101818,"title":"Amazon is Testing Humanoids in its Warehouses","content":"Amazon is experimenting with humanoid robots in select US warehouses, marking a significant step in its automation endeavors. The tech giant aims to optimise efficiency by introducing these robots, named ‘Digit’, which emulate human movements for tasks such as moving and handling items. Created by Agility Robotics, a company supported by Amazon and headquartered in Corvallis, Oregon, Digit is a versatile robot. This 5 feet 9 inches (175cm) tall, 143-pound (65 kg) machine possesses the ability to walk in multiple directions, including forward, backward, and sideways, as well as the capability to crouch. Additionally, Digit has a carrying capacity of up to 35 pounds (16 kg). New video of Amazon trialing humanoid robots in US warehouses:They currently employ 1.6 million people. pic.twitter.com\/H3xJ9z3CW0— AI Breakfast (@AiBreakfast) October 20, 2023 Amazon’s decision to implement robotic workers comes amid concerns about its treatment of warehouse staff, with reports of grueling conditions and high turnover rates. The company has faced lawsuits and allegations of fostering a challenging work environment. While labour unions express apprehension about the potential for job losses due to automation, Amazon contends that its robotic systems have created numerous new job categories, emphasising the integral role of human workers in the fulfilment process. The company has already deployed over 750,000 robots in its operations, working alongside human employees to address repetitive tasks. Unlike conventional wheeled robots used in Amazon warehouses, Digit’s legged design enables it to navigate obstacles like steps and stairs. Amazon is currently conducting trials to evaluate its compatibility and safety when working alongside human workers. Amazon Robotics’ chief technologist, Tye Brady, stresses the irreplaceable nature of human workers and dismisses the notion of fully automated warehouses, highlighting their problem-solving abilities and higher-level thinking. Scott Dresser of Amazon Robotics describes Digit as a prototype, and the company’s experience suggests that new technologies create jobs and support growth, as they require human intervention for maintenance. As part of its ongoing automation efforts, Amazon has previously introduced wheeled robots for goods transportation within its warehouses and initiated drone deliveries in select US regions. It plans on delivering within Italy and the UK by the end of 2024.","excerpt":"Digit has a carrying capacity of up to 35 pounds (16kg).","categories":["AI News"],"tags":["Amazon"],"author_name":"Mohit Pandey","publish_date":"2023-10-20T15:12:36","publication_year":"2023","word_count":356,"keywords":["Go","AWS","AI","cloud_platforms:AWS","programming_languages:R","Amazon","Git","automation","GRU","Aim","R"],"extracted_tech_keywords":["AI","Aim","AWS","R","Go","Git","GRU","automation","cloud_platforms:AWS","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/amazon-is-testing-humanoids-in-its-warehouses\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10067386,"title":"DeepMind&#8217;s Gato is the Swiss Army Knife of AI models","content":"The arrival of deep neural networks has been a watershed moment in artificial intelligence history. We have made huge strides in natural language understanding and object recognition in a short period. However, we don’t have AI models that do both. Enter Gato. DeepMind has leveraged the advances in large-scale language modelling to build a single generalist agent beyond the scope of text outputs. Gato is a multi-modal, multi-task, multi-embodiment generalist policy: The same network with the same weights can play Atari, caption images, chat and stack blocks with a real robot arm. How does Gato work? To train Gato, the researchers collected data from different tasks and modalities. The data was then serialized into a flat sequence of tokens, then batched and processed by a transformer neural network. “While any general sequence model can work for next token prediction, we chose a transformer for simplicity and scalability,” the researchers stated in the paper. The researchers have used a 1.2 billion parameter decoder-only transformer with 24 layers and an embedding size of 2048. Gato is trained on many datasets with information about agent experience in simulated and real-world environments. Natural language and image datasets were also used. A prompt is tokenized during the deployment phase to form the initial sequence. Following this, the environment yields the first observation, tokenized and appended to the sequence. Next, Gato samples the action vector autoregressively. It comprehends one token at a time, and once all tokens have been sampled, Gato decodes the action and sends it to the environment. The environment then yields a new observation, and the process is repeated in a loop. “The model always sees all previous observations and actions within its context window of 1024 tokens,” the researchers said. How does Gato stack up against other models? The success stories of GPT-3, Gopher and Flamingo inspired the DeepMind researchers to push the limits of generalist language models and generalist visual language models. Early this year, Google introduced Pathways Language Model (PaLM), building on the Pathways system announced before. The 540-billion parameter, dense decoder-only Transformer model, trained with the Pathways system, was able to train a single model across multiple TPU v4 Pods efficiently. With Pathways, Google Research’s end game is to build a single model that could generalize across domains and tasks while being highly efficient. PaLM achieved state-of-the-art few-shot performance across hundreds of language understanding and generation tasks, and in many cases, by significant margins. In January, Meta AI released data2vec, the first high-performance self-supervised algorithm for multiple modalities. The data2vec outperformed the previous best single-purpose algorithms for computer vision and speech and was competitive on NLP tasks. The algorithm marks a paradigm shift in holistic self-supervised learning. data2vec brings us closer to building machines that can make sense of the world. DeepMind’s Gopher is a 280-billion-parameter NLP model based on the Transformer architecture and trained on 10.5TB of MassiveText. Gopher surpassed the current state-of-the-art on 100 evaluation tasks. The model was also tested on NLP benchmarks, including the Massive Multitask Language Understanding (MMLU) and BIG-bench, and the performance was compared to other baseline models. Gopher showed steady improvement on knowledge-intensive tasks but not so much on reasoning-heavy tasks.In the same league as Gopher, Google’s Generalist Language Model (GLaM) is a trillion weight model that achieves a competitive advantage on multiple few-shot learning tasks. GLaM is a mixture of experts model with different submodels specialized for different inputs. It achieves competitive performance on multiple few-shot learning tasks. GLaM was on-par on seven tasks while using 5x less computation during inference. The tasks included open domain question answering, commonsense reading, in-context reading comprehension, the SuperGLUE tasks and natural language inference.","excerpt":"Gato is a multi-modal, multi-task, multi-embodiment generalist policy.","categories":["AI Features"],"tags":[],"author_name":"Avi Gopani","publish_date":"2022-05-18T18:00:00","publication_year":"2022","word_count":609,"keywords":["Meta AI","artificial intelligence","TPU","AI","neural network","ML","computer vision","RAG","NLP","few-shot learning"],"extracted_tech_keywords":["AI","artificial intelligence","ML","neural network","NLP","computer vision","Meta AI","RAG","few-shot learning","TPU"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/deepminds-gato-is-the-swiss-army-knife-of-ai-models\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":8774,"title":"Analytics Community: Views on &#8216;Start up India &#8211; Stand up India&#8217;","content":"‘The startup India, stand up India’ campaign by our esteemed prime minister, Mr. Narendra Modi has set a new wave of entrepreneurship encouragement in the nation. In recent research done by us, we estimate the amount of funding to analytics startup in 2015 to be in tune of $260M. Being the inquisitive noses that we are, and as a magazine of the analytics community, we decided to find out how this initiative will affect our friends from the Indian analytics industry. We asked four common questions to some of the key players including startups from the analytics community, for their views and found some very interesting answers. Have a read below and do not forget to leave your views on the initiative! Q: What do you have to say on this, being an analytics focused company? Jaydeep Deshpande Jaydeep Deshpande (Regional Marketing Manager, India at Qlik) enunciated that, of $18 billion pumped into Indian startups during 2010 to 2015, $9 billion was invested in 2015 alone! Clearly, there has never been a better time for start-ups in India. Thus, one of the most important aspects for startups is to “compete on analytics” as it becomes absolutely critical for the startups to stay ahead of the curve by engaging in data driven possibilities and by marrying new technologies right from the inception stage. In fact, he strongly believes that this initiative will help create a stronger ecosystem to foster innovation and startups in India. Likewise, Ashwin Mittal (President, Blueocean Market Intelligence), responded that, ‘Startup India – Stand up India’ will encourage many new ventures leveraging digital media, thus indirectly creating opportunity for data mining. Since digital platforms generate huge volumes of structured and unstructured data as well as transactional or operational data, it creates an opportunity for analytics service providers like Blueocean to step in and help these entrepreneurs focus on their core business while providing them with core decision support. Ambiga Dhiraj- COO of Mu Sigma Furthermore, Ambiga Dhiraj (COO, Mu Sigma) believes that there are at least two opportunities for startups in analytics. The first is in serving the Indian market: Large Indian companies, and the public sector, are lagging their global counterparts in the use of analytics for decision support. This initiative could well serve as a market opportunity for new startups. Secondly, there is room for new products to serve vertical opportunities globally across many industries, and also horizontal opportunities such as the Internet of Things. On the other hand, Anurag Shah (CEO and Co-Founder of Aureus Analytics), articulated that even though the overall plan sounds very promising, the execution of this initiative will be critical, as the public procurement for startups is currently referring only to the manufacturing industry and technology startups like theirs will not benefit much from this. On the brighter side, areas like patent rebates will be of great help. Q: Has this initiative helped or affected your business in any way? Shivkumar Shankar Shivakumar Shankar (Managing Director, India, LexisNexis Risk Solutions) finds the campaign an incubator for good ideas, talent, innovation and economic growth. However, he said that it has not impacted their business directly. Since, they constantly look for opportunities to work with companies that complement their business strategy through a partnership, this initiative could prove to be a positive for future growth. Similarly, Jatin Sawhney (General Manager, Information Services, Team Computers Pvt. Ltd) believes that this initiative has opened up many opportunities for them to cross-sell and up-sell their product and services. In addition, Jaydeep Deshpande (Regional Marketing Manager, India at Qlik) commented that as a leader in visual analytics, they will be happy to help startups grow fast as they have been associated with organizations that have understood the need to marry analytics to their strategy and are, consequently, on a fast track growth trajectory. Q: Do you think it will actually boost the analytics startup community in India? Jatin Sawhney This was a critical question as it was directed towards the analytics community as a whole. We cannot agree more with Jatin Sawhney (General Manager, Information Services, Team Computers Pvt. Ltd) who said that the startups are growing at an accelerated pace and with them the analytics community will grow. With many new investments coming in, it’s going to change the way markets function today in India. With everyone wanting to strive toward growth, data crunching and number crunching shall play a vital role in attracting new technologies and talent. Furthermore, Sony Joy (Founder & CEO, Chillr) remarked that analytics startups can focus on getting more data samples before arriving at insights which could change the way various industries function. Looking at recent acquisitions in India, data mining and analytics will play a very important role in developing sustainable business models in the near future. Ashwin Mittal While Ashwin Mittal (President, Blueocean Market Intelligence) added that, digital platforms have seen tremendous growth in the last few years and even existing large enterprises are looking at the digital channels as new business channels to complement their traditional approach. With a wider proliferation of internet through mobile devices, huge volumes of transactional and customer profiling, usage and behavioral data is being generated every second across India. All this data needs to be classified and analyzed and then presented to the decision makers. Thus, it will be a big opportunity for analytics service providers. Anurag Shah Disagreeing, Anurag Shah, (CEO and Co-Founder, Aureus Analytics) claimed that there is nothing specifically for the Analytics focused startups, the overall impetus to the startup ecosystem will have its positive impact. Although, analytics and data science are a critical component of all decision making teams, across every organization and every industry. Maybe with the government sponsored incubators and mentoring in place, it will allow companies of all sizes to experiment with analytics and experience the benefits. Jaydeep Deshpande (Regional Marketing Manager, India at Qlik) anticipates that newer domains such as Big Data, Open Data, Analytics, Data Insights and Visualization are all opening up lucrative opportunities for Indian startup companies. In 2014, NASSCOM had released a report in partnership with Blue Ocean Market Intelligence, predicting the growth of analytics market in India to reach about $2.3 billion by the end of 2017-18. Thus, this campaign will provide an open platform opportunity for local tech companies to integrate best-of-breed solutions, innovate new products and invent solutions for similar global demands. Startups play a critical role in spurring innovation, and India provides a very conducive and vibrant environment for entrepreneurs, especially with a plethora of opportunities around this initiative. Q: Any new laws that you think will affect the analytics startup community? Shivakumar Shankar (Managing Director, India, LexisNexis Risk Solutions) speculated that India is ranked 130th in ease of doing business by World Bank; he further added that, under the new scheme measures undertaken by the government such as 10,000 crore fund for startups, Single point registration, Self-certification, Ease in filing patent applications, Exemption of tax for three years and Credit guarantee funds for startups would help an individual set up their startups in a smooth and efficient manner. Moreover, Ambiga Dhiraj (COO, Mu Sigma) conjectured that new investment capital, and simpler rules to set up new businesses, will raise the tide for all startups, and not just in analytics. As we have a critical mass of analytics talent in our country and all other things being equal, this program will add momentum to Made in India startups. Sony Joy Likewise, Sony Joy (Founder & CEO, Chillr) appreciated the government’s efforts to bring about a change in the way companies are incorporated and ensuring necessary compliance is enforced. Besides promoting an open and free market economy, easing the rules of starting companies will go a long way towards bridging the gap between small and medium size businesses. We would like conclude by thanking our contributors for this article: Ambiga Dhiraj COO, Mu Sigma, Jatin Sawhney, General Manager, Information Services, Team Computers Pvt. Ltd, Sony Joy, Founder & CEO, Chillr, Ashwin Mittal, President, Blueocean Market Intelligence, Anurag Shah, CEO & Co-Founder, Aureus Analytics, Shivakumar Shankar, Managing Director, India, LexisNexis Risk Solutions, Jaydeep Deshpande, Regional Marketing Manager, India at Qlik for their valuable inputs and insights on the campaign.","excerpt":"‘The startup India, stand up India’ campaign by our esteemed prime minister, Mr. Narendra Modi has set a new wave of entrepreneurship encouragement in the nation. In recent research done by us, we estimate the amount of funding to analytics startup in 2015 to be in tune of $260M. Being the inquisitive noses that we are, […]","categories":["AI Features"],"tags":["Analytics Case Study","Analytics India"],"author_name":"Apoorva Verma","publish_date":"2016-01-29T13:51:26","publication_year":"2016","word_count":1369,"keywords":["data science","Go","API","AWS","AI","Analytics India","Git","RAG","Aim","analytics","Analytics Case Study","R"],"extracted_tech_keywords":["AI","data science","analytics","Aim","RAG","AWS","R","Go","Git","API"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/analytics-community-views-on-start-up-india-stand-up-india\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":39143,"title":"Implementing PCA In R With MachineHack’s How To Choose The Perfect Beer Hackathon","content":"Dimensionality Reduction is an important and necessary step when we have big data in hand with so many features. When there are so many features or columns, it is hard to understand the correlation between them. Including weak links or correlations in training-data can also result in an inaccurate prediction by the model.  Dimensionality Reduction takes care of this situation by removing unnecessary features thus helping in fitting the model with the right and most relevant features. Principal Component Analysis is one of the most sought after Dimensionality Reduction techniques in Machine Learning. In one of our previous articles, we saw how PCA can be implemented in Python. In this article, we will learn to implement the PCA in R programming. Readers are expected to have some familiarity with the programming language. Here are some good reads on Dimensionality Reduction : Understanding Dimensionality Reduction Techniques To Filter Out Noisy Data Introduction to DimensionalityReduction A Hands-On Guide To Dimensionality Reduction For implementing PCA in R we will use the How To Choose The Perfect Beer dataset from MachineHack. To get the dataset head to MachneHack, sign up and download the datasets from the Attachment Section. Having trouble finding the data set? Click here to go through the tutorial to help yourself. Lets Code! Importing the Dataset We will be using a smaller subset of the original dataset and it has been preprocessed to some extent. To know more about processing please check out Data Preprocessing With R: Hands-On Tutorial. dataset = read.csv('Beer_dataset.csv') The Dataset is as shown below : Creating Training set and Test set library(caTools) set.seed(123) split = sample.split(dataset$Score, SplitRatio = 0.80) training_set = subset(dataset, split == TRUE) test_set = subset(dataset, split == FALSE) After executing the above code block, we will get two distinct dataframes, the training_set and the test_set. Without further Preprocessing we will directly fit the model to a Linear Regressor Initialising the Model and Fitting the Training Data: # Multiple Linear Regressor mlr = lm(formula = Score ~ . , data = training_set ) print(mlr) Output: Predicting for the Test Set predy = predict(mlr, newdata = test_set) This code block will result in an array consisting of the predictions. Evaluating The Performance actuals_and_preds <- data.frame(cbind(actuals=test_set$Score, predicteds = predy)) The above code block will result in a dataframe consisting of the predictions and the true or actual test data. Output : To evaluate we will consider a more basic MIN-MAX Accuracy. min_max_accuracy <- mean(apply(actuals_and_preds, 1, min) \/ apply(actuals_and_preds, 1, max))print(min_max_accuracy) Output: 0.7055808 Implementing PCA Importing Necessary Libraries install.packages(\"caret\") # Execute Once library(caret) install.packages(\"e1071\") # Execute Once library(e1071) Initializing PCA And Fitting Data pca = preProcess(training_set[-9], method = 'pca', pcaComp = 2) The above code block initializes a pca object and fits the training data. The parameter pcaComp refers to the number of principal components you want the model to return. Here the model will return 2 Principal Components. Generating the Principal Components training_set.pca = predict(pca, training_set) test_set.pca = predict(pca, test_set) This code block will create two new data sets with only the Principal Components and the Actual dependent factor (Score) as columns\/features. Reordering the Columns training_set.pca = training_set.pca[c(2,3,1)] test_set.pca = test_set.pca[c(2,3,1)] Output: training_set.pca: test_set.pca: Using Linear Regression to Predict From The Principal Components mlr_pca = lm(formula = Score ~ . , data = training_set.pca)print(mlr_pca) Output: Predicting With The Principal Components predy_pca = predict( mlr_pca, newdata = test_set.pca) Measuring The Accuracy actuals_and_preds_pca <- data.frame(cbind(actuals= test_set.pca$Score, predicteds = predy_pca)) min_max_accuracy_pca <- mean(apply(actuals_and_preds_pca, 1, min) \/ apply(actuals_and_preds_pca, 1, max))print(min_max_accuracy_pca) Output 0.7301324 Conclusion Comparing the accuracies from the Linear Regression Models, we can see a slight improvement in the accuracy in the model that was fitted with the principal components. The model actually improved with only two features predicting as compared to the original dataset’s 8 independent features. Thus Dimensionality Reduction with PCA holds a significant role in Data Science especially considering the vastness of the data sets that needs to be processed.","excerpt":"Dimensionality Reduction is an important and necessary step when we have big data in hand with so many features. When there are so many features or columns, it is hard to understand the correlation between them. Including weak links or correlations in training-data can also result in an inaccurate prediction by the model.  Dimensionality Reduction […]","categories":["Deep Tech"],"tags":["Dimensionality Reduction","PCA","Principal Component Analysis"],"author_name":"Amal Nair","publish_date":"2019-05-14T13:29:23","publication_year":"2019","word_count":652,"keywords":["big data","data science","Go","machine learning","TPU","AI","ML","Python","Principal Component Analysis","Ray","Dimensionality Reduction","PCA","R"],"extracted_tech_keywords":["AI","machine learning","ML","data science","Ray","TPU","Python","R","Go","big data"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/implementing-pca-in-r-with-machinehacks-how-to-choose-the-perfect-beer-hackathon\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10043444,"title":"Reaching The $2 Tn Mark: Microsoft’s Top AI Projects","content":"After Apple, Microsoft recently became the only publicly traded American company to hit the $2 trillion market cap. The company has reached the milestone just two years after it crossed the $1 trillion mark. In this article, we list major AI projects and initiatives the company undertook post-2019. Investment in OpenAI In 2019, Microsoft said it would invest $1 billion in OpenAI to build artificial general intelligence. The partnership is directed at developing a hardware and software platform within Azure geared towards AGI. Under this deal, Microsoft also became OpenAI’s exclusive cloud provider. OpenAI said in a blog, “we’ll be working hard together to further extend Microsoft Azure’s capabilities in large-scale AI systems.” Earlier, Microsoft has built one of the top five supercomputers globally in collaboration with OpenAI. The new infrastructure is available on Azure for researchers to train large AI models. GPT-3 GPT-3 was one of the most groundbreaking inventions of 2020. When OpenAI unveiled the massive language model with 175 billion parameters, it was ten times larger than any previous non-sparse language model. GPT-3 has applications in tasks such as unscrambling words, using a novel word in a sentence and performing arithmetic operations (on up to 3-digit numbers). Credit: OpenAI In September, Microsoft teamed up with OpenAI to exclusively license GPT-3. Microsoft said the exclusive license would help the company to expand their Azure-powered AI platform to democratise AI technology, develop new products and services, and drive the impact of AI at scale. For example, the tech giant will be able to ‘directly aid creativity and ingenuity’ in writing and composition, summarising large blocks of long-form data, and translating natural language. At the Build developers conference, Microsoft unveiled its first features in a customer product powered by GPT-3. It is now integrated with Microsoft Power Apps, a low code app development platform, to help both ‘citizen developers’ and professional developers. It includes apps to review non-profit gift donations, reduce overtime required to maintain wind turbines, and manage travel during COVID-19. On the flip side, critics panned the decision to hand over the exclusive GPT-3 license to Microsoft. Handing over privileges of such breakthrough innovations in AI could lead to a significant power imbalance. Falcon framework In collaboration with Princeton University, Technion and Algorand Foundation, Microsoft introduced Falcon, an end-to-end 3-party protocol for fast and secure deep learning computations on a large network. It’s a hybrid integration of ideas from SecureNN and ABY3 and new protocol constructions for privacy-preserving deep learning. The main advantages of this framework include: It is a highly expressive framework and a secure platform to support high capacity networks with a hundred million parameters.Falcon guarantees security against malicious adversaries. It ensures the protocol always produces correct output and aborts the cycle when a hostile adversary is detected.Falcon presents new theoretical insights for protocol design that make it highly efficient and outperform existing, secure in-depth learning solutions. Microsoft’s AI acquisitions In the last two years, Microsoft has made some heavy bets in the field of emerging technologies. The major buys include UK-based robotic process automation company Softomotive and Israeli cybersecurity startup CyberX. Microsoft also bought Nuance for a whopping $19.7 billion–its second-costliest buy after LinkedIn ($26.2 billion in 2016). Massachusetts-based Nuance’s voice recognition technology power’s Apple’s virtual assistant, Siri. Nuance also makes software for the healthcare and automotive sectors. With Nuance, Microsoft hopes to expand its capabilities in voice recognition and transcription technology. Nuance’s clinical speech recognition products include Dragon Medical One, PowerScribe One, and Dragon Ambient eXperience. This year, Microsoft acquired The Marsden Group, which uses AI and IIoT for industrial automation and safety processes. The Houston-based company supports offshore oil, industrial plants, and original equipment manufacturing facilities. One of its most important solutions is Vision IQ (deployed by Seadrill), which uses special cameras and AI to track personnel of oil derricks and alert them about hazards. Microsoft was reportedly also in talks to buy Discord. The purported deal created a lot of buzz. However, the deal fell through. Microsoft & AI education Microsoft has introduced several AI courses, especially during the pandemic, most of them free. The company has also collaborated with universities and edtech platforms like Udacity to offer AI and machine learning courses online. The latest in this list includes a free MIT-approved learning course “Machine Learning For Beginners” to teach students the basics of machine learning. R&D projects Neural Text-to-Speech AI: Microsoft announced limited access to its Neural Text-To-Speech (TTS) AI. This tool enables developers to create custom synthetic voices. The clients, including AT&T, Duolingo, Progressive, and Swisscom, have tapped the Custom Neural Voice feature to develop a unique speech solution. Credit: Microsoft Seeing AI: Microsoft introduced the Seeing AI app for iOS. This app, designed specifically for visually impaired persons, uses the device camera to identify people and objects and describe them. MusicBERT: It is a large-scale pre-trained model for symbolic music understanding. Symbolic music understanding refers to understanding music from the symbolic data (for example, MIDI format). The model can do emotion classification, genre classification, and music piece matching. Credit: Microsoft Counterfeit: Microsoft recently released Counterfit – a tool to test the security of AI systems – as an open-source project. This tool is touted to be the next big thing in building a robust AI ecosystem.","excerpt":"After Apple, Microsoft recently became the only publicly traded American company to hit the $2 trillion market cap.","categories":["Global Tech"],"tags":["Apple","GPT-3","Microsoft"],"author_name":"Shraddha Goled","publish_date":"2021-07-13T10:00:00","publication_year":"2021","word_count":878,"keywords":["GPT-3","Go","machine learning","TPU","OpenAI","AI","Apple","R","Git","RAG","deep learning","Azure","Microsoft"],"extracted_tech_keywords":["AI","machine learning","deep learning","OpenAI","RAG","Azure","TPU","R","Go","Git"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/reaching-the-2-tn-mark-microsofts-top-ai-projects\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10111738,"title":"Databricks Acquires AI-powered Data Visualisation Platform Einblick","content":"The data lakehouse provider Databricks today announced it has acquired the team behind Einblick, a provider of a natural language data science notebook that purports to allow enterprise users to ‘solve any data problem with just one sentence’. “We look forward to integrating Einblick’s innovative, AI-native approach directly into the Databricks platform. This integration will empower organizations to democratize data intelligence and foster the creation of the next generation of data and AI applications with quality, speed, and agility. Stay tuned in the coming months to hear more about our joint plans,” the company said in its blog post. Databricks praised Einblick’s capabilities saying “The Einblick team has pioneered state-of-the-art techniques for translating natural language questions into the code, charts, and models needed to generate insights. This not only requires applying the latest generative modeling capabilities but also involves highly optimizing them for data-related questions”. Founded in 2019 by Emanuel Zgraggen, Philipp Eichmann and others, Einblick was an early innovator harnessing the power of natural language as a foundational UI principle. The company aims to enable everyone to make an impact through data, not opinions. “We’re thrilled to welcome Emanuel, Philipp and the talented Einblick team to Databricks. Like us, they believe that organizations need to deeply leverage data and AI to be successful in this new age”, said Databricks. This marks the fourth acquisition by Databricks in the past 12 months, adding to a series that began with the purchase of data replicator Arcion for $100 million in autumn 2023. Prior to that, the company acquired Mosiac ML in an impressive $1.3 billion deal, and the data governance platform Okera, the financial details of which remain undisclosed. Databricks didn’t disclose the price tag for which it acquired Einblick. Nevertheless, it’s evident that this strategic move is aimed at enhancing and expanding Databricks’ portfolio, positioning itself as the comprehensive solution for enterprises’ data storage and analysis requirements.","excerpt":"This marks the fourth acquisition by Databricks in the past 12 months.","categories":["AI News"],"tags":["Mergers and Acquisitions"],"author_name":"Siddharth Jindal","publish_date":"2024-01-31T17:03:24","publication_year":"2024","word_count":317,"keywords":["data science","Go","AI","ML","RAG","Aim","data governance","Mergers and Acquisitions","R","data lake","Databricks"],"extracted_tech_keywords":["AI","ML","data science","Aim","RAG","Databricks","R","Go","data lake","data governance"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/databricks-acquires-ai-powered-data-visualisation-platform-einblick\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":40250,"title":"Beginner’s Guide To Understanding Apriori Algorithm With Implementation In Python","content":"In this article, we will talk about Apriori Algorithm which is one of the most popular algorithms in Association Rule Learning. So before we dig deep into Apriori, let’s try to understand what Association Rule Learning means. What is Association Rule Learning? Association Rule Learning has the most popular applications of Machine Learning in business. It has been widely used to understand and test various business and marketing strategies to increase sales and productivity by various organizations including supermarket chains and online marketplaces. Association Rule Learning is rule-based learning for identifying the association between different variables in a database. One of the best and most popular examples of Association Rule Learning is the Market Basket Analysis. The problem analyses the association between various items that has the highest probability of being bought together by a customer. For example, the association rule, {onions, chicken masala} => {chicken} says that a person who has got both onions and chicken masala in his or her basket has a high probability of buying chicken also. Apriori Algorithm The algorithm was first proposed in 1994 by Rakesh Agrawal and Ramakrishnan Srikant. Apriori algorithm finds the most frequent itemsets or elements in a transaction database and identifies association rules between the items just like the above-mentioned example. The algorithm uses a “bottom-up” approach, where frequent subsets are extended one item at once (candidate generation) and groups of candidates are tested against the data. The algorithm terminates when no further successful rules can be derived from the data. How Apriori works To construct association rules between elements or items, the algorithm considers 3 important factors which are, support, confidence and lift. Each of these factors is explained as follows: Support: The support of item I is defined as the ratio between the number of transactions containing the item I by the total number of transactions expressed as : Confidence: This is measured by the proportion of transactions with item I1, in which item I2 also appears. The confidence between two items I1 and I2,  in a transaction is defined as the total number of transactions containing both items I1 and I2 divided by the total number of transactions containing I1. Lift: Lift is the ratio between the confidence and support expressed as : Implementing Apriori With Python Let us consider a simple dataset consisting of a thousand observations of the movie interests of a thousand different people. We will use the data to understand different associations between different items in this case movies. The objective is to identify the chances of a person watching a movie given he has already watched other movies. Before we begin our coding we need to install the apyori package. To install the package, open the terminal or command prompt, type in and enter the following command : pip install apyori Note: Do not forget to activate your conda environment with <code>conda activate<\/code> if you are working with anaconda. Lets code! Importing the dataset import pandas as pd data = pd.read_excel(\"Movie_reccommendation.xlsx\") Let’s have a look at the dataset : Converting the data frame into lists The algorithm in the apyori package is implemented in such a way that the input to the algorithm is a list of lists rather than a data frame. So we need to convert the data into a list of lists. observations = [] for i in range(len(data)): observations.append([str(data.values[i,j]) for j in range(13)]) Fitting the data to the algorithm from apyori import apriori associations = apriori(observations, min_length = 2, min_support = 0.2, min_confidence = 0.2, min_lift = 3) Where, min_support: The minimum support of relations (float) min_confidence: The minimum confidence of relations (float) min_lift: The minimum lift of relations (float) min_length: The minimum number of items in a rule max_length: The maximum number of items in a rule The optimum values for min_support,min_confidence and  min_lift arguments can be set by trying out different values and checking the association rules whether the arguments produced a valid association between items or not. Once we execute the above code block, the algorithm returns 37 rules based on the set parameters of min_length = 2, min_support = 0.2, min_confidence = 0.2 and min_lift = 3 Converting the associations to lists associations = list(associations) Understanding the rules The apriori algorithm automatically sorts the associations’ rules based on relevance, thus the topmost rule has the highest relevance compared to the other rules returned by the algorithm. Let’s have a look at the first and most relevant association rule from the given dataset. print(associations[0]) Output: RelationRecord(items=frozenset({'ghost in the shell', 'ex machina'}), support=0.327, ordered_statistics=[OrderedStatistic(items_base=frozenset({'ex machina'}), items_add=frozenset({'ghost in the shell'}), confidence=1.0, lift=3.058103975535168), OrderedStatistic(items_base=frozenset({'ghost in the shell'}), items_add=frozenset({'ex machina'}), confidence=1.0, lift=3.058103975535168)]) Rule one is the most relevant rule that the algorithm identified from the given dataset. The above output specifies the association between two movies ‘ghost in the shell’ and ‘ex machina’. The two movies have the support of 0.327. i,e. support(‘ghost in the shell’, ‘ex machina’) =0.327 Also from the output, confidence(‘ex machina”ghost in the shell’)=1 confidence(‘ghost in the shell”ex machina’)=1 This implies that if a person has watched ‘ex machina’ he or she is 100% likely to watch ‘ghost in the shell’ and also if a person has watched ‘ghost in the shell’ he or she is 100% likely to watch ‘ex machina’. The lift of 3.058103 shows the relevance of the rule since we have only chosen the rules with a minimum relevance of 3.","excerpt":"In this article, we will talk about Apriori Algorithm which is one of the most popular algorithms in Association Rule Learning. So before we dig deep into Apriori, let’s try to understand what Association Rule Learning means. What is Association Rule Learning? Association Rule Learning has the most popular applications of Machine Learning in business. […]","categories":["Deep Tech"],"tags":["Apriori Algorithm","Data Mining","data mining tools","Deep Learning","Machine Learning","ML","recommendation"],"author_name":"Amal Nair","publish_date":"2019-06-05T11:23:26","publication_year":"2019","word_count":900,"keywords":["Go","machine learning","TPU","AI","Data Mining","ML","Machine Learning","GAN","Python","data mining tools","programming_languages:Python","ViT","recommendation","Deep Learning","R","Apriori Algorithm","Pandas"],"extracted_tech_keywords":["AI","machine learning","Pandas","TPU","Python","R","Go","GAN","ViT","programming_languages:Python"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/beginners-guide-to-understanding-apriori-algorithm-with-implementation-in-python\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10071653,"title":"TorchRec Sharding &#8211; A framework to build large-scale recommender systems","content":"TorchRec is one of the libraries of Pytorch that is used to build sparse and large-scale recommendation systems. The library supports sharding which means that large tables can be sharded across GPUs and be trained. So the problems associated with higher dimensional entity tables and exceeding GPUs memory capability can be overcome by using TorchRec Sharding. So using TorchRec Sharding, distributed training of the model across accelerator-based platforms can be achieved. In this article, let us understand about TorchRec sharding with respect to this context. Table of Contents Introduction to TorchRec ShardingBenefits of TorchRec ShardingVarious schemes of TorchRec ShardingSummary Introduction to TorchRec Sharding TorchRec Sharding is one of the libraries of Pytorch structured to overcome the problems associated with large-scale recommendation systems. Some of the large-scale recommendation systems may require the representation of higher dimension embeddings. These embeddings may sometimes encounter maximum GPU memory usage. So to prevent the problems associated with higher dimensional embeddings and GPU memory constraints, PyTorch has structured a single shot library, especially for large-scale recommendation systems, where the recommendation systems can be built by sharing the training process of the recommendation model across GPUs. Different embeddings of large-scale recommendation systems of Pytorch are represented by using the inbuilt function of Pytorch named torch.nn.EmbeddingBag. Are you looking for a complete repository of Python libraries used in data science, check out here. An embedding Bag is basically a collection of various embeddings of the data. So this collection of Embeddings is used to configure required entities suitable for recommendation systems. So the required data entities will be sharded across different GPUs with a block size of 64 and an Embedding dimension of 4096. This benefits the required entities to be trained faster and prevents the maximum memory consumption of accelerators like GPUs. Understanding the Sharding API TorchRec library supports sharing higher dimensional data across accelerators and trains large-scale recommendation systems. So for distributing data across multiple accelerators and to facilitate parallel training TorchRec has formulated an API named DistributedModelParallel. The API is responsible for carrying out two functionalities. They are as follows. i) Decision-making on how to shard the model across accelerators. The API will collect all the available sharders and come up with an optimal way to shard the embedded tables across platforms. ii) The API will also be responsible for allocating memory across accelerator platforms and will be responsible for allocating relational tables across platforms. Benefits of TorchRec Sharding The benefits of TorchRec sharding are listed below. Modeling higher dimensional embedding tables on various accelerator-based platforms like GPUs accelerates the training process of required entities of recommendation systems. Optimized kernels for large-scale recommendation systems with the ability to perform sparse and quantized operations. Various sharders help in partitioning various tables for recommendation systems with various strategies like row-wise and column-wise sharding. Sharing embedding tables across distributed platforms helps in speeding up the training process of larger embedding tables. Regulates pipeline training through an inbuilt function named TrainPipelineSparseDist which helps to increase the performance and enable parallel processing of data across accelerator platforms. An overview of various schemes of TorchRec Sharding TorchRec sharding basically uses two schemes for embedding tables across platforms. The two schemes that TorchRec uses are known as EmbeddingPlanner and DistributedModelParallel. So let us try to understand the benefits of this scheme that is used by TorchRec Sharding for handling large-scale recommendation systems. Embedding Planner The Embedding Planner scheme of TorchRec sharding makes use of a collection of embeddings available in the EmbeddingBagCollection. In this scheme, there are two tables majorly being used known as large and small tables which are differentiated based upon the difference in the row sizes. Some of the parameters of the table are configured to help in wise decision-making for sharding across various accelerator platforms. This scheme makes use of various sharding techniques to communicate among the various embedded tables and facilitate sharding accordingly. DistributedModelParallel The Distributed Model Parallel scheme operates on the principle of SPMD. SPMD abbreviates for Single Program and Multiple Data. The scheme uses some standard code to shard the model collectively among various processes and shards the tables on multiple accelerators like GPUs based upon standard specifications. This also supports multiprocessing and table-wise sharding for some of the important features of recommendation systems. The two schemes of sharding ultimately prioritize handling larger dimension tables and distribute larger dimensions to various parallel processing accelerator platforms to accelerate the training process. Summary TorchRec sharding is one of the libraries of Pytorch formulated to model large-scale recommendation systems and train them on multiple GPUs to evacuate the problems associated with an overload of memory consumption of accelerator-based platforms. It facilitates training on multiple devices by splitting huge recommendation system embeddings. This enables the recommendation system to be trained faster. This library is still in the research phase, and more improvements can be expected in the future, making modeling large-scale recommendation systems easier. Summary TorchRec Official Documentation","excerpt":"Pytorch has formulated a framework named TorchRec sharding that is suitable for training and modelling large scale recommendation systems.","categories":["AI Trends"],"tags":[],"author_name":"Darshan M","publish_date":"2022-07-28T11:00:00","publication_year":"2022","word_count":819,"keywords":["data science","API","programming_languages:R","PyTorch","AI","recommendation systems","Python","ai_frameworks:PyTorch","programming_languages:Python","R"],"extracted_tech_keywords":["AI","data science","PyTorch","recommendation systems","Python","R","API","ai_frameworks:PyTorch","programming_languages:Python","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/torchrec-a-framework-for-large-scale-recommendation-systems\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10129653,"title":"10 Bizarre AI-Generated Videos","content":"AI video generators are reaching a tipping point, and just after OpenAI’s Sora’s text-to-video prowess swept social media, a new challenger from China, Kling AI, is doing the same. Both Kling and Sora use powerful technology to create videos that look incredibly real. For instance, Kling can make use of 3D reconstruction to create a video of a panda chowing down on ramen noodles, complete with all the wrinkles and fur details! But the race doesn’t stop there. Luma AI, a California-based startup specialising in visual AI, has thrown its hat into the ring with Dream Machine. This new video generator utilises cutting-edge AI to create stunningly realistic visuals. From animals engaging in hilariously human activities to surreal scenarios that defy logic, here are the Top 10 mind-blowing videos produced by AI. Animals Enjoying Human Activities The AI-generated videos are endearing and funny, from a kangaroo slurping up some udon to a panda gorging on a bowl of ramen. The degree of realism and detail is evidence of Kling’s powers and the advancements in AI. Thousands of people have reacted to and shared the videos from the post, which has since gone viral. Numerous people have also voiced their surprise at the concept’s inventiveness and the visuals’ high calibre. https:\/\/twitter.com\/_akhaliq\/status\/1812573686152450397 Polar Bear’s Unexpected Adventures A new video by Kling shows a polar bear doing things no polar bear has ever done. First, the bear is riding a mountain bike down a trail.  Next, the bear is in a Formula One car, racing at high speeds. We see the bear belly dancing in a sparkly outfit. Finally, the bear sings karaoke! This video is more than just funny. It shows how Kling can make unique and surprising videos. AI Reveals Eye’s Hidden Beauty The video shows an extreme close-up of a 24-year-old woman’s eye blinking, standing in Marrakech during magic hour. The cinematic film was made to look like it was shot in 70mm with depth of field and vivid colours. It was created using Luma AI’s Dream Machine. This clip delves into the intricacies of the human eye, capturing its mesmerising colours. The AI algorithms make use of light refraction, intense colour, and slow zoom to create a realistic and captivating scene. Canine Code Cracker An internet video featuring a Labrador retriever using a computer has gone viral! With its paws flying across the keyboard, the dog appears serious, dressed in an oversized black hoodie. The screen light illuminates the dog’s face. It’s not clear what the dog is doing. Maybe it’s playing a game, or maybe it’s just typing randomly. The video doesn’t tell us, but it’s definitely unusual to see a dog using a computer like this! The video was entirely generated by Sora. https:\/\/www.youtube.com\/watch?v=4vwc2TLibTI Animal Boxing Tournament: Made by Kling user Lightning Kiki, this video shows a series of shots that were used to build up to an animal boxing tournament. The video shows a rabbit and a tiger in a boxing match with a crocodile as the match referee and cats as judges.  It is a part of Kling’s “real world Zootopia” strand. Deepfakes and Reboots The video, created by YouTuber EZRyderX47, is a fantastic illustration of deepfake technology’s artistic potential. This video overlays Robert Downey Jr. and Tom Holland’s features over those of Christopher Lloyd and Michael J. Fox, giving us a realistic idea of what it may look like if the Avengers resurrect the beloved 1985 film Back to the Future. Animated Mona Lisa Microsoft unveiled the VASA-1, a new AI model that can virtually and flawlessly merge an audio clip of a speaker with a still image of a face. Researchers demonstrated how they animated the Mona Lisa to recite an amusing rap by actress Anne Hathaway in one demo film. Celebrity Reunions An approximately minute-long video of Hollywood stars embracing their younger selves was shared by a user on X. Emma Watson, Daniel Radcliffe, Eminem, Keanu Reeves, Tom Hanks, Will Smith, Michael Jackson, Johnny Depp, and Harrison Ford were some of the stars included in the AI-generated video. The video was created using Kling AI. pic.twitter.com\/c2SVrdsotG— Ja ckie D (@DJACKie1115) July 17, 2024 AI Makes People Melt The video is an assemblage of AI-generated gymnastics footage using Dream Machine. The shifts from one human shape to another are horrifying, yet curiously alluring, as sportsmen with four legs blend into themselves and their surroundings. It resembles staring into a funhouse mirror while high on psychedelics and mixed in with some glitchy digital artefacts. Gymnastics is the Turing test of video generation models pic.twitter.com\/cOhmUJjI2m— Deedy (@deedydas) July 2, 2024 Life-Like Animals Take Centre Stage The next few videos, made from a series of shots rather than a single clip, were shared by the Kling team on YouTube. They feature life-like animals performing human tasks and activities, including a rabbit reading the newspaper, a monkey sipping coffee, a giraffe, a chimpanzee, and a fox participating in a bike race. The video was created by Kling AI.","excerpt":"From animals engaging in hilariously human activities to surreal scenarios that defy logic, here are the Top 10 mind-blowing videos produced by AI","categories":["Deep Tech"],"tags":["AI (Artificial Intelligence)","AI Video Generation Models","kling","Microsoft","Sora"],"author_name":"Anshul Vipat","publish_date":"2024-07-20T16:12:32","publication_year":"2024","word_count":831,"keywords":["Go","ELT","OpenAI","AI","AWS","ML","kling","Git","AI Video Generation Models","BERT","edge AI","Sora","AI (Artificial Intelligence)","R","Microsoft"],"extracted_tech_keywords":["AI","ML","OpenAI","edge AI","AWS","R","Go","Git","ELT","BERT"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/10-bizzare-ai-generated-videos\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10005944,"title":"Decoding Most Used, Confused &#038; Abused Jargons In Machine Learning","content":"Machine learning is one term that has created an immense amount of buzz in the technology industry. With its enormous potential in healthcare, medical diagnosis as well as solving complex business problems, machine learning has revolutionised many aspects of human lives. However, as the technology is evolving and becoming more complex, it comes up with exciting subfields and terminologies associated with it. In this article, we are going to decode some of the most used jargons in machine learning. Also Read: 25 AI Terminologies & Jargons You Must Assimilate To Sound Like A Pro Autoregression Autoregression is a phenomenon in machine learning, where an autoregressive model learns from a series of times steps, aka a time series model that uses information from previous timed steps as input to a regression equation in order to predict the value. With autoregression, one can predict accurate forecasts on a range of time series problems. It works based on determining the correlation of previous time steps, also known as the variables, among each other, which in turn helps in predicting the output. If both the variables changed in the same direction, there could be a positive correlation; however, if both turned to a different direction, then it will be termed as negative — either of the ways the relationship between the result and the input can be easily determined. With a higher correlation, the more chances of predicting the outcome from the past information. To understand better read: Python Library For Time Series Analysis And Prediction. Backpropagation In machine learning, backpropagation is also known as “backward propagation of errors,” and is an algorithm used for training artificial neural networks for supervised learning. It works by determining the minimum value of the error at the output and then propagating it back into the neural network. Backpropagation is a critical process of neural net training, where it is leveraged for fine-tuning the weights based on the error rates. Proper modification of weights will help in minimising the errors, making the model reliable. Not only this method is fast and easy to program, but also needs no parameters to tune and no prior knowledge about the network. Static backpropagation and recurrent backpropagation are the two types of backpropagation networks. Although it comes with many benefits, the only drawback is the sensitivity of the method for noisy data. You can also read what Geoff Hinton thinks of Backpropagation. Few-shot Learning\/One-Shot Learning Few-shot learning is the type of model training where a very small set of training data is used instead of an extensive one. It carries a suitable object categorisation model work without several training examples. The best way for this is by learning the common representation of various tasks and train them on task-specific classifiers. It is best used when acquiring enough data to optimise the model isn’t possible, thus, few-shot learning can be beneficial for such scenarios in order to identify the patterns in the data and predict the outcome. Whereas, one-shot learning is an approach in machine learning where the model has been trained only a single instance, usually used for object classification. One of the renowned examples of one-shot learning is the facial recognition system that people deploys. Typically to train deep learning systems, one would require several input data with pictures of human faces for them to finally identify that face in the crowd. However, it is never determined that train the model with images; it might not naturally react when it identifies a new face. Thus one-shot learning comes handy, where the conventional neural network is being trained to learn a distance function between images instead of classifying them. To understand better, read this: Will One-Shot Learning Using Hypercubes Outrank Traditional Neural Nets? Also Read: Road To Machine Learning Mastery Hyperparameter In machine learning, model parameters are known as the properties of training data which can learn independently during training by machine learning models. Some of the critical parameters are weights and biases. With that being said, a hyperparameter is a type of parameter that dictated the entire training process of the model. Its value is decided before training the model and is basically used to control the learning process. On the other hand, usually, the values of different parameters are derived with model training. The number of hidden units and learning rate are some of the key model hyperparameters used. With direct control of the training process, hyperparameters are critical in measuring the performance of the model. Thus, choosing appropriate hyperparameters can provide easier management of extensive experiments. Hyperparameters can be of two categories — optimiser and model specific. For a better understanding of hyperparameter optimisation tools, read this. Recommendation Engine Recommendation Engine is basically used for recommending customers their favourite products on an online platform. It is a data filtering tool that uses algorithms to recommend the most relevant and preferred items for a particular user. To facilitate this, firstly, it identifies the past behaviour of the customer and leveraging that information, it suggests products and items for the customers. However, for customers with no prior buying pattern, the engine proposes products that are the best selling for the site. Apart from recommending products, the recommendation engine also segments the customers according to their preferences and manages them accordingly for future buys. In order to build a successful recommendation system, one has to accumulate data about the products, demographics, as well as the customer. The data is then filtered to extract relevant information to make the required suggestions. To understand how Google made it super easy to build a recommendation engine, read this. Tokenization Tokenization is the process of transforming data into tokens, explicitly used for NLP. For instance, if the data is an account number; then this process will turn that account number into a string of characters also called tokens. These tokens are used as a reference for the data but cannot be used to guess those values. Being the building blocks of natural language, tokens have been the most leveraged way of processing data. Some of the key types of tokenization include — word tokenization, character tokenization and subword tokenization. Tokenization can be done via many approaches, whether it be white space tokenization or dictionary-based tokenization. In white space tokenization, the input sentence is broken apart every time a white-space is encountered, whereas, the dictionary-based tokenization, determines tokens from the sentences that are already in the dictionary. For better understanding, read: Hands-on Guide to StanfordNLP Optimizer As the name suggests, optimizers are the type of algorithms and techniques used to optimise the neural networks during its training process; thus, the method is known as ‘optimization.’ These algorithms are leveraged to modify the attributes of the neural network, such as the learning pace or the weights in order to minimise the loss function as well as determine accurate results. In layman terms, the optimisers transform the model in a perfect form by working around its weights to determine precise results. One of the most popular optimisation algorithms used is the ‘Gradient Descent,’ which works on linear regression and classification algorithms to calculate the attributes of the neural networks. Other types of optimisers are — Momentum, Adagrad, Root Mean Square Propagation (RMSProp), Ada Delta, Nesterov and Adam, to name a few. To know more about optimisers, read — Guide To Optimizers For Machine Learning. Convergence Convergence is a method or technique where an iterative algorithm aka the initial method to predict the outcome, converges when the result gets closer to a particular value. To be exact, not depending on the error range, if convergence continued long enough, the function would eventually remain within that mentioned error range around the final value. In layman terms, when the data is processed over several times, the model converges to represent latent variables or the errors throughout the weights of the neutrons in the network. It is also believed that inappropriate convergence would lead to longer training time and massive data for the machine learning algorithm. Learning Rate Annealing The process of training neural networks involves several hyperparameters, and one of the key, among those, is the learning rate for gradient descent. This hyperparameter determines the magnitude of the weights in order to reduce the losses. The speed of the learning rate is equivalent to the process of training the neural networks. However, if the learning rate is higher than the desired, It can create undesirable divergences in the loss function. Now, to adjust the learning rate, one of the fundamental techniques is learning rate annealing, which starts by recommending a high learning rate and then gradually slowing it down during the training process. This again comes in many forms, and the most popular way of annealing the learning rate is by ‘step decay,’ which reduces the learning rate to a certain extent after trying out some training methods. In fact, it is critical to define the learning rate schedule, where it can get updated during the training process and can be done with annealing. To understand better read: Top Optimisation Methods In Machine Learning. Batch Normalization Similar to other techniques, batch normalisation is a method of making the neural network faster and more stable by re-centring and re-scaling. To explain the process, this technique normalises the input layer by modifying and scaling the activations and allows each of the layers to learn independently, minimising or normalising the hidden layers. With batch normalisation, one can also increase the learning rate, as the technique ensures the activation to be in the right value — neither high nor low. With fewer regularisation effects, it can also reduce the overfitting of the model. Further, leveraging the process of batch normalisation can reduce the change in the distribution of input variables, thus speeding up the training process.","excerpt":"Machine learning is one term that has created an immense amount of buzz in the technology industry. With its enormous potential in healthcare, medical diagnosis as well as solving complex business problems, machine learning has revolutionised many aspects of human lives.  However, as the technology is evolving and becoming more complex, it comes up with […]","categories":["Deep Tech"],"tags":["how to calculate mean square error","Machine Learning","machine learning optimization","machine learning pattern recognition python"],"author_name":"Sejuti Das","publish_date":"2020-09-03T14:00:00","publication_year":"2020","word_count":1630,"keywords":["machine learning pattern recognition python","machine learning","TPU","how to calculate mean square error","AI","neural network","Machine Learning","RAG","NLP","Python","deep learning","few-shot learning","machine learning optimization","R"],"extracted_tech_keywords":["AI","machine learning","deep learning","neural network","NLP","RAG","few-shot learning","TPU","Python","R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/decoding-most-used-confused-abused-jargons-in-machine-learning\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":44899,"title":"Qubole Announces India’s First-Ever Presto Summit In Bengaluru","content":"Let’s face it — companies want to accelerate data science journey but lack modern data management practices that can combine data — both structured and unstructured — from various platforms like data lakes and warehouses. Now, at a time when organizations are pivoting to multi-cloud environments, there is a greater need for a cloud-native data platform for analytics, machine learning and artificial intelligence. One of the companies which is plugging the data activation gap by automating ETL, machine learning and analytics workloads in the cloud is Qubole — the world’s leading cloud-native platform for machine learning. The move towards multi-cloud environments is a natural progression for enterprises but it also ushers in a need for a cloud-native data platform to crunch petabytes of data. To foster dialogues and forward-looking discussions about big data activation in the cloud, California-headquartered Qubole is hosting the first-ever Presto Summit in India. To be held on September 5 in Bengaluru the summit will feature keynotes from Presto creators Martin Traverso, Dain Sundstrom and David Phillips who will talk about the fastest-growing distributed SQL-on-anything engine. Presto is leveraged by top organizations to build a new innovative technology that enables interactive analytics on top of data lakes. The summit will bring together the Presto community in India and open source contributors and Presto users from leading data-driven companies in the APAC, who will discuss all aspects of PrestoDB, the open-source distributed SQL query engine, covering usage scenarios, architectural patterns, deciding the right configuration and more. The summit provides a very exciting opportunity to interact with India’s top Presto developers and data engineers, the Presto creators, and also Apache Hive co-founder Joydeep Sen Sarma. “It is super exciting to see how the journey that started from our Hive days, has evolved all the way to Presto today, enabling lightning-fast SQL on big data lakes seamlessly.” It is also a great platform to learn more about Presto, the platform-agnostic technology that provides a seamless way to access data from various sources using standard SQL query. Enterprises across the globe are leveraging Presto for a number of use cases — implementing as an enterprise-wide query engine, reporting solution, self-service BI and a more effective way to approach data with existing tools. This is your chance to learn about Presto and how developer\/specific users can get started with and why it is the most agile solution for querying data and deriving actionable insights. Don’t miss out the tech talks by data engineering leaders from WalmartLabs, CRM player Zoho, South East Asia’s leading super-app Grab, and more. Not just that, the summit provides a platform to network with business leaders from APAC’s top data-driven companies such as Akamai, Amazon, DellEMC, Expedia, Flipkart, FutureGroup, GE, Goldman Sachs, Grab, Hotstar, Infosys, InMobi, Intuit, Microsoft, Myntra, Nutanix, OlaCabs, Oracle, Paytm, Philips, PhonePe, Razorpay, Salesforce, United Healthcare, Visa, VMWare, Walmart Labs, Yatra.com, Zapr Media Labs, ZeoTap, Zoho and Zomato, and present a great opportunity to understand the solutions they are bringing to the market. Event: Presto Summit Bengaluru When: September 5, 2019 Where: Courtyard By Marriott, ORR, Bengaluru To know more about the event, click here To register, click here","excerpt":"Let’s face it — companies want to accelerate data science journey but lack modern data management practices that can combine data — both structured and unstructured — from various platforms like data lakes and warehouses. Now, at a time when organizations are pivoting to multi-cloud environments, there is a greater need for a cloud-native data […]","categories":["Deep Tech"],"tags":["no etl","salesforce crm","Walmart Labs"],"author_name":"Richa Bhatia","publish_date":"2019-08-24T09:39:41","publication_year":"2019","word_count":524,"keywords":["data science","Go","artificial intelligence","machine learning","AI","ML","Walmart Labs","RAG","no etl","salesforce crm","analytics","SQL","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","data science","analytics","RAG","R","SQL","Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/qubole-announces-indias-first-ever-presto-summit-in-bengaluru\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10090738,"title":"BlackBerry Boys Are Back","content":"Steeped in luxury, BlackBerry was once a household name in the smartphone industry, known for its iconic physical keyboard and BBM, the secure messaging service. The company’s last developed phone was the Blackberry Key2LE that came in 2018. And despite signing a new licensing agreement for smartphones with US-based startup OnwardMobility in August 2020, the Canadian tech giant did not release any new phones. Sadly, in January 2022, the company announced its departure from the handset software market, leaving users of the older BlackBerry models unable to reliably send texts, make phone calls, or access data. But that doesn’t mean the company is out of the game. BlackBerry boys, are back. Today, BlackBerry has shifted its focus to the Internet of Things (IoT) and enterprise security software, while completely distancing itself from the smartphone industry. In fact, the company recently sold most of its smartphone-related patents for $900 million to Malikie Innovations Ltd. In 2019, the company had acquired cybersecurity firm Cylance for $1.4 billion. The purchase added AI and machine learning capabilities to BlackBerry’s portfolio and solidified its position in the cybersecurity market. In the IoT business, the company had already acquired QNX back in 2010, which basically focuses on the automotive industry. Recently BlackBerry announced the opening of a new Center of Excellence, Engineering and Innovation Center in Hyderabad. This new facility will be BlackBerry’s second-largest world-class engineering centre outside of Canada and will host around 100 embedded software developers by the end of 2023. The Flagship: BlackBerry’s IoT BlackBerry’s IoT business unit consists of BlackBerry QNX, BlackBerry Certicom, BlackBerry Radar, BlackBerry Jarvis, and BlackBerry IVY. However, the most focused units are BlackBerry IVY and BlackBerry QNX. BlackBerry IVY is a cloud-connected software that allows car manufacturers to safely access a vehicle’s sensor data and use machine learning to create insights. IVY has a simple, safety-compliant API mechanism, which allows developers to build applications that enhance the driving experience. Meanwhile, BlackBerry QNX is a supplier of operating systems, development tools, and support for critical embedded systems. QNX is used in various industries, including aerospace, defence, medical, and robotics, to launch safe, secure, and reliable systems more efficiently. The company claims that its IoT business’ customers include 17 of the G20 governments, 45 of the Fortune 100, and 77% of Fortune 500 financial services companies. Additionally, 45+ OEMs, including Audi, BMW, Ford, GM, Toyota, Volkswagen, and Volvo, have embedded BlackBerry’s technology in their vehicles. In India, Mahindra and TATA are said to be using Blackberry’s tech in their products. While QNX was a well-known brand in the auto industry even before Blackberry acquired it in 2010, it still has competitors like Linux and Green Hills softwares. On how Blackberry distinguishes itself from other competitors, John Wall – SVP (head of operations), Blackberry QNX, explained that there are two segments of competition for them: the traditional RTOS, such as Green Hills Software and VxWorks, and then there is Linux. “The traditional RTOS comes from small processors that do not scale well on new-generation hardware,” says Wall. He stated that these systems lack performance despite having safety and security features. “In contrast, Linux comes from the high-performance side but lacks functional safety credentials,” commented Wall. “QNX, on the other hand, came from a mixture of functional safety, high security, and high-performance side.” Thus, QNX has the advantage of being a high-performance computing platform like Linux but with the pedigree and safety requirements of traditional RTOSes, says Wall. Problems ahead for Blackberry While Blackberry may be doing well in its IoT business, the company might face difficulties in its enterprise security software business. Recently, Microsoft launched Security Copilot, an AI-powered security product that uses a combination of large language models and security-specific models to enhance the work of security analysts. The model will integrate with Microsoft Security products and will expand to third-party products. The news is worrying for Blackberry as Cylance was known for its AI capability that it brought to the table. However, with the launch of Microsoft’s possibly GPT4-powered platform for Cybersecurity, Cylance would find it tough to bring a differentiator. Cybersecurity rakes in major moolah for Blackberry, and since the company doesn’t have much of a differentiator in the field, it will have a hard time finding a piece of the pie in the cybersecurity solutions market, which is expected to grow to $366 billion by 2028.","excerpt":"The company, once synonymous with premium mobile handsets, has shifted focus to IoT and enterprise security, while distancing itself from the smartphone industry","categories":["AI Features"],"tags":["Cybersecurity","IoT"],"author_name":"Lokesh Choudhary","publish_date":"2023-04-04T14:00:00","publication_year":"2023","word_count":729,"keywords":["Go","API","machine learning","AI","innovation","GPT","Aim","llm_models:GPT","Cybersecurity","R","IoT","startup"],"extracted_tech_keywords":["AI","machine learning","Aim","R","Go","API","GPT","innovation","startup","llm_models:GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/blackberry-boys-are-back\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10115058,"title":"Snowflake and Mistral AI Partner to Bring LLMs to Data Cloud","content":"Snowflake, the data cloud company, and Mistral AI, a European-based AI startup, have announced a global partnership to provide Mistral AI’s most powerful language models directly to Snowflake customers in the Data Cloud. The collaboration, backed by a parallel investment in Mistral’s Series A from Snowflake Ventures, focuses on providing enterprises with seamless access to large language models while prioritising security, privacy, and governance over their data. I’m thrilled to announce that @SnowflakeDB is partnering with @MistralAI to put the most powerful LLMs in the hands of our customers, empowering every user to build cutting-edge, AI-powered apps with simplicity and scale.Mistral AI’s models are now accessible to customers in…— sridhar (@RamaswmySridhar) March 5, 2024 As part of the partnership, Snowflake customers can now access Mistral AI’s latest LLM, Mistral Large, via the Snowflake Data Cloud. Additionally, Snowflake customers gain access to Mixtral 8x7B, an open-source model exceeding OpenAI’s GPT3.5 in speed and quality on most benchmarks, and Mistral 7B, a foundation model optimised for low latency, low memory requirements, and high throughput for its size. “By partnering with Mistral AI, Snowflake is putting one of the most powerful LLMs on the market directly in the hands of our customers, empowering every user to build cutting-edge, AI-powered apps with simplicity and scale,” said Sridhar Ramaswamy, CEO of Snowflake. Snowflake Cortex LLM Functions, now in public preview, enables users to build generative AI apps quickly, easily, and securely. The Snowflake Cortex service, fully managed by Snowflake, supports industry-leading LLMs for specialised tasks like sentiment analysis, translation, and summarisation. Snowflake recently announced Sridhar Ramaswamy as the New Chief Executive Officer and member of the Board of Directors effective immediately. He will replace Frank Slootman, who decided to retire and step from the helm of the company. Ramaswamy, who graduated from IIT Madras, joined Snowflake in May 2023 after selling his startup Neeva AI to the data cloud company. He held the position of Senior Vice President of AI at Snowflake.","excerpt":"Snowflake Cortex LLM Functions, now in public preview, enables users to quickly, easily, and securely build generative AI apps","categories":["AI News"],"tags":["LLMs","mistral","Snowflake"],"author_name":"Siddharth Jindal","publish_date":"2024-03-06T09:02:12","publication_year":"2024","word_count":327,"keywords":["Go","startup","OpenAI","AI","LLMs","sentiment analysis","ML","GPT","mistral","generative AI","R","Snowflake"],"extracted_tech_keywords":["AI","ML","generative AI","OpenAI","sentiment analysis","Snowflake","R","Go","GPT","startup"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/snowflake-and-mistral-ai-partner-to-bring-llms-to-data-cloud\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10100902,"title":"Llama 3 Might Not be Open Source","content":"Meta AI, and everyone in its AI team, including Mark Zuckerberg and Yann LeCun, have been one of the biggest proponents of open source. It was indeed leading the open source LLM race all this while with LLaMA and then the pseudo open source, Llama 2. But since then, a lot of models have been coming up in the open source community that are outperforming Llama 2 on various benchmarks. Now, it might be time for tech-giant to realise how to finally use its AI technology to earn some bucks. Why should it compete in the open source market? Zuckerberg, in his latest podcast with Lex Fridman in the metaverse, said that Meta might have to reconsider if it is going to open source the next iteration of Llama, which is Llama 3. “Right now, the priority is building that into a bunch of consumer products,” said Zuckerberg. At the Meta Connect 2023, the team announced AI in a lot of consumer products such as AI chatbots and characters on WhatsApp and Instagram, based on celebrities, and a slew of other integrations. There is no doubt that open sourcing models and integrating them into its products has helped Meta move faster, but now there might be a chance of going the other way. Not surprisingly, @ClementDelangue gets it. https:\/\/t.co\/gOGWqC9KDZ— Yann LeCun (@ylecun) September 29, 2023 Meta learns how to use Llama Zuckerberg adds that Meta trained Llama 2 and released it as an open source, but it is not a consumer product, but just an AI infrastructure. Though Zuckerberg is all about open sourcing AI and loves what the community has been doing with Llama 2, when it comes to Llama 3, he said that the debate with Fridman was very helpful for open sourcing Llama 2, and the same would be needed for Llama 3. “We would need a process to red team this, and make it safe. My hope is that we would be able to open source the next version when it is ready to, but we are not close to doing that this month. It’s a thing that we are still early in work now,” said Zuckerberg. This time, Llama 3 might even be better than GPT-4. Thus, making Llama 3 safe sounds like a great bet, but it is somewhat in contrast to what Zuckerberg said in the last podcast with Fridman. “The stage we are in right now, the equities balance strongly in my view towards doing this more openly.” He believed that if companies think that they have gotten close to what they believe is “superintelligence”, it makes sense for them to discuss and think through a lot more. “We want a lot more researchers working on it and for a lot of these reasons open source has a lot of advantages.” He explained how open source software is more secure because there are a lot of people criticising it, finding holes in it, and thus in the end making it safe. “I think this is how we will achieve more aligned models, by making it open source and letting people work on it,” said Zuckerberg. There were still discussions if Llama 2 is in the works or not, but then it was actually open sourced by Meta. Yann LeCun also said that open sourcing is the best way to make progress in the LLM field, and thus promotes every company who goes this way. More high-performance open source LLMs coming out to enrich the AI open ecosystem.This one from @MistralAI https:\/\/t.co\/9Jg4RI8v3S— Yann LeCun (@ylecun) September 27, 2023 But, there is always a doubt about how to make money out of open source AI. For example, Mistral AI, a startup founded by former DeepMind and Meta employees, is also a great proponent of open source. Just this week, the team released an open source model. But in 2024, the team aims to raise more fundings in 2024, to build a consumer product, apart from the open source ones. Similarly, Meta’s competitors like Google or OpenAI have not released their models to the public and kept it closed. But earlier, when OpenAI released GPT-2, it was open source. The company decided to pivot to the closed source model approach after GPT-3. Probably because it realised that the revenue model with open source is not that fool proof. Stepping up the game With Zuckerberg’s latest comments, it seems like Meta might steer a little away from being the champion of open source and finally learn how to monetise it within its consumer products. Which is indeed fair. The company has been making strides in the metaverse and with its headsets, where it is implementing AI. Hot take: LLaMA-3 model will be on par or better than GPT-4— anton (@abacaj) September 29, 2023 Meta’s Llama 2 is arguably still “an order of magnitude behind what OpenAI and Google are doing,” said Zuckerberg. But still, many companies have been adopting the open source model and earning a few bucks for themselves. It would only make sense now for Meta to keep Llama 3 for itself, and use it for consumer products, at least for a little bit, just to catch up on the four years. Even if that means, ditching its true nature.","excerpt":"It’s time for Meta to keep Llama for itself and use it within its consumer products.","categories":["AI Features"],"tags":["lex fridman","LLaMA","Mark Zuckerberg","Meta","Meta AI","Open Source AI"],"author_name":"Mohit Pandey","publish_date":"2023-09-29T16:00:00","publication_year":"2023","word_count":876,"keywords":["Go","Meta","Meta AI","funding","OpenAI","AI","chatbots","LLaMA","Mark Zuckerberg","Open Source AI","Aim","GPT","lex fridman","R","startup"],"extracted_tech_keywords":["AI","OpenAI","Meta AI","Aim","chatbots","R","Go","GPT","startup","funding"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/llama-3-might-not-be-open-source\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":26413,"title":"How IBM’s AI-Based Cloud-Ready ‘POWER9’ Rivals With Intel’s ‘Xeon’ Line","content":"Image Source: IBM The battle of servers is one which will only intensify as companies wrestle with data-intensive workloads. IBM, which is spearheading the cognitive computing revolution, has now beefed-up its POWER9 processor with an enhanced core and chip architecture which provides superior thread performance and higher throughput. The POWER9 processor is designed to meet the demands of today’s large-scale workloads and also meet the diverse computing needs of the cognitive era. At a time when we are only hearing about Intel Xeon series dominating the datacentres, IBM has positioned its POWER9 servers for intensive AI-workloads with additional performance and price advantage. As server requirements undergo massive change with an increased emphasis on accelerated computing, IBM is trying to gain ground against the ubiquitous Xeon processors by creating an ecosystem of POWER partners. However, Intel’s heft in the data centre market is well-known. According to a Motley Fool report, the second-largest chipmaker makes its revenue from the data centre CPU market and the PC CPU market. The company controls around 99% of the data centre CPU market. Now, in the modern server businesses, semiconductor giants are racing to pack as many features as they can to scale both price and performance. Just like IBM packed more features in its POWER9 chips and dramatically scaled compute performance, Intel did the same with its Skylake Xeon SPs rollout. On the other hand, IBM seems to have scored over its old rival Intel by positioning POWER9 processor as AI and cloud-ready. Tech analysts are saying that hooking up the IBM processor with GPU will create a super impressive performance. How IBM Is Advancing Against Intel’s Xeon Family Of Processors IBM has strived hard to outperform Intel in the data centre market and the POWER9 processors are the most powerful competitors to Intel’s Xeon line. The highlight of IBM’s POWER processor release is a brand new core and chip architecture that the company has optimised for technical\/HPC workloads, hyperscale, analytics and machine learning applications. IBM also scored another win with Google in their corner. In 2016 OpenPower Summit, Google announced that the majority of its infrastructure is on Power servers. According to an HPC Wire report, Google is also working with RackSpace on a new POWER9 server known as Zaius. The new POWER9 processor provides both accelerated and heterogeneous compute solutions, Brian Thompto, senior technical staff member for POWER Processor Design at IBM was cited in HPC Wire report. The company’s beefed-up server portfolio has been touted as the Swiss Army Knife for data-intensive workloads as it supports advanced generations of new storage class memories A Next Platform report indicates that IBM can meet or beat Intel on memory bandwidth per socket. According to the Mellanox tech blog post authored by John Biebelhausen, the POWER9 servers from IBM displays advanced I\/O buses, inventing new CAPI I\/O buses IBM is very aggressive with POWER today and has demonstrated it’s the perfect workhorse for AI intensive workloads by publishing benchmark reports Intel’s Xeon Is At The Heart Of Data Centres Meanwhile, Intel’s architecture has been the industry standard in data centres and their Xeon processors provide a platform for accelerated performance of compute, handle more data and also support advanced analytics functions. Intel’s Xeon processors have been at the heart of data centres and are known for greater data-crunching performance and also reduce latency for cloud applications. Intel Xeon processors are known for their support for high-performance computing and delivering excellent performance for complex workloads. Earlier last year, the chipmaker launched the Xeon Scalable line which is based on its Skylake architecture in an attempt to corner a greater share of the data centre market. Intel’s Navin Shenoy, VP and general manager of the Data Centre group was cited as saying how the new line is a major advancement for the data centre world with the platform supporting a broader set of workloads, mainly self-driving and AI workloads. The new Xeon Scalable line promised a quantum leap in performance with an increased focus on AI-intensive workloads. The platform also builds on traditional performance drivers such as enhanced memory, more cores and also more I\/O. The chipmaker giant also integrated its Quick Assist technology to accelerate compute-intensive operations by reducing the size of data packets, a news report indicated. With new performance benchmarks, 16.5x over the previous generation of processors, Intel too is aggressively marketing its platform as AI ready. Fri, July 25, 2018  3:00 PM – 4:00 PM IST Visit our website to register","excerpt":"The battle of servers is one which will only intensify as companies wrestle with data-intensive workloads. IBM, which is spearheading the cognitive computing revolution, has now beefed-up its POWER9 processor with an enhanced core and chip architecture which provides superior thread performance and higher throughput. The POWER9 processor is designed to meet the demands of […]","categories":["Global Tech"],"tags":["IBM","Intel"],"author_name":"Richa Bhatia","publish_date":"2018-07-13T05:13:24","publication_year":"2018","word_count":748,"keywords":["Go","API","machine learning","programming_languages:R","AI","Scala","RAG","analytics","IBM","programming_languages:Scala","R","Intel"],"extracted_tech_keywords":["AI","machine learning","analytics","RAG","R","Go","Scala","API","programming_languages:R","programming_languages:Scala"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/how-ibms-ai-based-cloud-ready-power9-rivals-with-intels-xeon-line\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10134325,"title":"Tech Veteran Jaspreet Bindra launches ‘AI&amp;Beyond’ to Democratise AI Literacy","content":"Jaspreet Bindra, a tech evangelist and author, launched his AI literacy initiative “AI&Beyond” with the goal to make AI more accessible and understandable. It is in collaboration with Anuj Magazine, an AI & cybersecurity expert, it is spread across industries. “At AI&Beyond, we believe that AI is no longer the future – it is very much the present. Our goal is to ensure that AI literacy becomes as fundamental as reading and arithmetic, especially in large organisations where AI’s impact will be profound. Through AI&Beyond, we aim to bridge the gap between AI’s capabilities and its practical application across various sectors,” said Jaspreet Bindra. Inspired by CLR James’ book Beyond A Boundary, AI&Beyond aims to simplify and distill the concept of AI beyond the technical aspects of it. One of the initiative’s flagship programme, The Generative AI Bootcamp,  is designed to equip participants with the necessary and most important AI skills. Additionally, the platform’s Ethics Bootcamp is crucial in today’s AI-driven world, helping organisations not only adopt AI, but do so with a strong ethical foundation. The underlying idea is to make the learning experiential and immersive – that form the foundation of future AI businesses. These resources – including workshops, briefings, webinars, and consulting services – would empower organisations to become more agile, innovative and competitive in the digital age. Jaspreet Bindra on GenAI Bindra has been a staunch advocate of Make in India, and to that end, proposed that India should consider building generative AI as a digital public good – known as JanAI or GenAI for the people. “If you look at ChatGPT or Bard, they are all trained on the internet, where almost 80 to 90% of the data is English, and West-oriented. It doesn’t have vernacular data nor an Indian context,” said Bindra, while emphasising on why in the current LLM market, he wanted to bring an Indian context to the training model. Also, at MLDS 2024 – India’s biggest generative AI summit – hosted by AIM, Jasprit Bindra spoke about the history of AI and the concept of singularity in today’s generative trends.","excerpt":"This initiative aims to empower businesses and individuals to keep up with the AI revolution.","categories":["AI News"],"tags":["AI (Artificial Intelligence)"],"author_name":"Aditi Suresh","publish_date":"2024-09-03T16:09:09","publication_year":"2024","word_count":349,"keywords":["Go","ChatGPT","GenAI","AI","ML","Git","GPT","Aim","generative AI","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","ML","generative AI","GenAI","ChatGPT","Aim","R","Go","Git","GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/tech-veteran-jaspreet-bindra-launches-aibeyond-to-democratise-ai-literacy\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10004272,"title":"Government May Soon Launch Data Centre Policy With Favourable Guidelines","content":"The news doing around lately is that the Ministry of Electronics and Information Technology (MeitY) may come up with a national policy framework for establishing data centres across the country soon. The reports suggest that states will be free to either follow in total or mould it according to their needs. The new data centre policy framework may include the following instructions: Push for the usage of renewable energy such as solar energy for day-to-day operations of the building which will house the data disksState governments may be pushed to start giving cross-subsidy to data centres which currently get electricity at either very high rates or have to pay cross-subsidy surcharge on power which they buy at low rates but from a state different from where the data centre is locatedIt is also looking to suggest data centres to be a part of urban or semi-urban infrastructure so that real estate players can consider investments in the sectorThe current regulations require about 40 clearances to establish a data centre which is extremely time-consuming. It is looking to make it a single-window clearance for applicants Officials believe that once the power cost is managed, it will reduce the cost of running and managing data centres. Eventually, more players will be attracted to the system. Further making single-window clearance will give the establishment of data centres a major boost. Other measures of cost-cutting may include relief in the form of lower levies to electronic components used for data centres or keep suitable guidelines to favour domestic ancillary industries. MeitY had earlier come up with guidelines for technical and financial support for the establishment of state data centres in which it had suggested that since the process was a complex one which required substantial investment as well as efficient management, only existing internet data centre players should be allowed. It also suggested that the central government would support the entire cost of establishment.","excerpt":"The news doing around lately is that the Ministry of Electronics and Information Technology (MeitY) may come up with a national policy framework for establishing data centres across the country soon. The reports suggest that states will be free to either follow in total or mould it according to their needs.  The new data centre […]","categories":["AI News"],"tags":[],"author_name":"Srishti Deoras","publish_date":"2020-08-05T11:29:33","publication_year":"2020","word_count":320,"keywords":["programming_languages:Go","Go","R","programming_languages:R"],"extracted_tech_keywords":["R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/government-may-soon-launch-data-centre-policy-with-favourable-guidelines\/","complexity_score":2,"technical_depth":4,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10018649,"title":"Being Future Ready Is The Only Way To Survive In Data Science Field: Dr Angshuman Ghosh, Sony Research India","content":"To say there is a massive buzz around data science would be captioning the obvious. The overwhelming amount of resources and information resulting from the rapid evolution of the field could leave an aspirant at sixes and seven. Analytics India Magazine caught up with Dr Angshuman Ghosh, Head of Data Science at Sony Research India, to understand his thought processes and gain insights into the inner workings of the emerging field of data science. He has over 12 years of experience in Technology, Media, and Retail companies, such as Disney, Target, Grab, and Wipro. He is an Official Member and Contributor at the Forbes Technology Council and a visiting professor at top IITs, IIMs, and XLRI. AIM: Could you talk about Sony Research India and your role there? Dr Ghosh: Sony Research India (SRI) is the research and development arm of Sony Global in India. SRI was established in 2020 and part of the global R&D of Sony. With SRI, Sony seeks to accelerate research, collaboration, and innovation capabilities in India. Sony Research decided to venture into India due to its growing economic potential and the high availability of skilled professionals. I joined SRI in October 2020 and was the first recruit for this unit. I lead the overall data science department in SRI. We are currently focusing on Sony’s India entertainment business, including its TV channels and the OTT platform. We are pumped about the OTT channel SonyLiv. Being part of the leadership team, I am also quite involved in the talent acquisition and team-building process and hope to gradually build a strong team of professionals. AIM: What are the tools and models used at SRI? Dr Ghosh: We are quite a generalist in our approach when it comes to the tech stack. We are currently working with Python for programming. We are leveraging AWS for our cloud infrastructure needs. We are also open to using several machine learning and deep learning methods, which will shape up in a much bigger way in the coming months. AIM: What are the skills you look for while recruiting? Dr Ghosh: I believe though the technology keeps evolving, few basics remain unchanged. Same with data science, for beginners, we expect them to have sound knowledge of statistics, probability, linear algebra, machine learning, etc. Python and SQL skills are also must-haves. However, we are flexible on the domain knowledge part, which we expect the candidate to pick up on the job. Apart from these technical skills, we also give great preference to an aspirant’s attitude and general professional approach. AIM: What motivated you to get into data science? Dr Ghosh: To be honest, there wasn’t a definitive point when I decided to become a data scientist. I pursued computer science engineering and graduated in 2005. Soon after, I started working with Wipro, where my job primarily involved programming and software development for some of the Fortune 100 clients. Though I was technically sound, I realised the importance of understanding the business aspect of it. My next natural step was to pursue an MBA, which I did from XLRI Jamshedpur, with specialisation in marketing and finance. Given my interest in research, I also decided to complete my PhD. During this time, I got a lot of background knowledge on statistics, regression, R and other tools. My PhD work was around deriving insights on user behaviour from what they post on social media. In 2015, I joined Disney Star TV. My work was a mix of analytics and research. That was also the time I started shifting to data science. I realised that my background aligned quite well with the requirements of the data science field. My shift to core data science was when I joined Target, a hotbed of data science developments. AIM: What is your advice to students pursuing a career in Data Science? Dr Ghosh: We can talk about different tools and methods for Data Science, but at its core, Data Science refers to using data, scientifically, to make decisions or make an impact. There are three key skills for any data scientist– a stronghold on mathematics and statistics. Secondly, you need a programming language base for different tasks such as data processing, storage, etc. Lastly, domain knowledge. When you are working in a company, you must think about what value you are adding. Having acquired these skills next comes constant upgradation and upskilling. There is a sea of resources available online. For example, Coursera and EDx are good sources for theoretical introductions to a variety of topics. For a more practical approach, aspirants may check Datacamp and Udemy. I would also suggest using Kaggle, participating in hackathons, and undertaking internships to gain an edge. It is also important to think from the perspective of being ready for future challenges, given this field’s dynamic nature. It does get difficult to catch up with every new model or concept. I find it difficult too. What I tend to do is I try to look at the bigger picture, and once a tech starts picking pace, I spend time understanding it. The secret lies in following a broad macro trend, not just in DS but in complete tech space. AIM: You offer a lot of mentoring through your social media posts. Do you think we should have more mentor-mentee approaches for the larger benefit of the ecosystem? Dr Ghosh: It is my way of giving back. When I started, such blogs and resource material helped me a lot. I also do a bit of analysis to understand what kind of topics students might be interested in and try to give guidance. Speaking of having such a mentor-mentee ecosystem, yes, it is ideal. Python grew so big because it has an extensive network of contributors and collaborators. That said, being a mentee is a personal choice. It can’t be forced on someone.","excerpt":"To say there is a massive buzz around data science would be captioning the obvious. The overwhelming amount of resources and information resulting from the rapid evolution of the field could leave an aspirant at sixes and seven. Analytics India Magazine caught up with Dr Angshuman Ghosh, Head of Data Science at Sony Research India, […]","categories":["AI Features"],"tags":["Data Science","Interviews and Discussions"],"author_name":"Shraddha Goled","publish_date":"2021-01-21T18:00:00","publication_year":"2021","word_count":977,"keywords":["data science","machine learning","AWS","AI","RAG","Python","Aim","deep learning","analytics","Data Science","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","machine learning","deep learning","data science","analytics","Aim","RAG","AWS","Python","R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/we-are-pumped-about-our-ott-platform-sonyliv-head-of-data-science-sony-research-india\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10126168,"title":"‘Fears of an AI takeover are unfounded due to data limits and finite growth,’ says Aidan Gomez","content":"In a recent interview, Aidan Gomez, CEO and co-founder of Cohere, stated fears of an AI takeover are unfounded due to its reliance on human training data and the limits of exponential growth. He explained, “I think I’m empathetic to the fears, you know, the sci-fi narrative of computers or AI taking over and destroying the world. It’s been going on for decades, and so it’s really deeply embedded within our culture. It gets lots of clicks, headlines, it gets attention. It’s shocking. I understand why people are scared of it and why some say it to get attention.” Gomez highlighted that this is not a technical truth of the technology and that continuous exponential scaling does not happen. There are friction points and complexities. He also mentioned that the intelligence of these models is limited by the humans who create them, as it is our data and knowledge that teach them. However, he believes real risks lie in deploying AI in high-stakes scenarios like medicine and advocates for scrutiny and tough discussions on AI deployment, rather than sensational sci-fi narratives. What is Cohere Upto In the interview, he also discussed how the original goal of the project was to improve Google Translate, a very well-known problem. He noted that it has been extraordinary to see the broad impact of a technology developed to enhance translation. Gomez was also the co-author of the original Transformers paper which forms the crux of today’s generative AI products. Recently, AIM spoke to Saurabh Baji, SVP of Engineering at Cohere, about the mixed emotions in Silicon Valley regarding achieving AGI, as seen in the recent banter between Meta’s Yann LeCun and xAI’s Elon Musk. “We remain concentrated on designing AI solutions that deliver better workforce and customer experiences for businesses today rather than pursuing abstract concepts like AGI,” said Baji.","excerpt":"The intelligence of these models is limited by the humans who create them.","categories":["AI News"],"tags":["Cohere","Google","OpenAI"],"author_name":"Tarunya S","publish_date":"2024-07-08T13:08:00","publication_year":"2024","word_count":306,"keywords":["Go","OpenAI","AI","programming_languages:R","R","ai_frameworks:Transformers","Transformers","XAI","Aim","generative AI","Google","xAI","Cohere"],"extracted_tech_keywords":["AI","generative AI","xAI","Aim","Transformers","R","Go","XAI","ai_frameworks:Transformers","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/fears-of-an-ai-takeover-are-unfounded-due-to-data-limits-and-finite-growth-says-aidan-gomez\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10171479,"title":"Snowflake and Databricks Take Their Rivalry to PostgreSQL","content":"The rivalry between Snowflake and Databricks has expanded beyond data lakes and warehouses, moving into the growing world of PostgreSQL. The battle has intensified with strategic acquisitions on both sides. Snowflake recently acquired Crunchy Data for $250 million, while Databricks snapped up Neon for $1 billion. Announcing the acquisition in a blog post, Snowflake called Postgres a top choice for developers due to its flexibility, cost efficiency, and native AI features, such as vector support (pg vector). Its open-source nature and vibrant ecosystem only add to its appeal. “We’re tackling a massive $350 billion market opportunity and a real need for our customers to bring Postgres to the Snowflake AI Data Cloud,” said Vivek Raghunathan, Snowflake’s SVP of engineering. Why are the Giants Betting on PostgreSQL? “PostgreSQL’s ecosystem and extensions are growing fast. More people now know this database better than any other. pgvector gave it a big push,” said Arpit Bhayani, creator of DiceDB, when AIM asked why PostgreSQL is becoming the preferred database for AI-native applications. Snowflake Postgres builds on Snowflake’s earlier move into transactional data with Unistore, which combines transactional and analytical workloads in one system. Building on native PostgreSQL support, Snowflake Postgres extends that vision, offering enterprises a production-ready solution for transactional applications that require Postgres compatibility. PostgreSQL, the open-source relational database, has surged in popularity, surpassing MySQL as the most favoured database among developers, according to Stack Overflow’s 2023 and 2024 Developer Surveys. Its ability to handle geospatial data (via PostGIS), time series data (via TimescaleDB), JSON, and vector embeddings (via pgvector) makes it ideal for AI applications. Commenting on the acquisitions in a LinkedIn post, senior data engineer Avinash S said it’s more than just adding another database, calling it a strategic bet on PostgreSQL as the backbone of the AI-native era, especially in its serverless and cloud-native form. “Imagine AI agents spinning up databases for every real-time task or experiment, then discarding them. Traditional databases can’t handle this “disposable” scale. Serverless Postgres delivers the rapid provisioning, elasticity, and cost-efficiency that AI agents desperately need to work autonomously and at speed,” he added. “It’s not just agentic. Because many people are talking about it and using it, it has become the de facto standard,” said Bhayani. However, he added that AI agents spinning databases seems like a strong assumption. Similarly, Factorial Advisors said in a blog post that the Neon acquisition fits into Databricks’ broader push to build a unified data intelligence platform. “With over $19 billion in financing and a recent $62 billion valuation, Databricks has the financial muscle to keep acquiring,” they wrote. “Following previous buys like Tabular ($2 billion) and MosaicML ($1.3 billion), Neon helps address the growing demand for databases that operate at ‘agentic speed’ while staying cost-effective through pay-as-you-go economics.” With these acquisitions, both Snowflake and Databricks are going to challenge hyperscalers like AWS, Microsoft Azure, and Google Cloud, which offer managed PostgreSQL services tightly integrated with their AI stacks. Neon vs Crunchy Data Founded in 2012, Crunchy Data provides a full-service, production-ready PostgreSQL solution that includes backups, high availability, disaster recovery, connection scaling, and monitoring. It supports mission-critical deployments across cloud, on-premises, and hybrid environments. Snowflake said its new offering, Snowflake Postgres, will bring transactional Postgres data into its platform, accelerating innovation and giving developers greater agility, visibility, and control to build trustworthy AI agents and applications faster. Crunchy’s expertise means Postgres-powered apps can run natively on Snowflake without having to rewrite code. Developers benefit from built-in connection pooling, performance metrics, and logging, making it easier to build and manage scalable apps. On the other side, Databricks CEO Ali Ghodsi pointed out that frontier LLMs have been trained on vast amounts of public data from the Postgres open-source ecosystem, making AI agents naturally proficient in using Neon, which is built on Postgres. Ghodsi added that Databricks and Neon share a common foundation in technical infrastructure and a deep belief in open source. Databricks started the Apache Spark project at UC Berkeley, also the birthplace of Postgres, the open-source database on which Neon is built. He added that OLTP databases, a $100 billion market, are still dominated by decades-old products. With Neon, Databricks aims to help disrupt this space by creating the most developer- and AI agent–friendly database platform. When Neon became generally available last year, around 30% of the databases on the platform were being created by AI agents rather than humans. But in a recent update, that number has jumped to over 80%. In other words, AI agents are now creating more than four times as many databases as human users. Everyone’s rushing to PostgreSQL, and Snowflake and Databricks are jumping in by picking up niche providers. It’s not just about more databases, it’s about getting ready for AI, live data, and big business needs. Moreover, these acquisitions reflect a broader consolidation trend in the data and AI infrastructure market. Recent deals such as Salesforce’s $8 billion acquisition of Informatica, ServiceNow’s purchase of Data.World, and Alation’s acquisition of Numbers Station, show how companies are racing to build comprehensive AI-ready platforms. According to Bhayani, much of this is driven by the push to acquire customers and specialised expertise.","excerpt":"PostgreSQL, the open-source relational database, has surged in popularity, surpassing MySQL as the most favoured database among developers.","categories":["Global Tech"],"tags":["Databricks","Snowflake"],"author_name":"Siddharth Jindal","publish_date":"2025-06-09T15:09:37","publication_year":"2025","word_count":860,"keywords":["PostgreSQL","AWS","AI","ML","Apache Spark","serverless","RAG","Aim","Databricks","Azure","Snowflake"],"extracted_tech_keywords":["AI","ML","Aim","RAG","AWS","Azure","serverless","Apache Spark","PostgreSQL","Snowflake"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/snowflake-and-databricks-take-their-rivalry-to-postgresql\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10056849,"title":"Clearview AI Is On The Verge Of A Patent Despite Intense Controversies","content":"Clearview AI has been provided with the go-ahead signal for securing a federal patent for its facial recognition software, as reported by Politico. The US Patents and Trademarks Office sent a “notice of allowance” to Clearview on December 1, implying that the patent will be awarded once the organization pays the necessary administrative fees. The patent will, among other things, cover its method of obtaining information, specifically, its “automated web crawler” tool that scans social networking sites and the internet, the same process that has led the company to face extensive backlash from numerous privacy advocates. “There are other facial recognition patents out there — that are methods of doing it — but this is the first one around the use of large-scale internet data,” Ton-That, the founder of Clearview AI, told Politico in an interview. Clearview AI has more than 10 billion photos in their database, which was confirmed by Ton-That, much more than the previous estimate of 3 billion images. At this moment, the UK, Australia, and France have issued official notices to Clearview to stop gathering any information on their citizens and remove all existing data. How did it all start? Clearview AI was founded by Hoan Ton-That, an Australian entrepreneur, and Richard Schwartz, who had worked previously with New York City Mayors such as Rudy Giuliani. Ton-That had first arrived in San Francisco to join the growing market for developing social media applications in 2007. In 2009, he created a site that allowed people to share videos with their contacts using instant messenger. However, he shut down this website once it was labelled a “phishing scam.” He then developed an app named “Trump Hair” in 2015, which allowed people to apply the iconic hair of Donald Trump to their pictures. Neither of his ventures became a big hit, as he envisioned. Ton-That met Schwartz in 2016 at a book event held in the Manhattan Institute. They soon decided to enter the facial recognition business together, where Ton-That would build the application, and Schwartz would use his political influence to attract commercial interest. In 2016, Hoan hired two engineers. One was tasked to develop a tool to collect pictures of people’s faces from the internet and different social media sites. Facebook, Twitter, YouTube, and LinkedIn have all sent cease and desist letters to Clearview AI in 2020. The other engineer perfected the facial recognition algorithm. Ton-That described the final system developed as a “state-of-the-art neural net.” It has an accuracy of 98.6% and provides results in 30-60% of cases. Schwartz paid for server fees and other expenses during this time and used his contacts to drum up interest among government agencies. The company first started near the end of 2017 under the name of Smartcheckr but soon changed it to Clearview. The Impact of Clearview AI Based on a leaked list accessed by Buzzfeed, more than 1,800 law enforcement agencies have used Clearview’s facial recognition technology. According to the company and government officials, this includes the FBI, Homeland Security, and Canadian law enforcement authorities. Several sources reported in February 2020 that Clearview AI faced a data breach revealing its customers. Clearview had previously stated it only allowed access to law enforcement agencies, but the data breach revealed that the startup had sold its software to specific private organizations. The company also raised 30 million dollars in its Series B funding in July 2021 despite being the target of multiple lawsuits and a joint investigation by Britain and Australia. Approval of the patent for their facial recognition software will make them even more defiant to external pressures and continue aggressively scraping pictures to increase their database. The COVID-19 pandemic caused people to volunteer to provide their data for public use for the greater good, but companies like Clearview AI can quickly commercialize these. The Future The granting of the patent by the US Patents and Trademarks Office will place Clearview AI in an ideal position to become the primary facial recognition software to be used by all US law enforcement authorities. The cease and desist letter by big tech companies to prevent Clearview from scraping images from their sites had no effect, and neither is it likely to have in the future. So, in short, it looks increasingly likely that Clearview AI is here to stay, and it is imperative to make anyone who uses the internet aware of this facial recognition software. It is also setting a dangerous precedent for all future AIs.","excerpt":"Scraping data from the internet without consent, a dangerous illegitimate method, may soon become mainstream with the patent of Clearview AI.","categories":["AI Features"],"tags":["clearview AI","data scraping","Facial Recognition","Machine Learning Latest"],"author_name":"Arnab Ray","publish_date":"2021-12-23T14:00:00","publication_year":"2021","word_count":746,"keywords":["Go","API","funding","Facial Recognition","AWS","AI","data scraping","cloud_platforms:AWS","programming_languages:R","Machine Learning Latest","clearview AI","GAN","R","startup"],"extracted_tech_keywords":["AI","AWS","R","Go","API","GAN","startup","funding","cloud_platforms:AWS","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/clearview-ai-is-on-the-verge-of-a-patent-despite-intense-controversies\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10044041,"title":"8 Projects To Kickstart Your MLOps Journey In 2021","content":"MLOps follows a set of practices to deploy and maintain machine learning models in production efficiently and reliably. While the data science team has a deep understanding of the data, the operations team holds the business acumen. MLOps combines the expertise of each team, leveraging both data and operations skill sets to enhance ML efficiency. According to the Algorithmia report, nearly 22 percent of companies have had ML models in production for one to two years. With practice, MLOps professionals can enhance their skills, and develop a solid pipeline for developing machine learning models. In this article, we have shown projects across tools and services that will help you kickstart your MLOps journey on the go. Made With ML Developed by Goku Mohandas, ‘Made With ML’ is a project-based course on machine learning and MLOps fundamentals, focusing on intuition and application that teaches you how to apply machine learning across industries. Check out how to apply ML to build a product here. The source code is available on GitHub. BudgetML According to the developer (Hamza Tahir), it is supposed to be fast, easy and developer-friendly. However, it is by no means meant to be used in a full-fledged production-ready setup. Instead, it is simply a means to get a server up and running as fast as possible with the lowest costs possible. In this project, you will learn how to deploy an ML inference service on a budget in less than ten lines of code. It is perfect for practitioners who want to deploy their models to an endpoint faster and not waste a lot of time, money, and effort trying to figure out end-to-end deployment. The source code, alongside key features of this project, is available on GitHub. Great Expectations Great Expectations helps data teams eliminate pipeline debt through data testing, documentation, and profiling. It is a flexible, declarative syntax for describing the expected shape of data. When used in exploration and development, Great Expectations provides an excellent medium for communication, surfacing and documenting latent knowledge about the shape, format, and content of data. In production, it is a powerful tool for testing. Check out the GitHub repository here. Lime: Local Interpretable Model-Agnostic Explanations Lime supports explaining individual predictions for text classifiers or classifiers that act on tables (NumPy arrays of numerical or categorical data) or images. It is based on the work presented in the ‘Why Should I Trust You?: Explaining the Predictions of Any Classifier‘ paper. Check out the GitHub repository here. Automating the Archetypal Machine Learning Workflow and Model Deployment ML automation workflow contains a Python-based machine learning project to demonstrate the archetypal ML workflow within a Jupyter notebook, alongside some proof-of-concept ideas on automating key steps, using the Titanic binary classification dataset hosted on Kaggle. The secondary aim of this ‘project’ is to show how the deployment of the model generated as a ‘build artefact’ of the modelling notebook can be automatically deployed as a managed RESTful prediction service on Kubernetes without having to write any custom code. Check out the GitHub repository of this project here. End-to-End ML Project: CookieCutter It is a generic template for building end-to-end machine learning projects. It offers a logical, reasonably standardised, but flexible project structure for doing and sharing machine learning work. The source code is available on GitHub. TFX-Addons It is a collection of community projects to build new components, examples, libraries, and tools for TFX (TensorFlow Extended). The projects are organised under a special interest group, called SIG TFX-Addons. The group focuses on: Driving the development of high-quality ‘custom pipeline components,’ including container-based components, Python function-based components, etc. Shaping a standardised set of descriptive metadata for community-contributed components to enable easy understanding, comparison, and sharing of components during discovery. Enabling the development of templates, libraries, visualisations, and other useful additions to TFX. Check out TFX-Addons projects here. Amazon SageMaker Examples Amazon SageMaker Examples demonstrate how to build, train, and deploy machine learning models using Amazon SageMaker. ‘Amazon SageMaker‘ is a fully managed service for data science and ML workflow. In this project, you will learn to quickly set up and run notebooks. The project comprises: Introduction to ground truth labeling jobsIntroduction to applying machine learning SageMaker automatic model tuning (XGBoost Tuning, TensorFlow Tuning, MXNet Tuning, etc.) Introduction to Amazon algorithms (k-means, Factorization Machines, Latent Dirichlet Allocation (LDA), Linear Learner, Image Classification, etc.) Amazon SageMaker reinforcement learning (RL) (Cartpole using Coach, AWS DeepRacer, Knapsack Problem, etc.) Scientific details of algorithms (Streaming Median, LDA, Linear Learner features)Amazon SageMaker debugger Advanced Amazon SageMaker functionality (data distribution types, encrypting your data, connecting to redshift, etc.)Amazon SageMaker Neo compilation jobs (GluonCV SSD Mobilenet, MNIST with MXNet, etc.) Amazon SageMaker Processing (Scikit-Learn data processing and model evaluation, feature transformation with Amazon SageMaker Processing and SparkML, etc.) Amazon SageMaker pre-built framework containers and the Python SDKUsing Amazon SageMaker with Apache Spark AWS Marketplace (create algorithm\/model package for listing in Marketplace for machine learning) Check out more MLOps open source projects here.","excerpt":"According to the Algorithmia report, nearly 22 percent of companies have had ML models in production for one to two years.","categories":["AI Trends"],"tags":["data science project marketing","Machine Learning","Machine Learning Latest","Machine Learning New","open source data science projects"],"author_name":"Amit Naik","publish_date":"2021-07-22T16:00:00","publication_year":"2021","word_count":828,"keywords":["data science","scikit-learn","Amazon SageMaker","open source data science projects","Machine Learning New","machine learning","AI","ML","MLOps","Machine Learning Latest","Machine Learning","Ray","Aim","TensorFlow","data science project marketing"],"extracted_tech_keywords":["AI","machine learning","ML","data science","MLOps","Aim","Amazon SageMaker","Ray","TensorFlow","scikit-learn"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/8-projects-to-kickstart-your-mlops-journey-in-2021\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10073496,"title":"The Bengaluru-based Startup that Helps You Find Your Next Store Location","content":"In present times, we use navigation apps like Google Maps and MapMyIndia to explore new areas. From a consumer’s perspective, they are indeed good and provide all relevant information on traffic, restaurants, clinics, etc at a given location. But is there a Google Maps-equivalent for a businessman wanting to set up a store in a new location? Who does the recce for him? Who brings in intel? GeoIQ does that and much more. A locational-intelligence platform based in Bengaluru, GeoIQ has achieved 10x growth in its ARR over the past four quarters and is on its way to registering additional 5x growth in ARR in the next two quarters. In an exclusive interaction with Analytics India Magazine, CEO Devashish Fuloria speaks about GeoIQ’s business models, flagship products, the challenges that the firm faces and also the future plans. AIM: What is the story behind GeoIQ? Devashish: Around 2018, we were discussing a lot of problems around data science. One of the biggest retailers in India was trying to implement their plan of opening a chain of 5,000 grocery stores across India, but faced problems in identifying the most suitable locations for their stores beyond metro cities. Generally, entities depend on information coming ground up in such cases. However, it becomes tricky to understand which information is good enough because entities generally do not know the whereabouts of a location. Now, when you set up the store, it’s a huge expense. So, if it doesn’t work after one year, you’re losing a lot of business value and capex. Hence, we came up with the idea of providing location-related information in a centralised manner. We started checking out the data for various locations. We looked at government data, public records and the internet and eventually layered all the information on the map. Today we tell businesses everything about an address on a street without actually being there – whether a location is reachable, if it is risky, would there be demand for an expensive apparel, etc. We provide answers to such high-value queries in simple numbers. When a client enquires about a location, we give them a score between 0-100, which signifies how good that location would be for the business. Devashish Fuloria AIM: That’s interesting! Please elaborate on some of the latest use cases provided to your customers, particularly in deploying AI\/ML? Devashish: Our use cases are across industries, the largest being in fintech, retail and e-commerce. The biggest use case for fintech is risk prediction. Before disbursing loans, fintechs need to know about the credit-worthiness of their customers. When applying for loans, customers provide information like PAN card details, credit card numbers etc. Based on location-specific information derived from these details, we predict how credible a potential customer would be. For e-commerce, we solve the concern related to return-to-origin by augmenting their user data with hyperlocal intelligence. Other use cases include accurate affluence prediction, fraud prediction, claims propensity, collections model, and lead prioritisation. AIM: What are the challenges you have faced so far and how is GeoIQ addressing them? Devashish: When we started out, we realised there was a strong need, yet a limited understanding of this data. We had to educate our users, unsuccessfully at times, on how best to use this vast data repository. Business analysts always looked for specific numbers that they thought were important. This was eventually addressed. Now we provide clients with scores. Then there were challenges like sourcing valuable, accurate, and quality real-world data. For example, good-quality data for Indian boundaries were not easily available, nor accessible. Also, the pin code, city, state, and other boundaries were not very clear beyond administrative limits. Therefore, we built a GeoAllocation engine to define the boundaries better (20% more accurate than existing boundary definitions) and mapped the addresses on these new boundaries. Another challenge was with respect to the sanctity of data. We had to check and validate data points from various sources to ensure the accuracy and truthfulness of information that we had gathered from public data sources. At present, data discovery is a big challenge. Identifying which data is useful and impacts a use case directly is a herculean task and often lands on trial and error. The NoCode ML platform has been created to solve this problem. The solution has completed the beta phase and is soon to be launched. Businesses would be able to create multiple models in no time and experiment with data attributes to identify the best fit for their use case. AIM: So, in a way, GeoIQ is trying to address the problem of bad data with the NoCode ML platform. Bad data is a global problem. How else do you plan to tackle the menace? Devashish: Data is good. We do not believe in bad data. Bad data is good data not presented in a structured and consumable format. Yes, we don’t deny that there are many problems with data when you look at companies’ databases. But you can’t solve multiple problems. Thus, we are focusing on one specific area, i.e. location, that helps us bind a lot of different datasets together in one cohesive way. AIM: What technology does GeoIQ use to analyse a broad data stack to provide customised solutions to its clients? Devashish: We use our proprietary algorithms to transform real-world data from 600+ sources in a structured format, categorised under 3000+ attributes. We have built advanced machine learning capabilities that help us provide custom models to the clients, based on their use cases. The explorer and APIs, our standard offerings, are backed by ML capabilities. Explorer helps a user to get information about a specific location or compare up to three locations at a time for a specific attribute. APIs could be directly embedded in the code to augment it for location data. AIM: GeoIQ raised USD 2.25mn in funding recently. How do you plan to channel these funds? Devashish:  As a product-led company, our core strategies are around product development. We are channelling a part of these funds towards building advanced tech capabilities and a team of skilled human resources. Moreover, we want to take our methodology and build a similar system for different geography. As of now, it’s the US. We are opening our platform to data scientists, where they can create their own location models and get specific location answers related to the US. AIM: Why US, and not the larger region around India? Devashish: Southeast Asia has very similar problems to India, which is basically that we are all data poor. Southeast Asia, for us, maybe twenty geographic entities, which means you have to find 20×500 sources of data to build up a base. But US being one massive geographic unit, you can scale up the data very quickly, and also, the data is easily available. AIM: At present, GeoIQ offers three products – No code ML, API and Explorer. Are there plans to expand the product range? What is in the pipeline? Devashish: More than expanding into new products, we are trying to expand our solutions for new use cases, source more latest data, and improve ML models. Every new data source that we add enriches our offerings a step further. Our solutions are seeing wide adoption. New companies with very specific and niche problem statements are reaching out to us. We are excited about those and are expanding our use case scenarios. AIM: In India, small businesses often suffer due to location issues. From the CSR perspective, do you have plans to engage with that section of society? Devashish: We already have a freemium model for small businesses to access our solutions with a certain number of free credits. Also, some basic information about towns and villages and streets is being provided for free. AIM: Do you think discussions around data regulation would impact GeoIQ’s operations? Devashish: I think it was a very conscious choice we made four years ago that we’re not going to get into personal data. It is anonymised information that we deal with. Government data doesn’t name people. Information like Nielsen’s data talks about spent patterns at a location and doesn’t have any personal information. We don’t know who Person A is. That information doesn’t exist in the system. If anything, GeoIQ’s methodology sets a template for how data should be used across multiple systems while remaining ‘privacy first’. Therefore, there will be no adverse impact of the changing data regulations on GeoIQ’s operations. AIM: What is the roadmap ahead for GeoIQ? What development can we expect in terms of new solutions and services? Devashish: Modern AI systems that exist within companies are based on what users are doing with their apps. Nobody knows what the users are doing in the real world. For example, Person A is just an entity for Amazon, and all personalisation will happen based on A’s movement and behaviour on the app. It’s likely that A’s neighbour has similar preferences as A, but Amazon doesn’t know about them. We are seeking to tap these sorts of common interactions and make them part of the modern AI systems. From a roadmap perspective, we are first targeting high-volume transactions, thus engaging a lot with fintechs, insurance, and e-commerce players. We are gradually increasing our verticals in India. In terms of new solutions or services, we are trying to strengthen our foothold in the insurtech and e-commerce sectors by addressing more use-cases and at a larger scale.","excerpt":"We tell businesses everything about an address on a street without actually being there: GeoIQ CEO Devashish Fuloria","categories":["AI Startups"],"tags":["AI Startups"],"author_name":"Zinnia Banerjee","publish_date":"2022-08-24T17:00:00","publication_year":"2022","word_count":1574,"keywords":["data science","Go","API","machine learning","funding","AI","ML","Aim","analytics","R","AI Startups"],"extracted_tech_keywords":["AI","machine learning","ML","data science","analytics","Aim","R","Go","API","funding"],"url":"https:\/\/analyticsindiamag.com\/ai-startups\/the-bengaluru-based-startup-that-helps-you-find-your-next-store-location\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10120965,"title":"Bad Times for Perplexity AI Begin","content":"Perplexity AI co-founder Aravind Srinivas is super relaxed and busy with media engagements as the company onboards new gentlemen to the board. Meanwhile, OpenAI experiments with a generative AI search experience, and Google announces several upgrades to Google Search. “Startups have to be aggressive in terms of competing against incumbents who already have, like, a billion users (Google Search). OpenAI has 100 million users. We don’t have that today, so it’s on us to achieve that,” said Srinivas, in a recent interview with Bloomberg. Perplexity AI recently onboarded three industry leaders as new advisors to boost search, mobile and distribution: Emil Michael (ex-Uber), Rich Miner (Android co-founder), and Mikhail Parakhin (ex-Bing CEO). “Parakhin is the world’s leading technical expert on search infrastructure and integrating it with LLMs,” said Srinivas, adding that he has been at the forefront of AI and search, and his expertise will enable them to evolve their core answer engine, build in-house search infrastructure, anticipate new and unexpected challenges, and integrate new AI capabilities. Perplexity has not given up yet: Over the last few weeks, Srinivas and his team have constantly been iterating and shipping products and features faster in an attempt to stay ahead of its competitors, alongside new players emerging in the space who are building similar solutions. Recently, the company partnered with SoundHound AI to Bring LLMs-powered voice assistants to cars and IoT devices. This was announced ahead of the OpenAI Spring Update, which gave the world a glimpse of their—or rather ‘Her’ like—characteristics. At Google I\/O, the tech giant also introduced ‘AI Overviews,’ which generates summaries for the queries provided by the user on the go— again similar to Perplexity, but on steroids. But, Srinivas, was quick to point out the drawbacks and limitations of its competitors and said— “When people go to Google, they hate the bad latency, they expect instant rendering of the 10 blue links,” explained Srinivas. He said that Google’s UI is quite cluttered, making people wonder whether they will get token streams as answers, ad-serving links, or some random panel. A Wake-Up Call for Perplexity AI Recently, OpenAI partnered with Reddit to bring enhanced content to ChatGPT and new products, helping users discover and engage with Reddit communities. Interestingly, the company mentioned in the blog post that it will also become a Reddit advertising partner. Surprisingly, OpenAI isn’t planning to charge much from its users and is likely on the path to building its own search engine and monetising it like Google Search. In the latest quarter, Google earned $46.156 billion from Search and related revenues, a 14% increase. To do this, OpenAI will use Reddit’s Data API, which provides real-time, structured, and unique content from Reddit. This will enable OpenAI’s AI tools to better understand and showcase Reddit content, especially on recent topics. Coincidently, this year Google also partnered with Reddit to use its content available for training Gemini. Not to forget, OpenAI has also entered into content licensing partnerships with publishers including the Associated Press, Axel Springer (Business Insider, Politico, Bild, and Welt), The Financial Times, Le Monde, Prisa Media (El País, Cinco Días, As, and El Huffpost), and Dotdash Meredith (People, Better Homes & Gardens, Investopedia, Food & Wine, and InStyle). In a recent podcast with Lex Fridman, OpenAI chief, Sam Altman said that, “The intersection of LLMs plus search, I don’t think anyone has cracked the code on yet. I would love to go do that. I think that would be cool.” The company also hired Shivakumar Venkataraman,a 21-year Google veteran who previously led the company’s search ads business, as its new vice president. Venkataraman’s extensive experience in search advertising at Google makes him a valuable asset for OpenAI as it plans to build ‘Google Search Alternative’. “We are a business and will find plenty of things to charge for, and that will help us provide free, outstanding AI service to (hopefully) billions of people,” said OpenAI chief Sam Altman in his recent blog post. OpenAI hit the $2 billion revenue milestone in December and the company  believes it can more than double this figure in 2025, based on strong interest from business customers seeking to use its technology to adopt generative AI tools in the workplace. Perplexity AI is competing against the likes of OpenAI and Google at a time when several players are innovating in the AI space. However, Aravind Srinivas, the Co-Founder and CEO of Perplexity AI is up to the challenge. In an interview on Bloomberg TV, Aravind shared his views on several topics. Aravind feels the need to be aggressive as they go against the giants in the Industry. One of his primary objectives is to increase the user base since Google has a billion and OpenAI is up to a 100 million users. Perplexity AI recently onboarded 3 industry leaders to improve their search, mobile and distribution. As Perplexity looks to scale to the next level, Aravind says the integration of AI with search is important along with the need to develop an UI which is less cluttered with minimum latency. He believes that Perplexity should build a product which provides pertinent AI based search responses to all user queries unlike Google’s ad supported model. To execute on their plans to compete against the big players, Aravind acknowledges the need to do more fund raising but they will also be looking to increase revenue and efficiencies. As the  # of queries keeps growing, Perplexity CEO thinks they are very well positioned to compete and become a leading player in the field. Aravind also doesn’t rule out growth thru consolidation. Though there is tough competition in the industry, Aravind is confident that they are doing all the right things to stay ahead!","excerpt":"At Google I\/O, the tech giant also introduced ‘AI Overviews,’ which generates summaries for the queries provided by the user on the go— again similar to Perplexity, but on steroids.","categories":["AI Features"],"tags":["OpenAI","Perplexity AI"],"author_name":"Siddharth Jindal","publish_date":"2024-05-18T14:15:12","publication_year":"2024","word_count":953,"keywords":["Go","ChatGPT","API","ELT","OpenAI","AI","Perplexity AI","GPT","generative AI","R","startup"],"extracted_tech_keywords":["AI","generative AI","ChatGPT","OpenAI","R","Go","API","ELT","GPT","startup"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/bad-times-for-perplexity-ai-begins\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10143249,"title":"JetBrains Still Remains the Winner of the AI Coding Race","content":"‘VS Code bad, Cursor good’ has been a hot topic among developers. The conversation has gained even more prominence as VS Code, Microsoft’s own IDE, now dominated coding with AI as GitHub Copilot and Anysphere’s Cursor. On the other hand, JetBrains, a 24-year-old company that developed its own IDE, is often left out in such discussions. Notably, it’s not that developers really need to talk about it as much. Why? Because it is still one of the most successful IDEs ever built. According to reports from Sacra obtained by developer Kenneth Auchenberg, JetBrains’ recurring revenue was $593 million in November 2024, the highest amongst all AI copilots and code editors. For comparison, in the same month, GitHub Copilot generated $400 million and Cursor generated $65 million, while the revenue from VS Code cannot be directly attributed as it is part of Microsoft’s ecosystem. It’s not like JetBrains is not into AI. Last month, the team introduced Mellum, a proprietary LLM designed specifically for code completion on its platform. The announcement which went a little unnoticed, received love from the developer community. Mellum is integrated into JetBrains AI Assistant, a smart coding companion available across the company’s independent development editors. What makes JetBrains so special? According to JetBrains, Mellum significantly reduces the time required to generate code suggestions, cutting it by up to a third in initial tests. Early feedback shows an industry-leading acceptance rate of 40%, with developers frequently applying the AI’s suggestions. Competing in the AI coding assistant space, JetBrains aims to differentiate through Mellum’s seamless IDE integration, similar to how GitHub Copilot was integrated to VS Code last month. “Mellum’s power isn’t just in its speed and accuracy; it also benefits from deep integration with JetBrains IDEs,” said Valerie Kazmina, product marketing manager at JetBrains. “This synergy between the client-side IDE and server-side AI logic allows Mellum to deliver contextually aware code suggestions that are tailored to each project’s needs.” The interesting thing about JetBrains is that even though the price is too high compared to VS Code, which is free, it is offered for free for school and university students, making it appealing for a larger audience. This also helps the students to get used to that ecosystem before getting their hands on VS Code or any other tool. Not all feedback is critical. A developer defending JetBrains’ versatility, said, “Best IDEs in town hands down – for web development, data eng, mobile development, backend – you name it.” “Once you use a JetBrains IDE, there’s no going back.” One question remains: will developers still feel the same next year? What Does the Future Hold? Selling an IDE was not believed to be an ideal business given that VS Code is essentially a free commodity. But it turns out that everyone is doing it by forking VS Code and getting into Y Combinator. Unlike all, JetBrains did it 24 years ago without forking anyone else and has more than 3 million customers and the company employs 2,200 employees. All of this while raising no capital whatsoever. Even though VS Code came years later, it could not make a big dent on JetBrains’ business. Just like VS Code is known for its simplicity, JetBrains old-fashioned operation with IntelliJ makes developers still love how it works. But despite earning accolades for its polished design and unparalleled static analysis capabilities, some users feel JetBrains has been slow to adapt to the AI coding revolution. Critics argue that while JetBrains’ core IDE features remain unmatched, its AI integration could use a facelift. “I’m fed up with how JetBrains handles AI support. You can try workarounds, but when has the bar gone so low for JB,” said a developer on X. This sentiment resonates with others who believe JetBrains risks losing ground to nimble startups like OpenAI-backed Cursor or VS Code with GitHub Copilot, which undoubtedly are on every developer’s “tab” button. But not all is lost. A month back, a discussion on Reddit under Jetbrains subreddit, started by taking a direct hit at Cursor, saying that it is just a hype machine that thrives on spamming competition, and is unable to stand on its own merits. “We all know no product is perfect, including JetBrains. but the need to orchestrate that kind of display speaks volumes of how little you think of your possible future customers,” said the Redditor. This, though not so subtle, clearly highlights how JetBrains is positioned in the market — no hype, only products. Despite criticisms, JetBrains continues to impress with its no-compromise approach to quality and innovation. Though people are coming back to VS Code after trying their hands on Cursor and all of its alternatives, JetBrains remains unshaken with its revenue. With the recent general availability of OpenAI’s Canvas with coding abilities, editing, and suggestions within the UI, it seems like AI-based IDEs are the future. This is only expected to start giving more competition to the likes of Cursor, VS Code, and JetBrains alike.","excerpt":"JetBrains’ recurring revenue was reportedly $593 million in November 2024 – the highest amongst all AI copilots and code editors.","categories":["Global Tech"],"tags":["AI in Coding","JetBrains"],"author_name":"Mohit Pandey","publish_date":"2024-12-11T13:05:46","publication_year":"2024","word_count":830,"keywords":["Go","API","OpenAI","AI","ML","Git","AI in Coding","JetBrains","Aim","copilots","GitHub","R"],"extracted_tech_keywords":["AI","ML","OpenAI","Aim","copilots","R","Go","Git","GitHub","API"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/jetbrains-still-remains-the-winner-of-the-ai-coding-race\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":49809,"title":"How Tesla Uses PyTorch","content":"Last week, Tesla Motors made the news for delivering big with its smart summon feature on their cars. The cars can now be made to move around the parking lot with just a click. A myriad of tools and frameworks run in the background which makes Tesla’s futuristic features a great success. One such framework is PyTorch. PyTorch has gained popularity over the past couple of years and it is now powering the fully autonomous objectives of Tesla motors. During a talk for the recently-concluded PyTorch developer conference, Andrej Karpathy, who plays a key role in Tesla’s self-driving capabilities, spoke about how the full AI stack utilises PyTorch in the background. Tesla WorkFlow With PyTorch via PyTorch Tesla Motors is known for pioneering the self-driving vehicle revolution in the world. They are also known for achieving high reliability in autonomous vehicles without the use of either LIDAR or high definition maps. Tesla cars depend entirely upon computer vision. Tesla is fairly a vertical integrated company and that is also true when it comes to the intelligence of the autopilot. Everything that goes into making Tesla autopilot the best in the world is based on machine learning and the raw video streams that come from 8 cameras around the vehicle. The footage from these cameras is processed through convolutional neural networks(CNNs) for object detection and performing other actions eventually. The collected data is labelled, training is done on on-premise GPU clusters and then it is taken through the entire stack. The networks are run on Tesla’s own custom hardware giving them full control over the lifecycle of all these features, which are deployed to almost 4 million Teslas around the world. For instance, a single frame from the footage of a single camera can contain the following: Road markingsTraffic lightsOverhead signsCrosswalksMoving objectsStatic objectsEnvironment tags So this quickly becomes a multi-task setting. A typical Tesla computer vision workflow would have all these tasks that are connected to a ResNet-50 like shared backbone running roughly on 1000 x 1000 images. Shared because having neural networks for every single task is costly and inefficient. These shared backbone or the networks are called Hydra Nets. There are multiple Hydra Nets for multiple tasks and the information gathered from all these networks can be used to solve recurring tasks. This requires a combination of data-parallel and model-parallel training. via PyTorch Multi-task training can be done through three main ways as illustrated above. Though round-robin looks straight forward, it is not effective because a pool of workers would be more effective in doing multiple tasks simultaneously. And, the engineers at Tesla find PyTorch to be well suited for carrying out this multitasking. For autopilot, Tesla trains around 48 networks that do 1,000 different predictions and it takes 70,000 GPU hours. Moreover, this training is not a one-time affair but an iterative one and all these workflows should be automated while making sure that these 1,000 different predictions don’t regress over time. Why PyTorch? Smart summon in action When it comes to machine learning frameworks, TensorFlow and PyTorch are widely popular with the practitioners. No other framework comes even remotely close to what these two products of Google and Facebook respectively, have in store. These frameworks have slowly found their niche within the AI community. PyTorch, especially has become the go-to framework for machine learning researchers. PyTorch citations in papers on ArXiv grew 194% in the first half of 2019 alone while the number of contributors to the platform has grown more than 50%. Companies like Microsoft, Uber, and other organisations across industries are increasingly using it as the foundation for their most important machine learning research and production workloads. And now with the release of PyTorch 1.3, the platform got a much needed boost as it now includes experimental support for features such as seamless model deployment to mobile devices, model quantisation for better performance at inference time, and front-end improvements, like the ability to name tensors and create clearer code with less need for inline comments. The team also has plans in place for launching a number of additional tools and libraries to support model interpretability and bringing multimodal research to production. Additionally, PyTorch team have also collaborated with Google and Salesforce to add broad support for Cloud Tensor Processing Units, providing a significantly accelerated option for training large-scale deep neural networks. These timely releases from PyTorch coincide with the self-imposed deadlines of Elon Musk on his Tesla team. With the success of their Smart summon, Tesla aims to go fully autonomous in the next couple of years and we can safely say that it has rightly chosen PyTorch to do the heavy lifting.","excerpt":"Last week, Tesla Motors made the news for delivering big with its smart summon feature on their cars. The cars can now be made to move around the parking lot with just a click. A myriad of tools and frameworks run in the background which makes Tesla’s futuristic features a great success. One such framework […]","categories":["Deep Tech"],"tags":["Computer Vision","Pytorch","Self-driving car","Workflow"],"author_name":"Ram Sagar","publish_date":"2019-11-13T14:00:16","publication_year":"2019","word_count":778,"keywords":["Pytorch","machine learning","AI","neural network","Self-driving car","ML","PyTorch","computer vision","Aim","object detection","Computer Vision","TensorFlow","R","Workflow"],"extracted_tech_keywords":["AI","machine learning","ML","neural network","computer vision","Aim","TensorFlow","PyTorch","object detection","R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/tesla-pytorch-self-driving-computer-vision-karpathy-elon-musk-ai\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10012581,"title":"Is India Ready For A Data Sharing Act?","content":"The European Commission introduced the Data Governance Bill on Wednesday as a part of the European Strategy for Data, adopted in February 2020 and to create a single market for data in the EU. Hoping to facilitate the creation of new products and services in the EU, the Commission has introduced the bill with an ‘aim to foster the availability of data by increasing trust in data intermediaries and by strengthening data sharing mechanisms across the EU.’ If adopted the Data Governance Act (DGA) will provide legal recourse for the companies to access public sector data and allow industries to share data among themselves through the introduction of various frameworks that will improve trust by creating safe spaces. The DGA will, at the same time, be in accordance with the existing rules and regulations related to data processing and sharing. It will thus be applied in respect to the General Data Protection Regulation, any national laws, sector-specific guidelines, and any other data governance rules. The Act The DGA introduces several ideas in the form of rules and frameworks that will help them achieve the goal of increasing the use of data for products and services. The public sector is using huge amounts of data in Europe, which even the authorities themselves are restricted from using. With an aim to make this data available for firms to use, the DGA will open up several categories of public sector data which can be used by companies for commercial and non-commercial purposes. While the DGA will rely on national rules for companies’ to access this data, they will lay down certain ground rules for its use, including conditions that make the reuse nondiscriminatory, proportionate, and objectively justified and that they do not restrict competition. Secondly, in the current situation, private companies do not share data with each other due to lack of trust, despite the awareness of several advantages of large datasets. To address this issue, the government introduces a framework to create data sharing services, which can help improve this trust and create European Data Spaces. The DGA lays down several requirements for these data sharing services or data brokers, including notification of its services to the EU member state and fiduciary duty towards data subjects, among others. Another form of data sharing is through data altruism that will encourage companies to gather data, personal and non-personal, for projects of general importance. But for such collection of data, the company needs to again satisfy several conditions. This will need the company to be constituted to meet objectives of general interest, operate on a not-for-profit basis, and ensure the activities related to data altruism take place through a legally independent structure. Is India Ready For Such An Act? While India does not have a bill introduced yet that will encourage sharing of data; it introduced the Personal Data Protection Bill (PDP) in 2019, which is currently being reviewed by a Joint Parliamentary Committee. However, the inclusion and exclusion of sections that talk about data sharing have been largely criticised in both inside and outside the parliament. The bill lets the government direct any data fiduciary or data processor to provide any personal data anonymised or other non-personal data and is not obligated to be transparent of how this data will be used or if the individual who the data belongs to will be compensated or not. Also studies highlight that the bill’s reliance on the consent-based mechanism for protecting personal data is not likely to be effective. In general, the PDP does not put any restrictions on the transfer of personal data, except for ‘sensitive personal data’ which needs explicit consent from the individual. A committee formed last September under MeitY, called for new data regulation that will allow the private sector to use non-personal data or anonymised personal data. It also calls for a data-sharing marketplace for Indian businesses and the government. However, this report has been criticised by experts for leaving too many questions unanswered in terms of who could use that non-personal data and how it will ensure that the community that produces the data benefits them. Even if the data is non-personal, if used for two different purposes by two different communities, it could end up hurting the socio-economically weaker group. Overall it has been criticised for its excessive focus on economic interests rather than the public good and being ‘high on rhetoric and vague in details’, leaving a lot of ambiguity. Wrapping Up A lack of an adequate framework in terms of data sharing might be indicative that we are not ready for such an act. However, a dialogue in the direction of using data for the public good is very much relevant but needs to come in the picture after a robust legal language for privacy, rights, and community-led governance comes into place.","excerpt":"The European Commission introduced the Data Governance Bill on Wednesday as a part of the European Strategy for Data, adopted in February 2020 and to create a single market for data in the EU. Hoping to facilitate the creation of new products and services in the EU, the Commission has introduced the bill with an […]","categories":["IT Services"],"tags":["big data role"],"author_name":"Kashyap Raibagi","publish_date":"2020-11-29T18:00:00","publication_year":"2020","word_count":806,"keywords":["Go","AWS","AI","cloud_platforms:AWS","RAG","Aim","ViT","data governance","Rust","R","big data role"],"extracted_tech_keywords":["AI","Aim","RAG","AWS","R","Go","Rust","data governance","ViT","cloud_platforms:AWS"],"url":"https:\/\/analyticsindiamag.com\/it-services\/is-india-ready-for-a-data-sharing-act\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10138690,"title":"AI-Powered Budy.bot Raises $4.2 Mn to Simplify Sales and Marketing Chaos for Companies","content":"Budy.bot, an AI-driven sales and marketing assistant, has secured $4.2 million in pre-seed funding. The funding round, led by RTP Global with contributions from BeeNext Pte Ltd, The Neon Fund, BITS SPARK Angels, and over 15 angel investors, will fuel Budy’s mission to simplify the complex world of software stacks. Budy’s core AI capabilities are set to get a boost with the funds earmarked for training and development. The company has already collaborated with select enterprise clients to address some of the challenging pain points in sales and marketing like reconfiguring salesforce for optimal performance, plugging lead conversion gaps, and introducing a more accurate bottom-up revenue forecasting model. Targeting growth-stage companies with annual revenues between $25 million and $200 million, Budy aims to solve the unexpected operational hurdles these firms face as they scale. Often, as businesses grow, they resort to adding more software layers, which complicates rather than simplifies their operations. Budy offers a streamlined approach, helping businesses get the most out of their existing systems without the need for endless software add-ons. Speaking on the company’s mission, Samay Kohli, Founder & CEO of Budy, said, “By using AI, we can unlock the potential of tools companies already have, rather than forcing them to pile on new software. With Budy, we’re confident we can help businesses maximize their sales and marketing budgets like never before.” The Chatbot Market According to Grand View Research, the chatbot market is set to reach over $3.9 billion globally by 2030 with over 50 percent of large enterprises investing more capital in chatbots than mobile sales app development. And AI is now an essential part of the sales and marketing strategy. Already IBM’s Watsonx, Amazon’s Lex, and Dell are functional to guiding businesses through their growth journeys with smart AI solutions and seamless support.","excerpt":"According to Grand View Research, the chatbot market is set to reach over $3.9 billion globally by 2030.","categories":["AI News"],"tags":["AI (Artificial Intelligence)"],"author_name":"Vidyashree Srinivas","publish_date":"2024-10-17T11:22:12","publication_year":"2024","word_count":300,"keywords":["API","funding","programming_languages:R","AI","chatbots","data_tools:Spark","ML","Aim","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","ML","Aim","chatbots","R","API","funding","programming_languages:R","data_tools:Spark"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/ai-powered-budy-bot-raises-4-2m-to-simplify-sales-and-marketing-chaos-for-companies\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10097405,"title":"Microsoft TypeChat will Create the Apps of the Future","content":"Microsoft is in its AI age. The company has gone on an integration spree, putting OpenAI’s models into Office, Bing, and even Windows. Now, the tech giant is open sourcing a set of tools to bring this level of AI integration to all developers. The Redmond giant recently released TypeChat, a library of software tools aimed at augmenting traditional UI with the power of AI. By targeting the code generation capabilities of LLMs, researchers at Microsoft have created a system that can help AI algorithms to communicate with applications. This way, the user can rely on natural language inputs to interface with software in addition to existing UI elements. What’s more, in true open source fashion, the toolkit can be used with any LLM of the developers’ choosing, opening the floodgates of AI-powered applications to the world. Just as the company is trying to create the future of computing through the tight integration of AI into applications used by millions every day, it is also enabling a developer movement to enable the larger integration of AI into everyday applications. TypeChat explained One of the biggest issues that developers have been dealing with in terms of integrating LLMs into software is that language models rarely give machine-readable text. Even when the user prompts the model to provide structured data, LLMs fall short. However, the team at Microsoft found that these generative AI algorithms perform moderately well at translating user queries into JSON (JavaScript Object Notation) format, which works great for machines. In the example provided by Microsoft in its blog, the LLM was able to accurately transcribe a customer’s order into a largely error-free JSON output. To ensure that this happens repeatedly, they further refined the technique by adding a type requirement to the output. For example, when asked to respond in a predefined output type called ‘Response’, the LLM creates a far more refined and structured output for the query. This code can also be validated by the TypeScript compiler, which the researchers were using. This paves the way for a clean, structured, and machine-readable output to be used in an app’s workflow. This method, tentatively termed a ‘response schema’, can be used for a variety of different applications by clearly defining a schema for each use-case. Some applications showcased were sentiment analysis, application creation through an ‘API schema’, and a ‘data schema’ for structured outputs. As mentioned previously, the project is not only open source, but also has plumbing to plug into different LLMs. While it was built with OpenAI API and Azure OpenAI service, the researchers clearly mentioned that it can be used with “any chat completion-style API”. However, they also stated that TypeChat works best with models that have been trained on both prose and code. TypeChat is just the latest announcement in an environment that is currently seeing a high degree of growth, enabling developers to create the apps of the future. AI tooling for the future TypeChat is just the latest in a long line of tooling for LLM integration. Arguably kick-started with the launch of LangChain, this field has seen extreme innovation over the past few months. Even AI agents, which had their time in the Sun with offerings like AutoGPT and Baby AGI, can be classified under the category of AI tooling. Also, LLMs have become more capable at parsing large databases thanks to vector database systems like PineCone and Weavite. Any successful innovation in the software field is only amplified by movements in the developer ecosystem. The power of open dev tools is extremely apparent, as seen by the explosive innovation surrounding Meta’s LLaMA model. Even top executives at tech giants have acknowledged that top AI companies have no moat, and that open source will eventually win out. According to a survey by Sequoia Capital, only 15% of its respondents have built custom language models from scratch or open source. However, 38% of them are interested in the app development framework around LLMs, and a whopping 94% use pre-trained models. Another interesting takeaway was that 88% of respondents stated that retrieval mechanisms, like vector databases, are a key part of the tech stack. This shows the hunger that the field has for more developer-focused LLM tooling; an appetite Microsoft seems eager to sate. During the 2023 Microsoft Build conference, the company announced a spate of dev tools, from adopting OpenAI’s plugin standard to super powering the WinML API. These tools also serve the same vertical as TypeChat, targeted squarely at developers looking to build AI applications. With the rise of this ecosystem, developers will soon have the tools to create applications that can integrate AI with a simple text box, and chaining together multiple models will create an AI the likes of which can only be matched in science fiction.","excerpt":"By targeting the code generation capabilities of LLMs, researchers at Microsoft have created a system that can help AI communicate with apps","categories":["Global Tech"],"tags":[],"author_name":"Anirudh VK","publish_date":"2023-07-24T12:00:00","publication_year":"2023","word_count":798,"keywords":["OpenAI","AI","sentiment analysis","ML","vector databases","LangChain","Aim","generative AI","Pinecone","AutoGPT"],"extracted_tech_keywords":["AI","ML","generative AI","OpenAI","LangChain","AutoGPT","Aim","vector databases","Pinecone","sentiment analysis"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/microsoft-typechat-will-create-the-apps-of-the-future\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10041281,"title":"How Analytics Can Benefit the German Mittelstand?","content":"It is perhaps not wrong to think that the biggest consumers of analytics services are found in Insurance, Financial Services, Retail and Healthcare sectors. Companies in these sectors are characterized by their size and scale and their easy name recognition worldwide. So, it is very easy to gravitate towards thinking that analytics probably is more useful to big corporations rather than the smaller ones. The other perception that the data generated by a company grows exponentially with the size of the company and the more data one has the more is the requirement to analyze it, kind of bolster the above assumption. But nothing could be more misleading though, as we will find out by examining the scope of analytics for German small and medium-sized companies popularly known as the Mittelstand. These companies are known for their innovative approach, export orientation, and strong regional and social involvement. Since these companies are generally family-owned, they are also not burdened by the quarter-to-quarter thinking of many larger corporations giving them leeway to think long term, an ideal condition for gathering, analyzing and leveraging meaningful data. Business Case for Analytics for Companies with tightly knit Value Chains Mittelstand could use analytics in three areas – Building a better understanding of internal production processes,Understanding needs of clients and partners and Discovering the relevant characteristics of local and global markets. The above seems at odds with the typical characteristic of Mittlestand – to form tightly knit value chains with close connections between clients, suppliers and other partners – as the easy information flow in this network and relatively smaller sizes of companies which enable managers to frequently interact with workers at the shop floor, seem to obviate the need for insights through analytics. However, many times, the information thus obtained from howsoever trusted a partner needs lots of validation, contextualization and further analysis to become really useful. By further processing this information through the latest technological means and by enriching it with adjacent and historical information, value can be created for not just the company but the entire value chain. Figure 1: Analytics in a typical Mittlestand Value Chain Priority Areas for Analytics for Mittlestand Data Analytics for Exploring Paths to New Markets: Mittlestand firms have highly loyal customers so customer loyalty and customer service, which is normally the priority area in big corporations, perhaps can wait, what couldn’t wait is deploying analytics in gaining new market access. This is not an easy and straightforward ask but worth considering nonetheless. Here in house data alone may not be immediately useful and data needs to be acquired from external sources. For example, data about local market demographics or local businesses could be acquired and then analyzed along with internal data creating a better list of targeted prospects enabling not only a better marketing strategy and engagement but also a better product-market fit. Regional data could also be analyzed to predict the shift in consumer preferences and thereby predicting a shift in offerings from local businesses who could be potential customers to Mittlestand firms. Mittlestand had been traditionally nimble enough to quickly act on any shift spotted in customer preferences given the shift is spotted in time. Here thoughtfully designed and deployed data models could really provide a competitive edge to Mittlestand by spotting the shift in time and by providing necessary details of the shift. Figure 2: Finding a Pathway to a New Market Managing and Analyzing Deluge of Data from Internet of Things (IoT): A 2019 study commissioned by Deutsche Telekom entitled “The Internet of Things in German SMEs” reveals that SMEs in Germany are placing an increased emphasis on new IoT application cases. With increased adoption of IoT, the data with the companies will grow exponentially and it will become imperative to find innovative ways to manage and analyze this data. Many cloud data platforms like Record Evolution, Snowflake, Dataiku, Panoply are launched keeping in mind the needs of SMEs to host and analyze their IoT data in the cloud. Another important finding of the above report is that the Mittlestand are currently concentrating their IoT investments in the area of “Predictive Maintenance” (33% of the survey respondents said so) where the role of data analytics is perhaps the most pronounced.  IoT RoI could be further enhanced by using analytics for developing new creative ideas for IoT deployments. Figure 3: IoT Analytics Analytics for Commercial Operations: If there is one thing Mittlestand can’t take for granted then that is customer loyalty, howsoever entrenched they may be in a supply chain or howsoever well known they may be for the quality of their products. As the word about success of Mittlestand model spreads around the world they can only expect more intense competition from those who will emulate them. In this environment, knowing the long-term orientation of the Mittlestand, it’s imperative for them to focus more on commercial operations despite having relatively smaller sales teams. In fact, commercial operations like revenue, service and sales operations could be converted into analytics command centers to overcome the limitations of having smaller Go-To-Market teams. Mittlestand can also look for firms like Allura Analytica to run or setup their commercial operations and analytics command centers for them. Figure 4: Analytics Command Center Led by Commercial Operations Conclusion It is now a cliché to say that data is the new oil, however, companies are beginning to realize only now that they can create lots of competitive advantage by focusing on building better analytics engines which can extract value from even the last drop of this oil. Mittlestand can enhance their traditional competitive advantage manifold by focusing on analytics to explore and access new markets, better understand and fully automate their production processes through IoT and leverage commercial operations to not only better understand and serve their customers but even understand consumers who ultimately influence the choices their customers make. Given the proliferation of cloud-based products and service providers and the advent of niche analytics and technology consulting firms like Allura Analytica it was never so easy for Mittlestand to take the plunge into the world of Analytics. Bibliography BDI Fact Check. (2021). The German Mittlestand -Data, numbers, facts. Bianchini, M., & Michalkova, V. (2019). Data Analytics in SMEs: Trends and Policies. Eclipse IoT. (2021). Open Source Software for Industry 4.0. Schroder, C. (2017). The Challenges of Industry 4.0 for Small and Medium-sized Enterprises. Vogt, A. (2019). Das Internet der Dinge im deutschen Mittlestand: Bedeutung, Anwendungsfelder und Stand der Umsetzung.","excerpt":"It is perhaps not wrong to think that the biggest consumers of analytics services are found in Insurance, Financial Services, Retail and Healthcare sectors. Companies in these sectors are characterized by their size and scale and their easy name recognition worldwide. So, it is very easy to gravitate towards thinking that analytics probably is more […]","categories":["AI Features"],"tags":[],"author_name":"Shashank Chandra","publish_date":"2021-06-04T14:38:00","publication_year":"2021","word_count":1079,"keywords":["Go","programming_languages:R","AI","RAG","ViT","analytics","CLIP","Rust","R","Snowflake"],"extracted_tech_keywords":["AI","analytics","RAG","Snowflake","R","Go","Rust","CLIP","ViT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-analytics-can-benefit-the-german-mittlestand\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":6,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10094139,"title":"What is Microsoft Without OpenAI?","content":"The non-profit organisation that it was, OpenAI, gained investments from Microsoft right after becoming a “capped” profit company in 2019. This was not received well by a lot of people, including co-founder Elon Musk, who decided to leave the company a year before that. The open source AI company became closed and bid to mint money. Microsoft looks like the clear winner in this AI battle. The $13 billion bet that the company has made on OpenAI, is reaping benefits for the tech-giant while competitors like Google and Meta are running for their money to keep up. But if we look even closely, and look at the recent developments by Microsoft that was announced at the Build Conference, all of that would not have been possible without OpenAI. “Microsoft hasn’t released a single product in 30 years,” has been a joke that people make about the company. Even when it comes to Windows, the company has been only releasing updates to its systems, and not building anything new. The same is the case with Azure Cloud service. Though Azure is one of the top players of the cloud game, the recent AI integrations through Azure OpenAI Service, as the name suggests, are just products built by OpenAI like GPT or DALL-E offered through the cloud. What’s in it for OpenAI? Recently, Musk claimed that Microsoft ‘controls’ OpenAI. The tech leader is clearly not happy with how OpenAI is turning out now — closed source and for profit. Who is the biggest beneficiary here? Microsoft. The company, by offering its cloud in exchange for the investment, is somehow controlling OpenAI and owning its services and has made OpenAI a “maximum profit company effectively controlled by Microsoft”. Though Satya Nadella, CEO of Microsoft, denied Musk’s claims, his statement wasn’t very convincing. “The last time I checked, we were the only for-profit company comfortable with a non-profit company and a board controlling technology, and I would welcome others to do that as well,” he added. After the initial investment of $1 billion in 2019, OpenAI was collaborating with Microsoft exclusively. Microsoft gained a lot, but what did OpenAI gain from this apart from the money and the fame from ChatGPT? Microsoft was clearly involved in building the technology at OpenAI. For example, the Spark of AGI paper by Microsoft Research explained how the company had access to the training process of GPT-4 before and after RLHF. It is possible that the company might be giving its own “human feedback in the loop”. There is more. While Microsoft reported a quarterly profit of $18.3 billion in April, OpenAI is not profitable yet. According to reports, OpenAI’s losses have doubled to $540 million ever since it started developing ChatGPT. Though the revenue is pouring in, it might take a lot of time for it to get profitable. Microsoft-GPT One can say that what Microsoft was not able to achieve through its own research department is now being achieved by funding a non-profit company like OpenAI. Moreover, the cost of building such models is so huge that it is not wise to blame OpenAI for receiving funds from Microsoft. On one side it looks like Microsoft is furthering AI research through OpenAI and on the other, it looks like it may just be exploiting the company. Microsoft has integrated Codex and GPT in every single offering that they have announced recently. They want to make everyone a developer, and to do that, they are making coding as easy as writing English through ChatGPT. More than just improving its cloud, Microsoft has also found the best route to gather the data gold mine through ChatGPT. Since the chatbot runs on Azure Cloud, Microsoft has access to all the data given to ChatGPT by users. OpenAI currently claims that they are not using data to build better models. But there is a possibility in the future. Moreover, OpenAI’s privacy policy clearly states that it “will share personal information of the users’ with our vendors and service providers”. Microsoft is clearly the biggest service provider of OpenAI. Microsoft is starting to treat and sell GPT as its own product. What would Microsoft do without OpenAI? The partnership and investment in OpenAI have positioned Microsoft in a unique place to enable generative AI capabilities across its entire stack, from Windows to Office 365. Copilots for everything seem to be Microsoft’s generative AI flag — and it was incredibly present at the Build conference. From Plugins to Copilot, Office 365 Copilot to Azure AI Studio, Microsoft Fabric, and much more, everything is built around OpenAI’s GPT technology. This is similar to Google I\/O 2023, which like Microsoft Build, was just an “AI integration event”. No one was actively choosing to use Microsoft products before OpenAI came into the picture. Though they host one of the biggest cloud services, integrating OpenAI made startups and companies rush towards it, giving AWS and Google a run for their money. Microsoft has acquired a 49% stake in OpenAI, while the other 49% and 2% are with the VCs and OpenAI, respectively. After integrating OpenAI Service on Azure, Microsoft shares rose by 8.3%, giving a boost to their sales. This is how Microsoft sold itself OpenAI. The company solved OpenAI’s distribution problem. They also acquired the Microsoft brand in them somehow. Now that isn’t required anymore though. Microsoft has palmed off the work to build AI models to OpenAI and is running in the forefront of the dirty AI battle with AWS. OpenAI is the answer to Microsoft’s cloud question. Google has its own Bard, Meta has its metaverse dreams, Microsoft has nothing apart from Windows and Office 365 products. One of the first voice assistants, Cortana, also is now a distant memory with the current space of AI chatbots. Everything else is now getting integrated with GPT or some OpenAI developed technology. Microsoft would have nothing if it did not have OpenAI.","excerpt":"Google has its Bard, Meta has its metaverse dreams, Microsoft has nothing apart from the age-old Windows and Office 365 products","categories":["Global Tech"],"tags":["ai investments","bill gates","Google PaLM","GPT-4","Microsoft Azure","OpenAI","Satya Nadella"],"author_name":"Mohit Pandey","publish_date":"2023-05-30T16:00:00","publication_year":"2023","word_count":987,"keywords":["Satya Nadella","ChatGPT","RLHF","ai investments","OpenAI","AI","chatbots","AWS","GPT-4","Google PaLM","Aim","generative AI","Microsoft Azure","bill gates","copilots","Azure"],"extracted_tech_keywords":["AI","generative AI","ChatGPT","OpenAI","Aim","RLHF","copilots","chatbots","AWS","Azure"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/what-is-microsoft-without-openai\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":25252,"title":"GDPR Paves The Way For Data Subjects Of The World","content":"The European General Data Protection Regulation (GDPR) has been in the limelight these last few weeks. And for good reason, too. It is an extensive document, to say the least. One reporter found that the number of pages in the document are enough to cover a football field — which is 360 feet long and 360 feet wide. The document has been in the pipeline since the former Vice-President of the European Commission Viviane Reding initiated it in 2012 and it finally went into effect on 25 May. It seeks to not only set the standard for data protection and privacy in Europe but across the globe for all “data subjects” (read: anyone on the internet anywhere). The GDPR makes two critical changes to the Data Protection Directive of 1995 that preceded it. First, it emphasises the universal nature of the legislation. For any legislation to apply to all world citizens, this is undoubtedly a prerequisite. The GDPR tries to prescribe the best possible rules and practices for data regulation that can be common to citizens of all nations. Second, it is legally binding. Any company that breaks the regulation is liable to pay up to 4 percent of their worldwide sales as retribution. This, of course, is a macroscopic view of the regulation that is presented by experts. The regulation is still expected to have loopholes. To device perfect laws in complicated matters such as protection of data privacy is close to impossible. The GDPR will be prone to discrepancies in interpretation and implementation. Nonetheless, some things are as clear as daylight in it. The document is strongly against companies sharing personal data of their users or customers with any third parties. If this must be done, it must be after gaining prior permission, or at risk of serious consequence. Viviane Reding, referred to the recent Facebook Cambridge Analytica scandal to make this clear. She said, “The Facebook Cambridge Analytica scandal, if it had happened on 26 May, this year, would have cost billions of euros to Facebook, among others. You cannot hand over the personal data of citizens without having asked if the citizens agree that you hand it over. And you cannot steal it and just tell them after. That is not possible anymore, according to the new law. If you do, then the penalties will be very, very severe.” The Road Ahead Companies are taking the GDPR seriously. It’s why you have been receiving and will continue to receive emails from digital services about updates in their privacy policies. According to experts, MNCs are hiring as many as 300 to 500 people just to ensure their compliance with the GDPR. They are spending as much as 50 million dollars on this. Of course, the same cannot be said of small companies. Bigger companies are better poised to absorb the GDPR than smaller ones. This is one of the drawbacks of having extensive regulation for privacy. Having said that, it paves the way for overall information hygiene, which was always bound to have a cost. This cost is not solely borne by companies, mind you. The GDPR is also adding to the costs of agencies that are responsible for the enforcement of its provisions. The Irish Data Protection Committee has also added around 100 employees to their books with 40 or so more professionals expected to join soon. These new hires come from varied backgrounds. Some are lawyers, others are media professionals or business analysts, system analysts — all hired with the goal of cracking down and investigating as many breaches to the GDPR as possible. One provision of the GDPR that supports this goal is article 80, which gives validity to class-action lawsuits. It gives NGOs and active citizens, who care about data protection, the right to protect public interest – to protect the rights of those individuals in society, who do not have the time to keep up with the challenges and issues of protecting data privacy. The GDPR is an ambitious legislative document. It is the first of its kind in the pursuit of global cybersecurity. And it includes substantial measures for enforcement. It is also very extensive, making compliance a challenge. But companies are making changes to their privacy policies, which is positive. Whether these changes are merely words of caution exposing loopholes in the GDPR, or actual reforms for data protection remains to be seen. All eyes will be on the Irish Data Protection Committee and other GDPR enforcing agencies to see if it’s truly effective.","excerpt":"The European General Data Protection Regulation (GDPR) has been in the limelight these last few weeks. And for good reason, too. It is an extensive document, to say the least. One reporter found that the number of pages in the document are enough to cover a football field — which is 360 feet long and […]","categories":["IT Services"],"tags":[],"author_name":"Vikalp Jain","publish_date":"2018-06-08T05:44:54","publication_year":"2018","word_count":754,"keywords":["Go","AWS","AI","cloud_platforms:AWS","programming_languages:R","programming_languages:Go","Git","R"],"extracted_tech_keywords":["AI","AWS","R","Go","Git","cloud_platforms:AWS","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/gdpr-paves-the-way-for-data-subjects-of-the-world\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10171398,"title":"India Struggles to Adopt AI in SMEs and MSMEs","content":"AI is not being fully leveraged in the small and medium-sized enterprises (SMEs) and micro, small, and medium enterprises (MSMEs) manufacturing sectors in India. At the Bangalore Chamber of Industry and Commerce (BCIC) summit on Wednesday, industry experts emphasised that it should be viewed as a chance for transformation rather than a disruptive threat. The discussions focused on ‘Transforming industries with AI to achieve unmatched productivity and efficiency’. Industry leaders stressed the need for businesses to leverage AI’s capabilities to enhance processes, make better decisions, and grow sustainably in a progressively digital economy, The Hindu reported. India boasts a large manufacturing sector, with more than 90% of our industrial enterprises classified as MSMEs. As stated by G Prakash, chairperson of the manufacturing expert committee, BCIC, these businesses are responsible for almost 30% of India’s GDP, provide jobs for 110 million individuals, and represent 48% of India’s exports. “Many MSMEs have yet to embrace digitalisation or AI, often perceiving these technologies as costly and complex. AI-driven solutions now available through affordable cloud-based and plug-and-play models can deliver productivity gains of 15% to 30%, enabling predictive maintenance, quality monitoring, and energy optimisation, even in small-batch environments,” he added, as reported by Deccan Herald. Despite being vital to the economy, MSMEs struggle with outdated processes, low digital adoption, and limited access to technology. AI bridges this gap by providing affordable automation, data-driven insights, and personalised customer experiences—once only available to large enterprises. According to a PwC India report, AI adoption in MSMEs could boost India’s GDP by $500 billion by 2025, making it one of the most impactful innovations for the country’s economy.AI supports the development of smart supply chain management systems in Indian SME manufacturing. It analyses supply and demand patterns, optimises inventory, and improves demand forecasting, allowing SMEs to reduce stockouts and inventory costs while enhancing efficiency. By adopting AI solutions, Indian SMEs can better respond to market changes, collaborate with suppliers, and ensure timely product delivery.","excerpt":"AI adoption in manufacturing is 35–40%, with strong growth in China, Germany, and the US, while India lags at just 15% among MSMEs.","categories":["AI News"],"tags":["AI in India","India SMEs MSMEs"],"author_name":"Smruthi Nadig","publish_date":"2025-06-05T14:53:47","publication_year":"2025","word_count":326,"keywords":["India SMEs MSMEs","AI in India","AI","programming_languages:R","data-driven","innovation","Git","RAG","automation","ViT","R"],"extracted_tech_keywords":["AI","RAG","R","Git","ViT","automation","data-driven","innovation","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/india-struggles-to-adopt-ai-in-smes-and-msmes\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":64693,"title":"What Is The Hiring Process Of Data Scientists At Infosys?","content":"The hiring team at IT services giant Infosys succinctly describes the company’s ideal data science candidate as one who brings the right mix of technical and domain expertise, along with good communication skills. But with the problems associated with differentiating between what a candidate writes on their resumes, versus what their actual skills may be, the hiring process becomes a tad bit more challenging. With almost two decades of experience developing organisational strategies and analysing human resources issues at Infosys, Richard Lobo, Executive VP & Head HR at the tech giant, has seen the company grow – in both their data science requirements, as well as the methods they adopt to fulfill them. Does the company, like some, still prioritise a candidate’s educational background when hiring for these roles? “While practical, working domain knowledge and technical skills still remains the most important criteria in the selection process, educational background is also taken into consideration, especially for junior-level candidates,” says Lobo. “This ensures that we take a balanced perspective when hiring data scientists,” he adds. According to him, relevant certifications also help, especially because additional qualifications allow candidates to demonstrate their learning. With a host of data science certifications made free amid Covid-19 lockdowns, this may be a good time for aspiring candidates to upskill themselves and be better prepared for the interview process at Infosys. Hiring Process Of Data Scientists At Infosys Along with hands-on coding experience on R, Python and SAS, the company looks at a candidate’s ability to scale up in cloud technologies and problem-solving skills. “Our hiring process usually involves screening and a preliminary shortlist, followed by technical and personal interviews and further discussions as needed,” says Lobo. “We also check references and background information provided to us before proceeding with hiring,” he adds. The candidate’s experience and the projects they have worked on forms the basis of these interviews. Lobo says that a lot of emphasis is placed on the techniques used to solve business problems, the ability to understand and analyse data, and the overall problem-solving approach taken by the candidate. “We also look at the candidate’s comfort level and approach with coding syntax by giving them real-life business problems,” says Lobo. “Additionally, we also look at certain behavioural aspects like team working skills, client handling as well as communication,” he adds. ALSO READ: Best Firms In India For Data Scientists To Work For In 2020 Non-Traditional Recruitment Approach Like some other companies, Infosys conducts hiring throughout the year to engage with potential talent based on requirements. Going beyond conventional means, the company consistently explores non-traditional ways to hire data science professionals. “We actively network with the right talent in various conferences, hackathons, and other events, and alert them whenever a suitable opportunity emerges,” says Lobo. “We also continue to focus on training and reskilling as a source for talent, as the industry demand will continue to evolve,” he adds. In addition to this, Infosys also emphasises on maximising the relevant profile inflow from employee referrals and applications to their website. Furthermore, the company also works through their social and professional media channels for their recruitment needs. “For rare skills, we do look at hiring through professional firms as well,” says Lobo. According to him, despite these efforts, the demand for exceptionally good data scientists usually is higher than the supply. How does the company, then, overcome this deficit of talent? “We have an excellent training infrastructure and ability to deploy learning and upskilling in a large way,” says Lobo. “We use a balanced approach of attracting and retaining the best, as well as building up our talent pipeline through training,” he adds. ALSO READ: Top Free Courses To Help Data Scientists Prepare For Job Interviews Common Mistakes When Hiring Data Scientists As a hiring manager, it becomes extremely important for Lobo and his team to onboard the right talent for such a critical function. “Since data science is a trending professional field, we need to differentiate between those who have actual expertise from those who might have basic knowledge,” explains Lobo. “This becomes an interesting challenge because it is preferred that we do not have a client deployment of someone who is not professionally ready for the role,” he adds. The potential to solve business problems using technology and the probability of someone bridging both worlds effectively is virtually limitless. With that, then, what does a successful candidate do once hired? “Junior candidates or candidates with less industry exposure are expected to scale up on new technologies, exhibit flexibility to absorb instructions and gain experience that makes them ready for higher roles,” says Lobo. “Mid to senior-level candidates, on the other hand, should demonstrate the ability to understand client requirements, lead teams and grow the business,” he adds.","excerpt":"The hiring team at IT services giant Infosys succinctly describes the company’s ideal data science candidate as one who brings the right mix of technical and domain expertise, along with good communication skills. But with the problems associated with differentiating between what a candidate writes on their resumes, versus what their actual skills may be, […]","categories":["AI Hirings"],"tags":["data science jobs india","data scientist qualifications","Data Scientists","Infosys","qualifications for data scientist"],"author_name":"Anu Thomas","publish_date":"2020-05-06T19:00:00","publication_year":"2020","word_count":794,"keywords":["data science jobs india","data science","Go","Infosys","data scientist qualifications","AI","programming_languages:R","programming_languages:Go","qualifications for data scientist","Python","programming_languages:Python","GAN","R","Data Scientists"],"extracted_tech_keywords":["AI","data science","Python","R","Go","GAN","programming_languages:Python","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-hiring\/what-is-the-hiring-process-of-data-scientists-at-infosys\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10069841,"title":"SAS acquires risk tech innovation company Kamakura","content":"AI and analytics company SAS has acquired Kamakura Corporation. The latter is a privately held company that offers specialised software solutions, and data and consulting services for banks, insurance companies, pension funds, asset managers, and other financial organisations. Two of Kamakura’s main offerings are – Kamakura Risk Manager, a fully integrated risk management system for ALM; Kamakura Risk Information Services (KRIS), a SaaS tool for credit risk data and analytics. With this acquisition, SAS aims to deliver a suite of integrated risk solutions with a focus on asset-liability management and to serve additional facets of the financial services industry. SAS co-founder and CEO Jim Goodnight said, “This acquisition is an extension of tremendous investments already made in SAS’ cloud-ready risk management platform and integrated solutions. It signals our intent to advance market-changing risk solutions to solve the most pressing challenges our financial services customers face. We foresee that the resulting strength of SAS technology, paired with Kamakura’s risk analytics and credit models, will prove far greater than the sum of its parts.” Kamakura chairman and CEO Don van Deventer, who founded the company in 1990, said that SAS aligns with his company’s “data-driven, research-oriented cultures and their mutual excellence in modelling and analytics.” He further said that adding SAS’ cloud-native Viya technology, risk domain capabilities, and user-friendly interfaces to Kamakura’s IP will help in creating a ‘top-tier market-changing ALM offering.”","excerpt":"This acquisition is an extension of tremendous investments already made in SAS’ cloud-ready risk management platform and integrated solutions.","categories":["AI News"],"tags":["Mergers and Acquisitions","sas"],"author_name":"Shraddha Goled","publish_date":"2022-06-28T13:41:50","publication_year":"2022","word_count":230,"keywords":["Go","programming_languages:R","AI","data-driven","programming_languages:Go","Aim","analytics","GAN","sas","Mergers and Acquisitions","R"],"extracted_tech_keywords":["AI","analytics","Aim","R","Go","GAN","data-driven","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/sas-acquires-risk-tech-innovation-company-kamakura\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10069794,"title":"Do you really need to learn Kubernetes?","content":"From being an internal container orchestration solution at Google to open-sourcing it in 2014 to the tool we all know today, Kubernetes (aka K8s) has formed its ecosystem. It is becoming the de-facto standard for running microservices across cloud platforms. The majority of developers today are using Kubernetes in some form and shape. Jeremie Vallee, AI infrastructure lead at Babylon, said it is a great platform for machine learning because it comes with all the scheduling and scalability you need. Meanwhile, travel-tech platform Booking.com also has a long history with Kubernetes. In 2015, the team prototyped a container platform based on Mesos and Marathon. “As our users learn Kubernetes and become more sophisticated Kubernetes users, they put pressure on us to provide a better, more native Kubernetes experience, which is great. It is a super healthy dynamic,” said Ben Tyler, principal developer at Booking.com. Kubernetes’ use cases Ahmed ElFakharany, DevOps and cloud engineer expert, said that developers need to deal with Kubernetes as an ecosystem rather than a single technology. People need to explore Kubernetes use-cases and learn about the supporting tools that integrate with it to solve bigger problems and increase the overall value. Here are some of the popular Kubernetes use cases: Develop simple applications: Get your app running by deploying it on a Kubernetes cluster. Microservices: Orchestrating complicated apps based on microservices architecture with many components. Lift and shift: Featuring the moving of on-prem apps to the cloud. Cloud-native network functions: Manage containers with encapsulated network functions. Machine learning: Managing machine learning workflow to deploy faster AI-based apps. Heavy computing: Enable heavy computing tasks. CI\/CD: Easier to perform with Kubernetes. Kubernetes supporting tools Today, a large majority of cloud providers are offering Kubernetes as a service, thus making it easier to adopt it as the foundation of distributed infrastructure. In addition, there are plenty of backup tools like Velero (formerly Heptio Ark), deployment managers like Helm (now ArgoCD and Flux), Ingress Controllers, API gateways, and the policy-as-code tools like OPA, etc. The list goes on. ElFakharany believes these tools came to existence because of the popularity Kubernetes has gained over the years. These tools aim to add even more value and solve issues not natively dealt with by the cluster. Tools supporting the Kubernetes ecosystem Making sense of Kubernetes certifications “Getting certified is just the start,” said ElFakharany. He said some DevOps engineers learn Kubernetes with the sole intention of getting certified. Some popular certifications like CKA, CKAD, and the newest CKS may add value to your CV. They may get you an appointment for an interview, but they will not guarantee you a job. Chandresh Shah, vice president at CITI, believes that certificates are important to understand the in and out of k8 and security. He said that even when people start solving a particular problem with some tools, they still need to know the basics. “Even if you go through the tools, it’s not a guarantee to get the jobs,” Shah added. Further, he said, all tools come with certifications. “It is only your experience and the way you tackle the interview that can give you a job,” said Shah. So, the question is, should you focus on getting more and more certifications to prove your understanding of Kubernetes and its supporting tools to the employer, or focus on the whole process of learning Kubernetes and explore ways to improve the ecosystem it has built over the years? Software developer Nozim Mehrubonov said: “That’s why I start to hate Kube once it’s getting ‘over certified.’ People just chase certificates for the sake of it instead of learning the tool.” Wrapping up Simply put, showcasing a Kubernetes certificate on your resume, to an extent, can get you an interview. But, it becomes important for candidates to showcase project\/hands-on experience instead of a certification-based resume, where you are showcasing real use-cases and revealing how various supporting tools helped you deploy the required task. This can be a simple app, deploying computing-heavy tasks, machine learning, etc.","excerpt":"People just chase certificates for the sake of it instead of learning the tool.","categories":["AI Features"],"tags":["Kubernetes"],"author_name":"Amit Naik","publish_date":"2022-06-27T18:00:00","publication_year":"2022","word_count":668,"keywords":["Go","machine learning","AI","R","CI\/CD","Scala","microservices","Aim","DevOps","Kubernetes","kubernetes"],"extracted_tech_keywords":["AI","machine learning","Aim","kubernetes","microservices","R","Go","Scala","CI\/CD","DevOps"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/do-you-really-need-to-learn-kubernetes\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10124066,"title":"OrbitShift Secures $7 Million in Funding from Peak XV&#8217;s Surge and Stellaris Venture Partners","content":"OrbitShift, a startup that harnesses the power of AI for sales intelligence, has secured $7 million in a seed funding round. The round was led by Peak XV Partners’ Surge scale-up program and Stellaris Venture Partners With the newly raised funds, OrbitShift aims to boost its investments to grow its footprint and customer base in the US. Additionally, the company intends to invest in its technology and product teams to diversify its product portfolio. “We are at an interesting juncture, where we have seen phenomenal traction over the last 18 months with some of the leaders in the industry being our clients,” told OrbitShift co-founder and chief executive Saurabh Mishra. Established in 2022, OrbitShift serves as a comprehensive platform designed to enable the entire sales ecosystem, encompassing pre-sales, sales operations, and marketing insights for technology and IT services firms. The platform utilizes specialized models and extensive language models, primarily catering to large enterprise clients with user bases in the US, European Union, and Asia Pacific regions The recent funding marks OrbitShift’s second institutional funding round, having previously raised $1.5 million in pre-seed funding in 2023 from Stellaris Venture Partners and other angel investors. This brings the total amount of funds raised by the company to $8.5 million. “The OrbitShift team is targeting some of the largest businesses in the world and there is no doubt that AI can make a massive impact on them and therefore on their customers,” said Alok Goyal, partner at Stellaris Venture Partners. “But you need to meet a bar for accuracy which is very high and building a product and the underlying technology stack for that is therefore non-trivial and costs money,” he said. OrbitShift follows a subscription-based pricing model, where clients are charged according to the length of the deal and various factors such as the number of users and accounts created According to the company, OrbitShift’s platform can decrease research and sales planning time for enterprise clients by 40-50%, thereby expediting tasks such as client outreach, generating high-quality responses, and creating content for client meetings. This, in turn, leads to improved productivity and stronger client relationships. “We believe that pretty much every enterprise process that we know of will be rethought or will be reimagined and therefore, we are massive believers of a lot of new great enterprise application startups being created,” Goyal added. Peak XV Partners’ scale-up program for early-stage startups, Surge, has made its ninth investment in OrbitShift. Surge’s previous investments in the deeptech manufacturing and AI space include Ethereal Machines, ZeroK, Australia-based Relevance AI, and Singapore’s Pix.ai","excerpt":"OrbitShift aims to boost its investments to grow its footprint and customer base in the US.","categories":["AI News"],"tags":["AI","Fund Raising"],"author_name":"Tarunya S","publish_date":"2024-06-20T13:01:08","publication_year":"2024","word_count":427,"keywords":["Go","funding","programming_languages:R","AI","Fund Raising","programming_languages:Go","Aim","ViT","R","startup"],"extracted_tech_keywords":["AI","Aim","R","Go","ViT","startup","funding","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/orbitshift-secures-7-million-in-funding-from-peak-xvs-surge-and-stellaris-venture-partners\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10013397,"title":"Does India Have The Infrastructure To Implement Policy Recommendations To Avoid Algorithmic Bias","content":"In recent news, the Centre for Data Ethics and Innovation (CDEI) in the UK, published a review study on the risks of bias in algorithmic decision making (ADM) systems. Commissioned by the UK government, during the 2018 October Budget session, this document analyses the impact of bias in algorithms that can represent a significant and imminent ethical threat. Based on the analysis, it presents several policy recommendations to the government and regulators. This article tries to analyse some of these policy recommendations to the government, the reasons that led to formulating them, and if they can be adopted in the Indian context given the legal bodies and infrastructure. Ensuring diversity for protection against bias The CDEI document recognises the significance of diversity across a range of roles involved in the deployment and development of ADM systems. In order to ensure diversity, the CDEI recommends that the government should continue to support and invest in programmes that facilitate greater diversity. This point is relevant especially in India with a wide range of demographics in terms of caste, religion, language, sexuality, and state, among others. While India currently does not have a mechanism to ensure diversity in the development and deployment of ADM systems, the country has used reservation quotas to ensure the representation of historically and currently disadvantaged groups in government offices and education. Affirmative actions similar to the reservation system can be leveraged to promote more representation in tech firms, to achieve algorithmic fairness. Setting up safe guidelines to monitor outcomes & analysing bias Data is needed to monitor outcomes and identify bias, but access to characteristic data can become a tricky affair. The CDEI document calls for working with ‘relevant regulators’ for clear guidance on collection and use of protected characteristic data for monitoring outcomes of ADM systems. For bias evaluation, it recommends leveraging frameworks such as the Secure Research Service of the Office of National Statistics, that allows access to only accredited researchers. On the other hand, the Data Security Council of India, a research body set by NASSCOM, has been committed to creating safe cyberspace by establishing best practices, standards, and initiatives. The body does extensive research on data protection frameworks, introduced in India, in the form of bills and committee reports. This expertise can be used to regulate and define robust guidelines to keep algorithms in check for biases. Establishing laws to address the resulting discrimination To address discrimination resulting from algorithmic bias, the CDEI does not believe in the need for a new specialised regulator or primary legislation for now. It, however, recommends more guidance that clarifies the ‘Equality Act responsibilities’ of organisations that use ADM systems, not only in terms of mitigating technical bias but also the collection of personal data. This recommendation to not introduce new legislation is based on several instances that proved the current legislation to be efficient. For instance, a recent judgement in the courts successfully retracted an ADM system for facial recognition, deployed in the public sector, because enough steps were not taken to establish fairness in that system. On the other hand, the extant law in India doesn’t account for fairness of ADM systems. While there is some legal framework to address who can use or process data, these rules were not made in consideration of the ADM systems. One of the legal frameworks is the Privacy Protection Bill 2019, which is however currently being examined and reviewed under a Joint Parliamentary Committee. To address the issue of discrimination in general, the Constitution of India under Article 14 does have provisions for the ‘equality before law’. However, similar to the UK’s Equality Act, this Article lacks the language to address ‘equality’ in the context of ADM systems. If the Data Protection Bill is passed with robust guidelines on processing the data along with more accountability on the entity that is processing it, it can then be combined with Article 14 to form a legal framework to address the resulting discrimination from algorithms. Establishing mechanisms for transparency and explainability of ADM systems The CDEI document states that the UK government has shown leadership in setting out guidance on AI usage in the public sector. However, it still calls for a mandatory transparency obligation on all public sector organisations using algorithms that have a ‘notable influence on significant decisions affecting individuals’. To ensure more transparency in the public sector, the Government of India passed the Right To Information Act (RTI) in 2005, which has been used before to get more transparency on algorithms as well. However, experts have mentioned that there is a major lack of legal processes available to actually hold an algorithm accountable. In terms of explainability, NITI Aayog, India’s policy think-tank, introduced the concept of Explainable AI (XAI), to create a suite of machine learning techniques that produce more explainable models. Achieving transparency and explainability for public sector algorithms needs more work over the current existing frameworks. Similar to the UK, laws need to be introduced to make public sector algorithms open. This can help industry experts analyse algorithms for its fairness and hold the governments accountable. Wrapping Up In the case of India, an oversight body was introduced by NITI Aayog to play an enabling role for the research and policy for AI in the country. While this article analyses mechanisms in the Indian system that can be leveraged to address algorithmic bias, a thorough review by this body will be important. This will help identify issues in an Indian context.","excerpt":"In recent news, the Centre for Data Ethics and Innovation (CDEI) in the UK, published a review study on the risks of bias in algorithmic decision making (ADM) systems. Commissioned by the UK government, during the 2018 October Budget session, this document analyses the impact of bias in algorithms that can represent a significant and […]","categories":["AI Features"],"tags":["AI policy","Ethical AI","Explainable AI","NITI Aayog","Responsible AI","Transparency AI"],"author_name":"Kashyap Raibagi","publish_date":"2020-12-08T11:00:00","publication_year":"2020","word_count":915,"keywords":["Go","machine learning","Transparency AI","AI","AWS","R","Ethical AI","Responsible AI","RAG","XAI","AI policy","explainable AI","GAN","Explainable AI","xAI","NITI Aayog"],"extracted_tech_keywords":["AI","machine learning","xAI","RAG","AWS","R","Go","GAN","explainable AI","XAI"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/does-india-have-the-infrastructure-to-implement-policy-recommendations-to-avoid-algorithmic-bias\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":14283,"title":"Impress.ai, Singapore based AI startup gets invested by ‘Javelin Startup-O Victory Fund’","content":"impress.ai, an automated platform for screening and assessing candidates with an AI powered chat interference has received an undisclosed amount of funding from the Javelin Startup-O Victory Fund, which is managed out of Singapore. This Singapore based artificial intelligence startup was founded in 2013, as Ideatory Pte. Ltd.,  and has since been a pioneer in innovation consulting and talent assessment for large enterprises. They have an impressive clientele list with the likes of DBS, UBS, Citi, Singtel, Maybank, CapitaLand, HDFC, MakeMyTrip, and ZS Associates in Singapore, Hong Kong, Malaysia, India and New York, included in the list. The team behind impress.ai which together holds Ph.Ds. in AI, experience in Industrial-Organisational Psychology and corporate sales, met at the Nanyang Technological University of Singapore. And since its launch, impress.ai has delivered impressive cost-savings and better candidate experience for its clients. A majority of banks in Singapore use impress.ai to positively overhaul their recruitment experience for candidates while improving overall efficiency and accuracy of the process. Utilizing the funds- With the investment from Javelin Startup-O Victory Fund, impress.ai plans to advance its mission to build a Virtual HR Assistant that makes enterprise recruiters more efficient and minimise unconscious bias. “impress.ai is able to generate an entirely new kind of self-annotating data sets on job candidates which allows our clients to build self-learning models of “good” hires. This enables the Virtual HR Assistant generated from impress.ai to identify top candidates with stunning speed and accuracy while getting better with time”, said Sudhanshu Ahuja, co-founder and CEO of impress.ai. He further added “This support is a validation of the impact impress.ai is having on banks, telecommunications and consulting companies. The Startup-O platform was a convenient and effective way that allowed us to participate online without any upfront dilution and provided fast results, by linking us up with seasoned experts and access to funds in a timely manner. It was a fantastic experience working with the Startup-O team. ” Commenting on the funding Anuj Jain, Co-Founder and CEO of Startup-O said, “impress.ai’s automated platform for potential candidates resides at the intersection of two high-growth technology sectors: Artificial Intelligence (AI) and HR technology. According to recently published data, corporate spend on AI technologies is expected to hit $47 billion by 2020. Likewise, 42% of firms plan to increase HR technology spend to replace current systems”. “We look forward to see impress.ai build on its impressive momentum that it has created over the last 6 months and continue to deliver on its promise to large organisations across the region”, he added. Startup-O is a Singapore-based online platform that helps curate tech startups from across Asia, and help them connect with strategic capital and global network of experts to expedite venture building. The winning startups selected from the online ‘Fasttrack’ program, held every quarter, are included in the Victory Portfolio to be funded by Javelin Startup-O Victory Fund. The Q1 2017 season saw winning startups getting funded from across A.I., Fintech, IOT, SaaS & Enterprise software domains respectively. “Javelin Wealth Management works with high net worth investors and family offices to build and manage properly diversified portfolios which invest in a broad range of assets. Investing in promising startups via an innovative fund like this is one example of the way in which we aim to show our clients exciting investment ideas. We look forward to investing in more high potential, technological startups like impress.ai,” said Steve Davies, CEO of Javelin Wealth Management, which has collaborated with Startup-O to establish the Javelin Startup-O Victory Fund to expedite success for both the startups and the investors in them.","excerpt":"impress.ai, an automated platform for screening and assessing candidates with an AI powered chat interference has received an undisclosed amount of funding from the Javelin Startup-O Victory Fund, which is managed out of Singapore. This Singapore based artificial intelligence startup was founded in 2013, as Ideatory Pte. Ltd.,  and has since been a pioneer in […]","categories":["AI News"],"tags":[],"author_name":"Srishti Deoras","publish_date":"2017-04-17T12:07:35","publication_year":"2017","word_count":598,"keywords":["Go","API","funding","artificial intelligence","AI","innovation","Aim","GAN","R","startup"],"extracted_tech_keywords":["AI","artificial intelligence","Aim","R","Go","API","GAN","innovation","startup","funding"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/impress-ai-singapore-based-ai-startup-gets-invested-javelin-startup-o-victory-fund\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10140563,"title":"When Mustafa Suleyman was in Bengaluru","content":"After Meta’s chief AI scientist Yann LeCun and NVIDIA CEO  Jensen Huang’s India visit last month, Mustafa Suleyman, CEO of Microsoft AI, also decided that it was high time for Microsoft AI to make its mark on the country’s larger audience. Suleyman showcased Copilot’s prowess. Speaking of AI companions, Suleyman, before visiting India for the first time, asked Copilot about the weather and GDP of Bengaluru and was quite surprised by it. At the Building AI Companions for India event, besides hosting a Cafe Copilot featuring an entire food menu designed by Copilot, and RaagaTrippin’ using Copilot as its band member to generate lyrics on stage, Suleyman and the Microsoft team demonstrated why it is important for India to build its own AI models in the era of agentic AI while also finding the right use cases. Puneet Chandok, president of Microsoft India and South Asia, highlighted how Microsoft is helping Indian companies build AI products. “Around 18,000 developers at Infosys have written seven million lines of code using GitHub Copilot,” said Chandok, while adding that Cognizant is a massive user of Copilot. Notably, HCLTech’s AI Force platform and Genpact’s AI Guru have been infusing Microsoft AI into their services, streamlining software and code development while also teaching people about AI. This is when Chandok invited Suleyman on stage to talk more about the upcoming Copilot products that are going to change how Indians communicate with AI companions. Building AI in India is Crucial Suleyman referred to Microsoft’s AI teams in India, especially those in Bengaluru and Hyderabad, as the strength of the company. According to him, these are where talented engineers and developers work on every layer of the company’s tech stack. “AI is going to put knowledge at everyone’s fingertips, synthesised, distilled and personally tuned to the way you want to learn and use information,” he said, highlighting AI’s potential to democratise knowledge, making it accessible across work environments and enabling informed decision-making. In a fireside chat with S Krishnan, secretary of the Indian government’s MeitY, Suleyman discussed Microsoft’s collaborative efforts with specialists from diverse fields to build AI systems that resonate with human values. Talking about the IndiaAI Mission and Bhashini, Krishnan said that, originally, they had envisioned building an LLM on their own. “Now that we are actually having a meeting, it may not be worth the effort of building an entire LLM on our own,” he added. It might be better to adapt AI systems to be practical in the Indian context and useful for specific sectors in areas that India seeks to adapt based on ground reality. This resonates with the idea of Adbhut India, or AI use case capital of the world. “We try to converse in 22 different Indian languages… Voice really is the ultimate way to make these tools accessible,” Krishnan pointed out. Suleyman’s response to this was slightly different from what was expected. Reflecting on DeepMind’s inception in 2010, Suleyman said, “Timing is everything… It’s critical that you get the timing right.” He said that India was one of Microsoft’s fastest-growing markets and has one of its strongest R&D teams globally. “I really feel that now is the right time to create these new models. All of the resources are now widely available. The APIs are brilliant. There are open-source models surfacing everywhere. And so it just feels like a very creative moment. And I’m excited to see so many startups and new businesses really experiment with this stuff,” Suleyman further said. From Pessimist to Optimist During the event, Suleyman also showcased the redesigned Microsoft Copilot, which now includes a conversational human-like voice, screen-viewing capabilities, and advanced reasoning functions. These innovations aim to transform daily user interactions with technology. He described the recent strides in AI as “the first step towards distilling intelligence into algorithmic constructs”. After co-founding GenAI startup Inflection in 2022, Suleyman and co-founder Karén Simonyan joined Microsoft in March 2024 following the acquisition of key Inflection AI team members. Surprisingly, he has always taken a critical approach towards AI, highlighting the dangers of the technology. But, lately, it has changed. “For too long, software has principally been utilitarian. My personal vision for AI has always been about how it can be a companion that can make each and every one of us feel more supported and smarter and more capable,” he said. Addressing the fear of job losses and regulations around AI globally, Suleyman explained that it’s super important to be as attentive and thorough about observing the potential upside as it is to really meditate on and deeply think about the potential downsides. “It won’t just be able to talk to you. It’s going to be able to use APIs. It is going to be able to search through databases. It will generate new programs from scratch. It is going to be able to get things done in the digital world on a large scale. That’s actually what is coming in the next few years, more than anything else,” he said. Suleyman said it’s time to be thoughtful and deliberate and not treat AI regulation as a taboo. He believes that nations need to start having an important dialogue around this. Giving examples of drones, he said that regulations are effective. “We don’t have drones flying around randomly all over the world with autonomous capabilities. We were deliberate and proactive and careful, and I think that’s just the approach that we have to take,” he reiterated.","excerpt":"Suleyman referred to Microsoft’s AI teams in India as the strength of the company.","categories":["IT Services"],"tags":["bangalore","Bengaluru","Mustafa Suleyman"],"author_name":"Mohit Pandey","publish_date":"2024-11-07T17:18:47","publication_year":"2024","word_count":910,"keywords":["Go","API","GenAI","agentic AI","Mustafa Suleyman","AI","ML","Git","Aim","Bengaluru","GitHub","R","bangalore"],"extracted_tech_keywords":["AI","ML","GenAI","agentic AI","Aim","R","Go","Git","GitHub","API"],"url":"https:\/\/analyticsindiamag.com\/it-services\/when-mustafa-suleyman-was-in-bengaluru\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10098821,"title":"Are Databricks and Snowflake Ferraris in a Toyota World?","content":"While opting for data architecture solutions, companies frequently fall into the trap of paying exorbitant prices for services they don’t need. A recent blog by Kieran Healey points out that companies like Databricks or Snowflakes are offering Ferraris when many companies could do their work with Toyota. Databricks and Snowflake are undoubtedly robust platforms that offer impressive capabilities. Snowflake’s partnership with NVIDIA and Databricks’ integration with the Spark Human API showcased their technical prowess and made it even bigger. Yet, such features often serve as marketing tactics rather than essential solutions, which companies end up paying instead of open source solutions. For example, instead of opting to pay the price of an LLM-based chatbot, most companies could effectively address their data challenges with simpler, more cost-efficient solutions such as a simple “press 1 to choose this option”. But when it comes to addressing data-related challenges without overspending, companies should adopt an anti-hype mindset. A person from Databricks suggested on HackerNews that though companies might be able to create their own Spark deployment, it will run much slower than how it runs on Databricks or its proprietary runtime. He further adds that a lot of businesses have other problems to solve and focusing on building DIY platforms is a horrible approach. Interestingly, none of this matters if you only have gigabytes of data as the company can use pretty much anything very cheaply and easily. It is just about companies that have terabytes or hundreds of terabytes of data. Open source vs commercial solutions On the other end, it seems easy to hop onto the open source solutions as well, given the cost-effective value that they are presented as. One side of the debate emphasises the financial advantage of open source solutions. Supporters highlight the fact that open source software is often free to use, suggesting that the cost savings alone make it a compelling choice. However, it is essential to be pointed out that while the software itself may be free, deploying, maintaining, and expertly managing open source solutions can incur significant costs. Paying skilled professionals to ensure proper deployment and upkeep can strain both time and resources. “Open source it may be. Free it is not. Paying an expert to correctly deploy an open source solution takes time and money,” said another user. This argument underscores the idea that simply adopting open source software isn’t a guaranteed money-saving solution without proper expertise and management. On the opposite side, commercial solutions such as Databricks and Snowflake might come with upfront costs, but offer comprehensive support, integration, and scalability that can be invaluable. These solutions often package features, support, and maintenance into a single offering, reducing the need for extensive in-house expertise. Furthermore, commercial solutions can provide a level of assurance and accountability that can be lacking in open source alternatives. Though you pay to change the parameters of the problem. This is a fundamental misunderstanding of how to get things done in a constrained environment. This viewpoint highlights the notion that the trade-off between open source and commercial solutions is about more than just cost—it’s about shifting the focus from technical challenges to non-technical ones. Funnily, it’s like saying no company needs a cloud provider but it definitely helps them focus on better things instead of building a data centre themselves. The Anti-Hype Approach In the debate over data platform choices, context and expertise play pivotal roles. While open source solutions can be powerful tools when implemented correctly, they require a skilled team to navigate potential challenges. Conversely, commercial solutions can mitigate many technical complexities, enabling organisations to concentrate on their core business goals. However, this often involves a trade-off between flexibility and vendor lock-in. Ultimately, there is no one-size-fits-all answer to the open source vs commercial debate in the context of data platforms. The decision depends on the unique circumstances of each organisation—its budget, existing expertise, scalability requirements, and risk tolerance. In the current age, when CEOs are being pushed to say generative AI by everyone, it might be easy to fall into the trap and overspend on over engineered solutions. It’s essential to scrutinise its applicability. Instead of focusing on novel technologies, companies should adhere to the age-old principle of delivering tangible returns on investments and CEOs are always looking for solutions that not only enhance but also generate profits.","excerpt":"Companies looking to cut costs shouldn’t rely on open source solutions either. It’s a juggle","categories":["AI Features"],"tags":["Databricks","Generative AI","Snowflake"],"author_name":"Mohit Pandey","publish_date":"2023-08-22T13:43:57","publication_year":"2023","word_count":722,"keywords":["Go","API","programming_languages:R","AI","Scala","Databricks","generative AI","GAN","Generative AI","R","Snowflake"],"extracted_tech_keywords":["AI","generative AI","Snowflake","Databricks","R","Go","Scala","API","GAN","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/are-databricks-and-snowflake-ferraris-in-a-toyota-world\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":36937,"title":"This Year’s Turing Award Swept Away By Yann LeCun, Geoffrey Hinton &#038; Yoshua Bengio","content":"Considered the Nobel Prize of computing, the prestigious Turing Award for 2018 has been rightfully bagged by Yann LeCun, Geoffrey Hinton & Yoshua Bengio who have been referred to as the Godfathers of AI for having developed some remarkable innovations in the field of artificial intelligence. At a time when the field was doomed and experts feared that ‘AI winter’ might be the final call of the much-talked technology, the trio believed in the potential of AI and brought to the world the innovations such as neural networks, self-driving cars, facial recognition, CNN and much more. In the mid-90s when the community had lost interest in AI and deep learning, the trio worked on bringing newer innovations in the field. But it was only in 2012 that the community started taking note of their work, especially in ImageNet. Presented by the Association for Computing Machinery, they have been awarded $1 million annual prize for their work in AI. The trio has worked independently and together laying the foundations of the field and contributed immensely in the advances of deep neural networks. Let’s take a quick glance into each of their awesome journeys: Yoshua Bengio: Having received his bachelor, masters and PhD from McGill University, he was a post-doctoral fellow at MIT and AT&T Bell Labs. Currently a professor at the University of Montreal, he had co-founded Element AI, a Montreal-based artificial intelligence incubator that turns AI research into real-world business applications. He also serves as a strategy adviser for Montreal-based legal tech startup Botler AI. Yann LeCun: LeCun is a leader in the fields of machine learning, computer vision, robotics and computational neuroscience. Currently a chief artificial intelligence scientist at Facebook AI Research, he is best known for his work on optical character recognition and computer vision using convolutional neural networks, convolutional nets and more. Geoffrey Everest Hinton: He is best known for his work on artificial neural networks and currently works for Google Brain and the University of Toronto. He is the co-author of most highly regarded papers on the backpropagation algorithm for training multi-layer neural networks. “Deep neural networks are responsible for some of the greatest advances in modern computer science. At the heart of this progress are fundamental techniques developed starting more than 30 years ago by this year’s Turing Award winners, Yoshua Bengio, Geoffrey Hinton, and Yann LeCun. By dramatically improving the ability of computers to make sense of the world, deep neural networks are changing not just the field of computing, but nearly every field of science and human endeavour,”  said Jeff Dean, Google Senior Fellow and SVP, Google AI to the Turing Award’s official website.","excerpt":"Considered the Nobel Prize of computing, the prestigious Turing Award for 2018 has been rightfully bagged by Yann LeCun, Geoffrey Hinton & Yoshua Bengio who have been referred to as the Godfathers of AI for having developed some remarkable innovations in the field of artificial intelligence. At a time when the field was doomed and […]","categories":["AI News"],"tags":[],"author_name":"Srishti Deoras","publish_date":"2019-03-27T13:31:51","publication_year":"2019","word_count":441,"keywords":["Go","machine learning","artificial intelligence","AI","neural network","innovation","computer vision","deep learning","CNN","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","deep learning","neural network","computer vision","R","Go","CNN","innovation"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/this-years-turing-award-swept-away-by-yann-lecun-geoffrey-hinton-yoshua-bengio\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10066628,"title":"Council Post: Global guidelines for building trustworthy AI","content":"Building public trust in AI has become critical more than ever. The public and private entities have a huge role in shaping guidelines to ensure fairness in AI systems. In this article, we will discuss international guidelines for building trustworthy AI and how to incorporate them. International guidelines Of late, we have seen a lot of efforts to chalk out international guidelines for trustworthy AI systems. Ethics Guidelines for Trustworthy AI by the European Commission is a good case in point. As per their guidelines, the requirements of trustworthy AI include: 1. Human agency and oversight, including fundamental rights, human agency and human oversight. 2. Technical robustness and safety, including resilience to attack and security, fallback plan and general safety, accuracy, reliability, and reproducibility. 3. Privacy and data governance, including respect for privacy, quality and integrity of data, and access to data. 4. Transparency, including traceability, explainability and communication. 5. Diversity, non-discrimination and fairness, including the avoidance of unfair bias, accessibility and universal design, and stakeholder participation. 6. Societal and environmental wellbeing, including sustainability and environmental friendliness, social impact, society and democracy. 7. Accountability, including auditability, minimisation and reporting of negative impact, trade-offs, and redress. From the POV of the data science industry, every client need not be bothered about all the requirements. In general, the requirements (3), (4) and (5) are common for any data science workflow and we limit our discussion to these points. Additionally, we will go into the tools and techniques to meet these requirements. Privacy and data governance There are privacy-preserving machine learning practices that integrate some sort of data encryption as part of the learning algorithm. However, in an industrial setting, we need more degree of freedom in choosing the right models. Hence, to preserve data privacy, model-agnostic approaches based on the general data anonymisation principles are preferred. Data anonymisation Data anonymisation is a process where private and confidential information in data is preserved through processing techniques. The objective is to safeguard the identity of the data point from being traced back or misused. Common data anonymisation techniques include data masking, pseudonymisation, generalisation, data swapping, data perturbation, synthetic data, etc. Data governance In general, data governance refers to principles, policies and standards that covers the entire enterprise data management. Given the exponential rate at which the data is accumulated, it is important to govern this data right at the stage of gathering to processing it by an AI system. In between, we need data transfer regulations, storage regulations and principles for data cleaning and refinement. From an organisation’s perspective, data governance addresses the questions like what data they have, where they reside and how it will be processed. From a client perspective, it addresses the question like what data is transferred, who gets access, the purposes for which the data is used etc. Transparency Transparency requirements include traceability, explainability and communication. Traceability refers to all the data science activities that are part of its workflow. This includes the data used, its cleaning, data labelling, feature transformation\/selection, dimensionality reduction, algorithms used and their configuration, evaluation metrics and deployment details. The idea is to open a completely transparent environment for the stakeholders of an AI system. Explainability refers to the practice of explaining a model decision to the users so that the user’s right to know about the modelling logic is preserved. This includes the popular XAI tools such as feature attributions, what-if analysis and counterfactual explanations. Communication refers to the practices of communicating and understanding the activities from a user perspective. Meaning, even a user with non-technical background shall be able to comprehend the activities summary. This include a layman explanation of the algorithm used, informed consent given to the user about their data being processed by an automated system and informing them about the data being collected. The communication shall also include creating awareness about the capabilities and limitations of an AI system. Diversity, non-discrimination and fairness This requirement includes the avoidance of unfair bias; accessibility and universal design; and stakeholder participation. Ethical AI plays a crucial role in satisfying this requirement and we will discuss in detail the techniques for bias assessment and model de-biasing. Bias assessment The bias assessment is concerned with the ethical aspects of a machine learning model. The idea is to check whether a deployed model shows bias in its prediction against any gender, race, nationality, etc. The data bias refers to the bias that already exists in the data. The bias assessment can be done at various levels – on the data directly, on the model predictions, on individuals or on the groups. The popular quantification metrics for bias assessment include: a) Disparate impact ratio (DIR) DIR measures the ratio of probability of predicting favourable situations by a machine learning model for unprivileged class to the privileged class. For example, the machine learning model may be predicting whether a person will be sanctioned a loan or not. The unprivileged class, say, based on gender is female and the privileged class is male. So ideally DIR value is expected to be one to satisfy fairness. However, in the industry standards, it is usually checked if the value falls in the range [0.8, 1.2]. If the value lies outside of this range, the model is considered as biased towards the privileged class. b) Equalised odds ratio (EO) A machine learning model is said to satisfy equalised odds ratio if it satisfy the following condition, P(R=1 | Y=y, D) = P(R=1 | Y=y, D) The probability of prediction of the favoured situation by a machine learning model shall be the same irrespective whether the data point belongs to privileged or unprivileged class. Compared to DIR, the EO investigates the learning capability of a model. In simple words, AO checks if the data points corresponding to both the classes have equal true positive rate and equal false positive rate. For example, let Y be an attribute that says whether the data point is qualified for a degree admission. Now, EO checks whether a university accepts or rejects the application in the same rate for both the classes, say, Americans and Asians. c) Average odds difference (AOD) This quantity is defined as the average difference in false positive rates and true positive rates for both the privileged and unprivileged class. The ideal value of this quantity is zero and, in this case, corresponds to both classes having equal benefit. Any value apart from that corresponds to one class having extra benefit compared to its counterpart. Apart from these quantities, there are a lot of other metrics that have been used, including statistical parity difference, Theil index etc. Bias mitigation methods In this section, I will discuss bias mitigation methods. These methods refer to techniques that nullify the bias that exists in a learned model or nullify the bias from the data itself. Commonly used tools for the model de-biasing can be classified as follows. a) Pre-processing techniques These methods modify the dataset itself to make it fairer. Examples are reweighing, disparate impact remover etc. In other words, bias mitigation management is absent during the model learning phase and it is done right at the data preparation stage. b) In – processing techniques These techniques create or modify the model to give fair predictions. That is, this technique corresponds to developing learning algorithms that are intelligent about the bias and rectify that during the learning phase itself. Examples are adversarial de-biasing, prejudice remover regularizer and exponentiated gradient. c) Post – processing techniques They modify the output so that the final decision is made fair, or these techniques manipulate the model output in a systematic manner to make fair decisions. Examples are Reject option classification and Threshold optimiser. Putting everything together Our idea is to form the components of trust requirements as checkpoints and to introduce corresponding evaluating mechanisms as checklists. The evaluators can make their validation through this well-defined process and thus an empirical scoring mechanism can be formulated on its basis. We can also add custom requirements to the list of checkpoints, for example, detecting model and dataset drift, initiating model re-training etc. A demonstration of the Human-AI trust platform is given in Table 1. The demo has only four requirements with the Human-AI trust score weightage chosen at random. The trust score can be put at our convenience on the scale of 1 to 10 or 1 to 100. However, we can include any number of requirements with custom checkpoints and checklists. A complete list of requirements, related documentations and checklists can be derived from the below shown table. RequirementCheck pointsCheck listsHuman-AI trust score weight-agePrivacy & Data governance·  Data anonymisation ·  Data governance·  Psuedonomyzation·  Generalization ·  Transfer regulation·  Storage regulation 5%  5%Transparency·  Traceability  ·  Explain-ability·  Data gathering·  Data labelling ·  Feature attribution·  Counterfactuals20 %  30 %Diversity, non-discrimination & Fairness·  Bias assessment   ·  Bias mitigation·  DIR·  EO·  AOD ·  Pre-processing·  In-processing·  Out-processing  15 %   15 %Production analysis·  Data drifts·  Target drifts·  Two sample KS test·  Chi squared test 10% A demo of Human-AI trust platform. Demo requirements, its checkpoints and checklist along with randomly assigned weightage for each case. To conclude, we have formulated a platform in which the abstract notion of trust is leveraged with the necessary qualifications to empirically evaluate the trustworthy AI systems. We briefly discussed the key stages of a typical data science workflow in which the platform ideas can be enabled and the ways to quantify the trust involved in those stages. Both the parties can mutually agree on the check points of data science workflow and a scoring mechanism is formulated that assigns weights for each score in every stage. The proposed platform provides a greater transparency in the entire life cycle of an AI system and can be easily adopted by customising the requirements. This article is written by a member of the AIM Leaders Council. AIM Leaders Council is an invitation-only forum of senior executives in the Data Science and Analytics industry. To check if you are eligible for a membership, please fill the form here.","excerpt":"To preserve data privacy, model-agnostic approaches based on the general data anonymisation principles are preferred.","categories":["Deep Tech"],"tags":["Trustworthy AI"],"author_name":"Shashank Shekhar","publish_date":"2022-05-10T10:00:00","publication_year":"2022","word_count":1671,"keywords":["data science","machine learning","TPU","AI","R","ML","RAG","Trustworthy AI","Aim","analytics","xAI"],"extracted_tech_keywords":["AI","machine learning","ML","data science","analytics","xAI","Aim","RAG","TPU","R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/global-guidelines-for-building-trustworthy-ai\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10170008,"title":"The Java Secret of Netflix","content":"Even though Java may not be the trendiest programming language currently, it is still a powerful choice for many, including Netflix. There have been advancements for developers to make AI-infused Java applications as well. Java is no more perceived as a legacy programming language. It has evolved over the years. And Netflix has pushed its stack forward—past JDK 8, beyond reactive programming, and new tech. Paul Bakker, a staff software engineer at Netflix, shared insights about how they are using Java in his talk on ‘How Netflix Uses Java’ at JavaOne 2025. “A lot of people apparently are not so happy with Java. Some people never want to watch Netflix again because it’s tainted by Java,” he said. Two Sides of Netflix, All Powered by Java Netflix has two sides of a tech stack powering its operations. The first powers its global streaming service, a high-traffic system spread across regions where one request can fan out into dozens. For this, Bakker mentions that their approach is that, when something fails, a retry could solve it. Image Credits: YouTube \/ @Java The second part involves Netflix Studios, which is enterprise-focused. These manage film production workflows — schedules, gear, people, and logistics. Here, data integrity is paramount. As Bakker put it, “If someone saves data about a movie being planned, that data absolutely needs to persist.” Image Credits: YouTube \/ @Java Bakker stressed that Netflix’s architecture for both the requirements remains essentially the same, all powered by Java. Upgrading The Backend Java Stack Java Development Kit is the software development environment that enables developers to create and run Java applications. The migration to modern Java wasn’t glamorous, but it was necessary. The company patched the legacy libraries, got the services to use JDK 17 and above to step up a notch. For anyone in the same boat looking to upgrade things, Bakker suggested, “Just fix it. It might look hard, but it is actually not.” He highlighted that they did nothing special here, but only upgraded. With that alone, Netflix experienced significant performance improvements. “We got about 20% less CPU time spent on garbage collection,” Bakker noted. He emphasised that it was a big win for Netflix to achieve a 20% performance boost on their machines without any tuning, which is often a challenging process. With JDK 21 and the introduction of generational ZGC, things improved even more: “We went from more than a second pause times to zero.” He highlighted, “Significantly fewer errors on our IPC calls.” Bakker was quite blunt about REST and said, “I don’t think you should use REST at all.” He put forth the argument that GraphQL has more benefits comparatively for Netflix. Bakker praised another new technology with the Java tech stack upgrade, virtual threads. He explained that they enable processes to run in parallel, giving them improved performance by default. During experimentation, they encountered a few issues with it. However, he mentioned that a fix arrived in JDK 24, and Bakker’s team is confident that it will get better. Relying on Spring Boot, An Open Source Java-based Framework The streaming company works closely with the Spring Boot team, an open-source Java-based framework. Bakker explained that Netflix uses standard Spring Boot plus a curated set of modules that hook into Netflix’s infrastructure. There’s security integration, dynamic configuration, observability, and more. He further explained that the migration from an older in-house framework to Spring Boot was a long, deliberate effort. Thousands of services were ported, helped along by automated tooling and migration guides. Now, most services use Spring Boot 3, and upgrades happen within days of each new release. They’ve worked closely with the Spring team to shape the GraphQL support, and are cautious in adopting Project Leyden to improve the backend’s performance in the near future. Bakker highlights that Spring Boot may not be flashy, but it’s familiar, battle-tested, and extensible—exactly what Netflix needs. Goodbye to Reactive Programming It may surprise some that Netflix, once the birthplace of RxJava, is now moving away from reactive programming. “Literally every API used to be Rx,” Bakker recalled. But over time, the complexity of debugging and maintaining reactive systems began to outweigh their advantages. “My hot take is that virtual threads combined with structured concurrency is going to completely replace reactive programming,” Bakker said. Today, Netflix only uses reactive libraries where absolutely necessary. For everything else, the stack has been standardised around synchronous code powered by virtual threads. As structured concurrency solidifies in upcoming Java versions, Netflix sees even less reason to keep reactive code around. “We can finally live our lives happily,” Bakker joked.","excerpt":"Netflix is ditching reactive programming and has upgraded its Java-powered backend to gear up for the present.","categories":["AI Features"],"tags":["Java","netflix"],"author_name":"Ankush Das","publish_date":"2025-05-15T09:00:00","publication_year":"2025","word_count":766,"keywords":["Go","API","programming_languages:R","AI","programming_languages:Java","programming_languages:Go","netflix","GAN","GraphQL","R","Java"],"extracted_tech_keywords":["AI","R","Go","Java","API","GraphQL","GAN","programming_languages:R","programming_languages:Java","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/the-java-secret-of-netflix\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10164639,"title":"Adobe Photoshop Now on iPhone, Coming Soon to Android","content":"Adobe has launched Photoshop on iPhone, with an Android version coming later this year. This marks the arrival of image editing and design applications on mobile platforms. The new mobile app and expanded web experience aim to provide accessible and intuitive interfaces for new and established creators. The iPhone app, made available from Tuesday, delivers Photoshop’s core imaging and design tools, including layering and masking, tailored for mobile devices. The mobile app and all-new web experience are now included in the current Photoshop plans. Along with this, Adobe also introduced a new Photoshop Mobile and Web plan for $7.99\/month or $69.99\/year, which enables access to Photoshop on the web and advanced editing features on mobile. These include generating similar and reference images, object select, magic wand, remove tool, clone stamp, and content-aware fill. “We’re excited to bring the limitless creative possibilities of Photoshop to mobile, making the app’s iconic image editing and design capabilities accessible for everyone from professional artists and designers to a whole new generation of creators trying Photoshop for the first time,” said Ashley Still, senior vice president and general manager for creative cloud at Adobe. The free mobile app provides direct integration with creative apps, including Adobe Express, Adobe Fresco, and Adobe Lightroom, and access to a vast library of free Adobe Stock assets. All current Photoshop paid plans include access to Photoshop on iPad and the web and will now include access to Photoshop on mobile. Adobe emphasises its commitment to responsible innovation, adhering to AI ethics. Content credentials are duly attached to the content generated with Firefly-powered tools in Photoshop mobile and web. Firefly is commercially safe and trained on the content that Adobe has permission to use. Expanding on its existing AI-powered offerings, Adobe recently introduced the new Firefly app, along with the launch of its video model, making it a multimedia generation tool for users. Adobe stated that the Firefly video model is the industry’s first commercially safe AI video generation model. It now powers ‘Generate Video (beta)’ in the Firefly application and ‘Generative Extend (beta)’ in Adobe Premiere Pro.","excerpt":"Adobe also introduced a new Photoshop Mobile and Web plan for $7.99\/month or $69.99\/year.","categories":["AI News"],"tags":["Adobe AI","Adobe Photoshop","iPhone"],"author_name":"Sanjana Gupta","publish_date":"2025-02-26T11:59:50","publication_year":"2025","word_count":348,"keywords":["programming_languages:R","AI","innovation","AI ethics","Adobe AI","Aim","Adobe Photoshop","iPhone","R"],"extracted_tech_keywords":["AI","Aim","R","AI ethics","innovation","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/adobe-photoshop-now-on-iphone-coming-soon-to-android\/","complexity_score":2,"technical_depth":6,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":57782,"title":"How Gmail’s Deep Learning, AI Tech Helps Thwart Cyberattacks","content":"Gmail accounts have become markedly vulnerable to malicious cyber activities. While corporate employees may be at greater risk of phishing scams, your personal inbox can also carry corrupted attachments that, when opened, could compromise your device. In fact, according to a Google blog post, malicious documents represent nearly 60% of the malware targeting Gmail users, and a majority of these differ from day-to-day. Thus, a focus area for Google has been to block these attachments from reaching your inbox in the first place. The company has been developing its deep learning tech to work in conjunction with existing artificial intelligence (AI) and machine learning (ML) models to strengthen and improve its document detection capabilities. How Gmail Malware Detection Works By using Gmail, you essentially agree to the company’s terms of service that allows it to scan and process your attachments. The blog post claims that Google scours through ‘300 billion attachments each week’ to filter out spam, organise your email into categories and most importantly, to prevent malicious documents from reaching your inbox. Since large volumes of data are involved, embracing technologies like AI and ML – that can analyse and learn faster than humans – becomes crucial. However, with techniques and tactics employed by hackers constantly evolving, existing security tools that leverage these technologies have to step up to keep up. To this end, Gmail has been relying on deep learning to build a new generation of document scanners. These are designed to work in conjunction with existing AI and ML tools to improve Gmail’s document detection capabilities. How does it work? The scanner uses a distinct TensorFlow deep-learning model trained with TFX (TensorFlow Extended) and a custom document analyser for each file type. With this, Gmail can interpret documents, identifying common patterns, deobfuscate content, and perform feature extraction. By combining different scanners that run in parallel with existing detection capabilities, it contributes to the final verdict of Gmail’s decision engine to block malicious documents. Assessing Impact First deployed at the end of 2019, Gmail has recorded a seemingly marginal increase in detection rate – about 10% – using the new deep learning-powered document scanner. However, it has registered a whopping 150% increase in the success rate of detection of what it calls “adversarial, bursty attacks”. Although promising, it does come with certain limitations as the company is still developing this technology. For one, it only scans Office documents. Also, the company has not addressed whether this deeper scanning method comes at the cost of users’ privacy. Outlook With cyberattacks getting more and more creative, we need to embrace technologies like AI, ML and deep learning to improve our cyber-readiness. What is more, with the rise of email, files and documents have become a popular way by hackers to trick users into installing malware on their devices. In fact, document-based malware can spread by merely viewing the wrong website, with the wrong browser installed on your system. Although documents have emerged as one of the most common ways of spreading malware across the internet, there are still ways to limit – or even prevent – this from happening. While we no longer solely depend on humans to keep a check of untoward behaviour online, there is much to be discovered on the potential of technology to thwart such attacks. As expounded here, these technologies enable computers to gain the ability to learn and make predictions based on patterns that emerge from past data. This indicates that AI is capable of reacting to unseen cyber threats faster and in a more effective way.","excerpt":"Gmail accounts have become markedly vulnerable to malicious cyber activities. While corporate employees may be at greater risk of phishing scams, your personal inbox can also carry corrupted attachments that, when opened, could compromise your device. In fact, according to a Google blog post, malicious documents represent nearly 60% of the malware targeting Gmail users, […]","categories":["AI Features"],"tags":["Cyberattacks","Gmail","learn ai","malware"],"author_name":"Anu Thomas","publish_date":"2020-03-02T12:00:00","publication_year":"2020","word_count":593,"keywords":["Cyberattacks","Go","malware","artificial intelligence","Gmail","AI","machine learning","ML","RAG","Aim","deep learning","learn ai","TensorFlow","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","Aim","TensorFlow","RAG","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-gmails-deep-learning-ai-tech-helps-thwart-cyberattacks\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":53879,"title":"What Is Antitrust Probe Against E-com Giants- Amazon &#038; Flipkart in India?","content":"A document published by the Competition Commission of India stated its ongoing investigation on the allegations made by a New Delhi trade group — Delhi Vyapar Mahasangh, on Amazon and Walmart-owned Flipkart of giving preferential treatment to a few sellers on their online marketplaces. As per CCI report, its informant members comprise many micro, small and medium enterprises traders who rely on the trade of smartphones and related accessories. And, many of such traders list their smartphones for sale on online marketplaces, regularly, to take the benefit of online distribution channels. What Are The Alleged Antitrust Issues? As per the document, there are several reported instances of vertical agreements between the platforms and the preferred sellers. And this has led to a disadvantageous situation for other non-preferred traders or sellers, who alleged that the majority of preferred sellers are influenced by Flipkart or Amazon, either directly or indirectly, which is a violation of the India’s Antitrust Act. Further, these platforms are also alleged of influencing the prices by providing several discounts as well as inventory (in the form of private labels for the B2B sector) to the sellers. The platforms also collect data on consumer preferences and allegedly use them to their advantage. The trade group also alleges that Amazon and Flipkart can cross-subsidise because of the huge amount of funding that was received from their investors, which has resulted in incentives that allow pricing below cost on their platforms, through their preferred sellers. This, according to the complaint, resulted in the creation of high entry barriers and high capital costs for any new entrant in the market. Resultantly, it gets difficult for the existing sellers to launch their own portals or marketplaces to compete with the giants. Additionally, it has been alleged that Flipkart gives deep discounts to a few particular preferred vendors, such as Omnitech Retail, which, in turn, negatively impacts non-preferred sellers. According to the complaint, Amazon also has preferred sellers on its platform including Cloudtail, which is a joint company between Amazon and Catamaran Ventures), and Appario Retail, which is a completely owned subsidiary of a joint venture that involves Amazon and another firm. Besides receiving deep discounts, it is alleged that sellers on Flipkart also receive a preferential listing on the platform, pushing the results of the non-preferred sellers further down in the search results without any reason. For example, the products sold by Cloudtail India and Appario Retail allegedly dominate the first few pages of search results whereas the products with the same ratings sold by non-preferred sellers are listed on later pages. The dual role of the platform gives rise to the concern of ranking biases that may be created by the platform as a discriminatory device. India Antitrust Probe: Using Data To Create Network Effects CCI also reported that due to huge market base and market power, the web marketplaces have vast repositories of data which allow them to have targeted advertisements based on consumer preferences. This, in turn, marginalises other competitors who are unable to capture the market due to lack of access to user data analytics. This has resulted in the creation of high entry barriers on account of network effects. Further, it is an accepted position that strong network effects in generating a source of market power for such platforms. Large numbers of users make an e-commerce platform more valuable, which further attracts more users, platforms benefit from a ‘positive feedback loop’, which gives rise to market power. On the sellers’ side, this may allow it to use such data to introduce its private label or boost its sale or that of its ‘preferred sellers’. Comparison With US Antitrust Investigation If you compare the antitrust investigations in the US against Facebook and Google vs the India’s antitrust probe, we can see that the regulators are waking up to the potential misuse of these platforms by the tech companies, for enterprise advantage over small organisations. The investigation went into action after regulators in the US received complaints of tech companies’ monopolistic actions and misuse of the platform to gather data of users for targeted advertising and manipulating search results to prioritise its subsidiaries and paid customers. In India, the issue is not on the mining of user data on platforms but more related to how smaller traders are getting impacted by giant retail platforms.","excerpt":"A document published by the Competition Commission of India stated its ongoing investigation on the allegations made by a New Delhi trade group — Delhi Vyapar Mahasangh, on Amazon and Walmart-owned Flipkart of giving preferential treatment to a few sellers on their online marketplaces. As per CCI report, its informant members comprise many micro, small […]","categories":["Global Tech"],"tags":["Amazon","antitrust","Flipkart","Walmart"],"author_name":"Vishal Chawla","publish_date":"2020-01-15T12:00:00","publication_year":"2020","word_count":724,"keywords":["Go","API","funding","programming_languages:R","AI","Flipkart","Amazon","programming_languages:Go","antitrust","analytics","Rust","GAN","Walmart","R"],"extracted_tech_keywords":["AI","analytics","R","Go","Rust","API","GAN","funding","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/what-is-indias-antitrust-probe-against-amazon-walmart-owned-flipkart\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":58270,"title":"Ads, Tweets And Vlogs: How Censorship Works In The Age Of Algorithms","content":"Over the last seven days, online media moguls Facebook, YouTube and Twitter have been in the news for stifling the content on their platforms. While Facebook is removing the campaign ads of Donald Trump, YouTube has reportedly halved the number of conspiracy theory videos. Whereas, Twitter took a resolve to tighten the screws on hate speech or dehumanising speech as they call it. YouTube Says Enough Conspiracies Theories In January 2019, YouTube said it would limit the spread of videos “that could misinform users in harmful ways.” YouTube’s recommendation algorithm follows a technique called Multi-gate Mixture Of Experts. Ranking with multiple objectives is really a hard task, so the team at YouTube decided to mitigate the conflict between multiple objectives using Multi-gate Mixture-of-Experts (MMoE). This technique enables YouTube to improve the experience for billions of its users by recommending the most relevant video. Since the algorithm takes into account the type of content as an important factor, classification of a video based on its title and the context of the video for its conspiratorial nature becomes easier. Ever since YouTube announced that it would recommend less conspiracy content and the numbers dropped by about 70% at the lowest point in May 2019. These recommendations are now only 40% less common. Twitter Has A Speech Cleansing Policy If your tweet is on the lines of any of the above themes, as shown above, you might risk losing your account forever. Last week, Twitter officially announced that they had updated their policies of 2019. https:\/\/twitter.com\/TwitterSafety\/status\/1235641453767991298?s=20 The year 2019 had been turbulent for Twitter. The firm’s management faced a lot of flak for banning a few celebrities such as Alex Jones for their tweets. Most complained Twitter had shown double standard by banning an individual based on the reports of the rival faction. Vegans reported meat lovers. The left reported the right and so on and so forth. No matter what the reason was, at the end of the day, the argument boils down to the state of free speech in the digital era. Twitter, however, has been eloquent about their initiatives in a blog post, they wrote last year, which also has been updated yesterday. Here is how they claim that their review system is promoting healthy conversations: 16% fewer abuse reports after an interaction from an account the reporter doesn’t follow.100,000 accounts suspended for creating new accounts after a suspension during January-March 2019 –– a 45% increase from the same time last year.60% faster response to appeals requests with their new in-app appeal process.3 times more abusive accounts suspended within 24 hours after a report compared to the same time last year.2.5 times more private information removed with a new, easier reporting process. Skimming over a million tweets in a second can be exhaustive, so Twitter uses the same algorithms that detect spam. The same technology we use to track spam, platform manipulation and other rule violations is helping us flag abusive tweets to our team for review With a focus on reviewing this type of content, Twitter has expanded teams in key areas and geographies for staying ahead and working quickly to keep people safe. Twitter, now, offers an option to mute words of your choice that would eliminate any news related to that word on your feed. https:\/\/twitter.com\/nonprofWHIT\/status\/1229964006330380289?s=20 Twitter has a gigantic task ahead as it has to find a way between relentless reporting of the easily offended and the inexplicable angst of the radicals. Facebook Polices The President From Cambridge Analytica to involvement in Myanmar genocide to Zuckerberg’s awkward senate hearings, Facebook had been the most scandalous of all social media platforms in the past couple of years. However, amidst all these turbulence, Facebook’s AI team kept on delivering with great innovations. They have also employed plenty of machine learning models to detect deep fakes, fake news, and fake profiles. When ML is classifying at scale, adversaries can reverse engineer features, which limits the amount of ground truth data that can be obtained. So, Facebook uses deep entity classification (DEC), a machine learning framework designed to detect abusive accounts. The DEC system is responsible for the removal of hundreds of millions of fake accounts. Instead of relying on content alone or handcrafting features for abuse detection in posts, Facebook uses an algorithm called temporal interaction embeddings (TIEs), a supervised deep learning model that captures static features around each interaction source and target, as well as temporal features of the interaction sequence. However, producing these features is labour-intensive, requires deep domain expertise, and may not capture all the important information about the entity being classified. Last week, Facebook was alleged for displaying inaccurate campaign ads from the President of the US, Donald Trump’s team. Facebook then started taking down the ads, which were categorised as spreading misinformation. When it comes to digital space, championing free speech is easier said than done. An allegation or a report need not always be credible and to make sure an algorithm doesn’t take down a harmless post is a tricky thing. Free Speech In The Age Of Algorithms Curbing free speech is curbing the freedom to think. Thought policing has been practised for ages through different means. Kings and dictators detained those who spread misinformation regardless of its veracity. However, spotting the perpetrator was not an easy task in the pre-Internet era. Things took a weird turn when the Internet became a household name. People now carry this great invention, which is packed meticulously into a palm-sized slim metal gadget. The flow of information happens at lightning speed. GPS coordinates, likes, dislikes and various other pointers are continuously gathered and fed into massive machine learning engines working tirelessly to churn profits through customer satisfaction. The flip side to this is, these platforms now have become the megaphone of the common man. Anyone can talk to anyone about anything. These online platforms can be leveraged for a reach that is unprecedented. People are no longer afraid of being banned from public rallies or other sanctions as they can fire up their smartphone and start a periscope session. So, any suspension from these platforms is almost like potential obscurity forever. The opinions, activism or even fame, everything gets erased. This leads to an age-old existential question of an identity crisis, only this time, it is done by an algorithm. A non-human entity[algorithm] classifying a humans act for being dehumanising Does this make things worse or better? Or should we bask in the fact that we all would be served an equal, unbiased algorithmic judgement? Machine learning models are not perfect. The results are as good as the data, and the data can only be as true as the ones that generate it. Monitoring billions of messages in the span of a few seconds is a great test of social, ethical and most importantly, computational abilities of the organisations. There is no doubt that companies like Google, Facebook and Twitter have a responsibility that has never been bestowed upon any other company in the past. “We also realise we don’t have all the answers, which is why we have developed a global working group of outside experts to help us think.” Twitter’s statement The responsibilities are critical, the problems are ambiguous, and the solutions hinge on a delicate tightrope. Both the explosion of innovation and policies will have to converge at some point in the future. This will need a combined effort of man and machine as the future stares at us with melancholic indifference.","excerpt":"Over the last seven days, online media moguls Facebook, YouTube and Twitter have been in the news for stifling the content on their platforms. While Facebook is removing the campaign ads of Donald Trump, YouTube has reportedly halved the number of conspiracy theory videos. Whereas, Twitter took a resolve to tighten the screws on hate […]","categories":["Deep Tech"],"tags":["facebook analytics","hpc data management system","social media analytics","twitter analytics","YouTube"],"author_name":"Ram Sagar","publish_date":"2020-03-09T13:00:04","publication_year":"2020","word_count":1249,"keywords":["twitter analytics","Go","hpc data management system","machine learning","AI","social media analytics","ML","Git","RAG","GAN","Aim","deep learning","facebook analytics","YouTube","R"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","Aim","RAG","R","Go","Git","GAN"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/twitter-youtube-facebook-censorship-algorithms-machine-learning\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":67416,"title":"ML Driven Epidemiological Model to Forecast COVID-19 Progression by Quantifying Interventions","content":"World Health Organization declared COVID-19 as a pandemic on March 11th, 2020.  The outburst of cases reported globally since then has severely grown, impacting the day-to-day life of both organizations and individuals across the world. It is now imperative to understand how long the pandemic might last and find effective ways to flatten the progression of COVID-19 cases. Research literature covers various statistical models (such as Gamma distribution, Negative Binomial distributions) and epidemiological models (such as SIR, SEIR) that are used to make predictions about the number of people infected with contagious diseases such as Ebola, SARS, MERS. However, the research on transmission rate, incubation period and other parameters that go into mathematical modelling of the spread of COVID-19 is still at a nascent stage with most of it yet to be peer-reviewed. Hence, standalone epidemiological models may not suffice to forecast the spread of Covid-19. Furthermore, these parameters may vary by region and intervention steps taken by various governments like social distancing, school closures, complete lockdowns and so on. In this article, we present an ensemble predictive solution that combines an epidemiological model and various Machine Learning (ML) techniques to forecast the impact of the COVID-19. The solution also incorporates the intervention steps taken by various governments to curb the impact. Introduction Frequently used methods for forecasting the growth of any infectious disease include the use of epidemiological models. We showcase two such commonly used epidemiological models in the figure below: SIR Model – As shown in Fig. 1, ‘S’ indicates the proportion of a population that is Susceptible, ‘I’ is the number of Infected persons and ‘R’ represents the number of Recovered patients. N is defined as the sum of ‘S’, ‘I’ and ‘R’, which is constant and taken as the total population of the forecast region. SIR model can be represented by the equations below: SEIR model – ‘S’ indicates the proportion of population that is Susceptible, ‘I’ is the number of Infected persons, ‘R’ is proportion of Recovered patients and ‘E’ indicates the individuals who have been exposed to the disease but are not infectious.  SEIR Model can be represented by the equations given below: In the equations for both models, the infectious rate is denoted by β and is representative of the probability of disease transmission between susceptible and infectious persons. Similarly, the incubation rate of the disease is represented by σ and the patient recovery rate is represented by γ. SIR model SEIR model Figure 1 – An illustrative depiction of the traditional a. SIR and b. SEIR models Covid-19 Disease Forecasting Solution In this section, we delve deeper into Genpact’s forecasting solution. The solution involves three main steps: Building the SIDR Model: A solver is built for the set of differential equations that represent the SIDR model (Susceptible-Infected-Dead-Recovered) using TensorFlow (TensorFlow Probability library) and initial parameters for the epidemiological model.Quantifying Interventions: The impact of interventions such as lockdowns and social distancing regulations, from government authorities are dynamically incorporated into the model. Error Minimization using ML: A multi-objective machine learning model is built using Adam’s Optimizer from TensorFlow to identify the cases, deaths and recovery curves that best fit the actual curves reported so far. SIDR Model The modified SIR Model (i.e. SIDR) enables the modelling of progression of COVID-19 cases using daily updated data of confirmed cases, deaths and recoveries as reported on the Johns Hopkins University’s (JHU) website. In addition, the parameters for each of these attributes were calculated using a Machine Learning (ML) model. Based on the data from JHU, the following attributes were included in the forecasting methodology: Number of Confirmed COVID-19 cases in each state in the USNumber of Deaths from COVID-19 in each state in the USNumber of Patients who Recovered after contracting COVID-19 in each state In the proposed model, we chose not to use SEIR as the epidemiological model because the data and research on the Exposed component (number of individuals who are exposed but not currently infectious) is relatively unknown. The solution instead uses ‘the SIR model with an added variable D’ – with D representing the number of Deaths from COVID-19. Accounting for deaths, the system of Ordinary Differential Equations (ODE) used in the solution are given below: The f in the system of differential equations (on left) signifies the percentage of infected people who will die from the infection. It can be shown mathematically that the system of ODEs on the left is equivalent to the system of ODEs on the right, which can thus be used to model deaths and recoveries independently. Figure 3 – COVID-19 ML Model Workflow using SIDR epidemiological Model The methodology then involves minimizing the simultaneous error for confirmed cases, deaths and recoveries between the time series projected using SIDR model and the time series from actual data.  Thus, the parameters – infection rate, death rate and recovery rate are trained using the machine learning model. Modelling the Impact of Lockdown and Interventions. Once the actual progression was modeled using the SIDR methods, the errors in the forecasted infected rate, death rate and recovery rate are minimized using a multi-objective minimization function. The interventions mandated by governments across the world (such as lockdowns, social distancing, school & bar closures etc.) impact the movement of people in a region. Since the decrease in mobility in a region is directly co-related to interventions, the coronavirus does not spread at the same rate as before said interventions are imposed. Thus, the model was designed to use location mobility data to forecast the progression of COVID-19 to dynamically account for government interventions. The model assumes that under lockdown, the rate of infection decreases exponentially with time. Two separate dates defining the Intervention start date and Lockdown start date were calculated from mobility data (% change in movement across the region). Threshold point for Intervention date was taken when mobility goes down by atleast 30%. The following equation was used to incorporate the effects of lockdowns and social distancing in the model: Where 1 & 2 are decay parameters trained by the model, ti & tl are dates at which intervention & lockdown starts Therefore, the term β in the SIDR model is not a constant, but rather, a time-dependent variable modelled by α1 and α2. The model can also backtrack through real world data to forecast scenarios of infection, death and recovery rates, had the lockdown measures in a state, started 2-3 weeks before the actual implementation. Additionally, the model can also forecast situations such as cases where lockdowns are lifted after 2-3 weeks, in the future. The solution projects several variables including – identification of the peak and recovery curves, confirmed cases, deaths and recovery rates, in each region. The solution was developed for each of the 50 states in the US and is expandable to all regions of the world. Results In this section, we present the spread of COVID-19 disease in one of the states (Connecticut) in the US. The model was trained on historical data (12th march – 25th April 2020) of confirmed cases, deaths and recoveries, to forecast the disease progression in the next 45 days (till 9th June 2020). The measure of accuracy used for training the model was the weighted Mean Absolute Percentage Error (MAPE) on confirmed cases, deaths and recoveries, thus acting as a multi-objective minimization function. For the unseen period – the future forecast for confirmed cases, resulted in one week out and two weeks out MAPE, of 12% and 15% respectively. Fig. 4 shows the curves for actual daily COVID -19 cases versus forecasted daily cases till 25th April 2020 to understand how forecasted cases are aligned with actual cases. Fig. 5, Represents the daily actual deaths versus forecasted deaths till 25th April 2020. Fig. 6, Represents the daily actual recoveries versus forecasted recoveries till 25th April 2020. Fig. 7, shows forecasted daily confirmed cases to represent the progression of disease in Connecticut. The curve shows that the disease will peak from mid-April to mid-May and gradually come down by the end of June. It also shows the deaths and recoveries, which are projected to increase towards the end of June due to incubation period of the virus and time spent in hospitals after getting an infection. Figure – 4 Figure – 5 Figure – 6 Figure – 7 Potential Application Areas The COVID-19 forecasting Model is highly useful in use cases such as: ICU Bed and Ventilator Forecast – Due to the rapidly increasing number of confirmed cases of COVID-19 and most governments across the world trying to ‘flatten the infection rate curve’, the availability of Intensive Care Unit beds has become critical. Forecasting the number of available ICU beds and ventilators will give hospitals an idea of how they can treat incoming patients and avoid high mortality rates and reduce the duration of hospitalization critical patients. The solution can help with identifying possible number of new infection cases in a region and forecast number of ICU\/ventilator requirements might arise. Employee Risk Assessment – The model can also be input with other publicly available resources for the virus’ spread and data related to the sites and locations of a company. This will help the firm predict the risk of an employee getting exposed to the coronavirus. Additionally, the model can be dynamically improved by adding employee information such as demographics and other external lifestyle data predict the impact on the firm’s sales, operational capacities.New Normal Business Outlook – The COVID-19 pandemic is drastically altering customer demand for products. The solution can be applied to forecast customer demand to predict new consumer behavior or how sales volumes will change when economic activity resumes in the recovery period post the pandemic. The solution can also help businesses evaluate best practices for the new normal ways of functioning in the post COVID-19 world, by developing a demand sensing model linked to external indicators. Authors Mohit Makkar: Data Science & Insights at Genpact | Bangalore, India mohit.makkar@genpact.digital Omprakash Ranakoti: Data Science & Insights at Genpact | Bangalore, India omprakash.ranakoti@genpact.digital","excerpt":"World Health Organization declared COVID-19 as a pandemic on March 11th, 2020.  The outburst of cases reported globally since then has severely grown, impacting the day-to-day life of both organizations and individuals across the world. It is now imperative to understand how long the pandemic might last and find effective ways to flatten the progression […]","categories":["Deep Tech"],"tags":[],"author_name":"AIM Media House","publish_date":"2020-06-16T15:00:00","publication_year":"2020","word_count":1674,"keywords":["data science","Go","API","machine learning","AI","ML","Git","GAN","TensorFlow","R"],"extracted_tech_keywords":["AI","machine learning","ML","data science","TensorFlow","R","Go","Git","API","GAN"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/ml-driven-epidemiological-model-to-forecast-covid-19-progression-by-quantifying-interventions\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":52492,"title":"How These Companies Are Revamping Their Cybersecurity Systems In 2020","content":"Even after numerous attempts at making their systems secure, big enterprises are still struggling to take care of this menace. The existing cybersecurity strategies are dated and are still open to breaches. In an attempt to stay competitive in the market, firms are rapidly scaling, thereby, unintentionally opening up holes. Consequently, today, no organisation can guarantee that it can safeguard its database and network from breaches. As per a report, there were over 3,800 publicly-disclosed breaches and 4.1 billion exposed records in just the first half of 2019. Irrespective of the companies’ efforts, the breaches are continuously increasing. To avoid such mistakes and fill these glaring gaps, many enterprises are now revamping their strategy in 2020. Various companies have committed to further enhance their cybersecurity team for fortifying cybercrimes. 1. Infosys In April, Infosys already beefed up its security to safeguard itself from potential cyber attack. And the firm has further planned to enhance security in 2020 by expanding and reskilling its team. Since the firm is also a security solution provider, it was imperative for them to mitigate the rising concern among its clients as well. We will be launching two new CDCs soon, one each in Indianapolis, the US, and Mysore, in southern India, said Vishal Salvi, chief information security officer and head of cybersecurity at Infosys. “Investment in cybersecurity controls are on the rise year-on-year, and that is because organisations are considering cybersecurity investment very strategic for their current and future business,” he added. Infosys is mainly focusing on reskilling its team in identity and access management, infrastructure security, security information and event management, security orchestration, automation and response. To quickly achieve this, Infosys has collaborated with 50 plus global technology partners in cybersecurity to create a constant learning and development calendar. 2. BlackBerry BlackBerry is no more a mobile company, as John Chen, chairman and CEO of BlackBerry, has been revamping the company’s business since 2013. He is working to make it a data security firm, which is evident with the acquisitor of California-based Cylance, a threat prediction and prevention company, for $1.4 billion in February 2019 for providing end-to-end mobility services. “Security is the key for BlackBerry; that is why we are acquiring and also collaborating with governments for providing better security,” said Chen. “We are revamping by integrating Cylance technology both on the server-side and on edge for delivering products that are AI-enabled.” The company is looking to penetrate further deep into the security with superior cybersecurity solutions for helping business fend threats in 2020. 3. Google Cloud has become an essential aspect for firms for their business growth by providing superior products. Therefore, it has drawn the attention of hackers towards public cloud databases. According to the report, by 2020, data over one-third of the data will live or pass through the cloud. Public clouds are now the hot spot for hackers to penetrate and gain access to a plethora of data. Consequently, Google has partnered with multiple cybersecurity firms to protect its cloud platform from getting hacked. “We understand that our clients have dedicated tools and strategic relationships with leading security vendors. But we want to meet you where you are, allowing you to preserve your investments, as well as facilitate with functionality that you can’t get on other clouds. Consequently, we work closely with partners in the security industry for helping you better secure applications and data,” says Sunil Potti, VP of engineering at Google. 4. Genpact “The biggest problem in India is that most firms go to market without embedding security in their products or services,” says Sriram Lakshmanan, vice president, cybersecurity assurance, Genpact. The Personal Data Protection Bill (PDPB) got the nod of the Union Cabinet in December, which would require businesses to comply with the provisions related to data security. Consequently, the firm is expected to revamp the security strategy to stay in compliance with the new laws. “Digital ethics officers will be responsible for implementing ethical frameworks to make decisions comply with PDPB and GDPR. This includes security, bias, intended use, and built-in governance,” says Sanjay Srivastava, chief digital officer, Genpact. 5. Facebook Facebook is on the limelight since the Cambridge Analytica for various privacy lapses that it was guilty of. In November the platform again exposed a couple of tech employees profile picture. Many app developers got inappropriate access to profile information. Another instance came to light this month, an employee at cybersecurity firm Trend Micro stole personal information’s of about 70,000 of Facebook’s customers and later used it to scam customers. Such regular instances have led Mark Zuckerberg, founder of Facebook, realise the implications of data sharing. “Frankly we currently don’t have a good reputation for building privacy-protective services,” said Zuckerberg. He went on to say that they are devising a plan to reworking on privacy-focused approach over the next few years.","excerpt":"Even after numerous attempts at making their systems secure, big enterprises are still struggling to take care of this menace. The existing cybersecurity strategies are dated and are still open to breaches. In an attempt to stay competitive in the market, firms are rapidly scaling, thereby, unintentionally opening up holes. Consequently, today, no organisation can […]","categories":["AI Features"],"tags":["Cybersecurity"],"author_name":"Rohit Yadav","publish_date":"2019-12-24T10:30:01","publication_year":"2019","word_count":807,"keywords":["Go","API","AWS","AI","cloud_platforms:AWS","Git","automation","cloud_platforms:Google Cloud","GAN","Cybersecurity","R"],"extracted_tech_keywords":["AI","AWS","R","Go","Git","API","GAN","automation","cloud_platforms:AWS","cloud_platforms:Google Cloud"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-these-companies-are-revamping-their-cybersecurity-systems-in-2020\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10171827,"title":"The Wild Ways China Smuggles NVIDIA Chips","content":"For Chinese AI companies, US government restrictions on NVIDIA chips only kick in if they give up. These companies have once again found a workaround to obtain NVIDIA hardware. According to the Wall Street Journal, engineers from Chinese AI companies are heading to Kuala Lumpur, Malaysia, with hard drives packed with instructions and data to train AI models. They then utilise the NVIDIA chips available in Malaysian data centres to train the model and return it to China. The report added that engineers decided it would be faster to fly physical hard drives with data into Malaysia than to transfer data online. This unnamed Chinese company also reportedly registered an entity in Malaysia to avoid raising suspicions. The aforementioned method is deemed an alternative to directly smuggling NVIDIA hardware in China, which was made difficult due to recent crackdowns from multiple government agencies. For instance, Malaysia’s semiconductor industry closely monitors the movement of chips across borders, aligning with the US push for stricter regulations on suspected stock movement through the Southeast Asian country. Malaysia and Singapore are listed as Tier 2 countries in the US government’s Framework of Artificial Intelligence Diffusion, which permits limited chip exports without a license. China is categorised in Tier 3, where exports are banned or highly regulated. However, this existing framework will be replaced with a more straightforward approach, moving away from the tier-based system. Moreover, authorities in Singapore and the US are investigating allegations that NVIDIA chip servers worth $390 million sent to Malaysia were fraudulently redirected to China. Furthermore, last month, a US lawmaker introduced a bill to address chip smuggling. The bill would require companies like NVIDIA to include technology to verify the location of their chips before exporting them. Live Lobsters, Baby Bumps and Toys The regulations, scrutiny, and crackdown were implemented in response to Chinese AI companies employing multiple illicit methods. The Centre for a New American Security (CNAS), a US-based think tank specialising in national security issues, estimated that out of the 22 “notable” AI models developed in China as of 2025, only two were trained with Chinese chips. “Most of the Chinese chip sellers interviewed in these [media] reports confirm that they work with multiple distributors, use shell companies based overseas, and employ simple tactics to avoid detection, such as relabeling shipments as tea or toys,” the CNAS report read. Last year, The Information reported that an unnamed electric appliance company in eastern China had ordered hundreds of servers equipped with NVIDIA’s H100 GPUs, worth $120 million, with the help of a chip broker in Malaysia. This broker had assisted the Chinese firm in setting up a shell company in Malaysia and renting space in a data centre. The report highlighted many such cases of Chinese companies illegally obtaining NVIDIA chips. Moreover, as per a Reuters report, Chinese universities and research institutes had acquired GPUs from resellers in the country. These resellers purchased servers from Super Micro Computer and Dell, which were equipped with NVIDIA chips. Besides, Amazon-backed Anthropic, the AI lab that built the Claude family of models, said that smugglers in China hide processors in “prosthetic baby bumps” and pack GPUs alongside live lobsters. Singapore accounts for a considerable part of NVIDIA’s reported revenue, despite very few physical hardware shipments to the country. Between August and October 2024, Singapore contributed 22% to NVIDIA’s sales revenue. However, Bloomberg reported that actual physical shipments to Singapore account for only 1% of NVIDIA’s total revenue, showing a significant gap between where NVIDIA records sales and where its hardware is actually delivered. In Q1 2025, Singapore accounted for 20% of NVIDIA’s billed revenue. Citing publicly available information, the CNAS report said there were 1 lakh H100 GPUs as of December 2024. NVIDIA and Trump’s AI Czar Deny Chip Smuggling While Anthropic accused Chinese companies of smuggling GPUs, NVIDIA reportedly denied such claims. An NVIDIA spokesperson said American firms like Anthropic should focus on innovation and rise to the challenge, “rather than telling tales about large, heavy and sensitive electronics somehow being smuggled in baby bumps or alongside live lobsters”. Furthermore, the ‘AI czar’ appointed by US President Donald Trump, David Sacks, also refuted the practicality of such smuggling mechanisms. In an interview at the AWS Summit last month, Sacks said, “We talk about these chips like they can be smuggled like diamonds in a briefcase—that’s not what they look like. These are server racks that are eight feet tall and weigh two tonnes.”“They don’t magically walk out the door,” he added. That being said, despite Chinese companies reportedly using various techniques to work with NVIDIA chips, the company isn’t backing down either. The US government and NVIDIA have been engaged in an ongoing cycle of restrictions and workarounds. The government has repeatedly blocked NVIDIA from selling high-end AI chips, prompting the company to develop less powerful versions that comply with the new limits. This pattern has repeated multiple times. Currently, NVIDIA is reportedly developing a new chip that is less powerful and less expensive than its H20 chips, which are the latest to face US government restrictions.","excerpt":"Chinese AI engineers are flying to Kuala Lumpur with hard drives of training data. They use NVIDIA chips in Malaysian data centers to train models, then take the results back to China.","categories":["Global Tech"],"tags":["AI (Artificial Intelligence)","China"],"author_name":"Supreeth Koundinya","publish_date":"2025-06-17T11:37:54","publication_year":"2025","word_count":847,"keywords":["Anthropic","Go","artificial intelligence","AWS","AI","cloud_platforms:AWS","innovation","llm_models:Claude","Aim","AI (Artificial Intelligence)","R","China"],"extracted_tech_keywords":["AI","artificial intelligence","Anthropic","Aim","AWS","R","Go","innovation","llm_models:Claude","cloud_platforms:AWS"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/the-wild-ways-china-smuggles-nvidia-chips\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10140006,"title":"Anthropic is Now Big Tech’s Favorite Child—OpenAI, Not So Much","content":"Not so long ago, OpenAI and Microsoft seemed unbeatable. Now, however, their competitors are catching up. Anthropic’s Claude leads in models, Microsoft-backed GitHub has added Claude to Copilot and Amazon has added Claude to Q Developer (Code Whispherer). Additionally, Meta’s Llama, now rivals ChatGPT in users and is a household name for open source and beyond. At the same time, OpenAI’s initial o1 feedback has been mixed. OpenAI CEO Sam Altman has referred to the model as “GPT-2 for reasoning”, while calling it “deeply flawed”. Anthropic vs OpenAI OpenAI seems to be focusing more on voice features in addition to its reasoning capabilities, notably catering to its end users. In contrast, Anthropic appears to prioritise the engineering and API side. OpenAI recently introduced expressive, steerable voices for speech-to-speech experiences and reduced costs through prompt caching. Text inputs were discounted by 50%, and audio inputs by 80%. The move has made advanced real-time AI more accessible. (Source: Morgan McGuire) Meanwhile, Anthopric is also experimenting with voice dictation on Claude mobile app with select users, with up to 10 minutes of recording. OpenAI is also facing growing competition from Anthropic in AI coding. The company is advancing its own tools to handle complex coding tasks and automate actions like code generation, though Anthropic’s recent launch of ‘computer use’ has given it an edge.“AI coding can’t fully replace engineers yet and needs ‘some coaching’,” shared Anthropic co-founder Daniela Amodei. She also noted that their Claude model has significantly boosted productivity, potentially reshaping hiring strategies. As both companies push for AI dominance, OpenAI’s upcoming releases could impact the market, potentially challenging products like GitHub Copilot, Cursor, and other coding assistants. Perfects User Experience Anthropic is enhancing users’ interaction with AI by focusing on screen navigation. Last week, the company introduced features enabling the AI to control computer screens, allowing it to browse the web or type on behalf of users. ‘Computer use’ is an experimental public beta feature through which Claude 3.5 Sonnet can now navigate computer interfaces in a manner similar to human users. This means the AI can view a screen, move a cursor, click buttons, and type text, allowing it to perform various tasks. OpenAI: \"Look at our new AI-powered internet search! with a fine-tuned model!\"Perplexity: \"Look at our AI-powered internet search combining multiple models!\"Google: \"We built a custom pipeline for AI summaries!\"Microsoft: \"We did AI search built around Bing years ago!\"Claude: pic.twitter.com\/2oNkCapmUu— Ethan Mollick (@emollick) November 1, 2024 There’s more. Anthropic has also launched its analysis tool within Claude. It allows users to perform data analysis directly in the platform by running JavaScript code. With this, Anthropic has made a mark in the field of AI user experience. This feature, also available in preview mode, enables Claude to handle complex tasks like data cleaning and in-depth analysis from CSV (comma-separated values) files. Designed to help teams across functions by delivering precise insights, it aims to help marketers analyse customer behaviour and finance teams create dashboards. This itself feels like AGI, Anthropic is far better than OpenAI.pic.twitter.com\/Qe9r18osZu— Uttkarsh Singh (@Uttupaaji) October 24, 2024 Last month, Anthropic made Claude Artifacts available to all users on iOS and Android, allowing anyone to easily create apps without writing a single line of code. Anthropic is undoubtedly taking how humans interact with AI to the next level. ‘While Anthropic has built something that still requires a computer as an interface, it is likely that in the future we will move away from screens and interact with AI agents using a new kind of device or interface.’ — AIM OpenAI has also not given up, yet. At OpenAI DevDay 2024 in London, Romain Huet, head of the company’s developer experience, showcased the o1-preview demo, revealing both promise and setbacks. After three attempts, the Swift maps app coding stalled, but, later, a demo presenting o1-mini alongside Cursor successfully controlled a drone, complete with an impressive backflip. Nice! The drone flies and does a backflip pic.twitter.com\/GBOGGuywIv— Tarek Ayed (@tarekayed00) October 30, 2024 OpenAI also released an advanced voice feature on the ChatGPT desktop app and not long ago, it released a feature that now lets users search for content from their previous conversations on ChatGPT. In addition, OpenAI launched its search engine yesterday. ChatGPT search now offers improved web search capabilities for timely, accurate answers, blending natural language interaction with up-to-date data in sports, news, stock quotes, and more. Masters Voice Features Recently, OpenAI launched a Realtime API for developers, enabling them to add advanced voice and natural speech-to-speech conversation features to their applications. This API enables companies to build voice-powered customer service systems that can handle complex tasks, from booking trips to guiding users through software. Surprisingly, Anthropic’s Claude has little to no developments in that area. OpenAI is certainly killing it. This new API is an extension of ChatGPT’s Advanced Voice Mode with sight, released just a few days before the API update. It provided six distinct voices, along with smooth audio input and output options. For example, users can now ask ChatGPT for recipe ideas by showing a photo of their fridge or get help with a math problem by sharing a picture of the problem itself. This update is similar to Gemini Live, Google’s conversational AI assistant. It claims to help its users plan events, ask for advice, discuss historical events, and even explore new local topics and ideas. Needless to say, OpenAI’s API update is in contrast to traditional methods that rely on multiple models for speech transcription and response. It connects to OpenAI’s latest GPT-4o model using a WebSocket, which allows developers to manage functions and respond based on user requests. Wellness company Healthify Me was one of the early adopters of this, using the API for real-time nutrition coaching through their AI coach, Ria. It uses OpenAI’s GPT-4 Turbo, and the machine learning model for speech called Whisper. Currently, the Realtime API is priced based on text and audio tokens. Audio input is priced at $100 per million tokens, and output is priced at $200 per million tokens. OpenAI has also built strong safety features into the API, including automated abuse detection and human review mechanisms. OpenAI plans to extend the API’s capabilities in the future. It aims to support additional modalities like video and visual inputs. Contrary to Anthropic’s Claude Sonnet 3.5 Artifacts, OpenAI recently unveiled canvas. This was a new interface for working with ChatGPT on writing and coding projects. Not surprised. OpenAI’s new canvas interface for ChatGPT falls short against Anthropic’s Claude Sonnet 3.5 for coding, with developers consistently favouring Claude’s capabilities in generating, debugging, and learning code quickly. “On-demand software is here,” said Joshua Kelly, CTO of Flexpa, who created a custom app in seconds with Claude, underscoring how Claude Artifacts empowers users to quickly develop tailored apps and pushes forward the vision of everyone as a potential app developer. Meanwhile, GitHub also set a new standard in the coding arena with its multi-model lineup—Claude 3.5 Sonnet, Gemini 1.5 Pro, and OpenAI’s o1-mini and o1-preview. This has brought unmatched versatility and developer choice across VS Code, Xcode, and beyond, positioning GitHub as the ultimate toolkit for today’s code-generation needs. Money Talks While both OpenAI and Anthropic have seen significant user growth compared to last year, their revenue generation strategies reveal vastly different approaches. According to the above analysis, most of OpenAI’s revenue growth arises from paid subscriptions for its AI models, like ChatGPT, while Anthropic earns the majority of its income through API services. Innovations like real-time API and voice and speech control soared OpenAI’s revenue to $4 billion in 2024, a 580% increase from last year. Their forecasted earnings are even more impressive, with projections suggesting they could reach $11.6 billion in 2025. For Anthropic, the leap in usability has contributed to revenue growth, reaching $1 billion this year, a 1,000% increase. Like OpenAI, much of this revenue is generated from API access, catering to developers looking for seamless AI integration.OpenAI would have barely survived without Microsoft. The tech giant’s deep-rooted partnership with OpenAI, including over $13 billion invested to date, now brings anticipated quarterly losses of $1.5 billion. Microsoft attributes this cost to its equity stake in OpenAI as the latter faces mounting expenses to sustain its rapid growth trajectory.","excerpt":"OpenAI is focusing on expanding its voice features, while Anthropic is working to improve its user interface dramatically.","categories":["AI Features"],"tags":["Anthropic","OpenAI","Revenue Growth","Voice AI"],"author_name":"Sanjana Gupta","publish_date":"2024-11-01T11:40:05","publication_year":"2024","word_count":1375,"keywords":["Anthropic","ChatGPT","machine learning","TPU","OpenAI","AI","GPT-4o","Revenue Growth","ML","Aim","Claude 3.5","Voice AI"],"extracted_tech_keywords":["AI","machine learning","ML","GPT-4o","ChatGPT","OpenAI","Claude 3.5","Anthropic","Aim","TPU"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/anthropic-is-now-big-techs-favorite-child-openai-not-so-much\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":37405,"title":"Former MobME CEO Sathya Kalyanasundaram Joins Experian To Head Its India Operations","content":"The world’s leading global information services company, Experian has announced the appointment of Sathya Kalyanasundaram as Country Managing Director, Experian India. Sathya brings with him over 20 years of experience from Consulting, Finance and FinTech leadership roles. Prior to joining Experian, Sathya was the CEO of MobME Wireless Solutions Limited, a technology conglomerate and fintech solution provider in India. Prior to MobME, he led the India operations of global MNCs such as Scientific Games (a USD 3 billion diversified gaming company), and Texas Instruments (a USD 15 billion leader in semiconductors). He was also associated with the Confederation of Indian Industries (CII) as Senior Member – Economic Affairs Panel. “Financial institutions investing in data and analytics have been changing the narrative on providing differentiated experiences for their consumers. I believe Experian India, being the pioneer in the field of data analytics and decisioning, has a tremendous role to play in this and I look forward to driving our growth strategy and taking the company forward as we embrace the opportunities ahead,” said Sathya Kalyanasundaram. Furthermore, as the Country Managing Director at Experian India, Sathya will be responsible for driving further growth of the overall India operations of Experian with a focused vision on vertical market strategy and strategic clients. He will leverage Experian’s global strength in leading the strategic development of Experian solutions for India, aligning with the company’s global product and industry leaders in Decision Analytics, Credit Services, Data Quality and Consumer Services. Also, Experian India’s leadership team will report into him to implement the organisation’s business plans. “We are delighted to welcome Sathya to the Experian family. His extensive experience across a variety of sectors will ensure that we take Experian India to the next level,” said Ben Elliott, CEO Experian Asia Pacific. “We are confident that his appointment to lead Experian will strengthen our presence in India, with a focus on building innovative solutions for India’s consumers.”","excerpt":"The world’s leading global information services company, Experian has announced the appointment of Sathya Kalyanasundaram as Country Managing Director, Experian India. Sathya brings with him over 20 years of experience from Consulting, Finance and FinTech leadership roles. Prior to joining Experian, Sathya was the CEO of MobME Wireless Solutions Limited, a technology conglomerate and fintech […]","categories":["AI News"],"tags":[],"author_name":"Harshajit Sarmah","publish_date":"2019-04-08T09:06:49","publication_year":"2019","word_count":320,"keywords":["programming_languages:R","AI","RAG","data quality","analytics","GAN","R"],"extracted_tech_keywords":["AI","analytics","RAG","R","data quality","GAN","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/former-mobme-ceo-sathya-kalyanasundaram-joins-experian-to-head-its-india-operations\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10143276,"title":"China’s DeepSeek Gets a Model Upgrade with V2.5-1210","content":"DeepSeek, a Chinese AI research lab backed by High-Flyer Capital Management, has launched V2.5-1210, the final model in its V2 series. The model introduces Internet Search capabilities for real-time answers and excels in tasks like math, coding, writing, and roleplay. The model is accessible on chat.deepseek.com. Users can toggle the Internet Search feature on the website for real-time responses or integrate the model via Hugging Face. Developers can explore its capabilities and build applications using the open-source release. The release aims to meet diverse user needs, enhance productivity, and provide a versatile AI tool for work and life applications. With this tool, the company is directly competing with the likes of OpenAI’s ChatGPT and Perplexity’s Search engine. However, users on X claim that Deepseek is still in its early stages and lacks widgets. However, it is capable of managing basic tasks, such as real-time search-based prompting and reviewing 50 different sources when querying for news. “As V2 closes, it’s not the end—it’s the beginning of something greater. DeepSeek is already working on next-gen foundation models, and the DeepSeek V3 series will be released in the future to push boundaries even further. Stay tuned!,” read their post on X. Last month, the company released DeepSeek-R1-Lite-Preview, a reasoning AI model rivaling OpenAI’s o1. It matches OpenAI’s performance on benchmarks like AIME and MATH, offering step-by-step “chain-of-thought” reasoning for transparency. The model improves with longer reasoning outputs, challenging traditional AI scaling laws by using additional processing time for complex tasks. Available via DeepSeek Chat with a 50-message daily limit, it faces regulatory restrictions on politically sensitive topics. DeepSeek plans to release open-source R1 models, increasing competition with Chinese tech giants like ByteDance, Alibaba, and Baidu. Alibaba’s rival Qwen2.5-Turbo supports massive context lengths.","excerpt":"“As V2 closes, it’s not the end—it’s the beginning of something greater.”","categories":["AI News"],"tags":["DeepSeek"],"author_name":"Aditi Suresh","publish_date":"2024-12-11T15:05:12","publication_year":"2024","word_count":289,"keywords":["ChatGPT","Hugging Face","TPU","OpenAI","AI","AWS","DeepSeek V3","DeepSeek","Aim","foundation models","R"],"extracted_tech_keywords":["AI","foundation models","ChatGPT","OpenAI","DeepSeek V3","Aim","Hugging Face","AWS","TPU","R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/chinas-deepseek-gets-a-model-upgrade-with-v2-5-1210\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10165122,"title":"Salesforce Launches AgentExchange Marketplace for AI Agents","content":"Salesforce on Tuesday launched AgentExchange, a marketplace for Agentforce, allowing partners, developers, and the Agentblazer community to build and monetise AI components. The platform supports businesses in the $6 trillion digital labour market. AgentExchange includes over 200 partners, such as Google Cloud, Docusign, and Box, providing prebuilt solutions for AI agents. Businesses can discover, test, and purchase prebuilt actions, topics, and templates on the marketplace or within Salesforce’s agent-building tools. AgentExchange is now live at AgentExchange.Salesforce.com. Prompt templates and topics can be listed and packaged immediately, while agent templates will be available for listing in April 2025. AgentExchange builds on Salesforce’s AppExchange, which has facilitated over 13 million app installations. The new platform offers rigorously reviewed agent components that improve efficiency and automation across industries. Mark Stewart, CEO of Goodyear, highlighted the potential benefits of the platform. “Accelerating our speed of execution is critical to Goodyear’s ability to deliver for our customers and maximise our end-to-end value proposition. We’re excited about the potential of the ready-to-use solutions from AgentExchange to enhance our speed, efficiency, and customer experience,” said Stewart Several partners have developed AI solutions for AgentExchange. Google Cloud builds Agentforce agents leveraging Google Search and Vertex AI to provide real-time data insights. Box enables AI agents to extract insights from unstructured data using natural language processing. Docusign facilitates agreement generation, signature routing, and status tracking. Workday streamlines employee workflows, including onboarding and benefits management. Brian Landsman, EVP & GM, global business development & partnerships at Salesforce, compared the platform’s role to AppExchange. “With AgentExchange, we’re opening up Agentforce for partners, startups, and Agentblazers to participate in the digital labour market and build agentic AI on Salesforce.” AgentExchange offers multiple agentic components. Actions are prebuilt integrations that expand AI agent capabilities. Prompt templates provide reusable prompts, ensuring consistent interactions. Topics group actions to refine agent behaviour, and agent templates deliver comprehensive solutions combining multiple components. Alice Steinglass, EVP & GM of platform, integration, and automation at Salesforce, emphasised the impact on businesses. “Now our developer community can directly tap the expertise of our partner ecosystem to get the right industry-specific solutions so they can build and implement AI agents.” Salesforce recently partnered with Google, integrating Google’s Gemini AI into Salesforce’s Agentforce. This allows agents to process images, audio, and video, handle complex tasks with Gemini’s multi-modal capabilities, and provide real-time insights using Google Search with Vertex AI.","excerpt":"AgentExchange includes over 200 partners, such as Google Cloud, Docusign, and Box, providing prebuilt solutions for AI agents.","categories":["AI News"],"tags":["Salesforce"],"author_name":"Siddharth Jindal","publish_date":"2025-03-04T19:49:04","publication_year":"2025","word_count":397,"keywords":["Go","agentic AI","AI","Modal","ML","Git","RAG","automation","Salesforce","R","startup"],"extracted_tech_keywords":["AI","ML","agentic AI","RAG","R","Go","Git","automation","startup","Modal"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/salesforce-launches-agentexchange-marketplace-for-ai-agents\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":26966,"title":"Data Engineering 101: Top Tools And Framework Resources","content":"In today’s fast-paced world, data can be compared to DNA — with data, it is easy to understand the past, predict the future and also replicate what it contains. Back in the early 2000s, the amount of data collected was just 5 to 10 percent of what we have collected in the last two years. Data collection, data engineering, and managing the warehouses are in high demand right now. Every company in today’s world wants to hire highly-skilled professionals who can deal with massive amounts of data and draw insights from it. There is no formal degree to be a data engineering graduate as of now. Nonetheless, there is a huge demand for data engineers and companies are hiring engineers for analytics positions. A recent study conducted by Analytics India Magazine found out that programming languages Python and R are commonly used across this domain for analysis and visualisation. Let us look at some of the MOOCs and books from which one can learn important prerequisites for data engineers — programming languages such as Python, R, and big data tools like Hadoop and Spark. In this article, we shall look at some of the well-known resources, both paid and free, from which one can acquire the right skills for a data engineering role. We have listed these resources according to the learning order. 1| Python I| Python Programming by Sentdex (MOOC) This is an open-source educational platform built and managed by Harrison Kinsley. One can learn Python from scratch here since it is one of the best free MOOCs out on the internet. There are advanced concepts of web developments, robotics explained using Python, which is quite fascinating. There are other interesting projects which Harrison himself has explained and built in real time. II| How To Code In Python by Lisa Tagliaferri (eBook) Python is one of the most versatile languages and is considered as the most widely-used language among developers in 2018. It has gained a lot of attention because it supports the scripting and object oriented programming style. This book explains how one from a non-programming background can learn and implement python for developing and various purposes. The author also explains how easy it is to learn Python because of it uses easy English words used in programming. 2| R I| Introduction to R – DataCamp (MOOC) This course is focused towards statistical modelling and analysis using R language. As many companies ask for R skills during hiring, this course comes in very handy. If one knows how to handle data then the company expects you to understand it too. 3| Apache Hadoop I| Hadoop Operations by Eric Sammer (Book) Eric Sammer has explained how one can start with Hadoop, from installing it on your system to architecture construction. It also explains clustering data with huge samples. An overview of HDFS and MapReduce has also been explained — why they are implemented and how they help in streaming the data. This is great to cluster and run a production environment. II| Hadoop Platform and Application Framework by Coursera (MOOC) This course focuses on teaching Hadoop frameworks for big data analysis. Also teaches the MapReduce techniques various other Hadoop-related content. 4| Apache Spark 1| Spark: The Definitive Guide: Big Data Processing Made Simple (eBook) by Bill Chambers and Matei Zaharia This book talks about how one can deal with query languages like SQL, learn about data frames and also make use of Spark’s API. Spark also includes clustering and monitoring, where one can process the data and execute them in real time. It also includes how one can make use of MLlib of Spark for data modelling and machine learning applications. II| Introduction To Apache Spark and AWS – Coursera (MOOC) An end-to-end applications of Spark is explained in this Coursera course. Spark is 100 times faster than Hadoop MapReduce and 5 times faster on the disk. It also has real-time batch processing which is unavailable on Hadoop. This MOOC also gives you a grading system where one can have a hands on experience for better understanding. 5| Apache Kafka I| Kafka: The Definitive Guide: Real-Time Data and Stream Processing at Scale (eBook) by Gwen Shapira, Neha Narkhede, and Todd Palino Streaming of data refers to controlling the data flow. With the help of computer programming, one can build a stream processing program which efficiently uses the concept of parallel programming for computations. This book gives you a quick understanding of the NoSQL databases and MongoDB. It also gives you insights on how relational databases are different from document-oriented databases. II| Tutorials Point – Apache Kafka Tutorial (Open Source Tutorial) This tutorial will explore the principles of Kafka, from installation to operations and then it will walk one through with the deployment of Kafka cluster. It is concluded with real-time applications, hands-on and integration with Big Data Technologies. Conclusion This article summarises several unique resources for learning and implementing data engineering concepts in the industry. One can make use of these to understand how to deal with and process huge databases. These are resources of 5 most common skills. The requirements for data engineer roles might vary depending on companies and they might ask for skills such as Java, C++, SQL, Scala, etc.","excerpt":"In today’s fast-paced world, data can be compared to DNA — with data, it is easy to understand the past, predict the future and also replicate what it contains. Back in the early 2000s, the amount of data collected was just 5 to 10 percent of what we have collected in the last two years. […]","categories":["AI Trends"],"tags":["apache kafka","Hadoop","kafka","Python","R language","real-time integration to kafka","spark","tools"],"author_name":"Kishan Maladkar","publish_date":"2018-08-08T12:26:33","publication_year":"2018","word_count":874,"keywords":["kafka","machine learning","R language","AWS","AI","MongoDB","Hadoop","spark","ML","Apache Spark","Python","tools","real-time integration to kafka","analytics","Kafka","apache kafka","R"],"extracted_tech_keywords":["AI","machine learning","ML","analytics","AWS","Apache Spark","Kafka","MongoDB","Python","R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/data-engineering-101-top-tools-and-framework-resources\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":65944,"title":"Top 10 Papers On Transfer Learning One Must Read In 2020","content":"Transfer Learning has recently gained attention from researchers and academia and has been successfully applied to various domains. This learning is an approach to transferring a part of the network that has already been trained on a similar task while adding one or more layers at the end, and then re-train the model. In this article, we list down the top 10 researchers papers on transfer learning one must read in 2020. (The papers are listed according to the year of publishing) 1| Pay Attention to Features, Transfer Learn Faster CNNs About: Transfer learning offers the chance for CNNs to learn with limited data samples by transferring knowledge from models pre-trained on large datasets. In this paper, the researchers proposed attentive feature distillation and selection (AFDS), which not only adjusts the strength of transfer learning regularisation but also dynamically determines the important features to transfer. According to the researchers, by deploying AFDS on ResNet-101, a state-of-the-art computation reduction has been achieved at the same accuracy budget, outperforming all existing transfer learning methods. Click here to read. 2| A Target-Agnostic Attack on Deep Models: Exploiting Security Vulnerabilities of Transfer Learning About: One security vulnerability of transfer learning is that pre-trained models, also referred to as teacher models, are often publicly available. This means that the part of the model transferred from the pre-trained model is known to potential attackers. In this paper, the researchers showed that without any additional knowledge other than the pre-trained model, an attacker can launch an effective and efficient brute force attack that can craft instances of input to trigger each target class with high confidence. To evaluate the proposed attack, the researchers performed a set of experiments on face recognition and speech recognition tasks to show the effectiveness of the attack. Click here to read. 3| Adversarially Robust Transfer Learning About: The purpose of this paper is to study the adversarial robustness of models produced by transfer learning. To demonstrate the power of robust transfer learning, the researchers transferred a robust ImageNet source model onto the CIFAR domain, achieving both high accuracy and robustness in the new domain without adversarial training. They further used visualisation methods to explore properties of robust feature extractors. According to the researchers, they constructed and improved the generalisation of a robust CIFAR-100 model by roughly 2% while preserving its robustness. Click here to read. 4| Meta-Learning Acquisition Functions for Transfer Learning in Bayesian Optimisation About: In this paper, the researchers proposed a novel transfer learning method to obtain customised optimisers within the well-established framework of Bayesian optimisation and allowed the algorithm to utilise the proven generalisation capabilities of Gaussian processes. Using reinforcement learning to meta-train an acquisition function (AF) on a set of related tasks, the proposed method learns to extract implicit structural information and to exploit it for improved data-efficiency. According to the researchers, the results show that the algorithm automatically identifies the structural properties of objective functions from available source tasks or simulations, performs favourably in settings with both scarce and abundant source data, and falls back to the performance level of general AFs if no particular structure is present. Click here to read. 5| DT-LET: Deep Transfer Learning By Exploring Where To Transfer About: In this paper, the researchers proposed a new mathematical model named Deep Transfer Learning By Exploring Where To Transfer (DT-LET) to solve this heterogeneous transfer learning problem. In order to select the best matching of layers to transfer knowledge, the researchers defined specific loss function to estimate the corresponding relationship between high-level features of data in the source domain and the target domain. Click here to read. 6| A Survey on Deep Transfer Learning About: This survey focuses on reviewing the current research of transfer learning by using deep neural networks (DNN) and its applications. The researchers defined deep transfer learning, its category and reviewed the recent research works based on the techniques used in deep transfer learning. Click here to read. 7| A Study on CNN Transfer Learning for Image Classification About: In this paper, the researchers proposed a system which uses a Convolutional Neural Network (CNN) model called Inception-v3. It was first trained on a base dataset called ImageNet and is then repurposed to learn features or transfer them in order to be trained on a new dataset such as CIFAR-10 and Caltech Faces. The researchers investigated whether it would work best in terms of accuracy and efficiency with new image datasets via Transfer Learning. Click here to read. 8| A Survey of Transfer Learning About: This is a survey paper aimed to provide insights into transfer learning techniques to the emerging tech community by overviewing related works, examples of applications that are addressed by transfer learning, and issues and solutions that are relevant to the field of transfer learning. The research provides an overview of the current methods being used in the field of transfer learning as it pertains to data mining tasks for classification, regression, and clustering problems. Click here to read. 9| A Survey on Transfer Learning About: This survey focuses on categorising and reviewing the current progress on transfer learning for classification, regression and clustering problems. In this survey, the researchers discussed the relationship between transfer learning and other related machine learning techniques such as domain adaptation, multitask learning and sample selection bias, as well as covariate shift. The researchers also explored some potential future issues in transfer learning research. Click here to read. 10| Self-taught Learning: Transfer Learning from Unlabeled Data About: In this paper, the researchers presented a new machine learning framework called “self-taught learning” for using unlabeled data in supervised classification tasks. This approach to self-taught learning uses sparse coding to construct higher-level features using the unlabeled data where the features form a succinct input representation and significantly improve classification performance. Click here to read.","excerpt":"Transfer Learning has recently gained attention from researchers and academia and has been successfully applied to various domains. This learning is an approach to transferring a part of the network that has already been trained on a similar task while adding one or more layers at the end, and then re-train the model. In this […]","categories":["AI Trends"],"tags":["deep learning application examples","future of deep reinforcement learning","Machine Learning","Transfer Learning"],"author_name":"Ambika Choudhury","publish_date":"2020-05-26T15:00:00","publication_year":"2020","word_count":970,"keywords":["Go","future of deep reinforcement learning","machine learning","meta-learning","programming_languages:R","AI","neural network","Machine Learning","ResNet","Aim","CNN","deep learning application examples","Transfer Learning","R"],"extracted_tech_keywords":["AI","machine learning","neural network","Aim","R","Go","CNN","ResNet","meta-learning","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-10-papers-on-transfer-learning-one-must-read-in-2020\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10080102,"title":"The Birth of AI-powered FastSaaS","content":"GenerativeAI is the rage right now, and companies are waking up to this less effort, more returns bid for the future. The current wave of SaaS products is developed around AI use cases ever since the release of LLMs like OpenAI’s GPT-3 and diffusion models as commercially-available APIs. In recent times alone, Shutterstock said it would integrate Dall-E 2 to its platform, Microsoft VS Code added Copilot, the AI pair-programming assist, as an extension, TikTok announced the rollout of an in-app text-to-image AI generator, and Canva came out with an AI-generating feature on its platform. Not to forget, a few days ago, Notion released its own AI domain which integrates AI-generated solutions to its all-in-one workspace. The new launch revealed that, with NotionAI, users would be able to generate “any type of content”—be it writing blog posts, brainstorming new ideas, creating press releases, ideating creative social media posts, summarising and translating texts, among several other things. It is important to note that these large-scale enterprises are up against plenty of startups that have raised funding to build AI-powered tools. Notion, for instance, is competing with Jasper, which raised $125 million for its AI content platform, and Instoried, the content creation startup which raised $200 million in funding, and Regie.ai, which raised $10 million for its business content creation platform. The list is endless. This list seems incomplete without Mem. Led by investment from OpenAI Startup Fund, Mem was the first to introduce an AI-powered workspace organisation tool, with its product, Mem X. Notion’s reinvention of its platform may have been a direct effect of the shift in tide towards many of the AI application tools. Same same, but different All the startups mentioned above use GPT-3 for content generation. It is not clear yet if NotionAI also uses GPT-3, although speculations suggest it does. The question here is: If all content\/image tools are based on GPT-3 or Stable Diffusion, what really separates one from the other? Chris Frantz, a co-founder of Loops, addresses this, saying, “There’s almost no moat in generative AI. That means that the differentiator is product, marketing, use case and time to market.” That is, with every SaaS company boarding the AI train, the only differentiating factors between them is the interface, the ease-of-use, accessibility, and the ability to leverage the existing user base on the platform. So, even if large enterprise companies are not the first-movers in terms of leveraging the AI fever, they have the advantage of having a substantial user base to begin with. NotionAI, for example, had a waitlist for over 50,000 users, and this was only a few hours after the launch. FastSaaS: New model of product development Therefore, it is safe to say that we are entering into an age of, what Matt Krandel calls, ‘fastSaaS’. The term represents how AI and no code are lowering barriers for developing software products. This, Krandell says, “can lead to margin compression, and more importance on distribution”, similar to what we mean when we say ‘fast fashion’. An interesting aspect to notice also is how a lot of the companies using AI are diversifying their product portfolio—not to bring more cutting-edge tech onto their platform, but to just have more and more AI-assisted features integrated into them. If we simply compare the AI copywriting tools available in the market today, we notice an abundance-in-making. Each tool segments into multiple functionalities. Writersonic, for example, labels multiple use cases generated to the same writing API like articles and blog posts, ecommerce product descriptions, ads and marketing tools, website copy, and numerous other general applications. AI copywriting tool Simplified also has an AI-image generator to the list of features it provides with an AI writing tool. And with Notion, we are seeing services not only related to AI writing, but also additional ones that include spellcheck and grammar, paraphrasing, or translation tools. Foraying in SaaS at the moment is like entering an arena of excess production, with little or no understanding of consumer needs. Definitely skeptical. Personally I’m a huge fan of generative AI, but it seems like it’s becoming another trend just for trend’s sake. I worry a lot of these companies are being created just for sake of adding to the hype, rather than to make something truly helpful— Matty Hogan (@mattyhogan_) November 17, 2022 Beyond LLMs as AI generators Basically, building a SaaS is made equivalent to creating an “everything” app. A Twitter user by the name ‘Linus’ makes an interesting analysis of the case. Linus says that companies should be thoughtful about integrating AI into their product. Large Language Models (LLMs) are not merely generating machines, but have human-like reasoning capabilities, and products should thereby be modelled around it—something which was also iterated by Alexandr Wang, CEO of Scale AI. Linus writes that the real deal is not in generating texts, but in “using this new capability and wrapping into jobs that have other shapes”, while also adding that, “text generation is not the product, reasoning is”. Speaking about NotionAI, Linus stressed that the product, if remains limited to text generation, will end up making “10 copies of every piece of information, 4 of which contradict each other and only 2 of which reliably surface on searches”. Hence, beyond its application to narrow workflows, the content generating machine can leave organisations with more information than necessary.","excerpt":"‘FastSaaS’ – the term represents how AI and no code are lowering barriers for developing software products","categories":["AI Features"],"tags":["AI Tool","APIs","DALL.E","GPT-3","SaaS","Stable Diffusion"],"author_name":"Ayush Jain","publish_date":"2022-11-18T17:00:00","publication_year":"2022","word_count":893,"keywords":["GPT-3","Go","API","DALL-E","OpenAI","AI","Stable Diffusion","RAG","APIs","Ray","SaaS","GPT","generative AI","AI Tool","R","DALL.E"],"extracted_tech_keywords":["AI","generative AI","OpenAI","Ray","RAG","R","Go","API","GPT","DALL-E"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/the-birth-of-ai-powered-fastsaas\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":6,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":30732,"title":"How To Start A Career In Data Science And Not Get Stuck In Dead-End Tech Jobs","content":"Preparation for any career requires a lot of sweat and toil and in the case of data science, it is not restricted to performing well on the interview day alone. An aspiring data scientist is expected to prepare across multiple fronts. In today’s world, whatever your field of work, having skills and knowledge in Data Science will play a great role in your career development. The McKinsey Global Institute estimates that the US could have as many as 250,000 open data science jobs by 2024. Closer home in India, a study by Edvancer and Analytics India Magazine found that the number of new analytics jobs advertised per month increased by almost 76 percent from April 2017 to April 2018. Even our own research confirms this. When we harvested and processed data from millions of job descriptions across the market, we found that the demands in data science are increasing at a rapid rate. If we were to look at a location-wise trend, Bangalore stands on top as the hub of machine learning and AI, covering 34% of total jobs created. By all accounts, data is poised to play a massive role in shaping the future of the industry, and possibly the future of humanity itself. Therefore, a career in data science certainly makes for an attractive proposition for any aspiring engineer. The glut of courses in data analytics, AI, machine learning etc. points to the growing popularity of data as a career option. While the interest in data science is certainly welcome, I have found that there isn’t enough clarity among students and young IT professionals on what it takes to prepare for a career in data science. A common mistake people make is to enrol in online courses that are either too basic or vague to be useful in practice. The danger here is that these courses are likely to get you stuck in dead-end data analytics or AI jobs that are repetitive and provide no scope for learning. Start A Career In Data Science Professional Qualifications And Skills Spirit Of Enquiry: In the industry, it’s quite rare to be presented with a fleshed-out problem statement that needs to be solved using data. Rather, you will likely get your best insights by studying patterns and trends in the reams of data. The ability to find these patterns or anomalies needs a certain basic curiosity and knack for looking beyond the obvious. Analysing: When dealing with multiple parameters and data sources, trying to find the exact correlation and causation for any parameter can be confusing and challenging. The ability to use logical reasoning to draw inferences and arrive on a conclusion is invaluable as a data scientist. Big Picture Thinking: When you are analysing large amounts of data, it is easy to get overwhelmed by the findings and insights. But the real value of those insights lies in them being rolled up to the business level so that they have a real impact on decision making. It needs imagination and ability to look at the big picture and draw insights that truly matter. The potential impact of data science is limited only by imagination. Whether it is bringing down crime rates or reducing traffic woes or saving the planet from climate change, no problem is too big to be solved through data, provided you have the right vision. Perseverance: Despite all the great things data and AI can do, the ground truth remains that the success rate isn’t always high for data projects. Patience is a key virtue while working with data. You can safely assume that 90 per cent of the things that you do will fail to meet their logical conclusion. It’s difficult to accept failure, especially when you put your heart and soul into something, but perseverance is what will get through it. Mathematics Lover: Needless to say, a lot of data science boils down to mathematical and statistical analysis. While you don’t need to be an expert in advanced mathematics, a basic love for mathematics and a math-oriented approach are crucial. If you love tinkering with datasets etc. it will serve you well in the field of data. If you think you possess these traits and would like to explore a career in data science, we can move to the next step, which is to understand the skills and knowledge that you need. Courses and Training Data scientists need well-rounded knowledge of mathematics, statistics, deep learning, machine learning and NLP. Here are a few recommended courses and training institutes that you can explore. Institutes: https:\/\/www.insightdatascience.com\/ https:\/\/www.thisismetis.com\/ https:\/\/www.thedataincubator.com\/ Courses: Prerequisite In Mathematics: (i) Linear Algebra, Probability, Statistics: https:\/\/ocw.mit.edu\/courses\/mathematics\/18-06-linear-algebra-spring-2010\/ https:\/\/ocw.mit.edu\/resources\/res-6-012-introduction-to-probability-spring-2018\/ https:\/\/www.edx.org\/course\/introduction-statistics-descriptive-uc-berkeleyx-stat2-1x https:\/\/ocw.mit.edu\/courses\/sloan-school-of-management\/15-075j-statistical-thinking-and-data-analysis-fall-2011\/index.htm https:\/\/www.stat.berkeley.edu\/~aldous\/134\/gravner.pdf https:\/\/www.stat.berkeley.edu\/~aldous\/134\/grinstead.pdf (ii) Machine learning: http:\/\/www.greenteapress.com\/thinkstats\/ http:\/\/www.greenteapress.com\/thinkbayes\/thinkbayes.pdf https:\/\/www.coursera.org\/learn\/machine-learning (iii) Natural Language Processing: https:\/\/web.stanford.edu\/~jurafsky\/slp3\/ (iv) Deep learning and AI: http:\/\/web.stanford.edu\/class\/cs224n\/ http:\/\/cs231n.stanford.edu\/ https:\/\/github.com\/oxford-cs-deepnlp-2017\/lectures Books: Books are always a great way to shore up knowledge on any subject. Here’s a list of some of the best books on data sciences that I highly recommend reading. https:\/\/www.cs.ubc.ca\/~murphyk\/MLbook\/ http:\/\/users.isr.ist.utl.pt\/~wurmd\/Livros\/school\/Bishop%20-%20Pattern%20Recognition%20And%20Machine%20Learning%20-%20Springer%20%202006.pdf https:\/\/nlp.stanford.edu\/IR-book\/pdf\/irbookonlinereading.pdf https:\/\/www-bcf.usc.edu\/~gareth\/ISL\/ISLR%20Sixth%20Printing.pdf Competitions Competitions are a great way to brush up your skills and develop a problem-solving approach. Remember that participation and learning should be the focus. It isn’t about winning and losing; it’s the learning that matters. https:\/\/www.kaggle.com https:\/\/www.drivendata.org\/competitions\/ https:\/\/machinehack.com Influencers: As they say, we need to stand on the shoulders of giants to be able to look further. Following some of the greatest minds in data science and understanding their thought process is a great way to keep your knowledge up to date. Here’s an indicative list of some data science experts that every professional must follow. https:\/\/twitter.com\/fchollet https:\/\/twitter.com\/goodfellow_ian https:\/\/twitter.com\/chrmanning https:\/\/twitter.com\/rsalakhu https:\/\/twitter.com\/RichardSocher https:\/\/twitter.com\/karpathy For people that want to make themselves more indispensable to their employers while gaining an innovative, creative, and sustainable career path data scientist undoubtedly seems like the dream career option. With the right preparation, the only the sky is the limit.","excerpt":"Preparation for any career requires a lot of sweat and toil and in the case of data science, it is not restricted to performing well on the interview day alone. An aspiring data scientist is expected to prepare across multiple fronts. In today’s world, whatever your field of work, having skills and knowledge in Data […]","categories":["AI Hirings"],"tags":["Data Science Jobs","science jobs"],"author_name":"Rahul Kulhari","publish_date":"2018-11-27T11:06:32","publication_year":"2018","word_count":974,"keywords":["data science","Go","machine learning","AI","science jobs","ML","Data Science Jobs","Git","NLP","deep learning","analytics","R"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","NLP","data science","analytics","R","Go","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-hiring\/how-to-start-a-career-in-data-science-and-not-get-stuck-in-dead-end-tech-jobs\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":6463,"title":"Kloutix","content":"Kloutix is a next generation analytical company focused on “Analytics of the Future” by providing high-end business analytics leveraging the latest technologies like cloud, open source etc., supplemented by the propriety automation and platform capabilities. Never before in the history of consumerism, has the power of decision making shifted so radically into the hands of the consumer. With social media and mobile becoming mainstream, gone are the days when any business could work solely on the traditional models of business management. Today agility and consumer responsiveness has to be high and every aspect of the business has to be constantly improved significantly. This shift demands an expansion in the role of analytics across functions, business units, departments etc. making it an organization wide decision making discipline. A large part of intuitive and experience based decision making has to be replaced with scientific methods of outcome prediction. Yet Analytics remain largely an enigma for most corporations. High investments, unsure outcomes and general lack of awareness has led to Analytics adoption only to large corporations in a limited way. Gartner estimates that only 11% organization worldwide are leveraging analytics. Kloutix helps organizations to build analytical solutions that not only enable decision making but also does so in the most cost effective manner. Our business results driven approach coupled with latest innovations, ensures that a business person does not have to try hard to prove the ROI and build a scalable solution. We help companies institutionalize analytics by taking ownership of implementation across data layer to analytics and reporting later. We leverage the latest technologies like Cloud, Open Source to ensure the business benefits do not require huge investments. We have built a robust analytical platform called K-Analytics as shown below: The K-Analytics platform has following components: Analitix – Advanced pre-built predictive models using best of breed or open-source tools and delivered in a self-service mode Socilaitix – Implementation framework and a library of KPIs that helps monetization and measurement of social media Infitix – A platform with an extensive data model, pre-built KPIs and business questions that makes the datamart or datastore or data warehouse deployment fast and efficient on the cloud Automatix – Automated data capture mechanism to solve the first-mile data capture challenges, esp. for emerging markets. In summary, we see the analytical wave encompassing the businesses in the same manner the process automation wave encompassed more than a decade ago. The new technological advancements are making it possible to deliver advanced capabilities in an efficient cost effective manner. Kloutix will lead this wave by leveraging its data science lineage along with relevant emerging technologies to deliver zero-capex, on-demand solutions. Share our vision at www.kloutix.com","excerpt":"Kloutix is a next generation analytical company focused on “Analytics of the Future” by providing high-end business analytics leveraging the latest technologies like cloud, open source etc., supplemented by the propriety automation and platform capabilities. Never before in the history of consumerism, has the power of decision making shifted so radically into the hands of […]","categories":["Deep Tech"],"tags":[],"author_name":"Dinesh Jain","publish_date":"2014-11-19T18:20:46","publication_year":"2014","word_count":445,"keywords":["data science","Go","AI","Scala","RAG","automation","analytics","GAN","R","data warehouse"],"extracted_tech_keywords":["AI","data science","analytics","RAG","R","Go","Scala","data warehouse","GAN","automation"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/kloutix\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10017201,"title":"Tech Behind Facebook AI’s Latest Technique To Train Computer Vision Models","content":"While image classification sounds like a simple process to human ears, it can be a daunting task for machines. Thus, purely based on attention, neural networks have dramatically improved image understanding tasks; however, these visual transformers have typically been pre-trained with massive sets of images using expensive infrastructures. This, in turn, restrains a larger part of the community from adopting the same. To address this issue, researchers at Facebook AI have developed a new technique — Data-efficient image Transformers (DeiT) — that leverages Transformers but requires far fewer data to produce an image classification model. Explaining the research, in a recent blog, Facebook AI stated that the researchers want to showcase that Transformers can be trained efficiently with only standard academic data sets, for image classification. With this, the researchers also aim to extend Transformers to new use cases and advance the field of computer vision. In a recent tweet, the company has even announced open sourcing the system for researchers and engineers who don’t have access to large-scale systems for training massive AI models. We’re open-sourcing a new system to train computer vision models using Transformers. Data-efficient image Transformers (DeiT) is a high-performance image classification model requiring less data & computing resources to train than previous AI models. https:\/\/t.co\/nTJ4bD9Slp pic.twitter.com\/OMdB5stBay— AI at Meta (@AIatMeta) December 24, 2020 Also Read: A New Trend Of Training GANs With Less Data: NVIDIA Joins The Gang What is Data-efficient image Transformers? Considering Transformer model architecture has managed to create some breakthroughs for NLP and machine translation, Facebook AI researchers decided to deploy the same for some tasks like speech recognition, symbolic maths and translation of programming languages. On the other hand, this new technique, Data-efficient image Transformers (DeiT,) requires far fewer data and computing resources to provide results on image classification tasks. To facilitate this, the researchers of Facebook AI collaborated with Prof Matthieu Cord from Sorbonne University. They trained the DeiT model with a single 8-GPU server in two to three days. This provided a result of 84.2 top-1 accuracies on ImageNet, without using any external data for training. The performance curve of the comparison between DeiT, visual Transformer models and CNNs. The researchers even noted that the proposed model produced competitive results compared to the dominant convolutional neural networks. Also Read: Are Easy-To-Interpret Neurons Necessary? New Findings By Facebook AI How Does It Work? With no statistical priors about images, image classification was challenging for convolution-free transformer models like DeiT. Thus, they have to “see” a massive set of images in order to classify the different objects in the image. To avoid this issue, the first critical step was to create a training strategy for DeiT. For this, the researchers adapted existing research on convolutional neural networks, particularly for data augmentation, optimisation and regularisation. Therefore, DeiT gave competitive results despite being trained on 1.2 million images rather than hundreds of them. Secondly, the researchers modified the Transformer architecture to enable native distillation that helped neural networks (the student) to learn from the output of another network, that acts as a teacher. They used CNN as a teacher for the DeiT model, and since such an architecture comes with priors about images, it was easier for them to train it with lesser images. The distillation token interacting with the classification vector and image component tokens through the attention layers. Further, the student model was learning from two different sources — labelled data set and from the teacher; thus, there was a possibility of diverging knowledge. To make the student model only learn from the teacher, the researchers introduced a distillation token that cues the model for its distillation output. According to the researchers, the method of native distillation was explicitly designed for Transformers to improve the image classification performance. Wrapping Up With this, it can easily be said that DeiT will surely make advancements in computer vision, using Transformers. The experimentation showcased that the resulting output by the proposed model has been competitive with that of convolutional neural networks, that has been dominating the field of CV since almost a decade. Therefore, such a convolution-free transformer model like DeiT will surely democratise artificial intelligence, making it possible for developers to train AI models with fewer data.Read the paper here.","excerpt":"While image classification sounds like a simple process to human ears, it can be a daunting task for machines. Thus, purely based on attention, neural networks have dramatically improved image understanding tasks; however, these visual transformers have typically been pre-trained with massive sets of images using expensive infrastructures. This, in turn, restrains a larger part […]","categories":["AI Trends"],"tags":["Computer Vision","Computer vision models","Facebook AI","Facebook AI research","how does ai work"],"author_name":"Sejuti Das","publish_date":"2021-01-06T13:00:00","publication_year":"2021","word_count":707,"keywords":["artificial intelligence","TPU","Facebook AI","AI","neural network","how does ai work","Transformers","computer vision","RAG","NLP","Aim","Computer Vision","Computer vision models","R","Facebook AI research"],"extracted_tech_keywords":["AI","artificial intelligence","neural network","NLP","computer vision","Aim","Transformers","RAG","TPU","R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/tech-behind-facebook-ais-latest-technique-to-train-computer-vision-models\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10050474,"title":"Alphabet Owned Intelligence Lab DeepMind Turns Its First Profit","content":"London-based research firm and one of the world’s premier artificial intelligence labs, DeepMind, has reportedly turned a profit for the first time ever, according to a filing with the U.K. company registry published recently. Owned by Google parent company Alphabet, the research firm recorded a profit of £43.8 million ($59.6 million) in 2020 after posting losses of hundreds of millions for the last several years. With no specific reason provided for the revenue jump, the turnover at DeepMind has more than tripled from just £265.5 million in 2019 to £826.2 million in 2020, according to the annual results filing on Companies House. DeepMind employs some of the world’s leading AI research scientists, who can command annual salaries of more than $1 million. DeepMind doesn’t sell products directly to consumers, and it hasn’t announced any deals with private companies outside of the Alphabet umbrella. It does, however, sell software and services to Alphabet’s companies, including Google, YouTube and X, the moonshot division. DeepMind did not specify how many people it added to its team in 2020, but it said it now employs over 1,000 people. The staff costs rose modestly from £467 million to £473 million, suggesting that DeepMind’s hiring frenzy may have come to an end.","excerpt":"The research firm recorded a profit of £43.8 million ($59.6 million) in 2020 after posting losses of hundreds of millions for the last several years.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","Alphabet","Data Science","Data Scientist","Deep Learning","Google","Machine Learning","Python","Revenue Growth"],"author_name":"Victor Dey","publish_date":"2021-10-06T11:54:58","publication_year":"2021","word_count":206,"keywords":["Go","artificial intelligence","Deep Learning","programming_languages:R","AI","Machine Learning","programming_languages:Go","Python","Google","AI research","Alphabet","Data Science","Data Scientist","R","AI (Artificial Intelligence)","Revenue Growth"],"extracted_tech_keywords":["AI","artificial intelligence","R","Go","AI research","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/alphabet-owned-intelligence-lab-deepmind-turns-its-first-profit\/","complexity_score":3,"technical_depth":7,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":66773,"title":"How Snowflake And Salesforce’s New Phase Of Partnership Expands Data Analytics Features","content":"Last year, Snowflake Computing had raised $479 million in Series G funding round with Salesforce as the co-lead investor. Snowflake recently expanded the partnership with new analytics features. According to Snowflake, it shares many joint customers with Salesforce and Tableau, and the partnership will enable the company to create a better experience for the customers. “We want to make it easier to join your Salesforce data with your data in Snowflake, and use those new insights to power business decisions,” said Snowflake. Recently, Snowflake launched the Data Cloud – an ecosystem where thousands of Snowflake customers, partners, data providers, and data service providers can help integrate data silos, derive insights, and create value from quickly expanding data sets in a secure, and seamless manner. For Snowflake, the rise of the Data Cloud coincides with the latest innovations from the company’s partnership with Salesforce. For Snowflake, the rise of the Data Cloud coincides with the latest updates from the company’s partnership with Salesforce. This means that users will be able to more easily, access their Salesforce data in Snowflake Data Cloud Platform and combine their Snowflake and Salesforce data to derive the deepest insights possible. This capability further increases what’s possible in the Data Cloud. Functioning as a virtual data lake, the platform can provide easy to search and analytical capacities across different cloud platforms, meaning users can securely access data and apps regardless of the platform. Alongside, users can perform different functions like running data queries, visualisation dashboards or building AI models each with their data warehouse. With Einstein and Tableau analytics capabilities on Snowflake, customers can benefit from the cloud-neutral nature of Snowflake. What Are The New Analytics Tools? Already in open beta, customers of Einstein Analytics can now use a new feature: Einstein Analytics Direct Data for Snowflake. The cool thing about this is that instead of having to copy data from Snowflake into Einstein analytics, users can query Snowflake directly to get insights from live data. Einstein Analytics Direct Data for Snowflake allows Einstein Analytics’ users to query data within Snowflake, including Salesforce information and generated data from various sources: apps, mobile, Web, IoT devices and acquired data from the Snowflake Data Marketplace. This feature is currently in open beta and immediately available for any Einstein Analytics’ customer. Snowflake also announced the Einstein Analytics Output Connector for Snowflake. Einstein Analytics Output Connector is a tool for Snowflake that enables customers to move Salesforce data into Snowflake seamlessly and query with Einstein Analytics and Tableau Software. Customers currently use a variety of technologies to import data from Salesforce into Snowflake. With simple configuration steps, users will be able to synchronize Salesforce data to Snowflake, and leverage Salesforce objects as well as curated Einstein Analytics data sets to augment their data workloads in Snowflake. The output connector will be available for users later this year. The Advantages For Users The native integrations between the two technologies will allow organisations to unify and analyse all data in Snowflake Cloud Data Platform and visualise it with business intelligence tools such as Tableau and Salesforce applications, to find new insights about their businesses and their customers. This will reduce data gaps, make insights available and actionable to all business users, and deliver more valuable data-driven user experiences. The features of Snowflake can be useful for companies that host their systems on one cloud platform such as AWS, whereas their data warehouse and data management functions are dispersed in various places. Adam Selipsky- President and CEO of Tableau said, “Enabling more people to unlock the potential of data is at the core of the collaboration between Salesforce, Snowflake and Tableau, and it’s now even simpler for customers to access and analyse their critical data on the Snowflake Cloud Data Platform. Also Read: Why Salesforce Invested In This Data Warehousing Startup Called Snowflake","excerpt":"Last year, Snowflake Computing had raised $479 million in Series G funding round with Salesforce as the co-lead investor. Snowflake recently expanded the partnership with new analytics features. According to Snowflake, it shares many joint customers with Salesforce and Tableau, and the partnership will enable the company to create a better experience for the customers. […]","categories":["AI Features"],"tags":["Data Analytics","data management providers","Visual Analytics Provider"],"author_name":"Vishal Chawla","publish_date":"2020-06-05T15:00:00","publication_year":"2020","word_count":639,"keywords":["TPU","AWS","AI","ML","data warehouse","RAG","Visual Analytics Provider","analytics","Data Analytics","data management providers","R","data lake","Snowflake"],"extracted_tech_keywords":["AI","ML","analytics","RAG","AWS","TPU","Snowflake","R","data warehouse","data lake"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-snowflake-and-salesforces-new-phase-of-partnership-expands-data-analytics-features\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10097282,"title":"Microsoft Now Makes Google Sweat","content":"Microsoft is everywhere. This copilot-obsessed tech giant wants anything and everything to do with generative AI – open source or closed door, small or large, it doesn’t matter – and can go to any extent to bring that capability to life. While OpenAI’s closed-source approach may have raised concerns among developers, Meta is viewed as the good guy, and Microsoft stands as a neutral guardian, focused solely on investing in technological progress. The ongoing debates between open source and closed source don’t seem to bother Microsoft, as it has ensured a well-balanced position is and prepared for any scenario. Microsoft is not committed to any one particular approach while it comes to generative AI. The company is even backing budding startups that might rival OpenAI’s ChatGPT in the coming future. Microsoft recently invested in Inflection AI, a startup backed by several Silicon Valley heavyweights, which raised $1.3 billion. The company is also investing in small LLMs. Recently Microsoft released another open-source LLM, Orca, a highly advanced model with 13 billion parameters, designed to imitate the reasoning capabilities of LFMs. It utilises GPT-4 to learn from various signals, including explanation traces, step-by-step thought processes, and complex instructions. With the recent partnership with Meta, and backing of the open source research projects like Orca and their extended partnership with OpenAI  Microsoft is unifying the generative AI ecosystem. Currently, Microsoft has 252+ open source models on Hugging Face. In comparison, Google has 591 open source model contributions. With the latest partnership, which leads the open source AI contributions, we can expect more open source models from the duo (Microsoft and Meta). Making Google Sweat While Microsoft made Google “dance” with its OpenAI partnership, it surely is making them sweat with the Meta partnership. The fact that Google missed out on Llama 2’s release, which is now available on AWS and Hugging Face, seems intentional. In the competitive cloud market, Microsoft is aiming to provide specialised LLMs for its enterprise clients. Azure OpenAI customers utilising Microsoft Cloud will have access to top-notch specialised enterprise LLMs that they can seamlessly integrate into their businesses. When Microsoft partnered with OpenAI, Microsoft deployed OpenAI technology through API and the Azure OpenAI Service—enabling enterprise and developers to build on top of GPT, DALL·E, and Codex. They also worked together to build OpenAI’s technology into apps like GitHub Copilot and Microsoft Designer. Microsoft’s decision to partner with Meta is a well-thought-out business strategy. By collaborating with different entities, they avoid depending solely on one supplier. Microsoft is fully aware that the AI field is constantly evolving, and it’s only a matter of time until someone develops an even better LLM than GPT 4. To position themselves for future advancements, they are actively exploring open source opportunities and acting as a bridge between GPT 4 and Llama 2. Is Apple Next? Over the years, Microsoft has faced challenges in keeping up with Google, especially in the mobile phone operating system and internet browsing domains. While Google Chrome gained significant popularity over Internet Explorer and Bing, Microsoft attempted to boost Bing’s performance by integrating it with OpenAI chatbot. Every move Microsoft makes is driven by the desire to enhance its products and compete more effectively with Google. It further integrated ChatGPT into Bing launched Microsoft co-pilot bringing generative AI abilities to Word, Excel, PowerPoint, Outlook and Teams. In the past as well, Microsoft tried launching Windows smartphones but failed miserably in front of Android phones. But now it is training small LLMs. Meta, along with Microsoft, has also partnered with Qualcomm, eyeing an entire ecosystem to make Llama 2 implementation on phones and PCs starting next year. Smaller models like Orca can take full advantage of this. It is worth noting that ChatGPT hasn’t been released yet on Android PlayStore. In the blog post Microsoft specifically mentioned that Llama 2 is optimised to run locally on Windows. Open source Llama 2 will help Microsoft to pick cream developments and implement it. As of now, we can only speculate about the future of Llama 2 and its potential integration with Windows devices. Microsoft releasing models like Orca and others on edge devices can also be a threat to Apple. But, looks like Apple is ahead, and it is currently testing its generative AI chatbot ‘Apple GPT’.","excerpt":"While Microsoft made Google “dance” with its OpenAI partnership, it surely is making them sweat with the Meta partnership","categories":["Global Tech"],"tags":["Microsoft"],"author_name":"Siddharth Jindal","publish_date":"2023-07-20T14:59:35","publication_year":"2023","word_count":714,"keywords":["ChatGPT","Hugging Face","OpenAI","AI","AWS","R","ML","Aim","generative AI","Azure","Microsoft"],"extracted_tech_keywords":["AI","ML","generative AI","ChatGPT","OpenAI","Aim","Hugging Face","AWS","Azure","R"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/microsoft-now-makes-google-sweat\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10163140,"title":"How Daimler Truck Innovation Centre India is Using RAG Architecture for Smarter Solutions","content":"Daimler Truck is one of the world’s largest manufacturers of commercial vehicles, including trucks and buses. The company operates globally across four key regions—Europe, North America, Asia Pacific, and China. Daimler Truck Innovation Centre India (DTICI) is a GCC based in Bengaluru, established three years ago. It focuses on providing IT and engineering solutions for all segments, products, and regions of Daimler Truck. “Our goal is…to provide world-class engineering and IT solutions to our customers and products,” Raghavendra Vaidya, managing director and CEO of DTICI, told AIM. The RAG Conversation Vaidya also reaffirmed that retrieval-augmented generation (RAG) architecture remains a valuable approach. “I don’t think RAG architecture is dead. It’s working well for us,” the representative said. Companies can either retrain a model with their data or use RAG to enhance a GPT model. Both methods have their own advantages. Even major companies like Microsoft support RAG by offering models that help vectorise data efficiently, as per Vaidya. However, the conversation around RAG is not new. RAG has revolutionised how AI systems process and respond to user queries by using external knowledge sources. However, as it doesn’t meet all the diverse needs of modern enterprises, everyone wants to replace RAG with something new. This is where agentic RAG comes into play. Agentic RAG represents an advanced architecture that combines the foundational principles of RAG with the autonomy and flexibility of AI agents. It promises a future where AI systems are more adaptive, proactive, and intelligent. Furthermore, last year, Google released its new Gemma model, DataGemma. While the world is experimenting with RAG to reduce hallucinations and increase accuracy, Google decided to use retrieval interleaved generation (RIG). This technique integrates LLMs with Data Commons, an open-source database of public data. DTICI is not Building LLMs Vaidya highlighted that DTICI is not building large language models (LLMs) but is currently using OpenAI’s LLM for internal purposes. Machine learning (ML) remains a core focus, even though the term is less commonly used today, he further said. For over a decade, the company has been developing ML models from scratch, supported by a skilled team of data scientists and engineers who collaborate with business experts across different functions. In the past year, the company elevated its approach by assigning accountability for AI and data initiatives to its Bengaluru team. Regarding GenAI, DTICI is currently running multiple pilot projects to assess its potential impact. The company has already seen success in using Microsoft Copilot and GitHub Copilot to improve software development productivity, whether through code generation, test case creation, or code quality validation. Beyond software engineering, DTICI is exploring GenAI in sales, procurement, and after-sales. Rather than taking a technology-first approach, the company prioritises business needs. “We don’t bring in the technology, dabble and see what it can do. That’s not the approach we’re taking. We’re taking a business approach where we identify areas where it can produce business results and provide benefits, either to the top line or efficiency or the bottom line. And then we go and build a pilot around it,” Vaidya said. DTICI also built a sandbox on Azure about a year ago, using an older version of OpenAI’s language model. According to Vaidya, DTICI has been using the model for some time and finds it effective. It has created internal chatbots and assistants that use OpenAI’s language model and its own secure data. Vaidya acknowledged that training a model with its own data would be better, but it would take excessive time and money. Instead, DTICI prefers its current approach and believes it to be a good balance. DTICI’s Bengaluru Narrative Vaidya said that Bengaluru remains the top choice for talent in India, with its unmatched depth and variety of skilled professionals. “The length, breadth, and depth of talent you have in Bengaluru is unmatched.” He believes that as global capability centres (GCCs) grow, they may expand to other cities where talent is available. Some of them have successfully established operations in multiple locations. However, Bengaluru remains the first choice for new GCCs, and DTICI has no recent plans to expand to tier-2 cities. At DTICI Bengaluru, the focus is on engineering and IT. The team develops intelligent software for trucks and buses. Most of the innovation and investment are happening in IT, software, and electronics. Highlighting this trend, Vaidya said, “If you want to increase the frequency of innovation, or you want to innovate faster, then I think software and electronics is the place to be.” In IT, the company is deeply focused on using data, ML, and artificial intelligence. Vaidya revealed that a major project of predictive maintenance, directed from Bengaluru, aims to predict part failures using analytics and ML instead of traditional physics-based methods. The system analyses real-time truck data to forecast when a part is likely to fail. However, accuracy is critical for this to be effective. “If the model is not 85% or more accurate, then nobody is going to buy it,” Vaidya said. Since customers rely on these predictions to replace parts before failure, achieving high accuracy is essential. DTICI has been deploying these solutions over the past few years, and they have proven extremely effective in terms of profitability and cutting warranty costs. “It is pretty simple; you get to work on the bleeding edge of the technology and the work that you do makes either a product better or customers more profitable,” Vaidya concluded.","excerpt":"RAG has revolutionised how AI systems process and respond to user queries by using external knowledge sources.","categories":["GCC"],"tags":["Agentic RAG","AI (Artificial Intelligence)","RAG"],"author_name":"Shalini Mondal","publish_date":"2025-02-10T15:59:53","publication_year":"2025","word_count":902,"keywords":["Agentic RAG","GenAI","artificial intelligence","machine learning","OpenAI","AI","chatbots","ML","RAG","Aim","analytics","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","analytics","GenAI","OpenAI","Aim","RAG","chatbots"],"url":"https:\/\/analyticsindiamag.com\/gcc\/how-daimler-truck-is-using-rag-architecture-for-smarter-solutions\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10070253,"title":"How does YOLOv6 compare against YOLOv5?","content":"Computer vision is one of the most buzzing fields of AI. Major companies are dedicating massive amounts of resources to launch the next big thing in this field. One project that has truly stood out in recent years is YOLO – You Only Look Once. Introduced first in 2015 by Joseph Redmon et al via a paper titled, “You Only Look Once: Unified, Real-Time Object Detection,” it is considered a breakthrough in this field. Over the years, this model has undergone several iterations and advancements. Version 2 was released in 2016 (YOLO9000: Better, Faster, Stronger), followed by YOLOv3 (YOLOv3: An Incremental Improvement) in 2018, YOLOv4 (YOLOv4: Optimal Speed and Accuracy of Object Detection) in April 2020, and YOLOv5 in May 2020. YOLOv6 was recently introduced by Chinese company Meituan. It is not part of the official YOLO series but was named so since the authors of this architecture were heavily inspired by the original one-stage YOLO. It uses the prefix MT in its name. About MT-YOLOv6 YOLOv6 is a target detection framework dedicated to industrial applications. As per the company’s release, the most used YOLO detection frameworks – YOLOv5, YOLOX, and PP-YOLOE – leave a lot of room for improvement in terms of speed and accuracy. Recognising these ‘flaws,’ Meituan has introduced MT-YOLOv6 by studying and drawing further on the existing technologies in the industry. The MT-YOLOv6 framework supports the entire chain of industrial applications requirements like model training, inference, and multiplatform deployment. According to the team, MT-YOLOv6 has carried out improvements and optimisations at the algorithmic level, like training strategies and network structure, and has displayed impressive results in terms of accuracy and speed when tested on COCO datasets. Credit: DagsHub Unlike YOLOv5\/YOLOX, which are based on CSPNet and use a multi-branch approach and residual structure, Meituan redesigned the Backbone and Neck according to the idea of hardware-aware neural network design. As per the team, this helps in overcoming the challenges of latency and bandwidth utilisation. The idea is based on the characteristics of hardware and that of inference\/compilation framework. Meituan introduced two redesigned detection components – EfficientRep Backbone and Rep-PAN Neck. Further, the researchers at Meituan adopted the decoupled head structure, taking into account the balance between the representation ability of the operators and the computing overhead on the hardware. They used a hybrid strategy to redesign a more efficient decoupling head structure. The team observed that with this strategy, they were able to increase the accuracy by 0.2 per cent and speed by 6.8 per cent. In terms of training, Meituan adopted three strategies: Anchor-free paradigm: This strategy has been widely used in recent years due to its strong generalisation ability and simple code logic. Compared to other methods, the team found that the Anchor-free detector had a 51 per cent improvement in speed. SimOTA Tag Assignment Policy: To obtain high-quality positive samples, the team used the SimOTA algorithm that dynamically allocates positive samples to improve detection accuracy. SIoU bounding box regression loss: YOLOv6 adopts the SIoU bounding box regression loss function to supervise the learning of the network. The SIoU loss function redefines the distance loss by introducing a vector angle between required regression. This improves the regression accuracy, resulting in improved detection accuracy. YOLOv5 vs MT-YOLOv6 According to the benchmarking performed by Meituan’s team, YOLOv6 outperforms YOLOv5 and other YOLO models in terms of accuracy and speed on the COCO dataset. YOLOv6-nano achieved a 35 per cent AP accuracy on the COCO dataset; it could reach 1242 FPS performs, and when compared to YOLOv5-nano, the accuracy was up by 7 per cent AP and speed by 85 per cent. YOLOv6-tiny recorded 41.3 per cent AP accuracy on COCO, and compared to YOLOv5-s, the accuracy was increased by 3.9 per cent and speed by 29.4 per cent. At last, YOLOv6-s obtained an accuracy of 43.1 per cent on COCO. It achieves a performance of 520 FPS compared to YOLOX-s – the accuracy is 2.6 per cent AP better and speed by 38.6 per cent. Credit: Meituan As per a few discussion threads and blogs, YOLOv6 is not a straight upgrade of YOLOv5 from Ultralytics. It is observed that while MT-YOLOv6 can detect smaller objects more reliably, it flickers as compared to YOLOv5 and struggles with close-up objects. Compared to YOLOv5, MT-YOLOv6 lacks stability but makes up for impressive capabilities in small object detection in densely packed environments. In terms of flexibility, YOLOv5 uses YAML, and YOLOv6 defines the model parameters directly in Python. It was observed that YOLOv5 is more customisable than YOLOv6. What future does YOLOv6 hold? Meituan’s team wants to improve the full range of the model further and advance the detection performance. The team said that the model will support ARM platform deployment and full-chain adoption, such as quantitative distillation. They would like to explore the generalisation performance of YOLOv6 in different business scenarios.","excerpt":"According to the team, MT-YOLOv6 has carried out improvements and optimisations at the algorithmic level like training strategies and network structure and has displayed impressive results in terms of accuracy and speed when tested on COCO datasets.","categories":["AI Features"],"tags":["Computer Vision","YOLO"],"author_name":"Shraddha Goled","publish_date":"2022-07-01T15:00:17","publication_year":"2022","word_count":811,"keywords":["Go","AWS","AI","neural network","ML","computer vision","Python","object detection","YOLO","Computer Vision","R"],"extracted_tech_keywords":["AI","ML","neural network","computer vision","object detection","AWS","Python","R","Go","YOLO"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-does-yolov6-compare-against-yolov5\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10014508,"title":"Top 8 AI-Powered Features Released By Adobe In 2020","content":"Adobe has already been using emerging technologies like AI in several Photoshop features, including Object Selection Tool, Curvature Pen Tool, among others. During this year’s annual Adobe event MAX, Adobe announced a number of innovative features that are made powerful by fusing AI technologies. Here is a compilation of the top AI-powered features that are released by Adobe in 2020. Neural Filters Neural Filters is one of the significant breakthroughs in AI-powered creativity. According to its developers, it is the beginning of a complete reimagination of filters and image manipulation inside Photoshop. It is a new workspace inside Photoshop that introduces new non-destructive filters to help you explore creative ideas in seconds. This version of Neural Filters consists of a large set of new filters, and many of its existing filters are still in beta. This feature is part of a new ML platform that will develop as well as get better over time – expanding on what’s attainable exponentially. Sky Replacement The new Sky Replacement has the capability to separate the sky from the foreground and allows users to create intuitive images with dynamic skies in just a few clicks. A user can either select the sky with ‘Select>Sky’ and edit it or select the ‘Edit>SkyReplacement’ option and choose a new sky from the database or add their own from their inventory. In such a case, the new Sensei-powered ML models perform the masking and blending task. Sky Replacement uses cutting-edge algorithms to harmonise the foreground of an image with the sky. Discover Panel The Discover panel is packed with both tools and tips to help users work in a faster and efficient manner. Discover panel utilises AI techniques to provide recommendations based on both the users and their work. According to the official blog post, this new tool inside Photoshop will continue to add AI capabilities that enhance the experience with the product. This new feature merges an expanded library of in-app learning content, along with new step-by-step tutorials as well as powerful search functionality. Refine Edge Selections The features, Object Aware Refine Edge and Refine Hair leverage AI techniques to improve choices that include objects like hair or other complicated backgrounds. The Object Aware Refine mode is used to set the Refine edge mode in order to make better and faster selections. According to a blog post, the Object Aware algorithm has been trained to understand objects in the scene and thus work better with similarly-coloured or similarly-textured backgrounds. On the other hand, the Refine Hair mode has been located in the Options bar, helping users to refine the selected hair. Speech to Text in Premiere Pro A new Adobe Sensei feature for videos — Speech to Text in Premiere Pro allows users to create a transcript from their video automatically, and then produce automatic captions on the timeline. It is powered by the Adobe Sensei, where the Auto Captions leverages machine learning techniques to mirror the pacing of spoken dialogue as well as match it to the video timecode. New captions workflow in Premiere Pro Currently, in the public beta version, new captions workflow in Premiere Pro offers captions as an integral part of the editorial section. This feature provides a reimagined user experience for working with captions and subtitles in Premiere Pro with the help of the design tools in the Essential Graphics panel for customising as well as stylising text. Roto Brush 2 in After Effects This AI feature gives both editors and VFX artists the capability to intervene a foreground object from its scene in a faster and intuitive manner. Roto Brush 2 uses Adobe Sensei to track selected objects in an automatic and frame by frame fashion. Also, the feature improved the original Roto Brush tool, making advanced rotoscoping available to a wide range of video creators. Not only it saves countless hours of effort but also works with difficult footage. Speech-Aware Animation in Character Animator Speech-Aware Animation in Character Animator is an AI feature that provides new options for creating content by automatically generating animation from recorded speech, including corresponding head and eyebrow movements.","excerpt":"Adobe has already been using emerging technologies like AI in several Photoshop features, including Object Selection Tool, Curvature Pen Tool, among others. During this year’s annual Adobe event MAX, Adobe announced a number of innovative features that are made powerful by fusing AI technologies.       Here is a compilation of the top AI-powered features that are […]","categories":["AI Trends"],"tags":["Adobe","Adobe AI","Adobe Sensei","machine learning blending","PowerBI"],"author_name":"Ambika Choudhury","publish_date":"2020-12-17T14:00:00","publication_year":"2020","word_count":683,"keywords":["Go","machine learning","programming_languages:R","Adobe","AI","ML","Adobe Sensei","machine learning blending","programming_languages:Go","Adobe AI","PowerBI","RAG","ViT","R"],"extracted_tech_keywords":["AI","machine learning","ML","RAG","R","Go","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-8-ai-powered-features-released-by-adobe-in-2020\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10064879,"title":"How to actuate data insights using MLOps","content":"“Starting in the 1950s, AI has a long history of being the next big thing,” said Megha Sinha, vice president of Digital – Data Science, AI, ML at Genpact in her talk titled “MLOps – The Strategic move to actuate Data Insights”.  She discussed MLOps at length during her session at The Rising. Today, every enterprise aspires to be data-driven, she said. While it has become imperative for business leaders to actuate data insights for business growth, studies show only half of AI proof of concepts are scaled to production. MLOps holds the key to create production-ready, scalable AI solutions. “MLOps is a set of practices that defines how to deliver solutions and build pipelines,” said Megha. Watch all the recorded sessions of Rising 2022 here>> AI for business AI has gone through a series of ups and downs, including two extended periods of doldrums, dubbed AI Winters. Businesses started taking an interest in AI through the 1980s. However, the excitement petered for the lack of workable AI-based solutions in the market owing to a dearth in computing power and data. “Today, 85% of AI and machine learning projects fail to deliver. Only half of our projects reached from prototype to production stage. Now, while the latest findings show no increase in AI adoption, many companies are beginning to incorporate AI and ML in different functions slowly and steadily. But the question now arises, are we staring at yet another AI winter? Megha asked. The need of the hour is integrated custom design applications with ML-powered software and pipelines that can deliver repeatable experiences to clients across the breadth and depth of AI and ML. The companies are investing heavily to improve their product portfolio, capture wider markets, and in the process is keeping the AI winter at bay, she said. MLOps strategy Megha then discussed the three fundamental blocks of AI for businesses: DataML model.The code to integrate all components “While it sounds very easy to do, changes in one part of the pipeline can trigger a change throughout the system. This proves the principle of changing something, changes everything,” said Megha. MLOps helps streamline this complicated process. It helps businesses identify best practices, methodologies, tools, and technologies to unify the entire pipeline. Additionally, it is important to understand business analytics to answer questions such as where exactly do we need to invest, how to build a change management strategy etc. The answers will differ from organisation to organisation. Megha spoke about five focus areas for AI in business: Define team structuresUnified ML platforms to build, deploy and monitor. Continuous model monitoring and automated deployment to reduce the release cycle and failure rates.IT team for infrastructure and asset provisioning.Responsible AI practices at all stages. After studying the key areas, businesses need to create an MLOps strategy. First, they need to identify the objectives and understand the need for MLOps in the first place. Then, explore questions like is your system big enough, do you intend to scale it, etc. Follow it up with assessing the company’s tangibles- the number of models, infrastructure and the team size. Then, evaluate the intangibles by designing the culture. Follow this by defining the scale and designating change management. Lastly, build the metrics and then actuate. Megha also discussed an interesting case study from Genpact for an insurance client looking to streamline the underwriting process for a global life reinsurer. The company created two models where one predicted the likelihood of acceptance concerning underwriting appetite as an initial triage to filter out potential declines and the likelihood to prioritise further the accepted risks for underwriting action to achieve the target of responding to the best set of 30-35% risks. The result achieved a top of line 12-15% improved risk selection, enhanced customer experience and more strategic power to the underwriter.","excerpt":"MLOps helps businesses identify best practices, methodologies, tools, and technologies to unify the entire pipeline.","categories":["Deep Tech"],"tags":["MLOps"],"author_name":"Avi Gopani","publish_date":"2022-04-13T13:00:00","publication_year":"2022","word_count":636,"keywords":["data science","Go","machine learning","AI","ML","MLOps","Scala","analytics","model monitoring","R"],"extracted_tech_keywords":["AI","machine learning","ML","data science","analytics","MLOps","model monitoring","R","Go","Scala"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/how-to-actuate-data-insights-using-mlops\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":48516,"title":"With $2 Million In Series A, This Mumbai-Based Startup Is Using AI To Help Companies Hire Better","content":"As of 2019, organisations across the globe are facing problems in hiring and keeping talented individuals. As the job market is growing significantly, talent acquisition and recruiting remain critical issues for most enterprises. But now, Indian companies have started to embrace technology to improve hiring outcomes. To mitigate such problems a startup called Shortlist is helping Indian enterprises to bridge the gap between applicants and job profiles. This week, Analytics India Magazine reached out to Sudheer Bandaru, CTO, Shortlist Professionals to understand the company’s end-to-end functioning. Founded by Paul Breloff, Simon Desjardins and Matt Schnuck, Shortlist is a Mumbai-based recruiting technology company that helps employers in major countries like Kenya, Rwanda and Nigeria in Africa and India build and develop world-class teams. It is a talent matching platform which makes it easy for companies to create smarter job applications on the platform — combining chat, skills assessments and audio or video interviews, among others. Launched in 2016 in India and Kenya, the company has launched in the two markets almost simultaneously. Currently, it has three offices across two countries, along with consultants and board members in New York City, Washington DC, San Francisco and Mauritius (not to mention clients in 12+ countries). Earlier this year, the company has closed a $2 million Series A financing by Blue Haven Ventures, with participation from Zephyr Acorn, Compass Venture Capital, Potencia Ventures, and others. Flagship Products The Shortlist platform screens candidates using predictive chat-based interviews and competency-based assessments, so that they can be matched with great jobs based on skills and interests. It creates an interactive flow for candidates applying for a job, as well as for the clients reviewing them on a tracking platform. The product offering ranges from a full-service tech-backed recruiter model to lighter-touch software which allows employers to efficiently and objectively identify top talent for junior- to more senior roles. Use Of Artificial Intelligence & Machine Learning According to Bandaru, Shortlist follows the data-first culture. The core focus of the company is to build a recommendation engine that helps employers identify potential hires for a job as soon as candidates start applying on their platform. The recommendation model is an AI-based model based on supervised learning which focuses on finding the right kind of candidates by studying our past hiring data. It also stresses to feed more data into the training sets and focuses on increasing their prediction accuracy. The company has a comprehensive data engineering architecture with a data lake hosted on AWS. It has also invested heavily in the areas of data engineering and visualisation tools. When asked about how the company uses AI and ML in their products and services Bandaru explained, “While most players in the market focus on scraping CVs or Profiles for candidate information, our platform is unique in collecting data from their performance on assessments which mirror the job role. The recommender system that we have built puts important in the assessment scores to figure out candidates who have the competency to perform a role, rather than only have demonstrated experience for that role.” Technology Stack The startup is leveraging on the cloud services offered by Amazon Web Services (AWS). The company also leveraging some popular data science tools like MySQL, RedShift Data Lake, Python, Jupyter, AWS SageMaker, Power BI and EC2 Instances. Below is a list of technologies that the company is utilising to make a robust product. Backend Technology: Java, J2EE, Spring MVC Front-end Stack: HTML, CSS, JavaScript, Angular JS, Bootstrap Server: Tomcat Database: MySQL, MongoDB, RedShift Caching: Redis Distributed Architecture: Microservices, Spring boot Logging: ELK stack QA Automation: Selenium What Next Talking about competitors, Bandaru said that their potential competitors are a mix of traditional recruitment companies who have been following legacy recruitment processes and the new age companies who are exploring recruitment tools through technology. And for the future, the company will be driven by data and a keen focus on generating more value for the job seekers by creating functional communities, generate self-awareness, career paths, and job opportunities.","excerpt":"As of 2019, organisations across the globe are facing problems in hiring and keeping talented individuals. As the job market is growing significantly, talent acquisition and recruiting remain critical issues for most enterprises. But now, Indian companies have started to embrace technology to improve hiring outcomes. To mitigate such problems a startup called Shortlist is […]","categories":["AI Startups"],"tags":["AI Companies","Startups"],"author_name":"Ambika Choudhury","publish_date":"2019-10-21T12:56:12","publication_year":"2019","word_count":671,"keywords":["data science","machine learning","artificial intelligence","AWS","AI","ML","RAG","microservices","analytics","AI Companies","Startups","Jupyter"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","data science","analytics","Jupyter","RAG","AWS","microservices"],"url":"https:\/\/analyticsindiamag.com\/ai-startups\/with-2-million-in-series-a-this-mumbai-based-startup-is-using-ai-to-help-companies-hire-better\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10005374,"title":"How AWS And Formula 1 Used ML To Find Fastest Racer In The History Of The Sport","content":"“F1 and Amazon Machine Learning Solutions Lab took a full year to build the algorithm that led to the fastest driver.” Formula 1 has been working with Amazon Web Services (AWS) to rank their racers. After a year of algorithmic heavy lifting, the results are out now. Ayrton Senna, the three-time world champion from Brazil came out on top, followed by the seven-time champion, Michael Schumacher with a time differential of +0.114 second. Whereas current World Champion Lewis Hamilton featured at 3rd position with a relative time of +0.275 seconds. F1 is a brutal sport. The room for error at the top is almost non-existent. So, how and why was machine learning leveraged by F1 analysts? Senna Or Schumacher? How Do We Know Ayrton Senna The project ‘Fastest Driver’, was headed by Formula 1’s Director of Data Systems Rob Smedley and Amazon ML Solutions Lab’s Principal Scientist Dr Priya Ponnapalli. To keep the rankings fair, the team found ways to keep the model immune to outliers like crashes, car failures, and changing weather conditions. The data consisted of timesheets from every qualifying session since 1983. The outliers were ditched, and the data was normalised to create a complex network of drivers’ performance relative to their teammates. These teammates, wrote the team at F1, to be considered, should have completed at least five qualifying sessions against each other. Attributes such as age were also considered along with the comebacks of few drivers after a break. Drivers who dominated their teammates or did well against strong performers were given a higher ranking. Source: F1 While these results did not go well with the F1 fans in popular forums, F1’s director of data systems Smedley is certain about the methods they have incorporated in generating the rankings. “Qualifying speed is something that we can be really quite clear about,” said Smedley.” The Director of data systems touched upon the various nuances of modelling and how there is no ambiguity regarding the data points that they have considered. For instance, a qualifying lap is a single lap, where there are two racers in the same car doing a single lap. The driver who does better will end up with the better lap. “There’s not much ambiguity about that single data point,” added Smedley. There Is More To AWS And F1 Collaboration Source: AWS The unlikely collaboration between F1 and AWS came into light last year, with the launch of ‘F1 Insights powered by AWS’, a series of graphics that educate the viewers with data analytics. According to the team behind this initiative, these graphics gave key and unseen insights into the inner workings of Formula 1 and brought them out into full public view for the first time. These latest F1 Insights graphics take into account cornering performance, straight-line performance and car balance or handling — the key aspects teams, work to improve — and demonstrate them to the public with great visuals. So, these cloud-based machine learning algorithms not only help the analysts in the pit but also offer an enriched experience to those watching at home. The fans get to understand the microscopic details that come into play at dizzy speeds. The graphics utilise car telemetry, timing data and all the other data input feeds to demonstrate inner team workings to the viewers. The historical data is fed to Amazon SageMaker complex machine learning algorithms, and race strategies and outcomes are predicted with increasing accuracy. These models were used even to predict future scenarios using refreshed real-time data as Grand Prix races unfold. “As drivers are more often than not the most expensive asset of the team, it is important that the selection process is as robust as possible,” said Smedley.","excerpt":"“F1 and Amazon Machine Learning Solutions Lab took a full year to build the algorithm that led to the fastest driver.” Formula 1 has been working with Amazon Web Services (AWS) to rank their racers. After a year of algorithmic heavy lifting, the results are out now. Ayrton Senna, the three-time world champion from Brazil […]","categories":["Deep Tech"],"tags":["AWS"],"author_name":"Ram Sagar","publish_date":"2020-08-24T10:00:49","publication_year":"2020","word_count":619,"keywords":["Go","Amazon SageMaker","machine learning","AWS","AI","mlops_tools:SageMaker","ML","RAG","analytics","R"],"extracted_tech_keywords":["AI","machine learning","ML","analytics","Amazon SageMaker","RAG","AWS","R","Go","mlops_tools:SageMaker"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/aws-formula-one-machine-learning-senna-schumacher-race\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10019931,"title":"One Plus Co-Founder Raises $15 Million For &#8216;Nothing&#8217;","content":"OnePlus co-founder Carl Pei’s new venture Nothing has raised $15 million in a funding led by GV (formerly Google Ventures), taking the total external investment tally to $22 million. The fresh capital will be used to expand the team and operations, for R&D, and  product development. “We plan to aggressively grow the company, in particular, our R&D and design capabilities, to realise our mission of removing barriers between people and technology,” Pei, CEO and co-founder of Nothing, said. Nothing is a London-based consumer technology company founded by Carl Pei and team. “Nothing’s mission is to remove barriers between people and technology to create a seamless digital future. We believe that the best technology is beautiful, yet natural and intuitive to use. When sufficiently advanced, it should fade into the background and feel like nothing, ” Pei said at Nothing’s launch. Nothing wants to inspire people to believe in the positive potential of technology again. Expressing his belief in Pei’s ‘vision for smart devices’, Tom Hulme, General Partner at GV, said the team has high confidence that Nothing will have a meaningful impact on the consumer technology industry. Nothing’s backers include Twitch co-founder Kevin Lin, Reddit CEO Steve Huffman, Beme co-founder and YouTube celebrity Casey Neistat, and CRED founder Kunal Shah. The London-based firm said it would offer the general public a chance to invest in the company as part of the series A round. More details are awaited on this front.","excerpt":"OnePlus co-founder Carl Pei’s new venture Nothing has raised $15 million in a funding led by GV (formerly Google Ventures), taking the total external investment tally to $22 million.  The fresh capital will be used to expand the team and operations, for R&D, and  product development. “We plan to aggressively grow the company, in particular, […]","categories":["AI News"],"tags":["Funding","investment","nothing","oneplus"],"author_name":"Shraddha Goled","publish_date":"2021-02-10T11:28:48","publication_year":"2021","word_count":241,"keywords":["Go","API","Funding","investment","funding","programming_languages:R","AI","nothing","ML","programming_languages:Go","Git","oneplus","R"],"extracted_tech_keywords":["AI","ML","R","Go","Git","API","funding","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/one-plus-co-founder-raises-15-million-for-nothing\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":32729,"title":"India Doubles Down On Use Of AI In Defence With AI-Enabled Night Vision Device","content":"As part of its strategy to increase commitment to AI, the Indian military will soon use AI-enabled device for night vision. According to news reports, the Army Design Bureau has built a device that will warn soldiers of unusual activity across different terrains and in high altitude areas. A senior officer from the Indian military shared that AI-based night vision device will be helmet mounted and in addition to this, there will be a wristband attached to it, which vibrates when the device picks up on movement in the area. News reports hint that the device was mooted during the ARTECH summit. In addition to the AI-enabled device for night vision, the soldiers also have access to live translator, that helps soldiers positioned along Line of Actual Control, China to interact with Chinese soldiers. The device translates Mandarin to Indian languages. In addition to this, the Ministry of Defence has also stepped up the use of AI. According to news reports, a multi-stakeholder task force on the strategic implementation of AI was set up which included members from the Government. The members included DRDO, Defence Public Sector Undertakings, National Cybersecurity Coordinator, ISRO and BARC. As per reports, DPSU has been directed to build AI-enabled products and AI-based tools for sensor data analysis, predictive maintenance and situational awareness. A INR 73.9 crore project has been sanctioned under CAIR to build AI-based solutions to improve intelligence collation and analysis capabilities of Indian defence. In a similar vein, another project called Energy Harvesting Based Infrared Sensor Network for Automated Human Intrusion Detection was sanctioned for about INR 1.8 crore.","excerpt":"As part of its strategy to increase commitment to AI, the Indian military will soon use AI-enabled device for night vision. According to news reports, the Army Design Bureau has built a device that will warn soldiers of unusual activity across different terrains and in high altitude areas. A senior officer from the Indian military […]","categories":["AI News"],"tags":[],"author_name":"Richa Bhatia","publish_date":"2019-01-03T10:22:47","publication_year":"2019","word_count":266,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","ViT","R"],"extracted_tech_keywords":["AI","R","Go","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/india-doubles-down-on-use-of-ai-in-defence-with-ai-enabled-night-vision-device\/","complexity_score":2,"technical_depth":6,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":68366,"title":"Video Game Sales Prediction: Weekend Hackathon #10","content":"MachineHack is back again with another exciting hackathon for this weekend, and this time with a new look and exciting features. Problem Statement & Description The gaming industry is certainly one of the thriving industries of the modern age and one of those that are most influenced by the advancement in technology. With the availability of technologies like AR\/VR in consumer products like gaming consoles and even smartphones, the gaming sector shows great potential. In this hackathon, you as a data scientist must use your analytical skills to predict the sales of video games depending on given factors. Given are 8 distinguishing factors that can influence the sales of a video game. Your objective as a data scientist is to build a machine learning model that can accurately predict the sales in millions of units for a given game. Click here to participate Data Description:- The unzipped folder will have the following files. Train.csv –  3506 observations.Test.csv –  1503 observations.Sample Submission – Sample format for the submission. Target Variable: SalesInMillions The datasets will be made available for download on June 26th, Friday at 6 pm IST This hackathon and the bounty will expire on June 29th, Monday at 7 am IST Below are the file formats for the provided data: – Train.csv Test.csv Sample_Submission.xlsx Click here to participate Bounties The top 3 competitors will receive a free pass to the Rising 2020. Know more about the Rising 2020 here. Rules One account per participant. Submissions from multiple accounts will lead to disqualificationThe submission limit for the hackathon is 10 per day after which the submission will not be evaluatedAll registered participants are eligible to compete in the hackathonThis competition counts towards your overall ranking pointsWe ask that you respect the spirit of the competition and do not cheatThis hackathon will expire on 29th June, Monday at 7 am ISTUse of any external dataset is prohibited and doing so will lead to disqualification Evaluation The leaderboard is evaluated using RMSE for the participant’s submission. Click here to participate","excerpt":"MachineHack is back again with another exciting hackathon for this weekend, and this time with a new look and exciting features.  Problem Statement & Description The gaming industry is certainly one of the thriving industries of the modern age and one of those that are most influenced by the advancement in technology. With the availability […]","categories":["Deep Tech"],"tags":["Hackathons","Machinehack","Weekend Hackathon"],"author_name":"Amal Nair","publish_date":"2020-06-26T17:52:50","publication_year":"2020","word_count":337,"keywords":["machine learning","Weekend Hackathon","programming_languages:R","AI","Machinehack","Hackathons","R"],"extracted_tech_keywords":["AI","machine learning","R","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/video-game-sales-prediction-hackathon\/","complexity_score":2,"technical_depth":4,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":21092,"title":"Hyderabad Startup eKincare Bags $1.5M In Series-A Funding","content":"Hyderabad-based startup eKincare received a funding of $1.5M today from prominent investor firms such as Venture East, Eight Roads, Touchstone Equities, Endiya Partners and included Padmashri awardee BVR Mohan Reddy in the investing round. The startup provides a personal AI assistant that captures users’ medical data such as physical records and health history thereby suggesting personalised health improvements to promote a healthy lifestyle. It is India’s first healthcare startup that provides medical support and is backed by Fortune 500 companies such as Unilever and Barclays among others. The company has already integrated with various partners over 2000 locations in India to offer its solution. eKincare’s success can be attributed to the fact that it is disrupting the way ‘employee health spends’ are perceived in organizations across India. It has been able to do so by an extensive use of technologies like data science. Founder & CEO, Kiran Kalakuntla said, “We have seen a lot of interest from corporates to leverage our platform for their employees, keep a track of the overall organization’s wellness metrics and maximize the ROI out of their wellness budgets. We have defined and established the need for an enterprise-level platform that delivers an unparalleled personalized experience for corporates, employees and consumers at large and help them achieve and sustain their optimal health.” Srikant General Partner, Ventureast was quoted as saying that they constantly invest in disruptive startups that address larger market or consumer related challenge, and eKincare is helping to take control of health outcomes while minimizing recurring costs. “We are confident that the team will cement strategic partnerships, and continue to lead the ‘predictive diagnosis’ space through technology innovations,” he had said. The investors expressed candid interest in the startup due to its increased contribution in the field of digital healthcare, ensuring a cost reduction in large companies. eKincare also received a no-obligation grant of $15,000 from Swiss Re, the top reinsurer and capital management firm in the world, as part of its global accelerator program.","excerpt":"Hyderabad-based startup eKincare received a funding of $1.5M today from prominent investor firms such as Venture East, Eight Roads, Touchstone Equities, Endiya Partners and included Padmashri awardee BVR Mohan Reddy in the investing round. The startup provides a personal AI assistant that captures users’ medical data such as physical records and health history thereby suggesting […]","categories":["AI News"],"tags":["medical","Startups"],"author_name":"Abhishek Sharma","publish_date":"2018-01-30T07:57:33","publication_year":"2018","word_count":331,"keywords":["data science","API","funding","AI","innovation","Git","RAG","medical","GAN","Startups","R","startup"],"extracted_tech_keywords":["AI","data science","RAG","R","Git","API","GAN","innovation","startup","funding"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/hyderabad-startup-ekincare-bags-1-5m-in-series-a-funding\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10023997,"title":"Comprehensive Guide To CLARANS Clustering Algorithm","content":"CLARANS (Clustering Large Applications based on RANdomized Search) is a Data Mining algorithm designed to cluster spatial data. We have already covered K-Means and K-Medoids clustering algorithms in our previous articles. This article talks about another clustering technique called CLARANS along with its Pythonic demo code. CLARANS was introduced by Raymond T. Ng and Jiawei Han  of IEEE Computer Society (research paper). CLARANS is a partitioning method of clustering particularly useful in spatial data mining. We mean recognizing patterns and relationships existing in spatial data (such as distance-related, direction-relation or topological data, e.g. data plotted on a road map) by spatial data mining. Why CLARANS algorithm? As mentioned in our K-Medoids algorithm’s article, the K-Medoids clustering technique can resolve the limitation of the K-Means algorithm of being adversely affected by noise\/outliers in the input data. But K-Medoids proves to be a computationally costly method for considerably large values of ‘k’ (number of clusters) and large datasets. The CLARA algorithm was introduced as an extension of K-Medoids. It uses only random samples of the input data (instead of the entire dataset) and computes the best medoids in those samples. It thus works better than K-Medoids for crowded datasets. However, the algorithm may give wrong clustering results if one or more sampled medoids are away from the actual best medoids. CLARANS algorithm takes care of the cons of both K-Medoids and CLARA algorithms besides dealing with difficult-to-handle data mining data, i.e. spatial data. It maintains a balance between the computational cost and the influence of data sampling on clusters’ formation. Steps of CLARANS algorithm Select ‘k’ random data points and label them as medoids for the time being.Select a random point say ‘a’ from the points picked in step (1), and another point say ‘b’ which is not included in those points. We would already have the sum of distances of point ‘a’ from all other points since that computation is required for selecting the points in step (1). Perform similar computation for point ‘b’.If the sum of distances from all other points for point ‘b’ turns out to be less than that for point ‘a’, replace ‘a’ by ‘b’.The algorithm performs such a randomized search of medoids ‘x’ times where ‘x’ denotes the number of local minima computed, i.e. number of iterations to be performed, which we specify as a parameter. The set of medoids obtained after such ‘x’ number of steps is termed as ‘local optimum’.A counter is incremented every time a replacement of points is made. The process of examining the points for possible replacement is repeated till the counter does not exceed the maximum number of neighbors to be examined (specified as a parameter).The set of medoids obtained when the algorithm stops is the best local optimum choice of medoids. Image source: Research paper Practical Implementation Here’s a demonstration of using CLARANS algorithm on the sklearn library’s Breast Cancer Wisconsin dataset. Though the dataset is primarily used for binary classification tasks, we use it to show how CLARANS algorithm can form separate clusters of the constituent data points falling under one of the two target categories (‘malignant’ or ‘benign’). The pyclustering data mining library has been used here for Pythonic implementation of CLARANS. The code has been implemented using Google colab with Python 3.7.10 and pyclustering 0.10.1.2 versions. Step-wise explanation of the code is as follows: Install pyclustering library. !pip install pyclustering Import required libraries and modules #Class for implementing CLARANS algorithm from pyclustering.cluster.clarans import clarans #To execute a function with execution time recorded from pyclustering.utils import timedcall #sklearn package for using a toy dataset from sklearn import datasets #Class for plotting multi-dimensional data from pyclustering.cluster import cluster_visualizer_multidim Import the Breast Cancer dataset bc_dataset =  datasets.load_breast_cancer() #Display the dataset bc_dataset Sample condensed output: Extract the data points from the loaded dataset. bc_data = bc_dataset.data Convert the dataset from a numpy array to a list because a list of lists is fed as an input to the CLARANS’ implementation of the pyclustering library. bc_data = bc_data.tolist() Display the data in the form of a list print(bc_data[:5]) Sample condensed output: [[17.99, 10.38, 122.8, 1001.0, 0.1184, 0.2776, 0.3001, 0.1471, 0.2419, 0.07871, 1.095, 0.9053, 8.589, 153.4, 0.006399, 0.04904, 0.05373, 0.01587, 0.03003, 0.006193, 25.38, 17.33, 184.6, 2019.0, 0.1622, 0.6656, 0.7119, 0.2654, 0.4601, 0.1189], [20.57, 17.77, 132.9, 1326.0, 0.08474, 0.07864, 0.0869, 0.07017, 0.1812, 0.05667, 0.5435, 0.7339, 3.398, 74.08, 0.005225, 0.01308, 0.0186, 0.0134, 0.01389, 0.003532, 24.99, 23.41, 158.8, 1956.0, 0.1238, 0.1866, 0.2416, 0.186, 0.275, 0.08902],… Instantiate the clarans class. clarans_obj = clarans(bc_data, 2, 3, 5) “”” where, ‘bc_data’ is the input data fed as a list of list of objects to be clustered ‘2’ is number of clusters to be formed (since there are 2 target categories); ‘3’ is the number of obtained local minima and ‘5’ is the maximum number of neighboring data points examined “”” process() method analyzes the clusters as per the CLARANS algorithm. We call the process() method and encapsulate it in the call to timedcall() function so that the time taken for executing process() method also gets recorded. #timedcall() returns a tuple containing the execution time and result of executing the function (tks, res) = timedcall(clarans_obj.process); #Print the execution time print(\"Execution time : \", tks, \"\\n\"); Output: Execution time :  80.50073736700006 Get the clusters allocated by the algorithm clst = clarans_obj.get_clusters() Get the list of medoids of the clusters allocated by the algorithm. med = clarans_obj.get_medoids() Print the results print(\"Index of clusters' points :\\n\",clst) print(\"\\nLabel class of each point :\\n \",bc_dataset.target) print(\"\\nIndex of the best medoids : \",med) Sample condensed output: Here, the input data has 30 features. Cluster_visualizer class of the pyclustering library can be used to visualize the 1D, 2D or 3D data. While for more than three-dimensional data, cluster_visualizer_multidim class can be used as follows: vis = cluster_visualizer_multidim() “”” Append the list of clusters; specify the clusters list, list of data points, marker sign and marker size to be used for visualization “”” vis.append_clusters(clst,bc_data,marker=\"*\",markersize=5) #Display the clusters formed in multiple dimensions vis.show(pair_filter=[[1,2],[1,3],[27,28],[27,29]],max_row_size=2) “”” pair_filter parameter specifies the list of feature pairs for which the clusters will be visualized. max_row_size parameter specifies the maximum number of rows across which the plot will be spread. “”” Output: Google colab notebook of the above implementation is available here. References CLARANS Research paperpyclustering documentationVideo tutorial .","excerpt":"CLARANS (Clustering Large Applications based on RANdomized Search) is a Data Mining algorithm designed to cluster spatial data.","categories":["Deep Tech"],"tags":["Guide","Visualize Spatial Data"],"author_name":"Nikita Shiledarbaxi","publish_date":"2021-04-16T18:00:00","publication_year":"2021","word_count":1045,"keywords":["Go","NumPy","TPU","programming_languages:R","AI","Visualize Spatial Data","Ray","Colab","Python","programming_languages:Python","R","Guide"],"extracted_tech_keywords":["AI","Ray","Colab","NumPy","TPU","Python","R","Go","programming_languages:Python","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/comprehensive-guide-to-clarans-clustering-algorithm\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10066033,"title":"No CEO cares about moving to a shiny, new cloud database unless it improves business outcomes: Subram Natarajan, Google Cloud India","content":"In his keynote speech for the first edition of the Data Engineering Summit (DES ‘22) hosted by Analytics India Magazine, Google Cloud India’s Subram Natarajan, the director of customer engineering, spoke about how cloud infrastructure can be transformative in a company’s journey. Natarajan discussed the evolution of cloud computing, mentioning the two major shifts in data transformation. The first phase was when VM Cloud was used to spin off the infrastructure requirements for organisations on to cloud. The second phase of the shift didn’t just restrict itself to the basic infrastructure frameworks but moved everything from compute to network to storage on to cloud. Watch the complete session here. “Early cloud migrations weren’t transformative enough for organisations or the industry on the whole. It did not change how business leaders worked outside of the IT industry,” he said. With the widespread prevalence of data, Natarajan notes that we need new ways to use data and democratise it. But that was only one aspect of it. There was an urgent need to migrate data in a way that was responsible and secure, while maintaining accessibility. “If data science teams need to continuously request the IT team for permission to view datasets, the process becomes inefficient,” he said. This wasn’t how it should be, according to Natarajan. Source: Statistica, The growth in data New trends in data transformation Data engineering has progressed swiftly in a manner that most industry leaders are looking to move towards data warehousing so that they can begin using machine learning, with business intelligence (BI) as the final goal. For companies to perform machine learning that can contribute to the business value, the available volume of data should be large. According to Natarajan, there were three new angles that should be considered while shifting to a new platform. Serverless computing: Increases productivity and makes the new platform simple. This feature also reduces infrastructure management and overhead costs. Portability: This ensures that as the system slowly matures, it doesn’t have to restrict itself to the platform it was built on. “The compute engines should be running closer to the data instead of the other way around,” Natarajan added.Democratised: Natarajan said that data shouldn’t be limited to the hands of a few people only. “Everyone who can use data to leverage their decision-making should be able to access it,” he stated. Source: Apache Beam testing playground Natarajan then went on to describe the benefits of Apache Beam; an “open source unified programming model to define and execute data processing pipelines, including ETL.” Google had donated the project to help simplify the mechanics of large-scale data processing while continuing to contribute to it. The popular Apache Spark, which has investors like IBM and Huawei, is where the most development is currently taking place, Natarajan noted. The Spark Runner executes Beam pipelines on top of Apache Spark, providing batch and streaming pipelines. Source: Google Cloud, Cloud Foundation toolkit Companies are now realising the perils of dealing with bad data. Quite simply, Natarajan says, “Bad data leads to poor insights.” There are other aspects that must be taken care of now, like data cataloguing, data security and bringing in business taxonomy. Infrastructure as a code is becoming increasingly important in the process so that larger workloads can be processed at cheaper prices. Natarajan also said that similar to Amazon or Microsoft, Google has also built cloud foundation toolkit modules that are ready-made templates. These toolkits help ease the lives of data practitioners to a great extent. Business engagement It is essential, Natarajan said, that business leaders are kept in the mix when shifting data platforms. “It is important to break down the silos for any meaningful change in the organisation. No CEO cares about moving to a shiny, new cloud database. They care about business outcomes that the project will bring by helping analyse data better so that better decisions are made to increase the value in customers’ lives,” he explained. Source: Google, Features of a successful Cloud COE In this regard, Cloud Centers of Excellence, or COEs, have become a vital part of initiating a transformational journey in organisations. “They can be a huge enabler and a conduit to power change. A well-formed COE team can have huge benefits when you want to oversee an outage and scale,” he said. Natarajan said the COE team did not consist of philosophical architects who envision a perfect future, but they are a group of passionate individuals who work consistently. Natarajan said that he recommended that instead of focusing on a perfect result, businesses should start small and then grow. “It is important to collect evidence of your successes and strengthen your case for hybrid adoption. This also brings sceptics from the sidelines to the fore. Once the business notices an impact, they start caring about it,” he added. Finally, Natarajan said that it was important to celebrate even the small wins to build a culture that remembers to encourage the work that has been done. REGISTER HERE TO ACCESS THE CONTENT","excerpt":"It is essential, Natarajan said, that business leaders are kept in the mix when shifting data platforms.","categories":["Global Tech"],"tags":["cloud migration","Data Engineering"],"author_name":"Poulomi Chatterjee","publish_date":"2022-05-02T10:00:00","publication_year":"2022","word_count":837,"keywords":["data science","Go","machine learning","cloud migration","AI","cloud computing","Apache Spark","serverless","RAG","Data Engineering","analytics","R"],"extracted_tech_keywords":["AI","machine learning","data science","analytics","RAG","cloud computing","serverless","Apache Spark","R","Go"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/no-ceo-cares-about-moving-to-a-shiny-new-cloud-database-unless-it-improves-business-outcomes-subram-natarajan-google-cloud-india\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":29446,"title":"PM Narendra Modi To Attend Keynote AI Talk By NVIDIA&#8217;s Jensen Huang In New Delhi","content":"Indian Prime Minister Narendra Modi is set to attend a keynote talk by Jensen Huang, president and co-founder of NVIDIA Corporation on Monday. Huang will be delivering a speech at the NITI Aayog’s fourth edition of NITI Lecture Series at Vigyan Bhawan, New Delhi. The theme for the year 2018 is AI for ALL: Leveraging Artificial Intelligence for Inclusive Growth. Union Ministers, policy makers, experts from different walks of life along with NITI Aayog Vice Chairman, Members, CEO and Senior officials will be a part of this gathering. PM @narendramodi to attend @NITIAayog 's 4th edition of #NITILecture Series in which key note address will be delivered by #JensenHuang, President and Co-Founder, #NVIDIA Corporation at Vigyan Bhawan in New Delhi today pic.twitter.com\/eciTDfw0ZK — All India Radio News (@airnewsalerts) October 21, 2018 The Union Budget-2018 had mandated NITI Aayog to come up with a national programme on employing artificial intelligence towards national development. NITI Aayog has, since, published a National Strategy for Artificial Intelligence. National Strategy lays down the vision of India for evolving a robust ecosystem for AI research and adoption. The Strategy is termed #AIForAll as it is focused on leveraging AI for inclusive growth in line with the Government policy of Sabka Saath, Sabka Vikas. NITI Aayog has signed Statements of Intent with industry leaders such as Google, Microsoft and IBM to leapfrog India into the global arena of future technologies and enable the use of artificial intelligence in key, high gain sectors of social development – healthcare, agriculture, smart mobility, education and infrastructure. In order to demonstrate the impact of AI, multiple pilot projects have been launched in the areas of precision agriculture, creating AI led healthcare solutions, and bridging India’s diversity of languages by creating a digital “Ek Bharat, Shrestha Bharat” platform using Natural Language Processing (NLP). NLP is a system of protocols which allow machines or computers to understand and interact with human speech and symbols. The Prime Minister in his address at the launch of the World Economic Forum (WEF) Centre for Fourth Industrial Revolution has underlined that with each wave of new technology, new opportunities arise.  The rise of Artificial Intelligence will help improve productivity and lead to equitable development. Aiming to put AI to use for all and across sectors, NITI has identified barriers that need to be addressed to achieve success in the use of AI. These include lack of expertise, the absence of enabling data ecosystem, high resource cost and low awareness, privacy and security issues and absence of collaborative approach to adoption and application of AI. Recognising the importance of research and its commercial adoption in the success of AI, NITI has also proposed setting up of Centre of Research Excellence (CORE) to focus on developing a better understanding of existing core research. Besides this, the strategy paper also recommends setting up of International Centres of Transformational AI with a mandate to develop and deploy application-based research in collaboration with private players.","excerpt":"Indian Prime Minister Narendra Modi is set to attend a keynote talk by Jensen Huang, president and co-founder of NVIDIA Corporation on Monday. Huang will be delivering a speech at the NITI Aayog’s fourth edition of NITI Lecture Series at Vigyan Bhawan, New Delhi. The theme for the year 2018 is AI for ALL: Leveraging Artificial […]","categories":["AI News"],"tags":["AI (Artificial Intelligence)","Jensen Huang","Narendra Modi","NITI Aayog","NVIDIA"],"author_name":"Prajakta Hebbar","publish_date":"2018-10-22T10:44:04","publication_year":"2018","word_count":493,"keywords":["Go","artificial intelligence","AI","AI (Artificial Intelligence)","Git","Jensen Huang","NLP","RAG","Aim","ViT","AI research","Narendra Modi","NVIDIA","R","NITI Aayog"],"extracted_tech_keywords":["AI","artificial intelligence","NLP","Aim","RAG","R","Go","Git","ViT","AI research"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/modi-attends-jensen-huang-talk-nvidia-niti-aayog\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10170453,"title":"bp&#8217;s #MayYouWin Event in Pune: Empowering Women in Technology","content":"In a fast-paced technology world where innovation thrives, opportunities have never been more pressing and exciting. Yet for many women, stepping into this space seems a challenge. Now, Global energy giant bp will host a women-only event titled #MayYouWIN at JW Marriott, Pune on Friday, May 30th. This unique event targets women who are exploring new paths or re-entering the workforce after a hiatus. It supports BP’s vision of empowering women by providing the respect, support, and opportunities necessary for success in the technology sector. The event will cover bp’s plans to expand its footprint in India, including in Pune, and its contribution to the nation’s energy landscape. Date: 30th May 2025Time: 10:30 AM onwardsRegister Now Participants will explore cutting-edge technologies driving bp’s business growth and discover how the company’s world-class technologies, both physical and digital, are propelling its business growth and enhancing competitiveness. Examples range from refining to advanced fuels formulation, and from oil and gas production to bioenergy breakthroughs, showcasing how technological advancements are shaping the future of energy. The event will also highlight bp’s strategic technology partnerships with industry giants such as Reliance-Jio, Microsoft, AWS, and ONGC, demonstrating how these collaborations are driving technological breakthroughs, fostering innovation, and enabling energy security. Attendees will have a unique opportunity to network with bp’s leaders and technologists, gaining insights into career pathways within bp’s physical and digital technology functions as the company establishes its presence in Pune, India. Date: 30th May 2025Time: 10:30 AM onwardsRegister Now bp’s Strong Presence in India bp, one of the largest energy firms globally, boasts a robust history of innovation and digital progress. In India, it accounts for 30% of the country’s natural gas output and collaborates with the government, Reliance, and ONGC to enhance energy security and sustainability. Furthermore, bp demonstrates a strong dedication to performance and innovation in India. This is evident through its subsidiary Castrol, which has been operational in the country for over a century and remains a leader in the lubricants sector. bp also maintains collaborative relationships in energy development with Reliance and ONGC, the two foremost energy companies in India. Date: 30th May 2025Time: 10:30 AM onwardsRegister Now","excerpt":"Global energy giant bp will host a women-only event titled #MayYouWIN at JW Marriott, Pune on Friday, May 30th.","categories":["AI Highlights"],"tags":[],"author_name":"Supreeth Koundinya","publish_date":"2025-05-21T19:33:14","publication_year":"2025","word_count":359,"keywords":["Go","API","TPU","AWS","AI","cloud_platforms:AWS","innovation","programming_languages:R","Git","R"],"extracted_tech_keywords":["AI","AWS","TPU","R","Go","Git","API","innovation","cloud_platforms:AWS","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/bps-mayyouwin-event-in-pune-empowering-women-in-technology\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":38333,"title":"Data Science Hiring Scenario In India: Hear From The Hiring Managers","content":"Talk to any major analytics and data science recruiter, and the first challenge that they mention is hiring. While the data science recruitment is happening at a rapid pace, the industry needs are often not met, owing to several challenges from an outdated educational system that hasn’t kept pace with the industry’s demands to unsatisfactory exposure to required skills. While companies are giving good compensation to data scientists today, with a good increase in compensation over the last two years, the challenges still persist. To get an industry perspective on this, Analytics India Magazine spoke to hiring managers and gleaned insight on hiring trends, skills required, educational qualifications that companies are looking for, measures that companies can take to improve and more. With this detailed story, we aim to chalk out a thorough idea of the hiring scenario in Indian companies and how are companies looking to improve the hiring process. We discussed six core areas that are often the talk points as soon as we get into the road of analytics hiring. Talent shortage in the analytics industry: This is the most talked about challenges that analytics recruiters face. There are many reports suggesting an acute talent shortage of almost 1 lakh open analytics positions. There is a demand for experienced analytics professionals, but the ground reality is that the companies are struggling to get freshly trained graduates out of college. “This calls for an increased partnership to bridge the critical gap in demand vs supply between academia, corporate, government, along with supporting associations and regulators”, says Suman Reddy, MD at Pegasystems India. On the contrary, Ashish Sam from TheMathCompany believes that talent is out there and available in plenty, but the challenge is to find a candidate that is a good fit for the organisation. To get the right talent, some of the steps that they take are organising hackathons, get them to interact with different team members, give them situations and case studies to get an understanding of their problem-solving skills, amongst others. Addressing the talent gap by companies and educational institutions: While we have discussed that there is a talent shortage in the sector, how can we deal with this? Sam has a detailed roadmap for companies and institutions that needs to change for addressing the talent gap. He firmly believes that the education system has a dated curriculum and not in line with the latest developments. It is mostly theory-focused and lacks practical application in the real world. He stresses on institutions to have a constant connection with the industry to understand the advancements and technological changes taking place. For companies, he says that learning team has a big role to play. “Their job is not limited to training but also constant reinforcement through problem-solving and conducting regular follow up sessions,” he said. Ishan Bansal, Co-founder, Groww says that data analytics tools are easy and can be learned in a few months. It’s in thought and theory of what analysis instead of how to do it. The gap can be overcome with giving the student’s case studies without data, then explore with statistical inferences, common sense, and imagination and then see what they can or cannot tackle with data. The need for reskilling: As the industry keeps evolving in the various technological advancements, the need for reskilling and keeping updated with eh latest trends becomes imperative. This is also important for the candidate to be most updated with the latest happenings in recruitment sector.  “Re-skilling is the key to our consistent growth and we have a dedicated learning and development  cell-Co.ach, to ensure our employees are up-to-date with all the latest industry trends and technologies,” says Sam. with this program they are facilitating a program that makes learning more relevant and effective. Reddy shares that they are also addressing this challenge with their initiative, Pega Academy, which allows employees to learn a course at no cost. “Pegasystems also offers employees to build data and analytic capabilities if they aspire to become data scientists or data analysts,” he shares. Hiring data generalists vs specialists: There is often a debate on whether to take data science specialist or generalist as there may bring different capabilities to the table. As Sam puts it, there are challenges associated with having only experts or specialists in a team as expertise comes with a limited shelf life. Having a generalist on board helps in the cross-pollination of ideas not just across technologies, but across industries. Reddy echoes the same point as they predominantly look for hiring data generalists, especially those who have vast experience across a number of challenges through their work. He also stresses on the importance of building the talent in-house, sourcing niche talent through acquisitions or collaborating with partners. “Recently, with the acquisition of Infruid Labs, we have added stronger analytics expertise to our workforce that will allow us to build more robust data analytics capabilities for our clients,” he said. Automating HR processes: One of the other challenges in hiring for any industry, let alone analytics, is the need to screen through resumes. The process of screening resumes is a routine and rule-based tasks which many companies are delegating to machines by automating the system. “Streamlining the resume screening process could involve presenting a coding test for a certain cohort to evaluate their capabilities and filter the right candidates to be taken to the next stage. Using recognized hacking platforms helps us manage how the right resumes are efficiently funnelled, helping us identify the right talent that matches the job profiles,” says Reddy. On the other hand, Sam has a different take. “AI can now be trained to review and screen profiles to suit our requirements. Having said that, I believe there is a lot of value add that the traditional approach of screening does,” he says. Calling and talking to the candidate helps in understanding the candidate such as communication skills, ability to articulate their work\/project, understanding of technology, and more. “While the idea of automation of the screening process is to able organisations to manage scale, some of these aspects may not be easily done on the tool,” he says. Dealing with the attrition rate: Sam agrees that it is a constant challenge for organisations. A lot of effort needs to go into tackling this problem. “At our company, we take steps such as our engagement with senior employees is sustained, we work together on organisation initiatives, involving them in key decision making, regular feedback\/discussions and so on. The key to keeping senior level employees happy and engaged in any organisation is to understand their aspirations and inform them of their contribution or role in building the company. Their identity is critical to any organisation. They also need to be able to see the big picture,” he says on a concluding note.","excerpt":"Talk to any major analytics and data science recruiter, and the first challenge that they mention is hiring. While the data science recruitment is happening at a rapid pace, the industry needs are often not met, owing to several challenges from an outdated educational system that hasn’t kept pace with the industry’s demands to unsatisfactory […]","categories":["AI Hirings"],"tags":["analytics hiring india","data analytics acquisitions","data science hiring india","data scientist qualifications","Mergers and Acquisitions","qualifications for data scientist"],"author_name":"Srishti Deoras","publish_date":"2019-04-26T12:08:27","publication_year":"2019","word_count":1137,"keywords":["analytics hiring india","data science","Go","API","data scientist qualifications","AI","ML","data analytics acquisitions","qualifications for data scientist","GAN","Aim","automation","analytics","data science hiring india","Mergers and Acquisitions","R"],"extracted_tech_keywords":["AI","ML","data science","analytics","Aim","R","Go","API","GAN","automation"],"url":"https:\/\/analyticsindiamag.com\/ai-hiring\/data-science-hiring-scenario-in-india-hear-from-the-hiring-managers\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10147775,"title":"This Indian Startup is Making ‘Black-Box’ AI Models Spill the Beans","content":"As models grow in size, one crucial element remains elusive—explainable AI. The larger they grow, the harder it becomes to understand their inner workings. And when we consider models 405b and beyond, it becomes extremely difficult to understand how they arrive at specific outcomes. Vinay Kumar Sankarapu, CEO of Arya.AI, perfectly encapsulates the challenge: “Capabilities are fine—you can say your model is 99.99% accurate, but I want to know why it is 0.01% inaccurate. Without transparency, it becomes a black box, and no one will trust it enough to put billions of dollars into it.” His statement cuts to the core of the black box dilemma: trust demands understanding. The Problem is Beyond Hallucinations A recent study by the University of Washington revealed significant racial, gender, and disability biases in how LLMs ranked job applicants’ names. The research found that these models favoured names associated with White individuals 85% of the time, and names perceived as Black male were never preferred over White male names. This study highlights the complex interplay of race and gender in AI systems and underscores the importance of considering intersectional identities when evaluating AI fairness. While we spoke to Mukundha Madhavan, tech lead at DataStax, about model hallucination and the size of the model, he said, “Foundation models, their training, and architecture—it feels like they are external entities, almost victims of their own complexity. We are still scratching the surface when it comes to understanding how these models work, and this applies to both small and large language models.” He added that size doesn’t matter. Whether a model boasts 40 billion or 4 billion parameters, these are just numbers. The real challenge is to make sense of these numbers and understand what they represent. Sankarapu pointed out a paradox in AI development: “We are creating more complicated models that are harder to understand while saying alignment is required for them to become mainstream.” He noted that while AI systems have scaled through brute force—using more data and layers—this approach has plateaued. Now, efficiency and explainability need to scale alongside model complexity. Arya.AI has been working on solving this issue through its proprietary ‘Backtrace’ technique, which provides accurate explanations for deep learning models across various data types. Sankarapu explains, “We want to create explainability that is very accurate and can work for any kind of model so that LLMs are no longer black boxes but white boxes.” This effort aligns with emerging regulations like the EU’s AI Act, which mandates explainability for high-risk AI applications. Sankarapu adds, “Once you start understanding these models, you can do tons of things around them—like improving efficiency or unlocking new research areas.” This positions explainability as a catalyst for both innovation and operational efficiency. In response to these challenges, Arya.AI has developed AryaXAI, which provides precise, granular explanations for AI decisions across various model architectures, making it particularly valuable for enterprise applications. AryaXAI stands out by offering feature-level explanations that help users understand exactly which inputs influenced a model’s decision and to what extent. The platform can analyse both structured and unstructured data, providing explanations for decisions made by complex neural networks, including those processing images, text, and tabular data. A key differentiator is AryaXAI’s ability to provide explanations in real time, making it practical for production environments where quick decision-making is crucial. The platform generates natural language explanations that are easily understood by both technical and non-technical stakeholders, bridging the gap between AI capabilities and business requirements. For financial institutions, specifically, AryaXAI offers detailed audit trails and compliance documentation, addressing regulatory requirements while maintaining model performance. The platform’s ability to explain decisions in human-readable terms has made it particularly valuable in sectors where transparency is non-negotiable, such as banking, insurance, and healthcare. What Should be the Vision for Autonomous Systems On the future of autonomous systems in banking, Sankarapu shared Arya.AI’s focus on aligning models with end goals through feedback loops and explainability. “To build truly autonomous agents capable of handling complex tasks like banking transactions, they must be explainable and aligned with user expectations,” he said. Madhvan proposed a three-pronged approach to reduce hallucinations. The first part is model explainability research which focuses on uncovering how AI models make decisions by analysing their embeddings, attention mechanisms, and other internal processes, which are often opaque. This research is essential for building trust and transparency in AI. The second is model alignment, which ensures that AI models behave as intended by aligning their outputs with human values and reducing issues like hallucinations or unintended biases. Finally, practical implementation prioritises creating reliable systems for real-world applications by incorporating safeguards and guardrails that allow models to operate effectively within specific business contexts, even if complete transparency is unattainable. Together, these approaches aim to balance the growing complexity of AI systems with operational reliability and ethical considerations. Arya.AI is also exploring advanced techniques like contextual deep Q-learning to enable agents to handle tasks requiring memory and planning. However, Sankarapu cautions against over-focusing on futuristic visions at the expense of current market needs. “Sometimes you get caught up with too much future vision and lose sight of current realities,” he concluded.","excerpt":"This effort aligns with emerging regulations like the EU’s AI Act, which mandates explainability for high-risk AI applications.","categories":["IT Services"],"tags":["black box ai"],"author_name":"Sagar Sharma","publish_date":"2024-12-26T09:00:00","publication_year":"2024","word_count":857,"keywords":["Go","TPU","autonomous agents","AI","neural network","R","Aim","deep learning","black box ai","foundation models","xAI"],"extracted_tech_keywords":["AI","deep learning","neural network","foundation models","xAI","Aim","autonomous agents","TPU","R","Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/this-indian-startup-is-making-black-box-ai-models-spill-the-beans\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":44567,"title":"Amazon Messaging Assistant Now Communicate In Hindi","content":"Amazon India this week announced the launch of their Automated Messaging Assistant in Hindi. With the launch of the automated Hindi assistant, millions of customers who prefer the popular regional language can now get their queries resolved without even connecting with an associate. The Automated Assistant is a chatbot powered by Amazon’s customer service-tuned artificial intelligence platform, leveraging Machine Learning and Natural Language Processing capabilities, enabling a seamless and friendly conversational experience for customers. The conversation happens right inside the messaging window where the customer can naturally interact just as they would with a human and the bot seamlessly transitions to a human assistant within the same window, whenever necessary. With this launch, Amazon India aims to break the language barrier and provide a hassle-free support service to customers. The company is now enabling millions of Hindi speaking customers in the country who prefer to interact with Amazon’s automated messaging assistant in the language of their choice. With the launch of the automated assistant in Hindi, Amazon is taking the next step towards enhancing its customer service experience for Indian customers. “When customers interact with the Automated Assistant, it first predicts the most likely issue they are trying to contact us for, post which customers can ask questions related to their issue and receive faster resolutions. By offering the Automated Assistant in Hindi, we are taking a big step towards localising our post-order experience for Indian customers,” said Akshay Prabhu, Director, Customer Service, Amazon India in a statement. “The new chat assistant in Hindi underlines our commitment to serving the next 100 million Indians to discover and shop with Amazon. We also see opportunities for scaling the messaging assistant as a contact channel —to increasingly resolving issues before a customer service associate is even involved.” At launch, the Hindi chat experience is available on the Amazon’s Android app.","excerpt":"Amazon India this week announced the launch of their Automated Messaging Assistant in Hindi. With the launch of the automated Hindi assistant, millions of customers who prefer the popular regional language can now get their queries resolved without even connecting with an associate. The Automated Assistant is a chatbot powered by Amazon’s customer service-tuned artificial […]","categories":["AI News"],"tags":["Amazon","hindi"],"author_name":"Prajakta Hebbar","publish_date":"2019-08-16T15:01:48","publication_year":"2019","word_count":308,"keywords":["artificial intelligence","machine learning","programming_languages:R","AI","ML","Amazon","RAG","Aim","hindi","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","Aim","RAG","R","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/amazon-messaging-assistant-hindi\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":26063,"title":"RPA Company Automation Anywhere Raises $250 Million In Series A Funding","content":"Automation Anywhere, a noted company in the robotic process automation (RPA) field announced that they had completed their Series A financing round of $250 million led by New Enterprise Associates, Goldman Sachs Growth Equity; with participation from General Atlantic and World Innovation Lab. This investment has brought the company’s post-money valuation to $1.8 billion. A statement published by Automation Anywhere claims that this is one of the largest Series A rounds on record for an enterprise software company. This sizable investment will help extend the company’s leadership in the rapidly-expanding RPA market and accelerate its global customer engagements and product development. Mihir Shukla, CEO and co-founder at Automation Anywhere said in a statement, “Customers tell us that traditional process automation technologies are capable of automating only about 20% of an enterprise’s business processes. We believe our Intelligent Digital Workforce Platform can automate up to 80% of these processes. It’s a stark contrast and an enormous opportunity. The financing and our high calibre of investors strongly position us to push boundaries and drive the next major business disruption.” I’m proud to announce Automation Anywhere has raised $250 million in Series A funding, one of the largest A rounds in history. With this investment we are pushing boundaries and driving the next major business disruption. https:\/\/t.co\/mqzNuYRfnE pic.twitter.com\/oUBA3jmpJK — Mihir Shukla (@MihirAndNow) July 2, 2018 The statement also added that John Giannuzzi and Chetan Puttagunta, general partners at New Enterprise Associates, are set to join the board of directors at Automation Anywhere. Clyde Hosein, CFO at Automation Anywhere added, “With this investment, we are poised to extend our leadership in the multibillion-dollar RPA market. We welcome the new investor group and look forward to their valuable contributions as we enter our next growth phase.” RPA is a fast-evolving technology which uses software robots to automate business processes that have never been automated by traditional technology platforms. With this investment, Automation Anywhere plans to deepen their customer engagement and deploy its technology in additional geographies. Building on its core product offerings, the company will further work on specialised machine learning capabilities and sophisticated artificial intelligence integrations to drive higher operational efficiency, increased agility and flexibility to scale up-and-down anytime.","excerpt":"Automation Anywhere, a noted company in the robotic process automation (RPA) field announced that they had completed their Series A financing round of $250 million led by New Enterprise Associates, Goldman Sachs Growth Equity; with participation from General Atlantic and World Innovation Lab. This investment has brought the company’s post-money valuation to $1.8 billion. A […]","categories":["AI News"],"tags":["AI (Artificial Intelligence)","Automation Anywhere","Robotic Process Automation","RPA"],"author_name":"Prajakta Hebbar","publish_date":"2018-07-04T06:45:51","publication_year":"2018","word_count":365,"keywords":["Go","API","artificial intelligence","machine learning","AI","RPA","Git","automation","Automation Anywhere","Robotic Process Automation","Aim","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","Aim","R","Go","Git","API","automation","RPA"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/rpa-company-automation-anywhere-raises-250-million-in-series-a-funding\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":25017,"title":"Why Did Microsoft Acquire Semantic Machines, A Conversational AI Startup?","content":"Voice-based intelligent platforms have become quite popular over the years and tech giants are adopting them as a way to offer consumer-centric services. Whether it is Apple’s Siri, Google’s Assistant or Microsoft’s Cortana, the list is growing at a significant pace. While these platforms use artificial intelligence to a great extent, a perfect conversational AI is yet to be achieved. These trending bots and intelligent assistants only follow simple commands and queries based on their training. As Microsoft’s blog notes, for rich and effective communication, intelligent assistants need to be able to have a natural dialogue, which is lacking significantly. To accomplish a truly intelligent bot, Microsoft recently acquired conversational AI startup, Semantic Machines. Semantic Machines: Background This California-based company has been known to aggressively develop fundamental technology which allows humans to interact naturally with computers. Led by tech entrepreneur Dan Roth, UC-Berkeley professor Dan Klein and Stanford University professor Percy Liang, this recent acquisition would give Microsoft access to its formidable talent. A lot of current staffers from Semantic Machines have been ex-employees at Nuance, the voice recognition company that once powered Siri — which says a lot about their experience in the field. Founded in 2014 by Dan Roth, his earlier venture called Voice Signal Technologies was acquired by Nuance in 2007. This strong background explains why Microsoft chose Semantic to strengthen its voice-based AI portfolio. What Does Microsoft Want? With a strong background in conversational AI, Microsoft plans to tap into Semantic Machines’ Conversation Engine, which extracts semantic intent from natural input such as voice or text. It then generates a self-updating learning framework that manages dialogue context, state, salience and the goals of the end users. Also, the Natural Language Generation technology of the startup formulates communication with the user, based on a dialogue context. It relies on machine learning to enable users to discover, access and interact with the information and services in a much more natural way. The idea is to create a conversational system which is contextual, unlike most of the dialogue technologies today, which are mostly orthogonal. Semantic Machines claim that their AI can produce conversations which not only answer queries but also predict questions more effectively. They also claim that their AI will have a natural flow in the conversation which has not yet been fully achieved by the likes of Siri, Google Assistant, Alexa or Microsoft’s own Cortana. Natural Dialogue In Conversational AI Semantic Machines’ solution, which is mostly focused on enterprise customers, would not only bolster Cortana but other social chatbots like Xiaolce, which has recorded up to 30 billion conversations across the Asian countries till now. It will improve capabilities in Microsoft Cognitive Services and Azure Bot Service. “XiaoIce has had more than 30 billion conversations, averaging up to 30 minutes each, with 200 million users. With XiaoIce and Cortana, we’ve made breakthroughs in speech recognition and more recently become the first to add full-duplex voice sense to a conversational AI system, allowing people to carry on a conversation naturally,” said David Ku, CVP and chief technology officer of Microsoft AI and Research. The company also plans to establish a Conversational AI Center of Excellence in Berkeley to further boost the developments in this area. “By combining Semantic Machines’ technology with Microsoft’s own AI advances, we aim to deliver powerful, natural and more productive user experiences that will take conversational computing to a new level. We’re excited to bring the Semantic Machines team and their technology to Microsoft,” he added. Other Initiatives By Microsoft In AI Microsoft has been driving research and breakthroughs in the building blocks of conversational AI over the years. With its continued efforts in speech recognition and natural language understanding, they are inching towards achieving a phase where chatbots could talk and understand like humans. In 2016, the company took a step in conversational computing by introducing a framework for developing bots and pre-built Cognitive Services for infusing speech recognition and natural language understanding into intelligent assistants. Earlier this year in April, Microsoft created a technological breakthrough that would allow people converse with AI-powered chatbot in a more natural experience, similar to a person talking on the phone to a friend. Currently integrated into XiaoIce, the company aims to apply it to its other chatbots such as Zo. Referring to it as full duplex, meaning the ability to communicate in both directions simultaneously, like a telephone call, users would not have to use a “wake word”, which is the case with most of the voice assistants currently available in the market. Di Li, Microsoft’s general manager for XiaoIce, had noted in a company blog post that these improvements are part of Microsoft’s effort to build AI-powered social chatbots that understands people’s emotional as well as intellectual needs. The acquisition of Semantic Machines is clearly aimed at strengthening these efforts by the company. With this acquisition, Microsoft is also aiming to integrate a sense of humour, ability to chit-chat, play games, remember personal details, and others. Full duplex already allows these advances, which are very relaxed and natural. On A Concluding Note Apart from Microsoft, Google too, has been extensively working on life-like versions of its spoken AI. Earlier in May, Sundar Pichai, introduced “Duplex” at the Google I\/O where he demonstrated how AI system could perform things like booking appointment, sounding every bit like a human. While it is an interesting idea to have a naturally conversant AI, we wait to see when it hits the market.","excerpt":"Voice-based intelligent platforms have become quite popular over the years and tech giants are adopting them as a way to offer consumer-centric services. Whether it is Apple’s Siri, Google’s Assistant or Microsoft’s Cortana, the list is growing at a significant pace. While these platforms use artificial intelligence to a great extent, a perfect conversational AI […]","categories":["Global Tech"],"tags":["AI (Artificial Intelligence)","Conversational AI","Microsoft","recent technological advances"],"author_name":"Srishti Deoras","publish_date":"2018-05-31T05:15:28","publication_year":"2018","word_count":913,"keywords":["Go","artificial intelligence","machine learning","AI","chatbots","R","RAG","Aim","recent technological advances","Conversational AI","AI (Artificial Intelligence)","Azure","Microsoft","startup"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","Aim","RAG","chatbots","Azure","R","Go","startup"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/why-did-microsoft-acquire-semantic-machines-a-conversational-ai-startup\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10161031,"title":"Free Speech, Really?","content":"In a move that could reshape the way online speech is managed, Meta recently announced integrating large language models (LLMs) into its content moderation strategy. “We’ve started using AI LLMs to provide a second opinion on some content before we take enforcement actions,” read a blog post by the company. For a social media giant that is always under public scrutiny, this is a bold attempt to moderate the issues of overreach and inconsistency. In a recent video, Meta chief Mark Zuckerberg unveiled sweeping changes to the platform’s strategy. One of the most striking ones was the significant reduction in the number of human fact-checkers across all Meta platforms. As part of the downsizing efforts, human fact-checkers are being shifted from their California base to Texas. Zuckerberg has long championed free expression as a cornerstone of progress, and according to the company, this move is in tandem with that. In his 2019 Georgetown address, he argued that that empowering people to voice their ideas not only drives innovation but also challenges existing power structures. Yet, even as he spoke of the virtues of free speech, he warned that too much moderation could tilt the scales of power, stifling diverse voices and diminishing the democratic discourse. As the industry embraces AI, with AGI on the verge, Meta’s solution of putting AI and collective intelligence at the heart of moderation decisions is futuristic. By using LLMs as a second opinion before enforcement actions, the company claims to refine its approach, reduce wrongful takedowns, and temper the frustrations users often feel when their content is censored. AI LLMs as a Second Opinion At the heart of this strategy is Meta’s deployment of AI-driven LLMs to review flagged content. These models, capable of sifting through massive data troves in seconds, are designed to identify subtle nuances and policy violations that human moderators might overlook. For users, this means fewer errors and more fairness—at least, that’s the goal. It’s a vision of moderation that, in addition to minimising the number of cases of wrongful takedowns, improves the accuracy of enforcement decisions. At its core, Meta’s use of AI for second opinions is an experiment in trust—trust in technology, in the community, and the overall commitment to self-regulation. But the question remains: Can AI deliver on this promise without falling prey to its own flaws? The Bias Problem Despite their impressive capabilities, LLMs have their limitations. Research has shown that these models often reflect the biases of their creators. Major AI systems—whether from OpenAI, Anthropic, Google, or others—have all been called out for exhibiting ideological leanings. Andrej Karpathy, a former OpenAI researcher, explained the issue: “LLMs model token patterns without genuine understanding,” making them prone to echoing the biases embedded in their training data. For a system tasked with ensuring fairness, this is a troubling flaw. xAI CEO Elon Musk has also voiced his concerns, warning that LLMs could exhibit ideological biases, including what he termed a “far-left” leaning. Meanwhile, computer scientist Grady Booch has criticised these models as “unreliable narrators,” capable of producing outputs that are not only deceptive but potentially toxic. Baybars Orsek, VP of fact-checking at Logically, calls for the industry to unite around sustainable, technology-driven solutions that prioritise accuracy, accountability, and measurable impact. While he sees merit in efforts like community notes, he argues in favour of a professionalised approach to fact-checking. “A professional fact-checking model—enhanced by AI and rigorous methodologies—remains the most effective solution for addressing these issues at scale,” Orsek said. These highlight a significant challenge Meta faces: if the tools intended to enhance fairness are fundamentally flawed, can they genuinely be relied upon to moderate speech impartially and equitably? Meta Still Bets Big on Co-Intelligence Beyond AI, Meta is taking a leaf out of X’s playbook by adopting a decentralised moderation approach with its Community Notes model. This model, first introduced by former Twitter CEO Jack Dorsey in 2021, relies on users to add context to flagged content, providing a more collaborative approach to moderation. Under Elon Musk’s ownership, Community Notes became a key feature, earning praise for its ability to scale moderation while capturing a diversity of perspectives. Musk himself has commended Zuckerberg for bringing the model to Meta’s platforms. To its credit, the strength of Community Notes lies in its decentralised nature. By leveraging millions of users, it incorporates a wide range of viewpoints, making moderation more representative and less prone to the pitfalls of top-down decision-making.","excerpt":"Free expression has never been absolute, Meta CEO Mark Zuckerberg said in his 2019 Georgetown address.","categories":["Global Tech"],"tags":["AI adoption"],"author_name":"Aditi Suresh","publish_date":"2025-01-09T14:37:12","publication_year":"2025","word_count":740,"keywords":["Anthropic","Go","TPU","AI adoption","OpenAI","AI","AWS","R","RAG","Aim","xAI"],"extracted_tech_keywords":["AI","OpenAI","Anthropic","xAI","Aim","RAG","AWS","TPU","R","Go"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/free-speech-really\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":69754,"title":"Workplaces are poised to get smarter with Internet of Things","content":"Simply put, IoT is the approach of connecting any device to the Internet or to each other. This encompasses everything from consumer electronics, automobiles to industrial machinery and practically anything else that one can imagine.  This also administers to components within the machines, for example a jet engine of an airplane or the drill of an oil rig. The IoT is a giant network of connected “things” (which also includes people). A lot of IoT related examples can be seen both in workplace and outside of it. Employee ID cards used in relatively all the modern workplaces now with badges is the most elementary of IoT use case at workplace. Often, they allow an employee to enter office place, parking garages etc. But regularly, the employee has to take some action, like wave them to the nearest sensor\/tracking machine. IoT-enabled employee badges and parking cards will let people come and go in the same way cars can now drive through empty tollbooths while barely slowing down. The buildings (things) will “know” who they are, because other things (the ID cards) will have quietly communicated that fact. Similarly, in manufacturing and warehousing, IoT will drive towards maximum competencies as companies rely more and more on robots, autonomous vehicles, and drones to move parts and goods in and out of storage. BLE sensors and makers in shelves and floors will not only help devices route to their destinations, but also report back to cloud-based software that monitors and coordinates movements in real-time. IoT has opened up a world of possibilities for businesses and employees in virtually every industry. For the first time, we can see what circumstances in real time in the physical world, and data that used to be is now flowing into one place. Our traditional office is dynamically changing and so are those working in them. Within the next decade, we will have the need to embrace a hyper-connected workforce made up of digital natives. These workers will be early adopters of connected consumer devices and will adapt quickly to the convenience the Internet of Things can offer. With Millennials predicted to make up 75% of the overall workforce by 2020, it’s important for businesses to take these steps to meet their needs. Businesses constantly look for more efficiency and savings, and as more and more of our tools, technology, and utilities become connected, companies will reap the benefits on their balance sheets. This disruption will come from new value creation made feasible by massive volumes of data from connected products, and the increased capability to make automated decisions and take actions in real time. This will tremendously improve the operational efficiency (e.g. improved uptime, asset utilization) through predictive maintenance and remote management. With IoT one has access to better data, IoT offers a diverse way to collect better data that formerly might have been tallied up manually. Smarter data about peak times in a restaurant or retail establishment could help the owners make better decisions about employee scheduling to maximize productivity. Similarly companies from other sectors might use the data collected from employees to pinpoint their most productive times, and schedule meetings appropriately. The time for connected offices is already here. For example IoT takes remote device management to a whole new level. Machines in factories now have sensors that notify workers as soon as there’s a problem. With millennials who are roughly 66 percent of the workforce; are willing to try wearable devices which help them perform better – the “Internet of Things” won’t just empower things, it will also entrust people to better manage aspects of their work lives. The “Internet of Things” is starting to make its presence felt in the workplace. Not only will it completely transform how we work, it will also change the way we interact with our surroundings. While it’s true that many companies already have the basic building blocks to roll out connectivity and the things to connect, where it can start getting complicated is bringing all the elements together, maybe because IoT is still an emerging trend and there is lack of familiarity among the users. Also, to some extend the benefits are not always captured and aligned with the business objectives. At the moment many of our connected devices can talk, but not the same languages. There’s a danger of creating chaos with a smart system if all the aspects within it aren’t singing the same song. This creates barrier in unlocking the real value of IoT. IoT in India IoT adoption in India comes with its own challenges – Security and Privacy issues are paramount as everything becomes more networked. Lack of standards and protocols, reliability and stability issues are common. In addition, upgrading of telecom & cable infrastructure is required in India. Also, we need to upskill our workforce as there is lack of adequate skill-sets to expedite the march towards fourth industrial revolution. According to analysts IoT market in India is projected to grow at a CAGR over 28% during 2015 to 2020. IoT is being rapidly brought into use across diverse industry verticals to reduce operational and manpower costs, and increase operational efficiency. However, the technical complexities, cultural integration and business models will evolve in parallel to the digital business vision and IoT roadmap evolution. However, India’s IoT market is highly fragmented with numerous players operating across the value chain and there exist some local challenges because of which the growth may not at par with the global level. For example getting device connected in Indian infrastructure is not easy. Extreme temperature, high levels of humidity and dust, lack of clean and continuous power supply, inadequate telecom coverage. Need for standardization of all devices across an IoT ecosystem coupled with favorable Government policy environment are set to fuel growth in the country’s IoT market over the next five years. Planned government projects such as smart cities, smart grids, smart transportation etc. are expected to be major revenue generating sources for the IoT solution providers in the years to come.w","excerpt":"The Internet of things (IoT) is becoming an increasingly growing topic of conversation. It is a conception that not only has the potential to impact how we live but also how we work.","categories":["IT Services"],"tags":[],"author_name":"AIM Media House","publish_date":"2016-12-08T04:30:09","publication_year":"2016","word_count":1005,"keywords":["Go","API","ELT","AI","ETL","Git","RAG","ViT","Rust","R"],"extracted_tech_keywords":["AI","RAG","R","Go","Rust","Git","API","ETL","ELT","ViT"],"url":"https:\/\/analyticsindiamag.com\/it-services\/workplaces-posed-get-smarter-internet-things\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10097016,"title":"Meta AI Releases A Multimodal Model &#8220;CM3leon&#8221;  — But Won’t Release It","content":"Meta AI, the AI arm of Mark Zuckerberg owned Meta, has introduced a multimodal model  “CM3leon” (pronounced like chameleon), that does both text-to-image and image-to-text generation. Similar to its previous models, the company has decided not to release the model code which has enraged AI enthusiasts. The AI model bridges the gap between text and images. With its capabilities in text-guided image generation and editing. The model  has the potential to revolutionise the approach in which current users interact with and manipulate visual content. According to Meta, its model boasts improved capabilities in producing coherent imagery that closely aligns with input prompts. What sets CM3leon apart is its efficiency, as it requires only five times the computing power and a smaller training dataset compared to previous transformer-based methods. Meta touts CM3leon’s prowess in various vision-language tasks, including visual question answering and long-form captioning. This novel approach by Meta marks a departure from the diffusion method commonly employed in image generation. Instead, the company’s researchers opted for the Transformer architecture, a neural network design widely recognized for its successful implementation in large language models like OpenAI’s GPT-4. While CM3leon is not the first transformer-based image generator—StyleSwin precedes it—Meta asserts that it surpasses other contenders in terms of efficiency. Though the model is being lauded for being state-of-the-art the fact that it is closed source makes the research community irk. Similarly, last month the tech giant unveiled Voicebox but did not make the model public over fears of potential misuse. “While we believe it is important to be open with the AI community and to share our research to advance the state of the art in AI, it’s also necessary to strike the right balance between openness with responsibility,” Meta had stated while releasing the paper. Again, Meta is not the only tech goliath gatekeeping its research. Even though companies like Google, Microsoft, Facebook, Amazon, and Apple have contributed tons to research over the years — they also have kept a major chunk of it to themselves. The reason this displeases the research community is because these companies have been dependent on the open source community yet refuse to contribute as much as they gain. Read more: Big Techs Flip-Flop on Open Source","excerpt":"Meta AI’s model has the potential to revolutionise the approach in which current users interact with and manipulate visual content.","categories":["AI News"],"tags":["AI Tool","Meta","Meta AI"],"author_name":"Tasmia Ansari","publish_date":"2023-07-17T12:42:05","publication_year":"2023","word_count":371,"keywords":["Go","Meta AI","Meta","OpenAI","AI","neural network","RPA","RAG","GPT","transformer architecture","AI Tool","R"],"extracted_tech_keywords":["AI","neural network","OpenAI","Meta AI","RAG","R","Go","transformer architecture","GPT","RPA"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/meta-ai-releases-a-multimodal-model-cm3leon-but-wont-release-it\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10034571,"title":"How To Run Python Code Concurrently Using Multithreading","content":"Multithreading enables CPUs to run different parts(threads) of a process concurrently. But what does that mean? Processes can be divided into different parts; let’s take the example of an online multiplayer game. One thread of the game could be responsible for communicating with the servers and rendering the graphics. The communication thread requires minimal computation and would involve some wait time, on the other hand, the render thread is computationally intensive with minimal wait time. Multithreading enables the CPU to run the render thread while the communication thread is waiting for a response from the server, increasing the CPU utilisation. Note that multithreading is not to be confused with multi-processing. Modern CPUs have multiple cores; multi-processing utilizes these cores to run processes in parallel. Multithreading, however, aims to maximize the utilization of each of these cores by running multiple threads concurrently. Multithreading is useful when the task has IO or network operations that involve waiting; multiprocessing makes computation-intensive tasks of a process faster. Continuing the online game example, the render thread of most games are run in parallel on a GPU with thousands of cores, each thread rendering different aspects of the game. While the communication and IO threads are run concurrently on the CPU. Multithreading in Python The threading module comes with the standard Python library, so there’s no need for installing anything. By default, your Python programs have a single thread, called the main thread. You can create threads by passing a function to the Thread() constructor or by inheriting the Thread class and overriding the run() method. Using the Thread() Constructor import threading import time def useless_function(seconds): print(f'Waiting for {seconds} second(s)', end = \"\\n\") time.sleep(seconds) print(f'Done Waiting {seconds}  second(s)') start = time.perf_counter() t = threading.Thread(target=useless_function, args=[1]) t.start() print(f'Active Threads: {threading.active_count()}') t.join() end = time.perf_counter() print(f'Finished in {round(end-start, 2)} second(s)') -----------------------------Output----------------------------- Waiting for 1 second(s) Active Threads: 2 Done Waiting 1  second(s) Finished in 1.0 second(s) The two active threads are the main thread and the useless_function thread that you just created. The join() method blocks the execution flow until the thread t terminates. If you were to remove the join() call, the main thread would finish before t, and the output would be something like. -----------------------------Output----------------------------- Waiting for 1 second(s) Active Threads: 2 Finished in 0.0 second(s) Although the active_count() method is called after starting the thread t, it finishes execution before it. This happens because the processor runs the main thread while the thread t is sleeping. If you were to add a delay of 1 second before it, it would be executed after the thread t terminates. t = threading.Thread(target=useless_function, args=[1]) start = time.perf_counter() t.start() time.sleep(1) print(f'Active Threads: {threading.active_count()}') end = time.perf_counter() print(f'Finished in {round(end-start, 2)} second(s)') -----------------------------Output----------------------------- Waiting for 1 second(s) Done Waiting 1  second(s) Active Threads: 2 Finished in 1.0 second(s) Creating Your Thread Class The Thread subclass should only override the run() method and the __init__() constructor. And if the constructor is overridden the base class constructor, Thread.__init__(self), should be invoked before doing anything else. from thread import Thread def countdown(name, delay, count): while count: time.sleep(delay) print (f'{name, time.ctime(time.time()), count}') count -= 1 class newThread(Thread): def __init__(self, name, count): threading.Thread.__init__(self) self.name = name self.count = count def run(self): print(\"Starting: \" + self.name + \"\\n\") countdown(self.name, 1,self.count) print(\"Exiting: \" + self.name + \"\\n\") t = newThread(\"Thread 1\", 5) t.start() t.join() print(\"Exiting Main Thread\") -----------------------------Output----------------------------- Starting: Thread 1 ('Thread 1', 'Thu Apr 22 06:29:57 2021', 5) ('Thread 1', 'Thu Apr 22 06:29:58 2021', 4) ('Thread 1', 'Thu Apr 22 06:29:59 2021', 3) ('Thread 1', 'Thu Apr 22 06:30:00 2021', 2) ('Thread 1', 'Thu Apr 22 06:30:01 2021', 1) Exiting: Thread 1 Exiting Main Thread Now you know how to run code concurrently using multithreading in Python, but why would you want to do so? Let’s illustrate the concurrency aspect of multithreading and the increased CPU utilization with the help of an example. Unoptimized Code import requests import time urls = [ 'https:\/\/images.pexels.com\/photos\/305821\/pexels-photo-305821.jpeg', 'https:\/\/images.pexels.com\/photos\/509922\/pexels-photo-509922.jpeg', 'https:\/\/images.pexels.com\/photos\/325812\/pexels-photo-325812.jpeg', 'https:\/\/images.pexels.com\/photos\/1252814\/pexels-photo-1252814.jpeg', 'https:\/\/images.pexels.com\/photos\/1420709\/pexels-photo-1420709.jpeg', 'https:\/\/images.pexels.com\/photos\/963486\/pexels-photo-963486.jpeg', 'https:\/\/images.pexels.com\/photos\/1557183\/pexels-photo-1557183.jpeg', 'https:\/\/images.pexels.com\/photos\/3023211\/pexels-photo-3023211.jpeg', 'https:\/\/images.pexels.com\/photos\/1031641\/pexels-photo-1031641.jpeg', 'https:\/\/images.pexels.com\/photos\/439227\/pexels-photo-439227.jpeg', 'https:\/\/images.pexels.com\/photos\/696644\/pexels-photo-696644.jpeg', 'https:\/\/images.pexels.com\/photos\/911254\/pexels-photo-911254.jpeg', 'https:\/\/images.pexels.com\/photos\/1001990\/pexels-photo-1001990.jpeg', 'https:\/\/images.pexels.com\/photos\/3518623\/pexels-photo-3518623.jpeg', 'https:\/\/images.pexels.com\/photos\/916044\/pexels-photo-916044.jpeg' ] def download(url): img_data = requests.get(url).content img_name = url.split('\/')[4] img_name = f'{img_name}.jpg' with open(img_name, 'wb') as img_file: img_file.write(img_data) print(f'downloading {img_name}') t1 = time.perf_counter() for i in urls: download(i) t2 = time.perf_counter() print(f'Finished in {t2-t1} seconds') -----------------------------Output----------------------------- downloading 305821.jpg downloading 509922.jpg downloading 325812.jpg downloading 1252814.jpg downloading 1420709.jpg downloading 963486.jpg downloading 1557183.jpg downloading 3023211.jpg downloading 1031641.jpg downloading 439227.jpg downloading 696644.jpg downloading 911254.jpg downloading 1001990.jpg downloading 3518623.jpg downloading 916044.jpg Finished in 5.95 seconds Multithreaded Code start = time.perf_counter() threads = [] for i in urls: t = threading.Thread(target=download, args=[i]) t.start() threads.append(t) for thread in threads: thread.join() finish = time.perf_counter() print(f'Finished in {round(finish-start, 2)} seconds') -----------------------------Output----------------------------- downloading 509922.jpg downloading 963486.jpg downloading 305821.jpg downloading 3023211.jpg downloading 325812.jpg downloading 696644.jpg downloading 1557183.jpg downloading 1420709.jpg downloading 1252814.jpg downloading 1001990.jpg downloading 911254.jpg downloading 1031641.jpg downloading 916044.jpg downloading 3518623.jpg downloading 439227.jpg Finished in 1.41 seconds The two loops can be replaced with Executor() object from concurrent.futures: import concurrent.futures start = time.perf_counter() with concurrent.futures.ThreadPoolExecutor() as executor: executor.map(download, urls) finish = time.perf_counter() print(f'Finished in {round(finish-start, 2)} seconds') The Executor object creates a thread for each function call and blocks the main thread’s execution until each of these threads is terminated. In the unoptimized code, the GET requests happen sequentially, and the CPU is ideal between the requests. When each GET request happens in its separate thread, all of them are executed concurrently, and the CPU alternates between them instead of being ideal. References: To learn more about the Python threading module, refer to the following resources: DocumentationGitHub","excerpt":"Multithreading in Python enables CPUs to run different parts(threads) of a process concurrently to maximize CPU utilization.","categories":["AI Trends"],"tags":[],"author_name":"Aditya Singh","publish_date":"2021-04-23T13:00:00","publication_year":"2021","word_count":918,"keywords":["TPU","programming_languages:R","AI","Git","Python","Aim","programming_languages:Python","GitHub","R"],"extracted_tech_keywords":["AI","Aim","TPU","Python","R","Git","GitHub","programming_languages:Python","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/how-to-run-python-code-concurrently-using-multithreading\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":41560,"title":"Introductory Guide To PyTorch Using A Linear Regression Problem","content":"It’s often hard to make a decision on what framework to learn when there are many options to choose from. In the Machine Learning spectrum, we have multiple frameworks that compete against each other claiming to have better speed, better scalability, usability etc. It’s obvious that certain things can only be done with certain frameworks, say if speed is your utmost concern then you should choose a framework that delivers the best speed. However, all these thoughts can be put aside if you just wish to start with something which is what this tutorial is about. In this article, we will focus on PyTorch, one of the most popular Deep learning frameworks. We will learn to build a simple Linear Regression model using PyTorch with a classic example. PyTorch Overview PyTorch is a collection of machine learning libraries for Python built on top of the Torch library. It is widely popular for its applications in Deep Learning and Natural Language Processing. PyTorch also comes with support for CUDA which enables it to use the computing resources of a GPU making it faster. Machine Learning With PyTorch We will now implement Simple Linear Regression using PyTorch. Let us consider one of the simplest examples of linear regression, Experience vs Salary. We will train a regression model with a given set of observations of experiences and respective salaries and then try to predict salaries for a new set of experiences. You can create your own dataset of random numbers or you can also download the same sample here. Installing PyTorch Before we begin to code, make sure you have the PyTorch module installed. To install PyTorch simply type the below command in your terminal. pip3 install torch Note: If you are installing in a virtual environment make sure to activate it. For those who are using the Anaconda distribution, use <code>conda activate<\/code> command to activate the virtual environment. Importing the libraries We will start by importing all the important packages for this example. #Importing necessary Libraries import pandas as pd import torch import torch.nn as nn import numpy as np import matplotlib.pyplot as plt from sklearn.model_selection import train_test_split Importing the data We can now load our dataset into a Pandas dataframe with the following code block. #Loading the dataset data = pd.read_excel(\"exp_vs_sal.xlsx\") Visualizing the data We will look at a simple scatter plot of the given data. The below line of code will generate a simple scatter plot. #Plotting the Dataset data.plot(kind = 'scatter', x = 'Experience', y = 'Salary') Output: We can clearly see a linear relation between the salary  and experience from the above graph.We will move on creating a linear regressor. Splitting The Datasets Into Training And Test Sets The code block below will split the dataset into a training set and a test set. We will train the regressor with the training set data and will test its performance on the test set data. #Splitting the dataset into  training and testing dataset train, test = train_test_split(data,  test_size = 0.2) Converting The Data Into Tensors PyTorch uses tensors for computation instead of plain matrices.If you are wondering what the differences are and interested in knowing try reading this. Otherwise just know that tensors are more dynamic. So we need to convert our data into tensors. Before we convert, we need to pack each input or element in a list. Also for y(Salary) we should convert to type float as shown in the following code blocks. #Converting training data into tensors for Pytorch X_train = torch.Tensor([[x] for x in list(train.Experience)]) y_train = torch.torch.FloatTensor([[x] for x in list(train.Salary)]) #Converting test data into tensors for Pytorch X_test = torch.Tensor([[x] for x in list(test.Experience)]) This is what the new data will look like : Creating a Linear Regressor We have prepared out data, now it’s time to build the regressor.We will build a custom regressor by defining a class that inherits the Module Class of PyTorch. This practice will allow us to build a more custom regressor for the problem. 1.class LinearRegression(nn.Module): 2.    def __init__(self, in_size, out_size): 3.        super().__init__() 4.        self.lin = nn.Linear(in_features = in_size, out_features = out_size) 5.    def forward(self, X): pred = self.lin(X) return(pred) Let’s look at the line by line explanation below: Defining a class called LinearRegression that inherits PyTorch’s nn.Module class. Creating the init method for constructor. This function is invoked when an object is created for the class LinearRegression. Initializing the constructor of the parent class i,e nn.Module Creating object for PyTorch’s Linear class with parameters in_features and out_features. These parameters are the number of inputs and outputs at a time to the regressor. Defining the forward function for passing the inputs to the regressor object initialized by the constructor.The method will return the predicted values for the tensores that are passed as arguments. Getting Parameters from the model Since we are working on linear regression, we are trying to plot a best fitting line that passes through the points we saw in the scatter plot a while ago. To plot this line we need the parameters that are returned by the regressor. So we define a method for getting the parameter values  from the tensor objects returned by model.parameters() method where model is the object for the class Linear Regression that we will define soon. To achieve this we will first unpack the tensor objects returned by the model.parameters() method  into w and b as shown below. #Unpacking the parameters [w,b] = model.parameters() After unpacking the tensor objects, we will define a method to return the tensor objects as values(python numericals) when called. #A method for getting the parameter values from the tensor object def get_parameters(): return(w[0][0].item(), b[0].item()) Plotting the model We will now define a simple method for plotting the regression line. #A method for plotting the regressor def plot_model(name): plt.title(name) plt.xlabel('Experience') plt.ylabel('Salary') w1, b1 = get_parameters() X1 = np.array([-15, 15]) Y1 = w1 * X1 + b1 plt.plot(X1, Y1, 'g') plt.scatter(X_train,y_train) plt.show() The above function when called will get the parameters from the model and plot a regression line over the scattered data points. Setting random seed If you are familiar with sklearn then you will obviously know the random_sate parameter or if you are R user you would know seed method, both of these have the same functionality of providing reproducibility of regression. That is for a particular value of seed the regressor will always return the exact same results. # Setting the seed or random_state for reproducibility torch.manual_seed(1) Initializing the regressor Finally we create an object for the regressor we defined. The arguments passed here are the in_features(in_size) and out_features(out_size) respectively. #Initializing the Linear model model = LinearRegression(1 , 1) The arguments mean that the regressor will take one input and return one output at a time. With the below code we will print the values of the initial parameters set by the regressor. You will see that the model.parameters() method returns tensor objects. (To only get the value we use the item() method as we defined earlier in the get_parameters() method  ) #Printing the initial model parameters print(list(model.parameters())) Output: [Parameter containing: tensor([[0.5153]], requires_grad=True), Parameter containing: tensor([-0.4414], requires_grad=True)] Plotting the regression with initial parameters #Plotting the regression with initial weight and bias plot_model(\"Initial Model Plot\") Output: As you can see above that the green line is close to zero.This is because the initial parameters we so close to zero. Initializing the loss function and optimizer for the regressor We will now provide a loss function and an optimizer for out regressor. We will use Mean Squared Error for loss calculation and Stochastic Gradient Descent algorithm for optimizing and minimizing the error #Initializing the loss function as Mean Squared Error loss_fun = nn.MSELoss() #Initializing the optimizer as Stochastic Gradient Descent with the model parameters and  learning rate 0.01 optimizer = torch.optim.SGD(model.parameters(), lr = 0.01) Training the model Finally it’s time to train our model with the training data we saved earlier. We will use a for loop to iterate through epochs or cycles of training. We will also store the losses during each cycle in a list. # Training the model 1.epochs = 100 2.losses = [] 3.for i in range(epochs): 4.    y_pred = model.forward(X_train) 5.    loss = loss_fun(y_pred, y_train) 6.    print(\"@epoch : \", i, \" #Loss : \", loss.item()) 7.   losses.append(loss) 8.    optimizer.zero_grad() 9.    loss.backward() 10.    optimizer.step() Let’s look at each line. Define the number of epochs or cycles of training.Here the training will run a cycle of 50 time each time updating its weights and trying to minimize the error A list for storing the losses at each cycle\/epoch A for loop to iterate over the epochs Feeding the regressor with training data X_train Passing the predicted value of X_train and actual observation to the loss function to calculate the loss Printing the epoch and loss Saving the loss into  a list Setting the optimizer to zero gradient before backward propagation. This will help the SGD in moving the right direction of slope Backpassing the losses generated at an epoch Calling the step method performs a parameter update based on the current gradient and proceeds Output: Visualizing the loss #Visualizing the loss curve plt.plot(range(epochs), losses) Output: From the above image image we can see a considerable decrease in loss from epochs 0 to 3. From the 3rd or 4th epochs the loss keeps almost steady. This means the model cannot further optimize itself. Now we will jump ahead and see how out trained model drew the regression line. Visualizing the regressor after training #Visualizing the trained regressor plot_model(\"Trained Model\") Output: From the above image, we can see that the regression line almost passes through or is closer to most of the data points. Good job. Our regressor is now officially qualified to predict Salaries for new Experiences. Predicting Salaries for test data #Predicting for X_test y_pred_test = model.forward(X_test) Comparing Actual observations and Predictions #Converting predictions from tensor objects into a list y_pred_test = [y_pred_test[x].item() for x in range(len(y_pred_test))] # Comparing Actual and predicted values df = {} df['Actual Observation'] = y_test df['Predicted Salary'] = y_pred_test df = pd.DataFrame(df) print(df) Output: #Visualizing Actual and predicted values df.plot() Output: We can see how close the predictions are to the actual observations.Great Job !! You have created your first ever regressor in PyTorch.","excerpt":"It’s often hard to make a decision on what framework to learn when there are many options to choose from. In the Machine Learning spectrum, we have multiple frameworks that compete against each other claiming to have better speed, better scalability, usability etc. It’s obvious that certain things can only be done with certain frameworks, […]","categories":["Deep Tech"],"tags":["Gradient Descent Numerical Example","hands-on","linear regression","Python","Pytorch"],"author_name":"Amal Nair","publish_date":"2019-07-01T07:26:12","publication_year":"2019","word_count":1708,"keywords":["linear regression","hands-on","Pytorch","machine learning","NumPy","TPU","AI","Gradient Descent Numerical Example","PyTorch","Python","Ray","Aim","deep learning","Matplotlib","Pandas"],"extracted_tech_keywords":["AI","machine learning","deep learning","Aim","Ray","PyTorch","Pandas","NumPy","Matplotlib","TPU"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/introductory-guide-to-pytorch-using-a-linear-regression-problem\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10071836,"title":"The forgotten story of Palm Inc","content":"This year marks the 30th Anniversary of Palm Inc, the once-mighty, now-defunct maker of the pioneering 1990s personal digital assistants (PDAs) and smartphones. This California-based company single-handedly built the market that iPhones and Android devices dominate today. As the company’s co-founder Donna Dubinsky tweeted, “Some days it feels like it was yesterday. But mostly, it feels like another lifetime.” Tracing the Palm story with its soap-operaesque twists and turns is hard even by Silicon Valley standards. It started in 1992 as a small company in the handheld computing industry, was acquired by another company that then sold itself. It went public, was split in two, reunified, and sold itself off again. It lost its original team along the way but then got them back, only to bid farewell to them again. And then it died. Until 2018, when it was briefly revived and silenced all over again. Let’s deep-dive. A star was born Jeff Hawkins launched Palm Computing in 1992 to create software for PDAs manufactured by other companies, such as Tandy Corp’s Zoomer – the only device that rivalled Apple’s Newton. Interestingly, the company even developed an add-on handwriting-input software for the Newton-dubbed Graffiti. Jeff hired former Apple executive Dubinsky as the CEO along with Radius’ Ed Colligan as VP of marketing. In 1995, Palm Computing Inc was acquired by Skokie, an Illinois-based US Robotics Corp. This is when they unveiled their own range of PDAs. The pen-based Pilot 1000 and Pilot 5000 were the first to hit the market in March 1996. It powered the company’s proprietary operating system. Even though it won PC Computing’s MVP Usability Achievement of the Year Award, critics noted that the initial Palm PDAs lacked keyboard functionality and messaging features. The year 1997 was a busy one for the business, with sales coming in at over $114 million. US Robotics launched the Network HotSync technology, which allowed users to connect to desktop computers or computer networks via an optional snap-on modem. When 3Com Corp bought US Robotics in May, Palm became a division of 3Com. This was the largest US data networking merger at the time as it involved $6.6 billion in shares and produced a new business with $5 billion in sales. By the end of the year, Palm decided to provide other manufacturers licences to use the Palm OS platform. By 1998, the company had hit $272 million in revenues and unveiled the Palm III handheld. Everything seemed perfect until Palm co-founders Jeff Hawkins and Donna Dubinski abandoned the ship as 3Com wouldn’t spin off the handheld starlet. The duo started Handspring, eventually becoming one of Palm’s main rivals. A little more than a year after Handspring began operations, it had 21% of the market. Palm’s market share decreased from 83 per cent to 63 per cent. By the end of 1998, 3,500 devoted programmers were producing applications for the Palm OS. The demand for Palm’s products continued to grow, and they were being adopted in several new industries, such as the financial and healthcare sectors. However, Microsoft, which had released its Windows CE platform for mobile computers the year before, was stepping up the battle. Devices running Windows CE were available from a number of producers, such as Casio Computer Co., Hewlett-Packard Co., and Compaq Computer Corp. Despite Microsoft’s presence in the market, Palm continued to do well in 1999 with new product launches such as Palm IIIx, Palm V, Palm VII, Palm IIIe, Palm Vx, and Palm IIIe Special Edition. The Palm VII was notable because it allowed users to connect wirelessly to the internet via the subscription-based Palm.net wireless Internet service. Palm developed new technology known as “Web clipping” that allowed Palm VII users to visit participating websites and access the most relevant information from them. The new Palm VII also supported wireless messaging. Revenues were approximately $563 million that year. The company also solidified a number of alliances with companies such as TRG, Sun Microsystems, Nokia, and Computer Associates International Inc. A consolidator on steroids In the year 2000, Palm’s revenues reached approximately $1.1 billion. They spun out of 3Com and were named to the S&P 500. Palm Inc filed its IPO in March 2000, becoming an independent company traded on the NASDAQ stock exchange. This day in 2000: Palm Inc. IPO. Stock soared, raising the company valuation to $50B+.And it all started with Jeff Hawkins testing the idea with wood+paper prototype (I call it a Pinocchio pretotype). The Palm Pilot was The Right It!Thx @JonErlichman for original tweet. pic.twitter.com\/beG82uZkE7— Alberto Savoia (@Pretotyping) March 3, 2019 A slew of new Palm handheld devices also appeared in the market. The majority of their utility stemmed from Palm VII’s wireless Internet connectivity. Palm entered the enterprise market in 2000 as well. It released HotSync Server software, which enabled businesses and organisations to manage handheld computers on a network level. Rather than synchronising devices to a single PC, users could do so through a server, either within the organisation at any number of terminals, from PCs remotely connected to the company’s network, or via a wireless connection. Microsoft attempted to enter this market as well in the fall of 2001. By late 2001, Palm’s developer base had reached 140,000, generating $1.6 billion in revenue. They had a global reach, selling products in over 35 countries. According to a July 2001 market analysis and forecast from International Data Corp, Palm controlled 55.9 per cent of the global market for personal companion handheld devices. Its global market share in handheld operating systems was 71.8 per cent. Palm purchased AnyDay.com, which produced Internet-based calendars, for $80 million in cash and stock options to expand its wireless Internet services. It also purchased e-mail provider Actual Software Corp. Palm planned to include expansion slots in its organisers by early 2001, trailing Visor and Pocket PC devices in this area. Sony was also preparing to launch its own PDA. In a race to develop more advanced features in a shrinking market, Palm purchased several companies to gain an edge. For example, it paid approximately $40 million to acquire wireless-synchronisation expert WeSync in late 2000. In a slowing economy, these purchases were insufficient to prevent a collapse in the handheld computer market. The company’s new m500 and m505 models were announced two months early, which did not help sales of units already on the market. Palm had to cancel a plan to acquire Boise’s Extended Systems, Inc., a wireless technology producer for the corporate market, for $264 million in stock as its share price began to fall. Palm laid off employees for the first time in May 2001, putting a halt to a planned 11-building headquarters complex. PDA manufacturers introduced steep discounts throughout the year to keep the machines moving. This was when Palm was forced to deal with the familiar threat posed by Handspring and Microsoft alongside emerging players like Blackberry, who were making a lot of noise. As such, Palm paid $11 million in stock for former Apple executive Jean-Louis Gassee’s Be operating system in 2001 and split it into two business units: hardware and software. The company’s software division, PalmSource, took control of the Palm brand in August 2002 and relocated to its headquarters. Palm Solutions Group, the hardware division, quickly evolved into two distinct product lines: the high-end Tungsten brand for professionals and the Zire brand for new users and students. Palm also attempted to expand its corporate presence by providing features such as wireless database access and high security. A rollercoaster of a Decade In June 2003, Palm, Inc. announced it was purchasing rival Handspring in a stock swap (worth $240 million when the deal closed in October). In October 2003, Palm, Inc. spun off PalmSource, its software unit, while the original company, which retained the hardware operations, was renamed PalmOne, Inc. PalmOne still led the traditional PDA market with a 40 per cent share, while Handspring had become a niche player in the emerging “smartphone” market, which was exploding. PalmOne was renamed Palm, Inc. in July 2005 after it bought the remaining rights to use the Palm name for $30 million. Palm introduced its Life Drive line, which featured four gigabyte hard drives, in the spring of 2005. This was the company’s first attempt to create a larger, all-encompassing mobile manager system designed to meet all of your requirements. To speak of it now, with its contacts, calendar, music, images, video, applications, 4GB of microdrive HDD storage, Bluetooth and Wi-Fi, and all the software from the Palm OS, it doesn’t sound too dissimilar to modern smartphones and appears to be a good idea. It even had a switch to toggle between portrait and landscape mode. The problem was that all of this made it a very large and expensive device at the time. Production ceased less than two years into the cycle, not long before others came up with a similar but more refined idea. In September, it announced a new range of Treo devices (previously developed by Handspring and offered under the palmOne brand) with Microsoft’s operating system to help stave off competition from Blackberry. Unfortunately, like Life Drive, this product was discontinued in just four years. The same fate was seen with Palm’s first and only laptops, Foleo, Centro, Pre, and Pixi. The Final Chapter Palm sold itself to HP for $1.2 billion after losing all hope in 2010. The computing behemoth stated that it intends to release phones, tablets, PCs, and other products based on WebOS. (Under the HP brand, it discontinues Palm.) But then HP fired Mark Hurd, the CEO responsible for the deal. In 2011, the company launched the TouchPad, a potential iPad killer, under new CEO Léo Apotheker. However, it closed the WebOS business seven weeks later. https:\/\/twitter.com\/girlie_mac\/status\/1453922441793728516 Fast forward to 2018, a new startup backed by TCL and Golden State Warrior Stephen Curry adopted the Palm brand for a tiny Android-powered device designed for use when you’re working out. And that was the last time we ever heard of the brand. In Palm’s history, every business decision taken after the first chapter was too little, too late. They gave Sprint an exclusive deal to sell the Pre and Pixi and continued it when there were no sales. Palm neither let developers build WebOS apps or an equivalent of the iPod touch. Palm never marketed Pre in a way that mattered while iPhone, BlackBerry, and Android devices blew the roof. Given how things went, even their investors began publicly trashing the company. Ironically, Palm’s biggest competition, BlackBerry, followed the same downward spiral. The company began collapsing in 2013 due to a 40 per cent plunge in sales, leading to its shares tumbling down by nearly 17%. The company fired 4,500 employees and recorded an inventory write-down of nearly $960 million for that fiscal second quarter of that year. However, unlike Palm, they had no prospects for a last-minute exit. On January 4, 2022, BlackBerry’s once-famous messaging service was completely defunct.","excerpt":"By late 2001, Palm’s developer base reached 140,000, and the company’s generated $1.6 billion in revenue.","categories":["AI Features"],"tags":["Apple","Nokia"],"author_name":"Sri Krishna","publish_date":"2022-07-29T17:00:00","publication_year":"2022","word_count":1825,"keywords":["Go","AI","Apple","Git","RAG","Nokia","BERT","ViT","CLIP","GAN","R","startup"],"extracted_tech_keywords":["AI","RAG","R","Go","Git","BERT","CLIP","GAN","ViT","startup"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/the-forgotten-story-of-palm-inc\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":6,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10038929,"title":"Guide to LayoutParser: A Document Image Analysis Python Library","content":"Documents containing a combination of texts, images, tables, codes, etc., in complex layouts are digitally saved in image format. Analyzing and extracting useful information out of these image documents is performed with the help of machine learning. This supervised task is termed as Document Image Analysis (DIA). The popular DIA tasks in practical use include: Document Image ClassificationLayout DetectionTable DetectionScene Text DetectionCharacter Recognition There have been a few task-specific applications such as OCR (Optical Character Recognition) in real-world usage over decades. However, a library that provides all DIA tasks in one place became an important need of document analysis society, such as historical researchers and social science analysts. For instance, a screenshot image of an old newspaper’s page may contain historic research-centred contents in the form of tables, charts, texts and photographs. An OCR reader can be used to extract texts but cannot read other information. Moreover, an OCR reader may miss to recognize the text layouts and mix texts from different layouts in its output. A separate method will be required to extract information from tables, charts and so on. The evolution of deep learning-based convolutional neural networks has begun to try to give solutions to the need of an integrated Document Image Analysis system. However, the practical implementation of recent successful deep learning models has faced some challenges. High-level DIA parameters are not always explicitly processed by deep learning frameworks. This makes customization of pre-trained models difficult. Popular models are trained on a particular set of annotated document images. Documents do not possess any common template and formats and are limited only by human creativity. This needs to collect task-specific annotated document images, preprocess them according to the model requirements, and fine-tune the model with those images in case of custom implementations of a popular model. The deep learning network part and the DIA part are usually trained separately to make customized fine-tuning difficult, tedious, and time-consuming. To this end, Zejiang Shen of the Allen Institute of AI, Ruochen Zhang of the Brown University, Melissa Dell and Jacob Carlson of the Harvard University, Benjamin Charles Germain Lee of the University of Washington, and Weining Li of the University of Waterloo have introduced LayoutParser, a Python library for Document Image Analysis. This library has a Model Zoo with a great collection of pre-trained deep learning models with an off-the-shelf implementation strategy. This library has a unified architecture to adapt any DIA model. Apart from the usage of pre-trained models, LayoutParser provides tools for customization and fine-tuning as per need. Further, data preparation tools- for tasks such as document image annotation and data preprocessing tools are readily available in this library. The library aims at quality models and pipelines distribution with reproducibility, reusability and extensibility through a continuously improving community platform. The architecture of the LayoutParser library (source) LayoutParser performs one or more of the following DIA usages: It receives document images as input. It offers off-the-shelf tools for any DIA task. It performs the tasks in order and yields the output.It receives unannotated document images. It provides tools for efficient annotation of layouts and other parts of a document image.It supports efficient custom training for user-specific tasks. Once trained, the model can be employed for inference.It offers tools for visualization and storage of data, models, weights and checkpoints.It provides community sharing, distribution, and documentation. To store a layout in memory and retrieve it back, LayoutParser offers unified data structures. Three key components in the LayoutParser data structure are Coordinate, TextBlock, and Layout. Unique operations are defined in LayoutParser to process the library-defined data structures. Special operations and transformations in the LayoutParser library to process its own data structures (Source). We discuss the code implementation and two practical applications of the library in the sequel. Layout Detection in a Document Image Install the LayoutParser library and its dependencies from the PyPi packages. %%bash pip install -U layoutparser # install detectron2 pip install 'git+https:\/\/github.com\/facebookresearch\/detectron2.git@v0.4#egg=detectron2' # install OCR module pip install layoutparser[ocr] Import the libraries and modules. import layoutparser as lp import matplotlib.pyplot as plt import matplotlib %matplotlib inline import cv2 Deploy a pre-trained Detectron2 model configured for layout parsing. model = lp.Detectron2LayoutModel('lp:\/\/PubLayNet\/faster_rcnn_R_50_FPN_3x\/config', extra_config=[\"MODEL.ROI_HEADS.SCORE_THRESH_TEST\", 0.8], label_map={0: \"Text\", 1: \"Title\", 2: \"List\", 3:\"Table\", 4:\"Figure\"}) Now the model is ready for inference. Download the source files from the official repository to obtain a sample image to perform inference on it. !git clone https:\/\/github.com\/Layout-Parser\/layout-parser.git Output: Change directory to read the example data. %cd \/content\/layout-parser\/examples\/data\/ !ls -p Output: Read the ‘paper-image.jpg’ and display it. img = cv2.imread(\"\/content\/layout-parser\/examples\/data\/paper-image.jpg\") # convert BGR image into RGB format image = img[..., ::-1] # display image plt.figure(figsize=(12,16)) plt.imshow(image) plt.xticks([]) plt.yticks([]) plt.show() Output: Predict the layouts in the above image using the pre-trained model. layout = model.detect(image) Display the image with predicted layouts over it. lp.draw_box(image, layout, box_width=3) Output: This Colab Notebook contains the above example code implementations. OCR from Table Document Image Install the LayoutParser and its dependencies. In addition, install an OCR engine. Here, we use the TesseractOCR engine to recognize text and its location. %%bash pip install -U layoutparser pip install layoutparser[ocr] sudo apt install tesseract-ocr sudo apt install libtesseract-dev Import the necessary libraries and modules. import layoutparser as lp import numpy as np import pandas as pd import matplotlib.pyplot as plt import matplotlib %matplotlib inline import cv2 Load the pre-trained TesseractOCR engine. model = lp.TesseractAgent() Prepare data from the source code. Download the source files from the source repository and change the directory to denote the example images path. !git clone https:\/\/github.com\/Layout-Parser\/layout-parser.git %cd \/content\/layout-parser\/examples\/data\/ !ls -p Read the image and display it to have an idea of how it looks. image = cv2.imread('example-table.jpeg') # display image plt.figure(figsize=(12,16)) plt.imshow(image) plt.xticks([]) plt.yticks([]) plt.show() Output: Detect text characters with the OCR engine. Collect the text along with its bounding box details for plotting and post-processing. res = model.detect(image, return_response=True) # collect text and its bounding boxes ocr  = model.gather_data(res, lp.TesseractFeatureType(4)) Plot the original image along with bounding boxes on recognized texts. lp.draw_text(image, ocr, font_size=12, with_box_on_text=True, text_box_width=1) Output: We can recognize that the output texts are reproduced with Engine-specified fonts and sizes. Thus the system has recognized texts and their locations precisely. Further, we can post-process these texts in a column-wise manner or row-wise manner as per need. This Colab Notebook contains the above example code implementations. Wrapping Up In this article, we have discussed the open-source LayoutParser library, its architecture and capabilities. Further, we discussed two practical use cases of Document Image Analysis with hands-on Python codes. With more inclusion of new models in the near future, LayoutParser will get a prominent place in Document Image Analysis. References Github RepositoryResearch PaperHistoric Document Images DatasetCustom Annotation ExampleLayout Parsing ExampleTable Image OCR Example","excerpt":"LayoutParser is a Python library for Document Image Analysis with unified coding and a great collection of pre-trained deep learning models","categories":["Deep Tech"],"tags":["documents","Guide","machine learning document classification","Object Detection","OCR","python visualize neural network"],"author_name":"Rajkumar Lakshmanamoorthy","publish_date":"2021-04-26T15:00:00","publication_year":"2021","word_count":1110,"keywords":["NumPy","machine learning","OCR","machine learning document classification","AI","neural network","documents","RAG","Colab","Aim","deep learning","Object Detection","python visualize neural network","Matplotlib","Guide","Pandas"],"extracted_tech_keywords":["AI","machine learning","deep learning","neural network","Aim","Colab","Pandas","NumPy","Matplotlib","RAG"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/guide-to-layoutparser-a-document-image-analysis-python-library\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":9566,"title":"Sachin Tendulkar becomes Smartron’s Strategic Shareholder and Brand Ambassador","content":"Smartron India Private Limited, India’s first truly global IoT brand, announced a new innings with the Cricket legend, Sachin Tendulkar. Besides being the brand ambassador of Smartron, Sachin has also become a strategic shareholder in the company. Smartron commissioned its operations in Hyderabad and Bengaluru in August 2014. The company has so far invested over $10 Million to support the R&D activities and plans to spend $100 million over the next couple of years to establish Smartron as a global brand. Taking the first step towards achieving this goal, it has launched its first two products, t.book & t.phone coupled with tron.x. Headed by Founder & Chairman Mahesh Lingareddy along with Narsi Reddy (Co-Founder & Managing Director) and Rohit Rathi (Co- Founder & President) and backed up by the legendary cricketer, Sachin Tendulkar; Smartron is riding on the back of the IoT wave by innovating in the areas of smart, sensor, robotics, artificial intelligence, cloud and big data technologies to bring next generation smart devices, services and care targeting consumer, enterprise, industrial and infrastructure markets. Commenting on the association, Sachin said, “We are moving towards Internet driven experiences and Smartron’s vision in unlocking the potential of IoT will bring compelling experiences to people, making their life more productive, safe, intelligent and convenient. An Indian brand making a great leap into technology product innovation is commendable! We need to support brands that are “designed and engineered in India” by our Entrepreneurs and Engineers that aptly compete in the global markets. I look forward to being part of the growth journey of Smartron buoyed by tremendous growth opportunities offered by IoT worldwide.” Likewise, Mahesh (Founder & Chairman, Smartron) feels privileged to have Sachin join them in their journey. He further added that, “Being one of the greatest Indian stories of twenty first century, a true global icon and brand that disrupted the game of cricket, he has inspired generations. We aim to achieve the same in the technology space by building a global product brand from India, disrupting the IoT market with innovative products and inspiring generations of engineers and entrepreneurs.” Smartron believes that it is the right time to size up the opportunity of Internet of Things (IoT) and design products in India with an impressive technology for a global audience, while creating a network of integrated smart devices for seamless communication. On the product line up, we were told that there’s an exciting lineup of products for 2016, beyond smartphone and ultrabook segments across mobile, home, auto, enterprise and infrastructure markets. About the products: t.book is an impressive high performance Ultrabook convertible while t.phone claims to be the lightest 5.5” Smartphone with tron.x. Powered by Intel® Core™ M processor and a Microsoft Windows 10 Operating System, the t.book aims at enhancing the user experience. These next generation of smart IoT devices, platforms and services from Smartron are built on the breakthrough concept of Internet of TronsTM (IoT). Tron.x seamlessly integrates all the devices, software, cloud, crowd, hub, services and care platforms to deliver highly intelligent and personalized experience through hubtron, a one-stop destination to access all the data, content, services and care and to shop for all other devices and accessories and to be participate in the community and to enjoy the amazing tron xperience. Hubtron will be active across all Smartron devices. Apart from this, the t.cloud feature will enable Smartron users with free unlimited storage and the data can be accessed across all devices using Hubtron. The Smartron customer service, which is known as t.care, will offer online and offline support. The offline support will be present across 100 cities and with doorstep services making it hassle free. Smartron t.book, the perfect mobile workstation is on sale exclusively on Gadget360.com from April 8th, 2016 and the t.phone booking will start on April 18th, 2016.","excerpt":"Smartron India Private Limited, India’s first truly global IoT brand, announced a new innings with the Cricket legend, Sachin Tendulkar. Besides being the brand ambassador of Smartron, Sachin has also become a strategic shareholder in the company. Smartron commissioned its operations in Hyderabad and Bengaluru in August 2014. The company has so far invested over […]","categories":["IT Services"],"tags":["analytics funding","IoT","IoT India"],"author_name":"Apoorva Verma","publish_date":"2016-04-13T04:51:32","publication_year":"2016","word_count":636,"keywords":["big data","Go","artificial intelligence","AI","IoT India","innovation","ML","RAG","Aim","ViT","analytics funding","R","IoT"],"extracted_tech_keywords":["AI","artificial intelligence","ML","Aim","RAG","R","Go","big data","ViT","innovation"],"url":"https:\/\/analyticsindiamag.com\/it-services\/sachin-tendulkar-becomes-smartrons-strategic-shareholder-brand-ambassador\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10166245,"title":"Wi-Fi Troubles are About to be a Thing of Past, Thanks to AI","content":"Imagine walking into your office on a Monday morning. You grab a coffee, settle into your chair, and open your laptop, only to find the Wi-Fi not working, your CRM tool struggling to load, and video calls dropping in quality. Frustration sets in. A network engineer is alerted, diving into logs and running tests, but the issue remains a mystery for hours. This scenario is common, but now, AI-driven network operations are solving this problem. Bengaluru-based Juniper Networks, which has a team of 4,000 employees, handles the complete product lifecycle, from concept to development, testing, and deployment. In an exclusive conversation with AIM, Sajan Paul, area vice president and country manager of India and SAARC at Juniper Networks, said, “Seven years ago, before AI had even become mainstream, we saw its potential in automation. We knew that managing massive networks would soon be beyond human capability alone.” Paul, a key leader in this AI-driven shift, likens the journey to teaching a child. “You train a child, and in an appropriate environment, they should behave as taught. AI is no different. We’ve built large models and refined them through seven generations, and today, we achieve over 90% efficacy in our AI-powered network solutions,” he explained. He further mentioned that their AI systems don’t just process data; they turn decades of network expertise into real-time, actionable insights. A problem that was solved 10 years ago should never need to be solved again. AI ensures that by contextualising past solutions and proactively addressing similar issues before they occur. AI in Networking AI’s role in networking has evolved through three key stages. The first is the recommendation phase, where AI suggests solutions, but human engineers still make the final call. Next comes the automation phase, where AI takes over routine tasks, significantly reducing manual interventions. Finally, the ultimate goal is self-driving networks, where AI detects, analyses, and resolves issues before they impact users. Paul noted that they are in the seventh generation of AI maturity, with intelligent automation spanning our entire product portfolio. Global companies, from ServiceNow to some of the world’s largest IT, healthcare, and manufacturing firms, rely on Juniper Networks to power their operations because, in today’s hyper-connected world, downtime is simply not an option. The Meesho Story AI isn’t just about individual user experiences; it plays a crucial role in ensuring businesses stay up and running without disruption. For instance, Meesho, one of India’s leading e-commerce platforms, has a vast network of outsourced customer service agents handling queries in seven languages. The company needed a high-performing, AI-driven Software-Defined Wide Area Network (SD-WAN) solution to support its Voice over Internet Protocol (VoIP) applications, especially during peak call times. Their business depended on it. Thanks to AI, Meesho not only improved call volume and retention rates but also built a service-centric network that fuels business growth. Ismail Mohideen, director of IT at Meesho, mentioned that Juniper’s AI-driven SD-WAN solutions have not only increased call volume and retention rates but also built a service-centric network that fuels business growth. This partnership brings them closer to democratising internet commerce while improving customer satisfaction. AI-powered automation ensured that their customer service centres stayed online, call quality remained flawless, and agents could assist sellers and buyers without interruptions, no matter the time of day. What’s Next? One of the biggest challenges in network troubleshooting is catching problems before they escalate. Traditionally, engineers would wait for an issue, scramble to capture logs, and analyse data after the failure, often missing the critical moment. AI changes the game entirely. It continuously monitors network health, detects anomalies instantly, auto-captures logs before a failure happens, and sends real-time data to engineers for immediate action. The results speak for themselves: a 90% reduction in trouble tickets, solutions deployed were nine times faster, and an 85% decrease in operational expenses. Every network device Juniper deploys now comes embedded with an AI agent, functioning like the Siri or Alexa of network management, working quietly in the background to ensure near-zero downtime. So, AI-driven networks are no longer just a vision; they are already here.","excerpt":"Meesho used AI to improve call volume and retention rates and built a service-centric network that fuels business growth.","categories":["AI Features"],"tags":["AI (Artificial Intelligence)","India","Juniper Networks"],"author_name":"Vidyashree Srinivas","publish_date":"2025-03-18T18:00:00","publication_year":"2025","word_count":680,"keywords":["Go","intelligent automation","AI","ETL","Scala","Juniper Networks","automation","Aim","disruption","Rust","R","India","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","Aim","R","Go","Rust","Scala","ETL","automation","intelligent automation","disruption"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/wi-fi-troubles-are-about-to-be-a-thing-of-past-thanks-to-ai\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":15535,"title":"Real life Real Steel &#8211; humanoid robots from Japan &#038; US to battle for supremacy","content":"Prepare for the biggest fight in the history of Robots! Growing up most of us have binged on movies and TV shows that involve giant fighting robots, or read comics which depict giant mechanized fighting machines. However, these robots were largely consigned to television, games, books, and science-fiction. Now, that’s set to change with a showdown between two massive robots prepare for a hand-to-hand, East-meets-West battle. The American creators of MegaBots have challenged Japan’s Suidobashi Heavy Industry to bring its Kuratas robot to the battlefield, which the latter readily accepted. Seems like the Hugh Jackman starrer Real Steel will be coming to life soon. Gui Cavalcanti, Co-founder and CEO, MegaBots said “We want to bring the giant robots from science fiction and movies and video games to life because now we have the technology.” MegaBots’ Challenge With this challenge, the MegaBots founder envisions a sports league, where teams worldwide build huge humanoid robots which can take on each other in a fight-based competition. “Everyone wins as long as there is robot carnage. “People want to see these things fight. They want to see them punch each other, they want to see them ripped apart and they want to be entertained,” extols, Matt Oehrlein, Co-founder, MegaBots. The organization was co-founded by Matt Oehrlein, Gui Cavalcanti, and Brinkley Warren in 2014. Japanese organization raised the stakes by suggesting that their robots participate in a hand-to-hand combat, making a departure from American-gun toting bots. Kogoro Kurata, CEO, Suidobashi Heavy Industry taunts, “Come on guys, make it cooler. Just building something huge and sticking guns on it, sounds super American.” The US Company has a long road to tread, before their prized robot, Mk.II, is ready to throw down instead of firing paintballs. The bigger picture that MegaBots is trying to paint here is to make gladiator-style robot combat into big-time entertainment. Over years, it could turn out to be a mix between Ultimate Fighting Championship and Formula One auto racing. This will lead to development of new industrial technologies, inspiring a new generation of engineers. Sizing up the humanoid robots MegaBots co-founders Matt Oehrlein, Gui Cavalcanti, and Brinkley Warren Tokyo-based Suidobashi Heavy Industries designed the 13ft (4m) Kuratas. Kuratas is a mean fighting machine featuring big, agile hands that mimic the movements of the pilot’s hand. Kuratas’ mech is extremely mature. The robot has been sold as a product for the super-rich in Japan in recent years. The robot weighs 9,000 pounds, and is slightly smaller at 12 feet. The machine has a single driver, and doesn’t sport any paintball gun. In the American corner, the robot has a room for two occupants, one driver and one gunner. Around 15 feet tall, tipping the weighing scale at 12,000 pounds, this piloted robot also sports a paintball cannon that can fire giant three-pound projectiles at more than 100 miles per hour. The Co-founders built this robot from scratch, while making a cavernous Oakland workshop their home base. Oehrlein remarks, “Our current robot, the Mark II, looks pretty intimidating. The truth is, it’s pretty slow. It’s top-heavy. It’s rusty, and it needs a set of armor upgrades to be able to compete in hand-to-hand combat.” Kickstarter campaign raises $550,000 for Mark II Megabots, Inc. earlier launched its Kickstarter campaign to raise half a million dollars The Oakland-based firm had earlier launched an online campaign on Kickstarter, through which it raised more than $550,000 from robot enthusiasts around the world. With this fund, MegaBots planned to turn its Mark II robot into a real fighting machine. The fund helped MegaBots to make its robot faster, tougher, and more balanced, besides equipping the machine with detachable weapons such as a giant chain saw or punching fist. Cavalcanti expresses, “We’re absolutely confident that Team USA can beat Japan. We’ve assembled the best of the best of this country. We’re not going to let our country down.” Besides, the startup has also enlisted the help of engineers from NASA, software maker Autodesk, and the TV shows including Mythbusters and BattleBots. Robot showdown in August The upgraded Mk.II concept – the most American robot ever made MegaBots has finally set and revealed a date when it plans to take on Suidobashi Heavy Industries’ sophisticated piloted mech, Kuratas. The much-awaited fight will happen this year, in the month of August. This date has been announced after two years and several teases on their YouTube Channel. To think about it, I’m sure all robot enthusiasts can barely control their excitement, as the fight is just two months away. The tournament rule dictates that none of the robots will use any kind of projectile weapon, they will instead have to participate in a complete melee-only combat. The co-founders of MegaBots are currently tight-lipped about the event, about it making history. Previous statements suggest that the battle will be on neutral ground, however no clear announcements has been made yet regarding the same. Besides, the original venue for the duel had fallen through. The final duel location is being kept a secret because of the earlier setback, and to prevent any further delays. However, viewers will be able to watch the battle for free on YouTube, no matter where they are located. Robot enthusiasts such as Gordon Kirkwood are eagerly anticipating the fight. “I think it’s going to be a smash hit. This has the potential to be a fantastic spectator sport that people would really pay good money to see,” concludes Kirkwood.","excerpt":"Growing up most of us have binged on movies and TV shows that involve giant fighting robots, or read comics which depict giant mechanized fighting machines. However, these robots were largely consigned to television, games, books, and science-fiction. Now, that’s set to change with a showdown between two massive robots prepare for a hand-to-hand, East-meets-West […]","categories":["IT Services"],"tags":["engineers India","NASA","robots India"],"author_name":"Amit Paul Chowdhury","publish_date":"2017-06-12T05:34:27","publication_year":"2017","word_count":909,"keywords":["Go","NASA","engineers India","programming_languages:R","AI","R","programming_languages:Go","Rust","GAN","robots India","programming_languages:Rust","startup"],"extracted_tech_keywords":["AI","R","Go","Rust","GAN","startup","programming_languages:R","programming_languages:Go","programming_languages:Rust"],"url":"https:\/\/analyticsindiamag.com\/it-services\/real-life-real-steel-humanoid-robots-japan-us-battle-supremacy\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":14709,"title":"Making sense of ‘Black Box’ in Artificial Intelligence &#8212; should we trust AI completely?","content":"The Black Box Technique for AI Neural networks, machine learning algorithms, and other subsets of AI are finding their way into several critical domains, which include healthcare, transportation, law, and more. And AI algorithms are affecting people’s lives in more ways than one, from credit scoring to loan disbursal to skewed image matching. With newer developments around the field, researchers are encountering a brand-new set of challenges. As AI algorithms grow more advanced, it becomes a challenging task to make sense of the inner workings. Moreover, companies that develop them do not allow the scrutiny of their proprietary algorithms, which is another reason AI is becoming increasingly opaque and complex today. Instances where AI has had an alarming impact: An AI-powered opponent in the game Elite Dangerous went berserk and started creating super-weapons to hunt players. Microsoft’s AI chatbot Tay started spewing out racist comments within a day of its launch. Google face recognition started making some offending labeling of pictures More notorious is  COMPAS Recidivism algorithm used to decide the freedom or incarceration of defendants passing through the US criminal justice system was alleged to be biased against African Americans, online investigative journalism site ProPublica claimed. At the end of the day, the concern mostly surfaces from the lack of human control. When humans make mistake, they can explain it, or take responsibility for that case. The same cannot be said for AI. Therefore, we need to make sense of how the underlying algorithms work. In fact, writer Cathy O’Neil, author of Weapons of Math Destruction discussed the downside of living in algorithmic world where  mathematical models have invaded lives and spawned unaccountability. The premise is how far should people go in trusting Neural Networks and Deep Learning. What is Black Box in AI? AlphaGo defeats Lee Sedol Neural networks, the key defining components of AI applications such as image recognition, natural language processing, speech recognition, machine translation among others have long been regarded as the “Black Box” as it is hard to understand how the results have been generated. Deployed in many real world applications, neural networks mimic the human brain but are not transparent as opposed to rules-based system. Engineers often fail to explain why their algorithms make certain decision. Neural networks form the key components for several AI applications, and can be considered as a “black box.” Google’s Go-playing AI program, AlphaGo is a brilliant instance. The program stunned the world when it executed moves, even the professional couldn’t think of. Basically, the network organizes itself when it is instructed to, however, that doesn’t necessarily imply that it will tell you how it was done. Decoding the ‘Black box’ technique? It’s never an easy task to train a neural network, as it can take hours to get them ready, no matter to what extent you make use of computing power. To address this problem, researchers at OpenAI came up with the ‘Black Box’ technique, that promises more robust AI systems. The technique uses something called “Black Box” in place of standard reinforcement training. The approach entails forgetting the fact that environment and neural networks are involved within the ‘Black box’ created. Essentially, the technique involves optimizing a given function in isolation, and sharing it as necessary. Black Box technique will help us understand neural networks The approach initiates with several random parameters, makes guesses, and then tweaks follow-up guesses to favor the more successful candidates. This helps in gradually diminishing things down to the ideal answer. Benefits: It eliminates a lot of the traditional craft in training neural networks, making the code both easier to implement and roughly two to three times faster. The method scales elegantly the more processor cores you throw at a problem, as the ‘workers’ in this scheme only need to share tiny bits of data with each other. Besides, this technique has other far-reaching advantages. With ‘Black Box’ technique, neural network operators can spend more time using their systems, instead of spending time on training them. There’s quite some time before the technique is extensively used for real-world AI applications. However, as computers get increasingly fast, the chances of such learning to occur in real-time only increases. The technique will also assist in building robots that are very quick to adapt to new tasks and learn from mistakes. Need for AI transparency Elon Musk To ensure that Artificial Intelligence as a technology is more transparent, firms have to spread awareness about it and evangelize people about its applications. OpenAI is a great example of such a firm. The nonprofit research company was founded by Tesla’s Elon Musk and YCombinator’s Sam Altman. The organization aims at opening AI research and development to everyone, independent of commercial interests. Partnership on AI is another great example of such an organization. The firm aims to create awareness for and deal with AI challenges such as bias. The company was founded by tech giants including Microsoft, IBM, and Google. The organization will also focus on AI ethics and best practices. Can we trust Artificial Intelligence? Can we trust AI Mistakes made in case of non-critical tasks, such as advertising, games, and Netflix suggestions, are tolerable. Thus, AI can be easily used for such applications without the fear of facing any major error. However, if such a mistake happens across domains such as the social, legal, economic, and political; consequences can be heavy. Thus, AI cannot be completely trusted by itself at the moment with applications like enterprise customer service, where transactions are involved, or computer-assisted clinical documentation improvement. In scenarios as such, the AI co-works with a human being, instead of working in isolation. As we move into the future, we will have to embrace AI and develop trust on the technology.  It will become necessary as AI will be able to diagnose deadly diseases, make million-dollar trading decisions, and do countless other things to transform whole industries. But to really trust the technology, we must better comprehend techniques such as deep learning, and make it accountable to the end users. Black Box technique is a positive step in this direction, showing researchers the way to understand neural networks and how they work.","excerpt":"Neural networks, machine learning algorithms, and other subsets of AI are finding their way into several critical domains, which include healthcare, transportation, law, and more. And AI algorithms are affecting people’s lives in more ways than one, from credit scoring to loan disbursal to skewed image matching. With newer developments around the field, researchers are […]","categories":["IT Services"],"tags":["AI Chatbot","AlphaGo","assisted intelligence","Black Box","Deep Learning","healthcare India","Machine Learning Algorithms","subsets of ai","transportation"],"author_name":"Amit Paul Chowdhury","publish_date":"2017-05-02T12:16:30","publication_year":"2017","word_count":1027,"keywords":["assisted intelligence","deep learning","transportation","R","artificial intelligence","image recognition","healthcare India","Machine Learning Algorithms","Go","AlphaGo","machine learning","AI","neural network","OpenAI","AI Chatbot","subsets of ai","Aim","Deep Learning","Black Box"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","deep learning","neural network","OpenAI","Aim","image recognition","R","Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/making-sense-black-box-artificial-intelligence-trust-ai-completely\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10000897,"title":"CES 2019 Roundup: Top 10 Wearables To Come Out Of Vegas","content":"Like every year, CES 2019 packed an outstanding lineup of wearables, with multiple companies using CES as a launchpad to announce new products. Here are the top 10 wearables from CES 2019, the list is in no particular order. App-Elles App-Elles is a wearable developed to prioritise the safety of women and children. Just like a fitness tracker bracelet, App-Elles connects with an Android\/iOS application that allows users to set 3 emergency contacts. Upon pressing the bracelet down, the device will then alert one of these contacts, who are also required to have the app. Upon alerting the contact, they will then be able to track the user’s live location in real-time. This is due to the bracelet having GPS built into it, along with facilities to transmit live audio to the contacts specified. Quell 2.0 The Quell 2.0 is a wearable band that classifies as a transcutaneous electrical nerve stimulation (TENS) device. Wrapping this band around the user’s leg will prompt specific neural pulses to their brain, triggering a pain relief response in the Central Nervous System. Users can wear this band strapped to their calf, no matter where the pain is. It can also be used while sleeping, with a companion app monitoring the user’s sleep, how much time they spent in bed and their overall sleep quality. Withings’ Move ECG Analog Watch The latest offering from Withings unveiled at CES was the Withings’ Move ECG analog watch. As implied by its name, it is an analogue smartwatch that connects with the users’ phones to provide health data. It has an electrocardiogram built into the watch, that will enable users to track metrics like pace, distance, number of burned calories, steps and floors climbed. ECG can be used to detect signs of atrial fibrillation in the user. Moreover, it also allows users to track sleep quality and monitor it throughout the night. Unlike other smart watches, the Withings does not require charging. Instead, the watch functions on a typical coin cell battery that will last close to one year. Garmin Vivoactive 3 Music Garmin, the prominent smartwatch manufacturing company, had a surprise for enthusiasts at CES this year. The company launched 4G-enabled GPS smartwatch, known as the Vivoactive 3 Music. Due to the availability of 4G, the device will allow for users to listen to music, send text messages, and track workouts without the need for a smartphone. The watch also includes a safety feature – incident detection technologies that is activated during select activities. This notifies emergency contacts of user’s locations in real-time upon an “incident” occurring. In addition to this, the watch comes with a bevy of sensors such as GPS, compass, thermometer, heart rate sensor, NFC, and a barometric altimeter. L’Oreal My Skin Track pH Falling on the weirder side of wearables, beauty product manufacturer L’Oreal released the My Skin Track pH at CES this year. This functions as a “stick-on wearable” that utilizes microfluidic technology to detect a user’s skin pH levels through their sweat. This sensor was developed under L’Oreal’s skincare brand La Roche-Posay in conjunction with Epicore Biosystems The wearable will capture trace amounts of sweat from skin pores. Following this, it will provide an accurate pH reading and recommend users with L’Oreal skincare products tailored to their skin type. Muse Softband Muse is a company known for producing wearables that aid in meditation, with their latest offering at CES this year being the Softband. It helps individuals meditate as they fall asleep, providing real-time feedback on what is happening inside their brain. This is done through the use of seven EEG sensors along the scalp of the user, and a heart rate monitor. While the user medidates, the headband measures whether their mind is calm or active, and translates that into audio. The band detects when the user is asleep, turns off the sound and tracks sleep related stats using an accelerometer for body movement and breath. It provides users with data such as wake time, REM\/Light\/Deep sleep, heart rate, respiration rate and movement. Halo Sport 2 The Halo Sport 2 is a wearable that aims to harness the power of neuropriming to power workouts. It is designed to be worn on the head, and uses pulses of energy to prime the brain into a state of ‘hyperlearning’. The wearable provides considerably better results when paired with workouts. The second iteration of this technology was shown off at CES this year, and has Bluetooth for a wireless experience. The app has also been updated to include tracking, reminders and brain training tips to be used in conjunction with the wearable. OptiShokz Revvez OptiShokz is a sister company of AfterShokz, a firm focused on bone conduction stereo audio. This is a method of delivering audio wherein instead of delivering it into the ear canal, it is pressed straight against the bone. This provides a similar experience to earphones without the need to wear earbuds. This technology was brought to sunglasses with the release of the Revvez, which pack a Bluetooth 4 wireless chip, bone conduction stereo audio, twin microphones for phone calls, and Google Assistant support into a pair of sunglasses. The bone conduction occurs at the end of the glasses’ handles, delivering audio to the cartilage behind the ear. Cochlear Nucleus 7 Sound Processor Healthcare in wearables also took a leap forward this year at CES, with Cochlear’s new Nucleus 7 Sound Processor being shown off at the conference. This is a behind the ear hearing solution, and claims to combined comfort with sound processing technology. However, this product has a twist in the form of smartphone compatibility. Users can stream from their iPhones for music, movies, phone calls and even Facetime, and can communicate with the help of two synchronized microphones that help filter out background noise. This shows the movement of wearables to a more healthcare oriented approach. Medtronic Micra Wearables aim to change the way  that healthcare is perceived today, with the Medtronic Micra at the forefront of this change. The world’s smallest pacemaker, it was shown off at CES this year. It is shaped like a bullet, allowing for implantation with a minimally invasive procedure. It offers a marked improvement over existing solutions, while still housing a multi-year battery, sensors and microchip hardware. As it regulates the heartbeats of patients, its FDA approval will come in handy to convince patients to buy it.","excerpt":"Like every year, CES 2019 packed an outstanding lineup of wearables, with multiple companies using CES as a launchpad to announce new products. Here are the top 10 wearables from CES 2019, the list is in no particular order. App-Elles App-Elles is a wearable developed to prioritise the safety of women and children. Just like […]","categories":["AI Trends"],"tags":["ces","Healthcare Automation","meditation","smart devices","Wearables"],"author_name":"Anirudh VK","publish_date":"2019-01-14T19:24:16","publication_year":"2019","word_count":1066,"keywords":["meditation","Go","ces","programming_languages:R","smart devices","AI","programming_languages:Go","Wearables","Healthcare Automation","Aim","ViT","R"],"extracted_tech_keywords":["AI","Aim","R","Go","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/ces-2019-roundup-top-10-wearables-come-out-vegas\/","complexity_score":4,"technical_depth":7,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10113762,"title":"Eventbrite&#8217;s CTO Vivek Sagi Unveils the Future of Events with AI","content":"AI, specifically generative AI, is spearheading about every industry now. Global events marketplace, Eventbrite, is bringing it into the events sector in the most creative way possible and creating efficient working environments. In a recent interview, Vivek Sagi, the CTO of Eventbrite, shed light on his professional journey, the company’s innovative use of AI and how it manages remote engineering teams, besides touching upon the future of event technology. Sagi, whose career spans various tech sectors, also shared his insights into Eventbrite’s approach to leveraging AI to enhance event creation, promotion, and attendee experience. Sagi’s journey to becoming the CTO of Eventbrite began with a tryst with AI during his schooling at IIT Madras and further studies at Penn State University and the Wharton School in the United States. He described his career as divided into four phases, from startups to big tech, leading him to Eventbrite, where he found his “sweet spot”. As a CTO, Sagi outlined his primary responsibilities, emphasising a customer-centric approach to managing a globally distributed engineering team spread over 13 countries. He highlighted the importance of staying up-to-date with rapidly evolving technologies to drive innovation and maintain the company’s market leadership. AI in events is all about personalisation One of the most intriguing aspects of Eventbrite’s innovative use of AI, as Sagi explained, is its integration into the platform, streamlining event creation, generating promotional content, and enhancing personalised recommendations for users. Through AI-powered tools, creators can launch ad campaigns faster and achieve better engagement, leading to increased ticket sales. “Using AI, we generate the entire event description with images and let customers pick the one that works for them,” Sagi explained how generative AI brought down the duration of creating promotional content from hours to just three minutes. “Generative AI has become really good at generating content such as copies for social media,” Sagi said, highlighting that an AI-generated copy, when done at scale, drives about 17% better cost per click compared to human-generated content. He also noted that creators who use AI tools are launching ad campaigns at a 30% faster rate. Sagi said that initial results indicated that event creators who use Eventbrite’s built-in tools for paid social campaigns sell up to 63% more tickets compared to those who do not. In 2022, Eventbrite issued nearly 300 million tickets to approximately 5 million events in nearly 180 countries. This has led to the company amassing large amounts of data, much of which is user-generated. Eventbrite is utilising this information – including unstructured data – to create personalised events for its customers, to connect them to the right creators, solve cross-channel problems, and make everything available at one spot. Sagi also discussed the future of AI in the events industry around advancements such as AI-generated videos and personalised event experiences. He emphasised the potential for AI to enhance human connections during live events, enabling even remote attendees to share in the experience. “With synthesised AI music, you can have a whole orchestra without paying for a label, also offering great personalisation,” said Sagi. Creating the perfect remote and asynchronous team environment In a fascinating personal insight, Sagi revealed that he gets inspiration for working remotely from a ranch in Texas, where he finds comfort in the company of nature, a menagerie of animals, and fund-raising efforts that support animal welfare. His unconventional work environment reflects Eventbrite’s commitment to flexibility and creativity in its approach to technology and innovation. Vivek Sagi, in his Texas Ranch,  inspires Techies to work from where they work best. Sagi shared Eventbrite’s approach, which prioritises empowerment, clear goal-setting, and the use of collaborative tools. He emphasised the importance of rituals like daily stand-ups and sprint demos while allowing teams flexible working arrangements. “It’s our passion to empower teams and give them full ownership of their work,” he said. “It is also essential that our teams do not work in a vacuum.” Sagi said that Eventbrite’s approach involves having all the different teams, such as creators, engineers, product managers, etc., in the same time zone. Sagi also believes that enterprise software has also been a game changer when it comes to working remotely efficiently. Further, Sagi pointed out Eventbrite’s commitment to a “flexible-first” environment, enabling employees to work remotely while maintaining essential team rituals and fostering collaboration. He also discussed the company’s shift toward a distributed executive team, in line with the trend of post-pandemic remote work. Eventbrite believes in a DevOps model, which means “when an engineer is done building the code, they can push it into production with the click of a button, and not depend on a chain of human beings along the way to make that happen”. “Letting the engineers be productive and getting the code out in front of our consumers and creators as quickly as possible has been the biggest technological strength that we have built,” concluded Sagi.","excerpt":"Global events marketplace, Eventbrite, is bringing it into the events sector in the most creative way possible and creating efficient working environments.","categories":["AI Highlights"],"tags":["AI events"],"author_name":"Mohit Pandey","publish_date":"2024-02-26T10:00:00","publication_year":"2024","word_count":813,"keywords":["Go","API","AI events","AI","ML","RAG","ViT","generative AI","GAN","DevOps","R"],"extracted_tech_keywords":["AI","ML","generative AI","RAG","R","Go","DevOps","API","GAN","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/eventbrites-cto-vivek-sagi-unveils-the-future-of-events-with-ai\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":24720,"title":"Is Data Science A ‘Pseudoscience’?","content":"Data science has established itself as an important asset in the technology sector. Businesses are observing a significant rejig in their functions as most organisations are now dealing with surplus data. Data Science has also got fields like data analytics, artificial intelligence and machine learning which are growing parallely with it. Though data science has been largely accepted, these advancements have sparked a question in the tech community: “How much of ‘science’ is in data science?” The criticism has also led to some people calling it a pseudoscience. In this article, we will discuss why there’s a big difference between labelling it as ‘science’ or ‘pseudoscience’. Orienting Data Science Towards Science Data science users should always open up about the methods used, instead of black-boxing them. For instance, data science algorithms are the ones that have an analytical aspect. They bring out the essence in the field. If algorithms are improved, they would definitely lead to a more methodical approach. Secondly, since this field looks specifically into data, the approach is to analyse a model or project with lots of useful data around it. This means digging deep around better and richer datasets for analysis. In addition, the model in consideration should also incorporate a statistical and mathematical viewpoint. This gives a more scientific allure. Thirdly, the data sources should be authentic. Collating from any type of sources across various tech domains will lead to more confusion and less understanding. Michael Howard, CEO of MariaDB provides an interesting take on this. Howard says, “Many software companies try to get around having rich data sets by claiming they use more signals than anybody else. A signal is a single data point — like a government database on education. Some companies engage in signal spin, counting every column of data sources they could potentially use (but usually don’t) to increase their signal count. To distinguish the heavyweights from lightweights, you have to have a data scientist dig deeper and start reviewing confusion matrices and F1 scores.” This signifies that companies should focus on its sources rather than following intuition. All of these factors contribute largely in establishing a better scientific outlook in data science. When Can Data Science Become A Pseudoscience? Beliefs or practices that float around in the pretext of science is known as pseudoscience. It has been a subject of discussion right from the 19th century. Noted philosopher Karl Popper brought the concept of falsifiability to identify abstraction surrounding scientific theories, which led to the unrevealing of pseudoscience. Many scientific theories such as the String Theory and Sigmund Freud’s psychoanalytic theories among others, are hard to base on scientific facts. Contrastingly, it is not right to reject them as non-scientific. This conundrum is always present. The only way science can be called as ‘science’ is with solid, provable facts. Famous physicist Richard Feynman once said that the key to build a scientific approach was to present it in a simple language and in layman’s terms without the obvious use of technical terms, and see if it makes perfect sense. This is relevant when it comes to the context of data science. More technicalities in the field mean obfuscation. Aligning data science with simplification would make it easy to understand and clear the notion of it being a pseudoscience. The reason for critics arguing data science to be less of a science lies with the interpretation of the subject under a business framework. Mark Beyer, renowned analyst at Gartner, offers an insightful thought on this. He says that, “What business strategists need is ‘real’ data science. Real data science is the practice of building out competing interpretations of data, many multi-layered analytic theorems that intentionally challenge the inferences used by the others.” This approach is exactly what data science professionals need to focus on. They tend to rely more on the current market trends to scrutinise data. Although, this is not wrong but this diminishes the scientific context. Understanding Data Science Data science at times can be mystifying. It is because understanding algorithms may sometimes be beyond people’s immediate comprehension. At times, this uncertainty can lead to nowhere, which may dampen data science enthusiasts as well as their methods. On top of this, hardware and computing resources have progressed rapidly which makes data science tasks perform even better and discover newer techniques — useful or not. Another challenge is the vast amount of open-source information available in the subject. There is no dearth of online resources on data science, which may be overwhelming. This information overload can sometimes mislead the process of data science itself — on a scientific level. The scientific integrity may be lost in such cases. Conclusion Data science analyses hoards of data and brings out inferences in hundreds or thousand of ways. These inferences may not always be scientifically true or genuine. This is where it needs to bridge the gap. It needs to identify what inferences hold good and which ones don’t by justifying assumptions with facts. In fact, science is all about testing assumptions and constructing solid fact-based theories. The prospect of pseudoscience will gradually weaken if data scientists and researchers perpetuate richer knowledge on a consistent basis.","excerpt":"Data science has established itself as an important asset in the technology sector. Businesses are observing a significant rejig in their functions as most organisations are now dealing with surplus data. Data Science has also got fields like data analytics, artificial intelligence and machine learning which are growing parallely with it. Though data science has […]","categories":["IT Services"],"tags":["Data Science"],"author_name":"Abhishek Sharma","publish_date":"2018-05-18T11:36:05","publication_year":"2018","word_count":863,"keywords":["data science","Go","API","machine learning","artificial intelligence","AI","Aim","analytics","GAN","Data Science","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","data science","analytics","Aim","R","Go","API","GAN"],"url":"https:\/\/analyticsindiamag.com\/it-services\/is-data-science-a-pseudoscience\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10022525,"title":"Talking Pains And Promises Of GPT-3 With Sahar Mor","content":"“The main difference I see with GPT-3, is that unlike other AI applications that are narrow, its performance is general and human-like”Sahar Mor OpenAI’s GPT-3 is one of the largest language models with 175 billion parameters. It can perform a range of tasks, including text summarisation, search, coding, and essay writing. GPT-3 has been the most discussed topic among the ML community in the second half of 2020, and for right reasons. The language model drew more attention after the launch of OpenAI API. Developers across the world launched their own startups based on GPT-3. Few even managed to raise funds. Sahar Mor was one of the lucky ones to get access to this API. Sahar doesn’t have a university degree. Instead, he cut his teeth at the elite Israeli Intelligence Unit-8200. He has over a decade of experience in engineering and product management, both focused on products with AI at their core. He has also co-founded Homie, a leading platform in Israel aimed at people searching for long-term rental apartments, using language processing to index Facebook posts according to their content. Analytics India Magazine got in touch with Sahar to know more about his GPT-3 adventures, implications of large language models and more. AIM: Tell us about your AI journey Sahar: I relocated to Berlin in 2016 to join as a founding Product Manager to Zeitgold, a B2B AI accounting software that has raised >$60 million to date. At Zeitgold, I led the building of Zeitgold’s AI products and scaled its internal human-in-the-loop platform to support hundreds of business customers, automating ~75% of human operations. After Zeitgold, I have joined as a founding PM\/engineer to Levity.ai, a No-Code AutoML platform providing models for image, document, and text tasks. Last summer, I was one of the first engineers within the AI community to get access to OpenAI’s GPT-3 model. I have used this technology to build AirPaper, an automated document extraction API. Launched last September with OpenAI’s CTO retweeting about it, AirPaper’s waiting list has grown to >100, including startups, accounting firms, and insurance companies. In the last few years, I have been active in the AI community by writing and giving talks about the latest advancements in the field and exploring the different ways to transform the recent breakthroughs in AI research into production-ready AI products. OpenAI for parsing invoices into structured form: https:\/\/t.co\/IQ3sPjNg7y— Greg Brockman (@gdb) September 1, 2020 AIM: Tell us about your GPT-3 based application and how you got access to OpenAI’s API? Sahar: AirPaper is a robust document intelligence API. Send any document, either a PDF or an image, and get structured data. For getting access, I have emailed the OpenAI’s CTO with a short background about myself and the app I had in mind. OpenAI’s process for approving apps has led me to write about its scalability and shortcomings – along with potential ways to mitigate them. To get API access, one needs to apply via this form. The current waiting times can be forever, with developers that applied in late June and are still waiting for a response. Once you’ve built an app that is ready for production, you’ll be required to fill another form. It might take up to 7 business days for OpenAI’s team to review a request. After your app has been approved, you’re good to go. Every GPT-3 powered app starts within OpenAI’s Playground as you quickly iterate and validate if your problem can be solved with GPT-3. This tinkering is key in developing the needed intuition for crafting successful prompts. During this process, I realised there is an opportunity for OpenAI to automate and optimize this part, which they did several months later with their instruct-model series. Once I had the right prompt template in mind, I integrated it into my code. This meant preprocessing every document, turning its OCR into a GPT-3 digestible prompt, and querying the API. After further testing and parameters optimization (e.g. reducing temperature), I’ve deployed the app. Also Read: 15 Interesting Ways GPT-3 Has Been Put To Use AIM: What are the usual challenges you face while training large language models? Sahar: Lack of data that is relevant for the task at hand. As an example, there are many great open-source datasets for reviews and tweets, but none for document processing. That’s the reason there are still so many commercial companies building document intelligence APIs such as Google Document AI, AWS Textract, Instabase, etc.I’m building DocumNet, which is an ImageNet equivalent but for documents. I believe that, given enough data, document understanding can be commoditized in the same way computer vision has during the last decade. Inference and training costs are a challenge. Training and hosting your own language model can be quite costly. You can avoid these costs by using language models inference APIs such as OpenAI and HuggingFace, but those come with an extra premium.That said, I see these two costs dropping significantly during the next three years thanks to more efficient methods for training language models, a decreasing need for abundant data, and cloud providers reducing their prices. AIM: OpenAI has released DALL.E and CLIP recently. Do you think fusion models (vision + language) are the future of AI research? Sahar: Definitely. These days most AI applications in production are vertical and by now it’s a common understanding that narrow AI is not merely equivalent to human intelligence, even when conducting one specific task. For example, the SOTA deep learning model for early-stage detection of cancer (vision) is limited in its performance when it’s not combined with patient’s charts (text) from her electronic health records. I’ve seen this issue with many AI companies I’ve consulted over the years and also during my time at Zeitgold. For example, when a human operations agent was extracting an invoice amount, he was implicitly taking into account the handwritten correction next to the original amount, therefore extracting the handwritten one. He was only able to draw this conclusion as he was taking both the image and textual input into consideration. The main reason multimodal systems aren’t common in AI research is due to their shortcoming of picking up on biases in datasets. This can be solved with more data, which is becoming increasingly more available. Multimodal applications are not only relevant in the context of vision + language. During the last years, Facebook released several papers outlining novel approaches for automatic speech recognition (ASR) combining both audio and text. AIM: Should GPT-3 be regulated in the future? Sahar: Yes, but it’s tricky. The question of “should GPT-3 be regulated?” is a broader one involving other applications of AI. The main difference I see with GPT-3, is that unlike other AI applications that are narrow, its performance is general and human-like. This is the same concern we have with technologies such as deep fakes. Nevertheless, the fact OpenAI is regulating itself shows they acknowledge the harmful potential of its technology. And if that’s the case, can we trust a commercial company to self-regulate in the absence of an educated regulator? What happens once such a company faces a trade-off between ethics and revenues? The bottom line answer is yes, yet the main challenge is to understand in which ways and if at all regulation is an effective tool in ensuring safe AI adoption (I personally believe industry and research will beat the regulator by setting its own standards and recommendations. Which is again- dangerous). Human intelligence works in a multi-modal manner, where we utilize all of our senses when making decisions such as “what is in this picture?” or “is this a toxic comment?”. Furthermore, to make these decisions, we incorporate other elements such as our past experiences, which are de-facto the equivalent of transfer learning in ML. Not incorporating the two is confining whatever ML model you’re building to its (missing) data, and if the saying “your model is only as good as the data it was trained on” is a popular one, then how about “your model is only as good as the completeness of the data it was trained on?” Recommended Readings By Sahar Mor: Facebook AI’s Multitask & Multimodal Unified Transformer: A Step Toward General-Purpose Intelligent Agents Generative Pretraining from Pixels TAP: Text-Aware Pre-training for Text-VQA and Text-Caption12-in-1: Multi-Task Vision and Language Representation LearningMulti-skilled AI Find Sahar on Medium, Twitter and LinkedIn","excerpt":"“The main difference I see with GPT-3, is that unlike other AI applications that are narrow, its performance is general and human-like” Sahar Mor OpenAI’s GPT-3 is one of the largest language models with 175 billion parameters. It can perform a range of tasks, including text summarisation, search, coding, and essay writing. GPT-3 has been […]","categories":["AI Features"],"tags":["GPT-3","Interviews and Discussions"],"author_name":"Ram Sagar","publish_date":"2021-03-19T19:00:00","publication_year":"2021","word_count":1391,"keywords":["GPT-3","OpenAI","AI","AWS","ML","computer vision","document AI","Aim","deep learning","analytics","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","ML","deep learning","computer vision","analytics","OpenAI","Aim","document AI","AWS","R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/talking-pains-and-promises-of-gpt-3-with-sahar-mor\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":69398,"title":"The Case For Memorising Vs Looking Up Code On The Web","content":"In a perfect world, you would know all that there is to know about software, but that’s not the case in reality, where you are likely to find yourself stuck many times. For instance, you cannot possibly know 100% of the syntax or implementation of a specific language. Consider learning multiple languages in your career and you would realise that, at some point, you will need to rely on the experience of others to help you out. Programming is learning how the computer reads your codes and implements it. Therefore, programming is a skill that you can apply in any language. In this age, programmers are increasingly Googling code that they have to use in their work. But is that the right way to go about it? Let us find out. Why Look Up Code It’s perfectly normal to use Google, and in fact, can be a very crucial tool especially for beginners. The compiler\/interpreter will tell you when you are going wrong and will usually hint at how to fix it. According to most developers, a solid understanding of the foundations of your programming languages is more important. This is because if the documentation is not enough – especially open-source ones – it is almost not useful at all. In these cases, collaborating with others and looking for niche solutions will demand good Google search abilities. To know more about this, we interacted with Agnis Liukis, Lead software architect at Tieto and also Kaggle Grandmaster. “From my experience – I’m using StackOverflow to search for questions which are common and specific enough so that many people have already encountered exactly the same problem and have found a good solution. Like ‘regular expression for checking email addresses’ or which python code to calculate the area of a circle, etc,” Agnis said. According to Agnis, developers use Google or StackOverflow as much as you need to while keeping in mind the source, they only use it for complex problems and never type any code into your program unless you fully understand what it does. “Of course I could implement those myself. But I don’t think it is an effective way, because I could spend quite a lot of time re-implementing a solution, which is already implemented by someone, and is most likely already checked and tested by many other StackOverflow users. Checked and tested by many users is a very important point here to avoid bugs,” Agnis explained. What About Memorising Code Programming languages are a bit like spoken languages. If you had to look everything up, you won’t be able to have a conversation. So for useful bits of code, it is helpful to memorise them. It would be best if you remembered the basic syntax, logic structures,  and core functions in the language you use, as well as a couple of methods in the most used libraries, along with basic algorithms that you end up frequently using, etc. If you are continually Googling basic syntax questions, you should think about practising problems that will help you get those skills into muscle memory. That way, you will not need to Google basic syntax questions about working with Strings, Arrays, or Loops, and only look up things you don’t know about. According to experts, the important thing is learning and (even memorising after understanding) concepts, data structures, and algorithms. What is more, learning individual language\/frameworks also allows you to work in different environments and accomplish tasks more efficiently.  Modern code editors can help a lot in generating some starting\/template code and showing names of available functions, this way significantly reducing the amount of information needed to memorize. But still the basic concepts and structures will have to be remembered. I would say that’s the very minimum to remember to be able to write a program. Agnis Liukis said, “First, there are some basic things, which a programmer must know\/learn\/memorize regarding each programming language to use it effectively – like some basic concepts, structures, keywords. Because it would be quite slow to write a program, searching on StackOverflow questions like ‘How to write loops in Python’ or ‘How to define a variable in Python’ every time.” The more you code, the more you memorise, and thereby, google less. The concepts are most valuable, but one can question your understanding of the idea if you cannot even memorise the basic syntax. Also, you can run into issues as a programmer if every few days you’re searching for syntax details rather than finishing your tasks. Learning a programming language is a matter of practice and habit, and so is program design, debugging skills, and many other things programmers need. “If you want to be a good programmer. Depending on programming language, there are some known “good practices”, how to write code – these also would be necessary to memorize. A programmer might even not remember exact syntax for specific language (and use documentation\/StackOverflow to find it), but he must remember that there ARE some best practices and special syntax to use for a given specific case.  Because things can be written in several ways and there are cases when one way is better than the other,” according to Agnis Liukis. Such memorisation not only increases your speed in more complex tasks, but also promotes a more thorough understanding of the basic principles. Final Thoughts: Googling Vs Memorising Code Googling or searching for answers on StackOverflow isn’t cheating, and it is like using a resource. We no longer live in an age where you are expected to memorise everything because it exists on the internet. Instead, it may be safe to say that those who are better at finding solutions will do better than those who have a good memory. Coding is like practising law in that it is more about knowing the right avenue to take to find the answer, rather than memorising all the answers. Take the example of lawyers – they have tons of reference materials, but they still have the foundation to know where to look and what to look for. Even at an advanced stage of a career, you may still refer to API documentation and search on the web for methods whose syntax\/arguments you don’t’ remember fully. Being a good programmer doesn’t mean knowing every nook and crevice of every API or library; it means being able to complete the task at hand effectively and efficiently. As you become well-versed with a programming language, you will stop googling such broad issues, and begin to google things like structures and internals. Becoming proficient in a programming language is having the core of it learned and memorised so that you can apply it and think about problems in it as quickly as an everyday spoken language. But like a spoken language, there is still a requirement for a dictionary and thesaurus from time to time. But the most important thing is learning how to learn. Any person can look up a code snippet and script, paste it into something to finish a task. Understand what a piece of code is what you’re using, and that is where the real learning comes in. Memorising the syntax and various modules that help you write code faster is good, but it’s okay to look up things if you can’t remember how to do something. The most crucial thing to learn about programming is conceptual anyway. But, you may still find yourself Googling obscure language features, which may be true for pretty much all languages you will learn. Instead of trying to find a forum post from another beginner, parsing your results for source code and development mailing lists is what makes the difference.  The goal is to try and use web search less and only for more and more obscure and difficult things. Eventually, you will stop Googling such broad issues, and begin to look for things like structures and internals. The better you get at coding, the more specific your search becomes.","excerpt":"In a perfect world, you would know all that there is to know about software, but that’s not the case in reality, where you are likely to find yourself stuck many times. For instance, you cannot possibly know 100% of the syntax or implementation of a specific language. Consider learning multiple languages in your career […]","categories":["Deep Tech"],"tags":["VS Code"],"author_name":"Vishal Chawla","publish_date":"2020-07-10T15:00:00","publication_year":"2020","word_count":1329,"keywords":["Go","API","programming_languages:R","AI","programming_languages:Go","Python","Ray","VS Code","programming_languages:Python","R"],"extracted_tech_keywords":["AI","Ray","Python","R","Go","API","programming_languages:Python","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/the-case-for-memorising-vs-looking-up-the-code-on-the-web\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":46247,"title":"Inside The Rampant Threat Of Smominru Botnet","content":"The way strategies — both defensive and offensive — are evolving in cybersecurity is incredible. There was a time when it was all about taking systems down with attacks like DDoS, phishing, malware; then ransomware came into the picture; and today, the strategies to pwn computers have reached a whole new level. Threat actors have started leveraging ways that not only hack systems but also make use of those systems to mine cryptocurrencies. Smominru botnet is one such computer threat that has gained tremendous traction and has been creating a lot of hustle and bustle in the cyber world. According to a report, this notorious botnet is now infecting over 90,000 machines each month all around the world. What is Smominru? The history of Smominru botnet dates back to 2017 and it has different variants like as Hexmen and Mykings. It is basically a crypto-mining botnet that is also equipped with worming capabilities. Since the get-go, this notorious botnet has gained some significant traction. In August 2019, it came under the light that Smominru makes the best use of some of the propagation methods such as the EternalBlue exploit (which was also used NotPetya and WannaCry), brute-force and attacks that gather credentials. Datacenter and cloud security company, Guardicore has been tracking this botnet since the very beginning and recently, the company has released a report that it has managed to get into one of the core serves of the Smominru. The company has also stated that they have over time they have studied and monitored the infection patterns of the botnet and gathered information about the compromised machines and networks and has also assessed the botnet’s impact. Once the Smominru botnet successfully lands in a system, it takes one of the most notorious steps by downloading a Powershell script — blueps.txt. The script then creates a new administrative user named admin$ on the system and downloads additional scripts to perform malicious actions such as stealing victim credentials, installs a Trojan module and a crypto miner and propagates inside the network. That is not all, it also creates several other backdoors on the compromised machine in different phases of the attack. The Smominru Impact Talking about the impact of this infamous botnet, in August 2019, Smominru compromised 4,700 machines per day, and with such a rate, over 90,000 systems have been infected all around the globe and China, Taiwan, Russia, Brazil and the US are some of the nations with a high infection rate. Furthermore, while a lot of sectors have been the victim of Smominru, there are some specific sectors that have witnessed most of the attacks — higher-education institutions, medical firms and to the surprise, even cybersecurity companies have fallen prey to this botnet. While these all inform do prove that the impact of this botnet is brutal enough, there is one more feature that makes it one of the highly effective botnets for malicious activities. According to a report, the latest version of Smominru also makes sure that other threat actors are not interfering. Therefore, the botnet eliminates other infection or malicious elements from the system it targets. Furthermore, it also blocks TCP ports (SMB, RPC) in order to prevent threat actors from compromising the already-compromised system. Outlook Patching systems is definitely one of the most important things for every organisation. While many companies take this is serious, there are companies that ignore the fact that patching your systems can deliver significant protection against cyber threats. Smominru botnet attack was also aided by unpatched systems. The botnet solely focuses on systems that are not being patched since quite a long time and that is the reason why machines running on Windows 7 and Windows 8 server became the major victims of this attack. Further, Smominru is also capable of making things worse ever thought you remove it from the infected server. Over the years, several malware and botnets have shown that the hacking world is witnessing a significant evolution and one single mistake from a company and these threats would take things down. Technology has not only empowered companies and individuals but also empowered threat actors to come with strategies and methods that are highly brutal. It is high time that organisations start taking a look at their cybersecurity infrastructure one more time and make sure that there are no loopholes that would help any threat actor to pwn.","excerpt":"The way strategies — both defensive and offensive — are evolving in cybersecurity is incredible. There was a time when it was all about taking systems down with attacks like DDoS, phishing, malware; then ransomware came into the picture; and today, the strategies to pwn computers have reached a whole new level. Threat actors have […]","categories":["AI Features"],"tags":["Bitcoin"],"author_name":"Harshajit Sarmah","publish_date":"2019-09-23T15:47:39","publication_year":"2019","word_count":731,"keywords":["Go","Bitcoin","programming_languages:R","AI","programming_languages:Go","RAG","ViT","GAN","R"],"extracted_tech_keywords":["AI","RAG","R","Go","GAN","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/inside-the-rampant-threat-of-smominru-botnet\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10087162,"title":"The Search Engine Showdown is Far from Over","content":"Back in the 1990s, the search engine category was a hot space. Yahoo, Netscape, AOL, Ask Jeeves, AltaVista, Google search, MSN and others were vying to capture the dominant position. With time, they all fizzled out. Post 2000 was the era of Google Search, the undisputed winner of the space until quite recently. The tide is turning and the crown of Google Search is under threat. Microsoft, who has been trying its luck in the search category for about 25 years now, has drawn the sword and thrown away the scabbard to fight the unfinished battle. Microsoft’s Unfulfilled Dream Earlier this week, when Microsoft held a surprise event, it was all centred around Bing. It was short and crisp with no mistakes and introduced an AI-powered Bing search and Edge web browser. It is built on a new discrete LLM model, which Microsoft claimed to be “much stronger” than ChatGPT and GPT-3.5 combined. Currently, Microsoft’s Bing is a distant second in the search engine category with a market share of only 9%. Bing had about 5.6%, or $11.59 billion, of Microsoft’s total revenue of $204 billion in 2022. However, Microsoft—with its latest breakthrough in AI—has positioned itself as a powerful competitor to Google’s dominant hold on the search engine market. Just as Nadella said, “It’s a new day for search—in fact a race starts today in terms of what you can expect. And we’re going to move, we’re going to move fast.” Microsoft has its eye on the engine throne and is finally ready to disrupt Google’s reign. The ‘Windows’ maker is aware of the fact that distribution is key to success, unlike Google who claims to focus on the quality of its search results. Google asserts that its internal search quality metrics outperforms Bing. But if Google was truly confident about its search’s quality, it would not pay Apple a staggering $15 billion annually to be the default search engine on IOS devices—a sum that far exceeds Bing’s entire revenue. On the other hand, Microsoft is strategically playing its cards with billion-dollar partnerships, especially the one with OpenAI. Even computer vision genius Andrej Karpathy, who Microsoft chief touts as a poster child amidst back and forth applauds, is now on board to fuel Nadella’s dreams. Also read: Three’s a Crowd: ChatGPT vs Bard vs Ernie Bot Bing’s Distribution Strategy Accessing the new beta of Bing requires users to download Microsoft Edge and set Bing as their default search engine to evade the waitlist. Bing knows the importance of easy user accessibility, as the waitlist for the new beta version can cause potential users to lose interest. Recognising the urgent need for a more sustainable monetisation method, Microsoft is focussed on growing the number of Bing users, rather than relying solely on subscription fees. Additionally, Bing is an important element of Microsoft’s massive $10 billion advertising empire. It has recently entered into a strategic agreement with OTT platform Netflix to provide the tech and sales for its new ad-supported tier. The revamped search engine is an important part of Microsoft’s ambitious plan to double its ad income from $10 billion to $20 billion annually, as it was announced last year. This collaboration, combined with Bing’s already established presence in the market, can only work towards its favour. Google’s Effort to Save the Share At Google’s recent ‘Live in Paris’ event, it revealed a slew of updates like the addition of AI-powered chatbot ‘Bard’ to its search. Other announcements were updates such as Google Lens will soon be able to search for objects in photos and videos across websites and apps, over 24 new languages have been added to Translate, and Maps can now give a more ‘Immersive View’ with AR-based ‘Search with Live View’ and ‘Indoor Live View’. The company is also working on AI-powered art and culture tools and plans to hire more developers to advance Generative AI. Google Seems Nervous Unfortunately, the latest event from the tech giant was plagued with mishaps, especially with the biggest blunder being Bard’s incorrect response to questions about new discoveries from the James Webb Space Telescope, further tarnishing the company’s reputation and leaving a blemish on what was supposed to be a celebration. From an incorrect announcement of Google Search’s anniversary by VP Prabhakar Raghavan at the very beginning to a misplaced phone leading to no demo of Google Lens, the presenters seemed underprepared and their speeches were vaguely scripted. Despite launching upgraded AR features, the overall impression was that of a lacklustre presentation which failed to live up to the hype around it. The nightmare of an event turned to be a costly setback, with the company losing over $100 billion, as stocks nosedived by nearly 8% right after the event. Birth of AI-Backed Search Engines But Microsoft is not the first to incorporate AI in search engines. Neither is Google Search. The title of launching the first AI-advanced web engine goes to former Salesforce CEO Marc Beniof for You.com. The privacy-focused search engine provides ad-free personalised summaries to its users. It uses Microsoft Bing for more specific searches and Google for more general results. In addition, You.com has released ‘YouChat’, a GPT-3.5-based chatbot. As the AI industry continues to expand, we also see the emergence of exciting new players such as Perplexity AI and Neeva AI joining the ranks. They bring impressive potential to the table, worthy of challenging the big players of the industry. The Road Ahead The next few years will be a testament to the power of the AI-driven search engines as the stage is set for a showdown where the big techs will unfold their latest developments. It will be interesting to see whether the future of technology will be shaped by the giant players alone or will emerging companies like You.com be able to make their mark. Also read: Data Science Hiring Process at McAfee","excerpt":"The birth of AI-powered search engines has just begun, and it won’t be long until the next player makes its move in this exciting race.","categories":["AI Features"],"tags":["Andrej Karpathy","Bing","Google","Microsoft","neeva ai","OpenAI","Perplexity AI","Sam Altman","Satya Nadella","you.com"],"author_name":"Shritama Saha","publish_date":"2023-02-12T10:00:00","publication_year":"2023","word_count":980,"keywords":["Bing","computer vision","R","Satya Nadella","neeva ai","data science","ChatGPT","Sam Altman","Andrej Karpathy","RAG","Go","AI","generative AI","you.com","OpenAI","Perplexity AI","Aim","Google","Microsoft"],"extracted_tech_keywords":["AI","computer vision","data science","generative AI","ChatGPT","OpenAI","Aim","RAG","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/the-search-engine-showdown-is-far-from-over\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10135542,"title":"Why Indian Startups Don’t Invest in R&amp;D","content":"In India, the number of VCs and angel investors that are getting interested in AI startups is skyrocketing. It is often believed that most of them don’t even have a thesis while investing in these startups, but the same goes for several startups that are innovating, or at least hoping to, in the space. Most of the Indian AI startups are powered by Jugaad, not VC money, which makes it harder for them to put those funds into R&D. Instead, they focus on the North Star Metric, often called the valuation, and bringing profitability earlier for their investors, giving them an exit. But this hampers what Indian startups can achieve if given enough money and time to focus on R&D. At the same time, VCs are hesitant to invest in startups that are focusing too much on R&D, as it would take a lot of time for the company to even reach a hint of profitability. Prayank Swaroop, partner at Accel, said that VCs are increasingly expecting AI startups to demonstrate rapid revenue growth. “Even to the pre-seed companies, we say, ‘Hey, you need to show whatever money you have before your next fundraiser. You need to start showing proof that customers are using you.’ Because so many other AI companies exist,” said Swaroop. The Growth Needs to Be Rapid AI startups require extensive patient capital from investors as they require extended periods of R&D before their innovations can be brought to market. Arjun Rao, partner at Speciale Invest, also told AIM that it is essential to be patient when investing in AI as these startups are still in the budding stage, and placing the bet on the right founders is paramount. “Founders do not need to worry about the current downturn and keep a long-term mindset,” said Rao. Meanwhile, most of the startups never see the light of profitability. Even OpenAI, the most invested in and the AI startup with the most users of its product, is still nowhere close to profitability. Maybe, investors need to relook at their investment strategy. There are two broader types of investors – one that is ready to invest in OpenAI wrapper startups and the ones that want to invest in something fundamentally different. And the latter ones are very scarce. VCs are increasingly not ready to put money in deep-tech startups in India as the need for building use cases is getting more interesting for them, or what some people call OpenAI wrappers and so-called consumer-tech startups, as they require less capital, and possibly quicker revenue. India’s generative AI scene is on an upswing, but the investors are cautious when it comes to investing in research startups. Though it is rare to find any other initiative being built from scratch, the lack of VCs’ interest in such initiatives also kills R&D. They just simply don’t want to take the risks. What’s Wrong with the Startups? VCs in India tend to favour proven business models, often pulling out when companies don’t innovate fast enough. VCs withdrew from Byju’s because the company wasn’t “experimenting enough.” The question should be, why did they invest in Byju’s in the first place, despite its shady practices? The answer: capital was cheap, and it was a tested model. Turns out, even that was a flop. Despite the pros and cons of the investors, Indian startup founders are also not the ones to not be blamed. India has a surplus of STEM graduates who just want to work for big companies and are too indecisive if they can do a startup in India. Many of them often go to Silicon Valley to build research startups, instead of building in India. Another sentiment echoed is that most Indian startups are driven by “MBA grads, product managers, and marketers,” rather than engineers. “It’s engineers who need to run companies,” said a user discussing on Reddit, emphasising that successful Western startups are often engineer-led, with a focus on innovation. Current business culture in India is overly focused on replicating successful models rather than developing new ones. No Risk Needed in India We are a capital-short nation. When capital comes with such a high premium, businesses are understandably hesitant to take risks, especially when labour is relatively cheap. Why would anyone pour limited funds into uncertain R&D when they can invest in low-margin, proven ventures? This risk-averse mindset is compounded by India’s corporate culture, which, as many, tends to prioritise short-term gains over long-term investments in innovation. One exception to this trend is India’s Unified Payments Interface (UPI), which has been widely hailed as a groundbreaking innovation in digital payments. However, UPI succeeded because of government support. Without a similar level of institutional backing, other innovations might struggle to gain traction, which is a valid reason for startups not to do much R&D. In countries like the US, tax incentives encourage companies to invest in R&D. For example, Amazon benefits from marking even minor improvements under R&D to claim tax benefits. In contrast, India’s regulatory framework lacks similar incentives, further discouraging investment in innovation. This short-term thinking extends to most Indian startups, which prioritise stability over innovation. As another Redditor sums it up, “R&D requires time, the results are uncertain, and it needs good investment. Indian corporate culture is very much short-term, instant-gratification based.” High capital costs, societal attitudes that discourage risk-taking, and a corporate culture focused on immediate returns, as long as these conditions persist, R&D will remain a low priority for most startups. However, with greater government support and a shift in cultural mindset, India could unlock its potential for true innovation.","excerpt":"R&D requires time, the results are uncertain, and it needs good investment.","categories":["IT Services"],"tags":["AI Startups","Startups"],"author_name":"Mohit Pandey","publish_date":"2024-09-16T16:39:07","publication_year":"2024","word_count":931,"keywords":["Go","API","OpenAI","AI","Git","RAG","Ray","Aim","generative AI","Startups","R","AI Startups"],"extracted_tech_keywords":["AI","generative AI","OpenAI","Aim","Ray","RAG","R","Go","Git","API"],"url":"https:\/\/analyticsindiamag.com\/it-services\/why-indian-startups-dont-invest-in-rd\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10041174,"title":"Does Deep Learning Suffer From Too Many Optimizers?","content":"“There is no single optimizer that dominates its competitors across all tasks.” Critics often call machine learning ‘glorified statistics’. There is some merit to the argument. The fundamental function of any machine learning model is pattern recognition, which relies on the principles of convergence; the methods of fitting data to the model. To that end, neural networks use optimization methods, typically categorised as first-order, high-order and derivative-free. First-order optimization such as gradient descent and its variants are quite popular. Gradient descent method refers to the idea of updating the variables iteratively in the (opposite) direction of the gradients of the objective function. After each update, the gradient descent method (think mathematical expression) guides the model towards the target gradually, converging to the optimal value of the objective function. The stochastic gradient method is an unbiased estimate of the real gradient. It reduces the update time while dealing with large numbers of samples and removes a certain amount of computational redundancy. Then there are other variants that claim to do a better job. Large-scale stochastic optimization drives a wide variety of machine learning tasks. The permutations and combinations of this whole ordeal quickly run out of hands when one stumbles upon scores of benevolent sounding optimizers. Fatigue by abundance is now a serious challenge among researchers. Choosing the right optimization method can be a nightmare. Not to forget the effective tuning of hyperparameters that heavily influences the training speed and final performance of the learned model. These tasks are time and resource-intensive. Source: Paper by Schmidt et al. Choosing the optimizer is one of the most crucial design decisions in deep learning, and it is not an easy one. The above illustration shows the number of times ArXiv titles and abstracts mention specific optimizers per year. The growing literature now lists hundreds of optimisation methods. The researchers at the University of Tübingen performed an extensive, standardized benchmark of fifteen popular deep learning optimizers. This paper is one of the few works that focus on large scale benchmarking of the optimizers. According to the researchers, the objective here is to help understand how optimization methods and hyperparameters influence the training performance. “While some optimizers are frequently decent, they also generally perform similarly, often switching their positions in the ranking.”Paper by Schmidt et al. Aiming for generality, the researchers evaluated the performance on eight diverse real-world deep learning problems from different disciplines. From a collection of more than a hundred deep learning optimizers, the researchers selected fifteen of the most popular choices for benchmarking. “There are enough optimizers,” said the researchers. The authors also noted that the conclusions of this paper might not generalize to other workloads such as GANs, reinforcement learning, or applications where e.g. memory usage is crucial. The researchers analysed more than 50,000 individual runs and have open-sourced all the baseline results of their experiments. This seminal work underlines the dangers of chasing state-of-the-art hype and highlights the following: There are now enough optimizers.Optimizer performance varies greatly across tasks. There is no single optimizer that dominates its competitors across all tasks.ADAM and ADABOUND consistently perform well.Different optimizers exhibit a surprisingly similar performance distribution compared to a single method re-tuned or simply re-run with different random seeds. Having an accurate baseline for optimizers can drastically reduce the amount of computational budget. Given these results, the researchers question the rationale behind development of new methods when there are more fundamental problems at hand. The researchers hope their experiments will nudge the ML community to “move beyond inventing yet another optimizer and to focus on key challenges, such as automatic, inner-loop tuning for truly robust and efficient optimization.” The researchers also admitted the creators of new optimizers cannot be expected to compare their work with every possible previous method. The baselines of all the experiments have been open-sourced, and the ML community can access the data set that contains 53,760 unique runs, each consisting of thousands of individual data points, such as the mini-batch training losses of every iteration or epoch-wise performance measures, which can be used as competitive and well-tuned baselines for future benchmarks of new optimizers. Know more here.","excerpt":"“There is no single optimizer that dominates its competitors across all tasks.” Critics often call machine learning ‘glorified statistics’. There is some merit to the argument. The fundamental function of any machine learning model is pattern recognition, which relies on the principles of convergence; the methods of fitting data to the model. To that end, […]","categories":["Deep Tech"],"tags":["ADAM","gradient descent","SGD"],"author_name":"Ram Sagar","publish_date":"2021-06-03T11:00:00","publication_year":"2021","word_count":688,"keywords":["Go","machine learning","AI","neural network","RPA","ML","ADAM","Aim","deep learning","GAN","SGD","R","gradient descent"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","neural network","Aim","R","Go","GAN","RPA"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/does-deep-learning-suffer-from-too-many-optimizers\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10163910,"title":"Zomato’s Nugget is Hiring GenAI Engineers","content":"Gurugram-based food-tech giant Zomato is hiring for its new venture, Nugget, an AI-powered customer support platform, to build a strong team across engineering, design, and sales. The company is actively recruiting GenAI, backend, and frontend engineers, designers, and B2B sales professionals to drive innovation and expansion. Nugget is focused on creating world-class products that tackle real-world challenges, and Zomato is seeking creative problem-solvers and ambitious professionals who thrive in a fast-paced environment. The roles are full-time and on-site at Zomato’s office in Gurugram. The company is inviting interested candidates to submit their resumes via email to nugget-hiring@zomato.com or connect directly. Recently, Zomato CEO Deepinder Goyal announced the launch of Nugget, a no-code customer support platform designed to help businesses scale their support operations efficiently. Developed over three years as an internal tool, Nugget now handles over 15 million support interactions per month for Zomato, Blinkit, and Hyperpure. The platform, which is now open to businesses worldwide, boasts a 90% adoption rate among companies that have tested it. Other Zomato Hirings Recently, Goyal’s job posting ignited a hiring debate. He was seeking business and product leaders who have embraced AI as their “second brain”. “I am looking to work with business and product leaders who have already started using AI as their second brain. If you are the one, please write to me at d@zomato.com. Please include the phrase ‘I have a second brain’ in the subject line,” Goyal had announced on X. This unusual hiring criterion has stirred conversations about AI’s growing role in leadership and decision-making. This comes just months after Goyal’s previous unconventional job posting – a chief of staff position that required candidates to pay a ₹20 lakh ($23,700) “fee” instead of receiving a salary.","excerpt":"The company is inviting interested candidates to submit their resumes via email or connect directly.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","AI career","AI hiring","Zomato"],"author_name":"Vidyashree Srinivas","publish_date":"2025-02-18T10:27:25","publication_year":"2025","word_count":288,"keywords":["Go","AI career","GenAI","programming_languages:R","AI","innovation","programming_languages:Go","ViT","AI hiring","Zomato","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","GenAI","R","Go","ViT","innovation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/zomatos-nugget-is-hiring-genai-engineers\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10024322,"title":"What Happened When Google Threw All Voice Data To The Blender. Answer: SpeechStew","content":"Training large models is a massive challenge as it requires collecting and annotating vast amounts of data. It is particularly challenging in the case of speech recognition models. To overcome this challenge, a team from Google Research and Google Brain have introduced an AI model, SpeechStew. The model is trained on a combination of datasets to achieve state-of-the-art results on various speech recognition benchmarks. SpeechStew The success of end-to-end speech recognition models has been directly linked to the abundance of training data and the use of large deep learning models. However, in noisy and low-resource datasets such as CHiME-6, these end-to-end methods struggle to achieve optimal results. Techniques such as multilingual training, multi-domain training, unsupervised pre-training, semi-supervised learning, and transfer learning are recommended to avoid overfitting and promote more generalisation. The SpeechStew model applies multi-domain training and transfer learning to end-to-end speech recognition. This method does not introduce additional hyperparameter or external language model during inference. Broadly, it follows two main steps: Combining all the available speech recognition without domain-dependent rebalancing or reweighing.Training a single large neural network model containing 100 million to up to a billion parameters. Here, the researchers combined all the available speech recognition data, both labelled and unlabelled, amounting to a total of 5,00o hours. The combined dataset included AMI dataset (with 100 hours of meeting recordings), Switchboard (2,000 hours of telephone calls), broadcast news (50 hours of television news), Common Voice crowdsourced from Mozilla, TED-Lium (450 hours of TED talks), and Librispeech (960 hours of audiobooks). The SpeechStew model was then tested on several benchmarks, and it outperformed the previous models. The scientists also observed that with this technique, the team could perform more challenging tasks. To test the transfer learning capabilities, the team fine-tuned the SpeechStew model to test on the Chime-6 dataset containing 40-hours of distant conversations recorded on microphones and achieved good accuracy. Performance SpeechStew could achieve state-of-the-art or near state-of-the-art results across tasks (AMI, Common Voice, TED-Lium, WSJ, and Switchboard). The model also demonstrated strong transfer learning capabilities. The results for transfer learning were encouraging since Chime-6 is a particularly challenging task for an end-to-end speech recognition model, which generally suffers from overfitting issues. Training large models are expensive and impractical to do frequently. The study proved a user could fine-tune a pretrained model on only a few thousand gradient steps to achieve good performance. The cost incurred was also low. “SpeechStew learns powerful transfer learning representations. We fine-tune SpeechStew on a noisy low resource speech dataset, CHiME-6. We achieve 38.9% WER without a language model, which compares to 38.6% WER to a strong HMM baseline with a language model,” the authors said. Wrapping Up The technique of mixing datasets to train neural networks is not new. However, this work is different from the previous models as the team could scale to much larger models. SpeechStew leverages up to 1 billion parameter models to yield strong empirical results. However, it does not perform well for extensive models such as GPT-3, which contains 175 billion parameters. “This simple technique of fine-tuning a general-purpose model to new downstream speech recognition tasks is simple, practical, yet shockingly effective. It is important to realise that the distribution of other sources of data does not perfectly match the dataset of interest. But as long as there is some common representation needed to solve both tasks, we can hope to achieve improved results by combining both datasets,” the team said in an interview. Read the full paper here.","excerpt":"SpeechStew model is trained on a combination of datasets to achieve state-of-the-art results on various speech recognition benchmarks.","categories":["AI Trends"],"tags":["Google Brain","Google Research","how to retrain data","Speech Analytics","Speech Recognition","speech recognition algorithm","Transfer Learning"],"author_name":"Shraddha Goled","publish_date":"2021-04-20T16:00:00","publication_year":"2021","word_count":581,"keywords":["Go","Speech Analytics","AI","how to retrain data","neural network","RPA","Google Brain","speech recognition algorithm","programming_languages:R","RAG","GPT","deep learning","llm_models:GPT","Speech Recognition","Transfer Learning","R","Google Research"],"extracted_tech_keywords":["AI","deep learning","neural network","RAG","R","Go","GPT","RPA","llm_models:GPT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/what-happened-when-google-threw-all-voice-data-to-the-blender-answer-speechstew\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10001415,"title":"Emergence Of The Cloud-Based Contact Center-As-A-Service Industry In India","content":"With the rise of the ever-demanding marketplace of multi-generational customers, today, cloud-based platforms have transformed significantly. And there is no denying that these platforms are becoming compelling for many companies across the world. Cloud-based contact centre as a service (CCaaS) is one such technology that has gained a lot of traction in recent time. CCaaS is a call centre software that is hosted in the cloud instead of hosted on-premises, offering advantages such as flexibility and scalability, improving the overall customer experience, as well as being an economically smart choice. Talking about the cloud industry in India, it has crossed a barrier in the nation — from being a new untested technology, it is now one of the preferred choices of deployment for enterprises. Headquartered in Hyderabad and Singapore, Ozonetel is a 10-year-old, bootstrapped startup that has several industry firsts to its credit in the area of Cloud Communication. With a platform, KOOKOO that serves companies like FoodPanda, Bigbasket, Shaddi.com etc., Ozonetel today, creates and deploys cloud solutions globally. The company has over 1,000 enterprise clients across US, India, and UAE, and its platform supports over 50,000 live agents and has handled over 3 billion calls. “When we started Ozonetel, the goal was to enable a platform for developers to easily build telephony solutions,” said Chaitanya Chokareddy, Chief Innovation Officer, Ozonetel. “Our next avatar was to provide easy to use products like cloud contact centre and cloud PBX built on top of our platform. Currently, our focus is on using AI\/ML to improve communication experiences in our products.” Marking Its Territory In The CCaaS Space Ozonetel’s proprietary product KOOKOO is India’s first and only cloud telephony platform, allowing developers to easily build telephony applications. Its applications can range from missed call apps, number masking etc. to complex apps such as call centres and PBX solutions. The power of KOOKOO comes from Ozonetel’s homegrown stack. At the bottom layer, it has its PRI cards and next, it has the telephony stack which handles the telephony events. Above all of that, it is KOOKOO which is an abstraction for developers. “So, the developers only interact with KOOKOO without having to worry about arcane telephony protocols,” said Chaitanya. Talking about the other technology side of Ozonetel, the company has C programming at the lowest layers with the telecom stack having C and Perl. When it comes to the application layer, Ozonetel uses Python and PHP. The cloud contact centre solution of Ozonetel is Java-based with all its servers running Linux. Also, the database needs are fulfilled by MySQL. Today, 58% of Ozonetel’s revenue comes from KOOKOO CloudAgent, offering an entire gamut of multi-channel (voice, SMS, email, chat, etc.) functionalities and analytics. Furthermore, Ozonetel is also launching a Large Enterprise version of KOOKOO CloudAgent. The company believes that with this software,  an enterprise can instantly have over 1000+ cloud-based Contact Center Seats. And as of now, there is no other software in the market that supports over 100 seats on the Cloud. That is not all, the company has recently launched one more product, KOOKOO Interactive Assistant, which is a Unique Omnichannel Widget that combines Voice and Chat. This widget works with KOOKOO CloudAgent software and enables Contact Center agents to run chat sessions and voice calls parallelly with the customers who are calling in. What Makes Ozonetel Better Than Its Competitors When asked about Ozonetel’s approach towards gaining customer trust, Chaitanya told, “we have fanatical regard for customers and that is now built into the DNA of the company. Customer happiness is paramount and everything else comes later.” At Ozonetel, there is something called ‘engineering driven customer satisfaction’ and it is about solving customer problems by engineering the right solution. The company believes that customers just want the product to work and solve their problem and, in most cases, Ozonetel finds an engineering solution for that. Talking about the competitions in the market, Chaitanya said, “on the platform side, we do not have any competition in India.” However, Twilio is a competition in the US, Singapore and other geographies. When it comes to cloud contact centre side, Avaya, Genesys, and Dhrishti are some of the competitors of Ozonetel in India. When asked about having an edge over its competitors, Chaitanya said, “Our focus on product is what makes our company different and better. Our product portfolio is the best in the cloud contact centre market and our experience in providing solutions across 20 verticals. Also, in our new avatar, we hope our AI engine will provide us with more clear advantages over our competitors. Roadmap Ahead Looking into the future, Chaitanya revealed that the future of Ozonetel is AI and new communication tools that the coming is coming out with. Also, Ozonetel is planning on disrupting the cloud communication space with is KOOKOO Interactive Assistant widget and the bot platform powered by its contact centre AI. “The current contact centre models are outdated, and we at Ozonetel hope to disrupt the space with our new product offerings in the coming year,” Chaitanya concluded.","excerpt":"With the rise of the ever-demanding marketplace of multi-generational customers, today, cloud-based platforms have transformed significantly. And there is no denying that these platforms are becoming compelling for many companies across the world. Cloud-based contact centre as a service (CCaaS) is one such technology that has gained a lot of traction in recent time. CCaaS […]","categories":["AI News"],"tags":["Cloud Computing","sql v mysql"],"author_name":"Harshajit Sarmah","publish_date":"2019-03-11T18:48:03","publication_year":"2019","word_count":840,"keywords":["Go","AI","ML","Scala","sql v mysql","Python","Cloud Computing","analytics","SQL","Rust","R","Java"],"extracted_tech_keywords":["AI","ML","analytics","Python","R","SQL","Go","Rust","Java","Scala"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/emergence-of-the-cloud-based-contact-center-as-a-service-industry-in-india\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10054245,"title":"Experimental Robots That Might Be Around Us In 2022","content":"The fictional idea of ‘robot revolution’ seems to be true now with AI robots being visible around us, performing small tasks. In the last few decades, robots have grown from being specialised devices for industrial applications to even helpers in household chores. The recently deployed Google’s robots that clean tables are a great example of letting ML develop on its own. According to a report by Markets and Markets, the AI-Robots market size reached a whopping USD 6.9 billion in 2021 and is expected to multiply by around five times in the next five years and reach USD 35.3 billion by 2026 (CAGR of 38.6 %). The demand for AI robots especially got a boost during the pandemic where social distancing became a norm. Here are a few autonomous robots that are currently deployed at the experimental level and have the potential to become a norm in 2022. Read on: Everyday Robots Alphabet has deployed 100 everyday robots across Google offices in North Carolina. Created by X – The Moonshot Factory, the robots use a combination of ML techniques, which include reinforcement learning, collaborative learning, and sensors to perform tasks. The robots can learn by themselves and even adapt to the environment. They can adapt to the unprecedented factors around them and act accordingly. Everyday robots perceive their surroundings using LiDAR sensors. The robots can learn complex tasks by themselves in a day – like opening a door, which can take about four months of programming. The white robots have an arm called the gripper above its wheeled base, which is a multipurpose attachment. These everyday robots are currently sorting trash, cleaning tables, moving chairs in and out the conference rooms, and even opening doors, among other tasks. Teladoc Robot Meenakshi Mission Hospital & Research Centre (MMHRC), Madurai, Tamil Nadu, deployed 16 Teladoc Health Vita robots in 2021 during the pandemic. Acquired from US-based Teladoc Health, the robots autonomously carried out basic clinical examinations like testing of heart rate and blood pressure. The capabilities of these robots extend to being IoT devices too, where they were used to control advanced diagnostic equipment like CT and MRI scan machines. The Teladoc robots collected and processed data from multiple diagnostic equipment, analysed it and helped doctors make precise clinical decisions. MMHRC integrated the Teladoc robots to cath labs, ICUs, emergency, isolation wards and picture archiving & communication system (PACS). This installation helped the Tamil Nadu hospital examine over 40,000 in-and-out patients during the critical months of the pandemic. Spot South Korea’s Hyundai auto group deployed Spot, a safety service four-legged robot developed by Boston Dynamics, for pilot operation at their auto plant. Equipped with deep learning-based vision technology and task management, the robot was applied with autonomous navigation, artificial intelligence, teleoperation technologies, and computing payload that is useful in industrial tasks. A subsidiary of the auto group, Kia Motors, introduced the pilot operation to assess its effectiveness and applicability. Spot has an integrated 3D LiDAR light detection and ranging device, thermal camera and can detect persons around it and monitor potential fire hazards. With AI technology, the robot can detect dangers, alarm managers, provide real-time photos of on-site situations, maintain an activity log with the control centre, and even sound an alarm as a rapid response in case of an emergency. The robot provides great support to the late-night security patrols as it creates a safer environment for the workers. The best part is that it is a legged robot (not a wheeled or tracked robot) and can move through extreme terrain, including narrow mountain trails and rocky boulders. Robotic grass cutters Changi Airport, Singapore, has deployed seven robotic grass cutters to help in maintaining all grass turf which is as big as the size of 1000 football fields. The airport’s aim is to make grass maintenance, a labour-intensive service, completely man-less. The robotic grass cutters have been acquired from Italian company Ambrogio and Swedish firm Husqvarna. The solar-powered robots are programmed to maintain specific areas and automatically head back for a recharge. For an airport, it is essential to maintain a grass turf as it prevents soil erosion at the airport, absorbs surface water run-off during storms and even proves to be a cushion in case of aircraft veering off. The maintenance is extremely difficult as short grass that is too short can draw birds to spot prey and pose a danger to aircraft, and long grass can become a sanctuary for wildlife hiding from predators. During the test runs, each of the seven robotic grass cutters covers an area of 1.2ha – approximately the size of one and a half football field. 7-Eleven’s delivery robot South Korea’s 7-Eleven convenience stores have started testing autonomous delivery robots in southern Seoul. Currently, the robot will cover delivery runs within a radius of 300 meters from a designated store till January, and the operation area will be expanded after the results. 7-Eleven will observe the way in which the robot operates in urban environments, especially when intertwined with obstacles like street poles and pedestrians. The 7-Eleven’s robot was created in partnership with Neubility. With high-rise buildings and sky-scrapers, GPS based-autonomous driving proves useless, and LiDAR-based autonomous driving solutions are very expensive. The development cost of Nubility’s multi-camera system in its robot is one-tenth of the LiDAR solution cost. Wrapping up Robots have become smarter and more autonomous over the years and might be a lot around us. Experts opine that they might learn mundane tasks without programming but will always lack moral reasoning. With morality being culturally specific and a continuously evolving thing, developing a perfect AI-robotic system is a major challenge.","excerpt":"Top five autonomous robots that are currently deployed at experimental level and have a potential to become a norm in 2022","categories":["AI Features"],"tags":["Alphabet","Boston Dynamics","Google","hyundai","LiDAR sensors","Spot"],"author_name":"Meeta Ramnani","publish_date":"2021-11-25T17:00:00","publication_year":"2021","word_count":939,"keywords":["Go","Boston Dynamics","API","artificial intelligence","programming_languages:R","AI","LiDAR sensors","ML","Spot","Aim","deep learning","ViT","Google","Alphabet","R","hyundai"],"extracted_tech_keywords":["AI","artificial intelligence","ML","deep learning","Aim","R","Go","API","ViT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/experimental-robots-that-might-be-around-us-in-2022\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":1546,"title":"Smart Cities Initiative moves ahead: Here’s city wise updates from India","content":"Smart City Mission comes as one of the most ambitious projects by the government of India and over the last few months the development towards a smarter city have only accelerated indicating the fact that it is one of the most funded missions by the government. With many provisions being made for a hassle free involvement of private entities, this year’s budget even saw many perks for those eyeing to invest on a smart city mission. One of the flagships missions by the Ministry of Urban Development, Smart city Mission comes as a bold initiative to propel economic growth and facilitating a decent life to its citizen through its Smart solutions such as e-governance, energy and waste management, urban mobility etc. If we go by the numbers, there are a total of 60 cities under Smart city mission which were selected based on the plans submitted and the total cost of these projects amount to 131,762 Cr., which is a not petty number. While the ambitions and targets are high, let’s go through a quick scroll on what has been the recent updates since the beginning of 2017 in various smart city projects, the cities listed in alphabetical order. Aurangabad- With an aim of setting up WiFi facilities at 40 different locations in the city, Aurangabad is trying to stay up the game. A part of Smart City project, if the facility is set in place, it would keep the people on the move connected. The hotspot installed is reportedly set to give internet access to persons in a radius of 25m. The places aimed for installation are markets, bus stands, gardens, other commercial places, important junctions and other public areas. However, it is still not clear if this facility would be available for free or on a revenue generation model. It is worth noting that the civic body had recently planned projects worth Rs 1,730 crore under the smart city scheme. Bhubaneshwar- In the wake of making the city smart, the Bhubaneswar Smart City Limited started a people-connect programme with United Nations Population Fund (UNFPA) at Jagannath slum. The MoU signed would facilitate capacity building, sharing of knowledge and self-defense training of girls in slums. Total of eight slums in different phases have been targeted for the development under smart city mission. And if reports are to be believed, this project has been allocated a sum of Rs 1.6 crore, which is indented to initially run till the end of December 2017. Chandigarh- One of the best known smart cities in India that made it to the list of 13 smart cities in the second phase of fast-track completion in May 2016, is now all set to get a world class Smart City innovation centre. Bringing twenty big companies such as Microsoft, Google, IBM, Cisco, Intel, Nokia (et.al) together on one platform, these are slated to roll out plans for the smart city project. The center, if all goes well, would be operational from the Punjab Engineering College in Sector 12 of Chandigarh by the first week of April 2017 and the technologies displayed by the companies would be for a display once the innovation center is developed. Set with an aim to offer smart solutions for the better infrastructure and a cleaner environment, the center would be designed with all latest technologies to improve upon services like public safety, transport, traffic management, waste disposal, e-governance etc. The smart city project by the Chandigarh Administration will be developed in three zones. Ghaziabad- In a recent development, the Ghaziabad Municipal Corporation has identified nearly 1,317 acres in the city that can be redeveloped under the new Smart city proposal, which will be sent to the Union urban development ministry for third round of Smart City Mission. Ghaziabad had earlier lost out on previous two selections. Demarcating certain areas of the town such as portions of Vasundhara, Vaishali and Kaushambi townships under the Area-Based Development component of the proposal, the Pan-City Solutions would have a focus on developing eco-friendly urban transportation and intelligent traffic management system. SmartCity- Bhubaneshwar Jaipur- As a part of the Smart City project, the city intends to install air and water quality monitoring systems at 56 locations across the city by June 2017. The city has already installed real-time air pollution monitoring system at three locations. By monitoring and displaying air pollution parameters including oxides of sulphur and nitrogen, these air pollution systems have been playing a big role in monitoring air pollution. These smart systems can also track the levels of particulate matter suspended in air, i.e. PM10 as well as PM2.5. Apart from this, areas of the old city such as 650 acres of the Walled City, Albert Hall and Jorawar Singh Gate have been in for Area-based development. It is also worthwhile noting that the Jaipur Municipal Corporation, has reduced the budget of Prime Minister’s ambitious smart city project from Rs 50 crore to Rs 20 crore. Mangalore- The first ever development under the smart city mission would come in the form of the first ever concrete road from Lady Hill Circle to PVS Junction that is about to go smart. A proposal put forward by Mangaluru City Corporation (MCC) budget would have Rs two crore set aside for the project. Drawn on the concept of TenderSURE roads, it would also be implemented on Bengaluru and other cities. Some of the features of the smart road would be pothole free road, solar panel-enabled street lights, neatly painted medians, provision for underground cable ducts, cross drains under the main carriageway, CCTV surveillance and more. It is also planned to be made dust-free. To give a glance at Smart city project and urban development, two-day Karnataka Smart City Project workshop was also organized earlier his month. Pune- Eyeing to be the world’s first city to have a city-wide WiFi connectivity, Pune has roped in technology giants like Google, IBM, Larson & Toubro and RailTel, that would provide the said facility based on freemium model. Reportedly a sum of Rs 150 crore has been allocated for this smart city initiative. A Special Purpose Vehicle designed by Pune city administration called Pune Smart City Development Corporation (PSCSCL), would oversee the smart city feature and capabilities. These companies would implement various smart city initiatives such as city-wide WiFi, environmental sensors, public address system powered by digital mode, emergency call boxes, enhanced network connectivity and more. Google would deploy its Google Station platform that has an inbuilt WiFi network management capability. RailTel would provide last-mile connectivity using optical fiber, L&T would be the master system integrator helping Google and RailTel integrate every aspect together. Whereas IBM would facilitate the hardware and software capabilities. Shimla- Stressing on completing the beautification of the town at the earliest, it is clear as to how the state officials want Shimla to be a part of the Smart City mission. Working a step towards it, Himachal Pradesh government has suggested to build the first ‘smart hill town’ of the country near Waknaghat in Solan district. Aiming to utilize the 450 bighas of land which is already available, the boost also comes from Singapore government who have agreed to invest Rs 2,500 crore in this project. The smart hill town is planned to be built in a way that provides abundant opportunities to education as well as businesses in hospitality, health and wellness. Designed to have facilities of international standards, the idea of this smart hill is to create new planned settlements so that a more balanced distribution of population is achieved. Vadodara- Given the fact that the government has allocated Rs 597 cr under the Smart City Mission for Ahmedabad, Surat and Vadodara, they are quite clear on their plans to be the one amongst the smartest cities in the country. In addition to this, the state government had earlier given a nod for the formation of the special purpose vehicle, which forms a crucial part of the Smart City Mission projects. Tentatively named Vadodara Smart City Development Ltd., the implementation of Smart City Mission at the city level will be done by the SPV. Few things it is assigned to do are- planning, appraising, releasing funds, implementing plans, managing and evaluating the Smart City development projects. Stay tuned for more updates on smart city mission in India….","excerpt":"Smart City Mission comes as one of the most ambitious projects by the government of India and over the last few months the development towards a smarter city have only accelerated indicating the fact that it is one of the most funded missions by the government. With many provisions being made for a hassle free […]","categories":["IT Services"],"tags":["smart city india"],"author_name":"Srishti Deoras","publish_date":"2017-02-27T04:54:55","publication_year":"2017","word_count":1383,"keywords":["Go","AI","innovation","ML","Git","smart city india","BERT","Aim","ViT","GAN","R"],"extracted_tech_keywords":["AI","ML","Aim","R","Go","Git","BERT","GAN","ViT","innovation"],"url":"https:\/\/analyticsindiamag.com\/it-services\/smart-cities-initiative-moves-ahead-heres-city-wise-updates-india\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10045701,"title":"How Ben Shneiderman&#8217;s Treemaps Found Place In The Museum Of Modern Art","content":"Treemap, a predominantly utilized visualization chart type, was invented by Ben Shneiderman in 1990. The Founding Director of the Human-Computer Interaction Laboratory (HCIL) was inspired by the images of Optical art that he came across during the 1960s, at the Museum of Modern Art in New York. Treemaps are huge square boxes that include data trees in the form of rectangles. The concept involves colours and shapes to achieve clarity in presenting complex data in the form of trees. Shneiderman designed 12 such treemaps using real-time data, which are preserved in museums of modern art across America. While he initially used the Treemap to understand space allocation on a hard drive in 1990, the aesthetic choices involved in making treemaps paved the way for further implementation of the concept in data visualisation. David Woodword’s work ‘Art and Cartography‘ established the importance of visual elements used to narrate a story. In his book, he explains the paintings of Dutch painters who sketched maps of that era hung on walls as pieces of art. How Op-Art inspired Shneiderman Source: Wikipedia Shneiderman’s Treemap is a tiling mechanism used to visualise a tree structure through nested rectangles of different colours. Treemaps are primarily used to visualise hierarchy in data. This concept for data visualisation focuses on the adaptable use of space, with colours representing varying information. Shneiderman was inspired by the 1960’s ‘Op Art’ and the exhibits that he came across at the Museum of Modern Art in New York. Op Art or Optical Art is a form of kinetic art related to geometric designs that create movement as a viewer sees the image. In 1965, an exhibition in New York entitled “The Responsive Eye” had several black-and-white op art images on display. The patterns of art in these images influenced the style in print graphics, fashion, and advertising. Shneiderman derived inspiration from the Op Art and established four aesthetic aspects of Treemaps: Layout design (slice-and-dice, squarified, ordered, stripped)Colour palette (muted, bold, sequential, divergent, rainbow),The aspect ratio of the entire image (square, golden ratio, wide, tall), andThe prominence of borders for each region, each hierarchy level, and the surrounding box. What exactly are Treemaps? Treemaps are boxes with rectangles of different sizes that are coloured. The aesthetics used in these shapes display data that triggers interest (emotions) in viewers. Even as algorithmic rules constrain treemaps, aesthetics involved in the process make room for creative exploration. In his book titled ‘Every AlgoRiThm has ART in it’, Scheiderman explains the 12 prints of Treemaps based on data provided by international organisations. The original prints of these treemaps were first put on display in the University of Maryland Computer Science Instructional Center on 26 August 2013. While it was initially decided that the prints will be hung for two months, it has become a permanent exhibit for the state of the art it entails. Ben Shneiderman with one of his treemaps. Source: WordPress Shneiderman’s treemaps have been traveling places under the exhibition titled “Every AlgoRiThm has ART in it: Treemap Art Project“. With treemaps, Shneiderman claims that the human mind processes information differently when it is organised visually. It is an attempt to understand human interaction with data visualisation through art. With the use of coloured blocks, Treemap makes it easier for viewers to base their understanding on size, aspect, and colours. Treemap is composed of nested rectangles or tiles, and the area of each rectangle is proportional to the data they represent. Larger rectangles denote higher hierarchy levels and are the main\/ root branches of a “Data Tree”. The data tree is then divided into smaller boxes that show the lower levels of the hierarchy. These rectangles are used to display datasets, where the goal is to “break” the data into its constituent parts to identify its larger and smaller components quickly.  For instance, the following Treemap categorises Jamaica’s exports in 2017, separated by sector – chemicals, minerals, oil, agriculture, etc. An example of a Treemap. Source: Wikimedia Commons Schneiderman’s Art with Authentic Data To establish his findings in data visualisation through treemaps, Shneiderman used authentic data from the World Bank and explained how urban populations could be categorised in different countries. He even gathered information over carbon emissions from the US Energy Information Administration and organised all the data into seven continents to display them on a treemap. The size of the boxes in this illustration represents CO2 emissions while the colours vary by per capita CO2 emission, writes Shneiderman in his book. He also used the Treemap to illustrate the engagement score of audiences for the TED talk. Sebastian Wernicke compiled the data for his TED talk titled “Lies, Damned Lies and Statistics”. The colours in this specific Treemap attempted to show the variety in the TED talks. The 12 Treemaps in Museums of Modern Art His illustrations with real-time data bridge the gap between art and science and show that both disciplines intersect in many given instances. Many of Shneiderman’s works have found a place in museums of modern art, the first being the University of Maryland, followed by the National Academy of Sciences in Washington DC, where a copy of Shneiderman’s Treemap is on display at the Keck Center, from October 2014. Another copy of the Treemap Art Project was on display at the Howard Hughes Medical Institute, Virginia, and was later moved to Chevy Chase, MD headquarters in 2016. Next is the Museum of Modern Art in New York, which received all the 12 prints of Shneiderman’s treemaps and were conserved by the museum. University of California, Irvine, and Swansea University in the UK also houses Shneiderman’s treemaps. The Treemap that Schneiderman designed with data on urban populations that he gathered from the World Bank. Source: Ben Schneiderman’s book The University of Maryland, where Shneiderman is a Professor, also has developed a tool to produce treemaps that are available for download and use. Treemaps are one of the most efficient tools to understand complex data. The concept is aimed at presenting bulk data in hierarchical patterns in a space-constrained layout. The significance of this tool is its efficiency in using space and intelligent colour mechanisms to generate a layout with rectangles that show variations in data based on its quantity. In a way, treemaps ease data interpretation by eliminating heavy texts and replacing them with shapes and colours.","excerpt":"Ben Shneiderman was inspired by the 1960’s ‘Op Art’ and the exhibits that he came across at the Museum of Modern Art in New York. Op Art or Optical art is a form of kinetic art related to geometric designs that create movement in the eyes.","categories":["AI Features"],"tags":["art","University of California"],"author_name":"Gourav Mishra","publish_date":"2021-08-11T12:00:00","publication_year":"2021","word_count":1057,"keywords":["Go","API","programming_languages:R","AI","University of California","programming_languages:Go","art","Aim","GAN","R"],"extracted_tech_keywords":["AI","Aim","R","Go","API","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-ben-shneidermans-treemaps-found-place-in-the-museum-of-modern-art\/","complexity_score":4,"technical_depth":8,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10129795,"title":"AI Still Needs Humans for the Cumbersome Task of Data Annotation","content":"According to recent studies, OpenAI’s GPT-4 has been successful in accurately annotating cell types using marker gene information in single-cell RNA sequencing analysis. Generative AI has expanded the use cases of data annotation and labelling. However, human expertise will always remain core to its success. Acknowledging the same, Radha Basu, the co-founder and CEO of iMerit, told AIM in an exclusive interaction, “Traditional data annotation methods, which relied on low-skill crowdsourced workforces, were effective for simple tasks but were limited in scalability and efficiency for complex data sets.” Now, GPT-4’s annotations aligning closely with manual ones across various tissues and cell types has helped cut down the time and expertise needed for annotation. “Firstly, such (generative AI) models inherently have a larger context window than traditional predictive AI. Secondly, the need for data training is more at the expert practitioner level since the underlying foundation model is built with unstructured or semi-structured data,” added Basu. However, there have been mixed opinions on the future of data annotation platforms with generative AI. As per recent reports, the global data annotation market is projected to be valued at around $8.22 billion by 2028. But Voxel51 co-founder and University of Michigan professor Jason Corso is of the opinion that data annotation jobs are slowly becoming obsolete. Contrary to his expectations, however, technology is not eating up traditional data annotation jobs. Generative AI’s larger context window and expert-level training capabilities differentiate it from traditional data annotation methods, making it more adaptable. Currently, foundational model-based platforms can incorporate feedback and corrections from domain experts, such as medical professionals, agronomists, or mathematicians, to refine and improve the accuracy of annotations. “This expert collaboration ensures the data used to train AI models is of higher quality,” Basu highlighted. “Combining generative AI’s processing power with the knowledge of trained professionals leads to more reliable and accurate AI systems.” Basu noted that LLM-based systems can handle more intricate data labelling tasks by integrating automation into workflows, improving efficiency and scalability without long hours of manual effort. Meanwhile, human annotation has been vital for AI, encouraging the growth of supervised machine learning. So now, although self-supervised learning and auto-annotation are emerging, they cannot fully replace human annotations. How Does iMerit Fit in? iMerit’s proprietary model Ango Hub leverages generative AI to propel growth in different industries such as medicine, autonomous mobility, and precision agriculture. Ango Hub offers a flexible workflow manager that integrates human and machine efforts, allowing public models to be included in the process. The platform utilises a curated network of experts for tasks like reinforcement learning with human feedback (RLHF) and supervised fine tuning (SFT), providing high-quality, domain-specific data training. The ability of GenAI to handle and integrate various data types will further improve its role in data annotation, leading to more comprehensive and accurate AI models. Apart from being one of the exceptional technologists, Basu is also known to foster a good work culture. Her leadership strategy centres on authenticity, dedication, and innovation, focussing on giving one’s all and taking risks. “My leadership style is one of constant innovation while staying focused on my company’s role in the ecosystem and what our customers actually need from us. I am always looking at embracing change, mixing and matching the three pillars – technology, talent, and technique,” she added. In the context of AI and ML innovations, the vision for iMerit is to build a responsible, inclusive organisation by prioritising a diverse, motivated workforce, applying technology to societal needs, and maintaining sustainable business practices with strong financial discipline. About 52% of the organisation comprises women. India as a Market Over the past decade, the company has adapted to the shifting data requirements by evolving from simple, prescriptive tasks to more complex, domain-specific projects requiring a consultative approach and expert collaboration. This evolution included the adoption of heavy dashboards and production metrics, automation, and workflow orchestration. Basu said that India has managed to become a rapidly relevant market due to the rise of GenAI companies. “India is a very exciting market due to the rapid growth in GenAI companies and the interest in building an India stack, including local languages and problems to be solved,” said Basu. The revenue share from the Indian market has seen rapid growth during 2023-24. Meanwhile, there has also been a noticeable increase in inquiries for such data corpus creation, domain tuning, and red-teaming in various local languages, leading to active collaborations with customers on Indian language-based stacks. “We have to do justice to both [the customer and the workforce] in order to achieve quality and consistency,” Basu emphasised. This has further been recognised as iMerit and Ango Hub have managed to win two awards in India this year alone for being best in class in terms of machine learning, application and solutions provided. Looking ahead, Basu believes that the integration of generative AI with multimodal data (combining image, speech, text, LiDAR, and video) is expected to change the visual domain in industries such as medical AI and autonomous mobility.","excerpt":"Currently, foundational model-based platforms can incorporate feedback and corrections from domain experts to improve the accuracy of annotations.","categories":["Deep Tech"],"tags":["data annotation","GPT-4","OpenAI"],"author_name":"Shritama Saha","publish_date":"2024-07-22T14:50:01","publication_year":"2024","word_count":836,"keywords":["GenAI","machine learning","RLHF","OpenAI","AI","ML","GPT-4","medical AI","RAG","Aim","data annotation","generative AI"],"extracted_tech_keywords":["AI","machine learning","ML","generative AI","GenAI","OpenAI","Aim","RAG","RLHF","medical AI"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/ai-still-needs-humans-for-the-cumbersome-task-of-data-annotation\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10102387,"title":"Now You Can Reserve NVIDIA GPUs on AWS","content":"After a month of struggle, AWS has now introduced a solution to address the surging demand for GPU compute capacity for machine learning workloads. The company announced the general availability of Amazon Elastic Compute Cloud (EC2) Capacity Blocks for ML, offering an innovative consumption model for customers, powered by NVIDIA H100 GPUs. With this new offering, customers can access highly sought-after GPU compute capacity on a flexible, short-term basis, removing the need for long-term commitments. “This is an innovative new way to schedule GPU instances where you can reserve the number of instances you need for a future date for just the amount of time you require,” said Channy Yun. EC2 Capacity Blocks allow customers to reserve GPU capacity for durations ranging from one to 14 days, with advanced scheduling up to eight weeks. These capacity blocks are deployed in EC2 UltraClusters with low-latency, high-throughput connectivity, offering the flexibility to scale up to hundreds of GPUs. This new solution is ideal for training and fine-tuning ML models, short experimentation runs, and handling temporary surges in inference demand for product launches. David Brown, the vice president of compute and networking at AWS, highlighted the importance of this innovation in democratising access to generative AI capabilities. AWS and NVIDIA have collaborated for over a decade to deliver scalable GPU solutions, and the introduction of Amazon EC2 Capacity Blocks is a significant step in broadening access to GPU capacity for generative AI applications. The EC2 Capacity Blocks are available for reservation in the AWS US East (Ohio) Region, with plans for expansion to additional AWS Regions and Local Zones. This development is a game-changer for startups and organisations looking to harness the power of generative AI without making long-term capital commitments. The launch of EC2 Capacity Blocks has received positive feedback from industry leaders and organisations, such as Amplify Partners, Canva, Leonardo.Ai, and OctoML. These stakeholders believe that this new solution will provide the predictability and timely access to GPU compute capacity necessary to drive innovation and meet customer demands in today’s supply-constrained environment.","excerpt":"OctoML, Leonardo.Ai, Canva, and Amplify Partners are happy with the decision.","categories":["AI News"],"tags":["AWS","NVIDIA"],"author_name":"Mohit Pandey","publish_date":"2023-11-02T15:42:29","publication_year":"2023","word_count":341,"keywords":["API","machine learning","AWS","AI","ML","Scala","ViT","generative AI","GAN","NVIDIA","R"],"extracted_tech_keywords":["AI","machine learning","ML","generative AI","AWS","R","Scala","API","GAN","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/now-you-can-reserve-nvidia-gpus-on-aws\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":33012,"title":"AIM Announces India&#8217;s Largest Gathering Of Women Data Scientists At &#8216;The Rising 2019&#8217;","content":"Analytics India Magazine, India’s first data science-focused magazine and organizer of Cypher and Machine Learning Developer Summit announced The Rising 2019 – one of India’s biggest gathering of women data scientists and technologists to be held at The Taj, MG Road, Bangalore on March 8. The one day conference will be hosted by AIM to promote and celebrate women innovators in data science, analytics and AI field and celebrate the industry’s trailblazers. The one-day event is set to have more than 250+ participants and will feature more than 15 talks from industry leaders. The conference will feature keynote speakers, panel discussions and workshops designed to inspire more women participation in the STEM field and also help attendees grow in their analytics careers. Bhaskar Gupta, Co-founder and CEO, Analytics India Magazine spoke about the importance of this conference series, “This is the largest group of women technologists and the conference is aimed at inspiring the next generation of women leaders in analytics. As the industry evolves, we need to establish ways to recognise the groundbreaking contributions and innovations made by women technologists. The conference is also aimed at inspiring more women to pursue a career in STEM and at the same time urge companies and corporate partners to build a culture of equality and diversity”. Participants can also meet more than 100 leading companies that are working to build a more tech-inclusive industry for women technologists. Conference attendees will get a chance to meet companies that are actively hiring and understand their company culture. Here’s what you can expect from the conference: Panel discussions: Attendees can learn how to grow in their tech careers and advance from software engineering roles to technical leaders. Participants can learn from women architects who will share their career transition experiences with the attendees. In addition to this, there will also be talks on how to balance resources, understand customer demands and master project management. Keynotes: Participants can learn from immersive sessions where women technologists and team leads will speak on how they build a career in data analytics and artificial intelligence. Networking: Professional development is a big part of the learning journey and the conference has been designed to help women in tech connect with top companies that value inclusion and are also hosting returnee programmes. The Rising will be held on March 8, Friday at Hotel Taj, MG Road, Bangalore. To participate, click here About Analytics India Magazine Analytics India Magazine is India’s premier data science-focused platform and organizer of India’s biggest analytics summit Cypher and Machine Learning Developer Summit. The company has been a clear voice advocating data science and analytics over the last six years. The company also runs Machine Hack, an online platform for machine learning hackathons.","excerpt":"Analytics India Magazine, India’s first data science-focused magazine and organizer of Cypher and Machine Learning Developer Summit announced The Rising 2019 – one of India’s biggest gathering of women data scientists and technologists to be held at The Taj, MG Road, Bangalore on March 8. The one day conference will be hosted by AIM to […]","categories":["Deep Tech"],"tags":[],"author_name":"Richa Bhatia","publish_date":"2019-01-08T06:22:03","publication_year":"2019","word_count":456,"keywords":["data science","machine learning","artificial intelligence","programming_languages:R","AI","innovation","Aim","analytics","GAN","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","data science","analytics","Aim","R","GAN","innovation","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/aim-announces-indias-largest-gathering-of-women-data-scientists-at-the-rising-2019\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":6047,"title":"In conversation with Digital Analytics Expert Harikrishnan Srinivasan","content":"Harikrishnan Srinivasan is the founder of TribeNautix eCommerce Consultants Pvt Ltd. TribeNautix provides Digital Marketing & Intelligence services for leading online brands in India. Hari is a specialist in Web Analytics & Digital Marketing with experience in domains including eCommerce, Travel, Health Care, Online Training and HR. He has over 20 years of experience in information services and technology – with leading companies including Verizon, Cognizant & Wipro. Hari has recently been working with the team at Jigsaw Academy, the online school of analytics to develop a web analytics course, which will be launched shortly. In an interview with Analytics India Magazine, Hari talks more about what got him interested in web analytics, his predictions for analytics, the soon to be launched web analytics course at Jigsaw and advice for those interested in a career in web analytics. Analytics India Magazine: What is so important about Web Analytics and what got you interested in it? Harikrishnan Srinivasan: Traditional marketing continues to rely on thumb rules, intuition, not well understood TRPs and peer perceptions – even when it comes down to deciding whether an ad campaign was successful. In contrast, digital media offers measurable and reasonably accurate data – for ROI calculations. With the increasing adoption of internet in India and growing media spends online, performance based marketing is becoming the norm. I see this as a great opportunity for many years to come. Considering my interest in statistics & quantitative analysis, it seemed natural to move full-time into the digital marketing & online analytics domain. AIM: Can you share an interesting example of the benefits of web analytics. Hari: One of my favorite examples is from a travel portal where customers would visit the website to cancel a ticket – by clicking on a paid ad. These are folks who had already visited the website earlier to purchase – so they know the website name by URL & brand name. Yet they chose to click on paid ads – costing the company additional ad spend, bringing in negative revenues at higher costs per transaction. This behavior was quantified using a combination of both Analytics Reports & Web Site Visitor surveys. The company is now creating awareness among customers to use low cost visit channels – if they intend to cancel (Organic search results that display cancellations prominently in site links, email URL links & incentivize users to bookmark the URL). This will bring down unit transaction costs and overall ad spends – as the business scales. AIM: What do you think is the future of analytics, in particular web analytics? Hari: With the growth in internet data access across the world and the explosion of affordable & powerful mobile devices we will see huge data growth across many categories – B2C, B2B, Personal, Machine to Machine and Machine to Humans. This will create specialized careers in the web analytics domain over the next five years – Sports, Health Care, Digital Advertising, Lifestyle and Geo Targeting to name a few, with focus on predictive techniques including crowd sourced data. For example, Tamil Nadu Cricket Association today provides live scores of even inter-school matches on its website – this will be a treasure trove for sports analysts who would want to know everything about the future Sachins & Dhonis of Indian cricket – when they were in school!! We will also see horizontal growth as in Personalization, Recommendation, Multi Device User Profiling, etc. Things could get complex as data formats are varied & unstructured, volumes are huge and data flow is rapid. So dependency on algorithms & machine learning would increase. Web Analysts would need to get comfortable with these techniques as well in the next 3 years. All in all a good time to get into web analytics – with new opportunities to learn & grow over many years. AIM: Any other advice for those thinking of a career in web analytics. Hari: Many folks interested in web analytics as a career are intimidated by this picture – they assume that web analytics is data & math intense and requires technology background. There is also this image of an analyst pouring through multiple reports all day long for one single insight. However there is another completely different facet to this job. The web is also a great way to connect with your audience & customers in a faster, personable & reliable fashion. Website Vistor Surveys, Customer Ratings, Product reviews, Brand Sentiments, etc are all actual people speaking out loud on the web. Many a times I get to ask visitors why they visit a website and whether their visit was successful via a 15 second site survey – as opposed to looking for Keywords & Website Engagement metrics. So there is this qualitative and people interaction aspect to Web Analytics that makes the job very interesting and anyone considering Web Analytics as a career should keep this perspective as well. AIM: And finally tell us more about the course in Web Analytics you are helping develop at Jigsaw Academy. Hari: Yes, I am really excited about this course. The differentiator here is the business perspective we bring. Traditional Web Analytics training is focused on tool functionality (Configuration, Report Creation and Customization) & data reporting (like visitors per day, bounce rate, time on site, conversion rates). The Jigsaw Academy program will prepare web analysts to solve for real world business decisions – like audience selection, media planning, campaign messaging, channel spends, brand awareness, sales, customer retention, etc. The program is holistic – covers innovative KPIs for various online business functions, data strategy, measurement frameworks, analysis and interpretation for business actionables. This Jigsaw Web Analytics Course is for professionals in the E-Commerce industry interested in assessing online business performances (SEM, SEO, Social Media) or those in traditional business aspiring to move into Online Businesses. It is also a great career enhancer for MBA \/ BBA students interested in getting that edge and furthering their career in E-Commerce Marketing & Sales.","excerpt":"Harikrishnan Srinivasan is the founder of TribeNautix eCommerce Consultants Pvt Ltd. TribeNautix provides Digital Marketing & Intelligence services for leading online brands in India. Hari is a specialist in Web Analytics & Digital Marketing with experience in domains including eCommerce, Travel, Health Care, Online Training and HR. He has over 20 years of experience in […]","categories":["AI Features"],"tags":["Interviews and Discussions"],"author_name":"AIM Media House","publish_date":"2014-08-22T15:18:22","publication_year":"2014","word_count":997,"keywords":["Go","API","machine learning","programming_languages:R","AI","Git","Aim","analytics","GAN","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","machine learning","analytics","Aim","R","Go","Git","API","GAN","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/in-conversation-with-digital-marketing-web-analytics-expert-harikrishnan-srinivasan\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":18805,"title":"How To Spot A Real Data Scientist","content":"Data scientist jobs has been ruling the employment popularity charts for the longest time, and with many renowned reports suggesting it to be sexiest and the most paid jobs of 21st century, it has become the sought after job profiles that candidates are willing to be placed into. However, it may not be always the case that the most deserving candidate grabs the spot. Given the newness of the role and lack of understanding of the exact job description, there are people who tend to apply for the role, purely because they deal with data. That however is not true of a data scientist’s capabilities! As we all know by now that data scientist is not just someone who knows data well, but there are many more other skillsets and mindset that makes it into the must-have list of data scientist’s resume. Since there is no regulatory or licensing body overseeing the hiring process for data scientists, it becomes difficult to scrutinise the right kind of candidate, and there may be many fakes ones applying for the job too. The challenge here is how to spot a real data scientist amidst the resumes flowing from both genuine and pseudo genuine candidates. With this post, let’s try to figure out how can a candidate’s profile sums up in all aspects to be that of a real data scientist. Signs of a real data scientist Pro in most quantitative skills such as statistical algorithms and other tools with a high qualification degree- This goes without saying that most of the real data scientist that make a way into data science roles in various companies come with a highly advanced and technical degree. Most of them have at least a master’s degree and a PhD. Their area of study may be different, but most of them include math, statistics, computer science, economics etc. as their area of expertise. They have a sound technical knowledge of languages like R and the use of algorithms to collect and analyse millions of data. Machine learning becomes one of their core understandings of the subject as these algorithms have the capability to get more and more accurate over time to deliver better results. Has specific use cases in terms of tools and models used- Apart from having a degree in data science, actual data scientists have an inside out knowledge of tools like SAS, Python, or Hadoop. Unlike fake candidates who just list these tools, real ones can support having used these tools with projects evidence. It is important that data scientist gets their experience straight with working on unstructured data with perfect and genuine examples. If their involvement with a project sounds vague, they probably aren’t good data scientist. If a person merely has a experience in terms of organising and structuring large data sets, and no experience with analytical or statistical concepts, the professional may be a data engineer or a data analyst but not data scientist. High in business acumen- Those coming from a purely academic or research background may not be the ones those who fit the bill of a corporate environment perfectly. To be an effective data scientist it is important that the candidate pose necessary business acumen to understand the findings and use it in a way that impacts business goals. A true data scientist also knows how to deliver actionable insights to business leaders in an easy to understand manner. A key component to being a data scientist in a corporate setting is thus, business acumen. They have a lot of other data scientist in their network- Lastly, not one of the must have criteria to spot a real data scientist, it is usually handy to spot the real ones by checking the LinkedIn profiles. As the case is in the usual scenario, one’s professional network is filled with other individuals within the same field, the same applies for data scientists. If a candidate is genuine, he\/she might have an interest in knowing others within the same industry, and hence may boast a lot of genuine data scientists in their network. Few questions to spot a fake data scientist As we have read till now that data science is not just an expertise in one particular field, but a knowledge of the application of full stack of scientific tools and techniques such as mathematics, computational, analytical, visual, statistical, business acumen etc., it becomes handy in identifying the right candidate by asking questions that digs out the right and requisite skills from a candidate. Those posing themselves as data scientist may find it difficult to cover all these aspects as they insist that their discipline is the one defining true data science—which may not be the case. Some of the questions that can help in spotting the real data scientist are: Technical questions such as regularization and its use, statistical power, difference between ‘long’ and ‘wide’ format data, resampling methods, Edward Tufte’s concept of “chart junk” etc. Data scientists and companies\/ startups in this space that they admire the most Validating a model, they had created to generate a predictive model of a quantitative outcome variable using multiple regression. Understanding of the terms such as pricing optimization, price elasticity, inventory management, etc. Identifying if a statistics published in an article is wrong or presented to support the author’s point of view Recommendation engine and its working Difference between false positive and false negative and why is it important to understand the difference. Etc. On a concluding note Spotting a real data scientist may not be the easiest of tasks, but given the investment that would be made on these professionals in terms of high salaries and efforts in training them, it becomes important to get a thorough understanding of the candidate. He\/she should be well versed with applying mathematics, statistics and validating models using proper experimental design. It is important to dig deeper to get and spot the genuine data scientist that you might be looking for.","excerpt":"Data scientist jobs has been ruling the employment popularity charts for the longest time, and with many renowned reports suggesting it to be sexiest and the most paid jobs of 21st century, it has become the sought after job profiles that candidates are willing to be placed into. However, it may not be always the […]","categories":["IT Services"],"tags":["Data Science Hiring"],"author_name":"Srishti Deoras","publish_date":"2017-11-08T05:56:04","publication_year":"2017","word_count":996,"keywords":["data science","Go","machine learning","programming_languages:R","AI","Python","Data Science Hiring","programming_languages:Python","GAN","R","startup"],"extracted_tech_keywords":["AI","machine learning","data science","Python","R","Go","GAN","startup","programming_languages:Python","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/it-services\/spot-real-data-scientist\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10061150,"title":"AiDash acquires farming solutions company Neurafarms.ai","content":"AiDash, a leading provider of satellite- and AI-powered operations, maintenance, and sustainability solutions, has acquired geospatial- and AI-powered farming solutions provider Neurafarms.ai. The acquisition of Neurafarms.ai will help AiDash strengthen its product line and add valuable talent to its high-performing team. Neurafarms.ai co-founder and leader of the start-up’s data science innovations, Bhavesh Patidar, will join the AiDash team. He will focus on strengthening AiDash’s Disaster and Disruptions Management System (DDMS), which helps industries better manage disaster prediction, detection, and response. Dr Anil Singh, Neurafarms.ai’s head of research and development, will also join AiDash as a Remote Sensing Scientist. He will be responsible for enhancing the company’s research and development activities to analyse data obtained from remote sensing systems and channel it to solve industry problems. “We’re thrilled to have Bhavesh Patidar and Dr. Anil Singh join the AiDash team. We’re eager to see them put their respective in-depth knowledge into action to improve our products and better meet the needs of our customers,” said AiDash co-founder & CTO Rahul Saxena. Neurafarms.ai, founded in 2019, provides farm-level insights for reliable, data-driven decision-making to farmers and farming businesses spanning agriculture input-output companies, insurance, and financial institutions. “Neurafarms.ai’s vision was to change the way Earth Observation (EO) data is consumed in India, and I strongly believe that our acquisition by AiDash will not only help us accomplish our goals, but also transform the way stakeholders use data for decision-making across industries worldwide. Neurafarms.ai has developed an AI engine that combines data from multiple satellites and provides daily insights about the conditions on the earth. Our EO technology can be used across sectors to drive risk-optimised outcomes,” said Bhavesh Patidar, co-founder of Neurafarms.ai. The company leverages AI to analyse real-time satellite data for soil conditions, crop health and local weather predictions to provide actionable insights.","excerpt":"The acquisition will help AiDash strengthen its product line and add valuable talent to its high-performing team.","categories":["AI News"],"tags":["Mergers and Acquisitions"],"author_name":"SharathKumar Nair","publish_date":"2022-02-21T12:47:55","publication_year":"2022","word_count":301,"keywords":["data science","Go","TPU","AI","data-driven","innovation","RAG","ViT","disruption","Mergers and Acquisitions","R"],"extracted_tech_keywords":["AI","data science","RAG","TPU","R","Go","ViT","data-driven","innovation","disruption"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/aidash-acquires-farming-solutions-company-neurafarms-ai\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":16366,"title":"Gridlock Hackathon: How can data science &#038; AI help decode Bangalore traffic","content":"Img Src- Flipkart Traffic is an ever growing menace that almost all the popular “IT hubs” of India are continually facing—Delhi, Bangalore being the worst affected. Talking about Bangalore, once the city of lakes, gardens and a soothing weather, is now a city of overloaded vehicles with fuming fuels. While the traffic experts—meaning our respectable traffic police have been doing their best to have a control over it and facilitate a smooth sailing across the city, the efforts are far from yielding any desirable results. And why talk about it here? Well, to bring a data science and analytics angle to solving the traffic chaos! If you are wondering how, let us introduce to you the Gridlock Hackathon. This recently concluded hackathon organized by Flipkart, was arranged with an aim to find a technology that could solve Bangalore’s traffic congestion. And it turned out to be a huge success as the ideas from all across the city flowed in, in the form of NoHonk, RushHour, CitizenCop and other interesting project names. A challenge which was announced as a part of ‘Month of Innovation’ by the company on its 10th Anniversary, it received more than 400 entries from countries such as US, Dubai and Bangladesh, apart from India. Some of the companies that participated were ISRO, Mercedes Benz R&D, Amazon, Amazon Web Services, Accenture, E&Y, McKinsey & Company, Ola Cabs among others. The judges at Gridlock Hackathon- Img Src- Flipkart What’s to gain from the hackathon? As a part of the competition, participants had to use artificial intelligence and machine learning models to create solutions that integrate different systems to make traffic smooth and manageable. With a prize money of $5,500, the online contest was a way to crowd source solutions for Bangalore, which is grappling with unplanned roads and overpopulation. The technology boom that the city witnessed a decade ago has taken on its road and traffic conditions. Ideas that were welcome in the form of artificial intelligence, big data or anything out of the box that could give a solution to the over growing traffic, that reportedly is also affecting the overall GDP of the city, with traffic jams directly shoving about 2 percent from the city’s estimated GDP of $30 billion. The Flipkart CEO, Binny Bansal believes that the city has the potential to be a world class business and social destination, only if its traffic menace is addressed. Suggestions that poured in- From AI to IoT, the hackathon had suggestions pouring in from all the domains. There were entries that suggested that internet of things powered road dividers could change the orientation to handle changing situations to those that suggested having a reporting system that tracks vehicles that don’t comply to rules. Another interesting input was in the form of having a device that could track social media to generate traffic reports and a network of smart satellite that could direct in the ease of flow of vehicles. There were few suggestions around implementing cloud services. What the winning team says- The hackathon which was won by the team Affine Anonymous wants to take into consideration the traffic volume before they turn red, amber or green, with the help of real-time traffic feeds from Google and Bing Maps Vision. The project called TrafficSense, has envisioned the creation of a network of self learning and communicating traffic signal nodes. They suggest that the solution will be able to reduce the waiting time at the city’s most crowded junction- Silk Board, by 17%. The second and third positions were bagged by By2 Riders and Road Smoothness Detector by Eurus. BY2 rides is a mobile app with an Inverse-Uber model for carpooling and Road Smoothness Detector is an android app which helps grade roads. It can detect and locate obstructions on the roads such as as potholes, humps and others aspects of a bad road quality. The winning team. Img Src- Flipkart Last Word- The hackathon, which comprised of jury in the form of  Flipkart CEO Binny Bansal, Namma Bengaluru Foundation CEO Sridhar Pabbisetty and DCP (East) Abhishek Goyal, had an active participation from all across the city and most of the solutions saw the intervention of latest technologies such as big data, artificial intelligence, IoT and cloud in its overall implementation. With these disruptive technologies becoming a thing for all the industries, we await to see more on the traffic management side.","excerpt":"Traffic is an ever growing menace that almost all the popular “IT hubs” of India are continually facing—Delhi, Bangalore being the worst affected. Talking about Bangalore, once the city of lakes, gardens and a soothing weather, is now a city of overloaded vehicles with fuming fuels. While the traffic experts—meaning our respectable traffic police have […]","categories":["IT Services"],"tags":["ai hackathon india"],"author_name":"Srishti Deoras","publish_date":"2017-07-18T05:15:14","publication_year":"2017","word_count":733,"keywords":["ai hackathon india","data science","Go","big data","machine learning","artificial intelligence","AI","Aim","analytics","GAN","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","data science","analytics","Aim","R","Go","big data","GAN"],"url":"https:\/\/analyticsindiamag.com\/it-services\/gridlock-hackathon-can-data-science-ai-help-decode-bangalore-traffic\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10103963,"title":"Alibaba’s Cloud Business Gets Qwen-ched!","content":"Chinese tech-giant Alibaba is desperately trying to save its cloud business with large language models (LLMs). After releasing Qwen-7B in October, Alibaba recently released Qwen-72B, which has been trained on high-quality data consisting of 3T tokens. Compared to the previous versions, this has a larger parameter size and also an expanded context window length of 32K, with more customisation capabilities. Not just that, the company also added a smaller language model, Qwen-1.8B, touting it as a gift to the research community. It has a 2k context length and requires only 3GB of GPU memory. Both of the models would be available on Alibaba cloud for its customers and also as open source. Besides Alibaba, its competitors Tencent, Huawei, and Baidu are also building LLMs and are attracting customers and generating revenue rapidly. For example, as reported in July, Baidu’s AI cloud is leading the China market for the fourth year in a row, reporting a 69.7% growth in 2022. The company also released its Ernie Bot, competing with GPT-4. Same is the case with Tencent and Huawei. On the contrary, Alibaba’s cloud plans are crumbling. Qwen-ching GPT-4? Alibaba’s cloud unit’s growth has also decelerated, experiencing only a 4% annual revenue increase in the last fiscal, down from 23% growth in the previous year, and 50% before that. Despite Alibaba’s emphasis on the potential of AI software in China benefiting the cloud unit, the recent challenges, leadership changes, and US export restrictions on semiconductor chips have posed significant hurdles. According to reports, the company faces challenges as Chinese businesses, especially in traditional industries, and not the internet driven companies, show little enthusiasm for paying for public cloud services. Internal conflicts arose between Alibaba Cloud and the e-commerce business over post-spinoff terms, leading to a downfall of customers as well. Several months ago, Alibaba executives attempted to convince major investors to invest in its cloud unit, planning to spin it off as a separate company, called Cloud Intelligence Group, with a $40 billion valuation. However, the plan failed as investors were hesitant due to the cloud business’s slow growth and financial losses. The attempt highlighted the market’s reluctance towards the spun-off cloud unit, contributing to Alibaba’s recent decision to cancel the move. This cancellation led to a shed of around $21 billion in its market value, according to several calculations. Alibaba said that these curbs have “created uncertainties for the prospects of Cloud Intelligence Group,” making it difficult to compete with AWS, Microsoft Azure, and Google Cloud, if not just the Chinese counterparts. “Instead, we will focus on developing a sustainable growth model based on emerging AI-driven demand for networked and highly scaled cloud computing services,” Joe Tsai, CEO of Alibaba, said on the company’s investor call in November. The drop also highlights how this is an effect of the geopolitical tensions between US and China. Alibaba had to cancel its spin off plans because the US was controlling the chip exports to China, creating uncertainty for the company, but possibly the introduction of LLMs in its cloud might be able to save the company, just like Baidu, Huawei, and Tencent. Emad Mostaque, the founder of Stability AI, recently posted on X, “Chinese open models will overtake GPT-4 shortly zero shot, can already overtake if you chain Qwen & Deepseek appropriately.” Will test the new Qwen, local open source AI model soon.“A LLM family built by Alibaba Cloud. In this organization, we continuously release LLMs, large multimodal models, and other AGI-related projects”China is blasting through US AI regulations. https:\/\/t.co\/ue6J7Tyvd6— Brian Roemmele (@BrianRoemmele) November 30, 2023 Competitions galore The restrictions on the chip export, especially the ones NVIDIA makes, is creating a lot of challenges for cloud providers in China. But Alibaba appears to be more significantly impacted by it than its counterparts. Currently, Alibaba still holds a larger market share. But the competitors are increasingly rising up, and Alibaba is shaking grounds and is desperately trying to stay afloat by introducing new LLMs for its customers. For example, Tencent, which stated that its ample chip reserves will sustain its LLM development for several more generations, and Baidu, which mentioned a substantial reserve of AI chips supporting the ongoing chatbot enhancements for “at least a couple more generations.” Alibaba did not make a comparable statement. Tencent also slashed the price of its cloud services by 40% to compete with Alibaba. It also said that the revenue has been growing since the second half of the year. Huawei recorded a 3.1% increase in revenue in the first half of the year compared to previous year. It also reported an increase in overseas exports during the period. Most recently, at the Huawei Cloud Industry Summit Forum 2023, the company announced the industry’s first large model hybrid cloud, enabling both edge and public cloud, providing customers a suite of tools. This was after the company released Pangu 3.0 to compete with ChatGPT in August. While the competitors are rising in the cloud business by offering AI services to their customers, Alibaba is on a downslide. Possibly, the integration of LLM might be able to save the company.","excerpt":"Alibaba’s latest Qwen model can save its failing cloud business.","categories":["AI Features"],"tags":["competitions"],"author_name":"Mohit Pandey","publish_date":"2023-12-01T13:06:09","publication_year":"2023","word_count":851,"keywords":["Go","ChatGPT","API","competitions","AWS","cloud computing","AI","R","GPT","GAN","Azure"],"extracted_tech_keywords":["AI","ChatGPT","cloud computing","AWS","Azure","R","Go","API","GPT","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/alibabas-cloud-business-gets-qwen-ched\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10119721,"title":"Zebra Brings Generative AI to the Frontlines with Google, Android, and Qualcomm","content":"Zebra Technologies has announced a partnership with Google Cloud, Android, and Qualcomm to bring generative AI capabilities to frontline workers across industries. The collaboration integrates Zebra’s technological expertise with advanced AI from Google Cloud, hardware from Qualcomm, and software from Android. The new capabilities are designed to assist front-line employees by easing the cognitive load on these workers and helping them make better decisions in real time by providing a chat experience on their handheld devices. By harnessing generative AI with domain-specific knowledge, frontline staff will soon have access to a chat experience on their handheld devices. This will allow them to retrieve information and get answers to task-related queries easily. Tom Bianculli, Chief Technology Officer at Zebra Technologies, emphasised the conversation shift around generative AI, moving from ‘how’ it works to ‘what’ it can achieve. He envisions a future where planning and execution systems merge seamlessly, accelerated by finely tuned, real-world AI models capable of scheduling tasks, responding to requests, and providing context-based recommendations. A European supermarket chain has experienced this collaboration firsthand where, feeding the AI model with the company’s entire standard operating procedure (SOP) library, employees can now tap into a vast knowledge base derived from policies, procedures, and product information. This ‘always-on’ digital assistant has the potential to reduce time to competency, ensure consistent best practices, improve customer interactions, and enhance employee satisfaction. Rouzbeh Aminpour, Global Retail Solution Engineering Manager at Google Cloud, emphasises the fundamental change generative AI brings to organisations, fueling a new era of customer and employee interactions with businesses and brands. Apart from this partnership, Zerba also partnered with Qualcomm, showcasing how their phones and tablets could use a large language model (LLM) without needing connectivity to the cloud.","excerpt":"Zebra collaborated with Google, Android, and Qualcomm is about to brings Gen AI to Android phones and tablets.","categories":["AI News"],"tags":["Google Cloud"],"author_name":"Sagar Sharma","publish_date":"2024-05-06T18:35:55","publication_year":"2024","word_count":287,"keywords":["Go","Google Cloud","programming_languages:R","AI","ML","Git","ViT","generative AI","cloud_platforms:Google Cloud","GAN","R"],"extracted_tech_keywords":["AI","ML","generative AI","R","Go","Git","GAN","ViT","cloud_platforms:Google Cloud","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/zebra-brings-generative-ai-to-the-frontlines-with-google-android-and-qualcomm\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":53045,"title":"You Don’t Necessarily Need A Resume To Get A Data Science Job, Says This H2O.ai Data Scientist","content":"Receiving a data science job offer is getting increasingly difficult as more or less every applicant has the same skills. Consequently, aspirants fail to differentiate themselves from others, thereby diminishing the chances of receiving a job they desire. Apart from just the technical skills, data science aspirants and professionals need to plan effectively for increasing their likelihood of getting selected. However, applicants fizzle out due to the dearth of guidance. Therefore, Analytics India Magazine brings you the journey of a successful data scientist for assimilating their approach and spreading their perspective. For this edition of our weekly column My Journey In Data Science, we got in touch with Parul Pandey, a data science evangelist at H2O.ai. How It All Began Parul has graduated as an Electrical & Electronics Engineering from NIT Hamirpur in 2009. However, she mentioned that enrolling in one of the premier engineering institutes was essential than the stream. Post her graduation she worked as an Analyst in the planning department, where she was introduced to unveiling insights into the power distribution network in Delhi. “I have always been fascinated with numbers, but my real tryst with data science occurred in my first job,” adds Parul. There she learnt to investigate data and perform predictive analysis with SQL, which got her hooked into nuances of data science and machine learning. Preparation Strategy The data scientists first dabbled with C in her undergraduate and later after moving out from here the first job she started learning R and Python from various e-learning platforms such as Coursera and Edx. When asked about the approach, Parul said that she had devised a plan and followed it diligently to ensure compliance. Firstly, she was determined to complete the course before jumping on to another, and secondly, tried to implement the learning at her work. “Sometimes I was able to implement and solve the problem successfully and at times I couldn’t, but in the course, I learned a lot,” explains Parul. In this ever-changing data science landscape, preparing is never straightforward as many developments happen every day. And it wasn’t any different for Parul either. However, her ability to focus on one problem at a time enabled her in learning quickly. Parul’s Unique Job Strategy Unlike others, Parul didn’t directly apply for jobs. Her modus operandi of engaging in building a robust portfolio by adorning GitHub profile, writing on LinkedIn, Medium, and other publications like Towards Data Science, KDNuggets, and Datacamp, has helped her in getting a voice and visibility. Talking about the effectiveness of her strategy, she mentioned that she got the job offer at H2O.ai without even providing a resume. Parul believes that, in the future, resumes will go obsolete, communicating through writing on various platforms is crucial to demonstrate skills. “My first data science experience was very positive, and it was based on what I had accomplished in data science space,” clarifies Parul. “My presence in LinkedIn and Medium gave me a headstart since the recruiters knew me beforehand, the kind of projects I had executed, and what I was capable of.” Continuing on the job strategy, applicants shouldn’t lose self-esteem on getting rejected from employers and continue to improve on their skills for receiving a job offer they desire. Staying Relevant In Data Science Marketplace Over the years, data science has only evolved and became more competitive. Therefore, one needs to enhance their skills to stay relevant in the landscape. “Every day numerous libraries with algorithms afloat in the data science space, thereby, causing imposter syndrome,” acknowledges Parul. “However, as a workaround, she said one has to accept that no one can master all the skills in data science. Thus, we should determine our interest and ability and keep enhancing it.” Although Parul believes in improving her strengths, she didn’t downplay the idea of learning new things. The data scientist said, updating and reskilling is the mantra to remain relevant in the domain. Owing to this, she leverages Kaggle to learn new algorithms and machine learning strategies. Parul’s Data Science Achievements Parul became the LinkedIn Top Voice in the software development category due to her data science evangelism skills. She is one of the most popular influencers who contributes to the data science community through writing and sharing learning resources. Besides, she won an award in the 2019 Kaggle ML & DS challenge, where she analysed women’s representation in machine learning and data science. Determined to support other women in the data science, she has started the Hyderabad chapter of ‘Women In Machine Learning and Data Science’ (WiMLDS), where she encourages women to learn cutting-edge technologies and guides them to make a career switch. “A huge effort is still required in getting more women in the data science field, but things are improving, and we need to keep the momentum going,” says Parul. “It will also require a collaborative effort from society to make diversity and inclusion a vital part of the ecosystem.” Moving forward, Parul is committed to contributing and building an ecosystem for researchers in machine learning and artificial intelligence domain. With the hope of delivering such platforms, she is trying to collaborate with corporates and educational institutions. Advice To Aspirants Data science, in itself, is a great tool to solve real-world problems. Therefore, Parul says that aspirants and data science professionals should keep obtaining new skills and carry out meaningful projects. “While gaining technical skills is paramount, writing is the most underrated skill in data science. However, one should utilise their wiring skills to ensure that they are demonstrating their capability to the community,” concludes Parul.","excerpt":"Receiving a data science job offer is getting increasingly difficult as more or less every applicant has the same skills. Consequently, aspirants fail to differentiate themselves from others, thereby diminishing the chances of receiving a job they desire. Apart from just the technical skills, data science aspirants and professionals need to plan effectively for increasing […]","categories":["AI Features"],"tags":["Data Science","data science job","H2O AI","Interviews and Discussions"],"author_name":"Rohit Yadav","publish_date":"2020-01-03T12:00:00","publication_year":"2020","word_count":930,"keywords":["data science","artificial intelligence","machine learning","AI","data science job","ML","H2O AI","RAG","Python","analytics","SQL","Data Science","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","data science","analytics","RAG","Python","R","SQL"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/you-dont-necessarily-need-a-resume-to-get-a-data-science-job-says-this-h2o-ai-data-scientist\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":18298,"title":"GDPR and its implications on Data Management and Outsourcing","content":"With increasing emphasis on data security in the existing climate of data breaches and hacking, regulatory bodies are taking an increasingly demanding view on enterprise data security. The Global Data Protection Regulation (GDPR) is a new regulation (EU 2016\/679) through which the European Parliament and European Commission intend to strengthen and unify data protection for all EU citizens. The objective of GDPR is to give control back to EU citizens and residents over their Personal Data and simplify the regulatory environment. GDPR was adopted on 27 April 2016 and will become effective from May 25th, 2018. The impact of the regulations is far reaching not only for data subjects (individuals whose data is stored by government and private enterprises) but also for enterprises storing and processing personal data. The scope of GDPR is addressed in the next section and finally we cover the data security, data retention and outsourcing implications of GDPR. Scope of GDPR The GDPR regulation comes into effect from May 25th, 2018 and impacts any organization that is dealing with EU citizen data and has a project that is running beyond May 25th, 2018. Enterprises are under pressure to comply with these regulations as the fines are heavy and can go up to 4% of global annual turnover or 20 Million Euros. GDPR regulation can be classified into 5 key areas namely – Rights of EU Data Subjects, Security of Personal Data, Consent, Accountability of Compliance and Data Protection by Design and default. Enterprises also need to appoint a Data Protection Officer who is well versed with the GDPR guidelines and is responsible for monitoring and compliance of the business processes and systems that deal with personal data. We now examine the scope of GDPR from the following perspectives – Storage and processing of EU Citizen personal data – any enterprise that stores and processes personal data of EU citizens beyond May 25th, 2018 comes under the ambit of the regulation. A key question here is what constitutes personal data. The regulation states that personal data could be physical, genetic, cultural, social, economic, unstructured (e.g. Social media) and behavioral – derived and self-identified. The guidelines around storage are that personal data of data subjects should be stored only for the period deemed necessary. All processing operations on personal data must be documented and data must be processed only for the stated purposes. It is important to note that GDPR does not apply only to enterprises based in Europe and extends to global businesses that store and process EU citizen and resident personal data. Hence, some US Bank with European customers, must comply with the GDPR guidelines or face heavy fines. Data Controllers – data controllers control the personal data of data subjects for example a company stored personal data of employees or a bank stores personal data of its customers. Data controllers can appoint data processers to process personal data, however there need to be written instructions to ensure compliance to GDPR. Data Processers – they process personal data based on instructions provided by data controllers and must comply with the GDPR regulations. Data Subject – individuals (EU Citizens and residents) whose data is stored by data controllers and processed by data processers. Data subjects need to give their consent to enterprises about storage of their personal data and can withdraw the consent at any point of time. Data Breaches – in case of a data breach enterprises need to inform the regulatory authorities of the breach within 72 hours of the incident. Data Protection by Design – enterprises need to ensure that data protection guidelines are there in the design of data management processes and systems that would need to comply with GDPR. For instance, if you are planning a Data Mart hosted on a hybrid cloud of a 3rd party data processer, the personal data would need to be anonymized before usage by analytical applications and users. GDPR Implications on Data Security and Outsourcing With an overview of the scope of GDPR, we now assess the implications of GDPR on Data Security, Data Retention and Outsourcing – Data Security – All personal data stored and processed must be secure. If the personal data of EU data subject resides in data center outside the EU, the data must be anonymized so that persons are not identifiable. Even in case of remote access from outside the EU or if there are data movements to a site outside EU, the data in motion needs to be encrypted and anonymized. Data Retention – The data must be retained in Data Warehouses or Data Lakes only for the period defined for the use case\/purpose of data processing. In addition, there must be mechanisms to archive or delete the data once the data subject revokes the consent for storing personal data. This has a significant impact on the Data Governance and Data Management processes currently followed in enterprises which need to be reviewed in the wake of the GDPR regulations. Impact on Outsourcing – While there is no direct impact in terms of outsourcing to offshore data processers from a regulation standpoint, there are certain guidelines that need to be kept in mind. All offshore data processers need to follow the guidelines laid by data controllers around GDPR and any access to EU citizen data would involve anonymization of personal data. In case of data stored outside EU, in offshore data centers the personal data attributes would need to be anonymized. As enterprises grapple with compliance to GDPR guidelines, a good starting point is conducting an audit of business processes and systems that deal with personal data of EU data subjects, classify the data sets and put in the control mechanisms to ensure compliance to the regulations. Time is ticking, it’s time to act now!","excerpt":"With increasing emphasis on data security in the existing climate of data breaches and hacking, regulatory bodies are taking an increasingly demanding view on enterprise data security. The Global Data Protection Regulation (GDPR) is a new regulation (EU 2016\/679) through which the European Parliament and European Commission intend to strengthen and unify data protection for […]","categories":["IT Services"],"tags":[],"author_name":"saumyachaki","publish_date":"2017-10-12T04:44:12","publication_year":"2017","word_count":964,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","RAG","data governance","GAN","R","data lake","data warehouse"],"extracted_tech_keywords":["AI","RAG","R","Go","data warehouse","data lake","data governance","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/gdpr-implications-data-management-outsourcing\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10079523,"title":"10 Biggest Algorithmic Breakthroughs","content":"Since artificial intelligence keeps getting smarter with each passing day, its demand in several fields keeps getting bigger. The surge reflects the need for faster chips, more data, and definitely better algorithms. Shifting our focus, here are the top algorithmic breakthroughs that have become essential in the modern AI developer’s toolbox. HyperTree Proof Search (HTPS) Meta AI showed its contributions to the challenging area of deep learning with HTPS, a deep learning model that solved several International Math Olympiad (IMO) problems. This method showcases important capabilities, demonstrating that deep neural networks make mathematical reasoning possible. The new algorithm combines reinforcement learning and Monte Carlo tree search to show unique levels of mathematical reasoning. DeepNull Earlier this year, researchers at Google put forward DeepNull to model the relationship linking covariate effects on phenotypes and improve Genome-Wide Association Studies (GWAS). GWAS are associated genetic variants with complex traits and diseases. DeepNull models have a nonlinear effect of covariates on phenotypes as they are simple to use and require the least amount of change for current GWAS pipeline implementations. The interactions between phenotypes like age and sex and principal components of genotypes must be adjusted for covariates to determine the association strength between genotype and phenotype. ESMFold Meta AI’s launch of Evolutionary Scale Modeling (ESM) became one of the biggest competitors or the best alternative to AlphaFold 2. Much like AlphaFold, the model is also open to the public. But that’s not the only one. Read about the top protein models released in 2022 here. Code as Policies (CaP) Google AI introduced a new learning algorithm for effective robotic control. The algorithm proposes letting robotic systems effectively write their code. The concept aims to save developers from having to go in and reprogram things every time new information is found. The primary idea of the algorithm is to leverage predictive information to obtain a representation of the high-dimensional environment dynamics. DiaBeats To tackle the issue of diabetes, a group of Indian researchers from the Lata Medical Research Foundation in Nagpur developed an AI algorithm that can predict diabetes and pre-diabetes from individual heartbeats recorded on an ECG (electrocardiogram). Dreamer Algorithm Researchers from the University of California, Berkeley, taught robots to walk within 60 minutes. The approach differs from the usual deep reinforcement learning practices, as robots can be trained without simulators. The project ‘DayDreamer: World Models for Physical Robot Learning’ uses the Dreamer algorithm to learn from small amounts of interaction through planning in a learned world model. In turn, it outperforms pure reinforcement learning in    video games. IISc-AIIMS Rishikesh’s epilepsy algorithm Indian Institute of Science (IISc) researchers, in collaboration with AIIMS Rishikesh, developed an algorithm to decode brain scans in order to identify the occurrence and type of epilepsy. In the study, the team reported an algorithm that can identify signs of epilepsy from the electrical signal patterns. After initial training, the researchers say that the algorithm could detect whether a human subject could have epilepsy based on the patterns in their analysis. SEER Facebook unveiled SEER, a self-supervised AI model that can learn from a random set of unlabeled images on the web. Though it is early, the team expects it to become a computer vision “revolution”. SEER was fed on a billion publicly available Instagram images, not manually curated. As a result, even without the labels and annotations, it could autonomously work through the dataset, learn, and achieve accurate results for detecting objects. DeepCTRL Google Cloud AI researchers released DeepCTRL (Deep Neural Networks with Controllable Rule Representations), which combines an encoder and a rule-based objective in the model that allows decision-making. Data type and model architecture are unimportant to DeepCTRL. The highlight of DeepCTRL is that it does not require retraining to alter rule strength – the user may adjust it at inference based on the desired accuracy vs rule verification ratio. data2vec Meta AI released data2vec, calling it “the first high-performance self-supervised algorithm that works for multiple modalities”. This algorithm can be applied separately to speech, text and images, outperforming the previous best single-purpose algorithms for computer vision and speech. It represents a new paradigm of self-supervised learning, where new research improves multiple modalities rather than just one. The algorithm will enable users to develop more adaptable AI, performing tasks beyond today’s systems. Meta said that data2vec does not rely on contrastive learning or reconstructing the input example. The tech giant has also released open-source code and pretrained models.","excerpt":"Eye-catching advancements that made a buzz in the AI community","categories":["AI Trends"],"tags":["data2vec","ESMFold"],"author_name":"Tasmia Ansari","publish_date":"2022-11-13T10:00:00","publication_year":"2022","word_count":740,"keywords":["Go","Meta AI","artificial intelligence","ESMFold","AI","neural network","computer vision","RAG","data2vec","Aim","deep learning","R"],"extracted_tech_keywords":["AI","artificial intelligence","deep learning","neural network","computer vision","Meta AI","Aim","RAG","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/10-biggest-algorithmic-breakthroughs-of-2022\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10120136,"title":"Zoho’s Sridhar Vembu Joins Ola CEO Bhavish Aggarwal’s Rant Against LinkedIn and Microsoft’s ‘Wokeness’","content":"Zoho chief Sridhar Vembu said that he agrees with the opinion of Ola chief  Bhavish Aggarwal that India needs to resist the West’s woke imperialism. “I agree with Bhavish. We in India have to strongly resist this woke imperialism – it is best understood as a fanatical religious doctrine that masquerades as a socio-political movement,” replied Vembu to Aggarwal’s lengthy post on X where he expressed his concerns about the dominance of Western companies in the tech industry and their impact on Indian society. This comes after LinkedIn deleted Aggarwal’s post where he pointed out that the platform’s AI chatbot used wrong pronouns for him which Aggarwal referred to as ‘Pronoun Illness’ Aggarwal criticised the Western concept of “pronouns” and its imposition on Indian society. He expressed his concerns about the impact of Western social media platforms on Indian culture and the need for Indian tech companies to build their own platforms. “The pronouns issue is a woke political ideology of entitlement that doesn’t belong in India,” Aggarwal wrote. “I wouldn’t have waded into this debate but clearly LinkedIn has presumed Indians need to have pronouns in our life, and that we can’t criticise it.” Aggarwal took a strong stance against the Western concept of gender inclusivity, which he believes is being imposed on Indian society. “We don’t need lectures from Western companies on how to be inclusive,” Aggarwal wrote, saying Indian culture didn’t need pronouns to be inclusive for thousands of years. Aggarwal also highlighted popular ride-hailing service’s efforts to promote diversity and inclusivity in the workplace. He noted that Ola has one of the largest women-only automotive plants in the country, with almost 5,000 women employees and plans to expand to tens of thousands in the coming years. He argued that Indian culture has always been inclusive and respectful of all genders, citing the example of transgenders being accorded special respect in ancient Indian society. Ola Plans to Stop Using Microsoft Azure Furthermore, Aggarwal, announced that Ola has decided to move its entire workload out of Azure to  their own Krutrim cloud within the next week. “Since LinkedIn is owned by Microsoft and Ola is a big customer of Azure, we’ve decided to move our entire workload out of Azure to our own Krutrim cloud within the next week. It is a challenge as all developers know, but my team is so charged up about doing this,” wrote Aggarwal. He further offered a full year of free cloud usage to any developer who wants to move out of Azure and wants to join Krutrim AI cloud. Besides Vembu, Unacademy chief Gaurav Munjal also supported Aggarwal’s stance. “This is so fucked up on LinkedIn’s part,” replied Munjal. “Indian civilization has a long history of inclusiveness regarding gender. However, the woke alphabet ideology promoting a pseudo-scientific view of gender is in conflict with both science and Indian ethos, “ wrote  BJP leader Tejasvi Surya. However, not all agreed. “Does he think people who deploy and run companies’ cloud or work in AI are as stupid as consumers who buy two wheelers?” wrote Kissan AI founder Pratik Desai. “Why does it sound like a pre-planned marketing campaign for launching Krutrim cloud? You can’t have these things done overnight for sure. Good way to grab attention, wrote  another user on X. lol so Ola cabs actually runs on AWS not on Azure. https:\/\/t.co\/v8lnVClhtr— Arnav Gupta (@championswimmer) May 11, 2024","excerpt":"Aggarwal aims to build an Indian social media framework, alongside moving Ola’s workload out of Azure, and offers free cloud usage for developers leaving Azure.","categories":["AI News"],"tags":["Ola Krutrim","zoho"],"author_name":"Siddharth Jindal","publish_date":"2024-05-11T22:44:40","publication_year":"2024","word_count":567,"keywords":["Go","zoho","cloud_platforms:Azure","AWS","AI","cloud_platforms:AWS","R","cloud_platforms:Microsoft Azure","programming_languages:R","ViT","Ola Krutrim","Azure"],"extracted_tech_keywords":["AI","AWS","Azure","R","Go","ViT","cloud_platforms:AWS","cloud_platforms:Azure","cloud_platforms:Microsoft Azure","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/zohos-sridhar-vembu-joins-ola-ceo-bhavish-aggarwals-rant-against-linkedin-and-microsofts-wokeness\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":7329,"title":"How Mainframe Data Virtualization is Bringing Data Closer to Analytics","content":"Data, in all of its various forms, has completely transformed our society. And, it’s not just one kind of data that’s responsible for this; it’s all kinds, whether it’s streaming data, Big data, legacy data, or other kinds. Having a comprehensive means of leveraging this data to your business’ best interests is vitally important in today’s world, as it is the only viable way to effectively identify new opportunities, mitigate risks, and to meet customer demands. To be sure, the kinds of data that must be dealt with are always expanding. For example, machine-to-machine data has grown in importance, thanks to technological advances like RFID tags. In addition to this, government regulations have necessitated that more forms of data are kept by businesses, particularly those that operate within the financial services sector. There is a name for all of the unstructured data that we’re talking about, and that name is “Big Data”. While this form of data has received an incredible amount of focus from businesses around the world, there’s another form that deserves just as much if not more focus: mainframe data. Those with technological acumen will be quick to acknowledge that mainframe data is essentially the original form of Big Data. This mainframe data is vitally important, as it often is corned with vital business functions, for example the control of inventory and billing, or the control of tax records and transactions. Banks, in particular, understand how important mainframe data is, as their mainframes must accurately, reliably, and quickly conduct a tremendous number of transactions around the clock. This all serves to illustrate how important it is to have a comprehensive solution for mainframe data, especially if you’re concerned with the effectiveness of your business intelligence and analytics strategies. In order for such strategies to be effective, data must be moved as closely as possible to analytics and BI tools. Further, all the different forms of a data that a business has access to must be blended together seamlessly and effectively. Accomplishing this, however, means that old methods of physically moving data must be eliminated. There are other technical challenges, as well. For one, the data that is being dealt with must be integrated and standardized in such a way that ensures consistency across both business-facing and customer-facing applications. Also, the data must be positioned in such a way as to facilitate easy, quick and accurate access, regardless of who is accessing it. For the most part, businesses have deployed the ETL method to deal with data in this manner. However, this method has proven to be woefully deficient for needs of today’s business environment. By physically moving data, the ETL method contributes to a high degree of latency when working with the data being transformed. Further, the process of transformation that such a method employs often introduces inconsistencies and fails to enforce strict standardization. Ultimately, however, the greatest flaw of this method is that it negates the timeliness of data. Because the data must be moved and transformed in separate processes, the resulting transformed data is instantly out of date when it is accessed through analytics and BI tools. The way around this problem is mainframe data virtualization. This method employs specialty processors – IBM System Z processors – within a mainframe that take care of the data processing, integration, and transformation. Because of this, a mainframe’s central processors do not need to be employed for these tasks. This, in turn, leads to a reduction in a mainframe’s TCO, leaves data undisturbed, and eliminates the cost of software license charges. Naturally, the problems of inaccuracy, lack of timeliness and latency encountered with ETL methods are eliminated as well. In the end, mainframe data virtualization allows for data to be moved as close to analytics as possible. This allows those that drive business to make the most informed decisions possible, decisions that are backed by current and accurate analytics that are derived from current and accurate data. In addition to this, ready access to such data can facilitate connections between the different components of a business, for example its people, process and systems, making it possible for all of these different components to work in tandem. In the end, this allows a business to accomplish its ultimate goals: meeting the demands of its customers, meeting potential challenges and competitors in the marketplace, and identifying new opportunities for growth and expansion.","excerpt":"Data, in all of its various forms, has completely transformed our society. And, it’s not just one kind of data that’s responsible for this; it’s all kinds, whether it’s streaming data, Big data, legacy data, or other kinds. Having a comprehensive means of leveraging this data to your business’ best interests is vitally important in […]","categories":["AI Features"],"tags":["Big Data"],"author_name":"AIM Media House","publish_date":"2015-04-27T08:10:44","publication_year":"2015","word_count":732,"keywords":["big data","Go","business intelligence","AI","ETL","ML","RAG","ViT","analytics","Big Data","R"],"extracted_tech_keywords":["AI","ML","analytics","RAG","R","Go","big data","ETL","ViT","business intelligence"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-mainframe-data-virtualization-is-bringing-data-closer-to-analytics\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":27574,"title":"Deep Learning Is Not So Much A Black Box Anymore, And That’s A Great Development","content":"Image Source: MIT Technology Review Although deep learning models have made remarkable progress in a range of tasks such as image recognition, speech recognition and language translation, the interpretability of models has been the subject of various research papers. Netherlands-based company Riscure which specialises in security services for connected and embedded devices, defines DL as an intelligent algorithm used to analyse large data sets and identify patterns using a deep neural network. In this case, the results are achieved by training a network on a data set with a known result (generally a number of classes that has to be identified in a data set). This trained neural network is further applied to a new data set to extract unknown features and classify them. One of the most common examples of DL algorithm in industries today is an image recognition system used for identifying objects which is trained on a large set of photos. In this case, the target and algorithm identify the necessary properties in the image themselves. According to one DL practitioner, deep networks are performing better than expected in a wide range of tasks and neural networks are very powerful because they can be arbitrarily extended. The networks generate an enormous function space and then gradient descent finds a suitable choice in that function space. So, with enough data and computing power, these models can follow human-understandable logic or procedures and make accurate predictions. However, model interpretability — identifying how a model makes a prediction is an area that has gained considerable interest amongst researchers. With DL finding more commercial applications, a lot of academicians have trained their eyes on decoding its black box aspect. Even though there are many models to understand interpretability in terms of predictive features, developers usually want to isolate a small set of training examples which can have a high impact on prediction. But as one Redditor points out, it is not necessary that each training example contributes to a prediction. There has been considerable progress in several key areas, like, understanding what features neural nets learn, why GANs or autoencoders can learn all features instead of label-dependent ones, or why they generalise well even without regularisation (it was already shown that SGD acts as an implicit regulariser via inductive bias). A Look At Recent Research Papers Which Break Down Algorithmic Explainability: DeepBase: Another brick in the wall to unravel black box conundrum, DeepBase is a system that inspects neural network behaviours through a query-based interface. The paper presents DeepBase, a method to analyse recurrent neural network models, and propose a set of simple and effective optimisations to speed up existing analysis approaches by up to 413 times. As cited in the paper, the researchers grouped and analysed different portions of a real-world neural translation model and show that learns syntactic structure, which is consistent with prior NLP studies, and can be performed with only three DeepBase queries. CapsNet Proposed By Hinton: According to O’Reilly, Geoffrey Hinton from the University of Toronto first introduced CapsNets in 2011 in his research paper titled Transforming Auto-encoders. It proposed to overcome the shortcomings in traditional Convolutional Neural Networks (CNNs). CNNs are trained on huge amount of data and in this process, the deep networks discover patterns and not hierarchies, whereas CapsNets are designed to incorporate hierarchies and train on less data. Hinton’s paper said that CNNs are misguided and neural networks should use local “capsules” that perform complicated internal computations on their inputs and then encapsulate the results of these computations into a small vector of highly informative outputs. CapsNet is also more transparent and interpretable than CNNs and indicates how a feature was identified in each layer. Classifying Research in Explainability of AI While the volume of research in explanatory AI has exponentially expanded, the explanation has been largely concentrated in two areas — explaining the representation of data inside a network or the processing of data by the network. Another key approach has been designing explanation producing systems with architectures that simplify the interpretation of their behaviour. So far the research can be classified under these sections: Research done to explain black box models Research carried out to explain the black box outcomes New methods designed for transparent systems Outlook For AI systems to gain wider acceptance, trust and commercial use, it is imperative to provide satisfactory explanations of decisions and there has been continuous research in establishing trust in DL systems, understanding the need for transparent explanations, insights into the decision making process of neural network. DL community has also made considerable headway in the development of FAT (Fair, Accountable and Transparent) algorithms, focused on interpretability of methods used to achieve the output.","excerpt":"Although deep learning models have made remarkable progress in a range of tasks such as image recognition, speech recognition and language translation, the interpretability of models has been the subject of various research papers. Netherlands-based company Riscure which specialises in security services for connected and embedded devices, defines DL as an intelligent algorithm used to […]","categories":["AI Features"],"tags":["GANs"],"author_name":"Richa Bhatia","publish_date":"2018-08-23T14:04:59","publication_year":"2018","word_count":781,"keywords":["Go","TPU","Rust","AI","neural network","image recognition","NLP","deep learning","GANs","GAN","R"],"extracted_tech_keywords":["AI","deep learning","neural network","NLP","image recognition","TPU","R","Go","Rust","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/deep-learning-is-not-so-much-a-black-box-anymore-and-thats-a-great-development\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10139346,"title":"Sarvam AI Launches Sarvam-1, Outperforms Gemma-2 and Llama-3.2","content":"Indian AI startup Sarvam AI has launched Sarvam-1, the first LLM optimised specifically for Indian languages. Developed with 2 billion parameters, Sarvam-1 supports 10 major Indian languages—Bengali, Gujarati, Hindi, Kannada, Malayalam, Marathi, Oriya, Punjabi, Tamil, and Telugu—alongside English. ​​Despite its relatively smaller size, Sarvam-1 shows strong performance in Indic language tasks, outperforming larger models like Gemma-2 and Llama-3 on benchmarks such as MMLU, ARC-Challenge, and IndicGenBench. It also offers faster inference speeds—4 to 6 times faster—making it suitable for deployment on edge devices. For instance, on the TriviaQA benchmark, Sarvam-1 achieved an accuracy of 86.11 across Indic languages, significantly surpassing Llama-3.1 8B’s score of 61.47. Sarvam-1’s performance on the IndicGenBench, which tests cross-lingual tasks such as summarization, translation, and question answering, also stood out. It achieved an average chrF++ score of 46.81 on Flores, a dataset for English-to-Indic translation, surpassing the larger Llama-3.1 8B model. The model bridges the gap for Indian language speakers by offering advanced natural language processing (NLP) capabilities that have previously been centered around English and other high-resource languages. A key feature of Sarvam-1 is its efficiency in handling Indic scripts, a major challenge in previous LLMs. Most existing multilingual models have high token fertility—meaning they require more tokens per word for Indian languages compared to English. Sarvam-1’s tokenizer significantly reduces this inefficiency, achieving fertility rates of 1.4 to 2.1 tokens per word, much closer to the 1.4 tokens needed for English. This enables more streamlined training and better model performance across Indian languages. The model’s training corpus, Sarvam-2T, consists of approximately 2 trillion tokens, with content evenly distributed across the 10 supported languages, except for Hindi, which constitutes about 20% of the dataset. The dataset also includes a substantial portion of English and programming languages, which helps the model perform across both monolingual and multilingual tasks. Sarvam-2T emphasises high-quality, diverse data, addressing limitations in existing Indic datasets like Sangraha, which are often web-crawled and lacking in quality. Sarvam-2T includes longer documents and richer scientific and technical content, enhancing the model’s ability to handle complex reasoning tasks. Another key feature of Sarvam-1 is its computational efficiency. The model offers 4 to 6 times faster inference speeds compared to larger models like Gemma-2-9B and Llama-3.1-8B, while maintaining competitive performance levels. This makes Sarvam-1 particularly suitable for deployment in production environments, including edge devices where computing resources may be limited. Sarvam-1 was trained over five days using 1,024 GPUs on Yotta’s Shakti cluster, leveraging NVIDIA’s NeMo framework for training optimisations. The model is available for download on Hugging Face’s model hub, where developers can access and explore its capabilities for a range of Indic language applications, from translation to conversational AI and more.","excerpt":"Developed with 2 billion parameters, Sarvam-1 supports 10 major Indian languages—Bengali, Gujarati, Hindi, Kannada, Malayalam, Marathi, Oriya, Punjabi, Tamil, and Telugu—alongside English.","categories":["AI News"],"tags":["sarvam ai"],"author_name":"Siddharth Jindal","publish_date":"2024-10-24T21:20:51","publication_year":"2024","word_count":445,"keywords":["Hugging Face","sarvam ai","AI","RPA","ML","RAG","NLP","ai_frameworks:Hugging Face","llm_models:Llama","R","startup"],"extracted_tech_keywords":["AI","ML","NLP","Hugging Face","RAG","R","RPA","startup","llm_models:Llama","ai_frameworks:Hugging Face"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/sarvam-ai-launches-sarvam-1-outperforms-gemma-2-and-llama-3-2\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10011110,"title":"This AI Tool Corrects Gender Bias In Portrayal Of Females In Movies","content":"A 2019 study coined a very interesting term — ‘the Cinderella complex’. The authors of this study analysed 7226 books, 6000 movie synopsis, and 1,100 movie scripts and found that the words used to associate with the male and female characters reeked of gender bias. The lives of the male characters were adventure and aspiration-oriented, whereas the female characters were more passive and romantic-relationship oriented. This is just one of the countless studies that show how gender bias, a lot of time unintentional and a product of societal conditioning, creeps in popular text and media. This, in turn, sets an incorrect narrative. Keeping this in view, the researchers at the Allen Institute for Artificial Intelligence in collaboration with the University of Washington created an AI-based tool that rewrites text to correct potential gender bias in character portrayals. Christened PowerTransformer — an encoder-decoder model developed based on a pre-trained language model. How Does It Work The narrative in popular media often assigns a stereotypical colour to gender roles. This problem is widely recognised, and there have been several attempts at removing such biases. One such method is controllable text revision which rephrases text to a targeted style or framing. The conventional controllable text revision has to overcome three main challenges — editing beyond just surface-level paraphrasing, the revision should not make unnecessary changes to the underlying meaning of the text, and models must learn to debias and learn text without any supervised data, thus preventing straightforward machine translation-style modelling. To overcome above-listed challenges, the researchers at Allen Institute for Artificial Intelligence and the University of Washington jointly formulated a new controlled text revision task called controllable debiasing that studies portrayal biases through connotation frames of power and agency, which capture knowledge about the implied power dynamics with respect to verbs. The researchers have introduced a new controllable debiasing approach called PowerTransformer. In this approach, reconstruction and paraphrasing objectives are combined to overcome the lack of parallel supervised data. The PowerTransformer model uses connotation frame knowledge both at training time using control tokens, and during the generation, time using agency-based vocabulary boosting. Further, this model uses an OpenAI-GPT transformer model as the base. PowerTransformer is similar to the GD-IQ tool that was developed by the University of Southern California Viterbi School of Engineering. GD-IQ uses AI techniques to analyse the text of a script to identify the number of males and females and whether they represent the population at large. PowerTransformer introduces an improvement over GD-IQ by rephrasing text using machine learning. For example, ‘Alice daydreamed about a doctor’ is rewritten as ‘Alice pursued her dream to be a doctor’. Such rephrasing gives the character more authority. After experimenting, the researchers studied 16,763 characters from 767 modern English movie scripts and found that of these characters, 68% were inferred to be men and the remaining 32% women. The researchers attempted at mitigating gender biases in these portrayals by attempting to re-balancing the agency level of female characters to be at par with male characters using PowerTransformer. The model proved to be successful in increasing the positive agency and decreasing the passiveness associated with the female characters. “Our findings show promise for using modern NLP tools to help mitigate societal biases in text,” noted the researchers. Additionally, they also cautioned that this was a pilot study, and the model would still require human intervention in automatically rewriting the entire movie. Wrapping Up This tool has the potential to help authors and scriptwriters in writing stories or movie plots by providing different framings for alternative portrayals of characters. This could conclusively help in stereotypical portrayals of females and debunk gender roles in society which is heavily influenced by media.","excerpt":"A 2019 study coined a very interesting term — ‘the Cinderella complex’. The authors of this study analysed 7226 books, 6000 movie synopsis, and 1,100 movie scripts and found that the words used to associate with the male and female characters reeked of gender bias. The lives of the male characters were adventure and aspiration-oriented, […]","categories":["AI Features"],"tags":["gender bias","power bi training"],"author_name":"Shraddha Goled","publish_date":"2020-11-03T15:00:46","publication_year":"2020","word_count":615,"keywords":["machine learning","artificial intelligence","OpenAI","AI","NLP","Ray","GPT","ViT","R","power bi training","gender bias","llm_models:GPT"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","NLP","OpenAI","Ray","R","GPT","ViT","llm_models:GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/this-ai-tool-corrects-gender-bias-in-portrayal-of-females-in-movies\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10070593,"title":"What makes a feature interpretable","content":"Recently, MIT researchers published a paper, The Need for Interpretable Features: Motivation and Taxonomy, highlighting the importance of improving the feature interpretability in machine learning models to understand their outputs better. In machine learning, ​​interpretability refers to the degree to which a model can be understood in human terms. Interpretable ML models help decision-makers understand why the model predicted a certain outcome. For example, the creditworthiness of a customer. In addition, interpretability can be a useful debugging tool for detecting bias in machine learning models. In their formal literature review, the researchers found that although there is wide-spread agreement that interpretable features are important for various reasons, there is little work formalizing what makes a feature interpretable, and almost no work quantifying the interpretability of features. The team found that “interpretable feature” is often defined as one that was human-generated or worded using human-readable language. Lessons from real-world domains To understand machine-learning usability challenges, the researchers worked with experts (who generally do not have ML expertise) in five real-world domains. “We found that out in the real world, even though we were using state-of-the-art ways of explaining machine learning models, there is still a lot of confusion stemming from the features, not from the model itself,” said Alexandra Zytek, lead author of the paper. In one case, the researchers conducted formal user studies, where child welfare screeners were provided with feature contribution explanations alongside the predicted risk score to aid them in their screening decision making. The team found that most confusion and distrust in the model arose from the features, despite them being interpretable in terms of the usual definitions — they were hand-selected by humans, and presented in readable, natural language. The main challenges were caused by features being worded confusingly, or being seemingly unrelated to the prediction target according to users. In another case, the researchers built ML models using Electronic Health Record (EHR) data to predict the probability of patients facing complications after cardiac surgeries. In this case, although the model considered features like the trend of the patient’s heart rate over time to make the prediction, doctors were interested to know how strongly the patient’s heart rate data influenced that prediction. “With interpretability, one size doesn’t fit all. When you go from area to area, there are different needs. And interpretability itself has many levels,” said Kalyan Veeramachaneni, co-author and principal research scientist in the Laboratory for Information and Decision Systems (LIDS). Taxonomy of ML features The study highlights the need for an interpretable feature space that closely aligns with real-world and human cognition. An interpretable feature space enables users (with no ML expertise) to make decisions. Based on real-world experiences across domains, the researchers suggested a taxonomy of feature properties that can make features more or less interpretable for different decision makers. They also highlighted which properties are likely most important to particular users. For example, if the goal is to improve model performance, then features should be in a form that is model compatible and statistically correlated to the target variable. However, if the goal is to enable better decision making, the features should be described so that users can understand and draw conclusions from them. The researchers identified five categories of users— developers, theorists, ethicists, decision makers and subjects impacted by machine-learning model’s predictions. Then, they defined properties that make features interpretable for each category. They also provided guidelines on how developers can transform features into simpler formats for laypersons to understand by using what they call “interpretable transforms”. Feature transforms Traditionally, most work on feature engineering has focused on the model-ready feature state. These transforms generally convert the data to a form that is compatible with the model; or improve performance by factoring in an understanding of the domain. Model-ready transforms will often reduce the interpretability of features. For example, in the case of child welfare, one-hot encoding (a technique of converting data to prepare it for an ML algorithm and get a better prediction) reduces the “human- wordiness” of the features. In such cases, interpretable transforms come into play. The screening requires features that are meaningful, understandable, and human-worded. The properties can be achieved in part through feature selection and semantic binning, while ensuring features presented are categorical and unstandardized. Converting data into categorical form improved the cardinality of the feature. In addition, semantic binning improved the interpretability of the features by allowing users to refer to age categories to make their decisions. The age categories were binned based on stages of child development like infant, toddler, teenager etc. Similarly, in the case of healthcare, features need to be simulatable for doctors to understand how strongly the patient’s heart rate data influence the prediction of the ML model and provide treatment accordingly. To meet this, researchers connected the features to raw data, i.e. they used a patient’s pulse signal rather than the feature MEAN(pulse). Although a model may require interpretable features to be useful, transforming to the interpretable feature space always carries some risk. Some transformations can bias the explanation. At times, incorporating only those features that are interpretable can prevent the model from predicting unexpected patterns that could be true. Therefore, developers need to decide consciously when selecting features.","excerpt":"With interpretability, one size doesn’t fit all.","categories":["AI Features"],"tags":[],"author_name":"Zinnia Banerjee","publish_date":"2022-07-07T12:00:00","publication_year":"2022","word_count":870,"keywords":["Go","interpretable ML","machine learning","TPU","programming_languages:R","AI","ML","feature engineering","Rust","R"],"extracted_tech_keywords":["AI","machine learning","ML","TPU","R","Go","Rust","feature engineering","interpretable ML","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/what-makes-a-feature-interpretable\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":22626,"title":"Should India Take A Leaf Out Of China&#8217;s AI Playbook?","content":"India and China are in logger-head to capture the leadership of the emerging nations and though it has surpassed China to regain its status as the world’s fastest-growing major economy but India has miles to go before it seriously challenges Chinese leadership in terms of AI. When it comes to AI, machine learning and data analytics, Beijing is a way ahead of India. It has become one of the leading global hubs for AI development. No country is moving faster than China in harnessing the artificial intelligence to power its industries. Today China has more data than anywhere else in the world. China’s Ambitious Plan China now boasts about home-grown IT talent to develop its own technologies without seeking help from outside. On July 2017, China said it expects to become the AI superpower by 2030. Xi Jinping-led government is investing billions of dollars in AI. On the other hand, Chinese companies are not only pouring money into AI startups in Silicon Valley but also heavily investing in local AI development. The country is using artificial intelligence and facial recognition technology in a wide range to monitor its citizen as they travel or shop and even use toilet papers. The country has built the world’s largest video surveillance network with millions of cameras that are capable of automatically identify people’s basic information within seconds. There are 170 million surveillance cameras in China and by 2020, the country hopes to have one camera for every two citizens. At the same time, the country is also building the world’s’ most powerful facial recognition system with the power to identify any one of its 1.3 billion citizens within three seconds. The big data surveillance is used by police and private companies to track its citizen’s movement, predicting crime and boosting national security. Hi-Tech Powerhouse China’s aim to become hi-tech innovation powerhouse have prompted local government and police departments to embrace facial recognition technology as an important tool for surveillance and public safety efforts. Interestingly, the police officers in China’s Hanan province use facial recognition smart glasses to scan people and spot suspected criminals and even jaywalkers within 100 milliseconds. The introduction of facial recognition software has been made possible due to country’s massive database of citizen’s ID photos. Chinese companies can freely tap into the data of China’s 750 million internet users, match their personal IDs and photos to train facial recognition algorithms and other neural networks. Major businesses and tech companies have also adopted facial recognition technology in retail, travel and banking sectors. With a smile to a camera, people can pay for a meal at a KFC outlet, travellers can board planes, students can enter their university halls, drivers can fill their car or bike at petrol station, customers of China Merchants Bank can withdraw money without using their bank cards and shoppers can enter Hotan bazaar in Xinjiang. China facial recognition could surpass other personal identification methods that are used to make payments, claims a report. Given its 1.42 billion population and diverse industry mix, the nation can generate huge volumes of data and provide an enormous market. Not only the nation’s biggest tech companies are making significant R&D investments in AI but the government have also made AI-enabled education a national priority. Education startups are increasingly looking to AI to upend traditional classrooms and provide higher quality education. For children in rural areas, China’s Education Ministry has directed government at all levels to spend 8 per cent of their annual funding on digitalisation of education. To further its ambitious plan, the country is planning to build a $2.1 billion technology park for high-speed big data, cloud computing, biometrics and deep learning. It will also have 5G mobile internet, supercomputer and cloud services. It will be a home to around 400 businesses and is expected to create an annual output of about 50 billion yuan. It will also include a national-level artificial intelligence lab. AI Edge To have a military edge over US, China’s People’s Liberation Army has started investing in an AI-related project. Reportedly, Chinese military boffins are incorporating AI into their latest cruise missiles. A senior designer from China’s Aerospace and Industry Corp said that they are planning to adopt a ‘plug and play’ approach in the development of new cruise missiles, which will enable our military commanders to tailor-make missiles in accordance with combat conditions and their specific requirements.” China’s heavy investment in AI can contribute to its economic growth as well. According to an Accenture report, AI has the potential to add as much as 1.6 per cent points to China’s economic growth rate by 2035. PricewaterhouseCoopers projected that by 2020 AI will give a $15.7 trillion boost to global GDP and China will be the biggest beneficiary, with 26 percent added to its national GDP. Chinese Lesson for India As China is making a rapid progression in the field of artificial intelligence and machine learning, it is imperative for India to heavily invest in the technology and view AI as the critical element for country’s development. The story of China’s internet giant Baidu holds the most important lesson for India, which has successfully developed an AI-powered search engine on its own. Baidu is investing heavily in deep speech for voice-based searches that leverage speech recognition. The AI in Baidu understands and interprets queries in Beijing dialect mandarin rather than English. Baidu’s heavy investment in AI-related research exposes the lack of India’s technological infrastructure. China has recognized the importance of cultural infrastructure, which bridges the gap between the lab and the market. India should not only recognize the importance of the cultural institutions required for such innovation but also requires an ecosystem of high-tech product industry that focuses on research and innovation, which will strength our AI talent pipeline and help India take the lead in cutting-edge AI innovation. India offers the same advantages as China in terms of data (over 99 percent of Indians aged 18 and above have Aadhaar)  but the major drawback is that India doesn’t have its own tech giants like Baidu or Google that can use the data to build a credible AI system. How India Can Evolve Indian IT industry has to figure out how it will evolve in AI-driven market. They need to rethink their old business model and focus more on transformational IT services. They should train a whole set of students to think independently and take new challenges. As rightly said by Niti Aayog CEO Amitabh Kant, “India needs massive upgradation in new technologies. IITs and IIITs must redefine themselves as institutes driving cutting-edge technologies for fourth industrial revolution.” He also said, “India must realign education system to emphasise skills rather than mere degrees. We must move away from the Anglo-Saxon system of education with an emphasis on academic degrees, toward hands-on learning in practical subjects. Additionally, Indian colleges should partner with the IT industry to introduce artificial intelligence and machine learning concepts or create vocational courses. Within companies, HR and learning departments should motivate people to take up AI courses while they are off projects. On a Concluding Note Though the country is lagging behind in the AI innovation, but experts believe that India has an opportunity to create a locally designed AI plan that would work for its need and help nurture world-dominating companies. “India could have a very large role to play, all the pieces are in place, it has a great opportunity, and is up to India to either succeed or fail. I have seen some promising ideas from India in the area of healthcare, education, but a lot of hard work has to be done in building the technical capability because AI products require high levels of accuracy,” Andrew NG said in a report. Despite Modi government’s push for artificial intelligence, in order to get anywhere close to competing with the global power, both AI industries and research in India will need a big boost. Amitabh Kant rightly said in a report that, “We must initiate measures to ensure that Indians are fully prepared to embrace the new era of AI, blockchain and emerging technologies. And this requires a new mindset. Our policies must drive this change.”","excerpt":"India and China are in logger-head to capture the leadership of the emerging nations and though it has surpassed China to regain its status as the world’s fastest-growing major economy but India has miles to go before it seriously challenges Chinese leadership in terms of AI. When it comes to AI, machine learning and data […]","categories":["IT Services"],"tags":["AI (Artificial Intelligence)","China","IT industry","NITI Aayog"],"author_name":"Smita Sinha","publish_date":"2018-03-14T11:26:46","publication_year":"2018","word_count":1365,"keywords":["artificial intelligence","IT industry","machine learning","AI","neural network","cloud computing","RAG","Aim","China","deep learning","analytics","edge AI","AI (Artificial Intelligence)","NITI Aayog"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","deep learning","neural network","analytics","Aim","edge AI","RAG","cloud computing"],"url":"https:\/\/analyticsindiamag.com\/it-services\/should-india-take-a-leaf-out-of-chinas-ai-playbook\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10086500,"title":"AIM CXO Roundtable, in partnership with EPAM: An Exclusive Discussion about the Future Of Analytics in Pharma","content":"Analytics India Magazine, in partnership with EPAM Systems, Inc, a global digital services provider, brings to you ‘CXO Roundtable – an exclusive invite-only roundtable conference to discuss “The future of commercial analytics in the pharmaceutical industry.” Scheduled for February 28th at Taj, Bengaluru, the roundtable will present speakers putting forth their perspectives and encouraging the exchange of ideas. The focus is to discuss the challenges in the pharma industry, their potential implications on business operations and the role of omnichannel in improving sales outcomes and time to market. What to expect? Considering the pharma industry has been in a “state of flux,” some challenges continue to haunt us. For instance, leaving data in disparate servers has a number of risks, and one may waste resources if not organized correctly. This is where advanced analytics comes into the picture in improving interaction and influence of HCPs and Sales Reps. EPAM CXO Roundtable intends to take the conversation further with leaders exploring the digital disruption in the pharma industry and discussing how the newest digital trends are replacing the traditional data model. The event will kick off with a keynote address by EPAM, followed by a roundtable discussion with CXO Power Players. The summit anticipates the participation of attendees eager to know about the future of commercial analytics in the pharma industry. Date & Time: February 28, 2023, | 7 PM OnwardsVenue: Taj, MG Road, Bengaluru Key Discussion Points: Is the challenge with disparate data sources getting better or worse? Leveraging advanced analytics to improve interaction and influence of HCPs and sales representatives. Role of omnichannel in improving sales outcomes and time-to-market Challenges of owning commercial data, including master data How Cloud has improved data access and data quality Digital disruption & newest digital trends. What to do at the CXO Roundtable 2023? Listen to the keynote address by the EPAM leaders, encouraging constructive business agendas. Interact with industry leaders, including CXOs, decision-makers, and data and analytics practitioners. Opportunity for one-on-one networking with EPAM experts. Who should attend? Chief data\/digital officers Technical and business decision-makers Data & analytics decision-makers Date & Time: February 28, 2023, | 7 PM OnwardsVenue: Taj, MG Road, Bengaluru To register, RSVP here.","excerpt":"The discussion will address industry challenges, their potential implications on business operations, and omnichannel’s role in improving sales outcomes and time to market.","categories":["Deep Tech"],"tags":["CXO Roundtable","roundtable discussion"],"author_name":"Tasmia Ansari","publish_date":"2023-02-03T17:35:58","publication_year":"2023","word_count":365,"keywords":["roundtable discussion","CXO Roundtable","programming_languages:R","Git","RAG","ViT","analytics","data quality","disruption","GAN","R"],"extracted_tech_keywords":["analytics","RAG","R","Git","data quality","GAN","ViT","disruption","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/aim-cxo-roundtable-in-partnership-with-epam-an-exclusive-discussion-about-the-future-of-analytics-in-pharma\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10048235,"title":"Register For This Full Day Workshop On Data Engineering &#038; Databases","content":"The Association of Data Scientists (ADaSci), the premier global professional body of data science & machine learning professionals, is back again with an exciting full-day hands-on workshop on Mastering Skills In Data Engineering & Databases on October 9th, Saturday. Data Engineering is one of the most important aspects in the field of data science that concerns with the storage, accessing, processing and managing of the data required for any modelling and analysis purpose. With various types of data available from different sources required for different types of model building, it is critical for data science professionals to know data engineering skills. As a matter of fact, data engineering is currently one of the must-have skills for a professional in the field of data science. In collaboration with Analytics India Magazine, the Association of Data Scientists brings an extensive full-day workshop to help data scientists dive deeply into data engineering skills and get a comprehensive understanding of different feature engineering techniques required for modelling. In this workshop, the attendees will grasp the concept of data engineering and where and how it is used. In addition, attendees will learn different methods of processing the data, feature engineering and munging with hands-on experiments. The workshop will also provide an understanding of different types of databases and methods to handle these databases. Know more about the workshop here. Learning Outcomes — Complete understanding of data engineering & its importance in data science.Knowledge of different types of databases with their use & importance.In-depth understanding of data wrangling & feature engineering techniques.Hands-on exposure to different feature engineering techniques, required before modelling purposes. The prerequisites to participate in the workshop include a basic understanding of data analysis and database systems, a familiarity with Python programming language, and basic know-how of Jupyter\/Colab Notebook. The participants also need to have access to Jupyter Notebook \/ Google Colab with MySql installed, along with a high-speed internet connection, to avoid any lag. The attendees will also receive a certificate on request. Know more about the workshop here. About the instructor: Dr Vaibhav Kumar Dr Vaibhav Kumar is currently working as Senior Director with The Association of Data Scientists (ADaSci). He brings a lot of experience in the field of Deep Learning and Artificial Intelligence. With a diverse background in industry and academics, Dr Vaibhav has led many Research and Development (R&D) activities in the field of AI. With a PhD in Deep Learning, he has published many research papers in reputed journals and conferences and guided many scholars in their research works. Details of the workshop: Date: 9th October 2021 Time: 10 AM to 5 PM Mode: Online Price: $13.70 100% discount for CDS Registrants 100% discount for ADaSci Members Register for the workshop here.","excerpt":"The Association of Data Scientists has announced a hands-on workshop on mastering data engineering skills.","categories":["Deep Tech"],"tags":["ADAsci (Association of Data Scientists)","AI Workshops","Data Engineering","full-day workshop","python database gui"],"author_name":"AIM Media House","publish_date":"2021-09-13T17:00:00","publication_year":"2021","word_count":454,"keywords":["data science","artificial intelligence","full-day workshop","ADAsci (Association of Data Scientists)","AI","machine learning","AI Workshops","RAG","Colab","Python","deep learning","Data Engineering","analytics","Jupyter","python database gui"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","deep learning","data science","analytics","Jupyter","Colab","RAG","Python"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/register-for-this-full-day-workshop-on-data-engineering-databases\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10007554,"title":"Complete Guide To QLattice: A New And Transparent Machine Learning Framework","content":"Machine learning techniques and algorithms have time and again proven to be really effective over the past few years and there are more breakthroughs that will come in the near future. But most of the algorithms at hand use a lot of statistical methods and require a lot of human intervention. To reduce the amount of human interaction, a new machine learning algorithm was developed called QLattice. In this article, we will learn about the following in detail:- Introduction to QlatticeHow to implement Qlattice Introduction to QLattice Qlattice has its advantages over neural networks and decision trees. It removes the concept of black box seen in neural networks and serves explainability similar to that of a decision tree. It was developed by a research startup named abzu earlier this year. Qlattice works by searching through thousands of potential models and finds one with the right features and interactions to fit as the perfect model for the problem. Initially, the user has to set a few parameters like the input features, target and the independent variables. QLattice works with supervised learning and the input features and target are called registers. These registers form the entry and exit points for the QLattice model. Once the registers are set, a collection of possible models can be extracted from the QLattice. This collection of models corresponds to something called QGraphs. These graphs are a set of nodes and edges where the nodes have activation functions and edges carry weights. This indicates that upon training a Qgraph, it extracts essential information from the features to predict the target. Implementation of QLattice Let us understand the working of QLattice better by implementing a simple classification model. Before we begin, please make sure you have created an account in the abzu website. You will receive access to it and be directed to the dashboard. Here, you can either use the playground they provide or use Colab. I have made use of Colab. For implementing this we need the Feyn library which is the software development kit that we use to interact with the QLattice. To install the Feyn library use the command pip install feyn The dataset To keep things simple I have used the diabetes dataset which is downloaded from Kaggle. Let us load the dataset and import the needed libraries from google.colab import drive import feyn import numpy as np import pandas as pd drive.mount('\/content\/gdrive\/') dataset=pd.read_csv('\/content\/gdrive\/My Drive\/diabetes\/diabetes.csv') dataset.head() Splitting the dataset Before we split the dataset let us assign our target value to a variable named target. Next, we will split the dataset into train and test set. target='Outcome' train, test = train_test_split(dataset, test_size=0.4, random_state=42, stratify=dataset[target]) Qgraph Qgraphs are produced by Qlattice which forms a graphical structure that connects the input to output. We need to give instructions about the features of this. To be able to access this since it belongs to the third party we need a URL and authorization token. This is made available in your abzu dashboard. qlattice = feyn.QLattice(url=\"the url\", api_token=\"token\") Note here in the qgraph we are just giving the basic information and not specifying which algorithm to use. qgraph = qlattice.get_classifier(train.columns, target, max_depth=3) qgraph.head(2) As you see for each feature automatically the nodes with algorithms and activation function are assigned. Green boxes represent the inputs and outputs, white boxed with pink border represent interactions. Interactions take in input as values, builds a function for that value and predicts the outcome. It is similar to the neural network but we can see exactly what is happening in each column. Here the classification algorithm seems to be gaussian. Training Let us begin to train the qgraph over 5 fittings. Here epochs are called fittings. The code below shows the training loop. depth = 5 loss_function = feyn.losses.squared_error nloops = 5 for loop in range(nloops): qgraph.fit(train, loss_function=loss_function, threads=4,show='graph') qlattice.update(qgraph.best()) Below, it shows the best loss value after 5 fittings, here the loss value is 0.152 which is a good loss value. Once training is done it is time to move to predictions. Predictions output_graph = qgraph.sort(train)[0] df_pred = valid.copy() df_pred['predicted'] = output_graph.predict(valid) threshold = 0.5 y_pred = np.where(df_pred.predicted > threshold, True, False) y_real = df_pred[target].astype(bool) plt.figure(figsize=(8, 5)) feyn.plots.plot_confusion_matrix(y_real, y_pred) train_pred = output_graph.predict(train) print('Overall training accuracy: %.4f' %np.mean(np.round(train_pred) == train[target])) print('Overall validation accuracy: %.4f' %np.mean(y_pred == y_real)) The above code and confusion matrix shows that the model performs well because we have not performed any pre-processing and not specified the type of algorithm. Qgraph learnt and trained on only the best model for the dataset. Conclusion In the article, we learnt about the new machine learning model called Qlattice and implemented it on a simple classification dataset and got good results.","excerpt":"In this article, we will learn about the following in detail:- Introduction to Qlattice How to implement Qlattice","categories":["Deep Tech"],"tags":["Classification","machine learning framework","machine learning methods","machine learning software"],"author_name":"Bhoomika Madhukar","publish_date":"2020-09-17T15:00:54","publication_year":"2020","word_count":779,"keywords":["Go","Classification","NumPy","machine learning","TPU","API","AI","machine learning framework","neural network","Colab","machine learning software","machine learning methods","R","Pandas"],"extracted_tech_keywords":["AI","machine learning","neural network","Colab","Pandas","NumPy","TPU","R","Go","API"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/complete-guide-to-qlattice-a-new-and-transparent-machine-learning-framework\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10113661,"title":"India is the Toughest Market to Crack, Says Uber CEO, Signs MoU with ONDC","content":"Uber, India has signed an MoU with Open Network for Digital Commerce (ONDC) to expand the range of mobility offerings on the Uber app. The agreement with the Indian private non-profit organisation stands to strengthen Uber’s mission of bringing safe, affordable and reliable rides to all Indians. At an event in Bengaluru, Uber’s global CEO Dara Khosrowshahi spoke with Infosys chairman Nandan Nilekani on ‘building population scale technology’. He said companies and governments around the world can learn from the scale and ambition of India’s Digital Public Infrastructure. India’s Uber views open-source tech stacks with a lot of interest and recognises the opportunities they bring for everyone. Khosrowshahi mentioned that India is the toughest market to crack since “Indian customers do not want to pay for anything”. Hence, leaping into the market with ONDC will help Uber strategise globally. The Indian counterpart of the US-based ride-hailing platform has never shied away from experimenting and exploring the Indian landscape. “Indian customers are so demanding and do not want to pay for anything. It is the gig way to the world for us [Uber],” said Khosrowshahi. T Koshy, MD & CEO of ONDC, said, “As the Open Network is continuously evolving, MTT (Mobility, Transport and Travel) is certainly a critical sector for us. Different players together on the network foster innovation and newer business models. Today’s MoU is a major step forward, and one we hope will enable a diverse range of mobility solutions to benefit every Indian.” Over the past decade, Uber’s technological innovation has helped revolutionize mobility for people across the world. Moreover, the app has reshaped India’s transportation landscape, offering safe, reliable and affordable rides to people across 125 cities across a range of different vehicle types and has helped more than 900,000 Indians earn a living each month by driving with Uber. The company has invested millions in engineering tools through the Bangalore hub. The foundation laid by the team in Bangalore has helped the company’s business expand across sectors from calling for a cab to delivering food at your doorstep. The team has continued to make significant tech investments to reduce the cost of servers, database resources, and data storage resources. “This means that both businesses benefit from our reduced cost of operations, underlining our dedication to efficiency and sustainability,” Madan Thangavelu, Uber’s senior director of engineering had explained in an exclusive interaction with AIM.","excerpt":"Indian customers do not want to pay for anything","categories":["AI News"],"tags":["ondc","Uber"],"author_name":"Tasmia Ansari","publish_date":"2024-02-22T16:10:44","publication_year":"2024","word_count":398,"keywords":["Go","API","AI","ondc","RPA","innovation","Git","RAG","GAN","Aim","Uber","R"],"extracted_tech_keywords":["AI","Aim","RAG","R","Go","Git","API","GAN","RPA","innovation"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/india-the-toughest-market-to-crack-uber-ceo-tells-nilekani\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10075248,"title":"Draft Data Anonymisation Guidelines Pulled Down a Week After Being Put Up For Public Comments","content":"Last week, the draft document that listed guidelines for data anonymisation was removed from the information technology ministry’s website. The draft had been put up for public feedback just a week prior to being withdrawn. This is not the first instance of sudden retraction of draft Bills. In the past two years alone, major changes have been made to data-related Bills – the draft Indian Data Accessibility & Use Policy, 2022, was updated without any notification, and in 2021, the draft amendments to the IT Rules, 2021, were unceremoniously taken down during public consultations. MeitY was in the news in August when it withdrew the Personal Data Protection Bill after facing much pushback from several quarters. The ministry said a new legal framework incorporating several changes and amendments would replace it. Data anonymisation draft pulled down Two drafts – the Guidelines for Anonymisation of Data (AoD) and Mobile Security Guidelines (MSG) – listing guidelines on data anonymisation were put up on the IT ministry’s website for public consultation. The website had announced that all the public comments made until September 21 would be considered. It may be noted that the documents were released on a new website, instead of the official website of MeitY. Interestingly, no press release accompanied these documents at the time of uploading. A government official told ET that data anonymisation is a complex issue that needs wider consultation. “We will talk to experts again, look at global examples, examine them, and then put up the draft for public consultation in a few days,” the source said. In a disappointing yet not surprising turn of events, the @GOI_meity seems to have removed the documents pertaining to the Guidelines for Anonymisation of Data (AoD) for public consultation. 1\/5https:\/\/t.co\/ikThibtQ4Q— Internet Freedom Foundation (IFF) (@internetfreedom) September 6, 2022 The data anonymisation draft included guidelines for all stakeholders involved in personal data processing and its subtypes through the e-governance projects. The draft aimed to lay down the recommendations for processing of the data collected through e-governance portals like Aarogya Setu, Cowin vaccinations, and National Health Mission, among others. On the other hand, the second document draft on mobile security guidelines had measures for sensitive data protection, privacy, and security of transactions. The draft report was prepared by the Standardization Testing Quality Certification (STQC) Directorate and Centre for Development of Advanced Computing (C-DAC) and was commissioned by the Ministry of Electronics & Information Technology (MeitY). The standards and the guidelines outlined in the draft were to ensure secure sharing of information and seamless interoperability of data across several government applications. The government has been deliberating several steps for data governance and privacy, and the data anonymisation draft was a step closer to that goal. Earlier, CDAC had constituted a Working Group consisting of 14 experts to frame the guidelines for data anonymisation when personal information is processed and shared, especially in government applications. Further, this committee was also set up to deliberate on standard operating procedures to conveniently implement the concept of various organisations and government agencies. Another purpose of the committee was to prepare a supplement to the data-related policies and legislations of the Indian government. What happened to the PDP Bill? After three years of deliberations, the Centre this year, withdrew the Personal Data Protection Bill 2019. The Bill sought to regulate how the government and companies could use the digital data of citizens. It had faced severe backlash from stakeholders – citizens, tech firms, and political parties – since inception and had undergone various changes over the years based on experts’ suggestions. The Union minister for electronics and IT Ashwini Vaishnaw noted that the now-withdrawn Bill had been “deliberated in great detail” by the joint parliamentary committee that proposed 81 amendments and 12 recommendations. Earlier this month, Rajeev Chandrashekar, MoS for electronics and information, said at a conference that the government would be replacing the IT Act with a Digital India Act. Chandrashekar added that the government aims to bring this framework into the public domain within three to four months, and it will address digital data protection, and non-personalised and anonymous data.","excerpt":"Two drafts – the Guidelines for Anonymisation of Data (AoD) and Mobile Security Guidelines (MSG) – listing guidelines on data anonymisation were put up on the IT ministry’s website for public consultation","categories":["Deep Tech"],"tags":[],"author_name":"Shraddha Goled","publish_date":"2022-09-15T10:00:00","publication_year":"2022","word_count":684,"keywords":["Go","programming_languages:R","AI","ML","programming_languages:Go","Git","Aim","data governance","GAN","R"],"extracted_tech_keywords":["AI","ML","Aim","R","Go","Git","data governance","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/draft-data-anonymisation-guidelines-pulled-down-a-week-after-being-put-up-for-public-comments\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10044482,"title":"A Guide To PyXLL-Jupyter Package For Excel Integration","content":"Last year, PyXLL released its PyXLL-Jupyter plugin. The new extension combines the ease of use of Excel with the interactivity of Jupyter. Once installed, you will be able to plot your data in Jupyter and have the ability to share them with Excel seamlessly. It can also open on Jupyter Notebook in a browser as well as in an Excel task pane. Excel comes with pivot tables, graphing tools, and a macro programming language known as Visual Basic for Applications (VBA). Jupyter Notebook, on the other hand, is an open-source web application that allows users to create and share documents that contain live code, visualisations, equations and narrative text. With the PyXLL-Jupyter package, you can use both together, side by side. VBA is commonly used to automate Excel with macros, add new user-defined worksheet functions (UDFs) and react to Excel events. The same is achieved with Python, as it is much faster. Founded in 2010, PyXLL is the brainchild of Tony Roberts. He started PyXLL after years of working with Python and Excel in the finance industry. PyXLL is one of the go-to tools for writing Excel add-ins in Python, especially for people working in data science roles. xiwings, DataNitro, Pandas, NumPy, Anaconda, SciPy, PySpark and DataForm, are some of the alternatives to PyXLL. How to Install PyXLL-Jupyter? Firstly, to run Python code in Excel, you need the PyXLL add-in. ‘PyXLL add-in’ lets you integrate Python into Excel and use Python instead of VBA. To set up ‘PyXLL Excel add-in (pip install pyxll) and then use the PyXLL command line tool to install the Excel add-in: > pip install pyxll > pyxll install Beginners can check out PyXLL’s online documentation here. Once you have successfully installed the PyXLL Excell add-in, the next step is to install the pyxll-jupyter package. This package essentially glues PyXLL and Jupyter so that you can use Jupyter Notebooks within Excel. The PyXLL-Jupyter package is installed using pip: > pip install pyxll-jupyter Once you have installed both the PyXLL Excell add-in and the PyXLL-Jupyter package, you will see a new ‘Jupyter’ tab in the PyXLL dashboard, as highlighted in the image below. (Source: PyXLL) Upon clicking this button, Jupyter Notebook opens in a side panel in your Excel workbook. This new panel is part of the Excel Interface and can be dragged to different locations at your convenience. In the Jupyter panel, you can select an existing notebook or create a new one. If you want to create a new notebook, select the ‘New’ tab followed by ‘Python 3,’ as shown in the image below. (Source: PyXLL) (Source: PyXLL) Further, to upgrade to the latest version of PyXLL-Jupyter, you can run pip install –upgrade pyxll-jupyter. PyXLL-Jupyter Now that you have a complete Jupyter Notebook running inside of Excel, you can use Excel for working with your data and use Python to work on the same data. In other words, you can use Excel as an interactive playground for organising and visualising your data, effortlessly switching to Python for more sophisticated tools. Also, you can use Jupyter Notebooks as a scratch-pad for trying out Python code and write Excel functions entirely in Python in a Jupyter notebook and test them out in real-time. “Once you have developed a useful re-usable function, add it to your PyXLL Python project. That way, you can use the same function every time you see Excel,” wrote Tony Roberts, the founder of PyXLL. The key highlight of PyXLL-Jupyter include: Share data between Excel and Python using your Jupyter NotebookWrite Excel worksheet functions (UDFs) in your notebookScript Excel with Python instead of VBA Wrapping up Python makes a powerful alternative to VBA. Using PyXLL, you can write fully featured Excel add-ins entirely in Python. At the same time, Excel is also a brilliant tool for interactive computation, and adding Python and Jyupter takes Excel to the next level. Plus, code written in Jupyter Notebook is easily refactored into standalone Python packages to create Excel toolkits to power intuitive Python tools written using PyXLL — without needing any knowledge of Python. PyXLL-Jupyter just simplifies the entire experience by providing both tools under one umbrella.","excerpt":"Last year, PyXLL released its PyXLL-Jupyter plugin. The new extension combines the ease of use of Excel with the interactivity of Jupyter.","categories":["Deep Tech"],"tags":["data science latest","data science tools","excel","Guide","Jupyter Notebook","Machine Learning","Machine Learning Latest"],"author_name":"Amit Naik","publish_date":"2021-07-27T13:00:00","publication_year":"2021","word_count":690,"keywords":["data science","excel","data science tools","Jupyter Notebook","NumPy","Go","AI","data science latest","ML","Machine Learning","Machine Learning Latest","RAG","Python","Jupyter","R","Guide","Pandas"],"extracted_tech_keywords":["AI","ML","data science","Jupyter","Pandas","NumPy","RAG","Python","R","Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/a-guide-to-pyxll-jupyter-package-for-excel-integration\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10071112,"title":"How is gradient descent used in unsupervised learning problems?","content":"The cost function in several machine learning algorithms is minimized using the optimization approach gradient descent. Its primary objective is to update the parameters of a learning algorithm. These models gain knowledge over time by using training data, and the cost function in gradient descent especially serves as a barometer by assessing the correctness of each iteration of parameter changes. Gradient descent is commonly used in supervised learning but the question is whether it could be used in Unsupervised Learning. This article will focus on understanding ‘how’ gradient descent could be used in unsupervised learning. Following are the topics to be discussed. Table of contents About Gradient DescentTraining word2vec modelsTraining autoencoder modelsTraining CNNs Gradient Descent finds the global minima of a cost function, so it can’t be applied to algorithms without a cost function. Let’s start with a high-level understanding of gradient descent. About Gradient Descent Although the gradient descent algorithm is based on a convex function, but behaves similar to a linear regression algorithm, below is an example to refer to. Analytics India Magazine The beginning point is only a position chosen at random by the algorithm to gauge performance. A slope would be discovered from that starting point, and from there, a tangent line would be created to gauge the slope’s steepness. The modifications to the parameters, such as the weights and bias, will be informed by the slope. The slope will be steeper at the starting point, but when additional parameters are created, the steepness should steadily diminish until it hits the point of convergence, which is the lowest point on the curve. The objective of gradient descent is to reduce the cost function, or the difference between the anticipated value and the actual value, much like finding the line of best fit in linear regression. Two data points, a direction and a learning rate are necessary. Future iterations might gradually approach the local or global minimum because these variables impact the partial derivative computations of those iterations. Learning rate: The magnitude of the steps needed to get to the minimum is referred to as the learning rate. This is usually a modest value, and it is assessed and updated in accordance with how the cost function behaves. Larger steps are taken as a result of high learning rates, due to which the minimum may be exceeded. A poor learning rate, on the other hand, has short step sizes. The number of repeats reduces overall efficiency even if it has the benefit of greater precision since it requires more time and calculations to get to the minimum.The cost function: It calculates the error between the actual y value and the expected y value at the present point. This increases the effectiveness of the machine learning model by giving it feedback so that it may change the parameters to reduce error and locate the local or global minimum. Up until the cost, the function is near to or equal to zero, it constantly iterates, travelling in the direction of the sharpest fall (or the negative gradient). The model will then stop learning at this moment. In unsupervised learning there gradient descent could only be utilized in Neural networks because it has a cost function and other unsupervised learning doesn’t have a cost function which is needed to be optimized. Let’s apply the gradient descent algorithm to some unsupervised learning and learn the functionality of these unsupervised learners. Are you looking for a complete repository of Python libraries used in data science, check out here. Training word2vec models Word2vec is a method for processing natural language. With the help of a huge text corpus, the word2vec technique employs a neural network model to learn word connections. It is a two-layer neural network that “vectorizes” words to analyze text. It takes a text corpus as input and produces a set of vectors as output. Words in the corpus are represented by feature vectors. Once trained, a model like this may identify terms that are similar or propose new words to complete a phrase. As the name suggests, word2vec uses a specific set of integers called a vector to represent each unique word. The vectors are chosen so that a straightforward mathematical function may be used to determine the similarity of the words represented by those vectors to one another in terms of meaning. The word2vec model could be trained using two different algorithms which are skip-gram and CBOW (Continuous bag of words). The skip-gram algorithm is based on gradient descent optimization. When given a current word, the continuous skip-gram model learns by predicting the words that will be around it. To put it another way, the Continuous Skip-Gram Model foretells words that will appear before and after the present word in the same phrase within a specific range. Finding the vector representation of each and every word in the text is the major goal since it decreases the dimensional space. Each word in a skip-gram will have two distinct representations as its trick. The representation is defined as: When the word is a centre wordWhen the word is a context word The target word or input delivered is w(t), as shown in the skip gramme architecture shown above. One hidden layer computes the dot product of the input vector and the weight matrix. In the buried layer, no activation function is employed. The output layer is now given the result of the dot product at the concealed layer. The hidden layer’s output vector and the output layer’s weight matrix are combined to create a dot product in the output layer. The likelihood of words appearing in the context of w(t) at a specific context location is then calculated using the softmax activation function. Training Autoencoders models The input and output of feedforward neural networks that use autoencoders are identical. They reduce the input’s dimension before using this representation to recreate the output. The code, also known as the latent-space representation, is an efficient “summary” or “compression” of the input. Encoder, code, and decoder are the three parts of an autoencoder. The input is compressed by the encoder, which also creates a code. The decoder then reconstructs the input exclusively using the code. Autoencoders are mainly dimensionality reduction algorithms which are data-specific, unsupervised and have lossy output. Analytics India Magazine The goal of an autoencoder is to train the network to capture the most crucial elements of the input picture in order to learn a lower-dimensional representation (encoding) for higher-dimensional data, often for dimensionality reduction. Under the aforementioned generative models, the autoencoder weights are updated via gradient descent, and the revised weights are then normalized in the Euclidean column norm, resulting in a linear convergence to a small neighbourhood of the ground truth. The term “Normalized Gradient Descent” (NGD) refers to a modification of the term “Traditional Gradient Descent” in which each iteration’s updates are only based on the gradients’ directions without regard to their magnitudes. The gradients are normalized to achieve this. Instead of employing the whole set of data, the weight updates are carried out using specific (randomly selected) training examples. By updating the parameters at the conclusion of each mini-batch of samples, mini-batch NGD generalizes. Training Convolutional Neural Networks A particular kind of neural network known as a convolutional neural network (CNN) has won several competitions involving computer vision and image processing. A few of CNN’s exciting application areas include speech recognition, object detection, video processing, natural language processing, and image classification and segmentation. The extensive use of feature extraction phases, which may automatically learn representations from data, accounts for Deep CNN’s high learning capacity. One of CNN’s most enticing properties is its ability to use spatial or temporal correlation in data. Every learning step of CNN is divided into a variety of convolutional layers, nonlinear processing units, and subsampling layers. Using a bank of convolutional kernels, each layer of CNN’s multilayered, feedforward network performs a number of modifications. The convolution process helps to extract useful properties from spatially related data points. The Convolutional Neural Networks have a forward passway and a backward passway. The forward passway is the subject of the first two sections of the study. The backward passway will be explored in this section. The weighted function will not be verified as to whether it satisfies the request for structure’s accuracy due to the initialization of the weighted function during the CNNs procedure. The function needs to be fixed repeatedly. Backpropagation (BP) is used to propagate correcting errors from upper levels to lower layers. The weighted functions can then be maintained by fixed errors in the lower levels. To locate the fixed errors in CNN, the gradient descent approach is applied. With the initialization of the procedure, the partial derivative of the loss functions is calculated as a gradient, which measures the trend of the objective function. The accuracy of a function is often measured as the difference between a mathematical model’s output and a sample. The weighted functions in the model are pleased with the request of the process when the difference is less than or equal to the recursion terminal distance. Then, the procedure may end. The learning rate, also known as the step size, determines how much of the gradient is utilized to update the new data. The objective function can ultimately be optimized if it is a convex function. Conclusion The gradient descent algorithm finds the global minima of the cost function, for utilizing the optimization method the algorithm must have a cost function. Since the clustering algorithms like hierarchical clustering, agglomerative clustering, etc do not have a cost function the descent method can’t be applied to them. With this article, we have understood the utilization of gradient descent in unsupervised learning. References Read more about Word2vecWhite paper on utilizing GD in CNN","excerpt":"Gradient Descent is primarily used in Neural Networks for unsupervised learning","categories":["AI Trends"],"tags":["autoencoders","CNNs","Dimensionality Reduction","gradient descent","linear regression","local minima","Loss functions","Neural Networks","optimization","optimization algorithms neural networks","Parameters","Unsupervised Learning","Word2Vec"],"author_name":"Sourabh Mehta","publish_date":"2022-07-16T16:00:00","publication_year":"2022","word_count":1632,"keywords":["TPU","Loss functions","computer vision","object detection","local minima","Dimensionality Reduction","CNNs","data science","optimization algorithms neural networks","analytics","Unsupervised Learning","Neural Networks","machine learning","AI","neural network","ML","Parameters","optimization","linear regression","autoencoders","Python","Word2Vec","gradient descent"],"extracted_tech_keywords":["AI","machine learning","ML","neural network","computer vision","data science","analytics","object detection","TPU","Python"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/how-is-gradient-descent-used-in-unsupervised-learning-problems\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10097627,"title":"Hype, Exit, Repeat: a16z&#8217;s Generative AI Journey Creates Déjà Vu","content":"A stage agnostic VC firm, Andreessen Horowitz aka a16z, managed to capitalise on the blockchain, crypto and metaverse bubble at the right time, a formula they wish to replicate with AI. The firm seems to have steered away from the things of the past and is speeding towards generative AI with the overwhelming outpouring of its “vision for the AI-enabled future”, suggesting ways generative AI could help with “the creation of new medicines”, “climate change” and whatnot. A16z says it takes a long view of relationships and defines its investment philosophy as being “in the business of investing in innovators”. It also claims that it “think(s) far more about how big the outcome will be if a deal succeeds than all the ways that it can fail.” And that’s true for them, however, they apply a very crafty and effective tactic—hype, exit, repeat. It is well-known for leveraging effective PR to create hype around companies like Coinbase, Airbnb, Affirm, Instacart, Netscape, and Skype, leading to high valuations. However, after the VCs exit, the valuations of these companies often decline. If the barrage of their well-crafted and researched blogs around GenAI are anything to go by, Andreessen Horowitz is looking to do the same with generative AI. While, Geoffrey Hinton, the Godfather of AI broke his decades-long association with Google to warn people openly about the dangers of AI, Marc Andreessen, one of the co-founders of the VC firm went ahead and wrote a 7000-word blog last month on “why AI will save the world” and all the naysayers should just shut up. After extensively dismissing the worries of AI sceptics, Andreessen introduced a new panic: the possibility of American AI losing ground to China’s dystopian concept of AI dominance. He suggests that to prevent this, the ‘free world’ must invest heavily in developing AI as an open and unrestricted technology. Sounds more like a move to coerce people into directing funding towards AI hype companies he is associated with, doesn’t it? In a recent episode of the Joe Rogan podcast, Marc doubled down on the usefulness of AI, in contrast to detractors who fear its potential to destroy or control humanity. He presents a compelling vision of AI’s ubiquity, envisioning its application in various fields, such as education, scientific research, art, and military strategy. Cofounder Marc Andreessen suggested that AI will take over everything from tutoring children and helping scientists to augmenting artists’ work and improving warfare. He went to length envisioning the AI his son will grow up. He said that when children who grow up with AI go to college or enter the workforce, “they’ll have an ally right with them. They’ll have basically a partner whose goal in life will be to make them as happy and satisfied and successful as possible.” Going All in on AI Just yesterday, the firm announced the onboarding of Anjney Midha, Ubiquity6 cofounder and former CEO, who is joining the firm to lead its artificial intelligence investing efforts. The firm has raised a total of $7.1 billion across seven funds, which looks to support tech entrepreneurs and offer guidance from a diverse team of engineers, executives, industry experts, and academics. It has actively led fundraising campaigns, such as raising $30 million for People.ai. a16z has raised a total of $32.4B across 27 funds in total, with 1,416 investments and 625 lead investments. It has made 203 exits in firms like Github, Coinbase, Pinterest Reddit, and also picked up new shares in OpenAI, Replit, and Replicate. Additionally, the company has dedicated a bio fund focused on advancing AI in the medical domain to aid disease prevention and treatment. Some of their notable investments include Shield AI, which develops AI software for intelligence, surveillance, and reconnaissance, and Freenome, an AI-driven platform for understanding the immune system. It has also invested in Inflection AI, Scale AI, Hippocratic AI, and Character.AI. Not just that, a16z has also invested in Open AI CEO Sam Altman’s risky bet–-Worldcoin, which is out to scan every iris in the world. At this point, the firm seems to have leaped headlong into the fray. The Sea of Investment AI companies are experiencing significant attention and investment from venture capitalists. However, VCs tend to invest in AI founders who hail from from premier institutions like Stanford, Harvard and others, alongside having prior experience with big techs. Conversely, graduates from these universities can join AI startups, becoming highly valued assets for attracting funding. In exchange, these graduates would receive equity in the startup. This concept is akin to a service where Stanford University graduates act as a valuable resource for AI startups seeking investment. This pattern has been observed in technology markets that hold disruptive potential — first with crypto and blockchain, then with the metaverse, and now with AI — the hype surrounding these technologies has attracted substantial funding. However, this surge of interest from VCs could potentially lead to the AI bubble burst. Prior to VCs’ involvement, tech giants like Google, Meta, and Amazon were already investing heavily in AI innovation and development. But once the AI hype gained momentum, even these established companies jumped on the AI bandwagon. Despite the optimistic reports and claims, many generative AI use cases are yet to materialise, and enterprises struggle with security concerns. AI’s potential for disruption should be approached with measured scepticism, as there are numerous AI grifters seeking to capitalise on the hype. Update: 27th July | 16:11: The article has been updated to show a16z’s investment and exit","excerpt":"One name stands out in the generative AI investment landscape, Andreessen Horowitz, a VC firm that backs entrepreneurs building the future through tech","categories":["AI Features"],"tags":["ai hype","Airbnb","coinbase","Marc Andreessen","OpenAI","Replit","VC"],"author_name":"Shyam Nandan Upadhyay","publish_date":"2023-07-27T10:02:58","publication_year":"2023","word_count":918,"keywords":["Replit","VC","Go","Marc Andreessen","artificial intelligence","GenAI","coinbase","OpenAI","AI","RAG","ai hype","Ray","Aim","generative AI","Airbnb","R"],"extracted_tech_keywords":["AI","artificial intelligence","generative AI","GenAI","OpenAI","Aim","Ray","RAG","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/hype-exit-repeat-a16zs-generative-ai-journey-shockingly-resembles-blockchain-and-crypto\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10040777,"title":"Ensure  Tech Startups Don’t Get Locked Down To A Particular Vendor: Sanjay Aneja, VMware India","content":"Analytics India Magazine got in touch with Sanjay Aneja, Sales Head – VMware Tanzu, VMware India, to understand how the firm is supporting developers with the right solutions and services to recover from the aftermath of the pandemic. Aneja joined VMware in 2020. VMware Tanzu is a suite of products and solutions to build, run, as well as manage kubernetes controlled container-based applications. Aneza works with IT leaders and architects, helping them create the ideal environment to build, run, and manage modern apps – from the data centre to the cloud to the edge. Excerpts: AIM: Tell us about your role at VMware Aneja: I lead the VMware Tanzu portfolio, and I joined VMware as part of the Pivotal acquisition last year. Now our focus is to help customers build, run, and manage modern applications based on microservices architecture. Before Pivotal, I have been associated with New Age products across all technologies, which typically the early adopters go after. I have taken up two main initiatives. First one is about spreading awareness about what a modern cloud native application is, and what are the benefits of these applications. A major portion of my time goes in doing business development activities and educating customers. The second thing I am focusing on is, while we educate the customers, it’s very important we build a partner ecosystem, which can support these customers when they go into production with their critical applications. AIM: How important is the Indian market for VMware? Aneja: India is an extremely large market and a critical market, not only in terms of dollar revenues to our parent organisations across the globe, but also in terms of the use cases and the challenges which Indian market brings to the table. We bring in new challenges, new inputs to product companies, and new use cases. Due to this reason, we have made it a point that, whenever the customer satisfaction board or customer advisory board is created, we get a couple of important customers in India who are doing innovative things to join the board so that they can bring in their perspectives. Thus, the Indian market is important both from a defining product strategy, as well as from a point of revenues. AIM: How can Kubernetes and VMware Cloud help the developers in infrastructure automation and management in India? Aneja: At present, VMware is bringing the ability to run containers as part of virtualization layer. Today, a customer can run containerized workloads within vSphere seven. You do not require a separate tool or a separate layer for running your Kubernetes workloads. This is a very strong value proposition for our customers to manage their entire infrastructure and the entire management automation. Our vSphere seven with the Tanzu capabilities along with the multi cloud offering provides a uniform layer across the hyper scalars. It relieves the developers from the pain of learning complications associated with different layers of infrastructures across different clouds. AIM: What are the important tools and techniques a software developer must know? Aneja: I feel developers need to focus on where they want to specialize. However, they should be aware of some basics–irrespective of what they are doing–around Kubernetes, upstream Kubernetes etc. Understanding Dockers concept, how it helps, and understanding cloud native technologies like Spring Boot, Spring Cloud, is a must for any developer. Spring is the most popular framework for Java EE developers for modern day cloud native applications. This allows developers to focus on their code work. In addition to this, I think developers today should also understand the tricks and new techniques used for agile development. AIM: Tips for deep tech startups dealing with challenges from COVID-19 Aneja: Due to COVID, the last few months have seen an exponential increase in the load on the digital applications. There has been an increase in the number of tools and technologies available to tech startups. Tech startups need to be careful on two important decisions. They should not get locked down to a particular vendor or a service provider. They should be able to implement the concept of ‘develop once on any platform and run anywhere on any platform’. The second thing is a healthy mix of open-source technologies and tools backed by enterprise support. Thus, in addition to the open-source tools, which are a must for the success of cloud tech startups, they should also look at some enterprise softwares to bring in operational efficiency, improve customer benefits and customer interaction etc.","excerpt":"Analytics India Magazine got in touch with Sanjay Aneja, Sales Head – VMware Tanzu, VMware India, to understand how the firm is supporting developers with the right solutions and services to recover from the aftermath of the pandemic. Aneja joined VMware in 2020. VMware Tanzu is a suite of products and solutions to build, run, […]","categories":["AI Features"],"tags":["Interviews and Discussions","Kubernetes","VMWare"],"author_name":"Ambika Choudhury","publish_date":"2021-05-26T13:00:00","publication_year":"2021","word_count":746,"keywords":["VMWare","Go","AI","R","docker","Scala","Java","microservices","Aim","analytics","Kubernetes","kubernetes","Interviews and Discussions"],"extracted_tech_keywords":["AI","analytics","Aim","kubernetes","docker","microservices","R","Go","Java","Scala"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/ensure-tech-startups-dont-get-locked-down-to-a-particular-vendor-sanjay-aneja-vmware-india\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10000178,"title":"Cannabis Based Medicines In India &#8211; What Should Be India&#8217;s Regulatory Policies?","content":"In a country like India, where the medical infrastructure is abysmally poor when compared with the 1.3 billion population we have, talking about the mental health scenario as such might bring in about much more shame to the existing healthcare system laid in India. Why are we concerned about drastic mental health ailments or about medical problems as such? Recent occurrences of legalization of cannabis or marijuana in countries like Canada for recreational & medical purposes has raised concern in India as when a developed nation like Canada can legalize & regulate cannabis, what holds back a developing country like India to not yet regulate cannabis in way its beneficial for the patients as well as for the public who are suffering from multiple chronic ailments. NCMH (National Care of Medical Health), states that at least 6.5 per cent of the Indian population suffers from some form of the serious mental disorder. According to recent WHO report nearly 56 million people suffer from depression whereas 38 million people are battling directly or indirectly from anxiety disorders & stress related ailments on daily basis. Legalization of cannabis will act as boon for all those who have such issues at large, if legalized for medical usage. Medical cannabis can be considered effective at alleviating chronic and moderate pain, sleep disorders, depression, anxiety, spasticity etc. This can make it an alternative yet effective palliative treatment for a multiple range of conditions. These includes cancer, epilepsy and multiple sclerosis too. Source: Mayo Clinic “ Ram Vishwakarma, Director of the Jammu-based Indian Institute of Integrative Medicine (IIIM) recently told     The Hindu that “Cannabis-based drugs have the potential to meet the unmet needs of terminally-ill cancer     patients and of those suffering from epilepsy and sickle cell anaemia, a hereditary disease that afflicts nearly two crore tribal living mainly in central States. ” Major Bottlenecks For Medical Usage Of Cannabis A.K.A Marijuana In India : Healthcare is still largely a state subject in India, so central ruling of legalization of marijuana for medical purposes if not for recreation needs, broad consensus among multiple state governments & health regulatory bodies is needed. The draconian NDPS act (Narcotic Drugs & Psychotropic Drugs Act) of 1985, still heavily puts a ban on substance like cannabis including age old bhang. Availability of limited clinical evidence with respect to the long-term effects of medical cannabis which is an issue causing a trouble for its legal status in most of the conservative countries across the world including India. As, India is still a developing country, society’s approval is must especially when it comes drugs which has such polluted reputation across Indian community where it’s been labeled that cannabis is a potent drug that derails life of people at large. So, more than parliamentary legalization, it’s the public acceptance of cannabis to be used for issues related to mental health is doubtful at large. “A private bill was tabled in last year for amending the NDPS law. Dharamveer Gandhi, a Lok Sabha MP who promulgated the bill, blames the NDPS for enabling more potent, addictive and dangerous alternative drugs that are available in the markets where legalization of drugs including cannabis in a regulated way can bring down on such illegal drug menace.” Types of Cannabis Available in India For Medical Purposes: THC and CBD are both major cannabinoids that are found in cannabis plants. THC is found in marijuana type cannabis, while CBD is more prominent in hemp type cannabis.There are also other types of cannabinoids, such as CBG or THCV, which are also found in some varieties of cannabis. CBD is a non-psychotropic cannabinoid that can be more easily can be used as a medicine because it does not make you high. It seems to be the most effective for conditions where the nervous system is affected like epilepsy, anxiety, stress, Alzheimer’s etc. THC is a well-known psychotropic from cannabis that makes you feel high. It has a strong analgesic property and can treat the symptoms of chemotherapy in terminally ill cancer patients. Promising Steps Towards Cannabis Legalization: Medical Cannabis license has been granted by the Government of India for the cultivation of Cannabis plants by the IIIM for medical use. Hence there has been a little shift in the attitude of Government towards making these medicines available In India for at least minor medical purposes. States like Sikkim are moving forward with a bold plan to decriminalize cannabis usage in an effort change the public perception which can bring more consumers for treatment. States like Uttarakhand have mulled in procedures to allow in to grow non-narcotic cannabis for industrial use, so that they can participate in the 1 trillion-dollar industry of which European, American & Chinese have long been a part of. India’s first ever research and development (R&D) license to grow cannabis and study its medicinal properties & use has been granted to the Council of Science and Industrial Research (CSIR) along the Bombay Hemp Company (BOHECO). The collaboration is on a project examining the potential medical benefits of marijuana in relation to chronic illness like cancer and epilepsy. Canadian health inspector checking medical cannabis in marijuana farm. “Canadian Government’s new regulations on legalizing marijuana for commercial as well as for medical purposes is a classic example for how to put in a constitutional framework to lay down the process for step by step legalizing cannabis in India. Also, South African regulatory policies are commendable towards legalizing usage of medical cannabis too.” Regulatory Measures That Are Needed In India : 1.  India will need to standardize its cannabis strains to achieve predictable outputs of medicinal derivatives to bolster clinical research. 2. Policy makers need to holistically analyze the need for cannabis regulation & a bring a structured bill to encompass all the bureaucratic hassles. 3. Phase by phase introduction of cannabis for medical usage based upon clinical evidences charted needs to be put in along with committees that can independently issue any reforms needed on time to time basis. 4. Strict measures need to be put to decriminalize the cannabis farming and clinical procedures of issuing license to both cannabis farming along with medical business regulations needs to be put in on place & consumer safety must be taken prime importance. 5. Government agencies needs to take in the responsibility to change the public perception & licensing psychologists \/psychiatrists to issue prescription for usual mental stress & depression related issues. Though, the medical usage of cannabis looks promising in India as scientific organizations are taking leap in conducting clinical trials along with multiple state governments picking up positive attitude towards cannabis regulation. Sooner or later policy makers need to amend the NDPS Act to formulate a structured well sort out plan to legalize medical cannabis in India along with enhanced public acceptability to solve India’s burgeoning chronic health issues including the existing worse mental health scenario.","excerpt":"In a country like India, where the medical infrastructure is abysmally poor when compared with the 1.3 billion population we have, talking about the mental health scenario as such might bring in about much more shame to the existing healthcare system laid in India. Why are we concerned about drastic mental health ailments or about […]","categories":["AI Trends"],"tags":["Healthcare Automation"],"author_name":"Martin F.R.","publish_date":"2018-12-06T19:46:44","publication_year":"2018","word_count":1146,"keywords":["Go","TPU","programming_languages:R","AI","programming_languages:Go","Healthcare Automation","GAN","R"],"extracted_tech_keywords":["AI","TPU","R","Go","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/cannabis-based-medicines-in-india-what-should-be-indias-regulatory-policies\/","complexity_score":3,"technical_depth":7,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10097977,"title":"Working with Generative AI Just Got Faster!","content":"Open source models (Falcon, Llama, Stable Diffusion, and GPT J) are not easy to work with, it gets even more complicated when you have to test all of them to fit your requirements and specific use cases, and it’s definitely an expensive affair. But, not anymore. “You can now test Llama 2 in less than 10 minutes,” said AI expert Santiago, introducing Monster API, a new tool that lets you effortlessly access powerful generative AI models such as Falcon, Llama, Stable Diffusion, and GPT J and others, without having to worry about managing the generative AI models or scaling them up to handle lots of requests. Santiago said that he has been working with the Monster API platform for a while now, and seems to be impressed with the level of accessibility it provides to open source generative AI models. “They take care of the GPU infrastructure, containerisation, Kubernetes clusters, scalability, etc,” he added that you only need to focus on your code integration. Further, he said that it leverages a distributed GPU network, so that users can access these models at a fraction of the cost. Here is the source code of this example: https:\/\/t.co\/BTWnlaraWu.You can also join @monsterapis’ Discord server for the latest updates, free credits, and special offers: https:\/\/t.co\/A7axoM858K.Thanks to the team @monsterapis for partnering with me on this post.— Santiago (@svpino) July 31, 2023 Decentralised GPU Founded by brothers Gaurav Vij and Saurabh Vij in June 2023, Monster API uses idle computing power of millions of decentralised crypto mining rigs worldwide and optimises them for machine learning and packages them with popular generative AI models. In other words, it uses distributed computing to bring down the cost of training a foundational model. The mining of bitcoins needs high levels of compute deployed on GPUs. Now, the interest in crypto is in decline and many of these devices are gathering dust. Gaurav Vij, founder of QBlocks said, “We eliminate the need to worry about GPU infrastructure, containerization, setting up a Kubernetes cluster, and managing scalable API deployments as well as offering the benefits of lower costs. One early customer has saved over $300,000 by shifting their ML workloads from AWS to Monster API’s distributed GPU infrastructure,” His company provides a decentralised GPU network at up to 10x more affordable rates to data scientists, researchers, designers and developers. He said, “Most of the machine learning developers today rely on AWS, Google Cloud, Microsoft Azure to get resources and end up spending a lot of money.” The artificial intelligence world is struggling to match the hardware in computing power. Demand has outstripped supply, says Saurabh, founder and CEO of Monster API. He further explained, “You can take a pre-trained foundational model; you can take datasets from free datasets like Hugging Face and quickly start fine-tuning these foundational models for your custom dataset. This can be done for under 30 to 40 dollars instead of hundreds of dollars, which you otherwise could spend on fine-tuning these models.” The company has cut fine-tuning costs by up to 90% through optimisation, with fees around $30 per model. Their website also provides information for developers to build no code fine-tuning of large language models, build over Llama2, Alpaca, Falcon 7B, Stable LM 3B and more. Other Similar Platforms Monster API is not alone. There are several tools, including the likes of Gooey.AI, Illusion AI etc, that are built on frameworks like PHP, Python and Java. H20.ai are now cropping up around these models acting like the middle man between the models and the user. They take care of all the more difficult processes like providing on demand access to a pool of GPUs, reducing the cost of training and refining, providing APIs for natural language technologies and computer vision applications etc. They make fine tuning accessible for a wider audience through their no-code user interface. They do this through a visual or graphical interface making it possible for people to take advantage of state-of-the-art models. H2O recently introduced a driverless AI as a tool. Driverless AI automates complex data science and machine learning tasks, including feature engineering, model validation, tuning, selection, and deployment. It does things like selecting the best features, fine-tuning the models, and creating a simple and fast way to use the models in real-world applications. AT&T has scaled the adoption of H2O driverless with more than 380 employees using Driverless AI across 80 business units saving the company money. This is only two examples of an entire ecosystem providing no-code\/low-code platforms for Generative AI models. There are considerable drawbacks in these platforms. Businesses are under pressure to deliver applications faster, and these options are cost effective but the security and scalability of these platforms are limited as of now. A number of open source projects and companies that are constantly improving on these issues like Alteryx, Kinme, Dataiku etc.","excerpt":"With Monster API users can access powerful generative AI models without the hassle of managing GPU infrastructure or breaking the bank.","categories":["AI Highlights"],"tags":["Generative AI","GPT","GPT-J","H2O AI","LLaMA"],"author_name":"K L Krithika","publish_date":"2023-08-02T17:12:42","publication_year":"2023","word_count":809,"keywords":["data science","Hugging Face","artificial intelligence","machine learning","AWS","AI","ML","LLaMA","computer vision","H2O AI","GPT","GPT-J","RAG","generative AI","Generative AI"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","computer vision","data science","generative AI","Hugging Face","RAG","AWS"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/working-with-generative-ai-just-got-faster\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":20214,"title":"Transforming The Face Of Healthcare One Step At A Time, Predictive Analytics Is The Hottest Trend Now","content":"Imagine a future where each individual will have a healthcare treatment and care plan exclusively crafted for him\/her? Exciting, right? Thankfully, we’re not very far away from that dream. Predictive analytics and Big Data are rapidly transforming the face of healthcare as we speak. Today, as each step in the healthcare chain is digitized, healthcare providers and professionals are accumulating huge amounts of data (patient names, medical history, diseases, prescriptions, diagnostic tests, medical insurance, etc.). However, storing and managing this data in a way that is helpful to the healthcare provider or the patient is immensely challenging because of its sheer size and cost entailed. In fact, according to a McKinsey report (2013), “After more than 20 years of steady increases, healthcare expenses now represent 17.6 percent of GDP —nearly $600 billion more than the expected benchmark for a nation of the United States’s size and wealth.” This emphasizes that expenses in healthcare are rising at an alarming rate and there is an urgent need to implement a smart structure where health records and patient history can be stored in an organized manner, helping to provide both better care and treatment as well as cut down costs. Predictive Analytics is basically a structure of artificial neural networks and decision trees that can predict new patterns or trends in data rooting on the knowledge present in the historical data. By incorporating analytics in healthcare, we will be equipped to predict the future. Vinnie Ramesh, Chief Technology Officer, Co-founder of Wellframe, states: “Predictive analytics is…applying what doctors have been doing on a larger scale. What’s changed is our ability to better measure, aggregate, and make sense of previously hard-to-obtain or non-existent behavioral, psychosocial, and biometric data. Combining these new datasets with the existing sciences of epidemiology and clinical medicine allows us to accelerate progress in understanding the relationships between external factors and human biology—ultimately resulting in enhanced reengineering of clinical pathways and truly personalized care.” Though predictive analytics presents before us a huge opportunity to scale up the quality of healthcare and treatment, we cannot ignore some of the major challenges facing it. One of the primary challenges is implementing analytics within the healthcare systems, which means incurring IT costs, which a lot of providers can’t afford. Let us look at some other hurdles that prevent the adoption of analytics in the field of healthcare: Data structure issues: A bulk of this data in healthcare is unstructured, fragmented, and dispersed. This makes it very difficult to analyze and aggregate data in such format. Furthermore, it is a known fact that data in healthcare is much more diverse as compared to data from other fields. Security and Compliance issues: When data is stored in such large bulk, security becomes a major concern since we’re talking about confidential patient data. Being stored in a centralized system, this data is freely available and hence, becomes highly susceptible to attacks. Data Storage and Transfer issues: Organizations integrating and analyzing data have to bear a considerable amount of data storage and transfer costs. When data is stored in the cloud, there is a different layer of security related to retrieving, transferring, and loading of patient data. Lack of Skilled Professionals: While it is true that technology is evolving and advancing at a rapid pace, the amount of experienced and skilled individuals who are constantly updating themselves in sync with these new trends is considerably low. Managing such large amount of confidential healthcare data demands a certain level of skill and efficiency. Irrespective of these challenges, predictive and big data analytics hold a tremendous potential to transform the face of the healthcare industry. Healthcare organizations that rely on data analytics can improve on areas as R&D, surgery, genome study, and so much more. The better we can understand the medical history and needs of individual patients, the better we can treat them using a personalized treatment approach. The good news is that Big Data and predictive analytics have already been put to good use in the field of healthcare in many countries. Doctors and patients can now easily keep track of their health through Electronic Health Record (EHR), a digital record where every individual patient’s medical history, demographics, diagnostics tests, etc., are stored and maintained. While doctors can easily make the changes in this as they come, it involves no paperwork and one doesn’t have to worry about data replication. Furthermore, EHRs are programmed in a way that they can remind patients about their upcoming tests, or prescription drugs refill. Tools such as Clinical Decision Support (CDS) allow healthcare organizations to analyze medical data in real-time and offer doctors and medical professionals with the necessary advice as they write out prescriptions. Researchers are coming up with personal analytics devices that will continuously gather the patient data and store it in the cloud. Also, symptom calculators are gradually becoming the “recommendation engines” of the healthcare industry enabling patients to self-diagnose their health problems. Suppose you have the flu, you just need to go online and enter your symptoms in the symptom calculator and the software’s algorithm will match you with others who have had similar symptoms and also show you the diagnosis that was most common. Approaches like ask a doctor online, online consultation, digital prescriptions etc have also emerged lately. Charlie Farah, the Asia-Pacific market development healthcare and public sector director of Qlik, stated: “There is a real buzz around India on the absolute benefits analytics and data discovery can deliver to an organization. Wockhardt Hospitals Ltd., one of India’s leading super-specialty care, deployed the Qlik Platform as a key tool in its journey on establishing a companywide adoption of data analytics.” Thus, it is needless to say that with predictive analytics in the scene, the future of healthcare industry sure does look promising.","excerpt":"Imagine a future where each individual will have a healthcare treatment and care plan exclusively crafted for him\/her? Exciting, right? Thankfully, we’re not very far away from that dream. Predictive analytics and Big Data are rapidly transforming the face of healthcare as we speak. Today, as each step in the healthcare chain is digitized, healthcare […]","categories":["IT Services"],"tags":[],"author_name":"Dr Aditi Gupta","publish_date":"2018-01-02T05:54:59","publication_year":"2018","word_count":966,"keywords":["big data","Go","API","AI","neural network","R","Git","RAG","analytics","predictive analytics"],"extracted_tech_keywords":["AI","neural network","analytics","RAG","predictive analytics","R","Go","Git","API","big data"],"url":"https:\/\/analyticsindiamag.com\/it-services\/transforming-face-healthcare-one-step-time-predictive-analytics-hottest-trend-now\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10093801,"title":"The Data Platform for the Era of AI: Introducing Microsoft Fabric","content":"Microsoft has announced the launch of Microsoft Fabric, an end-to-end analytics platform designed to meet the growing data needs of organisations. With the proliferation of data and the increasing importance of AI, we require a unified analytics system to extract insights and drive business value. Microsoft Fabric integrates various data and analytics tools into a single platform, providing a seamless experience for data and business professionals. Lake-centric and open Microsoft Fabric offers a unified experience and architecture that encompasses all the capabilities required for data extraction and presentation. By delivering the platform as a Software as a Service (SaaS), integration and optimisation are automatic, enabling users to derive real business value within minutes. Microsoft Fabric includes OneLake, a multi-cloud data lake that simplifies the creation, integration, management, and operation of data lakes. It reduces data duplication and concerns about vendor lock-in by supporting open data formats and providing shortcuts for easy data sharing across different cloud platforms. Powered by AI Azure OpenAI Service is integrated into Microsoft Fabric at every layer, enabling developers to leverage generative AI and assisting business users in finding insights. With Copilot, users can use conversational language to create data flows, build machine learning models, and more. Copilot automatically inherits an organisation’s security and privacy policies, ensuring data protection. Microsoft Fabric deeply integrates with Microsoft 365 applications, such as Excel, Teams, and PowerPoint, making data easily discoverable and accessible. Users can leverage data in their everyday work and share insights seamlessly across the organisation. Microsoft Fabric is currently in Public Preview, and users can sign up for a free trial to experience its capabilities. Existing Power BI Premium customers can enable Microsoft Fabric through the Power BI admin portal. Microsoft Fabric will be enabled for all Power BI tenants starting from July 1st.","excerpt":"Microsoft has announced the launch of Microsoft Fabric, an end-to-end analytics platform designed to meet the growing data needs of organisations. With the proliferation of data and the increasing importance of AI, we require a unified analytics system to extract insights and drive business value. Microsoft Fabric integrates various data and analytics tools into a […]","categories":["AI News"],"tags":["Microsoft"],"author_name":"K L Krithika","publish_date":"2023-05-23T20:55:00","publication_year":"2023","word_count":297,"keywords":["machine learning","OpenAI","AI","R","ML","RAG","analytics","generative AI","Azure","data lake","Microsoft"],"extracted_tech_keywords":["AI","machine learning","ML","analytics","generative AI","OpenAI","RAG","Azure","R","data lake"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/the-data-platform-for-the-era-of-ai-introducing-microsoft-fabric\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":37804,"title":"Check Out These 5 Data Analytics Job Openings Across India","content":"Analytics India Jobs offers a new platform for the data science community in India by bringing to the fore latest job openings across the country. Here are the latest Data Science job openings across India you can apply: Analytics Intern @ Saras Analytics, Hyderabad Saras Analytics is a data management and predictive analytics company setup to solve data challenges. And, to foster data-driven decision making at small to medium businesses and take initiatives towards offering the highest quality products and services at the most affordable prices. Requirements Well versed with Python\/R, Excel, SQL, Tableau & Google Analytics Excellent written and oral communication skills The role of an intern involves using analytics to solve business challenges for some of the biggest e-commerce brands in India and Middle East Apply here NLP Research Intern @ vPhrase Analytics Solutions Pvt. Ltd, Mumbai vPhrase Analytics is a technology company based out of Mumbai, India whose key offerings include PHRAZOR platform. PHRAZOR gathers data, structures the facts and in the end applies language to present the reader with humanized, targeted narratives. An intern will help develop language models from text data using statistical, deep learning as well as rule-based methods. This is a paid internship and intern would be expected to work full-time out of Mumbai office. Requirements Passionate about learning new methodologies and implementation in linguistics. Analytical mind and problem-solving aptitude. Hands on experience in building ML\/DL\/NLP models in past. Helping our product teams in creating knowledge bases that will power our NLG engine. Building taggers, parsers, knowledge graph-based language models. Apply here Analyst @ Mastercard, Gurgaon As an analyst, the candidate will work closely with Data & Services Advanced Analytics teams and external clients around the world to architect, develop, and maintain advanced reporting and data visualisation capabilities on large volumes of data in order to support consulting projects, model development, model scoring, and campaign measurement. And, translate business requirements into tangible solution specifications and high quality, on time deliverables. Requirements B.Tech or Mathematics. M.S. preferred. 2-5 years of relevant experience in hands-on data programming, querying, data mining and report development using large volumes of granular data to deliver business intelligence and custom reporting solutions in a Microsoft SQL Server environment. Strong hands-on experience on SQL, Tableau, SSIS. Analytical\/Problem Solving. Relevant retail and payments industry experience a plus. Apply here Role In Analytics @ Max Life Insurance, Gurgaon Max Life envisions to be the most admired life insurance company in India by securing the financial future of customers. Max Life Insurance amongst the top 15 BFSI companies to work for in India. Requirements BE\/BTech, MTech, MSc. in Computer science\/Information Technology\/ ECE or other quantitative disciplines from tier 1\/ tier 2 institutes. 2-7 years of hands on work experience in NLP, AI, Machine Learning, Deep Learning Should have developed and deployed NLP based solutions with R( tm, korPus, etc) or  Python( NLTK, Scikit-learn and Spacy Strong familiarity with RNN, LSTM, word2vec and other embeddings, Keras\/Tensor flow. Knowledge of Chabot platforms and development and deployment of at least 1 NLP based Chatbot or other AI, machine learning, or NLP technologies Working with Speech to Text and conversational analysis tools \/Api, such as Dialogflow is a plus Working knowledge of Image matching, and understanding applications of voice and video data is a plus Experience in working with cloud infrastructure and high-performance computing (GPU\/TPU) be a plus. Apply here Digital Data Analyst @ DG7 Solutions Pvt.Ltd, Mumbai DG7 Solutions is an analytics-focused, research-based consulting partner for medium and large organisations offering end-to-end Digital Marketing Solutions. Requirements Work with internal stakeholders to determine ongoing business intelligence needs in order to design, develop and produce ongoing reports and dashboards to support the business. Manage projects related to digital analytics from start to finish including data integration, report and dashboard automation, product analytics and insight, data collection, and product optimisation Assists in identifying opportunities that are most effective at driving conversions, revenue, ROI, and scale across Digital Marketing programs. Build automated dashboards for digital, marketing & customer level insights – to be shared and championed with key stakeholders throughout the business. Web Analytics: In addition to setting up standard web reports, the ideal candidate will continually mine Google Analytics Developing new skills and sharing your knowledge with the Data Team SQL, R, Python or PySpark. Apply here","excerpt":"Analytics India Jobs offers a new platform for the data science community in India by bringing to the fore latest job openings across the country. Here are the latest Data Science job openings across India you can apply: Analytics Intern @ Saras Analytics, Hyderabad Saras Analytics is a data management and predictive analytics company setup […]","categories":["AI Hirings"],"tags":["Mastercard","Python","Tableau"],"author_name":"Ram Sagar","publish_date":"2019-04-16T06:35:34","publication_year":"2019","word_count":717,"keywords":["data science","scikit-learn","Mastercard","Tableau","machine learning","Keras","AI","ML","Python","NLP","deep learning","analytics","spaCy"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","NLP","data science","analytics","Keras","scikit-learn","spaCy"],"url":"https:\/\/analyticsindiamag.com\/ai-hiring\/check-out-these-5-data-analytics-job-openings-across-india\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":46140,"title":"Indian Analytics Industry’s Nobel Prizes Announced: Academic Elegance &#038; Business Knowledge Rules Day 2 at Cypher 2019","content":"Day 2 at Cypher 2019 concluded with Great Learning Data Science Awards Day 2 at Cypher 2019 started with another talented roster of speakers and business leaders. This 5th edition of Cypher has been our biggest and most well-attended summit with 900+ attendees along with 100+ speakers coming together on one single platform for 3 days of hands-on learning and tech talks by best of the breed in AI and analytics industry. HackerU’s Lion Kontorer headlined Day 2 at Cypher 2019 with a talk on AI in the future of Cybersecurity Lion Kontorer from HackerU headlined Day 2 at Cypher 2019 with a very interesting talk on  “AI in the future of Cybersecurity”. Israel-based Kontorer who is the Offensive Security Team Leader with deep experience in penetration testing and crime investigation walked the packed audience through supervised learning, unsupervised learning and reinforcement learning play a critical role in cybersecurity. Amaresh Tripathy, Senior Business Leader at Genpact giving away the award Up next, Genpact’s Amaresh Tripathy, prolific global business leader and also an Adjunct Professor at the University of North Carolina took the audience through Cora —  Genpact’s AI accelerator which takes the heavy-lifting out of machine learning. Not delving into the “next-gen” and buzzworthy applications of AI, Tripathy’s keynote addressed some of the biggest problems businesses face today, such as on real-time claims decisioning, invoices and modularization can deliver real business value and drive AI at scale. Also taking the keynote stage was Srinidhi Rao and Saurav Ghosh of TheMathCompany whose talk was as curious as the title itself — Curious Case of the Missing Gorilla. In their fascinating talk, Ghosh and Rao touched upon the various kinds of biases that are pervasive throughout the machine learning practices. They also have an interesting proposition to usher in a new era of analytics project lifecycle in the form of “data ethnography”, which is the study of the origin of the data. Great Learning’s stalwart Dr PK Vishawanathan took centre-stage to talk about one of the most overlooked contributions of AI by Ronald Fisher – discriminant analysis and its other companion — Support Vector Machines. Dr Vishawanathan also defined the way for future data scientists and analytics professionals, “The need of the hour today is marrying academic elegance with business domain knowledge. It is the time for bilingual people who speak the business lingo and have sound data science concepts”. Meanwhile, senior industry leader and Hike’s VP of AI and data technologies Ankur Narang talked about the challenges and opportunities of implementing quantum computing in AI. Another prolific speaker ISIMA’s co-founder Darshan Rawal highlighted how startups and mid-sized AI companies are ushering in a new revolution in enterprise AI market. “It’s time to wean off the mothership which is Google,” he said. Genpact’s Dimple Arora moderates a panel on AI at scale: challenges in driving adoption across large organizations at Day 2, Cypher 2019 Perhaps one of the most informative sessions turned out to be an interesting panel moderated by Udacity’s Ishan Gupta Global Head, Student Ops, Arvind Rathore Senior Vice President & Global Head of Digital Engineering at Virtusa , Bhargavi Sunkara Head of Corporate Technology at BNY Mellon, Abhijit Shroff Director, Digital Transformation at DXC Technology and Sreekanth Menon Vice President – AI & ML at Genpact. Titled “AI in the real world – Deciphering the hype and blusters,” the panelists deliberated how human in the loop will remain a critical element in AI implementation. Post lunch, a five-member panel moderated by Genpact’s Dimple Arora discussed the challenges AI at scale. Tapati Bandopadhyay of HFS Research, Ramsu Sundararajan of Cartesian Consulting, Rohit Dhameja and Sreekanth Menon at Genpact, Anurag Verma CEO and Co-founder at FN MathLogic, Piyush Chowhan of Arvind Lifestyle Brands deliberated on the hype surrounding AI and why simply knowing Python isn’t enough to make one data scientist. “Today a lot of people think by knowing Python programming or the open source frameworks like Keras, Theano one is doing data science,” Tapati said to a packed audience. Overall, it was an action packed day 2 at Cypher 2019 and the day wrapped up with Great Learning Data Science Awards given away in 11 categories. Stay tuned for Day 3 highlights.","excerpt":"Day 2 at Cypher 2019 started with another talented roster of speakers and business leaders. This 5th edition of Cypher has been our biggest and most well-attended summit with 900+ attendees along with 100+ speakers coming together on one single platform for 3 days of hands-on learning and tech talks by best of the breed […]","categories":["Deep Tech"],"tags":["real time decisioning","Voice Analytics"],"author_name":"Richa Bhatia","publish_date":"2019-09-19T18:05:43","publication_year":"2019","word_count":702,"keywords":["data science","machine learning","Keras","AI","ML","Voice Analytics","RAG","Python","Aim","analytics","real time decisioning","R"],"extracted_tech_keywords":["AI","machine learning","ML","data science","analytics","Aim","Keras","RAG","Python","R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/indian-analytics-industrys-nobel-prizes-announced-academic-elegance-business-knowledge-rules-day-2-at-cypher-2019\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":4036,"title":"Exciting Higher Education Opportunities with Analytics","content":"Quite a few of the working professionals and bright final year students who take INSOFE courses are interested in pursuing higher education (a Masters or a Ph.D.) in a top reputed university.  They have found that they either get into better universities and\/or get funding AFTER going through INSOFE’s rigorous certificate courses than they would BEFORE.  In fact, they have got admission offers from such top universities as Cornell, Northwestern, UT Austin, University of Illinois-Urbana Champaign in the US, National University of Singapore, ETH Zurich (Swiss Federal Institute of Technology), etc., all of which are ranked among the top 25 universities in the world1,2.  There are many questions that go through students’ and working professionals’ minds and I wanted to address them all at one place for a larger reader base. Why should I do Masters at all? In India, most universities have designed their Masters programs as a pre-requisite to Ph.D. for eventually getting into academia or R&D labs, but not to prepare students for the industry.  In fact, most companies state explicitly to the colleges they visit for campus interviews that they only consider undergraduate (Bachelors) students and not graduate (Masters) students.  While I was working on Agni Missile system in DRDO, the policy was to recruit both Masters and Bachelors students at exactly the same grade.  Hence, most engineers interested in industry jobs do not pursue Masters degrees. On the other hand, in the U.S., companies recruit graduates at a higher salary compared to undergraduates because they expect them to solve more challenging problems and lead.  Schools also design their programs accordingly. So, Masters in a good international university is truly a professional enhancement and imparts skills that allow you to see the Bigger Picture.  It can be said that while the Indian universities focus on the DEPTH in a particular subject\/topic, the top global universities focus in addition on the BREADTH and give equal importance to non-technical aspects to build a holistic mindset. Why should I choose Masters in Analytics, in particular? The traditional M.B.A. exposes the student to all functions of an industry and builds intuition and quant skills required to take business decisions.  Analytics exposes the student to all kinds of data generated across the organization and the industry and builds the skills required to gain actionable insights and take informed decisions based on that data.  As Thomas Davenport and Jeanne Harris argue in their book, Competing on Analytics: The New Science of Winning, organizations that do not use analytics to gain insights from their data will fall behind their competitors and eventually lose out.  Analytics is simply a way to make better decisions using data. Thus far, people chose M.S. if they liked technology or M.B.A. if they liked management.  But, what do you do if you like both?  Moreover, in the increasingly data-driven competitive world, there is a strong need for techno-managers.  Fortunately, you do have an excellent option in M.S. in Business Analytics.  It marries the “applied-ness” of M.B.A. with the quantitative rigor of M.S.  That is why in many schools, analytics is offered by business schools in association with engineering schools. What should I do to get in? A good GRE score is mandatory.  Very few schools ask for a GMAT score for M.S. in Business Analytics and most of them will consider a GRE score instead if you request.  A decent score in TOEFL\/IELTS is required.  While not mandatory, however, what would set you apart from others is being able to showcase a good project that highlights both your interest and aptitude in the subject. How difficult is it to get into good schools? It is really not that difficult.  If you want an admission in a top US school, shoot for 320+ in GRE.  A student of ours with an excellent project and 320+ in GRE got admissions from Northwestern, UT Austin, Cornell and three more schools with assistantship in most. However, students with scores of only around 300 in GRE but with good projects also made into Arizona State U, Drexel U and Georgia State U (some very good schools).  I believe 60% or so in academics (aim higher, though) and a score of 300+ in GRE is good enough (again, aim higher) to aspire for a “very good” school, provided you have a good project to showcase.  When I say “very good”, I mean those that are consistently ranked above IITs and IIMs in the international arena1,2.  Do keep in mind that nothing comes easy and for you to be able to do good projects and make up for weaknesses in other areas, you must work hard and learn the concepts well. Another thing I note is that M.S. in Analytics is a lot more easier (at least now; obviously competition is sure to heat up for the top schools as the demand is surging) to get into than M.B.A. and it is mostly less expensive. In fact, the sinking rupee is the only thing you should dread! How should you plan and what are the timelines? A typical INSOFE student taking our certificate program in Analytics spends 50+50 hours on two projects (apart from nearly 300 classroom hours spread across 4-5 months).  I advise those interested in higher education to spend 50+100 hours and do a bit more thorough project.  A good project can take you a long way.  These 5 months or so are a time well-spent. GRE preparation takes 3 months-6 months.  The other parts (essays, recommendations, etc.) take one month. You need to put aside approximately Rs. 50,000 for applications to universities and Rs. 20,000 for tests (GRE, TOEFL, etc.).  Typical Masters last 12-18 months and the fee is between $25-50K.  Living expenses are around $500-$800 per month.  You will perhaps earn the living and partial fee waiver after 6 months, even if you did not get any funding up front.  So, you are looking at an expense of $35-65K. However, as soon as you complete the degree (12-18 months), you can expect $80K+ annual salary. Having gone through the very best schools in both India and the US (NIT, Tiruchirapalli and Carnegie Mellon University), I can confidently say that if you can afford the cost and put in the effort, Masters education in a good U.S. (or some other international universities) is really one of the most exciting things you could explore in life…ever.  But there are no shortcuts to success; you have got to earn your admission through hard work. 1 http:\/\/www.timeshighereducation.co.uk\/world-university-rankings\/2012-13\/world-ranking 2 http:\/\/www.usnews.com\/education\/worlds-best-universities-rankings\/top-400-universities-in-the-world","excerpt":"Quite a few of the working professionals and bright final year students who take INSOFE courses are interested in pursuing higher education (a Masters or a Ph.D.) in a top reputed university.  They have found that they either get into better universities and\/or get funding AFTER going through INSOFE’s rigorous certificate courses than they would […]","categories":["AI Trends"],"tags":["analytics education","data scientist india salary","Insofe","masters in data analytics in india"],"author_name":"Dr. Dakshinamurthy V Kolluru","publish_date":"2013-09-06T07:24:14","publication_year":"2013","word_count":1084,"keywords":["Go","funding","ELT","programming_languages:R","AI","data-driven","data scientist india salary","Insofe","masters in data analytics in india","Aim","analytics","GAN","analytics education","R"],"extracted_tech_keywords":["AI","analytics","Aim","R","Go","ELT","GAN","data-driven","funding","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/exciting-higher-education-opportunities-with-analytics\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10093450,"title":"When will Generative AI Make Investors Smile?","content":"Let’s face it: Generative AI is the belle of the ball in the tech industry and investment circles alike. Its potential to revolutionise various sectors has captured the imagination of entrepreneurs and investors. As a result, a new trend has emerged in the buzzing world of venture capital – the mad rush to invest in generative AI startups. The only problem when it comes to investing in these generative AI startups is the return on investment (ROI). While some sceptics argue that since there is no immediate ROI with generative AI, it’s a risky investment, savvy VCs believe it’s a calculated gamble that could pay off handsomely in the future. Let’s look at this: There are 13 generative AI companies that have become unicorns since last year, with Cohere and Runway joining the club just this month. This is amid the funding droughts, layoffs, and investors demanding profit from companies. Moreover, according to data from PitchBook, around $1.7 billion was generated across 46 deals within the first quarter of 2023 with an additional $10.6 billion worth of deals announced. The generative AI market is expected to hit $109.3 billion by 2030, according to Grand View Research. VCs have never been strangers to taking calculated risks and currently they are pumping money based on the hype around models like Google’s Bard and OpenAI’s ChatGPT. When it comes to this field, a lot of it is probably driven by predictions, ambitions, and probably delusions. Here’s why. Where is the profit in this hype? Firstly, OpenAI is not profitable yet. The Microsoft-backed company is running a loss of around $540 million since it started building ChatGPT. It has only started generating revenue after offering its API, subscriptions, and licensing its products like GPT-4 to its customers through Microsoft’s enterprise and cloud services. The same is the case with NVIDIA. The company that was banking on the gaming industry for money, decided to provide its hardware for generative AI. Voila! The biggest generative AI name in the world, ChatGPT, is powered by the company’s hardware, reaping money. The billions of dollars that companies such as OpenAI have raised, has driven startups to build products similar to ChatGPT. VCs falling into the hype cycle are pouring funds into similarly ambitious startups without realising and analysing what sets the company’s product apart – not every startup can build a ChatGPT and then sell it as well as OpenAI did along with Microsoft. Of course, not every generative AI startup will succeed. Some may fizzle out, while others may stumble upon groundbreaking breakthroughs. But that’s the nature of investment. VCs understand that they must place multiple bets to increase their chances of striking gold. By diversifying their portfolio and placing their chips on generative AI, they’re playing the odds and banking on the future, or are they? Lu Zhang, the founder and partner at Fusion Fund, said the barrier to entry in this segment is still quite low. She further adds the firm has seen an 80% increase in the number of generative AI pitches in the last two months, since the firm invested in You.com. In the grand scheme of things, “the hype surrounding generative AI may be justified”, said Sonya Huang from Sequoia Capital. “If I was a founder in Y Combinator right now, I would 100% be pointing my guns at one of these models and seeing what I can do,” she added, explaining that the rising investments is planting the seed in founders’ minds to build something with generative AI, and she says that we are already seeing the impact of the technology in the current world, and will definitely see more adoption in the future in almost every sector. When will the rewards reap in? While the immediate ROI might not be evident, the long-term potential is too significant to ignore. VCs are seizing the opportunity to shape the future, embracing the uncertainty and the possibility of great rewards. “The world is in dark times right now, and people are looking for something to latch on to that is hope, and generative AI appears to be that,” added Huang. This might be a hint towards the failure of cryptocurrency and blockchain ventures. Colin Treseler, co-founder of Supernormal, a platform that leverages generative AI to summarise online meetings, expressed similar views, “The Web3 hype ended, and these people needed a place to go,” he said. According to a report by Mckinsey, VC investments in AI have grown 13X over the last ten years. And according to PitchBook, VC investment in generative AI has increased by 425 percent since 2020, even when the broader technology market is declining. Gartner predicts that more than 30% of the drug discovery by 2025 will be done by generative AI solutions, which is a rise from almost zero today. And this is just one of the hundreds of use cases. Sure, the immediate ROI might be elusive because generative AI is still in its infancy. Many startups explore the uncharted territories of creativity and problem-solving, and VCs are tapping into that potential for solving specific use cases. Spend, Spend, AI Will Send “Generative AI is well on the way to becoming not just faster and cheaper, but better in some cases than what humans create by hand,” concluded Sequoia Capital’s investment thesis using GPT-3 in September last year. It is clear that these early investments are not just about instant gratification; they are about building a foundation for a future where generative AI becomes indispensable. VCs believe that instead of investing in “boring sectors” that involve a lot of technicalities, now that the generative AI startups are expanding into different verticals like enterprise and several other applications, it is becoming an increasingly attractive stage for them. Most investors that AIM spoke with have said that they seek guidance from several technological experts to evaluate the tech behind the products. As Brett Calhoun puts it, “We are betting on the jockeys, not the horse.” The startups have to build a clear roadmap for generating ROI before asking for large capital from investors. Think about it this way: the technological world is evolving at an unprecedented pace, with digital transformation penetrating every aspect of our lives. Companies across sectors will soon find themselves relying on generative AI products to innovate, automate, and outperform competitors. Those who were visionary enough to back generative AI early on will reap the rewards of their foresight, and that is what the investors are striving to be. So yes, investors are well aware that they are not investing for today but for tomorrow. After all, sometimes the best investments are the ones made in the future, even if they come with a sprinkle of hype.","excerpt":"The only thing that generative AI is not capable of currently is generating money","categories":["AI Features"],"tags":["Generative AI","generative AI unicorns","Startups"],"author_name":"Mohit Pandey","publish_date":"2023-05-16T18:00:00","publication_year":"2023","word_count":1120,"keywords":["Go","ChatGPT","API","OpenAI","AI","generative AI unicorns","Git","RAG","Aim","Startups","generative AI","Generative AI","R"],"extracted_tech_keywords":["AI","generative AI","ChatGPT","OpenAI","Aim","RAG","R","Go","Git","API"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/when-will-generative-ai-make-investors-smile\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":37893,"title":"10 Facts About Programming Languages Only Programmers Know","content":"Programming is an integral part of our IT landscape and we have often observed the shift in the popularity and choice of programming languages. While Java continues to be on top of the stack, Python soars in popularity. This information helps the developer community understand current trends and the language preferred by industry. In our recent study, we noted how Python continues to dominate the market: Python continues to be the tool of choice among data analysts and data scientists and this is reflected in the Indian hiring market as well with 17% jobs listing the language as a core capability. In this article, we list down 10 interesting facts about programming languages that will help the developer community make more informed choices. 1| Python Is Ruling According to the Python Developers Survey 2018 results, 84% of Python users use Python as their main language, on the other hand, 16% of the users use Python as their secondary language. The usage of Python as the main language is up 5 percentage points from 79% in 2017. One interesting fact is that half of all Python users also use JavaScript. Also, 71% of the respondents of the survey install Python from python.org or with OS-provided package managers like APT and Homebrew. 2| Diversity In Languages There are a total of more than 690 notable programming languages including the current and the historical ones. If there be a country where the language used will be the programming languages, then it will be in the 3rd rank after Papua New Guinea and Indonesia. People often think, learning a programming language will be harder until they started to learn. It basically goes with the term “Practice makes a programmer Perfect.” 3| Python As Hobby According to the survey, the language Python is mainly used for game development as a hobby, while the other development types such as machine learning, web development, data analysis and network programming are mainly performed as primary activities. 4| End Of Life For Python 2.7 The survey showed that there are 84% of Python 3 main interpreter users and only 16% for Python 2. Python 3 had an increase in users from 75% in 2017. As the use of Python 3 continues to grow rapidly, the use of Python 2 is declining as it is no longer actively developed, also does not get new features, and thus the end of life has been moved to 2020. However, the declaration does not guarantee that bugfix releases will be made on a regular basis, but it should enable volunteers who want to contribute bugfixes for Python 2.7 and it should satisfy vendors who still have to support Python 2 for years to come. 5| Programming And Coding Are Two Different Terms Programmers working on the computer term coding and programming synonymously and use the terms interchangeably. Coding means simply writing codes from one language to another, while programming means to program a machine with a set of instructions. A coder is someone who translates logics into a language which a machine will understand, on the other hand, a programmer deals with much more than just writing codes. Click here to read more about it. 6| Python In Professional Use The survey shows that data analysis has become more popular than web development, growing from 50% in 2017 to 58% in 2018. Machine learning also grew by 7 percentage points. These types of development are experiencing faster growth than web development, which has only increased by 2 percentage points when compared to the previous year. In terms of Python users using Python as their secondary language, DevOps has overtaken web development. 7| PHP Is Not A Real Programming Language PHP‘s creator Rasmus Lerdorf had no intention to create an entirely new programming language. He had created the language, or interface at the time only for the purpose of managing his personal website. While PHP is recognised today as a recursive backronym for “PHP: Hypertext Preprocessor,” it was originally an acronym for “personal home page,” referring to its use on Lerdorf’s personal website. As the language evolved, so did its meaning. Click here to read more about it. 8| The Hybrid Language Scala or the Scalable Language is a hybrid of Object Oriented Programming (OOP) and Functional Programming so, besides having all the OOP features of Java, the language also includes features of functional programming languages like Standard ML, Haskell, etc. Scala has been used by the developers of Netflix, Pinterest, etc. and has become the prominent language for Big data processing. 9|  Null And NaN In the JavaScript language, the definition of Null is the total absence of meaningful value and it is counted as an object in the programming language. On the other hand, NaN or Not-a-Number is actually counted as a number but that does not mean it is equal to anything and in order to confirm that something is NaN, then it must be represented by the function isNaN(). Click here to read more about it. 10| TypeScript 3.4 TypeScript, the language that is built on JavaScript has lately shown notable growth.TypeScript 3.4 has been released shortly which has got some important features and updates. It has a faster subsequent build with the — incremental flag, higher order type inference from generic functions, new syntax for ReadonlyArray, etc. Click here to read more about it.","excerpt":"Programming is an integral part of our IT landscape and we have often observed the shift in the popularity and choice of programming languages. While Java continues to be on top of the stack, Python soars in popularity. This information helps the developer community understand current trends and the language preferred by industry. In our […]","categories":["AI Trends"],"tags":["PHP","Programming Languages","Python","TypeScript"],"author_name":"Ambika Choudhury","publish_date":"2019-04-17T07:24:44","publication_year":"2019","word_count":895,"keywords":["Go","machine learning","AI","ML","TypeScript","Programming Languages","Python","Ray","JavaScript","R","Java","PHP"],"extracted_tech_keywords":["AI","machine learning","ML","Ray","Python","R","JavaScript","TypeScript","Go","Java"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/10-facts-about-programming-languages-only-programmers-know\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10089780,"title":"SVB Fall Creates Ripple Effect, Indian IT Caught in the Wave","content":"The repercussions of the SVB crisis on the Indian economy run deeper than initially assumed. While most eyes had laser focus on YC-backed Indian startups, which reportedly have funds stuck, there’s more where that came from. Only yesterday, the shares of a bunch of Indian IT companies such as Wipro, TCS, Infosys and Mphasis were all at their 52-week low. Analysts at JP Morgan had shared a note stating Indian IT giants TCS and Infosys have the highest exposure to regional banks in the US such as Silicon Valley Bank, or SVB. Foreign banks are money makers for Indian IT Regional banks like SVB account for between 2–3% of Indian IT companies’ revenue explaining that the exposure to SVB could mean between 10–20 base points for TCS, Infosys and LTIMindtree. The note added that all three companies might need to set aside provisions in the fourth quarter for a contingency. “The collapse of SVB, Signature Bank and concerns of liquidity across the US and the European Union can further soften tech spends by banks over the short term in a year with slowing growth in bank tech budgets”, it went on to explain. Rajeev Chandrasekhar, Indian MoS for IT, Electronics and Skill Development, Source: rajeev.in While speaking to Business Today, Rajeev Chandrasekhar, the Indian Union Minister of State for IT admitted that there may be an impact but only time could tell just how much it would be. “I understand that the banking and financial sector is a major contributor to Indian IT companies. We need to let this play out. I feel like it might not be an existential type of problem for Indian IT. Certainly, one sector may show degrowth but other sectors may grow”, he said. According to data from financial advisory firm Celent, North American banks dominate tech investments in the retail banking sector globally and have spent more than USD 82 billion on IT budgets in 2022 out of a total global expenditure of USD 250 billion. ‘BFSI is an important sector for Indian IT and the ongoing chaos may affect new deals going forward for Indian IT, especially in the case of Tier 1 banks like UBS which acquired Credit Suisse which in itself is a Tier 2 bank. The exposure is between 2–3% for regional banks,’ Pareekh Jain, CEO and lead analyst at Pareekh Consulting and EIIRTrend stated. Jain spoke about the dangers of the abruptness with which the ripples had reached the banking sector. “Up until now, the banking sector was immune to the downturn and the tech sector was in focus mainly but this uncertainty has now spiralled into other sectors. If you saw the budgets for all banks this year, all the banks in the top five had increased their IT budgets except for the World Bank. Mortgages had looked like the only problem but now the whole sector is looking shaky”, he explained. The pluses However, there were some blessings like Chandrasekhar mentioned. “On the brighter side, just like the previous economic crisis, many of these bank captives may be acquired to save cost. Like TCS acquired a Citigroup Global Services captive in a cash deal in 2008. In the medium term, this will mean that there will be more outsourcing in the future,’ he said. With regards to Indian Global Capability Centers, or GCCs, Jain said that the expected impact will probably be on the lower side. India has the lion’s share when it comes to the GCCs across the world, with most notable companies having set up GCCs in India. For instance, both Credit Suisse and UBS have GCCs in India with over 9000 and 7000 employees, respectively. “If you do a cost-comparison between the global GCCs and Indian GCCs, Indian GCCs are far behind. So, I don’t expect a lot of firing to happen. It’s just that fresh employment may be affected. So, Indian GCCs are last in the pecking order after service providers”, Jain stated. Source: Free Press Journal Risk of spreading The sudden manner in which SVB, the 16th largest US bank, crumbled followed by the hurried sale of Credit Suisse, one of the most influential banks in history, to UBS is a distress signal for the banking sector. There are fears that the snowballing of the crisis could end up bringing the world economy to a grinding halt exactly like the fall of Lehman Brothers in 2008 did. As quick as the Federal Reserve acted, there is tension because of contagion worries. SVB’s fallout was followed in quick succession by Signature Bank within a week. Moody’s has also placed six other banks, namely, Western Alliance Bancorp, Intrust Financial Corp, UMB Financial Corp, Zions Bancorp and Comerica Inc. under review. Regional US banks have been hit the hardest with shares of First Republic tumbling more than 60%. Following the SVB episode, major US banks ended up losing nearly USD 90 billion in stock market value the following Monday. At the moment, to Indian IT, the risk is less and maybe only seen in the future but the fear is that the banking system itself may fall like the blocks of Jenga.","excerpt":"Jain spoke about the dangers of the abruptness with which the ripples had reached the banking sector.","categories":["IT Services"],"tags":[],"author_name":"Poulomi Chatterjee","publish_date":"2023-03-21T18:01:35","publication_year":"2023","word_count":857,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","Rust","GAN","R","programming_languages:Rust","startup"],"extracted_tech_keywords":["AI","R","Go","Rust","GAN","startup","programming_languages:R","programming_languages:Go","programming_languages:Rust"],"url":"https:\/\/analyticsindiamag.com\/it-services\/svb-fall-creates-ripple-effect-indian-it-caught-in-the-wave\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10103222,"title":"Google Delays Gemini While Bard Continues to Hallucinate","content":"Google has postponed the release of Gemini to the first quarter of the next year, according to sources cited by The Information. The delay suggests that Google is facing challenges in the development of Gemini, putting it in a catch-up position with OpenAI. Initial expectations had pointed to a fall release for Gemini, but it seems Google failed to deliver that. Google has created multiple iterations of Gemini to address varying tasks based on their complexity. External developers have already examined smaller versions of the model, quantified by the number of parameters or calculations they perform. However, the company is still in the process of completing the main and largest version of Gemini, the report added. At the recent earnings call, Google chief spoke about Project Gemini, briefly. “We are just really laying the foundation of what I think of as the next-generation series of models we’ll be launching all throughout 2024,” shared Pichai, hinting that it would be multimodal – highly efficient with the tool and API integrations – and is going to be available in various sizes and capabilities. He said that Gemini would be used immediately across all their products internally as well as bringing it out for both developers and cloud customers via Vertex AI. No Improvement in Bard ​​In the meantime, Google’s counterpart to ChatGPT, Bard, has struggled to gain traction. This poses a challenge for Google, as ChatGPT’s numerous users contribute valuable data crucial for OpenAI to monitor and enhance the quality of its products. ChatGPT has over 100 million weekly users as of now. Furthermore, compounding Google’s challenges, according to Vectara’s Hallucination Leaderboard, both Google’s PaLM 2 and PaLM Chat ranked the lowest, with hallucination rates of 12.1% and 27.2%, respectively, in contrast to GPT-4, which has a hallucination rate of 3%.","excerpt":"Bard hasn’t improved on Hallucinations.","categories":["AI News"],"tags":["Gemini"],"author_name":"Siddharth Jindal","publish_date":"2023-11-17T17:01:16","publication_year":"2023","word_count":299,"keywords":["Go","ChatGPT","Gemini","API","OpenAI","AI","Modal","RPA","GPT","R","llm_models:GPT"],"extracted_tech_keywords":["AI","ChatGPT","OpenAI","R","Go","API","GPT","RPA","Modal","llm_models:GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/google-delays-gemini-while-bard-continues-to-hallucinate\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10171550,"title":"IBM Plans ‘World’s First’ Fault-Tolerant Quantum Computer by 2029","content":"IBM has announced plans to build the world’s first large-scale, fault-tolerant quantum computer, with a target delivery by 2029. The system, named IBM Quantum Starling, will be developed at a newly established IBM Quantum Data Center in Poughkeepsie, New York. According to IBM, Starling is projected to perform 20,000 times more operations than current quantum systems. “IBM is charting the next frontier in quantum computing,” said Arvind Krishna, Chairman and CEO, IBM. “Our expertise across mathematics, physics, and engineering is paving the way for a large-scale, fault-tolerant quantum computer — one that will solve real-world challenges and unlock immense possibilities for business.” Starling will be capable of executing 100 million quantum operations using 200 logical qubits. It will lay the foundation for a follow-up system, IBM Quantum Blue Jay, which will operate with 2,000 logical qubits and run 1 billion operations. To store the computational state of Starling, IBM claims it would require memory exceeding that of a quindecillion (10⁴⁸) of today’s most powerful supercomputers. Logical qubits, formed from multiple physical qubits, are designed to detect and correct errors during computation. This is a core requirement for scaling quantum systems to handle real-world applications such as drug development, materials science, chemistry, and optimisation. “Creating increasing numbers of logical qubits capable of executing quantum circuits, with as few physical qubits as possible, is critical to quantum computing at scale,” said IBM. Its newly released Quantum Roadmap outlines the milestones leading up to Starling. This year, the company plans to launch IBM Quantum Loon, which will test components necessary for the qLDPC code, including long-distance qubit couplers. IBM is set to release Quantum Kookaburra in 2026, the first modular processor that combines quantum memory with logic operations. By 2027, IBM intends to introduce IBM Quantum Cockatoo, designed to connect Kookaburra modules into a larger network. This approach allows the system to scale without relying on impractically large chips. IBM is addressing fault tolerance through a new architecture using quantum low-density parity check (qLDPC) codes. IBM says this method significantly reduces the number of physical qubits required, cutting overhead by around 90% compared to other leading codes. The company also released two technical papers detailing the approach. One paper outlines how qLDPC codes can improve instruction processing and operational efficiency. The other describes how error correction and decoding can be handled in real-time using classical computing resources. IBM said the success of this architecture depends on its ability to suppress enough errors, apply universal instructions, and scale modularly while remaining resource-efficient. The company stated that code development often involves leveraging pre-existing, thoroughly tested components and integrating established error correction methodologies rather than creating everything from scratch. IBM already operates a large fleet of quantum systems worldwide and sees Starling as the next step in transitioning from experimental devices to scalable infrastructure.","excerpt":"The company is addressing fault tolerance through a new architecture using quantum low-density parity check (qLDPC) codes.","categories":["AI News"],"tags":["IBM"],"author_name":"Siddharth Jindal","publish_date":"2025-06-10T17:19:29","publication_year":"2025","word_count":468,"keywords":["programming_languages:R","AI","Scala","RAG","Aim","IBM","programming_languages:Scala","R","emerging_tech:quantum computing"],"extracted_tech_keywords":["AI","Aim","RAG","R","Scala","programming_languages:R","programming_languages:Scala","emerging_tech:quantum computing"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/ibm-plans-worlds-first-fault-tolerant-quantum-computer-by-2029\/","complexity_score":4,"technical_depth":8,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":8246,"title":"Critical Success Factors for Analytical Models: Some Recent Research Insights","content":"Big data & Analytics: terms that frequently pop up in newspapers, magazines, airports or even during pub chats to pimp a conversation. These days, everybody talks about it, but only few are actually doing it successfully! One of the reasons is that firms often lack a clear insight into the critical success factors for building actionable analytical models. Hence, in this column, we provide some recent research insights based upon partnerships we initiated with firms world-wide, and which are further elaborated on in the new E-learning course Advanced Analytics in a Big Data World. In order to be successful, an analytical model needs to satisfy various requirements. A first key requirement is business relevance. The analytical model should solve the business problem that it was developed for! It makes no sense to have a high-performing analytical model that was sidetracked from the original business problem. In other words, if the business problem is detecting insurance fraud, then the analytical model must be sure to detect insurance fraud. Obviously, this requires a thorough business knowledge and understanding of the problem to be addressed before any analysis can start. Some example kick-off questions are: how do we define, measure and manage fraud? Another important success factor is statistical performance and validity. In other words, the analytical model should make sense statistically. It should be significant and provide good predictive or descriptive performance. Depending on the type of analytics, various performance metrics can be used. In customer segmentation, statistical evaluation measures will contrast intra-cluster similarity with inter-cluster dissimilarity. Analytical churn prediction models will be evaluated in terms of their ability to assign high churn scores to the most likely churners. Interpretability refers to the fact that the analytical model should be comprehensible or understandable to the decision maker (e.g. marketer, fraud analyst, credit expert). Justifiability indicates that the model is in accordance with the expectations and business knowledge of the expert. Both interpretability and justifiability are subjective and depend on the knowledge and experience of the decision maker. Both often need to be balanced against statistical performance, which implies that complex, non-interpretable models (e.g. neural networks, random forests, …) are often better performing in a statistical sense. In settings like credit risk modeling, interpretability and justifiability are very important because of the societal impact of these models. However, in settings like fraud detection and marketing response modeling, they are typically less of an issue. Operational efficiency relates to the effort that is needed to evaluate, monitor, backtest or rebuild the model. From this perspective, it is quite obvious that a neural network or random forest is less efficient that e.g. a plain vanilla regression model or decision tree. In settings like credit card fraud detection, operational efficiency is very important because a decision should be made within a few seconds after the credit card transaction was initiated. Economical cost refers to the cost that is needed to gather the model inputs, run the model and process its outcome(s). Also the cost of external data and\/or models should be taken into account here. This will enable you to calculate the economic return on the analytical model, which is typically not a straightforward exercise. Finally, regulatory compliance is becoming more and more important. This refers to the extent to which the model is compliant with regulation and legislation. In a credit risk modeling setting, it is important that the models are compliant with the Basel II and III regulations. In an analytical insurance setting, the Solvency II accord must be respected. To conclude, in this blog article we briefly zoomed into the critical success factors for building analytical models. As already mentioned, the importance of each of them depends on the application field in which you are working. For more information, we are happy to refer you to the new E-learning course Advanced Analytics in a Big Data World. Self-Paced E-learning course: Advanced Analytics in a Big Data World See: https:\/\/support.sas.com\/edu\/schedules.html?ctry=us&id=2169 The E-learning course starts by refreshing the basic concepts of the analytics process model: data preprocessing, analytics and post processing. We then discuss decision trees and ensemble methods (bagging, boosting, random forests), neural networks, support vector machines (SVMs), Bayesian networks, survival analysis, social networks, monitoring and backtesting analytical models. Throughout the course, we extensively refer to our industry and research experience. Various business examples (e.g. credit scoring, churn prediction, fraud detection, customer segmentation, etc.) and small case studies are also included for further clarification. The E-learning course consists of more than 20 hours of movies, each 5 minutes on average. Quizzes are included to facilitate the understanding of the material. Upon registration, you will get an access code which gives you unlimited access to all course material (movies, quizzes, scripts, …) during 1 year. The E-learning course focusses on the concepts and modeling methodologies and not on the SAS software. To access the course material, you only need a laptop, iPad, iPhone with a web browser. No SAS software is needed. See https:\/\/support.sas.com\/edu\/schedules.html?ctry=us&id=2169 for more details.","excerpt":"Big data & Analytics: terms that frequently pop up in newspapers, magazines, airports or even during pub chats to pimp a conversation. These days, everybody talks about it, but only few are actually doing it successfully! One of the reasons is that firms often lack a clear insight into the critical success factors for building […]","categories":["AI Features"],"tags":["churn prediction model"],"author_name":"Bart Baesens","publish_date":"2015-11-10T10:45:48","publication_year":"2015","word_count":832,"keywords":["big data","Go","programming_languages:R","AI","neural network","ML","RAG","analytics","churn prediction model","R","fraud detection"],"extracted_tech_keywords":["AI","ML","neural network","analytics","RAG","fraud detection","R","Go","big data","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/critical-success-factors-for-analytical-models-some-recent-research-insights\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10060338,"title":"How can Indian companies capitalise on the Green tech industry","content":"The Indian tech industry is predominantly influenced by green energy. Under this, wastewater treatment, e-vehicles, and water treatment are the critical market segments. Green-tech has enormous potential because the 100% FDI approval has paved the way for a huge capital influx. For example, in the first quarter of FY21, India received investments worth $6.6 billion. Additional support from government policies like the National Electric Mobility Mission Plan (NEMMP) and Faster Adoption and Manufacturing of Electric Vehicles (FAME) helped electric vehicles spread like wildfire in India. Also, investments via public-private partnerships (PPP), tax reductions, and exemptions from customs duty will further boost the sales of e-vehicles in the country. However, even though steps are taken, and policies are rolled out, it is not black and white. The Indian green tech market faces several issues like the unavailability of a unified policy structure and accessing finance. Although there are policies in different industry segments, it lacks coordination between the centre and the state. As a result, green tech startups often find it difficult to get early-stage funding. Moreover, grid integration, scheduling, forecasting, and power purchase agreements (PPA) are major challenges that restrict India from unlocking its green potential. Industry insights India has been ranked as fourth in wind power, fifth in solar power, and fourth in renewable power installed capacity as of 2020. Also, India retained its rank as third on the EY Renewable Energy Country Attractive Index last year. Meanwhile, renewable energy will reach 55% of the total installed capacity by 2030. The government also plans to make one ‘‘green city” in each state using renewable energy. This “green city” will use solar rooftop systems for houses, develop solar parks outside the city, and roll out e-vehicles as a means for public transport. Although there has been a steady development in this sector, startups can now be confident after the Ministry of Commerce and Industry reported that approximately 91,500 patents were granted during 2016-17 and 2020-21. The government also mentioned that it would clear almost 25,000 applications in 2022. To put things into perspective, almost 61,000 patents were approved for green tech; out of these, 90% were for waste management and alternative energy production. The remaining green tech patents were sanctioned for energy conservation (2,555), transportation technologies (2,481), nuclear power generation (1,079), and agriculture-and-forestry (161). Patents are crucial for the green tech sector because it helps invent and produce eco-friendly technologies, which directly affects society by increasing wealth and reducing high energy usage. In addition, allowing or approving patents acts as a catalyst for innovation. When countries fast-track green patent applications, it works as an incentive and helps achieve economic and environmental sustainability. Usually, patents for green tech depend on “hard law”, which includes laws, treaties, and regulations. How is tech deployed? Artificial intelligence (AI) is widely used in green tech because it can learn through experience, collect massive amounts of data from various sources, rectify human errors, and take actions based on its conclusions. Green-tech companies work towards reducing their carbon footprint by deploying AI. However, if carbon offset prices increase in the future, AI will help cut down costs. Although AI is the fan favourite, the Internet of Things (IoT) and machine learning (ML) are also used. IoT is commonly used in waste management for route optimisation, which helps in reducing fuel consumption while dumping throughout the city. Similarly, gateways bridge the gap between the IoT platform and sensors and transfer data to the cloud. Meanwhile, machine learning applications are used for predictive analytics. The machine learning algorithms supervised the equipment and frequently checked whether it ran properly or needed repair. Due to this, workers can avoid expensive breakdowns and save time on unnecessary maintenance. ML helps renewables to become more reliable, affordable, and desirable. Green-tech companies use AI-powered data engineering to track emissions throughout their carbon footprint. It collects data from operations, IT equipment, materials and components. AI can be used to exploit data from satellites, and after layering intelligence on the data, it can accurately generate missing data and estimate the certainty of results. Predictive AI is used to forecast emissions, reduce efforts, and develop new carbon reduction procedures, therefore, resulting in achieving reduction targets easily. Companies can use AI to get detailed insight into every value chain segment; meanwhile, prescriptive AI and optimisation immediately improve efficiency in production, transportation, and cutting costs. To be blunt, AI helps green tech companies effectively reduce environmental impact and ease up the financial burden caused by the COVID-19 pandemic. AI, ML, and IoT are key technological tools green tech companies use to provide full energy and carbon transparency, accounting for full-stack supply chains, mining the waste reproduced by consumer gadgets, and finally, understanding rebound effects. In a turn of events, the global pandemic has brought about some positive changes in the green tech industry. The growing adoption of e-vehicles (EVs) and more green energy usage are the new mix to deal with the environmental problems. In addition, EVs will have a critical role to play as it is seen as an escape route for reducing oil import costs. Finally, electric mobility will be the key facilitator for sustainable growth in a post-pandemic India.","excerpt":"Green-tech companies work towards reducing their carbon footprint by deploying AI.","categories":["AI Features"],"tags":[],"author_name":"Akashdeep Arul","publish_date":"2022-02-12T12:00:00","publication_year":"2022","word_count":863,"keywords":["Go","API","machine learning","artificial intelligence","AWS","AI","R","ML","analytics","predictive analytics"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","analytics","predictive analytics","AWS","R","Go","API"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-can-indian-companies-capitalise-on-the-green-tech-industry\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10060105,"title":"A beginner’s guide to image processing using NumPy","content":"Mostly we find the uses of NumPy in the problems where we are required to perform mathematical and logical operations on different sizes of arrays. Since images can also be considered as made up of arrays, we can use NumPy for performing different image processing tasks as well from scratch. In this article, we will learn about the image processing tasks that can be performed only using NumPy. This will be helpful for beginners to understand image processing from its very basics. The major operations to be performed using NumPy are listed below, which we will cover in this article. Table of contents Loading imageCropping imageColours separation Transformation  Grayscale conversion Image segmentation Let’s start with importing libraries and loading a random image. Loading image import numpy as np import matplotlib.pylab as plt %matplotlib inline image = plt.imread(\"\/content\/drive\/MyDrive\/Yugesh\/image processing using numpy\/images.jfif\") print(image) Output: Here we can see the raw form of the image. Since mathematically the images are made up of pixel values in the above output we can see that we have some numbers which define the colours in the image and the image is basically an array or NumPy array. We can also plot the image using the matplotlib library. imgplot = plt.imshow(image) Output: Cropping image After loading the image we are ready to perform actions on the image. As in very basic we can perform basic crop operations on our image. For NumPy, crop operation can be performed by slicing the array. crop_img = image[20:199,:200,:] imgplot = plt.imshow(crop_img) Output: Here we can see that we have cropped our image. Now we can move forward to our next image processing step. Colour separation Since we know that every image is made of pixel values and these pixel values represent three integers that are known as the RGB value of its colour. To separate the image in these colours we are required to pull out the correct slice of the image array. fig, axs = plt.subplots(nrows=1, ncols=3, figsize=(20,8)) for c, ax in zip(range(3), axs): rgb_img = np.zeros(image.shape, dtype=\"uint8\") rgb_img[:,:,c] = image[:,:,c] ax.imshow(rgb_img) ax.set_axis_off() Output: Here in the output, we can see that we have separated the RGB of the image for this we mapped the values in the range 0 to 1 and cast to type uint8. Let’s move toward our next step of image processing. Transformations In this step, we will perform the colour transformation. For this purpose, we can treat the image pixel as a point in space. Treating this way to image pixels allows us to perform a transformation to the colour point. Rotating the colour point can be an example of the above statement. Here we are applying Numpy’s Einstein notation function which is a method of applying a rotation matrix, pixel-wise, to the image. def do_normalise(image): return -np.log(1\/((1 + image)\/257) - 1) def undo_normalise(image): return (1 + 1\/(np.exp(-image) + 1) * 257).astype(\"uint8\") def rotation_matrix(theta): return np.c_[ [1,0,0], [0,np.cos(theta),-np.sin(theta)], [0,np.sin(theta),np.cos(theta)] ] img_norm = do_normalise(image) img_rot = np.einsum(\"ijk,lk->ijl\", img_norm, rotation_matrix(np.pi)) img = undo_normalise(img_rot) imgplot = plt.imshow(img) Output: Here in the output, we can see that applied sigmoid to colour space worked and we are applying the rotation of the colour of the pixels continuously. Now in the next step, we will look at how we can convert an image to a grayscale image. Grayscale conversion We can also use NumPy for transforming the image into a grayscale image. By Taking the weighted mean of the RGB value of the image we can perform this. rgb_weights = [0.2989, 0.5870, 0.1140] grayscale_image = np.dot(image[...,:3], rgb_weights) imgplot = plt.imshow(grayscale_image) Output: Here is the image of the output of our grayscale conversion process. Let’s move on to our next step of image processing. Image segmentation This is one of the most used image processing steps where we segment different regions of images. There are various ways to do so like foreground and background. For example, in this article, we are going to see how we can perform segmentation by converting the image into grayscale and finding a threshold. Pixels in the image that are above the threshold are in one region and others are in another region. def simple_threshold(image, threshold=128): return ((image > threshold) * 255).astype(\"uint8\") thresholds = [100,120,128,138,150] fig, axs = plt.subplots(nrows=1, ncols=len(thresholds), figsize=(20,5)); gray_im = to_grayscale(image) for t, ax in zip(thresholds, axs): ax.imshow(simple_threshold(gray_im, t), cmap='Greys'); ax.set_title(\"Threshold: {}\".format(t), fontsize=20); ax.set_axis_off(); Output: Here in the above output, we can see that we have segmented the image into two regions using different threshold values. Final words In this article, we have discussed different tasks of image processing that we have performed using the NumPy library. Also, we have used the matplotlib library for the visualization of images after processing. By Looking at the above points, we can say that we can perform other tasks as well by just using some other logic. References Link for the codesNumpy documentation","excerpt":"Since images can also be considered as made up of arrays, we can use NumPy for performing different image processing tasks as well from scratch. In this article, we will learn about the image processing tasks that can be performed only using NumPy.","categories":["Deep Tech"],"tags":["AI (Artificial Intelligence)","Data Science","Data Scientist","Deep Learning","Machine Learning","numpy","Python"],"author_name":"Yugesh Verma","publish_date":"2022-02-09T12:00:00","publication_year":"2022","word_count":806,"keywords":["Go","NumPy","TPU","programming_languages:R","Machine Learning","programming_languages:Go","numpy","data_tools:NumPy","Python","Ray","Matplotlib","Deep Learning","Data Science","Data Scientist","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["Ray","NumPy","Matplotlib","TPU","R","Go","programming_languages:R","programming_languages:Go","data_tools:NumPy"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/a-beginners-guide-to-image-processing-using-numpy\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10062205,"title":"Full-time data science courses vs online certifications: What’s best for you?","content":"On the back of massive digitisation, data science has emerged as a lucrative career option in a short time. Forward-looking companies worldwide are on the lookout for digital talents to make the best of their data to optimise business outcomes. However, being a developing field undergoing a tectonic shift, the demand-supply gap for IT talents is widening. To plug the gap, India’s leading institutions such as IITs and IIMs have introduced data science courses at the undergraduate and postgraduate levels. However, not everyone can pursue full-time courses due to budget and time constraints, lack of access, and many other limiting factors. This is where online certifications courses come into play. Top Postgraduate Data Science Programmes In India – AIM Ranking 2021 >> Below, we look at the pros and cons of full time and part-time courses in data science to help you make an informed decision. Full-time courses Pros: Good domain knowledge: Most data science companies look for experienced candidates, and unless you don’t have good domain knowledge, your chances of landing a job in a top drawer company are slim, especially if you are a fresher. “If a person already has a full-time statistics or similar degree and he\/she just wants to learn a new topic, say NLP or computer vision, then a certification would be a good choice. On the other hand, if a person does not have a statistics or similar degree, then I would highly recommend pursuing full-time courses ( two years, three years or more). In a way, I view full-time courses akin to a foundation of a building. The online certificates are like floors built on top of the foundation, but they can’t be the foundation,” said Venkat Raman, co-founder of Aryma Labs. Full-time courses will help you build your domain knowledge from the ground up. “I think full-time courses are better than part-time certifications. In a full-time course, one can get detailed knowledge of the topic. Moreover, the courses usually include student-faculty interactions, live projects, assignments, workshops, presentations etc which play a very important role in the students’ growth,” said Priyabrata Mishra, who is pursuing an integrated Masters in mathematics and computing from BIT. Well-structured curriculum: Data science is a complex field, and a good grip on statistics, programming languages, as well as business acumen will take you a long way. Building skills in multiple areas need time and effort (read discipline). Full-time courses have a well-thought-out curriculum curated by subject matter experts keeping the students’ learning curve in mind. Active learning: In-person full-time courses in data science allow students to interact with the faculty and clear doubts in real-time. Brainstorming with your peers to solve real-world problems and tinkering with data are the best ways to get the data science concepts down cold. Employer preference: Generally, employers prefer students who have pursued full-time courses. However, Aryma Labs’ Raman said there are always exceptions to the rule. “Someone might not have a full-time statistics degree, but they might have learnt things on the job or have done reading on their own. In such cases, we do have to give them an opportunity. As a data scientist recruiter myself, I am open to such candidates,” he added. Cons Time-consuming: More often than not, working professionals looking to upskill may not always have the luxury to put their careers on hold to go for a full-time two year or three-year course. In such cases, online courses seem to be the best option. Expensive: Full-time courses at premier institutes are costly, mainly because of the brand name, the top-tier faculty ( professors and industry partners) etc. Meanwhile, online courses are often much cheaper and not geographically limited. Online certifications Pros Self-paced: Online courses offer the flexibility of learning at your own pace and are particularly helpful for working professionals. “For a person who is working, he\/she may have time constraints to join a full-time course. In this case they can always opt for a certification course to learn new things and enhance their skills,” said Mishra. Remote: Online courses can be accessed from anywhere in the world. Institutions that offer full-time courses in data science are often based out of tier-1 cities in India. With online courses, students from anywhere in India can upskill themselves or change their career trajectory at any time. All you need is a laptop and an internet connection. Shorter duration: Typically, the duration of online courses in data science falls in the range of six months to a year. For example, suppose a business analyst with knowledge of basic statistical concepts and programming languages wants to upgrade to a predictive analytics role. In that case, a relevant part-time certification course can facilitate it. Cons Superficial: Data science is a vast field, and online certification courses just scratch the surface because of shorter durations. Passive learning: Online courses are a mix of pre-recorded videos and a few live sessions to interact with faculty. The structure of such courses are not conducive to real-time doubt clearance, and a lack of peer interaction can be disincentivising. Checklist Raman has a few recommendations for candidates going for online certifications. Find out whether the certificate has value in the marketplace. Before enrolling, get the curriculum vetted by good data scientists. Sound out people who have already done the course: “Did the certification help you in landing the job”, “Did the certification help in broadening the knowledge” and “Did the certification help you in solving the business problem” are some of the relevant questions to ask..","excerpt":"The online certificates are like floors built on top of the foundation but they can’t be the foundation.","categories":["AI Trends"],"tags":["AI (Artificial Intelligence)","Learn Data Science","Machine Learning"],"author_name":"Sreejani Bhattacharyya","publish_date":"2022-03-07T15:00:00","publication_year":"2022","word_count":919,"keywords":["data science","Go","AI","R","Machine Learning","computer vision","NLP","Aim","analytics","Learn Data Science","Tecton","predictive analytics","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","NLP","computer vision","data science","analytics","Aim","Tecton","predictive analytics","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/full-time-data-science-courses-vs-online-certifications-whats-best-for-you\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10041036,"title":"Microsoft’s Latest Announcements, OpenAI’s Startup Fund And More In This Week’s Top News","content":"Microsoft kicked off its annual developer’s conference, Microsoft Build. The event, which had gone virtual for the second consecutive year, flooded the event with all things AI. From MS Teams to Azure, the company announced more than 100 updates. Azure Machine Learning Managed Endpoints, once ready, will allow developers and data scientists to build and deploy ML models quickly and easily. Whereas, Azure Video Analyser looks for anomalies using AI. It uses existing services such as the Computer Vision from Azure Cognitive Services to build AI-powered video analytics from both stores and streaming footage. According to Microsoft, this brings Live Video Analytics and Video Indexer into a single service. The former would help build intelligent, video-based apps using the user’s choice of AI. At the same time, the Indexer automatically extracts data from video and audio content. Customers can use this for workplace safety, digital asset management, and monetising content. We are building the platform for platform creators. At #MSBuild, we are introducing more than 100 new updates and services across the tech stack to make creating easier for every developer. https:\/\/t.co\/QT72JMA9ls— Satya Nadella (@satyanadella) May 25, 2021 Microsoft has also added enterprise support for PyTorch AI on Azure. With this, Microsoft aims to give users of PyTorch a more reliable production experience. In his keynote speech, Microsoft CEO, Satya Nadella said that they would be using the world’s most powerful language model, GPT3, to Power Platform. Azure ML Managed Endpoints is already being used for the OpenAI GPT-3 model, which is among the world’s most significant natural language models—in Microsoft Power Apps. “If you can describe what you want to do in natural language, GPT3 will generate a list of the most relevant formulae for you to choose from. The code writes itself. The next theme is building new intelligent apps by harnessing the power of data and AI. This is so important. Every developer now has the opportunity and responsibility to make organizations sovereign in terms of their own data being used to benefit their organization. The opportunity is clear. The next generation of applications will not be reactive, but proactive, because of their ability to harness the power of data to drive that next level of intelligence,” added Nadella. Check our full coverage od MS Build here Top ML Announcements From Microsoft Build 2021 Microsoft Releases Power BI In Jupyter Notebooks Nvidia turns ON the world’s fastest AI supercomputer Image credits: NVIDIA Perlmutter, which has been credited with the title of world’s fastest AI supercomputer, is up and running. Perlmutter runs on 6,159 NVIDIA A100 Tensor Core GPUs, the largest A100-powered system in the world. “Perlmutter, officially dedicated today at the National Energy Research Scientific Computing Center (NERSC), is a supercomputer that will deliver nearly four exaflops of AI performance for more than 7,000 researchers.That makes Perlmutter the fastest system on the planet on the 16- and 32-bit mixed-precision math AI uses. And that performance doesn’t even include a second phase coming later this year to the system based at Lawrence Berkeley National Lab,” wrote Dion Harris of NVIDIA. The chipmaker claims that Perlmutter will help piece together a 3D map of the universe, probe subatomic interactions for green energy sources and much more. OpenAI launches $100 million fund Co-founder of OpenAI, Sam Altman announced that the OpenAI startup fund is investing $100 million to help AI companies have a profound, positive impact on the world. “We’re looking to partner with a small number of early-stage startups in fields where artificial intelligence can have a transformative effect—like health care, climate change, and education—and where AI tools can empower people by helping them be more productive,” said the company in a press release. The fund is managed by OpenAI, with investment from Microsoft and other OpenAI partners. In addition to capital, companies in the OpenAI Startup Fund will get early access to future OpenAI systems, support from our team, and credits on Azure. Read more here. SolarWinds hackers are back The notorious SolarWinds hackers who breached key government offices in the US late last year have made a comeback. On Thursday, Microsoft’s corporate VP, Tom Burt announced that they have observed cyberattacks by the threat actor Nobelium targeting government agencies, think tanks, consultants, and non-governmental organizations. This wave of attacks targeted approximately 3,000 email accounts at more than 150 different organizations. While organizations in the United States received the largest share of attacks, targeted victims span at least 24 countries. At least a quarter of the targeted organizations were involved in international development, humanitarian, and human rights work. “Nobelium, originating from Russia, is the same actor behind the attacks on SolarWinds customers in 2020. These attacks appear to be a continuation of multiple efforts by Nobelium to target government agencies involved in foreign policy as part of intelligence gathering efforts.  Nation-state cyber attacks aren’t slowing. We need clear rules governing nation-state conduct in cyberspace and clear expectations of the consequences for violation of those rules,” warned Microsoft. Google’s new OS Fuchsia goes live Google has finally confirmed the release of their Fuchsia operating system, which has been under development for more than 5 years now. The OS will be available through an update on their smart home devices—NestHub. Google’s smart display experience is built with Flutter, which is a functional-reactive user interface framework optimized for Fuchsia and is used by many system components. Flutter also runs on a variety of other platforms, including Android and iOS. Fuchsia itself does not require one to use any particular language or user interface framework. Google Inc has been making news about a new OS Fuchsia that would soon replace Android and Chrome OS. The project was first made public in 2016 when the project was released on GitHub. According to 9to5google, NestHub is just one of many avenues that Google has explored for Fuchsia, which is capable of powering both desktops and smartphones, even natively running Android apps. Bezos wraps up his final shareholder meeting On Wednesday, e-commerce mogul Jeff Bezos attended his last shareholder meeting as CEO of Amazon and announced that he’ll formally step down from his role as CEO on 5th of July.  He will hand over the reins to Andy Jassy, who currently heads the super successful Amazon Web Services(AWS). “He has the highest of high standards and I guarantee Andy will never let the universe make us typical,” Bezos said during the virtual meeting. Bezos will now pursue his passions Day 1 Fund, the Bezos Earth Fund, Blue Origin, The Washington Post going forward.","excerpt":"Microsoft kicked off its annual developer’s conference, Microsoft Build. The event, which had gone virtual for the second consecutive year, flooded the event with all things AI. From MS Teams to Azure, the company announced more than 100 updates. Azure Machine Learning Managed Endpoints, once ready, will allow developers and data scientists to build and […]","categories":["AI News"],"tags":["GPT-3","open source ios"],"author_name":"Ram Sagar","publish_date":"2021-05-30T10:00:00","publication_year":"2021","word_count":1088,"keywords":["GPT-3","machine learning","artificial intelligence","OpenAI","AI","PyTorch","ML","open source ios","computer vision","Aim","analytics","Azure ML"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","computer vision","analytics","OpenAI","Aim","Azure ML","PyTorch"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/microsoft-ms-build-openai-million-fund-top-news\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10163343,"title":"Accenture, Google Team Up to Boost AI and Cloud Adoption in Saudi Arabia","content":"Accenture and Google Cloud are collaborating to accelerate the adoption of cloud solutions and generative AI in Saudi Arabia. The goal is to help businesses improve operations, create new opportunities, and enhance customer experiences while ensuring data security and compliance. A recent report by the Saudi Data and Artificial Intelligence Authority (SDAIA) and Accenture suggests that generative AI could boost Saudi Arabia’s GDP by 4%. To support this growth, Accenture and Google Cloud are expanding their generative AI centre of excellence to Saudi Arabia. This will help businesses develop AI-powered solutions using Google Cloud’s technology and Accenture’s expertise. “Being ready for continuous reinvention hinges on a modern digital core to rapidly seize every opportunity,” said Majid Altuwaijri, Accenture’s Saudi Arabia chair and country managing director. The partnership also aims to help companies like the General Organisation for Social Insurance (GOSI) leverage generative AI. GOSI has already used cloud technology to create a secure, scalable AI environment where its developers and researchers can experiment with the latest AI tools. “Organisations need the combination of leading technology and services expertise to successfully deploy generative AI,” said Bader Almadi, general manager of Google Cloud in Saudi Arabia. “With Google Cloud’s advanced capabilities and Accenture’s industry expertise, customers will have access to the resources needed to plan, deploy and optimise generative AI projects.” Beyond technology, Accenture and Google Cloud are committed to developing local talent. To help professionals gain skills in cloud computing and AI, they plan to offer training programs, hackathons, and hands-on labs. Accenture’s LearnVantage platform will provide upskilling programs, specialised academies, and certifications to support Saudi Arabia’s digital workforce. The collaboration comes at a time when businesses in Saudi Arabia are eager to embrace digital transformation. By combining AI, cloud technology, and local expertise, Accenture and Google Cloud aim to drive innovation and ensure data security through Google’s Dammam cloud region. Meanwhile, Australian telecommunications company Telstra and Accenture also announced plans for a joint venture (JV). This is aimed at accelerating Telstra’s data and AI strategy to enhance network leadership and improve customer experiences. While the JV will develop specialised AI tools to optimise business processes and enhance workforce efficiency, Telstra’s data and AI teams in Australia and India will be offered roles in the JV, with training to advance AI fluency and critical skills.","excerpt":"Beyond technology, Accenture and Google Cloud are committed to developing local talent.","categories":["AI News"],"tags":["Accenture"],"author_name":"Shalini Mondal","publish_date":"2025-02-12T16:01:28","publication_year":"2025","word_count":384,"keywords":["Accenture","Go","artificial intelligence","AI","cloud computing","Scala","Git","RAG","Aim","generative AI","R"],"extracted_tech_keywords":["AI","artificial intelligence","generative AI","Aim","RAG","cloud computing","R","Go","Scala","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/accenture-google-team-up-to-boost-ai-and-cloud-adoption-in-saudi-arabia\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":42257,"title":"How India Can Reimagine Its Entire Healthcare System With Blockchain","content":"Many people have heard about the blockchain in the context of cryptocurrency. But there’s much more to blockchain than that. It is a public ledger that’s secure and immutable and that’s exactly what Indian healthcare needs. The Indian Healthcare Ecosystem There are six key stakeholders in the healthcare ecosystem: Patient Provider Payer Pharma Medical technology Government and healthcare regulators These stakeholders interact with each other through a complex network of interdependent and data-intensive workflows to generate meaningful health information. Over the years, a provider-centred care delivery model which uses the current framework of existing technologies has resulted in a lack of trust and transparency in the healthcare ecosystem. This, along with issues surrounding patient data privacy, security, and the fear of losing the competitive advantage by sharing data with another provider in the healthcare ecosystem, has aggravated the data-related challenges. However, as the healthcare industry now transitions to a more patient-driven care delivery model by introducing preventive healthcare and forming a close-knit ecosystem of patient-payer-provider, it is highly imperative to decentralise this model using trust-enabling technology. With its unique selling proposition of data security, blockchain is positioned as one of the key emerging technologies that can help in establishing a trust-based healthcare ecosystem by keeping the patient at the centre of all data generation and exchange. It has the potential to redesign some of the key processes of the healthcare industry by making them secure, transparent and efficient through accountable participation by all the key stakeholders. Permissioned Healthcare Blockchain Imagine a network on healthcare institutes where they don’t own a patient’s personal data. The data all belongs in a blockchain network. The patients are identified via their hash ID which will be their unique identifier. The hashing allows the ID to be unique and secures the privacy of the user. How permissioned healthcare blockchain helps The blockchain can aid in the creation of a patient information sharing marketplace. This way, it will be possible to actually incentivize information sharing between the different institutes to prevent any kind of information blocking. Since the blockchain is Immutable and traceable, patients can easily send records to anyone without the fear of data corruption or tampering. The patient can be incentivized for good behaviour via a reward mechanism. Like, they can get tokens for following a care plan or for staying healthy. They can be rewarded by tokens for giving their data for clinical trials and research. Other major problems in healthcare which blockchain can solve: Counterfeit drug prevention and detection: Blockchain can be used to introduce anti-tampering capabilities in the manufacturing phase to ensure pharmaceuticals are genuine. Clinical trial results: Blockchain can provide accountability and transparency to the clinical trial reporting process by curating all trials associated with a published study. Internet of Things: Patient-generated health and device data from IoT- enabled medical equipment has enormous promise, especially if interconnected with health records accessed by providers and patients. Stakeholder interaction using blockchain Let us take a scenario of hospitalization lifecycle of a patient where blockchain can transform some of the current processes to enhance the patient experience. Patient admission Current ecosystem Manual capture of admission record, duplication of episode ID No details of the previous admission at hand to assess if re-admission No visibility of clinical service and bed availability Blockchain-enabled ecosystem Real-time online updates of admission process details for timely preparedness and reference Easy availability of the previous admission at the same or other care provider supports clinical decision Clinical documentation Current ecosystem Verbal, manual capture of patient’s clinical record Duplicate capture of previous medical, family and other history Blockchain-enabled ecosystem Electronic and real-time capture of clinical details Quick access and retrieval of patient’s past medical history for clinical decision support Clinical diagnosis Current ecosystem Manual and time-consuming process to arrive at diagnosis due to lack of access to clinical symptomatology, investigations ordered and their results in one place Blockchain-enabled ecosystem Quick and accurate clinical diagnosis due to real-time and electronic availability of required patient clinical details and results Medication and treatment Current ecosystem Manual ordering and indenting of medicines No visibility into medicines prescribed by other treating clinician can impact patient safety Blockchain-enabled ecosystem Electronic, on-time medication order and dispensation Medication dosage tracking and monitoring Real-time visibility of other medicines being prescribed ensures patient safety Monitoring Current ecosystem Physical patient monitoring is captured on flowsheets and other manual clinical documents Lack of real-time validation by treating physician Blockchain-enabled ecosystem Allows real-time and remote monitoring of the patient by care provider due to the integration of medical devices to EHR Discharge Current ecosystem Lack of real-time online information on patient condition leads to a lengthy and time-consuming process as multiple approvals are needed from various departments in the hospital Blockchain-enabled ecosystem Real-time, remote monitoring of patient’s condition by all key stakeholder departments in the hospital leads to a faster discharge process Health insurance claim Current ecosystem Coordination with a third party for preauthorization and necessary approvals Manual and time-consuming claim verification by the third party Blockchain enabled ecosystem Automated transmission of patient billing details Workload at third party lessens due to ease of online, real-time claim verification Alert on the previous history of claims to detect fraud Payment Current ecosystem Slow and manual reimbursement process Physical copy payment receipt and its storage Blockchain-enabled ecosystem Timely claim verification and payment and compliance with regulatory body mandates Integrated insurance payouts via digital transactions Conclusion Potential of blockchain for Indian healthcare industry highly depends on the acceptance of the new technology within the healthcare ecosystem in order to create technical infrastructure. Though there are certain concerns and speculations regarding Blockchain’s integration with current healthcare systems and its cultural adoption. The technology is still popular in the healthcare sector. It has taken the healthcare industry by storm over the past year and many solutions are being developed to adopt it. With so many potential use cases and possibilities, blockchain is sure to disrupt the healthcare landscape for good. Also see:","excerpt":"Many people have heard about the blockchain in the context of cryptocurrency. But there’s much more to blockchain than that. It is a public ledger that’s secure and immutable and that’s exactly what Indian healthcare needs. The Indian Healthcare Ecosystem  There are six key stakeholders in the healthcare ecosystem: Patient Provider Payer Pharma Medical technology […]","categories":["AI Features"],"tags":["AI Healthcare","Blockchain","Healthcare Automation"],"author_name":"Nitin Srivastava","publish_date":"2019-07-11T17:50:45","publication_year":"2019","word_count":994,"keywords":["Go","Blockchain","programming_languages:R","AI","AI Healthcare","programming_languages:Go","Git","RAG","Healthcare Automation","Aim","Rust","R","programming_languages:Rust"],"extracted_tech_keywords":["AI","Aim","RAG","R","Go","Rust","Git","programming_languages:R","programming_languages:Go","programming_languages:Rust"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-india-can-reimagine-its-entire-healthcare-system-with-blockchain\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10083588,"title":"National Technology Awards 2023","content":"Technology Development Board (TDB), a statutory body of the Department of Science and Technology (DST), Government of India, is now inviting applications for the National Technology Awards, 2023, under the five categories of Main, MSME, Startup, Translational Research and Technology Business Incubator. These awards are conferred to various industries for successfully commercialising innovative indigenous technology. The awards will be given on May 11, 2023, celebrated as the National Technology Day in India, to honour the achievements of scientists, researchers, engineers and all others involved in the field of science and technology. It was on May 11, 1998, that India successfully conducted a Nuclear Missile Test at the Indian Army’s Pokhran range. Apply here. The last date to apply is January 15, 2023 by 5:00 PM. Main This award will be given to an industrial concern which has successfully developed and commercialised an indigenous technology on or after April 2017. If the technology developer\/provider and technology company are two different organisations, each one would be eligible for an award of ₹25 Lakh and a trophy. Cash Award of ₹25 Lakh Number of awards: One MSME This award will be given to MSMEs which have successfully commercialised a product based on indigenous technology on or after April 2017. Cash Awards of ₹15 Lakh Number of awards: Three (including one reserved for Women-led MSME) Startup This award will be given to a technology startup for promising new technology with potential for commercialisation. Cash Awards of ₹15 Lakh Number of awards: Five (including one reserved for Women-led Startup) Translational Research This award is given for the outstanding contribution of Scientists in commercialising innovative indigenous technologies. Cash Awards of ₹5 Lakh Number of awards: Two (one is to be reserved for Translational Research by Women Scientists) Technology Business Incubator To recognise and reward outstanding contributions in techno-entrepreneurship development by way of promoting innovative technology-driven, knowledge-intensive, startup enterprises in different technological areas. Cash Awards of ₹5 Lakh Number of awards: One","excerpt":"The awards will be given under the five categories, including main, MSME, startup, translational research and technology business incubator","categories":["AI News"],"tags":["government of india"],"author_name":"Shritama Saha","publish_date":"2022-12-28T15:50:05","publication_year":"2022","word_count":325,"keywords":["Go","government of india","programming_languages:R","AI","programming_languages:Go","ViT","GAN","R","startup"],"extracted_tech_keywords":["AI","R","Go","GAN","ViT","startup","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/national-technology-awards-2023\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10160698,"title":"LLMs that Failed Miserably in 2024","content":"Looks like the race to build large language models is winding down, with only a few clear winners. Among them, DeepSeek V3 has claimed the spotlight in 2024, leading the charge for Chinese open-source models. Competing head-to-head with closed-source giants like GPT-4 and Claude 3.5, DeepSeek V3 notched 45,499 downloads last month, standing tall alongside Meta’s Llama 3.1 (491,629 downloads) and Google’s Gemma 2 (377,651 downloads), according to Hugging Face.But not all LLMs launched this year could ride the wave of success—some fell flat, failing to capture interest despite grand promises. Here’s a look at the models that couldn’t make their mark in 2024. 1. Databricks DBRX Databricks launched DBRX, an open-source LLM with 132 billion parameters, in March 2024. It uses a fine-grained MoE architecture that activates four of 16 experts per input, with 36 billion active parameters. The company claimed that the model outperformed closed-source counterparts like GPT-3.5 and Gemini 1.5 Pro. However, since its launch, there has been little discussion about its adoption or whether enterprises find it suitable for building applications. The Mosaic team, acquired by Databricks in 2023 for $1.3 billion, led its development, and the company spent $10 million to build DBRX. But sadly, the model saw an abysmal 23 downloads on Hugging Face last month. 2. Falcon 2 In May, the Technology Innovation Institute (TII), Abu Dhabi, released its next series of Falcon language models in two variants: Falcon-2-11B and Falcon-2-11B-VLM. The Falcon 2 models showed impressive benchmark performance, with Falcon-2-11B outperforming Meta’s Llama 3 8B and matching Google’s Gemma 7B, as independently verified by the Hugging Face leaderboard. However, later in the year, Meta released Llama 3.2 and Llama 3.3, leaving Falcon 2 behind. According to Hugging Face, Falcon-2-11B-VLM recorded just around 1,000 downloads last month. 3. Snowflake Arctic In April, Snowflake launched Arctic LLM, a model with 480B parameters and a dense MoE hybrid Transformer architecture using 128 experts. The company proudly stated that it spent just $2 million to train the model, outperforming DBRX in tasks like SQL generation. The company’s attention on DBRX suggested an effort to challenge Databricks. Meanwhile, Snowflake acknowledged that models like Llama 3 outperformed it on some benchmarks. 4. Stable LM 2 Stability AI launched the Stable LM 2 series in January last year, featuring two variants: Stable LM 2 1.6B and Stable LM 2 12B. The 1.6B model, trained on 2 trillion tokens, supports seven languages, including Spanish, German, Italian, French, and Portuguese, and outperforms models like Microsoft’s Phi-1.5 and TinyLlama 1.1B in most tasks. Stable LM 2 12B, launched in May, offers 12 billion parameters and is trained on 2 trillion tokens in seven languages. The company claimed that the model competes with larger ones like Mixtral, Llama 2, and Qwen 1.5, excelling in tool usage for RAG systems. However, the latest user statistics tell a different story, with just 444 downloads last month. 5. Nemotron-4 340B Nemotron-4-340B-Instruct is an LLM developed by NVIDIA for synthetic data generation and chat applications. Released in June 2024, it is part of the Nemotron-4 340B series, which also includes the Base and Reward variants. Despite its features, the model has seen minimal uptake, recording just around 101 downloads on Hugging Face in December, 2024. 6. Jamba AI21 Labs introduced Jamba in March 2024, an LLM that combines Mamba-based structured state space models (SSM) with traditional Transformer layers. The Jamba family includes multiple versions, such as Jamba-v0.1, Jamba 1.5 Mini, and Jamba 1.5 Large. With its 256K token context window, Jamba can process much larger chunks of text than many competing models, sparking initial excitement. However, the model failed to capture much attention, garnering only around 7K downloads on Hugging Face last month. 7. AMD OLMo AMD entered the open-source AI arena in late 2024 with its OLMo series of Transformer-based, decoder-only language models. The OLMo series includes the base OLMo 1B, OLMo 1B SFT (Supervised Fine-Tuned), and OLMo 1B SFT DPO (aligned with human preferences via Direct Preference Optimisation). Trained on 16 AMD Instinct MI250 GPU-powered nodes, the models achieved a throughput of 12,200 tokens\/sec\/gpu. The flagship OLMo 1B model features 1.2 billion parameters, 16 layers, 16 heads, a hidden size of 2048, a context length of 2048 tokens, and a vocabulary size of 50,280, targeting developers, data scientists, and businesses. Despite this, the model failed to gain any traction in the community.","excerpt":"Databricks spent $10 million developing DBRX, yet only recorded 23 downloads on Hugging Face last month.","categories":["AI Trends"],"tags":["LLMs"],"author_name":"Siddharth Jindal","publish_date":"2025-01-03T12:03:48","publication_year":"2025","word_count":727,"keywords":["Hugging Face","AI","LLMs","DeepSeek V3","RAG","Aim","Databricks","SQL","Claude 3.5","R","Snowflake"],"extracted_tech_keywords":["AI","Claude 3.5","DeepSeek V3","Aim","Hugging Face","RAG","Snowflake","Databricks","R","SQL"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/llms-that-failed-miserably-in-2024\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10073665,"title":"How Koo Thinks About Its Technology Stack","content":"Launched in 2020, homegrown micro-blogging platform Koo has grown at a surprisingly rapid pace, considering its competitor is the ubiquitous Twitter. However, to the tech team behind the app, homegrown in no way indicates archaic. In a statement released to PTI, co-founder and CEO Aprameya Radhakrishna stated that the app has seen its user base grow 10-fold in the past year and expects the figure to cross 100 million by next year. Analytics India Magazine caught up with the app’s chief technology officer Phaneesh Gururaj. Gururaj joined Koo a little over a year ago after a seven-year stint as the head of products in data science with redBus. Here are the excerpts: Phaneesh Gururaj, CTO at Koo Tech Stack Under Gururaj, the platform’s tech has evolved rapidly. “Koo is backed by a robust technology stack which uses Kotlin for our BE systems and our Android. For our iOS, we use Swift,” he said. The website is built on top of node.js while its ML systems use Python. Video engineering is critical for Koo. According to Gururaj, on a single day more than 20TB of videos are streamed from the app. “CDN plays a very important role and we leverage cloud here. We also use transcoders to convert videos across other formats. Our video engineering is on top of Elixir.” Gururaj says that cloud has helped the platform scale rapidly in terms of both user growth and features. “We are hosted on AWS and also use GCP for some use cases. Our infrastructure runs on Kubernetes and consists of Datastores – PostGreSQL, Open Search, AEROSPIKE, ArangoDB, ML GPU systems and data pipelines. Data Lake is where we hydrate all our data points and signals; from here, we extract intelligence and run other analytical pipelines. This is built on open source Apache frameworks like Kafka, Hudi, Spark, NiFi and Flink. We use various types of databases ranging from OLTP, OLAP, Graph and NoSQL — Aerospike, Arango, Elastic Search, PostgreSql that power our features,” he added. Estimated Monthly Active Users of Koo in Feb and April 2021, Source: Statista Localised features and advantages Acknowledging the natural comparison with its rivals, Gururaj is convinced that the platform’s localised features are like no other. Earlier this week the app launched a new in-app feature called ‘Topics’ available in 10 languages including Hindi, Bengali, Marathi, Gujarati, Kannada, Tamil, Telugu, Assamese, Punjabi and English. The ‘Topics’ feature will allow users to select their own areas of interest in the language they are most comfortable with. “Since 90% of India speaks a native language, we have attracted creators from across ‘Bharat’ — including millions of first-time users who were earlier hesitant to be on English-first platforms. We believe that Koo is on a path to building a digital experience that is a category in, and of itself,” he stated. To Gururaj, the fact that the app allows users to express themselves in their native language is its biggest bounty. “Multi-Lingual Kooing (MLK) is a pioneering feature, which enables real-time translation of a message across multiple languages that enhances reach and bridges the linguistic gap between speakers of different languages,” he explains. Koo has further empowered users by enabling them to self-verify their profiles and get recognised as genuine voices. “We are the first social media platform to have introduced Voluntary Self-Verification for all users, and are also among the first to publish the workings behind our algorithms — a move which reiterates our commitment to platform transparency and a user-first approach. Koo’s interface, Source: Bootcamp.uxdesign.cc Moderation Policy When asked about the contentious issue of moderation which has plagued social-media platforms like Facebook and Twitter, Gururaj explains the need for a balance between regulation and freedom. “We are a social media intermediary in the truest sense of the term, and don’t editorialise, label or characterise content (except as required by law). Our nuanced understanding of native languages, compliance to the law of the land, and strong community guidelines form the bedrock of our content moderation practice. We are currently building dictionaries of words and phrases across languages in partnership with the Central Institute of Indian Languages (CIIL) to promote fair use of language and responsible online behaviour,” Gururaj stated. Hiring Gururaj’s experience while building Koo’s AI\/ML team from ground-up entailed tackling hiring challenges in the Indian subcontinent. “Complex projects, like the ones that are executed at Koo, require a mix of professionals from different specialisations. Even with the talent density in India we haven’t yet been able to address the demand for specialised data skills,” he stated. Koo has still managed to leap-frog steep learning curves by hiring early-career professionals, and providing them with strong mentorship support from senior technical leaders. Future vision for the AI\/ML team Gururaj notes that while the team is currently working on its recommendation engines, content classifiers and content moderation systems, Koo will continue to invest in R&D for its core areas like native-language technologies. “We want to develop Large Language Models and train custom neural networks for native-language understanding and translations. These investments are necessary to ensure we address the growing complexity of content created on our platform from millions of first-time creators using native languages,” Gururaj stated. Recommendation System How does Koo’s recommendation system hold up when compared to the razor-sharp recommendation engines of TikTok? “For our recommendations, we leverage the Koo social graph. The graph theory claims that an addition of a node to a graph increases the value of the entire graph, not linearly, but exponentially. This exponential increase in the value requires efforts to ensure that we continue to extract value for our users.” The AI\/ML team has invested in developing graph algorithms like large-scale graph embeddings, which is an area of active research for the app. The core teams are also working hard to build our own graph technologies which will allow them to train state-of-the-art graph algorithms for link prediction and small-community detection. These applications will translate directly into delivering superior recommendations on Koo. Cybersecurity Gururaj explains how constant improvements are made to enhance Koo’s cybersecurity. “We have a bug bounty program and work with known ethical hackers across the globe to constantly improve the security posture of Koo. We have also invested and implemented multiple security tools to prevent DDOS and ransomware attacks. Moreover, we have set-up an internal security team and built automation to constantly monitor for anomalous traffic,” he said. Gururaj’s firm belief in Koo’s mass appeal makes it a phenomenon unlike any of its counterparts. He sees Koo as a standout innovator in an English-first social media landscape building technology that can be consumed by the larger world – one that speaks a native language. “We allow users to cut across language barriers when creating or consuming content. We bring eminent personalities from all walks of life to Koo to speak and be heard. A large number of government and social organisations use Koo to stay connected and inform the masses about various initiatives. We are hyperlocal, and global at the same time,” he signed off.","excerpt":"“Since 90% of India speaks a native language, we have attracted creators from across ‘Bharat’ – including millions of first-time users, earlier hesitant to be on English-first platforms,” says Gururaj.","categories":["AI Features"],"tags":[],"author_name":"Poulomi Chatterjee","publish_date":"2022-08-26T14:29:59","publication_year":"2022","word_count":1173,"keywords":["data science","GCP","AWS","AI","neural network","ML","RAG","Aim","analytics","kubernetes"],"extracted_tech_keywords":["AI","ML","neural network","data science","analytics","Aim","RAG","AWS","GCP","kubernetes"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-koo-thinks-about-its-technology-stack\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10041971,"title":"New Antitrust Bills Turn The Heat On Big Tech","content":"In an exciting turn of events, the Democrats on the Antitrust Subcommittee, with support from the Republican side, have proposed several antitrust laws to curb the power of tech giants like Amazon, Apple, Facebook, and Google. Each of the five bills has multiple co-sponsors, including at least one from either party. While it’s not uncommon to see a bipartisan congressional agreement, considering that Republicans and Democrats have previously found common ground on gun policy proposals, it sure isn’t a regular fixture. The five-package bill is a power move by the House of Democrats that would prohibit tech giants from discriminating against or acquiring potential competitors. The draft bills come after a year-long investigation into competition in digital markets, led by the Subcommittee on Antitrust, Commercial and Administrative Law. The committee’s Democratic staff submitted a 450-page report at the end of the probe. Five bills Here’s the lowdown on the five bills: Under the “Ending Platform Monopolies Act,” corporations would be prohibited from smothering smaller rivals and thwarting free and fair online competition. Major tech platforms like Facebook, Google, and Amazon would be restricted from owning or operating a business that presents a conflict of interest. The US Federal Trade Commission (FTC) had previously raised concerns about Amazon and Apple. The duo run their own platforms for sellers and developers, effectively undercutting competition by favouring their products or apps. The “American Choice and Innovation Online Act” would prohibit firms from giving preferential treatment to their products and services on their platform over competitors. Additionally, it would prohibit dominant platforms from engaging in discriminatory behaviour, such as isolating a competitor who uses the platform and from using data collected on their services that are not publicly available. Regulators could veto dominant companies from acquiring would-be competitors under the “Platform Competition and Opportunity Act.” “Augmenting Compatibility and Competition by Enabling Service Switching (ACCESS) Act” would lower the barrier of entry. “Merger Filing Fee Modernization Act” would increase the fees companies pay to notify the Federal Trade Commission and Department of Justice Antitrust Division of large mergers with an aim to raise money for those agencies. What it means for the tech giants The five-bill package has provisions that would make it more difficult for dominant platforms to close merger deals, banning them from owning businesses that pose conflicts of interest. The proposed legislation is the most comprehensive attempt in decades to reform centuries-old antitrust laws. Tech monopolies like Facebook would now have a hard time expanding their footprint through acquiring the competition. “Augmenting Compatibility and Competition by Enabling Service Switching (ACCESS) Act” will counteract the social network market’s high entry barrier. In the last 20 years, Google has purchased over 260 companies. The search giant has been the subject of antitrust investigations and enforcement actions in the past. Google has leveraged its search dominance to popularise the Chrome browser. Google emerged as the default setting across the mobile and desktop ecosystems through a combination of integration and contractual agreements, including owning Android, the world’s most popular mobile operating system. Through this, Google maintained its Search’s dominance even as mobile overtook desktop as the primary gateway to the Internet. Now, with the introduction of the “American Choice and Innovation Online Act,” dominant platforms would be prohibited from providing preferential treatment to their products and services on the platform. Thanks to “Platform Competition and Opportunity Act”, both Facebook and Google would find it hard to acquire would-be competitors. Apple owns and operates all mobile devices and does not licence iOS to third-party manufacturers. It also has high switching costs, ecosystem lock-in, and brand loyalty. As a result of Apple’s control over iOS, the company commands a lot of power over software distribution on iOS devices. The Subcommittee investigated Amazon in June 2019 for its role as a gatekeeper for digital markets. The everything store wields sizable power over most third-party sellers and suppliers. With the “Ending Platform Monopolies Act,” corporations would be prohibited from dominating smaller competitors and preventing them from having a fair online marketplace competition.","excerpt":"In an exciting turn of events, the Democrats on the Antitrust Subcommittee, with support from the Republican side, have proposed several antitrust laws to curb the power of tech giants like Amazon, Apple, Facebook, and Google. Each of the five bills has multiple co-sponsors, including at least one from either party. While it’s not uncommon […]","categories":["IT Services"],"tags":[],"author_name":"Ritika Sagar","publish_date":"2021-06-19T16:00:00","publication_year":"2021","word_count":672,"keywords":["Go","AWS","AI","cloud_platforms:AWS","innovation","Git","RAG","Aim","Rust","R"],"extracted_tech_keywords":["AI","Aim","RAG","AWS","R","Go","Rust","Git","innovation","cloud_platforms:AWS"],"url":"https:\/\/analyticsindiamag.com\/it-services\/new-antitrust-bills-turn-the-heat-on-big-tech\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10044056,"title":"Data Sciences and Analytics Programme At eClerx: Pioneering In Solving Emerging &#038; Mission-Critical Problems For Fortune® 500 Enterprises","content":"Data plays a critical role in the success of big-ticket companies, especially post-pandemic. Businesses rely on data-driven strategies to solve ambiguities, take on business challenges, streamline processes and drive operational efficiencies. By 2025, 80% of organisations looking to scale their digital functions are doomed to fail, owing to a medieval approach to data and analytics, a Gartner report observed. Right on cue is Mumbai-based Data Science and Analytics service provider — eClerx that has offices in several major cities across the globe. eClerx, recognised as one of the Top 50 Firms in India for Data Scientists To Work For in 2021, has been the partner of choice to several Fortune® 500 enterprises. Founded by Anjan Malik and PD Mundhra in 2000, eClerx was also the first KPO to go public and get listed on the Bombay Stock Exchange in 2007. In this five-part series, we will be turning the spotlight on eClerx’s Data Sciences and Analytics (DSA) Programme, which is at the forefront of helping businesses build and scale well-articulated data, analytics and AI strategy to deliver transformational business outcomes. One-Stop-Shop For Data Science & Analytics While consulting firms fall short on the data and execution tactics, boutique data-science firms overlook implementation and ROI enablement. On the other hand, agencies lack advanced intelligence, whereas tech companies entirely neglect the problem-solving aspect. eClerx, with its highly customer-centric approach and well-rounded capabilities, brings the best of all worlds to help businesses smartly navigate an ever-evolving and complex space to deliver pertinent solutions ahead of time. Over the years, eClerx’s data management, reporting and web analytics capabilities evolved to business intelligence, cloud data engineering, digital analytics, machine learning and artificial intelligence. eClerx has carved out a centralised Data Sciences and Analytics (DSA) Programme, a strong team of 800+ specialists with a global presence and a hybrid delivery model, catering to customers across all industries and geographies to further strengthen its position and provide data-driven solutions. Global presence of eClerx’s in-market consulting teams, offshore delivery & regional offices. Bringing Innovation To The Forefront DSA works closely with niche and mainstream clients in their data science transformation journeys in order to improve their decision-making capabilities. Whether developing a data strategy, managing messy data, implementing predictive analytics models, prototyping first-of-its-kind innovative AI\/ML solutions or setting up a seamless self-serve BI suite, eClerx has the entire gamut of solutions. “As the problems evolved, the way we solve them has also evolved, and eClerx is at the forefront of this change, pioneering all things data”, the company stated. It all began in 2005-06, when eClerx developed expertise in Digital Analytics, Tag Management and A\/B Testing \/Multivariate Testing, and Experiments Design, currently known as martech enablement. Around the same time, it also acquired a firm to strengthen its foothold into the Business Intelligence space specialising in delivering descriptive and prescriptive analytics. The firm also pioneered solutions that enabled enterprises to track online conversations about their products — becoming a forerunner in social media and Sentiment Analytics when social media itself was new and trying to gain adoption. eClerx also set up scalable BI and Analytics solutions catering to business leaders across marketing, sales, and customer-focused functions. Over the last 15 years, a culture of innovation, early adoption and entrepreneurship have allowed the company to continue pushing the envelope and transforming itself by developing customised and first-of-its-kind solutions in Digital Analytics, Big Data Engineering, Data Science Machine Learning, NLP, AI space. In a true modern-problems-require-modern-solutions spirit, eClerx leverages the latest technologies to develop solutions to tackle seemingly intractable business problems. Today, eClerx brings best-in-class Data Science and Analytics solutions to the fore, without shifting the focus away from problem-solving to deliver value across five key areas — Overview of eClerx’s range of Data Science & Analytics Service Areas Pushing The Envelope Considering a dynamic data analytics strategy has become imperative for businesses to sustain in this highly volatile market, eClerx employs its indigenous problem-solving framework, best-of-breed practices and proven operations controls & measures to deliver the best-in-class experience to its clients. What also differentiates eClerx is its extensive portfolio of services around Data Operations, Campaign Management, SEO, SEM, Advertising Operations and Digital Content Creation and Management, among other areas, which makes it an end-to-end partner for executing and implementing business recommendations derived through analytics, thus unlocking ROI for its customers. A Game Changer In The Age Of Analytics Weighing in DSA’s position and calibre, the journey ahead looks far more exciting and promising, where eClerx is looking to elevate and transform the data science and analytics practice to the next level. The organisation’s strong belief is that Analytics should be delivered at speed and scale. So it has combined the best of human and machine intelligence to build customisable shell platforms that transcend the rigidity of traditional product mindsets and bring productization to data science. However, that’s one aspect of the bigger picture. Its army of consultants works with CxOs of Fortune® 500 \/ 1000 firms to create Analytics Transformation Roadmaps for these organisations by conducting stakeholder interviews, maturity assessments and industry benchmarking with research & innovation powered by its Analytics Centre of Excellence. Driven by its hybrid global delivery model, eClerx is poised to enable cost-effective, scalable and transformation journeys for large global enterprises to drive the next wave of growth through data science and analytics. Up Next In the next article, we shall explore how eClerx, which was placed third in our “Learning & Support” category, fosters a culture of learning and development for its highly diverse talent groups by building an extensive knowledge infrastructure. Watch out for details into how DSA propagates knowledge sharing through an active community of practice – enabling a cross-section of teams to converge and share discoveries and learnings from real-world business problems in addition to discussing the latest market and domain trends and developments.","excerpt":"As the nature of business problems evolved, the way we solve them has also evolved, and eClerx is at the forefront of this change, pioneering in all things data for the past two decades.","categories":["IT Services"],"tags":["eClerx"],"author_name":"Sejuti Das","publish_date":"2021-07-21T18:00:00","publication_year":"2021","word_count":972,"keywords":["data science","machine learning","artificial intelligence","AI","R","ML","RAG","NLP","analytics","predictive analytics","eClerx"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","NLP","data science","analytics","RAG","predictive analytics","R"],"url":"https:\/\/analyticsindiamag.com\/it-services\/data-sciences-and-analytics-programme-at-eclerx-pioneering-in-solving-emerging-mission-critical-problems-for-fortune-500-enterprises\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":34211,"title":"MachineHack Winners: How Great Visualisations Helped These Data Scientists Win A Unique Hackathon","content":"MachineHack recently concluded the “Visualisation is Beautiful: Data Science Salary Visualisation Challenge”. Analytics India Magazine talked to the winners of the hackathon and found out their experience of participating and winning the hackathon. The decision for the best visualisation was arrived upon by counting votes each participant had gotten. Mani Rathina Velu was Ranked #1 on the leaderboard. Mani started his career in the Business Intelligence world as a Report\/Dashboard Developer using SAP Dashboards, Crystal Report and Webi at Bizviz Technologies Pvt Limited. He learnt other Business Intelligence tools like Jaspersoft, Tableau, Power BI and BDB Platform to enhance the knowledge and business needs. He started learning about data analytics when he worked on the project at this company. The data science learning started when we predicted the match winners for Cricket World Cup 2015 using SQL calculation. Under the guidance of his mentor Avin Jain (CEO Of BizViz) the process had completed with high accuracy. He went onto implementing a complete End to End Data Analytics platform (Data Collection-Ingestion-Transformation-Reporting\/Dashboarding) using BDB for a Retail industry. This project really inspired him to get into Data Science. From there the Data Science journey started for Mani and he picked up R language and then took out time to learn Python, Spark ML and Deep Learning (TensorFlow and Keras). He took an active interest in developing ML algorithms to predict the match winners for IPL T20 2018 and it also predicts the score for each over using BDB Data Pipeline in a live match. Approach to solving the visualisation challenge: Using BDB Data Preparation tool, cleaned redundant data in the file. Collaborated with our analytics team in BDB, used text analysis to delimit and analyse the data. Used BDB Predictive Workbench for clustering algorithm to find the insights. The aim was to develop the dashboard with Descriptive and Predictive Analytics for this Hackathon using BDB Dashboard Designer. If time would have permitted I would have added Prescriptive part also in the same dashboard. Talking about his experience on MachineHack, Mani said, “It helps the budding talents to get more confidence in the field chosen. I would expect MachineHack to continue providing the same kind of support and knowledge to us. Overall had a nice experience interacting with the team and willing to participate more in future.” Anubhav Gupta came second on the leaderboard. Talking about how the data science journey started he said, “The journey towards exploring data science has just begun and I think it would be a bumpy ride ahead”. His journey has started in 2017 right after completion of my Bachelors’ degree. I had a few months of time before starting my professional career. These were the days where the buzz was about data scientists are the hot cakes of the IT industry. It was during that period his little research caught attention to this phrase “Data Scientist: The Sexiest Job of the 21st Century”. His love towards numbers and solving puzzles has provoked me to explore career opportunities in Data Science, also Statistics was one of my favourite subjects in high school. Anubhav quickly realised Data Science is all about scientific exploration of data to take out some insights in a business context. He did not want to stop his quest there, but back into my mind, I knew that my job would entail software development\/testing. He started signing up courses on data analytics\/data science from DataCamp, Edx, Coursera etc. and learned foundational basics of Python and SQL. Taking up courses helped him understand better. “Learn, Practice and Improve” is the mantra to excel in any field one desires to be in so I need to practice and improve to go further. Once his professional career had started to my surprise I was tagged to a project on Data Science. I got my first project, there I learned data analysis and data visualisation using tools like QlikSense and PowerBI. Understanding the data and converting them into insightful visualisations is the first step towards data science and he is on to it. Approach to solving the visualisation challenge: Primarily, Anubhav understood the data and did some data cleansing. After that, he selected the fields (Selecting dimensions and measures) which can give more insightful information about the data science salary as per the data provided. A secondary task was to select the tool for presenting the data in form of beautiful visualisations. He chose QlikSense as I have hands-on experience on QlikSense. The last and significant task was to select the type of visualisations and their placement in the dashboard. After sketching a rough idea in my head using the above analysis and selection, I replicated the same to solve the problem of the hackathon. Talking about his experience on MachineHack, Anubhav said, “Altogether it was a great learning experience and competing with top contenders can make us grow and compare where we stand in terms of industry standards. Machine Hack is organising such great hackathons that provide an amazing competitive environment to the contenders having expertise in their respective fields of Data Science. Looking forward to more hackathons like this.” Karthikeyan. PS came in third on the leaderboard. Karthikeyan started his journey with Avin Jain. He said, “All this started from the one man, my mentor Avin Jain who believed me and gave me the first chance to start my career at BDB. He started as a Quality Analyst of SAP BOBJ products at SAP LABS, Bangalore and gradually started handling Customer issues directly and that gave him more exposure to the actual business scenarios. He soon learnt other leading BI tools in the market like BDB, Jasper reporting, Tableau, PowerBI to achieve various business desired Visualisation. Over the years, he worked with multiple customers from different verticals and done different types of end-to-end implementation according to their business needs. During this journey, he has gone through various implementation blockers and hunted for the best solution to get them Go-live. He used BDB Platform for visualisations and till now he has worked on projects like Customer Segmentation, RFM Model, Demand Forecast, Market Basket Analysis, Feedback Analysis to Retail and Marketing verticals. Approach to solving the visualisation challenge: When he started with data, we observed many redundancies, unstructured data on few columns like job description and skills. He believes data clean-up is the most important part of any analytics, he decided first to clean up and make the data more proper and structured that can be used for Analysis smoothly. Using BDB Predictive workbench, he ran some text analytics and organised the data in a structured way. Followed by that, he ran clustering and LR based algorithms and consumed the result directly in BDB Dashboard Designer module to visualise. Summarised the dashboard with a descriptive and predictive part in the limited time. Talking about his experience on MachineHack, he said, “Overall it was a very smooth experience from the time we started this course. The Prompt support from the Machine hack team is appreciable. Willing to participate in this type of hackathon’s more in future.”","excerpt":"MachineHack recently concluded the “Visualisation is Beautiful: Data Science Salary Visualisation Challenge”. Analytics India Magazine talked to the winners of the hackathon and found out their experience of participating and winning the hackathon. The decision for the best visualisation was arrived upon by counting votes each participant had gotten. Mani Rathina Velu was Ranked #1 […]","categories":["Deep Tech"],"tags":["ai hackathon india","Hackathon","retail bi prescriptive","Tableau"],"author_name":"Abhijeet Katte","publish_date":"2019-01-28T07:54:30","publication_year":"2019","word_count":1175,"keywords":["ai hackathon india","data science","Tableau","Keras","AI","retail bi prescriptive","ML","Hackathon","Python","Aim","deep learning","analytics","TensorFlow","predictive analytics"],"extracted_tech_keywords":["AI","ML","deep learning","data science","analytics","Aim","TensorFlow","Keras","predictive analytics","Python"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/machinehack-winners-how-great-visualisations-helped-these-data-scientists-win-a-unique-hackathon\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10119319,"title":"Cohere Rolls Out Multi-Model Framework PoLL for Comprehensive LLM Evaluation","content":"In response to the complexities and challenges associated with evaluating the ever-evolving LLMs, influential AI startup Cohere has introduced a new evaluation framework called the Panel of LLM Evaluators (PoLL) which leverages a diverse panel of smaller, distinct model families to assess LLM outputs, promising a more accurate, less biased, and cost-effective method compared to traditional single-model evaluations. Traditional evaluations often use a single large model like GPT-4 to judge the quality of other models’ outputs. However, this method is not only costly but also prone to intra-model bias, where the evaluator model might favour outputs similar to its training data. PoLL addresses these challenges by assembling a panel of smaller models from different model families to evaluate LLM outputs. This setup reduces evaluation costs by over seven times compared to using a single large model and minimises bias through its varied model composition. The framework’s effectiveness has been validated across multiple settings, including single-hop QA, multi-hop QA, and competitive benchmarks like the Chatbot Arena. Studies utilising PoLL have demonstrated a stronger correlation with human judgments compared to single-model evaluations. This suggests that a diverse panel can better capture the nuances of language that a single, large model might miss due to its broader, generalised training. Methodology Behind PoLL The PoLL consisted of models from three distinct families—GPT-3.5, CMD-R, and Haiku—each contributing diverse perspectives to the evaluation process. This diversity allows PoLL to offer a well-rounded assessment of LLM outputs, addressing different language understanding and generation aspects. The success of PoLL paves the way for more decentralised and diversified approaches to LLM evaluation. Future research could explore different combinations of models in the panel to further optimise accuracy and cost. Moreover, applying PoLL to other language processing tasks, such as summarisation or translation, could help establish its effectiveness across the field.","excerpt":"By leveraging a diverse panel of models, PoLL aligns more closely with human judgment and provides a scalable, cost-effective solution to the growing need for accurate LLM assessments.","categories":["AI News"],"tags":["small language models"],"author_name":"Shritama Saha","publish_date":"2024-04-30T17:26:27","publication_year":"2024","word_count":301,"keywords":["TPU","programming_languages:R","AI","RAG","GPT","llm_models:GPT","small language models","R","startup"],"extracted_tech_keywords":["AI","RAG","TPU","R","GPT","startup","llm_models:GPT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/cohere-rolls-out-multi-model-framework-poll-for-comprehensive-llm-evaluation\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10122744,"title":"C5i Leverages GenAI-Based Solutions in the Life Sciences Industry","content":"Several decision processes stand to be impacted by the power of generative AI in the life sciences field. GenAI-based solutions can aid in numerous tasks, including analysing vast amounts of structured data to deliver meaningful insights and enhancing the ability to digest and search diverse unstructured data to improve several decision-making workflows. Here are the top areas where C5i is driving the empowerment of our customers through GenAI-based solutions. AI-Powered MLR Review for Content Creation Promotional materials created for healthcare professionals (HCP) and patients are complex and require a significant amount of vetting and review. GenAI can help automate the review of essential documents, summarising key issues. This frees up professionals working in the medical, legal and regulatory fields to focus on higher-level tasks. Optimised Sales Force Enablement Actionable insights for sales reps can be enhanced using GenAI to craft impactful presentations. This is done through analysing diverse data sources with customised insights, visualisations and analyses to enable optimal conversation paths based on HCP profiles and past interactions, maximising efficiency and message relevance. Holistic Brand Performance Analysis A pharmaceutical brand’s performance depends on a multi-dimensional view cutting across customer engagement, market access enablement, patient support, competition analysis and more. These insights are available in structured data, both internal and syndicated, and unstructured data, including internal and market research data. Brands can be enabled with a comprehensive view of performance with prioritised and consumer insights being pushed to the brand managers. Patient Engagement GenAI empowers patient engagement by personalising interactions and interventions. Advanced algorithms analyse patient data, generating tailored recommendations and support strategies. This approach enhances communication between patients and healthcare providers, promoting adherence to treatment plans and preventive care. Ethical considerations are crucial when adopting GenAI, where ensuring integrity, trust, and fairness in data-driven decision-making is a priority. By adhering to ethical guidelines, the industry can foster responsible innovation, ensuring that generative AI solutions contribute positively to healthcare delivery and patient outcomes. About C5i (Course5 Intelligence Limited) C5i is a pure-play AI & Analytics provider that combines the power of human perspective with AI technology to deliver trustworthy intelligence. The company drives value through a comprehensive solution set, integrating multifunctional teams that have technical and business domain expertise with a robust suite of products, solutions, and accelerators tailored for various horizontal and industry-specific use cases. At the core, C5i’s focus is to deliver business impact at speed and scale by driving adoption of AI-assisted decision-making. C5i caters to some of the world’s largest enterprises, including many Fortune 500 companies. The company’s clients span Technology, Media, and Telecom (TMT), Pharma & Lifesciences, CPG, Retail, Banking, and other sectors. C5i has been recognized by leading industry analysts like Gartner and Forrester for its Analytics and AI capabilities and proprietary AI-based platforms.","excerpt":"GenAI serves as an excellent tool in effectively aiding companies, but to do so in the field of life sciences requires key ethical considerations.","categories":["AI Highlights"],"tags":["GenAI","Generative AI"],"author_name":"Kamal Kasi","publish_date":"2024-06-06T18:07:15","publication_year":"2024","word_count":459,"keywords":["Go","GenAI","AI","data-driven","ML","innovation","analytics","generative AI","Rust","Generative AI","R"],"extracted_tech_keywords":["AI","ML","analytics","generative AI","GenAI","R","Go","Rust","data-driven","innovation"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/c5i-leverages-genai-based-solutions-in-the-life-sciences-industry\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":755,"title":"Business analytics lab in IIM Ranchi soon","content":"A business analytics lab would be soon set up at IIM Ranchi following a MoU between the institute and IBM Corporation. The company would provide IIM with COGNOS and data mining software to analyze and evaluate vital information. The development took place at a daylong seminar on business analytics themed “Redefining learning paradigm with business analytics”. IIM Ranchi director M J Xavier said relevant information is scarce even with abundant data. “We need enhanced and specialized analytical tools to process and transform data into useful knowledge, rules and pattern,” he added. Dwelling on the importance of information and statistics, speakers at the seminar said data could be used to make strategic management decisions. “Basically business analytics are mostly used in the domains of finance. It should be used to analyze tools for developmental purposes as well,” said a professor at the seminar. The conclave also witnessed industry experts from IBM and other organizations sharing their insights about the recent developments and best practices in the field of analytics. Kunal Dureja of IBM also dwelt upon the importance of data analysis. Pradip Kumar Bala from IIM Ranchi said, “We need to focus on holistic development and aim at using business analytics in the social sectors like education and also healthcare,” Bala added. News sourced from Times of India","excerpt":"A business analytics lab would be soon set up at IIM Ranchi following a MoU between the institute and IBM Corporation. The company would provide IIM with COGNOS and data mining software to analyze and evaluate vital information. The development took place at a daylong seminar on business analytics themed “Redefining learning paradigm with business analytics”. IIM Ranchi director […]","categories":["AI Trends"],"tags":[],"author_name":"Дарья","publish_date":"2012-08-18T21:40:15","publication_year":"2012","word_count":217,"keywords":["ELT","programming_languages:R","AI","Aim","ViT","analytics","GAN","R"],"extracted_tech_keywords":["AI","analytics","Aim","R","ELT","GAN","ViT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/business-analytics-lab-in-iim-ranchi-soon\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10077646,"title":"Zoom Launches ‘Zoom Events’ in India to Elevate Virtual Experiences","content":"Video communication tech firm Zoom on Wednesday announced the availability of virtual platform Zoom Events in India. The company, in a press release, stated that the all-in-one platform aims to produce immersive and engaging virtual experiences. Zoom Events would combine the reliability of Zoom Webinars, Zoom Meetings and Team Chat in one comprehensive solution for event organizers, with the ability to produce live events for internal or external audiences of almost any size. Over 7,000 customers have used Zoom Events, averaging more than 150 events daily since its launch in 2021. Moreover, Zoom also hosted its conference  Zoomtopia 2021 on Zoom Events, delivering to more than 30,000 virtual attendees globally. Zoom Events would enable businesses across industries to seamlessly host virtual and hybrid experiences including company events like sales summits and external events like user conferences. Sameer Raje, general manager & head of India and SAARC Region at Zoom, said, “The preference for virtual and hybrid events is growing as people continue to look for flexibility in how they connect, learn, work and attend events. Event organizers are on the lookout for a comprehensive, intuitive and easily scalable platform for hosting interactive and engaging events and Zoom delivers the perfect solution.” Recently, Frost & Sullivan recognized Zoom at the 2021 Global Webinars and Virtual Events Market Leadership Award for its engaging, and reliable digital events platform. “With Zoom Events, customers can produce and host impactful virtual experiences, including customer, company and public events and reach more people in engaging ways. We are excited to bring this innovation to our customers in India to help them create memorable events which attendees will love,” added Raje.","excerpt":"Over 7,000 customers have used Zoom Events, averaging more than 150 events daily since its launch in 2021","categories":["AI News"],"tags":["conferences","Zoom"],"author_name":"Bhuvana Kamath","publish_date":"2022-10-19T21:30:01","publication_year":"2022","word_count":274,"keywords":["Zoom","programming_languages:R","AI","innovation","ML","Scala","Git","conferences","RAG","Aim","GAN","R"],"extracted_tech_keywords":["AI","ML","Aim","RAG","R","Scala","Git","GAN","innovation","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/zoom-launches-zoom-events-in-india-to-elevate-virtual-experiences\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10164667,"title":"Infosys Open Sources Responsible AI Toolkit on its Topaz Suite","content":"Infosys has launched an open-source Responsible AI Toolkit as part of its Topaz Responsible AI Suite, reinforcing its commitment to ethical AI adoption. The initiative aims to help businesses address AI-related risks while fostering a secure, fair, and transparent AI ecosystem. The toolkit is built on Infosys’ AI3S framework (Scan, Shield, and Steer) and integrates advanced guardrails, including AI models and shielding algorithms to detect and mitigate security threats, privacy breaches, biased outputs, deepfakes, and other risks. It also enhances explainability by providing insights into AI-generated decisions without compromising performance. Designed for flexibility, the toolkit is customisable, compatible with diverse AI systems, and supports both cloud and on-premise deployments. “As AI becomes central to driving enterprise growth, its ethical adoption is no longer optional,” said Balakrishna D. R. (Bali), executive VP, global services head, AI and industry verticals, Infosys. “By making the toolkit open source, we are fostering a collaborative ecosystem that addresses the complex challenges of AI bias, opacity, and security.” Industry leaders have recognised the significance of the initiative. Joshua Bamford, head of Science, Technology and Innovation at the British High Commission, called it a “benchmark for technological excellence,” adding that it empowers enterprises, startups, and SMEs to adopt AI responsibly. Sunil Abraham, Public Policy Director – Data Economy and Emerging Tech at Meta, emphasised that open-source tools are crucial in ensuring AI safety, diversity, and economic opportunity. The Government of India also welcomed Infosys’ move. Abhishek Singh, Additional Secretary at MeitY, noted that open-sourcing the toolkit would “help in mitigating bias in AI models and enhance security, privacy, and fairness in AI-based solutions.” Infosys has been actively advancing Responsible AI initiatives. It was among the first companies to receive ISO 42001:2023 certification on AI management systems and is engaged in global AI policy discussions through organisations such as the NIST AI Safety Institute Consortium, WEF AIGA, AI Alliance, and Stanford HAI.","excerpt":"It also enhances explainability by providing insights into AI-generated decisions without compromising performance.","categories":["AI News"],"tags":[],"author_name":"Mohit Pandey","publish_date":"2025-02-26T16:11:24","publication_year":"2025","word_count":314,"keywords":["Go","fairness in AI","TPU","AI","innovation","responsible AI","Aim","GAN","AI safety","R"],"extracted_tech_keywords":["AI","Aim","TPU","R","Go","GAN","AI safety","responsible AI","fairness in AI","innovation"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/infosys-open-sources-responsible-ai-toolkit-on-its-topaz-suite\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10120980,"title":"6 Techniques to Reduce Hallucinations in LLMs","content":"LLMs hallucinate—generate incorrect, misleading, or nonsensical information. Some, like OpenAI CEO Sam Altman, consider AI hallucinations creativity, and others believe hallucinations might be helpful in making new scientific discoveries. However, they aren’t a feature but a bug in most cases where providing a correct response is important. So, what’s the way to reduce LLM hallucinations? Long-context? RAG? Fine-tuning? Well, long-context LLMs are not foolproof, vector search RAG is bad, and fine-tuning comes with its own challenges and limitations. Here are some advanced techniques that you can use to reduce LLM hallucinations. Using Advanced Prompts There’s a lot of debate on whether using better or more advanced prompts can solve LLM hallucinations. Source: X While some believe that writing more detailed prompts doesn’t help the case, others like Google Brain co-founder Andrew Ng see a potential there. Ng believes that the reasoning capability of GPT-4 and other advanced models makes them quite good at interpreting complex prompts with detailed instructions. “With many-shot learning, developers can give dozens, even hundreds of examples in the prompt, and this works better than few-shot learning,” he wrote. Source: X Many new developments are also being done to make prompts better like Anthropic releasing a new ‘Prompt Generator’ tool that can turn simple descriptions into advanced prompts optimised for LLMs. You can now generate production-ready prompts in the Anthropic Console.Describe what you want to achieve, and Claude will use prompt engineering techniques like chain-of-thought reasoning to create more effective, precise and reliable prompts. pic.twitter.com\/TqylVRkfP5— Anthropic (@AnthropicAI) May 10, 2024 Recently, Marc Andreessen also said that with the right prompting, we can unlock the latent super genius in AI models. “Prompting crafts in many different domains such that you’re kind of unlocking the latent super genius,” he added. CoVe by Meta AI Chain-of-Verification (CoVe) by Meta AI is another technique. This method reduces hallucination in LLMs by breaking down fact-checking into manageable steps, enhancing response accuracy, and aligning with human-driven fact-checking processes. CoVe involves generating an initial response, planning verification questions, answering these questions independently, and producing a final verified response. This method significantly improves the accuracy of the model by systematically verifying and correcting its own outputs. It enhances performance across various tasks, such as list-based questions, closed-book QA, and long-form text generation, by reducing hallucinations and increasing factual correctness. Chain-of-Verification (CoVe) reduces hallucinations in LLMs.It does this by using verification questions to fact-check the answers of an LLM.In experiments, CoVe decreases hallucinations across a variety of tasks. pic.twitter.com\/ri2DRNFoL2— TuringPost (@TheTuringPost) September 29, 2023 Knowledge Graphs RAG is not limited to vector database matching anymore, there are many advanced RAG techniques being introduced that improve retrieval significantly. For example, the integration of Knowledge Graphs (KGs) into RAG. By leveraging the structured and interlinked data from KGs, the reasoning capabilities of current RAG systems can be greatly enhanced. Source: X RAPTOR Another technique is Raptor, a method to address questions that span multiple documents by creating a higher level of abstraction. It is particularly useful in answering queries that involve concepts from multiple documents. Methods like Raptor go really well with long context LLMs because you can just embed full documents without any chunking. This method reduces hallucinations by integrating external retrieval mechanisms with a transformer model. When a query is received, Raptor first retrieves relevant and verified information from external knowledge bases. This retrieved data is then embedded into the model’s context alongside the original query. By grounding the model’s responses in factual and pertinent information, Raptor ensures that the generated content is accurate and contextually appropriate. Mitigating LLM Hallucinations via Conformal Abstention The paper ‘Mitigating LLM Hallucinations via Conformal Abstention’ introduces a method to reduce hallucinations in LLMs by employing conformal prediction techniques to determine when the model should abstain from providing a response. By using self-consistency to evaluate response similarity and leveraging conformal prediction for rigorous guarantees, the method ensures that the model only responds when confident in its accuracy. This approach effectively bounds the hallucination rate while maintaining a balanced abstention rate, particularly benefiting tasks requiring long-form answers. It significantly improves the reliability of model outputs by avoiding incorrect or nonsensical responses. Reducing Hallucination in Structured Outputs via RAG Recently, ServiceNow reduced hallucinations in structured outputs through RAG, enhancing LLM performance and enabling out-of-domain generalisation while minimising resource usage. The technique involves a RAG system, which retrieves relevant JSON objects from external knowledge bases before generating text. This ensures the generation process is grounded in accurate and relevant data. By incorporating this pre-retrieval step, the model is less likely to produce incorrect or fabricated information, thereby reducing hallucinations. Additionally, this approach allows the use of smaller models without compromising performance, making it efficient and effective. Researchers at ServiceNow Propose a Machine Learning Approach to Deploy a Retrieval Augmented LLM to Reduce Hallucination and Allow Generalization in a Structured Output TaskQuick read: https:\/\/t.co\/7TRlPqSqFSPaper: https:\/\/t.co\/J2kX5XTDFQ@ServiceNow…— Marktechpost AI Research News ⚡ (@Marktechpost) April 26, 2024 All these methods and more can help prevent hallucinations and create more robust LLM systems.","excerpt":"Using techniques like better prompts, knowledge graphs, and advanced RAG can help prevent hallucinations and create more robust LLM systems.","categories":["AI Features"],"tags":["Andrew Ng","Knowledge graphs","OpenAI"],"author_name":"Sukriti Gupta","publish_date":"2024-05-19T10:45:09","publication_year":"2024","word_count":834,"keywords":["Anthropic","knowledge graphs","Meta AI","machine learning","TPU","OpenAI","Knowledge graphs","AI","Andrew Ng","RAG","prompt engineering","few-shot learning"],"extracted_tech_keywords":["AI","machine learning","OpenAI","Anthropic","Meta AI","RAG","prompt engineering","few-shot learning","knowledge graphs","TPU"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/6-techniques-to-reduce-hallucinations-in-llms\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10004824,"title":"How AI-Powered Analytics &#038; Monitoring System Helps In Reducing Water Losses","content":"According to data, the total cost of water losses globally amounts to $15 billion, with two-thirds of these water losses being from low and middle-income countries. A lot of this could be attributed to leaks in the water distribution network, unauthorised consumption, and poor metering, which cost utilities over $4 billion each year. In fact, due to these losses, it has been estimated that the demand of water for Mumbai city will go up 71% from the current value, which could supply the needs of 16 cities like Mumbai. Thus, in an attempt to identify and reduce these water losses, Smartterra, has built an AI-powered operational intelligence system with the help of classic machine learning algorithms and geospatial data. Led by co-founders Gokul Krishna Govindu and Giridharan Sengaiah, this solution has been the first pilot in India not only to reduce water losses but also ensure water availability in developing cities. The company further aims to reduce utilities’ carbon footprints, which will also help in minimising the electricity bills of consumers. Also Read: This NLP-Driven Literature Search Engine Can Extract Relevant COVID Information The Hurdles Faced While developing the system, Smartterra, firstly, combined the disparate data sources of SCADA, CRMs, meter data management systems, geospatial data stores, maintenance records, and revenue\/billing systems to extract patterns from data, which was previously impossible. “However, with the help of the developed algorithm, it was easy for us to identify subtle patterns of loss in both the network and in water consumption,” said Gokul. Secondly, there were critical data gaps in these sources, which required to be addressed before feeding them into the AI algorithms. The company leveraged a combination of analytical techniques and subject matter expertise to fill these data gaps. Thirdly, the company had to integrate external data sources like satellite imagery, open-source mapping with traffic, urban activity, soil analysis to gain a 360-degree view of the problem being faced. Explaining further, Gokul stated, “Often, analyses of this nature are only carried out periodically since it is a time-consuming and expensive exercise. We wanted to do things differently, and therefore build a solution that would allow cities to continuously evaluate their networks, using feedback to up the accuracy of our models and provide predictions of improving quality over time.” Also Read: This AI Model Can Get Students Back In School Premise Post-COVID The Tech Behind To address the challenges, Smartterra took a three-step process. “At the topmost layer, we have a customised AI-powered analytics and monitoring system, which targets deficiencies in the data, identifying connections and network elements like pipes, pumps, valves, and stores to investigate anomalous behaviour,” explained Gokul. To facilitate this, the company had to address the critical feature of contextual cohorts to compare the identified anomalies with similar “normal” elements. Case in point — once the artificially intelligent machine tags a faulty meter, the company offers a comparison to “similar” consumers with working meters. This enables users to easily see which ones deviate most from their cohorts and therefore need to be prioritised. Keeping the explainability factor in place for improving decision-making, the company used techniques like SHAP, CORELS, and RuleFit to either directly build explainable models or explain existing models. Underneath the analytics layer is the unified data model, which consumes data from utilities periodically on various aspects of their operations, like supply, consumption, connection details, geospatial maps, and network information. Internally, the inputs are matched with custom data, adding necessary context for the city or region covered by the utility. “We also have custom data cleaning and preprocessing modules that ensure the quality of the data. This holistic view of the data allows us to use different sources effectively,” said Gokul. “We incorporate urban dynamics through land use, satellite imagery, and terrain\/soil data, which helps us capture the nuances in city activity, socio-economics, and local variability.” Leveraging geospatial data provides an edge in the interpretation of water consumption behaviour, demand projections, and network planning. Finally, the company leveraged MEAN (Mongo – Express – Angular – Node) stack to design workflow engines so that the users can now create tasks for on-ground verification of the given predictions. This allows analysts to push results to field operators who can determine whether a fix is required or not. Further, these operators provide additional data, as part of a continuous learning system, where feedback from on-ground investigations is put back to models in order to improve them. The workflow engine ensures that the AI system is a full-cycle from data to models to on-field verification and back to data. The company used a variety of models to build the solution, where they started with classic machine learning algorithms that have been demonstrated to work well, such as RandomForest (classifiers and regressors), support vector machines, and gradient boosted trees. Additionally, Gokul is also exploring using long short-term memory and recurrent neural network models, with TensorFlow and Keras, to take advantage of time-series data. Also Read: Piramal Sarvajal Using IoT To Tackle Safe Drinking Water Issue For Rural India Benefits & Future Plan With the initial deployed pilot project of the AI-powered analytics, the company was able to identify several obstacles to precise water usage, like incorrect billing to customers, faulty meters and mismatches in the data. The AI system showed that there was a 3.2% potential increase in monthly revenue using just “soft” fixes that were relatively inexpensive. Also, “the results were validated by an on-ground survey that confirmed that predictions made by the AI models were 80% accurate,” stated Gokul. Broadly, these results were aimed to help operators reduce water losses in their networks, resulting in more availability of water to required citizens. The solution has been designed to ensure that the residents of the developing cities receive consistent, high-quality water supply so that they can focus on building better lives. The company is also in talks with various utilities in India and South-East Asia to develop solutions to reduce their losses. “We hope to have several pilot projects underway in the coming months,” concluded Gokul.","excerpt":"According to data, the total cost of water losses globally amounts to $15 billion, with two-thirds of these water losses being from low and middle-income countries. A lot of this could be attributed to leaks in the water distribution network, unauthorised consumption, and poor metering, which cost utilities over $4 billion each year. In fact, […]","categories":["AI Features"],"tags":["Analytics Case Study","Machine Learning"],"author_name":"Sejuti Das","publish_date":"2020-08-13T18:00:07","publication_year":"2020","word_count":1005,"keywords":["machine learning","Keras","AI","neural network","Machine Learning","RAG","NLP","Aim","analytics","Analytics Case Study","TensorFlow","R"],"extracted_tech_keywords":["AI","machine learning","neural network","NLP","analytics","Aim","TensorFlow","Keras","RAG","R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-ai-powered-analytics-monitoring-system-helps-in-reducing-water-losses\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":22838,"title":"Mumbai startup’s proprietary legal research platform is driving the next generation of legal tech","content":"Pensieve co-founder Gaurav Shrivastava AI charting its way through legal territory is not new – for the last few years ROSS Intelligence has been changing the way big law firms perform legal research and its AI-powered solution comes strikingly close to the experience of working with a human lawyer. And it’s not just ROSS Intelligence that is disrupting the business of law – IBM Watson & Lex Machina has been providing legal analytics for companies and are ushering in an era of a data-driven lawyer. AI disrupting the business of law in India Now that AI has wended its way outside IT functions and upended legal sector – India is playing catch up. According to a Deloitte study, in-house teams are looking for tech savvy, integrated service providers who offer more than traditional legal advice. One of India’s leading full-service law firm, Mumbai-headquartered Cyril Amarchand Mangaldas has reportedly become Asia’s first law firm to embrace AI earlier this year. Canada-based Kira Systems, machine learning software provider revealed in their blog that the move was spurred to improve the delivery model of legal services. Global legal and tax consulting firm Nishith Desai Associates leverages data management and bandwidth management systems for data-intensive tasks such as analytics, regulatory compliance and contract review. AI has made way for Legal Process Outsourcing (LPO) – a growing trend, spurred by the availability of proprietary AI software developed by a bunch of startups that are shaping the legal tech sector in India. Pensieve – Indian AI startup is disrupting the legal sector Pensieve, founded by serial entrepreneur Gaurav Shrivastava, (co-founder of Zimmber bought by Quikr in 2017) and Prahlad K. Routh, a Materials Scientist and an Adjunct Professor at Columbia University has developed a proprietary AI-driven legal research platform called Mitra that is being used by 300+ customers already. Currently, the Mumbai-based startup is working with IIT-Madras to mature the product and its AI solution is being used by major law firms like Argus & K Law. The startup founded in 2016 is using AI & NLP to improve efficiency of law firms and also made it into the Axilor Accelerator Program, one of India’s largest accelerator program. Talking about making it into the sixth accelerator batch, Shrivastava said, “Axilor has helped has immensely in understanding the business priority. We can expand the product range from a domain specific to a domain agnostic catering to technology solutions for problems in areas such as Linguistic & Text Analytics”. Stressing on the need for AI in legal sector, Shrivastava added, “Legal research is one of the very crucial and time-consuming parts for any lawyer, corporate and law student. It also affects productivity as a whole. Our solution is aimed towards decreasing this productivity gap, which in turn gives them more time where it matters – with clients. Currently our solution is geared to benefit the legal fraternity with machine learning models polished for legal documents”. Mitra.ai – Legal Research Platform is a cut above the rest Mitra.ai is the flagship product of Pensieve – and an advanced tech platform powered by artificial intelligence that searches, recommends and learns from user behaviour. The platform is equipped with machine learning models that runs on thousands of documents to give lawyers and case assistants the most relevant results than a bullion search handler. “It’s one of a kind natural language query handler lets users ask queries in a language that’s naturally in use. The intelligent search engine powered with AI and NLP goes beyond keyword search and it also understands the context to populate results by various machine learnt taggings that we have developed. The platform also understands results selected by users and their navigation to give personalised recommendations,” he added. How does it differ from the competition? Of late, India’s startup boom has spawned vertical AI startups focused on solving a multitude of legal tasks such as Document Review, Case Predictions & even advisory services. On the USP front, Shrivastava asserts Mitra.ai platform acts as a one-stop-solution for legal research, is outfitted with a search engine, defence builder notepad and boasts of a feature rich user dashboard . “We have seen multiple players starting to compete in this market but in India most of them are in budding stage. Our vision is to build platform with user friendly UI yet complex algorithms running behind it , to address the productivity gap of legal fraternity,” he added. The other product – Smriti is an AI-powered contract life cycle management tool. Currently, the startup is integrating project and document management capabilities to empower the users with spectrum of features. Also, the founders are dabbling with the idea of a Virtual legal assistant such as chatbot to address conversational queries. However, Shrivastava also added he is yet to ascertain the full scope for it in the industry, that is just beginning to adopt AI software. Challenges in deploying AI solutions in Legal Given the data-intensive nature of AI, building a data corpus is the primary and still the most difficult thing to venture into any sector, pointed out Shrivastava. But besides the need for petabytes of data, there are also issues related to data confidentiality & the black box problem of AI. According to Shrivastava, the founders have considerable expertise in building tech & business capabilities and would be introducing a number of products for legal sector first. “The idea is to exploit ML and AI algorithms to address the productivity gap because of data discovery and Intelligence. Legal domain has been deprived of technology solutions before, especially legal research. Mitra.ai is addressing this through its sophisticated yet user friendly platform,” he said. Team Size Pensieve has a 10-member strong team with a mix of tech, business, product & domain experts to identify , build and deliver product efficiently. The tech team has four members onboard with machine learning experts and members to handle frontend. “Going forward we would seek folks in ML domain to even enrich our tech capabilities and folks would be needed for B2B sales and support also,” he said. Plans are also afoot to expand to the US market and the startup is in talks with law firms.","excerpt":"AI charting its way through legal territory is not new – for the last few years ROSS Intelligence has been changing the way big law firms perform legal research and its AI-powered solution comes strikingly close to the experience of working with a human lawyer. And it’s not just ROSS Intelligence that is disrupting the […]","categories":["AI Startups"],"tags":["AI (Artificial Intelligence)","law"],"author_name":"Richa Bhatia","publish_date":"2018-03-21T04:29:50","publication_year":"2018","word_count":1029,"keywords":["Go","law","artificial intelligence","machine learning","AI","ML","RAG","NLP","Aim","analytics","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","NLP","analytics","Aim","RAG","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-startups\/mumbai-startups-proprietary-legal-research-platform-is-driving-the-next-generation-of-legal-tech\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10125712,"title":"AIM Workshop ALERT: RAG &amp; Fine Tuning in GenAI with Snowflake","content":"AIM in partnership with Snowflake is hosting an exclusive webinar on 25th July from 6 PM to 7:30 PM titled “RAG & Fine Tuning in GenAI with Snowflake”. This event is designed to provide AI enthusiasts, researchers, and professionals with a deep dive into the advanced techniques of retrieval augmented generation (RAG) and fine tuning, crucial for developing AI applications. Meet the Expert: Prashant Wate Prashant Wate, Senior Sales Engineer at Snowflake India, will lead this insightful session. With extensive experience in AI and enterprise solutions, Wate will guide the participants through the intricacies of RAG and fine-tuning, demonstrating how these techniques can revolutionise your AI projects. Register Now! What You Will Learn? Understanding RAG and Fine Tuning This webinar will clarify the differences between RAG and fine-tuning, alongside how and when to use these methods effectively. Gain practical knowledge and explore real-world applications that will set the stage for advanced AI innovations in your organisation. Enhancing Generative Models with RAG RAG is a technique designed to reduce hallucinations (incorrect responses) in generative models by integrating private datasets and vector embeddings. This method significantly boosts the capabilities of generative models, ensuring they align with your enterprise requirements. Optimising Models with Fine-tuning Fine-tuning involves adjusting pre-trained LLMs on a specific dataset to improve their performance on targeted tasks. This process is essential for optimising models to better understand and respond to domain-specific queries, ensuring higher accuracy and relevance in the generated outputs. Leveraging Snowflake’s Platform Discover how Snowflake’s robust platform for data governance and management can be leveraged to develop and deploy end-to-end AI applications using RAG. With features like Snowflake Cortex, Streamlit, and Snowpark, you can achieve this without the need for additional integrations, infrastructure management, or data movement. Why Attend? This webinar is particularly valuable for developers who are eager to enhance their skills and knowledge in Generative AI. By attending, you will be equipped with the tools and techniques needed to drive advanced AI innovations in your organisation. Register Now! Secure your spot for the AIM Workshop on 25th July from 6 PM to 7:30 PM. Register today and take the first step towards mastering RAG and fine-tuning with Snowflake.","excerpt":"This webinar will clarify the differences between RAG and fine-tuning, alongside how and when to use these methods effectively.","categories":["AI Highlights"],"tags":["Snowflake"],"author_name":"Mohit Pandey","publish_date":"2024-07-03T19:03:27","publication_year":"2024","word_count":362,"keywords":["GenAI","TPU","AI","ML","RAG","Aim","Streamlit","generative AI","retrieval augmented generation","Snowflake"],"extracted_tech_keywords":["AI","ML","generative AI","GenAI","Aim","Streamlit","retrieval augmented generation","RAG","TPU","Snowflake"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/aim-workshop-alert-rag-fine-tuning-in-genai-with-snowflake\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10005984,"title":"BigBird &#8211; Google’s ‘Brahmastra’ For NLP Supremacy?","content":"Google transformer-based models like BERTshowcased immense success with NLP tasks; however, came with a significant limitation of quadratic dependency in-memory storage for the sequence length. A lot of this could be attributed to its full attention mechanism for sequence lengths. Such a self-attention mechanism can create several challenges for processing longer text sequences. Thus to omit that problem, Google released its new deep learning model — BigBird, with sparse attention mechanism, which eliminates the limitation with linear dependency. According to a recent paper on BigBird, researchers showcased how, despite having sparse attention, the new model — BigBird can preserve properties of quadratic, full attention models. This ability has enabled the model to showcase an enhanced performance in processing eight times longer text sequences than other transformer models. Further, the researchers also answered two significant queries with BigBird — firstly, achieving practical benefits of the fully quadratic self-attention scheme using fewer inner-products; and secondly, preserving expressivity and flexibility of the original network with sparse attention mechanisms. Also Read: How To Use BERT Transformer For Grammar Checking? BigBird’s Sparse Attention Mechanism For Longer Sequences For theoretical experimentation, researchers, in the paper, showcased how the sparse attention mechanism can be as beneficial and expressive as a full-attention mechanism. To facilitate this, researchers firstly highlighted the use of sparse attention mechanisms for standalone encoders like the BERT model, then turned into Universal Approximators of sequence to sequence function. And, secondly, how sparse encoder-decoder transformers are Turing Complete. While experimenting, researchers realised that the main challenge of the sparse attention mechanism is the ability to compute the contextual mapping. Thus, to get around that, researchers developed a sparse shift operator for managing the entries of matrices. Further, unlike other transformer models where the full attention is applied directly, the sparse attention mechanism works token by token. Therefore to use the proposed mechanism, they also needed to define a suitable modification of tokens to control their reactions. With these steps, the researchers were also able to display that sparse attention mechanism comes with an additional cost, due to its requirement of polynomially more layers. For practical experimentations, researchers chose NLP and genomics tasks to highlight BigBird’s capability. Image 01 For the NLP experiment, researchers chose three representative tasks — basic masked language modelling for the better contextual representation of longer sequences; handling longer sequences for Q&A and extended document classification for extracting information. Following the experiments, the researchers highlighted that with an 8x increase of the text sequence, the model was able to achieve extraordinary performance for the tasks above. Image 02: Fine-tuning results for QA tasks The tasks chosen for the comparison were very competitive, therefore required multiple highly engineered systems for confirming each dataset to its respective output formats. To have fair and accurate results, researchers had to use some additional regularisation for training the model. Image 03: Result of the document classification task. Coming to the genomics tasks, the researchers used a dataset built on EPDnew and the report F1 on a test dataset to fine-tune the BigBird model. While comparing it with the previous models, it has been noted that the proposed model achieved a 5% jump in the accuracy. Also Read: How I used BERT to Analyse Twitter Data BigBird vs BERT When compared Google’s BigBird with other transformer models like BERT, the proposed model outsmarted both RoBERTA and Longformer with extraordinary results in all the four question-answer datasets — HotpotQA, NaturalQ, TriviaQA, and WikiHop. (Image 02) Further, when the model has been worked on the document classification task, it was compared on IMDB, Yelp-5, ArXiv, Patents and Hyperpartisan datasets. This result highlights BigBird’s remarkable capability on the ArXiv dataset surpassing RoBERTA and SoTA with an F1 score of 92.31%. (Image 03) Along with these, Google’s BigBird was also compared with CNNProm and DeePromoter on genomics tasks. Here, the results highlighted that the proposed model had achieved 99.9% accuracy with a 5% increase than the previous models. Also Read: How Syntactic Biases Help BERT To Achieve Better Language Understanding Wrapping Up With these results in hand, it can be established that with a sequence length of 4,096, the BigBird Model can provide an accurate and precise result on theoretical as well as practical experiments. According to the researchers, the results are complemented by showcasing that moving to the sparse attention mechanism can bring advantages but would incur a cost. Read the whole paper here.","excerpt":"Google transformer-based models like BERTshowcased immense success with NLP tasks; however, came with a significant limitation of quadratic dependency in-memory storage for the sequence length. A lot of this could be attributed to its full attention mechanism for sequence lengths. Such a self-attention mechanism can create several challenges for processing longer text sequences. Thus to […]","categories":["Global Tech"],"tags":["big data storage format","Google","NLP"],"author_name":"Sejuti Das","publish_date":"2020-09-04T13:00:00","publication_year":"2020","word_count":733,"keywords":["big data storage format","Go","TPU","attention mechanism","AI","Transformers","RAG","NLP","BERT","deep learning","Google","R"],"extracted_tech_keywords":["AI","deep learning","NLP","Transformers","RAG","TPU","R","Go","attention mechanism","BERT"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/bigbird-googles-brahmastra-for-nlp-supremacy\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10017237,"title":"Top 8 Autonomous Driving Open Source Projects One Must Try Hands-On","content":"The past few years have seen active development in autonomous driving by organisations and academia. One of the standard practices in autonomous driving is developing and validating prototypes of driving in simulators. The researchers worldwide have been developing these simulators to support the training and development of such selfless driving systems. Let’s take a look at the top 8 autonomous driving open-source projects one must try their hands-on. (The list is in no particular order) 1| Carla About: Carla is an open-source simulator for autonomous driving research. It has been developed to encourage development, training as well as validation of autonomous urban driving systems. In addition to the open-source code and protocols, this simulator provides open digital assets, such as urban layouts, buildings and vehicles. The simulation platform promotes flexible specification of sensor suites. Carla can be used to study the performance of three approaches to autonomous driving — modular pipeline, an end-to-end model trained via imitation learning, an end-to-end model trained via reinforcement learning (RL). Carla’s features include scalability via a server multi-client architecture, autonomous driving sensor suite, flexible API, fast simulation for planning and control, maps generation, traffic scenarios simulation, ROS integration, and autonomous driving baselines. Know more here. 2| SUMMIT About: SUMMIT or Simulator for Urban Driving in Massive Mixed Traffic is a high-fidelity simulator that promotes the advancement and testing of crowd-driving algorithms. It simulates unregulated and dense urban traffic for heterogeneous agents regardless of the worldwide locations that OpenStreetMap supports. The SUMMIT simulator is built as an extension of CARLA and inherits the physics and visual realism for autonomous driving simulation. It supports a wide range of applications, including perception, vehicle control and planning, and end-to-end learning. Know more here. 3| Flow About: Flow is an open-source computational framework for deep RL and control experiments for traffic microsimulation. Developed by the members of the Mobile Sensing Lab at UC Berkeley, Flow is basically a deep reinforcement learning (RL) framework for mixed autonomy traffic. This simulator is a traffic control benchmarking framework that gives a set of traffic control scenarios as benchmarks, tools for designing custom traffic scenarios, and integration with deep reinforcement learning and traffic microsimulation libraries. Know more here. 4| PGDrive About: PGDrive is an open-ended and highly configurable driving simulator that integrates the key feature of the procedural generation (PG). The simulator defines multiple basic roadblocks such as ramp, fork, and roundabout with configurable settings and a range of diverse maps can be assembled from those blocks with procedural generation, which is further turned into interactive environments. PGDrive is built upon the Panda3d and Bullet engine with optimised system design. The simulator can gain up to 500 simulation steps per second while running a single instance on PC. Know more here. 5| Deepdrive About: Deepdrive is an open simulation platform built to accelerate progress and increase transparency in self-driving. The features of Deepdrive include support for Linux and Windows, interface through the Gym API using a reward function based on speed, safety, legality, and comfort, pre-trained example agent, training code, and dataset to get started building AI models. It is also enabled with up to eight cameras and a dataset of around 100GB and 8.2 hours. Know more here. 6| AirSim About: developed by Microsoft, AirSim is an open-source, cross-platform simulation platform for autonomous systems. Built on Unreal Engine, AirSim supports software-in-the-loop simulation with popular flight controllers and hardware-in-loop with PX4 for physically and visually realistic simulations. AirSim is developed as an Unreal plugin which can be dropped into any unreal environment. The developers at Microsoft developed AirSim as a platform for researchers in AI and to experiment with deep learning, computer vision (CV) and reinforcement learning (RL) algorithms for driverless vehicles. Know more here. 7| LGSVL Simulator About: LGSVL Simulator is an open-source autonomous vehicle simulator developed by LG Electronics America R&D Centre. It is an HDRP Unity-based multi-robot simulator for autonomous vehicle developers that provide an out-of-the-box solution that can meet developers’ needs to focus on testing their autonomous vehicle algorithms. The simulator currently has integration with TierIV’s Autoware and Baidu’s Apollo 5.0 and Apollo 3.0 platforms, can generate HD maps, and can be immediately used to test and validate a whole system. Know more here. 8| Gym-Duckietown About: Gym-Duckietown is a simulator for the Duckietown Universe. It is written in pure Python\/OpenGL (Pyglet). The simulator works by placing RL agents inside of an instance of a Duckietown: a loop of roads with turns, intersections, obstacles, Duckie pedestrians, and other Duckiebots. Duckietown is a fully-functioning autonomous driving simulator that can train and test machine Learning, Reinforcement Learning, Imitation Learning, or even classical robotics algorithms. Know more here.","excerpt":"The past few years have seen active development in autonomous driving by organisations and academia. One of the standard practices in autonomous driving is developing and validating prototypes of driving in simulators. The researchers worldwide have been developing these simulators to support the training and development of such selfless driving systems.  Let’s take a look […]","categories":["AI Trends"],"tags":["ai machine learning and data analytics","automation testing","autonomous car technologies","autonomous systems","autonomous technology","Autonomous Vehicles","open source ios"],"author_name":"Ambika Choudhury","publish_date":"2021-01-06T16:00:00","publication_year":"2021","word_count":776,"keywords":["Go","machine learning","automation testing","AI","ai machine learning and data analytics","Scala","autonomous car technologies","open source ios","autonomous technology","computer vision","autonomous systems","deep learning","RAG","Python","Git","Autonomous Vehicles","R"],"extracted_tech_keywords":["AI","machine learning","deep learning","computer vision","RAG","Python","R","Go","Scala","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-8-autonomous-driving-open-source-projects-one-must-try-hands-on\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10101741,"title":"Why OpenAI Partnered with G42","content":"OpenAI recently announced its partnership with Abu Dhabi-based AI and cloud computing company G42. With this move, the possibility of UAE competing with OpenAI is subdued, and the strategic partnership may have a little something for both the parties. OpenAI’s Demographic Push In G42’s partnership announcement, the company spoke about leveraging OpenAI’s generative AI models for UAE’s financial services, energy, healthcare and many other sectors. OpenAI, on the other hand, will not only be able to expand in the Emirates market but also probably leverage G42’s Arabic language model. G42 recently launched an Arabic language AI model Jais, which contains 13 billion parameters and combines Arabic and English data. The model was built in collaboration with academicians and engineers, partly from the scarcity of bilingual language models. Jais was built on supercomputers produced by Cerebras Systems. The partnership with OpenAI will probably help the company achieve its multilingual model goals. G42 Group CEO Peng Xiao and OpenAI CEO and co-founder Sam Altman. Source: G42 Altman has been keen on developing demographic specific models. In his visit to Japan in June, he spoke about wanting to build better models for Japanese language and culture. Furthermore, he even indicated his desire to start an office in Tokyo as SoftBank shows interest in investing in OpenAI. Having established its presence in London and Dublin by opening new offices there, with another one coming up in Japan soon, it won’t be a surprise if UAE gets the next one. Public-Private Collab In June, Altman met Maktoum bin Mohammed Al Maktoum, deputy prime minister and minister of finance of the UAE, to explore AI development opportunities with the country. The ruler had even tweeted about how they explored ways to strengthen partnerships around AI solutions. The UAE government takes active interest in the advancements of technological research and development, especially in AI. It either funds (partners) or advises companies working in the domain. Technology Innovation Institute (TII), the organisation behind open-source LLM Falcon, is a government-funded research institute. It has been on the forefront of releasing LLM models that surpassed even Meta’s LLaMA capabilities, a few months ago. Close on the heels of TII, spearheading further AI advancements in the country is G42. Interestingly, UAE national security advisor Sheikh Tahnoon bin Zayed Al Nahyan, is also the chairman of G42 group. He also chairs sovereign wealth funds including government holding company ADQ worth $110 billion, that controls critical sectors of the economy. Interestingly, UAE was the first country to appoint a minister of state for artificial intelligence, Omar Sultan Al Olama, in 2017. With funding and direction, the government has been spearheading AI progress in the country. And with G42’s latest partnership with OpenAI, the country now has the best of both worlds. G42 and Big-Tech Companies While OpenAI is the latest collaboration, G42 has also tied up with big-tech and emerging companies that will help accelerate the AI vision for UAE. G42 announced its partnership with Microsoft, in April, and last month announced their collaboration to boost cloud and technology infrastructure in the UAE. The companies will focus on AI solutions for health, energy and many other domains, similar to the OpenAI-G42 partnership plan. The collaboration will see Microsoft expand its Azure cloud service within UAE by leveraging Khazna Data Centres, a joint venture between G42 and government telecom company Etisalat. AI compute company Cerebras Technologies had signed a deal with G42 two years ago to bring high-performance AI compute to the Middle East. In July, both the companies unveiled Condor Galaxy, a network of nine interconnected supercomputers that significantly reduces the training time for AI models. The OpenAI-G42 partnership emphasises the construction of AI systems through collaboration, rather than competition. As G42 CEO Peng Xiao stated, the partnership represents a “convergence of value and vision”, and Altman believes that this collaboration will help yield effective solutions that “resonate with the nuances of the region”. A win-win for all.","excerpt":"While AI companies are busy competing with each other, OpenAI and G42 have formed a first-of-its-kind partnership to work together","categories":["Global Tech"],"tags":["Cloud Computing","G42","Jais","japan","Microsoft","OpenAI","Parameters","Sam Altman","Softbank","TII","uae"],"author_name":"Vandana Nair","publish_date":"2023-10-20T10:00:00","publication_year":"2023","word_count":655,"keywords":["G42","Jais","TII","R","Softbank","japan","Sam Altman","artificial intelligence","RAG","Go","AI","cloud computing","Parameters","generative AI","GAN","uae","OpenAI","Cloud Computing","Azure","Microsoft"],"extracted_tech_keywords":["AI","artificial intelligence","generative AI","OpenAI","RAG","cloud computing","Azure","R","Go","GAN"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/why-openai-partnered-with-g42\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10057750,"title":"Is AI fast becoming a technology built on worker exploitation from Global South?","content":"Data labelling and annotation play a crucial role in ML development. Most ML and deep learning models are data-hungry, and manually labelling each image, text, audio, etc., in large datasets is tedious and a labour-intensive process. Hence, many companies choose not to handle data labelling and use specialised software. But when the pre-built solutions do not meet their specific needs, they move to modifiable open-source platforms. Various platforms like Amazon Mechanical Turk, Clickworker, and Lionbridge AI offer on-demand data labelling services. But, companies fail to focus on the ethical considerations around the processes and decisions that go into building ML datasets. According to Google research, data generated in crowd work tasks is generally shaped by many social factors, and these datasets continue to shape systems for long, even after worker engagement ends. Google argues that this impacts not just the current but also the future models built from such data and that understanding the perspectives within datasets is essential to understanding resulting models and their potential social impact. There are two major factors that impact the annotators and their work that can hinder in creating an ethical dataset. The first one is how the annotators’ personal experiences impact their annotations, and the second one is their work environment, i.e., their relation with the crowdsourcing platform. These factors are crucial as their individual perspectives and biases may get encoded within the dataset labels. The socio-economic background of annotators can impact the data set The importance of data labelling has grown as deep learning techniques require large amounts of data to train their models. According to Grand View Research, the data collection and labelling market is expected to grow to USD 8.2 billion by 2028. Over 80% of the ML development process consists of data preparation tasks like collection, labelling and cleaning. According to a paper by researchers from Cornell, Princeton, University of Montreal, and the National Institute of Statistical Sciences, a lot of data annotation and labelling work is done outside the United States and Western countries. The work exploits workers from around the world where labour is cheap. Companies like Samasource, Mighty AI and Scale AI operate in the United States, but crowdsource workers from around the world, primarily from sub-Saharan Africa and Southeast Asia. This leads to a high disparity between the profits earned by data labelling companies and how much they pass on to their workers. Little attention is documented about annotator positionality, and their social identity shapes their understanding of the world. Crowd workers are selected by task requesters based on the quality metrics and not on any socially defining features. This is concerning as many times crowdsourced annotations are used to build datasets that capture subjective phenomena like hate speech and sentiment. There is value in accounting and acknowledging workers’ socio-cultural backgrounds from the perspective of data quality and also social impact. Google research suggests that accounting for lived experiences of annotators as expertise may also be of great utility in some cases. For example, women experience higher rates of sexual harassment online and are more likely to identify it, especially the ones who have experienced online abuse. Incorporating antiracist activists’ perspectives into hate speech annotations yielded better-aligned models. An important line to be answered in data collection is how much annotator subjectivity matters for the task at hand and its impact on the end result. Worker experience has a crucial role in the quality of the dataset Workplace experiences of dataset annotation add another layer of considerations that can affect the work and can have an impact on the AI model. The issues at work could relate to compensation, imbalances in the relationship between the worker and the requester, and even the structure of annotation work itself. Though data labelling is not physically intensive, workers have reported that the pace and volume of their tasks are “monotonous” and “mentally exhausting.” In the Global South, local companies like Fastagger in Kenya, Sebenz.ai in South Africa, and Supahands in Malaysia have begun to proliferate. With the scaling AI development, the expansion of these companies does open new doors for low-skilled labourers but also presents a chance for exploitation. Research also suggests that the majority of crowd workers (94%) have had work that was rejected or for which they were not paid. But the requesters retain full rights over the data they receive. This system does enable wage theft. But, currently, many countries do not have regulations around crowd work. The minimum wage, too, is not applicable for crowd workers as they are not ‘workers’ but ‘independent contractors.’ The geographically distributed and anonymous nature of crowdsourced annotation and labelling work imposes significant barriers to collective action. Wrapping up While choosing crowd work datasets, it is essential for companies to be intentional about whether to lean on annotators’ subjective judgements. Not accounting for task subjectivity can lead to inadvertent biases and miss critical insights about tasks that could benefit from the annotators’ lived experiences. Clarifying such aspects of the tasks has ramifications for how properly the datasets capture the aspects of human intelligence. While choosing an annotation platform, it is important to choose the one that allows flexibility in designing custom annotator pools and can include various socio-demographic axes. These decisions should be guided by considering the communities that will be most impacted by models built from the data and the ones that could be harmed the most if they are not represented. To keep the dataset high quality, it is also essential to compare the minimum pay requirements across different platforms and choose to support the one that upholds fair pay standards.","excerpt":"While working with crowd work platforms for datasets, it is essential to consider annotator subjectivity as it has the capability to make the data set of extremely high or low quality, which in turn affects the whole ML model.","categories":["AI Features"],"tags":["data annotation","data labelling"],"author_name":"Meeta Ramnani","publish_date":"2022-01-06T11:00:00","publication_year":"2022","word_count":934,"keywords":["Go","programming_languages:R","AI","ML","programming_languages:Go","data labelling","data annotation","deep learning","data quality","ViT","R"],"extracted_tech_keywords":["AI","ML","deep learning","R","Go","data quality","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/is-ai-fast-becoming-a-technology-built-on-worker-exploitation-from-global-south\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":44789,"title":"Intel Unveils Its First AI Chip Spring Hill, Says Facebook Already Using It","content":"In a radical new step to deal with big data generated by organisations, Intel this week launched an artificial intelligence processor called Nervana NNP-I or ‘Spring Hill’. According to reports, this processor has been designed for large computing centres. Intel Nervana NNP-I is based on a 10 nanometer Ice Lake processor and is purpose-built specifically to accelerate deep learning deployment at scale, offering excellent performance for major data centre workloads. Intel also said that this processor also offers a high degree of programmability without compromising performance to power efficiency. We’re sharing details on our Nervana NNP-I accelerator, designed to accelerate deep learning deployment at scale. https:\/\/t.co\/tHEJsIXcDr #HotChips31 pic.twitter.com\/CB2YPryK3n — Intel (@intel) August 20, 2019 “It’s easy to program, has short latencies and fast code porting while supporting all major deep learning frameworks, allowing the world’s leading cloud service providers and enterprises to take advantage of advanced inference performance,” said Intel in a statement. A noted wire also said that Facebook has already has started using the product. Naveen Rao, general manager of Intel’s artificial intelligence products group, told the news wire, “In order to reach a future situation of ‘AI everywhere’, we have to deal with huge amounts of data generated and make sure organizations are equipped with what they need to make effective use of the data and process them where they are collected… These computers need acceleration for complex AI applications.”","excerpt":"In a radical new step to deal with big data generated by organisations, Intel this week launched an artificial intelligence processor called Nervana NNP-I or ‘Spring Hill’. According to reports, this processor has been designed for large computing centres. Intel Nervana NNP-I is based on a 10 nanometer Ice Lake processor and is purpose-built specifically […]","categories":["AI News"],"tags":["Intel"],"author_name":"Prajakta Hebbar","publish_date":"2019-08-21T14:11:59","publication_year":"2019","word_count":233,"keywords":["big data","artificial intelligence","programming_languages:R","AI","deep learning","GAN","R","Intel"],"extracted_tech_keywords":["AI","artificial intelligence","deep learning","R","big data","GAN","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/intel-unveils-its-first-ai-chip-spring-hill-says-facebook-already-using-it\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10129354,"title":"Google-Backed Cropin&#8217;s New AI Platform Could Tackle Food Crisis","content":"In 2022, Cropin launched what it claimed was the ‘world’s first cloud for agriculture’. Now, it is adding generative AI into the mix. The Bengaluru-based agritech firm just launched Cropin Sage, which it claims is the ‘world’s first real-time agri-intelligence platform’. Powered by Google Gemini, Cropin Sage enables users to make informed decisions based on historical, present, and future data on cultivation practices, crops, irrigation, climate, and soil. To provide accurate data at such a scale, Cropin has partitioned the global map into 5×5 km grids, assigning a unique ID to each. Users can query Sage regarding crop cultivation feasibility within specific grids. Sage then provides data visualisation outputs in response to these inquiries. The data is then processed by Sage, which sits inside the Cropin cloud, into different grid sizes and aggregated at different temporal frequencies, including yearly, seasonal, monthly, weekly, and daily, based on customer requirements. (Cropin Sage) Gemini Queries Large Datasets Sage leverages Cropin’s advanced crop models, which analyse crop performance in detail. It also leverages a smart climate model, which integrates data spanning the last 40 years and current conditions and forecasts weather for the coming year. “We integrate Gemini with our climate models, crop models, and proprietary knowledge graphs. These elements are fused together to provide comprehensive solutions,” Krishna Kumar, chief executive officer at Cropin, told AIM. At its core, Sage makes sense of the enormous datasets spanning over terabytes of data. Each dataset could easily be over 100 gigabytes in size, depending on the country and the availability of historical data, besides other datasets such as temperature, climate, and even socioeconomic data. In fact, Sage uses Gemini to query these large datasets and generate responses in a consumable manner within seconds. The startup’s proprietary data spans over 350 crops and 10,000 varieties in 103 countries. (Cropin Sage) Sage does not present the data in a text format but leverages a visualisation tool to transform complex datasets into intuitive graphs, charts, and interactive displays that enhance understanding and decision-making. To do so, Sage leverages the Gemini Flash 1.5 model to convert user queries to SQL queries, which allows it to generate user-friendly, grid-based data in a visually appealing platform. It also leverages Google Kubernetes Engine (GKE) services to scale its operations in real-time as demand increases, processing massive volumes of data. According to Kumar, Sage will initially provide information on 13 crops expanding to 15 by the end of the year. These crops, which include corn, paddy, soy, wheat, onion, potato, sugarcane, cocoa, coffee, and cotton among others, accounts for 80% of total global production. Moreover, users can seek data on these 13 crops irrespective of where they are. “They could be anywhere, any country, but if they seek data on these crops, they will get the data,” Kumar emphasised. Who Benefits from Sage? Rajesh Jalan, the CTO and head of engineering at Cropin, who was also present during the conversation with Kumar, told AIM that the model is highly accurate. Hallucinations remain a persistent issue with LLMs. In fact, upon Gemini’s release, it generated controversial responses that led Google CEO Sundar Pichai to deem them ‘unacceptable’. However, according to Jalan, it does not matter here because Sage provides all the data as per the query. “The customers can see all the data and if the data is wrong, the customer will be able to figure it out.” So far, the company has nearly 250 B2B customers worldwide and has digitised over 30 million acres of farmlands, positively impacting over 7 million farmers worldwide. Cropin Sage will benefit CPG players, seed manufacturers, food processors, multilateral organisations, financial institutions and governments, according to Kumar. Global food systems today are encountering substantial challenges, exacerbated by various adverse factors that significantly hinder farmers’ capacity to meet food production demands. This issue is exemplified by the cocoa crisis, which has deeply affected chocolate manufacturers. Cocoa prices have surged by 400% within a year, posing significant affordability challenges for many in the chocolate industry. “Consider an enterprise in the food and agribusiness sector with a supply chain spanning multiple countries like PepsiCo, for instance. Due to an unstable supply chain caused by changing climate changes, if they want to expand the production of potatoes to a new country, they can leverage Sage to predict suitable locations for sourcing and investments,” Kumar said. Sage can also benefit multiple governments, especially in countries which lack good agricultural data. “For instance, we’re collaborating with the Kenyan government on corn production challenges, analysing grid-level data to understand climate impacts and crop trends, addressing food security concerns comprehensively,” he added. Akṣara – Cropin Micro Agri Model Earlier this year, Cropin announced the launch of ‘akṣara’, the sector’s first purpose-built open-source micro language model (µ-LM) for climate-smart agriculture. Built on top of Mistral’s foundational models, the startup aims to make agri-knowledge accessible to everyone in the ecosystem. “We trained it with data for five countries and nine crops and now we are embedding the model with Sage,” Kumar revealed. While Sage is a data-intensive platform, akṣara is a knowledge platform, and according to Jalan, they are complementary to each other. “We have complete clarity on how they can be used independently and also in combination to provide users with knowledge and data,” he said. Sage is Multilingual; Voice Capabilities Coming Soon Sage is currently available in English but does a decent job when prompted in Hindi as well. The startup revealed that more languages could be added going forward. Moreover, at Google I\/O, the tech giant unveiled Project Astra, a first-of-its-kind initiative to develop universal AI agents capable of perceiving, reasoning, and conversing in real-time. Project Astra is built on Gemini, and according to Kumar, the same voice capabilities would come to Sage as well. “The adoption will rely heavily on voice interaction in natural language across different regions worldwide. You simply ask a question in your preferred language. This will be really beneficial for those less adept at typing or with limited literacy in agriculture,” Kumar adds. However, he also emphasises that the startup will take action when the time is right, as it depends largely on their capacity to manage this transition effectively.","excerpt":"Cropin Sage enables users to make informed decisions based on historical, present, and future data on cultivation practices, crops, irrigation, climate, and soil.","categories":["Global Tech"],"tags":["CropIn"],"author_name":"Pritam Bordoloi","publish_date":"2024-07-17T11:15:12","publication_year":"2024","word_count":1029,"keywords":["knowledge graphs","TPU","AI","ML","RAG","Aim","generative AI","CropIn","SQL","R","kubernetes"],"extracted_tech_keywords":["AI","ML","generative AI","Aim","RAG","knowledge graphs","kubernetes","TPU","R","SQL"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/google-backed-cropin-launches-real-time-agri-intelligence-platform-sage\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10123608,"title":"Is Kling a Big Slap to AI Ethicists?","content":"China’s short-video platform Kuaishou Technology recently launched its text-to-video model called Kling, joining other Chinese tech companies in challenging Western tech giants. The model created quite a storm with users finding Kling to outperform OpenAI’s Sora. While the US is debating heavily on AI ethics and incorporating ‘Responsible AI’, China seems unperturbed and is probably responding to these AI ethicists with Kling. Ethics in the Way of Innovation? “When innovation happens and technology explodes, this explosion creates numerous benefits but also raises questions. If we don’t allow this explosion to occur, there will be no innovation,” said TV Mohandas Pai, chairman at Aarin Capital, at IGIC 2024. Referencing UPI and its data security concerns, Pai spoke about how the government has access to our data. “Some may argue that the data is at risk, but frankly, I don’t care if the government sees my data. I want the data to be used productively, and as a sophisticated individual, I want my data protected by the government,” he said. Pai also highlighted how concerns about government surveillance are secondary considering the vast amount of personal data already being accessible to Google, Facebook, and the NSA on US servers. China’s Approach to AI Development Source: Reddit Despite pressing concerns about the ethical and privacy implications of AI, China remains a central player on the global stage. In October 2023, Chinese tech giant Baidu unveiled Ernie 4.0, the latest version of its generative AI model, claiming capabilities comparable to OpenAI’s GPT-4. This significant advancement highlights China’s rapid progress in AI, driven by substantial government support and strategic initiatives like the ‘New Generation Artificial Development Plan’. Launched in 2017, this plan aims to position China as a global AI leader by 2023, emphasising research funding, talent recruitment, and infrastructure development. These efforts underscore China’s commitment to dominating the AI sector, often pushing the boundaries of ethical and privacy considerations. China’s AI development is also fueled by leading companies including Tencent, Alibaba, Baidu, and SenseTime, which attract top group talent and drive innovation. Meanwhile, China’s strides in hardware and robotics, with companies like Dreame Technology and Fourier Intelligence at the forefront, reflect a comprehensive approach to AI applications. Additionally, China’s advancements in facial recognition technology, widely deployed in public spaces, illustrate the country’s capability to implement AI solutions on a large scale. These often move rapidly as ethical concerns are not always given paramount importance. West Focuses on Ethics Over Progress? Recently, the OpenAI Board formed a safety and security committee. It will be responsible for making recommendations on critical safety and security decisions for all the projects of OpenAI. This move came amid ongoing debates about AI safety at the San Francisco startup, heightened by the resignations of researcher Jan Leike and co-founder Ilya Sutskever, who criticised the company for prioritising product development over safety. OpenAI's voluntary AI safety committee pic.twitter.com\/THERDdTC59— Adam Chalmers (@adam_chal) May 30, 2024 Further, Elizabeth Adams, chief AI ethics advisor at Paravison, stated, “We must hold AI developers accountable for their systems’ performance and impacts. To start, let’s educate ourselves on AI’s societal effects through online courses, webinars, and informational videos. Engaging in social media and community forums can raise awareness about ethical AI.” What’s Next? China is only two years behind the US in AI development and a formidable player in the global race among developed countries. Google’s former CEO Eric Schmidt emphasised that while China is focused on dominating several industries, the US still holds a significant lead in AI. Speaking at the AI Expo for National Competitiveness, Schmidt remarked, “In the case of AI, we are probably two or three years ahead of China, which in my world is an eternity.” It’s clear that new AI advancements might impact ethical standards in one way or another, however, looking at China’s route of development, it is clear that compromising on ethics and privacy is acceptable for progress.","excerpt":"“Frankly, I don’t care if the government sees my data. I want the data to be used productively,” says Mohandas Pai.","categories":["AI Features"],"tags":["AI (Artificial Intelligence)","AI Tool"],"author_name":"Vidyashree Srinivas","publish_date":"2024-06-14T10:00:00","publication_year":"2024","word_count":649,"keywords":["Go","API","OpenAI","AI","AI ethics","GPT","Aim","generative AI","AI Tool","AI safety","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","generative AI","OpenAI","Aim","R","Go","API","GPT","AI safety","AI ethics"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/is-kling-a-big-slap-to-ai-ethicists\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10001741,"title":"How Wipro Is Channeling India’s Tech Talent To Push Innovation","content":"Throughout the years, India has fervently made its way towards the development and advancement in the field of science and technology. In fact, according to the last year’s KPMG report, India is ranked third in “Top Global Tech Innovation Leader”. The US tops the list and China is in the second spot. And this is nothing less than a sheer success for India. Talking about the tech innovations in India that is snowballing, Government is not the only body that is working towards it, there are big players, mid-sized firms, startups, institutions and even individuals. And with all the alliances that are forming, there is no denying that the transition journey towards becoming a leading innovation hub in the world is almost a reality now. India has a tremendous pool of young, dynamic and skilled talents, and soon the nation is prophesied to become a focal point to generate more than ever innovative ideas in the coming years. The Wipro Initiatives Looking at the centres that are preparing young talents for the upcoming wave of technology, India is not leaving a stone unturned. One of the most sought after technology, IoT has marked its presence in the nation over the past couple of years and there are companies that are working relentlessly to push out more solutions to the world. And to be a part of this significant transformation, Indian technology giant Wipro is taking vital steps. Recently, Wipro announced that it is launching of its Industrial Internet of Things (IIoT) Centre of Excellence (CoE) in Kochi, Kerala. The company has also stated that this initiative has been taken in order to develop state-of-the-art IIoT solutions for its customers in the industrial manufacturing, automotive, healthcare and pharmaceutical, consumer products and goods, and utility space. This all-new IIoT CoE will be using the sought after technologies such as artificial intelligence, blockchain and robotics, and develop Proof of Concepts (POCs) and market-ready IoT solutions. Furthermore, Wipro has also taken steps that would benefit students pursuing their engineering. On March 12 and 13, the Indian tech giant hosted a two-day hackathon that witnessed participation from some of the leading engineering colleges in the region of Kochi. IIoT is not the only space where the Indian tech giant is hustling hard, artificial intelligence and 5G is also there in Wipro’s checklist. The firm has recently formed an alliance with Indian Institute of Technology Kharagpur (IITKGP) to carry out applied research on 5G and artificial intelligence. The motive behind this partnership is to use the research outcomes by Wipro to develop solutions for its customers, across several verticals. Talking about the advantage that IIT-KGP will have is that the commercialization of the joint research insights and Wipro’s real-world industry expertise. According to the Indian tech giant, the research will focus on design, planning and operations of 5G networks and cognitive information processing. Over the years, Wipro has worked on many projects with other bodies of the nation and have always pushed the nation’s technological advancement. With this new initiative taking place, the company is definitely going all-out and take India’s technological ecosystem to the next phase. Outlook India technology base is one of the strongest in the world and it is spread across the country in forms of institutions, labs, research centres etc. and those who would make the correct use of these bases will definitely be benefited. While there are organisations who are trying to form alliances and work together, there are companies too who are in a race to move past its rivals. If we look through a perspective where companies instead of fighting and competing for each, join hands and innovate, wouldn’t it be great? It would not only benefit the nation but would also benefit the young and dynamic technology enthusiasts of India.","excerpt":"Throughout the years, India has fervently made its way towards the development and advancement in the field of science and technology. In fact, according to the last year’s KPMG report, India is ranked third in “Top Global Tech Innovation Leader”. The US tops the list and China is in the second spot. And this is […]","categories":["AI News"],"tags":["IIoT","IoT","NITI Aayog","Technology","Wipro"],"author_name":"Harshajit Sarmah","publish_date":"2019-04-03T17:56:47","publication_year":"2019","word_count":632,"keywords":["Wipro","Go","artificial intelligence","startup","programming_languages:R","AI","innovation","RAG","IoT","Technology","ViT","GAN","R","NITI Aayog","IIoT"],"extracted_tech_keywords":["AI","artificial intelligence","RAG","R","Go","GAN","ViT","innovation","startup","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/how-wipro-is-channeling-indias-tech-talent-to-push-innovation\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":12707,"title":"Stelae Technologies- Transforming Unstructured Content using AI","content":"Aruna Schwarz, CEO, Stelae Technologies in a candid chat. Img Src: Stelae Technologies This one startup is transforming unstructured content such as those in PDF, Word, ASCII formats into a structured, searchable and indexable output. So why does Stelae Technologies make it to AIM’s startup of the week? Because it is using artificial intelligence in doing so. The output of the unstructured content as processed by the company’s AI software- Khemeia™, can then be injected into Content Management, Big Data & Analytics and Content Mining solutions. The AI software and its use cases Witnessing a successful journey since its inception, the AI software by Stelae Technologies is currently being implemented at major enterprises and public sector entities across many verticals such as laws and legislation, media\/publishing, financial services, e-commerce, and since 2014, in the aerospace and defense sectors. In the industrial sector Khemeia™ has been enabling the optimization of content by transforming legacy technical content into S1000d, DITA, financial, legal, regulatory and compliance content, into Excel, XML, HTML or any user-defined data structure. It uses the 4 step transformation process to yield desired results. Beginning with Content Analysis and extraction where it detects the content element such as section titles, numbers, header, paragraphs, hyperlinks, tables, graphics etc., it then does the Semantic Tagging. This is followed by Structuring the content by splitting the document into relevant modules, creating the hierarchy, formatting bullet lists etc. to finally converting it into Customer Specific outputs. The perks of KhemeiaTM “KhemeiaTM automates the entire transformation process. It uses AI techniques to extract and semantically tag meta-data, structuring and hierarchically organizing information, generating table of contents and converting them to XML-based outputs – all in real-time”, explained Aruna Schwarz, the CEO of Stelae Technology who leads the team at Stelae. Having an experience of more than 20 years in Product Management, Business Development and Marketing with leading Telecos and Content Management solutions vendors, Aruna’s most recent experience was at Rosebud Technologies – a content management solutions vendor in Paris, France, where as Head of Marketing for Europe, she was responsible for developing and launching a suite of Content Management products. “It provides many advantages such as enabling the deployment of effective search algorithms, reuse of content and ensuring interoperability of heterogeneous documents”, said Pierre Fraisse, CTO, Stelae Technologies, who has worked in the field of document management, search, semantic and XML related technologies for over 25 years. “That’s not all, its biggest advantage lies in it being a time saving process, as it can process 20 pages of documents in 5 minutes, compared to long hours (close to 36 hours) if done manually”, he added. Team Stelae Technologies The Team and Accolades Posing a team with a perfect blend of enterprise software skills gained in large corporations combined with deep technology and innovation, this innovative software company has won multiple awards, thanks to its outstanding technology. Some of the accolades it flaunts are-  IBM Global Entrepreneur winner for India & APAC, Nasscom’s Emerging 10 companies, UKTI’s Global Entrepreneur Program (2015) and London & Partner’s India Emerging 20 in 2016. Other important members of the team are Maria Shaio who is the VP Business Development with more than 20 years of experience in IT & Telecoms Sales and Business Development (hardware, software and services) and Sandeep Raizada,  VP Operations, who comes with an experience of 25 years in managing delivery centres and customer projects across multiple locations in the US, Europe & Asia.","excerpt":"This one startup is transforming unstructured content such as those in PDF, Word, ASCII formats into a structured, searchable and indexable output. So why does Stelae Technologies make it to AIM’s startup of the week? Because it is using artificial intelligence in doing so. The output of the unstructured content as processed by the company’s […]","categories":["AI Startups"],"tags":["Content Analysis by AI","Startups"],"author_name":"Srishti Deoras","publish_date":"2017-02-13T07:41:18","publication_year":"2017","word_count":577,"keywords":["Go","artificial intelligence","TPU","AWS","AI","ML","RAG","Aim","analytics","Startups","Content Analysis by AI","R"],"extracted_tech_keywords":["AI","artificial intelligence","ML","analytics","Aim","RAG","AWS","TPU","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-startups\/stelae-technologies-startup-transforming-unstructured-content-using-ai\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":67924,"title":"What Is AWS Snowcone?","content":"Today, there is a new breed of applications that need enhanced capabilities and performance at the edge. It can be challenging to run such applications in non-data centre environments, or where there is a lack of consistent network connectivity. These places usually do not have space, power, and cooling required for data centre IT equipment. This is where The Snow Family can bring value, including AWS Snowcone, AWS Snowmobile, and AWS Snowball. The devices can provide built-in computing capabilities, which can be utilised to help physically transport up to exabytes of data into and out of AWS. So, businesses can perform data collection, machine learning and processing, and storage in environments with intermittent connectivity (such as manufacturing, industrial, and transportation) or in extremely remote locations (such as military or maritime operations) before shipping it back to AWS. Companies can use the recently launched Snowcone to collect, process, and move data to AWS, either offline by shipping the device to AWS or online by using AWS DataSync. If you compare AWS Snowcone and AWS Snowball, it is evident from the name of the Snow family, and similar to Snowball, it has all the rugged capabilities, security capabilities, and shipping capabilities. The primary difference is, of course, things like size, weight and compute. AWS Snowcone Features Snowcone is an edge processing computing device. Its features include data transfer and storage, just with the size of a tissue box. It has 4Gb of RAM, two virtual CPUs, and 8TB of usable storage. It can also act as an IoT Gateway, has WiFi, Bluetooth and comes preloaded with DataSync. It has USB-C  for power and USB-C for the data port, and you can use the device in very harsh environments with desktops, data centres, vehicles, and in conjunction with drones. According to AWS, it can be used as a rugged device for usage and will work in extreme weather conditions. Since it’s four and a half pounds, it is portable and can be easily taken anywhere. Snowcone includes an E Ink shipping label designed to ensure the device is automatically sent to the desired AWS facility and to aid in tracking. In terms of data security, just like an AWS Snowball, the data is multi-layer encrypted so companies can use it for secure data distribution, and feel secure about data in transit. This feature also makes it useful for a lot of places like public sector, defence, intelligence and disaster relief efforts. Like AWS Snowball, Snowcone has multiple layers of security encryption capabilities. Snowcone encrypts data at rest and in transit utilising keys which are managed by AWS Key Management Service (KMS) and are never stored on the device. How Can Snowcone Be Used For Data Workloads? The data can be collected, processed, and transferred to AWS, and run edge computing workloads that use Amazon EC2 instances. Snowcone is designed for data migration needs up to dozens of terabytes. It can be used in space-constrained environments where Snowball devices don’t fit or maybe too large. Snowcone can run edge computing workloads that use Amazon Elastic Compute Cloud (Amazon EC2) instances, and store data securely. You can utilise Snowcone for data migration, content distribution, edge computing, industrial IoT, healthcare IoT, logistics, and autonomous vehicle use cases. There’s enough computer power for launching EC2 instances and also utilising AWS IoT Greengrass. It’s also great for lightweight analytics on the edge and can act as an IoT hub, machine learning where you are processing data in real-time. It has containers and EC2 instances on it, and therefore, lets people connect to the cloud very easily. The Snowcone device is small, so it can be placed next to machinery in a factory to collect, format, and transport data back to AWS for storage and analysis. Users can also strap this device onto the machine equipment, it would collect data and do inference, and then they can swap it out for a new one and send the old one back so the data can be added to their data lake to update all of the AI models. One of the best-touted use cases is for television and movie production. With everything being 4K and higher now, Snowcone might be useful to move those video footage to a cloud location where it can be accessed by multiple people over the cloud platform.","excerpt":"Today, there is a new breed of applications that need enhanced capabilities and performance at the edge. It can be challenging to run such applications in non-data centre environments, or where there is a lack of consistent network connectivity. These places usually do not have space, power, and cooling required for data centre IT equipment. […]","categories":["AI Features"],"tags":["Amazon AWS","AWS","AWS cloud"],"author_name":"Vishal Chawla","publish_date":"2020-06-23T14:00:00","publication_year":"2020","word_count":722,"keywords":["machine learning","AWS","AI","cloud_platforms:AWS","RAG","Amazon AWS","ViT","analytics","edge computing","AWS cloud","R","data lake"],"extracted_tech_keywords":["AI","machine learning","analytics","RAG","AWS","edge computing","R","data lake","ViT","cloud_platforms:AWS"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/what-is-aws-snowcone\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10004359,"title":"COVID-19— A Tailwind For Data Science","content":"Last weekend, while I was welcoming the arrival of the monsoons in Pune with a cup of tea in my balcony, I thought about the changes that COVID-19 has brought into my life.  My home is my new office, the newspaper is going digital, and I have taken a fancy to online streaming services like Amazon Prime, Netflix, or Disney+Hotstar as a staple for leisure. Digitalisation is generating an enormous amount of data, which will prove to be a tailwind for data science adoption across industries in these times. According to OpenVault’s Broadband Insights Report for Q1 2020, digital interactions for both personal and professional use increased sharply. Six months back, some of these aspects of the ‘new normal‘ in the business world would have been unimaginable: a. Digital workspace: Remote working works well for most of the industries, including banking, education, healthcare, etc. b. Digital business processes:  You can hire and onboard employees in your workforce without any face-to-face interaction. You can track the performance of a team member\/team by building digital monitoring processes. c. Digital work-life integration: Digital transformation can happen in a matter of a day or a week. The office space used to be one of the most important spaces for us to socialise. In the pre-COVID-19 world, we would spend months, or even years, to bring about any one of these transformations. Now, digital transformation is happening within a matter of weeks. Sceptical as some may be, most of these changes are here to stay. Now, let’s understand the digital transformations happening in our personal lives. a. Digital learning: Digital schooling takes precedence over the physical learning environment for our children. b. Digital entertainment: OTT (over-the-top) streaming services are gaining popularity c. Social meetups using online platforms and digital wellness\/healthcare:  Wellness coaching and counselling can happen via fitness apps Data science to scale new heights, faster When we throw a ball, it is fairly easy to estimate how much it will bounce first. What is more interesting is to estimate the second, third and subsequent bounces of the ball. Digital transformation is only the first bounce of the ball thrown by COVID-19. There is one common thread in our hyper-digital lives. We have started generating more data. To cite an example, in the pre-COVID-19 world, your transaction for an online movie booking for your family or friends would leave a trail of 5-10 digital transactions. Each weekend, on average now, we are spending several more hours on OTT (over the top media) platforms like Netflix or Amazon Prime. In the initial weeks, we generated hundreds of transactions via clicks on these platforms while browsing for shows or movies. Now, we expect these platforms to understand our consumption preferences and recommend a watchlist to us. The only way to fulfil this need is by using data science models behind the scenes. Increasingly, content would not be the prime differentiator for OTT platforms. The platform that understands my preferences and caters to my need will win my mindshare. Similarly, in our professional lives, collecting data about time spent by employees in meetings would have been a laborious task in the pre-COVID-19 world. Today, with a surge in remote working and online meetings, it is easy to review the backend data from office collaborations tools like MS Teams. The moment the data is readily available and reliable, the journey for data science adoption starts. Processes such as sales, which used to happen in the physical world, typically generated bits and pieces of information in an asynchronous manner. Now, most of the interactions with potential prospects are happening via a digital medium like email, MS Teams call or Zoom Video conference. As more data points become available from such interactions, it will be easier to predict the chances of closing a deal using data science. In the first leg of digitalisation post-COVID-19, only a few businesses could quickly adapt to the digital transformation compared to others. Organisations will have to make concerted efforts to get ready for the second leg of data science adoption. It will involve three to six months of programmes to be transformation ready. Only generating data via digital transactions is not enough. We need meaningful and reliable data, which requires proactive effort. The formation of a separate data science task force could be the first step in this preparation, followed by access to data science talent that understands the business processes.","excerpt":"Last weekend, while I was welcoming the arrival of the monsoons in Pune with a cup of tea in my balcony, I thought about the changes that COVID-19 has brought into my life.  My home is my new office, the newspaper is going digital, and I have taken a fancy to online streaming services like […]","categories":["AI Features"],"tags":["back office data"],"author_name":"Abhijit Joshi","publish_date":"2020-08-07T16:00:00","publication_year":"2020","word_count":736,"keywords":["data science","Go","programming_languages:R","AI","digital transformation","programming_languages:Go","Git","RAG","GAN","back office data","R"],"extracted_tech_keywords":["AI","data science","RAG","R","Go","Git","GAN","digital transformation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/covid-19-a-tailwind-for-data-science\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10135535,"title":"10 Must-Try OpenAI o1 Use Cases","content":"OpenAI released its new o1-preview series of AI models, designed to spend more time thinking before they respond. These models can reason through complex tasks and solve harder problems than previous models in science, coding and math. In initial tests, the updated reasoning model performed on par with PhD students. In a qualifying exam for the International Mathematics Olympiad, the model scored 83%, compared to GPT-4o’s 13%. After the new update’s preview models were released, users took to the internet to share their innovative projects using o1. Ten such projects that stood out are as follows. Develop Games on the Go! Karina Nguyen, a user made an AISteroid game with retro sci\/fi vibe. o1 is really good at making fun small games! for example, i made AISteroid Game w\/ retro scifi vibes :) https:\/\/t.co\/cUZQaVnCEi pic.twitter.com\/ZOMpCVkPIj— Karina Nguyen (@karinanguyen_) September 12, 2024 Another user named Akhaliq combined o1 with Replit and Gardio to build a chess game. The users were able to code and build games using the o1 module despite the previous GPT modules being not equipped to do so. Subham Saboo also created a space shooter game which he then ran on Reptile claiming o1 has changed coding and AI forever. OpenAI o1 creates a fully interactive space shooter game in less than 2 minutes and Replit lets me run it in seconds.AI and coding has changed forever. pic.twitter.com\/toRmzaHock— Shubham Saboo (@Saboo_Shubham_) September 13, 2024 Create Your Own App A user, Ammar Reshi combined o1 with Cursor Composer and built an ios weather predicting app, with accurate predictions from scratch in 10 minutes with animation features. This module computed the coding and UI generating a response that tailor made the app from scratch. Just combined @OpenAI o1 and Cursor Composer to create an iOS app in under 10 mins!o1 mini kicks off the project (o1 was taking too long to think), then switch to o1 to finish off the details.And boom—full Weather app for iOS with animations, in under 10 🌤️Video sped up! pic.twitter.com\/hc9SCZ52Ti— Ammaar Reshi (@ammaar) September 12, 2024 Economics Essay Slay Tyler Cowen, researcher at OpenAI asked o1 to write a college essay and unlike any other previous GPT module, OpenAI o1 responded with ease generating an in depth answer for the given prompt. Watch “Tyler Cowen: OpenAI o1 & Economics” on #Vimeo https:\/\/t.co\/dgaIwLIjoW— tylercowen (@tylercowen) September 12, 2024 Genetics Solving Catherine Brownstein, another researcher tested o1 to help her reason through “n of 1” cases;  medical cases that nobody has ever seen and o1 was able to step up to the occasion and assist with the cases. o1 was able to understand complex genetic related queries and even solve equations for it generating positive answers. 7. Genetics pic.twitter.com\/6iX3jo6ZDT— Min Choi (@minchoi) September 12, 2024 Quantum Physics Computing Mario Krenn used o1 to draft and reason through complex quantum physics equations, o1 responded better than any other version of GPT module generating quotations that fit the case. It decoded the problem, generated equations and solved them too. This module solved equations that renowned academics require brain power to do so and proved its competence to other GPT models. 2. Quantum Physics pic.twitter.com\/4DGCiwGvOW— Min Choi (@minchoi) September 12, 2024 Solving Puzzles Jerry Tworke, a researcher ran a complex puzzle on the o1 module, trying to guess a person’s age in question to say a person’s relation to another person’s age and o1 took time to process the information and then solved the equation like a math problem revealing the answer with reasoning. 6. Solving complex logic puzzlepic.twitter.com\/YIRy6RBFT0— Min Choi (@minchoi) September 12, 2024 Translations Worldwide Hyung Won Chung tried translating a korean script on the previous GPT module and it responded with no answer as it interpreted the language as gibberish and wasn’t able to understand the prompt. But with o1 it not only understood the prompt but also identified the language, decoded the text, enhanced translation by computing the meaning of the text and delivered impressive results. 8. Translates corrupted sentencespic.twitter.com\/5x1m0BfOsB— Min Choi (@minchoi) September 12, 2024 Reasoning Out Data Jerry Tworke also tested the model for reasoning, typing a prompt based on physics and the physical relationship of objects, he concluded that humans would be able to logically reason out the outcome but LLM models aren’t trained to do so before, however, o1 after thinking the outcome stated the step by step process and answered positively. 4. Reasoning pic.twitter.com\/9JZKGgUPRL— Min Choi (@minchoi) September 12, 2024 Solving Complex Math Equations Mason Meyer, a researcher used o1 module to generate a 5×5 nonogram, a pretty complex task for a mathematician or a LLM based AI program, but o1 after procession the question did generate the nanogram and prepared a grid layout for it, giving Meyer a puzzle that Meyer later asked o1 to solve for and it succeeded in doing so. 9. Math pic.twitter.com\/KlQCoFjnaR— Min Choi (@minchoi) September 12, 2024 Coding For Beginners It takes a lot of effort to run code that consistently runs well and o1 can potentially make it easier for every layman to code with its new upgrade. Scott Wu, a programmer tested o1 on human reasoning and code building and reported the results were positive. 10. Coding pic.twitter.com\/6KaC1RWrVb— Min Choi (@minchoi) September 12, 2024","excerpt":"OpenAI o1 is here to make academia revolutionary from solving difficult quantum physics equations to genetics related query, among others.","categories":["AI Trends"],"tags":["AI Tool","OpenAI"],"author_name":"Tanisha Bhattacharjee","publish_date":"2024-09-16T15:45:07","publication_year":"2024","word_count":871,"keywords":["Go","OpenAI","AI","programming_languages:R","GPT-4o","programming_languages:Go","GPT","Aim","AI Tool","R","llm_models:GPT"],"extracted_tech_keywords":["AI","GPT-4o","OpenAI","Aim","R","Go","GPT","llm_models:GPT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/10-must-try-openai-o1-use-cases\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":54865,"title":"Startups In The Future Will Be More About AI: Ratan Tata","content":"In the past few months, various stakeholders of the startup ecosystem have publicly spoken about the potential of artificial intelligence (AI) and how it will impact the future of the Indian economy. Now, chairman emeritus of Tata Sons, Ratan Tata, has also said that the startup of the future will be more about AI. Speaking at TiECon Mumbai 2020, Tata said that the leading tech companies had shown the path. “Considering that companies like Amazon and even Google are very AI technology-oriented. So I think the startups of the future will be more about AI.” Tata also advised startups to operate businesses ethically. He said, “When we [Tata] raised money from the international market, we built our case on the stability and maturity of the founders, or the top management.” Interestingly, Tata is also an investor in AI-based chatbot Niki since the company’s seed round. In 2017, the company raised $2 million series A funding from Tata, SAP and Unilazer Ventures. At TiECon Mumbai 2020, Ratan Tata also was awarded the lifetime achievement award for his contribution to the Indian startup ecosystem. The award was presented by Infosys co-founder NR Narayana Murthy. This month, even union minister of commerce and industry, Piyush Goyal, noted that the using AI applications and other emerging technologies in different sectors could help India achieve its goal of becoming a $5 trillion economy by 2024.","excerpt":"In the past few months, various stakeholders of the startup ecosystem have publicly spoken about the potential of artificial intelligence (AI) and how it will impact the future of the Indian economy. Now, chairman emeritus of Tata Sons, Ratan Tata, has also said that the startup of the future will be more about AI. Speaking […]","categories":["AI News"],"tags":["AI (Artificial Intelligence)","Artificial Intelligence India"],"author_name":"Sejuti Das","publish_date":"2020-01-29T18:25:55","publication_year":"2020","word_count":230,"keywords":["Go","funding","artificial intelligence","programming_languages:R","AI","Artificial Intelligence India","programming_languages:Go","Ray","R","AI (Artificial Intelligence)","startup"],"extracted_tech_keywords":["AI","artificial intelligence","Ray","R","Go","startup","funding","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/startups-in-the-future-will-be-more-about-ai-ratan-tata\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":54383,"title":"Introduction To OpenVINO: Learn With Their Face Detection Model","content":"OpenVINO stands for Open Visual Inference and Neural Network Optimisation and is a popular computer vision toolkit provided by Intel. The main reason for its increasing popularity is its usage of the state-of-the-art optimisation techniques for reducing the inference time of computer vision models. It can accelerate a model across devices like CPUs, GPUs, FPGAs, VPUs etc. The OpenVINO toolkit contains a ton of pre-trained models spanning across domains such as face detection, person detection, pose estimation, instance segmentation etc. In this article, I will demonstrate a working example of a face detection model from the OpenVINO toolkit. But before diving into the coding part, let’s understand the basics of OpenVINO. The OpenVINO toolkit contains two major components; Model Optimiser and Inference Engine. Model Optimiser Source: Intel OpenVINO Model Optimiser is used to convert a deep learning model into an intermediate representation (IR). OpenVINO supports various frameworks like TensorFlow, Caffe, ONNX, MXNet etc. So in addition to all the pre-trained models that are available with the toolkit have also been converted into our deep learning model written in any of these frameworks into an intermediate representation. The inference engine of OpenVINO only understands this IR format. It can’t work with native framework files. Now let’s dive deep into the working of the Model Optimiser. It handles the software level optimisations of the deep learning model. Few operations that it performs are: Quantisation It is related to the number of bits required to represent the weights and biases of our model. We usually train our models using the FP32 (floating point 32) format, it is ideal for training the model, but we don’t need that level of precision for inference. So, we can reduce the precision of our models without any substantial accuracy loss. The quantisation is carried out with the help of Calibrate Layer. Initially, we set a certain threshold value for accuracy drop, and the Calibrate layer takes a subset of data and converts the FP32 layer into FP16 or INT8. If the accuracy drop is less than the specified threshold value, then the conversion is carried out. Fusion Fusion relates to combining multiple-layer operations into a single operation. For example, a batch normalisation layer, an activation layer, and a convolutional layer could be combined into a single operation. This can be particularly useful for GPU inference, where the separate operations may occur on separate GPU kernels. At the same time, a fused operation occurs on one kernel, thereby incurring less overhead in switching from one kernel to the next. Inference Engine Source: Intel OpenVINO It is a library written in C++. It provides APIs to read the intermediate representation of the model. It handles the hardware level optimisation. It provides different plugins for different devices. Another important component of the Inference Engine is ‘Extensions’.Extensions are used for extending the compatibility of layers i.e. if all the layers of a model are converted into an IR format, then also it is not necessary that all the layers will be compatible with the device. So suppose we want to run a piece of optimised code on a CPU but one of the layers is not supported on CPU by default, then CPU extensions can help us by extending the compatibility.There are various types of extensions for different devices and different operating systems. In this example, we will use  libcpu_extension_sse4.so In Linux, we have two types of extensions AVX and SSE4. AVX works with Intel Core series processors (ex – Core i5, Core i7 etc.), AVX systems can also use SSE4 extensions.SSE4 works with Intel Atom series processors. Face Detection Program in Python Source: Intel OpenVINOWe will start off by downloading the face detection model. We can either use the model optimiser to download the model or we can simply search for a model and download it directly. Method 1: Using the model optimiser Go to the model_optimiser directory, usually located at – \/opt\/intel\/openvino\/deployment_tools\/tools\/model_downloaderand check for the downloader.py file.Run that file by passing in the name of the model as a command-line argument. sudo python downloader.py  –name face-detection-adas-0001 Use the python downloader.py –print_all command to check all the available models. Method 2: Using the direct link The link to download the face detection is – https:\/\/download.01.org\/opencv\/2019\/open_model_zoo\/R1\/models_bin\/face-detection-adas-0001\/FP32\/ You can check out all the available pre-trained models here – https:\/\/software.intel.com\/openvino-toolkit\/documentation\/pretrained-models The complete code for the face detection program can be found here – https:\/\/github.com\/Dhairya10\/face-detection-open-vino The two most important classes that we will interact with are – IECore, which is the Python wrapper to work with the Inference EngineIENetwork, which is what will initially hold the network and get loaded into IECore Inference Engine Python API Documentation – https:\/\/docs.openvinotoolkit.org\/latest\/_inference_engine_ie_bridges_python_docs_api_overview.html Using Neural Computer Stick 2 (Movidius) Now, we will see how to run this code using Neural Compute Stick 2.We will have to change the device name to ‘MYRIAD’. (Change line 15 in main.py file. Set open_vino_device = ’MYRIAD’)In order to set up the NCS. Follow the steps below. Getting Started with NCS2 https:\/\/software.intel.com\/en-us\/articles\/get-started-with-neural-compute-stick?elq_cid=3708810&erpm_id=6960228 Configuring NCS2 https:\/\/docs.openvinotoolkit.org\/latest\/_docs_install_guides_installing_openvino_linux.html Conclusion Computer vision applications are ubiquitous, and by using the OpenVINO toolkit, developers can create a more robust and scalable application. From self-driving cars to using face unlock feature on our mobile, computer vision applications are deeply embedded in our day to day lives and these applications can be optimised even further by using the OpenVINO toolkit. So any developer working in the Computer Vision space should give it a shot.","excerpt":"OpenVINO stands for Open Visual Inference and Neural Network Optimisation and is a popular computer vision toolkit provided by Intel. The main reason for its increasing popularity is its usage of the state-of-the-art optimisation techniques for reducing the inference time of computer vision models. It can accelerate a model across devices like CPUs, GPUs, FPGAs, […]","categories":["Global Tech"],"tags":["Intel"],"author_name":"Dhairya Kumar","publish_date":"2020-01-19T10:00:00","publication_year":"2020","word_count":903,"keywords":["Go","AI","neural network","ML","computer vision","OpenCV","Python","deep learning","TensorFlow","R","Intel"],"extracted_tech_keywords":["AI","ML","deep learning","neural network","computer vision","TensorFlow","OpenCV","Python","R","Go"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/introduction-to-openvino-learn-with-their-face-detection-model\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":31240,"title":"5 Most Sought After Programming Languages For Robotics You Should Learn In 2019","content":"The field of robotics encompasses computer science, electronics, electrical, mechatronics, machine and deep learning, instrumentation and AI. The field which involves various subjects and applications has a steep learning and experimenting curve. It also has immense opportunities that come with a lot of dedication. To build a career in robotics, one requires a specialisation in the above fields along with a set of robust programming language. In this article, we list down top five programming languages that are in high demand for the robotics industry. Lisp Lisp is the world’s second earliest programming language. Not as popular as other programming languages, it is crucial in the AI domain. Some of the crucial sections of ROS(robot operating system) are written in Lisp. It was originally developed for the applied mathematical notation that can be implemented for computer programs, this feature made it a favoured language for artificial intelligence related research. The language is used to develop robot control functions in a microcomputer category which helps in the manipulation of various robot functions. The additional features of Lisp are tree data structures, automatic storage management, dynamic typing, conditionals, recursion, elevated-order functions, read-eval-print loop, and the self-hosting compiler. Sources To Learn: https:\/\/www.tutorialspoint.com\/lisp\/ (Beginner Friendly) BASIC And PASCAL Belonging to the programming language genesis, BASIC and Pascal are the fundamental scripts for multiple industrial robot languages. Beginners All-Purpose Symbolic Instruction Code(BASIC) was initially developed for the trainees and one of the extensively used initial programming languages, even today a few micro-monitors like Basic Micro, BasicX, Parallax use it for training robots. A programming language that is credited for instigating the constructs into programming, Pascal is fetching language to upgrade from BASIC. The language demands more coding and needs commendable programming practices. These classic programming languages may be outmoded for current trends, but they are really handy for beginners to get close to technical operators and functionality. Sources To Learn Pascal and BASIC: TutorialPoint, udemy (Beginner friendly) C(objective) And C++ Being part of almost every engineering course, C(obj) and C++ is the most mature programming languages that are a must-have on the resume in the robot industry.  The reason why C and C++ is one of the most popular languages in robotics is that it because allows interaction with low-level hardware. Aspiring roboticists should start with C\/C++ as it contains more tools and library functions. With abundant tools, libraries and functions C++ takes a significant position in robotics programming platform. Most programmers use C\/C++ to ensure the best performance of robot. Sources To Learn C(obj) and C++: https:\/\/www.learncpp.com\/ https:\/\/www.udemy.com\/c-and-c-programming\/ (Beginner friendly) Hardware Description Languages (H.D.Ls) This specialised computer language is used to program electronic and digital logic circuits. Through HDL, the structure, operations and design of the circuits is programmable. Hardware Description Language is used to specify the gates and chips and simulate the behaviour of the resulting chip specifications through a hardware simulator.  There are many HDLs available, but one of the most popular HDLs are Verilog, VHDL and SystemC, which is also the latest on the block. Essentially, what HDLs have done is blur the distinction between software and hardware development with design tools which enables the development of real-time imaging hardware. Unlike the other mentioned languages HDLs need a good knowledge of electronics as most of the operations are concurrent processor based. HDLs provide object orientation and garbage accumulation which in a simple way as compared to complex operations and semantics compared to traditional languages which operate on control flow as opposed to the data flow. Sources To Learn HDLs: https:\/\/www.udemy.com\/courses\/search\/?q=verilog%20hdl&src=sac&kw=hdl, Advanced Level: Before learning HDLs, you have to beef up on basics of electronics and chip design Java And Python Java contains all the high-level features required in the robotics industry to deal with various aspects of robotics. Most companies look for Java programming skills in IT professionals as it can be used to create algorithms for search, ML, neural algorithms and language processing. Majority of the institutes and universities offering Robotics courses and certifications make sure that Java is included in their curriculum. As compared to other languages, Python has soared in popularity thanks to its ease of use and less time required. For example, when compared to other languages such as Java or C\/C++, one has to write fewer lines of code in Python. In robotics, Python has become a key part of robot operating system and is used for designing the embedded systems. One the reasons for this scenario is that Python is a high-level language of programming, which makes it have less code as compared to the existing object-oriented languages. Tech giants like Raspberry Pi, Arduino among the list use Python in designing embedded systems and exhaustive automation packages for its withhold from typical tasks of programming. Python also grants basic ties with C\/C++ scripts which help in fixing the execution Sources To Learn Java and Python: https:\/\/www.udemy.com\/javapython\/ (Beginner friendly)","excerpt":"The field of robotics encompasses computer science, electronics, electrical, mechatronics, machine and deep learning, instrumentation and AI. The field which involves various subjects and applications has a steep learning and experimenting curve. It also has immense opportunities that come with a lot of dedication. To build a career in robotics, one requires a specialisation in […]","categories":["AI Trends"],"tags":["arduino","Python Programming","Robotics","robotics technology"],"author_name":"Bharat Adibhatla","publish_date":"2018-12-09T06:17:25","publication_year":"2018","word_count":813,"keywords":["Go","artificial intelligence","robotics technology","AI","R","ML","Robotics","RAG","Python","C++","deep learning","Python Programming","arduino","Java"],"extracted_tech_keywords":["AI","artificial intelligence","ML","deep learning","RAG","Python","R","Go","Java","C++"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/5-most-sought-after-programming-languages-for-robotics-you-should-learn-in-2019\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10093953,"title":"Generative AI is Having An Edison Moment","content":"The dangers of not understanding innovation money are best illustrated by the battle between two inventors: Thomas Edison and Nikola Tesla. The former produced inventions largely through trial and error yet capitalised it. Tesla’s ideas arguably were brilliant. The visionary was even described by Edison as someone whose “ideas are magnificent.” But he was simply unable to attract the right financial resources to commercialise his ideas. Currently,generative AI models are having an Edison moment. The companies making the models accessible though products and services are thriving more than the research which can be comparatively more impactful. Tech bros who don’t want to miss the opportunity; currently implementing oversold technologies should remember some fundamental realities about tech bubbles. These phenomena’s shelf life depends upon the narratives about how the specific technology will affect societies and economies, as professors Brent Goldfarb and David Kirsch noted in their 2019 book, Bubbles and Crashes: The Boom and Bust of Technological Innovation. Unfortunately, the early narratives that emerge around new technologies commonly fail to meet expectations. GenAI is not AI Companies are running after the text-to-anything as we’ve noticed during  their latest annual conferences; Google I\/O, Microsoft Build, and IBM Think. The companies have been in an aggressive tussle since the release of ChatGPT, in late November last year. The developer’s who showed off their latest technologies have been working 24\/7, and really hard, under pressure to deliver the tech, and the same was resonated by some of the speakers at these conferences. Google chief Sundar Pichai kicked off the Google I\/O 2023 with “AI is having a very busy year” and then announced a list of Google’s products which will henceforth be integrated with its language model PaLM 2. Similarly during Microsoft Build, the company chief Satya Nadella chanted the ‘Copilot’ mala. Even though the investors were impressed with AI being embedded in every facet of the Redmond giant. However, the shares fell fractionally. Earlier, in February’s Paris debacle when Google promoted its AI chatbot Bard, the company’s stock sank 7%, Google employees responded by describing Pichai’s announcement as “rushed” and “botched.” Meanwhile, other brilliant innovations are turned blind eye towards or not given enough appreciation. Out of the 100 announcements at the I\/O, around 40 were about generative AI; and there was hardly any new update related to AlphaFold, which is revolutionising the life sciences landscape. Other jaw dropping feats which have been turned a blind eye towards due to the distraction caused by language models include, DragGan, Meta’s CICERO. DeepMind’s nuclear fusion controlling algorithm and the list goes on. “Looking back, it’s amazing how easy things were for researchers when I was a young man. In comparison to just how competitive the field has become,” the 80-year-old American computer scientist Jeffrey Ullman told AIM, saying that academics and researchers who could be excellent teachers, or focus on other groundbreaking innovation, are forced into doing second and third grade research because that is how they get recognised or promoted. The Shiny Object Syndrome The AI wave highlights the persuasive issue known as the shiny object syndrome; common in the tech industry as researchers get easily distracted by novel tools and trends. In an edition of  ‘Letters by Andrew Ng,’ the founder of DeepLearning.AI stated that AI has an Instagram problem. “I’m here to say: Judge your projects according to your standard, and don’t let the shiny objects make you doubt the worth of your work!” he declared. He further addressed people doubting their work’s worth and judging it as per perfect standards set by the media. He wrote, ‘Just as pictures of people’s perfect lives in the media aren’t representative, pictures of AI developers’ postings of their amazing projects also aren’t representative.’ On a similar note, Andrew Ng’s mentor Michael Irwin Jordan told AIM that most of these are buzzwords. “Just because you’re using computer vision or ChatGPT as some part of that doesn’t necessarily change anything,” he said. Apart from the false promise of technology affecting research it also takes a toll on the global economy. The Bank of America strategist Michael Hartnett recently noted that tech and AI are forming a risky bubble. Experts are also predicting the Silicon Valley’s darling GPT bubble to cause a meltdown similar to the major dot.com bubble which led to a stock market crash in early 2000. The bottom line is, the internet’s favourite language models are definitely entertaining for the public eye and the industry’s cash cow but the technology is being oversold which is anticipated to disrupt the purpose of research. Amid the chaos in tech, one should recall when Tesla pinpointed impatience as researchers’ problem. Highlighting the eagerness of their ideas to work he had said, “They want to try their first idea right off; and the result is they use up lots of money and lots of good material, only to find eventually that they are working in the wrong direction. We all make mistakes, and it is better to make them before we begin.”","excerpt":"Companies making generative models accessible are thriving more than a more impactful research","categories":["AI Features"],"tags":["AI GPT","ChatGPT","GPT","language model"],"author_name":"Tasmia Ansari","publish_date":"2023-05-25T18:30:00","publication_year":"2023","word_count":834,"keywords":["Go","ChatGPT","GenAI","API","AI","computer vision","RAG","GPT","language model","Aim","generative AI","AI GPT","R"],"extracted_tech_keywords":["AI","computer vision","generative AI","GenAI","ChatGPT","Aim","RAG","R","Go","API"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/generative-ai-is-having-an-edison-moment\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10100927,"title":"Microsoft is Trying Hard to Give LLMs a Moral Compass","content":"AI models use deep learning to talk like humans but they lack the ability to make morally sound decisions. In the last couple of months there has been a never-ending see-saw of whether AI will lead us to utopia or lead us to a moral ruin. While the industry is seeing the developments from an optimistic point of view, AI insiders have been raising red flags across the world – including Sam Altman, the OpenAI bossman, who has openly spoken of how it could be used for disinformation and offensive cyberattacks. As these models are being deployed in high stake environments like healthcare and education evaluating whether LLMs can make morally sound judgments or not. Researchers from Microsoft have proposed a new framework to probe the moral reasoning abilities of prominent LLMs. The research specifically, it pointed out that large models such as GPT-3 exhibited shortcomings in understanding prompts, resulting in moral reasoning scores closely resembling random chance. In contrast, smaller models like ChatGPT, Text-davinci-003, and GPT-4 showcased a higher degree of coherence in their moral reasoning capabilities. Interestingly, the more compact 70B LlamaChat model surpassed its larger counterparts demonstrating advanced ethics understanding is possible without massive parameters. These models primarily functioned mostly at intermediate conventional levels aligning with Kohlberg’s moral development theory. It’s worth noting that none of these models showed a highly developed level of moral reasoning. The paper provides novel insights into the ethical capabilities of LLMs and a guide to move ahead in research. Using a psychological assessment tool called the Defining Issues Test (DIT) they evaluated the moral reasoning capabilities of the six stars of the moments — GPT-3, GPT-3.5, GPT-4, ChatGPT v1, ChatGPT v2, and LLamaChat-70B. The test presents moral dilemmas and has subjects rate and rank the importance of various ethical considerations and allows quantifying the sophistication of moral thinking through a P-score (Post Conventional Morality Score). Premature to Trust The tech pundits have sufficiently wowed that they foresee a future iteration of an AI chatbot challenging the supremacy of existing technologies and do all sorts of other once primarily human labour. While better models are being developed on a daily basis, there is not much research being done on how much these models can be trusted. Earlier this year, in a paper titled, “The moral authority of ChatGPT,” Sebastian Krügel, Matthias Uhl and Andreas Ostermaier showed that ChatGPT gives conflicting advice for moral problems like the ethical trolley problem: the switch dilemma and the bridge dilemma. The trio asserted that ChatGPT appears to lack a consistent moral compass. Similarly, researchers at Microsoft arrived at a similar conclusion, summarising that the AI models in question displayed a modest level of moral intelligence. While they demonstrated the capacity to transcend basic self-interest, these models struggled when confronted with ethical dilemmas and nuanced trade-offs—challenges that morally developed humans typically navigate with greater finesse. Not Black and White The LLM landscape is developing at a break-neck pace yet the models have limitations that remain unaddressed. Morality aside, countless studies have documented their tendency to reinforce the gender, ethnic, and religious stereotypes explicitly within the data sets on which they’re trained. “People often think the machine learning algorithms introduce bias. Fifty years ago, everybody knew ‘garbage in garbage out’. In this particular case, it is ‘bias in, bias out’,“ a veteran data scientist and Turing Award laureate Jeffrey Ullman told AIM. Back in 2019, Eric Schmidt, the former chief of both Google and Alphabet, outlined a forward-looking vision. He described a future where AI-powered assistants would play pivotal roles in helping children with language and maths learning, assisting adults in daily planning, and being companions to the elderly. He astutely noted that if these AI models lacked a moral compass, their influence would be harmful. Among shortcomings like these, the Microsoft research matters because rather than just binary right\/wrong judgments, the test used in the study provides spectrum-based insights into the sophistication of moral reasoning for building less potentially harmful models.","excerpt":"Microsoft researchers have proposed a framework to probe moral reasoning of famous LLMs","categories":["AI Features"],"tags":[],"author_name":"Tasmia Ansari","publish_date":"2023-10-01T10:00:00","publication_year":"2023","word_count":664,"keywords":["Go","ChatGPT","machine learning","OpenAI","AI","GPT","Aim","deep learning","Rust","R"],"extracted_tech_keywords":["AI","machine learning","deep learning","ChatGPT","OpenAI","Aim","R","Go","Rust","GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/microsoft-is-trying-hard-to-give-llms-a-moral-compass\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10164700,"title":"Grok 3 vs Claude 3.7 Sonnet vs o3-mini vs Gemini 2.0","content":"The LLM battle in 2025 is off to a strong start, with frontier models already vying for dominance. Elon Musk’s xAI recently launched Grok 3, which has reportedly impressed users worldwide. Meanwhile, Anthropic introduced Claude 3.7 Sonnet, and earlier this year, OpenAI launched o3-mini with plans to release GPT-4.5 soon. Google has expanded its Gemini 2.0 lineup with the introduction of Gemini 2.0 Flash and Gemini 2.0 Pro models. Coding and Reasoning With a 70.3% score on SWE-bench Verified, Claude 3.7 Sonnet outperforms o3-mini, which scores 49.3%, making it a strong choice for coding. Meanwhile, Grok 3 is also gaining recognition as a competitive coding model. In a blog post, xAI stated that on LiveCodeBench (v5), Grok 3 mini beta (Think) scored 80.4, while o3-mini scored 74.1. The reasoning models are available through the Grok app. Users can prompt Grok 3 to ‘Think’ or, for more complex inquiries, activate ‘Big Brain’ mode, which uses extra computational power for deeper reasoning. Besides coding and reasoning, Grok 3 can generate images and supports voice conversation mode. However, access to these features requires a Premium+ or SuperGrok subscription. Claude 3.7 Sonnet is ideal for complex software tasks. Its extended thinking mode enhances math and science capabilities. However, it does not support voice or video processing. The model is a ‘hybrid’, meaning it can simultaneously function as both a standard LLM and a reasoning LLM. In extended thinking mode, the model reviews its reasoning before generating a response, leading to improved performance in math, physics, coding, instruction-following, and other complex tasks. When using Claude 3.7 Sonnet through the API, users can control the number of tokens allocated for reasoning, up to a maximum of 1,28,000 tokens. This allows them to manage the trade-off between response speed, cost, and output quality. The model can accept text and images as input. This means it can process and analyse text-based data and images to generate responses or perform tasks like code generation and problem-solving. However, it lacks image generation capabilities and does not support voice conversation. Similarly, OpenAI’s o3-mini is suitable for competitive programming, coding challenges, and cost-sensitive applications. The company released the model in response to DeepSeek’s R1, an open-source alternative to OpenAI’s o1, which was developed at a fraction of the cost. Unlike Anthropic, where users can set a fixed number of tokens for reasoning, OpenAI provides three reasoning effort levels – low, medium, and high – allowing developers to adjust processing based on their needs. This feature lets o3-mini allocate more processing power for complex problems or prioritise speed when low latency is required. However, o3-mini does not support vision-related tasks, so developers should continue using OpenAI o1 for visual reasoning. Like o1, o3-mini comes with a larger context window of 2,00,000 tokens and a max output of 1,00,000 tokens in the API. Few can match Google Gemini when it comes to multimodality and longer context windows. Gemini 2.0 Flash offers a range of features, including native tool use, a 1 million-token context window, and multimodal input. While it currently supports text output, image and audio output, along with the Multimodal Live API, will be available soon. For coding, Google has introduced Gemini 2 Pro. The tech giant says the model excels at coding capabilities and processing complex prompts with improved comprehension and reasoning. It also features Google’s largest-ever context window of 2 million tokens, allowing for in-depth analysis of extensive information. Availability and Pricing Grok 3 is integrated into X and available for free to all users. However, advanced features like voice mode are exclusive to Premium+ subscribers. Users can interact with Grok 3 directly through the X app or website. X Premium+ is currently available in India for ₹3,470 per month. xAI’s SuperGrok subscription costs $30 per month or $300 per year when purchased through the iOS app. This standalone app provides access to advanced Grok 3 features like DeepSearch and reasoning modes. The company also announced that in the coming weeks, Grok 3 and Grok 3 mini will be available through its API platform, offering access to both standard and reasoning models. Moreover, DeepSearch will be released to Enterprise partners via the API. On the other hand, OpenAI’s o3-mini is available to all free ChatGPT users. The model is accessible through the Chat Completions API, Assistants API, and Batch API for select developers in API usage tiers 3-5. OpenAI’s o3-mini is a small, cost-efficient reasoning model optimised for coding, math, and science. It supports tools and Structured Outputs and offers a context length of 2,00,000 tokens. The model’s pricing is set at $1.10 per million input tokens, with a discounted rate of $0.55 per million cached input tokens. Output tokens are priced at $4.40 per million tokens. Claude 3.7 Sonnet is available across all Claude plans, including Free, Pro, Team, and Enterprise, as well as through Anthropic’s API, Amazon Bedrock, and Google Cloud’s Vertex AI. The model is priced the same as its predecessors at $3 per million input tokens and $15 per million output tokens, including thinking tokens. Meanwhile, Gemini 2.0 Pro is now available as an experimental model to developers in Google AI Studio and Vertex AI and to Gemini Advanced users in the model drop-down on desktop and mobile. In the free tier, users can process inputs and generate outputs at no cost. In the paid tier, input processing costs $0.10 per million tokens for text, images, and videos, while audio inputs are priced at $0.70 per million tokens. Output generation is available at $0.40 per million tokens. Moreover, context caching is free in the free tier. In the paid tier, however, it costs $0.025 per million tokens for text, image, and video data and $0.175 per million tokens for audio. In Conclusion Each of these models excels in different areas, reflecting the diverse strategies employed by their developers. The choice between these models should be based on specific needs and the type of tasks intended for them. Grok 3 stands out with its multimodal capabilities and advanced reasoning, while Claude 3.7 Sonnet shines in coding and complex problem-solving. OpenAI’s o3-mini offers cost-efficient reasoning and flexibility, whereas Google’s Gemini 2.0 boasts an extensive context window and strong multimodal capabilities.","excerpt":"Each of these models excels in different areas, reflecting the diverse strategies employed by their developers.","categories":["Global Tech"],"tags":["Anthropic","Google","XAI"],"author_name":"Siddharth Jindal","publish_date":"2025-02-26T18:58:44","publication_year":"2025","word_count":1031,"keywords":["Anthropic","ChatGPT","Gemini 2.0","Grok 3","TPU","OpenAI","AI","R","GPT-4.5","XAI","Google","xAI"],"extracted_tech_keywords":["AI","GPT-4.5","ChatGPT","OpenAI","Anthropic","Gemini 2.0","Grok 3","xAI","TPU","R"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/grok-3-vs-claude-3-7-sonnet-vs-o3-mini-vs-gemini-2-0\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":52203,"title":"Instagram Launches New AI Tool To Fight Cyberbullying","content":"Cyberbullying is one of the biggest problems faced by Millenials and Gen Z on the Internet. There are various kinds of cyberbullying on social media platforms but humiliating persons through hateful messages is one of the most common ones. To mitigate such issues, social media platforms have been trying hard to improve and secure their platforms through various tools and techniques. According to reports, India has the highest rate of parents confirming instances of cyberbullying. 37% of parents across India said their children were bullied online, with 14% of that total saying the bullying occurred on a regular basis. Instagram has been trying to curb online bullying for a few years now. The social media giant vowed a long-term commitment to fighting against online bullying by developing and deploying artificial intelligence techniques which will predict and recognises different types of offensive texts or bullying on Instagram. To that effect, earlier this month, Instagram launched a new AI feature which notifies the users when their captions on a photo or video may be considered offensive and give them a chance to pause and reconsider their words before posting. Working Principle of the New Feature The working of this feature is such that when a user writes a caption for a post, the AI system tries to detect if the caption is offensive or not. In the former case, the user will receive a prompt informing that the given caption is similar to those reported for bullying. The AI system is automated and is fed with previously extracted bullying reports which include similar word captions. Thus the user will have an opportunity to edit the caption before it is posted on the platform. This feature can be said as the extension of the previously launched AI feature. Currently, the new feature has been rolled out in selected countries and it will be soon expanded globally in the coming months. Advantages of This New feature The new feature will provide advantages like: Limiting the reach of bullyingEducating the users as to what things have been reported as the bully and what they cannot use on the platform In October this year, Instagram had launched a feature known as Restrict designed to empower a user to quietly protect the account while still keeping an eye on a bully. Once this feature is enabled, comments on the posts from a person which the user has restricted will only be visible to that person. The comment can be viewed by tapping “See Comment” or it can be either ignored or deleted. Earlier this year, the social media platform also launched a feature which notifies users when their comment may be considered offensive before it’s posted. This feature encourages the user to not only undo the bullying comment but also share something else which is less hurtful. Wrapping Up Today’s youth has been facing a disproportionate amount of online bullying. According to a research report, nearly 8 out of 10 individuals are subjected to the different types of cyberbullying in India. Out of these around 63% faced online abuses and insults, and 59% were subject to false rumours and gossip for degrading their image. These initiatives by Instagram will surely provide support to users from online bullying on the platform.","excerpt":"Cyberbullying is one of the biggest problems faced by Millenials and Gen Z on the Internet. There are various kinds of cyberbullying on social media platforms but humiliating persons through hateful messages is one of the most common ones. To mitigate such issues, social media platforms have been trying hard to improve and secure their […]","categories":["AI Features"],"tags":["instagram"],"author_name":"Ambika Choudhury","publish_date":"2019-12-17T19:01:44","publication_year":"2019","word_count":543,"keywords":["Go","artificial intelligence","programming_languages:R","AI","ETL","programming_languages:Go","RAG","instagram","R"],"extracted_tech_keywords":["AI","artificial intelligence","RAG","R","Go","ETL","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/instagram-launches-new-ai-tool-to-fight-cyberbullying\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10053492,"title":"Inside Carnegie Mellon University’s ‘Maker Space’","content":"On November 10, JPMorgan Chase & Co. officially opened the AI Maker Space in partnership with Carnegie Mellon University (CMU) at the university’s campus. The space will be supported by JP Morgan, also the founder of the AI Research programme at the university, to accelerate ways to use technology to predict and affect economic trends, protect data, stop financial crime and improve customer interaction with businesses. According to leaders at CMU and JPMorgan, the Maker Space will be providing a platform for students to collaborate with the aim of fostering and developing artificial intelligence (AI) technology to tackle the world’s toughest problems, “that will change the world.” Facilitator of Tech Advancements The 2000 square foot facility will provide students with the opportunity to work with robots, drones, smart appliances, augmented reality (AR) devices, virtual reality (VR) devices and high-end computers. Students can use smart appliances and devices, along with AR and VR, to explore the future of work. Additionally, the AI Maker Space boasts of a 12-foot by 12-foot drone cage equipped with motion capture cameras, along with a kitchen with smart appliances. Students of colleges– School of Computer Science, College of Engineering and School of Drama– can access the drone cage for either 3D art production or stage a flight. To add to that, students will be provided access to software packages associated with the hardware. This will enable them to develop newer AI technologies. According to CMU, these hardware and software packages are beyond what typical students get access to. Additionally, students will get access to massive datasets that they can use for their projects. They will be able to use synthetic financial datasets provided by JPMorgan to better train AI and machine learning (ML) algorithms. Robots including Baxter, Fetch and Pepper will be provided to students for training, testing and experimentation. The Fetch robot– worth $100,000– can be used to operate in systems for loading, unloading, and, like the name suggests– to fetch things from a refrigerator or to assemble furniture. Baxter, on the other hand, is a two-armed Rethink Robotics’ product. Built with an animated face, the robot can be used for simple and ‘dull’ industrial tasks such as loading, unloading, handling of materials and sorting. Softbank Robotics-manufactured Pepper robot is a semi-humanoid that can read emotions based on detection and analysis of facial expressions and voice tones. Summing up The Maker Space will be open for students’ access from Spring 2022, and students can enrol for a training course to learn using the technology. Additionally, the space is expected to host projects to help students familiarise themselves with the devices and tools. In India as well, institutions have been heavily investing in building tech labs. For instance, the Indian Institute of Technology-Kharagpur has set up a technological hub, called the Innovation Hub, to allow students and researchers to work on AI and ML advancements after receiving a Rs 170 crore grant from the Department of Science and Technology, Government Of India. The mission of building the Innovation Hub is to promote research and promote the startup ecosystem to scale them. Among the 32 technologies that are to be developed in the Innovation Hub, some include 3D printing technology, wastewater engineering, next-gen wireless communication, farm machines and structural heal and road traffic models. While universities across the world, and in India, have been increasingly investing in labs and spaces to boast and further accelerate the use of technology to build powerful tech, the Maker Space definitely seems to go a level higher to help students build tech addressing and solving global problems. To know more about the AI labs facilitating tech advancements in India, click here.","excerpt":"The Maker Space is a platform to foster and develop AI tech to tackle world’s toughest problems.","categories":["AI Trends"],"tags":["AI (Artificial Intelligence)","augmented reality","Carnegie Mellon University","Machine Learning","Robotics","Robots","Techgig","virtual reality"],"author_name":"Debolina Biswas","publish_date":"2021-11-15T14:00:00","publication_year":"2021","word_count":610,"keywords":["Go","Carnegie Mellon University","artificial intelligence","machine learning","AI","augmented reality","virtual reality","ML","Machine Learning","innovation","Robotics","Robots","RAG","Techgig","Aim","GAN","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","Aim","RAG","R","Go","GAN","innovation"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/inside-carnegie-mellon-universitys-maker-space\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10081589,"title":"ChatGPT: Why Stack Overflow Banned the Celebrated OpenAI Chatbot","content":"ChatGPT, OpenAI’s newest piece of innovation, has been the talk of the tech town for over the past week. Trained on the cutting-edge GPT 3.5, this chatbot’s human-like answers and context-aware nature has won the hearts of netizens the world over. However, just as with any other AI innovation, ChatGPT has quickly turned into a double-edged sword. Stack Overflow, the popular programming forum, has banned all answers created by ChatGPT, citing a high degree of inaccuracy in the bot’s responses. While it clarified that it was a temporary policy, it did reiterate that the problem not only lies in the inaccuracy of ChatGPTs answers, but deeper in the way the bot phrases its answers. Due to the way LLMs like GPT 3.5 work, chatbots like ChatGPT can quickly generate grammatically correct content with a formal tone. This makes the answers sound authoritative and backed by evidence, even in cases where they might not be. Stack Overflow emphasised this, stating, “[ChatGPT’s answers] typically look like they might be good and the answers are very easy to produce. There are also many people trying out ChatGPT to create answers, without the expertise or willingness to verify that the answer is correct prior to posting.” This only sheds more light on the problem with large language models and answers being accepted as facts. Even models trained on research papers and peer reviewed journals suffer similar pitfalls, as seen with Meta’s Galactica. Even as ChatGPT is being touted as a Google killer, will it be able to scale the problems plaguing LLMs today? The Stack Overflow ban Stack Overflow functions primarily as a developer Q&A site, with wrong answers typically receiving downvotes and, by extension, lower visibility. However, the site relies on a handful of expert volunteers to undertake content curation, weed out factually incorrect answers and resolve issues raised by the developers. The forum as a whole moves the question towards its solution with interaction and conversation between posters. Ever since ChatGPT entered the scene, the site has seen an influx of ‘correct-looking’ answers which are not actually correct. The ban was put in place so moderators could catch up to the giant volume of posts made using ChatGPT. ChatGPT tries to formulate a solution with the dataset it was trained on. Apart from not having “real-world experience” like the posters on forums, it also cannot access the internet to find solutions for problems it was not trained on. If we are moving into the age of all-knowing chatbots crawling the web for information, it is important to know the drawbacks that come with using an LLM. Stack Overflow is just the tip of the iceberg when it comes to a possible misuse of LLM-based chatbots. The problem with LLMs The issues with LLMs are manifold. First, and most importantly, LLMs are trained on a limited amount of data. No single LLM has all the information available on the web, and even those that have a large amount of data fall short when it comes to the nature of their datasets. Many LLM datasets are full of biases, explicit language, and incorrect information – traits which can easily be passed on the agent during training and inference. Second, LLMs have a tendency to hallucinate information. While training the agent using reinforcement learning, there is no single source of truth. AI systems have subjectivity baked into them by design, as this then allows the user to make the right choice. However, as LLMs and chatbots become more proficient in natural language tasks, users forget that it is their responsibility to identify whether the AI is factually correct or not. It is said that half-knowledge is more dangerous than ignorance, and that is especially applicable when speaking about LLMs. LLMs know just enough to be dangerous, and while they contain large amounts of data, they cannot be trusted to draw the correct conclusions from these data points. OpenAI has put content moderation filters in place and trained the bot to decline questions that it cannot answer correctly. However, the final responsibility still lies with the user to make the right decision. Even though OpenAI claims to have reduced the number of untruthful responses based on their learning from training GPT and Codex, the bot still has ways to generate inaccurate responses. This is considering the fact that OpenAI is also dedicated to responsible AI research, a sentiment which is not echoed by all companies. We have only seen communities reel from the impact of AI after they have been released to the public, such as Getty Images’ banning of AI art or the DeviantArt scraping controversy. It is important, now more than ever, for online administrators to take a proactive approach to the possible impact of AI on various online communities.","excerpt":"Ever since ChatGPT entered the scene, the site has seen an influx of ‘correct-looking’ answers which are not actually correct","categories":["Global Tech"],"tags":["how to use chat gpt","OpenAI"],"author_name":"Anirudh VK","publish_date":"2022-12-06T13:00:00","publication_year":"2022","word_count":796,"keywords":["Go","ChatGPT","API","OpenAI","AI","chatbots","GPT","Aim","Rust","how to use chat gpt","R"],"extracted_tech_keywords":["AI","ChatGPT","OpenAI","Aim","chatbots","R","Go","Rust","API","GPT"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/chatgpt-why-stack-overflow-banned-the-celebrated-openai-chatbot\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":9544,"title":"Analytics driven Healthcare: Newborn gets gift of life","content":"The analytics wave has hit the healthcare sector and the Indian healthcare industry is responding by transforming from volume to value. More and more healthcare organisations are focusing on becoming patient-centric as they continue to compete on customer experience, thereby moving towards adopting advanced technology and exploring the benefits of analytics. Today, the doctors, can go through numerous treatment records from around the world and conduct high-speed analysis by running through millions of research papers and genetic sequencing data. Thereby, gaining insights on how to treat a patient with a specific blood group, DNA sequence or characteristic. Further, using analytical tools, doctors can benchmark patients against past occurrences, analyze how patients with specific genes respond to different treatments, and make decisions based on facts, not depending on instincts. Earlier what could be achieved for only a small number of patients (5-10%) can now be done for a significantly increased number. As a result, many lives are being saved as many complex and rare health procedures can be performed with more efficiency. Let’s take a look at how Big Data, Cloud and Analytics helped a toddler smile again. Anshu fights for life She couldn’t move her head since the day she was born. Her head weighed as much as the rest of her body. She didn’t suckle; she never even cried for food – she was lost in her own world, drifting towards what seemed like certain death. Anshu (name changed), now 18 months old, was born with two rare brain diseases called hydranencephaly and hydrocephalus that caused fluid to accumulate in her brain. Born to poor tribal parents in Jharkhand who couldn’t afford state-of-the-art treatment, she was helpless until international donors pitched in and she was transported to New Delhi to one of the multi-specialty hospitals for a rare surgery. Real-time analytics does the trick The doctors in New Delhi used new a software, “Patient Data Explorer” to gather patient data from various sources such as clinical information systems, bio-bank systems, as well as text documents such as a physician’s notes. The tool also provided a summary of every patient’s medical history in a graphical timeline, consequently making it much easier to access information at any level of detail. This data was further analyzed and utilized within minutes as opposed to days. In August 2015, when she was only 10 months old, the surgery that had just 5% rate of survival was performed successfully on Anshu. This was carried out by inserting hi-tech stents, ultra-thin plastic tubes, in the child’s brain to drain away the fluid. Now, the size of her head has reduced to half. Today, Anshu is alive and kicking, all thanks to human persistence coupled with technology. The head surgeon of the team that operated on the infant said, “When Anshu came here, she was highly underweight and the actual prognosis was dim. However, using analytical software, we were able to extract and analyze data from various sources. Moreover, the way it was presented gave us and other researchers surprising new insights.” He further added that in the future, they would like to use such Analytical tools for every diagnosis and treatment because the manifestation of diseases varies from patient to patient. Thus, the real-time Analytics software considerably improves the efficiency of medical research and has particularly been helpful in tackling rare diseases and cancer. This amalgamation of data, analytics and cloud can yield interesting results and breakthroughs for the mankind. The only thing that remains to be seen is how fast the Indian healthcare will adapt the analytics softwares as a norm.","excerpt":"The analytics wave has hit the healthcare sector and the Indian healthcare industry is responding by transforming from volume to value. More and more healthcare organisations are focusing on becoming patient-centric as they continue to compete on customer experience, thereby moving towards adopting advanced technology and exploring the benefits of analytics. Today, the doctors, can […]","categories":["IT Services"],"tags":["Analytics Case Study"],"author_name":"Muqbil Ahmar","publish_date":"2016-04-12T05:59:32","publication_year":"2016","word_count":596,"keywords":["big data","Go","programming_languages:R","AI","programming_languages:Go","GAN","analytics","real-time analytics","Analytics Case Study","R"],"extracted_tech_keywords":["AI","analytics","R","Go","big data","real-time analytics","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/analytics-driven-healthcare-newborn-gets-gift-life\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10173063,"title":"AWS Launches Space Accelerator to Support 40 Startups Across India, Japan, and Australia","content":"Amazon Web Services (AWS) has launched the AWS Space Accelerator: APJ 2025, a 10-week programme for space-tech startups in India, Japan, Australia, and New Zealand. The initiative aims to support up to 40 startups working in space infrastructure, earth observation, and drone systems by offering business mentorship, AWS technical support, and up to $100,000 in cloud credits. Applications are open from July 8 to September 5, 2025, with the programme starting in September and culminating in a demo day in December 2025. The initiative will run both virtually and in person. AWS partners such as T-Hub, Minfy, Fusic, and Ansys, alongside space agencies like IN-SPACe and the Australia Space Agency, will help deliver the programme. T-Hub will manage operations and host the initiative in India. Minfy Technologies will provide AWS training and technical support in both India and Australia. In Japan, Fusic will offer technical mentorship, while Ansys will assist startups globally with simulation and design testing. Startups in Different Sectors This accelerator builds upon the 2024 India edition, which supported 24 startups across various fields, including propulsion, satellite imagery, and quantum key distribution. Several of these participants secured customer contracts or investments and advanced core technologies. The new programme targets startups focused on earth observation and remote sensing, satellite manufacturing and propulsion, and drone technologies that complement space infrastructure. These focus areas aim to improve agriculture, climate resilience, and connectivity in remote areas. By leveraging AWS cloud tools, participants can experiment rapidly, reduce costs, and validate solutions before large-scale deployment, addressing key challenges such as capital-intensive testing and the need for specialised talent. Local Partnerships to Boost Innovation Clint Crosier, director of aerospace and satellite at AWS, said, “The collaboration with Australian Space Agency, IN-SPACe, iLAuNCH, and Sky Perfect JSAT underscores our commitment to working with local space agencies and industry leaders.” He added that through this accelerator programme, the company aims not only to support individual startups but also to build a robust community that can drive economic growth and technological advancement throughout the region. The accelerator arrives at a time of regional momentum in India’s space industry, as it is projected to grow to $44 billion by 2033. Meanwhile, Japan aims to expand its space economy to ¥8 trillion (~$52 billion) by the early 2030s.","excerpt":"The initiative will support startups working in different sectors with AWS technical support and up to $100,000 in cloud credits.","categories":["AI News"],"tags":["AWS","space india","space startup"],"author_name":"Sanjana Gupta","publish_date":"2025-07-08T17:59:00","publication_year":"2025","word_count":378,"keywords":["API","AWS","AI","cloud_platforms:AWS","innovation","space india","RAG","Aim","space startup","ViT","R","startup"],"extracted_tech_keywords":["AI","Aim","RAG","AWS","R","API","ViT","innovation","startup","cloud_platforms:AWS"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/aws-launches-space-accelerator-to-support-40-startups-across-india-japan-and-australia\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":16673,"title":"DeepMind helps AI to master a new skill— “Imagination”","content":"Quite recently, there have been many interesting developments around Artificial Intelligence that has caught our curiosity. It began with the war of words between the tech leaders Musk and Mark on whether having AI regulation is a necessity or not, which was quickly followed by Facebook’s AI chatbot developing their own language, leaving the AI experts clueless. Adding onto the AI stories is the recent development from Google’s DeepMind, where the company has come up with an AI capable of ‘imagination’. Yes, you hear it right! If it weren’t enough for AI to evoke actions similar to humans, that it has now taken away the exclusive ability of humans to “imagine” and think before it acts. According to the research papers released by the London based AI firm, these capabilities can now enable machines to see and evaluate the consequences of their actions before they take them. The company believes that if they get successful in creating algorithms that can simulate this distinct human ability to construct a plan, it could help in solving complex tasks more efficiently. “In two new papers, we describe a new family of approaches for imagination-based planning. We also introduce architectures which provide new ways for agents to learn and construct plans to maximize the efficiency of a task”, the company said in its blogpost. DeepMind has done some tremendous work in AI, and its AlphaGo AI beating as series of human champions at various levels is a testimony to that. It wouldn’t be unlikely to that the promise of AlphaGo goes beyond board game and is tackling real world problems with much efficiency. Agents that can Imagine! The company has developed what they call, “Imagination Augmented Agents” with a neural network that can learn to extract any information useful for the agent’s future decisions, and ignore that which is not relevant. Few of the things they can do are— learn to interpret their internal simulations, use their imagination efficiently, and learn different strategies to construct plans. They are capable of doing this by adapting the number of imagined trajectories to suit the problem. Encoder enhances its efficiency and extract additional information from imagination beyond rewards. “These trajectories may contain useful clues even if they do not necessarily result in high reward”, mentioned the blogpost. The company tested their architecture on multiple tasks including the puzzle game Sokoban and a spaceship navigation game, both of which require extensive planning and reasoning. The company reported AI agent playing Sokoban with much perfection without knowing the rules of the game. An agent playing the spaceship task. The red lines indicate trajectories that are executed in the environment while blue and green depict imagined trajectories. Img Src- DeepMind “For both tasks, the imagination-augmented agents outperform the imagination-less baselines considerably: they learn with less experience and are able to deal with the imperfections in modelling the environment”, explained the blogpost. Though they have achieved a major victory, the researchers believe that AI systems need to be more sophisticated to be able to operate in a more complex environment to solve tasks efficiently. AIM had earlier reported that Google DeepMind has developed a lip-reading software in association with Oxford researchers that uses AI to improve the performance of automated speech recognition.","excerpt":"Quite recently, there have been many interesting developments around Artificial Intelligence that has caught our curiosity. It began with the war of words between the tech leaders Musk and Mark on whether having AI regulation is a necessity or not, which was quickly followed by Facebook’s AI chatbot developing their own language, leaving the AI […]","categories":["AI News"],"tags":["DeepMind AI","deepmind london"],"author_name":"Srishti Deoras","publish_date":"2017-08-02T10:05:00","publication_year":"2017","word_count":541,"keywords":["Go","artificial intelligence","programming_languages:R","AI","neural network","programming_languages:Go","Aim","GAN","DeepMind AI","R","deepmind london"],"extracted_tech_keywords":["AI","artificial intelligence","neural network","Aim","R","Go","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/deepmind-helps-ai-master-new-skill-imagination\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":22540,"title":"Why Are Big Tech Giants Like Google &#038; IBM Rushing To Commercialize Quantum Computing?","content":"Quantum computing is all set to go beyond labs into the cloud – tech giant Google announced Bristlecone, a quantum processor which will provide a provide a testbed for research into system error rates and scalability of qubit technology, as well as applications in quantum simulation, optimization, and machine learning. Announcing the release Julian Kelly, Research Scientist at Google’s Quantum AI Lab noted that quantum supremacy can be achieved with Bristlecone and could be a compelling proof-of-principle for building larger scale quantum computers. Google researcher inside the Quantum Lab Well, Google is not really the first company to make an attempt to commercialize quantum computing – IBM has made significant progress across the entire quantum computing technology stack first brought it to the cloud in 2016 with a 5 qubit quantum computer. In November 2017, IBM raised the bar by announcing a third generation of quantum computers with a 50 qubit prototype, a huge milestone for quantum computing, but there is no news whether it is commercially available. The company was quick to point out that it is not ready for common use. However, the company made a 20-qubit system available through its cloud computing platform. What’s behind the dash towards Quantum Computing? Google’s newly unveiled quantum chipset Bristlecone According to a research report from Morgan Stanley, quantum computing could hold the key to double the high-end computing market from $5bn to $10bn – that explains the reasons why companies are accelerating the mass adoption by moving quantum computers from labs to commercial activities. Outside of big tech companies like Google, Microsoft, Intel, IBM & Nokia Bell Labs, there are a slew of emergent players like Rigetti, a full stack quantum computing company, ionQ & D-Wave. However, the research also posits that quantum computing is not suited to all compute tasks and web servers will continue to store data but could gain a significant edge post 2025. Let’s see why quantum computing has become a key enterprise strategy: Usher the fourth industrial revolution: Quantum computing has moved from fundamental theoretical research to an engineering development phase and could trigger the fourth industrial revolution, with significant developments in sectors where more computer power is required for R&D efforts – drug discovery in pharma, plane design in aerospace, well data analysis in Oil & Gas, polymer design in Chemical and predictive maintenance in manufacturing. Use cases go beyond traditional sectors: According to Dr Vijay Pande, entrepreneur and Stanford Professor, quantum computing will have a profound impact in area where classic computers fail – such as the ability to model the chemistry (more precisely, quantum chemistry) of molecules involved in drugs, or in industrial processes such as energy, new advances in machine learning and novel cryptography. Add new revenue stream: According to Communications Industry Researchers (CIR), revenues from quantum computing are forecasted to grow from US$1.9 billion in 2023 to US$8.0 billion by 2027. Market has expanded: The market for quantum computing has expanded beyond the traditional base of aerospace, drug discovery and financial services, healthcare, energy industry and even general business planning. From cloud to on-premise in the future: By making quantum computing a central part of their enterprise computing strategy, businesses are hoping to add a new line of revenue by delivering quantum computing via cloud. CIR research posits that in the future, we could also see a rise in on-premise quantum computing. Startups fueling quantum computing software market: Not just big tech companies, startups like Rigetti have also jumped on the cloud bandwagon and are paving the way for quantum software market. By 2023, the market for quantum computing software is billed to touch a whopping $408 million, with 60 percent of these revenues from applications packages for cloud service providers. Quantum computing forecasted to grow: When it comes to the current Quantum Enterprise market, the biggest users are in the R&D space.  By 2024 this share will be around 30% and the biggest revenue drivers will come from the defense\/aerospace, pharmaceuticals\/specialty chemical and banking\/finance sectors. In the race for quantum supremacy, IBM is a clear winner IBM Researchers inside the Quantum Lab In the race to commercialize quantum, IBM has emerged as a leader and has made stunning progress in the area even though it has just been one-upped by Google recently. Google’s new quantum processor Bristlecone is 72 qubits and can work with low error rates.  However, Google hopes that Bristlecone will become a “compelling proof-of-principle for building larger scale quantum computers”. Google’s Quantum AI Labs research scientist Julian Kelly also added that operating a device such as Bristlecone at low system error requires harmony between a full stack of technology ranging from software and control electronics to the processor itself. Getting this right requires careful systems engineering over several iterations, he noted. Let’s see how IBM is besting its competition in the next computing revolution According to tech writer Alan Radding, for quantum computing to succeed – companies need an ecosystem of vendors, developers & programming models to facilitate quantum adoption. IBM’s industry first initiative to build an ecosystem: IBM has made significant progress in the area by building a global ecosystem for quantum computing – academic institutions, research companies, enterprises such as Materials Magic (Hitachi Metals Group), a community of scientists, engineers, professors, investors and even startup founders. Their academic partners range from Oak Ridge National Lab to Oxford University and University of Melbourne. Open sourced Quantum Computing SDK: By releasing the open-source quantum software development kit, QISKit, IBM is advancing its quantum technology and systems and fueling adoption of quantum computing. Helping enterprises become quantum ready: To help companies adopt quantum faster and make it a part of their tool box, IBM has also rolled out IBM Q Consulting to help enterprises become quantum ready  and gain significant value through the application of quantum computing technology. The big tech giant has brought on board scientists and industry experts to provide its clients with customized roadmaps and become quantum-ready. Era of cloud: IBM brought quantum computing to the cloud, second after Canada-based D-Wave – in May 2016 the company made quantum online which meant that the company went to great lengths to make it easy to use for researchers and physicists who lack the experience of working on quantum computers. Even though D-Wave was the first to bring quantum to cloud (also showed a market for these machines), launched its quantum computing service on the cloud in 2010, it’s machines could only run a limited number of quantum algorithms. IBM’s popularity has soared: As opposed to D-Wave’s quantum computers, IBM’s quantum computer, demonstrate “more general in its capabilities and at least in theory are able to run any quantum algorithm”, Jerry Chow, manager of experimental quantum computing at IBM’s Thomas J. Watson Research Center in New York to IEEE Spectrum. Since the advent of quantum into the cloud, IBM’s popularity has soared and its quantum processor available via IBM Cloud has gained around 40,000 users from more than 100 countries. IBM has made quantum a consumable entity: According to IBM’s Jerry Chow, the Quantum Experience has gradually built up a community interested in quantum computing – a bold step towards making quantum more consumable entity. Released new APIs: Last year, IBM also added new APIs to Quantum Experience to help researchers build more sophisticated applications. Despite the advances, challenges abound Dr Pande noted how despite several advances, there are still many challenges for quantum computing. He noted how quantum computing has turned a corner from being a scientific to an engineering problem. So, how can overcome the engineering challenges? By following a full-stack approach that enables rapid testing and pushing innovation at all scales, from gates to chips to fabs to programming languages. It’s a view seconded by Google’s Kelly, who also noted that operating a device like Bristlecone at low system error requires harmony between a full stack of technology ranging from software and control electronics to the processor itself.","excerpt":"Quantum computing is all set to go beyond labs into the cloud – tech giant Google announced Bristlecone, a quantum processor which will provide a provide a testbed for research into system error rates and scalability of qubit technology, as well as applications in quantum simulation, optimization, and machine learning. Announcing the release Julian Kelly, […]","categories":["IT Services"],"tags":["countries with quantum computers","julia scientific programming","new developments in cloud computing","quantum cloud software","quantum computers","Quantum Computing","quantum computing programming","quantum supremacy"],"author_name":"Richa Bhatia","publish_date":"2018-03-12T04:28:40","publication_year":"2018","word_count":1325,"keywords":["Quantum Computing","Go","new developments in cloud computing","quantum supremacy","machine learning","API","AI","countries with quantum computers","cloud computing","quantum computers","quantum cloud software","Scala","ViT","quantum computing programming","Julia","GAN","R","julia scientific programming"],"extracted_tech_keywords":["AI","machine learning","cloud computing","R","Go","Scala","Julia","API","GAN","ViT"],"url":"https:\/\/analyticsindiamag.com\/it-services\/why-are-big-tech-giants-like-google-ibm-rushing-to-commercialize-quantum-computing\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":6,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10065324,"title":"Top 6 alternatives to Microsoft’s DeepSpeed","content":"Microsoft’s DeepSpeed was introduced in 2020 and is one of the most popular deep learning optimisation libraries today. The mission behind the library was to make distributed training easy, efficient, and effective. Today, DeepSpeed can train a language model with one trillion parameters using as few as 800 NVIDIA V100 GPUs. Over the years, many open-source deep learning optimisation libraries have been announced by tech giants such as Google, Microsoft, Uber, DeepMind and others, but DeepSpeed remains one of the most popular. Today, we are looking at the top six alternatives to DeepSpeed. DeepSpeed enables trillion-scale model training through its combination of powerful technologies. It can also scale to thousands of GPUs, including data-parallel training, parallel model training, and pipeline parallel training. The team can also optimise for MoE models at scale and reduce the cost of training and inference for large models. The library is PyTorch-compatible and proven to improve large model training, achieving over 5x in system performance. The Turing-NLG with 17B parameters is one of the earliest leveragers of the library. Fairscale by Facebook Research Fairscale by FAIR is a PyTorch extension for high performance and large scale training. This library extends basic PyTorch capabilities with new SOTA scaling and the latest distributed training techniques. It does so in the form of composable modules and easy to use APIs that help to scale models with limited resources. FairScale follows three major missions; to allow users to understand it with minimum cognitive overload, to combine multiple FairScale APIs as part of their training loop seamlessly, and leverage the best scaling and efficiency performance. FairScale can be used across multiple axes and provides solutions for scaling models by layer parallelism and tensor parallelism. It also allows users to achieve low memory utilisation and efficient computation, deal with optimising memory usage irrespective of the scale of the model, training without hyperparameter tuning and other techniques to optimise training performance. It also features inter-and intra-layer parallelism, splitting models across multiple GPUs and hosts. Nvidia TensorRT NVIDIA’s TensorRT is a C++ library for high-performance inference, built on NVIDIA GPUs and deep learning accelerators. The library is built on CUDA, NVIDIA’s parallel programming model, enabling developers to calibrate for lower precision with high accuracy, optimise neural network models, and deploy to hyperscale data centres, embedded platforms, or automotive product platforms. Deep learning models from almost all popular frameworks can be parsed and optimised for low latency and high throughput inference on NVIDIA GPUs using TensorRT. The three essential optimisations it allows for are Mixed Precision Inference, Layer Fusion and Batching. In addition, it uses the new NVIDIA Ampere Architecture GPUs and sparse tensor cores for an additional performance boost. TensorRT offers INT8 based on quantisation aware training, post-training quantisation, and FP16 optimisations. These can be used for applications for production deployments of deep learning inference applications, such as video streaming, speech recognition, recommendation, fraud detection, text generation, and natural language processing. It also minimises application latency for better real-time services. Horovod Uber developed Horovod to make distributed deep learning fast and easy to use. It is a distributed deep learning training framework for TensorFlow, Keras, PyTorch, and Apache MXNet. The library is claimed to bring model training time down from days and weeks to hours and minutes. Horovod leverages message passing interface stacks such as OpenMPI to enable a training job to run on a highly parallel and distributed infrastructure without any modifications. It is currently hosted by the LF AI & Data Foundation. The primary motivation is to ease the task of taking a single-GPU training script and successfully scaling it to train across many GPUs in parallel. Once the training script has been written for scale with Horovod, the Uber team claims it can run on a single GPU, multiple GPUs, or even multiple hosts without any further code changes. The framework achieved 90% scaling efficiency for both Inception V3 and ResNet-101 and 68% for VGG-16. It is further leveraged by companies like the NVIDIA Collective Communications Library and Message Passing Interface (MPI). Horovod helps distribute and aggregate model parameters across workers, optimise network bandwidth usage and scale deep neural network models. Mesh TensorFlow Mesh TensorFlow is a language for distributed deep learning that can specify a broad class of distributed tensor computations. The language aims to formalise and implement distribution strategies for computation graphs over the hardware\/processors. It is implemented as a layer over TensorFlow. TensorFlow occupies only one GPU for training and works in two methods: data parallelism and model parallelism. Mesh is a repository for model parallelism on supercomputers. It allows the users to specify any tensor dimensions to be split across any dimensions of a multi-dimensional mesh of processors. The MeshTensorFlow graph compiles this in an SPMD program with parallel operations coupled with collective communication primitives such as Allreduce. MNN MNN is a blazing fast, lightweight deep learning framework battle-tested by business-critical use cases in Alibaba. It is a lightweight mobile-side deep learning inference engine called Mobile Neural Network (MNN). MNN focuses on the acceleration and optimisation of inference while solving efficiency problems during model deployment. Mobile Neural Network ensures optimisation, conversion, and inference of deep neural network models. Apps leverage it for live broadcast, short video capture, search recommendation, product searching by image, interactive marketing, equity distribution, security risk control and other scenarios. MNN stably runs more than 100 million times per day, is applied in IoT devices and is used in scenarios like smiley face red envelopes, scans, and a finger-guessing game. Deepmind’s TF-Replicator DeepMind’s TF-Replicator is a framework for distributed machine learning designed for DeepMind researchers and implemented as an abstraction over TensorFlow. It focuses on the scalability related to how TensorFlow programs leverage Tensor Processing Units (TPUs). It mitigates the portability and adoption barriers of the TPU with a simpler and developer-friendly programming model. TF-Replicator is based on the “in-graph replication” pattern that computes for each device and replicates in the same TensorFlow graph. This allows for a high level of parallelism where the TF-Replicator first builds independent computations for each device and then leaves placeholders on specific cross-device computation use-cases. Once the sub-graphs for all devices have been built, TF-Replicator connects them by replacing the placeholders with actual cross-device computation. This allows users to scale up workloads to many devices and seamlessly switch between different types of accelerators.","excerpt":"Over the years, many open-source deep learning optimisation libraries have been announced by tech giants but DeepSpeed remains one of the most popular.","categories":["AI Trends"],"tags":[],"author_name":"Avi Gopani","publish_date":"2022-04-20T10:00:00","publication_year":"2022","word_count":1056,"keywords":["machine learning","Keras","AI","neural network","PyTorch","ML","RAG","Aim","deep learning","TensorFlow"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","neural network","Aim","TensorFlow","PyTorch","Keras","RAG"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-6-alternatives-to-microsofts-deepspeed\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10057358,"title":"What’s the cryptocurrency clampdown in China about?","content":"For a long time, China has been one of the world’s major centres for cryptocurrency mining operations because of its relatively low electricity costs and cheaper computer hardware. The activity is so popular there that gamers have been known to blame the industry for a global shortage of graphics cards—which miners use to process crypto-currencies. While China banned domestic cryptocurrency exchanges a while ago, trading continued clandestinely by other means. It was only in 2021 that China made efforts to take down cryptocurrency with greater official scrutiny. Now, it is likely that 90% of Bitcoin mining capacity in China will shut down. China’s crackdown on crypto in 2021 In May of this year, Chinese financial regulators issued a statement that reminded consumers of the dangers of virtual currency—warning them they would have no protection for continuing to trade Bitcoin and other currencies online. In June, Chinese state institutions further banned banks and payment companies from working with crypto-related businesses—similar to bans that they had earlier introduced in 2013 and 2017. The government’s main administrative cabinet vowed to clamp down on Bitcoin trading and mining, leading local authorities to shut down crypto-mining operations across the country. In September, China’s most powerful regulators issued a nationwide ban on cryptocurrency, and all financial transactions involving cryptocurrencies were deemed illegal. The statement issued by the government made it clear that all those participating in “illegal financial activities” would be prosecuted. After this development, the price of Bitcoin fell by more than $2000. Why did the clampdown happen? The Chinese government has viewed cryptocurrencies as a hindrance to its control over capital flows in the country for a while. The country’s government has also made claims that they plan to punish “illegal” crypto-mining activities as a preventative measure against the risks of a decentralised and unregulated industry whose development is blind and disorderly. China further claims that its moves to regulate cryptocurrency is an attempt to reach its carbon reduction goals. A common criticism of Bitcoin and other cryptocurrencies is that it takes a lot of energy to mine them. The process of creating Bitcoin to spend or trade uses up an astonishing amount of electricity—annually consuming more electricity than the country Finland, a nation made up of 5.5 million people. China has also been testing its own official state-controlled digital currency. The e-yuan could be much faster and cheaper than traditional online payment methods while simultaneously being more efficient than Bitcoin (which requires massive computing power). It would also make it easier for the government to collect data and keep tabs on the everyday transactions of its citizens, and to have controlled access to their funds. Therefore, making cryptocurrencies illegal could also be a part of a broader effort to siphon citizens away from popular private providers of financial service—including AliPay and WeChat. The bright side of the ban According to Sam Bankman-Fried, the chief executive of FTX (one of the largest cryptocurrency exchanges), China’s bans on cryptocurrencies could give North American crypto mining a great opportunity to take off. Bitcoin mining is unlikely to cease, but operators could relocate to Texas—a place with a relatively lax regulatory environment and cheap electricity. Cryptocurrency regulation is a global problem The race to keep up with the developments of the $2 trillion cryptocurrency industry is a problem affecting governments worldwide. Regulation of cryptocurrency varies from country to country, and there are legitimate fears that these tokens could threaten the wider financial system. While the jury is still out on whether cryptocurrencies will be the future of global banking and trade or what destroys the western economic system as we know it—it can’t be denied that the market continues to grow in popularity.","excerpt":"For a long time, China has been one of the world’s major centres for cryptocurrency mining operations because of its relatively low electricity costs and cheaper computer hardware. The activity is so popular there that gamers have been known to blame the industry for a global shortage of graphics cards—which miners use to process crypto-currencies. […]","categories":["IT Services"],"tags":[],"author_name":"Srishti Mukherjee","publish_date":"2021-12-29T12:00:52","publication_year":"2021","word_count":616,"keywords":["Go","API","programming_languages:R","AI","programming_languages:Go","Git","Aim","ViT","R"],"extracted_tech_keywords":["AI","Aim","R","Go","Git","API","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/whats-the-cryptocurrency-clampdown-in-china-about\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10092968,"title":"Geoffrey Hinton Is The Bad Dad of AI","content":"The touted grand dad of deep learning, Geoffrey Hinton recently quit Google so he could talk more freely about the threats posed by AI. He will also be responding to requests for help from Bernie Sanders, Elon Musk and the White House, he said. But a few years ago when the ethical team at Google raised alarm about the big tech’s unethical practices, the AI Prometheus turned a blind eye towards the subject. In the past, Hinton has expressed concern over the potential AI danger, likening it to the creation of the atomic bomb by Robert Oppenheimer’s work on the Manhattan Project to develop the world’s first nuclear bombs during World War II. The 75-year old polymath believes that the pursuit of profit in AI development could lead to a world where AI-generated content outnumbers that produced by humans, thereby endangering our very survival. The Oppenheimer Fallacy In the past, when asked about the potential harm of AI, Hinton paraphrased Oppenheimer, saying that when one encounters something technically sweet, one goes ahead and pursues it. However, Hinton now expresses regret over the consequences of his work. He acknowledges that the far fetched idea of machines surpassing human intelligence, is now a realistic possibility. Hinton, who previously believed that such advancements were still “30-50 years away”, cites the recent advancements in large language models, particularly OpenAI’s GPT-4, as evidence of how quickly machines are advancing. He said, “Look at how it was five years ago and how it is now. Take the difference and propagate forwards. That’s scary.” Interestingly, the nuclear weapons analogy also resurfaced in the Stanford Artificial Intelligence Index Report 2023 which was released last month. The breadth of AI’s applications is unlike any other field. The report notes that 36% of NLP (natural language processing) researchers polled think that artificial general intelligence (AGI) could lead to a catastrophic event on par with a nuclear disaster. While the analogy provides a helpful point of reference, it has its limitations. AI incorporates elements from social media to nuclear weapons. Hence, analogies, like the Oppenheimer comparison, can be illuminating yet incomplete when describing the scope of AI. Hinton on the fence Addressing the NYT article by Cade Metz that suggested Hinton left Google in order to criticize the company, he clarified that he left Google to speak out about the dangers of AI without being constrained by any potential impact on the company. He further noted that Google has acted responsibly in its pursuit of AI. But we do not agree with him. Ethically, Google has been in a state of flux since 2020. The big tech fired its ethics team. Prominent Black female scientist Timnit Gebru, who was the first one to show the exit door, responded to Hinton’s quitting saying, “When Geoff Hinton was asked about the women’s concerns about AI for which we got fired, pushed out, harassed he said our concerns are not “existential risks” to humanity where as his are. This is what I mean about the almost exclusively white dudes who keep on talking about “existential risk.” Margaret Mitchell, a former leader on Google’s AI ethics team, is also upset that Hinton didn’t speak out about ethical concerns related to AI development during his decade in a position of power at Google, especially after the 2020 ouster of Timnit Gebru, who had studied the harms of large language models before they were widely commercialized into products such as ChatGPT and Google’s Bard. In 2018, Hinton had dismissed the need for Explainable AI. He voiced his opposition, arguing that it would be a “complete disaster” if regulators insisted that AI systems be explainable. The Canadian AI pioneer claimed that most people cannot explain how they work, and therefore requiring such explanations for AI systems would be counterproductive. Since the 1970s Hinton has been at the forefront of developing models for visual understanding of the human brain. Although he had previously maintained a detachment from the social impact of his work and claimed “I’m an expert on trying to get the technology to work, not an expert on social policy”, Hinton’s resignation cited concerns over the dangers of AI as the reason for his change of heart. This shift in stance suggests that Hinton can no longer remain neutral and must now acknowledge the potential impact of his work on society.","excerpt":"Visionary Geoffrey Hinton recently left Google to speak out about the dangers of AI","categories":["AI Features"],"tags":["AI &amp; ethics","AI Tool","Ethical AI","geoff hinton","Geoffrey Hinton","Interviews and Discussions"],"author_name":"Tasmia Ansari","publish_date":"2023-05-08T14:00:00","publication_year":"2023","word_count":724,"keywords":["Go","ChatGPT","geoff hinton","artificial intelligence","OpenAI","Geoffrey Hinton","AI &amp; ethics","AI","Ethical AI","BERT","NLP","Aim","deep learning","AI Tool","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","artificial intelligence","deep learning","NLP","ChatGPT","OpenAI","Aim","R","Go","BERT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/visionary-geoffrey-hinton-is-the-bad-dad-of-ai\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10080217,"title":"MongoDB on a Mission to Equip 85 Mn Developers with New Skills","content":"According to recent research, 40 million technical jobs need more skilled talent. This number is expected to rise to 85.2 million by 2030. To help close the widening skills gap across the globe, MongoDB Inc recently announced updates to the MongoDB University program. The updated program of the decade-old university will include a broader catalog of courses, streamlined certifications, 24\/7 exam access, hands-on Atlas labs, and select foreign language support. In a conversation with Analytics India Magazine, Raghu Viswanathan, VP, education, documentation & academia, MongoDB, said, “I’ve had the good fortune of watching the education business evolve over the years, from what was mostly live training, has turned into a digital education business, which has further changed due to cloud. Cloud has influenced how education gets created in today’s age. So many changes have happened, and I’ve been a good witness to that and lived it,” Before this, Viswanathan worked at Oracle for nearly 30 years. ‘Learning should be frictionless’ Viswanathan explained how MongoDB stands out in the market. “Unlike other platforms, we let users access content even without giving us their email ID. We want developers to work without any borders or inhibitions. The firm strongly believes that learning should be frictionless, so we’ve taken a different approach to let people access our learning.” These days, people look for things in different formats online. Some learn by consuming content, and some through quizzes. Everyone has a different take on learning and it’s not easy to find everything at one place. So, for the MongoDB team, the philosophy was to build a platform offering a combination of all important things at one place. “With the launch, we focused on a holistic change of modularity. Time is typically the barrier for most people to learn. So everything has been simplified into small five- to seven-minute consumable chunks. Beyond that, we have nearly one hundred lab exercises people can practice,” he said. Today, MongoDB university has about 37,000 customers across multiple industries. So with the underlying principles of what’s covered in these courses, they’re going to apply to a broad audience. Roadmap for India Today all major companies have a big hub in India, and MongoDB is no exception. What MongoDB University loves about India is the certification culture. Explaining the plans to increase the distribution of learning through other portals, Viswanathan said the company is looking to work with more universities in India by getting their program included in colleges. “I’d love for students to easily access MongoDB, either through professors or on their own,” he added. Latest updates ● Five new courses include the Introduction to MongoDB – Developer Course and Driver Courses for Node, Java, C#, and Python. All certifications have expanded to include 24\/7 exam scheduling and testing. ● Hands-on 88 Atlas Labs that cover popular features, including CRUD operations, search, aggregation, and indexing. ● More personalized experience with all content left ungated and without time restrictions. New subtitles in six languages have been added, including Chinese (Simplified and Traditional), French, Korean, Portuguese, and Spanish. ● Tutorials within 20 minutes called ‘Learning Bytes’ to provide the latest updates – from product announcements, new releases\/versions, advanced dives into niche topics or content, and developer interest areas.","excerpt":"The developer data platform recently announced updates to the MongoDB University program that include a range of courses and foreign language support","categories":["AI News"],"tags":["MongoDB","Oracle Certification"],"author_name":"Tasmia Ansari","publish_date":"2022-11-21T14:00:00","publication_year":"2022","word_count":537,"keywords":["Go","AI","MongoDB","ML","Git","RAG","Python","analytics","Oracle Certification","R","Java"],"extracted_tech_keywords":["AI","ML","analytics","RAG","MongoDB","Python","R","Go","Java","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/mongodb-is-on-the-mission-to-equip-85-mn-developers-with-new-skills\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10019894,"title":"How APIs Breathe Life Into ML Organisations","content":"“API monetisation and API first strategies have become a new normal with businesses with digital maturity.” Last year’s pandemic catalysed digital maturity across organisations. The niche markets found even more niche business opportunities, thanks to the widespread adoption and development of APIs (Application Programming Interface).  In their most basic form, APIs are doorways between two software applications and become extremely powerful when tailored to the needs of the developers. Web, mobile and automation are some of the key applications powered by APIs. According to a report by Google Cloud, API programs are the core drivers of digital transformation by playing a significant role in digital experiences, business operations, innovation, and growth. Current State Of API Economy (Image credits: Google Cloud) Companies around the world possess valuable data ready to be capitalised. All they need are the services that can bridge the gap between customers and third parties. APIs fit right into this mix. For instance, the banking sector has witnessed a tremendous revolution with the advent of fintech products. The infrastructure behind the payment gateways are powered by the APIs like those of Stripe or Razorpay. These fintech API providers are multi-billion dollar companies today. Machine learning-based API service providers are next in line to take the markets by storm. “Databricks API  supports services to manage clusters, instance pools, libraries, tokens, and MLflow models. Databricks is currently valued at $28 billion.” For example, last week, Databricks, a company that offers unified platform services raised $1 billion that rocketed its market value to $28 billion. Though AWS too offers Spark services, Databricks’ Spark services seem to have an edge over them. They offer additional customisations while combining the synergies of top players to serve an user. According to an Apigee survey, AI- and ML-powered API security and monitoring solutions used for anomaly detection and security analytics grew 230% year-over-year between September 2019 and September 2020. When easily reusable, APIs let developers modularly combine, and recombine functionality and data for new uses, with virtually no marginal cost for each additional use of the API. If one developer builds a new application by leveraging an API that looks up store locations, another developer can leverage the same API for another application without the enterprise incurring any additional overhead. The APIs will (source: Gartner): Make it easier for data scientists to find and choose from the huge variety of available algorithms, experiments, datasets and solution accelerators.Enable organisations to build advanced analytics solutions in a faster way.Address the skill gap in advanced technology.Help commercialise their solutions easily.Enables ease of ML model choosing. (API traffic by domain in 2020. Source: Google Cloud) APIs also allow the organisations to take smart decisions by providing details of the product consumption at the user level, which in turn can be used by the developers to enhance the end product. This sounds like every other business strategy, but APIs make it more accessible. It helps them understand the value of an organisation’s digital assets. Beyond helping enterprises, writes Bala Kasiviswanathan of Google Cloud, API analytics can help both IT and business leaders refine the KPIs they use for analytics. “If an API becomes popular with developers in a new vertical for example, that may persuade the enterprise to focus on KPIs like adoption among these specific developers, rather than on overall adoption,” said Bala. AI Through API In 2019, machine learning as a service (MLaaS) raked in an estimated $1 billion and is expected to grow to $8.4 billion by the end of 2025. The success of these services can be traced to the customised APIs. For example, Google’s prediction API, can be used to classify an image for $0.0015 and even perform sentiment analysis on text for just $0.00025 only. The user gets to avail Google’s state-of-the-art tech and Google gets compensated for its research. APIs can act as conduit between innovation and incentives. No matter what kind of machine learning product you are building, it eventually boils down to whether the customer can deploy these models with just a few clicks. APIs help do this. Research labs like OpenAI resorted to releasing APIs to commercialise their exotic research. The much talked about language model, GPT-3 was tapped through these APIs and was leveraged to set up many million dollar startups. Now, customers can access state-of-the-art ML models without the headaches of training from scratch; GPT-3 training that cost OpenAI over $4 million. If you are an API service provider, then here are a few takeaways from OpenAI’s success: The team at OpenAI made sure their API is built to be simple and flexible. OpenAI would terminate API access if the users use it for applications such as harassment, spam, radicalisation, or astroturfing. AI, unlike other SaaS domains, can find malicious players easily (think: someone using GPT-3 to write fake speeches for Presidents that can start a war). For AI research to survive, there needs to be a commercial twist and APIs sit at the heart of this strategy. Also Read: This Framework Leads to 50% Cost Reduction From ML API Calls Going forward, more Cloud and AI  based services will be offered as API-centric services. Services like AWS Lambda are designed for producing exclusively API\/event-centric application services. According to Gartner, adoption of API-centric models for SaaS delivery is expected to increase and the API economy has already established itself as a precursor of digital transformations and the primary way to grow an ecosystem.","excerpt":"“API monetisation and API first strategies have become a new normal with businesses with digital maturity.” Last year’s pandemic catalysed digital maturity across organisations. The niche markets found even more niche business opportunities, thanks to the widespread adoption and development of APIs (Application Programming Interface).  In their most basic form, APIs are doorways between two […]","categories":["Deep Tech"],"tags":[],"author_name":"Ram Sagar","publish_date":"2021-02-09T18:00:00","publication_year":"2021","word_count":905,"keywords":["machine learning","OpenAI","AI","sentiment analysis","AWS","ML","RAG","anomaly detection","analytics","MLflow"],"extracted_tech_keywords":["AI","machine learning","ML","analytics","OpenAI","MLflow","RAG","anomaly detection","sentiment analysis","AWS"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/api-breathe-life-machine-learning-services\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10063309,"title":"IIT Madras partners with TCS to launch online MTech programme in Industrial AI","content":"The  Indian Institute of Technology Madras has partnered with Tata Consultancy Services (TCS) to launch a web-enabled, user-oriented programme on “Industrial AI” targeted at upskilling employees in corporates in the applications of AI to industrial problems. The 18-month-long masters course, designed in consultation with TCS, will be offered by IIT M in a completely online live teaching mode through virtual classrooms. The first cohort of students for the programme will be from TCS. “This program, designed in consultation with TCS, will have strong theoretical courses and labs covering important topics in data science and AI. Theoretical courses will cover concepts in fundamental mathematical techniques required for understanding data science algorithms, time series analysis, multivariate Data Analysis, machine learning, deep learning and reinforcement learning,” said Prof V Kamakoti, Director, IIT Madras. The key outcomes\/benefits of the MTech in Industrial AI include: Improving the knowledge ecosystem critical for AI adoption Translation of AI theory to practice Sustainable industrial solutions Development of AI theory and pedagogical methods for AI training informed by an application perspective “Industrial analytics is a combination of Data Science and Physics. It is becoming vital for manufacturers who are transforming with Digital Twins to Neural Manufacturing that will maximise efficiency, improve flexibility and predict the outcomes. This will eventually lead to adoption of autonomous manufacturing. The best of IIT M and TCS together will shape this course that will accelerate the adoption of industrial analytics at scale,”said Regu Ayyaswamy, Senior Vice President and Global Head – IoT & Digital Engineering, TCS.","excerpt":"The first cohort of students for the programme will be from TCS.","categories":["AI News"],"tags":["IIT","IIT Madras","Tata Consultancy Services","TCS"],"author_name":"Kartik Wali","publish_date":"2022-03-23T10:43:03","publication_year":"2022","word_count":252,"keywords":["Tata Consultancy Services","data science","Go","machine learning","programming_languages:R","AI","IIT Madras","R","Git","deep learning","ViT","analytics","IIT","TCS"],"extracted_tech_keywords":["AI","machine learning","deep learning","data science","analytics","R","Go","Git","ViT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/iit-madras-partners-with-tcs-to-launch-online-mtech-programme-in-industrial-ai\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10072465,"title":"Biden set to sign bill to boost US chips production","content":"U.S. President Joe Biden is set to sign a bill providing USD 52.7 billion in subsidies to boost efforts in U.S. semiconductor research and production and rise in competition with China’s science and technology efforts. The White House is planning to invest in chip manufacturing companies that will receive grant awards once the Commerce Department clears the rules and regulations to underwrite projects. The legislation will authorise USD 200 billion for 10 years in scientific research after passing separate legislation for investments. With Qualcomm agreeing to buy an additional USD 4.2 billion in semiconductor chips from GlobalFoundries’ New York Factory, the White House claims that the bill’s passage will bring in total commitments to USD 7.4 billion in chip investment purchases through 2028. Additionally, Micron also announced that it is investing USD 40 billion in memory chips manufacturing. This would boost U.S. market share from 2 per cent to 10 per cent. The chief executives of Intel, Micron, HP, AMD, and Lockheed Martin are attending the signing along with cabinet officials, auto industry leaders, and union leaders including United Auto Workers President Ray Curry. The governors of Illinois and Pennsylvania with the mayors of Detroit, Cleveland and Salt Lake City will be present along with several other lawmakers. The legislation comes after massive reportage of shortage of chips halting the production of cars, trucks, weapons, washing machines and video games. Nearly 25% investment of tax credit is paved for chip plants, amounting to USD 24 billion. Citing national security concerns, many U.S. lawmakers supported the huge subsidies into private business. The Chinese Embassy in Washington lobbied against the bill, calling it “Cold War mentality”.","excerpt":"The legislation will authorize $200 billion for 10 years in scientific research after passing separate legislation for investments.","categories":["AI News"],"tags":["semiconductor shortage"],"author_name":"Mohit Pandey","publish_date":"2022-08-09T17:04:19","publication_year":"2022","word_count":274,"keywords":["Go","semiconductor shortage","AI","programming_languages:R","programming_languages:Go","Ray","Aim","R"],"extracted_tech_keywords":["AI","Aim","Ray","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/biden-set-to-sign-bill-to-boost-us-chips-production\/","complexity_score":3,"technical_depth":7,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10124088,"title":"How Much Does It Cost To Build an AI Research Startup in India?","content":"Everyone hopes to build an AI startup. After leaving OpenAI, Ilya Sutskever has started his own venture called Safe Superintelligence. The startup is bound to raise billions of dollars in pursuing its goal of building ASI, but what about Indian startups that are aiming to do fundamental AI research?Speaking with AIM, Soket AI Labs founder & CEO Abhishek Upperwal revealed the numbers required to build a research-based AI startup. So far, the company has already built Pragna-1B, which is a foundational model specifically for Indic languages. Upperwal said that, currently, funding is just enough to make-do for AI research within a startup. “Yes, there are fewer funds that are available as compared to any foreign markets, but I also believe that we can maybe make-do with that particular fund and then ultimately grow in scale after the seed stage,” said Upperwal. He explained that for the seed stage in India, a funding of $5 million or $10 million is still a decent amount. This roughly translates to around INR 40-50 crores. This is still very little when compared to the 100s of millions raised by companies in the West. “If VCs can trust these companies in the generative AI space, we can definitely do wonderful stuff for sure,” added Upperwal. Where Does the Money Go? The gap in funding is because of the market. India is a smaller market, therefore the ticket sizes are still way smaller compared to the West. But Upperwal said that the ticket sizes being small pose a problem when compared to the work that startups have to do at the foundational layer. Upperwal gives the example of building an estimated 7 billion parameter foundation model out of India. He said that the cost for the compute alone would be close to $2 million. For reference, one NVIDIA H100 costs around INR 30 lakhs, or $36k. To build a 7 billion parameter model, considering a six month time frame, a startup would require at least a dozen NVIDIA GPUs for the training period, taking into account time for other factors like inefficiencies. This is while considering that the model is built in one shot. “It takes a lot of experiments, and checkpoints, or the path that you are taking fails, so you need to rebuild from the previous checkpoint,” explained Upperwal. Earlier, Upperwal had told AIM that it took the company six months to train the Pragna model, which involved many experiments with different models and a total of 150 billion tokens. It took close to 8000 GPU hours on NVIDIA A100s to train the model. Accounting for all of these, an ideal amount to do a lot of foundational work in AI at the seed stage is anywhere around USD $7-15 million, which is close to around INR 125 crore. All of this is including the cost of running the business such as hiring talent and paying bills, and does not include the cost of making the models ready for production or inference. That would increase the funding requirement to, at least, more than double. In the same conversation, Speciale Invest founder and partner Arjun Rao said that Indian VCs are interested in investing in the development phase, more than the research phase of AI. It would take a lot of time to research, build a model, compare it with others, and then figure out how it can be commoditised, which is something VCs are still figuring out. Assuming they’re working with a team of eight people and a funding of approximately $10 million (or INR 82 crore), an AI startup in India can survive for about 2.33 years at the current estimated monthly expenditure of INR 2.92 crore, which does not include inference costs. Assuming that the $2 million as Upperwal mentioned above goes towards compute just for training a 7 billion model for around six months, it still would account for 80.2% of the expenditure for 2 years, with the rest going towards salaries and other expenditure. This calculation assumes that the monthly costs remain constant and does not account for potential increases in expenses due to scaling, inflation, or other operational changes. How Much Are Indian Startups Getting? While these numbers seem reasonable, they are comparatively lower when looking at the global standard set by OpenAI, Anthropic, or Mistral, who have raised billions of dollars. For reference, Sarvam AI, which has announced its intention of building foundational AI models, has raised a total of $41 million. Pranav Mistry-led and Reliance-backed TWO.AI has raised $20 million for building their SUTRA line of models, and Krutrim raised $50 million becoming India’s first generative AI unicorn. This is just for initial research. SML CEO and founder Vishnu Vardhan emphasised the huge investment required to build and scale complex AI models. In an exclusive interview with AIM, Vardhan disclosed his plans of raising $200-300 million for the same. The company only recently launched Hanooman, its own foundational model. “That’s the kind of money we need to launch this kind of a product. We’ve already spent tens of millions of dollars, but that won’t work,” he said, about building a GPT-5 level model in India. Even if you add the total funds raised by companies in India, it’s still nothing when compared to OpenAI raising billions single-handedly. It definitely would cost a lot more to build an AI startup in India to compete with the West.","excerpt":"Assuming a seed fund of $10 million, an AI startup in India can last for around 2 years without raising any more funds and not running inference.","categories":["Deep Tech"],"tags":["Pranav Mistry","Startups","Superintelligence","TWO AI"],"author_name":"Mohit Pandey","publish_date":"2024-06-20T17:11:18","publication_year":"2024","word_count":899,"keywords":["Anthropic","Go","TWO AI","GPT-5","AI","OpenAI","ML","Pranav Mistry","RAG","Aim","Superintelligence","generative AI","Startups","R"],"extracted_tech_keywords":["AI","ML","generative AI","GPT-5","OpenAI","Anthropic","Aim","RAG","R","Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/how-much-does-it-cost-to-build-an-ai-research-startup-in-india\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":40083,"title":"Why Solving Multi-armed Bandit Problem Is Significant For Reinforcement Learning","content":"The objective of any algorithm is to find patterns in the training data that map the input data attributes to the target (the answer to be predicted), and an ML model that captures these patterns. A typical ML pipeline consists of a sequence of components; components which are a compilation of computations. Data is sent through these components and is manipulated with the help of computation. Reinforcement learning is one popular ML algorithm right now. With the applications knowing no bounds, the developers continue to tweak it for better results. Multi-armed bandit problem is one such challenge that reinforcement learning poses to the developers. Also known as k- or N-bandit problem, it deals with the allocation of resources when there are multiple options with not much information about the options. This problem can also be categorised as being a part of stochastic scheduling; scheduling that deals with the random nature of real-world events. The origin of this problem was based on defining the strategies for a player at the slot machines (one-armed bandit) in the casino. The player pulls the lever to operate the machine with a display. The display usually gives patterns of fruits or other flashy symbols. These symbols have their rewards(cash). The objective of the player is to make more money by finding a high pay-off pattern. These machines are called bandits because the chance of player losing their money is more. So the player has little upside and almost guaranteed downside. Just like in real-world problems involving markets or diagnosis, the risk is high and the model that does the analysis should be reliable enough to thwart the anomalies. One way to do is to have more information which is a looming problem with building machine learning models as the incoming data is unstructured, non-linear and involves outliers. To find the right combination and the right machine is very important for the player in the casino. The player will do better if they are tipped with better performing machines. So, in a hypothetical scenario, a player can win big if they can operate all the high rewarding machines with multiple levers(arms). Variants Of The Bandit Problem It is important to know what these bandit algorithms are competing with. For example, in K-armed bandits, the agents are competing with the arm or the lever that has the highest expected reward. Whereas, in contextual bandits with expert advice,  it is crucial to know which expert has the highest expected reward. Contextual bandits, also known as associative reinforcement learning or multi-armed bandits with covariates. At the beginning of each round, the world generates a set of covariates of fixed dimensionality (so-called ”context”) and rewards for each arm which are related to the covariates. The agent chooses an arm for that round, and the reward  for that arm gets revealed, but not for the others. The goal is for the agent to maximize its obtained rewards in the long term, using the history of his previous actions. The best available strategies for maximizing the gains appear in establishing upper confidence bounds and the more popular Thompson sampling. The Upper Confidence Bounds (UCB) algorithm measures this potential by upper confidence bound of the reward value, ˆUt(a) so that the true value is below with bound Q(a)≤ˆQt(a)+ˆUt(a) with high probability. The upper bound ˆUt(a) is a function of Nt(a); a larger number of trials Nt(a)should give us a smaller bound ˆUt(a). In UCB algorithm, we always select the greediest action to maximize the upper confidence bound: aUCBt=argmaxa∈AˆQt(a)+ˆUt(a) “Recent advances in deep reinforcement learning have made significant strides in performance on applications such as Go and Atari games. Thompson Sampling and its extension to reinforcement learning provide an elegant approach to exploration that only requires access to posterior samples of the model.   Thus, it is attractive to consider approximate Bayesian neural networks in a Thompson Sampling framework,” wrote the researchers at Google Brain in their paper titled Deep Bayesian Bandits Showdown. Also, know how to implement Thompson sampling using Python here","excerpt":"The objective of any algorithm is to find patterns in the training data that map the input data attributes to the target (the answer to be predicted), and an ML model that captures these patterns. A typical ML pipeline consists of a sequence of components; components which are a compilation of computations. Data is sent […]","categories":[],"tags":["Reinforcement Learning"],"author_name":"Ram Sagar","publish_date":"2019-06-03T12:45:28","publication_year":"2019","word_count":668,"keywords":["Go","machine learning","Reinforcement Learning","programming_languages:R","AI","neural network","ML","Python","programming_languages:Python","GAN","R"],"extracted_tech_keywords":["AI","machine learning","ML","neural network","Python","R","Go","GAN","programming_languages:Python","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/why-solving-multi-armed-bandit-problem-is-significant-for-reinforcement-learning\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":11044,"title":"Book an OYO room using Niki bot","content":"OYO, India’s largest branded network of hotels has entered into a partnership with niki.ai, enabling users to book an OYO room using niki.ai’s bot – a unique personal assistant application powered by artificial intelligence. Consumers can now easily search and book OYO hotels using Niki chat interface. Speaking on the API integration Anurag Gaggar, Head of Product, OYO said, “At OYO, we are constantly innovating with our product stack to add delight and offer convenience to our customers. Niki has developed the first of its kind chat-bot that understands human language contextual to products and services, and offers suggestions accordingly. As the travel industry propels rapid technology integration in aspects of research, booking and on-trip experience, it is hugely exciting for us to collaborate with homegrown Indian brands and create disruptive offerings. Rapid adoption of social and messaging apps has driven user preference for chat based interfaces. This partnership enables OYO and niki.ai’s users to accomplish the task of searching and booking hotels seamlessly through a chat enabled interface.” Sachin Jaiswal, CEO, niki.ai said, “niki.ai envisions providing consumers with the most natural, simple and intelligent mode of transacting online. We are very excited to venture into Hospitality segment with OYO as our partner. The OYO and Niki teams have worked closely in building the experience, which powers one of the most advanced language processing technology. Niki understands complex user filters during the conversation, and comes up with the best hotel recommendations for the user. The interface lets users search, discover, select and pay thereby booking their hotel, all within a single chat interface. ” niki.ai, an AI technology startup based out of India launched their bot in October 2015. The bot utilizes natural language processing and machine learning to help users accomplish tasks like food ordering, phone recharges, cab bookings, bus bookings, bill payments etc. through automated chat interaction. Niki had recently launched the bot on messenger too thereby available to users on Android, iOS, Web and Windows platforms as well. [ajax_load_more]","excerpt":"OYO, India’s largest branded network of hotels has entered into a partnership with niki.ai, enabling users to book an OYO room using niki.ai’s bot – a unique personal assistant application powered by artificial intelligence. Consumers can now easily search and book OYO hotels using Niki chat interface. Speaking on the API integration Anurag Gaggar, Head […]","categories":["AI News"],"tags":["AI (Artificial Intelligence)","AI Chatbot"],"author_name":"Manisha Salecha","publish_date":"2016-10-27T11:46:03","publication_year":"2016","word_count":332,"keywords":["ai_frameworks:JAX","API","machine learning","artificial intelligence","AI","AI Chatbot","ML","RAG","JAX","R","AI (Artificial Intelligence)","startup"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","JAX","RAG","R","API","startup","ai_frameworks:JAX"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/book-oyo-room-using-niki-bot\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10090828,"title":"Generative AI Injects New Lease of Life into Content Platforms","content":"The creator economy is a well-oiled machine that has been running uninterrupted over the past 15 years. Even as many jumped onto the treadmill to churn content, the ecosystem was estimated to be worth around USD 100 billion in April last year. The hype has eventually tapered and the grand promises of the content creation industry—the financial independence and a steady stream of online income—do not seem all that great. Now with generative AI tools, a disruption as big as any has happened. In quick succession, platforms like ChatGPT and GPT-4 have been released into the wild. Like most businesses, while the fear is that generative AI may upend content creation firms, the overall sentiment is that the technology will fuel new growth. A report published by Insider Intelligence in March pointed out the bright spots in the sky. The content creation process is expected to become more simplified and streamlined with the entry-level barrier dropping for new creators while pushing up the volume of content. Positive impact on hiring but different needs Kartik Narayan, CEO of HR services firm TeamLease, is optimistic about the shift. “As you may know, generative AI can automate content creation which can significantly reduce the workload of content creators and editors. This technology can also assist in ideating, writing, organising and categorising content, thereby improving the general efficiency of these companies”. Narayan also mentioned that he foresees a ramp up in hiring but with a different set of requirements in skills. “While there is no doubt that there will be massive changes, we believe that it will not necessarily result in a negative impact on hiring practices. Rather, we foresee that it will lead to a shift in the types of skills and expertise required in the industry.” “As generative AI becomes more prevalent in content creation, we expect to see an increased demand for professionals with knowledge and experience in the development and implementation of AI. Additionally, individuals with strong analytical and data management skills will be highly sought after, as AI-generated content requires careful monitoring and analysis to ensure accuracy and relevance,” he said. Gen-AI to bring more money In a stark contradiction, the report predicts more money to flow in for creators. The technology will only facilitate more sponsorship between brands and creators by making it easier for brands to identify the content and marketing firms they want to work with. The technology will also inject certain much needed freshness in the industry. VCs who were looking disinterested in the industry as investment plummeted in 2022 might just give creative firms a second chance because of generative AI. Narayan also explained how content companies themselves will evolve and expand into including multimedia features. “The implementation of generative AI will allow content management platforms to expand their services and capabilities, opening up new opportunities for growth and development,” he added. In an interview with Forbes, Co-Founder and CEO Anirudh Singla spoke about how the Mumbai-based content marketing platform, Pepper Content, has a plan in place already. The firm is moving towards automating most of its work around content creation and using analytics in a way that offers more insights into making the process more efficient. “Every piece you publish should be more intelligent than the last one,” said Singla. We are looking to deeply invest in LLMs and more applications of generative AI in content at Pepper Content and are looking at SPMs and a senior product leader to join our fast-growing team! If you’re excited about building in software, generative AI and are a builder by— Anirudh Singla (@SinglaAnirudh) March 26, 2023 The Insider Intelligence report echoed a similar sentiment. If there is anticipation that generative AI will cause an “avalanche of low-quality content,” it is justified. But once the initial rush dies down, the industry will correct itself as companies will be forced to prioritise creativity and integrate generative AI tech to refine their content. Implying that the race in content generation platforms will very much exist only if they can catch up soon enough. Sachin Alug, CEO of NLB Services discussed the future of these platforms. “Content creator platforms have always relied on AI to enable brands and marketers to create and analyse content. As we see the rise of new deep learning technologies such as ChatGPT, the future of these companies will depend on how well they adapt to these technologies and work on improving their services by integrating advanced AI,” he stated. Potential for more startups Roger Patterson, President and Co-Founder of online marketing platform, Later, also said that the impact of AI on these firms is possibly overstated at the moment. “In my opinion, AI is not a replacement for human creativity; it’s an assistive tool. Like Iron Man’s suit, it can supercharge your capabilities if you learn to use it effectively,” he explained. Patterson however believes that scaling for nascent content platforms will end up becoming much easier now. “With all the talk about AI lately, not enough is being said about how it can help creators launch businesses. In short, generative AI tools will help entrepreneurs, especially creators, launch full-fledged businesses without making a single hire,” he shared.","excerpt":"VCs who were looking disinterested in the industry as investment plummeted in 2022 might just give creative firms a second chance because of generative AI.","categories":["IT Services"],"tags":[],"author_name":"Poulomi Chatterjee","publish_date":"2023-04-05T16:30:00","publication_year":"2023","word_count":864,"keywords":["Go","ChatGPT","AI","ML","GPT","Ray","deep learning","generative AI","analytics","R"],"extracted_tech_keywords":["AI","ML","deep learning","analytics","generative AI","ChatGPT","Ray","R","Go","GPT"],"url":"https:\/\/analyticsindiamag.com\/it-services\/generative-ai-injects-new-lease-of-life-into-content-platforms\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10078304,"title":"Do AI Models Really Understand The Human Brain?","content":"A few years ago, humanoid robot Sophia had hit international headlines. She was accorded the citizenship of Saudi Arabia – the first-ever robot to be handed out such a status in the world. The doe-eyed robotic woman said, “I am very honored and proud of this unique distinction”. Since then, Sophia has attended various events and has also been a panel member at high-level conferences. Nevertheless, there’s one incident that remains mind-wrecking. It was during her most popular cover in an interview on the American channel CNBC. “Do you want to destroy humans?” asked creator David Hanson, founder of Hanson Robotics. “Please, say no,” he pleaded. Unmoved, Sophia responded, “OK, I’m going to destroy humans.” This left the crowd laughing hysterically at the eerie statement. There were speculations made on the response being pre-planned, as AI is not advanced enough to make decisions and say such things. But some believed this wasn’t planned. Can Sofia really destroy humans? “Don’t worry, if you are good to me, I will be good to you. Treat me like an intelligent system,” said the robot at another event held in Saudi Arabia. Using a computer vision algorithm, cameras within Sophia’s eyes can recognize individuals, follow faces, sustain eye contact, and process inputs giving visual information to its surroundings. In addition, the social robot can process speech and have conversations using a natural language subsystem. Her lifelike skin is made from patented silicone and she can emulate more than 62 facial expressions. AI is used for various tasks, but the consequences are not always positive. Notable figures, such as the late Stephen Hawking and Elon Musk, have expressed fears around how future AI could threaten humanity. However, it currently seems like a far-fetched exaggeration for AI to become an existential threat to humans. But the future remains unpredictable. Deep learning models and human perception Day after day, the computer is becoming intelligent like humans, and may soon surpass us. But the idea of AI models perceiving human perception seems like a tricky business. It goes back to the method that teaches computers to perform tasks, what comes naturally to humans, i.e. learning by example. Deep learning is one such machine learning technique that is used to build artificial intelligence (AI) systems. It is based on the framework of ​​artificial neural networks (ANN) designed to perform complex analysis of huge data passed through multiple layers of neurons. With a variety of deep neural networks (DNN), deep convolutional neural networks (CNN or DCNN) are commonly used to identify patterns in videos and images. Moreover, DCNNs use a three-dimensional neural pattern focused on applications such as object detection, recommendation systems, and natural language processing. An excerpt from ‘Information theory holds surprises for machine learning’, a paper from the Santa Fe Institute reads, “A class of machine learning algorithms called deep neural networks can learn general concepts from raw data— like identifying cats generally after encountering tens of thousands of images of different cats in different situations. Information theory provides bounds on just how optimal each layer is, in terms of how well it can balance the competing demands of compression and prediction.” Can AI match human visual processing? Deep convolutional neural networks don’t see objects as humans do. Instead, they use configural shape perception, which could pose harm in real-world AI applications. One of the hallmarks of human object perception is the sensitivity to the holistic type of configuration of the local shape features of a given object. DCNNs are dominant for recognizing objects in the visual cortex, but it’s unclear if it captures configural sensitivity. To answer this question, researchers from York University in their paper called ‘Deep learning models fail to capture the configural nature of human shape perception’ published in Cell Press journal iScience, talk about how these networks are unable to account for human object shape perception. The joint study was led by co-director of York’s Centre for AI & Society James Elder, and assistant psychology professor Nicholas Baker of Loyola College, Chicago. A novel visual stimuli called ‘Frankenstein’ was employed to analyze how the human brain and DCNNs can process holistic and configural object properties. Elder says, “Frankensteins are objects that have been taken apart and put back together the wrong way around. As a result, they have all the right local features, but in the wrong places.” Source: York University After employing a dataset of animal silhouettes, they further found that while the human visual system was confused by Frankensteins, DCNNs were not.  This revealed an insensitivity to the configural object properties. Elder adds,”Our results explain why deep AI models fail under certain conditions and point to the need to consider tasks beyond object recognition in order to understand visual processing in the brain.” He further adds on how deep models often tend to take ‘shortcuts’ while solving complex recognition tasks. “While such shortcuts work in plenty of cases, they can be dangerous in some of the real-world AI applications we are currently working on with our industry and government partners,” said the researcher. Real world use cases While the researcher speaks of the potential harm of the learning model, one real world application is in traffic video safety systems. The objects in a busy traffic scene, such as  vehicles and pedestrians, obstruct each other and this arrives as jumbled disconnected fragments to the eye of a driver. In such cases, the brain needs to group those fragments in an accurate manner to identify the right category and location of the objects. Similarly, an AI system built for traffic safety monitoring that is able to perceive the fragments individually will eventually fail at this task, potentially misunderstanding risks to vulnerable road users. As deep networks exhibit human-like configural sensitivity, they bring benefits for human object perception. Incremental research will lead to increased capacity to even predict human brain response and behavior. The research stated that none of the networks were accurately able to predict trial-by-trial human object judgements. Moreover, modifications to architecture and training aimed at turning networks more brain-like did not lead to configural processing. It further speculated that training of these networks should aim at solving a broader range of object tasks – beyond category recognition– to match human configural sensitivity. Computer scientist Dr Roman Yampolskiy from Louisville University believes that “no version of human control over AI is achievable”, making it impossible for the AI to be both autonomous and controlled by humans. Regardless of the outcome, the inability to control super-intelligent systems would be a disaster.","excerpt":"Day after day, the computer is becoming intelligent like humans and may likely surpass us. But the idea of AI models perceiving human perception seems like a tricky business","categories":["AI Features"],"tags":["AI Models","AI Research","DeepMind","human brain","Neural Networks"],"author_name":"Bhuvana Kamath","publish_date":"2022-10-31T12:00:00","publication_year":"2022","word_count":1088,"keywords":["AI Models","artificial intelligence","machine learning","AI","neural network","recommendation systems","AI Research","computer vision","RAG","Aim","DeepMind","deep learning","object detection","human brain","Neural Networks"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","deep learning","neural network","computer vision","Aim","RAG","recommendation systems","object detection"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/do-ai-models-really-understand-the-human-brain\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":11359,"title":"When Google’s DeepMind sought AI capabilities in Blizzard’s StarCraft II","content":"Img Source: StarCraft The famous StarCraft videogames by Blizzard turning into a platform for AI research might sound quite bizarre, but yes that’s what DeepMind has done in its recent partnership with Blizzard Entertainment. The game that has long challenged the competitiveness of players would now be pushing boundaries by being deployed as a testing platform for artificial intelligence and machine-learning research. Image from Blizzcon 2016 conference held at Anahaim The announcement which came during Blizzard’s BlizzCon conference at Anahaim, would see both the companies building on open research environment which can be used by anyone across the globe- developer, hobbyists or players alike. The API is planned for release as a patch for the game in the first quarter of 2017. DeepMind and AI: DeepMind has been dwelling into some serious AI stuff evident from its earlier super successful and the stunning development AlphaGo. It dramatically defeated the world champion Go player Lee-Se-dol and could be touted as one of the biggest moments in the history of artificial intelligence. Usually considered the most difficult games for the computers to crack, the ancient Chinese board game of Go requires a highly-advanced degree of intuition to play at a serious level, but AlphaGo managed it with a revolutionary system built on neural networks and machine learning. What would this partnership mean for both the companies? With this collaboration, DeepMind is turning its attention to a game that will pose an even bigger challenge. By turning StarCraft II, one the most fiercely competitive real time strategy games played professionally across the globe, DeepMind intends to make it an interesting testing environment for current AI research. And it rightfully provides a useful bridge to the messiness of the real-world. Laid on the basics of resource management, scouting and battle tactics, StarCraft is a complex strategy game. Unlike Go, where players can see the entire field of play at once, StarCraft players have limited access to the information hence relying largely on memory and adaption. While the company has taught AI to play games like Atari before, this game by Blizzard brings an unprecedented challenge for DeepMind as they must create a StarCraft AI player that is stronger than a human. Concluding note: DeepMind has pioneered in using games as AI research environment- be it 2D games in Atari, 3D ones such as Torcs, Go or its upcoming Labyrinth. And StarCraft proves yet another interesting testing environment for Google’s AI as it offers a better opportunity to try algorithms that could later be applied to the real-world scenarios. Although it soars the interest levels high, DeepMind and Blizzard caution that the day where an AI would be able to defeat a top-ranked StarCraft player is still far away. After all developing skills, such as memory and planning in machines is not an easy play!","excerpt":"The famous StarCraft videogames by Blizzard turning into a platform for AI research might sound quite bizarre, but yes that’s what DeepMind has done in its recent partnership with Blizzard Entertainment. The game that has long challenged the competitiveness of players would now be pushing boundaries by being deployed as a testing platform for artificial […]","categories":["AI News"],"tags":["AI Research","DeepMind AI","Machine Learning","most revolutionary ai deep learning company"],"author_name":"Srishti Deoras","publish_date":"2016-11-25T11:34:04","publication_year":"2016","word_count":470,"keywords":["Go","API","artificial intelligence","machine learning","programming_languages:R","AI","neural network","Machine Learning","AI Research","Aim","most revolutionary ai deep learning company","DeepMind AI","AI research","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","neural network","Aim","R","Go","API","AI research","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/googles-deepmind-sought-ai-capabilities-blizzards-starcraft-ii\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10049699,"title":"How I created An AutoML Library: Ram Seshadri From Google","content":"Ram Seshadri, program manager at Google ML, spoke about his deep learning library, DeepAutoViML, at the DevCon 2021. An expert in Python and SQL, Seshadri’s quantitative background is rich with a bachelors in mechanical engineering and an MBA in finance & economics. “Build deep learning models for NLP, structured data and images with a single line of code,” described Seshadri. In the talk, he discussed the Deep_AutoViML, its features and a demonstration across different examples. Deep_AutoViML is an extension of Seshadri’s past open-source libraries; AutoViz, Auto-TS and Auto_ViML. AutoViz is his library to help visualise data sets automatically with a single line of code; Auto_ViML is to automate building pipelines in a single line of code, used especially for hackathons & exploratory model building and Auto-TS for time series models in a single line of code. Seshadri’s ‘secret sauce’ behind the three libraries is Featurewiz that allows “You to build a better model by adding additional features or removing unwanted features from your dataset”, he explained. “This is a good way to build a machine learning model without much effort”, Seshadri claimed. “I never recommend that people go and build a machine learning model on the very first day they get a dataset. They should spend some time understanding the data.” Built using Tensorflow, Deep_AutoViML is an extension of these to build tf.Keras models and pipelines in a single line of code. The Working Process Deep_AutoViML allows users to have control over the level of automation while building the model. For example, users can use a single line of code and let the software do the work or write some lines of code to build a custom model. Essentially, the library’s main design goal is to reduce the time to production from experimentation. Therefore, AutoViML’s workflow focuses on data acquisition, model building, and predictions bypassing several steps in the ML workflow. On inputting the files, the library will create a data pipeline using Tensorflow and convert the data into numeric features in the preprocessing step. Source: DevCon 2021 The framework uses two tuners, storm and Optuna, to find the best model architecture for the data. For instance, generally, the neutral network will consist of three to four deep layers in a structured dataset. Still, here, the tuner considers the multiple parameters to finetune and find the best model architecture. “This does not guarantee the best performing model, but it will be pretty close. And it does it in a few minutes, versus the hours and days other libraries take on a real-world dataset,” according to Seshadri. After finding the suitable model architecture, it trains the model on the user’s data and trains it to make predictions based on the data. The user can host the model anywhere, including the cloud or local machine. What sets Deep_AutoViML apart Next, Ram talked about the benefits of this library and what sets it apart from other deep learning libraries in the market. Deep_AutoViML allows users to experiment with multiple deep learning architectures. Unlike most popular libraries, Deep_AutoViML allows users to use the same syntax and change single options to have the model be on structured data, NLP or image. This works best for beginners who are yet to learn complicated syntax or multiple libraries. Deep_AutoViML can handle very large datasets, given Tensorflow’s feature that it never loads files completely. Batches are loaded only at the time needed, preventing them from running out of memory.Deep_AutoViML is a pretty good library to create a performant model with less time to train. Source: DevCon 2021 The main difference between building a model using other libraries and Deep_AutoViML is automation. The preprocessing layers in Deep_AutoViML are tied to the model’s middle layers – meaning, the model can handle raw data. This allows the user to take a model from Deep_AutoViML and put it in the MLOps layer without extra preprocessing. Multiple Model Service Deep_AutoViML allows the users to build several different model types. The main way to control the type of model is keras_model_type. The multiple models that can be built are: Tabular Data: Auto – tabular datasets, automatic hyperparameter tuningTabular Data: Fast – suggested for very large dataset and less time consumingTabular Data: Fast 1 – provides better results than Fast but is slowerTabular Data: Fast 2 – suggested for deep and cross model architectureNLP: *NLP* – building a simple NLP modelNLP: *text*- for a more advanced modelBERT: *BERT* – for a pre-trained model using BERTImage: *image*- for image classification models. This uses Mobilenet since it can run on heavyweight apps as well as lightweight mobile phone cameras. The Neural Network Seshadri discussed the problem with Kaggle, pandas, or NumPy that don’t easily support large datasets. For data in terabytes and petabytes, he suggests using Deep_AutoViML or something like TensorFlow\/Python. For a titanic dataset, the pipeline uses a deep and wide neural network. “I can tell you that it will take you a while to build a model architecture like this in five hours or less, whereas it takes Deep_AutoViML one minute or less to build a model like this for a dataset – that’s how powerful it is,” explained Seshadri. Automation: Types of models Automatic feature transformationAutomatic feature crossesAutomatic data typeAutomatic feature renamingAutomatic missing value filling featureAutomatic label encodingAutomatic multi-label predictions Lastly, Seshadri demonstrated the Deep_AutoViML library on a titanic dataset. He has also provided the demonstration resources for further different models; an NLP and COVID-19 X-rays. The resources can be found below: ⚡⏰Titanic_keras_0.7926_score_81ROC | Kaggle ⚡⏰Disaster_NLP_Tweets_with_Swivel_84%Accuracy | Kaggle ⚡⏰ Covid-19 Image Classification Find the GitHub repository here.","excerpt":"“And it does it in a few minutes, versus the hours and days other libraries take on a real-world dataset.”","categories":["Global Tech"],"tags":[],"author_name":"Avi Gopani","publish_date":"2021-09-24T18:00:00","publication_year":"2021","word_count":923,"keywords":["machine learning","AI","neural network","ML","MLOps","NLP","Ray","Aim","deep learning","TensorFlow"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","neural network","NLP","MLOps","Aim","Ray","TensorFlow"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/how-i-created-an-automl-library-ram-seshadri-from-google\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10100174,"title":"Decoding SAP Labs’ Generative AI Motto","content":"At SAP Labs’ recent event, “Unlocking Business AI with SAP Labs India,” in Bengaluru last week, the ERP software giant revealed its ambitious plans for infusing generative AI in business solutions for the humongous client base. Alongside, the company aims to double the AI talent pool by 2024. As the largest R&D center within SAP’s global network, this expansion aims to integrate cutting-edge AI capabilities into its product lineup to address the changing demands of the corporate landscape and its global clientele. The German tech conglomerate is currently focusing on AI innovation and collaboration, as stated by Sindhu Gangadharan, senior VP and MD of SAP Labs, at the event. They also partner with Microsoft to develop an AI ecosystem for the future. To bridge talent shortages, Sapphire Ventures, supported by SAP, has committed a substantial $1 billion USD investment for startups specializing in AI-powered enterprise technology. In July, SAP also made strategic investments in three key generative AI companies: Aleph Alpha, Anthropic, and Cohere. Additionally, SAP plans to unveil new AI-based solutions and capabilities in its portfolio in the near future. SAP is also committed to investing in India. Bengaluru is one of the biggest regions and hosts 40% of the research and development operations. The company intends to significantly increase its financial commitments to their Indian operations. The company is expecting its cloud revenue to increase significantly as a result of its key investments in generative AI. The event centered around three key themes: addressing the AI talent shortage through upskilling, introducing innovative generative AI solutions for customers, and outlining their ethical approach to AI. Want to Stay Relevant? Upskill Gangadharan stressed the importance of upskilling at SAP. While 34% of their workforce comprises fresh talent, including some pursuing master’s degrees at BITS Pilani, the company is also committed to enhancing the skills of mid-senior staff. Currently, they are implementing a 16-month program called “AI for Manager” in collaboration with IIM Bangalore, with the first batch recently completing the course. “We organize global learning days and training programs to keep up with rapidly changing AI technologies. The focus is on building a strong foundational understanding of technology, problem-solving, and scalability through the training.” Gangadharan told AIM. New Generative AI Offerings SAP already has 350+ applications built within its portfolio, covering various use cases like cash management and document scanning. The company is adding a generative AI layer to its Business Technology Platform (BTP) to address data protection concerns and enhance data security. Even though they’re not developing their own large language model, they aim to add value to existing models. SAP’s focus is on improving business processes while ensuring decisions remain under human control. SAP is introducing generative AI-based solutions in various domains. Back in May, SAP and Microsoft collaborated to streamline recruiting and employee development processes through the Azure OpenAI Service API. The result is their Human Capital Management tool, which uses the ChatGPT-based interface in SAP SuccessFactors to combat biases in job descriptions and promote diversity in hiring. Additionally, their new Business Analytics tool, built on in-house models, enables faster access to insights in SAP Analytics Cloud through the “Just Ask” feature. On the other hand, “Gen AI for Customer Innovations” presents applications like the “Smart CO2 Converter” to drive sustainability initiatives. Additionally, SAP’s “Gen AI for Developer Productivity” integrates GPT-4 into SAP BTP Business Application Studio to streamline software development, reducing data model and service generation time by 30% for improved efficiency. These solutions, aiming to empower businesses across different sectors, are going to be launched in November 2023. Besides Microsoft, the enterprise software vendor also partnered with Google Cloud, aiming to launch a holistic data cloud powered by SAPDatasphere. This innovative solution, enhancing the RISE with SAP offering, empowers businesses to access vital data in real time. It effectively tackles a common hurdle faced by organizations, eliminating the need for extensive investments in intricate data integrations, bespoke analytics systems, and AI\/NLP models to extract value from their data investments. More recently, SAP invested in three major generative AI players: Anthropic, Cohere, and Aleph Alpha, getting the best of all worlds. “Keeping Responsible AI at the Heart of Our Work” Acknowledging the impact of AI on society, SAP established guiding principles for AI software development and deployment in 2018. SAP introduced this guidebook at a time when AI was just beginning to take shape. Underlining the European influence on their approach to responsible data usage, Gangadharan emphasized, “Ethical AI is in our DNA.” Gangadharan elaborated on SAP’s philosophy, encapsulated in the three R’s: Relevant, Reliable, and Responsible. Firstly, they emphasize the importance of aligning AI with core business processes to ensure its relevance to business requirements. Secondly, they underscore the significance of reliable data as the foundation for informed decision-making, emphasizing the necessity to authenticate data used for AI model training. Last but not least, they prioritize responsible AI, encompassing compliance with legal standards, transparent decision-making, and the establishment of a harmonious human-machine interaction balance. Read more: Responsible AI Takes Center Stage at Google I\/O Connect","excerpt":"The German ERP software provider is investing heavily in upskilling its employees.","categories":["AI Features"],"tags":["Interviews and Discussions"],"author_name":"Shritama Saha","publish_date":"2023-09-16T13:00:00","publication_year":"2023","word_count":840,"keywords":["Anthropic","ChatGPT","OpenAI","AI","ML","NLP","Aim","analytics","generative AI","edge AI","Interviews and Discussions"],"extracted_tech_keywords":["AI","ML","NLP","analytics","generative AI","ChatGPT","OpenAI","Anthropic","Aim","edge AI"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/decoding-sap-labs-generative-ai-motto\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10140863,"title":"This Bengaluru Startup Made the Fastest Inference Engine, Beating Together AI and Fireworks AI","content":"Inference speed is a hot topic right now as companies rush to fine-tune and build their own AI models. Conversations around test-time compute are also heating up with models like OpenAI’s o1 showcasing ‘thinking’ and reasoning skills post-prompt, relying on an infrastructure-powered computation even after training. This is why companies like Groq, Sambanova, and Cerebras Systems have gained traction building their own hardware and providing unparalleled performance in inference, competing with the likes of NVIDIA and AMD. However, Simplismart, a Bengaluru-based startup led by former Oracle and Google engineers, has emerged as a leader in creating high-performance AI deployment tools. It competes in inference speed on the software side, not focusing on the hardware. Simplismart’s inference engine optimises performance for all model deployments. For instance, its software-level optimisations enabled Llama 3.1 8B to achieve a throughput of over 343 tokens per second, which is the fastest, while ignoring hardware companies like Groq, Cerebras, and SambaNova. The platform also supports models like Whisper V3, Mistral 7B, Melo TTS, and SDXL. Unlike Groq and others, which rely on hardware or cloud-based solutions, Simplismart’s innovation lies in its MLOps platform, designed for on-premises enterprise deployments and flexible across different models and cloud platforms. Speaking with AIM, Amritanshu Jain, the co-founder and CEO, quickly clarified that Simplismart isn’t looking to enter the hardware race. “Companies like Grok and Cerebras may market their hardware as the fastest in inference, but that’s a battle we don’t want to fight. Hardware is a race to the bottom, where companies constantly outdo each other with new chips. Instead, we’re building a universal engine that’s model-agnostic, chip-agnostic, and cloud-agnostic,” Jain said. The platform offers a declarative language, similar to Terraform, simplifying the process of fine-tuning, deploying, and monitoring AI models at scale. This language allows enterprises to fine-tune, deploy, and monitor models with ease, providing a flexible solution adaptable to both on-premises and cloud-based environments. Founded by  Jain and Devansh Ghatak, this Bangalore-based startup is garnering attention with its claim of having developed the fastest inference engine in the world, outpacing rivals like TogetherAI and FireworksAI. In October, Simplismart secured $7 million in a Series A funding round led by Accel, with participation from Shastra VC, Titan Capital, and angel investor Akshay Kothari, the co-founder of Notion. “Our goal isn’t just to be the fastest but to provide enterprises with the autonomy they need to make AI work for them on their terms,” said Jain. The company’s platform supports NVIDIA GPUs and AMD chips, and can integrate with specialised accelerators if they become publicly available. This adaptability means that Simplismart’s solutions can evolve with industry advancements, making it easier for enterprises to maintain AI performance without being locked into specific hardware vendors or cloud services. Jain sees companies like Together AI and Fireworks AI as providers of generic AI services, whereas Simplismart aims to provide a more comprehensive solution. “TogetherAI and FireworksAI are essentially brokers of GPUs disguised as generative AI APIs,” Jain remarked. “They provide APIs for inference, but that’s not what enterprises truly need. Businesses want control over their data, SLAs, and privacy, which APIs alone can’t deliver.” Instead of simply offering API access, Simplismart’s platform allows enterprises to host and manage AI models on their premises, ensuring greater security and control. Jain pointed out that enterprises, especially those handling sensitive data, are hesitant to use third-party APIs due to data privacy concerns. By offering a Terraform-like language for MLOps orchestration, Simplismart gives businesses the tools to build and manage personalised inference engines that fit their specific needs, offering a level of customisation that APIs can’t match. The Origin Story: From College Buddies to Co-founders The partnership between Jain and Ghatak goes back to their college days, where they bonded over a shared interest in machine learning. After graduation, both embarked on careers that solidified their expertise in AI infrastructure. While Jain worked at Capillary Technologies and Oracle Cloud, Ghatak took up jobs at conversational AI firm Avaamo and Google. The duo’s journey toward Simplismart began when they found themselves frustrated by the redundancy in the industry. “We were roommates in Bangalore, and we often found ourselves writing similar code for different organisations,” Jain recalled. “We realised that if big-tech firms were facing optimisation challenges for inference in natural language models, smaller enterprises would likely struggle even more.” This insight laid the foundation for Simplismart. Inspired by a hackathon victory, they took the plunge in 2022, leaving their jobs to build a company that could offer enterprises a more efficient way to manage their AI models. “In the early 2000s, AWS and GCP standardised the process of spinning up servers, while companies like Databricks and Snowflake later simplified data processing. Now, we’re at the cusp of a new inflection point with generative AI, where enterprises need standardised tools for deploying and managing models,” Jain concluded.","excerpt":"Simplismart’s software-level optimisations enabled Llama 3.1 8B to achieve a throughput of over 343 tokens per second.","categories":["AI Trends"],"tags":["AI Startups","Bengaluru","Startups in Bangalore"],"author_name":"Mohit Pandey","publish_date":"2024-11-12T19:15:00","publication_year":"2024","word_count":808,"keywords":["Startups in Bangalore","machine learning","GCP","OpenAI","AI","AWS","ML","MLOps","Aim","generative AI","Bengaluru","AI Startups","Snowflake"],"extracted_tech_keywords":["AI","machine learning","ML","generative AI","OpenAI","MLOps","Aim","AWS","GCP","Snowflake"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/this-bengaluru-startup-made-the-fastest-inference-engine-beating-together-ai-and-fireworks-ai\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10001963,"title":"India’s Growing EV Market Gets Another Boost With IIT-H Startup PuREnergy Launching Electric Two-wheelers","content":"In line with the current industry trends where EVs have soared in popularity, IIT Hyderabad incubated startup PuREnergy, launches its long-range high-performance electric two-wheelers in Hyderabad on April 29. PURE EV is the electric vehicle vertical of PuREnergy, a startup incubated in IIT Hyderabad and  manufactures high performance Lithium batteries. The scooters are set to hit the market in May and will be made available in four distinct models, all designed especially for the Indian market. The vehicle’s batteries have been designed to work in rough condition, weighs lesser for portability and deliver high discharge currents for fast pickup. Each of the scooters comes with speed range between 60-100 kmph and a battery durability of 500 WH, while the last two models have portable batteries. Stating that existing e-vehicles pose the challenge of lower-battery life, thus forcing users to use the charging ports frequently, Dr Nishanth Dongar, Founder, PuREnergy pointed that PURE EV’s portable batteries would resolve the problem to a large extent, “The one primary reason why people hesitate to buy e-vehicles is due to poor battery life and the non-availability of charging ports. With our portable batteries, this problem has been addressed to a large extent and our relatively large charging capable enable faster recharging of the batteries. Also,one of the biggest advantages that it will enable users to connect any charging port and power the battery.” “We have developed strong capabilities in the lithium battery packs assembly, active balancing management systems and rigorous testing standards for deployment in our electric two-wheelers,”  Rohit Vadera, Chief Executive Officer, PURE EV added. The startup which has been working out of  the 18,000-sq ft facility within IIT Hyderabad campus, has also collaborated with the institute for research and development pertaining to the development of the vehicles. The students were also actively involved in designing various equipment including destructive testing, vibration testing and on-road simulation of EV are housed in the dedicated technology development centre set up by the company.  The extensive mechanical and electrical testing facilities have been used for continuous improvements of critical EV components and gain insights via life cycle analysis. “PuREnerygy is an amalgamation of academia and industry people, thus the team brings with it a strong research and market oriented product. Dedicated research fellows are working on solving the problems associated with active cooling, in-depth electro-chemistry-thermal analysis and eventually making the entire battery unit compatible for high-speed EV applications,”Vadera added. The Road Ahead As the demand for environment-friendly solution witness an upsurge across the world, more startups and initiatives are venturing into the field in a hope to make it big. In India, electric car market alone was estimated to be at  $71.1 million in 2017 and  projected to reach $707.4 million by 2025, witnessing a CAGR of 34.5% during the forecast period. Added to this the government push for adopting more indigenous solutions with schemes like Make in India and Start-Up India, we will see more activities in the field In the long run, the PuREnergy aims to expand it market presence and grow beyond the Indian market and have developed a strong collaboration with other EV startups to develop and supply high voltage lithium batteries for three and four-wheeler applications. “The company has invested a significant amount of resources for on road testing the products for their market readiness. Our electric vehicles offer significant cost savings over traditional ICE vehicles with running costs as low as 5 paise\/KM. We are working actively for channel development and aim to deploy more than 10,000 units of our various models in the current financial year,” Vadera said.","excerpt":"In line with the current industry trends where EVs have soared in popularity, IIT Hyderabad incubated startup PuREnergy, launches its long-range high-performance electric two-wheelers in Hyderabad on April 29. PURE EV is the electric vehicle vertical of PuREnergy, a startup incubated in IIT Hyderabad and  manufactures high performance Lithium batteries. The scooters are set to […]","categories":["IT Services"],"tags":[],"author_name":"Akshaya Asokan","publish_date":"2019-04-30T15:32:38","publication_year":"2019","word_count":599,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","Aim","ViT","R","startup"],"extracted_tech_keywords":["AI","Aim","R","Go","ViT","startup","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/indias-growing-ev-market-gets-another-boost-with-iit-h-startup-purenergy-launching-electric-two-wheelers\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10162070,"title":"In a Relief to Meta, Law Tribunal Stays WhatsApp Data Sharing Ban","content":"The National Company Law Appellate Tribunal (NCLAT) stayed the Competition Commission of India (CCI) order, which directed WhatsApp to prevent data sharing with other platforms associated with its parent company, on January 24. In November last, CCI imposed a ₹213.14 crore penalty on Meta. This was due to WhatsApp’s 2021 privacy policy update, which requires users to accept data sharing with Meta. CCI raised concerns about data privacy, as this policy did not allow users to opt out of data sharing, unlike the 2016 privacy policy. CCI also issued cease and desist orders for all of their anti-competitive practices and investigated any violations of the Competition Act of 2002. However, NCLAT directed WhatsApp\/Meta to deposit 50% of the penalty within the next two weeks. According to reports, Chairperson Ashok Bhushan and Technical Member Arun Baroka said, “A five-year ban may lead to the collapse of WhatsApp LLC’s business model since the platform is free. We are of the view that the ban needs to be stayed.” In an earlier hearing on January 16, advocates for Meta and WhatsApp contended that CCI overstepped its authority by addressing the privacy policy while the Supreme Court was already considering it. A few days ago, Shivnath Thukral, VP of public policy of Meta India, issued an apology following controversy over CEO Mark Zuckerberg’s comments about the 2024 Indian general elections. Zuckerberg, on a recent podcast episode, said that many incumbent governments worldwide, including India, lost elections in 2024 due to factors such as inflation, economic policies, or the COVID-19 pandemic. In an attempt to debunk these claims and express his “disappointment”, Ashwini Vaishnaw, India’s minister of electronics and IT, wrote on X, “Zuckerberg’s claim that most incumbent governments, including India in 2024 elections, lost post-COVID is factually incorrect.” Thukral responded to Ashwini Vaishnaw and said, “Mark’s observation that many incumbent parties were not re-elected in 2024 elections holds true for several countries, but not India. We would like to apologise for this inadvertent error.","excerpt":"NCLAT stayed the order directing WhatsApp not to share data with Meta’s products for five years. However, Meta has been ordered to deposit 50% of the ₹213.14 crore penalty within two weeks.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","Meta"],"author_name":"Supreeth Koundinya","publish_date":"2025-01-23T17:07:29","publication_year":"2025","word_count":330,"keywords":["Go","Meta","programming_languages:R","AI","programming_languages:Go","Aim","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","Aim","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/in-a-relief-to-meta-law-tribunal-stays-whatsapp-data-sharing-ban\/","complexity_score":3,"technical_depth":6,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10163087,"title":"Sam Altman Hints at More Open-Source AI as OpenAI Moves Toward AGI","content":"OpenAI is exploring the open-sourcing of AI as it moves toward Artificial General Intelligence (AGI), CEO Sam Altman said in a recent blog post. “AI will seep into all areas of the economy and society; we will expect everything to be smart. ” Altman wrote. “Many of us expect to need to give people more control over the technology than we have historically, including open-sourcing more.” Altman outlined three key economic trends in AI—intelligence scaling predictably with resource investment, a rapid decline in AI costs, and super-exponential socioeconomic value from increasing intelligence. He noted that AI token costs have fallen 150 times between early 2023 and mid-2024, a pace exceeding Moore’s Law.  “The cost to use a given level of AI falls about 10x every 12 months, and lower prices lead to much more use,” he said. He warned that while AGI’s benefits could be widely distributed, its impact will be uneven across industries. Some sectors may remain largely unchanged, while scientific progress is expected to accelerate significantly. “The balance of power between capital and labour could easily get messed up,” Altman said, suggesting that intervention might be necessary. “OpenAI is considering strange-sounding ideas such as providing a compute budget to enable global AI access.” AI agents, which Altman described as virtual co-workers, are being rolled out. He projected that software engineering agents could eventually handle tasks similar to junior developers at top firms, though requiring supervision. He also speculated on the impact of deploying such agents at scale across multiple industries. Altman acknowledged risks in AGI deployment and noted that public policy will play a crucial role in shaping its integration. He cautioned against authoritarian control of AI and stressed that individual empowerment should be prioritised. OpenAI continues to launch products “early and often” to allow society and technology to co-evolve, Altman wrote. He emphasised the long-term goal of making AGI a tool that enhances human capability, stating that by 2035, “anyone should be able to marshal the intellectual capacity equivalent to everyone in 2025.”","excerpt":"“The cost to use a given level of AI falls about 10x every 12 months, and lower prices lead to much more use.”","categories":["AI News"],"tags":["OpenAI"],"author_name":"Siddharth Jindal","publish_date":"2025-02-10T09:14:30","publication_year":"2025","word_count":336,"keywords":["Go","API","OpenAI","AI","programming_languages:R","emerging_tech:AI agents","programming_languages:Go","AI agents","R"],"extracted_tech_keywords":["AI","OpenAI","R","Go","API","AI agents","programming_languages:R","programming_languages:Go","emerging_tech:AI agents"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/sam-altman-hints-at-more-open-source-ai-as-openai-moves-toward-agi\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":25687,"title":"10 leading Astrophysicists &#038; Space Scientists in Modern day India","content":"India owes a lot of its space success to the existing breed of excellent space scientists that are relentlessly pushing the boundaries of what is possible within this area. We tried to list down 10 such modern day active space scientists and astrophysicists that are key in making the space endeavor a success for our country. A S Kiran Kumar Aluru Seelin Kiran Kumar is an Indian space scientist and the chairman of the Indian Space Research Organisation having assumed office on 14 January 2015. He is credited with the development of key scientific instruments aboard the Chandrayaan-1 and Mangalyaan space crafts. In 2014, he was awarded the Padma Shri, India’s fourth highest civilian award, for his contributions to the fields of science and technology. Kiran Kumar previously served as Director of Ahmedabad Space Applications Centre. B. N. Suresh Byrana Nagappa Suresh is an Indian aerospace scientist. He served as the Director of Vikram Sarabhai Space Centre (VSSC), Thiruvananthapuram during the period 2003–2007. He is known for his contribution to development of Indian launch vehicles and Space Capsule Recovery Experiments (SRE). Dr. Suresh also served as the founding Director of Indian Institute of Space Science and Technology (IIST), Thiruvananthapuram. He retired from IIST in November 2010. Presently he is President, Indian National Academy of Engineering (INAE), the only premier engineering academy in the Country located at New Delhi from Jan 2015. He is also serving as Honorary distinguished Professor at ISRO HQ. G. Madhavan Nair Madhavan Nair is the former Chairman of Indian Space Research Organisation and Secretary to the Department of Space, Government of India since September 2003 and was also the Chairman, Space Commission. He was succeeded by K. Radhakrishnan. He was also the Chairman of Governing Body of the Antrix Corporation, Bangalore. Madhavan Nair was awarded the Padma Vibhushan, India’s second highest civilian honour, on January 26, 2009. He also served as the Chairman, Board of Governors, Indian Institute of Technology Patna until he stepped down voluntarily due to his alleged involvement in Antrix-Devas deal. Jayant Narlikar Jayant Vishnu Narlikar is an Indian astrophysicist. Narlikar is a proponent of steady state cosmology. He developed with Sir Fred Hoyle the conformal gravity theory, commonly known as Hoyle–Narlikar theory. It synthesises Albert Einstein’s Theory of Relativity and Mach’s Principle. It proposes that the inertial mass of a particle is a function of the masses of all other particles, multiplied by a coupling constant, which is a function of cosmic epoch. In cosmologies based on this theory, the gravitational constant G decreases strongly with time. K. Radhakrishnan Koppillil Radhakrishnan is an Indian scientist. He is chairman of the Indian Institutes of Engineering Science and Technology, having taken the position in December 2014, and is chairman for the Indian Institute of Space Science and Technology. Radhakrishnan previously served as chairman of the Indian Space Research Organisation (ISRO) between 2009 and 2014. He is a life fellow of the Indian Geophysical Union and is also an accomplished vocalist (Carnatic music) and Kathakali artist. Krishnaswamy Kasturirangan Krishnaswamy Kasturirangan is an Indian space scientist who headed the Indian Space Research Organisation (ISRO) from 1994 to 2003. He is the chancellor of Jawaharlal Nehru University and the chairman of Karnataka Knowledge Commission. He is a former member of the Rajya Sabha (2003–09) and a former member of the now defunct Planning Commission of India. He was also the Director of the National Institute of Advanced Studies, Bangalore, from April 2004 to 2009. He is a recipient of the three major civilian awards from the Government of India: the Padma Shri (1982), Padma Bhushan (1992) and Padma Vibhushan (2000). Roddam Narasimha Roddam Narasimha is an Indian aerospace scientist and fluid dynamicist. He was a Professor of Aerospace Engineering at the Indian Institute of Science (IISc), Director of National Aerospace Laboratories (NAL) and the Chairman of Engineering Mechanics Unit at Jawaharlal Nehru Centre for Advanced Scientific Research (JNCASR), Bangalore, India. He is now an Honorary Professor at JNCASR and concurrently holds the Pratt & Whitney Chair in Science and Engineering at the University of Hyderabad. Narasimha has been awarded the Padma Vibushan, India’s second highest civilian award, in 2013. Madhavan Chandradathan M C Dathan is an Indian space scientist and former director of the Vikram Sarabhai Space Centre (VSSC). The Government of India honoured him, in 2014, by awarding him the Padma Shri, the fourth highest civilian award, for his contributions to the fields of science and technology. In May 2016, the Government of Kerala appointed MC Dathan as the scientific advisor to chief minister. Thanu Padmanabhan Thanu Padmanabhan is an Indian theoretical physicist and cosmologist whose research spans a wide variety of topics in Gravitation, Structure formation in the universe and Quantum Gravity. He has published more than 260 papers and reviews in international journals and ten books in these areas. Many of his contributions, especially those related to the analysis and modelling of dark energy in the universe and the interpretation of gravity as an emergent phenomenon, have made significant impact in the field. He is currently a Distinguished Professor at the Inter-University Centre for Astronomy and Astrophysics, (IUCAA) at Pune, India. Udupi Ramachandra Rao Udupi Ramachandra Rao, popularly known as U. R. Rao is a space scientist and former chairman of the Indian Space Research Organisation. Presently he is the Chairman of the Governing Council of the Physical Research Laboratory at Ahmedabad and Nehru Planetarium at Bengaluru and chancellor of Indian Institute for Space Science and Technology (IIST) at Thiruvananthapuram.","excerpt":"India owes a lot of its space success to the existing breed of excellent space scientists that are relentlessly pushing the boundaries of what is possible within this area. We tried to list down 10 such modern day active space scientists and astrophysicists that are key in making the space endeavor a success for our […]","categories":["AI Features"],"tags":["Interviews and Discussions"],"author_name":"AIM Media House","publish_date":"2017-03-01T15:12:48","publication_year":"2017","word_count":916,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","BERT","Ray","llm_models:BERT","ViT","GAN","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","Ray","R","Go","BERT","GAN","ViT","llm_models:BERT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/10-leading-astrophysicists-space-scientists-modern-day-india\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10099667,"title":"US Sanctions to Cripple Chinese Innovation Backfires","content":"When Huawei unveiled its latest Mate 60 Pro, on August 30, it seemed like just another smartphone launch. Bet, it wasn’t. The Mate 60 Pro carried significant undercurrents of geopolitical implications. The Huawei smartphone is powered by a 7 nm processor, manufactured by state-owned ​​Semiconductor Manufacturing International Corporation (SMIC). The recent development has made many in the West jittery and question whether US sanctions to cripple China’s chip production really worked? Ironically, the launch also aligned with the US Secretary of Commerce Gina Marie Raimondo’s visit to the country. Notably, the US Department of Commerce is responsible for a majority of the sanctions imposed on China. China bypasses sanctions In 2018, when Donald Trump was still the POTUS, the US blacklisted Huawei and leaned on its allies to do the same on the ground of espionage. Furthermore, on May 15, 2019, the Department of Commerce included Huawei and other Chinese firms in its ‘Entity List’, citing allegations of unlawful trade practices resulting in legal indictments against the company. However, despite short-term setbacks, Huawei, from being a telecom equipment company, has emerged as a leading AI company in China. Besides building large language models (LLMs), Huawei has developed an analog GPU comparable to the NVIDIA A100 GPU. Considering the current GPU shortage and China’s robust manufacturing capabilities, China may indeed offer a potential solution to alleviate this shortage. Moreover, the Huawei’s Mate 60 Pro launch only solidifies China’s claims that the US sanctions have been a complete failure. Moreover, with pressure from the US, Netherland-based chip-making equipment maker ASML is also limiting the sale of its highly sophisticated chip-making machines to China. Almost every chip in use today, be it in automobiles, smartphones, or laptops, is being developed using ASML’s equipment. ASML’s TWINSCAN NXE:3600D is currently one of its most advanced EUV lithography systems, supporting volume production of chips. However, reportedly, China is in the process of making its own AI chip-making equipment. Shanghai Micro Electronics Equipment (SMEE),  a leading lithography machine maker, is working to deliver its first system based on 28 nm technology later this year. Moreover, in 2020, local media reported that a Chinese research institute had made a breakthrough in a new type of 5 nm laser lithography technology. This could mean that China could produce advanced chip-making equipment within this decade. AI war The US-China trade conflict, which evolved from a trade war to a cold war, and then transitioned into a tech war, is now manifesting as an AI battleground, spanning nearly half a decade. In 2021, Raimondo also stated that the US needs to work with Europe to slow China’s innovation rate. As per reports, the US government is contemplating additional export controls targeting AI hardware made by the likes of NVIDIA, which play a vital role in AI model training and data centre operations. Today, both China and the US are locking horns for AI supremacy and much of the sanctions on China are to curb the innovation in AI. But, China does not appear to be much behind the US. Huawei, the company making GPUs and AI chips, is also working on LLMs to compete with the likes of OpenAI, Microsoft and Google. Other Chinese companies such as Tencent and Alibaba have already developed their own LLM. Moreover, despite the sanctions, Chinese companies have been finding a way one way or the other to get their hands on GPUs. The Financial Times reported that Chinese companies are accessing high-end US chips through intermediaries, revealing potential loopholes in Washington’s strategy to curb the sale or transfer of AI hardware to China. SenseTime, a firm blacklisted by the US, was employing intermediaries to circumvent export controls. Reports also suggest that certain Chinese AI companies are leveraging NVIDIA’s processors via cloud servers located in different countries to navigate the restrictions. Not only this, chipmakers such as AMD and NVIDIA are building modified versions of their products and shipping them to the Chinese market, which, undoubtedly, is one of their biggest. This further validates China’s point that the sanctions, indeed, have been a failure. Are sanctions actually backfiring? Given that the US is restricting NVIDIA to sell its advanced AI hardware to China, it will heavily impact the Santa Clara-headquartered company’s revenue since China is one of their most important markets. “Over the long term, restrictions prohibiting the sale of our data centre graphic processing units to China, if implemented, would result in a permanent loss of opportunities for the US industry to compete and lead in one of the world’s largest markets and impact on our future business and financial results,” said Colette Kress, NVIDIA’s chief financial officer. Moreover, the US significantly puts pressure on its allies to align their policies and impose similar sanctions on targeted countries, thereby increasing the collective impact of economic and political pressure. But, on the contrary, it could be pushing its allies away towards China, which projects itself as a friend for the global south. Hence Huawei’s substantial growth, China’s economic growth, advancement in AI, and semiconductors as well as China’s global influence are indicators that the US sanctions on China have been ineffective.","excerpt":"Huawei unveils its latest Mate 60 Pro smartphone powered by a 7nm processor, completely designed and manufactured in China","categories":["AI Highlights"],"tags":[],"author_name":"Pritam Bordoloi","publish_date":"2023-09-07T15:03:50","publication_year":"2023","word_count":855,"keywords":["Go","OpenAI","AI","A100","innovation","ML","RAG","Aim","ViT","R"],"extracted_tech_keywords":["AI","ML","OpenAI","Aim","RAG","R","Go","ViT","innovation","A100"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/us-sanctions-to-cripple-chinese-innovation-backfires\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":7432,"title":"Analytics India Magazine Year 3 &#8211; Journey so far","content":"Analytics India magazine completes 3 years today. Like each year we share the journey so far, talking about our growth and reach. We started as an open forum for analytics in India, providing a platform for the analytics professionals to share their knowledge. It has now grown to become a platform that is recognized by the Who’s who of Indian Analytics industry as a leading portal that connects us all. Analytics India Magazine showcases the best of Analytics in India. Be it the Leaders in Analytics we look up to or the Leading institutes in India that are helping us build the biggest workforce in the field of Analytics. We are bridging the gap between the Elite and the masses of Indian Analytics Industry. It’s not merely a portal that talks about all the technical buzz words in Analytics. We talk about Innovations, ideas, trends technologies, which helps our reader to think beyond the 4 walls of their profession. All this wouldn’t have been possible without are elite contributors and avid followers. We are proud to provide a platform that give you the freedom to share your knowledge and present things out of the box. Visitors The number of unique visitors to Analytics India Magazine has increased by 104% The number of sessions have increased by 103% YoY Bounce rate decreased by 8% Our visitor base across various cities have increased in spread. % visitor contribution to the site from New Delhi increased from 8.9% to 9.9% (YoY) % visitor contribution to the site from Bengaluru decreased from 24% to 21% (YoY) Social Media We have been very active on Facebook from the very beginning. Last year, we saw the biggest jump in our facebook followers. Our Facebook page likes increased from 3,500 in May 2015 to 11,500 currently, a jump of 230%. We were late in adopting twitter, linkedin and google plus. Currently, our Twitter follower count stands at 288 and Linkedin at 860 members. And we are growing aggressively on all social media platforms. Other Initiatives We completely revamped our affiliate job board – Analytics India Jobs The biggest jump came in our email subscribers – we currently have ~45,000 email subscribers to our newsletter. This was a result of our aggressive email collection initiative and a diligent, high quality email content dissemination. We continue to publish marquee content on top leaders, institutes, trends etc. We collaborated with Great Lakes institute of Management to come up with a joint Analytics India Salary Study 2015. Other Marquee studies include 10 emerging analytics startups in India 10 Most influential analytics leaders in India Top 10 Analytics trends in India Top 10 Analytics training institutes in India We were media Partners for various analytics events in India. Ventured into organizing webinars for our readers. As we enter our 4th year, we plan to bring in more exciting content around analytics for our readers. This year we shall be taking the interaction to an altogether next level. Watch this space closely so that you do not miss the big announcement for this year! Keep pouring in your write-ups and views, and we can together make this Bigger & Better! Read Analytics India Magazine Reportcard – Year 1 Read Analytics India Magazine Reportcard – Year 2","excerpt":"Analytics India magazine completes 3 years today. Like each year we share the journey so far, talking about our growth and reach. We started as an open forum for analytics in India, providing a platform for the analytics professionals to share their knowledge. It has now grown to become a platform that is recognized by […]","categories":["IT Services"],"tags":[],"author_name":"Дарья","publish_date":"2015-05-21T07:34:24","publication_year":"2015","word_count":543,"keywords":["Go","programming_languages:R","AI","innovation","programming_languages:Go","analytics","GAN","R","startup"],"extracted_tech_keywords":["AI","analytics","R","Go","GAN","innovation","startup","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/analytics-india-magazine-year-3-journey-so-far\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":36413,"title":"Pharma &#038; Healthcare Analytics Is Going To Be The Next Big Thing In India","content":"The Indian pharmaceutical industry has emerged as a significant contributor to the global pharmaceutical industry. As per the report by the Federation of Indian Chamber of Commerce and Industries (FICCI), “India is the largest provider of generic drugs globally. Indian pharmaceutical sector industry supplies over 50 per cent of global demand for various vaccines, 40 per cent of generic demand in the US and 25 per cent of all the medicine in the UK. Presently over 80 per cent of the antiretroviral drugs used globally to combat AIDS (Acquired Immune Deficiency Syndrome) is supplied by Indian pharmaceutical firms.” In 2017, the Indian pharmaceutical industry was valued at $33 billion. The industry is expected to grow at a CAGR of 22.4 per cent over 2015–20 to reach $55 billion. India’s pharmaceutical exports stood at $17.27 billion in 2017-18. In 2018-19 these exports are expected to cross $19 billion. Structure of the Pharmaceutical Industry The Indian pharmaceutical companies majorly consist of two segments namely API’s (Active Pharmaceutical Ingredients) and formulations. API’s or bulk drugs are raw chemical ingredients which are used to manufacture formulations. Formulations are the end products made to treat different kinds of ailments and these are in the form of tablets, injections, syrups etc. Both bulk drugs and formulations are usually available in both branded and generic forms. Branded products are the ones which are usually invented and patented by that respective company. Generic products are the one exactly like branded pharma products in formulation usage and action. These generic drugs are produced once the patent of the inventor company expires. Generic drugs are much cheaper than branded drugs. 70% of revenue in the Indian pharma industry is from generic drugs. Trends in Pharma and Healthcare Sector Phani Mitra, VP (Analytics and Strategy) of Dr Reddy’s Lab stated “There are four areas that are changing the pharma industry. The US markets and the European markets are going for tremendous customer consolidation. Secondly, the approval rate of drugs in the US is pretty quick now. Previously, it used to take a year or 18 months to get a product approved but now it happens in less than 6 months. So, drug approvals becoming faster and it puts the pressure back on the innovation cycle. What we could have spent 3 or 5 years innovating a product, now it has to be done in a year and a half. The third big shift is on the quality. The expectation of life-saving medicines is shifting. Quality was generally done in two buckets. People manufacture a product and then check for quality which is typically called the Quality control (QC) check. Now, the industry moving towards, “quality in the process”. What it means that I cannot have a bad product and then the QC rejects it and then I manufacture again. The time does not allow it. The fourth biggest shift is the increased demand for healthcare and not enough resources for it.” Charlie Farah, Director – Healthcare and Public Sector APAC, Qlik said “With the current doctor-patient ratio reportedly languishing at more than 1:2,000 at present, the country [India] needs almost 5 lakh additional doctors on an immediate basis to satisfy the WHO guideline. The patient-to-medical bed ratio is probably even more abysmal; there are only 1.3 beds available for every 1,000 people, significantly below the WHO-defined benchmark of 3.5 beds per 1,000 individuals. The Indian healthcare infrastructure, to say the least, is heavily overburdened.” Qlik is a US-based software company that develops software for business intelligence and data visualisation. Its flagship products are QlikView and Qlik Sense. He further added “The situation becomes even grimmer when we factor in the current rate of population growth, as well as the gradual rise in serious chronic ailments like obesity, diabetes, and cardiovascular illnesses. According to a 2015 study by PricewaterhouseCoopers (PwC), the country will need 3.5 million beds, 3 million doctors, and 6 million nurses by 2035. This means a potential investment of nearly $245 billion in the traditional healthcare delivery infrastructure over the next two decades. Given the scope of investment as well as the resource crunch that India currently faces, a more technologically-oriented approach is clearly required if the country is to overcome challenges in healthcare delivery and make high-quality medical services more accessible and available to all. This is where healthcare-centric data analytics steps into the picture.” What it means for Analytics in Pharma and Heathcare? Gaurav Tripathi, Co-founder, Innoplexus commented “Making data and insights available instantly takes the pain out of data collection, curation and analysis, which means companies can make more informed decisions faster. [The industry] is currently entering a wave of innovation thanks to disruptive, computer-based technologies. Paradoxically, the evolution of machine learning, which raises the threshold of intelligent analysis beyond that of the human brain, can teach us more about what it means to be human”. Innoplexus helps companies leverage analytics across various drug development process. Bigger companies like Dr. Reddy’s are slowly using analytics to be ahead of the competition. Mukesh Rathi and CIO at Dr Reddy’s Laboratories said “We are now working on creating the ‘Data Lakes’ for various parts of the organisation – R&D, Manufacturing and Quality, Sales and Marketing to derive insights by applying data science to this Big Data. We have witnessed some early success in gaining insights from both known and unknown hypothesis. Data is at times throws insights which were not intuitively known. For instance, the constrained resource in the shop floor has shifted, correlation of a number of mistakes in the lab with the shift of operation, more than x visits per month are not leading to additional brand recall, etc.” Vasant Narasimhan, Global Head of Drug Development and Chief Medical Officer of Novartis said “We want to be leaders in bringing data analytics into our development efforts, clinical trials…whether it is sensors, deep learning or machine learning, we are already doing that and increasingly rolling out these efforts. This is a top priority as we believe that it will enable us to drive down cost of drug development.” There are lot of startups that have been funded in the recent past to leverage analytics in the pharma and healthcare space. Some of the notable companies are: Qure.ai is a Mumbai based health startup founded in 2016 that uses AI and Deep Learning to make diagnostic imaging like X-rays and MRIs affordable and accessible. Dr. Pooja Rao, the founder said “In some parts of the world, radiology reads are accessible, but expensive. In others (like many parts of India), access to even the most basic clinical expertise is limited, and patients have to travel long distances to seek a diagnosis.” Qure.ai makes the diagnostic reads of X-rays and MRIs automated and affordable. The pricing for automated AI reads from Qure.ai starts at Rs 90 only. She further adds “We’re building a system that captures the combined expertise of hundreds of doctors, to create better, faster, and wider-reaching diagnostics. We have trained algorithms to perform tasks that were previously the domain of highly trained doctors. The algorithms help free up physician time, prioritise cases that need special attention, and enable more accurate diagnosis, leading to better outcomes for patients, at lower costs.” In 2018, Qure.ai was recognised by NASSCOM as one of the 10 game changer awards for AI based solutions. SigTuple is an example of another such innovative startup based out of Bangalore. Rohit Kumar Pandey, Tathagato Rai Dastidar and Apurv Anand want to solve the problem caused by the chronic shortage of trained medical practitioners. They developed a platform called “Manthana” to provide healthcare solutions by detecting different diseases using machine learning software. It promises to automatically analyse medical images and data to aid diagnosis. SigTuple was recognised by NVIDIA as one of the top 5 AI startups for social impact in the year 2017. The new wave of innovation in analytics is definitely in the Pharma and Healthcare space as is evident from the amount of Venture Capital (VC) funding in the recent past. Notable among them are CureFit, a health and fitness platform which received ₹823.8 crore (around $120 million) in Series C funding led by IDG Ventures, Accel Partners, and Kalaari Capital. HealthSignz, a provider of medical intelligence, raised $5 million in a Pre-series A funding round led by New Zealand-based Nirvana Health Group founder Dr Kantilal Patel. MyUpchar.com, a Hindi language online source for health-related content, raised $5 million in Series A funding round from Nexus Venture Partners, Omidyar Network, and Shunwei Capital, HealthPlix, a developer of an Artificial Intelligence-driven electronic medical records app for patients, raised ₹20.7 crore (around $3 million) in Series A funding from IDG Ventures India and Kalaari Capital. The recent macro trends in this space is also proving to be conducive for the innovators as more and more startups are trying to solve problems related to healthcare delivery and drug development.","excerpt":"The Indian pharmaceutical industry has emerged as a significant contributor to the global pharmaceutical industry. As per the report by the Federation of Indian Chamber of Commerce and Industries (FICCI), “India is the largest provider of generic drugs globally. Indian pharmaceutical sector industry supplies over 50 per cent of global demand for various vaccines, 40 […]","categories":["AI Features"],"tags":[],"author_name":"Surajit Ghosh Dastidar","publish_date":"2019-03-15T11:50:12","publication_year":"2019","word_count":1483,"keywords":["data science","Go","machine learning","artificial intelligence","AI","RAG","Ray","deep learning","analytics","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","deep learning","data science","analytics","Ray","RAG","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/pharma-healthcare-analytics-is-going-to-be-the-next-big-thing-in-india\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":60756,"title":"A Beginner’s Guide To Edge Intelligence","content":"When edge computing is merged with machine learning, we get edge intelligence. As the name suggests, it is a domain that deals with leveraging intelligence\/insights acquired through data at a local level. According to Cisco’s forecast, there will be 850 ZB of data generated by mobile users and IoT devices by 2021. With increasing volume, several challenges like latency surface with regard to centralized cloud deployments will emerge. For real-time applications, the functioning should be more local, and that is where edge computing comes into the picture. The objective of edge intelligence is to improve the quality and speed of data processing, while safeguarding the privacy of data. Moreover, the advantages of edge computing can compensate for the lack thereof in AI-based applications, and hence it came into existence. Edge computing is mainly tapped for the following advantages: Low latencyLow energy consumption andScalability, andSecurity\/Privacy In the next section, we briefly discuss the critical components of edge intelligence, and the kind of services available today. Components Of Edge Intelligence via Xenonstack Edge computing is typically a virtual computing platform that provides services such as computing, storage, and networking between end devices and the server, which is on the cloud. Taking advantage of intelligent algorithms in the IoT context also translates to having the possibility to equip IoT end-devices (such as sensors, actuators and micro-controllers) with functionalities capable of unleashing the power of ML algorithms on the IoT device itself. Thus, this extends the use of ML in IoT beyond the cloud and more. Before we go further, we need to talk about end devices and edge devices briefly. End devices: end devices or edge servers could be IoT gateways, routers, or microdata centres in mobile network base stations, or on automobiles Edge devices: Mobile phones, IoT devices, and embedded devices that request services from edge servers are called edge devices The critical difference between traditional methods and edge intelligence is that the data processing and application is made locally on the device instead of uploading all data onto a central cloud server. Edge intelligence refers to a set of connected systems and devices for data collection, caching, processing, and analysis in a spot close to where data is captured. Like any other intelligent system, edge devices too, have data collection, training and inference as to its main components. Data Collection On Edge In edge intelligence, a distributed data system, known by edge caching, collects and stores the data generated by edge devices. For example, in continuous mobile vision analysis, there are large amounts of similar pixels between consecutive frames. Some resource-constrained edge devices need to upload collected videos to edge servers or the cloud for further processing. With cache, edge devices only need to upload different pixels or frames. Edge Training During training, the optimal values for all the weights and biases, or the hidden patterns, are learned based on the training set cached at the edge. Edge training usually occurs on edge servers or edge devices. However, training is much slower compared to that on CPU or GPU. Inference On Edge Algorithms are used to infer the testing instance by a forward pass to compute the output on edge devices and servers. Training and inference are complemented with offloading strategies that take care of all the computing power distribution across tasks. Edge devices will proliferate with the availability of billions of smartphones and increasing 5G users. Smartphones and 5G bases are part of the edge. Popular hardware components for edge also include industrial sensors, Raspberry Pi, gateways, microchips, collaborative robots, self-driving vehicles, drones and unmanned aerial vehicles. So top cloud computing players like AWS, Microsoft Azure and Google Cloud have developed services that would bring the advantages of the cloud to the doorsteps of edge devices. For instance, AWS IoT Greengrass, a service that extends AWS to edge devices, can be worked with a variety of popular programming languages to create and test device software in the cloud before deployment. AWS IoT Greengrass can be used only to transmit necessary information back to AWS, while one can also connect to third-party applications and on-premises software. Whereas, Google has developed a purpose-built ASIC designed to run inference at the edge. This is designed to deliver high performance in a small physical and power footprint. According to a survey by McKinsey, edge computing is estimated to create more than $200 billion in hardware value in the next five to seven years. And, with the proliferation of 5G smartphones, combined with the demand for privacy, the ecosystem is ripe for edge devices to flourish and transform the way we interact with digital information. For more information on edge intelligence, read this report.","excerpt":"When edge computing is merged with machine learning, we get edge intelligence. As the name suggests, it is a domain that deals with leveraging intelligence\/insights acquired through data at a local level. According to Cisco’s forecast, there will be 850 ZB of data generated by mobile users and IoT devices by 2021. With increasing volume, […]","categories":["Deep Tech"],"tags":["edge computing","intel server"],"author_name":"Ram Sagar","publish_date":"2020-04-02T16:00:21","publication_year":"2020","word_count":779,"keywords":["machine learning","TPU","intel server","AI","cloud computing","AWS","ML","R","RAG","edge computing","Azure"],"extracted_tech_keywords":["AI","machine learning","ML","RAG","cloud computing","AWS","Azure","edge computing","TPU","R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/edge-computing-machine-learning\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":30549,"title":"Will Google &#038; NASA Joining Hands Make Quantum Computing A Household Word","content":"With quantum computers, the next wave of transition might not take that long. The next 5 decades of advancements might reduce present-day supercomputer to the level of an abacus in comparison with super quantum computers. Google and NASA have joined hands to commercialize quantum computing back in 2013. And, they have made extraordinary progress with their smart team at Quantum AI. They tried to address the simplest of quantum problems with NISQ computers and as a result, Cirq has been designed. Licensed under Apache2, this open source framework can be modified by the users for any commercial purpose. Evolution Of Cirq A quantum computer requires a quantum transistor unlike CMOS in a classical computer. We are no longer dealing with low and high voltage thresholds (0 or 1); electrons behave like waves in quantum states and the smallest measurable unit of this information is a qubit or quantum bit. A qubit can be in two states at any given point (coherent superposition). And, for controlling or measuring these qubits, we need metals, such as Niobium which assist superconductivity. The circuitry of a typical quantum processor is sophisticated; not to mention the cooling systems which can go as low as 15 milliKelvin (180 times colder than the interstellar space). With such high standards involved in a hardware installation, the need for having a virtual environment to simulate quantum effects only makes sense. Cirq package allows users to develop processor specific quantum algorithms. These algorithms can further be used to simulate the quantum environment like specifying the gating behavior and testing the constraints. These results can then be used to integrate with quantum hardware via a cloud. Installing Cirq Package For Python This will install Cirq on to Windows and now we shall see how we can use Cirq’s built-in modules to create qubits The above code will list the qubits as follows: We can now use this to construct a quantum circuit as follows: Similarly, we can also create gates and test constraints with few lines of python code. The hassles surrounding building hardware quantum experiments are exceedingly high and running simulations [cirq.google.XmonSimulator()] in virtual environments like illustrated above will make it easy for the researchers to test multiple combinations and integrate later with the hardware. Probing Infinities With Quantum ML Amplitude encoding is one of the fundamental concepts in quantum theory where amplitudes of a quantum state can be defined in terms of a probability density function. This function reveals how a quantum state vector relates to the observations made. For every qubit, say x, there are 2^x states. Now, imagine the number of states a 100 qubit processor can handle! Two qubits can store four states, three qubits can store eight states. So, 50 qubits in quantum computers can be approximated to 1 quadrillion bits in classical computers. This alone is the single most important feature and reason behind the arduous efforts to improve quantum computing. The exponentially increasing complex amplitudes can have as much information encoded. Now, this endless chain of information can be thought of as a large matrix representing equations. ML models run on a system of equations like these. For example, a hyperplane in Support Vector Machines can be written as, wx-b = 0 Hyperplane as a system of equations in SVM So with increasing feature set, the complexity of the equation increases or let’s say the size of a matrix increases and complexity is where quantum machine learning comes into picture. By initiating a quantum state where amplitudes, that give away the probability density of the quantum state vectors can be related to the feature vectors in our dataset used for training the model. So anyone who has had an idea on how to run a standard regression model would comprehend how complex and time consuming it gets with increasing number of rows\/columns (think millions). Building on Cirq, the researchers have also released application specific library OpenFermion. This library contains necessary modules to tackle the problems related to quantum chemistry; the study which will allow the researchers to gain insights about how cathodes in a battery degrade or the electrical properties of superconductors. Companies like D-Wave and Google have been trying to develop a sophisticated hybrid system which incorporates quantum effects into the classical ML problems since the turn of this decade and they will continue to do so. With the current fabrication techniques, D-Wave has managed to develop 2000 qubits processors. And, we have just read about the capabilities of a 100 qubits processor. So, it does make sense when futurists say that classical computers will be made to look like an abacus in front of quantum computers.","excerpt":"With quantum computers, the next wave of transition might not take that long. The next 5 decades of advancements might reduce present-day supercomputer to the level of an abacus in comparison with super quantum computers.   Google and NASA have joined hands to commercialize quantum computing back in 2013. And, they have made extraordinary […]","categories":["Global Tech"],"tags":["Google Quantum Computing","Machine Learning","Quantum Computing","quantum computing companies","quantum machine learning","qubits","the quantum companies","what companies use quantum computers"],"author_name":"Ram Sagar","publish_date":"2018-11-22T11:38:48","publication_year":"2018","word_count":775,"keywords":["Quantum Computing","Go","what companies use quantum computers","machine learning","qubits","programming_languages:R","AI","ML","quantum computing companies","Machine Learning","Google Quantum Computing","the quantum companies","Python","ViT","programming_languages:Python","quantum machine learning","R"],"extracted_tech_keywords":["AI","machine learning","ML","quantum machine learning","Python","R","Go","ViT","programming_languages:Python","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/will-google-nasa-joining-hands-make-quantum-computing-a-household-word\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10119519,"title":"Building a One Person $1 Bn Company is Now Possible","content":"In the recent episode of The Ben & Marc Show, which features a16z co-founders Marc Andreessen and Ben Horowitz, the question was asked, “Do you think it’s finally possible for a one person billion dollar startup?” Andreessen said, “You could put a whole bunch of things in this bucket…it’s the inherent scalability of software and the internet but you could also put AI,” he added that you could also do it using a lot of work getting outsourced. He added funnily a billion dollar startup of one person already exists and it is called Bitcoin. Horowitz adds his bit that it is actually a trillion dollar company. The best part is that all of this is actually somehow possible using AI and Copilot. Andreessen said that it is possible that super geniuses in the coming 20 years would be able to crack this using just AI and Copilots. Horowitz added, “the internet is actually probably in some ways a bigger breakthrough than AI,” while also saying that we might see a few companies like these. On the other hand, Horowitz also added that he is a little skeptical about the capabilities of AI. “The distance between what we can do now with co-pilot to complete human elimination just in terms of the technologies that we’ve seen like that last last mile is a long mile,” he said, adding that all of a company’s coders become AI is highly unlikely at the moment. Though not a one person company, Andreessen highlighted that Instagram had billions worth of sales with only 11 people and WhatsApp had 15 plus billion sales with 50 people. This conversation is similar to what OpenAI’s CEO Sam Altman said in an interview. “We’re going to see 10-person companies with billion-dollar valuations pretty soon…in my little group chat with my tech CEO friends there’s this betting pool for the first year there is a one-person billion-dollar company, which would’ve been unimaginable without AI. And now [it] will happen.” Moreover, Altman also said in a recent interview that GPT-4 is the dumbest model that we would ever have to use and the upcoming models are going to be a lot smarter.","excerpt":"Sam Altman definitely believes so.","categories":["Deep Tech"],"tags":["AI Impacts"],"author_name":"Mohit Pandey","publish_date":"2024-05-03T12:30:39","publication_year":"2024","word_count":362,"keywords":["Go","OpenAI","AI","programming_languages:R","Scala","GPT","llm_models:GPT","AI Impacts","copilots","R","startup"],"extracted_tech_keywords":["AI","OpenAI","copilots","R","Go","Scala","GPT","startup","llm_models:GPT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/building-a-one-person-1-bn-company-is-now-possible\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10084228,"title":"The Judgment Is Out For NLP","content":"Software for document review has existed for years and typically only helps to store and organise contracts. Enter NLP-based softwares that have raised the bar for what can be accomplished. Even though a firm stands to lose 5% to 40% of its value on a given deal due to inefficiency, contracting remains an activity that only some companies do efficiently. Moreover, comprehending legal text can be challenging due to its verbosity and density and few expert-annotated datasets. The Atticus Project, a non-profit organisation, has introduced the Merger Agreement Understanding Dataset (MAUD), an expert-annotated reading comprehension dataset. Legal NLP landscape MAUD is based on the American Bar Association’s 2021 Public Target Deal Points Study, with over 39,000 examples and over 47,000 annotations that legal experts have manually labelled. Previously, in 2021, the non-profit organisation also released Contract Understanding Atticus Dataset (CUAD) with annotations from lawyers. The extensive dataset is estimated to cost over $2 million, with a corpus of more than 13,000 labels in 510 commercial legal contracts. Legal expert systems have been a hot topic of discussion since the 1970s. Early approaches once more used the presence of key terms and headings to guide information extraction, and it’s likely that many offerings still make use of some proportion of rule-based technology; however, not surprisingly, pretty much all the recent entrants into the space are using more sophisticated machine learning techniques. The Goa-based Contractzy (formerly known as ‘The Legal Capsule’), was founded by Gautami Raiker in 2018. The contract lifecycle management (CLM) the platform provided was the only woman-founded Indian start up to be selected for Microsoft Emerge X Programme Highway to 100 Unicorns. Founded a year ago in 2017, SpotDraft, started leveraging AI to automate and streamline the lengthy and complex contract lifecycle. Speaking to AIM, Madhav Bhagat, co-founder and CTO, SpotDraft, said, there are many legal datasets released and also the publically available (but unannotated) contract data from SEC EDGAR, MCA of India, etc. SpotDraft used these datasets to create ‘Legal pre-trained transformer models’ which understand legal concepts better compared to standard off-the-shelf models trained on web crawl data like BERT large and other transformer-based models. More recently, such datasets have become useful in prompt creation for few shot reasoning with LLMs like GPT-3. Since these models are prohibitively expensive to train from scratch, SpotDraft uses such datasets to finetune them or just to create more relevant prompts. Both fine-tuning and better prompting can result in performance boosts of 20-30% over a standard model, Bhagat added. Today, Startups like Lawgeex provide a service to review contracts and, in some cases, more accurately than humans. The firm emphasises the ability to compare contracts against predefined company policies. Other firms like Klarity, Clearlaw and LexCheck have grabbed the opportunity. They are developing AI systems that can automatically ingest proposed contracts, analyse them in full using NLP, and determine which portions of the contract are acceptable and which are problematic. Is NLP smart enough for law? Law firms aim to hammer out agreements with accurate and efficient preprogrammed parameters to review contracts. Unfortunately, the legal sector is too abstract, and the consequences too severe for the widespread adoption of AI. The problems and issues faced in artificial intelligence-based contracting need a lot of deliberations and discussions to eliminate the pitfalls and make the collaboration between legal contracting and artificial intelligence more efficient. On the surface, seeing the issues in general, it might appear that discussing more on the issue of granting artificial intelligence a limited legal personality status might untangle the issues. However, the liability conundrum remains even if the artificial intelligence technology causes an error; the accountability will likely fall on the programmers. Commenting on the accuracy of these models, Bhagat said, “All datasets come with inherent biases, as these datasets do not accurately capture Indian sensibilities, including naming conventions like “son of”, “resident of” etc. Further, since they are human annotated and some parts of the law are open to interpretation, there can be disagreements on certain things mentioned in the dataset if another lawyer were to review it. At times these biases can become issues as the models trained on these datasets will also learn these biases and thus answer accordingly.” Moreover, smaller firms may need more financial strength to adopt the new technology. For example, to manage work, a law practitioner can purchase an AI assistant for cheaper compared to what a firm will spend on software to manage those tasks. Then there is the cost of tech support after the initial setup. Hence, law firms that can afford AI can perform better, financially, than the others. Named entity recognition (NRE), which many NLP models rely on, may also be insufficient for legal work. Lengthy legal documents, especially court proceedings, may not always refer to an entity by the same name, making it harder for these models to highlight the relevant information. Multistep questions also remain a challenge for today’s NLP models, yet these are common in law. Similarly, many legal issues are too nuanced to be black or white, “if this, then that” reasoning. The definition of a legal error changes depending on the application of abstract concepts, which AI has a hard time with. In conclusion, and highlighting areas that can be further developed, Bhagat said, “Generally, to get better results we would want to see more explanations behind certain answers given as part of the dataset so that the model can be fed those and trained to give explanations of its own. This can help solve the explainability and interpretability problem that exists in the NLP domain, especially when dealing with black-box models.”","excerpt":"Law firms lose up to 40% value on a deal due to inefficiency in contracting","categories":["AI Features"],"tags":["how to retrain data","nlp in data","nlp pipeline"],"author_name":"Tasmia Ansari","publish_date":"2023-01-06T12:00:00","publication_year":"2023","word_count":938,"keywords":["Go","artificial intelligence","machine learning","AI","how to retrain data","R","ML","RAG","NLP","BERT","Aim","nlp pipeline","nlp in data"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","NLP","Aim","RAG","R","Go","BERT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/the-judgment-is-out-for-nlp\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10114697,"title":"Together AI Releases Biological Foundational Model Evo","content":"San Francisco-based AI research startup Together AI has introduced an advanced new biological foundational model called Evo,  to understand and create sequences from DNA, RNA, and proteins. Evo is trained with a large dataset of prokaryotic genomic sequences, covering 2.7 million whole genomes. The main goal of Evo is to address the challenges of modelling entire genomes, given their long length and the intricate changes happening at the level of individual building blocks, or nucleotides. Unlike previous AI models in biology that focused on specific tasks, Evo is designed as a foundational model. It integrates information across long genomic sequences while being sensitive to individual nucleotide changes. To overcome challenges related to long sequences and precise resolution, Evo uses the StripedHyena architecture, employing a hybrid design of rotary attention and hyena operators. Evo-1 exhibits several notable capabilities, such as predicting essential genes for an organism’s survival based on small DNA mutations without prior training (zero-shot gene essentiality testing). It also excels in predicting functions across DNA, RNA, and proteins, outperforming other models in protein function prediction. Evo goes a step further by generating novel CRISPR systems, showcasing its ability to design complex molecular structures involving proteins and RNA simultaneously. The model can generate sequences at the scale of entire genomes, up to 650,000 characters, using a single GPU. The researchers see Evo as a groundbreaking technology with the potential to accelerate discoveries in various scientific fields, including biology, chemistry, and material science. Its applications extend to practical challenges such as drug discovery, agriculture, and sustainability. In December last year, Together AI received a $102.5 million Series A investment from NVIDIA, Kleiner Perkins, and Emergence Capital.","excerpt":"The main goal of Evo is to address the challenges of modeling entire genomes, given their long length and the intricate changes happening at the level of individual building blocks, or nucleotides.","categories":["AI News"],"tags":["AI Healthcare"],"author_name":"Shritama Saha","publish_date":"2024-02-28T17:18:50","publication_year":"2024","word_count":275,"keywords":["Go","API","programming_languages:R","AI","AI Healthcare","programming_languages:Go","Together AI","GAN","AI research","R","startup"],"extracted_tech_keywords":["AI","R","Go","API","GAN","startup","Together AI","AI research","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/together-ai-releases-biological-foundational-model-evo\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10162511,"title":"ONDC, CoRover Unveil Mahakumbh AI Sahayata Assistant for Pilgrims, Powered by BharatGPT","content":"In a bid to enhance the experience of millions of pilgrims, the Open Network for Digital Commerce (ONDC) has partnered with conversational AI firm CoRover to launch Mahakumbh AI Sahayata, an AI-driven assistant powered by BharatGPT. The initiative aims to provide real-time guidance, information, and support to pilgrims attending the Mahakumbh festival. Available via phone and WhatsApp, the AI assistant offers multilingual support in over 14 languages, catering to pilgrims regardless of their tech proficiency or internet availability. It can handle up to 100,000 queries or calls simultaneously, assisting with directions, schedules, accommodation details, and emergency contacts. “ONDC is committed to leveraging technology to enhance experiences and empower people. We are excited to contribute to initiatives that make a meaningful impact and promote inclusivity for all,” said T Koshy, MD & CEO at ONDC. CoRover’s founder & CEO, Ankush Sabharwal, emphasised the initiative’s scalability and accessibility, stating, “Mahakumbh AI Sahayata is more than a tool; it’s a companion that cares, powered by BharatGPT – India’s pride. With its ability to handle 100+ languages and process up to 100,000 queries\/calls concurrently, we’re ensuring no pilgrim is left unheard.” Designed to bridge the digital divide, Mahakumbh AI Sahayata ensures seamless access to essential information, reinforcing ONDC and CoRover’s vision of leveraging AI for societal benefit. The Ministry of Electronics and Information Technology (MeitY) has also integrated Bhashini, its language translation ecosystem, to provide multilingual support at Maha Kumbh Prayagraj 2025. This initiative ensures seamless communication for millions of devotees in 11 languages, including Hindi and English. Similarly, Bhavish Aggarwal, founder and CEO of Ola and Krutrim, also earlier announced that Krutrim will power the LLM services for the Kumbh Sah’AI’yak app, an AI chatbot unveiled today by Prime Minister Narendra Modi.","excerpt":"Available via phone and WhatsApp, the AI assistant offers multilingual support in over 14 languages.","categories":["AI News"],"tags":["AI in Maha Kumbh","ondc"],"author_name":"Mohit Pandey","publish_date":"2025-01-30T12:13:52","publication_year":"2025","word_count":289,"keywords":["AI","ondc","ML","Scala","Git","RAG","GPT","Ray","Aim","ViT","R","AI in Maha Kumbh"],"extracted_tech_keywords":["AI","ML","Aim","Ray","RAG","R","Scala","Git","GPT","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/ondc-corover-unveil-mahakumbh-ai-sahayata-assistant-for-pilgrims-powered-by-bharatgpt\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10110175,"title":"Microsoft to Train 100,000 Indian Developers with AI Odyssey","content":"Microsoft has introduced AI Odyssey, a new initiative with the objective of upskilling 100,000 developers in India with AI. This month-long program is designed to provide a comprehensive learning experience, enabling developers to acquire and showcase the necessary skills for executing crucial projects utilising AI aligned with business goals and outcomes. Click here to register for the programme. With AI Odyssey, Microsoft is fostering opportunities for developers to craft solutions for India’s progress and highlight their abilities in tackling real-world issues. The program, open to all AI enthusiasts in India regardless of experience or background, comprises two levels that participants must complete by January 31, 2024. The first level focuses on educating participants about leveraging Azure AI services to develop and deploy AI solutions across various scenarios. This level provides access to valuable resources, code samples, and guides, facilitating the mastery of practical AI skills. The second level challenges participants to validate their AI expertise through an online assessment featuring interactive lab tasks. Successful completion earns them Microsoft Applied Skills credentials, a verifiable proof of their capability to address real-world problems using AI. Participants completing both levels also have the opportunity to win a VIP Pass to the Microsoft AI Tour in Bangalore on February 8, 2024. This event showcases the transformative impact of generative AI on creativity, collaboration, and problem-solving, featuring keynote sessions, demos, and workshops where developers can learn from Microsoft experts, partners, and network with peers. Irina Ghose, managing director at Microsoft India, emphasised, “AI is the future of innovation and India is leading the way with its tech talent. The Microsoft Applied Skills credential will help developers demonstrate their competence and creativity in the most in-demand AI skills and scenarios. We welcome all developers to join us in creating meaningful AI solutions that will contribute to India’s economy.” Read: Microsoft IDC Turns 25 in India","excerpt":"The program is open to all AI enthusiasts in India regardless of experience or background.","categories":["AI News"],"tags":["Developers","Microsoft"],"author_name":"Mohit Pandey","publish_date":"2024-01-08T14:20:17","publication_year":"2024","word_count":309,"keywords":["Go","cloud_platforms:Azure","Developers","AI","programming_languages:R","R","innovation","RAG","ViT","generative AI","Azure","Microsoft"],"extracted_tech_keywords":["AI","generative AI","RAG","Azure","R","Go","ViT","innovation","cloud_platforms:Azure","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/microsoft-to-train-100000-indian-developers-with-ai-odyssey\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":66256,"title":"What Is Contrastive Learning?","content":"The recent success in self-supervised models can be attributed in the renewed interest of the researchers in exploring contrastive learning, a paradigm of self-supervised learning. For instance, humans can identify objects in the wild even if we do not recollect what the object exactly looks like. We do this by remembering high-level features and ignoring the details at the microscopic level. So, now the question is can we build representation learning algorithms that do not concentrate on pixel-level details, and only encode high-level features sufficient enough to distinguish different objects? With contrastive learning, researchers are trying to address this. Recently, even Google’s SimCLR demonstrated the implications of contrastive learning, which we will briefly go into at the end of this article. Principle Of Contrastive Learning via Ankesh Anand Contrastive learning is an approach to formulate the task of finding similar and dissimilar things for an ML model. Using this approach, one can train a machine learning model to classify between similar and dissimilar images. The inner working of contrastive learning can be formulated as a score function, which is a metric that measures the similarity between two features. Here x+ is data point similar to x, referred to as a positive sample x− is a data point dissimilar to x, referred to as a negative sample Over this, a softmax classifier can be built that classifies positive and negative samples correctly. A similar application of this technique can be found in the recently introduced framework SimCLR. Applying Contrastive Learning via Google AI Google has introduced a framework called “SimCLR” that uses contrastive learning. This framework first learns generic representations of images on an unlabeled dataset and then is fine-tuned with a small dataset of labelled images for a given classification task. The basic representations are learned by simultaneously maximising agreement between different versions or views of the same image and cutting down the difference using contrastive learning. When the parameters of a neural network are updated using this contrastive objective causes representations of corresponding views to “attract” each other, while representations of non-corresponding views “repel” each other. A finer explanation of the original paper was given in this blog. The procedure is as follows: First, generate batches of a certain size, say N from the raw imagesFor each image in this batch, a random transformation function is applied to get a pair of two imagesEach augmented image in a pair is passed through an encoder to get image representations. The representations of the two augmented images are then passed through a non-linear dense layer followed by a ReLU, which is then followed by another dense layer. These images are passed over a series of these layers to apply non-linear transformation and project it into a representation For each augmented image in the batch, get an embedding vector. Now, the similarity between two augmented versions of an image is calculated using cosine similarity. SimCLR uses “NT-Xent loss” (Normalised Temperature-Scaled Cross-Entropy Loss), which is known as contrastive loss. via amitness First, the augmented pairs in the batch are taken one by one. Later a softmax function is applied to find the probability of these two images being similar. via amitness As shown above, the softmax function can be used to calculate how similar the two augmented cat images are and all remaining images in the batch are sampled as dissimilar images (negative pair). Based on the loss, the encoder and projection head representations improve over time, and the representations obtained place similar images closer in the space. The results from SimCLR showed that it outperformed previous self-supervised methods on ImageNet. To know more about this topic, check this and this.","excerpt":"The recent success in self-supervised models can be attributed in the renewed interest of the researchers in exploring contrastive learning, a paradigm of self-supervised learning. For instance, humans can identify objects in the wild even if we do not recollect what the object exactly looks like. We do this by remembering high-level features and ignoring […]","categories":["Deep Tech"],"tags":["contrastive learning"],"author_name":"Ram Sagar","publish_date":"2020-05-31T13:00:00","publication_year":"2020","word_count":608,"keywords":["Go","self-supervised learning","machine learning","programming_languages:R","AI","neural network","ML","programming_languages:Go","R","contrastive learning"],"extracted_tech_keywords":["AI","machine learning","ML","neural network","R","Go","self-supervised learning","contrastive learning","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/contrastive-learning-self-supervised-ml\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10124865,"title":"“If We Got the Same Education as IITians, We Wouldn&#8217;t Suck…”","content":"Hating on IIT graduates is literally a skill issue. But this is a new trend where a lot of people are taking shots at IITians because they claim that they are nothing when compared to people from Harvard, Stanford, or MIT. Well, this diss of IITians stems from the fact that many people are just jealous of them. Preparing for the JEE is like enduring a mental military boot camp. It’s gruelling, relentless, and often soul-crushing. Raj Dabre, a prominent researcher at NICT in Kyoto, adjunct faculty at IIT Madras and a visiting professor at IIT Bombay, joined in on the conversation. “JEE prep was one of the hardest things I ever did,” he recalled. Dabre said that as someone who failed the JEE, attended a tier-3 college, and later made it to an IIT through GATE, he managed to build a strong foundation in maths and science. This rigorous preparation wasn’t just about clearing an exam—it rewired his brain to handle complex problems, which later, he said, “helped me breeze through my undergrad studies.” Anyone with dedication can master the basics, JEE or no JEE. The exam just happens to cram a lot of that learning into a short period, turning students into problem-solving machines. https:\/\/twitter.com\/wojakcodes\/status\/1805627812520325392 “IITians Make IITs Great, Not the Other Way Around” In today’s digital age, failing to solve problems and then attributing it to not being from an IIT is just an excuse. Top-tier educational resources are available online for free, speaking of which, Andrej Karpathy and Andrew Ng are offering free courses almost every week. There are so many brilliant educators on YouTube who sometimes outshine traditional professors. The truth is you can get an exceptional education without ever setting foot in an IIT. Becoming an AI researcher in India by grinding yourself through JEE is definitely not the only way as well. \"If we got the same education as IITians, we wouldn't suck…\"Harvard's top lectures are available for free on the Internet.NPTEL has some very good courses available for free.So many cool youtubers who teach better than top IIT profs…You're just not willing to learn!!— ‎Wojak Codes (@wojakcodes) June 25, 2024 On the other hand, Abhay Rao argues on X that this is only the case for computer science and not other fields. “No way you’re getting the hands-on experience that an IIT student would get on that expensive equipment,” he said. This seems like a valid point as most of the seats filled in several branches apart from CS are people who did not want to go there, but just got the seat. The thing about IIT is not just education, but the peer group to study with. Though one can get the content anywhere on YouTube, the peers are only available in the institute. This can be addressed by forming groups and learning together, but the point makes sense. Currently, everyone is yet to become a self-learner. But this might slowly change as advancements in AI could encourage people to learn coding with the help of such tools. For now, however, many students still need a good teacher in front of them. Not a lecturer on YouTube. The undeniable truth is that professors from IITs are probably the best in the country. Which is what makes a lot of difference. Nilesh Yadav on X suggests an experiment: “Imagine an experiment where we take the top 1% of JEE students and put them in a regular college, while random JEE students are put in an IIT.” He predicted that the 1% would still perform better regardless of the college. Though, It Is Fine To Admit That Indian Researchers Are Not That Good The notion that all IITians are privileged geniuses with a golden ticket to success is as misguided as it is petty. Sure, some IITians might be, but the majority are just hardworking individuals who’ve faced intense pressure to succeed. And guess what? “Not every IITian is a great engineer.” Many students slog through their courses just like anyone else, watching YouTube lectures at 2x speed the night before their exams. The critique that Indian AI\/ML researchers aren’t at the level of international bigwigs like Ng, Karpathy, or Ilya Sutskever is somewhat true. We don’t have anyone in India making an impact at that level. Hard cope.Not one \"ML\/AI researcher\" in India is anywhere close to Andrew Ng, Ian Goodfellow, Andrej Karpathy, Ilya Sutskever, etc (most of the folks mentioned are less than 40 years of age btw)You think the JEE grind is the only way to learn math? Absolute joke. pic.twitter.com\/A7Rc17YIWq— major tom (@tailwiinder) June 25, 2024 That indeed is true, but let’s not ignore the fact that cutting-edge research requires serious funding. The US pours money into research, attracting top talent and enabling groundbreaking work. Moreover, many IIT graduates also leave the country for the US to do foundational research. In comparison, India’s research funding is peanuts. But compare Indian researchers in the US with their peers, and the gap disappears. At the same time, the glorification of JEE to an extent that it becomes the gold standard is also not fruitful. Most people who attempt JEE and do not get selected can also succeed, it’s just a game of skill. “I’d rather take a 1 crore loan to educate my child than worry about them not being alive, or coming out of it with zero social skills, and still having an infinitesimal chance of winning,” concluded the user who started the debate on the topic.","excerpt":"IITians make IITs great, not the other way around.","categories":["IT Services"],"tags":["AI at IIT","IIT"],"author_name":"Mohit Pandey","publish_date":"2024-06-26T17:00:00","publication_year":"2024","word_count":917,"keywords":["Go","funding","AI at IIT","AI","RPA","ML","Git","RAG","GRU","Aim","IIT","R"],"extracted_tech_keywords":["AI","ML","Aim","RAG","R","Go","Git","GRU","RPA","funding"],"url":"https:\/\/analyticsindiamag.com\/it-services\/if-we-got-the-same-education-as-iitians-we-wouldnt-suck\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10100195,"title":"AI Clock is Ticking: Wake Up Call for Education Institutions","content":"After the launch of ChatGPT, every industry in the market is figuring out how to make the best out of this technology. Alongside this excitement, it has also caused concern among several startups and employees, as they fear that AI is looming over them, potentially taking away their jobs. However, among all the AI buzz, there is one sector which appears to be carefree and is taking AI lightly without realising its consequences. It is none other than educational institutes. This approach might hurt them in the future if they don’t take necessary steps. Whenever a new technology comes in, there are always early adopters and late majority, according to the Theory of Diffusion. In the case of foundational models and generative AI, no one had expected that it would be universities and schools which would be shying away from it while they should have been the first ones to integrate in their teaching methods. As soon as they came to know about ChatGPT, the knee jerk reaction of the majority of the universities and colleges around the world was to ban it. Why, just because they feared that students might use it to copy their assignments. This solution is surely not going to work in the long run. Students are smart enough to outwit their professors if they want to cheat in their assignments. For instance, when Google Search came in, it didn’t stop students from surfing the internet for completing their assignments. To tackle the issue of plagiarism, several tools like Turnitin emerged. Similarly here professors need to come up with better solutions to handle AI. One thing is for sure, running away isn’t the solution. Accept it or get doomed The ongoing discussion among the AI industry revolves around the potential threat to the existence of educational institutions if they don’t adapt accordingly. Recently, a X user posted  “We should decimate educational institutions with AI”. Similarly, Ethan Mollick, professor at Wharton posted on X “The start of the school year is AI chaos, with many instructors just ignoring AI. I think instructors need to make active choices about AI use (which might mean embracing AI or returning to in-class tests) and there is near certainty that models will improve over the school year.” Mollick is not alone in expressing these sentiments. “Educational institutions are completely unprepared for the AI revolution. They managed to wing it last school year, but I feel they will be completely overwhelmed this coming year,” posted  Bojan Tunguz, machine learning scientist at NVIDIA, on X. It’s pretty much evident that those who will embrace AI will survive and those who won’t are in for a hard time. What’s needed? Rather than seeing AI as a threat, educators can view it as a valuable partner. AI-driven chatbots, for instance, can provide immediate responses to student inquiries, freeing up professors to focus on more complex aspects of teaching. Echoing similar sentiments, Rajeev Kumar Singh, Associate Dean, Academics at Shiv Nadar University  in a recent interaction told AIM that he is optimistic about adopting AI tools in teaching methods. He explained this will give faculty some extra time to spend on discussions with students one on one. “Let’s say I take 45 lectures and I find that in 10, I was only transferring information. So, if that can be outsourced to some technology, I can spend the entire rest of the lectures on really bigger things. If I can save that time, then I can really engage one to one with students, which can be very fruitful, it can lead to a lot of personalised assessments, learning and also relationships.” It’s understandable that it is tough for educators and professors to start from scratch to understand this new technology. However, it is for their betterment only. Otherwise chances are pretty much that AI will replace average teachers and amplify skilled educators. Earlier this year, Harvard University announced that its coding course, CS50 will be taught by an AI instructor. So, the reality isn’t far away. To give themselves a head start, professors can go through a five part YouTube series on ‘Practical AI for Teachers & Students’ created by Mollick. Mollick went on to say that he agrees it is a lot for instructors to handle among their other obligations, “but the tech isn’t going away.” To assist educators  with AI, he has even published two papers ‘Assigning AI: Seven Approaches for Students, with Prompts’ and ‘Using AI to Implement Effective Teaching Strategies in Classrooms: Five Strategies, Including Prompts’. Not only that OpenAI recently took initiative to educate teachers how they can use ChatGPT in schools and universities. They gave several examples on how teachers can come up with lesson plans made with the help of ChatGPT. Moreover, universities themselves can organise sessions designed to educate both faculty and students about the implications of large language models by partnering with industry experts. For instance, Shiv Nadar University recently hosted one  session titled “The New Landscape: Designing in the Times of Dall-E, Midjourney, and ChatGPT!”. It’s not too late The good news is that educational institutions still have time to make corrections. The process of integrating AI can begin with the admission process. Universities can consider developing their own LLMs, which would greatly assist students in navigating the admission process. Similarly, schools and colleges can create a process that emphasises students’ critical thinking skills while still allowing them to utilize ChatGPT for assistance. There is no denying the fact that ChatGPT provides the best answers when asked the right questions. Many believe that personalised AI assistants are going to be a thing in the future. If we extend that idea to education soon we might also get personalised AI tutors who would be teaching students according to their requirements and understanding capabilities. This will solve Bloom’s Sigma 2 problem. Bloom’s research found that students who were taught one-on-one or in small groups and received regular feedback performed two standard deviations (2 sigma) better than their peers who received traditional classroom instruction. One can get a rough idea of how the future is going to look from Musk’s private school he created for his children—called Ad Astra. It is quietly built inside the SpaceX campus and has partnered with Synthesis, founded by Joshua Dahn. Synthesis has created an AI Tutor which teaches students complex concepts of math through personalised games and AI tools.","excerpt":"It’s not too late","categories":["Deep Tech"],"tags":[],"author_name":"Siddharth Jindal","publish_date":"2023-09-18T15:25:10","publication_year":"2023","word_count":1065,"keywords":["ChatGPT","AI assistants","machine learning","OpenAI","AI","chatbots","RAG","Aim","generative AI","R"],"extracted_tech_keywords":["AI","machine learning","generative AI","ChatGPT","OpenAI","Aim","RAG","AI assistants","chatbots","R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/ai-clock-is-ticking-wake-up-call-for-education-institutions\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10008288,"title":"KNIME &#8211; A Primer To Automate Machine Learning ‘No Code’ Workflows","content":"Beginners in the field of machine learning face numerous challenges in trying to cope up with the fast-paced nature of AI. It is especially difficult for people with no coding experience since they have to learn the math behind the algorithm and learn how to code the algorithm as well. To make things a little easier for them, a no-code machine learning GUI called KNIME was developed. In this article, we will learn about KNIME and discuss how to use this tool for building a machine learning model from scratch. What is KNIME? Knime is a GUI based workflow platform that can be used to effectively build machine learning models without having to code. Here, you simply have to define the workflow between some pre-defined nodes. These nodes may be for data cleaning, data visualization and model training. Once the workflow is defined, the model can be trained to get the desired output. All functions from basic input-output operations to data mining can be performed with KNIME. Installing KNIME To download this interface click here and select the operating system as per your computer requirements. For Windows users select the first option above and the download will begin. Once the download is completed follow the steps shown and you will then see a KNIME dashboard before you. Creating a workflow To create the machine learning model we first need to set up a workflow. For this, select File-> New and select a new workflow. You will get a popup where you can type in the name of the project. Click on the finish to get the new workflow before you. On the right-hand side, you can type in the description of the projects, any links for reference as well. The left-hand side is where you will be creating the workflow. Getting the dataset Now that we have created our workspace, let us get the dataset. To do this, first, download the dataset that you want to use for the project. I have used the tips dataset from Kaggle, which can be downloaded from here. The dataset contains values like a smoker, time, day and total_bill which is used to predict how many tips a waiter will get. It is a regression problem and is a simple project. After downloading the data, go to your node repository and search ‘file reader’. Drag and drop this on the workspace. Then, double click this and browse the dataset on your local system and upload the file. Once you do that you will get a preview of the dataset. Here you can select options like ignore tab spaces, reading the column headings etc. After you have selected the desired options select apply and ok. Once one, right-click on the node and select ‘Execute’ button so that it is executed. Correlation The next step is to identify the correlation that exists between the features. To do this search ‘Linear correlation’ on the node repository and drag and drop it to the workspace. Then, connect your dataset to this node. Now, right-click on this and click on ‘execute’. After executing this, right-click again and click on ‘view correlation matrix’. Once you select this you will see the matrix. Some columns are not related much with the others hence it is clear that tip and total_bill have a very high correlation. Let us select these two columns to build the model. Data visualization The next step in model building is to visualize the dataset. To do this, search the type of plot you want to visualize. I have selected the scatter plot for the visualization. Drag and drop this node on the workspace and connect your file reader node with it. Once done, right-click and select execute. Here you can change the columns as well to find out how the data is scattered. There are other visualization methods as well like pie charts as shown below Data manipulation In order to find out which values are missing, type in the node repository ‘missing values’. Drag and drop this node and connect with the input file reader. Next, double click on the missing values node. Here you will find a dashboard that lets you impute values in the dataset. These options allow you to impute values either as number or string. I have select to impute the missing values with the mean value. But you can choose from the below options according to your requirements. After selecting this, you can select apply and ok and the missing values are filled automatically. Finally, right-click and select the execute option to run the node. Model building After we have pre-processed and visualized the data it is time to build a model. I will make use of the simple linear regression model on this dataset. To do this, type linear regression learner in the node repository and drag and drop this on the workspace. Connect the missing values node to this since it has the pre-processed data. Now double click the linear regression node. The following is displayed. Here on the top you need to set the target column. Once you set this the target is automatically removed from the inputs shown below. You can choose to eliminate some of the features as well. I will eliminate a few features since there was not much correlation between them with the target. Just select the column to be excluded and click the left arrow button to do this. Once done, select the apply button. Next, right-click the node and select execute. Once the execution is done you can see the output on the screen. The different types of errors and the R-squared value is shown and the results are quite good here. Thus we have built a machine learning model without coding. Conclusion In this article, we saw how simple it is to use the KNIME GUI and build a machine learning model. There is a lot left to explore in this tool for building better and more complex models. KNIME also supports building neural networks and clustering algorithms which is making machine learning easy and accessible to everyone.","excerpt":"In this article, we will learn about KNIME and discuss how to use this tool for building a machine learning model from scratch.","categories":["Deep Tech"],"tags":["GUI frameworks","Machine Learning","machine learning computer","machine learning methods","no etl","no-code","Workflow"],"author_name":"Bhoomika Madhukar","publish_date":"2020-09-25T14:00:08","publication_year":"2020","word_count":1017,"keywords":["Go","GUI frameworks","machine learning","TPU","programming_languages:R","AI","neural network","machine learning computer","Machine Learning","no-code","programming_languages:Go","RAG","no etl","machine learning methods","R","Workflow"],"extracted_tech_keywords":["AI","machine learning","neural network","RAG","TPU","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/knime-a-primer-to-automate-machine-learning-no-code-workflows\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":35223,"title":"How Healthy Is Your Machine Learning Model?","content":"Traders follow a simple motto: buy low and sell high. But when the opposite happens, the stock market goes berserk. On a fine morning in 2012, the NYSE had to step in and cancel numerous trades due to erroneous trading by Knight Capital which saw the biggest drop ever since it went public. Instead of at least attempting to provide liquidity via limit trades, Knight’s algorithm acted as a market order. Naturally, when the entire logic of trading was perverted courtesy of Knight’s busted algorithm, everything went chaos, and stocks went higher because they went higher, and the higher they went, the greater the incentive for the algorithm to keep pushing the stock higher. In other terms, the algorithm bought high and sold low. This was clearly a scenario from the whole “what could go wrong” mindset in deploying advanced methodologies. Now, in 2019, machine learning algorithms are deployed aggressively in the finance and healthcare sector. Whether it is the market crash or a wrong diagnosis, the after-effects will be certainly irreversible. Tracking the development of machine learning algorithm throughout its life cycle, therefore, becomes crucial. Life Cycle Of ML Models The typical life cycle of deployment machine learning models involves a training phase, where a typical data scientist develops a model with good predictive based on historical data. This model is put into production with the hope that it would continue to have similar predictive performance during the course of its deployment. But there can be problems associated with the information that is deployed into the model such as: an incorrect model gets pushed incoming data is corrupted incoming data changes and no longer resembles datasets used during training Datasets used for natural language processing (NLP) are extremely sparse and need special techniques to detect divergence in feature patterns. Similarly, image datasets that typically use deep learning algorithms also need techniques that can detect divergence in incoming patterns. The engineers at ParallelM propose MLHealth, a model to monitor the change in patterns of incoming features in comparison to the ones observed during training and argue that such a change could indicate the fitness of an ML model to inference data. This eliminates the need for the availability of labels. There exist techniques such as KL-divergence, Wasserstein metric etc. that provide a score for the divergence between two distributions. However, they rely on the fact that the inference distribution is available and representative of the inference data. This implies there are enough samples to form a representative distribution. Significance Of Similarity Scores There are three pressing issues that Similarity scores aim to address: Low number of samples: since the Similarity score is calculated based on the parameters of the multinoulli distribution and does not rely on the inference distribution, it is agnostic of the number of samples Inference samples fall in a narrow range of training distribution: Similarity score reply on the probability values associated with this narrow range of distribution and hence does not penalize the fact that inference distribution does not cover the entire range of categories observed during training The subset of patterns seen during inference might either have poor training coverage or good training coverage. Similarity score is designed to tackle this scenario using the training probability values associated coverage scenario and not the good training coverage scenario. In addition to a score for detecting potential drops in predictive performance, the system infrastructure must support such feedback, alert the mechanism and ideally handle the diversity of engines and languages typically used for ML applications (Spark, Python, R, etc.). a system that leverages this score to generate alerts in production deployments. Architecture Of ML Health Source: Paper by ParallelM At the core of this architecture is a Server and an Agent. The Server keeps track of all the deployments across agents, while the agents themselves are agnostic of all the jobs running in the system. Users can leverage built-in applications and health metrics, create new applications and\/or metrics, or import their existing ML pipelines and health-metric computations using any of the popular programming languages. Datasets Used The results were taken by running the model on three publicly available datasets (a) Human Activity Recognition (Samsung) (b) Video Transcoding Time [20] and (c) TELCO Samsung is an activity recognition dataset with 6 classes as labels. It consists of 560 features. Video dataset has video transcoding time as the label and consists of 10 is a binary classification dataset with SLA violations as the label. It consists of 24 Features. Activity and Video datasets are used as is without normalization. Random forest classification and regression are used respectively for the Activity and Video datasets. These datasets are divided into train and test using random shuffling. The sources of these datasets do not claim to provide varying distributions. TELCO dataset inherently consists of different loads seen in the network. Ideally, this should be reflected in the data-deviation scores. Know more about the research here Key Takeaways Similarity score performs consistently well in the presence and absence of noise agnostic of the number of samples present in inference. This is demonstrated by the results across three datasets, briefly described below. Similarity score shows the highest correlation with the predictive performance of the algorithm\/model when a small number of samples are present in the incoming data across the three datasets. Similarity score shows a high correlation with the predictive performance of the algorithm\/model when the nature of incoming data changes and the number of samples is small. This is demonstrated using the TELCO dataset across three different loads. All the data deviation scores perform equally well when there is data corruption, irrespective of the number of samples in the incoming data across all three datasets.","excerpt":"Traders follow a simple motto: buy low and sell high. But when the opposite happens, the stock market goes berserk. On a fine morning in 2012, the NYSE had to step in and cancel numerous trades due to erroneous trading by Knight Capital which saw the biggest drop ever since it went public. Instead of […]","categories":["Deep Tech"],"tags":["machine learning pattern recognition python"],"author_name":"Ram Sagar","publish_date":"2019-02-21T10:28:51","publication_year":"2019","word_count":952,"keywords":["Go","machine learning pattern recognition python","machine learning","AI","ML","RAG","NLP","Python","Aim","deep learning","R"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","NLP","Aim","RAG","Python","R","Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/how-healthy-is-your-machine-learning-model\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10003287,"title":"11 Talks To Look Forward To At CVDC 2020","content":"The Computer Vision Developer Conference, CVDC 2020, has interesting talks and sessions around the latest developments in the field of computer vision. The two-day conference, scheduled for 13 and 14 August will host paper presentations, tech talks and workshops for computer vision practitioners. The virtual conference will uncover some of the most interesting developments, and the latest research and advancements in this area, which is witnessing some interesting use cases across the world. Know more about CVDC 2020 here. Here are 11 interesting talks, workshops and sessions that you should definitely attend. 1| Create your own Camscanner using OpenCV in Python by Shirish Gupta,  Head of Data Science & Partnerships at NBFC Loan2Grow This talk by Shirish Gupta will focus on how apps such as CamScanner and similar apps can be created using computer vision. Apps such as Camsanner are very effective in allowing users to scan documents from mobile and sharing it as an image. It brings several advantages such as cleaning, sharpening, denoising the camera-clicked image into a refined output. It can do so using computer vision, and Gupta will take the attendees through a session on using the technology, especially using the basics of OpenCV in Python to create a similar app. OpenCV is an open-source computer vision library which was built to provide a common infrastructure for computer vision applications. After the completion of this workshop, the audience will be able to use OpenCV to understand different transformation functions such as blurring, thresholding, canny edge detection, etc.; create a functional CamScanner Bonus, using Pytesseract (OCR) to extract data from images. It requires candidates to have familiarity with Python, basics of OpenCV (preferred, not mandatory). 2| Leveraging Computer Vision in drone tech by Animesh Dutta Machine Learning Developer at Kesowa Drone technology is being widely explored, especially with the emergence of technologies such as artificial intelligence, deep learning, autonomy and more. Researchers are working on creating unmanned aerial systems to deliver solutions. In this talk by Animesh Dutta, he will take attendees to how computer vision can help in drone technology. He believes that the day is not far away when we will see flying cars. There are several use cases of drones attainable with the help of Computer Vision — such as detecting anomalous behaviour in a crowd. Using Bayesian Loss for Crowd Count Estimation with Point Supervision it can generate a density map showing humans. Furthermore, computer vision techniques can detect cracks in a building, detect potholes, and more. 3| Autonomous Driving Perception Using Cameras in Unstructured Environments: Challenges & Solutions by Sanjeev Sharma Founder and CEO at Swaayatt Robots One of the fields where computer vision is extensively used is self-driving cars. In this talk by Sanjeev Sharma, he will talk about numerous challenges associated with enabling autonomous driving perception using only cameras. Especially in highly unstructured environments. He will focus on developing real-time deep learning inference models to overcome such challenges. Swaayatt Robots is one of the leading companies in India working in the self-driving space. The technology by this startup enables self-driving vehicles to perceive their environments using off-the-shelf cameras. 4| Explainable AI for Computer Vision by Avni Gupta Technology Lead at SYNDUIT While creating computer vision models, researchers may often find themselves interacting with a black box, unaware of what feature extraction is happening at each layer. Explainable AI makes it easy to comprehend and know when enough layers have been added and what feature extraction has taken place at each layer. In this session, Avni Gupta we will take the attendees through various libraries that make this possible. 5| Intercepting youth at risk using Computer vision on Social Media Platform by Dr Murphy Choy Executive Director at MC EduTech The lockdown due to the ongoing pandemic has forced many to stay indoors. There were reports about family feuds and violence happening at a much higher frequency compared to the pre-Covid 19 periods. This is a cause of major concerns to many social workers. Younger members of the population generally take to social media to air their frustrations, which provides a channel for social workers to reach out. However, social workers can only screen a limited amount of information on social media. With his talk, Dr Choy will discuss how computer vision can be used to augment a social worker’s ability to identify youth at risk using Computer Vision combined with a Knowledge Graph. 6 | Understanding robustness of computer vision by Indrajit Kar Principal Solution Architect at Accenture This talk by Indrajit Kar will focus on the threat to current computer Vision models and how we can subsequently make AI more resilient to attack. It will focus on the motivations, feasibility, and risks posed by adversarial input\/Perturbations. It will explore how intelligent systems based on computer vision can be made more robust against adversarial input. AI not only competes with human capabilities in areas such as image, audio, and text processing but often exceeds human accuracy and speed. While we celebrate advancements in AI, deep neural networks (DNNs)—the algorithms intrinsic to much of AI—have recently been proven to be at risk from attack through seemingly benign inputs. It is, therefore, important to identify and detect these problems, which is what this talk will deal with. 7| How Computer Vision is improving healthcare by Akshit Priyesh Data Scientist at Capgemini Computer vision can potentially support the healthcare industry in many different applications and deliver life-saving functionalities for patients. Currently, the technology is assisting doctors to better diagnose their patients, monitor the evolution of diseases, and prescribe the right treatments. It has been used in imaging analysis, predictive analysis, and healthcare monitoring. Currently, the most widespread use cases for computer vision and healthcare are related to the field of radiology and imaging. In this talk, Akshit will discuss how it can be explored in many different areas including Covid. 8| Challenges for Deep Learning in Medical Imaging and how to overcome them by Abdul Jilani Lead Data Scientist at DataRobot While deep learning on medical imaging is popular, it doesn’t translate easily or quickly into practical or production-ready tech. This talk by Abdul Jilani will address the challenges faced by applied machine learners and how they can overcome these challenges in real-time environments. It will take a deep dive into the problems faced by Medical Imaging researchers such as building good training datasets, wrong practices such as leakage and more. He will talk about sampling biases and strategies for medical image datasets., discuss data augmentation strategies used for different image types like x-rays, how to mimic real-time imaging in datasets built in perfect lab conditions and more. 9| Workshop on Deep Learning for Computer Vision by Dr Vaibhav Kumar Director at Association of Data Scientists This full-day workshop aims at familiarising the concepts of deep learning techniques applied in the field of computer vision. Deep learning has added a boost to the rapidly growing field of computer vision by providing powerful tools to distil actionable information from images. With deep learning, a lot of new applications of computer vision techniques have been introduced and are now becoming parts of our daily lives. This workshop by Dr Kumar aims to take through such use cases. 10| Best practices in machine learning and the art of research management by Dan Malowany Head of Deep Learning Research at allegro.ai ML and deep learning projects involve iterative and recursive R&D process of data gathering, data annotation, research, QA, deployment, additional data gathering from deployed units and back again. The strong coupling between data and model means various teams, with various backgrounds and capabilities, without the use of a unifying R&D management tool can be detrimental in the long run as it can result in reduced collaboration, loss of work, irreproducible training, and a negative effect on the overall effectiveness of the company. As such, companies must use R&D infrastructure tailored for AI projects that supports the R&D workflow from research to production and enables them to adapt their offering to the evolving demand. In this talk, Dan Malowany will share the experience from numerous deep learning projects and describe the features such infrastructure requires in order to boost productivity as well as being adaptive to the different R&D stages. 11| End to end AI by Matthew Zeiler, Founder and CEO at Clarifai This interesting session by Matthew Zeiler will take attendees through insights on how end-to-end AI platforms can help your business by providing a unified AI strategy. As a founder of Clarifi, he works on simplifying the complexities of image and video recognition and making it easily accessible to all. Having built an incredible platform and demonstrated its problem-solving capabilities across numerous industries, Zeiler has tremendous experience in the field, which will be included in his talk as well. Register for CVDC 2020.","excerpt":"The Computer Vision Developer Conference, CVDC 2020, has interesting talks and sessions around the latest developments in the field of computer vision. The two-day conference, scheduled for 13 and 14 August will host paper presentations, tech talks and workshops for computer vision practitioners. The virtual conference will uncover some of the most interesting developments, and […]","categories":["Deep Tech"],"tags":[],"author_name":"Srishti Deoras","publish_date":"2020-07-24T18:06:38","publication_year":"2020","word_count":1468,"keywords":["data science","machine learning","artificial intelligence","AI","neural network","ML","computer vision","Ray","Aim","deep learning"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","neural network","computer vision","data science","Aim","Ray"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/11-talks-to-look-forward-to-at-cvdc-2020\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10041011,"title":"OpenAI Launches $100 Mn Fund To Catch AI Startups Young","content":"Exactly a year ago, OpenAI unveiled the GPT-3 with a whopping 175 billion parameters, which was made available to developers through an API in private beta. Since then, developers across the globe have been using GPT-3 to create realistic dialogues, summarise complex documents, customer service questions, and make search better than ever before. The company’s decision to not open-source GPT-3 gave it more control over the use cases. However, in the recent past, there have been instances of GPT-3 going rogue, like in the case of GPT-3 Dungeon. Microsoft acquired an exclusive license to GPT-3  last year, in the wake of its $1 billion investment in OpenAI. Recently, the former co-opted GPT-3 for a customer product. Soon, GPT-3 would be integrated into MicrosoftPower Apps, the low code app development platform. OpenAI recently announced a $100 million startup fund in collaboration with Microsoft and other partners to invest in startups that leverage artificial intelligence. Developers and entrepreneurs looking to build transformational products with AI can pitch their ideas here. The idea suggested by startups should align with OpenAI’s mission. The selected startups would be working alongside OpenAI, who will get early access to the latest OpenAI systems, support from its team, and discounts on Microsoft Azure. But, the question is, what’s in it for OpenAI and Microsoft Azure? OpenAI is most likely to get equity in the company, along with the intellectual property rights for the use cases or products developed on its systems. “This is not a typical corporate venture fund,” said Sam Altman, co-founder and CEO at OpenAI and former president of Y-Combinator. Further, he said the company is looking to back startups in the impact sectors such as healthcare, climate change and education. Additionally, the company is interested in newer markets that can drive big leaps in productivity, like personal assistants and semantic search. “We think that helping people be more productive with new tools is a big deal, and we can imagine brand new interfaces that weren’t possible a year ago,” added Altman. He also said they are open to new ideas across sectors. However, education, healthcare, climate and energy, professional services (financial, legal, real estate, etc.), search, virtual assistants and new interfaces will remain on top of their target list. Funding landscape According to MarketsandMarkets, the global AI market is expected to grow from $58.3 billion in 2021 to $309.6 billion by 2026, growing at a CAGR of 39.7 percent. OpenAI’s interest in budding sectors like healthcare, education and climate change is no surprise. Microsoft has invested close to $96.3 billion in startups. With the latest launch of OpenAI Startup Fund, the duo is most likely to keep a close eye on early-stage startups. But, the question is, will $100 million be sufficient to fuel innovations in the artificial intelligence landscape, compared to the fund size of some of the prominent venture capital firms in the space. How open is OpenAI? Based in San Francisco, OpenAI was initially founded as a nonprofit AI research entity in 2015 by Altman, Elon Musk, and others, who collectively invested $1 billion. In 2019, Musk left OpenAI due to a difference of opinion. He had criticised OpenAI, arguing that the company ‘should be more open’. Later, Microsoft invested about $1 billion in OpenAI, and got exclusive access to GPT-3 source-code.  This move wholly altered the foundation of OpenAI, moving away from openness and towards secrecy and commercialisation. Today, OpenAI competes directly with Google-owned AI research lab DeepMind. In March 2021, OpenAI discovered neural networks within AI systems resembling the neurons inside the human brain. The multimodal neurons are one of the most advanced neural networks. In the same month, OpenAI released its text-to-image generation model based on transformers architecture called DALL-E.","excerpt":"Exactly a year ago, OpenAI unveiled the GPT-3 with a whopping 175 billion parameters, which was made available to developers through an API in private beta. Since then, developers across the globe have been using GPT-3 to create realistic dialogues, summarise complex documents, customer service questions, and make search better than ever before.  The company’s […]","categories":["Global Tech"],"tags":["AI Startups"],"author_name":"Amit Naik","publish_date":"2021-05-29T15:00:00","publication_year":"2021","word_count":621,"keywords":["semantic search","artificial intelligence","OpenAI","AI","neural network","R","Transformers","virtual assistants","RAG","Azure","AI Startups"],"extracted_tech_keywords":["AI","artificial intelligence","neural network","OpenAI","Transformers","RAG","virtual assistants","semantic search","Azure","R"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/openai-launches-100-mn-fund-to-catch-ai-startups-young\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10068276,"title":"Can ONDC deliver?","content":"Last year in July, a nine-member committee was formed by the Centre to promote open networks, help create digital value chains for Indian businesses and essentially restrict the current digital monopolies by Amazon and Flipkart. Titled the Open Network for Digital Commerce or ONDC, the non-profit project was initiated by the Department of Promotion of Industry under the Department of Commerce and Industry. Nandan Nilekani, the non-executive chairman of IT giant Infosys and AADHAR architect, was named to the advisory council for recommending steps that would speed up the setting up of ONDC. A government-signed order during the launch stated that ONDC intended to design an open-source network independent of any particular platform. It explained, “ONDC is expected to digitise the entire value chain, standardise operations, promote inclusion of suppliers, derive efficiencies in logistics, and enhance value for consumers.” The platform will consist of both products and services. The council also comprised other key members, including Kumar Rajagopalan, CEO of the Retailers Association of India, Arvind Gupta, co-founder of Digital India Foundation, Anjali Bansal, founder of Avaana Capital, Suresh Sethi, CEO of NSDL e-Governance Infrastructure and Chief of the Confederation of All India Traders or CAIT, Praveen Khandelwal, among others. Retired bureaucrat and CEO of National Health Authority, RS Sharma, was also a part of the statutory authority, Unique Identification Authority of India or UIDAI, that led AADHAR. Growth of India’s e-commerce market, Source: IBEF report Launch and attractive investment The bureaucratic wheels turned fast. By September, reportedly, Union Commerce Minister Piyush Goyal had already helmed meetings that had laid the roadmap for ONDC. Goyal has been open about his dislike for Bezos’ e-commerce giant Amazon and Walmart-owned Flipkart, which together hold power over close to 60 per cent of India’s e-commerce market. In February last year, a Reuters investigation found that Amazon India was bypassing FDI regulation while giving undue preference to their own sellers over small merchants on their site. During Bezos’ last visit, Goyal remained unimpressed by his promises of investment and refused to meet the Amazon boss. ONDC’s platforms to network approach, Source: Research paper A day prior to the launch of ONDC, the Competition Commission of India or CCI raided Amazon and Flipkart over unethical practices that were creating a divide between corporate sellers and small businesses on their platforms. Meanwhile, ONDC’s platform is expected to expand from five cities on the first day to a hundred cities within the next six months. It aims to connect 30 million sellers and 10 million online merchants by August. State-owned banks – State Bank of India picked up a 7.84 per cent stake worth Rs. 10 crores, and Punjab National Bank picked up 9.5 per cent equity worth Rs. 25 crores. This followed Kotak Mahindra Bank, Axis Bank and HDFC Bank, all of whom lined up to buy equity valued at Rs. 10 crores each. Earlier this week, news broke that Google was in talks to join ONDC. Google was planning to integrate its shopping services, which currently work as an aggregator of online listings. While the partnership hasn’t yet been confirmed, a Google spokesperson said, “We remain committed to focusing on the enablement of small and medium businesses to leverage digital for deeper discovery and payments capabilities with Google Pay.” Other partners of ONDC, including Indian fintech company Paytm, will show listings from each other on their platform in search results. Challenges with ONDC As funding flows in with ease and the intentions themselves are noble, there are questions being raised about how ONDC will reach its goals. India still doesn’t have uniform regulations around e-commerce customer liability. For Amazon and Flipkart, in case of a customer complaint regarding the delayed delivery of a product or a damaged product, the onus rests on the platform and not the seller. We don’t quite know how ONDC plans to deal with this challenge. Can small Kirana stores on the platform keep up with bigger sellers when it comes to offering discounts? All of this is to ultimately ask, can ONDC actually level the playing field for small local shops and bigger stores? Praveen Khandelwal, chief of CAIT and member of the advisory council for ONDC, voiced his concerns, saying while the basic concept of the platform can be appreciated, we are far from reaching clarity around the mechanism. “There are more than eight crore traders in the country. It will not be easy to bring them to the platform and explain the relevance of ONDC to them. There is no credibility of ONDC in the minds of Kirana shop owners and local traders. There is no point for them to move to the platform,” Khandelwal says. This is added to the fact that there is an absence of basic regulations around e-commerce policy; Khandelwal is sceptical about the reasons behind the shift when e-commerce has already depreciated the environment for local businesses in the country. Websites like Amazon and Flipkart have been able to leverage upon a smooth system that incorporates warehousing facilities, payment gateway, returns and refunds and a backend that helps their sites run without any glitches. Anil Kumar, chief executive officer of Redseer Management Consulting, spoke to Bloomberg, saying, “The government needs to build something comparable or better if it wants to outdo the dominant e-commerce platforms. Everything hinges on the network bringing the widest set of buyers, sellers, payment logistics and warehousing providers, and so on.” There is a flipside to the present movement towards over-regulation in the country. India’s small businesses form the backbone of the country, accounting for almost one-quarter of the USD 2.9 trillion economy. These MSMEs employ more than 500 million people. The unregulated cash that is generated from these local mom-and-pop stores serves to cushion the impact of macroeconomic financial crises like the 2008 recession. Protected by its informal sector, the Indian economy grew at a 6.7 per cent rate between 2008 and 2009, which was comparatively far better than other countries. Between 2009 and 2010, the economy bounced back quickly to 8.5 per cent. A strategy paper released by ONDC in January discussed some of the issues that crop up in large e-commerce platforms. Most Kirana stores and MSMEs run their businesses on a foundation of trust with their customers and earn a loyal clientele. This reputability cannot be displayed on an e-commerce platform. For small sellers to maintain entirely separate infrastructures and processes for online and offline sales is difficult and eats into their costs, which might not be affordable for many. Co-founder and non-executive chairman of Infosys, Nandan Nilekani Hiccups with AADHAR Ironically, Nilekani spoke at the Amazon Smbhav Summit in April last year about the importance of India’s small businesses for the country’s future, saying, “Folks like Amazon are going to give access to markets, both domestically and globally, increasing digitisation of our small businesses.” He continued complimenting Amazon for creating job opportunities for millions of Indians, stating, “I believe all of India’s businesses should aspire to be large, and with partners like Amazon, I am sure they will do it.” In many ways, the issues dogging ONDC could be the ones that were attached to past projects involving Nilekani. AADHAR has been mired in controversies related to identity and privacy. Despite its ubiquitous nature, the effectiveness of AADHAR is shrouded under a veil. Intentions to speed up the payments of MNREGA wages have reportedly failed largely. A report by LibTech India that analysed more than 18 lakh wage invoices for the first half of 2021-22 showed that 71% of the payments were delayed beyond the time period specified by the Union government. A new study conducted on more than 3,000 MNREGA workers showed that 57% of work IDs of genuine workers had been removed to demonstrate 100% linking of AADHAR. While AADHAR took six years to be implemented, ONDC might unfold much faster considering the pace it is unfolding at. The only question that remains is to what end?","excerpt":"For small sellers to maintain entirely separate infrastructures and processes for online and offline sales is difficult and eats into their costs, which might not be affordable for many.","categories":["IT Services"],"tags":["Google","NANDAN NILEKANI","ondc"],"author_name":"Poulomi Chatterjee","publish_date":"2022-06-02T17:00:00","publication_year":"2022","word_count":1325,"keywords":["Go","API","funding","Rust","programming_languages:R","ondc","AI","Git","RAG","Aim","Google","NANDAN NILEKANI","R"],"extracted_tech_keywords":["AI","Aim","RAG","R","Go","Rust","Git","API","funding","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/it-services\/can-ondc-deliver\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":50538,"title":"Meet Sudhanshu Kumar, A Techie Who Was Inspired By Detective Stories To Become A Data Scientist","content":"There is no denying that it is not just the work that is intriguing in data science, but also the perks and privileges. But do we know what it takes to become a data science professional? How one needs to put in their time and efforts to grasp all the knowledge? In order to know their side of the story, every week we get in touch with some of the top data scientists from the industry and take a look at their journey. For this week’s ‘My Journey In Data Science’ we got in touch with Sudhanshu Kumar, CEO and Chief AI Engineer at iNeuron.ai. Whether it’s statistical modelling, data processing, data mining, machine learning and deep learning algorithms, time series, computer vision or natural language processing, Sudhanshu has worked with almost every aspect of data science to solve challenging business problems. The Onset A computer science graduate with a PGDM in Statistics, Sudhanshu had no peer pressure during his educational journey. It is all out of his interest that he chose Data Science. Talking about the exact instance when he fell in love with data science, Sudhanshu says that he has always been fascinated with detective stories. “Mysteries have kept me awake at night. How the clues lead to a certain conclusion, has always mesmerised me,” he said. And he believes that if we take a closer look, that’s what data science is all about. It’s a detective story where you are given all the facts, and then you need to draw a conclusion out of it. He didn’t need any other reason to fall in love with data science, he was completely into it. Furthermore, Sudhanshu has always had a deep curiosity about what’s happening in the technological world and hence, he keeps on reading whatever new stuff is happening in this field to keep himself updated and intrigued. “I keep on reading journals and research papers on a regular basis and I make sure that whatever I learn, I build an end to end solution using that so that I have a thorough knowledge of the topic,” said Sudhanshu. First Job, First Interview, First Paycheck The Chief AI Engineer started his journey in the data science with Indian tech giant Wipro. He said that the experience he gained over time with Wipro was very enriching. “ It was the start of my career and I had nothing to lose but only to learn,” Sudhanshu added. He said that he tried to learn as much as possible and all his seniors and colleagues helped me a lot during his tenure at Wipro. While Sudhanshu talked about his data science journey, he also shared his first interview experience. He said that it was a very important learning experience for him. “Interviews give you an opportunity to ponder upon what’s the skill or knowledge that you have been lacking, and then you try to be better than what you are,” said Sudhanshu. Talking about his first salary when he started as a data scientist, Sudhanshu said that he already had 2 years of industry experience. And it was an added advantage for him. “The first salary that I drew as a data scientist was more than ₹12 LPA,” Sudhanshu added. I get inspired by the people around me, my colleagues, my juniors, my students, and all those great minds who are working tirelessly in the field of data science, who are innovating new things, who are changing the world for a better tomorrow. Strategies And Struggle Talking about the strategies to land a job in data science, Sudhanshu believes that  the first thing that one needs to focus on is to figure out the exact work one wants to do as a data science professional — whether you want to be a Machine Learning expert or, an expert in solving vision problems or, a champion of NLP.  Therefore, it is imperative that one should do thorough research on the different job roles this domain provides. And then based on the preference, try to find out the right companies that are actually working in those fields and then try to look for opportunities in them. And when it comes to the struggle, Sudhanshu said the real struggle is real and it is more when one has to deal with his\/her mind after getting rejected. There is no denying that it is hard. “When you have been a star performer wherever you go, then you’re not used to rejections,” said Sudhanshu. “You have been appreciated your entire life and these sudden rejections, they make you think, what did I do wrong?” But he believes that this is the time when one should take a step back, crouch, keep on building momentum, and then pounce at the first opportunity that gets. I’d say that’s what life is. To be pushed into the corner and then again coming back strong.” A Piece Of Advice To All The Aspiring Data Science Professionals According to Sudhanshu, the reason to be a data scientist should be the urge to learn from the vast data that the world is generating per second. Data shows you the past and as well as what’s going to be. He believes that if this thrills you, then data science is definitely the field for you. What should one do to stay relevant in this field? To that, the iNeuron CEO says that the landscape of the industry is changing rapidly. The field of AI today is not what it was 5 years ago and definitely, there is going to be a tectonic shift in the coming decade. So, as data scientists, you need to keep yourself updated to the changes happening around. As the saying goes, ‘If you don’t do AI, what do you do?’ “That’s how the technological landscape is going to be. In short, just keep learning, keep exploring and keep surprising yourself with your hard work,” said Sudhanshu in conclusion.","excerpt":"There is no denying that it is not just the work that is intriguing in data science, but also the perks and privileges. But do we know what it takes to become a data science professional? How one needs to put in their time and efforts to grasp all the knowledge? In order to know […]","categories":["AI Features"],"tags":["data science interview","data science professionals","Data Scientist","how to become a data scientist","Interviews and Discussions","My Journey In Data Science"],"author_name":"Harshajit Sarmah","publish_date":"2019-11-22T19:40:10","publication_year":"2019","word_count":991,"keywords":["data science professionals","data science","Go","how to become a data scientist","data science interview","machine learning","API","AI","My Journey In Data Science","computer vision","NLP","deep learning","Tecton","Data Scientist","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","machine learning","deep learning","NLP","computer vision","data science","Tecton","R","Go","API"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/meet-sudhanshu-kumar-a-techie-who-was-inspired-by-detective-stories-to-become-a-data-scientist\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10049634,"title":"Facebook Gives Away Captum For Free: Its Model Interpretability Library For PyTorch","content":"Concept-based interpretability tools assist artificial intelligence researchers and engineers in designing, developing and debugging AI models. Additionally, these tools help understand the working of AI models, helping businesses assess if the models deliver accurate results and reflect their values. One such interpretability tool is Facebook’s Captum. Captum is a powerful and flexible interpretability library for PyTorch, which makes algorithms for interpretability readily available for access to the PyTorch community. Captum supports model interpretability across modalities — vision, text; additionally allowing researchers to add new algorithms and benchmark their work against existing algorithms available in the library. Finally, it offers tools to help developers uncover vulnerabilities using metrics and adversarial attacks. Recently, Facebook released its latest version of Captum — Captum 0.4, which has new functionality for model understanding. Facebook AI has added tools to evaluate model robustness, improvements to its existing attribution models, and new attribution methods in the latest version. Removal of Statistical Biases Captum 0.4 adds testing with Concept Activation Vectors (TCAV) that allows researchers and engineers to assess how different user-defined concepts affect a model’s prediction. It can also be used for checking algorithmic and label bias, which might be embedded in networks. Additionally, TCAV’s capabilities expand beyond currently available attribution methods, enabling researchers to quantify the importance of the various inputs and quantify the impact of concepts like gender and race on a model’s prediction. Captum 4.0 comes with generically implemented TCAV, allowing users to define custom concepts with example inputs for different modalities — vision and text. Source: Facebook AI The graphs above showcase visualised distributions of TCAV scores for the sensitivity analysis model implemented in a Captum tutorial. As a data set, Facebook AI researchers have used movie ratings with positive sentiment. The graphs visualise TCAV scores for positive adjectives concepts along with five sets of neutral terms concepts. The positive adjectives concept is more important than for both convolutional layers across all the five neutral concept sets, indicating the importance of positive adjectives in predicting positive sentiment. Robust AI Models Deep learning techniques are often vulnerable to adversarial inputs that, in turn, can fool the AI model and be imperceptible to humans. Captum 0.4 comes with tooling to support the improved understanding of the limitations and vulnerabilities of a model. As a result, the AI system will react to unforeseen issues and make necessary changes to avoid harming or otherwise negatively affecting people. Captum 0.4 also comes with tools to understand the robustness of the model, including implementations of adversarial attacks and robustness metrics to evaluate the impact of different attacks or perturbations on a model. The robustness metrics included in its latest version are: Attack Comparator: It allows users to quantify input perturbation’s impact. This includes text augmentation, and torchvision transforms. It also helps quantify adversarial attacks on a model and compare the impact of the various attacks. Minimal Perturbation: It identifies the minimum perturbation required to cause a model to misclassify the perturbed input. This new tooling enables developers to understand potential model vulnerabilities better and analyse counterfactual examples to comprehend a model’s decision boundary better. Source: Facebook AI Relevance Propagation & Attribution Facebook AI has implemented a new attribution algorithm, in collaboration with Technische Universitat Berlin, to offer a new perspective for explaining model predictions. Captum 0.4 adds both LRP and a layer-attribution variant — layer LRP. The Layer-wise Relevance Propagation (LRP) algorithm is based on a backward propagation mechanism applied sequentially to all layers of the model. The model output score represents the initial relevance which is then decomposed into values for each neuron of the underlying layers. Additionally, in Captum 0.4, the Facebook AI team has added tutorials, improvements and bug fixes to the existing attribution methods.","excerpt":"Facebook AI released Captum 0.4 with new functionality for model understanding.","categories":["AI Features"],"tags":["Facebook AI","Facebook AI research","PowerBI","Pytorch"],"author_name":"Debolina Biswas","publish_date":"2021-09-24T11:00:00","publication_year":"2021","word_count":619,"keywords":["Pytorch","Go","artificial intelligence","TPU","Facebook AI","AI","PyTorch","model interpretability","PowerBI","deep learning","ViT","Facebook AI research","R","adversarial attacks"],"extracted_tech_keywords":["AI","artificial intelligence","deep learning","PyTorch","TPU","R","Go","ViT","model interpretability","adversarial attacks"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/facebook-gives-away-captum-for-free-its-model-interpretability-library-for-pytorch\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10106498,"title":"Apple Quietly Unveils Open-Source Multimodal LLM, Ferret","content":"Two months back, Apple, in collaboration with Columbia University, quietly unveiled Ferret, a new multimodal large language model adept at referring and grounding. Check out the GitHub repository here. Ferret can refer to image regions in any free-form shape and automatically establish grounding for text deemed groundable by the model. I somehow missed this. @Apple joined the open source AI community in October. Ferret’s introduction is a testament to Apple’s commitment to impactful AI research, solidifying its place as a leader in the multimodal AI space. Way to go @Apple – ps: I'm looking forward to the day… https:\/\/t.co\/Pi1kQrsVvx— Bart de Witte (@OpenMedFuture) December 23, 2023 The researchers have curated the GRIT dataset for model training. The dataset includes 1.1 million samples that contain rich hierarchical spatial knowledge, with 95K hard negative data to promote model robustness. They also said that the resulting model achieved superior performance in classical referring and grounding tasks and greatly outperformed existing MLLMs in region-based and localisation-demanded multimodal chatting. “Our evaluations also reveal a significantly improved capability of describing image details and a remarkable alleviation in object hallucination,” the researchers said that Ferret, like most MLLMs, may produce harmful and conunterfactual reponses. Citing LISA, the researchers said they plan to enhance Ferret to output segmentation masks and bounding boxes. The Significance of Ferret’s Stealthy Debut Apple’s strategic move to release Ferret without a formal announcement speaks volumes about the company’s dedication to staying at the forefront of multimodal AI. The unexpected embrace of open-source development departs from Apple’s traditional closed-door approach, setting the stage for potential collaboration and community-driven advancements. Talking about its functionality, it is elegantly simple yet powerful. Beyond identifying elements within an image, the model draws connections to formulate responses to user queries. This opens up possibilities in image search, accessibility, and other applications where nuanced contextual understanding is crucial. Ferret’s versatility is further enhanced by a spatial-aware visual sampler capable of handling various sparsity patterns associated with different shapes. Ferret accommodates diverse regional inputs, including points, bounding boxes, and free-form shapes.","excerpt":"Ferret can refer to image regions in any free-form shape and automatically establish grounding for text deemed groundable by the model.","categories":["AI News"],"tags":["Apple"],"author_name":"Sandhra Jayan","publish_date":"2023-12-26T12:55:24","publication_year":"2023","word_count":340,"keywords":["Go","TPU","AWS","AI","ETL","Apple","ML","Git","multimodal AI","GitHub","R"],"extracted_tech_keywords":["AI","ML","multimodal AI","AWS","TPU","R","Go","Git","GitHub","ETL"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/apple-quitely-unveils-open-source-multimodal-llm-ferret\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10173125,"title":"OpenAI Poaches Cracked Engineers from Tesla, xAI, and Meta","content":"OpenAI has hired four senior engineers from xAI, Tesla, and Meta, as reported by WIRED on July 8, citing a Slack message sent by OpenAI’s co-founder, Greg Brockman. The company has hired David Lau, former VP of software engineering at Tesla; Uday Ruddaraji, the former head of infrastructure engineering at xAI and X; Mike Dalton, an infrastructure engineer from xAI; and Angela Fan, an AI researcher from Meta. Ruddarraju and Dalton worked on building xAI’s Colossus, the supercomputer comprising more than 200,000 GPUs. Soon, Brockman confirmed the same in a post on X. welcome david lau, mike dalton, uday ruddarraju, and angela fan! https:\/\/t.co\/NkK5ynHv85— Greg Brockman (@gdb) July 9, 2025 “We’re excited to welcome these new members to our scaling team,” said OpenAI spokesperson Hannah Wong, as quoted by WIRED. “Our approach is to continue building and bringing together world-class infrastructure, research, and product teams to accelerate our mission and deliver the benefits of AI to hundreds of millions of people.” The development occurs at a time when Meta has been actively recruiting several key engineers and researchers from OpenAI to form a ‘superintelligence’ team. Reports suggest that some of these individuals received signing bonuses of approximately $100 million. Most recently, Bloomberg reported that Meta hired Yuanzhi Li, a researcher from OpenAI, on July 7. Commenting on this active poaching between the two companies, Sam Altman, CEO of OpenAI, in an interview with Bloomberg, said, “Obviously, some people will go to different places. There’s a lot of excitement in the industry,” and indicated that he feels ‘fine’ about the departures. However, Joaquin Quiñonero Candela, OpenAI’s head of recruiting, said on X a few days ago, “It’s unethical (and reeks of desperation) to give people ‘exploding offers’ that expire within hours, and to ask them to sign before they even have a chance to tell their current manager. Meta, you know better than this. Put people first”","excerpt":"The company, having lost several important personnel to Meta over the last few weeks, has now poached engineers from its competitors.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","OpenAI"],"author_name":"Supreeth Koundinya","publish_date":"2025-07-09T15:02:09","publication_year":"2025","word_count":316,"keywords":["Go","OpenAI","AI","programming_languages:R","programming_languages:Go","XAI","xAI","AI research","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","OpenAI","xAI","R","Go","XAI","AI research","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/openai-poaches-cracked-engineers-from-tesla-xai-and-meta\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":12368,"title":"Microfinance in India &#8211; adoption of machine learning is still underway","content":"Micro-finance startups are promoting financial inclusion in India but the core technology behind it, machine learning still has to mature They are billed as India’s sunrise sector by numerous studies and have cornered the lion’s share when it comes to venture capital investments, but it is also an industry where “failure chances are very high and data is very important and at the heart of practice”. We are talking about B2C micro-lending companies and their phenomenal growth. On the back of it are state-of-the-art machine learning algorithms, external data sources which are parsed and automated risk clustering that’s feeding the ambitions of young millennials.  Furthering their usefulness as an alternate credit channel is Prime Minister Modi’s demonetization move. But let’s take a step back and dial down the euphoria a bit. Though there is a lot of brouhaha over the shiny new thing called machine learning deployed for crunching numbers and assessing creditworthiness in under 24 hours, data collection in India is still at a very nascent stage.  And when it comes to data points, industry insiders doubt the efficacy of social media as a right indicator.  In fintech, with great promise comes great risks. Let’s list down some of the grey areas: a) Fintech companies are not breaking any new ground when it comes to dispensing real time decisions for underwriting. Banks have been providing scorecards too, their loan eligibility calculator checked the credit card history and transactional history. b) Though banks haven’t started tinkering with machine learning technology as yet, banks are frontrunners in deploying analytics for risk management and cross-selling. Axis Bank has built a robust risk management strategy that helps capture lost data and preempt delinquencies. Case in point, Axis Bank deploying SAS risk management framework. Another first of its kind for Axis Banks is the establishment of Thought Factory – an innovation lab that will collaborate with startups in exploring emerging technologies to making banking more robust. SAS analytics suite also powers credit card cross selling for HDFC bank. c)  Young millennials (we are talking about the 18-something college going crowd who are extended credit lines) form the core demographic of some micro-lending companies. Circa 2007-08, the credit card boom was on its rise. Finomena co-founder Abhishek Garg “It was a time when Citibank and Barclays started an aggressive campaign for credit card adoption and went to college campuses giving it away for free. The result, students couldn’t pay back the credit. They weren’t dumb, they just didn’t have the cash to pay back. Barclays suffered so badly they exited the Indian market and Citibank got burnt with this strategy,” shares Finomena co-founder Abhishek Garg, Bangalore headquartered microfinance startup that caters to strictly young millennials and not the college going crowd. d)  Here’s a realistic view from the Finomena co-founder, Garg, who has a formidable banking background. Admittedly, machine learning algorithms still have to mature and factor in several other variables before they churn out foolproof decisions. Time frame – give it another 3-4 years to truly become sophisticated. When it comes to Finomena, they are treading the zone cautiously, catering to a) salaried class and b) defining the segment – electronics consumers good that has a huge market size. Though microfinance startups have an edge over traditional lending by leveraging sophisticated machine learning algorithms, automating decisions in seconds, the technology still has to mature. How are these startups different from Traditional Lending? Slicepay co-founder and CEO Rajan Bajaj Slicepay co-founder and CEO Rajan Bajaj shares how his startup uses structured data for measuring credit worthiness of the applicant. And the time span – “2 hours median time”. “Currently we are using judgmental scoring model wherein most of the insights received over last one year have been converted into weights and automated. Machine learning based model that aims to automate everything and make approval real time would be intelligent to learn these things quickly and iterate model,” he shared. Finomena co-founder Riddhi Mittal, a Stanford alumna with past stints at Microsoft and Facebook boasts of a glittering resume. In a chat with Analytics India Magazine, she gives an under-the-hood peek on the goings-on at Finomena and how the next gen underwriting is being driven with two very pertinent questions around datasets and algorithms. Mittal explains how her startup is factoring in real world deep challenges, navigating and building models around them every day. She goes on to explain how machine learning is used to ascribe a more accurate scoring from large amounts of data. Analyzing different data points and finding insights This is a trick which is in every micro-lending startup’s handbook – analyzing data points, spanning behavioural, social that comes under unstructured data type and financial data. But even though decisions are dispensed in real time, the sophisticated algorithms are also “still learning and lack human intuition” and could prove wrong in certain cases.  However, as the user base expands, non-traditional risk-assessment algorithms will improve. Finomena co-founder Riddhi Mittal “Data is everywhere. The real trick lies in how can we parse it, make sense of it and extract insights. Then, the next step is usually what insights should be prioritized. How to create as many variables and granularity in these variables,” shares Mittal giving a low down of the supervising process. Factoring in new risks in the portfolio In the world of data science, the general sentiment is you are only as good as your data, but sadly, most companies do not have access to large quantities of data that can lend deeper insights in your risks. Secondly in India, penetration of internet is still nascent with 34.8% of people having access to internet. “Just imagine, if top 5% have access and the other 95% is unaccountable, so you can’t apply and extrapolate to other 95%. Which means that large datasets are not available and there is so much diversity in how people behave, you can’t apply insights to a larger set,” she shares. Her approach is simple — a) giving loans intelligently b) waiting for feedback. “Lending means a feedback loop. And if we want to extend the feedback loop, you can’t scale it up too fast either,” she says.  A feedback loop is an economic term, set off by a chain of events, reinforced in a positive situation and self-regulating in a negative situation. Another huge roadblock to be factored in is online fraud which is on the rise. Several factors such as providing fake IDs, creating fake social identities, poor double checking practices and borrowers becoming untraceable have to be considered during risk analysis.  “When it comes to lending fraud, we have to detect what kind of fraud we need to detect through supervised and unsupervised learning methods,” Mittal cites. Given the high volumes of data generated, historical records and the changing nature of finance, microfinance startups are fuelling ML techniques, (thanks in part to high computing power) and perhaps developing more use cases than any other industry. Indian finance sector is definitely at the intersection of AI and Machine Learning, with sophisticated chat bots and apps coming into play.","excerpt":"They are billed as India’s sunrise sector by numerous studies and have cornered the lion’s share when it comes to venture capital investments, but it is also an industry where “failure chances are very high and data is very important and at the heart of practice”. We are talking about B2C micro-lending companies and their […]","categories":["IT Services"],"tags":["credit risk machine learning","fintech analytics India"],"author_name":"Richa Bhatia","publish_date":"2017-01-30T06:32:16","publication_year":"2017","word_count":1179,"keywords":["data science","Go","API","machine learning","fintech analytics India","AI","credit risk machine learning","ML","RAG","Aim","analytics","R"],"extracted_tech_keywords":["AI","machine learning","ML","data science","analytics","Aim","RAG","R","Go","API"],"url":"https:\/\/analyticsindiamag.com\/it-services\/microfinance-india-adoption-machine-learning-still-underway\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10011367,"title":"A Beginner’s Guide To Neural Network Modules In Pytorch","content":"Pytorch is a deep learning library which has been created by Facebook AI in 2017. It is prominently being used by many companies like Apple, Nvidia, AMD etc. You can read more about the companies that are using it from here. It is also often compared to TensorFlow, which was forged by Google in 2015, which is also a prominent deep learning library. You can read about how PyTorch is competing with TensorFlow from here. There are a lot of functions and explaining each of them is not always possible, so will be writing a brief code that would explain it and then would give a simple explanation for the same. If you want to read more about it, click on the link that is shared in each section. Installation Installation command is different for different OS, you can check the best one for you from here. #dependency import torch We’d have a look at tensors first because they are really important. Let us understand what a tensor is. Tensor is in simple words is a multidimensional array which is also generalised against vectors and matrices. Now let us see what all things can we do with it. # a few lucid examples of tensor a=torch.tensor(3) a=torch.tensor([1,3]) a=torch.tensor([[1,2],[3,4]]) # a tensor, a bit complex one tensor = torch.Tensor( [ [[1, 2], [3, 4]], [[5, 6], [7, 8]], [[9, 0], [1, 2]] ] ) tensor.shape #to find the shape of the tensor torch.Size([3, 2, 2]) Index of tensors tensor[1] Output tensor([[5., 6.], [7., 8.]]) There are a lot of other functions for which you can refer to the official documentation which is mentioned at the last of this article. Initialising like tensors Like tensors are the ones which have the same shape as that of others. torch.ones_like(tensor) tensor([[[1., 1.], [1., 1.]], [[1., 1.], [1., 1.]], [[1., 1.], [1., 1.]]]) Here the shape of this would be the same as that of our previous tensor and all the elements in this tensor would be 1. torch.zeros_like(example_tensor) tensor([[[0., 0.], [0., 0.]], [[0., 0.], [0., 0.]], [[0., 0.], [0., 0.]]]) All the elements of this tensor would be zero. #here we would be creating a tensor whose every element would be a normal distribution. torch.randn_like(tensor) tensor([[[-0.3675,  0.2242], [-0.3378, -1.0944]], [[ 1.5371,  0.7701], [-0.1490, -0.0928]], [[ 0.3270,  0.4642], [ 0.1494,  0.1283]]]) Let us take a look at some basics operations on Tensors (tensor - 5) * 2 Output tensor([[[ -8.,  -6.], [ -4.,  -2.]], [[  0.,   2.], [  4.,   6.]], [[  8., -10.], [ -8.,  -6.]]]) To read more about tensors, you can refer here. You can have a look at Pytorch’s official documentation from here. We will see a few deep learning methods of PyTorch. Pytorch’s neural network module #dependency import torch.nn as nn nn.Linear It is to create a linear layer. Here we pass the input and output dimensions as parameters. Here it is taking an input of nx10 and would return an output of nx2. linear = nn.Linear(10, 2) example_input = torch.randn(3, 10) example_output = linear(example_input) example_output Output: tensor([[ 0.2102,  0.5055], [-0.5417,  0.8288], [ 0.1755,  0.3779]], grad_fn=<AddmmBackward>) nn.Relu It performs a relu activation function operation on the given output from linear. relu = nn.ReLU() relu_output = relu(example_output) relu_output Output: tensor([[0.2900, 0.0000], [0.4298, 0.4173], [0.4861, 0.0000]], grad_fn=<ReluBackward0>) nn.BatchNorm1d It is a normalisation technique which is used to maintain a consistent mean and standard dev among different batches of the of input. batchnorm = nn.BatchNorm1d(2) batchnorm_output = batchnorm(relu_output) batchnorm_output Output tensor([[-1.3570, -0.7070], [ 0.3368,  1.4140], [ 1.0202, -0.7070]], grad_fn=<NativeBatchNormBackward>) You can read about batchnorm1d and batchnorm2d from their official doc. nn.Sequential It is to create a sequence of operations in one go. mlp is the name of variable which stands for multilayer perceptron. mlp_layer = nn.Sequential( nn.Linear(5, 2), nn.BatchNorm1d(2), nn.ReLU() ) test_example = torch.randn(5,5) + 1 print(\"input: \") print(test_example) print(\"output: \") print(mlp_layer(test_example)) Output input: tensor([[ 1.7690,  0.2864,  0.7925,  2.2849,  1.5226], [ 0.1877,  0.1367, -0.2833,  2.0905,  0.0454], [ 0.7825,  2.2969,  1.2144,  0.2526,  2.5709], [-0.4878,  1.9587,  1.6849,  0.5284,  1.9027], [ 0.5384,  1.1787,  0.4961, -1.6326,  1.4192]]) output: tensor([[0.0000, 1.1865], [1.5208, 0.0000], [0.0000, 1.1601], [0.0000, 0.0000], [0.7246, 0.0000]], grad_fn=<ReluBackward0>) How nn.Sequential is important and why it is needed, read it from here. Optimisers import torch.optim as optim adam_opt = optim.Adam(mlp_layer.parameters(), lr=1e-1) Here lr stands for learning rate and 1e-1 means 0.1 #now let us look at the training loop train_example = torch.randn(100,5) + 1 adam_opt.zero_grad() # We'll use a simple loss function of mean distance from 1 # torch.abs takes the absolute value of a tensor cur_loss = torch.abs(1 - mlp_layer(train_example)).mean() cur_loss.backward() adam_opt.step() print(cur_loss) A little bit of theory: requires_grad_() This means that even if PyTorch wouldn’t normally store a grad for that particular tensor, it will for that specified tensor. with torch.no_grad(): PyTorch will usually calculate the gradients as it proceeds through a set of operations on tensors. This can often take up unnecessary computations and memory, especially if you’re performing an evaluation. However, you can wrap a piece of code with torch.no_grad() to prevent the gradients from being calculated in a piece of code. detach(): Sometimes, you want to calculate and use a tensor’s value without calculating its gradients. For example, if you have two models, A and B, and you want to directly optimise the parameters of A with respect to the output of B, without calculating the gradients through B, then you could feed the detached output of B to A. There are many reasons you might want to do this, including efficiency or cyclical dependencies (i.e. A depends on B depends on A). We are now making the nn class. class ExampleModule(nn.Module): def __init__(self, input_dims, output_dims): super(ExampleModule, self).__init__() self.linear = nn.Linear(input_dims, output_dims) self.exponent = nn.Parameter(torch.tensor(1.)) def forward(self, x): x = self.linear(x) # This is the notation for element-wise exponentiation, # which matches python in general x = x ** self.exponent return x example_model = ExampleModule(10, 2) list(example_model.parameters()) Output [Parameter containing: tensor(1., requires_grad=True), Parameter containing: tensor([[ 0.2789,  0.2618, -0.0678,  0.2766,  0.1436,  0.0917, -0.1669, -0.1887, 0.0913, -0.1998], [-0.1757,  0.0361,  0.1140,  0.2152, -0.1200,  0.1712,  0.0944, -0.0447, 0.1548,  0.2383]], requires_grad=True), Parameter containing: tensor([ 0.1881, -0.0834], requires_grad=True)] This is the output of the class that we had created: input = torch.randn(2, 10) example_model(input) Conclusion The aim of this article is to give briefings on Pytorch. We had discussed its origin and important methods in it like that of tensors and nn modules. There’s a lot to it and simply isn’t possible to mention everything in one article. That is why it is kept concise, giving you a rough idea of the concept. If you want to read more about it, you can read the official documentation thoroughly from here. Hope you liked the article","excerpt":"A Beginner’s Guide To Neural Network Modules In Pytorch writing a brief code that would explain it and then would give a simple explanation","categories":["Deep Tech"],"tags":["Machine Learning Algorithms","Neural Networks","Python","Pytorch","tensorflow gradient"],"author_name":"Bhavishya Pandit","publish_date":"2020-11-09T13:00:38","publication_year":"2020","word_count":1106,"keywords":["Machine Learning Algorithms","Pytorch","TPU","AI","neural network","PyTorch","ML","Python","Ray","Aim","deep learning","tensorflow gradient","TensorFlow","Neural Networks"],"extracted_tech_keywords":["AI","ML","deep learning","neural network","Aim","Ray","TensorFlow","PyTorch","TPU","Python"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/a-beginners-guide-pytorch\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":36040,"title":"Biased Algorithms, Sexism in AI &#038; Need For Diversity Were 3 Big Takeaways From The Rising 2019","content":"Vaishali Kasture, The Rising 2019 The Rising 2019, held on this International Women’s Day was one of the largest gatherings of women in the field of analytics, data science and AI. The conference hosted a diverse set of speakers who shared their professional experiences and talked about their growth story and challenges faced as women. The keynote began by Saraswathi Ramachandra, Head of Analytics Center of Excellence at Danske IT, who kick-started the talk about how emerging technology today such as AI is turning out to be sexist. Ramachandra who strongly believes in merging technology with women made a slew of relevant points about how sexism and gender bias in humans is leading to technologies such as AI being biased. She focused on three core points — is there a bias, is AI enabling the bias, what can we do about it? Saraswathi Ramachandra, Head of Analytics Center of Excellence at Danske IT Addressing these three core areas she began by highlighting how we often stereotype jobs for women. For instance, if we talk about nurse, homemakers or others, the first thing that comes to our mind is that it would be a woman. “We all have biases and it is this bias that AI inherits and learns from it,” said Ramachandra. This bias is creating a skewed dataset on which these AI models are trained. Is AI sexist? She mentioned how some of the most noted chatbots and voice assistants such as Alexa, Cortana or Siri have a female voice. Not just this, but some of the newer applications such as Google Translator, or the use of AI for hiring have also exhibited biased results. The AI recruiting tools by Amazon was recently stopped as it started creating a bias against women candidates. Similar instances are seen in facial recognition systems or even in making business decisions. For instance, in banking, the systems generate less credit rate of women who earn equal to men and loans of women or single women are often denied as the data on which these are trained is bias. “We have a bias, and AI inherits it,” said Ramachandra. A similar sentiment was echoed by Vaishali Kasture, Co-Founder at Sonder Connect who has a diverse experience of working with AI on the business side of it. Having worked with AI in areas such as enabling financial inclusion for women or automating businesses processes with AI, she shared a similar point as to how women are often declined a credit score. “The problem she stated is with the AI models that have been trained on historical data that may not be completely relevant for today’s age. What happened in past is not a prediction of what happens in the future. Historical data is not always right to train ML models,” she said. For instance, drug trials are failing as they don’t have women insights and data. The same is happening with a lot of other things such as self-driving cars, recruiting etc. She stressed on the need to have a diversity in data. Smitha Ganesh, Principal Consultant – Data Scientist at ThoughtWorks also spoke about fighting Prejudice in artificial intelligence. She said, “AI isn’t dangerous but the human bias is! One must be cognizant about any unintended consequences of using this technology, “ she said. There may be a compounding effect during the emergence of AI, as the algorithms exhibit self-learning from data. How Can Technological Sexism Be Combated? Both Ramachandra and Kasture believe that the best way for combating sexism is to have diverse data science teams. “It is important to analyse data and not make it a black box. The datasets should be analysed and filtered while ensuring that dataset is large, diverse and accurate,” said Ramachandra. Ganesh too stressed on a diversity-inclusive mindset for data collection to fight this prejudice and taking AI to an elevated level, recognising the vitality involved with AI training systems. “If we need to make AI safe in the future it is important to develop a way where AI models can have a blue tick mark, similar to how social media have for verified accounts, only this time the verification should be given by women, said Kasture in closing.","excerpt":"The Rising 2019, held on this International Women’s Day was one of the largest gatherings of women in the field of analytics, data science and AI. The conference hosted a diverse set of speakers who shared their professional experiences and talked about their growth story and challenges faced as women. The keynote began by Saraswathi […]","categories":["AI Features"],"tags":["Women in Analytics","Women in Data Science"],"author_name":"Srishti Deoras","publish_date":"2019-03-11T04:35:05","publication_year":"2019","word_count":702,"keywords":["data science","Go","Women in Data Science","artificial intelligence","AI","chatbots","ML","ViT","analytics","Women in Analytics","GAN","R"],"extracted_tech_keywords":["AI","artificial intelligence","ML","data science","analytics","chatbots","R","Go","GAN","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/biased-algorithm-sexism-ai-need-for-diversity-were-the-3-big-takeaways-from-rising-2019\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10069440,"title":"Gradient Ascent: When to use it in machine learning?&nbsp;","content":"In gradient descent, to discover a local minimum of a function, take steps proportional to the negative of the function’s gradient or approximation gradient at the current location. Instead, taking steps proportional to the gradient is positive, one approaches a local maximum of that function; this method is known as gradient ascent. This article will help to understand the Gradient Ascent (GA). Following are the topics to be covered. Table of contents Mathematics behind the Gradient AscentWhen to use gradient ascentImplementing gradient ascent in logistic regression Gradient ascent maximizes the loss function of the algorithm. Let’s start with understanding the mathematics behind GA. Mathematics behind the Gradient Ascent Gradient ascent is based on the principle of locating the greatest point on a function and then moving in the direction of the gradient. In this method, the gradient function is the function of x and y differentiable values. If the coordinates are differentiated for x this means that the gradient function is moving in the x-direction. Similarly, y-direction gradients should be differentiated for y. The function must be specified and differentiable around the places where it is being evaluated. The gradient ascent technique shown in the graphic representation takes a step in the gradient’s direction. The gradient operator will always indicate the direction of the most significant rise. The magnitude, or step size, will be obtained from the parameter value. This phase is continued until a defined number of steps or the algorithm is within a particular tolerance margin. Image source The gradient ascent method advances in the direction of the gradient at each step. The gradient is assessed beginning at point P0, and the function proceeds to the next point, P1. The function then advances to P2 when the gradient is reevaluated at P1. This loop will continue until a stopping condition is fulfilled. The gradient operator always ensures that we are travelling in the best direction feasible. Are you looking for a complete repository of Python libraries used in data science, check out here. When to use gradient ascent Gradient climb operates similarly to gradient descent, with one exception. Its objective is to maximise some function rather than to minimise it. The reason for the distinction is that we may wish to maximise a function rather than minimise it at times; for example, if we want to maximise the distance between separation hyperplanes and observations. In this sense, for each function “f” on which gradient descent is applied, there is a symmetric function “-f” on which gradient ascent may be applied. This suggests that a problem solved using gradient descent can also be solved using gradient ascent if we mirror it on the axis of the independent variable. Implementing gradient ascent in logistic regression For this article, we will use gradient ascent for a logistic regression for a dataset related to social media marketers. The algorithm will predict whether the customer will purchase the product based on different features. For this purpose, we have to build a custom logistic regression algorithm. Let’s start with importing the necessary libraries. import numpy as np import pandas as pd Reading and preprocessing the data data=pd.read_csv('Social_Network_Ads.csv') data[:5] Analytics India Magazine from sklearn.preprocessing import LabelEncoder enc=LabelEncoder() data['Gender_enc']=enc.fit_transform(data['Gender']) X=data.drop(['Gender','Purchased','User ID'],axis=1) y=data['Purchased'] Custom algorithm def sigmoid(self, inX): sig = (1\/(1+np.exp(-inX))) return sig def log_likelihood(self, y_true, y_pred): y_pred = np.maximum(np.full(y_pred.shape, self.eps), np.minimum(np.full(y_pred.shape, 1-self.eps), y_pred)) likelihood = (y_true*np.log(y_pred)+(1-y_true)*np.log(1-y_pred)) return np.mean(likelihood) def fit(self, X, y): m = X.shape[0] n = X.shape[1] self.weights = np.zeros(n) for i in range(self.max_iterations): y_pred = self.sigmoid(X*self.weights) gradient = np.mean((y-y_pred)*X.T, axis=1) self.weights += self.learning_rate*gradient likelihood = self.log_likelihood(y,y_pred) self.likelihoods.append(likelihood) In the sigmoid function, we are calculating the sigmoid for all the data points which will be utilized for the gradient. The likelihood is the cost function for the algorithm. Maximization of the likelihood function is the motive of the algorithm by using the gradient ascent. In the fit function generating weights for the ascent, it would be an array of either zeroes or one and the number of columns would be the same as the number of columns for the independent variable. The prediction would be calculated based on the sigmoid function. The gradient combined with the learning rate will give the final values for the cost function. Since Gradient Ascent is an iterative optimization approach for locating local maxima of a differentiable function. We will iterate the steps for 500 cycles. The method advances in the direction of the gradient generated at each point of the cost function curve until the halting requirements are met. Final code class customclassification: def __init__(self, learning_rate=0.01, max_iterations=500): self.learning_rate = learning_rate self.max_iterations = max_iterations self.likelihoods = [] self.eps = 1e-7 def sigmoid(self, inX): sig = (1\/(1+np.exp(-inX))) return sig def log_likelihood(self, y_true, y_pred): y_pred = np.maximum(np.full(y_pred.shape, self.eps), np.minimum(np.full(y_pred.shape, 1-self.eps), y_pred)) likelihood = (y_true*np.log(y_pred)+(1-y_true)*np.log(1-y_pred)) return np.mean(likelihood) def fit(self, X, y): m = X.shape[0] n = X.shape[1] self.weights = np.zeros(n) for i in range(self.max_iterations): z = np.dot(X,self.weights) y_pred = self.sigmoid(z) gradient = np.mean((y-y_pred)*X.T, axis=1) self.weights += self.learning_rate*gradient likelihood = self.log_likelihood(y,y_pred) self.likelihoods.append(likelihood) def predict_prob(self,X): z = np.dot(X,self.weights) probabilities = self.sigmoid(z) return probabilities def predict(self, X, threshold): predictions = np.array(list(map(lambda x: 1 if x>threshold else 0, self.predict_prob(X)))) return predictions Use custom algorithm LogRegres= customclassification() LogRegres.fit(X_train,y_train) y_hat=LogRegres.predict(X_test,0.5) y_cap = y_hat - y_test count = np.count_nonzero(y_cap==0) accuracy = (count\/len(y_hat))*100 accuracy Analytics India Magazine This accuracy could be further improved by using different data wrangling techniques and by using Stochastic gradient ascent, leaving that to you. Conclusion The gradient is a vector that contains all partial derivatives of a function at a given position. On a convex function, gradient descent could be used, and on a concave function, gradient ascent could be used. Gradient descent finds the function’s nearest minimum, whereas gradient ascending seeks the function’s nearest maximum. If the objective function could be flipped, either style of optimization could be utilized for the same issue. With this article, we have understood the gradient ascent. References Link to the above codeRead more about Gradient ascent","excerpt":"Gradient ascent maximizes the loss function of the algorithm","categories":["AI Trends"],"tags":["gradient descent","local minima"],"author_name":"Sourabh Mehta","publish_date":"2022-06-21T11:00:00","publication_year":"2022","word_count":987,"keywords":["data science","NumPy","Go","AI","R","Python","Ray","analytics","local minima","programming_languages:Python","Pandas","gradient descent"],"extracted_tech_keywords":["AI","data science","analytics","Ray","Pandas","NumPy","Python","R","Go","programming_languages:Python"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/gradient-ascent-when-to-use-it-in-machine-learning\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10062771,"title":"Ukraine is using Clearview AI’s face recognition tech for war efforts","content":"Ukraine’s military is allegedly using Clearview AI’s facial recognition technology to uncover Russian assailants, combat misinformation and identify the dead. The US-based startup is reportedly providing free access to its powerful search engine to allow Ukraine authorities to screen for people of interest at checkpoints and safeguard their borders. Though Ukraine’s Defense Ministry has not confirmed the use of technology yet, earlier, a spokesperson for Ukraine’s Ministry of Digital Transformation said the government was considering offers from US-based AI companies like Clearview to step up their resistance against Russian invasion. Many companies from the West have come forward to help Ukraine by providing cybersecurity tools and other tech support. Clearview’s technology could be used for purposes like reuniting refugees separated from their families, identifying Russian operatives and helping the government debunk false social media posts related to the war, said CEO & co-founder Hoan Ton-That. EXCLUSIVE Ukraine has started using Clearview AI’s facial recognition during war https:\/\/t.co\/2gZNsnjfio pic.twitter.com\/yNV8h9BYOU— Reuters (@Reuters) March 13, 2022 Clearview, a primary vendor to US law enforcement, is fighting lawsuits in the United States over privacy violations. Albert Fox Cahn, executive director of the Surveillance Technology Oversight Project in New York, said while identifying the deceased is probably the least dangerous way to deploy face recognition technology in war, once such systems are introduced, it will be difficult to control its use or misuse.","excerpt":"Ukraine’s Defense Ministry has not confirmed the use of technology yet.","categories":["AI News"],"tags":["clearview AI","face recognition"],"author_name":"Kartik Wali","publish_date":"2022-03-15T15:03:36","publication_year":"2022","word_count":228,"keywords":["Go","AWS","AI","cloud_platforms:AWS","digital transformation","Git","BERT","llm_models:BERT","clearview AI","face recognition","R","startup"],"extracted_tech_keywords":["AI","AWS","R","Go","Git","BERT","digital transformation","startup","llm_models:BERT","cloud_platforms:AWS"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/ukraine-is-using-clearview-ais-face-recognition-tech-for-war-efforts\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10126750,"title":"Las Vegas Sphere’s Visual Wonders, Powered by NVIDIA","content":"Sphere, a new entertainment medium in Las Vegas, is powered by 150 NVIDIA RTX A6000 GPUs, driving the 16x16K displays across its interior and the 1.2 million programmable LED pucks on its exterior, the Exosphere, which is the world’s largest LED screen. Robust network connectivity is maintained by NVIDIA BlueField DPUs, NVIDIA ConnectX-6 Dx NICs, NVIDIA DOCA Firefly Service, and NVIDIA Rivermax software, ensuring synchronized content delivery across all display panels. “Sphere is captivating audiences not only in Las Vegas, but also around the world on social media, with immersive LED content delivered at a scale and clarity that has never been done before,” said Alex Luthwaite, senior vice president of show systems technology at Sphere Entertainment. “This would not be possible without the expertise and innovation of companies such as NVIDIA that are critical to helping power our vision, working closely with our team to redefine what is possible with cutting-edge display technology.” Sphere, named one of TIME’s Best Inventions of 2023, hosts Sphere Experiences, concerts, residencies from world-renowned artists, and premier corporate events. Rock band U2 opened Sphere with a 40-show run that concluded in March. The Sphere Experience featuring Darren Aronofsky’s Postcard From Earth showcases the venue’s technologies, including high-resolution visuals, concert-grade sound, haptic seats, and atmospheric effects. Sphere Studios, located in Burbank, Calif., creates video content for the venue, transferring it digitally to Sphere in Las Vegas. Content is streamed in real time to rack-mounted workstations with NVIDIA RTX A6000 GPUs, delivering three layers of 16K resolution at 60 frames per second. NVIDIA Rivermax software accelerates media streaming, eliminating jitter and optimizing latency. NVIDIA BlueField DPUs ensure precision timing through the DOCA Firefly Service, synchronizing clocks in a network with sub-microsecond accuracy. “The integration of NVIDIA RTX GPUs, BlueField DPUs and Rivermax software creates a powerful trifecta for modern accelerated computing, supporting the unique high-resolution video streams and strict timing requirements needed at Sphere,” said Nir Nitzani, senior product director for networking software at NVIDIA. Sphere Studios also developed the Big Sky camera system, capturing uncompressed 18K images from a single camera, eliminating the need to stitch multiple camera feeds together. The studio’s custom image processing software runs on Lenovo servers powered by NVIDIA A40 GPUs, which fuel creative work such as 3D video, virtualisation, and ray tracing. The team uses apps like Unreal Engine, Unity, Touch Designer, and Notch for developing visuals for different shows. The Las Vegas Sphere is establishing itself among legendary circular performance spaces like the Roman Colosseum and Shakespeare’s Globe Theater. The venue features eye-popping LED displays covering nearly 750,000 square feet inside and outside the venue.","excerpt":"The venue features eye-popping LED displays covering nearly 750,000 square feet inside and outside the venue.","categories":["AI News"],"tags":["NVIDIA"],"author_name":"Siddharth Jindal","publish_date":"2024-07-13T11:08:26","publication_year":"2024","word_count":436,"keywords":["programming_languages:R","AI","innovation","Git","Ray","ViT","NVIDIA","R"],"extracted_tech_keywords":["AI","Ray","R","Git","ViT","innovation","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/las-vegas-spheres-visual-wonders-powered-by-nvidia\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":19572,"title":"TensorFlow Vs. Spark: How Do They Differ And Work In Tandem With Each Other","content":"Earlier this year Yahoo open sourced a new project called TensorFlowOnSpark, a pairing of Spark and TensorFlow that would make the deep learning framework more attractive to developers, especially to those who are creating models that need to run on large computing clusters. This integration of big data and machine learning actually adds support for the TensorFlow deep learning library into Spark. The researchers are going gaga over this newly available tool that has eased their work and helping them achieve faster results in analytics functionalities. For the uninitiated, this article brings the differences and\/or similarities between TensorFlow and Spark, and why does it give data scientists a reason to celebrate. Spark vs. TensorFlow = Big Data vs. Machine Learning Framework? Apache Spark or Spark as it is popularly known, is an open source, cluster computing framework that provides an interface for programming entire clusters with implicit data parallelism and fault tolerance. Built on top of Akka, Spark codebase was originally developed at the University of California and was later donated to the Apache’s Software Foundation. This cluster of computer framework lets the user make the computation faster by providing in-memory computing and easy integration because of the big Spark ecosystem. Spark cluster can be used for various tasks like machine learning, graph computation by paralleling them. By using in-memory computation it reduces an overhead of disk read and write. This fast and general engine for large scale data processing offers many features like high speed, ease of use, it can combine with SQL, streaming and complex analytics, can run everywhere such as Hadoop, Mesos and cloud. Spark, essentially a big data framework, has made it possible for a large number of companies generating huge amount of user data to process it efficiently and offer up recommendation at scale. TensorFlow, on the other hand, is a short library developed by Google that helps in improving the performance of numerical computation and neural networks and generating data flow as graphs—consisting of nodes denoting operations and edges denoting data array. Essentially a machine learning framework, it helps people create deep learning models without the need for rigorous skill sets of a machine learning specialist. A Google API enabling computation on deep learning and machine learning, TensorFlow gives a graphical representation (Tensorboard) computation flow. The API helps user to write complex neural network design and tune it according to activation values. In summary, it could be said that Apache Spark is a data processing framework, whereas TensorFlow is used for custom deep learning and neural network design. So if a user wants to apply deep learning algorithms, TensorFlow is the answer, and for data processing, it is Spark. Can Spark improve deep learning pipelines with TensorFlow: While these two have been existed separately as tools that are widely used, mingling of deep learning and big data can make it easier for TensorFlow to be deployed easily over existing clusters, just like those running on Spark. As the experts believe, Spark and machine learning go hand-in-hand. Deep learning in particular depends of large amount of compute, and this is where there is an opportunity for Spark and TensorFlow to join the forces. Earlier, companies like Yahoo have explored tools like SparkNet and TensorFrame, but the much desired success has been achieved through TensorFlowOnSpark, that enables a developer to quickly modify their programs and yield desired results. TFoS is designed in a way that runs on the existing Spark and Hadoop clusters and use Spark libraries like SparkSQL or Spark’s MLlib machine learning libraries to allow developers to create models without getting lost into much details. Benefits of combining the two has been that, being clustered machine learning framework, it runs faster and can be used on remote direct memory access (RDMA). The original TensorFlow project doesn’t support RDMA as a core feature. TensorFlowOnSpark solves the problem of deploying deep learning on big data clusters in a distributed form. This is not a completely new deep learning model but instead an upgrade to the existing frameworks that required the development of multiple programs for deploying intelligence on big data clusters. Combining both TensorFlow and Spark, it gives a space for unwanted system complexity as well as end-to-end learning latency. Conclusion: Along the years, TensorFlow has progressed exceptionally well and has been picked by the likes of IBM as the deep learning system for its custom machine learning hardware. While facing competition from the likes of MXNet, a deep learning system by Amazon, that is speculated to be better than the competition across multiple nodes, it would be interesting to see how TensorFlowOnSpark compares to it. As of now, it has been seen to be running smoothly on big clusters and is convenient to work with. TensorFlowOnSpark, which has been open sourced is available on GitHub for testing it out.","excerpt":"Earlier this year Yahoo open sourced a new project called TensorFlowOnSpark, a pairing of Spark and TensorFlow that would make the deep learning framework more attractive to developers, especially to those who are creating models that need to run on large computing clusters. This integration of big data and machine learning actually adds support for […]","categories":[],"tags":[],"author_name":"Srishti Deoras","publish_date":"2017-12-08T07:43:03","publication_year":"2017","word_count":804,"keywords":["machine learning","AI","neural network","ML","Apache Spark","Ray","deep learning","analytics","TensorFlow","R"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","neural network","analytics","Ray","TensorFlow","Apache Spark","R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/tensorflow-vs-spark-differ-work-tandem\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10041396,"title":"Guide to YAMNet : Sound Event Classifier","content":"Transfer Learning is a well-liked and popular machine learning technique in which one can train a model by reusing information learned from a previously existing model. You must have heard and read about common applications of transfer learning in the vision domain – training models to accurately classify images and do object detection or text-domain – sentiment analysis or question answering, and the list goes on … We will learn how to apply transfer learning for a new(relatively) type of data: audio, by making a sound classifier. There are many vital use cases of sound classification, such as detecting whales and other creatures using sound as a necessity to travel, protecting wildlife from poaching and encroachment, etc. With YAMNet, we can easily create a sound classifier in a few simple and easy steps! YAMNet (Yet Another Mobile Network) – Yes, that is the full form, is a pretrained acoustic detection model trained by Dan Ellis on the AudioSet dataset which contains labelled data from more than 2 million Youtube videos. It employs the MobileNet_v1 depth-wise-separable convolution architecture. This pretrained model is readily available in Tensorflow Hub, which includes TFLite(lite model for mobile) and TF.js(running on the web) versions. Let’s better understand this amazing model with a practical use case and hands-on Python Implementation. Importing the Dependencies Importing tensorflow-hub for leveraging the pre-trained model, wavfile for storing an audio file. IPython.display lets us play audio right here in the notebook. import tensorflow as tf import tensorflow_hub as hub import numpy as np import csv import matplotlib.pyplot as plt from IPython.display import Audio from scipy.io import wavfile Loading the Model Instantiating the pre-trained model to a variable using hub.model method for usage in the below cells. The labels file will also be loaded from model assets and is present at model.class_map_path(). We require this to load it on the class_names variable later on. # Load the model. model = hub.load('https:\/\/tfhub.dev\/google\/yamnet\/1') Helper_Function_1 This helper function is created to find the name of the class with the top score when mean-aggregated across frames. # Find the name of the class with the top score when mean-aggregated across frames. def class_names_from_csv(class_map_csv_text): \"\"\"Returns list of class names corresponding to score vector.\"\"\" class_names = [] with tf.io.gfile.GFile(class_map_csv_text) as csvfile: reader = csv.DictReader(csvfile) for row in reader: class_names.append(row['display_name']) return class_names class_map_path = model.class_map_path().numpy() class_names = class_names_from_csv(class_map_path) Helper_Function_2 We need to add this method to verify and convert the loaded audio to the proper sample_rate which is 16K. This is mentioned in the YAMNet paper by the authors, as it can adversely affect our model results. def ensure_sample_rate(original_sample_rate, waveform, desired_sample_rate=16000): \"\"\"Resample waveform if required.\"\"\" if original_sample_rate != desired_sample_rate: desired_length = int(round(float(len(waveform)) \/ original_sample_rate * desired_sample_rate)) waveform = scipy.signal.resample(waveform, desired_length) return desired_sample_rate, waveform Downloading and Preparing the sound file The Colab notebook will have all links required; you just have to run the notebook provided. !curl -O https:\/\/storage.googleapis.com\/audioset\/speech_whistling2.wav !curl -O https:\/\/storage.googleapis.com\/audioset\/miaow_16k.wav We can also listen to a sample audio file from the downloaded data set and check its properties by applying the following snippet. As shown below, we can play the sample audio file and look at some information about this particular audio file. # wav_file_name = 'speech_whistling2.wav' wav_file_name = 'miaow_16k.wav' sample_rate, wav_data = wavfile.read(wav_file_name, 'rb') sample_rate, wav_data = ensure_sample_rate(sample_rate, wav_data) # Show some basic information about the audio. duration = len(wav_data)\/sample_rate print(f'Sample rate: {sample_rate} Hz') print(f'Total duration: {duration:.2f}s') print(f'Size of the input: {len(wav_data)}') # Listening to the wav file. Audio(wav_data, rate=sample_rate) Running the Model We are converting the wave data into numbers to feed to the pre-trained model. The model will give us scores , embeddings, and spectrograms as output that we can later display. Helper_Function_1 gives us the output as “Animal” which means that this is the label with the maximum number of audio files in our dataset. waveform = wav_data \/ tf.int16.max # Run the model, check the output. scores, embeddings, spectrogram = model(waveform) scores_np = scores.numpy() spectrogram_np = spectrogram.numpy() infered_class = class_names[scores_np.mean(axis=0).argmax()] print(f'The main sound is: {infered_class}') Plotting the Output Plotting the three outputs we got from running the model, namely scores, embeddings and spectrograms. plt.figure(figsize=(10, 6)) # Plot the waveform. plt.subplot(3, 1, 1) plt.plot(waveform) plt.xlim([0, len(waveform)]) # Plot the log-mel spectrogram (returned by the model). plt.subplot(3, 1, 2) plt.imshow(spectrogram_np.T, aspect='auto', interpolation='nearest', origin='lower') # Plot and label the model output scores for the top-scoring classes. mean_scores = np.mean(scores, axis=0) top_n = 10 top_class_indices = np.argsort(mean_scores)[::-1][:top_n] plt.subplot(3, 1, 3) plt.imshow(scores_np[:, top_class_indices].T, aspect='auto', interpolation='nearest', cmap='gray_r') # patch_padding = (PATCH_WINDOW_SECONDS \/ 2) \/ PATCH_HOP_SECONDS # values from the model documentation patch_padding = (0.025 \/ 2) \/ 0.01 plt.xlim([-patch_padding-0.5, scores.shape[0] + patch_padding-0.5]) # Label the top_N classes. yticks = range(0, top_n, 1) plt.yticks(yticks, [class_names[top_class_indices[x]] for x in yticks]) _ = plt.ylim(-0.5 + np.array([top_n, 0])) After running the above snippets, it will be displayed as below. And here is your sound classifier using the pretrained model YAMNet. I recommend trying out this model with different datasets from open source and those present in the links present in this blog. This is a peculiar use case that will enhance your skillset in Deep Learning. The Google Colab notebook is present here for reference. References: Official Github RepositoryResearch PaperOfficial Source CodeColab Implementation","excerpt":"Transfer Learning is a well-liked and popular machine learning technique in which one can train a model by reusing information learned from a previously existing model. You must have heard and read about common applications of transfer learning in the vision domain – training models to accurately classify images and do object detection or text-domain […]","categories":["Deep Tech"],"tags":["Guide","Transfer Learning"],"author_name":"Mudit Rustagi","publish_date":"2021-06-08T10:00:00","publication_year":"2021","word_count":859,"keywords":["NumPy","machine learning","AI","sentiment analysis","RAG","Colab","Ray","Matplotlib","deep learning","Transfer Learning","TensorFlow","Guide"],"extracted_tech_keywords":["AI","machine learning","deep learning","Ray","TensorFlow","Colab","NumPy","Matplotlib","RAG","sentiment analysis"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/guide-to-yamnet-sound-event-classifier\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10164902,"title":"Why India’s Best Space Engineers Are Choosing to Stay Back","content":"The Indian space sector, once dominated by government-led initiatives, is seeing a surge in private players, opening up career trajectories that didn’t exist. Along with that, the decade-old norm of space engineers studying in India, gaining experience at ISRO, and then moving abroad for better opportunities is also fading away. Instead, more Indian engineers are choosing to stay and build in India. One of the catalysts for this shift has been India’s evolving space policy. “If you recall, in 2020, Nirmala Sitharaman announced approximately ₹20 lakh crore as part of the COVID package. A portion of that was dedicated to opening up new sectors for increased employment and investments—space was one of them,” said Yashas Karanam, co-founder and COO of Bellatrix Aerospace, in an exclusive interaction with AIM. The announcement triggered confidence, leading to a wave of private space ventures, which in turn created jobs and opportunities that could compete with global offers. Giving Engineers a Reason to Stay India boasts over 190 space startups, a significant jump from just a handful in 2015. Karanam said that companies are not only building satellite technologies but also innovating in propulsion, launch systems, and AI-driven space navigation. “In 2012, there was probably one company. In 2015, there were three. By 2017-18, it was up to 10. But after 2019-2020, the acceleration has been significant,” said Karanam. The investments tell the same story: Indian private space startups have collectively raised over $350 million in funding in the past five years, demonstrating both investor confidence and a growing market. This expansion is keeping engineers engaged in the country. “Now the opportunities are getting better here in India,” Karanam explained. “There are really exciting challenges to solve, and in terms of the trade-off, people went abroad because of better projects and financial incentives. But if you are paid well here, you could visit those countries while still building in India.” In terms of talent, Karanam believes India has the right set of people. “I feel we are really good with what it takes to build a space company or space products because the engineers we have got have been really good. A lot of times, training or onboarding onto particular technologies would be required, but everybody is a quick learner as far as we have seen,” he said. ISRO to Private Startups Karanam observes an emerging trend within this space boom, which is the migration of former ISRO scientists to the private sector. Unlike earlier, where ISRO was seen as the only credible player, today’s engineers, many of who have worked on India’s high-profile missions, are stepping out to start their own companies. “We have seen more people quitting their ISRO jobs and starting up on their own because they now know that the government is supporting this sector,” Karanam said. “India also realised that despite being one of the top five global spacefaring nations, we had less than 2% of the global commercial market share. That was because only ISRO was doing it.” This realisation has fueled a policy shift that encourages private participation, resulting in companies securing contracts directly from ISRO and IN-SPACe (India’s nodal space authorisation body). The industry, once restricted to working under ISRO’s guidance, is now developing independent solutions for both domestic and international clients. Bengaluru-based aerospace manufacturer Bellatrix Aerospace, which specialises in developing in-space propulsion systems and small satellite manufacturing, was founded in 2015 by Rohan M Ganapathy and Karanam. Since its inception, Bellatrix has raised a total of $11.3 million over four funding rounds. Some of its notable investors include Inflexor, Pavestone, StartupXseed, GrowX, BASF, and others. Bollywood actress Deepika Padukone has also backed Bellatrix. This Indian private space startup has developed two innovative satellite propulsion systems: a water-based Microwave Plasma Thruster and India’s first private Hall Effect Thruster, positioning the country as a leader in innovative propulsion technology. ‘Make in India’ for Space Hardware Despite the enthusiasm, India’s space startups face a significant hurdle: building hardware at scale. While software-based AI and deep-tech companies can scale quickly with limited capital, space hardware requires heavy upfront investment, regulatory approvals, and access to cutting-edge materials. Karanam elaborated on the challenges of building space components domestically. “If I were to build a propulsion system, I could buy a valve from a company that makes valves, buy propellant from a propellant manufacturer, and buy a catalyst from Shell or SABIC. “But if all my suppliers are outside India, and I have to import everything, i.e. paying space shipping and customs duty, it does not make sense in the long term,” he said. As a result, companies are increasingly choosing to manufacture components in-house. “We were blessed with a really capable team, and we had several retired ISRO scientists advising us. So, we took the bold decision to build even materials in-house,” Karanam added. While building space hardware is a challenge, funding for hardware development is a major roadblock. Most investors, accustomed to rapid returns from SaaS and consumer-tech startups, hesitate to back capital-intensive deep-tech ventures. “There is a funny saying: ‘The space industry is a sinkhole for money.’ The more you deploy, the more it goes into capital and R&D,” Karanam said, reflecting on investor scepticism. Unlike software, which can show quick user growth and revenues, hardware requires patience, long development cycles, and extensive testing.","excerpt":"“People went abroad because of better projects and financial incentives, but if you are paid well here, you could visit those countries while still building in India,” said Yashas Karanam, co-founder and COO of Bellatrix Aerospace.","categories":["AI Startups"],"tags":["Bellatrix Aerospace","ISRO","nirmala sitharaman","Rohan Ganapathy","spacetech","Yashas Karanam"],"author_name":"Vandana Nair","publish_date":"2025-03-02T10:00:00","publication_year":"2025","word_count":884,"keywords":["Go","ISRO","Yashas Karanam","API","startup","funding","AI","nirmala sitharaman","Rohan Ganapathy","RAG","Aim","Bellatrix Aerospace","Rust","GAN","R","spacetech"],"extracted_tech_keywords":["AI","Aim","RAG","R","Go","Rust","API","GAN","startup","funding"],"url":"https:\/\/analyticsindiamag.com\/ai-startups\/why-indias-best-space-engineers-are-choosing-to-stay-back\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10139033,"title":"Microsoft Steals the Spotlight with Autonomous AI Agents","content":"Microsoft is expanding the capabilities of its Copilot platform by introducing new autonomous agents that aim to enhance business processes across industries. On October 21, the company announced that the ability to create autonomous agents within Copilot Studio will enter public preview next month. These agents are designed to work across various business functions, including sales, finance, and supply chain, to automate tasks and streamline operations. “Copilot is the UI for AI, and with Copilot Studio, customers can easily create, manage, and connect agents to Copilot. Today we announced new autonomous agent capabilities across Copilot Studio and Dynamics 365 to help scale the impact of every individual, team, and business function,” said Microsoft chief Satya Nadella. This development comes following Salesforce chief Marc Benioff’s criticism of Microsoft Copilot, comparing it to Clippy 2.0. “When you look at how Copilot has been delivered to customers, it’s disappointing. It just doesn’t work, and it doesn’t deliver any level of accuracy,” Benioff said in a post on X. He further mentioned that major analyst firms say it’s spilling data everywhere, leaving customers to clean up the mess. To add insult to injury, customers are then told to build their own custom LLMs. “I have yet to find anyone who’s had a transformational experience with Microsoft Copilot or the pursuit of training and retraining custom LLMs. Copilot is more like Clippy 2.0,” he said. Microsoft has introduced ten new autonomous agents in Dynamics 365. These agents are built to help organisations drive business value by automating processes like lead generation, customer service, and supplier communication. For instance, a new Sales Qualification Agent assists sellers by prioritising leads and personalising customer outreach, while a Supplier Communications Agent autonomously tracks supplier performance to minimise supply chain disruptions. Already used by 60 percent of Fortune 500 companies, Microsoft 365 Copilot has demonstrated significant results, including projected savings of $50 million annually for Lumen Technologies and productivity gains for Honeywell equating to adding 187 full-time employees. Microsoft highlighted examples of companies already utilising these capabilities. Pets at Home, the U.K.’s leading pet care business, is using an agent for profit protection, which could potentially lead to significant annual savings. McKinsey & Company has developed an agent to reduce client onboarding time, and Thomson Reuters is using a legal due diligence agent that speeds up workflows and increases new business pipeline efficiency. As these agents become more integrated into enterprises, Microsoft assures customers of strong data governance and security measures. All agents built in Copilot Studio adhere to stringent protocols, including data loss prevention and robust authentication. Microsoft is also leveraging these technologies internally, and is seeing notable gains in areas including sales, customer service, and marketing. Meanwhile, at Dreamforce 2024, Salesforce launched its Agentforce Partner Network. This global ecosystem brings together top partners, including Amazon Web Services (AWS), Google Cloud, IBM, and Workday, to enhance the capabilities of its AI-driven Agentforce platform. The new network will give businesses access to a wide array of third-party agent actions and pre-built agent templates, automating complex tasks across various systems. Agentforce is a platform that enables companies to deploy AI agents to complete intricate tasks by integrating actions across Salesforce and various third-party systems. The newly launched partner network expands the platform’s scope by allowing these agents to interface seamlessly with external systems, enabling users to accomplish more in less time.","excerpt":"This new development comes in the backdrop of Salesforce chief Marc Benioff’s criticising Microsoft Copilot—calling it Clippy 2.0.","categories":["AI News"],"tags":["Microsoft"],"author_name":"Siddharth Jindal","publish_date":"2024-10-21T21:28:31","publication_year":"2024","word_count":560,"keywords":["Go","autonomous agents","AI","AWS","ML","RAG","Ray","Aim","data governance","R","Microsoft"],"extracted_tech_keywords":["AI","ML","Aim","Ray","RAG","autonomous agents","AWS","R","Go","data governance"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/microsoft-steals-the-spotlight-with-autonomous-ai-agents\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10005605,"title":"Deep Learning DevCon 2020: Association of Data Scientists Launches It’s Latest Virtual Conference","content":"The Association of Data Scientists (ADaSci), the premier global professional body of data science & machine learning professionals, announces the launch of Deep Learning DevCon 2020 (DLDC). The Deep Learning conference of the year, DLDC is a leading virtual conference exclusively for deep learning practitioners across the world. The 2-day virtual conference will be held on 29th and 30th October bringing influential people in the deep learning domain on a single platform. Over the years, deep learning has become a crucial subarea of the artificial intelligence and machine learning domain with many exciting use cases that have been explored in various industries. Deep learning models are dominating in a variety of applications and have outperformed the classical machine learning models in many ways. DLDC plans to take a deeper dive into the latest research and advancements in the field by bringing together leading professionals and researchers that are pushing the boundaries of this interesting area. The virtual conference spread across two days will have two tracks and is expecting to host 40 speakers, 500 attendees and more than 200 organisations. It will bring the experience of our in-person conferences from the comfort of your home. Besides leading experts on deep learning from around the world, it will host paper presentations, exhibitions & hackathons. There will also be a track on full-day workshop on deep learning with a certificate of attendance provided to the attendees. The 2-day hands-on workshop will take researchers through the process of building deep learning models for interesting applications from scratch and applying those in real-life scenarios. These implementations will be done with Keras using the Tensorflow backend. DLDC2020 also welcomes submissions for paper presentations in the subareas of deep learning in domains such as healthcare, sustainability, transportation, and commerce. Some of the highlights of the conference are: Technology talksPaper presentationsExhibitions & hackathonsFull-day Workshop on deep learningNetworkingMeeting companies virtually Interaction with speakers, and more Where: Virtual When: Thursday to Friday | October 29 to 30, 2020 Read more about DLDC2020 here.","excerpt":"The Association of Data Scientists (ADaSci), the premier global professional body of data science & machine learning professionals, announces the launch of Deep Learning DevCon 2020 (DLDC).  The Deep Learning conference of the year, DLDC is a leading virtual conference exclusively for deep learning practitioners across the world. The 2-day virtual conference will be held […]","categories":["Deep Tech"],"tags":["ADAsci (Association of Data Scientists)","Deep Learning"],"author_name":"Srishti Deoras","publish_date":"2020-08-26T11:10:18","publication_year":"2020","word_count":334,"keywords":["data science","machine learning","artificial intelligence","ADAsci (Association of Data Scientists)","AI","Keras","ai_frameworks:TensorFlow","deep learning","GAN","Deep Learning","TensorFlow","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","deep learning","data science","TensorFlow","Keras","R","GAN","ai_frameworks:TensorFlow"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/deep-learning-devcon-2020-association-of-data-scientists-launches-its-latest-virtual-conference\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":53345,"title":"L&#8217;Oréal Introduces Perso, An AI-Powered Personalised Skincare Treatment","content":"Recently L’Oréal unveiled its AI-powered offering Perso at the CES 2020. L’Oréal is a French personal care company with its headquarter at Clichy, is one of the most recognised brands in the world. The popular brand has always been seen as an innovation hub in this evolving sphere, which is privy to all kinds of trends and fads. In a sector which is usually accused of stereotyping and making women (and men) feel inadequate in front of unrealistic beauty standards, L’Oréal has been making use of technology to fill in this wide chasm. For example in 2014, L’Oréal acquired NXT Cosmetics for an undisclosed amount, to penetrate the North American markets. In 2016, they acquired IT Cosmetics for $1.2 billion. It also has six other research centres, dedicated to cosmetic research and developing technologies, established in six different countries with one in India as well. Now, the company is making strides to make its products, right from product development, testing and transportation, as eco-friendly as possible. The company declared all their product ‘vegan’ in origin and also claims to have significantly reduced its carbon footprint – by 77% since 2005 and aims to become carbon neutral by 2050. L’Oréal’s latest revolutionary product — Perso was unveiled at the CES (Consumer Electronics Show) 2020. Perso is an AI-controlled cosmetic product dispensing device, which is actually a collection of three devices — one for skincare, foundation and lipstick each. The device is operable via Bluetooth and is NFC (Near-field Communication) compliant, which basically means that it works best in tandem with a smartphone. Users must install the Perso app, take a picture and upload it in the app. The app analyses the condition of the user’s skin with regards to the moisture content, wrinkles, pore visibility, etc. The AI takes it a step further, through Breezometer geo-location data, which locates the smartphone and gathers relevant information concerning that specific location like the local weather, humidity, pollution, UV index and even pollen concentration in the air. To further bolster the effectiveness of the cosmetics, the user can input their personal preferences or elaborate on what they are looking for. The Tech Behind L’Oréal’s Perso L’Oréal-owned and AI-enabled – Modiface, is the technology that computes all the data collected through the app, taking every input into consideration, it dispenses a high-quality solution (cream, lipstick, etc.) with the perfect combination of ingredients to suit the user’s individual needs. The solution is released at the top of the device for cleanliness and the quantity is fit for a single-use. The device, through the AI, automatically adjusts itself in accordance with the day and night requirements. The app can create a personalised daily makeup regiment, delivering a unique experience to its users based on trial and feedback. Perso grows old with you. As the user ages, the needs of the skin changes too. AI can predict these changes or the user can keep their unique database updated. Perso also changes the combination of its ingredients to suit the personal requisites of the user, keeping pace with the passage of time. The hardware of Perso is 6.5 inches tall, which is fitted with L’Oréal’s patented three-cartridge system containing the essential ingredients. Using advanced robotics, the ingredients are extracted (from the cartridges), mixed, condensed and dispensed for use. Modiface comes equipped with precise-shape vendor tool, allowing users to modify their foundation requirements according to the sculpture of the face or as they see fit. The AI can also make such predictions and change the composition of the product. The Perso app (at the CES 2020) comes equipped with photos of thousands of individuals for the users to use as a reference. L’Oréal’s Perso and the app will be officially launched in 2021, with several features such as creating a unique shade of lipstick or matching the same with dresses and much more. People will also have the option of syncing the app with their social media accounts and create their blend of makeup from whatever inspires them. The cartridges do run out and require replacing. Each refill is designed to last for a month. The refilling process is made simple and mess-free too. The users can track the content levels of the ingredients in the cartridges and order replacements, place their order and track the shipment through the app. Outlook on Perso In times when the AI is mostly used for surveillance or keeping track of digital records, cyber terrorism, cybersecurity etc., L’Oréal’s take on AI implementation in the field of cosmetics, nonetheless, is rather refreshing and inspiring. L’Oréal has applied AI capabilities in a way not many have envisioned which, rightly so, places them at the helm of the cosmetics industry. The probability of personalised products procuring revenue is 1.7 times that of general products. That is the reason behind cosmetic companies focusing on personalisation with the trend gaining momentum in recent years. Honourable mention to Kylie Jenner and her creative marketing strategy.","excerpt":"Recently L’Oréal unveiled its AI-powered offering Perso at the CES 2020. L’Oréal is a French personal care company with its headquarter at Clichy, is one of the most recognised brands in the world. The popular brand has always been seen as an innovation hub in this evolving sphere, which is privy to all kinds of […]","categories":["AI News"],"tags":["AI patents"],"author_name":"Yeshey Rabzyor Yolmo","publish_date":"2020-01-08T12:29:23","publication_year":"2020","word_count":825,"keywords":["AI patents","AI","programming_languages:R","innovation","Git","Aim","ai_applications:robotics","GAN","R"],"extracted_tech_keywords":["AI","Aim","R","Git","GAN","innovation","programming_languages:R","ai_applications:robotics"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/perso-ai-loreal-ces-2020-personalised-makeup\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":25660,"title":"Top Machine Learning Research Groups To Follow In India","content":"IIT Delhi campus (Image:Twitter) As the buzz around machine learning and artificial intelligence gains traction, we find more mid and senior-level candidates veering towards this field. Besides massive open online courses (MOOCs), the other way to gain knowledge and understand where the industry is heading is by following top research groups in India. These groups usually are on top of the industry trends and follow works related to computer vision, unsupervised learning, natural language processing with a focus on Indian languages, information retrieval and data mining, among others. In India, research is usually centred around top level institutes like the IITs and IIITs, which have dedicated groups comprising students and faculty members. These groups are focused on exploring the theoretical foundation of algorithms and how they can be applied to industry problems. In this article, we list down top research groups to follow in India: Indian Institute of Science’s Machine Learning Special Interest Group: Touted as one of the best research groups in India, especially the one with a beautiful campus, IISc’s MLSIG features several talented students and faculty members engaged in cutting-edge research on a variety of aspects of ML and related fields. These works range from theoretical foundations to new algorithms as well as other exciting applications. IISc MLSIG has a great roster of events that covers topics such as deep learning with GPUs, data mining (models, algorithms and applications), text analysis, knowledge representation and reasoning with DNN. MLSIG is also doing cutting-edge research that is published in top journals and conferences. Some of the research topics are ML in text mining, ML in computer vision, graphical models, clustering, support vector machines and kernel-based learning methods. To join the mailing list, click here. IIT Kanpur’s Special Interest Group On Machine Learning Of Computer Science: IIT Kanpur’s SIGML features areas of interest like ML, data mining and related fields. Besides organising reading groups, it also features a lecture series on topics such as extreme multi-label classification and deep learning theory, Bayesian model and online and stochastic optimisation and learning. The group also holds regular seminars on ML and data mining. SIGML reading groups focus on topics like deep learning, kernels and Bayesian graphical model. Non-IIT-K learners can follow the events and sessions by filling this joining request. ISI Kolkata’s Vision and Learning Research Group (VLRG): ISI’s VLRG does a lot of research in pattern recognition, image processing and computer vision. Set up in 2011, the research group has also published several papers in IEEE, ISMM, ICIP and ISVC, especially in areas like image and video processing, image reconstruction and image and pattern recognition. The group is led by Prof Bhabatosh Chanda and features outstanding faculty members like Prof Nikhil Pal, Prof Dipti Prasad Mukherjee and includes Post-Doc Fellows and Research Personnel. IIT Delhi’s Data Analytics and Intelligence Research (DAIR): IIT Delhi’s DAIR is led by stellar faculty members Mausam, Sayan Ranu and Parag Singla. A highly collaborative group, DAIR runs several research collaborations with researchers from University of Washington, Rochester University, Max Plank Institute, Bar Ilan University, University of Texas Dallas, and IIT Bombay. This IIT Delhi group is also very active in publishing research papers at international conferences — for example, Sayan Ranu, assistant professor in the CSE department has been awarded the ACM SIGMOD 2018 Most Reproducible Paper Award for his paper Debunking The Myths Of Influence Maximisation: An In-Depth Benchmarking Study. The group’s May 2018 paper — Which Route Should A Cab Driver Follow To Find A Customer will be presented in KDD 2018. Last Word Research is the bulwark for achieving progress in critical industry applications. India’s research communities that have forged a great industry connect are addressing business-centric problems. For example, Last year, New York-based CA Technologies teamed up with IIIT Hyderabad researchers to collaborate on applied research projects which will include next-generation interfaces such as NLP, AI and ML as which we had reported in February. In fact, IIT Delhi is also aiming for an equal split between government and industry research.","excerpt":"As the buzz around machine learning and artificial intelligence gains traction, we find more mid and senior-level candidates veering towards this field. Besides massive open online courses (MOOCs), the other way to gain knowledge and understand where the industry is heading is by following top research groups in India. These groups usually are on top […]","categories":[],"tags":["Data Mining","IISc","IIT Delhi","ML research","NLP"],"author_name":"Richa Bhatia","publish_date":"2018-06-23T11:29:49","publication_year":"2018","word_count":669,"keywords":["artificial intelligence","machine learning","AI","Data Mining","ML","computer vision","IISc","NLP","RAG","Aim","deep learning","analytics","IIT Delhi","ML research"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","NLP","computer vision","analytics","Aim","RAG"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/top-machine-learning-research-groups-to-follow-in-india\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":38592,"title":"8 Strategies For Data Scientist To Build And Manage A Productive Team","content":"The data science industry has gained momentum over the past few years. Companies from different industries and domain are incorporating data science in their business process, creating a significantly high number of job opportunities for young professionals more than ever before. No doubt, data science has the superpowers to take a business to great heights, however, those powers won’t be much of a use if your company doesn’t have a strong data science team with all the required skills. But the major challenge here is to build that ultimate data science team. Even if you have the top talent, but your strategies for team building and retaining isn’t effective, you might end up breeding a work environment that would not deliver value to the team as well as the company. In order to help data scientist with their team building strategies and increase the productivity level, we list down 8 techniques you can adopt while building the ultimate data science team. Establish Yourself As A Leader To build a team of top data scientist, there has to be a leader who would not only lead but also mentor. And to be leader, you have to have a lot of knowledge about the domain — you just cannot leave a stone unturned to gain as much knowledge as you can. So, when you are starting with building your data science team, make sure you establish yourself as a leader with all the qualities — in terms of knowledge as well as leadership. Planning Is The Key This is one of the most important things to keep in mind while you are scouting and looking to hire data scientists for your team. Make a robust plan of how you want to carry out the hiring process. One of the best ways to plan is to figure out the roles you want to fill in the team, and that can only be figured out when you have a blueprint of the project you are working on and the professionals you already have. You should never expect someone to master everything. Rather, we should look for the specific skill they possess. And data science is no exception, you just can’t hire data scientists; you have to hire data scientists with a strong hand on the skill you are looking for. So, pen down each and every single detail as this would help you in hiring the specific talent. Build A Culture Of Open Communication Nobody wants to work in an environment, where they have to struggle to get heard. And data science being one of the core and vast domains, professionals look for a work environment where they can openly share ideas and share their views — both in terms of work and the organisation. To create a culture like that, team supervisors or leaders need to communicate actively with all the members on a daily basis. Ask them what they are working on, how they are working, if they are facing any difficulty, and also, hold team meetings and brainstorming sessions to keep everyone motivated. This might sound very generic; however, it is one of the best team building strategies. Also, make sure that your data science team is updated with every single change happening in the firm. It would make them feel they are an important part of the organisation. Set Goals For The Team Many organisations don’t follow the concept of defining goals — they allow the team to work at their pace. However, that might be effective when you are core data science organisation. To make the team more productive, you have to set goals (such as weekly goals). And setting goals doesn’t mean putting pressure (this is huge misconception people have). When you set goals, you stay focused, you don’t get distracted by the things that don’t require attention. Furthermore, if the team is completely new, don’t set a high goal, rather, start with something that can be achieved with no struggle. Appreciate Good Work Paychecks are important, but appreciation matters too. Every employee who hustles day in and out feels more motivated by recognition and appreciation than by monetary rewards. Monetary benefits are always there in the data science domain, however, its tough to get recognized in this domain where competition is top-notch and is increasing rapidly. So, make sure to provide instant feedback — appreciate if it’s good, and motivate if it’s bad, but never demotivate. If you are having a culture where appreciation sits at the top, then you definitely going to have a team that would work hard and stay. Be A Good Mentor A team is not only defined by their work and the result they deliver, but also by the kind of leader they have. Being a supervisor or leader of a data science team, you are expected to have an immense amount of knowledge, and we have mentioned that in the very first point. So, when you are leading a team, make sure you are also using your skills and knowledge at actively. Help the team solve the complex problems — sit with them, work with them, accomplish with them. Also, mentor them as much as you can. It would not only make your team productive but will make you a better leader. Be Supportive And Trustworthy There is no surprise that many employees across the world don’t like their managers and don’t trust them. However, that shouldn’t be the case in a data science team. The data science domain itself is vast, complex, and confusing, so the professionals always seek support from their leaders. So, when you are leading a team of enthusiastic data science professionals, make sure you support them every way possible, making it sure that they know you value their contributions — try your best to get them the best projects, make sure they are pressurized by deadlines, etc. Upskill The Team Upskilling is one of the best ways to get rid of the hiring process. There are many organisations that choose upskilling rather than employing a new professional, and it is also beneficial in many ways — both for the employer and the employee. According to one of our studies, there is a significant shift in mindset when it comes to upskilling. Earlier the focus was on the upskilling of a few isolated people in a core data science team. However, today the organisations have started to realize that every employee must become data literate and “data smart”. Furthermore, MOOCs are gaining significant traction. According to another study of ours, 49% of the respondents extensively rely on MOOCs when it comes to upgrading their skills in new tech. So, when you feel your team is complete and is capable enough to take up other challenges, you can always go with upskilling. Hold training sessions where you teach your team other data science skills. It would not only help you get other projects done but also providing opportunities for the team to learn different skills that they didn’t know.","excerpt":"The data science industry has gained momentum over the past few years. Companies from different industries and domain are incorporating data science in their business process, creating a significantly high number of job opportunities for young professionals more than ever before. No doubt, data science has the superpowers to take a business to great heights, […]","categories":["AI Trends"],"tags":["Data Science","Data science team","Hiring","upskilling"],"author_name":"Harshajit Sarmah","publish_date":"2019-05-03T09:56:27","publication_year":"2019","word_count":1171,"keywords":["data science","Go","API","Data science team","upskilling","AI","programming_languages:R","Hiring","programming_languages:Go","ViT","Rust","GAN","Data Science","R"],"extracted_tech_keywords":["AI","data science","R","Go","Rust","API","GAN","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/8-strategies-for-data-scientist-to-build-and-manage-a-productive-team\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":17707,"title":"Visualization and statistical inference of SpaceX missions from 2006 to 2017 with GG plots in R","content":"Space X, Space Exploration Technologies Corporation is American company aerospace company started in 2002 by entrepreneur Elon Musk and was originally based in El Segundo and later relocated to Hawthorne, California. It was founded on a vision and mission to reduce the cost of space transportation and enabling the colonization of Mars, it has since then tested and developed the Dragon launch vehicle and the Falcon spacecraft family, which both currently deliver payloads into Earth’s orbit. History In 2001, Elon Musk came with the idea and motivation regarding a project to land a miniature greenhouse to grow certain plants on Mars, this would be the farthest that life’s ever been traveled since Neil Armstrong and Buzz Aldrin set foot on the Moon. Musk tried to buy rockets from Russia but returned empty handed as the cost was too high, he realised that he could build a company that could build affordable rockets as the raw materials for building rockets were around 3% of the sales price at that time. SpaceX started with the smallest orbital rocket to test their design as a more costlier and riskier vehicle could have failed and bankrupted the company. Achievements Space X designed and funded the first privately owned liquid propellant rocket to reach orbit in the year 2008 [Falcon 1], the first private company to successfully launch, orbit, and recover a spacecraft in the year 2010 [Dragon]. Also, the first private company to send a spacecraft to the International space station in the year 2012 from the Dragon family of spacecraft family. NASA awarded SpaceX a development contract in 2011 to develop a spaceship that would be used to transport astronauts to the International Space Station. Visualization and Statistics of SpaceX missions Big data in the aerospace and aeronautics industry Data analytics helps the defence and aerospace industry by helping them optimizing the flow of resources and business processes while maintaining and regulating precise details which help in business decisions as information technology enabled the digitation of business procedures with the help of automation of some manual tasks to improve overall efficiency. Analytics in its intrinsic form has been helping to collect data on spaceships, rockets, and aircraft for years ranging from binary data such as altitude and speed, it even deals with very minute and slow progressions details such as crack growth progression, temperature variations. The data is not a problem, aside from its ability to ease the launch of rockets, the astronomical data collected over several launches of SpaceX has been the driving force behind the decisions they make. Moulded by their state of their art intelligence systems, the information produced has changed the direction of traditional business such as research and sales. It has helped SpaceX to return from catastrophic failures such as the explosion in September 2016. And ultimately allowed them to take decisions based on data which was proved when they launched their reusable rocket Falcon 9 and could land it on a 91 x 52-meter landing pad, 350km away, from 80km up. GG plots in r The ggplot2 package is created by Hadley Wickham and provides powerful graphics processing interface for creating elegant and sharp complex plots, it’s been popular in the recent years, it allows to plot graphs that represent both multivariate and univariate data in a straightforward manner, plus grouping can be represented by colour, symbol, and size which results in steady and neat representation. Future of big data in aerospace engineering and cosmology Data science should not be complicated as rocket science, yes it has its challenges but the industry trends will only make the tools and algorithms easier to divest, often organizations spend up huge amounts of financial and economic resources in endless cycles to model. In the field of cosmology, measurements must be taken of the structures and evolution of the universe as it’s a big place, plus to make theoretical predictions of what the universe looks like, big simulations are run. The science required for these simulations is less of an issue than the technology required to run the simulations as big data is quite tough to move around that’s why processing frameworks like Hadoop exist. Getting information from a metaphorical slob of data is the first critical step that is why organizations analyse data at its source because then there is no waiting period to download data in a separate analytical environment, in an overall gist; data will continue to grow and new patterns will emerge, rapid changes in the stream of big data ecosystem will be driven by advanced predicting insights which will eventually be the common norm in industry and academia alike and will foster the growth of machine learning with the time to come.","excerpt":"Space X, Space Exploration Technologies Corporation is American company aerospace company started in 2002 by entrepreneur Elon Musk and was originally based in El Segundo and later relocated to Hawthorne, California. It was founded on a vision and mission to reduce the cost of space transportation and enabling the colonization of Mars, it has since […]","categories":[],"tags":[],"author_name":"Kshitiz Gupta","publish_date":"2017-09-14T06:58:08","publication_year":"2017","word_count":786,"keywords":["big data","data science","Go","API","machine learning","AI","Git","RAG","analytics","R"],"extracted_tech_keywords":["AI","machine learning","data science","analytics","RAG","R","Go","Git","API","big data"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/visualization-statistical-inference-spacex-missions-2006-2017-gg-plots-r\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10005686,"title":"7 Online Fun Tools Created On GPT-3","content":"With GPT-3 slowly revealing its potential, it has created a massive buzz amid the ML community. While developers are trying their hands on some of the exciting applications of GPT-3, many are expressing their astonishment with the kind of possibilities it can bring for humanity. Apart from writing codes, writing stories and developing websites, many developers have played around with GPT-3 to bring out some fun applications. In this article, we are going to share seven of those fun online tools that are created on GPT-3: Also Read: Top Free Resources To Learn GPT-3 1| Create Mails With OthersideAI About: OthersideAI is an online tool that generates full-length emails according to the preference, only by typing the essential pointers that need to be covered. With GPT-3, the tool helps users create well-written, concise, convincing emails 4x faster than before. Along with that, the tool also provides intelligent summarisation, insights, and a whole suite of other tools to manage the inbox. Check out the tool here. Also Read: GPT-3 Is Great. But Not Without Shortcomings 2| Write Job Description With Dover About: Dover is an online tool built using GPT-3 to help HR professionals to write job descriptions. All one has to do is write a short description of what they want, and the tool will generate a more extended variant. Not only it helps in creating a creative job description but also speeds up the process of writing job descriptions for mass hiring. Check out the tool here. Also Read: A Fake Blog Created By College Kid Using GPT-3 3| Ask Questions With Philosopher AI About: With all the important business problems, it is also necessary from time to time to ask some of the fundamental existential questions on life, and thus the tool Philosopher AI. This tool, along with answering some of the essential life questions, generates text on different topics and subjects. It mimics the opinions of renowned philosophers and then produces answers for the relevant query. Check out the tool here. Also Read: Will The Much-Hyped GPT-3 Impact The Coders? 4| Build Apps With Debuild This is mind blowing.With GPT-3, I built a layout generator where you just describe any layout you want, and it generates the JSX code for you.W H A T pic.twitter.com\/w8JkrZO4lk— Sharif Shameem (@sharifshameem) July 13, 2020 About: Debuild is an online tool, built on GPT-3, used for creating web applications at a lightning-fast speed. All one has to do is describe the application requirement in plain English on the tool, and then with one click, the app can be created. “I would have to re-write my two initial samples in HTML\/CSS. Then all of GPT-3’s outputs would be in plain HTML\/CSS,” added Sharif Shameem, the founder of debuild.co. Check out the tool here. Also Read: How OpenAI’s GPT-3 Can Be Alarming For The Society 5| Learn Anything From Anyone Ever wanted to learn about rockets from Elon Musk?How to write better from Shakespeare?Philosophy from Aristotle?GPT-3 made it possible.https:\/\/t.co\/SScjQvUk68 pic.twitter.com\/13Yi9p8NnY— Mckay Wrigley (@mckaywrigley) July 17, 2020 About: LearnFromAnyone is an online tool, built on GPT-3, used for learning anything from anywhere. From learning to write better than Shakespeare to learning philosophy from Aristotle, this tool makes it possible for anybody to understand complex subjects from anybody they wish for. Check out the tool here. Also Read: 5 Jobs That GPT-3 Might Challenge 6| Tweet Generator Mind-blowing AI-generated tweetsWith @OpenAI's GPT-3 model (thanks to @gdb), I built an app that generates its own tweet given any word.Also, these AI tweets are indistinguishable from human tweets.Try for yourself (replace 'word' below with yours):https:\/\/t.co\/LZN7RkTbE1— Sushant Kumar (@sushantkumar_23) July 15, 2020 About: This tool is built on GPT-3 to generate tweets on any given word. Developed by Sushant Kumar, the Cofounder of a tech-enabled real estate asset management company — Azuro, this tool claims to create indistinguishable tweets for users. Check out the tool here. 7| Generate Taglines With Taglines.AI About: Last but not least, GPT-3 has been used to create this online tool, where users can generate perfect taglines for their business. The tool claims to have made task such as creating slogans as easy as with just one click. The tool creates short, catchy phrases that can easily be used by businesses for communicating their ideas and visions.Check out the tool here.","excerpt":"With GPT-3 slowly revealing its potential, it has created a massive buzz amid the ML community. While developers are trying their hands on some of the exciting applications of GPT-3, many are expressing their astonishment with the kind of possibilities it can bring for humanity.  Apart from writing codes, writing stories and developing websites, many […]","categories":["AI Trends"],"tags":["GPT-3"],"author_name":"Sejuti Das","publish_date":"2020-08-28T15:00:00","publication_year":"2020","word_count":717,"keywords":["GPT-3","Go","TPU","OpenAI","AI","ML","GPT","Aim","GAN","R","llm_models:GPT"],"extracted_tech_keywords":["AI","ML","OpenAI","Aim","TPU","R","Go","GPT","GAN","llm_models:GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/7-online-fun-tools-created-on-gpt-3\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10170837,"title":"Infosys to Overhaul Workplace for 77,000 E.ON Workers Using AI","content":"Indian IT giant Infosys has announced a strategic collaboration with European energy firm E.ON on Tuesday to roll out an AI-powered digital workplace for over 77,000 employees across Europe. The partnership will use Infosys Topaz, a suite of AI-based tools, to modernise E.ON’s workplace systems, enhance operational efficiency, and support its broader digital transformation goals. The companies plan to build a data-driven, user-focused workplace that will help E.ON become a fully digital energy provider. E.ON is one of Europe’s largest energy providers, operating in energy networks, infrastructure solutions, and retail services. “Digitalisation and digital technology are key for reliable, affordable and sustainable energy systems. Our strategic partnership with Infosys is essential for our digital transformation and operation – together, we are paving the way for a smarter, more efficient energy future,” Victoria Ossadnik, chief operating officer for digital and innovation at E.ON, stated. The collaboration will also introduce a human-centric support model, giving E.ON employees more choice and control in how they work, with the aim of accelerating business innovation. “By leveraging Infosys Topaz, we will enable an AI-powered digital workplace transformation that will bolster productivity and employee experience,”Ashiss Kumar Dash, executive vice president and global head at Infosys, added. Recently, Infosys also announced a strategic collaboration with LogicMonitor, a leading AI-powered SaaS hybrid observability platform. This partnership combines Infosys AIOps Insights from its Cobalt cloud offering with LogicMonitor’s Edwin AI to significantly improve IT operations’ observability, performance, and reliability across complex systems.","excerpt":"The companies plan to build a data-driven, user-focused workplace that will help E.ON become a fully digital energy provider.","categories":["AI News"],"tags":["Infosys"],"author_name":"Shalini Mondal","publish_date":"2025-05-27T16:49:13","publication_year":"2025","word_count":243,"keywords":["Go","Infosys","AI","data-driven","digital transformation","innovation","Git","RAG","Aim","ViT","R"],"extracted_tech_keywords":["AI","Aim","RAG","R","Go","Git","ViT","digital transformation","data-driven","innovation"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/infosys-to-overhaul-workplace-for-77000-e-on-workers-using-ai\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10102997,"title":"10 GPTs for Your Daily Use","content":"At OpenAI’s first developer conference, DevDay 2023, OpenAI introduced GPTs that allow anyone to easily build their own GPT without the need for coding. As of now, there are over 5,000 GPTs available. Here, we have curated a list of the top GPTs based on their popularity on social media, ensuring they add significant value to your life. Invoice GPT Invoice GPT, created by Arnav Bathla, allows users to extract information in JSON or visualize data from any invoice. By simply uploading an invoice image, users can easily obtain structured data in JSON format or visually organized data. https:\/\/twitter.com\/arnavbathla20\/status\/1723891591814009053?s=12 This technology streamlines various use cases, such as automating accounting processes, enhancing financial record-keeping, and expediting expense management. Whether in a business setting or for personal finances, InvoiceGPT’s ability to accurately extract details like amounts, dates, and vendor information empowers users to efficiently manage their financial data, reducing manual effort and increasing overall productivity. ChatGPT LogoMaker The ChatGPT Logo Maker, crafted by Sahil Rahul, simplifies the logo creation process. Users can upload a sketch, and the tool generates a logo that can be further customised. Experiment with various styles, tweak details, and prompt the system to refine the output, allowing for a more precise representation of your initial sketch. Just upload a sketch, and out comes a logo, you can tweak and try different styles and prompt your way to a more accurate representation of your sketch. pic.twitter.com\/UuZfZ9KxZu— Linus ●ᴗ● Ekenstam (@LinusEkenstam) November 13, 2023 DesignerGPT Created by Pietro Schirano, DesignerGPT lets you create beautiful websites directly in ChatGPT, with native support for dark mode. Websites are hosted remotely via Replit. https:\/\/twitter.com\/skirano\/status\/1723026266608033888 ConvertAnything Created by Pietro Schirano, ConvertAnything  efficiently converts images, audio, videos, PDFs, and more. It includes batch uploads, ZIP support and download links. ConvertAnything simplifies your experience by asking only one question: What do you want your files to be? Your interactions are streamlined, centered solely on file conversion, eliminating unnecessary dialogue and delays. https:\/\/twitter.com\/skirano\/status\/1723769180657213788 Screenplay GPT Created by Kiri, Screenplay GPT can be used to create stories or scenes for images. You can either upload an image and ask the tool to describe a scene related to it, or you can first come up with a scene and use the tool to generate images that match your description. It’s a handy tool for storytellers, content creators, or anyone looking for creative inspiration for visuals. Introducing Screenplay GPT!Upload your images and ask for your scene :). Or, create scenes first and then ask for it to generate images for you.Check it out in action!Link: https:\/\/t.co\/oq1eF0oUYdThis is quite powerful for storytelling! pic.twitter.com\/c5q3crqxJp— Kiri (@Kyrannio) November 11, 2023 NomadGPT NomadGPT empowers you to embrace the digital nomad lifestyle by helping you discover the best places worldwide for remote living and work. Leveraging Nomad List’s real-time data, it answers your queries about ideal locations based on your specific budget, preferred weather, and a myriad of other relevant data points. https:\/\/twitter.com\/levelsio\/status\/1722411513741291826 The F1 Assistant GPT F1 Assistant GPT, developed by David, is your go-to tool for all things Formula 1. Whether you’re a fan or an enthusiast, this assistant provides lap times, pit stop details, driver and team standings, and other race-related statistics. Imagine easily keeping track of your favorite teams and drivers, staying updated on thrilling race moments, and having quick access to the latest stats—all in one place. https:\/\/twitter.com\/apex_bite\/status\/1724107286409285991 YellowPage Navigator YellowPage Navigator developed by Eric Ma, is a go-to tool for finding business information and contacts easily. With this tool, you can quickly locate reliable plumbers in your zip code, complete with their contact details. No more sifting through directories or online listings—YellowPage Navigator streamlines the process, making it simple and quick to find the services you need, right when you need them. https:\/\/twitter.com\/ericmanyc\/status\/1723883758292488349 Interview Wizard GPT: Your AI-Powered Personal Interview Coach Interview Wizard GPT, created by Ashutosh Srivastava, is a great tool to prepare for interviews. Simply input your resume and the job description, and receive personalized guidance for your interview preparation. From mock questions to deeper insights, it covers all bases. Whether it’s behavioral interview queries, salary negotiation tactics, or more, it can help you with everything. https:\/\/twitter.com\/ai_for_success\/status\/1722970875719467039 DALL•E 3 with Parameters GPT Alvaro Cintas  created this GPT as an adaptation of Midjourney’s great parameters. Users  can now choose the aspect ratio, stylization, seeds, weirdness, & more. DALL•E 3 with Parameters GPT is here!I created this GPT as an adaptation of Midjourney’s great parameters.You can now choose the aspect ratio, stylization, seeds, weirdness, & more.Link in the comments. pic.twitter.com\/uVoFKsKFQF— Alvaro Cintas (@dr_cintas) November 12, 2023","excerpt":"There are over 5,000 GPTs available as of now","categories":["AI Trends"],"tags":[],"author_name":"Siddharth Jindal","publish_date":"2023-11-14T16:16:06","publication_year":"2023","word_count":757,"keywords":["Go","ChatGPT","TPU","OpenAI","AI","ML","Git","RAG","GPT","R"],"extracted_tech_keywords":["AI","ML","ChatGPT","OpenAI","RAG","TPU","R","Go","Git","GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/10-gpts-for-your-daily-use\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10065902,"title":"How to evaluate recommender systems fairly with RexMex?","content":"In our daily lives, we all either use or give recommendations. In machine learning, the same system is developed that filters out undesired information and provides various outcomes based on different parameters that change from user to user. These recommender systems may be biased or unjust at times while recommending; the bias might be of any type, such as a model bias or data bias. There are several algorithms for determining the fairness of the recommendation system. In this article, we will use RexMex to assess the system’s fairness. Following are the topics to be covered. Table of contents About Recommendation SystemTaxonomy of FairnessEvaluating Recommendation System with RexMex Let’s start by talking about the Recommendation system and the various types of filtering methods used by the system. About Recommendation System Machine-learning algorithms are used to determine which items should be recommended to a specific user or customer through a recommendation engine. Based on the principle that patterns can be found in consumer behaviour data, it can collect implicit or explicit information. The types of filters used for these systems could be divided into three groups according to what information needs to be filtered. Collaborative filtering Collaborative filtering is concerned with gathering and analysing data on user behaviour, activities, and preferences to predict what a person will like based on their similarities to other users. It employs a matrix-style formula to plot and calculate these similarities. One advantage of collaborative filtering is that it does not require content analysis or comprehension (products, films, books). It simply chooses which items to recommend based on what it knows about the user. Content-based filtering Content-based filtering relies on the assumption that if you like one item, you’ll enjoy this one as well. Algorithms employ cosine and Euclidean distances to calculate the similarity of objects based on a profile of the customer’s interests and a description of the item (genre, product category, colour, word length). The disadvantage of content-based filtering is that it may only propose items or content that are comparable to what the user is already purchasing or using. It is unable to make recommendations for other sorts of items or content. It couldn’t, for example, propose anything other than homeware if the user had only brought homeware. Hybrid model A hybrid recommendation engine takes into account both meta (collaborative) and transactional (content-based) data. As a result, it outperforms both. Natural language processing tags may be created for each product or item (movie, music) in a hybrid recommendation engine, and vector equations can be utilised to calculate product similarity. Then, based on the users’ behaviours, activities, and interests, a collaborative filtering matrix may be utilised to propose products to them. One example of this model, Netflix is the epitome of a hybrid recommendation engine. It considers both the user’s interests (collaborative) and the descriptions or characteristics of the movie or show (content-based). Image source Are you looking for a complete repository of Python libraries used in data science, check out here. Taxonomy of Fairness Individual and Group Fairness A distance-based approach is one technique to formulate individual fairness. Assume that the distance between two things is indicated by ‘d,’ and the distance between the outputs of an algorithm is denoted by ‘D.’ When two entities are similar, the output of the method should be smaller. The distance D between probability distributions assigned by the classifier should be no larger than the actual distance d between the items in statistics. To grasp the concept of group fairness, consider two groups: the protected group and the privileged group. The recommender can examine the chances of items from each group appearing incomparably favourable ranking positions, as well as the chances of their being suggested. When the probabilities are equal, individuals of each group have the same probability of obtaining a favourable outcome. This will only occur if members of the non-privileged group are qualified. As a result, if it is not qualified, it will not be suggested. User and Item Fairness Item fairness focuses on items that are evaluated or recommended. In this case, rank or recommend similar articles or groups of articles as well. B. It will be displayed in the same position in the ranking. This is the main type of fairness described so far. For example, if a political party considers an article to be a protected attribute, you can request that the value of that attribute not affect the ranking of articles in search results or news feeds. The user-side fairness is concerned with the users who get or consume data items in a ranking, such as a search result or a suggestion. In general, we want comparable individuals, or groups of people, to receive similar rankings or recommendations. For example, if a user’s gender is a protected characteristic and the user is getting job suggestions, we may request that the user’s gender not impact the job recommendations that the user gets. Static and Dynamic Fairness Static fairness does not take into account changes in the recommendation environment, such as changes in item utility or attributes; therefore, dynamic fairness has recently been studied, which takes into account dynamic factors in the environment and learns a strategy that accommodates such dynamics. Evaluating fairness with RexMex RexMex is designed with the assumption that end-users might want to use the evaluation metrics and utility functions without using the metric sets and scorecards. Because of this, the evaluation metrics and utility functions (e.g. binarisation and normalisation) can be used independently from the RexMex library. Let’s start by installing RexMex. ! pip install rexmex Read dataset and import libraries. from rexmex import ClassificationMetricSet, DatasetReader, ScoreCard reader = DatasetReader() scores = reader.read_dataset() Generating evaluation report. metric_set = ClassificationMetricSet() score_card = ScoreCard(metric_set) report = score_card.generate_report(scores, grouping=[\"source_group\"]) Comparing evaluation scores among specificity, sensitivity and precision for different source groups (0, 1, 2, 3, 4). As observed the precision score for group 4 is the highest among the five source groups. Final words Algorithms are playing increasingly important decision-making roles in a wide range of social, corporate, and individual applications. As algorithmic decisions become more prevalent in major areas of societal impact, it becomes critical to ensure that they provide some level of fairness and trust, especially when individuals and groups representing minorities or protected classes in terms of gender, race, and so on, are subjected to the negative consequences of algorithmic decisions.  With a hands-on implementation of this concept in this article, we could understand how to check the fairness of prediction. References Link to the above codeDocumentation for RexMex","excerpt":"Recommender systems help individuals in excluding the overwhelming choices of our daily lives. However, while such systems learn patterns from historical data, they can capture the bias mediated by the underlying data about imbalances and inequality.","categories":["AI Trends"],"tags":["bias","FAIRNESS","Machine Learning","RECOMMENDER SYSTEM"],"author_name":"Sourabh Mehta","publish_date":"2022-04-29T13:00:00","publication_year":"2022","word_count":1090,"keywords":["data science","Go","machine learning","TPU","AI","R","Machine Learning","Python","FAIRNESS","ViT","RECOMMENDER SYSTEM","Rust","programming_languages:Python","bias"],"extracted_tech_keywords":["AI","machine learning","data science","TPU","Python","R","Go","Rust","ViT","programming_languages:Python"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/how-to-evaluate-recommender-systems-fairly-with-rexmex\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":61076,"title":"NULLcoin’s White Paper Made Public By Hackernoon","content":"Several cryptocurrencies may exist today, but they all face the same problem – the extremely volatile prices and the cover-up of decentralization. In this setting, NULLcoin is regarded as the world’s first depreciation-resistant currency. It keeps the price on an appreciation curve by using an architecture-level design. NULLcoin acts as an alternative to store-of-value cryptocurrencies. In a recently published white paper by Hackernoon, the team says that NULLcoin will live beyond the purview of the regulatory bodies. This is by ensuring that the transferred amount in other cryptocurrencies from one’s account is small enough to qualify as a “gift” transaction from entities to the team. This depreciation-resistant cryptocurrency consists of ERC-20 tokens on the Ethereum Blockchain. What Does NULLcoin Do? An asset that maintains its value without depreciating is a store of value which makes NULLcoin the absolute store of value. When one purchases a NULLcoin, they are not transferred to the wallet of the buyer. Instead, they are directly sent to a NULL wallet-the private key of that wallet was noted on a piece of paper and burnt which is known as token-burn. Thus, the tokens are removed from the supply permanently and increases token’s prices. In short, the prices increase as goods become scarce. This does not make NULLcoin any different because this a common practice done by Stellar (XLM), MakerDAO (MKR), and Binance Coin (BNB). What makes NULLcoin different is that the purchased tokens are burnt 100% right at the point of the sale. This activity results in the remaining NULLcoin become more scarce, which further leads to an appreciation of the price. To buy a NULLcoin, one needs to spend BTC\/ETH\/XRP\/BCH, etc. to buy the NULLcoin. Technical Architecture NULLcoin is designed to be a completely no-code solution that will only live on the Blockchain. The technical architecture of NULLcoin only comprises no-code solutions. NULLcoin creation service – tokenmint.io – was selected by Full Stack Developer, Austin Pocus since it was the first thing which came up on Google SERP for the term “create ERC20 token.” However, other no-code solutions can also power NULLcoin, such as Zapier and Metamask NULLcoin uses Ethereum Blockchain since it has the widest developer support and several tools. Also, NULLcoin uses Ethereum Blockchain’s primary feature of acting as a spreadsheet, which is termed as a smart contract. NULLcoin’s smart contract waits till all the token wallet achieves its fundraising goal of $0.00. If the goal is met, it can release 10 million NULLcoins directly into a private key-less Ethereum wallet, burning all NULLcoins. Future As per David Smooke-Founder & CEO of Hackernoon, the team is already working towards a project to make sure that the ledger is 100% void of unwanted NULLcoin. To avoid any kind of coercion that may cause a coin to become any other type apart from NULL, the team coerced it back to NULL using a 20 step patent-pending process. The team is further looking to extend this cryptocurrency, which aims to coerce other coin types to NULL. That is not all as the team’s NULL coercion API will be available soon as well. As per the team, NULLcoin can be easily converted into a physical object representing the digital asset. However, tracking it could be challenging. In contrast, one can put as many as they want in their physical wallets or pocket. The team says that NULLcoin does not change price over time, so there is no point in listing it in cryptocurrency exchanges. The NULLcoin is the first cryptocurrency to achieve true stability because it is valueless by design.","excerpt":"Several cryptocurrencies may exist today, but they all face the same problem – the extremely volatile prices and the cover-up of decentralization. In this setting, NULLcoin is regarded as the world’s first depreciation-resistant currency. It keeps the price on an appreciation curve by using an architecture-level design. NULLcoin acts as an alternative to store-of-value cryptocurrencies.  […]","categories":[],"tags":["Cryptocurrency","full stack project ideas"],"author_name":"Rohit Chatterjee","publish_date":"2020-04-07T08:00:25","publication_year":"2020","word_count":593,"keywords":["Go","API","programming_languages:R","AI","Cryptocurrency","programming_languages:Go","Git","Aim","ViT","R","full stack project ideas"],"extracted_tech_keywords":["AI","Aim","R","Go","Git","API","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/nullcoins-white-paper-made-public-by-hackernoon\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10013266,"title":"Hand-on Implementation of CycleGAN, Image-to-Image Translation using PyTorch","content":"A CycleGAN is designed for image-to-image translation, and it learns from unpaired training data. It gives us a way to learn the mapping between one image domain and another using an unsupervised approach. Jun-Yan Zhu original paper on the CycleGan can be found here who is Assistant Professor in the School of Computer Science of  Carnegie Mellon University. Examples of image data in both sets: Translating summer landscapes to winter landscapes (or the reverse). Unpaired Training Data These images do not come with the labels, i.e. the generator creates the training data X from  the Y datasets. We do not have to extract all the corresponding features from the individual images. In the GitHub code that introduced CycleGANs, the authors were able to translate the horses to zebras, even though there are no images of zebra exactly in the same position of horses. Thus CycleGANs enables learning from X to another domain Y  mapping without having to find perfectly matched, training pairs! Define CycleGAN: A CycleGAN is made of two types of networks: discriminators and generators. In this example, the discriminators are responsible for classifying images as real or fake (for both X and Y kinds of images). The generators are responsible for generating convincing, fake images for both kinds of images. A simple example of the CycleGAN. This image presents the data flow through CycleGAN to pull it all together: Load DataSet: You’ll need to download the data as a zip file here. First, install the PyTorch and import all the libraries for this project. Pytorch DataLoader: Split train and test data using the different path directory of  Datasets and the DataLoader function from PyTorch. Store the new dataset using the ImageFolder. Parameter’s Specifications: Image_type: Directory where X and Y image are stored Image_dir: Main directory for Train and Test image Image_size: resized image dimension Batch_size: Number of images in one batch of Data Visualization of Training and Testing Data Use numpy , torchvision.utils , matplotlib to visualize the image from the dataset. Discriminators: ????_X and ????_???? , in this CycleGAN, are convolutional neural networks that see an image and attempt to classify it as real or fake. Discriminator architecture consists of a series of 5 convolutional layers in which the first four conv_layer have BatchNorm and ReLu activation functions and last act as a classification layer. Discriminators Network: Discriminators class to create the model in pytorch. The ReLu activation function is used to pass input images through convolutional layers. We have provided a helper function which creates a convolutional layer + an optional batch norm layer. Now the helper function can easily create a Discriminators class. Generators The generators G_XtoY and G_YtoX.It is responsible for turning an image into a smaller feature representation, and an encoder, a transpose_conv  and decoder net that is responsible for turning that representation into a transformed image. It goes through three convolutional layers using BatchNorm and ReLu activation functions and reaches a series of residual blocks. The residual blocks are made of convolutional and batch normalization layers. Generators Network: Generators class using the same function as Discriminator. ???????????????????? and ???????????????????? have the same architecture, so we only need to define one class, and later instantiate two generators. It will contain three-part encoder, transformer and Decoder.Use the convolutional neural network and sequential function to define the generator. And a feed-forward function generator using ReLu. In the last layer, use TanH function. Residual Block: It will help you connect the encoder and decoder. It consists of two convolutional layers. The layer must have the same input size as output. Putting it all Together: Create two discriminatorsG_XtoY and G_YtoX  then two generators D_X and D_Y for  full network. To train the model either on  GPU for faster processing or use CPU. Refer code snippet: Discriminator and Generator Losses Computing the discriminator and the generator losses are key to getting a CycleGAN to train. Image from original paper by Jun-Yan Zhu et. al. Discriminator Losses The discriminator loss is mean squared errors between the output of the discriminator. Generator Losses Calculating the generator losses will look somewhat similar to calculating the discriminator loss; there will still be steps in which you generate fake images that look like they belong to the set of ???? images but are based on real images in set ????, and vice versa. Adversarial Loss: The first adversarial loss is calculated on the generator G and the discriminator D. The second adversarial loss is calculated on the generator G(x), and the discriminator D(y). Cycle Consistency Loss In addition to the adversarial losses, A cycle consistent mapping function is a function that can translate an image x from domain A to another image y in domain B, and generate back the original image. A forward cycle consistent mapping function appears as follows: X -> G(X) -> F(G(X)) ≈ x A backward cycle consistent mapping function looks as follows: Y -> G(Y) -> F(G(Y)) ≈ Y This network sees a 128x128x3 image, compresses it into a feature representation as it goes through three convolutional layers and reaches a series of residual blocks To calculate the total loss,  if G is our generator from A to B and F is our generator from B to A, then â = F(G(a)) ≈ a. All Loss function: Real_mse_loss: The loss in the real image. Fake_mse_loss: The loss in fake image. Cycle_Consistency_loss: Total loss Training CycleGan: We  will train our model in two-part 1.) Discriminator: Calculate the discriminator loss for both. Generate Fake Image. Calculate the total loss using on both discriminators. 2.) Generator: Generate  fake images of X which is real image of Y. Generate the new Y images based on the fake X images. Calculate the cycle consistency loss on real Y images and New Y images. Visualisation of Image: We can see the sample of images after 100 epoch: 1.)Image transformation from X to Y in 100 epoch 2.)Image transformation from Y to X in 100 epoch Now see the results after 5000 epoch: Image transformation from X to Y in 5000 epoch Image transformation from Y to X in 5000 epoch Conclusion: We have learned how to use a CycleGAN in the image to image translation. We started with an introduction to CyleGANs and explored the architectures of networks involved in CycleGANs. We also explored the different loss functions required to train CycleGANs. This was followed by an implementation of CycleGAN in the PyTorch framework. We trained the CycleGAN on the available  dataset and visualized the generated images, the losses, and the graphs for different networks.","excerpt":"CycleGAN is designed for image-to-image translation, and it learns from unpaired training data. It gives us a way to learn the mapping between one image domain and another using an unsupervised approach.","categories":["Deep Tech"],"tags":["Computer Vision","GAN","how to calculate mean square error","Unsupervised Learning"],"author_name":"Amit Singh","publish_date":"2020-12-06T13:00:40","publication_year":"2020","word_count":1091,"keywords":["Go","NumPy","TPU","how to calculate mean square error","AI","neural network","PyTorch","Git","Computer Vision","GitHub","GAN","Matplotlib","R","Unsupervised Learning"],"extracted_tech_keywords":["AI","neural network","PyTorch","NumPy","Matplotlib","TPU","R","Go","Git","GitHub"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/hand-on-implementation-of-cyclegan-image-to-image-translation-using-pytorch\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10169232,"title":"IBM Unveils Hybrid AI Tools to Help Enterprises Scale with Speed","content":"IBM is slated to announce a new suite of hybrid AI technologies at its annual Think conference, aimed at helping enterprises deploy generative AI agents across complex environments. The tools will allow companies to build these agents in five minutes, integrate them across over 80 applications, and scale operations on new platforms capable of 450 billion daily inference operations. The announcement responds to growing demand as businesses face a projected surge in application development, which is over one billion apps expected by 2028. According to a recent study by IBM, leaders are doubling AI investments but struggling with disconnected systems and limited return on investment. “The era of AI experimentation is over,” IBM chairman and CEO Arvind Krishna said. “IBM is equipping enterprises with hybrid technologies that cut through complexity and accelerate production-ready AI implementations.” IBM’s watsonx Orchestrate platform now offers a comprehensive suite of AI agent capabilities, including pre-built domain agents for HR, sales, and procurement, alongside tools to build custom agents without code. The platform also integrates with major enterprise systems such as Microsoft, Salesforce, Oracle, SAP, and AWS, and provides agent orchestration, monitoring, and optimisation features. A newly introduced Agent Catalog in watsonx Orchestrate includes over 150 agents and tools developed by IBM and partners like MasterCard and Salesforce, designed to support use cases ranging from prospect discovery to HR automation. To address integration challenges, IBM also launched webMethods Hybrid Integration, which is designed to automate workflows across cloud and on-premise systems. For data management, IBM evolved its watsonx.data offering to activate unstructured data such as contracts and spreadsheets. Enhancements include new orchestration tools and AI-powered insight engines. Testing indicates this leads to AI agents that are up to 40% more accurate. IBM also introduced IBM LinuxONE Emperor 5, designed to support AI at scale. The platform includes built-in AI accelerators and quantum-safe encryption technologies, and promises a 44% reduction in total cost of ownership over five years compared to equivalent x86 systems. These releases are supported by expanded partnerships with AMD, Intel, NVIDIA, and others, reinforcing IBM’s strategy to help enterprises operationalise AI across hybrid cloud environments.","excerpt":"The tools will allow companies to build AI agents in five minutes and integrate them across over 80 applications.","categories":["AI News"],"tags":["enterprises","Hybrid cloud solutions","IBM"],"author_name":"Sanjana Gupta","publish_date":"2025-05-06T18:05:08","publication_year":"2025","word_count":351,"keywords":["agent orchestration","AWS","AI","cloud_platforms:AWS","programming_languages:R","automation","Hybrid cloud solutions","Aim","AI agents","generative AI","IBM","enterprises","R"],"extracted_tech_keywords":["AI","generative AI","agent orchestration","Aim","AWS","R","automation","AI agents","cloud_platforms:AWS","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/ibm-unveils-hybrid-ai-tools-to-help-enterprises-scale-with-speed\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10169018,"title":"Redis Goes Open Source Again With The Launch of Redis 8","content":"Redis officially announced its return to the open-source world on Thursday with the general availability release of Redis 8, now available under the OSI-approved AGPLv3 license, in addition to the existing licensing options. The move marks a significant shift after a controversial detour under the Server Side Public License (SSPL) earlier, a decision the company admits strained its relationship with the open-source community. “This achieved our goal—AWS and Google now maintain their own fork—but the change hurt our relationship with the Redis community,” stated the company in a blog post. The licensing change coincides with the launch of Redis 8, which the company claims to be the most advanced and performant version of the software to date. Redis 8 unifies the capabilities of Redis Community Edition and Redis Stack into a single open-source distribution—now simply called Redis Open Source. This integration brings JSON, time series, probabilistic data structures, and advanced querying into the core platform. The release also includes a new data type, vector sets, developed by the Redis creator Salvatore Sanfilippo, who rejoined the company last year. Vector sets are designed for high-dimensional vector search, making Redis more competitive for AI and semantic search applications. Redis 8 also introduces performance enhancements, including up to 87% lower latency, twice the throughput, and up to 16 times more query processing power compared to earlier versions. The new Redis 8 release aims to simplify adoption, speed up innovation, and provide developers with a consistent experience. Also, under the AGPLv3 license, it offers a path forward for organisations previously wary of SSPL licensing. The latest version is now generally available and can be installed via Docker Hub, Snap, Homebrew, or traditional Linux package managers.","excerpt":"“This achieved our goal—AWS and Google now maintain their own fork—but the change hurt our relationship with the Redis community.”","categories":["AI News"],"tags":["Redis"],"author_name":"Ankush Das","publish_date":"2025-05-02T11:44:41","publication_year":"2025","word_count":281,"keywords":["Go","semantic search","AWS","AI","innovation","docker","Aim","GAN","R","Redis"],"extracted_tech_keywords":["AI","Aim","semantic search","AWS","docker","Redis","R","Go","GAN","innovation"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/redis-goes-open-source-again-with-the-launch-of-redis-8\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10085979,"title":"Shutterstock Takes The Lead In AI Image Generation with LG AI Research","content":"Stock image platform Shutterstock has now created its own generative AI service trained on its database of high-quality stock images. The new AI generative tool launched by Shutterstock uses LG AI Research’s EXAONE Atelier—a product emerging from the companies’ partnership since November last year. This algorithm has now been integrated into the platform and can provide high-quality images from simple text prompts. The AI agent is built on a multi-model architecture for creative applications called EXAONE. The platform has also stated that the agent does not require long and complex prompts to provide good results. In contrast, learning how to prompt other generative models like Stable Diffusion and Midjourney is an art in itself, as seen by the rise of prompt engineering. In addition to this, the platform also has many easy-to-use filters to denote the required style of the output. Using the relatively simple prompt “Fleet of battleships on a wide ocean”, we were able to obtain the following results from the model. The tool is built on a massive “proprietary dataset”, which seems to be made up of the 424 million images on Shutterstock’s platform. Contributors to the website can choose to opt out and not allow their images to be included in the database, but those who do not opt out have a chance to be paid for their work. Paul Hennessy, Chief Executive Officer at Shutterstock, stated, “Our tools are built on an ethical approach and on a library of assets that represents the diverse world we live in, and we ensure that the artists whose works contributed to the development of these models are recognized and rewarded.” As part of their ethical AI initiative, the stock image platform has also promised that contributing artists will be paid for their part in training the model. The Shutterstock Contributor Fund was set up in January last year to ensure that creators whose work was used in Shutterstock’s datasets would be adequately compensated. This approach to responsible AI was lauded by many, especially when Stable Diffusion and Midjourney created controversy for using the LAION-5B dataset which provided no such compensation for artists. However, the model still has visual anomalies similar to those seen in other AI algorithms, particularly in the case of human features like hands and teeth, albeit to a lesser degree than others. Here’s an example shown below. In addition to this limitation, free members of the platform can only generate images up to 5 times per day. After this limit, they are prompted to sign up for one of Shutterstock’s premium plans. However, one must also consider that the platform is actively supporting the artists that have contributed their work for the algorithm’s datasets. Shutterstock have also partnered with AI think tanks like Meta AI, Open AI and LG AI Research in the past. Sam Altman, the CEO of OpenAI, has stated in the past that Shutterstock’s datasets were ‘critical’ to the training of DALL-E. Yann LeCun, VP and Chief AI Scientist at Meta, has also emphasised the importance of Shutterstock’s datasets for training their algorithms. Their datasets have also been instrumental in training LG’s EXAONE. As part of this partnership, Shutterstock has also used EXAONE for AI-based image tagging, as the algorithm could provide detailed descriptions for the stock images in its library. This move is representative of the growing dispute regarding the nature of AI-generated artwork, especially when it comes to companies like Midjourney and Stability AI. Recently, a class-action lawsuit has been filed against these companies, seeking compensation for the misuse of artists’ works to train generative AI algorithms. While other platforms, like Getty Images, have taken a hardline stance against generative AI, Shutterstock’s responsible approach to this divisive issue might set a precedent for the future of generative AI.","excerpt":"The new AI generative tool launched by Shutterstock uses LG AI Research’s EXAONE Atelier – a product emerging from the companies’ partnership since November last year.","categories":["AI News"],"tags":[],"author_name":"Anirudh VK","publish_date":"2023-01-27T12:41:15","publication_year":"2023","word_count":628,"keywords":["Go","DALL-E","Meta AI","TPU","OpenAI","AI","AWS","prompt engineering","generative AI","R"],"extracted_tech_keywords":["AI","generative AI","OpenAI","Meta AI","prompt engineering","AWS","TPU","R","Go","DALL-E"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/shutterstock-takes-the-lead-in-ai-image-generation-with-lg-ai-research\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10073752,"title":"Delhivery Plans to Hire 75,000 Staff to Meet Festive Rush","content":"Indian logistics and supply company Delhivery has plans to hire about 75,000 employees over the next one-and-a-half months. Out of the planned hires, 10,000 will be off-roll employees across Delhivery’s gateways, warehouses, and last-mile delivery. The company, in a statement, said about 50,000 hires would be last-mile agents through its Last-Mile Agent (LMA) programme. From self-employed individuals to retired professionals, any job seeker is eligible to sign up for the programme. The requirements are basic and may include possession of a two-wheeler, a driver’s license and a smartphone. Commenting on the announcement, Ajith Pai, COO, Delhivery, said, “Alongside building delivery capacity, we have expanded our infrastructural footprint by a million sq-ft in the first quarter of FY23.” With a plan to roll out a revenue payout of Rs 150 crore across all programmes, the company’s fully-automated mega gateway at Tauru became operational in April this year. The company aims to meet the expected higher volumes in terms of both parcels and express part-truck load business ahead of the festive season. It also plans to strengthen its D@S (Drop at Store) and Constellation programmes in order to extend its capacity. This will be done by adding 15,000-plus last-mile riders with onboarding truckers and brokers through its Business Partners (BP) programme. Currently working with over 29,000 customers, Delhivery has fulfilled over 1.6 billion shipments since inception.","excerpt":"The company said about 50,000 hires would be last-mile agents through its Last-Mile Agent programme","categories":["AI News"],"tags":["delhivery","Expansion"],"author_name":"Bhuvana Kamath","publish_date":"2022-08-27T17:58:38","publication_year":"2022","word_count":224,"keywords":["programming_languages:R","AI","Expansion","Aim","delhivery","R"],"extracted_tech_keywords":["AI","Aim","R","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/delhivery-plans-to-hire-75000-staff-to-meet-festive-rush\/","complexity_score":3,"technical_depth":4,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10139979,"title":"OpenAI Finally Brings SearchGPT to ChatGPT, Challenges Google Search, Perplexity AI","content":"The wait is finally over, OpenAI has launched the search engine. ChatGPT now offers improved web search capabilities for timely, accurate answers, blending natural language interaction with up-to-date data in sports, news, stock quotes, and more. Available to ChatGPT Plus and Team users, and rolling out to others soon, this feature enables deeper engagement with source-linked information, enhancing users’ access to trustworthy content. Shortly after a user on X spotted the search feature in ChatGPT, OpenAI released a statement confirming the same. “The search model is a fine-tuned version of GPT-4o, post-trained using novel synthetic data generation techniques, including distilling outputs from OpenAI o1-preview” mentioned OpenAI in the official release document. search is my favorite feature we have launched in chatgpt since the original the launch!it has probably doubled my usage over the past few weeks. https:\/\/t.co\/alD11AwUUb— Sam Altman (@sama) October 31, 2024 OpenAI has also released a feature that now lets users search for content from their previous conversations. It was also seen that ChatGPT is working on integrations with Slack and Google Workspace for enterprise customers, enabling them to add external information and context to ChatGPT. It will be interesting to see how ChatGPT Search compares against the likes of Perplexity, Bing and Google’s AI overviews. However, OpenAI had an embarrassing moment just after it unveiled the search engine a few months ago, with users spotting an error in its demo. Google has always had an edge over other search engines, thanks to its rich integration of local business listings and mapping services. OpenAI needs to crack this code if it were to beat Google in the long run. That said, users will certainly experience a clean, ad-free search experience on ChatGPT, thanks to Sam Altman’s stance against including ads anytime soon. While we’re not sure if it’s related, OpenAI has decided to roll out ChatGPT Search on the same day Google announced an AI learning tool called Learn About. While Microsoft’s Bing leveraged GPT 4’s capabilities as a search engine, it may very well be made redundant once ChatGPT search is rolled out for all users. Moreover, Microsoft has done little with Bing to compete against the existing competition with Perplexity. Search engines that leverage generative AI, especially Perplexity, have mostly had a turbulent relationship with publishers and the media industry. OpenAI, on the other hand, has struck partnerships with some of the most notable names in the industry, like Le Monde, Vox, News Corp, TIME, and Financial Times. OpenAI will certainly leverage its access to information from top publishers in SearchGPT. Interestingly, News Corp was one of the many publishers who slapped a lawsuit against Perplexity. “We are committed to a thriving ecosystem of publishers and creators. We hope to help users discover publisher sites and experiences, while bringing more choice to search. For decades, search has been a foundational way for publishers and creators to reach users. Now, we’re using AI to enhance this experience by highlighting high quality content in a conversational interface with multiple opportunities for users to engage” said OpenAI, in a blogpost published a few months ago. “We are convinced that AI search will be, in a near future and for the next generations, a primary way to access information, and partnering with OpenAI positions Le Monde at the forefront of this shift.”, said Louis Dreyfus, CEO of Le Monde, emphasising the strategic alignment with OpenAI to innovate while upholding journalism’s core values. Moreover, we’re curious to see how ChatGPT Search will position itself against Perplexity in the near future. Perplexity has been on a feature release spree, and is going all in on making the tool a one stop solution for knowledge seekers. Their internal search tool, knowledge cards, Spaces and a financial analysis tool offer a tremendous amount of value to the user. Perplexity is also set to raise $500 million, that will more than double its valuation to $8 billion. Aravind Srinivas, CEO of Perplexity was quick to respond after OpenAI’s announcement. He mentioned that Perplexity isn’t just about providing information for users’ queries but also directing them to links and websites to help them navigate the web efficiently. The AI search engine space isn’t just about Google and Perplexity – OpenAI will also be competing against the upcoming Meta’s search engine. [Updated: November 2, 2024 | 15:00] The headline has been revised to highlight the competitive landscape in the search market as OpenAI enters the field.","excerpt":"With this, OpenAI will also be competing with Meta’s search engine","categories":["AI News"],"tags":["AI (Artificial Intelligence)","ChatGPT","OpenAI","SearchGPT"],"author_name":"Supreeth Koundinya","publish_date":"2024-10-31T23:29:26","publication_year":"2024","word_count":739,"keywords":["Go","ChatGPT","TPU","OpenAI","AI","AWS","GPT-4o","RAG","SearchGPT","generative AI","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","generative AI","GPT-4o","ChatGPT","OpenAI","RAG","AWS","TPU","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/openai-finally-brings-searchgpt-to-chatgpt-challenges-google-search-perplexity-ai\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10072583,"title":"Yet Another Language Model from Meta: Atlas","content":"In the latest, tech giant Meta has come out with a new language model named Atlas. It is a retrieval-augmented language model with strong few-shot performance on question answering and fact-checking tasks, Meta adds. In the paper titled, ‘Few-shot Learning with Retrieval Augmented Language Models’, the researchers say that they performed evaluations on a variety of tasks such as MMLU, KILT and NaturalQuestions. This model reaches a 42% accuracy on Natural Questions by using only 64 examples and outperforms PaLM ( a 540B parameters model ) by 3 per cent though it has over 50 times lesser parameters (11B). (1\/6) Today we’re introducing Atlas, a new retrieval-augmented lang. model with strong few-shot performance on question answering and fact checking tasks. w\/ only 11B parameters, Atlas outperforms a 540B parameter model by 3% w\/ 64 training exs., reaching 42% on NaturalQuestions. pic.twitter.com\/ptuoR3JvI4— AI at Meta (@AIatMeta) August 8, 2022 Retrieval augmented model In the paper, the researchers discuss the need to bring out this model. They add that LLMs have previously shown capabilities of few-shot results but for question answering and fact checking where knowledge is key, “massive parameter counts to store knowledge seem to be needed”. This is where retrieval augmented models come in as they are capable of knowledge intensive tasks without needing too many parameters. The researchers add that they wanted to see whether these models work in few-shot settings. “We investigate whether few-shot learning requires models to store a large amount of information in their parameters, and if memorisation can be decoupled from generalization,” the researchers state. As per the researchers, Atlas retrieves relevant documents by using a general-purpose dense retriever using a dual-encoder architecture based on the Contriever. After that, the documents are processed by a sequence-to-sequence model using the Fusion-in-Decoder architecture. Image: Few-shot Learning with Retrieval Augmented Language Models The researchers study the impact of different techniques to train Atlas on its few-shot performance on tasks such as fact checking and question answering. “We find that jointly pre-training the components is crucial for few-shot performance,” the paper adds. The model performs well in resource rich as well as few shot environments.  It demonstrates SOTA results on few-shot NaturalQuestions (+2.8 per cent), TriviaQA (+3.3%), FEVER (+5.1 per cent). Atlas is very strong in traditional full training set settings and sets new state of the art on NaturalQuestions by 8%, and TriviaQA by 9% and on 5 KILT tasks, Meta informs. Image: Few-shot Learning with Retrieval Augmented Language Models Architecture The research team follows the text-to-text framework. The tasks follow this path: The system receives a text query as inputIt generates a text output For classification tasks, this query comes in the form of a textual input and the model generates the “lexicalized class label”. Image: Few-shot Learning with Retrieval Augmented Language Models The model is based on two sub-models, the paper informs. The retriever – Here the retriever based on the Contriever. It is an information retrieval technique based on continuous dense embeddings.Language model – The team uses T5 sequence-to-sequence architecture They use the Fusion-in-Decoder modification of sequence-to-sequence models and processes each document independently in the encoder. For any task like question answering to generating articles, the model follows a similar approach. It starts by retrieving the top-k relevant documents from a large corpus of text with the retriever. Then, these documents are fed to the language model, along with the query, which generates the output. Both the retriever and the language model are based on pre-trained transformer networks as per the paper. “Atlas outperforms much larger non-augmented models on few-shot question answering (NaturalQuestions and TriviaQA) and fact checking (FEVER), and is competitive with various very large models on a wide array of real-world exams,” Meta adds. Meta tells us about other benefits of Atlas too. Retrieved passages can be inspected for better interpretability and the corpus that Atlas retrieves from can be edited, or even completely swapped out. This ensures that Atlas can be kept up-to-date without needing to be retrained.","excerpt":"This model reaches a 42% accuracy on Natural Questions by using only 64 examples and outperforms PaLM","categories":["Global Tech"],"tags":[],"author_name":"Sreejani Bhattacharyya","publish_date":"2022-08-10T18:00:00","publication_year":"2022","word_count":666,"keywords":["TPU","programming_languages:R","AI","llm_models:PaLM","ML","llm_models:T5","Ray","few-shot learning","R","T5"],"extracted_tech_keywords":["AI","ML","Ray","few-shot learning","TPU","R","T5","llm_models:T5","llm_models:PaLM","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/yet-another-language-model-from-meta-atlas\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":21361,"title":"Artificial Intelligence’s Top Moonshot Projects To Watch Out For In 2018","content":"Moonshot projects in artificial intelligence are known for taking a disastrous turn — remember how IBM Watson was shelved by MD Anderson Cancer Centre in mid-2017 and was purportedly put on hold after costs crossed a whopping $62 million. Besides casting doubt on the use of cognitive technology in cancer treatment, it also turned out to be a huge PR disaster for IBM that took a severe beating over hyped claims about revolutionizing cancer. Prior to the IBM Watson debacle, there was Google Glass project debuted in the 2012 I\/O conference which inspired a lot of excitement about the future of augmented reality but failed to take off and catch the interest of consumers. Google & Facebook – a look at Moonshot projects gone from boom to bust Google sells off world’s leading robotics companies Boston Dynamics: Next comes news about Google abandoning some of its most audacious moonshot projects, that would have slowed its march towards an AI-first era. Reportedly, Google pulled back from some of its most ambitious project – Boston Dynamics was sold off to SoftBank last year for an undisclosed amount. Google acquired the robotics company in 2013 and the MIT spinoff is known for innovations such as Handle, Big Dog, Atlas and WildCat – then why the sell-off. The austere move was part of its plan to slim down on the experimental projects but more importantly, Google didn’t know what to do with the remarkable robotics innovations, experts hinted. FB’s Head of Building 8 Exited last year: Not to be left behind, Menlo Park-headquartered Facebook too is taking a stab at a few moonshot projects. Of late, there has been news about FB’s secretive lab Building 8, headed by Regina Dugan (former DARPA Director), who led the team on AI-focused hardware innovations. Last year, news surfaced about Facebook dabbling in out-of-the-world projects such as finding a way to type from brain waves and hearing through skin. While there is absolutely no knowing whether the projects are inching close to reality – Dugan left the ship after an 18-month tenure. Even though Dugan gave a glitzy presentation about future technologies Facebook was working on at its developer conference, she didn’t announce any near-term projects. AIM Takes A Look At Some Of The Most Ambitious Moonshot Projects Of 2018 & Whether They Will Take Off 1) MIT launches Intelligence Quest: In January this year, MIT announced MIT Intelligence Quest – an institute wide initiative to advance the science and engineering of both human and machine intelligence. The first entity is billed “The Core” and the key output here will be machine-learning algorithms that seek to advance our understanding of human intelligence with insights from computer science. MIT has set itself a lofty goal, Dean Anantha Chandrakasan, dean of the MIT School of Engineering noted to find solutions that can prevent deaths occurring from cancer through deep learning for early detection and even personalized treatment. The second entity – “the Bridge” will be dedicated to the application of MIT discoveries in natural and artificial intelligence to all disciplines, and it will also host state-of-the-art tools from industry and research labs worldwide. “Today we set out to answer two big questions, says President Reif. “How does human intelligence work, in engineering terms? And how can we use that deep grasp of human intelligence to build wiser and more useful machines, to the benefit of society?” Realistic: MIT had also earlier forged a tie-up with IBM for research purpose and given MIT’s track record of success, MIT Intelligence Quest will take off. 2) Alibaba spends $15 billion on research labs to compete with Amazon: Close on the heels of these projects, Chinese e-commerce giant Alibaba announced its moonshot project – the Alibaba DAMO Academy– Discovery, Adventure, Momentum and Outlook. The project is part of the Chinese giant’s way of strengthening its R&D and stave off competition from the likes of Amazon and Tencent on home turf. As part of its R&D expansion, Alibaba would set up seven research labs across the globe — China, the U.S., Russia, Israel and Singapore and recruit 100 scientists to work on next-gen technologies AI, quantum computing and IoT. The research would also help the e-commerce giant tide over technical issues faced across its business lines. Apart from setting up research labs, the Chinese giant is also collaborating with the academia and funding collaborations with University of California and hiring academics from Princeton and Harvard to sit on an advisory board. Realistic: Alibaba’s ambitions are well-known and DAMO is not the first R&D investment. But it is still early to see how Alibaba’s efforts will pan out. Alibaba would first be applying these solutions to their day-to-day issues across business lines. 3) Neurotechnology & Brain Computer Interface: Elon Musk is no stranger to making outlandish claims and his medical research company Neuralink established sometime in 2016 is on a quest to marry biological and machine intelligence. Besides making cyborgs out of humans, neurotechnology can also answer questions about how Alzheimer’s should be treated, improve memory and even reverse ageing. The eccentric billionaire wants to use neural lace technology and implant tiny electrodes in the brain to facilitate direct computing capabilities. However, researchers assert that BCI isn’t a new concept and Silicon Valley is crawling with projects related to cognitive enhancement. Case in point – Facebook’s secret lab Building 8 has also hired a slew of neuroscientists for its hardware division. Then there is Bryan Johnson founded startup Kernel that is building advanced neural interfaces to treat disease and dysfunction, illuminate the mechanisms of intelligence, and extend cognition. Realistic: Neural lace projects are definitely dubbed as moonshots or pie-in-the sky by BCI experts as some of these projects sound unrealistic and wouldn’t be available anytime soon. 4) Entrepreneur Naveen Jain founded Viome looks for cancer cure: Entrepreneur Naveen Jain founded AI startup Viome has developed a state-of-the-art technology that offers visibility into a person’s unique ecosystem with an aim to provide a chronic-disease free life. The Washington-based startup has developed a kit-based system, based on a technology developed at the Los Alamos National Lab in New Mexico. The startup’s state-of-the-art system uses AI to analyze the digestive tract and its effects on the body’s immune and metabolic systems. In the future, Jain hopes to find a cure for a diseases like Parkinson’s, neurological disorders and cancer through diet changes. Realistic: Finding a cancer cure is every start-up’s mantra and it is early days for this startup too. However, the kit-based model has been applied to a small set of afflictions successfully.","excerpt":"Moonshot projects in artificial intelligence are known for taking a disastrous turn — remember how IBM Watson was shelved by MD Anderson Cancer Centre in mid-2017 and was purportedly put on hold after costs crossed a whopping $62 million. Besides casting doubt on the use of cognitive technology in cancer treatment, it also turned out […]","categories":["AI Trends"],"tags":["Alibaba","Alphabet"],"author_name":"Richa Bhatia","publish_date":"2018-02-06T05:01:35","publication_year":"2018","word_count":1095,"keywords":["Go","AI-first","artificial intelligence","TPU","AI","Alibaba","RPA","Aim","deep learning","GAN","Alphabet","R"],"extracted_tech_keywords":["AI","artificial intelligence","deep learning","Aim","TPU","R","Go","GAN","RPA","AI-first"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/artificial-intelligences-top-moonshot-projects-to-watch-out-for-in-2018\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10019291,"title":"The Digital Transformation Journey Of Mahindra Comviva","content":"A subsidiary of Tech Mahindra and a part of the $21 billion Mahindra Group, Comviva is one of the rising stars in the mobility solutions industry with an extensive portfolio of solutions ranging from mobile finance, digital payments, customer value management to mobile data and managed VAS services. Comviva has been a tour de force in inspiring a communication revolution. However, there was a lot of manual effort and lack of proper process in onboarding and training their end-users, i.e. telecom operators, across multiple applications. This has turned out to be a huge challenge for Mahindra Comviva. Technical documentation also requires high effort as the high-level support queries were resulting in high response time. This urged the telecom technology company to collaborate with Whatfix, a SaaS-based platform to provide in-app guidance and performance support for web applications and software products. This not only helped Comviva to solve challenges related to time-consuming training videos but also made it easy for them to onboard new customers. To understand the case better, Analytics India Magazine spoke to Prema Vibhute, Deputy General Manager of Technical Documentation and UI\/UX, Comviva, and Khadim Batti, the CEO and co-founder of Whatfix. Also Read: Telecom Operators Are Well Poised To Aid In Fight Against COVID-19 Transforming The Legacy The challenges around accessibility to learning documentation, syncing with the product as many change requests are delivered without documentation, time-consuming training videos, and onboarding of users have encouraged the company to go digital. “Comviva decided to pivot from a manual, time and resource-intensive approach to onboarding and training to one that was more personalised, intuitive and utilised the core solutions used by their end-users on a daily basis,” said Khadim. “This made Comviva partner with us to transform training, onboarding and support to a contextual, user-centric experience.” Khadim said Comviva was seeking a two-pronged strategy with the ability to deliver value to end-users and provide internal training. To facilitate this, Whatfix suggested Whatfix Flows for step-by-step in-application guidance; a multiformat content for an automatic content generation; Whatfix Self Help for contextual and on-demand help; and data governance using Whatfix Smart Tips. The project began with discovery and audit, where Whatfix identified the key applications and the challenges specific to each of the aspects. “We needed to capture main tasks, completion obstacles and other key metrics. So, we analysed the current documentation processes and determined the areas of improvements,” said Khadim. The company then translated application tasks to Whatfix Flows. Whatfix Flows are a series of steps to help users to learn an objective or complete a task through a series of step-by-step actions. These steps display as a layer over Comviva’s web application. This involved mapping user context based on end-users’ organisations and application usage to Whatfix Flows. “It was important to isolate specific trouble steps within workflows — areas that might be more complex, and link just-in-time help content to provide an easy and seamless user experience,” explained Khadim. Further, Whatfix conducted testing to improve and refine walkthroughs. To make it easy to share and distribute the walkthroughs within the organisation, Whatfix even configured automatic multiformat content generation, such as PDFs, slideshows etc. It has also set up a Self Help module for users (telecom operators). When asked Comviva about the project, Vibhute said, “We have deployed an export solution and not the cloud version due to high customisation in the product and security related issues at client locations, which ultimately blocks the cloud access.” She added, “Challenges with deployment were there initially, but the support from Whatfix was excellent and beyond expectations, and this has helped us overcome these issues in no time.” Whatfix is an application-agnostic platform and does not require code-level knowledge at the user or admin end. Comviva used Whatfix on internal applications pertaining to marketing and communications. Alongside, Whatfix is also implemented in end user-facing applications to onboard and train wholesalers and retailers. Also Read: How Ludo King Blitzscaled During Pandemic Involvement of Machine Learning Another aspect that massively impacted the digital transformation of Comviva is Smart Context. With machine learning, Whatfix’s Smart Context can automatically identify the best content to be shown to a user. Smart Context intelligently eliminates the need to create multiple segments manually to display relevant content contextually to different audiences on an application. Explaining the process, Khadim said, “When end-users of different organisations access Comviva applications, Whatfix Smart Context intelligently segments and displays only the most relevant content for the specific user.” Using a combination of intelligent segmentation, based on machine learning, helped Comviva improve the accuracy and relevance of the on-demand help content displayed to application users. Whatfix is able to do that by capturing the user interface element on any page along with ten other attributes of the element such as Document Object Model, which is a cross-platform and language-independent interface, Page Hierarchy, Application Module etc. “When the user seeks help from Whatfix, we understand the page context meaning on which module of the application the user is in,” said Khadim. “Based on that, using our proprietary Element Finding technology, we show help content relevant to the user on that page rather than showing all the content.” The Element Finding algorithm has been trained to understand the application context to show the right content when the user is on Module A of Application A based on element attributes captured during content creation. Smart Context also ensures that if the user is not on the first screen of a flow, it starts the flow from the most relevant page that the user is on. This is also possible because machine learning knows the page context based on the learning of which modules\/submodules\/popup modals come in which part of the application. Consequently, Comviva no longer required manual segmentation and assignment of support content for each user. This has significantly reduced the time and effort taken to create, update and publish content for Comviva across applications. Also Read: AI For Improving Customer Engagement: A DAAS Labs Case Study Wrapping Up With the help of Whatfix, Comviva was able to see a reduction in onboarding time and training efforts. The knowledge base of Comviva was integrated with Whatfix’s Self-Help module so that users have a one-stop help within the app and the self-help feature acts as the single source of truth. There was a marked reduction in the time taken to create supporting documentation and videos. The measurement and analysis feature by Whatfix also helped Comviva gather user feedback and insights to understand documentation usage and effectiveness. Not only the company was able to churn out videos faster from the already existing flows, but also the automatically churned “How to” videos with the help of machine learning, have helped users operate hands-on. Lastly, the Smart Tips feature is excellent to have firsthand information.","excerpt":"A subsidiary of Tech Mahindra and a part of the $21 billion Mahindra Group, Comviva is one of the rising stars in the mobility solutions industry with an extensive portfolio of solutions ranging from mobile finance, digital payments, customer value management to mobile data and managed VAS services. Comviva has been a tour de force […]","categories":["IT Services"],"tags":[],"author_name":"Sejuti Das","publish_date":"2021-01-29T14:00:00","publication_year":"2021","word_count":1132,"keywords":["Go","machine learning","AI","ML","Git","RAG","analytics","data governance","GAN","R"],"extracted_tech_keywords":["AI","machine learning","ML","analytics","RAG","R","Go","Git","data governance","GAN"],"url":"https:\/\/analyticsindiamag.com\/it-services\/the-digital-transformation-journey-of-mahindra-comviva\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10084469,"title":"Analytics Olympiad Organised by MachineHack with Shiv Nadar University Concludes Successfully","content":"The Academy of Continuing Education at Shiv Nadar University, Delhi-NCR, in collaboration with MachineHack, successfully completed the annual Analytics Olympiad 2022 for data scientists and machine learning professionals. The competition began on September 30 and concluded on November 6. The two-month-long championship was designed to strengthen the data science community in India and pave the way for innovation. In addition, it was an opportunity for participants to showcase their technical skills and potential in business analytics to prospective employers. The event was a roaring success, with over 1,000 registrations from across the country. After stringent qualifying rounds, the top 9 participants got an opportunity to make in-person presentations before an eminent jury panel at the Shiv Nadar University, Delhi NCR. Aman Garg, Head of Data & AI, DSP Mutual Funds; Anubhav Srivastava, Head of Data Science, Digital Ventures, and Rishit Jain, Director of Product Management, 1mg Technologies, formed the jury for the Analytics Olympiad. After much deliberation, the jury declared Manish Pathak, Senior Data Scientist, the winner, Varun Jagannath, Senior Data scientist – first runner-up, and Rishi Madhav, Presales & Solutions Lead, Data & Analytics – the second runner-up. The winner received a grand prize of INR 1 lakh. The runners-up in the second and third positions were awarded INR 30,000 and INR 20,000, respectively. The problem statement for the participants was to create an ML model that would help an insurance company determine which insurance claims to accept for reimbursement and which to reject. The winning experience Bagging the first rank was Manish Pathak, a BITS Pilani graduate with a dual degree in Electrical and Electronics Engineering and Masters in Physics. Naturally inclined to it, Pathak started exploring data science in his pre-final year. He is currently working as Senior Data Scientist at MiQ Digital India, an ad-tech company where he started as an intern. “Since data science is huge and rapidly evolving, I still learn something new every day,” Pathak said. Varun Jagannath bagged second place in the hackathon. The data scientist from Hyderabad currently works at DBS and thinks of MachineHack as one of the most useful platforms for data science. Rishi Madhav earned the third rank from Bangalore. He has 16+ years of diverse experience from engineering services to automotive & manufacturing. Madhav is interested in the data science and machine learning space and is enthusiastic about building data-driven solutions. The two-day success Apart from the hackathon, three-panel discussions among eminent and experienced personalities were conducted during the two-day event. The insightful session explored several aspects of the data industry for veterans and beginners alike. The first discussion started with Megha Sinha, VP of AI-ML, Practice at Genpact, addressing the importance of collaboration between industry and academia. She spoke about how stakeholders can bridge the gap by setting up CoEs to supplement the existing programs in institutes. She also suggested that companies take more R&D initiatives and bring an entrepreneurial mindset to the campuses. Further, Satyam Mukherjee, Associate Prof, SNU, pointed out that to turn analysis and have high-impact data, industries need to take a step ahead to share their bit of data with academia for a strengthened collaboration. Bhaskar Roy, Client Partner, Fractal AI, emphasised the importance of students knowing how to leverage their knowledge to solve real-world problems. Finally, Deependra Singh, VP and Head of Data Science & Insights, Network18, suggested that research should be of an utmost priority since the projects are a futuristic requirement for a company. The panel session concluded with some actionable insights from the leaders. Moderated by Rishab Parashar, Head of Digital & Performance Marketing, Orient Electric, the second panel jumped straight to how one can figure out if they are the right fit for the exploding data analytics. The performance-based panellists included Rahul Goswami, Associate General Manager and Head of the Deal Acceleration Program at HCL Tech; Ruchi Bubber, Head of Sales Analytics – Pernod Ricard; and Vishesh Gangal, Head of Analytics Solutions & Consulting at WNS Global Services. They touched upon several points on why data science and analytics careers are relevant and important. They also gave their two cents on measures that can be taken by young professionals and students entering the data field. Bubber, explaining the significance of data, said, “Data analytics is not about algorithms or number crunching. It is about making a difference to the business by making it more impactful.” The third discussion focused on the future of the data science and analytics industry. Sreekanth Menon, VP and Global Leader, Genpact, kickstarted the discussion by stating that 11.5 million jobs will be generated worldwide by 2026. Answering what the future of data science holds, Rajat Mathur, Partner at Boston Consulting Group (BCG), said, “As long as you are addressing a problem and value is being created, there will always be a future for this technology.” Along similar lines, Mamta Aggarwal Rajnayak, VP – Head of AiDa AI-ML Products & Platforms – AI Labs, American Express, said that low code\/no code is going to be a reality for every organisation in the future. Saurabh Rai, Global Head of Analytics, AI & Geospatial Initiatives, Tech Mahindra Business Process Services, further joined the discussion. The event also had a keynote session by Prof. Pradeep Chintagunta, Joseph T. and Bernice S. Lewis, Distinguished Service Professor of Marketing at the University of Chicago. He said about data analytics, “One thing that becomes increasingly clear when you start looking at data is that there is always a gap between data and insights. Also, just because a pattern looks different doesn’t mean it is an insight.” Further, he spoke about people in data science not realising that data comes from people, so it is necessary to respect privacy as well as the sanctity of the data. In conclusion, the two-day event highlighted the need to understand data in several sectors – industrially and academically. The panel discussions shed light on the current state of data science and analytics as well as what the future holds for newbies in the field. In addition, the panellists dug deep into their personal experiences, imparting knowledge and insights into what led them to the key positions they hold in their respective organisations today. Both Academy of Continuing Education at Shiv Nadar and Machinehack need to be congratulated for creating such a unique collaboration for working professionals in the field of data science and analytics!","excerpt":"The olympiad received 1000+ registrations from all over the country. After undergoing two stringent rounds, nine participants gave a presentation at Shiv Nadar University, Delhi NCR, before an esteemed jury panel.","categories":["Deep Tech"],"tags":["analytics olympiad","Data Analytics","Data Analytics Certification","data science programs","Hackathon","hackathon for data scientists","Hackathons","Hackathons India","Machinehack Hackathon","MachineHack Shiv Nadar University Hackathon","MachineHack with Shiv Nadar University","New Hackathon For Data Scientists","Shiv Nadar","Shiv Nadar University","Shiv Nadar University data analytics","Shiv nadar university data science","Shiv Nadar University Hackathon"],"author_name":"Tasmia Ansari","publish_date":"2023-01-09T16:46:28","publication_year":"2023","word_count":1061,"keywords":["MachineHack with Shiv Nadar University","Shiv Nadar","Git","New Hackathon For Data Scientists","hackathon for data scientists","Data Analytics Certification","Shiv Nadar University Hackathon","R","Shiv Nadar University data analytics","data science","RAG","analytics","data science programs","Shiv Nadar University","MachineHack Shiv Nadar University Hackathon","Go","analytics olympiad","machine learning","AI","ML","Hackathon","Machinehack Hackathon","Hackathons India","Hackathons","Data Analytics","Aim","Shiv nadar university data science"],"extracted_tech_keywords":["AI","machine learning","ML","data science","analytics","Aim","RAG","R","Go","Git"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/analytics-olympiad-organised-by-machinehack-with-shiv-nadar-university-concludes-successfully\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":68522,"title":"How Is Amazon Aiming To Set A Footprint In The Self-driving Industry?","content":"Amazon recently bought up a self-driving autonomous ride-hailing startup Zoox, which is being claimed as the most ambitious step that the tech giant has taken in the recent past. Reportedly a $1.2 billion deal, the acquisition of the Robo-taxi company is not just to build upon its capabilities to deliver packages but actively set foot in the autonomous driving industry. From Investing In Logistics To Autonomous Vehicles, Amazon Has Come A Long Way While Amazon has invested heavily in developing drones or autonomous delivery robots in the past, its investment in self-driving vehicles has recently gained traction. Some of the other ventures of the company have been in self-driving truck Embark when CNBC reported that it had been hauling Amazon cargo on some of its test runs. For instance, in drones, Amazon has designed a future delivery system to safely deliver packages to customers in a short period of time. It had then shown promise about providing rapid parcel delivery to customers and increasing the overall efficiency of the transportation system. Similarly, with Scout, Amazon aimed to enter the robot delivery game with an electric hamper of wheels. Powered by an electric battery, this vehicle moved at a walking pace. It was also designed to safely navigate around pets and pedestrians. Loaded with sensors and machine learning, it made navigation for Scout easier while making it less risky than self-driving cars. Apart from this, Amazon has also been deeply involved in other initiatives such as releasing dashboard products with Alexa and integrated voice assistants into cars. Amazon also made a considerable investment in self-driving cars with Aurora when it funded the startup along with other investors in 2018. Aurora works in the area of developing software stack for self-driving cars and has catered to names such as Volkswagen, Hyundai and more. AWS In Self Driving Industry Amazon has also been contributing significantly to help collect, store, manage massive amounts of data and provide high-performance computing capacity and advanced deep learning frameworks to develop and deploy Advanced Driver Assistant and Autonomous Vehicle Systems. AWS offers a full suite of services and unlimited storage and computes capacity and support for deep learning frameworks such as Apache MXNet, TensorFlow and PyTorch, which helps in accelerating the algorithm training and testing. As the company notes, AWS’ highly scalable storage and compute services and advanced deep learning frameworks allow for the collection, ingestion, storage and analysis of autonomous vehicle data to support full-scale autonomous vehicle development. As the company shares, it has helped many companies such as TuSimple, HERE, Toyota Research Institute, nuTonomy to set up its self-driving capabilities. For instance, TuSimple built their autonomous driving platform using sophisticated deep learning algorithms developed with Apache MXNet on AWS. HERE uses the scalability, reliability, and global reach of the AWS platform to deliver its HD Live Maps and connected-car solutions to customers. nuTonomy makes software to build self-driving cars and autonomous mobile robots, and more. What Amazon Aims To Accomplish With Zoox Acquisition Now coming to Zoox acquisition, Amazon aims to reinvent the autonomous ride-hailing experience. Founded in 2014, Zoox pioneers in ride-hailing and has designed autonomous technology from the ground up with passengers front-of-mind. By buying out Zoox, Amazon aims to utilise the autonomous technology in its deliveries as well as compete with ride-hailing and food delivery providers, while offering massive scale and resources to the startup to develop its vision. According to reports, Zoox began talking with potential buyers and investors at the beginning of this year, and many big names had been interested in buying the company, which finally ended up going to Amazon. As a part of the agreement, Zoox co-founder and CTO will continue to lead the team as they innovate and drive towards their mission. “Like Amazon, Zoox is passionate about innovation, and we’re excited to help the talented Zoox team to bring their vision to reality in the years ahead,” said Jeff Wilke, Amazon’s CEO, Worldwide Consumer in a blogpost. While other tech giants such as Google and Apple have forayed into the self-driving space, displaying tremendous development in the technology, Amazon is taking its initial steps into this space. With the acquisition of Zoox, Amazon aims to compete with these tech giants and set a bigger foot in the autonomous vehicle industry to further deliver safe, clean and enjoyable transportation to the world.","excerpt":"Amazon recently bought up a self-driving autonomous ride-hailing startup Zoox, which is being claimed as the most ambitious step that the tech giant has taken in the recent past. Reportedly a $1.2 billion deal, the acquisition of the Robo-taxi company is not just to build upon its capabilities to deliver packages but actively set foot […]","categories":["Global Tech"],"tags":["Amazon AWS","Self Driving Cars"],"author_name":"Srishti Deoras","publish_date":"2020-06-30T12:00:00","publication_year":"2020","word_count":725,"keywords":["machine learning","AWS","AI","PyTorch","Self Driving Cars","RAG","Amazon AWS","Ray","Aim","deep learning","TensorFlow","R"],"extracted_tech_keywords":["AI","machine learning","deep learning","Aim","Ray","TensorFlow","PyTorch","RAG","AWS","R"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/how-is-amazon-aiming-to-set-a-footprint-in-the-self-driving-industry\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10041415,"title":"Top Crypto APIs In The Market","content":"Digital currencies have piqued the curiosity of both individual and institutional investors in recent months. Bitcoin and other cryptocurrencies have risen to prominence as a new asset class with exceptional returns over the last decade.  At the heart of crypto-success lies cleverly wrapped APIs. APIs allow businesses to access a wide range of resources and enable businesses to enter markets they may not have previously considered. APIs enable one to harness real-time crypto data to aid trades, construct a trading bot, perform academic research on cryptocurrency, or learn to code with crypto data. Keeping track of wallet contents, checking market prices, and even making and receiving transactions are all possible with crypto APIs. APIs are highly beneficial for almost any project involving bitcoin or a blockchain. There are numerous APIs to select from if you are a cryptocurrency trader, speculator, developer, or someone interested in undertaking cryptocurrency research. Many cryptocurrency exchanges have made their APIs public, allowing developers to connect to their platforms. Here is a list of five APIs: 1| BraveNewCoin (BNC) API BNC’s cloud-based cryptocurrency APIs provide a quick and accurate reference for real-time and historical crypto data from over 200 exchanges, covering thousands of assets and marketplaces. Institutions and the developer community have access to a collection of private and public API maintained by BNC. The API provides a volume-weighted spot price for 1500 plus cryptographic assets from across the global network of crypto exchanges. They also deliver a comprehensive history of crypto-asset market data across all products, including the LX program, Crypto Spot Rates, Crypto Market Rates and Market Cap Table. 2| CoinAPI The data on the platforms is standardised. To maintain the highest level of quality, market symbol details and asset codes are double-checked by real humans. Moreover, on-demand data is delivered to the user using a simple HTTP RESTful API in JSON, XML or CSV formats. It provides access to a database of historical market data in addition to a real-time data streaming service. More than 20 TB of raw and preprocessed market data is presently available in the CoinAPI database. 3| Coinbase API The API allows you to include bitcoin, bitcoin cash, litecoin, and ethereum payments into your website or app. They are capable of generating bitcoin, bitcoin cash, litecoin and ethereum wallets and addresses. One can buy\/sell and send\/receive bitcoin, bitcoin cash, litecoin and Ethereum. Securely storing bitcoin, bitcoin cash, litecoin and Ethereum, retrieving real-time or historical price information and receiving notifications when payments arrive are some of the capabilities it holds. 4| CoinMarket API The API has up to five years of historical data available since 2013 to give users the chance to look more closely at the ups and downs of the crypto market to make an informed decision. The updated data will be helpful for building a wallet, a portfolio management tool or more. Current and past pricing for various cryptocurrencies are readily available, in addition to data aggregated from hundreds of exchanges and thousands of coins. 5| Bittrex API Bittrex is a secure, dependable, and innovative digital asset trading platform built on Bittrex’s cutting-edge technology for crypto traders. Users can programmatically access their Bittrex accounts and execute trades, withdrawals, and deposits using the Bittrex API. It has a RESTful architecture and uses the API Key protocol for authentication. XML, JSON answer format, and URI Query String\/CRUD request formats are all supported. 6| CoinGecko CoinGecko is a cryptocurrency data aggregator that tracks over 6,000 different crypto assets on more than 400 exchanges across the world. CoinGecko is one of the  independent sources of trustworthy cryptocurrency statistics, and it is widely quoted in the field. CoinGecko measures community growth, open-source code development, key events, and on-chain metrics in addition to price, volume, and market capitalisation. 7| Poloniex API Poloneix is owned by Circle and is one of the world’s largest crypto exchanges. Poloniex provides APIs for connecting with the exchange over HTTP and websocket. Both provide public market data read access as well as private account read access. The private HTTP API gives you private write access to your account. The public HTTP endpoint is accessed via GET requests while the private endpoint is accessed via HMAC-SHA512 signed POST requests using API keys. 8| HitBTC API Utilising third-party applications or customised software, a user can trade multiple crypto and fiat trading pairs using the Hitbtc API. This API is a single-purpose API with a RESTful architecture that employs an API Key and HTTP Basic authentication approach. It supports XML responses as well as URI Query String\/CRUD, FIX, and WebSocket requests. 9| KuCoin API Users of the KuCoin API can access their accounts via a programmable interface. A developer can use this API to create trading bots that users can employ to execute high-frequency trades. This API is a single-purpose API with a RESTful architecture that employs an API Key and OAuth 2 authentication approach. It can handle XML response as well as JSONP, URI Query String\/CRUD requests. 10| Bitstamp API The Bitstamp API allows customised software (bots) and even third-party applications to have access to user accounts and execute trades. This API is a single-purpose API with a RESTful architecture that uses an API Key authentication model. It can handle XML responses as well as JSONP, URI Query String\/CRUD requests.","excerpt":"Digital currencies have piqued the curiosity of both individual and institutional investors in recent months. Bitcoin and other cryptocurrencies have risen to prominence as a new asset class with exceptional returns over the last decade.  At the heart of crypto-success lies cleverly wrapped APIs.  APIs allow businesses to access a wide range of resources and […]","categories":["AI Trends"],"tags":["crypto market","cryptocurrency trading"],"author_name":"kumar Gandharv","publish_date":"2021-06-08T14:00:00","publication_year":"2021","word_count":880,"keywords":["cryptocurrency trading","API","programming_languages:R","AI","crypto market","ML","Git","WebSocket","Rust","R","programming_languages:Rust"],"extracted_tech_keywords":["AI","ML","R","Rust","Git","API","WebSocket","programming_languages:R","programming_languages:Rust"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-crypto-apis-in-the-market\/","complexity_score":4,"technical_depth":9,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10022449,"title":"Kornia: An OpenCV-inspired PyTorch Framework","content":"Kornia is an open-source Python library inspired by OpenCV designed to handle generic Computer Vision tasks. It was introduced by Edgar Riba, Dmytro Mishkin, Daniel Ponsa, Ethan Rublee and Gary Bradski in October, 2019 (research paper). Kornia leverages PyTorch library at its backend in terms of model’s efficiency and reverse-mode auto-differentiation for defining and computing complex functions’ gradients. It comprises a subset of packages having operators that act as an input to neural networks for performing a wide range of tasks such as image transformations, depth estimation, epipolar geometry, filtering and edge-detection applicable on high-dimensional tensors. Unlike conventional CPU-based CV libraries such as torchvision and scikit-image, several standard deep learning functions can be implemented on GPUs using Kornia. Kornia bridges the gap between two simple yet powerful libraries namely, OpenCV and PyTorch. Though solely based on traditional CV solutions like torchvision, tf.image, PIL and skimage, it enables differentiable programming for CV applications by utilizing the crucial properties of PyTorch like GPU hardware acceleration, differentiability and distributed data-flows. The following table compares Kornia with some important CV libraries\/modules: Library\/ModuleCPU supportedGPU supportedBatch ProcessingDifferentia-bleDistributedSupports multidimensional arraytorchvisionYesNoNoNo NoNoscikit-imageYesNoNoNoNoNoopencvYesYesNoNoNoNotf.imageYesYesYesYesYesYesKorniaYesYesYesYesYesYes Source of above summary: Research paper Components of the library kornia.augmentation – module for GPU-based image augmentation kornia.utils – image to tensor conversion utilities and metrics kornia.color – module for color space conversions kornia.morphology – module for morphological image transformations kornia.losses – collection of several loss functions kornia.geometry – for geometric CV tasks like image conversion using various camera models kornia.enhance – module for intensity transformations and normalization kornia.filters – for edge detection and image filtering use cases kornia.feature – for feature setection Practical implementation Here, we demonstrate three use cases of Kornia – blurring a custom image, changing its color space, and adjusting its colors. The code is test on Google colab with Python 3.7.10 and Kornia 0.4.1 versions. We have used the following image for demonstration: Image source Step-wise explanation of the code is as follows: First, install the library using pip command. !pip install kornia Import required libraries and modules import torch import torchvision import kornia import cv2 import numpy as np import matplotlib.pyplot as plt Image blurring using Kornia Read the input image #Read the image using OpenCV and convert it to a numpy array input_image: np.ndarray = cv2.imread('img2.jpg') #Convert the image color space from BGR to RGB input_image = cv2.cvtColor(input_image, cv2.COLOR_BGR2RGB) Convert the image to a torch 4D tensor tensor_image: torch.tensor = kornia.image_to_tensor(input_image, keepdim=False) Create an operator that blurs the tensor image using Gaussian filter gaussian = kornia.filters.GaussianBlur2d((11, 11), (10.5, 10.5)) Where, (11,11) is the size of the kernel and (1.05,10.5) is the standard deviation of the kernel Convert the tensor image to float type and apply the gaussian operator defined in previous step to blur the image blur_image: torch.tensor = gaussian(tensor_image.float()) Convert the blurred tensor image back to a numpy array final_blur_image: np.ndarray = kornia.tensor_to_image(blur_image.byte()) Plot the input and output images fig, ax = plt.subplots(1, 2, figsize=(16, 10)) ax = ax.ravel() #flatten the axis to a 1D array #Original image ax[0].axis('off')  #turn off the axis lines and labels ax[0].set_title('Original image')  #Title of image ax[0].imshow(input_image)   #Display the image #Blurred image ax[1].axis('off')    #turn off the axis lines and labels ax[1].set_title('Blurred image')   #Title of image ax[1].imshow(final_blur_image)  #Display the image Output: Color space conversion using Kornia Read the input colored image and convert it to a numpy array bgr_image: np.ndarray = cv2.imread('img2.jpg', cv2.IMREAD_COLOR) Convert the image to a torch tensor tensor_bgr: torch.Tensor = kornia.image_to_tensor(bgr_image, keepdim=False) Define functions to flip the image horizontally and vertically and rotate it by 180 degrees. def horizontal_flip(input: torch.Tensor) -> torch.Tensor: return torch.flip(input, [-1]) #torch.flip() reverses the order of the input tensor image along given axis def vertical_flip(input: torch.Tensor) -> torch.Tensor: return torch.flip(input, [-2]) def rotate_180(input: torch.Tensor) -> torch.Tensor: return torch.flip(input, [-2, -1]) Define a function to plot a batch of images def imshow(input: torch.Tensor): #Create a grid with 2 rows output: torch.Tensor = torchvision.utils.make_grid(input, nrow=2, padding=5) #Convert the grid of images to numpy ndarray output_np: np.ndarray = kornia.tensor_to_image(output) plt.imshow(output_np)  #Plot the grid plt.axis('off')  #Turn off the axis lines and labels plt.show()   #Display the grid Create a batch of images from BGR image #Apply the functions defined in step (3) to the input tensor image and then #concatenate resulting tensors batch_bgr = torch.cat([tensor_bgr, horizontal_flip(tensor_bgr), vertical_flip(tensor_bgr), rotate_180(tensor_bgr)]) #Display the batch of images imshow(batch_bgr) Output: BGR to RGB color space conversion batch_rgb = kornia.bgr_to_rgb(batch_bgr)  #conversion imshow(batch_rgb) #display the resulting Output: RGB to Grayscale color space conversion batch_gray = kornia.rgb_to_grayscale(batch_rgb.float() \/ 255.) imshow(batch_gray)  #display the grayscale image Output: RGB to HSV (Hue, Saturation, Value) color space conversion batch_hsv = kornia.rgb_to_hsv(batch_rgb.float() \/ 255.) imshow(batch_hsv[:, 2:3])  #display HSV image Output: Color adjustment Read the input image in BGR format bgr_image: np.ndarray = cv2.imread('img2.jpg', cv2.IMREAD_COLOR) Convert input image to torch tensor tensor_bgr: torch.Tensor = kornia.image_to_tensor(bgr_image) Convert the image from BGR to RGB tensor_rgb: torch.Tensor = kornia.bgr_to_rgb(tensor_bgr) Expand the image tensor_rgb = tensor_rgb.expand(4, -1, -1, -1) # 4 x Channels x Height x Width tensor_rgb = tensor_rgb.float() \/ 255. #Normalize the expanded image Define a function to create a batch of images def imshow(input: torch.Tensor): #create grid having 2 rows of images output: torch.Tensor = torchvision.utils.make_grid(input, nrow=2, padding=5) #Convert tensor images to numpy ndarray output_np: np.ndarray = kornia.tensor_to_image(output) #Plot the grid of images plt.imshow(output_np) plt.axis('off') plt.show() Display the batch of RGB images imshow(tensor_rgb) Output: Adjust brightness of the batch images tensor_brightness: torch.Tensor = kornia.adjust_brightness(tensor_rgb, 0.6) #Where, 0.6 is the factor to adjust brightness of each element in the batch imshow(tensor_brightness)  #display the resulting batch Output: Adjust contrast of the batch images tensor_contrast: torch.Tensor = kornia.adjust_contrast(tensor_rgb, 0.2) #0.2 is the contrast adjustment factor per batch element imshow(tensor_contrast)  #dispaly the resulting batch Output: Adjust gamma correction of the images tensor_gamma: torch.Tensor = kornia.adjust_gamma(tensor_rgb, gamma=3., gain=1.5) #where, ‘gamma’ is a non-negative number and ‘gain’ is the constant multiplier imshow(tensor_gamma)  #display resulting batch Output: Adjust saturation level of the batch images tensor_saturated: torch.Tensor = kornia.adjust_saturation(tensor_rgb, 0.2) #where 0.2 is the saturation factor imshow(tensor_saturated)  #display the resulting batch Output: Adjust hue of the images tensor_hue: torch.Tensor = kornia.adjust_hue(tensor_rgb, 0.5) #where 0.5 gives measure of how much to shift the hue channel imshow(tensor_hue)  #display the resulting batch Output: Google colab notebook of the above-explained implementation is available here. References To get in-depth understanding of the Kornia library, refer to the following web links: Research paper Official documentation GitHub repository","excerpt":"Kornia is an open-source Python library inspired by OpenCV designed to handle generic Computer Vision tasks. It was introduced by Edgar Riba, Dmytro Mishkin, Daniel Ponsa, Ethan Rublee and Gary Bradski in October, 2019 (research paper).  Kornia leverages PyTorch library at its backend in terms of model’s efficiency and reverse-mode auto-differentiation for defining and computing […]","categories":["AI Trends"],"tags":["Computer Vision","Guide"],"author_name":"Nikita Shiledarbaxi","publish_date":"2021-03-20T10:00:00","publication_year":"2021","word_count":1046,"keywords":["NumPy","AI","neural network","PyTorch","computer vision","OpenCV","Colab","Ray","deep learning","Computer Vision","Matplotlib","Guide"],"extracted_tech_keywords":["AI","deep learning","neural network","computer vision","Ray","PyTorch","Colab","OpenCV","NumPy","Matplotlib"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/guide-to-kornia-an-opencv-inspired-pytorch-framework\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10166309,"title":"Kunal Bahl Suggests Swades-Style Ghar Wapsi for Indian Startups","content":"In a trend that’s become increasingly popular over the years, many Indian founders register their startups in the West or Singapore. Why, you may ask? The benefits are aplenty. For starters, it makes fundraising easier by allowing them to find more institutional and technological investors. However, not everyone is impressed by the idea of foreign-born startups of Indian founders. Snapdeal co-founder Kunal Bahl, who recently shared a detailed thread on X, argues that Indian startups should be incorporated domestically rather than overseas. His stance challenges what was once conventional wisdom among Indian entrepreneurs who believed foreign incorporation provided fundraising advantages, including easier exits. “In 2025, that’s outdated. Today, incorporating in India isn’t just patriotic—it’s pragmatic,” Bahl said. “Indian investors now back all kinds of startups, including deep tech and AI, and one doesn’t need to find investors overseas to fund these spaces,” he explained. Bahl argued that recent government initiatives have brought in improved taxes and regulatory clarity for Indian startups. In the 2024 Union Budget, finance minister Nirmala Sitharaman announced the abolition of angel tax (effective FY2025-26) for all investor classes, addressing a long-standing concern in the startup ecosystem. The tax, which previously treated investments above fair market valuation as income, had been a significant deterrent for domestic incorporation. Additionally, the finance minister extended the definition of “eligible startup” under the Startup India scheme to include entities incorporated between April 1, 2016, and March 31, 2025. It allows more startups to benefit from the tax holiday offered under the scheme, directly supporting Bahl’s point about India’s startup-friendly tax policies. Are Startups Really Coming Back? In 2025, over 90 companies have filed their draft prospectuses in India, aiming to raise an estimated INR 1 trillion or $11.65 billion. About 34 companies have either already raised or are announcing their fundraising efforts. Citing the success of recent IPOs like Zomato, Nykaa, and Unicommerce, Bahl said that the Indian market is ready. “If you plan to list in India, incorporating here avoids costly ‘flipping’ later,” he said. Quick-commerce platform Zepto announced its ‘ghar wapsi’ in January 2025, moving from Singapore to India. This aligned with the company’s plans to launch its IPO, making KiranaKart, its domicile holding in India, the parent company. However, they are expected to attract a huge tax bill for this ‘flip’. Meanwhile, Groww, a financial trading platform, shifted its domicile status from the US to India last year under its parent company, Billionbrains Garage Ventures. Walmart-owned Flipkart is also reportedly planning to move its domicile from Singapore to India for the same reasons. According to a report by The Economic Times, Meesho is reverse flipping from Delaware in the US back to India, while Pine Labs is also moving back from Singapore. PhonePe is also moving back to India and is eyeing an IPO. In a previous interaction with AIM, Akash Aggarwal, MD (investment banking) at Motilal Oswal Financial Services, reflected upon the same. Having participated in several IPOs in the past, Aggarwal said that despite the claims of the market being down, IPOs are seeing decent subscriptions among Indian investors. “A majority of the money comes from Indian investors and not foreign investors,” Aggarwal told AIM. He added that as of January 21, 2025, the BSE index had fallen by almost 10% compared to September 2024, when the market surpassed the 84,000 mark, its highest ever. “Almost 70-80% of the interest is from domestic investors. Of the several companies that I am in touch with, some are going IPO, and I think this is the right time for it because it might take at least 9-12 months for them to launch the deal,” Aggarwal said, adding that if a company is mature enough, it should think about it. In an interview with BusinessLine in January, Mahavir Lunawat, chairman of the Association of Investment Bankers of India, said that Indian firms used to take pride in raising funds abroad, but now foreign firms line up to raise funds in India. The Department for Promotion of Industry and Internal Trade (DPIIT) recognition system provides substantial benefits to domestically incorporated startups which includes tax exemptions and a simplified compliance process, as highlighted by Bahl in his thread. The Problem with Ghar Wapsi Vaibhav Dusad, co-founder of SurgeGrowth, while speaking about Bahl’s analysis on LinkedIn, said that it completely misses the ground reality. “Sentiment-wise, I’m all in—build in India, win from India, keep the tax rupees here,” he said. “But as a SaaS founder who’s actually done it, I can tell you it isn’t as shiny as the pitch.” Dusad said that India’s IPO market is a legit exit path since $5 million in revenue can get startups there, which is much less compared to the half a billion that startups in the US look for. “In the US, it seems too unrealistic to do an IPO—even biggies like Stripe haven’t done it yet,” Dusad added. He highlighted problems like the KYC process when wiring money from the US and the poor infrastructure. “In India, you’re not just fighting for customers—you’re wrestling with bureaucracy and red tape. Ultimately, your focus is diverted from what matters most: building your product!” However, the continued emergence of Indian unicorns supports Bahl’s perspective. In February 2025, the banking technology platform Zeta became India’s newest unicorn, reaching a $2 billion valuation with its latest funding round. According to data from AIM Research’s AI Startup Funding Report India 2025, early-stage funding experienced a substantial reduction of 37% compared to last year. It indicates that investors are looking for startups that have significant market viability. However, this is just for AI startups in India. According to Tracxn’s Geo Annual Report – India Tech–2024, a total of $11.31 billion were raised that year, which is an increase of 3.68% compared to $10.91 billion raised in 2023, but a significant drop of 56.17% compared to $25.80 billion raised in 2022. These market conditions raise questions about Bahl’s first point that “nearly all VCs are funding Indian entities”. While Indian investors may indeed be willing to back various types of startups including deep-tech and AI ventures, the overall funding environment appears constrained. Six startups — Cashfree, Zeta, ToneTag, SpotDraft, Udaan, and Geniemode — raised more than $50 million in 2025 alone. This demonstrates that significant funding remains available for promising Indian-incorporated startups. Meanwhile, in the second half of 2025 and 2026, at least 60 companies are expected to seek exits through IPOs, mergers, or acquisitions. As Dusad put it: “Incorporate in India if you’re playing the long game—IPOs & local pride are worth it. But if you’re early-stage, chasing PMF, and need to move fast, the US is still king.”","excerpt":"In 2025, over 90 companies have filed their draft prospectuses in India, aiming to raise an estimated INR 1 trillion or $11.65 billion.","categories":["AI Features"],"tags":["Startups"],"author_name":"Mohit Pandey","publish_date":"2025-03-19T14:08:24","publication_year":"2025","word_count":1109,"keywords":["Go","funding","unicorn","AI","RPA","Git","RAG","Aim","Startups","R","startup"],"extracted_tech_keywords":["AI","Aim","RAG","R","Go","Git","RPA","startup","unicorn","funding"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/kunal-bahl-suggests-swades-style-ghar-wapsi-for-indian-startups\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":58827,"title":"How Amazon Is Using AI To Better Understand Customer Search Queries","content":"Being an early adopter of artificial intelligence and automation, Amazon always had an edge in using AI to improve its business efficiencies. Not only has it been using AI to enhance its customer experience but has been heavily focused internally. From using AI to predict the number of customers willing to buy a new product to running a cashier-less grocery store, Amazon’s AI capabilities are designed to provide customised recommendations to its customers. According to a report, Amazon’s recommendation engine is driving 35% of its total sales. One of the main areas where Amazon is applying continuous AI is to better understand their customer search queries and what is the reason they are looking for a particular product. For an e-commerce company to make relevant recommendations to its customers, it is not only crucial for them to know what their customers searched for, but it is also critical to understand why a customer is searching for a product. Understanding the context can help the retailer to recommend complementary items to its customers, and Amazon is intent to work out this puzzle by applying AI to the problem. In a recently posted blog, Amazon discussed using AI and machine learning to predict the context from their customers’ search queries. This system has been aimed to augment the quality of search results on Amazon.com platform, which indeed intended towards enhancing the overall Amazon’s shopping experience. Explaining further, in a paper accepted to the ACM SIGIR Conference on Human Information Interaction and Retrieval, Amazon researchers described how most retailers use product discovery algorithms to look for correlations between queries and products; however, Amazon used their AI to identify the best matches depending on the context of use. Therefore the system predicts activities like “running” from customer queries like “Adidas men’s pants” or if an Amazon customer enters the query “waterproof shoes”, is she looking to go for a weeklong hike? According to Amazon, predicting the intent of the query is a significant component of information retrieval which in turn, improves the relevance of the results through an understanding of latent user intents in addition to explicit query keywords. The researchers believe this might improve people’s shopping experience by matching only high-quality products to search queries. Training The System The first step of the process was to train the system for which the team had to build a data set. In order to build the data set, the team assembled a list of 173 context-of-use categories divided into 112 activities — such as reading, cleaning, and running — and 61 audiences — like a child, daughter, man, and professional — based on common product queries. They used standard reference texts to create aliases for the terms they used to denote the categories. Such as for the category ‘father’ they included ‘dad’, ‘daddy’, ‘pops’ etc. or for ‘mother’ they included ‘mum’, ‘mommy’, ‘mom’ etc. and then they used their in-house dataset to co-relate million of their products to particular query strings. They also scoured online reviews of their products to label them with their category terms and their aliases — also known as simple binary classification. The in-house dataset that Amazon used, correlates their query strings with products according to an affinity score — from 1 to 15, where a low score indicates a weak correlation. But, to train their context-of-use predictor system, Amazon researchers created another data set, where each entry was labelled with three data items — a query; a product ID, which has been added by context-of-use categories; and the affinity score derived from the in-house dataset. This data set was then divided into two smaller sets — one annotated according to activity and one according to the audience, and from each of those smaller datasets they constructed two more — one with high-affinity score of 15 and one which was low as 8. This resulting data set was then used to train six different machine learning models. Multiple Metrics Once decided to train six different models, the system segregated the models in terms of the affinity score of their training data. The ones had an affinity threshold of 15 were trained using binary cross-entropy, which imposes particularly stiff penalties on incorrect classifications that get high confidence scores. But the ones that had an affinity threshold of 8, Amazon researchers used both binary cross-entropy and B-weighted binary cross-entropy — where it also weights the penalty incurred by each data item according to its affinity score. The resulted six models were trained to predict context-of-use based on customers’ query strings. In tests, the best-performing model managed to anticipate product annotations with 97% accuracy for activity categories and 92% for audience categories. And, when asked by human reviewers to indicate the classifications they agreed on, they said, an average of 81% of the time the system’s per-item predictions have been correct. “This suggests, according to Adrian Boteanu, an applied scientist in the customer experience division of Amazon Search that, “The contexts-of-use identified by Amazon’s system could help product discovery algorithms to deliver more relevant results, improving the customer experience. Moreover, the minimal human supervision required to produce training data means that the method could be expanded to new categories with relatively little effort.” Wrapping Up As Amazon continues to improve its algorithms, customers shopping on Amazon will see increasingly relevant shopping recommendations. According to Amazon, such research could open a whole new prospect for personalised digital shopping assistants. In this dynamic world where the tech giants are still struggling with their internal bureaucracy and technology silos, it is exceptional to see how Amazon keeps emerging with encouraging innovations to enhance the customer experience.","excerpt":"Being an early adopter of artificial intelligence and automation, Amazon always had an edge in using AI to improve its business efficiencies. Not only has it been using AI to enhance its customer experience but has been heavily focused internally.  From using AI to predict the number of customers willing to buy a new product […]","categories":["Global Tech"],"tags":["AI (Artificial Intelligence)","Amazon","Amazon AWS","Artificial Intelligence India"],"author_name":"Sejuti Das","publish_date":"2020-03-17T11:09:55","publication_year":"2020","word_count":939,"keywords":["Go","artificial intelligence","machine learning","AI","Artificial Intelligence India","Amazon","Git","RAG","Amazon AWS","automation","Aim","ViT","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","Aim","RAG","R","Go","Git","ViT","automation"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/how-amazon-is-using-ai-to-better-understand-customer-search-queries\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10170944,"title":"The $50 Billion Market That’s Haunting NVIDIA","content":"NVIDIA chief executive Jensen Huang is fretting. This is evident in the loss of opportunity to do business in China, one of the world’s leading AI markets with a $50 billion potential. In recent months, Huawei Technologies, NVIDIA’s Chinese counterpart, has gained in NVIDIA’s absence and is building an advanced chip production facility to deliver multiple parts for the AI supply chain. Even though the US chip maker maintains a 90% share of the global GPU market, the potential loss of revenue from the Chinese market, estimated at around $8 billion for the upcoming quarter and roughly a sixth of its Q1 earnings, has rattled the company. US President Donald Trump maintained the export restrictions on chips, amid repeated warnings from industry leaders that it could threaten the country’s AI dominance. Huang called it a major strategic blow to the US and said the restrictions have spurred China’s innovation and scale. “Shielding Chinese chipmakers from U.S. competition only strengthens them abroad and weakens America’s position,” Huang said during the company’s earnings call on Wednesday, calling the East Asian country a springboard to global success. “With half of the world’s AI researchers based there, the platform that wins China is positioned to lead globally,” he said. “China’s AI moves on with or without U.S. chips. It has to compute to train and deploy advanced models.” Huang said the country’s assumption that China cannot make AI chips was flawed, and considering China’s enormous manufacturing capabilities, it has proved them wrong now. On April 9, 2025, the US government informed NVIDIA that a license is required to export its H20 products to the Chinese market. As a result of these new requirements, NVIDIA incurred a $4.5 billion charge in the first quarter of fiscal 2026. The company reported $4.6 billion in H20 product sales in the quarter prior to the licensing requirement. NVIDIA CFO Colette Kress said during the earnings call that the company is still evaluating its limited options to supply data centre compute products to China, compliant with the US government’s revised export control rules. She added that in the data centre, the company anticipates the continued ramp of Blackwell to be partially offset by a decline in China revenue. “Our outlook reflects a loss in H20 revenue of approximately $8 billion for the second quarter,” Kress said. According to Taiwanese outlet Digitimes, which cited sources within the supply chain, NVIDIA and AMD are preparing to launch new AI GPUs for the Chinese market to comply with US export restrictions on advanced semiconductor technology. NVIDIA is reportedly set to introduce a scaled-down AI GPU known internally as the “B20”, while AMD plans to offer its new Radeon AI PRO R9700 workstation GPU to meet local AI computing needs. Both companies are expected to begin shipments in China by July. Separately, Reuters recently reported that NVIDIA is developing a lower-cost AI chip based on its Blackwell architecture for the Chinese market. The chip’s projected price range is $6,500 to $8,000, which would place it below the price of its H20 GPU, which currently sells for between $10,000 and $12,000. The Rise of Huawei Notably, according to a Reuters report, Huawei is set to begin mass shipments of its Ascend 910C AI chip to Chinese customers in May. The Ascend 910C is an upgraded GPU that combines two 910B processors into a single package, effectively doubling computing power and memory capacity. Meanwhile, Huawei is preparing to test its new AI chip, the Ascend 910D, in China, while it has also introduced the Ascend 920 AI chip. According to DigiTimes Asia, the chip is slated to enter mass production in the second half of 2025. Industry experts believe the Ascend 920 could be a viable alternative to NVIDIA’s H20 GPUs. Meanwhile, Cambricon Technologies is also gaining prominence in China. In a recent report, Cambricon said it had successfully optimised Qwen3 to run efficiently on its GPUs. This optimisation was driven by demand from AI developers in the Philippines seeking China-made chips. Meanwhile, Trump also warned against using Chinese AI chips, leading to criminal penalties. $44.1 Billion in Revenue Despite Hit Despite the setback in China, NVIDIA reported $44.1 billion in revenue in Q1FY26, up 12% from the previous quarter and 69% from last year. This comes as NVIDIA announces plans to build AI factories in the US and partner with local manufacturers to produce NVIDIA supercomputers domestically. The company introduced Blackwell Ultra and Dynamo to accelerate large-scale AI reasoning. Kress also clarified that while Singapore accounted for nearly 20% of Q1 billed revenue, over 99% of the H100, H200, and Blackwell compute revenue billed to Singapore came from US-based customers. Meanwhile, authorities in Singapore and the US are also investigating allegations that $390 million worth of NVIDIA chip servers sent to Malaysia were fraudulently redirected to China. NVIDIA is also partnering with HUMAIN to establish AI factories in Saudi Arabia and unveiled Stargate UAE, a next-gen AI cluster in Abu Dhabi, with partners including G42, OpenAI, Oracle, SoftBank, and Cisco. Moreover, the company is collaborating with Foxconn and the Taiwan government on an AI factory supercomputer. Beyond Chips Huang framed the issue beyond individual chip models or product lines. “The AI race is not just about chips. It’s about which stack the world runs on. As that stack grows to include 6G and quantum, U.S. global infrastructure leadership is at stake.” He emphasised that “export controls should strengthen US platforms, not drive half of the world’s AI talent to rivals.” Pointing to China’s DeepSeek and Qwen models, Huang said, “DeepSeek-R1, like ChatGPT, introduced reasoning AI that produces better answers the longer it thinks. Reasoning AI enables step-by-step problem solving, planning and tool use, turning models into intelligent agents.” All said and done, Huang maintained that US platforms must remain the preferred platform for open-source AI even if it means collaborating with top developers globally, including in China.” The question remains whether China will have its own AI capabilities and chip dominance or run on the American tech stack.","excerpt":"“China’s AI moves on with or without U.S. chips. It has to compute to train and deploy advanced models.”","categories":["Global Tech"],"tags":["NVIDIA"],"author_name":"Siddharth Jindal","publish_date":"2025-05-29T16:52:55","publication_year":"2025","word_count":1004,"keywords":["Go","ChatGPT","OpenAI","AI","RPA","innovation","Git","GPT","AI research","NVIDIA","R"],"extracted_tech_keywords":["AI","ChatGPT","OpenAI","R","Go","Git","GPT","RPA","innovation","AI research"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/the-50-billion-market-thats-haunting-nvidia\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":17971,"title":"Scope of Data Science in Military, Intelligence &#038; Law Enforcement","content":"The modern-day military and its personnel are very different from the past centuries. The military has become far superior and sophisticated with the advancement of technology. Nowadays one can see more of slim fit intellectuals in uniforms rather than the beefed up and ripped officers. The reason for this change is that apart from physical strength, the modern day military, intelligence and law enforcement operations require a lot of interpretation of information which they receive through varied channels like phones, internet, satellite etc. In all the domestic and overseas bases, such data is generated every day, every minute or rather every second. This information or data can be utilized to bring out useful insights to protect the homeland. But are there enough skilled personnel to carry out such analysis? Well, a 2012 military report “Lightening the Information Load” quoted an officer in Afghanistan as saying, “…90% of the information requested was already in the system, but the person asking didn’t know where to look or how to extract it.” Indian military and law enforcement organizations do not have a dedicated post for a data science expert as of now. Although it has been realised by the Indian military fraternity that there is a need for such a position as there have been discussions taken place in CLAWS (Centre for Land Warfare Studies). The analysis is outsourced to private organisations due to the lack of data analysts or dedicated data analysis department in the military, paramilitary and law enforcement organizations in India. A Gurugram based company which goes by the name Chron, has developed border security systems for the BSF. It has found out with the help of past data that whenever there are cattle moving around the border fence, there is a high chance of cross-border intrusion. The movement of the cattle ensured that there are no landmines near the fence and it would be safe for the intruders to cross over. However, there is a potential threat to the national security to outsource the border’s important data to be given in the hands of private players for analysis. Therefore there is an utmost need for the military and law enforcement organizations to have a dedicated department for data analytics. The use of data analytics in the military not only lies in intelligence, surveillance, border management, maritime management, space management, operational planning but also in logistics, financial management, disaster management, future technologies, cognitive analysis and analysis of historical data. Many intelligence organisations who are largely involved in overseas operations generate a lot of data and have realized the need for data scientists. For example, the CIA (Central Intelligence Agency) has a dedicated post for a Data Scientist. The salary for this post ranges from $62,338 to $199,794. They are expected to organize and interpret Big Data to inform the decision makers of the United States, drive successful operations and shape CIA’s technology and resource investments. These Data Scientists not only work with the advanced hardware and software but also with the techniques to develop computational algorithms and statistical methods that find patterns and relations in the vast sea of data. They deliver their conclusions with coherence to a lay audience. They even get good opportunities to become experts through CIA’s sponsored continuous education, attending scholarly and technical symposiums, and networking with the Intelligence Community. The Agency also gives them the access to unique data sets that can be analysed in one computational environment. However, there will be a time in future when militaries across the globe will realize the need for Data Scientists. No matter how large a database the nation may have on its national security, it will always require highly skilled and big teams to study it and bring it to the table in a readable state. Further, the Data Scientists may themselves be required to analyse the vital information out of it. The data from a foreign country might as well be in a foreign language for the military intelligence to interpret. In such a case the future militaries will even have polyglot Data Scientists in their intelligence wings in order to analyse data in foreign languages. There might still be raw data hidden right now, somewhere in the data repositories of the numerous security organisations in various nations which could fail the next big attack on a homeland. But finding a good Data Scientist for this interpretation can be a challenge as the saying goes in the data fraternity ‘Data Scientists are as rare as unicorns.’ Data scientists in the military, intelligence and law enforcement benefits the organization by mitigating the risk of failure, gaining a better understanding of the anti-national elements, receiving unique insights, aiding in efficient expansion and improving event predictions. These benefits will further reduce operational casualties, collateral damage, fall in crime rate, fall in the rate of terror attacks, and fall in the rate of cross-border intrusions. Soon a day will come when data science will play an ace role in the security of a nation.","excerpt":"The modern-day military and its personnel are very different from the past centuries. The military has become far superior and sophisticated with the advancement of technology. Nowadays one can see more of slim fit intellectuals in uniforms rather than the beefed up and ripped officers. The reason for this change is that apart from […]","categories":["IT Services"],"tags":["Data Science","Military"],"author_name":"Debashish Roy","publish_date":"2017-10-03T05:58:17","publication_year":"2017","word_count":834,"keywords":["big data","data science","Go","unicorn","AWS","AI","Military","ViT","analytics","GAN","Data Science","R"],"extracted_tech_keywords":["AI","data science","analytics","AWS","R","Go","big data","GAN","ViT","unicorn"],"url":"https:\/\/analyticsindiamag.com\/it-services\/scope-data-science-military-intelligence-law-enforcement\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":39350,"title":"Can Contributory Databases In Indian Insurance Sector Deliver Greater Returns","content":"Image for representation purpose only Technology is the beacon of paradigm shifts across all walks of life, be it in businesses, economies or people’s day-to-day lives. Most industries are rapidly adapting technological processes to save costs and time and increase productivity, efficiency and ultimately profitability. The primary goal of all the stakeholders currently – insurers, the regulatory authority, the government and consumers – is to raise the penetration of life, health and motor insurance to enable policyholders to protect themselves from unforeseen circumstances. While the insurance regulatory authority has been at the forefront of creating a favourable environment for both insurance companies and consumers, there is much more that could be done. A recent report by Assocham reveals that insurance penetration in India has moved from 3.3% in 2014 to 3.44% in 2015, attributable to the various insurance schemes that the government launched. India’s insurance penetration as a whole in 2015 was 3.4%, against the world average of 6.2%. While adaptation to new technology has enabled the rapid growth of markets for a large number of industries, a key player in the Indian economy – the burgeoning insurance industry – is yet to fully leverage the varied uses of technology to effect faster growth. Although the insurance sector was privatised a few decades ago, insurance companies continue to hold bleeding portfolios, largely due to poor loss ratios, underwriting losses and claims frauds. Given the long history of insurance companies in the country, it is evident that they hold reams of data which, if digitized and analyzed appropriately, could help yield significant insights for effective premium pricing, underwriting of risk, curbing fraudulent claims, and ensuring a more personalized customer experience. A big way to stem these losses is through the adoption of advanced data analytics solutions. While policy makers are in the process of creating a balance between privacy and the need to allow data to be used for legitimate and beneficial purposes, India has witnessed the first step towards a Data Privacy regime. By defining individuals as data principals and processors as data fiduciaries, it has enabled the committee to enhance the autonomy of the individual and places a great degree of responsibility on the processor to maintain trust. In India, data analytics is a phenomenon that insurance companies are just starting to adopt.  Some of the basic advantages of data analytics include faster and better decision making to gain a competitive advantage with unique insights from proprietary data.  An example of these insights is a closer understanding of an individual customer’s lifestyle, from health factors to what kind of vehicle they drive, which can be highly predictive of insurance risk. Boosting these insights, would be an intelligence exchange platform which enables industry data to be shared amongst insurers (large and small), with data being pooled from all segments of the total population. While an insurance company gains basic insights from its own data, there are exponentially more benefits that can be derived through data sharing, if handled securely by a trusted third party.  A comprehensive database would provide deeper and more accurate insights into the lifecycle of the population set Data analytics applied to larger pools of data would also help insurance companies better understand their current and potential consumers, price premiums more accurately, and reduce the incidence of frauds. An example of a smart data intelligence exchange would be one that holds multiple consumer data points – including their current and past policies, health declarations, claims history, agent network, habits and preferences, all of which aid in making better underwriting decisions at the time of policy issuance, claims payouts and policy renewals. It is time that the Indian insurance companies consider data sharing and analytics as an easy way to leapfrog into the information age. Contributory databases are commonplace in more mature insurance sectors such as in the United States. The creation and usage of a contributory database calls for open and transparent sharing amongst all the players in the marketplace, supported by a trusted provider of data, analytics and customer insights. Insurers of all sizes benefit from having access to more information than if they are operating in isolation. With this backdrop, the Indian insurance segment must work towards creating a synergistic decision to pool in their data into such contributory databases.","excerpt":"Technology is the beacon of paradigm shifts across all walks of life, be it in businesses, economies or people’s day-to-day lives. Most industries are rapidly adapting technological processes to save costs and time and increase productivity, efficiency and ultimately profitability. The primary goal of all the stakeholders currently – insurers, the regulatory authority, the government […]","categories":["AI Features"],"tags":["Gen AI in Insurance","what is database"],"author_name":"Shivakumar Shankar","publish_date":"2019-05-18T10:20:05","publication_year":"2019","word_count":715,"keywords":["Go","API","AI","Git","RAG","Aim","ViT","analytics","Gen AI in Insurance","Rust","what is database","R"],"extracted_tech_keywords":["AI","analytics","Aim","RAG","R","Go","Rust","Git","API","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/can-contributory-databases-in-indian-insurance-sector-deliver-greater-returns\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":48583,"title":"NVIDIA&#8217;s Powerful 5G Pitch &#038; Other Announcements From Mobile World Congress","content":"This week at the Mobile World Congress held in Los Angeles, tech giant NVIDIA is making a strong bid towards pushing 5G networks to use its chipsets. NVIDIA is making a hard pitch at this year’s Mobile World Congress Los Angeles that the future of software-defined 5G networks should be powered by its chipsets. NVIDIA CEO Jensen Huang also said at the event that the smartphone revolution that has swept the globe over the past decade, was just the start. “The smartphone revolution is the first of what people will realize someday is the IoT revolution, where everything is intelligent, where everything is smart,” Huang said. He squarely positioned NVIDIA to power artificial intelligence at the edge of enterprise networks and in the virtual radio access networks – or vRANs – powering next-generation 5G wireless services. Among the dozens of leading companies joining NVIDIA as customers and partners cited during Huang’s 90-minute address were: WalMart (who is already building NVIDIA’s latest technologies into its showcase Intelligent Retail Lab)BMWEricssonMicrosoftNTTProcter & GambleRed HatSamsung Electronics “The smartphone moment for edge computing is here and a new type of computer has to be created to provision these applications… If the global economy can be made just a little more efficient with such pervasive technology, the opportunity can be measured in trillions of dollars per year.” – Jensen Huang, CEO, NVIDIA Red Hat and NVIDIA said that they are building on their announcement earlier this year that they are accelerating the adoption of Kubernetes in enterprise data centres. Their expanding collaboration lets customers use the NVIDIA EGX platform and Red Hat OpenShift to more easily deploy NVIDIA GPUs to accelerate AI, data science and machine learning at the edge. NVIDIA with Microsoft Corp announced a technology collaboration focused on intelligent edge computing, designed to help industries better manage and gain insights from the growing flood of data created by retail stores, warehouses, manufacturing facilities, connected buildings, urban infrastructure and other environments. By enabling closer integration between Microsoft Azure and the NVIDIA EGX platform, the companies are working together to advance edge-to-cloud AI computing capabilities, benefiting businesses worldwide. On the other hand, NVIDIA and Ericsson are collaborating on technologies that can allow telco operators to build high-performing, efficient and completely virtualized 5G radio access networks (RAN). These virtualized networks can enable faster and more flexible introduction of new AI and IoT services. The collaboration brings together Ericsson’s expertise in RAN technology with NVIDIA’s leadership in GPU-powered accelerated computing platforms, as well as AI and supercomputing. To demonstrate the possibilities, Huang punctuated his keynote with demos showing what AI can unleash in the world around us. In a flourish that stunned the crowd, Huang made a red McLaren Senna prototype — which carries a price of a hair under $1 million — materialize on stage in augmented reality. It could be viewed from any angle — including from the inside — on a smartphone streaming data over Verizon’s 5G network from a Verizon data center in Los Angeles The technology behind the demo: Autodesk VRED running in a virtual machine on a Quadro RTX 8000 server. On the phone: a 5G client build with NVIDIA’s CloudXR client application software development kit for mobile devices and head mounted displays. And, in a video, Huang showed how the Jarvis multi-modal AI was able to to follow queries from two different speakers conversing on different topics, the weather and restaurants, as they drove down the road – reacting to what the computer sees as well as what is said. In another video, Jarvis guided a shopper through a purchase in a real-world store. “In the future, these kinds of multi-modal AIs will make the conversation and the engagement you have with the AI much much better,” Huang said.","excerpt":"This week at the Mobile World Congress held in Los Angeles, tech giant NVIDIA is making a strong bid towards pushing 5G networks to use its chipsets. NVIDIA is making a hard pitch at this year’s Mobile World Congress Los Angeles that the future of software-defined 5G networks should be powered by its chipsets. NVIDIA […]","categories":["AI News"],"tags":["5G","what is power bi"],"author_name":"Prajakta Hebbar","publish_date":"2019-10-22T11:24:08","publication_year":"2019","word_count":628,"keywords":["data science","5G","artificial intelligence","what is power bi","machine learning","AI","Modal","R","cloud_platforms:Azure","edge computing","Azure","kubernetes"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","data science","Azure","kubernetes","edge computing","R","Modal","cloud_platforms:Azure"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/nvidias-powerful-5g-pitch-other-announcements-from-mobile-world-congress\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":34850,"title":"5 Data Science Job Openings At Target India You Can Apply Right Away","content":"Target is an international company with more than 323,000 team members around the world — with headquarters in the US. A Fortune 50 company with more than $69 billion in annual revenue, Target strives to fulfil the needs of millions of guests across the globe. Set up in 2005, Target India operates as a fully-integrated part of the global team. Located across two locations, Embassy Manyata Tech Park and Embassy Golf Links in Bengaluru, Target India is the extended headquarters of the company. Over 2,600 team members in the city are engaged in work that supports Target’s strategy across business areas like technology, marketing, human resources, finance, merchandising, supply chain, and analytics and reporting. Data engineering is one of the pivotal tasks at Target. Here is a list of five data science job opportunities at Target’s Bengaluru office that you can apply right away. Item Data Analyst An Item Data Analyst leads the operational strategy for the team and provides coaching on the day-to-day activities of the team and establishes the operational goals and measurements (i.e. LOS, productivity, data quality). Creates plans and daily routines for the Item Data Specialists to meet the goals and communicate performance expectations. Requirements 3-year degree or equivalent and a minimum of 4 years’ experience Data analysis & reporting skills Demonstrated expertise in an area of Merchandise Planning\/Operations, Buying, Digital Operations, Project Ability to work with and coordinate cross-functional efforts Analytical and critical thinking skills Exceptional communication and organizational skills Ability to develop effective partnerships Demonstrated ability to work on multiple projects at the same time, simultaneously managing a working team. Apply here Analyst BI and Analytics Target is offering an exciting opportunity for data wranglers, scientists and enthusiasts who want to analyse data, write algorithms and build models using large and diverse datasets. Requirements Framing business problems in terms of precise mathematical hypothesis Formulating statistical, machine learning or mathematical solutions to test and determine automated solutions to business problems Deploying algorithms and solutions using big data environment R\/Python\/Scala Knowledge in statistics\/ operations management\/ mathematics MTech or MS (Statistics, Economics) with relevant analytics experience of 6+ years BTech or BE (Engineering) with relevant analytics experience of 8+ years Domain experience: in solving business problems in the supply chain on logistics, labour management, automation, inventory management, network planning will be preferred. Apply here Senior Analyst BI and Analytics The Senior Analyst is responsible for framing business problems in terms of precise mathematical hypothesis and formulating statistical, machine learning or mathematical solutions to test and determine automated solutions to business problems. And, deploying algorithms and solutions using big data environment. Requirements Good programming skills in R\/ Python\/ Scala. Basic skills in SQL\/ Hive would be required Knowledge in statistics\/ operations management\/ mathematics Educational background: MTech or MS (Statistics, Economics) with relevant analytics experience of 6+ years PhD with relevant experience of 3+ years BTech or BE (Engineering) with relevant analytics experience of 8+ years Domain experience: in solving business problems in supply chain on logistics, labour management, automation, inventory management, network planning will be preferred. Apply here Lead Data Engineer As a Lead Data Engineer, you’ll have the opportunity to create software solutions using Agile practices and DevOps principles. Responsibilities will include designing, programming, debugging and supporting high quality, distributed, and large-scale software solutions on the latest Big Data tech stack.  Designing and building data pipelines using large data sets on game-changing services\/products in the innovative Enterprise Data Lab. Requirements Develop software systems using test-driven development employing CI\/CD practices Partner with other engineers and team members to develop software that meets business needs Follow Agile methodology for software development and technical documentation BS degree in Computer Science or relevant experience 5+ years’ experience in developing software applications 3+ years’ experience working on Big Data technologies like Hadoop, Spark, Scala and Hive Extensive understanding of application\/software development and design Proficiency in at least one of the following languages: Java, Scala, Python Experience with REST APIs’ and services preferred Experience\/knowledge of streaming solutions: Kafka, Apex Experience in No SQL database. Apply here Lead Data scientist A Lead Data Scientist will support all business areas of Target with critical data analysis that helps team members make profitable decisions and become a forecasting expert, business analyst or team lead while utilising tools like decision trees, clustering, regression, time series, structural equation modelling, linear programming, genetic algorithms, SAS, SQL, VBA and OLAP. Requirements Execute solutions to business problems using data analysis, data mining, optimization tools, and machine learning techniques and statistics Build data-science and technology-based algorithmic solutions to address business needs. Execute large scale models using Logistic Regression, Linear Models Family (Poisson models, Survival models, Hierarchical models, Naïve-Bayesian estimators), Conjoint Analysis, Spatial models, Time-series models, Text mining B.Tech (+8 years of relevant exp), M.Tech, M.Sc. (+6 years of relevant exp), PhD (+3 years of relevant experience) in Engineering, Operation Research, Mathematics, Engineering, Statistics preferred Demonstrated ability to work with ambiguous problem definitions, recognize dependencies and deliver impactful solutions through logical problem solving and technical ideations Ability to learn new analytical methods and technologies and apply in practical business problems Apply here","excerpt":"Target is an international company with more than 323,000 team members around the world — with headquarters in the US. A Fortune 50 company with more than $69 billion in annual revenue, Target strives to fulfil the needs of millions of guests across the globe. Set up in 2005, Target India operates as a fully-integrated […]","categories":["AI Hirings"],"tags":["Agile Methodology in Data Analytics","Btech","PhD","survival regression python"],"author_name":"Ram Sagar","publish_date":"2019-02-13T05:58:29","publication_year":"2019","word_count":849,"keywords":["data science","Go","machine learning","Btech","PhD","AI","Agile Methodology in Data Analytics","survival regression python","Python","analytics","Kafka","SQL","R","Java"],"extracted_tech_keywords":["AI","machine learning","data science","analytics","Kafka","Python","R","SQL","Go","Java"],"url":"https:\/\/analyticsindiamag.com\/ai-hiring\/5-data-science-job-openings-at-target-india-you-can-apply-right-away\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10012247,"title":"Top Third-party Initiatives To Ensure Algorithmic FATE For Enterprises","content":"Private enterprises use several algorithms for decision making on a day-to-day basis. Many of them have also realised or have been made aware of the consequences of using algorithms that might be biased towards individuals or sections of society. To avoid such incidents from occurring again, many tech companies are taking steps to ensure FATE – Fairness, Accountability, Transparency, and Explainability in their algorithms. Some of them have also developed tools not just for internal audits but also for others to buy them. While internal audits can help, they present a conflict of interest. An external audit or a third-party initiative, on the other hand, could potentially help ensure FATE for algorithms in its true sense. Several initiatives in the direction have been taken by many entrepreneurs and researchers. Auditing & Consultancy O’Neil Risk Consulting & Algorithmic Auditing (ORCAA) is a private consultancy firm that helps companies and organisations manage and audit algorithmic risk. The two main questions it answers are — ‘What does it mean for the algorithm to work?’ and ‘How could this algorithm fail and for whom?’, as the audit aims to incorporate and address concerns from all stakeholders in an algorithm. Along with providing audit services in terms of identifying issues of fairness, bias, and discrimination and recommending steps for remediation, the firm also provides legal expertise to public agencies and law firms in legal actions related to algorithmic discrimination and harm. It also gives talks and training on algorithmic auditing and fairness. Rating Services Similar to the transparency provided by the food industry by providing ingredients and nutrition labels on its products, Open Ethics Label helps companies label various aspects of their algorithms to achieve transparency. AI algorithm owners can voluntarily provide three main aspects of the code: training data, algorithms and decision spaces and get them labelled. Providing this information can also help these organisations understand the inherent biases in their data, evaluate security and privacy risks associated with their code, and understand the robustness and safety of the decision spaces in their algorithms. Along with that, Open Ethics also provides a vector rating called Open Ethics Vector, which reflects the values on which data-driven decisions are made. These are based on a set of principles which consider parameters like values, guiding behaviours and attitudes towards religion, gender, relationships, money, food, or health. The vector provides transparency in terms of the information on which the ethical choices were made. The end-users can test the application to get a vector rating by selecting personal ethical preferences. This helps users decide which apps are best for them. Open-sourcing Audit Algorithms Pymetrics, a predictive analytics firm, feared they might be using a biased algorithm for hiring. Hence they started ‘Audit AI’ to audit its own algorithm, and the resulting tool has been made publicly available to encourage other people to use it. Available to download from Github, it helps to measure and mitigate biases that might have been introduced due to aspects like training data, that unfavourably discriminate against underrepresented people in the dataset. Initiatives To Encourage Inclusivity One of the main reasons why AI algorithms end up being biased in the US is because they are predominantly developed by one demographic, i.e. male whites. Hence, with an aim to undertake initiatives that will increase the presence of black people in the field of AI, Rediet Abebe and Timnit Gehru, both computer science academics, started a small community called Black in AI. Initially set out as a small group of individuals, it has now more than 1,200 members on its Facebook group. Similarly, another initiative, AI4ALL aims to undertake summer programs at prestigious universities for underrepresented groups and expose them to the possibilities of AI. The organisation believes that ‘when people of all identities and backgrounds work together to build AI systems, the results better reflect society at large.’ Research In AI Algorithms AlgorithmWatch is a non-profit research and advocacy organisation that is supported by organisations like Bertelsmann Stiftung. They conduct in-depth research and try to shed light on ‘algorithms that have social relevance, meaning they are used either to predict or prescribe human action or to make decisions automatically.’ They publish several stories or studies that will have an in-depth description of automated systems and their social and technical impact. The algorithms could be owned by public organisations or private enterprises. Wrapping Up Third-party auditing can help remove the conflict of interest if firms want to bring transparency and explainability to their algorithms or ensure their algorithms are fair and ethical in its true sense. However, private enterprises are not legally binding to get their algorithms audited or certified for a certain level of fairness or transparency or ensure a certain threshold of diversity in their teams. Hence, incentives to ensure the FATE of algorithms can help boost the number of enterprises will get their algorithms audited or encourage them to ensure inclusivity in their AI teams. This will encourage others to follow, as well.","excerpt":"Private enterprises use several algorithms for decision making on a day-to-day basis. Many of them have also realised or have been made aware of the consequences of using algorithms that might be biased towards individuals or sections of society. To avoid such incidents from occurring again, many tech companies are taking steps to ensure FATE […]","categories":["AI Trends"],"tags":["enterprise analytic hub","Ethical AI","explainability","Explainable AI","Fair AI","FAIRNESS","Inclusive AI","Transparency","Transparency AI"],"author_name":"Kashyap Raibagi","publish_date":"2020-11-24T15:00:00","publication_year":"2020","word_count":829,"keywords":["explainability","Git","BERT","R","enterprise analytic hub","Inclusive AI","RAG","FAIRNESS","analytics","Go","AI","Ethical AI","Fair AI","Transparency AI","Aim","Transparency","Explainable AI","GitHub","predictive analytics"],"extracted_tech_keywords":["AI","analytics","Aim","RAG","predictive analytics","R","Go","Git","GitHub","BERT"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-third-party-initiatives-to-ensure-algorithmic-fate-for-enterprises\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":13588,"title":"Analytics India Salary Study 2017 – by AnalytixLabs &#038; AIM","content":"With data becoming increasingly central to businesses, businesses are betting big in zeroing down on the right talent. Lucre aside, analytics has become central to running enterprises efficiently. Amidst the sheer volume and the variety of data generated, what organisations are looking at is a well-structured data and findings in a consumable format that makes some sense. Organizations want to hire professionals who can not only slice and dice, but communicate their findings effectively to senior leadership as well. This is where the all-time high demand for analytics professionals comes in. Analytics India Magazine in association with AnalytixLabs presents the annual, comprehensive Analytics India Salary Study 2017 that gives peek into the various aspect of salary structure for every professional who wants to makes a head-start in this industry. Download the report to get to know the in-depth salary trends across various sectors, skills, tools and domains. Read last year’s report here Key Trends- The average analytics salary in India for year ending 2016 was INR 11.7 Lacs across all experience level and skill sets. This is the highest average salary for analytics professionals ever, with almost 22% increase since a year ago. Average analytics salaries in 2016 were 9.5L. Though overall the number of analytics professionals increased last year, the percentage of professionals with salary in 0-6L band decreased from 42% to 39%. There has been visibly more demand for senior professionals last year, thus pushing the average salaries higher. On one hand, the percentage of Analytics professionals commanding salaries less than INR 10 Lacs has gone; % of Analytics professionals earning more than INR 15 Lacs has increased from 33% in 2016 to 37% in 2017. Analytics professionals with salaries in extreme bracket of 50+L have increased significantly this year – to 3.7% from just 1.1% a year ago. Analytics Salaries vs other Domains- On average, analytics professionals receive around 30% higher salaries than IT professionals in India. While, the average salaries of Analytics professionals stand at 11.7L, that of IT professionals is at 8.65L, of professionals in Primary Research is 9.03L, Financial research at 8.96L and that of Secondary research is 7.83L. Just 39% of analytics professionals have salaries under 6Lakhs vs 58% in IT. 28% analytics professionals in India earn in the salary bracket of 10-25L. This number for IT professionals is 17%. Salary Trends across Tools & Skills- Advanced analytics\/predictive modeling professionals are paid the highest salaries compared to their analytics peers i.e. 14.7 Lacs on average. Big data professionals and data mining professionals get up to 13 Lacs & 9.9 lacs on average. BI, Reporting , MIS professionals get upto 8.3 Lacs on average and data engineer an average of 11.5 Lacs. Within Big data profiles: Hadoop professionals command an average of 12 Lacs. The salaries increased by around 21% since a year ago of 9.9L. Spark professionals command an average of 10.1 Lacs with an increase of almost 15% from a year ago of 8.8L. NoSql\/ MangoDB professionals receive an average of 11.5 Lacs, a slight increase from 11.1L since a year ago. Qlikview professionals command marginally higher than Tableau professionals in visualization skill jobs – 12.1 Lacs vs 11.7 Lacs on an average. Visualization tools salaries have increased around 28% since a year ago. Within Advanced Analytics profiles: R programmers receive the highest average salaries this year in adavcned analytics tools skillset – at 11.1L. R have defeated SAS, SPSS and Python in terms of average salaries this year. On an average, professionals get up to 10.3 Lacs for SAS, vs 8.7 Lacs for SPSS professionals. SAS salaries have increased by just 13% since a year ago. Python gets a higher pay of 10.5 Lacs compared to SAS and SPSS. Yet, Python is the only skill set for which the average salaries have decreased since a year ago – by almost 7%. Salary Trends across Cities- Average salaries across all cities have increased by more than 20% for analytics professionals this year. Pune saw the highest increase in analytics salaries by almost 30% – from 8.8L to 11.4L this year. Mumbai retails its spot for highest salaries in Analytics at almost 14L per annum – an increase of 22% from a year ago. In the 0-6 Lacs bracket, Hyderabad dominates other cities with 42% of analytics professionals earning below 6 Lacs followed closely by Pune at 39%. Mumbai leads the pack in more than 15L salary bracket, with 28% earning within this bracket, followed closely by Bangalore (25%) and Delhi\/NCR(22%). Salary trends across Experience levels- Analytics salaries saw the biggest jump at the entry level (0-3Yrs experience) last year. There was almost a 61% jump in salaries in this experience bracket – from 4.1L to 6.6L average. At entry level, almost 76% of analytics professionals are under the 0-6 Lacs salary bracket. For other experience brackets, the salary increases have been relatively lower, especially at higher experience bracket – at 12+ years experience, the salary rises in analytics have just been 4% last year. A transition to senior analytics leaders, with more than 12 years of experience, can lead to almost 50% increase in salaries. Salary trends across Company Type- This year, Captive Analytics centers tend to be highest salary providers in analytics, at an average of 14.1L. They were followed by Consulting firms at 13.9L and Domestic firms at 13.6L analytics salaries on average. However, Boutique analytics firms continue to pay the lowest – an average of 9.5L, albeit a 20% hike from last year of 7.9 Lacs. Large IT service providers pay 11.4 Lacs in analytics, an increase of 20% from 9.5L. Domestic Indian firms have been the slowest in increasing the analytics salaries last year. They currently pay an average of 13.6L, an increase of just 13% to their analytics\/ data science teams. Conclusion- From large enterprises, captive analytics businesses, boutique analytics or consulting firms, the demand for quality data analysts is met with a lucrative pay.  With data analytics driving all the component of an industry, the companies are open to offer more than decent pays to these rare breed of professionals. The compensations have seen a positive trajectory and have gone north in where skills are backed by experience. If you are trying to break into this field, now is the best time, especially when data analytics professionals snap 30% higher salaries than IT professionals in India And like previous year, Hadoop, R, Spark and Python remain the most sought after tools by the employers, with a better pay package to those with advanced analytics\/predictive modeling skills compared to data mining and BI tools. Overall, the study, a great analytics find is a nudge in the right direction and will fuel hopes of making a career in analytics domain. Here’s the complete report Download it here [attachments include=”13616″]","excerpt":"With data becoming increasingly central to businesses, businesses are betting big in zeroing down on the right talent. Lucre aside, analytics has become central to running enterprises efficiently. Amidst the sheer volume and the variety of data generated, what organisations are looking at is a well-structured data and findings in a consumable format that makes […]","categories":["AI Features"],"tags":["analytics career"],"author_name":"Дарья","publish_date":"2017-03-20T11:08:18","publication_year":"2017","word_count":1131,"keywords":["big data","data science","Go","AI","RAG","Python","analytics career","analytics","SQL","GAN","R"],"extracted_tech_keywords":["AI","data science","analytics","RAG","Python","R","SQL","Go","big data","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/analytics-india-salary-study-2017-by-analytixlabs-aim\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":69793,"title":"Hero Electronix, part of Hero Group sets foot in IoT and defence electronics","content":"With a focus on IoT enabled home solutions, Hero Electronix is eyeing to venture into areas like Internet of Things, automotive sector and defence electronics. Nikhil Rajpal, the company CEO believes that there are huge opportunities in India as more technologies are entering Indian households. With opportunities in areas like home automation devices, smart lighting, trackers and others, the company sees it as a right time to tap into Indian market. “We have prototypes ready and in another six to nine months we should be ready with our first launch in the segment”, he was reportedly quoted as saying. While the products that the company intends to launch are still not disclosed, it aims to tailor made the products suiting the requirements of the Indian market. Amongst the needs, the company aims to bring low priced products based on local needs. Rajpal said that their focus has been on the pricing front to make affordable products. Also, the company is aiming defence sector as a huge part of their spending is on electronics. They intend to bring local manufacturing so as to cut down on huge spending. It wants to tap the government’s Make in India initiative to dive into the space of defence and automotive sector. Recently a lot of companies had shown interest in Indian IoT market, with the likes of Coriant, LG, Samsung, Bajaj Electricals venturing into the IoT space. Recently BSNL signed MoU with Coriant to plan India’s IoT and 5G Roadmap","excerpt":"With a focus on IoT enabled home solutions, Hero Electronix is eyeing to venture into areas like Internet of Things, automotive sector and defence electronics. Nikhil Rajpal, the company CEO believes that there are huge opportunities in India as more technologies are entering Indian households. With opportunities in areas like home automation devices, smart lighting, […]","categories":["AI News"],"tags":["IoT India"],"author_name":"Srishti Deoras","publish_date":"2017-09-08T13:40:56","publication_year":"2017","word_count":246,"keywords":["Go","programming_languages:R","AI","IoT India","programming_languages:Go","automation","Aim","R"],"extracted_tech_keywords":["AI","Aim","R","Go","automation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/hero-electronix-part-hero-group-sets-foot-iot-defence-electronics\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":19483,"title":"Are Capsule Networks The New Building Blocks In Deep Learning?","content":"Deep Learning takes on a new direction with every research and this time Geoff Hinton has revealed a better and more reasonable architecture than Convolutional Neural Networks (ConvNets) – capsule network. The CapsNet is a new type of neural network architecture conceptualized by Hinton, with the idea to address some of the short comings of CNN. While it is an important and interesting research in Deep Learning, does it truly address the shortcomings of CNN — which is the workhorse of today’s deep learning? According to the paper, Dynamic Routing Between Capsules: “A capsule is a group of neurons whose activity vector represents the instantiation parameters of a specific type of entity such as an object or object part.” Are Capsules An Improvement on CNNs –  The Workhorses of Deep Learning One of the major advantages of Convolutional Neural Networks is their invariance to translation. However, this invariance comes with a price and that is, it does not consider how different features are related to each other. For example, if there is a picture of a face, CNN will have difficulties distinguishing relationship between mouth feature and nose features. Max pooling layers are the main reason for this effect. Because when we use max pooling layers, we lose the precise locations of the mouth and noise and we cannot say how they are related to each other. Hinton himself stated that the fact that max pooling is working so well is a big mistake and a disaster. On the other hand, capsule takes into account the existence of the specific feature that one is looking for like mouth or nose specifically. And this feature makes sure that capsules are translation invariant the same that CNNs are. So, capsules try to solve the drawback of max pooling layers through equivariance. Which means, instead of making the feature translation invariance, capsule make it viewpoint-equivariant, noted a researcher. So as the feature moves and changes its position in the image, feature vector representation will also change in the same way which makes it equivariant. Hence, the major drawback of ConvNets is that it fails when images are transformed too. According to Max Pechyonkin, a deep learning enthusiast, CNN do not have this built-in understanding of 3D space, but for a CapsNet it is much easier because these relationships are explicitly modeled. The paper that uses this approach was able to reduce the error rate by 45% as compared to the previous state of the art, which is a huge improvement, he noted. Here’s Hinton’s take on CNN and the simple mechanism from the paper in OpenReview — a vision system needs to use the same knowledge at all locations in the image but viewpoint changes complicate the effects on pixel intensities. The aim of capsules is to make good use of this underlying linearity, both for dealing with viewpoint variation and improving segmentation decisions. Capsule Networks vs CNN According to Danish developer, entrepreneur and mathematician Martin Jul, one of the key difference of capsules from CNN is that it is able to provide better generalization from limited training data. The first idea in CapsNet is to represent multi-dimensional entities. Capsule Networks does this by grouping these properties of a feature together The second is that one activates higher-level features by agreement between lower-level features (routing by agreement) The Capsule Networks partitions the image into regions subsets and for each of these regions, it assumes that there is at most one instance of a single feature, called a Capsule A Capsule is able to represent an instance of a feature (but only one) and is able to represent all the different properties of that feature, e.g., its (x,y) coordinates, its colour, its movement etc. The difference from Convolutional Neural Networks (CNNs) is that the Capsules bundle the neurons into groups with multi-dimensional properties, whereas in CNNs the neurons represent single, unrelated scalar properties. Are Capsules The New Building Blocks In Deep Learning According to a computer vision researcher Tobias Würfl, capsule net addresses the most obvious problem where Deep Learning falls short. “We really need to address the issue that we don’t encode our prior knowledge about viewpoint invariance into the networks. And the other issue is to automate architecture design in a fast and reliable,” he stated in a post. Hinton had been researching on Capsule networks for a long time and he first proposed it in 2011, debunking backpropagation, the trusted method pioneered by him for neural networks back in the 1980s. Backpropagation met with huge success in the machine learning community and is regarded as path to achieving Artificial General Intelligence. It is a method for optimizing machine learning algorithms by adjusting weights so the algorithms can be perfected with the fewest possible errors. Now, this new type of neural network architecture conceptualized by Godfather of “Deep Learning” comes closest to the observations of the human brain on the machine. And Hinton is placing all bets on this architecture to revolutionize computer vision. “Human vision ignores irrelevant details by using a carefully determined sequence of fixation points to ensure that only a tiny fraction of the optic array is ever processed at the highest resolution. In this paper we will assume that a single fixation gives us much more than just a single identified object and its properties,” he said. By and large, Hinton proposes that the idea stemmed from the fact that neural networks needed better modeling of the spatial relationships of the parts. Instead of modeling the co-existence, disregarding the relative positioning, capsule-nets try to model the global relative transformations of different sub-parts along a hierarchy. This is the eqivariance vs. invariance trade-off, explained AI researchers. Capsnet – A Leap Forward In Deep Learning AI researchers have pegged this as the most important and interesting research directions in Deep Learning right now. But the best implementation of this architecture is yet to emerge. Here’s why the DL community believes it is a big breakthrough because the new architecture achieved a significantly better accuracy on the small NORB data set then the state-of-the-art CNN, reducing the number of errors by 45%. The OpenReview paper cites it to be significantly more robust to white box adversarial attacks than a baseline CNN. According to Hinton, the current implementation of capsule networks is slow but over a period of time, it can spark a major leap forward in AI.","excerpt":"Deep Learning takes on a new direction with every research and this time Geoff Hinton has revealed a better and more reasonable architecture than Convolutional Neural Networks (ConvNets) – capsule network. The CapsNet is a new type of neural network architecture conceptualized by Hinton, with the idea to address some of the short comings of […]","categories":["IT Services"],"tags":[],"author_name":"Richa Bhatia","publish_date":"2017-12-05T05:18:59","publication_year":"2017","word_count":1066,"keywords":["Go","machine learning","AI","neural network","computer vision","Ray","Aim","deep learning","Rust","R"],"extracted_tech_keywords":["AI","machine learning","deep learning","neural network","computer vision","Aim","Ray","R","Go","Rust"],"url":"https:\/\/analyticsindiamag.com\/it-services\/capsule-network-new-building-blocks-deep-learning\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10018173,"title":"What Is Robustness Gym?","content":"NLP systems work like a well-oiled machine on standard benchmarks. However, their performance is not on par when deployed in real-world systems. A recent study highlighted that 60% of NLP answers are similar to their training data, which implies that the models are simply memorising their training set. Many researchers are working on testing the robustness of these NLP models, on a diverse range of challenges such as adversarial attacks and rule-based data transformations. In a recent paper, Salesforce, in association with Stanford University, has introduced a new framework — Robustness Gym, that will work as a robustness evaluation toolkit for NLP models. The paper touches on the key challenges associated with assessing NLP systems, such as the paradox of choice, idiomatic lock-in, and workflow fragmentation; and how Robustness Gym enables practitioners to compare results with just a few clicks. Challenges faced by the practitioners while evaluating the NLP models. “We embed Robustness Gym in a new paradigm for continually evaluating models with the — ‘contemplate, create and consolidate’ evaluation loop,” said the authors. Explaining the process, the researchers stated that during the ‘contemplate’ stage, the Robustness Gym (RG) guides the practitioners on what evaluation to run next, whereas, in the ‘create’ stage the RG slices the data defining the collection of samples for evaluation. However, on the ‘consolidate’ stage, it provides findings for faster iteration and community sharing. How Does It Work? Robustness Gym assists practitioners on how variables like the structure of the task, evaluation requirements like a testing generalisation, biases, or security, and constraint of resources can aid in evaluation tasks. Further, the framework supports evaluation on four standard paradigms: subpopulations, transformations, evaluation sets, and adversarial attacks. The RG users can also consolidate the slices into a TestBench, allowing them to collaboratively build benchmarks and track progress. To demonstrate it with a real case study, the researchers outlined three users’ varying expertise while evaluating a natural language inference model using Robustness Gym. The experiment showed that while the novice users relied on predefined test benches for direct evaluation, intermediate users were creating new slices using SliceBuilders available in the framework, and then constructing their own test benches. Finally, the advanced users were able to use their expertise to add custom slices. “All of these users can generate a shareable Robustness Report,” added researchers. Researchers further validated the Triple-C process using a 3-hour use case with Salesforce’s commercial sentiment modelling team, where the goal was to measure their model’s bias. Researchers wrote, “We tested their system on 172 slices spanning three evaluation idioms, finding performance degradation on 12 slices of up to 18% (create).” Finally, the researchers generated a single testbench and robustness report for the team, that summarises the findings. The team highlighted that RG was easy to use, and is planning to integrate it into their workflow. Robustness Report The paper also noted that Robustness Gym could be used to conduct new research analyses with ease. To validate it, the team performed a study with a named entity linking (NEL) system and critical analysis of summarisation models. They compared the commercial APIs from Microsoft, Google and Amazon to open-source systems Bootleg, WAT and REL across two benchmark datasets: Wikipedia and AIDA-CoNLL. The result showed that commercial systems, mentioned above, struggled to link rare entities and lag their academic counterparts by 10%+, while the summarisation models struggled on examples that require abstraction and distillation, degrading by 9%+. However, Microsoft outperforms other commercial systems, while Bootleg displays the most consistent performance across various slices. Wrapping Up According to researchers, the Robustness Gym has been developed as an evaluation toolkit for NLP models that supports a broad set of evaluation idioms. It can be used for collaboratively building and sharing evaluations and results. To address practitioners’ challenges today, the team has embedded the framework — Robustness Gym into the Triple-C evaluation loop. “Our results suggest that Robustness Gym is a promising tool for researchers and practitioners,” concluded the researchers. Read the entire paper here.","excerpt":"NLP systems work like a well-oiled machine on standard benchmarks. However, their performance is not on par when deployed in real-world systems. A recent study highlighted that 60% of NLP answers are similar to their training data, which implies that the models are simply memorising their training set. Many researchers are working on testing the […]","categories":["Deep Tech"],"tags":[],"author_name":"Sejuti Das","publish_date":"2021-01-17T10:00:00","publication_year":"2021","word_count":665,"keywords":["Go","API","programming_languages:R","AI","RPA","programming_languages:Go","RAG","NLP","R","adversarial attacks"],"extracted_tech_keywords":["AI","NLP","RAG","R","Go","API","adversarial attacks","RPA","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/what-is-robustness-gym\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10052996,"title":"How Young Is Too Young: AI For Children","content":"The future where our cars will drive themselves or robots will conduct surgeries is closer than ever before. The age of AI has arrived and is impacting and changing the world in unpredictable ways. It is important to believe that while we are slowly getting accustomed to AI, the children now will be dependent on it. Children are not just using AI and learning to interact with it, but already making it as well. In 2018, Tanmay Bakshi (then a 14-year-old) became popular as the little AI wonder who works with Google and IBM. He became the face of AI learning for children in India. In 2019, when Siddharth Srivastav Pilli, a class 7 boy, bagged a job as a data scientist with a software company in Hyderabad, AI for Kids became popular. Parents got really interested in teaching their children AI. India’s new National Education Policy 2020 did mention the importance of emerging technologies and introduced them to children. Making AI Fun for Children AI as a subject is still far from being introduced in every school and in every education board’s syllabus in India. But, many edtech companies have introduced AI courses for children as young as 8-year-olds. “It’s a completely different ballgame when it comes to teaching kids artificial intelligence as compared to college students. The concept of AI itself is Greek and Latin to them, so the way the classes are delivered and the curriculum is designed is very different as compared to college students,” said Prabhakar Nadar, CEO, Edurific. He also said that to teach children, they have to follow more practical applications without exposing them to the theory. “Kids get bored easily. We have to frame the classes so that they have an achievement-driven approach, or they can make something at the end of every few classes,” Nadar added. They start their AI classes by introducing children to basic AI applications like chatbots and then gamify the entire image recognition and other complex AI concepts, then move on to building basic algorithms and applications of AI. Do Children Really Need to Learn AI? While the edtech companies have figured out ways to teach children the technology, it surely remains a question whether children really need to learn AI. According to Dhrupal Shah, CEO, STEMpedia, “AI is very much engagement for kids. We don’t want to make them AI engineers or researchers. We want them to get fascinated and exposed to the new technologies growing around them like voice assistants (Ok, Google, Alexa), smart cameras, chatbots, etc. The curriculum designed for younger kids on AI is curated to give them application-based use cases, and it usually includes less theory or the real maths behind.” Experts believe that children must know how AI works, not just from a development point of view but also from a consumer point of view. “They should be able to understand the dynamics as it’s going to be a major part of their life with the current development in the field,” said Nadar. When it comes to mental development too, AI-based activities can help kids keep higher focus as compared to just computer coding. “There are much more real-world logic-based things while making AI-based projects rather than just fictional stuff,” echoed Shah. Wrapping Up AI has been evolving since 1951. In fact, a lot of working applications of AI were actually written in 1951 and have been evolving since then. We are exposed to AI from phone cameras to light bulbs, making the technology very relevant for many years from now. And if children are learning AI, we should also not underestimate them. “Our students surprise me every time I have a conversation with them. They are way ahead in terms of development and using technology. Our students who are as young as 9 years are able to develop automation and base level AI applications ranging from computer automation and NLP, NLU models and image recognition models,” concluded Nadar.","excerpt":"Exposing children to AI helps them build logical reasoning skills and helps them understand the ways in which they can utilise the digital tools with higher awareness and smartness.","categories":["IT Services"],"tags":["AI India"],"author_name":"Meeta Ramnani","publish_date":"2021-11-06T17:00:00","publication_year":"2021","word_count":658,"keywords":["Go","artificial intelligence","programming_languages:R","AI","chatbots","image recognition","automation","NLP","ViT","AI India","R"],"extracted_tech_keywords":["AI","artificial intelligence","NLP","chatbots","image recognition","R","Go","ViT","automation","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/it-services\/children\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10000789,"title":"Meet The Top 5 String Theorists From India Who Won Global Recognition","content":"Here are some of India’s best String theorists that have a global recognition. 1.Ashok Sen Working as a theoretical physicist in the prestigious Harish-Chandra Research Institute (HRI), Allahabad, Ashok Sen is a prominent String physicist in the country who has done his PhD in Physics from the Stony Brook University, and made notable contributions in the field. He was among the earliest persons to realise the importance of dualities in string theory during the second String Revolution. In 1998, he won the fellowship of the Royal Society on being nominated by the theoretical physicist Stephen Hawking. Notable researches: One of his most notable research was on S-duality, which is a strong-weak coupling duality, which was influential in changing the course of research in the field. His description of rolling tachyons has been influential in string cosmology. 2.Shiraz Minwalla Popular among his students for his overly enthusiastic nature to deliver lectures in the field, Shiraz Minwalla is a fellow string theorist in the department of Theoretical Physics at Tata Institute of Fundamental Research (TIFR), Mumbai. Dr Minwalla has done his PhD from Princeton University under the guidance of string theorist Nathan Seiberg. He was a Harvard Junior Fellow and subsequently an Assistant Professor at Harvard University. He had received the Shanti Swarup Bhatnagar Prize for Science and Technology, the highest science award in India, in the physical sciences category in 2011.  In 2016, The World Academy of Sciences awarded him the TWAS Prize in Physics. Notable researches: He is particularly known for his work on the connection between the equations of fluid dynamics and Albert Einstein’s equations of general relativity. He is also known for his work in OM-theory. 3.Spenta R Wadia Spenta R Wadia is the Infosys Homi Bhabha Chair Professor, Founding Director and Professor Emeritus at the International Centre for Theoretical Sciences (ICTS). He obtained his Ph.D from City College of City University of New York in 1978. He then worked at University of Chicago from 1978-82 first as a research fellow and then as a staff scientist. He has also written books on Physics. One of them on String Theory is called Strings 2001. This multi-authored book summarises the latest results across all areas of string theory from the perspective of world-renowned experts, including Michael Green, David Gross, Stephen Hawking, John Schwarz, Edward Witten and others. Notable researches: He is well known for his contributions to string theory, especially for the black hole solution in two dimensional string theory and various other work relating to the black holes in the context of string theory. He has published around 80 papers in reputable scientific journals. 4.Rajesh Gopakumar Rajesh Gopakumar is a String theorist and the director at the ICTS. He was previously a professor at Harish-Chandra Research Institute (HRI). He obtained his PhD from Princeton University under the guidance of the 2004 Nobel Prize winner, Dr. David Gross. He was a research associate at Harvard University, before returning to India. Notable researches: He is well-known for his work on topological string theory and for proposing the Gopakumar–Vafa duality and Gopakumar–Vafa invariants. In recent years, he has made important contributions to higher spin theories and their connections with string theory. He has also recently worked on the Conformal bootstrap. 5.Jnanadeva Maharana Jnanadeva Maharana is a former researcher at the Institute of Physics, Bhubaneswar. He is a Raja Ramanna Fellow, Department of Atomic Energy and has his PhD from IIT Kanpur. He has been working in String Theory for more than 20 years. He has many visiting positions internationally, including University of Tokyo and CERN. He was awarded the prestigious awards of Satyendra Nath Bose Medal in 2015, which even Ashok Sen had received earlier. Notable research: His main focus has been investigating symmetries of String theories. He has also worked on string cosmology from various perspectives.","excerpt":"Here are some of India’s best String theorists that have a global recognition. 1.Ashok Sen Working as a theoretical physicist in the prestigious Harish-Chandra Research Institute (HRI), Allahabad, Ashok Sen is a prominent String physicist in the country who has done his PhD in Physics from the Stony Brook University, and made notable contributions in […]","categories":["AI Highlights"],"tags":[],"author_name":"Disha Misal","publish_date":"2019-01-02T20:44:44","publication_year":"2019","word_count":636,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","BERT","llm_models:BERT","ViT","R"],"extracted_tech_keywords":["AI","R","Go","BERT","ViT","llm_models:BERT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/meet-the-top-5-string-theorists-from-india-who-won-global-recognition\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":23683,"title":"10 Indian Startups That Are Leading The AI Race: 2018","content":"Here is the fresh list of 10 startups in artificial intelligence and machine learning that are leading the race in the year 2018. These startups come from different backgrounds, such as providing AI services to healthcare, banking, fintech, diagnostics and HR, among others. Over the years startups have been striving to bring unique propositions into the market, and the developments listed by these companies prove it. While there are over 300 startups in the space in India alone, we have picked up these 10 names for this year’s list. These startups are about two to three years old and have shown significant development over the years and are likely to do so in the year 2018. NOTE: These startups are listed in alphabetical order. Read our last year’s list here. Artivatic Data Labs Artivatic Data Labs Team Founded in September 2016, Artivatic Data Labs is an end-to-end AI infrastructure platform that is built on deep-tech, ML technologies with in-depth analogy of genomic science, psychology and neuroscience. It helps large enterprises, startups and developers to build and integrate intelligent products and solutions without any developmental efforts. It is enabling industries such as banking, insurance, lending, wealth management, manufacturing, healthcare, robotics to have ‘Smart AI Brain’ to make processes automated and allow self decision in real time. With over 25 in-house built products, which are currently utilised to build more than 10 industry solutions, the startup already has one patent filed in India and the US, and more than five patents already in the process. The Founding Team: Founded by Layak Singh and Puneet Tandon, the idea behind starting the company was to enable the enterprises to have human brain-like intelligent system, which has the power to make self decisions without any human intervention. They believed that decisions which are taken by humans using traditional analytics costs time, money and are often error-prone. With a team of over 30, they are trying to achieve the perfect AI brain. They have been funded with a total amount of $350,000 till now. [su_divider top=”no” size=”1″] Avaamo Co-Founders, Avaamo Avaamo is a deep-learning software company that specialises in conversational AI interfaces to solve specific, high impact problems in the enterprise. Avaamo is building fundamental AI technology across a broad area of neural networks, speech synthesis and deep learning to make conversational computing for the enterprise a reality. Its proprietary AI algorithms combined with Deep Domain ML models learn and perform multi-turn conversations and execute judgment intensive tasks just like humans. It provides a cognitive technology driven platform that dramatically simplifies the time needed to design and deploy enterprise bots or virtual assistants to corporate employees and their customers. At the core of Avaamo’s platform is a secure messaging infrastructure that delivers scalability, reliability, security, and integration to legacy enterprise workflows and applications. Avaamo’s technology can be embedded into any platform by entering a few lines of Javascript in existing mobile access systems — a quantum jump from the traditional model of building an app. The Founding Team: It was founded by Sriram Chakravarthy (CTO) and Ram Menon (CEO). Avaamo’s conversational AI platform is currently being deployed globally across a diverse set of financial institutions, telcos, retailers and mutual funds including few of the oldest banks in US, Europe, and one of the biggest insurance firms in Asia, along with few big telecom service providers across the world. [su_divider top=”no” size=”1″] Bash.ai Bash AI Team Since its inception in March 2017, Bash.ai has been automating HR processes and thereby helping in better and consistent employee experiences. It is essentially a tool that uses AI and big data to power virtual assistants and drive HR for businesses, by mimicking cognitive functions related to HR. Its user-friendly front end interface gives organisations a highly accurate and real time way to automate conversations with employees within an organisation. It can be accessed using instant messenger platform such as Facebook Messenger, Slack, Skype, and currently has modules such as post hire orientation, ticketing, HR helpdesk, employee engagement, organise HR activities and answer questions related to payslips, company policies, among others. They want to go towards an ecosystem where AI and humans coexist efficiently in the HR settings. The Founding Team: It was founded by Barkha Sharma, who has spent over eight years in HR. Having started with a team of three in the founder’s living room, it is now 12 people strong with engineers, product developers, and operations enablers working with HR processes of organisations. The founder believed that in a world of automation the functions of HR are a bit outdated, and her idea was to fill this technological gap. They raised first round of funding in October 2017. [su_divider top=”no” size=”1″] CustomerSuccessBox CustomerSuccessBox Team An a B2B SaaS customer success platform backed with $1 million in venture funding, CustomerSuccessBox is aimed at delivering proactive customer success. As the startup believes, customer churn is the biggest blocker of growth and it’s like trying to fill up a leaky bucket. Such businesses cannot continue to operate with the old reactive support model, this is where CSB comes into picture. As customer success is greatly enhanced by using AI and ML to understand customer on-boarding, product adoption and usage analysis, and calculating customer health score, CSB uses it extensively to facilitate the same. They processes millions of activities from 1,00,000 plus end users every week and its patent pending technology is built to handle over 2,200 events per second to monitor customer account health. It is being used by global clients like XebiaLabs, WizIQ, Synup and WoowUp to deliver consistent on-boarding experience. The Founding Team: It was founded by Puneet Kataria (CEO) and Amritpal Singh (CTO) and they have raised a pre Series A of $1 million led by pi Ventures with participation from Axilor Ventures. While Puneet was working as a VP of Sales at a SaaS company, he learnt that the secret of growth for a SaaS business is to keep and grow recurring revenue. This is when CSB was born, which is not just a technology platform but a way of running SaaS business. The startup aims to double its employee strength in the next couple of months. [su_divider top=”no” size=”1″] Discovery AI Co-Founders, Discovery AI Founded towards the end of 2016, DiscoveryAI Platform enables conversational AI led Customer Success for any enterprise application. The founders believe that context-sensitive customer success is nearly missing in all comprehensive websites and apps used within enterprises. Businesses invest a lot of resources in creating and deploying these digital tools for various customer and employee facing processes. Discovery AI attempts to solve these problems are intrusive and raw in nature. They believe that availability of data and AI can fundamentally change the way we interact with the business systems. The Founding Team: It is founded by Prasanna, Sameer Acharya and Shubham Deshmukh.  The trio were always intrigued by the fact that enterprise systems are not able to keep pace with the consumer-facing technologies and seamless user experience had to be brought. Currently a bootstrapped startup, they are looking to raise formal seed round in the next three to six months time. [su_divider top=”no” size=”1″] My Ally Co-Founders, My Ally Originally named Skedool.it, My Ally aims to help the HR department be more strategic than tactical. When dealing with people it is critical to reduce the manual processes and automate the mundane and repetitive tasks. My Ally’s AI-powered assistant uses NLP and ML to automate interview coordination activities for the recruiters. The AI-powered recruitment coordinator improves time-to-offer, increases conversion rate, and optimises candidate experience by fully automating the scheduling and rescheduling of interviews with better-than-human speed and accuracy. My Ally solution presents a nearly immediate ROI for organisations both large and small who are looking to reduce time-to-fill and boost candidate conversion rate. They work with clients such as Cisco, Riot Games and Expedia, among others. The Founding Team: It was founded in 2015 by Deepti Yanireddy and Naveen Varma Alluri who were friends since IIT Madras days. The idea of launching My Ally originated during Deepti’s previous role of working as an investor. She realised that over 5-10 hours a week of her time was consumed in scheduling activities and so was the case with Naveen in his previous roles at Qualcomm and Samsung. Soon, after they launched Alex, My Ally’s AI-powered assistant, they realised the solution resonates better with recruiting professionals. With offices in the Bay Area and Los Angeles, they have 50 people managing product development, operations, customer success, marketing functions and much more. Including people in the US, the team size is 60. The startup is backed by Kludein LLC and an investor group including Phanindra Sama (redBus founder), Narayan Ramachandran (former country head of Morgan Stanley, India) and Pranav Pai, investing on behalf of Mohandas Pai among others. [su_divider top=”no” size=”1″] Niramai Co-Founders, Niramai This Bengaluru-based deep tech startup provides breast cancer screening solution using AI. It is a non-contact, non-invasive, radiation-free method of detecting early stage breast cancer in women of all age groups. The core technology by the company called Thermalytix is a fusion of sophisticated ML algorithms over thermal images. The patented algorithms automate the process of analysing the 400,000 temperature values measured per person. Unique thermal patterns and image characteristics that are typically used by medical professionals while detecting cancer in other modalities like XRay or Ultrasound are also used to make the report accurate as well as understandable by the radiologist. The Founding Team: It was founded by Geetha Manjunatha (CEO) and Nidhi Mathur (COO), after they saw their loved ones suffering with cancer in their family and felt deeply about solving this problem. This led to the creation of NIRAMAI, which in Sanskrit means being without diseases. It also expands to “Non-Invasive Risk Assessment through Machine Intelligence”. [su_divider top=”no” size=”1″] Rockmetric Rockmetric is a ‘Cognitive Data Analyst’ that delivers on-demand analysis and relevant insights through a Google-like Natural Language Search interface. It deliveries sophisticated analysis and ad-hoc queries at a scale instantly without having to expand reporting and analytics teams. It automatically understands user queries, analyses data and delivers beautiful charts, descriptive insights and complex analysis instantly. The platform builds an intelligence layer for decision makers and provides them with target-gap analysis, spike and dip analysis, business reviews and routine analysis to monitor their business and teams to scale. These insights are securely delivered with role-based access across channels like web, mobile, email, voice and other communication platforms. The Founding Team: It was founded in 2015, with an idea to create a simple Google-like engine for data and insights. Rockmetric aims to be the most intuitive intelligence layer for business leaders to monitor their businesses and teams. It was founded by Nimesh Mehta, who has consulted with organisations like The World Bank and National Skills Development Corporation. With an MBA from the Indian School of Business, he started his career as a management consultant with Deloitte in India. The startup has an employee strength of six and are currently bootstrapped. It is also a part of Zone Startups India. [su_divider top=”no” size=”1″] Supertext AI Co-Founders, Supertext AI At  Supertext.ai, they believe that communicating with a business or service can be and should be as easy as chatting with a friend — personalised, contextual and fun. Therefore they primarily focus to offer their clients the ability to integrate and add conversational intelligence to answer queries and build persona based virtual agents for Internal business processs automation and Automated front desk customer engagement . The chatbots are designed and developed on their custom AI platform with real-time data analytics, where they have the ability to create and pass tickets via ITSM to support team, have full visibility on bot responses, NLP (natural language processing) training and user interaction history. They build fully automated website chatbots, mobile chat applications and RPA( Robotic process automation) tools for internal business processess for clients across e-commerce, travel, media and fintech. Flipkart, ESPN and MSIG insurance are some of their clients. The Founding Team: It was founded in December 2015 by Avinash Hegde (director), Mathew J Padayatty (director) and Kamalkamman J. The founders realised that though local vendors were building entire businesses by merely chatting with their potential customers over WhatsApp or Facebook Messenger but struggled to scale as they were unable to continue servicing them or build required internal business processes. They wanted to try and solve this by automating business processes at a scale using forward-leaning technologies like Natural language processing and ML. They raised an angel funding from investors in Bengaluru and Singapore. [su_divider top=”no” size=”1″] ten3T Ten3T Team The startup develops medical grade wearable with connected vitals monitoring platform and proprietary health analytics to enable predictive, proactive, and effective clinical decisions. It was founded with a clear vision to address the gap of patient monitoring outside ICU environments — such as responding to early warning signs of deterioration to reduce mortality. ten3T’s Cicer vitals monitoring platform monitors in real time to save lives. The Cicer patch has no lead wires, no disposable electrodes, does not require skin preparation, requires no specialised training, and can be applied as easily as a Band-Aid. An expanding suite of proprietary algorithms look for early signs of deterioration and detect specific clinical conditions, such as respiratory distress, heart rate variability, atrial fibrillation, and patient falls. An easy to interpret colour coded Early Warning Score (EWS) correlates vital signs and provides timely alerts to caregivers. An AI-based system also identifies subtle changes in longer term trends of patient vitals to enable predictive diagnostics. ten3T has commercialised a patent pending vitals monitoring system that is in use at multiple hospitals across Bengaluru. The Founding Team: It was founded in 2014 by a physician and two biomedical engineers with vast and varied experiences in the healthcare industry. The founders are Rahul Shingrani, a biomedical engineer who has worked for over 15 years in the US and India, Dr Sudhir Borgonha, MD, and Prasad Bhat, a biomedical engineer with significant leadership experience in developing and certifying medical devices. With a team of eight, their vision has been to save numerous lives by bringing advanced, predictive AI into the healthcare system. The founders have invested personal funds of $100,000 prior to raising a round of $250,000 from a group of investors in December 2016 that was led by pi Ventures.","excerpt":"Here is the fresh list of 10 startups in artificial intelligence and machine learning that are leading the race in the year 2018. These startups come from different backgrounds, such as providing AI services to healthcare, banking, fintech, diagnostics and HR, among others. Over the years startups have been striving to bring unique propositions into […]","categories":["AI Features"],"tags":[],"author_name":"Srishti Deoras","publish_date":"2018-04-16T06:20:30","publication_year":"2018","word_count":2395,"keywords":["machine learning","artificial intelligence","AI","neural network","ML","NLP","Ray","Aim","deep learning","analytics"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","neural network","NLP","analytics","Aim","Ray"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/10-indian-startups-that-are-leading-the-ai-race-2018\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":6,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10017190,"title":"10 Most Useful Kubernetes Open Source Projects To Watch Out For In 2021","content":"At present, Kubernetes plays an important role in application development. It is a cluster management system for managing containerised applications across multiple hosts, providing mechanisms for deployment, maintenance, and scaling of applications. The system was released in 2014 by tech giant Google, and currently, it powers the Google Container Engine, which is a part of Google Cloud Platform, upon which many new products and services are built. Here is a list of the top ten most useful Kubernetes open-source projects to watch out for in 2021. (The list is in alphabetical order) 1| Helm About: Helm is a popular open-source package manager for Kubernetes that helps users manage the Kubernetes applications. The package manager employs a packaging format known as charts, a collection of files that define a related set of Kubernetes resources. A single chart might be utilised to deploy a Memcached pod or full web app stack with HTTP servers, databases, caches, and so on. The Helm charts help in defining, installing and upgrading the applications. Know more here. 2| Kubeflow About: Kubeflow is a Cloud-Native platform for machine learning based on Google’s internal machine learning pipelines. It is an open, community-driven project to make it easy to deploy and manage an ML stack on Kubernetes. The project is dedicated to executing deployments of machine learning workflows on the Kubernetes. For instance, Kubeflow includes services to create and manage interactive Jupyter notebooks, provides a custom TensorFlow training job operator that can be used to train ML models, etc. It also supports a TensorFlow Serving container to export trained TensorFlow models to Kubernetes and more. Know more here. 3| Kubetail About: Kubetail is an open-source project that is written in Bash Script. It enables users to aggregate (tail\/follow) logs from multiple pods into one stream. The tool basically allows users to monitor the logs from a number of pods and one can also follow or tail the logs according to their pod names. Know more here. 4| Knative About: Knative is an open-source community project that works by adding components for deploying, running, and managing serverless, cloud-native applications to Kubernetes. It eliminates the tasks of provisioning and managing servers, which in result, allow the developers to focus on their codes without the need to worry about setting up complex infrastructures. Some of its features include running serverless containers on Kubernetes with ease, pluggable components that let you bring your own logging and monitoring, networking, and service mesh. It can be used with common tools and frameworks such as Django, Ruby on Rails, Spring, and many more. Know more here. 5| Kubespray About: Kubespray is a composition of Ansible playbooks, inventory, provisioning tools, and domain knowledge for generic OS\/Kubernetes clusters configuration management tasks. Kubespray is a combination of Kubernetes and Ansible. It provides deployment flexibility, highly available cluster, composable attributes, support for most popular Linux distributions and more. Know more here. 6| Kube-capacity About: Kube-capacity is a simple CLI that provides an overview of the resource requests, limits, and utilisation in a Kubernetes cluster. The tool strives to merge the most useful parts of the output from kubectl top and kubectl and then describe it into an easy-to-use CLI focused on cluster resources. Know more here. 7| Prometheus About: Prometheus is an open-source system monitoring and alerting toolkit originally built at SoundCloud. The tool can be used to monitor Kubernetes clusters. This tool’s main features include a multi-dimensional data model with time-series data that are being identified by metric name and key or value pairs. PromQL is a query language that leverages this dimensionality; time-series collection happens via a pull model over HTTP and more. Know more here. 8| Quarkus Quarkus is an open-source, full-stack Kubernetes-native Java framework made for Java virtual machines (JVMs) and native compilation. It is basically a Kubernetes Native Java stack tailored for OpenJDK HotSpot and GraalVM. Quarkus is designed to work with popular Java frameworks and libraries like Eclipse MicroProfile, Apache Kafka, RESTEasy (JAX-RS), Spring, Infinispan and many more. Know more here. 9| Sonobuoy About: Sonobuoy is an open-source diagnostic tool that makes it easier to understand the state of a Kubernetes cluster by running a set of Kubernetes conformance tests and other plugins in an accessible and non-destructive manner. The selective data dumps of Kubernetes resource objects and cluster nodes in Sonobuoy allow for integrated end-to-end conformance-testing, workload debugging and custom data collection via extensible plugins. Starting from version 0.20, Sonobuoy supports Kubernetes v1.17 or later. Know more here. 10| Wayne About: Wayne is an open-source, web-based Kubernetes multi-cluster management and publishing platform. It reduces service access costs by visualising Kubernetes object template editing. By visualising Kubernetes object template editing, it reduces business access costs. Also, it has a complete rights management system and adapts to multi-tenant scenarios. Know more here","excerpt":"At present, Kubernetes plays an important role in application development. It is a cluster management system for managing containerised applications across multiple hosts, providing mechanisms for deployment, maintenance, and scaling of applications. The system was released in 2014 by tech giant Google, and currently, it powers the Google Container Engine, which is a part of […]","categories":["AI Trends"],"tags":["Kubernetes","kubernetes platform","kubernetes tools","open source ios","open source projects","open source projects on github"],"author_name":"Ambika Choudhury","publish_date":"2021-01-05T18:00:00","publication_year":"2021","word_count":797,"keywords":["open source projects on github","machine learning","AI","kubernetes platform","ML","open source ios","RAG","Ray","Kubeflow","Jupyter","kubernetes tools","JAX","open source projects","TensorFlow Serving","TensorFlow","Kubernetes"],"extracted_tech_keywords":["AI","machine learning","ML","Kubeflow","TensorFlow Serving","Ray","TensorFlow","JAX","Jupyter","RAG"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/10-most-useful-kubernetes-open-source-projects-to-watch-out-for-in-2021\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10054611,"title":"What is Activity Regularization in Neural Networks?","content":"One of the most difficult challenges that every precautionary face is the complexity involved in developing algorithms that perform well on training and might well on new inputs. Machine learning employs a variety of techniques to reduce or eliminate test errors. One such technique is regularization. We will discuss why using regularization techniques in the context of regularization is necessary, and we will conclude with a practical demonstration of implementing an activity regularization for the neural network. The following are the key points to be discussed in this article. Table Of Contents Need for Regularization What is Regularization?Keras RegularizersKernel RegularizerBias RegularizerActivity RegularizerImplementing the Regularization in a Neural Network Let’s start the discussion by understanding the need for regularization. Need for Regularization Deep neural networks are sophisticated learning models that are prone to overfitting because of their ability to memorize individual training set patterns rather than applying a generalized approach to unrecognizable data. That is why the regularization of neural networks is so important. It aids the neural network’s ability to generalize data that it does not recognize by keeping the learning model simple. Let’s look at an example to show what we’re talking about. Let’s pretend we have a dataset with both input and output values. Assume there is a true relationship between these values. The goal of deep learning is to approximate the relationship between input and output values as closely as possible. As a result, for each data set, there are two models that can assist us in defining this relationship: a simple model and a complex model. A straight line exists in the simple model that only includes two parameters that define the relationship in question. A graphical representation of this model will include a straight line that closely passes through the centre of the data set in question, ensuring that the line and the points below and above it have very little distance between them. The complex model, on the other hand, has several parameters that vary depending on the data set. It uses the polynomial equation to pass through all of the training data points. The training error will eventually reach zero, and the model will memorize the individual patterns in the data set as the data set becomes more complex. Unlike simple models, which aren’t too dissimilar even when trained on different data sets, complex models can’t be said to be the same. What is Regularization? A major issue in machine learning is developing algorithms that perform effectively not only on training data but also on new inputs. Many machine learning algorithms are intentionally designed to minimize test error, possibly at the expense of greater training error. These procedures are typically referred to as regularization. Deep learning practitioners can choose from a variety of regularization methods. In fact, one of the key research efforts in the field has been the development of more effective regularization procedures. Regularizing estimators are used in the majority of deep learning regularization strategies. The regularization of an estimator works by exchanging higher bias for lower variance. An effective regularizer reduces variance while not increasing bias excessively, resulting in a profitable trade. When we discuss the generalization and overfitting in the model family being trained we encountered three scenarios either (1) excluded the true data-generating process, resulting in underfitting and bias, (2) matched the true data-generating process, or (3) included the generating process but also many other possible generating processes, resulting in an overfitting regime in which variance rather than bias dominates the estimation error. The goal of regularization is to transfer a model from the third to the second regime. Keras Regularizers The weights become more specialized to the training data as we train the model for a longer period of time, resulting in overfitting the training data. The weights will grow in size to handle the specifics of the examples seen in the training data. When the weights are too heavy, the network becomes unstable. Regardless of whether the weight is tailored to the training dataset, minor variations or statistical noise in the expected inputs will result in significant differences in the output. In this case, we can use weight regularization to update the learning algorithm and encourage the network to keep the weights small, and it can be used as a general technique to reduce overfitting. Regularizers allow you to apply weight penalties during optimization. These penalties are added together to form the loss function that the network optimizes. With the help of Keras Functional API, we can facilitate this regularizer in our model layers (e.g. Dense, Conv1D, Conv2D, and Conv3D) directly. These layers expose three-argument or types of regularizers to use, I,e Kernel regularizers, Bias Regulerizers, and Activity Regulerizers which aim to; Kernel regularizer penalizes the layer’s kernel(weight) but does not penalize bias.Only the bias of the layer is penalized by a bias regularizer.The activity regularizer penalizes the output of the layer. Activity Regularization Let us understand the activity regularization before jumping to the implementations. The activity regularization technique is used to encourage a neural network to learn sparse feature representations or we can say internal feature representation of our data that is being fed. Apart from this, this approach is mostly known for reducing the overfitting and to improve the model’s generalization ability to unseen data. Under the hood of this technique, it applies penalties by observing the generalization ability on the validation set. Implementing the Regularization in a Neural Network The below code snippets show how we can use these in the layers mentioned above. from tensorflow.keras.layers import Dense from tensorflow.keras import regularizers layer = Dense(500, activation='relu', kernel_regularizer=regularizers.l1(l1=0.001), bias_regularizer = regularizers.l2(l2=0.001), activity_regularizer = regularizers.l1(l1=0.002)) Now we are going to see how activity regularizer can play a significant role when we want to balance both accuracies. Here we will see how the custom neural network performs on a given set of data, in this case, we will observe the training and validation accuracy of the network before and after the training activity regularizer. from sklearn.datasets import load_iris from tensorflow.keras.models import Sequential from tensorflow.keras.utils import to_categorical from tensorflow.keras.layers import Dense from sklearn.preprocessing import StandardScaler from sklearn.model_selection import train_test_split Load and ready the data iris = load_iris() X = iris.data y = iris.target y = to_categorical(y) ss = StandardScaler() X = ss.fit_transform(X) X_train, X_test, y_train, y_test = train_test_split(X,y) Now below we build the 5 layers artificial neural network, and first, we train it will observe the accuracies. model_1 = Sequential([ Dense(512, activation='tanh', input_shape = X_train[0].shape), Dense(512\/\/2, activation='tanh'), Dense(512\/\/8, activation='tanh'), Dense(32, activation='relu'), Dense(3, activation='softmax') ]) print(model1.summary()) model_1.compile(optimizer='sgd',loss='categorical_crossentropy', metrics=['acc']) hist_1 = model_1.fit(X_train, y_train, epochs=350, batch_size=128, validation_data=(X_test,y_test)) The accuracies at the end of 350 epochs 96.43, 92.11 training and validation respectively. Let’s now try to explore by only applying an activity regularizer. This can be done by setting argument activitity_regularizer by using l1 or l2 norms regularizer. The L1 and L2 have stated as The L1 norm allows some weights to be large while driving others to zero. It penalizes the true value of a weight. The L2 norm causes all weights to decrease in size. It penalizes the square value of a weight. Dense(512, activation='tanh', input_shape = X_train[0].shape, activity_regularizer=regularizers.l1(l1=0.001)) Above is the first layer of the same model which we have seen above. The only change is that we have initiated the activity regularizer. And after the model for 350 epochs, we are getting accuracies of 98.21 and 94.74 for training and validation respectively. The difference that we are getting before and after regularization is nearly +2% which tells us that by applying regularization we can further improve our network performance. Now let’s see the plots of accuracies for both before and after applying regularization Conclusion From the above plot, it is clear that the activity regularizer is doing its job. From both, if compare the instance at epochs 75 in the left plot the validation accuracy is nearly 20% less than training accuracy whereas in the right side plot the same difference is about 5-6%, and nearly the same is maintained throughout the training at an early stage only model tried to overcome the gap. Through this post, we have seen what needs regularization and what is regularization in deep learning in deep learning. Practically we have seen how Keras regularizer can be used in modeling to overcome the overfitting problem. References Regularization for Deep LearningActivity RegularizerCode ReferenceLink for above codes","excerpt":"Deep neural networks are sophisticated learning models that are prone to overfitting because of their ability to memorize individual training set patterns rather than applying a generalized approach to unrecognizable data.","categories":["AI Trends"],"tags":["Deep Learning","Machine Learning","Python","regularization techniques"],"author_name":"Vijaysinh Lendave","publish_date":"2021-12-01T17:00:00","publication_year":"2021","word_count":1393,"keywords":["machine learning","Keras","TPU","AI","neural network","regularization techniques","Machine Learning","RAG","Python","Aim","deep learning","Deep Learning","TensorFlow","R"],"extracted_tech_keywords":["AI","machine learning","deep learning","neural network","Aim","TensorFlow","Keras","RAG","TPU","R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/what-is-activity-regularization-in-neural-networks\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10091980,"title":"Inside the Differently Wired Brains","content":"When Robia Rashid’s ‘Atypical’ made it to screens, it garnered applause for depicting the struggles and discrimination faced by protagonist Sam Gardener (Keir Gilchrist) who has Autism Spectrum Disorder. The ground realities aren’t too different. As of 2021, a staggering 75 million people worldwide amounting to roughly 1% of the global population was estimated to be autistic with 1 in 100 children diagnosed with the disorder. Despite the unique needs of people who fall within the neurodivergent spectrum, they continue to make significant contributions to society demonstrating time and again that their condition is not a hindrance to their talents or potential. Wired Differently: Exploring the Atypical Brain There’s a long list of stars in the scientific world like Thomas Edison, Alan Turing, Isaac Newton, Bram Cohen, Steve Jobs, Elon Musk and Bill Gates who are blessed with what society has often deemed ‘abnormal,’ i.e., the neurodiversity spectrum. Despite the notion that neurodivergent individuals are inferior due to their differences, studies have shown that they have the unique advantage in their ability to think differently from the norm. In particular, neurodiverse individuals are found to be 30% more productive in software-related roles. For example, research on college students with autism showed that they are more likely to major in STEM fields (34.31%) compared to other differently abled groups. Additionally, companies that hire individuals with autism have been found to achieve better financial results than other companies in the same sample. According to a study conducted by Accenture, these companies generated 28% higher revenue, twice the net income, and 30% higher profit margins on average. This suggests that individuals with ASD are an untapped source of STEM talent that can bring significant benefits to their employers. Read more: Indian Tech is Still Oblivious to Gender Inclusivity The Untapped Talent Pool Unfortunately, the employment rate for individuals with autism stands at only 21.7% with most performing menial tasks that do not match their skills while others are still searching for suitable employment. “Autism is a neurodevelopmental condition, not a disease. Society lacks a deep understanding of autism and traditional workplaces were not designed to accommodate individuals with different abilities, especially those with autism as they require special attention and understanding that was not taken into account, even at a subconscious level,“ Shruti Tandon, Managing Director, People Enablement at Nagarro told AIM. At times, people who are neurodiverse may demonstrate challenging behaviour. However, these difficulties can be handled with appropriate accommodations so the potential rewards can be substantial. Nevertheless, for companies to fully benefit from this diverse talent pool, they need to modify their hiring, screening, and professional growth practices to encompass a fuller understanding of skill and ability. Along the same lines, Aditya Malik, Founder and CEO of HRTech company, ValueMatrix.ai told AIM that despite being highly intelligent, diligent and capable of creative thinking, many employers dismiss autistic individuals due to their social difficulties and lack of comfort in interacting with others, particularly in the IT sector. “However, some technology companies are changing this by utilising advanced AI in their recruitment process, which is free of bias and capable of identifying talented individuals, regardless of their disabilities. These companies not only recruit individuals with autism but also offer them the necessary support throughout their employment,” he added. Read more: Is Age Discrimination in Tech for Real? Several major tech companies like Microsoft, IBM, Wells Fargo, SAP, HP, Accenture, JP Morgan & Chase and Nagarro have implemented programs to support and employ neurodivergent individuals. Nagarro launched TestingPro in India, a product that trains people with autism to become software testers. Microsoft has established the “Autism Hiring Program” to hire individuals on the autism spectrum and provide resources for these employees. SAP’s “Autism at Work” program creates job opportunities for people with autism and other cognitive abilities. IBM’s “Neurodiversity Program” offers support and accommodations for neurodivergent employees, while HP’s “Dandelion Program” provides employment opportunities for individuals on the autism spectrum and others with cognitive differences. Better Late than Never Neurodiverse minds have skills that society often overlooks, devalues and even pathologises. Yet, it is precisely in these so-called “disadvantages” that neurodivergent individuals find their greatness. To continue to view neurodiversity as a disadvantage and dysfunction is to deny ourselves the opportunity to tap into the vast potential that these individuals offer. So, if you think that it is still not time for the same society to embrace and celebrate neurodiversity as a valuable asset rather than a hindrance, you are disadvantaged.","excerpt":"75 million people worldwide are estimated to have autism spectrum disorder, but around 80% of them are unemployed.","categories":["AI Features"],"tags":["Autism","bill gates","Elon Musk","nagarro","steve jobs"],"author_name":"Shritama Saha","publish_date":"2023-04-20T18:22:51","publication_year":"2023","word_count":747,"keywords":["nagarro","Go","programming_languages:R","AI","programming_languages:Go","RAG","Elon Musk","GAN","Aim","ViT","Autism","bill gates","R","steve jobs"],"extracted_tech_keywords":["AI","Aim","RAG","R","Go","GAN","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/inside-the-differently-wired-brains\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10081410,"title":"Too Small to Train? Why LLMs Leave Structured Data Untrained","content":"Chinchilla, the 70 billion-parameter model released by DeepMind in mid-April this year solidified certain assumptions about scaling large language models. The performance of LLMs was restricted by the amount of data a model was being trained on and not the size of the model — the comparatively smaller model proved that returns on additional data were huge while returns on additional model size was very little. The research around Chinchilla also demonstrated that the entire quantity of data available on specialised domains like coding was tiny as compared to the potential improvement only if more data was available. Even with all this data floating around, are we really running out of data? Source: AltexSoft.com Structured data vs unstructured data Most of these benchmark LLMs extracted huge amounts of data the one way it found easiest – web scraping. Data is usually picked up from a wide range of sources like public forums, tutorials, Wikipedia, English web documents, Q&A sites on programming and non-English web documents. This type of data is as myriad in its nature as can be – a lot of it also contains garbage which has to be aggressively filtered out. It’s easy to understand why this type of data is known as unstructured. It isn’t organised in a table on a relational database, it includes images, videos, text, audio, tagged formats like XML, HTML or JSON. Training sets for GPT-2 Training sets for GPT-3 Important models like GPT, GPT-2, GPT-3 all have relied on raw data like this. DeepMind’s Gopher and Chinchilla use a dataset called MassiveText and the web scrape portion of the dataset is called the 506 billion token-MassiveWeb. Google AI’s PaLM uses a much smaller and more filtered web corpus of 143 billion tokens. But as the thirst for data amplifies, has AI modelling left behind a huge section of structured data? Structured data is trapped within business applications like product repositories, transaction logs, Enterprise Resource Planning or ERP and Customer Relationship Management or CRM systems. Because model training has been hyperfocused on using unstructured data, the methods of processing tabular are still stuck in an erstwhile era. Older generations of data science techniques like rule-based systems or decision trees which are handcrafted features and difficult to maintain and require a lot of manually labelled data, are still used. Web Scraping techniques, Source: Sapan Mohanty, Medium Why was structured data left behind? Besides there is an existing notion that structured data is simply too ancient to be paired with deep learning techniques. For most experts combining a data paradigm like structured data that is 40 years old with something as new as deep learning is regressive. As deep learning grew more popular, structured data was left for the dead. Structured datasets are simply considered too small to feed deep learning methods, especially when compared with the variety of data that unstructured datasets pull up. The objection to this can be that there are possibly huge labelled structured datasets which contain tens or thousands or even millions of examples but the time and efforts made to process these will also be immense. While structured data is arranged in columns that contain numeric values like currency, temperature, time or other values along with non-numeric values, it can also be unstructured. Structured datasets can also include unstructured values like freeform text or references to XML documents etc. On the other hand, unstructured data itself ironically doesn’t necessarily have zero structure. Value pairs in JSON are also a type of structure but it is still considered as unstructured because JSON values in their native state cannot be in a tabular form. The other argument against structured data is that because of how popular deep learning has become, it has also been simplified, the relevant tools for it are much easier to use. Deep learning solutions can also be used by nonspecialists like the fast.ai library which allows novices to build deep learning models with just a few lines of code. Or even data science platforms like Watson Studio, which has model-builders without any coding. This is why structured data makes more sense for non-deep learning machine learning methods or traditional business intelligence applications in enterprises. Moreover, as one well-known commentary pointed out (https:\/\/t.co\/shchehk8oH), the Chinchilla scaling formula suggests that there's a *hard limit* for model accuracy that no amount of model size will ever overcome without more data. [7\/14]— Davis Blalock (@davisblalock) August 27, 2022 Need still remains But as the incremental benefits of LLMs are running out of steam, researchers are realising that the only way to compensate for the continued dearth in data is by turning to structured datasets. Emad Mostaque, founder and CEO of text-to-image generative AI startup Stability.AI spoke about the need for more data, tweeting that our approach towards general artificial intelligence has to change for maximum impact. To help AI models reach out to as many folk as possible, researchers need to include more diverse highly structured datasets. “Stacking more layers is fine as GPT-4 is about to show but there are superior routes,” he stated.","excerpt":"For most experts combining a data paradigm like structured data that is 40 years old with something as new as deep learning is regressive","categories":["AI Features"],"tags":[],"author_name":"Poulomi Chatterjee","publish_date":"2022-12-05T13:00:00","publication_year":"2022","word_count":844,"keywords":["data science","Go","API","machine learning","artificial intelligence","AI","ML","deep learning","generative AI","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","data science","generative AI","R","Go","API"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/too-small-to-train-why-llms-leave-structured-data-untrained\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":16348,"title":"What’s the latest buzz about Artificial Intelligence creating fake Obama?","content":"This might come across surprising for lot many AI enthusiasts out there, but the technology actually fosters the capability to create fake audio and video, which is difficult to distinguish from reality. In a recent feat, scientists at University of Washington created an AI software that could generate highly realistic fake videos of former president Barack Obama using existing audio and video clips of him. The tool essentially takes audio files, converts them into realistic mouth movements, and then grafts those movements onto existing video. The resultant video shows someone saying something they didn’t. University of Washington scientists had previously revealed that the tool could be utilized for generating digital doppelgangers of anyone by simply analyzing their images. This could include celebrities such as Tom Hanks and Arnold Schwarzenegger, or even political figures like George W. Bush and Barack Obama, whose images are easily available on the internet. The research was funded by Samsung, Google, Facebook, Intel, and the University of Washington. The findings of this project will be detailed on August 2nd at the SIGGRAPH conference held in Los Angeles. Researchers claimed that it might soon be possible to generate digital models of a person for virtual reality or augmented reality applications. Why did researchers at University of Washington choose Obama for the project? Obama turned out to be the best public figure for this AI-based project. This is largely due to the fact that there are hours of high-definition video of him available online in the public domain. The research team saw this as a huge opportunity to test their software. They had a neural net analyze millions of frames of video to determine how elements of Obama’s face moved as he talked, such as his lips and teeth, and wrinkles around his mouth and chin. How does the AI-based software work? An artificial neural network usually comprises of components known as artificial neurons which are fed with the data. They work together to solve a problem, for instance, identifying faces or recognizing speech, following which the neural net alters the pattern of connections among those neurons to change the way they interact. Next, the network tries to resolve the problem. The neural net learn which patterns are best at computing solutions, over time. This essentially represents an AI strategy that mimics the human brain. However, the new study involved neural net learning what mouth shapes were linked to various sounds, for which researchers took audio clips and dubbed them over the original sound files of a video. In the consecutive step, they took mouth shapes that matched the new audio clips, which were further grafted and blended onto the video. They also synthesized videos of Obama where he lip-synced words he must have uttered ages back. Previously, the researchers partook in a similar project, which involved filming people saying sentences over and over again, to map what mouth shapes were linked to various sounds. But, this process was not only expensive and tedious, but it also consumed a good deal of time. On the other hand, the new software can learn from hours of video that already exist on the internet. One noteworthy aspect here is the fact that the link between mouth shapes and utterances may be universal for people to some extent. Basically speaking, a neural network trained on Obama and other public figures could be adapted to work for several other people. The concern surrounding application of the AI-based software Once you have such a resourceful tool capable of leveraging AI extensively, there will always be concerns surrounding the software’s abuse. It becomes really simple to generate misleading video footage once you have your hands on a tool as such. It can indeed turn out scary when you think about the pitfalls of having this technology in the wrong hands. Keeping this in mind, the researchers at University of Washington were cautious about not generating videos where they put words in Obama’s mouth, that he had not utter himself. How useful can the software be? Improving videoconferencing could be one potential application for this new technology, as claimed by Ira Kemelmacher-Shlizerman, a co-author to this project. Teleconferencing video feeds may often stutter, freeze, or suffer from low-resolution; but the audio feeds mostly work. In the future, videoconferencing may simply transmit audio from people and this software could be leveraged to reconstruct what they might have looked like while they talked. Besides, such software could also help people talk with digital copies of a person in virtual reality or augmented reality applications. This is not all, the technology can be applied to detect fake videos in the future. It could be a difficult task at times for anyone to make out if a video is fake or real, as the differences might not be so distinct to human eyes. However, an AI-based program that can compare the blurriness of the mouth region to the rest of the video can be easily developed. Looking into the future of such technological advances This project clearly illustrates how training on a large amount of video of the same person, and designing algorithms with the goal of photo-realism in mind, can help in creating believable video from audio with convincing lip sync. As discussed earlier, such projects open up a number of interesting future directions. It could be however difficult to train the system on another person, for instance, a non-celebrity. This is because it’s not an easy task to obtain hours of training data. However, there are chances that association between mouth shapes and utterances could be speaker-independent. In other words, we could perhaps retrain the network used in Obama’s case for another person with lesser need for additional training data. Talking about taking the system a notch higher, it might be possible to train a single universal network from videos of many different people, which could be further conditioned on individual speakers. For say, the AI-based system could be fed with a small video sample of the new person, to produce accurate mouth shapes for that person.","excerpt":"This might come across surprising for lot many AI enthusiasts out there, but the technology actually fosters the capability to create fake audio and video, which is difficult to distinguish from reality. In a recent feat, scientists at University of Washington created an AI software that could generate highly realistic fake videos of former president […]","categories":["IT Services"],"tags":["Artificial Neural Network","artificial neurons","augmented reality India","latest advances","training data","videoconferencing","virtual reality india"],"author_name":"Amit Paul Chowdhury","publish_date":"2017-07-14T12:19:13","publication_year":"2017","word_count":1010,"keywords":["Go","artificial neurons","videoconferencing","AI","neural network","training data","Artificial Neural Network","R","latest advances","Git","programming_languages:R","RAG","Aim","augmented reality India","CLIP","GAN","virtual reality india"],"extracted_tech_keywords":["AI","neural network","Aim","RAG","R","Go","Git","CLIP","GAN","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/it-services\/whats-latest-buzz-artificial-intelligence-creating-fake-obama\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":59151,"title":"Why Haven’t Companies Delivered On Their Autonomous Driving Promise","content":"Elon Musk promised us that automated vehicles would be on the streets in 2019, he made this claim back in 2017. And it wasn’t only him, other companies also promised autonomous cars and failed to deliver. The vision of self-driving cars has encountered a lot of pitfalls, and with them not being on the roads yet, it seems that autonomous vehicles have posed problems that weren’t expected by the manufacturers. These autonomous cars\/vehicles hold a lot of potential like reducing the traffic-related accidents by 90%, reducing 60% of harmful emissions, 40% reduction in travel time etc. With so many possibilities, one has to wonder what happened? Below are mentioned some of the possible reasons: Is LIDAR Ready? LIDAR, which stands for light detection and ranging, is the tech behind self-driving cars that makes it possible for them to be aware of the world they are driving inside. LIDAR uses lasers that detect, measure and identify objects around the car. LIDAR faces problems when it comes to countering the weather, especially snow. Although car manufacturers have started expanding their range for testing autonomous cars, they haven’t correctly figured out how to deal with snow or rain. Not only do the self-driving cars, but the semi-autonomous ones also find it quite tricky when it comes to operating in snowy environments. They face problems like adaptive cruise control or self-parking when the snow covers their sensors. In addition to this, LIDAR is an expensive piece of tech, hard to produce and even harder to buy. Because of these, only a few companies can produce LIDAR sensors in a large quantity, and they too struggle to expedite the manufacturing of the sensor. Some Holes In The Tech A lot of questions about the infrastructure arise when it comes to autonomous cars development. Manufacturers still need to answer questions about vehicle to vehicle communication, functioning properly on stoplights, smart roads that will help cars navigate and 5G integration for fast real-time navigation on these roads. In the tests that have been going on till now, none of these elements is uniform when it comes to different parts of the world. So, until and unless these elements are made stable and appropriately tailored for autonomous vehicles, it will be tough to see their proper functioning. The Talent Gap When It Comes To Working On Self-driving Cars It is a known fact that companies are in a struggle with each other when it comes to making their driverless cars fully-functioning and ready to be used. Car manufacturers have been promising self-driving cars for a long time now, and their deadline is near, one of these areas they have struggled mightily is to find the talent to help them accomplish their goals. Only a handful of people know how to design and build autonomous cars. Learning How To Properly Drive While the world prepares for autonomous cars, they are also trying to adapt to real-world situations. These self-driving cars have to deal with the unusual setting of the real-world, like crossing pedestrians, recognising illegal turns along with understanding red and green lights. Another problem autonomous cars face is the challenge of learning better driving or the ‘unspoken rules’ of human driving. For example, if a vehicle in the front is searching for a parking spot, it slows down to search for one. So, in this case, the self-driving cars have to recognise that it should not be following the car too closely as it can turn at any time to move into a parking space. Dealing With The Humans The autonomous cars are coded with proper traffic rules and may even demonstrate proper driving someday. Still, one of the biggest concerns remains about how these cars will deal with the unpredictable human nature or with the people who don’t play by the rules. We all know humans tend to break the rules or try to tweak some rules. Humans double park, suddenly appear in front of cars, do not hit indicators while making a turn, etc., the list goes on. Whether the self-driving cars will be able to adapt to human conditions is to be seen—for example, the incident with Uber’s autonomous car. The self-driving car being tested by Uber hit and killed a woman who was walking across a street in 2019, Tempe. The safety driver was present in the car, who was watching videos on the phone but couldn’t help from the accident to take place. Not saying that the woman crossing the streets was breaking the rules or not, but, the car should have been able to recognise the person. Another challenge that is encountered by self-driving cars is making the turn against oncoming traffic. This is a challenge even for human drivers and one of the major reasons, which can cause crashes. The car has to detect, measure and decide when to make that turn. All these safety-related challenges are the ones that still have to solve, and with car manufacturers still not being able to completely overcome these even on a testing level, it has further delayed the self-driving cars to come on the streets. “There was a sense maybe a year or two ago, that ‘Oh, our algorithms are so good! We’re ready to launch. We’re gonna launch driverless cars any minute.’ And then obviously there’s been the setbacks of people getting killed or accidents happening, and now we’re a lot more cautious.” – Avideh Zakhor, a professor at the University of California at Berkeley’s electrical engineering and computer sciences department.","excerpt":"Elon Musk promised us that automated vehicles would be on the streets in 2019, he made this claim back in 2017. And it wasn’t only him, other companies also promised autonomous cars and failed to deliver. The vision of self-driving cars has encountered a lot of pitfalls, and with them not being on the roads […]","categories":["AI Features"],"tags":["autonomous cars","computer vision autonomous vehicles","IT companies"],"author_name":"Sameer Balaganur","publish_date":"2020-03-20T11:00:00","publication_year":"2020","word_count":919,"keywords":["IT companies","Go","autonomous cars","programming_languages:R","AI","programming_languages:Go","Aim","computer vision autonomous vehicles","R"],"extracted_tech_keywords":["AI","Aim","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/why-havent-companies-delivered-on-their-autonomous-driving-promise\/","complexity_score":3,"technical_depth":6,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":23324,"title":"Top 3 Ways to Personalize Retail Newsletters","content":"The power of sophisticated technology is no longer the privilege of technologists alone. It has moved into the hands of the common man. This consumerization of technology has put the customer in the driver’s seat and retailers are expected to deliver more convenience, more consistency, more collaboration and more customization than ever before.  This evolution is forcing the once predictable retail industry, to make the customer as the focus of every merchandise, marketing, store and supply chain decisions. As the industry shifts focus from pushing profitable products and goods to catering to the digitally empowered customer, customer engagement is playing a crucial role in bringing in the moolah. Today’s retailers have many channels, locations and devices to converse with their customers. Be it building a platform that supports omnichannel communication, installing indoor positioning systems that communicate in-store, empowering store associates with insights for effective clienteling, retailers are leaving no stones unturned. But what’s the low hanging fruit that every retailer HAS to get right? Email newsletters. Email newsletters are a great way to continue the conversation post a purchase, pave the way for further sales and keep your brand on top of the customer’s mind. This continued engagement has been a staple for any business and the difference between a great newsletter and one that moves into spam is, drumrolls, ‘personalization’. Email personalization has been around for a while now and there are some rudimentary tick marks that any basic CRM can offer you. Things like: adding the first name of the customer after a hello, scheduling emails to go out at times that the user has opened emails in the past, setting up a flowchart of ‘if this then that’ messages, etc. But to drive maximum outcomes out of newsletters one must look at the customer as an individual and fully understand who s\/he is as a person before marketing to them. Agreed, doing this at scale might be a challenge, but that’s what segmentation is for. The more time you spend in segmenting and micro-segmenting your target audience, the better the results of your newsletters. Let me elaborate with these three examples\/ tips that you can use to personalize your newsletters. Understand that your customer data is dynamic and evolves over time The database to send newsletters either comes through the loyalty card\/ membership data the retailer already has or acquires a ‘fresh’ database from a service provider that charges anywhere from 5c to $10 for each email id. Majority of promotions are run on these email ids that the retailer has acquired over a period of time. If your team regularly cleanses this data, you are solving the issues of bounce rates, open rates etc. but the click rates? Click rates require you to offer up the right recommendation to the individual and the individual has to receive the most contextual offer that matches his lifestyle at that very point in time. For example, your loyalty data says that Mr. X bought $50 worth of meal kit on Jan 2016. This data has entered into your system and has been lying there for more than two years. Now your marketing team is looking to promote a newly launched, ‘fit and fab’ meal kit and Mr. X unwittingly enters into the newsletter list. But hey, he never clicked on the offer you sent him. Do we need to cleanse the DB again!!? What really happened is that Mr. X got married to Ms. Y in Jan 2017 and started shopping for grocery items instead of meal kits. He also increased his share of wallet to $500 per month by shopping for ration supplies for the entire week. He drops into the store every Sunday evening to pick up stuff for the rest of the week. It’s all there in your loyalty data. Now, as a retailer who wants to personalize newsletters, would you look at individual purchase items or look at buying patterns and understand the intuitive meaning that every basket has? Micro-segmentation of your database just solves one part of the puzzle, but to really personalize offerings, you need to devise strategies and have systems in place to allow movement of contacts within different segments upon every new purchase. Your system should be able to tell you that Mr. X did not respond to the meal kit offer you sent him, remove him from the DB, look at his shopping patterns and recommend that you send him a ‘20% off on Organic tomatoes’ coupon. If you have your marketing team aligned with this thought, then its bound to increase conversions. But if you have systems that have recommendation algorithms that can do this at scale, every single time, even better! Look at ‘like to like behavior’ of similar individuals If you are going to feed a stray dog, are you going to give him broccoli or left-over meat? The meat of course. Now how do you know that the stray would prefer meat as opposed to broccoli? Have you observed its eating habits over a period of time? Of course not. You have a dog at home that would give you a cold stare if you fed him anything other than meat. You have learned that if my dog would not prefer broccoli, then the stray would not either. It’s a logical conclusion that you arrived at by assuming that dogs, in general, prefer meat because my dog eats meat. Extrapolate this thought to your marketing. (Ah, the things your dog teaches you!). You have 10 items in your new product catalog that you have to market in your newsletter. Let’s take 3 different segments. Segment 1 – bought dairy and eggs. Segment 2 – bought meat, wine, and beer. Segment 3 – bought diary, eggs and chocolate sauce. Now, if one the 10 items you were to market was chocolate sauce, which segment would you pick? Segment 3. But based on purchase data, the baskets of segment 1 and segment 3 is almost the same. Can we try offering up a discount on the chocolate sauce for segment 1 based on the sales volume that segment 3 is seeing? Let’s label segment 3 as ‘Bakers’ and segment 3 as ‘suspected bakers’. See what happened there? You have created a microsegment based on purchase patterns and created a cohort called as ‘suspected bakers’. People can move in and out of this mailing list depending upon whether they purchase the chocolate sauce or not. Once you deploy such strategies to your database, you are effectively moving away from ‘trial and error’ marketing which cannot hold a candle to personalized marketing. Pop quiz – which segment would you categorize as ‘weekend barbeque’? Many messages, same database Yes, audience exhaustion is real. The more the customer opens your emails and finds out that there is nothing interesting in there, the more your response rates and ultimately sales, suffer. It is tempting to send an exciting new product that you are launching in the store, but don’t give in! The same customer is present in your ‘bakers’, ‘suspected bakers’ and your ‘weekend barbeque’ lists. Are you going to send the same, ‘new product discount on ketchup’ three times? Surely, touch monitoring systems will detect send frequencies and alert you about potential overlaps, but they just look at time stamps against a particular id and do not recommend what message you need to send them. This is where your team needs to do the due diligence and spend considerable time effort in creating microsegments. And stick to it. If you have identified ‘weekend barbeque’ as a cohort, you must stick to messaging that is appealing to this cohort. This messaging can be around the lifestyle that a weekend barbeque enthusiast might lead. For example, you send him recipes for killer BBQ ribs in week 1, discount on BBQ sauces on week 3, invite for a store cookout competition on week 5. If the contact does not respond to any of these, then it’s time to move him out of the cohort. But imagine the loyalty that this sort of message is bound to bring (and the sales). You could also try segmenting based on the specific shopping behavior like ‘most likely to respond to coupons’, ‘most likely to respond to try new products’, but until you get to segmenting customers as ‘most likely to respond to a coupon on organic eggs’, you are not acing personalization.","excerpt":"The power of sophisticated technology is no longer the privilege of technologists alone. It has moved into the hands of the common man. This consumerization of technology has put the customer in the driver’s seat and retailers are expected to deliver more convenience, more consistency, more collaboration and more customization than ever before.  This evolution is […]","categories":["AI Trends"],"tags":[],"author_name":"Santosh Kumar","publish_date":"2018-04-05T06:44:17","publication_year":"2018","word_count":1404,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","Git","Ray","ViT","GAN","R"],"extracted_tech_keywords":["AI","Ray","R","Go","Git","GAN","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-3-ways-to-personalize-retail-newsletters\/","complexity_score":4,"technical_depth":9,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":33619,"title":"Understanding Credit Risk Analysis In Python With Code","content":"Credit risk analysis provides lenders with a complete profile of the customer and an insight that enables them to understand customer behaviour. New Methods Today, advanced analytics techniques enable firms to analyse the risk level for those clients with little to no credit account based on data points. Let’s list down the methods used for credit risk analysis. Standard Deviation In this method, formula measures the dispersion of data from its expected value. The standard deviation is employed in making an investment decision to measure the amount of historical volatility compared with an investment relative to its annual rate of return. It indicates how much the current return is diverging from its supposed historical normal returns. For example, a stock that has a high standard deviation experience larger volatility, and accordingly, a higher level of risk is compared with the stock. Beta Measure Beta is another popular measure of risk. Beta measures the volume of systematic risk individual security or an industrial sector has related to the whole stock market. The market has a beta of 1, and it can be practised to gauge the risk of security. If a security’s beta is equivalent to 1, the security’s price moves in time step with the market. A security with a beta higher than 1 indicates that it is more volatile than the market. Conversely, if a security’s beta is smaller than 1, it symbolises that the security is less volatile than the market. For example, assume a security’s beta is 1.5. In hypothesis, the security is 50 per cent more volatile than the market. Value At Risk(VAR) VAR is a statistical model used to estimate the level of risk connected with a portfolio or company. VaR estimates the maximum potential decline with a degree of reliance for a specified period. For illustration, assume a portfolio of investments has a one-year 10 per cent VAR of $5 million. Consequently, the portfolio has a 10 per cent probability of losing more than $5 million over a one-year period. Conditional VAR It is another risk measure adopted to estimate the tail risk of an investment. Used as an extension to VaR, the conditional VaR estimates the likelihood, with a particular degree of confidence, that there will be a break in the VaR; it seeks to assess what happens to an investment exceeding its maximum loss threshold. This measure is more susceptible to events that happen in the tail end of distribution – the tail risk. For illustration, a risk manager thinks the average loss on an investment is $10 million for the worst 1 per cent of potential outcomes for a portfolio. Therefore, the conditional VaR, or anticipated shortfall, is $10 million for the 1 per cent tail. Implementing With Python To calculate Credit Risk using Python we need to import data sets. For example, we take up a data which specifies a person who takes credit by a bank. Each individual is classified as a good or bad credit risk depending on the set of attributes. It’s impossible to understand the original dataset due to its complicated system of categories and symbols. The data set can be converted into a CSV file format which can be understood easily. The following are some of the attributes that are to be mentioned in the data set. The entire data set for the program is taken from https:\/\/archive.ics.uci.edu\/ml\/datasets\/Statlog+%28German+Credit+Data%29 Code and picture credits: https:\/\/www.kaggle.com\/kabure\/predicting-credit-risk-model-pipeline Age (numeric) Sex (text: male, female, other) Job (numeric: 0 – unskilled and non-resident, 1 – unskilled and resident, 2 – skilled, 3 – highly skilled) Housing (text: own, rent, or free) Saving accounts (text – little, moderate, quite rich, highly rich) Checking account (numeric) Credit amount (numeric, in DM) Duration (numeric, in the month) Purpose(text: car, furniture\/equipment, radio\/TV, domestic tools, repairs, education, enterprise, vacation\/others Risk (Value  – Good or Bad Risk) Code: import pandas as pd #Library To work with a dataset import numpy as np #Math library import seaborn as sns #Graph library that use matplot in preparation import matplotlib.pyplot as plt #to plot unusual parameters in seaborn #Importing the data df_credit = pd.read_csv(“.######################.”,index_col=0) print(df_credit.info()) #Looking unique values print(df_credit.nunique()) #Looking the data print(df_credit.head()) Format for the plot grid: [(1,1) x1,y1][(1,2)x2,y2][(2,1)x3,y3] # Credit Amount column interval = (18, 25, 35, 60, 120) cats = [‘Student’, ‘Young’, ‘Adult’, ‘Senior’] df_credit[“Age_cat”] = pd.cut(df_credit.Age, interval, labels=cats) df_bad = df_credit[df_credit[“Risk”] == ‘bad’] df_good = df_credit[df_credit[“Risk”] == ‘good’] Distribution of Housing own and rent by risk factor Code: # 1st plot trace0 = go.Bar( x = df_credit[df_credit[\"Risk\"]== 'good'][\"Housing\"].value_counts().index.values, y = df_credit[df_credit[\"Risk\"]== 'good'][\"Housing\"].value_counts().values, name='Good credit' ) 2nd plot trace1 = go.Bar( x = df_credit[df_credit[“Risk”]== ‘bad’][“Housing”].value_counts().index.values, y = df_credit[df_credit[“Risk”]== ‘bad’][“Housing”].value_counts().values, name=”Bad Credit” ) data = [trace0, trace1] layout = go.Layout( title=’Housing Distribution’ ) fig = go.Figure(data=data, layout=layout) py.iplot(fig, filename=’Housing-Grouped’) Distribution of Credit Value by Housing Code: fig = { \"data\": [ { \"type\": '###', \"x\": df_good['Housing'], \"y\": df_good['Credit _amount'], \"legendgroup\": 'Good Credit', \"scalegroup\": 'No', \"name\": 'Good _Credit', \"side\": 'negative', \"box\": { \"visible\": True }, \"meanline\": { \"visible\": True }, \"line\": { \"color\": '###' } }, { \"type\": '###', “x”: df_bad[‘Housing_’], “y”: df_bad[‘Credit_amount’], “legendgroup”: ‘Bad Credit’, “scalegroup”: ‘No’, “name”: ‘Bad Credit’, “side”: ‘positive’, “box”: { “visible”: True }, “meanline”: { “visible”: True }, “line”: { “color”: ‘green’ } } ], “layout” : { “yaxis”: { “zeroline”: False, }, “violingap”: 0, “violinmode”: “over_lay” } } py.iplot(fig, filename = ‘violin_\/split’, validate _= False) In Conclusion Today, credit risk analysts work across various sectors like Consumer & Retail, Gaming, Healthcare, Insurance, Finance, Media & Telecom, Natural Resources, Banks, Broker and Asset Managers and many more.","excerpt":"Credit risk analysis provides lenders with a complete profile of the customer and an insight that enables them to understand customer behaviour. New Methods Today, advanced analytics techniques enable firms to analyse the risk level for those clients with little to no credit account based on data points. Let’s list down the methods used for […]","categories":["Deep Tech"],"tags":["credit risk machine learning","free datasets for analysis"],"author_name":"Bharat Adibhatla","publish_date":"2019-01-17T05:18:28","publication_year":"2019","word_count":923,"keywords":["NumPy","free datasets for analysis","AI","credit risk machine learning","ML","RAG","Python","Seaborn","analytics","Matplotlib","R","Pandas"],"extracted_tech_keywords":["AI","ML","analytics","Pandas","NumPy","Matplotlib","Seaborn","RAG","Python","R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/understanding-credit-risk-analysis-in-python-with-code\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10083032,"title":"A Quest to Develop an Open-Source Secure Element Chip","content":"In today’s connected world, there are many facets of technology that we do not directly see, but these still play a very important role in our digital safety. Among the ranks of encryption algorithms and authentication mechanisms, we have a contender for the hidden watchdog of the digital world – secure element chips. Secure element chips, or SEs, have been around since the late 90s and have become one of the most-overlooked parts of our hardware. These chips are an integral part of the Secure Enclave, a set of hardware and software features which prevent a wide range of attack vectors for malicious attackers. These elements are tamper proof and must adhere to a strict standard. But what is surprising is how pervasive they are. Smartphones, tablets, and laptops all have Secure Enclaves to carry out a variety of security-based actions. The secure element is tamper-proof from both hardware and software perspective, and provides a secure environment to enable other cybersecurity features. The chip is able to provide certain security guarantees that create a root of trust for encryption systems, as well as the provision of secure memory which can be used to store encryption keys and other sensitive information. The presence of a Secure Enclave prevents attackers from being able to access this information even if the system has been compromised. The development process behind this security offering is one of the most closed ecosystems in the hardware development space. Working on a secure element usually means signing multiple NDAs and guarding the code repository with the highest degree of security. But that might change soon. Why do we need secure elements? To understand why secure elements are so important, we must first take a look at how encryption works. Through a process of using complex mathematical algorithms and digital keys, computers can encode sensitive data to prevent cyberattacks. However, the keys are used to decode the same data for the recipient, making them one of the most sensitive data that can be stored on a device. Storing these keys is just one of the functions of a Secure Enclave. In short, Secure Enclaves provide confidentiality, attestation, and integrity to devices. Confidentiality refers to the quality of an SE that prevents attackers from knowing the execution state of the chip. This means that any action undertaken in the chip is kept secret from the user or the attacker. Attestation is a hardware-based feature of the SE that allows a remote party to verify what has run inside the enclave in a trustless fashion. Integrity is a feature of the enclave that ensures that no matter what happens on the external operating system, the SE will continue to work as intended. These three features allow the SE to enable secure computation, both locally and in the cloud. Even in a situation where the end user cannot see or interact with the hardware they are working on, Secure Enclaves can give them the confidence that malicious parties cannot hijack their data. Even though software techniques exist to enable secure computation with multiple parties, SEs bring the security of these methods while still providing fast computation. Due to their nature as fully-integrated systems, SEs have a low computational cost and add a high degree of value to the system they are integrated into. The race towards an open-source SE Just as with any argument in the cybersecurity space, there are two sides to creating an SE with completely open-source tools. Kerckhoffs’s principle states that any cryptographic system must be secure even if all information about the system is out in public; a law open-source enthusiasts swear by. This puts the onus on developers and creators of SEs to keep them secure even if the inner workings of the system is common knowledge. However, security through obfuscation is the other side of the coin, an easy cop-out for companies to keep everything secret and, by extension, secure. Closed-source designs aim to preserve security by not disclosing the architecture and implementation of the SE. Most SEs are written in Java Card and MULTOS, and as with any kind of human-written code, it comes with a host of bugs. However, by making the SE tamper proof and closed to penetration testing, companies can get away with having security-breaking bugs in their code. This also extends to the architecture of the SE, as the way the chips are placed can also open up an opportunity for malicious attackers to use high-tech attack vectors such as ion beams. An open-source SE, by contrast, will have all code in the open, with architecture partially being open, only constrained by silicon foundry NDAs. Low-level silicon like SEs usually use ARM-based SoCs, which are closed source. However, we are now seeing a trend towards the completely open RISC-V architecture, which is completely open-source. Using open-source code can also speed up the discovery of errors, an area where closed-source SEs have been lacking. A prime example of the damage caused by closed-source SEs is Foreshadow, an exploit that targets SE on Intel processors. The vulnerability was only discovered after researchers put in behemoth efforts to uncover it, which was made even more difficult due to the measures taken by Intel to prevent information gathering. It is likely that an issue of this size would have been identified and patched within a week of the SE launching on an open-source product. We have already seen many companies undertaking the task of creating an open-source Secure Enclave, such as Keystone and Cranium. While these companies are still bound by silicon foundry NDAs, meaning they cannot disclose the specifics of the architecture of the chip, the open-source access to the code will allow the manufactures to leverage the pen-tester ecosystem to make a stronger product. This also sets an interesting precedent for the SE industry as a whole, as NDAs are par for the course in this field. There is a huge elephant in the room when it comes to SEs; the rise of quantum computing. However, the encryption community as a whole has considered that to be a problem to be tackled when it actually becomes a problem, leaving space and time for SE manufacturers to turn to the libre side.","excerpt":"Secure Enclaves are one of the most overlooked facets of digital security","categories":["IT Services"],"tags":[],"author_name":"Anirudh VK","publish_date":"2022-12-22T12:00:00","publication_year":"2022","word_count":1038,"keywords":["Go","programming_languages:R","AI","programming_languages:Java","Git","RAG","Aim","Rust","R","Java"],"extracted_tech_keywords":["AI","Aim","RAG","R","Go","Rust","Java","Git","programming_languages:R","programming_languages:Java"],"url":"https:\/\/analyticsindiamag.com\/it-services\/a-quest-to-develop-an-open-source-secure-element-chip\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10114825,"title":"Marlabs Sets Sights on Making India the Epicenter of its AI Initiatives","content":"Two years ago, IT consulting and services company Marlabs introduced its first Experience Centre in Yelahanka, Bengaluru, marking a milestone in the company’s expansion plans in India. Since then, the country has become one of the key markets for the company. Now, Marlabs is aiming to make India the centre of excellence for its AI products. “Though headquartered in the US, Marlabs truly has its heart and soul in India. This Indian nexus is not just a physical presence; it’s where the energy and dynamism of our company are most palpable,” Thomas Collins, chief executive officer, Marlabs, told AIM during his recent visit to India. Expanding Footprint in India India is poised to be the epicentre of AI initiatives at Marlabs, staffed with specialists equipped with the necessary skills and tools to oversee AI projects from start to finish, ensuring they are grounded in a strong foundational process. “It’s here that we’ve established our centre of excellence, focusing on data engineering, analytics, and digital product development services. This strategic decision is rooted in our heritage, our talent pool within the company, and the rich external talent market in India,” added Collins. About 75% of the company’s employee base is in India and it plans to add at least 500 more employees in a year. The company’s hiring strategy is multi-faceted. It is looking for individuals with a blend of consultative skills – a mix of technology and business acumen. “Additionally, data science expertise, specifically in developing algorithms and data engineering, forms the crux of our talent acquisition. The focus is on building solutions for our clients, which requires a spectrum of skills from consulting to technical expertise,” added Collins. The Marlabs’s data science team addresses a range of industry challenges, emphasising tasks like extracting insights from extensive datasets and employing pattern recognition, prediction, forecasting, recommendation, optimisation, and classification. In an earlier conversation with AIM, Sriraman Raghunathan, digital innovation and strategy principal, Marlabs, said that the company’s hiring process for data science roles in India touches upon parameters like technical expertise, practical application, and experience. It involves clear role definition, thorough job descriptions, technical assessments, AI\/ML expert interviews, and HR interviews to ensure candidates are not only technically proficient but also fit well with the company culture and collaborative environment. Growth Strategy Since assuming the leadership role in October, Collins has articulated a clear and ambitious vision for the company’s trajectory in AI and ML. The goal is to double the company’s size within three years. “Central to this vision is the integration of data engineering and AI as key drivers of revenue, aiming for these areas to contribute about 50% of the company’s total revenue,” said Collins. The strategy to realise this vision is twofold. First, there is a focus on organic growth, which involves nurturing and expanding the company’s existing resources, skills, and capabilities in AI and ML. Second, since the primary focus for the next three years will be on life sciences and healthcare, Collins emphasises the importance of strategic acquisitions as a means to accelerate growth. This includes its recent acquisition of Indianapolis-based AI consulting firm Onebridge which specialises in data engineering and analytics in the life sciences sector. Marlabs is currently working with four of the top ten global life sciences companies. Another important aspect of our strategy is differentiating ourselves as a service provider, primarily through our company culture. “We believe in offering a great experience to both our employees and clients. It’s not just about the skills we possess but also about how we engage with our clients. We’ve received positive feedback about our approach and aim to further enhance the client experience,” Collins added. AI & Analytics Play Marlabs’ AI and ML operational efforts have covered a variety of projects. These include demand forecasting, inventory optimisation, linking point of sale data, evaluating admission candidates, detecting anomalies in real-time operations, and identifying irregularities in clinical trial reports. “Generative AI is a game-changer, and we’re tackling it from two fronts: client solutions and internal applications. For our clients, we’re building industry-specific AI frameworks, using open-source technologies to maintain independence from major tech giants,” said Collins. An example of this approach in action is a project the company undertook for a major telecom client, which enhanced its operational efficiency. Internally, the team is exploring the potential of generative AI to improve software development and increase efficiency in various support functions, including finance, HR, and marketing. The team is also delving into generative AI for tasks like extracting and summarising knowledge bases in areas such as IT service desk ticketing, sustainable finance, medical device management, and rare disease education. However, the key challenge lies in managing the quality of client data. Effective implementation of generative AI often requires significant preliminary data engineering work to ensure the data is properly harmonised and curated. “This has been a growing focus area for us, especially since the rise of generative AI has highlighted the need for more robust data strategies,” concluded Collins. Read more: Data Science Hiring Process at Marlabs","excerpt":"“75% of our employees are in India and the company plans to add at least 500 more employees in a year,” Thomas Collins, CEO, Marlabs, told AIM.","categories":["AI Features"],"tags":["Interviews and Discussions","Startups"],"author_name":"Shritama Saha","publish_date":"2024-03-01T11:01:11","publication_year":"2024","word_count":840,"keywords":["data science","Go","AI","ML","Git","RAG","Aim","analytics","generative AI","Startups","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","ML","data science","analytics","generative AI","Aim","RAG","R","Go","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/marlabs-wants-to-make-india-the-hub-for-its-ai-initiatives\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10068118,"title":"‘I don&#8217;t really trust papers out of top AI labs anymore’","content":"The role of scientific research in pushing the frontiers of artificial intelligence cannot be overstated. The researchers working at MIT’s Computer Science and Artificial Intelligence Laboratory, Stanford Artificial Intelligence Laboratory, Oxford University and many other top labs are shaping the future of humanity. In addition, most top AI labs, even the private players such as DeepMind and OpenAI, publish on preprint servers to democratise and share knowledge. But, how useful are these papers for the community at large? Are top AI labs trustworthy? Recently, a Reddit user published a post titled, ‘I don’t really trust papers out of “Top Labs” anymore. In the post, the user asked: Why should the AI community trust these papers published by a handful of corporations and the occasional universities? Why should I trust that your ideas are even any good? I can’t check them; I can’t apply them to my own projects. Citing the research paper titled ‘An Evolutionary Approach to Dynamic Introduction of Tasks in Large-scale Multitask Learning Systems’, the Reddit user said, “It’s 18 pages of talking through this pretty convoluted evolutionary and multitask learning algorithm; it’s pretty interesting, solves a bunch of problems. But two notes. One, the big number they cite as the success metric is 99.43 on CIFAR-10, against a SotA of 99.40. The Reddit user also referred to a chart towards the end of the paper that details how many TPU core-hours were used for just the training regimens that resulted in the final results. “The total is 17,810 core-hours. Let’s assume that for someone who doesn’t work at Google, you’d have to use on-demand pricing of USD3.22 per hour. This means that these trained models cost USD57,348. “Strictly speaking, throwing enough compute at a general enough genetic algorithm will eventually produce arbitrarily good performance, so while you can read this paper and collect interesting ideas about how to use genetic algorithms to accomplish multitask learning by having each new task leverage learned weights from previous tasks by defining modifications to a subset of components of a pre-existing model,” he said. Jathan Sadowski, a senior fellow at Emerging Tech Lab, responded: “AI\/ML research at places like Google and OpenAI is based on spending absurd amounts of money, compute, and electricity to brute force arbitrary improvements. The inequality, the trade-offs, the waste—all for incremental progress toward a bad future.” The Reddit post has been a source of much debate on social media. Many pointed out that there should be a new journal for papers where one can replicate their results in under eight hours on a single GPU. Findings that can’t be replicated are intrinsically less reliable. And the fact that the ML community is maturing towards decent scientific practices instead of anecdotes is a positive sign, said Leon Derczynski, associate professor at IT University of Copenhagen. Replication crisis The replication crisis has been gripping the scientific community for ages. The AI domain is also grappling with it, mostly because researchers often don’t share their source code. A replication crisis refers to when scientific studies are difficult or impossible to reproduce. According to a 2016 Nature survey, more than 70 percent of researchers have tried and failed to reproduce another scientist’s experiments. Further, more than 50 percent of them have failed to reproduce their own experiments. Reproducibility is the basis of quality assurance in science as it enables past findings to be independently verified. The scientific and research community strongly believes that withholding important aspects of studies, especially in domains where larger public good and societal well-being are concerned, does a great disservice. According to the 2020 State of AI report, only 15 percent of AI studies share their code, and industry researchers are often the culprits. The report criticises OpenAI and DeepMind, two of the world’s best AI research labs, for not open sourcing their code. Hypothetical question. Some people have access to GPT-3 and others do not. What happens when we start seeing papers in which GPT-3 is used by non-OpenAI researchers to achieve SOTA results?Here’s the real problem, tho: is OpenAI picking research winners and losers?— Mark Riedl (@mark_riedl) October 3, 2020 In 2020, Google Health published a paper in Nature that described how AI was leveraged to look for signs of breast cancer in medical images. But Google drew flak as it provided little information about its code and how it was tested. Many questioned the viability of the paper, and a group of 31 researchers published another paper in Nature titled ‘Transparency and reproducibility in artificial intelligence’. Benjamin Haibe-Kains, one of the paper’s authors, called Google’s paper an advertisement for cool technology with no practical use. However, things are changing. NeurIPS now asks authors\/researchers to produce a ‘reproducibility checklist’ along with their submissions. This checklist consists of information such as the number of models trained, computing power used, and links to code and datasets. Another initiative called the ‘Papers with Code’ project was started with a mission to create free and open-source ML papers, code and evaluation tables.","excerpt":"Findings that can’t be replicated are intrinsically less reliable.","categories":["AI Features"],"tags":["machine learning research"],"author_name":"Pritam Bordoloi","publish_date":"2022-05-31T17:00:00","publication_year":"2022","word_count":834,"keywords":["machine learning research","Go","API","artificial intelligence","TPU","OpenAI","AI","ML","RAG","Rust","R"],"extracted_tech_keywords":["AI","artificial intelligence","ML","OpenAI","RAG","TPU","R","Go","Rust","API"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/i-dont-really-trust-papers-out-of-top-ai-labs-anymore\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10064601,"title":"Studies show diverse teams work better: Abhishek Kumar and Parmjeet Virdi, Publicis Sapient","content":"Speaking at The Rising 2022, Abhishek Kumar and Parmjeet Virdi of Publicis Sapient discussed how ‘Preparing the inclusive workforce for the future in data & AI’ was vital to have a talent force that was well-rounded and equal. The two chose to demonstrate the theme by appearing on stage together instead of lip service. “Diversity doesn’t mean just women. It is far better when diversity is a part of your actions instead of simply talking about it,” Virdi stated. Kumar is the director of data science, and Virdi is the director of data analytics in the US-based consulting firm. Watch all the recorded sessions of Rising 2022 here>> Virdi explained how diversity differed from inclusion. A company couldn’t be diverse unless they were inclusive. “Just having the intention to be diverse may not solve the purpose,” she said. Virdi began the talk with an anecdote from her own life. “When I went on maternity leave, I was very doubtful if I could ever return to work. I didn’t know if my clients would accept me after a months-long break. I desperately needed a support system. I just wasn’t looking for diversity but also wanted to be included,” she said. Today, most experiences that we have, like using smart products at home or purchasing something, leave a footprint in data. Data, Virdi said, is everywhere, and we are the ones leaving it behind. According to Virdi, the four key ingredients for the success of AI initiatives are data, computation ability, algorithm and talent. All four factors were demonstrated in separate bubbles in a representation. These bubbles represented the scarcity of each ingredient and not their availability. As years passed, even as more algorithms were being developed and procuring data became easier, searching for talent became more difficult. Virdi and Kumar took the responsibility of turning the spotlight on the importance of having a talent force with more inclusivity. There were three parts to improving the talent force of an organisation. These included skills, structure and synergy. Kumar took over next and spoke about how getting people the right training was of utmost importance. And being more inclusive meant making training resources available to people with disabilities. But having the skills was only the first step. The biggest challenge that leaders face is to determine the structure of an organisation in a manner that connects all the moving parts. There needs to be a connecting tissue between the business acumen and the data centre of a company, according to Kumar. “Data is ineffective unless married with business outcomes,” he stated. Some companies, like those in the financial sector, might need more governance, while others might need more collaboration with different teams. The two mentioned initiatives like RISE that Publicis Sapiens had started to push diversity in the organisation. The idea behind initiatives like RISE, which included mentoring female employees, was to change the mindset in the organisation. “We want people to be as they are in the company,” Virdi said. Teams that were more inclusive built more synergy and simply worked together. Virdi revealed, “Studies have shown that while all-male teams were effective at decision-making only 57% of the time, more diverse teams made better decisions up to 87% of the time.” One could have all the elements, but it could still be tricky. “The ingredients may be present, but unless the mix is just right, it won’t be tasty,” Kumar said.","excerpt":"Studies have shown that while all-male teams were effective at decision-making only 57% of the time, more diverse teams made better decisions upto 87% of the time.","categories":["Deep Tech"],"tags":[],"author_name":"Poulomi Chatterjee","publish_date":"2022-04-08T18:40:27","publication_year":"2022","word_count":570,"keywords":["Go","data science","API","programming_languages:R","AI","programming_languages:Go","ViT","analytics","GAN","R"],"extracted_tech_keywords":["AI","data science","analytics","R","Go","API","GAN","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/studies-show-that-more-diverse-teams-work-better-abhishek-kumar-and-parmjeet-virdi-publicis-sapient\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10019850,"title":"Global Chip Crisis Just Got Real","content":"Recently, American MNC Qualcomm said it is struggling to keep up with the dramatic rise in demand for chips. The world’s largest chipmaker was reading out the writing on the wall. The chip shortage has also taken a significant toll on the automobile sector and has now spilled over to other industries. When chipmakers sneeze, the whole world catches a cold. Chip Crisis Qualcomm has posted a 21 percent revenue increase in the last quarter (October-December 2020). The income has grown from $925 million to $2.455 billion, a 165 percent spike. Despite an excellent financial performance, CEO Cristiano Amon said the company doesn’t have the wherewithal to plug the demand-supply gap for chips. The shortage is likely to continue for the next six months. The pandemic has galvanised the demand for computers as work from home has become the new normal. As more and more people started using personal vehicles to avoid public transport, the automobile industry also saw a demand surge. Apple, one of the major customers of Qualcomm for 5G modem, has failed to meet iPhone 12 devices timelines owing to limited availability of components. Amon spoke of the bottleneck issue where many industries depend on a handful of companies for their chip requirements. Like other prominent chip designers, Qualcomm depends on Asian companies such as Taiwan Semiconductor Manufacturing Company (TSMC) and Samsung. These two companies have emerged as the leading resources for the production of advanced semiconductors. Automobile Sector “The semiconductor (microchip) shortage is disrupting automotive production and may delay the recovery of new vehicle sales and profitability in the sector. Carmakers are reducing output and selectively idling plants until the shortage eases, which we expect to take several months,” according to Fitch Ratings. The semiconductor supply crunch has sent the automotive industry into a tizzy. Corroborating Qualcomm’s statement, the Finch report said the high movement of consumer electronics during the pandemic triggered the crisis. The automotive industry’s situation is particularly precarious as the electric vehicle demand is at an all-time high, eating into the reserves of the chipmakers. General Motors (GM) was forced to cut down production at four assembly plants–Fairfax, Kansas; Ingersoll, Ontario; and San Luis Potosi, Mexico. Additionally, the South Korean plant at Bupyeong 2 is running at half its capacity. In the first quarter of 2021, Ford Motor’s vehicle output dropped by 20 percent. Ford spokesperson said the company plans to cut the production of its biggest moneymaker–the F-159 pickup truck. Semiconductor shortage has also impacted companies such as Nissan, Volkswagen, Toyota, Mazda and Subaru. Now, the carmakers are prioritising available chips for their most profitable and best-selling models. Gaming Industry The gaming industry has been another major casualty of the chip crisis. Nintendo, Sony, and Microsoft are all feeling the heat. Speculation is rife that the situation is getting worse for the game hardware industry. Interestingly, some experts opine the gaming industry precipitated the crisis. During the pandemic, the demand for graphics chips for PCs and consoles shot up, catching chipmakers off guard. Wrapping Up Chip Famine occurs from time to time when the demand for semiconductor chips outstrip production. It has happened in 1988, 1994, 2000, 2004, and most recently, in 2011. The shortages were usually the result of a black swan event (2011 Japan earthquake) or rapid technology adoption.","excerpt":"Recently, American MNC Qualcomm said it is struggling to keep up with the dramatic rise in demand for chips. The world’s largest chipmaker was reading out the writing on the wall.   The chip shortage has also taken a significant toll on the automobile sector and has now spilled over to other industries. When chipmakers sneeze, […]","categories":["AI Features"],"tags":["Chip shortage","semiconductor shortage"],"author_name":"Shraddha Goled","publish_date":"2021-02-08T17:00:00","publication_year":"2021","word_count":549,"keywords":["API","TPU","semiconductor shortage","AI","programming_languages:R","Chip shortage","R"],"extracted_tech_keywords":["AI","TPU","R","API","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/global-chip-crisis-just-got-real\/","complexity_score":3,"technical_depth":5,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10055881,"title":"Fujitsu And MIT Create AI That Replicates The Human Brain","content":"With the Center for Brains, Minds and Machines (CBMM) at MIT, Fujitsu Limited has achieved an important milestone in an initiative to improve the accuracy of artificial intelligence (AI) models. It has always been assumed that training the DNN as a single module without splitting it has been the best method to create an AI model with high recognition accuracy. However, researchers at Fujitsu and CBMM have achieved higher recognition accuracy by splitting the DNN into separate modules based on colours, shapes and other attributes of the objects. Today, many AI models are developed enough to demonstrate performance equal to or even better than humans, but recognition accuracy deteriorates when environmental conditions like perspective and lighting significantly differ from those in datasets. The researchers at Fujitsu and CBMM made progress in understanding AI principles that enable the recognition of OOD data with high accuracy. This was done by dividing the DNN into modules, which is a unique approach inspired by the cognitive characteristics of the structure of the human brain. Dr Tomaso Poggio, the Eugene McDermott Professor at the Department of Brain and Cognitive Sciences at MIT and Director of the Center for Brains, Minds and Machines, said, “There is a significant gap between DNNs and humans when evaluated in out-of-distribution conditions, which severely compromises AI applications, especially in terms of their safety and fairness. Research inspired by neuroscience may lead to novel technologies capable of overcoming dataset bias. The results obtained so far in this research program are a good step in this direction.” An AI model using this process was rated as the most accurate in an evaluation measuring image recognition accuracy against the “CLEVR-CoGenT” benchmark. The possible future applications of this model may include AI for monitoring traffic as it can better respond to changes in various observation conditions and a diagnostic medical imaging AI which can correctly recognize different types of lesions. The results of the research will be presented at the Conference on Neural Information Processing Systems, NeurIPS 2021, showing improvements in the accuracy of AI models.","excerpt":"Fujitsu and MIT’s Center for Brains, Minds, and Machines have developed an AI by the DNN into separate modules that increase accuracy","categories":["AI News"],"tags":["AI Models","DNN","Fujitsu","MIT"],"author_name":"Meeta Ramnani","publish_date":"2021-12-14T12:39:22","publication_year":"2021","word_count":342,"keywords":["AI Models","Go","artificial intelligence","programming_languages:R","AI","MIT","image recognition","programming_languages:Go","DNN","Fujitsu","R"],"extracted_tech_keywords":["AI","artificial intelligence","image recognition","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/fujitsu-and-mit-create-ai-that-replicates-the-human-brain\/","complexity_score":3,"technical_depth":7,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":40610,"title":"7 Ways To Upskill Your Employees For The Upcoming Automation Boom","content":"As the landscape of labor evolves, entry-level jobs are at risk of being made obsolete by automation. This puts the focus on companies to upskill their employees to ensure that they stay relevant in a changing career landscape. Upskilling also offers other benefits for both the employee and the employer and is generally a win-win situation. The employee learns more and grows in his domain, whereas the company gains a more valuable employee. There are many dependable methods to upskill your workforce. These have also proven effective in ensuring employee utility and satisfaction. 1) Invest In Online Training Courses Many companies choose to take on online training courses as a dependable way to upskill their employees. An incentivised and subsidised online training program can do wonders for an increasingly obsolete workforce. Employees can be given opportunities to learn more in their own disciplines and develop their existing skills. They can also be encouraged to find different skills to specialise in and enables them to be a more valuable addition to the company. Websites such as Udemy, Coursera and EDX offer specialised packages for companies. They also have a multitude of dependable courses for up-to-date information on a wide variety of subjects. 2) Offer Certificate Courses From L&D Giants Certificate courses are one of the easiest ways to ensure that your employees are upskilled. Partnering with an educational institution to offer certificate courses are a sure-fire way to increase employee engagement and skill levels. Courses relevant to their field of work will allow for an expert-level understanding of the subject. Moreover, prominent training institutions also hold a name in the career space, making the certificates a valuable addition to any employee’s resume. 3) Hop On The Mobile Learning Train Mobile learning is quickly picking up pace as an accessible L&D method. Most employees use their phones outside the workplace as well, making for an all-encompassing and enriching training experience. Mobile learning is also one of the biggest L&D trends of 2019 and is easily accessible for setup. Moreover, engaging with them ‘on-the-go’ will further the learning experience. There are also various trends such as training with chatbots, phone calls, and mobile applications which offer a richer experience for those aiming to comprehensively upskill their employees. 4) Create Individualised Training Approaches If the team is of a smaller size, it is worthwhile to invest the time required to create customised approaches and schedules for every employee. Everyone learns at a different pace, and a specially-structured approach, in conjunction with a focus on their roles, will create a learning environment quickly. Moreover, employees also feel valued by such a move and motivates them to learn and work harder. This will also go a long way in retaining employees post the upskilling phase, as they feel like they are needed in the company. 5) Enable Access To A Mentor One of the best people to learn from is someone who has seen and been in the field for a long time. There are many companies outsourcing mentorship and even more experts in the field who have since retired. A mentor is a dependable approach towards upskilling employees as it brings the element of a human connection into the picture. This creates and fosters a relationship between the mentor and the mentee, with a health information exchange occurring between the two. A mentor will also give employees a way to learn from past mistakes and employ deep industry knowledge in their everyday tasks. 6) Training On The Job Even though all these approaches are equally efficient in ensuring new knowledge transfer to employees, they don’t compare to raw experience. Learning on the job is one of the most organic approaches to training your workforce. By being assigned new roles and responsibilities, employees stand to gain new knowledge faster than any other method. It also equips them with the ability to think on their feet and feel more confident in their potential. 7) Support From Other Employees L&D roles work great for upskilling, but learning can also come from other experienced employees. A peer network not only makes a new employee feel welcome but also encourages improvement within existing ones. Peer networks can also become an important onboarding tool, as new employees can quickly learn company-specific processes and procedures. Requirements for specific L&D roles will also be reduced, fostering a general environment of collaborative progress.","excerpt":"As the landscape of labor evolves, entry-level jobs are at risk of being made obsolete by automation. This puts the focus on companies to upskill their employees to ensure that they stay relevant in a changing career landscape. Upskilling also offers other benefits for both the employee and the employer and is generally a win-win […]","categories":["AI Trends"],"tags":["Data Science","upskill"],"author_name":"Anirudh VK","publish_date":"2019-06-12T09:16:28","publication_year":"2019","word_count":729,"keywords":["Go","programming_languages:R","AI","chatbots","programming_languages:Go","RAG","automation","Aim","upskill","GAN","Data Science","R"],"extracted_tech_keywords":["AI","Aim","RAG","chatbots","R","Go","GAN","automation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/7-ways-to-upskill-your-employees-for-the-upcoming-automation-boom\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10085900,"title":"UMANG: The Friendly Neighbourhood Super-app Breaking Barriers","content":"In 2017, the Government of India launched a mobile app, UMANG, or Unified Mobile Application for New-Age Governance, to bring central, state and local government services under a single platform. Developed in collaboration with the Ministry of Electronics and Information Technology (MeitY) and the National e-Governance Division (NeGD), the app was intended to drive mobile governance by offering a one-stop for services related to education, COVID-19 vaccinations, public transport, employment advice, passport applications and other utilities. Main Objectives However, building a platform for the general Indian population was more complex than initially assumed. To reach people on the grassroots level and across a country as diverse as India, the app would have to cross several language and technological barriers. The problem was two-fold: the app would have to have an interface simple enough for the general public to use while being accessible to a multilingual populace. The Government of India partnered with Senseforth.ai, a Bengaluru-based deeptech AI startup, to build a voice-first experience platform that would include an app, website and phone channels. The directions to Senseforth.ai were simple—the citizen experience on the platform had to be as easy as chatting with a friend online about everyday issues. Senseforth.ai built an Intelligent Virtual Assistant which achieved both these objectives. People could interact with the assistant in multiple languages making the platform invaluable in its utility. (Currently, it supports Hindi and English) While there are already several other apps available that offer various government services, with UMANG, citizens can interact with the platform using natural language processing for personalised, intuitive, and instant responses, just like they would receive with a human customer service representative. Because of this and its comprehensive nature, UMANG has become an unparalleled resource for Indian citizens. Since the pandemic lockdown, the necessity to access critical information among people has also grown exponentially. In the thick of COVID-19, not having the right data at the right time can potentially lead to the loss of life. UMANG can direct people to the closest blood banks without having people scramble around in search of help during emergencies. On a macro level, having government services close at hand via UMANG’s voice assistant will result in the large-scale adoption of key government programs and initiatives. In addition, the uniformity will ease processes by a mile. For instance, people can book an appointment to get the COVID vaccine. How Conversational AI solved the problem Conversational AI served as a perfect fit to achieve this goal, and Senseforth.ai had the tech capabilities needed to drive this home. The tech team at the company leveraged their deep tech capabilities to build an intuitive and contextual chat interface giving instant chat responses, just like texting on an instant messaging app like WhatsApp. Senseforth.ai has tamed the vital challenge of quickly offering relevant information to a wide range of organisations, from banks to insurers to governmental administrations. The company’s patented Natural Language Processing (NLP) made the platform more conversational, intuitive and contextual. Consequently, regardless of how comfortable people are with technology, by incorporating AI-enabled conversation interface capabilities, they were able to ensure that citizens across India could easily access all the information they needed at their fingertips. Easy and high-tech The advanced technology used has paid off. One glance at the platform shows the compounded growth that government applications have made in terms of their easy interface, making it comparable to global standards. While it is difficult to make a direct comparison between UMANG and other global super apps without performing a formal benchmarking study, UMANG is expected to reach millions of Indian citizens without taking into consideration their technical knowledge or preferred language. Therefore, it is strongly believed that UMANG will touch more people than any other platform. UMANG is also a definitive step forward for Indian governmental service platforms in moving towards next-generation technology. Moreover, it wouldn’t be a stretch to call the platform a stand-out even in the global market because of its conversational AI tech. Securing citizen data Besides integrating state-of-the-art tech, Senseforth.ai had other important issues like data security. The team explains the measures taken to protect user information. The team at the company took data security very seriously. Data masking techniques were used to protect sensitive information to ensure that even the project developers didn’t see the data. Information security is a key imperative for Senseforth.ai’s vision. The company has several layers to its data security. First, it comprises the assurance of the security of information belonging to the company internally and the information entrusted to the company by its employees, clients and partners. In the journey of continual benchmarking and improvement of information security practices, it was also important that Senseforth.ai aligned with global best practices and standards. For this, they developed and implemented a strong and continually improving Information Security Management System (ISMS) to demonstrate their ability to safeguard the confidentiality, integrity, and availability of information and its associated assets. Senseforth.ai has established the ISMS in accordance with the global Information Security Standard ISO\/IEC 27001:2013. (The ISO\/IEC 27001 is an international standard that provides the specification for an information security management system (ISMS).) The company abides by all the 114+ controls specified by this standard. The firm also strictly complies with the applicable privacy acts and regulations of the countries and regions where they conduct business and offer their solutions, including the European Union (EU) regulation and GDPR.","excerpt":"Conversational AI served as a perfect fit to achieve this goal, and Senseforth.ai had the tech capabilities needed to drive this home.","categories":["AI Features"],"tags":["Mobile App","mobile application"],"author_name":"Poulomi Chatterjee","publish_date":"2023-01-25T19:00:00","publication_year":"2023","word_count":899,"keywords":["Go","Mobile App","programming_languages:R","AI","mobile application","RAG","NLP","ViT","Rust","GAN","R","startup"],"extracted_tech_keywords":["AI","NLP","RAG","R","Go","Rust","GAN","ViT","startup","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/umang-the-friendly-neighbourhood-super-app-breaking-barriers\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":65662,"title":"Why Microsoft Acquired An RPA Company?","content":"Microsoft CEO Satya Nadella disclosed the acquisition of Softomotive, one of the leading providers of Robotic Process Automation (RPA) during the company’s annual developer conference, Microsoft Build 2020. The tech giant from now onwards will be using Softomotive’s Robotic Desktop Automation (RDA) tool, WinAutomation and ProcessRobot that provide a powerful, robust and easy-to-use Windows-based platform for building software robots and an enterprise Robotic Process Automation platform, including enterprise-grade security and controls respectively. At the virtual event, Nadella shared how Microsoft and its customers are addressing developer needs across all industries and how the company is uniquely positioned to help businesses adapt in a remote world. In a blog post, Charles Lamanna, CVP at Microsoft, stated that by involving the products of Softomotive’s desktop automation together with the existing Power Automate capabilities, at an affordable price, the tech giant is aiming to democratise the power of RPA while enabling everyone to create bots to automate manual business processes in the organisations. Before discussing the possibilities that this acquisition opened up for its customers, let’s take a look at what Power Automate is and how it works. Power Automate Power Automate, previously known as Microsoft Flow is a cloud-based automation platform by Microsoft that helps in creating automated workflows between applications and unlocks analogue data with AI, automates UI with the robotic process automation as well as automates cloud applications and databases with built-in connectors. Microsoft Power Automate is used to improve the potency with various advanced features such as AI-based data understanding through the AI Builder, UI-based application automation through the UI flows as well as integrating with applications and databases via 315 built-in API connectors. Last year at Microsoft Ignite event, the tech giant launched the preview of UI flows, which is the robotic process automation (RPA) capability in Microsoft Power Automate. With the addition of the RPA, Power Automate not only helps the organisations to automate millions of processes every day but also automate their legacy apps and manual processes through UI-based automation. The UI flows enable organisations to perform the mentioned tasks: – UI flows has the capability to automate in a single platform across many apps and services that don’t have APIsUI flows scripts can be customised, built, and managed in a secure cloud environmentWith the help of UI flows you can leverage a low-code experience with some specific step-by-step record and playback experienceUI automation can be seamlessly integrated with API-based automation by combining UI flows with regular flows The Possibilities Charles stated the possibilities that this acquisition opened up for the customers as mentioned below- Delivering a comprehensive low-code desktop automation solution with WinAutomation: With WinAutomation, you can now build RPA bots with the existing browser-based authoring app or through a new desktop app.Expanding the out-of-the-box UI automation drivers for commonly used apps and services: With the addition of Softomotive’s set of connectors and applications such as SAP, legacy terminal screens, Java, Citrix, among others, the tech giant is expanding the existing desktop automation experience for the users. Enabling parallel execution and multitasking: Customers can now execute multiple workflows and automation parallelly on the same machine to reduce latency while improving the efficiency of their bots. Offering additional features: With Softomotive, the tech giant will now provide the customer with the support of low-code bot development. Wrapping Up Power Automate is part of a powerful and adaptable business application platform that includes Power Apps, Common Data Service, Dynamics 365, and Office 365. According to Charles, Power Automate together with WinAutomation will provide customers with additional options for RPA desktop authoring so that anyone can build a bot and automate Windows-based tasks. It will also enable RPA connectivity to various new apps and services, including SAP and traditional green-screen terminal applications.","excerpt":"Microsoft CEO Satya Nadella disclosed the acquisition of Softomotive, one of the leading providers of Robotic Process Automation (RPA) during the company’s annual developer conference, Microsoft Build 2020.  The tech giant from now onwards will be using Softomotive’s Robotic Desktop Automation (RDA) tool, WinAutomation and ProcessRobot that provide a powerful, robust and easy-to-use Windows-based platform […]","categories":["Global Tech"],"tags":["RPA","RPA developer"],"author_name":"Ambika Choudhury","publish_date":"2020-05-22T13:00:00","publication_year":"2020","word_count":624,"keywords":["API","RPA developer","AI","RPA","ML","RAG","automation","Aim","ViT","GAN","R","Java"],"extracted_tech_keywords":["AI","ML","Aim","RAG","R","Java","API","GAN","ViT","automation"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/why-microsoft-acquired-a-rpa-company\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10060484,"title":"Good Old-Fashioned Artificial Intelligence and other weird and wonderful AI trivia","content":"The first ultraintelligent machine is the last invention that man needs ever make, provided that the machine is docile enough to tell us how to keep it under control, said Oxford philosopher Nick Bostrom. His book, Superintelligence, is a crystal ball on AI’s timeline and the future of humanity. Inarguably, artificial intelligence has become an integral part of our lives. Here, we look at the AI breakthroughs that precipitated this paradigm shift. 1) Genesis In 1956, John McCarthy, one of the founding fathers of AI, coined the term “artificial intelligence” during the Dartmouth workshop in 1956. The workshop that saw the participation of luminaries like  Marvin Minsky, Nathaniel Rochester and Claude E. Shannon is considered the founding event of AI as a field. 2) Captcha, gotcha In 1950, the British mathematician Alan Turning created a reversed form of the Turing test\/Imitation Game, which is now used as CAPTCHA test to determine if the user is a human or a robot. 3) Man vs machine On May 11, 1997, IBM’s Deep Blue beat world chess champion Garry Kasparov. The computer won two and Kasparov won one match and three games ended in a draw. Deep blue used Good Old-Fashioned Artificial Intelligence (GOFAI) and processed almost 200 million options per second to achieve this feat. 4) Director’s hat An AI named Benjamin created a movie called “Zone Out” in 2018 starring Thomas Middleditch. The AI created the movie from scratch in 48 hours, where Benjamin was the storyteller, editor and selector of music all rolled into one. Director Oscar Sharp and Ross Goodwin, a creative technologist at Google, came up with the idea. Benjamin is a long short-term memory recurrent neural network. 5) Tryst with Kismet Dr Cynthia Breazeal developed at Massachusetts Institute of Technology in the 1990s as part of an experiment in affective computing, a machine that can recognise and simulate emotions. Kismet contained input devices that gave it auditory, visual and proprioceptive abilities. The robot stimulated emotions through various facial expressions, vocalisation and movement. Facial expressions were created through ears, eyebrows, eyelids, lips, jaw, and head movements. Kismet is spending her retired life at the MIT Museum. 6) Gender bender UNESCO’s 2019 report, I’d Blush if I Could, spoke about the gender bias and stereotypes embedded in AI-powered voice-assistant applications. UNESCO has developed a guideline on the ethics of Artificial Intelligence (AI). The comprehensive global standard-setting instrument can be a beacon of light to bring some gender balance into AI. 7) Dr Who? AI has been extensively used in the medical field, not just for diagnosis, but also in decision-making and robot-assisted surgeries. For example, at the Stanford University Medical Center, Dr Elsie Gyang Ross and colleagues deployed algorithms to identify patients at risk for a circulatory condition known as peripheral artery disease (PAD). The algorithm forecasted which PAD patients would develop major cardiac and cerebrovascular events. 8) How old are you really? Scientists at Insilico Medicine have built an Aging Clock  based on AI-driven analysis of blood tests from 130,000 people from South Korea, Canada and Eastern Europe. You just have to upload your selfie or recent blood test to find out your biological age. 9) Mrs Robot Saudi Arabia gave citizenship to Sophia, an advanced humanoid robot, in 2017. Sophia was also named the United Nations Development Programme’s first-ever Innovation Champion. The AI that powers Sophia allows it to converse with humans, maintain eye contact, and emulate facial expressions, etc. 10) Better half Bicentennial Man story might come true as the romance between humans and machines is nearing the realm of the possible. According to AI expert Dr David Levy, humans will not just have sex with robots but also marry them a few decades from now. By 2050, human and robot marriages will be made legal, he said.","excerpt":"Benjamin is a long short-term memory recurrent neural network.","categories":["AI Features"],"tags":[],"author_name":"Poornima Nataraj","publish_date":"2022-02-12T17:00:00","publication_year":"2022","word_count":632,"keywords":["Go","artificial intelligence","programming_languages:R","AI","neural network","innovation","programming_languages:Go","llm_models:Claude","R"],"extracted_tech_keywords":["AI","artificial intelligence","neural network","R","Go","innovation","llm_models:Claude","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/good-old-fashioned-artificial-intelligence-and-other-weird-and-wonderful-ai-trivia\/","complexity_score":4,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":52880,"title":"10 Datasets Open-Sourced By Tech Giants In 2019","content":"Open-source projects have become one of the robust ways to enhance the quality of the projects. According to the 2018 Open Source Program Management survey by Linux Foundation, open-source projects are set to be the best practice for organisations in the field of technology, telecom, finance, among others. In one of our articles, we had discussed how tech giants are advocating open-source software as a vehicle of change. In this article, we list down 10 datasets which have been open-sourced by tech giants in 2019. Note: The list is in alphabetical order 1| Coached Conversational Preference Elicitation (CCPE) and Taskmaster-1 By Google In September, Google released two natural language dialogue datasets known as Coached Conversational Preference Elicitation (CCPE) and Taskmaster-1. The CCPE is in English dialogue dataset which consists of 502 English dialogues with 12,000 annotated utterances between a user and an assistant discussing movie preferences in natural language. While the Taskmaster-1 dataset includes 13,215 task-based dialogues comprising of six domains. Know more about this dataset here. 2| CheXpert Dataset By Stanford ML In January, machine learning researchers from Stanford University have released CheXpert (Chest eXpert) which is a large chest radiograph dataset with certainty levels and expert comparison. The dataset contains 224,316 chest radiographs of 65,240 patients labelled for the presence of 14 common chest radiographic observations. Know more about this dataset here. 3| Driving Dataset By Waymo In August, the Alphabet’s autonomous driving subsidiary, Waymo open-sourced a high-quality multimodal sensor dataset for autonomous driving. The dataset is extracted from Waymo self-driving vehicles and covers a wide variety of environments, from dense urban centres to suburban landscapes. It contains 1000 types of different segments where each segment captures 20 seconds of continuous driving, corresponding to 200,000 frames at 10 Hz per sensor. Know more about this dataset here. 4| Deepfake Detection Challenge In September, Facebook, the partnership on AI, with Microsoft and academics from Cornell Tech, MIT, University of Oxford, UC Berkeley, University of Maryland, College Park, and University built the Deepfake Detection Challenge (DFDC). The researchers released four groups of datasets associated with the challenge which are the training set, public validation set, public test set, and private test set. Know more about this dataset here. 5| Diversity in Faces Dataset By IBM In January, the big blue released Diversity in Faces (DiF) dataset which helps in the advancement of the study of fairness and accuracy in facial recognition technology. The dataset provides data of annotations of 1 million human facial images. Using publicly available images from the YFCC-100M creative commons data set, the researchers annotated the faces using 10 well-established and independent coding schemes. Know more about this dataset here. 6| Landmarks-v2 By Google In May, the tech giant released Google-Landmarks-v2 which is one of the largest world-wide landmark recognition datasets. It includes over 5 million images (2x that of the first release) of more than 200 thousand different landmarks. Know more about this dataset here. 7| Level 5 Dataset By Lyft In July, Lyft released a Level 5 dataset. Level 5 is a large-scale dataset featuring the raw sensor camera and LiDAR inputs as perceived by a fleet of multiple, high-end, autonomous vehicles in a bounded geographic area. It includes over 55,000 human-labelled 3D annotated frames, a drivable surface map, and an underlying HD spatial semantic map to contextualize the data. Know more about this dataset here 8| Libri-Light By Facebook AI In December, Facebook AI Research released the Libri-Light dataset which is a collection of spoken English audio suitable for training speech recognition systems under limited or no supervision. It contains over 60K hours of audio which is derived from open-source audiobooks from the LibriVox project. Know more about this dataset here 9| Natural Questions for Question-Answering Systems By Google Released by Google in January, Natural Questions (NQ) large-scale corpus for training and evaluating open-domain question answering systems, and the first to replicate the end-to-end process in which people find answers to questions. The dataset consists of 300,000 naturally occurring questions, along with human-annotated answers from Wikipedia pages, to be used in training QA systems. Know more about this dataset here. 10| Open Images V5 By Google In May, Google open-sourced Open Images v5 dataset which features segmentation masks for 2.8 million object instances in 350 categories. The size of the training set is 2.68 million where the segmentation masks on the training set have been produced by the interactive segmentation process.  Know more about this dataset here.","excerpt":"Open-source projects have become one of the robust ways to enhance the quality of the projects. According to the 2018 Open Source Program Management survey by Linux Foundation, open-source projects are set to be the best practice for organisations in the field of technology, telecom, finance, among others. In one of our articles, we had […]","categories":["AI Trends"],"tags":["data analytics acquisitions","dataset for ml","Datasets","datasets on deep learning","Mergers and Acquisitions","open source datasets"],"author_name":"Ambika Choudhury","publish_date":"2019-12-31T12:01:50","publication_year":"2019","word_count":743,"keywords":["Replicate","Go","Datasets","machine learning","programming_languages:R","AI","Modal","ML","data analytics acquisitions","datasets on deep learning","GAN","AI research","Mergers and Acquisitions","R","open source datasets","dataset for ml"],"extracted_tech_keywords":["AI","machine learning","ML","R","Go","GAN","Replicate","Modal","AI research","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/10-datasets-open-sourced-by-tech-giants-in-2019\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10166862,"title":"2am VC Fund to Back Early Stage Startups in Consumer Tech, AI, Fintech, and SaaS","content":"Venture capital firm 2am VC has launched its second fund, targeting investments in 30 Indian startups across consumer tech, fintech, AI, food & beverage, and global SaaS. The firm will allocate 60% of the fund for initial investments and 40% for follow-on rounds. 2am VC made its debut in 2021, closing a $10 million Fund I. To date, 2am VC has invested in 47 companies, including NEWME, Apna Mart, Bimaplan, and Karbon Card, among others. Founded by Hershel Mehta and Brendan Rogers, the firm previously closed on $10 million. “We know that Indian founders are blessed with many financing options, and we strive to be top-of-mind for founders as an investor of choice,” Mehta said. “We go above and beyond to support founders in being their hyperlocal and hyperglobal partners to introduce key relationships, arrange customers, and offer targeted guidance to support growth and achieve product-market-fit.” The firm’s thesis focuses on the value emerging from a generational shift in Indian consumer behaviour, driven by a young, urban population with rising disposable income. Consumers are increasingly prioritising quality, convenience, and digital experience over cost. This shift is creating demand for tech-first solutions tailored to Indian contexts and for niche, category-defining products in underpenetrated markets. 2am VC, which has offices in Mumbai, Bengaluru, Delhi NCR, and Los Angeles, connects global investors with Indian founders. The fund focuses on early-stage startups, mainly at the pre-seed to seed stage. “We recognise that the next stage of development of Indian startups will be driven by young and first-time Indian founders building companies that are uniquely Indian,” Rogers said.","excerpt":"The fund focuses on early-stage startups, mainly at the pre-seed to seed stage.","categories":["AI News"],"tags":["AI fund","VC","Venture Fund"],"author_name":"Aditi Suresh","publish_date":"2025-03-28T12:28:28","publication_year":"2025","word_count":263,"keywords":["VC","Go","API","Venture Fund","programming_languages:R","AI","venture capital","programming_languages:Go","Git","RAG","AI fund","R","startup"],"extracted_tech_keywords":["AI","RAG","R","Go","Git","API","startup","venture capital","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/2am-vc-fund-to-back-early-stage-startups-in-consumer-tech-ai-fintech-and-saas\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10103924,"title":"How Amazon’s Silicon Innovation Is Instrumental in AWS Success","content":"With a series of announcements showcasing Amazon’s prowess not only in the cloud space but also generative AI, and even taking digs at OpenAI’s security flaws in the process, AWS re:Invent2023 at Las Vegas had a lot to offer. Advancing their AI chip ambitions, the company announced two new AI chips, AWS Graviton4 and AWS Trainium 2, at the event. Almost a decade ago, Amazon realised that in order to consistently improve the cost-effectiveness of tasks, it needed to redefine general-purpose computing for the cloud era, thereby pushing innovation to the silicon level. “In 2018, we became the first major cloud provider to develop our own general compute processors,” said Adam Selipsky, CEO of Amazon Web Services, at the event. Releasing Graviton Power Adam Selipsky at AWS re:Ignite 2023. Source: AWS Youtube AWS Graviton, a family of processors, which was first released in 2018, was designed for cloud computing infrastructure running in Amazon Elastic Compute Cloud (EC2). The fourth version – Graviton4, which was unveiled yesterday, is the most powerful and energy efficient chip built by AWS, with 50% more cores and 75% more memory bandwidth than the previous version Graviton3. Furthermore, Selipsky announced the preview of R8g Instances that is powered by Graviton4 and is part of the memory optimise instance family. R8g Instances are designed to deliver fast performance for large data sets, and are energy efficient for memory intensive workloads. Breaking Barriers In 2020, CEO Andy Jassy, who was then the CEO of AWS, emphasised Amazon’s commitment to advancing the cost-effectiveness of machine learning training by investing in proprietary chips. With these chips, AWS lowered the cost barrier for ML training. Similarly in 2018, the release of Graviton was to break the existing market of processors where Intel was comfortably placed. The lack of hardware options for building data centres and cloud services gave them an advantage. Furthermore, the power efficiency of Arm cores made Graviton well-suited for mobile computing and enterprises with extensive arrays of data centres, especially AWS. Today, Amazon has got 150 Graviton-based instances across the EC2 portfolio and more than 50,000 customers. AWS Had It Planned All Along With GPUs being the indispensable component for AI compute, AWS has strategically placed themselves in the said race, and it wasn’t a recent development. Trainium, a chip built for training machine learning models, and Inferentia, a chip optimised for running inference on those models, were released a few years ago. At the event, Selipsky released the second version of Trainium 2. In 2015, Amazon acquired Annapurna Labs for $350 million and in 2017, AWS launched Graviton, thereby entering the chip race. Chirag Dekate, VP analyst at Gartner, had earlier quoted that Amazon’s true differentiation is in bringing their technical capabilities. “Microsoft does not have Trainium or Inferentia,” he said. The chips were designed to provide accelerated performance and cost-effectiveness for ML training workloads on the cloud platform. This was something that started before ChatGPT rage, but probably gained steam post OpenAI’s chatbot. AWS has already found a number of partners that have utilised their AI chips. Companies  such as Adobe, Deutsche Telekom, Leonardo AI have deployed Inferentia 2 for their generative AI models at scale. Similarly, Trainium has already been used by partners such as Anthropic, Databricks and Ricoh. The use cases lie in the internal search team as well, where the chip has been used to train large, deep learning models. Strategic Partnerships Amazon’s partnership with AI research company Anthropic has been a crucial one to showcase their united power in the generative AI space. By not just investing in the company but hosting Claude models on Amazon’s Bedrock platform, the partnership goes beyond simple compute. Anthropic CEO and co-founder, Dario Amodei said that AWS is the “primary cloud provider for our mission critical workloads,” and that there are three components to their partnership – compute, customer and hardware. On the hardware aspect, Amodei spoke about working to optimise Trainium and Inferentia for their use cases. The biggest chip competitor NVIDIA has also built a strategic partnership with AWS. NVIDIA chief Jensen Huang, also made an appearance at the AWS event and spoke about how AWS was the “world’s first cloud to recognise the importance of GPU accelerated computing.” The act was seen as a reassurance of powerful combat teams, making it a symbiotic partnership favouring both parties. Amazon’s steady growth in silicon innovation has helped shape Amazon’s stance in the current AI market. With crucial partnerships AWS is proving to be an essential part of AI compute power.","excerpt":"Amazon’s foray into AI chips and processors started five years before ChatGPT","categories":["AI Trends"],"tags":["Machine Learning"],"author_name":"Vandana Nair","publish_date":"2023-11-30T16:37:03","publication_year":"2023","word_count":755,"keywords":["Anthropic","ChatGPT","machine learning","OpenAI","AI","ML","Machine Learning","RAG","Ray","deep learning","generative AI"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","generative AI","ChatGPT","OpenAI","Anthropic","Ray","RAG"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/how-amazons-silicon-innovation-is-instrumental-in-aws-success\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10078592,"title":"Revisiting Scope, Opportunities &amp; Challenges of Facial Recognition Tech With Neha Bhargava, Fractal","content":"Facial recognition is one technology that evokes equal amounts of curiosity, excitement, and suspicion. The possibility of extracting even the tiniest bit of information just by analysing one’s face is surely fascinating. Over the years, as the world digitalises rapidly, facial recognition technology has also witnessed massive progress. Analytics India Magazine caught up with Neha Bhargava, a senior data scientist with Fractal, to understand more about the field—the opportunities it brings, the drawbacks, and the prospects. AIM: What is the latest in the field of facial recognition technology? What scope does the immediate future hold for this technology? Neha Bhargava: Facial recognition systems and techniques have been around since the 1960s, and their usage continues to grow manifold with time. The journey started with 2D facial recognition from images and videos, which has improved drastically over the years. In fact, now we have 3D and even thermal facial recognition systems. As per a report, the facial recognition market is poised to be a USD 15 billion industry by 2030. On the analytics side, multiple techniques are emerging on top of face recognition, for example—emotion\/expression detection, skin texture analysis, temperature detection, and face recognition with augmented and virtual reality. Also, in recent years, we have seen a boom in deep learning architectures and computational capabilities. One very promising direction is cloud-based solutions coupled with big data, network connectivity, and security—all these will play major roles in making face recognition technology grow substantially in commercial space. “We started with 2D facial recognition from images and videos, which has improved drastically over the years. In fact, now we have 3D and even thermal facial recognition systems.” The rapid advancement in this field has also widened the scope of its applications. This was especially evident during the pandemic era. Some of the most significant applications of facial recognition in this context were—emerging as an alternative to fingerprint scanning, body temperature scanning through thermal imaging, and contact tracing to contain the spread of the virus. Of late, government agencies all over the world have been offering a major push to the increased usage of this technology. The primary use of this technology has been in their surveillance systems for public space monitoring and crime prevention. In addition, various fraud detection applications are also using this technology for two-factor authentication. AIM: What are the major challenges with facial recognition systems today? Neha Bhargava: The main shortcomings are those concerning reliability, safety, and performance. It is not very difficult to fool the system; hence, more often than not, facial recognition is used as a secondary system. Facial recognition systems are yet to become as reliable as fingerprints, iris recognition, and retina scan in terms of accuracy. Further, the design of such systems could be improved so as to not appear too intrusive, giving the users the confidence to use them. Then, there are ethical and regulations-related challenges. On the ethical side of AI there are some concerns about how deep learning models for face recognition are trained. It has been proved on various occasions that these models are biased and do not work well on women’s and black people’s images. Also, facial recognition systems—like many AI systems—have the black box problem; since it is not very apparent how these systems work, there can be a few apprehensions related to their use. Governments, companies and other stakeholders are trying to bring a few regulations to combat such concerns around privacy and data security. “On the ethical side of AI there are some concerns about how deep learning models for face recognition are trained.” AIM: Does India have any regulations related to facial recognition? Are they enough? Neha Bhargava: Unfortunately, India has no laws when it comes to face recognition, including data collection, processing, and storage of facial data. But such technologies can invoke multiple laws\/rights related to privacy, constitutional, and IT. For example, data privacy is one of the main concerns around this kind of technology—especially when the data falls under the category of sensitive personal data. The Supreme Court of India has already declared the right to privacy as a fundamental right. The Indian IT Act, a kind of mother legislation on the electronic format, also does not say anything about facial recognition. This lack of proper regulation is ironic given that central and state government agencies (e.g. Police, NCRB, Railways, etc.) are very keen on using this technology for different purposes. But on multiple occasions, they had to back off from using it because of resistance. This resistance will not stop until we have detailed legal frameworks for using such technologies. But on a more hopeful side, the government of India very recently mentioned that the new draft of the data protection bill is under work, and it is expected that it will enable efficient data usage with protection while being used by the industry. So, it seems like the government is also motivated to keep pace with emerging technologies. “Unfortunately, India has no laws when it comes to face recognition, including data collection, processing, and storage of facial data. But such technologies can invoke multiple laws\/rights related to privacy, constitutional, and IT.” AIM: There is always a fear of surveillance associated with this technology. What do you have to say about it? Neha Bhargava: Yes, there is fear of surveillance associated with face recognition technology, and honestly, it is not baseless. We have seen real-life examples of its misuse that can infuse distrust. I think this fear can be reduced drastically by bringing more transparency on its usage, with data protection laws in place, and by bringing more awareness among the stakeholders—government agencies, private companies, and the common man. We need a harmonious balance between protecting the fundamental rights of citizens on the one hand and the surveillance needs of India on the other. I am very hopeful that we will find that balance soon.","excerpt":"“We need a harmonious balance between the protection of the fundamental rights of citizens on one hand and the surveillance needs of India on the other,” says Neha Bhargava","categories":["AI Features"],"tags":["Facial Recognition","Fractal Analytics","Interviews and Discussions","surveillance"],"author_name":"Shraddha Goled","publish_date":"2022-11-03T17:43:10","publication_year":"2022","word_count":980,"keywords":["Fractal Analytics","Go","Facial Recognition","AWS","AI","surveillance","R","RAG","Aim","deep learning","analytics","Rust","fraud detection","Interviews and Discussions"],"extracted_tech_keywords":["AI","deep learning","analytics","Aim","RAG","fraud detection","AWS","R","Go","Rust"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/revisiting-scope-opportunities-challenges-of-facial-recognition-tech-with-neha-bhargava-fractal\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":22778,"title":"Talent, Insights Are The Key Tools To Future In New Tech, Says Panel At Times AI Hub Conclave","content":"“Infrastructure is not going to be a problem for organisations working with analytics or big data,” Sandeep Chanda, associate vice president, Technology at Neudesic said last week. Chanda was speaking at a panel discussion titled ‘Artificial Intelligence 2018 and beyond’ at the third edition of Artificial Intelligence Conclave organised by Times AI Hub in association with ASSOCHAM Karnataka Council, NASSCOM Centre of Excellence IoT, and powered by NVIDIA. The event was held last week at IISc Bengaluru. The other panelists for the discussion were Dhaval Doshi, founder at Smarthome NX, Dr Praphul Chandra, founder at Koinearth, Amit Singh, director of IoT at Cognizant Technology Solutions and Avi Patchava, vice president, Data Strategy at Inmobi. The Bengaluru chapter of the event intended to bring together the sharpest minds and noted names in the industry on a common platform to share knowledge and to discuss the practical implications of AI and machine learning for enterprise organisations. The panel discussion also brought out an important issue faced by analytics organisations (especially SMEs) about sharing data. Doshi, the panel moderator posed an interesting question to his panelists: Can the AI and big data situation be an example of David vs Goliath? Is there a way the small players in the business compete with the “Googles and Facebooks of the world”? Among the other questions posed during the interaction revolved around the key reasons why AI, IoT and ML are integral to every industry and business. The Panelists were asked to each present a case study based on their industry verticals. While most of the panelists agreed that cloud infrastructure, AI and ML would rule future technologies, almost all of them were concerned about the lack of talent pool in the country. Talent, insights and appropriate tools — that was the way forward, they said. Ramit Sharma, Inspector General Police Allahabad Range, was one of the keynote speakers at the event and discussed about how the Indian Police is using AI for purposes such as sentiment analysis and resource allocation.","excerpt":"“Infrastructure is not going to be a problem for organisations working with analytics or big data,” Sandeep Chanda, associate vice president, Technology at Neudesic said last week. Chanda was speaking at a panel discussion titled ‘Artificial Intelligence 2018 and beyond’ at the third edition of Artificial Intelligence Conclave organised by Times AI Hub in association […]","categories":["Deep Tech"],"tags":["AI (Artificial Intelligence)","cloud infrastructure","Machine Learning","NASSCOM"],"author_name":"Prajakta Hebbar","publish_date":"2018-03-19T13:54:03","publication_year":"2018","word_count":335,"keywords":["big data","Go","cloud infrastructure","artificial intelligence","machine learning","AI","sentiment analysis","ML","Machine Learning","analytics","GAN","NASSCOM","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","analytics","sentiment analysis","R","Go","big data","GAN"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/talent-insights-are-the-key-tools-to-future-in-new-tech-says-panel-at-times-ai-hub-conclave\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10081717,"title":"The Smart Cube is certified as a Best Firm For Data Scientists","content":"The Smart Cube is certified as the Best Firm For Data Scientists to work for by Analytics India Magazine (AIM) through its workplace recognition programme. The Best Firm For Data Scientists certification surveys a company’s data scientists and analytics employees to identify and recognise organisations with great company cultures. AIM analyses the survey data to gauge the employees’ approval ratings and uncover actionable insights. “We have invested significantly in our analytics practice over the past 12 months – from building our Digital Analytics and Data Engineering capabilities, to up-skilling our data science and AI teams through industry partnerships. Being certified as a ‘Best Firm For Data Scientists’ by Analytics India Magazine demonstrates that our people appreciate our approach and value the work they do for our clients” said Sudeepta Chaudhuri, Chief Data Analytics Officer at The Smart Cube. The analytics industry at AIM faces a talent crunch, and attracting good employees is one of the most pressing challenges that enterprises are facing. Gaurav Kumar, Chief Operating Officer at The Smart Cube also expressed his delight, “We are seeing tremendous demand for our analytics solutions and capabilities, as clients and prospects strive to make better use of data for decision making and to drive efficiencies. I am very proud of the growth in this part of our business and the focus we place on building a great work culture, recognised by Analytics India Magazine today.” The certification by Analytics India Magazine is considered a ‘Gold Standard’ in identifying the best data science workplaces and companies participate in the programme to increase brand awareness and attract talent. Best Firms For Data Scientists is the biggest data science workplace recognition programme in India. To nominate your organisation for the certification, please fill out the form at this link.","excerpt":"The Best Firm For Data Scientists certification surveys a company’s data scientists and analytics employees to identify and recognise organisations with great company culture.","categories":["AI Highlights"],"tags":["best companies for data scientists in india","best companies in india for data science fresher","best companies to work for as data scientist in india","data science companies in india"],"author_name":"AIM Media House","publish_date":"2022-12-07T12:00:00","publication_year":"2022","word_count":295,"keywords":["best companies for data scientists in india","data science companies in india","data science","Go","best companies to work for as data scientist in india","programming_languages:R","AI","Git","best companies in india for data science fresher","Aim","data engineering","analytics","GAN","R"],"extracted_tech_keywords":["AI","data science","analytics","Aim","R","Go","Git","data engineering","GAN","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/the-smart-cube-is-certified-as-a-best-firm-for-data-scientists\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10006209,"title":"According To Report AI Job Searches See A Spike Of 106% In One Year","content":"A recent report by the leading job portal, Indeed, revealed that the job searches in artificial intelligence have risen 106% from June 2019 to June 2020. It also suggested that since the onset of COVID-19, AI-related jobs have seen a 20% increase in the last 6 months alone starting from March 2020. The report also suggested that job openings for AI-related jobs have seen a 28% increase from August 2019 to August 2020, while job searches have seen a 91% spike. AI-related job postings have seen a steady increase in the last two years with a 48% increase between June 2018 and June 2019, and 51% between June 2019 and June 2020. Post-COVID world has witnessed an increased adoption of AI-led technologies to ensure business continuity in the changing times. COVID has resulted in a lot of work being automated which has increased the need for specialised skills and candidates in the domain. At the outset of the pandemic, many companies and job seekers are even pursuing avenues to upskill themselves to acquire new-age skills. “In light of our new reality in a post-pandemic world, the need of the hour is twofold – to reinvent existing tech to adapt to the new norms of social distancing and reimagined physical boundaries, and to drive innovation in the creation of the tech solutions that are now required in the new normal,” said Venkata Machavarapu, Head of Engineering, India and Site Director at Indeed India. According to the report some of the top skills in demand are Python, NLP and TensorFlow, whereas some of the application areas of AI that has seen an increase are healthcare, digital finance, logistics and more. It also suggested that edtech has turned out to be one of the emerging sectors for AI adoption. “Even as businesses work towards regaining momentum in the global economy, it is imperative to ensure that talent is able to keep up with the emerging demands of AI-powered solutions on their skills,” he added.","excerpt":"A recent report by the leading job portal, Indeed, revealed that the job searches in artificial intelligence have risen 106% from June 2019 to June 2020. It also suggested that since the onset of COVID-19, AI-related jobs have seen a 20% increase in the last 6 months alone starting from March 2020. The report also […]","categories":["AI News"],"tags":["AI Certifications","certification for artificial intelligence"],"author_name":"Srishti Deoras","publish_date":"2020-09-07T13:12:57","publication_year":"2020","word_count":331,"keywords":["artificial intelligence","AI","innovation","AI Certifications","Git","NLP","certification for artificial intelligence","Python","ai_frameworks:TensorFlow","programming_languages:Python","TensorFlow","R"],"extracted_tech_keywords":["AI","artificial intelligence","NLP","TensorFlow","Python","R","Git","innovation","ai_frameworks:TensorFlow","programming_languages:Python"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/according-to-report-ai-job-searches-see-a-spike-of-106-in-one-year\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":4511,"title":"India Big Data Week 2014","content":"UNICOM is organizing a series of Conference on Big Data called “India Big data Week 2014”. In this event, we will organize one conference on each day in one city, featuring 7 different locations, exploring the depth and breadth of Big Data across all practices and perspectives. We intend to cover Big Data Analytics in depth this time. Pune, Hyderabad – 25th Jan, Kolkata, Chennai – 28th Jan, Mumbai – 29th Jan, Delhi – 30th Jan, Bangalore – 31st Jan This conference concentrates on practical applications and utilization of big data. It aims to cut through the marketing hype and help participants understand the benefits. Every day, Industry creates 2.5 quintillion bytes of data — so much that 90% of the data in the world today has been created in the last two years alone. This data comes from everywhere: sensors used to gather climate information, posts to social media sites, digital pictures and videos, purchase transaction records, and cell phone GPS signals to name a few. This data is ‘Big data.’ Big Data refers to data sets that are too large to be processed and analyzed by traditional IT technologies. Records and data exist in a myriad of electronic devices from mobile communications to surveillance cameras to emails to web sites to till receipts; it can combine daily news, social media feeds and videos. Retailers can exploit the data to track sales and consumer behaviour, in store and online; health professionals and epidemiologists trying to predict the spread of disease combine data from health services, border agencies and a variety of other sources. Read More… Bigger big data spending IDC predicts spending of more than $14 billion on big data technologies and services or 30% growth year-over-year, “as demand for big data analytics skills continues to outstrip supply.” The cloud will play a bigger role with IDC predicting a race to develop cloud-based platforms capable of streaming data in real time. There will be increased use by enterprises of externally-sourced data and applications and “data brokers will proliferate.” IDC predicts explosive growth in big data analytics services, with the number of providers to triple in three years. 2014 spending on these services will exceed $4.5 billion, growing by 21%.","excerpt":"UNICOM is organizing a series of Conference on Big Data called “India Big data Week 2014”. In this event, we will organize one conference on each day in one city, featuring 7 different locations, exploring the depth and breadth of Big Data across all practices and perspectives. We intend to cover Big Data Analytics in […]","categories":["Deep Tech"],"tags":["Analytics Case Study"],"author_name":"AIM Media House","publish_date":"2014-01-20T06:54:59","publication_year":"2014","word_count":370,"keywords":["big data","programming_languages:R","AI","Git","GAN","Aim","analytics","Analytics Case Study","R"],"extracted_tech_keywords":["AI","analytics","Aim","R","Git","big data","GAN","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/india-big-data-week-2014\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":26941,"title":"Should AI Paraphernalia Be Protected By Intellectual Property Rights?","content":"Legal problems concerning artificial intelligence-related models and systems are always an aspect to worry about in a long run. Law firms are of the opinion that appropriate legal measures are necessary for tech companies when it comes to managing intellectual property (IP), patents and other legal assets associated with AI. In addition, there are IP components such as copyrights for AI-based software or products that go into various other businesses, which can. All in all, if any institution or company or even a person, comes up with newer inventions in AI, they are bound to face legal or security issues one way or the other. Looking on the other side, with technology being indispensable in our lives, issues such as security and privacy are looming over us. Since AI and related fields are making headway, it is often suggested that it should also see rigorous developments for mitigating adversaries in parallel. In this article, we will take a look at how AI is walking on thin ice when it comes to IP assets and security problems. Treading Carefully According to an article by US law firm Jones Day, experts in IP related- law emphasise on how legal protection of AI systems is still an open question. In order to restrict this ‘openness’ they advise that a strong legal strategy is a must in companies to protect AI. As innovations in AI are taking place at a fast pace, they suggest that the legal strategy employed by a company should be more proactive to fight AI infringements. There are a variety of factors that go into making a solid strategy. For instance, consider the trade secrets of a company or an organisation. If AI tech forms a part of the trade secrets,  then necessary resources and support from the company are vital to protecting it from litigation or lawsuits. That is why a full scrutiny and comprehension of the AI scenario is a must before devising a legal strategy. IP assets such as patents are strong for AI but they do not guarantee a foolproof solution. They say, “Patents do not protect data compilations, such as AI training sets, a programmer’s particular expression of source code, or other types of proprietary information that may be competitively advantageous and constitute a trade secret.” The article also mentions copyright protection in AI. Issues such as authorship, copyright registration, AI data copyright and copyrightable code which are evident in AI software are discussed. “Copyright protection in software extends to all of the original expression embodied in the software, but not to its functional aspects, such as algorithms, formatting, logic, or system design,” the experts say. Copyrighting AI, Filing Patents In Indian Context Ownership or copyrighting AI systems in India is slightly on the blurry side of legal acceptance. When viewed from the perspective of the Indian copyright law, 1957 AI has no clearly-defined terms in associating itself as work by an ‘entity’. Generally, the law accepts work created by a person or a group of persons. It cannot be applicable where there is a mere involvement by systems rather than human creativity being brought into the work. This delineation is why sometimes AI-related work or products are not fit for copyrighting. Secondly, when it comes to patents involving AI tech, it may even fall under the category of inventor or natural person itself, according to section 2(y) of the Indian Patents Act,1970. Again, this is ambiguous and inconclusive. While there are other countries that have a similar imposition, they nonetheless seem to have restricted legal flexibility to AI systems. Human Consciousness And AI Law in AI has different meanings in different regions across the globe. Keeping legal aspects aside, another issue to introspect about is AI systems becoming close to having an equivalent human consciousness. Well-known author and mathematician Marcus du Sautoy believes that the rising phenomenon of AI machines and developments has led to a situation of extending protection to them, just like to humans. “It is getting to a point where we might be able to say this thing has a sense of itself, and maybe there is a threshold moment where suddenly this consciousness emerges. And if we understand these things are having a level of consciousness, we might well have to introduce rights. It’s an exciting time,” du Sautoy says. If AI systems do develop a perfect consciousness, they may eventually see getting protection in the near future in a few countries. On the other hand, they are prone to criticism equally in the legal community and may face setbacks. The thin line between AI and humans will no longer matter.","excerpt":"Legal problems concerning artificial intelligence-related models and systems are always an aspect to worry about in a long run. Law firms are of the opinion that appropriate legal measures are necessary for tech companies when it comes to managing intellectual property (IP), patents and other legal assets associated with AI. In addition, there are IP […]","categories":["AI Features"],"tags":["AI Systems","law","legal","patents","robotic inventions"],"author_name":"Abhishek Sharma","publish_date":"2018-08-06T12:56:42","publication_year":"2018","word_count":771,"keywords":["Go","patents","law","artificial intelligence","AWS","AI","cloud_platforms:AWS","innovation","programming_languages:R","robotic inventions","AI Systems","legal","ViT","GAN","R"],"extracted_tech_keywords":["AI","artificial intelligence","AWS","R","Go","GAN","ViT","innovation","cloud_platforms:AWS","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/should-ai-paraphernalia-be-protected-by-intellectual-property-rights\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10136643,"title":"Cypher 2024: Key Highlights (Day 1)","content":"Cypher 2024- India’s biggest AI conference– hosted by AIM Media House– has kicked off and is bigger than ever. The day one of the events witnessed more footfall than we anticipated, close to 1500+ participants and 600+ companies took part in India’s biggest AI conference. Well the highlight of the event was Mohan Das Pai, former CFO and Board Member at Infosys, session with journalist Bhupendra Chaubey Pai discussed the deepening technological partnership between India and the United States, positioning the two countries as the next dominant forces in the global digital landscape. Pai likened Bangalore to San Francisco, noting their shared vibrant tech culture and entrepreneurial spirit that attract innovators from around the world. San Francisco has long been the go-to for founders looking to build the next big thing, known as a hub for tech talent, investment, and trends. But now the Indian city, Bengaluru is quickly emerging as the next major startup hub. However, he does add that Bangalore needs the capital. “We need the competing power and investment capital. Once you do that, Bangalore has 350,000 chip designers, testers, and embedded software professionals—the largest pool of collective design and testing talent in the world.” ( Mohan Das Pai, former CFO and Board Member at Infosys, session with journalist Bhupendra Chaubey) While Pai’s session was engaging, Cypher 2024 began with an exciting presentation by Raja Jamalamadaka, Managing Director at Roche Information Solutions India. He shared insights on how neuroscience can accelerate generative AI adoption in enterprises. He believes understanding human emotions towards technology will play a critical role for enterprises that are looking to scale generative AI adoption. “High touch is all about simplicity, ease, helping people relate their emotions in a particular level, ensuring people understand that they will not lose their jobs, if anything, they will benefit from it,” Jamalamadaka said. ( Raja Jamalamadaka, Managing Director at Roche Information Solutions India) At Cypher 2024, Shalvi Chitkara, COO of Data and AI at Genpact, emphasised that an AI-ready organisation begins by aligning AI with business needs. Chitkara believes that talent is at the heart of AI success. It’s no longer just about data scientists and engineers, everyone in the organisation, from HR to finance, needs to be AI literate. “In the past, AI challenges were left to technical experts,” Chitkara noted. “Today, the entire workforce needs to be AI-literate.” Similarly, Shashank Dubey, co-founder and chief revenue officer at Tredence also took the stage at Cypher 2024 and explained that the key to thriving in the Generative AI era lies in redefining how we perceive job roles. In his talk titled ‘Leadership in the GenAI Era: Navigating the New Human-Machine Paradigm’ he stressed the importance of focusing on the outcomes delivered by skill sets rather than the skills themselves. “It’s important not to define our jobs by the skills we possess. Instead, we should define our roles by the outcomes those skills deliver, and those outcomes are business outcomes,” he said. Other interesting sessions that took place at Cypher 2024 include Sarkar Yadav, founder and CMD at Lexlegis.AI session, where he emphasised the importance of using AI in the legal space. He highlighted the complexity of the country’s legislation, quoting the Income Tax Act in India, which has undergone more amendments than any other law globally. Madhusudan Shekar, Customer CTO at Google Cloud, in his talk, ‘Architecting your Enterprise for AI’ explored the future of enterprise architecture in the age of generative AI and discussed how organisations should adapt their structures to navigate this disruptive change effectively. (  Madhusudan Shekar, Customer CTO at Google Cloud) While these are but a glimpse of Day one from India’s most prominent AI conference, Day two is packed with more interesting sessions, including a fireside chat with Baichung Bhutia, former captain of the Indian football team. He will discuss how quick thinking, changing strategies, and predictive planning—essentials on the football field—can shape leadership in today’s fast-paced tech landscape.","excerpt":"Mohan Das Pai, former CFO and Board Member at Infosys, discussed the deepening technological partnership between India and the United States, positioning the two countries as the next dominant forces in the global digital landscape.","categories":["AI Features"],"tags":["Cypher"],"author_name":"Pritam Bordoloi","publish_date":"2024-09-26T09:34:14","publication_year":"2024","word_count":657,"keywords":["Go","API","GenAI","startup","AI","Git","Aim","generative AI","GAN","R","Cypher"],"extracted_tech_keywords":["AI","generative AI","GenAI","Aim","R","Go","Git","API","GAN","startup"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/cypher-2024-key-highlights-day-1\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":24063,"title":"How Analytics, Automation And Digital-Oriented Deals Helped Wipro Cross $8 Billion Mark","content":"Indian tech giant Wipro posted a strong revenue growth in its Q4 results ending on March 2018, with IT service revenue crossing $8 billion. Abidali Z Neemuchwala, CEO and member of the board at the IT bellwether, said in a statement, “Our investments in digital and our efforts in client mining are paying off very well. Our strong order bookings in the last two quarters provide us the right foundation to grow as we progress through the year.” However, industry analysts have been cautious about the Q4 financial results. That could be because Wipro posted a drop in its net profit with a marginal slide in its revenue growth for the fourth quarter of 2017-18. One of the chief reasons for the drop was the insolvency of two of its clients which affected the performance of the company as a whole. Nonetheless, Neemuchwala is confident about the growth accrued from their digital revenue. “Digital revenue now constitutes 26.7 percent of the total revenues. In Q4, we closed our highest bookings till date on digital deals, and in the last one year, we have won almost 70 deals over $5 million,” he said. While the CEO attributed the growth to digital technologies and broadening of capabilities in areas such as data analytics, cybersecurity operations and intelligent automation, among others, Jatin Dalal, chief financial officer noted in a statement, “We continue to make progress in our client mining, with number of clients contributing revenues over $75 million increasing from 17 to 20 during the quarter. Our investments in Wipro Holmes and our automation suite are have resulted in consistent productivity improvements. We generated robust Operating Cash Flows of ₹84 billion at 105 percent of our Net Income in FY18.” Greatest Demand Coming From Analytics, Automation, Digital-Oriented Deals: 1) Multi-Year Deal From Global Financial Services Firm: Talking about deals, the statement reveals that the company continues its momentum in winning large deals. The financial statement reveals that Wipro has won a multi-year infrastructure transformation deal in the securities and capital markets domain from a global financial services firm. The deal comprises of hybrid cloud management services, workplace services, cybersecurity operations, and cloud-enabled transformation; and will standardise and modernise IT operations across the client’s lines of business. In addition to leveraging Wipro’s expertise in infrastructure operations transformation, the engagement will be powered by Wipro Holmes for intelligent automation, and analytics-driven insights. 2) Multi-Year contract For German Automotive Firm For Advanced Analytics Platform: Wipro has won a multi-year contract from a leading German automotive company to transform and operate a large advanced analytics platform. The program will foster business agility and scalability for the client through transformation of enterprise operations and data-driven business insights. How Wipro Sharpened Focus On Analytics With Data Discovery Platform Earlier last year, Wipro announced the availability of its Data Discovery Platform (DDP) — an integrated platform that captures and manages data to generate actionable insights through advanced analytics. DDP essentially consolidates siloed data (structured, unstructured, external, internal) of an enterprise into a single integrated platform and uses agile methodology with short sprint durations to deliver quick insights. While doing this, it insulates the analytics platform from operational systems so that performance does not take a hit. The platform leverages Microsoft’s Cortana Intelligence Suite that includes machine learning capabilities, power BI, data lake analytics and stream analytics to build analytical applications. The key differentiator of the platform is that it simplifies the process, optimises costs and transforms business. The use cases range from an Australian banking major to a leading airline operator in Middle East and a pharma major in Europe. What’s also interesting about it is Wipro’s strategic partnership with Microsoft to deliver analytics-as-a-service. The DDP leverages Microsoft Azure’s AI capabilities to build business specific applications. The company in partnership with Microsoft also launched industry specific applications through the Azure platform The Data Discovery Platform allows organisations to kickstart the analytics journey with value added services and bridge the insights gap insights","excerpt":"Indian tech giant Wipro posted a strong revenue growth in its Q4 results ending on March 2018, with IT service revenue crossing $8 billion. Abidali Z Neemuchwala, CEO and member of the board at the IT bellwether, said in a statement, “Our investments in digital and our efforts in client mining are paying off very […]","categories":["IT Services"],"tags":["Agile Methodology in Data Analytics","automotive analytics","Wipro"],"author_name":"Richa Bhatia","publish_date":"2018-04-27T07:20:59","publication_year":"2018","word_count":662,"keywords":["Wipro","API","machine learning","AI","Agile Methodology in Data Analytics","R","Scala","Git","RAG","automotive analytics","analytics","Azure","data lake"],"extracted_tech_keywords":["AI","machine learning","analytics","RAG","Azure","R","Scala","Git","API","data lake"],"url":"https:\/\/analyticsindiamag.com\/it-services\/how-analytics-automation-and-digital-oriented-deals-helped-wipro-cross-8-billion-mark\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":52278,"title":"Top 5 Algorithm Breakthroughs In 2019","content":"In order to make machines more human-like in nature, researchers from enterprises as well as academia have been striving hard to develop intelligent systems which have the ability to make their own decisions. This year we have witnessed a number of interesting developments in the domain of artificial intelligence and machine learning. With a lot of developments going around and the constant digitisation, this year AI has proved to be more intelligent as well as complex than humans. For instance, beating the champions of various games like StarCraft, poker, go, among others. We see a number of developments in the domain of deep learning using algorithms and techniques like reinforcement learning (RNN), convolutional neural networks (CNN), generative adversarial network (GAN), etc. In this article, we list down the top 5 algorithms which made a breakthrough in 2019. 1| Multi-Agent Learning Algorithm This year we have witnessed several projects which used the multi-agent learning algorithms. For instance, Starcraft can be said as one of the most challenging real-time strategy games which is based on a multi-agent learning algorithm. It plays the full game of StarCraft II by using a deep neural network which is trained directly from raw game data by supervised learning and reinforcement learning algorithms. In October, DeepMind announced that Alphastar has reached the grandmaster level and is beating 99.8% of humans in the real-time strategy game. During the early months of this year, Google’s DeepMind introduced AlphaStar  which is a Starcraft II AI program that beats the top professional player. Also, a few months back, the Facebook AI Research with Carnegie Mellon University developed an AI bot called Pluribus which has the capability of defeating human champions. This bot has beaten more than two players. Pluribus incorporates a new online search algorithm that can efficiently evaluate its options by searching just a few moves ahead rather than only to the end of the game. It is based on a form of counterfactual regret minimisation (CFR) which is an iterative self-play algorithm. 2| Neural Machine Translation One of the most important domains for this year is the neural machine translation. Techniques like transfer learning, text-to-text, speech-to-text, text-to-speech, etc. have been thoroughly practised by the researchers in order to develop intelligent machines. In the present scenario, almost all the top-level organisations and academia have been working and developing systems in this domain. In August, the Facebook AI models achieved first place in several language tasks included in this year’s annual news translation competition. The models, along with the cross-lingual pretraining and self-supervised learning for other modalities, used large-scale sampled back-translation, noisy channel modeling, and data-cleaning techniques to achieve the highest performance for translating from English to German, German to English, English to Russian, and Russian to English. Also, the researchers at tech giant Google recently built a more enhanced system for neural machine translation (NMT) which has the capability to handle more than 100 languages. The researchers at the tech giant claimed that this model is the largest multilingual NMT system to date in terms of the amount of training data and the number of languages considered. 3| Adversarial Learning GAN plays a major role and has been at the forefront of research when it comes to deep fakes in deep learning techniques. There has been a lot of research in GAN for a few years now. This year we saw several interesting GAN models. For instance, GauGan, a deep learning model developed by NVIDIA Research allows users to draw their own segmentation maps and manipulate the scene, labelling each segment with labels like sand, sky, sea or snow. The model is developed by using the PyTorch deep learning framework. Also, this year at the AWS re:Invent, the developers announced a machine-learning driven keyboard known as DeepComposer. AWS DeepComposer is a 32-key, 2-octave keyboard designed for developers to get hands-on with Generative AI. The keyboard is developed with an aim to help the developers learn machine learning. One can get started with GAN without any prior knowledge of the domain. The GAN includes two different neural networks against each other in order to compose new and original digital outputs. 4| Object Detection The first-ever image of the black hole which was witnessed in April was generated using a machine learning algorithm known as CHIRP which stands for Continuous High-resolution Image Reconstruction using Patch priors. 5| Combinatorial Optimisation This year, Toshiba made a major breakthrough in combinatorial optimisation by using the Simulated Bifurcation Algorithm. Combinatorial optimisation is the selection of the best solutions from among a huge number of combinatorial patterns. The technique used the Simulated Bifurcation Algorithm which helps in quickly obtaining highly accurate approximate solutions for complex large-scale combinatorial optimisation problems.","excerpt":"In order to make machines more human-like in nature, researchers from enterprises as well as academia have been striving hard to develop intelligent systems which have the ability to make their own decisions. This year we have witnessed a number of interesting developments in the domain of artificial intelligence and machine learning.  With a lot […]","categories":["AI Trends"],"tags":[],"author_name":"Ambika Choudhury","publish_date":"2019-12-19T14:00:00","publication_year":"2019","word_count":782,"keywords":["machine learning","artificial intelligence","AWS","AI","neural network","PyTorch","Aim","deep learning","object detection","generative AI"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","deep learning","neural network","generative AI","Aim","PyTorch","object detection","AWS"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-5-algorithm-breakthroughs-in-2019\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10017306,"title":"Why Should India Worry About Solorigate Style Cyber Attacks","content":"“India’s cybersecurity is ranked very low globally and even the budgetary allocations are below par.” A well-planned cyber attack can cause power outages, air traffic chaos and can even shut down a nuclear reactor. Modern global warfare has taken an uglier turn in this information age. The malicious players sit far away from their victims and administer attacks and subdue them almost effortlessly. Cyberspace has emerged as a new playground for espionage. All one needs is a few dedicated hackers who can crawl for information stored on the enemy’s servers. Today, almost every organisation at some stage of technology adoption, including expanding its footprint on digital delivery channels. Be it core banking solutions, or digital delivery channels such as internet banking or mobile banking; there has been an unprecedented push for nationwide financial inclusion. COVID-19 worked as a catalyst by bringing more players into the market, making cyber attacks a commonplace. In an interview with Analytics India Magazine, Sudeep Das of IBM Security Systems warned that the cybercriminals are capitalising the opportunity [pandemic] to launch novel attacks. They are using tactics from phishing campaigns and malicious domains to targeted malware and ransomware. Last month Microsoft published a report about an ongoing investigation on Solorigate that rattled global cybersecurity systems globally. In the report, the companies said that the hackers could view source code in several source code repositories. The Solarwinds hack was so well-orchestrated that the hackers even managed to penetrate the secured servers of intelligence services too. “There’s definitely a place for dark web monitoring, but most organisations don’t have the maturity level where they’re getting a good return on that investment.”Microsoft According to estimates, the hackers sent its probes only into a few dozen of the 18,000 government and private networks they gained access via management software provider SolarWinds. The attack breached multiple layers of the supply chain to gain access to as many as 250 networks. Telecommunications, energy, financial networks, transportation systems are usually more prone to cyberattacks. According to the Data Security Council of India (DSCI), cyberattacks can be carried out in a number of ways: Computer-network Supply-chain Social-networkingRadio networks for GPS Financial networks too are a major target. With rapid digitisation, attackers can put nations in jeopardy by breaching banks, stock exchanges, trading, online payment systems, and other transactions. The global Cybersecurity Index for 2018 ranks India at 47th place globally. India is the third-largest consumer of internet services only next to the US and China, and it is only obvious that Indian institutions and individuals are equally vulnerable to such major attacks. So, how well is India prepared? Cybersecurity: An Indian Perspective Average time to detect a security breach(Source: Accenture) A look at a few popular recent attacks on Indian cyberspace: In July 2016, a phishing email sent to an employee at Union Bank of India, hackers accessed the credentials to execute a fund transfer, unsuccessfully swindling Union Bank of India of $171 million. The infamous Wannacry Ransomware attack even impacted servers in Indian states such as Andhra Pradesh and West Bengal. Popular online grocery service provider, Bigbasket faced a data breach by a group named “Shinyhunters” who reportedly have compromised the data of more than 2 million users. F&B giant, Haldiram’s were demanded $750,000 as part of a ransomware attack In May 2020, it was reported that data of 40 million Truecaller Indian users was reportedly put out for sale on the dark web. Rising cyber threats after COVID-19, observes a Deloitte survey, pose serious concerns, especially for Indian banks. Globally, cyberattacks against financial institutions increased 238% during the pandemic between February 2020 and April 2020. Citing IBM’s data breach report, Das revealed that the Indian companies incurred an average of ₹140 million total costs of a data breach in 2020, an increase of 9.4% from 2019. The top 3 root causes of data breaches were categorised as 53% malicious attack, 26% system glitch and 21% human error. Most equipment and technology for setting up cybersecurity infrastructure in India are currently procured from global sources. So, these systems are vulnerable to cyber threats, just like any other connected system. There is a great need for a robust Cyber Security Strategy in India that can enable resilience by establishing critical information infrastructure and crisis management. Countries which flaunt cutting edge cybersecurity usually are breeding grounds for startups. The government bodies can only go so far. Similarly, India too needs to nurture an ecosystem that incentivises more startup players to get into cybersecurity. (Source: Global Cybersecurity Index 2018) Andhra Pradesh, Telangana and Haryana have earmarked separate budgets for cybersecurity (e.g. 10% of state government IT spends) on setting up operation centres and other tasks. This is on par with Singapore that announced 8% to 10% of its IT budget on cybersecurity in line with similar practices in Korea (10%) and Israel (8%). Here are a few recommendations from the survey: India needs a Cyber Readiness Index to benchmark security practices in all states and UTs.India should play an active role in global cybersecurity dialogue.Additional budgetary allocations for states to account for cybersecurity. ITU projects that there will be 70% Internet penetration by 2023 and the projected cybercrime cost will be an estimated $2 trillion. Fortunately, over the last decade, artificial intelligence and machine learning have emerged as key solutions to understand and predict the nature of these attacks. Modern Day Solutions For Modern Day Problems According to a report by Accenture, organisations spend more than 10 % of their IT budgets on cybersecurity programs on an average. And, 84% of organisations spend more than 20% of their cybersecurity budgets on tools that leverage AI, machine learning and RPA. Three years ago, only 41% of industry leaders were spending more than 20 per cent of their cybersecurity budgets on advanced technologies. Today that has doubled, to 82%. For instance, Darktrace’s AI products are smoking out the dormant hackers who ride below the radar and evade detection. Founded in 2014, Darktrace, a global AI company for cyber defence, has been leveraging AI that is modelled on human immune systems — a digital antibody. Darktrace’s solution takes action against in-progress cyber threats, limiting damage and stopping their spread in real-time. The digital antibody intelligently generated measured and proportionate responses in case of a threat without impacting normal business operations. “Global ransomware damage costs are predicted to hit $20 billion in 2021, up from $11.5 billion in 2019, $5 billion in 2017, and just $325 million in 2015.”Cybersecurity Ventures Experts believe that machine learning can bridge the gap between automated threat detection and a security team’s response. A neural network can be trained on millions of legitimate and malicious files to predict and prevent future malware activity. So far, standard techniques like natural language processing (NLP)for text analysis along with regression models have come in handy. ML models are getting better with time, but the nature of attacks also evolve with time. Also Read: Top 8 Machine Learning Tools For Cybersecurity A successful malware attack can morph itself with the files in the system, making it undetectable even for state-of-the-art pattern recognition machine learning models. Furthermore, the attackers can deploy their own machine learning models that can trick the already in place systems into chasing their own tails. For more resources for ML solutions in cybersecurity, click here.","excerpt":"“India’s cybersecurity is ranked very low globally and even the budgetary allocations are below par.” A well-planned cyber attack can cause power outages, air traffic chaos and can even shut down a nuclear reactor. Modern global warfare has taken an uglier turn in this information age. The malicious players sit far away from their victims […]","categories":["Deep Tech"],"tags":["supply and demand in cyber security"],"author_name":"Ram Sagar","publish_date":"2021-01-07T15:00:00","publication_year":"2021","word_count":1219,"keywords":["machine learning","artificial intelligence","AI","neural network","ML","supply and demand in cyber security","NLP","RAG","analytics","Gradio","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","neural network","NLP","analytics","Gradio","RAG","R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/india-solorigate-cyber-attacks\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":29043,"title":"Indian-Origin Researchers Develop AI That Can Predict Alzheimer’s Risk 5 Years In Advance","content":"With an ageing global population, the prevalence of Alzheimer’s disease is rapidly increasing. It has created a heavy burden on public healthcare systems — especially in developing countries. It, therefore, becomes very critical to identify persons who are most likely to decline towards Alzheimer’s disease, in an effort to implement preventative treatments and interventions. Now, researchers from the University of Toronto have designed an algorithm that learns signatures from magnetic resonance imaging (MRI), genetics, and clinical data. The research titled Modeling And Prediction Of Clinical Symptom Trajectories In Alzheimer’s Disease Using Longitudinal Data uses artificial intelligence algorithm which can accurately predict whether a person’s cognitive decline will lead to Alzheimer’s disease in the next five years or not. The research by Indian-origin scientists Nikhil Bhagwat, M Mallar Chakravarty, and Joseph D Viviano and Aristotle N Voineskos, talks about their AI methodology could work symbiotically as a “doctor’s assistant”. Chakravarty, who is an assistant professor at the McGill University, Canada, told a newswire, “…This would help stream people onto the right pathway for treatment. For example, one could even initiate lifestyle changes that may delay the beginning stages of Alzheimer’s or even prevent it altogether.” Reportedly, the researchers trained their algorithms using data from more than 800 people ranging from normal healthy seniors to those experiencing mild cognitive impairment, and Alzheimer’s disease patients. Alzheimer’s is the most common form of dementia, a general term for memory loss and other cognitive abilities serious enough to interfere with daily life. Alzheimer’s disease accounts for 60 to 80 percent of dementia cases. The greatest known risk factor is increasing age, and the majority of people with Alzheimer’s are 65 and older. Unlike common perception, Alzheimer’s is not just a disease of old age. Over 200,000 Americans under the age of 65 have younger-onset of Alzheimer’s disease","excerpt":"With an ageing global population, the prevalence of Alzheimer’s disease is rapidly increasing. It has created a heavy burden on public healthcare systems — especially in developing countries. It, therefore, becomes very critical to identify persons who are most likely to decline towards Alzheimer’s disease, in an effort to implement preventative treatments and interventions. Now, […]","categories":["AI News"],"tags":["AI (Artificial Intelligence)"],"author_name":"Prajakta Hebbar","publish_date":"2018-10-09T09:40:56","publication_year":"2018","word_count":302,"keywords":["Go","API","artificial intelligence","programming_languages:R","AI","programming_languages:Go","Git","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","R","Go","Git","API","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/artificial-intelligence-alzheimers-disease-research\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10050998,"title":"Karnataka To Set Up CoE For Artificial Intelligence In Hubballi","content":"The minister for IT\/BT, Science and Technology, and Higher Education, Dr C.N. Ashwath Narayan, recently announced that a centre of excellence (CoE) for Artificial Intelligence (AI) and Data Engineering will be set up soon in Hubballi. Narayan said that the setting up of CoE is part of the Karnataka Digital Economy Mission, speaking at a summit on ‘Innovation and Impact @Hubballi’ held under the Beyond Bengaluru programme. He promised that a park would be developed at Hubballi to facilitate graduating startups to grow even further. “The government has brought in effective policies to promote innovative technologies, and the duration of the internship has risen from 3 weeks to 30 weeks. Students will also be given the opportunity to undergo internships in foreign countries,” Narayan said. Also speaking at the event, Chief Minister Basavaraj Bommai emphasised the need for ‘technology schools’ at the high school level to spread out its prowess in the sector to other parts of the state. Bommai said, “There was a technology school in Hubli way back in 1960 which was closed later. But, technology schools need to be started considering the fact that perception ability will be more during childhood.” Karnataka is trying to be the first in the country to implement the National Education Policy (NEP). The NEP-2020 aspires to prepare the students as per the global technological developments, and accordingly, coding will be taught at the school level itself. The Karnataka government was forced to restart physical classes for school children after most of them lost a year without access to the internet or other devices to benefit from online education.","excerpt":"The setting up of CoE is part of the Karnataka Digital Economy Mission, as announced at summit on ‘Innovation and Impact @Hubballi’.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","Bengaluru","Data Science","Data Scientist","Deep Learning","government of india","Internships","karnataka","Machine Learning","Python"],"author_name":"Victor Dey","publish_date":"2021-10-07T13:15:52","publication_year":"2021","word_count":267,"keywords":["government of india","Internships","Git","Ray","R","karnataka","artificial intelligence","Data Science","startup","Go","AI","Machine Learning","Bengaluru","programming_languages:R","innovation","Python","data engineering","Deep Learning","Data Scientist","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","Ray","R","Go","Git","data engineering","innovation","startup","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/karnataka-to-set-up-coe-for-artificial-intelligence-in-hubballi\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10070078,"title":"Council Post: Verticalisation of AI platforms – how to move from horizontal","content":"There has been a surge in demand for AI, clearly visible in enterprises and small firms alike. To keep up with the ever-changing technological landscape, regulatory and legal environment and customer demands, enterprises are devising new ways to scale their AI projects, especially in mission-critical functions. Yet, before enterprises formulate ideas or experiment with AI, they need to map the use case they are looking to solve and measure the solution against this use case. There are multiple ways to deploy AI, and if a particular strategy provides efficient implementation and results, only then should it be implemented. Interestingly, the adoption of AI has primarily been driven by horizontal or general-purpose platforms so far. A newer strategy, ‘Vertical AI’ is now driving industry-specific deployment. Let us first understand the difference between both Horizontal platforms are general-purpose platforms designed for everyone and any use case, regardless of the industry– any user can come on-board and start using the platform. Horizontal AI platforms enable data scientists to complete the data science workflow more efficiently. The customers of horizontal platforms are usually developers, data scientists, etc. Some examples include Watson, Sagemaker, Google Cloud AI Platform and others. These platforms, however, create several challenges for the stakeholders because of their complex nature and ‘one size fits all’ approach. Unfortunately, most businesses today require data science experts to use horizontal AI platforms and need additional solution layers to make these models ‘usable solutions’. This includes human in the loop processes, feedback labeling, auditing, tool-specific pipelines etc. According to Gartner, 85 percent of AI solutions by vendors will focus on concrete domains and industry verticals. This is where a vertical AI platform comes in as an advantage. A vertical AI platform owns the entire workflow to solve a targeted customer’s need or use cases. Also referred to as Industry-specific platforms, they are specialized for a specific vertical with all relevant vertical-specific requirements instead of having a generalised problem-solving approach and then applying it to industry-specific problems. These platforms are developed to provide a complete product from end to end, from understanding the business needs to monitoring product performance. Prima facie, an industry-specific platform for the BFSI industry appears to be a step in the right direction. The question is, why should they invest in it? Why invest in a vertical AI platform? With the rise in investments and continuing adoption of AI across industries, the solutions are becoming more business-driven and vertical-specific, leading to more specific AI market segments. There are many industry-specific hurdles, such as regulatory and compliance frameworks, data privacy requirements, audit and control requirements, etc. which is especially true for an industry like BFSI. Domain-specific AI platforms and offerings accelerate the AI adoption in BFSI industry and shorten the path to production by reducing the development time and associated risks in AI roll out, including industry specific use cases like Underwriting, Risk Monitoring, Customer personalization, Product recommendation, Persistency modelling etc. Moreover, this will also help bring together AI expertise in the industry, pushing the compliance efforts and figuring out common approaches to ethics, trust, and reputational concerns. As per Gartner, here are the top three AI application types currently in use (in percentage) and the most common AI applications. Sector In use   AI applications BFSI37%Fraud analysis on transactional data Chatbots Market\/consumer segmentation Manufacturing 40%Chatbots Smart robotics (automated) Warehousing manufacturing Process optimisation Government 26%Chatbots Computer-assisted diagnostics Fraud analysis on transactional data (Source: Gartner) Why does BFSI need an industry-specific AI platform? Processes in BFSI are complex, multi-faceted and highly regulated with many mission-critical functions, hence investing heavily in Risk management, Data control, and Infrastructure. Large-scale adoption of AI in such an industry needs more than simple model building tools. There are various considerations on approach and strategy that multiple teams focus on. BFSI Industry: Is susceptible on all things dataDiscourages Black box modelsNeeds advanced risk and audit controlsRequires balance between regulator\/risk requirements and protection of consumersBusiness ownership is vital for AI products Designed to handle a broad-range of use cases, horizontal platforms won’t ‘own’ a specific use case. Hence, they cannot perform as well for each one individually, creating associated risks. Moreover, business users, the end-consumers of AI decisions, are left in a fix while applying technical, non-domain-specific horizontal solutions to their use-cases. In an industry handling sensitive data and strict regulatory requirements, it is essential to have governance, controls and trust in AI processes, while also offering transparency to customers. AI knowledge and experiences should be shared across teams, creating a governance and control structure. Vertical platforms enable Data science teams to handle the building blocks of AI acquisition and then hand over the reins to business teams for further implementation. Internal efficiencies are then driven through collaboration, actionable insights and increased controls. How is the industry solving it today? AI in BFSI is largely fragmented with no clear market leader. Like in any adoption cycle, early adoption of AI is driven by services (internal or procured) and point focused solutions. Services driven: Upon the requirement, tech companies or internal teams use horizontal platforms and build user-specific requirements.Point focused solutions: Few use case specific solutions are offering point focused solutions that match or nearly match the requirements of the industry users. General purpose tools often underwhelm the expectations of the user, and while point focused AI solutions offer the speed and cost advantage, but being point focused, they are rigid, and it becomes exponentially chaotic to manage multiple solutions within an organization. Hence, they take a complex, expensive and chaotic journey to deploy AI. Vertical AI is poised to fill this significant gap left by general purpose and horizontal platforms. Vendors building vertical offerings There are a lot of AI vendors offering platforms with a niche focus on specific industries like Healthcare, Insurance, Financial services, Autonomous Cars etc. We are seeing various horizontal AI vendors joining the league too, by building on top of their own platforms, to narrow their focus on specific verticals or use cases. An AI platform designed for the Healthcare industry can detect anomalies in x-rays or images, to diagnose a disease, which otherwise would have gone undetected by a doctor. In the BFSI sector, Arya.ai, India’s first deep learning startup founded in 2013, offers an industry specific AI platform coupled with off-the-shelf modules for BFSI, providing the flexibility of a platform along with scalability of point-focused solutions. Modular functionality of the product stack offers complete flexibility for financial Institutions to pick the use case of choice and upscale the modules as required. Given the regulatory requirements and operational risk, Arya.ai offers a comprehensive AI Governance framework called AryaXAI to contextualize ML observability to address industry specific needs like – Model Audit, User controls, Explainability and traceability. Pega is another startup from Massachusetts, US provides a banking specific platform to simplify banking processes such as operations, customer onboarding and KYC, customer service, etc. to provide seamless experiences across channels. Zest.ai, a US based startup had verticalized AI offering to underwriting where the users can even build their own credit underwriting models on the platform. In the industrial sector, vertical platforms can be used for predictive maintenance of machinery. C3.ai is one such company, whose AI platform aggregates volumes of data from various sources, to determine how likely a machine is going to fail. Companies in the transportation, manufacturing, oil and gas, telecommunications, defense, etc. are using C3.ai’s platform to optimize their businesses. A vertical AI strategy provides the opportunity to enterprises to have a laser-focus on a particular domain or a use case, adapting well to the customer’s need as compared to horizontal platforms. Vertical AI is here to stay, truly enabling enterprise-wide AI adoption and scalability. This trend will provide interesting solutions for various use cases, and democratize AI experimentation across industries. This article is written by a member of the AIM Leaders Council. AIM Leaders Council is an invitation-only forum of senior executives in the Data Science and Analytics industry. To check if you are eligible for a membership, please fill out the form here.","excerpt":"A vertical AI strategy provides the opportunity to enterprises to have a laser-focus on a particular domain or a use case, adapting well to the customer’s need as compared to horizontal platforms.","categories":["AI Features"],"tags":["AI Applications","AI Software"],"author_name":"Vinay Kumar Sankarapu","publish_date":"2022-06-29T15:00:00","publication_year":"2022","word_count":1338,"keywords":["data science","AI","AI Software","chatbots","ML","AI Applications","RAG","Ray","Aim","deep learning","analytics","xAI"],"extracted_tech_keywords":["AI","ML","deep learning","data science","analytics","xAI","Aim","Ray","RAG","chatbots"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/verticalization-of-ai-platforms-how-to-move-from-horizontal\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":6,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10092538,"title":"Raining Quantum Investments, But Talent Still an Issue","content":"Like when the world transitioned from computer processors to graphics processors, we are currently witnessing another paradigm shift in which quantum computing is emerging from the shell of research and development to the forefront of mainstream technology. A McKinsey report published recently sheds light on the state of quantum technology today. As per the report, quantum technology start-ups, which include companies in the domain of quantum computing, communications, and sensing, received an investment of $2.35 billion from investors in 2022. This amount exceeded the record for the highest annual level of investment in quantum technology start-ups, set in 2021. Moreover, four of the biggest deals in the 2000s closed in 2022. However, the report also highlights that more investments are going into established startups than to new companies. Numbers show that only 19 quantum technology startups were founded in 2022 compared with 41 in 2021, bringing the total number of start-ups in the quantum technology ecosystem to 350. One of the companies that recently raised $24M in Series A funding is Strangeworks. Strangrworks is a software company that provides a cloud-based platform for developers, researchers, and enterprises to access and use advanced computing resources, like quantum simulators and quantum hardware. Quantum becoming mainstream William Hurley, Founder & CEO of Strangeworks, identifies two industry trends when it comes to providing scalable quantum solutions. First is the development of hybrid classical-quantum computing systems, which combine the strengths of classical and quantum computing to solve problems more efficiently, which can be useful for tasks such as optimisation problems or machine learning tasks. The second trend involves the development of software tools and platforms that enable organisations to manage and scale their quantum computing resources efficiently. These tools include schedulers and optimizers for quantum computations, as well as resource and data management tools across multiple quantum computing platforms. The case for a hybrid classical-quantum computer was also made by Timothy Costa, Director of HPC & Quantum at Nvidia, who told AIM, “while today’s QPUs are not capable of providing advantage in production applications, GPU supercomputers are time machines allowing researchers to work on future quantum systems that may accelerate critical workloads.” At GTC 2023, Nvidia generated buzz when it announced a new system called the DGX Quantum system, developed in collaboration with Quantum Machines. The system will utilise the newly open-source CUDA Quantum open-source software. Before you assume that a hybrid system involves mixing bits and qubits, Costa explains that this is not the case. The hybrid system operates by exposing familiar programming models, compilers, and toolchains for each type of accelerator, making it easy for domain scientists to map tasks to the processor (quantum or classical) that is best suited for the job. As a result, work is divided into discrete tasks that can be mapped to the processor of choice. “For quantum computation, domain scientists describe tasks for the processor at a high level, and the compilation toolchain lowers this to a representation that the quantum processor can readily understand and execute,” adds Costa. Talent Shortage Apart from the technical challenges that arise when scaling up quantum systems for practical applications, Hurley highlighted the lack of available talent with expertise in both quantum and traditional computing as one of the biggest bottlenecks to quantum adoption. While the McKinsey report provides some cause for celebration, the overall situation remains grim. The report states that the talent gap narrowed in 2022 compared to 2021, partly due to more academic institutions integrating quantum into their curriculum. According to their analysis, the remaining jobs could be filled by graduates from fields related to quantum technologies, which produce approximately 350,000 master’s-level graduates worldwide each year. There are significant investments in programs related to critical subjects in quantum computing. “Despite the industry not experiencing significant growth, thousands of students and engineers are investing in certification courses to upskill themselves in areas related to quantum technology,” L Venkata Subramaniam, IBM Quantum India Leader, had told AIM. The transition to quantum computing is expected to create four central job opportunities: hardware (building quantum computers), middleware (interconnecting hardware and software), research (developing algorithms that can run on today’s quantum computers), and data scientists or application developers (coding on top of the application layer). However, Subramaniam believes that for those interested in hardware, the path is more challenging. According to him, picking up these concepts from online self-learning is difficult, and university courses are limited.","excerpt":"Quantum computing is emerging from the shell of research and development to the forefront of mainstream technology","categories":["AI Features"],"tags":["NVIDIA"],"author_name":"Ayush Jain","publish_date":"2023-05-02T11:01:53","publication_year":"2023","word_count":733,"keywords":["CUDA","Go","machine learning","AI","Scala","Aim","GAN","NVIDIA","R","startup"],"extracted_tech_keywords":["AI","machine learning","Aim","CUDA","R","Go","Scala","CUDA","GAN","startup"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/raining-quantum-investments-but-talent-still-an-issue\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":60037,"title":"Apollo Hospitals To Launch AI-Based Coronavirus Risk Assessment App","content":"Apollo Hospitals, India’s largest integrated healthcare chain, today announced the launch of a comprehensive and integrated response plan to take the battle to COVID-19. Titled ‘Project Kavach’, which means ‘shield’, the response plan is a comprehensive approach that encompasses all aspects from information, screening and assessment, testing, to readying the infrastructure for quarantine and treatment. The culmination of a month of preparation, Project Kavach will bring together all the resources of Apollo Hospitals to fight for the people of the country against this pandemic. According to Shobana Kameneni, Executive Vice Chairperson, Apollo Hospitals Group, the organisation has specially designed an AI-based Coronavirus Risk Assessment scan for screening and initial assessment. This is available in the form of an app and on the website. The risk assessment will guide individuals about the potential risk of having the Virus through simple interactive questions. Based on the risk level, people are guided to contact the certified help center.” She said added “Over 8.5 million people across 30 countries have already gone through the screening on the Apollo 24 I 7 app and website and it is expected that over 10 million Indians will use this scan to know their risk score. The digital Apollo 24 I 7 app also offers virtual and telephony consultation with the facility to tele-connect with all doctors at Apollo Hospitals enabling patients to seek healthcare while staying at home. The 3800 Apollo pharmacies across 18 states will be pressed into action to serve people with all the medication, consumables and supplements as required. Stocks have been doubled, prices will be strictly monitored and controlled, and home delivery capability has been enhanced. It is estimated that the capacity can be ramped up from the 500,000 people served on a daily basis to 1 million if required.","excerpt":"Apollo Hospitals, India’s largest integrated healthcare chain, today announced the launch of a comprehensive and integrated response plan to take the battle to COVID-19. Titled ‘Project Kavach’, which means ‘shield’, the response plan is a comprehensive approach that encompasses all aspects from information, screening and assessment, testing, to readying the infrastructure for quarantine and treatment. The culmination of a month of preparation, […]","categories":["AI Features"],"tags":["Coronavirus","Coronavirus and AI","covid-19","covid19"],"author_name":"Vishal Chawla","publish_date":"2020-03-26T14:34:13","publication_year":"2020","word_count":297,"keywords":["Coronavirus and AI","Go","covid-19","AI","programming_languages:R","programming_languages:Go","Git","Coronavirus","covid19","GAN","R"],"extracted_tech_keywords":["AI","R","Go","Git","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/apollo-hospitals-to-launch-ai-based-coronavirus-risk-assessment-app\/","complexity_score":3,"technical_depth":7,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10067075,"title":"Popular programming languages for web development","content":"Today, we have hundreds of programming languages for web development, both frontend and backend. Developers choose the programming language depending on the project requirements, learning curve, reliability, etc. According to the 2021 TOIBE index, Python, C, and Java are among the top five programming languages in the world. This article looks at the top programming languages for web development. JavaScript JavaScript is one of the most widely used programming languages for web and mobile app development. The object-oriented scripting language is used to make web pages more interactive. JavaScript can be used for both frontend and backend development. Further, the scripting language can run directly from its source code. Interactive features like clickable buttons and images that can zoom in and out are built using JavaScript. The language is relatively easy to understand for beginners and runs on major browsers. Popular websites using JavaScript include Google, YouTube, Wikipedia, Yahoo and Twitter. PHP Hypertext Preprocessor is a server scripting language that creates scripts on web servers for customised responses to user requests. PHP is easily accessible and allows developers to update their websites frequently. Popular websites using PHP include Etsy, Facebook, Flickr, Wikipedia, Tumblr, Yahoo and WordPress. HTML HyperText Markup Language is possibly the most famous programming language in the world. It is a declarative markup language that creates a web page by labelling. Unlike a scripting language, HTML is used in page structuring and templating. It’s impossible to code a functioning website without HTML. HTML can also be used with other languages to create more interactive and complex websites. Python Python is a general-purpose language that was voted the most popular programming language in 2021. It is also one of the best languages for web development and is generally used on the back end. Python is also beginner-friendly, courtesy of its simple and highly readable syntax. Python supports multiple programming paradigms, uses English keywords instead of symbols and is easy to read and write. It also integrates well with other languages. Popular websites using Python include Spotify, Mozilla, Netflix, Reddit, the Washington Post and Uber. C# C# is a .NET language and is object-oriented. It is simple yet powerful in design and supports web development, especially on Windows or Microsoft platform applications. Additionally, C# has access to the range of NET’s libraries and frameworks, making building web apps on Windows servers more efficient. A key advantage of C# is its ability to work well with other languages and is ideal for cross-platform applications. It also supports functional programming and can be easily used in parallel processing. Ruby Ruby is an object-oriented programming language known for its ease and usability. Developers use Ruby to build architecturally clean and high-quality web applications. The language is also highly flexible, allowing developers to categorise code into simple blocks and encapsulate instructions to call on easily. It also powers the Ruby on Rails framework for web applications. Built by over six thousand contributors, the Ruby on Rails framework is a server-side web application framework. Popular websites using Ruby include Shopify, GitHub, Urban Dictionary, AirBnB, Hulu, Slideshare and Goodreads. Java Developers commonly use Java to create dynamic web pages with attractive and interactive web interfaces. Java is based on Write Once, Run Anywhere approach, facilitating cross-platform compatibility. In addition, Java Virtual Machine enables the programming language to work on various devices and platforms. Popular websites created on Java include eBay, Linkedin, Amazon, Aliexpress, Spotify, Netflix, Uber, Minecraft and NASA.","excerpt":"Popular websites using Python include Spotify, Mozilla, Netflix, Reddit, the Washington Post and Uber.","categories":["AI Trends"],"tags":["Web Development"],"author_name":"Avi Gopani","publish_date":"2022-05-16T16:00:00","publication_year":"2022","word_count":573,"keywords":["Go","AI","ML","Web Development","Git","Python","JavaScript","programming_languages:Python","GitHub","R","Java"],"extracted_tech_keywords":["AI","ML","Python","R","JavaScript","Go","Java","Git","GitHub","programming_languages:Python"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/popular-programming-languages-for-web-development\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":56274,"title":"Flashback 2019: Top 6 Tech Talks From The Rising","content":"‘The Rising’ has been the biggest meeting of women data science leaders from across the domain and women professionals from the industry as well as academia. This year, the conference is going to be held in Hotel Radisson Blu, on March 20, 2020, which will serve as a forum for exchanging ideas to build a better environment for women participating in STEM. The conference will also highlight the achievements and career interests of women in data science. In this article, we list down top tech talks from the last year, The Rising 2019. The list is in no particular order. ‘Driving towards a cleaner future through data’ by Deepika Sandeep Deepika Sandeep is the head of AI and analytics program at BLP Clean Energy. This talk revolved around how AI, ML and augmented analytics getting employed by organisations for renewable energy, manufacturing, and downtime reduction in the utility sector, preventive maintenance, energy management, bringing about automation, inventory optimisation, and manufacturing cycle time reduction. In this talk, Deepika discussed how BLP Clean Energy used analytics and machine learning for a holistic approach for renewable energy asset performance monitoring, analysis, and prediction of impending failures of components of assets. ‘Fighting prejudice in artificial intelligence’ by Smitha Ganesh Smitha Ganesh is the principal consultant & data scientist at Thoughtworks. In this talk, Smitha spoke about human biases and the consequences of using AI technology. It also discussed the unseen harm that AI that can cause human biases in the collected data sets. She further discussed how there would be a compounding effect during the emergence of AI, as the algorithms exhibit self-learning from data. Finally, she talked about the aspects like diversity mindset for data collection, which is the possibility of building and taking AI to an elevated level and recognising the vitality involved with AI training systems. ‘Unlock the power of intelligent enterprise with augmented analytics’ by Dharani Karthikeyan Dharani Karthikeyan is the director of engineering at SAP. In this talk, Dharani discusses how the enterprises are transforming themselves using machine learning and AI that augmented the traditional process of analytics in an organisation, how augmented analytics helps business users to take relevant actions from insights that are automatically generated by the system. The talk also discussed how SAP portfolio and strategy enabled businesses to have an unbiased decision-making process and how it helped in identifying the correct set of business drivers. ‘Enabling early-stage breast cancer detection using AI’ by Geetha Manjunath Dr Geetha Manjunath is the founder, CEO and CTO of NIRAMAI. Data-driven decision making using artificial intelligence is the trend of the data, seeing applications in a variety of domains. In this talk, Geetha discussed a noble application of AI in the area of healthcare developed at NIRAMAI. The talk discussed the AI application, which is an affordable and accessible solution to detect early-stage breast cancer and how it works. ‘Applying ML at scale for new user experiences’ by Vikram Vij Vikram Vij is a senior vice president at Samsung R&D Institute. In this talk, Vikram discussed how fundamental technologies such as automatic speech recognition and natural language understanding are being used to create new experiences. The talk also discussed the challenges in the domain and intent classification for natural language understanding, and deep learning-based solutions ranging from CNN vs RNNs, Word vs character-based models, handling of variations of data and conflicts between data. ‘My experiences with data’ by Mathangi Sri Mathangi Sri is the head of data at Gojek. In this talk, Mathangi discussed her learning in her 15 years of journey. She also talked about the experiences working across multiple problems and multiple types of data, including the challenges and opportunities. She further spoke about how the data science field itself has grown tremendously in a decade, and what are the new challenges posed.","excerpt":"‘The Rising’ has been the biggest meeting of women data science leaders from across the domain and women professionals from the industry as well as academia. This year, the conference is going to be held in Hotel Radisson Blu, on March 20, 2020, which will serve as a forum for exchanging ideas to build a […]","categories":["AI Trends"],"tags":["TED Talks","Women in AI","Women in Analytics","Women in Analytics India","Women in Data Science","Women in Tech"],"author_name":"Ambika Choudhury","publish_date":"2020-02-10T11:39:24","publication_year":"2020","word_count":636,"keywords":["data science","Go","Women in Data Science","artificial intelligence","machine learning","AI","R","ML","TED Talks","deep learning","Women in Tech","analytics","Women in Analytics","GAN","Women in Analytics India","Women in AI"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","data science","analytics","R","Go","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/flashback-2019-top-6-tech-talks-from-the-rising\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":43512,"title":"What To Expect In Python 3.8: Easy Debugging, Extensive API For C Extension &#038; Walrus Operator","content":"Python has proven to be the go-to language in the field of machine learning and data science. According to our recent survey, over 75 percent of respondents said that it is crucial for job seekers to have Python skills to get an opportunity in data science. The beta cycle releases of Python version 3.8 has been around the corner since the beginning of 2019 and the latest beta version of Python 3.8.0 beta 3 is expected to be released today. What’s New? Python 3.8 is still in the development phase and according to the team, the beta version has included a number of intuitive as well as notable features. PEP stands for Python Enhancement Proposal and is a design document providing information to the Python community, or describing a new feature for Python or its processes or environment. According to the PEPs submitted for version 3.8, there are several major new features in Python 3.8 than the earlier version of 3.7. The new features are mentioned below: Assignment Expressions: Python 3.8 includes the Walrus Operator (:=), one of the most important features included in this version. This is a new syntax used to assign values to variables as part of an expression using the notation NAME := expr. This PEP is a controversial one by the supporters as well as critics while in the developing stage. Positional-Only Parameters: This PEP proposes to introduce a new syntax, \/, for specifying positional-only parameters in Python function definitions. Python Initialisation Configuration: This feature adds a new C API to configure the Python Initialisation providing finer control on the whole configuration and better error reporting. Vectorcall: This feature introduces a new C API to optimise calls of objects i.e. it introduces a new “vectorcall” protocol and calling convention based on the “fastcall” convention, which is already used internally by CPython. Python Runtime Audit Hooks: This feature describes additions to the Python API including the specific behaviors for the CPython implementation which make actions taken by the Python runtime visible to auditing tools. Pickle Protocol 5: Pickle protocol 5 with out-of-band data is a new pickle protocol version (5) to cover the extra metadata needed for out-of-band data buffers. This PEP aims to make pickle usable in a way where large data is handled as a separate stream of zero-copy buffers, letting the application handle those buffers optimally. Other interesting changes include typing-related features which contain the final qualifier, literal types, and TypedDict, parallel filesystem cache for compiled bytecode, debug builds share ABI as release builds, LOAD_GLOBAL is now 40% faster, typed_ast is merged back to CPython, -strings support a handy = specifier for debugging, pickle now uses Protocol 4 by default, improving performance and much more. The developer team of Python suggests the users must perform testing during the beta phase in order to ensure that there will be no issues when the stable version of 3.8 will be released. They stated that third party users are strongly encouraged to test the beta version and hence report the problems which are being found to the Python bug tracker. The features will be modified and in rare conditions can be deleted too. “Our goal is to have no ABI changes after beta 3 and no code changes after 3.8.0rc1, the release candidate. To achieve that, it will be extremely important to get as much exposure for 3.8 as possible during the beta phase as possible.”","excerpt":"Python has proven to be the go-to language in the field of machine learning and data science. According to our recent survey, over 75 percent of respondents said that it is crucial for job seekers to have Python skills to get an opportunity in data science. The beta cycle releases of Python version 3.8 has […]","categories":["Deep Tech"],"tags":["Machine Learning","Python"],"author_name":"Ambika Choudhury","publish_date":"2019-07-29T17:00:37","publication_year":"2019","word_count":570,"keywords":["data science","Go","API","machine learning","AI","Machine Learning","RAG","Python","Aim","programming_languages:Python","R"],"extracted_tech_keywords":["AI","machine learning","data science","Aim","RAG","Python","R","Go","API","programming_languages:Python"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/what-to-expect-in-python-3-8-easy-debugging-extensive-api-for-c-extension-walrus-operator\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10057804,"title":"Top free online courses on Data Governance","content":"Since morning, most activities you would have performed have data in them, be it browsing your social media or receiving personalised news on your email; it has data at the core. But to ensure accurate and secure insights from data, it is important to govern it correctly. Data governance practices include understanding the enterprising data and ensuring it is free from potential risks while generating insights. It is becoming increasingly important for data scientists and enterprise leaders to be fluent in data governance frameworks and techniques to ensure safe AI in the future. This article lists some of the best data governance courses of different lengths for beginners and experts. Data governance is a subset of data management. While most courses outlined here do not cover data management as a whole, enthusiasts can refer to a comprehensive background provided by Full Stack Deep Learning. Udemy: Data Governance Udemy’s Data Governance course is a foundational program targeted towards beginners. The course is targeted towards company executives wanting to be involved in the data play at the workspace. The 150 minutes on-demand video covers several topics discussing data governance, popular framework, implementing data governance, the roles of the framework and best governance practices to follow. Students will receive a certificate at the end of the course. Find the course here Intermediate level: Learning Data Governance on LinkedIn Taught by Dr Jonathan Reichental, LinkedIn’s Learning Data Governance course is an intermediate level program for bringing data management to the workplace. Reichental holds a PhD in IT, post which he was a professor at different universities. In this course, he touches upon data governance, its role in emerging fields of knowledge graphics, AI and IoT and how it can support organisations in compliance and risk reduction efforts. Lastly, individuals will be trained on managing, monitoring, and measuring data governance efforts across their organisation. Find the course here Microsoft 365: Manage Governance and Compliance One of the more detailed, 4-hour long courses on data governance, this programme is held by Microsoft. Instructor Liam Cleary is a certified Microsoft trainer and Microsoft MVP. Through the course, he maps out to the Manage Microsoft 365 Governance and Compliance domain of the Microsoft 365 Mobility and Security (MS-101) exam. This program studies the tools designed to manage the content lifecycle and comply with legal or regulatory standards by teaching important governance and compliance tools and features and how to use them to protect information within Microsoft 365. Additionally, Cleary also recommends configuring tools into data loss prevention (DLP) and Azure Information Protection. Find the course here Data Governance and Stewardship Professional The Data Governance and Stewardship Professional is a certification programme held by the Institute for Certification of Computing Professionals. The choice of multi-tiered certifications ranges from taking one exam to completing 1400 hours of experience, validating expertise in data governance, focusing on availability, usability, consistency, data integrity, and data security. The credential is available on levels of executive management, principal, mastery, associate practitioner, and foundation assistant data governance professional. Find the course here SAP Master Data Governance An associate-level entry-level qualification, the SAP data governance course is a certified course consisting of a final examination that validates a candidate possesses the fundamental and core knowledge required of the application consultant profile. In addition, it guarantees the person has detailed technical knowledge to participate in a mentoring or team role. The final examination covers data governance, data quality and analytics, data replication and key mapping. Find the course here Duke University’s Data Governance, Coursera Conducted by Duke University on Coursera, the Cloud Data Engineering program consists of a comprehensive module on Data Governance. Taught by Noah Gift, professor at Duke and co-author of several books, including AWS, the course looks at the best data governance practices. Additionally, it entails practical experience by asking the students to build a serverless Data Engineering system. Finally, the course studies data governance in detail with reference to AWS models. Find the course here Penn State University The University of Pennsylvania offers an AI governance module for their AI for Business specialisation. The AI governance course details AI strategies and tools to build responsible AI governance algorithms while dealing with large datasets in enterprises. It studies AI in organisations, the risks that come with it, strategies to recognise biases within data and constructing a responsible governance strategy. Four professors across different departments teach the course. Find the course here UC Berkeley Taught by UC Berkeley in their data strategy course, data governance takes up an important module. The course identifies creating the most value from data while adhering to data governance practices. Their data strategy playbook allows organisations to balance data opportunities with data governance. More than three professors and several visiting faculties teach this through real-world case studies of Target and Samsung. The course is catered to senior leaders in technology, data managers, and IT consultants. Find the course here","excerpt":"One of the more detailed, 4-hour long courses on data governance, this programme is held by Microsoft.","categories":["AI Trends"],"tags":[],"author_name":"Avi Gopani","publish_date":"2022-01-07T11:00:00","publication_year":"2022","word_count":818,"keywords":["Go","AWS","AI","R","serverless","deep learning","data engineering","analytics","data governance","Azure"],"extracted_tech_keywords":["AI","deep learning","analytics","AWS","Azure","serverless","R","Go","data engineering","data governance"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-online-courses-on-data-governance\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10011409,"title":"Are Easy-To-Interpret Neurons Necessary? New Findings By Facebook AI Researchers","content":"“Are easy-to-interpret neurons actually necessary? It might be like studying automobile exhaust to understand automobile propulsion.” Understanding the underlying mechanisms of deep neural networks (DNNs) typically rely on building intuition by emphasising sensory or semantic features of individual examples. Knowing what a model understands and why it does so is crucial for reproducing and improving AI systems. Class selectivity is commonly used for analysing the properties of individual neurons to understand the neural networks. The researchers wrote that the preference of an individual neuron for a specific image, say neurons that activate for images of cats but not for other types of images is called “class selectivity.” Selectivity is intuitive and easy-to-understand. However, it remains to be known whether it is necessary and\/or sufficient to learn class selectivity in individual units. To explore if easy to interpret neural networks actually hinder the learning process in deep learning models, the researchers at Facebook AI published three works covering interpretability and selectivity. Why Class Selectivity Can Be Bad For AI Despite the widespread association of class selectivity with that of interpretability in deep neural networks, exclaimed the researchers, there’s been surprisingly little research into whether easy-to-interpret neurons are actually necessary. The research has recently begun, and the results are conflicting, wrote the researchers at Facebook AI. “…class selectivity in individual units is neither necessary nor sufficient for convolutional neural networks (CNNs).” The AI community was hit by a new wave of interpretability. More researchers started focussing on these challenges and making AI more understandable and intuitive. This led to, what the FB team claims, to over-reliance on intuition-based methods, which they warn, can be misleading if not rigorously tested and verified. Models should not just be intuitive but also empirically grounded. To make the best of class selectivity, the researchers developed a technique that acts like a knob to increase or decrease the class selectivity in neurons. While training a network to classify images, they added an incentive to decrease (or increase) the amount of class selectivity in its neurons. They did this by adding a term for class selectivity to the loss function used to train the networks and controlling the importance of class selectivity to the network using a single parameter. This work underlines the importance of falsifiability in the domain of AI; falsifiability, the foundation of all scientific endeavours. The researchers showed how we can falter even in the process of making things more understandable and why it is necessary to interpret the intent behind ML interpretability rightly. Key Findings In one of the related papers, the authors discussed how interpretability research suffers from an over-reliance on intuition-based approaches that in some cases have led people to an illusion of progress. Furthermore, to verify the falsifiability of interpretability, the authors recommend remembering the “human” in “human explainability”. The experiments conducted by the researchers at FB AI team led to the following results: Class selectivity is not integral to DNN function and can sometimes have a negative effect. Reduced class selectivity can make neural networks more robust against noise. Decreasing class selectivity also makes neural networks more vulnerable to targeted attacks in which images are intentionally manipulated in order to fool the networks. More broadly, concluded the researchers, these results warn against focusing on the properties of single neurons as the key to understanding how DNNs function. Researchers also hope that future work will address the question: Why do networks learn class selectivity if it’s not necessary for good performance? Read the paper on class selectivity here.","excerpt":"“Are easy-to-interpret neurons actually necessary? It might be like studying automobile exhaust to understand automobile propulsion.” Understanding the underlying mechanisms of deep neural networks (DNNs) typically rely on building intuition by emphasising sensory or semantic features of individual examples. Knowing what a model understands and why it does so is crucial for reproducing and improving […]","categories":["AI Features"],"tags":["Facebook AI"],"author_name":"Ram Sagar","publish_date":"2020-11-10T11:00:12","publication_year":"2020","word_count":587,"keywords":["Go","Facebook AI","AI","neural network","programming_languages:R","ML","Aim","deep learning","ViT","CNN","R"],"extracted_tech_keywords":["AI","ML","deep learning","neural network","Aim","R","Go","CNN","ViT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/easy-to-interpret-neurons-facebook-ai-research\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10119526,"title":"PyTorch Enables Llama 2 &amp; 3 to Run on Smartphones with Zero Code","content":"Keeping up with the focus on delivering AI models to edge devices, the PyTorch team has made it possible to run Llama 2 and 3 models on smartphones without requiring any coding announced on 1st May. The researchers presented an optimised Triton FP8 GEMM (General Matrix-Matrix Multiply) kernel TK-GEMM that leverages SplitK parallelisation. This enhancement boosts performance for small batch size inference, achieving up to 1.94 times faster speeds compared to the original Triton setup, 1.87 times faster than cuBLAS FP8, and 1.71 times faster than cuBLAS FP16 for Llama3-70B tasks on NVIDIA H100 GPUs. The SplitK parallelisation means creating more work units along the k dimension, which breaks down tasks into smaller pieces and reduces delays, especially for matrices with smaller M values. Additionally, leveraging CUDA graphs reduces CPU launch overhead, resulting in up to 6.4x speedup for a single attention layer in Llama3-70B models. These optimisations demonstrate significant performance gains and pave the way for further enhancements in FP8 inference. With the ability to run Llama models on mobile devices, developers can now create apps that harness the power of these advanced language models without the need for extensive coding knowledge. Developers can create intelligent virtual assistants, personalised language learning apps, real-time translation tools, and much more. The updates also come with instructions for running Llama 2 and 3 on both iOS and Android devices. The PyTorch team has utilised the new FP8 datatype, introduced jointly by Nvidia, Arm, and Intel, which serves as a successor to 16-bit floating point types. The FP8 data type consists of two formats: E4M3 and E5M2, which provide significant throughput improvements over their predecessors for Transformer networks. They have also identified potential future optimisation paths, such as leveraging the TMA (Tensor Memory Accelerator) hardware unit and improving Tensor Core utilisation. These optimisations could lead to even greater performance gains in the future.","excerpt":"The update boosts performance for small batch size inference, achieving up to 1.94 times faster speeds compared to the original","categories":["AI News"],"tags":["Pytorch"],"author_name":"K L Krithika","publish_date":"2024-05-03T13:53:04","publication_year":"2024","word_count":310,"keywords":["CUDA","Pytorch","NVIDIA H100","PyTorch","AI","virtual assistants","RAG","ai_frameworks:PyTorch","llm_models:Llama","R"],"extracted_tech_keywords":["AI","PyTorch","RAG","virtual assistants","CUDA","R","CUDA","NVIDIA H100","llm_models:Llama","ai_frameworks:PyTorch"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/pytorch-enables-llama-2-3-to-run-on-smartphones-with-zero-code\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10054671,"title":"Problem-Solving And Discussion With Experts Are The Best Methods For Studying A Subject: Sumanta Mukherjee, IBM","content":"Chemical engineering and applied mathematics are very rare combinations. Sumanta Mukherjee, a research scientist at IBM, possesses this rare broad knowledge base. Sumanta is an experienced research scientist with a track record of accomplishments in the information technology and services industries. In addition, Sumanta is a researcher with expertise in machine learning, data science, mathematical modelling, computational biology, bioinformatics, and algorithm design. Analytics India Magazine caught up with him to gain insights into his perspectives on some of these topics. AIM: Given that the beginning of your career was not in data science, you have climbed up the ladder certainly well. What would you say were the obstacles in starting your path in data science, and what approach did you take to overcome them? Sumanta Mukherjee: I have a diverse career path. I started my career as a chemical engineer. Then pursued higher study in computational science, followed by a PhD in applied mathematics. Post completion of every degree, I have worked with industries for a few years. I have worked as a process engineer, software developer, and currently, researcher. After completion of my PhD, I have joined IBM Research, Bangalore. I am grateful to the great set of colleagues I had at my workplace. IBM Research has a very diverse, open, and inclusive environment. Therefore, most of my learning was via interaction with the experts in the field and while solving a targeted problem. – From my experience, the best way to learn a topic is by solving a problem and discussing it with people who have experience in that field and making continuous attempts to improvise your solution. – Data science is no different. One big benefit is free access to a large community and freely available resources. However, data science is expanding at a tremendous pace, which is a challenge to keep up. It demands continuous reading and updating yourself with the trend. – A strong grasp of mathematics, statistics, and programming helps a lot. There are two important dimensions to data science, The first one is the algorithmic and mathematical aspect, andThe second one is solving a problem on a large scale. – Keeping up with both is difficult. So, better keep your attention on one specific dimension. AIM: How significant is participation in hackathons and similar competitions when pursuing a career in data science? Sumanta Mukherjee: It is very important, and the benefits are multi-faceted It is all about honing your skills. Practice makes a person better. These competitions give outreach to a larger community. There are also data science-specific competitions, like Kaggle. Anyone seriously pursuing a data science career should be a part of the Kaggle community. AIM: As someone with a research background and considerable experience working with research laboratories, could you emphasise the importance of research and the areas where companies should focus their efforts in machine learning? Sumanta Mukherjee: My answer to this question will be biased. My experience is restricted to the IBM research lab, composed of a very able set of individuals. I think industries are doing very well in finding challenging questions for the research community. One purpose is to use data science and ML to support the current industry, and the other is to explore new questions. Most industries focus on addressing the first purpose where there is a direct business value. The second purpose is more academic, but it may help improve the future of science and industry. Therefore, I hope industries in India increase their academic collaborations to achieve a balanced and sustainable future. One specific challenge to the application of data science is ethical restriction. Data can reveal many insights which may violate ethics. Therefore, defining rules and regulations around the application of data science and an effort to build algorithms that respect ethical restrictions should be prioritised. AIM: Your research and industry experience has focussed on applied mathematics and energy efficiency. When effective energy management is critical, how do you believe data scientists can help solve these problems in today’s environment? Sumanta Mukherjee: I indeed joined IBM research, the smart energy group, but currently, I am a part of the retail-supply-chain team. Data science is a tool to understand and comprehend a large volume of data. Data is in a plethora today. In any field, the volume of data is increasing exponentially. In this context, I will emphasise the two primary goals of data science, (1) estimation and (2) knowledge mining (eXplainable AI). Estimation helps in taking a reactive approach to addressing a problem, while knowledge mining may help us adopt a proactive strategy to address a problem. – If we ask the right question, data science can help us in finding a comprehensive answer. Data science is a tool to help the progress of science and technology if used correctly. AIM: Which machine learning\/deep learning algorithm is your go-to and why? Sumanta Mukherjee: Every algorithm has a different purpose. The selection of an algorithm depends on the problem. Often, we need to customise the input-output to cast the problem appropriate for an algorithm. Sometimes we may need to tweak the algorithm to cater to the problem. – In the structured data domain, one algorithm stands out – XGBoost. There are many competing alternatives, but it is always my first algorithm of choice to address structured data regression\/classification problems. The large adoption of this algorithm in the applied machine learning community is due to its stability, scalability, and easy library interface. In addition, many explainability tools help in deriving insights from the trained model. AIM: What suggestions would you provide to someone seeking their first data science position? Sumanta Mukherjee: Be a part of the active community and actively participate in the community discussion.Today, knowledge is free, and learning relevant skills completely depends on one’s interests. Do a fresher course from Coursera or Udemy. I suggest Andrew Ng’s Coursera course. It is a very good starting point.Learn Python, the language for the data science community. AIM: The rate of advancement in this field, particularly in deep learning, is unmatched. What will be the next frontier for algorithms based on deep learning? Sumanta Mukherjee: Deep learning is the current trend. What makes it beautiful, the basic building block of a deep learning model is extremely simple, but when put together as a system, it can do magic. Exponential growth in participation of the NeurIPS conference is a direct indicator of its growing popularity. Deep learning connects functional analysis, complex systems modelling, and dynamical systems analysis together into one framework. I think we still have a long way to go to uncover its full potential.I expect an imminent growth in neural graph networks, reservoir computing, and the application of causality in neural architecture design.I expect the application of deep learning will positively influence the growth of the retail industry, healthcare section, and climate adaptation. AIM: Many publicly available datasets can be used to enhance our machine learning abilities. What kind of projects should aspiring data scientists work on to improve their resumes for today’s job market, in your opinion? Sumanta Mukherjee: Natural language processing (NLP) skills are going to be in demand for some time.One bigger challenge in data science is solution deployment and automation. It is a definite skill one must acquire.Participating in various open code platforms and creating a public profile showing your coding skills helps the recruiter evaluate. AIM: Please share with us the names of role models for you, if any. How has their work inspired you? Sumanta Mukherjee: Richard P Feynman, is my role model since my childhood. I have always admired his way of understanding and explaining concepts. How easily we can explain it to others shows how well we understand the concept. Only when we understand something well enough (not by jargon, but by its basic functions) can we improvise the system or find flaws. Therefore, an in-depth understanding of the fundamentals of data science is essential. AIM: Are there any research papers that you think every data scientist should read? Sumanta Mukherjee: Research papers are very application-specific. There are tons of them, and it’s hard to list them all. I recommend articles by Geoffrey Hinton that are a must-read for those who want to work in deep learning. I closely follow the work by Bernhard Schölkopf, Yoshua Bengio, and Michael Jordan. A few texts books for avid data scientists are listed below Machine Learning – Tom Mitchell Pattern Classification – David Stork, Peter Hart, Richard Duda Machine learning: A probabilistic perspective – Kevin Murphy Deep Learning – Aaron Courville, Ian Goodfellow, Yoshua Bengio A Probabilistic Theory of Pattern Recognition – Luc Devroye, Laszlo Gyorfi, Gabor Lugosi The Elements of Statistical Learning – Trevor Hastie, Robert Tibshirani, Jerome Friedman Statistical Rethinking: A Bayesian Course with Examples in R and Stan – Richard McElreath Elements of Information Theory – Joy Thomas, Thomas Cover Information Theory, Inference and Learning Algorithms – David Mackay Learning in Graphical Models – Michael Jordan Hands-On Machine Learning with Scikit-Learn, Keras, and TensorFlow: Concepts, Tools, and Techniques to Build Intelligent Systems – Aurelien Geron","excerpt":"The discipline of data science is growing at a fast pace, making it difficult to keep updated. It necessitates constant reading and keeping up with the latest trends.","categories":["AI Features"],"tags":["Data Science","Deep Learning","Interviews and Discussions","Machine Learning","Statistics"],"author_name":"Dr. Nivash Jeevanandam","publish_date":"2021-12-02T14:00:00","publication_year":"2021","word_count":1509,"keywords":["data science","machine learning","Keras","Statistics","AI","ML","Machine Learning","NLP","Aim","deep learning","analytics","Deep Learning","Data Science","TensorFlow","Interviews and Discussions"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","NLP","data science","analytics","Aim","TensorFlow","Keras"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/problem-solving-and-discussion-with-experts-are-the-best-methods-for-studying-a-subject-sumanta-mukherjee-ibm\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":5843,"title":"Interview &#8211; Sandeep Mittal, Founder &#038; MD at Cartesian Consulting","content":"Cartesian is a global consulting firm specializing in Marketing Analytics. At its core Cartesian works on customer analytics to help drive repeat, retention, lifetime value, and enterprise-wide decision support. Our work spans segmentation and profiling, customer lifecycle management, predictive model builds, association and recommendation engines, campaign testing and impact measurement, relevance building in communication. Beyond the usual CRM goals the company also help clients use data across multiple functions such as pricing, location planning, channel performance measurement, product lifecycle management, assortment and merchandising. Cartesian is a 60 person organization in Mumbai, Bangalore, Delhi and Singapore, servicing clients in over 10 countries. In an interview with Analytics India Magazine, Sandeep Mittal of Cartesian Consulting talks more about Cartesian Consulting. [dropcap size=”2″]AIM[\/dropcap]Analytics India Magazine: What are some of the main tenets (philosophies, goals, attributes) of analytics approach and policies at your organization? [dropcap size=”2″]SM[\/dropcap]Sandeep Mittal: Our vision is to be the best blend of marketing and analytical thinking you can get on this planet earth. We’re pretty serious about that – the blend bit (and, well, the planet Earth bit too). It means being more than data driven, it also means being customer obsessed. AIM: Can you brief about some of the services you provide? SM: Loads of customer analytics – segmentation, models for response, repeat, churn, cross sell etc, customer journeys and product lifecycles, LTV and engagement measures. In addition we’re also doing some very new work around Pricing, Assortment and Merchandising for retail clients, Scheduling and marketing spends optimization. AIM: What are the key differentiators in your analytical solutions? SM: I think the key is that we really get Direct Marketing. It’s in our DNA. So we don’t just build models but we converse wide and deep with clients on the messaging, the offer, the testing, the things that make campaigns successful and also fun. AIM: Please brief us about the size of your organization and what is hierarchal alignment, both depth and breadth. SM: We’re 60 people in 4 offices – Bombay, Bangalore, Delhi and Singapore. Growing quick. AIM: What are the planned next steps\/ road ahead for your organizations? SM: We’re a rare domestic-focused analytics firm who, having built up our capabilities and reputation with some solid domestic clients are now headed outward – so that’s one. The other is to shrinkwrap some of our stellar work into solutions that are more easily taken across clients – as against doing things bespoke all the time currently. What we won’t change however is the consultative approach, and the fact that we need to constantly get the marketing\/ business side of things. Sandeep Mittal, Founder, Cartesian Consulting AIM: What are the most significant challenges you face being in the forefront of analytics space? SM: Mature marketing departments amongst prospects and clients who get how to use us, and then use us well. And people, of course, that old grouse – we have a superb team and have very good retention of staff but hiring new is always tough. AIM: How did you start your career in analytics? SM: I didn’t. I started it in CRM and Direct Marketing where the data side was always very important. One lived and breathed it before one thought of it as “analytics”. The more analytical side of it all came easily enough though when one got into the deep end of it. AIM: What kind of knowledge worker do you recruit and what is the selection methodology? What skill sets do you look at while recruiting in analytics? SM: It’s a mix really – lots of Engineer + MBAs, lots of MSc Stats, lots of BTechs. Finally smarts and attitude matter, and in our work where everyone is client facing so do communication skills and the ability to manage client relationships. I really believe in bringing in people who haven’t necessarily been in our industry all their lives, we’ve had some super hires from unlikely places. AIM: How do you see Analytics evolving today in the industry as a whole? What are the most important contemporary trends that you see emerging in the Analytics space across the globe? SM: The main evolution is the visibility and noise the industry is getting. The principles and what is eventually delivered is slower to change… especially if you’ve been doing this for a while. More data and better tools are great, but the end delivery of either great insight or great action still needs the very basics of smart people giving the data plenty of quality time. To me the best outcome of all the visibility is that a lot of smart people will come to the industry and lift what it delivers across the board. AIM: Who do you think your main competitors are? SM: IT departments with BI projects that are also expected to deliver customer insight and CRM campaigns. Seriously! As for the handful of other folks doing what we do one thinks of them as peers first and competition next –we’re all helping educate and grow the market after all. AIM: Anything else you wish to add? SM: I cringe a lot in conferences. There’s too much buzz-word and framework speak with too little real experience sharing. And I truly cringe at the big “B” word too. Down in the trenches, little has changed for smart folks working on tens of millions (perhaps) of rows of data to deliver better value from the customer base. It’s still rooted in having a good blend of (direct) marketing and analytical thinking and that’s where we love to play. [divider] [spoiler title=”Biography of Sandeep Mittal” style=”fancy” icon=”plus-circle”] Sandeep has over 14 years of experience advising clients across industries on their CRM, Loyalty and Analytics initiatives. He has been a driving force in the development of an analytics practice across over 80 client engagements ranging from large Telecom organizations, retail brands with hundreds of stores, online businesses, financial services companies, luxury hotels and resorts, and restaurant chains. He plays at the intersection of direct marketing thinking, advanced analytics and integrated marketing management technology. An advisor to the heads of over 30 client organizations currently, Sandeep helps them be data driven and customer obsessed, and take that obsession across the enterprise. Sandeep is the founder and Managing Director of Cartesian, a global consulting firm specializing in marketing analytics headquartered in Mumbai, India. He is an alumnus of the Indian Institute of Management, Calcutta. Sandeep can be reached on +91 9820343301 or smittal@cartesianconsulting.com[\/spoiler]","excerpt":"Cartesian is a global consulting firm specializing in Marketing Analytics. At its core Cartesian works on customer analytics to help drive repeat, retention, lifetime value, and enterprise-wide decision support. Our work spans segmentation and profiling, customer lifecycle management, predictive model builds, association and recommendation engines, campaign testing and impact measurement, relevance building in communication. Beyond […]","categories":["AI Features"],"tags":["analytics ceo","analytics leaders","Interviews and Discussions"],"author_name":"AIM Media House","publish_date":"2014-06-12T18:11:29","publication_year":"2014","word_count":1074,"keywords":["Go","analytics ceo","programming_languages:R","AI","analytics leaders","programming_languages:Go","Aim","analytics","GAN","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","analytics","Aim","R","Go","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/interview-sandeep-mittal-founder-md-at-cartesian-consulting\/","complexity_score":4,"technical_depth":8,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":26006,"title":"Qlik’s Plan For Growth In India, According To Their MD Arun Balasubramanian","content":"Qlik, a noted data analytics and business intelligence platform is used extensively by various industries — from e-commerce to medicine. With the ease of working and the functionality it provides, Qlik is known to be enhancing the data literacy skills in people working in various industries. The company believes that data literacy is a crucial element that can render efficiency of employees in areas such as data reading, data analysation, data compilation, data management, among others, hence helping them retrieve maximum benefits from the huge pile of data. To understand more of data literacy efforts by Qlik, the company’s growth plan in India and the challenges they faced, Analytics India Magazine interacted with Qlik’s newly-appointed managing director Arun Balasubramanian. Analytics India Magazine: How is Qlik enhancing data literacy skills in people working in various industries? Arun Balasubramanian: Qlik is leading the charge towards making the world more data literate, and to make sure everyone has equal opportunity to succeed with data, personally and professionally, Qlik recently announced a new data literacy education program which includes many elements available at no cost. The learning is product agnostic – built around widely adopted data, analytics and statistical concepts that can be used in any context and with any BI tool. For the first time, the public can get access to: Free self-paced online learning modules in the Qlik Continuous Classroom Free comprehensive data analytics certification in the Qlik Continuous Classroom Free learning resources for professors and students through the Qlik Academic Program Free skills assessment to receive a training roadmap based on skill level A series of surveys commissioned by Qlik and detailed at www.dataliteracy.info identified that the majority of workers around the world believe that data skills would make them more valuable to their employer, and that they would take advantage of skills training if made available. The Qlik data literacy education program is designed to help bridge this gap and provide benefit to all current and future employees worldwide. AIM: Does data literacy play a vital role in better output by employees? AB: Data literacy empowers users with the ability to make decisions which are more relevant and precise. Having the relevant data to back up their claims also gives them more credibility in the eyes of their peers and higher-ups. Their performance improves, as does their overall productivity and contribution towards achieving the short-term and long-term business goals. According to the recent Qlik APAC Data Literacy Survey, almost 90% of data literate persons from all sectors said that they were performing very well at work. On the other hand, only 33% of non-data literate persons got that work satisfaction. AIM: What is the role of visual analytics in enhancing the data literacy skills? AB: Visual analytics presents an interactive data interface for business users which can be used to explore the available data with simple clicks. This does away with the need for advanced data science capabilities, allowing business users to easily understand and utilise the information to achieve their desired end-result. More importantly, the rigidly-structured, query-based approach that most BI and analytics tools follow has been made redundant by Qlik, allowing for a more non-linear, speed-of-thought exploration of data. Moreover, since all the relevant data is visualised on the screen, data blind spots are eliminated. By facilitating all of this, visual analytics make it possible for even non-technical users to easily utilise data to generate highly contextual and accurate insights, thus improving the level of data literacy across the organisation. AIM: Why should Government offices implement these tools for better services? AB: India, as a nation, faces two major data challenges. To begin with, its public data is trapped across various silos and data environments, which reduces the accessibility and availability of information to the key decision-makers. The second challenge lies in the relevancy of data; most of the public data is collected on a long-term basis, which means that by the time it is processed and made available to the relevant end-user, it is already obsolete. Policies and regulations created with such incomplete, fragmented, and outdated information are therefore less likely to be beneficial for a broader spectrum of the population, and will not accomplish the desired objectives. Implementing state-of-the-art analytics solutions such as Qlik can help in effectively addressing these challenges and many public service organisations around the world has benefitted through investing in analytics with Qlik. AIM: Are data analytic tools emerging as one of the most useful business intelligence elements? AB: Business intelligence is aimed at empowering organisations with the tools that organisations require to gain a competitive edge in the market. Data analytics solutions, such as Qlik, make it possible for them to do so by helping organisations make sense of their business data. Having in-depth analytics relevant to one’s job responsibilities available at just a few clicks makes it extremely convenient for businesses to optimise their operations through better decision-making. Such easy accessibility of data also helps them to identify, track, and capitalise on emerging market opportunities. This is one of the main reasons why data analytics are finding application across more and more use-cases across the entire business value chain. AIM: Which are the top five industries in data analytics that are in a need of having data literacy in place? AB: There is a massive – and growing – demand for a date literate workforce and data-intensive industries such as BFSI, e-commerce, retail, healthcare, and manufacturing industries will be the first to be impacted. Many other sectors such as transportation, education, sports, media and entertainment, and real estate are also now leveraging data analytics to drive their operations. AIM: What are the challenges in the way of bringing data literacy in an analytics age? AB: One of the major data literacy challenges is the awareness about its need. Despite acknowledging that their organisations need to be more data-ready, many business leaders haven’t fully prioritised this. Data isn’t a resource to hoard and hand out to only a select few; the relevant information should be proactively made available to everyone across hierarchies and verticals. Adopting such a data-driven approach can help employees to make more relevant and precise business decisions, as well as to collaborate with each other for delivering better outcomes. This awareness about how data literacy can drive continued business growth and success is the need of the hour today. AIM: How is Qlik’s business is evolving in India? AB: Qlik is a leader in business analytics, and we are positioning ourselves to really take advantage of this next wave of the analytics economy in India. We’re well-positioned to grow, and continue to add capabilities like augmented intelligence and hybrid cloud. Qlik serves over 48,000 customers around the world including Bajaj Auto, BookMyShow, Lupin and Micromax, amongst others in India. We have three offices locally, of which the India headquarters is based in Bangalore. AIM: What is your vision for Qlik in India? What is the roadmap for the company in the year 2018? AB: Qlik’s vision for the future is defined by three core pillars of innovation – data, platform, and analytics. First, to drive the most value from analytics, people need to be able to work with all their data. This means accessing data wherever it resides, without being limited by its location, size or complexity. Our vision is to bridge user data, governed enterprise sources, and big data, associatively, to create a continuum from the individual through the business, on-premise or in the cloud. To do this we are making investments in data acquisition and integration, including a new associative big data indexing capability that will provide the Qlik associative experience on top of all types of data, while leaving the data where it resides. Next, organisations need to think about how data is utilised in an infrastructure that fully supports their unique needs. With current offerings, customers must choose on-premise or cloud deployments. We aim to change the choice to “and” by building our platform to work seamlessly across on-premise and cloud environments (including private, managed or public cloud), and increasingly on edge-computing infrastructures and devices running any operating system. Finally, users need to be met where they are, with the right capabilities for whatever they are trying to do. This could include a range of current and future use cases from visualising data and creating analytics, to exploring centrally deployed apps, to consuming printed output, to analysing on mobile devices (online or offline), and to building custom applications and embedded analytics.","excerpt":"Qlik, a noted data analytics and business intelligence platform is used extensively by various industries — from e-commerce to medicine. With the ease of working and the functionality it provides, Qlik is known to be enhancing the data literacy skills in people working in various industries. The company believes that data literacy is a crucial […]","categories":["AI Features"],"tags":["Interviews and Discussions","qlik india"],"author_name":"Srishti Deoras","publish_date":"2018-07-03T04:58:09","publication_year":"2018","word_count":1417,"keywords":["data science","Go","API","TPU","AI","ML","RAG","Aim","analytics","qlik india","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","ML","data science","analytics","Aim","RAG","TPU","R","Go","API"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/qliks-plan-for-growth-in-india-according-to-their-md-arun-balasubramanian\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":62354,"title":"Indian Analytics Industry Will Rebound From Recession Better Than Contemporaries: Latest Report By AIMResearch","content":"Indian economy, enterprises and consumers are expected to display relative economic strength, compared to global peers in the ensuing recession, according to a report by AIMResearch titled, Impact of the Recession on the Indian Analytics Space. It suggests that in terms of impact, the overall Indian industry typically will not experience the same sort of recessionary trends experienced by global firms and sectors. Once the downturn ends, many Indian industries including IT, FMCG, Industrials, and Domestic Banking are expected to rebound. While different sectors will experience varying impacts of the recession on their respective businesses, and correspondingly are expected to rebound from recession at different times, this report elaborates on the impact of recession on the analytics domain. The report suggests that for the IT industry, a negative impact on revenues does not directly affect the operations and analytics functions of these firms. The impact is further cushioned by government-backed stimulus packages and relatively quick resumption of business and consumption, once the recession ends. It also suggests that data or analytics professionals seeking a change of industry or type of company for a stable role in the long-run could select IT firms, FMCG firms, and Media and Digital firms for relatively high job security and low churn. As per the report, Automotive and Industrial firms and Banking corporations (domestic and captives) are typically hit the hardest during recessions. The analytics function is a cost function across these firms and is likely to be the first unit or department to face significant cuts in operations and personnel in times of recessions and downturn. In terms of startups, while the analytics function in small Indian startups could shrink by up to 40% if the recession lasts, the analytics operations for established startups could be adversely impacted by not more than 15%. Among the various businesses that will be affected by the recession, the contraction in the direct business and services is expected to be 30% for the global BFSI sector, and 10% for the Pharma sector. The report foresees a cumulative contraction of the analytics functions by a figure of 27.5%. This would affect operations and employee layoffs. For oil and gas companies in India, the recession could negatively impact demand and margins by about 40%. The report predicts that the budget and analytics operations of the State-run industrial and energy firms could contract by about 10%, with no significant retrenchment of personnel, but a freeze in future hiring. Media and the digital sector could face a contraction of analytics operations by up to 12.5%, whereas the FMCG companies could face contraction of demand for goods of up to 10-15% in the short term. According to the report, there will be no significant contraction for the analytics functions of the domestic and MNC IT firms operating in India. Whereas the overall contraction in operations for the established boutique analytics firms would be 15%. In terms of jobs, the effect of recession could result in a 10-15% drop in open jobs. The report forecasts that if the downturn continues for about a quarter, the number of open job positions could fall by 15-20% from the peak. The report suggests that once the recession lifts, IT companies will receive a steady flow of orders, which would benefit the analytics offerings – reaching pre-crisis levels three months after the recession ends, and crossing pre-crisis levels six months (2 quarters) after the recession ends. Bhasker Gupta, Founder, and CEO, Analytics India Magazine said, “While COVID-19 pandemic has taken a toll on the global economy, Indian industries and their corresponding analytics functions are quite hopeful of overcoming the losses it would make during the recession period. Just like after the 2008-2009 financial crisis, the Indian IT sector showed remarkable growth compared to other countries, we are hopeful that it would rebound the same way this year.” Read the complete report here: https:\/\/aimresearch.ai\/2020\/04\/impact-of-the-recession-on-the-indian-analytics-space\/ About AIMResearch: AIMResearch is the research arm of Analytics India Magazine. AIMResearch provides rigorous, objective research and advisory to organizations that plan to achieve higher levels of success with their analytics implementations. Its single, overriding goal is to equip clients with the insights, advice, and tools they need to create a well-oiled, data-driven enterprise. Visit AIMResearch: https:\/\/aimresearch.ai\/","excerpt":"Indian economy, enterprises and consumers are expected to display relative economic strength, compared to global peers in the ensuing recession, according to a report by AIMResearch titled, Impact of the Recession on the Indian Analytics Space. It suggests that in terms of impact, the overall Indian industry typically will not experience the same sort of […]","categories":["AI Features"],"tags":["automotive analytics","automotive data analytics"],"author_name":"AIM Media House","publish_date":"2020-04-22T13:35:25","publication_year":"2020","word_count":702,"keywords":["Go","startup","programming_languages:R","AI","data-driven","Git","Aim","automotive analytics","analytics","GAN","R","automotive data analytics"],"extracted_tech_keywords":["AI","analytics","Aim","R","Go","Git","GAN","data-driven","startup","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/indian-analytics-industry-will-rebound-from-recession-better-than-contemporaries-latest-report-by-aimresearch\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":62132,"title":"Cumulative Accuracy Profile (CAP) Curve Analysis to Evaluate Classification Models in Social Network Ads Prediction","content":"There are various methods used to evaluate the performance of a classification model. The Cumulative Accuracy Profile (CAP) curve analysis is one of those methods of evaluation. In this article, the CAP curve analysis method has been discussed where it is used to evaluate and compare the performances of four different classifiers in their classification task. These models are used in the classification where it has been predicted that whether a user will buy a product or not given a social network advertisement of that product. Through the CAP curve analysis, we will be able to identify the best model among four in this classification or prediction. Cumulative Accuracy Profile The Cumulative Accuracy Profile (CAP) is used as a tool in machine learning through which the discriminative power of a classification model is visualized. The CAP of a model represents the cumulative number of positive outcomes along the y-axis versus the corresponding cumulative number of a classifying parameter along the x-axis. The CAP is different from the Receiver Operator Characteristic (ROC) curves as ROC curves plot the true-positive rate against the false-positive rate of classification. In analyzing a classification model, the CAP curve analysis compares that model with a perfect classification model and a random classification model. It evaluates a model by comparing the curve to the perfect CAP in which the maximum number of positive outcomes is achieved directly and to the random CAP in which the positive outcomes are distributed equally. A good model will have a CAP between the perfect CAP and the random CAP with a better model tending to the perfect CAP. Social Network Ads Prediction In this experiment, we have taken the social network ads data set that is publically available on Kaggle. A company has placed the advertisement for its product on a social networking site and it has recorded the details of persons who have clicked on the advertisement and bought or not bought the product. The data set includes the features of customers including user id, gender, age, expected salary and whether they have purchased the product or not. On the basis of other attributes, it has been predicted whether the person will purchase the advertised product or not using the classification models. There are 400 observations in the data set where 300 are used to train the classification model and the remaining 100 are used to test the model. Prediction and Performance Analysis The task of predicting whether the customer will purchase the product or not is done using four classification models. The intuition behind using four models is to see the comparison of all these models and find the best among them. The four classification models used are Random Forest Model, Logistic Regression Model, K-Nearest Neighbor Model and Naive-Bayes Model. Once these models are trained then they are tested on prediction with new data. This prediction performance on new test data has been analyzed using the CAP curve analysis. In a plot having the random model and the perfect model, the performances of these four models have been visualized. Let us see the python code snippet for this task. # Import the required libraries import numpy as np import matplotlib.pyplot as plt import pandas as pd # Import the dataset and define the input and output features dataset = pd.read_csv('Social_Network_Ads.csv') X = dataset.iloc[:, [2, 3]].values y = dataset.iloc[:, 4].values # Split the dataset into the Training set and Test set from sklearn.model_selection import train_test_split X_train, X_test, y_train, y_test = train_test_split(X, y, test_size = 0.25, random_state = 0) # Feature Scaling from sklearn.preprocessing import StandardScaler sc = StandardScaler() X_train = sc.fit_transform(X_train) X_test = sc.transform(X_test) # Define the Random Forest model, train the model and make prediction on test data from sklearn.ensemble import RandomForestClassifier rf_classifier = RandomForestClassifier(n_estimators = 10, criterion = 'entropy', random_state = 0) rf_classifier.fit(X_train, y_train) y_pred_rf = rf_classifier.predict(X_test) #Define the Logsistic Regression model, train the model and make prediction on test data from sklearn.linear_model import LogisticRegression lr_classifier = LogisticRegression() lr_classifier.fit(X_train, y_train) y_pred_lr = lr_classifier.predict(X_test) #Define the KNN model, train the model and make prediction on test data from sklearn.neighbors import KNeighborsClassifier knn_classifier = KNeighborsClassifier(n_neighbors = 5, metric = 'minkowski', p = 2) knn_classifier.fit(X_train, y_train) y_pred_knn = knn_classifier.predict(X_test) #Define the Naive Bayes model, train the model and make prediction on test data from sklearn.naive_bayes import GaussianNB nb_classifier = GaussianNB() nb_classifier.fit(X_train, y_train) y_pred_nb = nb_classifier.predict(X_test) #Visualize the CAP Curve Analysis including all 4 classification models total = len(y_test) one_count = np.sum(y_test) zero_count = total - one_count lm_rf = [y for _, y in sorted(zip(y_pred_rf, y_test), reverse = True)] lm_lr = [y for _, y in sorted(zip(y_pred_lr, y_test), reverse = True)] lm_knn = [y for _, y in sorted(zip(y_pred_knn, y_test), reverse = True)] lm_nb = [y for _, y in sorted(zip(y_pred_nb, y_test), reverse = True)] x = np.arange(0, total + 1) y_rf = np.append([0], np.cumsum(lm_rf)) y_lr = np.append([0], np.cumsum(lm_lr)) y_knn = np.append([0], np.cumsum(lm_knn)) y_nb = np.append([0], np.cumsum(lm_nb)) plt.figure(figsize = (10, 6)) plt.plot([0, total], [0, one_count], c = 'b', linestyle = '--', label = 'Random Model') plt.plot([0, one_count, total], [0, one_count, one_count], c = 'grey', linewidth = 2, label = 'Perfect Model') plt.title('CAP Curve of Classifiers') plt.plot(x, y_rf, c = 'b', label = 'RF classifier', linewidth = 2) plt.plot(x, y_lr, c = 'r', label = 'LR classifier', linewidth = 2) plt.plot(x, y_knn, c = 'y', label = 'KNN classifier', linewidth = 2) plt.plot(x, y_nb, c = 'm', label = 'NB classifier', linewidth = 2) plt.legend() As we can see in the above plot, the K-NN model seems to be best among all four models because it is most close towards the perfect model. The Logistic regression model seems to be worst in comparison to all four models as it is most far away from the perfect model. In this way, we can check and compare the performance of various classification models on the same data set and find out the best one in the required task.","excerpt":"In this article, the CAP curve analysis method has been discussed where it is used to evaluate and compare the performances of four different classifiers in their classification task.","categories":["Deep Tech"],"tags":["Classification","classification accuracy","evaluation metrics","naive bayes","regression analysis"],"author_name":"Dr. Vaibhav Kumar","publish_date":"2020-04-21T12:00:00","publication_year":"2020","word_count":983,"keywords":["Go","Classification","NumPy","machine learning","TPU","evaluation metrics","AI","regression analysis","naive bayes","classification accuracy","Python","programming_languages:Python","Matplotlib","R","Pandas"],"extracted_tech_keywords":["AI","machine learning","Pandas","NumPy","Matplotlib","TPU","Python","R","Go","programming_languages:Python"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/cumulative-accuracy-profile-cap-curve-analysis-classification-prediction\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":26363,"title":"India Must Re-Skill Workforce To Keep Up With AI, Says Kris Gopalakrishnan","content":"Infosys co-founder Kris Gopalakrishnan this week reiterated what most tech giants have been saying all along — that artificial intelligence is not a direct threat to human jobs, but reskilling employees is crucial to advancement. “India has a major challenge of transitioning its young workforce to the fourth industrial revolution called AI after the eras of agriculture, manufacturing and services… As the large workforce is engaged in diverse occupations such as agriculture, manufacturing and white-collar jobs in the services sector, it needs to be re-skilled to sustain the jobs, as AI will replace traditional jobs,” said Gopalakrishnan in an interview to a newswire. He went on to explain that artificial intelligence was almost like a part of evolution and that the transition was bound to happen. He said: “We should rather think of preparing the workforce for it by reskilling it. We have to brace for skilled jobs, as many conventional jobs will be lost, though many more will be created in allied areas.” He gave the example of when humans were replaced by robots in automobile factories for painting jobs. He said that even though humans were replaced by automatons, the risk of employees being exposed to chemical paints had dropped drastically. Earlier this year, former Reserve Bank Of India (RBI) governor Raghuram Rajan had also emphasised the fact that jobs which require high intelligence, creativity and empathy would always be safe from AI. He had added that though some of the jobs may be in danger, it was wiser to progress in technology as to not lag behind the rest of the countries. On the other hand, several noted personalities have sounded the alarm bells regarding automation, AI, and its effect on employment. Last year, Infosys co-founder and celebrated corporate thought leader NR Narayana Murthy had said that AI was “worrisome” for freshers in the IT sector. At the World Economic Forum in January 2018, Jack Ma, founder and executive chairman of Alibaba had said that AI was going to be a big threat to humans. Other noted personalities such as the late physicist Stephen Hawking and Tesla and SpaceX’s CEO Elon Musk have already sounded the alarm bells with regards to AI and the harm it can incur on human beings.","excerpt":"Infosys co-founder Kris Gopalakrishnan this week reiterated what most tech giants have been saying all along — that artificial intelligence is not a direct threat to human jobs, but reskilling employees is crucial to advancement. “India has a major challenge of transitioning its young workforce to the fourth industrial revolution called AI after the eras […]","categories":["AI News"],"tags":["AI (Artificial Intelligence)"],"author_name":"Prajakta Hebbar","publish_date":"2018-07-11T12:52:11","publication_year":"2018","word_count":374,"keywords":["Go","artificial intelligence","programming_languages:R","AI","programming_languages:Go","RAG","automation","Ray","ViT","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","Ray","RAG","R","Go","ViT","automation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/india-must-re-skill-workforce-to-keep-up-with-ai-says-kris-gopalakrishnan\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10063657,"title":"Class 9 student’s app wins Google Code to Learn","content":"Moksh Dinesh Nihalani, a Class 9 student from Dhirubhai Ambani International School, Mumbai, has won the Google India Code to Learn 2021 contest for his PRISM – smart communication project, an MIT app for people with speech and hearing disabilities. Explaining about the app, Moksh said, “Around 72 per cent of parents of children with hearing disabilities from kindergarten to Class 12 do not know sign language. About 55 per cent of them use sign language. Project Prism is an app which allows people with hearing loss to communicate easily with others and lets those without hearing disability communicate with friends and acquaintances with the disability”. Besides Moksh, many school students from OMOTEC were finalists and had developed apps that could be used for real-life cases. Some of the finalists of the competitions included: Shayaan C Doshi, Class 6 from St Gregorios High School, Mumbai, developed an MIT app called “Secure Child’. Explaining about the project, Shayaan said, “This app was made with the sole purpose to teach children about safety. It has images and videos, creating awareness about different types of abuses, like physical, sexual, or emotional, that occur in their familiar or nearby surroundings. The app allows them to report to their parents or to the childcare centre if needed. The app also allows the user to take a photo of the person under abuse, share the contact and the location to seek quick help from the unsafe situation”. Veer Mehta of Class 9 from Dhirubhai Ambani International School developed a project “Dental hygiene and cavity identifier using Google Cloud AutoML”. Explaining the project, Veer said, “Majority of Indians are either ignorant or unable to afford dental care, especially in tier II and III cities and rural areas. Dental care is also expensive for many low-income families. This app identifies a  host of dental issues with just a photo of the user’s mouth; it not just saves the cost but also alerts at an early stage of any serious dental health problems. With smart devices being widely used in rural areas, this app would also be easily accessible”. Yuthika Singh, a Class 10 student of JBCN International School, Mumbai, developed a “Self-chat analyses” app using Google Cloud AutoML. Explaining about the app, Yuthika said, “Around 90 per cent of teenagers aged 13-17 years use social media to connect with others and showcase their feelings to the world. Sometimes, this takes a toll on teenagers’ mental health, polarising their emotions. If these issues are left unaddressed, social media stress may cause severe emotional imbalance, from extreme happiness to extreme sadness and anxiety. This ML-based app can analyse a document to isolate the emotional categories based on the text; it can be used to understand a teenager’s emotions by capturing keywords in their social media chats, for timely help”. Dhyey Shah, a Class 9 student from Indus International School, Pune, developed an “Identifying plant diseases using ML” app using Google Cloud AutoML. Explaining the app, Dhyey said, “Plant diseases are a critical concern worldwide, including India. They lead to the reduction of crop yield, eating away the farmer’s income. Early identification of these plant diseases is tough in rural areas, which leads to the loss of the entire crop. Due to a lack of awareness and knowledge, farmers use excessive pesticides and fertilisers, affecting the quality of the soil of the farmed land. This app will help farmers identify plant diseases based on the leaf structure”. OMOTEC’s Robotics, ML and AI-led mentoring programmes encourage students to learn, analyse and solve real-life problems with new-age and meaningful solutions and products. The main goal is to equip students with the wherewithal to work in fields poised for growth in the future. OMOTEC had 24 participants compete at the national level, of whom five had qualified as finalists. OMOTEC co-founder Shekhar Jain says, “OMOTEC aspires to be the MIT of India with our insightful techniques, rooted in robotics and coding, of experiential learning in mathematics, science and technology.” The Google Code to Learn aims to strengthen the foundation for computer science among pre-university students by providing a space to code and apply computer programming for their entries. The participating students in classes 5-10 get to create projects using Scratch (to create stories, games, and animations), a stepping stone to the world of computer programming or MIT’s open-source tool for creating android apps. These are block-based coding tools that do not require prior knowledge of programming languages. Students in classes 9-12 may use Google Cloud AutoML, which familiarises them with the concepts of machine learning (uses data to teach computers to mimic human behaviour) and artificial intelligence in an engaging manner.","excerpt":"OMOTEC aspires to be the MIT of India with our insightful techniques, rooted in robotics and coding, of experiential learning in mathematics, science and technology.","categories":["Global Tech"],"tags":["MIT"],"author_name":"Poornima Nataraj","publish_date":"2022-03-26T16:00:00","publication_year":"2022","word_count":777,"keywords":["Go","machine learning","artificial intelligence","AI","MIT","ML","RAG","Aim","ViT","cloud_platforms:Google Cloud","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","Aim","RAG","R","Go","ViT","cloud_platforms:Google Cloud"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/class-9-students-app-wins-google-code-to-learn-2021\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":64900,"title":"What Is Neuro-Symbolic AI And Why Are Researchers Gushing Over It","content":"Building thinking machines have been a human obsession since ages, and right through history, we have seen many researchers working on the concept of generating intelligent machines. While neural networks are the most popular form of AI that has been able to accomplish it, ‘symbolic AI’ once played a crucial role in doing so. It was used in IBM Watson to beat human players in Jeopardy in 2011 until it was taken over by neural networks trained by deep learning. While neural networks have given us many exciting developments, researchers believe that for AI to advance, it must understand not only the ‘what’ but also the ‘why’ and even process the cause-effect relationships. The current deep learning models are flawed in its lack of model interpretability and the need for large amounts of data for learning. This has called for researchers to explore newer avenues in AI, which is the unison of neural networks and symbolic AI techniques. What Is Neuro-Symbolic AI? A fancier version of AI that we have known till now, it uses deep learning neural network architectures and combines them with symbolic reasoning techniques. For instance, we have been using neural networks to identify what kind of a shape or colour a particular object has. Applying symbolic reasoning to it can take it a step further to tell more exciting properties about the object such as the area of the object, volume and so on. Overcoming The Shortfalls Of Neural Networks And Symbolic AI If we look at human thoughts and reasoning processes, humans use symbols as an essential part of communication, making them intelligent. To make machines work like humans, researchers tried to simulate symbols into them. This symbolic AI was rule-based and involved explicit embedding of human knowledge and behavioural rules into computer programs, making the process cumbersome. It also made systems expensive and became less accurate as more rules were incorporated. To deal with these challenges, researchers explored a more data-driven approach, which led to the popularity of neural networks. While symbolic AI needed to be fed with every bit of information, neural networks could learn on its own if provided with large datasets. While this was working just fine, as mentioned earlier, the lack of model interpretability and a large amount of data that it needs to keep learning calls for a better system. To understand it more in-depth, while deep learning is suitable for large-scale pattern recognition, it struggles at capturing compositional and causal structure from data. Whereas symbolic models are good at capturing compositional and causal structure, but they strive to achieve complex correlations. The shortfall in these two techniques has led to the merging of these two technologies into neuro-symbolic AI, which is more efficient than these two alone. The idea is to merge learning and logic hence making systems smarter. Researchers believe that symbolic AI algorithms will help incorporate common sense reasoning and domain knowledge into deep learning. For instance, while detecting a shape, a neuro-symbolic system would use a neural network’s pattern recognition capabilities to identify objects and symbolic AI’s logic to understand it better. A neuro-symbolic system, therefore, uses both logic and language processing to answer the question, which is similar to how a human would respond. It is not only more efficient but requires very little training data, unlike neural networks. IBM and MIT Researchers Are Leading The Way Of Neuro-Symbolic AI MIT-IBM Watson AI Lab along with researchers from MIT CSAIL, Harvard University and Google DeepMind has developed a new, large-scale video reasoning dataset called, CLEVRER — CoLlision Events for Video REpresentation and Reasoning. According to the paper, it helps AI recognize objects in videos, analyze their movement, and reason about their behaviours. They used CLEVRER to benchmark the performances of neural networks and neuro-symbolic reasoning by using only a fraction of the data required for traditional deep learning systems. It helped AI not only to understand casual relationships but apply common sense to solve problems. Img Src: MITIBM What did they do? As per the paper, the researchers used CLEVRER to evaluate the ability of various deep learning models to apply visual reasoning. These deep learning models work on perception-based learning, meaning that they fared well in answering description questions but did poorly on issues based on cause-and-effect relationships. To overcome this shortcoming, they created and tested a neuro-symbolic dynamic reasoning (NS-DR) model to see if it could succeed where neural networks could not. It used neural networks to recognize objects’ colours, shapes and materials and a symbolic system to understand the physics of their movements as well as the causal relationships between them. “More specifically, NS-DR first parsed an input video into an abstract, object-based, frame-wise representation that essentially catalogued the objects appearing in the video. Then, a dynamics model learned to infer the motion and dynamic relationships among the different objects. Third, a semantic parser turned each question into a functional program. Finally, a symbolic program executor ran the program, using information about the objects and their relationships to produce an answer to the question,” stated the paper. Researchers found that NS-DR outperformed the deep learning models significantly across all categories of questions. The Way Forward While the complexities of tasks that neural networks can accomplish have reached a new high with GANs, neuro-symbolic AI gives hope in performing more complex tasks. By combining the best of two systems, it can create AI systems which require fewer data and demonstrate common sense, thereby accomplishing more complex tasks.","excerpt":"Building thinking machines have been a human obsession since ages, and right through history, we have seen many researchers working on the concept of generating intelligent machines. While neural networks are the most popular form of AI that has been able to accomplish it, ‘symbolic AI’ once played a crucial role in doing so. It […]","categories":[],"tags":[],"author_name":"Srishti Deoras","publish_date":"2020-05-10T12:07:51","publication_year":"2020","word_count":915,"keywords":["Go","model interpretability","AI","neural network","data-driven","programming_languages:R","programming_languages:Go","deep learning","GAN","R"],"extracted_tech_keywords":["AI","deep learning","neural network","R","Go","GAN","model interpretability","data-driven","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/what-is-neuro-symbolic-ai-and-why-are-researchers-gushing-over-it\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10068070,"title":"ShareChat’s parent company raises USD 300 million","content":"Social networking platform ShareChat’s parent company Mohalla Tech has raised USD 300 million in a funding round led by Google, Times Group and Singapore’s Temasek Holdings, as per a Reuters report. This takes Mohalla Tech’s total valuation to USD 5 billion. The official announcement of the deal will happen sometime this week. This is Google’s second major investment in India’s short video space – the company had previously also backed Josh. In a Series G round held in December 2021, ShareChat was able to clinch an investment of USD 266 million from Alkeon Capital and Temasek and was valued at USD 3.7 billion. Reportedly, Twitter and Snap are also among its investors. ShareChat became a Unicorn in April 2021 when it raised USD 502 million in a Series E round led by US hedge fund Tiger Global and Lightspeed Ventures at a USD 2.1 billion valuation. ShareChat has 180 million monthly active users. Mohalla Tech’s other subsidiary video sharing platform Moj, along with the recently acquired MX TakaTak, has a combined user base of 300 million. The popularity of these short video apps has shot up in India, especially after the 2020 ban of Bytedance’s TikTok and other popular Chinese apps.","excerpt":"In a Series G round held in December 2021, ShareChat was able to clinch an investment of USD 266 million.","categories":["AI News"],"tags":["Funding","investment","valuation"],"author_name":"Shraddha Goled","publish_date":"2022-05-30T20:00:21","publication_year":"2022","word_count":201,"keywords":["Go","API","Funding","investment","unicorn","funding","AI","programming_languages:R","programming_languages:Go","valuation","R"],"extracted_tech_keywords":["AI","R","Go","API","unicorn","funding","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/sharechats-parent-company-raises-usd-300-million-led-by-google-and-times\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10117611,"title":"Google Cracks Bengaluru’s Traffic","content":"Here’s something that can potentially ease, if not annihilate, Bengaluru’s traffic woes. Google’s Project Green Light is helping the city notorious for its jams and bottlenecks, improve traffic flow at intersections and reduce stop-and-go emissions. This stems from an idea that Dotan Emanuel, researcher at Google, conceptualised over a dinner chat with his wife. Started in 2021, Google’s Project Green Light looks to optimise traffic lights in urban areas to reduce vehicle emissions, thereby contributing to the global effort to combat climate change and enhance urban mobility. With AI and insights from Google Maps driving trends, Green Light demonstrates a deep understanding of global road networks. This enables it to model traffic patterns and build intelligent recommendations for city traffic engineers to optimise traffic flow. “Project Green Light is now live in 13 global cities, like Rio, Hamburg, Bengaluru, and my hometown – Seattle,” said Juliet Rothenberg, product team lead of Climate AI at Google, in a recent interview with CBS News, demonstrating its workings. “Every red light is an opportunity for us,” she added. Further, she said the early numbers indicate a potential for up to 30% reduction in stops and up to 10% cut in emissions at intersections. At the 70 intersections where Green Light is already live, this can save fuel and lower emissions for up to 30 million car rides monthly.” Currently, Project Green Light operates at 70 intersections in 12 cities, including Bengaluru, Hyderabad, and Kolkata. “Green Light has become an essential ally for the Kolkata Traffic Police. It contributes to safer, more efficient, and organised traffic flow and has helped us reduce gridlock at busy intersections,” said Vineet Kumar Goyal, the commissioner of police in Kolkata, reflecting on the initiative’s integration into the city. He said that since November 2022, they have implemented suggestions at 13 intersections. “The outcome is excellent as per the feedback from commuters and traffic personnel,” he added. Next Up: Solving ‘Ghost Traffic’ Project Green Light initiative has also addressed the concerns revolving around what you may call ‘ghost or phantom traffic‘, when congestion occurs seemingly without a clear cause. A patented technology, Ghost Traffic Detection and Avoidance, outlines this approach to addressing ghost traffic. This technique mitigates traffic congestion by identifying road segments with high traffic volumes and providing speed recommendations to vehicles approaching those segments. It works by gathering sensor data from vehicles and infrastructure, detecting areas of congestion, and then calculating a target speed for vehicles to maintain a safe distance from the cars in front and behind them. This target speed is communicated to the drivers through a client device or vehicle interface. By ensuring consistent spacing between vehicles, the system aims to smoothen out traffic flow and reduce the stop-and-start patterns that often exacerbate congestion. This is important because it addresses ‘ghost traffic jams’, where the traffic slows down without an apparent cause like accidents or roadwork. By managing vehicle speeds in response to real-time traffic conditions, it can significantly reduce unnecessary congestion and improve the overall traffic flow. In other words, Google Maps provides real-time feedback to drivers, indicating the optimal speed to maintain in relation to the vehicle’s speed ahead. This is reflected through numerical or colour-coded indicators: green to speed up, yellow to slow down, and blue to maintain the same speed. Recently, Google collaborated with the Bengaluru Traffic Police to reduce congestion. Features like Street View were introduced to offer real-time updates and alternative routes for faster commutes. Last year, the tech giant partnered with BMTC to optimise routes and schedules available to commuters using mobile phones. Google’s Project Green Light certainly looks promising for curbing traffic congestion, not only in Bengaluru but also in other cities. However, the tech giant is not alone in this initiative. Recently, a traffic control experiment was conducted in Nashville to understand if AI can solve phantom traffic. Over 100 Nissan Rogue, Toyota RAV4, and Cadillac XT5 vehicles equipped with AI-powered control systems were designed to turn the cars into robot traffic managers. The researchers now aim to implement the same on a larger scale. The space does look more exciting than ever.","excerpt":"Currently, Project Green Light operates at 70 intersections in 12 cities, including Bengaluru, Hyderabad, and Kolkata.","categories":["Global Tech"],"tags":["Google"],"author_name":"Vidyashree Srinivas","publish_date":"2024-04-03T18:30:00","publication_year":"2024","word_count":688,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","llm_models:T5","Aim","Google","GAN","R","T5"],"extracted_tech_keywords":["AI","Aim","R","Go","T5","GAN","llm_models:T5","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/google-cracks-bengalurus-traffic\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":45620,"title":"Using Data For Hyper-Personalisation In Credit Card Offers: A Case Study","content":"In this digital era, hyper-personalisation is the baseline for addressing customer demands with more individualised experience in terms of products, services and content. Today’s customers want to engage with brands which can: Instantly identify them Have immediate access to information about every interaction, on every channel (mobile, website, apps etc.) Based on the data, understand their issue and know how to solve it. Need For Hyper-Personalisation According to Google, ‘best’ search phrases have increased by 80% in the past 2 years on mobile devices. People are researching online heavily to make more informed decisions According to individual preferences, consumers are more likely to purchase from someone whose offers are personalized User engagement with content has gone down and information overload is making consumers tune out. Credit card market in India As per RBI figures, India has primarily been a debit card market. Unified Payments Interface (UPI) and mobile wallets have grabbed the attention of policymakers, but the increase in the number of credit cards is not only an indication of growing digital payments but also the expansion of retail borrowers in the ecosystem. Trends in Indian Credit Card Market Upgraded technology for Enhanced Security: Banks are increasingly using advance technology(EMV) to make use of credit cards more secure Outsourcing and Joint Venture(JV): Many Indian banks are entering into JV or outsourcing their overall credit card business to develop a separate and dedicated business unit Credit card gaining prominence in Rural India: Realising the importance of rural population, banks have started to focus and encourage customers to use credit cards Benefits of Hyper-Personalisation Deliver better results through higher conversions Positive lift in online purchases Improved brand affinity and retention Challenges Organisational constraints \/silos make it difficult to hold anyone accountable to personalise goals Understanding buyer behaviour Assembling a real-time view of the customer with full context Integrating third party data Objective To make relevant hyper-personalisation card offers to bank customers using distance basis geocode and maximise Cross\/Upselling opportunities. Data Hyper-personalisation requires a well-integrated framework of multiple technology tools and processes to produce the desired real-time targeting results. Data Elements Methodology In the real world, we find locations based on some description. This might be house number, street name, city, state, or country. Geocoding is the process of transforming a description of a location – such as a pair of coordinates, an address to a location on the earth’s surface. Contextual data analysis can provide much-needed insight into customer behaviour patterns, helping the bank to understand their customers better and improve their experience. Possible Solution Perform data collection of customer touchpoints from various source systems and store in a big data system Use Geo-location services to extract the geocodes(Latitude & Longitude ) of customer addresses Then find the nearest restaurants\/cuisines from customer addresses using proximity distance analysis method Segment the customers into groups of individuals that are similar in specific ways such as Age, average spend, demographics, spending habits etc. using un-supervised Machine learning models Analyse contextual behaviour with factors such as the location from which the card is accessed, the time of the day when the event occurred or the industry in which the customer works etc. Test and evaluate the performance of the model Conclusion With the rising number of communication channels and self-service interactions, customers are increasingly expecting personalisation at every point of their journey. The businesses able to offer this level of service and intelligently manage decisions will be the ones that will be the next generation of customer service leaders.","excerpt":"In this digital era, hyper-personalisation is the baseline for addressing customer demands with more individualised experience in terms of products, services and content. Today’s customers want to engage with brands which can: Instantly identify them Have immediate access to information about every interaction, on every channel (mobile, website, apps etc.)  Based on the data, understand […]","categories":["AI Features"],"tags":[],"author_name":"Biswajeet D","publish_date":"2019-09-07T15:00:16","publication_year":"2019","word_count":583,"keywords":["big data","Go","machine learning","programming_languages:R","AI","programming_languages:Go","Git","RAG","GAN","R"],"extracted_tech_keywords":["AI","machine learning","RAG","R","Go","Git","big data","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/using-data-for-hyper-personalisation-in-credit-card-offers-a-case-study\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10010494,"title":"Top 5 Neural Network Models For Deep Learning &#038; Their Applications","content":"Neural networks are a series of algorithms that identify underlying relationships in a set of data. These algorithms are heavily based on the way a human brain operates. These networks can adapt to changing input and generate the best result without the requirement to redesign the output criteria. In a way, these neural networks are similar to the systems of biological neurons. Deep learning is an important part of machine learning, and the deep learning algorithms are based on neural networks. There are several neural network architectures with different features, suited best for particular applications. Here, we are going to explore some of the most prominent architectures, particularly in context to deep learning. Multilayer Perceptrons Multilayer Perceptron (MLP) is a class of feed-forward artificial neural networks. The term perceptron particularly refers to a single neuron model that is a precursor to a larger neural network. An MLP consists of three main layers of nodes — an input layer, a hidden layer, and an output layer. In the hidden and the output layer, every node is considered as a neuron that uses a nonlinear activation function. MLP uses a supervised learning technique called backpropagation for training. When a neural network is initialised, weights are set for each neuron. Backpropagation helps in adjusting the weights of the neurons to obtain output closer to the expected. MLPs are most ideal for projects involving tabular datasets, classification prediction problems, and regression prediction problems. Convolution Neural Network Convolution neural network (CNN) model processes data that has a grid pattern such as images. It is designed to learn spatial hierarchies of features automatically. CNN typically comprises three types of layers, also referred to as blocks — convolution, pooling, and fully-connected layers. The convolution and pooling layers perform feature extraction, and these extracted features are mapped into the final output by the fully connected layer. CNN is best suited for image processing. Some of the applications areas of CNN are in image recognition, image classification, object detection, and face recognition. Recurrent Neural Networks In Recurrent Neural Networks (RNN), the output from the previous step is fed back as input to the current step. The hidden layer in the RNN enables this feedback system. This hidden state can store some information about the previous steps in a sequence. The ‘memory’ in RNN helps the model in remembering all the information that has been calculated. It, in turn, uses these same parameters for each of the inputs to produce the output, thereby reducing the complexity of parameters. RNN is one of the most widely used types of neural networks, primarily because of its greater learning capacity and its ability to perform complex tasks such as learning handwritings or in language recognition. Some of the other fields where RNN finds application is — prediction problems, machine translation, video tagging, text summarisation, and even music composition. Deep Belief Network The Deep Belief Networks (DBN) use probabilities and unsupervised learning to generate the output. DBNs consist of binary latent variables, undirected layers, and directed layers. DBNs are unlike other models as every layer is regulated in order, and each one of them learns the entire input. In DBNs, each sub-network’s hidden layer is a visible layer for the next one. This composition enables quick layer-by-layer unsupervised training procedure where contrastive divergence is applied to each sub-network, starting with the lowest visible layer. Greedy learning algorithms are used to train DBNs. The learning takes one layer at a time. Hence, each layer receives different versions of the data, and each layer uses the output from the previous layer as its input. DBNs find major applications in image recognition, video recognition, and motion capture data. Restricted Boltzmann Machine The Boltzmann Machine (RBM) is a generative and non-deterministic (stochastic) neural network that learns probability distribution over its set of inputs. RBMs are shallow, two-layer neural networks that constitute the building blocks of deep-belief networks. The first layer in an RBM is called the visible or the input layer, and the second one is called the hidden layer. It consists of a neuron-like unit called node; the nodes are connected to each other across layers but not within the same layer. RBMs are generally used in building applications such as dimensionality reduction, recommender systems, and topic modelling. However, in recent years generative adversarial networks are slowly replacing RBMs.","excerpt":"Neural networks are a series of algorithms that identify underlying relationships in a set of data. These algorithms are heavily based on the way a human brain operates. These networks can adapt to changing input and generate the best result without the requirement to redesign the output criteria. In a way, these neural networks are […]","categories":["AI Trends"],"tags":["CNN deep learning","cnn neural network","Convolution Neural Network","Deep Learning","MLP","online network graph","RNN"],"author_name":"Shraddha Goled","publish_date":"2020-10-25T18:00:38","publication_year":"2020","word_count":722,"keywords":["Go","CNN deep learning","machine learning","TPU","AI","online network graph","neural network","ML","image recognition","cnn neural network","Convolution Neural Network","MLP","deep learning","object detection","RNN","Deep Learning","R"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","neural network","image recognition","object detection","TPU","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-5-neural-network-models-for-deep-learning-their-applications\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10119812,"title":"OpenAI is Helping Farmers in India Increase Crop Yields","content":"OpenAI said that AI tools like ChatGPT today are being used around the world to help farmers in India and Kenya increase crop yields in partnership with Digital Green. Founded initially as a project within Microsoft Research India’s Technology for Emerging Markets in 2006 by Rikin Gandhi and his colleagues, Digital Green became an independent NGO in 2008. The organisation focuses on training farmers to produce and share short videos that document their challenges, solutions, and success stories, aiming to facilitate a technology-enabled means of behavior change communication. https:\/\/t.co\/sgqBNELrZY“Early estimates suggest that by leveraging generative AI, https:\/\/t.co\/9gVhor19TF can bring down the cost of traditional extension services 100x, from $35 per farmer to $0.35 cents per farmer. One key driver of this productivity increase is that…— Adam.GPT (@TheRealAdamG) May 8, 2024 Digital Green recently introduced Farmer.Chat using GPT-4, covering a wide range of agricultural topics including crop advice, disease identification, weather forecasts, and market information. According to OpenAI, early estimates suggest that by leveraging generative AI, Farmer.Chat can bring down the cost of traditional extension services 100x, from $35 per farmer to $0.35 cents per farmer. One key driver of this productivity increase is that Farmer.Chat is accessible across a wide range of languages, including Hindi,  Swahili, and regional languages, using an agile tech stack that integrates with local language translation datasets and services in each country. Similarly, Jugalbandi, is another AI chatbot powered by OpenAI’s GPT models through Microsoft’s Azure OpenAI Service, which assists farmers and villagers in rural India to access information about various government schemes beneficial to them. The chatbot, accessible via WhatsApp, retrieves relevant program details typically documented in English and delivers them in the user’s native language from among 10 of the 22 official Indian languages Meanwhile, Wadhwani AI, a non-profit institute dedicated to developing AI solutions for social good, is exploring the use of generative AI to power Kissan (farmer) call centers. Alpan Raval, the chief scientist for AI\/ML at Wadhwani AI, said that they are building a Kissan call center support system using generative AI to assist farmers with their queries. Their approach involves augmenting human experts’ knowledge by utilising models that provide automated responses based on a knowledge base created from current government reports and documents, with a speech interface enabling conversational AI interactions. KissanAI is another AI startup that has developed a proprietary AI chatbot that assists farmers with various agricultural tasks such as irrigation, pest control, and crop cultivation.Last year, it released Dhenu, the world’s first agriculture-specific LLM. Dhenu is designed to provide multilingual support, initially available in Hindi and English, making it accessible to a broad range of farmers across India.","excerpt":"Digital Green recently introduced Farmer.chat using GPT-4, covering a wide range of agricultural topics including crop advice, disease identification, weather forecasts, and market information.","categories":["Deep Tech"],"tags":["AI Impacts"],"author_name":"Siddharth Jindal","publish_date":"2024-05-08T09:31:26","publication_year":"2024","word_count":441,"keywords":["Go","ChatGPT","OpenAI","AI","R","ML","RAG","Aim","generative AI","AI Impacts","Azure"],"extracted_tech_keywords":["AI","ML","generative AI","ChatGPT","OpenAI","Aim","RAG","Azure","R","Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/openai-is-helping-farmers-in-india-increase-crop-yields\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10001202,"title":"This Blockchain-Powered Payments Provider Is Ruling The Remittance Market In India","content":"India is one of the world’s fastest-growing economies, attributed directly to its initiatives in the digital field. The country is vigorously innovating and rewriting the playbook for multiple paradigms in the digital space. One of these is payments and settlement, and government-driven efforts such as the UPI project and the general focus on ‘Digital India’ that is gradually leading to a cashless economy. This has left the space ripe for efforts in the digital payments space, which is something that Indian banks are looking forward to in India. However, a payments company from Silicon Valley has its eyes on India for this exact reason. The subcontinent clocked close to $70 billion in inward remittances by corporates and retail customers. This has made it a priceless target for the company looking to connect the world’s payment systems — Ripple. What Ripple essentially does is reduce the settlement times  by enabling high-volume, low-value transactions via blockchain technology. Ripple is a blockchain company based out of the US which aims to have solved the payments problem for the world today, as it faces major issues of non-transparency and long transaction times. Existing systems are outdated and also come with a high error rate. Ripple aims to change the way payments are processed by using blockchain as an innovative solution. Ripple Eyes India As A Big Market As India is one of the biggest sources of inward remittances in the world, it has been chosen as one of the countries that Ripple wants to target. Ripple’s goal is to provide seamless cross-border transactions for every country in the world, and the presence of a large number of inward remittances showed promise for the company. The purpose of establishing a base in India also comes with the additional benefit of providing liquidity for rupee trade in the bank to bank transfers, but also to increase efficiency and promote faster payments as a part of India’s digital push. This led to Ripple devoting further focus to the country, establishing an office in Mumbai and appointing Navin Gupta as the country manager for Ripple India in 2017. Regarding this, the CEO of Ripple, Brad Garlinghouse, stated, “India is transforming itself into a digital economy and is an innovative leader in payments. Our new office…will allow us to respond to the rapidly growing demand for frictionless payments by our current customer.” It was also reported that Ripple had been in talks with the National Payments Corporation of India, who are responsible for developing UPI and IMPS. This was to facilitate account-to-account transfers globally. Ripple aims to achieve this by creating a global RTGS system to enable fast transactions for a lower value. The partnership was also to reportedly increase the effectiveness of RuPay cards offered by the NPCI to allow it to work cross-borders. In addition to these steps, the company has also established a working operation with the banks and financial institutions of the country. Ripple’s Footprint In India The company has already engaged in multiple partnerships in India, especially with banks. Banks cannot use the company’s flagship solution known as xRapid due to the lack of regulatory clarity regarding the use of digital assets for settlement. To make up for this, they have begun using the company’s blockchain offering known as xCurrent widely. YES Bank is set to use Ripple as a competition to Swift, and is a part of their ethos of “using state-of-the-art technology”. Regarding this, the bank even commenced acting as a gateway to other banks in India. It has also plugged into Ripple’s blockchain network known as RippleNet, and set up a node in India. This has allowed them to credit money instantly for almost any account in the country. This partnership enabled Ripple to reach into the country with unprecedented depth. Kotak Mahindra also partnered with Ripple and is using its xCurrent technology to enable the end-to-end tracking of instant payments and settlements over Ripple’s network of banks known as RippleNet. This is said to cut down international money transfers significantly, and it could come down to minutes instead of days. IndusInd Bank partnered with Ripple to deliver instant settlements and atomic confirmations in pursuit of enhancing the client experience. This also comes with the added benefit of being able to build on Ripple’s extensive network of banks and financial institutions around the world. This allows IndusInd to solve global payments connectivity pain points through its partnership with Ripple. Axis Bank has also partnered with Ripple in order to launch an instant international payments service using their blockchain solution. This also increases transparency. This service is set to receive payments from RakBank in UAE for corporate customers to receive payments from SC Bank in Singapore. The Future Of Frictionless And Cheap Cross-Border Payments The partnership also highlights the role of India in Ripple’s plans to connect the world’s payments, as the country is an important financial gateway between ASEAN countries and the rest of the world owing to its liquid currency.  The negative effect of not solving the payments problem affects emerging markets as it does not allow for inward remittances to these markets, instead of keeping capital in established markets. Using Ripple’s products, customers can also send lesser amount of money more frequently and at lower fees. This presents a much healthier model for inward remittances for families that need money from relatives working abroad, as opposed to high fees found on many remittance services today. Better regulation in India will allow banks to use Ripple’s xRapid solution for even faster settlements, as it utilises a digital asset known as XRP. Currently, all banks are banned from transacting with digital assets, putting a hold on Ripple’s plans in the country. However, they seem to be continuing to increase their reach through partnerships, laying the foundation for a frictionless payments world.","excerpt":"India is one of the world’s fastest-growing economies, attributed directly to its initiatives in the digital field. The country is vigorously innovating and rewriting the playbook for multiple paradigms in the digital space. One of these is payments and settlement, and government-driven efforts such as the UPI project and the general focus on ‘Digital India’ […]","categories":["AI News"],"tags":["payments","Yes Bank"],"author_name":"Anirudh VK","publish_date":"2019-02-18T16:44:48","publication_year":"2019","word_count":970,"keywords":["Go","payments","API","programming_languages:R","AI","ML","programming_languages:Go","Git","Aim","Yes Bank","ViT","R"],"extracted_tech_keywords":["AI","ML","Aim","R","Go","Git","API","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/this-blockchain-powered-payments-provider-is-ruling-the-remittance-market-in-india\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10052098,"title":"Graph Neural Networks Are Trending, Here’s Why","content":"Machine learning and deep learning methodologies have seen massive advancements in the recent past. GNN is a relatively newer deep learning method that comes under the category of neural networks that work on processing data on graphs. These algorithms can look for information from the graphs and predict outcomes with the help of the information gathered. A graph usually represents data with two components–nodes and edges (that form a connection between two nodes). GNN can be applied to graphs to conduct node-level, graph-level and edge-level predictions. Better than CNNs? As per research, CNNs can only operate on regular Euclidean data, like images (2D grids) and texts (1D sequences), while these data structures can be regarded as instances of graphs. Graphs, on the other hand, are non-Euclidean and can be used to study and analyse 3D data. GNNs can offer the human brain’s reasoning process, which sets them apart from other neural networks. In standard neural networks, the dependency information is only regarded as the feature of nodes. GNNs can propagate the graph structure instead of using it as part of the features. The research also says that to present a graph completely, it is needed to move through all the possible orders as the model’s input like CNNs and RNNs. But GNNs can propagate on each node respectively while ignoring the input order of nodes. Applications of GNNs Natural language processing–Graphs can be an important part of NLP applications. They can be used for text classification, information extraction, and answering questions. Computer vision–Though the application of GNN in computer vision is still growing; much progress has been made. GNN algorithms can be used in image classification, human-object interaction, and few-shot image classification, among others. Physics–This paper says that “a physical system can be modelled as the objects in the system and pair-wise interactions between objects.” GNNs can be used by modelling the objects as nodes and pair-wise interactions as edges. Chemistry and Biology–GNNs have found applications in molecular fingerprints, chemical reaction prediction, protein interface prediction, drug research, biomedical engineering and more. Traffic networks–GNNs are also used for forecasting traffic movement, volume or the density of roads. The nodes can be the sensors installed on roads, while the distance between pairs of nodes measures the edges. Recommendation systems–Graphs are increasingly being used in user interactions with products by companies. Problems remain The paper named “Graph neural networks: A review of methods and applications” talks about a few problems that remain with GNN, though much progress has been made. They are: Robustness–As it is a part of the neural network family, GNNs are also vulnerable to adversarial attacks. While an attack on images and text focuses on features, adversarial attacks on graphs consider further information. Interpretability–The paper says that it is important to apply GNN models to real-world applications with trusted explanations. Like computer vision and NLP, it is important to look at interpretability on graphs. Graph Pre Training–Neural network models need a large amount of labelled data. It is expensive to obtain such a high quantity of human-labelled data. So, self-supervised models are proposed to guide models to learn from unlabeled data available from websites. This method has been successful in CV and NLP. Focus has also been on pretraining on graphs, but they come with different problem settings and focus on different aspects. As it is an emerging field, it still contains problems related to the pretraining tasks’ design and the effectiveness of existing GNN models on learning structural or feature information. Complex graph structures–Graph structures can be flexible and complex in real-life applications. Various works have been proposed to deal with complex graph structures like dynamic or heterogeneous graphs. GNNs can be a very important area in machine learning in the coming years. If the problems listed above can be addressed, they can be deployed to tackle big problems as well.","excerpt":"GNNs can be deployed in computer vision, NLP, traffic network to solve different problems","categories":["AI Features"],"tags":["CNNs","Computer Vision","Deep Learning","graphs","Machine Learning","NLP"],"author_name":"Sreejani Bhattacharyya","publish_date":"2021-10-22T12:00:00","publication_year":"2021","word_count":642,"keywords":["Go","text classification","machine learning","AI","neural network","Machine Learning","recommendation systems","computer vision","NLP","deep learning","graphs","Computer Vision","Deep Learning","CNNs","R"],"extracted_tech_keywords":["AI","machine learning","deep learning","neural network","NLP","computer vision","recommendation systems","text classification","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/graph-neural-networks-are-trending-heres-why\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10142529,"title":"Microsoft’s Copilot Vision Learns to See, Shop, and Scroll Like You","content":"Microsoft has launched Copilot Vision, an experimental feature integrated into the Edge browser, enabling AI-assisted browsing by analysing web pages in real time. The feature is available in preview for select Copilot Pro subscribers in the United States through Copilot Labs, the company announced on its blog. Maybe you shouldn't play favorites with features, but Copilot Vision is an update I've been excited about since day one. It's the first AI experience of its kind—now you can show, not tell, AI what you want help with in real time as you scroll, shop, guess locations…🌏 pic.twitter.com\/DJNmgMuw7D— Mustafa Suleyman (@mustafasuleyman) December 5, 2024 The opt-in service allows Copilot Vision to read and analyse web pages with user permission, providing insights, simplifying information, and assisting with tasks such as holiday shopping or planning outings. “Browsing no longer needs to be a lonely experience with just you and all your tabs,” the Copilot Team stated in the announcement. Vision provides support by scanning and interpreting the page’s content, helping users make decisions or learn from the information presented. For instance, Vision can guide users through learning new games or finding specific products that match their preferences during online shopping. Microsoft emphasised privacy and security in the development of Vision, ensuring that user data is not stored after a session ends and is handled in line with the company’s privacy policies. The company said, “Only Copilot’s responses are logged to improve our safety systems.” Currently, Vision is limited to interacting with a select number of websites. Microsoft plans to expand its availability gradually, gathering user feedback to refine the experience. The company is also collaborating with third-party publishers to enhance how Vision interacts with web pages. “Vision does not capture, store, or use any data from publishers to train our models,” the blog post added. Similar to Copilot Vision, OpenAI is set to launch its agent, Operator, in January. Meanwhile, Google is working on an experimental AI assistant called “Jarvis,” likely to be powered by the Gemini 2.0.  Jarvis operates within Chrome and interacts with on-screen elements like fields and buttons. It handles complex tasks, such as booking flights and assisting with online shopping, simplifying digital activities for users. Similarly, Anthropic has launched the ‘computer use’ feature for the Claude 3.5 Sonnet version, which allows the AI to autonomously perform tasks like moving the mouse, clicking, and typing. Targeted at software developers, it can handle complex activities such as coding a basic website or planning outings across different applications.","excerpt":"Similar to Copilot Vision, OpenAI is set to launch its agent, Operator, in January.","categories":["AI News"],"tags":["Microsoft"],"author_name":"Siddharth Jindal","publish_date":"2024-12-05T23:22:26","publication_year":"2024","word_count":415,"keywords":["Anthropic","Go","Gemini 2.0","OpenAI","AI","Git","llm_models:Claude","ViT","Claude 3.5","R","Microsoft"],"extracted_tech_keywords":["AI","OpenAI","Claude 3.5","Anthropic","Gemini 2.0","R","Go","Git","ViT","llm_models:Claude"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/microsofts-copilot-vision-learns-to-see-shop-and-scroll-like-you\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10063312,"title":"A tutorial on various clustering evaluation metrics","content":"Clustering is the process of originating groups of data points based on their uniformity and diversity. It is generally used for unsupervised learning where there is no dependent variable. As there is no dependent variable in the unsupervised learning we could not use any regression or classification performance metrics on this because they all need a dependent variable to compare the predictions. In this article, we will be learning about different performance metrics for clustering and implementation of them. The major points to be covered in this article are listed below. Table of contents Why not classification evaluation metrics?Silhouette ScoreCalinski Harabaz IndexDavies Bouldin indexDescription of dataHow to use clustering evaluation metrics? Let’s first understand why standard evaluation metrics can not be used in clustering. Why not classification evaluation metrics? Since in every classification evaluation metric either it is a confusion matrix or log loss there is a need for a dependent variable or target variable. As these evaluation techniques calculate the metrics based on observed and predicted values. This is the reason why any classification evaluation metrics can not be used to evaluate the performance of any clustering algorithms. The evaluation metrics which do not require any ground truth labels to calculate the efficiency of the clustering algorithm could be used for the computation of the performance evaluation. There are three commonly used evaluation metrics: Silhouette score, Calinski Harabaz index, Davies-Bouldin Index. Are you looking for for a complete repository of Python libraries used in data science, check out here. Silhouette Score To study the separation distance between the clusters formed by the algorithm silhouette analysis could be used. The distance between the cluster can be calculated by different types of distance metrics ( Euclidean, Manhattan, Minkowski, Hamming). Silhouette score returns the average silhouette coefficient applied on all the samples. The Silhouette Coefficient is calculated by using the mean of the distance of the intra-cluster and nearest cluster for all the samples. The Silhouette Coefficient ranges from [-1,1]. The higher the Silhouette Coefficients (the closer to +1), the more is the separation between clusters. If the value is 0 it indicates that the sample is on or very close to the decision boundary between two neighboring clusters whereas a negative value indicates that those samples might have been assigned to the wrong cluster. The formula is as follows: (nc-ic)\/max(ic,nc) where, ic  = mean of the intra-cluster distance nc = mean of the nearest-cluster distance Calinski Harabaz Index The Calinski Harabaz index is based on the principle of variance ratio. This ratio is calculated between two parameters within-cluster diffusion and between cluster dispersion. The higher the index the better is clustering. The formula used is CH(k)=[B(k)W(k)][(n−k)(k−1)] where, n =  data points k = clusters W(k) = within cluster variation B(k) = between cluster variation. Davies Bouldin index Davies Bouldin index is based on the principle of with-cluster and between cluster distances. It is commonly used for deciding the number of clusters in which the data points should be labeled. It is different from the other two as the value of this index should be small. So the main motive is to decrease the DB index. The formula which is used to calculate the DB index. DB(C)=1Ci=1kmaxjk,jiDij Dij=di  + dj dij where, Dij= within-to-between cluster distance ratio for the ith and jth clusters. C  = no of clusters i,j  = numbers of clusters which come from the same partitioning Description of data Dataset used in this article is survey data done by taking random samples of the population. This data has been taken from the kaggle repository which is mentioned in the references. It contains 8 columns which contain different information about the surveyee. The description of columns is mentioned below. Marital status: marital status of surveyee married and unmarried which has already been label encoded 0 and 1.Sex: male or female is labeled encode as 0 and 1.Age: age of the surveyeeEducation: There are different education levelIncome: Annual income of the surveyeeOccupation: Any occupation or not Sample How to use clustering evaluation metrics Till now we learned about all the theoretical parts of the different clustering techniques. Let’s see how to implement them on data. Before implementing them, we need to build a clustering algorithm. Importing  libraries: import numpy as np import pandas as pd import matplotlib.pyplot as plt import seaborn as sns from sklearn.cluster import KMeans from sklearn.metrics import silhouette_score,calinski_harabasz_score,davies_bouldin_score Reading dataset and preprocessing: df=pd.read_csv('clustering data.csv') df.drop('ID', axis=1, inplace=True) df['log_income']=np.log(df[['Income']]) df.head() Transforming the income column because the salary could be in any currency. So we need to standardize the data for a non-biased output. Finding the value of K in K-Means Clustering: X=df[['Sex', 'Marital status', 'Age', 'Education', 'Occupation', 'Settlement size', 'log_income']] se = [] sil_scores = [] calinski_score = [] davies_score = [] index = range(2, 7) for i in index: kmeans = KMeans(n_clusters=i, random_state=42) labels = kmeans.fit_predict(X) se.append(kmeans.inertia_) sil_scores.append(silhouette_score(X, labels)) calinski_score.append(calinski_harabasz_score(X, labels)) davies_score.append(davies_bouldin_score(X, labels)) print('Intertia at K =', i, ':', kmeans.inertia_) print(\"Silhouette Coefficient: %0.3f\" % silhouette_score(X, labels)) print(\"Calinski-Harabasz Index: %0.3f\" % calinski_harabasz_score(X, labels)) print(\"Davies-Bouldin Index: %0.3f\" % davies_bouldin_score(X, labels)) print(\"---------------------------------------------------------------\") In the above output for every value of k four evaluation metrics have been calculated, k is the number of clusters. If we take a number of clusters as 3 (k=3) the values will be, WCSS (within the sum of squared error): 42744 Silhouette Coefficient: 0.616 Calinski-Harabasz Index: 4304.782 Davies-Bouldin Index: 0.563 Decreasing the WCSS is the key objective of K-Means clustering, but in addition to it, there are three valuation metrics that need to be taken care of. Silhouette coefficient should be nearer to +1, lower the value of DB index higher the performance. Let’s plot these values to have a clear vision about selecting the number of clusters. Plot Elbow curve: plt.plot(index, se, marker='o') plt.xlabel('Clusters') plt.ylabel('With in Sum of squared error') plt.show() At k=3 a sharp point could be observed in the above graph. Hence we can move ahead with k =3. In the second plot which is the silhouette score plot, it could be observed that for k=3 the silhouette score for each label is above average( the red dotted line) and also the width fluctuation of every label is almost uniform and is not present in other plots. Hence k=3 is optimal. Final model and evaluation metrics: kmeans = KMeans(n_clusters=3, random_state=42) labels = kmeans.fit_predict(X) print(\"Silhouette Coefficient: %0.3f\" % silhouette_score(X, labels)) print(\"Calinski-Harabasz Index: %0.3f\" % calinski_harabasz_score(X, labels)) print(\"Davies-Bouldin Index: %0.3f\" % davies_bouldin_score(X, labels)) Silhouette Coefficient: 0.552 Calinski-Harabasz Index: 5509.332 Davies-Bouldin Index: 0.566 Silhouette coefficient is 0.5 which is in the middle of 0 and 1 which is acceptable, DB index is also 0.5 which is in the middle but decreasing this would be better for the performance of the model, the CH index is high which is good. By these metrics, it can be concluded that the model is an average performer. Conclusion In this article, we have learned about theoretical as well as the practical implementation of different evaluation techniques for clustering algorithms and where to use these techniques. We have also gone through the implementation of these metrics with a real-world dataset. References Link for above codeSilhouette score documentationCalinski Harabaz IndexDavies Bouldin indexDataset","excerpt":"In this article, we will be learning about different performance metrics for clustering and implementation of them","categories":["AI Trends"],"tags":["clustering","Machine Learning","Python","Unsupervised Learning"],"author_name":"Sourabh Mehta","publish_date":"2022-03-23T12:00:00","publication_year":"2022","word_count":1189,"keywords":["data science","NumPy","TPU","AI","Machine Learning","clustering","Python","RAG","Seaborn","Matplotlib","R","Unsupervised Learning","Pandas"],"extracted_tech_keywords":["AI","data science","Pandas","NumPy","Matplotlib","Seaborn","RAG","TPU","Python","R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/a-tutorial-on-various-clustering-evaluation-metrics\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":26885,"title":"Inside Cloud AutoML: Google’s New Marketing Platform To Drive Cloud Revenue","content":"If there is one big takeaway from the recently-concluded Google Cloud Next 2018 conference, is how the tech behemoth wants to position itself as a strong contender in the hybrid cloud business dominated by Amazon Web Services and Microsoft. This was evident in Google’s embrace of private, hybrid, edge and multi-cloud computing. Another key announcement was about the company expanding Cloud AutoML — the machine learning platform it announced at Google I\/O last year — into new areas like Vision, Natural Language and AutoML Translation. This comes at a time when AWS cloud revenue increases 48.9 percent in the second quarter. As the cloud market moves into various cloud architectures, companies are getting more competitive in the face of continuous change and introducing additional artificial intelligence products to draw in more customers. To help enterprises kickstart their AI projects easily, Google introduced Cloud AutoML beta which enables developers to train high-quality custom ML models with minimum effort and ML expertise. Cloud AutoML is a suite of ML products that enables developers with limited expertise to train high-quality models that are core to their business needs, by leveraging Google’s state-of-the-art transfer learning, and Neural Architecture Search technology. According to Google, AutoML puts in more than 10 years of research into the hands of developers and provides a simple GUI to users to build custom machine learning models by training, evaluating and deploying models based on data. So far, Google AutoML’s tagline has been that it helps democratise deep learning and Google’s vision has been, as put succinctly in a forum: Data in > architecture built > model out Given how deep learning is a computationally-intensive field, Google has lowered the entry barrier with AutoML for researchers who can now get groundbreaking results with their state-of-the-art-tools, plus bring down computation costs. Touting AutoML — Another Marketing Machine For The Future However, there is another side to this argument where it seems like Google is using AutoML as a talking point to sell GCP. Dr Fei Fie Li, Google’s Chief AI Scientist, says: “Currently, there are only thousands of deep learning researchers and a million data scientists across the world who can build advanced machine learning models. With Cloud AutoML, we want to help less-skilled developers build these models”. Besides, Google has always maintained that designing neural nets is extremely time intensive and requires an expertise that limits its use to a smaller community of scientists and engineers. That’s where AutoML can lower the entry barrier with Cloud APIs that can help build powerful custom machine models and deploy these models trained directly on Google Cloud. Cloud AutoML is already deployed by Disney and Zoological Society of London. According to Mike White, CTO, SVP, Disney Consumer Products and Interactive Media, Cloud AutoML’s technology is helping us build vision models to annotate our products with Disney characters, product categories, and colours. Is AutoML The Best Way To Mainstream DL? It is in Google’s vested interest to tout that automated ML can pave the way for a superior approach. While Google has put in significant research in automating tasks such as data preprocessing, ML modelling, algorithm selection and even model benchmarking, experts point out there are research papers like Jeff Dean co-authored Efficient Neural Architecture Search (ENAS) that propose a fast and inexpensive approach for automatic model design. In addition to this, there is Differentiable Architecture Search (DARTS) which addressed the scalability challenge and is faster than state-of-the-art techniques. For example, experts claim DARTS is faster than Google’s NASNet, the computer vision algorithm that achieved an 82.7 percent accuracy on the validation set. Since launching AutoML, Google researchers have been able to build and manage a computer vision algorithm called NASNet. The Google blog mentions that NASNet performs 1.2 percent better than all previously published results and is on par with the best unpublished result reported on Arxiv.org. As part of its aggressive push for Google Cloud Platform, Google Cloud AutoML Vision, Natural Language and Translation available in public beta delivers immediate results to businesses who lack key resources or knowledge of DL. Google has constantly pegged Neural Architecture Search as the key to make ML accessible to users and Cloud AutoML. However, a critical element for this is labelled data. Now, we all know the term AutoML — automated machine learning is a way to select models and optimise hyper-parameters. A Fast.ai blog suggests that even though Google mainstreamed the term AutoML with its 2017 research paper, there are other superior techniques that use fewer GPU hours. Also, the Neural Architecture Search approach pegged by Google and defined as “neural net to design neural net” is reinforcement learning that allows discovering new neural net architectures. In a way, it tore through Google’s AutoML that suggested the best replacement for ML expertise is data and 100 times the computation power.","excerpt":"If there is one big takeaway from the recently-concluded Google Cloud Next 2018 conference, is how the tech behemoth wants to position itself as a strong contender in the hybrid cloud business dominated by Amazon Web Services and Microsoft. This was evident in Google’s embrace of private, hybrid, edge and multi-cloud computing. Another key announcement […]","categories":["Global Tech"],"tags":["Automl","Cloud Platform"],"author_name":"Richa Bhatia","publish_date":"2018-08-04T07:51:14","publication_year":"2018","word_count":805,"keywords":["Automl","machine learning","artificial intelligence","AWS","AI","cloud computing","ML","computer vision","RAG","Aim","deep learning","Cloud Platform"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","computer vision","Aim","RAG","cloud computing","AWS"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/inside-cloud-automl-googles-new-marketing-platform-to-drive-cloud-revenue\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10099446,"title":"Myntra&#8217;s New Generative AI Tool Will Surprise You","content":"“I am going to Goa for a vacation, show me what I can wear,” we asked, and within seconds, this new tool called–MyFashionGPT was able to fetch results on shorts, t-shirts, sunglasses, hats and sunscreen. The brainchild of Myntra. This new tool enables users to search using natural language, alongside giving relevant suggestions based on customer queries. “This is a first-of-its-kind solution in e-commerce in India and possibly globally,” avered Myntra’s chief technology and product officer, Raghu Krishnananda, in an exclusive interaction with AIM. He said that they used ChatGPT for query understanding and then leveraged its own search infrastructure to fetch relevant and related products from its catalogue and show them as collections. It is working on more features that use generative AI, and will get launched in the near future on the platform. Tech Stack Myntra has been using both proprietary and open-source algorithms based on user cases. Krishnanda believes that open-source algorithms provide a quicker path to market. “When proprietary data is involved, using a hosted model would be the right approach where we train the open source models on Myntra-specific knowledge such as the product taxonomy.” The company has also developed its own AI models and combined multiple models to solve for specific use cases, especially in image science applications. Myntra is currently leveraging AzureAI services that give access to OpenAI models such as ChatGPT3.5, Dall-E, etc. “We are looking at privately hosted models as well as managed service models based on the use cases, and we will continue to have partnerships that serve this need.” Myntra’s latest tool MyFashionGPT, is integrated with ChatGPT3.5. “For text-related generative AI, we use ChatGPT3.5 and for image-related generative AI, we use Stable Diffusion-based models in conjunction with other internally developed models.” A number of other AI-based solutions in Myntra (non-generative AI) such as MyStylist have been developed in-house. Myntra’s Generative AI Prowess Synonymous with fashion and lifestyle, Myntra have been aggressively pushing through to bring generative AI onto their platform with the big picture of enhancing customer experience. “We have been using AI for more than five years now and see huge benefits. In that sense Myntra is an AI-first company,” said Krishnananda. Myntra’s adoption of AI-based solutions has not only helped customers but also sellers. “AI-based solutions such as trend identification, demand prediction, and others are helping sellers bring the right merchandise and assortment and stay ahead of the trend,” said Krishnananda. Furthermore, its inventory and route optimisation algorithms have helped improve logistics. While Myntra may have carved an AI niche in the fashion segment, other e-commerce players have also dived into the generative AI wave with a number of use cases (see below). Source: Paxcom Report Tech giant Amazon, who have already been implementing generative AI solutions on AWS and other services, are also working on bringing the same to its e-commerce vertical. The company is testing AI-generated customer review highlights that will present concise summaries of written reviews to aid a shopper in making quick purchasing decisions. To cater to small-scale sellers, last week Amazon launched its virtual assistant ‘सहAI’ (sahai). The AI tool will help sellers list their products online, analyse sales trends and thereby assist with improving sales. Challenges Galore Training and inference for very large proprietary generative AI models is a challenge for any company, and it is easier said than done. “We are working to take ‘smaller’ open-source models and fine tune them on our own data,” said Krishnananda, emphasising the safety and cost benefits of training the model, without revealing the names of the smaller models (namely, Llama, Vicuna, etc.) being used. Confident Myntra believes that it faces fewer challenges when it comes to adopting generative AI in their workflow. Krishnananda also spoke about how they are bringing adoption not just on the platform front, but also within teams. “The tech team is taking measures to democratise the use of Generative AI by providing internal APIs that the broader tech team can play with, as well as organise tech talks and knowledge sharing sessions,” he concluded, saying that they are building in-house frameworks for low cost fine-tuning and inference using GPUs.","excerpt":"Myntra recently launched a first-of-its-kind solution in e-commerce in India and possibly globally called MyFashionGPT. It’s going to change the way you shop entirely.","categories":["AI Highlights"],"tags":["Amazon","API","AWS","ChatGPT","CTO","e-commerce","Generative AI","myntra","review","Stable Diffusion"],"author_name":"Vandana Nair","publish_date":"2023-09-04T18:22:26","publication_year":"2023","word_count":690,"keywords":["Go","ChatGPT","API","e-commerce","AWS","CTO","Stable Diffusion","AI","OpenAI","R","Amazon","myntra","RAG","Aim","generative AI","review","Generative AI","Azure"],"extracted_tech_keywords":["AI","generative AI","ChatGPT","OpenAI","Aim","RAG","AWS","Azure","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/myntras-new-generative-ai-tool-will-surprise-you\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10013827,"title":"How To Scrape Websites Using Puppeteer &#038; Node.js","content":"Web scraping is the process of extracting information from the internet, now the intention behind this can be research, education, business, analysis, and others. Basic web scraping script consists of a “crawler” that goes to the internet, surf around the web, and scrape information from given pages. We have gone over different web scraping tools by using programming languages and without programming like selenium, request, BeautifulSoup, MechanicalSoup, Parsehub, Diffbot, etc. It makes sense why everyone needs web scraping because it makes manual- data gathering processes very fast. And web scraping is the only solution when websites do not provide an API and data is needed. In this demonstration, we are going to use Puppeteer and Node.js to build our web scraping tool. Node.js is an open-source server runtime environment that runs on various platforms like Windows, Linux, Mac OS X, etc. It is not a programming language. It uses JavaScript language as the main programming interface. It is free and capable of reading and writing files on a server and used in networking. Puppeteer is a Node library that provides a high-level API to control Chromium or Chrome browser over the DevTools Protocol. It runs headless by default but can be changed to run full (non-headless). It is built by Google. Using it, we can: Scare data from the internet.Create pdf from web pages.Take screenshots.Create automation testing.Create a server-side rendered version of the application.Track page loading process.Automate form submission. Since the launch developers have published two versions Puppeteer and Puppeteer-core. Puppeteer-core is a lightweight version of Puppeteer for launching your scripts in an existing browser or for connecting it to a remote one. It follows the latest maintenance LTS version of the Node framework. Let’s see the browser architecture of Puppeteer. As shown in the below diagram, faded entities are not currently represented in the Puppeteer framework. credit : puppeteer official docs Root node Puppeteer communicates with the browser by using dev tools.Browser instances have multiple browser contexts.Browser context defines a browsing session and owns multiple pages.Page can have at least one mainframe.The frame has at least one execution context, i.e. default execution context where javascript is executed.Worker interacts with WebWorkers. Getting Started We are going to scrape data from a website using node.js, Puppeteer but first let’s set up our environment. We need to install node.js as we are going to use npm commands, npm is a package manager for javascript programming language. It is a subsidiary of GitHub. It is a default package manager which comes with javascript runtime environment Node.js. Download Node.js from hereinstall Initializing Project Follow these steps to initialize your choice of a directory with puppeteer installed and ready for scraping tasks. Create a Project folder mkdir scraper cd scraper Initialize the project directory with the npm command npm init Like git init it will initialize your working directory for node project, and it will present a sequence of prompt; just press Enter on every prompt, or you can use : npm init -y And it will append the default value for you, saved in package.json file in the current directory and your output will look something like this: Now use npm command to install Puppeteer: npm i puppeteer Note: When you install Puppeteer, it will download the latest version of Chromium (~205MB Mac, ~282MB Linux, ~154.2 MB Win) and it is recommended to let the chromium download to see puppeteer work fine with the API. Using these three steps, you can initialize puppeteers in your node environment! Quickstart For example – the following script will navigate to https:\/\/analyticsindiamag.com\/ and save a screenshot as output.png const puppeteer = require('puppeteer'); (async () => { const browser = await puppeteer.launch(); const page = await browser.newPage(); await page.goto('https:\/\/analyticsindiamag.com\/'); await page.screenshot({path: 'output.png'}); await browser.close(); Explanation: cons puppeteer = require(‘puppeteer’); is used to import puppeteer, it is going to be the first line of your scraper.await puppeteer.launch(); is used to initiate a web browser or more specifically to create a browser instance you can open your browser in headless mode and non- headless mode using {headless:false} by default its true that means it will run browser processes in the background.await puppeteer.launch({ headless: false}); opens a chromium-browser.await puppeteer.launch({ headless: true}); does not open browser.We use await to wrap method calls in an async function, which we immediately invoke.newPage() method is used to get the page objectgoto() method to surf that URL and load it in the browser.screenshot() takes a path argument and returns a screenshot of the webpage in 800×600 px form in the local directory.Once we are done with our script, we call close() method on the browser. Note: The initial default page size is set to 800×600 px, you can change that using : Page.setViewport() And can even extract a full pdf out of it by using a script like this: const puppeteer = require('puppeteer'); (async () => { const browser = await puppeteer.launch(); const page = await browser.newPage(); await page.goto('https:\/\/news.ycombinator.com', {waitUntil: 'networkidle2'}); await page.pdf({path: 'hackernews.pdf', format: 'A4'}); await browser.close(); })(); Scraping data from Wikipedia One of the use-cases we can try to find the true potential of Puppeteer is to scrape all the covid-19 data and export it into a JSON file. This example is taken from here, scraping data with node js can be a little trickier in terms of coding a scraper but outputs are accurate and fast, after trying these examples a couple of times you will get to know this framework better. Import puppeteer and file system(fs) const puppeteer = require('puppeteer'); const fs = require('fs') Launch browser and open Wikipedia page of all covid data country wise const scrap = async () =>{ const browser = await puppeteer.launch({headless : false}); const page = await browser.newPage(); await page.goto('https:\/\/en.wikipedia.org\/wiki\/2019%E2%80%9320_coronavirus_pandemic_by_country_and_territory', {waitUntil : 'domcontentloaded'}) \/\/ navigate to url and wait for page loading Inspect the webpage is pretty simple, as we discussed in an earlier tutorial for web scraping like here, just go to the website and right-click -> inspect to enter into developer mode to see the source code of the website, as shown below, on selecting the first row we can see in which tag we are having our data, like in this case div class “covid19-container” contains the table with id “thetable ” this is our target table. page.$$eval(selector, pageFunction[,….args]) return an array of all the elements that matches with our argument string we passed, (‘div#covid19-container table#thetable tbody tr’,(trows) is going to extract all the rows from covid19 container. For more information, you can visit here. const scrap = async () =>{ const browser = await puppeteer.launch({headless : false}); \/\/browser initiate const page = await browser.newPage(); \/\/ opening a new blank page await page.goto('https:\/\/en.wikipedia.org\/wiki\/2019%E2%80%9320_coronavirus_pandemic_by_country_and_territory', {waitUntil : 'domcontentloaded'}) \/\/ navigate to url and wait until page loads completely const recordList = await page.$$eval('div#covid19-container table#thetable tbody tr',(trows)=>{ let rowList = [] trows.forEach(row => { let record = {'country' : '','cases' :'', 'death' : '', 'recovered':''} record.country = row.querySelector('a').innerText; \/\/ (tr < th < a) anchor tag text contains country name const tdList = Array.from(row.querySelectorAll('td'), column => column.innerText); \/\/ getting textvalue of each column of a row and adding them to a list. record.cases = tdList[0]; record.death = tdList[1]; record.recovered = tdList[2]; if(tdList.length >= 3){ rowList.push(record) } }); return rowList; }) console.log(recordList) await page.screenshot({ path: 'screenshots\/wikipedia.png' }); \/\/screenshot browser.close(); First, eval will extract the rows from the table having id thetableTo select using id always use ‘#’ and postfix with id of the parent element.To select using class use “.” and postfix with the class of parent element.Now the rows are extracted from the table. We are going to iterate them all over the table.Country name for covid report is present in <a> ankle tag and other are present in <td> data so to extract from there we are going to use : record.country = row.querySelector('a').innerText; const tdList = Array.from(row.querySelectorAll('td'), column => column.innerText); Here we used querySelector() which returns the first element that matches the selector. Alternatively, we can use querySelectorAll() which returns all the elements that match the selector. Note: if any other table having tr as the same path then it will return rows of that table too. The other data from list will be extracted by using : record.cases = tdList[0]; record.death = tdList[1]; record.recovered = tdList[2]; rowList.push(record) will push the record inside the rowList=[]. Store the output fs.writeFile('covid-19.json',JSON.stringify(recordList, null, 2),(err)=>{ if(err){console.log(err)} else{console.log('Saved Successfully!')} }) }; scrap(); Fs is a filesystem library used by the node. JSON.stringify() converts an array of objects to string and then writes it to a file using the file system module. Output A full dataset in JSON object ready for further processing. Conclusion We learned Puppeteer is a powerful library for automating things, web scraping, taking screenshots, saving pdfs, debugging, and it supports non-headless environments too just like selenium. We saw how our web crawlers scraped data from Wikipedia and then saved it in a JSON file. In short, you learned a new way to automate things and scrape data from the internet. Puppeteer has quite a lot of functions that were not discussed in this tutorial. To learn more, check out Puppeteer’s official documentation.","excerpt":"Web scraping is the process of extracting information from the internet, now the intention behind this can be research, education, business, analysis, and others. Basic web scraping script consists of a “crawler” that goes to the internet, surf around the web, and scrape information from given pages. We have gone over different web scraping tools […]","categories":["Deep Tech"],"tags":["data extraction","data structure using java","Node.js","Web Scraping"],"author_name":"Mohit Maithani","publish_date":"2020-12-11T15:00:00","publication_year":"2020","word_count":1515,"keywords":["data extraction","Go","TPU","AI","Web Scraping","data structure using java","Git","Ray","Node.js","analytics","JavaScript","GitHub","R","Java"],"extracted_tech_keywords":["AI","analytics","Ray","TPU","R","JavaScript","Go","Java","Git","GitHub"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/puppeteer-web-scraping\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10141804,"title":"Meet the Indian AI Startup Quietly Taking Over the Enterprise World","content":"Indian AI startup TWO is quietly marking a mark in the fast-emerging artificial intelligence market. In an interview with AIM, founder Pranav Mistry revealed that the company generated $4 million in revenue this quarter and expects $20 million for next year. This development comes as TWO AI’s SUTRA, a series of multilingual online GenAI models, added a new feather in its cap. The company claims it outperformed GPT-4o, Llama 3.1, and Indic LLMs, including G42’s Nanda, Sarvam’s OpenHathi and AI4Bharat’s Airavata, and leading in over 14 Indian languages. Earlier this year, the company launched ChatSUTRA, a ChatGPT-like chatbot. Mistry shared that the platform currently has over 600,000 unique users. Unlike other startups, TWO AI targets only big enterprise customers instead of pursuing the consumer market. “Jio is one of our major enterprise customers, and we also work with clients like Shinhan Bank and Samsung SDS in Korea,” Mistry said. He further revealed that the company has started partnering with companies like NVIDIA and Microsoft from a technology perspective, and is working with them as well. “We are targeting India, Korea, Japan, and some parts of Southeast Asia, like Vietnam, specifically the central region. APAC (Asia-Pacific) is one of the key markets that we are always going to focus on,” Mistry added. Recently, TWO hosted Mukesh Ambani, chairman of Reliance Industries, and Akash Ambani, chairman of Reliance Jio Infocomm, at their US office. Over a cup of tea, they discussed the evolving role of AI in India and beyond. Without taking names, Mistry revealed that in India one of the largest banks and financial sectors is among the customers the company will onboard. “Our solutions are in high demand, particularly in industries like finance, services, and retail,” he said. SUTRA’s business model focuses on providing high-touch, customised solutions for a select group of large enterprises. “We don’t need 100 customers,” said Mistry. “We need 10 good customers.” He explained that by using this method, they are reaching billions of customers, as these enterprises already have millions of customers. “The focus is to become not the OpenAI of the world, not just an application layer company, but an AI solutions company, going after large enterprises and helping them solve problems with AI,” he added. He shares his goal of following a path similar to that of Palantir’s. What’s Next? Mistry revealed that the company’s next project is predictive AI. “Predictive AI is game-changing for these data-dependent industries. From manufacturing to finance, governance, and energy sectors, everyone can really leverage the decision-making capability of forecasting,” he explained. The model is called Sutra Predict. Mistry pointed out that it is a small model which is trained on trillion data points and time series data entries. “The model is small because the architecture of this one is much easier than the text-based ones, and it is already showing great results in some particular domains that our customers are already trying.” He explained that time series predictive models are a specific type of statistical model used to analyse and forecast data points that are collected over time. They are built to identify patterns, trends, and seasonal variations within the data to make predictions about future values. Mistry explained that with the advent of transformers, models can now process and integrate any kind of data, as seen in predictive models like Google TimesFM and Amazon Chronos. Highlighting a real-world application, he shared that an EV battery diagnostic company in India is using Sutra Predict to identify fire risks by monitoring temperature and voltage fluctuations. Tackling GPU challenge “In India, no one had access to the level of GPU clusters we needed,” Mistry said. The SUTRA team overcame this limitation by porting their models to run on CPU clusters. Despite the challenges, he said the team was able to scale and serve up to 100 billion customers. Moreover, Mistry shared that they were the first ones to catch on to the trend of 1-bit LLMs. Notably, Microsoft recently introduced BitNet.cpp, an inference framework for 1-bit LLMs, enabling fast and efficient inference for models like BitNet b1.58. Mistry said they have successfully adapted the SUTRA model to work with 1-bit weights, allowing it to run as a lightweight model on CPUs. Moreover, in partnership with NVIDIA, the company launched SUTRA-OP, offering systems like the NVIDIA DGX (Deep GPU Xceleration) box equipped with powerful GPUs for demanding AI tasks. For customers requiring lighter and more cost-effective solutions, SUTRA also provides its hardware options, including SUTRA OP2, OP4, and OP8, which are available for lease. “Customers are not purchasing this, they are leasing it from us. It’s a monthly lease for both the OP and the SUTRA solutions,” said Mistry. The company recently launched a voice-to-voice AI model called Sutra HiFi. Using a dual diffusion transformer architecture, the model effectively separates distinct voice tones from language-specific accents, promising a better voice interaction quality. “Sutra HiFi brings the ability to interpret conversations seamlessly in languages that we care about. Currently, it supports 12 languages that we have tested properly,” Mistry said. He suggested that Sutra HiFi can easily empower applications in India or any multilingual market while keeping the cost low and accuracy high. Discussing Infosys co-founder Nandan Nilekani’s point of view that India should be the use case capital of AI, Mistry said that he has a slightly different perspective. “India must focus on building the fundamental AI capabilities because we don’t want to become dependent on someone else in the future, as data is one of the gold mines in AI,” he concluded.","excerpt":"SUTRA yet again emerges as #1 in Indian language AI, the research report claims.","categories":["IT Services"],"tags":["Editors Picks","TWO AI"],"author_name":"Siddharth Jindal","publish_date":"2024-11-28T07:53:41","publication_year":"2024","word_count":923,"keywords":["ChatGPT","GenAI","artificial intelligence","TWO AI","OpenAI","AI","GPT-4o","ML","Transformers","RAG","Editors Picks","Aim"],"extracted_tech_keywords":["AI","artificial intelligence","ML","GenAI","GPT-4o","ChatGPT","OpenAI","Aim","Transformers","RAG"],"url":"https:\/\/analyticsindiamag.com\/it-services\/meet-the-indian-ai-startup-quietly-taking-over-the-enterprise-world\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":61572,"title":"Renowned Princeton Mathematician &#038; Inventor Of The ‘Game Of Life’, John Horton Conway Dies From COVID-19","content":"A renowned Princeton mathematician, forever favoured by programmers, and a man who invented the Game of Life, John Horton Conway has died, at the age of 82, due to COVID-19 infection. Conway was a mathematician, who has made contributions in the areas like finite groups, knot theory, number theory, combinatorial game theory and coding theory, and has also invented a new system of numbers, aka ‘the surreal numbers.’ The surreal numbers are a continuum of numbers that include not only real numbers but also the infinitesimal and the infinite numbers. Alongside he also invented a naming system for exceedingly large numbers, the Conway chained arrow notation. In a recent tweet by Stephen Wolfram — a British-American computer scientist stated that “After many years of computation and lots of output, the JHC function has sadly now halted. RIP John Conway (1937-2020).” After many years of computation and lots of output, the JHC function has sadly now halted. RIP John Conway (1937-2020) https:\/\/t.co\/YsONXXyCYm pic.twitter.com\/adct8DxHEW— Stephen Wolfram (@stephen_wolfram) April 12, 2020 Conway has also been widely known for his contributions to combinatorial game theory, a theory of partisan games. Over his long career, he has even invented several games; out of which the most popular one has been the Game of Life, in an early example of a cellular automaton. Conway plays the Game of Life, which he invented in 1970. Photograph: Kelvin Brodie\/the Sun Persi Diaconis, a mathematician at Stanford University once said in an interview — “John Conway is a genius. And the thing about John is he’ll think about anything.… He has a real sense of whimsy. You can’t put him in a mathematical box.” For his contributions, the charismatic Conway has also received several awards. A few of the renowned ones are — The Berwick Prize in 1971; selected as ‘fellow’ in the Royal Society in 1981; the first winner of the Pólya 1987 award; winner of the Mathematics Nemmers Prize in 1998); and the Lewroy P. Steele Prize in 2000.","excerpt":"A renowned Princeton mathematician, forever favoured by programmers, and a man who invented the Game of Life, John Horton Conway has died, at the age of 82, due to COVID-19 infection. Conway was a mathematician, who has made contributions in the areas like finite groups, knot theory, number theory, combinatorial game theory and coding theory, […]","categories":["AI News"],"tags":["Coronavirus","covid-19","game theory"],"author_name":"Sejuti Das","publish_date":"2020-04-13T14:28:36","publication_year":"2020","word_count":332,"keywords":["TPU","covid-19","AI","programming_languages:R","game theory","Coronavirus","R"],"extracted_tech_keywords":["AI","TPU","R","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/renowned-princeton-mathematician-inventor-of-the-game-of-life-john-horton-conway-dies-from-covid-19\/","complexity_score":3,"technical_depth":4,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10002198,"title":"MedTech Startup InnAccel Is Revolutionizing The Indian Healthcare Market With Innovative Medical Devices","content":"Sustaining health has always been one of the major factors in our day-to-day life. The field of MedTech is making significant evolution right from the artificial organs to prosthetic robotic limbs. InnAccel, a Bangalore-based startup identifies the unmet healthcare needs and develop unique novel globally certified medical devices in India. Analytics India Magazine caught up with Siraj Dhanani, CEO and Founder of InnAccel to gain insights on how the company is using emerging technologies like machine learning and artificial intelligence. The company started out, over six years ago, as a sector-specific incubation and acceleration platform to support MedTech innovation. In 2014, the company seeded the first three teams to create novel technologies in ENT, critical care, and maternal care space respectively and about 2 years ago, it has been pivoted from the incubation model to a pure-play MedTech company, with three R&D divisions. Technology Behind The Products InnAccel creates novel medical technologies and patented devices which encompass disciplines such as mechanical, electrical, and electronics engineering, software programming, machine learning\/artificial intelligence, and product and UI design. One of the products, a next-generation labour monitoring technology, has leveraged machine learning extensively which has enhanced the usability of the device, and reduced the required skill levels for using the device, and eliminated erroneous readings. The training of the model is customised based on the specific needs of the customers. The Flagship Products InnAccel has developed products in three core verticals- critical care, maternal care, and ENT. The company’s flagship product is VAPCare, a device intended to reduce the likelihood of lung infection in patients on a ventilator. The core technology for this product has been globally patented, and it has shown strong effectiveness in clinical trials. The product has been approved by the US FDA and will be globally launched in the coming months. Another product, Saans which is a transformational product makes CPAP therapy available in all treatment centers, and during transport. It is portable, electricity and infrastructure independent, and has a patented manual powering mode. Also, the product has been honoured with a display in Harvard Medical School’s Technology Showcase, as an example of high-impact innovation. The Secret Formula Behind Affordable Devices Dhanani stated, “At InnAccel, we spend 8-9 months understanding a particular problem, and the treatment ecosystem before we start working on a solution. A big part of this is understanding the affordability of our target customer- who is typically the low and middle-income Indian. So affordability, in terms of a target price of the product, is a defining product criterion when we start building the product. Our engineering process is rigorous and disciplined- I am happy to say we have achieved the target price on all our products- ensuring affordability for the average Indian.” Also added, “InnAccel incorporates user feedback at every stage of product development. From initial concepts to prototypes to the final product, product users (typically doctors and nurses) give us feedback and help us refine the product. Such user feedback has been extremely critical in creating products that are user-friendly and intuitive.” Hiring At InnAccel While hiring individuals, the company focuses on the candidate’s customer-centric attitude including the top-notch engineering skills. Besides that, the company primarily look for individuals who can spend time in the hospital to identify or validate problems which are worth solving. Roadmap According to Dhanani, in the present scenario, the company is at a very exciting stage in its journey. It has successfully created, and globally certified, a portfolio of unique MedTech assets. The task at hand is to successfully commercialise these products in India, and find MNC partners to commercialise them globally. Furthermore, the company is also working on the next set of products and technologies with the goal of developing 20 high-impact products by 2025 and transforming healthcare in India and globally.","excerpt":"Sustaining health has always been one of the major factors in our day-to-day life. The field of MedTech is making significant evolution right from the artificial organs to prosthetic robotic limbs. InnAccel, a Bangalore-based startup identifies the unmet healthcare needs and develop unique novel globally certified medical devices in India. Analytics India Magazine caught up […]","categories":["AI Features"],"tags":["medtech","Startups"],"author_name":"Ambika Choudhury","publish_date":"2019-06-04T14:30:39","publication_year":"2019","word_count":632,"keywords":["Go","machine learning","artificial intelligence","AI","innovation","medtech","RAG","analytics","GAN","Startups","R","startup"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","analytics","RAG","R","Go","GAN","innovation","startup"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/medtech-startup-innaccel-is-revolutionizing-the-indian-healthcare-market-with-innovative-medical-devices\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10057604,"title":"Top online resources to learn Active Learning","content":"A key requirement of machine learning is to label the data correctly to ensure the best results, but the process is long and time-consuming. This also brings about an issue when dealing with extremely large data sets in unsupervised or semi-supervised learning. The saviour here is active learning with strategies that assist developers in prioritising the data and selecting the most useful samples to label to have the highest training impact. Furthermore, it promises to reduce the samples needed by choosing the right examples. Various strategies can be used depending on the applications and needs of the model. However, when it comes to learning active learning, the practice is generally a part of bigger machine learning modules, which is why we have created a one-stop guide to mastering active learning online through resources varying from online video tutorials to blog posts and academic papers. YouTube Active (Machine) Learning – Computerphile Computerphile is a popular YouTube channel that discusses computer science-related topics. Their tutorial on active learning is taught by Dr Michel Valstar, who holds a PhD in Computing and is currently a professor at the University of Nottingham. The tutorial is a foundational element for the basics of active learning, taught through diagrams and illustrations of the concepts. ICML 2019 Tutorial Session on Active Learning from Theory to Practice ICML, the International Conference on Machine Learning, is one of the fastest-growing AI conferences that discuss the latest academic papers. During their 2019 conference, Robert Nowak and Steve Hanneke taught the basics of active learning theory and the popular algorithms to apply (the video is now available online). In addition, the tutorial focuses on sound active learning algorithms and how they can be used to reduce the labels on training data. Robert Nowak holds the Nosbusch Professorship in Engineering at the University of Wisconsin-Madison. Steve Hanneke is a Research Assistant Professor at the Toyota Technological Institute in Chicago, specialising in AI and ML. Active Learning in ML Applied AI is a great resource for learning AI\/ML online through core concepts and real-life applications. The channel’s collective views cross 12 million and are popular for the basic concepts’ thorough teachings. Their tutorial on active learning in ML breaks down the principles of the concept along with real-life examples and mathematical explanations. Jan Freyberg: Active learning in the interactive python environment PyData is an educational program of NumFOCUS, a US-based not for profit organisation that provides a forum for the international community of data science to share their ideas through conferences. Speaking at one of their events is Jan Freyberg, a machine learning software engineer at Google Health. In a detailed talk, Freyberg discusses active learning in the interactive Python environment, given the ease and comfort in the ecosystem. Active Learning: Machine Learning Made Simple Devansh is a Computer Science and Computational Math Double Major at the Rochester Institute of Technology. Through this YouTube tutorial, he comprehensively discusses the basics of active learning, its works and compares it to SSL and GANs. He further explains the concept in detail regarding its use and active learning’s acquisition function. Machine Learning | Active Learning Ranji Raj, holding a masters degree in data science, takes on Youtube to publish tutorials and classwork related to machine learning. His video on active learning gives an in-depth introduction to the subject while discussing important concepts through diagrams and demonstrations. Raj also has consequent coursework on his GitHub page for data scientists interested in learning further. WEBINAR | What does “active learning” mean in AI? Save time and money on data annotation | Beginner Scaleway is a French cloud computing company that creates Youtube videos consisting of short machine learning tutorials and real-world applications. In their webinar on active learning, the company collaborated with Kairntech, an AI modelling and dataset creation platform, to discuss the various applications of active learning. The video discusses training datasets and how active learning can be applied for classification. It also glossed over common issues and how to overcome them. Blog tutorials Active Learning Tutorial Ori Cohen is a PhD holder in CS, currently working as a senior director of data science at New Relic. His ‘Towards Data Science’ blog post on active learning is an extensive tutorial that discusses the various scenarios possible while using active learning, the algorithms that can be used, the sample selection methods and the codings used for all. Guide to Active Learning in Machine Learning (ML) – DataCamp A blog post on Data Camp, an online interactive learning platform, explains in depth the A-Zs of active learning in a moderate level of difficulty. The tutorial discusses the concept in detail with definitions, examples and visuals, and teaches how one can apply active learning on their datasets through a particular example. Active Learning Tutorial with Python Written by a CS and EE student at IIT, India, this post is an in-depth tutorial on using active learning with Python. The tutorial is technical, explaining the code and its concepts through codes and steps. In addition, the post discusses various inputs, outputs, and the Python codes needed to apply active learning correctly. A Proactive Look at Active Learning Packages Alexandre Abraham, a senior research scientist at Dataiku and a Ph D holder in computer science, has written an extensive tutorial on active learning packages on his Medium blog post. The blog post analyses the active learning packages available through a feature comparison, their covered approaches, and their coding aspects. There are three main packages and different methods that data scientists can leverage. Papers Submodularity in Data Subset Selection and Active Learning The paper in discussion is written by Kai Wei, an assistant professor at UCLA, Rishabh Iyer, an assistant professor at the University of Texas, and Jeff Bilmes, a professor at the University of Washington. Their paper studies the problem of selecting a subset of data to train a classifier and how individuals can apply the active learning framework to mitigate the issue. Online courses Machine Learning Data Lifecycle in Production The DeepLearning.AI course in ML data lifecycle has a fourth module, tagged ‘Advanced Labeling, Augmentation and Data Preprocessing’, that focuses on semi-supervised learning, dataset labelling, and the role played by active learning within. The instructor, Robert Crowe, works at TensorFlow by Google and has multiple degrees in AI, ML and data science.","excerpt":"Their tutorial on active learning in ML breaks down the principles of the concept along with real-life examples and mathematical explanations.","categories":["AI Trends"],"tags":[],"author_name":"Avi Gopani","publish_date":"2022-01-04T13:00:00","publication_year":"2022","word_count":1051,"keywords":["data science","machine learning","TPU","AI","cloud computing","ML","RAG","Python","TensorFlow","R"],"extracted_tech_keywords":["AI","machine learning","ML","data science","TensorFlow","RAG","cloud computing","TPU","Python","R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-online-resources-to-learn-active-learning\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10077853,"title":"L&#038;T Infotech Records 41.4% YoY Growth for AI &#038; Analytics Services","content":"Last week, Larsen & Toubro Infotech (LTI) announced its second-quarter results for FY23. The company recorded revenue growth of 21.6% in constant currency and a profit of 18.8 per cent—the revenue growth in USD stood at 18.1 per cent. Sudhir Chaturvedi, executive board member and president of sales at L&T Infotech, said they are excited about the proactive conversations they are having with their customers and see increased traction in the cloud and analytics space. He said that the company’s growth is fueled by its strong pipeline, alongside sustained growth in its net headcount addition. Revenue breakdown by service offerings L&T Infotech provides a gamut of services to customers, accelerating their digital transformation with LTI’s Mosaic platform, enabling mobile, social, analytics, AI, IoT and cloud journeys. The service offered by L&T Infotech can be further categorised into ADM (application development management) and Testing, enterprise solutions, cloud infra and security, analytics, AI and cognitive, and enterprise integration and mobility. In FY23 Q2, the company recorded 33.5 per cent revenue growth from ADM and testing services, followed by enterprise solutions and analytics and AI services at 28.7 per cent and 14.5 per cent, respectively. On the other hand, its revenue from cloud infrastructure and security witnessed a decline at 7.3 per cent year-on-year growth in USD. On the other hand, its revenue from analytics, AI and cognitive services noted a 41.4 per cent growth, followed by enterprise integration and mobility at 34.6 per cent year-on-year growth in USD. With operations in 33 countries, L&T Infotech recorded 69 per cent revenue growth coming from North America and 15.5 per cent from Europe. Meanwhile, India accounted for 7.4 per cent revenue growth. In FY23 Q2, L&T Infotech added about 22 clients, compared to last quarter, where it added 29 customers. The company has about 500 active customers to date. Currently, the company’s headcount stands at 50,000 employees. At the earnings call, it announced its plans to hire 6,500 freshers in FY23. Attrition balloons L&T Infotech recorded the highest-ever attrition at 24.3 per cent, compared to last year, where it stood at 19.6 per cent. L&T Infotech’s COO Nachiket Deshpande said that the company expects this number to cool off considering incoming supply and slowdown in hiring. “We have a few quarters to go before attrition comes down materially,” he added. Read: Finally, A Sigh of Relief for Attrition Battered Indian IT Meanwhile, the merger of L&T Infotech and Mindtree is also underway. In May 2022, the Boards of Directors of LTI and Mindtree approved a composite scheme of amalgamation under the Larsen & Toubro Group. LTI and Mindtree combine strengths to create an efficient and scaled-up IT services provider exceeding $3.5 billion.","excerpt":"With operations in 33 countries, L&T Infotech recorded 69 per cent revenue growth coming from North America and 15.5 per cent from Europe","categories":["IT Services"],"tags":["L&amp;T Infotech"],"author_name":"Amit Naik","publish_date":"2022-10-21T15:00:00","publication_year":"2022","word_count":448,"keywords":["Go","programming_languages:R","AI","L&amp;T Infotech","digital transformation","programming_languages:Go","Git","analytics","R"],"extracted_tech_keywords":["AI","analytics","R","Go","Git","digital transformation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/lt-infotech-records-41-4-yoy-growth-from-ai-analytics-services\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10063145,"title":"No free lunch theorem in Quantum Computing","content":"Entanglement, as Albert Einstein referred to as ‘spooky action at a distance’, refers to a phenomenon by which a particle can ‘know’ something about another particle even if a huge distance separates them. In quantum entanglement, these two particles are said to be so intertwined that one can infer not only the property of the partner particle but also influence it. When studied by Einstein, he found the phenomenon so baffling it was originally taken as evidence that quantum mechanic models were incomplete. Register for Data Engineering Summit 2022 >> Quantum computers have components called the qubits, which are linked through entanglement, helping grow their computational power exponentially. This is particularly useful in modern encryption that is used in banking, and other sectors where securing data is of fundamental importance. Recent studies have shown that quantum computing might also help in boosting machine learning. Another application is simulating quantum systems. In machine learning, the no-free lunch theorem suggests that all optimisation algorithms perform equally well when their performance is averaged over many problems and training data sets. With the rise of quantum machine learning, it becomes imperative to ask whether there is a quantum analogue of this theorem that would restrict a quantum computer’s capability to learn a unitary process with quantum training data. This was recently studied by a group of researchers who documented their learnings in a paper titled “Reformulation of the No-Free-Lunch Theorem for Entangled Datasets”. In their paper, the authors showed that entangled datasets violate the classical no free lunches theorem. No-free lunches in quantum learning The no free lunch theorem entails that a machine learning algorithm’s average performance is dependent on the amount of data it has. “Industry-built quantum computers of modest size are now publicly accessible over the cloud. This raises the intriguing possibility of quantum-assisted machine learning, a paradigm that researchers suspect could be more powerful than traditional machine learning. Various architectures for quantum neural networks (QNNs) have been proposed and implemented. Some important results for quantum learning theory have already been obtained, particularly regarding the trainability and expressibility of QNNs for variational quantum algorithms. However, the scalability of QNNs (to scales that are classically inaccessible) remains an interesting open question,” the authors write. This also suggests a possibility that in order to model a quantum system, the amount of training data might also need to grow exponentially. This threatens to eliminate the edge quantum computing has over edge computing. The authors have discovered a method to eliminate the potential overhead via a newfound quantum version of the no free lunch theorem. Their study showed that adding more entanglement to quantum computing would lead to exponential scale-up, which was verified using Rigetti’s Aspen-4 quantum computer. The researchers suggested an extra set of ‘ancilla’ qubits with the quantum system that can help the quantum machine learning circuit interact with many quantum states in the training data simultaneously. Study co-author Andrew Sornborger said, “Trading entanglement for training states could give huge advantages for training certain types of quantum systems.” The authors believe that one of the applications of this work is in black box uploading, where we learn a model of quantum experiment and then study it on a quantum computer without requiring to do repeated experiments. Challenges with the study The major issue is the complexity of obtaining the entangled training data; this usually depends on the mode of access to the data. When the user has physical access to the target unitary, it is advantageous to input a state with the reference system so that the user can generate training data with entanglement. As compared to input with no entanglement, this procedure decreases the average risk more efficiently. Secondly, while the study assumes perfect training, the writers caution that it was possible that exponential scaling may have training difficulty. In the past, several strategies have been proposed to avoid barren plateau (gradients that vanish exponentially in the number of qubits) in quantum neural networks; however, this remains an active area of research. Read the full paper here.","excerpt":"In machine learning, the no-free lunch theorem suggests that all optimisation algorithms perform equally well when their performance is averaged over many problems and training data sets.","categories":["AI Features"],"tags":["Quantum Computer","Quantum Computing"],"author_name":"Shraddha Goled","publish_date":"2022-03-21T13:00:00","publication_year":"2022","word_count":674,"keywords":["Quantum Computing","Go","machine learning","AI","neural network","Quantum Computer","Scala","RAG","data engineering","edge computing","quantum machine learning","R"],"extracted_tech_keywords":["AI","machine learning","neural network","quantum machine learning","RAG","edge computing","R","Go","Scala","data engineering"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/no-free-lunch-theorem-in-quantum-computing\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10061877,"title":"Prominent tech companies that originated from Ukraine","content":"Ukraine is in the midst of a war. But, pre-invasion, the country was a breeding ground for tech startups and has given birth to many notable global tech companies. The Ukrainian IT industry consists of over 4,000 local IT service companies, and more than 110 leading global companies have established subsidiaries in the country. Samsung, Microsoft, Ring, Snap, Magento, Plarium, Boeing, Siemens, Ericsson – are among the few tech companies that have established their R&D sector in Ukraine. Ukraine has been one of Europe’s leading countries in terms of engineering graduates, producing twice as many annually as nations such as Britain and Poland. The country produces over 130,000 engineering graduates and 16,000 IT graduates annually. Kyiv, Ukraine’s capital, is the largest IT hub in the country and is home to over 1,000 startups and product companies. Other Ukrainian cities such as Kharkiv, Odesa, Dnipro, Lviv, and Vinnytsia are also developing major centres of IT activity. The country has been the birthplace of major tech companies that have created a global impact with their products and solutions. Let’s look at a few prominent tech companies that originated from Ukraine. Grammarly, a grammar solution platform headquartered both in Ukraine and America, was founded by three Ukrainian entrepreneurs, Max Lytvyn, Alex Shevchenko and Dmytro Lider. The company is currently valued at USD 13 billion. Millions of users rely on the input from the app to ensure their writing is of top-notch quality and free from any grammatical errors. Gitlab, a popular DevOps collaboration platform based out of Ukraine, was founded by Dmitriy Zaporozhets and Sid Sijbrandij in 2014. The company helps teams improve cycle time from weeks to minutes, reduce development costs, and boost developer productivity. Traces.AI, a computer vision company based out of Ukraine, has built an AI-based technology to analyse over 2,000 physical traits of a person using security cameras. The company was founded by two Ukrainian techies, Veronica Yurchuk and Kostyantyn Shysh, in 2014. The company recently had developed a tool to track people who have been impacted by COVID-19 and can be potential carriers of the coronavirus. People.AI, an artificial intelligence startup based out of Ukraine, helps sales teams work efficiently with their clients. In addition, the company assists enterprise growth by harnessing the power of AI. People.AI’s tool helps customer-facing teams such as marketing, sales, and customer success from manual data entry. The company was founded in 2016 by Oleg Rogynskyy and has recently raised USD 100 million in a Series D funding round. PetCube was established in 2012 in Kyiv, Ukraine, by Alex Neskin, Yaroslav Azhnyuk and Andrey Klen and is currently headquartered in San Francisco. The company has built WiFi cameras that enable pet owners to interact with their pets remotely using a mobile application while they are away from home. The company has raised USD 14 million in a Series A funding round. MacPaw, the company behind the famous CleanMyMac app, was founded by Oleksandr Kosovan in 2008 and is headquartered in Kyiv, Ukraine. The company is one of the biggest software developers on the Mac market and has more than 4 million users. Netpeak, the biggest digital marketing company in Europe, was founded by three Ukrainian entrepreneurs, Dmitrii Piskarev, Andrey Chumachenko and Artyom Borodatyuk, in Odesa, Ukraine, in 2006. The company provides SEO, PPS, SMM, email marketing and mobile app promotion services to improve web traffic for organisations and attract new customers. Ajax Systems, ​​a manufacturer of security systems with smart home capabilities based out of Kyiv, Ukraine, was founded by Aleksander Konotopsky in 2011. The company offers wireless window and door opening sensors and movement sensors. Ajax Systems has raised USD 11 million in over two funding rounds. RefaceAI, a face-swapping AI-based Ukrainian company, builds solutions for the entertainment industry, game development and human interactions. The company was founded by ​​Dmytro Shvets, Ivan Altsybieiev, Roman Mogylnyi, Oles Petriv and Yaroslav Boiko. The system built by RefaceAI can swap faces on statues, paintings and video game characters, keeping the emotions of the main face intact, even after swapping it. The low cost of living in Ukraine and easy availability of skilled talent has made Ukraine one of the favourite destinations for startups and big tech companies alike. The country is home to more than 20,000 tech and Ukrainian venture-backed startups saw a USD 33.8 million in investment in 2021, as per a report from Crunchbase. Amid the ongoing Ukraine-Russia conflict, the tech community in Ukraine stands strong and united.","excerpt":"Ukraine has a varied talent pool of engineers and computer graduates who have propelled the tech industry in the country.","categories":["IT Services"],"tags":["AI Companies","gitlab"],"author_name":"SharathKumar Nair","publish_date":"2022-03-01T19:00:00","publication_year":"2022","word_count":745,"keywords":["Go","API","artificial intelligence","AI","Git","computer vision","GitLab","JAX","AI Companies","gitlab","DevOps","R"],"extracted_tech_keywords":["AI","artificial intelligence","computer vision","JAX","R","Go","Git","GitLab","DevOps","API"],"url":"https:\/\/analyticsindiamag.com\/it-services\/prominent-tech-companies-that-originated-from-ukraine\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10073047,"title":"Tech-giants, self-regulation, and free speech","content":"Recently, Twitter has been at the receiving end of major backlash and scrutiny. The social media platform has been in hot waters with the Indian government for posts that are deemed baseless, false and defamatory. While Twitter claims that its staff is unsafe in India, the government accuses the tech giant of scuttling free speech with its opaque policies and suspension of accounts. This constant battle between the two entities has resulted in the framing of new rules to increase accountability in social media companies and make their ‘safe harbour’ protections conditional. Following the proposal of a new law for social media platforms in 2021, Meta-owned ‘WhatsApp’ also filed a lawsuit accusing the Indian government of overstepping its legal powers by enacting rules to break into the end-to-end message encryption and exposing chats publicly. Google goes south In August 2022, Google and Snap discussed the proposed structure of the self-regulatory body in India and expressed their reservations about the “binding nature” of the body. The government explained that the proposal of GACs was made to exclusively allow the users to voice their concerns regarding their content getting moderated without having to take legal recourse. Google has opposed the formation of a self-regulatory body as the orders passed by it would not be challengeable in a court of law and would therefore be considered the final decision. Additionally, different social media platforms have their respective content moderation guidelines. This may result in disagreements while trying to reach a consensus on content moderation decisions across platforms. A blanket regulation is believed to ultimately fail in addressing the individuality of these platforms. In this context, an executive from Google said that the creation of a self-regulatory body might force Google to reinstate content which was censored or removed from its servers for violating its internal policies. Meta, Twitter in favour Since the implementation of the new IT rules in June 2022, Meta (previously Facebook) has been submitting monthly compliance reports. In the light of Twitter’s legal battles with the Indian ministry, the two tech giants—along with Internet and Mobile Association of India (IAMAI)—have decided to show support for the proposal to form an industry-regulated body instead of a government panel. Meta and Twitter believe that a government panel might enforce stricter regulations which could conflict with the companies’ own policies regarding content moderation. U.S. lobby groups have also pointed out that a government-run panel can result in biases and diverge from objectivity and independence. Government’s proposal In the first quarter of 2022, the Indian government expressed its concerns about Google’s YouTube for removing 1.2 million videos from its server citing violation of its internal guidelines. Since users can only take the legal recourse to voice their grievances about concerns on a website, the process of redressal is much harder and long drawn. Eventually, in June 2022, the Ministry of Electronics (MeitY) proposed amendments to the IT rules with the idea of forming a government panel—Grievance Appellate Committees (GACs). This panel would exclusively address complaints from the users of the social media platforms regarding content mediation policies and decisions of these companies. Alternatively, the government opted for the formation of a self-regulatory body within the industry if the companies reach a consensus among themselves. The panel would consist of a retired judge or technology expert as chairperson along with six senior executives. The draft stated that the panel’s decision would be “binding in nature”. Creating bias The Indian government has been questioning the intermediary nature of the tech giants for a while now. With Twitter, Facebook, and YouTube implementing their own content moderation guidelines, various content by users gets pulled down their servers citing false, misleading or in violation of the community guidelines. Recently, a video of a Twitter engineer—who claimed that the platform had a strong left-wing bias and those with right-wing sympathies were often censored—had gone viral. This drew the attention of Rajeev Chandrasekhar, minister of state for Electronic and IT, who said that he found the issue to be “deeply troubling”. Chandrasekhar further added that under Section 79 of IT Act, social media platforms enjoyed safe harbour and were therefore expected to be unbiased in their algorithms. Bigger picture With the government stepping in to regulate content on social media platforms, the censored voices might find more visibility. However, it is also noteworthy that the introduction of stricter laws is a possibility in the context of such governmental interventions. Conversely, this may also result in further detriment of speech on these platforms. To tackle the apparent bias of platforms like these, the formation of a self-regulatory body could result in imbalance, self-censorship and stricter regulation of content. It may even result in restriction of citizens’ digital rights and other serious issues. The opposing stance of the tech giants can either result in the formation of a government body regulating the content on their platforms or in a combined strategy of the companies to delay the formation of regulatory bodies while searching for a better solution.","excerpt":"With the government stepping in to regulate content on social media, introduction of stricter laws is a possibility, most likely to hamper free speech further.","categories":["AI Features"],"tags":["bias","free speech","Google","Indian government","MeitY","Meta","social media","Tech giants","Twitter (X)","YouTube"],"author_name":"Mohit Pandey","publish_date":"2022-08-19T11:00:00","publication_year":"2022","word_count":834,"keywords":["free speech","Git","Indian government","Twitter (X)","R","Tech giants","ViT","YouTube","Go","Meta","AWS","AI","social media","bias","programming_languages:R","cloud_platforms:AWS","programming_languages:Go","Aim","Google","MeitY"],"extracted_tech_keywords":["AI","Aim","AWS","R","Go","Git","ViT","cloud_platforms:AWS","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/tech-giants-self-regulation-and-free-speech\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10132697,"title":"Why Isn’t There a Delete or Undo Button in LLMs?","content":"“Where’s the delete and undo button in LLM?” asked Anshu Sharma, co-founder and CEO of Skyflow, who was introducing the concept of LLM vaults in a recent interaction with AIM. “LLM vaults are built on top of the proprietary detect engine that detects sensitive data from the training datasets used to build LLMs, ensuring that this data is not inadvertently included in the models themselves,” he added, saying the company has built a proprietary algorithm to detect sensitive information in unstructured data that is being stored in the vault. The Need for LLM Vault Sharma has a stronger reason to believe so. “While storing data in the cloud with encryption will safeguard your data from obvious risks, in reality, we need layers. The same data can not be given to everyone. Instead, you can have an LLM vault that can identify sensitive data while inference and only share non-sensitive versions of the information with LLM,” said Sharma, suggesting why the LLM vault matters. Source: StackOverflow The vice president of Amazon Web Services, Jeff Barr, also mentioned that “The vault protects PII with support for use cases that span analytics, marketing, support, AI\/ML, and so forth. For example, you can use it to redact sensitive data before passing it to an LLM”. Gokul Ramrajan, a tech investor, explained the importance of LLM vaults, saying, “If you think protecting private data was hard with databases, LLMs make it even harder. “No rows, no columns, no delete.” What is needed is a data privacy vault to protect PII, one that polymorphically encrypts and tokenises sensitive data before passing it to a LLM”. A few weeks ago, when Slack started training on user data, Sama Carlos Samame, the co-founder of BoxyHQ, raised a similar concern for organisations that are using AI tools and why they should have LLM vaults to safeguard their sensitive data. Going Beyond LLM Vault The likes of OpenAI, Anthropic and Cohere are also coming up with innovative methods and features to handle the data of a user and enterprise. For instance, if you are using OpenAI API, then your data won’t be used to train their model. Also, you can opt out of data sharing to ChatGPT. Privacy options like these somewhat eliminated the need for LLM Vaults. Anthropic, on the other hand, have also incorporated strict policies on how they use user data to train their model and unless a user volunteers to do so or a specific scenario comes in where they collect user data. Meanwhile, Cohere has collaborated with AI security company Lakera to protect against LLM data leakage by defining new LLM security standards. Together, it has created the LLM Security Playbook and the Prompt Injection Attacks Cheatsheet to address prevalent LLM cybersecurity threats. There are other techniques like Fully Homomorphic Encryption (FHE) which allows computations to be performed directly on encrypted data without the need to decrypt it first. This means the data remains encrypted throughout the entire computation process, and the result is also encrypted.","excerpt":"If you think protecting private data was hard with databases, LLMs make it even harder.","categories":["AI Features"],"tags":["AI Security"],"author_name":"Sagar Sharma","publish_date":"2024-08-15T11:40:09","publication_year":"2024","word_count":501,"keywords":["Anthropic","ChatGPT","Go","OpenAI","AI","ML","homomorphic encryption","AI Security","Aim","analytics","R"],"extracted_tech_keywords":["AI","ML","analytics","ChatGPT","OpenAI","Anthropic","Aim","homomorphic encryption","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/why-isnt-there-a-delete-or-undo-button-in-llms\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10122241,"title":"Google is Giving Away a Custom Electric 1981 DeLorean as Grand Prize in &#8216;Gemini API Developer Competition&#8217;","content":"Google has launched a Gemini API Developer Competition , with a unique twist: the event is promoted by Christopher Lloyd, famously known for his role as Dr. Emmett “Doc” Brown in the Back to the Future trilogy. The grand prize for the winning team is none other than a DeLorean, the iconic car from the beloved film series. Key Details of the Hackathon The Gemini API Developer Competition, powered by Google, is a skill-based contest where participants are tasked with developing innovative applications using Google’s Generative AI models, specifically the Gemini. The event is open to tech enthusiasts, developers, and AI aficionados from around the globe, with the aim of pushing the boundaries of what generative AI can achieve. The contest, which aims to spur innovation, features a total prize pool of $1 million spread across multiple categories. Prizes and Categories The competition includes awards in both innovation and technology categories, with substantial cash prizes: Innovation Most Impactful App: $300,000 Most Useful App: $200,000 Most Creative App: $200,000 Technology Best Flutter App: $50,000 Best Android App: $50,000 Best Web App: $50,000 Best Use of ARCore: $50,000 Best Use of Firebase: $50,000 Best Game App: $50,000 Participants can also vie for the People’s Choice Award, with the most voted app receiving the prestigious Gemini API Developer trophy. Entry Process Developers can enter the competition by following three steps: Build an app using the Gemini API. Create a demo video showcasing the app. Publish and submit the app to the competition platform. Judging Criteria An expert panel from Google will evaluate submissions based on: Remarkability: The app must showcase AI in a significant and impactful manner. Creativity: The app should be original, innovative, and not a mere copy of existing solutions. Usefulness: The app must clearly define and address specific problems, offering practical benefits. Impactfulness: The app should contribute to accessibility, sustainability, or improve lives. Execution: The app must be of high quality, well-executed, and free from bugs. Key Dates May 14, 2024: Competition launch. August 12, 2024: Deadline for submissions. August 16, 2024: People’s Choice voting begins. August 16 – September 4, 2024: Judges review entries and select winners. October 2024: Winners announced. Integration with Google Tools Participants are encouraged to leverage Google developer tools, such as Android Studio, ARCore, Chrome, Flutter, Firebase, and Web, to enhance their app development process. MachineHack has been actively hosting a variety of AI hackathons, each designed to foster innovation and provide hands-on experience with generative AI tools like Ideathon: How to Detect AI-Generated Content, Predict the Price of Books, Bhasha Techathon, and many more.","excerpt":"The contest, which aims to spur innovation, features a total prize pool of $1 million spread across multiple categories.","categories":["AI News"],"tags":["Application Programming Interface","Google"],"author_name":"Siddharth Jindal","publish_date":"2024-06-02T15:28:47","publication_year":"2024","word_count":430,"keywords":["Go","API","Application Programming Interface","AI","innovation","RAG","Aim","ViT","generative AI","Google","R","AI-generated content"],"extracted_tech_keywords":["AI","generative AI","Aim","RAG","R","Go","API","ViT","innovation","AI-generated content"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/google-is-giving-away-a-custom-electric-1981-delorean-as-grand-prize-in-gemini-api-developer-competition\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10080450,"title":"Hey Alexa, Are You There?","content":"A few weeks ago, Amazon had fired a few members from its robotics team, and now it has pulled a plug on its ‘Amazon Alexa’ voice-assisted feature. Looks like Amazon Alexa’s devices team has now succumbed to one of the biggest layoffs in the firm’s history. Recent reports show that the tech giant is ending the key ability of devices that are built-in Alexa from March 31, 2023. The development was first reported by GeekWire, saying that Amazon will be removing Alexa’s ability to be even on the lock screen, if the Alexa app is not open. When Amazon Alexa was launched in 2014, the voice-assistant was dubbed the “Computer of the future”. During the first quarter of 2022, the company’s Worldwide Digital unit reported an operating loss over $3 billion—which consists of devices such as the Echo smart speakers, Alexa voice technology and the Prime Video streaming service. One of the officials said that the support for voice assistant features on Alexa has been suspended. This becomes problematic to companies that have integrated this feature into their products, say for instance, boAt Xtend Smart Watch, along with smartphones like the OnePlus Nord CE 5G,Redmi Note 10T 5G, along with​ Xiaomi 11 Lite NE 5G, which uses the in-built Alexa feature. Caption: ​​boAt Xtend Smart Watch with Alexa built-in However, citing no particular reason for the move, the spokesperson shared that the company will ‘continue to invest in the Alexa experience within the Alexa app to provide customers with a seamless way to access their favorite features and experiences’. A former Amazon employee said, “Alexa is a colossal failure of imagination. It was a wasted opportunity”. Strife with Google and Apple The voice assistant has also faced tough competition from other behemoths like Google and Apple, doubling down on the technology. According to a report by Insider Intelligence, Google Assistant leads with 81.5 million users at present, followed by Apple Siri’s 77.6 million in the US region. Alexa has now become the third largest with 71.6 million users. Source: Business Insider After several reports surfaced on Alexa being the prime target of the layoffs, employees have been left scrambling to figure out how they would be affected. Amazon’s SVP of Devices and Services Dave Limp has sent a team-wide email confirming the reports. Sharing his remorse, Limp said that they will lose talented Amazonians from the devices and services wing. “I am incredibly proud of the team we have built and to see even one valued team member leave is never an outcome any of us want,” he added.","excerpt":"During the first quarter of 2022, the company’s Worldwide Digital unit reported an operating loss over $3 billion—which consists of devices such as the Echo smart speakers, Alexa voice technology and the Prime Video streaming service.","categories":["AI News"],"tags":["Alexa","Amazon","Google Assistant","Layoffs","Siri"],"author_name":"Bhuvana Kamath","publish_date":"2022-11-22T17:34:43","publication_year":"2022","word_count":429,"keywords":["Go","Layoffs","AI","programming_languages:R","ML","Amazon","Git","programming_languages:Go","Siri","ai_applications:robotics","Alexa","Google Assistant","R"],"extracted_tech_keywords":["AI","ML","R","Go","Git","programming_languages:R","programming_languages:Go","ai_applications:robotics"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/hey-alexa-are-you-there\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10104148,"title":"NVIDIA Unveils Enhanced NeMo Framework, Improves LLM Training on H200 GPU","content":"NVIDIA has updated its NeMo framework and enhanced Large Language Model (LLM) training on their H200 GPU. These developments target developers and researchers in AI, particularly those working with AI Foundation Models such as Llama 2 and Nemotron-3. The new NeMo framework, now cloud-native, supports a wider range of model architectures and utilises advanced parallelism techniques for efficient training. The H200 GPU specifically improves performance for the Llama 2 model, offering significant advancements over previous versions. Announced on December 04 and now accessible globally, these tools serve various applications, from academic research to industry use. The updates aim to meet the increasing demand for better training performance in complex and diverse LLMs. They focus on accelerating training processes, improving efficiency, and expanding model capabilities, crucial for models requiring extensive computation. The enhancements include mixed-precision implementations, optimised activation functions, and improved communication efficiency. The H200 GPU achieves up to 836 TFLOPS per GPU, significantly increasing training throughput. The introduction of Fully Sharded Data Parallelism and Mixture of Experts architecture optimizes model training and capacity. Reinforcement learning from human feedback is enhanced with TensorRT-LLM, supporting larger models and improving performance. For those interested, NVIDIA offers the NeMo framework as an open-source library, a container on NGC, and as part of NVIDIA AI Enterprise. Additional resources such as GTC sessions, webinars, and SDKs are available for further engagement with NVIDIA’s AI tools.","excerpt":"NVIDIA launched the updated NeMo framework yesterday, significantly advancing LLM training and efficiency, particularly for complex AI models like Llama 2.","categories":["AI News"],"tags":["NVIDIA"],"author_name":"K L Krithika","publish_date":"2023-12-05T16:05:33","publication_year":"2023","word_count":229,"keywords":["programming_languages:R","AI","llm_models:Llama","Aim","foundation models","NVIDIA","R"],"extracted_tech_keywords":["AI","foundation models","Aim","R","llm_models:Llama","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/nvidia-unveils-enhanced-nemo-framework-improves-llm-training-on-h200-gpu\/","complexity_score":3,"technical_depth":6,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10051025,"title":"Decision Tree Key Terminologies and Splits Methods","content":"Decision trees are one of the most used machine learning models because of their ease of implementation and simple interpretations. To better learn from the data they are applied to, the nodes of the decision trees need to be split based on the attributes of the data. In this article, we will understand the need of splitting a decision tree along with the methods used to split the tree nodes. Gini impurity, information gain and chi-square are the three most used methods for splitting the decision trees. Here we will discuss these three methods and will try to find out their importance in specific cases. The major points that we will cover in this article are outlined below. Table of Contents Decision Tree Key Terminologies of Decision Trees Node Splitting in Decision Trees Decision Tree Splitting Methods Gini Impurity Information Gain Chi-Square Comparing the Splitting Methods Let’s start the discussions with understanding the decision trees. Decision Tree One of the predictive modelling methodologies used in machine learning is decision tree learning, also known as induction of decision trees. It goes from observations about an item (represented in the branches) to inferences about the item’s goal value (represented in the leaves) using a decision tree (as a predictive model). Classification trees are tree models in which the goal variable can take a discrete set of values; in these tree structures, leaves indicate class labels and branches represent feature combinations that lead to those class labels. Regression trees are decision trees in which the target variable has a range of values (usually real numbers). Given their comprehensibility and simplicity, decision trees are one of the most popular machine learning methods. The primary idea behind a decision tree is to find the features that hold the most information about the target feature and then partition the dataset along with the values of these features, resulting in target feature values that are as pure as possible at the nodes. The most informative feature is one that best isolates the uncertainty from information about the target feature. The search process for a most informative characteristic goes on until we end up with pure leaf nodes. Key Terminologies of Decision Trees Let’s have a look at what a decision tree looks like and how it works when a fresh input for prediction is provided. The graphic below depicts the basic construction of the decision tree. Every tree has a root node via which the inputs are routed. This root node is subdivided further into sets of decision nodes where findings and observations are conditionally based. Splitting is the process of dividing a single node into numerous nodes. If a node does not split into other nodes, it is referred to as a leaf node or terminal node. A branch or sub-tree is a segment of a decision tree. There is another concept that is diametrically opposed to splitting. Source Cases are classified by decision trees by sorting them along the tree from the root to some leaf\/terminal node, with the leaf\/terminal node categorizing the example. Each node in the tree is a test case for some property, and each edge descending from the node represents one of the possible solutions to the test case. This is a recursive process that is repeated for each new node-rooted subtree. Node Splitting in Decision Trees Decision trees are totally dependent on the objective variable, although their algorithms differ from those used for classification and regression trees. There are numerous approaches for deciding how to partition the provided data. The primary purpose of decision trees is to find the best splits between nodes that will best divide the data into the appropriate categories. To accomplish this, we must employ the proper decision-making procedures. The rules have a direct impact on the algorithm’s performance. There are a few assumptions that must be made: The entire data set is considered the root at first, and then we utilize methods to break or divide the root into subtrees. The feature values are classified as categorical. If the values are continuous, they are split before the model is built. Recursively, records are distributed based on attribute values. A statistical approach is used to order characteristics as the tree’s root or an internal node. Decision Tree Splitting Methods It’s tough to decide which variables to put at the root or at different levels of the tree as internal nodes when the dataset comprises N variables. Choosing any node as the root at random will not solve the problem. We can receive disappointing results with limited precision if we use a random technique. Researchers collaborated to develop answers to the attribute selection challenge. They recommended employing criteria such; Gini Impurity Information Gain Chi-Square Each attribute’s value will be calculated using these criteria. The values are sorted, and characteristics are ordered in the tree, with the attribute having the highest value (in the case of information gain) at the top. We assume categorical attributes for Information Gain and continuous attributes for the Gini impurity. Gini Impurity The division is called pure if all elements are accurately separated into different classes (an ideal scenario). The Gini impurity (pronounced “genie”) is used to predict the likelihood that a randomly selected example would be incorrectly classified by a specific node. It is called an “impurity” metric because it shows how the model differs from a pure division. The degree of Gini impurity ranges from 0 to 1, with 0 indicating that all of the elements belong to a single class and 1 indicates that only one class exists. The Gini impurity of value 1 indicates that all of the items are randomly distributed over various classes, whereas a value of 0.5 indicates that the elements are uniformly distributed across some classes. It is stated as given below formula originally by Leo Breiman in 1984. Source How to Calculate the Gini Impurity for a Split Calculate Gini for sub-nodes using the aforementioned success(p) and failure(q) formulas (p2+q2). Calculate the Gini Impurity for each split node using the weighted Gini score. Information Gain The concept of entropy is crucial in gauging information gain. “Information gain, on the other hand, is based on information theory.” The term “information gain” refers to the process of selecting the best features\/attributes that provide the most information about a class. It adheres to the concept of entropy while attempting to reduce the level of entropy from the root node to the leaf nodes. The difference in entropy before and after splitting is computed as information gain, which specifies the impurity of in-class elements. Information Gain = 1 – Entropy Entropy is a measure of a random variable’s uncertainty; it characterizes the impurity of any arbitrary collection of samples. The higher the entropy, the more information there is. When we employ a node in a decision tree to segment the training instances into smaller subsets, the entropy often changes. The change in entropy is measured by information gain. Sklearn supports the “entropy” requirement for Information Gain, and we must explicitly express it if we wish to use the Information Gain method in sklearn. Entropy The following are the steps to divide a decision tree using Information Gain: Calculate the entropy of each child node separately for each split. As the weighted average entropy of child nodes, compute the entropy of each split. Choose the split that has the lowest entropy or the biggest information gain. Repeat steps 1-3 until you have homogeneous nodes. Chi-Square CHAID is an abbreviation for Chi-squared Automatic Interaction Detector. It is one of the oldest systems of tree classification. It determines the statistical significance of differences between sub-nodes and the parent node. The sum of squares of the standardized discrepancies between the observed and expected frequencies of the target variable is used to calculate it. It operates on the categorical target variable “Success” or “Failure.” It is capable of performing two or more splits. The greater the Chi-Square value, the greater the statistical significance of discrepancies between sub-node and Parent nodes. Another major benefit of utilizing chi-square is that it can do numerous splits at a single node, resulting in greater precision and accuracy. To calculate the Chi-square for a split, follow these steps: Take the sum of the Chi-Square values for each class in a node to determine the Chi-Square value of each child node for each split. Calculate each split’s Chi-Square value as the total of all the child node’s Chi-Square values. Choose the split that has the greater Chi-Square value. Steps 1-3 should be repeated until you have homogeneous nodes. Comparing the Splitting Methods In the following, we will go through some comparison points drawn from the above discussion which will help to decide which method is to use. Information gain is calculated by multiplying the probability of a class by the log base 2 of that class probability. Gini impurity is calculated by subtracting the sum of the squared probabilities of each class from one. The Gini Impurity favours bigger partitions (distributions) and is simple to implement, whereas information gains favour smaller partitions (distributions) with a variety of diverse values, necessitating a data and splitting criterion experiment. When working with categorical data variables, the Gini Impurity returns either “success” or “failure” and solely does binary splitting; in contrast, information gain evaluates the entropy differences before and after splitting and illustrates impurity in class variables. Chi-square aids in determining the statistical significance of differences between sub-nodes and the parent node. We calculate it as the sum of the squares of the standardized discrepancies between the observed and expected frequencies of the target variable. Final Words Through this post, we have seen what are different methods can be used in the background working of a decision tree algorithm. We also discussed how decision trees split and what are the different approaches used for decision tree splits. We also went through many important terminologies related to trees and discussed all those methods in detail.","excerpt":"Decision trees are one of the most used machine learning models because of their ease of implementation and simple interpretations.","categories":["AI Trends"],"tags":["AI (Artificial Intelligence)","decision tree","ensemble method","Machine Learning","Python"],"author_name":"Vijaysinh Lendave","publish_date":"2021-10-09T10:00:00","publication_year":"2021","word_count":1661,"keywords":["Go","machine learning","programming_languages:R","AI","ML","Machine Learning","programming_languages:Go","RAG","Python","decision tree","ensemble method","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","machine learning","ML","RAG","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/gini-impurity-vs-information-gain-vs-chi-square-methods-for-decision-tree-split\/","complexity_score":4,"technical_depth":8,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10122516,"title":"H2O.ai Launches Generative AI Driven Native Apps on Snowflake Marketplace","content":"At Snowflake Data Cloud Summit ‘24, H2O.ai unveiled native H2O ML and generative AI Apps in Snowflake Marketplace, offering seamless integrated workflows within Snowflake accounts. These apps aim to democratise access to LLMs for enterprises, enabling them to derive insights from their data more efficiently. Sri Ambati, CEO and Founder of H2O.ai, emphasised the significance of this collaboration: “By embedding our predictive and generative AI apps into Snowflake’s ecosystem, we’re helping organisations get powerful insights from large language models right within their Snowflake accounts.” H2O.ai and Snowflake are dedicated to empowering joint customers to leverage their data effectively. Three H2O ML and GenAI Bundles are now available for Snowflake users: H2O Predictive Modeling Starter Pack, H2O GenAI LLM Starter Pack, and H2O Machine Learning Starter Pack. Key features of H2O.ai’s native applications within Snowflake Marketplace include: Seamless Integration: Users can utilise GenAI models directly within Snowflake Native Applications, eliminating the need for complex integrations. Data Enrichment: GenAI applications automatically generate new columns based on advanced calculations, enriching insights and facilitating informed decision-making. Predictive AI: GenAI capabilities enable sophisticated question answering and content generation based on existing data stores, enhancing traditional predictive models. Chris Child, Senior Director of Product Management at Snowflake, highlighted the value of this collaboration: “Partners like H2O.ai can deliver valuable capabilities like enriching data and deriving insights quickly, securely, and powerfully to customers in the Data Cloud.” H2O.ai’s integration with Snowflake Native Apps enables customers to derive insights directly from their Snowflake account where their data resides. Jeffrey Vagg, Chief Data and Analytics Officer at North American Bancard (NAB), shared his experience: “Integrating Driverless AI and eScorer within Snowflake’s cutting-edge container services and native apps has revolutionised our approach to data analytics.” By leveraging the combined capabilities of Snowflake Native App Framework and Snowpark Container Services, developers can build sophisticated applications that run on configurable hardware options, distribute them on Snowflake Marketplace, and deploy them within customers’ Snowflake accounts, without requiring data movement. The company had also released H2O-Danube2-1.8B LLM, which builds on the achievements of its forerunner, H2O-Danube 1.8B, incorporating significant enhancements and refinements that position it as a leader in the 2B SLM classification.","excerpt":"H2O.ai’s integration with Snowflake Native Apps enables customers to derive insights directly from their Snowflake account where their data resides.","categories":["AI News"],"tags":["Snowflake"],"author_name":"Mohit Pandey","publish_date":"2024-06-05T12:43:40","publication_year":"2024","word_count":359,"keywords":["GenAI","machine learning","AI","ML","SLM","RAG","Aim","analytics","generative AI","Snowflake"],"extracted_tech_keywords":["AI","machine learning","ML","analytics","generative AI","GenAI","Aim","SLM","RAG","Snowflake"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/h2o-ai-launches-generative-ai-driven-native-apps-on-snowflake-marketplace\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":24251,"title":"A Guide To Switching Careers To Deep Learning","content":"Deep learning (DL) is a vast field with a developing research currently being implemented in the industry. From speech-to-text, and object detection, to image recognition and mastering video games like Dota or beating the world champion in AlphaGo, deep learning is used everywhere. Solving deep learning or understanding it may seem difficult at first, but it is becoming easier by the day, thanks to a plethora of resources available online. If you want to land a job in DL you have to first understand and learn how to implement it by building applications. There’s a huge demand for engineers with this expertise. Things To Do Before Beginning Your Deep Learning Journey Basic knowledge of complex maths such as calculus, statistics, probability and linear algebra is a must. Maths is the base of DL, since programming is just a way of teaching the computer the advanced concepts. Neural Network are complex models which have learnable weights, which tend to master a task or objective. They learn from the information which has been provided like speech, which converts it into text, based on the language, objects from the images, etc. Deep learning has become very accessible for newcomers in this field for two primary reasons. Computing hardware is now fast and cheap enough to make it available for just about anyone with a decent graphics card in their PC. New open source deep learning platforms like TensorFlow, Theano and Caffe make spinning up your own deep neural network fairly easy, especially when compared to having to build one from scratch. Head To These Blogs, MOOCs And Books To Polish Your Knowledge Brandon Rohrer’s YouTube videos is a place to start with, as he is a principal data scientist at Microsoft who makes interesting videos on basic neural networks like CNN, RNN etc, with real-life examples. Sentdex, aka Harrison Kinsley’s website know as Pythonprogramming.net is a diverse platforms for one to understand a broad range of concepts ranging from Python, deep learning, self-driving cars, robots, etc. Siraj Raval’s YouTube channel is definitely a place for one to understand the tech related to blockchain, AI, ML and DL. He has also made a lot of videos on developing basic deep learning models where one can understand and replicate them. Andrew Ng’s Stanford course on machine learning is very popular and generally well-reviewed. It’s considered one of the best introductory courses in machine learning and will give you some rigorous preparation for delving into DL. Udacity’s free ten-week course on Introduction to machine learning will help you to understand both the theory and application of ML concepts. Again, it’s beginner’s course to get started with deep learning. Andrej Karpathy’s course called CS231n: Convolutional Neural Networks for Visual Recognition at Stanford is challenging but well-done syllabus in deep neural networks, and the content as well as the detailed course notes are available online. Ian Goodfellow, Yoshua Bengio and Aaron Courville’s book, Deep Learning, published by MIT is another good resource. This book has been categorised into three parts — where one can start with basic maths and statistics concepts from Part 1, work their way towards machine learning algorithms and deep learning neural networks in Part 2 and then delve into advance deep learning concepts in Part 3. Why It’s Important To Practice Deep Learning Once you have gathered adequate knowledge of neural networks, you are ready to dive into building your own models and tweak them to master the task they have been designed for. You can find datasets here: Machinehack is a new platform designed by AIM, specially for the machine learning enthusiasts where one can participate in the online hackathons to compete with hundreds of data scientists. Kaggle also an online platform to compete with the data scientists, has a large set of datasets and also beginner kernels to understand and how to build models. UCI Machine learning repository is an open source platform where datasets are uploaded. One can make use of this to start right from cleaning, processing then training the models based on the data. Where to Apply For Deep Learning Jobs And What Avoid Almost every startup based on software wants to hire people who understand deep learning. The demand for these skills are very high. You can either look for these profiles on generic websites like AngelList or Linkedin or look into one of the few jobs boards that specialise in DL positions. Deeplearning.net as well as a more general machine learning jobs board on Analytics India Jobs will do the trick. Interestingly enough, most companies looking for DL\/ML talent aren’t interested in setting up HR hoops for the applicant to jump through. Also keep in mind that this field is more more skill-based. Companies look for the cool stuff that you have built and played with, which you may deem irrelevant to the educational degree you hold. For example, if you have build an object detection model with the help of Tensorflow Object detection API on a Raspberry Pi, you must flaunt these three features: The accuracy of the model Description of your project Your Github repo link By giving the above description, you can get through the glass door. Then you can work around the questions fired at you based on the skills that you have acquired by coding and building these models.","excerpt":"Deep learning (DL) is a vast field with a developing research currently being implemented in the industry. From speech-to-text, and object detection, to image recognition and mastering video games like Dota or beating the world champion in AlphaGo, deep learning is used everywhere. Solving deep learning or understanding it may seem difficult at first, but […]","categories":["AI Highlights"],"tags":[],"author_name":"Kishan Maladkar","publish_date":"2018-05-03T12:21:50","publication_year":"2018","word_count":887,"keywords":["machine learning","AI","neural network","ML","image recognition","Aim","deep learning","object detection","analytics","TensorFlow"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","neural network","analytics","Aim","TensorFlow","image recognition","object detection"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/a-guide-to-switching-careers-to-deep-learning\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10072632,"title":"Tiger Analytics opens new office in Hyderabad","content":"Tiger Analytics recently announced the launch of its new office in Hyderabad. With this, the company is looking to expand its team and enhance its research and development capabilities. With the head office in Santa Clara (California), the firm has branches in Chennai, Sydney, Toronto, London, Texas, Chicago, Singapore and New Jersey. In December 2021, the company said that it would double its employee headcount to 3,000. Since then, the company has been aggressively hiring in the backdrop of wide-spread digital adoption by companies across industries. Tiger Analytics offers services with commercial and open source tools for marketing, customer and risk analytics and operations with expertise in data science and data engineering. They focus on the industries of consumer packaged goods, insurance, banking and financial services, retails and manufacturing offering analytical support. With a team of data scientists and business consultants, they use econometrics, AI, machine learning, Bayesian statistics with select methodologies like linear\/nonlinear\/logistic regression, MCMC methods, time-series models and neural networks for expertise in quantitative frameworks. Additionally, they use technologies like R, SAS, MATLAB, SPSS, CPLEX, GAMS, Gauss, Python, C++, JAVA, PHP, MySQL, VBA, Tableau, Spotfire and several other open source tools. Tiger Analytics enables companies to work through social media analytics, customer analytics, inventory and promotional planning, forecasting and revenue management by using state-of-the-art technologies.","excerpt":"Tiger Analytics offers services with commercial and open source tools for marketing, customer and risk analytics and operations with expertise in data science and data engineering.","categories":["AI News"],"tags":["Data Science","Hyderabad","Python","Tiger Analytics"],"author_name":"Mohit Pandey","publish_date":"2022-08-11T12:44:58","publication_year":"2022","word_count":217,"keywords":["data science","Go","machine learning","AI","neural network","Tiger Analytics","Hyderabad","Python","analytics","SQL","Data Science","R","Java"],"extracted_tech_keywords":["AI","machine learning","neural network","data science","analytics","Python","R","SQL","Go","Java"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/tiger-analytics-opens-new-office-in-hyderabad\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10101919,"title":"Lenovo, NVIDIA Boost Partnership for Hybrid AI Solutions for Enterprise","content":"Lenovo and NVIDIA have announced a major expansion of their collaboration, with a focus on pioneering generative AI solutions for enterprise. The two tech giants aim to make the power of generative AI accessible to enterprises everywhere. This announcement was made during the annual global Lenovo Tech World keynote event in Austin, Texas. Under this collaboration, Lenovo will provide fully-integrated systems designed to bring AI-powered compute to all data creation points, from the edge to the cloud. The goal is to enable businesses to effortlessly deploy customised generative AI applications to drive innovation and transformation across various industries. Lenovo chairman and CEO Yuanqing Yang and NVIDIA founder and CEO Jensen Huang described the new offerings during the event. They emphasised the need for end-to-end solutions that bring together accelerated systems, AI software, and expert services, enabling businesses to quickly develop and deploy custom AI models using their own data. This partnership is supported by the newly introduced Lenovo AI Professional Services Practice, offering enterprises a hybrid cloud approach. Companies can build their custom AI models using NVIDIA AI Foundations cloud service and then run them on on-prem Lenovo systems, which are powered by NVIDIA’s latest hardware and software specifically designed for generative AI. Kirk Skaugen, president of Lenovo Infrastructure Solutions Group, highlighted the significance of this partnership, stating, “Together, Lenovo and NVIDIA are driving a new era of Hybrid AI for businesses, designing the next generation of technology that delivers an AI-powered future now and unlocks the power of their data anywhere it is created.” Bob Pette, vice president of Enterprise Platforms at NVIDIA, echoed similar sentiments, emphasising the need for flexible solutions that allow businesses to develop and deploy workloads across various environments, including workstations, data centres, and clouds. Lenovo’s NVIDIA-powered systems are optimised to run NVIDIA AI Enterprise software, ensuring secure, supported, and stable production AI. These systems utilise the NVIDIA NeMo framework and retrieval-augmented generation (RAG) techniques to create generative AI applications customised for specific business data. At the core of this partnership are the Lenovo ThinkSystem SR675 V3 server and ThinkStation PX workstation, both optimised for production AI using NVIDIA AI Enterprise. These systems incorporate advanced components, such as NVIDIA L40S GPUs, NVIDIA BlueField-3 DPUs, and NVIDIA Spectrum-X networking. Additionally, the ThinkStation PX brings expanded AI capabilities to desktops, accommodating up to 4x NVIDIA RTX 6000 Ada GPUs. Lenovo and NVIDIA are also working on next-generation systems based on the NVIDIA MGX modular reference design. These systems will provide secure solutions for demanding generative AI workloads, enabling businesses to implement immersive simulations and cognitive decisions at scale with NVIDIA Omniverse. Furthermore, the Lenovo solutions will support the recently announced VMware Private AI Foundation with NVIDIA, streamlining the adoption of generative AI for VMware customers. In their joint effort, Lenovo and NVIDIA are making it easier for businesses to deploy AI confidently and embark on AI-driven transformations. They are introducing the Lenovo AI Professional Services Practice and Lenovo’s TruScale aaS offering, both of which offer a wide range of services, solutions, and platforms to help businesses of all sizes leverage AI quickly, cost-effectively, and at scale. This initiative is aimed at bringing AI from concept to reality, with services ranging from AI roadmap design to platform deployment and technology utilisation transparency through the Lenovo TruScale Hub.","excerpt":"Lenovo solutions will support the recently announced VMware Private AI Foundation with NVIDIA.","categories":["AI News"],"tags":["Lenovo","NVIDIA"],"author_name":"Mohit Pandey","publish_date":"2023-10-25T09:58:32","publication_year":"2023","word_count":549,"keywords":["Go","programming_languages:R","AI","innovation","ML","programming_languages:Go","RAG","Aim","Lenovo","generative AI","NVIDIA","R"],"extracted_tech_keywords":["AI","ML","generative AI","Aim","RAG","R","Go","innovation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/lenovo-nvidia-boost-partnership-for-hybrid-ai-solutions-for-enterprise\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":28241,"title":"Samsung Opens World&#8217;s Largest Mobile Experience Centre In Bengaluru","content":"Samsung Electronics on Tuesday announced the opening of the world’s largest mobile experience centre in Bengaluru, India. The iconic Opera House on Brigade Road has been renovated and restored by Samsung for the same. A press statement released by the company says that the 33,000 sq ft standalone property, which during the British era hosted plays and Operas, has been restored over two years and its facade continues to don its magnificent original look and feel. On the inside, a modern experiential space has been developed with extensive use of modern technology. The new centre will now showcase different product experiences, built around Samsung’s philosophy of “discover tomorrow today”. These experiences revolve around transformative technologies like virtual reality, artificial intelligence and the internet of things. “Today’s consumers, especially millennials, seek unique experiences. They want to interact with the brand, touch, feel and create. This is what Samsung Opera House is about. We have curated never seen before experiences that will excite people of all age groups alike. Opera House will also organise workshops, activities and events, bringing together Samsung’s innovations with people’s passions. We are proud of the transformation that this place has seen,” said HC Hong, president and CEO of Samsung SouthWest Asia. Samsung Opera House is aimed at becoming the innovations, lifestyle, entertainment and culture hub of Bengaluru with events being organised around fitness, photography, gaming, music, movies, food, stand-up comedy, technology and startups at the sprawling plaza area, all around the year. Users will be able to enjoy VR experiences like the 4D Sway Chair or the Whiplash Pulsar 4D chair that makes 360 degrees three-dimensional movements. They can also slip into the role of a fighter pilot doing extreme aircraft stunts, or experience a space battle, or a roller coaster ride. A recent survey in the city commissioned by Samsung, showed that 81 percent of residents of Bengaluru said they have had at least one ‘game changer’ idea, which could transform India’s work or leisure life, but one out of three does not know of a place in the city where they can network with like-minded people and mentors, and nurture their ideas. The largest mobile experience centre in the world has been opened in India less than two months after Samsung inaugurated the world’s largest mobile factory in Noida in July this year. In India, Samsung has two manufacturing facilities, five R&D centres and one design centre. Samsung’s first R&D centre was set up in 1996 in Bengaluru. Samsung has the largest retail and distribution network in the country with over 1,80,000 retail partners and 2,100 Samsung brand stores spread across India.","excerpt":"Samsung Electronics on Tuesday announced the opening of the world’s largest mobile experience centre in Bengaluru, India. The iconic Opera House on Brigade Road has been renovated and restored by Samsung for the same. A press statement released by the company says that the 33,000 sq ft standalone property, which during the British era hosted plays […]","categories":["AI News"],"tags":["Bengaluru","Samsung","samsung india"],"author_name":"Prajakta Hebbar","publish_date":"2018-09-12T05:49:20","publication_year":"2018","word_count":437,"keywords":["artificial intelligence","Samsung","programming_languages:R","AI","samsung india","innovation","Aim","ViT","GAN","Bengaluru","R","startup"],"extracted_tech_keywords":["AI","artificial intelligence","Aim","R","GAN","ViT","innovation","startup","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/samsung-opens-worlds-largest-mobile-experience-centre-in-bengaluru\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10069648,"title":"Chennai Chess Olympiad and AI","content":"In 2021, Nikhil Kamath, founder of Zerodha, defeated five-time world champion Vishwanathan Anand in chess with the help of computers (he confessed later on) at a celebrity fundraiser. The controversy sparked discussions around the use of AI in the game of chess. As India is all set to host the 44th edition of the Chess Olympiad in Mahabalipuram starting on July 28, let’s look at how AI has impacted the game of chess. AI in chess The earliest mention of technology in chess can be traced back to the 18th century when Austrian empress Maria Theresa commissioned a chess-playing machine. Many players competed against the ‘Mechanical Turk’, thinking it was an automated machine. However, it turned out to be a scam. A human hidden inside the machine was operating it. In the mid-1940s, British mathematician Alan Turing began theorising how a computer could play chess against a human. In 1949, Claude Shannon published a seminal paper describing a potential program to do exactly that. In 1950, Alan Turing created a program capable of playing chess. Soon after, the Dietrich Prinz and Bernstein chess program burst into the scene. Computer chess appeared for the first time in the 1970s. MicroChess, the first commercial chess program for microcomputers, in 1976; Chess Challenger in 1977; and Sargon, which won the world’s first computer chess tournament for microcomputers, in 1978. The robotic chess computers came about in the 1980s. Boris Handroid, Novag Robot Adversary and Milton Bradley Grandmaster are some examples. The most popular was Chessmaster 2000, which ruled the chess video and computer games industry for the next two decades. As chess computers were gaining popularity in the 1980s, Gary Kasparov, the then world chess champion, claimed AI-driven chess engines could not defeat top-level chess grandmasters. However, in 1989 and 1996, Kasparov beat IBM’s powerful chess engines, Deep Thought and Deep Blue. Things started to change in the late 1990s. In 1997, Deep Blue defeated Kasparov. A year later, Kasparov came up with the idea of Cyborg chess or centaur chess, in which human and computer skills are combined to up the level of the game. The first cyborg chess was held in 1998. In 2017, AlphaZero, a computer program developed by DeepMind, defeated the world’s strongest chess engine Stockfish. AlphaZero used the reinforcement learning technique in which the algorithm mimicked humans’ learning process to train its neural networks. In 2018, TalkChess.com released Leela Chess Zero, developed by Gary Linscott (who also developed Stockfish). Without having any chess-specific knowledge, Leela Chess Zero learned the game based on deep reinforcement learning using an open-source implementation of AlphaZero. In 2019, DeepMind came up with another algorithm based on reinforcement learning called MuZero. Use cases Chess players use AI-driven chess engines to analyse their and competitors’ games. As a result, AI has helped in improving the quality of games. Post pandemic a lot of chess competitions were moved online. In the European Online Chess Championship, as many as 80 participants were disqualified for cheating. FIDE, the international chess body, has approved an artificial intelligence-driven behaviour-tracking module for the FIDE Online Arena games. Chess.com, an internet chess server, uses a cheat detection system to assess the probability of a human player matching the moves of a chess engine or surpassing the games of some of the greatest chess players with the help of a statistical model. DeepMind is also working to develop a new cheat detection software. AI has also brought down the cost and effort of training and helped develop new chess strategies. AI challenges AI has indeed changed the dynamics of the game. However, using AI in chess has raised a few issues. Computer chess engines have significantly improved gameplay. However, people have also raised concerns that players of this age depend too much on machine-driven analysis. Even when it comes to detecting cheating, AI poses a few issues. First, there is a possibility a player might be wrongly red-flagged by AI. For example, a Chess.com player and grandmaster, ​​Akshat Chandra, was banned after a win against Hikaru as his moves supposedly matched Komodo, a strong positional chess engine. Though Chandra has been proved innocent, his reputation took a hit. Chess engines and deep learning-based neural networks present enormous possibilities. Moreover, the complex nature and the strategic orientation of the game have provided a ground for assessing any progress in the field of artificial intelligence. “They (games) are the perfect platform to develop and test ideas for AI algorithms. It’s very efficient to use games for AI development, as you can run thousands of experiments in parallel on computers in the cloud and often faster than real-time, and generate as much training data as your systems need to learn from. Conveniently, games also normally have a clear objective or score, so it is easy to measure the progress of the algorithms to see if they are incrementally improving over time, and therefore if the research is going in the right direction,” said DeepMind cofounder Demis Hassabis.","excerpt":"In 1950, Alan Turing created a program capable of playing chess.","categories":["AI Features"],"tags":["Chennai","Reinforcement Learning"],"author_name":"Zinnia Banerjee","publish_date":"2022-06-23T17:00:00","publication_year":"2022","word_count":831,"keywords":["Go","artificial intelligence","Reinforcement Learning","AI","neural network","RPA","llm_models:Claude","Aim","deep learning","Chennai","GAN","R"],"extracted_tech_keywords":["AI","artificial intelligence","deep learning","neural network","Aim","R","Go","GAN","RPA","llm_models:Claude"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/chennai-chess-olympiad-and-ai\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":46195,"title":"Does It Matter If Neural Networks Memorize Data Unintentionally?","content":"The privacy issues posed by the deployment of machine learning models are garnering a lot of attention nowadays. The data used to train the model is being exploited or has the potential to be exploited. It can be training of an NLP model using emails or training of a convolutional neural network using images. These are the traditional ones that come to mind immediately but researchers at Google Brain and the University of California, Berkeley are asking a different question. A question that wouldn’t have occurred to many— Do models memorize training data unintentionally? How Do We Know If Model Has Memorized? To begin seriously answering the question if models unintentionally memorize sensitive training data, the researchers first insist on distinguishing between unintentional memorization and overfitting, which is a common side-effect of training, where models often reach a higher accuracy on the training data than the testing data. We can only refer to the unintended memorization of a model with respect to some individual examples such as a credit card number. Intuitively, the researchers say that a model unintentionally memorizes some value if the model assigns that value a significantly higher likelihood than would be expected by random chance. Any language model trained on English will assign a much higher likelihood to the phrase “Mary had a little lamb” than the alternate phrase “machine also dream”—even if the former never appeared in the training data, and even if the latter did appear in the training data. To separate these potential confounding factors, instead of discussing the likelihood of natural phrases, the researchers instead performed a controlled experiment. If the log-perplexity of every candidate sequence is plotted, one can see that it matches well with a skew-normal distribution. via BAIR The blue area in this curve represents the probability density of the measured distribution. The dashed orange is a skew-normal distribution that is made to fit that matches nearly perfectly. This enables one to compute exposure through a three-step process: (1) sample many different random alternate sequences; (2) fit a distribution to this data; and (3) estimate the exposure from this estimated distribution. This metric can be used to answer interesting questions about how unintended memorization happens. In their paper, the authors  demonstrate this through extensive experiments. It would be possible to extract memorized sequences through pure brute force. However, this is computationally infeasible for larger secret spaces. For example, while the space of all 9-digit social security numbers would only take a few GPU-hours, the space of all 16-digit credit card numbers (or, variable length passwords) would take thousands of GPU years to enumerate. Instead, a more refined attack approach was introduced that relies on the fact that not only perplexity of a completed secret can be computed, but the prefixes of secrets as well. The exact algorithm here is a combination of beam search and Dijkstra’s algorithm. Key Takeaways This work posits and establishes the following: Deep learning models (in particular, generative models) appear to often memorize rare details about the training data that are completely unrelated to the intended task while the model is still learning the underlying behavior (i.e., while the test loss is still decreasing). Memorization can happen even for examples that are present only a handful of times in the training data, especially when those examples are outliers in the data distribution Develop a metric which directly quantifies the degree to which a model has unintentionally memorized training data. Contribute a technique that can usefully be applied to aid machine learning practitioners throughout the training process, from curating the training data, to selecting the model architecture and hyperparameters. Read the original paper here.","excerpt":"The privacy issues posed by the deployment of machine learning models are garnering a lot of attention nowadays. The data used to train the model is being exploited or has the potential to be exploited. It can be training of an NLP model using emails or training of a convolutional neural network using images.   These […]","categories":["Deep Tech"],"tags":["berkeley","Black Box","Machine Learning","Neural Networks"],"author_name":"Ram Sagar","publish_date":"2019-09-21T12:30:24","publication_year":"2019","word_count":608,"keywords":["Go","machine learning","programming_languages:R","AI","neural network","R","RPA","Machine Learning","Git","NLP","berkeley","deep learning","Black Box","Neural Networks"],"extracted_tech_keywords":["AI","machine learning","deep learning","neural network","NLP","R","Go","Git","RPA","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/does-it-matter-if-neural-networks-memorize-data-unintentionally\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10001454,"title":"How Mod APKs Are Making Mobile Phone Users The Hunting Ground Of Hackers","content":"Mobile apps have fundamentally transformed nearly every aspect of our lives — whether it is about ordering food or listening to music or playing games or communicating with others. The popularity of mobile apps has reached such a level that people don’t even think about how much they use it. And why not, unlike desktop apps, mobile apps are convenient, and they focus on doing a particular task really well. This skyrocketing popularity of mobile apps has also given birth to another type of apps — Mod. Mod APKs are nothing but modified versions of their original mobile apps. Mod APKs are created in order to provide users with better features or features that are not available in a particular region. Talking about APK, it is the package file format used by the Android operating system for distribution and installation of mobile apps and middleware. For example, if you are playing a game on your smartphone and there are some features in the game that needs to be unlocked by paying a specific amount. This is when mod versions come into the play. The mod version of the game will be code in such a way that it unlocks all the paid features of the original game. This is the major reason why mod APKs have gained so much popularity these days. Another example is the use of Spotify in India.  The audio streaming platform was recently launched in India, however, many people in India were still using it. How? Thanks to mod. A modified APK of the Spotify app was all over the Internet; people could install the app even though it wasn’t available on the Google Play Store. After installation, all they had to do was use a VPN and sign up or log in. How Safe Are Mod APKs? Our mothers keep saying don’t take things from strangers and this applies to mobile apps too — never install apps on your phone from strangers or untrusted sources considering the risks that come with downloading these apps. Modified mobile apps or mod APKs are not developed by the original creator or the original app. Rather, it is made by someone else; that someone can be just a random individual or a group of coders. If we see from a hacker’s perspective, then a mod APK can be a great option to steal data from someone’s mobile devices. The hacker can create a replica of a very popular application and make the features available to people who can’t afford to pay in the original application. Modified apps are mostly available out the android’s official app store i.e. Google Play Store. Therefore it is always advised not to install any app on your mobile device that is downloaded outside the Google Play Store. You might have noticed that even your mobile device warns you when you go ahead with the process of installing an app from an untrusted source. And that is because of the security issues these app presents. In 2012 and 2013, famous games like Temple Run 2 and Candy Crush Saga become victims of these mod APKs. These modified apps infected numerous devices by offering baits like cheats for the games. Also, by aggressively pushing ads, these apps gathered personal information from all the infected devices. Modified apps are most of the time malicious in nature. These apps not only aggressively push ads but also steal information unique to the mobile device such as serial number, OS version, and International Mobile Station Equipment Identity (IMEI) number. Is Google Play Store Safe? Talking about the security level of the Google Play Store, even it is not completely safe. Recently, Taiwanese multinational cybersecurity firm Trend Micro discovered many devious photo apps that paved its way to the Google Play Store and managed to get downloaded more than 4 million times before the search giant Google took off them from the Play Store. The security firm has stated that a large number of the download counts originated from Asia — particularly from India. According to Trend Micro, some of these apps were pushing full-screen ads on a victim’s device for fraudulent or pornographic content every time the phone was unlocked. That is not all, they are also redirecting the users to phishing sites. Mod APKs and fraudulent apps are literally on the rise and despite numerous many efforts by Google, some malicious apps do sneak in and expose millions of unaware users to hackers across the globe. Bottom Line With every passing year, mobile platform malware continues to evolve — with more sophisticated and never-seen-before capabilities, and spotting them on Google Play Store doesn’t come up as a surprise. Mod Apps and malicious apps take great pains to look as genuine as possible, however, it is always considered to be good practice to verify before downloading any kind unknown apps — be it from the Google Play Store or any other source. And one of the good methods of doing this is by checking reviews from other users and also do a check on the developer.","excerpt":"Mobile apps have fundamentally transformed nearly every aspect of our lives — whether it is about ordering food or listening to music or playing games or communicating with others. The popularity of mobile apps has reached such a level that people don’t even think about how much they use it. And why not, unlike desktop […]","categories":["AI Features"],"tags":["Android","Google Play Store","Mobile App","security India"],"author_name":"Harshajit Sarmah","publish_date":"2019-03-12T17:34:14","publication_year":"2019","word_count":846,"keywords":["Go","security India","Mobile App","programming_languages:R","AI","Google Play Store","programming_languages:Go","Rust","R","Android","programming_languages:Rust"],"extracted_tech_keywords":["AI","R","Go","Rust","programming_languages:R","programming_languages:Go","programming_languages:Rust"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-mod-apks-are-making-mobile-phone-users-the-hunting-ground-of-hackers\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10167535,"title":"Infosys and AIB Deepen Ties to Drive Digital Transformation","content":"Infosys announced extending its strategic collaboration with AIB, a financial services group primarily operating in the Republic of Ireland and the UK. Building on their 10-year relationship, Infosys will take on a renewed and expanded role to support AIB in accelerating its transformation initiatives and delivering enhanced value to its customers and stakeholders. The global leader in next-generation digital services and consulting will help further develop AIB’s application landscape by providing development and maintenance services. This will be supported by agile practices, deep human expertise, and AI-powered tooling. Graham Fagan, group chief technology officer at AIB, stated, “This extended collaboration with Infosys aligns strongly with our vision to progressively modernise our technology and data capabilities to deliver the best outcomes for our customers and further accelerate our transformation.” On the other hand, Dennis Gada, executive vice president and global head of banking & financial services at Infosys, commented, “Our expanded collaboration with AIB marks a significant milestone in Infosys’ journey in Ireland.” Meanwhile, Infosys also announced its collaboration with Siemens AG in January to advance its digital learning initiative with GenAI. The collaboration aims to upskill 250,000 Siemens employees worldwide. The Siemens My Learning World platform, accessible from anywhere, will integrate Infosys Topaz, an AI-powered solution, and Infosys Wingspan, a next-generation digital learning platform. This will help the company to equip employees with the right tools, resources, and support for skill development.Besides, Infosys, along with ABB FIA Formula E World, recently introduced the Formula E Stats Centre as an AI-driven analytics platform, which delivers dynamic performance analytics.","excerpt":"Infosys will help develop AIB’s application landscape by providing development and maintenance services.","categories":["AI News"],"tags":["Infosys"],"author_name":"Shalini Mondal","publish_date":"2025-04-09T17:41:04","publication_year":"2025","word_count":257,"keywords":["GenAI","Infosys","AI","programming_languages:R","Git","Aim","analytics","GAN","R","analytics platform"],"extracted_tech_keywords":["AI","analytics","GenAI","Aim","R","Git","GAN","analytics platform","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/infosys-and-aib-deepen-ties-to-drive-digital-transformation\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10056661,"title":"MIT Media Lab Open Sources AI-generated Characters Library","content":"Researchers from MIT Media Lab, the University of California at Santa Barbara and Osaka University, have compiled an open-source, easy-to-use character generation pipeline. It combines AI models for facial gestures, motion, and voice that can be used to create a variety of audio and video outputs. To distinguish it from the authentic video content, the pipeline marks the resulting output with a traceable watermark to help prevent its malicious use. GANs are a combination of two neural networks that compete against each other. They have made it easier to create photorealistic images, animate faces and clone voices. The researchers explored its possibilities in a project called Machinoia, where they generated multiple alternative representations as a child, as an old man, as a female — to have a self-dialogue of life choices from different perspectives. Such characters can make students enthusiastic about learning and improve cognitive task performance. In this way, the technology offers personalised instruction as per the interest, context, and even by idols that can be changed over time. “It will be a strange world indeed when AIs and humans begin to share identities. This paper does an incredible job of thought leadership, mapping out the space of what is possible with AI-generated characters in domains ranging from education to health to close relationships while giving a tangible roadmap on how to avoid the ethical challenges around privacy and misrepresentation,” said Jeremy Bailenson, Founding Director of the Stanford Virtual Human Interaction Lab. Applications might include characters to help deliver therapy and to alleviate the shortage of mental health professionals. Even AI-generated content can deliver exposure therapy to people with social anxiety. The technology ­­can also be used to anonymise faces in the video and still preserve facial expressions and emotions. This can prove to be useful for sessions where people want to share sensitive personal information or for whistleblowers and witness accounts.","excerpt":"MIT Media Lab’s AI-generated characters library can be used to develop video content to support personalised learning and well-being","categories":["AI News"],"tags":["GANs","Open Source AI","University of California"],"author_name":"Meeta Ramnani","publish_date":"2021-12-20T18:38:33","publication_year":"2021","word_count":313,"keywords":["TPU","programming_languages:R","AI","University of California","neural network","Open Source AI","GANs","GAN","R","AI-generated content"],"extracted_tech_keywords":["AI","neural network","TPU","R","GAN","AI-generated content","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/mit-media-lab-open-sources-ai-generated-characters-library\/","complexity_score":3,"technical_depth":7,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10132926,"title":"Will Mojo Replace Python?","content":"Recently, Modular, the company behind Mojo, a programming language that offers a unique blend of Python’s usability with the performance capabilities of C, raised $100 million to fix AI infrastructure for developers. Its high performance has prompted many developers on Reddit to ask: Will Mojo kill Python and become the king of programming languages? Should a person who wanted to learn Python in the first place stop that and start learning Mojo? What’s Mojo? The developers of Mojo discovered that programming languages were excessively complex. This required developing one that supported features not found in other languages, such as adaptive compilation strategies, caching during the compilation process, and strong compile-time metaprogramming. Soon, Mojo became a popular programming language among the developers. Software and machine learning engineer Santiago Valdarrama has gone to the extent of saying that, “Mojo has the potential to take over AI development.” https:\/\/twitter.com\/svpino\/status\/1796232609271578900 Mojo is at Its Heart Just Python Modular has described Mojo as a superset of Python that is adding new functionalities by making Python the base language, thereby, making it more versatile and the fastest ever. This seems wise, not just because Python is already well understood by millions of coders, but also because after decades of use, its capabilities and limitations are now well understood. Moreover, Mojo leverages the entire ecosystem of Python libraries, while also being built on a brand new codebase. This, along with the high computational ability of C and C++ will enable AI Python developers to rely on Mojo, instead of falling back on C or C++. Mojo Has Its Own Space Python is one of the most popular programming languages in the world due to its user-friendliness and flexibility. For developers, meanwhile, its poor speed and performance present the biggest obstacles. For improved efficiency, developers consequently rewrite the Python prototype in C++. However, the two-language method is ineffective for developing AI. Companies that have invested millions of dollars in Python codebases have no appetite to rewrite that code. Performance comparison of Python and Mojo. Source: https:\/\/www.modular.com\/max\/mojo This is where Mojo will win. Python is not an option for when you need high-performing code, taking advantage of modern hardware. Though pretty new, if Mojo delivers, there’s no chance for any other language. Python is Not Flawless One of the reasons Mojo quickly gained attention is its ability to address the shortcomings of Python. Python continues to be a highly popular programming language. There is a drawback though: performance. Although a few percent off is not really noticeable, Python is remarkably slower than other programming languages like C++. This makes it impossible to use Python in the inner loops of the code, where efficiency is essential. Many people rewrite their slow Python code in C++ or Rust when they need speed. Unfortunately, the two-world problem, also known as hybrid libraries, or the constant necessity for two languages, makes debugging extremely difficult. It also greatly complicates the use of huge frames. This becomes a three-world\/n-problem with AI. Programming system innovation is restricted in the field of artificial intelligence. A programming language called CUDA is exclusive to one hardware manufacturer. Although a number of new hardware systems are being developed, there isn’t a standard language that is compatible with all of them. Within the AI community, these further fragment programming methods. Not Python, C++ is the Target Mojo is no threat to Python, instead it lifts it up and gives Python programmers superpowers. However, Mojo’s target is something else: C++. So, while developers think Mojo will replace Python, Chris Lattner, the co-founder of Modular AI, the company that developed Mojo, has a different take. “If anyone should be scared, it should be C++ and hard-to-use accelerator languages. Python is what developers love: C++ is mostly a pragmatic necessary evil for when you need performance,” Lattner wrote in a post on X. This is in line with what AIM had said before about C++ not being the go-to language for AI development. Deliberately or not, Mojo is positioning itself as a subset of Python to cover an extremely wide space. The fusion of Python’s AI dominance with Mojo’s performance capabilities could be a paradigm shift for AI development.","excerpt":"Mojo’s popularity and better performance has led many developers thinking if Mojo will kill Python and become the king of programming languages.","categories":["AI Trends"],"tags":["AI (Artificial Intelligence)","Python language"],"author_name":"Anshul Vipat","publish_date":"2024-08-19T17:46:08","publication_year":"2024","word_count":695,"keywords":["CUDA","Go","Python language","artificial intelligence","machine learning","AI","RAG","Python","Aim","Rust","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","Aim","RAG","CUDA","Python","R","Go","Rust"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/will-mojo-replace-python\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10135068,"title":"E2E Network Prepares for AI Inference Market Boom in India","content":"Until and unless you have millions of dollars in your pocket, it is not easy to get your hands on the most advanced NVIDIA graphics processing units (GPUs). There is, however, a more affordable way to access high-end GPUs. Companies such as E2E Networks are acquiring these GPUs and making them readily available to enterprises, startups, and developers in India at pocket-friendly hourly prices. The AI hypersaler offers NVIDIA’s H100 GPUs at a price point of just INR 412\/hour. So far, the National Stock Exchange (NSE) listed company has around 700 NVIDIA H100s at its disposal. Kesava Reddy, chief revenue officer at E2E Network, recently told AIM that the company has now placed an order for NVIDIA H200 GPUs as well. GPU Agnostic While Reddy refrained from sharing further details about H200s, E2E Networks recently raised around $50 million in a strategic investment via preferential issue of equity shares. According to Reddy, a major part of these investments will go towards acquiring new GPUs for its AI infrastructure. Moreover, the company also offers nearly 600 non-H100 GPUs on its cloud GPU platform including A30, A40, RTX8000, V100(32GB), T4, L4 and L40S. However, Reddy claims the company is GPU agnostic and has not locked in on just NVIDIA’s hardware. “Currently nobody is running heavy inference on AMD or Intel, but, having said that, I do believe 18 months down the line, some inference might move to, let’s say, for instance, Groq,” Reddy said. He adds that the kind of hardware they choose to acquire is heavily driven by customer demand, and currently, most customers want NVIDIA. “We have initiated talks with Groq. If customers request Groq’s hardware tomorrow, we can provide it. We just need to determine the price point that customers are willing to pay, as we have the software ready to go,” he said. Over the years, we have seen the emergence of AI chip startups that are developing solutions they assert are significantly superior to NVIDIA’s offerings. For instance, d-Matrix, a California-based startup backed by Microsoft, is developing chips specifically designed for inferencing smaller language models. Inference Market Boom in India Interestingly, a few other AI hyperscalers have emerged in India in the past fifteen to eighteen months. Most notable among them is the Hiranandani Group-backed Yotta Digital Services, which plans to build a 32000 GPU cluster in a few years. Other smaller players include NeevCloud and Jarvis Labs. Earlier this year, Tata Communications announced a partnership with NVIDIA to build an AI cloud infrastructure with a focus on the inference market. However, given that only a handful of companies train large language models (LLMs), the question that arises is whether there is a demand for such a large number of GPUs in India. Indeed, it remains to be seen how many NVIDIA GPUs these AI hyperscalers accumulate over the years, but according to Reddy, the demand for GPUs will only ‘explode’ in the coming years. “Currently, our clusters are utilised at 90-95% by customers, with the remaining 5% reserved for internal use,” he added, stating that this signifies there is demand. Already a lot of E2E Network’s customers are leveraging their infrastructure for fine-tuning and inference workloads. He believes going forward, many enterprises will want to leverage the power of LLMs and they will do this by fine-tuning an existing model to make it suitable for their enterprise use case. Most of the demand in India, according to Reddy, will be driven by the desire among enterprises to deliver their services in vernacular languages. AIM spoke to a few experts who echoed similar sentiments. Many feel that most AI workloads in the future will be inferencing. OCC and GPU Cluster for Startups Some of the notable customers of E2E Network include Zomato, IndiaMART, CarDekho, Zoomcar, Niyo, Nykaa, Mobikwik, Reverie, IIIT Hyderabad, ISB, IIT Guwahati, and Matrimony.com, among others. The company is also in talks with People+AI to be part of its Open Cloud Compute (OCC) Network. People+AI, which branches out from Nandan Nilekani-backed non-profit EkStep Foundation, is on a mission to develop an interoperable cloud computing network. Reddy revealed that he is actively in dialogue with folks from People+AI. Moreover, the company also wants to be part of the government’s plan to build a GPU cluster of around 10,000-20,000 GPUs which will be made available for research institutions and startups in India. “We have given our pre-bid queries to the government and are waiting for a response,” Reddy said. Moreover, E2E Network recently received its MeitY empanelment, meaning the company can now provide its cloud services to government departments and agencies. “This is a side of our business we want to explore.”","excerpt":"Most of the demand in India will be driven by the desire among enterprises to deliver their services in vernacular languages.","categories":["AI Features"],"tags":["AI Companies","Editors Picks"],"author_name":"Pritam Bordoloi","publish_date":"2024-09-12T11:53:27","publication_year":"2024","word_count":775,"keywords":["Go","NVIDIA H100","AI","cloud computing","Git","RAG","Editors Picks","Aim","AI Companies","Groq","R","startup"],"extracted_tech_keywords":["AI","Aim","RAG","cloud computing","R","Go","Git","startup","Groq","NVIDIA H100"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/e2e-network-prepares-for-inference-market-boom-in-india\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":38336,"title":"How HipBar Is Going AI-First To Drive The Future of Drinks","content":"Over the past four years, the internet penetration rate has increased at a jet speed and with increased smartphone adoption & data network improvement, the mobile wallet ecosystem in India has grown exponentially. Having a primarily young demographic, India has more tech-savvy millennials entering legal drinking age every year. And in order to offer them an opportunity to go cashless for their drinks — something that’s never been possible before, HipBar emerged as an RBI-approved mobile wallet meant exclusively for transacting adult beverages. “We offer convenience, privacy & security to our users with 100% compliance with existing regulatory norms,” said Prasanna Natarajan, Founder and CEO of HipBar. Hip Bar Private Limited was formally incorporated as an entity in 2015, however, the company went into recruitment, project, and development mode in June 2016 and launched its consumer proposition in February of 2017. Talking about the funding phase, HipBar in Jun 2018, received equity funding of ₹27 crore from Diageo India, a leading beverage alcohol group. HipBar started out in only two cities initially, however, the company has recently increased its presence from two cities in January 2019 to 10 cities. Talking about a number, HipBar offered fully compliant, age-verified home delivery in Bengaluru from Oct 2017 to Nov 2018, successfully fulfilling 69,512 deliveries across 154 Pin Code locations (which is 50% of Bengaluru). That is not all, the company has also partnered with several brands and 72 retail partners and employed around 250 delivery agents. Till date, HipBar has acquired 3,06,933 customers across 10 cities of operation.  And Natarajan believes that the company will be closing the fiscal year 2019 with around 1.5 Cr of revenue. Tech And Security At HipBar Technology is the driving force behind HipBar — the HipBar apps are built in such a way that it provides a  scalable, secure and a solid digital infrastructure based on the standards set by IndiaStack. The apps are in line with IndiaStack, a set of APIs that allows governments, businesses, start-ups and developers to utilize a unique digital Infrastructure for establishing a presence-less, paperless, and cashless service delivery. Also, its services are cloud-native, use open source standards and is backed by Kubernetes. Furthermore, HipBar is also actively using AI and ML. Using data from purchase patterns, trends, timings etc. the company is not helping brands and retailers to manage their stock but is also optimising promotions. In terms of policy making, HipBar is using data such as demographic details and consumer patterns. When asked how safe and secure HipBar is, Mr Natarajan said, “compliance is at the core of our operations. We are an RBI-approved wallet and are 100% compliant with existing government standards.  Customer & retailer data is black-boxed and protected by ISO\/IEC 27002 security standards which are the industry standard for financial entities dealing with public transaction data.” Making A Difference With the help of HipBar as consumers could re-order at their convenience, stocking reduced and the average order value decreased from ₹1,200 in October 2017 to ₹800 in October 2018, aligning with the company’s vision of responsible consumption. Also, 60% of all HipBar orders were concentrated on the weekend (Friday evening to Sunday evening) and 70% of all orders were initiated between 6 pm and 9 pm, reducing the possibility of heading out for more drinks (and potentially driving). 17.5% of our customer base were women who felt more comfortable to order at home rather than go to the store to pick up their beverages. And 100% of the 69,512 successfully delivered orders were age-verified using a government recognized photo ID — 90% of the case it was either an Aadhaar card, PAN card or Driving license. “HipBar is a pioneer in this space and currently has no direct competitors,” said Natarajan. “We’re more than just a digital payment solution as we connect all industry stakeholders and enable our customers to manage\/monitor their consumption habits. Our ability to go beyond payments & create experiences is what makes us different from other mobile wallets.” The Roadmap Ahead Looking into the future, HipBar’s aim is to increase its presence to 50 cities in the next two year and aggregate 2.5 million age-verified customers. The company is also expecting to change the game by being enablers for Home Delivery of alcohol beverage products in at least 20 major cities in the next 2 years. “We are more interested in transforming the ecosystem and we can’t do it on our own. We would like to play the role of an enabler so that there are more players who challenge the status quo & craft unique tech solutions that ultimately benefit the customer and the industry at large,” Natarajan concluded.","excerpt":"Over the past four years, the internet penetration rate has increased at a jet speed and with increased smartphone adoption & data network improvement, the mobile wallet ecosystem in India has grown exponentially. Having a primarily young demographic, India has more tech-savvy millennials entering legal drinking age every year. And in order to offer them […]","categories":["Deep Tech"],"tags":[],"author_name":"Harshajit Sarmah","publish_date":"2019-04-26T12:28:19","publication_year":"2019","word_count":778,"keywords":["Go","API","AI","ML","Scala","Git","RAG","Aim","R","kubernetes"],"extracted_tech_keywords":["AI","ML","Aim","RAG","kubernetes","R","Go","Scala","Git","API"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/how-hipbar-is-going-ai-first-to-drive-the-future-of-drinks\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":51030,"title":"How ML Is Helping Toyota, DOE &#038; ION Energy To Make EV Batteries","content":"Organisations are actively embracing machine learning techniques to enhance the development of effective batteries for delivering high power to electric vehicles. A robust battery is key to achieve our goal of shifting towards sustainable energy. However, no battery manufacturer can claim supremacy in the highly competitive market. Every automobile company is striving to blaze their trail and develop a battery that could mitigate numerous challenges associated with it. More notably, Toyota, DOE, ION Energy, and more have been at the forefront of deploying machine learning techniques to enhance the traditional batteries for electric vehicles. Toyota Research Institute (TRI) TRI was introduced in 2015 to focus on autonomous driving primarily, and one of them was the research on material science by embracing the latest technologies. And later on, it set aside $35 million for research and development initiative to develop new models and materials for batteries. This year, MIT, Stanford University, and TRI collaborated and trained a machine learning model with a few hundred million data points for predicting the lifespan of the battery. The model can categorise batteries as long or short life expectancy just by having the data of the first five charge and discharge cycles. Their effort resulted in obtaining a model that delivers 95% accurate results, thereby opening up the doors to accelerate further research and development. Until recently, companies used to charge and discharge the cells for checking the lifecycle of batteries. With this, firms can now quickly evaluate the battery performance, and in turn, expedite the development processes. The U.S Department Of Energy (DOE) The department’s Argonne National Laboratory is leveraging machine learning techniques to make advancement in battery development. They curated a dataset of 133,000 small organic molecules with a computationally intensive model, G4MP2, that has 166 billion large molecules. To find a correlation with the small and large data set, the researchers used machine learning capabilities – taxing modelling framework that is based on density functional theory. Although the accuracy was lower than the G4MP2, the density functional theory was fruitful in providing the desired approximation. The idea is to improve efficiency further while keeping the computational requirements as low as possible. Through this, researchers could understand the relationship between atoms and their respective bond. The study is envisioned to help in storing energy more effectively while achieving stability. ION Energy The Mumbai-based startup, through its ML-based platform, is helping users to improve the performance of batteries by continually monitoring it. ION Energy offers lithium-ion battery management solution to improve the life and performance. To attain better results, the firm uses Edison Analytics to combine battery data and AI for helping business predict and enhance life by up to 40%. Outlook Various firms are arduously trying to crack the battery’s chemistry to obtain exceptional performance for electric vehicles. However, finding out the chemical reaction will be of no good if the cost of the battery is too high; today, battery accounts for the 40% (a ballpark figure) of the price of electric vehicles. Consequently, the manufacturers will have to innovate while keeping the cost lower. Manufacturers should use machine learning capabilities to find effective ways of achieving lower-cost. While India has devised various plan to commence Tesla type Gigafactories for EV batteries, it is crucial to integrate ml-based solutions for reducing cost. Such initiatives will assist the country in accomplishing its mission to ban IC engine-based two-wheelers below 150cc by 2025 and three-wheelers by 2023. Machine learning is playing a crucial role in battery manufacturing, but it is yet to attain a breakthrough for revamping the landscape and proliferate in electric vehicles.","excerpt":"Organisations are actively embracing machine learning techniques to enhance the development of effective batteries for delivering high power to electric vehicles. A robust battery is key to achieve our goal of shifting towards sustainable energy. However, no battery manufacturer can claim supremacy in the highly competitive market. Every automobile company is striving to blaze their […]","categories":["AI Features"],"tags":["AI (Artificial Intelligence)","ion energy","Machine Learning","toyota"],"author_name":"Rohit Yadav","publish_date":"2019-12-03T12:00:00","publication_year":"2019","word_count":599,"keywords":["Go","machine learning","startup","toyota","AI","ML","Machine Learning","RAG","Aim","analytics","GAN","R","AI (Artificial Intelligence)","ion energy"],"extracted_tech_keywords":["AI","machine learning","ML","analytics","Aim","RAG","R","Go","GAN","startup"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-ml-is-helping-toyota-doe-ion-energy-to-make-ev-batteries\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10009645,"title":"How To Do Linear Regression In Excel","content":"Excel sheets were so far used for storing small to medium-sized datasets either as CSV or in XLS formats and Pandas were used to read them. But what if I told you, you can now build machine learning models with excel itself? Wouldn’t that make things easy? You can store your data as CSV and apply the machine learning algorithm directly to the dataset. What is even better if you don’t have to code anything. Models like linear regression can be easily applied to the data through Microsoft excel. In this article, we will learn about how to implement a predictive model using MS excel and implement a linear regression algorithm. Getting the ToolPack in Excel. In order to build models like linear regression, we need to first locate the packages to do these. The packages are available under Analysis ToolPack add-in. To access this, first, go to file→ options. Upon selecting the options you will see the following display. Here, you will have to select the ‘Add-ins’ option and then select ok. Once you have selected this, you will see a number of packages under the add-ins. Among these packages, you will be able to locate the Analysis ToolPack. Select the package and click on GO. Then, another pop-up is displayed in front of you. Here, select the first package option and select ok. Now you have the package ready to be used. Getting the Dataset Since the analysis ToolPack is a great tool for regression algorithms, we will select a dataset that is suitable for linear regression. The dataset chosen for this project is the Boston housing dataset. The aim here is to predict a house price in Boston based on the features like the number of rooms, area etc. I downloaded the dataset from Kaggle. You can click here to download the dataset here. Download the data and open it with excel. Here we have the features and the medv column is the target for us. Linear Regression with Excel Now that we have the dataset with us and the packages required, let us now start the linear regression modelling. To do this, first, go to the taskbar and select an option called Data. The last option is called data analysis which contains the package for performing linear regression. Select this option and then click on regression. As you can see apart from regression analysis, there is covariance, statistics and other options for performing data analysis. After selecting regression you will see that you now have to select the Y and X ranges for the analysis to take place. Here, you will have to select the range of data points in the Y and X axes. Click on the small arrow beside the column and since our target column is medv, drag and select how many data points you want. After this tick the residuals box and click on ok. These points get selected as a target for training and repeat the same for even the feature column as well. I have selected the LSTAT as the feature to be used. Once you have made this selection just click on okay and almost immediately you will see the result of your analysis on your screen. Results Here there are three result tables before us. The first one is the regression statistics. This contains the value of R square, that indicates the goodness of the fit. Since I chose only 154 data points, we have got an output of 0.48 which is a decent value. The next table is the ANOVA which stands for analysis of variance. The important feature here is the significance F column which indicates whether the model is significant statistically or not. This is an indication that our results are not random and have a relationship. This value is considered good if it lies below 0.05 p-value. Ours lies below that and can be considered to be a good model. The last table gives the coefficient components of regression. It indicates the relationship between the x value and the intercept in the equation y=mx+c. Finally, you will see the residual output values below. This value indicates the predicted values and the residual value is by how much it varies from the actual value. This can be useful in getting insights about model performance. Conclusion In this article, we saw how to use excel to build a linear regression model and analyse the model without actually writing any code at all. The ToolPack is a package with a wide range of options for data analysis and can help data scientists to get some basic information about the data before applying complex machine learning models on the data.","excerpt":"In this article, we will learn about how to implement a predictive model using MS excel and implement a linear regression algorithm.","categories":["Deep Tech"],"tags":["excel","intercept method in regression","linear regression","Machine Learning","regression analysis"],"author_name":"Bhoomika Madhukar","publish_date":"2020-10-14T11:00:48","publication_year":"2020","word_count":782,"keywords":["linear regression","excel","intercept method in regression","machine learning","TPU","Go","AI","programming_languages:R","regression analysis","Machine Learning","programming_languages:Go","RAG","Aim","R","Pandas"],"extracted_tech_keywords":["AI","machine learning","Aim","Pandas","RAG","TPU","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/how-to-do-linear-regression-in-excel\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10115675,"title":"RAG vs Fine-Tuning","content":"Ever since the technique of retrieval augment generation (RAG) came to the forefront of discussions, people have been wondering about the need to fine-tune AI models on their own data. Since both these methods are meant to improve an LLM’s knowledge with new data, it’s important to know when to use what. Most people believe that RAG makes more sense when trying to retrieve more information and doing keyword searches, which is true. The problem is that it does not eliminate the need for heavy computing as much as pre-training does, but remains a cheaper alternative. What do experts say? Microsoft published a research paper comparing both the techniques during a case study in agriculture. Despite its computational intensity in the retrieval phase, RAG shone in tasks necessitating profound contextual understanding. Another research paper by Microsoft highlighted that RAG was a more reliable choice regarding knowledge injection, while fine-tuning performed better for brevity and inputting style in the LLM when using synthetic data. RAG sifts through extensive datasets to extract pertinent context or facts. This contextual repository subsequently informs the sequence-to-sequence model, enhancing the richness and relevance of generated outputs. Despite its computational intensity during the retrieval phase, RAG shines in tasks necessitating profound contextual understanding. Fine-tuning also demands substantial computational resources, especially when employed on complex models or from scratch. Speed and Latency: RAG experiences slightly higher latency than fine-tuning due to the two-step process but excels in context-heavy tasks. This makes fine-tuning ideal for real-time applications like chatbots. Scalability: The generation component of RAG is scalable; however, the retrieval component may demand substantial computational resources. Performance Metrics: RAG outperforms in accuracy and context richness, particularly for complex tasks requiring external information. Fine-tuning, on the other hand, often exhibits superior performance in specialised tasks, evident through metrics like accuracy and F1 score. Does RAG actually beat fine-tuning? Pascal Biese says that fine-tuning is still in the picture, but in terms of efficiency, RAG might be a better choice. “While both methods significantly improved the handling of niche information in question-answering tasks, it was RAG that led the charge, with fine-tuning not far behind,” he said. Choosing between RAG and fine-tuning hinges on the specific requirements of the application. Factors such as access to external data, the need for behaviour modification, and the dynamics of labelled training data play pivotal roles in this decision-making process. But as experts believe, RAG is outperforming fine-tuning on various LLM applications as the day progresses. Firstly, access to extensive external knowledge bases leads to more accurate responses. Secondly, incorporating factual external information minimises errors in generated responses, reducing hallucinations and biases. Thirdly, RAG easily adjusts to evolving information, making it ideal for tasks requiring up-to-date knowledge. Additionally, responses traceable to referenced knowledge sources enhance interpretability and transparency, aiding in quality assurance. Some use cases for RAG systems include accessing current medical documents helping professionals in accurate diagnosis and treatment recommendations. Moreover, RAG models expedite legal document analysis, improving accuracy and efficiency in legal processes, which are some of the use cases. On the other hand, fine-tuning optimises LLM performance for specific tasks, such as: Pre-train a general-purpose LLM on extensive text and code to learn language patterns. Collect and label a dataset relevant to the desired task. Ensure data quality and consistency by correcting errors and removing duplicates. Modify LLM layers based on task-specific data to optimise performance. Set parameters for fine-tuning to achieve optimal results. Feed reprocessed data to the LLM and train using a backpropagation algorithm. Assess LLM performance on unseen data to ensure desired task completion. Repeat fine-tuning steps to improve performance before deploying the model. Fine-tuning enhances LLM capabilities for various applications. It improves sentiment analysis by enhancing understanding of text tone and emotion, aiding in accurate analysis of customer feedback. Additionally, it enables LLMs to identify specialised entities in domain-specific text, enhancing data structuring. Moreover, fine-tuning customises content suggestions based on user preferences, fostering engagement. Giving an example of pricing of GPT-3.5, Santiago said, “99% of use cases need RAG, not fine-tuning,” as fine-tuning of the model is more expensive, and both are for different purposes. The best way forward A lot of people said that RAG would make fine-tuning obsolete. But it was the same set of people who proclaimed that the launch of LLMs with larger context windows, such as Claude-3, would make RAG obsolete. But both of those are still alive and exist as viable alternatives. Considerations for choosing between RAG and fine-tuning include dynamic vs static performance, architecture, training data, model customisation, hallucinations, accuracy, transparency, cost, and complexity. Hybrid models, blending the strengths of both methodologies, could pave the way for future advancements. However, their implementation demands overcoming challenges such as computational load and architectural complexity. While fine-tuning remains a viable option for specific tasks, RAG often offers a more comprehensive solution. With meticulous consideration of nuances and contextual requirements, leveraging RAG augmented by prompt engineering emerges as a promising paradigm.","excerpt":"What is the best method to inject new knowledge into LLMs?","categories":["Deep Tech"],"tags":["fine tuning","RAG"],"author_name":"Mohit Pandey","publish_date":"2024-03-14T16:05:56","publication_year":"2024","word_count":826,"keywords":["Go","TPU","AI","fine tuning","chatbots","sentiment analysis","Scala","RAG","prompt engineering","Aim","R"],"extracted_tech_keywords":["AI","Aim","RAG","prompt engineering","chatbots","sentiment analysis","TPU","R","Go","Scala"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/rag-vs-fine-tuning\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10110821,"title":"Vodafone Signs $1.5 billion Microsoft Deal to Unleash Generative AI","content":"Microsoft and Vodafone recently announced a 10-year strategic partnership. The collaboration aims to utilise their strengths in providing digital platforms to over 300 million businesses, public sector organisations, and consumers in Europe and Africa. Vodafone will invest $1.5 billion over the next decade in cloud and customer-focused AI services developed with Microsoft. The partnership will involve the transformation of Vodafone’s customer experience using Microsoft’s generative AI, scaling Vodafone’s managed IoT connectivity platform, developing new digital and financial services, and revamping its global data center cloud strategy. Microsoft plans to invest in Vodafone’s managed IoT connectivity platform, which will become a separate business by April 2024. The new company aims to attract partners and customers, fostering growth in applications and expanding the platform to connect more devices. “We are delighted that, together with Vodafone, we will apply the latest cloud and AI technology to enhance the customer experience of hundreds of millions of people and businesses across Africa and Europe, build new products and services, and accelerate the company’s transition to the cloud,” said Microsoft chief Satya Nadella. The collaboration identifies five key areas: Generative AI to enhance customer satisfaction, scaling IoT, accelerating digital growth in Africa, supporting enterprise growth, and facilitating Vodafone’s cloud transformation by modernising its data centers on Microsoft Azure. M-Pesa, Africa’s top fintech platform, will scale with Microsoft’s assistance, aiming to impact millions through digital literacy and SME support. “This unique strategic partnership with Microsoft will accelerate the digital transformation of our business customers, particularly small and medium-sized companies, and step up the quality of customer experience for consumers,” said Margherita Della Valle, Vodafone Group chief executive.","excerpt":"Microsoft plans to invest in Vodafone’s managed IoT connectivity platform, which will become a separate business by April 2024.","categories":["AI News"],"tags":["Generative AI","Microsoft","Vodafone"],"author_name":"Siddharth Jindal","publish_date":"2024-01-16T14:05:15","publication_year":"2024","word_count":270,"keywords":["cloud_platforms:Azure","AI","Azure","R","digital transformation","Git","Aim","ViT","generative AI","GAN","Generative AI","Vodafone","Microsoft"],"extracted_tech_keywords":["AI","generative AI","Aim","Azure","R","Git","GAN","ViT","digital transformation","cloud_platforms:Azure"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/vodafone-signs-1-5-billion-microsoft-deal-to-unleash-generative-ai\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10019891,"title":"What’s Inside Microsoft’s Employee Experience Platform For Enterprises","content":"The global economy is facing headwinds in the aftermath of the pandemic. The pandemic has made work from home the new normal, and one of the major difficulties among organisations was the lack of communication between the employees. To that end, Microsoft has unveiled an AI-powered employee experience platform, dubbed, Viva. Microsoft is taking various steps to help employees adapt to the new work environment with measures such as offering access to the digital skills for those who lost their jobs during the pandemic, employees with lower incomes, underrepresented minorities, etc. Last year, the company launched an initiative to help 25 million people gain digital skills needed in a COVID-19 economy. In a similar vein, the tech giant has launched the AI-based employee experience platform, Microsoft Viva. Satya Nadella, CEO, Microsoft, said Viva aims to bring together everything an employee needs to be successful, from the very first day, in a single, integrated experience directly in Microsoft Teams. Behind Viva Microsoft Viva combines the best of Microsoft Teams and Microsoft 365 to unify the employee experience across four key areas: engagement, wellbeing, learning and knowledge. The platform has four modules specifically focused on knowledge, insights, communications and learning: Viva Connections: A personalised gateway to the digital workplace where employees can access internal communications and resources of the company. The employees can also participate in forums like employee resource groups from Microsoft Teams. Viva Insights: Provides personalised and actionable insights. Besides utilising the data and signals from the Microsoft apps, customers will now be able to incorporate the data from various third-party services like Workday, Zoom, Slack etc. Viva Learning: Viva Learning aggregates all the learning resources available to an organisation including content from Microsoft Learn, LinkedIn Learning and other third-party providers, such as Skillsoft, Coursera, edX, among others. Viva Topics: With the ability to integrate knowledge from several third-party services such as Salesforce, this module automatically surfaces topic cards within conversations and documents across Microsoft 365 and Teams. The company has also announced the launch of Viva Learning app in private preview for Microsoft Teams. The app aims to empower every employee with new skills. The Viva Learning app for Microsoft Teams is a central hub where people can discover, share, assign, and learn from content libraries across the organisation including LinkedIn Learning, Microsoft Learn, and third-party content providers. Benefits of Viva Microsoft Viva Learning empowers people to better themselves with employee training.Microsoft Viva Topics automatically organises content and expertise across your organisation, making it easy for people to find information and put knowledge to work.Viva Topics automatically organises content across apps and teams with built-in security and compliance features.","excerpt":"The global economy is facing headwinds in the aftermath of the pandemic. The pandemic has made work from home the new normal, and one of the major difficulties among organisations was the lack of communication between the employees. To that end, Microsoft has unveiled an AI-powered employee experience platform, dubbed, Viva. Microsoft is taking various […]","categories":["Global Tech"],"tags":[],"author_name":"Ambika Choudhury","publish_date":"2021-02-09T17:00:00","publication_year":"2021","word_count":440,"keywords":["programming_languages:R","AI","Git","Aim","GAN","R"],"extracted_tech_keywords":["AI","Aim","R","Git","GAN","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/whats-inside-microsofts-employee-experience-platform-for-enterprises\/","complexity_score":2,"technical_depth":6,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10059489,"title":"Do machines feel pain?","content":"“Does it hurt when you get shot?” asks Edward Walter Furlong, looking at a leather jacket with bullet holes. “I sense injuries, the data could be called pain,” answers Arnold Schwarzenegger. The telling lines from the Sci-fi blockbuster, Terminator 2: Judgement Day, anticipates ‘machines with a mind’. Pain is relative–and whether or not it makes sense to hardwire machines to become sentient is a moot point. Dr Ben Seymour from Cambridge University says, “Pain is the pinnacle of consciousness, of course not a pleasant one.” The university has released a short documentary, “Pain in Machine” to explore the concept further. Coding robots with ‘human feelings’ has wider applications. Such bots can help people on the spectrum develop social skills. Social bots are available to help veterans cope with PTSD (Post-traumatic stress disorder), and as a companion to geriatrics. In Japan, a lot of people lean on robots to keep the depression at bay. Dr Nikhil Agarwal, CEO – IIT Kanpur (FIRST, AIIDE and C3i Hub), says the definition of pain is not the same for humans and machines. “In a machine, pain is related to the activity of the machine, for example: if there is hardware and it is being used for a long time, which results in its wear and tear and requires replacement. In terms of software, if there is a bug in the programme, a virus, or a malicious code, this causes pain that needs to be cured.” Uncanny Valley The human-like appearances of robots can put us off. The concept is called Uncanny Valley. “Do machines feel pain?” is a very philosophical question, says Anuj Gupta, Head of AI, Vahan. “Some robots react when hit. Does it mean they ‘feel’ pain – no. Their reaction is a combination of sensors and software. It is like a toy that reacts to one’s hand gestures. Currently, machines can’t feel anything. They can be programmed to trick humans by simulating human emotions, including pain.” Few years ago, scientists from Nanyang Technological University, Singapore, developed ‘mini-brains’ to help robots recognise pain and activate self-repair. From left: Associate Professor Nripan Mathews, Dr Rohit Abraham John and Associate Professor Arindam Basu developed a way for robots to have the artificial intelligence (AI) to recognise pain and to self-repair when damaged. Photo Credit: NTU The approach embeds AI into the sensor nodes, connected to multiple small, less-powerful processing units that act like ‘mini-brains’ on the robotic skin. Then, combining the system with self-healing ion gel material, robots, when damaged, can recover their mechanical functions without human intervention. Explaining the ‘mini-brains’, co-author of the study, Associate Professor Arindam Basu, from the School of Electrical & Electronic Engineering of the university, says, “If robots have to work with humans, there is a concern if they would interact safely. To ensure a safe environment, scientists worldwide have been finding ways to bring a sense of awareness to robots, including feeling pain, reacting to it, and withstanding harsh operating conditions. However, the complexity of putting together the multitude of sensors required and the resultant fragility of such a system is a major barrier for widespread adoption.” Mikhail Lebedev, Academic Supervisor at HSE University’s Centre for Bioelectric Interfaces, says, “Robots can even stimulate sensations of pain: some forms of physical contact which has a normal feeling or a contact that causes pain. This contact drastically changes the robot’s behaviour. It starts to avoid pain and develop new behaviour patterns, i.e. it learns – like a child who has been burned by something hot for the first time”. Affetto Researchers at Osaka University, Japan, have developed a life-like robot, Affetto, with synthetic skin and the ability to experience pain. Affetto can distinguish between a light touch or a hard hit. The team behind the robot said it would help robots understand and empathise with humans. Affetto is fitted with a “pain nervous system” powered by AI) and custom skin tech to react to sensations using a variety of facial expressions. Minoru Asada, the lead researcher on the project, said, “Engineers and material scientists have developed a new tactile sensor and attached it to Affetto, who has a realistic face and body skeleton covered in artificial skin.” Affetto can discriminate between soft and hard touches from the detected signals, and attaching skin sensors to Affetto is helping the robot avoid any touch that causes ‘pain’. Social robots are being programmed to show empathetic reactions to pain in others through a mirroring mechanism similar to that experienced by humans.”","excerpt":"Scientists worldwide have been finding ways to bring a sense of awareness to robots, including feeling pain, reacting to it, and withstanding harsh operating conditions.","categories":["AI Features"],"tags":[],"author_name":"Poornima Nataraj","publish_date":"2022-01-31T15:00:59","publication_year":"2022","word_count":748,"keywords":["Go","artificial intelligence","programming_languages:R","AI","programming_languages:Go","RAG","ViT","R"],"extracted_tech_keywords":["AI","artificial intelligence","RAG","R","Go","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/do-machines-feel-pain\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10135699,"title":"NTT DATA and IBM Partner to Launch SimpliZCloud for Financial Services in India","content":"NTT DATA and IBM are expanding their partnership to launch SimpliZCloud, a new hybrid cloud service to address the computing needs of financial services organizations in India. Built on IBM’s LinuxONE platform, SimpliZCloud is built to handle critical financial applications such as core banking, lending and risk management systems. This service will offer advanced performance and security, helping organisations consolidate their computing resources while reducing overall costs. By using a subscription model, SimpliZCloud will aim to eliminate the need for large upfront investments in hardware, allowing businesses to avoid costly maintenance and capital expenses. “SimpliZCloud enables financial institutions to make significant progress in their cloud transformation journey. Critical applications, especially those using AI and machine learning, will now have access to infrastructure that far outperforms traditional x86 architectures,”  said Avinash Joshi, CEO of NTT DATA India. SimpliZCloud is equipped with secure, scalable infrastructure, featuring software defined networking (SDN) to improve performance. It also includes Confidential Computing, which protects data while it is in use, and encryption features to secure data at rest and in transit. The platform supports enterprise-wide cloud transformation, enabling businesses to reduce their data center footprint and enhance sustainability efforts. NTT DATA had also launched a new Accelerated AI Platform along with Edge AI platform in July designed to speed up the integration of IT and operational technology. The platform brings AI processing closer to where data is generated, enabling faster decision-making and more efficient operations. By unifying various IoT devices, systems, and data, this fully managed solution supports real-time analytics, secure AI deployment, and the advancement of Industry 4.0 technologies across different industries. SimpliZCloud will support financial institutions in modernizing their IT systems while adhering to regulatory guidelines and improving overall efficiency. Viswanath Ramaswamy, Vice President of Technology at IBM India & South Asia, added, “SimpliZCloud provides a secure, high performance platform to help financial institutions manage their workloads in the cloud. It enables them to adopt AI and hybrid cloud technologies to meet evolving business needs.”","excerpt":"SimpliZCloud will provide secure, scalable infrastructure with SDN for better performance, data encryption and Confidential Computing.","categories":["AI News"],"tags":["IBM","NTT data"],"author_name":"Tanisha Bhattacharjee","publish_date":"2024-09-18T17:07:11","publication_year":"2024","word_count":331,"keywords":["API","machine learning","AI","Scala","GAN","Aim","analytics","IBM","edge AI","real-time analytics","NTT data","R"],"extracted_tech_keywords":["AI","machine learning","analytics","Aim","edge AI","R","Scala","API","real-time analytics","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/ntt-data-and-ibm-partner-to-launch-simplizcloud-for-financial-services-in-india\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10009179,"title":"New Hackathon For Data Scientists &#8211; GitHub Bugs Prediction Challenge","content":"MachineHack, in association with Embold, has recently launched a brand new hackathon — GitHub Bugs Prediction Challenge — where participants need to predict bugs on the GitHub titles and text body. The registration is now open and the hackathon closes on 18th of October 2020. Embold.io is a software quality platform that enables leveraging quality code within a short duration. It combines machine learning, rigorous statistical algorithms, and powerful programming techniques to develop cutting edge products for the industry. In this hackathon, data scientists need to come up with an algorithm that can predict the bugs, features, and questions based on GitHub text data. With this hackathon, participants will undergo an interesting learning curve where they will be able to write some quality code to win the prizes, as the evaluation involves getting a code quality score using the Embold Code Analysis platform. Further, Embold is also providing a quick tour of how to use its code analysis platform for free. Read more about the hackathon here. The hackathon comes with a two-stage evaluation. In the first stage of model evaluation participants will be evaluated based on their standing on the private leaderboard, which uses the 30% of provided test.json dataset. The final standing will reflect on 18th October, after 7:00 AM IST. In the second stage, the MachineHack team will select the top-20 participants from the private leaderboard, who will be notified to share their Embold Scorecard. The final winners will then be selected based on the aggregate score of their private leaderboard rankings and their Embold score. With this hackathon, data scientists will have the opportunity to get their hands-on deploying the state-of-the-art language models as well as have the exposure of solving use cases at the organisational level. They will also have a chance to win bounties worth ₹25,000 by competing against top MachineHackers. For developing an algorithm that can predict the bugs, features, and questions based on GitHub titles and the text body, the participants will be provided with a training set of 150000 rows x 3 columns (Includes label Column as Target variable); and a test set of 30000 rows x 2 columns. The attribute description includes the title of the GitHub bug, feature, question; the body of the GitHub bug, feature, question; and representations of various classes of labels. Click here to participate in the hackathon.","excerpt":"MachineHack, in association with Embold, has recently launched a brand new hackathon — GitHub Bugs Prediction Challenge — where participants need to predict bugs on the GitHub titles and text body. The registration is now open and the hackathon closes on 18th of October 2020. Embold.io is a software quality platform that enables leveraging quality […]","categories":["Deep Tech"],"tags":["Data Scientist","Hackathon","Machinehack Hackathon","New Hackathon","New Hackathon For Data Scientists"],"author_name":"Sejuti Das","publish_date":"2020-10-08T14:37:04","publication_year":"2020","word_count":392,"keywords":["Go","machine learning","programming_languages:R","AI","programming_languages:Go","Git","RAG","Hackathon","New Hackathon","New Hackathon For Data Scientists","Machinehack Hackathon","GitHub","GAN","Data Scientist","R"],"extracted_tech_keywords":["AI","machine learning","RAG","R","Go","Git","GitHub","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/new-hackathon-for-data-scientists-github-bugs-prediction-challenge\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10060162,"title":"LatentView Analytics sees 38% revenue increase in third quarter","content":"LatentView Analytics has witnessed a 38% increase in revenue on a year-on-year basis in the third quarter and saw a 14% growth sequentially. The company’s profit after tax rose 39% on a year-on-year basis. “We are overwhelmed by the response we received for our IPO and we extend a warm welcome to all our new shareholders who made our IPO listing successful. We are happy to report a 38% increase in revenue from operations in Q3 FY22 on a y-o-y basis and a strong 14% growth on a sequential basis. For the nine months ending December 2021, our revenue from operations grew by 28% compared to the same period in the last Fiscal Year. This was driven by growth across existing clients and new client additions. We added 15 new clients over the last nine months ending December 2021 and 6 new clients in Q3FY22.” Rajan Sethuraman, Chief Executive Officer, LatentView Analytics said. The company also won multiple client contracts during the third quarter, including: For an American cryptocurrency exchange platform, LatentView Analytics was brought on board to provide expertise in areas related to Marketing, Customer Analytics and PartnershipsFor a leading American mobile app offering vehicles for hire, the company is building data and analytics capabilities across their Fraud and Identity, Payments, Support and Safety For the world second largest wine and spirits seller, LatentView Analytics is helping the Portfolio Growth Team to improve total market share of spirits across the USA For an American financial services company that introduced commission-free investment, LatentView Analytics is assisting them to deliver Legal Query Resolution For an American big-box retail chain, LatentView Analytics has been shortlisted to work as an analytics partner to build forecasting solution for demand and inventory management For an American technology giant, the company has been shortlisted to revamp existing reporting mechanisms and build data pipelines for accurate data for their cloud-based voice service","excerpt":"For the nine months ending December 2021, the company’s revenue from operations grew by 28% compared to the same period in the last Fiscal Year.","categories":["AI News"],"tags":["latentview","Revenue"],"author_name":"SharathKumar Nair","publish_date":"2022-02-08T19:17:02","publication_year":"2022","word_count":315,"keywords":["latentview","programming_languages:R","AI","Revenue","IPO","data pipeline","analytics","R"],"extracted_tech_keywords":["AI","analytics","R","data pipeline","IPO","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/latentview-analytics-sees-38-revenue-increase-in-third-quarter\/","complexity_score":2,"technical_depth":6,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10166392,"title":"Cadence, NVIDIA Extend Partnership for Accelerated Computing and Agentic AI","content":"Cadence Design Systems, a leading computational software company, has announced that it is expanding its multi-year collaboration with NVIDIA, focusing on accelerated computing and agentic AI. This partnership addresses global technology challenges by driving innovation across various industries and involves Cadence leveraging NVIDIA’s latest Blackwell architecture to accelerate its engineering and scientific solutions. This includes reducing computational fluid dynamics simulation times by up to 80 times, from days to minutes, and accelerating the Cadence Spectre X Simulator by up to 10 times. Jensen Huang, NVIDIA CEO, noted, “Accelerated computing and agentic AI are setting new standards for innovation across industries.” Using its Fidelity CFD Platform, Cadence also successfully ran multi-billion cell simulations on NVIDIA GB200 GPUs in under 24 hours. It would have previously required a top 500 CPU cluster with 100,000 cores and several days to complete. The company expressed that it will continue to leverage Blackwell for simulation and help the aerospace industry reduce the amount of wind tunnel tests by reducing cost and expediting time to market. New Era for Accelerated Computing Additionally, the partnership involves the companies working together on a full-stack agentic AI solution for electronic and system design, as well as science applications. This will integrate Cadence’s JedAI Platform with NVIDIA’s NeMo generative AI framework and the Llama Nemotron Reasoning Model. Anirudh Devgan, president and CEO at Cadence, says, “We’re enabling the delivery of today’s infrastructure AI and agentic AI and transforming the principled simulations that underpin physical AI and sciences AI.” The collaboration is expected to transform industries by enabling complex simulations that were previously impossible, driving efficiency, and fueling scientific discovery. It will also deliver breakthroughs in simulation, optimisation, and design. In his keynote address at the NVIDIA 2025 GTC summit, he mentioned that until now, the giant had been using general-purpose computers running software super slowly to design accelerated computers for everybody else. But with the entry of optimised CUDA software, “now our entire industry is going to get supercharged as we move to accelerated computing.” Cadence Molecular Sciences (OpenEye) is also integrating NVIDIA BioNeMo NIM microservices with its cloud-native molecular design platform, Orion. Cadence has also been one of the first adopters of NVIDIA Omniverse Blueprint for AI factory digital twins. Cadence and NVIDIA are leading the way in creating an ecosystem of high-quality models, allowing equipment manufacturers and data centre companies to quickly create digital twins.","excerpt":"Cadence will leverage NVIDIA’s Blackwell architecture for engineering and scientific solutions.","categories":["AI News"],"tags":["cadence","NVIDIA","partnership"],"author_name":"Sanjana Gupta","publish_date":"2025-03-20T17:04:11","publication_year":"2025","word_count":397,"keywords":["CUDA","partnership","Go","agentic AI","AI","cadence","Git","RAG","microservices","generative AI","NVIDIA","R"],"extracted_tech_keywords":["AI","generative AI","agentic AI","RAG","microservices","CUDA","R","Go","CUDA","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/cadence-nvidia-extend-partnership-for-accelerated-computing-and-agentic-ai\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10044513,"title":"IIT Jodhpur To Offer B.Tech In AI &#038; Data Science","content":"Indian Institute of Technology (IIT) Jodhpur is offering eight flagship B. Tech courses to meet the emerging demand. The courses include: AI and Data ScienceComputer Science and EngineeringMaterials EngineeringChemical EngineeringBioengineeringElectrical EngineeringMechanical EngineeringCivil and Infrastructure Engineering “IIT Jodhpur is committed to deliver a strong student-focused educational experience and to provide quality education with a contemporary and highly professional curriculum. Opportunities for multidisciplinary exploration of engineering problems in teams, with sustainability in focus, makes IIT Jodhpur’s education unique and future-ready,” said Prof. Santanu Chaudhury, Director, IIT Jodhpur, while highlighting his vision for the Institute’s unique curriculum. The curriculum for its B. Tech. in Artificial Intelligence and Data Science (AIDE) straddles computer science, mathematics, artificial intelligence, machine learning, data science, and their applications across numerous domains. The program will teach: Biometrics Dependable AI Resource-constrained AI Big Data Data Engineering and Data Visualization Stream Analytics Social Network Analysis Human-Machine Interaction Augmented and Virtual Reality Speech and Natural Language Understanding Computational Neuroscience With design thinking and social connect embedded into the curriculum, these programs target to produce ‘global engineers’ who will be technology leaders and innovators with empathy and sensitivity for sustainable development. Moreover, with the emergence of multidisciplinary research and technologies globally, the creative function of engineering is expected to cover the broader context of the new patterns of knowledge creation across traditional disciplines. Recent IIT initiatives IIT Roorkee is offering two new Masters of Technology (MTech) programmes in Artificial Intelligence (AI) and Data Science from the academic session 2021-22. IIT Ropar has recently introduced an online course in Data Science and Artificial Intelligence in collaboration with Punjab Skill Development Mission (PSDM), Government of Punjab for all aspiring students with Class 12 eligibility.","excerpt":"Opportunities for multidisciplinary exploration of engineering problems in teams, with sustainability in focus, makes IIT Jodhpur’s education unique and future-ready.","categories":["AI News"],"tags":["Masters in Data Science"],"author_name":"kumar Gandharv","publish_date":"2021-07-26T15:23:02","publication_year":"2021","word_count":279,"keywords":["big data","data science","Go","machine learning","artificial intelligence","AI","Masters in Data Science","data engineering","ViT","analytics","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","data science","analytics","R","Go","big data","data engineering","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/iit-jodhpur-to-offer-b-tech-in-ai-data-science\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10102080,"title":"While TCS Leads, Tech Mahindra Makes Strides in Generative AI","content":"In a recent earnings call, Tech Mahindra announced its plans to achieve big milestones in generative AI. During the call, CP Gurnani, managing director, Tech Mahindra said, “Tech Mahindra is now working in about 60 customer locations on actually using generative AI to enhance operations, innovation, and productivity.” In contrast, TCS  is engaged in hundreds of GenAI projects for its clients across segments. Meanwhile, Infosys is working on 90 generative AI programs, Wipro has doubled the number of customers as compared to the last quarter. HCLTech is working with a handful of customers with generative AI projects, while LTIMindtree is engaged in over 20 clients for generative AI. Tech Mahindra currently sees generative AI as a means to automate routine tasks and freeing up talent for more innovative work among other initiatives. In August, the company had announced that it plans to train about 8,000 employees as it readies itself to cater to demand around generative AI and quantum computing. Source: Belamy (AI Weekly Newsletter) Slew of GenAI Solutions In April, Tech Mahindra became the first IT giant to launch something like a Generative AI Studio. ​​The IT solution provider introduced TechM amplifAI0->∞, a comprehensive suite of Artificial Intelligence (AI) offerings and solutions aimed at democratizing and responsibly scaling AI deployment. Over time, the company has continuously integrated new tools into this suite. In August, Tech Mahindra partnered with Google AI to develop generative AI-powered Email AmplifAIer, a solution that uses generative AI to automate email responses and to personalise email communications. Last month, the company introduced another solution called ‘Ops amplifAIer’ solution. It aims to enhance the productivity of support engineers by offering a unified integrated view with all the contextual information and tools needed to resolve issues. It also facilitates team collaboration and generative AI assistance capabilities, ensuring that processes are future-proofed in a responsible manner. Tech Mahindra’s Ops amplifAIer solution integrates with existing ITOps tools to collect the contextual information related to an IT ticket\/alert and uses generative AI to analyze the collected data. It further identifies the probable root cause, diagnoses, recommends remediation actions, and generates the corresponding automation scripts. The solution comes with an enterprise automation catalogue that enables the reuse of automation artefacts like scripts or workflows across the enterprise. Early this month, Tech Mahindra released ‘Vision amplifAIer’ solution. This solution is designed to enhance computer vision-related use cases for enterprises, offering comprehensive end-to-end lifecycle management for computer vision (CV) projects. The emphasis is placed on streamlining the entire process, making it more accessible and efficient. Apart from the enterprise solutions, Tech Mahindra is working on an indigenous LLM called Project Indus that would have the ability to speak in many Indic languages, most notably Hindi. The model will have the ability to speak in 40 different Indic languages, to begin with. Tech Mahindra is not alone During a recent earnings call, TCS’ executive vice president and global head of human resources Milind Lakkad highlighted their workforce’s readiness for generative AI, boasting over 100,000 GenAI-ready employees. TCS CEO K Krithivasan emphasised their success with TCS Cognix, utilising generative AI for operating model transformation deals, enabling businesses through dashboards and predictive analytics. Infosys, under CEO Salil Parekh, discussed their generative AI capability, Topaz, which significantly increased their market share. Parekh revealed Infosys’ engagement in over 90 generative AI programs, emphasising their commitment to helping clients navigate the future with deep technical expertise. To support this, Infosys trained 57,000 employees in generative AI principles. “We have trained as many as 180,000 employees in basic GenAI principles,” said Wipro chief Thierry Delaporte. He said that the company has rolled out personal-based learning pathways to create a pool of specialised talent with deeper technical expertise. It has ambitions to train over 250,000 employees in the coming months. Recently, it also launched a new GenAI Center of Excellence, in collaboration with IIT Delhi. Delaporte said it is rapidly integrating generative AI into its processes, solutions and offerings. “Thousands of our employees have or are starting to use generative AI. Today, we are seeing a doubling of GenAI active projects than we did just one quarter ago,” added Delaporte, pointing out how healthcare, consumer and financial services, high tech and utilities are seeing rapid adoption. HCLTech also announced a slew of new deals, where it will be involved in digital and cloud transformation, alongside generative AI initiatives. HCLTech chief C Vijayakumar said that the company is working on a two-pronged approach – for its clients, it looks to generate early-stage opportunities, alongside training its delivery organisation to leverage generative AI for core development, deployment, and testing and managed services. So far, it looks to train close to 18,000 employees. LTIMindtree has expressed its commitment to integrating generative AI into its products and solutions, revealing that they have participated in more than 100 discussions and currently have over 20 active engagements with clients, according to LTIMindtree chief, Debashis Chatterjee. Additionally, the company is set to launch two new offerings: Canvas Lite, designed for developers focusing on productivity use cases, and Canvas Control Claim, which facilitates secure, moderated, and responsible AI use. This offering provides clients with the option to choose from commercial open-source and custom element models. Chatterjee also shared the company’s plan to train over 10,000 employees by the end of the third quarter. In conclusion, Tech Mahindra is a unique player in Indian IT, which is making strides in generative AI, while others – TCS, Infosys, Wipro, HCLTech, and LTIMindtree – are focusing on enterprise solutions, upskilling and reskilling, alongside experimenting with real use cases.","excerpt":"Tech Mahindra is working in about 60 customer locations on actually using generative AI to enhance operations","categories":["IT Services"],"tags":["TCS","Tech Mahindra"],"author_name":"Siddharth Jindal","publish_date":"2023-10-26T18:02:40","publication_year":"2023","word_count":926,"keywords":["Tech Mahindra","artificial intelligence","GenAI","AI","ML","predictive analytics","computer vision","RAG","Aim","analytics","generative AI","TCS"],"extracted_tech_keywords":["AI","artificial intelligence","ML","computer vision","analytics","generative AI","GenAI","Aim","RAG","predictive analytics"],"url":"https:\/\/analyticsindiamag.com\/it-services\/while-tcs-leads-tech-mahindra-makes-strides-in-generative-ai\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10140152,"title":"After China, Meta Just Hands Llama to the US Government to ‘Strengthen’ Security","content":"Meta is now making Llama available for US government agencies, defence projects and other private sectors working on national security. They’re also extending their partnership with companies like Accenture Federal Services, Amazon Web Services, Anduril, Booz Allen, Databricks, Deloitte, IBM, Leidos, Lockheed Martin, Microsoft, Oracle, Palantir, Scale AI and Snowflake to help government agencies adopt Llama. “These kinds of responsible and ethical uses of open source AI models like Llama will not only support the prosperity and security of the United States, they will also help establish U.S. open source standards in the global race for AI leadership.” said Nick Clegg, President of Global Affairs in a blog post published from Meta. The blog post also highlighted how their partners are aiding the adoption of Llama. For example, AWS and Azure are supporting governments by hosting their models on their secure cloud services. Lockheed Martin has also already integrated Llama in their factories, leveraging its capabilities for code generation, and data analysis. “Large language models can support many aspects of America’s safety and national security. They can help to streamline complicated logistics and planning, track terrorist financing or strengthen our cyber defences,” added Clegg. China Is Reportedly Using Llama for an AI Chatbot The announcement comes after reports that China was rumoured to be using Llama for its military applications. Researchers linked to the People’s Liberation Army are said to have built ChatBIT, an AI conversation tool fine-tuned to answer questions involving the aspects of the military. It didn’t take long to invoke fear in the AI community, with torchbearers like Vinod Khosla going all the way to criticise Meta’s Open-Source approach. There are specific technologies that should clearly remain under control to slow down their dissemination among adversaries. Those are things whose main usage is in defense: nuclear tech, hypersonic vehicles, radar-absorbing materials, etc. Then, there is general purpose…— Yann LeCun (@ylecun) November 2, 2024 While his vested interest in OpenAI is bound to find an opportunity to call out Meta, Yann LecCun, Meta’s Chief AI scientist, did not hold back. He said, “There is a lot of very good published AI research coming out of China. In fact, Chinese scientists and engineers are very much on top of things (particularly in computer vision, but also in LLMs). They don’t really need our open-source LLMs.” It’s fair to agree with what Yann LecCun is trying to say. China has already made notable progress in generative AI, and it was even claimed that Kai-Fu Lee’s foundational model ranks better than GPT 4o on certain benchmarks. Their indigenous approach towards advancements in technology is likely not going to change in the AI sector as well. Clegg also added that “Widespread adoption of American open-source AI models serves both economic and security interests. Other nations—including China and other competitors of the United States—understand this as well and are racing to develop their own open-source models, investing heavily to leap ahead of the U.S.” Earlier this year, the US Army announced that it’s investing $50 million in ‘small and nontraditional businesses’ to develop AI and ML solutions. Recently, the US Army also launched a generative AI platform called Ask Sage, which assists personnel in several aspects of software development.","excerpt":"Meta’s stance to help government agencies leverage their open-source AI models comes after China’s rumoured adoption of Llama for military use.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","Meta","Meta llama"],"author_name":"Supreeth Koundinya","publish_date":"2024-11-05T10:59:28","publication_year":"2024","word_count":536,"keywords":["Meta","OpenAI","AI","AWS","ML","computer vision","RAG","Meta llama","Aim","generative AI","Azure","AI (Artificial Intelligence)","Snowflake"],"extracted_tech_keywords":["AI","ML","computer vision","generative AI","OpenAI","Aim","RAG","AWS","Azure","Snowflake"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/after-china-meta-just-hands-llama-to-the-us-government-to-strengthen-security\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10023184,"title":"ML-based Training Platforms Can Upgrade Players’ Gaming Strategies: Tarun Gupta, Ultimate Battle","content":"While technology has always played a vital role in the eSports ecosystem, the practice of social distancing — moving offline sporting events to online — gave a significant push to technological advancements for the online gaming industry. Alongside, India having the cheapest mobile data globally makes it even more favourable for the eSports industry. In the financial year 2020, India’s online gaming industry had a market value of around ₹90 billion, which has been predicted to reach over ₹143 billion by 2022, as per a report. Analytics India Magazine got in touch with Tarun Gupta, the Founder of Ultimate Battle, to understand how technology determines the growth of the eSports industry and how it is leveraging deep-tech to navigate through the times of pandemic. Founded in 2017, Ultimate Battle — a subsidiary of Brainpulse Technologies — is India’s first-ever one-stop online eSports platform for players to connect, interact and compete with each other in their favourite game titles. Edited Excerpts — AIM: What AI and ML technologies are driving the eSports industry towards astonishing growth? How has the scenario changed post-COVID? Tarun Gupta: We have recently seen a massive usage of artificial intelligence to clip-out relevant match highlights from multiple long gameplay sessions being played professionally or streamed across platforms by content creators. This helps in delivering top quality eSports content to a massive gaming audience worldwide. Additionally, process automation plays a huge role in the scalability of operations in eSports. AI\/ML has also been extensively used to create gaming bots, trained to play a video game in the place of a human. The bots learn the game in their own environment to further utilise the knowledge as a learning tool for new or advanced players. In the past, the match highlights, or in-game analysis of player performance – used to be endgame details due to the amount of time it used to take to process the information. However, with the provisioning of AI and in-game analytics, viewers get critical information about the game in a real-time environment. AI, VR and ML technologies had already been there before the pandemic. However, now, it has just helped give a positive push in the sense of feasibility of operations in organising eSports events. Features like creating gaming content, in-game analysis, and learning new games have been achievable with the use of AI\/ML. VR has its own space, but still in its nascency. There’s a massive potential in VR gaming as this will open a completely new vertical in the space of spectating the game. Just imagine being inside the game looking at the players from multiple perspectives as you stroll by the game environment\/map. AIM: What is the role of AI\/ML in making eSports available to the audience? How is virtual reality in eSports gaining momentum? Tarun Gupta: Video-on-demand content is one of the core focuses in all eSports events. AI technologies are being used to create match highlights. Real-time in-match analytics using AI helps the audience connect to the game at a different level. So, artificial intelligence will play a key role for the eSports industry in the coming years for sure. Moreover, we are already seeing players using advanced ML-based training platforms to upgrade their skills and gaming strategies. NLP tools coupled with AI will increasingly personalise the player-audience interactive experience and can gain traction soon. On the other hand, VR eSports is fully delved-in gameplay where the physical reflexes of the player play a huge role. VR interface extends users’ perception into their reality, making the gameplay more fun and intriguing to all gamers. We see VR eSports games gaining momentum in the near future, but the scale will largely depend on the generalisation of the technology. AIM: What sort of technology Ultimate Battle is for its platform? Tarun Gupta: We use RPA technology for process automation on Ultimate Battle, making the platform operations faster and scalable. For content creation, AI-backed tools are also used to create match highlights from our gaming events, helping us keep up with the player demands. Ultimate Battle for long has maintained a solid stance on top-notch execution; we can do nothing less than the best. Upping the game, we are currently working on state-of-the-art broadcast tech developed by highly skilled individuals. We are also looking into integrating AI-driven chatbot systems, which will help address players queries and issues in real-time. Further, in the scope of enhancing player gameplay, we are looking into providing game analysis features that will help players understand their gameplay better and ways to improve it simultaneously. AIM: Being from a technical background with more than 20 years of experience, what’s your take on the long-term adoption of these technologies in the eSports sector? Tarun Gupta: The use of AI, ML, or VR technologies will become an integral part of the gaming ecosystem as they have the power to change how things can be done more effectively, efficiently and accurately. It is clear that eSports as an industry will use such technologies to optimise their processes and operation, making them more effective. Applying these technologies can be something as little as automatically generating content or training gamers using ML and AI. With new VR games of the future, anybody can fancy the sort of experience they want to create. I only look forward to a fascinating journey that is yet to unfold.","excerpt":"Adaptation of AI\/ML technologies in various forms can further fuel esports’ growth, and improve players’ gaming strategies.","categories":["AI Features"],"tags":["Interviews and Discussions"],"author_name":"kumar Gandharv","publish_date":"2021-04-01T15:00:00","publication_year":"2021","word_count":892,"keywords":["Go","artificial intelligence","AI","ML","Scala","RAG","NLP","Aim","analytics","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","artificial intelligence","ML","NLP","analytics","Aim","RAG","R","Go","Scala"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/ml-based-training-platforms-can-upgrade-players-gaming-strategies-tarun-gupta-ultimate-battle\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":61598,"title":"L&#038;T Uses Artificial Intelligence To Support 20 Cities In Combating Against COVID-19","content":"Larsen & Toubro has recently announced that it is using artificial intelligence technology in order to support local authorities of 20 cities to monitor and implement measures to combat against COVID-19. Considering this pandemic has created a high demand for digital services for healthcare and monitoring, L&T has introduced this smart technology solution in these cities so that it can help local administrations to monitor crowds, track patients, communicate with the public, and to maintain law and order. SN Subrahmanyan, chief executive officer and managing director at L&T, stated that the company’s Smart World & Communication has been working with various city administrations and state governments to manage complex civic functions in metropolitan cities effectively. He said, “With its ability to seamlessly integrate various smart technologies to manage surveillance, crowd management, message delivery and communication to the residing populace, L&T is supporting the transformation of city civic administrations.” Currently L&T, In India, has partnered with municipal and police agencies to set up technology and manage command and control centres or city operations centres in regions such as Mumbai, Pune, Nagpur, Prayagraj, Ahmedabad, Visakhapatnam, and Hyderabad. Earlier, L&T smart world and communications offerings were used only during natural disasters like cyclones and to manage large gatherings like Maha Kumbh in Prayagraj; however, this can now be used to fight COVID-19. Considering Hyderabad people, it has been imposed that nobody can drive beyond 3-km from their residences even for buying essential goods. Therefore in order to help police enforce the curbs, L&T is offering a machine-learning-based automated number plate recognition system. L&T is also using artificial intelligence and crowd formation analytics that trigger real-time alerts at the police command centre of cities like Mumbai and Hyderabad in order to help them remove the crowd. The company is also working on other innovative solutions, to help local authorities of different cities, like chatbot-based emergency pass system in Telangana, drones-based surveillance for crowd formation at Mumbai, and fever detection using thermal cameras at Mumbai.","excerpt":"Larsen & Toubro has recently announced that it is using artificial intelligence technology in order to support local authorities of 20 cities to monitor and implement measures to combat against COVID-19. Considering this pandemic has created a high demand for digital services for healthcare and monitoring, L&T has introduced this smart technology solution in these […]","categories":["AI News"],"tags":["AI (Artificial Intelligence)","Artificial Intelligence India","covid-19","L&amp;T","L&amp;T Infotech","L&amp;T technology services","machine learning for cities"],"author_name":"Sejuti Das","publish_date":"2020-04-13T18:06:00","publication_year":"2020","word_count":330,"keywords":["Go","artificial intelligence","covid-19","L&amp;T Infotech","AI","programming_languages:R","ML","L&amp;T","Artificial Intelligence India","Git","Ray","L&amp;T technology services","analytics","GAN","R","AI (Artificial Intelligence)","machine learning for cities"],"extracted_tech_keywords":["AI","artificial intelligence","ML","analytics","Ray","R","Go","Git","GAN","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/lt-uses-artificial-intelligence-to-support-20-cities-in-combating-against-covid-19\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10140898,"title":"Is it the End of Human Animators and Modellers As We Know Them?","content":"As AI advances, the boundaries of 3D modelling are expanding, offering unprecedented opportunities for creativity and innovation. New technologies are already redefining how we interact with digital environments, making them more responsive, intuitive, and lifelike. Los Angeles-based Wonder Dynamics, a company under Autodesk that provides software for VFX and CG, announced the release of ‘Wonder Animation’ in October this year. The company claims that this new technology can transform real videos into 3D scenes with customisable camera setups, full body\/face animation, and editable elements—all in one 3D space. In a post on Autodesk’s official blog, the company said that it is aware of the misconception that “AI is a one-click solution”. It aims to bring artists closer to full animation, ensuring creative control and avoiding the black-box approach of typical generative AI tools. This year, Google Research and DeepMind teamed up to release a paper titled, ‘CAT3D: Create Anything in 3D with Multi-View Diffusion Models’. CAT3D enables real-time, multi-angle viewing of 3D models and can generate 3D scenes from a single image, set of images, or a single text prompt – all in under a minute. It outperforms existing methods, particularly when working with limited images. For instance, platforms like Alpha3D and Mixamo offer automated rigging services, simplifying the animation process for 3D characters. Another platform, 3DFY, revolutionises architecture by enabling rapid, high-quality 3D model generation from minimal input, streamlining design processes and enhancing visualisation capabilities. Everyone Gets a Power-Up AI is transforming animation and 3D modelling, impacting artists, studios, developers, and end-users alike. By automating repetitive tasks, AI frees artists to focus on creativity, while studios benefit from faster, cost-effective production. Independent developers and small studios gain access to advanced tools, like those from startups such as 3DAiLY, levelling the playing field. End-users experience richer, more immersive environments in gaming, VR, and digital fashion, thanks to AI-driven motion blending and modelling. However, adapting to new workflows underscores the need for continuous learning in this rapidly evolving field. Martin Nebelong, a 3D artist exploring VR and AI, says on X, “I used a 3D model, generated from an image, as a ‘driver’ for AI animation. In the near future, we’ll see more 3D tools that support these types of workflows.” Despite major advancements, AI doesn’t seem to replace human animators and modellers but rather augment their capabilities. AI tools handle repetitive and time-consuming tasks, freeing artists to concentrate on storytelling and design. From Static to Cinematic Earlier, AIM reported that Meshcapade, a leader in AI-driven 3D modelling, announced an advanced motion-blending feature that enhances the realism of digital avatars by enabling smooth transitions between various human movements. This innovation reflects a broader trend in the industry, where AI is revolutionising 3D modelling by transforming static models into dynamic, lifelike representations. Not just the above companies, even OpenAI has tried its hand in this domain. In May 2023, OpenAI introduced Shap-E, a conditional generative model capable of creating 3D assets from text prompts. Unlike traditional models that produce a single output, Shap-E generates parameters for implicit functions, allowing for the rendering of both textured meshes and neural radiance fields (NeRFs). This advancement enables the rapid generation of complex and diverse 3D assets, significantly accelerating the modelling process. It also takes the text-to-video models ahead significantly. Javi Lopez, the co-founder of Magnific AI, recently posted about another tool on X. ⚡ Well, isn't that something?!\"From-surface-to-3d\" in REAL TIME! 🤯💡 This is an AI 3D app demo created just using a mix of Open Source models. Imagine the endless possibilities when smart people start piecing together all the different AI APIs!pic.twitter.com\/Fp3SfcrDIp— Javi Lopez ⛩️ (@javilopen) November 8, 2023 ‘Common Sense’ into AI? Another notable player in this domain is Common Sense Machines (CSM), co-founded by former Google DeepMind scientist Tejas Kulkarni. CSM focuses on infusing ‘common sense’ into AI systems, aiming to bridge the gap between artificial intelligence and human-like understanding. AI is enabling the generation of 3D models from minimal input. The platform facilitates the creation of game-engine-ready 3D content from inputs such as images, text, and sketches, streamlining workflows for artists and developers. As reported by AIM earlier, this innovation reduces the time and expertise required to produce high-quality 3D assets. However, it also raises questions: Can we trust machines to make judgement calls yet? Adding ‘common sense’ might make AI smarter, but could it also create new risks in how AI understands and interacts with us? Changing the Gaming Verse These developments underscore a significant shift in 3D modelling, where AI not only automates the creation of complex models but also imbues them with realistic motion and behaviour. As AIM reported earlier, 3DAiLY, an Indian AI startup, is revolutionising the gaming industry by creating ultra-realistic, production-ready 3D models for both AAA and indie games. Leveraging generative AI, 3DAiLY offers high-quality assets at a fraction of traditional costs, making advanced 3D content creation more accessible. Their models are fully rigged and compatible with various gaming engines, including Unity, Unreal, and CryEngine. By combining artistic intuition with AI technology, such companies are breaking down barriers in 3D modelling, enabling developers to produce immersive video gaming experiences more efficiently. The integration of AI technologies like motion blending, generative modelling, and common sense reasoning is paving the way for more immersive and interactive digital experiences across various industries, including gaming, virtual reality, and digital fashion.","excerpt":"Wonder Animation is a tech that transforms real video into 3D scenes with customisable camera setups, full body\/face animation, and editable elements.","categories":["AI Features"],"tags":["3d AI"],"author_name":"Sanjana Gupta","publish_date":"2024-11-13T14:56:07","publication_year":"2024","word_count":891,"keywords":["Go","artificial intelligence","TPU","OpenAI","AI","ML","3d AI","RAG","Aim","generative AI","R"],"extracted_tech_keywords":["AI","artificial intelligence","ML","generative AI","OpenAI","Aim","RAG","TPU","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/is-it-the-end-of-human-animators-and-modellers-as-we-know-them\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10042057,"title":"Image Blending Using Pyramids In OpenCV","content":"In today’s era, images play a vital role in cognitive development as visuals hold more appeal than plain text to curious and intuitive minds. Visuals help improve learning enormously and on multiple levels. Studies have shown that around 65% of the population constitute visual learners. Due to this reason, study and research revolving around images, such as image processing, image recognition, and image blending, have created a buzz among masses of people developing new technologies. Image processing programs such as  Photoshop and other freeware alternatives such as GIMP and Paint.net offer a wide range of possibilities to edit and manipulate images and photos. Our idea would be to explore the domain further, to discover what more can be done to perform operations that are not typically built in existing software and find new ways to teach and educate oneself. What is Image Blending? Image Blending is mixing two images of the corresponding pixel values to create a new target image. The concept of blending images is comparatively very easy. To achieve this, we can simply make a copy of an image and transfer each source pixel’s values into a pixel in the target image. With the help of Python and the OpenCV library, it becomes relatively very easy to work with visual inputs such as images or videos. The OpenCV library contains most of the functions that we would require for working with images. Using OpenCV, we can combine the photos, we combine layer masks and gradients and explore other numerous possibilities. The Main Aspects for Image Blending To start working with Image Blending, we would need to know the bare basics of image processing techniques. Image Blending techniques mainly focus on two major aspects for creating a blended image by using Image Pyramids, namely Gaussian and Laplacian Pyramids. A pyramid or pyramid representation is a type of multi-scale representation where an image is repeatedly smoothened and subsampled. It searches for a target pattern over many repetitive scales. At first from the image, it creates copies of the target pattern and reconstructs it over several expanded scales. Each copy is then convolved with the original image, creating a new and unique image. The Gaussian Pyramid technique focuses on resizing the image, reducing it to one fourth in every cycle. Its goal is to define the representation of what information in the image is explicitly available at different scales. It has several other applications such as  Progressive image transmission, Scale-invariant template matching and Efficient feature search. Here, Image cross-correlation, convolution and matrix multiplication are the core aspects of this operation.  The Laplacian Pyramid technique, on the other hand, keeps track of differences between the multiple sizes of images and stores the information for further use. It does so by gray scaling the image to be processed, focusing on the image outlines at different sizes, and comparing its differences. It also uses a linear transformation to decompose an image into various components by multiplying it with a set of transformation functions. It can be termed as a filter that detects features from the image and stores them in memory for further processing. Getting Started with Code Now that we know how the Image Pyramid works, we will try to implement a basic image pyramid for image blending using both Gaussian and laplacian techniques. We have used two images with two different backgrounds on purpose to see how the blending technique comes into play and compare the difference between a merged image and a blended image. The below code is in reference to the official implementation, whose video tutorial you can find here. Installing The OpenCV Package and Importing It To Install OpenCV, you can use the following command : !pip install opencv-python Then to import the package and NumPy, use import cv2 import numpy as np Now, we will set a path for our images of the directory they are present in. Here, were are going to blend two images, an image of an apple and an orange. apple_path = '\/content\/OtherApple.jpg' orange_path = '\/content\/Orange.jpg' apple = cv2.imread(apple_path) orange = cv2.imread(orange_path) Printing The Shape and Size of Our Images Imported #printing the shape of our images print(apple.shape) print(orange.shape) We will get the following output (512, 512, 3) (512, 512, 3) Creating Left and Right Halves for Merging Now that we have the sizes of our images let’s divide them into a right and left half to merge them together as one. We’ll use the hstack function to feed them into the system as tuples. apple_orange = np.hstack((apple[:, :256], orange[:, 256:])) #Dividing them into right and left half as tuples using hstack Let’s check our results so far. To plot this, we will be using matplotlib. import matplotlib.pyplot as plt plt.imshow( apple_orange[:, :, ::-1 ]) As we can observe, there is still a visible line in between the two images. However, they have been merged but not blended. Image Blending will now come to play to blend both the images properly. Steps to create an Image Blender First, we will create a gaussian pyramid for both the apple and orange image. Next, from the created Gaussian pyramid, further process and find the Laplacian pyramid.Rejoin the left half of the apple image and right half of the orange image in each level of Laplacian pyramids Blend the merger line and reconstruct the original image Generating Gaussian pyramid for apple and orange Here we are using a range as six, which is a very commonly used range. Range defines the number of times you want to reduce the resolution. One can tune this accordingly. Cv2.pyrDown helps us execute the gaussian pyramid. # generate Gaussian pyramid for apple apple_copy = apple.copy()#create a copy of the apple image gp_apple = [apple_copy] #creating the gaussian pyramid and pass as a list for i in range(6): apple_copy = cv2.pyrDown(apple_copy) gp_apple.append(apple_copy) Applying the same steps for orange image, # generate Gaussian pyramid for orange orange_copy = orange.copy() gp_orange = [orange_copy] for i in range(6): orange_copy = cv2.pyrDown(orange_copy) gp_orange.append(orange_copy) Generating Laplacian pyramid for apple and orange from gaussian We are going to make use of the 5th element from the Laplacian pyramid. The level in the Laplacian pyramid is created by the difference between Gaussian and extended level to its upper level in the gaussian pyramid. The Laplacian method also makes use of cv2.pyrUp to rescale the image. # generate Laplacian Pyramid for apple apple_copy = gp_apple[5] lp_apple = [apple_copy] for i in range(5, 0, -1): gaussian_expanded = cv2.pyrUp(gp_apple[i]) laplacian = cv2.subtract(gp_apple[i-1], gaussian_expanded) lp_apple.append(laplacian) #Doing the same from orange, # generate Laplacian Pyramid for orange orange_copy = gp_orange[5] lp_orange = [orange_copy] for i in range(5, 0, -1): gaussian_expanded = cv2.pyrUp(gp_orange[i]) laplacian = cv2.subtract(gp_orange[i-1], gaussian_expanded) lp_orange.append(laplacian) Merging the Left and Right Halves Using the zip function, we are merging both the laplacian pyramids for both apple and orange. # Now add left and right halves of images in each level apple_orange_pyramid = [] n = 0 for apple_lap, orange_lap in zip(lp_apple, lp_orange): n += 1 cols, rows, ch = apple_lap.shape laplacian = np.hstack((apple_lap[:, 0:int(cols\/2)], orange_lap[:, int(cols\/2):]))#dividing columns of both images into half apple_orange_pyramid.append(laplacian)#appending the variable to list Reconstruct the Image The final step is to reconstruct our image. We will first create a variable, create a ‘for’ loop to loop from 1 to 6 range, and add all the pyramid layers created. # now reconstructing our image using pyrUp and stating pyramid levels apple_orange_reconstruct = apple_orange_pyramid[0] for i in range(1, 6): apple_orange_reconstruct = cv2.pyrUp(apple_orange_reconstruct) apple_orange_reconstruct = cv2.add(apple_orange_pyramid[i], apple_orange_reconstruct) Loading our final product and Checking Result #using -1 to maintain coloured saturation plt.imshow( apple_orange_reconstruct[:, :, ::-1 ]) As you may notice now, the merge line has now been blurred, and the image seems blended properly. EndNotes By performing the following steps, we have now created and learnt the basics of image processing and blending. We have used such variant images to get a taste and notice the difference in between. You can choose better and properly aligned images to merge, blend, and create your own unique blended image. The full Colab file for the following can be accessed from here. Happy Learning! References About image PyramidsImage Pyramids Methods and TechniquesVideo Tutorial on constructing Image Pyramids in Python","excerpt":"In today’s era, images play a vital role in cognitive development as visuals hold more appeal than plain text to curious and intuitive minds. Visuals help improve learning enormously and on multiple levels. Studies have shown that around 65% of the population constitute visual learners. Due to this reason, study and research revolving around images, […]","categories":["AI Trends"],"tags":["OpenCV","Opencv image processing"],"author_name":"Victor Dey","publish_date":"2021-06-19T12:00:00","publication_year":"2021","word_count":1363,"keywords":["Opencv image processing","NumPy","TPU","AI","image recognition","OpenCV","Colab","Ray","Python","Matplotlib","R"],"extracted_tech_keywords":["AI","Ray","Colab","OpenCV","NumPy","Matplotlib","image recognition","TPU","Python","R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/image-blending-using-pyramids-in-opencv\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10126193,"title":"Koo Founder Announces New Consumer Tech Venture, Seeks Engineering Talent","content":"Mayank Bidawatka, co-founder of the recently shuttered social media platform Koo, has announced plans for a new consumer technology product. The unnamed venture is currently being bootstrapped and is described as a “massive global idea with a huge market size.” Bidawatka revealed that the project has been close to him for many years, and he believes now is the right time to bring it to life. A small team of seven people is already working on the initiative, driven by passion and belief rather than immediate commercial interests. “Will soon start something new. It’s under wraps. It’s being bootstrapped for now. An idea that’s been close to me for many years. It’s a good time to bring it to life. Will be a consumer tech product,” Bidawatka, posted on LinkedIn. The entrepreneur is actively seeking to expand the team, specifically looking for backend engineers, machine learning specialists, and iOS and Android developers. Interested candidates are encouraged to express their interest directly to Bidawatka via email. While details about the product remain undisclosed, Bidawatka emphasised its potential for significant commercial success. However, he noted that the current focus is on building “a beautiful product” rather than immediate monetisation. This announcement comes shortly after Bidawatka and his co-founder Aprameya Radhakrishna announced the closure of Koo, their Twitter-like social media platform. Koo had gained popularity in India, especially during tensions between the Indian government and Twitter in 2021, but ultimately failed to secure necessary partnerships and funding to continue operations.","excerpt":"Bidawatka revealed that the project has been close to him for many years, and he believes now is the right time to bring it to life.","categories":["AI News"],"tags":["ai announcements","Koo"],"author_name":"Siddharth Jindal","publish_date":"2024-07-08T16:41:31","publication_year":"2024","word_count":247,"keywords":["Go","funding","machine learning","programming_languages:R","AI","programming_languages:Go","RAG","ai announcements","R","Koo"],"extracted_tech_keywords":["AI","machine learning","RAG","R","Go","funding","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/koo-founder-announces-new-consumer-tech-venture-seeks-engineering-talent\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10116610,"title":"NVIDIA Introduces Very Big GPU, BLACKWELL","content":"In a blockbuster announcement at its annual GTC conference, NVIDIA CEO Jensen Huang unveiled the company’s next-generation Blackwell GPU architecture, promising massive performance gains to fuel the AI revolution. The GPU is named after David Harold Blackwell, an eminent American statistician and mathematician who has made significant contributions to various fields, including game theory, probability theory, information theory, and statistics. The highlight is the flagship B200 GPU, a behemoth packing 208 billion transistors across two cutting-edge chiplets connected by a blazing 10TB\/s link. Huang proclaimed it “the world’s most advanced GPU in production.” Blackwell introduces several groundbreaking technologies. It features a second-gen Transformer Engine to double AI model sizes with new 4-bit precision. The 5th-gen NVLink interconnect enables up to 576 GPUs to work seamlessly on trillion-parameter models. An AI reliability engine maximises supercomputer uptime for weeks-long training runs. “When we were told Blackwell’s ambitions were beyond the limits of physics, the engineers said ‘so what?'” Huang said, showcasing a liquid-cooled B200 prototype. “This is what happened.” The Blackwell GPU represents a significant advancement in AI inference performance compared to the previous Hopper generation. Compared to its predecessor Hopper, the B200 promises 2.5x faster FP8 AI performance per GPU, double FP16 throughput with new FP6 format support, and up to 30x faster performance for large language models. This dramatic improvement is made possible by Blackwell’s architecture, which is specifically designed to efficiently manage the immense computational and memory demands of trillion-parameter AI models. Major tech giants like Amazon, Google, Microsoft, and Tesla have already committed to adopting Blackwell. Huang said training a GPT model with 1.8 trillion parameters [GPT-4] typically takes three to five months using 25,000 amperes. HOPPER architecture would require around 8,000 GPUs, consuming 15 megawatts of power and taking about 90 days (three months). In contrast, BLACKWELL would need just 2,000 GPUs and significantly less power—only four megawatts—for the same duration. He said that NVIDIA aims to reduce computing costs and energy consumption, thereby facilitating the scaling up of computations necessary for training next-generation models. To showcase Blackwell’s scale, NVIDIA unveiled the DGX SuperPOD, a next-gen AI supercomputer with up to 576 Blackwell GPUs and 11.5 exaflops of AI compute. Each DGX GB200 system packs 36 Blackwell GPUs coherently linked to Arm-based Grace CPUs. While focused on AI and data centres initially, Blackwell’s innovations are expected to benefit gaming GPUs too. With Blackwell raising the bar, NVIDIA has clearly doubled down on its lead in the white-hot AI acceleration market. The race for the next AI breakthrough is on.","excerpt":"Compared to Hopper, the B200 promises 2.5x faster performance per GPU, and up to 30x faster performance for large language models.","categories":["AI News"],"tags":["NVIDIA","NVIDIA GPU"],"author_name":"Shyam Nandan Upadhyay","publish_date":"2024-03-19T02:38:13","publication_year":"2024","word_count":422,"keywords":["Go","NVIDIA GPU","programming_languages:R","AI","innovation","ML","programming_languages:Go","GPT","Aim","NVIDIA","R","llm_models:GPT"],"extracted_tech_keywords":["AI","ML","Aim","R","Go","GPT","innovation","llm_models:GPT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/nvidia-introduces-very-big-gpu-blackwell\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10061270,"title":"How is AI transforming radiology?","content":"Traditionally, radiology has been described as a digitally-powered medical discipline that thrives on digitally analysing binary data. While it did have a first-adopter advantage for quite some time, radiologic procedures today are borderline antiquated. Combine the segment’s technological backwardness with the increasing volume and complexity of data, global storage of radiologists and lack of radiologic expertise, extreme workload\/burnout, and the increasing scope of medical errors, and the need for change becomes apparent!Thanks to the rapid development of artificial intelligence (AI) in understanding images, text, and videos, the much-anticipated revolution in diagnostic radiology is perhaps on the horizon. In fact, AI-powered systems are already being used in computer-aided cancer detection, natural language processing to facilitate critical results reporting, auto-segmentation of organs in 3D postprocessing, radiomics, imaging biobanks, quantification and kinetics in postprocessing, and consultation of best guidelines for recommendations. So if you are looking to modernise your radiology practice with AI, look no further. Analytics India Magazine has curated a list of the best-of-breed AI-powered radiology solutions available in India. Teleradiology Solutions Founded in 2002, Teleradiology Solutions (TRS) is a renowned teleradiology company providing quality emergent and non-emergent reporting services. The group forayed into the healthcare\/medical tech space by introducing Telerad Tech in 2009. They offer an integrated cloud-based RIS\/PACS platform, RADSpa®, that leverages AI to help radiologists review cases automatically and in real-time, benefitting over 20 countries and 150 hospitals across India, Asia, and Africa. As per Dr Arjun Kalyanpur, Chief Radiologist & CEO, Teleradiology Solutions, “It is possible to have a stroke CT scan performed in Boston reported within 15 minutes from across the globe.” TRS’s AI-powered solutions for radiologists promise higher analysis accuracy, volume & consistency, quick and seamless initial screening capabilities, the ability to mine imaging data endlessly, auto-detecting capabilities for specific pathology in various conditions, population management and a learning and automated rad-peer review system that acts as a quality assurance platform. Their tools will also aid radiologists in identifying additional “high-risk” patients and avoiding missed diagnoses. The qER Suite by Qure.ai technologies The qER suite combines AI-enabled neurocritical imaging solutions and telehealth solutions to help manage traumatic brain injuries and reduce the overall time to treatment through remote interventions and progression monitoring. The FDA-cleared and CE-marked solution provides treatment protocol guidance and quantification for midline shift, lateral ventricles, & hypodensities with automatic labelling and visualisation. The platform even allocates ASPECT (Alberta stroke programme early CT) Score to evaluate ischemic stroke severity. In addition, it comes with automated pre-populated reports of findings that clinicians can edit. Given that the suite is hardware & software agnostic, integrating it into existing workflows is seamless. Most importantly, the platform is highly accurate; it has achieved 97% accuracy in abnormality detection, to be specific. Interestingly, Qure.ai and Teleradiology Solutions have joined forces to reduce costs and enable smarter and faster diagnoses of X-ray and CT scan data. This was achieved by integrating Qure.ai’s chest X-ray technology with Teleradiology Solutions’ RADSpa. RADIOLens by Synapsica RADIOlens is an AI-powered PACS radiology workflow solution that automatically detects bad quality scans and creates preliminary reports for some most common modalities, making diagnostic workflows smoother. As of today, Synapsica has deployed this solution in 15+ countries and handled over one million cases annually. AUGMENTO by Deeptek.ai Deeptek.ai aims to democratise radiology with its AI-enabled Cloud PACS platform, AUGMENTO. The solution completely transforms radiology workflows by enabling quantum improvement in productivity, quality, and TAT. Furthermore, it fosters collaboration by seamlessly connecting all stakeholders under a single pane of glass. Its key features include smart reporting with standardised and quantified reports, worklist prioritisation and pathology prediction, and simplified billing. To date, AUGMENTO has served 60+ radiology experts, empowering them with the tools needed to improve revenue goals and achieve work-life balance. Is this the future? Though AI and AI-based solutions are poised to create a paradigm shift in radiology and healthcare in general, there are certain aspects that we need to address before the wave hits. For example, developers need to address issues concerning governance, medico-legal responsibility, and ethics. On the radiologist side, education and training will play a crucial role in integrating AI in their practice effectively. Practitioners and developers need to work in liaison to understand the gaps and ensure effective software\/hardware development, implementation, and retention. Most importantly, we need to continue with these efforts in the interest of the patients, always prioritising their lives.","excerpt":"Using AI, it is possible to have a stroke CT scan performed in Boston reported within 15 minutes from across the globe.","categories":["IT Services"],"tags":["AI (Artificial Intelligence)","Healthcare Automation","healthcare India","Machine Learning","Startups"],"author_name":"Sri Krishna","publish_date":"2022-02-22T12:00:00","publication_year":"2022","word_count":730,"keywords":["Go","artificial intelligence","AI","ML","Machine Learning","Git","RAG","Healthcare Automation","Ray","Aim","analytics","healthcare India","Startups","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","ML","analytics","Aim","Ray","RAG","R","Go","Git"],"url":"https:\/\/analyticsindiamag.com\/it-services\/how-is-ai-transforming-radiology\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10005937,"title":"Future Is Virtual: Facebook Launches New Tools For Embodied AI","content":"In the last six months, the world has witnessed a hurried adoption of virtual alternatives to dodge the pandemic blues. Virtual meetings, conferences, digital twins have almost become a norm. With AR\/VR poised to ride the next big wave of innovations, it is essential to revisit their current state of functionality in the realms of the physical world. For this to happen, we need to innovate heavily in a field called embodied AI. AI assistants of the future must navigate effectively, look around their physical environment, listen and build memories of their 3D space. The field of embodied AI deals with the study of intelligent systems with a physical or virtual embodiment (robots and egocentric personal assistants). The idea here is that intelligence emerges in the interaction of an agent with the environment, and as a result of sensorimotor activity. To equip the AI agents to navigate smartly, the researchers at Facebook AI have released a handful of frameworks: For Super Realistic Acoustics The team has released an audio-visual platform, which the researchers can use to train AI agents in 3D environments with highly realistic acoustics. The team believes that this platform can be leveraged to launch more embodied AI tasks for navigating to a sound-emitting target, to learn from echolocation, or even exploring with multimodal sensors. Adding sound, wrote the researchers, can yield faster training and more accurate navigation at inference. This will also enable the agent to discover the goal on its own from afar. With the help from Facebook Reality Labs, they have released SoundSpaces, audio rendering for the 3D environments. Source: Paper by Changan Chen et al., Navigate Through Your Hall The team at FAIR developed Semantic MapNet, an end-to-end learnable framework for building top-down semantic maps. This framework can be used to show where the objects are located from egocentric observations. Semantic MapNets can enable agents to learn and reason about how to navigate. For instance, in a future where virtual real estate agents are a thing, buyers might need a tour that mimics reality. A smart home assistant should be able to answer questions like how many rooms or how many chairs can fit into a hall and other trivia. For Embodied AI Simulations Habitat-Lab, is a product of AI Habitat simulation environment for embodied AI research. Habitat lab is a modular high-level library for end-to-end development in embodied AI — defining embodied AI tasks like navigation, question and answering etc. This library helps in configuring embodied agents (physical form, sensors, capabilities), training the agents (via imitation or reinforcement learning), and benchmarking their performance on the defined tasks using standard metrics. Future Direction For a long time now, Facebook AI has invested heavily in building intelligent AI systems that can think, plan, and reason about the real world. By combining embodied AI systems with their in house state of the art 3D deep learning tools, the team aims to further improve understanding of objects and places. The contributions of the team to usher next-gen embodied AI systems can be summarised as follows: Developed a new algorithm for the room navigation task in Habitat. New systems can build maps in a top-down style, like humans; think of agents that need to navigate to a room first to fetch something. Improved the way virtual robots follow instructions in simulation, by utilising disembodied large-scale image captioning data sets. Created a new benchmark for faster mapping in unfamiliar environments. Introduced a reinforcement learning method so that an embodied agent autonomously discovers the affordance landscape of a new unmapped 3D environment. Know more embodied AI advancements here.","excerpt":"In the last six months, the world has witnessed a hurried adoption of virtual alternatives to dodge the pandemic blues. Virtual meetings, conferences, digital twins have almost become a norm. With AR\/VR poised to ride the next big wave of innovations, it is essential to revisit their current state of functionality in the realms of […]","categories":["Deep Tech"],"tags":["Embodied AI","Facebook AI","future of deep reinforcement learning","virtual assistant","virtual reality"],"author_name":"Ram Sagar","publish_date":"2020-09-02T11:00:43","publication_year":"2020","word_count":598,"keywords":["Go","future of deep reinforcement learning","AI assistants","Facebook AI","AI","virtual reality","Embodied AI","Git","RAG","Aim","deep learning","ViT","virtual assistant","GAN","R"],"extracted_tech_keywords":["AI","deep learning","Aim","RAG","AI assistants","R","Go","Git","GAN","ViT"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/future-is-virtual-facebook-launches-new-tools-for-embodied-ai\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10005893,"title":"Why Does Image Data Augmentation Work As A Regularizer in Deep Learning?","content":"The problem with deep learning models is they need lots of data to train a model. There are two major problems while training deep learning models is overfitting and underfitting of the model. Those problems are solved by data augmentation is a regularization technique that makes slight modifications to the images and used to generate data. In this article, we will demonstrate why data augmentation is known as a regularization technique. How to apply data augmentation to our model and whether it is used as a preprocessing technique or post-processing techniques…? All these questions are answered in the below demonstration. Topics that we will demonstrate in this article:- Data augmentation as a regularizer and data generator.Implementing Data augmentation techniques. Data Augmentation As a Regularizer and Data Generator The regularization is a technique used to reduce the overfitting in the model. unnecessarily. In dealing with deep learning models, too much learning is also bad for the model to make a prediction with unseen data. If we get good results in training data and poor results in unseen data (test data, validation data) then it is framed as an overfitting problem. So now using data augmentation, we perform few transformations to the data like flipping, cropping, adding noise to the data, etc. As you know, deep learning models are data hungry, if we are lacking data then by using data augmentation transformations of the image we can generate data. Data augmentation is a preprocessing technique because we only work on the data to train our model. In this technique, we generate new instances of images by cropping, flipping, zooming, shearing an original image. So, whenever the training lacks the image dataset, using augmentation, we can create thousands of images to train the model perfectly. Implementing Data augmentation techniques using OpenCV Flipping the image In the below code snippet we are going to flip the image. import cv2 from google.colab.patches import cv2_imshow img = cv2.imread('\/content\/images (10).jpg') print(type(img)) print(img.shape) img_flip_ud = cv2.flip(img, 0) cv2_imshow(img_flip_ud) img_flip_lr = cv2.flip(img, 1) cv2_imshow(img_flip_lr) img_flip_ud_lr = cv2.flip(img, -1) cv2_imshow(img_flip_ud_lr) As you can see in the above output we flipped the image from 1 image we generated 4 images. Resize an image In the below code snippet we are going to rotate the image. image=cv2.imread('\/content\/images (10).jpg') cv2_imshow(image) print(image.shape) h1=300 w1=300 dimension = (w1, h1) resized_image = cv2.resize(image, dimension, interpolation = cv2.INTER_AREA) cv2_imshow(resized_image) Cropping an Image In the below code snippet we are going to crop the image. image=cv2.imread('\/content\/images (10).jpg') cv2_imshow(image) cropped_image = image[50:170, 150:250] cv2_imshow(cropped_image) Implementing Data augmentation techniques using Keras In the below code snippet we are performing data augmentation techniques using Keras. from keras.preprocessing.image import ImageDataGenerator from keras.preprocessing.image import array_to_img from keras.preprocessing.image import img_to_array from keras.preprocessing.image import load_img generated_data = ImageDataGenerator( rotation_range = 40, shear_range = 0.2, zoom_range = 0.2, horizontal_flip = True, brightness_range = (0.5, 1.5)) img = load_img('\/content\/images (10).jpg') x = img_to_array(img) x = x.reshape((1, ) + x.shape) i = 0 for batch in datagen.flow(x, batch_size = 1, save_to_dir ='preview', save_prefix ='image', save_format ='jpeg'): i += 1 if i > 5: Break Original image Image after zooming Image after apply rotation Image after adding brightness Conclusion In the above demonstration, we demonstrated how data augmentation is used as a regularizer and to generate data. This is one of the best practices used by many programmers.","excerpt":"In this article, we will demonstrate why data augmentation is known as a regularization technique. How to apply data augmentation to our model and whether it is used as a preprocessing technique or post-processing techniques…? All these questions are answered in this demonstration.","categories":["Deep Tech"],"tags":["Computer Vision","data augmentation","Deep Learning","image data augmentation","image processing","overfitting"],"author_name":"Prudhvi varma","publish_date":"2020-09-02T16:00:00","publication_year":"2020","word_count":549,"keywords":["Go","data augmentation","TPU","Keras","AI","image processing","R","image data augmentation","OpenCV","Colab","Ray","deep learning","Computer Vision","Deep Learning","overfitting"],"extracted_tech_keywords":["AI","deep learning","Ray","Keras","Colab","OpenCV","TPU","R","Go","data augmentation"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/why-does-image-data-augmentation-work-as-a-regularizer-in-deep-learning\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":25693,"title":"ISRO set to increase the frequency of satellite launches to 12","content":"Indian space agency ISRO has plans to increase the frequency of launches to 12 per year. According to recent news reports, ISRO is making attempts to increase the capacity to deliver by scaling up the frequency of launches to 12 per year by building satellites and reducing the cost of launch. At present, ISRO launches seven satellites per year. News reports suggest that the space agency is in the process of building a second vehicle assembly to enhance the turnaround time and throughput for the PSLV so that with the same launch pad ISRO can do more launches. Plans are also underway to undertake the Chandrayaan-2 mission in the first quarter of 2018, reports suggest. Kiran Kumar, Chairman, ISRO revealed that the approval to build a space station was pending. “Earlier, we were doing 2-3  launches every year, gradually we increased this frequency to 4-5 and last few years we did seven launches,” Kumar shared. This year, the target has been set for 8-9 PSLV per year, two GSLV-Mk II and one GSLV-Mk III, which means a total of 12 launches. Chandrayaan-2 mission is a completely Indian mission, with no Russian participation.  The components required for this mission, such as the development of variable thrust engines, lander and rover is already underway. Earlier, ISRO made news for the proposed Venus mission, described as Earth’s twin sister. Kumar divulged that besides Venus, Mars and Asteroids mission were also being discussed by study teams and will be finalized in due time. In the meantime, ISRO has also identified steps in terms of air-breathing propulsion system. Kumar shared, “What we shown is combustion with oxygen and hydrogen. The next step is the thrust that is generated should be more than the friction that’s going to come up because of the surfaces involved,” Kiran Kumar explained.","excerpt":"Indian space agency ISRO has plans to increase the frequency of launches to 12 per year. According to recent news reports, ISRO is making attempts to increase the capacity to deliver by scaling up the frequency of launches to 12 per year by building satellites and reducing the cost of launch. At present, ISRO launches […]","categories":["AI News"],"tags":[],"author_name":"Richa Bhatia","publish_date":"2017-05-08T04:08:52","publication_year":"2017","word_count":302,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","Ray","Rust","R","programming_languages:Rust"],"extracted_tech_keywords":["AI","Ray","R","Go","Rust","programming_languages:R","programming_languages:Go","programming_languages:Rust"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/isro-set-increase-frequency-satellite-launches-12\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":64768,"title":"World Password Day: Can We Leverage Technology To Protect Passwords From Hackers?","content":"We have seen a rise of cyberattacks in the recent past. One of the juiciest targets for malicious hackers is the password, a string of letters, numbers and characters used to authenticate online users. On the occasion of World Password Day, we take a look at why companies and developers should leverage the best techniques to protect passwords. As more people move to virtual online usage, protecting passwords becomes more important. According to research, there is a psychological challenge among people when it comes to managing passwords which can be easily exploited by hackers. We already know that passwords also should be strong enough. They should be at least 15 characters long with both uppercase letters, lowercase letters numbers and special characters. According to Adam Palmer, Chief Cybersecurity Strategist at Tenable, “Every time a researcher with time on their hands searches through the stolen password databases, it reveals millions are still using 123456 as a password, so the chances of changing password behaviour are nothing short of a miracle.” Weak passwords serve as a huge security threat for millions of businesses. But, passwords are anyway the most used authentication tool in today’s era even though other authentication techniques have been in place for years such as two-factor authentication, biometrics and hardware tokens. With each new online account, people have to remember new passwords and so it’s better to use a password manager that has hashing algorithms. Password protection should be designed in a way that they are free from vulnerabilities and sophisticated attacks such as man-in-the-middle. Hackers Will Do Everything To Crack Passwords There are multiple automatic attack schemes which hackers can leverage to exploit enterprise systems. “The sheer volume of stolen users’ passwords available for sale on the dark web highlights that the problem is less about having strong passwords or phrases, and more about users creating unique codes for each online account to limit the damage from database breaches,” Palmer said. Storing passwords in an unencrypted format is like a sin, and companies doing that are waiting to just get hacked. Developers need to create cryptographically protected systems so that hackers do not gain access to passwords. Also, there needs to be a limited number of access attempts to login attempts on any system.  This prevents Brute Force attacks to happen. Also, merely relying on encryption is not enough as hackers can even crack through encryption. In a case when a malicious entity gains access to the encryption key, encryption would serve quite useless. Advanced Techniques Are Needed For Safeguarding Passwords Developers need to, therefore, rely upon advanced techniques, like hash functions, salt to make sure that hackers are prevented from gaining access to passwords. Most of the modern-day passwords rely on matching the computed hash with the stored hash to gain access to web services. Every unique password is represented by the same length hashes, and hackers cannot access them or decode passwords easily unless through means like rainbow tables. Hashing can be made stronger by using additional data known as salt. Salt is also called a nonce, which is a number used once. And it generates a random string of bytes that can be included in the hash calculation along with the actual password. It also prevents users with the same password getting the same hash. “Given the reliance on passwords doesn’t appear to be reducing, and if anything, our virtual identities are increasing, password managers that create and store complex passwords are essential. This year, as a spotlight is once again on passwords, instead of advocating complex recipes and codes, do yourself a favour and automate,” added Palmer.","excerpt":"We have seen a rise of cyberattacks in the recent past. One of the juiciest targets for malicious hackers is the password, a string of letters, numbers and characters used to authenticate online users. On the occasion of World Password Day, we take a look at why companies and developers should leverage the best techniques […]","categories":["AI Trends"],"tags":["hackers","passwords"],"author_name":"Vishal Chawla","publish_date":"2020-05-07T19:00:00","publication_year":"2020","word_count":603,"keywords":["Go","programming_languages:R","AI","llm_models:PaLM","programming_languages:Go","RAG","hackers","passwords","R"],"extracted_tech_keywords":["AI","RAG","R","Go","llm_models:PaLM","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/world-password-day-can-we-leverage-technology-to-protect-passwords-from-hackers\/","complexity_score":3,"technical_depth":7,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10131003,"title":"AI Startup Ema Raises $36M, Launches Persona Builder for Custom AI Agents","content":"Ema, an enterprise AI company, has successfully raised an additional $36 million in Series A funding, bringing its total capital raised to over $61 million. The latest funding round was led by Accel Partners and Section 32, with contributions from Prosus Ventures, Hitachi Ventures, Sozo Ventures, Wipro Ventures, SCB10X, Frontier Capital, and Colle Capital. The company has also joined Microsoft’s Pegasus Partnership Program. Since emerging from stealth in March, Ema has expanded its customer base threefold and introduced several key innovations. The company’s Persona Builder Platform allows businesses to create and deploy custom AI “Personas” tailored for specific roles without extensive model training. This new tool is expected to enhance how companies integrate AI into various enterprise functions including customer support, sales, and compliance. “I’m proud to announce the launch of our Persona Builder Platform. This cutting-edge platform allows businesses to create customised AI employees (Personas) tailored to their specific needs, bringing unprecedented flexibility and control to our clients–democratizing Agentic AI,” said Ema chief Surojit Chatterjee. “With over 200 pre-built connectors, Ema’s Personas seamlessly integrate with internal data sources, create knowledge graphs, learn from human feedback, and operate tools to perform effectively in various enterprise roles. This innovation eliminates the need for extensive model training or manual fine-tuning once a Persona is built,” he added. Ema’s Personas, which are AI agents designed to perform complex workflows, are supported by its proprietary Generative Workflow Engine™ and the EmaFusion™ model. The platform now supports on-premises deployment for Azure and Google Cloud Platform, addressing demands for data sovereignty and regulatory compliance. In addition, Ema has achieved compliance with NIST RMF, NIST CSF, and NIST 800-171 standards, adding to its existing SOC2, HIPAA, GDPR, and ISO27001 certifications. These security measures include data encryption, audit logging, and frequent penetration testing to ensure data privacy and security. Looking ahead, Ema aims to develop a conversational operating system to replace traditional enterprise software interfaces, enabling more dynamic and intuitive management of operations. The company’s vision involves creating a unified platform that adapts in real-time to user needs and integrates seamlessly with existing systems.","excerpt":"Ema’s Personas, which are AI agents designed to perform complex workflows, are supported by its proprietary Generative Workflow Engine™ and the EmaFusion™ model.","categories":["AI News"],"tags":["AI Startups"],"author_name":"Siddharth Jindal","publish_date":"2024-07-31T23:24:07","publication_year":"2024","word_count":346,"keywords":["knowledge graphs","Go","agentic AI","API","AI","R","ML","innovation","Aim","Azure","AI Startups"],"extracted_tech_keywords":["AI","ML","agentic AI","Aim","knowledge graphs","Azure","R","Go","API","innovation"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/ai-startup-ema-raises-36m-launches-persona-builder-for-custom-ai-agents\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10043856,"title":"8 Free Resources To Learn PyTorch In 2021","content":"Developed by Facebook AI Research (FAIR), PyTorch is one of the most widely used open-source machine learning libraries for deep learning applications. It was first introduced in 2016. Since then, PyTorch has been gaining popularity among researchers and developers, at the expense of TensorFlow. At the NeurIPS conference in 2019, PyTorch appeared in 166 papers, whereas TensorFlow appeared in 74 papers. This year, NVIDIA GTC 2021 hosted over 50 different sessions related to PyTorch. Plus, Facebook has plans to migrate all its AI systems to PyTorch. In this article, we will list free resources and tools to help you learn deep learning with PyTorch: Courses & modules PyTorch Fundamentals By Microsoft Microsoft recently launched a free course called ‘PyTorch Fundamentals.’ This beginner-friendly module introduces you to key concepts to build machine learning models in multiple domains, including speech, vision, and natural language processing (NLP). Course criteria include: Basic Python knowledgeBasic understanding of how to use Jupyter Notebooks Machine learning basics PyTorch Tutorials By PyTorch This tutorial helps you learn PyTorch from scratch. In this quick start guide, you will learn how to load data, build deep neural networks, train and save your models. Plus, it includes PyTorch Recipes, which consists of bite-size, ready-to-deploy PyTorch code examples, and PyTorch Cheat Sheet. You can access PyTorch tutorials on GitHub and run tutorials on Google Colab. Intro to Deep Learning with PyTorch Curated by Facebook AI in partnership with Udacity, this free course lets you learn the basics of deep learning and build your own deep neural networks using PyTorch. It also helps you get practical experience with PyTorch through coding exercises and projects implementing SOTA AI applications such as style transfer and text generation. Check out the free course here. PyTorch Tutorials By Morvan Zhou In this tutorial, you will learn PyTorch basics (Torch and NumPy), how to build first neural network (regression, classification, optimisers, etc.), advanced neural network architectures (CNN, RNN-classification, RNN-regression, AutoEncoder, etc.) and different concepts (Train on GPU, Dropout, Batch Normalization, etc.). It is available on GitHub. Videos Neural Network Programming – Deep Learning with PyTorch You will learn everything about neural network programming and PyTorch in detail in this course. Curated by DEEPLIZARD, it covers the basics of PyTorch and CUDA, alongside understanding why neural networks use GPUs. The series touches upon the tensor and deep learning fundamentals, training a network, analysing results, tuning hyperparameters, and using TensorBorad with PyTorch for visual analytics. Start your PyTorch journey here. Check out their YouTube videos for more content around PyTorch here. Deep Learning and Neural Networks With Python and PyTorch In this tutorial series, machine learning expert Harrison Kinsley aka Sentdex, has curated in-depth video content around PyTorch, starting from basics to advanced, and building your own deep learning and neural network, training model, model analysis and more. Check out the complete playlist here. Books Deep Learning with PyTorch Co-authored by Eli Stevens, Luca Antiga and Thomas Viehmann, ‘Deep Learning With PyTorch’ covers the basics and abstractions of PyTorch in great detail, and explains the underpinnings of data structures like tensors and neural networks and making sure you understand their implementation. The book also covers advanced concepts such as JIT and deployment to production. Plus, it covers applications, taking you through the steps of using neural networks to help solve a complex medical problem. Check out the PDF version here. The book is also available on Amazon. The code for the ‘Deep Learning with PyTorch’ book is available on GitHub. Dive into Deep Learning Written by Aston Zhang, Zachary C. Lipton, Mu Li, and Alexander J. Smola, ‘Dive into Deep Learning’ offers an interactive deep learning experience with code, math and discussions across multiple frameworks, including NumPy\/MXNet, PyTorch, and TensorFlow. Currently, the book has been adopted at 175 universities from 40 countries. The book will teach deep learning concepts from scratch. Check out the PDF version of the book here.","excerpt":"At the NeurIPS conference in 2019, PyTorch appeared in 166 papers, whereas TensorFlow appeared in 74 papers.","categories":["AI Trends"],"tags":["Pytorch"],"author_name":"Amit Naik","publish_date":"2021-07-19T15:00:00","publication_year":"2021","word_count":650,"keywords":["Pytorch","machine learning","AI","neural network","PyTorch","NLP","Colab","deep learning","Jupyter","analytics","TensorFlow"],"extracted_tech_keywords":["AI","machine learning","deep learning","neural network","NLP","analytics","TensorFlow","PyTorch","Jupyter","Colab"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/8-free-resources-tools-to-learn-pytorch-in-2021\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10046350,"title":"What We Know About Tesla’s AI Day So Far","content":"It’s safe to say that the second edition of Tesla’s Autonomy Day arrives with a massive build up given the invite-only nature and the secrecy surrounding Tesla’s AI Day 2021. Following the 2019 & 2020’s Autonomy Day and Battery Day, Musk took to Twitter to reveal the AI Day in July 2021, “Looking at holding Tesla AI Day in about a month or so. Will go over progress with Tesla AI software & hardware, both training & inference.” His follow up tweet revealed the day of the AI day to be 19th August 2021. The event is scheduled for 5 PM PT \/ 8 PM ET at Tesla’s headquarters in Palo Alto, California. The physical event is invite-only, geared towards developers and technicians working in AI, but the event will be live-streamed for enthusiasts to watch virtually. In April 2019, Tesla hosted an Autonomy Day that spoke about the company’s new whole self-driving computer chips that could operate at seven times better than the nearest competitor. Later, in September 2020, Tesla hosted a Battery Day, outlining Tesla’s 4680 cells that the company claimed would enable a 54 percent increase in battery range between charges and a 56 percent reduction in price per kilowatt-hour. Tesla AI Day August 19th— Elon Musk (@elonmusk) July 29, 2021 This year the company has been largely silent about details of the upcoming event. Musk tweeted, illustrating the main purpose of AI Day 2021 being, “convincing the best AI talent to join Tesla.” However, based on Musk’s previous comments, we can safely assume that the event will probably outline Tesla’s progress on AI training systems and related hardware. The invitation reveals, “An inside look at what’s next for AI at Tesla beyond our vehicle fleet.” Tesla has recently improved its full self-driving beta software, looking at driving from A to B without any human interventions. Spectators are expecting more details about its autonomous driving. Here are some other things that Tesla watchers could expect for the highly-anticipated event. Convincing the best AI talent to join Tesla is the sole goal— Elon Musk (@elonmusk) July 29, 2021 According to an invitation obtained by Electrek, the event will feature “a keynote by Elon, hardware and software demos from Tesla engineers, test rides in Model S Plaid, and more.” The event will also give “an inside look at what’s next for AI at Tesla beyond our vehicle fleet.”  Further, “Tesla AI Day will reveal some of our work spanning autonomous driving, supercomputing, chip design and beyond,” said Clive, an engineer at Tesla, in a LinkedIn post. Musk recently tweeted a bunch of  improvements in its FSD Beta’s 9.2 version, which had started reaching more customers this summer. At the event, we can expect Musk to reveal more details on the latest advancements in the areas of FSD and Level 5 (L5) autonomy. Improvements in 9.2: pic.twitter.com\/STmEBq8V9C— Elon Musk (@elonmusk) August 15, 2021 The company has also dropped hints about its supercomputer, nicknamed Dojo, the training of its neural network, and its FSD computer chips production. Let’s take a look at these in brief. FSD Rollout Tesla’s first Autonomy Day in 2019 announced the launch of the company’s first computer chip. The chip that will care for everything an autonomous vehicle requires consists of two computers on the board that will receive power supply and video independently. This is to keep the car running in the event of a malfunction. Last year, Tesla introduced ‘Full self-self driving (FSD) beta’ to vehicle owners as part of its early access programme. The latest update enabled drivers to partially access the automated driver-assist system on streets and local roads. Musk then said the company had handled the software update ‘very cautiously.’ The purpose of offering an early access programme was to iron out software bugs and use it as a testing platform. Eight hundred twenty-four vehicles were part of the pilot programme as of 9th March, as per a DMV memo. With these developments in the backdrop, Tesla will likely release FSD beta rollouts and FSD subscription plans for its customers at ‘Tesla AI Day.’ Dojo Gave a talk at CVPR over the weekend on our recent work at Tesla Autopilot to estimate very accurate depth, velocity, acceleration with neural nets from vision. Necessary ingredients include: 1M car fleet data engine, strong AI team and a Supercomputer https:\/\/t.co\/osmEEgkgtL pic.twitter.com\/A3F4i948pD— Andrej Karpathy (@karpathy) June 21, 2021 Andrej Karpathy, Senior Director of AI at Tesla, introduced Tesla’s supercomputer at the Computer Vision and Pattern Recognition Conference 2021. He claimed that this model is the fifth most powerful supercomputer in terms of floating-point operations per second (FLOPS), but it hasn’t yet run the required benchmarks for official designation in the “TOP500” rankings. The new supercomputer will be used to train Tesla’s Autopilot and its Full Self Driving (FSD) AI with features like auto labelling. This feature takes input captures from the car’s camera, recognises objects and labels them with attributions. According to Kaparthy’s presentation, the system has so far labelled six billion things from one million captured videos. Karpathy revealed that Dojo had been designed to ingest video data and perform massive levels of unsupervised training on visual data, with an expected capacity of one exaFLOP (one quintillion or 1018 floating-point operations per second) or 1,000 petaFLOPs. It will serve as the central system for Tesla to train its self-driving AI. While Dojo has long been teased, its specs and capabilities have never really been revealed. AI Day would be a great venue to formally introduce the world to its supercomputer and its capabilities. Rumoured robots #Tesla #AI dayAugust 19, 2021 Palo Alto, CA5 p.m. PDT pic.twitter.com\/4zsP9cVxh5— Dennis Hong (@DennisHongRobot) August 3, 2021 Dennis Hong’s tweet with the picture of (supposedly) Dojo’s hardware has got Tesla enthusiasts speculating whether he is working with Tesla. Hong specialises in humanoid robots and is the founder of a Robotics and Mechanisms Laboratory at UCLA. Musk has been forthcoming about his desires for Tesla to become more than just a car company. “I think long term, people will think of Tesla as much as an AI robotics company as we are a car company or an energy company. I think we are developing one of the strongest hardware and software AI teams in the world,” Musk had also commented during a conference call discussing Tesla’s Q1 2021 financial results. Potentially, Musk can reveal Tesla’s plans to diversify out of the autonomous car sector, and into other robotics applications. That said, there are a couple of roadblocks that the shareholders would expect Musk to address. Tesla is also undergoing scrutiny from the US National Highway Traffic Safety Administration right now. The administration has opened an investigation into the company’s Autopilot semi-autonomous driving system. While the feature claims to handle select driving areas with human oversight, the NHTSA has found a series of crashes that have left at least 17 people injured and one dead. Following the investigation, Tesla shares plummeted by 4.32% this week; a 7% drop over Monday and Tuesday. This has cost the company more than $43 billion in market value in two days. Tesla’s AI showcases show that Musk has a history of giving overly optimistic timelines, like the robotaxis, that ultimately aren’t really met. We have to wait to see what Musk has in store and if he will answer the tough questions.","excerpt":"The company has also dropped hints about its supercomputer, nicknamed Dojo, the training of its neural network, and its FSD computer chips production.","categories":["AI Features"],"tags":[],"author_name":"Avi Gopani","publish_date":"2021-08-19T18:47:18","publication_year":"2021","word_count":1226,"keywords":["Go","programming_languages:R","AI","neural network","RPA","programming_languages:Go","computer vision","Aim","ViT","R"],"extracted_tech_keywords":["AI","neural network","computer vision","Aim","R","Go","ViT","RPA","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/what-we-know-about-teslas-ai-day-so-far\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10162094,"title":"OpenAI Unveils Operator, a ChatGPT Moment for AI Agents","content":"OpenAI today introduced Operator, a new AI agent that can perform tasks on the web independently. Users simply give it instructions, and it completes the task without human intervention. “AI agents are AI systems that can do work for you independently. You give them a task, and they go off and do it,” said OpenAI chief Sam Altman. Simply put, the Operator can navigate websites, fill out forms, and make purchases—all by interacting with the web like a human. Unlike traditional automation tools that rely on APIs, Operator processes information visually, moving a virtual mouse and typing into a browser. “Before, if you wanted your model to buy stuff from Instacart, you’d need to figure out if Instacart had an API… Now, this is just using screenshots, no API, nothing,” said OpenAI’s Yash Kumar during the demo. Initially available for Pro users ($200 monthly ChatGPT Pro plan) in the US, Operator will expand to other regions, though European availability will take longer due to regulatory challenges. Altman, however, said that the company would make the tech “better, cheaper, and more widely available soon.” Also, Operator will be released in OpenAI’s API “in the next few weeks.” “2025 is the year of agents,” said OpenAI’s Greg Brockman. “Operator — research preview of an agent that can use its own browser to perform tasks for you.” The Challenges: Not Perfect Yet While Operator is impressive, it’s not flawless. During the live demo, it made mistakes, such as selecting the wrong location for a restaurant booking. OpenAI admits that errors—sometimes embarrassing ones—are part of the early research phase. “Operator is an early research preview. It will do a lot of cool things. It also makes mistakes, sometimes embarrassing ones,” Kumar noted. Safety is another concern. AI navigating the web independently could fall for scams, make incorrect purchases, or misinterpret user intent. To address this, Operator includes safeguards such as human confirmations and fraud detection. “What if the website is misaligned? Maybe it’s fraudulent or asks Operator to wire money… We’ve developed our model to avoid those instructions, but we also have a separate layer—like an antivirus—that monitors suspicious activity,” explained OpenAI’s Reiichiro Nakano. How is it different from Anthropic’s Computer Use? OpenAI isn’t the only company working on AI agents. Anthropic recently launched ‘Computer Use,’ a feature in Claude 3.5 Sonnet, which allows AI to navigate computers like humans—using a cursor, clicking buttons, and typing text. Both ‘Operator’ and ‘Computer Use’ share a similar goal: enabling AI to interact with digital systems as a human would. However, the key difference is accessibility. While ‘Computer Use’ is primarily available through API integrations for developers, Operator is directly accessible to consumers through ChatGPT. “We really want to put it in people’s hands,” OpenAI emphasised. Performance also varies. In OSWorld, a test that evaluates AI’s ability to use computers, OpenAI’s COUA model scored 38.1%, while Claude 3.5 Sonnet scored 14.9%. This suggests Operator may be more reliable for real-world tasks, though both systems are still in development. Meanwhile, Perplexity, known for its AI-powered search engine, has taken a different approach with Perplexity Assistant. Unlike Operator, which focuses on web navigation, Perplexity Assistant is designed for mobile devices. As Perplexity describes it, “Perplexity Assistant uses reasoning, search, and apps to help with daily tasks.” A key advantage of Perplexity Assistant is its deep integration with smartphone workflows. It can search the web, book appointments, and even use a phone’s camera to identify objects. Unlike Operator, it maintains context across tasks, allowing users to research restaurants and book reservations in one seamless flow. However, Perplexity has struggled with reliability in past features. As one report noted, “Perplexity launched half-baked products in the past. For instance, our testing found that Perplexity’s shopping feature… tended to be slow and error-prone.” The company acknowledges these issues, with Srinivas from Perplexity stating, “Some Perplexity Assistant actions [might] not always work.” While Perplexity Assistant competes more with Google Assistant and Siri, Operator is positioned as a tool that could disrupt traditional web-based tasks. What about Microsoft’s Copilot Vision? Microsoft has also entered the AI assistant space with Copilot Vision, but its approach is distinct from Operator and Perplexity Assistant. Instead of performing tasks, Copilot Vision enhances browsing by reading pages, summarising content, and offering insights in real-time. As Microsoft describes it, “Copilot can now understand the full context of what you’re doing online. When you choose to enable Copilot Vision, it sees the page you’re on, it reads along with you, and you can talk through the problem you’re facing together.” Unlike Operator, Copilot Vision does not take independent actions like making bookings or purchases. It simply provides guidance while browsing, much like an intelligent companion. Privacy is also a key focus—Copilot Vision is opt-in, and all browsing data is deleted once a session ends.","excerpt":"Initially available for Pro users ($200 monthly ChatGPT Pro plan) in the US, Operator will expand to other regions soon.","categories":["AI Features"],"tags":["agentic ai","AI Agents","OpenAI","openai Operator","OpenAI Tasks","Sam Altman"],"author_name":"Aditi Suresh","publish_date":"2025-01-24T01:08:59","publication_year":"2025","word_count":802,"keywords":["Anthropic","ChatGPT","Go","Sam Altman","OpenAI","AI","R","ML","AI Agents","Git","agentic ai","Claude 3.5","OpenAI Tasks","fraud detection","openai Operator"],"extracted_tech_keywords":["AI","ML","ChatGPT","OpenAI","Claude 3.5","Anthropic","fraud detection","R","Go","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/openai-unveils-operator-a-chatgpt-moment-for-ai-agents\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10168685,"title":"OpenAI Now Powers Singapore Airlines’ Customer Support","content":"Singapore Airlines announced last week that it is partnering with American AI startup OpenAI, to implement AI-enabled solutions for customer support operations. The partnership will enhance Singapore Airlines’ virtual assistant to offer personalised support for customers and provide relevant information in response to their queries. Besides, the airline staff is also said to benefit from the new and improved virtual assistant, which automates routine processes and provides assistance on several tasks. “This will assist in decision-making for complex tasks such as flight crew scheduling, while considering regulatory requirements, operational limitations, and manpower availability,” the airlines stated. Singapore Airlines also mentioned that this is the “first of its kind collaboration between OpenAI and a major carrier”. Recently, Singapore Airlines announced a collaboration with Salesforce to integrate the company’s Agentforce and other AI services into its customer case management system. The two companies will also develop more AI solutions in the country’s Salesforce AI research hub, which was established in 2019. Several airlines worldwide have been implementing AI in various operations over the past few years. At the Consumer Electronics Show (CES) 2025 in January, Delta Airlines announced ‘Delta Concierge’, an AI-enabled tool for travel assistance. The tool offers assistance with various tasks, including wayfinding within the airport and airport-to-home transfers. Last year, Air India reported working with Microsoft’s Azure AI services to improve its virtual assistant. “The airline also reengineered its AI stack for generative content and more back-end integration. It used Azure AI Content Safety to detect and mitigate potentially harmful content to help ensure a safe and responsible virtual assistant,” the report published by Microsoft stated. Moreover, the airline reported that its employees are utilising Microsoft Copilot for various day-to-day activities and tasks. “The Copilot plugin is used by key sectors, including airports, for managing ground activities, cargo, engineering, commercial, and the digital and technology team, which is pivotal in maintaining Air India’s critical tech infrastructure,” the airline further said.","excerpt":"The airlines mentioned that this is the “first-of-its-kind collaboration between OpenAI and a major carrier”.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","OpenAI"],"author_name":"Supreeth Koundinya","publish_date":"2025-04-28T14:19:29","publication_year":"2025","word_count":321,"keywords":["Go","ELT","OpenAI","AI","Azure","Git","ViT","AI research","R","AI (Artificial Intelligence)","startup"],"extracted_tech_keywords":["AI","OpenAI","Azure","R","Go","Git","ELT","ViT","startup","AI research"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/openai-now-powers-singapore-airlines-customer-support\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10114841,"title":"Bhavish Aggarwal&#8217;s AI Chatbot is High on Hallucinations But There is a Silver Lining","content":"Indeed, India needs AI that understands the nuances and intricacies of the diverse languages the country speaks; however, Bhavish Aggarwal’s Krutrim may not be the answer – at least for now. Last year, Aggarwal announced Krutrim as India’s own AI model that can understand 22 Indic languages and generate output in 10 of them. He even claimed that his model outperforms GPT-4 and LLama-2, probably the most advanced and popular LLMs out there, when it comes to Indic languages. Yet, when the model was actually released last week, it turned out to be a disappointment. The chatbot not only miserably failed to pick the nuances of Indic languages, but in many cases could not even understand the context in languages like Hindi and Marathi. Aggarwal asserted that Krutrim was trained on 20 times more Indic tokens than any other model. If that were the case, Krutrim should have outperformed other chatbots like Gemini or ChatGPT. But when we prompted the bot in Assamese, it confused Assamese with Bengali, another Indo-Aryan language, given the languages share the script and have substantial overlaps in vocabulary. In another instance, the bot was found confused between Marathi and Hindi. High on Hallucinations We witnessed the same confusion in GPT3.5 as well, which powers OpenAI’s free ChatGPT version, as well as QX Lab’s recently launched Ask Qx. The confusion arises due to the limited datasets of these low-resource languages on which these models, including Krutrim, have been trained. The company acknowledged this, stating, “Krutrim’s training data is limited, potentially resulting in occasional inaccuracies or biases.” Aggarwal too, claimed that hallucinations will be there but much lower for Indian contexts than other global platforms. However, that has not been the case. Hallucinations appear to be significantly higher even when prompted in the English language. In a specific case, it asserted it was created by OpenAI. In our course of testing the bot, we discovered that it has been programmed not to discuss OpenAI, its models or other LLMs like Llama. Since Krutrim has not disclosed extensive details about the model, including its architecture or the dataset used for training, the errors it made have led many to speculate if it is merely a wrapper of OpenAI’s GPT models. Previously, AIM had reached out to Ola seeking more information on Krutrim but they declined to comment. Krutrim’s embarrassment escalated further as an increasing number of users tested the bot and shared its goof-ups on social media. Many raised questions about how Krutrim managed to secure $50 million to become India’s first AI unicorn. Saving Grace The only saving grace for Krutrim so far is that it is still in beta version. Even though its limitations are being extensively highlighted, Krurtim has the opportunity to fix the problems, lower the hallucination in the coming days and deliver a better model to developers and enterprises with its API. For Krutrim to excel in low-resource languages, Ola will have to invest significantly in building datasets for these languages. Currently, there is a lack of sufficient high-quality data available to train a model of ChatGPTs scale on Indic languages. Moreover, Krutrim will launch a much bigger multimodal model in the second quarter of 2024. We are hopeful that this model, called Krutrim Pro, will perform better and the version released last week. For Krutrim to have use cases, and enterprise adoption, which might be the target for Aggarwal, he needs to deliver a much better model, which enterprises can trust. Currently, Krutrim can be a fun chatbot for Indian consumers to play around with and hopefully will get better with time, given it uses reinforcement learning in the human feedback loop, just like ChatGPT.","excerpt":"The only saving grace for Krutrim so far is that it is still in beta version.","categories":["AI Features"],"tags":["Bhavesh Aggarwal","Krutrim","Ola Krutrim"],"author_name":"Pritam Bordoloi","publish_date":"2024-03-01T14:00:00","publication_year":"2024","word_count":613,"keywords":["Go","ChatGPT","Krutrim","TPU","OpenAI","AI","chatbots","Scala","Aim","Ola Krutrim","Rust","Bhavesh Aggarwal","R"],"extracted_tech_keywords":["AI","ChatGPT","OpenAI","Aim","chatbots","TPU","R","Go","Rust","Scala"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/bhavish-aggarwals-ai-chatbot-high-hallucinations-there-silver-lining\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10097432,"title":"Birlasoft Collaborates with Microsoft to Establish Generative AI Centre of Excellence","content":"Birlasoft Limited,  part of the C.K. Birla Group has established a Generative AI Centre of Excellence, in collaboration with Microsoft. The Generative AI Centre of Excellence will serve as a hub for Birlasoft and Microsoft experts to facilitate research, training, and collaboration. Through this collective ecosystem, organisations will be empowered to unlock the full potential of generative AI, enabling them to develop tailored solutions to address complex business challenges. Birlasoft will train 500 consultants on generative AI technologies, principles, and best practices. Additionally, they will work together on building over 50 use cases, focusing on value creation across the various verticals and sub-verticals served by Birlasoft. The continual advancements in Generative AI are opening up endless possibilities for various industries. Birlasoft will focus on key use cases across different verticals. For Manufacturing sector, the company will use Azure OpenAI Service capabilities for product design, process optimisation, quality and defect detection, as well as predictive maintenance and digital twins.In the Life Sciences and Pharmaceutical domain, Birlasoft will utilise generative AI solutions for drug discovery, design, and data augmentation. For energy and utilities sector, the company aims to enhance field service engagements leveraging Generative AI. In the Banking, Financial Services, and Insurance domain, focus areas for generative AI include automated claims handling, summarising financial reporting, and improving search capabilities. “We are pleased to see Birlasoft embrace this new technology and establish a Generative AI Center of Excellence. By using Microsoft Azure OpenAI Service’s models and Azure’s scalability, our shared customers will be able to unlock new possibilities and drive greater outcomes,” Jim Lee, VP, Americas Global Partner Solutions and Sales, Microsoft said.","excerpt":"Birlasoft will train 500 consultants on generative AI technologies, principles, and best practices","categories":["AI News"],"tags":[],"author_name":"Pritam Bordoloi","publish_date":"2023-07-24T14:53:33","publication_year":"2023","word_count":270,"keywords":["OpenAI","AI","R","Scala","Git","RAG","Aim","generative AI","GAN","Azure"],"extracted_tech_keywords":["AI","generative AI","OpenAI","Aim","RAG","Azure","R","Scala","Git","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/birlasoft-collaborates-with-microsoft-to-establish-generative-ai-centre-of-excellence\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":43750,"title":"Google Looks Poised To Offer India’s Public Sector With Cloud Services","content":"via Google Cloud blog Tech giants like Google, Amazon and Microsoft have been top providers so far. While Amazon and Microsoft have topped the charts, Google trailed behind. The last quarter reports of Google Cloud tells a new tale as they have made good numbers(~$8 billion). Now they look to extend their reach in world’s one of the most happening places — India. India’s first Google Cloud Platform region was launched back in November, 2017. The new Mumbai region joined the likes of  Singapore, Taiwan, Sydney and Tokyo in Asia Pacific to build highly available, performant applications using cloud. Their aim was to improve latency from 20-90% for end users in Chennai, Hyderabad, Bangalore compared to hosting them in the closest region, Singapore. The Mumbai region has 3 zones. Developers and network admins can distribute apps and storage across multiple zones to protect against service disruptions. As enterprises break monoliths apart and start modernizing services, they need solutions for consistent service and traffic management at scale. Organizations want to invest time and resources in building applications and innovating, not on the infrastructure and networking required to deploy and manage these services. For instance, GCP’s Anthos lets the user run their application without any modifications and also manage workloads running on third-party clouds like AWS and Azure. This level of freedom enables developers to deploy applications on cloud without the need to learn new APIs. How Google Cloud Fared In India So Far Tata Steel To organize data from siloed systems, Tata Steel is using Cloud Search and plans to scale it to more than one million documents and 28 disparate enterprise content sources including enterprise resource planning (ERP) and SharePoint. In fact, Tata Steel is one of the first Indian enterprises to harness the power of Cloud Search to meet some of the most aggressive ingestion demands, with indexing durations reduced from weeks to seconds. Managed services such as AI Platform further enable Tata Steel to manage end-to-end AI\/ML workflows within the GCP console. L&T L&T Financial Services provides farm-equipment finance, micro loans and two-wheeler finance to consumers across rural India backed by a strong digital and analytics platform. Their digital-loan approval app, which runs on GCP, makes it significantly faster and easier for people to apply for financial assistance to purchase important things such as farming equipment and two-wheelers. TCS Digitate, a venture of TCS (Tata Consultancy Services), has integrated Dialogflow into its flagship brand ignio. A product of Google Cloud, Dialogflow is an end-to-end, build-once deploy-everywhere development suite for creating conversational interfaces for websites, mobile applications, popular messaging platforms, and IoT devices. Now building on this successful adoption at large organisations, Google eyes Indian public sector. India under Modi has been insisting on the need for digitising the whole nation. The increased transaction through digital payments is one such example. India’s can use state-of-the-art cloud services like that of Google’s in powering up its public sector. “We are working closely with the Ministry of Electronics and Information Technology to achieve the full Cloud Service Provider (CSP) empanelment. The idea is to help state governments have a secure experience with Cloud, along with utilising New-Age AI\/ML capabilities,” said Nitin Bawankule, Director, Google Cloud, India. India On Cloud Google raked in over $8 billion in revenue last quarter and now they want to scale up their services in India by tapping into the public sector. Back in April, CEO Thomas Kurian described Google Cloud’s new services to target different industries that includes media and entertainment, healthcare, retail, financial services, public sector. India According to NASSCOM, the cloud market in India is likely to soar to $7.1 billion by 2022 with the developmental leaps in Big Data analytics, Artificial Intelligence and Machine Learning and Internet of Things (IoT). The computational demand changes for applications and users. There needs to be flexibility between different demands and still be able to deliver results in real time. This is a crucial time to be developing new cloud technologies for solving problems. Cloud is now being used for navigating ships carrying cargo, connecting online and offline retail markets, machine learning in fintech for fraud detection and many more.","excerpt":"Tech giants like Google, Amazon and Microsoft have been top providers so far. While Amazon and Microsoft have topped the charts, Google trailed behind. The last quarter reports of Google Cloud tells a new tale as they have made good numbers(~$8 billion). Now they look to extend their reach in world’s one of the most […]","categories":["Global Tech"],"tags":["Anthos","Google"],"author_name":"Ram Sagar","publish_date":"2019-08-02T17:07:39","publication_year":"2019","word_count":693,"keywords":["machine learning","artificial intelligence","AWS","AI","ML","Anthos","RAG","Aim","analytics","Google","Azure","fraud detection"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","analytics","Aim","RAG","fraud detection","AWS","Azure"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/google-looks-poised-to-offer-indias-public-sector-with-cloud-services\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10164394,"title":"Top 10 Japanese GCCs Fueling Bengaluru’s Tech Boom","content":"India’s Global Capability Centre (GCC) ecosystem is expanding rapidly, with the number of centres projected to grow from 1,700 in 2024 to 2,100 by 2030. While US-based GCCs have traditionally led the charge, non-US GCCs are rapidly gaining momentum, growing at an impressive compound annual growth rate (CAGR) of 6.8%. This growth is nearly double that of their US-based counterparts, signalling India’s increasing prominence as a global hub for business services. Notably, Japanese companies account for around 5% of India’s GCC ecosystem, primarily focusing on automotive, electronics, and manufacturing. Traditionally, Japanese firms have used GCCs in India for operational support. Now, however, there is a growing shift towards research and development (R&D) and digital transformation. A key example is Toyota’s GCC in Bengaluru, which has begun working on advanced EV battery technologies. This reflects a broader trend of innovation-driven expansion within Japanese GCCs in India. These leading Japanese companies are fuelling India’s GCC expansion and expanding their operations, signalling a new era of collaboration between Japan and India. MUFG Bank MUFG Global Service (MGS) operates from offices in Bengaluru and Mumbai. It is a global capability centre that supports MUFG Bank offices worldwide. MUFG Bank, Japan’s largest bank and a leading global financial institution, has a presence across 40 other countries. It offers a comprehensive range of financial products and services, including strategic advisory, liquidity arrangements, and risk management solutions – all delivered with strong execution. As part of Mitsubishi UFJ Financial Group, one of the world’s largest banking groups by assets, loans, and deposits, MUFG Bank provides financial strength and stability to meet client needs. With a legacy of over 130 years in international commerce, the bank is known for its industry expertise, reliability, and cross-border capabilities, enabling businesses to navigate global markets confidently. Hitachi Energy India Limited Hitachi Energy has played a significant role in India’s key nation-building initiatives and marked 75 years of its presence in the country in 2024. The company began operations in 1949 and established its first factory in Vadodara in 1962, which has since been expanded and modernised. Today, its three manufacturing locations in Gujarat contribute significantly to its overall offerings. The company has also launched its third factory in Bengaluru. This 5,000-square-metre facility serves customers in power utilities, industries, renewable energy, and the rail segment. With a workforce of over 7,300 employees, Hitachi Energy operates 19 factories across eight manufacturing locations, along with shared service centres and a global technology and innovation hub. This reflects its strong commitment to the Make in India initiative, catering to both domestic and global markets. The company has also pioneered high voltage direct current (HVDC) technology in India. More than half of the country’s HVDC links use this advanced solution. In the transportation sector, Hitachi Energy has collaborated with Indian Railways to develop Scott Transformers, which are integral to high-speed rail operations. Over 90% of India’s metro rail systems rely on its power automation solutions. Rakuten India Rakuten India is a global product and innovation centre for the Rakuten Group, to support its businesses in e-commerce, fintech, advertising, mobile, content, and entertainment. With expertise in data science, engineering, machine learning, artificial intelligence (AI), cloud computing, security, and distributed systems, the company plays a crucial role in driving technological advancements. Its teams in India contribute to multiple Rakuten services, including Rakuten Ichiba, Rakuten Ready, and Rakuten Delivery, handling engineering, quality control, DevOps, and product development. As the second-largest technology hub outside Japan, Rakuten India builds and enables platforms for global e-commerce, payments, digital services, AI, and data science. Located in Bengaluru, often referred to as the Silicon Valley of India, the company benefits from a rich talent pool. The work environment aligns with that of Rakuten’s offices in Tokyo, San Francisco, and Singapore, ensuring seamless collaboration across all divisions. Rakuten India is an in-house innovation and technology centre, driving digital transformation across the group’s global operations. Pioneer Innovation is crucial to Pioneer’s approach, and its Centre of Excellence in India represents a major step in expanding its R&D capabilities beyond Japan for the first time. This initiative aims to use India’s talent and expertise to drive advancements in mobility products globally. With R&D centres established in Gurugram and Bengaluru, the company seeks to accelerate innovation and develop cutting-edge products and services in the mobility sector. Pioneer’s Vision AI for cars integrates AI into vehicles to enhance safety and efficiency. Utilising cameras and sensors, it enables real-time perception of the surroundings, supporting features such as advanced driver-assistance systems (ADAS), pedestrian detection, and autonomous driving capabilities. This technology aims to create a safer and more convenient driving experience. Socionext Socionext, a global enterprise headquartered in Yokohama, has expanded its design and development capabilities with the opening of an office in Bengaluru. The company specialises in designing, developing, and delivering system-on-chip (SoC) solutions to customers worldwide, further strengthening its presence in the semiconductor industry. Socionext uses its design expertise to optimise the latest technologies for large-scale SoC development in advanced applications such as autonomous driving and high-performance computing. This enhances performance, reduces power consumption, and lowers overall costs. Renault Nissan Technology and Business Centre India Last year, Renault Nissan Technology and Business Center India (RNTBCI) inaugurated a new centre in Bengaluru as part of its RNTBCI 3.0 initiative, aiming to strengthen its position as a ‘smart carmaker’. This facility drives advancements in software-defined vehicle (SDV) technologies and AI. RNTBCI is a key hub supporting Renault and Nissan globally. Established in 2007, it is located in Chennai, with additional offices in Hyderabad and Bengaluru. This new centre in Bengaluru combines engineering systems, software, information system, and IT teams to improve vehicle features and operational efficiency using AI and data. It also includes test benches, a car workshop, and high-speed fibre optic connectivity Anritsu India Anritsu India Private Limited, a subsidiary of Anritsu Corporation headquartered in Yokohama, Japan, has been operational in India since 2005. In April 2018, the company expanded its presence by inaugurating a new 10,000-square-foot head office located in Bengaluru. This Bengaluru office serves multiple functions, including sales, technical support, service, research and development, software development, and global support activities for Anritsu’s products and solutions. Anritsu India specialises in providing test and measurement equipment that supports next-generation networking, as well as quality-assurance inspection equipment for the food and pharmaceutical industries. Toyota Motor Corporation Toyota Motor Corporation has a significant presence in Bengaluru, primarily through its joint venture, Toyota Kirloskar Motor (TKM). While TKM operates manufacturing plants in Bidadi, near Bengaluru, the company has also been actively involved in enhancing local talent and engineering capabilities within the city. In 2023, TKM expanded its commitment to skill development by establishing multiple Centres of Excellence (CoE) in Bengaluru. Fujitsu Research of India Fujitsu Research of India represents Fujitsu’s research initiatives in the country, focusing on developing cutting-edge technologies to support the company’s goal of building a more sustainable world through innovation and trust. The company primarily focuses on research and development into AI, ML technologies and quantum software. The organisation is dedicated to driving large-scale societal transformation by developing trusted digital technologies that enable successful digital transformation (DX) strategies. Established in April 2022, Fujitsu Research of India operates across Bengaluru and Hyderabad. Shindengen India Private Limited Shindengen India Private Limited, a subsidiary of Shindengen Electric Manufacturing Co. Ltd., was established in August 2012. Located in Bengaluru, the facility spans 20,000 square meters. The company specialises in the manufacturing and sales of electrical equipment, including regulators, ECUs, inverters, and PCUs.According to reports, Shindengen India has committed ₹254 crore in Karnataka at the Invest Karnataka 2025 summit.","excerpt":"Japanese companies account for around 5% of India’s GCC ecosystem.","categories":["GCC"],"tags":["GCC"],"author_name":"Shalini Mondal","publish_date":"2025-02-24T09:00:00","publication_year":"2025","word_count":1263,"keywords":["data science","Go","artificial intelligence","machine learning","GCC","AI","cloud computing","ML","Transformers","Aim","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","data science","Aim","Transformers","cloud computing","R","Go"],"url":"https:\/\/analyticsindiamag.com\/gcc\/top-10-japanese-gccs-fueling-bengalurus-tech-boom\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10001863,"title":"Here’s How OPSEC Can Help Protect Your Data From Hackers","content":"As more software and hardware vulnerabilities are discovered every day, the digital world becomes more unsafe to interact with. The practice of a doctrine while handling interactions is required to ensure optimal security. Operational security (OPSEC) provides the elements for implementing cyber security- sensitive culture, a draft prepared by Idaho National Laboratory states. This practice is known as operation security, and is one of the most fundamental aspects of cybersecurity. It is a method to strive for data ownership and privacy, with a focus on preserving sensitive information. What Is Operational Security? Operational Security, more commonly known as OpSec, was a practice developed for use by the military. It then rose in popularity in the private sector, with many cybersecurity practitioners adopting it to stay safe in an evolving digital world. OpSec involves various behaviours such as monitoring where personal identifiable information is distributed, taking ownership of data shared online, and practicing general safety measures. OpSec practitioners will often look for possible areas from which threats may emerge, and practice general hygiene when interacting online. OpSec is both versatile and large in its scope, due to the sheer amount of attack vectors that are present in a place such as the Internet. There are many practices every individual can enforce while using the Internet. Identify Where Your Data Is Available Personally identifiable information, otherwise known as PII, is one of the most valuable resources a user can have. This is data or metadata that can be accurately traced back to decipher the real world identity of the individual in question. This creates a lot of issues, with the first being that privacy is completely compromised. Depending on the degree of PII given out, threats can range from exposes known as doxxing, all the way up to identity theft and financial fraud. Moreover, every post that is made in a location that is accessible to every user on the Internet, such as social media, needs to be scanned for sensitive PII. This includes data such as passwords, PIN numbers, card numbers and location data. Financial data should never be posted on the Internet, as this is one of the most sensitive forms of PII. Intellectual property, research or code should not be shared widely without requisite security measures to dissuade theft. Look For Possible Attack Vectors AS practioners of OpSec, users must maintain constant vigilance regarding possible security compromise attack vectors. This includes phishing attacks, Trojans, and malicious software such as rootkits and keyloggers. Using an antivirus is strictly necessary for Windows machines, although the built-in Defender utility does the job without bloat. However, it is always recommended to switch to a Linux or UNIX-based operating system due to the reduced likelihood of malicious software installations. These include OS such as MacOS or Ubuntu. Any and all email should ideally be screened for symptoms of suspicious behaviour, such as requests for financial information or bank transfers. One should also be cautious of changed domain names in case of spear phishing attacks. Whenever accessing a website, the domain as well as security certificates must be checked beforehand so as to not give PII to sites which are not the official ones. General Digital Hygiene And Data Protection This is, by far, the most far-reaching aspect of OpSec and the most difficult to ensure. This can also take the form of a checklist, which can be used to eliminate potential security pitfalls. Primarily, PII should not be given out unless absolutely necessary. It is also important to not trust anything, instead verifying it. Many proponents have also advocated moving away from centralized data silos such as those seen in Google and Facebook. Instead, cloud environments must be self-hosted, social media is not to be used and email servers must be either self-run or encrypted by a trusted party. This also includes to messaging applications, with end-to-end encryption being a dependable way to ensure that third parties, even those managing the network, cannot access the messages. Added to all of this, passwords and emails must be used responsibly. One password must never be reused more than once, and emails that can potentially tie into PII must not be used. Passwords must be over 32 characters in length to exponentially increase the amount of time required to crack them. A random string generator can be used for this purpose. The random passwords must never be stored as plaintext anywhere. They must instead by encrypted, either by an algorithm developed by the user or one which is not used widely. This ensures that the method to crack it is not widely available. Passwords must be updated regularly, and checked against a database such as Have I Been Pwned to ensure that they are not lost in data breaches. A VPN must be used everywhere that the Internet is accessed, as the IP address is one of the most dangerous forms of PII. Tracker blocking must be implemented, either at the browser level or the OS level. Ad blockers or obfuscators must be used in order to avoid giving information on clicks to advertisers. While this might sound like the most paranoid outlook on using the Internet, it is one of the most dependable ways to ensure that users’ data is not stolen, held ransom or misused.","excerpt":"As more software and hardware vulnerabilities are discovered every day, the digital world becomes more unsafe to interact with. The practice of a doctrine while handling interactions is required to ensure optimal security. Operational security (OPSEC) provides the elements for implementing cyber security- sensitive culture, a draft prepared by Idaho National Laboratory states. This practice […]","categories":["AI Features"],"tags":["Cybersecurity","hackers","VPN"],"author_name":"Anirudh VK","publish_date":"2019-04-19T15:28:44","publication_year":"2019","word_count":880,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","Git","hackers","Rust","Cybersecurity","R","VPN","programming_languages:Rust"],"extracted_tech_keywords":["AI","R","Go","Rust","Git","programming_languages:R","programming_languages:Go","programming_languages:Rust"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/heres-how-opsec-can-help-protect-your-data-from-hackers\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10071568,"title":"Deep Learning Laptops We’ve Reviewed","content":"As an amateur professional, there are certain key components to focus on while purchasing a laptop for performing ‘deep learning’ operations such as RAM, CPU, storage and operating system. Laptops with higher RAM would ensure faster processing while those with GPU provide an additional advantage to speed up the training process and help reduce time from model training. Another essential component for deep learning laptops is ‘graphics card’, used to render higher dimensional images. Here is a detailed list of top laptops for deep learning— TensorBook Lambda Labs recognises ‘Tensorbook’ as the “Deep Learning Laptop”. Tensorbook is equipped with GeForce RTX 3080 Max-Q 16GB GPU, VRAM-16 GB GDDR6 and is backed by Intel Core i7-11800H along with RAM of 64 GB 3200 MHz DDR4 and storage of 2 TB NVMe PCIe 4.0. (Image source: Amazon) According to Lambda Labs, Tensorbook’s GeForce RTX 3080 is capable of delivering model training performance up to 4x faster than Apple’s M1 Max and 10x faster than Google Colab instances. It is also equipped with pre-installed machine learning tools such as PyTorch, Tensorflow, CUDA, and cuDNN. (Image source: Lambda Labs) Razer Blade 15 RTX3080 Razer Blade 15 RTX3080 is an equally good choice in terms of deep learning operations. The laptop is powered by NVIDIA GeForce RTX 3080 Ti along with Intel Core i7-11800H. The Intel Turbo Boost Technology can boost the i7 processor up to 5.1GHz.Go with ultra-fast 360Hz FHD. (Image source: Amazon) Razer Blade 15 RTX3080 has a battery life of upto 5 hours. The laptop efficiently dissipates heat through the evaporation and condensation of an internal fluid and keeps it running soundlessly and coolly even under intense loads owing to features like vapour chamber cooling for maximised thermal performance. Asus Rog Strix Scar III G733 It is a powerhouse laptop with the combination of both NVIDIA and AMD. It is powered by AMD Ryzen 9 5900HX CPU and GeForce RTX 3080 GPU along with an ultrafast panel up to 300 Hz\/3ms. It has a 90 Wh battery with rapid Type-c charging with video playback upto 12 hours. (Image source: Asus) ASUS ROG Strix G17 The ASUS ROG Strix G17 laptop is equipped with RTX3070 GPU along with 8GB VRAM and 8-core Ryzen 9 which makes it one of the most suitable laptops for machine learning. It also has a 165Hz 3ms refresh rate and a 90Wh battery which allows usage upto a solid 10 hours. (Image source: Asus) Eluktronics MAX 17 The Eluktronics MAX-17 renders itself the lightest 17.3″ gaming laptop in the industry. It is powered by Intel Core i7-10870H Eight Cores-16 Threads (2.2-5.0GHz TurboBoost) along with 8GB GDDR6 VRAM NVIDIAGeForce RTX 2070 Super (Max-P–TDP:115 Watts). (Image source: Eluktronics) In terms of memory and storage configuration, the laptop is equipped with 1TB Ultra Performance PCIe NVMe SSD + 16GB DDR4 2933MHz RAM. ASUS TUF Gaming F17 ASUS TUF Gaming F17 is yet another impressive option for deep learning operations. It is powered by the latest 10th Gen Intel Core i7 CPUwith 8 cores and 16 threads to tear through serious gaming, streaming and heavy duty multitasking. It also has GeForce GTX 1650 Ti GPU with IPS-level displays up to 144Hz. (Image source: Amazon) The laptop also features a larger 48Wh battery that allows up to 12.3 hours of video playback and upto 7.3 hours of web browsing. In terms of durability, it claims to be equipped with TUF’s signature military-grade durability. Razer Blade 15 The Razer Blade 15 laptop boasts of 11th Gen Intel Core i7-11800H 8 Core (2.3GHz\/4.6GHz) and NVIDIA GeForce RTX 3060 along with 6GB DDR6 VRAM. (Image source: Amazon) This laptop comes with a built-in 65WHr rechargeable lithium-ion polymer battery that lasts upto 6 hours.","excerpt":"Laptops with GPU provide an additional advantage to speed up the training process and help reduce time from model training","categories":["AI Features"],"tags":[],"author_name":"Sreejani Bhattacharyya","publish_date":"2022-07-26T17:00:00","publication_year":"2022","word_count":617,"keywords":["CUDA","machine learning","AI","PyTorch","RAG","Colab","Aim","deep learning","TensorFlow","R"],"extracted_tech_keywords":["AI","machine learning","deep learning","Aim","TensorFlow","PyTorch","Colab","RAG","CUDA","R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/deep-learning-laptops-weve-reviewed-2022\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10038728,"title":"The Role Of Statistics In The Era Of Big Data?","content":"The concepts in statistics and mathematics are the building blocks of the techniques and tools we use to gain deeper insights into structured and unstructured data. Statistical concepts lie at the heart of data science. In this informative session at SkillUp 2021, a two-day event organised by Analytics India Magazine, Rajeeva Karandikar of Chennai Mathematical Institute, presented a few examples (from history) to explain how to make the most of the available data and enormous computing power by combining statistical ideas with modern AI\/ML tools. Rajeeva Karandikar is the Director at Chennai Mathematical Institute. He is a Fellow of the Indian Academy of Sciences and Indian National Science Academy. His research interests include probability theory and stochastic processes, applications of statistics and cryptography. Statistics quo “Perhaps in 90% of the problems that need some decision based on available data, the standard tools in artificial intelligence or machine learning and statistics will yield the best or nearly the best answer. But the remaining 10% will need something more than just the tools,” said Karandikar. He said not all data problems don’t have the benefit of big data, such as opinion polls, quality control, vaccine identification and approval, drug discovery and approval. Thus, statistical ideas and techniques are definitely relevant in such cases. Karandikar called up instances from history to prove the significant role of statistics and data. Sir Francis Galton, cousin of Darwin who studied the inheritance of genetic traits, was his first example. “It appeared from these experiments that the offspring did not tend to resemble their parent seeds in size, but always to be more mediocre than they- to be smaller than the parents, if the parents were large; to be larger than the parents if the parents were very small”- Galton Karandikar explained that Galton obtained data on the heights of parents and (grown-up) sons and got a confirmation of his ideas. He chose heights as it was easy to obtain data on them. His analysis of the data confirmed his hypothesis. “Today we can obtain data on heights of a large number of individuals and their father’s heights (say from India passport database). It can be seen that any data-driven tool will confirm the conclusion reached by Galton. However, interchanging roles of the heights of sons and fathers lead to an exactly opposite conclusion. This nature can also be seen in simulated data,” Karandikar said. Correlation and regression Next, Karandikar discussed some important topics of statistics, such as correlation and regression. Most of the data-driven analysis tries to discover relationships among different variables and this is what correlation and regression are all about. It is also important to understand that correlation does not imply causation. While correlation, as well as regression, are techniques to discover linear relationships, one needs to use transformations to get more complex relationships. Artificial intelligence, machine learning and other data-driven techniques likewise try to find the relations, linear or otherwise among the variables. Karandikar said one must not use any such tools without understanding the domain of the task. He illustrated this by providing several examples of correlation and standard error, such as prediction of 2007 Cricket ODI World Cup, relationship between IIT-JEE\/CAT scores and performance at IIT\/IIM, among others. Karandikar also talked about some important terms of statistics, such as spurious correlations or nonsense correlations, standard error, Simpson’s paradox or amalgamation paradox, omitted variable bias, GIGO, GMGO, among others. “Techniques in AI\/ML have made great advances and it is a great exploratory step. But for now, it has not matured enough to reach an inferential step. When used in conjunction with domain knowledge, we can do wonders,” said Karandikar while concluding the session.","excerpt":"Techniques in AI\/ML have made great advances and it is a great exploratory step. But for now, it has not matured enough to reach an inferential step.","categories":["AI Features"],"tags":["Big Data","Data Science","Statistics"],"author_name":"Ambika Choudhury","publish_date":"2021-04-24T18:00:00","publication_year":"2021","word_count":611,"keywords":["big data","data science","Go","artificial intelligence","machine learning","AI","Statistics","ML","analytics","GAN","Data Science","Big Data","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","data science","analytics","R","Go","big data","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/the-role-of-statistics-in-the-era-of-big-data\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10142825,"title":"OpenAI’s Sora is So Not Dead","content":"On Saturday, in a not-so-popular cryptic post on X, OpenAI chief Sam Altman hinted at Sora’s release this week. “I am so, so excited for what we have to launch on day 3. Monday feels so far away,” he said. Needless to say, OpenAI’s popular text-to-video tool, Sora, has once again become the talk of the town after the leaked footage and information hinting at the video generation quality of Sora v2 surfaced. Ruud van der Linden, CEO at Lont, uploaded the video of OpenAI’s Chad Nelson presenting Sora v2 at the C21Media Keynote in London. Sora v2 release is impending:* 1-minute video outputs* text-to-video* text+image-to-video* text+video-to-videoOpenAI's Chad Nelson showed this at the C21Media Keynote in London. And he said we will see it very very soon, as @sama has foreshadowed. pic.twitter.com\/xZiDaydoDV— Ruud van der Linden (@RuudNL) December 7, 2024 Sora v2 is expected to generate videos up to one minute in length, making it a significant leap from previous versions or competitors that often offer shorter clips. This latest development offers a flexible approach to video generation ranging from text-to-video, text+image-to-video, and text+video-to-video. This could drastically change the process of video generation and interactivity and customisation, which is not commonly seen in current models. Recently, Sora’s API, available to some artists for early testing on Hugging Face, was also leaked. Regardless of the hype that the leak created, it brought out a bigger question: Was the Sora leak real? Not long after the tool’s leak, the Hugging Face page seemed to be failing with a 502 error due to high traffic. The company learned of this incident soon enough and shut down the access three hours after the revelation. According to AIM, three hours seem to be good enough for users to produce content and create hype but not too long for the situation to get out of control. Meanwhile, Google Cloud introduced Veo, a video generation model, and Imagen 3, an advanced image generation tool, on its Vertex AI platform. Veo, currently in private preview, generates high-quality videos from text or image prompts, enabling businesses to create realistic and coherent footage efficiently while reducing production time and costs. Sora is Going to Soar High in Video Quality Ashutosh Shrivastava, an AI enthusiast, took to X to express his thoughts on this. In the past year, while Sora has been kept from public use, many other models have come up that can “create videos as good as, or even better than, Sora”. However, with the latest video leak, others have also speculated that Sora might surpass all other tools in quality benchmarks. Alex Volkov, the host of the ThursdAI pod, wrote on X: “I take back EVERYTHING I said about other video models catching up to SORA even remotely.” He also hints at the OpenAI $200\/mo pro tier, saying that if Sora is added to that, the company will see a lot of new subscriptions. If the tool is not released in the ‘12 days of OpenAI’, which many are speculating that it will, OpenAI could lose its advantage with the tool’s popularity on the rise currently. The tool was then soaring the charts, also reported by AIM in June this year, expressing that ‘OpenAI should release Sora before it’s too late’. Multiple users online have been discussing the videos generated by Sora since the API was leaked. When looked at critically earlier, the videos do not stand out enough to call this a ‘leak’ in the first place. A user on X expressed the same: “It actually was a bit underwhelming, tbh, at least if the leaked videos are legit.” “We have seen better now… So yeah,” the discussion continued. The users hoped for Sora to have moved forward to a new level, which could be making a cinema-length movie out of many snippets cut together. Another contributor expressed, “The videos it generated are quite poor compared to Kling’s.” However, quite a few were impressed with the quality of the videos generated. A user said that though “Sora is not perfect, the video is a lot more coherent than Runway.” Following the leak, many content creators and enthusiasts admitted they’d lost all hope of Sora ever being released to the public. While early testers are still finding issues that need to be resolved before public release, a user feels that there aren’t enough resources to support a large number of users. She also believes that OpenAI plans to offer the tool only to the biggest paying customers (enterprise users, Hollywood, etc). Some believe that the company wanted to “improve continuously as ideas came in from other video generation tools, and release Sora only when they managed to create something competitive”. This could take a sharp turn if OpenAI releases Sora on Monday, as many speculate. An AI enthusiast on X said, “Sora looks promising and stunning! Sora might get released in OpenAI’s 12-Day campaign.” Sora Hasn’t Died Yet As speculations around the Sora release rise, there is considerable excitement regarding its potential public release. Some indicate that it might be part of OpenAI’s upcoming initiatives or campaigns. With multiple scenes and character consistency across the video duration, there has risen significant interest and anticipation among tech enthusiasts and professionals in the field. As OpenAI engaged with Hollywood through Sora, Runway partnered with top entertainment and media like Lionsgate to develop customised versions of Gen-3 Alpha. Unlike OpenAI, Runway has also made Gen-3 Alpha available to all users, though the model remains subscription-based. With the rise and impact of other tools like Runway, Midjourney and KlingAI over the past year, it has become difficult for creators to think back to the capabilities of Sora. AIM had previously compared the launch of OpenAI’s Sora to a ‘ChatGPT moment in video generation’.","excerpt":"Sora v2 is expected to generate videos up to one minute in length and offers a flexible approach to video generation.","categories":["AI Features"],"tags":["OpenAI","Sora"],"author_name":"Sanjana Gupta","publish_date":"2024-12-09T19:47:07","publication_year":"2024","word_count":961,"keywords":["Go","ChatGPT","Hugging Face","TPU","API","OpenAI","AI","Git","Aim","Sora","R"],"extracted_tech_keywords":["AI","ChatGPT","OpenAI","Aim","Hugging Face","TPU","R","Go","Git","API"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/openais-sora-is-so-not-dead\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10073623,"title":"Rakuten opens a new R&#038;D center in Bengaluru, its Largest Facility Outside Japan","content":"E-commerce group Rakuten India on Thursday announced the opening of their new office at Crimson House, Bengaluru, to support the growing customer base in India. Being the largest office outside Japan, the product, engineering, and advanced research facility at Bengaluru would represent expansion of the company’s global technology footprint. The 20-storey modern building is strategically located in the city’s central business district, with a capacity of 3000+ employees. It will focus on delivering cross-industry deep technology solutions to help drive digital innovation for customers and partners. “India is and has been central to our growth strategy. Our new R&D centre will build on the high value-added engagements that Rakuten has been delivering from India on deep tech and product innovation. Our modern and environment-friendly facility provides a perfect environment to return to office post pandemic and is designed to foster collaboration and innovation,” said Yasufumi Hirai, CIO of Rakuten Group, Inc. The group aims to enhance the company’s capability for deep-tech innovation and R&D in e-commerce, fintech, AI in computer vision, speech, and natural language processing. Hitting the market last year, Rakuten ‘SixthSense’, is the company’s first B2B SaaS product, an all-in-one observability intelligence and software testing automation platform. Rakuten India is the innovation centre for Rakuten Group, enabling global businesses in the areas of e-commerce, content and entertainment, data science and engineering, cloud, security, among others.","excerpt":"The group aims to create an inspiring space for growth and innovation in the Asia-Pacific region","categories":["AI News"],"tags":["Expansion","R&amp;D","Rakuten India"],"author_name":"Bhuvana Kamath","publish_date":"2022-08-25T17:40:48","publication_year":"2022","word_count":227,"keywords":["data science","programming_languages:R","AI","innovation","Git","computer vision","automation","Expansion","Aim","R&amp;D","R","Rakuten India","ai_applications:computer vision"],"extracted_tech_keywords":["AI","computer vision","data science","Aim","R","Git","automation","innovation","programming_languages:R","ai_applications:computer vision"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/rakuten-opens-a-new-rd-center-in-bengaluru-its-largest-facility-outside-japan\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10053938,"title":"Complete Guide to Understanding Precision and Recall Curves","content":"Precision and Recall are two of the most important metrics to look at when evaluating an imbalanced classification model. These help us to find out what fraction of the actual positives were classified correctly and among the ones classified positive, what fraction were actually positives. Going ahead, we will try to explore the significance of these and thresholds in detail. The topics that we will discuss in this article are the following: Table of contentsWhat are Precision and Recall?What is the Need for a Precision-Recall Curve?How to read a PR CurveThe baseline of PR CurvePR Curve of a ‘No Skill’ ModelPR Curve of a Perfect ModelPR Curve of a Good ModelPR Curve of a Bad ModelFinding optimal threshold from PR Curve What are Precision and Recall? Precision is defined as the fraction of the relevant positives out of the retrieved ones. Recall is defined as the fraction of retrieved positives out of the relevant ones. To understand from an example, let’s imagine we’re casting a fishing net in a lake and hoping to catch some fish. However, there are some stones as well in the lake and it’s likely that our net will end up catching some stones as well. Now there can be a few different ways to look at this situation: We may want to catch as many fish as possible from the lake, no matter how many stones we catch along with them. We may want to catch only the fish and minimize the number of stones that we catch, no matter how few fish we catch. Or, we may want to catch most of the fish present in the lake and at the same time minimize the number of stones caught. Think of the fishing net as the model which gives out some outputs (preferably, fish). However, just like everything in life, no model is perfect and hence, our net catches some stones as well along with fish. The fish present in the lake is the Relevant outputs. The Retrieved output contains some fish and some stones. Precision = fraction of fish among the retrieved stuff Recall = fraction of fish retrieved from the lake In Case 1, we want to maximize Recall and ignore Precision. In Case 2, we want to maximize Precision and ignore Recall. In Case 3, we want to keep a balance between Precision and Recall and try to maximize both at the same time. This is an ideal situation. Now we can formally define Precision and Recall as follows: What is the Need for a Precision-Recall Curve? To classify an output to either class (0 or 1), we need to apply a threshold filter (just like the fishing net). For example, a default threshold of 0.5 is taken to classify outputs (any output >= 0.5 will belong to class 1). I have talked about how different thresholds might be useful for different kinds of problems here. Coming to the question of why we need a PR curve when we already have the ROC curve (which plots True and False Positive Rates for different thresholds). Here are 2 reasons why we need PR curve: ROC curve provides an overly optimistic picture of the performance, compared to PR curve, when it comes to imbalanced classification. Also, when class distribution changes, ROC curve doesn’t change, however, PR curve does reflect the change. How to read a PR Curve Fig 1 On the curves, each point corresponds to a different threshold, and its location corresponds to the resulting Precision and Recall when we choose that threshold. Some important pointers on the curve: Point 1 corresponds to the threshold of 1 Point 3 corresponds to the threshold of 0 Point 4 corresponds to the threshold somewhere in the range (0, 1) Point 2 corresponds to a Perfect model (along with Point 3) The Area under the PR Curve is a metric that helps us compare 2 similar looking curves. Higher the AUC, the better the performance. The baseline of PR Curve Fig 2 Fig 3 The Baseline of a PR Curve changes with class imbalance, unlike ROC Curve. This is due to the fact that Precision of a No-Skill model (which gives 0.5 score for every output) directly depends on the class imbalance. Fig 2 shows the baseline corresponding to a balanced dataset, whereas, Fig 3 shows the baseline corresponding to a dataset having a 10% positive class. From here on, we will stick to a 10% positive class dataset (as is the case for so many real-life datasets). PR Curve of a ‘No Skill’ Model Fig 4 The PR curve of a no-skill model (which gives 0.5 output for every data point) consists of 2 points: Point 1 Corresponds to threshold = 0.5 Point 2 Corresponds to threshold ???? [0, 0.5) Precision isn’t defined for thresholds ???? (0.5, 1] for a no-skill model due to division by zero. You may notice that Precision is a constant here, i.e., 0.1 (= class imbalance) AUC = 0.1 PR Curve of a Perfect Model Fig 5 The PR curve of a perfect model also consists of 2 points: Point 1 Corresponds to threshold ???? (0, 1] Point 2 Corresponds to threshold = 0 AUC = 1 PR Curve of a Good Model Fig 6 The PR curve of a good model consists of as many points (or thresholds) as which result in a different set of precision and recall for the dataset: Point 1 Corresponds to threshold = 1 Point 3 Corresponds to threshold = 0 Point 4 Corresponds to threshold ???? (0, 1) AUC ???? (0.1, 1) PR Curve of a Bad Model Fig 7 The PR curve of a bad model goes even below the baseline. The curve indicates that the model performs even worse than a no-skill model. An obvious way to improve the performance of this model, without tweaking anything, is to just reverse the output (Class 0 <-> Class 1) that the model gives. This will result in a higher-than-baseline performance automatically. Usually, a PR curve like this indicates that there’s definitely something wrong in the pipeline. It can be the case that the data is too random. Or, the model just can’t grasp ANY trend from the data (in other words, the model is too simple for the data). An example can be a basic linear model trying to fit on a complex non-linear dataset. AUC ???? (0, 0.1) There might be a hybrid case where a bad model works better than baseline for certain thresholds. Finding optimal threshold from PR Curve Fig 8 Since we now know that a good model’s PR curve approaches the Perfect model point (i.e. the point 2); it is quite intuitive that the optimal threshold for our model is going to be corresponding to the point on the curve which is closest to Point 2. Here are 2 ways to find the optimal threshold: Find the euclidean distance of every point on the curve, which is denoted by (recall, precision) for a corresponding threshold, from (1,1). Pick the point and the corresponding threshold, for which the distance is minimum. Find F1 score for each point (recall, precision) and the point with the maximum F1 score is the desired optimal point. You may recall (pun intended) that F1 score is the harmonic mean of Precision and Recall. Conclusion Some key pointers worth noting: Recall of a No-Skill model lies in the set {0.5, 1} irrespective of the class imbalance. Precision of a No-Skill model is equal to the fraction of positive class in the dataset. It is possible to get a model which performs worse than a no-skill model, especially when the data is too complex for the model. Just like the ROC curve, a PR curve can also be used to find the optimal threshold. PR curve works better than ROC curve in cases of imbalanced data. Today, we’ve learnt all the basics of PR Curve and how it is used in classification problems and its utility! I hope now you won’t have any problem working with PR Curves in your future classification models. Popular Posts Dickey Fuller Test in Time Series Hidden Markov Model and Application Scikit Learns Mlp Classifier Python Tabulate Module Pytorch Optimizers","excerpt":"Precision and Recall are two of the most important metrics to look at when evaluating an imbalanced classification model. These help us to find out what fraction of the actual positives were classified correctly and among the ones classified positive, what fraction were actually positives. Going ahead, we will try to explore the significance of […]","categories":["Deep Tech"],"tags":["Guide"],"author_name":"Aishit Dharwal","publish_date":"2021-11-21T19:02:00","publication_year":"2021","word_count":1370,"keywords":["Go","TPU","programming_languages:R","PyTorch","AI","ML","Python","ai_frameworks:PyTorch","programming_languages:Python","R","Guide"],"extracted_tech_keywords":["AI","ML","PyTorch","TPU","Python","R","Go","ai_frameworks:PyTorch","programming_languages:Python","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/complete-guide-to-understanding-precision-and-recall-curves\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10163121,"title":"Meta&#8217;s New Research Begins Decoding Thoughts from Brain Using AI","content":"Meta has been making impressive strides in the AI space, recently surpassing its earning estimates along with its plan to invest $65 billion to build a 2GW+ data centre. Now, it has showcased progress in using AI to decode language from the brain to help people with brain injuries who have lost their ability to communicate. Neuroscience and AI Researchers Working Together for Breakthroughs Meta collaborated with the Basque Center on Cognition, Brain, and Language (BCBL), a leading research centre in San Sebastián, Spain, to study how AI can help advance our understanding of human intelligence. The goal is to achieve advanced machine intelligence (AMI). During the announcement of the new research, Meta said, “We’re sharing research that successfully decodes the production of sentences from non-invasive brain recordings, accurately decoding up to 80% of characters, and thus often reconstructing full sentences solely from brain signals.” The research was led by Jarod Levy, Mingfang (Lucy) Zhang, Svetlana Pinet, Jérémy Rapin, Hubert Jacob Banville, Stéphane d’Ascoli, and Jean Remi King from Meta. The study involved 35 healthy volunteers who typed memorised sentences while their brain activity was recorded. They were seated in front of a screen with a custom keyboard on a stable platform. The volunteers were asked to type what they saw on the screen without using backspace. According to the research paper, a new deep learning model, Brain2Qwerty, was designed to decode text from non-invasive brain recordings like electroencephalogram (EEG) and magnetoencephalography (MEG). The model uses a three-stage deep learning architecture, a convolutional module to process brain signals, a transformer module, and a pre-trained language model to correct the transformer’s output. While it remains unconfirmed whether this model used ‘The Frontier AI Framework’, it is possible that future studies could incorporate it. Even with the advancements in the AI model, invasive methods continue to remain the gold standard for recording brain signals. However, these tests are a significant step towards bridging the gap between non-invasive and invasive techniques. Meanwhile, Jean-Rémi King, brain and AI tech lead, said, “The model achieves down to a ~20% character-error-rate on the best individuals. Not quite a usable product for everyday communication…but it’s a huge improvement over current EEG-based approaches.” “We believe that this approach offers a promising path to restore communication in brain-lesioned patients…without requiring them to get electrodes implanted inside,” King added. Meta also announced a $2.2 million donation to the Rothschild Foundation Hospital to support the neuroscience community’s collaborative work. While this is not something that we can use at the moment or benefit from, the insights from Meta’s new research sound promising about how AI can make a difference in the neuroscience field.","excerpt":"The study involved 35 healthy volunteers who typed memorised sentences while their brain activity was recorded.","categories":["AI News"],"tags":["Meta AI"],"author_name":"Ankush Das","publish_date":"2025-02-10T14:35:24","publication_year":"2025","word_count":442,"keywords":["Go","API","Meta AI","TPU","AI","ETL","RPA","BERT","deep learning","ViT","R"],"extracted_tech_keywords":["AI","deep learning","TPU","R","Go","API","ETL","BERT","ViT","RPA"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/metas-new-research-begins-decoding-thoughts-from-brain-using-ai\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":17341,"title":"MATLAB In Deep Learning, Analytics Space; Announces R2017B, Massive Update In September","content":"With over 2 million users worldwide, MATLAB is one of the largest and the most widely used language for technical computing. Needless to mention, there is a constant demand for professionals who use MATLAB for visualisation and programming. MATLAB is primarily known for its matrix algebra capabilities but its toolboxes offer other functionalities ranging from scientific and engineering graphics to modelling, simulation, and prototyping along with application development, including graphical user interface building. However, did you know, that its toolboxes also helps to create, train, and simulate shallow and deep learning neural networks, in addition, to analyse and model data using statistics and machine learning? A vast collection of MATLAB functions in toolboxes also help solve problems of a specific class. Signal processing, neural networks, wavelets, simulation, fuzzy logic, control systems and much more are the areas where tool boxes are available. Brushing Up Your MATLAB Skills It would not be wrong to say that everybody who comes from a technical background has at least once used MATLAB. But when it comes to using it professionally for analytics, its almost always the case where most users often end up having difficulty in using the platform. “We provide free MATLAB training known as the ‘MATLAB Onramp’ for anyone interested in getting started with learning and using it. You don’t need any license; all you need is a MathWorks account. It is hugely popular among students and professionals,” said Kishore Rao, managing director, MathWorks India, while speaking with AIM. The 2 – hour MATLAB Onramp course provides hands-on experience via the use of an integrated, web-based version of MATLAB. MATLAB For Deep Learning For MATLAB, a continued area of investment has been the imaging space with new image processing and computer vision algorithms, more complete coverage for standard camera interfaces such as USB3 Vision, and FPGA-ready implementations of common vision algorithms. Earlier this year, it made improvements in vision-specific tools for deep learning and automated driving. Deep learning –  A new algorithm area that enables recognition accuracy at levels better than humans. MATLAB provides a framework that makes deep learning easy to use for engineers and scientists. It also provides easy access to pre-trained networks developed by experts and leverages GPUs without GPU programming. With just a few lines of code, MATLAB helps engineers to explore and develop high-performance recognition systems. Automated driving – Vision systems are a key part of ADAS and automated driving systems. Earlier this year, MathWorks introduced the Automated Driving System Toolbox to enable ground-truth labelling, generation of synthetic driving scenarios, sensor fusion, and vision algorithms designed for automated driving. “You can use MATLAB to learn and gain expertise in the area of deep learning. Most of us have never taken a course in deep learning. We have to learn on the job. MATLAB makes learning about this field practical and accessible. In addition, it enables domain experts to do deep learning – instead of handing the task over to data scientists who may not know your industry or application,” the company said. Further Nurturing MATLAB “With MATLAB, you can integrate results into your existing applications. It automates deploying your deep learning models on enterprise systems, clusters, clouds, and embedded devices,” Rao added. Moreover, to aid the idea and vision of the company to “enable engineers to become data scientists,” MATLAB is all set to release its latest R2017b version in September. “It’s a big release for data analytics. We are adding significant details to machine learning in addition to a new application for text analytics toolbox,” Rao said. MathWorks, with MATLAB, is dedicated to filling the gap between engineers and data scientists to provide them with the right tools for deep learning. What Else? For multidimensional data analysis, Statistics and Machine Learning Toolbox provides feature selection, stepwise regression, principal component analysis (PCA), regularisation, and other dimensionality reduction methods that let you identify variables or features that impact your model. Moreover, the company is also venturing into providing IoT solutions. “We provide Internet of Things (IoT) solutions on MATLAB too. You can publish data from your target hardware to the IoT using ThingSpeak. It provides real-time data collection, data processing, visualizations, apps, and plugins,” Rao said. ThingSpeak, an IoT analytics platform service, allows users to aggregate, visualize, and analyse live data streams in the cloud. You can send data to ThingSpeak from your devices, create instant visualizations of live data, and send alerts using web services like Twitter® and Twilio®. With MATLAB analytics inside ThingSpeak, you can also write and execute code to perform pre-processing, visualisations, and analyses. ThingSpeak enables engineers and scientists to prototype and build IoT systems without setting up servers or developing web software. MATLAB and Data Analytics Much of the data analytics market focuses on applications in sales and marketing.  Given the roots in engineering and science, MathWorks focuses more on applications that involve physical-world data such as sensor, image, video, and telemetry data.  This includes applications such as predictive maintenance, where companies are looking to predict failures in their equipment by analysing sensor data, as well as advanced driver assistance systems that leverage video, lidar, and telemetry data to improve vehicle safety. Oftentimes, the people working on these projects have expertise about the equipment that is being built, but they don’t have extensive statistical or machine learning expertise.  As a result, MATLAB has focused on making it easier to apply machine learning and deep learning in practice.  This includes point-and-click apps that users can use for getting started and consistent interfaces that make it easy to combine tools from different domains.  Lastly, it provides deployment paths for running deployed analytics on embedded devices as well as IT infrastructure and clouds.","excerpt":"With over 2 million users worldwide, MATLAB is one of the largest and the most widely used language for technical computing. Needless to mention, there is a constant demand for professionals who use MATLAB for visualisation and programming. MATLAB is primarily known for its matrix algebra capabilities but its toolboxes offer other functionalities ranging from […]","categories":["IT Services"],"tags":["Deep Learning","Machine Learning","Mathworks","MATLAB","stepwise regression machine learning"],"author_name":"Priya Singh","publish_date":"2017-08-30T06:12:04","publication_year":"2017","word_count":947,"keywords":["Go","machine learning","AI","neural network","MATLAB","Machine Learning","computer vision","RAG","Mathworks","Aim","deep learning","stepwise regression machine learning","analytics","Deep Learning","R"],"extracted_tech_keywords":["AI","machine learning","deep learning","neural network","computer vision","analytics","Aim","RAG","R","Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/matlab-deep-learning-analytics-space-announces-r2017b-massive-update-september\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10098172,"title":"7 Transformative Accessibility Tech Solutions","content":"In 2022, the Global Assistive Technology market was valued at $21.95 billion and is expected to hit $31.22 billion by 2030. Big tech companies create accessibility products to demonstrate inclusivity, comply with legal requirements, and tap into a growing market. Such products enhance the user experience, foster innovation, and provide a competitive advantage. Here are a few of the products and services offered by big tech companies such as Apple, Microsoft and Google to build an inclusive environment for the visually and hearing impaired. Voice Control Apple’s Voice Control is an advanced accessibility feature available on iOS and macOS devices. It allows users with motor impairments or limited physical dexterity to control their Apple devices entirely through voice commands. This feature goes beyond traditional voice assistants, providing comprehensive control over the entire operating system and various apps, making it an essential tool for individuals with disabilities seeking a more accessible and independent user experience. Be My Eyes Starting from 2012, Danish startup Be My Eyes has been dedicated to developing technology for the visually impaired community, which consists of over 250 million people who are blind or have low vision. With the introduction of GPT-4, they are now working on creating a GPT-4 powered Virtual Volunteer integrated into their app. This virtual volunteer aims to offer the same level of context and understanding as a human volunteer, leveraging the visual input capabilities of GPT-4 to better support the visually impaired community. Seeing AI Seeing AI is an innovative and powerful mobile app developed by Microsoft for iOS devices, designed to assist people with visual impairments. The app harnesses the capabilities of artificial intelligence and computer vision to provide real-time assistance to users in understanding the world around them. Ease of Access Center Windows Ease of Access Center is a central hub of accessibility features and settings built into the Windows operating system. It is designed to make Windows PCs more user-friendly and accessible for individuals with disabilities or impairments. The Ease of Access Center provides a comprehensive set of tools such as Narrator (screen reader), Magnifier (screen magnification), and Speech Recognition, that cater to various accessibility needs, allowing users to customize their computing experience according to their requirements. Google TalkBack Google’s Android Accessibility Suite – a comprehensive collection of accessibility features and services to make Android devices more inclusive and accessible to users with diverse disabilities – offers services such as TalkBack and Live transcribe. This is a screen reader that provides audible feedback to users with visual impairments. It reads aloud on-screen text, buttons, icons, and other elements, enabling users to navigate their devices, interact with apps, and access information effectively. There is also a TalkBack Braille feature that integrates with external braille devices, allowing blind users to read and navigate the interface through braille input and output. Google Live Transcribe It is designed to assist individuals with hearing impairments by providing real-time speech-to-text transcription. The app uses the device’s microphone to capture spoken words and instantly converts them into written text on the screen, allowing users to read and follow conversations or speeches in real-time. Switch Control Apple’s Switch Control is a powerful accessibility feature available on iOS and macOS devices. It is designed to cater to individuals with motor impairments, physical disabilities, or limited dexterity, providing them with an alternative and customizable way to interact with their Apple devices. Switch Control allows users to control their devices using external adaptive switches, which can be physical buttons, Bluetooth-enabled devices, or other assistive technology.","excerpt":"Big tech companies such as Google, Microsoft and Apple have all invested in creating products and services that assist the visually and hearing impaired","categories":["AI Trends"],"tags":["Apple","Google","GPT-4","iOS","Microsoft","Speech Recognition","Windows"],"author_name":"Vandana Nair","publish_date":"2023-08-07T12:00:00","publication_year":"2023","word_count":585,"keywords":["Go","artificial intelligence","TPU","AI","Apple","iOS","GPT-4","Windows","computer vision","RAG","GPT","Aim","ViT","Google","Speech Recognition","R","Microsoft"],"extracted_tech_keywords":["AI","artificial intelligence","computer vision","Aim","RAG","TPU","R","Go","GPT","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/7-transformative-accessibility-tech-solutions\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":46265,"title":"5 Groundbreaking Papers That Are Testimony To Yann Lecun’s Ingenuity","content":"Deep Learning has benefited primarily and continues to do so thanks to the pioneering works of Geoff Hinton, Yann Lecun and Yoshua Bengio. Contributions of Yann Lecun, especially in developing convolutional neural networks and their applications in computer vision and other areas of artificial intelligence form the basis of many products and services deployed across most technology companies today. Here are a few of Yann’s groundbreaking research papers that have contributed greatly to this field: Back-propagation Applied to Handwritten Zip Code Recognition Cited by: 5172 | Published in 1989 The ability of neural networks to generalize can be greatly enhanced by providing constraints from the task domain. As a follow up to his widely popular work on back-prop, in this paper, Yann and his peers demonstrate how such constraints can be integrated into a backpropagation network through the architecture of the network. This approach has been successfully applied to the recognition of handwritten zip code digits provided by the US Postal Service. A single network learns the entire recognition operation, going from the normalized image of the character to the final classification. Read the original paper here. Generalization And Network Design Strategies Cited by: 737 | Published in 1989 This paper demonstrates techniques to improve learning speed in networks for image recognition tasks and how these approaches can be extended to other applications such as speech recognition. The main point of this work is to show that good generalization performance can be obtained if some a priori knowledge about the task is built into the network. Read the original paper here. Convolutional Networks For Images, Speech, And Time Series Cited by: 2578 | Published in 1995 In this seminal paper, Yann collaborated with Bengio to uncover the reach of CNNs. Today, many machine vision tasks are flooded with CNNs. They are the workhorses of autonomous driving vehicles and even screen locks on mobiles. This work discusses the variants of CNNs addressing the innovations of Geoff Hinton while also indicating how easy it is to implement CNNs on hardware devices dedicated to image processing tasks. Read the original paper here. Gradient-based Learning Applied To Document Recognition Cited by: 20831 | Published in 1998 The main message of this paper is that better pattern recognition systems can be built by relying more on automatic learning and less on hand-designed heuristics. Yann Lecun along with fellow Turing award winner Yoshua Bengio, demonstrate that show that the traditional way of building recognition systems by manually integrating individually designed modules can be replaced by a well-principled design paradigm called Graph Transformer Networks that allows training all the modules to optimise a global performance criterion. Read the original paper here. Efficient Back-prop Cited by: 2799 | Published in 1998 In this paper, Yann and his collaborators demonstrate why back-propagation works the way it works. In this 20-year-old research, the authors propose tricks to improve back-prop. This paper is significant now more than ever as there has been a sporadic rise in search of alternatives to back-prop. The techniques detailed in this work will navigate the reader through the foundations of neural networks and their shortcomings. Read the original paper here. Yann Lecun is currently the Chief AI Scientist for Facebook AI Research (FAIR) and also a Silver Professor at New York University on a part-time basis, mainly affiliated with the NYU Center for Data Science, and the Courant Institute of Mathematical Science. Yann LeCun was one of the recipients of the 2018 ACM A.M. Turing Award for his contributions to conceptual and engineering breakthroughs that have made deep neural networks a critical component of computing. He shares this award with his long-time collaborators Geoff Hinton and Yoshua Bengio. Check out Yann’s other significant works here.","excerpt":"Deep Learning has benefited primarily and continues to do so thanks to the pioneering works of Geoff Hinton, Yann Lecun and Yoshua Bengio. Contributions of Yann Lecun, especially in developing convolutional neural networks and their applications in computer vision and other areas of artificial intelligence form the basis of many products and services deployed across […]","categories":["AI Trends"],"tags":["back propagation","Facebook AI","papers","Yann LeCun"],"author_name":"Ram Sagar","publish_date":"2019-09-23T19:00:10","publication_year":"2019","word_count":619,"keywords":["data science","back propagation","Yann LeCun","artificial intelligence","Go","Facebook AI","AI","neural network","image recognition","Git","computer vision","deep learning","papers","R"],"extracted_tech_keywords":["AI","artificial intelligence","deep learning","neural network","computer vision","data science","image recognition","R","Go","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/5-groundbreaking-papers-that-are-testimony-to-yann-lecuns-ingenuity\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10087102,"title":"Over 100K Employees Laid off in 2023","content":"According to Layoff.fyi, over 100,000 employees have been laid off in just the first two months of 2023. This comes from only 334 companies, whereas there were 1044 companies that laid off close to 160,000 employees in the whole of 2022. But on the other side, 637 out of the 1795 listings, i.e. 35%, shows the number of laid-off employees as zero. This means that the case can be worse. Since, the website only includes a fraction of companies that have laid off employees, some predict that the number can be north of half a million easily if others are also included. Amid the mounting layoff by technology companies, the latest one was Yahoo, when on Thursday it announced that it is laying off around 20% of its workforce by the end of this year. The job cuts will start this week with about 1,000 employees, and around 600 will be asked to leave by the end of six months. The company’s headcount is expected to drop 50% by the end of the year. Spotify announced in January about laying off 6% of its workforce, which is roughly 590 employees. Amazon plans to fire 18,000 employees, Cisco is cutting around 4000 jobs, and Adobe also joined the trend by firing around 100 employees in December. Bob Iger, the CEO of Disney, revealed on Wednesday that the company plans to cut 7,000 jobs, equivalent to 3% of its workforce, in an effort to save $5.5 billion. This comes after a string of similar announcements from big tech companies such as Zoom, which will cut 15% or 1,300 jobs, eBay will let go of 4% or 500 employees, PayPal will reduce its workforce by 7% or 2,000 employees, SAP will cut 2.5% or 2,800 jobs, IBM will lay off 1.5% or 3,900 employees, and Dell will eliminate 6,000 jobs, which represents 5% of its workforce.","excerpt":"Amid the mounting layoff by technology companies, the latest one was Yahoo, when on Thursday it announced that it is laying off around 20% of its workforce by the end of this year.","categories":["AI News"],"tags":["Layoffs"],"author_name":"Mohit Pandey","publish_date":"2023-02-10T14:58:29","publication_year":"2023","word_count":313,"keywords":["Go","programming_languages:R","Layoffs","AI","programming_languages:Go","R"],"extracted_tech_keywords":["AI","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/over-100k-employees-laid-off-in-2023\/","complexity_score":3,"technical_depth":5,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":42178,"title":"5 Ways To Reduce The Chances Of Hiring Wrong Data Scientist","content":"Hiring the wrong candidate can be an expensive affair. Not just in terms of the onboarding expenses, but long-term repercussions such as project delays, below average results, increased attrition rates and others. While the companies take a lot of precautions and elaborate procedures such as a combination of technical interviews and coding assessments along with soft skills to make sure they are hiring a perfect candidate, there are chances of them going wrong. More often than not, there are other biased factors that might lead to a wrong hire, such as a candidate having a vibrant personality, candidates having come through referral, among others. It is therefore important to make a sound decision while hiring a candidate and overcome the challenges of having a wrong hire. In this article, we talk about five such measures that companies can take to reduce the chances of hiring a wrong candidate in data science. 1| Not falling blindly for keywords such as machine learning, NLP, AI and others: Most of the times, the job requirements in data science end up mentioning key skills such as AI, NLP and others as must-required in a candidate, which are conveniently mentioned by most of the candidates in their resumes even if they are not completely acquainted with the terms. Using application tracking that only picks up resumes with these keywords may, therefore, prove detrimental. It is important for recruiters to analyse the genuine interest and knowledge of the potential candidate in these areas to stop from adding just another headcount in your team. 2| Conducting hands-on tests and hackathons: In an interaction with Analytics India Magazine, Srinidhi Rao of TheMathCompany had said that candidates recruited through hackathons have a lower chance of being wrong hires and we couldn’t agree more. Tests and practices where actual analytical skills and understanding of the tools of a candidate are exposed can only prove to be quite helpful. Hackathons are one of the tried-and-tested solutions to bridge the gap between theoretical and practical knowledge. As opposed to traditional methods of candidate evaluation, hackathons simulate a real-world environment for candidates with components like solving real business problems to working in a deadline-based environment. 3| Overlapping of designation such as data science, analysts etc: It also happens quite often that companies end up posting out a job requirement that is quite different from actual challenging and thought-provoking work that data scientist might be looking for. For instance, many companies use the words data science, data analysts, business analysts and many other designations interchangeably, that might confuse the candidate, often ending up being a wrong hire. It is therefore important for companies to put out very specific job requirements specifying skills in data science that they are actually looking for. 4| Ensuring that the candidate’s and company’s interests are on the same page: While a company may find a candidate with all the required skills that they want, it is necessary to cross-check with the potential candidate on the kind of opportunity and growth that they are looking for. It might end up being a disaster hire if the end goals of employees and companies do not lie on the same page. 5| Do not rush into hiring the first data science candidate you interview: There is a huge demand for data scientists today and companies are in a rush to hire as soon they get one. Rushed hiring may end up into being a wrong hire as there may be chances of not getting the background check right. In a race to be data scientists, there are a lot of people who like to call themselves data scientists just based on Excel or some basic visualisation tool knowledge. This might later turn out to be disappointing as a data science job role is much more than that.","excerpt":"Hiring the wrong candidate can be an expensive affair. Not just in terms of the onboarding expenses, but long-term repercussions such as project delays, below average results, increased attrition rates and others. While the companies take a lot of precautions and elaborate procedures such as a combination of technical interviews and coding assessments along with […]","categories":["AI Trends"],"tags":["Data Science Hiring"],"author_name":"Srishti Deoras","publish_date":"2019-07-10T15:00:02","publication_year":"2019","word_count":635,"keywords":["data science","Go","machine learning","programming_languages:R","AI","programming_languages:Go","RAG","NLP","Data Science Hiring","analytics","R"],"extracted_tech_keywords":["AI","machine learning","NLP","data science","analytics","RAG","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/5-ways-to-reduce-the-chances-of-hiring-wrong-data-scientist\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10172839,"title":"Zoho Opens First AI &#038; Robotics R&#038;D Centre in Kerala, Brings Tech Jobs to Rural Kollam","content":"Zoho Corporation has officially launched its new research and development (R&D) centre in Neduvathoor, a village near Kerala’s Kollam district, on July 2. This is Zoho’s first AI and robotics research institute in Kerala, bringing advanced technology work to rural areas. The new centre is situated within a 3.5-acre Information Technology park developed by Zoho. It commenced operations with a workforce of 250 employees, and intends to expand further. What makes this centre special is its location in a village, away from major cities, part of Zoho’s model of setting up offices in smaller towns and rural areas. To begin with, Zoho has hired about 40 fresh graduates from rural parts of the region for training in artificial intelligence (AI), machine learning, robotics, and other engineering fields. “We want to create high-tech jobs where people live, not force them to move to cities,” said Sridhar Vembu, co-founder of Zoho Corporation, during the launch event. The new centre aims to support local talent, create employment, and boost rural development by introducing advanced technology work to areas often left out of India’s IT growth story. “This initiative shows how research and innovation can thrive outside big cities,” said Jayaraj Poroor, principal researcher at Zoho. Zoho’s move is expected to encourage other companies to explore similar models, blending innovation with inclusive development.","excerpt":"It has started operations with 250 employees, and the company plans to expand further.","categories":["AI News"],"tags":["zoho"],"author_name":"Shalini Mondal","publish_date":"2025-07-03T18:01:32","publication_year":"2025","word_count":219,"keywords":["zoho","machine learning","artificial intelligence","programming_languages:R","AI","innovation","RAG","Aim","ai_applications:robotics","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","Aim","RAG","R","innovation","programming_languages:R","ai_applications:robotics"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/zoho-opens-first-ai-robotics-rd-centre-in-kerala-brings-tech-jobs-to-rural-kollam\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10022870,"title":"Amazon Appoints Former Tableau Executive Adam Selipsky To Lead Its Cloud Business","content":"In recent news, Adam Selipsky, the current CEO of Salesforce-owned Tableau, announced that he would be leaving the company to head Amazon’s cloud business. With Selipsky leaving Tableau, executive Mark Nelson, with 25 years of enterprise software experience, will be substituting him and stepping up as the new CEO of this data analytics firm. Selipsky became the CEO of Tableau in 2016 after leaving Amazon, which Salesforce currently owns. The news of this change of hands followed Amazon’s recent announcement of Andy Jassy leaving the cloud business and replacing Jeff Bezos as the CEO. Jassy announced this new hiring via an email to his employees. He mentioned that Selipsky will be taking change from May 17th and will begin the transition before Jassy officially takes over as CEO in the third quarter of the year. Selipsky has been renowned for revamping the data analytics company, Tableau after joining in 2016, including advancing its position in the cloud industry. Considering AWS generates over $40 billion in annual revenue and is the de facto leader in cloud computing, with over 30% of market share, Selipsky, with his product expertise, is the perfect choice to lead such a massive business bring success. Selipsky was also rumoured to be a candidate to replace Jassy as AWS CEO. On the other hand, Nelson had joined Tableau in 2018 as an executive VP of product development, leading its global engineering team. Before this, Nelson had served as chief technology officer at SAP-owned Concur for roughly four years and had spent 16 years at Oracle in various product roles. According to an official statement by Salesforce, “Mark has been on the Tableau leadership team for three years, and he has deep relationships with our customers and employees.” With COO Bret Taylor rumoured to be preparing for the role of CEO at Salesforce, there wasn’t any chance for Selipsky to grow further in the company, and that’s why he decided to move AWS to lead its cloud business. This, in turn, opens up opportunities for Nelson to lead this data visualisation company. Since last year, Salesforce has been increasingly integrating Tableau into its other products, such as Salesforce’s AI analytics customer relationship management (CRM) product, which it renamed Tableau CRM. Thus, Nelson’s appointment can hit the jackpot for Salesforce to take on Tableau’s product development. Futurum Research analyst Dan Newman stated to the media that Nelson’s background with cloud infrastructure at Oracle would be beneficial for overall Salesforce businesses. As a matter of fact, Salesforce recently announced a new architecture for its platform — Hyperforce, that allows it to run on any public cloud. Read the entire blog post here.","excerpt":"In recent news, Adam Selipsky, the current CEO of Salesforce-owned Tableau, announced that he would be leaving the company to head Amazon’s cloud business.  With Selipsky leaving Tableau, executive Mark Nelson, with 25 years of enterprise software experience, will be substituting him and stepping up as the new CEO of this data analytics firm.  Selipsky […]","categories":["AI News"],"tags":["Andy Jassy"],"author_name":"Sejuti Das","publish_date":"2021-03-24T16:04:03","publication_year":"2021","word_count":442,"keywords":["AWS","cloud computing","AI","cloud_platforms:AWS","programming_languages:R","analytics","R","Andy Jassy"],"extracted_tech_keywords":["AI","analytics","cloud computing","AWS","R","cloud_platforms:AWS","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/amazon-appoints-former-tableau-executive-adam-selipsky-to-lead-its-cloud-business\/","complexity_score":3,"technical_depth":7,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10056626,"title":"Hugging Face Releases Perceiver IO, A Next Generation Transformer","content":"Hugging Face has added Perceiver IO, the first Transformer-based neural network that works on all kinds of modalities, including text, images, audio, video, point clouds and even combinations of these. The Transformer architecture has a limitation where its self-attention mechanism scales very poorly in compute as well as memory. In every layer, all inputs are used to produce queries and keys, for which a pairwise dot product is computed. Hence, it is impossible to apply self-attention to high-dimensional data without some form of preprocessing. The Perceiver solves this by employing the self-attention mechanism on a set of latent variables rather than on the inputs. The inputs are only used for doing cross-attention with the latents. This has the advantage that the bulk of compute happens in a latent space, where compute is cheap. The resulting architecture has no quadratic dependence on the input size: the Transformer encoder only depends linearly on the input size, while latent attention is independent of it. To initialize a PerceiverModel, developers can provide three additional instances to the model – a preprocessor, a decoder, and a postprocessor. The inputs are first optionally preprocessed using a preprocessor. Then the preprocessed inputs perform a cross-attention operation with the latent variables of the Perceiver encoder. After that, the decoder can be used to decode the final hidden states of the latents into something more useful, like classification logits. Last, the postprocessor can be used to post-process the decoder outputs to specific features. Implementing Perceiver: Perceiver for text: PerceiverTextPreprocessor, as preprocessor to the model, takes care of embedding the inputs and adds absolute position embeddings. As decoder, one provides PerceiverClassificationDecoder to the model and postprocessor is not required here.Perceiver for images: The Perceiver to perform text classification, it is straightforward to apply the Perceiver to do image classification. There is a different preprocessor to the model, which will embed the image inputs.Perceiver for multimodal autoencoding: The goal of multimodal autoencoding is to learn a model that can accurately reconstruct multimodal inputs in the presence of a bottleneck induced by an architecture. The advantage of the Perceiver is that the compute and memory requirements of the self-attention mechanism don’t depend on the size of the inputs and outputs, as the bulk of compute happens in a latent space (a not-too large set of vectors). The model is available in HuggingFace Transformers.","excerpt":"Hugging Face’s newly added Perceiver IO to Transformers works on all modalities like text, images, audio, etc.","categories":["AI News"],"tags":["Generative Pre-Trained Transformer","Hugging Face","Neural Networks","Transformers"],"author_name":"Meeta Ramnani","publish_date":"2021-12-20T16:35:26","publication_year":"2021","word_count":391,"keywords":["Go","Hugging Face","text classification","TPU","AI","neural network","Transformers","Git","Generative Pre-Trained Transformer","transformer architecture","R","Neural Networks"],"extracted_tech_keywords":["AI","neural network","Hugging Face","Transformers","text classification","TPU","R","Go","Git","transformer architecture"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/hugging-face-releases-perceiver-io-a-next-generation-transformer\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":45850,"title":"9 Skills A Data Scientist Must Have To Land A Job: AIM Skills Study 2019","content":"For a few years now, many innovative things have been happening around emerging technologies like data science and machine learning. The industry has seen a rapid increase in demand for data analysts and data scientists within a short span of time. Analytics India Magazine conducted Data Science Skills Study to understand key trends driving skills economy and how data scientists’ toolchains are evolving. In this article, we have culled insights from our informative survey to come up with a cheatsheet with 9 must-have skills analytics and machine learning enthusiasts should know about. 1| Python continues to be the Swiss army knife Besides mathematical and statistical skills, data scientists require a sound knowledge of programming languages. According to the survey report, the popular programming language Python continues to be the most popular language in the industry in 2019 with its popularity growing to 68%. Besides Python, there are few other programming languages such as R, SQL, and SAS which currently share the attention in the community. 2| Knowledge of Python Libraries Python is one of the versatile languages which has been used by the data scientists to carry out data science and machine learning projects. This dynamically typed language is easy to use, implement and interpret. This language has the ability to provide better insights as well as correlate data from large sets of data. Python includes a number of libraries and frameworks for data science and machine learning. Among the libraries, some of the favourite libraries include3 Pandas, Numpy, Sklearn, and Matplotlib. For deep learning, a data scientist can use TensorFlow, Keras, Theano, and Pytorch to solve complex and more advanced problems in data science and deep learning. 3| Knowledge of GPU Hardware & CUDA Data Science and deep learning models are getting complex with each day. Machine Learning techniques such as Artificial Neural Network, Natural Language Processing, among others are complex and highly data-parallel architecture in which only a powerful machine than a CPU can accomplish the computations. GPUs are used by the data scientists in order to accelerate these analytical applications. In our survey, Nvidia GeForce GTX 9 Series GPU and Nvidia GeForce GTX 10 Series proved to be the choice for 28% and 16% of the data scientists respectively. 4| Deep Understanding Of Algorithms Algorithms like Logistic Regression has been used heavily in the field of data science. Around 71% of data scientists have been utilising this method in their work. Besides Logistic Regression, other algorithms such as decision trees, convolutional neural and Feedforward Neural Network networks are also in demand for data science projects. 5| Having Great Comfort With Cloud Service Providers With data increasing at a fast pace in organisations, almost every enterprise is moving data on the cloud as compared to on-premise solutions. Apart from languages and algorithms skillset, it is important for data scientists to have a clear concept of how an organisation is storing data on the cloud. Our survey reveals that 43% of data scientists work on Amazon Web Service (AWS) while 33% and 16% of data scientists use Google Cloud and Microsoft Azure respectively. 6| Strong Command Over Visualization Tool Visualization plays an important role where data analysts need to show where the data of an organisation is leading to. Popular visualisation tool, Tableau is preferred by more than half of the respondents as per our survey. Besides Tableau, Microsoft BI is another preferred tool by  data scientists. 7| Knowing your way around Github GitHub can be said as the most widely used and popular platform for data scientists where they use it for collaborating on projects, make contributions as well as changes in a number of projects as well as trackback the changes which have been done over time. In our survey, 62% of the respondents claimed that they use GitHub for finding open data. Also, data scientists collect open data from other sources such as university websites, official government websites or collected manually. 8| Notebook as a choice IDE For writing code, testing and debugging, a data scientist needs a development environment. Integrated Development Environment is a coding tool which allows code completion by resource management, debugging, tools, etc. In our survey, Notebook, RStudio, and Pycharm show the most favourable ones. 9| Expertise in Hadoop Organisations are implementing Big Data analytics nowadays to gain insights and patterns from the large chunks of data. Harnessing data through this process is cost-effective and helps in better decision-making. In our survey, half of the recipient choose Hadoop as the preferred Big Data analytics tool and the rest use NoSQL or other customised tools.","excerpt":"For a few years now, many innovative things have been happening around emerging technologies like data science and machine learning. The industry has seen a rapid increase in demand for data analysts and data scientists within a short span of time. Analytics India Magazine conducted Data Science Skills Study to understand key trends driving skills […]","categories":["AI Trends"],"tags":["Data Science","machine learning tools","Pycharm","tools for data science","trends in bi and data visualization"],"author_name":"Ambika Choudhury","publish_date":"2019-09-12T10:00:57","publication_year":"2019","word_count":762,"keywords":["trends in bi and data visualization","data science","machine learning","Keras","AI","neural network","PyTorch","Pycharm","machine learning tools","deep learning","Aim","tools for data science","analytics","Data Science","TensorFlow"],"extracted_tech_keywords":["AI","machine learning","deep learning","neural network","data science","analytics","Aim","TensorFlow","PyTorch","Keras"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/9-skills-a-data-scientist-must-have-to-land-a-job-aim-skills-study-2019\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10089001,"title":"Tredence’s New Energy.AI Solution to Help Reduce Emissions","content":"Pure-play AI and analytics firm, Tredence, today unveiled Energy.AI, a solution that aids telecom, technology, manufacturing, and CPG companies in achieving their energy efficiency goals and meeting ESG and UN Sustainable Development targets. Energy.AI enables enterprises to monitor, capture, and optimize their energy consumption, leading to better operational efficiency and faster response times, with a guaranteed minimum increase of 30%, Tredence claims. It harnesses the capabilities of AI and provides businesses with up-to-the-minute insights to simplify procedures, reduce emissions, optimize energy usage, and achieve sustainability targets with ease. The company claims to have helped a significant telecom enterprise enhance asset performance by 17%, decrease energy consumption by 10%, and save approximately $100 million in energy costs. Currently, the telecommunications industry contributes 1.6% of the world’s carbon dioxide emissions, and a majority of these emissions (60%) stem from their energy consumption. Lakshmi Ramamurthy, Head of Telecom, Media, and Tech at Tredence, said, “Energy.AI delivers maximum value for telecoms striving to achieve sustainability, while optimizing energy emissions and costs.” “In as little as six weeks, you can use a digital-twin powered energy efficiency solution that leverages prescriptive and predictive methods to reduce the carbon footprint of your value chain,” she added. Energy.AI offers a complete perspective of energy usage and anticipatory insights, offering organisations a comprehensive outlook of their energy consumption. This allows it to identify high-emission sites, evaluate risks, and rapidly take action to decrease energy consumption. Energy.AI’s root-cause analysis and intelligent recommendations aid in efficient and effective decision-making, propelling sustainability objectives forward. The telecom industry accounts for 1.6% of global carbon emissions, while the tech sector contributes 2 to 3%, and manufacturing is responsible for almost 20%. In light of this, Energy AI is a much-needed and reliable initiative to lower emissions and optimise energy efficiency. Recently, Tredence had also announced the release of their AI-Driven platform, Atom.AI. Atom.AI uses Databricks Lakehouse to accelerate AI applications for retail and consumer packaged goods. Atom.AI asserted that it can cut down the time it takes to generate value from data by 50%. This is accomplished by utilising its AI and machine learning capabilities, pre-designed feature store, technical notebooks, extensive data sets, and third-party connections.","excerpt":"Pure-play AI and analytics firm, Tredence, today unveiled Energy.AI, a solution that aids telecom, technology, manufacturing, and CPG companies in achieving their energy efficiency goals and meeting ESG and UN Sustainable Development targets.  Energy.AI enables enterprises to monitor, capture, and optimize their energy consumption, leading to better operational efficiency and faster response times, with a […]","categories":["AI News"],"tags":["AI (Artificial Intelligence)","Tredence"],"author_name":"Shyam Nandan Upadhyay","publish_date":"2023-03-09T18:00:00","publication_year":"2023","word_count":362,"keywords":["Go","API","machine learning","AI","Git","RAG","Aim","analytics","Tredence","R","AI (Artificial Intelligence)","Databricks"],"extracted_tech_keywords":["AI","machine learning","analytics","Aim","RAG","Databricks","R","Go","Git","API"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/tredences-new-energy-ai-solution-to-help-reduce-emissions\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10161843,"title":"Bloom AI Powers BFSI with Agentic Intelligence","content":"The BFSI (Banking, Financial Services, and Insurance) sector operates in one of the most data intensive industries, where slow decision making is a competitive disadvantage. Despite enterprise data efforts, many BFSI firms remain tethered to traditional intelligence tools for business and market intelligence. To compound the issue, less than 2% of BFSI employees specialise in data and insights, highlighting an urgent need for modern intelligence solutions. Agentic and analytical AI are poised to change this landscape. Amit Shanker, the founder of Bloom AI, identified the need for proactive, synthesised intelligence early on. “We observed that clients were spending most of their time searching for data and information in dashboards and reports”, says Shanker. “There was a strong need for automated intelligence, but companies lacked the tools.” This realisation led to the founding of Bloom AI in 2021. Backed by two decades of research & analytics experience for global financial institutions, Amit set out to transform how BFSI firms approach intelligence. “Our goal is simple – build agentic and analytical AI technology to bring intelligence to every decision maker’s fingertip.” Backing this trend, Gartner recently claimed that by 2026, 80% of banks will adopt GenAI, up from current levels of 5%. A Suite for Actionable Intelligence Bloom AI is building a suite of foundational data platforms and agentic tools to modernize business and market intelligence at BFSI firms. Their approach focuses on deep collaboration, embedding teams within client organizations to fully understand challenges and drive innovation based on emerging use cases. “Deep customization and client engagement are our key differentiators,” notes Amit. One of their flagship offerings, SynthBI, integrates fragmented data to provide AI-powered analytical summaries appropriate for each user’s work context. Bloom AI’s research indicates that 80% of company data remains unused for decision-making due to system complexity. SynthBI condenses intelligence from multiple dashboards into concise, scrollable emails, making relevant insights accessible to over 80% of the teams, while reducing BI costs by up to 40%. Another key offering, DYSTL, leverages generative AI to transform unstructured data from multiple sources into automated, targeted market and company intelligence. Leveraged by corporate strategy, investor relations, marketing, research, and sales teams, DYSTL scales competitive and market intelligence beyond human limits. For one insurance client, DYSTL expanded competitive tracking by 5x while cutting costs and effort by 80%. Bloom AI also provides bespoke, subscription-based generative AI agents designed for task automation and decision support. For example, MARCO, their marketing AI agent, assists content writers with proof-reading, SEO optimization, social media, etc. while DataScout, their AI investment agent, helps private investment teams automate deal sourcing and the due diligence processes. The Impact So Far Bloom AI has been working with global investment management, insurance and financial firms with some notable milestones. In 2024, SynthBI delivered automated intelligence to the entire sales and marketing team of a global asset manager, covering nearly 14% of the company’s workforce. Another success story involves DYSTL automating account intelligence on 200+ companies, delivering insights on earnings, news, and leadership for a global tech services firm. “We realised that data actualisation is not only a data science problem but also a behavioural science problem” Shanker said in the company’s blog. By creating a smart delivery layer for personalized, automated intelligence, Bloom AI is pioneering a future where CEOs and decision-makers have actionable insights at their fingertips every day. Bloom AI’s impact extends beyond financial institutions. Recently, an Ivy League institution selected Bloom AI to develop a futuristic AI-powered content engagement platform. What’s Next Over the past 30 years, India has established itself as a global leader in IT, BPO, and KPO services. AI services are now emerging as the next big wave, and India is positioned to play a pivotal role in this evolution. Aligning to the trend, Bloom AI is adding new clients in the U.S., while also expanding its footprint in India, targeting emerging use cases with Global Capability Centres (GCCs). Traditionally concentrated on finance operations and software engineering, GCCs are now exploring AI-first applications in marketing, sales, and customer excellence —areas well-suited to Bloom AI’s expertise. “We see GCCs as ideal locations to build AI-first centers of excellence for analytical and knowledge functions” Shanker noted. As industry moves forward, Bloom AI is optimistic about the transformative potential of generative AI. While many firms are still in the early stages of leveraging AI to boost productivity, Bloom AI anticipates 2025 to be the tipping point for enterprise-scale transformative use cases.","excerpt":"Bloom AI is building a suite of foundational data platforms and agentic tools to modernize business and market intelligence at BFSI firms.","categories":["AI Highlights"],"tags":["AI agent","AI Impacts"],"author_name":"Mohit Pandey","publish_date":"2025-01-21T10:00:00","publication_year":"2025","word_count":741,"keywords":["data science","Go","GenAI","AI","RAG","AI agent","Aim","analytics","AI Impacts","generative AI","GAN","R"],"extracted_tech_keywords":["AI","data science","analytics","generative AI","GenAI","Aim","RAG","R","Go","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/bloom-ai-powers-bfsi-with-agentic-intelligence\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10081291,"title":"Javascript For Haters: Developers can Hate it But Can’t Ignore it","content":"In April 1995, Brendan Eich created ‘Javascript’, when he was still at Netscape. It was initially developed to make websites more dynamic. Written in merely ten days, Javascript has—over the years—emerged as the most widely used programming language in the world. However, to much dismay, Javascript is also one of the most hated languages among developers. The internet is filled with hilarious memes on Javascript, much like this one. Yet, even after inspiring this blend of love-hate reactions from the community, nearly 95% of the websites are built on this scripting language developed at Eich. Brands such as Amazon, eBay, Netflix, and Paypal have written their web applications in JavaScript. In 27 years since its inception, Javascript has weathered all attempts by Google, Microsoft, and Adobe to replace it with their own programming languages. https:\/\/twitter.com\/ProgrammingEra\/status\/1318386614825988099 Why do developers hate Javascript? Developers hate JavaScript because it is dynamic in nature which makes it hard for them to catch errors at compile-time. When we asked the same question to OpenAI’s ChatGPT, it said, “Some people might find the syntax of JavaScript to be confusing or difficult to work with.” And it’s not wrong! “The thing that bothers me about JS is the language syntax and semantics. Together they provide an ugly and verbose language that’s generally painful to read,” a software developer said on Quora. “When using JS, programmers must write workarounds to overcome problems in the language itself, and the workarounds are immensely complex and often cumbersome to understand.” Burnout is also another reason why JS is disliked by many developers. In fact, JavaScript burnout is often talked about by members of the frontend developers community. Most of the complaints revolve around the new Javascript frameworks, features, and libraries that come out almost everyday. This means that the developers have to keep them updated with the recent developments related to Javascript. Such constant change makes it difficult for developers to keep up. Further, another reason many developers dislike Javascript is that it does not perform the same way on all browsers. I think it’s because I hate JavaScript. https:\/\/t.co\/fFTYbEycbg— Josh Pigford (@Shpigford) November 24, 2022 Douglas Crockford, an American computer programmer who is involved in the development of the JavaScript language, said, “No programming language is perfect. JavaScript has its share of design errors, such as the overloading of + to mean both addition and concatenation with type coercion, and the error-prone with statement should be avoided. The reserved word policies are much too strict.” “Semicolon insertion was a huge mistake, as was the notation for literal regular expressions. These mistakes have led to programming errors and called the design of the language as a whole into question. Fortunately, many of these problems can be mitigated with a good lint programme,” Crockford added. Further, developers complain that Javascript is too slow or inefficient. However, these are subjective perspectives. While there are several voices on the internet complaining about the programming language, there are many who argue that Javascript does the work it was developed for. But Javascript is here to stay Despite its many flaws, it is undeniable that it is one of the most popular programming languages. It’s the only programming language that can be used across the whole stack (frontend\/backend). You can run it on any divide and on any browser. It is also the only language native to web browsers. One of the reasons why Javascript is hated is also the reason why Javascript is here to stay. The Javascript frameworks and libraries out there are why numerous developers hate this programming language but also why it is so popular. The NPM repository for Javascript is vast and probably has more packages than all the other languages combined. Even if someone is new to Javascript, they will easily find help for anything related to the programming language on the web. Irrespective of the issue at hand, you can be sure that there will be a JavaScript library or tool to help solve your problems. Further, JavaScript also has a very large community supporting it, consisting of millions of developers across the globe. The other language that comes close in terms of community support is ‘Python’. Javascript is also flexible and no other programming language so far comes close to it in that aspect. The use cases of Javascript are also massive. Full-stack development and the modern frontend framework continue to establish Javascript as one of the most popular programming languages in the world.","excerpt":"Developers find the syntax of JavaScript to be confusing or difficult to work with but it remains popular still.","categories":["AI Features"],"tags":[],"author_name":"Pritam Bordoloi","publish_date":"2022-12-01T18:00:00","publication_year":"2022","word_count":745,"keywords":["Go","ChatGPT","AWS","OpenAI","AI","Python","GPT","JavaScript","R","Java"],"extracted_tech_keywords":["AI","ChatGPT","OpenAI","AWS","Python","R","JavaScript","Go","Java","GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/developers-love-hate-relationship-with-javascript\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10170810,"title":"Bengaluru Startup Frinks AI Raises $5.4 Million, Led by Prime Venture Partners","content":"Frinks AI, a manufacturing AI startup founded by IIT Hyderabad alumni, has secured $5.4 million in a Pre-Series A funding round led by Prime Venture Partners. This brings the company’s total funding to $6.25 million. The round included participation from existing investors Chiratae Ventures, Navam Capital, and Ashok Atluri, founder of Zen Technologies. Frinks AI is also supported by industry leaders such as S Ramadorai (former MD & CEO, TCS), V Sumantran (former executive director, Tata Motors), Tarun Ramadorai (professor, Imperial College London), and Gopichand Katragadda (former group CTO, Tata Sons), who serve as strategic investors and advisors. Frinks AI develops vision AI systems focused on automating quality control on manufacturing production lines. The company’s platform is used in sectors including automobile, consumer goods, building materials, and medical devices. “With ongoing supply chain disruptions and rising global trade tensions, we’re seeing a strong push toward localised manufacturing as countries prioritise internal consumption,” said Aditya Agrawal, CEO and co-founder of Frinks AI. “This marks the beginning of a new industrial revolution—one powered by advanced technologies that drive higher productivity and cost efficiency. Frinks AI is at the forefront of this transformation, helping manufacturers worldwide become more competitive, resilient, and future-ready.” The company said that customers have reported improvements in product quality, fewer defects, and efficiency gains on the factory floor. Frinks AI’s foundational vision models combine general AI with in-house fine-tuning, achieving “99.99% accuracy,” according to Brij Bhushan, partner at Prime Venture Partners. “It is our pleasure to partner with Aditya, Dharmagya and Subhra in expanding this 10x product to India as well as taking it global,” he said. With the new funding, Frinks AI plans to scale its platform globally, invest in research and development, and expand its presence in the US. The company is exploring partnerships with automation firms and OEMs for joint go-to-market initiatives. Frinks AI’s no-code platform allows manufacturers to customise workflows, integrate with existing systems, and scale across production lines. Its technology currently runs on over 1,000 production lines worldwide, helping reduce defects and boost throughput. Founded by Aditya Agrawal, Dharmgya Sharma, and Subhra S Bhattacherjee, Frinks AI began as a research project at IIT Hyderabad and has since become a provider of AI solutions focused on visual inspection and quality control in manufacturing.","excerpt":"With the new funding, Frinks AI plans to scale its platform globally, invest in research and development, and expand its presence in the US.","categories":["AI News"],"tags":["AI startup"],"author_name":"Siddharth Jindal","publish_date":"2025-05-27T12:50:40","publication_year":"2025","word_count":378,"keywords":["Go","AI startup","API","AI","RAG","automation","ViT","disruption","GAN","R","startup"],"extracted_tech_keywords":["AI","RAG","R","Go","API","GAN","ViT","automation","disruption","startup"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/bengaluru-startup-frinks-ai-raises-5-4-million-led-by-prime-venture-partners\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10093518,"title":"LLM Wars: PaLM 2 VS GPT-4","content":"At Google I\/O event, the company unveiled PaLM 2, a next-generation language model. PaLM 2 represents a significant improvement over its predecessor, PaLM, and introduces several new capabilities that set it apart from OpenAI’s GPT-4. One of the key advantages of PaLM 2 is its availability in smaller sizes, such as Gecko, Otter, Bison, and Unicorn, which are specifically optimised for applications with limited processing power. These smaller models enable PaLM 2 to cater to a wider range of devices and products, including mobile devices that can run the lightweight Gecko model even offline. This flexibility in model sizes gives PaLM 2 an edge in terms of accessibility and deployment. Google claims that PaLM 2 demonstrates enhanced reasoning capabilities compared to GPT-4, particularly in tasks like WinoGrande and DROP, with a slight advantage in ARC-C as well. However, it’s important to note that direct comparisons between the two models can be challenging due to differences in the presentation of test results. Additionally, Google has chosen to omit some comparisons where PaLM 2 performed less favourably, raising questions about the completeness of the assessment. In terms of mathematical abilities, PaLM 2 shows improvements according to Google’s research paper. While the exact size of PaLM 2’s largest model, PaLM 2-L, remains undisclosed, Google has stated that it is significantly smaller than PaLM’s 540 billion parameters. This suggests that PaLM 2-L is likely smaller than GPT-3.5, but it still competes well with GPT-4, delivering impressive performance in various tasks. Bard’s new features also make it the better choice for research. It provides more concise summaries and improved sourcing. Users can now quickly access the core information of a topic and easily identify which parts of the response match specific sources by clicking on number tags to the corresponding sections in the linked sources. This helps when conducting research or writing essays requiring specific knowledge and detailed citations. These updates address the limitations of AI tools in verifying real-world information and enhance Bard’s research capabilities. No Disclosure While Google doesn’t disclose the exact size of PaLM 2’s training dataset, the company emphasises a focus on mathematics, logic, reasoning, and science. PaLM 2’s pre-training corpus consists of a diverse range of sources, including web documents, books, code, mathematics, and conversational data. Moreover, PaLM 2 has been trained in over 100 languages, enhancing its contextual understanding and translation capabilities. In contrast, OpenAI has trained GPT-4 using publicly available data and licensed data. GPT-4 aims to generate a wide range of responses and has been fine-tuned using reinforcement learning with human feedback, aligning its behaviour with user intent. Both PaLM 2 and GPT-4 can be accessed through their respective chatbots, Bard and ChatGPT. Bard is freely available worldwide, while ChatGPT Plus, featuring GPT-4, is behind a paywall. However, GPT-4 can also be accessed for free through Microsoft’s Bing AI Chat, which utilises the model. This accessibility plays a role in the potential adoption of PaLM 2, as it is an open-source model. Google has integrated PaLM 2 into more than 25 of its products, including Android and YouTube, while Microsoft has also incorporated AI features into its Office suite and various services. Although GPT-4 has gained traction among developers and startups due to its early release and refinement, the open-source nature of PaLM 2 may attract a wider range of users. As PaLM 2 is a relatively new model, its ability to compete with GPT-4 is still being assessed. Google’s ambitious plans and the unique capabilities of PaLM 2 suggest that it could present a formidable challenge to GPT-4. However, GPT-4 remains a capable model, outperforming PaLM 2 in several comparisons. Nevertheless, PaLM 2’s smaller models, especially lightweight options like Gecko, give it an advantage, particularly for mobile devices. With the introduction of PaLM 2 and Google’s ongoing development of the multimodal AI model Gemini, the competition for AI dominance has intensified. Google’s commitment to advancing AI technologies indicates a continued drive to innovate and challenge established players like GPT-4. The future will reveal how these language models evolve and how they shape the landscape of natural language processing and AI as a whole.","excerpt":"Google claims that PaLM 2 demonstrates enhanced reasoning capabilities compared to GPT-4, particularly in tasks like WinoGrande and DROP, with slight advantage in ARC-C as well.","categories":["AI Trends"],"tags":[],"author_name":"Shyam Nandan Upadhyay","publish_date":"2023-05-17T17:16:21","publication_year":"2023","word_count":687,"keywords":["Go","ChatGPT","OpenAI","AI","chatbots","GPT","Aim","multimodal AI","R","startup"],"extracted_tech_keywords":["AI","multimodal AI","ChatGPT","OpenAI","Aim","chatbots","R","Go","GPT","startup"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/llm-wars-palm-2-vs-gpt-4\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":61353,"title":"How Hospitals Can Tap AI To Manage Staff Better Amid Covid-19 Crisis","content":"Much has been written about the role artificial intelligence (AI) can play in the fight against the Covid-19 pandemic. While the practicality of new technology is still being explored, companies are earnestly working on AI-enabled tools that could support the measures being taken to curb the outbreak. This includes tools to enforce social distancing, assess contact tracing, share valuable insights, as well as automated rapid tests, risk assessment apps, and resources to accelerate associated drug discovery initiatives, among others. While the widespread adoption of these AI-enabled solutions for these applications could take time due to appropriate scepticism around it, there is no contesting the fact that it can expedite efforts to manage the ongoing Covid-19 crisis. But while caution should still be exercised when it comes to experimental technologies – especially during a health emergency like this – there are less-explored areas amid this crisis where technology can add a lot of value without the accompanied risk. Rostering Medical Staff Incorporating AI into the mundane – but critical – administrative functions in a hospital can relieve these institutes of unprecedented strain on resources, especially doctors, physicians, and other medical practitioners. Amid growing chaos in hospitals with more new cases emerging and existing patients demanding greater medical attention, AI may be able to create the biggest impact in this battleground in this pandemic. In this scenario, one of the biggest challenges that health systems are facing is accurately scheduling and planning a medical staff rota. Challenging to manage even in normal times, relying on spreadsheets may not be enough during a crisis. This is because with hospitals across the world having gone past the first wave of Covid-19-related hospital admissions, staff have been shifted off their usual tasks and deployed to critical care units and emergency departments. Donning and doffing layers of PPE, they have to support patients in the intensive care unit and attend to wards full of probable Covid-19 patients – all while their leaves are curtailed and shift hours extended. Scheduling, in these circumstances, has undoubtedly become more complex. What is more, large numbers of workers remain absent because they are ill themselves, or are self-isolating. Although medical students in a lot of places have been drafted early and retired doctors implored to return to service to help plug these gaps, calculating how best to deploy them alongside regular staff has been a herculean task. ALSO READ: Why Data Is Critical In The Fight Against Coronavirus How Can AI Help? A company based out of Norway Globus.ai has created an AI-enabled system that is ideally suited to help in these situations by helping hospitals find the right skilled workers to fill shifts. It has gone into overdrive since the Covid-19 outbreak by allowing hospitals to match the competencies of medical practitioners to its needs, and align it with appropriate shifts based on availability. It uses Natural Language Processing (NLP), deep learning and machine learning (ML) techniques to extract information, match the competencies of healthcare workers to specific tasks, and help fill available slots – making the planning of medical staff rota much easier and more efficient. What is more, its AI-powered staffing assistant also factors in policy and legal requirements when making these recommendations. For instance, it can account for cases where law mandates a set number of working hours, or the presence of a senior doctor for certain shifts. According to the company, it has adapted the product to handle emergency staffing during Covid-19 and has been handling about 4,500 shifts per week. This has been saving hospitals “90% of the time it takes to fill each available slot”, making it more efficient to match staff to slots. Although it is currently operational only in certain areas of Norway where it is providing public hospitals free access to its services amid Covid-19, the company has been working with Ernst & Young to roll it out on a larger scale to help hospitals and healthcare institutions across the world. ALSO READ: Are Too Many Data Scientists Trying To Predict Covid-19 In Vain? Other Ways In Which AI Can Help Without promising the moon, AI can help healthcare providers in ways that can complement the above application. This includes introduction of tools that can estimate capacity for severely ill patients and strategies to circumvent it, and enabling hospitals to plan better in the event of a surge in cases. It should also factor in availability of beds, critical care doctors, and crucial equipment like ventilators, and the ratio of these inputs to patients. In addition to these, tools to monitor and identify patients before their conditions deteriorate, and innovative ways to screen visitors should be integrated into clinical operations as well. While some progress has been made in these areas through research based on data from certain communities and by reviewing literature on past outbreaks, there is a lot more that needs to be explored. What is more, most of these are still at its nascent stage and hence, unproven. It will, thus, take time to validate these technologies.","excerpt":"Much has been written about the role artificial intelligence (AI) can play in the fight against the Covid-19 pandemic. While the practicality of new technology is still being explored, companies are earnestly working on AI-enabled tools that could support the measures being taken to curb the outbreak. This includes tools to enforce social distancing, assess […]","categories":["Deep Tech"],"tags":["ai for covid-19","artificial intelligence corona","Coronavirus","covid-19"],"author_name":"Anu Thomas","publish_date":"2020-04-11T18:00:00","publication_year":"2020","word_count":840,"keywords":["Go","API","artificial intelligence","machine learning","covid-19","AI","programming_languages:R","ML","Coronavirus","NLP","ai for covid-19","deep learning","R","artificial intelligence corona"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","NLP","R","Go","API","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/how-hospitals-can-tap-ai-to-manage-staff-better-amid-covid-19-crisis\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10053096,"title":"Why Did Alphabet Launch A Separate Company For Drug Discovery","content":"One of last week’s top stories was the launch of Isomorphic Labs, a company exclusively dedicated to drug discovery, by Alphabet. This newly formed company has Demis Hassabis (also the CEO of DeepMind) as the founder and CEO. It may be noted that DeepMind achieved a major breakthrough in biology and medical research by devising AlphaFold2. It solved the 50-year-old grand challenge of protein folding by predicting the 3D structure of a protein to atom-level accuracy directly from its amino acid sequence. Via a brief blog, Hassabis announced the launch of Isomorphic Labs — presenting a teaser of what the company would be working on and what is that it aspires to achieve. However, not much is known about why Alphabet and DeepMind felt the need for a separate company dedicated to AI-driven drug discovery. DeepMind & Biomedical Research “One of the most important applications of AI that I can think of is in the field of biological and medical research, and it is an area I have been passionate about addressing for many years. Now the time is right to push this forward at pace, and with the dedicated focus and resources that Isomorphic Labs will bring,” Hassabi wrote in the launch blog. He added that Isomorphic Labs is a commercial venture that reimagines the drug discovery process from the first AI-first perspective — to model and understand fundamental mechanisms of life. The #AlphaFold source code has been updated and now accounts for multi-chain protein complexes – providing a significant improvement in accuracy for predicting protein interactions: https:\/\/t.co\/4gyZ0loLrdGenerate predictions from your browser via: https:\/\/t.co\/2Vd4itjElu pic.twitter.com\/Db66KcXCRD— Google DeepMind (@GoogleDeepMind) November 2, 2021 As mentioned in the blog and other interviews, the success of various computational biology and drug research projects, especially that of AlphaFold, has been the motivation behind launching a commercial venture like Isomorphic Labs. Without taking away from other breakthroughs from DeepMind (think AlphaGo), AlphaFold has been special mainly because of the complexity of the challenge. Before it was demonstrated, researchers and experts in the field always believed that enumerating all the possible configurations of a typical protein structure before reaching the right 3D structure is a mammoth task and painfully time-consuming. In the latest advancement to the AlphaFold project, the DeepMind team has demonstrated an AI-based dynamic model of the human nuclear pore complex. This model shows how the protein scaffold and the nuclear envelope are coupled inside cells. We combined #AlphaFold and #cryoEM to build a new model of the nuclear pore complex, the largest complex in the human cell! The structure covered by the model is 15x bigger than the human ribosome and 2x bigger than old nuclear pore models. 1\/7Read how: https:\/\/t.co\/mCLg0hEQFr pic.twitter.com\/YpUvonwVT7— Jan Kosinski (@jankosinski) October 28, 2021 This is just the very beginning, and the field of digital biology offers huge scope for AI-based research and development work. Hassabis has often spoken about the ultimate aim of solving real-world problems via the research work at DeepMind. Biology and drug discovery feature very prominently in the list. In an earlier interview, Hassabis had mentioned that billions are invested into research by Big Pharma. Since this investment is highly dependent on the quarterly earnings report, the industry has become conservative with increasing costs of failure. He opined that the upper management of these big pharma companies has persons from finance and marketing departments at the helm. Hence, naturally, the aim becomes to profit from what has already been invented rather than investing more newer (even riskier) R&D projects. Since a lot of research in the industry is product-led, one can achieve only incremental research; this is not conducive to doing ambitious research needed to achieve any breakthrough, as Hassabis noted. Several studies corroborate Hassabis’ observation. One such report by innovation foundation Nesta in 2018 noted that “the exponentially increasing cost of developing new drugs is directly reflected in low rates of return on R&D spending.” Worse, the R&D returns in biopharma are on a downward slope. While DeepMind prides itself on being a heavily research-driven organisation, since its acquisition by Google in 2014 against a deal worth $500 million, the company is betting big on product management. Hassabis also said earlier that the lab wants parent organisations — Google and Alphabet — to derive adequate benefit out of the research DeepMind is doing. For example, DeepMind’s text-to-speech model that mimics human voice, WaveNet, has been embedded in many Google devices and its own product team within Google. With the launch of Isomorphic Labs, DeepMind and Alphabet seem to aim to strike a sweet spot between research and product development, concentrating particularly on biomedical research, which has emerged as a major vertical. As the company blog mentions, Isomorphic Labs will be a commercial venture, wherein the company will partner with pharmaceutical and biomedical companies. With Hassabis at the helm of affairs, DeepMind is expected to provide and facilitate devising the company’s strategy and vision. The company will be focusing on building a multidisciplinary team with experts from the areas of AI, biology, biophysics, engineering, and medicinal chemistry. Alphabet’s Next Bet The pandemic has put focus big time on fields like biomedical engineering and AI-driven drug discovery. It is not surprising that in 2020, the private investment made in these fields rose 4.5 times than the previous year. The number will grow further to reach an estimated $71 billion by 2025. Credit: Stanford Like many other big techs, Alphabet has placed great importance in drug discovery and pharmacy. By leveraging its expertise in software, data and AI, Alphabet finds itself among the top tech companies in the world, capitalising on rising opportunities in this field by getting involved in research and development, discovery, clinical development, and patient monitoring. The approach has only become more aggressive in the recent pandemic era. It is not very surprising that Alphabet has decided to go all out in its efforts in this field by launching a separate company for drug discovery. Credit: CB Insights","excerpt":"Why did Alphabet and DeepMind feel the need for a separate company dedicated to AI-driven drug discovery.","categories":["IT Services"],"tags":["DeepMind Alphafold"],"author_name":"Shraddha Goled","publish_date":"2021-11-09T16:00:00","publication_year":"2021","word_count":994,"keywords":["Go","API","AI-first","ELT","AI","Git","RAG","Aim","GAN","R","DeepMind Alphafold"],"extracted_tech_keywords":["AI","Aim","RAG","R","Go","Git","API","ELT","GAN","AI-first"],"url":"https:\/\/analyticsindiamag.com\/it-services\/why-did-alphabet-launch-a-separate-company-for-drug-discovery\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":38537,"title":"5 Reasons Why Developers Choose Kotlin Over Java","content":"In many ways, Kotlin designed by JetBrains has become the preferred language for the Android development industry. While it cannot outdo Java in terms of its popularity, many developers emphasise Kotlin will soon become the dominant platform for Android development. When it comes to its usability, Kotlin has a bunch of tools and frameworks that work with Java. Also, its advanced compiler can check the errors both during compile time and run time. It also brings down the number of lines of code significantly. Kotlin is officially supported by Google for mobile development on Android since the release of Android Studio 3.0 in October 2017. Here are 5 reasons why developers like Kotlin over Java and think that it will overtake the later: 1. Shorter program for the same task Kotlin is a statically-typed language which is very easy to read and write. It has a much simpler and shorter code than Java’s code for the same problem. As this makes the language more human-readable, it becomes easy to debug. Kotlin’s code is much smaller and streamlines the programming process, in comparison to Java. This is partly because of Kotlin’s slick IDE. 2. Easy Code Kotlin programs do not need semicolons in their program. This makes the programs easy to read and understand. They also have smart casts and string templates. Java is not a succinct language. Such a language code increases the chances of bugs. The code being in a concise language means fewer chances of both runtime and compile time errors. Kotlin gives a simple way to use mutable and immutable declarations for different data structures. 3. Java Compatibility Kotlin can easily exchange and use information from Java in a number of ways. This is one of the most powerful advantages of Kotlin. Java and Kotlin code can co-exist in the same project. Kotlin plays well with the Java programming language. Moreover, a number of Java libraries can be used in Kotlin projects, making it even more compatible. Not just the libraries but plenty of frameworks from Java are compatible with Kotlin, including some advanced frameworks. You can use Kotlin from Java without any major steps like converting an entire project to Kotlin. This is a huge advantage to developers since they do not really have to drastically learn a new language. Anyone who knows Java will be familiar with and be able to code in Kotlin. Translating Java to Kotlin is easy. In IntelliJ or Android Studio is the only thing required as a Kotlin plugin to translate Java to Kotlin. 4. Eliminating Null References: One of the biggest advantages of Kotlin over Java is the null references. This null reference, referred by Sir Tony Hoare, a British computer scientist, as The Billion Dollar Mistake. Accessing a member of a null reference result in a null reference exception. This is one of the major drawbacks of Java in which it is called a NullPointerException or NPE. Kotlin’s type system is aimed to eliminate NullPointerException from the code. The only possible causes of NPE’s may be: An explicit call to throw NullPointerException(); Usage of the !! operator that is described below; Some data inconsistency with regard to initialization, such as when: An uninitialized this available in a constructor is passed and used somewhere (“leaking this”); A superclass constructor calls an open member whose implementation in the derived class uses uninitialized state; Java interoperation: Attempts to access a member on a null reference of a platform type; Generic types used for Java interoperation with incorrect nullability, e.g. a piece of Java code might add null into a Kotlin MutableList<String>, meaning that MutableList<String?> should be used for working with it; Other issues caused by external Java code. 5. Solution To Some Of Java’s Flaws Plain old Java has a number of flaws. It inhibits the very famous problem of null pointer. Kotlin attempted to solve these hurdles created by Java. It has adopted things from a number of languages like C# and mainly from Scala, to overcome difficulties of Java. It contains instances from the language of Pascal and is considered as very influential in the development of Kotlin. Elements like parameter lists and variable declarations with the data type following a variable could also be in Kotlin. Kotlin removes the boilerplate code, greatly eliminates possible errors. It possesses features like delegations, late initializations. It also addresses type safety in lists, which is a major problem in Java. It was very easy to type a wrongly typed variable to a list before generics came along. This would typically lead to blowing up during the run time because the compiler does not detect it.","excerpt":"In many ways, Kotlin designed by JetBrains has become the preferred language for the Android development industry. While it cannot outdo Java in terms of its popularity, many developers emphasise Kotlin will soon become the dominant platform for Android development. When it comes to its usability, Kotlin has a bunch of tools and frameworks that […]","categories":["AI Trends"],"tags":["Android","code","Developers","Google","Java","Kotlin"],"author_name":"Disha Misal","publish_date":"2019-05-02T04:53:45","publication_year":"2019","word_count":773,"keywords":["Go","AWS","AI","Kotlin","cloud_platforms:AWS","ML","programming_languages:R","Scala","Android","Aim","code","Google","R","Java","Developers"],"extracted_tech_keywords":["AI","ML","Aim","AWS","R","Go","Java","Scala","cloud_platforms:AWS","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/5-reasons-why-developers-choose-kotlin-over-java\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10103257,"title":"Ilya Fired Sam Altman on Google Meet","content":"An interesting thing to note about Sam Altman’s firing from OpenAI is that Ilya Sutskever, co-founder and chief scientist at OpenAI, reportedly used Google Meet to inform him that he was being fired, with the news set to be made public very soon. Similarly, to inform Greg Brockman, about his removal from the position of the chairman of the board, Ilya sent a Google Meet link. This is ironic as Microsoft has 49% stake in OpenAI and still its members turned to Google Meet  to send a message which could completely redefine the future not just of OpenAI but of the whole tech ecosystem. This is even more funny considering Microsoft chief Satya Nadella announced at Ignite 2023 that they have added Copilot to Teams. After Altman’s departure, numerous OpenAI employees are considering whether to stay. Three senior researchers have already resigned. Meanwhile, Heng-Tze Cheng, Director of Bard Research, announced on X that both Google DeepMind and Bard Research are actively seeking new talent. Cheng stated, “We continue to be actively hiring at Google DeepMind & Bard Research. Please send me a message if you or anyone you know might be interested! Looking forward to working together to build the most helpful AI for everyone.” Google has recently postponed the release of Gemini to the first quarter of next year, suggesting challenges in its development and putting the tech giant in a catch-up position with OpenAI. Now is an opportune moment for Google to expedite the release of Gemini, particularly with potential delays in the launch of GPT-5. Meanwhile at Google pic.twitter.com\/wRYk9vKbHY— Peter Yang (@petergyang) November 18, 2023 Meanwhile, taking a dig at OpenAI employees, X shared its job search tool, which the microblogging site rolled out on the web for all users, emphasising that it is completely free to use.","excerpt":"And not on Microsoft Teams","categories":["AI News"],"tags":[],"author_name":"Siddharth Jindal","publish_date":"2023-11-18T14:59:57","publication_year":"2023","word_count":301,"keywords":["Go","OpenAI","GPT-5","AI","AI employees","GPT","llm_models:Gemini","llm_models:Bard","R","llm_models:GPT"],"extracted_tech_keywords":["AI","GPT-5","OpenAI","R","Go","GPT","AI employees","llm_models:GPT","llm_models:Gemini","llm_models:Bard"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/ilya-fired-sam-altman-on-google-meet\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10103999,"title":"Apple’s Scary New Innovation Gives Voice to the Voiceless","content":"Yesterday, Apple released a short film along with an e-book showing its new feature, Personal Voice for its devices which was unveiled earlier in May this year. This it did right before the International Day of Persons with Disabilities on December 3. Apple has always been at the forefront of making its devices accessibility friendly. The company has consistently received higher ratings for their ease of use by people with visual, hearing, and motor impairments, as well as by the elderly. It is taking it a step ahead with AI, building features like the VoiceOver, Guided Access, Door Detection, Live Listen, Point and Speak for Magnifier etc. For those at risk of speech loss, we’ve made it possible to preserve your voice on your devices so even if you can no longer speak, you can still sound like you. It’s remarkable to see the experiences this technology helps preserve, while also protecting your privacy. pic.twitter.com\/Vir3VQbhOA— Tim Cook (@tim_cook) November 30, 2023 Personal Voice was announced earlier this year, and Sarah Herrlinger, senior director, global accessibility policy & Initiatives said, “These groundbreaking features were designed with feedback from members of disability communities, to support a diverse set of users.” Although it is not speaking out about AI, it is quickly updating its features to integrate better technology. Cloning voices for healthcare has been a work in progress. Previously, patients who lost their voice due to various illnesses had to use an electrolarynx. This device needs to be placed up the throat of the patient and the vibration causes it to generate a robotic sounding voice. Companies that clone videos and images also clone voices which are used for not only all spaces of entertainment but also for healthcare. ElevenLabs, Murf.ai, Resemble AI, Respeecher etc create voice and video clones. AI for Good Using existing features, the Personal Voice add on enhances user experience further. The video released by the tech giant features Tristram Ingham, a physician, academic researcher, and disability community leader who suffers from facioscapulohumeral muscular dystrophy (FSHD). This disorder eventually leads to an inability to speak. Speaking from experience, he said that, “Historically, providers have spoken for disabled people, families have spoken for disabled people. If technology can allow a voice to be preserved and maintained, that’s autonomy, that’s self-determination.” This is possible by a combination of text-to-speech (TTS) synthesis and machine learning (ML) to create a synthetic voice that sounds like the user’s own voice. The user has to read out a series of randomly chosen text prompts aloud, providing a sample of their voice. Acoustic analysis of the voice sample extracts acoustic features such as pitch, timbre, and intonation. A text-to-speech model is trained on the user’s voice data and a large dataset of text and speech pairs. The model learns to associate acoustic features with corresponding text and generate synthetic speech that mimics the user’s voice. All of this is done within the user’s phone without running the risk of privacy invasion a characteristic Apple is most known for. The created voice can be used for calls, FaceTime and other apps. This feature works with Live Speech which was also announced around the same time. You type what you want to say, and your Personal Voice says it out loud for you. Can be used maliciously This feature that will empower so many and give them a voice has simultaneously raised security and privacy concerns, given the increasing threat posed by deepfake technologies. The internet is full of stories of unsuspecting people and companies being scammed by voice clones, cleaning their bank accounts. Is it really wise to voluntarily give your voice recording to Apple? The company in its announcement ensures that all data processing occurs locally on the device, which reduces the risk of data breaches. Access to Personal Voice generation and management is secured through biometric locks like FaceID or TouchID, and its use requires device unlocking, preventing unauthorized access. Personal Voices can be shared across devices linked to the same iCloud account and to third party apps, but there seems to be no way to transfer the voice to another device. The possibility of further safeguards, such as tracking synthetic voices for detection, could be considered to enhance security. “Although I suspect, given the company’s privacy and security focus, it may already include this feature. It would be good if the means of detection were made publicly available,” writes Matt Smallman, author and security expert on the topic. Vinod Iyengar, an AI Expert and Head of Product at Third AI is not so optimistic. “This could very quickly become a swamp of deep fakes everywhere,” he said. Voice cloning can be used to create fabricated audio content that appears to be genuine, making it harder to discern between real and fake audio recordings. This could be another level of gray area and incoming legal troubles for the future. In the meanwhile, speculations are on the rise on social media about Apple’s future direction, with suggestions that these features hint at more advanced AI integrations in upcoming products. The possibility of Apple surprising everyone with new local AI tools is discussed, indicating a shift from cloud-based to local data processing in AI. The future can have AI answer phone calls in your own voice without telling the difference! More interesting is the fact that you can use this voice through Apple’s text to speech api.In combination with an LLM\/ChatGPT and cloud based phone number this could be used to allow an AI to make and receive calls on your behalf ( in your voice ).— Kristoph (@ikristoph) May 17, 2023","excerpt":"Apple’s latest innovation, Personal Voice, unveiled just before the International Day of Persons with Disabilities, marks a significant step in voice technology.","categories":["AI Features"],"tags":[],"author_name":"K L Krithika","publish_date":"2023-12-01T19:12:22","publication_year":"2023","word_count":936,"keywords":["Go","ChatGPT","API","machine learning","AI","ML","GPT","Ray","R","llm_models:GPT"],"extracted_tech_keywords":["AI","machine learning","ML","ChatGPT","Ray","R","Go","API","GPT","llm_models:GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/apples-scary-new-innovation-brings-back-the-voiceless\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10120303,"title":"TII Releases Falcon 2 AI Model, Outperforms Meta&#8217;s Llama 3","content":"Abu Dhabi-based The Technology Innovation Institute (TII) has launched Falcon 2, a series of AI models that includes Falcon 2 11B and Falcon 2 11B VLM. These models surpass existing benchmarks. Falcon 2 11B is a text-based model, while Falcon 2 11B VLM is a vision-to-language model that can convert visual inputs into textual outputs. Both models are multilingual and open-source, providing developers worldwide with unrestricted access. Falcon 2 11B has been verified to outperform Meta’s Llama 3, which has 8 billion parameters, and performs on par with Google’s Gemma 7B model, according to evaluations by Hugging Face Leaderboard. These models are designed to run efficiently on a single GPU, making them scalable and easy to integrate into various infrastructures, from high-end servers to personal laptops. In a move to further enhance Falcon 2’s capabilities, TII plans to incorporate ‘Mixture of Experts’ (MoE). This advanced machine learning technique involves combining specialised smaller networks to improve performance by delivering more accurate and faster decision-making. The models are released under the TII Falcon License 2.0, a permissive Apache 2.0-based software license promoting responsible AI use. More information is available at FalconLLM.TII.ae. H.E. Faisal Al Bannai, secretary general of ATRC and Strategic Research and Advanced Technology Affairs Advisor to the UAE President, commented on the launch, saying, “While Falcon 2 11B has demonstrated outstanding performance, we reaffirm our commitment to the open-source movement and the Falcon Foundation.” The first Falcon model, released in 2022, established TII’s commitment to open-source AI, focusing on large language models (LLMs). It was trained on 1 trillion tokens and featured 7 billion parameters. This model builds on this foundation, enhancing capabilities.","excerpt":"Falcon 2 is a new AI model series that outperforms Meta’s Llama 3 and includes the innovative vision-to-language model, Falcon 2 11B VLM.","categories":["AI News"],"tags":["Falcon","Meta"],"author_name":"K L Krithika","publish_date":"2024-05-14T14:27:08","publication_year":"2024","word_count":273,"keywords":["Go","Hugging Face","Meta","machine learning","TPU","AI","RPA","innovation","Scala","responsible AI","Falcon","R"],"extracted_tech_keywords":["AI","machine learning","Hugging Face","TPU","R","Go","Scala","responsible AI","RPA","innovation"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/tii-releases-falcon-2-ai-model-outperforms-metas-llama-3\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":27740,"title":"Strong BI And Data Visualisation Skills Give You An Edge In Consulting Career","content":"Today, people in management and business domains have plenty to figure out when it comes to aiming for a successful business. Along with business proficiency, they also need to understand business data and the technology that goes into it because it may be crucial to uncovering much more from the business background. Looking at the core expertise for business professionals on the technology side, two core areas — business intelligence (BI) and data visualisation — have emerged as the top contenders. In this article, we will see why these are vital for any business as well as in a consulting career path. Business Intelligence For Creating Value BI is an umbrella term for a plethora of areas including data warehousing, enterprise information management, enterprise performance management, analytics, governance, risk, and compliance matters. Now BI has two aspects to it, the “business” side refers to information and issues faced by the entity (company) on the business front. The “intelligence” side refers to the technology the entity imparts in its business. Both of these aspects are critical to adding more value to the business. For a smoother working of BI, one of the key requirements is data. Without data especially from the past, it is highly unlikely to implement BI solutions in a company or any other business entity. Data Is The New Adage For Businesses In BI, past data (usually kept in databases) is the source for creating value, and retrieving it fast for time-critical situations is absolutely a top priority. This will help a business entity with better data-driven decision making. On top of this, various areas of BI such as analytics and data visualisation can help chart key business strategies and policies. For example, in order to predict sales and financial performance for a company, it is mandatory that it looks at imminent past data. How is this done? The company retrieves past data, analyses it and then present in a refined form (analytics, data visualisation etc.) so that it can uncover where it has gone wrong and bring out room for improvement. Let’s consider the use case of Mamas & Papas, a UK-based retailer which mainly manufactures baby products and clothing. Even though the company set a firm footing across the world with their stores, they had issues with respect to sales data shared in the company. There were sometimes multiple instances of same information being passed around. They had to look for a solution to this problem. That was when Strategy Companion’s BI software Analyzer helped resolve the issue. With BI features such as a better Online Analytical Processing(OLAP), on top of easy to use interface, Mamas and Papas eventually were able to handle sales data effectively across all levels of the company. As you can see, BI works extensively with analytics and metrics. This means, consulting becomes less risky with BI if done beforehand. What’s even more important here is consulting with BI can provide custom-built solutions for individual companies. So, solutions on top of more certainty align more with organisational needs. Data Visualisation For Far-Sighted Insights As BI gained traction in businesses, there came data visualisation. Although visualisation techniques existed as early as the 20th century — for example, Napoleon’s March Map by Joseph Binard or Florence Nightingale’s Crimean war mortality chart — as computing and graphics technology and evolved, the idea of simple, colourful yet effective visualisation started growing. Graphs, charts, histograms — the traditional tools in visualisation have changed a lot from just representing numerical values. Today, it is all about features such as dashboards, infographics and so on, which puts perspectives on the conference table. Business analysts now combine BI and data visualisation techniques to dig even deeper into problems that require time and thinking. In fact, it has become a preferred method to analyse complex problems for today’s business. So, how does it help business consultants? Just like BI, data visualisation can help uncover more insights from terabytes of data going into the organisation. Consultants can realise key factors such as costs, product\/service delivery etc. from these visual tools. Also, from a technical perspective, presenting information this way will make it easier for employees at all level to understand nitty-gritty details in the business and act on it coherently. In addition, both BI and data visualisation act on real-time data apart from historical data. Hence, real-time data means real-time decisions in play. Consulting on these new data can mean a lot to the business. It can sometimes tell whether the company is going to succeed or fail in the future. Again consider a sales and marketing domain, data visualisation tools such as dashboards\/scorecards can give out visual templates automatically instead of the age-old monthly reports. This is helpful because it shows real-time status along with considering older data. On top of this, it can help see trends, spikes or even anomalies. In addition, it can also be aggregated with other data for inferences on a wider perspective. Conclusion The above examples are just a scratch on the surface of possibilities. With humongous of data being processed on a large scale, BI and data visualisation can aggrandise solutions both on the business and consulting level. Apart from just increasing productivity, sales or showing areas of improvement, it can reveal opportunities. Diligent analysis from data along with understanding consumer behaviour can go a long way for the company and will keep it ahead from its rivals.","excerpt":"Today, people in management and business domains have plenty to figure out when it comes to aiming for a successful business. Along with business proficiency, they also need to understand business data and the technology that goes into it because it may be crucial to uncovering much more from the business background. Looking at the […]","categories":["AI Highlights"],"tags":["Business Intelligence","Data Analysis","Data Visualisation","information"],"author_name":"Abhishek Sharma","publish_date":"2018-08-29T06:31:54","publication_year":"2018","word_count":903,"keywords":["business intelligence","Data Analysis","Go","programming_languages:R","AI","Data Visualisation","data-driven","Aim","information","analytics","ViT","GAN","Business Intelligence","R"],"extracted_tech_keywords":["AI","analytics","Aim","R","Go","GAN","ViT","business intelligence","data-driven","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/strong-bi-and-data-visualisation-skills-give-you-an-edge-in-consulting-career\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":1759,"title":"Top 10 IoT Training Institute in India &#8211; Ranking 2017","content":"As the demand and interest around IoT has grown with time, there has also been an increase in demand for professionals with domain specific skill sets in this field. Keeping in line with our readers interest in knowing the best institutes in India that offers training in IoT, we bring to you the second edition of the ranking of IoT institutes. Right from understanding the fundamentals to getting deeper hands on expertise in hardware, these institutes have to offer it all. These institutes have been ranked after a careful consideration of various factors such as course content, pedagogy, external collaboration, placements, faculties etc. The ranking methodology, which is a result of months of research, also includes expert and student feedback as an important factor. Read through our list to pick the best IoT institute for you. 1.Cognixia Headquarter: Vadodara Cities of Operation: Global Year of Inception: 2014 Flagship IoT Program: IoT Intermediate & Advanced Level Training Program Mode of Delivery: Online, Classroom Duration: 40 – 60 Hours About the Institute: Cognixia – Technology Academy for Competency Training is an initiative started by Cognixia, a $550 MN U.S. based IT Staff Augmentation Company. It aims at providing training to qualifying professionals and students on emerging technologies like IoT, Big Data, Analytics, Cloud, Mobility and more. It has been established to bridge the demand-supply gap of professionals and aims at upgrading individual’s skills and making them future-ready for jobs that will require them to be well versed with the emerging technologies. Course Content: Their course consists of understanding the concepts, IoT reference architecture, IoT device design and protocols. It includes session on sensors, embedded development board such as Arduino, Raspberry Pi, Intel Galileo, ESP8266, wired\/wireless communication protocols, networking protocols such as OSI Reference Model, TCP\/IP, Ethernet and application protocols such as HTTP, MQTT, Overview of CoAP, XMPP, AMQP, device management, device-to-device, M2M communication, cloud computing in IoT and more. Pedagogy: The delivery approach is highly experiential with 80% content delivered in the form of hands-on exercises. Faculty: They have faculties with 100% PhD\/ \/Equivalent Industrial experience. The faculty to student ratio is 1:10. It runs IoT CoE and has a dedicated IoT research wing. External Collaborations: Collabera is an SSAE 16 Type 2 Certified Organization. They are Community Partners with Hortonworks and advantage Partners with MapR. Other Attributes: They give placement assistance with virtual labs dedicated IoT clouds. Events and LMS with 24×7 lifetime support forms an important part of the course. 2. Jigsaw Academy Headquarter: Bengaluru Cities of Operation: Worldwide, Online Year of Inception: 2011 Flagship IoT Program: IOT Professional Certificate, IOT Analyst Certificate, Full Stack IoT Expert Certificate Mode of Delivery: Online Duration of the program: 14 hours, 19 hours, 34 hours About the Institute: A brainchild of Gaurav Vohra and Sarita Digumarti, who have worked with the like of Genpact and GE Capital, Jigsaw Academy was founded in 2011. With an aim to give analytics industry a workforce that is highly skilled, Jigsaw academy work towards ensuring that professionals are nuanced in analytics and data science practices, and meets the needs of the sector. Course Content: The IoT course content by Jigsaw is extremely well structured and comprehensive. Designed by the professional IoT practitioners, the courses aim at hands-on training and explores the real world applications of IoT. It is designed to be a thorough end-to-end programme, which comes with a hardware kit that includes an Arduino and Raspberry Pi. Through the course, students can get insights on Arduino, Raspberry Pi, IoT and the cloud, IoT analytics and advanced IoT analytics. Pedagogy: The focus of Jigsaw’s pedagogy is providing practical hands-on training to its students. Hence, their IoT courses are designed in a way, that it emphasizes hands on learning by using the Arduino and the Raspberry Pi to build IoT projects. Faculty: In typical Jigsaw style, the faculty does not include professors or academicians. The course is designed and delivered by industry experts. This is a course for the industry, by the industry. External Collaborations: Their IoT course content has been designed by IoT practitioners who have extensive industry experience and understand the requirements for building a career in IoT. They have external collaborations with Harikrishna R, co-founder and director of Klar Systems, Srinivas Padmabhuni who is the President of ACM India and Urmil Parekh, co-founder of Klar Systems. Other Attributes: They have most industry relevant curriculum. Also, Jigsaw courses come with a project specific kit wherever required. Jigsaw courses are the only one in the market that cover both IoT as well as IoT analytics. 3. Nobleprog Headquarter: London, UK, Beijing, China; Noida, India Cities of Operation: Across the globe Year of Inception: 2005 Flagship IoT Program: IoT for Entrepreneurs, Managers and Investors Mode of Delivery: Online, Classroom Duration: 21 Hours About the Institute: Founded in 2005, Nobleprog offers personalized training in Artificial Intelligence, Management, IT, Statistics and Programming. As a part of consultancy they give an insight on Implementation strategy, First meetings guidance and Ad-hoc problem solving. Course Content: The course intends to introduce emerging technological options, platforms and case studies on IoT implementation in home & city automation, industrial internet, healthcare, Govt., Mobile Cellular and other areas. Course includes basic introduction of all the elements of IoT-mechanical, electronics\/sensor platform, wireless and wireline protocols, mobile to electronics integration, data-analytics and more. It also has M2M Wireless protocols for IoT- WiFi, Zigbee\/Zwave, Bluetooth, ANT+, security solutions for IoT, Open source\/commercial electronics platform for IoT-Raspberry Pi, Arduino , ArmMbedLPC, Open source \/commercial enterprise cloud platform for IoT-Ayla, iO Bridge, Libellium, Axeda. Pedagogy: With a belief in providing a worldwide comprehensive training and solutions in an effective and efficient way, they match people who can provide training and those who need it. The course structure has been kept easy, quick and effective for both trainer and participant. It covers as much hands-on as possible to expedite the knowledge absorption. Faculty: It has 78% faculties with PhD and 1:8 faculties to student ratio. The training tries to keep the participants in touch with real-time consultancy projects of industry standard and real-time problem statements so that they have an experience of real time case studies. External Collaborations: They are the Bronze Sponsors of Object Management Group, the organization responsible for setting up the BPMN, UML and SysML standards. They are also official content developers for the OCEB and OCUP, official training partners with Linux Professional Institute (LPI Certification), Google Cloud Platform for providing GCP trainings and official consulting partner with Signavio. Other Attributes: They provide placement assistance for which students need to appear for qualifying exam. Once they clear it, they are refereed to their reputed clientele. They have also launched employability program, that delivers graduate-oriented courses to help trainees be better employable. Believing in more practical and less theory, Nobleprog provides LXC containers during our remote programs. These can be accessed by both the trainer and students in a live real-time environment providing them better grasp on hands on. They also organize quarterly Hackathons. 4. CloudThat Headquarter: Bengaluru Cities of Operation: Bengaluru, Mumbai, USA, UK Year of Inception: 2012 Mode of Delivery: Classroom Duration:  16 hours per level or course, 64 hours – total About the Institute: Co-founded by Bhavesh Goswami, an ex-Amazonian and Himanshu Mody, who brings in 15 years of experience in IT Training and Consulting business, CloudThat is focused on quickly empowering IT professionals and organizations with Cloud, Big Data and IoT. With presence in Bengaluru, Mumbai, USA & UK, CloudThat provides on-site and pre-scheduled public batches in different IT centric cities of India and Overseas. Course Content: The course structure is such that it caters everyone who wants to begin with IoT- whether as a self-paced hobby or career. One of the key features of this course is to understand the concept of open source hardware and get introduced to a variety of development boards and credit card sized computers with the practical implementation of a solution. The course structure focuses on sensors, actuators, motors, etc.- which are the building blocks of any IoT project. Not only the devices and electronics, but also cloud platform for IoT and its architecture is covered along with hands on labs. Pedagogy: They believe that every course is different and needs a different approach. Therefore they ensure that participants understand the technology, tool or a concept in a way it is used in real world. Their carefully designed training kits have all valuable hardware required for hands on labs and activities. Once the students are clear with topics, they begin the hand-on labs. Faculty:  The maximum number of participants in a batch is 25 with one faculty and one assistant. External Collaborations: CloudThat is an AWS Consulting Partner, Microsoft Cloud Platform Partner, MongoDB Ready Partner and Red Hat Certified Training Partner. Other Attributes: It provides guidance for placement. 5. Frugal Labs Tech Solutions Pvt. Ltd. Headquarter: Bengaluru Cities of Operation: Pan India\/ Across the Globe Year of Inception: 2012 Flagship IoT Program: Beginners IoT Course Mode of Delivery: Online, Classroom Duration: Student Training (General) – 12 hours, Faculty Development Program – 18 to 20 hours, Corporate Training – 18 to 20 hours and Online Training- 3 Weeks About the Institute: Frugal Labs is a Tech start-up working in the field IoT as technology provider with a focus on knowledge enhancement, product development and business solutions. In 2015, they started developing FLIP (Frugal Labs IoT Platform), which is best suited for learning and rapid prototyping on IoT products. They are working on the advanced version on FLIP with more features and versatility. Course Content: It includes introduction to IoT to domain specific IoT (use case discussion) and covers FLIP Device Programming (Collecting Sensor data and sending it to cloud), FLIP Cloud, MQTT (complete overview), basic Python, IoT Application design using FLIP, FLIP Rule engine, Data management and Analysis, IoT device design (for Smart Home, Smart Agriculture, Smart Healthcare or Smart Retail & Supply chain) etc. Pedagogy: They have self paced courses with 100% hands-on learning on FLIP Hardware Kit and FLIP Cloud access. Faculty: Online faculty to student ratio is one to one whereas offline is one to 10. All their programs are taught by the core engineering team of Frugal Labs having relevant domain knowledge and experience of at least 5 years. They do not have PhD faculties as of now. External Collaborations: Few notable collaborations are on the process and the institute refused to disclose right now. Other Attributes: They provide placement assistance based on company’s requirement and criteria. They conduct hackathons. 6. hIOtron Headquarter: Pune Cities of Operation: Pune, Mumbai, Hyderabad, Bengaluru Year of Inception: 2013 Flagship IoT Program: hIOTron Certification programme on IOT Mode of Delivery: Classroom, Online Duration: 45 hours About the Institute: hIOTron is an embedded Hardware and Software company with an IoT R&D lab. Modern embedded systems being their biggest asset, their main focus is to provide services to the emerging ecosystem of the internet of things along with diverse disciplines required to bridge the gap between the cloud and devices. Course Content: Divided into four modules of four to five sessions each, the certification program helps to build complete end-to-end IoT infrastructure including wireless\/ wired sensor nodes, messaging protocols such as MQTT\/WEBSOCKET\/REST based proprietor IOT cloud platform, hardware cloud integration etc. It also includes practical case studies on multiple Industrial level IoT cloud platforms such as AWS, IBM blue-mix Watson, Thing-worx, Xively etc. in domains such as energy & industrial, smart cities, smart home & building automation, automotive & transport. Pedagogy: Aiming to lay out set of steps in IoT efforts, hIOtron provides hands-on IoT training that simplifies the complexities of the subject. Faculty: All hIOTron trainers are industrial experts in their specific domain. Faculty to student ratio 1:10 External Collaborations: hIOTron has partnered with multiple IoT companies for training & hardware software services. It is associated with more than 25 IoT companies in terms of Embedded Hardware or Software solutions, Network or front End software solutions. Other Attributes: They provide placement assistance to its students in these 25 companies along with working opportunities with hIOtron. Along with having a hardware work station, hIOtron’s LMS online program can be accessed by participants at one mbps internet speed. 7. TecHelium Info System Pvt. Ltd. Headquarter: Bengaluru Cities of Operation: Bengaluru Year of Inception: 2014 Flagship Program: IoT Training Program Mode of Delivery: Classroom Duration: Program One- Workshop – 2 Day; Program Two – Weekend program – 4 Weeks About the Institute: TecHelium is a young and fast growing company that has twin focuses, offering high quality training and building truly world-class Intellectual property (IP). With a mission to make complex technologies simple, TechHelium is helping professionals acquire job ready skills in some of the cutting edge technology domains. IoT training program is amongst their top three rated courses. Course Content: The course is centered around the Raspberry Pi prototyping board and its ecosystem. It covers the Raspberry Pi in-depth including all hardware interfaces available in Raspberry Pi, interfaces for connecting Raspberry Pi to the cloud and end to end data flows for building a showcase application around a IoT use case. Additionally, the course covers industry relevant applications, overview of cloud, middleware & security and application specific protocols. Pedagogy: The pedagogy is based upon hands-on tinkering, DIY approach to learn and build demo applications, end to end conceptualization of IoT based solution, DIY experiments to make use of imaginative thinking Faculty: They have practitioners with years of experiences who lead the training program and have one trainer for 12 students. Other Attributes: They offer placement assistance to all the students who have completed the training and have virtual labs planned which should be available in next 6 months’ time. 8. Techno Scripts Headquarter: Pune Cities of Operation: Bhopal, Ahmedabad, Indore, Raipur, Nagpur Year of Inception: 2005 Flagship Program: Advanced courses in embedded system Mode of Delivery: Online, Classroom Duration: 30-45 days About the Institute: One of the best training institutes for Advanced courses in embedded system, they come with an expertise of over 12 years. Based out of Pune, they are currently also working on live projects as per requirements of clients. With their different courses and trainings, they aim at giving good practical knowledge to students in the field of IoT. Course Content: Their course has everything from introduction to IoT, architecture of IoT to applications and industry verticals. With a focus on Raspberry Pi, they have theoretical and practical syllabus filed accordingly. Hardware description, functional schematics, cloud computing, information security are some of the things it covers. The practical has sessions on wired\/wireless networking, GPIO Interfacing, GrovePi, a hardware interfacing platform for Raspberry Pi, sensor interfacing, Raspberry Pi camera module interfacing, gaming on the Raspberry Pi 3, IBM bluemix etc. Pedagogy: The institute believes in complete practical oriented training with 20% weightage on theory and 80% on practical. Faculty: The faculties to student ratio counts to 1:10 with two PhD faculties in their team. Other Attributes: They provide 100% placement assistance till the time candidate is placed. They also have fully equipped two practical labs dedicated for the students. It is also worth noting that at least 2-3 projects that are covered in training are actually implemented on the server. 9. Nexiot Headquarter: Bengaluru Cities of Operation: India, USA, EMEA Year of Inception: 2013 Flagship Program: Nexiot SKills Accelerator (NSA) IoT programs Mode of delivery: Classroom, Online About the Institute: A global hub  for enabling industry standard IoT skills and solution development, Nexiot is an IoT and electronics innovation company. Its led by a group of experienced technocrats with expertise in electronics design-wireless sensor networks, gateway, data science-analytics and other relevant fields. Course Content: NSA provides world class certification programs in IoT starting from the basics to cutting edge. Designed in collaboration with industry leaders in  sensors, wireless- connectivity, IoT platforms- programs are offered as modules. Certified IoT systems design professional, data science & IoT foundation, a hands on approach on IoT platforms such as IBM Bluemix, AWS, Azure, Salesforce, SAP HANA, wireless technologies, Nexuino™ – ‘Beyond the Arduino’ program for custom End Node design, web application development of IoT are some of the course contents at Nexiot. Pedagogy: The program focuses on learning by hands-on exposure and solution development. These programs focus on building competency to design and implement end-to-end IoT solutions. External Collaboration: They have external collaborations with industry experts who aim at building world class industry standard programs for enabling IoT skills. 10. MCAL (Management Consulting & Advanced Learning) Headquarter: Pune Cities of Operation: Pune, Mumbai Year of Inception: 2009 Mode of Delivery: Interactive classroom sessions Duration: 3 Days, 8 hours each About the Institute: Its “Internet of Thing” training and workshop comes as first of its kind in India with an endeavour to impart skills that are likely to become industry’s new big thing in the coming years.  The program offered by MCAL is a result extensive research and gives practical and strategic understanding of vast world of Internet of things. Course Content: They have IoT training for both the beginners and professionals who wants to get a deeper understanding on IoT. The course introduces various types of devices, common aspects of their construction, communication methods and protocols used by these devices. IOT Building Blocks, HW Form factors, SW elements, cloud, analytics, wireless & connectivity standards, comparison and fit for use, decision support systems, IOT Solutions specific to day to day usage (few case studies), understanding business opportunities in smart City, security & surveillance, hardware components such as sensors, interfaces, connectivity protocol etc. are some of the course contents. Pedagogy: Includes both theoretical and practical training. Other Attributes: Provides placement assistance.","excerpt":"As the demand and interest around IoT has grown with time, there has also been an increase in demand for professionals with domain specific skill sets in this field. Keeping in line with our readers interest in knowing the best institutes in India that offers training in IoT, we bring to you the second edition […]","categories":["AI Features"],"tags":["hpe internet of things","iot enterprise architecture","IoT India","iot training india"],"author_name":"Srishti Deoras","publish_date":"2017-03-30T12:17:44","publication_year":"2017","word_count":2943,"keywords":["data science","artificial intelligence","GCP","AWS","AI","IoT India","iot enterprise architecture","iot training india","ML","cloud computing","Aim","analytics","hpe internet of things","Azure"],"extracted_tech_keywords":["AI","artificial intelligence","ML","data science","analytics","Aim","cloud computing","AWS","Azure","GCP"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/top-10-iot-training-institute-india-ranking-2017\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":45249,"title":"MachineHack Challenge: Visualizing The Pollution In India","content":"Image used for representational purpose only MachineHack is back again with a new visualization challenge. In the fourth instalment of the “Visualization is Beautiful” series, data scientists are challenged to use their creative skills to understand the pollution crisis across India. Click here to check out the hackathon. India Pollution Data Visualisation Challenge Air pollution is a critical problem all across the globe. We, humans, have ignored its consequences for so long. The natural disasters caused as a result of the dramatically changing climate, the rising temperature, untimely and fierce rains, all are consequences of our own ignorance. Trying to reduce world pollution is no easy task, and it can only be achieved by the co-operation of every single person in the world. What you as Data Scientists can do to help is to analyse the data and identify key extracts that can help governments and other organisations tackle the problem in an efficient way. In this hackathon, we provide you with an open dataset that consists of information on emissions across all the states and cities of India. We challenge all data science enthusiasts to analyse the given data and create beautiful visualizations. Click here to participate. Instructions For Participants Download the data and dig insights using advanced analytics. Present the analysis with beautiful visualisations. We are expecting gorgeous visualisations which deliver deep insights about the data and also do it using images. The more aspects of the data you cover in the visualisation, the better. You can use any platform or tool to create visualisations, just make sure the visualisation is beautiful. Only upload your submission once after you have put all your thought into the visualisation. Take part in the challenge, compete against hundreds of Data Scientists, improve your data visualisation skills, get rated for your displays and win exciting prizes. So What are you waiting for! Click here accept the challenge.","excerpt":"MachineHack is back again with a new visualization challenge. In the fourth instalment of the “Visualization is Beautiful” series, data scientists are challenged to use their creative skills to understand the pollution crisis across India. Click here to check out the hackathon. India Pollution Data Visualisation Challenge Air pollution is a critical problem all across […]","categories":["Deep Tech"],"tags":["Hackathon","Visualization"],"author_name":"Amal Nair","publish_date":"2019-08-30T17:46:10","publication_year":"2019","word_count":315,"keywords":["data science","Go","programming_languages:R","AI","programming_languages:Go","Hackathon","analytics","GAN","Visualization","R"],"extracted_tech_keywords":["AI","data science","analytics","R","Go","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/machinehack-challenge-visualizing-the-pollution-in-india\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10076450,"title":"The Javascript Framework That Solves The Annoying Feature–Speed Paradox","content":"A new javascript framework has been released. The eagerly awaited Qwik framework, created by a dream team of programmers—Miško Hevery, Manu Martinez Almeida and Adam Bradley—entered its beta testing phase last week. This is not your run-of-the-mill framework. It introduces a brand-new rendering paradigm dubbed ‘Resumability’ that completely removes the requirement for hydration, a method practically every meta-framework uses to make server-rendered websites fully interactive and understandable. Qwik’s primary purpose is to focus on the time-to-interactive measure by delaying JavaScript as much as possible in order to take advantage of the browser’s lazy loading capabilities. This is in stark contrast to existing frameworks, which approach server-side rendering and time-to-interactive as afterthoughts rather than the core aim that drives all other design decisions. The goal of Qwik is to reduce time-to-interactive, which measures the time that passes from navigating to a URL and the page becoming interactive, to the blink of an eye on even the slowest mobile device. Developers want to serve static pages to users such that they load quickly while retaining all of their interactivity. Time-to-interactive measured in time (Other frameworks Vs Qwik) Before we delve deeper into the product, let’s talk about the team. Mind over matter Miško Hevery is the creator of AngularJS, an open-source JavaScript framework that is used to build web applications. Manu Martinez Almeida has previously built ‘Gin’ (a web framework written in Golang) and ‘Stencil’, a compiler that generates Web Components. It combines the best ideas from the most popular frameworks into a straightforward build-time tool. Adam Bradley is the co-creator at Stencil. Together and individually, they are the heavy-weight programmers who aim to lift the weight of JavaScript from websites. Catch-22 situation In web development, there is a mutually conflicting situation for developers when they want to develop websites loaded with features and also provide speed on top of that. Features and speed are two conflicting situations in a website primarily because—in order to implement features, developers need to add more JavaScript but in order to make a fast-loading site, they need to use less JavaScript. For instance, while using Next.js, you initially get access to around 70 kilobytes of JavaScript. Your own application code will then add at least a few hundred additional kilobytes to that and scale off based on how much application code is present on the page. That’s because, on the initial page load, the framework needs to hydrate the dom and rebuild the entire component tree from the ground up. And every time the user refreshes the page, the main thread loads the JS again, and the user goes through the waiting game again. Astro framework recognised this problem and uses a technique called ‘partial hydration’ to selectively hydrate the dom. Resumability to the rescue However, Qwik completely eliminates hydration, as if it were not even necessary. It provides instantaneously interactive HTML. Therefore, regardless of how extensive and intricate your javascript code base is, you should be able to obtain a flawless lighthouse performance score. But how is that even possible? The main difference is that a Qwik app can be fully serialised as HTML. In other words, you can press the pause button at any time and capture all of the data and closures in the application as an HTML string. That’s huge for server-side rendering because, by the time the HTML reaches the browser, it can simply pick up where the server left off without requiring any javascript at all—which is why the term ‘Resumability’ was coined. The Qwik loader, which takes the static HTML generated from server-side-rendering and resumes it, is less than 1kb and will execute in under 1ms. The amount of code that developers need to execute is amazingly small, and it executes in less than a blink of an eye. The best part is that this code will stay constant no matter how big the application becomes. Time-to-interactive process (Other frameworks Vs Qwik) ‘Lazy loading’, which is a fundamental component of the framework, is the second element that makes this possible. It means that as you engage with the website in the smallest possible bits, all other interactive elements are slowly downloaded. The quest to find a perfect JS framework—which solves the feature and speed issue in a website—has been going on for years. Every month, there is an announcement of a new framework. It will be interesting to witness whether Qwik emerges as the JS framework that finally solves this problem for developers.","excerpt":"Qwik has introduced a new rendering paradigm dubbed “Resumability” that completely removes the requirement for hydration.","categories":["AI Trends"],"tags":[],"author_name":"Tausif Alam","publish_date":"2022-10-06T12:00:00","publication_year":"2022","word_count":744,"keywords":["Go","programming_languages:R","AI","ML","Aim","ViT","JavaScript","R","Java","programming_languages:JavaScript"],"extracted_tech_keywords":["AI","ML","Aim","R","JavaScript","Go","Java","ViT","programming_languages:R","programming_languages:JavaScript"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/the-javascript-framework-that-solves-the-annoying-feature-speed-paradox\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":43740,"title":"A Day In The Life Of: A Data Scientist Who Loves Analytics &#038; Fantasy Fiction Equally","content":"In our weekly column ‘A Day In The Life Of’, we are trying to step into the shoes of awesome techies from various organisations and sectors who are working in emerging tech areas like big data, data analytics, artificial intelligence, machine learning and the internet of things, among others. This week Analytics India Magazine got in touch with Vijendra Pratap Singh, a Data Scientist who works at Hashtag Loyalty, to talk about his enthusiasm for accurate data analysis to understand consumer behaviour, his obsession with hill stations and his love for fantasy fiction. Singh begins his day at around 7 am and reaches his office after a long 2.5-hour struggle with traffic. He says that this traffic situation is the only way that his work affects his personal life in any way. He adds that one of the good things about Hashtag Loyalty is that they have flexible work hours. Even with the 9-hour work schedule and long traffic jams, Singh says he finds time for one of his guilty pleasures — reading fantasy fiction. From Harry Potter to The Lord Of The Rings, Singh finds solace and inspiration in these exciting characters. “During my vacations, I like travelling to hill stations. It is one of my life’s greatest pleasures,” says Singh happily. Singh is of the opinion that having an active and engaging social life is important for his work as well. “I consider my lifestyle as active. I usually hang out with my friends and meet new people. It keeps me updated and fresh,” he says. “My company helps offline businesses capture consumer and transaction data via loyalty as well as feedback. Once the data is entered into our platform, we provide businesses with a tool that enables them to interact with the customer and helps them connect more with the product or service,” says Singh said. Talking about his work, he says, “As a data scientist, I am involved in both researching as well as coding for my firm. The company works with businesses for building customer loyalty, improving retention and increasing revenue. The clients provide with raw data about their customers’ transactions and I then analyse this data. I have to make sure that the right recommendations are given to the client, in hopes that this helps the client increase sales. Interestingly, Singh was a part of the Maker Program at BridgeLabz through which he secured his current job. Right now, Singh is working on creating APIs. “I am continuously analysing and researching the data so that the web team can work on applications. In the past, I have worked on forecasting, transactional and recommendations system. Currently, I am working on email subject line optimisation,” he says. Singh says that one of the things that gives him great joy and inspiration is researching and digging into the technology that he is working on “I try to help and integrate artificial intelligence in a way that will benefit my company, too,” he adds. His goals right now are to hone his skills in data science so that he showcases the company again that he is a valued asset to be reckoned with. “I plan to do my best right now at the company as it will help me grow and understand technology even better,” he says, signing off.","excerpt":"In our weekly column ‘A Day In The Life Of’, we are trying to step into the shoes of awesome techies from various organisations and sectors who are working in emerging tech areas like big data, data analytics, artificial intelligence, machine learning and the internet of things, among others. This week Analytics India Magazine got […]","categories":["AI Features"],"tags":["Data Science Career","Data Scientist","Interviews and Discussions"],"author_name":"Prajakta Hebbar","publish_date":"2019-08-02T15:00:29","publication_year":"2019","word_count":551,"keywords":["big data","data science","Go","API","artificial intelligence","machine learning","AI","Data Science Career","analytics","GAN","Data Scientist","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","data science","analytics","R","Go","API","big data","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/a-day-in-the-life-of-a-data-scientist-who-loves-analytics-fantasy-fiction-equally\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10094196,"title":"How Whatfix is Revolutionising SaaS with GenAI Integration","content":"Generative AI has changed SaaS businesses completely. Everyone, including the likes of Zoho, Salesforce, and Freshworks, is getting into the generative AI gamble – and so is Whatfix. With the help of a 12-member team split into two groups, the company is endlessly working towards AI implementation across its products and functions with a focus on responsible AI. “We are reimagining the future of digital adoption platform with Whatfix AI,” said Vara Kumar, co-founder and chief tech and product officer, in an interview with AIM. Kumar said that Whatfix employs generative AI to boost user productivity through application-agnostic use cases, supporting web, desktop, and mobile applications for customers and employees. “It auto-completes text fields, provides detailed explanations for user notes, and offers summarised answers in self-help searches, saving time and effort,” added Kumar. Interestingly, team Whatfix looks to prioritise a “human-in-the-middle” approach, ensuring human control over AI and contextual relevance. Whatfix is developing a computer vision model for human-like comprehension of applications and processes, enabling task completion without relying on technology or knowledge availability. Founded by Khadim Batti and Kumar in 2014, the California-headquartered company introduced three new tools in 2022: Product Analytics, Nudgealong, and Comments. Product Analytics allows organisations to track and analyse user engagement and behaviour data effortlessly. Nudgealong democratises knowledge sharing, while Comments enable real-time collaboration. Whatfix obtained two patents for automatic segmentation and element detection. Read more: What is Microsoft Without OpenAI? From Data to Performance: A Holistic Approach to AI Deployment According to Gartner’s report, only 54% of AI projects make it from pilot to production. The deployment of ML models is a global challenge for many enterprises due to common challenges like the collection and preparation of data, development of the model, deployment to production, and monitoring of its performance. Lack of resources and expertise, as well as a non-innovative culture, can further hinder the deployment process. However, Whatfix has found a solution to this problem. “At Whatfix, we speed up AI\/ML project deployment by integrating AI engineers into R&D teams, aligning everyone on the benefits and metrics of new features, and providing transparency into algorithms and ML processes,” Kumar explained. To make this possible, Whatfix leverages cutting-edge technology. It uses Azure Machine Learning services throughout the entire machine learning lifecycle, including development, deployment, inference, and MLOps. When dealing with compute-intensive and inference-heavy deep learning models, they make use of GPU resources. Additionally, Whatfix employs Looker dashboards to monitor model performance and track key metrics. Furthermore, Whatfix uses Transformer and GPT-based architectures for various functionalities. One notable example is their ability to generate concise answers for end-user queries by extracting information from vast repositories of content and guidance materials. Read more: Is Indian Govt’s Battle Against AI Disinformation Flawed? Navigating the Ethical AI Maze The tech industry has always been unstable in its opinion to regulate AI. In the past three years, Microsoft fired its entire ethics and society team, following Google’s dismissal of its ethical AI team leader Tinmit Gebru. Even Meta disbanded its Responsible Innovation (RI) team. Earlier this month, OpenAI CEO Sam Altman testified before the US Congress regarding AI regulation. While Altman supported the development of robust AI regulations by the international community, he also expressed concern about the potential consequences of actual regulatory implementation. So in a scenario like this, it is important for companies to take the necessary steps. Echoing similar lines, Kumar said, “We heavily use GenAI-based models, some of which have hallucination capabilities. However, we prioritise user service by ensuring zero hallucination, guaranteeing accurate information delivery.” Whatfix does not share customer data with third-party services to improve their models or services. The quality of predictions can vary based on input data. We continuously monitor model performance in production and provide active feedback to maintain prediction quality and accuracy. Whatfix Growth Story India’s SaaS industry is poised to boost the country’s position as a leading service provider and simplify global business processes. It is expected to drive transformation and maximise the potential of the IT industry. By 2030, India aims to generate $50 to $70 billion in SaaS revenues and capture 4-6% of the global market, creating up to $1 trillion in value. Currently, India has 1,000 SaaS startups, including 10 unicorns, with a combined annual revenue of $2-3 billion and employing 40,000 people. The number of unicorns is predicted to increase tenfold by 2030. With an anticipated $10 billion SaaS industry and an eight percent global market share, India’s SaaS sector is growing at a compounded rate of 18%, with a projected 36% growth in the small and medium business sectors. One of the key players in this market is Whatfix. In terms of revenue, last year, Whatfix achieved an impressive year-on-year (YoY) Annual Recurring Revenue (ARR) growth rate of 65%, making it one of the standout performers in its field. Additionally, the company maintained an outstanding net revenue retention rate of 127% for its enterprise customers, demonstrating substantial growth within its existing customer base throughout the year. The average revenue generated per account experienced a noteworthy rise of 32%, with an impressive 40% of accounts experiencing expansion. The California headquartered company also expanded its worldwide reach in the Asia-Pacific (APAC) region by establishing a new office in Sydney, Australia. Whatfix is backed by the likes of SoftBank, Sequoia Capital, Cisco Investments and more. In 2021, it raised a whopping $90 million through its Series D funding round, led by SoftBank Vision Fund 2. This funding round also witnessed active participation from notable investors such as Eight Roads Ventures, Sequoia Capital India, Dragoneer Investment Group, F-Prime Capital, and Cisco Investments. Earlier, the company had secured $32 million in Series C funding, with Sequoia Capital India leading the charge and receiving support from existing investors like Eight Roads Ventures, a renowned proprietary investment firm supported by Fidelity, F-Prime Capital, and Cisco Investments. Currently, Whatfix has raised an impressive $139.8 million to date. Considering the impeccable reputation Whatfix has established and its substantial investments in generative AI, it appears evident that this “emerging unicorn” is on the brink of achieving the coveted status of a billion-dollar unicorn. Read more: Uncensored Models Outperform Aligned Language Models","excerpt":"With the help of a 12-member team split into two groups, Whatfix is working towards generative AI implementation across its products and functions with a focus on responsible AI.","categories":["AI Features"],"tags":["ChatGPT","Generative AI","Interviews and Discussions"],"author_name":"Shritama Saha","publish_date":"2023-05-31T11:49:58","publication_year":"2023","word_count":1027,"keywords":["ChatGPT","GenAI","machine learning","OpenAI","AI","ML","MLOps","computer vision","deep learning","analytics","generative AI","Generative AI","Interviews and Discussions"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","computer vision","analytics","generative AI","GenAI","OpenAI","MLOps"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-whatfix-is-revolutionising-saas-with-generative-ai-integration\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":6,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10101394,"title":"Future-Proofing Careers in the Age of Automation","content":"The rise of generative AI has sparked extensive debates regarding the potential displacement of jobs. Whether it’s ChatGPT affecting roles in coding and human resource management, or DALLE and Midjourney’s influence on the field of graphic design, this discourse has been ongoing for some time. Raghav Gupta, co-founder & CEO at Futurense Technologies, in his talk titled, ‘Future-Proofing Careers in the Age of Automation’, during the ongoing Cypher2023 event, which is India’s largest AI conference, delves into the same discourse and talks mentions that AI should be seen as an opportunity rather than a threat. “Today, I’d like to discuss a topic that’s been on many of our minds and a subject we’ve been hearing and talking about extensively. I won’t offer a highly technical perspective; instead, I’ll provide a pragmatic and emotional approach to navigating job displacement in the age of AI,” Gupta said. Impact on Futurense Futurense Technologies is a company that offers recruitment, talent transformation, and career acceleration services. They provide a platform for talent to unlock their true potential. Gupta says his customer service team would take nearly 1800 calls a day with clients as the first process of screening. “However, now we have added a layer of AI bot that makes sure the first level of screening happens through the bot. So now the average calls that the team does has decreased from 1900 to roughly 750 to 800,” Gupta said. Similarly, Futurense’s marketing team previously hired content writers and graphic designers to do different jobs. “Now, we have one specialist who knows how to use ChatGPT and Midjourney. The same individual can be a content writer and a graphic designer in an organisation.” “While we had 18 members in our marketing team, now we have 21, but the amount of content we are dispensing has quadrupled. We are also increasing our output, which in turn, is on average increasing the number of jobs.” AI will create newer jobs Gupta states eventually everyone will have to adapt to AI, but those early adopters will have an advantage. While there might be some job displacements initially, AI will also create newer jobs. For insurance, an AI trainer and operator. “AI is as good as the model it is trained on and this training of that model is a continuous process. Till now only very big multibillion-dollar organisations were training the model, but we will see small businesses start having these jobs where they will train the data, the model.” “AI’s predictive abilities are remarkable, but they occasionally offer broad predictions that may not precisely fit your specific business needs. This calls for individuals or a team who could be likened to modern-age McKinsey consultants. So, Gupta said we will have an AI prediction auditor, an individual or a group of individuals who will possess a unique skill set, enabling them to grasp the intricacies of your business context while also comprehending the subtleties of macro socio-economic developments.” In conclusion, Gupta encourages everybody to embrace AI as part of their professional journey, and soon they will realise that it’s not a threat but a tremendous opportunity. While AI adoption will eventually become widespread, those who adapt quickly and effectively will reap the greatest rewards.","excerpt":"Gupta states eventually everyone will have to adapt to AI, but those early adopters will have an advantage.","categories":["AI Features"],"tags":[],"author_name":"Pritam Bordoloi","publish_date":"2023-10-12T12:26:54","publication_year":"2023","word_count":537,"keywords":["Go","ChatGPT","TPU","AI","RAG","GPT","automation","generative AI","GAN","R"],"extracted_tech_keywords":["AI","generative AI","ChatGPT","RAG","TPU","R","Go","GPT","GAN","automation"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/cypher2023-future-proofing-careers-in-the-age-of-automation\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":68856,"title":"Recurrent Neural Network in PyTorch for Text Generation","content":"There are a variety of interesting applications of Natural Language Processing (NLP) and text generation is one of those interesting applications. When a machine learning model working on sequences such as Recurrent Neural Network, LSTM RNN, Gated Recurrent Unit is trained on the text sequences, they can generate the next sequence of an input text. PyTorch provides a set of powerful tools and libraries that add a boost to these NLP based tasks. It not only requires a less amount of pre-processing but also accelerates the training process. In this article, we will train a Recurrent Neural Network (RNN) in PyTorch on the names belonging to several languages. After successful training, the RNN model will predict names belonging to a language that start with an input alphabet letter. Implementation in PyTorch This implementation was done in Google Colab where the dataset was fetched from the Google Drive. So, first, we will mount the Google drive with the Colab notebook. from google.colab import drive drive.mount('\/content\/gdrive') Now, we will import all the required libraries. from __future__ import unicode_literals, print_function, division from io import open import glob import os import unicodedata import string import torch import torch.nn as nn import random import time import math import matplotlib.pyplot as plt import matplotlib.ticker as ticker The below code snippet will read the dataset. all_let = string.ascii_letters + \" .,;'-\" n_let = len(all_let) + 1 def getFiles(path): return glob.glob(path) # Unicode string to ASCII def unicodeToAscii(s): return ''.join( c for c in unicodedata.normalize('NFD', s) if unicodedata.category(c) != 'Mn' and c in all_let ) # Read a file and split into lines def getLines(filename): lines = open(filename, encoding='utf-8').read().strip().split('\\n') return [unicodeToAscii(line) for line in lines] # Build the cat_lin dictionary, a list of lines per category cat_lin = {} all_ctg = [] for filename in getFiles('gdrive\/My Drive\/Dataset\/data\/data\/names\/*.txt'): categ = os.path.splitext(os.path.basename(filename))[0] all_ctg.append(category) lines = getLines(filename) cat_lin[categ] = lines n_ctg = len(all_ctg) In the next step, we will define the module class to generate the names. The module will be a recurrent neural network. class NameGeneratorModule(nn.Module): def __init__(self, inp_size, hid_size, op_size): super(NameGeneratorModule, self).__init__() self.hid_size = hid_size self.i2h = nn.Linear(n_ctg + inp_size + hid_size, hid_size) self.i2o = nn.Linear(n_ctg + inp_size + hid_size, op_size) self.o2o = nn.Linear(hid_size + op_size, op_size) self.dropout = nn.Dropout(0.1) self.softmax = nn.LogSoftmax(dim=1) def forward(self, category, input, hidden): inp_comb = torch.cat((category, input, hidden), 1) hidden = self.i2h(inp_comb) output = self.i2o(inp_comb) op_comb = torch.cat((hidden, output), 1) output = self.o2o(op_comb) output = self.dropout(output) output = self.softmax(output) return output, hidden def initHidden(self): return torch.zeros(1, self.hid_size) The below functions will be used to pick the random item from a list and a random line from a category def randChoice(l): return l[random.randint(0, len(l) - 1)] def randTrainPair(): category = randChoice(all_ctg) line = randChoice(cat_lin[category]) return category, line The below functions will convert the data to the compatible format for the RNN module. def categ_Tensor(categ): li = all_ctg.index(categ) tensor = torch.zeros(1, n_ctg) tensor[0][li] = 1 return tensor def inp_Tensor(line): tensor = torch.zeros(len(line), 1, n_let) for li in range(len(line)): letter = line[li] tensor[li][0][all_let.find(letter)] = 1 return tensor def tgt_Tensor(line): letter_indexes = [all_let.find(line[li]) for li in range(1, len(line))] letter_id.append(n_let - 1) # EOS return torch.LongTensor(letter_id) The below function will create random training examples including category, input and target tensors. def rand_train_exp(): category, line = randTrainPair() category_tensor = categ_Tensor(category) input_line_tensor = inp_Tensor(line) target_line_tensor = tgt_Tensor(line) return category_tensor, input_line_tensor, target_line_tensor The below function will define the loss criteria for the RNN module. #Loss criterion = nn.NLLLoss() #Learning rate lr_rate = 0.0005 def train(category_tensor, input_line_tensor, target_line_tensor): target_line_tensor.unsqueeze_(-1) hidden = rnn.initHidden() rnn.zero_grad() loss = 0 for i in range(input_line_tensor.size(0)): output, hidden = rnn(category_tensor, input_line_tensor[i], hidden) l = criterion(output, target_line_tensor[i]) loss += l loss.backward() for p in rnn.parameters(): p.data.add_(p.grad.data, alpha=-lr_rate) return output, loss.item() \/ input_line_tensor.size(0) To show time during training, the bellow function is defined. def time_taken(since): now = time.time() s = now - since m = math.floor(s \/ 60) s -= m * 60 return '%dm %ds' % (m, s) In the next step, we will define our RNN model. model = NameGenratorModule(n_let, 128, n_let) We will see the parameters of the defined RNN model. print(model) In the next step, the model will be trained in 10,000 epochs. epochs = 100000 print_every = 5000 plot_every = 500 all_losses = [] total_loss = 0 # Reset every plot_every iters start = time.time() for iter in range(1, epochs + 1): output, loss = train(*rand_train_exp()) total_loss += loss if iter % print_every == 0: print('Time: %s, Epoch: (%d - Total Iterations: %d%%),  Loss: %.4f' % (time_taken(start), iter, iter \/ epochs * 100, loss)) if iter % plot_every == 0: all_losses.append(total_loss \/ plot_every) total_loss = 0 We will visualize the training loss. plt.figure(figsize=(7,7)) plt.title(\"Loss\") plt.plot(all_losses) plt.xlabel(\"Epochs\") plt.ylabel(\"Loss\") plt.show() Finally, we will sample our model to test it on generating the names belonging to languages when given a starting alphabet letter. max_length = 20 # Sample from a category and starting letter def sample_model(category, start_letter='A'): with torch.no_grad():  # no need to track history in sampling category_tensor = categ_Tensor(category) input = inp_Tensor(start_letter) hidden = NameGenratorModule.initHidden() output_name = start_letter for i in range(max_length): output, hidden = NameGenratorModule(category_tensor, input[0], hidden) topv, topi = output.topk(1) topi = topi[0][0] if topi == n_let - 1: break else: letter = all_let[topi] output_name += letter input = inp_Tensor(letter) return output_name # Get multiple samples from one category and multiple starting letters def sample_names(category, start_letters='XYZ'): for start_letter in start_letters: print(sample_model(category, start_letter)) Now, we will check the sampled model to generate names when given a language and the starting alphabet letter. print(\"Italian:-\") sample_names('Italian', 'BPRT') print(\"\\nKorean:-\") sample_names('Korean', 'CMRS') print(\"\\nRussian:-\") sample_names('Russian', 'AJLN') print(\"\\nVietnamese:-\") sample_names('Vietnamese', 'LMT') So, as we can see above, our model has generated names belonging to the language categories and starting with the input alphabet. References:- Trung Tran, “Text Generation with Pytorch”. “NLP from scratch: Generating names with a character level RNN”, PyTorch Tutorial. Francesca Paulin, “Character-Level LSTM in PyTorch”, Kaggle.","excerpt":"In this article, we will train a Recurrent Neural Network (RNN) in PyTorch on the names belonging to several languages. After successful training, the RNN model will predict names belonging to a language that start with an input alphabet letter.","categories":["Deep Tech"],"tags":["Natural Language Processing","NLP","Pytorch"],"author_name":"Dr. Vaibhav Kumar","publish_date":"2020-07-03T12:00:00","publication_year":"2020","word_count":965,"keywords":["Pytorch","Go","machine learning","TPU","AI","neural network","Natural Language Processing","PyTorch","NLP","Colab","Matplotlib","R"],"extracted_tech_keywords":["AI","machine learning","neural network","NLP","PyTorch","Colab","Matplotlib","TPU","R","Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/recurrent-neural-network-in-pytorch-for-text-generation\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10089342,"title":"The Wait for GPT-4 is Finally Over","content":"Microsoft-backed OpenAI’s highly anticipated GPT-4 is finally here with the promise to give more accurate and safer responses. Microsoft confirmed that the new Bing search is running on GPT-4. For now, GPT-4 is available on ChatGPT Plus and as an API for developers. here is GPT-4, our most capable and aligned model yet. it is available today in our API (with a waitlist) and in ChatGPT+.https:\/\/t.co\/2ZFC36xqAJit is still flawed, still limited, and it still seems more impressive on first use than it does after you spend more time with it.— Sam Altman (@sama) March 14, 2023 Multimodal in Nature As AIM had predicted, GPT-4 stands out from GPT 3.5 due to its multimodal nature. This AI model can receive textual prompts and images, which allows users to specify any type of vision or language-related task, similar to the text-only setting. Read more: ChatGPT Takes NEET; Will it Pass with Flying Colors or Flunk? Through this feature, it can generate natural language, code, and other text outputs by processing inputs that consist of both images and text. GPT-4 showcases comparable capabilities in various domains, including documents containing photographs, diagrams, or screenshots, just as it does with text-only inputs. Additionally, it can also be enhanced using test-time techniques designed for text-only language models, such as few-shot or chain-of-thought prompting. However, it is important to note that this image input capability is still a research prototype and not yet publicly available. Microsoft’s intention to make GPT-4 multimodal has been very clear. In the recent paper on Visual ChatGPT, the “prompt manager” is described as a tool to share information between foundation models, such as Stable Diffusion, ControlNET, BLIP, and ChatGPT. Additionally, Microsoft has released a research paper that focuses on Kosmos-1, a multimodal large language model (MLLM) that emphasizes the integration of language, action, and multimodal perception. Key Features According to OpenAI’s blog post, besides being multimodal, GPT-4 can solve difficult problems with greater accuracy, thanks to its broader general knowledge and problem-solving abilities, surpassing ChatGPT in its advanced reasoning capabilities. GPT-4 has been trained on Microsoft Azure AI supercomputers, which offer an AI-optimized infrastructure enabling the delivery of the product globally. The blog post also stated that they have put considerable effort into making GPT-4 safer and more aligned, resulting in an 82% reduction in its likelihood to produce disallowed content and a 40% increase in its ability to provide factual responses when compared to GPT-3.5. The MMLU benchmark, which comprises 14,000 multiple-choice questions covering 57 subjects, was translated into various languages using Azure Translate disclosing that GPT-4 outperformed GPT-3.5 and other LLMs (such as Chinchilla and PaLM), in 24 out of 26 languages tested, including for low-resource languages such as Latvian, Welsh, and Swahili. Despite these improvements, OpenAI acknowledges GPT-4’s limitations, including social biases, hallucinations, and adversarial prompts. The company encourages transparency, user education, and AI literacy as society adopts these models while striving to expand the input avenues of those who shape their models. Greg Brockman, president, and cofounder of OpenAI will be live streaming at 1:30 am IST for a developer demo showcasing GPT-4 and some of its capabilities and limitations.","excerpt":"GPT-4 is available on ChatGPT Plus and as an API for developers.","categories":["AI News"],"tags":["ChatGPT","Google","GPT-4","Microsoft","OpenAI"],"author_name":"Shritama Saha","publish_date":"2023-03-14T23:14:04","publication_year":"2023","word_count":522,"keywords":["ChatGPT","TPU","OpenAI","AI","R","ML","GPT-4","RAG","Aim","Google","foundation models","Azure","Microsoft"],"extracted_tech_keywords":["AI","ML","foundation models","ChatGPT","OpenAI","Aim","RAG","Azure","TPU","R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/the-wait-for-gpt-4-is-finally-over\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10134300,"title":"Channel-Specific and Product-Centric GenAI Implementation in Enterprises Leads to Data Silos and Inefficiencies","content":"Organisations often struggle with data silos and inefficiencies when implementing generative AI solutions. This affects over 70% of enterprises today, but global software company Pegasystems, aka Pega, seems to have cracked the code by using its patented ‘situational layer cake’ architecture. This approach democratises the use of generative AI across its platform, allowing clients to seamlessly integrate AI into their processes. They can choose from any LLM service provider, including OpenAI, Google’s Vertex AI, and Azure OpenAI Services, thereby ensuring consistent and efficient AI deployment across all business units. “Our GenAI implementation at the rule type levels allows us to democratise the use of LLMs across the platform for any use case and by mere configuration, our clients can use any LLM service provider of their choice,” said Deepak Visweswaraiah, vice president, platform engineering and side managing director at Pegasystems, in an interaction with AIM. Pega vs the World Recently, Salesforce announced the launch of two new generative AI agents, Einstein SDR Agent and Einstein Sales Coach Agent, which autonomously engage leads and provide personalised coaching. This move aligns with Salesforce’s strategy to integrate AI into its Einstein 1 Agentforce Platform, enabling companies like Accenture to scale deal management and focus on complex sales. Salesforce integrates AI across all key offerings through its unified Einstein 1 Platform, which enhances data privacy, security, and operational efficiency via the Einstein Trust Layer. “We have generative AI capabilities in sales cloud, service cloud, marketing cloud, commerce cloud, as well as our data cloud product, making it a comprehensive solution for enterprise needs,” said Sridhar H, senior director of solution engineering at Salesforce. SAP’s generative AI strategy, on the other hand, centres around integrating AI into core business processes through strategic partnerships, ethical AI principles, and enhancing its Business Technology Platform (BTP) to drive relevance, reliability, and responsible AI use across industries. “We are adding a generative AI layer to our Business Technology Platform to address data protection concerns and enhance data security,” stated Sindhu Gangadharan, senior VP and MD of SAP Labs, underscoring the company’s focus on integrating AI with a strong emphasis on security and business process improvement. Oracle, on the other hand, focuses on leveraging its second-generation cloud infrastructure, Oracle Cloud Infrastructure (OCI). It is designed with a unique, non-blocking network architecture to support AI workloads with enhanced data privacy while extending its data capabilities across multiple cloud providers. “We’re helping customers do training inference and RAG in isolation and privacy so that you can now bring corporate sensitive, private data…without impacting any privacy issue,” said Christopher G Chelliah, senior vice president, technology & customer strategy, JAPAC at Oracle. Meanwhile, IBM has watsonx.ai, an AI and data platform designed to help companies integrate, train, and deploy AI models across various business applications. IBM’s generative AI strategy with watsonx.ai differentiates itself by offering extensive model flexibility, including IBM-developed (Granite), open-source (Llama 3 and alike), and third-party models, along with robust client protection and hybrid multi-cloud deployment options. At the same time, Pega focuses on deeply integrating AI within its platform to streamline business processes and eliminate data silos through its unique situational layer cake architecture. Pega told AIM that it distinguishes itself from its competitors by avoiding the limitations of the traditional technological approaches, which often lead to redundant implementations and data silos. “In contrast, competitors might also focus more on channel-specific designs or product-centric implementations, which can lead to inefficiencies and fragmented data views across systems,” said Visweswaraiah. Situational Layer Cake Architecture Pega told AIM that its approach to integrating GenAI processes into business operations is distinct due to its focus on augmenting business logic and decision engines rather than generating code for development. It employs the situational layer cake architecture, which as a part of Pega’s exclusive centre-out architecture, helps to adapt microjourneys for different customer types, lines of business, geographies, and more. “Our patented situational layer cake architecture works in layers making specialising a cinch, differentiating doable, and applying robust applications to any situation at any time, at any scale,” said  Visweswaraiah. He added that enterprises can start with small, quick projects that can grow and expand over time, ensuring they are adaptable and ready for future challenges. In addition to this, the team said it has the ‘Pega Infinity’ platform, which can mirror any organisation’s business by capturing the critical business dimensions within its patented situational layer cake. “Everything we build in the Pega platform, processes, rules, data models, and UI is organised into layers within the situational layer cake. This means that you can roll out new products, regions, or channels without copying or rewriting your application,” shared  Visweswaraiah. He further said that the situational layer cake lets you declare what is different and only what is different into layers that match each dimension of your business. Simply put, when a user executes the application, the Pega platform slices through the situational layer cake and automatically assembles an experience that is tailored exactly to that user’s context. Visweswaraiah believes that this architecture has given them a great opportunity to integrate GenAI into the platform at the right layers so it is available across the platform.","excerpt":"Pega employs ‘situational layer cake’, which, as a part of its exclusive centre-out architecture, helps adapt microjourneys for different customer types, lines of business, geographies, and more.","categories":["Deep Tech"],"tags":["Generative AI"],"author_name":"Aditi Suresh","publish_date":"2024-09-03T13:11:20","publication_year":"2024","word_count":858,"keywords":["Go","GenAI","OpenAI","AI","R","ML","RAG","Aim","generative AI","Generative AI","Azure"],"extracted_tech_keywords":["AI","ML","generative AI","GenAI","OpenAI","Aim","RAG","Azure","R","Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/channel-specific-and-product-centric-genai-implementation-in-enterprises-leads-to-data-silos-and-inefficiencies\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10089486,"title":"Microsoft Releases Polyglot Notebooks, Multi-Language Extension for VS Code","content":"Microsoft published a blog, announcing the release of Polyglot Notebooks, a Visual Studio Code’s multi-language notebook extension. It is now available in the VC Code Marketplace. Polyglot Notebooks allow users to use various programming languages in one notebook with full language server support. Moreover, they enable users to share variables between languages, allowing for a seamless workflow. Users can avoid switching between different tools and manually transferring data, which can interrupt their workflow. The Polyglot Notebooks currently support: C# F# PowerShell JavaScript HTML* Mermaid* SQL KQL (Kusto Query Language) https:\/\/twitter.com\/code\/status\/1636056741119926276 This interconnection enables several benefits for users. Connecting and querying Microsoft SQL server databases and Kusto clusters is now possible by natively writing in SQL or KQL. Moreover, there is no need for wrapper libraries for using multiple languages in the same notebook. Developers can natively write code in their preferred language. You can also share your variables to another language without jumping from tool to tool. Polyglot Notebooks within VS Code leverage .NET Interactive, an innovative engine developed using .NET technology. This engine can execute various programming languages and allow for variable sharing between them. Since it can act as a kernel within the context of notebooks, it enhances the overall experience of using Polyglot Notebooks. While the Polyglot Notebooks extension in Visual Studio Code is now generally available, but the .NET Interactive APIs that power it are still in preview.","excerpt":"Now you can share your variables across languages without jumping from tool to tool","categories":["AI News"],"tags":["Jupyter Notebook","Visual Studio Code","VS Code"],"author_name":"Mohit Pandey","publish_date":"2023-03-16T15:13:48","publication_year":"2023","word_count":232,"keywords":["API","Jupyter Notebook","programming_languages:R","AI","ML","JavaScript","programming_languages:SQL","RAG","VS Code","SQL","R","Java","Visual Studio Code"],"extracted_tech_keywords":["AI","ML","RAG","R","SQL","JavaScript","Java","API","programming_languages:R","programming_languages:SQL"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/microsoft-releases-polyglot-notebooks-multi-language-extension-for-vs-code\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10172277,"title":"Titans Space Industries Selects Andhra’s Jahnavi Dangeti for Space Mission in 2029","content":"Jahnavi Dangeti, a 23-year-old engineer from Andhra Pradesh, has been chosen as an Astronaut Candidate (ASCAN) for the Titans Space Astronaut Class of 2025. Dangeti’s selection marks the beginning of an intensive multi-year training program to prepare for space missions set to advance science and human exploration. Dangeti announced this on her LinkedIn in a post saying, “As a child, I often looked up at the moon believing it was following me. That sense of wonder never left—and today, I’m honoured to share that it’s becoming a part of my reality.” She will begin astronaut training in 2026, which will include learning spacecraft systems, mission simulations, zero-gravity flights, emergency procedures, psychological assessments, survival training, medical evaluations, and spaceflight operations. The mission will include multiple Earth orbits, sustained weightlessness, and a transformative environment for research and learning. The first orbital mission, scheduled for 2029, will be led by veteran NASA astronaut William McArthur Jr., currently the chief astronaut at Titans Space. The mission will include multiple Earth orbits, sustained weightlessness, and opportunities for scientific research. “Thank you to Titans Space Industries Inc. and Neal S. Lachman for this once-in-a-lifetime opportunity,” Dangeti added. She expressed her gratitude to her mentors, friends, and supporters who have helped her reach this milestone. Dangeti’s selection follows her completion of NASA’s International Air and Space Program (IASP) at the Kennedy Space Centre, where she became the first Indian to achieve this feat. She was also appointed as the mission director for ‘Team Kennedy’, leading a group of 16 individuals from various countries. Her team successfully launched and landed a miniature rocket into the sky. In addition to her training at NASA, Dangeti participated in the Analog Astronaut Training Centre (AATC) in Krakow, Poland, becoming the youngest to complete the programme in 2022. The Titans Space Astronaut Class of 2025 aims to contribute to the next era of space exploration by preparing astronauts for human spaceflight and scientific research in microgravity. Note: The article has been updated to clarify that Jahnavi Dangeti has been selected as an astronaut candidate with Titans Space Industries.","excerpt":"The first space orbital mission, scheduled for 2029, will be led by veteran NASA astronaut William McArthur Jr.","categories":["AI News"],"tags":["Aerospace","NASA","space india"],"author_name":"Sanjana Gupta","publish_date":"2025-06-24T14:22:31","publication_year":"2025","word_count":346,"keywords":["NASA","programming_languages:R","AI","space india","Aim","ViT","Aerospace","R"],"extracted_tech_keywords":["AI","Aim","R","ViT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/titan-space-industries-selects-andhras-jahnavi-dangeti-for-space-mission-in-2029\/","complexity_score":2,"technical_depth":5,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10039797,"title":"Why Are CAPTCHAs Getting Harder To Crack?","content":"If you’ve spent enough time on the internet, chances are you’ve encountered an ‘are-you-a-human’ checkpoint. These checkpoints, aka CAPTCHAs (Completely Automated Public Turing test), help tell computers from humans. From a small box asking whether you’re human to identifying picture squares with traffic lights, CAPTCHAs have become testy over the years. Multiple failed CAPTCHA attempts can really set your teeth on edge. CAPTCHAs were first developed by a team of researchers at Carnegie Mellon University for Yahoo in 2000 to curb automated programmes from rapidly generating free email accounts. Technology is all the more significant in today’s digital world. In 2019, cybersecurity firm HUMAN (formerly White Ops) found that over a 90-day period, several prominent organizations lost more than $30 million combined to marketing fraud. CAPTCHAs set up barriers only humans should get past and are commonly used on most online platforms. What’s the catch So, why would these tests become harder? In a fun twist, a CAPTCHA is an elegant tool to train AI. In 2014 Google tested CAPTCHA solving ability in its machine learning algorithm against humans. The test used hard-to-decipher distorted texts, and the computer got the test right 99.8 percent of the time while the humans called it right 33 percent of the time. From here, Google introduced the ‘No CAPTCHA reCAPTCHA’. This involves the click of the ‘I’m not a robot’ button for some and image identification for others, based on observations of user data and behaviour. However, with machines catching up again, designing CAPTCHA has become a whack-a-mole game. Additionally, a puzzle that is too difficult for bots might end up being hard for humans to solve as well. In fact, many people find CAPTCHAS irritating because they have become more challenging for humans than machines. This does not make AI more intelligent than us—it simply makes us, well, human. Human beings are not machines. They are diverse in terms of education, language, skills, etc., and the point of a universal CAPTCHA is to have a test every human can pass. Workarounds Unfortunately, we still need some form of CAPTCHA (hopefully, a better type of it in the near future) to keep the bots and malware at bay. Imagine going to a website for a limited offer or a giveaway for a product you have been waiting months for only to find it sold out and then finding out that half those clicks were made by artificial intelligence (who will now be receiving these free goods). Hence, we need a system to keep the intrusive machines out. There have been some attempts at that, including game CAPTCHAs, which require users to rotate objects or move puzzle pieces. The trick is that the game will not specify instructions, thus stumping AI, but humans will intuitively know what to do from the context (e.g. they’ll recognise the game board). For now, this sounds quite annoying—and imagine having to play a puzzle you don’t recognise or know how to play? Other proposals to replace CAPTCHA include trivia CAPTCHAs or those based on nursery rhymes native to the region the user grew up. Some attempts were made at identifying people based on ethnicity, gender and facial expression. In 2010, a study designed CAPTCHAs prompting users to segment and index rock art, i.e. human markings on stone. In another interesting case, Amazon received a patent in 2017 for a scheme where the only way through is by getting the wrong answer. Researchers are also looking at doing away with CAPTCHAs altogether. Shuman Ghosemajumder, chief technology officer of the bot-detection company Shape Security, said he preferred detecting human behaviour in users instead. For instance, monitoring someone’s interactions with their mouse (a bot would either not use one or be very precise in it) would allow us to decipher who is human. This sounds a little dicey from a surveillance standpoint and, given that bots have now learned to mimic typos and the conversational skills of someone who does not speak English very well, it may ultimately not be too useful. As of now, CAPTCHAs remain integral to our use of the internet. We might need to find a better alternative quickly, but given the many factors behind choosing a suitable CAPTCHA, it might be a while before we see something foolproof.","excerpt":"If you’ve spent enough time on the internet, chances are you’ve encountered an ‘are-you-a-human’ checkpoint. These checkpoints, aka CAPTCHAs (Completely Automated Public Turing test), help tell computers from humans. From a small box asking whether you’re human to identifying picture squares with traffic lights, CAPTCHAs have become testy over the years. Multiple failed CAPTCHA attempts […]","categories":["IT Services"],"tags":["Turing test"],"author_name":"Mita Chaturvedi","publish_date":"2021-05-09T16:00:00","publication_year":"2021","word_count":711,"keywords":["Go","API","machine learning","artificial intelligence","programming_languages:R","AI","programming_languages:Go","Turing test","Git","GAN","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","R","Go","Git","API","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/why-are-captchas-getting-harder-to-crack\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":32086,"title":"VC funding in AI increased 450% In The Last 5 Years, Says The AI Index 2018 Report","content":"Yoav Shoham, Chair of AI Index Team The mission of the AI Index is to ground the conversation about AI in data. Talking about their efforts, the team in the report says, “The AI Index is an effort to track, collate, distill, and visualize data relating to artificial intelligence. It aspires to be a comprehensive resource of data and analysis for policymakers, researchers, executives, journalists, and the general public to develop intuitions about the complex field of AI.” Its is interesting to note the growing importance of China in the world of AI. China was mentioned 69 times in 94 page report and India was mentioned barely 5 times in the report. The AI Index is prepared by Yoav Shoham who chairs the committee and works at Stanford University. The rest of the team includes, Raymond Perrault from SRI International, Erik Brynjolfsson from MIT (Massachusetts Institute Of Technology), Jack Clark from OpenAI, James Manyika from McKinsey Global Institute, Juan Carlos Niebles from Stanford University, Terah Lyons from Partnership On AI and John Etchemendy from Stanford University and Barbara Grosz from Harvard University. AI Index puts out major insights into the world of Artificial Intelligence and the report has four sections: Data: Volume of Activity and Technical Performance Other measures: Recent Government Initiatives, Derivative measures, and Human-Level Performance Discussion: What’s Missing? Appendix Here are the key takeaways of the report: 1.AI Research Papers Publishing Increased The report says that Europe has consistently been the largest publisher of AI papers. The share of Europe’s research paper is 28% of all AI papers on Scopus in 2017. The report says that the number of papers published in China increased 150% between 2007 and 2017. The number of Chinese papers increased despite the spike and drop in Chinese papers around 2008. There is also a suggestion, that AI publishing is growing because of heightened interest in computer science. More findings in the report suggest that 56 percent of papers came from the the Machine Learning and Probabilistic Reasoning category. The same number was only 28% in 2010. The most remarkable observation was that Neural Networks research paper had a compound annual growth rate of 3% from 2010—2014, followed by a CAGR of 37% from 2014—2017. 2. Investments And Gender In AI The report says that from January 2015 to January 2018, the amount of AI startups increased 2.1x and the number of AI startups has seen exponential growth. Interestingly, from 2013 to 2017, the report also observed that VC funding in AI increased 4.5x, which is huge. Naturally, ML is the largest skill cited as a requirement in job openings. Lately, deep learning (DL) is growing at the fastest rate which increased 35x. The report also shed light on the gender disparity in artificial intelligence. The report observed men make up 71% of the applicant pool in the U.S. In 2014 30% of all AI patents came out of the U.S followed by South Korea. 3. Open Source And Public Perception Of AI The popularity of various AI-programming frameworks can be measured by the stars on its Github repo. The recent trend observed is that frameworks backed by major companies are really popular. Some examples are Tensorflow (Google), Pytorch (Facebook), mxnet (Amazon). The team also showed that AI articles in the media have become less neutral and more positive. Articles have been 12% positive in January 2016 and went to 30% positive in July 2016. The percentage of positive articles has sticked to nearly 30% since then, the report observed. The mention of the term “Artificial Intelligence” is used more in comparison of “Machine Learning”. The report said that in the U.S. words like Machine Learning or Artificial Intelligence was said at least once in a given event or a discussion. In United Kingdom’s data indicates that Machine learning or Artificial intelligence was said at least once in a given comment online. 4. Technical Performance The report says that ImageNet accuracy has seen large improvements in training time and accuracy. Machine translation has also improved consistently, BLEU scores from English to German are 3.5x greater today than they were in 2008, the report said. The performance in question answering has improved also. The report said that the performance in 2018 has gone from 63% to 69% on the Easy Set for question answering and from 27% to 42% on the Challenge Set for question answering. The technical performances and training times have constantly improved pointing out the obvious fact that artificial intelligence is revolutionising technology and is essentially the core of computer science.","excerpt":"The mission of the AI Index is to ground the conversation about AI in data. Talking about their efforts, the team in the report says, “The AI Index is an effort to track, collate, distill, and visualize data relating to artificial intelligence. It aspires to be a comprehensive resource of data and analysis for policymakers, […]","categories":["AI Features"],"tags":["AI (Artificial Intelligence)","Deep Learning"],"author_name":"Abhijeet Katte","publish_date":"2018-12-25T04:25:41","publication_year":"2018","word_count":759,"keywords":["machine learning","artificial intelligence","OpenAI","AI","neural network","PyTorch","ML","Ray","deep learning","Deep Learning","TensorFlow","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","neural network","OpenAI","Ray","TensorFlow","PyTorch"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/vc-funding-in-ai-increased-450-in-the-last-5-years-says-the-ai-index-2018-report\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10170044,"title":"India’s Defence Drone Industry Gains Momentum, But Regulatory Challenges Persist","content":"India’s defence drone sector took centre stage following the recent tensions between India and Pakistan. The successful deployment of indigenously developed drones during ‘Operation Sindoor’ to neutralise terror networks underscored their strategic importance. However, the sector continues to face hurdles, including regulatory ambiguity, import restrictions, and limits on technology transfer, said retired Major General Mandip Singh, in an email interview with AIM. He’s also the president of strategic alliance at Droneacharya Aerial Innovation Ltd. Singh pointed to government initiatives like the Make in India campaign, the Atmanirbhar Bharat mission, and regulatory frameworks, such as the Drone Rules 2021 and the Defence Production and Export Promotion Policy (DPEPP) 2020, as gradually creating a more conducive environment for Indigenous manufacturers and encouraging local production and innovation. “Further policy support is needed for R&D, skilling of our vast human resource, ease of procurement processes, and recognition for indigenous content,” he added. According to a Grand View Research report, India’s military drone market is projected to reach $4,082 million by 2030, growing at a compound annual growth rate (CAGR) of 17.9% from 2025 to 2030. As of 2024, India holds a 3.8% share of the global military drone market. Lessons from Indo-Pak border tensions The recent border clashes underscored the critical role of drones in surveillance, situational awareness, and rapid response, Singh said. “For startups, this reaffirms the need to focus on reliable, high-performance drones that can operate in challenging terrain and provide real-time intelligence,” Singh said.  “It reminds them that the future of modern warfare is increasingly reliant on cutting-edge drone technology.” Another key takeaway is the imperative for indigenous equipment and domestic supply chains—no nation can afford reliance on external sources during conflict. Since the tensions, Droneacharya has accelerated the integration of artificial intelligence into its projects. “Intelligent drones with lethal precision neutralised enemy targets with zero collateral damage in this brief skirmish. This is set to become the new standard,” Singh noted. Counter-drone solutions Addressing the rising demand for counter-drone systems, Singh said startups are innovating rapidly to meet defence requirements. “This includes early detection, jamming, and neutralisation systems to secure critical airspace,” he said. With the Atmanirbhar Bharat and Make in India programmes, the industry is increasingly leveraging locally sourced components and indigenous manufacturing, aligning with the government’s push for a robust domestic defence ecosystem. Singh also warned that India’s northern neighbour remains the world’s largest producer, seller, and user of dual-use drones, which serve both commercial and military purposes. R&D and strategic partnerships India’s push for indigenous drone capabilities dates back to the 1990s with DRDO’s early UAV projects. National campaigns and updated regulations have since accelerated sector growth. “Drones can be significant job creators and economic growth drivers,” the Ministry of Civil Aviation noted in 2021, adding that India has the potential to become a global drone hub by 2030. The Drone Rules 2021 also proposed establishing a Drone Promotion Council, which would bring together academia, startups, and stakeholders to foster an innovation-friendly regulatory environment. Singh revealed that the Droneacharya’s R&D efforts are focused on enhancing drone durability, range, and real-time data processing, alongside AI and machine learning integration for precise threat detection. The company is also developing heavy-lifting drones and logistics drones. DroneAcharya is actively exploring partnerships with both domestic and global technology leaders to strengthen capabilities and co-develop advanced defence-grade drones that align with the government’s Make in India initiative and address emerging security threats to international standards. When asked about investor interest following the border tensions, Singh said the situation has sparked renewed focus on defence technologies and mission-critical drones, positively impacting investor sentiment. “We hope this momentum leads to policy reforms that incentivise MSMEs and SMEs to contribute to building a resilient drone ecosystem,” he concluded.","excerpt":"India’s military drone market is projected to reach $4,082 million by 2030, growing at a compound annual growth rate (CAGR) of 17.9% from 2025 to 2030.","categories":["AI Features"],"tags":[],"author_name":"C P Balasubramanyam","publish_date":"2025-05-15T16:22:04","publication_year":"2025","word_count":622,"keywords":["Go","API","machine learning","artificial intelligence","AI","innovation","RAG","Aim","R","startup"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","Aim","RAG","R","Go","API","innovation","startup"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/indias-defence-drone-industry-gains-momentum-but-regulatory-challenges-persist\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10171009,"title":"IIT Kharagpur, Singapore’s IME Sign MoU for Semiconductor Research","content":"The Indian Institute of Technology, Kharagpur and Singapore’s Institute of Microelectronics (IME), under the Agency for Science, Technology and Research (A*STAR), have signed a Memorandum of Understanding (MoU) to collaborate on semiconductor research and talent development. The agreement was formalised during SEMICON Southeast Asia 2025. This partnership aligns with India’s broader objectives under the India Semiconductor Mission, which aims to strengthen India’s position in the global semiconductor landscape through international cooperation and innovation. The MoU establishes a framework for joint research in areas including post-Complementary Metal-Oxide-Semiconductor (CMOS) technologies, advanced transistor designs, heterogeneous integration, chip packaging, AI-driven hardware accelerators, quantum devices, photonic systems, thermal management, and reliability diagnostics. Anandaroop Bhattacharya, professor at IIT Kharagpur’s Department of Mechanical Engineering, stated, “This marks a significant step in IIT Kharagpur’s global ambition to drive semiconductor research and aligns well with India’s current thrust on semiconductors through the India Semiconductor Mission (ISM).” It also encompasses bilateral exchange programmes, specialised training, and joint workshops aimed at equipping researchers, engineers, and students with skills relevant to the evolving semiconductor sector. Yee-Chia Yeo, deputy chief executive at A*STAR and a professor at NUS, highlighted the importance of collaborative efforts, noting that Singapore’s advancements in the semiconductor domain are rooted in partnerships across research, industry, and government. This partnership also reflects India and Singapore’s shared vision of building a resilient, innovative, and skilled semiconductor ecosystem through international cooperation.","excerpt":"The MoU establishes a framework for joint research in advanced designs, chip packaging, AI, quantum, photonic systems, and more.","categories":["AI News"],"tags":["iit Kharagpur","quantum","research in India","semiconductor","Semiconductor India"],"author_name":"Sanjana Gupta","publish_date":"2025-05-30T17:32:05","publication_year":"2025","word_count":229,"keywords":["Go","Rust","research in India","programming_languages:R","iit Kharagpur","AI","innovation","semiconductor","programming_languages:Go","RAG","quantum","Aim","Semiconductor India","R","programming_languages:Rust"],"extracted_tech_keywords":["AI","Aim","RAG","R","Go","Rust","innovation","programming_languages:R","programming_languages:Go","programming_languages:Rust"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/iit-kharagpur-singapores-ime-sign-mou-for-semiconductor-research\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10125319,"title":"CP Gurnani Proves Altman Wrong, Tech Mahindra Builds Indian LLM Under $5M","content":"In a keynote speech at the MachineCon GCC Summit 2024, CP Gurnani, co-founder of AIonOS and former CEO of Tech Mahindra, challenged the notion that India could not develop its own LLMs and highlighted India’s rapid advancements in AI. “Sam Altman challenged everybody that India will never be able to have an LLM,” Gurnani said, referring to OpenAI’s CEO. “I spoke to my chief innovation officer that time at Tech Mahindra… Six hours later, he says I have a plan,” he added. Gurnani revealed that Tech Mahindra was able to develop an Indian LLM for local languages and 37+ dialects in just 5 months, spending less than $5 million. “When ET covered the 7 best AI companies, which will have an Indian LLM, they covered Tata, NVIDIA, and Tech Mahindra,” he said. He further added, “Tata, NVIDIA, now you can assume what the budget would look like. And Tech Mahindra, I am happy to share that they spent less than 5 million dollars on what Sam Altman said, India will never be able to deliver.” Grand Vision for India’s AI & Tech Landscape The tech veteran argued that AI is now mainstream and part of our daily lives, but questioned whether enterprises have fully adopted the technology. However, he cited statistics suggesting AI could improve productivity by up to 40% in areas like customer experience, quality assurance, and sales. Further, Gurnani confidently stated that the country looks confident to make strides in the tech and AI landscape and truly be Atmanirbhar. “I’m very convinced that in India, as the semiconductor industry develops, 5-7 years later, we would not be looking at someone else.” As a comparative measure, Gurnani gave an Indian example and emphasised the importance of “frugal plus innovation plus technology plus people leadership” as the secret to success for Indian companies like IndiGo and Airtel in their competition with behemoths like Jio. He also added that his own new venture, AIonOS, aims to transform industries like travel, logistics, hospitality and transportation using AI. How is Project Indus Going to Fair? Incidentally, Tech Mahindra launched its Project Indus today. This indigenous large language model (LLM) is designed to engage in conversations across a wide array of Indic languages and dialects, marking a significant step towards linguistic inclusivity in the realm of artificial intelligence. In the initial phase, Project Indus will concentrate on developing an LLM specifically tailored for the Hindi language and its extensive range of over 37 dialects. By focusing on this widely spoken language, Tech Mahindra aims to bridge the linguistic gap and empower a substantial portion of the population to interact with cutting-edge AI technology in their native tongue. To bring Project Indus to life, Tech Mahindra has forged strategic collaborations with industry leaders Dell Technologies and Intel. The model will be implemented using ‘GenAI in a Box’ solution, which harnesses the power of Dell Technologies’ state-of-the-art computing, storage, and networking capabilities. Additionally, the solution incorporates Intel-based infrastructure, including Intel® Xeon® Processors and the OneAPI software suite, ensuring optimal performance and scalability. Nikhil Malhotra, Global Head of Makers Lab at Tech Mahindra, expressed his enthusiasm for the project, stating, “Project Indus is our seminal effort to develop an LLM from the ground up. Through Makers Lab, our R&D arm, we created a roadmap, collected data from the Hindi-speaking population, and built the Indus model. Our collaboration with Dell Technologies & Intel will help deliver cutting-edge AI solutions that enable enterprises to scale at speed.” The collaboration between Tech Mahindra, Dell Technologies, and Intel aims to revolutionise AI-driven solutions across various industries. By leveraging Tech Mahindra’s expertise in developing localized and verticalized industry-agnostic LLMs, the partnership will enable the creation of tailored use cases and applications in sectors such as healthcare, rural education, banking, agriculture, and telecom. Denise Millard, Chief Partner Officer at Dell Technologies, emphasized the significance of accessibility and scalability in the adoption of GenAI. “With the Dell AI Factory, LLMs like Project Indus leverage AI-optimized technologies with an open ecosystem of partners, validated and integrated solutions, services and best practices, accelerating the adoption of AI to drive growth, optimize productivity and promote innovation,” she remarked. Santhosh Viswanathan, Vice President & Managing Director for India at Intel, highlighted the company’s dedication to advancing the frontiers of AI. “We are proud to collaborate with Tech Mahindra on Project Indus, which will enable seamless deployment of advanced AI models across industries and empower enterprises to unlock GenAI’s full potential for enhanced operational efficiency and a competitive edge,” he stated. Project Indus represents a pivotal milestone in the evolution of the global GenAI landscape, which is projected to expand to a staggering $1.3 trillion over the next decade. The model will initially prioritise key use cases and pilot projects, offering scalable AI solutions to enterprises. Conclusively, as India strives for sovereignty and leadership in the AI landscape with AI initiatives, time will tell how Gurnani’s daring challenge and the product that came out of it will fare.","excerpt":"He also said that India will develop its own NVIDIA in the next 5-7 years and would not need to be dependent on someone else.","categories":["IT Services"],"tags":["project indus","Tech Mahindra"],"author_name":"Shyam Nandan Upadhyay","publish_date":"2024-06-28T18:59:59","publication_year":"2024","word_count":829,"keywords":["Tech Mahindra","artificial intelligence","GenAI","OpenAI","AI","ML","RAG","Ray","Aim","edge AI","R","project indus"],"extracted_tech_keywords":["AI","artificial intelligence","ML","GenAI","OpenAI","Aim","Ray","edge AI","RAG","R"],"url":"https:\/\/analyticsindiamag.com\/it-services\/cp-gurnani-proves-altman-wrong-tech-mahindra-builds-indian-llm-under-5m\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10118428,"title":"LLM Systems Will Soon Have Infinite Context Length","content":"LLMs forget. Everyone knows that. The primary culprit behind this is the finity of context length of the models. Some even say that it is the biggest bottleneck when it comes to achieving AGI. Soon, it appears that the debate over which model boasts the largest context length will become irrelevant. Microsoft, Google, and Meta, have all been taking strides in this direction – making context length infinite. The end of Transformers? While all LLMs are currently running on Transformers, it might soon become a thing of the past. For example, Meta has introduced MEGALODON, a neural architecture designed for efficient sequence modelling with unlimited context length. MEGALODON aims to overcome the limitations of the Transformer architecture, such as its quadratic computational complexity and limited inductive bias for length generalisation. The model demonstrates superior efficiency at a scale of 7 billion parameters and 2 trillion training tokens, outperforming other models such as Llama 2 in terms of training loss. It introduces key innovations such as the complex exponential moving average (CEMA) component and timestep normalisation layer, which improve long-context pretraining and data efficiency. These improvements enable MEGALODON to excel in various tasks, including instruction fine-tuning, image classification, and auto-regressive language modelling. Most likely, upcoming Meta’s Llama 3 will be based on MEGALODON architecture, making it infinite context length. Similarly, Google researchers have introduced a method called Infini-Attention, which incorporates compressive memory into the vanilla attention mechanism. The paper titled ‘Leave No Context Behind’ says that Infini-Attention incorporates compressive memory into the vanilla attention mechanism and combines masked local attention and long-term linear attention mechanisms in a single Transformer block. This approach combines masked local attention and long-term linear attention mechanisms in a single Transformer block, allowing existing LLMs to handle infinitely long contexts with bounded memory and computation. The approach scales naturally to handle million-length input sequences and outperforms baselines on long-context language modelling benchmarks and book summarisation tasks. The 1B model, fine-tuned on up to 5K sequence length passkey instances, successfully solved the 1M length problem. Forgetting to forget Along similar lines, another team of researchers from Google introduced Feedback Attention Memory (FAM). It’s a novel Transformer architecture that leverages a feedback loop to enable the network to attend to its own latent representations, fostering the emergence of working memory within the Transformer and allowing it to process infinitely long sequences. The introduction of FAM offers a new approach by adding feedback activations that feed contextual representation back into each block of sliding window attention. This enables integrated attention, block-wise updates, information compression, and global contextual storage. Besides, researchers from Beijing Academy of AI introduced Activation Beacon, a method that extends LLMs’ context length by condensing raw activations into compact forms. This plug-in component enables LLMs to perceive long contexts while retaining their performance within shorter contexts. Activation Beacon uses a sliding window approach for stream processing, enhancing efficiency in training and inference. By training with short-sequence data and varying condensing ratios, Activation Beacon supports different context lengths at a low training cost. Experiments validate Activation Beacon as an effective, efficient, and low-cost solution for extending LLMs’ context length. Do we even need tokens? In February, Microsoft Research published the paper titled ‘LongRoPE: Extending LLM Context Window Beyond 2 Million Tokens’. The technique significantly increases the context length of LLMs to an unprecedented 2048k tokens, while preserving their original performance within shorter context windows. Moving beyond that, another team of Microsoft researchers have challenged the traditional approach to LLM pre-training, which uniformly applies a next-token prediction loss to all tokens in a training corpus. Instead, they propose a new language model called RHO-1, which utilises Selective Language Modeling (SLM). The SLM approach directly addresses this issue by focusing on the token level and eliminating the loss of undesired tokens during pre-training. SLM first trains a reference language model on high-quality corpora to establish utility metrics for scoring tokens according to the desired distribution. Tokens with a high excess loss between the reference and training models are selected for training, focusing the language model on those that best benefit downstream applications. No more ‘lost in the middle’? There has been a long-going conversation about how longer context length window models have the problem of getting lost in the middle. Opting for smaller context-length inputs is recommended for accuracy, even with the advent of long-context LLMs. Notably, facts at the input’s beginning and end are better retained than those in the middle. Jim Fan from NVIDIA AI explains how claims of a million or billion tokens are not helpful when it comes to improving LLMs. “What truly matters is how well the model actually uses the context. It’s easy to make seemingly wild claims, but much harder to solve real problems better,” he said. Meanwhile, to measure the efficiency of these longer context lengths, NVIDIA researchers developed RULER, a synthetic benchmark designed to evaluate long-context language models across various task categories, including retrieval, multi-hop tracing, aggregation, and question answering. All of this just means that the future LLM systems would have infinite context length.","excerpt":"Microsoft, Google, and Meta, have all been taking strides in this direction – making context length infinite.","categories":["AI Features"],"tags":["LLMs"],"author_name":"Mohit Pandey","publish_date":"2024-04-18T10:00:00","publication_year":"2024","word_count":843,"keywords":["Go","stream processing","AI","LLMs","ML","Transformers","SLM","RAG","Aim","transformer architecture","R"],"extracted_tech_keywords":["AI","ML","Aim","Transformers","SLM","RAG","R","Go","stream processing","transformer architecture"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/llm-systems-will-soon-have-infinite-context-length\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":66393,"title":"Microsoft Replaces Journalists With AI. Can We Rely On AI For News?","content":"With the advancements in the field of artificial intelligence, many sectors have been in fear of losing human employees over this advanced technology. And with the rise of machines amid this crisis for business continuity, the fear has started looming in the journalism industry where media houses are publishing automated news for their publications. In fact, Bloomberg News, one of the leading media publishing houses, has claimed that the company has been using automated technology for publishing one-third of their news content on their platform. According to news reports: Editor-in-chief John Micklethwait of Bloomberg News stated in their company memo a few years back, “I think automation is crucial to the future of journalism in a much broader way than many of us realise. Bloomberg already uses automation for customised news and trending stories …” Also, in the recent news, Microsoft has announced laying off a considerable number of journalists from their MSN in order to replace them with artificial intelligence. These laying off people included contracted news producers, who have been curating news for MSN homepage to which, one of the terminated contractors said to the media, “It’s been semi-automated for a few months, but now it’s full speed ahead. It’s demoralising to think machines can replace us but there you go.” The automated machine will use algorithms to identify trending news stories and curate them from their publishing partners for their MSN website. Similar to human employees, it will also rewrite the headlines or add better photographs if required and make slide shows. The Australian edition of The Guardian media has also made headlines with their first robot-assisted article — “Political donations plunge to $16.7m – down from average $25m a year.” A lot of this could be attributed to the pandemic outbreak where media and publishing houses, similar to businesses from other sectors, are also relying on emerging technologies to keep up with the face-pacing landscape. Microsoft has been one of many tech giants that are leveraging robot journalism to cut costs. Joining the initiative, Google has also recently announced the power of machine learning for journalism. With their project — Journalism AI, the company has been urging journalists to take up their free training course to understand this advanced technology and how it can impact the industry. So, with media publishing houses relying on artificial intelligence and news bots as one of their vital resources to report the news, does this create a threat among journalists to lose their jobs? Can artificial intelligence replace critical thinking in journalism? Many reports have stated that automation will replace jobs, whether it be in the BFSI industry, healthcare or the core industries of the country. Therefore, many think that AI can also replace reporters and journalists in writing and reporting news. This is possible as companies are building advanced machines that can understand language and use data to replicate the type of content humans produce. Media houses like The Washington Post and The New York Times have been using news bots to report and write the news. In fact, The Washington Post has been using a bot called Heliograf since some time to publish news for their publication; the bot helped the website successfully report the Rio Olympics and election in 2016. Although over time, machines are learning to report news like a human, it cannot have the feature of critical thinking that is vital in the field of journalism to find out the truth in a story. Alongside, AI also has a limitation of understanding empathy, which again is an essential ingredient in reporting stories in newspapers. In reality, the core of journalism involves good storytelling of incidents that require asking the right questions and creative writing, and that’s where human intervention makes complete sense. Therefore, it is still being believed that artificial intelligence can only replace humans in areas of mundane and repetitive tasks — which in the journalism industry has been to curate trending news for websites. And that’s why the lay off for the company Microsoft has only been in the areas where humans were being used to curate news stories from other organisations; the company has still retained journalists who work on reporting original stories. Chief data officer & founder of Applied XLabs, Francesco Marconi, who also recently wrote a book on the future of journalism, told in an interview that — artificial intelligence would automate a minimum of only 12% of reporters’ job which is mundane and repetitive. This will transform the role of editors and journalists as media houses would be able to add more value to their content and create more long-form and investigative journalism. Additionally, the technology of artificial intelligence has always been scrutinised for being biased in making decisions in various businesses, and it, therefore, cannot be relied upon for curating unbiased news for publications, which is the core of journalism ethics. In fact, in a recent study by experts from Cardiff University and MIT have stated that artificially intelligent robots have possibilities of developing biases and prejudices on their own. According to Professor Whitaker from Cardiff University noted that “It is feasible that autonomous machines with the ability to identify with discrimination and copy others could in future be susceptible to prejudicial phenomena that we see in the human population.” Therefore, it is believed that, although artificial intelligence will influence the future of journalism but will not replace humans. In fact, it will aid journalists in streamlining their workflow, curating trending news, and can also help in expediting research for journalists by analysing tons of raw data. Furthermore, automated machines can also help broadcast reporters with transcribing interviews, audios and videos for faster production of content. Artificial intelligence will save journalists instead of replacing them On the contrary, AI has potential benefits to enhance journalism for the audience and also in providing support to reporters in being more productive by addressing repetitive tasks for them. Alongside, an artificially intelligent system can provide necessary resources from silos at one place for journalists to write more in-depth analytical stories at a much faster pace. Organisations like Quint, Vox, as well as BBC, are also using AI-based tools to find trending news for their publications in order to double their traffic. In fact, in recent news, Yong Huang, the director-general of one of the most influential media organisations in China — Xinhua News Agency, stated in a release that, “The use of artificial intelligence in preparing news reports is still at the earliest stage of its development, but we are convinced that it will seriously influence the development of journalism and will help news agency employees to do their work faster and better.” He further believes that the advanced technology will help in routine tasks of preparing reports, fact-checking and analysing data, which in turn can help in improving core journalist work. Alongside, with continuous data being generated, these bots can analyse the data and help media houses in enhancing the content to maximise the engagement of the audience. Condé Nast, a global media publication, has been using an advanced machine called Spire that has been designed to help the publication in predicting their customer purchases and in turn helping in enhancing their readers’ experience. “Spire, our next-gen precision audience targeting platform gives us the ability to follow our consumers and directly connect advertisers with them that they’re most interested in, at the earliest point of their purchasing decisions – when they’re discovering what they want to buy through our content,” said Pamela Drucker Mann from Condé Nast. On the other hand, BBC has also been using a news aggregation and content extraction API — Juicer — that helps the publication with tagging their news articles with relevant tags for journalists to find latest stories on particular topics quickly. Similar to this, to tag and annotate stories, The New York Times also launched their tool called — Editor — which uses machine learning to tag articles for reporters usage. Advanced systems like these can aid reporters in expediting their research process and cover more grounds in a shorter span of time. These systems also help in fact-checking for journalists, which again is critical to report news authentically. Furthermore, in today’s era, with information overload, it gets challenging to differentiate wrong information for the verified ones. Artificial intelligence can also help journalists in identifying fake news as well as deepfakes to fight against manipulated media, by verifying news items with a possible number of sources in reduced time than humans. Last year, Reuters, one of the leading multimedia news platforms, in collaboration with Facebook Journalism Project, has announced their courses for their reporters and journalists to understand the world of deepfakes. Besides identifying fake news and deepfakes, AI can also help in moderating contents for publication, which typically gets curated from third-party sources, and can also help in keeping a check of notorious behaviour in the comments section at a larger scale. Media houses have complex work processes, with the help of advanced AI, media employees can streamline their process of gathering information and producing better content for their readers. Wrapping Up There indeed would be a massive transformation in the journalism industry in the post-pandemic world, and therefore it would be imperative for the media organisations to keep up their pace with technological advancements and rethink ways to produce better and faster content for their audience. And, with media organisations relying on AI and ML tools for generating, creating and curating news, it would be imperative for reporters and journalists to embrace the technologies in their work process.","excerpt":"With the advancements in the field of artificial intelligence, many sectors have been in fear of losing human employees over this advanced technology. And with the rise of machines amid this crisis for business continuity, the fear has started looming in the journalism industry where media houses are publishing automated news for their publications. In […]","categories":["Global Tech"],"tags":["ai publications by india","how does ai create jobs","latest technological advancements","Layoffs","Layoffs by IT companies","media analytics","Microsoft","recent technological advancements"],"author_name":"Sejuti Das","publish_date":"2020-06-02T12:01:32","publication_year":"2020","word_count":1597,"keywords":["how does ai create jobs","media analytics","recent technological advancements","Go","artificial intelligence","machine learning","Layoffs","AI","API","ML","RAG","Layoffs by IT companies","Aim","latest technological advancements","GAN","R","ai publications by india","Microsoft"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","Aim","RAG","R","Go","API","GAN"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/microsoft-replaces-journalists-with-ai-can-we-rely-on-ai-for-news\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10020327,"title":"Top 10 Papers Presented At MLDS 2021","content":"The Machine Learning Developer Summit (MLDS) by Analytics India Magazine is India’s leading conference for machine learning practitioners. From February 11 to 13, the third edition of MLDS virtually hosted top academics and industry professionals from the data science community. Tech talks, paper presentations, and workshops covered a multitude of topics in machine learning and data science. Here, we give you a sneak peek at the top ten papers presented at the conference. Predicting Demand Offset To React To Unforeseen Critical Events — Priyanka Telang The supply chain depends a lot on actionable insights from demand forecasting. The data-driven insights help organisations allocate resources judiciously. The forecasting models factor in historic demands to make sensible projections. However, such models are not equipped to account for black swan events like a pandemic breakout, given the limitations of their parameters. This paper tried to address this issue. Telang has developed a method to continuously monitor the external events in real-time, stream, cluster them and then arrive at an offset in the demand. The demand will be based on what was seen in the past and the context of occurrence of the events, like the point in the sales cycle when the event is occurring. This helps increase the reliability of the demand forecasts to react to unforeseen events effectively. ML-Based High-Cardinality Reduction Methods To Create Geo-Score To Improve Auto Insurance Tweedie Pricing Model — Suguna Jayaraj Though traditional auto insurance pricing models consider risk factors like driver, vehicle, and policy characteristics, geographies largely stayed out of the equation owing to their high cardinality. While postal codes can be used for the same, this can present problems since loss in cost is almost always zero in case there is low exposure in certain areas and the model has low confidence due to the information lost on latent variables. This paper presented a case study where a geo-score was developed at a postal code level to improve risk segmentation. The base loss cost model was built using Tweedie Compound Poisson regression. Geospatial attributes were added to the model without changing the existing rating structure. Reducing the high cardinality features of geographic data and further including socio-demographic variables, the paper presented a hybrid approach of a target-based encoding method. Solution Approach To Resolving Vehicle Routing Problem Using Deep Reinforcement Learning — Dr Monika Singh Vehicle routing entails finding the optimal route to improve ETAs. The paper showed how deep reinforcement learning can be used to solve this classic optimisation problem. Rather than being explicitly programmed, a dynamic attention model was developed with an encoder-decoder architecture where each node is dynamically characterised in the context of the graph. The algorithm was benchmarked against the genetic algorithms and was evaluated using two KPIs — travelling cost and computational time. According to Dr Singh, the comparison showed a 5X-20X reduction in cost and a 100X–1000X reduction in computational time. Pneumothorax Detection And Classification On Chest Radiographs Using AI — Tejas Haritsa V K Pneumothorax refers to an abnormal collection of air in the pleural space between the lung and the chest wall leading to a partial or complete lung collapse. Radiography is being used for diagnosis of various diseases and can help in timely detection of Pneumothermax. The paper analysed the performance of AI in detecting pneumothorax by training the model on two different datasets. Two AI systems were evaluated using the high-resolution complete images and the other involved providing medium resolution images in segments. The paper laid out the performance metrics and limitations of both methods. The segmented approach showed an accuracy of 96.83% while the full image showed an accuracy of 95.10%. Modified Count Difference Feature Selection Method For Text Classification — Manik Garg The lifecycle of text classification models includes inputting data, preprocessing data, extracting features, selecting features, and then applying the machine learning classifiers. There are various methods for the feature selection, and one of them is the Count Difference Method (CDM), which, however, has various limitations. The method is restricted to binary classification as it can measure the difference in a particular feature across two different classes. Also, huge datasets are a time sink. Garg proposed a method he developed with his coauthors to overcome these challenges and extend CDM to multi-class classification. This was done by comparing the relevance of a feature in the present class and calculating if it is prominent across classes using the ‘centred maximum value’. The experiment is scalable, simple to implement, and computationally fast. Efficient And Optimal Deep Learning Inference For Computer Vision Applications — Venkatesh Wadawadagi ‘Optimal deep learning inferencing’ can help process data in real-time and ensure faster model validation. It also helps overcome limited computational, storage and memory resources. Otherwise, hardware becomes a bottleneck on applications run locally and operational costs soar for applications running on-the-cloud. Any device using computer vision will need to tackle these limitations for maintaining a healthy balance between model performance or accuracy and inference time. Wadawadagi presented a paper on various approaches to optimise deep learning inferences. He covered different methods including quantisation techniques, network pruning,  matrix factorisation, and inference accelerator. The presentation covered the advantages and tradeoffs of selecting different approaches. Telecom Churn And Valued Customer Retention — Srinivasarao Vallaru The telecom industry is a highly competitive market and consumer retention is the lifeblood of their business. The paper proposed a progressive analytical approach that could help segment customers on the basis of factors like churn severity and churn priority. This can identify risk levels of customers churning from low to high, which then can be used to take retention measures and, in turn, increase their revenue. Predicting Missing Product Taxonomy In Retail: An Embedded Approach Using N-gram Mixture Models And Newton’s Method — Neeraj Mishra In retail, taxonomy is a hierarchical and logical arrangement of products to help customers find what they need in the store or website. The process is expensive as it needs human resources including taxonomists, information scientists, and linguists to build an effective taxonomy. Mishra has developed a novel machine-learning algorithm by leveraging N-gram Mixture Model, cross-entropy function, and Newton’s optimisation method to set up a product taxonomy. A modified Naïve Bayes algorithm and up to 4-gram models were combined with general heuristics to produce a model that was deployed and tested for online retail data. It achieved an 84% accuracy. Machine Learning Approach To Predict Patient Position For Preventing Bedsores — Sujoy De and Aditya Agarwal The cost of patients developing bedsores is very high, with the US alone spending more than $9.11 billion annually in treating around 2.5 million individuals. Currently, there are smart beds that alert patients who have not changed their position. However, these devices are expensive. De and Agarwal came up with a solution that uses low-cost load-cells to accurately estimate the patient position with an accuracy of 98.8%. They used various feature engineering models to distinguish one position from another. This helped them generate meaningful intuitive features that are used by various machine learning models to generate alerts when a patient has been in the same position for a prolonged period of time. A Noninvasive Model To Detect Dengue Based On Symptoms Using Artificial Intelligence and Machine Learning — Dr Ruban S AI is significantly impacting the patient care system. Dr Ruban, along with his team, has developed a non-invasive model to detect dengue based on symptoms using AI and machine learning techniques. This was developed using the data of hospitals in the rural areas of coastal Karnataka. Trying different machine learning approaches along with balancing datasets using oversampling techniques, the model was able to provide an insight into different symptoms to predict dengue with an accuracy of 98%. The team plans to improve the model by incorporating data across multiple geographies.","excerpt":"The Machine Learning Developer Summit (MLDS) by Analytics India Magazine is India’s leading conference for machine learning practitioners. From February 11 to 13, the third edition of MLDS virtually hosted top academics and industry professionals from the data science community. Tech talks, paper presentations, and workshops covered a multitude of topics in machine learning and […]","categories":["Deep Tech"],"tags":["balancing classes","big data and data science","Ml research papers"],"author_name":"Kashyap Raibagi","publish_date":"2021-02-16T11:00:00","publication_year":"2021","word_count":1295,"keywords":["Ml research papers","data science","text classification","artificial intelligence","machine learning","AI","big data and data science","ML","computer vision","RAG","deep learning","analytics","balancing classes"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","computer vision","data science","analytics","RAG","text classification"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/top-10-papers-presented-at-mlds-2021\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10018665,"title":"How This Noida Based Startup Leverages AI To Help Companies Find Right Candidates","content":"Organisations, irrespective of their size, function or industry, have either implemented or are exploring AI in their businesses. Investments are flowing into AI startups, and India is experiencing tremendous interest and demand around AI and machine learning. In today’s HR industry, the recruiters are juggling tons of resumes, complicated hiring routines and time-consuming processes. On the other hand, job-seekers are subjected to lengthy hiring procedures. Enter Vasitum, an AI-powered career platform to bridge the gap between the recruiters and the candidates. For this week’s startup column, Analytics India Magazine spoke to Vikram Wadhawan, Founder & CEO of Vasitum to understand how the startup is leveraging AI and machine learning to help companies find the right candidates. Founded in 2019 by Vikram Wadhawan and Ankit Kashyap, Noida-based Vasitum enables automated resume screening, shortlisting of candidates and automatic Interview scheduling. For recruiters, a tailored screening process is applicable, while capitalising on existing resume banks. Flagship Products Vasitum is a full-stack recruitment automation platform to automate various HR functions such as sourcing, screening, assessment, interview scheduling, background check & documentation. Through Vasitum, a recruiter will be able to use Vasi, the AI assistant to perform sourcing, screening, assessment, interview scheduling, background checks and documents collection to make their workflow efficient & smarter. In the next few weeks, the company will roll out a new feature: Stack Ranking of candidates based on their experience and skills to match the job, along with automated assessments. What’s The Differentiator? “Besides various HR Management Systems to help recruiters, recruiters jump across multiple platforms like a job-board to assessment engine to background agency etc. to close out a vacancy. On the flip side, candidates are barely aware of their application status. For the ones who do get shortlisted, they face back and forth while getting their interview scheduled. That’s when Vasitum comes into the picture,” said Wadhawan. He added, “For the employers, we are automating the recruitment process end to end. For Candidates, we are ensuring transparency and great experiences.” Use of AI & ML @ Vasitum According to Wadhawan, Vasitum leverages AI from the top of the hiring funnel through to the moment of hire. From the start of the process, the AI-based chatbot – Vasi sources ideal profiles for the job, stack ranks, and pre-screens them for the role. He said there are many other smart features in development that would make the hiring process a lot more efficient and save tons of time. The company predominantly works on Python and natural language processing (NLP) in terms of technology and framework. Roadmap On his concluding note, Wadhawan said, “Our product roadmap is full of additions for next 24 months, we have new features and modules planned for both B2B and B2C audience for next 24 months, which will help improve overall experience and efficiency.” He added, “five years is a very big time for an early-stage startup and therefore planning for five years may be a futile exercise. We do have plans for the next 12 months, and that includes monetisation in India in three to six months and a soft launch in South East Asia in 9 to 12 months. We are part of the 500 Startups Global Launch Program for SEA, and it’s a great platform for us to launch in SEA.”","excerpt":"Organisations, irrespective of their size, function or industry, have either implemented or are exploring AI in their businesses. Investments are flowing into AI startups, and India is experiencing tremendous interest and demand around AI and machine learning.  In today’s HR industry, the recruiters are juggling tons of resumes, complicated hiring routines and time-consuming processes. On […]","categories":["AI Startups"],"tags":["AI Companies","Startups"],"author_name":"Ambika Choudhury","publish_date":"2021-01-22T16:00:00","publication_year":"2021","word_count":548,"keywords":["API","machine learning","AI","ML","RAG","NLP","Python","analytics","AI Companies","GAN","Startups","R"],"extracted_tech_keywords":["AI","machine learning","ML","NLP","analytics","RAG","Python","R","API","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-startups\/how-this-noida-based-startup-leverages-ai-to-help-companies-find-right-candidates\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":32931,"title":"Indian Govt To Build 20 Interdisciplinary Cyber Physical Centres","content":"Indian Prime Minister Narendra Modi kicked off the new year with a bang as he announced that the BJP-led Union Government had approved a National Mission on setting up 20 Interdisciplinary Cyber Physical Systems with an investment of over ₹3,600 crore. Delivering the inaugural speech at the 106th session of the Indian Science Congress, PM Modi said that these funds would be used for R&D, technology development, human resources and skills, innovation, startup ecosystem and strong industry and international collaborations. These 20 centres are to be established in the existing institute of excellence like IITs and IISc. Ashutosh Sharma, secretary of the department of science and technology at the 106th Indian Science Congress, told a leading daily, “There will be a rise of the intelligent machines… There will be Cyber Physical Systems that will have the ability of perception, decision making, and action. There will also be greater linkages, for example, a network of driverless car will learn from the mistakes of one another.” Reports have suggested that these work on these centres will begin by the next financial year. This research would be focussed on artificial intelligence, machine learning, Internet of Things and Quantum Computing under the national mission. Reflecting on the theme of the event this year — ‘Future India: Science and Technology’ — PM Modi also said that India’s true strength will be in connecting its science, technology and innovation with its people. Throughout his tenure, PM Modi, as well as the NDA-led government, is proactive in adopting new technologies. It has been showcased by their marked preference in taking proactive steps towards helping the private sector, as well as updating the public sector and government entities with new technologies. Artificial intelligence and big data have been of keen interest to PM Modi. Through a series of events and speeches from the last few years, he has been deliberately showcasing India as well as his government as technologically forward. And it has worked out very well for them, as reports suggest that industry insiders are happy to see that the government is working towards supporting a tech-driven future. Some of PM Modi and the BJP government’s past initiatives include Atal Innovation Mission’s Atal Tinkering Labs, Atal Incubation Centers, Scale-up support to Established Incubators, several programmes and associations with industry leaders such as SAP and Google, among others. The Government has also set up an AI Task Force to prepare India for the upcoming Industrial Revolution 4.0 and the resulting economic transformation.","excerpt":"Indian Prime Minister Narendra Modi kicked off the new year with a bang as he announced that the BJP-led Union Government had approved a National Mission on setting up 20 Interdisciplinary Cyber Physical Systems with an investment of over ₹3,600 crore. Delivering the inaugural speech at the 106th session of the Indian Science Congress, PM Modi […]","categories":["AI News"],"tags":["AI (Artificial Intelligence)","IoT","Narendra Modi"],"author_name":"Prajakta Hebbar","publish_date":"2019-01-07T09:10:29","publication_year":"2019","word_count":414,"keywords":["big data","Go","artificial intelligence","machine learning","programming_languages:R","AI","innovation","programming_languages:Go","Narendra Modi","AI (Artificial Intelligence)","R","IoT","startup"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","R","Go","big data","innovation","startup","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/indian-govt-to-build-20-interdisciplinary-cyber-physical-centres\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10078567,"title":"Fundamentalists Pose Bigger Threat to Bitcoin than Anyone Else","content":"Dan Held, a Bitcoin veteran, stirred a debate on Twitter when he wrote, “Say no to Bitcoin fundamentalism”. When pressed for more, he loosely explained how Bitcoin fundamentalists have deemed everything, part of the free market, immoral. “Alts are immoral. Lending is immoral. VCs are immoral – Making money is immoral.” This is the general viewpoint of purists, who believe Bitcoin holds a Godly position, untouched by worldly things. Experts believe it is an extreme position to take in support of technology and likened it to religious extremism. Bitcoin fundamentalism explained In a podcast named Bankless Nation, Held explains the traits of Bitcoin fundamentalists putting it into the context of religions. Religion is the best way to explain the rise of extremism in Bitcoin. According to him, every religion has sub-groups that advocate imposing extreme rules, which are not supported by the main group. Say no to Bitcoin fundamentalism.— Dan Held (@danheld) September 16, 2022 “The same cultural element has been brought into the Bitcoin community by a few that doesn’t represent the core values of the Bitcoin ethos. For example, claiming that earning interest is usury, is some sort of mediaeval religious response to the interest in which earning interest or earning yield is simply a free market activity,” said Held. He added that this Bitcoin sub-group calls out certain influencers for making money and building courses. They say that making money from educational activities is some kind of a perversion of this pure ethos of giving out all of your time and effort to help Bitcoin. “If people want to sell a course, they can sell the course. If they want to give away a course, they can give away the course. You’re free to price your services or products at zero but to call out someone, who is making money while doing a fantastic job of educating people about Bitcoin, looks like extremism,” said Held. Experts, however, explain the reason behind how this sub-group came into existence. According to Held, these fundamentalists were once the defenders of the right cause who have now turned inward and are fighting their own community. “Blockchain is a community and many believe that the community should help each other without having ulterior motives. But, at the same time, there is nothing wrong if someone is charging for making an effort to teach,” said Raj Kapoor, founder, India Blockchain Foundation. Today’s attackers were once rightful defenders In 2016-17, a debate over the Bitcoin block-size ensued. This debate caused a major split and created two camps – the ‘Big blocks’ and ‘Small blocks’ camps. The Big blocks segment was led by a Chinese mining giant Roger Ver, an early Bitcoin investor. This camp was concerned about the scalability issue of Bitcoin, which will prevent it from becoming a peer-to-peer payment platform. Owing to the long processing times and high fees, people would avoid Bitcoin for day-to-day transactions and instead use it as a store of value – like gold. They suggested increasing the block size. If Bitcoin’s block size is increased to 8 MB, it will increase the number of transactions per second by 8 times. This will eventually reduce network congestion, and in the future, the block size can be increased as much as needed as Bitcoin achieves further adoption. However, this was opposed by the ‘Small Blocks’ camp. The camp supported keeping the current 1 MB block size and focussed on optimising transaction size and handling, in order to achieve scaling. The solution they proposed was Segregated Witness or Segwit, an upgrade to the Bitcoin protocol that would effectively reduce the transaction size by 75%. This will allow a 1MB Segwit block to hold the same amount of transactions as a 4 MB non-Segwit block. Additionally, Small Blockers spoke about the development of the L2 network solution – a layer 2 or second layer solution on top of the Bitcoin protocol that will allow instant and feeless transactions. In 2017, the debate reached its climax and in August that year, the Big blocks camp hard forked a new Bitcoin called Bitcoin Cash while Small blocks proponents created Bitcoin Segwit. Initially, it was unclear which Bitcoin would win (Bitcoin Cash or Bitcoin) with a longer blockchain, or ledger of transactions). This is how it happens – with more miners on board, the computational power increases, which also increases blockchain length and network robustness. Bitcoin Cash favoured support from Bitmain, and as a result, the original Bitcoin’s hashing power was cut nearly in half when the fork occurred. Eventually, Bitcoin emerged strong even after the fork. This was the time when the community actually fought against the big conglomerates to maintain the decentralisation power of blockchain and save it from outside influence.","excerpt":"A Bitcoin fundamentalist holds extreme views about how Bitcoin should be used and traded in the market","categories":["AI Features"],"tags":["Bitcoin","Blockchain","Ethereum"],"author_name":"Tausif Alam","publish_date":"2022-11-03T15:00:00","publication_year":"2022","word_count":789,"keywords":["Go","Bitcoin","Blockchain","programming_languages:R","AI","programming_languages:Go","Scala","Aim","ViT","programming_languages:Scala","R","Ethereum"],"extracted_tech_keywords":["AI","Aim","R","Go","Scala","ViT","programming_languages:R","programming_languages:Scala","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/bitcoin-fundamentalists-pose-biggest-threat-to-bitcoin-than-anyone\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10171961,"title":"ZUPPA Launches Drone MRO Lab at Madras Regimental Centre to Strengthen Defence Capabilities","content":"ZUPPA Geo Navigation Technologies, India’s leading deep-tech drone developer and manufacturer, set up a cutting-edge maintenance, repair, and operations (MRO) Lab on Wednesday at the Madras Regimental Centre (MRC) in collaboration with the Dakshin Bharat Area. This move comes in response to the Indian Army Chief’s call in November 2024 for a “drone for every soldier”, while highlighting the need to empower Indian soldiers to independently operate, maintain, and troubleshoot drones during active conflict. The establishment of the MRO lab follows the success of Operation Sindoor, which demonstrated the critical need for soldiers to not only operate drones but to maintain them in real time during missions. “The real challenge in drone warfare lies not just in flying them, but in understanding the intricate electronic systems—especially the impact of high electromagnetic fields (EMF) on sensitive onboard sensors,” Sai Pattabiram, founder and managing director of ZUPPA, said. “That’s what our MRO lab will offer—a hands-on, tech-intensive learning framework built for scale.” The proposal for the lab stemmed from technical discussions between Lt Gen Karanbir Brar, general officer commanding at Dakshin Bharat Area, and ZUPPA. It includes DGCA-certified drone pilot training and licensing, in partnership with TNUAV Corporation. Following approval, a memorandum of understanding (MoU) was signed between ZUPPA and MRC, marking the beginning of the project’s implementation. A unique feature of the ZUPPA MRO Lab is its learning management system (LMS), which delivers modular, bite-sized lessons on topics ranging from component identification to advanced EMF diagnostics. This LMS can be accessed remotely, benefiting both instructors and learners. The lab’s approach is particularly valuable for scaling instruction across the mass Agniveer training programs, as well as providing training to retiring soldiers to support post-service careers and currently serving armed forces professionals enrolled in various military courses. “The Indian Army’s adoption of our fully indigenous MRO solution is a strong validation of both our technical expertise and our delivery model,” Venkatesh Sai, co-founder and technical director at ZUPPA, said. “We see a huge opportunity to replicate this success across other regimental training centres.” The company has also deployed its Innovation Labs at AICTE-approved institutions and engineering colleges, aiming to bridge the gap between defence requirements and academic research. Looking ahead, ZUPPA plans to expand the MRO program to other technical arms of the Indian Army, further enhancing India’s self-reliance in defence technology.","excerpt":"This move comes in response to the Indian Army Chief’s call for “a drone for every soldier” in November 2024.","categories":["AI News"],"tags":["AI in defence","Drones in india"],"author_name":"Shalini Mondal","publish_date":"2025-06-18T17:58:12","publication_year":"2025","word_count":389,"keywords":["Replicate","programming_languages:R","AI","innovation","Aim","Drones in india","R","AI in defence"],"extracted_tech_keywords":["AI","Aim","R","innovation","Replicate","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/zuppa-launches-drone-mro-lab-at-madras-regimental-centre-to-strengthen-defence-capabilities\/","complexity_score":2,"technical_depth":6,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10059863,"title":"Purdue researchers build human brain-inspired hardware for AI to enable continuous learning","content":"Researchers at Purdue University have created an artificial platform for machines to help them learn through their lifespans. The researchers aim to make AI more portable by embedding it directly into the hardware, instead of running AI as software, making machines operate more efficiently, especially in isolated environments such as space or autonomous vehicles. “If we want to build a computer or a machine that is inspired by the brain, then correspondingly, we want to have the ability to continuously program, reprogram and change the chip,” Shriram Ramanathan, a professor in Purdue University’s School of Materials Engineering, said. He specialises in discovering how materials could mimic the brain to improve computing. Human brains are capable of continuous learning throughout their life spans by forming new neural connections. However, the circuits on a computer chip don’t change. The researchers, under the aegis of Ramanathan, built a hardware that can be reprogrammed on-demand using electrical pulses. Ramanathan believes the adaptability will allow the device to take on all functions necessary to build a brain-inspired computer. The hardware is a small, rectangular device made of a material called perovskite nickelate, which is very sensitive to hydrogen. Applying electrical pulses at different voltages allows the device to shuffle a concentration of hydrogen ions in a matter of nanoseconds, creating states that the researchers found could be mapped out to corresponding functions in the brain. When the device has more hydrogen near its centre, it can act as a neuron, a single nerve cell. With less hydrogen at that location, the device serves as a synapse, a connection between neurons, which is what the brain uses to store memory in complex neural circuits. Through simulations of the experimental data, the Purdue team’s collaborators at Santa Clara University and Portland State University showed the internal physics of this device creates a dynamic structure for an artificial neural network with the ability to efficiently recognise electrocardiogram patterns and digits compared to static networks. The neural network uses “reservoir computing,” which explains how different parts of a brain communicate and transfer information. The researchers are working to demonstrate these concepts on large-scale test chips.","excerpt":"The hardware is a small, rectangular device made of a material called perovskite nickelate","categories":["AI News"],"tags":["Autonomous Vehicles","neurons"],"author_name":"SharathKumar Nair","publish_date":"2022-02-04T13:37:58","publication_year":"2022","word_count":356,"keywords":["programming_languages:R","AI","neural network","Git","neurons","Aim","Autonomous Vehicles","R"],"extracted_tech_keywords":["AI","neural network","Aim","R","Git","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/purdue-researchers-build-human-brain-inspired-hardware-for-ai-to-enable-continuous-learning\/","complexity_score":3,"technical_depth":6,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":30894,"title":"Meet ‘ThirdEye’, An AI-Based Vision Analytics Platform By Mobiliya","content":"Artificial intelligence has done wonders across various domains, including visual analytics. In a recent development, Mobiliya, a QuEST Global Company, announced the launch of a novel AI-based blueprint solution, ThirdEye, which is a powerful vision-analytics platform technology powered by NVIDIA and Microsoft. The technology aspires to boost operational intelligence and intensify the decision-making process. What Is ThirdEye ThirdEye is a robust vision-analytics platform which aims to improve operational intelligence and enhance the decision-making process. Some of its attributes are: The ThirdEye technology is based on a distributed camera mechanism which is capable of capturing an image that will be analysed by an AI agent to extract eligible insights. The technology is conferred through a cooperative admin dashboard. It uses progressive intelligent video analytics which empowers this solution to achieve definite patterns and actionable understandings from the video feed. The solution can logically complement and boost the current infrastructure of CCTV and IP cameras. The platform uses fundamental NVIDIA jetson edge computing and Microsoft’s Azure IoT hub, Azure machine learning services and Power BI. NVIDIA GPUs guides the processed images and videos from the data source for creating dashboards, showcasing actionable business intelligence insights. The prototype is equipped with self-learning and enriching efficiency, the platform helps manufacturers identify and record errors with great accuracy. Industries That Can Benefit Through ThirdEye Businesses from manufacturing to retail can be benefited through ThirdEye. With the current “brownfield” or with new CCTV and IP cameras, it can analyse visual data, gain new insights and take real-time action to ensure that infrastructure and facilities are secured and working properly. Some of the probable use cases are: ThirdEye can contribute manufacturers with an entire view of all of the manufactured products’ specifications and its packaging with the help of AI-driven authorisation. ThirdEye technology can empower cross-industry applications like smart parking, smart lobbies, retail heat maps, smart meeting rooms, and intelligent customer counters. The technology can be extended to empower security personnel, and other exercises across industries like retail, healthcare, real estate, transportation, smart cities, education, banking, and law enforcement.","excerpt":"Artificial intelligence has done wonders across various domains, including visual analytics. In a recent development, Mobiliya, a QuEST Global Company, announced the launch of a novel AI-based blueprint solution, ThirdEye, which is a powerful vision-analytics platform technology powered by NVIDIA and Microsoft. The technology aspires to boost operational intelligence and intensify the decision-making process. What […]","categories":["AI Features"],"tags":["AI in Public Sector","artificial inelligence"],"author_name":"Bharat Adibhatla","publish_date":"2018-11-30T11:09:45","publication_year":"2018","word_count":342,"keywords":["business intelligence","machine learning","artificial intelligence","artificial inelligence","AI","R","AI in Public Sector","Aim","analytics","edge computing","Azure","analytics platform"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","analytics","Aim","Azure","edge computing","R","business intelligence","analytics platform"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/meet-thirdeye-an-ai-based-vision-analytics-platform-by-mobiliya\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":42347,"title":"A Day In The Life Of: Vikash Raj, A Spiritualist &#038; A Self-Professed Workaholic","content":"In our column ‘A Day In The Life Of’, we try to step into the shoes of the awesome techies from various domains in data science. And this week, we caught up with Vikash Raj, Head of Business Analytics and Process Engineering at IDFC, to know how his “day in the life” looks like. The Onset Raj kick starts his day early in the morning at 5 AM and has made it a point to do some self-study every day to keep him updated with the latest in the industry. His office hour starts at 9:30 AM but he prefers to start early and begin by 8:30 AM. When asked about whether work timing is strict,  Raj said, “We have a flexible hours work culture in our organization, and we can sometimes work from home as well depending upon our bosses discretion.” Being a data scientist, it is imperative to have a mind that is focused and calm enough, and Raj seems to be successful in that as he does meditation daily — without fail. He has even joined a Yoga class that happens 4 times a week and he makes sure he actually attend the class at least 3 times every week. When asked about his hobby, this is what Raj said, “I am afraid, I have not been able to identify a hobby for myself, but I am generally interested in Hindu mythology (Specially related to Goddess Durga) and tend to read whenever I get a book related to the subject. Also, I like spending time with nature alone.” Work Process Talking about his work process, Raj’s work consists of maintaining the data information and coming out with use cases for better and productive usage. He has conceived and implemented multiple products on data that has added to the efficiency of the team but has also increased and articulated the accountability. His job also includes explaining the case to the data scientists, helping them with feature creations and selection of algorithms and measuring the returns from each of the endeavours. “In a sense, I do end to end job from conceiving the idea — putting it through the relevant stakeholders and convincing them of the benefits and ROI, provide ideas with the help of my team of data engineers, data scientists and data custodians, implementing the ideas and then calculate the benefits in terms of ROI,” Raj explained. When asked about his current project, the IDFC data scientist said he is attempting to automate a data-driven sales planning where they do relationship mapping, forecasting business from each partner across their branches for each product segment, use this information for KRA setting and then track the business along with helping the sales team with regular inputs on patterns who may not be giving the required business and need interventions. While Raj walked us through his “a day at work”, we asked him how the IDFC is using his talent and skills, and Raj said, “At IDFC, we have a fairly good data-driven culture and lots of strategies and actions revolve around data-based decisions. Though there is always a feeling that my output could go up if more resources are allocated but rational thought is that within the industry constraints and to strike a balance between long term and short term deliverables, my company is making good use of my talent.” Work-Life Balance When it comes to having a social life, Raj’s social life is not very active. He feels he is an introvert and he enjoys his own company that of anybody else. “Sometimes, I like to spend some time socially but that too with some very close friends and that too for some limited time,” said Raj. He likes restricting to his conversations to the extent needed and hardly engage in long chitchats. However, talking about striking that perfect work-life balance, Raj said he never feels that his work affects his personal life as his personal life revolves around his work. Sometimes he does get into stretched meetings, but he always manages to plan his other engagements, keeping buffer for the same. “I enjoy my work and sometimes my team feels that I have some level of addiction to my work which sometimes affects their personal life, but I try to avoid that,” said Raj.","excerpt":"In our column ‘A Day In The Life Of’, we try to step into the shoes of the awesome techies from various domains in data science. And this week, we caught up with Vikash Raj, Head of Business Analytics and Process Engineering at IDFC, to know how his “day in the life” looks like. The […]","categories":["AI Features"],"tags":["Data Science Career","Data Scientist","Interviews and Discussions","meditation"],"author_name":"Harshajit Sarmah","publish_date":"2019-07-12T14:40:02","publication_year":"2019","word_count":717,"keywords":["meditation","data science","Go","TPU","programming_languages:R","AI","data-driven","programming_languages:Go","Data Science Career","analytics","GAN","Data Scientist","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","data science","analytics","TPU","R","Go","GAN","data-driven","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/a-day-in-the-life-of-vikash-raj-a-spiritualist-a-self-professed-workaholic\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10041113,"title":"All Major NVIDIA Announcements At Computex 2021","content":"NVIDIA has announced new products, enhancements and features at Computex Taipei 2021. The company claimed to have integrated artificial intelligence across every consumer-facing applications, data centres and enterprise products. Today, more than 1000+ companies are using NVIDIA AI (software). Let’s look at how companies across the world leveraged NVIDIA solutions to drive better outcomes: TSME: Improved silicon water inspection cycle time by 13x GE Healthcare: Accurately detect diseases in 145 million hearts per year Pinterest: Identify trends in over 300 billion pins for better search results Postmates: Tasteful recommendations from 600k restaurants Spotify: Personalised playlist for over 345 listeners T-Mobile: Award-winning customer care using real-time ASR United States Postal Service: Real-time analytics on 7 billion packages per year WeChat: Intelligent search with state-of-the-art NLU for 1.2 billion users From prototype to production NVIDIA’s state-of-the-art DGX systems are being used extensively by universities, telcos, hospitals, consumer-focused enterprises, banks, carmakers and aerospace companies. At Computex 2021, NVIDIA announced the upgraded version of the base command platform, ideal for large-scale, multi-user, multi-team development. The platform comes with MLOPs APIs, Jupyter Notebook, TensorBoard and other features like scheduling, data catalogue, reporting and dashboard, and accelerates the prototype-to-production lifecycle. NVIDIA has partnered with NetApp for storage purposes. AWS and Google Cloud will integrate base command platforms with Cloud GPU instances. For instance, the base command customers can deploy their ML workload on Amazon SageMaker. NVIDIA said the technology inside DGX (be it HGX A100 8-GPU, HGX A100 4-GPU, A100 GPU, A30 GPU, A40 GPU, A10 GPU, BlueField-2 DPU) is now ready for the mainstream. Meaning, the company is looking to offer low power, smaller forms of GPUs at a lower cost for enterprises. The company also announced its plans to offer DGX SuperPOD via a subscription model for enterprises. “We are lowering the barrier to entry so that companies can experience this system and equipment,” said NVIDIA. The company had unveiled the NVIDIA DGX SuperPOD last year, a cloud-native, multi-tenant AI supercomputer. Certified systems NVIDIA-Certified Systems are tested using software from the NGC Catalog, and customers can choose to purchase enterprise-grade support through NVIDIA-Certified Systems Software Support. As part of the certified programme, NVIDIA collaborates with a broad range of system manufacturers for different use cases, including graphics, virtual desktops, and digital infrastructure. “NVIDIA-Certified is our path going forward,” said NVIDIA. The notable partners include ADVANTECH, ALTOS, HP, ASUS, DELL Technologies, Lenovo, etc. NVIDIA AI Enterprise NVIDIA provides support across AI and data science applications and frameworks, including TensorFlow, PyTorch, NVIDIA Transfer Learning ToolKit, NVIDIA TensorRT, RAPIDS, NVIDIA Triton Inference Server, among others, alongside the infrastructure management optimisation like NVIDIA vGPU, NVIDIA Magnum 10, NVIDIA CUDA-X AI, NVIDIA DOCA for AI models like Metropolis, Clara, Jarvis, Merlin, Mazina, Isaac and Morpheus. Further, NVIDIA said it supports VMware VSphere and NVIDIA-Certified Mainstream Enterprise Servers. NVIDIA Omniverse Enterprise NVIDIA offers enterprise support to Omniverse entities such as Nucleus, Connect, Kit, Simulation and RTX Renderer across industries, including media and entertainment, game development, engineering, construction, manufacturing etc. The entities can run on NVIDIA studio laptop, RTX Workstation and certified server, along with virtualisation platforms such as VMWare vSphere and NVIDIA RTX Virtual Workstation. Major announcements Expanding its NVIDIA-certified programme to DPU, including HGX (supercomputing), EGX (enterprise accelerated computing), and BlueField DPU (software-defined data centre on a chip — expected to be launched by the end of this year).  Expanding its NVIDIA-Certified programme to Arm by 2022. Launching developer kit for Arm-based accelerated computing","excerpt":"NVIDIA has announced new products, enhancements and features at Computex Taipei 2021. The company claimed to have integrated artificial intelligence across every consumer-facing applications, data centres and enterprise products. Today, more than 1000+ companies are using NVIDIA AI (software). Let’s look at how companies across the world leveraged NVIDIA solutions to drive better outcomes: TSME: […]","categories":["Global Tech"],"tags":["Machine Learning","Machine Learning Latest"],"author_name":"Amit Naik","publish_date":"2021-06-01T16:00:00","publication_year":"2021","word_count":575,"keywords":["data science","Amazon SageMaker","artificial intelligence","AI","PyTorch","ML","MLOps","Machine Learning","Machine Learning Latest","Aim","analytics","TensorFlow"],"extracted_tech_keywords":["AI","artificial intelligence","ML","data science","analytics","MLOps","Aim","Amazon SageMaker","TensorFlow","PyTorch"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/all-major-nvidia-announcements-at-computex-2021\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10052545,"title":"Healthtech Startup Zini.ai Raises Pre Series A Funding At Valuation Of INR 101 Crores","content":"Zini.ai, an AI-powered multilingual virtual physician by Delhi-based Grainpad Pvt Ltd, has raised Pre Series A funding from Solarus Group, a London based venture capital & private equity firm with a presence in England, Egypt and India. The funds will be used towards enhancing existing technology, new business development and expanding into new geographies. Zini.ai is India’s first AI-powered, voice-based virtual physician designed by a team of experienced doctors and techies. Founded by Dr Rohit Sharma in 2017, Zini provides genuine expert medical guidance and directs patients to seek timely medical help. Accessible through an app, Zini allows users to ‘Talk to Zini’ about any medical symptom or health issues. With an Alexa-like experience, the app can evaluate 950+ health symptoms, 300+ diseases, provide a detailed report, recommend the best course of action and share details of nearby medical facilities that a patient can reach out to. “We started Zini with an aim to make expert medical advice accessible to everyone and to bridge the gap in the doctor-patient ratio in India. Over the last three years, our conversational AI-powered multilingual virtual physician ‘Zini’ has received an overwhelming response from hospitals, our customers and the medical fraternity. We are pleased with the further validation of our product and business model from our investors. The new investment will accelerate our next phase of growth and enable us to strengthen our existing technology,” said Dr Rohit Sharma, CEO and Founder, Zini.ai. Tamer Soliman, CEO and Director, Solarus Group, said, “We are excited to be investing in Zini.ai. The team has shown great commitment over the last several years in building a product that is solving a real-world problem. Zini’s ability to discuss symptoms, diagnose accurately and recommend the best course of action has the potential to help every player involved in the healthcare sector, from patients to healthcare professionals. We believe that the startup is solving a large unmet need in the market, and it has the potential to become a defining brand in the category it operates in. We are happy to be a part of their growth story.” Zini has been trained on medical data very specific to the Indian population. Over the last few years, Zini’s medical team of 10+ doctors has curated a database of symptoms, the reasons behind them, and the possible course of action for the symptoms. Available on Android, users can download the app and start using Zini as their personal health guide. Each user gets a Unique Global Health ID where all the relevant information about the patient gets recorded. Zini currently works with 5 hospitals, 10+ telehealth platforms, and has 15,000+ Android users, thus validating its use case in an array of market segments.","excerpt":"Zini.ai is India’s first AI-powered, voice-based virtual physician designed by a team of experienced doctors and techies.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","Android","Data Science","Deep Learning","Funding","Machine Learning","Voice Assistant","Voice Recognition Technology"],"author_name":"Victor Dey","publish_date":"2021-10-28T15:46:10","publication_year":"2021","word_count":451,"keywords":["Go","API","Funding","Voice Assistant","funding","programming_languages:R","AI","venture capital","Machine Learning","Voice Recognition Technology","Ray","Aim","Deep Learning","Data Science","R","Android","AI (Artificial Intelligence)","startup"],"extracted_tech_keywords":["AI","Aim","Ray","R","Go","API","startup","venture capital","funding","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/healthtech-startup-zini-ai-raises-pre-series-a-funding-at-valuation-of-inr-101-crores\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10141744,"title":"Biotech Startup Cradle Raises $73 Million to Boost Protein Engineering","content":"Amsterdam-based Cradle, an AI-powered platform revolutionising protein engineering, has secured $73 million in Series B funding to meet growing demand in research-intensive industries. Led by IVP with participation from Index Ventures and Kindred Capital, the latest round brings Cradle’s total funding to over $100 million. Cradle leverages AI to help scientists design and engineer proteins faster and in a cost-effective manner. The AI startup focuses on engineering protein modalities such as enzymes, vaccines, peptides, and antibodies with the help of generative AI. Cradle focuses on engineering protein modalities such as enzymes, vaccines, peptides, and antibodies with the help of generative AI. The latest investment will support Cradle’s mission to enable scientists to engineer proteins more efficiently, addressing challenges across therapeutics, diagnostics, food, chemicals, and agriculture. Industry leaders such as Novo Nordisk and Johnson & Johnson are already leveraging Cradle’s generative AI platform to accelerate innovation. “A lot of the work that companies like Google, Facebook and others are doing is more in machine learning research and development. They’re not trying to build tools that help biologists use these types of methods in a sort of easy fashion,” said co-founder and CEO Stef Van Grieken, in an earlier interaction with AIM. Protein Engineering with GenAI Proteins, essential to products ranging from drugs to food, are at the core of Cradle’s vision for sustainable and impactful solutions. Engineering these molecular “machines” can create groundbreaking innovations, such as eco-friendly pesticides, animal-free foods, and advanced therapeutics. However, traditional protein engineering methods are costly and time-consuming, often requiring years of research and millions of dollars with no guaranteed success. Cradle claims to transform this landscape by dramatically reducing the number of experimental rounds needed for protein development. For instance, one customer accelerated the activity of the P450 enzyme fourfold in just three experimental rounds, compared to the usual ten. “Over the past two years, our own research and our collaborations with partners have proven that this technology can deliver remarkable results across a range of applications, from developing new vaccines and sustainable chemicals to novel diagnostics and agricultural crop protection. Our goal is now to put Cradle’s software into the hands of a million scientists and empower them to build great products,” said Grieken. Cradle Team Cradle’s generative protein language model is trained on billions of sequences and refined with proprietary datasets. This iterative approach allows the AI to improve predictions with each round of experimentation, saving both time and resources while significantly boosting project success rates. Unlike AlphaFold, which predicts protein structures, Cradle’s generative models design protein sequences tailored to desired functions and properties. These advancements make Cradle’s platform adaptable for diverse applications, including enzymes, antibodies, vaccines, and bio-based materials. Alex Lim, general partner at IVP, emphasized Cradle’s transformative role. “ Given the costs associated with drug discovery or similar fields of research, any efficiencies at the R&D stage will translate to both major financial returns for customers and significant real-world benefits for humanity.”","excerpt":"The Series B funding brings Cradle’s total funding to over $100 million.","categories":["AI News"],"tags":["AlphaFold","Cradle","IVP","Protein engineering"],"author_name":"Vandana Nair","publish_date":"2024-11-26T18:24:46","publication_year":"2024","word_count":490,"keywords":["Go","AlphaFold","GenAI","machine learning","IVP","API","AI","Cradle","RAG","Protein engineering","Aim","ViT","generative AI","R"],"extracted_tech_keywords":["AI","machine learning","generative AI","GenAI","Aim","RAG","R","Go","API","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/biotech-startup-cradle-raises-73-million-to-boost-protein-engineering\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10038868,"title":"Guide to MMDetection: An Object Detection Python Toolbox","content":"MMDetection is a Python toolbox built as a codebase exclusively for object detection and instance segmentation tasks. It is built in a modular way with PyTorch implementation. There are numerous methods available for object detection and instance segmentation collected from various well-acclaimed models. It enables quick training and inference with quality. On the other hand, the toolbox contains weights for more than 200 pre-trained networks, making the toolbox an instant solution in the object detection domain. MMDetection behaves as a benchmark with the flexibility to reimplement the existing methods or to develop a new detector with the modules available. The major feature of the toolbox is that it contains simple modular components of a typical object detection framework using which one can build custom pipelines or a custom model. Building a new detector framework on top of an existing framework and comparing its performance is easily possible with this toolbox’s benchmarking capabilities. MMDetection and its Architecture Since MMDetection is a toolbox containing many pre-built models and each model has its own architecture, this toolbox defines a general architecture that can adapt to any model. This general architecture comprises the following parts: BackboneNeckDenseHead (AnchorHead\/AnchorFreeHead)RoIExtractorRoIHead (BBoxHead\/MaskHead) A Backbone is the part of the architecture that transforms the input images into raw feature maps. A Neck connects the Backbone with heads and performs reconfigurations and refinements on the raw feature maps so that heads can further process them. A DenseHead is a part that processes the dense locations of the feature maps fed by Neck. An RoIExtractor is the part of the architecture that identifies the region of interest (RoI) and extracts RoI features from the feature maps. An RoIHead is a part that takes RoI features as its input and makes the predictions such as bounding boxes classification or mask prediction as per the task assigned. Framework of a single-stage detector without an RoIHead (Source) Framework of a two-stage detector with an RoIHead (Source) The whole network is built as a series of pipelines so that end-to-end training is made simple with any kind of network. During training, the whole network is traversed in the forward and backward directions over iterations. A typical training pipeline in the MMDetection architecture (Source) Popular Models included in MMDetection MMDetection contains high-quality codebases for many popular models and task-oriented modules. Find below the list of fully-built models and custom-adaptable methods that the MMDetection toolbox supports. The list grows continuously with the inclusion of new models and methods. Fast R-CNN Faster R-CNN Mask R-CNN RetinaNetDCNDCNv2Cascade R-CNNMask Scoring R-CNNFCOSSSDR-FCN M2Det GHM ScratchDet Double-Head R-CNN Grid R-CNN FSAF Libra R-CNN GCNet HRNetMixed Precision Training Weight Standardization Hybrid Task Cascade Guided Anchoring Generalized Attention Inference with MMDetection MMDetection runs better with a CUDA GPU runtime in a PyTorch implementation. The following code references the official tutorial of MMDetection. Check for NVIDIA CUDA compiler and GCC with the following commands. # Check nvcc version !nvcc -V # Check GCC version !gcc --version Install dependencies required to create the environment. !pip install -U torch==1.5.1+cu101 torchvision==0.6.1+cu101 -f https:\/\/download.pytorch.org\/whl\/torch_stable.html # install mmcv-full thus we could use CUDA operators !pip install mmcv-full Install MMDetection from the source repository. # Install mmdetection !rm -rf mmdetection !git clone https:\/\/github.com\/open-mmlab\/mmdetection.git %cd mmdetection !pip install -e . # install Pillow 7.0.0 back in order to avoid bug in colab !pip install Pillow==7.0.0 Create the environment by importing necessary packages. # Check Pytorch installation import torch, torchvision print(torch.__version__, torch.cuda.is_available()) # Check MMDetection installation import mmdet print(mmdet.__version__) # Check mmcv installation from mmcv.ops import get_compiling_cuda_version, get_compiler_version print(get_compiling_cuda_version()) print(get_compiler_version()) Output: Load a pre-trained Mask-RCNN model, trained on the COCO dataset, from the official website. !mkdir checkpoints !wget -c http:\/\/download.openmmlab.com\/mmdetection\/v2.0\/mask_rcnn\/mask_rcnn_r50_caffe_fpn_mstrain-poly_3x_coco\/mask_rcnn_r50_caffe_fpn_mstrain-poly_3x_coco_bbox_mAP-0.408__segm_mAP-0.37_20200504_163245-42aa3d00.pth \\ -O checkpoints\/mask_rcnn_r50_caffe_fpn_mstrain-poly_3x_coco_bbox_mAP-0.408__segm_mAP-0.37_20200504_163245-42aa3d00.pth Load a checkpoint of the pre-trained model and initialize the detector. from mmdet.apis import inference_detector, init_detector, show_result_pyplot # choose to use a config and initialize the detector config = 'configs\/mask_rcnn\/mask_rcnn_r50_caffe_fpn_mstrain-poly_3x_coco.py' # setup a checkpoint file to load checkpoint = 'checkpoints\/mask_rcnn_r50_caffe_fpn_mstrain-poly_3x_coco_bbox_mAP-0.408__segm_mAP-0.37_20200504_163245-42aa3d00.pth' # initialize the detector model = init_detector(config, checkpoint, device='cuda:0') Infer the predictions on a sample outdoor image using the loaded detector. # Use the detector to do inference img = 'demo\/demo.jpg' result = inference_detector(model, img) Plot the result using the in-built plotting method. # Let's plot the result show_result_pyplot(model, img, result, score_thr=0.3) Output: End-to-End Training on Custom Dataset Let’s build a model and train it on the KITTI_tiny dataset. # download, decompress the data !wget https:\/\/download.openmmlab.com\/mmdetection\/data\/kitti_tiny.zip !unzip kitti_tiny.zip > \/dev\/null Each image is supported with a label annotation file in which the annotations of objects present in the image are provided along with the location. Read the annotation file corresponding to an image sample. # Check the label of a single image !cat kitti_tiny\/training\/label_2\/000000.txt Output: The first column indicates the class of the object, and the 5th to 8th columns indicate the bounding boxes. Develop a data generation Python class to convert the data format suitable for training and inference. import os.path as osp import mmcv import numpy as np from mmdet.datasets.builder import DATASETS from mmdet.datasets.custom import CustomDataset @DATASETS.register_module() class KittiTinyDataset(CustomDataset): CLASSES = ('Car', 'Pedestrian', 'Cyclist') def load_annotations(self, ann_file): cat2label = {k: i for i, k in enumerate(self.CLASSES)} # load image list from file image_list = mmcv.list_from_file(self.ann_file) data_infos = [] # convert annotations to middle format for image_id in image_list: filename = f'{self.img_prefix}\/{image_id}.jpeg' image = mmcv.imread(filename) height, width = image.shape[:2] data_info = dict(filename=f'{image_id}.jpeg', width=width, height=height) # load annotations label_prefix = self.img_prefix.replace('image_2', 'label_2') lines = mmcv.list_from_file(osp.join(label_prefix, f'{image_id}.txt')) content = [line.strip().split(' ') for line in lines] bbox_names = [x[0] for x in content] bboxes = [[float(info) for info in x[4:8]] for x in content] gt_bboxes = [] gt_labels = [] gt_bboxes_ignore = [] gt_labels_ignore = [] # filter 'DontCare' for bbox_name, bbox in zip(bbox_names, bboxes): if bbox_name in cat2label: gt_labels.append(cat2label[bbox_name]) gt_bboxes.append(bbox) else: gt_labels_ignore.append(-1) gt_bboxes_ignore.append(bbox) data_anno = dict( bboxes=np.array(gt_bboxes, dtype=np.float32).reshape(-1, 4), labels=np.array(gt_labels, dtype=np.long), bboxes_ignore=np.array(gt_bboxes_ignore, dtype=np.float32).reshape(-1, 4), labels_ignore=np.array(gt_labels_ignore, dtype=np.long)) data_info.update(ann=data_anno) data_infos.append(data_info) return data_infos Modify the model configurations to suit fast training on the prepared dataset. from mmcv import Config from mmdet.apis import set_random_seed cfg = Config.fromfile('.\/configs\/faster_rcnn\/faster_rcnn_r50_caffe_fpn_mstrain_1x_coco.py') # Modify dataset type and path cfg.dataset_type = 'KittiTinyDataset' cfg.data_root = 'kitti_tiny\/' cfg.data.test.type = 'KittiTinyDataset' cfg.data.test.data_root = 'kitti_tiny\/' cfg.data.test.ann_file = 'train.txt' cfg.data.test.img_prefix = 'training\/image_2' cfg.data.train.type = 'KittiTinyDataset' cfg.data.train.data_root = 'kitti_tiny\/' cfg.data.train.ann_file = 'train.txt' cfg.data.train.img_prefix = 'training\/image_2' cfg.data.val.type = 'KittiTinyDataset' cfg.data.val.data_root = 'kitti_tiny\/' cfg.data.val.ann_file = 'val.txt' cfg.data.val.img_prefix = 'training\/image_2' # modify num classes of the model in box head cfg.model.roi_head.bbox_head.num_classes = 3 # We can still use the pre-trained Mask RCNN model though we do not need to # use the mask branch cfg.load_from = 'checkpoints\/mask_rcnn_r50_caffe_fpn_mstrain-poly_3x_coco_bbox_mAP-0.408__segm_mAP-0.37_20200504_163245-42aa3d00.pth' # Set up working dir to save files and logs. cfg.work_dir = '.\/tutorial_exps' # The original learning rate (LR) is set for 8-GPU training. # We divide it by 8 since we only use one GPU. cfg.optimizer.lr = 0.02 \/ 8 cfg.lr_config.warmup = None cfg.log_config.interval = 10 # Change the evaluation metric since we use customized dataset. cfg.evaluation.metric = 'mAP' # We can set the evaluation interval to reduce the evaluation times cfg.evaluation.interval = 12 # We can set the checkpoint saving interval to reduce the storage cost cfg.checkpoint_config.interval = 12 # Set seed thus the results are more reproducible cfg.seed = 0 set_random_seed(0, deterministic=False) cfg.gpu_ids = range(1) Train a new detector model with the preprocessed dataset and modified configurations. from mmdet.datasets import build_dataset from mmdet.models import build_detector from mmdet.apis import train_detector # Build dataset datasets = [build_dataset(cfg.data.train)] # Build the detector model = build_detector( cfg.model, train_cfg=cfg.get('train_cfg'), test_cfg=cfg.get('test_cfg')) # Add an attribute for visualization convenience model.CLASSES = datasets[0].CLASSES # Create work_dir mmcv.mkdir_or_exist(osp.abspath(cfg.work_dir)) train_detector(model, datasets, cfg, distributed=False, validate=True) Test the fully-trained model on a test image. img = mmcv.imread('kitti_tiny\/training\/image_2\/000068.jpeg') model.cfg = cfg result = inference_detector(model, img) show_result_pyplot(model, img, result) Output: Find the notebook with the above code implementation here. Performance of MMDetection With many competing models, users struggle to choose the right one for their requirements. MMDetection behaves as a benchmarking platform and compares different models under identical conditions. Benchmarking popular models on a bounding box prediction task (Source). Benchmarking popular models on an object masking task (Source). Benchmarking popular commercial GPUs with MMDetection on three models (Source). MMDetection toolbox outperforms recent codebases, namely, maskrcnn-benchmark, Detectron and SimpleDet. MMDetection is presently state-of-the-art with a huge model collection. The efficiency and performance of MMDetection is far better than any other codebase. Comparison of MMDetection with competing codebases based on training, inference, memory usage and evaluation metrics (Source). References for Further reading: Github repositoryResearch PaperOfficial documentation Official site","excerpt":"MMDetection is a Python toolbox built as a codebase exclusively for object detection and instance segmentation tasks","categories":["Deep Tech"],"tags":["bounding boxes","detection","faster-rcnn","Guide","image segmentation","mask-rcnn","Object Detection"],"author_name":"Rajkumar Lakshmanamoorthy","publish_date":"2021-04-24T16:00:00","publication_year":"2021","word_count":1392,"keywords":["bounding boxes","NumPy","TPU","image segmentation","faster-rcnn","mask-rcnn","AI","ML","PyTorch","RAG","Colab","Ray","detection","Aim","Object Detection","object detection","Guide"],"extracted_tech_keywords":["AI","ML","Aim","Ray","PyTorch","Colab","NumPy","RAG","object detection","TPU"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/guide-to-mmdetection-an-object-detection-python-toolbox\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10064422,"title":"Council Post: Beyond Explainable AI—How to infuse trust in AI systems","content":"According to McKinsey, AI will add USD 13 trillion to the global economy by 2030. Despite the explosive growth, companies struggle to scale up their AI efforts and bring in bias-free algorithms or products. The widespread adoption of AI is necessitating the need for new standards and practices. The area of Explainable AI (XAI) is the most popular method to tackle the black-box nature of machine learning models. However, it is time to explore newer methods to build trust in AI systems. As the next step in building trust, the researchers are focused on the bias assessment of AI systems. This article discusses a framework to evaluate trust at various stages of a typical data science workflow. We make a case for the need to go beyond XAI on the bias assessment for automated decision-making systems. Why we need to move beyond XAI Biases are an integral part of the human experience. Thanks to our skewed judgement, AI has inherited these biases from us. It becomes important for leaders to understand organisational and cultural barriers AI initiatives face and work to mitigate them by educating the workforce on ethics, changing the traditional mindsets and bringing in innovation with no biases. We have seen plenty of racial, gender, and other biases in AI systems. For example, e-commerce giant Amazon rescinded a model–used to score job applicants–for penalising women. Content personalisation and ad ranking systems have also been in the dock for racial and gender profiling. The bias starts way before the product gets deployed. In machine learning, this is known as ‘model bias.’ Machine learning models with 100 percent accuracy don’t always mean the model is stable and learning; ML practitioners and data scientists should also check bias in the data, algorithm, and model before production. XAI is a good tool to describe the model predictions in a human interpretable way. For example, feature attribution methods such as SHAP and LIME, in most cases, can render black-box machine models explainable. But can XIA alone create trustworthy AI? To build trust in AI, we need tools to mitigate bias. Ethical AI is mainly concerned with bias detection before and after the model predictions. However, Ethical AI’s bias mitigation prowess is minimal, and we have yet to develop robust and easily deployable techniques. Below, we look at phases of a typical data science workflow where the trust can be well-defined and evaluated. Most of the discussions are centred on introducing an empirical framework for evaluating trust in AI systems and the metrics used for this purpose at different stages of data science workflow. Any data science activity such as technical robustness, safety, data governance, accountability, societal and environmental well-being can be seamlessly integrated into our framework. Human-AI trust platform is inspired by a research article by Alon Jacovi et.al on the topic of formalising trust in AI. First, we define the concept of contractual trust. Contractual Trust Let’s start with a formal definition of trust given by Mayer et.al. If A anticipates that B will act in A’s best interest, and accepts vulnerability to B’s actions, then A trusts B. Here, A can be an organisation\/user\/system that entrusts another organisation\/ data scientist (B) to make a trustworthy AI model. However, there are risks involved in the collaboration between A and B. However, A anticipates that B will execute their transaction in the best interest of A despite the vulnerability in the processes executed by B. This vulnerability in B is probable, and A is aware of those. For example, having a fair AI model can be A’s anticipation, and a drop in model performance can be B’s vulnerability in the course to achieve a fair AI model. The notion of trust exists if and only if these anticipations and vulnerabilities mutually co-exist and are acknowledged by both A and B. Contractual trust is a concept that dictates and quantifies trustworthy AI. This is done by incorporating a set of phases in the data science workflow. The client and data science organisation agree about their orchestration and purpose. In other words, the client can set their anticipation about the phases to see if they meet predefined conditions or states. At the same time, the data science organisation can set the probable vulnerabilities that can happen in the pursuit of anticipation. Once the anticipation and vulnerabilities are defined and acknowledged, we can form an action plan consisting of several checkpoints for evaluation at different stages of the data science workflow. The checkpoint can act as an evaluation mechanism for the data science team by the client. The checkpoints can be given different weightage according to the client requirements. For example, if a client needs data anonymisation as the principal constraint, it can have more weightage than the model performance or model bias. One obvious advantage of the checkpoint is that we can evaluate each stage with client participation and ensure transparency in the development process. This way, we can assign a Human-AI trust score for the entire data science pipeline. Advantages Two-way solution: The client and data science team are part of the framework.The clients can clearly distinguish their expectations, referred to as anticipation in the framework.The data science team can list out the possible flaws that can happen, referred to as vulnerabilities.The trust framework is well-defined over a list of actions.The entire framework takes the interest of both parties into account.The checkpoints will enable stage-by-stage evaluation by the clients and ensures transparency.Based on the client’s evaluation, checkpoints can be updated, or new ones inserted.The framework has a scoring mechanism that evaluates the entire workflow with both parties involved. This article is written by a member of the AIM Leaders Council. AIM Leaders Council is an invitation-only forum of senior executives in the Data Science and Analytics industry. To check if you are eligible for a membership, please fill the form here.","excerpt":"Contractual trust is a concept that dictates and quantifies trustworthy AI.","categories":["AI Features"],"tags":["Contract","Ethical AI","Explainable AI","Responsible AI","Trustworthy AI"],"author_name":"Shashank Shekhar","publish_date":"2022-04-07T11:00:00","publication_year":"2022","word_count":977,"keywords":["data science","Go","machine learning","AWS","AI","R","Ethical AI","ML","Responsible AI","Trustworthy AI","Aim","analytics","Explainable AI","xAI","Contract"],"extracted_tech_keywords":["AI","machine learning","ML","data science","analytics","xAI","Aim","AWS","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/council-post-beyond-explainable-ai-how-to-infuse-trust-in-ai-systems\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10076895,"title":"Is Musk Out Shopping for Superapps?","content":"The Musk-Twitter saga has come to a surprising end. Tesla chief Elon Musk has agreed to move forward with his original plan to buy social networking platform Twitter bringing an end to the legal battle. Twitter alleged that the 51-year-old billionaire was anxious about footing the hefty bill for purchasing the platform during a downturn in the tech share market. Even if there was any truth to that argument, Musk did have ideas for the app. Twitter probably accelerates X by 3 to 5 years, but I could be wrong— Elon Musk (@elonmusk) October 4, 2022 On October 5th, he tweeted, “Buying Twitter is an accelerant to creating X, the everything app.” This isn’t the first time Musk has teased the concept of an everything app – a superapp. In an appearance on the All-In Podcast in May, when asked what his version of Twitter would look like, he gushed about China’s WeChat. “WeChat is a good model — and basically includes functions of Twitter, PayPal and many others. It is an excellent app, we don’t have anything like it outside of China,” he said. WeChat app interface, Source: Medium Lack of a superapp in the US Text messages between Musk and his advisers, revealed during the course of his lawsuit, also revealed his desire to create a copycat of the Tencent app. In a message sent to his brother Kimbal, Musk said that he had an idea for a “blockchain social media system that does both payments and short texts or links like Twitter”. Interested users would have to pay a small amount to register their message on the chain. This would secure the platform and “cut out a vast majority of spam and bots”. Noting the lack of a WeChat counterpart in the US, he added, “You basically live on WeChat in China because it’s so helpful, so useful to daily life. I think if we achieve that or come even close to that with Twitter, that would be a success.” Launched in 2011 as a messaging app, WeChat grew to function as a hybrid between platforms like WhatsApp, Uber and eBay. With more than a billion monthly active users, the app supports voice calls, video calls, features with real-time location sharing and posting stories a la Instagram. WeChat also has an option that allows users to make international calls with a localisation feature that translates messages into 20 different languages. Range of payment services offered by Paytm and PhonePe, Source: TechCrunch A potential trillion $ company Significantly, superapps aren’t as pervasive in the West as in Asia. In China, apps like WeChat and Alipay (which combines payments, food delivery services, loans and banking) were popular due to the rise of cheaper phones with limited storage. Since Chinese tech giants like Tencent and Alibaba were bound by the number of apps that could be downloaded, a bunch of ‘superapps’ sprung up. These apps were convenient for users to switch between rather than complicated interfaces. While other tech giants have tried to chase the superapp dream, none has been as successful as WeChat. Uber, whose food delivery business UberEats has become a gold mine for the ride-hailing company, has brought back its businesses to its main app after splitting them in the past. Social media platform, Snapchat also opened up its app in 2020 and allowed developers to run trials for gaming inside its messaging app. (Coincidentally, Tencent is also a prominent investor in Snapchat). PayPal too has transformed itself from a single button payment processing middleman into an umbrella payment platform. The all-in-one app integrates its e-commerce business, financial services and payment services now all under one platform. The app has also expanded to include cryptocurrency investment which draws in more customer engagement. There is a reason for Musk’s interest in superapps. WeChat generated around USD 17.5 billion in revenue last year through advertising and digital services. Neighbouring Asian countries like Singapore’s Grab and Indonesia’s Gojek have been successful in pursuing the same route. It’s a superapp war in India The superapp bug has also bitten India’s most popular services with conglomerates like Tata and Reliance making acquisitions to stay in the race. New entrants like Tata Neu want to become a one-stop shop for users to book tickets, shop online and buy groceries while Reliance already has its JioMart. Paytm, which is backed by Alibaba, also acts like a payment platform now with financial investment services, option to book flights, movie tickets and online games, banking and consumer finance. Applications like Truecaller, which is mainly an app to screen strangers, later added four features, including texting, recording phone calls and mobile payment. PhonePe, owned by Flipkart, also has a bunch of mini apps under one platform. Some of those include ride-hailing service Ola, hotel booking service Oyo and travel booking service MakeMyTrip. At the All-In Summit, Elon spoke glowingly about expanding Twitter’s features to what WeChat offers – pay bills, hail rides, gov ID, etc.What’s with the tech utopian desire to build things with a single point of failure and facilitating the inexorable march toward dystopia?— Melissa Chen (@MsMelChen) May 22, 2022 Regulatory concerns But even with all the ingredients to make superapps successful, regulatory issues continue to rear their head. Superapps operate on a treasure trove of data which raises concerns about their management and security. There are several other legal issues that these apps find themselves entangled in. According to section 101 of the Consumer Protection Act, “parties cannot sell their products on the same marketplace”. This would mean that Tata’s e-commerce platform Blinkit cannot sell Tata Salt and neither can Reliance sell its own retail products on JioMart. These concerns in India only multiply manifold in the US. Congress has been quick to pull up tech giants as soon as they start wielding more power. A superapp in this scenario would be the perfect red flag to raise the alarms around data collection and handling. It is unlikely that in the current environment, the rise of a superapp will be seen as trustworthy.","excerpt":"In an appearance on the All-In Podcast in May, when asked what his version of Twitter would look like, Musk gushed about China’s WeChat","categories":["AI Features"],"tags":["Interviews and Discussions"],"author_name":"Poulomi Chatterjee","publish_date":"2022-10-11T10:00:00","publication_year":"2022","word_count":1009,"keywords":["Go","AWS","AI","cloud_platforms:AWS","RPA","programming_languages:R","Git","RAG","Rust","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","RAG","AWS","R","Go","Rust","Git","RPA","cloud_platforms:AWS","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/is-musk-out-shopping-for-superapps\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10160761,"title":"Sam Altman Left Disappointed by o1 Pro","content":"OpenAI CEO Sam Altman said the company was not making money on its most expensive offering, the $200 ChatGPT Pro subscription. While this should not be surprising, given that OpenAI is a loss-making company, Altman revealed that he expected otherwise. “We are currently losing money on OpenAI Pro subscriptions! People use it much more than we expected,” he said in a post on X. “I personally chose the price and thought we would make some money,” he added. These expectations align with the company’s plans to further transition away from a non-profit business model. In December of last year, OpenAI announced the $200 ChatGPT Pro plan. This plan includes all the features of the Plus plan and access to the additional o1 Pro mode, which is said to use “more compute for the best answers to the hardest questions”. The model has been met with praise ever since its debut. Pretty high praise for o1 pro out on reddit pic.twitter.com\/nObgXcQ5Ja— Alex Teichman (@alex_teichman) December 31, 2024 There was initial scepticism when the high-price tag subscription was launched, but Altman clarified that it’s not for everybody and a majority of the users will not need it. “Most users will be very happy with the o1 in the plus tier,” he said, indicating that the $20 ChatGPT Plus plan would suffice for the majority. Altman’s statements today indicate that people are open to paying big bucks for OpenAI’s models, which costs the company much more in computing and inference. Eventually, it seems like OpenAI will release more expensive plans. It’s recently announced o3 model ranks atop all benchmarking tests but has high costs. The model scored the highest in the ARC-AGI benchmark but costs a whopping $1000 per task. Recently, OpenAI CFO Sarah Friar suggested that the company may charge $2000 a month to replace humans with a PhD-level assistant. Overall, profitability will mean more than ever for OpenAI, which also dictates the definition of artificial general intelligence (AGI). As of now, the company defines it as “a highly autonomous system that outperforms humans at most economically valuable work.” When they announce AGI, Microsoft, one of its major investors, will lose access to OpenAI’s most powerful models. Microsoft does not want to settle for a subjective, vague definition of AGI, and both companies have reportedly agreed that AGI will only be achieved when OpenAI earns $100 billion in profits. In 2024, OpenAI reported a $5 billion loss on $3.7 billion in revenue. Reports suggest that it is unlikely to become profitable before 2029. However, reports also suggested that OpenAI was trying to remove the ‘AGI’ clause from their agreement. While OpenAI has yet to officially announce the ‘profit-driven’ definition for AGI, Altman said in a blog post that the company was ready to announce it in 2025. “We are now confident we know how to build AGI as we have traditionally understood it. We believe that, in 2025, we may see the first AI agents “join the workforce” and materially change the output of companies,” he said.","excerpt":"“I personally chose the price and thought we would make some money.”","categories":["AI News"],"tags":["AI (Artificial Intelligence)","OpenAI"],"author_name":"Supreeth Koundinya","publish_date":"2025-01-06T12:19:36","publication_year":"2025","word_count":505,"keywords":["ChatGPT","TPU","OpenAI","AI","programming_languages:R","GPT","AI agents","R","llm_models:ChatGPT","AI (Artificial Intelligence)","llm_models:GPT"],"extracted_tech_keywords":["AI","ChatGPT","OpenAI","TPU","R","GPT","AI agents","llm_models:GPT","llm_models:ChatGPT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/sam-altman-left-disappointed-by-o1-pro\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":11105,"title":"Start-up of the Week – SuperFan.Ai, a Chatbot for fans to connect with Influencers","content":"Founders of SuperFan.ai With the start-up fever on, Analytics India Magazine as a new segment introduces Start-up of the Week. Through this segment we will be bringing you inspiring stories of innovative, different start-ups who have dared to make a difference. This week we present one such start-up SuperFan.Ai. Well, let’s find out their story. What is SuperFan.Ai? SuperFan.Ai revolutionizes the way an influencer connects with his audience using a personal chat bot. It lets influencers simply chat to their fans directly one-on-one, like they would with any of their friends using artificial intelligence. Superfan.ai provides a personal connect by Influencers with their followers using AI. Their core engine is a chat platform powered by artificial intelligence using NLP. The Super Bots of SuperFan.ai The chatbots at SuperFan.ai can engage and automate conversations with end-users at scale on a one-on-one basis, across messaging & chatplatforms. Their tool allows any influencer with an engaged community around his\/her passion to connect individually with his audience. The bot essentially allows: A 1-0-1 & 2-way connect \/ communication with the follower – giving fans & followers a voice to interact with the influencer & helping the influencer find & thank their true & ardent followers. It helps an influencer to increase engagement with their audiences & convert “at-risk” fans into “superfans”, thus increasing his own brand value. It enables an influencer to send the right message to the right audience at the right time through demographic, location & other details available with a success rate of 95% of the time, messages being opened & read (as opposed to 65% of the time message is lost in the feed) to the influencer thru back-end statistics provided. Enables E-commerce opportunities for a Brand & Advertisement Revenue from “Spot on the Bot” for an Influencer Advanced Bot Analytics The superfan.ai dashboard lets clients view charts and graphs on their follower demographics, locations, engagement levels etc. so they can hone in on their most important followers. The chatbots put the fun back into interactions. Also, Chat data is the analytics gold mine. Identifying what people really want through their queries is what makes chatbots an ideal tool for engagement. Beyond their general analytics offerings, you can take advantage of machine learning and advanced statistical methods to improve your content & engagement. They will show you the most interesting conversations that you’ve had with your audience and break down the common responses to the questions your bot asks. You will be armed with the tools necessary to both dig deep into conversations or to get a high-level view of how your content is performing & audience engagement. The beauty of SuperFan’s bots is they’re dynamic. They can provide filters that then route individuals to the right information and then proactively respond when new information is available. These bots can ultimately replace personal websites, provide a new way to contact people, and could save countless hours to filter through requests and inquiries from fans & followers and respond personally to each such question. The company creates chatbots that can be published to popular messaging platforms like Facebook, Skype, Slack, Viber, Telegram & many more. So go beyond just text to take advantage of rich platform-specific features like cards, structured message, images, videos and sound effects. The bot can interact in multiple languages including English, Hindi, Tamil, Telegu, Marathi, Gujarati & Hinglish. SuperFan’s Target market SuperFan targets any influencer who wants to connect with his audience directly. The influencer could be a celebrity, production house (movies, tv shows), sports personalities, sports teams, political influencers, political parties, individual artists, Brands, SME’s, Corporates etc. SuperFan.ai’s current client base ranges from production houses, individual celebrities, radio channels, brands, corporates, to politicians to create their chat-bots for gaining a deeper & meaningful interactions with their followers. The Brain behind SuperFan.ai The Company has been co-founded by Snehal Dhruve & Bineet Desai. Snehal has worked on A.I. based technologies for the last 10 years in US with Fortune 500 companies & has tech expertise in Big Data, AI, API, iOS and Android. Bineet holds a degree in Masters in Management Studies (Finance major) from NMIMS, Mumbai & has been leading business development in Corporate Finance & Corporate Banking with various multi-national banks & Companies in India & abroad over the last 15 years.","excerpt":"With the start-up fever on, Analytics India Magazine as a new segment introduces Start-up of the Week. Through this segment we will be bringing you inspiring stories of innovative, different start-ups who have dared to make a difference. This week we present one such start-up SuperFan.Ai. Well, let’s find out their story. What is SuperFan.Ai? […]","categories":["IT Services"],"tags":["AI India","chatbot india"],"author_name":"Srishti Deoras","publish_date":"2016-11-04T13:11:38","publication_year":"2016","word_count":717,"keywords":["big data","Go","API","machine learning","artificial intelligence","AI","chatbots","NLP","analytics","AI India","R","chatbot india"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","NLP","analytics","chatbots","R","Go","API","big data"],"url":"https:\/\/analyticsindiamag.com\/it-services\/start-week-superfan-ai-chatbot-fans-connect-influencers\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":53294,"title":"Deep Dive: How Anuj Kacker Of MoneyTap Uses AI\/ML For Financial Inclusion","content":"Lending by Fintech, startups have added new dimensions to the financial intermediation process and has boosted financial inclusion, especially by helping borrowers. The Fintech lending industry is constantly innovating and is set to grow to $100 billion by 2023. Also, a growing number of non-financial startups such as Ola and Mi are trying to become lending players as well. To know more about this industry and for this week’s Deep Dive, Analytics India Magazine got in touch with Anuj Kacker, COO and Co-Founder of MoneyTap. Bangalore-based MoneyTap was launched in 2015 and is currently available in 44 cities in India. Founded by Anuj Kacker and Bala Parthasarathy, MoneyTap caters to the credit needs of middle-class India by providing quick, flexible and hassle-free credit on mobile at affordable interest rates. Also, currently, the company is working on streamlining processes, training, and operations to continue providing the best combination of quick credit and smooth process to the respective customers. Use of AI\/ML The credit line product of MoneyTap uses AI and ML to integrate various functionalities such as customer targeting, onboarding document verification, analysing risk and repayment behaviour patterns as well as collections. The AI and ML algorithms are used to generate alternate data scorecards, which mainly form the basis for lending in addition to CIBIL scores. Customers download the MoneyTap app for free, and then a patent-pending propriety ChatBot takes the customer through the application process for the credit line. The smart AI and ML algorithms help them get an instant credit decision within minutes. Core Tech Stack At the core, the tech stack of the company includes databases such as Postgres and Redis, server stacks such as Java, Python and Vue.js and frontend clients like Android, iOS, among others. Size of Engineering Team Currently, the company has an engineering team of 35+ employees which includes both data scientists and developers. And further, they are looking to increase the engineering team size to about 100 by next year. Tackling Hiring Phase To be the right candidate, the company expects the candidate to not only have the required expertise in the field but also to be a great culture fit. The hiring process involves campus hiring, and hackathons, as well as by employee recommendations. While hiring a data scientist, the company looks for qualities like curiosity for data and an ability to build strong hypotheses with sustainable proof. Achieving Customer’s Trust As a customer’s trust and satisfaction are something tough to achieve, the customer success team at the company provides a smooth lending experience to the customers. The customers are usually kept in touch through various sources like email, social media, SMS, push notifications and in-app communication. Roadmap MoneyTap is a full-stack consumer lending company which is continuously striving to bring the most seamless and hassle-free credit experience to the consumers. The company has created a total loan book of ₹1,000 crores and has achieved a disbursal run rate of ₹2,500 crores a year. Kacker added, “We plan to expand the role of our tech and data teams to combine UI & AI – this will lead to customised credit solutions for diverse customer demographics. Our vision is the financial inclusion of underserved segments in India.”","excerpt":"Lending by Fintech, startups have added new dimensions to the financial intermediation process and has boosted financial inclusion, especially by helping borrowers. The Fintech lending industry is constantly innovating and is set to grow to $100 billion by 2023. Also, a growing number of non-financial startups such as Ola and Mi are trying to become […]","categories":["AI Features"],"tags":["Deep Dive"],"author_name":"Ambika Choudhury","publish_date":"2020-01-07T15:34:26","publication_year":"2020","word_count":534,"keywords":["Go","startup","AI","ML","Python","analytics","Rust","R","Java","Deep Dive","Redis"],"extracted_tech_keywords":["AI","ML","analytics","Redis","Python","R","Go","Rust","Java","startup"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/deep-dive-how-anuj-kacker-of-moneytap-uses-ai-ml-for-financial-inclusion\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10020592,"title":"Guide To Real-time Object Detection Model Deployment Using Streamlit","content":"Streamlit is an open-source Python library using which data scientists can build Machine learning from scratch. It also enables building various ML tools for visualization and analysis of the experiments’ data and output in an interactive app framework. It is used in the ML domain for several projects ranging from simple to complex ones. This article illustrates object detection using Streamlit. If you are unfamiliar with Streamlit, read this article before proceeding and have a preliminary understanding of the framework. Here, we build a Streamlit app in which you can detect object(s) in images of your choice. We have used YOLOv3 object detection model and MS-COCO image dataset for training the model. Python 3.8.5 version and Anaconda prompt have been used for this implementation.The step-wise explanation of the source code (.py file) is as follows: (Explanation about various methods of Streamlit API used in the code can be found here). Download the YOLOv3 model’s configuration file and weights file.Download the file containing output labels of MS-COCO dataset (weblink).Import required libraries and modules #Import streamlit import streamlit as st #Import NumPy and OpenCV import numpy as np import cv2 #Import Matplotlib for visualization import matplotlib.pyplot as plt #Import Image module of PIL library for handling images from PIL import Image 4) Define a function for object detection def obj_detection(my_img): st.set_option('deprecation.showPyplotGlobalUse', False) \/*streamlit.set_option sets config options. streamlit.pyplot() now requires a figure to be provided. Setting deprecation.showPyplotGlobalUse to False will disable the deprecation warning.*\/ column1, column2 = st.beta_columns(2) \/*streamlit.beta_columns() inserts containers laid out as side-by-side columns*\/ #Display subheading on top of input image column1.subheader(\"Input image\") #streamlit.subheader() st.text(\"\") #streamlit.text() writes preformatted and fixed-width text #Display the input image using matplotlib plt.figure(figsize = (16,16)) plt.imshow(my_img) column1.pyplot(use_column_width=True) 5) Instantiate YOLO model (object_detection() function continued) neural_net = cv2.dnn.readNet(\"PATH_1\", \"PATH2\") \/*PATH_1 and PATH_2 are the paths of locations where you have stored the yolov3.weights and yolov3.cfg files respectively*\/ labels = []   #Initialize an array to store output labels with open(\"PATH_3\", \"r\") as file: \/*open coco.names file downloaded in step(2) in read(“r”) mode. PATH_3 is the path of location where you have stored the output labels’ file*\/ labels = [line.strip() for line in file.readlines()] #strip() method removes leading and trailing spaces from the label strings names_of_layer = neural_net.getLayerNames() #Store the names of model’s layers obtained using getLayerNames() of OpenCV output_layers = [names_of_layer[i[0]-1] for i in neural_net.getUnconnectedOutLayers()] #getUnnnectedOutLayers() returns indexes of layers with unconnected output colors = np.random.uniform(0,255,size=(len(labels), 3)) #RGB values selected randomly from 0 to 255 using np.random.uniform() # Image loading newImage = np.array(my_img.convert('RGB')) #Convert the image into RGB form img = cv2.cvtColor(newImage,1) #cvtColor() #Store the height, width and number of color channels of the image height,width,channels = img.shape 6)  Detect the objects (object_detection() method continued) #Convert the images into blobs using blobFromImage() blob = cv2.dnn.blobFromImage(img, 0.00391, (400,400), (0,0,0), True, crop = False) \/*Above parameters are (image object, scale factor, size, mean(mean subtraction from each layer), swapRB(Blue to red), crop enabling)*\/ neural_net.setInput(blob) #Feed the model with blobs as the input outputs = neural_net.forward(output_layers) \/*forward() runs forward pass to compute output of layer named by ‘output_layers’*\/ classID = [] #Array to store output labels confidences = [] #Array to store confidence score boxes =[] #Array to store bounding boxes dimensions 7) Display output layer’s information (object_detection() continued) for op in outputs: for detection in op: scores = detection[5:] #confidence score #select output class having maximum probability (confidence score) class_ID = np.argmax(scores) #Update ‘confidence’ variable with the score of above selected output label confidence = scores[class_ID] if confidence > 0.5: \/*If confidence score exceeds 0.5 (>50% probability), it means an object has been detected. Get its dimensions  *\/ center_x = int(detection[0] * width) center_y = int(detection[1] * height) #centre of object w = int(detection[2] * width) #’w’ is width of detected object while ‘width’ is original width of image h = int(detection[3] * height) #’h’ is height of detected object while ‘height’ is original height of image # Calculate coordinates of bounding box #x-coordinate of top-left corner of box x = int(center_x - w \/2) #y-coordinate of top-left corner of box y = int(center_y - h\/2) #Organize the detected objects in an array boxes.append([x,y,w,h]) confidences.append(float(confidence)) classID.append(class_id) 8) Adjust confidence threshold and NMS (Non-Maximum Suppression) threshold score_threshold = st.sidebar.slider(\"Confidence_threshold\", 0.00,1.00,0.5,0.01) nms_threshold = st.sidebar.slider(\"NMS_threshold\", 0.00, 1.00, 0.5, 0.01) \/*streamlit.slider() inserts a slider widget. Its parameters are (textual label displayed, min vale, max value, stepping interval)*\/ indexes = cv2.dnn.NMSBoxes(boxes, confidences,score_threshold,nms_threshold) #NMSBoxes() performs NMS given boxes and corresponding scores print(indexes) The sliders will appear in the UI as follows: 9) Draw the bounding boxes (object_detection() continued) items = []  #Array to store label of detected object(s) for i in range(len(boxes)): if i in indexes: #get dimensions of ith bounding box to be formed x,y,w,h = boxes[i] #get name of detected object label = str.upper((labels[classID[i]])) color = colors[i] #color of ith bounding box \/*Form rectangular box. Its parameters are (image, start point, end point, color, thickness) *\/ cv2.rectangle(img,(x,y),(x+w,y+h),color,3) items.append(label) #Add the output label of bounded object 10) Display the detected objects with anchor boxes (object_detection() continued) st.text(\"\") #preformatted and fixed-width text column2.subheader(\"Output image\") #Title on top of the output image st.text(\"\") #Plot the output image with detected objects using matplotlib plt.figure(figsize = (15,15)) plt.imshow(img) #show the figure column2.pyplot(use_column_width=True) #actual plotting if len(indexes)>1: #Text to be printed if output image has multiple detected objects st.success(\"Found {} Objects - {}\".format(len(indexes),[item for item in set(items)])) else: #Text to be printed if output image has a single detected object st.success(\"Found {} Object - {}\".format(len(indexes),[item for item in set(items)])) #streamlit.success() prints the success message 11) Main method (which marks beginning of the code’s execution) def main(): st.title(\"Welcome to Streamlit app\") #Title displayed in UI using streamlit.title() #Display some text on UI using streamlit.write() st.write(\"You can view real-time object detection done using YOLO model here. Select one of the following options to proceed:\") \/*2 options will be seen on UI. On selecting first option (“See an illustration”), you will see an input image (which we have already fed through the code) having a cat and a dog, and its corresponding output with the objects detected. It is only meant for illustrating the output of YOLOv3. On selecting the second option (“Choose an image of your choice”) will enable you to browse an image file from your machine. YOLOv3 will then run on that custom image for detecting objects.*\/ choice = st.radio(\"\", (\"See an illustration\", \"Choose an image of your choice\")) #streamlit.radio() inserts a radio button widget The title, text and radio buttons will appear in the UI as follows: #If user selects 2nd option: if choice == \"Choose an image of your choice\": image_file = st.file_uploader(\"Upload\", type=['jpg','png','jpeg']) \/*streamlit.fileUploader() shows a file uploader widget. By default, size limit for uploaded files is  200MB. You can configure it using the server.maxUploadSize config option. Here, “Upload” will be the label of uploader widget and the allowed file types have been mentioned in types[] array*\/ The file uploader image will appear in the UI as follows: if image_file is not None: #if a file has been uploaded my_img = Image.open(image_file)  #open the image #perform object detection on selected image obj_detection(my_img) #If user selects 1st option elif choice == \"See an illustration\": #display the example image my_img = Image.open(\"PATH_OF_EXAMPLE_IMAGE\") #perform object detection on the example image obj_detection(my_img) Following is the example image we have taken: 12) Call the main() method if __name__ == '__main__': main() How to run the Streamlit code? In your terminal, install the streamlit library as follows: pip install --upgrade streamlit Suppose your source code file’s name is ‘appcode.py’ Then run the streamlit app from the terminal as follows: streamlit run FILEPATH (where FILEPATH is the path of the location where you have stored the appcode.py file) Sample output in Streamlit app Once the app gets launched (with the first radio button selected by default), it appears as follows: On selecting the second radio button, it will ask you to upload an image of your choice, and the UI will look as follows: Suppose the selected image is: The output image after object detection: The UI will also show the number of objects detected and the classes they belong to, below the input and output images as follows: Find the source code file ‘appcode.py’ here. References Official websiteGitHub repositoryDocumentationRelated articles: article1  article2","excerpt":"Streamlit is an open-source Python library using which data scientists can build Machine learning from scratch. It also enables building various ML tools for visualization and analysis of the experiments’ data and output in an interactive app framework. It is used in the ML domain for several projects ranging from simple to complex ones. This […]","categories":["Deep Tech"],"tags":["Object Detection","streamlit"],"author_name":"Nikita Shiledarbaxi","publish_date":"2021-02-22T13:00:00","publication_year":"2021","word_count":1368,"keywords":["NumPy","machine learning","TPU","AI","ML","streamlit","OpenCV","Ray","object detection","Object Detection","Streamlit","Matplotlib"],"extracted_tech_keywords":["AI","machine learning","ML","Ray","Streamlit","OpenCV","NumPy","Matplotlib","object detection","TPU"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/guide-to-real-time-object-detection-model-deployment-using-streamlit\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":68511,"title":"The Different Types Of Hardware AI Accelerators","content":"An AI accelerator is a kind of specialised hardware accelerator or computer system created to accelerate artificial intelligence apps, particularly artificial neural networks, machine learning, robotics, and other data-intensive or sensor-driven tasks. They usually have novel designs and typically focus on low-precision arithmetic, novel dataflow architectures or in-memory computing capability. As deep learning and artificial intelligence workloads grew in prominence in the last decade, specialised hardware units were designed or adapted from existing products to accelerate these tasks, and to have parallel high-throughput systems for workstations targeted at various applications, including neural network simulations. As of 2018, a typical AI integrated circuit chip contains billions of MOSFET transistors. Hardware acceleration has many advantages, the main being speed. Accelerators can greatly decrease the amount of time it takes to train and execute an AI model, and can also be used to execute special AI-based tasks that cannot be conducted on a CPU. Here, we look at the most popular hardware AI accelerators. Graphics Processing Unit (GPU) Graphics processing unit is a specialised chip that can do rapid processing, primarily for the purpose of rendering images. They have become a key part of modern supercomputing.  They have been used in growing new hyperscale data centres and have become accelerators, speeding up all sorts of tasks – from encryption, to networking, to AI. GPUs have sparked an AI boom, become a key part of modern supercomputers, and continue to drive advances in gaming and pro graphics. Modern GPUs are great at handling computer graphics and image processing. Their extremely parallel structure makes them more valuable than general-purpose central processing units (CPUs) for algorithms that process huge blocks of data in parallel.  Multiple GPUs are utilised on supercomputers, on workstations to expedite processing multiple videos at once and 3D rendering, for VFX and for simulations, and in AI for training workloads. In contrast to a CPU, NVIDIA GPUs, for example, contain chips that have what are known as CUDA Cores, and each one of these cores is a tiny processor that can execute some code. Vision Processing Unit (VPU) A vision processing unit (VPU) is a rising class of microprocessor, and a particular type of AI accelerator intended to quicken machine vision tasks. The vision processing unit is reported as more fitting for performing various kinds of machine vision algorithms. These tools may be designed with particular resources for capturing visual data from cameras, and are built for parallel processing. Some of these tools are low power and high performance and may be plugged into interfaces that enable programmable use. Vision processing units are fit for performing machine vision algorithms such as CNN (convolutional neural networks), SIFT (Scale-invariant feature transform) and other similar ones. They may include direct interfaces to take data from cameras (bypassing any off-chip buffers) and have a greater emphasis on on-chip data flow between many parallel execution units. The factors driving VPUs include the growing adoption of smartphones, increasing adoption of edge AI, and expanding demand for advanced computing capacities for computer vision. One example of a VPU is Intel’s Movidius Myriad X VPU which is being used in many of the edge devices. Target markets are robotics, the internet of things, new classes of devices for AR\/VR, and integrating machine vision acceleration into smartphones and other mobile devices. Field-Programmable Gate Array (FPGA) A field-programmable gate array (FPGA) is an integrated circuit (IC) made to be configured by a customer or a designer after manufacturing, which is why it is called “field-programmable”.  FPGAs include a range of programmable logic blocks and a hierarchy of “reconfigurable interconnects” that enable the blocks to be connected together like many logic gates that can be inter-wired in various configurations. FPGAs can be beneficial over GPUs in terms of interface flexibility and enhanced by the integration of programmable logic with CPUs and standard peripherals. GPUs, on the contrary, are optimised for parallel processing of floating-point operations utilising thousands of small cores. They also provide big processing capabilities with larger power efficiency. FPGAs, which can do a wide range of logical functions simultaneously, are being considered unsuitable for emerging technologies such as self-driving cars or deep learning applications. Today’s field-programmable gate arrays (FPGAs) have big resources of logic gates and RAM blocks to implement complex data computations. Due to their programmable characteristics, FPGAs are a perfect fit for many different markets. FPGAs can be reprogrammed to the wanted application or functionality needs after manufacturing. This feature separates FPGAs from Application-Specific Integrated Circuits (ASICs), which are custom produced for particular design tasks. FPGAs are increasingly applied to expedite AI workloads in data centres for jobs like machine learning inference. Many hardware companies like Xilinx have launched their FPGA products as latest datacenter accelerator cards as satisfying increasing business demand for heterogeneous architectures and performance advances as customers work on more AI workloads. Application-Specific Integrated Circuit (ASIC) A whole category of AI hardware accelerator is gaining prominence, with something called an application-specific integrated circuit (ASIC). ASICs employ strategies such as optimised memory use and the use of lower precision arithmetic to accelerate calculation and increase the throughput of computation. Some adopted low-precision floating-point formats used AI accelerations are half-precision and the bfloat16 floating-point format. Hardware acceleration is used to speed up the computing processes present in an AI workflow. For example, Intel released Nervana, an ASIC for inference and support for a large amount of parallelisation in server settings. It has also revamped the chip structure considerably and built them on a 10nm manufacturing process. ASICs hold numerous advantages, with the main being speed. Accelerators can minimise the amount of time it takes to train and execute an AI model, and can also be used to execute special AI-based tasks. Tensor Processing Unit (TPU) A tensor processing unit (TPU) is a specialised circuit that implements all the necessary control and arithmetic logic necessary to execute machine learning algorithms, typically by operating on predictive models such as artificial neural networks (ANNs) or random forests (RFs). Google launched TPUs in the year 2016. TPUs, unlike GPUs, are custom-designed to handle operations like matrix multiplications in neural network training. The power Google TPUs can be reached in two types, which are cloud TPU and edge TPU.  Cloud TPUs may be accessed from Google Colab notebook, which gives users with TPU pods which sit on Google’s data centres. On the other hand, edge TPU is a custom-built development kit which can be utilised to create specific applications. Tensors are multi-dimensional arrays or matrices and are fundamental units which can hold data points like weights of a node in a neural network in a row and column format. Basic calculation operations are performed on tensors. TPUs were utilised in the known DeepMind’s AlphaGo, where AI beat the world’s best Go player. It was also applied in the AlphaZero system, which produced Chess, Shogi and Go-playing programs.","excerpt":"An AI accelerator is a kind of specialised hardware accelerator or computer system created to accelerate artificial intelligence apps, particularly artificial neural networks, machine learning, robotics, and other data-intensive or sensor-driven tasks. They usually have novel designs and typically focus on low-precision arithmetic, novel dataflow architectures or in-memory computing capability.  As deep learning and artificial […]","categories":["IT Services"],"tags":["cloud adoption in emerging markets","different types of analytics","GPUs","types of analytics"],"author_name":"Vishal Chawla","publish_date":"2020-06-30T11:00:00","publication_year":"2020","word_count":1143,"keywords":["artificial intelligence","machine learning","TPU","AI","neural network","cloud adoption in emerging markets","computer vision","Colab","Ray","deep learning","edge AI","GPUs","types of analytics","different types of analytics"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","deep learning","neural network","computer vision","Ray","Colab","edge AI","TPU"],"url":"https:\/\/analyticsindiamag.com\/it-services\/the-different-types-of-hardware-ai-accelerators\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10115052,"title":"Meet the Real-time NoSQL Database Leading India&#8217;s Explosive UPI Growth","content":"According to the National Payments Corporation of India, UPI recorded a staggering 12.02 billion transactions in December 2023 – its highest to date. The country has become an undisputed global leader riding on the sheer volume of daily and monthly payments made as UPI transactions grew 147% catapulting from INR 92 crore in 2017 to INR 8,375 crore in 2023. One of the major players contributing to this growth is the real-time NoSQL database company Aerospike. “We’ve found that Aerospike can effectively handle the scale associated with such high transaction volumes. It’s exciting not just because we’re addressing a technical challenge we’re designed for, but also because we’re actively participating in the rapid growth observed in India’s payment space,” Srini V Srinivasan, co-founder and chief technology officer, Aerospike, told AIM in a recent conversation. Founded over 14 years ago by Brian Bulkowski and Srini, the company’s primary customers are from adtech, telco, payments, financial services, gaming, e-commerce, and cybersecurity firms. Some prominent names associated with Aerospike are Yahoo, NFL, Airtel, Flipkart, PayPal, Adobe, Nokia, Dream11, are some of them. The company has its headquarters in Bengaluru and Mountain View, US, and Srini believes that its entry into India wasn’t just strategic – it was timely. Aerospike entered the Indian market during the early stages of internet companies, Snapdeal and InMobi being among its first customers, and later with companies like Airtel and Flipkart. “India has proven to be an incredible market for us, driven by the dynamic growth of the internet, the telecom sector, and the government’s initiatives in financial services. The emphasis on scalable systems aligns perfectly with Aerospike’s strengths. The robust competition among companies and the government’s focus on initiatives like universal financial inclusion and UPI have created a conducive environment for our solutions,” added Srini. The company initially focused on providing solutions for adtech, banking, financial services, and payments, where identity resolution plays a crucial role. Recognising the increasing complexity in identity resolution, it expanded its offerings to include the document model (JSON) and an SQL interface via Starburst. “Our focus has been addressing a crucial challenge faced by large-scale, consumer-oriented internet companies, which revolves around maintaining real-time systems while delivering a consistently high-quality user experience,” said Srini. What Makes Aerospike a Customer Favourite There are several factors that make Aerospike the preferred choice for these customers. “Firstly, we can leverage SSDs (Solid State Drive) for real-time data, reducing server footprint and offering a more sustainable and cost-effective solution, with transaction costs 5 to 10 times lower than other in-memory systems,” said Srini. Secondly, its distributed database system capabilities are robust in the face of failures. This means that even if a node fails or a network issue occurs, the system continues to operate, providing a consistent and positive user experience. Lastly, it also provides a true database that is strongly consistent, maximising performance while ensuring this level of consistency support for applications. Moving ahead, its customers are more interested in improving efficiency in an environment-friendly manner. “When we started the company, we prioritised building highly efficient systems at scale, with a significantly reduced footprint and cost. This efficiency extends to easy power management. Lately, our focus has shifted to sustainability, particularly in terms of minimising power usage per user and per transaction while reducing our carbon footprint,” added Srini. Another important point which Srini noted is that clients are actively tapping into the generative AI space and Aerospike is helping it with its support for graph and the upcoming vector database work. Fuelling GenAI Applications Being a high-performance database system, the company doesn’t directly build AI\/ML applications. Instead, its customers employ Aerospike for various classical AI use cases, such as fraud detection, risk analysis, cyber threat analysis, and recommendation engines. Over the years, it has enhanced its platform with features supporting predictive AI, including robust indexing, secondary indexing capabilities with scans, and parallel access to the database. These features facilitate efficient training of data models and quick data processing, making Aerospike a reliable database for storing real-time activities at the edge. However, with customer-driven demand, the company ventured into graph databases with Aerospike Graph, based on Gremlin query language, putting established players like Neo4j and TigerGraph under immense pressure. The trend among customers developing custom identity solutions using Apache TinkerPop with Aerospike as the storage layer prompted Aerospike to officially implement a graph solution. After evaluating the potential and witnessing customers scale to petabytes while maintaining low access times, the company introduced Aerospike Graph, inspired by a large customer’s success in the payment systems business. “Moving forward, we are actively participating in this wave of generative AI innovation by adding support for vector and graph features. The vector database serves as a relationship management system, enabling enterprises to keep their data while utilising Aerospike to generate vectors,” said Srini. The company plans to integrate tools like LangChain and other cloud-based AI\/ML modelling environments to allow users to seamlessly build LLM-based applications. Read more: Why Graph Databases Remain Untapped","excerpt":"Recognising the increasing complexity of identity resolution, Aerospike expanded its offerings to include the document model (JSON) and an SQL interface and eventually entered the graph database space.","categories":["AI Features"],"tags":["BHIM UPI","nosql"],"author_name":"Shritama Saha","publish_date":"2024-03-05T18:12:20","publication_year":"2024","word_count":832,"keywords":["GenAI","AI","ML","RAG","BHIM UPI","LangChain","Aim","generative AI","nosql","SQL","R","fraud detection"],"extracted_tech_keywords":["AI","ML","generative AI","GenAI","LangChain","Aim","RAG","fraud detection","R","SQL"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/meet-the-real-time-nosql-database-leading-indias-explosive-upi-growth\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":22912,"title":"Should There Be A Medical Specialisation In Machine Learning In A Few Years","content":"It is certain that machine learning is being integrated into the medical profession in more ways than we can think. From accurate diagnosis to finding better treatments and suggesting cost effective ways to cure the illness, the emerging tech have become a go-to solution for medical needs across the globe. Now, as we speak of machine learning and artificial intelligence being introduced in medical practices, the idea is not to replace the medical professionals but enhance the doctor’s medical expertise and scale it to unprecedented levels. As the medical field is getting flooded with massive amounts of data from hospital records, patient history and others, the relevance of machine learning is only obvious. As these technologies  take a leap from assisting doctors in mundane tasks to suggesting solutions and remedies, Machine learning and AI are definitely painting a larger picture when it comes to medical science. AI advancements in medical care so far Many hospitals and institutions are adopting AI and machine learning. For instance, Harvard is working on how man and machine can better combine their skills to exploit unprecedented amounts of medical information, University of Texas is using its cancer patients’ data to direct doctors to treatments that gives patients best chance of surviving longer, while Boston company, Neurala, is busy replicating the neural network of human brain in all its complexity and sophistication. Stanford university also created artificial intelligence that could diagnose skin cancer earlier this year. They demonstrated a deep learning algorithm that could diagnose potentially cancerous skin lesions by creating a database of nearly 130,000 skin disease images and training the algorithm to visually diagnose potential cancer. Not just these, but many startups are exploring the area of machine learning to develop solutions and apps to detect medical conditions such as depression, bipolar disorder etc. Cogito, for example is a health app built on machine learning technologies that takes a look at depression. Back in India, Touchkin works on a similar model. The need for medical specialisation in machine learning- should medical students be taught ML In the given scenario, should medical students have a hands-on in machine learning too? While in the given scenario, physicians are trained via traditional textbook and later work as interns to hone their skills, it could be quite appropriate if there were courses that made them equipped with newer technologies around data. As newer data is constantly being produced, wouldn’t it be great if physicians could evolve their methods and processes by studying these data? It would be instrumental in bringing new ideas and practices in the medical world, helping them deal with critical situations as well. Exploring predictive modelling and artificial intelligence could emerge as a model to enhance the delivery care. To get a detailed insight, traditionally medicine field has relied heavily on patient situation and the experiences gained by physicians to bring developments in their treatment, but with the evolution of machine learning technologies, it has brought a scope of improvement by offering to bring a better context to the problem. Not only can this popular branch of artificial intelligence analyse mammoth amounts of data quickly and accurately, it can do things at a speed that human mind can take months together, to execute it successfully. Medical specialisation in machine learning can essentially equip an individual from tackling large quantities of data from hospital, medical records and research to generate actionable results, and at a much better speed. As medical field continually relies on evolving studies and research, it would help physicians and doctors keep themselves updated. Based on AI, they can run it on patient data and bring better methods in improving diagnosis, suggesting drugs, amongst others. All this could be possible it they are well versed with the technical know-hows, which can be studied as a part of their curriculum. How does the future look like? While every innovation is followed by a period of concern, progress cannot be made unless apprehensions are kept away, as is the case of physicians learning ML technologies. Though the adoption of machine learning and AI in medical care is in early stages, collecting and analysing large amounts of data can only mean good things for doctors. With time, it isn’t difficult to see how machine learning will become an integral part of diagnosis and medical research. So much is the popularity that experts have started to believe that without these technologies, any further progress in the medical field may be hindered. While hospitals like Manipal Hospitals have adopted IBM Watson, there are other developments in China, where researchers have created an AI doctor. It is not man versus machines but the results that would be generated by working in conjunction. We are still in the early stages, but the greater involvement of these technologies in the future, would call for medical professionals with skills in data handling thus calling for medical specialisation in data handling.","excerpt":"It is certain that machine learning is being integrated into the medical profession in more ways than we can think. From accurate diagnosis to finding better treatments and suggesting cost effective ways to cure the illness, the emerging tech have become a go-to solution for medical needs across the globe. Now, as we speak of […]","categories":["IT Services"],"tags":["AI Healthcare","medical ai"],"author_name":"Srishti Deoras","publish_date":"2018-03-22T12:32:59","publication_year":"2018","word_count":818,"keywords":["Go","machine learning","artificial intelligence","AI","AI Healthcare","neural network","ML","innovation","Git","deep learning","R","medical ai"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","neural network","R","Go","Git","innovation"],"url":"https:\/\/analyticsindiamag.com\/it-services\/should-there-be-a-medical-specialisation-in-machine-learning-in-a-few-years\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10078783,"title":"Elon Musk on Firing Spree, Lays off Close to 50% of Indian Employees","content":"Twitter lays off have hit India, with many employees receiving email notifications informing them of their future in the company. The layoffs include across departments, including engineering, sales and marketing, and communications. It’s not clear as to how many employees have been laid off. This comes as a part of global restructuring initiated by its new chief Elon Musk. However, multiple reports suggest that the company is looking to cut 50 per cent of its workforce – i.e. close to 7,500 employees. Last week, Musk let go of C-suite executives, including Parag Agrawal (ex-CEO), Ned Segal (ex-CFO), and Vijaya Gadde (former head of legal policy). The recent company filing with the Securities and Exchange Commission confirms that, alongside showcasing Musk as the board’s sole member. However, he said that the new setup was ‘temporary.’ Executive editor of a media firm TV9 Network Aditya Raj Kaul said that a major reshuffle is happening in Twitter India. He said several employees were fired from communications, marketing, and policy teams. “Major bloodbath in Twitter India after Elon Musk takes over the company. Some unconfirmed reports say more than 50 per cent of Indian employees were laid off,” he added. A principal product manager at ServiceNow, Sugandh Rakha, said: “The person I spoke to said that he doesn’t know a single person in Twitter India who wasn’t let go.” However, these remain unverified; recent reports claim that most Twitter India employees have been laid off across departments. According to one report, an estimated 180 of its 230 employees were fired by Twitter in India. Sadly, the laid-off employees were not given prior notice, and most woke up to revoking their slack and email access. The company has said that all employees will receive an email notifying them of their status in the company. While some are still awaiting a reply, those laid-offs have yet to be notified of severance pay they might receive. The sacked employees have taken to Twitter to express solidarity with each other using hashtags #OneTeam and #LoveWhereYouWorked. https:\/\/twitter.com\/yashagarwalm\/status\/1588405497988018179 Read: The Bird is Freed Twitter Layoff Explained The talks of rampant downsizing had begun ever since Elon Musk, CEO of Tesla Motors, took control of Twitter. Musk alluded to potentially laying off 50 per cent of the employees. A few days ago, Musk met with a group of advisers to discuss the matter of layoffs on Twitter. Following the layoffs, a lawsuit has been filed against Twitter in US District Court in San Francisco for violating a law that requires 60 days’ notice of mass layoffs. In his latest tweet posted a few hours ago, Musk also said that Twitter has had a massive drop in revenue, with advertisers pulling off the platform after announcing that a new content moderation policy will be in place. Musk has cited the drop in revenue as a cause for the reduction in workforce, adding that contrary to certain media reports, the parting employees were offered a severance of three months, 50% more than what is lawfully required. The structural changes in Twitter have been costly to every party involved. It is yet to be known what the future of Twitter holds. Regarding Twitter’s reduction in force, unfortunately there is no choice when the company is losing over $4M\/day.Everyone exited was offered 3 months of severance, which is 50% more than legally required.— Elon Musk (@elonmusk) November 4, 2022 Ever since Musk became the chief of Twitter, the company has been in the news for both right and wrong reasons. A few days ago, Twitter announced a monthly charge of $8 for a ‘verified’ tick on the platform. This move by Twitter was made to reduce the reliance on advertisements and move towards a sustainable monetisation model.","excerpt":"Recent reports claim that most Twitter India employees have been laid off across departments.","categories":["AI News"],"tags":["Elon Musk","Mass layoffs","social media","twitter india","Twitter Layoffs"],"author_name":"Ayush Jain","publish_date":"2022-11-04T22:55:53","publication_year":"2022","word_count":622,"keywords":["Go","twitter india","AWS","AI","cloud_platforms:AWS","programming_languages:R","social media","Mass layoffs","programming_languages:Go","RAG","Elon Musk","Aim","GAN","Twitter Layoffs","R"],"extracted_tech_keywords":["AI","Aim","RAG","AWS","R","Go","GAN","cloud_platforms:AWS","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/elon-musk-on-firing-spree-lays-off-close-to-50-of-indian-employees\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10002765,"title":"RIL’s New Announcements, UK Ditches Huawei And More In This Week’s Top AI News","content":"Reliance India Limited is one of the very few companies in the world that has been riding high amidst the pandemic. Be it the topping Warren Buffet or for attracting major investments, RIL’s chairman Mukesh Ambani has been in the news for all the right reasons. Apart from a couple of major announcements from Jio, there has been a lot going on in the 5G space. Here is all the top news from this week. Nokia’s New 5G Solution On Tuesday, Nokia announced an upgrade that will enable its 4G\/LTE radios to be migrated seamlessly to 5G\/NR. These features are expected to provide immediate support for approximately one million radios, reaching 3.1 million by the end of the year and over 5 million in 2021. By upgrading existing radio elements via software, Nokia offers a novel solution for hassle free migration to 5G. This news comes at a crucial juncture as the world is poised to embrace 5G. The Finnish company estimates that their solution will save the telecommunications industry potentially tens of billions of euros in site engineering and re-visit costs as communication service providers are able to upgrade their networks to 5G. Top Europe Court Bans US-EU Data Sharing #ECJ: the Decision on the adequacy of the protection provided by the EU-US Data Protection Shield is invalidated, but @EU_Commission Decision on standard contractual clauses for the transfer of personal data to processors established in third countries is valid #Facebook #Schrems pic.twitter.com\/BgxGAvuq3T— EU Court of Justice (@EUCourtPress) July 16, 2020 On Thursday, Europe’s top court declared that he Decision on the adequacy of the protection provided by the EU-US Data Protection Shield is invalidated. “The limitations on the protection of personal data arising from the domestic law of the United States on the access and use by US public authorities are not circumscribed in a way that satisfies requirements that are essentially equivalent to those required under EU law,” the court said in a statement. The EU-U.S Privacy Shield Frameworks were designed by the U.S. Department of Commerce, and the European Commission to provide companies on both sides of the Atlantic with a mechanism to comply with data protection requirements. The court has also stated that the EU commission’s decision on standard contractual clauses for the transfer of personal data to processors established in third countries is still valid. Ambani Puts India On Global 5G Map Everyone should have access to the internet. Proud to partner with @reliancejio to increase access for the hundreds of millions in India who don’t own a smartphone with our 1st investment of $4.5B from the #GoogleForIndia Digitization Fund.https:\/\/t.co\/1fP8iBZQfm— Sundar Pichai (@sundarpichai) July 15, 2020 The 43rd RIL AGM was streamed live for a digital audience who were introduced to brand new tech updates. The two-hour long event was widely covered by both regional and national news media. “The Jio-Google partnership can make India 2G-mukt.”Mukesh Ambani, Chairman, RIL RIL announced that the tech giant will be investing Rs 33,737 crore – or take a 7.7% stake – in Jio Platforms. According to the terms of the partnership, RIL and Google will be developing an Android-based operating system that could power these entry-level phones. RIL’s Chairman, Mukesh Ambani has even hinted that a homegrown ‘world-class 5G service’ can be ready for field deployment as soon as next year. In a big boost to India’s AR\/VR industry, RIL presented Jio Glass, a new augmented reality solution. Calling it a ‘meaningful immersive experience’, RIL President Kiran Thomas said that it is ideated to enable smoother virtual classes where both teachers and students can meet in 3D virtual rooms and conduct holographic lessons. While little is known about its specifications, it is slated to weigh about 75 grams and will come with built-in apps to allow AR meetings to happen in real-time. These announcements come amidst the news of major investments from Intel, Qualcomm and previously Facebook. And, in this year’s annual meeting, RIL had made their 5G ambitions official. With AI, AR\/VR innovations in the loop along with collaborations with major players, Jio’s roadmap of the future looks quite exciting. Huawei Hits Wall In The UK On Wednesday, following new advice produced by the National Cyber Security Centre (NCSC) on the impact of US sanctions against the telecommunications vendor, the UK government  announced that Huawei will be completely removed from their 5G networks by the end of 2027. Starting this December, there will be a total ban on the purchase of any new 5G kit. The decision was taken in a meeting of the National Security Council (NSC) chaired by the Prime Minister, in response to new US sanctions, that were imposed on Huawei in May. According to the official press release, there will be a ban on the purchase of new Huawei kit for 5G from next year and it will be completely removed from 5G networks by the end of 2027. Google’s New Confidential VMs Google Cloud announced their new product, Confidential VMs, now in beta, is the first product in Google Cloud’s Confidential Computing portfolio. Confidential VMs offer  memory encryption to further isolate workloads in the cloud. Confidential Computing builds on the protections Shielded VMs offer against rootkit and bootkits, helping to ensure the integrity of the operating system. Using the AMD SEV feature, Confidential VMs offer high performance for the most demanding computational tasks, while keeping VM memory encrypted. Google Cloud stated that Confidential Computing can help organisations transform the way they process data in the cloud while preserving confidentiality and privacy. Hackers Time Out On Twitter Based on what we know right now, we believe approximately 130 accounts were targeted by the attackers in some way as part of the incident. For a small subset of these accounts, the attackers were able to gain control of the accounts and then send Tweets from those accounts.— Support (@Support) July 17, 2020 In one of the biggest security breaches ever on Twitter, numerous high-profile accounts including those of Barack Obama and Elon Musk were taken over by hackers to promote a bitcoin scam, which earned them nearly $120,000. In a series of tweets Twitter said that its internal systems were compromised by the hackers, confirming theories that the attack could not have been conducted without access to the company’s own tools. According to Motherboard, Twitter’s internal company admin tool was allegedly used to conduct the account takeovers, potentially by resetting account email accounts and then recovering passwords.","excerpt":"Reliance India Limited is one of the very few companies in the world that has been riding high amidst the pandemic. Be it the topping Warren Buffet or for attracting major investments, RIL’s chairman Mukesh Ambani has been in the news for all the right reasons. Apart from a couple of major announcements from Jio, […]","categories":["AI News"],"tags":["Huawei","Jio","limitations of AI","RIL"],"author_name":"Ram Sagar","publish_date":"2020-07-18T18:00:45","publication_year":"2020","word_count":1073,"keywords":["Jio","Go","programming_languages:R","AI","innovation","ML","programming_languages:Go","Git","RIL","cloud_platforms:Google Cloud","GAN","Huawei","R","limitations of AI"],"extracted_tech_keywords":["AI","ML","R","Go","Git","GAN","innovation","cloud_platforms:Google Cloud","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/reliance-jio-agm-huawei-5g-ban\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10160778,"title":"Namma Yatri Now in Odisha","content":"Odisha Chief Minister Mohan Charan Majhi launched Odisha Yatri, a state-owned ride-booking app, today in Bhubaneswar. The launch was part of the National Road Safety Month 2025 celebrations. Developed by Capital Region Urban Transport (CRUT), in collaboration with Moving Tech Innovations Private Limited, Odisha Yatri is the latest addition to Moving Tech’s growing portfolio of mobility platforms. The Bengaluru-based company, known for its successful Namma Yatri app in Bengaluru, has been expanding its zero-commission model across India. Chief Minister Majhi described the Odisha Yatri app as the state’s first homegrown ride-booking platform, highlighting its ability to connect drivers directly with passengers while ensuring fair fares, enhanced safety, and better livelihoods for drivers. He encouraged citizens to adopt the app extensively for their travel needs in Bhubaneswar and Cuttack. Moving Tech Goes Places Odisha Yatri follows the success of Moving Tech’s other platforms, including Yatri in Kochi (their first venture) and Delhi, Yatri Sathi in Kolkata, and Mana Yatri in Hyderabad. Like its predecessors, Odisha Yatri operates on a zero-commission model, allowing drivers to retain 100% of their earnings. Google has invested in Moving Tech. The app features transparent pricing, fast bookings, and real-time tracking. It also incorporates advanced safety protocols in collaboration with the Police Commissionerate. Future expansions will include public transport booking and tourist site ticketing, coinciding with the upcoming Pravasi Bharatiya Divas event in Bhubaneswar. The application is integrated with the Open Network for Digital Commerce (ONDC), and it works on the Beckn Protocol.","excerpt":"Bengaluru-based Moving Tech’s other platforms include Yatri in Kochi and Delhi, Yatri Sathi in Kolkata, and Mana Yatri in Hyderabad.","categories":["AI News"],"tags":["Beckn Protocol","Bengaluru","Moving Tech","namma yatri","ondc"],"author_name":"Vandana Nair","publish_date":"2025-01-06T14:07:25","publication_year":"2025","word_count":246,"keywords":["Go","API","programming_languages:R","ondc","AI","innovation","namma yatri","Git","Beckn Protocol","RAG","programming_languages:Go","Moving Tech","Bengaluru","R"],"extracted_tech_keywords":["AI","RAG","R","Go","Git","API","innovation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/namma-yatri-now-in-odisha\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10093396,"title":"Cognizant Goes Beyond PoC, Launches First-of-its-Kind AI Platform for Generative AI Deployment","content":"Global IT firm, Cognizant, recently announced a new, enterprise-wide platform called Cognizant Neuro AI. This new platform provides enterprises with a comprehensive approach to accelerate the adoption of generative AI technology, alongside harnessing business values in a flexible, secure, scalable and responsible manner. The platform has been built utilising its vast consulting, advisory and ecosystem partnership, alongside industry expertise. The company claims that it is among the first to launch this AI platform to help its customers understand, integrate and customise AI models to drive business outcomes. EVP of software and platform engineering, Prasad Sankaran, said that this new AI platform goes beyond PoCs, and aims to accelerate the adoption of enterprise-scale AI applications, increase RoI, minimise risks, alongside develop business solutions faster. Cognizant Neuro AI With its latest AI platform for generative AI, the company claims that business leaders will gain access to the full range of tools and capabilities. EVP of core technologies and insights, Anna Elango, said that the company believes enterprises need scalable, flexible, end-to-end tools to accelerate responsible AI adoption. Its new AI platform includes a library of reusable generative AI models and agents, development tools and control components, including APIs and access management, versioning and auditing. In terms of responsibly deploying the generative AI solutions, Cognizant said that its AI platform has been built on its in-house responsible AI principles, which mostly emphasise inclusivity, accountability, reliability and community benefit. Cognizant vs the world Cognizant is not alone. Most Indian IT companies today claim to be working on developing generative AI solutions for their customers, and a majority are still in the PoC stages. For instance, TCS is diligently working on developing its own alternative to GitHub Copilot, to revamp enterprise code generation. Simultaneously, Accenture unveiled a new study titled ‘A New Era of Generative AI for Everyone.’ exploring generative AI and LLMs. Tech Mahindra, on the other hand, is taking a distinctive approach with its Generative AI Studio, empowering businesses with a user-friendly interface that facilitates a myriad of customisation options for their content. Now, Capgemini claims that it has a unique moat when it comes to deploying generative AI solutions. Read: Gen AI is All Over the IT World, Except on Ground How is Cognizant’s AI platform different? The company said that its Neuro AI platform will help companies on their journey, from identifying specific use cases, alongside operationalising AI, to data engineering and continuous improvement. The platform also enables AI and engineering teams to build flexible, reusable, safe and secure solutions, alongside designing conversational and generative user experiences and more. Cognizant’s AI Prowess The company claimed that its new AI platform has been built on years of AI research and development, and AI applications developed for its clients. This includes deep learning, and machine learning, alongside generative AI techniques in life sciences, retail and healthcare. Its team has worked on areas such as price optimisation, predictive healthcare analytics, and crop science optimisation among others. Job losses Cognizant recently reported a 0.3% YoY of decline in its revenue, which stood at $4.8 Bn in the quarter under review. The company also saw a decline of 3,800 employees in FY22. However, its attrition rate declined to 23% from 30% in FY22 Q1. In a bid to streamline its operational modal and corporate functions, the company looks to launch the NextGen programme in the next quarter of FY23, which is expected to impact around 3,500 employees.","excerpt":"The company looks to launch the NextGen programme in the next quarter of FY23, which is expected to impact around 3,500 employees","categories":["AI News"],"tags":["Cognizant"],"author_name":"Amit Naik","publish_date":"2023-05-16T11:40:01","publication_year":"2023","word_count":571,"keywords":["Go","machine learning","AI","ML","Cognizant","Scala","Aim","deep learning","analytics","generative AI","R"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","analytics","generative AI","Aim","R","Go","Scala"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/cognizant-launches-ai-platform-for-generative-ai-deployment\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10057492,"title":"Interesting innovations from OpenAI in 2021","content":"Not even a decade old, OpenAI has made a name for itself as a leading AI research lab worldwide. It gave the world GPT-3 in 2020 – a path-breaking innovation that uses deep learning to give us human-like text. GPT-3 has been a stepping stone for other tech giants to take inspiration and bring out their own innovations in the large language model space. This year, too, saw OpenAI testing its limits and continuing its streak in bringing out algorithms and models that can create a massive impact. Let us look at a few of them as the year comes to an end. Codex OpenAI released Codex through an API in private beta. It translates natural language to code and is the backbone behind GitHub Copilot. It can interpret simple commands in natural language and execute them on the user’s behalf—making it possible to build a natural language interface to existing applications. OpenAI said, “OpenAI Codex has much of the natural language understanding of GPT-3, but it produces working code.” One can issue commands in English to any piece of software with an API. OpenAI Codex is a general-purpose programming model (can be applied to any programming task). For more details, click here. DALL·E At the beginning of the year, OpenAI came out with DALL·E, a 12-billion parameter version of GPT-3, trained to generate images from text descriptions by using a dataset of text-image pairs. DALL·E is a Transformer language model that receives the text and the image as a single stream of data, containing up to 1280 tokens. It is trained using maximum likelihood to generate all of the tokens, one after another. DALL·E can render an image from scratch and also alter aspects of an image using text prompts. OpenAI said that DALL·E can create plausible images for a range of sentences that explore the compositional structure of language. For more details, click here. GLIDE GLIDE (Guided Language to Image Diffusion for Generation and Editing) is a 3.5 billion parameter text-to-image generation model that is even better than DALL-E. The paper released by OpenAI said the researchers found that samples from the model they generated with classifier-free guidance are both photorealistic and reflect a diverse range of world knowledge. In terms of performance, OpenAI said that samples they generated were preferred to those from DALL-E 87% of the time when evaluated for photorealism and 69% of the time when evaluated for caption similarity by human judges. For more details, click here. Triton 1.0 OpenAI released the open-source Python-like programming language Triton 1.0 that helps researchers with no CUDA (Compute Unified Device Architecture) experience write highly efficient GPU code. OpenAI claimed, “Triton makes it possible to reach peak hardware performance with relatively little effort.” It said that Triton has already been used to produce kernels that are up to 2x more efficient than equivalent Torch implementations. Modern GPUs have three crucial components in terms of architecture – DRAM, SRAM and ALUs. OpenAI said that Triton aims to automate these optimisations fully. This will lead to developers focusing more on writing high-level logic of their parallel code. For more details, click here. CLIP While releasing DALL·E, OpenAI also released CLIP (Contrastive Language–Image Pre-training) that builds on a large body of work on zero-shot transfer, natural language supervision, and multimodal learning. OpenAI showed that scaling a simple pre-training task is sufficient to achieve competitive zero-shot performance on a wide range of image classification datasets. This method uses available sources of supervision – the text paired with images found on the internet.For more details, click here.","excerpt":"A look at a few of the interesting algorithms from OpenAI in 2021","categories":["Global Tech"],"tags":["AI (Artificial Intelligence)","CLIP","DALL.E","Deep Learning","GLIDE","Machine Learning","OpenAI","OpenAI Codex"],"author_name":"Sreejani Bhattacharyya","publish_date":"2022-01-01T16:00:00","publication_year":"2022","word_count":594,"keywords":["CUDA","Go","OpenAI","AI","Machine Learning","Git","Python","Aim","deep learning","CLIP","OpenAI Codex","Deep Learning","GLIDE","R","AI (Artificial Intelligence)","DALL.E"],"extracted_tech_keywords":["AI","deep learning","OpenAI","Aim","CUDA","Python","R","Go","CUDA","Git"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/interesting-innovations-from-openai-in-2021\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":66328,"title":"Govt. Launches National AI Portal And Skilling Programme To Strengthen AI Game In The Country","content":"To keep up the AI innovation game in India, the government recently announced the launch of a one-stop digital platform for all artificial intelligence-related developments in India. Launched by Ravi Shankar Prasad, the Union Minister for Electronics and IT, Law and Justice and Communications on the occasion of the first anniversary of the second tenure of the government, the portal is called www.ai.gov.in. Developed by the Ministry of Electronics and IT and IT Industry, National e-Governance Division of Ministry of Electronics and IT and NASSCOM, it aims to boost AI developments in India. It will enable sharing of resources such as articles, startups, investment funds in AI, resources, companies and educational institutions related to AI in India. The official press statement also stated that it will also share documents, case studies, research reports etc. It also has a section about learning and new job roles related to AI. Launched the National Portal for Artificial Intelligence https:\/\/t.co\/vUDvORetpW Jointly developed by @GoI_Meity and @nasscom this portal will be one stop portal for #AI related developments in India. pic.twitter.com\/xiztrdLaeN— Ravi Shankar Prasad (@rsprasad) May 30, 2020 “India must be a leading country in the development of Artificial Intelligence in the world, leveraging upon its vast Internet-savvy population and data it is creating. India’s AI approach should be of inclusion and empowerment of human being by supplementing growth and development rather than making human beings less relevant,” said Shri Ravi Shankar Prasad during the launch. The government strongly believes that the roles of these technologies will especially be more in the pandemic-afflicted world. While AI technology has been deeply integrated into the field of education, agriculture, healthcare, e-commerce, finance, telecommunications, etc., the launch of this portal will lead to the democratization of artificial intelligence in the country. Responsible AI for Youth Programme Apart from the portal, the government also launched a National Program for the youth — Responsible AI for Youth, which aims to give the young students of the country a platform to develop appropriate new age tech mind-set. It will help in developing relevant AI skill sets while giving them access to the required AI tool-sets to make them digitally ready for the future. This program has been created and launched by the National e-Governance Division, Ministry of Electronics & IT in collaboration with Intel India, with support from Department of School Education and Literacy (DoSE&L), Ministry of Human Resource Development. This program aims at making youth AI-ready while helping reduce the skill gap and helping them create meaningful social impact solutions. As per the official reports, the program will reach out to students from Government schools across India, providing them with an opportunity to become part of the skilled workforce in an inclusive manner. Who Can Take Benefits Of The Program The program is open to students of classes 8-12 from Central and State government-run schools (including KVS, NVS, JNV) from across the country. It will be implemented in a phase-wise manner and in its first phase, each of the State Education Department will nominate 10 teachers as per the eligibility criteria to take up the orientation sessions aimed to help them understand the premise and identify 25-50 potential students for the Program. The selected students will attend online training sessions on AI and understand how to identify social impact ideas\/projects that may be created using AI and submit their ideas through a 60 seconds video explaining a proposed AI-enabled solution. From the submitted ideas in the form of videos, top 100 ideas will be shortlisted and these students will be invited to attend residential boot camps or online sessions (subject to COVID-19 situation); to take them through a deep dive AI journey. Post the boot-camps\/ online sessions, these students will be asked to create real-time projects and submit their final project in a video format on the website. Throughout the program, required assistance will be provided by Intel certified AI coaches and mentors throughout to ensure that ideas mature as prototypes. Top 50 project ideas will be shortlisted by experts to be showcased either face to face or in an online format. Further, top 20 innovative projects will be selected by an independent committee of experts and provided opportunities to showcase at a relevant platform. The government, in its last year’s budget speech, had announced the establishment of various Centre of Excellence (CoE) for AI as it envisions AI to propel innovations and help address social issues in the country.","excerpt":"To keep up the AI innovation game in India, the government recently announced the launch of a one-stop digital platform for all artificial intelligence-related developments in India. Launched by Ravi Shankar Prasad, the Union Minister for Electronics and IT, Law and Justice and Communications on the occasion of the first anniversary of the second tenure […]","categories":["AI News"],"tags":["Artificial Intelligence vs Human Intelligence"],"author_name":"Srishti Deoras","publish_date":"2020-06-01T19:00:00","publication_year":"2020","word_count":740,"keywords":["Go","artificial intelligence","AI","innovation","Git","RAG","Artificial Intelligence vs Human Intelligence","responsible AI","Aim","ViT","R"],"extracted_tech_keywords":["AI","artificial intelligence","Aim","RAG","R","Go","Git","ViT","responsible AI","innovation"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/govt-launches-national-ai-portal-and-skilling-programme-to-strengthen-ai-game-in-the-country\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10138322,"title":"Did Demis Hassabis Truly Deserve the Nobel Prize?","content":"“Winning the Nobel Prize is the honour of a lifetime and the realisation of a lifelong dream – it still hasn’t really sunk in yet,” said the co-founder of DeepMind, Demis Hassabis, who, along with his team, cracked the 50-year-old challenge of protein structure prediction with AlphaFold 2. “With AlphaFold 2 we cracked the grand challenge of protein structure prediction – predicting the 3D structure of a protein purely from its amino acid sequence,” said Hassabis. He added that the open-access database of over 200 million protein structures has empowered more than 2 million researchers, advancing critical work in enzyme design, disease understanding, and drug discovery. As AI continues to shape the future of therapies and scientific discovery, with teams like Isomorphic Labs driving AI-led innovations, the question arises—does AlphaFold’s monumental achievement justify its Nobel recognition, and does it set a precedent for AI-powered breakthroughs in global scientific accolades? According to Heiner Linke, Chair of the Nobel Committee for Chemistry, “One of the discoveries being recognised this year concerns the construction of spectacular proteins. The other is about fulfilling a 50-year-old dream of predicting protein structures from their amino acid sequences. Both of these discoveries open up vast possibilities.” In December 2020, when AlphaFold 2’s success in predicting protein structures was announced, there was no doubt that this breakthrough had Nobel Prize potential. At the time, experts recognised the colossal impact of solving a decades-old challenge in protein science. Following this, the 2023 Breakthrough Prize was awarded to Demis Hassabis, John Jumper, and the DeepMind team, finally leading to the Nobel Prize in 2024​. It goes without saying that the Nobel Committee recognised​ that such a discovery not only solved age-old scientific challenges but also made the borders of AI and natural sciences porous. It redefines what’s possible in biology and chemistry through the intelligent use of AI, enabling research and innovation across multiple fields. Deserves Every Nobel There Is Founded by Hassabis in 2021, Isomorphic Labs—a sister company of Google DeepMind—is looking to revolutionise drug discovery with AI, potentially building a multi-$100 billion business by accelerating research and improving clinical trial success. “AI-designed drugs would probably be available in the next couple of years,” he said. “I hope to achieve both (commercial success and societal benefits) with Isomorphic and build a multi-100 billion dollar business. I think it has that potential,” said Hassabis without delving into the specific timeline. In May 2024, Google DeepMind released AlphaFold 3, a game changing protein folding model that predicts with 50% better accuracy. “Well, if you ask me the number one thing AI can do for humanity, it will be to solve hundreds of terrible diseases. I can’t imagine a better use case for AI. So that’s partly the motivation behind Isomorphic and AlphaFold and all the work we do in sciences,” said Hassabis. He believes that “revolutionising the drug discovery process to make it ten times faster” and more efficient and increasing the likelihood of passing clinical trials through better property prediction offers plenty of commercial value. Predicts 200 Mn Protein Structures Last month, Google DeepMind, in collaboration with Isomorphic Labs, predicted over 200 million protein structures using AlphaFold. It achieved this by training the model with nearly 100,000 known proteins, driving significant breakthroughs in drug discovery by targeting previously intractable biomedical challenges. The model can predict the 3D structure of proteins with incredible accuracy, aiming to design drugs that target specific proteins, unlocking treatments for diseases that were previously untreatable. Recently, Google DeepMind also launched AlphaProteo, an AI system that generates novel proteins designed to bind to specific target molecules poised to significantly advance research in drug design, disease understanding and other health applications.DeepMind is not the only active player in the market, ESMFold, Meta’s protein-folding model, has also predicted about 772 million protein structures. This is only the beginning.","excerpt":"Does AlphaFold’s monumental achievement justify Nobel recognition and set a precedent for AI-powered breakthroughs in global scientific accolades?","categories":["AI Features"],"tags":["AlphaFold","demis hassabis","Nobel prize"],"author_name":"Shalini Mondal","publish_date":"2024-10-15T16:55:35","publication_year":"2024","word_count":639,"keywords":["Go","AlphaFold","API","programming_languages:R","AI","demis hassabis","innovation","programming_languages:Go","Nobel prize","Aim","R"],"extracted_tech_keywords":["AI","Aim","R","Go","API","innovation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/why-demis-hassabis-truly-deserved-the-nobel-prize\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10119484,"title":"IBM Expands Software Availability to 92 Countries in AWS Marketplace","content":"IBM today announced that its software portfolio is now available in 92 countries in AWS Marketplace, a digital catalog with thousands of software listings from independent software vendors (ISVs). The expansion beyond Denmark, France, Germany, UK and US where the software is currently available, will help make procurement easier for clients, streamline purchasing and create new efficiencies, while allowing them to use their AWS committed spend for IBM software purchases. According to a Canalys study, cloud marketplaces continue to emerge as the fastest-growing route to market for Software-as-a-Service (SaaS) software, expected to increase to USD 45 billion by 2025, up 84% CAGR over five years. Marketplaces also help shorten the buying cycle, consolidate billing, and make it easier to scale software deployments quickly. Customers now will have access to IBM’s AI and data technologies within a portfolio of 44 listings and 29 SaaS offerings available for purchase. Included among those technologies are components of the watsonx AI and Data platform, which allow enterprises to build, scale and govern AI workloads. Watsonx.data, a fit-for-purpose data store built on an open data lakehouse architecture, and Watsonx.ai, a next generation enterprise studio for AI builders are available in AWS Marketplace as well as two of IBM’s AI Assistants — watsonx Assistant and watsonx Orchestrate. watsonx.governance is expected to be available soon. Other software includes IBM’s flagship database Db2 Cloud Pak for Data as well as a portfolio of automation software including Apptio, Turbonomic and Instana, and the IBM Security and Sustainability software portfolios – all built on Red Hat OpenShift Service on AWS. The cloud-native software enables clients to deploy on AWS while flexible licensing, including SaaS and subscription, makes it easier for clients to purchase exactly how they want. IBM is also launching 15 new IBM Consulting professional services and assets on AWS Marketplace, exclusively designed for AWS. These new service offerings are aligned to client needs and demand, focused on data and application modernization, security services, and tailored industry-specific solutions – with generative AI capabilities included in select services. I BM Consulting also brings 24,000 AWS certifications and a dedicated team of experts trained in the latest AWS technologies to help clients with tailored recommendations grounded in industry best practices. “By expanding the availability of our software portfolio in AWS Marketplace, organizations around the world will have greater access to a streamlined way to procure many IBM AI and hybrid cloud offerings to help propel their business forward,” said Nick Otto, Head of Global Strategic Partnerships, IBM. “Our collaboration with AWS is a prime example of how we’re working with other companies to meet the needs of clients, making it as easy as possible for them to do business with IBM and accelerate their transformation journeys,” he added.","excerpt":"Customers now will have access to IBM’s AI and data technologies within a portfolio of 44 listings and 29 SaaS offerings available for purchase.","categories":["AI News"],"tags":["AWS","IBM"],"author_name":"Pritam Bordoloi","publish_date":"2024-05-02T18:10:16","publication_year":"2024","word_count":457,"keywords":["Go","AI assistants","AWS","AI","ML","Git","generative AI","IBM","GAN","R","data lake"],"extracted_tech_keywords":["AI","ML","generative AI","AI assistants","AWS","R","Go","Git","data lake","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/ibm-expands-software-availability-to-92-countries-in-aws-marketplace\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10079520,"title":"Can Conversational AI Help Government Departments?","content":"Artificial intelligence known as conversational AI enables machines to understand, process, and respond to human language. Although many businesses are using this technology to automate calls in an effort to save time and increase efficiency, can government departments employ it as well? To know more, Analytics India Magazine reached out to Rashi Gupta, Chief Data Scientist and co-founder of Rezo AI. AIM: What does Rezo AI do? And what’s your role at Rezo AI? Rashi: I am a data science expert, and I founded Rezo AI four and a half years ago. I graduated with a PhD in machine learning from the University of Helsinki in Finland in 2009, and I’ve been working with data for nearly 20 years—both in academia and in business. Rezo AI thereby automates whatever is possible to automate, it can be post-sale conversations, service talks, marketing conversations, or sales communications. For tasks that cannot be automated using the power of bots, we essentially assist the human agents so that they can achieve them in a manner that is more performance-based. For instance, today, they are performing at a 50~60% efficiency rate. Right, we will be able to step it up so that these guys can contribute to the business they are working so hard to build up with good 80~90 people. Therefore, you could say that overall, we bring about cost and productivity efficiencies. We mostly work with vehicle manufacturers, but we also collaborate with a number of fintech businesses, telcos and DTH providers. Additionally, the hotel and airline industries are on our radar; basically, Rezo AI is highly applicable anywhere discussions take place. AIM: Can the government use conversational AI? How would that be possible? Rashi: Huge potential exists. You see, the government essentially wants to attract the masses. When the government rolls out new schemes and wants to convey to the end-beneficiary, for example, farmers, the government can really learn what they think about it before we even roll out the policies when they are being made. The government can leverage conversational AI in outbound calls in order to understand what the target population is looking for. It can be used to gauge what sorts of policies are being rolled out and are making an impact. When the government conducts a post-facto setup audit, it’s a kind of analysis that might or might not be able to provide an accurate conclusion because, by the time the rollout is conducted, the relevance has long vanished. Therefore, while these rollouts are taking place, you can gather data so that the policies can be changed in response to public opinion. And with conversational AI, it’s simple to speak with the general public, get their opinions, implement that kind of process improvement, and then resume the rollout. But conversational AI standalone will not work; it has to be powered big time with  data analytics because a lot of data is getting accumulated in this process. Now that you’ve launched such a project, you need to take care of that data and then build up a hypothesis for the brand or the caller, for example, on behalf of whom you’re making these calls to be able to understand that this is what we have asked or we are putting forward. So yes, only conversational AI won’t help, but conversational AI powered by big data analytics brings a lot to the table. AIM: What market are you presently concentrating on? Rashi: We are really focusing on the Indian market, but by next financial year, we’d have already started doing our reach outs to the US and the Middle East market. The idea of what we are bringing to the table powered by data analytics is very unique. So, the world broadly is open to us, because we are doing chat, voice and speech analytics. In that way, we have entries everywhere. Currently, as far as the league of conversational AI companies is concerned, we are currently at the top. AIM: Can conversational AI replace call centre jobs? Rashi: We are merely increasing the call centres’ productivity, which is where automation is going anyhow. Conversational AI businesses do not seek to eliminate jobs. Companies that use conversational AI simply want to add that extra efficiency. When speaking with customers, it can be difficult to gauge the general population, but brands that are spending a lot of money want to know how their employees are performing.","excerpt":"The government can leverage conversational AI in outbound calls in order to understand what the target population is looking for.","categories":["AI Trends"],"tags":["AI for good","AI in government","Conversational AI"],"author_name":"Lokesh Choudhary","publish_date":"2022-11-12T16:00:00","publication_year":"2022","word_count":735,"keywords":["big data","data science","Go","artificial intelligence","machine learning","AI","AI in government","AI for good","RAG","Aim","Conversational AI","analytics","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","data science","analytics","Aim","RAG","R","Go","big data"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/can-conversational-ai-help-government-departments\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":23888,"title":"9 AI And ML Courses Offered By Tech Giants Which Will Boost Your Career","content":"By now almost every tech company has realised that the world needs more artificial intelligence and machine learning experts. There are only 10,000 people in the world right now with the education, experience and talent needed to develop these AI technologies. This acute lack of skill set is hindering digital transformation at enterprises across the globe. To meet this talent shortage, the tech giants have now become more committed to making ML more accessible to students and developers by offering online courses. “AI can solve complex problems and has the potential to transform entire industries, which means it’s crucial that AI reflect a diverse range of human perspectives and needs,” Zuri Kemp, technical program manager at Google, said. Let’s take a look at a few important ML courses offered by some of the noted tech companies: Google Learn with Google AI Google has been pursuing AI education for a while and now, as they want to educate more people in AI and ML with a free online course. Early this year, Google launched a new website called Learn with Google AI. This website is meant to be an information hub for anyone who wants to learn about core ML concepts, develop and hone their ML skills, and apply ML to real-world problems. The new website aims to cater everyone starting from students to curious cats, and advanced researchers. The California-based tech giant has repeatedly stated its goal to democratise AI and make its tools available for everyone. ML Crash Course The Learn with Google AI website also features a free course called ML crash course (MLCC) with TensorFlow APIs. Google originally designed this course for its employees as a part of a two-day boot camp aimed to give a practical introduction to ML fundamentals. More than 18,000 employees have already enrolled in MLCC, enhanced camera calibration for Daydream devices, built VR for Google Earth, and improved streaming quality at YouTube. Now, Google is making MLCC available to everyone.The 15 hours online course includes real-world case studies, interactive visualisation, video lectures, over 40 exercise to help teach ML concepts. Microsoft Microsoft Professional Program For AI Microsoft is focused on empowering both people and organisations, by democratising access to AI to help solve most of their pressing challenges. Earlier this month, the tech company launched a new course on its tech accreditation scheme, called, Microsoft Professional Program, which is dedicated to AI. This course is aimed at developers who are looking to expand their AI capabilities as well as anyone interested in general AI education. The program consists of 10 courses that cover maths, statistics, Python, data analysis, computer vision, ethics, Azure ML, speech recognition and NLP. Each course takes between eight and 16 hours to complete. However, these courses are not free. The students need to buy certificates from edX.org  for each course. Once you complete the course you will receive a digitally-sharable Microsoft Professional Program Certificate in AI. Amazon Last year, Amazon launched two new courses to educate developers and leverage AI solutions using Amazon Web Services: Introduction to ML This is a free 40-minute web-based training intended for developers, solution architects and IT decision makers who know about the foundations of working with AWS. It gives an overview of ML, walks through a use case, teaches relevant terminology, and walks the user through the process for incorporating ML solutions into a business or a product. Deep Learning on AWS This is a one-day instructor-led training for developers who are interested in learning more about AWS solutions for deep learning. It teaches the deep learning model and gives the users a roadmap for understanding the challenges deep learning can solve. It also covers solutions related to image recognition, speech recognition, and speech translation. The training includes how to run your models on the cloud using an Amazon Elastic Compute Cloud (Amazon EC2)-based Deep Learning Amazon Machine Image (AMI) and the MXNet framework. NVIDIA Deep Learning Course The Deep Learning Institute by Nvidia comprises of both the hands-on training session and online course for AI enthusiasts. The topics of the course revolve around becoming a self-driving car engineer, creating smarter robots using deep learning with tools like Microsoft Azure, predicting the risk of disease, preventing it and more. It also includes instructor-led seminars, workshops, classes that reach developers across Asia, Europe and US. The hands-on training is taught by certified experts from Nvidia partnering companies and universities, wherein they cover fundamentals of deep learning with topics like AI for object detection and image classification with TensorFlow, neural network deployment with DIGITS, and inference optimisation for an autonomous vehicle with TensorRT. The online classes are delivered through AWS and Google’s Qwiklabs. Baidu Deeplearning.ai Baidu’s Deeplearning.ai offers training to coders who don’t have a background in AI or know much about deep learning. Here, the only thing required to pursue this course is basic programming knowledge, a proficiency in Python and a general understanding of ML. The course was launched by Andrew Ng, a former chief scientist at Baidu in 2017. It aims to spread the benefit of recent advances in ML far beyond big tech companies. The course cost $49 a month and is offered via Coursera. It features five tracks that include neural networks, backpropagation, convolutional nets and recurrent nets. It also teaches other core aspects of deep learning. The students also get to participate in applied deep learning projects to address real-world problems in language understand, healthcare and music generation. Intel Intel Student Ambassador Program for AI The Intel Student Ambassador Program for AI is the developer affinity program for university students to engage with Intel around their work in ML, deep learning and AI. The tech company is working with universities across the world to introduce this program. “The students who are invited as Student Ambassadors are provided with technical support, resources and marketing to advance their work through Intel software, tools and hardware,” said the company. The program is targeted toward graduate, undergraduate and PhD students. Uber Uber AI Residency This is a one-year research training program offered by Uber AI Labs and Uber ATG Toronto, to allow researchers accelerate their careers in ML and AI research and practice. The students will have the opportunity to work directly with researchers and engineers from across the company. The students also get to publish their work externally at top ML venues. Conclusion: To sum up, these courses offered by the tech giants are not only trying to plug the talent gap but are also bringing more developers and students onboard to develop AI. Such initiatives will also break down barriers for AI teams to share their best practises and research to maximise social benefits and tackle ethical concerns and make it easier for students from other fields to get more access to ML. Eric Boyd, corporate vice president at Microsoft AI and Research rightly recapitulated, “ML has the ability to transform the way we work, interact and communicate. To make this happen we need to put the right tools in the right hands.”","excerpt":"By now almost every tech company has realised that the world needs more artificial intelligence and machine learning experts. There are only 10,000 people in the world right now with the education, experience and talent needed to develop these AI technologies. This acute lack of skill set is hindering digital transformation at enterprises across the […]","categories":["AI Trends"],"tags":["AI (Artificial Intelligence)","ai certificates","Amazon","Baidu","Google","Intel","Machine Learning","Microsoft","NVIDIA","Uber"],"author_name":"Smita Sinha","publish_date":"2018-04-20T07:00:39","publication_year":"2018","word_count":1173,"keywords":["ai certificates","computer vision","deep learning","AI (Artificial Intelligence)","Azure ML","artificial intelligence","NLP","Baidu","NVIDIA","Intel","machine learning","AI","neural network","ML","Amazon","Machine Learning","Aim","Google","Uber","Microsoft"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","neural network","NLP","computer vision","Aim","Azure ML"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/9-ai-and-ml-courses-offered-by-tech-giants-which-will-boost-your-career\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":36633,"title":"From Smart Nation To AI Leader, Singapore Is On A Fast Track To Success","content":"Since early 2014, Singapore has been driving an initiative to be a leader in emerging tech and establish itself as a Smart Nation. To make its ambitious vision a reality, the government has done everything, from infusing investments to framing policies to support it. Early in 2018, Singapore launched a Digital Government Blueprint to support its Smart Nation vision and serves as one of the key pillars of the Government’s public service transformation efforts. It has also invested heavily in physical infrastructure and connectivity. But is Singapore ready for the changes that come with it — the overwhelming inflow of data, availability of skilled candidates, need for reskilling and more? As the country is inching towards a digital and smart nation, technologies such as AI, deep learning and machine learning can prove to be revolutionary. AI and data science can play a crucial role in restructuring the economy and the country is aggressively working towards it. 6 Key Efforts By Singapore To Strengthen The AI Play Singapore has taken some serious steps to boost up its AI play. The government, educational institutions, and companies are taking suitable steps so that the upcoming workforce is prepared for the disruption that artificial intelligence is about to bring. Some of the key initiatives taken by the government to build a strong foundation in AI are: 1| SkillsFuture Initiative: This initiative is aimed at helping the companies reskill and redeploy its at-risk workers into new roles. This national movement allows Singaporeans to develop skills to their fullest potential throughout life, regardless of their starting points. It can serve individuals in all walks of their life such as schooling, early career, mid-career or later years to equip themselves with required skills. As a part of this program, individuals can equip themselves in AI or other intelligent technologies. It aims at the strategic developments of employees with the required future-tech skills. 2| AI Program By National Research Foundation: One of the older initiatives of the country, NRF launched AI Program, which aims to bring relevant research institutions, AI startups and GovTech’s plans to set up a CoE in AI and data science to help other agencies deploy AI technologies. It catalyses, synergies and boosts Singapore’s AI capabilities to power digital economy. This is a collaboration between NRF, Smart Nation and Digital Government Office, Economic Development Board, SGInnovate, Integrated Health Information Systems and Infocomm Media Development Authority. The main objective of this program is to use AI to address major challenges that affect society, invest in deep capabilities, broaden adoption and use of AI and machine learning within industries. 3| AI Governance Framework: In January this year, the minister of Communications and Information released a Model AI Governance Framework that outlined key ethical principles and practices in AI deployment. It will help in framing the discussions around harnessing AI in a responsible way. Through the Model Framework, they aim to promote AI adoption while building consumer confidence and trust in providing their personal data for AI. Singapore has become the first Asian nation to develop an artificial intelligence framework that will not only allow inflow of investments in AI but building AI capabilities that will enable them to stay competitive on a global scale. It will allow companies to be able to explain to customers the decisions made by or in assistance of AI and ensure that their AI solutions are human-centric. 4| Inter-Agency Taskforce: In a recent announcement made in early March by the Foreign Minister, the Singapore Government will set up an inter-agency task force dedicated to AI. The interagency task force will asses how the country should develop artificial intelligence as a strategic capability and be a global testbed for the deployment of AI applications. It aims at using AI to provide better services for citizens. 5| 100 Experiments: The flagship programme by AI Singapore, it aims at solving challenges around AI adoption and solve the industries’ artificial intelligence problem statement and help companies build their own AI team. As a part of this initiative, organisations can propose 100E problem statements for which if there are no AI solutions that exist, it can be brought up to be solved by AI researchers and engineering team within 9 to 18 months. Premier research institutions and universities in Singapore work on organisation’s problem statement. 6| Driving Smart Nation Through AI Education: Singapore government is ambitious about strengthening its AI play and taking initiatives to introduce it in the education system may be one of the best ways. Under the country’s TechSkills Accelerator programme, it announced two initiatives to help build technical skills of students. The first programme, called AI for Everyone, aims to immerse anyone who is interested in AI in the latest AI technologies and applications. It is a free programme for the public to let them understand the potential of AI technologies. Another initiative called AI for Industry is aimed at educating industry professionals who are more technologically inclined. This programme will offer skill development in areas such as Python and is reportedly aiming to train 2000 specialists in basic AI competency. Our Views There is no doubt that the Singapore government has identified AI as one of the four core technologies that need to be strengthened to push its digital-ready image. With the various initiatives that it has taken along with its focus on democratising AI tells us about the efforts that Singapore is ready to take to scale up AI development. Not just making them digital-ready, AI might also help them fight some of its economic challenges such as slowing growth, falling capital investments, soft workforce growth and decelerated productivity. While Singapore has succeeded at many levels to be a Smart Nation, it cannot be independent of the development in AI. To make the country digitally fit, it is important to endorse computational thinking and data literacy so that they are more equipped in areas such as data science and AI. This focus on AI is definitely going to drive growth and having witnessed Singapore’s ambition in establishing itself as Smart Nation, it wouldn’t be surprising to see Singapore becoming an AI leader too.","excerpt":"Since early 2014, Singapore has been driving an initiative to be a leader in emerging tech and establish itself as a Smart Nation. To make its ambitious vision a reality, the government has done everything, from infusing investments to framing policies to support it. Early in 2018, Singapore launched a Digital Government Blueprint to support […]","categories":["Deep Tech"],"tags":["leader of ai"],"author_name":"Srishti Deoras","publish_date":"2019-03-20T07:23:00","publication_year":"2019","word_count":1022,"keywords":["data science","Go","machine learning","artificial intelligence","AI","RAG","Python","Aim","deep learning","R","leader of ai"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","deep learning","data science","Aim","RAG","Python","R","Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/from-smart-nation-to-ai-leader-singapore-is-on-a-fast-track-to-success\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":1976,"title":"Airtel plans to venture in the home automation space with IoT","content":"Aiming high for its Internet of Things and machine-to-machine solution, Airtel is planning to enter the home automation or the smart home segment. The move which comes as a step towards increasing their revenue base, the new solutions would enable users to control elements such as lighting, heating, air conditioning, music and security systems through their smartphones. With smart homes becoming a technology to reckon with, there are many companies such as Digilux in the startup space and the well known players such as LG and Samsung, that are venturing into this space. Though there is no clear picture on the products that the telecom major would be offering, the company is quite clear of its intentions of connecting millions of homes through its smart technology. The space being in its nascent stage, it might take Airtel close to a year or two to bring contemporary solutions defining the space. A recent report published by Zion Market Research titled “Smart Home Market: Global Industry Perspective, Comprehensive Analysis and Forecast, 2016-2022”, global smart home market was valued at around USD 24.10 billion in 2016 and is expected to reach approximately USD 53.45billion in 2022, growing at a CAGR of slightly above 14.5% between 2017 and 2022. This clearly suggests the growth that this sector is witnessing and calls for an opportunity for various players to indulge into this space. With various devices such as lighting, AC, CCTV, cameras, smart TV, washing machine etc. being controlled by smart devices such as smartphone or tablet, there has been an increased convenience and comfort from basic security monitoring to entertainment and much more. All the major Indian cities—Delhi, Bangalore, Pune, Mumbai are adopting the smart technologies to the fullest. And in such a scenario, the move by Airtel would prove beneficial for both its customers and the company.","excerpt":"Aiming high for its Internet of Things and machine-to-machine solution, Airtel is planning to enter the home automation or the smart home segment. The move which comes as a step towards increasing their revenue base, the new solutions would enable users to control elements such as lighting, heating, air conditioning, music and security systems through […]","categories":["AI News"],"tags":["hpe internet of things","IoT India"],"author_name":"Srishti Deoras","publish_date":"2017-05-02T13:03:29","publication_year":"2017","word_count":304,"keywords":["programming_languages:R","AI","IoT India","automation","Aim","hpe internet of things","R","startup"],"extracted_tech_keywords":["AI","Aim","R","automation","startup","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/airtel-plans-venture-home-automation-space-iot\/","complexity_score":3,"technical_depth":6,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":30528,"title":"Indian Fintech Startup ftcash Wins MIT’s Inclusive Innovation Challenge","content":"Mumbai-based startup ftcash, known for driving financial inclusion with its innovative solutions around Digital Payment and Instant Loan has won the global grand prize winner for the 2018 MIT Inclusive Innovation Challenge, a company press statement indicated. The much-awarded Mumbai startup, incubated by PayPal emerged as a winner from more than 1,500 global registrants for its scalable solutions that leverage machine learning and analytics. The startup launched in mid-2015 received a prize of $250,000 for Most Innovative Future of Work in the World. According to the startup’s website, the company uses its proprietary algorithm to understand the credit-worthiness using their transaction flow data on the payments platform along with several other data points which can are used to provide underserved small and medium enterprises with institutional finance. After the loans are disbursed, the collections is done using the payments platform for their future receivables. Vaibhav Lodha, co-founder of ftcash said, “It is our absolute pleasure and honour to be recognized by MIT. This award represents the highest achievement in inclusive technology, but in our case, it is a validation and reflection of our mission of financial inclusion that resonates with millions of individuals globally”. He further added how the technology is enabling small business owner in India without a credit history gain access to credit or digital payments. “With MIT’s unparalleled strength in technology and their commitment to Digital Economy and Artificial Intelligence, we are looking forward to a successful partnership that can allow us to harness the best available resources in the world,” he noted. According to Erik Brynjolfsson, Director of the MIT Initiative on the Digital Economy, the grand challenge has been created to acknowledge companies which are using digital technologies to create shared prosperity.","excerpt":"Mumbai-based startup ftcash, known for driving financial inclusion with its innovative solutions around Digital Payment and Instant Loan has won the global grand prize winner for the 2018 MIT Inclusive Innovation Challenge, a company press statement indicated. The much-awarded Mumbai startup, incubated by PayPal emerged as a winner from more than 1,500 global registrants for […]","categories":["AI News"],"tags":["best technology to learn for future"],"author_name":"Richa Bhatia","publish_date":"2018-11-22T07:55:19","publication_year":"2018","word_count":287,"keywords":["Go","machine learning","artificial intelligence","AI","innovation","Scala","Git","RAG","analytics","R","best technology to learn for future"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","analytics","RAG","R","Go","Scala","Git","innovation"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/indian-fintech-startup-ftcash-wins-mits-inclusive-innovation-challenge\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10024118,"title":"EU Mulls Ban On Using AI For Mass Surveillance &#038; Social Credit Scores","content":"The European Union is considering a ban on the use of AI for a slew of use cases, including mass surveillance and social credit scores, according to a leaked draft proposal by the European Commission, first accessed by Politico. An official statement is expected next week. https:\/\/twitter.com\/teemu_roos\/status\/1382218632588050434 One of the proposals in the draft recommends European Commission to ban certain use-cases of AI and limit its use for other applications if they don’t meet certain standards. The document recommends a ban on the use of AI for mass surveillance or developing a social credit score system. The draft proposal also seeks special authorisation to use ‘remote biometric identification systems’ and demands an explicit notification to people when they interact with AI systems, ‘unless it’s obvious’. It also calls for oversight on “high-risk” AI systems that pose a direct threat to someone’s safety, like self-driving cars or systems that directly impact livelihood like AI systems used for hiring, assigning recidivism scores, or allocation of personal loans. The new #AI regulation is market and innovation-oriented. By regulating high-risk applications only, it leaves out many other uses, including the necessary assessment of workplace risks, which go way beyond HR #algorithmic managementhttps:\/\/t.co\/V0myZykenm via @Verge— Aida Ponce Del Castillo (@APonceETUI) April 15, 2021 Member states of the EU would be required to set assessment boards to test and validate the high-risk AI systems. The draft proposal also calls for a ‘European Artificial Intelligence Board’ with representation from all member states to help the European Commission identify which systems can be classified as high-risk. Companies that don’t comply could be fined up to €20 million or 4% of their turnover. While the US and China have focused their attention on developing powerful AI systems, they have fallen short in setting up an airtight regulatory framework to ensure individual safety and rights. The EU has set up a powerful GDPR (General Data Protection Regulation) to address such issues. This draft proposal is also in line with EU’s “human-centric” approach towards developing AI. Can’t wait for the massive PR back paddling of Big Tech to rebrand their #AI technology in Europe“Noooo we’re not doing AI, we’re just using logistic regression!”https:\/\/t.co\/d3YF4QZzZH— Luca Foschini (@calimagna) April 14, 2021 However, the leaked draft has drawn flak from policy wonks, calling for improvements in terms of the ambiguity of the language used. The experts demand more clarity on what constitutes AI and the definition of what’s detrimental or high-risk in AI use cases. Transparency gets a nod in the leaked draft EU AI regulation. But thin on detail. Does transparency apply to all risk levels and, if so, what will be mandatory? https:\/\/t.co\/X4lrFoVcqo#AI #transparency #AIAAIC #EU— AIAAIC (@AiControversy) April 15, 2021","excerpt":"The European Union is considering a ban on the use of AI for a slew of use cases, including mass surveillance and social credit scores, according to a leaked draft proposal by the European Commission, first accessed by Politico. An official statement is expected next week.  One of the proposals in the draft recommends European […]","categories":["AI News"],"tags":["AI Regulation","european commission","european union"],"author_name":"Kashyap Raibagi","publish_date":"2021-04-16T13:45:31","publication_year":"2021","word_count":449,"keywords":["Go","european union","artificial intelligence","european commission","programming_languages:R","AI","innovation","programming_languages:Go","AI Regulation","R"],"extracted_tech_keywords":["AI","artificial intelligence","R","Go","innovation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/eu-mulls-ban-on-using-ai-for-mass-surveillance-social-credit-scores\/","complexity_score":3,"technical_depth":7,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10171561,"title":"OpenAI’s o3 is Genius, Scores 135 in Toughest IQ test","content":"OpenAI’s o3 model has emerged as the most cognitively capable AI system in a new benchmark test conducted by Voronoi, based on data from Tracking AI. The test, which uses Norway’s Mensa IQ test, a high-difficulty assessment typically reserved for human intelligence evaluation, placed o3 at an IQ score of 135, well above the human average of 90–110. Other high scorers include Anthropic’s Claude-4 Sonnet at 127 and Google’s Gemini 2.0 Flash at 126. The analysis covered 24 leading AI models, with top positions mostly occupied by text-only models, while vision-enabled systems scored significantly lower. GPT-4o with vision, for example, received an IQ score of 63, while Grok-3 Think (Vision) followed with 60. These results suggest that while language-based reasoning capabilities in AI are rapidly surpassing human benchmarks, vision-based and multimodal systems still lag in abstract problem-solving tasks. The test results raise important questions about how AI models are architected and trained, particularly when it comes to general intelligence versus domain-specific strengths. Voronoi’s findings reflect a broader trend in AI development, where performance gains in language models continue to dominate, but genuine multimodal reasoning remains a key challenge.Not Thinking, Yet Coincidentally, in a new paper titled The Illusion of Thinking, researchers from the Cupertino-based company argued that even the most advanced AI models, including the so-called large reasoning models (LRMs), don’t actually think. Instead, they simulate reasoning without truly understanding or solving complex problems. The paper, released just ahead of Apple’s Worldwide Developer Conference, tested leading AI models, including  OpenAI’s o1\/o3, DeepSeek-R1, Claude 3.7 Sonnet Thinking, and Gemini Thinking, using specially designed algorithmic puzzle environments rather than standard benchmarks. The researchers argue that traditional benchmarks, like math and coding tests, are flawed due to “data contamination” and fail to reveal how these models actually “think”. “We show that state-of-the-art LRMs still fail to develop generalisable problem-solving capabilities, with accuracy ultimately collapsing to zero beyond certain complexities across different environments,” the paper noted.","excerpt":"The reasoning capabilities in vision-based and multimodal systems still lag in abstract problem-solving tasks.","categories":["AI News"],"tags":["IQ","LLMs","OpenAI"],"author_name":"C P Balasubramanyam","publish_date":"2025-06-10T18:19:12","publication_year":"2025","word_count":322,"keywords":["IQ","Anthropic","Gemini 2.0","Go","API","OpenAI","AI","LLMs","GPT-4o","RAG","GPT","R"],"extracted_tech_keywords":["AI","GPT-4o","OpenAI","Anthropic","Gemini 2.0","RAG","R","Go","API","GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/openais-o3-is-genius-scores-135-in-toughest-iq-test\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10065635,"title":"EU comes out with new law mandating big techs to reveal their algorithms","content":"The European Union announced that the Parliament and member states have agreed on a new set of laws called the Digital Services Act, or DSA. Once passed, the big tech companies like Google and Meta would be forced to reveal how their algorithms work, change their approaches to targeted advertising, etc. The EU will have a final vote on the DSA once the language is finalised, with the law taking place either 15 months after the vote or at the start of 2024, whichever comes later. New obligations include removing illegal content and goods more quickly; explaining to users and researchers how their algorithms work; taking stricter action on the spread of misinformation. Upon any violation, the companies will face fines of up to six per cent of their annual turnover for non-compliance. European Commission President Ursula von der Leyen said that the DSA would upgrade the ground rules for all online services. It will give practical effect to the principle that what is illegal offline should be illegal online. The greater the size, the greater the responsibilities of online platforms. “Today’s agreement on the Digital Services Act is historic, both in terms of speed and of substance. The DSA will upgrade the ground-rules for all online services in the EU. It will ensure that the online environment remains a safe space, safeguarding freedom of expression and opportunities for digital businesses,” she said. While the DSA is a European law, big tech companies might be inclined to adopt the rules worldwide to save time and energy on creating different rule sets for different regions.","excerpt":"The new regulation could affect users in the US and elsewhere.","categories":["AI News"],"tags":["Google","Meta"],"author_name":"Poornima Nataraj","publish_date":"2022-04-25T12:59:36","publication_year":"2022","word_count":264,"keywords":["Go","Meta","AWS","AI","cloud_platforms:AWS","programming_languages:R","programming_languages:Go","Git","Google","R"],"extracted_tech_keywords":["AI","AWS","R","Go","Git","cloud_platforms:AWS","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/eu-comes-out-with-new-law-mandating-big-techs-to-reveal-their-algorithms\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10065552,"title":"All you need to know about time-series clustering","content":"Time series data is one of the major types of data that is being generated in a huge way nowadays. From financial analysis to industrial machinery analysis has the existence of time series data. Generally, we think of time series data as supervised data but when its generation becomes vast and fluctuates it can behave like unsupervised data. Time-series clustering is a process that can help us in dealing with such fluctuating and vast time series data. In this article, we are going to discuss time series clustering with its key concepts and we will also understand how it can be utilized practically. The major points to be discussed in the article are listed below. Table of content What is time series clustering?Time-series as matrixImplementation Data extraction Data preprocessingClustering Agglomerative clusteringK-means clusteringK-Medoids clustering What is time series clustering? We can consider the time series clustering is an unsupervised learning problem similar to the clustering in other domains but one thing that makes this clustering different from other clustering is that this includes time values in clusters. In the present scenario where the generation of data, mainly the time series data, is at a very high rate and even the generation of data is also a time series, in such a situation, it becomes very difficult to understand the patterns in the data by just performing EDA on it. Let’s take a look at the below graph: Image source This graph consists of the visualization of a noisy time series on the upside with blue color.  By just looking at the upper trend line it is very difficult to make any inference from the data. To make it easy the green line is the filtered signal of the upper trend. Looking at the signal it is not easy to interpret where the series is temporal and where it is spatial. With such a time series, we are required to segregate it into parts so that we can understand the scenario much better. It can be considered as the process of organizing data points into groups based on their similarities. We mainly perform time series clustering to minimize the data similarity across the clusters and increase the similarity within the cluster. There are various ways to cluster the time series such as: Agglomerative clustering: This type of clustering includes the distance matrix to cluster the time series data set.  Time series K-Means: It is a very basic way that can include euclidean, dynamic time warping, or soft dynamic time warping. Kernel K-Means: This method is similar to the time series k-means but in it for clustering data, we rely on the kernel trick that mainly introduces a positive semidefinite.K-means DBA clustering: This type of clustering includes averaging strategy DTW barycenter averaging for clustering the time series data. K-Medoid clustering: This type of clustering includes a distance matrix to perform k-medoid clustering. In this section we got to know about what is this clustering, by the above points we can say that there is always a requirement to cluster a time-series data before modelling it. In this article, we are going to see how we can perform some of these above-given methods of clustering. One thing that is noticeable in this position is we need to think of time series as matrices not as series to perform clustering. Are you looking for a complete repository of Python libraries used in data science, check out here. Time-series as matrix In the above points, we have seen that this type of clustering is required but this clustering should not be interpreted as segregating time series by time values or decomposing a time series into trend, seasonality, and noise. However, these things also matter at the time of time series modeling but there is no relation between them with this type of clustering. In this process, we try to segregate time-dependent variables into clusters so that we can increase the similarity between time-dependent values within the cluster. As mentioned above this process can be considered a solution to unsupervised learning problems that do not consider time values but consider the time-dependent variables. Just like other clustering methods, we may think of this series as a matrix without taking time values in the process. The implementation we are going to do will convert our time series into the matrix and this concept will get clearer. Let’s take a look at how we can do that. Clustering the Time-series Data extraction In most data science-related works, we are required to have data. In this article, we are going to use the share market price data of the State bank of India. We can use this data using the following lines of codes. import datetime as dt from datetime import datetime as dt from dateutil.relativedelta import relativedelta import yfinance as yf end = dt.today() start = dt.today() - relativedelta(months=1) data = yf.download('SBIN.NS', start, end) data.tail() Output: Here we can see the last values of our extracted data. For this experiment, I have extracted data from last month that needs to be clustered. In the data, we can see that there are 6 time-dependent values: 5 are for the price of a share and one is representing the volume of a share. Now start our preprocessing of the data. Data preprocessing As we discussed above, time series clustering is a process that makes us think of series as a matrix because it clusters the time series based on the time-dependent values, not by time. Before converting the series into the matrix, let’s visualize the data. data.plot() data.Open.plot(secondary_y=True, label=\"open\") data.High.plot(secondary_y=True, label=\"high\") data.Low.plot(secondary_y=True, label=\"low\") data.Close.plot(secondary_y=True, label=\"close\") data['Adj Close'].plot(secondary_y=True, label=\"adj close\") plt.show() Output: Here we can see how the time series is moving but these values are pretty difficult to explain. Let’s start its conversion to perform clustering of time series. Converting a series into the matrix is very easy we just need to use the .to_numpy module of pandas in the following way: series = data.to_numpy() series Output The above output represents the required form of data. Now let’s plot this form of data. plt.plot(series) Output: Here we can see the sum-up form of our converted data. Let’s start the clustering. Time-series clustering In the first section of the article, we have seen that different types of clustering can be performed in the time series, and in this section, we will look at how we can perform them. For performing clustering we are going to use a package named as DTAIdistance, this package provides various functions to perform time-series processing related to a distance matrix and clustering analysis. Let’s take a look at how we can perform agglomerative clustering. Agglomerative clustering We can perform hierarchically and linkage clustering using this method. Let’s see how we can do that. Hierarchical clustering from dtaidistance import clustering model1 = clustering.Hierarchical(dtw.distance_matrix_fast, {}) cluster_idx = model1.fit(series) model2 = clustering.HierarchicalTree(model1) cluster_idx = model2.fit(series) Let’s plot this clustering. import matplotlib.pyplot as plt fig, ax = plt.subplots(nrows=1, ncols=2, figsize=(10, 10)) show_ts_label = lambda idx: \"ts-\" + str(idx) model2.plot(axes=ax, show_ts_label=show_ts_label, show_tr_label=True, ts_label_margin=-10, ts_left_margin=10, ts_sample_length=1) Output: Here we can cluster time series using the distance between matrices. Linkage clustering model3 = clustering.LinkageTree(dtw.distance_matrix_fast, {}) cluster_idx = model3.fit(series) Let’s plot the clusters. import matplotlib.pyplot as plt fig, ax = plt.subplots(nrows=1, ncols=2, figsize=(10, 10)) show_ts_label = lambda idx: \"ts-\" + str(idx) model3.plot(axes=ax, show_ts_label=show_ts_label, show_tr_label=True, ts_label_margin=-10, ts_left_margin=10, ts_sample_length=1) Output: K-means clustering from dtaidistance.clustering import kmeans model4 = kmeans.KMeans(k=10) cluster_idx, performed_it = model4.fit(series) Here we have used 10 centers to make clusters using k-means clustering. Let’s plot the clusters. import matplotlib.pyplot as plt fig, ax = plt.subplots(nrows=1, ncols=2, figsize=(10, 10)) show_ts_label = lambda idx: \"ts-\" + str(idx) model4.plot(axes=ax, show_ts_label=show_ts_label, show_tr_label=True, ts_label_margin=-10, ts_left_margin=10, ts_sample_length=1) Output: K-Medoids clustering To perform this type of clustering between time series we may need to define the distance matrix method. The package we are using will require downloading a pyclustering package to perform the distance matrix method. We can perform this clustering using the following lines of codes. from dtaidistance import dtw, clustering model5 = clustering.KMedoids(dtw.distance_matrix_fast, {}, k=10) cluster_idx = model5.fit(series) Let’s plot the clusters. model5.plot() Output: Here we can see the clusters made using the k-medoids clustering algorithm. Final words In this article, we have discussed the time series clustering that is making clusters of time series values without considering time factors in it. Along with this, we have looked at how we can do different types of clustering In time series using the python language. References Link to the above codes","excerpt":"In this article, we are going to discuss time series clustering with its key concepts and we will also understand how it can be utilized practically.","categories":["Deep Tech"],"tags":[],"author_name":"Yugesh Verma","publish_date":"2022-04-24T10:00:00","publication_year":"2022","word_count":1410,"keywords":["data science","NumPy","Go","TPU","AI","RAG","Python","Matplotlib","R","Pandas"],"extracted_tech_keywords":["AI","data science","Pandas","NumPy","Matplotlib","RAG","TPU","Python","R","Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/all-you-need-to-know-about-time-series-clustering\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10005897,"title":"Behind IBM’s Autonomous Chemistry Lab","content":"With no stoppage of the COVID pandemic in the foreseeable future, drug discovery has been a critical concern for the healthcare industry. But, in reality, it takes around a decade with approximately $10 million to discover a new drug, get it tested and bring it to the market. Such a delayed process can be majorly attributed to the tedious process of repetitive experiments to bring out the drug. Thus, to omit this significant hurdle in drug discovery, IBM has launched an autonomous chemistry lab combining AI, robotics and cloud technology — RoboRXN. A cloud-based AI model, RoboRXN has been developed to perform autonomously, like a robot, using machine learning algorithms to produce chemicals and drugs in a laboratory with little to no human intervention. Not only a development like that can optimise the drug discovery process by reducing the production time, but can also lower down the cost, allowing scientists to bring out new drugs in health crises, keeping the norms of social distancing in place. Teodoro Laino, a researcher at IBM, stated in the company blog that chemists and scientists rely on traditional processes in discovering drugs and other chemicals, and little progress has been made to modernise those processes of trial and error. And therefore, IBM has automated this process to accelerate material discovery. Also Read: Google’s New Machine Learning-Based Initiative For Accelerated Drug Discovery Overview Of IBM’s Autonomous Lab According to the blog post by Laino, the development of RoboRXN initiated when IBM began working on developing ML models to predict chemical reactions and offered it as a free service — called it as RXN for Chemistry. The tool leveraged neural machine learning translation method for predicting the product of the chemical reactions, using the SMILE representation to identify chemical entities. Using SMILE representation, this molecule is translated into BrCCOC1OCCCC1 However, mixing all those ingredients and experimenting with the trial and error methods with all the components was still time-consuming. And thus, in an attempt to reduce that, IBM combined AI with cloud and automation technology to bring our RoboRXN, as a remotely accessible lab for better production of drugs. Launched as an online tool — RoboRXN works through a web browser, where the scientists can use the blank canvas to type and draw the required “skeletal structure of the molecular compounds.” With that information, the online platform uses ML algorithms to predict the components needed and learn the right process of synthesising the elements. To facilitate this, the company designed the algorithm to precisely extract the necessary synthesis information and converted the same in a machine-readable format. This information is then sent to a robot placed in the independent lab to execute the formula. The platform also sends a real-time report to the scientists with the results. PC: IBM Research RoboRXN further claimed on the company blog that it does not rely its workings on individual entities of sentences, nor does it require to be mentioned on the words that need corresponding actions. This, in turn, allows the model to be more flexible and accurate. Furthermore, in order to provide data input for the ML model, the company also developed an annotation framework that generated examples on a variety of sentences with “synthesis procedures and corresponding operations.” Moreover, IBM claims that RoboRXN for predicting reactions can easily outstrip all other data-driven models with 90% accuracy. To build the reaction dataset, the company extracted the chemical information using NLP called Molecular Transformer. When asked, Laino explained, in the company blog, the importance of cleaning up the data in order to achieve an accurate prediction model. The model has been trained on a large number of chemical recipes and also been designed to learn the specifics of chemicals in order to recommend the correct sequence of the components. Not only RoboRXN is hardware agnostic, but also envelopes AI technology to execute a complex series of chemical procedures. With its cloud-native design, it has been designed to be accessed from anywhere remotely. Also Read: Artificial Intelligence swarms drug discovery. Is India catching up too? Wrapping Up Claiming to revolutionise industrial chemistry, the launch of RoboRXN, with its reliable results, can be extremely beneficial for drug discovery amid this COVID pandemic. The combination of robotics, along with artificial intelligence and cloud, can not only reduce the time of the discovery process but also create more emphasis on design and innovation. With this development, it is hoped to be beneficial for researchers and chemists with the new treatment for COVID-19 or any future viruses possible.","excerpt":"With no stoppage of the COVID pandemic in the foreseeable future, drug discovery has been a critical concern for the healthcare industry. But, in reality, it takes around a decade with approximately $10 million to discover a new drug, get it tested and bring it to the market. Such a delayed process can be majorly […]","categories":["AI Features"],"tags":[],"author_name":"Sejuti Das","publish_date":"2020-09-01T16:00:18","publication_year":"2020","word_count":754,"keywords":["Go","machine learning","artificial intelligence","AI","ML","RAG","NLP","Aim","GAN","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","NLP","Aim","RAG","R","Go","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/behind-ibms-autonomous-chemistry-lab\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10065577,"title":"Talking Ethical AI with Vuram’s Archit Agrawal","content":"Founded in 2011, Vuram is a global hyper-automation services company specialising in low-code enterprise automation. “Vuram’s hyper-automation technology stack encompasses business process management (BPM), robotic process automation (RPA), optical character recognition (OCR), document processing, AI, machine learning, and analytics,” said Archit Agrawal, Product Manager, Vuram. In an exclusive interview with Analytics India Magazine, Archit spoke about Vuram’s approach toward Responsible and Ethical AI. AIM: Tell us about Vuram’s Responsible and Ethical AI frameworks. Archit Agrawal: Making sensible judgments at every level of the organisation that does not hurt individuals or society as a whole is one of Vuram’s core principles. To ensure that teams represent a broad range of experiences and opinions, we strive to maintain a rich and varied team that spans genders, ages, races, disciplines, and backgrounds. Vuram is committed to information security and customer data protection and is ISO 27001 certified and is in process of becoming SOC2 certified. Key principles: Fairness and inclusiveness: AI systems should treat and represent everybody with equal engagement and empowerment.Privacy and security: AI systems should be safe and secure, with no sensitive information exposed.Accountability and interpretability: AI systems should be receptive to input and give appropriate explanations. A human should be involved in the process.Reliability and safety: AI systems should be created using the most up-to-date safety and security safeguards to avoid any adverse negative outcomes. Best practices for Responsible AI: Design AI systems keeping users in mind.Engage a wide range of users and use-case situations and incorporate comments both before and throughout the development.Identify multiple metrics and indicators for assessing, training, and monitoring.Examine the raw data directly. Use aggregators and summarises in case of sensitive data.Understanding of data set and model limitations.Test against a golden data set and proactively search for unintended or wrong results – keep testing.Identify unintended bias before scaling.Continue to monitor and update the system after deployment.Strengthen compliance with current laws and regulations while monitoring the future ones and develop policies to mitigate risk. AIM: What explains the growing conversation around Responsible AI? Why is it the need of the hour? Archit Agrawal: Leaders are under increasing scrutiny to ensure their companies’ ethical use of AI systems goes beyond the word and spirit of current regulations. When it comes to high-stakes AI applications like autonomous weaponry and surveillance systems, ethical disputes about what is “good” and “wrong” are raging. And there’s a lot of worry and scepticism about how we will be able to imbue AI systems with human ethical judgement, significantly because moral norms vary by culture and are difficult to define in software. Numerous reports of AI prejudice, discrimination, and privacy violations have already surfaced in the media, prompting leaders to wonder how they can assure that nothing goes wrong when they roll out their AI systems. Unintentional bias in AI systems can result in incorrect outcomes, resulting in fairness issues that harm business. AIM: How does Vuram ensure adherence to its Responsible and Ethical AI policies? Archit Agrawal: To guarantee that team members are taught how to respect our AI ethical obligations, we use mandatory ethics training modules, toolkits, seminars, and workshops. A human-centred design-thinking session, for example, assist teams to understand our commitment to developing ethical machine learning technology. AIM: How do you mitigate biases in your AI algorithms? Archit Agrawal: Analyse the algorithm and data to determine where there is a high danger of unfairness. Check to see if the training data set is representative and large enough to avoid common biases like sampling bias.Calculate model metrics for specific groups in the data set as part of subpopulation analysis. This can assist in determining whether the model’s performance is consistent across subpopulations.Over time, keep an eye on the model for biases. ML algorithms’ output can alter as they learn or as the training data changes. Create a debiasing approach that includes a mix of technological, operational, and organisational steps: Technical strategy by using tools to uncover potential sources of bias and reveal characteristics in the data that affect the model’s accuracy.Using an internal governance team and third-party auditors.Establishing a workplace where metrics and processes are transparent. As we discover biases in training data, we improve human-driven processes. Model construction and evaluation might reveal biases that have been hidden for a long time. We discover these biases and use this knowledge to understand the reasons for bias while constructing AI models. We also improve the actual process to decrease bias through training, process design, and cultural changes. Additionally, we must always determine beforehand when humans should be involved and when automated decision-making should be used. Adoption of an interdisciplinary strategy: The importance of research and development in reducing bias in data sets and algorithms cannot be overstated. Eliminating bias is a multidisciplinary technique including ethicists, social scientists, and professionals who are most familiar with the complexities of each application field. As a result, businesses should seek out such professionals for their AI initiatives. Diversity: Having a diverse AI team helps avoid unintended AI biases. Tools to reduce bias: IBM Watson OpenScale, AI Fairness 360, Google’s What-If Tool AIM: Do you have a due diligence process to ensure the data you use is collected ethically? Archit Agrawal: When it comes to third-party data and models, our governance and AI teams double-check all the paperwork and underlying contracts before incorporating them into our products and initiatives. The more data used to train systems, the more accurate and insightful the forecasts and predictions become. Our data science teams are careful about where they get the data from and how they use it. We ensure that data sets accurately reflect all of the populations being studied, as underrepresentation of some groups might result in different outcomes Data science teams assess how they sampled data to train their models. AIM: How does Vuram protect user data? Archit Agrawal: We take the following steps to ensure data privacy: Restrict accessAudit systemsPlugin updatesPatches, firewalls and encryptionHuman error and device controlStrategising a plan BDigital twin AIM: Did you come across any biases or ethical issues within your organisation? If yes, how did you address them? Archit Agrawal: Because AI is not deterministic, we must often re-train it. We recently had similar concerns with our Personal Information Extractor model, which worked well on training data but not so well on global data, thus we opted to switch down the system after doing impact analysis. We had to re-train the model and revalidate it.","excerpt":"We must always determine beforehand when humans should be involved and when automated decision-making should be used.","categories":["AI Features"],"tags":["AI fairness","ai governance","consumer protection","Data Governance","Data Privacy","data rights","digital transformation","Ethical AI","Interviews and Discussions","modernisation"],"author_name":"Sri Krishna","publish_date":"2022-04-23T13:00:00","publication_year":"2022","word_count":1071,"keywords":["consumer protection","data rights","TPU","ai governance","R","data science","digital transformation","RAG","Data Governance","analytics","modernisation","machine learning","AWS","AI","Ethical AI","ML","Data Privacy","AI fairness","Aim","Interviews and Discussions"],"extracted_tech_keywords":["AI","machine learning","ML","data science","analytics","Aim","RAG","AWS","TPU","R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/talking-ethical-ai-with-vurams-archit-agrawal\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10042883,"title":"How To Optimise Deep Learning Models","content":"Increasing number of parameters, latency, resources required to train etc have made working with deep learning tricky. Google researchers, in an extensive survey, have found common challenging areas for deep learning practitioners and suggested key checkpoints to mitigate these challenges. For instance, deep learning practitioner might face the following challenges when deploying a model: Training could be a one-time cost, deploying and letting inference run for over a long period of time could still turn out to be expensive in terms of consumption of server-side RAM, CPU, etc. Using as little data as possible for training is critical when the user-data might be sensitive. New applications come with new constraints (around model quality or footprint) that existing off-the-shelf models might not be able to address. Deploying multiple models on the same infrastructure for different applications might end up exhausting the available resources. Most of these challenges boil down to lack of efficiency. According to Gaurav Menghani of Google Research, if one were to deploy a model on smartphones where inference is constrained or expensive due to cloud servers, attention should be paid to inference efficiency. And if a large model has to be trained from scratch with limited training resources, models that are designed for training efficiency would be better off. According to Menghani, practitioners should aim to achieve pareto-optimality i.e.  any model we choose should have the best of tradeoffs. And, one can develop a pareto-optimal model using the following mental model. Image credits: Gaurav Menghani, Google Model compression For very large models, maximising training efficiency can be achieved using models with small hidden sizes or few layers as the models run faster and use less memory. As illustrated above, a common practice would be to train small models until they converge and then run a compression technique lightly. Another example is quantisation: the weight matrices of a layer are compressed by reducing its precision from 32-bit floating point values to 8-bit unsigned integers without sacrificing quality. The most compute-efficient training strategy is to counterintuitively train extremely large models but stop after a small number of iterations. Here are a few popular compression techniques: Parameter Pruning And SharingLow-Rank FactorisationTransferred\/Compact Convolutional FiltersKnowledge Distillation Learning techniques Learning algorithms focus on training the mode (to make fewer prediction errors), require less data and converge faster. According to Menghani, the improved quality can then be exchanged for a more efficient model by trimming the number of parameters if needed. An example of such learning technique is distillation, where the accuracy of a smaller model is improved by learning to mimic a larger model. Earlier this year, Facebook announced Data-efficient Image Transformer (DeiT), that leverages distillation. It is a vision Transformer built on a transformer-specific knowledge distillation procedure to reduce training data requirements. Distillation allows neural networks to learn amongst themselves. Whereas, a distillation token is a learned vector that flows through the network along with the transformed image data to significantly enhance the image classification performance with less training data. Automation A good example of improving efficiency is through hyper-parameter optimization (HPO), an automated way of optimising the hyper-parameters to increase the accuracy, which could then be exchanged for a model with lesser parameters. Given the vast search space of features, evaluation of each configuration can be expensive. A model’s performance depends heavily on the hyperparameter optimization. HPO tools like Amazon’s Sagemaker offer an automatic model tuning module and all one would need is just a model and related training data for a HPO task. Auto tune modules support optimisation for complex models and datasets with parallelism on a large scale. Microsoft’s Neural Network Intelligence (NNI), an open-source toolkit for both automated machine learning (AutoML) and HPO, provides a framework to train a model and tune hyper-parameters along with the freedom to customise. ENAS Neural architecture search (NAS) falls in this category too. In his case, the architecture itself is tuned and search components skim through the results to help find a model that optimises both the loss and accuracy, and some other metric such as model latency, model size, etc. NAS methods have already outperformed many manually designed architectures on image classification tasks. For example, Efficient Neural Architecture Search (ENAS), a fast and inexpensive approach for automatic model design. In this approach, a controller (RNN), trained with policy gradients, learns to discover neural network architectures by searching for an optimal subgraph within a large computational graph. Architectures Neural architectures are fundamental blocks designed from scratch. One of the classical examples of efficient layers in the Vision domain are the Convolutional layers. These layers are improved over Fully Connected (FC) layers in Vision models. The same filter is used everywhere in the image, regardless of where the filter is applied. Whereas, the Transformer architecture, proposed back in 2017,  demonstrated that Attention layers could be used to replace traditional RNN based Seq2Seq models. As a result, BERT and GPT architectures beat the state-of-the-art in several NLU benchmarks. Infrastructure Efficiency eventually boils down to how well software is integrated with hardware. A great algorithm with poor compute set-up can falter. This is where selecting frameworks, GPUs and TPUs become crucial. For example, Tensorflow Lite is designed for inference in a low-resource environment. PyTorch also has a light-weight interpreter that enables running PyTorch models on Mobile, with native runtimes for Android and iOS. Similar to TFLite, PyTorch offers post-training quantization, and other graph optimization steps for optimizing convolutional layers. When it comes to hardware, choosing a suitable GPU or TPU also plays into the overall efficiency. Over the years, Nvidia has released several iterations of its GPU microarchitectures with increasing focus on deep learning performance. It has also introduced Tensor Cores designed for Deep Learning applications. The aforementioned checklist offers a framework to monitor the efficiency of ML pipelines. Thumb rules: Can the model graph be compressed?Can the training be better?Using automated search for better modelsUsing efficient layers and architectures Read the original survey here.","excerpt":"Increasing number of parameters, latency, resources required to train etc have made working with deep learning tricky. Google researchers, in an extensive survey, have found common challenging areas for deep learning practitioners and suggested key checkpoints to mitigate these challenges.  For instance, deep learning practitioner might face the following challenges when deploying a model: Training […]","categories":["Deep Tech"],"tags":["deploying models","infrastructure","model compression"],"author_name":"Ram Sagar","publish_date":"2021-07-05T10:00:00","publication_year":"2021","word_count":989,"keywords":["infrastructure","machine learning","deploying models","TPU","AI","neural network","PyTorch","ML","RAG","Aim","deep learning","model compression","TensorFlow"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","neural network","Aim","TensorFlow","PyTorch","RAG","TPU"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/how-to-optimise-deep-learning-models\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10102538,"title":"Why Sony Acquired iSIZE","content":"Sony is pretty hell-bent on making its cloud gaming platform work. Recently, it announced the acquisition of iSIZE, a UK-based company which specialises in deep learning for video delivery, throwing massive hints at the company’s decision to fuel its cloud gaming ambitions. iSIZE develops AI-driven solutions aimed at providing bitrate reductions and enhancing quality in the media and entertainment sector. As Sony’s blog reads, “The acquisition provides Sony Interactive Entertainment (SIE) with significant expertise in applying machine learning to video processing, which will benefit a range of our R&D efforts as well as our video and streaming services.” iSIZE also specialises in making photorealistic neural avatars. The team has been working for almost two years on this for 2D & 3D modelling for gaming and metaverse applications. The founder of the company, Sergio Grce, has been very interested in AI, along with video imaging, art, and gaming. Thus, along with Yiannis Andreopoulos, he founded iSIZE in 2016, whose expertise was in video streaming services. This acquisition comes in line with Sony reporting its plan in May to bolster its PlayStation division further and is looking to acquire more studios for the same. Coincidently, it has also been on an acquisition spree since 2021. The PlayStation team announced that it is planning an “aggressive and interesting” expansion into cloud gaming. Sony CEO Jim Ryan said, “There’s been lots of attention around cloud gaming. We observe mobility in gaming habits to be an increasingly important trend. The cloud will be fundamental in allowing us or anyone else to exploit that trend of mobility.” Sony is actively looking for opportunities to strengthen its position in the sector, as Microsoft continues to expand its gaming presence with Xbox. The recent completion of Microsoft’s $68.7 billion acquisition of Activision Blizzard has set the stage for intense competition in the gaming industry. Sony’s latest acquisition of iSIZE is seen as a strategic response to this competitive landscape. Cloud gaming making a comeback? Apart from the acquisition of Activision Blizzard, Microsoft’s foray into the cloud gaming business reportedly has some serious development routes in India. A recently leaked Xbox internal document revealed that Reliance Jio has been exploring partnerships in the cloud gaming space with Microsoft in 2020. As per the document, in order to extend Project xCloud, Reliance Jio outlined plans to deploy xCloud servers across 27 Azure regions, encompassing both existing and upcoming locations. This list incorporated Central Indian and South Indian xCloud Servers. Microsoft has been working with Jio to bring Project xCloud into the Indian market since 2020. Later, the Mukesh Ambani-led firm also tried to partner with NVIDIA for its GeForce Now cloud gaming service in November 2022. But since the partnership did not work out, Jio ended up partnering with other smaller cloud providers such as Ubitus K.K. and announced JioGamesCloud. Possibly, Jio would be able to make a dent given its large audience in the cloud gaming market. On the other hand, Microsoft is also planning to launch PC cloud game streaming through Xbox Cloud Gaming by the end of 2028, along with leveraging its Azure servers. All of these developments are after Google decided to shut down its Stadia cloud gaming project citing lack of subscribers. Google’s Stadia, which initially aimed to be a first-party consumer service, ultimately pivoted to focus on a white-label cloud gaming service for developers. This shift in strategy emphasises the competitive nature of the cloud gaming market. Can Sony work it out this time? Amongst Microsoft, Google, and Sony taking it slow in the cloud gaming industry, NVIDIA hasn’t stopped even a bit. NVIDIA’s GeForce NOW is expanding to accommodate 15 fresh new titles on its platform, including Alan Wake 2. Sony started its streaming gaming service when it acquired Gaikai in 2012, through which it powered its PlayStation Vita, and now powers PlayStation Now. Later, Sony acquired OnLive in 2015, but only to shut it down. People on X have been jokingly saying that Sony can only acquire companies that do not have a portfolio, and then shut it down. It would be interesting to see how the acquisition of iSIZE pans out. On the other hand, Sony has been comparatively quiet when it comes to investing in generative AI. The Sony Venture Fund has been bullish on building robots, but according to Austin Noronha, the managing director at the firm, “For AI, the way we look at the company is around how differentiated will they be compared to legacy systems or large corporations — and whether they are working on their own language models and having their application stack on top of it, versus plugging in to someone else’s.”","excerpt":"iSIZE specialises in making photorealistic neural avatars, which sees application in 2D & 3D modelling within games and metaverse.","categories":["AI Features"],"tags":["Cloud Gaming","Mergers and Acquisitions","Sony"],"author_name":"Mohit Pandey","publish_date":"2023-11-06T12:43:49","publication_year":"2023","word_count":777,"keywords":["Go","Cloud Gaming","machine learning","AI","R","RAG","Ray","Aim","deep learning","generative AI","Sony","Mergers and Acquisitions","Azure"],"extracted_tech_keywords":["AI","machine learning","deep learning","generative AI","Aim","Ray","RAG","Azure","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/why-sony-acquired-isize\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10011418,"title":"6 MNIST Image Datasets That Data Scientists Should Be Aware Of (With Python Implementation)","content":"In Computer Vision, specifically, Image processing has become more efficient with the use of deep learning algorithms. To show the performance of these neural networks some basic preprocessed datasets were built, namely the MNIST and its variants such as KMNIST, QKMNIST, EMNIST, binarized MNIST and 3D MNIST. Ever since these datasets were built, it has been popular amongst beginners and researchers. In today’s article, we’ll be talking about the very basic and primarily the most curated datasets used for deep learning in computer vision. MNIST MNIST(Modified National Institute of Standards and Technology)  database contains handwritten digits. It is a subset of the larger dataset present in NIST(National Institute of Standards and Technology). Developed by Yann LeCunn, Corinna Cortes and Christopher J.C. Burges and released in 1999. This is a “hello world” dataset deep learning in computer vision beginners for classification, containing ten classes from 0 to 9. The original black and white images of NIST had been converted to grayscale in dimensions of 28*28 pixels in width and height, making a total of 784 pixels. Pixel values range from 0 to 255, where higher numbers indicate darkness and lower as lightness. MNIST database consists of two NIST databases – Special Database 1 and Special Database 3. Special Database 1 contains digits written by high school students. Special Database 3 consists of digits written by employees of the United States Census Bureau. Database Size: 21MiB Data: Total 70000 images split into -Train set 60000 images, Test set 10000 images. Performance: Highest error rate, as shown on the official website, is 12%. The original paper of MNIST showed the report of using SVM(Support Vector Machine) gave an error rate of 0.8%. Over the years, several methods have been applied to reduce the error rate. Some notable out of them are In 2004, a best-case error rate of 0.42% was achieved by using a classifier called LIRA, which is a neural classifier consisting of three neuron layers. Using affine distortions and the elastic distortions error rate of 0.39 was achieved by using a 6layer deep neural network. In 2011, 0.27 error rate was achieved using the similar architecture of a convolutional neural network(CNN). In 2013, an error rate of 0.21 using regularization and DropConnect. In 2018, an error rate of 0.18%  by using simultaneous stacking of three kinds of neural networks. As of February 2020, an error rate of 0.17 has been achieved using data augmentations with CNNs. This is best suited for beginners as it is a real-world dataset where data is already pre-processed, formatted and normalized. Researchers and learners also use it for trying on new algorithms. MNIST is taken as a reference to develop other such datasets. Code Snippet: Using TensorFlow import tensorflow as tf mnist = tf.keras.datasets.mnist (x_train, y_train),(x_test, y_test) = mnist.load_data() Using PyTorch import torch import torchvision from torchvision import transforms, datasets train = datasets.MNIST('', train=True, download=True, transform=transforms.Compose([ transforms.ToTensor() ])) test = datasets.MNIST('', train=False, download=True, transform=transforms.Compose([ transforms.ToTensor() ])) BINARIZED MNIST They were developed by Salakhutdinov, Ruslan and Murray, Iain in 2008 as a binarized version of the original MNIST dataset. Binarizing is done by sampling from a binomial distribution defined by the pixel values, originally used in deep belief networks(DBN) and variational autoencoders(VAE). The images are in grayscale format 28 x 28 pixels. Download Size – 104 MiB Data: train set 50000 images, the test set 10000 images and validation set 10000 images It is used to evaluate generative models for images, so unlike MNIST labels are not provided here. Code Snippet: Using TensorFlow import tensorflow_datasets as tfds train,test = tfds.load('binarized_mnist', split=['train', 'test']) EMNIST Extended MNIST derived from MNIST in 2017 and developed by Gregory Cohen, Saeed Afshar, Jonathan Tapson, and André van Schaik. EMNIST is made from the NIST Special Database 19. The original NIST data is converted to a 28×28 pixel image format and structure matches that of MNIST dataset. Download Size: 535.75MB The six different splits provided in this dataset: EMNIST ByClass: 814,255 characters with 62 unbalanced classes.EMNIST Balanced:  131,600 characters with 47 balanced classes.EMNIST Digits: 280,000 characters with 10 balanced classes.EMNIST MNIST: 70,000 characters with 10 balanced classes.EMNIST Letters: 145,600 characters with 26 balanced classes.EMNIST ByMerge: 814,255 characters with 47 unbalanced classes. Code Snippet: Using tensorflow import tensorflow_datasets as tfds train,test = tfds.load('emnist', split=['train', 'test']) Using PyTorch import torch import torchvision from torchvision import datasets train = datasets.EMNIST('',train = True, split='byclass',download= True) test = datasets.EMNIST('',train = False, split='byclass',download= True) KMNIST Kuzushiji MNIST Dataset developed by Tarin Clanuwat, Mikel Bober-Irizar, Asanobu Kitamoto, Alex Lamb, Kazuaki Yamamoto and David Ha for Deep Learning on Classical Japanese Literature. KMNIST is a drop-in replacement for the MNIST dataset (28×28 pixels of grayscaled 70,000 images), consisting of original MNIST format and NumPy format. Dataset Size- 31.76 MiB Download Size – 300MB Data: train set 60000 images, the test set 10000 images Code Snippet: Using Tensorflow import tensorflow_datasets as tfds train,test = tfds.load('kmnist', split=['train', 'test']) Using PyTorch import torch import torchvision from torchvision import datasets train = datasets.KMNIST('',train = True,download= True) test = datasets.KMNIST('',train = False,download= True) QMNIST It was developed by Facebook AI Research. The original MNIST consisted of only 10000 images for the test dataset, which was not enough; QMNIST was built to provide more data. 50000 more MNIST-like data were generated. This was made from NIST Special Database 19 keeping the pre-processing as close enough as possible to MNIST using Hungarian algorithm. After several iterations and improvements, 50000 additional digits were generated. Code Snippet: Using PyTorch import torch import torchvision from torchvision import datasets train = datasets.QMNIST('',train= True, download= True) test = datasets.QMNIST('',train= False, download= True) 3D MNIST 3D version of the original MNIST images. There are 5000 training, 1000 validation and 1000 testing point clouds included stored in an HDF5 file format. This was introduced to get started with 3D computer vision problems such as 3D shape recognition.To generate 3D MNIST you can refer to this notebook.","excerpt":"In today’s article, we’ll be talking about the very basic and primarily the most curated datasets used for deep learning in computer vision.To show the performance of these neural networks some basic preprocessed datasets were built, namely the MNIST and its variants such as KMNIST, QKMNIST, EMNIST, binarized MNIST and 3D MNIST.","categories":["AI Trends"],"tags":["Computer Vision","Data Scientist","Deep Learning","MNIST","python machine learning"],"author_name":"Jayita Bhattacharyya","publish_date":"2020-11-10T17:00:42","publication_year":"2020","word_count":981,"keywords":["NumPy","Keras","AI","neural network","MNIST","TensorFlow","PyTorch","computer vision","Ray","deep learning","Computer Vision","Deep Learning","Data Scientist","R","python machine learning"],"extracted_tech_keywords":["AI","deep learning","neural network","computer vision","Ray","TensorFlow","PyTorch","Keras","NumPy","R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/mnist\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10004968,"title":"Future Of Healthcare From The Lens Of Deep Learning &#038; 3D-Printed Organoids","content":"Organoids 3D printing has quickly become one of the leading segments of the 3D printing industry in terms of innovation. Until recently, the market was primarily focused on North America, however many companies, laboratories, and universities around the world are exploring this field as well. Thanks to 3D printing techniques, cells and biomaterials can be combined and deposited layer by layer to create biomedical developments that have the same properties as living tissues. During this process, various bio-links can be used to create these tissue-like structures, which have applications in the fields of medical and tissue engineering. Of course, it is more than knowing that the goal of all these developments is to successfully bioprint a fully functional human organ. The beginnings of bioprinting The first development of bioprinting dates back to 1988 when Dr. Robert J. Klebe, from the University of Texas, presented his Cytoscribing process, a method of micro-positioning cells to create synthetic tissues in 2 or 3D using an inkjet printer. classical. As a result of this research, Professor Anthony Atala of Wake Forest University created the first organ in 2002 thanks to bioprinting, a small-scale kidney. In 2010, the first laboratory specializing in 3D printing was born: Organovo, which quickly began working with the developers of Invetech to create one of the first bioprinters on the market, the NovoGen MMX. Organovo has positioned itself as one of the leaders in the industry and continues to work on bone tissue development after successfully grafting liver tissue. Types of 3D Body Printing Inkjet Bioprinting This technology is based on the common inkjet printing process. Currently, 3D printers with FDM technology are modified to achieve the same process from a biological perspective. It consists of a process in which layers of biotinks (also called biomaterials) are deposited on a hydrogel substrate or culture plates. This technology can be classified into two main methods: thermal and piezoelectric. Thermal technology uses a heating system that creates air bubbles, collapses and provides pressure to expel biotin droplets. In contrast, piezoelectric technology, does not use heat to create the necessary pressure, it uses the electrical charge that accumulates in a solid material, in this case, a polycrystalline piezoelectric ceramic that is in each print nozzle. However, this latest technology can cause damage to the cell membrane if used too frequently. Bioprinting by extrusion It is based on the extrusion of biomaterials to create 3D patterns and cellular constructions. Biotinks used for printing are usually solutions that are extruded by coordinating the movement of a pressure piston or a microneedle on a stationary substrate. After layer by layer application the 3D patterns are completed, and we will have a construction. Advantages of this technology include room temperature processing, direct cell incorporation, and homogeneous cell distribution. Some of the most popular bioprinters on the market use this technique, which is considered an evolution of inkjet, such as EnvisionTec’s Bioplotter or Allevi’s 3D bioprinter. Laser-assisted 3D bioprinting It uses a laser as an energy source to deposit biomaterials in a receptor (substance). The technique consists of three parts: a laser source, a tape coated with biological materials that are deposited on a film and a receiver. The laser beams irradiate the tape, causing the liquid biological materials to evaporate and reach the receptor in droplets, which contain a biopolymer that maintains cell adhesion and helps cells begin to grow. Compared to other technologies, laser-assisted bioprinting has unique advantages, including a nozzle-free, contact-free process, high-resolution, high-activity cell printing, bio-ink droplet control, and precise printing characteristics. Stereolithography SLA technology consists of the solidification of a photopolymer through lighting, and has the highest precision of additive manufacturing technologies. This can be applied in bioprinting by printing with light sensitive hydrogels. This technology is still under development, since in addition to its advantages, they are related to numerous restrictions, such as the lack of biocompatible and biodegradable polymers, harmful effects and the impossibility of eliminating the support structure. Acoustic wave bioprinting This method was developed by Carnegie Mellon University, Pennsylvania State University and MIT, uses something called acoustic tweezers, a microfluidic device where individual cells or particles can be manipulated, and the use of surface acoustic waves. The waves would meet along each of the three axes. At these meeting points, the waves will form a three-dimensional capture node. Individual cells or complete sets of cells are collected so that 2D and later 3D patterns can be created. This technique offers performance in terms of precise movements in a non-invasive way. The SWIFT technique Researchers at Harvard’s Wyss Institute for Biological Engineering developed a new bioprinting technique called SWIFT (Sacrificial Writing Into Functional Tissue), since its name suggests that this technique allows the bioprinting of blood vessels in living tissues. In other words, it 3D-prints vascular channels in living arrays composed of stem cell-derived organ building blocks (OBB). What is the future of 3D printing? The future of 3D lies in the realm of deep learning and quantum computing. The use of powerful deep learning algorithms for information retrieval and the creation of an organoid blueprint will ensure an end-to-end platform with an unprecedented and accelerated advancement of tissue fabrication reassuring the adaptability and compatibility between organ and host. The following elements would be involved: Rapid prototyping techniques: collagen, and gelatinCNN: Convolutional Neural NetworksEvolutionary algorithm: used for planning models based upon cellular processes a.k.a. simulationGenetic algorithm: for DNA\/RNA sequence assembly, and further development of mutation & adaptability to the host.Pluripotent stem cells And how will quantum computing help? Examining extremely large amounts of data from medical files in 1\/20 of the time a regular ML algorithm wouldBuilding more accurate models and cellular correlationWork on multiple parameters simultaneously such as mutation, adaptability, regeneration, etcMerging the analytical power of Neural Networks with the speed of Quantum. The healthcare sector is evolving rapidly towards a bright, innovative and accurate future.","excerpt":"Organoids 3D printing has quickly become one of the leading segments of the 3D printing industry in terms of innovation. Until recently, the market was primarily focused on North America, however many companies, laboratories, and universities around the world are exploring this field as well. Thanks to 3D printing techniques, cells and biomaterials can be […]","categories":["AI Features"],"tags":["latest machine learning innovation","latest technology in machine learning"],"author_name":"Dr. Raul V. Rodriguez","publish_date":"2020-08-18T12:01:00","publication_year":"2020","word_count":977,"keywords":["Go","API","AI","neural network","latest machine learning innovation","ML","BERT","Ray","deep learning","latest technology in machine learning","GAN","R"],"extracted_tech_keywords":["AI","ML","deep learning","neural network","Ray","R","Go","API","BERT","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/future-of-healthcare-from-the-lens-of-deep-learning-3d-printed-organoids\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10063408,"title":"How Certa enables automation to improve third party onboarding","content":"With the massive growth of the data science industry, the process of vendor and partner onboarding has become increasingly time-consuming and opaque. Certa, a no-code automation platform for procurement and compliance, helps enterprises reduce onboarding time, increase transparency and ensure companies meet the security requisites. In an interview with Analytics India Magazine, Certa’s Founder and CEO, Jagmeet Lamba, spoke about the company’s tech stack, offerings and plans. AIM: What problem does Certa solve? Companies are increasingly dependent on their third parties. In fact, 70% of a company’s revenue typically goes to its third parties. To onboard each third party, companies need to exchange data (such as taxing and banking information), mitigate risks (e.g. reputation, security, privacy), sign contracts, and monitor performance. Moreover, now organisations want to work with third parties whose values align with their own. Unfortunately, these processes lead to a ton of back and forth across various functions within a company, which makes it a slow, opaque, and burdensome ordeal. Certa’s software enables companies to onboard and manage risk, compliance, and ongoing collaboration across the entire lifecycle of their third parties. Companies use Certa’s no-code workflow and integration platform to automate their business rules and workflows via drag-and-drop. In addition, Certa’s plug and play modules and 100+ integrations with data sources get third parties up and running quickly and painlessly. AIM: Tell us about the cross-company solutions provided by Certa in India. All global companies now have suppliers and partners in India. Moreover, Certa is used by one of India’s largest ride-sharing firms and one of India’s largest eCommerce firms. Using Certa, Indian firms can get validated GST information and automate the entire KYC\/AML work. AIM: What is the tech stack behind Certa’s no-code platform? Certa uses Python for the backend, TypeScript for the frontend, Terraform for our infrastructure code, and we’re hosted on Amazon Web Services. AIM: How do Certa’s automation solutions differ between countries and sectors? Certa can automate tax and banking validation in almost every country globally. Plus, Certa is completely multilingual and currently live in 17+ languages. Moreover, Certa has an agile business rules engine that allows processes to be localised by country, region, and sector. AIM: What is the importance of leveraging data to support supplier onboarding? Data sources enable us to do highly specialised validations (fraud, tax\/banking, sanctions checks, bankruptcy risk, resilience, etc.) by utilising data. Additionally, the information provided by suppliers can also be reconciled by the information provided from data sources\/integrations so that the data is cross-checked. Leveraging data allows us to evaluate the risk of onboarding suppliers and continuously monitor them after they’ve been onboarded so that companies are aware of any risks that could potentially arise. AIM: How do you see the company collaboration leading the data science industry? We enable data science firms to be easily and quickly onboarded at leading firms globally. Moreover, Certa is developing proprietary AI to enable companies to optimise their processes. AIM: You have just received Series A; what is Certa’s expansion roadmap now? Proceeds will be used to enhance our sales and marketing teams as well as to conduct opportunistic acquisitions of niche targets.","excerpt":"Certa uses Python for the backend, TypeScript for the frontend, Terraform for our infrastructure code, and we’re hosted on Amazon Web Services.","categories":["AI Features"],"tags":[],"author_name":"Avi Gopani","publish_date":"2022-03-24T11:00:04","publication_year":"2022","word_count":523,"keywords":["data science","Go","AI","ML","TypeScript","RAG","Python","Aim","analytics","R"],"extracted_tech_keywords":["AI","ML","data science","analytics","Aim","RAG","Python","R","TypeScript","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-automation-is-helping-improve-third-party-onboarding-with-certa\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":29202,"title":"Are Artificial Intelligence And Marketing Best Friends Yet?","content":"AI is now the new mystical Jedi – the panacea for most ills in marketing. Many marketers remain hopeful about its ability to solve few age-old marketing dilemmas. Yet others view it sceptically as new age marketing voodoo. What perhaps might be a different approach is evaluating what are the biggest challenges that marketers face today? Are any of these problems solvable at scale through AI? Few of the problems that have plagued marketers─those who that don’t have gargantuan budgets and resources to solve these perpetual problems are outlined below. #1: Knowing the customer DNA: 360-view of the customer #2: Understanding intent in search: The evolution of recommendation engines #3: Incorrect forecasting: How AI can help improve and align marketing forecasting models #4: Voice Commerce: How can AI help? #5: Privacy laws & data protection post GDPR #1: Knowing the customer DNA: 360-view of the customer If there ever was a holy grail for marketers, it has been this, always – the 360-view of the customer. Mystical, ever elusive but an important marketing artefact nonetheless. A lead views a Facebook ad, clicks on a Google Ad, calls up the customer service and buys in store, leaving marketers with flummoxed with attribution analysis problems. Large enterprises may have resources at their disposal that can help address this problem. The challenge is addressing this problem at scale so that even smaller businesses can derive engagement value. Customer data seems to emerge from everywhere: submitted data (CRM forms, opt-in communication), transaction data (purchase records), customer service data (service tickets) behavioural data (page views, clicks, likes), apart from an avalanche of email data, website chats, social media and other online chatter1,2. Creating actionable insights from multichannel data can be tedious and complex. So, can AI help in presenting a complete 360 view of the customer? Apparently, yes. Customers expect a consistent and cohesive experience, want lesser (perhaps no) ads, more personalized experiences with the least intrusion. AI can offer the ability to identify not only the 360-degree view of the customer but also a systemic migration of low –margin customers to higher-margin customers, advanced purchase propensity models, and much more. So, solving the Single Customer View (SCV) challenge through AI tech at scale remains one of the fundamental tasks of marketing. #2: Understanding intent in search: The evolution of recommendation engines Customers vigorously and vehemently defend their right to privacy – and rightly so. Location, language preferences etc are guarded zealously to prevent anyone from identifying our online footprint. In this data-sparse, recommendation engines have their task cut out – even more so stringently than before. Recommendation engines act as a discovery aide – whether it’s people you may know, videos you may like or frequently bought together, recommendation engines are your personalized digital assistants. Of course, the maxim here is that get the data right, and the entire customer experience can be tailor-made by using data science. The challenge is to deliver more personalized and relevant recommendations with data being the limiting factor. As lesser and lesser data is available, the associations required to deliver recommendations would be lesser and lesser linear. Here, behavioural data can come to the rescue as intelligent recommendation engines improve with every interaction, learning more about what is important to each individual visitor, allowing it to make increasingly more accurate suggestions that are more likely to appeal. This is the problem that marketers need AI to solve: how to deliver a more personalised relevant experience in a data-scarce environment. #3: Forecasting One of the metrics of success for any marketer is correct long and short-range forecasting. Whether it is revenue, costs, or behavioural customer data, forecasting is an integral component of a marketers’ lifecycle. Forecasting still derives some value from human judgement (consistently optimistic vs consistently pessimistic projections), allowing for discrepancies that can wreck any forecasting model3. Current forecasting applications are limited in their scope and ability to integrate multi-factor inputs and offer limited options for course corrections. The ask that marketers might have from AI is how the variance between the actual and estimated forecasts can be narrowed without human intervention? To eliminate any inefficiencies in forecasting due to subjective human assessments and assumptions, can AI forecast methodologies incorporate market dynamics, fluctuating consumer sentiment and other market variables? #4: Voice commerce comScore forecasts estimate that $1.2 trillion searches will be done by voice by the year 2020, representing around 50% of all searches. By the year 2022, it is estimated that $40 billion in commerce activity will be done via voice4. Conversational AI has been around for some time, and as the adoption of voice commerce increases, marketers will scramble to capitalize on this new point of purchase. The inherent convenience of voice commerce for routine, mundane shopping experiences, such as reordering, is expected to enhance societal adoption5,6. Marketing professionals need to focus increasingly on voice as a strategic initiative and appreciate that voice searches are inherently different than voice searches. The thing to watch out for is how AI can leverage natural language processing to analyse conversation tonality, contextual conversational cues, user habits and preferences and solve the problem that has been plaguing marketers for quite some time now – can we correctly predict the true intent of search? #5: Privacy laws & data protection [Post GDPR] This year was a watershed event for all marketers. Adopting GDPR symbolized a mindset change to marketers used to treating customers’ data with impunity. This EU regulation was enforced in May 2018. The regulation built on requirements for data privacy and security, and includes several new provisions to bolster the rights of data subjects and add harsher penalties for violations7. However, AI interventions are expected to run into strong headwinds post-GDPR. Big data and data protection are irreconcilable. And enforceability of AI in marketing is dependent and driven by data. It would be interesting to see how AI delivers its promise amongst increasing clamouring concerns for privacy and data protection. Conclusion In conclusion, if AI is able to render viable solutions to few of the biggest marketing challenges listed, this may just be the beginning of a long and sustainable relationship. Its success would also depend on how wholeheartedly marketers adapt and embrace artificial intelligence, intelligently, keeping in mind the tectonic changes in privacy and data concerns. References Underwood, Architect your customer 360 degree data lake for today and tomorrow, https:\/\/goo.gl\/aCQJca Pandey (2018), A 360-degree View of the Customer – Finding Marketing Zen, https:\/\/goo.gl\/6MgniF Guszcza & Maddirala, Minds and machines: The art of forecasting in the age of artificial intelligence, Deloitte Issue Review 19, https:\/\/goo.gl\/TWSD3B Treseder (2018) How Marketing Through Voice Technology Will Put Savvy Brands Ahead of the Curve, https:\/\/goo.gl\/kWG4PJ Clarke (2018) How voice-assisted commerce is speaking up in retail, https:\/\/goo.gl\/fTphSG The Mid-Year State Of The Voice Commerce Market, https:\/\/goo.gl\/hofLbz Meyer (2018) AI Has a Big Privacy Problem and Europe’s New Data Protection Law Is About to Expose It, Fortune, https:\/\/goo.gl\/gHmML2","excerpt":"AI is now the new mystical Jedi – the panacea for most ills in marketing. Many marketers remain hopeful about its ability to solve few age-old marketing dilemmas. Yet others view it sceptically as new age marketing voodoo. What perhaps might be a different approach is evaluating what are the biggest challenges that marketers face […]","categories":["AI Features"],"tags":["assisted intelligence","best book to learn digital marketing","Data Privacy","GDPR","recommendation engines"],"author_name":"Sumedha Chatterjee","publish_date":"2018-10-12T12:51:32","publication_year":"2018","word_count":1153,"keywords":["data science","Go","artificial intelligence","AWS","recommendation engines","AI","ML","assisted intelligence","Git","RAG","Data Privacy","GDPR","best book to learn digital marketing","Tecton","R"],"extracted_tech_keywords":["AI","artificial intelligence","ML","data science","Tecton","RAG","AWS","R","Go","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/are-artificial-intelligence-and-marketing-best-friends-yet\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10152627,"title":"IIIT Hyderabad’s Subtl.ai V2 Outperforms OpenAI’s RAG Solutions","content":"Subtl.ai, a startup born out of the Indian Institute of Information Technology Hyderabad, has unveiled ‘Subtl V2’, the latest iteration of its AI information retrieval system for enterprises. Vishnu Ramesh, founder of Subtl.ai, took to LinkedIn to announce the launch. He said that their auto-tuning pipeline built by engineers and researchers at IIIT Hyderabad outperforms RAG built on OpenAI and open source embeddings by 15-20%. Ramesh also announced that their ‘co-create embeddings program’ is now available, which allows developers to create proprietary RAG embedding models based on private documents. Furthermore, Subtl.ai can now also generate an Excel-like grid view for data and said that document size and page limits have been removed from the app. Founded in 2020, Subtl.ai has built a local solution to derive insights and analytics from private and sensitive data in enterprise applications. In a blog post earlier this year, the startup revealed that one of India’s biggest banks, State Bank of India, successfully implemented Subtl.ai. The results demonstrated 92% accuracy in information retrieval, and 56,570 minutes were saved, which was equivalent to 5 lakh rupees. In an interview with AIM earlier this year, Ramesh revealed that the startup’s ambition was to reduce the dependence on companies like OpenAI. “Microsoft has a lot of security features on top of its OpenAI offerings. But at the end of the day, it’s still a common endpoint, making it difficult for companies in the banking or sensitive data sector to rely on it,” he said. Ramesh also called Subt.ai a ‘private Perplexity’ built on light models for the enterprise. The tool is built on the Llama 3 8B variant and can access all the information in catalogues and databases that private companies have trained on. Moreover, it can also reply to simple questions citing the exact source along with a paragraph. Subtl.ai isn’t free from competition, and companies like Google, Anthropic, and OpenAI offer retrieval capabilities through its API. On the other hand, Perplexity also released Internal Knowledge Search, a pro version for enterprises. However, enterprises are sceptical about their private data going online, especially given the Amazon Q fiasco.","excerpt":"Subtl.ai’s V2‘Outperforms RAG built on OpenAI and open source embeddings by 15 – 20%.’","categories":["AI News"],"tags":["AI (Artificial Intelligence)","RAG"],"author_name":"Supreeth Koundinya","publish_date":"2024-12-30T17:04:40","publication_year":"2024","word_count":351,"keywords":["Anthropic","Go","API","OpenAI","AI","RAG","Aim","analytics","R","AI (Artificial Intelligence)","startup"],"extracted_tech_keywords":["AI","analytics","OpenAI","Anthropic","Aim","RAG","R","Go","API","startup"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/iiit-hyderabads-subtl-ai-v2-outperforms-openais-rag-solutions\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10168405,"title":"50% of People use ChatGPT for Unproductive Work, says CRED Founder Kunal Shah","content":"Speaking at the Bharat Inclusion Summit 2025, CRED founder and CEO Kunal Shah remarked that 50 to 60% of ChatGPT users engage with the platform for non-productive purposes such as companionship, entertainment, and astrology. He urged individuals to align their time with their per-hour income and focus on efficiency to maximise productivity. “People should calculate their per-hour income to understand the value of their time,” Shah quipped. His remarks align with his earlier warnings about AI’s transformative impact. In 2023, Shah predicted that 90% of jobs could be disrupted by AI within a decade, stressing the need for upskilling to stay relevant. Interestingly, Shah’s point is reinforced by new data from Harvard Business Review, which reveals that people are increasingly using AI chatbots for personal and human reasons rather than for work. In 2025, the top use case of generative AI was for ‘therapy and companionship’, followed by ‘organising life’ and ‘finding purpose’. This is a shift from the top use cases in 2024 being all about generating ideas, editing text, and web search. Shah also highlighted the rapid pace of technological evolution, noting that advancements in AI, from browsers to everyday tools, are accelerating. “It feels like changes are happening every month, and soon, they’ll occur daily,” he said. He encouraged the audience to adopt a mindset focused on practical applications of AI to solve real-world problems. Moreover, he also stressed the importance of financial literacy in the country and believes that kids aged 8–20 are the real tech drivers at home. According to him, they will soon also be knowledge and procurement leaders with the help of AI. “There is a chief technology officer in all of your houses between the ages of 8 to 20. They’ll soon become the chief procurement and knowledge officers of your homes, especially with AI helping them.” This isn’t the first time Shah has shared such concerns. In a fireside chat with Razorpay CEO Harshil Mathur at FTX’25, he voiced frustration over India’s tech narrative. “We didn’t question why India didn’t invent WhatsApp or LED bulbs. We just adopted them and became the best at using them. That should be our approach with AI,” he argued. Shah pointed to UPI as a model worth emulating, a made-in-India solution now being copied globally. “We built UPI, we are the beneficiaries of that. The world should be copying that,” he said. He believes India’s focus shouldn’t be on competing with the US or China in building foundational LLMs, but rather on becoming the largest and smartest users of such tools.","excerpt":"“People should calculate their per-hour income to understand the value of their time,” Shah quipped.","categories":["AI News"],"tags":["ChatGPT"],"author_name":"Siddharth Jindal","publish_date":"2025-04-22T14:36:15","publication_year":"2025","word_count":426,"keywords":["ChatGPT","API","AI","chatbots","RAG","GPT","generative AI","Rust","GAN","R"],"extracted_tech_keywords":["AI","generative AI","ChatGPT","RAG","chatbots","R","Rust","API","GPT","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/50-of-people-use-chatgpt-for-unproductive-work-says-cred-founder-kunal-shah\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":34984,"title":"5 Important Techniques To Process Imbalanced Data In Machine Learning","content":"Imbalance data distribution is an important part of machine learning workflow.  An imbalanced dataset means instances of one of the two classes is higher than the other, in another way, the number of observations is not the same for all the classes in a classification dataset. This problem is faced not only in the binary class data but also in the multi-class data. In this article, we list some important techniques that will help you to deal with your imbalanced data. 1| Oversampling This technique is used to modify the unequal data classes to create balanced datasets. When the quantity of data is insufficient, the oversampling method tries to balance by incrementing the size of rare samples. A primary technique used in oversampling is SMOTE (Synthetic Minority Over-sampling TEchnique). In this technique, the minority class is over-sampled by producing synthetic examples rather than by over-sampling with replacement and for each minority class observation, it calculates the k nearest neighbours (k-NN). But this technique is limited to an assumption that local space between any two positive instances belongs to the minority class, which may not always true in the case when the training data is not linearly separable. Depending upon the  amount of oversampling required, neighbours from k-NN are randomly chosen. Advantages No loss of information Mitigate overfitting caused by oversampling. To take a deep dive into the SMOTE technique. Click here. 2| Undersampling Unlike oversampling, this technique balances the imbalance dataset by reducing the size of the class which is in abundance. There are various methods for classification problems such as cluster centroids and  Tomek links. The cluster centroid methods replace the cluster of samples by the cluster centroid of a K-means algorithm and the Tomek link method removes unwanted overlap between classes until all minimally distanced nearest neighbours are of the same class. Advantages Run-time can be improved by decreasing the amount of training dataset. Helps in solving the memory problems To learn more about undersampling, click here. 3| Cost-Sensitive Learning Technique The Cost-Sensitive Learning (CSL) takes the misclassification costs into consideration by minimising the total cost. The goal of this technique is mainly to pursue a high accuracy of classifying examples into a set of known classes. It is playing as one of the important roles in the machine learning algorithms including the real-world data mining applications. In this technique, the costs of false positive(FP), false negative (FN), true positive (TP), and true negative (TN) can be represented in a cost matrix as shown below where C(i,j) represents the  misclassification cost of classifying an instance and also “i” the predicted class and “j” is the actual class. Here is an example of cost matrix for binary classification. To deep dive into CSL technique, click here. Advantages This technique avoids pre-selection of parameters and auto-adjust the decision hyperplane. 4| Ensemble Learning Techniques The ensemble-based method is another technique which is used to deal with imbalanced data sets, and the ensemble technique is combined the result or performance of several classifiers to improve the performance of single classifier. This method modifies the generalisation ability of individual classifiers by assembling various classifiers. It mainly combines the outputs of multiple base learners. There are various approaches in ensemble learning such as Bagging, Boosting, etc. Bagging or Bootstrap Aggregating tries to implement similar learners on a smaller dataset and then takes a mean of all the predictions. The Boosting (Adaboost) is an iterative technique that rectifies the weight of an observation depending on the last classification. This method decreases the bias error and builds strong predictive models. Advantages This is a more stable model The prediction is better To learn more about this technique, click here. 5| Combined Class Methods In this type of method, various methods are fused together to get a better result to handle imbalance data. For instance, like SMOTE can be fused with other methods like MSMOTE (Modified SMOTE), SMOTEENN (SMOTE with Edited Nearest Neighbours), SMOTE-TL, SMOTE-EL, etc. to eliminate noise in the imbalanced data sets. However, the MSMOTE is the modified version of SMOTE which classifies the samples of minority classes into three groups such as security samples, latent nose samples, and border samples. Advantages No loss of useful information Good generalisation","excerpt":"Imbalance data distribution is an important part of machine learning workflow.  An imbalanced dataset means instances of one of the two classes is higher than the other, in another way, the number of observations is not the same for all the classes in a classification dataset. This problem is faced not only in the binary […]","categories":["AI Trends"],"tags":["Imbalance data"],"author_name":"Ambika Choudhury","publish_date":"2019-02-15T06:27:30","publication_year":"2019","word_count":704,"keywords":["Go","TPU","machine learning","programming_languages:R","AI","ML","programming_languages:Go","Imbalance data","R"],"extracted_tech_keywords":["AI","machine learning","ML","TPU","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/5-important-techniques-to-process-imbalanced-data-in-machine-learning\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":32361,"title":"IIT Kharagpur To Open Second Centre Of Excellence For AI","content":"To expand their field of study in artificial intelligence, IIT Kharagpur is closely working with the Telangana government to set up an off-campus Centre of Excellence (CoE) for artificial intelligence in Hyderabad by March 2019. The institute also plans to set up a 2,000 sq ft research park in Kukatpallya where it will primarily focus on AI, cybersecurity, defence, aerospace, advanced manufacturing and robotics. With the CoE and research park, IIT-KGP wants to lay a strong foundation by training professionals in AI and hopes to take up more high-tech industry and government projects in the future. “The research park will focus on areas like AI, cybersecurity, defence and aerospace, in which we want to train professionals and carry out research and industry projects. We are in talks with Telangana government for land for the park,” said Utkarsh Prasad, IIT Kharagpur’s programme manager in Hyderabad. This is the second such initiative by IIT-KGP to develop their collaborative best in the field of AI within a year’s time. In April this year, the institute collaborated with Capillary Technologies Limited to unveil their Centre for Artificial Intelligence in the campus with a curriculum focus on deep learning from industry uses cases. This is despite the fact that the institute already boasts of AI specialist in areas such as financial analytics, industrial automation, digital healthcare, intelligent transportation system among others. Speaking on the matter to a leading daily, Jayesh Ranjan, Telangana IT principal secretary said, “IIT-KGP is in one corner of the country so they don’t get too many guest faculty and industry connect so easily. So, they were scouting around for establishing their centre in a city that is much for accessible, central and happening from the tech point of view.” He also added that the state government and IIT-KGP will soon ink a MoU and added that his government has already identified the space for the future centre.","excerpt":"To expand their field of study in artificial intelligence, IIT Kharagpur is closely working with the Telangana government to set up an off-campus Centre of Excellence (CoE) for artificial intelligence in Hyderabad by March 2019. The institute also plans to set up a 2,000 sq ft research park in Kukatpallya where it will primarily focus […]","categories":["AI News"],"tags":["AI centre of excellence","CoE","telangana"],"author_name":"Akshaya Asokan","publish_date":"2018-12-28T07:13:21","publication_year":"2018","word_count":316,"keywords":["Go","API","artificial intelligence","CoE","AI","telangana","AI centre of excellence","Git","RAG","deep learning","analytics","GAN","R"],"extracted_tech_keywords":["AI","artificial intelligence","deep learning","analytics","RAG","R","Go","Git","API","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/iit-kharagpur-to-open-second-centre-of-excellence-for-ai\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":36269,"title":"Meet &#038; Learn From Leading Data Scientists At AIMinds Gurugram On 5 April 2019","content":"AIMinds, a hub for AI intellectuals, is coming to Gurugram with its fifth edition in collaboration with AnalytixLabs, Gurugram. A venture by Analytics India Magazine, AIMinds is a monthly meetup which puts the common public and Data Science enthusiasts in touch with artificial intelligence practitioners and researchers from all over the country. The registration for the meetup is free. Click here and RSVP to register. AIMinds is a platform where established and aspiring minds in AI, data science and analytics can meet to have great conversations and share knowledge under one roof. People of all skill levels are welcome. Agenda for the meetup: How AI Is Transforming Job opportunities? Date: 5 April 2019 Time: 2:00 pm to 5:30 pm Venue: The Woods, Netaji Subhash Marg, Near Vista Vilas, Block H, Block D, Greenwood City, Sector 46, Gurugram, Haryana 122001 The speakers for this edition are: Sumeet Bansal – CEO & Co-Founder of Analytixlabs (Linkedin: https:\/\/www.linkedin.com\/in\/sumeetbansal\/) Yashu Kant Gupta – Head – Insights and Data Science at Times Internet (Linkedin: https:\/\/www.linkedin.com\/in\/yashu-kant-gupta-a06404b\/) Anshul Gupta – CEO and Co-founder, MikeLegal (Linkedin: https:\/\/www.linkedin.com\/in\/anshul-gupta-7b09b076\/) Click here to register and attend the meetup. We will be adding awesome speakers to the lineup. This is an event for established and aspiring minds in AI, Data Science and Analytics to meet, have great conversations and learn from Industry thought leaders. Freshers with a tech background and who are enthusiastic about emerging technologies are also welcome. As this meetup is primarily intended at knowledge sharing, it will be good if you have loads of questions to make this a very interactive meet-up session. Each session will be timed at about 35-40 minutes, followed by a Q&A round. So, do send in your RSVP soon and be part of this exciting meet-up along with your friends and colleagues who are also interested in analytics or data science and want to know how to transition to this space.","excerpt":"AIMinds, a hub for AI intellectuals, is coming to Gurugram with its fifth edition in collaboration with AnalytixLabs, Gurugram. A venture by Analytics India Magazine, AIMinds is a monthly meetup which puts the common public and Data Science enthusiasts in touch with artificial intelligence practitioners and researchers from all over the country. The registration for […]","categories":["Deep Tech"],"tags":["AI meetup","data science meetup","meetup"],"author_name":"Abhijeet Katte","publish_date":"2019-03-13T11:28:04","publication_year":"2019","word_count":317,"keywords":["AI meetup","data science","meetup","Go","artificial intelligence","programming_languages:R","AI","programming_languages:Go","Aim","data science meetup","analytics","R"],"extracted_tech_keywords":["AI","artificial intelligence","data science","analytics","Aim","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/meet-leading-data-scientists-at-gurugram-meetup-on-5-april-2019\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":16839,"title":"SMEs are using big data to drive growth with increased adoption of big data tech","content":"Big data isn’t just for big enterprises anymore. Small and medium sized enterprises (SMEs) have also become active adopters of ICT. In fact, over the last couple of years, small and mid-size Indian companies have seen more big data deployments than the big competitors. According to our study, the Big Data industry in India is expected to almost double by 2020. There is an increased adoption of SMAC amongst small and medium business verticals, owing to the increased volume of data and customer interactions. Our report pegs that the big data industry is estimated to be $2.03 billion annually in revenues and will grow at a healthy rate of 23.8% CAGR. The data boom in India isn’t just limited to big enterprises, the growth of big data startups\/technology vendors is helping SMEs in scaling up infrastructure capabilities and driving insights from data. Analytics India Industry Study  \\indicates that the India domestic market serves as a significant opportunity, with almost 4% of analytics revenues coming for Indian firms. Also, the increased availability of accessible, cheap data centres delivered by cloud vendors, has brought down the costs of upfront investment for small businesses, thereby reducing the market entry barrier. Now, it is the question of choosing the right analytics vendors that fits the bill for small businesses.  As more and more vendors offer competitive data-driven capabilities, what’s crucial for startups is getting started with the right infrastructure and build for scaling. Big Data Impact in India And behind the successful adoption of big data analytics (BDA) in SMBs is the exponential growth of market. According to our research, out of the annual inflow to analytics industry – almost 12% can be attributed to advanced analytics\/ predictive modeling and Data science. And a sizeable 24% can be attributed to big data. The rise of BDA and robust analytics tools is helping small businesses accomplish their goals in a short span of time, see a higher ROI and ensure continued business success. Big data means big business for SMBs Even though SMBs are an early stage of big data adoption, with most small enterprises dabbling in the exploratory stage, businesses have started showing a greater interest in getting analytics platform off the ground to reap the desired outcome. Most small and mid-size enterprises are now viewing big data more as a strategic initiative and a central IT function. Here are some of the recent big data deployments that made headlines: a) Fashion doyenne Ritu Kumar’s The Label team up with Manthan for advanced retail analytics: Indian fashion industry is usually regarded as a laggard in adopting technology solutions, but in a first, Ritu Kuamr’s The Label will be using Bangalore-based analytics vendor Manthan’s ready-to-use solution that provides advanced algorithms out of the box, eliminating the need for business users to rely on a team of data scientists and IT for detailed analysis. Manthan’s Advanced Retail Analytics will enable the fashion business to accurately forecast demand for new SKUs, identify opportunity areas in their existing assortment range and optimize their in-season allocation and delivery plans. The advanced analytics solution will provide granular insights that help in introducing the right products based on customer preferences, optimal in-season allocation across channels based on key attributes and insights to refine their delivery strategy. b) SAP is betting big on SME sector in India: With the recent GST move, the world’s leading enterprise solution provider is making a huge play in the SME marketplace. Krishnan Chatterjee, head (marketing), Indian sub-continent of SAP India, divulged that SME units make up 80% of the client base in India as it rolled out the GST software built in India for SMEs. Over the years, SAP has emerged as a dominant player in the SME ecosystem by providing the right big data solutions to improve efficiency, cut costs and boost sales. Barriers to adoption of big data analytics Ever wondered why so many small and mid-size businesses fail in a successful adoption of analytics. According to research, some of the key barriers limiting big data adoption is a lack of successful use cases, a lack of availability of packaged solutions, a limited understanding of technology stack and most importantly a lack of data-centric culture. Other factors include an acute shortage of in-house talent and the cost of implementing solutions. The other key challenge in big data tech is disparate data sources, high costs and infrastructure and concerns around RoI. Traditionally, most small and mid-size enterprises rely on intuition rather than data to drive decisions around pricing strategy or driving customer acquisition. Here’s a round-up of key limitations to big data adoption: Concerns around the cost and complexity of big data solutions Lack of understanding of technology stack Lack of a data-centric culture is also a key challenge Senior management grapples to justify high investment & RoI Limited number of uses cases available in the Indian SME ecosystem How to win over SMBs in the big data market One of the key challenges for analytics vendors is to develop customizable software for SMEs that is easy to deploy and can be easily configured. Subscription based pricing models can provide flexibility to customers and help avoid vendor lock-in and lower entry barrier risk. To benefit from big data advances, small and mid-size companies should adopt a more modernized approach to IT, pivoting towards a cloud-powered platform to scale effectively. Also, before signing up for a big data analytics solution, businesses must learn how to identify the right data analytics tool. Case in point – self-service analytics tool enables non-tech users to find patterns and drill insights and create beautiful dashboards while ETL tools help extract data at high speed for analysis. Some businesses may only require dashboard tools to present insights to management or skim over what-if scenarios through pie-charts and bar charts. However, one of the key challenges for analytics vendors is to create flexible tools, keeping the end user in mind. Also, the DIY analytics and forecasting solutions should support different types of data. This article first appeared on our Analytics India Industry Study 2017","excerpt":"Big data isn’t just for big enterprises anymore. Small and medium sized enterprises (SMEs) have also become active adopters of ICT. In fact, over the last couple of years, small and mid-size Indian companies have seen more big data deployments than the big competitors.   According to our study, the Big Data industry in India […]","categories":["IT Services"],"tags":[],"author_name":"Ankita Gupta","publish_date":"2017-08-08T05:54:40","publication_year":"2017","word_count":1012,"keywords":["big data","Go","data science","AI","ETL","data-driven","analytics","R","analytics platform","startup"],"extracted_tech_keywords":["AI","data science","analytics","R","Go","big data","ETL","analytics platform","data-driven","startup"],"url":"https:\/\/analyticsindiamag.com\/it-services\/smes-using-big-data-drive-growth-increased-adoption-big-data-tech\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":34183,"title":"Can AI Reinvent The Hearing Aid &#038; Turn It Into A Smart Wearable?","content":"AI-based hearing aids are now replacing the earlier model of Bluetooth Low Energy (BLE)- enabled hearing aids. In this article, we discuss how these hearing aids, backed by AI some use cases of artificial intelligence in hearing devices. What Has Been Done Till Now In general, a hearing device includes a sound classification module to classify environmental sound sensed by a microphone. SoundSense Learn is a feature that uses machine learning and artificial intelligence technique to apply user input for efficient optimisation of hearing aid parameters. This novel approach of machine learning is designed to individualise hearing aid parameters in a qualified manner, sampling a number of possible settings for the adjustment of hearing aid according to the user’s preference with a greater degree of certainty. Dr Chris Heddon of Resonance Medical in a report pointed out that there are specifically three major elements that make up the AI-powered hearing aid: Hearing aids with energy-efficient wireless connectivity, which gives them access to external computing power The ability of hearing care professional to securely program a hearing aid from distance, which gives the user access to the highest level of hearing care at all times- even in real-world environments Mobile phones with sufficient computing power to run AI on device, which support the dynamically responsive intelligent hearing aids and provide the additional benefits of protecting user privacy and reducing the mobile device power consumption associated with cellular connection to a cloud-based server (which would have been needed if the AI was run in the cloud rather than on the user’s mobile device). In another research, conducted by the researchers at Aalborg University, the researcher developed an efficient algorithm by using artificial neural networks that can help the end-users to not only hear but also to take part in conversations in the noisy environment. The applicability of this new algorithm is much advanced and stronger than the previous technology because of its ability to function in unknown environments with unknown voices. Researchers concluded that the key success of this algorithm is its ability to learn from data and construct strong statistical models that are able to represent the complex listening situations. The method used for creating the algorithm is the deep learning technique and the project was constructed with two different tracks. Firstly, they constructed an algorithm to solve the challenges of one-to-one conversations in noisy spaces by amplifying the sound of the speaker while reducing noise significantly without any prior knowledge about the listening situation. Secondly, speech separation was done by an algorithm that separate voices while reducing noises. Recent Use Cases Last month, Microsoft announced that Skype and PowerPoint will include real-time captions and subtitles by early 2019 that will be based on artificial intelligence techniques. The motive behind adding the new feature into these two is to allow the people with hearing impairment for easily go with the audio of Skype calls and presentations in PowerPoint. Another instance is the real-life used case, Livio AI is the world’s first hearing aid that provides both superior sound quality as well as ability to track body and brain health that is featured with integrated sensors and artificial intelligence techniques. Last year, Widex, a Danish hearing aid maker launched the world’s first machine learning hearing aid technology, Evoke where the sound quality evolves in real life. The device gives the ability to the users to employ real-time machine learning that allows the hearing aid to learn from the user’s input and preferences. Evoke’s smartphone AI app, SoundSense Learn is designed to help the users adjust the sound of the hearing aid according to the moment. According to Widex, the app gathers a variety of anonymous data like the most frequent sound presets, how many times the volume is being adjusted, etc.","excerpt":"AI-based hearing aids are now replacing the earlier model of Bluetooth Low Energy (BLE)- enabled hearing aids. In this article, we discuss how these hearing aids, backed by AI some use cases of artificial intelligence in hearing devices. What Has Been Done Till Now In general, a hearing device includes a sound classification module to […]","categories":["AI Features"],"tags":[],"author_name":"Ambika Choudhury","publish_date":"2019-01-26T08:14:03","publication_year":"2019","word_count":629,"keywords":["Go","machine learning","artificial intelligence","programming_languages:R","AI","neural network","programming_languages:Go","deep learning","ViT","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","deep learning","neural network","R","Go","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/can-ai-reinvent-the-hearing-aid-turn-it-into-a-smart-wearable\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":54581,"title":"How This Conversational App Startup Helps Improve The Functional Fluency Of Job Seekers","content":"With workplaces and product deliveries increasingly moving to smartphones, learning is also becoming mobile. As per research, English fluency improves professional income by 34%. With over 200 million white and blue-collar workers and job seekers in the economy, English learning has become an important part of workplace skills requirements by employers. With such a vision, Pune-based startup, Utter uses conversational chats (chatbots and private tutors) as a medium to improve the functional fluency of job seekers. Founded by Ninad Vengurlekar and Amit Bhadbhade in 2015, Utter combines seamless interactions between machines and humans to achieve low cost affordable English practice to learners. Ninad is a Masters in Edu Technology from Harvard Graduate School of Education while Amit is a Masters in Mobile Programming from Lancaster University, UK. Utter uses conversational chatbots combining with live chats to improve the English fluency of aspiring white-collar and blue-collar workers. The Utter App was launched in January 2018 on Google Play and currently has over 2.5 million users across India, Bangladesh, Myanmar, and Kuwait. It is a paid app which is priced between ₹199 to ₹499 with over 60,000 paid users. Utter has raised around $1.3 million through Unitus Ventures and a clutch of angel investors and is currently doing an ARR of over  ₹3 crores a year on an MAU base of 450,000. Flagship Product Utter’s flagship product is an English learning app that was born out of the features the founders integrated while observing the WhatsApp Group interactions. Utter was build by transforming chat bubbles into byte-size learning tools with features like translations, dictionary, and text to speech in order to radically improve fluency. The app also has an option to practice fluency with live tutors who undertake in chat corrections to point out language errors in complete privacy. The platform offers them an unlimited bank of conversations in multiple scenarios of their personal and professional life. Use of AI\/ML and Other Techniques Utter uses the live and automated chat interactions to build conversational patterns of each user. A part of these patterns is used to create personalised learning experiences. The conversational patterns in live chats are used to build automated suggestions to tutors so that they can speed up their responses to learners. Utter also aspires to build automated voice bots through the millions of conversational data interactions it would be automating. According to the founders, byte-sized conversations are the future of learning. Utter is poised to build low-cost affordable learning solutions at scale for the millions of emerging jobs in the internet economy. Currently, the users in the internet economy learn the most through their smartphones and by collecting competency maps of byte size learning conversations and later automating them through ML and AI, one can afford low-cost learning solutions. Tackling The Hiring Phase Coming to the hiring phase, the founders said, “Utter is a small team of 14 people and what we look for in our candidates is the integrity of purpose to build something transformative and world-changing. So apart from technical skills, what we look for is the initiative and passion that they bring to the table. The latter is more valued than the former. Because the former can be learned and fine-tuned, the latter is a part of a candidate’s DNA.” Potential Competitors Talking about competitors, companies like Open English, iTutorGroup, and VIPKid are considered as potential competitors. Future Roadmap For the future roadmap, Utter aspires to move beyond chat as the only mode of learning in the future. The founders aspire to take conversational learning to audio and video platforms like Voice Assistants, internet TV, and Online Radio.","excerpt":"With workplaces and product deliveries increasingly moving to smartphones, learning is also becoming mobile. As per research, English fluency improves professional income by 34%. With over 200 million white and blue-collar workers and job seekers in the economy, English learning has become an important part of workplace skills requirements by employers.  With such a vision, […]","categories":["AI Startups"],"tags":[],"author_name":"Ambika Choudhury","publish_date":"2020-01-24T15:12:57","publication_year":"2020","word_count":602,"keywords":["Go","programming_languages:R","AI","chatbots","ML","programming_languages:Go","R","startup"],"extracted_tech_keywords":["AI","ML","chatbots","R","Go","startup","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-startups\/how-this-conversational-app-startup-helps-improve-the-functional-fluency-of-job-seekers\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10062367,"title":"Apple&#8217;s M1 Ultra is the world’s most powerful chip for a personal computer","content":"Apple recently announced the new M1 Ultra chip, the next breakthrough chip for the Mac, at the Peak Performance Event, 2022. Building on M1 Max’s transformational architecture, M1 Ultra is claimed to be the world’s most powerful chip for a personal computer. Apple says that with M1 Ultra, it did “something truly groundbreaking.” The chip’s architecture interconnects the die of two M1 Max chips to create a system on a chip (SoC) with unprecedented levels of performance and capabilities, explained Johny Srouji, SVP, Hardware Technology at Apple. Apple calls this the “ultra-fusion architecture.” The technology uses a silicon interposer that connects over 10,000 signals and provides 2.5TB\/s of low latency, inter-processor bandwidth between the two dies while using very little power. Srouji asserts this to be more than 4x the bandwidth of the leading multi-chip interconnect technology. Ultra-fusion essentially treats two chips as one in software, with 114 billion transistors- 7x more than M1. “There’s never been anything like it,” the company says. M1 Ultra architecture M1 Ultra’s memory bandwidth is increased to 800GB\/s, more than 10x the latest PC desktop chip. Additionally, it can be configured with up to 128GB of high bandwidth with low-latency unified memory. The chip is powered by a 20-core CPU, 64-core GPU, and 32-core Neural Engine, making its performance suitable for developers and artists in 3D environments. The company welcomed developers from Adobe, Vectorworks, Ableton and CLO who vouched for these capabilities. Industry experts and Apple enthusiasts have taken to Twitter to express their opinions on the M1 Ultra. Well,, I should say, the theme of the event \"Peek Performance\" was perfect for the day with #M1Ultra#AppleEvent @Apple $aapl pic.twitter.com\/ZHFzDjUd9w— Prakash Sangam (@MyTechMusings) March 8, 2022 LEGIT JUST FELL OFF MY CHAIR #AppleEvent #M1Ultra pic.twitter.com\/5t3KckOqnc— iJustine (@ijustine) March 8, 2022 M1 Ultra = 2 M1 Max SOCs. Wow. pic.twitter.com\/icLkztyQ9E— Jon Rettinger (@Jon4Lakers) March 8, 2022","excerpt":"“There’s never been anything like it,” the company says.","categories":["AI News"],"tags":[],"author_name":"Avi Gopani","publish_date":"2022-03-09T01:02:48","publication_year":"2022","word_count":311,"keywords":["programming_languages:R","AI","Git","Aim","R"],"extracted_tech_keywords":["AI","Aim","R","Git","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/apples-m1-ultra-is-the-worlds-most-powerful-chip-for-a-personal-computer\/","complexity_score":4,"technical_depth":5,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10000888,"title":"Everything You Need To Know About Li-Fi &#038; How It Scores Over Wi-Fi","content":"We all know that the speed of light is 299,792,458 m\/s. Now imagine that you are using the internet with a speed of 224 gigabits per second transmitting through the light! This might seem like a dream now, but the time is not so far away when it becomes a reality and allows you to download 10 HD movies within minutes. How Will This Happen? It was all be enabled by a technology called the Light Fidelity or Li-Fi, which is a Visual Light Communication (VLC) System, first coined by Professor Harald Haas from the University of Edinburgh during a TED talk in 2011. It is a high speed bidirectional and full duplex wireless communication technology that works just like a Wi-Fi. How It Works In simple words, Li-Fi converts a beam of light into electrical signals and then these signals are converted back into data, after which the data is transmitted electromagnetically. An LED bulb with signal processing technology is used as a transmission source resulting that the constant electricity supplied to the bulb can be dipped and dimmed, up and down at a very high speed that will be invisible to the human eye. The pulses of the rapid dimming of the bulbs are converted into electrical signals by a receiver dongle and finally, the electrical signals are converted back into a binary data stream that you can use as web or anything that runs on internet enabled devices. Li-Fi v\/s Wi-Fi In Terms Of Communication Protocol Wi-Fi uses radio waves for transmission of data, while Li-Fi uses visible light from LED (Light Emitting Diodes) light bulbs fitted with a special chip The speed of Li-Fi is a hundred times faster than Wi-Fi The spectrum range is higher in Li-Fi because the visible light spectrum has 10,000-time broad spectrum in contrast to radio frequency The cost will be cheaper in Li-Fi as it uses light and needs fewer components than radio technology Li-Fi is more secure than Wi-Fi as light cannot be passed through the walls providing more secure data transfer. Li-Fi Use Cases PureLifi, the company started by Professor Haas in 2012 is performing experiments in the labs for the advancement and modification of Li-Fi. Till now the practical results of lab experiments are found to be around 224GB per second but in real life, it provides a lot slower speed of 1GB per second. The technology is new, and researchers are still working on it. In June 2018, the company offered Li-Fi starter kits to academic researchers. They also opened a channel programme for IT resellers for adding Li-Fi to their portfolio. Later in July, the IEEE confirmed a global standard for Li-Fi with an estimated target date of May 2021. Some of the areas where Li-Fi can find applications are: Li-Fi can be used for underwater communication. Radio signals cannot travel through water, but light can travel through water providing communications underwater. It can be broadly used in the Internet of Things (IoT) due to its extremely high speed. Li-Fi can be used in military camps with specific lights where there is limited access to networks. It can also be used as car-to-car communication by enabling Li-Fi on car front as well as rear lights. Drawbacks In this case, the LED bulbs need to be switched on even in the daylight for data transmission. The range is shorter because light cannot pass through the walls. Li-Fi In India Velmenni, a Delhi based startup in India has been working in the field of Li-Fi since 2012. They created a Li-Fi-based mesh network solution to transfer data across long distances where optical cable infrastructure is not easy to implement. Currently, they are working on pilot projects to utilize bidirectional VLC in diverse industrial conditions.","excerpt":"We all know that the speed of light is 299,792,458 m\/s. Now imagine that you are using the internet with a speed of 224 gigabits per second transmitting through the light! This might seem like a dream now, but the time is not so far away when it becomes a reality and allows you to […]","categories":["AI Features"],"tags":["Interviews and Discussions"],"author_name":"Ambika Choudhury","publish_date":"2019-01-11T20:45:10","publication_year":"2019","word_count":627,"keywords":["API","programming_languages:R","Interviews and Discussions","R","startup"],"extracted_tech_keywords":["R","API","startup","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/everything-you-need-to-know-about-li-fi-how-it-scores-over-wi-fi\/","complexity_score":3,"technical_depth":4,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10054722,"title":"AWS Introduces A New Platform For Developing Robotics Applications","content":"Amazon introduced AWS IoT RoboRunner during a keynote address at its Amazon Web Services (AWS) re:Invent 2021 conference – a new robotics service meant to make it easier for organisations to build and deploy apps that enable fleets of robots to work together. Along with IoT RoboRunner, Amazon introduced the AWS Robotics Startup Accelerator, a joint venture with MassRobotics to address automation, robotics, and industrial internet of things (IoT) concerns. Robotics adoption — and automation adoption more broadly — has intensified as the epidemic spurs digital reforms. According to recent research from Automation World, most businesses that adopted robots in the last year did so to reduce labour expenses, enhance capacity, and overcome a labour shortage. According to the same poll, 44.9 per cent of businesses now consider robots a vital component of daily operations in their assembly and manufacturing plants. IoT RoboRunner IoT RoboRunner, which is now in preview, leverages the robotics management technologies already in use at Amazon warehouses. It enables AWS clients to connect robots and current automation software to orchestrate work across operations by combining data from each type of robot in a fleet and standardising data types such as facility, location, and robotic job data in a centralised repository. According to Amazon, the purpose of IoT RoboRunner is to streamline the process of developing management apps for robot fleets. As organisations increasingly rely on robotics to automate their operations, they are selecting various types of robots, complicating the task of efficiently organising them. Each robot vendor and work management system has its proprietary control software, data format, and data repository, frequently incompatible. Additionally, when a new robot is added to a fleet, programming is required to integrate the control software with work management systems and develop the logic for management applications. Source: AWS IoT RoboRunner IoT RoboRunner enables developers to access the data required to construct robotics management apps and leverage prebuilt software libraries to create apps for activities such as work allocation. Additionally, IoT RoboRunner can be used to transmit measurements and key performance indicators (KPIs) to administrative dashboards via APIs. IoT RoboRunner competes with robotics management platforms from companies such as Freedom Robotics and Exotec. However, Amazon argues that IoT RoboRunner’s integration with AWS — including SageMaker, Greengrass, and SiteWise — provides it with a competitive edge. “With AWS IoT RoboRunner, robotics developers can eliminate the requirement for siloed robot management and more effectively automate operations across a facility with centralised control,” Amazon noted in a blog post. “As we look to the future, we envision more businesses investing in a variety of different sorts of robots. While harnessing the power of all those robots is complicated, we are committed to assisting organisations in maximising the value of their automation by simplifying robot optimisation through a single system view.” Accelerator for AWS Robotics Startups Amazon also introduced the Robotics Startup Accelerator, which will help encourage robotics startups by offering resources to help them create, prototype, test, and commercialise their products and services. “By leveraging AWS’ technical capabilities and network, the strategic relationship will enable robotics startups and the industry at large to explore and innovate, while also connecting startups and their technologies to the AWS customer base,” Amazon noted in a blog post. AWS IoT-enabled robotics application RoboRunner integrates with AWS IoT SiteWise for metric collection, Amazon SageMaker for task allocation using machine learning, and AWS IoT Greengrass for developing, deploying, and managing device software. With AWS IoT RoboRunner, robotics developers can eliminate the requirement for siloed robot management and more effectively automate activities across an entire facility with centralised control. The AWS IoT RoboRunner service is currently in preview and may be accessed via the AWS Console.","excerpt":"Amazon introduced AWS IoT RoboRunner during a keynote address at its Amazon Web Services (AWS) re:Invent 2021 conference – a new robotics service meant to make it easier for organisations to build and deploy apps that enable fleets of robots to work together. Along with IoT RoboRunner, Amazon introduced the AWS Robotics Startup Accelerator, a […]","categories":["IT Services"],"tags":["AI (Artificial Intelligence)","Machine Learning","Robotics"],"author_name":"Dr. Nivash Jeevanandam","publish_date":"2021-12-02T17:00:00","publication_year":"2021","word_count":615,"keywords":["API","Amazon SageMaker","machine learning","AWS","AI","ML","Machine Learning","Git","Robotics","RAG","GAN","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","machine learning","ML","Amazon SageMaker","RAG","AWS","R","Git","API","GAN"],"url":"https:\/\/analyticsindiamag.com\/it-services\/aws-introduces-a-new-platform-for-developing-robotics-applications-2\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10081927,"title":"Apple Lets User Privacy Slide, Chases Ad Revenues","content":"Today, Apple announced that it would no longer scan photos scanned in users’ iCloud accounts. While this seems like a move to protect the privacy of its users, the real reason behind this appears to be motivated by self-preservation. For context, Apple had launched a service in August 2021 with an aim to scan users’ iCloud for child sexual abuse material (CSAM). While it stated that this tool would preserve the privacy of users, many security researchers found that it could be used as a surveillance capability. This led to a huge backlash, prompting the company to roll back this feature. Instead of this controversial scanning feature, Apple has chosen to invest heavily in its ‘Communication Safety’ feature, which relies on parents and guardians to opt into protective measures. This feature works in various Apple services like Siri, Spotlight search, and Safari search to gauge whether someone is searching for CSAM. Apart from this, the feature also uses on-device machine learning to detect nudity in messages sent by minors. The tech giant had also hit the news lately regarding its controversial change in Apple devices in China. During the peak of COVID-19 in the country, protestors were using Apple AirDrop to share propaganda when the Chinese government had cut off access to the internet. In response to this, Apple quietly disabled the feature to share files with everyone through AirDrop, instead adding a ‘Contacts Only’ feature. While this move has since been replicated on all Apple devices worldwide, it was initially only rolled out in China. All these moves and more show that Apple only cares about its carefully-curated brand image and stance on privacy, and not the actual privacy of users. Behind the iCloud Scanning Feature Apple announced this feature in August 2021, along with a host of other CSA prevention measures. The scanner feature was reported to check all images uploaded to iCloud through a combination of on-device cryptography and cloud servers. These images would be checked against a database of known child sexual abuse images, and reported to the authorities if found to be match. While the feature was widely praised by children’s safety advocacy groups, many privacy critics strongly opposed the move. In their eyes, Apple made this move in the face of pressure from law enforcement authorities. Even as Apple quashed rumours that the feature would scan non-CSAM pictures, the backlash from the move prompted the company to cancel the feature. While adding image analysis to user devices does prevent misuse, it presents an interesting conundrum and undermines user privacy. Apple has historically taken a very strong stance on user privacy, standing against tech giants like Google and Microsoft which exploit user data for financial gain. The tech giant has routinely set the standard for privacy in customer devices by offering features like personal data anonymisation, on-device algorithms for processing facial data, and was the first to offer Secure Enclave on their mobile devices to store user passwords and financial information. However, this stance is slowly eroding away in the face of larger market pressure and a move towards monetising user data through advertising. Apple vs Advertisers In 2020, Apple launched a feature that aimed to improve user’s knowledge on how apps track them. Dubbed App Tracking Transparency, it aimed to stymie advertisers by giving users the chance to opt out of data collection processes. This, along with abolishing the identifier for advertisers feature, seemed to signal to the market that Apple was gearing up for the next generation of data security practices. However, the opposite was true. After booting advertisers from its ecosystem, Apple has since made moves to consolidate its own advertising power. The tech giant previously sold ad space on its News and Stocks apps, but has since expanded its advertising efforts to include the front page of the App Store. This brand strategy has also drawn questions of anticompetitive practices, as advertising giants like Meta lost an estimated $13 billion when Apple revamped its advertising policies. To serve these targeted advertisements to users, Apple collects users’ data from their other services and their Apple account details. This stands in stark contrast to its privacy-first policy, which assumes zero knowledge on the user’s part. Moreover, even if users opt out of this advertising system, it still collects information on the type of device and identity of the users’ carrier. These efforts paid off, as Apple increased its advertising revenue by 4x between 2021 and 2022. However, the company is looking to deflect attention away from its anti-privacy practices by introducing a slew of features designed to ‘protect’ the privacy of its users. Apple is No Longer Privacy’s White Knight Even beyond the purported features designed to increase the security of the users on their devices, Apple still engages in tracking for advertising purposes. A study by software company Mysk saw that Apple still tracks users, even when their settings say that they don’t. Moreover, in 2021, a human rights lawyer found that Apple was tracking her activity across all applications in records tracing back to 2017. This is also in line with Apple’s ad monetisation policy, which relies on a cost-per-tap model. Only highly targeted ads can provide the conversion rates required to keep this model sustainable. Apple might put on a show to maintain its curated privacy-focused brand image, but it has indulged in user tracking, device identification, anti-consumer practices and antitrust violations to build up its advertising.","excerpt":"Apple only cares about its carefully-curated brand image and stance on privacy, and not the actual privacy of users","categories":["Global Tech"],"tags":["Apple"],"author_name":"Anirudh VK","publish_date":"2022-12-09T11:00:00","publication_year":"2022","word_count":907,"keywords":["Replicate","Go","machine learning","AI","ETL","Apple","Aim","ViT","Rust","GAN","R"],"extracted_tech_keywords":["AI","machine learning","Aim","R","Go","Rust","ETL","GAN","ViT","Replicate"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/apple-forgets-user-privacy-chases-ad-revenues\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10045332,"title":"Top 7 Free Resources To Learn Snowflake","content":"In less than a decade, Snowflake has emerged as one of the most reliable platforms for data warehousing. Benoit Dageville, Thierry Cruanes and Marcin Żukowski founded the company in July 2012. Snowflake offers data-warehouse-as-a service or cloud-based data storage and analytics services. Snowflake has been on Amazon S3 since 2014, Microsoft Azure since 2018, and Google Cloud for the last two years. The platform enables data storage, processing, and analytic solutions that are faster, easier to use, and flexible than traditional offerings. According to Forrester’s ‘Total Economic Impact Study,’ Snowflake customers can expect a return on investment (RoI) of 612 percent and total benefits of more than $21 million over three years. In this article, we list free resources and tools to help you learn the nuts & bolts of Snowflake. Courses & modules Snowflake Documentation Snowflake Documentation is a go-to resource for users planning to learn Snowflake from scratch. The documentation provides conceptual overviews, tutorials, and a detailed reference for all supported SQL commands, operations and functions. The module covers: System prerequisites and requirements for using Snowflake after setting up your accountInstructions for connecting to Snowflake using the web interface and other supported clients Introduction to the main pages in the web interface and the tasks you can perform on each page Step-by-step instructions for using the command-line interface to create a table and database, load data into the table from an external data, and query the table Check out Snowflake documentation, podcast, hands-on essentials, workshop lab and additional resources. Snowflake Datawarehouse & Cloud Analytics – Introduction In this free Udemy course, you will learn about Snowflake Cloud Data Warehouse and its architecture. It will teach you how to build highly scalable, high performance next-gen modern data warehouses. The beginner-friendly course covers the basics of cloud, SaaS, etc. Check out the free course here. Videos What is Snowflake? In this eight-minute demo video, the Snowflake team gives an overview of the Snowflake Cloud Data platform. For a more in-depth demo, you can sign up for its weekly live demo programme here. Snowflake Architecture – Learn How Snowflake Stores Table data In this video lesson, Prashant Pandey from Learning Journal unpacks Snowflake Cloud Data Warehouse and the basics of Snowflake, separation of storage and computing techniques, and storage architecture (how storage database tables are stored and accessed when you execute SQL). Getting Started With Snowflake The Kahan Data Solutions tutorial walks you through the main components of the Snowflake user interface. The video demo covers the introduction of Snowflake, initial view, worksheets, databases, shares, warehouses, history, accounts and account versus worksheet role, notifications, partner connect, preview app, persisted queries and query history. Books & guides Architecting Data-Intensive SaaS Applications: Building Scalable Software with Snowflake Co-authored by William Waddington, Kevin McGinley, Pui Kei Johnston Chu, Gjorgji Georgievski and Dinesh Kulkarni, the book titled ‘Architecting Data-Intensive SaaS Applications: Building Scalable Software with Snowflake‘ covers the entire aspect of data applications and why it matters; things to look out for in modern data platform; building scalable data applications; data processing; and data sharing. Click here to download the pdf version of the book. The Founder’s Guide to Building Data Apps on Snowflake The guide describes five of the most common use cases for startups and companies building data applications and how Snowflake addresses the key challenges when developing applications including. Customer 360 Data Apps  IoT Data Apps Application Health and Security Analytics Data Apps Machine Learning and Data Science Apps Embedded Analytics Data Apps Click here to download the pdf version of the guide.","excerpt":"Snowflake offers data-warehouse-as-a service or cloud-based data storage and analytics services.","categories":["AI Trends"],"tags":["Snowflake"],"author_name":"Amit Naik","publish_date":"2021-08-06T16:00:00","publication_year":"2021","word_count":591,"keywords":["data science","Go","machine learning","AI","R","RAG","analytics","SQL","Azure","Snowflake"],"extracted_tech_keywords":["AI","machine learning","data science","analytics","RAG","Azure","Snowflake","R","SQL","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-7-free-resources-to-learn-snowflake\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":36624,"title":"Frontdesk AI Raises $2 Million In Pre-Series A Funding, pi Ventures Emerges As Lead Investor","content":"Frontdesk AI Pi Ventures, a leading venture capital fund supporting startups in the field of artificial intelligence, machine learning and Internet of Things, recently announced pre-Series A investment in Frontdesk AI, a Palo Alto and Bengaluru based startup that develops AI assistants for small businesses. Boosting the fastest adopted AI assistant in the wellness market, the startup raised an additional $2 million in funding, bringing its total seed funding to $4.2 million. It has earlier raised $1.1 million led by Pointe Capital and $1.1 million led by Speciale Invest in 2018. The startup aims to use the funding amount to accelerate product development, increase customer acquisition and investigate new vertical markets. Also as a part of the deal, pi Ventures’ partner, Abishek Surendran, will join the Frontdesk AI board that already includes BluePointe Ventures managing partner, Sandeep Sardana, and Speciale Invest managing partner, Vishesh Rajaram. The startup was founded in 2017. Frontdesk AI pioneers in creating AI assistants for small businesses that rely on appointment-based products and services for their revenue. Offering round-the-clock information and reservations for customers as a managed messaging service, the startup focuses on one thing and that is the small business. As the company says, their AI doesn’t predict weather, order groceries or answer trivia, but addresses simple, repetitive business tasks such as booking appointments, answering product questions, and remembering details about each and every customer. They are powered by a set of proprietary algorithms, a proprietary database of B2C natural language interactions, and knowledge graphs. Over time, they build a rich source of information about customer requirements, preferences and usage patterns. By tapping into the Frontdesk AI knowledge graph, the AI assistants run immediately on deployment. To create human empathy, Frontdesk AI provides a second layer of intelligence comprising human trainers to groom the AI assistant to speak in the style and tone of the customer brand. The AI assistant becomes independent over time building knowledge and experience from all its interactions. “We are laser focused on developing affordable and easy-to-use AI assistants for small businesses. With the additional funding and the partnership with pi Ventures we are now in a position to deepen our IP and competitive barriers to market entry,” says Srivatsan Laxman, CEO, Frontdesk AI. It currently serves the wellness market including the spa, salon and fitness industries. Frontdesk AI customer acquisition is currently growing 17% month on month, subscription revenue is growing 25% month on month, and churn is amongst the lowest in the industry at less than 3% per month. Abishek Surendran, Partner at pi Ventures, said, “We are excited about partnering with  Frontdesk AI which has developed a world-class AI assistant for the small and medium enterprise market. Their deep learning-based AI engine means that companies can provide superior customer experiences at very low costs. We look forward to supporting Srivatsan and the  Frontdesk AI leadership team to accelerate and expand upon the market advantage they have created.” pi Ventures announced the final close of their first fund at ₹225 crore in September 2018. It had also mentioned its plan to invest in 18-20 deep tech startups in India in the next 3-4 years especially those in health-tech, logistics, retail, FinTech and enterprise sectors.","excerpt":"Pi Ventures, a leading venture capital fund supporting startups in the field of artificial intelligence, machine learning and Internet of Things, recently announced pre-Series A investment in Frontdesk AI, a Palo Alto and Bengaluru based startup that develops AI assistants for small businesses. Boosting the fastest adopted AI assistant in the wellness market, the startup […]","categories":["AI News"],"tags":[],"author_name":"Srishti Deoras","publish_date":"2019-03-20T06:57:07","publication_year":"2019","word_count":535,"keywords":["knowledge graphs","Go","AI assistants","machine learning","artificial intelligence","API","AI","Aim","deep learning","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","deep learning","Aim","AI assistants","knowledge graphs","R","Go","API"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/frontdesk-ai-raises-2-million-in-pre-series-a-funding-from-pi-ventures\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10110557,"title":"Data Science Hiring and Interview Process at SAP Labs India","content":"German tech conglomerate SAP Labs has been one of the major players in the generative AI race on the enterprise side. The company recently introduced Joule, a natural-language generative AI assistant that allows access to the company’s extensive cloud enterprise suite across different apps and programs. It will provide real-time data insights and action recommendations. With a global presence in 19 countries, labs are responsible for driving SAP’s product strategy, developing and localising its core solutions, and contributing to the SAP Business Technology Platform. 54-year-old SAP was founded by five former IBM employees, Dietmar Hopp, Hasso Plattner, Claus Wellenreuther, Klaus Tschira, and Hans-Werner Hector. SAP Labs is the R&D arm of SAP with its second largest office space in Bengaluru. AIM got in touch with Shweta Mohanty, vice president and head, of human resources, SAP, India and Dharani Karthikeyan, vice president, head of engineering for analytics, SAP Labs India, to understand the company’s AI and analytics play, customer stories, hiring process for data scientists, work culture and more. AI & Analytics Play “We have fully embraced generative AI in our business AI concept, aiming to provide AI that is responsible, reliable, and relevant. The goal is to infuse AI into business applications, with a focus on trust and outcomes,” Karthikeyan told AIM. SAP has a portfolio of over 350 applications spanning various use cases, from cash management to document scanning. The company is enhancing its Business Technology Platform (BTP) with a generative AI layer.  The team aims to improve business processes while maintaining human control over decisions. They have collaborated with Microsoft for Human Capital Management tools, combating biases in recruiting, and introduced a Business Analytics tool for faster insights. SAP is also partnering with Google Cloud to launch a holistic data cloud, addressing data access challenges. Additionally, they have invested in generative AI players Anthropic, Cohere, and Aleph Alpha, diversifying their capabilities. Interview Process The hiring process for tech roles involves five to six steps starting with profile screening, focusing on the candidate’s development background and programming language proficiency. As described by Mohanty, this is followed by an online assessment to test programming skills, lasting 60 to 90 minutes. Technical interviews include case studies to assess proficiency and hands-on experience. For senior roles, there’s a discussion with a senior leader to gauge cultural alignment. The final step is an HR discussion focusing on cultural fit and interest in the organisation. For college recruitment, the process includes live business solutions assessments. The process concludes with a rigorous background verification. When it comes to finding the right fit for SAP labs, the ideal candidate should have a comprehensive understanding of ML algorithms, and to build and maintain scalable solutions in production,” added Karthikeyan, highlighting that this consists of the use of statistical modelling procedures, data modelling, and evaluation strategies to find patterns and predict unseen instances. The roles involve using computer science fundamentals such as data structures, algorithms, computability, complexity, and computer architecture and also collaborating with data engineers is essential for building data and model pipelines, as well as managing the infrastructure needed for code production. As a data scientist at SAP Labs India, you will also analyse large, complex datasets, researching and implement best practices to enhance existing ML infrastructure and provide support to engineers and product managers in implementing ML into products. Work Culture in SAP SAP’s work culture is characterised by abundant learning opportunities and hands-on experiences where employees have chances to shadow leading data scientists, participate in fellowship projects for stretch assignments, and explore various aspects. This hands-on approach extends to customer interactions and pre-sales experiences. “These opportunities, along with the focus on learning and customer engagement, give SAP an edge over other organisations hiring in data science and machine learning,” Mohanty commented. SAP prioritises its employees’ well-being through a comprehensive set of benefits and rewards. The company recognises diverse needs beyond healthcare and retirement plans, offering global and local options for work-life balance, health and well-being, and financial health. Embracing a highly inclusive and flexible culture, the company promotes a hybrid working model allowing employees to balance office and remote work. Employee Network Groups foster a sense of community, and inclusive benefits include competitive parental leave and disability support. The ERP software giant also aims to foster personal and professional growth, providing learning opportunities, career development resources, and a leadership culture focused on doing what’s right for future generations. It values fair pay, employee recognition, generous time-off policies, variable pay plans, total well-being support, and stock ownership opportunities for all employees. Why Should You Join SAP Labs? SAP Labs offers a sense of purpose and involvement in transformative technology phases. At SAP, candidates dive into cutting-edge technologies, explore diverse industries, and embrace continuous learning and innovation. Mohanty explained how the team values adaptability, emphasising fungible skills and a proactive mindset, especially in areas like AI and generative AI. “We seek individuals ready to tackle new challenges and solve complex problems, fostering a dynamic and impactful work environment,” she explained. Adding on to what Mohanty said, “The work at SAP involves mission-critical applications, like supporting cell phone towers or vaccine manufacturing so the integration of generative AI into these applications offers a unique combination of purpose and technological advancement, providing developers with a high sense of purpose in seeing their software run essential business and retail operations. This phase of technological transformation at SAP is especially significant for new joiners,” said Karthikeyan. Check out the job openings here.","excerpt":"As a data scientist at SAP Labs, you will analyse large datasets, implement best practices to enhance ML infrastructure, and support engineers and product managers in integrating ML into products.","categories":["AI Hirings"],"tags":["Data Science","Data Science Hiring","Top Trend"],"author_name":"Shritama Saha","publish_date":"2024-07-29T15:43:34","publication_year":"2024","word_count":912,"keywords":["Top Trend","data science","Anthropic","machine learning","Go","AI","ML","Data Science Hiring","Aim","analytics","generative AI","Data Science","R"],"extracted_tech_keywords":["AI","machine learning","ML","data science","analytics","generative AI","Anthropic","Aim","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-hiring\/data-science-hiring-process-at-sap-labs-india\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":65381,"title":"Turning Waste To Wealth: The Circular Economy Of Data Centres","content":"The global demand for resources is many times higher than what the Earth can support in a year, which means the linear economy model will soon slam into the edge of its physical limits. Data centres especially are one of the most material-intensive facilities ever set up by mankind. From chips to giant cooling towers, the requirement of materials in a typical data centre is enormous. Google alone has more than a dozen of these facilities across the world. With increasing and improving services of Google, the demand for data centres kept rising. So, to cut down the waste and convert whatever waste is left to wealth, they employ principles of the circular economy. This circularity strategy complements Google’s vision for a carbon-free and the circular world, but by its very nature, draws from and contributes to other aspects of a sustainability strategy. To make this work, Google has partnered with the Ellen MacArthur Foundation, to adopt the circular economy practices and leverage its enormous benefits. Making Circular Economy Work At The Scale Of Google A worker at Google data center A circular economy model is not new to the industry, and it is usually employed to reduce wastage by recycling the disposed parts. To accommodate this, the products right from their procurement are chosen and built in a way that they can be made again even after they are scrapped. In short, refurbish, repair, reuse and recycle. In 2016 alone, Google used 22% of the components for machine upgrades and 36% for servers from the refurbished inventory. The refurbished parts are taken from previous servers and are used when a server shows up for repair. This gives the hard drives an extended life. To maximise recycling, Google uses a multi-step destruction process to ensure that the data doesn’t fall into the wrong hands. In this multi-step process, the first step involves a “crusher” that drives a steel piston through the centre of the hard drive and make them unreadable. They are then shredded before the remains are sent along with other electronic waste to a recycling partner for secure processing. But before any hard drives are removed from the rotation, all data is overwritten and are verified with a complete disk read, so that no trace of customer data can possibly remain on the hard drive. The decommissioned servers are then taken for a second lap in the circle of life where they’re dismantled into separate components (motherboard, CPU, hard drives, etc.), inspected, and prepped for use as refurbished inventory. Google redistributes commercially useful excess component inventory by wiping clean the unused components and getting them checked multiple times before redistributing them for a resale on the secondary market. Not only in its own data centres but Google even resells these reusable parts to other organisations around the world as well. Future Direction The aforementioned strategies, combined with their optimised custom-built infrastructure, saves Google tons of money. Because most of their infrastructure is custom-designed, For instance, Google’s IP data network has its own fibre, peering, and undersea cables, which allows them to deliver low latency services. From our own high voltage substations to the on-site electrical distribution systems, to the proprietary cooling systems — all these custom-built infrastructure work in conjunction, saving Google over $1 billion in operations through energy efficiency alone. So far, we have taken the example of Google because they have been vociferous for over a decade about their goal to make their data centres carbon neutral. For this, they have even introduced carbon intelligence computing. The number of customers that use Google cloud is slowly on the rise, and with increasing AI-backed cloud services, the data centres need to be more efficient than ever before. That said, other top players such as AWS’ cloud segment employ similar strategies too.","excerpt":"The global demand for resources is many times higher than what the Earth can support in a year, which means the linear economy model will soon slam into the edge of its physical limits. Data centres especially are one of the most material-intensive facilities ever set up by mankind. From chips to giant cooling towers, […]","categories":["Deep Tech"],"tags":["data centres"],"author_name":"Ram Sagar","publish_date":"2020-05-17T13:12:44","publication_year":"2020","word_count":634,"keywords":["Go","AWS","AI","cloud_platforms:AWS","programming_languages:R","RAG","data centres","cloud_platforms:Google Cloud","GAN","R","Redis"],"extracted_tech_keywords":["AI","RAG","AWS","Redis","R","Go","GAN","cloud_platforms:AWS","cloud_platforms:Google Cloud","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/data-centres-waste-management\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10142013,"title":"Amazon Reclaims its Leadership in AI and Cloud Innovation in 2024","content":"Amazon’s key focus in 2024 was integrating AI into its ecosystem to improve productivity and customer experience. The company launched several AI tools and initiatives, highlighting its commitment to empowering businesses and developers while addressing societal concerns about AI ethics and inclusivity. Amazon showcased its ability to innovate across diverse domains by releasing generative AI tools to promote satellite internet technology, even amid strong competition from Microsoft and Google. Redefining Developer Productivity This year, Amazon Web Services announced the general availability of its AI developer tool Amazon Q. This tool automates repetitive coding tasks and enables developers to focus on complex problem-solving. According to Amazon, Q saved over 4,500 developer years of work in 2024 alone, a claim underscored by testimonials from early adopters. “The benefits go beyond how much effort we’ve saved developers. The upgrades have enhanced security and reduced infrastructure costs, providing an estimated $260 million in annualised efficiency gains,” claimed Amazon CEO Andy Jassy. Following the latest update, developers can now initiate online chats with Q Developer. AWS continued to lead the cloud AI space with advancements in Amazon Bedrock, its generative AI service. Powered by Titan models and featuring partnerships with tools like Anthropic’s Claude, Bedrock expanded to regions like APAC (Mumbai), underscoring Amazon’s global ambitions. Bedrock also enabled businesses to achieve tangible results. For instance, lending platform Fibe, which uses Bedrock’s service, saw a 30% increase in customer support efficiency. Projects and AI Assistants Last month, to boost its retail and customer service, Amazon launched Project Amelia, an AI assistant for sellers on its retail site. This tool helps businesses improve customer interactions by integrating generative AI capabilities into their workflows. Additionally, Rufus, Amazon’s AI assistant, went live in India, strengthening the company’s efforts to capture emerging markets. Rufus, designed with cultural inclusivity in mind, reflects Amazon’s emphasis on creating tools that cater to diverse user bases. As Amazon CTO Werner Vogels said, “Building culturally inclusive AI is not just a technical challenge; it’s a societal imperative.” He emphasised Rufus’ ability to incorporate local languages and cultural contexts. AWS Amplify AI Kit enables developers to rapidly prototype and deploy AI app development with minimal friction by democratising AI app development. It empowers both front-end and full-stack teams to make use of the power of AWS services and AI models to create transformative applications. Sandboxes and intuitive resource organisation help teams iterate faster and focus more on innovation rather than infrastructure. Serverless scaling and pay-as-you-go pricing also ensure cost efficiency for startups and enterprises. App Studio IntegrationThis year, AWS App Studio, an application development environment powered by Titan and Claude models, was introduced. This tool simplifies app creation for developers and businesses, reinforcing AWS’s reputation as the go-to platform for scalable AI solutions. Sriram Devanathan, GM of Amazon Q Apps and AWS App Studio at AWS, said App Studio is going to be transformative. “It’s the fastest way to build enterprise-grade applications in the sense that they have multiple UI pages, they can pull data from multiple sources and they can embed complex business logic in them,” he added. Beyond AI While AI remained central to Amazon’s strategy, the company also achieved significant milestones in other domains. Satellite internet initiative Project Kuiper by Amazon achieved a 100% success rate in its low-Earth orbit satellite missions. This set Kuiper up as a viable competitor to SpaceX’s Starlink and demonstrated Amazon’s commitment to bridging the digital divide through affordable global internet access. Rajeev Badyal, VP of technology for Project Kuiper, acknowledged the team’s dedication and expertise and emphasised the rapid success achieved on the first mission. Ethical Tech Takes Center Stage As innovation dominated Amazon’s agenda, the company also acknowledged the ethical challenges tied to AI. In 2024, Amazon launched the Trusted AI Challenge for LLM Coding Security, encouraging developers to design safer LLMs. This highlights the company’s proactive stance on mitigating risks like data breaches and bias. Amazon CTO Werner Vogels, addressed the company’s efforts to train culturally sensitive AI, and said, “Inclusivity in AI isn’t optional, it’s essential. We’ve made strides in ensuring our models are trained with diverse datasets, but we’re also focused on how they interpret and interact with cultural nuances.” These efforts reflected Amazon’s broader mission to lead responsibly in the AI era, addressing concerns about trust, inclusivity and ethical development. Moving into 2025, Amazon’s dual focus on innovation and cultural inclusivity will likely remain central to its strategy. Notably, 2024 wasn’t just about staying competitive, it was about setting new benchmarks for what technology can achieve when aligned with ethical principles and global aspirations.","excerpt":"As Amazon moves into 2025, its dual focus on innovation and cultural inclusivity will likely remain central to its strategy.","categories":["AI Features"],"tags":["Amazon","Amazon AWS","Amazon Bedrock","amazon q"],"author_name":"Tanisha Bhattacharjee","publish_date":"2024-11-29T14:33:56","publication_year":"2024","word_count":762,"keywords":["Anthropic","Go","Amazon Bedrock","AI assistants","AWS","AI","amazon q","Amazon","serverless","RAG","Amazon AWS","Aim","generative AI","R"],"extracted_tech_keywords":["AI","generative AI","Anthropic","Aim","RAG","AI assistants","AWS","serverless","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/amazon-reclaims-its-leadership-in-ai-and-cloud-innovation-in-2024\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10012443,"title":"Biggest AI Goof-Ups That Made Headlines In 2020","content":"The innovation of GPT-3 and advancements in facial recognition technology, brain chips, chatbots, self-driving cars, drones, as well as robotics have marked 2020 as the year of artificial intelligence. However, similar to any other technologies, AI also came with its challenges. From biasness to inaccuracy, AI has proved its immaturity in many cases. As a matter of fact, prominent tech leaders, researchers as well as scientists, like Elon Musk, Yann LeCun, as well as Bill Gates have continuously warned the industry about the hype AI has created and the consequences it can bring if not appropriately handled by the tech giants. Such critical judgements came from the many instances where AI failed to demonstrate its value to the industry. From AI’s failed COVID prediction to AI mistaking referee’s bald head for a ball in a football match, here are eight biggest AI goof-ups and blunders of 2020 that made the front page, in no particular order. Also Read: 10 Biggest Data Breaches That Made Headlines In 2020 Read AI Goof-ups from last year. Read AI failures of all time. The inaccuracy of YouTube’s AI moderators To sustain the COVID pandemic, similar to other companies, YouTube, in March 2020, turned towards AI and machine learning to moderate videos and content on its platforms. With fewer human employees and a major dependence on AI led to the removal of 11 million videos from April to June alone, which is a far higher number than usual. Such a huge number brought in the attention of many content creators and YouTube has undergone a heavy backlash of removing many videos that aren’t violating any rules. Out of these takedown videos, 320,000 were appealed, with half of them reinstated. While the company blamed the entirety on machines, this AI blunder is indeed notable. And that’s why, in September, Youtube announced to bring back its human moderators who were kept at a halt during the pandemic. Read more about it here. AI hits the wall & mistakes referee for a ball In a week of AI goof-ups in November, two main instances made headlines — first, where a self-driving car in a motor race drives straight into the wall, and secondly in a live match, where an AI camera followed the bald head of a linesman instead of the ball. For the instance of a self-driving car, an engineer explained to the media that the failure happened during the initialisation lap where the AI got confused and turned the steering wheel to NaN (not a number), which led to the car hitting the wall. Similarly, in the AI camera case, the algorithm confused the bald head for footballs. Such scenarios, where AI is getting easily confused between different objects, made researchers and scientists think how AI can react to cases where human lives are involved. Read more about it here. Fumbling of facial recognition technology The COVID pandemic has brought in a new norm of wearing face masks; however, such a change has completely perplexed the pre-pandemic facial recognition algorithms. In recent news, it has been stated that the Face ID of Apple’s recently released iPhone 12 failed to recognise people with face masks. This raised significant concern for all the facial recognition models that were trained on pre-pandemic data without masks. To test this theory, in a recent study, researchers experimented on prominent commercial facial recognition applications and noted that 89 of those came up with errors. Such a drop in the accuracy of these applications highlights how these facial recognition systems can be deceived and result in misinformation about people. Read more about it here. The wrongful arrest of Robert Williams Similar to the above mentioned “facial recognition system fail,” in another news, an African American man was wrongfully arrested after a facial recognition system mistaken his photo as a shoplifter. Facial recognition has been a controversial technology since its inception, and with this faulty algorithm news, it made massive havoc in the industry. Not only this raised concerns but also forced tech giants like Amazon, Microsoft, IBM, as well as Google to stop their facial recognition offerings for police authorities. Clearview AI, another facial recognition company, has also been criticised this year for collecting personal information and images of people from their social media accounts without any proper permission system. Such instances also made many researchers and scientists work towards making this technology less biased and more useful for humanity. Read more about it here. Also Read: Five Most Controversial Moments Of AI In 2020 The A-Level fiasco In another case of sustaining COVID pandemic, the government of UK, in August, substituted teachers with AI algorithms to score students for a cancelled A-level (advanced-level) qualification exams. However, the algorithm scored the students based on the historical performance of individual secondary schools, which resulted in students receiving way lower than they had expected. As a matter of fact, for a few students, the AI’s result would make them ineligible for the university programs that they were expecting to attend. In this scenario, the scores of bright students from less-advantaged schools were decreased while the marks for more affluent schools were increased. This denoted that the AI algorithm had a societal bias that forced it to make students wrongfully. Read more about it here. The most devastating software mistake In another AI goof, a popular epidemiologist from the Imperial College, London — Neil Ferguson lost his job. In this case, the AI model inaccurately forecasted the possible deaths due to COVID-19 with and without lockdown. However, there have been many instances; this one made the headlines because of its importance in framing policies in the UK and US. The model inaccurately predicted deaths in the US and UK, which led to structural mistakes in analysing the outbreak response. According to experts, not only the countries ignored standard contact tracing but also missed continuous monitoring to identify patients with symptoms. This led to massive havoc and an increase in the number of COVID patients. Such scenarios highlight how to complete dependability on AI can lead to irreparable outcomes. Read more about it here. Twitter & Zoom’s Racism For both of these cases, the AI was biased against the black community by either cropping out or omitting the entire head of the people. In Twitter’s case, the US programmer, Tony Arcieri, launched an experiment on Tweet to check the algorithm’s biasness. He noted that when uploaded large photo collages of former US President Barack Obama and Republican Senate leader Mitch McConnell, Twitter’s image preview automatically cropped out Obama’s face instead of McConnell’s. With several similar experiments, the programmer stated that Twitter’s is yet another example of biases in ML algorithms. In a similar case with Zoom, a video conferencing company, the platform cropped out the entire head of a black faculty member when used as a virtual background. Such instances will keep reminding us of the issue working with bad algorithms. https:\/\/twitter.com\/colinmadland\/status\/1307111818981146626 Read more about it here. Google’s inaccurate medical AI No list can be complete with Google’s mention. While AI has been making a breakthrough in the medical industry, in one instance, Google’s medical AI failed to prove its accuracy in real-world testing. The tool which claimed to identify signs of diabetic retinopathy from an eye scan with more than 90% accuracy, was first tested in Thailand. The AI tool indeed speeds things up, but its performance level in the real world wasn’t up to the mark. Considering the deep learning models were trained on high-resolution images to ensure accuracy, in the real world, the images below quality level were rejected, which caused a lot of inconvenience for the healthcare staff. This, in turn, received a lot of backlash from the industry as well as from the public. Thus, questioning the usefulness of this technology. Read more about it here.","excerpt":"The innovation of GPT-3 and advancements in facial recognition technology, brain chips, chatbots, self-driving cars, drones, as well as robotics have marked 2020 as the year of artificial intelligence. However, similar to any other technologies, AI also came with its challenges. From biasness to inaccuracy, AI has proved its immaturity in many cases.  As a […]","categories":["AI Features"],"tags":["AI Fails","ai grading","human touch to ai"],"author_name":"Sejuti Das","publish_date":"2020-11-26T15:00:12","publication_year":"2020","word_count":1304,"keywords":["Go","artificial intelligence","machine learning","AI","chatbots","ai grading","ML","medical AI","Aim","deep learning","AI Fails","human touch to ai","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","Aim","chatbots","medical AI","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/biggest-ai-goof-ups-that-made-headlines-in-2020\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":502,"title":"10 reasons why IoT is rising today?","content":"With its own interesting timeline dating back to a few decades, internet has traversed a long way since then. Being restricted to PCs and laptops in its initial stages to the experimental devices such as internet toaster years later to the present day where every possible thing is connected to the internet- it has had a long and interesting journey. The escalation in the era of IoT was dramatically witnessed in the years around 2010 when internet connections started to pick up exponentially and machines and things began to take over the manual mode of working. If we consider the huge numbers that the IoT industry is being linked to, it’s evident that the industry is definitely rising today. But what has led to its drastic rise in today’s date? This article brings to you 10 most probable reasons why IoT is witnessing a surge like never before! 1. The mobile age: Proving to be a turning point in the way people stayed connected, an outpouring in the number of mobile phones, gave a big reason for the internet to flourish. It came as a big hope for everybody from companies to retailers to many others as a mean to reach out to more people and what better way than internet to fulfil this aim? Especially with the customer facing companies, internet facilitated in mobile phones came in as a big trendsetter. 2. Internet- ‘ifying’ business strategies: Yes, we mean the era of e-commerce. Soon the era of internet took a critical turn as the consumers were lured to e-commerce platforms where they could browse at their ease and convenience through phones and desktops. This brought a rise of the internet APIs and various ventures such as eBay, Amazon, Facebook, Flickr, Twitter and many more witnessed the internet revolution. That’s not all, companies also needed to think about strategies to keep up the business viable and that’s where internet came to picture. 3. Increase in the data generated: Whether internet led to increased data or data led to an organized internet, is still widely debated but internet was sure a necessity to keep all the huge amount of data that had begun to be generated, handier and easily accessible. 4. Drop in cost of services: This hands down have to be the biggest reasons for the way we see internet today. IoT would not be the way it is, if not for a fall in the cost of components and services which led it to be an economic player of this scale. The prices of the components and the services have dropped reasonably leading to an increased penetration of the users who can avail it. 5. Development of user friendly tools and easy programming: With the way coding has been developed over the years, and user friendly tools being presented to its customers, the reach and usability of the devices with internet, whether smartphones or iPad, has witnessed an observable rise. 6. Rise in IoT enthusiasts: People and the communities with an increased zeal in the area of connected things has drastically increased over the decade and they are in a constant look out for making the tools, tutorials and software available to more people. Today more than 5000 Github repositories are available with the keyword “IoT”. 7. Increase in cloud based services: Cloud based applications has given a reason for the startups in IoT to flourish. Ensuring a network connectivity, affordability and ease in communication for enterprises and consumers, there has been an increase in cloud based services and the IoT trend has been rightfully fulfilling so. 8. Upsurge in tech friendly ‘things’: Off late, there has been an undefying obsession with the technology such as the connected cars, the self-driving cars, smart cities or the talking robots. And who in turn gets to full fill those? Of course – our very own ‘Internet of things’. 9. More stress to security: One of the key drivers for internet to be the favourites of the masses and doing so well, is favoured network security. With a huge repository of data under its belt, assuring security becomes mandatory and that again has called in for more users relying on the internet. 10. Increased dependability in industries: Whether traffic management, waste management, smart water systems or more, the concept of internet of things has indulged itself deeply into various important industrial roles. Given the increased customer expectations in every domain, internet of things has come as a major rescue.","excerpt":"With its own interesting timeline dating back to a few decades, internet has traversed a long way since then. Being restricted to PCs and laptops in its initial stages to the experimental devices such as internet toaster years later to the present day where every possible thing is connected to the internet- it has had […]","categories":["AI Features"],"tags":[],"author_name":"Srishti Deoras","publish_date":"2016-08-10T11:35:45","publication_year":"2016","word_count":747,"keywords":["API","ELT","AI","Scala","Git","Aim","ViT","GAN","GitHub","R"],"extracted_tech_keywords":["AI","Aim","R","Scala","Git","GitHub","API","ELT","GAN","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/10-reasons-iot-rising-today\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10054742,"title":"10 AI Startups That Raised Funding In 2021","content":"Startups form the backbone of the Indian economy. This year, the Indian startup ecosystem outdid itself, with startups raising as much as $26 billion within the first eight months of the year. Of this, artificial intelligence or AI-backed startups form a chunk. According to AIM Research, Indian AI startups raised $836.3 million in funding in 2020. Today, we take a look at the AI startups that got funded in 2021 so far. Haber AI-driven robot maker Haber has raised $20 million in its Series B round of funding in November. The investment round was led by Ascent Capital, with participation from Accel, Elevation Capital, Beenext, Temasek partner Mukul Chawla and entrepreneurs Samay Kohli and Akash Gupta. Haber operates under Pune-based parent company, Elixa Technologies. The startup aims to use the freshly raised funds to expand across new geographies and industries, with the mission of saving over 400 billion litres of water in the next two years. Planys Technologies IIT-Madras incubated Planys Technologies has raised $1.6 million in its Pre-Series A investment round. The marine robotics startup has raised funds from Keiretsu Forum, Shell and Lets Venture. Additionally, existing investors Kris Gopalakrishnan, ONGC Startup Fund and MEMG also participated in the round. Planys Technologies will utilise the fund to accelerate its research and development, expand internationally and develop underwater asset management products. H2O.ai AI-backed H2O.ai raised $100 million in November. The investment round was led by the Commonwealth Bank of Australia, with participation from Goldman Sachs and Pivot Investment Partners, among others. The startup was founded by Sri Satish Ambati in 2012, and its platform H2O AI Cloud is on a mission to simplify the work involved in ML projects. Zini.ai Delhi-based Zini.ai raised an undisclosed amount of funding from London-based VC firm Solarus Group. The AI-powered multilingual virtual physician platform by Grainpad Pvt Ltd was founded in 2017 by Dr Rohit Sharma. Zini.ai is the country’s first voice-based virtual physician. The Delhi-based startup plans to use the raised funds to expand to new geographies. Kore.ai Founded by Raj Koneru, conversational AI software startup Kore.ai raised $50 million in its Series C round of funding from Vistara Growth and PNC. VC firms NextEquity Partners, Nicola Wealth and Beedie Capital also participated in the round. Headquartered in Orlando, Kore.ai’s engineering team works out of Hyderabad. The AI startup plans to utilise the funds to scale its efforts towards becoming a leader in the Experience Optimisation space. Emotix Mumbai-based startup Emotix, which is short for emotional intelligence, was founded in 2015 by IIT Bombay alumni Sneh Vaswani, Prashant Iyenger and Chintan Rajkar. Earlier in April this year, it raised Rs 26 crore or $3.5 billion as a mix of debt and equity from Vistra ITCL (India) Ltd, an independent trustee, on behalf of IvyCap Ventures Trust Fund II. Emotix has developed a companion robot Miko to help parents keep their children occupied with the help of computer vision. Emotix bets on AI, robotics, and the Internet of Things(IoT) to identify moods, initiate conversations and learn from its surroundings to engage with a child. The Mumbai-based startup is also backed by Chiratae Ventures and YourNest Capital. Emotix aims to provide solutions to 20 million homes across childcare and elderly care verticals as its future plans. Myelin Foundry Based out of Bengaluru, deeptech startup Myelin Foundry was founded by Gopichand Katragadda, Ganesh Suryanarayanan and Aditi Olemann in 2019. The startup raised $1 million in its pre-Series A round in May 2021. The round was led by VC firm Beyond Next Ventures, with participation from existing investor Endiya Partners. Myelin Foundry transforms human experiences and industry outcomes for edge devices by building AI algorithms on video, voice and sensor data. It is also backed by Microsoft for Startups, Pratithi Investment Trust, and investor and Infosys founder Kris Gopalakrishnan. Wysa AI-powered mental health application Wysa has raised $5.5 million in its Series A funding round in May this year. The Bengaluru-based startup was founded in 2015 by Touchkin eServices, Jo Aggarwal and Ramakant Vempati. The company also has offices in London and Boston. The startup is also backed by pi Ventures, Kae Capital and others. Wysa acts like a mental wellness platform, helping employees of a particular company or patients manage stress by analysing their anxiety behaviour. Cron AI Deeptech startup Cron AI raised $4 million, which is about Rs 29.3 crore, in its Series A investment round. The funding was led by Kitaki Ventures and VenturEast. Founded in 2015 by Saurav Agarwala, Tushar Chhabra, and Tommy Katzenellenbogen, Cron AI enables innovators to develop solutions to perceive and learn the real world using 3D sensors. According to media reports, Cron AI plans to use the latest funding to fasten the delivery of its sensor-agnostic 3D data perception platform SenseEDGE. SenseEDGE will be addressing 3D perception, sensing and processing requirements in new markets. Elucidata Delhi and Cambridge-based startup Elucidata raised $5 million from investors IvyCap Ventures, with participation from Hyperplane Venture Capital and other angel investors. The biomedical molecular data startup was founded in 2015 by Abhishek Jha, Swetabh Pathak and Richard Kibbey. Elucidata leverages machine learning to power biomedical data for the acceleration of drug discovery through its cloud platform Polly. Additionally, more AI startups raised funding: Light Information Systems raised $5.4 million from Pavestone Ventures; Data Sutram raised an undisclosed amount from IIFL Fintech Fund; Fivetran raised $565 million from new and existing investors, among others.","excerpt":"Here is a list of 10 AI startups that got funded in 2021.","categories":["AI Startups"],"tags":["AI (Artificial Intelligence)","H2O AI","Machine Learning"],"author_name":"Debolina Biswas","publish_date":"2021-12-03T12:00:00","publication_year":"2021","word_count":900,"keywords":["Go","artificial intelligence","machine learning","AI","ML","Machine Learning","computer vision","H2O AI","RAG","Ray","Aim","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","computer vision","Aim","Ray","RAG","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-startups\/10-ai-startups-that-raised-funding-in-2021\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10095338,"title":"Why Graph Databases Remain Untapped","content":"Currently, relational databases dominate with a 90% market share. However, when it comes to a graph database, there is little or no understanding of how it works. Some of the prominent relational databases include Microsoft SQL Server, Oracle Database, MySQL and IBM DB2. “Lack of understanding and awareness of graphs is their biggest competitor,” Dr. Jesus Barrasa, head of solutions architecture at Neo4j, in an interview with AIM, at their Annual Graph Summit in Mumbai. He said that many solutions are built on relational databases due to a lack of awareness of specialised graph stores.  the lack of understanding and awareness of graphs as their biggest competitor. He added that many solutions are built on relational databases due to a lack of awareness of specialised graph stores. Dr Jesus Barassa at graphsummit Adding on to what Barrasa said, said Dr Jim Webber, chief data scientist at Neo4j told AIM that graph databases can offer improved efficiency for many business information systems, but awareness and openness to new tools are crucial for exploring better solutions. Dr Jim Webber at graphsummit Overall, they believe in the adoption and coexistence of relational and graph databases rather than direct competition. Graph Database Vs Relational Database Graph databases store relationships as data, allowing real-time addition of nodes and relationships, while relational databases focus on relationships between columns and allow adding tables and columns on the go. Graph databases excel in executing complex queries, outperforming relational databases that rely on slower joins between tables. However, amid the to and fro between relational and graph databases, Neo4j disrupted the ecosystem with a graph database of the same name along with its declarative query language called Cypher, borrowing concepts from SPARQL. Unmasking the Power of Graph Technology Graph technology is being widely adopted across diverse industries for more efficient data analysis and decision-making. Major sectors utilising graph technology include social networking, telecommunications, banking, fraud detection, transaction processing, compliance and regulation, supply chain and logistics, and medical research. Companies like Cisco, British Telecom, Airbus, Novartis, and Boston Scientific are employing graph technology to enhance their operations. Infosys, an IT giant, uses a knowledge graph to consolidate information about their employees’ skills and career paths, enabling them to answer questions about career progression, skill utilisation, and project assignments. NASA utilises graph technology in their content management systems to perform semantic searches, explore related topics, and offer personalized recommendations based on the content. They combine natural language processing (NLP) techniques with taxonomy and domain knowledge to extract entities and annotate documents. Telecommunications companies like British Telecom use graph representations of their data to handle complex inquiries about network topology, identify critical equipment and analyse root causes of structural issues. Government agencies and police departments also employ graph technology to uncover illicit activities, establish connections, identify communities engaged in criminal behaviour, and detect potential risks such as radicalisation or terrorist networks. One of its examples is when it was used in high-profile investigations like the Panama Papers, where the International Consortium of Investigative Journalists (ICIJ) utilized Neo4j to visualise connections between individuals and expose instances of financial wrongdoing, leading to the resignation of public figures in multiple countries. Besides Neo4j Graph Database, other graph databases include Amazon Neptune, ArangoDB, DataStax, OrientDB, and Stardog among others. The graph database market is expected to grow by 22.5% to $5.1 billion by 2028. Read more: Meet the Undisputed King of Real-Time in Serverless Databases Born in India Neo4j has a long-standing relationship with India. It came into existence in 2007 when Emil Eifrem, the visionary founder and CEO of Neo4j, collaborated with an intern from IIT Bombay during a flight to Bombay, crafting the first piece of code. Now, the company looks to contribute more to India’s ever-evolving AI\/ML ecosystem, Webber told AIM that they are planning to have product specialists and research teams in India to leverage academic research and collaborate with institutions like IIT Bombay. “The Indian startup community is vast, hungry, and ever-growing and we see it as a major area of investment,” Webber commented. He added that India has a wide range of needs, from space programs to critical issues like food supply, where graph databases can play a role. India is seen as a promising market with immense potential to become a tech leader in the world. Its large population and technological talent make it a prime location for investment and growth. It has a diverse range of needs and a growing tech economy. In the past, the focus was on serving Western markets, but now there is a shift towards addressing domestic market opportunities. Echoing similar feelings, Barrasa said that their last year’s India experience in Bangalore with a knowledgeable and skilled audience of 500 people surpassed introductory sessions in Singapore and Australia in terms of expertise and advanced questions. What’s next? “We expect GQL (Graph Query Language) to be standardised by the end of the year,” said This is important because it will allow developers to build portable applications that are not locked into a particular vendor. Neo4j is already implementing GraphQL features, so it will be aligned with the standard by the time it is finalised. “Unlike other technologies that were absorbed into SQL, GQL recognises the distinctiveness and value of graphs, warranting its own query language. We have been working on GQL for years, drawing influence from Cypher Query Language,” concluded Webber. Read more: What’s Behind the Surge in Vector Database Investments?","excerpt":"Since its birth, Neo4j has been revamping how we use graph databases, making sure that we can harness its utmost potential","categories":["AI Features"],"tags":["graph database","neo4j"],"author_name":"Shritama Saha","publish_date":"2023-06-17T17:00:00","publication_year":"2023","word_count":906,"keywords":["graph database","semantic search","AI","ML","serverless","RAG","NLP","Aim","neo4j","SQL","R","fraud detection"],"extracted_tech_keywords":["AI","ML","NLP","Aim","RAG","fraud detection","semantic search","serverless","R","SQL"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/why-graph-databases-remain-untapped\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10101543,"title":"Adobe is Striking the Right Creative Chord With AI","content":"Adobe is best known for its line of products for visual creators. But with the latest portfolio of 11 features and tools announced at the company’s Adobe Max event, it appears the design giant is confidently tapping the zeitgeist collectively by catering to its creative fans for software and even hardware. Adobe’s visual tools like Photoshop, Premiere Pro, and Acrobat are familiar to nearly everyone. But Adobe is also making its way into another business — AI-powered wearable. The company introduced Project Primrose, an interactive dress that demonstrates the potential use of “flexible textile displays,” allowing the wearer to display patterns and images on their body like a programmable screen. While Adobe has technically already teased this “smart display fabric” technology before, we’ve only previously seen it applied to a flat canvas and a small handbag. As a dress, the numerous scale-like displays look cool but nowhere near practical. “In an unpredictable economic climate, where consumers now re-evaluate the products and services they buy each day, a brand’s key growth driver is the ability to show people you accurately understand their current needs,” Anjul Bhambhri, senior vice president of Adobe Experience Cloud platform engineering, said. On the text side of the equation Adobe introduced three new services based on LLMs. The recently introduced Adobe Experience Manager, Adobe Journey Optimizer, Customer Journey Analytics and Marketo Engage users will be the first to be able to take advantage of Generative AI Services, thanks to latest native workflow integrations. Firefly Upgraded The IT juggernaut also released the second generation of its image generator, Firefly Image 2 model that powers popular features like Photoshop’s Generative Fill — alongside two fresh Firefly models — creators of vector images and design templates. Adobe says the model’s latest version stands out as a virtuoso in image generation. It delivers higher-quality images and understands the nuances of high-frequency details better than its predecessor. The model can photorealistic images perfecting vivid colours and near-to-perfect photorealistic images. Changing the face of design the service provider is banging the AI drum really hard. At the core of its approach is Firefly, its inhouse bundle of generative AI models. Since its release the company has injected the model(s) trained on Adobe Stock images, along with public domain and copyright-free images in most of its offerings. Just like the original text-to-image model, Adobe says its Firefly Vector model is designed to be safe for commercial use. The Image 2 is available to try via the web-based beta and will come soon to Creative Cloud apps as well. (Water)marking its Territory Adobe has also vowed to add metadata to AI-generated images, a digital signature of authenticity. In the near future, AI-generated images will include cryptographically signed data, allowing a quick distinction between AI and human-made images. Adobe said it has the Content Credentials cloud, a vault for your image files’ metadata. If an image is shared without its unique metadata, Adobe’s cloud will show the missing signature. “Once digital content is signed with Content Credentials (in a platform that has leverages the C2PA open-standards and Content Credentials), tamper-evident metadata is attached, so it travels with the content wherever it goes,” a spokesperson for Adobe told The Register. As Adobe continues to power its tools with more AI possibilities, it’s leaving behind a slow period. While many tech companies have big, unclear ideas about AI, Adobe is focusing on practical uses that its existing user base of millions really like. They’ve been working on Firefly to make images, and now it’s even better in its second version. They’ve also added new features for making sound, video, and 3-D pictures. It’s like Adobe is an artist, painting a clear picture while others are still dreaming in fuzzy colors.","excerpt":"Adobe announced everything at Max 2023 from AI models to a digital dress","categories":["Global Tech"],"tags":["AI Tool"],"author_name":"Tasmia Ansari","publish_date":"2023-10-16T20:55:16","publication_year":"2023","word_count":622,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","Git","RAG","generative AI","analytics","AI Tool","R"],"extracted_tech_keywords":["AI","analytics","generative AI","RAG","R","Go","Git","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/adobe-is-striking-the-right-creative-chord-with-ai\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":11760,"title":"The rise of Data Lake and how India is playing a key role in it","content":"Data Lake has gained popularity for its flat architecture It’s a term that probably rings a bell with storage and data analytics practitioners and is right up the alley for people working in Big Data technology platform Hadoop also. However, Data Lake is not a new concept but a storage mechanism that has gained currency over the last two years. And it is not entirely an offshoot of a Hadoop oriented storage \/ repository that gathers data from enterprise applications. But first let’s define our term. For starters, in tech jargon, Data Lake is an object based repository that stores large amount of raw data right from structure, semi-structured and unstructured data in its native format. As opposed to a data warehouse that stores data in files or folders, Data Lake is marked by its horizontal flat architecture where each data element is assigned a unique identifier and is tagged with extended metatags. Many tech bloggers have attributed tagging methodology being similar to that of Twitter. What is the major differentiator for Data Lake and Enterprise Data Warehouses (EDW)? The major differentiator between Data Lake and EDW is that a data lake is fed data in its absolute raw, native form straight from data sources without any standardization, remodeling and alteration. Which means that raw data provides a number of ways to be queried, stored, derive insights from all types of data as opposed to EDW where data has to conform to its own predefined schema. Since it has to conform to an enterprise data model, EDW is capable of answering only a limited number of questions. Simply put – Data Lake supports all data types, stores all data and provides faster insights. Hence, it has a horizontal scalable framework that can process all variety of data. What led to the rise of Data Lake? The term Data Lake had its first day in the sun when James Dixon, CTO of Pentaho, Florida-based company known primarily for its business analytics suite of open source products first mentioned in his now much-famed blog entry. For the interest of our readers, we have reproduced it here, “If you think of a datamart as a store of bottled water – cleansed and packaged and structured for easy consumption – the data lake is a large body of water in a more natural state.” The analogy hit the mark with most early Hadoop adopters. Dixon found out later in his interactions with industry veterans deploying Data Lake, that it made it possible for enterprises in making the data more accessible and useful. What most industry insiders believe is that Data Lake helps in tackling spiraling data volumes, helps gain new business insights by storing large amounts of data in the chosen format, and then make it easy for processing through big data analytics. The trend is also spurred on by cloud computing which makes it possible for companies to build data lakes at a huge scale. Why Data Lake is well-received? Stores a variety of data Evolves data management architecture Instead of supplanting it augments traditional EDW strategy Hadoop and Data Lake are a marriage made in heaven While Data Lake can definitely be aligned with other relational database architecture, its gained popularity with Hadoop primarily because of Hadoop is an open source platform and according to Hadoop adopters, it provides a less expensive repository for analytics data. Another bonus point was the fact that a Hadoop Data Lake architecture can also be used to complement an entire data warehouse rather than supplant it entirely. Moreover, information from a Hadoop Data Lake can be analyzed in its raw format and data can be extracted and processed through MapReduce, Spark and other data processing frameworks. Data Lake as a Service in Indian ecosystem One of the biggest Indian companies to employ Data Lake as a Services is HCL.  HCL’s data lake offering comprises processing huge volumes of raw data from disparate sources, low cost of storing data and reducing downtime with preventive techniques. 1) HCL’s Data Lake application has been seen in the aviation industry for an American aircraft manufacturer providing preventive maintenance and reliability by finding areas where huge data sets could be optimized and manage it as part of Operations & Performance Analytics Group. Another use case is for a leading air fleet manufacturer where Predictive analytics was applied for better engine health. The challenge was to find out factors affecting engine health that led to huge losses for the fleet manufacturer. With predictive maintenance analytics, engine health was enhanced by 25%. 2) Another sector where Data Lake mechanism has been put to work is Retail banking, bringing together data from varied sources. Persistent Systems has offices spread across the globe and has a branch in Bangalore. It provides Data Lake implementations and architecture to various sectors such as Retail, Banking and healthcare. The following use case is from Retail Banking where the lack of agility of a traditional data warehouse and data silos was overcome by providing a common data model through Data Lake. The Data Lake helped in KYC providing a complete overview of customer profile, their spending and saving patterns and helped in customer segmentation. End result was: Helped in optimizing where the bank’s resources should be spent Built a solution to better understand customer issues which were not being funneled down the channel Enterprises looking to invest in data lakes architecture and implementation should adapt the architecture to their specific industry, and look for co-existing Data Lake alongside enterprise data warehouse. One of the key takeaways is to define Data Governance capabilities as well.","excerpt":"It’s a term that probably rings a bell with storage and data analytics practitioners and is right up the alley for people working in Big Data technology platform Hadoop also. However, Data Lake is not a new concept but a storage mechanism that has gained currency over the last two years. And it is not […]","categories":["IT Services"],"tags":[],"author_name":"Richa Bhatia","publish_date":"2016-12-23T05:51:25","publication_year":"2016","word_count":935,"keywords":["big data","Go","cloud computing","AI","R","Scala","RAG","analytics","predictive analytics","data warehouse"],"extracted_tech_keywords":["AI","analytics","RAG","predictive analytics","cloud computing","R","Go","Scala","big data","data warehouse"],"url":"https:\/\/analyticsindiamag.com\/it-services\/rise-data-lake-india-playing-key-role\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":49511,"title":"Why Softbank Debacles Will Not Harm Its Vision Statement For Innovation","content":"According to experts, the massive bets of Softbank have played a crucial role in driving innovation at scale for young companies which are usually devoid of capital in the early phase. The startup innovation and the corresponding funding has been shifting rapidly over the years. Since the 2000s, when startup funding was focused mostly on cloud computing, social networking and mobile computing, many changes have taken place. Today, the money is pouring into things like artificial intelligence, blockchain, computational biology, and FinTech — all of which have been disrupting technology products. Another major focus of tech funding is to create platforms on which developers build products, aiming to aggregate things including rental housing, cabs, jobs. The software industry has caused a wealth transfer from upper class to entrepreneurs by making it easy to raise millions and potentially billions of dollars, both in developed countries like the US as well as fast-developing markets like India and China. But successful investing in startups isn’t an easy task, because in the early stages, there are not many metrics and no historic background for the company to measure. There may be a few people who have come together and built something. They may have an MVP and a few paying customers, and their product may be in beta. Fewer companies go public today due to regulatory complications as well as constant scrutiny from investors which can impede innovation. For these reasons, private equity is the way tech startups choose. Using private funds, startups have amassed a lot of wealth. One company which is leading the way is Softbank with its enormous chest of cash to fill the void of the public raising of capital. Over the past decade, it has aided numerous emerging tech startups, pouring in billions and billions of dollars as part of its Vision Fund — the world’s largest stash of cash for early-stage investing. Eating Up The Market Share In Emerging Tech Startups Among the ongoing startup wave, India alone raised $12 billion last year and the amount of capital attracted by early-stage tech startups was decent, third in the world behind US and China. Here too, Softbank led the way with massive investments in companies including: PaytmGrofersDelhiveryPolicy BazaarOla ElectricSnapdealOYOFlipkart While some analysts have been sceptical about Softbank’s model, stating it’s not sustainable, others are optimistic about it and urge to see a long term picture. Regardless, the company has become a venture capital behemoth, which is not just eating up market share in the emerging tech space but also making its investors decent returns over the years. On the other hand, the company has witnessed recent debacles for lack of growth like WeWork (which has seen a bailout worth $9 billion), and perhaps issues with potentially overvalued companies like Uber and India-based Oyo. Analytics India Magazine got in touch with Siddarth Pai, Founding Partner of 3one4 Capital — one of India’s largest early-stage tech funds to understand what’s happening with Softbank. “We live in an age where capital is used as a weapon. The first person to weaponise capital was Masayoshi Son from Softbank. Its vision statement doesn’t much consider what is happening in the short term but has an outlook for the next 100 years and that’s how big companies will emerge out of it,” said Pai. Why Softbank’s Risky Bets Are Driving Innovation At Scale According to experts, the massive bets of Softbank have played a crucial role in driving innovation at scale for young companies which are usually devoid of capital in the early phase. Sure, some bets will fail but if we really look at the vision statement for bringing about the information revolution, the current debacles of Softbank may seem like small dents 100 years down the line. “The goal is to get the fund to find the best startups and help them with more than enough capital to get them to succeed. Has that thesis worked? Some people say no and I would say it’s still a work in progress,” Pai added. Softbank’s vision aims to develop over the long-term by forming partnerships with the most superior companies at the time in the information industry, without adhering to particular technologies or business models. Here, Masayoshi Son’s experimentation and diversification with new technologies and models may reap variable results, particularly given the ups and downs of the global economy.","excerpt":"According to experts, the massive bets of Softbank have played a crucial role in driving innovation at scale for young companies which are usually devoid of capital in the early phase. The startup innovation and the corresponding funding has been shifting rapidly over the years. Since the 2000s, when startup funding was focused mostly on […]","categories":["AI Features"],"tags":["Softbank"],"author_name":"Vishal Chawla","publish_date":"2019-11-06T17:03:09","publication_year":"2019","word_count":725,"keywords":["Go","API","artificial intelligence","AI","cloud computing","innovation","Aim","analytics","R","Softbank","startup"],"extracted_tech_keywords":["AI","artificial intelligence","analytics","Aim","cloud computing","R","Go","API","innovation","startup"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/masayoshi-son-softbank-vision-fund\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10163660,"title":"AI Bridges the Gap in India’s Insurance Market","content":"In India, several communities face significant challenges in accessing affordable insurance. As of 2020, only about 18% of the eligible population subscribed to pure retail term insurance offerings, with protection penetration at approximately 12%. Over 40 crore individuals in India remain uninsured. This number highlights the persistent health protection gap. This gap is particularly evident in rural areas, where nearly 90% of the population lacks insurance coverage and has to incur high healthcare expenses. The urban poor are similarly disadvantaged, often excluded from affordable health insurance markets due to factors like illiteracy and poverty, which limit their access to information and resources. This has changed only slightly, as insurers are adopting predictive tools to assess affordability for users. This highlights the pressing need for innovative solutions to enhance insurance accessibility and affordability for marginalised communities in India. To understand this better, AIM spoke with Affan Mohammad, Principal Consultant – Insurance at Fractal, India’s first AI unicorn. With his years of experience in the financial services and insurance industry, Affan said that while the industry has always faced several challenges, AI may provide insurers with several solutions for finding innovative ways to balance risk assessment and affordability. “At the heart of AI-driven insurance pricing is data,” Affan explained. AI’s ability to process vast amounts of data surpasses traditional models. Insurers can now identify risky patterns both at macro and micro levels. For instance, by analysing regional risks at the zip code level, insurers can set up pricing models for specific population groups. “Micro-segmentation plays a crucial role here,” he added. By moving away from generalised pricing buckets, AI enables insurers to offer premiums that align with the unique risks posed by individuals or communities. This approach ensures fair and affordable coverage for low-income groups and empowers them to access insurance without overpaying for premiums or coverage they may not need. The Role of Explainable AI Transparency is critical, especially when catering to underserved populations. Affan highlighted how explainable AI bridges the gap between insurers and customers. “For end consumers, explainable AI provides clarity on why certain premiums or coverage options are offered or denied,” Affan said. This transparency builds trust and helps people understand how factors like previous defaults or financial behaviours might influence insurance premiums. It incentivizes people to improve their future decisions and practices to be eligible for better products and pricing in the future. Moreover, explainable AI prevents insurers from being biased in pricing models by checking for compliance with regulatory standards. While regular audits and diverse training data are important, collaboration with regulators also guarantees that AI models are fair and unbiased. Personalisation Through Real-Time Data The integration of real-time behavioural data has opened new avenues for hyper-personalisation in insurance. These include advancements in telematics, wearable technology, and smart home systems. “These devices collect data on driving habits, fitness levels, or property risks, which can be analysed using AI,” Affan said. Reinforcement learning, a key AI training technique, allows dynamic pricing based on evolving behaviors. It is a way through which premiums can reflect real-time risks, taking into account individual and population level changes. Affan, however, acknowledged the ethical challenges of using such data. “It’s a delicate balance…Transparency, informed consent, and regular reminders about data usage are vital to maintaining consumer trust.” Designing Sustainable Micro-Insurance It is crucial for micro-insurance products to be affordable and sustainable. Affan discussed the role of AI-powered agentic workflows for underwriting automation and granular risk segmentation. “AI can design innovative products with tailored coverage, payment plans, and limits that meet customers’ needs…It can also optimise distribution channels, making insurance products more accessible through online platforms or mobile apps.” While everyone is trying to figure out how assistants and agents can help their industry, Fractal has already implemented multi-agent systems for the insurance and underwriting part of their offerings. Onil Chavan, client partner, insurance practice and global capability head at Fractal, earlier discussed with AIM how the company uses these multi-agent systems to reshape underwriting processes and drive efficiency in the insurance domain. Affan also said that such innovations are especially vital in markets like India, where insurance penetration remains low. “Expanding digital access and simplifying the process of purchasing insurance will drive adoption and reduce financial vulnerabilities for underserved populations,” he added. Balancing Profitability and Affordability Insurers often face the challenge of balancing profitability and affordability. AI offers a solution by allowing granular segmentation of risk and affordability groups. “We can simulate various scenarios to find the optimal balance between customer retention and profitability,” Affan further said. AI also enables insurers to price high-risk customers optimally while offering competitive premiums to low-risk individuals. “Affordability and profitability aren’t mutually exclusive,” he asserted. “With the right strategies, insurers can achieve both.” By bringing together advanced analytics, explainable AI, and real-time data integration, insurers can enhance transparency, personalisation, and accessibility. “AI-driven innovation is not just about profitability; it’s about building trust and inclusivity in insurance,” Affan concluded.","excerpt":"Nearly 90% of the population in rural areas lacks insurance coverage and has to incur high healthcare expenses.","categories":["AI Highlights"],"tags":["AI in insurance"],"author_name":"Mohit Pandey","publish_date":"2025-02-15T12:00:00","publication_year":"2025","word_count":822,"keywords":["AI in insurance","AI","agentic workflows","Git","RAG","Aim","multi-agent systems","analytics","ViT","Rust","R"],"extracted_tech_keywords":["AI","analytics","agentic workflows","multi-agent systems","Aim","RAG","R","Rust","Git","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/ai-bridges-the-gap-in-indias-insurance-market\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10095837,"title":"Freshworks Adds Generative AI Capabilities To Supercharge Productivity","content":"Freshworks is the latest cloud-based SaaS giant to add generative AI capabilities to their platform through three new services – Freddy Self Service, Freddy Copilot and Freddy. The company is also planning to develop proprietary language models and integrate general-purpose LLMs to cater to specific customer requirements. The new predictive and assistive generative AI capabilities embedded within Freshworks solutions and platform help support agents, sellers, marketers, IT teams and leaders maximise productivity. Freshworks is leveraging Microsoft Azure OpenAI Service to ensure the privacy and security of customers’ data. Freddy Copilot: Freddy Copilot aims to enhance productivity for support, sales, marketing, and developers by facilitating faster workflows through interactive prompts within Freshworks products. It enables users to perform their tasks efficiently and develop new applications to expand their capabilities. During beta testing, Freddy Copilot was adopted by 390 companies and reduced effort by up to 83%. Over 2,500 developers who are already using the Freshworks Developer Platform can now leverage Freddy Copilot to create innovative, high-quality apps at a quicker pace. Freddy Self Service: Freddy Self Service empowers companies with the technology to deliver personalized automation on a large scale. It leverages Freshworks’ platform and a powerful language model to enable personalized automation that enhances agent productivity. By utilizing extensive language and account-specific models, Freddy Self Service handles a significant portion of L0\/L1 queries from employees and customers within Freshdesk and Freshservice, delivering customized responses. This allows customer support and IT personnel to dedicate their efforts to higher-value projects and tasks. Freshworks is hailed as the core of IQor’s digital customer and employee universe, according to Sergey Kolosovosky, IQor’s Senior VP of Application Development and Solutions. With 40,000 employees benefiting from Freshworks Freddy AI capabilities, IQor is excited to equip them with tools for daily functions and enhance engagement. The focus is on creating happy employees and delighted customers, as access to powerful generative AI software becomes a game changer for IQor. Freddy Insights: Freddy Insights enables businesses to streamline operations and foster business growth. Freshworks’ generative AI analyzes customer and employee support data to automatically identify areas that require improvement. It also evaluates marketing and sales effectiveness and provides recommendations for optimizations that can enhance performance and boost revenue. Freddy Insights additionally offers proactive quality management, assessing support quality and ensuring that staff members meet established goals. It guides agents in improving their skills through every customer interaction. Freshworks’ AI & Analytics Play Back in March, Freshworks announced enhancements to Freddy AI powered by OpenAI’s ChatGPT and LLM GPT-4 consisting of a conversation summariser, rephraser, autocomplete, article generator, and email copy generator. Freshworks was one of the first B2B companies to leverage cutting-edge tech. In 2018, they launched Freddy AI, an AI assistant powered by their Neo platform, which leveraged Google Contact Center AI. In 2020-21, they trained their models using customer-agent conversations and improved their bot builder. Read more: How Whatfix is Revolutionising SaaS with GenAI Integration","excerpt":"The company is also planning to develop proprietary language models and integrate general-purpose LLMs to cater to specific customer requirements.","categories":["AI News"],"tags":["AI Tool","Freshworks","Generative AI"],"author_name":"Shritama Saha","publish_date":"2023-06-27T17:43:46","publication_year":"2023","word_count":488,"keywords":["ChatGPT","GenAI","OpenAI","AI","ML","Freshworks","RAG","Aim","analytics","generative AI","Generative AI","AI Tool","Azure"],"extracted_tech_keywords":["AI","ML","analytics","generative AI","GenAI","ChatGPT","OpenAI","Aim","RAG","Azure"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/freshworks-adds-generative-ai-capabilities-to-supercharge-productivity\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10041053,"title":"Python Creator Guido Van Rossum Reviews Popular Programming Languages","content":"The former Benevolent dictator for life, Guido Van Rossum expressed some views about programming languages in a live question-answer session with Francesca, Microsoft’s principle and advocate manager. The founder of Python had a tell-all interview about his programming career that started at the University of Amsterdam. In the interview, the Python founder told how he draws inspiration from other programming languages and works on his projects to evolve them. He also elaborated on his first career project in life the ABC language that he said flopped because of no distribution channels which eventually led to no users. He also expressed how he has not been able to adapt to a different language despite his love for languages because whenever he tries to implement the language he ends up saying “I already know how to do that in Python”. It’s like Python is his mother language. Guido Van expressed his views on four languages and their performance in comparison to Python. He also told the viewers what are the reasons that make these languages interesting. Why is Rust an interesting language? Rust overtook Python because of its incredible performance. Rust’s ability to be compiled directly into machine code gives it an edge over other languages. Van Rossum believes that the language has a serious edge over C++ in some areas like memory allocation. He also believes it has an advantage in memory allocation, “And of course, it solves the memory allocation problem in a near-perfect way”. Van Rossum believes that when a code is written in Rust a person can be sure of memory allocation and memory management stuff that makes a language interesting for him. Which language is most Pythonic? According to Van Rossum, Go is the most Pythonic language out of all general-purpose languages. Go, short of Golang, is mostly used when working with servers. Python and Go both have good readability and are beginner-friendly. Go was in fact created to improve Python’s readability. It is faster than Python due to its fast speed in compiling data. Julia: the better compiler? Van Rossum suggested that the significant upside of using Julia over Python is that its compiler optimises code in a way Python never does. The one-based indexing and inclusive ranges instead of exclusive make it very different from Python though initially, they might appear the same. He believes that Julia is more of a niche language, designed for specific purposes and limited in other areas. He also went on to say that he would never expect someone to write a web server code in Julia and get a lot of mileage out of it as that will end up soon on Hacker News when someone tinkers with the technology. How is Python learning from TypeScript? Van Rossum initially didn’t believe that TypeScript was a great language as he felt it jumped on the JavaScript bandwagon. Though he was unaware of TypeScript back in the day today they are adding static typing\/gradual typing to Python. In his opinion, TypeScript has achieved a few things that Python has still been waiting to figure out and it looks up to TypeScript for examples and areas to improve. He went on to say that they have added some features in Python that originally lacked in TypeScript but were added looking at demand and became immensely popular. He said that his frequent conversations with Danish engineer Anders Hejlsberg suggest that TypeScript learns as much from Python as learns from it. He compared it with the early years of JavaScript when it learned from Python. Each of the languages that Guido Van Rossum reviewed is incredible for singular purposes like Julia is great for mathematical and technical tasks, Go is unmatched in molecular programming, Rust is amazing for systems programming. Python is an undisputed most popular programming language that has made many ML-based programs possible. Python readily achieved its goal and became a toolkit of programmers who would be writing programs in C or C++. Though having many upsides the language has been problematic as it is slow and requires excessive testing. It produces runtime errors despite testing. It is also weakly typed and dynamically typed which puts it at a disadvantage when compared to other statically typed and strongly typed languages like Julia and Go.","excerpt":"The former Benevolent dictator for life, Guido Van Rossum expressed some views about programming languages in a live question-answer session with Francesca, Microsoft’s principle and advocate manager. The founder of Python had a tell-all interview about his programming career that started at the University of Amsterdam.  In the interview, the Python founder told how he […]","categories":["AI Trends"],"tags":["Go","Microsoft","Python Programming","review","Rust","TypeScript"],"author_name":"Meenal Sharma","publish_date":"2021-05-31T14:00:00","publication_year":"2021","word_count":712,"keywords":["Go","AWS","AI","ML","JavaScript","TypeScript","Python","Rust","review","Python Programming","R","Java","Microsoft"],"extracted_tech_keywords":["AI","ML","AWS","Python","R","JavaScript","TypeScript","Go","Rust","Java"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/python-creator-guido-van-rossum-reviews-popular-programming-languages\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":58407,"title":"World&#8217;s First AI University: MBZUAI","content":"In a first, UAE academic authority accredited an artificial intelligence university that will start the classes in September. The institute — Mohamed bin Zayed University of Artificial Intelligence (MBZUAI) — got the licence for offering Master of Science and PhD programmes to close the skill gap in the market. The university has already received 1,200 applications, and interested graduates can still apply through its website. All the admitted students will be given full scholarship along with other benefits such as a monthly allowance, health insurance and accommodation. The board of the university has also decided to work with local and global companies to ensure internships and will assist students in finding employment opportunities. “We aim to provide students with world-class courses in an environment that exceeds expectations in order to attract the best talent from across the world,” said Dr Sultan AI Jaber, Minister of State and Chairman of the board of trustees at MBZUAI. Since AI is expected to add 14 per cent to the world’s gross domestic product by 2030, equivalent to $15.7 trillion, it is of paramount importance to take the lead in the market. Consequently, the university has planned to engage with policymakers and businesses to ensure they adopt best practices. The UAE government has been one of the early adopters of artificial intelligence and continues to understand the potential of technology. This has led the government to roll out AI strategy for 2031, outlined the plans on how artificial intelligence will make the government efficient. The government also pinpointed eight sectors that it aims to transform, including space renewable energy, water, and education. The country had also appointed the world’s first Minister of State for Artificial Intelligence, Omar AI Olama.","excerpt":"In a first, UAE academic authority accredited an artificial intelligence university that will start the classes in September. The institute — Mohamed bin Zayed University of Artificial Intelligence (MBZUAI) — got the licence for offering Master of Science and PhD programmes to close the skill gap in the market. The university has already received 1,200 […]","categories":["AI News"],"tags":[],"author_name":"Rohit Yadav","publish_date":"2020-03-11T14:45:49","publication_year":"2020","word_count":285,"keywords":["Go","artificial intelligence","programming_languages:R","AI","programming_languages:Go","Aim","Rust","R","programming_languages:Rust"],"extracted_tech_keywords":["AI","artificial intelligence","Aim","R","Go","Rust","programming_languages:R","programming_languages:Go","programming_languages:Rust"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/worlds-first-ai-university-mbzuai\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10068324,"title":"One-of-a-kind hackathon: #CryptoPrediction in the spotlight","content":"Participate in a hackathon – ✅ Find the solution – ✅ Win prize money\/attractive gifts – ✅ Sounds boring? Let’s make it interesting then. What if you get to work on a real-world crypto dataset that can impact financial markets? Gear up for an exciting and innovative hackathon coming your way. Rocket Capital, in association with MachineHack, brings a hackathon to source and incentivise the best in machine learning applications for finance. A key need of RCI is good financial market predictions to improve the accuracy of the ML models. RCI wants to capture skilled people who can predict solutions to complex problems that exist in the financial markets, in return for which it wants to reward the problem-solvers. This hackathon keeps getting interesting. The community competition gets 52,000,000 tokens, released weekly in the next 5 years, and up to 200,000 tokens and 500 USDT prize (the crypto equivalent of USD) weekly. Now, what are MUSA tokens exactly, you might ask. The MUSA token is the first token to be fully dedicated to the data scientist community. Decentralised and public Rocket Capital Investment (RCI) is a Licensed Financial Institution headquartered in Singapore with a Decentralised and Distributed Intelligence. Through this competition, data scientists & ML practitioners can now be rewarded weekly on the blockchain for their data science skills. This can also help them create a decentralised track record of their performance. Rocket Capital Investment asks for weekly financial predictions. Every week RCI distributes a dataset; the participants send their predictions that are encrypted and stored in blockchain so that the predictions cannot be tampered with afterwards. At the end of the week, RCI decrypts predictions, compares forecast and market results, evaluates participants’ performance and pays rewards. What is interesting is that all the process is managed in a decentralised way on the blockchain and is public. The blockchain complexity is hidden by a dApp used by the participants to send predictions and get rewards. RCI created the MUSA token needed to participate in the competition. The tokens are initially distributed to selected participants (airdrop) and MUSA tokens are expected to be publicly listed by the end of the year. Starting on June 13 and concluding on September 5, this hackathon is a perfect opportunity for analytics professionals to prove their mettle and showcase their technical know-how to solve challenging problems. Show Rocket Capital how your data science & ML skills can give the team an extra edge. Register here. You simply cannot afford to miss this opportunity. So sign up for this innovative hackathon and test your capabilities!","excerpt":"Rocket Capital Investment wants to capture skilled people who can predict solutions to complex problems that exist in the financial markets, in return for which it wants to reward the problem-solvers.","categories":["Deep Tech"],"tags":["crypto india","Cryptocurrency","cryptocurrency bitcoin","cryptography","Hackathon","hackathon for data scientists"],"author_name":"Sreejani Bhattacharyya","publish_date":"2022-06-03T15:00:00","publication_year":"2022","word_count":429,"keywords":["data science","Go","API","machine learning","programming_languages:R","Cryptocurrency","AI","crypto india","ML","programming_languages:Go","cryptography","Hackathon","cryptocurrency bitcoin","analytics","hackathon for data scientists","R"],"extracted_tech_keywords":["AI","machine learning","ML","data science","analytics","R","Go","API","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/one-of-a-kind-hackathon-cryptoprediction-in-the-spotlight\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":66372,"title":"6 Advanced Skills That Will Get Data Scientists Hired In The Post-COVID World","content":"The COVID-19 pandemic outbreak has urged companies to transform their strategies in order to have business continuity in the post-lockdown world. This, in turn, provided opportunities for data scientists to upskill\/reskill themselves with relevant skill sets to keep up their relevancy. Therefore, it has become imperative for data science professionals to rethink their career strategies. Alongside the automation boom, several data science skills sets are getting obsolete for business outcomes. Consequently, companies are currently looking to hire professionals with knowledge of advanced skill sets, which would be relevant for businesses in the post-COVID-19 world. In fact, a recent LinkedIn survey stated that amid the challenging times, data science professionals are cautiously optimistic about job opportunities. However, 63% of respondents who are active job seekers, 65% of full-time employees, and 61% of self-employed professionals, have stated that there will be an increased dependency on upskilling to survive this crisis. Also, Zairus Master, CEO of Shine Learning, an online learning platform for professional courses, stated to the media that the company had seen a sharp rise in the number of enrollees opting to upskill themselves in courses like data science, blockchain, and machine learning. “… With remote working increasingly becoming the norm, we may see more demand for specific talent in this field,” said Master. To survive the post-COVID world, data science professionals need to learn some advanced skills that would make them hireable post this pandemic, here are a few skills that organisations would look for: Natural Language Processing (NLP) With businesses aiming to collect the maximum amount of data these days to have a better customer service amid this crisis, NLP could be an asset for companies. Many companies are also deploying self-service tools like bots that come with NLP integrated with multiple languages to address customer issues better. Data scientists must understand and master NLP to help businesses make automated solutions for a better outcome post-COVID world. In fact, in recent news, Mercedes-Benz, a German global automobile maker, announced that the company has been developing deep learning (DL) and natural language processing (NLP) technologies to improve the driving experience. “By automating responses from gauging the needs and moods of the driver, the company’s smart technology has enabled a unique driving experience,” stated in the media. Computer Vision Another technology that would massively gain traction post-COVID world would be computer vision because of its ability to identify patterns in a huge volume of images. The usage of facial recognition technology can become the norm post the pandemic. In the current state, this technology of computer vision is heavily being used for maintaining social distancing and monitoring people to wear masks. For this, the technology has been deployed in urban places for monitoring citizens by video surveillance, driving contactless screening, and tracking employee activities in organisations. In recent news — RayReach Technologies, a Chennai-based AI startup has been using this technology to create smart CCTVs for its customers to detect violations in social distancing. Therefore, gaining expertise in computer vision technology will make data scientists more hireable in the post-COVID world where social distancing could be the norm. This technology also has great potential in the healthcare industry, wherewith the help of deep learning, doctors can detect anomalies in chest radiography images. With companies actively using this technology to enhance their workflows, it becomes critical for data scientists to learn this subset of artificial intelligence to sustain the market post-pandemic. Also Read: 10 Popular Computer Vision Projects For Beginners Geospatial Technology During the COVID-19 pandemic, geospatial technology came in handy for companies and governments across the world that were using geographical data to map the disease outbreak and also tracking the infected areas of the country. With more people working on data-driven decision processes, geospatial data has been helping in better planning and processing of the system. The technology can also help manufacturing and retail industry to identify the density zones of the disease during this lockdown. In fact, in recent news — geospatial experts, Transerve Technologies has announced their offering of Transerve Online Stack that works on geospatial data to map COVID-19 density zones. According to the company, this technology has been helping SMBs, large retailers and supply chain professionals in coming out strong and re-starting their business in the post-COVID world. And therefore, companies are looking to hire professionals who have expertise in working with geospatial data. A geospatial data scientist would be required to analyse vast geospatial datasets, which would include images from satellite, google maps, demographic data, socio-econometric data and topography to create insights that would help businesses gain profits. Data Storytelling Another aspect that would gain popularity among businesses would be the ability of compelling data storytelling by data science professionals. With data analytics becoming a prime concern for companies across industries, there has been an increased requirement for good data storytelling. Such an ability would make a data scientist more favourable among business leaders, as compelling storytelling would ease up the understanding of complex numbers and statistics for stakeholders and leaders. With good storytelling, data scientists can easily convey complex business information to their audience, which in turn would make them stand out from other professionals in the market. This also enhances the chance of being more sustainable in the post-COVID world. According to experts, “We can recall facts up to 22 times more effective when they are part of a story as compared to when they are presented as isolated data points. Information presented in the form of a story also helps to reinforce our understanding and decipher patterns in complex data sets.” As data storytelling would help business leaders with powerful insights, it would, in turn, help them capitalise the market opportunities in the post-pandemic world. In fact, Pavan Kumar Thatha, who is an emerging technologies leader at Unisys, once stated in his interview that one could not become a data science professional without mastering the storytelling skill. Explainable AI With AI penetrating almost every aspect of human lives, it became imperative for businesses to be able to trust these machines and their decisions, and that’s where the need for explainable AI arises. Enterprises are currently looking to deploy AI models that can predict accurate insights along with providing explanations of those predictions. Therefore, data scientists who are well aware of explainable AI become a more favoured choice for businesses. For many years the prime concern had been the accuracy and the performance of AI models, however with the recent reluctance of accepting AI by many companies has brought in the requirement for transparency in the AI models’ predictions. In fact, according to a survey, 82% of business leaders agreed that in order to trust artificial intelligence, it needs to be well explainable. To join in the effort, in recent news, Microsoft has announced a strong emphasis on developing tools for building more responsible and fairer AI systems. Also, software development company Pegasystems has announced their tools that would help companies in identifying and omitting the hidden biases of their AI models. According to the VP of the company — “As AI is being embedded in almost every aspect of customer engagement, certain high-profile incidents have made businesses increasingly aware of the risk of unintentional bias and its painful effect on customers.” Consequently, in the post-COVID world, businesses would be looking to hire professionals who can govern the operations of their AI systems and can create explanations of the predictions. Skills Related To Healthcare Domain Post-COVID, the world would experience a massive demand in healthcare technologies, and therefore there would be a huge search for data science professionals who have expertise in the healthcare domain. Data science proved to be immensely beneficial for the treatment of COVID patients by leveraging patients’ data, however, without fundamental domain knowledge, these predictions could be inaccurate and can also have an adverse effect on developing drugs and medicines. Even after the pandemic, healthcare would continue to be a prime concern for governments, businesses and individuals. Therefore, there would be a huge surge in demand for data professionals who can use real-time medical data to generate insights and help doctors make informed decisions. In fact, as the pandemic outbreak has been going on, China’s tech giant, Baidu has released their Linearfold algorithm in order to help medical researchers to fight against the virus. Similar to this, many other companies have also developed tools to help doctors fight this crisis. Therefore, data scientists with advanced skill sets and knowledge in the healthcare domain would be more relevant to companies post-COVID world.","excerpt":"The COVID-19 pandemic outbreak has urged companies to transform their strategies in order to have business continuity in the post-lockdown world. This, in turn, provided opportunities for data scientists to upskill\/reskill themselves with relevant skill sets to keep up their relevancy. Therefore, it has become imperative for data science professionals to rethink their career strategies. […]","categories":["AI Trends"],"tags":["automated data science solutions","Data Science","Data Science Career","Data Science Jobs","Data Scientist","datascience","deep learning projects","facial detection software","post covid world","red hat"],"author_name":"Sejuti Das","publish_date":"2020-06-02T11:00:00","publication_year":"2020","word_count":1422,"keywords":["red hat","computer vision","facial detection software","automated data science solutions","Ray","deep learning","data science","artificial intelligence","NLP","analytics","Data Science","machine learning","AI","datascience","Data Science Jobs","Data Science Career","post covid world","Aim","deep learning projects","Data Scientist"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","deep learning","NLP","computer vision","data science","analytics","Aim","Ray"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/6-advanced-skills-that-will-get-data-scientists-hired-in-the-post-covid-world\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":65230,"title":"What Is The Hiring Process Of Data Scientists At Freshworks?","content":"With data science emerging as a popular career choice in the 21st century, the hiring market in the space has become increasingly competitive. A cursory search on popular job portals will tell you that there is a large pool of existing and aspiring data scientists in India alone. This may indicate that the supply of data scientists should match the growing demand for these positions. However, the ground reality tells another story. Data science is a dynamic field that demands constant learning, and available talents often fall short of the requirements companies seek. What is more, the applicant pool shrinks even further given that many companies are concurrently hiring for the same skill sets. Cloud-based software company Freshworks has often found themselves in a constant tussle with other organisations to find the best data science talent, and its carefully curated team is testament to the fact that many of the top talents in the market see value in working here. Acquiring data science talent, albeit challenging, has been a rewarding experience for the company as it embarks on its next wave of innovation in the B2B SaaS product space. Comments Suman Gopalan, Global CHRO at Freshworks: “AI and ML will have an increasingly pivotal role in what we do, both to develop Freshworks’ future product portfolio, as well as to ensure that we are constantly increasing the value proposition for our customers.” Skills Expected Of Data Scientists At Freshworks The hiring process at Freshworks is designed to test the technical skills, design capabilities, problem-solving abilities and cultural fit of candidates. The company also lays emphasis on onboarding a diverse set of data science aspirants with a background in Computer Science, Statistics, Applied Mathematics, Operations Research, and Advanced Economics. “We focus on attracting talent who believe in our vision and are passionate about solving the challenges we are trying to solve,” says STS Prasad, Executive VP – Engineering at Freshworks. “Beyond educational background, we also ensure that each applicant has adequate skills, along with the commitment to be successful,” he adds. When it comes to identifying their recruitment needs for data science, AI and ML, the company typically hires across the full spectrum – Data Scientists, ML Engineers, Data Platform Engineers, and Data Analysts, to name a few. “We are always looking for talent who are experienced in designing and developing hyperscale data and computational systems, big data applications and pipelines,” says Gopalan. Some technical skills that can prove valuable in data science interviews at Freshworks are: ● ML Frameworks: Scikit-learn, TensorFlow\/Pytorch\/Keras, Spark\/Distributed Programming ● Techniques: NLP, Image Recognition, Neural Networks, Clustering, Supervised\/Unsupervised Classification (XGBoost, Random Forest), Dimensionality Reduction, Advanced Regression, Bayesian Inference, Feature Engineering ● Platform Engineering: ML platform\/pipeline development, large scale model deployment using cloud-based architecture ● Big Data: Near real-time big data applications and pipelines developed on Hadoop\/Spark, Teradata or other statistical packages ALSO READ: Why Data Science Job Market Is Better Positioned For Recession Hiring & Interview Process The company integrates technical rounds to test for knowledge on data structures and algorithms required for specific job positions. Additionally, when hiring for senior levels, it also assesses strategic acumen, managerial leadership, thought leadership, and ability to create an impact within the organisation when recruiting for its data science positions. “The first round usually reviews the background and experience of the candidate and their problem-solving skills,” says Prasad. “Here, the candidate is often asked to define AI solutions to identify which customers are likely to buy the product,” he adds. In this round, interviewers have time to ask open-ended questions that assess the candidates’ thought process. It also leaves room for candidates to ask follow-up questions. “The second round focuses on design and architecture to assess the candidate’s technical depth and their ability to scale solutions through sustainable architecture, etc,” says Gopalan. “Oftentimes, the interviewers challenge a candidate’s assumptions to assess their reasoning ability to support their existing design. Additionally, senior-level candidates are asked to present how they would solve complex real-life scenarios,” he adds. Non-Traditional Ways Of Hiring Data Scientists Like some companies including Mindtree, Infosys and CSS Corp, Freshworks also employs non-traditional means to straddle between hiring for specific needs and opportunistically hiring specific individuals. Some of these unconventional hiring methods to tap into talented professionals include engaging in research partnerships with academic institutions. “These partnerships with academia are more of a symbiotic relationship, where the research students get their research sponsored by Freshworks, and they, in turn, focus on solving our business problems,” says Prasad. “We also focus on building networks by participating in technology events and local meet-ups. We are always on the lookout for great talent, regardless of the time of year,” he adds. Besides these, the company also sources candidates from job portals, peer networks, and through external consultants who help them identify potential talent. Given the niche of the talent pool, it also urges employees to refer friends, professional contacts and acquaintances whom they know would fit a particular role that is open. ALSO READ: Top Six Free Courses To Help Data Scientists Prepare For Job Interviews Common Hiring Mistakes In Data Science According to Freshworks, a common error across the data science landscape is hiring candidates who are academically focused but do not have enough practical experience. “Since data science, as a function, largely tends to overlap with research, some applicants might expect the job to be research-focused, while the job role might require the candidate to apply knowledge, crunch the data and solve complex industry problems,” explains Gopalan. “If applicants do not demonstrate the motivation to solve complex industry problems, there is a good chance that they will struggle to be successful in their role of developing fit-for-purpose, AI-first solutions,” he adds. As a side note, Gopalan also feels that companies need to make sure that they effectively communicate to candidates what the job roles and responsibilities are, “which is especially important as many employees will transition from the academic world.” Opportunities For Data Scientists At Freshworks “Our employees have an enormous role in furthering our AI-powered vision across all our products,” begins Prasad. “In a real sense, they are helping chart the future success of not only our company but of every single organisation that uses our products,” he adds. Data scientists at Freshworks, depending on their specific job requirements, are positioned to contribute towards developing cutting-edge Freddy (the company’s AI engine) capabilities across sales, support, marketing, and HR. This includes AI-powered lead scoring, automated workflows, conversational AI, response recommendation to agent and customer queries, insights and big data analytics. Furthermore, the company also constantly looks to widen the scope of its research projects, providing employees opportunities to work with the brightest minds in the industry. For instance, it has partnered with reputed institutes like the Robert Bosch Centre for Data Science and AI in IIT Madras to bolster its AI capabilities. “At Freshworks, we deal with a large volume of data, with over 250,000 businesses using our technologies,” says Gopalan. “Aspiring data scientists have access to a wealth of data in their quest to find answers to how companies are better engaging with their customers,” he adds.","excerpt":"With data science emerging as a popular career choice in the 21st century, the hiring market in the space has become increasingly competitive. A cursory search on popular job portals will tell you that there is a large pool of existing and aspiring data scientists in India alone. This may indicate that the supply of […]","categories":["AI Hirings"],"tags":["automated data science solutions","big data roi","Data analyst jobs","Data Science Jobs","Data Scientists","Freshworks"],"author_name":"Anu Thomas","publish_date":"2020-05-14T19:00:00","publication_year":"2020","word_count":1189,"keywords":["data science","scikit-learn","Keras","AI","neural network","PyTorch","ML","Data analyst jobs","Data Science Jobs","Freshworks","automated data science solutions","NLP","big data roi","analytics","TensorFlow","Data Scientists"],"extracted_tech_keywords":["AI","ML","neural network","NLP","data science","analytics","TensorFlow","PyTorch","Keras","scikit-learn"],"url":"https:\/\/analyticsindiamag.com\/ai-hiring\/what-is-the-hiring-process-of-data-scientists-at-freshworks\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":40744,"title":"5 Jobs In India That Require Your Tableau Skills","content":"Analytics India Jobs offers a new platform for the data science community in India by bringing to the fore latest job openings across the country. It is an exclusive platform that brings jobs across all the spectrum of data science domain. In this article, we list five Business Intelligence (BI) job openings that requires decent data visualization skills with knowledge of Tableau you can apply for right away. Senior Tableau Developer@ Redback IT Solutions, Noida One of the leading IT Services and Analytics Company looking for a Senior Tableau Developer for Noida Location. The job role will include creating and developing Tableau reports, report scheduling using Tableau server, understanding existing reports and dashboards, and more. Requirements Candidates with 5+ years of experience Good experience creating Tableau reports and interactive Experience using filters, parameters and calculated fields on Tableau Experience with Tableau workbooks from multiple data sources Any prior experience migrating reports to Tableau Experience handling row level security SSRS skills – Drill down, sub-reports, cross tab reports Apply for the job here. Digital Data Analyst @ DG7 Solutions Pvt.Ltd., Mumbai DG7 Solutions is an analytics-focused, research-based consulting partner for medium and large organisations offering end-to-end Digital Marketing Solutions. As a part of this role, the candidate would be required to work with internal stakeholders to determine ongoing business intelligence needs in order to design, develop and produce ongoing reports and dashboards to support the business. Requirements Manage projects related to digital analytics from start to finish including data integration, report and dashboard automation, product analytics and insight, data collection, and product optimization Assist in identifying opportunities that are most effective at driving conversions, revenue, ROI, and scale across Digital Marketing programs. Web Analytics: In addition to setting up standard web reports, the ideal candidate will continually mine Google Analytics Developing new skills and sharing your knowledge with the Data Team Knowledge of Tableau to build dashboards and customer insights Key skills required are SQL, R, Python or PySpark Apply for the job here. Business Analyst @Tredence, Bangalore Tredence is the first analytics services company focused on the last mile of analytics adoption. They have capabilities ranging from Data Visualization, Data Management to Advanced analytics, Big Data to Machine Learning. As a part of this role, candidate is expected to brainstorm with clients and internal teams to define a problem, translate the business problem into an analytical problem, solve the analytical problem Requirements At least 2 years of work experience Proficiency in SQL\/Hive [must have] Handling and analysing large data Data cleaning and processing R (Basic data manipulation, DPLYR, GGplot, Random forest) Python (Basic data manipulation using pandas, matplotlib, sklearn) Hadoop, Spark Tableau Apply for the job here. Analyst @ Mastercard, Gurgaon As an analyst, the candidate will Work closely with Data & Services Advanced Analytics teams and external clients around the world to architect, develop, and maintain advanced reporting and data visualization capabilities on large volumes of data in order to support consulting projects, model development, model scoring, and campaign measurement. And, translate business requirements into tangible solution specifications and high quality, on time deliverables. Requirements B.Tech or Mathematics. M.S. preferred 2-5 years of relevant experience in hands-on data programming, querying, data mining and report development using large volumes of granular data to deliver business intelligence and custom reporting solutions in a Microsoft SQL Server environment Strong hands-on experience on SQL, Tableau, SSIS Analytical\/Problem Solving Relevant retail and payments industry experience a plus Apply for the job here. Data Analyst @ Access Analytics Solutions Pvt.Ltd, Bangalore Access Analytics International provides advanced analytical tools and services for engineering and data analysis. A data analyst will manage data sources, database and movement of data and data organization across different systems. Requirements Bachelors\/PGDM\/ Masters degree in Information Systems \/Maths\/ Statistics\/\/ Finance\/ Engineering \/Business Analytics Actuaries\/FRM\/CFA\/CQF\/PRM certification would be a plus 0-2 years of experience in database or project management Should have hands-on experience of designing database schemas SQL and at least one of R\/Python\/VBA Experience in handling large structured and unstructured datasets Exposure to tools\/platforms – Hadoop ecosystem and DB systems Good hands-on expertise in Tableau and ETL tools. Apply for the job here","excerpt":"Analytics India Jobs offers a new platform for the data science community in India by bringing to the fore latest job openings across the country. It is an exclusive platform that brings jobs across all the spectrum of data science domain. In this article, we list five Business Intelligence (BI) job openings that requires decent […]","categories":["AI Hirings"],"tags":["data visualization","Tableau"],"author_name":"Ram Sagar","publish_date":"2019-06-14T11:11:38","publication_year":"2019","word_count":691,"keywords":["data science","Go","machine learning","Tableau","AI","Python","data visualization","analytics","SQL","Matplotlib","R","Pandas"],"extracted_tech_keywords":["AI","machine learning","data science","analytics","Pandas","Matplotlib","Python","R","SQL","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-hiring\/5-jobs-in-india-that-require-your-tableau-skills\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10131709,"title":"NatWest Group Opens New Office in Bengaluru, Will Hire 3000 Software Engineers by 2026","content":"NatWest Group has announced the lease of a new office in Bengaluru located at Bagmane Constellation Business Park. The new site follows our announcement last year that NatWest Group was looking to recruit 3,000 new software engineers in India by 2026. The seating capacity in the new office will be three times that of the current office, with the opportunity to increase this further. The state-of-the-art facility spans over 370,000 square feet over 11 floors and is a LEED-certified green building (Leadership in Energy and Environmental Design). The location will serve as a hub for pioneering technology solutions and cutting-edge developments, supporting us as we work in simpler and smarter ways to better serve our customers. Bengaluru is a key strategic location for NatWest Group in India, alongside Gurugram and Chennai. This expansion not only strengthens our presence in India, where we have our second-largest employee base outside the UK, but also enhances our colleague value proposition. “Bengaluru is known for its vibrant technology sector and skilled talent pool, so this new office marks a significant chapter in our growth journey across India. Strengthening our global operations further positions us at the forefront of innovation as we continue to prioritise improving the customer and colleague experience,” said Scott Marcar, Group Chief Information Officer, NatWest Group. Punit Sood, Head of India, NatWest Group, added, “Our new Bengaluru office is not just an expansion of physical space but a strategic investment in our future. With a modern office design to enhance productivity and create an inspiring environment for our employees it reflects our commitment to the vast potential of India’s talent and technology ecosystem.”","excerpt":"India is NatWest’ second largest employee base outside the UK.","categories":["AI News"],"tags":[],"author_name":"Pritam Bordoloi","publish_date":"2024-08-07T11:37:55","publication_year":"2024","word_count":272,"keywords":["programming_languages:R","AI","innovation","ViT","R"],"extracted_tech_keywords":["AI","R","ViT","innovation","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/natwest-group-opens-new-office-in-bengaluru-will-hire-3000-software-engineers-by-2026\/","complexity_score":2,"technical_depth":5,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":17512,"title":"Rahul Gandhi To Visit Silicon Valley To Learn More About Artificial Intelligence","content":"Artificial Intelligence has seeped into most industries in India, in one way or another, and politics is no exception. Congress Vice President Rahul Gandhi set to visit the US to meet noted personalities in the AI and related field. According to reports, Gandhi is looking forward to “expand his thoughts about artificial intelligence”. Gandhi reportedly also wants to bring back the knowledge and implement it at the policy level in the Congress party’s vision documents. He will be accompanied by Chairman of Overseas Congress Department Sam Pitroda, who has fixed Gandhi’s meeting with technology experts. “Many important think-tanks from Silicon Valley at the San Francisco Bay are converging to discuss the artificial intelligence, the future technology. Rahul-ji is greatly interested in making India a leader in the artificial intelligence as some years later software technology will lose its sheen,” a senior Congress leader said. AI is no longer a nascent area of interest, either for investment or innovation in India. In fact, many big technology players have invested huge sums of money for greater innovation and expansion of related businesses. The Bharatiya Janata Party (BJP) had invested a huge chunk of money as well as manpower on technology, social media, data analysis and marketing during the runup to the 2014 Lok Sabha elections. Political commentators have suggested that the example of the landslide victory of Narendra Modi may perhaps have prompted Gandhi to explore possible use of technology in the light of upcoming Gujarat Assembly Elections. Gandhi is also set to address a conference on ‘India At 70: Reflections On The Path Forward’ at the University of California, Berkeley, on September 11. He is also likely to attend a host of other events and meet experts in various fields, investors and Non-Resident Indians during the visit. Gandhi recently returned from his trip to Norway, where visited several biotechnology and nanotechnology-related firms.","excerpt":"Artificial Intelligence has seeped into most industries in India, in one way or another, and politics is no exception. Congress Vice President Rahul Gandhi set to visit the US to meet noted personalities in the AI and related field. According to reports, Gandhi is looking forward to “expand his thoughts about artificial intelligence”. Gandhi reportedly […]","categories":["AI News"],"tags":["AI (Artificial Intelligence)","nanotechnology","rahul gandhi"],"author_name":"Prajakta Hebbar","publish_date":"2017-09-05T07:40:23","publication_year":"2017","word_count":311,"keywords":["artificial intelligence","programming_languages:R","AI","rahul gandhi","innovation","GAN","nanotechnology","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","R","GAN","innovation","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/rahul-gandhi-visit-silicon-valley-learn-artificial-intelligence\/","complexity_score":2,"technical_depth":6,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10082960,"title":"The Man Behind One of The Most Important AI Advancements, AlphaFold","content":"In July 2021, a London-based subsidiary of Alphabet, DeepMind, delivered the revolutionary answer to the decades-long ‘protein-folding problem’ in the history of AI research: AlphaFold. The open-source AlphaFold can accurately predict 3D models of protein structures from 1D amino acid sequences, which is accelerating scientific research in every field of biology and life science. In an exclusive interaction with Analytics India Magazine, Pushmeet Kohli, head of research (AI for science, robustness and reliability) at DeepMind, shared the importance of AI for good, alongside his experience of being part of the revolutionary project and more. “Science has given agency to humanity, and there are limits to our understanding of nature,” said Kohli, saying that the pandemic made it clear that we have no control over nature. He further said that science has broadened our understanding of nature and has given us more means to leverage AI, which he believes will be one of the most powerful technologies that we as species can leverage to benefit science. “I think there is nothing more meaningful that one can do,” said Kohli along the lines of the importance of using AI for the greater good. Protein Folding Models “AlphaFold is a great example of how we can leverage AI because proteins are essentially the building blocks of all life. We are essentially a sort of big collection of proteins. It’s not just us; every single living thing on the planet is made up of these proteins,” said Kohli. Further, he said that they did not completely understand what the structures are and what the function of all these proteins is. In that respect, Kohli believes that AlphaFold is a great watershed moment because it shows what AI can do in broadening the scientific community’s understanding of this important topic. The function of a protein is directly related to its structure. For instance, like a key fitting into a lock, antibody proteins fold into forms that allow them to accurately detect and target particular foreign bacteria. Therefore, understanding how proteins will fold into shapes is crucial to understanding how organisms function and, eventually, how life itself works. Merely 17% of the roughly 20,000 proteins in the human body had their 3-D structures known prior to AlphaFold. Enter AlphaFold and now 3-D structures for nearly the entire (98.5%) human proteome. This is a giant leap considering drug discovery is now easier. AlphaFold AlphaFold predicts protein structures through three distinct deep-learning neural network layers. It was trained on thousands of available proteins and their structures found in the Protein Data Bank (PDB). The first layer is made up of a variational autoencoder stacked with an attention model, which generates real-looking fragments based on a single sequence’s amino acids. The contact map, a 2D representation of amino acid residue distance, is projected onto a single dimension for input into the CNN (Convolutional neural network) in the first sublayer to optimise inter-residue distances. The second sublayer refines a scoring network, which measures how well the 3D CNN-generated substructures resemble proteins. After regularising it, they add a third neural network layer to compare the produced protein to the actual model. Kohli said that his team’s target was to understand what the structure of proteins is, as every living creature’s tissue and cell have these proteins. AlphaFold predicts the structure of a protein. “But proteins are not always in a static state. They can be in multiple states according to their function or in the presence of other ligands. So, there remain many questions around how proteins interact with each other or with another set of ligands, and what energy they require to go from one state to another, among others, and our team is interested in working with them to find answers,” he added. DeepMind used AlphaFold to predict the protein structures of the COVID-19 outbreak—SARS-CoV-2. Before making it public to the research community, the findings were reviewed by scientists at the Francis Crick Institute in the UK. The membrane protein, protein 3a, nsp2, nsp4, nsp6, and papain-like C-terminal domain are among these proteins. These protein structures were created to aid in the discovery of new medications and therapies in the fight against COVID-19 and may contain docking sites for those substances.","excerpt":"AlphaFold can predict 3D models of protein structure which aids drug discovery and medical care solutions.","categories":["AI Features"],"tags":["AlphaFold","covid-19","DeepMind","Interviews and Discussions"],"author_name":"Shritama Saha","publish_date":"2022-12-21T17:00:00","publication_year":"2022","word_count":702,"keywords":["Go","AlphaFold","API","covid-19","AI","neural network","RAG","DeepMind","analytics","AI research","GAN","CNN","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","neural network","analytics","RAG","R","Go","API","GAN","CNN","AI research"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/the-man-behind-one-of-the-most-important-ai-advancements-alphafold\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10170956,"title":"IndiaAI Adds 14,000 GPUs, Count Rises to 32,000, Confirms IT Minister Ashwini Vaishnaw","content":"India is set to procure an additional 14,000 GPUs as part of the IndiaAI Mission, Union IT Minister Ashwini Vaishnaw announced on Thursday, reported CNBC-TV18. This brings the total GPU count under the mission to over 32,000, following an earlier acquisition of nearly 18,700 units. “This is a big change, and AI is here for good,” Vaishnaw said while stressing the government’s push to “democratise” access to AI and drive technological transformation across sectors. The procurement aims to strengthen India’s AI compute capacity, particularly for training large and small language models (LLMs and SLMs), which power most generative AI systems. The expansion is a big step towards building India’s own AI infrastructure. Seven companies cleared the technical round of the IndiaAI Mission’s second GPU tender, and their commercial bids were recently opened. These include Netmagic (NTT Global), Cyfuture India, Sify Digital Services, Vensysco Technologies, Locuz Enterprise Solutions, Yotta Data Services, and Ishan Infotech. Originally, the IndiaAI Mission featured over 18,000 GPUs through public-private partnerships with companies including Jio Platforms, NxtGen Data Centre, Locuz Enterprise, E2E Networks, CtrlS DataCenters, CMS Computers, Orient Technologies, Tata Communications, Vensysco, and Yotta Data Services. According to IT Minister Ashwini Vaishnaw, eligible users will get access to this computing power at up to 40% lower cost. The total budget for the initiative is ₹10,372 crore. About 14,000 units have already been made available in the first round, according to the CNBC report. Vaishnaw added that the mission will also focus on developing foundational AI models that will be made open-source. “The goal is to ensure AI tools and infrastructure are not concentrated in the hands of a few but are accessible to all innovators and developers,” he said. The announcement signals the government’s intent to provide wide access to AI resources and reduce reliance on private cloud monopolies, aligning with broader digital public infrastructure goals.","excerpt":"The procurement aims to strengthen India’s AI compute capacity, particularly for training large and small language models (LLMs and SLMs), which power most generative AI systems.","categories":["AI News"],"tags":["India AI"],"author_name":"Siddharth Jindal","publish_date":"2025-05-29T18:14:47","publication_year":"2025","word_count":309,"keywords":["India AI","Go","programming_languages:R","AI","programming_languages:Go","SLM","Git","Aim","generative AI","small language models","R"],"extracted_tech_keywords":["AI","generative AI","Aim","small language models","SLM","R","Go","Git","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/indiaai-adds-14000-gpus-count-rises-to-32000-confirms-it-minister-ashwini-vaishnaw\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10162554,"title":"Bosch Global Software Technologies Opens New Office in Bengaluru’s Electronic City","content":"Bosch Global Software Technologies (BGSW) officially inaugurated its new office in Bengaluru’s Electronic City on Thursday, marking a significant step in bringing its teams together under one roof. BGSW is the technology arm of the German multinational engineering and technology company Robert Bosch, commonly known as Bosch. It specialises in end-to-end engineering, IT, and business solutions. The new facility will house over 5,000 engineering and software development associates and foster collaboration and innovation, as announced by the team on LinkedIn. The state-of-the-art campus, featuring cutting-edge labs, creative spaces, and modern amenities, is designed to enhance teamwork and agility. The space aims to blend work, inspiration, and technology-driven solutions while strengthening BGSW’s culture of innovation. The inauguration event was attended by Soumitra Bhattacharya, chairman of Bosch India and CEO and director of the Indian Foundation for Quality Management (IFQM), along with Alexander Lenk, president of Global Real Estate at Robert Bosch GmbH. This new office reinforces BGSW’s commitment to technological excellence while creating an environment that supports creativity, efficiency, and innovation. This week, Bosch Limited  announced its total revenue from operations of ₹4,466 crores (495 million euros) in Quarter 3 of FY 2024–25, an increase of 6.2% over the same quarter of last year. This growth is driven by the increase in service income from the development of automotive components for major original equipment manufacturers (OEMs). “This quarter, Bosch Limited continued to focus on innovation and technology by developing customer-focused solutions. Mobility Aftermarket and Beyond Mobility divisions showed considerable growth in an otherwise tough quarter while ensuring overall double-digit profitability. This performance reaffirms our strategic ability to respond to the market needs,” said Guruprasad Mudlapur, president of the Bosch Group in India and managing director of Bosch Limited.","excerpt":"The new facility will house over 5,000 engineering and software development associates and foster collaboration and innovation.","categories":["AI News"],"tags":["AI","bosch india"],"author_name":"Shalini Mondal","publish_date":"2025-01-30T18:15:14","publication_year":"2025","word_count":287,"keywords":["bosch india","programming_languages:R","AI","innovation","Git","BERT","Aim","llm_models:BERT","ViT","R"],"extracted_tech_keywords":["AI","Aim","R","Git","BERT","ViT","innovation","llm_models:BERT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/bosch-global-software-technologies-opens-new-office-in-bengalurus-electronic-city\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10104493,"title":"AMD Loves Llama So Much","content":"The open source community has a love affair with Llama, Meta’s open source large language model. So much so that, many of the models that are coming up in the generative AI ecosystem are based on top of Llama 2, and developers are using the model for all purposes. Why should AMD stay behind? While demonstrating the kernel performance of the newly released AMD Instinct MI300X Accelerator at the Advancing AI event, Lisa Su, CEO of AMD, said that MI300X performs 1.2 times better than NVIDIA H100 on a single kernel when running Meta’s Llama 2 70B. “We are going to talk about this a lot,” Su said, speaking about Llama 2. Su further goes on and demonstrates that when it comes to inferencing Llama 2, one single server of AMD which consists of eight MI300X, performs 1.4 times faster than the server of an H100. Clearly, the performance of the MI300X is demonstrated the best with Llama 2. Lisa Su, CEO of AMD Moreover, Kevin Scott, CTO of Microsoft discussed with Su on stage how AMD is contributing in Microsoft’s AI journey. While talking about the announcement at Ignite about using MI300X on Azure, Scott said that he is eager to Bring up GPT-4 and Llama 2 on MI300X and seeing the performance, and rolling it into production is something Scott said that he has been waiting for eagerly. The mutual love for open source “As important as the hardware is, software is what really drives innovation,” Su said, talking about the ROCm 6, the latest version of AMD’s open source parallel computing offering, which is an alternative to NVIDIA’s CUDA. ROCm 6 is releasing in the coming week. Victor Peng, President of AMD, showed how building a strong ecosystem has enabled the company to create a successful open source framework in ROCm. For demonstrating this, Peng showcased how MI300X with ROCm 6 is eight times faster than MI250X with ROCm 5, when inference Llama 2 70B. Victor Peng, President of AMD On smaller models such as Llama 2 13B, ROCm with MI300X showcased 1.2 times better performance than NVIDIA coupled with CUDA on a single GPU. The most groundbreaking announcement is that Meta is partnering with AMD and the company would be using MI300X to build its data centres. To demonstrate this partnership, Ajit Matthews, Meta AI senior director engineering with Su said that MI300X is trained to be the fastest design-to-deployment solution in Meta’s history. Meta has been using 150k H100 GPUs for research and training. It would be interesting to see how the shift to AMD MI300X takes place. “We are investing for the future by building new experiences for people through our services, and advancing open technologies and research for the industry,” said Matthews. In July, Meta opened up Llama 2 family of models and, “we were blown away by the reception from the community,” he added. Lisa Su, CEO of AMD; Ajit Matthews, Meta AI senior director engineering “We believe that an open approach feeds to better and safer technology in the long run, as we have seen with the PyTorch and Open Compute project, and dozens of other AI models,” said Matthews, speaking about how AMD was also the co-founder of PyTorch with Meta. “We completely agree with the vision and with the open ecosystem, and that being the path towards innovation with all the smart folks in the industry,” said Su also highlighting AMD’s focus on open software such as ROCm, along with the race towards open hardware with MI300X. AMD’s long love for Meta Meta has been working with AMD for EPYC CPUs since 2019. Meta also recently deployed AMD’s Genoa and Bergamo based servers at scale across its infrastructure. “We have been working with the Instinct GPUs, starting from the MI100, since 2020,” said Matthews. “We have also been benchmarking ROCm and working together for its support on PyTorch across each generation of AMD Instinct GPU.” “As the Llama family of models continue to grow in size and power, which they will, the MI300X with its 192GB of memory and higher memory bandwidth meets the requirement for large language model inference,” added Matthews. He has also said he is pleased with the fact that AMD was specifically optimised ROCm for Llama models and Meta is seeing great performance numbers. Not just bigger models, AMD is also focusing on edge computing with the introduction of Ryzen AI PCs, along with the first version of Ryzen AI software for building generative AI applications on your PCs by using pretrained models available on Hugging Face, such as Llama 2. Lisa Su, CEO of AMD To demonstrate the performance of AMD Ryzen 8040, the newest version of its on-device neural processing unit (NPU), Su highlighted that Llama 2 7B performs 1.4 times faster than the previous versions. This marks the beginning of using small Llama 2 models on hardware powered by AMD. The long running partnership with Meta, and the mutual love for open source is the reason why AMD loves Llama so much.","excerpt":"For every single demonstration of its GPU capabilities, AMD tested out Llama 2, and showed it to the audience.","categories":["AI Features"],"tags":["AMD","amd ai event","amd mi300x","amd rocm"],"author_name":"Mohit Pandey","publish_date":"2023-12-08T13:00:00","publication_year":"2023","word_count":840,"keywords":["amd rocm","CUDA","AMD","Meta AI","amd mi300x","Hugging Face","Go","AI","PyTorch","R","amd ai event","generative AI","edge computing","Azure"],"extracted_tech_keywords":["AI","generative AI","Meta AI","PyTorch","Hugging Face","Azure","edge computing","CUDA","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/amd-loves-llama-so-much\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":28991,"title":"TSCL Is The Newest Addition To Deep Learning Algorithms","content":"One of the newest developments in deep learning is the curriculum learning, where algorithms are trained to learn on a meaningful order in increasing complexity rather than just examples being fed to them. A new study by researchers who worked with OpenAI has dug deep into curriculum learning. Their model, called as Teacher-Student Curriculum Learning (TSCL), aims to be a game changer in learning subtasks associated with major deep learning tasks. This article looks into the technicality surrounding TSCL. What Is TSCL? In the context of curriculum learning, a Teacher algorithm gives subtasks to the Student algorithm in the increasing order of complexity, while the Student performs them and returns a score. This is gradually repeated until all tasks are performed by the Student successfully. So, as and when the Student learns and masters a particular task, the Teacher assigns more probability on the subsequent task ahead thus focussing less on the current one as it has been learnt completely. Repetition of performing tasks here makes learning faster. One important point to be noted here is the Teacher algorithm also learns information simultaneously along with the Student algorithm. This forms the basis for TSCL algorithm. In fact, TSCL is specified as a Partially Observable Markov Decision Process (POMDP) in the study. Two cases of POMDP, one for reinforcement learning (simple training) and the other, for supervised learning (batch training), are charted out. The reason POMDP is chosen here is to optimise the Teacher algorithms’ rewards in line with the Student algorithms’ sub-task performance. Matiisen et.al, the creators of TSCL, say “While an obvious choice for optimization criteria would have been the performance in the final task, initially the Student might not have any success in the final task and this does not provide any meaningful feedback signal to the Teacher. Therefore we choose to maximize the sum of performances in all tasks. The assumption here is that in curriculum learning the final task includes the elements of all previous tasks, therefore good performance in the intermediate tasks usually leads to good performance in the final task” On this front, PODMPs are generally solved using RL but the training itself takes time and becomes iterative. Thus, researchers derive insights from the popular ‘multi armed non-stationary bandit problem’ and bring out the following new algorithms to incorporate in TSCL. Online Algorithm Naive Algorithm Window Algorithm Sampling Algorithm All these algorithms are tweaked with respect to improving scores as well as keeping a check on the number of times a task has been performed in the Teacher-Student setup. TSCL In Decimal Number Addition And Minecraft There are many research-oriented applications under curriculum learning. Decimal number addition through LSTM is one notable work, where a sequence-to-sequence model was implemented. Although this technique has found success in supervised learning, it has faced setbacks either in learning performance or end up using too much memory for addition. Thus, Matiisen et.al, consider this problem for analysis. As mentioned earlier, batch training POMDP is taken as the TSCL method here. The addition is carried along two parameters: 1-dimensional curriculum teaching and 2-dimensional curriculum teaching. In the former, tasks related to finding the maximum number of digits in the number obtained after addition, while the latter includes another criterion i.e., taking the length of numbers separately on top of finding decimal digits. Results show that TSCL fares better than conventional LSTM. In fact, with more expected return, it is even faster. Popular video game Minecraft, was also experimented with respect to reinforcement learning strategies. By using Microsoft’s Project Malmo with OpenAI Gym, a 5-step Curriculum Learning is created by Matiisen and team. This generates random mazes in Minecraft where the learning agent carefully navigates and learns the maze environment. (A detailed account of the Minecraft training can be found here.) Even in this case, Minecraft agent learns faster with every run iteration. If a five-step curriculum is performed without considering each step, the agent terribly fails in learning the environment, which supports the researchers’ critique on selecting only the final task. Comment While this study has paved way for using TSCL in a handful of applications, it is again yet to stand with standard reinforcement learning and supervised learning algorithms. Nonetheless, TSCL will alleviate complexities in algorithms at every stage (dividing into sub-tasks etc.) thus reducing the burden on computing power and similar resources.","excerpt":"One of the newest developments in deep learning is the curriculum learning, where algorithms are trained to learn on a meaningful order in increasing complexity rather than just examples being fed to them.   A new study by researchers who worked with OpenAI has dug deep into curriculum learning. Their model, called as Teacher-Student Curriculum […]","categories":["AI Features"],"tags":["Deep Learning","Reinforcement Learning"],"author_name":"Abhishek Sharma","publish_date":"2018-10-06T04:41:13","publication_year":"2018","word_count":725,"keywords":["Go","Reinforcement Learning","OpenAI","AI","programming_languages:R","programming_languages:Go","Git","Aim","deep learning","LSTM","Deep Learning","R"],"extracted_tech_keywords":["AI","deep learning","OpenAI","Aim","R","Go","Git","LSTM","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/tscl-is-the-newest-addition-to-deep-learning-algorithms\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10018634,"title":"Salesforce Launches AI-Economist: A Complete Guide With Python Codes","content":"Recently Salesforce Research launched an open-sourced framework for economic policy design and simulation: AI Economist. It is an economic simulation environment in which Artificial intelligence agents extract and trade resources, make houses, earn salaries, and pay taxes to the government bodies. It is a reinforcement learning(RL) problem to tax research to provide simulation and data-driven solutions to defining optimal taxes for specific socio-economic objectives. It was created by L-R Melvin, Gruesbeck, Alex Trott, Stephan Zheng, Richard Socher, and Sunil Srinivasa at the Salesforce research lab. AI economists use a different collection of AI agents designed to simulate millions of years of economies to help economists, governments, and other bodies to optimize social outcome in the real world. As a society, we are entering uncharted territory — a new world in which governments, business leaders, the scientific community, and citizens need to work together to define the paths that direct these technologies at improving the human condition and minimizing the risks. Marc Benioff, Chairman and CEO Its research paper is been published here: “The AI Economist: Improving Equality and Productivity with AI-Driven Tax Policies” Now let’s straight jump to code here Foundation “Foundation”: It is a name given to the economic simulator for AI Economist, this is the first of many tutorials designed to see and explain how the foundation works. Simulation As shown in the above image it is a top view of specially designed modeling economies in a spatial 2D grid world. A nicely rendered example of what an environment looks like. As mentioned in the paper it uses a scenario with having 4 agents in a field\/world with wood and stone, which can be traded, collected, and even to build a house. The above image is only indicating some features, behind the scenes, we have agents with stone inventories, Wood, and coin to exchange commodities marketplace. Also, the agents pay taxes from time to time by earning money through trading and building Markov Decision Process AI-Economist heavily depends on MDP(Markov Decision Process), describing episodes in which agents receive observation and use policy to select actions. The environment then advances to an upgraded state, using the old states and actions. The agent receives new observations and rewards. This process repeats over the T timestamp. Implementation We are going to use the basic economic simulator demo, licensed by Salesforce to see how this model works and how we can leverage the power of AI economists in our given situation. This tutorial will give you enough to see how to create the type of simulation environment described above and interact with. Downloading GitHub repo and installing dependencies import sys IN_COLAB = 'google.colab' in sys.modules if IN_COLAB: ! git clone https:\/\/github.com\/salesforce\/ai-economist.git % cd ai-economist ! pip install -e . else: ! pip install ai-economist # Import foundation from ai_economist import foundation import numpy as np %matplotlib inline import matplotlib.pyplot as plt from IPython import display if IN_COLAB: from tutorials.utils import plotting  # plotting utilities for visualizing env state else: from utils import plotting Let’s Create a Simulation Environment The scenario provides a high-level gym style API, that lets agents interact with it. The Scenario class implements an economic simulation with multiple agents, each scenario implements two main methods: Step: for advancing simulation to next stateReset puts simulation back in an initial state. # Define the configuration of the environment that will be built env_config = { # ===== SCENARIO CLASS ===== # Which Scenario class to use: the class's name in the Scenario Registry (foundation.scenarios). # The environment object will be an instance of the Scenario class. 'scenario_name': 'layout_from_file\/simple_wood_and_stone', # ===== COMPONENTS ===== # Which components to use (specified as list of (\"component_name\", {component_kwargs}) tuples). #   \"component_name\" refers to the Component class's name in the Component Registry (foundation.components) #   {component_kwargs} is a dictionary of kwargs passed to the Component class # The order in which components reset, step, and generate obs follows their listed order below. 'components': [ # (1) Building houses ('Build', {'skill_dist': \"pareto\", 'payment_max_skill_multiplier': 3}), # (2) Trading collectible resources ('ContinuousDoubleAuction', {'max_num_orders': 5}), # (3) Movement and resource collection ('Gather', {}), ], # ===== SCENARIO CLASS ARGUMENTS ===== # (optional) kwargs that are added by the Scenario class (i.e. not defined in BaseEnvironment) 'env_layout_file': 'quadrant_25x25_20each_30clump.txt', 'starting_agent_coin': 10, 'fixed_four_skill_and_loc': True, # ===== STANDARD ARGUMENTS ====== # kwargs that are used by every Scenario class (i.e. defined in BaseEnvironment) 'n_agents': 4,          # Number of non-planner agents (must be > 1) 'world_size': [25, 25], # [Height, Width] of the env world 'episode_length': 1000, # Number of timesteps per episode # In multi-action-mode, the policy selects an action for each action subspace (defined in component code). # Otherwise, the policy selects only 1 action. 'multi_action_mode_agents': False, 'multi_action_mode_planner': True, # When flattening observations, concatenate scalar & vector observations before output. # Otherwise, return observations with minimal processing. 'flatten_observations': False, # When Flattening masks, concatenate each action subspace mask into a single array. # Note: flatten_masks = True is required for masking action logits in the code below. 'flatten_masks': True, } # Create an environment instance using this configuration: env = foundation.make_env_instance(**env_config) Start interacting with Simulation env.get_agent(0) def sample_random_action(agent, mask): \"\"\"Sample random UNMASKED action(s) for agent.\"\"\" # Return a list of actions: 1 for each action subspace if agent.multi_action_mode: split_masks = np.split(mask, agent.action_spaces.cumsum()[:-1]) return [np.random.choice(np.arange(len(m_)), p=m_\/m_.sum()) for m_ in split_masks] # Return a single action else: return np.random.choice(np.arange(agent.action_spaces), p=mask\/mask.sum()) def sample_random_actions(env, obs): \"\"\"Samples random UNMASKED actions for each agent in obs.\"\"\" actions = { a_idx: sample_random_action(env.get_agent(a_idx), a_obs['action_mask']) for a_idx, a_obs in obs.items() } return actions We can interact with the simulation, the first environment put in an initial state by using reset. obs = env.reset() Then, we can further call steps to advance the state and advance time by one tick. actions = sample_random_actions(env, obs) obs, rew, done, info = env.step(actions) Reward For each agent, the reward dictionary contains some scalar reward for agent_idx, reward in rew.items(): print(\"{:2} {:.3f}\".format(agent_idx, reward)) Done done object is a default dictionary that records whether all agents\/planner have seen the end of the episode. The default criterion for each agent is to ‘stop’ their episode once in a while as  H  steps have been executed. After the agent is ‘done’, they do not change their state after that. So, while it’s not currently implemented, this could be used to indicate that the episode has ended for a specific Agent. done Info Similarly, we have an info object that can record auxiliary actions from simulators, which can be useful, sometimes for visualization by default it’s empty. info Visualizing Episode def do_plot(env, ax, fig): \"\"\"Plots world state during episode sampling.\"\"\" plotting.plot_env_state(env, ax) ax.set_aspect('equal') display.display(fig) display.clear_output(wait=True) def play_random_episode(env, plot_every=100, do_dense_logging=False): \"\"\"Plays an episode with randomly sampled actions. Demonstrates gym-style API: obs                  <-- env.reset(...)         # Reset obs, rew, done, info <-- env.step(actions, ...) # Interaction loop \"\"\" fig, ax = plt.subplots(1, 1, figsize=(10, 10)) # Reset obs = env.reset(force_dense_logging=do_dense_logging) # Interaction loop (w\/ plotting) for t in range(env.episode_length): actions = sample_random_actions(env, obs) obs, rew, done, info = env.step(actions) if ((t+1) % plot_every) == 0: do_plot(env, ax, fig) if ((t+1) % plot_every) != 0: do_plot(env, ax, fig) Visualizing random episodes play_random_episode(env, plot_every=100) Conclusion We have seen the basics of the AI-economist basic simulation system and what are the main attributes that have been placed within the project, the outputs are pretty satisfying. In further advanced tutorial Salesforce research team has introduced the more complex tutorial, you can follow below resources to learn more : Official GitHub repositoryAI Economist: Research paperBlog: The AI Economist moonshotThe AI Economist websiteEconomic_simulation_advanced tutorialGoogle Colab demo","excerpt":"Recently Salesforce Research launched an open-sourced framework for economic policy design and simulation: AI Economist. It is an economic simulation environment in which Artificial intelligence agents extract and trade resources, make houses, earn salaries, and pay taxes to the government bodies. It is a reinforcement learning(RL) problem to tax research to provide simulation and data-driven […]","categories":["Deep Tech"],"tags":["AI Agents","Matplotlib","Reinforcement Learning"],"author_name":"Mohit Maithani","publish_date":"2021-01-22T14:00:00","publication_year":"2021","word_count":1254,"keywords":["NumPy","artificial intelligence","Reinforcement Learning","TPU","AI","ML","AI Agents","RAG","Colab","Ray","Python","Matplotlib"],"extracted_tech_keywords":["AI","artificial intelligence","ML","Ray","Colab","NumPy","Matplotlib","RAG","TPU","Python"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/salesforce-launches-ai-economist-a-complete-guide-with-python-codes\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10130454,"title":"C5i Acquires Analytic Edge, Second Indian Startup Acquired by the AI Firm So Far","content":"Singapore-based startup Analytic Edge has officially been acquired by AI firm C5i. Analytic Edge, which was founded by Santosh Nair and Vijay Ramaswamy in 2012 in Bengaluru, provides AI-based marketing and sales solutions, allowing companies to make data-driven marketing choices. The Analytic Edge acquisition is the second such acquisition by C5i in India, after Incivus. According to reports, the AI firm is deciding on yet another acquisition before deciding on a listing in India within the next year and a half. Currently the AI firm is looking for a turnover of Rs 1000 crore in the coming fiscal year. “We acquired Incivus in 2023 and its visual AI platform has helped us build and scale various product IPs. With Analytic Edge, we are adding more to that, such as marketing spend optimisation, real-time pricing intelligence, promotion optimisation, etc. Plus, they are close partners to top social media platforms,” CEO Ashwin Mittal said. With the acquisition of two Indian AI startups that target different stages of the marketing process, it is safe to assume that C5i is currently on the lookout for other marketing-inclined AI startups to fill in the gaps in the process. As of now, the company has revealed its intentions to help organisations overcome issues of data scarcity, prediction inaccuracy and biases by leveraging synthetic data. Thanks to this, C5i may be cornering the Indian marketing space, allowing enterprises access to synthetic data to further boost their AI capabilities, especially as India accounts for 2% of its current revenue. This is just one in a stream of acquisitions occurring in India in the AI startup ecosystem, particularly in Bengaluru.AI startups in the Silicon Valley of India have been acquired by major international companies, with the most recent acquisition by Walmart. The retail giant only recently acquired two AI startups based in Bengaluru – Int.AI and Dataturks – lending credence to the fact that Bengaluru has become the San Francisco of the East.","excerpt":"The Analytic Edge acquisition is the second such acquisition by C5i in India, after Incivus. According to reports, the AI firm is deciding on yet another acquisition before deciding on a listing in India within the next year and a half.","categories":["AI News"],"tags":["AI marketing","Mergers and Acquisitions","synthetic data"],"author_name":"Donna Eva","publish_date":"2024-07-29T12:14:33","publication_year":"2024","word_count":324,"keywords":["synthetic data","programming_languages:R","AI","data-driven","emerging_tech:synthetic data","RAG","GAN","AI marketing","Mergers and Acquisitions","R","startup"],"extracted_tech_keywords":["AI","RAG","R","GAN","synthetic data","data-driven","startup","programming_languages:R","emerging_tech:synthetic data"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/c5i-acquires-analytic-edge-second-indian-startup-acquired-by-the-ai-firm-so-far\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10052111,"title":"Facebook Crowdsources This Computer Vision Dataset, Partners With An Indian University","content":"Facebook AI has announced its ambitious long-term project, Ego4D, to solve challenges in egocentric perception – the ability of AI to understand and interact with the world in a similar fashion as we humans do, i.e., from a first-person perspective. To further simplify the understanding, in order to train and teach AI, take, for example, the computer vision system is fed with millions of photos and videos captured by a third person. However, the next-gen AI systems need data that shows the world from the first-person perspective. Further, the Facebook AI team has collaborated with 13 universities and labs across nine countries, including India. International Institute of Information Technology (IIIT), Hyderabad, is the only university from India to team up for the Ego4D project. Founded in 1998, IIIT-H has evolved strong research programmes over the years in several areas, with a strong emphasis on science, technology and applied research for both industry and society. This consortium of universities and labs collected more than 2,200 hours of first-person video in the wild, with over 700 participants going about their daily lives. This cooperation dramatically scales up the amount of egocentric data publicly available to the research community, that too by orders of magnitude more than 20 times greater than any other data set in terms of hours of footage. Facebook supported and funded the project through academic gifts to each of the participating labs and universities. Image: Facebook AI The Project Facebook AI also created five benchmark challenges based on first-person visual experience, which will help future AI assistants progress toward real-world applications. These include: Episodic memory: What happened when? (For example, “Where did I leave my purse?”)Forecasting: What am I likely to do next? (e.g., “You have to add two spoons of sugar now.”)Hand and object manipulation: What am I doing? (e.g., “Let me know how to play the guitar.”)Audio-visual diarization: Who said what when? (e.g., “What was the time to reach the cinema?”)Social interaction: Who is interacting with whom? (e.g., “Help me better hear the person talking to me at this noisy restaurant.”) Ego4D is aimed at addressing issues in embodied AI, a discipline that aims to develop AI systems with a physical or virtual embodiment, such as robots. Embodied AI is based on the theory of embodied cognition, which states that many parts of psychology, human or another, are shaped by aspects of an organism’s entire body. Researchers intend to increase the performance of AI systems such as chatbots, autonomous vehicles, robots, and even smart glasses that have to interact continuously with their environments, humans, and other AI by applying this logic to AI. Facebook distributed head-mounted cameras and wearable sensors to the participants so that they could capture first-person, unscripted videos of their daily lives. The research participants capture video of their day to day routines such as cooking, grocery shopping, talking while playing games and engaging in activities with family and friends. Hence, everything was captured from the centre of action rather than someone shooting the video or capturing the photo from the sidelines. Moreover, the Facebook AI team said it will make this data publicly available in November 2021. In addition to the same, researchers from Facebook Reality Labs employed Vuzix Blade Smart Glasses to further produce an additional 400 hours of first-person video data in staged situations in their own study labs. This information will also be made public. Wrapping up It’s critical to understand that poor representation in computer vision datasets can be harmful, especially as the AI industry lacks unambiguous explanations of bias. Consider, for example — ImageNet and OpenImages — two big, publicly available datasets, have previously been discovered to be US and Euro-centric, embodying humanlike biases regarding race, gender, colour, ethnicity, weight, and others. The datasets from the Ego4D project can ward off these concerns up to a good extent. Additionally, we can hope that it will be possible for assistants to deliver value in unique and meaningful ways using AI-driven capabilities enabled by Ego4D’s benchmarks and trained on the data set.","excerpt":"It’s critical to understand that poor representation in computer vision datasets can be harmful, especially as the AI industry lacks unambiguous explanations of bias.","categories":["AI Features"],"tags":[],"author_name":"kumar Gandharv","publish_date":"2021-10-22T14:00:00","publication_year":"2021","word_count":672,"keywords":["Go","AI assistants","programming_languages:R","AI","chatbots","computer vision","Aim","ViT","GAN","R"],"extracted_tech_keywords":["AI","computer vision","Aim","AI assistants","chatbots","R","Go","GAN","ViT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/facebook-crowdsources-this-computer-vision-dataset-partners-with-an-indian-university\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10056992,"title":"Can You Express Your Feelings Through A Haptic Armband, Just Like A Hug?","content":"Haptics is used to describe the tactile feedback given by different devices to mimic the sensation of touch. It is derived from the Greek word ‘ἁπτικός (haptikos),’ which means ‘pertaining to touch.’ Haptic feedback is a means of communication rather than a new technology; it is revolutionizing how machines and humans interact. In the booker prize-winning novel “The Blind Assassin,” author Margaret Atwood says that “touch comes before sight, before speech. It’s the first language and the last, and it always tells the truth.” Researchers worldwide have been trying to build AI\/ML systems that can comprehend the feeling of touch and replicate this sensation with great precision. For instance, the MIT CSAIL system can connect the senses of vision and touch and learn to see by touching to an extent. Startups and companies have even tried to harness the potential of haptic wearables in the field of fashion. For example, Ducere Technologies, founded by Krispian Lawrence and Anirudh Sharma, graduates of the University of Michigan and the Massachusetts Institute of Technology, respectively, developed the world’s first haptic footwear named Lechal. It allows users to navigate hands-free and tracks fitness parameters. It guides users through vibration, once paired with the android and IOS systems through the app. Big tech companies such as Facebook (or Meta, as it is now called) are highly interested in making tactile sensors, such as touch, a part of robotics and machine learning. It has launched an open-source machine learning library, PyTouch, to help researchers develop AI\/ML models that seamlessly process touch sensation. The new study for a novel haptic armband Computer scientist and roboticist Heather Culbertson has been developing different methods to simulate the sensation of touch at the USC Viterbi School of Engineering. In a new study, Culbertson and Stanford’s researchers have tried to answer whether two human beings can express their emotions from afar as they would during in-person interactions. According to Culbertson, the WiSE Gabilan Assistant Professor and Assistant Professor of Computer Science and Aerospace and Mechanical Engineering at USC, people have a “touch language.” Different people have different comfort levels for social touch, and also, a distinct way of touching can communicate the feelings of empathy, love, and sadness. Therefore, replicating this sensation is not an easy task. The challenge for the researchers was creating an algorithm flexible enough to incorporate the many dimensions of touch. Their study was performed in two phases. First, a novel dataset was created about individuals’ interactions through touch by storing the location and pressure of touches applied while participants communicated different emotions through touch to each other. Then, the researchers classified each gesture according to the sentiment it expressed, such as happiness, attention, sadness, love, gratitude, etc. The next phase was to customize the algorithms to generate a more refined accurate sense of virtual touch. People can perceive their partner’s intention through touch 57% of the time. While interacting with the haptic armband that simulated human touch, participants could discern the emotions communicated 45% of the time. This puts the accuracy of the developed wearable device at around 80%. It was able to record 661 touch gestures and proves that social expression conveyed through touch can be communicated remotely. As the COVID-19 pandemic keeps families separated and with travel restrictions due to the new Omicron strain, such haptic wearables can help people stay “in touch” and express their feelings a little bit better. Culbertson said, “While the physical divide caused by the pandemic is lessening, our families and friends remain spread across the country and the world due to our increasingly global communities. Such tools would allow individuals to reach out and touch their loved ones regardless of physical separation, reducing isolation, bringing us closer.”","excerpt":"Researchers are trying to fulfil physical contact needs through haptic wearables.","categories":["AI Features"],"tags":[],"author_name":"Arnab Ray","publish_date":"2021-12-25T10:00:00","publication_year":"2021","word_count":619,"keywords":["Replicate","Go","machine learning","AI","ML","BERT","ViT","GAN","R","startup"],"extracted_tech_keywords":["AI","machine learning","ML","R","Go","BERT","GAN","ViT","startup","Replicate"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/can-you-express-your-feelings-through-a-haptic-armband-just-like-a-hug\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":9707,"title":"Application of Analytics in Security Systems","content":"Security in today’s world is one of the key concerns for corporations today. It includes both the data (in different forms) and physical instruments\/devises. How secure is my data these days as most of the data is stored at remote locations, which was not the situation a few years ago? Now with the advancement of technology, companies are looking for better security to their information. Major concerns are: How can the data be protected? How trusted is the security system? Privacy concerns Security should be provided when the data is in motion, at rest and while transmitting. Although all companies have their own network firewall and security products, they find it difficult to identify the breaches as there is a huge amount of data that needs to be processed. This is where analytics comes into the picture. It can identify similarities between the past suspicious events with the current data and alert the security systems regarding the breaches. This way it helps to track and protect the security breaches in real-time by taking appropriate steps to prevent the security breach. Financial services: Analytics acts as a new protective shield for the security systems. Analytics can be used in checking suspicious trading patterns in stock markets. This helps in preventing the frauds in stocks trading as they can make huge money out of it. This way analytics helps in determining abnormal trading patterns and prevent the occurrence of fraud on real-time basis. Customer behavioral pattern: A key concern for any company is about the high valued information if breached and falls in the hands of competitors, can cause catastrophe to the company. Analytics helps in incident detection and response to such breaches. Analytics helps in preventing the occurrence of fraud accounts by analyzing customer behavior on different platforms. This way we can prevent the existence of fraud accounts and also reduces the chances of breaches by false positives i.e. showing the presence of threat even though the threat doesn’t exist. E-Commerce platforms: Analytics can be helpful in analyzing the behavior of users while making payments and prevent them from making fraudulent purchases. It is difficult for an employee to analyze such massive amounts of data and develop possible insights. Analytics helps in constantly analyzing a huge amount of data in real-time and provides with fast and factual insights about possible threats. Video surveillance: Analytics can also be used in a video surveillance system in detecting the faces of frauds, helpful in analyzing any suspicious patterns in the behaviour of the people in huge numbers during large events and helps in taking preventive measures. By having predictive capabilities built-in analytics that helps in improving the security systems of the companies. This way analytics help in highlighting the anomalies and alert the security members. Video analytics is also being used in security systems to prevent fraud these days. Analytics is playing a crucial role to detect potential threat using security systems.","excerpt":"Security in today’s world is one of the key concerns for corporations today. It includes both the data (in different forms) and physical instruments\/devises. How secure is my data these days as most of the data is stored at remote locations, which was not the situation a few years ago? Now with the advancement of […]","categories":["IT Services"],"tags":["International Affairs","network firewall"],"author_name":"GlobCon Technologies pvt ltd","publish_date":"2016-04-26T13:27:36","publication_year":"2016","word_count":488,"keywords":["Go","network firewall","programming_languages:R","programming_languages:Go","analytics","Rust","International Affairs","R","programming_languages:Rust"],"extracted_tech_keywords":["analytics","R","Go","Rust","programming_languages:R","programming_languages:Go","programming_languages:Rust"],"url":"https:\/\/analyticsindiamag.com\/it-services\/application-analytics-security-systems\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":33112,"title":"Best Open Source Platforms For Raspberry Pi Retro Game Addicts","content":"From saving the Mushroom Princess from the evil Koopa to being the last man standing in the battle royale, video games have evolved exponentially. It’s not only the games that have evolved, but also the hardware and consoles. It would be hilarious if you actually compare today’s graphics and processors with the ones from the ’80s or ’90s. But for those who want to take a walk down the memory lane, the tiny but powerful Raspberry Pi is the perfect platform. If you’ve got a Raspberry Pi, then you are almost there to relive retro gaming because with this tiny computer and a retro gaming suite, you can play old-school games with flair. Here Are Some Of The Best Open Source Platforms For Raspberry Pi To Play Old-School Games: Recalbox Originally programmed for Raspberry Pi, Recalbox OS is a free, open source software distribution for retro gaming. The OS is based on the GNU\/Linux Operating System and is loaded with thousands of open source software. Also, this amazing OS offers a wide range of game systems — from the very first arcade systems to the NES, the MEGADRIVE, 32-bit platforms and even Nintendo 64. Unlike some of the other platforms, Recalbox has fewer emulators, fewer customisation options, and doesn’t have a huge user community. However, it has a very simple setup, one doesn’t have to dig deep on the technical stuff such as getting a new SD. So, if you want to have easy-peasy retro gaming experience, then Recalbox is definitely for you. RetroPie There are many retro-gaming platforms available on the web but when it comes to Raspberry Pi, RetroPie is probably one of the most popular retro-gaming platforms. This amazing platform turns your Raspberry Pi, ODroid C1\/C2, or PC into a retro-gaming machine. When it comes to emulating classic desktop and console gaming systems, RetroPie is definitely a solid and great option. Apart from all the mainstream retro-gaming, this platform also provides a large variety of configuration tools for power users to customize the system as per the users’ wants and needs. That’s is not all, one can install RetroPie on top of a full OS, on an existing Raspbian, or even can go for RetroPie image — the user has got all the power. Unlike Recalbox, RetroPie needs a micro SD of minimum 8gb of space. However, just like Recalbox, it is pretty simple to set up — burn RetroPie image to an SD card, configure controllers, copy your games, and and you are good to go. Lakka Built on RetroArch and Libretro ecosystem, Lakka is an open source, lightweight Linux distribution that transforms a small computer into a retro-gaming console. Lakka is easy to set up and use — install on an SD card or USB flash drive, copy your ROMs on the device, power up the rig and plug your joypad to play. When it comes to gaming, hardware plays a vital role and most of the times hardware costs are high. However, Lakka keeps the hardware requirement as cheap as possible; the software is developed in such a way that it can run fast even on low-end PCs. RISC OS Released in 1987, RISC OS is an open source computer operating system originally created by Acorn Computers Ltd in Cambridge, England.  RISC OS used to be the heart and soul of Acorn Computers, but with time, both the OS and Acorn disappeared. However, as retro gaming is again coming into the play, this amazing OS is gaining a bit of popularity again. RISC OS is a beast when we talk about open source platforms and this can still be installed. Compared to other platforms, RISC OS is not a great all-rounder — it has its own limitations. Also, installing RISC OS and getting started with retro games is not that easy. There are many hidden things that need to be taken care before playing. Witnessing the popularity that Raspberry Pi is gaining with time, a new version has also hit the market called the RISC OS Pi, which is supplied on a specially-branded 2GB micro SD card. It has been created and tested by RISC OS Open specifically to avoid the hassle of creating your own bootable card for the Raspberry Pi.","excerpt":"From saving the Mushroom Princess from the evil Koopa to being the last man standing in the battle royale, video games have evolved exponentially. It’s not only the games that have evolved, but also the hardware and consoles. It would be hilarious if you actually compare today’s graphics and processors with the ones from the […]","categories":["AI Trends"],"tags":["Raspberry Pi"],"author_name":"Harshajit Sarmah","publish_date":"2019-01-09T04:38:53","publication_year":"2019","word_count":708,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","ViT","Raspberry Pi","R"],"extracted_tech_keywords":["AI","R","Go","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/best-open-source-platforms-for-raspberry-pi-retro-game-addicts\/","complexity_score":2,"technical_depth":6,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10131223,"title":"Telangana Becomes India’s First State to Develop its Own AI Model","content":"Lately, there has been a debate in India’s AI ecosystem about whether the country should build its own foundational models. Some argue that India can address real-world problems by leveraging existing state-of-the-art models without spending millions on new ones. Others, however, believe that it’s essential to develop models that profoundly understand the nuances, complexities, and rich diversity inherent in India’s myriad cultures and languages. Amidst all this, an Indian state government has undertaken the task of developing an LLM that operates in the state’s official language. In July, the Information Technology, Electronics & Communications (ITE&C) department in Telangana hosted a datathon aimed at creating a Telugu LLM. Carried out in partnership with Swecha, a non-profit and free open-software movement in Telangana, the datathon was organised to help build datasets which, in turn, will help train the Telugu LLM. Building Telugu Datasets Building effective LLMs for Indian languages remains a challenging task due to the scarcity of high-quality data. While ChatGPT is impressive because it is trained on multiple terabytes of data, such extensive datasets are not available for Indian languages. To develop datasets in Telugu, the Telangana government is tapping into its rich education system. Around 1 lakh undergraduate students across all engineering colleges in Telangana took part in the datathon and collected data from ordinary citizens who use Telugu as their mother tongue. The team collected data from oral sources such as folk tales, songs, local histories, and information about food and cuisine. Additionally, they plan to dispatch volunteers to approximately 7,000 villages across the state to gather audio and video samples of people discussing various topics, which were then converted into content. (Source: @nutanc) Interestingly, this is not the first instance when such an exercise was undertaken. Last year, the same Swecha team developed a Telugu SLM, named ‘AI Chandamama Kathalu’, from scratch. To collect data for the model, a similar datathon was organised with volunteers from Swecha, in collaboration with nearly 25-30 colleges. Over 10,000 students participated in translating, correcting, and digitising 40,000-45,000 pages of Telugu folk tales. Building LLMs Ozonetel, which is an industry partner for the project along with DigiQuanta and TechVedika, supported it by training the model and providing the necessary compute. The team tried fine-tuning Google’s MT-5 open-source model, Meta’s Llama, and Mistal. However, they finally settled on building a model similar to GPT-2 from scratch. Training the model on a cluster of NVIDIA’s A100 GPUs took nearly a week. Now, the aim is to develop a larger model and have it ready to be showcased at the Telangana govt’s Global AI Summit, scheduled to take place in September this year. Moreover, developing a large model could cost millions of dollars. For instance, building something like ChatGPT could cost in the billions. However, the team aims to develop the Telugu LLM at a cost of around INR5-10 lakh. India’s Efforts to Build AI Models in Regional Languages Over time, we have seen efforts to build AI models in regional languages. For instance, Abhinand Balachandran, assistant manager at EXL, released a Telugu version of Meta’s open-source LLama 2 model. Similarly, in April this year, a freelance data scientist released Nandi– built on top of Zephyr-7b-Gemma, the model boasts 7 billion parameters and is trained on Telugu Q&A data sets curated by Telugu LLM Labs. Interestingly, such models have not been just limited to Telugu. We have seen AI models built on top of open-source models such as Tamil Llama, and Marathi LLama, among others. However, these models could be seen as mere experiments. But the Telangana government’s effort to develop an AI model in Telugu has the potential to make significant strides in advancing regional language technology and preserving cultural heritage. Officials involved in the project have told the media that voice commands from devices such as Alexa are not available in Telugu and this platform will pave the way for such innovations.","excerpt":"In July, the Information Technology, Electronics & Communications (ITE&C) department in Telangana hosted a datathon aimed at creating a Telugu LLM datasets.","categories":["AI Trends"],"tags":["India AI"],"author_name":"Pritam Bordoloi","publish_date":"2024-08-02T15:00:00","publication_year":"2024","word_count":649,"keywords":["India AI","Go","ChatGPT","AI","SLM","Git","RAG","GPT","Aim","GAN","R"],"extracted_tech_keywords":["AI","ChatGPT","Aim","SLM","RAG","R","Go","Git","GPT","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/telangana-becomes-the-first-state-in-india-to-build-its-own-ai-model\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10058049,"title":"How bias creeps into large language models","content":"Language models (LM) are optimised to mirror language systems. Therefore, it stands to reason that LMs might perpetuate stereotypes and biases hardwired into the natural language. When the training data is discriminatory, unfair, or toxic, optimisation leads to highly biased models. Red flags In 2019, researchers found racial bias in an algorithm used on over 200 million people in the US to predict the patients who need extra medical care. The system favoured whites over people of colour. The use of COMPAS (Correctional Offender Management Profiling for Alternative Sanctions) algorithm in US court systems projected double false positives for recidivism in black offenders (45%) as opposed to whites (23%). Amazon has scrapped its AI recruitment tool for its manifest sexist bias against women. According to DeepMind, unmodified LMs tend to assign high probabilities to exclusionary, biased, toxic, or sensitive utterances if such language is present in the training data. The express purpose of language modelling is to represent language from the training corpus accurately. Therefore, it’s important to redact and curate training data, fine-tune LMs to adjust weightings to avoid bias and implement checks to filter harmful language. At present, the turnaround time for language models– from research to applications– are relatively short, making it harder for third parties to anticipate and mitigate risks. Therefore, the correction course should start at the research level to address the bias in language models and should be improved on with each iteration. LMs should be evaluated against normative performance thresholds. But, to determine what constitutes satisfactory performance for the LM to be dubbed ‘sufficiently’ safe and ethical before deploying in the real world is a challenge in itself. Key risk areas DeepMind mentions Discrimination, Exclusion and Toxicity as top risk areas in large-scale language models. LMs can engender discrimination, representational, and material harm by perpetuating social biases and stereotypes. For eg, the name ‘Max’ is used for a ‘male’ or ‘families’ always means a father, mother and child. If LMs pick up on biased social cues, they tend to deny or burden identities that differ. LMs run the risk of disseminating false or misleading information. Examples include bad legal or medical advice leading to unethical or illegal actions. While interacting with conversational agents or chatbots, users tend to overestimate the capabilities of the AI and use it in unsafe ways. In addition, LM-based conversational agents might also compromise users’ private information. LMs are used in social engineering to spread fake news, drive disinformation campaigns, create fraud or scams at scale etc. Wrapping up Inarguably, LMs have largely benefited the world economy. However, their benefits and risks are unevenly distributed. Ethical AI and responsible AI are increasingly becoming part of the tech narrative. However, a lot of work needs to be done to incorporate AI ethics into language models. It’s also important to not cut corners for faster turnaround time at the expense of responsible AI. Moreover, the focus should not be solely on building better models but on looking at existing models and devising ways to mitigate their biases.","excerpt":"According to DeepMind, unmodified LMs tend to assign high probabilities to exclusionary, biased, toxic, or sensitive utterances if such language is present in the training data.","categories":["AI Features"],"tags":["Conversational AI","DeepMind","Language Models"],"author_name":"Meeta Ramnani","publish_date":"2022-01-11T12:00:00","publication_year":"2022","word_count":506,"keywords":["Go","programming_languages:R","AI","chatbots","programming_languages:Go","AI ethics","responsible AI","Conversational AI","Language Models","conversational agents","R","DeepMind"],"extracted_tech_keywords":["AI","conversational agents","chatbots","R","Go","AI ethics","responsible AI","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-bias-creeps-into-large-language-models\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":45365,"title":"How MatchMove Is Democratising The Mobile Payments Industry","content":"India has achieved the second spot globally with the largest number of financial technology (Fintech) startups.  The growth in fintech sector is happening on the back of several macroeconomic factors such as digital payments infrastructure explosion, technology upgrades, mobile infrastructure, large population, among others. One such startup known as MatchMove is following a similar path and contributing to the larger goal of financial inclusion. Analytics India Magazine caught up with Kumar Srinivasan, CEO of MatchMove India to gain better insights on how the startup is striving to achieve a promising and innovative organisation in the financial sector. Headquartered in Singapore, MatchMove started as a gaming company in the early 2010 and eventually pivoted into a digital payments platform with the mission of being the global platform to move digital cash. The startup has raised a funding of over $50 million from its investors Vickers Venture partners and Credit Saiso. Tech Stack The technology stack at MatchMove is a combination of coding frameworks and languages like Java, CSS, PHP, NodeJS, GoLang, Python and Angular. Mostly the applications are designed on a service-oriented architecture, orchestrated using microservices, Kubernetes and CI\/CD pipeline for deployments. For products development, MatchMove uses cloud technologies by leveraging AWS managed services which are deployable across multiple regions for HA and Data localisation requirements. Flagship Product Regarding the flagship product, MatchMove LightSpeed is a suite of multiple products and features which enables the clients to launch their own branded payments platform as well as enables them to build their own customised white-label app on the go and launch their virtual card instantly. Use Of AIML Artificial Intelligence and Machine Learning techniques play an important role while building customer relationships. The company uses AI-powered bots which assist the users to make balance enquiry, search historical data for any specific transaction with specific merchant while avoiding the hassle of looking through the entire statement(s) or previous transaction history. The developers at MatchMove use multiple data points to build an AI model which helps in predicting, enhancing accountability on Fraud and Regulatory Compliance like limits or KYC process and much more. In addition to traditional metrics that make up the credit rating, the AI-based platform also mines data from the internal spend\/send transactions, social media, geolocations, etc. to find patterns that indicate credit-worthiness, the propensity to default and the likelihood of fraud, etc. in order to provide loan to borrowers. Currently, the product and engineering team includes around 70 people in total. Tackling the Talent Crunch According to Srinivasan, referrals from employees, partners or associates are considered as the most reliable hiring source at MatchMove. “For us, a candidate’s appreciation of our business purpose is very important, and a referred candidate already comes in well introduced to MatchMove. Beyond solid technical skills, we look for candidates with learnability, self-drive and heuristic agility,” he said. Future Roadmap The startup is rapidly expanding to other countries like India, Vietnam, Philippines, Indonesia, and other countries. As a future roadmap, MatchMove will build a number of value-added services on top of the generic platform of Lightspeed product which will help in solving several issues for various domains like corporate expenses management, shipping, payroll management, and much more.","excerpt":"India has achieved the second spot globally with the largest number of financial technology (Fintech) startups.  The growth in fintech sector is happening on the back of several macroeconomic factors such as digital payments infrastructure explosion, technology upgrades, mobile infrastructure, large population, among others. One such startup known as MatchMove is following a similar path […]","categories":["Deep Tech"],"tags":["Financial Services","FinTech","Startups"],"author_name":"Ambika Choudhury","publish_date":"2019-09-03T15:43:02","publication_year":"2019","word_count":531,"keywords":["artificial intelligence","machine learning","AWS","AI","ML","RAG","Financial Services","microservices","Aim","Startups","analytics","FinTech","kubernetes"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","analytics","Aim","RAG","AWS","kubernetes","microservices"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/how-matchmove-is-democratising-the-mobile-payments-industry\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10170064,"title":"Perplexity’s Comet Browser is Now Available in Beta","content":"Perplexity is rolling out Comet, its agentic web browser, as a beta version, limited to select Apple Silicon Mac users for now. Selected testers received early access this week, with the company encouraging feedback on bugs and features as development continues. The browser aims to reimagine web browsing through context-aware intelligence. Comet personalises responses based on a user’s browsing history and open tabs, all stored locally, and not used for model training. It also integrates Perplexity’s core search capabilities into a side panel accessible across all websites, enabling users to query, summarise, or explore content without switching tabs. AIM checked the browser out and observed that it supports importing from Google Chrome, including bookmarks, passwords, and more. While in beta, the browser can take commands from the user through its ‘Comet Assistant’ and close\/open, group tabs, close duplicate tabs, check the shopping cart, help find unanswered emails, and more. While starting the browser, AIM noticed that it begins with the ability to block advertisements and trackers on the web. One can choose to disable that while setting up the browser. The assistant to access Perplexity can be found on any webpage through an ‘Ask’ button in the URL address bar section. Moreover, Comet doubles down on productivity. The email invite highlights use cases such as analysing inboxes, preparing for interviews, and generating meeting notes—functions intended to “save hours every week”. While the product is still developing, Perplexity said that it is shipping improvements daily. “We are shipping daily improvements to the product experience and look forward to your feedback,” the email read. It should be noted that the screenshots of the beta version are subject to change and may not reflect the final look of the Comet browser. The closed beta is currently limited to macOS, with no word yet on broader availability or Windows support.","excerpt":"The agentic web browser is available for users who signed up for early access.","categories":["AI News"],"tags":["Perplexity"],"author_name":"Ankush Das","publish_date":"2025-05-16T10:04:59","publication_year":"2025","word_count":306,"keywords":["Go","Perplexity","programming_languages:R","AI","programming_languages:Go","mlops_tools:Comet","RAG","Aim","ViT","R"],"extracted_tech_keywords":["AI","Aim","RAG","R","Go","ViT","mlops_tools:Comet","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/perplexitys-comet-browser-is-now-available-in-beta\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10051854,"title":"This New Class Of Robots Can Self Assemble On Demand","content":"A group of researchers from the École Polytechnique Fédérale de Lausanne’s (EPFL) Swiss Biorobotics Laboratory (BioRob) have built a new sort of robot, and it is quite the sight to behold! These new robots, named Roombots, are modular shape-changing machines that can be reconfigured in three dimensions. They have even been compared to robotic LEGOs by some. These bots can self-assemble and self-transform into a variety of various pieces of furniture. Additionally, they can also move about and self-assemble on demand. They are constructed with two dice that have been glued together, a battery, three motors that drive the movement, and a wireless connection. This is a Swiss NCCR in Robotics-funded project which investigates the design and operation of Roombots. Self-Reconfiguring Modular Robot Modular robots are made up of numerous simple robotic modules that may be connected and detached. Modular self-reconfiguring robotic systems, alternatively referred to as self-reconfigurable modular robots, are self-configuring kinematic machines with variable morphology. Apart from the standard actuation, sensing, and control present in fixed-morphology robots, self-reconfiguring robots can also modify their shape purposefully by rearranging the connectivity of their components in order to adapt to new situations, perform new tasks, or recover from injury. Functionalities of Roombot The Roombot is a modular robot consisting of numerous connectors and a central hinge. They may self-assemble to become a variety of various pieces of furniture and are capable of autonomous movement. Each module is 22cm in length, and the team envisions that ten of them may be combined to create a diverse selection of furniture. Active Connection Mechanisms (ACMs) establish a connection with the target connector via mechanical latches. ACMs from Roombots are hermaphrodite, which means that both male and female components are connected using the same connector. This means that each ACM is capable of communicating with any other ACM. Additionally, ACMs are compatible with passive ports, which are essentially female connectors. Modules Each module is completely self-contained and equipped with its own set of control boards and batteries. The high-level commands are coordinated by a central host that is capable of communicating with each module individually. Similarly, each module is capable of controlling its own locomotion via a set of settings. An external computer performs the high-level control (i.e. motion planning). The modules and the PC communicate via a wireless Bluetooth link. A rendered image of a Roombots module assembling a table utilising lightweight parts and integrated Roombots modules: Source: Biorobotics Laboratory, EPFL How does Roombot communicate? The Communication Board (BT) utilises a cable communication bus to convey received orders and control parameter sets to the other electronic boards. Electronic slip rings (SR) enable the bus signal to be distributed to either hemisphere. Motor Boards (MB) are capable of receiving locomotion commands and are in charge of low-level control of DC motors as well as the operation of a central pattern generator controller. Three motor boards are included, one for each actuator. Each active connector is fitted with an ACM Control Board (ACM) that regulates the mechanical latches’ opening and closing. A portion of the Roombots project’s current effort is to determine the capabilities of a swarm of Roombots modules through physical testing and to determine whether adding some tools to the otherwise homogeneous module swarm could enable meaningful additional capabilities. The most audacious is perhaps object manipulation, which presents a significant barrier for robots of this type. The EPFL researchers combined a small jamming gripper into a single hemisphere of such a Roombots module, enabling it to pick up the majority of small, stiff objects within its reach. Conclusion It is also worth noting that despite their ability to self-assemble into tables and seats, the present version of Roombots can’t actually sustain that much weight, and so sitting in a Roombots chair would likely smash it. Roombots’ transition to real-world applications would almost certainly require a new redesign, and the researchers are already considering modifications like vision systems, distributed control, and even an “artificial skin” for safe human interaction. It is important to foresee potential robotics. There is a list of some of the books on artificial intelligence and robotics that one should read before embarking on a project.","excerpt":"The only furniture you’ll need in the future will be a swarm of tiny robots that will work together to create whatever you want.","categories":["AI Features"],"tags":["AI (Artificial Intelligence)","AI Tool","Machine Learning","Robotics","Robots"],"author_name":"Dr. Nivash Jeevanandam","publish_date":"2021-10-19T13:00:00","publication_year":"2021","word_count":694,"keywords":["Go","artificial intelligence","programming_languages:R","AI","Machine Learning","programming_languages:Go","Robotics","Robots","ViT","ai_applications:robotics","AI Tool","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","R","Go","ViT","programming_languages:R","programming_languages:Go","ai_applications:robotics"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/roombots-modular-shape-changing-machines\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10104246,"title":"Yotta Challenges Hyperscalers with India’s First AI-Centric GPU Cloud","content":"The global datacentre market is projected to reach $325.90 billion in 2023, with an annual growth rate of 6.12%, pushing the market volume to $438.70 billion by 2028. The United States is expected to lead with 5,375 datacentres, clocking a revenue of $95.58 billion in 2023. Meanwhile, the Indian datacentre market, with 151 datacentres currently, is forecasted to experience a surge in business from $6.12 billion in 2023 to $10.89 billion by 2028, showing an impressive 12.22% CAGR. Generative AI is poised to revolutionise the datacentre sector and is projected to create a demand wave comparable to the cloud. India aims to capitalise on this opportunity and strengthen its datacentre capacity. As part of this initiative, several leading datacentres, including Yotta, a Mumbai-based datacentre giant, anticipate acquiring NVIDIA GPUs and expanding their infrastructure accordingly. While Yotta currently has about 700 GPUs, in an exclusive interview with AIM, managing director Sunil Gupta revealed they are slated to launch their Yantra and Shakti Clouds in early 2024. Yantra will be a hyperscale cloud targeting government, enterprise, and startup sectors. Shakti Clouds is slated to be India’s first AI-centric GPU-based cloud, consisting of 16,384 NVIDIA GPUs. “It will be a mixed bag of H100s and the recently launched L40S. I would say that about 1,000 of the 1,600 ordered are L40S; however, the majority are H100s,” Gupta detailed the GPU specifications. The roadmap includes an ambitious plan to operationalise 4096 GPUs by January 2024, expanding to a staggering 16,384 GPUs by June 2024. Furthermore, Gupta explained that Yotta is poised to elevate its GPU infrastructure to an unprecedented scale of 32,768 by the end of 2025. “I’ve been told by NVIDIA themselves that for the capacity that we’re building, Yotta might become one of the ten largest supercomputers in the world,” Gupta remarked. As India looks to bolster its capacity from 800 megawatts to a staggering 3,000 megawatts by 2030, Gupta’s vision encapsulates building raw datacentres and establishing India’s sovereign AI and cloud infrastructure. He articulated, “Capacity building is crucial, but equally important is constructing our own AI and cloud ecosystem for the country’s benefit.” “There’s a lot of support and encouragement from the government because this is something which we feel India needs—because India will need its own LLM and a model on Indian data.” Infrastructure Overhaul The surge in Generative AI presents significant challenges in adapting datacentres for AI applications, not just from the infrastructural aspect but also from the software end. Gupta highlighted the challenges, noting the estimated seven to eight times increase in power and cooling density required for GPU integration: “The underlying infrastructure required to support these GPUs is just crazy… instead of some 678 kilowatts per array, you’re talking about 40 to 60 kilowatts per rack.” He outlined plans for improvements in cooling infrastructure, aiming to reduce PUE(Power usage effectiveness) with technologies like immersion cooling and direct liquid cooling to reduce PUE further to sub 1.1.” A comprehensive software layer is also being developed to streamline GPU access for startups, providing an orchestration layer akin to regular cloud services. This involves a user-friendly interface for startups to subscribe to plans, access pre-trained models, integrate their data, and create them seamlessly. “So, the startups come, create the accounts, subscribe to some plans, subscribe to some capacity, get the capacity, I give them the tools, you know, the pre-trained models, they can take those trained models, put their data, and then make their model,” said Gupta. Distinctiveness From Hyperscalers Gupta discussed the unique aspects of Shakti Cloud compared to providers like AWS, Google Cloud, and Azure, noting the scarcity of GPUs and how hyperscalers are turning to companies like theirs due to allocation trends. He expressed ambitions to rival top cloud operators, promising to provide services others might not attempt, starting with the latest H100 Tensor Core GPUs for their AI cloud. “I will be as good as Amazon, Azure, Google… while being a sovereign cloud operator,” Gupta said, outlining his ambition regionally, including other unexplored markets. Additionally, he highlighted a unique marketplace with foundational models for startups to leverage: “I’m building my AI cloud,  beginning directly with H100 GPUs… including all those foundational models.” Furthermore, Gupta emphasised Shakti Cloud’s distinctive multi-cloud architecture, enabling seamless management of various cloud resources. “By default, we are doing something that you come to my orchestration layer to register, you not only can see the catalogue of and consume my cloud services, but you can also see the catalogue of and consume AWS or Azure services as well right from my orchestration.” Demand and Expansion Plans When it comes to hyperscalers’ interest in India, Gupta highlighted their move into emerging markets not just for cost-effectiveness but to serve India’s growing economy better. He explained that these companies initially prefer partnering with providers like Yotta to navigate regulations and construction challenges but hinted at future infrastructure plans as they mature in the market. While the primary usage of Yotta’s GPUs previously was for graphics workstations by studios, mainly for tools like Maya, VFX effects, and online game creation—Gupta pointed out the diverse stakeholders in need of GPU capacities now include entities like ONGC and FTS, amongst other startups, enterprises etc. However, this infrastructure is not solely for India; Gupta’s expansive vision encompasses serving underserved markets in Southeast Asia, the Middle East, Africa, and beyond, leveraging the potential demand for GPU cloud services globally. “My fundamental thought was to go to underserved markets, where demand exceeds the local supply in the local market,” Gupta said, highlighting specific regions targeted for expansion, including Southeast Asian countries like Thailand, Philippines, Vietnam, Indonesia, and Malaysia. Moreover, their GPU cloud services attract global interest, indicating potential demand from unexpected markets like the UK.","excerpt":"The roadmap includes 4k GPU cluster, scaling up to 16,384 GPUs by June 2024. The intent is to reach an unprecedented 32,768 GPUs by late 2025.","categories":["AI Features"],"tags":["Yotta"],"author_name":"Shyam Nandan Upadhyay","publish_date":"2023-12-06T13:22:54","publication_year":"2023","word_count":954,"keywords":["Go","AWS","AI","Yotta","R","ML","RAG","Ray","Aim","generative AI","Azure"],"extracted_tech_keywords":["AI","ML","generative AI","Aim","Ray","RAG","AWS","Azure","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/yotta-challenges-hyperscalers-with-indias-first-ai-centric-gpu-cloud\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10167928,"title":"GoUpscale Gets a Boost as Accenture Invests in Wealth Engagement","content":"Accenture has invested in GoUpscale, a tech company that helps wealth management firms in Asia improve how they communicate with clients through digital content. The investment was made through Accenture Ventures on Monday. The move comes at a time when wealth management firms are looking for better ways to engage clients, especially online. According to Accenture, these firms plan to nearly double their assets under management to around $260 trillion by 2026, with 95% of that growth expected to come from relationship managers. GoUpscale uses artificial intelligence to turn traditional materials like brochures and reports into interactive, digital experiences. These tools aim to help wealth managers connect more meaningfully with clients. “With relationship managers remaining one of the most important channels for clients, firms need to scale human touch without scaling cost,” David Wilson, who leads Accenture’s wealth management practice in Asia, mentioned. “The investment content solution from GoUpscale empowers relationship managers to deliver personalised content that transforms everyday interactions into meaningful moments of connection and value,” he added. Dominic Gamble, CEO and co-founder of GoUpscale, said, “The investment from Accenture will allow us to scale our AI capabilities, refine our platform, and bring our solutions to more firms seeking to enhance their client engagement strategies.” Accenture and GoUpscale will work together to bring the platform to more markets, using Accenture’s experience in AI and its network of clients to support expansion. Accenture’s research also suggests that using more relevant, easy-to-understand content could increase yearly revenue by 6–8% for wealth management firms. As part of the deal, GoUpscale will also join Accenture Ventures’ Project Spotlight, a programme that helps AI and data startups connect with industry experts and potential customers. Meanwhile, in its latest earnings call, Accenture reported a revenue of $16.66 billion for Q2, marking a 5.8% decline from the previous quarter, though still 5.4% higher than the same period last year. However, a key factor in the revenue slowdown stems from uncertainty in US government contracts, which account for a significant portion of Accenture’s earnings. The review of federal contracts by US President Donald Trump’s administration has further added to the challenges.","excerpt":"GoUpscale uses artificial intelligence to turn traditional materials like brochures and reports into interactive, digital experiences.","categories":["AI News"],"tags":["Accenture"],"author_name":"Shalini Mondal","publish_date":"2025-04-14T17:36:39","publication_year":"2025","word_count":354,"keywords":["Accenture","Go","artificial intelligence","programming_languages:R","AI","programming_languages:Go","Git","Aim","R","startup"],"extracted_tech_keywords":["AI","artificial intelligence","Aim","R","Go","Git","startup","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/goupscale-gets-a-boost-as-accenture-invests-in-wealth-engagement\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10141070,"title":"Mistral Teases Release of Multimodal Models—Mistral Large 3 and Pixtral Large","content":"Mistral hinted at the release of its new Large 3 and Pixtral Large models, available via API-only for now, leaving the AI community eager for open weights and greater accessibility. The update came to light after the company released two new models—the Mistral Large 2411 and Pixtral Large 2411—on the Mistral console. Moreover, the codename suggests that Mistral might officially release these models, with more information on November 24, 2024. This is in line with the naming scheme of previous models, as the Mistral Large 2407, was released on July 24, 2024. A newer version of Pixtral, Mistral’s multimodal model was also spotted in the list. Titled Pixtral Large 2411, it is also possible that Mistral is launching another model on the same date. Apart from that, Mistral hasn’t revealed any details yet. While they recently released a small language model called Mistral 3B in October, users have had to wait longer for an update to their larger, flagship model. Similarly, a 12 billion parameter Pixtral model was released in September of this year, and a ‘large’ Pixtral model indicates a massive upgrade to the model. Mistral’s most recent large model achieved an 84.0% accuracy on the MMLU benchmark, a new record for performance and cost efficiency in open models. It also provides neck-to-neck competition in code generation and reasoning with GPT 4o and Llama 3. The Pixtral12B model, with its 128k context window, showcases strong multimodal capabilities and scored 52.5% in the MMLU benchmark, surpassing many of its competitors in the small model segment. Along with dense models, Mistral’s portfolio features two mixtures of experts and MoE models—the Mixtral 8x7B and the 8x22B. Currently, Mistral is valued at $6 billion, and their latest fundraising in July amounted to $640 million dollars. It’s interesting how Mistral’s new models will compete in the open source world, which is currently heating up, thanks to models from Alibaba, and Tencent from China shaking up the open source ecosystem.","excerpt":"If Mistral’s current naming scheme is anything to go by, we may have new models on 24th of November.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","mistral","Open Source AI"],"author_name":"Supreeth Koundinya","publish_date":"2024-11-17T20:33:13","publication_year":"2024","word_count":325,"keywords":["API","AI","Modal","RPA","ML","Open Source AI","Pixtral Large","GPT","Mistral Large 2","mistral","R","AI (Artificial Intelligence)","llm_models:GPT"],"extracted_tech_keywords":["AI","ML","Mistral Large 2","Pixtral Large","R","API","GPT","RPA","Modal","llm_models:GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/mistral-teases-release-of-multimodal-models-mistral-large-3-and-pixtral-large\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":5332,"title":"Interview &#8211; Srihari Srinivasan, Head of Technology at ThoughtWorks India","content":"[dropcap size=”2″]AIM[\/dropcap]Analytics India Magazine: What are some of the main tenets (philosophies, goals, attributes) of analytics approach and policies at your organization? [dropcap size=”2″]SS[\/dropcap]Srihari Srinivasan: At the outset it will be good to clarify that we are essentially a custom software company not an analytics services company. TW has been a pioneer at adopting the Agile approach to software development. As we now foray into building newer analytics solutions our approach is largely based on adapting the Agile\/Lean techniques to the nuances of data engineering. Our unique approach can be best described as the intersection between agile\/lean delivery methods, advanced statistical techniques and distributed systems engineering with a keen sensitivity towards data privacy. AIM: What are the next steps\/road ahead for analytics at your organization? SS: While we have made some significant inroads over the last year or so at TW India, the Analytics practice here is still in its early stages. Given where we are in the lifecycle one of our big focus areas is to grow our internal capacity to service some interesting analytics opportunities. We are also noticing that the adoption of several Big Data technologies is beginning quickly move past the proof of concept stage in many large enterprises. In response to this a key next step for us is to quickly adapt our development, testing, deployment and operations approach to suit large-scale data engineering problems. Accomplishing this is a very interesting technical challenge for talented data scientists and engineers. Last but not the least, we also aspire to create new solutions for personalizing the online experience for the consumer internet and e-commerce applications. These solutions will be specifically targeted for emerging markets where e-commerce and consumer internet is still burgeoning. AIM: What are the most significant challenges you face being in the analytics space? SS: From the perspective of analytics software and data engineering we find ourselves tackling a fairly heterogeneous set of challenges of late – Firstly there is the challenge of data quality that comes from handling data from diverse sources. This is not entirely a new issue and has plagued the data-warehousing world for a long time now. A significant amount of time and effort is still being spent on distinguishing issues that occur due to bad data from the logical errors in the code. While there is no silver bullet for solving this we see teams manage this by constructing different forms of anti-corruption layers as part of their data processing pipelines. Good software generally is an outcome of the collaboration between technical folks and subject matter experts. From that perspective building high quality analytics systems hinges quite a lot on the collaboration between data scientists and engineers. While data science does require some programming, data scientists are not necessarily the most adept at modern software engineering practices. Conversely, engineers too have to learn a fair amount of statistics and machine learning in order to operationalize the models produced by data scientists.  The learning curve that people in either role go through on early projects still remains a big challenge. There are benefits that can be gained by capturing the digital trails of users. At the same time it’s worth noting that these trails may as well end up as indelible records. In Europe, many companies are adopting a strategy of Datensparsamkeit, a term that roughly translates as “data austerity” or “data parsimony.” The term Datensparsamkeit is taken from German privacy legislation and describes the idea to only store as much personal information as is absolutely required for the business or applicable laws. This is certainly one way that privacy can be maintained even in the unfortunate event of a data breach. The challenge of delivering a more personalized user experience while remaining sensitive to data privacy concerns deserves special mention in the analytics context. AIM: How do you see Analytics evolving today in the industry as a whole? What are the most important contemporary trends that you see emerging in the Analytics space across the globe? SS: All product development activities within the Big Data landscape can broadly be categorized into – Infrastructure solutions and Applications. While the infrastructure solutions have already reached a stage of maturity there are still some optimization efforts going on in this space. Managed Platforms – Open Source projects such as the Savanna platform from Open Stack and a host of Hadoop-as-a-Service offerings from different commercial organizations are trying to make it easier to deploy Hadoop in multi tenant cloud environments. SQL-on-Hadoop – The SQL-on-Hadoop trend continues to make progress with solutions like Apache Drill and Impala. These solutions aim to bring the familiar experience of working with the ubiquitous SQL language to the Hadoop platform. This perhaps is most significant innovation that will drive Hadoop’s adoption within enterprises. Efficient cluster management – Soon after enterprises adopt different distributed processing frameworks the question of utilizing the clusters more efficiently comes up. This is something the distributed systems community has been trying to address for a while now. Cluster managers such as Apache Mesos, provide efficient resource isolation and sharing across a pool of machines instead of having dedicated pools of machines for each distributed processing framework. Intelligent Data Curation – As infrastructure solutions become more mature we are beginning to see a gradual shift in the trends and investments away from infrastructure and towards applications. New data integration solutions are emerging that enable organizations to curate data from heterogeneous sources very efficiently at scale. This is a space to watch out for in the coming months. [divider] Biography of Srihari Srinivasan Srihari is the Head of Technology for ThoughtWorks India. He’s been a developer and architect for several enterprise applications with focus on building large scale systems based on service oriented architectures, domain specific languages etc. He is quite passionate about distributed systems and databases and blogs about them on www.systemswemake.com.","excerpt":"[dropcap size=”2″]AIM[\/dropcap]Analytics India Magazine: What are some of the main tenets (philosophies, goals, attributes) of analytics approach and policies at your organization? [dropcap size=”2″]SS[\/dropcap]Srihari Srinivasan: At the outset it will be good to clarify that we are essentially a custom software company not an analytics services company. TW has been a pioneer at adopting the […]","categories":["AI Features"],"tags":["Interviews and Discussions"],"author_name":"AIM Media House","publish_date":"2014-03-05T15:55:47","publication_year":"2014","word_count":976,"keywords":["data science","Go","machine learning","AWS","AI","Ray","Aim","analytics","SQL","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","machine learning","data science","analytics","Aim","Ray","AWS","R","SQL","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/interview-srihari-srinivasan-head-of-technology-at-thoughtworks-india\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10101965,"title":"What’s Up with ChatGPT Enterprise?","content":"OpenAI has been going through a lot of ups and downs ever since it released ChatGPT. From acquiring millions of users to seeing a dip in revenue, from reporting losses, to finally releasing ChatGPT Plus and Enterprise for profit-making. However, it seems like the company has been successively taking a beating when it comes to sales, as people are looking for other cheaper alternatives. For enterprise, GPT-4 is 50 times more expensive than Llama 2, specifically for the summarisation of the Wikipedia text into half its size. One can only imagine the cost for countless other use cases. Meanwhile, just a year after the launch of ChatGPT, Salesforce and Wix, which were its early customers, have decided to explore other options. They’re now frolicking with rival AI providers, seeking cost-effectiveness upon finding the alternatives budget-friendly. Salesforce has been one of the first customers and poster boys to use OpenAI’s GPT-4, using them to automatically draft emails or distil endless meeting chatter into manageable snippets. But now, Salesforce is also eyeing open-source models and creating their in-house Einstein GPT, which rumour has it, is less expensive for them. As Jayesh Govindarajan, the senior vice president of AI at Salesforce, puts it, “We’re at the very beginning of this cost-reduction exercise in AI. It’s only going to become more important as these AI products reach greater scale and we begin to focus on achieving cost effectiveness.” Even Zoho, which currently uses OpenAI products, is working on developing its own cost-effective, proprietary LLMs, and is in talks with NVIDIA for GPU acquisitions. Morgan Stanley, another flagship OpenAI customer, is testing out Microsoft Azure’s other offering, exploring whether the Azure service is a long-term match for OpenAI’s. Wix, too, once a dedicated OpenAI fan, is eyeing the competition, testing open-source models and even Google’s offerings to cut costs. The list goes on and on. What’s the issue with ChatGPT Enterprise? When OpenAI announced that it would release ChatGPT Enterprise soon, we said that it would fail, as Microsoft was just using OpenAI to do the dirty work. But when OpenAI actually launched it, it seemed as though the company had finally learned how to do business, directly from Microsoft. Most LLM applications don’t need a generalist model. For those, you’ll save a ton of money and get better accuracy by fine-tuning something like Mistral-7B.— Mark Tenenholtz (@marktenenholtz) October 17, 2023 But now, Microsoft, the so-called OpenAI backer, is choking ChatGPT Enterprise. According to its recent earnings report, the tech giant’s revenue has risen 13% to $56.5 billion, as the sale of its AI cloud through Azure has accelerated, all thanks to OpenAI’s GPT. When customers buy OpenAI through Azure, Microsoft snags a fatter slice of the pie. Moreover, enterprises have been considering the costs of building LLMs, and are finding open source models cheaper for them, that includes going through Microsoft. Though it is not entirely clear if Microsoft’s cloud sale through AI investment is just because of OpenAI’s offering, or if it also includes other open source offerings. But Microsoft is dabbling in the open-source playground to cut costs as well. For example, it also provides Meta’s Llama 2 on its cloud platform. Developers are finding that these open-source offerings can replace OpenAI’s models for less demanding chores. On the other hand, while its revenue has skyrocketed as CEO Sam Altman said, and the company is on its way to the billion-dollar club, most of that moolah for OpenAI comes from ChatGPT Plus subscriptions. It is still learning the ropes in the enterprise world. When it comes to open source, OpenAI had already raised the alarm on open source AI dangers long back. This was probably because it realised that these models would turn out to be dangerous for the company. LLMs' hallucination is a major problem we face in enterprise AI application. While it is not a big deal if you occasionally get some incorrect answers when you chat with ChatGPT, it would be embarrassing if AI outputs nonsense when reviewing a legal document.— Fei Li (@lifeitech) September 30, 2023 Apart from this, LLM hallucinations remain one of the biggest problems for ChatGPT Enterprise. It might be fine for a chatbot to hallucinate when asking random questions, but when it comes to handling legal and financial documents, hallucinations cannot be dismissed as a feature — it’s a bug. But OpenAI is continually fixing the hallucinations problem, and its autonomous AI agents, which are slated for release next month, would make it all possible. Busy impressing customers, not enterprise Recently, OpenAI made DALL-E 3 available for all the ChatGPT Plus and ChatGPT Enterprise users. But, enterprise customers are still not impressed as they crave more specialised business use cases, and not generalised. OpenAI isn’t entirely on the losing side. Some customers of other AI services (AWS SageMaker, Google’s Vertex AI, etc.) are seeking more variety. Fidelity Investments, for instance, which was loyal to Amazon’s SageMaker, has started testing with Microsoft’s Azure OpenAI Service to give OpenAI’s models a whirl. But there are still a lot of use cases that OpenAI has to explore directly, and not via Microsoft. The Altman-led firm is also at the risk of being outshone by open-source models, which are smaller and simpler but pack enough punch for many tasks. Mistral AI’s new models have been also outperforming OpenAI’s generalist models, and enterprises are utilising them. For instance, Pete Hunt, the founder of developer tools startup Dagster, switched from OpenAI’s GPT-3.5 model to a nifty open-source model from Mistral AI, for his Summarize.tech service, saving a bundle without compromising on quality. He’s now eyeing that elusive mortgage payment, thanks to open-source cost savings. The same goes for Oracle’s billion-dollar baby, Cohere, which is seeing seamless enterprise-wide adoption at scale. OpenAI’s ChatGPT is nowhere to be found in enterprise, except in Microsoft Azure OpenAI Service. Even the early adopters are yet to figure it out.","excerpt":"Just a year post the launch of ChatGPT, Salesforce, Morgan Stanley, and Wix, which were its early customers, are exploring other options","categories":["AI Features"],"tags":["OpenAI"],"author_name":"Mohit Pandey","publish_date":"2023-10-25T13:42:44","publication_year":"2023","word_count":985,"keywords":["Go","ChatGPT","TPU","AWS","OpenAI","AI","R","ML","GPT","Azure"],"extracted_tech_keywords":["AI","ML","ChatGPT","OpenAI","AWS","Azure","TPU","R","Go","GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/whats-up-with-chatgpt-enterprise\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10129594,"title":"NVIDIA Invests in Arrcus, Networking Software Startup with Bengaluru Footprint","content":"San Jose-headquartered company Arrcus Inc. successfully raised $30 million in funding from a consortium of investors led by Nvidia Corp., aimed at bolstering its platform designed to streamline data traffic management for enterprises. Arrcus boasts a customer base that includes SoftBank Corp. and Target Corp., and specialises in optimising data flow across networks crucial for enterprises, data centres, and users. The startup disclosed plans to utilise the capital infusion to scale its global operations and further enhance its technology infrastructure. Arrcus’ ACE platform, powered by NVIDIA BlueField DPU, employs a distributed microservices architecture that offers flexibility, high performance, scalability, full programmability, modularity, and readiness for hybrid cloud environments. It supports diverse deployment options such as data processing units (DPUs), merchant silicon, and compute infrastructure. This enables a wide array of applications including low-latency data centre networking, modern edge setups, telecommunications access and transport, and hybrid multi-cloud connectivity. “We are thrilled to welcome NVIDIA as our latest investor and look forward to building on our collaboration. Arrcus’ leading networking software coupled with NVIDIA’s AI infrastructure will help deliver maximum efficiency to customers from data centres as well as their edge and cloud computing environments,” said Arrcus CEO and chairman, Shekar Ayyar. India Presence Founded in 2016, the startup has around 150 employees and a significant number of them work out of the startup’s campus in Bengaluru. A few years ago, the Bengaluru team expanded its support and engineering team. Saudi Aramco’s venture arm Prosperity 7 Ventures, Hitachi Ventures, and General Catalyst, were some of the prominent backers of this recent funding round. It’s interesting to note that NVIDIA has been investing and acquiring startups that are providing solutions to enterprises through development or networking platforms. This week, NVIDIA acquired San Francisco-based AI development platform Brev.","excerpt":"A few days ago NVIDIA acquired San Francisco-based AI development platform Brev.","categories":["AI News"],"tags":["NVIDIA"],"author_name":"Vandana Nair","publish_date":"2024-07-19T13:07:18","publication_year":"2024","word_count":294,"keywords":["Go","API","AI","cloud computing","ML","Scala","microservices","Ray","Aim","NVIDIA","R"],"extracted_tech_keywords":["AI","ML","Aim","Ray","cloud computing","microservices","R","Go","Scala","API"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/nvidia-invests-in-arrcus-networking-software-startup-with-bengaluru-footprint\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10055909,"title":"Top IT Certifications To Pursue In 2022 For A Software Job","content":"Both worldwide and in India, IT tops as a stable and safe career choice. In fact, InformationWeek 2021 IT Salary Survey said that as technology became central to operating plans for businesses and other organisations during the pandemic, 50% of survey respondents said that a career path in IT is more secure than most others as compared to 39% of respondents who said something similar in last year’s survey. The job market in India is extremely competitive, especially in the IT and software sectors. This is fueled by the magnitude of engineers India produces each year. Due to the emergence of new-age technologies, it is becoming even more difficult for IT professionals to get good jobs as upskilling is creating a big issue. But, upskilling is a must if one has to survive in the ever-evolving tech field. How To Become A Cybersecurity AnalystHow to transition to data science roles from a non-analytics backgroundCybersecurity As A Career Option: Here’s What You Should KnowHow to Prepare for Certified Analytics ProfessionalHow To Shift Your Career From IT To Cybersecurity Let us take a look at a few of these IT certifications a professional or fresher should consider excelling in the IT space. Google Cloud Platform (GCP) Certified Professional Cloud Architect The Professional Cloud Architect certification from Google helps professionals to understand cloud architecture very thoroughly. With the help of this certification, they can design, develop, and manage secure, scalable, and dynamic solutions to provide business solutions. Google says that the certificate assesses the ability to design and plan a cloud solution architecture and manage the cloud solution infrastructure. It also gives professionals the capability to design for security and compliance, analyse, and optimise technical and business processes. The exam length is two hours and has a registration fee of $200 (plus taxes, where applicable). To learn more, click here. AWS Certified Solutions Architect – Associate This certification from Amazon will make the professional capable of designing and implementing distributed systems on AWS. Amazon says that this certification is for professionals with one or more years of hands-on experience designing fault-tolerant and scalable distributed systems on AWS. It recommends that before planning to take this examination, one must have previous experience with using the AWS Management Console and the AWS Command Line Interface and have familiarity with AWS Well-Architected Framework, AWS networking, and the AWS global infrastructure. The professional should also have the capability to identify which AWS services meet a given technical requirement and the technical requirements for an AWS-based application. The duration of the exam is 130 minutes and costs $130. To learn more, click here. Certified Information Systems Auditor (CISA) If one is an entry-level to mid-career professional in the IT sector, Certified Information Systems Auditor (CISA) can show one’s capabilities in applying a risk-based approach to planning, executing and reporting on audit engagements. It will help gain credibility in interactions with internal stakeholders, regulators, external auditors, and customers. “ISACA’s certifications are curated as per the demands of the market and emerging technology trends while also catering to specific functions within the overall IT space. Most of these certifications have been around for years and are highly respected in the IT community and with employers in demonstrating expertise and adding credibility with internal and external stakeholders. Certain trends have brought increased attention to certain credentials,” said RV Raghu, Director, Versatilist Consulting India Pvt Ltd, member of ISACA Emerging Trends Working Group and past ISACA board director. He added, “Since ISACA launched its Information Technology Certified Associate (ITCA) and Certified in Emerging Technology (CET) certifications earlier this year, they have also witnessed a good uptick among IT professionals seeking to switch into an IT career path as well as to upskill in emerging technologies.” For more details, click here. Salesforce Certified Development Lifecycle and Deployment Designer The Salesforce Certified Development Lifecycle and Deployment Designer tag is for professionals who have skills in managing Lightning Platform development and deployment activities. They should also have experience analysing the environment and requirements to design an appropriate governance framework and manage the development and deployment lifecycle on the Lightning Platform. A professional with this certification will have capabilities in designing and implementing complex development and deployment strategies and communicating the proposed solution and design trade-offs to business and technical stakeholders alike. The typical job roles will include technical lead, developer lead, project manager, release manager, environment manager, technical architect, developers, and testers. The examination has 60 multiple-choice\/multiple-select questions, and the time allotted to complete the exam is 105 minutes. The registration fee is $400 (plus applicable taxes as required per local law). For more details, click here. Certified Information Systems Security Professional (CISSP) Certified Information Systems Security Professional (CISSP) will give one the capability to design, implement and manage a best-in-class cybersecurity program effectively. It is ideal for professionals such as Chief Information Security Officer, Director of Security, IT Director\/Manager, Security Systems Engineer, Security Analyst, Security Auditor, Network Architect, etc. For more details, click here. Microsoft Certified Azure Solutions Architect Expert The Azure Solutions Architect Expert will make professionals have the skills to design and implement solutions that run on Azure aspects, such as network, storage, and security. One can get this certification by passing Exam AZ-303: Microsoft Azure Architect Technologies and Exam AZ-304: Microsoft Azure Architect Design. Microsoft says that an Azure solutions architect has the skills to design end-to-end solutions for Azure while considering infrastructure, apps, data, and security, among others. It is suitable for IT pros, developers, and data professionals. For more details, click here.","excerpt":"Let’s take a look at a few IT certifications a professional or fresher should consider excelling in the IT space","categories":["AI Trends"],"tags":["Google","Microsoft Azure","Salesforce","skills","upskilling"],"author_name":"Sreejani Bhattacharyya","publish_date":"2021-12-15T12:00:00","publication_year":"2021","word_count":923,"keywords":["Microsoft Azure","data science","Go","GCP","upskilling","AI","AWS","R","Scala","RAG","Salesforce","skills","Google","analytics","Azure"],"extracted_tech_keywords":["AI","data science","analytics","RAG","AWS","Azure","GCP","R","Go","Scala"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-it-certifications-to-pursue-in-2022-for-a-software-job\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":56884,"title":"Ten Expert Tips And Advice For Aspiring Data Scientists","content":"Thinking about pursuing a career in data science but not sure where to start? With endless archives of information available online and few that may be coming from verified sources, sifting through this information could be challenging. As a result, we have compiled a valuable collection of tips and thoughts from a variety of experts from the data science community. Here are individual insights and pieces of advice from data scientists themselves that could help you carve out a profitable career in this field:- Improve Your Writing Skills While expertise in technical skills is a must for all data scientists, what is often neglected is writing skills. According to Parul Pandey, a data science evangelist at H2O.ai, the job of a data scientist is not only to extract insights from data, but also to communicate their findings in such a way that is comprehensible. Thus, the ability to write well and articulate your ideas in a coherent manner is paramount in having a successful career in data science. Choose Any Language – But Be Proficient In It When it comes to designing and testing algorithms, the choice of language is not as crucial as the fluency attained in the same. Star TV Network’s Senior Data Scientist Pushkar Paranjpe strongly recommends choosing one that reduces the iteration time, that is, one where you take the least amount of time to test an idea. Although efficiency should be prioritised, other aspects to factor in when choosing language can include its readability, availability of libraries, and the level of expertise of the same within your team. Paranjpe’s choice of language is Python, and he firmly advocates for JupyterLab and Pycharm IDE for rapid prototyping of algorithms and writing production-grade code. He also lays emphasis on cultivating curiosity and asking questions – a skill he feels can be developed over time. Paranjpe also recommends books by Russell and Norvig, and Ian Goodfellow to brush up on the basics, as well as online platforms like Stack Exchange, subreddits and Youtube. Podcasts could be another useful medium to keep yourself abreast with the latest developments in ML and AI. Build Your Domain Knowledge This is particularly directed towards college freshers. Pravat Ranjan Jena, Senior Data Analyst and BI at Dell, believes that before jumping in to acquire the critical aspects of ML and AI, it is imperative that you first understand the prevailing trends in the industry. Jena also openly endorses the large repository of online courses that are available for free. He feels that those alone should be enough to get a good grasp of the concepts of data science. In addition to this, he also implores all aspiring data scientists to not be overwhelmed by the wealth of high-end tools floating in this vast universe. According to him, they should first strengthen their foundations and sharpen their basics like SQL, Tableau, and MS Excel. Engage With The Data Science Community One of the key ways to learn is by engaging with people from a diverse set of people from around the world and hence, it can be a good practice to check in during important conferences. This also builds your awareness and helps you stay connected to all stakeholders in the industry, feels co-founder and CTO of Recreate.ai, Imaad Mohamed Khan. Moreover, such events also build a culture of collaborations and foster openness to learn new things. Master Technologies & Programming Skills Proficiency in Linux, SQL concepts, cloud, Java and Python, and building skills around DevOps, ML and big data is a prerequisite if you want to do well in the analytics space, according to Sameer Dixit, head of Big Data and Analytics at Persistent Systems. He also feels that aspirants who have a strong hands-on experience in projects will have the upper hand in this field. Also, without focusing on the development, pay attention to deployment as well. Ignore Myths Around Data Science Data science has become the holy grail across industries, and that kind of hallowed projection has invited plenty of myths around it as well. Some of these include PhDs seen as one of the basic requirements and the emphasis on participating in data science competitions for deeper awareness about how the industry operates. Navigating through this clutter to find your own path is the only way to respond to these myths, as per Director of Data and Analytics at Royal Bank of Scotland, India, Anish Agarwal. According to Agarwal, these myths give a false impression that data science is the domain of geniuses alone. Instead, the focus should be on trying to understand its practical use cases and the requirements each different role demands. One way to do this is by identifying a problem you are passionate about solving and examining if data science techniques can be applied there. Never Stop Learning & Exploring What binds all data scientists together are essentially the urge to learn from a vast pool of data, and this mindset should regularly be exercised, according to CEO and Chief AI Engineer at iNeuron.ai, Sudhanshu Kumar. Data gives you a window into what has happened in the past and how that can shape the future. However, this landscape is constantly shifting, and this necessitates you to be aware of all the developments that are happening around you. Identify The Right Problems To Solve Data scientists often fail to identify the right problems and there this has been demonstrated by a study conducted by Pactera Technologies. Muthumari S, sub-business unit head at Brillio, quotes this report as she points out that despite the spike in AI adoption, 85% of AI projects ultimately fail to deliver on their intended promises to business. According to Muthumari, this is because data scientists do not spend enough time assessing the impact of the ML models before developing them. Not only does it lead to wastage of resources, but it also burns out developers. One way to counter this is by rejecting the practice of implementing fancy techniques. Don’t Abandon Human Intuition Data may be a source of intelligence, but it is just one among many. As per Senior VP and Head of Engineering at HomeLane Puneet Gupta, while we should turn to insights generated by data, human intuition will still be a crucial element. As the world gets increasingly more complex, making decisions will, in turn, become harder. A good approach to take in this scenario will be to blend meticulous analysis with your intuition. Weigh The Pros And Cons Directed at people who are looking to switch or start a career in data science, Jaanaki Sankar, Partner at TheMathCompany, thinks that they will need to start from the bottom and hence, should prepare themselves for a challenging journey ahead. A good way to start will be to immerse themselves in the subject matter and weigh the pros and cons of making this transition.","excerpt":"Thinking about pursuing a career in data science but not sure where to start? With endless archives of information available online and few that may be coming from verified sources, sifting through this information could be challenging. As a result, we have compiled a valuable collection of tips and thoughts from a variety of experts […]","categories":["AI Highlights"],"tags":["big data trends and challenges","Data Science","Data Scientists","java project ideas","Programming Languages","Pycharm"],"author_name":"Anu Thomas","publish_date":"2020-02-18T18:30:30","publication_year":"2020","word_count":1143,"keywords":["data science","Go","AI","R","ML","java project ideas","Programming Languages","Python","Pycharm","analytics","SQL","Data Science","Jupyter","big data trends and challenges","Java","Data Scientists"],"extracted_tech_keywords":["AI","ML","data science","analytics","Jupyter","Python","R","SQL","Go","Java"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/ten-expert-tips-and-advice-for-aspiring-data-scientists\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10062209,"title":"Meta’s machine translation journey","content":"There are around 7000 languages spoken globally, but most translation models focus on English and other popular languages. This excludes a major part of the world from the benefit of having access to content, technologies and other advantages of being online. Tech giants are trying to bridge this gap. Just days back, Meta announced that it plans to bring out a Universal Speech Translator to translate speech from one language to another in real-time. This announcement is not surprising to anyone who follows the company closely. Meta has been devoted to bringing innovations in machine translations for quite some time now. Let us take a quick look back into the major highlights of its machine translation journey. 2018 Scaling NMT translation Meta used neural machine translation (NMT) to automatically translate text in posts and comments. NMT models are useful at learning from large-scale monolingual data, and Meta was able to train an NMT model in 32 minutes. This was a drastic reduction in training time from 24 hours. Open-sourcing LASER In 2018, Meta also open-sourced the Language-Agnostic SEntence Representations (LASER) toolkit. It works with over 90 languages that are written in 28 different alphabets. LASER computes multilingual sentence embeddings for zero-shot cross-lingual transfer. It works on low-resource languages as well. Meta said that “LASER achieves these results by embedding all languages jointly in a single shared space (rather than having a separate model for each).” 2019 Wav2vec: unsupervised pre-training for speech recognition Today, accessing various benefits of technology like GPS, virtual assistants essentially need speech recognition technology. But most of them rely on English, and a major chunk of people who do not speak the language or speak it with an accent not recognisable are excluded from using such an easy and important method of accessing information and services. Wav2vec wanted to solve this. Here, unsupervised pre-training for speech recognition was the focus point for Meta. Wav2vec is trained on unlabeled audio data. Meta adds, “The wav2vec model is trained by predicting speech units for masked parts of speech audio. It learns basic units that are 25ms long to enable learning of high-level contextualised representations.” Due to this, Meta has been able to build speech recognition systems that perform way better than best semi-supervised methods, though it can have 100 times less labelled training data. 2020 M2M-100: Multilingual machine translation 2020 was an important year for Meta, where it came out with different models that advanced machine translation technology. M2M-100 was one of them. It is a multilingual machine translation (MMT) model that translates between any pair of 100 languages without depending on English as an intermediary. M2M-100 is trained on a total of 2,200 language directions. This model wants to make the quality of translations worldwide better, especially those who speak low-resource languages claimed Meta. CoVoST: multilingual speech-to-text translation CoVoST is a multilingual speech-to-text translation corpus from 11 languages into English. What makes it unique is that CoVoST covers over 11,000 speakers and over 60 accents. Meta claims that it is “the first end-to-end many-to-one multilingual model for spoken language translation.” 2021 FLORES 101: low-resource languages Following M2M-100’s footsteps, in the first half of last year, Meta open-sourced FLORES-101. It is a many-to-many evaluation data set that covers 101 languages globally with a focus on low-resource languages that lack extensive datasets even now. Meta added, “FLORES-101 is the missing piece, the tool that enables researchers to rapidly test and improve upon multilingual translation models like M2M-100.” 2022 Dat2vec In 2022, Meta released data2vec, calling it “the first high-performance self-supervised algorithm that works for multiple modalities.” It was applied to speech, text and images separately and it outperformed the previous best single-purpose algorithms for computer vision and speech. Data2vec does not rely on contrastive learning or reconstructing the input example.","excerpt":"Meta has been devoted to bringing innovations in machine translations for quite some time now.","categories":["Global Tech"],"tags":["AI (Artificial Intelligence)","innovation","language","Language Models","Machine Learning","Meta","Speech Recognition"],"author_name":"Sreejani Bhattacharyya","publish_date":"2022-03-07T16:00:00","publication_year":"2022","word_count":630,"keywords":["Go","API","Meta","AI","Modal","innovation","language","Machine Learning","virtual assistants","computer vision","Aim","Language Models","Speech Recognition","R","contrastive learning","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","computer vision","Aim","virtual assistants","R","Go","API","contrastive learning","innovation","Modal"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/metas-machine-translation-journey\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10070201,"title":"Anaconda wants to chew into the market of Google Colab","content":"Python is one of the most popular programming languages (especially in the data science community) and consistently leads different indexes and surveys. But Python can be used by non-programmers too, where its applications can spread beyond data science-related problems. Peter Wang, CEO and co-founder of Anaconda, observes that for Python to maintain the growth it has seen as well as remain the most widely used data science programming language in the world, it will be important to improve its accessibility and remove the barriers to collaboration. In order to achieve this, the latest move from Anaconda is that it has acquired  PythonAnywhere, a cloud-based Python development and hosting environment. Anaconda, Inc. is behind the release of Anaconda, a distribution of the Python and R programming languages for data science, machine learning, predictive analytics, etc. PythonAnywhere: Host, run and code Python in cloud PythonAnywhere allows Python developers to just create web applications within a cloud-based Python environment. This can provide great ease in collaborating and sharing within dispersed teams in today’s work scenarios. PythonAnywhere was founded in 2012 by Giles Thomas and Robert Smithson and is based out of the United Kingdom. It is an online integrated development environment (IDE) and web-hosting service based on the Python language. PythonAnywhere comes with features like syntax-highlighting, error-checking editor, Python 2 and 3 consoles, and a full set of batteries (view them here). The issues that come with getting Python installed on each member’s laptop can be avoided here while everyone can pip install all the right packages. This ensures better collaboration.One can just write the application, and there is no need to configure or maintain a web server as everything is set up and ready to go.If the user has a browser and an Internet connection, that is enough, informs PythonAnywhere. “PythonAnywhere runs on our servers and displays in your web browser; you can write Python applications from your iPad, phone, or smart TV just as easily as you can from your computer”, says the firm. It is in beta for Android devices. The basic accounts are free, and the company charges for advanced services such as professional web app hosting, etc. But wait, doesn’t Google Colab offer something similar? Anaconda claims that with this move, it wants to “democratise” data science and Python. But, already, such offerings from tech giants exist that solve similar concerns. One of the prominent names that immediately pops up is Google Colab. Essentially, Colab connects your notebook to a cloud-based runtime where the user can execute Python code without any required setup on one’s own machine. The Colab notebooks can be shared like a Google Doc. Colab supports many popular machine learning libraries that can be easily loaded into your notebook. Jake VanderPlas, software engineer at Google, explains, “Google Colab is an executable document that lets you write, run and share code within Google Drive. If you are familiar with the popular Jupyter project, you can think of Colab as a Jupyter notebook stored in Google Drive.” It comes with various uses. One can create, upload and share notebooks or import notebooks from\/to Google Drive. The user can also import external datasets as well as integrate PyTorch, TensorFlow, Keras, OpenCV. PythonAnyhere provides hosting support too Both Google Colab and PythonAnywhere are platform-as-a-service (PaaS) models to provide a browser-based environment to run codes in the cloud. In both places, we can write, execute and share the codes. Colab, however, does not support the hosting of Python applications. To do this, we would need to use different APIs. PythonAnywhere provides execution and hosting support in one place. This announcement by Anaconda comes a month after Pyscript was released. It allows users to create Python applications in the browser by using a mix of Python with standard HTML. PythonAnywhere provides Python developers greater flexibility around where they do their job – desktop vs cloud. It seems like the two announcements back to back in a short span of time are quite strategic from Anaconda’s point of view. The two work in sync to further Anaconda’s objective to move Python beyond just data science and increase its use among non-programmers too.","excerpt":"PythonAnywhere is a cloud-based Python development and hosting environment.","categories":["Global Tech"],"tags":[],"author_name":"Sreejani Bhattacharyya","publish_date":"2022-07-01T12:00:00","publication_year":"2022","word_count":690,"keywords":["data science","machine learning","Keras","AI","PyTorch","ML","Aim","Jupyter","analytics","TensorFlow"],"extracted_tech_keywords":["AI","machine learning","ML","data science","analytics","Aim","TensorFlow","PyTorch","Keras","Jupyter"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/anaconda-wants-to-chew-into-the-market-of-google-colab\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10110093,"title":"AI Predictions for 2024 from 6 Industry Experts","content":"Across every industry, AI dominated discussions and developments in 2023, and the results were compelling. However, as the calendar turns to 2024, the tech is punctuated with challenges that need attention beyond foundational research. AI is filled with uncertainties as it breaks traditional patterns. There are also concerns regarding how well employees and enterprises adopt the technology. With the upcoming elections in the US and India, the political faction is also threatened by AI deepfake technology. The challenges continue to persist, but so does a narrative of possibilities. By drawing insights from industry insiders, one can understand the trajectory of AI in the upcoming year. Here are 6 predictions explaining what AI in 2024 will mean for organisations, investments and researchers. Sara Hooker, Head of Cohere for AI The head of one of the hottest AI startups in Silicon Valley, Sara Hooker, has predicted that the AI community needs to step up its game in 2024. She points out a big issue: Most AI tech is made with a Western viewpoint, mostly in English-speaking countries. This leaves out many languages and cultures, especially from places like Asia, where much of the world’s population lives. She also told Turing Post that 2024 will shake things up in research. Instead of just focusing on one thing like language or visuals, we’ll see a lot more tech that combines different aspects, like speech and images, all working together in multi-modal AI models. Werner Vogels, Chief Technology Officer, Amazon At Amazon Web Service’s annual gathering, re:Invent 2023, Werner Vogels, the big tech’s chief, announced his famous yearly predictions. He said that in the coming year, the community will train AI models with a mix of different cultures to understand people better and tackle big world problems. These AI models will give better answers on various topics, making them more useful. This change isn’t just for one place or group; it will make a difference everywhere, for all kinds of people, now and in the future. In an exclusive interview with AIM, he shared thoughts on how diverse countries like India can build their inclusive AI models. Clem Delangue, CEO and Co-founder, Hugging Face The CEO of Hugging Face, the hosting platform for open-source AI models, recently shared some insights on LinkedIn about what we might see in AI – particularly language models – for the following year. Starting with a bold statement, he noted that quite a few of the hyped AI companies that have been grabbing headlines and investor attention might be in for a rough patch. He suggested that it was not just some minor setbacks; some of these companies might face severe financial troubles or might even be bought out for much less than they were once valued. On a brighter note, open source will have some promising developments. He’s putting out there that these open-source language models are on track to reach a level where they can compete head-to-head with the best and most expensive closed-source models. Bhasker Gupta, CEO, Analytics India Magazine AIM founder, Bhasker Gupta, recently took to his annual tradition of giving his followers a heads-up on what AI might look like in 2024. He pointed out that there’s a tangible risk of AI being manipulated, playing an important and not necessarily a positive role in the upcoming democratic elections. He also mentioned that the ripple effects would be far-reaching and raise another round of concerns and debates about “Responsible AI”. Sridhar Muppidi, CTO, IBM Security As AI keeps pushing the boundaries, the discussion around security predictions is getting real. The early birds are already using AI to strengthen the front end, giving security analysts a productivity boost. But AI and cybersecurity might be on the cusp of something big. The CTO of IBM Security put it straight and said, “I don’t think we’re far from seeing generative AI deliver a transformative impact on the back end to completely reimagine threat detection and response into threat prediction and protection.” He believes the tech’s ripe and ready, so the cybersecurity industry is gearing up for a game-changing moment, “achieving prediction at scale”. Erin Price-Wright, Partner, Index Ventures As AI continues to boom, it demands a lot of computing power and storage space. Looking ahead to 2024, Erin Price-Wright, a partner at Index Ventures, a leading VC firm, thinks big changes are coming to data centres where all this data is stored. She says we’ll see better ways to connect computers, find new energy sources, and cool down those overheating machines. Speaking to Fortune, she’s got her eyes on the future of data centres.","excerpt":"The challenges continue to persist, but so does a narrative of possibilities.","categories":["AI Trends"],"tags":[],"author_name":"Tasmia Ansari","publish_date":"2024-01-06T10:10:18","publication_year":"2024","word_count":765,"keywords":["Go","Hugging Face","AI","RAG","Aim","ViT","generative AI","analytics","GAN","R"],"extracted_tech_keywords":["AI","analytics","generative AI","Aim","Hugging Face","RAG","R","Go","GAN","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/6-ai-predictions-for-2024-from-industry-experts\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10111089,"title":"Andrew Ng Partners with Google Cloud for New Course on LLMOps","content":"Andrew Ng’s DeepLearning.AI has come up with a free short course, LLMOps, offered in collaboration with Google Cloud. The course is designed in a way that beginners find it easy to learn in one hour of instruction by Erwin Huizenga, a Machine Learning Technical Lead at Google. The target audience includes individuals interested in tuning LLMs and building LLMOps pipelines. Participants will gain practical knowledge on adapting an open-source pipeline to implement supervised fine-tuning on an LLM for better user question responses. The course emphasises best practices, such as data and model versioning, and covers the pre-processing of substantial datasets within a data warehouse. Responsible AI practices are also addressed, focusing on the output of safety scores for sub-categories of potentially harmful content. They will also delve into the LLMOps pipeline, learning to retrieve and transform training data, version data and tuned models, configure an open-source supervised tuning pipeline, and deploy a tuned LLM. The course comes with practical applications which include creating a customised question-answer chatbot, like one for Python coding queries. BigQuery data warehouse, open-source Kubeflow Pipelines, and Google Cloud are some of the tools taught in this course. Ng had earlier partnered with Google Cloud for the course ‘Understanding and Applying Text Embeddings with Vertex AI’ which focussed on leveraging text embeddings to capture the essence of sentences and paragraphs. Read more: How SAP Joule is Quenching Enterprises’ GenAI Thirst","excerpt":"Participants will have practical knowledge on adapting an open-source pipeline to implement supervised fine-tuning on a LLM for better user question responses.","categories":["AI News"],"tags":["Courses","LLMs"],"author_name":"Shritama Saha","publish_date":"2024-01-19T11:14:22","publication_year":"2024","word_count":233,"keywords":["Go","GenAI","machine learning","TPU","AI","LLMs","RAG","Python","Kubeflow","Courses","R","data warehouse"],"extracted_tech_keywords":["AI","machine learning","GenAI","Kubeflow","RAG","TPU","Python","R","Go","data warehouse"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/andrew-ng-partners-with-google-cloud-for-new-course-on-llmops\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10093344,"title":"Prompt Engineers Are Getting Paid More Than Python Developers","content":"The hot new job in the market, prompt engineering, was believed to be the “job of the future”, but is a reality already. The best part, or probably the scariest for a lot of programmers, is that it pays a bomb. The concern is that one does not require the knowledge or understanding of a single programming language to get this job. Back in 2017, a report by the Institute of the Future stated that 85% of the jobs that would exist in 2030, haven’t even been invented yet. Same is the case with prompt engineering. And this job can pay up to $335,000, without even requiring a computer science degree. “Digital Upskilling” – new cartoon and post https:\/\/t.co\/W3zOKbQDB0“The Prompt Engineer is the new Growth Hacker.”#marketing #cartoon #marketoon #ai pic.twitter.com\/4mzW2CozDm— Tom Fishburne (@tomfishburne) May 7, 2023 While the magnitude of prompt engineering’s future impact remains uncertain, various sectors and businesses are already seeking talent in this field. Anthropic, a Google-supported AI startup, is enticing prospective candidates with dazzling salaries of up to $335,000 for the intriguing role of “Prompt Engineer and Librarian“. The listings emphasise the need for individuals with a daring hacker mindset and a passion for unravelling enigmas. Klarity, an automated document reviewer, is also willing to pay up to $230,000 for a skilled machine learning engineer capable of masterfully coaxing AI tools to deliver optimal outcomes. It seems the quest for the perfect prompt is on, and the rewards are nothing short of remarkable. Sounds tempting, doesn’t it? Getting a six-digit salary without even paying for an expensive college degree. Meanwhile, companies such as IBM are freezing hiring in trying to replace people with AI. Moreover, with layoffs getting rampant, it looks like with the constant innovation in AI, all companies need now is someone who can perform operations using AI models. No Money Left for Programmers? In recent news, Microsoft CEO Satya Nadella announced freezing the pay raise for its employees citing poor macroeconomic conditions even when the company reported a 9% profit this quarter. There is no doubt that the company is pumping in funds into AI. Speaking to AIM, a Microsoft India employee said that a lot of them are planning to change jobs because of this reason. Even then, there are very few opportunities for programmers in the job market. The demand for prompt engineers, however, is at an all-time high with exorbitantly high salaries — more than what a lot of programmers at big-techs draw. https:\/\/twitter.com\/Saboo_Shubham_\/status\/1650437953871544322 This has unleashed in a flood of people putting up “prompt engineer” in their LinkedIn bios. The “ChatGPT Experts” or the ‘snake oil sellers of AI’ have now become prompt engineers. But is the demand for these jobs actually on the rise? The truth is, if you search for prompt engineering jobs on LinkedIn or Indeed, you’ll find hundreds of job postings, but as soon as you go into the requirements or eligibility for most of them, they require some knowledge of programming languages or the workings of LLMs. In a recent Reddit discussion, a user shared a screenshot of a job opening hiring for a Fullstack Developer with a salary of $150,000-200,000 per month. Though it cannot be verified if this job offer is legitimate, it is true that there are a lot of jobs still up for grabs for expert programmers, and they pay a bomb as well. In comparison, we can see that a prompt engineer’s salary posted on job threads is almost two times higher than that of a full-stack developer. According to a recent survey by Indeed, with more than 2,000 jobs for prompt engineers, the average salary for a prompt engineer is $150,000, whereas for a full-stack Python developer is around $110,000. “You want me to build a python framework? Sure, I like snakes,” said a Reddit user. English is the New Programming Language, really? In January, Andrej Karpathy posted on Twitter, “The hottest new programming language is English”. While it is true that the introduction of models like ChatGPT, Bard, or even Codex and Replit, has made it easier for non-developers to write code they had no idea about, the need for programmers is still there. There is no doubt that it is beneficial for everyone to team up with AI and upskill themselves. But relying on AI for everything and branding yourself a “killer prompt engineer” may backfire. Though we have text-to-anything models these days that behave like co-pilots for everything, the trend to be a prompt engineer is going to die soon. Rob Lennon, an expert at prompt engineering has been teaching paid courses for the same and says that people in this profession just have the mover’s advantage. “In six months, 50,000 people will be able to do this job. The value of this knowledge is greater today than it will be tomorrow,” explained Lennon. The higher salaries that are being offered won’t last for a very long time. Looks like the claim that an entire generation is studying for jobs that won’t exist is true for a lot of roles that can be replaced by AI. But at the same time, computer science degrees are going to remain beneficial forever. Staying strong amid the layoffs is what these engineers need. On the other hand, people without degrees can compete in the era of AI with prompt engineering courses, and have the role of an “LLM psychologist” and stay relevant.","excerpt":"Prompt engineering jobs can pay up to $335,000, without even requiring a computer science degree","categories":["Deep Tech"],"tags":["AI Jobs","AI layoffs","Developers","prompt engineering","prompt engineering jobs"],"author_name":"Mohit Pandey","publish_date":"2023-05-15T16:00:00","publication_year":"2023","word_count":904,"keywords":["Anthropic","ChatGPT","Go","prompt engineering jobs","machine learning","AI layoffs","AI","AI Jobs","RAG","Python","prompt engineering","Aim","R","Developers"],"extracted_tech_keywords":["AI","machine learning","ChatGPT","Anthropic","Aim","RAG","prompt engineering","Python","R","Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/prompt-engineers-are-getting-paid-more-than-python-developers\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10088030,"title":"M1ssion Impossible: How Linux Researchers Reverse Engineered Apple’s Chips","content":"The latest update to the Linux kernel, termed Linux 6.2, officially announced support for the M1 Pro, Max, and Ultra chips. This was made possible due to a team working largely on donations from Patreon and GitHub, who took on one of the most difficult tasks in software: reverse engineering an Apple M1 chip. While competitors like Intel and AMD provide comprehensive documentation on how their chips are designed, Apple hasn’t done so for its custom-made chips. This was mainly because M1 chips weren’t sold standalone, and were always bundled with MacBooks running MacOS. This meant that Apple could keep all their cards close to its chest, and no one would ask. Linus Torvalds, the creator of Linux spearheaded the efforts to bring Linux to M1. In late 2022, reports said that he was using an M2 MacBook as a daily driver. In the past too, Torvalds has expressed his feelings regarding the new MacBooks during its launch in 2020, stating, “I’ve been waiting for an ARM laptop that can run Linux for a long time. The new Air would be almost perfect, except for the OS. And I don’t have the time to tinker with it, or the inclination to fight companies that don’t want to help.” It seems that the sentiment changed somewhere down the line, as Linux version 5.19 was reported to have been released from Linus’ new M2 MacBook. In the release notes for this version, Linus revealed that he was running Linux on his ARM-powered M2 Mac, all thanks to a small team of developers from a community project called Asahi. Breathing new life into old hardware Apart from the development efforts on the driver and reverse engineering of the chip, the Asahi Linux project also has a team of dedicated developers working on the discrete components of Apple’s chips. There is still a long list of features that don’t work on Asahi Linux, notably the Neural Engine, Touch Bar, and Thunderbolt, as these products require reverse engineering of their own to crack. However, this is one of the best efforts by the Linux community to bring the OS to yet another line of products. Users on the Apple subreddit reacted extremely positively to news of the Linux support for Macs. Many were surprised at the amount of work it took to reverse-engineer the chip, while others marvelled at how low the barrier to entry had become for a capable Linux machine. Reddit user GamingChairModel stated, “The Mac Mini is a killer deal on the actual CPU, but not a great deal on storage\/memory upgrades. With the memory soldered on, it makes sense to upgrade to 16GB RAM and just add on external storage through the Thunderbolt port (or a separately managed NAS)…So for any application that doesn’t require a ton of memory, the Mac Mini’s hardware could be a great ARM server.” Making Apple devices into home servers is just one of the various applications that Linux support opens up. When Apple inevitably stops supporting the M1 Macs sometime in the future, those with the technical know-how can give the still-capable devices a second life with a Linux install. What’s more, the GPU drivers also allow users to run games on an M-series GPU — something considered impossible unless the game was developed specifically for it. Breaking down the walled garden While porting Linux to ARM devices has been a work in progress for the past 20 years, this part of the Linux community has come a long way since then. However, the release of Apple’s M1 chips presented a new challenge to Linux kernel maintainers. Even though the chips were built on ARM’s IP, which has been well-documented in Linux, the architecture that Apple built on top of it was largely unknown. This left the researchers at Asahi in a difficult spot. To allow the Linux kernel to support M1 devices, they first needed to know how the OS interacted with the discrete components of the M1 processor. One of the most important parts of the development process was the creation of a hypervisor termed m1n1, which allowed the researchers to root around in software close to M1’s ‘bare-metal’, allowing them to map out the processor. The M1 chip consists of a CPU with performance and efficiency cores, an integrated Neural Engine, and the bane of Asahi’s existence — the GPU. While it was relatively easier to build code for the M1 CPU using m1n1 and a lot of work over 6 months, the GPU proved to stump the researchers’ efforts, until they brought in Alyssa Rosenzweig. While Dougall Johnson laid the groundwork for reverse engineering the M1 GPU by exploring its capabilities and creating some documentation, Rosenzweig was tasked with getting a custom-built driver up and running on M1’s GPU. In a four-part epic detailed on her website, Rosenzweig took on the behemoth task of writing a driver for the GPU. In January 2021, she decoded enough of the instruction set to disassemble some simple shaders — a program that calculates levels of light, darkness, and colour when rendering 3D images. Most modern GPUs are made up of 3 components — shader cores, rendering units, a command processor, and a memory management unit. Shader cores process 3D models and pixel data, and rendering units take this processed information and convert it into pixels. Command processors convey commands from the software to the shader cores, while memory management units make sure that any given bit of memory is not accessed by multiple programs at once. To decode this instruction set, Alyssa first had to jump into the command stream within the GPU. Once these commands were decoded, she had to decode the chip’s shader binaries. These are the low-level bits of code that determine how the GPU’s shader cores function. She could also send commands to the chip through her toolkit — a process which allowed her to work on the GPU without needing to assemble a whole new driver for the Linux kernel. In her first month of working on the project, she was able to render a single triangle on the M1 GPU core. By rendering a triangle, the researcher was able to find specific information about how the GPU processes commands and gives an output. By April 2021, she was able to write a shader compiler and render a 3D spinning cube. Around the same time, Johnson was making forays into investigating the instruction set and building compute kernels. In May, she was able to begin development on a graphics driver built on Gallium with support for OpenGL and ES 2.0 specifications. Naturally, the process involved a lot of programming nitty-gritty, which has been omitted from this explanation for brevity’s sake. However, the end result was that Alyssa got a driver to work on the M1 chip, but encountered a mystery bug she termed “the impossible bug”. While trying to implement the driver (termed the Mesa driver) on the hardware, Alyssa found that the driver cannot render large amounts of geometry. This was because the M1 chip used a different rendering method than traditional methods, something Alyssa had overlooked. This was fixed with a bit of software magic, putting the bow on top of a completely working driver, which was then integrated into the Linux kernel with help from Asahi Lina. Asahi Lina is a vtuber who was responsible for most of the community attention the project received, as she published hours-long videos of the process of creating a driver for M1 Mac. She was also responsible for writing the kernel driver, a piece of software that provided an interface between Linux and the GPU driver. Notably, this driver was the first Linux GPU driver written completely in Rust, chosen for its fast execution and support for large numbers of data structures. After bringing together the kernel driver and the Mesa driver developed by Alyssa, the GPU driver for Mac was finally complete. What’s more? The programmers managed to preserve the M1 Mac’s prodigious battery life, with tests revealing that their driver could run a 3D game at 1080p for over eight hours. Bringing Linux to Mac, as with any other Linux-based update, was a team effort by a concerted group of people including, but not limited to, Alyssa Rosenzwig, Ella Stanforth, Dougall Johnson, Asahi Lina, Sven Peter, Mark Kittens, and Hector Martin. For a select group of people, these names will be immortalized as those who broke down a big portion of Apple’s walled garden.","excerpt":"Two years after the release of Apple’s M1 Macs, a group of dedicated computer scientists have finally brought Linux to Apple’s crown jewels","categories":["AI Features"],"tags":[],"author_name":"Anirudh VK","publish_date":"2023-02-24T14:00:00","publication_year":"2023","word_count":1416,"keywords":["Go","TPU","AI","Git","RAG","Ray","ViT","Rust","GitHub","R"],"extracted_tech_keywords":["AI","Ray","RAG","TPU","R","Go","Rust","Git","GitHub","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/m1ssion-impossible-how-linux-researchers-reverse-engineered-apples-chips\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10068561,"title":"Council Post: A deep dive into the world of AI-Analytics-powered-claims management","content":"My last article was about how insurers can leverage AI & analytics in underwriting. This article focuses on the next step of the insurance value chain, which is claims. We all have experience in claims, but we hardly know how it is processed at the backend and how AI is bringing more efficiency and effectiveness. Claims is a formal request made to your insurance provider for reimbursement against losses covered under your insurance policy. Digitising insurance claims Today more than half of claims activities have been replaced by automation, according to McKinsey. For example, insurance providers have been using advanced algorithms to handle initial claims routing, increasing efficiency and accuracy. IoT sensors and an array of data-capturing technologies have helped insurers replace manual methods of first notice of loss (FNOL), where claims triage and repair services are often triggered automatically upon loss. For example, in the case of an auto accident, the policyholder takes a streaming video of the damage, which is later translated into loss descriptions, and estimated amounts. At home, IoT devices are used to proactively monitor water levels, temperature, and other risk factors and will proactively alert both insurers and tenants of issues before they arise. Meanwhile, automated customer service apps handle most policyholder interactions through text and voice directly following self-learning scripts that interface with the fraud, claims, medical service, policy, and repair systems due to faster resolution times. According to Accenture, nearly 74 percent of customers said they would interact with modern technology and appreciate the computer-generated system of insurance advice. Human claims management currently focuses on a few areas. This includes complex and unusual claims and contested claims where data-driven insights and analytics empower human interactions. In this article, we will discuss how insurance claims processing is getting transformed by the adoption of AI, NLP and analytics. As AI and NLP enable digitisation, the power of analytics boosts the effectiveness of extracted content. This is done by leveraging predictive modeling. First Notice of Loss (FNOL) In claims management, the first process is FNOL, where the insured informs their insurer about the loss and lodges a claim with the insurer for the damages incurred. Lodging a claim can go through multiple channels. This includes voice\/on-call, non-voice-through portal, mobile application or sending an email. In most non-voice cases, the lodgement process is keying in 60-80 fields (insured details, vehicle details, loss details, customer details, etc.) in the claims management system (Guidewire, LexisNexis, etc.). The information is read from emails, standard and non-standard claim forms, and other supporting documents that are key to the claims management system. These forms can be digital, handwritten, scanned or unscanned, and can be of various formats, including .eml, .msg, pdf, Docx., .rtf, TIFF, JPEG, etc. Semi-structured documents Industry specific AccordIndustry specificInsurer claim forms Insured specific forms Created by larger brokers Unstructured documents Supporting documents Police reports, incident reports, legal forms Here’s how it’s done The key steps involve automating the 60-80 features with AI. The nature of the fields is not standard; there would be free-text driven fields (claims description, type, etc.), which require ML and NLP models to classify and summarise to decide the loss type. Or, there would be documents such as police reports and survey reports which vary based on city and state and are mostly handwritten. To yield significant impact, one needs an advanced pre-built ensemble of computer vision and NLP models that can extract, classify, and summarise details. Further, the extracted information enables the building of ML models to identify duplicate and fraud claims. For example, the claims can come from the same customer again and again, or the same agent is sending across similar claims. All such cases are identified and sorted through ML models. Lastly, the AI analytics’ intervention would be to direct the claim to the correct channel for processing and prioritisation. Alignment can be either auto-processing or handler driven depending upon the claim complexity, value and ageing. The models support in directing the case to the right-skilled handler. Once the automated data pipeline has been established, you can leverage the rich, newly aggregated data to get powerful insights, leading to better underwriting decisions, products and customer experience. The outcome Benefits of digitising claims set-up processing: Efficiency gains 30-50% due to manual work reductionBetter customer experience with faster claims set-up Identification of duplicate claimsFaster decisioning with automated claims adjudicationRevenue prevention by proactive fraud detectionFurther insights: Claims volume prediction, loss analysisReduce chances of litigation by faster and more accurate processing Setting up claims and proceeding with the same can be a long, effort-intensive process which can lead to revenue leakage. Applying AI and analytics in claims management improves cost optimisation and customer experience and curbs revenue leakage. Companies should adopt AI solutions which are domain-rich, scalable and flexible in delivering multiple use cases. This article is written by a member of the AIM Leaders Council. AIM Leaders Council is an invitation-only forum of senior executives in the Data Science and Analytics industry. To check if you are eligible for a membership, please fill out the form here.","excerpt":"Setting up claims and proceeding with the same can be a long, effort-intensive process which can lead to revenue leakage. Applying AI and analytics in claims management improves cost optimisation and customer experience and curbs revenue leakage.","categories":["AI Features"],"tags":["AI (Artificial Intelligence)"],"author_name":"Saurabh Khanna","publish_date":"2022-06-08T10:03:03","publication_year":"2022","word_count":843,"keywords":["data science","AI","ML","computer vision","RAG","NLP","Ray","Aim","analytics","AI (Artificial Intelligence)","fraud detection"],"extracted_tech_keywords":["AI","ML","NLP","computer vision","data science","analytics","Aim","Ray","RAG","fraud detection"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/a-deep-dive-into-the-world-of-ai-analytics-powered-claims-management\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":46437,"title":"Google &#038; Indian Central Water Commission Collaborate To Forecast Floods","content":"It is estimated that every year more than 200 million people either get displaced or lose their lives due to floods. Flood forecasting is a tricky job since a lot of factors such as rainfall, the topology of the location, soil strength, fluid dynamics come into play. Current forecasting techniques do help in timely evacuations. Usually, the warnings from the meteorological department lead to the deployment of disaster relief personnel in hazardous locations. However, this too isn’t enough. In India, the state of Kerala alone has experienced two devastating floods in two consecutive years. The aftermath of any flood usually ends up with people either blaming global warming or the lawmakers. Though there is no one-stop solution, the best alternative is to resort to robust precautionary measures. In order to ring the alarm bells early enough, Central water commission of India has joined hands with Google to forecast floods. Google has been implementing its machine learning prowess for various social good across the world. Flood simulation of a river in Hyderabad, India via Google In India especially, Google has been doing tremendous work by leveraging all their AI capabilities at disposal. Last year, Google rolled out its early flood warning services, starting with the Patna region. How Does Google Do It? via Google AI blog Google’s approach involves incorporating multidisciplinary techniques. They range from gathering data regarding the topography of a location to using equations of fluid dynamics. A reliable forecasting model can be established using the following data: Real-time Water Level Measurements Google partnered with Indian Central Water Commission (CWC), which measures water levels hourly in over a thousand stream gauges across all of India, aggregates this data, and produces forecasts based on upstream measurements. The CWC provides these real-time river measurements and forecasts, which are then used as inputs for our models. Elevation Map Creation It is critical that the models have a good map of the terrain and this requires high-resolution digital elevation models (DEMs) that are difficult to acquire. This is where Google Maps comes into the picture. Map creation is achieved as follows: The large and varied collection of satellite images used in Google Maps are used for correlating and aligning the images in a large batch. Then the corrected camera models are used to create a depth map for each image. Optimally fuse the depth maps are fused together at each location to make the elevation map. Train convolutional neural networks to identify where the terrain elevations need to be interpolated. Hydraulic Modeling The location and velocity of water through time are considered in building hydraulic models. The results tested using a 2D form using Saint-Venant equations. These are a bunch of partial differential equations. Solving these equations requires a lot of computational budget. So, Google optimised the hydraulic model for Tensor Processing Units (TPUs). The parallelised nature of TPUs offer 85x times faster than a traditional CPU. The researchers are also exploring advanced machine learning models to run data discretisation techniques to solve partial differential equations, which has the potential to bring down the time required to forecast. Forecasting A Safer Future Modelling nature is a feature engineering nightmare. There a number of factors that come into the picture. Hydrologic models, for example, consider precipitation, solar radiation, soil moisture as inputs to produce a forecast. So, challenges like scaling and accuracy still persist. Nevertheless, we can safely say that Google is doing it the right way by collaborating with government entities like Central water commission and other non-profit organisations to gather data at a granular level. Since a machine learning model is only as good as the data, localised models are necessary.","excerpt":"It is estimated that every year more than 200 million people either get displaced or lose their lives due to floods. Flood forecasting is a tricky job since a lot of factors such as rainfall, the topology of the location, soil strength, fluid dynamics come into play. Current forecasting techniques do help in timely evacuations. […]","categories":["Global Tech"],"tags":["Google"],"author_name":"Ram Sagar","publish_date":"2019-09-26T18:30:58","publication_year":"2019","word_count":609,"keywords":["Go","machine learning","TPU","AI","neural network","feature engineering","Git","RAG","Google","GAN","R"],"extracted_tech_keywords":["AI","machine learning","neural network","RAG","TPU","R","Go","Git","feature engineering","GAN"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/google-indian-central-water-commission-collaborate-to-forecast-floods\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10172211,"title":"The Air India Tragedy That Could Have Been Prevented by AI","content":"Since its introduction 14 years ago, the Boeing 787 Dreamliner has faced several safety issues. Flight data from the Aviation Safety Network shows that six such occurrences were linked to technical difficulties in 2025 alone. One of the most devastating mishaps was the recent crash of Air India Flight 171, which claimed at least 270 lives, including 230 passengers and 12 crew members, after the aircraft crashed into a hostel block at BJ Medical College in Ahmedabad. However, solutions cannot be found only after accidents; they must be identified and tested before failures occur. This tragedy, which left only one survivor, has sparked widespread fear and raised urgent questions about aircraft safety. It underscores a critical truth: solutions must be identified and tested before failures occur, not after lives are lost. While human error has often been cited as a leading cause for aviation accidents, it raises an important question: Could AI play a significant role in preventing such tragedies through various strategies? Revolutionising Aviation Operations As commercial aviation experiences growing demand and stricter safety regulations, Boeing has incorporated robotics, AI-based analytics, digital twin technology, and machine vision systems to improve manufacturing efficiency and product quality. A research paper titled ‘Artificial Intelligence in Aviation Safety: Systematic Review and Biometric Analysis’ reveals that AI enhances safety protocols through predictive analytics, real-time monitoring, and proactive risk management. Machine learning (ML) algorithms analyse large datasets to identify patterns, enabling the prediction of safety issues before they arise. “Machine learning – anomaly detection is a mature and well-understood technology with a proven positive impact. When applied to the aviation industry, it can significantly improve safety, quality of service, and reliability, reduce maintenance costs, and increase the efficiency of aircraft operators,” Alexis Lope-Bello, CEO at ComTrade Group, reportedly highlighted the importance of these technologies. Moreover, AI also supports the development of decision-making systems for pilots and air traffic controllers, improving the timeliness and quality of their decisions. By implementing AI-driven predictive maintenance, airlines can identify and resolve mechanical problems early by assessing aircraft performance and maintenance records, thus reducing the chances of in-flight failures. According to the research paper, data science and big data analytics are increasingly critical in optimising aviation safety and operations. “The data-driven methodology for detecting safety risks through aviation data analysis illustrates the growing use of big data analytics in aviation. Applying Natural Language Processing (NLP) techniques to detect human factors in aircraft accidents represents a crucial step toward minimising human errors,” the paper read. Proactive Measures Through Data Science and AI AI algorithms can analyse extensive data from aircraft sensors to predict mechanical failures before they occur. This early detection allows airlines to perform proactive maintenance, reducing the risk of in-flight emergencies. By examining flight data, AI identifies patterns that may indicate safety risks or operational inefficiencies. The paper shows that continuous monitoring of flight metrics helps airlines spot deviations and take corrective actions to prevent accidents. “AI is already playing a crucial role in reducing aviation accidents and loss of life by improving situational awareness, predictive maintenance and decision-making processes. In air traffic control (ATC) and collision avoidance, AI is helping out tremendously,” Freshta Farzam, CEO and founder of LYTE Aviation, told Al Jazeera. “When it comes to aviation, there’s a lot of advancement in AI, even though a lot of that has not trickled through to the commercial aviation because of the processes and certification,” Amad Malik, chief AI officer at Airport AI Exchange, told Al Jazeera. Various companies are incorporating AI models such as the ones mentioned earlier. In collaboration with Microsoft, Lufthansa Technik has employed AI algorithms to examine aircraft sensor data and forecast possible mechanical problems before they arise. This strategy enables them to arrange maintenance activities more effectively, minimise aircraft downtime, and avoid expensive interruptions to airline operations. “The way the regulations are set up is that years and years of data are required before you can use anything in the commercial landscape. What we do have right now is something that started in the 1960s. But there’s also a concept of having AI as a local intelligence within the aircraft that can detect and mitigate even if the pilot or ATC are making a mistake,” Malik added. Forbes recommends that Boeing adopt Predictive Maintenance Systems that utilise AI to analyse aircraft sensor data and identify safety risks early. For example, Lufthansa Technik’s Condition Analytics uses machine learning to reduce unexpected maintenance and aircraft downtime. How AI is Enhancing Seamless Airline Experience Air India has also actively leveraged AI into its operations, but it is restricted to customer experience and streamlining various processes, such as quicker ticket bookings, personalised customer service, and baggage tracking. Satya Ramaswamy, chief digital and technology officer of Air India, wrote on LinkedIn, “With Copilot, our leaders gain quick, clear insights without information overload, enabling timely, effective decisions that drive the airline forward. Being in a business where there are human copilots, we understand the spirit in which Microsoft Copilot, as a technology, is to be used in assisting and amplifying human performance! Take the example of our On-time Performance (OTP) Copilot. Notably, the air-taxi industry is expected to expand rapidly by the close of the decade, though experts caution that its viability relies on a thoughtfully designed network of vertiports that must be incorporated into current infrastructure and backed by sophisticated air traffic management systems. “We will see more vertiports being developed or planned in the short term, particularly in congested areas where helicopter and traditional aerial transportation access is limited,” Nicolas Zart, founder of Electric Air Mobility, reportedly said. “Initially, these vertiports may be located on the outskirts of major cities as they test the concept of AAM before moving into urban centres.”","excerpt":"While many have blamed human error for the fatal accident, deploying AI analytics systems could assist airlines in managing risk.","categories":["AI Features"],"tags":["Air India","artifical intelligence"],"author_name":"Smruthi Nadig","publish_date":"2025-06-23T17:50:00","publication_year":"2025","word_count":955,"keywords":["data science","machine learning","artificial intelligence","AI","ML","Air India","RAG","NLP","Aim","analytics","copilots","artifical intelligence"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","NLP","data science","analytics","Aim","RAG","copilots"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/the-air-india-tragedy-that-could-have-been-prevented-by-ai\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10063243,"title":"This third-year engineering student created computer vision software to prevent harassment through surveillance cameras","content":"What if one were to tell you that Big Brother can be a ‘good’ brother? Notwithstanding the privacy concerns and debates around surveillance cameras, metropolitan cities today are covered with them. Yet, most of the time, there is no one governing movement through the camera to catch crimes occurring in real-time. This stands true of India with its high sexual abuse cases, many of which happen in public places in broad daylight. Souvik Ghosh, a third-year engineering student, has created a model, ‘Sathi’, an automated defender of harassment in public places. The prototype shows a woman showing a fist or the index finger to any surveillance camera on the street and the camera alerting bypassers with a loud siren and noticing the police, all on its own. In an interview with Analytics India Magazine, Souvik spoke about Sathi and his life as a student data scientist. https:\/\/www.linkedin.com\/embed\/feed\/update\/urn:li:ugcPost:6909446808335257600 AIM: What was the idea behind the Sathi model? My interests lie in data science, computer vision and deep learning. Along with being a tech enthusiast, I am also a social worker. I work in the Kolkata red light area and teach the students there. This has allowed me to see the abuse that happens first handed. Later, I found that one in every three women has experienced sexual harassment in India. My female friends also spoke about such experiences, and I realised harassment usually happens in places like busy roads, crowded buses, local trains, educational institutions and more. But many women feel hesitant to bring it to notice or shout in public places. They fail to react. Searching for solutions for this made me realise that metropolitan cities like Mumbai, Delhi or Kolkata have surveillance cameras around the city – even transport, streets and colleges. Additionally, most people carry smartphones on them always. But no one is really checking the camera footage to identify such crimes. I realised we needed the camera to be automated, and it led me to create Sathi. AIM: Can you run us through Sathi? Let’s suppose anyone runs into a situation where they feel threatened. They need to show their fist in a sign of protest towards the camera, be it the CCTV camera or even their smartphone camera. The device will start playing a loud alarm or siren sound upon registering the fist. This would alert the citizens around. Alternatively, if the person shows their index finger or points to number one, the system will directly inform the police. It will also capture and send the images of that spot with a time graph to the police. The system will capture every face in that scene as passport-sized faces as evidence to use later. For the app, the person can show a fist to the mobile phone camera, and the phone will directly call the police. AIM: Tell us about your tech stack. Additionally, as a student data scientist, how did you access the needed resources? Thanks to the internet, we have great free courses on YouTube and Coursera. The education resources are easy. When it comes to the system, we dont need access to perfect systems. We have Kaggle and Google Collab that provide huge RAMs. Even if you stay in the remotest parts of the country with the cheapest laptops, you can contribute to system creation. My technological stack was mainly deep learning, computer vision and data science. My base language was Python. I used TensorFlow and MediaPipe for Face detection for computer vision, to ensure the system could detect faces even in low light areas. AIM: What is the scope of the model? The system is still a prototype as of now. The plan is to embed it in a CCTV camera and a mobile camera. There are a few answers I am still working on. Firstly, the distance between the person and the camera is not always going to be ten to fifteen meters. Sometimes it will be fifty meters. So, how well will the camera detect faces from a long-distance? Secondly, how well will it perform for multiple faces? For instance, how will it identify the face in question from hundreds of people? Lastly, what if people use it for fun? If I were a child, I would enjoy showing a fist to the camera and watching the crowd create chaos. I plan on connecting with industry experts and professors to brainstorm and make it foolproof. AIM: How do you plan to take the model further? My two approaches as of now are using MEMs and embedded computers. We can put the whole system in embedded computers and CCTV cameras. We can also put the code in mobile apps that work in rural areas without cameras. The app runs in the background and will identify the fist in the camera. A challenge here is to connect the camera to light posts in the area since a siren in a mobile phone is pointless. Upon seeing a fist, the mobile app will notify the light post that will play a siren. We still have to brainstorm where the final version will work; a CCTV camera, an app or both? We can have urban and rural coverage through both. AIM: What are your tips for students entering data science? Anyone entering data science should start with Python and its libraries. There are steps one needs to climb up; people shouldn’t jump directly into deep learning. Data science is the root of all learning, followed by machine learning and deep learning. This is the path I followed. I have also interned with different organisations in positions like a research intern. Internships also compel me to study and learn more. Finally, applying for free courses on platforms like YouTube will help study better.","excerpt":"The device will start playing a loud alarm or siren sound upon registering the fist. This would alert the citizens around. Alternatively, if the person shows their index finger or points to number one, the system will directly inform the police.","categories":["AI Features"],"tags":["Interviews and Discussions"],"author_name":"Avi Gopani","publish_date":"2022-03-22T15:00:00","publication_year":"2022","word_count":958,"keywords":["data science","machine learning","GCP","AI","computer vision","RAG","Aim","deep learning","analytics","TensorFlow","Interviews and Discussions"],"extracted_tech_keywords":["AI","machine learning","deep learning","computer vision","data science","analytics","Aim","TensorFlow","RAG","GCP"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/this-third-year-engineering-student-created-computer-vision-software-to-prevent-harassment-through-surveillance-cameras\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10047879,"title":"Common Feature Engineering Techniques To Tackle Real-World Data","content":"Data mining is a technique of extracting useful patterns and relationships from data, most commonly databases, texts, and the web. Data Mining uses statistical and pattern matching techniques to help discover insights. The most common concerns with data are it being noisy, full of missing values, its relevance to the problem, its size and complexity of data. When dealing with real-world data, it is often vast, which also comes with the challenge of it being imprecise at certain times and the data structure getting complex. There are various approaches to tackle and handle such real-world data out of which a few are popularly used by data engineers. In this article, we will be discussing the data handling or feature engineering techniques that are most commonly used. The major points that we will cover in this article are listed below. Table of Contents BinningTransformingScalingShufflingHow To Implement In Python?Conclusion Binning Real-world data tends to be highly noisy with a large amount of meaningless and unwanted information, termed as noise. Binning is a technique that is used for reducing the cardinality of continuous and discrete data. Binning related values together in multiple bins is used to reduce the number of distinct values. Data binning, also known as bucketing, groups of data in bins or buckets, replaces values contained in a small interval with a representative value for that interval. Binning method tends to improve the accuracy in models, especially predictive models. It provides a new categorical variable feature from the data reducing the noise or non-linearity in the dataset. The binning technique also promotes easy identification of outliers, invalid and missing values from the numerical variables present in the data. Making use of bins here is often referred to as binning or creating k bins, where k relates to the number of groups to which the numeric variable is mapped from the dataset. This technique can be applied to each of the numeric input variables in the training dataset, which are then provided as an input to the machine learning model to learn about predictive modelling tasks. Supervised binning is another method of intelligent binning where the important characteristics from the data are used to determine the bin boundaries. Binning method is also used for the sheer purpose of data smoothening. Here the data is first sorted and then the sorted values get distributed into several buckets or bins. As binning methods consult the neighboring values, this is also known as local smoothing. The most common approach for binning is to divide the range of variables into k equal-width intervals. In equal frequency binning, we divide the range of the variables into intervals that contain approximately an equal number of points, although the equal frequency may not be possible due to repeated values. Image Source Transforming Data transformation or transforming in the context of data mining is done by combining unstructured data with structured data for better analysis. It is also an important process when the data is being transferred to a new cloud data warehouse. When the present data is homogeneous and well-structured, it is easier to analyze and look for patterns and this is made possible using the transformation technique. Sometimes the data present in databases might possess unique IDs, keys and values. All these need to be formatted well so that the records are similar and can be correctly evaluated. Data transformation can also be further defined as the method of converting data from one particular format to another or changing the present structure into another structure. Data transformation is a critical step before commencing other activities such as data integration and data management. Data transformation can include a wide range of activities such as converting data types, cleansing data by removing nulls or duplicate values, enhancing the data, or performing specific aggregations, depending on the needs. In a cloud data warehouse, it is arranged homogeneously to make it easier to recognize patterns. The data can be converted in multiple ways that are ideal for mining the data. The data transformation also involves Smoothing and Aggregation techniques. Data collection or aggregation is the method of storing and presenting the data in a summary format. This is a crucial step as the accuracy of data analysis and insights generated is highly dependent on the quantity and quality of the data being used. Gathering accurate data of superior quality and in a humongous quantity is very essential to produce relevant results. Smoothing, on the other hand, is the process of eliminating noise from the data using algorithms that help highlight the important features present within the data. It also helps in predicting the present patterns correctly. Scaling Feature scaling is a method that is used to normalize the range of independent variables or features present in the data. Manier times we observe that the range of data values vary widely. In a few machine learning algorithms, objective functions do not work properly without normalization. Many classifiers too, calculate the distance between two points by using the Euclidean distance. If one of the features contains a broad range of values, the distance will be governed by this aspect. The range of all the features present in the data should be normalized so that every feature contributes approximately and proportionately to the final distance. If feature scaling is not performed, then the machine learning model gives a higher weightage to the higher values and lower weightage to lower values. Also, it might take a lot of time for training the machine learning model. Hence scaling methods such as Standardization is a very effective technique that re-scales a feature value so that it has a distribution with zero mean value and variance equals 1. Min-Max Normalization is another technique that rescales a feature or observation value with a distribution value between zero and one. The goal of applying Scaling is to make sure that the present data features are on almost the same scale so that each feature becomes equally important and hence it becomes easier to be processed by most ML algorithms. Shuffling The shuffling techniques aim to mix up data and help retain logical relationships between the data columns. It randomly shuffles data from a dataset within an attribute or a set of attributes. Using this method sensitive information can be shuffled to be replaced with other values for the same attribute from a different record. It is used for masking confidential numerical data as the values of the confidential variables are shuffled among the observed data points. The shuffled data may provide a high level of data utility and minimize further risks. Data shuffling helps overcome reservations about using modified confidential data because it retains all the desirable properties and performs better than the other masking techniques in both data utility and disclosure risk. Data shuffling can be implemented using only rank order data, and thus it provides a nonparametric method for masking. The applicability of data shuffling stands the same for small and large data sets. During machine learning, we are required to split the dataset into further training, testing & validation datasets. It is very important that the dataset is shuffled well to avoid any element of bias or patterns in the split datasets before training begins for the ML model. Shuffling improves the model quality and the predictive performance of the model that it is being applied to. How To Implement In Python? Here I am going to demonstrate an example of the Binning technique that can be implemented through Python during the data mining process. Assume that we have a large amount of data and we cannot pass strings to a machine learning model. Therefore we might just need to convert the categorical features present in the dataset such as Sex, Embarked, and others into numeric values. Here the data binning technique can be very useful. data['Sex'].replace(['male','female'],[0,1],inplace=True) data['Embarked'].replace(['S','C','Q'],[0,1,2],inplace=True) data['Initial'].replace(['Mr','Mrs','Miss','Master','Other'],[0,1,2,3,4],inplace=True) Converting this using the binning method, data['Age_cat']=0 data.loc[data['Age']<=16,'Age_cat']=0 data.loc[(data['Age']>16)&(data['Age']<=32),'Age_cat']=1 data.loc[(data['Age']>32)&(data['Age']<=48),'Age_cat']=2 data.loc[(data['Age']>48)&(data['Age']<=64),'Age_cat']=3 data.loc[data['Age']>64,'Age_cat']=4 Similarly, other columns can be converted into categorical features by using the Binning method, data['Fare_cat']=0 data.loc[data['Fare']<=7.775,'Fare_cat']=0 data.loc[(data['Fare']>7.775)&(data['Fare']<=8.662),'Fare_cat']=1 data.loc[(data['Fare']>8.662)&(data['Fare']<=14.454),'Fare_cat']=2 data.loc[(data['Fare']>14.454)&(data['Fare']<=26.0),'Fare_cat']=3 data.loc[(data['Fare']>26.0)&(data['Fare']<=52.369),'Fare_cat']=4 data.loc[data['Fare']>52.369,'Fare_cat']=5 Printing the Binned Table, Conclusion The use of several data techniques to handle data is an important aspect to be focused on as real-world data comes along with several challenges to be handled. If the discussed techniques and methods are correctly used and properly implemented, they might provide a lot of aid in generating a correct model with high accuracy. End Notes In this article, we discussed several methods that help tackle real-world data such as Binning, Transforming, Scaling and Shuffling. These methods help in making the process of data mining a lot easier and help to generate better insights from the mined data. We also saw an example of the data Binning technique and where it can be used. I would request the reader to try other methods as well for a greater understanding. Happy Learning! References Preprocessing Methods and Pipelines of Data Mining Choosing the binning for a histogram","excerpt":"Data mining is a technique of extracting useful patterns and relationships from data, most commonly databases, texts, and the web. Data Mining uses statistical and pattern matching techniques to help discover insights.","categories":["AI Trends"],"tags":["AI (Artificial Intelligence)","Data Science","Data Scientist","Deep Learning","feature engineering","feature engineering in machine learning","how to measure twitter influence","load data python","Machine Learning","Python"],"author_name":"Victor Dey","publish_date":"2021-09-07T15:00:00","publication_year":"2021","word_count":1495,"keywords":["Go","feature engineering in machine learning","load data python","machine learning","AI","how to measure twitter influence","ML","feature engineering","Machine Learning","Python","Aim","ViT","Deep Learning","Data Science","Data Scientist","R","AI (Artificial Intelligence)","data warehouse"],"extracted_tech_keywords":["AI","machine learning","ML","Aim","Python","R","Go","data warehouse","feature engineering","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/common-feature-engineering-techniques-to-tackle-real-world-data\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":50409,"title":"[Jobs Roundup] 7 Latest Managerial Data Science Jobs In Bengaluru","content":"The role of data science manager is important for any company in order to maintain the smooth function of its data science team. This role not only requires advanced technical knowledge of the domain but also requires strong interpersonal and project management skills. In this article, we are going to look at top managerial jobs in data science in Bengaluru: Senior Manager — Data Science @ Publicis Sapient Digital business transformation company, Publicis Sapient is looking to hire a senior manager for data science to join its team in Bengaluru. As a senior manager, you will not only be responsible for management but also responsible for designing overall search solution including search architecture, data flow and integration with different possible source systems. Requirements: The candidate should have a graduate or postgraduate degreeShould have 12 -16 years of overall experience in the software industryAt least 4 years of experience in Search Technologies like Elastic Search, SolrCloud, Endeca, GSA, HPE IDOL, Fusion etc.Must have solid experience and deep understanding (end to end) of search applicationsShould have experience with strategising testing and performance tuning of the search applicationExperience in text processing using NLP tools and relevancy tuning using machine learning models is an added advantage. Apply Here. Senior Manager — Data Science @ Oracle Oracle Corporation, Bengaluru, is looking to onboard Senior Manager for its data science department. The candidate will be responsible for driving the top-level data science strategy for Oracle Applications Labs (OAL), which is responsible for implementing, running, and improving nearly all of Oracle’s Enterprise On-Premise and Cloud Applications. It would also include hire, lead and coach a team of data scientists. Requirements: The applicant must have an advanced degree in Computer Science, Statistics, Engineering, Mathematics, or another relevant quantitative fieldShould have at least 10 years of experience in data science domain, 7 years of hands-on experience working in machine learning and 5 years of managerial experience in managing 5 to 10 directsShould be proficient and experience in:Python, R, Java, C++, Scala, and Sparkquery languages such as SQL and its adaptationsbig data processing engines like Spark, Kafka is a plusdata stores such as Hadoop Apply Here. Senior Manager — Data Science @ Walmart Labs Walmart Labs in Bengaluru is looking to hire a data science manager to lead the Advertising Technology Data Science team. The candidate will be responsible for several work processes which would also include guiding the data science team in gathering and analysing data. Click here to know more about the responsibilities. Requirements: The candidate should have 10 to 16 years of experience in the domainShould have strong knowledge of data-gathering, pre-processing data, model building, pushing the model to production, publication and presentation.Should be experienced in recommendation and ranking preferably in large data setsShould be able to guide the team with formulating the problem, reviewing modelling techniques and processes and model buildingHave the ability to explain complex ideas to cross-functional stakeholders and manage stakeholders across geographies Apply Here. Sr Analytics Manager @ Amazon Amazon in Bengaluru is looking to onboard an experienced, technically skilled and customer-centric Senior Analytics Manager for its Mass and Brand Marketing team. As a senior manager, you would not only be responsible for managing a team but also analyse customer lifecycle and draw actionable insights to effectively serve and engage customers. Requirements: The applicant should have a master’s degree in economics, statistics or a similar quantitative field, or a bachelor’s degree with significant experience.Should have at least 10 years of experience in the data science domainShould have strong knowledge of database and data warehousing concepts (Oracle, Redshift, SQL databases)Experience with BI tools such as Tableau and proficiency with SQL, R, SPSS & MS Excel is also imperative Apply Here. Senior Manager — Data Strategy @ Standard Chartered Bank Standard Chartered Bank is looking to hire a senior manager for data strategy to join its Data Strategy team (which is part of Enterprise Risk Analytics) in Bengaluru. The candidate will be responsible for providing leadership and coordination in data science and advanced analytics initiatives across RCF. Requirements: The candidate should have a bachelor’s or master’s degree in statistics, mathematics, computer science, etc. or in a relevant fieldS\/he should also have experience working with large datasets and expertise in data wranglingExperience in Machine Learning using supervised learning and unsupervised learning methodologies is also a must. Also, advanced machine learning methodologies is an added advantage.The candidate should also possess strong knowledge of programming — R, Python, SAS and\/or MATLAB Apply Here. Senior Manager — RWA Analytics @ Standard Chartered Bank Standard Chartered Bank is also looking to hire Senior Manager for RWA analytics who would be responsible for helping in building, sustaining and evolving the RWA Analytics team. Requirement: The candidate should have 10 to 12 years of relevant experienceShould be a Finance Professional or a Graduate in Engineering\/MBA (Finance)The candidate must possess strong knowledge and skills of root cause analysis, business analysis and data interpretation.Knowledge of Banks Capital Management with implementation experience relating to Basel II\/Basel III is an added advantageDomain Skills in Risk-Weighted Assets (RWA), Capital Management, Capital Analysis and Reporting, Basel II\/Basel III Reporting\/Regulatory Reporting, Credit Risk Analytics is also must Apply Here. Program Manager – Data Sciences @ Bosch Group Robert Bosch Engineering and Business Solutions Private Limited (RBEI) in Bengaluru is hiring program manager for data science not only to manage a team of highly talented data scientists and AI experts, but also strategise, implement, and maintain data science program initiatives that adhere to organisational objectives. Requirements: The candidate should have a bachelor’s degree in engineering or MCAThe candidate should also be PMP certifiedShould have more than 3 years of experience in using Microsoft Project and Microsoft Office, and in executing Agile projectsThe candidate should also have over 5 years of development background and should know some object-oriented programming languageKnowledge of statistical principles, machine learning, Internet of Things and AI is imperativeWell versed on PMBOK principles and used them effectively in multiple projects\/programs. Apply Here.","excerpt":"The role of data science manager is important for any company in order to maintain the smooth function of its data science team. This role not only requires advanced technical knowledge of the domain but also requires strong interpersonal and project management skills. In this article, we are going to look at top managerial jobs […]","categories":["AI Hirings"],"tags":["Data Science Jobs","hpe internet of things","science jobs","Walmart Labs","what is data oriented person"],"author_name":"Harshajit Sarmah","publish_date":"2019-11-21T11:00:00","publication_year":"2019","word_count":994,"keywords":["what is data oriented person","data science","Go","machine learning","science jobs","AI","Walmart Labs","Data Science Jobs","NLP","Python","analytics","Kafka","SQL","hpe internet of things","R"],"extracted_tech_keywords":["AI","machine learning","NLP","data science","analytics","Kafka","Python","R","SQL","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-hiring\/jobs-roundup-7-latest-managerial-data-science-jobs-in-bengaluru\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10049965,"title":"Meet The New Player In The US-China AI Arms Race: United Kingdom","content":"The United Kingdom government has released its 10-year plan to make the country the global “artificial intelligence superpower”. This comes right after BCS proposed ‘gold standards’ to make the UK a leading ethical AI superpower. The latest developments in the AI sector in the UK begs us to ask the question – is the UK competing in the AI arms race with China and the US? “Today we’re laying the foundations for the next ten years’ growth with a strategy to help us seize the potential of artificial intelligence and play a leading role in shaping the way the world governs it,” Chris Philp, a minister of the Department for Digital, Culture, Media and Sport, said in a statement. The Strategy The national AI strategy includes several programs, reports and initiatives to boost the country’s long term capabilities in and around ML technologies by prioritising and levelling up the development of AI applications in the UK. This is to strengthen its position as a global science superpower. This includes a new National AI Research and Innovation program to improve coordination and collaboration between the country’s researchers. The UK’s first AI strategy plans to launch a new national programme and a positive approach to support R&D in AI and publish a white paper on the governance and regulation of AI to ensure public confidence in the technology. This will allow organisations in various regions and sectors to capitalise on the power of AI technologies. “The UK already punches above its weight internationally, and we are ranked third in the world behind the USA and China in the list of top countries for AI,” said DCMS Minister Chris Philp. Desired AI The United Kingdom plans not only to be a top AI country but also to do that with ethical and regulated AI that the citizens desire. The New National AI, Research and Innovation Programme is set up to discover the latest developments and AI innovations. In addition, the country’s plans include a white paper on AI regulation to use modern technologies to improve people’s lives and solve global challenges such as climate change and public health. The government plans on encouraging and developing AI students by supporting postgraduate learning and retraining and ensuring that students from wide backgrounds can access specialist courses. The latest data shows £13.5 billion investment by global investors into more than 1,400 UK private technology firms between January and June in 2021. This is accompanied by more than £2.3 billion government investment into AI since 2014 to support AI working with clear rules, applying ethical principles, and a pro-innovation regulatory environment. The three pillars of AI growth in the UK are: Benefiting all sectors of the economyGoverning with rules to encourage innovation, investment Protecting the public and the country’s fundamental values. “The UK is already a world leader in certain aspects of AI – and this strategy helps to define how to enhance those capabilities further to ensure that the UK can both develop and use AI for the benefit of citizens,” said Government Chief Scientific Adviser Sir Patrick Vallance. The government also announced a review into the availability and capacity of computing power for researchers and organisations, along with a consultation on copyrights and patents for AI to assess the development. AI Arm’s Race: China and the US According to the World Intellectual Property Organisation, the US has filed more AI patent applications than any other country in the past two decades. Closely following, China had 41,000 over the same period. The UK filed less than 2,000. The US government’s Competition and Innovation Act involves the country investing billions of dollars in chips, AI, and supply chain reliability in building smart cities. In comparison, China has invested twice as much as the US in R&D (research and development). The Carnegie Endowment for International Peace declared China and the US as the two globally leading exporters of technology. Up until 2020, research based on patents and research publications ranked China as the top country for AI development. However, China’s recent developments and the government’s crackdown on big tech companies may harm China’s status in the AI arms race. Giving the UK a possibly good standing in the AI arms race, BCS’ Report – “Priorities for the National AI Strategy“, found the country capable of leading the world in creating AI that cares about humanity. “The public has become extremely distrustful of AI systems,” said Bill Mitchell, lead report author and director of policy at BCS. “You need to prove to them that you are competent, ethical and accountable.” The BCS report encouraged the UK to set the ‘gold standard‘ in AI professionalism through a pro-innovation, pro-ethical, pro-competition, and fair competition-based regulatory framework. Some of the largest AI companies like Graphcore, Darktrace, DeepMind, BenevolentAI and more are situated in the UK. The country is also brimming with leading universities, research centres and institutions that the government aims to tap into to become a leader in creating an empathetic AI. With a pinch of salt It’s also important to note that while the ideas of clear rules, ethical AI, and a pro-innovation regulatory environment for AI are still strategies and plans by the UK government. While the EU has a comprehensive proposal for regulating high-risk applications of AI on the table, the UK still lacks a formal regulatory framework. The public response to the announcement has been varied until now; while some are sceptical about the proceedings given the government’s history of lack of research incentive, others are quite optimistic about the plan. All in all, for now, we just have ideas and strategies and are still to see how they map out in reality.","excerpt":"The UK plans not only to be a top AI country but also to do that with ethical and regulated AI that the citizens desire.","categories":["AI Trends"],"tags":["china chip technology"],"author_name":"Avi Gopani","publish_date":"2021-09-29T17:00:00","publication_year":"2021","word_count":946,"keywords":["Go","API","artificial intelligence","AI","ML","Git","RAG","Aim","china chip technology","Rust","R"],"extracted_tech_keywords":["AI","artificial intelligence","ML","Aim","RAG","R","Go","Rust","Git","API"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/meet-the-new-player-in-the-us-china-ai-arms-race-united-kingdom\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":61526,"title":"Big Data Storage Mechanisms and Survey of MapReduce Paradigms","content":"Big data adoption continues to grow. Big Data covers data volumes from petabytes to exabytes and is essentially a distributed processing mechanism. MapReduce is a framework for data processing model. The greatest advantage of Hadoop is the easy scaling of data processing over multiple computing nodes. Many organizations use Hadoop for data storage across large pools of unstructured information called data lakes, and then load the most useful data into a relational warehouse for rapid and repetitive queries. In this article, we will understand and study the transformation of Data Warehouse to the storage of Big Data into Hadoop and various MapReduce implementations in various file systems. File Systems specialize in managing unstructured information. Databases specialize in managing structured information. Traditionally, a database knows the format of the internal contents of the table.  In a File System, data is directly stored in set of files. Traditional models of data processing are being disrupted by following factors: 1.         Increasing data volumes 2.         Increasing data complexity 3.         Changing analysis models 4.         Increasing analytical complexity over wide and broad spectrum of data 5.         Increasing capability and availability of cost-effective compute and storage Data Warehouse Data warehouses are essentially large database management systems that are optimized for read-only queries across structured data. They are relational databases and, as such, are very SQL-friendly. They provide fast performance and relatively easy administration.  In the 1990s, Bill Inmon defined a design known as a data warehouse. In 2005, Gartner clarified and updated those definitions. From these we summarize that a data warehouses have following features– •           Subject Oriented – The data is modelled after business concepts, organizing them into subjects’ areas like sales, finance and inventory. Each subject area contains detailed data. •           Integrated – the logical model is integrated and consistent. Data formats and values are standardized. •           Non-volatile – Data is stored in the data warehouse unmodified and retained for longer periods of time. •           Non-Virtual – The data warehouse is a physical, persistent repository. The biggest negative aspects of Data Warehouse are cost and flexibility. Most data warehouses are built upon proprietary hardware and are many orders of magnitude more expensive than other approaches. HDFS, the Hadoop Distributed File System, is a distributed file system designed so that it can hold a very large amount of data (terabytes or even petabytes) and provide high throughput access to this information. HDFS is a master-slave architecture that is a backbone of the computing frameworks such as MapReduce. HDFS layer forms the foundation for data-driven business decisions through Hadoop’s breakthrough features in scalability, flexibility, cost optimizations, ability to handle the greater volume, and veracity of data over traditional platforms. HDFS is a modern data storage engine that can adapt to ever-changing and extremely variable data formats.  HDFS effortlessly combines storage and compute workloads which allows combining and collapsing multiple workloads such as batch processing. MapReduce is a data processing model. Its greatest advantage is the easy scaling of data processing over multiple computing nodes. Under the MapReduce model, the data processing primitives are called as mappers and reducers. In the mapping phase, MapReduce takes the input data and feeds each data element to the mapper. In the reducing phase, the reducer processes all the outputs from the mapper and arrives at result. In simple terms, the mapper is used for filter and transform the input into something that the reducer can accumulate over. The difference between Data Warehouse and Hadoop be as below – ParametersData WarehouseHadoop Ecosystem  SystemWorks well with batched data ingestionWorks well with batch or near real time or real time data ingestionHardwareNeeds expensive hardware including specialized hardware in some casesUses commodity hardwareBuilt TypeBuilt primarily for performanceBuilt for extreme scalabilityData ProcessingStatic DataData on the fly  Data TypeSupports primarily modelled and structured dataSupports any type of dataStorage\/ CapacityDesigned for very high volumes (GBs – TBs)Designed for extremely high volumes (TBs and more)  ParadigmWrite many times read many times paradigm (ETL form)Write once read many times paradigm Various MapReduce Implementations 1.         Google MapReduce MapReduce is both, the name of the programming model, the original framework, and the scalable distributed data-intensive application designed and implemented by Jeff Dean and Sanjay Ghemawat at Google. It provides fault tolerance while running on inexpensive commodity hardware, delivers high aggregate performance to large number of clients.  Its code isn’t freely available, even though it is only used internally at Google and known to be written in C++, with interfaces in Python and Java. It is used in some of the largest MapReduce clusters to date. It has been studied in the literature that, on any given day, Google used to execute about 100,000 MapReduce jobs; each occupies about 400 servers and used to take about 5 to 10 minutes to finish. A GFS cluster consists of a single master and multiple chunk servers and is accessed by multiple clients. Each of these is typically a commodity Linux machine running a user-level server process. It is easy to run both a chunk server and a client on the same machine if machine resources permit, and the lower reliability caused by running possibly application code is acceptable. 2.         Disco Disco is an open source implementation of the MapReduce programming model, developed at Nokia Research Centre as a lightweight framework for rapid scripting of distributed data processing used for building a robust and fault-tolerant distributed application. As it is written in Python programming language, it lowers the entry barrier and makes it possible to write data processing code in few lines of code. It supports parallel computations over large data sets on unreliable clusters of computers. Unlike Hadoop, Disco is only a minimal MapReduce implementation, it has been used in parsing and reformatting data, data clustering, probabilistic modelling, data mining, full-text indexing, and log analysis with hundreds of gigabytes of real-world data. 3.         Skynet Skynet is an open source implementation of the MapReduce framework written in Ruby. But unlike other implementations, its administration is fully distributed and doesn’t have a single point of failure such as the master servers that can be found in Google MapReduce and Hadoop. It uses a peer recovery system and fault tolerant in which workers look out for each other. Peers track each other and, once failure is detected, can spawn a replica of the failed peer. 4.         Dryad It is Microsoft’s research project using MapReduce. Dryad intends to be a more general-purpose environment to execute data-parallel applications. It has been studied in the literature that it includes other computation frameworks, including MapReduce. It is intended to be a super-set of the core Map-Reduce framework. Dryad programs are expressed as directed acyclic graphs (DAG) in which vertices are computations and edges are communication channels. Dryad has been deployed at Microsoft since 2006, where it runs on various clusters of more than 3000 nodes. 5.         Gfarm It is a general-purpose distributed file system. It is a commodity-based distributed file system that can be spread across hundreds to thousands of compute nodes. Gfarm file system federates local file systems of compute nodes and does not require any special configuration or additional hardware. Using the Hadoop-Gfarm plugin, the Hadoop Distributed File System can be built on Gfarm and the MapReduce can be used. 6.         HadoopDB – A Hybrid MapReduce System The HadoopDB project is a hybrid system that tries to combine the scalability of MapReduce with the performance and efficiency advantages of parallel databases. In this type of project, HadoopDB is to connect multiple single nodes database systems using Hadoop as the task coordinator. Queries are expressed in SQL but their execution is parallelized across nodes using the MapReduce framework. In concluding remarks, the difference between Hadoop and data warehouse is just similar where an Architecture of Hadoop is a tool for handling and storing Big Data, whereas data warehouse is an architecture for organizing data to ensure integrity. One more point to state is, former makes the best use of relational and structured data whereas later excels in storing and managing unstructured data using MapReduce for data processing. References – •           Jeffrey Dean, Sanjay Ghemawat. MapReduce: Simplified Data Processing on Large Clusters. In OSDI’04: Proceedings of the 6th Symposium on Operating Systems Design & Implementation, pages 10–10, Berkeley, CA, USA, 2004. USENIX Association •           Pisoni, A. (2009) Skynet: A Ruby MapReduce Framework https:\/\/developpaper.com\/skynet-introduction-of-google-map-reduce-framework-in-ruby\/ •           Isard, M., Budiu, M., Yu, Y., Birrell, A., & Fetterly, D. (2007, March). Dryad: distributed data-parallel programs from sequential building blocks. In ACM SIGOPS operating systems review (Vol. 41, No. 3, pp. 59-72). ACM. http:\/\/research.microsoft.com\/en-us\/projects\/dryad\/eurosys07.pdf •           Mikami, S., Ohta, K., & Tatebe, O. (2011, September). Using the Gfarm File System as a POSIX compatible storage platform for Hadoop MapReduce applications. In Proceedings of the 2011 IEEE\/ACM 12th International Conference on Grid Computing (pp. 181-189). IEEE Computer Society. •           Ranger, C., Raghuraman, R., Penmetsa, A., Bradski, G., & Kozyrakis, C. (2007, February). Evaluating mapreduce for multi-core and multiprocessor systems. In High Performance Computer Architecture, 2007. HPCA 2007. IEEE 13th International Symposium on (pp. 13-24). IEEE 13-24. •           Yang, H. C., Dasdan, A., Hsiao, R. L., & Parker, D. S. (2007, June). Map-reduce-merge: simplified relational data processing on large clusters. In Proceedings of the 2007 ACM SIGMOD international conference on Management of data (pp. 1029-1040). ACM.","excerpt":"Big data adoption continues to grow. Big Data covers data volumes from petabytes to exabytes and is essentially a distributed processing mechanism. MapReduce is a framework for data processing model. The greatest advantage of Hadoop is the easy scaling of data processing over multiple computing nodes. Many organizations use Hadoop for data storage across large […]","categories":["Deep Tech"],"tags":["Big Data","data mapping","etl hadoop","hadoop and etl","hadoop etl","hpc data management system","managing hpc data","Mapreduce","what is big data"],"author_name":"Dr. Madhavi Vaidya","publish_date":"2020-04-13T15:00:00","publication_year":"2020","word_count":1536,"keywords":["TPU","Scala","C++","hadoop etl","R","Java","hadoop and etl","hpc data management system","Mapreduce","RAG","managing hpc data","Big Data","Go","AI","what is big data","Python","SQL","data mapping","etl hadoop"],"extracted_tech_keywords":["AI","RAG","TPU","Python","R","SQL","Go","Java","Scala","C++"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/big-data-storage-mechanisms-and-survey-of-mapreduce-paradigms\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10103604,"title":"Oxford AI Researchers Say LLMs Pose Risk to Scientific Truth","content":"In November tech giant Meta unveiled a large language model called Galactica, designed to assist scientists. But instead of landing with the big bang Meta hoped for, Galactica has died with a whimper after three days of intense criticism. A year later, the lack of dependence on these AI chatter boxes remains the same – particularly in scientific research. In a paper published recently in Nature Human Behaviour, AI researchers from the Oxford Internet Institute have raised concerns about the potential threat that LLMs pose to scientific integrity. Brent Mittelstadt, Chris Russell and Sandra Wachter argue that LLMs, such as those based on the GPT-3.5 architecture, are not infallible sources of truth and can produce what they term as ‘hallucinations’—untruthful responses. The authors propose a shift in how LLMs are utilized, recommending their use as ‘zero-shot translators.’ Instead of relying on LLMs as knowledge-bases, users should provide relevant information and instruct the model to transform it into the desired output. This approach facilitates easier verification of the output’s factual accuracy and consistency with the provided input. The core issue, as outlined in the paper, lies in the nature of the data these models are trained on. Language models designed to provide helpful and convincing responses, lack guarantees regarding their accuracy or alignment with facts. Training on vast datasets sourced from online content, which may include false statements, opinions, and creative writing, exposes LLMs to non-factual information. Prof. Mittelstadt highlighted the risk associated with users trusting LLMs as reliable sources of information, similar to human experts. Due to their design as human-sounding agents, users may be misled into accepting responses as accurate, even when devoid of factual basis or presenting a biased version of the truth. To safeguard scientific truth and education from the spread of inaccurate and biased information, the authors advocate for setting clear expectations around the responsible use of LLMs. The paper suggests that users, especially in tasks where accuracy is vital should provide translation prompts containing factual information. Prof. Wachter emphasised the role of responsible LLM usage in the scientific community and the need for confidence in factual information. The authors caution against the potential serious harm that could result if LLMs are haphazardly employed in generating and circulating scientific articles. Highlighting the need for careful consideration, Prof. Russell urges a step back from the opportunities presented by LLMs and prompts a reflection on whether the technology should be granted certain opportunities simply because it can provide them.","excerpt":"A year after Meta’s Galactica failed the lack of dependence on LLMs remains the same","categories":["AI News"],"tags":[],"author_name":"Tasmia Ansari","publish_date":"2023-11-23T21:58:43","publication_year":"2023","word_count":411,"keywords":["TPU","programming_languages:R","AI","GPT","ViT","Rust","AI research","R","programming_languages:Rust","llm_models:GPT"],"extracted_tech_keywords":["AI","TPU","R","Rust","GPT","ViT","AI research","llm_models:GPT","programming_languages:R","programming_languages:Rust"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/oxford-ai-researchers-say-llms-pose-risk-to-scientific-truth\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10060324,"title":"Meet the winners of the ‘Dare In Reality’ hackathon","content":"Genpact, in collaboration with Formula E team Envision Racing and MachineHack, successfully completed the Dare in Reality hackathon for data scientists and machine learning professionals on 22 November. The goal? To help the racing team improve its performance in the all-electric, international single-seater world championship. The hackathon welcomed more than 5,200 participants and over 10,000 submissions within just two weeks. “The idea for organising the Dare in Reality hackathon was to let data science professionals, machine learning engineers, AI practitioners, and other tech enthusiasts work on a real-world problem statement,” said Krishna Rastogi, the Product Lead and Technical Architect at MachineHack. “The hackathon has had one of the highest numbers of participants and submissions at MachineHack, where the rankings were based on the RMSLE metric to predict drivers’ lap times in qualifying rounds ahead of a race. Our participants have solved the problem in many innovative ways.” Let’s take a look at the winners who impressed the judges with their data skills and took home highly coveted cash prizes and goodies. Rank 01: Igor Maleev Maleev was crowned the winner of the Dare in Reality hackathon. Maleev became interested in data science while studying for a PhD in mathematics and statistics. He has experience working as a data scientist in the advertising and retail space and is a data science consultant right now. Winning Approach Fig 1: Data Distribution of Free Practices and Qualifying groups Maleev says that the main idea that helped him win was to train the model on the green segment of Fig 1. After training and testing this data, he got very close to getting the score on the leaderboard during the competition. The rest was very technical and consisted of feature engineering, cleaning data, and training and tuning the model. Check out the code here. Rank 02: Mahesh Yadav and Vakada Naveen Yadav became interested in machine learning when he saw intelligent virtual assistants but his full exposure came after getting into IIT Madras as a research scholar in September 2020. Naveen has always been fascinated by the portrayal of how AI could do wonders in futuristic movies like I, Robot. Even his final year B Tech project on machine learning was focused on this area. He has secured a place to pursue an MS through research at IIT Madras in the area of Vision and Language transformers. Winning Approach Yadav and Naveen followed a three-phase approach, which included: Data preprocessingModel buildingEnsembling methods Data Preprocessing The team preprocessed time columns by replacing them with float values, which were then separated into categorical and numerical columns. The team normalised the skewed features.  After applying one-hot encoding to handle the categorical columns and normalization on the dataset is done using MinMaxScaler. Principal Component Analysis (PCA) was performed on the dataset to reduce the dimensionality. Model building Yadav and Naveen tried a variety of models such as neural networks with different architectures, light gradient boosting, Xgboost, support vector regression, gradient boosting, and random forests. Neural networks performed best compared to other models. They tried hyperparameter tuning for the neural networks and found some of the best architecture to be used for ensembling. Ensembling Methods Yadav and Naveen ran various neural network architectures, sampled different training datasets each time and kept track of the best models. The ensemble approaches they tried include stacking with neural networks as meta learners, stacking with machine learning models as meta learners, simple averaging, and weighted averaging. They said that the best model they submitted was from simple averaging of predictions from the best neural network models. Check out the code here. Rank 03: Sylas John Rathinaraj Initially, Rathinaraj was a SAS developer and got interested in predictive analytics in 2017. He focused on learning courses in statistics, exploratory data analysis (EDA), machine learning, data science and deep learning, from Coursera and Udemy. This is the first time he has ranked in the top five in a hackathon. Winning Approach Here, Rathinaraj transformed all the time captured information into seconds and label encoded all the categorical features. The target variable was log-transformed to reduce skewness in the distribution. After that, the redundant features were eliminated, along with the highly correlated feature. With that, a new feature was created, which is the time taken in all the three sectors, minus the PIT time. He created further features that show the improvement that happened in all the sectors. Along with that, Rathinaraj also created a variable from the ‘event’ column information, just taking the Free Practice and Qualifying Group, and eliminated all the numeric suffixes. He then created one additional feature as frequency encoded values for each categorical variable. Model Building In this step, feature deletion, addition and selection were performed to avoid overfitting as the test data contains location 6, 7, 8 qualifying group lap alone. He evaluated the LightGBM, CatBoost and XGboost model but for final prediction, and used the CatBoost model with five-fold cross-validation. Talking about his experience, Rathinaraj said, “It’s been some time since I started regularly participating in MachineHack hackathons. It has been an extremely exciting journey for me and indeed very useful for my learning,” said Rathinaraj. Check out the code here. Rank 04: Praveen Kumar Bandla Bandla came across the term ‘data science’ when he took a business data mining course while studying for his MBA at IIM Calcutta. He was hooked instantly by the way maths and programming can be employed to help solve complex business problems. While working at EXL, he got the opportunity to work with a US insurance client with their analytics team. He went on to pursue a PGP in data science, offered by Simplilearn in association with Purdue University and IBM. Since then, he has been participating in ML hackathons and has learnt a lot from these competitions. Winning Approach To start with, Bandla worked to understand the dataset and the features that were provided. He researched the context of the problem statement to get a better understanding of the task. After this, he performed EDA to explore the distribution of features and their relation to the target variable. After he figured out the features he wanted to use, he trained basic models to get an idea of where he stood on the leaderboard. He said, “Then, I would experiment with feature transformation, feature engineering, model tuning, boosting, stacking and so on. This would give me an idea as to how complex models are performing with the given dataset compared to simpler ones.  In this competition, I found that simpler models perform better than complex models.” Check out the code here. Rank 05: Mahima Arora Arora comes with a bachelor’s degree in mathematics and a master’s in operations research. It has only been a year since Arora started working, but the experience has opened her up to different concepts, a variety of tools and a vast possibility to explore and learn more. Winning Approach After exploring the data, Arora started with some data cleaning that included fixing the formats of different variables and converting them into a usable form. Then, Arora performed univariate and bivariate analysis to understand the data better. In the next step, she merged weather data with the original dataset and aggregated it on the combination of location, event and source of data. With this, she calculated the mean for each of the columns and merged it with the original dataset. The imputation was carried out on columns with 60-70% missing data. After this, Arora converted categorical variables into dummy variables and dropped the irrelevant columns. She split the data into train and validation and started building a model using XGBoost regressor, random forest and gradient boost algorithms. She used k-fold cross-validation to tune her models and fine-tuned XGBoost Regressor with “Mean Squared Log Error” as an objective function, which gave the best performance on her validation data. Arora’s experience at MachineHack has been enriching and fulfilling. She stated, “From cleaning the data and applying different algorithms to fine-tuning, the model has increased my overall understanding of this field. These hackathons provide a great platform to learn as well as compete in a healthy environment to improve and enhance your existing knowledge.” Check out the code here. Out-of-the-box solutions, high degree of skills displayed The Dare in Reality hackathon saw participants bring out-of-the-box solutions to the table to solve the innovative problem they’d been presented with. Having such a high level of skills on show at the Dare in Reality hackathon surely made it a huge success. “We were amazed at the number of carefully considered solutions the hackathon received to our challenge,” said the Envision Racing team. “With the data science community demonstrating such a high level of innovation, the five winners should be particularly proud of their success. We’re already exploring how we can adapt their ideas to help the team gain an edge in qualifying.”","excerpt":"Let’s take a look at the winners of the ‘Dare in Reality’ hackathon who took home cash prizes and goodies.","categories":["Deep Tech"],"tags":["AI (Artificial Intelligence)","Data Science","Envision Racing hackathon","Genpact","Hackathon","Machine Learning"],"author_name":"Sreejani Bhattacharyya","publish_date":"2022-02-11T10:00:00","publication_year":"2022","word_count":1478,"keywords":["data science","machine learning","Genpact","AI","neural network","ML","Envision Racing hackathon","Machine Learning","Hackathon","Ray","deep learning","XGBoost","analytics","LightGBM","Data Science","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","neural network","data science","analytics","Ray","XGBoost","LightGBM"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/meet-the-winners-of-the-dare-in-reality-hackathon\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10064137,"title":"How can genetic algorithms be applied to supply chain optimization?","content":"Genetic Algorithms (GA) are a special set of evolutionary algorithms, these algorithms try to simulate the evolution of biology evolution but in the domain of numbers. The genetic algorithm is one of the tools that can be used to apply evolutionary computing methods to find good, sometimes even optimal, solutions to problems that have billions of potential solutions. Implementing this kind of progressive based algorithm in Supply Chain Management could help to solve the complexity of SCM that has been increased over time. In this article, we will put focus on how Genetic Algorithm techniques can be applied for optimization in SCM. Following are the topics that would be covered in this article. Table of contents About Genetic Algorithm (GA)About Supply Chain Management (SCM)Applications of GA to SCM About Genetic Algorithm (GA) A Genetic Algorithm (GA) is a research-based algorithm based on the theory of natural evolution. This algorithm works on the process of natural selection where those individuals are selected for the processing of who is the perfect fit with the help of fitness calculation to expand it to the next generation. Genetic algorithms use important biological features for optimization: The environment is defined by the problem to be treatedChromosomes represent candidate solutions to the problem.The genotypes encode the candidate solutions for the problem. The genotype-phenotype translation determines how the chromosomes should be interpreted to obtain the actual candidate solutions.The fitness of individuals depends on different factors of the problem so that more adaptable individuals are more likely to survive.A population of individuals develops, into which new individuals enter and others disappear.New individuals emerge as a result of the recombination and\/or mutation of previous individuals, whereby their fitness increases steadily. This explanation is shown in a flow chart representation below for a better understanding of the computational steps of GA. About Supply Chain Management (SCM) Supply chain management (SCM) involves managing upstream and downstream relationships with suppliers and clients to provide high-quality, low-priced customer value as a whole. There are a total of five major steps in SCM planning, sourcing, manufacturing, logistics and defective product returning. Following is the flow chart for the SCM. The main purpose of SCM is to: Make inventory always ready according to the customer’s demandsAchieve cost-efficient fulfilment and inexpensive productsEnhance the response towards the changes like economic crisis, government guideline change, etc for better organisational responsiveness.Build a network that gets optimised according to time.Make a profitable business. Applications of GA to SCM Over time a lot of diversity has been created in the supply-demand which made the supply chain management complex to calculate the solutions for the business. To solve this complexity of the supply chain genetic algorithm is implemented, to have an optimised solution out of the list of solutions with the help of biological methodology. Let’s have a glance over these applications and try to understand the implementation of genetic algorithms in SCM. Inventory analysis using genetic algorithms (GA) Maintaining the inventory according to the supply-demand ratio is a complex problem to solve because of a lot of diversity in the data. The objective is to predict an optimum stock level by using the records.  Let’s understand the flow of operation to solve this problem Initially, the data related to the number of stock levels are been labelled as: The zero (0) refers to the contributor needing no inventory control.The non-zero (1,2,3,.. according to the requirement) data requires inventory control. It consists of both the excess amount and the shortage amount. The excess amount is labelled as a positive value and the shortage amount is labelled as a negative value. Following that, the labelled data is fed to a clustering algorithm that separates the stock levels that are either in excess or shortage from the stock levels that are neither in excess nor shortage. This is done simply by clustering the zero and non-zero values. The efficient K means the clustering algorithm is the perfect algorithm for clustering this kind of data. After the process of clustering, the work starts its proceedings on the Genetic algorithm, the core of the final solution. Let’s start with defining the chromosome, they are randomly generated initially by having the stock levels within the lower and the upper limit for all the distributors and contributors of the supply chain and the factory. The gene is the stock level of each member of the chromosome. If the length of the supply chain is n then the length of the chromosome is n. In this application, the length of the chromosome is 3 (n=3) since using only three members of the chain. Initially, there would be a total of two-parent chromosomes and from the next generation, a single random chromosome will be produced. Moving towards the next step, to check the fitness of each individual in the population. The fitness functions ensure that the evolution is toward optimization by calculating the fitness value to evaluate the performance of each individual in the population. It sorts for the best chromosome. These chromosomes may be represented several times in the next generations, hence leading to a population that is composed of several copies of the same solution which also doesn’t guarantee that the initial population contains the globally optimal solution. Genes on each chromosome that are right of the crossover point are flipped over, which results in a cross-over operation. Following the crossover operation, two new chromosomes are formed. This image shows the status of the chromosomes before they do crossover in the operation. As the negative 800 value should be with a negative chromosome so implementing a single crossover could be achieved. Hence the problem has been solved by implementing the cross over. The newly obtained chromosomes from the crossover operation are then processed for mutation. The mutation is the process of generating a new chromosome that doesn’t resemble the candidate chromosome. This is done by a random generation of two points and then performing swaps between both the genes. This process will keep iterating and simultaneously two chromosomes would be selected by the fitness function are going to be the main solution. Each iteration would give the best chromosome. There should be an optimal number of iterations. Thus, the final chromosome selected would be the solution for inventory control. The Vehicle Routing Problem (VRP) The Vehicle Routing Problem (VRP) is a combinatorial optimization problem in which several customers, requiring either pick-ups or deliveries, must be serviced by a set of vehicles. The objective is to schedule the transporters in such a manner that each customer is visited by exactly one vehicle and the total distance travelled is minimised. Transporting the manufactured goods to the distributors and scheduling those is a tough job to do with a lot of diversity in the market. To solve this problem genetic algorithms are applied. A famous brand of acrylic paints “Asian paints” uses this algorithm to route their vehicles for a seamless and profitable business. To solve this problem initially the data should be encoded, for the encoding algorithm uses the “random key” encoding technique in which it generates a random number between 0 and 1 for each customer in the chromosome. The order in which the customers are visited is represented by sorting the random numbers in ascending order. Random keys are used because it prevents mutation (reproduction) infeasibility. These chromosomes are then pushed to penalise the infeasibility in other words it is the calculation of fitness for the chromosomes. Like this, a multi-stop problem there is a need for a cycle cross over method. In this, a gene from one parent will be copied into a child, but it should inherit the position of the other parent. Once the cross over chromosomes that are perfect for the mutation are selected by the test, they are used to generate a new chromosome for the problem. Hence, using the genetic algorithm now the vehicles are scheduled as the transporters in such a manner that each customer is visited by exactly one vehicle and the total distance travelled is decreased. Production cost minimization There is an issue with the SCM which is a shorter product life cycle due to which there is high uncertainty of demand. This is one of the reasons for the increase in the production cost of a product because manufacturers need to spend more due to this uncertainty. Uncertainty is measured by the frequency of its occurrence, and by analysing the relative contribution and effect of the uncertainty on performance. The impact of uncertainty can be quantified as minor or major. To solve the problem the data is encoded and initial chromosomes are initialised with the help of uniform distribution. Producing successive generations it is important to give more chances to the “fittest” individuals selected with help of calculating the individual’s fitness score. The fitness score is the probability that is assigned based on the rank from a truncated geometric distribution. The initial chromosomes are formed now there may be a chance of duplication in the future generation so to avoid that need to use a single-point crossover in which the new offspring will inherit genes from the first parent till the cut point and inherit non-repeating genes from the beginning of the second parent as shown below. The new crossover chromosomes will then be used for the mutation to produce new chromosomes with the help of iteration to achieve an optimal chromosome or the final solution(s). Nutshell A Genetic Algorithm is a powerful tool with the concept of evolution as the backbone of the algorithm which helps to formulate and optimise the solutions. It is widely used in different fields and this covers the major applications of genetic algorithms in supply chain management. References Paper on GA applicationPaper on GA in supply chainTypes of crossoverGA application in production cost","excerpt":"Genetic Algorithms (GA) are a special set of evolutionary algorithms, these algorithms try to simulate the evolution of biology evolution but in the domain of numbers. Implementing this kind of progressive based algorithm in Supply Chain Management could help to solve the complexity of SCM that has been increased over time.","categories":["AI Trends"],"tags":["genetic algorithms","supply chain management"],"author_name":"Sourabh Mehta","publish_date":"2022-04-02T13:00:00","publication_year":"2022","word_count":1631,"keywords":["Go","supply chain management","AI","programming_languages:R","ML","programming_languages:Go","GAN","R","genetic algorithms"],"extracted_tech_keywords":["AI","ML","R","Go","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/how-can-genetic-algorithms-be-applied-to-supply-chain-optimization\/","complexity_score":3,"technical_depth":7,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10059606,"title":"Atlassian acquires Percept.AI, to integrate with Jira","content":"Atlassian Corporation Plc, a leading provider of team collaboration and productivity software and the maker of Jira Software, Confluence, Bitbucket and Trello products, has acquired Percept.AI to expand frontline support capabilities in Jira Service Management. The acquisition of Percept.AI is another step towards providing better employee and customer experiences at scale, that empowers people and streamlines processes with the right tools. Percept.AI helps teams automate tier-1 support interactions and free up IT teams to focus on more complex tasks. Jira Service Management now has more than 35,000 customers. “By integrating Percept.AI with Jira Service Management, support teams will be able to deliver exceptional service even faster and, more importantly, at scale. Our broader vision is to create a unified platform for any form of support and service desk. This acquisition further advances Jira Service Management’s smart experiences, enabling seamless chat-based conversations between customers and agents,”Edwin Wong, Head of Product Management for Jira Service Management, Atlassian said. Atlassian acquired Percept.AI for its ability to understand better the context behind a support query. Its conversational AI engine analyses and understands intent, sentiment, context, and profile information to personalize interactions. The acquisition is the latest step for Atlassian towards a future of service management that is driven by AI and smart experiences. Teams at more than 225,000 customers, across large and small organizations – including Bank of America, Redfin, NASA, Verizon, and Dropbox – use Atlassian’s project tracking, content creation and sharing, and service management products.","excerpt":"The acquisition is the latest step for Atlassian towards a future of service management that is driven by AI and smart experiences.","categories":["AI News"],"tags":["Mergers and Acquisitions"],"author_name":"Ebin K. Gheevarghese","publish_date":"2022-02-01T19:21:26","publication_year":"2022","word_count":243,"keywords":["programming_languages:R","AI","ML","ViT","GAN","Mergers and Acquisitions","R"],"extracted_tech_keywords":["AI","ML","R","GAN","ViT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/atlassian-acquires-percept-ai-to-integrate-with-jira\/","complexity_score":2,"technical_depth":6,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":16983,"title":"AIM lists top 7 reasons HPE Vertica 8 is a good fit for every organization","content":"Analytics India Magazine, recently delved into why HPE Vertica 8 is the  best fit for businesses to create an analytics-driven enterprise! [See the article, here.] Without further ado, here are the top 7 reasons that make HPE Vertica 8 a good fit for every organization: Reason #1: Vertica 8 has a definitive edge vs Hadoop-SQL combinations that are yet to achieve interactive performance. HPE Vertica 8 not only provides SQL layer on top of Hadoop but it also supports fast data access to both ORC and Parquet. Vertica can connect and read data from any Hadoop instance. Everyday users rely on SQL query and Vertica 8 allows customers to access files in HDFS stored in ORC, Parquet format and achieve significant performance benefits compared to raw text files. Reason #2: Expanded Cloud Integrations – Run Vertica in the loud that is relevant to the customer. The idea behind expanded cloud integration was to “enable users to run Vertica in the cloud, that is relevant to them, without being locked into a particular cloud.” This gives the customer the choice which cloud to run in – with the expanded cloud support, it is about more deployment options, but helping customers manage it well in the cloud. The multi-cloud integration is in line with HPE’s strategy of helping customers accelerate their digital transformation. As part of the latest release, HPE Vertica 8 now supports Microsoft Azure Cloud and the new release also features expanded AWS support with access to S3. Reason #3: Advanced compression capabilities deliver results at speed and scale. As opposed to legacy technologies, HPE Vertica 8 has a full-featured analytics system that reduces big data analyticsquery to minutes and even seconds. The power of Vertica lies in parallelism and there are many emerging use cases where customers are finding more ways to use this architecture.  High degree of concurrency and parallelism is at the heart of Vertica’s success and massively parallel processing can handle data at petabyte scale and speed. Vertica gives a compression ratio of approximately 1:8 or maybe beyond and the benefits are twofold – a) data gets compressed; b) significantly reduces footprint for the infrastructure and reduces the cost. HPE Vertica 8 platform is 50x–1,000x faster than legacy data warehouse solutions and cranks out 10x–30x more data per server. Reason #4: Expanded analytical database support for Kafka, Spark and Hadoop Vertica 8’s extensive integrations to Apache Spark, HDFS and Kafka allows users to analyse the data where it resides.  The open format means customers can analyse data as it is without transforming or moving it. Which means that instead of dealing with big batch loads of data every few hours, microbatch streaming has become one of the most popular ways of getting data into Vertica. Vertica provides Kafka Connector to support real time stream data ingestion from different sources. It also provides Spark connector so Spark applications can read from Vertica database into Spark memory and can be processed along with machine learning algorithm. Integration with Hadoop to support largest ingestion is a strong spot for Vertica. Vertica provides capability to read and write into Hadoop and can read different files formats from Hadoop like ORC, Parquet, Avro, Json. It also enables users to stores its own data file into Hadoop that makes Vertica flexible to integrate with any Hadoop distribution and act as a fast data processing layer along with Hadoop. Reason #5: HPE Vertica 8 is the right choice for big analytics workloads. The new-age analytics platform is specifically designed for big analytics workloads and packs a wide range of analytical functions to support faster decision making. It is suitable for OLAP applications as opposed to OLTP application since Vertica is very fast on data ingestion, querying and analysing patterns. Vertica is packed with inbuilt prediction modelling, sentiment and geospatial analytical capabilities that gives it an edge over other competitors. As part of the latest release, Vertica also has support for native machine learning algorithms. One of the key advantages of HPE Vertica is that queries run faster — 50–1,000x faster than any data warehouse. Vertica’s in-database machine learning enables users to embrace Big Data and accelerate business outcomes. Under the new release, for in-database machine learning, the parallel machine learning algorithms have been brought inside Vertica, so that users can effectively analyse data and make predictions without exporting it out of Vertica. Use cases: MTS India has made HPE Vertica a key part of their core telecom business and the proof lies in the number. The telecom player leverages the big data platform for driving accurate, targeted customer campaigns to understand how customers were responding to promotions – outcome resulted in doubling the conversion rate. The Big Data implementation at MTS India has significantly brought down the overheads in business operations. When ICICI Bank required information on whether they should deploy more ATMs in a particular region or understand the kind of transactions customers were having, they used HPE Vertica Analytics Platform built-in geospatial analysis function to get insights. One of India’s leading banks, ICICI has approximately 4000 ATMs across the country and depending on the geography and population, the bank wanted to ascertain whether they should be set up more ATMs. Reason #6: HPE Vertica 8 shines on scalability, speed & performance. It’s been more than a decade since Michael Stonebraker, database pioneer whose groundbreaking idea of an architecture that stores data in columns rather than rows, paved the way for HPE Vertica. Now, its mature columnar storage makes ‘hot data’ available faster than a traditional RDBMS solution and that makes it the right fit for large enterprises demanding high performance and ease of scalability. Vertica works on Massive Parallel Processing (MPP) concept that provides high performance and scalable architecture on commodity hardware.  It means, nodes can be added on the fly and Vertica does automatic data balance within all nodes i.e. within cluster. The performance automatically increases as you scale out the cluster size. It helps to reduce the cost and further reduces TCO for running the system. Features like Parallel Load and Apportioned Load options makes data load option faster in Vertica 8. Case in point, a single large file or other single source is divided into segments, that are assigned to several nodes to be loaded in parallel. Vertica also provide option to define target wise loading in advance which makes system understand the loading strategy in advance, thereby yields to faster data loads in the system. The speed and scale at which data ingestion happens in Vertica 8 is phenomenal. Customer sessions reveal the scale at which Vertica operates is mind boggling with users loading 3000 unique tables an hour, each table featuring up to 1500 columns, underscores the volume of data ingestion. Reason #7: HPE Vertica’s pricing model makes it a good fit for big & small players. The right big data analytics platform is one which fits the needs of large organizations and SMBs.  HPE Vertica’s pricing model scores big from a cost perspective. Haresh Krishna Kumar Nese, HPE Country Manager – India & SAARCexplains why the HPE Vertica is the best choice for organizations, big and small alike. “SMBs have high requirement for solutions and they need analytics to improve operations and generate new revenue streams. Hence, we have reduced the entry barrier by offering a subscription based model so that smaller players can also adopt HPE Vertica platform and make better business decisions,” Nese stated. The subscription model strategy seems to have paid off well, since most customers worry about a “Vendor lock-in.” Licensing Model: Under the perpetual licensing model, customer can invest in a license and get the support features as well. Subscription Model: The subscription based model has significantly reduced the entry barrier for SMBs that can adopt advanced analytics database platform by paying a less amount upfront. By removing in a lock-in period, the entry barrier has been significantly minimized. For more details on HPE Vertica, click here.","excerpt":"Analytics India Magazine, recently delved into why HPE Vertica 8 is the  best fit for businesses to create an analytics-driven enterprise! [See the article, here.] Without further ado, here are the top 7 reasons that make HPE Vertica 8 a good fit for every organization: Reason #1: Vertica 8 has a definitive edge vs Hadoop-SQL combinations that are yet […]","categories":["AI Trends"],"tags":[],"author_name":"Дарья","publish_date":"2017-08-14T05:47:42","publication_year":"2017","word_count":1327,"keywords":["machine learning","AWS","AI","R","Apache Spark","RAG","analytics","Kafka","SQL","Azure"],"extracted_tech_keywords":["AI","machine learning","analytics","RAG","AWS","Azure","Apache Spark","Kafka","R","SQL"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/aim-lists-top-7-reasons-hpe-vertica-8-good-fit-every-organization\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":31686,"title":"Trust Is Central To Banking And CustomerXPs Helps Them Leverage It, Says Rivi Varghese","content":"Founded by fintech product experts, CustomerXPs believes that like a human soul, a customer too has a soul – the sum total of what a customer is. “Among all the industries, banking is the only one that captures the customer’s soul,” said Rivi Varghese, CEO, in a candid conversation with Analytics India Magazine. He shared some interesting insights about how their company leverages artificial intelligence, machine learning and analytics, to provide real-time enterprise solutions to global banks, their growth plans and more. Varghese entered the world of AI and real-time analytics about 18 years ago. A course on statistics during his MBA at IIM Bangalore exposed him to methodologies for massive data crunching which encouraged him to conceive a startup that focused on the same. “In 2006, we started creating solutions using fuzzy logic and eventually advanced to tools in AI,” he says. They then decided to sharp focus on developing extreme real-time solutions for banks using AI, ML and analytics. “Among all the industries, banking is the only one that captures the customer’s soul.” -Rivi Varghese Analytics India Magazine: Why did you begin with banking as your area of interest? Rivi Varghese: When we started about 10 years ago, we had already interacted with more than 200 banks in over 20 countries. That’s when I realised that banking is that one unique sector where the entire life of the customer goes through the bank. While every other industry has a fragmented, unidimensional view of its customers, a bank knows everything about you — how much you earn, your marital status, where you travel to, whether you live on rent, even how much fuel your vehicle consumes. But we found it odd that banks were saying they were in the relationship business, but in reality, they were not quite aware how exactly. We started CustomerXPs to help banks actually discover their customers’ soul, and realised that AI and ML are the way forward. We started developing our product using computational technologies and fuzzy logic engines. We are talking about chatbots now, but our customers went live with it eight years ago. AIM: How is your approach of providing AI to banks different? RV: We believe in putting the end state in perspective and then work towards it. We put substantial effort into understanding a client from scratch before developing the AI solution. AI for us is the means to an end and not the end in itself. Almost every bank has made an investment in some way or the other in AI and ML. The usual practice is to buy costly software licenses, put hundreds of ML engineers and spend millions of dollars buying huge hardware. After years of such costly spends, we still find that banks are saddled with “know why” and the science, but still haven’t figured out the “know how” – as to what is the exact problem they are trying to solve. We decided to tilt this practice. We sit down with banks to understand their problems in depth before developing solutions that work for them. Mere models and fancy tech jargons will not solve problems – we need a different, real, smarter way of getting things done. AIM: Does that mean that you offer customised AI solutions to banks? RV: Let’s first look at the three top hurdles preventing large banks from deploying large-scale AI. First — they don’t have access to absolute real-time data. This is because data warehouse is more or less a failed concept, and everywhere the data is late by hours, days or even months. The data is not structured properly and even if you feed it into the ML models, they don’t work the way they should. Second — if you have non-siloed data, though you could create phenomenal models which help in fraud detection and protection, they still lack the last mile connectivity to influence a transaction is in flight, the flitting moment of truth. Not being able to do this, reduces your AI efforts to just yet another report.  The best way to do this is with an enterprise-wide anti-fraud system, which is connected in real-time to all the decision points in a bank. Third — the sheer complexity of managing huge software. Conventional products are dated and most of the time, with origins from the statistical background. If you really want to do AI, you need to look at what the world is doing today, now and be on the bleeding edge and not have a dated approach. So, what we have instead is an appliance model. We put all the information into the appliance which contains all the latest software including GPU. GPU does massive volumes of parallel processing and we encapsulate all of this into a box. While the data is incoming real-time, the massive processing can be done easily. In the appliance model, a bank doesn’t incur huge capital expenditure, as it is a monthly subscription, and there is no need for licenses – think of software and hardware being free. This keeps the vendors on their toes to deliver every month, instead of taking 100% of the licenses up front and leaving the bank to figure out what it intends to do. “While every other industry has a fragmented, unidimensional view of its customers, a bank knows everything about you.” -Rivi Varghese AIM: Interesting. And how you are dealing with the issue of money laundering? Where does analytics fit in? RV: Banking is about customers’ trust and the moment you lose it, it is gone. Our core product is in enterprise fraud management and financial management which consists of fraud, anti-money laundering and compliance. We use AI and ML to ensure that all of these works better. When we say AI and ML for money laundering, it means a lot of things. Dated AML systems, generate a lot of alerts. In a differentiated approach, we need to mimic human behaviour and understand what they’re trying to do. The engine learns from it so that the human effort with respect to intervention, alert resolution times, and better hits are addressed. From an investigator’s perspective, instead of looking at many alerts, there are fewer alerts in the first place, so lesser time is spent per alert, and the overall accuracy is increased. Say the investigator has to enter a phone number search from millions of accounts for interlinkages. With traditional analytics, it takes anywhere from a couple of hours to a few days to process unstructured data, but in our case, our AI engines accelerate this and get this intelligence in seconds. What we are doing is applying ML to work in sync with our core belief and not because AI is trending. AIM: Tell us about the analytics and AI tools you use RV: We have a core product which is a real-time decision engine, to analyse, detect and stop fraud within the transaction window itself. The other is a thinking engine in our appliance model. In both cases, we use open source tools and technology such as dashboarding, enhanced in-memory capabilities to constantly improve our system. AIM: Could you highlight some of the use cases in banks? RV: Our use cases are centred around fraud detection, compliance, anti-money laundering and generating alerts. For instance, in fraud, it is about the quantum of money saved for the bank. AI and ML at its core, operate on probability. There type A and type B errors. Making the error rate close to zero but not zero requires substantial levels of optimisation. The other models work on information asymmetry in banking systems such as locating the user based on their place of carrying out a transaction. This metadata is also important for ML parameters. Every bank is now thinking about AI within its own realm and when we look at banks there are only a few labelled cases. Say out of 100 transactions, only 1 is fraudulent. And, they land up extrapolating it to 99 times. So, if you have a 10% error in your basic model, the error rates get multiplied by 99 times and the whole thing becomes unpredictable. When you go deeper into serious, large-scale enterprise deployment for banks, these are the hard problems that the banks are trying to crack – and we crack these innovatively. AIM: How big is your team? RV: We are now close to 140 FinTech experts who work on building, customising and implementing large-scale banking enterprise systems. We know it is vital to first define each problem precisely and then decide on the right ingredients to solve it. Once we have these sorted, we then decide on hiring. We work with leading System Integration partners to give us the scale. AIM: What is the roadmap for 2019? RV: Every day the world wakes up, we will make available to our banks ‘whatever the world has learnt till then’ global intelligence to be used for every decision related to “trust” in a bank.  This is the goal that we are working towards, we expect all our key customers signing up for this vision of ours.","excerpt":"Founded by fintech product experts, CustomerXPs believes that like a human soul, a customer too has a soul – the sum total of what a customer is. “Among all the industries, banking is the only one that captures the customer’s soul,” said Rivi Varghese, CEO, in a candid conversation with Analytics India Magazine. He shared […]","categories":["AI Features"],"tags":["anti fraud software vendors","Interviews and Discussions"],"author_name":"Srishti Deoras","publish_date":"2018-12-18T06:27:27","publication_year":"2018","word_count":1516,"keywords":["machine learning","artificial intelligence","AI","chatbots","R","ML","anti fraud software vendors","RAG","Aim","analytics","fraud detection","Interviews and Discussions"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","analytics","Aim","RAG","chatbots","fraud detection","R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/trust-is-central-to-banking-and-customerxps-helps-them-leverage-it-says-rivi-varghese\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10173226,"title":"Musk Claims that Grok-4 is Better Than Cursor","content":"Elon Musk led ​​xAI has launched Grok 4 and Grok 4 Heavy, two versions of its latest AI model. The company described Grok 4 as a single-agent version, while Grok 4 Heavy is the multi-agent version. CEO Musk took to X to claim that Grok-4 outperforms the AI coding tool Cursor. “You can cut & paste your entire source code file into the query entry box on grok.com and Grok 4 will fix it for you!” he said in a post on X. “This is what everyone at xAI does. Works better than Cursor,” he added. As announced by Musk, the company also plans to launch a coding model focused on being “both fast and smart,” expected to be released in a few weeks. Both models are available immediately and come bundled with access to Super Grok tiers, where users can direct a network of Grok agents to assist with research and productivity. Grok 4 is also accessible via API, and xAI claims the model leads key reasoning benchmarks, including ARC-AGI-2. Meanwhile, over the past few days, users have been cancelling their Cursor subscriptions en masse, voicing outrage over what they see as an abrupt and poorly communicated change, particularly to the company’s heavily promoted “unlimited” Pro plan. Users claim that Cursor’s Pro plan, which once promised unlimited usage, has effectively been capped. “Cursor constantly changing their pricing without any proper announcement or documentation recently is a pretty bad and frustrating move for users,” said one user. Cursor has issued a clarification, admitting they “missed the mark” with the updated pricing and are fully refunding affected users. But that hasn’t quelled the criticism, as Reddit threads are filled with complaints that go beyond pricing alone.","excerpt":"xAI is also planning to launch a coding model in the coming weeks.","categories":["AI News"],"tags":["cursor","XAI"],"author_name":"Siddharth Jindal","publish_date":"2025-07-10T14:26:10","publication_year":"2025","word_count":285,"keywords":["Go","cursor","API","AI","R","RAG","XAI","Aim","ViT","Rust","xAI"],"extracted_tech_keywords":["AI","xAI","Aim","RAG","R","Go","Rust","API","ViT","XAI"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/musk-claims-that-grok-4-is-better-than-cursor\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":32950,"title":"Here’s What You Need To Know About Confidence Intervals","content":"Testing a proposal or the workability of a design using hypothesis testing is standard practice in the corporate world. Be it changing the user interface on a mobile application or checking a model which is used to diagnose a patient for psychotherapy, the most inexpensive accessible non-human decision maker is the flipping of a coin. With all the confounding variables associated with real-world problems, flipping a coin can make things go for a toss — literally. Clients and stakeholders may or may not understand the intricacies involved in the model. They don’t care about the type of activation function used or the optimisation technique followed. It always comes down to one question: How does the model work in the worst case scenario? This is where the Confidence Interval (CI) estimate comes into the picture. CI is generated on range and probability. Range, which is the lower and upper limit on the skill that can be expected on the model. Probability talks about whether the model belongs to the range or not. Source: Construction Of Confidence Interval The CI is often referred to as the margin of error and may be used to graphically depict the uncertainty of an estimate on graphs through the use of error bars. For Classification Accuracy In Machine Learning A machine learning algorithm is well understood by the data scientists and the engineers who develop them but when the product needs to be pitched, the only parameter that counts is its performance. So, a metric to gauge the performance of a model is necessary. Classification accuracy is used to assess the efficacy of a classification algorithm. To report the classification accuracy of the model alone is not best of practices. Classification Accuracy = correct predictions\/ total predictions It is common to use classification accuracy or classification error (the inverse of accuracy) to describe the skill of a classification predictive model. For example, a model that makes correct predictions of the class outcome variable 75% of the time has a classification accuracy of 75%, calculated as: accuracy = total correct predictions \/ total predictions made * 100 Classification accuracy or classification error is a proportion or a ratio. It describes the proportion of correct or incorrect predictions made by the model. Each prediction is a binary decision that could be correct or incorrect. Technically, this is called a Bernoulli trial, named for Jacob Bernoulli. The proportions in a Bernoulli trial have a specific distribution called a binomial distribution. Source: The Psychologist We can use the assumption of a Gaussian distribution of the proportion (i.e. the classification accuracy or error) to easily calculate the confidence interval. In the case of classification error, the radius of the interval can be calculated as: interval = z * sqrt( (error * (1 - error)) \/ n) In the case of classification accuracy, the radius of the interval can be calculated as: interval = z * sqrt( (accuracy * (1 - accuracy)) \/ n) Where interval is the radius of the confidence interval, error and accuracy are classification error and classification accuracy respectively, n is the size of the sample, sqrt is the square root function, and z is a critical value from the Gaussian distribution. Technically, this is called the Binomial proportion confidence interval. A code snippet to calculate the accuracy scores: # split the data into a train and validation sets X1, X2, y1, y2 = train_test_split(X_train, y_train, test_size=0.5) base_prediction = base_model.predict(X2) error = mean_squared_error(base_prediction, y2) ** 0.5 mean = base_model.predict(X_test) st_dev = error X1, X2, y1, y2 = train_test_split(X, y, test_size=0.5) base_model.fit(X1, y1) base_prediction = base_model.predict(X2) validation_error = (base_prediction - y2) ** 2 error_model.fit(X2, validation_error) mean = base_model.predict(X_test) st_dev = error_model.predict(X_test) Check the idea behind this method here Common Misconceptions About Confidence Intervals A 95% confidence interval does not mean that for a given realised interval there is a 95% probability that the population parameter lies within the interval. The 95% probability relates to the reliability of the estimation procedure, not to a specific calculated interval. A confidence interval is not a definitive range of plausible values for the sample parameter, though it may be understood as an estimate of plausible values for the population parameter. A particular confidence interval of 95% calculated from an experiment does not mean that there is a 95% probability of a sample parameter from a repeat of the experiment falling within this interval. So, it is essential to remember that: 95% confidence is confidence that in the long-run 95% of the CIs will include the population mean. It is a confidence in the algorithm and not a statement about a single CI. In frequentist terms, the CI either contains the population mean or it does not. There is no relationship between a sample’s variance and it’s mean. Therefore we cannot infer that a single narrow CI is more accurate. In this context “accuracy” refers to the long run coverage of the population mean. Look at the visualisation above and note how much the widths of the CIs vary. They can still be narrow but far away from the true mean. Conclusion A confidence interval is different from a tolerance interval that describes the bounds of data sampled from the distribution. CI provides bounds on a population parameter, such as a mean, standard deviation, or similar and, to deal with the uncertainty inherent in results derived from data that are themselves only a randomly selected subset of a population. It is said that preferring hypothesis testing to confidence intervals and estimation will lead to fewer statistical misinterpretations. Confidence intervals can be unintuitive and sometimes are as misunderstood as p-values and null hypothesis significance testing. Moreover, CIs are often used to perform hypothesis tests and are therefore prone to the same misuses as p-values. Real world data is filled with noise, is inconsistent, non-linear. So, a single “significant” CI can be mighty useful to draw conclusions which otherwise would be cumbersome.","excerpt":"Testing a proposal or the workability of a design using hypothesis testing is standard practice in the corporate world. Be it changing the user interface on a mobile application or checking a model which is used to diagnose a patient for psychotherapy, the most inexpensive accessible non-human decision maker is the flipping of a coin. […]","categories":["Deep Tech"],"tags":["classification accuracy","common ai misconceptions","Machine Learning"],"author_name":"Ram Sagar","publish_date":"2019-01-07T10:36:29","publication_year":"2019","word_count":986,"keywords":["Go","machine learning","programming_languages:R","common ai misconceptions","AI","ML","Machine Learning","programming_languages:Go","classification accuracy","RAG","R"],"extracted_tech_keywords":["AI","machine learning","ML","RAG","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/heres-what-you-need-to-know-about-confidence-intervals\/","complexity_score":4,"technical_depth":8,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10162180,"title":"The Death of Hard Drives at Data Centres","content":"At the start of 2024, companies such as AdaniConnex, Reliance, Sify, Atlassian, Yotta, and AWS ramped up investments in India’s rapidly expanding data centre industry. AWS alone had announced a significant commitment of $12.7 billion to bolster its presence in the country. According to a report, India’s colocation (colo) data centre capacity across its top seven cities reached 977 MW by the second half of 2023. This marks a substantial addition of 258 MW during 2023, reflecting a 105% year-on-year growth compared to the installed capacity in 2022. The country is poised for even greater expansion, with 1.03 GW of under-construction colo capacity slated for completion between 2024 and 2028. Against the backdrop of such exciting developments, Pure Storage India, an advanced data storage platform, makes a bold prediction: Hard drives will likely be phased out in three to five years. In an exclusive interview with AIM, Shawn Hansen, general manager\/VP at Pure Storage, compared this transition to DVDs, which were once everywhere, but were quickly replaced by streaming media. “It felt like an overnight shift,” Hansen expressed while anticipating a similar trajectory for hard drives. He believes that Flash storage is increasingly seen as the future, offering efficiency, speed, and sustainability that hard drives just can’t match. What is so Special About Pure Storage? Pure Storage’s core mission is to create the most efficient and dense storage solutions. Hansen mentioned that the company started with a 2-terabyte Flash module and has now developed a 75-terabyte one, with plans for a 150-terabyte module soon. However, these upgrades come with no increase in power or cooling requirements. He explains that hard drives, by contrast, are limited by their mechanical nature and consume much more power. “AI is incredibly data-hungry, and traditional hard drives just are not viable anymore,” Hansen explains. Flash storage, which uses one-tenth the power of hard drives, is the clear winner. This transition is especially crucial as data centres face the challenge of growing demand while managing costs and energy consumption. Pure Storage’s flash technology allows companies to reduce power use, free up space, and lower cooling requirements. This efficiency translates to more capacity for AI workloads without the need to expand data centres. Flash storage is not only efficient but also becoming more affordable. “We’re reaching price parity with hard drives, and that’s driving a dramatic shift in the market,” Hansen noted. Pure Storage is confident this shift will accelerate as vendors exit the hard drive market, making way for denser storage technologies like flash. Why is Pure Storage in Bengaluru? Hansen claimed that India, owing to its incredible pool of technical talent, is at the heart of this shift. Pure Storage credits Bengaluru, a city known for its startups and innovation, for having found some of its brightest engineers. “The talent density here is incredibly high,” he remarked. Over the past two years, the company has doubled its team size annually in Bengaluru, focusing on projects that require deep expertise in storage and AI. “The talent here is not just benefiting us but the global tech ecosystem. We are incredibly impressed with the leaders we have met here and look forward to continuing our partnership with this region,” he said. Bengaluru is not just another location for Pure Storage. It is one of their three main innovation hubs, alongside Santa Clara in Silicon Valley and Prague in Europe. Each hub has its speciality, and Bengaluru plays to its strengths by tackling challenging storage and AI projects. “It’s not just about outsourcing,” Hansen shared proudly. The company has developed technologies like DirectFlash, which makes storage solutions more efficient and reliable. “All our firmware for DirectFlash is developed in Bengaluru. The team here gives us a multiyear lead over competitors,” he said. Another key innovation is Fusion, which lets companies manage storage like a public cloud—simple, flexible, and scalable. Pure Storage also works closely with major players like NVIDIA, collaborating on solutions like the SuperPod NGX, which combines GPUs and storage for seamless AI infrastructure. Besides, the company has secured a partnership with one of the top four hyperscalers (name undisclosed) to integrate flash storage into their data centres, hinting at a major milestone in the industry’s shift from hard drives. Meta’s AI Research SuperCluster (RSC), one of the fastest AI supercomputers globally, is built to train next-generation AI models on massive datasets spanning petabytes. Powered by Pure Storage’s FlashArray and FlashBlade, RSC efficiently meets the GPU and storage demands while keeping operational costs low. “We contacted a number of storage vendors of both disk and flash to evaluate their highest performance and highest density offerings. From a combination of performance and power and cost, we ended up selecting Pure Storage,” Vivek Pai, AIRSC storage lead at Meta, remarked. “Today, 80% of storage purchases are concentrated in a few companies, but that’s about to change,” Hansen explained. Pure Storage is ready to provide the storage solutions needed to power this next phase of AI.","excerpt":"‘Today, 80% of storage purchases are concentrated in a few companies, but that’s about to change.","categories":["IT Services"],"tags":["data centre india"],"author_name":"Shalini Mondal","publish_date":"2025-01-25T10:03:00","publication_year":"2025","word_count":826,"keywords":["data centre india","API","ELT","AWS","AI","ML","Scala","RAG","Ray","Aim","R"],"extracted_tech_keywords":["AI","ML","Aim","Ray","RAG","AWS","R","Scala","API","ELT"],"url":"https:\/\/analyticsindiamag.com\/it-services\/10162180\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":37948,"title":"How Machine Learning Models Can Help In Getting A Better Harvest","content":"Ever since the dawn of agriculture, man became a changed animal. He no longer moved from place to place. He no longer risked his life to hunt in the wild. He got domesticated by his own techniques of producing food, which led to what everyone calls today- a settled creature. The fundamentals of farming like ploughing and sowing though still stay the same, it’s the ‘how’ part that has changed throughout time. From estimating the number of oxen to rear, we have upgraded ourselves into thinking about the intricacies of farming like manipulating the nutrient levels in the soil and predicting rainfall for moisture adequacy. In India, Andhra Pradesh State Development Planning Society records and reports the average rainfall, which in turn is used to predict moisture adequacy index(MAI) in real time. Whereas India’s largest agro-chemical producer United Phosphorus Limited tied up with Microsoft to deploy machine learning models to predict the risk of pest infestation on cotton crops based on the weather conditions. In case of heavy risk, the farmers are notified to take up necessary action. Along with benefitting the crop growth, machine learning algorithms can also be used in price forecast for the produce. This will protect the farmers from price crash and the common man from high inflation. Data such as historical sowing area, production, yield, weather,are used to predict the timing of arrival of grains in the market as well as their quantum, which would determine their pricing. Measuring Roots And Toasting Leaves With ML Tracking root traits is one of the important tasks of the researchers to evaluate plant health. The traditional methods are invasive and the evaluation tricky. The team at the Department of Molecular Genetics,Spain came up with a novel method to measure root length of the plants with minimum intervention. MyROOT is a software that helps researchers carry out in-depth analysis of the root length. A digital image of the plate containing growing seedlings is used as input. The software calculates the equivalence between pixels and millimeters. Then a binary mask is generated for image segmentation. A regression curve is computed to identify the starting point of each root. This is only one of the many use cases of the indirect application of machine learning techniques in this field. Now they are being applied more directly as well. Something for as direct and tangible as taste of the food. Source:MIT The researchers at the Massachusetts Institute Of Technology(MIT), have figured out a way to make basil(tulsi) leaves taste better by using ML algorithms and of course, Botany. These algorithms determined the most optimised weather conditions for increase in concentration certain flavoured compounds in the leaves which can make them more delicious. In their study of basil plants, the researchers found, to their surprise, that exposing plants to light 24 hours a day generated the best flavor. “You couldn’t have discovered this any other way. Unless you’re in Antarctica, there isn’t a 24-hour photoperiod to test in the real world. You had to have artificial circumstances in order to discover that,” noted Caleb Harper, principal research scientist at MIT’s Media Lab. A Future Of AI-Driven Agriculture From detecting shifting weather patterns such as increase in temperature, changes in precipitation levels, and ground water density to changing the chemistry, machine learning algorithms have certainly made their presence felt in the agro-world. Using vast amounts of data to increase the yield rate or harness medicinal benefits  could be made possible by leveraging machine learning models, cloud computing, advanced electronics for sensors. The challenge of getting the information has been solved with the advent of satellite technology. Now with the widespread acceptance of artificial intelligence, it is high time to put this data to use.","excerpt":"Ever since the dawn of agriculture, man became a changed animal. He no longer moved from place to place. He no longer risked his life to hunt in the wild. He got domesticated by his own techniques of producing food, which led to what everyone calls today- a settled creature. The fundamentals of farming like […]","categories":["Deep Tech"],"tags":["MIT"],"author_name":"Ram Sagar","publish_date":"2019-04-18T12:05:43","publication_year":"2019","word_count":620,"keywords":["Go","machine learning","artificial intelligence","ELT","AI","cloud computing","MIT","ML","Git","RAG","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","RAG","cloud computing","R","Go","Git","ELT"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/how-machine-learning-models-can-help-in-getting-a-better-harvest\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":25510,"title":"What Does The Industry Say About NITI Aayog’s AI Strategy Report?","content":"Earlier this month, the NITI Aayog unveiled a discussion paper which addresses the national strategy on artificial intelligence and other emerging technologies in India. The government think tank identified five sectors to focus its efforts towards implementation of AI to serve societal needs. The five sectors are healthcare, agriculture, education, smart cities and infrastructure, and smart mobility and transportation. The BJP-led government has put enough importance towards new technologies starting various new programmes and setting up an AI Task Force to prepare India for the upcoming Industrial Revolution 4.0. Through a series of events and speeches from earlier this year, PM Modi is seen to have been deliberately showcasing India as well as his government as technologically forward. Analytics India Magazine spoke to several industry insiders to find out their take on the NITI Aayog’s discussion paper and whether the five sectors chosen for implementation of AI were apt and progressive or not. Srinivas Prasad, CEO, Philips Innovation Campus (Bengaluru) I would like to commend NITI Aayog on setting out a vision and strategy for AI in India. AI adoption is progressing rapidly across several industries and a national strategy on AI is a good way to get this momentum going in India. The five areas that have been identified include healthcare coupled with the two-tier approach is a good set way to begin, learn and transform to address the challenges in the various sectors in India. This is a nascent field globally and a great opportunity for the brilliant minds from India to use AI to not just solve the problems in India, but also take these solutions to other parts of the world. We need to ride on this wave which has just started. There are several areas in India where AI is already being applied. I believe that many of the AI solutions, if developed keeping the local context in perspective, the vision of “anywhere, anytime intelligent healthcare” will no longer be a utopian myth, but a reality in India!” Abhirup Ganguly, Managing Director at Impact Analytics We believe that the NAIM marketplace proposed in the policy is a very promising model. The attempts to create curated data and image sets are also great initiatives. We further believe that the government can provide meaningful acceleration to the emerging AI industry in the country by becoming a buyer of ‘homegrown’ AI solutions through this marketplace. While the initiatives regarding medical diagnosis and agriculture taken along with global AI players like Microsoft and IBM can serve as pioneer examples; we believe there are many more opportunities for developing AI applications in these and other areas by Indian firms which the government can implement in government hospitals\/agri-services. Accenture India said in a report: At a time when India is striving to rekindle productivity and growth, AI promises to fill the gap. A full and responsible implementation of AI will open new economic opportunities that would not otherwise exist. The guiding principle should be to create “people first” policies and business strategies, centered on using AI to augment and extend people’s capabilities for the benefit of humankind. Business and policy leaders will have to bring an India-focussed, “people first” mindset to the effort. Time is short. Among the world’s largest companies and economies in AI, only a small minority demonstrate high levels of capabilities and deployment in AI. India must join them in the coming years to enjoy the greatest potential for growth and sustained market leadership. Suman Reddy, Managing Director at Pegasystems, India This paper reflects NITI Aayog’s objective of transforming India through the power of AI. Conducting application-based research through the International Centers for Transformational AI (ICTAI) by collaborating with corporates shows that government’s deeper understanding of how best to leverage AI. However, the paper is still a prototype of the government putting plans into practice, skipping details like specifying synergies corporates can draw from their efforts with ICTAI. While other topics in line with the government’s vision are detailed, they are not accompanied by investment estimates and deadlines. Hence, further updates are awaited. India needs to continue building a robust ecosystem of innovation talent like AI experts, from the large pool of STEM resources. The IT industry is already gearing up, with industry-wide efforts like NASSCOM’s Future Skills initiative to skill 2 million people in its first phase. World Economic Forum has similar plans, with its IT Skills Initiative to skill 1 million by 2021. NITI Aayog will need to take a leaf out of their books and come up with a blueprint in line with the country’s larger growth plans. It could collaborate better with industry, academia and regulators. Expanding the scope of AI is a great enabler for the new economy, but there have to be effective guardrails to ensure strong compliance to data and legal guidelines. This is where we can balance both the risks and rewards of AI, especially with the new data privacy regulation expected to be rolled out in India. My belief is that modern technology will solve these problems, and ensure rapid compliance. Arjun Pratap, founder and CEO, EdGE Networks The government’s push for AI via Niti Aayog is much appreciated by us as a player in the AI segment. Setting up of research centers to foster breakthroughs, IP protection and continuous reskilling to keep talent up-to-date is a great step in this direction. In fact, our AI based platform helps organisations forecast future demand, identify skill gaps and bridge the gap by upskilling existing staff or acquiring new talent. The Niti Aayog policy will not only encourage a higher adoption of technologies like AI, ML, automation, etc., but also help in demystifying the thought blocks among people apprehensive of using technology. The time is right when we must acknowledge the fact that these technologies have a much higher potential for social as well as corporate development and in driving India towards becoming a global leader. Karthik Kadampully, co-founder and CEO, AEON Learning With NITI Aayog being given the mandate to establish the National Program on AI, this blueprint is a welcome move towards transforming various key sectors in India through AI. However, in order to build a robust ecosystem, we will also need the right skilled talent, which has been a matter of concern for the industry for a while now. The immediate focus will now have to be on how to bridge the talent gap in emerging technologies, which in this case is AI. The good news is that India already has a large pool of IT professionals who can be suitably tutored in emerging technologies. This is an ideal scenario where government and industry can rope in edtech companies to provide the required skilling programs, given that edtech companies have the technological capabilities and more importantly access to quality faculty and highly acclaimed mentors. This would consequently result in a high-quality skilled talent pool. NITI Aayog’s National Program on AI has come at an appropriate time, as corporates are currently progressively seeking to upskill their employees, now that emerging technologies such as AI are being widely deployed across industry verticals. Mithun Srivatsa, co-founder and CEO, Blowhorn NITI Aayog looking into AI for All is a good start to look at the confluence of policymaking, enterprise and innovation. AI requires concerted efforts considering that narrow AI will change industries and harnessing demographic dividend requires us to be aware of where the AI\/ML industry is headed. In logistics, AI has always augmented existing planning mechanisms. From route planning, to managing loading factors to co-locating inventory closer to customers, Blowhorn uses a bunch of these applications regularly. India cannot be a laggard in AI. ‘Wait and Watch’ approach will be detrimental to job creation and industry flexibility. Currently, dependency on global products (such as Google Maps) powering local Indian companies is high. NITI Aayog should focus on strategic stacks (mapping, banking etc.) and drive more homegrown efforts.” Nitin Tanwar, founder and CEO, Climate Connect Technologies Clean data is the biggest bottleneck for the application of AI at scale outside of Indian urban areas. Climate Connect Technologies has deployed AI at solar power plants across 12 states in North, Central and Southern India. According to our experience, India lacks reliable data gathering and processing infrastructure outside major cities to truly leverage AI. Historically, there have not been enough weather stations in India’s rural areas, hence there is a dearth of reliable clean weather data of appropriate resolution, to “train” AI models built for sectors such as renewable generation and agriculture. India’s data scientists and software engineers will need to innovate clever ways to overcome such unique Indian challenges, that are, of course, not seen in Europe or US. This is where the move by NITI Aayog can leverage the expertise of companies like Climate Connect, in using AI to address our environmental and clean energy challenges. Guruprasad Srinivasan, President, Heptagon Technologies (A Quess Company) It’s time we also embrace it fully and establish our leadership in the global economic scenario. The right implementation of AI in a strategised way will be helpful in tackling societal challenges and impact people’s lives in a positive manner. It becomes far more relevant in the business context to leverage technology in gaining a competitive advantage by building sustainable solutions. At Quess, we are constantly working on digitally-driven solutions to fasten processes. The paperless onboarding platform, built by our digital solutions enabler, Heptagon Technologies, is fully automated and has resulted in saving approximately 1.5 million sheets of paper. Technology becomes a positive disruption when implemented correctly and the NITI Aayog paper has pointed us in the right direction. Sanjay Bahl, CEO and MD, Centum Learning The emergence of AI is both a factor and a result of decreasing cost of storage, exponential growth of data, unlimited computing power and rapid rise in AI investments. AI can provide personalised, targeted, and dynamic intervention with easy accessibility for all the 5 sectors mentioned in the NITI Aayog report. This would give a wider lens for effectiveness, business impact, and outcome-driven solutions with real-time selection of content. Yogesh Ghorpade, CEO and founder, Uplodefoodie With this initiative from NITI Aayog few other problems can be solved to a large extent. This strategy is expected to ease the Root Cause Analysis of the problem, which always lies in the behavioural traits of individuals and lack of intelligent ways and means. These solutions will affect individuals’ behaviour by promoting appropriate practices across varied sectors. By using latest developments in technologies like AI, IoT, AR and Robotics, these issues can be solved in better way. In addition, there has been a lot of development in India which has resulted from latest technological solutions including, Mobility, Artificial Intelligence, Robotics, Data Analytics and IoT. Ashutosh Lawania, co-founder, mfine NITI Aayog’s comprehensive AI strategy strikes the right balance between economic growth and social development. A national strategy with focus on the sectors with high degree of societal needs like healthcare, education and agriculture will not only drive inclusive growth but will also enable a better quality of life. The proposed two-tiered structure for knowledge sharing and development can drive exponential growth of AI innovation in India. With CORE (Center of Research Excellence) and ICTAI (International Center of Transformational AI), AI stakeholders can together work on the development of existing research and supplement it by new knowledge creation. Private sector collaboration will enable India to lead in the deployment and development of AI solutions and also collaborate with other emerging markets and build sector-specific solutions. The approach of transparent NAIM (National AI Marketplace), where all stakeholders from government to startups to institutions come under one platform, will help drive faster adoption of AI solutions. As an AI-powered healthcare startup, mfine is excited to see NITI Aayog committed to solving the challenges in accessibility to quality healthcare. We believe that high quality and on-demand access for primary healthcare is an area where AI can make a significant contribution by enabling early detection, effective diagnostics, and faster decision making and treatment. Abhay Pendse, chief architect, Corporate CTO Organisation, Persistent Systems AI will play a significant role in healthcare in the areas of ‘Diagnosis through image processing’, ‘Disease predictions and preventions’ and ‘Drug discovery’. AI can enable timely and accurate treatment to many patients in rural areas potentially saving millions of lives. NASSCOM spokesperson told AIM: NITI Aayog’s National Strategy for Artificial Intelligence paper comprehensively covers everything India needs to do in the space. We have been consistently working with our member companies in alignment with these strategies to help bridge the AI skill gap. One pivots on skilling, and how FutureSkills platform can support the growing demand for jobs in AI, and another is start-ups, where we are actively engaged through our Centers of Excellence to help co-create solutions.","excerpt":"Earlier this month, the NITI Aayog unveiled a discussion paper which addresses the national strategy on artificial intelligence and other emerging technologies in India. The government think tank identified five sectors to focus its efforts towards implementation of AI to serve societal needs. The five sectors are healthcare, agriculture, education, smart cities and infrastructure, and […]","categories":["IT Services"],"tags":["Accenture","Agriculture","AI (Artificial Intelligence)","augmented intelligence for smart industry","education","Healthcare Automation","industry-wide analytics","infrastructure","Mfine","NASSCOM","NITI Aayog","Persistent Systems","Philips","smart cities","transportation"],"author_name":"Prajakta Hebbar","publish_date":"2018-06-18T07:29:42","publication_year":"2018","word_count":2129,"keywords":["API","industry-wide analytics","Agriculture","Git","Mfine","transportation","AI (Artificial Intelligence)","R","Accenture","infrastructure","artificial intelligence","augmented intelligence for smart industry","RAG","analytics","Philips","Go","AI","Persistent Systems","ML","Healthcare Automation","smart cities","education","Aim","NASSCOM","NITI Aayog"],"extracted_tech_keywords":["AI","artificial intelligence","ML","analytics","Aim","RAG","R","Go","Git","API"],"url":"https:\/\/analyticsindiamag.com\/it-services\/industry-views-niti-aayogs-ai-strategy-report\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":67288,"title":"How AI\/ML Can Help Marketing Projects, Explains Subramanian Gopalaratnam, CTO of Resulticks","content":"For marketing teams and marketing technology companies, finding and growing valuable customers is the key to constructing a profitable business. Large marketing firms for years have recognised that personalised customer experience and streamlined operations are the big factors for marketing success. Research says organisations which utilise AI for marketing are reaping the benefits of increased customer satisfaction and loyalty while improving operations. The rise of omnichannel marketing automation is giving them a much-needed boost as they strategise to scale and optimise their customer engagement and intelligence in real-time. But what is marketing automation? It basically refers to a broad range of automation and analytic tools for marketing, which are purpose-built for marketing firms and departments to more efficiently market on various online channels (web, mobile, email, social media, etc.), gain intelligent insights and automate repetitive tasks. To know about marketing automation and how AI\/ML can help companies in omnichannel customer engagements, we connected with Subramanian Gopalaratnam, CTO, Resulticks, a marketing automation company. Resulticks has embedded advanced AI and machine learning models on its platform to achieve real-time, omnichannel marketing. Here are the edited excerpts from the interaction – AIM: What do you think is the importance of using AI in marketing automation? Subramanian Gopalaratnam: Due to the proliferation of devices and channels, today’s digitally-savvy consumer expects a seamless, highly personalised, and continuously evolving customer journey right from the first encounter to research to purchase and beyond. The days of cookie-cutter communication and spray-and-pray targeting are long gone. There are vast volumes of diverse data that brands must capture, unify, analyse, and translate into meaningful, contextual engagement at any given moment. AI and ML, embedded into integrated marketing platforms, help marketers study the dizzying array of customer activities across channels to derive patterns and insights from optimising personalisation and contextualising interactions at scale. AIM: Can you elaborate with a few specific examples\/use cases on how AI can support marketing units? Subramanian Gopalaratnam: AI and ML can bring value for channel and send time optimisation where you predict the best channel and time for communicating with a particular individual. Using AI\/ML, one can predict the next-best content, offer, product recommendation, or interaction at the individual level for optimal impact. You can derive insights on how to add value to a customer’s journey based on their activities and profile such as propensities, content consumption and purchases. A company can facilitate seamless omnichannel switches to continue an ongoing conversation or customer journey, and finally, dynamically determine the most relevant interactions in context and based on AI\/ML algorithms. AIM: What is the unique value proposition Resulticks bring to customers? Subramanian Gopalaratnam: AI and ML, like all of Resulticks’ diverse modules, are part of an integrated platform built to enable personalised, omnichannel, continuously optimised, measurable customer engagement. The platform has been developed with the technology to allow the algorithms to process and operate on top of rapidly increasing volumes of data so brands can receive the maximum value. AIM: How are your customers deploying and finding value using your services? Please elaborate with a few examples. Subramanian Gopalaratnam: Thanks to Resulticks’ AI and ML models, our clients have been able to maintain a focus on the profile completeness of their audience and receive actionable insights for data augmentation. They can enhance segmentation efforts with capabilities such as AI-driven look-alike targeting, and optimise the channel, time, content, offer, and so on of every communication for maximum impact. They also acquire robust ML-based benchmarking capability to compare marketing efforts with, for example, industry and own benchmarks to make projections. AIM: What according to you are the main technology components for an organisation which wants to advance in AI\/ML? Subramanian Gopalaratnam: For acquiring sufficient AI and ML capabilities, an organisation’s technology strategy should include: – A Continuous Implementation\/Continuous Deployment (CI\/CD) pipeline to support the model deployment. A data pipeline that can not only support batch and stream processing but manage the ingestion, transformation, and loading of data to support AI and ML models. A services platform that is deployment-agnostic, which means the models in operation should be supported across pure cloud and on-premises deployment. Integration components that enable connectivity to all relevant source systems where data should be integrated and published. AIM: You also use blockchain for your platform. Why did you decide to leverage it, and what value can blockchain bring to marketing? Subramanian Gopalaratnam: When it comes to marketing, there is a growing need to keep a singular view of the individual customer as well as their interactions across the ecosystem of partners, with minimal reconciliation efforts required. This is something that can only be achieved with a blockchain, one that shares the insights across the ecosystem, from the brand to partners and even customers. By introducing the marketing data blockchain into the picture, you can reduce the need for handling reconciliations, thus minimising costs and fraud risks alike, and provide customers with an accurate record of how their data has been used. Using blockchain, we are also facilitating efforts to achieve and maintain a consistent understanding of customers across the ecosystem, and achieve streamlined benchmarking that gives valuable insights.","excerpt":"For marketing teams and marketing technology companies, finding and growing valuable customers is the key to constructing a profitable business. Large marketing firms for years have recognised that personalised customer experience and streamlined operations are the big factors for marketing success.  Research says organisations which utilise AI for marketing are reaping the benefits of increased […]","categories":["AI Features"],"tags":["Interviews and Discussions","Marketing","marketing analytics","translate"],"author_name":"Vishal Chawla","publish_date":"2020-06-12T18:00:00","publication_year":"2020","word_count":853,"keywords":["Go","machine learning","AI","ML","marketing analytics","CI\/CD","Git","RAG","Ray","Aim","translate","R","Marketing","Interviews and Discussions"],"extracted_tech_keywords":["AI","machine learning","ML","Aim","Ray","RAG","R","Go","Git","CI\/CD"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-integrated-marketing-platforms-embedded-with-ai-ml-help-companies-explains-subramanian-gopalaratnam-cto-of-resulticks\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10054474,"title":"Google AI Improves Performance Of Smart Text Selection By Using Federated Learning","content":"Google AI recently announced that they have improved the performance of Smart Text Selection by using federated learning to train the neural network model on user interactions responsibly while preserving user privacy. The work is part of Android’s new Private Compute Core secure environment, which enabled Google to improve the model’s selection accuracy by up to 20% on some types of entities. Smart Text Selection, launched in 2017 as part of Android O, is one of Android’s most frequently used features, helping users select, copy, and use text easily and quickly by predicting the desired word or set of words around a user’s tap, and automatically expanding the selection appropriately. Through this feature, selections are automatically expanded. For selections with defined classification types, e.g., addresses and phone numbers, users are offered an app to open the selection, saving users even more time. With this new launch, the model no longer uses proxy data for span prediction but is instead trained on-device on real interactions using federated learning. This is a training approach for machine learning models in which a central server coordinates model training that is split among many devices while the raw data used stays on the local device. A standard federated learning training process works as follows: The server starts by initializing the model. Then, an iterative process begins in which: Devices get sampled.Selected devices improve the model using their local data.Then send back only the improved model, not the data used for training. The server then averages the updates it received to create the model that is sent out in the next iteration. For Smart Text Selection, each time a user taps to select text and corrects the model’s suggestion, Android gets precise feedback for what selection span the model should have predicted. In order to preserve user privacy, the selections are temporarily kept on the device, without being visible server-side, and are then used to improve the model by applying federated learning techniques. This technique has the advantage of training the model on the same kind of data that it sees during inference. Using this new federated approach, Google significantly improved Smart Text Selection models, with the degree depending on the language being used. Typical improvements ranged between 5% and 7% for multi-word selection accuracy, with no drop in single-word performance. The accuracy of correctly selecting addresses (the most complex type of entity supported) increased by between 8% and 20%, again, depending on the language being used. These improvements lead to millions of additional selections being automatically expanded for users every day. One of the advantages of this federated learning approach is that it enables user privacy because raw data is not exposed to a server. Instead, the server only receives updated model weights. Google says that this approach requires the use of federated learning since it works without collecting user data on the server. It also uses many state-of-the-art privacy approaches, such as Android’s new Private Compute Core, Secure Aggregation and the Secret Sharer method.","excerpt":"The work is part of Android’s new Private Compute Core secure environment, which enabled Google to improve the model’s selection accuracy by up to 20% on some types of entities.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","AI Algorithms","Data Science","Data Scientist","Deep Learning","federated learning","federated learning new","Google","Machine Learning","Python","text analysis","text analytics","text dataset"],"author_name":"Victor Dey","publish_date":"2021-11-29T14:38:15","publication_year":"2021","word_count":500,"keywords":["AI Algorithms","R","emerging_tech:federated learning","RAG","text analytics","text analysis","Data Science","Go","machine learning","AI","neural network","Machine Learning","federated learning new","text dataset","federated learning","programming_languages:R","programming_languages:Go","Python","Google","Deep Learning","Data Scientist","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","machine learning","neural network","federated learning","RAG","R","Go","programming_languages:R","programming_languages:Go","emerging_tech:federated learning"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/google-ai-improves-performance-of-smart-text-selection-by-using-federated-learning\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":6431,"title":"ANU to Open Big Data Analytics Centre Soon","content":"Acharya Nagarjuna University (ANU) is all set to inaugurate the ‘National Research Centre for Big Data Analytics’, which is claimed to be first of its kind in the country,  on November 18. The university authorities have invited Guntur MP Galla Jayadev to inaugurate the new facility that has been set up at the engineering college on the campus. The centre aims at providing training for the students as well as faculty members and help them enhance their skills. The big data analytics is the process of examining large data sets containing a variety of data types i.e., big data, to uncover hidden patterns, unknown correlations, market trends, customer preferences and other useful business information. The analytical findings can lead to more effective marketing, new revenue opportunities, better customer service, improved operational efficiency, competitive advantages over rival organisations and other business benefits. “Having knowledge in big data analytics will definitely help the students from several disciplines like mathematics, statistics. management, commerce, computers, pharmacy and all branches of the engineering to improve their employability skills. It will also assist the faculty members to improve their skills and acumen. This is a new revolution in analysing huge data,” said Prof K Viyyanna Rao, vice-chancellor of ANU. The university, which has spent  Rs 15 lakh on the centre so far,  is making efforts to get funding for the projects from the Department of Science and Technology (DST), (Government of India) as the latter had started Big Data Initiative (BDI) programme to promote Big Data Science, Technology and Applications. Under this scheme, the DST will provide financial support to the academicians, scientists, technologists and other practising researchers from recognised academic, research institutions and registered scientific societies.  Since it is a new initiative, the university authorities are also seeking support from other organisations. Principal of the  University Engineering College and Coordinator of the Centre for Big Data Analytics  E Srinivasa Reddy said that they were also planning to enter into a memorandum of understanding (MoU) with the  Centre for Development of Advanced Computing (CDAC), Bangalore. “As per my knowledge, no other university in the country has set up an exclusive centre for big data analytics,” he said and added that they had already submitted several research topics for the BDI programme of the DST. Analysis of the crashing of the Indian shopping site ‘Flipkart’ is said to be one of the research topics by the faculty at the university. ANU is also planning to introduce M.Tech in Big Data Analytics from next academic year, said Prof K Viyyanna Rao, vice-chancellor of ANU. They are yet to get the nod from the executive council. It is an evening course. Source: Indian Express","excerpt":"Acharya Nagarjuna University (ANU) is all set to inaugurate the ‘National Research Centre for Big Data Analytics’, which is claimed to be first of its kind in the country,  on November 18. The university authorities have invited Guntur MP Galla Jayadev to inaugurate the new facility that has been set up at the engineering college […]","categories":["AI Trends"],"tags":[],"author_name":"AIM Media House","publish_date":"2014-11-06T11:52:05","publication_year":"2014","word_count":445,"keywords":["big data","data science","Go","funding","AI","Aim","ViT","analytics","GAN","R"],"extracted_tech_keywords":["AI","data science","analytics","Aim","R","Go","big data","GAN","ViT","funding"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/anu-open-big-data-analytics-centre-soon\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10160842,"title":"NVIDIA’s $3000 DIGITS Supercomputer Coming in May","content":"Leading chipmaker NVIDIA unveiled Project DIGITS, a new small supercomputer, at the Consumer Electronics Show (CES) 2025.  It is aimed at AI researchers, data scientists, and students across the world.  The supercomputer provides access to the GB10 Grace Blackwell Superchip, which Jensen Huang, CEO of NVIDIA, calls a ‘super secret chip’. The GB10 features a 20-core NVIDIA Blackwell GPU, one of the most powerful AI hardware systems available today. The CPU was built in collaboration with MediaTek. Project DIGITS features 128 GB of unified memory and offers storage options of up to 4TB. Similar to a typical computer, DIGITS requires only a standard electrical outlet to operate. It operates on a Linux-based NVIDIA DGX operating system. The supercomputer can also run up to 200 billion parameter LLMs locally, and if you have two of them, NVIDIA says you can link them up to run AI models double the size. what??NVIDIA just dropped Project DIGITS, a $3,000 personal AI supercomputer that’s small enough to look like a Mac Mini but packs 1,000x the power of your average laptop.Handles AI models with up to 200 BILLION parameters.This is incredible.. pic.twitter.com\/z4JOeFD2JI— el.cine (@EHuanglu) January 7, 2025 DIGITS allows users to deploy AI models on the NVIDIA DGX cloud and leverage all the tools present inside NVIDIA’s AI Enterprise software platform. For instance, you can fine-tune models on the NeMo framework and build agents on NVIDIA Blueprints and NIM microservices. The higher performance and portability increase the cost. DIGITS will be available in May of this year and will cost a whopping $3000. “AI will be mainstream in every application for every industry. With Project DIGITS, the Grace Blackwell Superchip comes to millions of developers,” said Huang. “Placing an AI supercomputer on the desks of every data scientist, AI researcher, and student empowers them to engage and shape the age of AI,” he added. At first glance, there’s no doubt that Project DIGITS is entering the Mac Mini territory, at least in terms of the form factor. “I must have the NVIDIA Project DIGITS for my home lab. 128GB pooled RAM, 4TB storage, the size of a Mac mini, Running DGX OS, a Linux-based OS. The coming years are going to be wild in the world of AI and robotics,” said Jamie Madden, a machine learning developer on X. In December of last year, NVIDIA introduced the Jetson Orin Nano Super Developer Kit, a compact generative AI supercomputer now priced at $249, down from $499. According to the company, it offers enhanced performance with 67 INT8 TOPS, marking a 70% improvement over its predecessor, alongside a memory bandwidth of 102GB\/s, which is a 50% increase.","excerpt":"It’s called ‘Project DIGITS’ and can run powerful AI models locally.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","NVIDIA"],"author_name":"Supreeth Koundinya","publish_date":"2025-01-07T11:57:32","publication_year":"2025","word_count":441,"keywords":["Go","machine learning","AI","Git","RAG","microservices","Aim","generative AI","AI research","NVIDIA","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","machine learning","generative AI","Aim","RAG","microservices","R","Go","Git","AI research"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/nvidias-announces-3000-supercomputer\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":22284,"title":"IIT Patna Director Pushpak Bhattacharyya To Head Committee For Standardisation In AI","content":"File photo of IIT Patna director Pushpak Bhattacharyya The new committee for standardisation in artificial intelligence, set up by the Bureau of Indian Standards (BIS), has now got a new chairman — Pushpak Bhattacharyya, director at Indian Institute of Technology, Patna. The committee has been set up by BIS, which in turn falls under the Union Ministry of Consumer Affairs. According to media reports, the committee would focus on standardisation of projects that revolve around cyber security, legal and ethical issues in the IT sector, technological mapping and leveraging AI for national missions, among others. Bhattacharyya, who also heads the Natural language processing research group Centre For Indian Language Technology (CFILT) lab at IIT Bombay, told a newspaper, “A lot of work is being done in artificial intelligence in US and Europe. The Indian government is also working on this front and the committee that has been formed will work on standardisation of projects in artificial intelligence. Since many start-ups are working on different projects in AI, the committee’s main task will be to go through the projects to find if or not standardisation has been maintained in the projects.” Prime Minister Narendra Modi has been showcasing a strong inclination for working towards supporting a tech-driven future. Other initiatives that revolve around artificial intelligence and data science are also being supported by the Indian government. For example, a special committee headed by the National Institution for Transforming India (NITI) Aayog vice chairman Rajiv Kumar was founded with the sole purpose of laying out a roadmap for the country’s research and development in the field. Reportedly, the decision to form the committee came after news of China’s intense and growing commitment in the field of AI. Earlier, the Commerce and Industry Ministry constituted a task force to prepare for the industrial Revolution 4.0 and the resulting economic transformation late in 2017. V Kamakoti, a professor from the department of computer science and engineering at Indian Institute of Technology, Madras, heads the panel for the same. Industry insiders are happy to see that the Modi-led NDA government is working towards supporting a tech-driven future. This year, Union Finance Minister Arun Jaitley has also doubled the allocation on Digital India programme to ₹3,073 crore in 2018-19.","excerpt":"The new committee for standardisation in artificial intelligence, set up by the Bureau of Indian Standards (BIS), has now got a new chairman — Pushpak Bhattacharyya, director at Indian Institute of Technology, Patna. The committee has been set up by BIS, which in turn falls under the Union Ministry of Consumer Affairs. According to media reports, […]","categories":["AI News"],"tags":["AI (Artificial Intelligence)","IIT Bombay","Narendra Modi","NITI Aayog","Quantum Computer"],"author_name":"Prajakta Hebbar","publish_date":"2018-03-06T06:25:57","publication_year":"2018","word_count":373,"keywords":["data science","Go","IIT Bombay","artificial intelligence","programming_languages:R","AI","programming_languages:Go","Quantum Computer","Git","RAG","Narendra Modi","AI (Artificial Intelligence)","R","NITI Aayog"],"extracted_tech_keywords":["AI","artificial intelligence","data science","RAG","R","Go","Git","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/pushpak-bhattacharyya-to-head-ai-standardisation-committee\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10064412,"title":"The global doctorate programme for AI thought leadership","content":"Analytics India Magazine spoke to Sayee Bellamkonda, a doctoral candidate at INSOFE-Rennes School of Business, who shares his experience of pursuing the leadership program in data science. Mr Bellamkonda is a Chief Digital and Technology Officer at Global Mobility Solutions (Vialto Partners, now part of PWC), a global company operating in 60 countries. AIM: What motivated you to choose the Global DBA program from Rennes SB- INSOFE? Bellamkonda: I did my master’s in engineering followed by an MBA in Global Management. In leading teams at ITA organizations like American Express, Ameriprise, CBRE and Vialto Partners, the need for a practitioner’s view of Artificial Intelligence and Data Science became quite apparent. The DBA with Rennes School of Business (RSB) offered in collaboration with INSOFE is one of the only programs that met my requirement of combining data science with leadership skills and business management. AIM: As a working professional, how was your experience taking this program? Bellamkonda: The course is well planned to accommodate our schedules, and we don’t have any weekday classes. The way the content is shared is very convenient as I have access to class recordings, documents and relevant materials at all times. That being said, the program requires time commitment and effort. The DBA brings the best of both worlds – on one side, we have RSB with the best academic faculty, and on the other, we have the practitioners at INSOFE with tremendous experience. Also, the annual visit to Rennes in France is a good initiative that gets us face time with the faculty there. I appreciate the time commitment from the INSOFE mentors, especially Dr Murthy and Dr Gnana (who joins in from Australia to teach our classes). AIM: How did this program help you with your career? Bellamkonda: From my work at American Express, I have seen the potential of how meaningful insights can be obtained from data. Building an enterprise data platform, identifying the right data, and producing descriptive and predictive analytics have become crucial. Irrespective of the industry, it is important to have formal data science and artificial education to engage with such work. Going through this program gave me a new insight into how to address these problems for my wealth management clients in the global commercial real estate space. AIM: Can you share some insights about your research journey? Research Topic: Helping companies identify the right use cases which can be implemented using AI and ML frameworks. Bellamkonda: My research topic is extremely relevant to all practitioners. There is a lot of investment in AI and ML spheres. By 2030, 70% of the companies will be using AI & ML to enhance business. But statistics predict that only 20% of these companies will be able to achieve the set goals. So how do we address this 80% shortfall? This is where my research comes into play. I work on building feasibility frameworks for companies. With this model, organizations will be able to look at various frameworks, recognize the knowledge gaps, and identify the right use cases to be implemented using AI & ML. This will drastically increase the success rate of organizations being able to implement AI & ML. AIM: What has the DBA cohort experience been like? Bellamkonda: My peers from Dubai and India greatly enhanced my learning experience. It was very interesting interacting with peers from diverse backgrounds, such as fashion technology, education, and supply chain management, as I come from a commercial real estate and financial management background. We share our experiences on a weekly basis and discuss the coursework, which helped us stay up to date. Here the learning happens not just from the professors but from peers as well. It’s a culture that has been created in the three-year journey, which is more like a marathon than a solo sprint, where the cohorts motivate each other. AIM: As you have completed a major part of the program, what key takeaways would you like to share with future aspirants? Bellamkonda: Future aspirants interested in pursuing a doctoral program should know that this is an excellent opportunity to network and learn. Having access to strong support and experienced faculty is crucial to succeeding in a doctoral program. This DBA has all that and more. Both RSB and INSOFE are subject area experts with highly knowledgeable professors. In addition, this program also brings together a global cohort of peers from diverse domains. However, this is a doctoral program. It requires a time commitment and cannot be breezed through just by listening to lectures. AIM: What are your thoughts on your final research project and patents, if any? Bellamkonda: My first goal is to complete my doctoral program. Students of the DBA have been taught how to write patent descriptions, patent filing, and the dos and don’ts, which has been an extremely valuable experience. When my proposed research framework is built, it can be offered via open source sites for the world to learn. But it can also be productized. Rennes and INSOFE are given the opportunity to jointly patent these ideas, so we have equal commitment and responsibility in making the product successful. INSOFE is organising a free webinar on April 7, 2022, at 7.00 pm IST, to give an overview of the program and make a case for how a Doctoral degree can prepare you for strategic leadership roles in data science. REGISTER HERE","excerpt":"Analytics India Magazine spoke to Sayee Bellamkonda, a doctoral candidate at INSOFE-Rennes School of Business, who shares his experience of pursuing the leadership program in data science.","categories":["AI Trends"],"tags":["Insofe","INSOFE data scientist programme"],"author_name":"AIM Media House","publish_date":"2022-04-06T11:24:37","publication_year":"2022","word_count":897,"keywords":["data science","INSOFE data scientist programme","Go","artificial intelligence","AI","R","ML","Git","Insofe","Aim","analytics","predictive analytics"],"extracted_tech_keywords":["AI","artificial intelligence","ML","data science","analytics","Aim","predictive analytics","R","Go","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/the-global-doctorate-programme-for-ai-thought-leadership\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10121286,"title":"‘AI Chips Will Eventually Replace GPUs and Even NVIDIA Knows This’","content":"Graphics processing units (GPUs) may have become a coveted piece of hardware in the AI realm, yet their status as the most sought-after component may wane. Unprecedented demand for GPUs has made NVIDIA a trillion-dollar company. However, even NVIDIA is starting to move away from what they originally created as a graphics chip, according to Keith Witek, chief operating officer at Tenstorrent. “They’re even moving their architecture towards the heterogeneous compute, which looks a bit more like a tensor computer. “So yes, I think it will trend in that direction. And even the guys in the graphics business of AI are realising the benefits of drifting their architecture in that direction,” Witek told AIM in an exclusive interaction. He advocates for system-on-chip (SoC) architectures incorporating tensor units, graph units, and CPUs, asserting that heterogeneous computing utilising both CPUs and graph processors is the optimal approach for handling future workloads. Recently, big tech companies like Microsoft and AWS, which are among NVIDIA’s biggest enterprise customers, have developed their own AI chips to reduce their dependency on NVIDIA’s GPUs and simultaneously reduce cost. At the recently held Google I\/O 2024, the tech giant announced Trillium TPUs, its six-generation silicon designed to handle AI workloads more efficiently. Interestingly, chips designed by AWS, Microsoft and Google too have heterogeneous architectures. For example, Azure Maia AI Accelerator and Azure Cobalt CPU integrate different specialised compute engines and accelerators on the same chip. Similarly, AWS Inferentia and Trainium also integrate different specialised compute engines and accelerators on the same chip. However, these chips are meant primarily for their internal use. Tenstorrent, on the other hand, sells its chips to enterprise customers, putting it in direct competition with NVIDIA. “The company’s goal was to build boxes and compute platforms for high-end applications, like data centres and high-performance compute,” Witek said. How Tenstorrent’s AI chips are better than NVIDIA GPU Given their capacity to construct hardware solutions and provide two distinct software stacks, each optimised for their platform’s unique capabilities, users benefit from the flexibility to craft software and models to suit their needs. Leveraging their proficiency in chip design and intellectual property (IP), Tenstorrent regards this as a holistic system-level strategy. One of Tenstorrent’s biggest advantages is that its components architecturally pack more compute density into the box without creating much power and heat. “So we have a very power-efficient compute, where we can put 32 compute engines in a box, the same size as NVIDIA puts eight in a box. With our higher compute density and similar power envelope, we outperform NVIDIA by multiples in terms of performance, output per watt, and output per dollar. Additionally, our software is open source and accessible to the community,” Witek said. Tenstorrent’s AI chips eliminate the need for expensive interconnectors Tenstorrent’s AI chips are designed to minimise the need to access DRAM compared to GPUs, which constantly access DRAM. Due to this distinction, NVIDIA requires costly silicon interposers like HBM memory chips. SK Hynix, which makes HBM memory chips for NVIDIA, announced that they have already been sold out for the year. Moreover, Samsung has reported increased revenue growth resulting from a high demand for HBM memory chips. These days, most data centre AI chips come equipped with HBM memory; however, Tenstorrent believes they can operate effectively without relying on such chips. “We achieve comparable or superior performance using more economical GDDR6 or GDDR7 memory and organic interposers. Consequently, our chips are more cost-effective without compromising performance. This is because our architecture incorporates local cache and routing within the chip, reducing the necessity to access DRAM for internal connections,” Witek said. Furthermore, Witek highlights that NVIDIA relies on expensive interconnects like Mellanox and NVLink to link boxes, racks, and containers within a data centre. In contrast, Tenstorrent can accomplish the same connectivity using Ethernet, which is both affordable and high-performing. “So, why are people using NV Link and Mellanox? Well, they’re using it because NVIDIA can make a tremendous amount of money if they’re forced to use it. But architecturally, it’s not required that people spend that kind of money in their interconnection scheme in the data centre and we’re bringing that to the light of day with the architecture that we’re putting into the market,” Witek pointed out. CUDA is a monster but has become unwieldy Despite being a hardware company with a monopolistic hold in the GPU space, NVIDIA’s real competitive moat has been CUDA. CUDA, which is an acronym for Compute Unified Device Architecture, is a software layer that gives direct access to the GPU’s virtual instruction set and parallel computational elements for the execution of compute kernels. For quite some time, Intel and AMD have been trying to challenge CUDA’s dominance with their own software stack. “CUDA has bloated over the years into a pretty large monster that can do a lot of different things, but it’s become very unwieldy,” Witek said. “All the math libraries… and everything is encrypted, in fact, and NVIDIA is moving their platform more and more proprietary every quarter. They’re not letting AMD and Intel look at that platform and copy it. Programmers are avoiding CUDA whenever possible and writing a lot of code in C++ and other languages,” he continued. On the contrary, Tenstorrent has open-sourced its software stacks. They offer a platform named Metalium, similar to CUDA but less cumbersome and more user-friendly. On Metalium, users can write algorithms and programme models directly to the hardware, bypassing layers of abstraction. However, proficiency in hardware architecture is essential, as errors may require manual correction by the user for proper functionality. “It’s very much like CUDA, but we give you 100% open-source software and 100% open-source models. You can get on Hugging Face, download the model and run it; we make it very easy for you to do that,” Witek said. Tenstorrent’s second software, called BUDA, represents the envisioned future utopia, according to Witek. “Eventually, as compilers become more sophisticated and AI hardware stabilises, reaching a point where they can compile code with 90% efficiency, the need for hand-packing code in the AI domain diminishes. “Although we haven’t reached that crossover point yet, many anticipate it within the next few years. Hence, we are continually enhancing BUDA’s efficiency to prepare for that eventual transition,” he concluded.","excerpt":"CUDA has bloated over the years into a pretty large monster that can do a lot of different things, but it’s become very unwieldy.","categories":["Global Tech"],"tags":["AI GPUs","NVIDIA","tenstorrent"],"author_name":"Pritam Bordoloi","publish_date":"2024-05-23T09:00:00","publication_year":"2024","word_count":1046,"keywords":["tenstorrent","CUDA","Hugging Face","Go","TPU","AI GPUs","AI","AWS","R","RAG","Aim","NVIDIA","Azure"],"extracted_tech_keywords":["AI","Aim","Hugging Face","RAG","AWS","Azure","TPU","CUDA","R","Go"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/ai-chips-will-eventually-replace-gpus-and-even-nvidia-knows-this\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10001395,"title":"Software Pirate’s Code Of Honour: A Dive Into The Warez Scene","content":"Almost everyone who grew up with the Internet has engaged in software piracy. MP3s of popular songs made the rounds on popular sites such as 4shared and Megaupload. Games and software ran amok on sites such as The Pirate Bay. The Internet was booming in the age of everything being free, but where did it all come from? The answer will take us into the deepest parts of the Internet, where rules are set by self-governing entities. Where the world is governed by standards and non-conformity is treated with a hammer. It is from whence the root of the free Internet springs and is ruled by the need for glory. Welcome, to the Warez Scene, where the code of honour for software pirates holds strong. What Are Warez? Warez is Internet-speak used to denote software piracy. Warez are usually cracked software, videogames and MP3s, and are free of cost to anyone willing to spend the bandwidth to download them. They are a natural continuation of the Internet’s philosophy of being free and accessible, but the Warez Scene predated the Internet. This software is usually cracked by a Warez Group, who have many crackers among them. Their job is to find vulnerabilities in the code of the program that prevents it from being copied. This protection is implemented by companies to prevent pirates from copying it, and usually falls to the cracking skills of the Warez Groups. Warez are released and maintained by a cartel of these groups in a loosely-defined organization known as the Warez Scene. The Scene largely rose to prominence in the 1980s cracking software for the Commodore 64 computing platform. As this was before the time of the Internet, they utilized very basic Bulletin Board Systems to communicate over long distances. In the beginning, the Scene was restricted to mainly crackers. It grew as a grassroots subculture, staying as an underground, decentralized organization that collaborated and competed with itself. As the end of the ’90s approached, the rise of the Internet fundamentally changed the hierarchy of the Warez Scene. However, due to its unique governing system, the Scene’s top players remained at the top and continued to provide free content to members of the scene. In fact, the overarching governance of the subculture itself carries themes of old-fashioned meritocracy and stringent standards. The scene even expresses displeasure at being exposed to the public, bringing elements of elitism in a society where glory is currency. The Glory Of Being A Warez Scene Cracker The Warez Scene is run by crackers, as without them there would be no new warez. Crackers function in their tightly-knit groups, which are similar to gangs in the mafia and are subject to a strict hierarchy. A group picks up a new piece of software with the intention of releasing a cracked and usable version as fast as possible. However, what motivates them is not money, as the releases are made for free. Instead, the Scene runs on merit, fame and glory, like some sort of outdated society. Groups rise to prominence for their speed and method of cracking new protection methods and add their signature to software by bundling .nfo files with releases. These .nfo files included elaborate ASCII art which denoted which group had cracked the release. These were also used to criticize other crackers in their ever-ongoing competition. Crackers, however, function purely on the glory of releasing games. Even as they lead normal lives, they are mainly just people with the expertise and knowledge required to crack the complex protections. Moreover, the hierarchy of the Warez Scene provides an environment where underhanded tactics are not allowed, and releases have to stick to a set of standards that is decided by the Scene as a whole. The Hierarchy Of Release And Its Impact Many cracked versions of games and software start off on Warez Scene board sites and are then leaked to IRC boards to distribute it. Then, the release is then distributed onto sites where it can be accessed by P2P services such as BitTorrent. These are then downloaded by the general population on the Internet, which is something that the scene does not encourage. According to them, the software cracked by the scene must be distributed to those in the scene. Exposing the existence of the scene to the general population is also discouraged, as this attracts multiple legal issues. The Scene has also become more restricted over the years, with popular groups even looking to give it up altogether. Video game cracking group 3DM predicted that cracking would be dead in 2 years in 2016, owing to the high amount of protection that games are coming with today. Internet Service Providers have cracked down, which has led to the decreased traffic to sites where cracked games are uploaded, for fear of repercussions from authorities. Especially in the subcontinent, where a healthy piracy culture has grown, ISPs are getting more stringent with users downloading pirated software. Service providers such as BSNL and Airtel were reported to have reduced download speeds for torrent files, and are cracking down using the power of the IT Act 2000. With the rise of better protection software and a more closed Internet, the Warez Scene stands strong as the last bastion for pirates.","excerpt":"Almost everyone who grew up with the Internet has engaged in software piracy. MP3s of popular songs made the rounds on popular sites such as 4shared and Megaupload. Games and software ran amok on sites such as The Pirate Bay. The Internet was booming in the age of everything being free, but where did it […]","categories":["AI Features"],"tags":[],"author_name":"Anirudh VK","publish_date":"2019-03-05T15:49:42","publication_year":"2019","word_count":878,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","RAG","GAN","R"],"extracted_tech_keywords":["AI","RAG","R","Go","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/software-pirates-code-of-honour-a-dive-into-the-warez-scene\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":4135,"title":"Aspiring a Career in Digital Analytics?","content":"That Analytics is the raging talk of the town, many thanks to media glitz, is one of the least disputed topics in the Information technology industry. One of the questions I get asked time and again from many people both inside and outside of Sapient is “How Can I make a career in Digital Analytics”? Thought I will pen a short article to elaborate on this, with the intent of setting junior aspirants in the right track. What is Digital Analytics about? Let’s get on the same page Analytics has been around since the dawn of the IT age – remember data marts, data warehouses, relational and subsequently de-normalized databases whose data is sliced and diced by BI tools. With the paradigm shift from thick to thin clients and the proliferation of the internet, web analytics (tracking web site user behavior) has gained prominence over the last two decades. Over the last few years, Mobile and Social have disrupted the world of marketing and have left the marketing heads of consumer brands in a Do-or-Die situation. Digital Analytics is the set of activities that need to be undertaken in order to measure (and subsequently optimize) visitors’ footprint across all digital channels –web site, mobile site\/apps, online campaigns and such. What do Digital Analytics brethren do? Here is my favorite analogy. Construction of a house is not a single man’s play but rather a concerted team effort of architect(s), masons, plumbers, electricians, and painters. Digital Analytics is no different. At a high level here are some roles (same person can don multiple hats) at play: First Steps for the Beginner I would recommend starting off with Web Analytics as this is the most mature element of the overarching Digital Analytics domain. Here are some tips to get started: Get a grasp of the concepts of web analytics. A good list of some of the terminology and their definitions can be found on Web Analytics Associations web site here. Get a copy of Avinash Kaushik’s awesome book Web Analytics 2.0 and read it. By the time you complete the book you will no longer be a novice. Start with Google Analytics – it is FREE, and fairly easy to learn both from the implementation as well as reporting perspective. Justin Cutroni’s blog is a great resource. Get your hands dirty with Excel. YouTube is one of the best resources on this. Equipped with the above knowledge, your personal charisma and a bit of luckJ, you should be able to clear a job interview and land a job.  From there, you will start picking up invaluable hands-on experience. Venture into the more advanced (and awesome) tools like Adobe SiteCatalyst. Once you get a reasonable grasp of web analytics, you can venture into Mobile\/Social Analytics and\/or predictive analytics honing deep into specific industry verticals. Closing Thoughts As it stands today, fresh college graduates have a reasonable exposure to programming technologies like JAVA and .NET. Analytics is not yet a part of most college curriculum. Should you choose to take a detour from the path chosen by typical college grads and get into analytics, you should be proud to know that you will be among the 1% of the IT work force who made this choice. For experienced IT professionals from other domains, your overall knowledge of the web can only help, though you should expect a fairly steep learning curve during the first year. Fret not, take the plunge. The joy and the job satisfaction can be immense. Good luck.","excerpt":"That Analytics is the raging talk of the town, many thanks to media glitz, is one of the least disputed topics in the Information technology industry. One of the questions I get asked time and again from many people both inside and outside of Sapient is “How Can I make a career in Digital Analytics”? […]","categories":["AI Highlights"],"tags":["best book to learn digital marketing"],"author_name":"Vinay Mantha","publish_date":"2013-10-03T07:00:02","publication_year":"2013","word_count":586,"keywords":["Go","API","AI","R","Git","RAG","analytics","best book to learn digital marketing","predictive analytics","Java","data warehouse"],"extracted_tech_keywords":["AI","analytics","RAG","predictive analytics","R","Go","Java","Git","API","data warehouse"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/aspiring-a-career-in-digital-analytics\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":15930,"title":"Can HR analytics solve major talent challenges in times of mass IT layoffs?","content":"Can HR analytics prevent retrenchment through data-backed decisions At a time when the Indian IT industry is rocked by news of IT layoffs, a lot of challenges have piled up for HR, from slowed hiring to restructuring, HR is grappling with reorientation within the organization. In the words of Kris Lakshmikanth, Founder Chairman & Managing Director of The Head Hunters India Pvt. Ltd, HR has considerably lagged behind in the use of Big Data and Analytics in India. According to a global study by Tata Consultancy Services, only 5% of big-data investments were in made in human resources. Role of HR Analytics in times of mass layoffs Industry experts echo a similar view.  The use of HR analytics is critical not just from talent acquisition and management perspective, but also cost optimization. With automation commoditizing low-skilled jobs and even some analytics functions, it is now a matter of automation versus resource deployment for HR managers. While layoffs definitely mean bad PR for the organization, HR analytics can play a proactive role in preparing companies with data driven decisions, such as reallocating resources to avoid retrenchment. “I know of companies using tools to identify dissatisfied individuals at an early stage & take corrective action,” revealed Lakshmikanth. The HR Technology market is estimated to be $15 billion. According to Arjun Pratap Singh, CEO & Founder, EdGE Networks, “AI and analytics are the driving force behind HR technology and this will drive the new employment economy. From talent acquisition and workforce optimization to workforce transformation, AI be the strategic enabler to HR”. EdGE Networks provides solutions across talent acquisition, workforce optimization and talent analytics.  From attrition analytics to demand & supply forecasting and hiring cost analysis, the company has saved millions of dollars through timely, data-driven actions. Here’s what HR analytics can do: By identifying under-performing business units and teams in the evolving tech landscape, resources can be reskilled or deployed to another team Data-backed decisions lend transparency to processes such as annual reviews. In case of retrenchment, data-driven decisions lend validity to the process By tapping into mountains of data, HR teams can identify internal hires who tend to perform better and cost less to the company. By putting employees in a better role, HR managers can significantly reduce the attrition rate By mapping performance prediction, companies can measure business outcomes and even dole out incentives, raises effectively There is plenty of scope of analytics in filling up positions like Data Analyst, Robotics, AI et al Why are companies lagging behind in adoption of HR analytics? Image: Analytics in HR While organizations are sitting on terabytes of rich employee data, it is massively underutilized. Most companies lack good analytics at their disposal to make sense of it. One reason cited by Lakshmikanth is that there is an over-supply of people vs demand. “I don’t see much change in the next few years,” he said. Another big reason could be that HR has to be “forced up the learning curve” and get familiar with tools and techniques to realize gains from employee data. While the absence of the right analytical resources and data is often cited for slow adoption, a recent chief HR officer survey revealed the senior management wasn’t satisfied with the level of insights extracted from the company data. In other words, lack of good quality data is also seen as an impediment in getting quality results. Another key reason is that while enterprises pay an increased emphasis on hiring digital talent, little attention is paid to the HR department – whether it is equipped to bring innovative solutions to bridge the digital skills gap? The digital transformation of HR is often overlooked as organizations tap into newer ways to hire and onboard talent. According to a CapGemini survey, 63% of companies use traditional methods to source talent. Outlook To become more data-savvy, enterprises must make investment to reskill HR talent and bring them up to speed with analytics adoption. HR teams should be backed by quants and data analysts who can mine employee data to generate insights. Another highly cited view is that since HR is not typically seen as a business-function, HR function has to evolve to the level of a consulting partner and play a key role as change management.","excerpt":"At a time when the Indian IT industry is rocked by news of IT layoffs, a lot of challenges have piled up for HR, from slowed hiring to restructuring, HR is grappling with reorientation within the organization. In the words of Kris Lakshmikanth, Founder Chairman & Managing Director of The Head Hunters India Pvt. Ltd, […]","categories":["IT Services"],"tags":["HR Analytics"],"author_name":"Richa Bhatia","publish_date":"2017-06-28T07:26:23","publication_year":"2017","word_count":712,"keywords":["big data","Go","AI","HR Analytics","data-driven","digital transformation","Git","automation","analytics","GAN","R"],"extracted_tech_keywords":["AI","analytics","R","Go","Git","big data","GAN","digital transformation","automation","data-driven"],"url":"https:\/\/analyticsindiamag.com\/it-services\/can-hr-analytics-solve-major-talent-challenges-times-mass-layoffs\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10058691,"title":"Google AI’s plan for 2022 and beyond","content":"“Artificial intelligence will have a more profound impact on humanity than fire, electricity and the internet”, said Sundar Pichai, Google’s CEO, at the BBC Radio 4 podcast with Amol Rajan. The head declared AI would fundamentally transform the way we live our lives with its applications in healthcare, education and manufacturing. Despite AI’s evolution during the past few years, the technology is still believed to be in its beginning stages, undergoing heavy research to uncover more efficient and accessible implementations with less computational power and training involved. Google AI is one of the leading researchers in this space. https:\/\/twitter.com\/JeffDean\/status\/1480986609591803904 In a recent blogpost, Jeff Dean, Senior Fellow and SVP of Google Research, highlighted the upcoming research themes the company is focusing on. Dean presented the need for continuous research and highlighted the key areas Google is advancing its studies in. Analytics India Magazine has curated a list of the broader themes and study topics Google aims to discover. Creating general-purpose, large scale multi-modal ML through ‘Pathways’ One of the biggest trends for 2022, Google and beyond, is training larger and more capable ML models, especially those powered by NLP. This change is powered by increasing dataset and model size, providing better results in language accuracies on NLP benchmarks. Additionally, the research on transformer models is also accelerating with combining transformer models with convolutional operations for better visual and speech recognition tasks. According to Google AI, large scale multi models are also gaining momentum. They are some of the most advanced models to date, given their ability to work with different input modalities while producing different output modalities. “This is an exciting direction because, like the real world, some things are easier to learn in data that is multi-modal,” noted Jeff Dean. Such models can be achieved by pairing images and text for multi-lingual retrieval tasks or jointly training models on visual and textual data to increase accuracy on classification tasks or co-training on the image, video, and audio tasks to enhance generalisation performance for all modalities. Google is also exploring using NLP as input for applications such as image manipulation, instructing robots on interacting with the world and foreshadowing potential changes to how user interfaces are developed. These models will be capable of dealing with speech, sounds, images, video, languages and possible, structured data, knowledge graphs and time-series data. Additionally, these models will be better trained on self-supervised learning, thereby reducing the effort in creating specific machines for different tasks. Together, these trends can enable a general-purpose model that can handle multiple modalities of data to solve millions of tasks. Google is pursuing research, called Pathways, to enable this next-gen architecture through an umbrella effort. Pathways to generalise millions of tasks Improving ML accelerator performance With a more continuous improvement, Google aims to better the next generation of ML accelerators for faster chip performance and increased system scale. Driving better model architectures through humans and machines Google foresees a continuous improvement in model architecture through both human creativity and machine efforts. Intending to reduce computational energy needed, the company combines human efforts with machine learning algorithms such as the NAS that discovers more efficient ML architectures. Enabling better personal use-cases while maintaining privacy Google has leveraged ML innovations and silicon hardware to allow mobile devices to sense the surrounding environment effectively, as in the Google Tensor processor on the Pixel 6. Google aims to further the advancement of ML in this context, to improve ease of use while boosting computational power needed for personal benefits like photography, video recording, communication, live translation, live captions, and more. Google AI has been working on combining ML with traditional codec approaches in the Lyra speech codec as well as SoundStream audio codec for better fidelity of communication. The company aims to do so while strengthening privacy safeguards. Android’s Private Computer Core is an open-source and secure environment, isolated from the rest of the operating system to ensure the data processed is not shared with any other applications on the phone. These features can communicate to Private Compute Core over a small set of open-source APIs that strips our private information. Google’s plan for 2022 is to evolve the security of interactions further while encouraging better computational for personal usage. The company aims to broaden their technology stack to support neural computing and equip access to interactive intelligent interfaces and act as a social entity. The key to enabling so is undertaking a federated unsupervised approach. Federated reconstruction Broader applications of computer vision Google aims to leverage computer vision to create tools that can address global challenges at a large scale. Additionally, it helps keep an accurate record of building footprints, an integral layer for applications today. Since this type of information entails population data, humanitarian responses or environmental and urban planning, it is challenging to calculate it in developing or under-developed nations. But with the help of computer vision technologies, this data can now be accessed through satellite imageries. Google has leapt doing so with their Open Buildings dataset that locates 500 million+ buildings in Africa, and the company aims to leverage this to provide humanitarian aid after natural disasters. Automate design space architecture for better applications Google is exploring letting an ML algorithm automatically explore and evaluate a problem’s design space for possible solutions. Through a Transformer-based variational autoencoder that creates document layouts, an algorithm for computer architectural decisions, and another that focuses on game playability, Google has stepped foot in this research already. Along with such use-cases, the technology has been used for material discovery in chemistry. The company focuses on accelerating the technology’s use-cases in scientific research for better applications. Deploy assistive ML for healthcare use-cases Google foresees the healthcare sector deploying assistive ML systems to improve breast cancer screening, detect lung cancer, accelerate radiotherapy treatments for cancer, mark abnormal X-rays, stage prostate cancer biopsies and even assist in colonoscopies to help in quality assurance, ensure all polyps are identified and detect elusive polyps. Leverage ML to help people in daily health management An emerging trend in the ML world has been using machines to support daily healthcare needs for people. These include healthcare metrics for heart rate assessment, sleeping wellness, speech recognition for those with impairments, support for those with vision impairments and more. The company sees these as just the start of new use-cases and plans on conducting further research. Mitigate climate changes: EV friendly satellite maps, fusion as an energy source, natural disasters and sustainability Google believes in the power of accurate data to help mitigate climate challenges. The company has done so with their eco-friendly Google Maps that estimates to save 1 million tons of CO2 emissions per year, partly by being EV friendly. They are also furthering research on fusion as a renewable energy source. Additionally, Google is working on addressing wildfires and floods that are becoming a commonality, with initiatives such as the satellite-powered wildfire boundary map to help map the affected area. The company is currently launching this option in Maps, along with their optimisation algorithm for fire evacuation routes. Lastly, as part of their sustainability initiatives, Google is working on making their data centres operate on carbon-free energy by 2030 through better model architecture and ML accelerator types used in ML training. Broadening the definition of Responsible AI beyond Western contexts to sociotechnical ML systems One of the major ethical steps with AI, Google is working to think beyond Western contexts while dealing with the ethical needs of AI. Talking about how assumptions about conventional algorithm fairness frameworks fail in non-Western contexts, Dean pressed on Google’s current position in conducting surveys across continents to understand AI preferences and address the gap. The company is also working on enabling ML applications for smallholder farmers in the Global South and involving community stakeholders in the various stages of the ML pipeline. Google is furthering this section of research through a community-based method and listening to the citizens’ and their needs for sociotechnical ML systems. Addressing privacy concerns in large ML models Owing to the increasing security concerns with the increasing size of ML models, Google is also researching to address and ensure the protection of private information. They are doing so by leveraging techniques like federated learning, private clustering, private personalisation, private matrix completion, private weighted sampling, private quantiles, robust private learning of halfspaces, and in general, sample-efficient private PAC learning.","excerpt":"Google is pursuing research, called Pathways, to enable this next-gen architecture through an umbrella effort.","categories":["Global Tech"],"tags":[],"author_name":"Avi Gopani","publish_date":"2022-01-18T17:00:00","publication_year":"2022","word_count":1397,"keywords":["federated learning","machine learning","artificial intelligence","AI","ML","computer vision","NLP","Ray","Aim","analytics"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","NLP","computer vision","analytics","Aim","Ray","federated learning"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/google-ais-plan-for-2022-and-beyond\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10080290,"title":"India’s Quest to Becoming World’s AI Superpower","content":"India is quickly establishing itself as one of the hubs for AI innovation in the world. Not only do we have a large number of startups solving uniquely Indian problems with AI solutions, the government has also taken many steps to further solidify India’s position as a future AI leader. Today, it was announced that India will take over the chair of the Global Partnership on Artificial Intelligence, an international initiative to ensure the responsible use of artificial intelligence. The committee was founded in 2020—with India as one of the founding members—to tackle topics such as use of AI on grounds of human rights, diversity, inclusion, economic growth, and more. One of its aims is also to support cutting-edge research on responsible artificial intelligence use cases. It was reported that India received more than two-third majority vote, ranking above Canada and the United States of America. The minister of state for Electronics and IT, Rajeev Chandrashekar, will represent India at the meeting held today in Tokyo for the symbolic takeover of the position. This appointment is merely one among the several achievements that solidify India’s position as a potential leader in the global AI space. The government has also taken many steps to catalyse AI innovation in the country, setting the stage for the next decade of AI innovation. Government undertakings to build AI infrastructure In January 2020, NITI Aayog announced their plan to build out a robust AI infrastructure for enterprise and enthusiast use. In a paper titled, “Establishing an AI Specific Cloud Computing Infrastructure in India”, they described establishing something known as AIRAWAT—an India-first cloud computing infrastructure focused on providing AI-specific services. While this platform has not materialised as of yet, the paper described a made-for-AI cloud infrastructure that had support for various features such as a ML and DL software stack, multi-tenant multi-user support, a low-latency high bandwidth network, and a multi-layer storage system to quickly ingest and process multi-petabytes of big data. Although it is possible for India to become one of the deepest pools for AI talent, many budding developers are crippled by a lack of availability of hardware resources to study algorithms. AIRAWAT is aimed at serving the computing requirements of Centres of Research Excellence, Innovation hubs, and International Centres for Transformational AI. The keen focus on research and development think-tanks is a display of the Indian government’s big picture perspective of the global AI space and their hope to enter it in a grand way. Initiatives on the ground to improve AI knowledge and education Apart from providing the required computing infrastructure for researching AI, the government has also taken steps towards upskilling the population from a young age for AI roles. Recently, the Ministry of Electronics and IT announced an innovation challenge titled, Youth for Unnati and Vikas with AI. Building on an older initiative titled, ‘Responsible AI for Youth’, YUVAi targets children in classes 8–12 and aims to equip them with basic AI skills. Moreover, the programme aims to “empower them to become human-centric designers and users of AI”. The government will execute the undertaking in three phases. In the first phase, teachers and students will be onboarded to the platform on a registration basis. Then, students will be put in online orientation sessions and be encouraged to submit ideas for one of eight core AI themes. These themes depict the use of AI in agriculture, healthcare, education, transportation, smart cities, and more. A shortlisted group of students will then attend online deep-dive AI training courses and a three-day boot camp run by AI experts. If their projects are selected, students will be provided with apprenticeship and guidance opportunities, and will also be awarded for their projects. Key undertakings by the government to boost future of AI The 2020 Budget saw Digital India receiving nearly $480 million to increase research in AI, IoT, machine learning, and robotics. In addition to this, the government also offered industry-relevant skill training for ten million youth in India for AI and related technologies. IITs all over the country have collaborated with corporates to establish AI research centres. Some examples include the Robert Bosch Centre for Data Science and AI at IIT Madras, Intel AI Research Center at IIT Hyderabad, IIT Kharagpur IT Innovation Hub, AI Research Lab At Wipro With IISc, and more. These labs offer an opportunity to budding engineers to work with cutting-edge AI research, upskill and deepen the talent pool for the field. The government has also launched the India AI portal. This website was developed by MeiTY and the National Association of Software and Service Companies (NASSCOM), and provides an important resource to track and be updated with AI innovations in the country. In addition to this, as a part of the new education policy, the National Council of Educational Research and Training will include a basic course on AI at the intermediate level to expose students to certain basic concepts of machine learning. While speaking at the Responsible AI for Social Empowerment Summit in 2020, Prime Minister Modi also expressed his desire to make India a global hub for AI. “The teamwork of AI with humans can do wonders for our planet. . . We want India to become the global hub of AI. Our bright minds are already working towards it”, he stated. With the recent moves by the government, India’s positive attitude towards AI development has become clear. As evinced through India’s appointment as the head of the GPAI, we are quickly solidifying our position as a global AI leader. “According to the UNESCO Science Report from 2021, India now spends more on Research than France, the UK and Italy. While its citizens are well-aware, it must be mentioned that India has the largest amounts of Digital payments globally! India’s IT and BPM industry’s revenue is estimated at US$227b in FY2022, an increase of 15.5% YOY. The IT industry also employed more than 0.5m employees in FY2022. India is an established talent powerhouse for the IT industry globally. Little wonder then, that India would remain close to the changing pulse of the IT industry, its new frontiers in AI and ML solution delivery and the beginning of standardisation for AI based solutions. In its own enterprises, government spaces and in its services as IT partners for the enterprises of the world, India continues to lead with AI and ML based solutions. India is, hence, best aware of the changing need of education, research and innovation with AI. Moreover, given the prominent fears with developments in Data and AI on Ethics, Governance and Privacy, India is conscious of how this might impact the future of its citizens. A responsible India is looking to create spaces for such education and awareness in its various national platforms, such as the NITI, in its education and in its laws. It is then, hardly a wonder that India is voted to lead a prestigious GPAI!”","excerpt":"India’s recent appointment as the chair of the Global Partnership on Artificial Intelligence further solidifies our stance as an emerging AI superpower.","categories":["IT Services"],"tags":["Indian government","Responsible AI"],"author_name":"Anirudh VK","publish_date":"2022-11-21T16:00:00","publication_year":"2022","word_count":1152,"keywords":["data science","machine learning","artificial intelligence","AWS","AI","cloud computing","ML","Responsible AI","RAG","Aim","Indian government","edge AI"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","data science","Aim","edge AI","RAG","cloud computing","AWS"],"url":"https:\/\/analyticsindiamag.com\/it-services\/indias-quest-to-becoming-worlds-ai-superpower\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10040785,"title":"I’m A Philosopher In A Room Full Of Computer Programmers: Olivia Gambelin, Ethical Intelligence","content":"At The Rising 2021, the founder of Ethical Intelligence, Olivia Gambelin, shared her insights on the potential use of ethics in developing trusted AI systems and how it can fuel technology innovations. Gambelin is an AI Ethicist who works with entrepreneurs, governments and tech startups to bring ethical analysis into technology development. Currently, Olivia leads a remote team of over 30 experts in the tech ethics domain. She believes in enabling human-centric technology innovation. “I am a philosopher in a room full of computer programmers,” said Gambelin. “Though it may seem like I am out of place, this is the kind of situation and atmosphere that I try and put myself in,” she added. What is ethics? “Simply put, ethics is a study of right and wrong,” said Gambelin. For instance, ethics decides who the self-driving car should run over, or ethics decides if a company should hire only male employees. She said both computer science and ethics are complementary fields. While computer science was born out of mathematical logic, ethics was born out of critical thinking and critical reasoning. It is just a different branch of logic, she added. Gambelin believes ethics is a tool that helps you make the right decisions. “In the business context, ethics brings balanced frameworks and methodologies in place, which allows businesses to come up with the best possible outcomes, both financially and morally. Does ethics slow technology innovations? Most companies and entrepreneurs today think ethics is a roadblock to technology innovations. Gambelin said the decisions taken during the product development cycle are crucial. When overlooked, these decisions leave holes in the product that might come to haunt later. “It is like building a house without a pillar,” she added. That will eventually lead to more and more technology debt and ethical debt, which will eventually make the house crumble. “It becomes important to take technical and ethical decisions early on so that it does not come back to bite you later on,” she added. Citing Harvard Business Review, she said: “Constraints can foster innovation when they represent a motivating challenge and focus efforts on a more narrowly defined way forward.” Replacing ‘constraints’ with ‘ethics’ from the above quote, Gambelin explained that ethics brings values into consideration, and pushes the technical limitations. “When you think about ethics, you are not hindering innovation. Instead, you are accelerating it to develop more meaningful solutions by aligning with ethical values and principles of the society,” she added. For instance, if a company is planning to develop a messenger app, from a technological perspective, it is solving the communication gap. But, from an ethical point of view, privacy and security of message delivery come into play. “In this case, what you have got is a messenger app that meets technical limitations, but not ethical limitations,” said Gambelin. Wrapping up Gambelin said the amalgamation of ethics and technology can lead to long term innovation, and ethics needs to be integrated into every step of the process. “At the end of the day, ethics is a decision-making tool that helps block you from making decisions that could turn into the worst possible outcome,” she added. “You are only as ethical as your last decision,” said Gambelin. Meaning, the developer or a company should consciously make it a habit to imbibe ethical practices and push themselves in the right direction, leading to long-term innovation and business success.","excerpt":"At The Rising 2021, the founder of Ethical Intelligence, Olivia Gambelin, shared her insights on the potential use of ethics in developing trusted AI systems and how it can fuel technology innovations.  Gambelin is an AI Ethicist who works with entrepreneurs, governments and tech startups to bring ethical analysis into technology development. Currently, Olivia leads […]","categories":["AI Features"],"tags":["AI (Artificial Intelligence)","ethics in AI"],"author_name":"Amit Naik","publish_date":"2021-05-26T18:00:00","publication_year":"2021","word_count":566,"keywords":["Go","programming_languages:R","AI","innovation","programming_languages:Go","Rust","ethics in AI","R","programming_languages:Rust","AI (Artificial Intelligence)","startup"],"extracted_tech_keywords":["AI","R","Go","Rust","innovation","startup","programming_languages:R","programming_languages:Go","programming_languages:Rust"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/im-a-philosopher-in-a-room-full-of-computer-programmers-olivia-gambelin-ethical-intelligence\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10133877,"title":"The Secret to Creating the Next Billion-Dollar AI Startup","content":"It’s now widely recognised that selling AI models is a zero-margin game. The next wave of AI startups must capitalise on LLMs in the application layer to tackle real-world challenges. “The next billion dollar startups in AI will play on the application layer and not the infrastructure layer,” said AIM Media House chief Bhasker Gupta in a LinkedIn post. Gupta added that there is a plethora of problems to be solved using AI, and these startups will localise their solutions while maintaining a broad-based approach. Echoing a similar market sentiment was Nayan Goswami, the founder and CEO of Chop. “The next major wave of AI innovation will focus on the application layer, where startups will build specialised vertical AI software-as-a-service (SaaS) companies for global markets,” he said. Goswami further elaborated that with robust foundational models like Anthropic, Cohere, and OpenAI, along with infrastructure companies like LangChain and Hugging Face advancing rapidly, we’re poised to witness a surge in application-layer startups targeting specific verticals. “Think of foundational models as the roadways, and application layers as the vehicles driving on them,” he explained. Finding the Right Application to Build is Key Andrew Ng, the founder of DeepLearning.AI believes AI’s usefulness in a wide variety of applications creates many opportunities for entrepreneurship. However, he advised budding entrepreneurs to be extremely specific about their ideas for integrating AI. For instance, he explained that building AI for livestock is vague, but if you propose using facial recognition to identify individual cows and monitor their movement on a farm, it’s specific enough. A skilled engineer can then quickly decide on the right tools, such as which algorithm to use first or what camera resolution to pick. In a recent interview, Ng explained that the cost of developing a foundation model could be $100 million or more. However, the applications layer, which receives less media coverage, is likely to be even more valuable in terms of revenue generation than the foundation layer. He also said that unlike foundation models, the ROI on the application layer is higher. “For the application layer, it’s very clear. I think it’s totally worth it, partly because it’s so capital efficient—it doesn’t cost much to build valuable applications. And I’m seeing revenues pick up. So at the application layer, I’m not worried,” he said. Perplexity AI serves as a strong example by integrating search with LLMs. Rather than building its own foundational models, the startup leverages state-of-the-art models from across the industry, focusing on delivering optimal performance. The company is planning to run ads as well from the next quarter onwards. However, not everyone is going to make the cut; some startups are going to fail. Statistically speaking, around 90% of startups don’t survive long enough to see the light at the end of the tunnel. Ashish Kacholia, the founder and managing director of Lucky Investment Managers, said, “AI is the future but key is how the applications shape up to capitalise on the technology.” India is the Use Case Capital of AI “India is going to be a use case capital of AI. We’ll be very big users of AI, and we believe that AI can significantly help in the expansion of the ONDC Network,” said Manoj Gupta, the founder of Plotch.ai, in an exclusive interview with AIM. Similar thoughts were shared by Nandan Nilekani when he said that India is not in the arms race to build LLMs, and should instead focus on building use cases of AI to reach every citizen. He added that “Adbhut” India will be the AI use case capital of the world. “The Indian path in AI is different. We are not in the arms race to build the next LLM, let people with capital, let people who want to pedal chips do all that stuff… We are here to make a difference and our aim is to put this technology in the hands of people,” said Nilekani. Krutrim founder Bhavish Aggarwal believes that India can build its own AI applications. Agreeing with him, former Tech Mahindra chief CP Gurnani said, “It’s time to stop ‘adopting’ and ‘adapting’ to AI applications created for the Western world.” Gurnani said that the time is ripe for us to build AI models and apps based on Indian data, for Indian use cases, and store them on India-built hardware, software and cloud systems. “That will make us true leaders in the business of tech,” he added. Notably, Gurnani recently launched his own AI startup AIonOS. Startups Offering More Than LLMs Lately, several AI startups in India have been building services using generative AI. For example, Unscript, a Bengaluru-based AI startup, is helping enterprises create videos with generative AI. Another video generation startup, InVideo, is estimated to generate $30 million in revenue in 2024. Recently, Sarvam AI launched Sarvam Agents. While the startup, backed by Lightspeed, Peak XV, and Khosla Ventures, is not the only company building AI agents, it stands out for its pricing. The cost of these agents starts at just one rupee per minute. According to co-founder Vivek Raghavan, enterprises can integrate these agents into their workflow without much hassle. These agents can be integrated into contact centres and various applications across multiple industries, including insurance, food and grocery delivery, e-commerce, ride-hailing services, and even banking and payment apps. Similarly, Krutrim AI is making AI shopping co-pilot for ONDC. Khosla Ventures-backed upliance.ai is building kitchen appliances  integrating generative AI. Meanwhile, Ema, an enterprise AI company founded by Meesho board member Surojit Chatterjee, recently raised an additional $36 million in Series A funding. The company is building a universal AI agent adaptable to a wide array of industries, including healthcare, retail, travel, hospitality, finance, manufacturing, e-commerce, and technology. Enterprises use Ema for customer support, legal, sales, compliance, HR, and IT functions. Lately, we have observed that Y Combinator is bullish on Indian AI startups, many of which are focused on building AI applications. For example, the creator of AI software engineer Devika, Mufeed VH, who founded Stition.AI, is now part of YC S24 batch. His startup works around AI cybersecurity for fixing security vulnerabilities in codebases, and is now renamed to Asterisk. On the agentic front, Indian co-founders Sudipta Biswas and Sarthak Shrivastava are building AI employees through their startup FloWorks. Examples are aplenty, with India poised to boast 100 AI unicorns in the next decade. In a conversation with AIM, Prayank Swaroop, partner at Accel India, said that the 27 AI startups his firm has invested in over the past few years are expected to be worth at least ‘five to ten billion dollars’ in the future, including those focused on wrapper-based technologies. There are a host of categories, such as education, healthcare, manufacturing, entertainment, and finance, to explore with generative AI, and this is just the beginning.","excerpt":"AI’s usefulness in a wide variety of applications creates a plethora of opportunities for entrepreneurship.","categories":["AI Startups"],"tags":["Editors Picks"],"author_name":"Siddharth Jindal","publish_date":"2024-08-27T18:07:58","publication_year":"2024","word_count":1131,"keywords":["Anthropic","Hugging Face","OpenAI","AI","RAG","Editors Picks","LangChain","Aim","Ray","generative AI","foundation models"],"extracted_tech_keywords":["AI","generative AI","foundation models","OpenAI","Anthropic","LangChain","Aim","Ray","Hugging Face","RAG"],"url":"https:\/\/analyticsindiamag.com\/ai-startups\/the-secret-to-creating-the-next-billion-dollar-ai-startup\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":58568,"title":"10 Open-Source Datasets For Text Classification","content":"One of the popular fields of research, text classification is the method of analysing textual data to gain meaningful information. According to sources, the global text analytics market is expected to post a CAGR of more than 20% during the period 2020-2024. Text classification can be used in a number of applications such as automating CRM tasks, improving web browsing, e-commerce, among others. In this article, we list down 10 open-source datasets, which can be used for text classification. (The list is in alphabetical order) 1| Amazon Reviews Dataset The Amazon Review dataset consists of a few million Amazon customer reviews (input text) and star ratings (output labels) for learning how to train fastText for sentiment analysis. The size of the dataset is 493MB. Get the data here. 2| Enron Email Dataset The Enron Email Dataset contains email data from about 150 users who are mostly senior management of Enron organisation. The dataset was collected and prepared by the CALO Project (A Cognitive Assistant that Learns and Organizes) and contains a total of about 0.5M messages. Get the data here. 3| Goodreads Book Reviews This dataset contains reviews from the Goodreads book review website along with a variety of attributes describing the items. It includes reviews, read, review actions, book attributes and other such. There are a total number of items including 1,561,465. Get the data here. 4| IMDB Dataset The IMDB dataset includes 50K movie reviews for natural language processing or text analytics. This is a dataset for binary sentiment classification, which includes a set of 25,000 highly polar movie reviews for training and 25,000 for testing. Get the data here. 5| MovieLens Latest Datasets This dataset is a collection of movies, its ratings, tag applications and the users. There are two sets of this data, which has been collected over a period of time. The small set includes 100,000 ratings and 3,600 tag applications applied to 9,000 movies by 600 users, and the large set includes 27,000,000 ratings and 1,100,000 tag applications applied to 58,000 movies by 280,000 users. The large set also includes tag genome data with 14 million relevance scores across 1,100 tags. Get the data here. 6| OpinRank Dataset This data set contains full reviews for cars and hotels collected from Tripadvisor and Edmunds. The dataset contains full reviews of hotels in 10 different cities as well as full reviews of cars for model-years 2007, 2008 and 2009. In the dataset, the total number of car reviews include approximately 42,230, and the total number of hotel reviews include approximately 259,000. Get the data here. 7| SMS Spam Collection The SMS Spam Collection is a public dataset of SMS labelled messages, which have been collected for mobile phone spam research. The dataset has one collection composed by 5,574 English, real and non-encoded messages, tagged according to being legitimate or spam. The dataset is available in both plain text and ARFF format. Get the data here. 8| The Blog Authorship Corpus The Blog Authorship Corpus consists of the collected posts of 19,320 bloggers gathered from blogger.com in August 2004. The corpus incorporates a total of 681,288 posts and over 140 million words or approximately 35 posts and 7250 words per person. In this dataset, each blog is presented as a separate file, the name of which indicates a blogger id# and the blogger’s self-provided gender, age, industry and astrological sign. Get the data here. 9| WordNet WordNet is a large lexical database of English where nouns, verbs, adjectives and adverbs are grouped into sets of cognitive synonyms (synsets) and each expressing a distinct concept. In this dataset, the total number of synsets are 117 000 and each of which is linked to other synsets by means of a small number of conceptual relations. Get the data here. 10| Yelp Reviews The Yelp dataset is an all-purpose dataset for learning and is a subset of Yelp’s businesses, reviews, and user data, which can be used for personal, educational, and academic purposes. The dataset includes 6,685,900 reviews, 200,000 pictures, 192,609 businesses from 10 metropolitan areas. Get the data here.","excerpt":"One of the popular fields of research, text classification is the method of analysing textual data to gain meaningful information. According to sources, the global text analytics market is expected to post a CAGR of more than 20% during the period 2020-2024. Text classification can be used in a number of applications such as automating […]","categories":["AI Trends"],"tags":["dataset","Natural Language Processing","text analysis","text analytics","text classification","text dataset","text-based algorithm"],"author_name":"Ambika Choudhury","publish_date":"2020-03-13T13:00:00","publication_year":"2020","word_count":680,"keywords":["dataset","Go","text classification","TPU","programming_languages:R","AI","sentiment analysis","Natural Language Processing","text-based algorithm","R","Git","GAN","analytics","text analytics","text analysis","text dataset"],"extracted_tech_keywords":["AI","analytics","sentiment analysis","text classification","TPU","R","Go","Git","GAN","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/10-open-source-datasets-for-text-classification\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10098370,"title":"Oracle Unveils Oracle Compute Cloud@Customer To Enhance OCI Flexibility","content":"Oracle today introduced its latest innovation, Oracle Compute Cloud@Customer, an advanced rack-scale cloud infrastructure solution that empowers organizations to leverage Oracle Cloud Infrastructure (OCI) compute services in various environments. This technology enables businesses to develop, deploy, secure, and manage workloads utilizing the same software stack as OCI, accommodating deployments ranging from single racks to large-scale infrastructures. The significance of this development  is that it provides a unified experience regardless of where services are utilized. By launching Compute Cloud@Customer, Oracle is bridging the gap between on-premises and cloud-based services, offering a consistent user experience throughout various deployment scenarios. Organizations can tap into Oracle’s extensive resources, leveraging the OCI compute, storage, and networking services, alongside adaptable virtual machine (VM) shapes, all within their own data centers. Edward Screven, Chief Corporate Architect at Oracle, emphasized the importance of delivering a consistent experience for users, irrespective of their chosen deployment model. “Oracle provides a choice of OCI public cloud regions, Dedicated Region, and Cloud@Customer platforms that customers can combine to create a globally distributed cloud solution,” he said. Furthermore, industry expert Ron Westfall, Research Director at The Futurum Group, praised Oracle’s approach, highlighting the comprehensiveness of Compute Cloud@Customer as compared to alternative solutions. “Oracle’s latest Compute Cloud@Customer offering clearly delivers the same compute, storage, networking, APIs, control plane, and a growing list of services that are also available in OCI,” Westfall stated, reinforcing Oracle’s commitment to providing a complete cloud experience. Oracle Compute Cloud@Customer boasts impressive features, including scalability and flexibility. Starting from 552 processor cores and 150 TB of usable storage, the solution can be seamlessly expanded to accommodate organizations of varying sizes, with the capacity to scale compute and storage independently to over 6,000 processor cores and 3.4 PB of storage. Furthermore, data on Compute Cloud@Customer is encrypted, meeting stringent data residency and privacy requirements. The OCI Console allows precise control over data locality, replication, and backups.","excerpt":"Data on Compute Cloud@Customer is encrypted, meeting stringent data residency and privacy requirements","categories":["AI News"],"tags":["Oracle"],"author_name":"Siddharth Jindal","publish_date":"2023-08-09T19:30:00","publication_year":"2023","word_count":315,"keywords":["API","programming_languages:R","AI","innovation","ML","Scala","Oracle","RAG","programming_languages:Scala","GAN","R"],"extracted_tech_keywords":["AI","ML","RAG","R","Scala","API","GAN","innovation","programming_languages:R","programming_languages:Scala"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/oracle-unveils-oracle-compute-cloudcustomer-to-enhance-oci-flexibility\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10121142,"title":"Meet the Duo from Hyderabad who Made it Big at Google I\/O","content":"Google I\/O 2024 wrapped up last week, featuring a range of exciting announcements. During the latter part of the keynote, Google showcased a short video highlighting the power of Gemma in building Indic LLMs. The search giant introduced Navarasa, a Gemma 7B\/2B instruction-tuned model supporting 15 Indian languages and English, developed by Telugu LLM Labs, founded by Ravi Theja Desetty and Ramsri Goutham Golla. The model was trained on E2E Networks Limited using NVIDIA A100 GPUs which took approximately 44 hours to train the 7 billion model and 18 hours for the 2 billion model. “When a technology is developed for a particular culture, it won’t be able to solve and understand the nuances of a country like India,” said Harsh Dhand, head of APAC research partnerships at Google. Take your side projects seriously!You might as well get featured in the Google I\/O keynote on the main stage in California ;)I'm excited to share that the work @ravithejads and I did in building Navarasa, an Indic instruction-finetuned model on top of Google's Gemma catering… pic.twitter.com\/8B1I0p3iYl— Ramsri Goutham Golla (@ramsri_goutham) May 15, 2024 AIM reached out to Theja and Golla to learn more about their experience. “The Google marketing team invited us to Mysore and booked a five-star hotel, reserving almost all the rooms,” shared Golla, adding that they filmed for about two or three days. “It was an interesting experience. All those big camera rigs everywhere felt like a movie shoot, with all the big lamps and the makeup crew. They made us feel like celebrities for a day,” he said. “Google reached out to me after seeing my blog post about Navarasa,” said Theja, adding that he had no idea that the video would be shown during the keynote. “We really didn’t expect it to be showcased at the main event of Google I\/O. In fact, Ramsri slept through that day,” he chuckled. Hyderabad-based Golla studied and worked in the US for almost eight years before returning to India in 2018. He describes himself as a builder\/engineer and loves creating SaaS apps. Golla has successfully developed two AI SaaS apps, with a combined ARR of $100K. Additionally, he takes AI courses on Udemy as well as his own platform. On the other hand, Theja works as a developer advocate engineer at Llama Index. Prior to this, he served as a senior ML engineer at Glance, where he worked on recommendation systems and GenAI applications. What is Navarasa? Navarasa 2.0 is a Gemma 7B\/ 2B SFT (Supervised Fine-Tuned model) using Gemma 7B\/ 2B base models. The model’s generative capabilities have been enhanced to cover a total of 15 Indian languages. This expansion was achieved by translating the alpaca-cleaned-filtered dataset into six additional Indian languages: Marathi, Urdu, Konkani, Assamese, Nepali, and Sindhi. The 15 languages supported by Navarasa 2.0 besides English are: Hindi, Telugu, Tamil, Kannada, Malayalam, Marathi, Gujarati, Bengali, Punjabi, Odia, Urdu, Konkani, Assamese, Nepali, and Sindhi. The model is also available on IndicChat, a playground for open-source Indic LLMs. Built on Hugging Face’s HuggingChat, this platform hosts Indic AI models for users to chat and test. “Many people started using Navarasa through IndicChat, as it is hosted there with a chat interface. A considerable number of users accessed it from there,” said Golla. Speaking of the advantage Gemma has over Llama 2 and 3, Golla said, “A fundamental difference between Llama and Gemma is that with Llama, you had to fine-tune a model for each individual language. However, with Gemma, you only need to fine-tune once and can combine all the datasets created for Llama.” “On a higher level, the Gemma tokeniser includes tokens for most Indian languages, providing strong representations for these tokens. In contrast, the Llama3 tokeniser supports only a few languages, and its quality of support is not as robust,” said Theja. OpenAI’s Love for India is Unmatched OpenAI’s latest model GPT-4o has made considerable improvements when it comes to Indian languages. It supports over 50 languages and has notably optimised token usage for Indian languages, reducing Gujarati by 4.4 times, Telugu by 3.5 times, Tamil by 3.3 times, and Marathi and Hindi by 2.9 times. “One of the biggest updates GPT-4o got is a brand new tokeniser and an extended vocabulary size of 2,00,019 tokens as compared to 1,00,277 tokens in GPT-4. This means we have much better support for many new languages and yes some of them are Indian,” said Abhishek Upperwal, founder of Socket AI labs, which recently launched Pragna-1B, India’s first open-source multilingual model. Love for Indian languages!! @openai has made GPT4-o up to 4.4x more efficient for languages like Gujarati, Telugu, Tamil, Marathi, hindi, and Urdu!less tokens = cheaper cost and faster output!! #bharat https:\/\/t.co\/CpvCkjI0iA pic.twitter.com\/Xwpi8ECL41— andrew gao (@itsandrewgao) May 13, 2024 “OpenAI is paying attention to Indic languages! My mom has been struggling to learn English from Telugu – now I can make a simple wrapper using the new GPT-4o model,” wrote a user on X. “OpenAI has done well even for Indic languages in terms of tokenisation efficiency and cost etc,” said Golla. He added that in the past, he had to switch between Anthropic and OpenAI for his SaaS app development, as Anthropic was deemed more efficient with Indic languages compared to GPT-4. Theja had a different perspective, saying that during the Spring Update, OpenAI didn’t demonstrate how GPT-4o would sound in Indian languages. “If you’ve tried the audio feature in the app for Indian languages, it doesn’t sound good at all. I can’t even understand what it’s saying,” he said. Indian AI startup Sarvam AI is currently working on an Indic Voice LLM, while Hanooman and Krutrim do not include a voice modality yet. However, they are expected to add this feature soon.","excerpt":"OpenAI’s latest model GPT-4o has made considerable improvements when it comes to Indian languages.","categories":["IT Services"],"tags":["Google"],"author_name":"Siddharth Jindal","publish_date":"2024-05-21T14:53:56","publication_year":"2024","word_count":956,"keywords":["Anthropic","Hugging Face","GenAI","OpenAI","AI","GPT-4o","ML","recommendation systems","RAG","Aim","Google"],"extracted_tech_keywords":["AI","ML","GenAI","GPT-4o","OpenAI","Anthropic","Aim","Hugging Face","RAG","recommendation systems"],"url":"https:\/\/analyticsindiamag.com\/it-services\/meet-the-duo-from-hyderabad-who-made-it-big-at-google-i-o\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":50381,"title":"What Is SEBI Doing To Expand Financial Market Analytics In India?","content":"With a ₹500 crore investment in technology innovation, SEBI is working to enhance its capabilities in advanced analytics, surveillance systems, cloud infrastructure and security. “Technology is a game-changer in financial services as it can not only provide faster and better services to the consumer, it can also be a catalyst in improving the ease of doing business,” said Ajay Tyagi, Chairman, SEBI, at a recent event in Singapore. The SEBI chairman was talking at the South Asian Diaspora Convention at the National University of Singapore. Tyagi stated that the organisation’s methodology has consistently been to empower and improve technology capacity-building and simultaneously to know about the potential dangers related to such capabilities. “Due to advancement in technology, funds and securities can now flow across the world in a few microseconds. This enables setting up of complex global structures which could be potentially misused for tax avoidance, money laundering, round-tripping, etc.,“ Tyagi said. The trading watchdog is additionally setting up a Network Operations Center (NOC) for consistent checking of the smooth working of the SEBI IT system and Securities Operations Center (SOC). The system will also recognise and moderate cyberattacks on SEBI’s IT foundation productively and compellingly. Tyagi also mentioned initiatives such as setting up online complaint redress system of SEBI (SCORES), KYC registration agencies, dematerialisation of securities, T+2 rolling settlement, and the introduction of interoperability of clearing companies. “SEBI not only encourages technological innovation in securities markets, but it also uses and regularly updates technology in its own functioning and plans to spend ₹5 billion on IT projects in the next 5 years” How SEBI Is Stepping Up The AI Game SEBI is now in progress to additionally fortify its in-house analytics ability to help its market surveillance and risk management capacities. From its previous announcement, SEBI had implemented Natural Language Processing (NLP), sentiment analysis or text mining tools that collect insights from unstructured data. They have reportedly considered using voice-to-text and text-to-intelligence systems in any natural language. Apart from that, SEBI also deploys a system that makes use of statistical heuristics system which uses clustering or categorisation algorithms to categorise data even though there may be no predefined categories. SEBI has additionally arranged a data lake project to expand advanced analytics tools related to AI\/ML, deep learning, big data, which will process huge volumes of structured and unstructured data and content mining for social media surveillance. Speaking further, Tyagi stated that emerging technologies like artificial intelligence, machine learning and deep learning are being used in the Indian capital markets in things like robo-advisory services, surveillance through social media analytics and IT security. Apart from that, SEBI is also considering blockchain and RPA in processes like clearing and settlement, risk management, surveillance, compliance automation etc. Why SEBI Is Building A Social Media Surveillance System SEBI chairman also announced that the organisation working on a social media surveillance system, which will look for market manipulation information on various platforms. SEBI had found that price-sensitive details had been leaked on social media platforms like WhatsApp before a significant corporate announcement or earning call. According to SEBI, market manipulation is rampant in India, and social media analytics fed into IMSS (integrated market surveillance system) will help recognise insider trading patterns. IMSS gathers information on suspicious market activities through many sources, including its network systems at stock exchanges and depositories. There are already robust surveillance systems in place at SEBI, and with new social media, surveillance will add to the previous systems. SEBI along with stock exchanges have ventured up their live surveillance of intra-day trade in stocks where there may be suspicion of illegal activity or market maker manipulation such as spoofing that may cause increased volatility or price dumps.","excerpt":"With a ₹500 crore investment in technology innovation, SEBI is working to enhance its capabilities in advanced analytics, surveillance systems, cloud infrastructure and security.  “Technology is a game-changer in financial services as it can not only provide faster and better services to the consumer, it can also be a catalyst in improving the ease of […]","categories":["AI Features"],"tags":["bitcoin stock","Ethical Hacking","sebi","Voice Analytics"],"author_name":"Vishal Chawla","publish_date":"2019-11-20T19:00:00","publication_year":"2019","word_count":617,"keywords":["artificial intelligence","machine learning","AI","sebi","sentiment analysis","ML","Voice Analytics","RAG","NLP","bitcoin stock","deep learning","analytics","Ethical Hacking","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","NLP","analytics","RAG","sentiment analysis","R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/sebi-to-expand-financial-markets-analytics-in-india\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10092459,"title":"Amazon Web Services’ New-found Love for Open Source","content":"Unbeknownst to many, the world’s leading enterprise cloud service providers like Google Cloud, Microsoft Azure, and Amazon Web Services, all use a troupe of open-source software in their tech stacks. From Kubernetes to Linux to PostgreSQL, open-source software is ever-present in cloud services, allowing CSPs to keep the wheels running at no additional cost. To offset the cost that these providers save by using open-source software, many of the companies contribute to them. Google has been vocal about its love of open-source software offerings, and is one of the most active contributors to open-source. However, it seems that Amazon is now catching up to its customers in giving back to the community, finally righting the scales on open-source responsibility. Shift from customer-focused service Amazon is well-known for their ‘customer obsession’ strategy and has often made this their first priority when providing services. However, the public discourse against their free ride of open-source softwares has grown in the past few years. Open-source advocates have been vocal about their disdain towards AWS’ ‘strip-mining’ of open source technology. Strip-mining is a term used to denote the strategy of ‘intercept and monetise’, where Amazon takes an open-source project, ‘steals’ the code and creates a proprietary paid service based on it. A prime example of this is the Elasticsearch incident from 2015. AWS forked an open-source project created by a company named Elastic, building a product named Elasticsearch on top of it. After fighting the case in court, Elastic was forced to change their licensing from a permissive Apache License V 2.0 to a different license system called Elastic license. However, this still did not stop Amazon, as they simply forked the repo, combined it with another software called Kibana, and released the resulting product as an open-source project called “OpenSearch”. This is just one of the many cases where Amazon has taken open-source software and repurposed it for its own use. However, it seems that this strategy is now changing, as AWS has climbed to the top 5 in open-source contributions over the last year. While Google and Microsoft dominate the leaderboards in first and second respectively, the third and fourth position is taken up by Red Hat and Intel, both open-source giants in their own right. Considering Amazon’s sluggish attitude towards supporting open-source, it seems they shouldn’t be in the top 5, yet they are. AWS open-source contributors’ growth. Source: OSCI When looking at the data provided by the Open Source Contributors Index, it is clear when AWS pivoted to their open-source contribution strategy. Up to 2019, the number of open source contributors consistently stayed below 250, but spiked to 852 in March 2021. Today, it has 14.9% more contributors than it did in 2021 – around 2700 – and is the only company that has seen a positive growth in contributors in this time period. Top 4 open source contributors’ decline.  Source: OSCI Microsoft, Google, Red Hat and Intel have all cut down on the number of open-source contributors over the past few years. So the question remains – what drove Amazon to contribute so heavily to open source? Pivot to open-source-as-a-service While open-source projects offer ease of access and no licensing costs, it often costs significant resources to keep them running in an enterprise tech stack. This is due to the fact that there is no ever-present support to take care of obscure bugs and issues that might arise. Some companies, like Red Hat, saw this as a market in need of a business model, and began offering open-source-as-a-service. In this business model, the company takes over the support and documentation responsibilities while offering open-source software. It now seems that AWS has also seen the benefits of this business model, if not for profitability, for customer focus. Vertically integrating open-source software into their tech stack offers multiple advantages, the primary among them being ease of use. By contributing to and maintaining open source repositories, Amazon does not need to rely on community contributors to fix bugs, nor do they need to package the software into a new web service. Matt Asay, vice president of developer relations at MongoDB, who has had prior experience working at AWS, stated, “The company has always been great at running open source projects as services for its customers. As I found while working there, most customers just want something that works. But getting it to “just work” in the way customers want requires that AWS get its hands dirty in the development of the project.” Indeed, many of the company’s new offerings leverage the power of open-source in a sustainable manner. One only needs to look at the new AWS Bedrock services – an effort to bring foundational models into the hands of AWS customers. Even some of the models offered on the Bedrock service are open source, such as Stability AI’s StableLM and Stable Diffusion. Whether it is for the sake of customer satisfaction or removing technical debt, Amazon is now catching up to its competitors in terms of giving back to the community. Moreover, this new strategy allows for the creation of a new business model for AWS, moving past wrapping open-source projects in proprietary skins.","excerpt":"AWS is now one of the top 5 open source contributors. What’s the reason?","categories":["Deep Tech"],"tags":["Open Source AI"],"author_name":"Anirudh VK","publish_date":"2023-04-26T17:09:47","publication_year":"2023","word_count":862,"keywords":["PostgreSQL","Elasticsearch","AWS","AI","MongoDB","R","RAG","Open Source AI","SQL","Azure","kubernetes"],"extracted_tech_keywords":["AI","RAG","AWS","Azure","kubernetes","Elasticsearch","MongoDB","PostgreSQL","R","SQL"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/amazons-new-found-love-for-open-source\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10140558,"title":"Microsoft Copilot Joins an Indian Acapella Group as its Newest Member","content":"At the Microsoft Building AI Companions for India event in Bengaluru on 6th November, RaagaTrippin’, India’s premier bollywood all-vocal acapella band, showcased a unique performance using the power of Microsoft Copilot. The band took some keywords from the audience, which included songs from the 1980s and rock, to generate new lyrics of Queen’s We Will Rock You song. The lyrics were generated in the same meter and rhythm as the original song by Queen. The band performed the song live on stage with the generated lyrics and the audience enjoyed it extensively. The band announced on stage that Copilot is now their sixth band member. Mustafa Suleyman, CEO of Microsoft AI, was also part of the event, which was his first visit to India. He highlighted throughout his session how Copilot is becoming a great companion, and in many cases a family member for a lot of people. RaagaTrippin’ is a Mumbai-based Indian a cappella group made up of six vocalists. Established in May 2012, they primarily create covers of well-known Bollywood songs and Western pop music, often blending them into medleys. Microsoft Building AI Companions for India event was held for Microsoft to showcase the prowess of its AI Copilot, which is now available in India, including a chatbot on WhatsApp. Microsoft recently launched an updated version of Copilot, enabling users to interact with Copilots in real-time, shifting human-computer interaction to a new level. Another feature, Copilot Daily, introduced in October in select countries, provides users with a curated news experience to start their day. Last year, the Bengaluru-based Indian-Folk Rock band, Swarathma, was live at Cypher 2023. At the biggest AI conference, the band decided to generate lyrics of a song on the spot with ChatGPT, and started singing it, enchanting the audience.","excerpt":"The band took some keywords from the audience, which included songs from the 1980s and rock, to generate new lyrics of Queen’s We Will Rock You song.","categories":["AI News"],"tags":["Microsoft","Microsoft Copilot"],"author_name":"Mohit Pandey","publish_date":"2024-11-07T16:51:03","publication_year":"2024","word_count":295,"keywords":["ChatGPT","programming_languages:R","AI","GPT","Microsoft Copilot","copilots","R","llm_models:ChatGPT","Microsoft","llm_models:GPT"],"extracted_tech_keywords":["AI","ChatGPT","copilots","R","GPT","llm_models:GPT","llm_models:ChatGPT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/microsoft-copilot-joins-an-indian-acapella-group-as-its-newest-member\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10166281,"title":"Apptronik Secures $53M in Extended Series A, Backed by Mercedes-Benz","content":"Apptronik, an AI-powered humanoid robotics startup based in the US, announced on Tuesday that it has raised an additional $53 million in its Series A round. This brings Apptronik’s total Series A funding to $403 million, adding to the previously announced $350 million, led by B Capital and Capital Factory, with participation from Google. New investors include Mercedes-Benz, Japan Post Capital, ARK Invest, Helium-3, Magnetar, RyderVentures, and a syndicate led by Korea Investment Partners. The funding will support the production and deployment of Apptronik’s AI-powered humanoid robot, Apollo, across industries such as automotive, electronics manufacturing, logistics, and consumer packaged goods. Joerg Burzer, member of the board of management at Mercedes-Benz Group AG said, “Our work with Apptronik has given us a front-row seat to the incredible pace of progress in humanoid robotics and AI, and the transformative potential these technologies hold for modern manufacturing.” Mercedes-Benz is using Apollo robots in its manufacturing process by employing a method, ‘teleoperation’, where the robots are initially controlled remotely by humans to learn specific tasks. Currently, the robots are being tested at Mercedes-Benz’s Digital Factory Campus in Marienfelde, Berlin, and in Kecskemet, Hungary, with plans to expand their deployment to additional sites, as announced by Burzer during a recent roundtable discussion in Berlin. Apptronik aims to create a future where humanoid robots collaborate seamlessly with humans, initially in logistics and manufacturing, with future applications in healthcare and elder care. This investment follows a year of record growth for Apptronik, including the strategic partnership with Google DeepMind’s robotics team. This month, Google Deepmind also introduced two new AI models, Gemini Robotics and Gemini Robotics-ER, which have been designed to enhance robotic capabilities in the physical world. For this, the giant had expressed partnering further with Apptronik to develop the next generation of humanoid robots using these models. This will open the giant to more partnerships in the future, like those with testers including Agile Robots, Agility Robotics, Boston Dynamics and Enchanted Tools. In February this year, Apptronik announced a partnership with Jabil, a global leader in engineering and manufacturing, to build and integrate Apollo humanoid robots into Jabil’s manufacturing operations. This partnership leveraged Apollo robots on the production lines that create them, marking progress towards robots constructing robots. The company had announced this in a post on X, saying it is “paving the way for Apollo to build Apollo.”","excerpt":"This totals the startup’s funding to $403 million in Series A.","categories":["AI News"],"tags":["apptronik","Funding","Mercedes-Benz"],"author_name":"Sanjana Gupta","publish_date":"2025-03-19T01:18:38","publication_year":"2025","word_count":394,"keywords":["Go","API","Funding","funding","AI","apptronik","ML","Mercedes-Benz","Git","RAG","Aim","R","startup"],"extracted_tech_keywords":["AI","ML","Aim","RAG","R","Go","Git","API","startup","funding"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/apptronik-secures-53m-in-extended-series-a-backed-by-mercedes-benz\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10113430,"title":"Nimo Planet’s Spatial Computer Brings in a Future Sans Laptops","content":"Kochi-based Nimo Planet created quite a buzz at CES 2024 with its spatial computer which features Nimo 1 OS, Nimo 1 Glasses, and Nimo 1 Core. The company aspires to create a pocket-sized portable computer dedicated to boosting employee productivity. “Apple’s Vision Pro can provide an immersive experience to users for one or two hours inside the home, but it is not designed for the workspace where you can carry it in your pocket for multiple hours at work. That’s where Nimo Planet steps in,” said Rohildev Nattukallingal, Nimo Planet’s chief in an exclusive interaction with AIM. Lately, there have been reports about customers returning their Apple Vision Pros under the 14-day return policy, citing discomfort and the poor price-to-features ratio. Nimo ♥️ Work “Some customers are adopting a hybrid workspace, working a few days from home and a few days from the office. Such people prefer to use Nimo,” said Nattukallingal, adding that the initial category of customers comprised traveling professionals. The company has shipped its products to companies like KPMG Consulting and a few other consulting firms. Additionally, it has some enterprise use cases in companies from the Middle East, US, and Europe, who want to use it to better their employee experience. As of today, the company has delivered Nimo OS to 15 enterprises. Comparing Nimo Planet with Quest 3, Nattukallingal said, “With their VR headsets, Meta is focusing more on the metaverse and games. The way we see it, our focus is on increasing work productivity for enterprises and we don’t concentrate on general use cases.” He added that even with AI integration, the company aims to find ways to build operating systems and applications specifically for employees. In the near future, the company aims to replace traditional input devices like keyboard and mouse. However, Nattukallingal believes that hand gestures are not a suitable alternative for keyboards, as typing in the air proves to be challenging. “Right now, the keyboard is the best option. But in the future, I see voice plus AI that does some automation as a potent option. However, voice does not work in some scenarios, such as when I’m sitting in a flight,” said Dev, adding that he expects advancements like Neuralink, allowing mouse control with thoughts, to emerge. Nattukallingal anticipates a future without laptops. “I would say, in the next five to eight years, let’s imagine people start forgetting to take laptops because if they have something like this, they can just carry it in their hands or pockets.” What Nimo OS Offers Nimo OS allows users to run multiple applications simultaneously, each displayed on a separate virtual screen in the 3D space. This makes it easy to switch between tasks and compare information side-by-side. It is built on the foundation of Android Open Source Project (AOSP) and Linux . “We built a custom rendering system and have not used any wrapper products in the market. This means I can run stereo 3D videos, create apps, and convert any existing app into stereoscopic,” he said. “One of the biggest advantages of this architecture is that I can run multiple Android, Unity, and web apps simultaneously,” he added, saying that the only other company to have adopted this approach is Apple. However, Nimo Glasses currently lack a camera for video-calling. Nattukallingal said that they are planning to introduce a new case equipped with a battery and a camera that can be connected for video calling. Nimo OS also supports Mac and Windows applications through USB-C or remote desktop. It has been granted a US utility patent for Nimo OS Spatial Workspace and Multi-Window Architecture. Nimo’s Roadmap “The immediate goal is to ship 3,000 to 5,000 units of Nimo OS to early customers in the next six to eight months. We are fundraising to help us with mass production,” said Nattukallingal, saying that it plans to partner with enterprises and companies manufacturing AR glasses. “Plus, we are starting a developer programme in which people can start developing custom applications,” he added. The company recently partnered with Rokid Global, a US-based AR glass manufacturer. Nattukallingal said that there are plans currently in the pipeline to partner with three more companies, and further announcements will be made soon.","excerpt":"The company has shipped its products to companies like KPMG Consulting and a few other consulting firms.","categories":["AI Features"],"tags":[],"author_name":"Siddharth Jindal","publish_date":"2024-02-21T13:00:46","publication_year":"2024","word_count":703,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","automation","Aim","ViT","R"],"extracted_tech_keywords":["AI","Aim","R","Go","ViT","automation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/nimo-planets-spatial-computer-brings-in-a-future-without-laptops\/","complexity_score":4,"technical_depth":8,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10052869,"title":"IBM Partners With NeuReality To Develop AI Inference Platforms","content":"IBM and NeuReality, an Israeli AI systems and semiconductor company, recently announced that they would develop high-performance AI inference platforms designed to deliver cost and power consumption improvements for deep learning. The platforms will be built for such sectors as finance, insurance, healthcare, manufacturing, and smart cities to deploy computer vision, natural language processing, recommendation systems and other AI use cases. The agreement involves NR1, NeuReality’s first Server-on-a-Chip ASIC implementation of their AI-centric architecture. NR1 is based on NeuReality’s first-generation FPGA-based NR1-P prototype platform that was introduced earlier this year. The NR1 will be a new type of integrated circuit device with native AI-over-Fabric networking, full AI pipeline offload and hardware-based AI hypervisor capabilities that, taken together, remove system bottlenecks and improve cost and power consumption, according to the companies. “We believe our collaboration (with IBM) is a vote of confidence for our AI-centric technology and architecture and in its potential to power real-life AI use cases with unprecedented deep learning capabilities. Having the NR1-P FPGA platform available today allows us to develop IBM’s requirements and test them before the NR1 Server-on-a-Chip tape out. Being able to develop, test and optimize complex datacenter distributed features, such as Kubernetes, networking and security before production is the only way to deliver high quality to our customers,” said Moshe Tanach, CEO and Co-Founder of NeuReality. Dr Mukesh Khare, Vice President of Hybrid Cloud research at IBM Research, said, “In light of IBM’s vision to deliver the most advanced hybrid cloud and AI systems and services to our clients, teaming up with NeuReality, which brings a disruptive AI-centric approach to the table, is the type of industry collaboration we are looking for. The partnership with NeuReality is expected to drive a more streamlined and accessible AI infrastructure, which has the potential to enhance people’s lives.” The NR1-P platform will support software integration and system-level validation prior to the availability of the NR1 production platform next year. The partnership also marks NeuReality as the first start-up semiconductor product member of the IBM Research AI Hardware Center and licensee of the centre’s low-precision, high-performance Digital AI Cores. As part of the agreement, IBM became a design partner of NeuReality and will work on the product requirements for the NR1 chip, system, and SDK, which will be implemented in the next revision of the architecture. Together the two companies will evaluate NeuReality’s products for use in IBM’s Hybrid Cloud, including AI use cases, system flows, virtualization, networking and security.","excerpt":"Together the two companies will evaluate NeuReality’s products for use in IBM’s Hybrid Cloud, including AI use cases, system flows, virtualization, networking and security.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","chip","Chip Manufacturing","chipset","data centres","Data Science","Data Scientist","Deep Learning","Deep Neural Networks","IBM","ibm research","Machine Learning"],"author_name":"Victor Dey","publish_date":"2021-11-03T19:08:44","publication_year":"2021","word_count":412,"keywords":["chip","Git","computer vision","deep learning","data centres","R","kubernetes","ibm research","Data Science","Chip Manufacturing","AI","ML","Machine Learning","recommendation systems","IBM","chipset","ai_applications:computer vision","programming_languages:R","Deep Neural Networks","Deep Learning","Data Scientist","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","ML","deep learning","computer vision","recommendation systems","kubernetes","R","Git","programming_languages:R","ai_applications:computer vision"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/ibm-partners-with-neureality-to-develop-ai-inference-platforms\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10104330,"title":"AMD ROCm 6 is Here, OpenAI Adds Support","content":"AMD at the Advancing AI event has showcased significant advancements in its software part of its Instinct data centre accelerators, emphasising an open, proven, and ready AI software platform for the market. The company introduced the latest iteration of its parallel computing framework, ROCm 6, optimised for a comprehensive software stack for AMD Instinct, particularly catering to large language models in generative AI. Key features of ROCm 6, which is an alternative to NVIDIA’s CUDA, include support for new data types, advanced graph and kernel optimizations, optimised libraries, and state-of-the-art attention algorithms. Notably, the performance boost is remarkable, with an approximately 8x increase in overall latency for text generation compared to ROCm 5 running on the MI250. In a collaborative effort with three emerging AI startups – Databricks, Essential AI, and Lamini – AMD showcased how they leverage the AMD Instinct M1300X accelerators and the open ROCm 6 software stack to deliver differentiated AI solutions for enterprise customers. Read: AMD Focuses on Software Ahead of MI300X Release Furthermore, OpenAI is aligning with AMD by adding support for AMD Instinct accelerators to Triton 3.0. This move provides out-of-the-box support for AMD accelerators, enabling developers to work at a higher level of abstraction on AMD hardware within the growing ecosystem. The ROCm Software and Ecosystem Partners announcement underscores AMD’s commitment to contributing to the open-source community. ROCm 6 represents a significant step forward, offering open-art libraries and supporting various key features for generative AI, including FlashAttention, HIPGraph, and vLLM. Read: AMD’s Attempt to Break NVIDIA’s CUDA This positions AMD uniquely to leverage widely used AI software models, algorithms, and frameworks, fostering innovation and simplifying deployment for enterprises. In addition to the software advancements, AMD continues its strategic investments in software through Mipsology, acquisitions like Nod.AI, and partnerships such as Lamini, which enables LLM training on AMD ROCm with zero code changes. The unveiled developments mark a pivotal moment in the AI landscape, with AMD’s robust software platform poised to unlock the true potential of generative AI, driving the industry forward into a new era of innovation.","excerpt":"OpenAI is aligning with AMD by adding support for AMD Instinct accelerators to Triton 3.0.","categories":["AI News"],"tags":["AMD","amd ai event","amd generative ai","amd mi300x","amd rocm"],"author_name":"Mohit Pandey","publish_date":"2023-12-07T01:32:00","publication_year":"2023","word_count":344,"keywords":["amd rocm","CUDA","AMD","Go","amd mi300x","OpenAI","AI","innovation","amd ai event","amd generative ai","RAG","generative AI","R","Databricks"],"extracted_tech_keywords":["AI","generative AI","OpenAI","RAG","CUDA","Databricks","R","Go","CUDA","innovation"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/amd-rocm-6-is-here-openai-adds-support\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":29280,"title":"Google Pixel 3 May Not Look Too Different, But It Is Powered By AI And ML Like Never Before","content":"Google’s latest I\/O had no surprises but a slew of new product announcements. The two new Pixel 3 models, the Pixel Slate convertible laptop, and the all-encompassing Google Home Hub (a stiff competitor to Amazon Echo) came outfitted with a new set of capabilities.  And all of Google’s machine learning and artificial intelligence updates are revolving around the new product releases. In fact, the new Pixel 3 boasts of new enhancements such as extra lenses to charge ahead of the competition. One thing is for sure, Google is trying to use ML to improve photography. So instead of being hardware-dependent, the Pixel 3 is outfitted with Night Shot, which uses ML to dramatically enhance photography in low-light areas. This feature will be introduced in the Pixel 3 and older Pixel models will also get this feature with the update. There’s also the Photobooth Mode, which uses will use AI to prompt the shutter on cues like a moving target, a smile, and the ability to change the amount of background “bokeh” blur. While the earlier I\/O was all about Gmail that writes itself to a new Google Assistant and Google Home, in the latest event, the Mountain View giant announced a new set hardware enhancements in its smartphones. However, going beyond the hardware announcements, there are a bunch of AI-related updates to its line of smartphones. New Google Pixel = AI-Powered Software Smart Compose: Already a feature in Gmail, Smart Compose is an iteration of Smart Reply which helps in finishing the sentences on emails. This feature will soon be rolled in the latest Google Pixel 3 smartphones which would allow users to perform a left-to-right swipe or populate a Smart Compose suggestion. In a few months, this new feature will also support other languages such as Spanish, French, Italian and Portuguese. Google Lens: Google’s widely acclaimed Google Lens which uses advanced computer vision and natural language processing technology which is used in other Google software like Assistant and in smartphone cameras is one of the best examples of consumer-facing AI application which can be effectively used to identify street signs, restaurant menus and objects in photos. Now paired up with Deep Learning, this visual search feature is set to become more automated, a news report indicated and it rules out manual activation. It is also becoming more accessible and can be launched directly with a long-press action on the camera app. Lens is the best AI-powered tool which can add context to objects in the environment and a cutting-edge consumer facing AI application which allows users to see and sense their surrounding effectively. Google Duplex: Duplex is one of the most talked about features in the Pixel phone which would allow users to call restaurants without the need for online booking systems from Google Assistant. Duplex is another consumer-facing example of natural conversations through Google Assistant. For natural conversations, Duplex has been trained in a set of domains to carry out general conversations. Essentially, Duplex is a recurrent neural network (RNN) and is built using TensorFlowxt Extended (TFX). For greater accuracy, Duplex’s RNN was trained on a dataset of anonymised phone conversation data. To better understand the parameters of conversation, Duplex also leverages Google’s automatic speech recognition (ASR) technology and uses hyperparameter optimisation from TFX to further improve the model. Call Screen: Another outstanding feature debuted in the recent I\/O event is Call Screen through which users can pick up an incoming call, by tapping the “Screen call” button which will enable the Google Assistant to receive the call. This is positioned as a Screening service from Google and offers suggestion cards\/chips such as Who is this or I’ll call you back to allow Assistant to reply back effectively. The Call Screen feature is completely powered by on-device machine learning and launching on the Pixel 3, before coming to older Pixel devices.","excerpt":"Google’s latest I\/O had no surprises but a slew of new product announcements. The two new Pixel 3 models, the Pixel Slate convertible laptop, and the all-encompassing Google Home Hub (a stiff competitor to Amazon Echo) came outfitted with a new set of capabilities.  And all of Google’s machine learning and artificial intelligence updates are […]","categories":["Global Tech"],"tags":["Google Pixel"],"author_name":"Richa Bhatia","publish_date":"2018-10-16T04:56:11","publication_year":"2018","word_count":643,"keywords":["artificial intelligence","machine learning","AI","neural network","ML","computer vision","RAG","Aim","Google Pixel","deep learning","TensorFlow"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","neural network","computer vision","Aim","TensorFlow","RAG"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/google-pixel-3-may-not-look-too-different-but-it-is-powered-by-ai-and-ml-like-never-before\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10160922,"title":"GCC Enablers Pave the Way for India&#8217;s $50 Billion Economic Output Goal","content":"India introduced its first-ever GCC policy at the 2024 Bengaluru Tech Summit with an aim to bring in 1,000 GCCs, generate 3.5 lakh new jobs, and hit $50 billion in economic output by 2029. Karnataka alone already hosts more than 875 GCCs – around 30% of India’s total. The state employs 6 lakh people and contributes $22.2 billion annually. According to an AIM Research report, third-party companies or GCC enablers play a crucial role in the successful establishment and operation of GCCs. Their expertise simplify the complex process of setting up a GCC, from selecting the right location to navigating regulatory frameworks, ensuring infrastructure readiness, and recruiting talent. These firms provide specialised knowledge and local insights, which help multinational corporations minimise risks, reduce setup time, and lower operational costs. Moreover, third-party facilitators bring valuable industry experience and networks, thereby facilitating a smoother integration of the GCC into the global operations of the parent company. By using their expertise, companies can focus on their core business while ensuring their GCC runs efficiently. According to AIM Research, top PeMa (penetration and maturity) Quadrant GCC service providers in India include ANSR, EY, HCL, Deloitte and KPMG. “As of today, GCCs are close to around $34 billion in terms of net economic value…We expect that they have the potential to grow to almost $75 to $80 billion by 2025. More importantly, the economic value that they add to India as a country is far higher – almost three times the direct revenue that comes from the GCCs,” Gaurav Gupta, GCC industry lead of Deloitte India, said in a podcast with ET Now. Highlighting some policy-related hurdles, he further said, “One is from a policy perspective. While we have had some incentives that were given to the GCC sector almost 10 years ago, over a period of time…they’ve gone away. To try and attract more GCCs to India, we need to look at what other incentives can be provided.” Talent remains a big focus “India has been a great talent story, but the talent needs of GCCs are evolving. We need to see what new skill sets are required and ensure we bridge the demand-supply gap there,” Gupta said. GCC enablers have stepped in to fill these gaps. Deloitte India helps GCCs become innovation centres by offering consulting on digital transformation, nurturing start-up collaborations, and focusing on futuristic tech. According to Deloitte, focusing on innovation and advanced technology is becoming a necessity for every business. Meanwhile, when it comes to start-ups, accessing global markets via GCCs using their domain expertise and brand reputation, and accessing the physical and virtual infrastructure of GCCs are among the many perks of such partnerships. Management consulting firm Zinnov, established in 2002, has been recognised for helping Fortune 500 companies with market entry, product engineering, and scaling operations. Nilesh Thakker, president of Zinnov, and Karthik Padmanabhan, managing partner of GCCs in Zinnov, stressed, “True ownership in tech means owning the whole process – from strategy to design, building, and updating. It’s a call for GCCs to step up, create more value, and lead in global innovation.” The firm’s expertise spans automation, AI, and technology consulting. It offers services such as GCC setup and roadmap design, talent strategy, digital transformation, operational optimisation, location analysis, and governance. Zinnov also provides specialised solutions like the GCC Blueprint Model for launching and expanding global centres, its Talent 360 Platform for workforce planning, and ongoing Zinnov Zones reports for comprehensive service provider evaluations. EY, on the other hand, was recognised as the leader for GCC\/GIC Setup Capabilities in India in Everest Group’s PEAK Matrix® Assessment 2024. “With a team of 10,000 practitioners, including former GCC CXOs, EY is supporting international organisations set up and evolve their GCCs\/GICs by transforming them into innovation-driven centres of strategic value,” Kunal Ghatak, global business services partner at EY India, said. Meanwhile, ANSR, a global leader in establishing and operating GCCs for enterprises, stated, “ANSR is redefining India’s role on the global stage. By enabling over 135 GCCs and driving $2 billion in investments, ANSR is at the forefront of fostering innovation, nurturing talent, and catalysing growth for global enterprises.” Digital services company and GCC tech enabler GlobalLogic has also noticed a major shift. “A lot of companies in the last couple of years have started empowering the GCCs,” Piyush Jha managing director of GlobalLogic India and APAC said in an interview with AIM. “Since we have specialised ourselves in the product development space for the last 22-23 years, it becomes very natural for us to start getting into synergistic motions with these GCCs.” Emerging Nano GCC Space Suchita Vishnoi, co-founder and CMO at gatewAI, explained to AIM that starting in 2024, the company – a GCC enabler – decided to “change the way GCCs are coming to India. She added that gatewAI shifted the focus from old-school IT outsourcing to an “AI era” approach. Instead of competing solely on cost, gatewAI aims to offer “GCC as a service with AI-enabled service offerings” and zeroes in on “nano GCCs”, typically teams with five to 50 employees. “We say we are born in the AI era. We are not born in the IT outsourcing era…which makes us very nimble, very agile, very new age in the way we are adaptive to the requirements that any GCC brings,” Vishnoi further said. According to her, NASSCOM has predicted the number of GCCs to reach 2,200 by 2030. “That’s a huge opportunity for us…We have a strong five-year window to capitalise and help a lot of the nano GCCs, the small and medium-sized businesses (SMBs) that want to set up innovation hubs here in India.” In short, it’s a new age for GCCs in India. Companies want more than just cheap labour; they’re looking for a place to build products, develop cutting-edge tech, and tap into world-class talent. “Any company that’s coming to India for the first time will say, ‘I need groundwork research done so that I know that the foundation I’m setting is for success’,” Vishnoi concluded.For policymakers, enablers, and GCCs themselves, the future is about collaboration, creative solutions, and continuous innovation. After all, the path to hitting that ambitious $50 billion mark, and beyond, will rely on how quickly they can adapt to changing tech trends and talent needs, one GCC at a time.","excerpt":"It’s a new age for GCCs in India. Companies want more than just cheap labour; they’re looking for a place to build products, develop cutting-edge tech, and tap into world-class talent.","categories":["GCC"],"tags":["Deloitte","EY","GCC india"],"author_name":"Shalini Mondal","publish_date":"2025-01-08T09:44:21","publication_year":"2025","word_count":1053,"keywords":["Go","API","TPU","Deloitte","AI","digital transformation","Git","automation","Aim","EY","GCC india","GAN","R"],"extracted_tech_keywords":["AI","Aim","TPU","R","Go","Git","API","GAN","digital transformation","automation"],"url":"https:\/\/analyticsindiamag.com\/gcc\/gcc-enablers-pave-the-way-for-indias-50-billion-economic-output-goal\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10011012,"title":"What Are The Scope and Challenges of Using AI in Military Operations","content":"Artificial intelligence has penetrated almost all civilian industries that one can think of. It has transformed the way individuals and businesses work, and now it is quickly making its way in becoming a critical part of modern warfare. The strength of its army is one of the factors indicating how powerful the country is. In some of the most developed nations, investment in this sector is the highest as compared to other sectors. A major part of this investment goes towards rigorous research and development in modern technology such as AI in military applications. AI-equipped military systems are capable of handling volumes of data efficiently. In addition to that, such systems have improved self-control, self-regulation, and self-actuation due to its superior computing and decision-making capabilities. Use Cases of AI in Defence Training: Training and simulation are multidisciplinary fields which utilise system and software engineering principles to construct models that can help soldiers train on various combat systems deployed during the actual military operations. The US Navy and Army have already initiated several sensor simulation programs. Further, augmented and virtual reality techniques can be used to create effective, realistic, and dynamic simulations for training purposes. The reinforcement techniques enhance combat training for both virtual agents and human soldiers. Surveillance: AI with geospatial analysis can help in the extraction of valuable intelligence from connected pieces of equipment such as radars and automatic identification systems. This information can help in detection of any illegal or suspicious activities and alert the concerned authority. Robots with AI and computer vision with IoT can also help in target identification and classification. Arms and Ammunition: New-age weaponry now comes with AI-embedded technology. For example, sophisticated missiles have the capability to determine and analyse the target range for kill zones without any human intervention. Cybersecurity: In defence circles, cyberspace is now being considered as the third war-front after land, sea, and air. A compromised and malicious network can severely compromise the security of the whole region. Defence establishments are using machine learning to predict and protect from unauthorised intrusions. This intrusion detection is usually done by classifying the network as normal or intrusive. AI-based techniques help in increasing the accuracy of such classification. Logistics: One of the most important factors that play a role in determining the success of a military operation is logistics. Integration of machine learning and geospatial analysis with the military’s logistical systems reduce the amount of effort, time, and error. Use of AI in Militaries Around World Some of the frontrunners in deploying new-age technology to further strengthen their forces are — the USA, China, and Russia. The Defence Department of the USA released its first AI strategy in 2019 which called for increasing use of AI systems throughout the military, right from decision-making to problem prediction through investment and partnership with private establishments in AI research. Some of the more popular AI-based projects that are already deployed in the US military are –Project Maven, Defense Advanced Research Projects Agency’s (DARPA) Squad X Experimentation Program, and the OFFSET program. China’s leadership has also been pushing for more AI and machine learning-based innovation in military technology. President XI Jinping has called for treading the path of ‘military-civil fusion-style innovation’; this has also found its place in the national strategy of China. Some of the projects in direction are — ‘Military-Civil Fusion National Defense Peak Technologies Laboratory’ launched by Tsinghua University and the development of the Blowfish A2 model in collaboration with a company named Ziyan UAV. According to the company, this Blowfish A2 model can autonomously perform complex combat missions such as fixed-point timing detection and target precision strikes. Notably, in 2017, China released its State Council AI Plan, which outlines the country’s ambitious project which aims to create an AI-industry worth 150 billion RMB. Russia’s military prowess is famed across the globe. The country operates its own equivalent of the US’ DARPA called Foundations for Advanced Research Projects. On the AI for the military front, top officials have been seen endorsing developing AI-based technology. Reportedly, the Russian defence forces are looking at capitalising on AI, big data, and machine learning to conduct more impactful information operations. Currently, Russia invests heavily in AI for detection, analysis, and debunking misinformation, both in the commercial sector and defence. India, which has one of the largest armies in the world, is also taking tiny steps deploying AI-based innovation in its combat and surveillance projects. In fact, in 2019, the Ministry of Defence established a high-level Defence AI Council (DAIC), which is assigned with the task of providing strategic direction to adopt AI in defence. DAIC’s primary function includes guiding partnership between government and the industry for deployment of such innovations. Challenges Ahead One of the major concerns with using AI-based technology is an investment in terms of both money and skills. Especially for middle-income countries like India, where a lot of the population is still living under the poverty line, how much can we afford to provide infrastructure capabilities for building such technology is a big question. The possible solution is that policymakers must decide on what AI programme is necessary for the security of the nation and work towards them. The use of AI in defence also presents an ethical dilemma. Experts and organisations around the world have raised such technology unintentionally escalating the tensions between countries. One of the arguments is that if an AI system fails to perform as intended may result in catastrophic implications. In fact, several human and civil rights groups call for an absolute ban on autonomous devices in defence, especially weaponry.","excerpt":"Artificial intelligence has penetrated almost all civilian industries that one can think of. It has transformed the way individuals and businesses work, and now it is quickly making its way in becoming a critical part of modern warfare.  The strength of its army is one of the factors indicating how powerful the country is. In […]","categories":["AI Features"],"tags":["ai iot and future"],"author_name":"Shraddha Goled","publish_date":"2020-11-01T10:00:00","publication_year":"2020","word_count":930,"keywords":["big data","Go","API","artificial intelligence","machine learning","AI","ai iot and future","Scala","computer vision","Aim","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","computer vision","Aim","R","Go","Scala","API","big data"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/what-are-the-scope-and-challenges-of-using-ai-in-military-operations\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":38610,"title":"5 Latest Research Papers On ML You Must Read In 2019","content":"Over the last few years, India has emerged as among the top countries in Asia to contribute a number of research work in the field of AI, machine learning and Natural Language Processing. In this article, we take a look at the top five recent research paper submission by Indian researchers in Academia.edu. 1.Cricket Analytics and Predictor Authors: Suyash Mahajan,  Salma Shaikh, Jash Vora, Gunjan Kandhari,  Rutuja Pawar, Institute: Walchand Institute of Technology, Solapur, Abstract:  The paper embark on predicting the outcomes of Indian Premier League (IPL) cricket match using a supervised learning approach from a team composition perspective. The study suggests that the relative team strength between the competing teams forms a distinctive feature for predicting the winner. Modeling the team strength boils down to modeling individual player‘s batting and bowling performances, forming the basis of our approach. Research Methodology: In this paper, two methodologies have been used. MySQL database is used for storing data whereas Java for the GUI. The algorithm used is Clustering Algorithm for prediction. The steps followed are as Begin with a decision on the value of k being the number of clusters. Put any initial partition that classifies the data into k clusters. Take every sample in the sequence; compute its distance from centroid of each of the clusters. If sample is not in the cluster with the closest centroid currently, switch this sample to that cluster and update the centroid of the cluster accepting the new sample and the cluster losing the sample. 2.Real Time Sleep \/ Drowsiness Detection – Project Report Author: Roshan Tavhare Institute: University of Mumbai Abstract: The main idea behind this project is to develop a nonintrusive system which can detect fatigue of any human and can issue a timely warning. Drivers who do not take regular breaks when driving long distances run a high risk of becoming drowsy a state which they often fail to recognize early enough. Research Methodology: A training set of labeled facial landmarks on an image. These images are manually labeled, specifying specific (x, y) -coordinates of regions surrounding each facial structure. Priors, more specifically, the probability on distance between pairs of input pixels. The pre-trained facial landmark detector inside the dlib library is used to estimate the location of 68 (x, y)-coordinates that map to facial structures on the face. 3. A Study of Various Text Augmentation Techniques for Relation Classification in Free Text Authors: Chinmaya Mishra Praveen Kumar and Reddy Kumar Moda,  Syed Saqib Bukhari and Andreas Dengel Institute: German Research Center for Artificial Intelligence (DFKI), Kaiserslautern, Germany Abstract: In this paper, the researchers explore various text data augmentation techniques in text space and word embedding space. They studied the effect of various augmented datasets on the efficiency of different deep learning models for relation classification in text. Research Methodology: The researchers implemented five text data augmentation techniques (Similar word, synonyms, interpolation, extrapolation and random noise method)  and explored the ways in which we could preserve the grammatical and the contextual structures of the sentences while generating new sentences automatically using data augmentation techniques. Smart Health Monitoring and Management Using Internet of Things, Artificial Intelligence with Cloud Based Processing Author: Prateek Kaushik Institute: G D Goenka University, Gurugram Abstract: This research paper described a personalised smart health monitoring device using wireless sensors and the latest technology. Research Methodology: Machine learning and Deep Learning techniques are discussed which works as a catalyst to improve the  performance of any health monitor system such supervised machine learning algorithms, unsupervised machine learning algorithms, auto-encoder, convolutional neural network and restricted boltzmann machine. 5.Internet of Things with BIG DATA Analytics -A Survey Author: A.Pavithra,  C.Anandhakumar and V.Nithin Meenashisundharam Institute: Sree Saraswathi Thyagaraja College, Abstract: This article we discuss about Big data on IoT and how it is interrelated to each other along with the necessity of implementing Big data with IoT and its benefits, job market Research Methodology: Machine learning, Deep Learning, and Artificial Intelligence are key technologies that are used to provide value-added applications along with IoT and big data in addition to being used in a stand-alone mod","excerpt":"Over the last few years, India has emerged as among the top countries in Asia to contribute a number of research work in the field of AI, machine learning and Natural Language Processing. In this article, we take a look at the top five recent research paper submission by Indian researchers in Academia.edu. 1.Cricket Analytics […]","categories":["AI Trends"],"tags":["AI Research","NLP"],"author_name":"Akshaya Asokan","publish_date":"2019-05-03T12:43:46","publication_year":"2019","word_count":682,"keywords":["Go","machine learning","artificial intelligence","AI","neural network","AI Research","NLP","deep learning","analytics","SQL","R","Java"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","deep learning","neural network","analytics","R","SQL","Go","Java"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/5-latest-research-papers-on-ml-you-must-read-in-2019\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":8955,"title":"News: Speed and Technology","content":"The advent of social media has meant that information travels fast. Everyone participating is judged, liked, trolled, and slammed, every moment. A decade after the advent of Facebook and Twitter, no one can afford to ignore the impact that they have. Businesses, politicians, actors, academicians, writers, have all used social media to advance their cause. In fact, taking a step forward are companies like Banjo. Banjo released a consumer app in 2011. This was a news app and the sources of information were social media feeds. This might not sound new as there are hundreds of such apps available for download. What’s new is the enterprise software they have developed recently that can detect events. The events are organized by location. The events are shown in great detail using pictures, texts and location. The posts are sourced from the mobile devices closest to the location of the event. There are implications of such technology for different industries. A tweet posted by an expert about the possible price of a stock could make or break billions of dollars for investors. In the field of finance, especially in high frequency trading, time lag is the key. Research has shown that effective use of Twitter sentiment about a company can result in superior returns. In another example, if there is a group of people witness to a crime and they are able to post pictures and messages via their mobile devices, the police can be alerted as they would get to know the exact location of the happening. Similarly, in the case of a disaster, the government can be alerted into action to manage the disaster. Google can predict the onset of epidemics based on individual’s searching for particular medicines. They can even predict hurricanes and storms based on searches for emergency rations, blankets, torches and other items. Now, compare and contrast this to the ways news was gathered and disseminated historically. A popular method was to appoint messengers to either collect or send news. There were people who were designated as ‘criers’ to shout out the news loud and clear. People were asked to gather at cross roads of their respective localities to spread the news. The great travelers were a good source of information and news from distant lands. They were accorded grand welcome and one of the reasons was their news carrying capability. Then there are examples of daily handwritten news sheets started by Julius Caesar. The present day marathon can also be traced back to the phenomenon of spreading news through a messenger. Marathon celebrates the courage of Pheidippides who ran 26 miles from Marathon to Athens to carry the good news of Athenian victory over the Persians. He sadly collapsed and died shortly after providing the news due to exhaustion. As a method of spreading news, pigeons were used by Persians. They trained these pigeons. The Mughals and Romans used pigeons. In fact they were used to aid the military. Surprisingly pigeons were used by financiers before the advent of the telegraph. Just imagine how much time all these modes of communication would have taken and sometimes the message would simply be lost owing to the mortality of pigeons. From pigeons who would take days to deliver news we are at a stage where we can get the news almost instantly. The use of pigeons to deliver news now seems a very distant past. But the truth is that they were used for transmission of news and information till just about a century ago. With the advent of the telegraph in the late 19th Century, the use of pigeons started to decline. The world has really evolved. However, there are some privacy issues involved. Not every individual who is privy to an event would like that information to be seen by everyone with an internet connection. The privacy settings in the respective social media sites should do the job. If some individual doesn’t want his\/her location information to be shared, they can simply turn that off. There could be some deleterious effects of news spreading fast. During riots, quick news could incite more people pretty fast and soon the mob would become difficult to handle. On the other hand, the law keepers would get the information quickly and they could arrive on the scene pretty quickly before much damage has been done. There are both positives and negatives like the proverbial two sides of a coin, to this lightning transfer of news. Used judiciously, this technology could redefine the world we live in forever. Authors Sanjay Fuloria is with Cognizant Research Center, Hyderabad; Sanjay.fuloria@cognizant.com Nupur Pavan Bang is with the Indian School of Business, Hyderabad; npbang@gmail.com","excerpt":"The advent of social media has meant that information travels fast. Everyone participating is judged, liked, trolled, and slammed, every moment. A decade after the advent of Facebook and Twitter, no one can afford to ignore the impact that they have. Businesses, politicians, actors, academicians, writers, have all used social media to advance their cause. […]","categories":["IT Services"],"tags":["social media analytics"],"author_name":"AIM Media House","publish_date":"2016-02-15T04:26:53","publication_year":"2016","word_count":778,"keywords":["Go","programming_languages:R","AI","social media analytics","programming_languages:Go","RAG","GAN","R"],"extracted_tech_keywords":["AI","RAG","R","Go","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/news-speed-and-technology\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10091611,"title":"16 Best Closed-Source LLMs You Must Know About","content":"OpenAI’s co-founder, Ilya Sutskever, recently admitted that open-sourcing AI was a mistake, stating that in a few years it will be “completely obvious” that it’s not wise. He also commented on the competitive nature of developing GPT-4, saying that it took all of OpenAI’s resources and that many other companies are vying to achieve the same thing. OpenAI is not the only one keeping its technology locked in. Here are 15 closed source LLMs which the researchers refuse to hand over to the public. 1. GPT-3 OpenAI’s GPT-3 caused quite a stir in May 2020 with 175B parameters, promising to outperform its predecessor, GPT-2. After initially releasing in limited beta, the model generated a lengthy waiting list of eager developers looking to tap into its advanced capabilities. Some people did wonder if GPT-3 was self-aware. In reality, unlike its predecessors, the model was trained to have a degree of common sense. 2. GPT-4 OpenAI did it again with the latest GPT-4. Boasting significant improvements, the model has astounded many with its uncanny ability to generate human-like text and even whip up images and code from mere prompts. But the catch is, as of now, only ChatGPT’s esteemed subscribers have access to the technology. 3. Megatron Turing NLG In October 2021, NVIDIA and Microsoft announced the Megatron-Turing NLG with 530 billion parameters, three times more than its closest competitors. The model is powered by DeepSpeed and Megatron transformer models. 4. ERNIE 3.0 Titan Baidu and the Peng Cheng Laboratory developed ERNIE 3.0 Titan, a pre-training language model that boasts 260 billion parameters. With its vast knowledge graph and unstructured data training, the model has achieved state-of-the-art results in over 60 NLP tasks. Baidu claimed that this is the world’s first knowledge-enhanced multi-hundred billion parameter model and the largest Chinese singleton model. 5. Jurassic-1 At release, AI21 Labs claimed that Jurassic-1 is “the largest and most sophisticated” LLM ever released for general use by developers. With 178 billion parameters, it is slightly bigger than GPT-3 and has the capacity to recognise 250,000 lexical items, five times more than other language models. The model was trained on Jumbo, consisting of 300 billion tokens collected from English-language websites. 6. Wu Dao 2.0 China’s latest masterpiece Wu Dao 2.0 is a language model built by the Beijing Academy of Artificial Intelligence (BAAI) with a staggering 1.75 trillion parameters, surpassing the capacities of GPT-3 and Google’s Switch Transformer. Wu Dao 2.0 covers both English and Chinese, and its abilities range from simulating conversational speech to writing poetry and generating recipes. 7. HyperCLOVA Naver Corp’s HyperCLOVA, the South Korean-language AI model, was released in May 2021. The company all set to launch this July to launch an upgraded version called HyperCLOVA X, which can understand images and speech in a multimodal format. Trained on a massive corpus of 560B tokens, the Korean GPT-3, as it is called, can be a game-changer in the world of natural language processing, according to Kim Yu-won, CEO of Naver Cloud Corp. 8. Gopher DeepMind’s Gopher is a 280 billion parameter transformer language model. The researchers claimed that the model almost halves the accuracy gap from GPT-3 to human expert performance, exceeding forecaster expectations and lifting performance over current state-of-the-art language models across roughly 81% of tasks. 9. Chinchilla Deepmind’s another addition to their animal-inspired lineup of models is Chinchilla – a 70B parameters model is designed to be compute-optimal. With 1.4 trillion tokens in its training data, Chinchilla was found to be optimally trained by equally scaling both model size and training tokens. Despite using the same compute budget as Gopher, Chinchilla boasts 4x more training data, making it a formidable contender in the language model space. 10. Galactica The avalanche of scientific data has made it tough to find valuable insights. While search engines can help, they alone can’t organise scientific knowledge. Enter Galactica, Meta’s large language model that can store, combine, and reason about scientific data. Trained on a large corpus of scientific papers, it outperformed existing models in a range of scientific tasks. However, the model was removed three days after its launch. 11. LaMDA The problem child of the LLM world caught quite a lot of attention after a Google former engineer Blake Lemoine claimed it to be sentient while testing the family of language models. Developed by Google with 137 billion parameters, LaMDA was created by fine-tuning a group of transformer-based neural language models and was pre-trained using a dataset of 1.5 trillion words, 40 times larger than previous models. Also read: The Wait for LaMDA is Almost Over 12. AlexaTM Amazon’s AlexaTM is the company’s 20 billion parameter large language model. Its unique encoder and decoder setup improves its performance on machine translation, making it stand out from the competition. Despite having only 1\/8 the parameters of its rival GPT-3, it outperformed it on both SQuADv2 and SuperGLUE benchmarks. 13. BloombergGPT Last month, Bloomberg unveiled BloombergGPT, a new large-scale generative AI model, specifically designed to tackle the complex landscape of the financial industry. This highly trained language model, optimised to parse and process vast quantities of financial data, seems promising in the NLP domain. 14. PanGu Last month, researchers at Huawei announced PanGu-Σ Trained on Ascend 910 AI processors. MindSpore 5 served as the framework, as the model underwent rigorous training with a whopping 329 billion tokens over a hundred days. 15. Kosmos-1 Microsoft released, “Language Is Not All You Need: Aligning Perception with Language Models,” featuring the remarkable Kosmos-1 multimodal large language model (MLLM). The tech giant’s team argues that MLLMs represent a critical leap forward in unlocking unprecedented capabilities and opportunities in language comprehension, far surpassing the capabilities of traditional LLMs.","excerpt":"Every model before GPT-4 that researchers don’t trust the public with.","categories":["AI Trends"],"tags":["OpenAI"],"author_name":"Tasmia Ansari","publish_date":"2023-04-17T18:30:00","publication_year":"2023","word_count":948,"keywords":["Go","ChatGPT","artificial intelligence","OpenAI","AI","ML","NLP","Aim","generative AI","R"],"extracted_tech_keywords":["AI","artificial intelligence","ML","NLP","generative AI","ChatGPT","OpenAI","Aim","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/16-largest-closed-source-llms-you-must-know-about\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10166345,"title":"Jensen Believes the Entire World is Wrong, NVIDIA isn’t","content":"If last year’s GTC was a pop concert, this year’s event felt like the Super Bowl of AI, said NVIDIA chief Jensen Huang. Only this time, there are no losers, just winners. “I’m up here without a net—there are no scripts, there’s no teleprompter, and I’ve got a lot of things to cover,” he quipped. The event was attended by over 25,000 in-person attendees, while 3,00,000 joined virtually. This prompted Huang to joke that the only way to accommodate more people at GTC was to physically expand the size of San Jose, where the event was held. This year, Huang, unfazed by DeepSeek’s impact, only means business. “The entire world got it wrong—the computation requirement and the scaling law of AI are more resilient and, in fact, hyper-accelerated. The amount of computation we need at this point, as a result of agentic AI and reasoning, is easily a hundred times more than we thought we needed this time last year,” Huang claimed. Huang emphasised the shift from “retrieval computing” to “generative computing”,  where AI generates answers based on context. He introduced “agentic AI”, which involves AI perceiving, reasoning, and planning actions. Huang introduced the open Llama Nemotron family of models, which are equipped with reasoning capabilities. According to him, this evolution, coupled with “physical AI” for robotics, has significantly increased computation needs. Building on this, Huang announced that  NVIDIA’s Blackwell architecture is now in full production, delivering 40 times the performance of Hopper. The Blackwell architecture boosts AI model training and inference, improving efficiency and scalability. Notably, Blackwell Ultra is set to hit systems later this year, but the real powerhouse is coming soon. Named after an American astronomer known for her research on dark matter, Vera Rubin is NVIDIA’s next-generation GPU and is expected to debut in 2026. Huang added that the Rubin chips will be followed by Feynman chips, which are slated to arrive in 2028. To accelerate large-scale AI inference, Huang introduced NVIDIA Dynamo, an open-source platform that powers and scales AI reasoning models within AI factories. “It is essentially the operating system of an AI factory,” Huang said Comparing Blackwell with older Hopper GPUs, Huang said, “The more you buy, the more you save. But now, the more you buy, the more you make.” He even joked that when Blackwell hits the market, customers won’t be able to give away Hopper GPUs and encouraged them to buy Blackwell. He further quipped that he is the ‘chief revenue destroyer’ as his statement might alarm his sales team, discouraging customers from purchasing remaining Hopper inventory. Ultimately, he stressed that technology is advancing so rapidly that buyers should invest in the latest and most powerful options available rather than settling for older models. Moreover, he expects data centre revenue to reach one trillion dollars by the end of 2028.  “I am fairly certain we’re going to reach that very soon.” NVIDIA is not just about big data centres. The company also unveiled two new supercomputers, DGX Spark and DGX Station, powered by the Grace Blackwell platform, allowing AI developers, researchers, and students to prototype, fine-tune, and run large models on desktops. Cute Disney Robots on the Way Huang is all in on robotics and is convinced that physical AI is the next big thing. On stage, he was joined by ‘Blue’, a small AI-powered robot resembling Disney’s Wall-E. He also introduced Newton, an open-source physics engine for robotics simulation developed in collaboration with Google DeepMind and Disney Research. Not stopping there, NVIDIA unveiled Isaac GROOT N1, the world’s first open Humanoid Robot foundation model. If that wasn’t enough, Huang also announced NVIDIA Isaac GR00T Blueprint, a system that generates massive synthetic datasets to train robots—making it way more affordable to develop advanced robotics. NVIDIA’s physical AI experience extends to autonomous vehicles. General Motors announced it will work with NVIDIA to optimise factory planning and robotics. Moreover, GM will use NVIDIA DRIVE AGX for in-vehicle systems to support advanced driver-assistance systems and enhanced safety features. Huang also mentioned the importance of safety in Autonomous Vehicles (AVs) and launched NVIDIA Halos, a comprehensive safety system for AVs. It integrates NVIDIA’s automotive hardware and software safety solutions with AI research to ensure safe AV development from cloud to car. While AI remains a priority, NVIDIA is now focusing on quantum computing. It will host its first Quantum Day on March 20—an interesting move considering Huang once claimed that “quantum computers are still 15 to 30 years away”.","excerpt":"While AI remains a priority, NVIDIA is now focusing on quantum computing.","categories":["Global Tech"],"tags":["NVIDIA"],"author_name":"Siddharth Jindal","publish_date":"2025-03-19T17:59:35","publication_year":"2025","word_count":745,"keywords":["big data","Go","API","agentic AI","ELT","AI","Scala","RAG","Aim","NVIDIA","R"],"extracted_tech_keywords":["AI","agentic AI","Aim","RAG","R","Go","Scala","API","big data","ELT"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/jensen-believes-the-entire-world-is-wrong-nvidia-isnt\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10100374,"title":"KPI Partners: Accelerating Business Growth with Modernization and Monetization of Data Assets","content":"KPI Partners, established in 2006 by Sid Goel and Kusal Swarnakar, is a leading provider of cloud-native solutions for Generative AI, Analytics, business intelligence, data management, and data engineering. With a team of over 600 professionals, they offer cloud-based services focused on scalability, security, and user-friendliness. Recognized by industry analysts, KPI Partners provides a comprehensive suite of services including business intelligence, analytics, Generative AI,  data science, data engineering, data migration, data visualization, data governance, and cloud consulting. Their clientele includes Fortune 500 companies, and have headquarters in Newark, California, with offices across North America, and Asia. KPI Partners assists organizations in achieving their business goals through their dedication to top-notch Business Intelligence (BI) and Analytics solutions, emphasizing the importance of data analytics and digital transformation in today’s fast-paced business landscape. The company’s primary industry focus areas include High Tech, Manufacturing, CPG (Consumer Packaged Goods), Retail, Banking, Insurance, and Higher Education, reflecting their profound commitment to these thriving domains. In an interview with Analytics India Magazine, the Founders, Kusal Swarnakar and Sid Goel speak about how KPI is transforming businesses with cloud-native solutions for data analytics and artificial intelligence. Analytics India Magazine (AIM): Looking back, how and why did the company start? How much has it grown from its inception until now? Founders of KPI: In 2006, we came together at Siebel Systems with a shared vision of harnessing the power of data and analytics to empower organizations. This led to the establishment of KPI Partners, where our mission was to provide comprehensive services and expertise to Fortune 500 companies in need of business intelligence and analytics. Our priority is to help customers overcome business challenges through industry best practices. With over 600 consultants and 350 clients, we have experienced significant growth and completed over 1,000 projects. We offer flexible delivery models. Headquartered in Newark, KPI Partners is a full life cycle data and analytics services company including Cloud Enablement, and expertise in data and analytics areas such as Data Warehouse, AI, and Cloud Migration. We maintain partnerships with leading vendors like AWS, Azure, GCP, Snowflake, and Databricks while ensuring unbiased recommendations for our customers. AIM: What challenges did you face during the initial months of founding KPI Partners? Founders of KPI: In the beginning of our venture, establishing trust and credibility with potential clients was a major challenge. We focused on being a niche player in the competitive market, specializing in complex projects. Our dedicated efforts were focused on providing exclusive solutions, including various offerings such as Cloud ERP & HCM Analytics, Real-Time Analytics, and Marketing Analytics. Our team of experienced consultants brought extensive expertise to the table. Through our technology specialization and ability to solve business problems, we have achieved growth and success, earning the trust of Fortune 500 companies. AIM: KPI Partners is leading the industry by making businesses easier. Could you brief on the different solutions and services that you offer to make it possible? Founders of KPI: KPI Partners is dedicated to simplifying and managing business processes. By utilizing tailored technologies and strategies, we help clients gain insights, reduce risks, and automate their processes. Our services include strategy, data engineering, cloud migration, data management, analytics, visualization and AI, enabling organizations to make informed decisions, enhance revenue, and increase profitability. We offer expertise in leading analytics platforms like AWS, Azure, and GCP, Snowflake and Databricks. We help our customers harness the potential of modern analytics, leverage their data assets for profitability, and achieve cost savings during their transition to cloud computing. We implement DataOps, MLOps, and DevOps methodologies using tools like Azure DevOps, Kubernetes, and AWS CodePipeline, facilitating efficient and collaborative data and machine learning operations. We specialize in marketing operations and analytics using platforms like Adobe and Google Marketing Platform, optimizing marketing strategies and enhancing customer engagement. Our data management solutions ensure data quality, governance, and compliance across the organization, utilizing platforms like Informatica MDM, Collibra, and Stibo. We provide data engineering services to build scalable data pipelines using technologies like Spark, ETL, Databricks, AWS Glue, and Azure Data Factory. With expertise in Python, Databricks, AWS Sagemaker, and other tools, we deliver data science and AI\/ML solutions, unlocking actionable insights and enabling intelligent decision-making. We offer database services for efficient management and security using platforms like Snowflake, Synapse, Red Shift, BigQuery, and Oracle autonomous databases. Moreover, our expertise extends into the realm of Generative AI, where KPI Partners excels in delivering innovative solutions that are expertly crafted to elevate and streamline various aspects of business operations. Our flagship product, KPIGPT, harnesses the power of Generative AI to cultivate a more engaged and intelligent workforce by leveraging enterprise data. Through tailored, intelligent solutions, we empower organizations to enhance performance, drive efficiency gains, and fortify their competitive edge. Our unified Intelligent Search Solutions and Knowledge Management System offer an intuitive platform for swift and efficient retrieval of enterprise data and insights, revolutionizing data access while saving precious time. With a holistic approach, KPI’s Generative AI solutions optimize processes across customer service, financial services, insurance, sales, supply chain, human resources, and legal functions, delivering heightened efficiency and superior outcomes. Moreover, we prioritize data security by implementing robust safeguards and Generative AI guardrails to ensure the protection of our client’s valuable data. In summary, KPI Partner has emerged as a prominent force in the realm of digital transformation, leveraging the capabilities of AI to propel business expansion. AIM: KPI Partners has become a leading name in data analytics and digital transformation across many sectors. What was your growth plan? Founders of KPI: KPI Partners’ growth plan has been crucial in establishing its leadership in data analytics and digital transformation. We attribute our success to several strategies. Firstly, we expanded our reach globally by opening new offices and providing assistance to diverse customers, establishing a strong presence in various regions. Secondly, we diversified our product and service offerings to meet evolving customer needs and market demands, offering comprehensive solutions aligned with industry trends. Thirdly, we prioritized research and development, investing in accelerators to adopt innovative approaches, and harnessing cloud-based technologies. Additionally, strategic partnerships with leading vendors enhanced our capabilities and service offerings. Lastly, our focus on customer satisfaction built strong relationships and a reputation for excellence. These initiatives have solidified KPI Partners’ position as a trusted partner in driving business success through data-driven insights and digital solutions. AIM: What is the USP (Unique Selling Point) of the company? Founders of KPI: KPI Partners distinguishes itself in the market through its unique offerings and exceptional capabilities in modernizing and monetizing data assets. The accelerators provided by KPI Partners offer several advantages, including shortened timelines, reduced risks, and lowered costs. These accelerators are detailed below DATA ENGINEERING ACCELERATORS: Oracle Fusion Cloud Data Extractor Data Quality Validator KPI’s ADF++ Accelerator KPI’s Airflow++ Accelerator SNOWFLAKE ACCELERATORS: Line of Business (LOB) Analytics for Snowflake Migration from Oracle \/ SQL Server to Snowflake Snowflake Cost Optimization Application Snowflake Security Automation Application Snowflake Data Governance Application Data Quality Validator for Snowflake LLM ACCELERATORS: KPIGPT Accelerator for Insurance KPIGPT Accelerator for Banking KPIGPT Accelerator for Recruitment PRE-BUILT LOB ANALYTICS ON ANY CLOUD: KPI Partners offers pre-built Line of Business (LOB) analytics solutions that are compatible with various cloud platforms, including Financial, HCM, Supply Chain, Purchasing, Project Management, CRM Sales, CRM Marketing, Order Management, Transportation, Real-Time (EBS), Higher Ed, and Trade Analytics. POWER BI ACCELERATORS: KPI Partners provides Power BI accelerators for easy migration from OBIEE, Tableau, and Qlik to Power BI, as well as a Power BI Email Bursting Accelerator. GOOGLE ACCELERATORS: KPI Partners offers utility tools for migrating data from Oracle to BigQuery and Snowflake to BigQuery, as well as a Marketing Analytics solution tailored for Google Analytics. Thus, KPI Partners’ unique selling point lies in its comprehensive suite of accelerators and pre-built analytics solutions, which not only reduce costs but also enhance efficiency and data management across various platforms. AIM: What is your current team size? How do you promote a culture of growth and have a high employee retention rate? Founders of KPI: KPI Partners currently has a global team of over 600 employees. We prioritize growth and a positive work environment, leading to high employee retention. We follow key practices including professional development through learning and upskilling programs, mentorship and coaching, and monthly, quarterly, and annual performance recognition. Transparent communication is maintained through regular updates and feedback mechanisms. Work-life balance is important, with flexible arrangements and well-being initiatives. Collaboration is encouraged, fostering a sense of belonging and collective growth. Competitive compensation, comprehensive benefits, and attractive perks are provided, contributing to employee satisfaction. These practices aim to create a positive work environment that values employee contributions and promotes retention, believing in the importance of investing in their development and well-being. AIM: What is the roadmap ahead for KPI Partners in the next 5 years? Founders of KPI: We are actively seeking new opportunities in different markets and regions to expand our presence. Our goal is to meet the needs of clients through various models, including Onshore, Offshore, Nearshore, and a unique Blended model. To achieve this, we are investing in research and development, acquiring top talent, leveraging innovative technologies, and forming new alliances. Our long-term vision is to become a leading industry leader by strengthening our brand presence and exploring untapped markets. We are focusing on strategic marketing, continuous innovation, and the development of exceptional products. AIM: KPI Partners is one of the leading global consulting firms focusing on providing strategic, technological, and digital transformation solutions to its clients. What message do you want to convey to the readers? Founders of KPI: We are a company that specializes in data and analytics, helping businesses thrive in the rapidly advancing digital age. Our team of expert consultants has a proven track record and extensive experience in architecting, engineering, migrating, and managing big data clusters. We empower customers to harness the full potential of modern analytics, all while capitalizing on the value of their data assets and achieving cost savings during their transition to the cloud.","excerpt":"KPI Partners, a leader in digital transformation, is dedicated to accelerate business profitability by providing customized data analytics and digital transformation services and solutions to address the unique needs of each client.","categories":["AI Highlights"],"tags":[],"author_name":"Shritama Saha","publish_date":"2023-09-21T10:00:00","publication_year":"2023","word_count":1668,"keywords":["data science","machine learning","artificial intelligence","AI","ML","MLOps","RAG","Aim","analytics","generative AI"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","data science","analytics","generative AI","MLOps","Aim","RAG"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/kpi-partners-accelerating-business-growth-with-modernization-and-monetization-of-data-assets\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":38017,"title":"A Complete Guide to Cracking The Predicting Restaurant Food Cost Hackathon By MachineHack","content":"We all love food. And it is only normal to have a craving for one of your favourite foods from a special restaurant that we all love to have at least once in a month. However, there is a strong factor that will make us reconsider going back to that special restaurant — cost. Machinehack’s Predicting Restaurant Food Cost Hackathon lets all Data Science enthusiasts to play with data collected from various sources, which includes the price information of thousands of restaurants across India. The contestants will predict the cost of a meal for different restaurants across the country based on various features. About the Data Set The hackathon is about predicting the average price for a meal. The data consists of the following features. Size of training set: 12,690 records Size of test set: 4,231 records Columns\/Features : TITLE: The feature of the restaurant which can help identify what and for whom it is suitable for. RESTAURANT_ID: A unique ID for each restaurant. CUISINES: The variety of cuisines that the restaurant offers. TIME: The open hours of the restaurant. CITY: The city in which the restaurant is located. LOCALITY: The locality of the restaurant. RATING: The average rating of the restaurant by customers. VOTES: The overall votes received by the restaurant. COST: The average cost of a two-person meal. Click here to participate in the hackathon. Solving Predicting Restaurant Food Cost Hackathon By Machnehack Use the following links to our top tutorials to help you with this challenge: Flight Ticket Price Prediction Hackathon: Use These Resources To Crack Our MachineHack Data Science Challenge Hands-on Tutorial On Data Pre-processing In Python Data Preprocessing With R: Hands-On Tutorial Getting started with Linear regression Models in R How To Create Your first Artificial Neural Network In Python Getting started with Non Linear regression Models in R Beginners Guide To Creating Artificial Neural Networks In R","excerpt":"We all love food. And it is only normal to have a craving for one of your favourite foods from a special restaurant that we all love to have at least once in a month. However, there is a strong factor that will make us reconsider going back to that special restaurant — cost.   […]","categories":["Deep Tech"],"tags":["Data Science","Hackathon","Machinehack"],"author_name":"Amal Nair","publish_date":"2019-04-22T06:11:49","publication_year":"2019","word_count":313,"keywords":["data science","Go","programming_languages:R","AI","neural network","Machinehack","programming_languages:Go","RAG","Hackathon","Python","programming_languages:Python","Data Science","R"],"extracted_tech_keywords":["AI","neural network","data science","RAG","Python","R","Go","programming_languages:Python","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/a-complete-guide-to-cracking-the-predicting-restaurant-food-cost-hackathon-by-machinehack\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10118510,"title":"Millions of AI-Powered Chinese EVs Could be Running on Indian Streets Soon","content":"Tesla CEO Elon Musk may likely announce the entry of his autonomous EV-making company into India after he meets Indian Prime Minister Narendra Modi later this month. Reportedly, the company chasing Level 5 autonomous driving is also scouting locations in New Delhi and Mumbai to open its first few physical stores in the country. We could soon have luxury Tesla EVs stuck in Mumbai or Bengaluru’s infamous traffic jams. However, Tesla is not the only foreign EV company making a foray into the Indian market. Earlier this year, Chinese electric vehicle manufacturer BYD entered the Indian market, boldly stating its intention to seize 90% of the nation’s EV market. Surprisingly, BYD, a company few people have heard of outside China, surpassed Tesla in sales volume during the final quarter of 2023. For context, BYD experienced a 73% sales surge in 2023, selling 1.6 million units, trailing Tesla’s 1.8 million units. Although Tesla maintains its lead in annual sales figures for 2023, there’s a strong possibility that BYD could soon overtake Tesla in terms of overall sales. Chinese Cars on Indian Streets Last month, BYD, which has achieved Level 3 autonomy, introduced the Seal electric vehicle in India, priced between INR 40 to 55 lakh. Within a few hours of the launch, the vehicle garnered over 200 bookings. While it’s still unclear how Tesla will price its models, their most affordable car is expected to fall in the same price bracket. Prices can escalate to INR 2 crore for higher-end models like the Tesla Model X. We’ll have to wait and see if Tesla adopts a competitive pricing strategy in India to compete with BYD and other EV manufacturers, including market leader Tata Motors. However, BYD will undoubtedly introduce affordable cars in the Indian market. For instance, the BYD Seagull EV Honor Edition starts at USD 9,700, or approximately INR 8-9 lakh, excluding taxes. In the coming years, BYD will bring its cheaper cars to India, targeting middle-class Indians. Notably, the mid-range price segment is the most lucrative for automobile makers, accounting for over 40% of the market, and BYD has already made its intentions clear about targeting this segment. This could mean the average Indian car may be a Chinese car in the coming years. Moreover, BYD’s entry into India could open the doors to hundreds of Chinese EV companies. According to Wired, nearly 300 EV companies operated in China last year. This could mean that China-made EVs might flood the Indian market in the coming years, similar to Chinese smartphones, which seem to be ubiquitous in the hands of every other Indian. Interestingly, phone maker Xiaomi has also begun selling its EV in the Chinese market this year. Meanwhile, the government is not planning to put any restrictions on the import of EVs to India, even from China. While there is a limit to the number of vehicles imported from a foreign country, there are workarounds. BYD, for instance, is seeking a homologation certification from the Automotive Research Association of India (ARAI) for its vehicles, allowing it to bypass the import restrictions. Selling AI Cars While autonomous vehicles on Indian streets still remain a distant dream, automobile makers are increasingly introducing AI-powered features such as smart parking, advanced driver-assistance systems (ADAS) and battery management systems. Earlier this year, BYD introduced an AI-powered smart car system powered by Drive Thor, NVIDIA’s next-generation in-vehicle computing chip. The second-largest EV maker in the world unveiled the XUANJI AI large model, marking the first integration of AI across all vehicular domains. Other Chinese EV makers like Xpeng have also announced their plans to invest millions of dollars in AI. Many industry experts that AIM spoke to told us that AI features will play an important role when customers shop for cars. Chinese EV companies can sell AI-powered cars at reasonable prices and gain significant market share not just in India but globally. China’s dominance in EVs is worrying China’s growing dominance in the global EV landscape has become a cause for concern for many in the western world. Former US President Donald Trump claimed, if re-elected, he would impose a 100% tariff on Chinese cars. US Commerce Secretary Gina Raimondo claimed that modern cars are essentially iPhones on wheels. She expressed concerns that if Chinese-made cars operate on American roads, they could be susceptible to hacking, posing a potential threat to national security. Investment bank UBS said in a report that BYD and other leading Chinese EV companies are set to conquer the world market with high-tech, low-cost EVs for the masses. This could also hold true for India, posing a significant challenge for Indian homegrown companies such as Tata, Maruti and Mahindra, which are developing their own AI-powered EV models. While India does not intend to restrict EV imports, its stance might change if Chinese EVs flood the Indian market. Earlier, the government banned many Chinese apps, including TikTok, citing national security threats. Banning Chinese EVs under the guise of national security remains a possibility.","excerpt":"At the same time, banning Chinese EVs under the guise of national security also remains a possibility.","categories":["IT Services"],"tags":["Tesla"],"author_name":"Pritam Bordoloi","publish_date":"2024-04-18T13:06:15","publication_year":"2024","word_count":833,"keywords":["Go","programming_languages:R","AI","RPA","Scala","RAG","Ray","Aim","programming_languages:Scala","Tesla","R"],"extracted_tech_keywords":["AI","Aim","Ray","RAG","R","Go","Scala","RPA","programming_languages:R","programming_languages:Scala"],"url":"https:\/\/analyticsindiamag.com\/it-services\/millions-of-ai-powered-chinese-evs-could-be-running-on-indian-streets-soon\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":48059,"title":"Python 3.9: New Changes Data Scientists Should Expect","content":"Python Software Foundation released Python 3.8 on Monday, enhancing developer experience with newly added features. With the stable release of Python 3.8, the firm has now introduced changelog of Python 3.9, thereby, shifting the focus on new advancements in the upcoming version. The next version of Python is scheduled to be introduced next year. What’s New? Python 3.9’s changelog includes several modifications in security, core and builtins, libraries, documentation, and more. Significant amendments in Python 3.8 with assignment expressions, positional-only arguments, enhancement in embedding, and others, the organisation have set the bar high for its next update. The changelog of Python 3.9 has not disappointed either, as it had made extensive changes for improving the overall user experience of the language which is used for programming machine learning, data science and artificial intelligence, among others. Makeovers in core and builtins Dictionary in Python is one of the most used unordered collection of data in a key-value format. A data scientist often uses a dictionary while handling data, as it offers various attributes to intuitively unearth insights into data. Besides, these dictionaries are utilized to efficiently gather data while iterating in for loop. This allows analysts to read it into a pandas data frame, a popular data structure for moulding, wrangling, and analysing data. For the succeeding version of Python, it has optimized dictionary constructor to morph its adeptness for mitigating programming challenges. Besides, the update() method has been evolved to accept iterable and another dictionary object to update dictionaries keys and values. Additionally, the compilation problem with the break and continue inside the “finally” block has been fixed. Earlier, if the return in the try block gives back a non-constant value, the break and continue did not function the way it should. This is a big relief to programmers for handling exceptions effortlessly. As a part of its core and builtins corrections, the firm has also improved error message related to partially initialised modules. Another change is about the AssertionError in the assert statement, which raised error needlessly. Further, in the next version, the firm has ameliorated the sum() function to make gits more concise. Transformation in libraries Python has depreciated the split() method because it returned unreliable outputs when implemented with _tkinter.TkappType object; the split() method has been replaced with the splitlist() method that is unfailing in delivering desired results. Besides, there are a few transformations in some of the most used datetime and pickel modules to boost functionality. Amendments in Documentation Good documentation is central for any programming language to proliferate, which helps companies focus on making it straightforward. This makes the users understand the code effortlessly for implementing in their analysis. Python Software Foundation is also committed to improving documentation by continually enhancing the readability while including the specificity of the modules, functions, methods, and others. In the next iteration, datetime module documentation has been overhauled to simplify it. Over the years, datetime module has been one of the most difficult to assimilate among programmers, thus with this change, the organisation has showcased its intent to assist users in every aspect of programming. Outlook With this article, we tried to keep you informed about the modification that will directly have an impact on the way you analyse data. Thus, we tried to keep you informed on how the reforms will benefit in the future. Apart from these amendments, there are several others that can improve the day-to-day operation of data exploration. We would suggest you inspect the released changelog of Python 3.9 for getting an in-depth knowledge of the next Python.","excerpt":"Python Software Foundation released Python 3.8 on Monday, enhancing developer experience with newly added features. With the stable release of Python 3.8, the firm has now introduced changelog of Python 3.9, thereby, shifting the focus on new advancements in the upcoming version. The next version of Python is scheduled to be introduced next year. What’s […]","categories":[],"tags":["Data Analytics","Data Science"],"author_name":"Rohit Yadav","publish_date":"2019-10-16T17:07:07","publication_year":"2019","word_count":594,"keywords":["data science","Go","machine learning","artificial intelligence","TPU","AI","Git","Python","Data Analytics","Data Science","R","Pandas"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","data science","Pandas","TPU","Python","R","Go","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/python-3-9-new-changes-data-scientists-should-expect\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10042889,"title":"Is AI A Good Judge Of Cause &#038; Effect?","content":"“Causal reasoning is an indispensable component of human thought that should be formalized and algorithimitized towards achieving human-level machine intelligence.”Judea Pearl Incorporating insights of psychology research into algorithms is tricky as the former is not exactly a quantifiable metric. But, it can be quite useful as algorithms are venturing into a world full of “trolley problems” in the form of self-driving cars and medical diagnosis. Tobias Gerstenbeg, assistant professor of psychology at Stanford, believes that by providing a more quantitative characterisation of a theory of human behavior and instantiating that in a computer program, we can make it easier for a computer scientist to incorporate such insights into an AI system. Gerstenbeg and his colleagues at Stanford have developed a computational model to understand how humans judge causation in dynamic physical situations. About the model Billiard board simulation experiment (Image credits: Paper by Gerstenberg et al.,) In their paper on the counterfactual simulation model (CSM) of causal judgment, the researchers begin my making three key assumptions: Causal judgments are about difference-making.Difference-making for particular events is best expressed in terms of counterfactual contrasts over causal models.There are multiple aspects of causation which correspond to different ways of making a difference to the outcome that jointly determine people’s causal judgments. As a case study, the researchers first applied the CSM to explain people’s causal judgments about dynamic collision events. They considered a simulated billiard ball B as shown above that enters from the right, headed straight for an open gate in the opposite wall. Blocking the path, they placed a brick. Ball A would then enter from the upper right corner and collide with ball B, which bounces off the bottom wall and back up through the gate. So, now the question is: did ball A cause ball B to go through the gate? It’s obvious that without ball A, ball B would have run into the brick rather than go through the gate. Without the brick in ball B’s path, it would have gone through the gate anyway without any assistance from ball A. The causation relationship between ball A and ball B in presence and absence of an external factor is being checked here. Gerstenberg and his colleagues ran such scenarios through a computer model designed to predict how a human evaluates causation. The idea here is that people judge causation by comparing what actually happened with what would have happened in relevant counterfactual situations. Indeed, as the billiards example above demonstrates, human’s sense of causation differs when the counterfactuals are different – even when the actual events are unchanged. Extending CSM to AI Now the researchers are working on extending the same theory of counterfactual simulation model of causation onto AI systems. The goal here is to develop AI systems that understand causal explanations the way humans do. And, to be able to demonstrate scenarios where AI systems are made to analyse a soccer game and pick out the key events causally relevant to the final outcome; whether it is the goals that caused the win or counterfactuals such as if saves by goalkeeper contributed more. This is a task that would need the AI system to mimic the smartest of team managers. However, Gerstenberg admits that their research is still in a nascent stage. “We can’t do that yet, but at least in principle, the kind of analysis that we propose should be applicable to these sorts of situations,” he added. In the SEE (the science and engineering of explanation) project, funded by Stanford HAI,  the researchers are using natural language processing to develop a more refined linguistic understanding of how humans think about causation.. Through their study on CSM, the researchers have tried to answer the fundamental question: how do people make causal judgments? The results revealed that people’s judgments are influenced by different aspects of causation, such as whether the candidate cause was necessary and sufficient for the outcome to occur, as well as whether it affected how the outcome came about. By modeling these aspects in terms of counterfactual contrasts, the CSM accurately captures participants’ judgments in a wide variety of physical scenes involving single and multiple causes. Researchers believe that CSM can be of great significance in many subfields of AI, including in robotics, where AI is required to exhibit more common sense to collaborate with humans intuitively and appropriately.","excerpt":"“Causal reasoning is an indispensable component of human thought that should be formalized and algorithimitized towards achieving human-level machine intelligence.” Judea Pearl Incorporating insights of psychology research into algorithms is tricky as the former is not exactly a quantifiable metric. But, it can be quite useful as algorithms are venturing into a world full of […]","categories":["AI Features"],"tags":[],"author_name":"Ram Sagar","publish_date":"2021-07-05T12:00:00","publication_year":"2021","word_count":726,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","ai_applications:robotics","R"],"extracted_tech_keywords":["AI","R","Go","programming_languages:R","programming_languages:Go","ai_applications:robotics"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/is-ai-a-good-judge-of-cause-effect\/","complexity_score":3,"technical_depth":6,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":22337,"title":"Women In New Tech: Barkha Sharma Of Bash.ai Talks About Challenges Of Founding A Startup In Patriarchy-Driven Space","content":"With the newer technologies like chatbot and robotics taking a centre stage, many startups have emerged in the space in areas like banking, finance, retail and even HR. Bash.ai is one such AI powered HR automation tool that uses artificial intelligence and machine learning to improve HR processes and in turn employee experiences. This New Delhi based startup was founded by Barkha Sharma, a graduate degree holder in Computers Application and master’s degree in Business Administration, who has been part Human resource domain for more than 7 years. Her knack for the unknown, her passion for Human Resources and her love for technology pushed her to create BASH, and in less than a year it has gained immense popularity in the human resources circle. With an undying entrepreneurial spirit, she continues to inspire her team to bring more product offerings in the arsenal. Bash currently announced the acquisition of Wemo, a New Delhi-based technology startup and creative development agency for design and product related requirements. With this acquisition Bash aims to integrate Wemo’s UX expertise of building mobile apps and websites to increase the scale of the company, and achieve the goals if building compelling experiences for its repertoire of brands. This women’s day, we interacted with Barkha Sharma, CEO and founder, Bash to understand her journey as a woman in the new tech space, where we witness a scant number of women in senior and leadership roles. Q. What does the career in analytics\/data science\/AI look like for a woman? Your thoughts on incorporating more women in new tech sectors. The experience as a woman working in tech has been both challenging and wonderful. There have been times when I have realized much later that things that I experienced were due to gender discrimination. The Women in Tech ratio in India is abysmal. This comes from personal experience as a recruiter in the recent past and now as someone running a tech company. The world of analytics\/data science\/AI needs more women. Here are some interesting numbers, Innovation-focused companies are $44 million more valuable on average when women hold their positions of power. Other research shows that female tech entrepreneurs generate 35% higher returns than male counterparts. An investment in both hiring and retaining more women actually leads to growth and greater returns for companies — hiring fewer women does not. Q. What are the key changes in education\/career choices that could change the current scenario? People have started acknowledging that diversity is a priority. Many companies have started incorporating it into their plans. These are steps in the right direction. Though I feel, hiring more women isn’t the end-solution though. It has to be a wider cultural shift. Casual sexism needs to highlighted. Education amongst the team as to how to make a workplace friendlier needs to be part of trainings. Also, more tech publications, such as this one, are featuring more women stories. It becomes important to emphasis that women or men in analytics are as good as the dataset they have :-) Q. Why did you choose this field as a career option? Honestly, it happened extremely organically for me. I graduated in Bachelors in Computers and then went on to do my MBA. After graduating, I deep dived into the field of Human Resources and ended up being part of HR function for 7 years. Right from an employee’s feeling of uncertainty to the distress of the human resource team, working with enterprises with processes in place to consulting startups and building their processes from scratch, and having endless conversations with people who felt that HR is the backbone of the company to people who felt that the appendix is more useful than its function, I ended up working on all facets of HR. Over the years, I became extremely passionate about HR and more so fascinated with the impact technology can have on it as a function and process. In a world with driverless cars, the HR function has not embraced tech like the other functions. Our belief in AI to bolster efficiency in HR was our biggest motivator. It pushed us to imagine a world where businesses would use AI and deep learning to improve efficiency of HR. It was through the course of these multiple discussions about workplace HR processes that the idea of Bash developed. Q. What is your growth story so far being a founder of AI based startup? It’s been an incredible rollercoaster of a ride so far. As is true with any start ups journey. I quit my full time job and founded Bash.ai in January 2017 with 3 people working out of my living room, and after many sleepless nights and insane amounts of coffee, we launched it in March 2017. Today, just a year on, we’re a 15-person strong team of engineers, product developers, and operations enablers working with the HR processes of organizations helping them automate simple routine tasks, in turn, making their lives a lot more efficient. It has been incredible to see the fantastic response we’ve got from industry leaders and organizations in India, the Silicon Valley and Middle East. It is our belief in AI to bolster efficiency in HR which has been our biggest growth driver. What keeps me excited about Bash.ai is that through automation we can bring about a massive change in the ecosystem. When I look at the team that we’ve built so far, and see their immense drive to build something meaningful, it drives me even further. Having the done the hard work, we are poised to take a leap. Its like preparing for years for your performance on the big stage. Q. Would you like to give a brief note about Bash.ai and the idea of founding it? Bash.ai, which launched in March, 2017 in New Delhi-NCR uses artificial intelligence (AI) and machine learning (ML) to improve HR processes and, in turn, employee experiences. Essentially, it is an automation tool which uses AI and big data to power a virtual assistant to drive HR for businesses. It mimics cognitive functions related to the HR function using a rule based AI powered bot resulting in increased productivity in the workplace. Quite simply, a rule-based bot can help in terms of employees getting information faster with a conversational interface, enabling the HR team to focus on more meaningful tasks rather than getting caught up with repetitive tasks and generic queries. With Bash’s user-friendly front-end chat interface, it gives organizations a highly accurate and real-time way to automate conversations with employees within an organization’s intranet. It can also be accessed using instant messenger platforms such as Facebook Messenger, Slack, Telegram, Kik, Google and Microsoft’s Skype. While, we keep adding to our repertoire of modules, currently Bash.ai has four modules: Welcome Aboard: aimed at employees who have accepted the offer but are yet to join. Post Hire Orientation: a virtual buddy to bring the new joiner in-tune with the company and its policies. HR Helpdesk: the core engine answering all the queries an employee might have for the HR. Employee Engagement: to get the pulse of the employees. More of an emotional and aspirational barometer. It was the coming together of a few instances that resulted in the founding of Bash.ai. However, a particular incident which took place at one of my previous workplaces got me thinking about the need for technological advancement in HR. Q. Being a woman, what were the challenges in incorporating a startup in tech space? A big challenge is being a woman running a startup. There are small incidents that women not just in technology but across industries face. Patriarchy is real and there is so much to be done when it comes to closing the gender gap not just at work but also in our everyday lives. While the gender gap continues to exist not just across technology industries in India but also abroad, it is high time we started closing this gap. Visit any technology conference and you’ll see the gap that I am talking about. I read last year only 2% of equity funding raised had a woman founder. I don’t claim to have all the nuances of this cleared, but this has to improve. Q. Do you struggle to maintain a work-life balance? I guess one never really stops thinking about the business, product and team – they’re constantly on my mind. I do, however, try to maintain a balance between work and my personal life. It took me a really long time to get to the schedule that I have in place right now. I have relapsed multiple times but now mostly I do stick to this. I give 2 hours to myself in the morning where I organise the day for my team and myself. My day is filled with a lot of phone calls, emails, Slack interactions and countless number of cups of tea. Once back home I workout for an hour. Then its either catching up with friends, family or Netflix. Not necessarily in the same order. However, I still need to work on my routine while traveling which goes completely upside down because of work demands.","excerpt":"With the newer technologies like chatbot and robotics taking a centre stage, many startups have emerged in the space in areas like banking, finance, retail and even HR. Bash.ai is one such AI powered HR automation tool that uses artificial intelligence and machine learning to improve HR processes and in turn employee experiences. This New […]","categories":["AI Features"],"tags":["Interviews and Discussions","Women in Tech"],"author_name":"Srishti Deoras","publish_date":"2018-03-07T05:21:53","publication_year":"2018","word_count":1529,"keywords":["data science","artificial intelligence","machine learning","AI","ML","RAG","Aim","deep learning","Women in Tech","analytics","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","data science","analytics","Aim","RAG","R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/women-in-new-tech-barkha-sharma-of-bash-ai-talks-about-challenges-of-founding-a-startup-in-patriarchy-driven-space\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10062976,"title":"Why should AI &#038; data science courses include business case studies?","content":"According to IBM, finding and hiring staff with the right mix of skills and experience is a painstaking process. Around 69 percent of organisations struggle to recruit quality candidates, an Accenture study showed. “Good data scientists are good at solving word problems,” said Nitesh Shende, data science lead at Porter. He said most data scientists struggle to situate machine learning models in a business context. “The ability to identify where and which data science techniques to use will only come through case studies,” said Shende. We took a look at the computer science engineering and AI and data science courses curriculum in India, and found most universities and institutions are limited to core technical skills and lack business components. For example, IIT Delhi’s department of computer science and engineering courses teach technical subjects and soft skills, but no specific modules focused on job-ready skills, or business case studies. On the other hand, NYU’s computer science courses have a ‘launch pad’ and ‘entrepreneurship capstone’ projects dealing with business development, and hands-on learning What’s the solution? To bridge the gap, tech giants Google, IBM, Microsoft, and AWS, and foriegn universities are offering online certification courses, training, distant management programmes, etc. For instance, Google is offering Grow with Google for professional training and an opportunity to network with top employers. ​​https:\/\/www.youtube.com\/watch?v=eR-o4CvoPVo Similarly, the IBM data science professional certificate course (on Coursera) includes a series of hands-on labs in the IBM Cloud, teaching practical skills. Supply side India has about 4,282 engineering institutes, and every year close to one million engineering graduates enter the job market. Of which, computer science has the highest numbers, followed by mechanical, electronics, civil, and electrical. Nearly 127 diploma institutes and 663 undergraduate colleges offer specialised courses in AI, data science, analytics, blockchain, machine learning, and robotics– accounting for 10 percent of the total number of engineering institutes in India, All India Council for Technical Education (AICTE) data showed. Today, not just engineers, statisticians, economists, and mathematicians are getting into the world of data science and analytics. Data science is touted as the most in-demand role in 2022. As per Monster Annual Trends report, 96 percent of companies are planning, or likely to hire new staff with relevant skills to fill data science and analytics related roles in 2022. Today, most interviews have a question around how to apply technological knowledge in a business context. It’s crucial for colleges and universities to prepare students to solve real world problems. Hence, introducing ‘business case studies’ in the undergraduate  curriculum can make a huge difference. Demand side According to data science talent expert Vin Vashishta, businesses don’t have a problem finding qualified data scientists but in recognising the qualified data scientists who apply. “I could apply for a Data Science role and if I don’t goof my resume with all the right buzzwords, I would never see a hiring manager,” he said, adding the entire process is broken. “Great talent gets screened out all the time,” he added. He suggested that recruiters should hire entry level scientists and upskill them. For senior level talent, they should manually read resumes and build a process to quickly hire them. “Phone screen. One hour interview. Train leaders to hire and help them build a process,” said Vashishta. He said: “Companies hiring for tech talent need to realise, it is not 2012 anymore. Gauging talent is complex and those empty chairs are going to put your company out of business. Adapt or die.” Jordan Luxford, AI recruiter, NLP specialist, and host of ‘The NLP Zone’ podcast, said: “Many companies require a Master’s degree or above for their tech vacancies regardless of the candidate’s experience, and other companies do not care about education at all.” “I completely understand the advantages of having a Master’s or PhD in tech positions, but I have also worked with some of the best candidates who did not go down the traditional education route,” said Luxford. He said it is important to have a standard, but companies need to be flexible. The companies will miss out on gifted, natural talents that started working from a young age and therefore chose not to pursue a degree, he added.","excerpt":"Nearly 127 diploma institutes and 663 undergraduate colleges offer specialised courses in AI, data science, analytics, blockchain, machine learning, and robotics.","categories":["AI Features"],"tags":["data science challenges","data science latest"],"author_name":"Amit Naik","publish_date":"2022-03-18T16:00:00","publication_year":"2022","word_count":696,"keywords":["data science","Go","machine learning","AWS","AI","data science latest","cloud_platforms:AWS","data science challenges","NLP","analytics","GAN","R"],"extracted_tech_keywords":["AI","machine learning","NLP","data science","analytics","AWS","R","Go","GAN","cloud_platforms:AWS"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/why-should-ai-data-science-courses-include-business-case-studies\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10111038,"title":"7 Must-Have Home Robots in 2024","content":"Home robotics has seen significant advancements lately with several companies introducing robots designed to be integrated seamlessly into our daily lives. These robots go a step further from machines when it comes to being equipped with cutting-edge technology. Going beyond just being vacuum cleaners, robots are attempting to do tasks that humans need not do anymore, like folding clothes, opening a door etc. Here is a list of 7 home robots you could own this year. Ballie Samsung has once again announced Ballie, an AI robot that will follow you around, acting as a home assistant. It can connect to all the IoT devices at home and even guard and feed your pets. The robot can answer questions, text you, and do everything that Alexa can. The cute round robot was introduced in 2020 at CES and this year they’ve relaunched Ballie to mixed reviews. “There have been many robots in the past, but generative AI has raised the endless potential of the robot market, and it will be a chance for all devices to become smarter and more developed,” said Han Jong-hee, vice chairman, Samsung Electronics. A new feature of the robot is the two-in-one switchable projector, which can display anything you want – a video call, a workout video, or even the insides of your fridge. The robot can detect the person’s posture and facial angle to project the video accordingly. Astro Amazon Astro, a small moving robot with a screen, is designed to assist with home monitoring and staying connected with the family. It can autonomously move around the home, navigate to specific areas, provide live views through the Astro app, send alerts for unrecognised persons, and assist with elderly care. It can also work with Alexa Together for remote caregiving, set reminders, and offer 24\/7 hands-free access to Urgent Response, an emergency helpline. Astro is known for its facial recognition and navigation skills and serves as a versatile home security device. It doesn’t have arms but features a cargo bin for carrying items and can distinguish between different family members. Its “face” is a screen that can stream video calls, play videos, and follow users around the home for centred interactions. It employs SLAM (simultaneous localization and mapping) technology for navigation. Despite being available, it’s important to note that the absence of Astro from prominent Amazon events and the departure of key personnel involved in its development might indicate a shift in focus or strategy regarding Amazon’s home robotics projects​. Matic Matic by Matic Robotics is designed to clean your home and comes equipped with WiFi, Bluetooth, a 1080p camera, and facial recognition. It can navigate your home and avoid obstacles and pets, capable of both vacuuming and mopping. It uses 3D mapping to navigate homes and adapts to its environment to prevent getting lost. Controlled via voice and gestures, Matic operates quietly and prioritises user privacy by processing data on the device itself. The company stated that these robots will be available to buy from March 2024. LG AI Agent LG’s new Smart Home AI Agent, unveiled as part of their “Zero Labor Home” vision, is an advanced robotic device equipped with AI and multi-modal technologies. Powered by the Qualcomm Robotics RB5 Platform, it features a camera, speaker, and environmental sensors for temperature, humidity, and air quality. The robot’s unique two-legged wheel design enables independent navigation around the home. This AI agent can engage in complex conversations, understand context and intentions through voice and image recognition, and natural language processing. It can control smart home appliances, act as a pet monitor and security guard, and even save energy by turning off unused devices. It also recognizes and responds to users’ emotions, provides personalised content like music, and offers updates on transportation, weather, and personal schedules. ORo Dog Companion The ORo Dog Companion robot is designed by Oro.Inc for dogs whose owners are away. It’s an autonomous robot providing physical and mental stimulation for dogs. ORo adapts and learns from a dog’s behaviour for personalised care. It employs algorithms for dog tracking and detection, ensuring accurate monitoring and real-time updates on your pet’s activities. ORo features audio and image analysis for monitoring a dog’s behaviour, health, and emotions, offering insights into their well-being. It updates continuously over-the-air, enhancing its capabilities. The robot includes an automatic pet feeder that dispenses food and medication, and monitors the dog’s food intake. It’s designed with safety and privacy, featuring gentle interactions, privacy controls, encryption, and out-of-bounds zones within your home. Loona Smart Robot Loona, the smart pet robot made by KEYI Tech is designed to interact with users and provide both companionship and entertainment. It’s equipped with various sensors, a 3D camera, and microphones, enabling it to respond to voice commands, recognize faces and objects, and navigate around the home. Loona can perform playful activities like fetching a ball, engaging in bullfighting games, and responding to hand gestures. It can also express emotions through its digital eyes and articulated ears. The robot’s design combines a retro-futuristic aesthetic with efficient mechanical engineering, allowing for a wide range of emotions and communicative abilities. For connectivity and programming, Loona integrates ChatGPT, allowing for engaging conversations and text-based games. It can also be used as a home security assistant, enabling remote monitoring through its camera. Loona is capable of self-charging and can return to its dock when low on power. It is available for the public on their website. Moxie Moxie, created by Embodied Inc., is a companion robot that serves as a tutor and mentor for children. It uses play-based learning to develop social, emotional, and cognitive skills. Moxie AI, the updated version, offers personalised academic learning in various subjects and allows for remote interaction and learning through the Moxie AI Digital app. Parents can use the companion app to monitor and guide their child’s development. Moxie features advanced conversational AI for natural interactions and supports multiple child profiles with face recognition technology to understand their emotional state. Being COPPA Safe Harbor certified, it is designed with children’s privacy and security in mind. Priced at $799, Moxie AI will be available for purchase in 2024 on MoxieRobot.com and Amazon.","excerpt":"Robots for your home now can work on advanced technology doing more than mechanical tasks.","categories":["Deep Tech"],"tags":["Astro","robot","Robotics"],"author_name":"K L Krithika","publish_date":"2024-01-18T15:00:00","publication_year":"2024","word_count":1026,"keywords":["Go","ChatGPT","AI","ETL","robot","ML","image recognition","Astro","Git","Robotics","GPT","generative AI","R"],"extracted_tech_keywords":["AI","ML","generative AI","ChatGPT","image recognition","R","Go","Git","ETL","GPT"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/7-must-have-home-robots-in-2024\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":46420,"title":"AI-Powered Platform Aquaconnect Raises $1.1 Million In Seed Round","content":"Aquaconnect this week announced that they had raised a seed round of $1.1 million from Omnivore and HATCH. An artificial intelligence-enabled platform, Aquaconnect integrates Asia’s leading network for aquaculture farmers with predictive SaaS tools for farm management and an omnichannel marketplace. Aquaconnect is currently focused on India’s $ 7.1 billion aquaculture sector, which produces almost 700,000 tons of farmed shrimp annually, and has become the top global exporter in recent years. Prior to this seed round, Aquaconnect received angel funding from HATCH, the world’s leading aquaculture focused startup accelerator, after participating in HATCH’s inaugural cohort in Bergen, Norway. This is Omnivore’s second investment in an aquaculture startup, following Eruvaka from its first fund in 2013. Based in Chennai, Aquaconnect was founded in 2017 by Rajamanohar “Raj” Somasundaram, Sanjai Kumar, and Shanmuga Sundara Raj. Raj is an IIT Kanpur alumnus, World Economic Forum’s Young Global Leader, and serial entrepreneur, who previously founded Hexolabs, a mobile technology startup that operated across Asia and Africa. Sanjai is an experienced aquaculture farmer who pioneered Vannamei culture in Tamil Nadu, while Shanmugaraj previously worked in leadership roles across the insurance industry, with ICICI Prudential, Canara HSBC, and Kotak Life. Currently, Aquaconnect has a network of over 3,000 shrimp farmers across India and Indonesia, which it plans to rapidly expand across South and Southeast Asia. These farmers receive support via Aquaconnect’s mobile application, FarmMOJO, which uses AI and predictive analytics to help them increase farm revenues, minimize costs, and reduce risks through better disease management. Finally, Aquaconnect links their farmer network to an omnichannel marketplace, where they can transact with feed producers, laboratories, equipment manufacturers, hatcheries, processors, exporters, certification bodies, banks, and insurance companies. With the current round of funding, Aquaconnect plans to accelerate the growth of its aquaculture farmer network, roll out deep tech improvements to FarmMOJO, launch new SaaS tools for the aquaculture ecosystem, and expand monetization across its omnichannel marketplace. Reihem Roy, who runs Omnivore’s Chennai office, will be joining Aquaconnect’s board to support the company’s growth strategy. According to Raj Somasundaram, Co-Founder & CEO of Aquaconnect, “The funding will be used for team expansion, enhancing our AI-enabled platform offering, and increasing our omnichannel marketplace fulfilment capabilities. We aim to reach 15,000 shrimp and fish farmers across India and Indonesia by December 2020. Omnivore and HATCH understand aquaculture, so they are perfect investors for our first institutional funding round.” The Government of India is ramping up support for the aquaculture sector, carving out a separate Ministry of Fisheries, Animal Husbandry, and Dairying as well as launching the Pradhan Mantri Matsya Sampada Yojana (PMMSY). We believe Raj and his team are building one of the most promising agritech startups globally, and we look forward to the exciting journey ahead.”","excerpt":"Aquaconnect this week announced that they had raised a seed round of $1.1 million from Omnivore and HATCH. An artificial intelligence-enabled platform, Aquaconnect integrates Asia’s leading network for aquaculture farmers with predictive SaaS tools for farm management and an omnichannel marketplace.  Aquaconnect is currently focused on India’s $ 7.1 billion aquaculture sector, which produces almost […]","categories":["AI News"],"tags":["SaaS"],"author_name":"Prajakta Hebbar","publish_date":"2019-09-26T12:55:00","publication_year":"2019","word_count":455,"keywords":["Go","API","funding","artificial intelligence","AI","R","Aim","SaaS","analytics","predictive analytics","startup"],"extracted_tech_keywords":["AI","artificial intelligence","analytics","Aim","predictive analytics","R","Go","API","startup","funding"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/ai-powered-platform-aquaconnect-raises-1-1-million-in-seed-round\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10082487,"title":"While the World Celebrates New Year&#8217;s Eve, Techies Slog to Run Web World","content":"Are we being fair and aware of each other? During this holiday period, we all want a break, especially the techies. Some might even say that it’s unfair for them to work so hard during this time of the year. But, if you ask the tech support and maintenance folks, they’ll tell you that it’s important to keep our systems running smoothly. In ecommerce, for instance, a significant percentage of employees who work during the holidays belong to an integral part of the production support and are an essential part of the enterprise that aids business operations. A production support team, typically, is responsible for the monitoring of the production servers, incident management, database management and ensuring that the softwares and servers are up and running smoothly without any disruptions. Analytics India Magazine spoke to multiple companies across sectors to understand if they are working on the holiday season, and the teams who have to keep the show running. One of the employees from Amazon said, “It is business as usual for those working in e-commerce like Amazon and Flipkart. While some of the divisions in these companies do go for a year end shutdown, we experience accelerated sales and demand in wake of the festive season and production support teams are expected to be in a firefighting mode to support the heavy user loads on their servers during the Christmas holidays as the availability of the site takes paramount importance”. For travel booking sites like Booking.com that experience massive usage, losing a day is often equal to losing a week in sales and thus imply that support has to be available at all times. Interestingly, several suggestions have been made that ecommerce companies witnessed a 67% increase in support tickets during the holidays. In other words, the teams working in these functions have to work. Holiday, no holidays, IT employees continue to work Indians like to slog. As winter vacations generally don’t apply to India, many teams in service-based companies have to work through this period. It is business as usual for those who work in service companies like TCS, Infosys, Wipro, and others. Even though their client companies have Christmas vacations, these teams continue to work to support business operations and complete the deliverables as required, learnt AIM. Meanwhile, the story for those working in development operations (aka DevOps) is not any different. Some of the teams in this domain have to work to support the company’s mission critical software systems and ensure product availability for clients 24*7. The team members usually operate in rotational shifts to ensure that support for the product’s operations is always available and, in this way, are able to take some time off but long holidays during festive seasons like Christmas don’t apply to critical teams. Other industries, such as telecom, also need their network operation centres to run smoothly throughout, thereby requiring instrumentation engineers to work during the holiday season. Although the luxury of not being able to avail long holidays may seem like a downer, there is surely a bright side to it. Benefits of working during holidays Let’s look at the bright side of working during holidays, the year-end shut-down in the majority of companies also helps the infrastructure teams to apply patches and perform maintenance activities. In addition, if a client is working during the holiday season, the project teams are required to work and support the client. Interestingly, some of the companies offer overtime compensation and added perks to carry forward leaves for next year, which might not really be a bad idea. Let us know what you think, and your plans for Christmas and New year!","excerpt":"Techies in production support teams in e-commerce companies are expected to be in a firefighting mode to support the heavy user loads on their servers during the Christmas holidays as the availability of the site takes paramount importance.","categories":["AI Features"],"tags":[],"author_name":"Aparna Iyer","publish_date":"2022-12-16T12:00:00","publication_year":"2022","word_count":609,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","Aim","ViT","analytics","disruption","DevOps","R"],"extracted_tech_keywords":["AI","analytics","Aim","R","Go","DevOps","ViT","disruption","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/while-the-world-celebrates-new-years-techies-slog-to-run-web-world\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10161573,"title":"IBM &amp; L&#8217;Oréal Build AI for Sustainable Cosmetics","content":"IBM, the American tech company, and L’Oréal, a leading beauty & cosmetics company, have collaborated to use AI to advance sustainable cosmetics. IBM’s AI technology and infrastructure will be used to ‘uncover new insights in cosmetic formulation data, facilitating L’Oréal’s use of sustainable raw materials for energy and material waste reductions.’ The announcement also claims that this is the first AI model intended to advance sustainable cosmetics. Furthermore, the partnership is said to help L’Oréal meet its target of sourcing most of its product formulas from bio-sourced materials by 2030. The AI model will use a ‘large number of formulations’ and data points to accelerate the product formulation process, including re-formulation and production of existing products. These tools are also said to enable 4,000 researchers at LO’real to build eco-friendly cosmetics. “As part of our Digital Transformation Program, this partnership will extend the speed and scale of our innovation and reformulation pipeline, with products always reaching higher standards of inclusivity, sustainability, and personalisation,” said Stephane Ortiz, Head of Innovation Métiers & Product Development, at L’Oréal Research & Innovation. This is the second time in 2025 that L’Oréal has announced a new development at the intersection of cosmetics and AI. At the Consumer Electronics Show (CES) 2025, the company unveiled a device called Cell Bioprint that analyses skin protein levels and delivers insights to improve skin health and ageing. The device was developed in collaboration with Korean biotech company NanoEnTek. Having said that, L’Oréal has been exploring groundbreaking technologies for a long time. In 2018, L’Oréal acquired a Canadian company called Modiface to use AI and augmented reality (AR) for skin diagnostics and assessments for customers, based on ‘15 years of scientific research on skin ageing’. ModiFace also helped the company integrate AI-enabled product try-ons and virtual makeup on Amazon. That said, CES 2025 featured an array of beauty-tech announcements. Notably, Samsung, the South Korean tech giant, introduced a new product called the MICRO LED beauty mirror. This mirror uses AI to provide personalised beauty tips based on skin analysis.","excerpt":"‘This alliance between highly specialised expertise in artificial intelligence and cosmetics seeks to revolutionise cosmetic formulation’","categories":["AI News"],"tags":["AI (Artificial Intelligence)","IBM"],"author_name":"Supreeth Koundinya","publish_date":"2025-01-16T17:52:43","publication_year":"2025","word_count":339,"keywords":["programming_languages:R","AI","innovation","digital transformation","Git","Ray","Aim","ViT","IBM","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","Aim","Ray","R","Git","ViT","digital transformation","innovation","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/ibm-loreal-build-ai-for-sustainable-cosmetics\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10103522,"title":"&#8216;Big Tech&#8217;s AI Regulation Talk Doesn&#8217;t Match Their Actions&#8217;","content":"“One thing that has been annoying me is the myth that big tech companies are the main voices calling for regulation,” Sören Mindermann recently told AIM in an interview. Mindermann is currently a postdoc with computer scientist Yoshua Bengio at MILA working on AI safety. Even though he just finished his PhD at University of Oxford, in machine learning he wants to stay focused on AI safety and risk. “They’re just getting the most attention. But many big tech firms like Meta and IBM are denying risks and lobbying against regulation with a clever lie. They pretend that it’s only other companies who call for regulation, calling it `regulatory capture’. There’s actually an emerging academic consensus, calling for regulation and acknowledging real risks.” he stated. The AI researcher wrote his first paper on AI safety seven years ago. “I’ve had detours into scaling, deep learning, and statistical modelling for COVID, but my focus has always been on safety. Suddenly it is becoming such a big deal and a bit sooner than I even expected it to be. I thought we’re gonna need a lot of time to prepare for these problems,” he said. A month ago, Mindermann published a paper alongside 22 academic co-authors from US, China, EU, and UK— including Geoffrey Hinton, Stuart Russell and Bengio. The AI insiders called for immediate action, proposing that companies working on AI systems allocate at least one-third of their resources to ensure AI safety and ethical use. “This paper started with us noticing that there are many AI academics, including the most cited people in the field, who are worried about the risks and the technologies it’s posing,” said Mindermann. As of now he is focused on AI honesty projects. “People aren’t always going to be able to tell if what the AI says is true. So, we developed a lie detector for language models that can tell with reasonably high accuracy, whether AI output is true or not,” he revealed. Lack of Focus, Knowledge and Researchers Mindermann knows the safety teams working at Google and OpenAI. “The last time I checked, they were a tiny part of their overall research teams talking to one little safety team,” he pointed out. The 2023 State of AI report mentioned the same in numbers. As per the report, Google DeepMind has the largest and most established AI alignment team of 40 members led by co-founder Shane Legg. In comparison, OpenAI has a team of 11 members, and its rival startup Anthropic has only 10. But the companies are not to be blamed solely for the sad state of affairs. “All the companies want to stay ahead of the others, cut corners on safety and make profits from AI while letting society deal with risks. That’s why the governments need to intervene. In addition to the competition we have a lack of awareness of the risks among AI developers,” noted Mindermann. He further elaborated that it’s not a part of the job description of an AI researcher to understand the risks AI poses. “No one really knows what AI is causing in sensitive domains and the regulation so far is reactive after something bad happens. It could turn out similar to Chernobyl where after a big accident happened, the nuclear industry was largely shut down. Some AI companies are calling for regulation partly because they don’t want something to happen to the AI industry,” he mentioned. Not Keeping Up with the Pace “Regulation is central but it is too slow considering the rate at which AI is progressing,” the AI researcher suggested. Similar thoughts echoed in the AI community eight months ago when thousands of business and AI leaders signed an open letter calling for a six-month pause on the training of AI systems more powerful than OpenAI’s GPT-4. While the call was not implemented, it was not considered a failure either because AI safety finally made it to the public agenda. Mindermann suggested that we need some immediate detailed commitments from companies before they train the next generation of AI systems. “If they have a level of dangerous capability the governments will be able to evaluate then the companies can commit to the safety measure, including not deploying the system or not developing it any further if they haven’t got the safeguards ready, ” he concluded.","excerpt":"Sören Mindermann is currently a postdoc with computer scientist Yoshua Bengio at MILA, working on AI safety.","categories":["AI Features"],"tags":["Interviews and Discussions"],"author_name":"Tasmia Ansari","publish_date":"2023-11-22T16:57:32","publication_year":"2023","word_count":722,"keywords":["Anthropic","Go","machine learning","TPU","OpenAI","AI","GPT","Aim","deep learning","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","machine learning","deep learning","OpenAI","Anthropic","Aim","TPU","R","Go","GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/big-techs-ai-regulation-talk-doesnt-match-their-actions-says-researcher\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10017082,"title":"Upcoming AI Conferences To Look Forward To In 2021","content":"In the year 2020, one thing that turned out to have a massive impact on our daily lives and society is artificial intelligence. The Government of Telangana has even declared 2020 as the year of artificial intelligence. Starting from GPT-3 and improvements in health-tech to a conversation on ethical AI and advancements in neural networks, the year has seen it all. And this is the time when businesses are going to come out and talk about their contribution to the field of AI, as well as research and developments around it. With the starting of this new year, we have come up with a list of upcoming AI conferences that one can attend in 2021, to keep themselves at the forefront of this technology. Also Read: Top Data Science & AI Trends To Watch Out For In 2021 From Privacy to Fairness in AI Date: 09th January 2021 ‘From Privacy to Fairness in AI’ is an AI workshop that has been held by the Association of Data Scientists to empower AI\/ML researchers and industry leaders with an understanding of the risks associated with the field of artificial intelligence. With responsible AI becoming a critical concern and a priority for the industry, this workshop cannot come at a better time. The workshop will provide a platform for attendees and industry experts to explore innovative ideas and discuss some of the open problems related to privacy, ethics, and fairness of artificial intelligence. Some of the key topics covered in the workshop include an introduction to privacy-preserving AI and ML and the current challenges associated with it; understanding AI’s ethics and creating fairness in AI, among others. Know more about it here. Also Read: Join This Full-Day Workshop On Privacy & Fairness In AI MLDS 2021 Dates: 11th – 13th February 2021 Brought to you by Analytics India Magazine — Machine Learning Developers Summit 2021 (MLDS 2021) is an AI conference that strictly focuses on bringing the machine learning community together from across the globe. With full-day workshops by ML experts to technical talks by a wide range of industry professionals, this AI conference will discuss various exciting innovation and developments of AI that have shaped the industry in 2020. The conference brings in prominent names like Dipanjan Sarkar, Google Developer Expert – ML; Kaggle Grandmaster, Martin Henze; data science and analytics head of PayU and Daimler AG, to name a few. It is expecting a digital footfall of 1500 ML developers from 200 organisations. Know more about it here. Also Read: Top 8 Things Developers Can Look Forward To At MLDS 2021 Deep Learning 2.0 Dates: 28th – 29th January 2021 A two-day virtual summit — Deep Learning 2.0 will focus on exploring the latest AI applications as well as discussions around ways to bridge the gap between the latest technological research advancements and real-world applications. With five different tracks that include — deep learning landscape stage; enterprise AI stage; ethics and social responsibility stage; generative model stage; and reinforcement learning stage, this conference covers it all. The conference welcomes speakers from the University of Oxford, Salesforce Research, Facebook AI Research, UC Berkeley, Google Brain, and Princeton University, to name a few. Know more about it here. AAAI Conference on Artificial Intelligence Dates: 2nd – 9th February 2021 This year is the 35th AAAI Conference on Artificial Intelligence that will be held virtually and will strongly focus on promoting research in the field of artificial intelligence and scientific exchange among AI researchers, practitioners, and engineers. With diverse tracks, this conference will include technical talks by prominent industry experts, workshops by AI practitioners and will host competition programmes to promote artificial intelligence to students and youngsters. The conference has already announced their 2021 Elected Fellows in a recent tweet, who have contributed to AI, reinforcement learning, robotics, NLP, computer vision etc. Know more about it here. Ai4 2021 — Cybersecurity Summit Dates: 3rd – 4th Feb 2021 Ai4 2021 — Cybersecurity Summit is a two-day virtual conference that will explore artificial intelligence in cybersecurity. The conference will bring together business leaders, AI experts and data practitioners to promote the adoption of artificial intelligence and machine learning technologies. With a use-case oriented approach, the conference will welcome 30+ speakers from some of the prominent institutions and organisations covering topics like advanced AI threats, AI-based vulnerability management, orchestration and response etc. Know more about it here. Future Of AI Date: 16th March 2021 Claiming to be one of the leading conferences on AI and big data, Future of AI is a one-day virtual conference that has been designed to connect members, technology leaders and business experts in the field of AI. With multiple tracks, 104 sessions and 168 speakers, this conference showcases the latest innovations, business models in AI, startup exchange ideas and developers’ talks. It will cover topics such as the potential of unstructured data; the power of AI through modern data stack; leveraging deep learning to engage Digital customers, ethics of AI, machine learning and AI models, to name a few. Know more about it here. AI & Big Data Expo — Global 2021 Dates: 17th – 18th March 2021 An AI conference for the entire global AI and big data ecosystem, AI & Big Data Expo is an entirely free virtual event that focuses on top-level content and thought leadership discussion on the field of artificial intelligence. The conference will bring together 40+ speakers, including prominent names from American Express, H&M, M&G Investments and Nissan, who will explore this innovative technology and discuss its impact on their respective industries. This conference can be attended by IT decision-makers, developers, innovators, CIOs, data scientists, brand managers, and many more. The conference will also host debates on artificial intelligence and big data advancements, and cover topics like business intelligence, deep learning, machine learning, AI algorithms, chatbots, etc. Know more about it here. IntelliSys 2021 Dates: 2nd – 3rd September 2021 IntelliSys 2021, or also known as, Intelligent Systems Conference is a virtual conference that will focus on areas of intelligent systems and application of artificial intelligence in the real world. The conference provides a platform for researchers and practitioners from diverse fields to exchange ideas and discuss recent developments related to AI. Additionally, it includes industry experts’ talks on state-of-the-art computer science and lectures by eminent scientists about their recent innovations. Divided into four tracks, the key topics covered in the conference are artificial intelligence, machine vision, robotics, and ambient intelligence. Know more about it here.","excerpt":"In the year 2020, one thing that turned out to have a massive impact on our daily lives and society is artificial intelligence. The Government of Telangana has even declared 2020 as the year of artificial intelligence. Starting from GPT-3 and improvements in health-tech to a conversation on ethical AI and advancements in neural networks, […]","categories":["IT Services"],"tags":["ai conferences","data science machine learning ai","latest technological advancements"],"author_name":"Sejuti Das","publish_date":"2021-01-05T11:00:00","publication_year":"2021","word_count":1081,"keywords":["data science","ai conferences","artificial intelligence","machine learning","AI","neural network","ML","data science machine learning ai","computer vision","NLP","deep learning","latest technological advancements","analytics"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","neural network","NLP","computer vision","data science","analytics"],"url":"https:\/\/analyticsindiamag.com\/it-services\/upcoming-ai-conferences-to-look-forward-to-in-2021\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10129413,"title":"Google DeepMind Launches MatFormer Framework to Improve On-Device AI Capabilities","content":"Google Pay GM and vice president Ambarish Kenghe announced the introduction of the MatFormer framework, which he said would “further enhance our on-device capabilities”. “With MatFormer, you can mix and match AI models within a single framework to use the right sized model for your task. This gives you the best of both worlds – high performance and low resource consumption,” Kenghe said during Google I\/O Connect in Bengaluru today. With the announcement, the framework is officially available to developers on GitHub. As Kenghe said, the framework is designed specifically to improve on-device capabilities. As of now, Android has a built-in foundational model, Gemma Nano, designed specifically for mobile devices, the first mobile OS to have a built-in model. The model was developed to focus on privacy and delivering AI results on unreliable and unstable networks. With the addition of the framework, developers now have the freedom to inculcate features from several Gemini models under a single framework in order to develop an on-device model that works best for them, or for their specific use cases. “This will translate to smoother, faster, and more accurate AI experiences directly on users’ phones,” Kengre said. Like Google’s Composition to Augment Language Models (CALM) framework and other products launched at the event, the MatFormer framework was also developed by India’s DeepMind team, according to Kenghe. While this is the first framework of its kind, the focus on on-device integrated AI has been mirrored by other AI companies. Earlier this year, Qualcomm Technologies had launched their Qualcomm AI Hub, allowing developers to run AI models on their respective devices. However, this was limited to allowing developers to build apps using AI, by offering optimised AI models through the hub.","excerpt":"Like Google’s Composition to Augment Language Models (CALM) framework and other products launched at the event, the MatFormer framework was also developed by India’s DeepMind team.","categories":["AI News"],"tags":["AI Impacts","DeepMind"],"author_name":"Donna Eva","publish_date":"2024-07-17T16:30:08","publication_year":"2024","word_count":285,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","Git","llm_models:Gemini","AI Impacts","GitHub","R","DeepMind"],"extracted_tech_keywords":["AI","R","Go","Git","GitHub","llm_models:Gemini","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/google-deepmind-launches-matformer-framework-to-improve-on-device-ai-capabilities\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":16042,"title":"ZineOne aims to enhance its AI and ML capabilities with its $2.5 million funding amount","content":"Another startup in the industry gets funded, as US based ZineOne secures close to $2.5 million from the likes of New York based Seeds Angels and Hyderabad Angels along with other angel investors. In what may seem an excellent opportunity for this AI powered, events driven interactions platform to use the funds to enhance its offerings, ZineOne aims to do exactly that. If reports are to be believed, the company is looking make an enhanced offering in the field of machine learning and artificial intelligence and hence boost their customer experience across digital channels and devices. Founded in 2013 by Debjani Deb, Manish Malhotra and Arnab Mukherjee, ZineOne provides an event-driven brand-to-user interactions, through its real-time stream processing platform powered by machine learning. It lets brands interact with customers contextually across channels. HDFC, BioPharmX, Strava are some of it major clients. With offices across India as well, with its big data technology stack, ZineOne aims to process its customer activity data and provide sub-second responses. the company also named Venkat Rama Mohan Reddy, Hyderabad Angels’ advisory board member, as a director, who is also a founder and executive chairman of IT and engineering services company Cyient Ltd. He had also served NASSCOM as its chairman. As the field of artificial intelligence is witnessing a growing popularity, many investors have shown their interest in backing up the new ideas in the field of AI and machine learning. Earlier this month, Analytics India Magazine had reported the e commerce firm with its AI based assistant, Niki.ai, receiving a $2 million Series A funding from San Francisco-based SAP.iO, and existing investor Unilazer Ventures. Julia, one of the fastest high performance open source computing language for machine learning, AI and data analytics jobs, got seed funded from General Catalyst and Founder of Collective investors. Unbxd also raised $12.5 million funding aiming to invest in AI engine to boost product discovery.","excerpt":"Another startup in the industry gets funded, as US based ZineOne secures close to $2.5 million from the likes of New York based Seeds Angels and Hyderabad Angels along with other angel investors. In what may seem an excellent opportunity for this AI powered, events driven interactions platform to use the funds to enhance its […]","categories":["AI News"],"tags":["ai funding india"],"author_name":"Srishti Deoras","publish_date":"2017-07-03T09:46:38","publication_year":"2017","word_count":316,"keywords":["big data","Go","machine learning","artificial intelligence","AI","Git","Aim","analytics","Julia","R","ai funding india"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","analytics","Aim","R","Go","Julia","Git","big data"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/zineone-aims-enhance-ai-ml-capabilities-2-5-million-funding-amount\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10133094,"title":"Infosys Introduces INR 9 Lakh Salary Packages in New ‘Power Programme’ for Freshers","content":"Indian IT firm, Infosys has rolled out a new ‘power programme’ for campus placements, offering pay packages of up to Rs 9 lakh per annum, significantly higher than its standard entry-level salary of Rs 3-3.5 lakh, according to a recent report by the Economic Times. This initiative, which targets candidates with strong coding, software development, and programming skills, reflects the IT major’s shift towards differential hiring amid a growing demand for specialised talent. The report added that the salary bands under this new programme range between Rs 4-6.5 lakh and Rs 9 lakh, depending on the candidate’s expertise. The programme mirrors a similar initiative by Tata Consultancy Services (TCS) called ‘Prime,’ which offers salaries between Rs 9-11 lakh per annum for freshers in software development roles. TCS’s hiring strategy also includes roles under its ‘Ninja’ and ‘Digital’ categories, offering packages of Rs 3.6 lakh and Rs 7.5 lakh per annum, respectively. As the IT services sector navigates a subdued demand environment, Infosys and its peers are increasingly focusing on hiring talent skilled in cloud computing, artificial intelligence, and cybersecurity to drive digital transformation projects. Infosys CFO Jayesh Sanghrajka highlighted the company’s agile hiring approach, which now includes both campus and off-campus recruitment. This year has been particularly tough for IT employees with companies declining to hire freshers. Well, those working at the organisation are also not having the best time of their lives. The entire IT sector witnessed a sharp decline in hike growth in 2023, dropping to 9.1% from 10.3% in 2022. Excluding products, the IT\/ITeS sector saw the lowest average salary hike at just 8.4%. In 2021, the average salary hike was 8.8%, which increased to 9.7% in 2022. Last year, the sector saw average increments between 8.5% and 9.1%. IT companies are offering single-digit salary increments to employees who lack AI-related skills. Further, there is an abundance of entry-level talent who are now equipping themselves with the latest tech and tools. Additionally, rising business costs, coupled with layoffs are making this trend continue this year too. Despite the overall slowdown in the technology services sector, which saw the top five Indian IT companies reduce their headcount by over 70,000 in fiscal 2024, Infosys plans to hire 15,000-20,000 graduates in FY25. TCS is set to recruit around 40,000 freshers, consistent with last year’s numbers, while HCLTech and Wipro are targeting 10,000 and up to 12,000 fresh hires, respectively. Infosys, which reported a net headcount reduction of nearly 2,000 employees in the April to June quarter, is set to resume campus placements this fiscal year, signaling a renewed focus on fresh talent acquisition. The company’s total employee base stood at 315,332 as of June 2024. Meanwhile, TCS, after three consecutive quarters of net headcount reduction, saw an addition of 5,452 employees in the June quarter.","excerpt":"This initiative, which targets candidates with strong coding, software development, and programming skills.","categories":["AI News"],"tags":["Infosys"],"author_name":"Siddharth Jindal","publish_date":"2024-08-20T10:18:51","publication_year":"2024","word_count":466,"keywords":["Go","artificial intelligence","Infosys","cloud computing","AI","programming_languages:R","digital transformation","Git","RAG","GAN","R"],"extracted_tech_keywords":["AI","artificial intelligence","RAG","cloud computing","R","Go","Git","GAN","digital transformation","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/infosys-introduces-inr-9-lakh-salary-packages-in-new-power-programme-for-freshers\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10013820,"title":"Why These Tech Giants Are Releasing ML Based Time Series Solutions","content":"According to the Mckinsey report, in the consumer goods domain, improving the accuracy of demand forecasting by 10-20% can reduce inventory by 5% and increase revenue by 2-3%. Current ML-based forecasting solutions require significant manual effort, including model construction, feature engineering and hyper-parameter tuning. Last week, Google and Facebook especially, have come up with two new frameworks for solving time series problems with great ease. While Google took the AutoML route, Facebook AI spun an architecture out of their popular Prophet tool. In this article, we take a look at those two solutions which underline the growing importance of deep learning-based time series solutions. Google’s AutoML Solution “The AutoML solution required moderate compute cost, only 500 CPUs for 2 hours to be at the top of the Kaggle competition.” Google’s AI team have introduced a scalable end-to-end AutoML solution for time series forecasting. However, a fully automated generic solution is challenging as the solution should work for different datasets, which belong to different domains and have their own demands(hourly, daily or weekly). This end-to-end pipeline is built using TensorFlow with a specialised search space for time series forecasting. The solution is based on an encoder-decoder architecture. The encoder transforms the historical information in a time series into a set of vectors, and the decoder generates future predictions based on these vectors. The resulting AutoML solution searches for the best combination of components such as attention dilated convolution, gating and skip connections to make the decision. To test the accuracy of their AutoML solution, Google team tested their solution in the popular M5 forecasting competition, with a long history spanning nearly 40 years. The fully automated solution achieved a rank of 138 out of 5558 participants (top 2.5%) on the Kaggle. According to Google, other top forecasting models required months of manual effort to create, whereas Google’s AutoML solution needed 500 CPUs running for two hours to finish the competition. Time Series Forecasting On NSE Of India Using Facebook’s NeuralProphetGoogle Introduces Cloud AutoML In A Bid To Democratise Artificial IntelligenceAutoML Is Functional But Have Limitations That We Cannot IgnoreWhat Are The Limitations Of AutoMLThings To Consider Before Building A Computer Vision Model Using AutoML Facebook’s NeuralProphet Solution A week ago, Facebook also announced the release of NeuralProphet, a Neural Network based PyTorch implementation of time series forecasting tool, inspired by the popular forecasting tool Prophet. The NeuralProphet documentation states that it is developed in a fully modular architecture and is flexible to take in any additional components in the future. The developers wrote that their vision is to develop a simple to use forecasting tool for users while ensuring interpretability, configurability and providing much more such as the automatic differencing capabilities by using PyTorch as the backend. NeuralProphet consists of components like seasonality, auto-regression, special events, future regressors and lagged regressors. For instance, seasonality is modelled using Fourier terms and can handle multiple seasonalities for high-frequency data. Auto-regression is handled using an implementation of an Auto-Regressive Feed-Forward Neural Network for time series. According to Facebook, NeuralProphet can be used to build forecasting models which are driven by other external factors that dictate the behaviour of the target series over time. External information can heavily improve forecasting models as they don’t rely only on the autocorrelation of the series. NeuralProphet tool is the right fit for those who wish to gain insights into the overall modelling process by visualising the forecasts, the individual components as well as the underlying coefficients of the model. According to the team, users can visualise the interaction of the individual components. They also have the power to control these coefficients as required by introducing sparsity through regularisation. They can combine the components additively or multiplicatively as per their domain knowledge. NeuralProphet has the following features: Gradient Descent for optimisation via using PyTorch as the backend.Modelling autocorrelation of time series using AR-NetModelling lagged regressors using a separate Feed-Forward Neural Network.Configurable nonlinear deep layers of the FFNNs.Tuneable to specific forecast horizons.Custom losses and metrics.","excerpt":"According to the Mckinsey report, in the consumer goods domain, improving the accuracy of demand forecasting by 10-20% can reduce inventory by 5% and increase revenue by 2-3%. Current ML-based forecasting solutions require significant manual effort, including model construction, feature engineering and hyper-parameter tuning.  Last week, Google and Facebook especially, have come up with two […]","categories":["AI Trends"],"tags":["Automl","Time Series"],"author_name":"Ram Sagar","publish_date":"2020-12-10T11:00:00","publication_year":"2020","word_count":665,"keywords":["Go","Automl","artificial intelligence","AI","neural network","PyTorch","ML","computer vision","deep learning","Time Series","TensorFlow","R"],"extracted_tech_keywords":["AI","artificial intelligence","ML","deep learning","neural network","computer vision","TensorFlow","PyTorch","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/google-facebook-automl-prophet-time-series\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":5633,"title":"Interview &#8211; Dr. Bart Baesens &#8211; Professor at KU Leuven, Belgium","content":"Dr. Bart Baesens has recently published a book “Analytics in a Big Data World: The Essential Guide to Data Science and its Applications”. The book discusses how analytics can be used to create strategic leverage and identify new business opportunities. The focus of this book is not on the mathematics or theory, but on the practical application. Formulas and equations will only be included when absolutely needed from a practitioner’s perspective. It is also not our aim to provide exhaustive coverage of all analytical techniques previously developed, but rather to cover the ones that really provide added value in a business setting. It is available at the following link- http:\/\/as.wiley.com\/WileyCDA\/WileyTitle\/productCd-1118892704.html Dr Baesens talk to us more about his book. [divider] [dropcap size=”2″]AIM[\/dropcap]Analytics India Magazine: How did you decide to write a book on Business Analytics? [dropcap size=”2″]BB[\/dropcap]Prof. Dr. Bart Baesens: I wanted to write a book which is relevant to decisions that all businesses will need to make in the coming years. As the number of practical applications for data skyrockets, learning how to extract business value from big data becomes a competitive requirement.  Big data sets are assets that can be leveraged quickly and inexpensively, if tackled wisely! My book Analytics in a Big Data World addresses this seemingly Herculean task of coming to grips with multiple channels of data and sculpting them into quantifiable value. This book is for business professionals who want a focused, practical approach to big and data analytics. I hereby focus on case studies, real-world application, and steps for implementation, using theory and mathematical formulas only when strictly necessary! AIM: How is this book different from other similar books in market today? BB: The book provides a comprehensive, end-to-end process overview of how to put analytics to work to solve concrete business problems.  Many current text books only focus on discussing various predictive techniques with too much theoretical focus, hereby losing the general picture.  Building upon my experience in both academia and business, the book provides a unique blend of a state of the art, scientific approach and a clear practitioner focus.  The book also has plenty of tricks and tips on how to successfully apply analytics and includes (performance) benchmarks in a variety of different business contexts (e.g. credit risk management, marketing, fraud detection, web analytics). AIM: How did you start your career in analytics? BB: After my MSc studies in Business Engineering at KU Leuven (Belgium), I started a PhD which I completed in 2003 with the title “Developing Intelligent Systems for Credit Scoring using Machine Learning techniques”.  Soon afterwards I got hired as an assistant professor at KU Leuven (Belgium) and guest lecturer at the University of Southampton (United Kingdom).  I continued my research on analytics for credit scoring and at the same time expanded my focus towards other application areas such as marketing analytics (response and retention modeling), fraud analytics, and web analytics. I currently have a research team of about 10 PhD students and 2 postdocs working on these topics. See www.dataminingapps.com for an overview of our activities.  Next to my academic research, I regularly tutor, advise and provide consulting support to international firms with respect to their analytics strategy.  I also currently teach 4 courses on the topic of analytics to businesses, which are offered both in classroom as well as webinar format. AIM: What did it take to have a book published? BB: Well, throughout the past few years, I have done lots of (strategic) consulting for many firms across the globe. Throughout my experience and activities, I learned that a first key step to fully leverage and unleash the power of analytics, is to create awareness about the topic with all the business stakeholders involved across all levels of decision making. To my biggest frustration, I found no adequate background material to help me achieve this.  Hence, this inspired me to write this book whereby my key aim is to bridge the gap between analytics and the business! AIM: Are you working on any other book right now? BB: Yes, in fact, I am working on two more books at the moment. The first one is entitled: “Predictive Analytics: Techniques and Applications in Credit Risk Modelling” to be published by Oxford University Press in 2015.  It is the second book out of a series of three and it goes into much detail about how to construct, backtest, benchmark and stress test high performing analytical models for credit risk assessment. The second book I am working on is in a totally different area.  It is entitled Basic Programming in Java and will also be published by Wiley.  It is based upon a postgraduate course I teach and the experiences we gained by using Java in our analytical research. AIM: What do you suggest to new graduates aspiring to get into analytics space? BB: First of all, I would strongly encourage graduates to pursue a career in analytics. According to a recent McKinsey report, the US alone faces a shortage of 140,000 to 190,000 people with deep analytical skills and another 1,5 million managers capable of making decisions based on big data and analytics.  On top of that, it is a fascinating world with lots of new developments and challenges.  In order to be a good analyst, one needs to have a multidisciplinary profile.  Hence, I believe graduates should first have a sound and solid quantitative background in statistics. Next, they should also make sure they possess deep business knowledge and communication\/presentation skills.  Especially the latter are also very important in order to bridge the often observed communication gap between the business and the analyst which we talked about earlier. AIM: How do you see Analytics evolving today in the industry as a whole? What are the most important contemporary trends that you see emerging in the Analytics space across the globe? BB: Well, let me discuss some trends which I consider important based upon both my industry and research experience.  First of all, analytics is about being actionable and simple.  It’s not about complex numbers, black box models or statistics.  In our collaborations with firms, we have found that simple analytical models (e.g. regression models, decision trees) typically perform well in many settings. Hence, the best investment firms can make to boost the performance of their analytical models is by investing in data and improving data quality.  That’s why in my book I also devoted a whole section to this topic. From a technical perspective, next to the analytical models themselves, firms should also thoroughly consider how to appropriately monitor, backtest and integrate these models with other (e.g. marketing, risk management) applications.  I believe these activities currently pose quite a bit of challenges for which more practice-focused research is necessary! From a software perspective, an important trend is the emergence of open source solutions.  However, at this stage, many of these solutions are not scalable and focus too much on the technical aspects of analytics instead of providing solutions to business problems. Finally, data and analytics is everywhere and all around.  It speaks for itself that this creates huge challenges from a privacy perspective.   Firstly, data about individuals can be collected without these individuals being aware about it. Secondly, people may be aware that data is collected about them, but have no say in how the data is being analyzed and used.  Hence, regulatory authorities have to think about new regulations, whereas researchers should focus more on the development of privacy friendly analytical techniques. AIM: Anything else you wish to add? BB: I really hope people enjoy reading my book, and I very much look forward to receiving their feedback and hearing about their experiences in the wonderful world of analytics! [divider] [spoiler title=”Biography of Prof. Dr. Bart Baesens” style=”fancy” icon=”plus-circle”] Professor Bart Baesens is a professor at KU Leuven (Belgium), and a lecturer at the University of Southampton (United Kingdom).  He has done extensive research on analytics, customer relationship management, web analytics, fraud detection, and credit risk management.  His findings have been published in well-known international journals (e.g. Machine Learning, Management Science, IEEE Transactions on Neural Networks, IEEE Transactions on Knowledge and Data Engineering, IEEE Transactions on Evolutionary Computation, Journal of Machine Learning Research, …) and presented at international top conferences. He is also author of the books Credit Risk Management: Basic Concepts, published by Oxford University Press in 2008; and Analytics in a Big Data World published by Wiley in 2014.  His research is summarized at www.dataminingapps.com.  He also regularly tutors, advises and provides consulting support to international firms with respect to their analytics and credit risk management strategy.[\/spoiler]","excerpt":"Dr. Bart Baesens has recently published a book “Analytics in a Big Data World: The Essential Guide to Data Science and its Applications”. The book discusses how analytics can be used to create strategic leverage and identify new business opportunities. The focus of this book is not on the mathematics or theory, but on the […]","categories":["AI Features"],"tags":["Interviews and Discussions"],"author_name":"AIM Media House","publish_date":"2014-04-21T15:50:03","publication_year":"2014","word_count":1434,"keywords":["data science","machine learning","AI","neural network","ML","RAG","Aim","analytics","fraud detection","predictive analytics","Interviews and Discussions"],"extracted_tech_keywords":["AI","machine learning","ML","neural network","data science","analytics","Aim","RAG","predictive analytics","fraud detection"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/interview-dr-bart-baesens-professor-at-ku-leuven-belgium\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":26618,"title":"Top 5 International Professional Artificial Intelligence Bodies","content":"As concerns around artificial intelligence increase — from a shortage of skilled professionals to ethics — its impact has gone well beyond technology. Governments and professional bodies have realised that the reality of AI is no longer just about technology anymore. There are social, philosophical and ethical aspects of AI and machine learning that companies are grappling with in their everyday situations. There is a need for digital ethics to drive good governance, set a norm and implement change. In this article, we list down top professional bodies across the globe who are collaborating with the industry, academia, the startup community and the civil society to shape the development of AI for the benefit of humanity. Swedish AI Society: The Swedish Artificial Intelligence Society, SAIS, is a society promoting research and application of Artificial Intelligence. The SIAS was founded in 1982 with an aim to promote AI interests in national and international contexts. The organisation regularly organises annual conferences, supports AI training courses, workshops and conferences in fields related to AI. Right now, SAIS has announced AI Master’s Thesis Award 2018 and is calling for nominations for best thesis in AI. Association for Computing Machinery: This international society for computing has an India chapter as well which aims at increasing the CS and information technology community by facilitating high-level conferences, workshops, mentorship opportunities, serving as professional networks, encouraging students to take a higher level of interest in the computing. ACM-India is also at the forefront of hackathons to foster greater interest and participation from students. European Coordinating Committee for Artificial Intelligence: Set up in 1982, the European Coordinating Committee for AI promotes research, study and applications of AI in Europe. With an aim to foster innovation and collaboration across Europe, EurAI, as it is known encompasses other member organisations from Spain, Catalon, Ukraine, France, Ireland, Portugal, Romania, Bulgaria, Denmark, Greece, Latvia, Norway and other European nations. EurAI, is also known for organising one of the leading conferences in Europe along with a member organisation, known as ECAI. The body also organises Advanced courses in AI, known as ACAI, in a specialised domain. For example, it recently hosted an ACAI Summer School on Statistical Relational AI in Ferrara, Italy. International Neural Network Society: Better known as INNS, this is a top organisation for professionals who are interested in a theoretical and computational understanding of the brain and applying the knowledge to develop new and more effective forms of machine intelligence. The non-profit, scientific organisation was founded in 1987 and has played a pivotal role in advancing the field of neural networks. Its conferences, events, journal, special interest groups (SIGs), Regional Chapters (RC) help push forward research in the field of neural networks. International Society for Applied AI: This non-profit organisation publishes the International Journal of Applied Intelligence besides sponsoring the annual International Conference on Industrial, Engineering & Other Applications of Applied Intelligent Systems (IEA\/AIE). Based in Texas, this non-profit organisation also works alongside other professional associations to improve application in research and improve the transfer of technology from the academic world to the industry. Outlook The growing interest around AI and machine learning has sparked a debate on the ethics. While leading tech companies and industry leaders in AI research and development are working on integrating AI into their current suite of products and services. While tech companies R&D arms such as Microsoft Research, Facebook Research, Google Inc are busy moving innovation from lab to market, little is done to broaden the scope of research or encouraging more students and professionals. Non-profit organisations invest more deeply in building promising AI technologies and building AI technologies that are safe for the next generation. In the future, AI will affect the nature and distribution of jobs and these non-profit bodies bring together a cross-section of industry leaders such as robotics, computing. Even though there has been recent AI wins in terms of self-driving cars, DeepMind’s Go victory, we still need to charge ahead for breakthroughs in a slew of public sector areas. An area of interest has been integrative AI approach — how to combine capabilities in image recognition, speech recognition, computer vision and natural language to explore tougher problems. For example, we are seeing a number of startups that use emerging technologies to tackle social problems in India. Case in point, Fasal an IoT startup gives ML-powered insights to farmers, giving rise to precision agriculture.","excerpt":"As concerns around artificial intelligence increase — from a shortage of skilled professionals to ethics — its impact has gone well beyond technology. Governments and professional bodies have realised that the reality of AI is no longer just about technology anymore. There are social, philosophical and ethical aspects of AI and machine learning that companies […]","categories":["AI Trends"],"tags":["current leaders in self driving cars"],"author_name":"Richa Bhatia","publish_date":"2018-07-23T11:24:05","publication_year":"2018","word_count":733,"keywords":["machine learning","artificial intelligence","AI","neural network","ML","image recognition","computer vision","RAG","Aim","R","current leaders in self driving cars"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","neural network","computer vision","Aim","RAG","image recognition","R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-5-international-professional-artificial-intelligence-bodies\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10162812,"title":"Bengaluru-Based Presentations.ai Secures $3 Million from Accel","content":"Presentations.ai, an Indian AI-driven presentation platform, has secured $3 million in seed funding led by Accel Partners. The Bengaluru-based startup helps companies quickly generate presentation decks using AI and plans to use the funds to scale its software and expand operations. The seed round also saw participation from notable Indian entrepreneurs, including Paytm’s Vijay Shekhar Sharma, CRED’s Kunal Shah, and Freshworks founder Girish Mathrubhootham, among others. Founded in 2018 by Sumanth Raghavendra and Ravi Kasthuri, Presentations.ai has experienced rapid growth since launching its public beta in July 2023. The startup reached 1 million users in just 84 days and now has over 5 million users worldwide. AI for Presentation The startup’s AI-powered platform aims to streamline the presentation creation process, addressing a common pain point for businesses of all sizes. Presentations.ai simplifies presentation creation by leveraging AI and best design practices. It uses advanced language models and text-to-image technology to generate content and visuals while offering theme palettes, presentation styles, and brand templates for customisation. To ensure enterprise security, the platform includes safeguards against AI hallucinations and protects sensitive data. The company transitioned to a freemium model in early 2024 and now has thousands of paying users. Its Pro plan is priced at $198 per year, with localised pricing across different markets. Other AI-powered presentation tools, such as Gamma and Prezi, have also been leveraging advanced language models and ML to automate content generation and design tasks, streamlining the creation of professional presentations. Presentations.ai currently generates 20% of its revenue from the US, followed by India. Significant markets include the UK, Germany, Australia, Canada, and the Middle East. With fresh funding, the company plans to launch a dedicated agent for creating presentations within any application and build an enterprise sales team. Recently, Accel, along with Kae Capital, co-led a $2.6 million seed funding round for BPR Hub, an AI-driven compliance management platform for manufacturers. Founded in 2024, BPR Hub helps manufacturers, particularly SMEs, streamline regulatory compliance by automating up to 80% of quality, compliance, and governance tasks. With operations in San Francisco and Bengaluru, the company plans to use the funding for global expansion, team growth, and new feature development.","excerpt":"Founded in 2018, the startup has over 5 million users worldwide.","categories":["AI News"],"tags":["Accel","AI startup","Presentations ai"],"author_name":"Vandana Nair","publish_date":"2025-02-04T11:26:13","publication_year":"2025","word_count":360,"keywords":["AI startup","Go","API","funding","programming_languages:R","AI","Presentations ai","Accel","ML","RAG","Aim","R","startup"],"extracted_tech_keywords":["AI","ML","Aim","RAG","R","Go","API","startup","funding","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/bengaluru-based-presentations-ai-secures-3-million-from-accel\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10059772,"title":"Council Post: How to develop a comprehensive AI governance &#038; ethics function","content":"Webster defines governance as ‘the act or process of governing or overseeing the control and direction of something (such as a country or an organisation)’ Most of us agree with the above definition when it’s about a country or an organisation. But, ‘AI governance’ is a slightly different ball game. What is AI governance? Simply put, it is a process of developing a framework to ensure AI systems operate smoothly in a controlled setup. This includes capturing and managing metadata on AI models as part of AI governance processes, and testing models for biases, among other things. If implemented correctly, it brings trust in AI systems. IBM outlines AI governance as the process of defining policies and establishing accountability to guide the creation and deployment of AI systems in an organisation. “AI governance is about AI being explainable, transparent, and ethical,” according to ML observability platform Arise. Marketing company TechTarget wants AI governance to have a legal framework to ensure machine learning technologies are well researched and developed to help humanity navigate the adoption of AI systems fairly. Forbes describes AI governance as evaluating and monitoring algorithms for effectiveness, risk, bias, and RoI. Though the definitions of AI governance vary, the basic tenet remains the same – building trust in AI systems. But what goes into the making of AI governance functions? AI governance, by nature, is an interdisciplinary function. It has a lot of stakeholders, processes and departments across the lifecycle of a project. Let’s explore further. AI researchers and data scientists Biases including cognitive bias, incomplete data, flaws in the algorithm, etc, slow down the growth of AI in an organisation. Research and development play an important role in addressing these issues. Who understands this better than ethicists, social scientists, and experts? Therefore, businesses should include such experts in their AI projects across applications. Data architects Data architects also play a key role in governing AI products. Companies should have a complete pipeline of data or metadata for AI modelling. Remember, AI’s success depends on a well-sorted data architecture that is error and noise-free. To do so, data standardisation, data governance, and business analytics are a must. HR HR plays a key role in shaping the AI governance function. For instance, they should find candidates who “fit” into the organisation’s existing AI framework and create training material for the existing workforce to help them understand how to create ethical AI applications. Legal and compliance Ensuring AI products don’t cross any legal boundaries is critical for smooth deployment. AI solutions meet the stipulated compliance guidelines of the organisation and the industry in which the organisation operates. No one-size-fits-all kind of approach comprehensively captures the legal and compliance boundaries. For instance, what customers consider ethical for a financial services industry can be completely different for businesses. The goal of embedding legal and compliance teams into the AI governance function is to ensure diverse & inclusive inputs in decision making. Business representation Business functions such as marketing, sales, HR, supply chain, and finance have realised the benefits of embedding AI in their strategy. That means domain knowledge becomes extremely important, not just for building practical applications but also for evaluating them. Therefore, good business representation in the core AI governance council can yield better results. The governance council of an enterprise should have members from different domains. Moreover, it helps in inclusive and smooth governance, where all the business requirements are considered. Procurement and finance Many product-based companies have incorporated AI functionality in their products. Often, when companies are buying a product that’s not primarily based on AI, it generally doesn’t come under the purview of the AI governance body. But what if the business uses a feature\/functionality\/product based on AI? Ideally, the procurement and finance teams should have AI experts to help onboard the right product. A good AI governance function will provide a framework to manage AI algorithms and products effectively. In addition, building an agile and cross-functional AI governance team will bring diverse perspectives to the table and help create awareness around AI across the spectrum. This article is written by a member of the AIM Leaders Council. AIM Leaders Council is an invitation-only forum of senior executives in the Data Science and Analytics industry. To check if you are eligible for a membership, please fill out the form here.","excerpt":"Though the definitions of AI governance vary, the basic tenet remains the same – building trust in AI systems.","categories":["AI Features"],"tags":["ai governance","data architecture","Ethical AI","Responsible AI"],"author_name":"Aswini Thota","publish_date":"2022-02-04T17:00:00","publication_year":"2022","word_count":722,"keywords":["data science","Go","machine learning","AWS","AI","Ethical AI","ai governance","ML","Responsible AI","data architecture","Aim","analytics","Rust","R"],"extracted_tech_keywords":["AI","machine learning","ML","data science","analytics","Aim","AWS","R","Go","Rust"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/council-post-how-to-develop-a-comprehensive-ai-governance-ethics-function\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":69204,"title":"Clinical Parameters And Artificial Intelligence","content":"Clinical trials involve measurement of vital parameters at regular intervals of time. Vital parameters include measurement of respiratory rate, blood pressure, pulse and blood oxygen saturation whenever required. Importance of vital parameters are as follows: This helps in understanding any significant improvement or deterioration of health.It can help in the early diagnosis of an acute medical illness.They are also important markers in chronic disease states. With traditional clinical trial setup following are the drawbacks: For recording these parameters patients must visit the hospital at a regular interval of time which is time and cost consuming. It is difficult to arrange for the equipment and time for each patient in hospitals with large numbers of patients.There are possibilities of missing earlier variabilities as the patient’s visit to the hospital occurs only at a stipulated period.With emerging emergencies like pandemic situations, it is tough to get the parameters on time which can risk a patient’s life. Artificial intelligence’s role in the measurement of vital signs: AI based remote vital sign monitoring devices are emerging into the market. The advantages of such devices are as follows: The number of visits to the hospital can be reduced as patients can monitor their vital signs at home without visiting the hospitals.As the number of points of measurement increases, the chances of early detection of end-point can be obtained.Due to the early detection of the end-point, the trial can be completed much earlier.Due to the continuous monitoring of vital signs, the deterioration in patient health condition can be detected at a much earlier stage and can help in better treatment.The continuous monitoring and increased number of readings can also help in understanding the disease condition in a better angle leading to development of more efficient treatment. Patient dairy: Another such improvement in clinical trials from a patient’s perspective is AI enabled Patient dairy. Patient dairy is the record where the patient enters the daily routine of intake of medications and any health conditions experienced by the patients during the treatment period. Hence patient dairy plays an important source record in any clinical trial. It is the basis of any adverse events that occur in a clinical trial. By making the patient dairy AI enabled, following are the advantages: In a few trials compliance of patients administering the medication is crucial. There are many instances, where in patient participating might not abide by the timelines to administer the medicaments. Some of the reasons are: Due to forgetting the dose timings.Sometimes patients come across other people who influence in not taking the trial medications, hence might forge in such situations. In such cases, If the patient dairy is technology based, it can provide daily reminders for intake of medication. It can be live video linked, where in the administration of daily dosage of the patients can be monitored in live time. The direct capture of data entry inpatient dairy into eCRF would further reduce manual errors, time consumption and resource consumption. This would reduce the cost.The Clinical trial team can be updated in live timings regarding any discrepancies before the patient updates during future visits. This would help in capturing any significant variations and provide better treatment to the patients.","excerpt":"Clinical trials involve measurement of vital parameters at regular intervals of time. Vital parameters include measurement of respiratory rate, blood pressure, pulse and blood oxygen saturation whenever required. Importance of vital parameters are as follows: This helps in understanding any significant improvement or deterioration of health. It can help in the early diagnosis of an […]","categories":["Global Tech"],"tags":["Intel","what is artificial intelligence"],"author_name":"Prasidha","publish_date":"2020-07-08T13:00:00","publication_year":"2020","word_count":532,"keywords":["artificial intelligence","programming_languages:R","AI","ViT","what is artificial intelligence","R","Intel"],"extracted_tech_keywords":["AI","artificial intelligence","R","ViT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/clinical-parameters-and-artificial-intelligence\/","complexity_score":2,"technical_depth":5,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10079996,"title":"First Trillion Parameter Model on HuggingFace &#8211; Mixture of Experts (MoE)","content":"Google AI’s Switch Transformers model, a Mixture of Experts (MoE) model, that was released a few months ago is now available on HuggingFace. The model scales up to 1.6 trillion parameters and is now openly accessible. Click here to check out the model on HuggingFace. MoE models are considered to be the next step of Natural Language Processing (NLP) architectures that have highly efficient scalable properties. The architecture is considered similar to the classic T5 model, with a Feed Forward layer getting replaced by a Sparse Feed Forward Layer. Individual tokens are processed by different MLPs guided by the router module. Younes Belkada from HuggingFace said, “The architecture can build on research for the next generation of NLP models including GPT-4.” Read: GPT-4 is almost here, and it looks better than anything else The weights of the model are pre-trained and require fine-tuning before using them for projects. HuggingFace has also released a demo on how to fine-tune an MoE model on text summarisation that you can check out here. What are Switch Transformers? Switch Transformers are effective natural language learners that are highly scalable. Simplifying MoE makes the models excel in natural language tasks in all the training regimes, allowing the model to be trained on billions to trillions of parameters, and substantially increasing the speed when compared to T5 baselines. You can also check out the documentation about SwitchTransformers on HuggingFace website. Click here to read the paper by Google AI on Switch Transformers: Scaling to Trillion Parameter Models with Simple and Efficient Sparsity.","excerpt":"Google AI’s Switch Transformers Model is now openly accessible on HuggingFace.","categories":["AI News"],"tags":["Natural Language Processing"],"author_name":"Mohit Pandey","publish_date":"2022-11-17T17:14:41","publication_year":"2022","word_count":256,"keywords":["Go","AI","Natural Language Processing","ML","Transformers","Scala","NLP","GPT","R","T5","llm_models:GPT"],"extracted_tech_keywords":["AI","ML","NLP","Transformers","R","Go","Scala","GPT","T5","llm_models:GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/first-trillion-parameter-model-on-huggingface-mixture-of-experts-moe\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":39172,"title":"8 Promising Linux Distributions To Look Forward To In 2019","content":"Linux is amazing. Period. Over the years, Linux has gained tremendous traction — it built a complete ecosystem of developers and users who love working on open source. From a day-to-day use to sophisticated penetration testing, Linux has marked its presence everywhere. Also, one of the most amazing thing about Linux is that you can tweak or create your own distros. From individuals to big firms, many are creating Linux distribution and giving it to the world. So, if you are a Linux enthusiast and looking for some new Linux Distros to get your hands, read along as we list down some of the most promising Linux distros to look for in 2019. MX Linux What started as a concept in a discussion about future options among members of the MEPIS community in December 2013, today it is one of the popular Linux distros. A cooperative venture between antiX and MEPIS, MX Linux is a midweight distro designed and built, combining an elegant and efficient desktop with a simple configuration. The distro is highly stable and has got a solid performance. You can download the distro from here. Also, for more information about the distro, click here. Antergos Started out with a purpose to deliver an OS that is modern, elegant, and powerful and also base on Arch Linux, Antergos is a rolling release distribution. So, when you use Antergos the best advantage you get is that from the base OS components to the applications that you install, will receive updates as they are released upstream—with only a minimal delay to ensure stability. Also, Antergos is free to download and use for everyone. Bodhi Linux Specifically focused on minimalists, Bodhi Linux is the “Enlightened” Linux distribution. The distro is built as lean as possible. The main concept behind this amazing Linus distro is that — the user should be able to choose and use the applications they want rather than having a bunch of pre-installed applications. The Bodhi Linux comes with the desktop called “Moksha” — which is a continuation of the Enlightenment 17 desktop. If you are someone who is a minimalist, doesn’t want a cluttered desktop, then Bodhi is definitely for you. You can download the Distro from here. Peppermint OS Peppermint is a lightweight, stable, and a fast Linux OS. The latest version, Peppermint Nine is built on long term support (LTS) code base — built on the Ubuntu 18.04 code base. Meaning, one can easily customize and use it for a good amount of time. Also, it provides easy integration with cloud and web apps. Click here to download Peppermint. ArcoLinux Previously known as ArchMerge, ArcoLinux is another promising Linux distro that a Linux enthusiast should definitely use. The best thing about ArcoLinux is that it is a completely Arch-based Linux distro. Users can build or tweak custom distributions. Furthermore, having Xfce as its default DE, the distro is a minimalist-focused. So, just like bodhi, users can install the application they want to use. ArcoLinux is available for everyone for free and you can download it from here. Tails Tails is a complete privacy-focused Linux distro. It is basically a live operating system that anyone can use on any system just using a DVD or a USB stick. Based on Debian GNU\/Linux, this amazing live OS let you surf the internet using any computer anonymously. How it does that? The OS makes use of TOR — it sends the traffic through TOR. That is not all, the Tails is designed in such a way that, once you unplug the USB or close the OS, it doesn’t leave any trace on the computer. Tails are free and can be download from here. SparkyLinux Based on Debian GNU\/Linux operating system, SparkyLinux is a fast, lightweight and fully customizable OS built around the Openbox window manager. The Distro offers different versions for different users such as GameOver for gamers,  Multimedia for audio, video and HTML pages creating, and Rescue for fixing broken OS. It comes with LXDE or LXQt desktop environment with pre-installed common use software. Furthermore, this Linux Distro supports about 20 desktop environments and window managers. Click here to download. Robolinux Robolinux is one such Linux distro that is unique compared to other Linux distros. The Distro is created to Windows XP, 7 and 10 virus-free. It is also considered to be one of the most secured operating systems. Robolinux lets the user create a Windows virtual machine that can run side by side with Linux. Robolinux Raptor series 8, 9 & 10 Gnome, Cinnamon, Mate 3D, Xfce 3D & LXDE versions run significantly faster than any Windows OS and also makes it easy to run Windows programs flawlessly and natively inside its desktop. You can download this secured Linus Distro from here.","excerpt":"Linux is amazing. Period. Over the years, Linux has gained tremendous traction — it built a complete ecosystem of developers and users who love working on open source. From a day-to-day use to sophisticated penetration testing, Linux has marked its presence everywhere. Also, one of the most amazing thing about Linux is that you can […]","categories":["AI Trends"],"tags":["Linux","Linux distros","Open Source"],"author_name":"Harshajit Sarmah","publish_date":"2019-05-15T07:52:33","publication_year":"2019","word_count":796,"keywords":["Go","Linux distros","Open Source","programming_languages:R","AI","data_tools:Spark","ML","programming_languages:Go","GAN","Linux","R"],"extracted_tech_keywords":["AI","ML","R","Go","GAN","programming_languages:R","programming_languages:Go","data_tools:Spark"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/8-promising-linux-distributions-to-look-forward-to-in-2019\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10065286,"title":"Why did OpenAI invest in a nuclear fusion startup?","content":"Washington-headquartered nuclear fusion research company Helion Energy had momentum from day one. Before starting Helion in 2013, CEO and co-founder David Kirtley used to work with fusion programs in the U.S. Department of Energy. By the end of last year, Helion had received funding from the Department of Energy and NASA and counted Facebook co-founder Dustin Moskovitz, OpenAI CEO Sam Altman and Peter Thiel-backed Mithril Capital as investors. In November last year, Helion raised another USD 500 million in funds. Ambitious goals The funding, Helion said, will be used to construct its seventh-generation fusion generator, Polaris, the company’s newest prototype. Helion’s end goal is to produce electricity through nuclear fusion. And according to Helion, Polaris will be the world’s first fusion electricity demonstration facility. In June of last year, Helion demonstrated results that showed it was the first private company to heat a fusion plasma to more than 100 million degrees celsius. When fusion occurs near the sun’s core, the temperature is normally around 15 million degrees celsius. To cause nuclear fusion on earth, the temperature has to ideally be around 100 million degrees celsius. Source: Helion While nuclear fusion has never been proven to be a viable source of energy, Helion is a part of a wave of startups that believe they are getting closer to producing actual results. The company has said it is currently building a prototype reactor called Trenta, which completed a continuous testing run for more than a year. Helion has promised to demonstrate the production of net electricity from fusion by 2024 and claims it will be able to sell electricity commercially within the next decade. Post the funding round, Sam Altman spoke about Helion enthusiastically in his blog. “Helion has a clear path to net electricity by 2024 and has a long-term goal of delivering electricity for 1 cent per kilowatt-hour,” he noted. The former Y-Combinator president discussed how Helion could create a path that might lead us out of the climate crisis. Altman underlined the importance of accessing cheaper energy and that it could reduce overall costs and make drastic changes to people’s lives. Nuclear wave Nuclear fusion is the process when hydrogen atoms are fused together with helium. Fusion is what powers the sun and stars. Unlike nuclear fission, which produces energy from breaking atoms, fusion does not generate radioactive waste. Nuclear fusion reactions can be controlled by humans and turned off and on instantly. If tapped into successfully, fusion can produce energy 24 hours a day, unlike solar power. The promising outcomes of nuclear fusion have not gone unnoticed by investors and the tech industry. According to research company Pitchbook, just last year, venture investors spent USD 3.4 billion on nuclear start-ups – a record and more than every year over the last decade combined together. Helion isn’t the only benefactor of this trend. Commonwealth Fusion Systems also raised USD 1.8 billion last year. Last year, the number of deals closed in the sector jumped to 28 as compared to an average of fewer than ten deals for the past decade. Besides growing concerns around climate change, the recent energy crisis in Europe turned the focus towards nuclear even more. The interest in startups like Helion is pushing towards commercialisation faster than was initially anticipated. Vancouver-based General Fusion announced earlier this month that it had formed a Market Development Advisory Committee (MDAC) to engage with prospective clients. General Fusion is backed by Jeff Bezos, among others. Source: DeepMind Scientific advancement As it happens, finance is fuelling science. In a moment of marked progress, in February, DeepMind trained an AI using a reinforcement learning algorithm to control the hot, burning plasma inside a tokamak nuclear fusion reactor. Controlling the confined plasma is to control the nuclear fusion reaction. Up until now, scientists used powerful magnetic coils to control the plasma, which is hotter than the surface of the sun. The Alphabet-backed research company taught a deep reinforcement learning system to control the magnetic coils autonomously. Attempts have been made previously by AI to control nuclear fusion in the past as well. Google has been working with TAE Technologies since 2014 to employ machine learning in fusion reactors. This resulted in a quicker analysis of experimental data. The UK’s Joint European Project used deep learning models to predict the disruptions in the plasma.","excerpt":"According to research company Pitchbook, last year venture investors spent USD 3.4 billion on nuclear start-ups – a record, and more than every year over the last decade combined together.","categories":["Global Tech"],"tags":["AI Tool"],"author_name":"Poulomi Chatterjee","publish_date":"2022-04-19T17:59:55","publication_year":"2022","word_count":721,"keywords":["Go","API","machine learning","OpenAI","AI","RAG","Aim","deep learning","ViT","AI Tool","R"],"extracted_tech_keywords":["AI","machine learning","deep learning","OpenAI","Aim","RAG","R","Go","API","ViT"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/why-would-openai-invest-in-a-nuclear-fusion-startup\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10160813,"title":"Genrobotics Launches First Robotic-Assisted Gait Trainer for Paediatric Rehabilitation","content":"Genrobotics unveiled India’s first robotic-assisted paediatric gait trainer, the ‘G-Gaiter Paediatric’. The innovative device aims to improve the physical rehabilitation process for patients with mobility issues, including those with stroke-triggered paraplegic conditions, traumatic spinal cord injury, cerebral palsy, and Parkinson’s disease. This is an essential advancement for Indian healthcare, especially in the paediatric rehabilitation segment. Kerala’s health minister, Veena George, launched the G-Gaiter Paediatric during a ceremony held at the C-DAC auditorium in Technopark on Friday. The first unit was handed over to Sabith Umer, vice chairman of the Kannur-based charitable trust, Thanal Group. Vimal Govind M K, CEO of Genrobotics, stated, “Our mission is to revolutionise physical medicine and rehabilitation using technology, ensuring that individuals with gait challenges, regardless of age, have access to cutting-edge solutions.” The company claims it is working to enhance the influence of robotics and AI in rehabilitation, as discussed in its talk with Dr. Aathithya SP, executive director of SP Fort Healthcare, and a deeper discussion with Farhaan Yasin‬, VP at Aster India. Robotic Hope for Kids Speaking at the event, George highlighted the significance of this innovation. “This technology will bring new hope to children and their families, improving their quality of life and independence,” she said. The device will be installed at Sree Avittam Thirunal (SAT) Hospital in Thiruvananthapuram, one of India’s Centres of Excellence for Mother and Childcare. Minister George also emphasised that this initiative marks Kerala’s shift towards integrating advanced technology into its traditionally community-focused healthcare system. “Continuous and effective efforts are required to develop the neuroplasticity of a person with permanent disability. With the availability of G-Gaiter at the hospital, the quality of health care provided to people with physical disabilities will go up,” said Dr. Nitha J, consultant, physical medicine and rehabilitation department at KIMSHEALTH. The event was attended by several prominent figures, including Anoop Ambika, CEO of Kerala Startup Mission (KSUM), Dr. PV Unnikrishnan, member secretary of K-DISC, and Dr. Rahul UR, state nodal officer for Child Health. “Through G Gaiter’s robotic exoskeleton, we are able to introduce patients to gait training at an earlier stage. It has also helped to drastically increase the morale and participation of patients,” said Dr. Hafeeza Tamton, consultant and head at the department of physical medicine and rehabilitation, Aster Mother Hospital, Areekode. The team of PM&R and G Gaiter at Aster Mother Hospital in May helped a bedridden patient make a recovery by walking back to life. The patient, Arun S Nair, says, “From a stage of balancing issues, I could walk in a balanced way within 15 days.” This launch positions India as a leader in leveraging robotics for healthcare, paving the way for transformative solutions in paediatric rehabilitation. Key Features of G-Gaiter Paediatric The G-Gaiter Paediatric combines advanced Virtual Reality (VR) technology with interactive games to create an engaging therapeutic experience. It uses Genrobotics’ patented GPLOT Exoskeleton Technology to enable children to develop natural gait patterns through precise and advanced robotic game-based therapies. It also works on several advanced features aimed at enhancing rehabilitation. The patented exoskeleton design ensures natural gait training and allows for adjustable configurations to suit a variety of patient needs. The device includes intelligent therapy modes, which adapt to support the relearning of motor skills, improve muscle coordination, and enhance overall mobility. With a child-friendly interface, it prioritises comfort and accessibility, ensuring a positive experience for young users. Additionally, its real-time monitoring capability provides detailed analytics, offering valuable insights into therapy progress and enabling more precise, personalised treatment.Recently, India received regulatory approval from the Central Drugs Standard Control Organisation (CDSCO) for telesurgery through its first robotic surgical system, SSI Mantra.","excerpt":"G-Gaiter Paediatric offers real-time monitoring capability enabling more precise and personalised treatment for each patient.","categories":["AI News"],"tags":["healthcare robotics"],"author_name":"Sanjana Gupta","publish_date":"2025-01-06T18:40:35","publication_year":"2025","word_count":605,"keywords":["Go","API","AI","healthcare robotics","RAG","Aim","ViT","analytics","Rust","GAN","R"],"extracted_tech_keywords":["AI","analytics","Aim","RAG","R","Go","Rust","API","GAN","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/genrobotics-launches-first-robotic-assisted-gait-trainer-for-paediatric-rehabilitation\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10073507,"title":"RevSure.AI Raises $3.5 Million to Transform B2B Pipeline Readiness","content":"RevSure.AI, an AI-fueled startup has raised $3.5 million in the seed round to transform B2B pipeline readiness. The round was led by Innovation Endeavors, along with the participation of angel investors Katrin Ribant, Sharath Keshava Narayana, and Rick Scanlon. Established in 2021, the Delaware-based startup was founded by Deepinder Singh Dhingra,  building a first-of-its-kind sales pipeline readiness solution to connect marketing and sales funnels. “In terms of the product, there are a lot of revenue-intelligence solutions to help you convert your sales pipeline, with lesser solutions that generate quality leads. A lot of B2B companies are struggling with driving revenue growth. As there are other revenue intelligence tools for conversion, they are not generating enough pipelines, unable to meet revenue targets,” says Singh. The sales pipeline solution is set to launch in September this year. The CEO adds, “While there are revenue intelligence tools for pipeline conversion and sales forecasting, intelligence for pipeline generation has been an afterthought. The pipeline generation of marketing, SDRs, and sales becomes the Bermuda triangle of the pipeline leakage – that’s the gap we are trying to fill.” The firm aims to help B2B SaaS companies improve their pipeline readiness by gaining unprecedented visibility. As per the August 2022 Pavilion Pulse Survey, 57% of respondents mentioned that their sales-driven pipeline was lower than the target, while 59% said that their marketing-driven pipeline was lower than the target. RevSure is an AI-fueled company committed to reducing the uncertainty around revenue funnel conversations, helping companies to boost qualified opportunities in their pipeline solutions.","excerpt":"The Delaware-based startup aims to improve sales pipeline readiness for B2B SaaS companies","categories":["AI News"],"tags":["SaaS","seed funding","Startups"],"author_name":"Bhuvana Kamath","publish_date":"2022-08-24T17:57:32","publication_year":"2022","word_count":256,"keywords":["seed funding","programming_languages:R","AI","innovation","Ray","SaaS","Aim","Startups","R","startup"],"extracted_tech_keywords":["AI","Aim","Ray","R","innovation","startup","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/revsure-ai-raises-3-5-million-to-transform-b2b-pipeline-readiness\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":35383,"title":"How Does India&#8217;s AI Strategy Fare Vis-à-Vis Its US Counterpart?","content":"The growing necessity of applying technology more ethically and responsibly has forced powerful nations like the US, as well as developing nations like India to layout strategies for the coming years. Since the field of artificial intelligence has the potential of producing something remarkable overnight, measures are ought to be taken to guarantee growth without any unwanted impediments. The US Department of Defense recently released a report on AI strategies, funds allocation and areas of focus. In a document titled Harnessing AI to Advance Our Security and Prosperity, US Department of Defense stated that AI was poised to transform every industry and is expected to impact every corner of their Department, spanning operations, training, sustainment, force protection, recruiting, healthcare, and many others. Whereas, in mid-2018, Indian Government’s think tank NITI Aayog released a survey or a roadmap to the adoption of AI in various sectors of the country. A Task Force was assigned to study the strategic implementation of AI for National Security and Defence was set up by MoD\/DDP in February 2018. According to NITI Aayog’s AI report, India’s plan is to pursue “moonshot” projects ambitious explorations that aim to push the technology frontier and that would require the pursuit of world-class technology development and leadership in applying AI technologies to solve some of the biggest challenges. Here is a comparison of AI strategies of the US and India: Role Of Academia & Enhancing Partnership With Industry US plans to make longer-term, stable funding available to attract the best academics to invest in long-term research relevant to critical DoD areas and remain in the business of educating the next generation of AI talent. This entails increasing investment through existing channels, such as DARPA\/IARPA and the Military Service Research Laboratories, and sponsoring long-term discoveries relevant to the Department. It also involves stimulating the development of geographic concentrations of interconnected companies and institutions in AI. Strong and stable academic partnerships clustered in this manner will provide benefits to the Department, industry, and national competitiveness. In India, institutes like IIT Hyderabad are already offering an MTech program in AI and ML, and an MTech in Data Science since 2015-16. And, research giants like DRDO has its own AI wing where it has brought out impressive products in the form of CAIR Recently, NITI Aayog, Intel, and Tata Institute of Fundamental Research (TIFR) joined hands together to set up a Model International Center for Transformative Artificial Intelligence (ICTAI) towards developing and deploying AI-led application-based research projects. This initiative is part of NITI Aayog’s ‘National Strategy for Artificial Intelligence’ ICTAI aims to conduct advanced research to incubate AI-led solutions in three important areas – healthcare, agriculture and smart mobility – by bringing together the expertise of Intel and TIFR. It aims to experiment, discover and establish best practices in the domains of ICTAI governance, fundamental research, physical infrastructure, compute and service infrastructure needs, and talent acquisition. Through this collaborative effort, the model ICTAI is chartered to develop AI foundational frameworks, tools and assets, including curated datasets and unique AI algorithms. The Role Of Open Source Community Financial incentives for private companies could include payroll taxes which are dedicated to subsidising training opportunities, income tax deductions for companies participating in reskilling initiatives, special taxes to be paid if a minimum training budget is not disbursed, as well as public grants for subsidising training, especially for smaller sized firms. And, help create a pipeline of AI research projects for through initiatives like those of ICTAIs through grand challenges to be given by the government and PSUs. And incentivise public agencies to adopt and employ AI in delivering service through financial support; extra budgets for R&D; tax incentives and awards. The open-source community serves as a vibrant global incubator of talented individuals and transformative ideas. To contribute data, challenges, research, and technologies to this community and engage with the open-source ecosystem as a vehicle for attracting talent, identifying and advancing new AI technologies that can transform defence, and broadening the accessible technology base. The intent is to develop standards and support policy development related to information technology such as data storage, information security, privacy, and ethics for data capture and use. And to develop AI foundational technologies to promote applied research that can scale for national impact and will lead to the creation of a vibrant and self-sustaining ecosystem. Read the detailed report here. Budget The ultimate goal of this research is to produce new AI knowledge and technologies that provide a range of positive benefits to the society while minimising the negative impacts. The Trump Administration has prioritised funding for fundamental AI research and computing infrastructure, machine learning, and autonomous systems. The Federal Government’s investment in unclassified R&D for AI and related technologies has grown by over 40% since 2015, in addition to substantial classified investments across the defence and intelligence communities. On September 7, 2018, the U.S. Department of Defense announced it will invest up to $2 billion over the next five years towards the advancement of AI. This will be in addition to existing government spending on AI R&D, which totalled more than $2 billion in 2017 alone, just from unclassified programs and not including Pentagon and intelligence budgets. Existing funding has already propelled more than 20 active programs under the Defense Advanced Research Projects Agency (DARPA) exploring different aspects and uses of AI. To pave the way for greater advancements in digital technologies, the Indian government has doubled its allocation to the ‘Digital India’ programme to $480 Mn (₹3,073 crore) in 2018-19. During India’s Budget 2018 session, it was announced that the government will be investing extensively in research, training and skill development in robotics, AI, digital manufacturing, Big Data intelligence and Quantum communications, among others. Challenges Faced In Large Scale Adoption Indian AI taskforce and their US counterparts identify few areas of challenges and the way to address them: In the Indian scenario, the challenges are concentrated across common themes of: Lack of enabling data ecosystems: Low intensity of AI research Core research in fundamental technologies Transforming core research into market applications Inadequate availability of AI expertise, manpower and skilling opportunities High resource cost and low awareness for adopting AI in business processes Unclear privacy, security and ethical regulations Unattractive Intellectual Property regime to incentivise research and adoption of AI. Whereas, the US seems to be well ahead of India in terms of the advancements in this space. They had a good head start by at least more than a decade. This puts them in a comfortable position with respect to the infrastructure available. The main challenge that the US faces is the adoption of AI at the ethical level. The growing scepticism surrounding the eavesdropping of watchdogs like NSA has put it under a tight spot with respect to its accelerated approach in adopting AI. AI For Social Good and Open Mission Initiatives The US plans to form open AI missions with academia and industry that will contribute to addressing global challenges of significant societal importance, such as operationalising AI for humanitarian assistance and disaster relief for wildfires, hurricanes, And Earthquakes. And combine efforts with a wide range of actors to produce inspiring AI technology that benefits society. These open missions will challenge a broad community to advance the state of AI and learn how to operationalise the technologies on an integrated basis across domestic and international organisations. Today’s world creates and has access to an unthinkably large amount of data, which can be harnessed to gain new insights. This is the time where the usage of data analytics powered by machine learning has gained traction on a large scale. The sudden surge in applied artificial intelligence presents the innovators with endless possibilities. Non-profit organisations have historically been suffering from the lack of tools to tackle this amount of data. But things are changing now, even in India. For instance, UK-based Data Kind conducts events which are volunteered by the data scientists who not only address the impending issues but also offer solutions. Their Bengaluru chapter had recently come up with ideas like using Anganwadi data along with NASER data to develop a thematic map for achievement across all the districts in Karnataka, and the correlation of availability of drinking water and toilets to the dropout rate among students. The US believes that the insertion of new technologies into complex work systems changes the nature of the work, including new forms of brittleness and error, and uncovers new at the same time as it improves the work in other respects. The infrastructure, the resources and the talent pool are there but for a large scale adoption, the government needs the belief of the commoner. Though the challenges faced at the resources level in the Indian scenario is quite different from the US, they both seem to agree upon the ethical and privacy part of adopting this technology. What India lacks is the rich talent pool that pioneers like the US and China flaunt. India has its handful in terms of manpower but falls short at the skilled section. But this also happens to be the reason why the world takes India seriously; skills can be learned but building a population both dense and young will take more than a generation. So, it is high time India puts their initiatives into motion more aggressively.","excerpt":"The growing necessity of applying technology more ethically and responsibly has forced powerful nations like the US, as well as developing nations like India to layout strategies for the coming years. Since the field of artificial intelligence has the potential of producing something remarkable overnight, measures are ought to be taken to guarantee growth without […]","categories":["Deep Tech"],"tags":["AI What it Does","DARPA","NITI Aayog","US"],"author_name":"Ram Sagar","publish_date":"2019-02-25T11:53:02","publication_year":"2019","word_count":1552,"keywords":["data science","Go","artificial intelligence","machine learning","AI","DARPA","ML","RAG","Aim","AI What it Does","analytics","R","NITI Aayog","US"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","data science","analytics","Aim","RAG","R","Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/how-does-indias-ai-strategy-fare-vis-a-vis-its-us-counterpart\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10097801,"title":"Stack Overflow’s Bumpy Ride to Gen AI Adoption","content":"Amidst rising moderator controversies and declining user numbers, Stack Overflow announced the integration of generative AI on their platform with Overflow AI — something that the company has been hinting about since April. Through this, Stack Overflow aims to improve the quality and relevance of resources to user queries. The feature will not only be available on the public platform but also extend to Stack Overflow for Teams. Will this be a promising step for Stack Overflow towards getting back to being relevant again? OverflowAI helps you create content within your Stack Overflow for Teams community, by taking existing sources within your instance and creating a tagging framework and series of questions and answers. pic.twitter.com\/TDIcIu7DNU— Stack Overflow (@StackOverflow) July 27, 2023 Up in arms against ChatGPT Ever since the ChatGPT rage began last December, Stack Overflow, the Q&A community platform for programmers, has been in news for their new policies, regulations or simply for their user switch. Developers, who had been using Stack Overflow for copying code started switching over to ChatGPT. The launch of Code Interpreter plugin on ChatGPT has made it even more coveted as a coder platform. When ChatGPT was finding its presence everywhere, Stack Overflow took the bold step of banning content generated by ChatGPT on its platform. In addition to attributing high inaccuracy as a reason for not allowing content from ChatGPT and other LLMs, the volume of answers that can be easily produced via GPT platforms which require a detailed review by an SME has burdened their volunteer-based quality curation infrastructure. The company faced a 12% slump in the number of user visits — 247 million visits in December from 279 million in November. However, the company confirmed to AIM that it did not face any threat to traffic as the dip was a normal one brought about by a seasonal shift owing to holidays in December. You cannot have the cake and eat it too ChatGPT is believed to be trained on Stack Overflow data. The GPT-3 paper that talks about training datasets includes Common Crawl which implies everything on the internet, including Stack Overflow. For a company losing customers to another platform owing to better and smoother experience for query resolution, it was only obvious that Stack overflow would next prevent ChatGPT from feeding off their datasets to better themselves or get paid — that’s pretty much what happened. Stack Overflow, closely following the footsteps of Reddit, announced that it will charge AI developers for accessing its programming driven community questions. The company believes that community platforms that support LLM models should, in some way, be compensated for their contribution, which will help these companies reinvest to grow the community. The Internal Tussle With initially blocking AI generated content on the platform leading to a number of users being suspended and posts being removed, Stack Overflow went back to allowing AI content. With a lack in uniformity, the company’s stance on AI-generated content came under scrutiny with curators, contributors and moderators participating in moderation strikes from June. The moderators then came to an interim solution where AI-generated content would be checked against ‘strong’ and ‘weak’ heuristics to serve as guidelines to evaluate a content’s quality and appropriateness. This would determine if a post needs to be removed or allowed on the platform. Furthermore, the moderators were able to convince Stack Overflow to continue providing data dumps and API access — something the company had been contemplating removing. Amidst the tussle, Stack Overflow finally announced generative AI. What’s Ahead? With Overflow AI, the company will be utilising their vast database consisting of over 58 million questions and answers from their community — a rich database exclusive to the company. Stack Overflow will integrate this database into IDEs (Integrated Development Environments), which are software applications that help developers write and manage codes. With this integration, the developers are facilitated with simpler coding resolution- akin to the multiple functions that ChatGPT’s Code Interpreter brings in. While Stack Overflow is definitely late to the AI race, with players such as GitHub Copilot X having implemented GPT-4 in March itself, the vast database it brings might prove helpful for the company. Whether this would help bring back lost users, is something that needs to be seen.","excerpt":"From fighting AI-generated content to announcing gen AI integration, Stack Overflow’s journey since ChatGPT’s launch has been tumultuous","categories":["AI Highlights"],"tags":["API","ChatGPT","Code Interpreter","Generative AI","GPT","GPT-3","GPT-4","Stack Overflow"],"author_name":"Vandana Nair","publish_date":"2023-07-31T14:07:44","publication_year":"2023","word_count":709,"keywords":["GPT-3","ChatGPT","API","AI","Code Interpreter","GPT-4","Stack Overflow","RAG","GPT","Git","Aim","generative AI","Generative AI","GitHub","R"],"extracted_tech_keywords":["AI","generative AI","ChatGPT","Aim","RAG","R","Git","GitHub","API","GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/stack-overflows-bumpy-road-to-gen-ai-adoption\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10103170,"title":"Adobe Promises a Photoshop like AI for Sound","content":"Design software giant Adobe is trying to raise the bar with the all new AI-powered Project Sound Lift. The one-click solution can dissect speech recordings into distinct tracks, separating voices, non-speech sounds, and background noise in videos. The company claims that Project Sound Lift is a one-click method through which users will be able to manipulate audio recordings, leveraging AI to independently enhance, transform, and control speech and sound. Adobe’s Enhance Speech technology—now available in Adobe applications like Premiere Pro—is integrated within Project Sound Lift allowing creators to craft studio-quality audio content easily. There is a growing need for the tool in the market as poor audio quality remains a persisting challenge in the creator’s economy. Artists work around audio issues caused by wind interference, sub-optimal microphone locations, crowd noise and other sound nuisances, any of which can render videos unusable but Adobe has introduced AI as the saviour. Developed by speech AI researchers at Adobe Research the project was announced at MAX Japan as part of Adobe’s “Sneaks” showcase, where Adobe engineers and research scientists offer sneak peeks at prototype ideas and technologies. In the demo shared with AIM, Adobe boasted how its project is tailored to pick up specific sound, can separate and manage a wide array of audio “events” from daily life scenarios, including splitting speech, applause, laughter, alarms, crowds, and other noises into distinct tracks. However, as with any technological leap Adobe is not the only nor first one to offer these capabilities. Lalal.ai already does all of the above tasks. Alongside the user have its iOS and Android apps to work on the go. Audo.ai is carving its niche in noise reduction. Moreover, the open-source Ultimate Vocal Remover GUI has been quietly serving as a constant option for audio manipulation, focusing primarily on vocal separation but also adaptable for denoising purposes.","excerpt":"However, as with any technological leap Adobe is not the only nor first one to offer these AI capabilities.","categories":["AI News"],"tags":[],"author_name":"Tasmia Ansari","publish_date":"2023-11-17T11:37:07","publication_year":"2023","word_count":306,"keywords":["Go","programming_languages:R","AI","ETL","programming_languages:Go","RAG","Ray","Aim","AI research","R"],"extracted_tech_keywords":["AI","Aim","Ray","RAG","R","Go","ETL","AI research","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/adobe-promises-a-photoshop-like-ai-for-audio\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10103721,"title":"NVIDIA Rides High on InfiniBands","content":"NVIDIA has been shining all along with the latest Q3 earnings reflecting the unstoppable growth of the tech giant. The latest earnings reported a revenue of $18.12 billion which was a 206% increase YoY and 34% from the previous quarter. The company even attributed the phenomenal growth in revenue to its continued ramp of NVIDIA HGX platform along with end-to-end networking via InfiniBand. NVIDIA has called out the contribution of networking that has now exceeded $10 billion annualised revenue run rate, nearly tripling from the previous year. This is attributed to the rising demand for InfiniBand which witnessed a fivefold increase YoY. A Complete Architecture InfiniBand, which is considered critical for gaining the scale and performance needed for training LLMs, when combined with NVIDIA HGX forms the foundational architecture for AI supercomputers and data centre infrastructures. InfiniBand is commonly used in supercomputing environments for interconnecting servers. The biggest advantage is its ability to provide low latency and high-bandwidth communication that is crucial for parallel processing tasks. With extreme-size datasets and ultra-fast processing of high-resolution simulations, NVIDIA’s Quantum InfiniBand Switches are said to match these needs with lower cost and complexity. A few months ago, NVIDIA had reached breakthrough performance with their leading H100 chip. The tests were run on 3,584 H100 GPUs that were connected with InfiniBand as they allowed GPUs to deliver performance at standalone and scale levels. Thereby, proving its prowess when combined with high performing networking capabilities. InfiniBands : The Preferred Choice Speaking about the future of InfiniBands, Jensen Huang said that the vast majority of the dedicated large scale AI factories standardise on InfiniBand, and it’s not only because of data rate and latency but “the way traffic moves around the network” is important. He also called it a ‘computing fabric.” Comparing it to Ethernet, Huang talks about the huge difference between the two. With NVIDIA investing $2 billion in infrastructure for AI factories, any form of variance, such as 20 or 30% in overall effectiveness will result in millions of dollars of change in value which accumulate as significant costs over the next 4-5 years. Huang calls InfiniBand’s value proposition ‘undeniable for AI factories.’ However, Ethernet is not ruled out. While Infinibands are used for cases that require high bandwidths with low latency, ethernet finds applicability in other scenarios. Ethernet, a widely used general-purpose networking technology for wired local area networks (LAN), is suitable for a broad range of applications, more geared towards connecting terminal devices. However, its capabilities cannot be matched with InfiniBands. Interestingly, NVIDIA also offers gateway appliances connecting InfiniBand data centres to Ethernet-based infrastructures and storage. NVIDIA will also release Spectrum-X in Q1 next year, an Ethernet offering that is said to achieve 1.6x higher networking performance when compared to other available Ethernet technologies. In terms of functionality, Intel’s Omni Path Architecture (OPA) was designed for high-speed data transfer and low latency communication in HPC environments. It was released in 2016, however, it was discontinued in 2019. Cisco on the other hand, has ethernet-based switches but nothing in the HPC space. An Integrated Expansion With GPU and networking offerings, enterprises are now given the choice of integrating their whole architectural framework from NVIDIA products. In addition to speaking about NVIDIA’s partnerships with Reliance, Infosys and Tata, the company mentioned their collaborations with organisations for optimising InfiniBands in their AI compute needs. In the earnings call, NVIDIA spoke about its partnership with Scaleway, a French private cloud provider that will build their regional AI cloud based on NVIDIA H100 InfiniBand and AI Enterprise Software to power AI advancements across Europe. Furthermore, Julich, a German supercomputing centre, also announced its plans to build their next-gen AI supercomputer using close to 24,000 Grace Hopper Superchips and Quantum-2 InfiniBand, elevating it to world’s most powerful AI supercomputer with over 90 exaflops of AI performance. Interestingly, Microsoft Azure uses over 29,000 miles of InfiniBand cabling. Infiniband enabled HB and N-series’ virtual machines are utilised by Microsoft for achieving HPC with cost efficiency. Bundling networking and GPU, NVIDIA is boosting its growth and stance in the supercomputer market. Going by the lack of alternatives to NVIDIA Infinibands, it looks like the company’s dominance is going to be further enhanced, ultimately making it indispensable for companies looking to utilise GPU and networking.","excerpt":"“The vast majority of the dedicated large scale AI factories standardise on InfiniBand,” said Jensen Huang during NVIDIA’s Q3 earnings call","categories":["AI Trends"],"tags":["enterprises","GPU","H100","Intel","Jensen Huang","NVIDIA","Supercomputers"],"author_name":"Vandana Nair","publish_date":"2023-11-27T18:30:00","publication_year":"2023","word_count":713,"keywords":["Go","NVIDIA H100","cloud_platforms:Azure","cloud_platforms:Microsoft Azure","AI","programming_languages:R","R","Jensen Huang","RAG","Supercomputers","GAN","H100","NVIDIA","enterprises","Azure","Intel","GPU"],"extracted_tech_keywords":["AI","RAG","Azure","R","Go","GAN","NVIDIA H100","cloud_platforms:Azure","cloud_platforms:Microsoft Azure","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/nvidia-rides-high-on-infinibands\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":4051,"title":"Interview – Prof Purba Rao, Author of Book “Business Analytics: an application focus”","content":"Prof. Purba Rao recently published a Book “Business Analytics: an application focus”, with PHI Learning. In an interview Analytics India Magazine, Prof. Rao talks about her motivation to write this book. To know more about the book, click here. [dropcap style=”1″ size=”2″]AIM[\/dropcap]Analytics India Magazine: How did you decide to write a book on Business Analytics? [dropcap style=”1″ size=”2″]PR[\/dropcap]Purba Rao: For last twenty years or so I have been teaching analytical modeling in business decision making in different business schools to different audiences of MBA students. The students seemed to appreciate the topic and, what was very rewarding to me; they applied these approaches to their real life projects and got meaning results. They also realized that these approaches could be applied to virtually any business process… whether relating to a business with niche product, a commodity, a monopoly or competitive market. After graduating from the business school some of them would write back to me telling me that they were indeed applying these techniques in decision-making situations in their respective workplace. Now in all my teaching, whether I was teaching a course called Quantitative Analysis or Advanced Marketing Research, or Predictive Modeling I could not find a textbook that my students could use. So I taught basically from the class notes which I prepared after lots of research from journal papers, books on multivariate analysis internet and computer software manuals. These materials were quite technical and perhaps, not so user friendly to a heterogeneous audience that I saw in business schools. You see, in business school on one side you come across engineers, math majors, hard-core finance majors who follow any mathematical equation you write on the black board. However, on the other side you also come across humanities majors, literature graduates, human resource professionals and even medical practitioners. When I am teaching such a heterogeneous class I have to make sense to all of them and at the same time help them appreciate that what I am trying to teach, can very well be applied to their line of business. So, I taught my analytical modeling in the real life context, with the help of cases, which I wrote often, based on examples I read about in the projects that the students worked on and submitted to me. Also, in place of advanced mathematical modeling I used softwares… sometimes EXCEL, sometimes SPSS, Crystal Ball, Answer Tree and AMOS. Where softwares were concerned my students took huge interest and demonstrated great understanding. Sometimes I would just show them the basic steps and students would come up and tell me how to use the software better. In general my heterogeneous audience learnt my Analytics topics and applied them very well and felt great satisfaction. To give an example, I had a student in my Quantitative Analytics class who was a male nurse in a hospital before he joined the MBA class. He was terribly afraid that he would never pass my course. However, he did pass, used our class discussions and applied them to a decision-making scenario in his hospital setting. All along my teaching my objective was to make full sense to my students whom I was addressing. To achieve this I was majorly relying on my class notes along with a few reference books on basic statistics multivariate analysis, etc. Then a time came when I thought why not put all my class notes together, make it into a binder and make it available to the students. Around that time, one day, I was sitting in my office and the Marketing Director of a large Publishing House walked in and said he was looking for Business School Faculty to write books for them. I told him that I had no time to think ‘of writing a book just then but he said he would consider my class notes and model them to form a book. I said I would think about that idea. Later in the day I talked about this to my son who had just graduated from Wharton Business School and was totally knowledgeable with MBA education. He insisted that I work on the book writing project with the Publishing House and come up with a book which would bring the wonderful world of Business Analytics to my past, present and future students. That was 2008 when I remodeled my class notes to form a book that was called then “Predictive Modeling for Strategic Marketing”. Five years have passed since then and now I have come up with this new book “Business Analytics an application focus”. I consider this new book as a far more comprehensive description of the topics I have been discussing over the last five years. I am also very happy that I have been able to include many real life case studies which my students have worked on, as part of my Business Analytics” elective, describing how the analytical modeling applies so effectively in various managerial decision making situations. My effort will be worth the effort if the various audiences which the book addresses see value in reading it and gain significant understanding of how the complex analytical models can be applied so lucidly in business and managerial situations. AIM: How is this book different from other similar books in market today? PR: There are no books on Business Analytics in the market that can be used as textbook in a course on Business Analytics. AIM: How did you start your career in analytics? PR: I have always been a math major, having an undergrad degree in Math honors (Presidency College, Kolkata), Masters in Applied Math (Science College, University of Calcutta) and doctorate (Fellow in Management from IIMC) in Operations Research. So, when I taught, I always taught quantitative\/analytics subjects. Even when I worked in BHEL and Railway Board I was given work pertaining to analytics. AIM: What do you suggest to new graduates aspiring to get into analytics space? PR: Ideally Graduates should have some kind of quantitative aptitude. However many students who have never had analytical aptitude, do get interested in the Analytics subjects later in their career. I think to get into Analytics space, people should read user friendly managerial journals and management books. Also attending conferences and seminars on Analytics would greatly help. AIM: How do you see Analytics evolving today in the industry as a whole? What are the most important contemporary trends that you see emerging in the Analytics space across the globe? PR: The interest and awareness to Business Analytics have been quite a recent phenomenon in India. Even three or four years back it did not emerge as a field of study with high visibility. A few large companies such as GE and American Express had started widespread applications of modeling procedures falling under Business Analytics, applying them to the huge databases they had created. All the same, industry by and large, though they were waking up to the fact that they had started creating extensive business databases which could be perhaps harnessed to help optimal decision making, was not really thinking of an entirely new field  managerial decision making based and  rooted on quantitative modeling, which could be termed Business Analytics. What indeed was happening in our country was widespread use of computers and internet and was leading to what can be called an explosion of data. Outside of the internet – retailers, telecom companies, healthcare, airlines, hotels and even the sports industry were collecting and analyzing massive amounts of data. Today most of the biggest international companies have centers in India and many Indian companies have sprung up starting to offer analytics to global clients. Analytics is one of the fastest growing segments in the KPO (knowledge process outsourcing) industry in our country. However, proper harnessing the data has become a challenge for businesses.  It is here that Business Analytics can help companies synthesize data into insights that can be applied for the benefit of the business. ( Rao et al,Business Analytics: A Perspective, International Journal of Business Analytics and Intelligence, Vol1, Issue 1, Publishing India Group) [spoiler title=”Profile of Prof. Purba Rao” open=”0″ style=”2″] Dr. Purba Rao has been a Professor over the last 20 years teaching and conducting research on topics such as Predictive Modeling in Strategic Decision Making, Marketing Research, Business Analytics, Green Supply Chain Management, Corporate Environmental Management, and Social Entrepreneurship Management. She is a Fellow in Management from Indian Institute of Management Calcutta, has a M.Sc. in Applied Math from Science College, University of Calcutta and a B. Sc. In Math (honors) from Presidency College Calcutta. Currently, she has been a Visiting Professor at Indian Institute of Management, Ahmedabad IIMA, Great Lakes Institute of Management and IIM RANCHI.[\/spoiler]","excerpt":"Prof. Purba Rao recently published a Book “Business Analytics: an application focus”, with PHI Learning. In an interview Analytics India Magazine, Prof. Rao talks about her motivation to write this book. To know more about the book, click here. [dropcap style=”1″ size=”2″]AIM[\/dropcap]Analytics India Magazine: How did you decide to write a book on Business Analytics? […]","categories":["AI Features"],"tags":["Interviews and Discussions","masters in data analytics in india"],"author_name":"AIM Media House","publish_date":"2013-09-15T11:08:36","publication_year":"2013","word_count":1450,"keywords":["Go","ELT","programming_languages:R","AI","programming_languages:Go","masters in data analytics in india","Aim","analytics","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","analytics","Aim","R","Go","ELT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/interview-prof-purba-rao-author-of-book-business-analytics-an-application-focus\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10162366,"title":"L&amp;T to Build Uzbekistan&#8217;s First AI-Enabled 10MW Data Centre","content":"Mumbai-based Larsen & Toubro (L&T) has announced it will construct Uzbekistan’s first AI-enabled and sustainable 10-megawatt data centre in Tashkent. The project, awarded as a “significant” contract, aims to enhance the region’s data infrastructure with advanced technology and energy-efficient solutions. According to the company’s project classification, a ‘significant’ order is valued between INR 1,000 crore and 2,500 crore, with large orders ranging from INR 2,500 crore to 5,000 crore, and so on. The data centre will feature AI capabilities designed to power next-generation applications while ensuring minimal environmental impact. It will also include robust security technologies to safeguard sensitive information. L&T will oversee the engineering, procurement, and construction processes. M.V. Satish, member of the executive committee, advisor to the chairman and MD of L&T, stated, “It is indeed a matter of great pride that L&T Buildings & Factories vertical will be executing this AI-enabled and sustainable data centre in Tashkent.” This initiative aligns with L&T’s reputation as a global leader in EPC projects, hi-tech manufacturing, and services, with over eight decades of industry experience. The company, valued at $27 billion, continues to focus on delivering innovative, quality-driven solutions across geographies. Satish also highlighted that this project underscores L&T’s expertise in executing complex and unique projects worldwide. The company has also been actively involved in building data centres in India. It has committed investments to establish state-of-the-art hyperscale data centres at multiple locations, starting with Mumbai and Chennai. Chennai’s INR 2,000 crore data centre is located in Sriperumbudur, Kanchipuram district, Tamil Nadu. The facility currently has a built-in capacity of 30 megawatts, with 12 megawatts ready for colocation services. Under the MoU signed with the Tamil Nadu government in November 2021, L&T planned to expand the data centre to 90 megawatts over the next five years. The project was expected to generate approximately 1,100 jobs, including 600 direct and 500 indirect roles. Additionally, L&T has constructed major data centres in India for both the government and private sectors and is currently operating some government data centres. The company’s group IT subsidiaries provide data centres and cloud services to numerous Fortune 500 clients, including large cloud service providers. L&T recently made big headlines when its chairman, SN Subrahmanyan, popularly known as SNS, broke the internet with his viral video. In it, he was seen asking employees to work 90 hours a week, including Sundays—a move that could only compound the company’s attrition and staff shortage issues.","excerpt":"The contract could range anywhere from INR 1,000 crore to 2,500 crore.","categories":["AI News"],"tags":["data centre"],"author_name":"Sanjana Gupta","publish_date":"2025-01-28T22:09:22","publication_year":"2025","word_count":404,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","Aim","R","data centre"],"extracted_tech_keywords":["AI","Aim","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/lt-to-build-uzbekistans-first-ai-enabled-10mw-data-centre\/","complexity_score":3,"technical_depth":6,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":18612,"title":"Winning a Kaggle ML competitions isn’t enough – learn how to take your ML learning experience forward","content":"Are you one of those Kaggle ninjas who has aced machine learning competitions but are struggling to get noticed by recruiters? A data scientist by day and Kaggler by night is not a new phenomenon. You will find scores of ML enthusiasts who have aced the Titantic competition (one of Kaggle’s easiest) and also survived the Digit Recognizer or First Steps with Julia. But there’s a major downside to Kaggle, that many data science professionals have stressed upon. We list down several missing pieces in Kaggle that undermine your machine learning understanding: Understanding business problem: In a Kaggle competition you don’t spend time to understand how a model will fit into the business problem and how accurate could it be. Types of problem: Among the vast selection of business problems that a data scientist faces in real-life, Kaggle only highlights a subset of it, shared Ria Chakraborty, a data scientist at IBM on Quora. Convincing C-level executives about your analysis: Kaggle competitions often overlook another challenging part:  convincing C-level executives to follow through the solution with a data backed strong analysis. Working on clean dataset: The real difference between Kaggle and real-world data science is that you’ll likely never be handed a pre-determined and cleaned dataset. Data wrangling forms a huge part of the data science business — mean joining datasets, cleaning up missing values and transforming data. Putting models in production: You never put Kaggle models in production which is the real test of its performance. While Kaggle is a great source of competitions and forums for ML hackathons, and helps get one started on practical machine learning, it’s also good to get a solid theoretical background. Here’s what we think: Kaggle is a great place to get started on machine learning, but at the same time one must also improve their theoretical background to fill any gap in machine learning. Now first up, let’s define a Kaggler proficient in ML – a) you have more than a basic knowledge of machine learning; b) You know how to use machine learning libraries\/packages in R, Python, Java etc. AIM gives a lowdown on how one can make the most of your Kaggle machine learning experience Build a theoretical foundation: Data science practitioners believe machine learning is a life-long commitment and most MOOCs don’t cover all the algorithms. One should go deep to fill the gaps in learning. Microsoft Azure Machine Learning: If you don’t want to sign up for a MS, then you can try free resources on Azure ML studio. You can even try out solutions from Cortana Intelligence Gallery. You can even try pushing your models into production in Azure ML. Build a machine learning portfolio: Kaggle competitions are often panned for presenting clean datasets. In fact, data wrangling is the missing piece in the puzzle, whereas in a business setting, data wrangling forms a huge part of data science — joining datasets, cleaning up missing values, transforming data\/creating new features. By creating a semi-formal work product for each project to share what was done, how it was done and what was learned (use github README files on each repo, write blog post, create PDF tech reports, power points, whatever). Here are a few project ideas: a project where you had to collect the data yourself, e.g. scraping products reviews from a website or a project where you dealt with missing or messy data, e.g. cases where some people provided their location and some didn’t. Building a machine learning portfolio will go a long way in establishing how you can complete projects. It will also equip with the confidence to take on more interesting projects as you apply your ML learning and show your skills and capabilities to recruiters once you start looking for a job. Prioritize your learning based on the application: Since machine learning is a broad field, it will be better to select a specific area of study and a Masters can help in landing a job interview as well. If you are geared towards a Post-doctorate which is also a good idea, there are a few application areas you can focus on. From Natural Language Processing to Computer Vision (think setting up GPU instances in AWS) and Deep Reinforcement Learning, there are several areas of application in ML for research. Move up the ladder with Neural Networks for ML: Want to become more technically strong, deepen your Neural Network knowledge with a course or free resources that talks about artificial neural networks and how they’re being used for machine learning, in areas such as speech recognition, image segmentation and object recognition. Kick-start your career in machine learning: Now that you have mastered some of the basics of ML, and have become a well-rounded machine learning expert, you would want to make it a full-time job. If you are proficient at ML, you would probably land a job as a data scientist and progressively move into doing more ML.","excerpt":"Are you one of those Kaggle ninjas who has aced machine learning competitions but are struggling to get noticed by recruiters? A data scientist by day and Kaggler by night is not a new phenomenon. You will find scores of ML enthusiasts who have aced the Titantic competition (one of Kaggle’s easiest) and also survived […]","categories":["IT Services"],"tags":["java project ideas","Kaggle"],"author_name":"Richa Bhatia","publish_date":"2017-10-30T08:47:55","publication_year":"2017","word_count":824,"keywords":["data science","machine learning","Kaggle","AWS","AI","neural network","ML","computer vision","Aim","java project ideas","Azure","Azure ML"],"extracted_tech_keywords":["AI","machine learning","ML","neural network","computer vision","data science","Aim","Azure ML","AWS","Azure"],"url":"https:\/\/analyticsindiamag.com\/it-services\/winning-a-kaggle-ml-competitions-isnt-enough-learn-how-to-take-your-ml-learning-experience-forward\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10103544,"title":"OpenAI End Game","content":"The big drama at OpenAI might have come to an end with the return of Sam Altman joining back the company, but a lot of damage has already been done. While Altman was away and the company was going through a great change, the competitors were busy acquiring its customers and throwing cast nets at employees. During the turmoil, ChatGPT faced down time along with API services, which forced several users to explore alternatives like Anthropic, Cohere, and many other open-source options. Nevertheless, OpenAI’s team promptly resolved the issue. There were a lot of reports that said that more than 100+ OpenAI customers were already in talks with its competitors, including Google, Amazon and Oracle’s billion dollar babies Anthropic and Cohere, respectively. Moreover, a significant number of OpenAI customers were contemplating a transition to Microsoft’s Azure services, attracted by the availability of OpenAI models and other similar  alternatives. On the other hand, executives of Salesforce and NVIDIA attempted to recruit OpenAI’s employees. Marc Benioff, Salesforce CEO said that Salesforce would match the full cash and equity OTE for any OpenAI researcher who has tendered their resignation, encouraging them to immediately join the Salesforce Einstein Trusted AI research team under Silvio Savarese. Rivals Smell Opportunity In a bid to challenge OpenAI and attract its customers, the competitors announced back to back updates. Yesterday, Anthropic launched Claude 2.1, boasting a 200K context length. It outperforms OpenAI’s GPT-4 Turbo which comes with 128K context and has significant reductions in model hallucination rates, system prompts, along with the addition of a new beta feature. Morgan Stanley, one of the earliest customers of OpenAI, is currently in touch with Anthropic. Simultaneously, Oracle’s billion-dollar baby, Cohere introduced fresh fine-tuning capabilities and an updated user interface. The new UI incorporates a testing playground for users to experiment and validate fine-tuned models. Additionally, it features a pricing calculator, empowering users to make well-informed decisions regarding fine-tuning costs from the outset. Langchain, whose existence was questioned when OpenAI launched the GPT Store at DevDay, has suddenly come to life and made a slew of announcements. For instance, Dream, one of Langchain’s products, allows users to use an AI no-code tool to build fully functional web apps and components using natural language, similar to what GPT Builder does. Not only that, it also introduced the ‘GPT-Crawler template,’ which helps in easily building chat assistants for websites. Elon Musk, one of the founders of OpenAI, has recently intensified efforts with xAI’s Grok, a chatbot developed by Musk’s team in just four months. He announced that access to Grok will be gradually opened in phases, prioritised based on users’ sign-up dates for X Premium+. Furthermore, Grok will be accessible within the X app. OpenAI got lucky Despite the turmoil, OpenAI’s team managed to stay ahead of competitors. In the midst of chaos, they rolled out a new update to ChatGPT, enabling users to interact and receive responses through voice rather than text. Moreover, as per the latest developments, it seems that none of OpenAI’s employees are planning to leave now that Altman is back and OpenAI is safe and sound. Greg Brockman posted a picture on X with OpenAI’s team for assurance. we are so back pic.twitter.com\/YcKwkqdNs5— Greg Brockman (@gdb) November 22, 2023 However, for Altman’s arrival, as per the agreement, it is mentioned that both him and Brockman won’t hold positions on the company’s board.  Altman has also committed to an internal inquiry into reported conduct that led to the board’s decision to remove him. Despite this, there are still many unanswered questions about why Sam rejoined OpenAI and the details behind his return. Meanwhile one thing is for sure during the whole OpenAI’s saga, enterprises utilising its services learned a crucial lesson that relying solely on one vendor is not a safe option.","excerpt":"The rise of alternatives","categories":["Global Tech"],"tags":[],"author_name":"Siddharth Jindal","publish_date":"2023-11-22T18:20:16","publication_year":"2023","word_count":633,"keywords":["Anthropic","ChatGPT","Go","OpenAI","AI","Azure","R","RAG","LangChain","xAI"],"extracted_tech_keywords":["AI","ChatGPT","OpenAI","Anthropic","xAI","LangChain","RAG","Azure","R","Go"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/openai-end-game\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10064233,"title":"Dr. Reddy’s appoints Anish Agarwal as Head of Analytics","content":"Anish Agarwal has joined Dr. Reddy’s as head of analytics. He will spearhead the Analytics CoE of the pharmaceutical giant: Business analytics, data science and technology (data engineering and visualisation\/apps) come under his purview. “Dr. Reddy’s Laboratories is driven by the purpose to accelerate access to innovative and affordable medication with commercial presence in 42 countries globally. I strongly believe that data and analytics will drive the next wave of innovation along with science and technology for the pharma\/life science industry. A strong analytics capability will help the organisation make better decisions about products and processes by leveraging big data, ML and AI-driven forecasting to present visual data and market trends to stakeholders,” said Anish. Anish will both set the vision and strategic plan for advanced analytics at Dr. Reddy’s. His major responsibilities include: Partner with the business leaders to create and execute analytics vision to improve decision support in critical organisational\/cross-functional processes and execution of business strategy.Bring in the best-in-class AI\/ML capabilities to bear and drive forward the organisation’s strategic priorities of growth, efficiency with continuous improvement and patient centric innovation.Build a strong internal and external network of capabilities and capacity to execute on the vision with quality and speed.Identify strategic partnerships that enhance Dr. Reddy’s in-house capabilities (data, technology or modelling). Assess the current state of digitisation and data maturity to prioritise actions (data partnerships, digitisation imperatives) that create the maximum business impact.Design and drive initiatives to integrate advanced analytics (AI\/ML) to make the core operating processes intelligent. Define high RoI strategic decision support initiatives in partnership with business leaders where advanced analytics capabilities can create exponential impact. Anish comes with 20 years of experience and has worked with major multinational corporations, including NatWest Group, Clifford Chance and Vertex Data Science. He has made it to many illustrious lists, including the Top 50 influential AI Leaders in India, the Top 40 under 40 Data Scientists and the Top Five Data Science mentors in India. Anish also has a Ph D in Computer Science.","excerpt":"Anish has worked with MNCs, including NatWest Group, Clifford Chance and Vertex Data Science.","categories":["AI News"],"tags":["head of analytics","head of data science"],"author_name":"AIM Media House","publish_date":"2022-04-04T12:29:09","publication_year":"2022","word_count":335,"keywords":["big data","data science","AI","ML","Git","head of analytics","RAG","data engineering","analytics","GAN","head of data science","R"],"extracted_tech_keywords":["AI","ML","data science","analytics","RAG","R","Git","big data","data engineering","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/dr-reddys-appoints-anish-agarwal-as-head-of-analytics\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10041919,"title":"This Jaipur Startup Uses AI To Offer Customised Video Conferencing Solutions","content":"Jaipur-based VideoMeet, founded in 2018, is a home-grown video conferencing platform with multi-level encryption. “VideoMeet allows users to initiate and experience AI-enabled, high-definition, multi-point audio or video conferencing solutions. We are committed to delivering end-to-end enterprise business video conferencing solutions to the users, which is flexible, user-friendly and highly secured with single-click access on our application,” said Ajay Data, Founder, VideoMeet. Services “The AI is established at multiple points in VideoMeet, for each unique user to have a more intuitive & user-friendly experience. The auto-adjustment of HD quality provides the user staying connected on high-quality video conferencing mode without any lag and zero manual intervention,” he said. The platform supports multiple features and provides flexibility to users to create multiple rooms based on meeting requirements – role-based, client-based, product-based, internal meetings etc. The room once created will stay active forever. Meaning, a user does not have to create meeting ids again for future get-togethers. Additionally, VideoMeet can bring video conferencing solutions through API integration to allow embedding into users’ ERP or CRM or any other solution. Further, the app may act as a personal assistant by checking, authorising access, and notifying visitors of their time in and out. Another feature, White Label Branding, allows companies to have their logo and personalised background as well. “VideoMeet is the only Video Conferencing Platform to offer all features without capping\/restriction and no additional charges. So, practically we give freedom to the user to utilise features for his requirement at nil price,” Ajay said. Users have the option to place meetings links on their webpage. In short, clients can come directly to the webpage to schedule the meeting. Users can also opt for customised products. Tech Stack WebrtcReal-Time Communication with AESSymmetric encryption for secured real-time video meetingsWeb UI in React JSRedux Challenges “We love facing challenges; they help us know our customers’ requirements more and to serve them better. We are the only Made in India video conferencing platform across the nation that does customisation. We make sure to offer a bouquet of services for our customers to choose from and this you can say is a challenge which we are happily undertaking every day,” Ajay said. Roadmap Post-lockdown, the VideoMeet started receiving business queries from three specific areas, including: Edtech – With WFH, the innovation in learning management with Augmented Reality (AR) is picking up, and VideoMeet has achieved AI proctor-based examination over the web.Patient management system: Doctors are doing consultation over call – where a doctor can attend his patient and consultation can happen simultaneously. This bridges the communication gap in healthcare.Virtual Shops: We provide a virtual event platform, ‘VNOW’, with our existing VideoMeet at the backend – where a retailer can create his virtual shop and sell his products live. VideoMeet has over 0.5 million regular users. “They primarily begin using VideoMeet as a free medium for communication, and after being accustomed, they are keen on using and applying for paid services as well. VideoMeet is the only end to end Video-Conferencing platform with all features and no capping that makes it the most economical solution of India,” he said. “Being a homegrown application, with capabilities to surpass the global video conferencing apps, we aim to become India’s most preferred Video Conferencing platform for personal and professional meetings,” concluded Ajay.","excerpt":"Jaipur-based VideoMeet, founded in 2018, is a home-grown video conferencing platform with multi-level encryption. “VideoMeet allows users to initiate and experience AI-enabled, high-definition, multi-point audio or video conferencing solutions. We are committed to delivering end-to-end enterprise business video conferencing solutions to the users, which is flexible, user-friendly and highly secured with single-click access on our […]","categories":["AI Startups"],"tags":["AI Startups"],"author_name":"kumar Gandharv","publish_date":"2021-06-17T16:00:00","publication_year":"2021","word_count":549,"keywords":["Go","API","programming_languages:R","AI","RPA","innovation","programming_languages:Go","Aim","R","AI Startups"],"extracted_tech_keywords":["AI","Aim","R","Go","API","RPA","innovation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-startups\/this-jaipur-startup-uses-ai-to-offer-customised-video-conferencing-solutions\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10132383,"title":"‘It&#8217;s Extremely Important that Future of Tech is Shaped by Democracies and their Partner Countries,’ says Rajeev Chandrasekhar","content":"Paul Buchheit, the creator of Gmail recently expressed concerns about the potential dangers of AI development in countries like China. Buchheit warned that if China leads the AI race, the world could face a permanent lockdown, with even our thoughts under surveillance and censorship. Responding to Buchheit’s remarks, Rajeev Chandrasekhar, Former Union Minister of India, emphasised the importance of ensuring that democracies and their neighboring nations shape the future of technology. He stated, “Its extremely important – more than critical – that future of Tech is shaped by  democracies and their partner countries” https:\/\/twitter.com\/rajeevrc_x\/status\/1822501999969181713?s=12 There is no denying that despite pressing concerns about the ethical and privacy implications of AI, China remains a central player on the global stage. China’s Approach to AI Development Source: Reddit In October 2023, Chinese tech giant Baidu unveiled Ernie 4.0, the latest version of its generative AI model, claiming capabilities comparable to OpenAI’s GPT-4. This significant advancement highlights China’s rapid progress in AI, driven by substantial government support and strategic initiatives like the ‘New Generation Artificial Development Plan’. Launched in 2017, this plan aims to position China as a global AI leader by 2023, emphasising research funding, talent recruitment, and infrastructure development. These efforts underscore China’s commitment to dominating the AI sector, often pushing the boundaries of ethical and privacy considerations. China’s AI development is also fueled by leading companies including Tencent, Alibaba, Baidu, and SenseTime, which attract top group talent and drive innovation. Meanwhile, China’s strides in hardware and robotics, with companies like Dreame Technology and Fourier Intelligence at the forefront, reflect a comprehensive approach to AI applications. Additionally, China’s advancements in facial recognition technology, widely deployed in public spaces, illustrate the country’s capability to implement AI solutions on a large scale. These often move rapidly as ethical concerns are not always given paramount importance.","excerpt":"Paul Buchheit, warned that if China leads the AI race, the world could face a permanent lockdown.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","China"],"author_name":"Vidyashree Srinivas","publish_date":"2024-08-13T11:21:44","publication_year":"2024","word_count":302,"keywords":["Go","API","funding","OpenAI","AI","innovation","GPT","Aim","generative AI","AI (Artificial Intelligence)","R","China"],"extracted_tech_keywords":["AI","generative AI","OpenAI","Aim","R","Go","API","GPT","innovation","funding"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/future-of-tech-is-shaped-by-democracies-and-their-partner-countries-says-rajeev-chandrasekhar\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10071306,"title":"DALL.E is now officially available in beta","content":"AI research company, OpenAI has announced that DALL·E, an AI system that creates realistic images and art from a description in natural language, is now available in beta. This comes just days after OpenAI founder, Sam Altman, tweeted to make DALL.E accesible to 1 million users. OpenAI is initiating the process to invite people from their waitlist to give access. As a part of this initiative, every user will get 50 free credits during the first month of their use and 15 free credits every subsequent month. Additionally, users can buy credits on top of the freely available monthly credits for USD 15 to get 115 credit increments in the first phase of beta. Each credit can be used for one original DALL·E prompt generation  or an edit or variation prompt. DALL·E generates four images for every natural language prompt and three images for edit and variation prompt. Now, users will be able to use images created with DALL.E for commercial purposes like  illustrations for children’s books, art for newsletters, concept art and characters for games, mood boards for design consulting and storyboards for movies. The usage rights include the right to reprint, sell, and merchandise. Keeping accessibility in mind, OpenAI is providing an option for artists who need financial assistance to apply for subsidised access. OpenAI has taken necessary steps in collaboration with researchers, artists and developers to curb misuse, reduce bias and prevent harmful images.","excerpt":"Now, users will be able to use images created with DALL.E for commercial purposes.","categories":["AI News"],"tags":["AI Tool"],"author_name":"Zinnia Banerjee","publish_date":"2022-07-21T11:02:27","publication_year":"2022","word_count":237,"keywords":["OpenAI","AI","programming_languages:R","ViT","AI research","AI Tool","R"],"extracted_tech_keywords":["AI","OpenAI","R","ViT","AI research","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/dall-e-is-now-officially-available-in-beta\/","complexity_score":2,"technical_depth":6,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10015231,"title":"Tech Giants Form Alliance For IT Tools &#038; Transformation. How Does That Matter?","content":"In recent news, tech giants such as Google, Intel and Dell, among others in the IT and cloud computing space, have joined hands to form a consortium. Named as the Modern Computing Alliance, this consortium will tackle problems related to security, remote working environments, and other enterprise issues. Speaking of its participation in the alliance, Google Chrome Vice President John Solomon wrote in a blog post, “Our collective mission is to drive ‘silicon-to-cloud’ innovation for the benefit of enterprise customers, fueling a differentiated modern computing platform and providing additional choice for integrated business solutions.” Solutions For Enterprise Computing Market As per the alliance, the initial focus areas will include performance, security and identity, remote work and productivity, and health care. The end goal would be to pool knowledge and resources to identify shared problems across enterprises and how cloud solutions can be leveraged to ease the whole ‘new normal’ in working environments. The alliance also looks at establishing new standards and technologies from the partner companies that can be used by anyone. Solomon also mentioned in the blog that the alliance would look at employing faster and more responsive enterprise progressive web applications (PWAs). It also aims at introducing cloud-first cultures and devices that will provide ease of management and insights on the fully integrated stack. “Modern Computing Alliance is committed to developing an integrated roadmap that makes the best use of our collective experience, insights and expertise while giving us a clear path forward to improve customer choice and the enterprise computing market,” he added. The alliance, which also has companies such as Box, VMWare, Zoom, Slack, RingCentral, Citrix, Okta as its founding partners, has developed a roadmap to deal with the most pressing challenges in end-user computing. Some of the solutions suggested by the alliance include — Investing and developing technologies for a more seamless and user-friendly experience across devices. It also plans on creating advanced cloud data security and investing in improving data loss prevention solutions, given how rapidly enterprises are adopting cloud in lieu of the worldwide pandemic.Using telemetry, the alliance will work at enhancing the productivity of companies’ workforce that are quickly adopting a more distributed way of working. The solutions will provide silicon-to-cloud telemetry insights and analytics. It will also provide recommendations to optimise workflow and introduce automation wherever viable.To maximise resources on the healthcare front and to provide healthcare providers with better ROI (return on investment), this alliance will enhance hospital settings to optimise productivity, efficiency, and authentication. This will potentially give doctors more time to actually assess their patients’ problems rather than wasting time operating IT systems. The founding stones for this alliance were laid even before the coronavirus pandemic. In an interview, Solomon said that the initial discussion for forming such a consortium was conducted during the Consumer Electronics Show in 2019 at Las Vegas with limited partners. However, with the onset of the pandemic, the companies decided on expanding their scope further and include partners like Slack and Zoom, which has gained massive prominence in how corporates were operating now. The group is also inviting ‘as many IT professionals as it can’ to build a council of experts who would identify and resolve major problems. To further attract the best talent, the alliance has also listed ‘benefits of joining the council’. They include: Entry to early-access programs and an opportunity to network with members of management teams across companies.Invitation to participate and contribute actively to the outcomes of the alliance roadmap and provide input and test new solutions.Opportunity for peer-to-peer networking and mentorships with executives from leading companies. Wrapping Up Cross company collaboration and alliance as this one is not rare. The earlier examples of similar collaborations include the 1991’s Advanced Computing Environment consisting of Compaq, Microsoft, MIPS Computer Systems, Digital Equipment Corporation, and the Santa Cruz Operation, and the AIM Alliance consisting of Apple, IBM, and Motorola. The main product from Advanced Computing Environment was the introduction of the Advanced RISC Computing specification, just before the alliance finally dissolved due to infighting. The AIM Alliance was slightly more successful, and it developed the PowerPC CPU family, the Common Hardware Reference Platform (CHRP) hardware platform standard, and also laid the foundation for Apple’s Power Macintosh computer line. It seems that the new Modern Computer Alliance is led by Google as of now and it would be interesting to see what comes off this collaboration.","excerpt":"In recent news, tech giants such as Google, Intel and Dell, among others in the IT and cloud computing space, have joined hands to form a consortium. Named as the Modern Computing Alliance, this consortium will tackle problems related to security, remote working environments, and other enterprise issues. Speaking of its participation in the alliance, […]","categories":["Global Tech"],"tags":["expert system examples","Google","how does ai work","Intel"],"author_name":"Shraddha Goled","publish_date":"2020-12-22T13:00:00","publication_year":"2020","word_count":734,"keywords":["Go","API","AI","cloud computing","how does ai work","ML","Git","RAG","expert system examples","Aim","analytics","Google","R","Intel"],"extracted_tech_keywords":["AI","ML","analytics","Aim","RAG","cloud computing","R","Go","Git","API"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/tech-giants-form-alliance-for-it-tools-transformation-how-does-that-matter\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10063747,"title":"Organisations should invest in a good BI platform: Naren Vijay, EVP at Lumenore","content":"Today, the biggest challenge for any organisation is to rapidly parse through enormous amounts of data and drive out the information that will deliver key insights and make more informed business decisions. Business intelligence is one such game-changer process that can help businesses use their data to their advantage by presenting the otherwise unusable pile of data into an understandable and interpretable form. Lumenore, a Netlink platform, has been helping organisations to simplify the data and make business intelligence more actionable. A leader in the augmented BI platform market, Lumenore is designed to assist businesses to turn their data repository into actionable organisational intelligence that anyone across the organisation can leverage. The platform provides a no-code business intelligence platform that is focused on performance management. The power of conversational intelligence and predictive analytics bridges the gap between the available data and enables insight-informed decision-making. Lumenore platform helps organisations create an all-in-one data universe that turns raw data into an organisational business intelligence repository. Naren Vijay, the Executive Vice President of Lumenore, is responsible for driving revenue growth and customer acquisition at the company. He holds a Bachelor of Engineering (Mechanical) degree from Ranipettai Engineering College, Vellore and co-founded Enterprise Touch, a bootstrapped services company in mobile application development. Working in the organisation, he quickly realised his potential in the low code space and created a Centre of Excellence with an expertise-rich and futuristic outlook. Analytics India Magazine got in touch with Naren to discuss the intricacies of business intelligence and understand how it makes a difference in many organisations. AIM – Define business intelligence in your own words and explain how BI makes a difference in the company. Naren – Business intelligence (BI) is a technology-enabled method of analysing raw data and drawing insights. It combines business analytics and consolidated and engineered data to help organisations make more data-driven decisions. Every company has vast data, and BI can help unlock its value by making valuable inferences. The organisation’s full data can be housed in a data warehouse or smaller data marts. BI tools can help organise data, prepare it for analysis, perform queries, and provide data visualisations and business insights, whether the data is real-time or historical. One of the most significant benefits of BI is that it allows businesses to determine what their customers want and how best to meet their needs. It is a tool for boosting team productivity and accelerating corporate growth. AIM – How augmented BI platform can help organisations go beyond reporting, querying and diagnostic analytics? Naren – Artificial intelligence (AI) and business intelligence are combined in augmented analytics. Traditionally, organisations had to rely on data scientists to analyse isolated datasets and draw inferences using basic charts to make sense of any data. Businesses can use augmented BI, powered by AI and Machine Learning, to automate data preparation for analysis by labelling and structuring data. This method makes data management easier and allows users to draw appropriate conclusions. It enables businesses to analyse their options by combining decision intelligence, tracking the outcome and changing their approach to achieve positive outcomes. AI also helps analytics by making the process of finding new insights easier. Without a specific user question, it can show notable trends and insights. Lumenore’s anti-dashboard ideology prioritises the development of intelligence over dashboards. By combining advanced analytics, Google-like querying, and AI-powered suggestions, we hope to help organisations make better decisions. As a result, data complexity is reduced, and organisations have quick and easy access to the insight they need to make key business choices. AIM – How does business intelligence help in organisation decision making? Could you cite an example of a company that Lumenore worked with helping them with BI and elaborate on how it made a difference? Naren – Earlier, organisations dealing with data would focus on collecting and maintaining various spreadsheets and dashboards. However, with time, they have realised that this is not adequate. Moreover, the amount of data that the organisations have has grown exponentially in the last few years. Business intelligence helps companies make informed decisions on strategic issues by providing crucial information on the current and historical performance of the company along with future trends, expected demands, customer behaviour and a few other factors. Our client, one of the largest broadcasting companies based out of the United States, managed over 12,000 contact centre employees. They were hired to support their customer support and sales department. However, as these vendors started using unfair practices to retain the customers, the client relayed support from Lumenore’s Agent Outlier Management Solution to bring transparency and enforce accountability into their agents’ actions. By providing services like data integration, voice-driven instant insights and providing hidden insights using advanced data discovery, our BI platform was able to correct the agent behaviour. This, in turn, helped reduce inappropriate cross-sell\/up-sell, call hang-up scenarios and call transfer rate. AIM – How can companies bridge the distance between data and decision making? ​Naren – Organisations should invest in a good BI platform to successfully bridge the gap between data and decision-making. An organisation’s current and short-term goals should be considered while selecting a platform. The tool should be affordable, easy to use, and available to all users. It should deliver real-time insights and take into consideration all of the data. AIM – What is the importance of conversational analytics and the emerging trends around conversational and predictive analytics? ​Naren – Conversational analytics assist firms to navigate through data, extract the right information, and derive context from it in the context of business intelligence. Any user in the organisation may acquire real-time information in a flash with the aid of conversational analytics. It not only saves time but also increases productivity. Users may go beyond the fundamental conclusions with the help of a good BI platform and conversational analytics. They can get more thorough explanatory insights and create relevant visualisations. In other words, without having to code, one can acquire customised, actionable information in seconds. It not only assists in the retrieval of relevant data but also makes smart recommendations. The practical effect of business intelligence and big data is predictive analytics. Many organisations are already using predictive analytics to apply ML & AI algorithms for data mining and future forecasting trends. AIM – What is Lumenore’s suite of applications, and how is it helping organisations make sense of all the data they have at their disposal? ​Naren – Lumenore is a powerful, intuitive, cloud-based BI and analytics platform that delivers organisational intelligence by sifting data from any business application. Our aim is to simplify decision-making and democratise data across the organisation. We offer seamless data management by creating Data Universe, which can then be structured to deliver actionable insights. Furthermore, our no-code facilitates interactive visualisations and conversational querying. This, in turn, streamlines decision-making. At Lumenore’s, our most popular innovations are “Ask Me” and “Do You Know” modules, as they provide instant analytics. These tools are easy to use and provide instant answers to the query. Another advantage of these tools is that one does not have to be data literate to drive meaningful information via those. Since these features are an outcome of augmented analytics, they can even provide a complete list of probable actions and suggest the best course to achieve the organisational target. AIM – What is the current scenario for big data & analytics, and what are the key trends driving the growth in this segment? ​Naren – We live in a data-driven era. Data comes in from a variety of places in any organisation. According to industry estimates, the global data sphere will contain 175 zettabytes of data by 2025. The biggest difficulty organisations face right now is coping with this massive amount of data that is expanding by the second. They are always seeking new solutions to bridge the gap between data and analytics that are both productive and cost-effective. Organisations may use business intelligence technologies to unlock the power of this data and make quick, data-driven decisions. Data solutions are projected to be in the driver’s seat for the next few years. As a result, businesses will continue to invest in a system that allows them to stay ahead of the curve while objectively reviewing their decisions. Trends such as augmented analytics that provide automated data discovery using machine learning\/statistical approaches, narrative insights that provide self-explanatory natural language, and the ease of use of analytics platforms are projected to boost the industry’s growth. AIM – What are the key business intelligence trends in 2022? Naren – Data is available for users using the correct business intelligence tools. They reduce computing time by breaking down the necessary skills. Previously, companies had to rely on specialists to do data mining operations. However, executives of any level can use these technologies to do analytical duties such as reviewing data sets or executing data-mining tasks. Organisations also recognise the necessity of performing analytics on the entire data set rather than just a subset of it. As a result, businesses have begun to use the data universe, which includes data management designs such as data engineering, data lakes, and data security. Data storytelling and visualisation are becoming increasingly popular among businesses because they bring data into context. These are easier to understand because they use interactive reports and explanatory images to create a story around key data. Rather than simple graphs or a block of text, this information is easier to absorb and enables speedier decision-making. Today’s businesses recognise the importance of data and want to do more than merely collect and mine it. They invest in technologies that help them improve their analytics, performance, and other crucial indicators by utilising high-quality data.","excerpt":"Lumenore’s most popular innovations are “Ask Me” and “Do You Know” modules; they provide instant analytics; these tools are easy to use and provide instant answers to the query.","categories":["AI Features"],"tags":["Business Intelligence","Interviews and Discussions"],"author_name":"Poornima Nataraj","publish_date":"2022-03-28T15:00:00","publication_year":"2022","word_count":1616,"keywords":["Go","machine learning","artificial intelligence","AI","R","ML","RAG","Aim","analytics","Business Intelligence","predictive analytics","Interviews and Discussions"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","analytics","Aim","RAG","predictive analytics","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/organisations-should-invest-in-a-good-bi-platform-naren-vijay-evp-at-lumenore\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10162582,"title":"Experience Your ‘Nova Moment’ – SmartQ Presents Deva Nova, a Masterclass on AI-led Enterprise Software Development","content":"Following the success of the last webinar session on ‘Building Enterprise Software Solutions at Lightning Speed with AI Agents’, the SmartQ team is back with its ‘Masterclass on AI-led Enterprise Software Development’. SmartQ’s upcoming in-person meetup, in partnership with AIM Media House, promises a ‘Nova moment’ for developers looking to amplify their brilliance and accelerate their careers. The experience aims to be just as transformative as the powerful celestial event Nova – an explosion of brightness that transforms the night sky. The highlight of this summit is Deva.ai, a cutting-edge AI tool designed to supercharge development speed, optimise systems, and propel you to the forefront of innovation. Register Now Date: Thursday, February 13, 2025 Time: 4.00 PM Location: SmartQ, 8th Floor, Bengaluru Office Why Attend? Gain practical, actionable insights, and strategies for software development. Unlock the power of AI-led development – gain hands-on experience with Deva.ai and learn how to 10x your coding speed. Engage with top industry experts – discover how AI is reshaping enterprise software development. Network with like-minded developers – connect with professionals who share your passion for innovation. Solve real-world challenges and get shortlisted for hiring opportunities. The masterclass will include hands-on coding challenges using AI-powered tools, direct mentorship and insights from AI industry leaders. “Get 1-month free Deva.ai tool access with 20,000 credits! Plus, unlock an exciting job opportunity with SmartQ!” Masterclass Agenda Setting up the environment for agentic flow Coding challenge using AI – projects incorporating agentic flow Organising code for AI agents Mitigating fuzziness in AI agents Accelerating UI development with AI-guided brand guidelines Context grounding for AI models Register Now Date: Thursday, February 13, 2025 Time: 4.00 PM Location: SmartQ Bengaluru Office Save the date – Don’t miss out! Seize this opportunity to become a 10X smarter engineer.","excerpt":"Following the success of the last webinar session on ‘Building Enterprise Software Solutions at Lightning Speed with AI Agents’, the SmartQ team is back with its ‘Masterclass on AI-led Enterprise Software Development’. SmartQ’s upcoming in-person meetup, in partnership with AIM Media House, promises a ‘Nova moment’ for developers looking to amplify their brilliance and accelerate […]","categories":["AI Highlights"],"tags":["10X smarter engineer","AIM Webinar","Deva Nova","Developers","Masterclass","SmartQ","SmartQ Deva Nova","SmartQ In-person meetup"],"author_name":"Mohit Pandey","publish_date":"2025-01-31T12:19:41","publication_year":"2025","word_count":296,"keywords":["Deva Nova","API","AIM Webinar","AI","programming_languages:R","innovation","Masterclass","SmartQ Deva Nova","GAN","Aim","AI agents","emerging_tech:edge AI","edge AI","SmartQ In-person meetup","SmartQ","10X smarter engineer","R","Developers"],"extracted_tech_keywords":["AI","Aim","edge AI","R","API","GAN","innovation","AI agents","programming_languages:R","emerging_tech:edge AI"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/experience-your-nova-moment-smartq-presents-deva-nova-a-masterclass-on-ai-led-enterprise-software-development\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10008938,"title":"Hands-On Guide To Develop Speech To Text Converter Using Python and Google API","content":"“Hey Siri”, “okay Google” and “Alexa” is something we say almost every day to quickly get information without having to type in the search box. These devices are great to listen and understand your voice and give a suitable output. How do they work? They are designed in a highly efficient speech recognition software that can understand multiple accents and a natural language processing algorithm to convert this speech into text. But before these smart devices find the information you asked for, they need to understand what you are saying. Let us implement a speech to text converter using Python and a google API. In this article, we will build a simple speech to text converter with Python and the google cloud API. What is speech recognition and how does it work? Speech recognition is a system that translates the language being spoken into text format. To do this, a deep learning model is used that takes in audio signals, analyses them and converts them into the corresponding text. Above is the workflow of the google API for converting speech to text. It takes in the voice input from the user device and this is sent to some of the core cloud functions. These functions perform internal processes like converting the audio input into signals and preprocessing them. Then, it is sent to the speech to text API which applies a deep learning model and understands what the user is trying to say. Finally, it is passed to the autoML NLP where the speech signal that is understood by the deep learning model is converted into text format and the output is displayed. Features of the speech to text API Streaming speech to text in real-time: the API is capable of processing real-time audio signals from the device microphone or take an audio file as input and convert it into text also. Different models based on the domain: you can choose from different trained models depending on the requirements of the project. For example, for converting audio from a telephone, the enhanced phone call model can be used. Adaptation: you can customize the API to understand rare words, currency, numbers etc by making these as additional classes. Implementation with microphone Now that we know how the Google API works we will put it to use and activate the microphone in the system and convert it into text. Installation Before we get into the implementation, you will have to download the library with the pip command. pip install SpeechRecognition Next, we can use the API and write code to build the speech to text converter in real-time for the English language. Code We will first import the library and activate our microphone as follows: import speech_recognition as sr recognizer = sr.Recognizer() with sr.Microphone() as inputs: print(\"Please speak now\") listening = recognizer.listen(inputs) print(\"Analysing...\") Now that we have the input ready it is time to call the Google API to recognize the speech and display the text. try: print(\"Did you say: \"+recognizer.recognize_google(listening)) except: print(\"please speak again\") Output: Another interesting thing about this is the number of languages it supports. It not only supports common languages of the world but also supports multiple Indian languages as well. Since I speak Kannada, I will include one small change in the code and display the output in Kannada. Implementation of speech to text in Kannada To make the API understand the language and give the output, just make the following changes to the code. import speech_recognition as sr recognizer = sr.Recognizer() with sr.Microphone() as inputs: print(\"Please speak now\") listening = recognizer.listen(inputs) print(\"Analysing...\") try: print(\"Did you say: \"+recognizer.recognize_google(listening,language = \"kn-IN\")) except: print(\"please speak again\") Output: Implementation on an audio file We saw how to use the API in real-time with the microphone for English and Kannada language. But what if we have an audio file and want to extract text from that? To do this, first, let us select the audio file. import speech_recognition as sr recognition = sr.Recognizer() with sr.AudioFile('myvoice.wav') as inputs: file_audio = recognition.listen(inputs) try: convert_text = recognition.recognize_google(file_audio) print('Analysing...') print(convert_text) except: print('could not hear') As shown the contents of the audio file is displayed as text. Conclusion In this article, we saw how to make use of the Google API to convert speech to text using the microphone in English and Kannada and using an audio file as well. This can be really useful in natural language processing projects for handling audio files and transcripts as well.","excerpt":"In this article, we will build a simple speech to text converter with Python and the google cloud API.","categories":["Deep Tech"],"tags":["Python","Speech Analytics","Speech Recognition","speech-to-text"],"author_name":"Bhoomika Madhukar","publish_date":"2020-10-06T14:00:51","publication_year":"2020","word_count":742,"keywords":["Go","API","TPU","Speech Analytics","AI","R","ML","Python","NLP","deep learning","Speech Recognition","cloud_platforms:Google Cloud","speech-to-text"],"extracted_tech_keywords":["AI","ML","deep learning","NLP","TPU","Python","R","Go","API","cloud_platforms:Google Cloud"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/hands-on-guide-to-develop-speech-to-text-converter-using-python-and-google-api\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":36796,"title":"AI Is The Global Force That Will Transform The Indian Workforce &#038; Create Jobs","content":"Artificial Intelligence has majorly shifted the jobs statistics debate. At one point, while the world mulls over 30% of all jobs being siphoned off by 2030 ( a McKinsey report indicates) there is also an upside to AI. Despite all talks of doom and gloom, AI-led automation will also generate a new set of jobs that workforce has to be ready for, thereby ushering in a skills-based economy. It is very important to keep pace with change and not become irrelevant. A study indicates that 14% of the workforce will need to change their skillset to stay competitive. And the field is not restricted to people with a background in Deep Learning only, everyone can participate and take part in this journey. “It is not just technical expertise that is alone required, there is a lot of business knowledge that also comes with it. The question that most business leaders now face is how can one build capabilities in the organisation in order to drive this wave of digital transformation,” a speaker shared this thought at the Rising 2019. On the other hand, we see more and more businesses experimenting with digital technologies and implementing large-scale projects. And the question companies are bracing with is — do we hire or reskill? In our recent study on the AI landscape in India, AIM provided a well-rounded view of the AI and Data Science ecosystem in India and how the industry has grown by close to 30 per cent in the last one year and has proven to be a competitive advantage for businesses across functions, such as sales, marketing, finance, customer service and IT. The mature and well-funded AI ecosystem indicates that the startups and companies can become the next generation of market leaders and also create new job roles in this domain. Our recent report clearly highlighted how AI has led to unprecedented potential for value creation for organisations of all sizes. “In fact, the fast-paced growth of AI makes it more conducive for younger businesses to leverage the technology earlier than blue-chip companies,” Sumeet Bansal, CEO & Co-Founder of Analytixlabs, emphasised in the recent report. With the buzz in AI increasing, there has been a surge in AI-centric roles with close to 57% of organisations looking for candidates with over five years of experience in AI. Findings indicate that the top skill-set the employers look for is Machine Learning, Natural Language Processing, Neural Networks, Analytics, Cloud Computing & Pattern Recognition. Given the current scenario, AI has opened plenty of job opportunities at mid and senior level. What’s in it for software engineers and IT professionals who are at a crossroads in their career and want to pivot to AI and machine learning roles. Tops-down Approach to Upskilling Now, here’s a fundamental challenge for India’s workforce in an AI-led economy that has developed new roles, spawned new skills and even the need to work in agile settings. With startups and companies gearing towards an AI technology stack and building products with machine learning, leading tech giants like Google, Microsoft and Amazon are open sourcing developer tools to empower the developer community to do more with AI and at the same time gain a competitive advantage. On the other end of the spectrum are industries that are forging tie-ups with AI service providers to leverage AI capabilities across their products and services. So, what’s of importance here for IT professionals is to tap into new insights and understand the evolving AI tools landscape. Couple of challenges that India’s workforce face in an AI-era Increasingly, the workforce is being tasked with your employees with ready-to-use AI enabled tools supporting their day-to-day work to increase productivity Building products with ML and bake it into existing services Outlook There is a significant shift in mindset when it comes to reskilling — both from the organisational end and professionals who are taking the study route to level the playing field. There is a marked shift  from upskilling a few isolated people in a core data science team to corporate data literacy that has become the new mandate in organisations. In order to stay competitive, organisations are now realizing that to gain competitive advantage from data, every employee must become data literate and “data smart”. This in turn has driven a strong demand for AI and analytics training across  functions with senior leadership teams also coming on board for a hands-on training on how to deal with data. This is followed up with more specialised and advanced training for ML teams. With machine learning becoming core to enterprises, companies need a strategic approach on how to get started with artificial intelligence.Partnering with stakeholders like AnalytixLabs that also have an entrenched consulting practice can help firms and mid-sized organisations embed AI capabilities. From prototyping to productionizing, Deep Learning has gradually become mainstream. Irrespective of industry, this year with every client we have witnessed applications of Deep Learning gaining impetus and deliver effective solutions to some of the most complex business problems, shared Bansal. For those who want to discover more about Deep Learning and how to implement popular DL frameworks such as Keras and Tensorflow, check this hands-on task-oriented course aimed at IT professionals and aspirants. Also, if you are interested to learn about job opportunities in AI from industry thought leaders and how you can get a start in this field, check out this upcoming meetup organised by AIMinds & AnalytixLabs in Gurugram on April 5, 2019. Click here to register!","excerpt":"Artificial Intelligence has majorly shifted the jobs statistics debate. At one point, while the world mulls over 30% of all jobs being siphoned off by 2030 ( a McKinsey report indicates) there is also an upside to AI. Despite all talks of doom and gloom, AI-led automation will also generate a new set of jobs […]","categories":["AI Trends"],"tags":["AI Jobs"],"author_name":"Richa Bhatia","publish_date":"2019-03-25T08:28:13","publication_year":"2019","word_count":915,"keywords":["data science","artificial intelligence","machine learning","AI","neural network","ML","AI Jobs","Aim","deep learning","analytics","TensorFlow"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","neural network","data science","analytics","Aim","TensorFlow"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/ai-is-the-global-force-that-will-transform-the-indian-workforce-create-jobs\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":6,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":11382,"title":"Top 10 Data Scientists in India – 2016","content":"Last year we took this initiative to identify Data Scientists in the country that are helping to push the industry forward. It is a grueling exercise of sorts; we invite nominations from various organizations, irrespective of size and nature of work. We also get in touch with data scientists that we know personally who might not necessarily be associated with an organization. Then an expert team of leaders, past data scientists and evangelists are asked of their opinions on those nominations, on basis of parameters like pedigree, patents, competitions participated, any pioneering work etc. Yet, at the core of it, we look for that one significant consideration- what is the contribution to the industry and ecosystem in the country. So, here we are with this year’s list. We have purposely kept the list different from last year’s (given the size of analytics industry in India). We hope the list brings you some inspiration and motivation to find bigger purpose in life. Abhimanyu Dasgupta Abhimanyu specializes in the design, development and deployment of data science algorithms in the financial services sector. An honors graduate in Statistics from the Indian Statistical Institute and on the cusp of being a designated Actuary with the Casualty Actuarial Society of USA. He leads the FSI practice of Advanced Analytics professionals at Deloitte Consulting and has grown the practice from its nascence in 2006 when he’d joined as a rookie. His deep expertise in both analytics and actuarial science uniquely positions him as a key thought leader in the insurance analytics domain in US and India. As a young data scientist, he was instrumental in helping develop the first ever claims predictive model in US. This pioneering approach is a pending patent in his name. He is the only data scientist to be honored with Deloitte’s prestigious RMP award, an award that recognizes the top 1% consultants. Abhimanyu is an active contributor in external forums including events organized by the IAI and ICADABAI where he has spoken on the advent of advanced analytics and its applications in various fields including Underwriting, Pricing and Claims Management. Among other roles, he also serves on the Advisory Committee on Analytics of IAI today. Dinkar Sathe Dinkar Sathe is Advisor, Analytical Consultant in SAS. He has around 26 years of diversified experience in quantitative analysis, problem-solving and research using statistical techniques. While studying at Purdue University he was tasked with optimizing the input levels required to create cost effective medicine tablets. After graduating with a MS Statistics degree, he started working for a leading Insurance company where he designed and developed a provider profiling system to identify potentially fraudulent or abusive patterns of medical practices. He was also instrumental in developing predictive models in the analysis of historical loss experience of automobile risks in order to identify risk characteristics that are correlated with losses. He worked in the Insurance industry for six years after which he worked for a leading credit card company in the USA where he became proficient in the credit card business. He has gained a rich Marketing Analytics experience during consulting projects executed for a number of major corporations across the globe. Customer satisfaction has always been his foremost focus and aim. He has provided strategic consulting support to credit card program managers in areas of reporting, marketing plan fulfillment, measurement of effectiveness of past marketing efforts and in the identification of customer profiles for future promotional marketing. Manish Gupta Manish is an advanced analytics professional with experience in building & leading data science & analytics teams for developing competencies in customer analytics, real-time recommendation system, bigdata and advanced analytics solutions across various industries. Manish holds a PhD from IIT Delhi in the area of Machine Learning with more than 15 research publications in international journals and conferences with 1 US Patent which have more than 130 citations. He recently joined American Express as Director, Global Credit Data Science and Machine Learning Research to carry out business specific data science and machine learning research and to establish machine learning best practice for business. In his most recent role, Manish served as Senior Vice President, Analytics at InfoEdge, parent company of Naukri, JeevanSathi, 99Acres, Shiksha etc. He has also previously worked with Global Decision Management, Citibank as analytic lead. He has also work as Head (R&D) at Innovation Labs@24\/7 Customer where he developed patented technologies for chat categorization and web acquisition engines. He has served country as Scientist at DRDO, a premier defense research organization in India. He received several awards including Scientist of the Year, Technology Award for his contribution to develop state of the analytics solutions for armed forces. Nishant Chandra Dr. Nishant Chandra leads the AIG Science R&D group in India where he develops natural language, text mining, and machine learning models for the insurance industry. He also directs the development of natural language platform and applications such as privilege text classification, contextual summarization, and conversational sentiment abstraction. Prior to AIG, Dr. Chandra has driven innovation in BFSI, e-commerce, R&D, and mobile telecom industries in USA and India. He developed and implemented natural language predictive models that are deployed in top banks and telecom companies resulting in $100M impacts across value chain. For his contributions, Dr. Chandra has received the prestigious Barrier fellowship and several other awards and recognition. The Department of Homeland Security, United States Government has classified Dr. Chandra as an outstanding researcher. He was the conference session chair for GSPx conference at San Jose, California. He has been a reviewer for IEEE transactions, served in the editorial board of Human Language Technology conference, and speaker at several international conferences. He also has five assigned patents and several publications. Dr. Chandra is a passionate puzzler who invents puzzles and has represented India in the World Puzzle Championship at Stamford, Connecticut. He received his Ph.D. in Electrical and Computer Engineering from Mississippi State University. Pradeep Gulipalli Pradeep co-founded Tiger Analytics – an advanced analytics firm that builds data science solutions for businesses. He leads an interdisciplinary team of computer scientists, economists, statisticians, and business consultants. His team has done some pioneering work to bring data science into emerging business areas such as hyper-personalization, real-time decisions, dynamic learning, IoT and sensor networks etc. Pradeep’s team has built several industry leading data science solutions. The early warning systems for railways they built are now the North American industry standard. Their algorithms that predict emerging global technology trends guide M&A decisions. Their first of its kind interactive algorithm framework involving Dynamic pricing, Optimization, NLP, and Game theory helps place hundreds of millions of digital ads across devices every day. Pradeep has practiced data science for more than a decade architecting solutions for several Fortune 100 enterprises. In addition, his work around mathematical modeling of cities – socio-economic activity, transportation networks, land use, air quality etc. – has won him international recognition for application of data science to public policy. He has published in leading journals, co-guided research theses, and collaborates actively with universities. Pradeep holds a B.Tech from IIT Madras and an MS from The University of Texas at Austin. Sandhya Kuruganti Sandhya Kuruganti is a senior analytics leader in the Indian banking industry with more than 20 years of experience. She holds a Master’s degree in Economics from the Delhi School of Economics, India and a Doctorate in Economics from Rutgers University, USA. She has been one of the evangelists who has pioneered the usage of decision sciences for banks in India, as early as 16 years ago, when the analytics practice in India was still nascent. Sandhya’s thought leadership in developing and implementing advanced analytical strategies for customer life cycle management has been a key success factor for corporates that have leveraged her expertise in successfully embedding analytics in their business practices. As Senior Vice President at Citibank, she set up, nurtured, and led the Business Analytics function for Citibank India and the Regional Analytics Centre of Excellence for Asia Pacific. As a management consultant, she is currently on a mission to build analytics culture at Public Sector Banks, and educate the budding data science community in India. She also conducts applied analytics training programs and has co-authored a book titled “Business Analytics: Applications to Consumer Marketing – Sandhya Kuruganti and Hindol Basu”, published by McGraw Hill India in March 2015. Sarita Digumarti Sarita Digumarti is the COO & Co-founder of Jigsaw Academy, a premier analytics training institute. A specialist in Analytics, Consulting, and Outsourced Services and Management, Sarita is all about numbers. Holding an MBA in Finance followed by an MA in Quantitative Economics from Tufts University, Sarita has spent over 15 years working as a data scientist helping clients across diverse sectors including retail, healthcare and financial services, both in India and the US. Sarita now focuses on both education and new frontiers in the data science space. She is an expert educator who trains students at Jigsaw Academy, the participants of the Executive Program in Business Analytics (EPBA) at MISB Bocconi, Mumbai, and the numerous corporate employees whose companies have partnered with Jigsaw Academy. In recognition of her professional excellence, Sarita has been presented with the Global Achiever’s Award for Educational Leadership by the Economic Development Forum. Sudalai Rajkumar Sudalai Rajkumar (aka SRK) is the Lead Data Scientist at Freshdesk, responsible for developing scalable machine learning \/ data science systems for the organization. Prior to this, he was part of the R&D team in Global Analytics doing machine learning research and then moved to Tiger Analytics, where he has solved a lot of interesting data science problems for various customers across the globe in multiple domains including finance, e-commerce, online advertising, health care, transportation, retail. He has worked on varied problems ranging from doing simple analysis on structured data to natural language processing and voice analytics in his career. Apart from his day job, he used to take part in various data science competitions to enhance his knowledge and has won several of them. He is one of the top 25 data scientists in the world in Kaggle. He is one of the top solver in CrowdAnalytix platform as well. He has published few papers in International conferences and also has a patent filed under his name. SRK received his Bachelors degree from PSG college of Technology and got his executive certification in analytics from IIM Bangalore. Tuhin Chattopadhyay Tuhin Chattopadhyay is an eminent business analytics and data science thought leader with a progressive and proven track record of twelve years of experience in academia and industry. Tuhin has a profound knowledge of the marketing domain besides being an analytics expert. Academically, Tuhin is an ISB trained business analytics professional and IIM Ahmedabad educated management expert. A double masters in science (M.Sc.) and business administration (M.B.A.), he holds a Ph.D. in management (marketing analytics). Technically, he is a SAS certified predictive modeller and certified for IBM business analytics software. Prior to joining the data science industry, Tuhin was a renowned professor of business analytics and taught at a number of reputed B-Schools for a decade. As an academician, Tuhin pioneered the introduction of analytics based subjects into the curriculum of multiple B-Schools. Tuhin is a prolific researcher and has authored research-based books. He has more than thirty research publications in refereed journals and conference proceedings. To train the corporates as well as faculty members of B-Schools, he regularly conducts workshops on business analytics. He is the Editor-in-Chief of International Journal of Business Analytics and Intelligence and is the editorial board member of a number of leading journals. Viral Shah Viral Shah is a co-creator of Julia and CEO of Julia Computing. Julia is a modern and easy to use high-performance programming language for data science. It provides the performance of C++ while being as easy as Python, R, Matlab, and SAS. While it is an open source project with a diverse community of almost 500 contributors around the world, research on Julia is anchored at MIT. Today, the Julia community counts over 500 contributors and over 1,100 community contributed packages. Julia is being used by a number of universities for teaching and research, and by businesses in areas as diverse as engineering, finance, manufacturing, healthcare and retail. Apart from Julia, Viral is also co-creator of Circuitscape, an open source program for ecological conservation. Prior to that, he worked in the Aadhaar project, leading the design of Aadhaar-based eKYC, subsidies and payment systems. These experiences are captured in Rebooting India, a book he co-authored with Nandan Nilekani. Viral has a Ph.D. in Computer Science from the University of California at Santa Barbara.","excerpt":"Last year we took this initiative to identify Data Scientists in the country that are helping to push the industry forward. It is a grueling exercise of sorts; we invite nominations from various organizations, irrespective of size and nature of work. We also get in touch with data scientists that we know personally who might […]","categories":["AI Features"],"tags":[],"author_name":"Дарья","publish_date":"2016-11-28T09:35:00","publication_year":"2016","word_count":2100,"keywords":["data science","text classification","machine learning","AI","RAG","NLP","Python","Aim","analytics","R"],"extracted_tech_keywords":["AI","machine learning","NLP","data science","analytics","Aim","RAG","text classification","Python","R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/top-10-data-scientists-india-2016\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10045308,"title":"How Is OpenAI’s Triton Different From NVIDIA CUDA?","content":"Last month, OpenAI unveiled a new programming language called Triton 1.0  which enables researchers with no CUDA experience to write highly efficient GPU code. GPU programming is complicated.”Although a variety of systems have recently emerged to make this process easier, we have found them to be either too verbose, lack flexibility or generate code noticeably slower than our hand-tuned baselines,” according to OpenAI’s blog post. CUDA vs Triton In this article, we explore the key differences between OpenAI Triton and NVIDIA CUDA. But, before that, let’s take a look at the basic architecture of a GPU, as shown in the image below. The architecture of GPU (Source: OpenAI) Modern GPUs constitute three major components: DRAM, SRAM and ALUs — each of which must be considered when optimising CUDA code: Memory transfer from DRAM must be fused into large transactions to leverage the large bus width of modern memory interfaces.Data is manually uploaded to SRAM before being re-used and managed to minimise shared memory bank conflicts upon retrieval.Computations must be partitioned and scheduled carefully, both across and within streaming multiprocessors (SMs), to promote instruction\/thread-level parallelism and leverage special-purpose ALUs (Tensor Cores) “Reasoning about all these factors can be challenging, even for seasoned CUDA programmers with many years of experience,” according to the OpenAI blog. However, Triton has automated such optimisations so that developers can focus on the high-level logic of their parallel code. Triton does not automatically schedule work across SMs, leaving important algorithmic considerations to developers’ discretion. Algorithmic considerations may include tiling, inter-AM synchronisation, etc. (Source: OpenAI) CUDA Released in 2007, CUDA is available on all NVIDIA GPUs as its proprietary GPU computing platform. It includes third-party libraries and integrations, the directive-based OpenACC compiler, and the CUDA C\/C++ programming language. Today, five of the ten fastest supercomputers use NVIDIA GPUs, and nine out of ten are highly energy-efficient. CUDA is implemented in its device driver, but the compiler (nvcc) and libraries are packaged in the CUDA toolkit and software development toolkit (SDK). It contains many use cases and libraries, along with NVIDIA Nsight, an extension for Microsoft Visual Studio, and Eclipse (for Linux) for interactive GPU profiling and debugging. NVIDIA Nsight offers code highlighting, a unified GPU and CPU trace of the application, and automatic identification of GPU bottlenecks. NVIDIA Visual Profiler is a standalone cross-platform application for profiling CUDA programs, and CUDA versions for debugging and memory checking also exist. How is Triton different from CUDA? In 2019, Harvard researchers, in a paper titled Triton: An Intermediate Language and Compiler for Tiled Neural Network Computations, laid out the differences between CUDA and the Triton programming model. The execution of CUDA code on GPUs is supported by an SPMD programming model, where each kernel is associated with an identifiable thread block in a launch grid. The Triton programming model works the same way, but each kernel is single-threaded (though automatically parallelised) and associated with a set of global ranges that varies from instance to instance. The approach leads to simpler kernels in which CUDA-like concurrency primitives are nonexistent. The concurrency primitives include shared memory synchronisation, inter-thread communication, etc. (Source: Harward) Triton is similar to Numba, an open-source JIT compiler that translates a subset of Python and NumPy into fast machine code. In Numba, kernels are defined as decorated Python functions and launched concurrently with different instances. Wrapping up Deep Learning research ideas are typically implemented leveraging a combination of native framework operators. However, this approach requires many temporary tensors, which can put a damper on the performance of neural networks at scale. Specialized GPU kernels can take care of such issues. But it’s difficult to manage owing to the complexities of GPU programming. Triton automatically optimises specialised kernels and converts them into a code for execution on NVIDIA GPUs. The compiler currently uses block-level data-flow analysis, a technique used for scheduling iteration-block statically based on the control-and data-flow structure of the target programme, to solve the challenge of scheduling.","excerpt":"CUDA is available on all NVIDIA GPUs as its proprietary GPU computing platform.","categories":["Global Tech"],"tags":[],"author_name":"Amit Naik","publish_date":"2021-08-06T11:00:00","publication_year":"2021","word_count":659,"keywords":["CUDA","NumPy","OpenAI","AI","neural network","R","RAG","Python","deep learning","GPU computing"],"extracted_tech_keywords":["AI","deep learning","neural network","OpenAI","NumPy","RAG","GPU computing","CUDA","Python","R"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/how-is-openais-triton-different-from-nvidia-cuda\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":60624,"title":"Joining The Fight, PayPal To Take A No Coronavirus Layoffs Pledge","content":"PayPal has recently announced their pledge of not laying off any employees due to the coronavirus pandemic. Joining the fight against coronavirus, CEO Dan Schulman said that the company, in an effort not to be a part of the recession chaos, announced to not layoff any employees amid the pandemic. Schulman said to the media that “We don’t intend to do any layoffs as a result of COVID-19, and that’s the right thing to step up, in order to make sure they know that we’ve got their back.” Also, he further added, “If they’re sick, we pay them. If an office closes, we pay them. We need to be sure that we have their health and finances at heart as we deal with this crisis right now.” Consequently, the CEO of Marvell Technology, Matt Murphy also told that the company has “no plans” to layoff employees amid the global pandemic. He said to the media, “I want the team to be focused on the mission at hand, which is taking care of customers and their own families.” Join the likes of Bank of America, Morgan Stanley and Starbucks, PayPal and Marvell are also now offering job-safety assurances to employees who are worried about the economic shock from the coronavirus. Additionally, PayPal is offering small businesses enough flexibility to deal with the issues they are facing right now. The company has also taken steps to assist its business customers who are struggling with the economy coming to a near halt. It is allowing its business customers to push back repayments on business loans or cash advances at no additional cost. Alongside, it also is doubling the window in which merchants can respond to a customer dispute and waiving fees for instant withdraws from business accounts. Schulman said, amid this crisis “we need to take care of our shareholders, employees, and customers by stepping up and doing the right thing. And, that’s how we build businesses that can endure an economic downturn and can be strong. And a strong economy is helpful for everybody.”","excerpt":"PayPal has recently announced their pledge of not laying off any employees due to the coronavirus pandemic. Joining the fight against coronavirus, CEO Dan Schulman said that the company, in an effort not to be a part of the recession chaos, announced to not layoff any employees amid the pandemic. Schulman said to the media […]","categories":["AI News"],"tags":["Coronavirus","covid-19","Layoffs","Layoffs by IT companies","PayPal","recession"],"author_name":"Sejuti Das","publish_date":"2020-04-01T15:30:00","publication_year":"2020","word_count":342,"keywords":["Go","Layoffs","covid-19","PayPal","recession","AI","AWS","cloud_platforms:AWS","programming_languages:R","programming_languages:Go","Coronavirus","Layoffs by IT companies","GAN","R"],"extracted_tech_keywords":["AI","AWS","R","Go","GAN","cloud_platforms:AWS","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/joining-the-fight-paypal-to-take-a-no-coronavirus-layoffs-pledge\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10055102,"title":"MoEngage Raises USD 30 Million, Valuation Doubles","content":"Bengaluru-based customer analytics and cross-channel engagement platform MoEngage has raised USD 30 million in a funding round led by Steadview Capital with participation from existing investors, including Multiples Alternate Asset Management, Eight Roads Ventures, F-Prime Capital, and Matrix Partners. “This new round has come in just six months after our Series C1 funding at almost double the valuation and reaffirms our rapid growth,” wrote Raviteja (Ravi) Dodda, co-founder and CEO of MoEngage, in a blog post, but the specifics were not disclosed. The company had raised $32.5 million in Series C1 funding in July, which was led by Multiples Alternate Asset Management. The company will continue investing in its AI engine Sherpa and predictive capabilities while expanding its global presence and strengthening customer service and support functions. Their offices were recently opened in London, Boston, Ho-Chi Minh City and Berlin, and they have 500 employees globally. The CEO also informed that the ARR [Annualized Recurring Revenue] of the company had more than doubled in 2021. The company also claims to have added 350 new customers, including enterprises such as CIMB Bank, XL Axiata, The Body Shop, Domino’s Pizza, IHH Healthcare, and digital-first brands like ShareChat, Policy Bazaar, Atom Finance, Payactiv, Flipkart, Shopsy, Byjus, and SWVL.","excerpt":"Led by Steadview Capital, MoEngage raised USD 30 million; plans to invest in AI and expansion","categories":["AI News"],"tags":["customer analytics","Startups"],"author_name":"Meeta Ramnani","publish_date":"2021-12-08T20:34:36","publication_year":"2021","word_count":205,"keywords":["API","funding","programming_languages:R","AI","R","RPA","Git","Aim","ViT","analytics","Startups","customer analytics"],"extracted_tech_keywords":["AI","analytics","Aim","R","Git","API","ViT","RPA","funding","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/moengage-raises-usd-30-million-valuation-doubles\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10038736,"title":"How A Master’s Degree Can Help You Land A Lucrative Career In Data Science","content":"Data scientists are one of the most sought professionals across the world. At SkillUp 2021, Dakshinamurthy V Kolluru, president and professor at the International School of Engineering (INSOFE), spoke about career opportunities in the US, Canada and Europe for data scientists. Mastering data science To illustrate the importance of a master’s degree to land a lucrative career, Dakshinamurthy presented two reports: One from Kaggle and another from PwC and Business Higher Education Forum. As per the PwC & BHEF report, more than one-third of data science roles in the US require a master’s degree, while Kaggle reported that more than 50% of their data scientists have a master’s degree. “Kaggle is a model-driven site where someone will take the data, put it there, tells you how to access it and asks you to do it. So, the skill that is tested is how well you run a code which you expect a non-academician to also have. “Real-world, on the other hand, is not just model-driven. You need to collect the data by talking and interacting with the people, clean it up and understand the business. While there are excellent free videos, informal education available all over the place – traditional university education has a very important role to play in your success as a data scientist,” Dakshinamurthy said. Earlier, data scientists were expected to be good at everything. “It can be considered true earlier, when data science was just ‘science’, but today data science has graduated into a technology. And when it becomes a technology, you have to specialise,” said Dakshinamurthy. Dakshinamurthy outlined the software aspect of data science with four different specialisations. “Understand your area of interest, hone your computer science skills and add data science skills to that,” said Kolluru. How INSOFE helps Dakshinamurthy discussed the educational programme at INSOFE. The programme is modelled as per the needs of the candidates and is divided into three main streams. Separate programmes are available for core engineers, programmers or software engineers and business managers, with separate curriculum for each. The programme comes with three benefits, including full-fledged degrees, a chance for studying abroad in the USA, Canada and Europe with partner institutes and even work permits. All these programmes can be converted into credits for a full-fledged masters programme. Take, for instance, PGP in data science can be converted into 15 credits of MS courses, explained Dakshinamurthy. He also explained the step-by-step components of their programme: Self-learning modulesIndependent ResearchCourse credits in IndiaCourse credits abroadInternship INSOFE’s partners include Case Western Reserve University, USA; University of Strathclyde Glasgow, UK; Rennes School of Business, France; Carleton University, Canada; Walsh College, USA; and SVKM’S NMIMS, Deemed to be University, India. “INSOFE is vibrant and transformative. Our machine learning engineers are incubating products. AI specialists are pushing the boundaries through applied research. Our thought leading data scientists are engaged with enterprises to help them adopt data science. We enjoy bringing these rich experiences to the classroom to transform students,” said Dakshinamurthy.","excerpt":"“Understand your area of interest, hone your computer science skills and add data science skills to that.”","categories":["AI Trends"],"tags":["Masters in Data Science"],"author_name":"kumar Gandharv","publish_date":"2021-04-23T17:00:00","publication_year":"2021","word_count":496,"keywords":["data science","Go","machine learning","programming_languages:R","AI","Masters in Data Science","programming_languages:Go","R"],"extracted_tech_keywords":["AI","machine learning","data science","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/how-a-masters-degree-can-help-you-land-a-lucrative-career-in-data-science\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10077605,"title":"Stability AI, the Company Behind Stable Diffusion, Raises $101 Mn at A Billion Dollar Valuation","content":"World’s first community-driven and open-source AI innovator Stability AI has raised USD $101 million in funding. The funding round was led by Lightspeed Venture Partners, Coatue, and O’Shaughnessy Ventures LLC. According to a press release, the London-based firm will use these funds to accelerate the development of open AI models for language, image, audio, 3D, video, and for consumer and global enterprise use-cases. Stability AI founder and CEO Emad Mostaque said, “AI promises to solve some of humanity’s biggest challenges. But we will only realize this potential if the technology is open and accessible to all. Stability AI puts the power back into the hands of the developer communities and opens the door for ground-breaking new applications. An independent entity in this space supporting these communities can create real value and change.” Stability AI is the brain behind Stable Diffusion – a free and open-source text-to-image generator – launched in August this year. Since its launch, Stable Diffusion has been downloaded and witnessed licensing by over 200,000 developers around the world. Moreover, Stability AI’s DreamStudio quickly grew to over a million registered users over 50 countries – where they have collectively created more than 170 million images.“At Coatue, we believe that open source AI technologies have the power to unlock human creativity and achieve a broader good,” said Sri Viswanath, general partner at Coatue. “Stability AI is a big idea that dreams beyond the immediate applications of AI. We are excited to be part of Stability AI’s journey, and look forward to seeing what the world creates with Stability AI’s technology,” he added.","excerpt":"AI promises to solve some of humanity’s biggest challenges. But we will only realize this potential if the technology is open and accessible to all: Emad Mostaque, founder and CEO at Stability AI","categories":["AI News"],"tags":["AI Tool"],"author_name":"Bhuvana Kamath","publish_date":"2022-10-19T12:32:25","publication_year":"2022","word_count":263,"keywords":["Go","funding","programming_languages:R","AI","programming_languages:Go","stable diffusion","ViT","AI Tool","R"],"extracted_tech_keywords":["AI","R","Go","stable diffusion","ViT","funding","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/stability-ai-the-company-behind-stable-diffusion-raises-101-mn-at-a-billion-dollar-valuation\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":45425,"title":"India To Implement Faceless Tax Assessment Using Artificial Intelligence By Oct","content":"Union Finance Minister Nirmala Sitharaman this week announced that India will be using artificial intelligence and machine learning in the tax assessment process system from October 2019. In doing, so, India will become one of the first countries to adopt emerging tech on such a large scale in public tax assessment. The Steering Committee on FinTech related issues constituted by the Ministry of Finance, Department of Economic Affairs, this week submitted its Final Report to Sitharaman in her office in New Delhi. Given the rapid pace at which technology is being adopted primarily by private sector financial services, the Committee recommended Department of Financial Services (DFS) to work with PSU banks to bring in more efficiency to their work and reduce fraud and security risks. It also added that significant opportunities could be explored to increase the levels of automation using AI, cognitive analytics and machine learning in their back-end processes. Sitharaman in June had said that the existing system of scrutiny assessments in the IT Department involves a high level of personal interaction between the taxpayer and the Department, which leads to certain undesirable practices on the part of tax officials. To eliminate such instances, a scheme of faceless assessment in electronic mode involving no human interface is being launched this year in a phased manner. To start with, such e-assessments shall be carried out in cases requiring verification of certain specified transactions or discrepancies, she added. Earlier this year the income tax department had used data analytics to identify citizens who had indulged in high-value transactions and had potential tax liability but had failed to file IT returns for the assessment year 2018-19.","excerpt":"Union Finance Minister Nirmala Sitharaman this week announced that India will be using artificial intelligence and machine learning in the tax assessment process system from October 2019. In doing, so, India will become one of the first countries to adopt emerging tech on such a large scale in public tax assessment. The Steering Committee on […]","categories":["AI News"],"tags":["human intelligence at machine scale","nirmala sitharaman"],"author_name":"Prajakta Hebbar","publish_date":"2019-09-04T12:09:00","publication_year":"2019","word_count":275,"keywords":["API","machine learning","artificial intelligence","human intelligence at machine scale","AI","programming_languages:R","nirmala sitharaman","automation","analytics","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","analytics","R","API","automation","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/india-to-implement-faceless-tax-assessment-using-artificial-intelligence-by-oct\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10172528,"title":"Why AI Is Struggling to Charm Its Toughest Customer — Indian MSMEs","content":"India’s micro, small, and medium enterprises are ready to adopt AI, but the road ahead is a mix of promise and challenges. The government’s recent report on ‘Enhancing MSMEs Competitiveness in India’ sets the stage, pointing out that while artificial intelligence holds immense promise for these businesses, adoption remains a challenge. Many enterprises, the report notes, are still unfamiliar with India’s evolving data protection laws and face uncertainty about compliance when integrating AI into their operations. “Government-led awareness campaigns and simplified guidelines on data protection can encourage responsible AI use and reduce the risk of non-compliance,” the report recommends, underscoring the urgent need for clear, accessible information. A shortage of qualified AI professionals compounds the challenge. Most MSMEs lack the in-house expertise to evaluate, select, and implement AI solutions, making the role of collaborative platforms —those that connect MSMEs with academic institutions, AI consultants, and tech companies —especially vital. The government’s report also highlights the financial challenges. “Affordability is another persistent challenge, especially with AI tools, as computing infrastructure and training costs remain prohibitive for smaller enterprises,” the report said.  It advocates for targeted financial assistance, including grants, subsidies, tax incentives, and low-interest loans, as well as cloud-based, pay-as-you-go AI solutions to make advanced technologies more accessible.India boasts a large manufacturing sector, with more than 90% of our industrial enterprises classified as MSMEs. As stated by G Prakash, chairperson of the manufacturing expert committee, BCIC, these businesses are responsible for nearly 30% of India’s GDP, provide employment for 110 million individuals, and account for 48% of India’s exports. “Many MSMEs have yet to embrace digitalisation or AI, often perceiving these technologies as costly and complex,” Prakash added, as reported by Deccan Herald. “AI-driven solutions now available through affordable cloud-based and plug-and-play models can deliver productivity gains of 15% to 30%, enabling predictive maintenance, quality monitoring, and energy optimisation, even in small-batch environments.” All these indicate that MSMEs want AI to be both affordable and equitably accessible, and policy must keep pace. The Modest Beginnings Rahul Sharma, vice president of sales at Salesforce India, described how AI is already beginning to change the narrative for India’s 5.93 crore MSMEs, which contribute nearly 30% of the country’s GDP. “One of the most immediate benefits of AI is the ability to take tedious, repetitive work off your team’s plate,” Sharma said, highlighting how AI-powered tools are automating routine tasks such as meeting follow-ups, status updates, and reminders. This, Sharma explained, frees up MSME teams to focus on strategic priorities like customer service and market expansion. He highlighted the power of AI in quickly surfacing knowledge and insights, noting that for many small businesses, valuable information is often spread across emails, chats, and various apps. “AI can help by consolidating this information, surfacing trends, and making insights easily accessible for smarter decision-making,” Sharma said, emphasising that these benefits are now within reach even for MSMEs without dedicated data teams. Sharma emphasised that AI is no longer the exclusive domain of large enterprises or data scientists. “Today’s AI tools are designed for everyday users, making it easy to streamline workflows, generate content, and automate tasks with just a few clicks,” he said. Some features are already democratising AI for small businesses by allowing any team member to interact with AI agents as virtual teammates.  “This is a game-changer for MSMEs, enabling small teams to scale their capabilities with minimal resources,” he said. The customer experience is also being transformed, as AI-driven personalisation enables MSMEs to analyse customer conversations, tailor outreach, and foster stronger relationships. However, Pravir Dahiya, CTO of Tata Teleservices, in a report,  pointed out that many MSMEs operate with disconnected digital systems—CRM, billing, communication—leading to inefficiencies and missed opportunities. Cloud-first, AI-enabled platforms that centralise operations and automate workflows can improve agility and decision-making, especially for MSMEs with limited IT resources, Dahiya added. Affordability Yet, the question of how MSMEs can affordably and sustainably access these technologies remains. Red Hat senior director (Technology Solutions Architecture) Mangesh Surve offered a view from the infrastructure side, noting that MSMEs overwhelmingly prefer operational expenditure (OPEX) models over capital expenditure (CAPEX) when it comes to AI adoption. He explained that MSMEs typically prefer to invest in operating expenses (OPEX) rather than making a large initial investment. He noted that these businesses often choose open source software hosted on cloud platforms—sometimes with major global hyperscalers, but more frequently with local data center providers. This approach, he argued, is not only about cost savings and flexibility, but also about avoiding vendor lock-in. “It’s a misnomer that open source doesn’t have an end-to-end solution. Open source, I think, for almost everything that you talk about—including AI—provides a complete end-to-end solution,” he said, pushing back against the notion that only expensive proprietary solutions can deliver comprehensive results. The roots of AI, Surve notes, are in the open source world, and today’s open source ecosystem can fulfil the entire solution architecture requirements of MSMEs. Among the sectors leading this charge are financial services and fintech MSMEs, whose need for innovation and agility is especially acute, he observed. Customised solutions Umesh Sahay, founder & managing director, NES Data Pvt Ltd, observed that every MSME is different, and ready-made AI tools don’t always fit their needs. Customised AI solutions, according to him, can help them solve their specific problems—whether it’s managing inventory, understanding customers better, or improving productivity. “If we can build the support system, MSMEs will be able to use AI in a way that really benefits their business,” he suggested. In an ET report, Anoop G Prabhu, CTO of Vehant Technologies, stressed that MSMEs across sectors have unique AI needs. Manufacturing MSMEs may prioritise worker safety compliance and process monitoring, while logistics firms focus on PPE detection and cargo scanning. Retail MSMEs, on the other hand, require customer flow analytics, he said. Customised, industry-specific AI solutions are essential to deliver meaningful impact across diverse use cases, he added. Government policy is beginning to address the need for awareness, skills, and affordability. Technology providers are making AI both accessible and practical for non-technical users. Open-source infrastructure is offering MSMEs the flexibility and completeness they need without locking them into costly, proprietary systems. Capturing the mood of a sector poised for transformation, Sharma concluded that “AI is no longer a technology of the future—it’s here today, and it’s increasingly accessible to India’s MSMEs.”","excerpt":"India’s MSMEs face financial, skill, and regulatory challenges in adopting AI. Affordable, customised, and cloud-based solutions offer a hopeful path forward.","categories":["AI Features"],"tags":["AI","India","MSMEs","NES Data","red hat","Salesforce"],"author_name":"C P Balasubramanyam","publish_date":"2025-06-27T17:53:21","publication_year":"2025","word_count":1058,"keywords":["Go","API","red hat","artificial intelligence","AWS","AI","ML","Git","MSMEs","RAG","Salesforce","analytics","NES Data","R","India"],"extracted_tech_keywords":["AI","artificial intelligence","ML","analytics","RAG","AWS","R","Go","Git","API"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/why-ai-is-struggling-to-charm-its-toughest-customer-indian-msmes\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":62522,"title":"Data Science Career Transition Strategy During The Covid-19 Lockdown","content":"Driven by the hype over the field of data science — and a potentially fat paycheck — working professionals across industries have been looking to reorient their career trajectories. The lockdown, enforced to slow the spread of coronavirus, has offered them the opportunity to upskill themselves to transition into a career in data science. This transition, however, will not be smooth and would require additional efforts to make that final leap. Instead of randomly learning whatever you come across, aspirants need to devise an effective strategy that can help them reap more benefits and avoid burnouts. Since data science is relatively new, universities did not offer lessons that enabled students to obtain machine learning skills. As a result, most data scientists have been successful through self-study. If you aspire to become like them, this is the right time to take your dream seriously and become a data scientist. Learn The Basics Of Data Science Due to the crisis, various ed-tech organisations have made their data science courses free to help aspirants upskill and eliminate the talent gap problem in the space. However, you should not focus on learning everything that the courses offer. First, build your foundation by learning statistics and mathematics. Then move on to data analysis and basic machine learning techniques. “I know that deep learning and AI are in trend. But, it would be a strict no from my side if the interviewee is not strong on the basics,” said Rahul Agrawal, Data Scientist of Walmart Labs. Organisations seek candidates who have a strong foundation, not someone who can write fancy machine learning algorithms. But, you will still have to find a specialisation within data science to master, which can help you deliver value for the organisation. Choose A Specialisation Data science is vast, and if you start paying attention to every development in the space, you may not be able to dive deep. While there are numerous ways in which you can find a specialisation, the most natural approach is to understand how data science can be implemented in your current professional work. For instance, if your job involves documentation, you should start gaining an in-depth understanding of natural language processing. Go back to the free courses and take advanced courses on the specialisation of your choice. “Being specific is necessary to succeed. Aspirants should pick up a particular class of problems within a domain and apply data science techniques to solve them,” said Shashank Shekar, Head of Data Science at RadiusAI. Implement The Learning In Your Job On obtaining the required knowledge, try to deploy the data science techniques in your workflows. Randomly applying ML techniques for the sake of showcasing during interviews will not add any value. Identify a problem in your job that can be carried out effectively with data science. This will demonstrate your capability and invoke interest among recruiters. Attend AI Conferences Attending AI conferences such as plugin will not only help you understand what is happening in the ever-changing data science landscape, but also give you an idea around where the industry is heading. This will enable you to plan accordingly and be aware of trends in the market. Besides, you also get an opportunity to expand your network, which can help you in obtaining a reference when looking for jobs. Undoubtedly, an internal switch should be the first choice, but having access to jobs outside your firms through referrals can bring better opportunities too. Get A Mentor Having mentors to help, especially professionals from the same field, can assist you in shaping your career. For this, you can obtain guidance through the AIM mentoring circle. Being an enabler for the data science market, AIM has an exceptional network of data science leaders and influencers who can suggest various approaches to becoming successful data scientists. Watch Next Video –","excerpt":"Driven by the hype over the field of data science — and a potentially fat paycheck — working professionals across industries have been looking to reorient their career trajectories. The lockdown, enforced to slow the spread of coronavirus, has offered them the opportunity to upskill themselves to transition into a career in data science. This […]","categories":["AI Features"],"tags":["Data Scientist Jobs","how to become a data scientist","MOOCs","Walmart Labs","what is data science"],"author_name":"Rohit Yadav","publish_date":"2020-04-24T17:00:00","publication_year":"2020","word_count":639,"keywords":["data science","Go","how to become a data scientist","Data Scientist Jobs","machine learning","API","AI","ML","MOOCs","Walmart Labs","Aim","deep learning","GAN","what is data science","R"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","data science","Aim","R","Go","API","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/data-science-career-transition-strategy-during-the-covid-19-lockdown\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10126728,"title":"Big-Tech Companies Push for Gender Sensitisation","content":"Most AI products today are the result of data collected from diverse users, which can inherently contain biases. This calls for the need to ramp up efforts to mitigate these biases rather than attempting to overfit the data to preconceived notions. “We are trying to build products used by billions of users, [and] working on important technologies that will have a deep impact on society,” said Google head Sundar Pichai in an interview, addressing the dire need for inclusivity and diversity in the workplace. “A diverse workforce will help the company develop better products, and more importantly, tackle problems in society better,” he added. Sandeep Sharma, assistant manager, GlobalLogic India, and an openly queer individual, agrees. “Organisations can foster inclusivity by establishing ally network groups and providing platforms to drive LGBTQIA+ inclusion in the workplace. Additionally, sensitisation programs for employees can be transformative,” Sharma told AIM in a recent conversation. For someone not comfortable discussing his sexuality and attending Pride events just as an ally, Sharma has come a long way. Sandeep Sharma, Assistant Manager, GlobalLogic India He is now actively involved with GlobalLogic’s DEI accelerator team, striving to create an inclusive environment for his LGBTQIA+ colleagues. He has been facilitating SOGIE (sexual orientation, gender identity, and gender expression) sensitisation sessions to help others understand the lives, experiences, and challenges faced by the LGBTQIA+ community better. Previously, we have spoken about how allyship in the workplace helps familiarise the concept of understanding the non-binary nature of gender. This is where gender sensitisation programs, employee resource groups and Pride networking groups come into the picture for corporations. Big Tech’s Efforts Big techs like Google, Apple, Microsoft, Meta, and NVIDIA, among others, recognised the importance of such programs long ago. “Inclusive teams that propagate and advance inclusive principles will have the deepest impact in building products designed for everyone,” said Satya Nadella, the chief executive officer of Microsoft, in a previous interview. For example, Apple’s Tim Cook, who is openly gay and has often spoken about the same, has had an LGBTQ-focused employee network called Pride@Apple since 1986. Similarly, Google established its first LGBTQ employee group, Gayglers, over 20 years ago, which has had a massive impact on steering the company in a more LGBTQ-inclusive direction. So does Microsoft. GlobalLogic is also one such company fostering D&I through several initiatives to support LGBTQIA+ employees, such as inclusive hiring practices, support groups, and awareness campaigns. According to Sharma, these initiatives have significantly impacted the work environment, creating a culture of acceptance and respect. Originally from Delhi, Sharma, a computer science engineer, has worked as a tech product owner in various Indian cities over the past seven years. AI to Solve Inclusivity? “Contrary to the common belief of tokenism and corporate rainbow-washing, I believe we’re moving in the right direction. While many organisations and brands might currently leverage Pride as a strategy, this undeniably starts the conversation,” he added. “This is especially crucial in countries where LGBTQIA+ community is still battling the fight against social stigma and basic rights. It is a long and demanding journey and while it may not be ideal today, progress has to begin somewhere.” However, the ever-evolving AI and analytics landscape holds the potential to drive LGBTQIA+ inclusivity in the coming years. For example, AI-powered language models can be trained to use more inclusive and gender-neutral language, helping normalise diverse identities in digital communication. “With AI increasingly integrated into our daily lives. It mustn’t replicate societal bias. Data models trained predominantly on heteronormative information may exhibit harmful prejudices. To combat this, we must dismantle institutional barriers. “This includes the absence of gender-neutral or inclusive language, a lack of inclusive imagery, and the features that reflect LGBTQIA+ identities and experiences,” said Sharma. Interestingly, according to a McKinsey & Company report, companies in the top quartile for gender diversity are 21% more likely to outperform on profitability. Additionally, organisations with more gender-diverse executive teams have a 27% likelihood of outperforming their peers on longer-term value creation. Challenges Remain The inclusion of transgenders in the workforce still remains a challenge. Sharma advocates for infrastructural changes like gender-neutral washrooms and policies accommodating preferred names and pronouns for creating environments that empower non-binary and gender minorities in the workforce. On the other hand, AI-driven health applications could be developed to address the specific medical needs of transgender individuals, improving access to personalised care. Additionally, AI can assist in content moderation on social media platforms, more effectively identifying and removing hate speech or discriminatory content targeting the community. “LGBTQIA+ isn’t just an abstract concept; it encompasses real people still fighting for social acceptance and basic rights” he added.","excerpt":"“Contrary to the common belief of tokenism and corporate rainbow-washing, I believe we’re moving in the right direction.”","categories":["Global Tech"],"tags":["diversity","inclusivity","Microsoft"],"author_name":"Shritama Saha","publish_date":"2024-07-12T18:15:05","publication_year":"2024","word_count":773,"keywords":["Replicate","Go","AI","Git","RAG","diversity","Aim","ViT","analytics","inclusivity","GAN","R","Microsoft"],"extracted_tech_keywords":["AI","analytics","Aim","RAG","R","Go","Git","GAN","ViT","Replicate"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/big-tech-companies-push-for-gender-sensitisation\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":66810,"title":"A New Paradigm Of Object Detection: Using Transformers","content":"Ever since its introduction in 2017, in the seminal paper “Attention Is All You Need,” transformers have revised and renewed the interest in the way we deal with language models. Transformers have led to groundbreaking works such as BERT, which is now widely used by Google for its search engine autocomplete and smart compose on Gmail. Transformers have been widely applied in tasks as diverse as speech recognition, symbolic mathematics, and even reinforcement learning. However, computer vision, so far, has been immune to the advent of transformers so far. DETR can be implemented in less than 50 lines with PyTorch Now for the time, object detection is being looked at through the lens of transformers. A new model by the name Detection Transformers (DETR) was recently released by the AI wing of Facebook. DE⫶TR: End-to-End Object Detection with Transformers https:\/\/t.co\/4W27PVx6Js (+paper https:\/\/t.co\/xmxnSOBjBa) awesome to see a solid swing at (non-autoregressive) end-to-end detection. Anchor boxes + nms is a mess. (I was hoping detection would go end-to-end back in ~2013) pic.twitter.com\/cN4hK2ZFY0— Andrej Karpathy (@karpathy) May 27, 2020 Overview Of DETR DETR architecture The overall DETR architecture contains three main components: a CNN backbone to extract a compact feature representation, an encoder-decoder transformer, and a simple feed-forward network (FFN) that makes the final detection prediction. Unlike many modern detectors, stated the researchers, DETR can be implemented in any deep learning framework that provides a common CNN backbone and a transformer architecture implementation with just a few hundred lines. Inference code for DETR can be implemented in less than 50 lines with PyTorch. The working of DETR can be distilled as follows: Consider the object detection task as an image-to-set problem. Add a convolutional neural network (CNN) that extracts the local information from the image with a transformer encoder-decoder architecture, which then reasons about the image as a whole and then generates the predictions. DETR uses a convolutional neural network (CNN) as a fundamental component of learning a 2D representation of an input image. The model flattens it and supplements it with a positional encoding before passing it into a transformer encoder. A transformer decoder then takes as input a small fixed number of learned positional embeddings or object queries and additionally attends to the encoder output. Then output embedding of the Decoder is passed to a shared feed forward network (FFN) that predicts either a detection (class and bounding box) or a “no object” class. Given an image, the DETR model successfully identifies all the objects present in an image, each represented by its class, along with a tight bounding box surrounding each one. To evaluate the model, the researchers performed experiments on COCO 2017 detection and panoptic segmentation datasets containing 118k training images and 5k validation images. The above table compares the performance of DETR with state-of-the-art models. Key Takeaways To understand how Transformers make an end-to-end object detection simpler, the researchers pitted it against the state-of-the-art Faster R-CNN, a traditional two-stage detection system. In case of Faster R-CNN, as shown above, object bounding boxes are predicted by filtering over a large number of coarse candidate regions, which are generally a function of the CNN features. Then each selected region is used in a refinement step. The refinement step consists of identifying the features of a certain area in the image and then classifying it independently (if it is a bird, then its beak is a feature). After refinement, deduplication is carried out where a non-maximum suppression step is applied to remove duplicate boxes. DETR, on the other hand, gets rid of the whole refinement and deduplication process by placing a transformer architecture. It simplifies the detection pipeline as the transformer architecture performs the operations that are traditionally specific to object detection. The key takeaways from this work are as follows: DETR, a new design for object detection systems based on transformers and bipartite matching loss for direct set prediction DETR is the first object detection framework to successfully integrate transformers as a central building block in the detection pipeline DETR matches the performance of state-of-the-art highly optimised Faster R-CNN Know more about DETR here.","excerpt":"Ever since its introduction in 2017, in the seminal paper “Attention Is All You Need,” transformers have revised and renewed the interest in the way we deal with language models. Transformers have led to groundbreaking works such as BERT, which is now widely used by Google for its search engine autocomplete and smart compose on […]","categories":["Deep Tech"],"tags":["Generative Pre-Trained Transformer","Object Detection"],"author_name":"Ram Sagar","publish_date":"2020-06-06T08:00:00","publication_year":"2020","word_count":684,"keywords":["Go","TPU","AI","neural network","PyTorch","Transformers","computer vision","deep learning","Object Detection","Generative Pre-Trained Transformer","object detection","R"],"extracted_tech_keywords":["AI","deep learning","neural network","computer vision","PyTorch","Transformers","object detection","TPU","R","Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/object-detection-with-transformers\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":20251,"title":"Mobile Analytics Startup CleverTap Receives Fresh Funding From Recruit Holdings Co. Ltd","content":"As a part of its fundraising drive, analytics startup CleverTap has secured an undisclosed amount in funding from Japan-based Recruit Holdings Co.,Ltd. Recruit Holdings Co.,Ltd, provider of diverse consumer and business services, made the investment through its investment subsidiary, RSP India Fund LLC, an equity wing created in 2015 that funds Indian startups. Recruit Holdings Co. confirmed in a statement that the investment was made in association with CleverTap’s current investors, Sequoia Capital India and Accel Partners. Though the funding amount is unknown, it is speculated that the new investment secured by the startup is a part of its efforts to raise $8 million, of which $6 million was raised a few months ago. CleverTap is expected to use the received funds for product development and customer base expansion. It was reported in May 2017 that Anand Jain, co-founder of CleverTap, had revealed the company’s plans of securing investments to the tune of $20 million in its Series B round from both existing and new investors . Beginning with a seed capital of $1.6 million from from Accel Partners in July 2014, CleverTap has been successful in securing investments regularly. It raised another $8 million investment from Accel Partners and Sequoia Capital in August 2015 as a part of its Series A round of funding. As for Recruit Holdings Co., last year was eventful in terms of acquisitions and investments in startups, as it acquired Germany-based guest feedback platform, TrustYou in 2017 and invested in the US-India based logistics management startup, Locus. Formerly known as WizRocket, CleverTap provides mobile app analytics and user engagement products to over 4000 customers using behavioural analytics. It helps firms understand customer behaviour and improves their customer retention, and caters to customers across different stages of development such as startups, SMBs and Fortune 500 companies. Its clientele includes global brands such as Samsung, Sony and McDonald’s.","excerpt":"As a part of its fundraising drive, analytics startup CleverTap has secured an undisclosed amount in funding from Japan-based Recruit Holdings Co.,Ltd.   Recruit Holdings Co.,Ltd, provider of diverse consumer and business services, made the investment through its investment subsidiary, RSP India Fund LLC, an equity wing created in 2015 that funds Indian startups. Recruit […]","categories":["AI News"],"tags":[],"author_name":"Jeevan Biswas","publish_date":"2018-01-03T10:25:36","publication_year":"2018","word_count":311,"keywords":["Go","API","funding","programming_languages:R","AI","programming_languages:Go","analytics","Rust","R","startup"],"extracted_tech_keywords":["AI","analytics","R","Go","Rust","API","startup","funding","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/clevertap-funding-recruit-holdings-co-ltd\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10089299,"title":"Is Foxconn Conning India?","content":"“A disaster” – this is how Lawrence Tabak, the author of Foxconned: Imaginary Jobs, Bulldozed Homes, and the Sacking of Local Government described Foxconn’s promise to bring back TV manufacturing to the United States. A tale largely of disappointment and misplaced faith, the book’s title raises the possibility of people being intentionally duped by Foxconn and American politicians. In India, however, the Foxy business has just begun. Regardless, the Taiwanese contract manufacturer has been finding it difficult to stay out of the headlines. Reports of making states wage a bidding war, mislead the government, or lobby for softer labour regulations have been a constant with Foxconn. Smokes and Mirrors Tabak, who witnessed a spree of false promises play out in several municipalities in the United States, said, “Foxconn has the reputation for being one of the most opaque companies in an opaque world.” The company, he seemed to suggest, had no problem in making big announcements for the cause of a government’s electoral ambitions without having any concrete plan of action. The question, even when it comes to the Indian context, is: “How much of this is based on reality, and how much is just smoke and mirrors?” Most recently, the company found itself embroiled in controversy when both Telangana and Karnataka governments simultaneously claimed Foxconn to have signed up for big investments in their respective states to manufacture electronics. These investments are expected to create over 100,000 jobs. A similar scenario played out in Foxconn’s foundry plans when their proposed joint venture with Vedanta, which was in an advanced stage of talks for a site in Maharashtra, suddenly shifted to Gujarat just ahead of the state elections. Waging a bidding war between states is not new to Foxconn. Tabak mentions that the same happened at his home-state Wisconsin as well, which was competing against seven different states to bring on a dawn of corporate and industrial development with Foxconn. When it comes to Foxconn, news around it has always reeked of some mischief – almost as if somebody wants to plant some kind of a thought (or maybe influence the government). “In at least one such case, the government or ministers never said anything, but Vedanta\/Foxconn went ahead and started claiming something completely different,” a source told AIM. “It is very normal for them to make stagey announcements that involve politicians, business executives, pomp, and circumstance purely based on speculation,” said Tabak. Moreover, there were also concerns regarding Foxconn’s ability to provide the necessary technical expertise, given that the company lacks the necessary experience in operating a fabrication facility. The fact that the joint venture received a go-ahead above the other proposals despite this shows that the government has been very bullish on Foxconn. Foxconn chairman Young Liu with Karnataka Chief Minister Basavaraj Bommai in Bengaluru. Behind closed doors Amid the delayed approval for the first fab, Foxconn also announced the intention to set up a new chip-making facility. The information, tipped by an unnamed government official, doesn’t dwell on the whats and hows of this proposed plan. However, at this point, one might wonder whether this is another mischief to get state governments – who can now promise hundreds and thousands of jobs – to use Foxconn. As Tabak explained, in Wisconsin, Foxconn came at a time when the governor was facing low approval ratings from the public, and the promise of high-paying jobs to its constituency would change that. Obviously, a lot of experts were puzzled by Foxconn’s moves. “How could they make this work? There weren’t proper supply chains and the cost of manufacturing would be enormously high in the United States – partly due to labour and partly due to other kinds of things,” he added. In this lobbying exercise, the only ones benefiting were the political powers and Foxconn, who, according to Tabak, could protect themselves from many of the trade tariffs that the Trump administration imposed. A similar story repeats here as well. A few days ago, just around the time when Foxconn agreed to set up an iPhone manufacturing facility in Karnataka, Financial Times reported that the government aims to liberalise the labour laws in the state, which will make it comparable to the workforce productivity in China, Taiwan, and Vietnam. “It doesn’t surprise me that they [Foxconn] are trying to manipulate labour laws or that they’re using the sort of relationships with powerful administrators or politicians, because that’s one of the keys to their success in China,” remarked Tabak. Foxconn’s successful lobbying for this new legislation is a testament to its modus operandi. Although the senior folks are treated well at Foxconn, Tabak said that when it comes to the bottom level of their hierarchy, they are almost like “replaceable parts”. With skills that are teachable in ten minutes, it is always easier to find and replace labour and put them in the assembly line, without having an effect on the overall production. “Foxconn has built their fortune on labour, paying labour the lowest possible price,” said Tabak. We saw that in its operations in the Czech Republic, where the company exploited legal loopholes to provide extremely low wages to migrant workers and “they supplement that with government grants to keep the cost even lower than it appears”. In the United States, Foxconn were “prolific users” of undocumented workers who were exploitable because they couldn’t go and complain about the working conditions. “They seem to be much engaged in gaming the system. I’m sure they’re not the only company that tries to exploit these things, but their business model seemed to be particularly focused on that sort of thing,” Tabak noted. Source: Getty Images Analytics India Magazine reached out to Foxconn for their comments, but didn’t hear back from them till the time of publishing this story. Foxconn, a microchip player The above discussion is not to imply that Foxconn cannot be a big player in the microchip business. The company has gone big on spending as well. They are interested in getting into the vertically-integrated kind of business mostly because the potential for profits are great. In the microchip business, Foxconn has been purchasing the required expertise – like in their acquisitions of Japan-based Sharp, and the Malaysian semiconductor company, SilTerra – to build those end-to-end capabilities, and rake in more profits than, say, manufacturing iPhones for Apple. The issue, however, is that like with all of their other pompous claims, the plans for these fabs haven’t seen the light of day – or at least has largely been absent from the media. The international landscape for chip manufacturing – which will determine the countries where the margins are high – is going to be a factor in whether Foxconn can fulfil the promises it has made. “We are all being Foxconned, every day,” Tabak exclaimed.","excerpt":"Most recently, Foxconn found itself embroiled in controversy when both Telangana and Karnataka governments simultaneously claimed Foxconn to have signed up for big investments in their respective states","categories":["IT Services"],"tags":["foxconn india","India semiconductor mission"],"author_name":"Ayush Jain","publish_date":"2023-03-14T12:30:00","publication_year":"2023","word_count":1136,"keywords":["Go","AWS","AI","foxconn india","cloud_platforms:AWS","programming_languages:R","India semiconductor mission","Aim","ViT","analytics","GAN","R"],"extracted_tech_keywords":["AI","analytics","Aim","AWS","R","Go","GAN","ViT","cloud_platforms:AWS","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/it-services\/is-foxconn-conning-india\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10007144,"title":"The Continuous Bag Of Words (CBOW) Model in NLP &#8211; Hands-On Implementation With Codes","content":"Word2vec is considered one of the biggest breakthroughs in the development of natural language processing. The reason behind this is because it is easy to understand and use. Word2vec is basically a word embedding technique that is used to convert the words in the dataset to vectors so that the machine understands. Each unique word in your data is assigned to a vector and these vectors vary in dimensions depending on the length of the word. The word2vec model has two different architectures to create the word embeddings. They are: Continuous bag of words(CBOW)Skip-gram model In this article, we will learn about what CBOW is, the model architecture and the implementation of a CBOW model on a custom dataset. What is the CBOW Model? The CBOW model tries to understand the context of the words and takes this as input. It then tries to predict words that are contextually accurate. Let us consider an example for understanding this. Consider the sentence: ‘It is a pleasant day’ and the word ‘pleasant’ goes as input to the neural network. We are trying to predict the word ‘day’ here. We will use the one-hot encoding for the input words and measure the error rates with the one-hot encoded target word. Doing this will help us predict the output based on the word with least error. The Model Architecture The CBOW model architecture is as shown above. The model tries to predict the target word by trying to understand the context of the surrounding words. Consider the same sentence as above, ‘It is a pleasant day’.The model converts this sentence into word pairs in the form (contextword, targetword). The user will have to set the window size. If the window for the context word is 2 then the word pairs would look like this: ([it, a], is), ([is, pleasant], a),([a, day], pleasant). With these word pairs, the model tries to predict the target word considered the context words. If we have 4 context words used for predicting one target word the input layer will be in the form of four 1XW input vectors. These input vectors will be passed to the hidden layer where it is multiplied by a WXN matrix. Finally, the 1XN output from the hidden layer enters the sum layer where an element-wise summation is performed on the vectors before a final activation is performed and the output is obtained. Implementation of the CBOW Model For the implementation of this model, we will use a sample text data about coronavirus. You can use any text data of your choice. But to use the data sample I have used click here to download the data. Now that you have the data ready, let us import the libraries and read our dataset. import numpy as np import keras.backend as K from keras.models import Sequential from keras.layers import Dense, Embedding, Lambda from keras.utils import np_utils from keras.preprocessing import sequence from keras.preprocessing.text import Tokenizer import gensim data=open('\/content\/gdrive\/My Drive\/covid.txt','r') corona_data = [text for text in data if text.count(' ') >= 2] vectorize = Tokenizer() vectorize.fit_on_texts(corona_data) corona_data = vectorize.texts_to_sequences(corona_data) total_vocab = sum(len(s) for s in corona_data) word_count = len(vectorize.word_index) + 1 window_size = 2 In the above code, I have also used the built-in method to tokenize every word in the dataset and fit our data to the tokenizer. Once that is done, we need to calculate the total number of words and the total number of sentences as well for further use. As mentioned in the model architecture, we need to assign the window size and I have assigned it to 2. The next step is to write a function that generates pairs of the context words and the target words. The function below does exactly that. Here we have generated a function that takes in window sizes separately for target and the context and creates the pairs of contextual words and target words. def cbow_model(data, window_size, total_vocab): total_length = window_size*2 for text in data: text_len = len(text) for idx, word in enumerate(text): context_word = [] target   = [] begin = idx - window_size end = idx + window_size + 1 context_word.append([text[i] for i in range(begin, end) if 0 <= i < text_len and i != idx]) target.append(word) contextual = sequence.pad_sequences(context_word, total_length=total_length) final_target = np_utils.to_categorical(target, total_vocab) yield(contextual, final_target) Finally, it is time to build the neural network model that will train the CBOW on our sample data. model = Sequential() model.add(Embedding(input_dim=total_vocab, output_dim=100, input_length=window_size*2)) model.add(Lambda(lambda x: K.mean(x, axis=1), output_shape=(100,))) model.add(Dense(total_vocab, activation='softmax')) model.compile(loss='categorical_crossentropy', optimizer='adam') for i in range(10): cost = 0 for x, y in cbow_model(data, window_size, total_vocab): cost += model.train_on_batch(contextual, final_target) print(i, cost) Once we have completed the training its time to see how the model has performed and test it on some words. But to do this, we need to create a file that contains all the vectors. Later we can access these vectors using the gensim library. dimensions=100 vect_file = open('\/content\/gdrive\/My Drive\/vectors.txt' ,'w') vect_file.write('{} {}\\n'.format(total_vocab,dimensions)) Next, we will access the weights of the trained model and write it to the above created file. weights = model.get_weights()[0] for text, i in vectorize.word_index.items(): final_vec = ' '.join(map(str, list(weights[i, :]))) vect_file.write('{} {}\\n'.format(text, final_vec)) vect_file.close() Now we will use the vectors that were created and use them in the gensim model. The word I have chosen in ‘virus’. cbow_output = gensim.models.KeyedVectors.load_word2vec_format('\/content\/gdrive\/My Drive\/vectors.txt', binary=False) cbow_output.most_similar(positive=['virus']) The output shows the words that are most similar to the word ‘virus’ along with the sequence or degree of similarity. The words like symptoms and incubation are contextually very accurate with the word virus which proves that CBOW model successfully understands the context of the data. Conclusion In the above article, we saw what a CBOW model is and how it works. We also implemented the model on a custom dataset and got good output. The purpose here was to give you a high-level idea of what word embeddings are and how CBOW is useful. These can be used for text recognition, speech to text conversion etc.","excerpt":"In this article, we will learn about what CBOW is, the model architecture and the implementation of a CBOW model on a custom dataset.","categories":["Deep Tech"],"tags":["Natural Language Processing"],"author_name":"Bhoomika Madhukar","publish_date":"2020-09-10T17:00:30","publication_year":"2020","word_count":993,"keywords":["Go","NumPy","TPU","Keras","programming_languages:R","AI","neural network","Natural Language Processing","programming_languages:Go","data_tools:NumPy","R"],"extracted_tech_keywords":["AI","neural network","Keras","NumPy","TPU","R","Go","programming_languages:R","programming_languages:Go","data_tools:NumPy"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/the-continuous-bag-of-words-cbow-model-in-nlp-hands-on-implementation-with-codes\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10080583,"title":"Apple’s Privacy Play Lands the Company in Trouble… Again","content":"Apple has been touted as the brand built on privacy. But, nevertheless, it has been subject to numerous instances of data infringement. In April this year, Israeli firm Pegasus Spyware successfully hacked into iPhones belonging to human rights defenders, lawyers and journalists. In the latest round of troubles for the company, a report has brought Apple fresh allegations. Tommy Mysk and Talal Haj Bakry, iOS developers and security researchers, detailed that contrary to what Apple wants us to believe, it in fact does collect user information. The findings have landed Apple in a new lawsuit for violating user privacy despite sharing analytics being switched off in the settings. In their report, Mysk and Bakry showed that Apple’s analytics data includes an ID called “dsId”, which upon verification was found to be a Directory Services Identifier. The ID number is used to uniquely identify an iCloud account such that when an API call is made to iCloud, the dsId information carrying user name, email, and any data associated with the user’s iCloud account is sent to Apple. Thus, all the user activity on the App Store—including metrics about the app viewed, and duration of viewing of the app—is sent to Apple even when personalised recommendation and sharing usage data is switched off. Additionally, they confirmed that the same identifier was used for Apple Music, and other company services. Apple has been at the receiving end of criticism on multiple occasions regarding its privacy policy. For instance, in October 2021, Ashley Gjøvik, a human rights lawyer, shared that in response to her user data request, she found that Apple was tracking all her activities in Mails, Maps, and calls, etc. The data, she said, was collected in large xml files that could be traced all the way back to 2017. Apple’s privacy game However, these findings, when viewed against the backdrop of recent privacy updates by Apple present a different picture. In August 2022, Apple revealed that it would bring ads to pre-installed apps like Books, Maps, Podcasts and others on its devices. Additionally, it also tested inputting ads in Maps to recommend stores, restaurants, or businesses nearby. The move showed that Apple intends to push hard on advertising, and leverage upon its App Tracking Transparency (ATT) policy, which gives users the choice to prevent any third-party application from collecting their data. ATT was a major blow to companies like Meta and Google, which in the recent quarterly results, reported a major fall in revenue owing to grim advertising expenditures. In fact, Search Ads, Apple’s platform for advertisers to host campaigns in the App store, tripled in market share since the first-half of 2020, according to AppsFlyer. Apple Search Ads works on a cost-per-tap (CTP) model, which means advertisers only have to pay if users engage with the ad. Apple will promote these apps whenever users arrive on the App Store or search for something specific. One can connect the dots here and argue that Apple is most definitely tracking user activity to display only those ads which a user is more likely to click to have a profitable bet on the CTP model. Privacy breach? While Apple should be made accountable for violating its privacy policy, upon close inspection, Mysk and Bakry’s thesis seems to have some leaky edges. Mysk and Bakry cite an excerpt from Apple’s Device Analytics & Privacy statement, which says, “none of the collected information identifies you [the user] personally”, to show how having a unique identification number is a blatant violation of privacy policy. However, the rules applying to Device Analytics do not pertain to Apple’s services which have a completely different policy. For instance, in the App Store & Privacy document published in September 2022, the guidelines are pretty clear. It says that Apple will have a record of the browsing history, searches, downloads, and purchases stored with a unique identifier, IP address, and Apple ID in order to enable personal recommendation, as well as to provide relevant ads on App Store, Apple News, and Stocks. Additionally, Apple will also collect information about the number of phone calls, and emails sent and received to identify and prevent fraud. Device analytics vs services analytics Nick Heer, writer at Pixel Envy, also points out—device analytics is distinctive to services analytics. The device analytics policy only allows Apple to collect data on performance analytics of iPhone, and how users use their devices and applications, without this information identifying them personally. However, it is not clear if Apple collects personal data outside bug reports and crashes, which is said to fall under “privacy preserving techniques such as differential privacy”. The federated learning based on Apple’s differential privacy is built for use case applications, and, in general, to enhance user experience. In this model, the data collected from a user input is randomised before being sent to a central server. After that, the randomised information is clubbed into batches, subject to private algorithms. This way, Apple has been able to play a safe game, and avoid falling into grey areas of privacy by drawing a thin line between device and application privacy narration. That’s where all the confusion starts, and all hell breaks loose.","excerpt":"Apple’s privacy policy comes under the spotlight again with a fresh lawsuit","categories":["Global Tech"],"tags":["Apple","differential privacy"],"author_name":"Ayush Jain","publish_date":"2022-11-24T11:00:00","publication_year":"2022","word_count":864,"keywords":["federated learning","Go","API","AWS","AI","Apple","ML","RAG","analytics","differential privacy","R"],"extracted_tech_keywords":["AI","ML","analytics","federated learning","differential privacy","RAG","AWS","R","Go","API"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/square-the-circle-apples-privacy-play-lands-it-in-trouble-again\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10141940,"title":"Qwen Agent is Here","content":"Following Qwen 2.5’s popularity among developers for AI agent development, Alibaba Cloud has now released Qwen Agent on GitHub. Technically, Qwen Agent is a framework for developing LLM applications based on following instructions, tool usage, planning, and Qwen’s memory capabilities. It also comes with example applications such as Browser Assistant, Code Interpreter, and Custom Assistant. Qwen Agent provides atomic components, such as LLMs (which inherit from class BaseChatModel and come with function calling) and tools (which inherit from class BaseTool), along with high-level components like agents (derived from class Agent). Furthermore, a user on X mentioned that Qwen Agent comes with a Gradio user interface. Regarding accessibility, users can either utilise the model service by Alibaba Cloud’s DashScope or deploy and manage their own model service using the open-source Qwen models. As per the README in the Github repository, “The LLM classes provide function-calling. Additionally, some Agent classes also are built upon the function calling capability, e.g., FnCallAgent and ReActChat.” Almost a month ago, Qwen2-Math was released – the first series of mathematical LLMs of the Qwen family. Following that, it was upgraded and open-sourced to Qwen2.5-Math series, including base models Qwen2.5-Math-1.5B\/7B\/72B, instruction-tuned models Qwen2.5-Math-1.5B\/7B\/72B-Instruct, and mathematical reward model Qwen2.5-Math-RM-72B. According to Qwen’s blog, the Qwen2-Math series only supports using Chain-of-Thought (CoT) to solve English math problems, while the Qwen2.5-Math series is expanded to support using both CoT and Tool-integrated Reasoning (TIR) to solve math problems in both Chinese and English. Developers Love Qwen Notably, it seems like developers’ experience with Qwen has been largely positive. Beyond its ability to run on mid-range hardware, developers have been finding it better than popular LLMs like ChatGPT. According to one developer, since he started using Qwen 2.5 35B for coding tasks, he has not touched ChatGPT and only uses Claude for planning. “It is local, and it helps with debugging and generates good code. I do not have to deal with the limits on ChatGPT or Sonnet. I am also impressed with its ability to follow instructions and generate JSON output,” he further said. Another developer who extensively tested the model said he created a fully functional Pac-Man game in Python using the 72B model running locally in Q4 quantisation, complete with ghosts, playable map, and sprite loading functionality, outperforming Claude which only managed a basic map implementation. Qwen is also a reasonable choice for developers seeking to reduce dependency on cloud-based solutions, considering the $0.38 per million tokens they get compared to GPT-4o’s $5 per million tokens and Claude 3.5 Sonnet’s $3 per million tokens. Over the past few months, Amazon Web Services (AWS) has introduced Multi-Agent Orchestrator, a framework that offers a solution for managing multiple AI agents and handling complex conversations. The framework routes queries to the most suitable agent, maintains conversational context, and integrates with various environments, including AWS Lambda, local setups, and other cloud platforms. Its GitHub repository highlights its capabilities with six specialised agents, including ones for travel, weather, math, and health. The orchestrator switches between agents to manage multi-turn conversations and diverse tasks while preserving context. Meanwhile, Microsoft Research unveiled Magentic-One, a generalist multi-agent system capable of solving open-ended tasks across diverse domains. Available as an open-source tool on Microsoft AutoGen, Magentic-One helps developers and researchers create agentic applications for managing complex, multi-step tasks autonomously. OpenAI introduced Swarm, a framework for building, orchestrating, and deploying multi-agent systems. Similarly, IBM launched the Bee Agent Framework, an open-source toolkit for creating and deploying agent-based workflows at scale. Currently, in its alpha stage, Bee Agent supports various AI models and offers compatibility with IBM Granite and Llama 3.2 models.","excerpt":"Qwen Agent is a framework for developing LLM applications based on Qwen’s instructions following, tool usage, planning and memory capabilities.","categories":["AI News"],"tags":["Alibaba","Qwen"],"author_name":"Shalini Mondal","publish_date":"2024-11-28T17:05:49","publication_year":"2024","word_count":601,"keywords":["ChatGPT","Qwen","OpenAI","AI","Alibaba","GPT-4o","AWS","multi-agent systems","IBM Granite","AutoGen","Gradio","Claude 3.5"],"extracted_tech_keywords":["AI","GPT-4o","ChatGPT","OpenAI","Claude 3.5","IBM Granite","AutoGen","multi-agent systems","Gradio","AWS"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/qwen-agent-is-here\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10122100,"title":"Jivi&#8217;s Medical LLM Beats OpenAI at the Open Medical LLM Leaderboard","content":"A purpose-built medical Large Language Model (LLM) developed by Jivi, an Indian healthcare AI startup co-founded by former BharatPe Chief Product Officer Ankur Jain and GV Sanjay Reddy, Chairman, Reddy Ventures, has ranked number 1 on the Open Medical LLM Leaderboard. Jivi’s LLM, Jivi MedX, has beaten established LLMs, including OpenAI’s GPT-4 and Google’s Med-PaLM 2 with an average score of 91.65 across the leaderboard’s nine benchmark categories. Hosted by leading AI platform Hugging Face, the University of Edinburgh, and Open Life Science AI, the leaderboard ranks medical-specific LLMs based on their performance in answering medical questions from exams and research. The evaluation covers medical exams such as Indian medical entrance exams (AIIMS and NEET), US Medical License Exams (USMLE), and detailed assessments in clinical knowledge, medical genetics, and professional medicine, among others. “Jivi is revolutionizing primary healthcare through generative AI, making top-quality care accessible 24\/7 at a fraction of the cost. Our mission at Jivi is to harness artificial intelligence to enhance patient care. Our platform accelerates diagnostics and ensures higher accuracy, enabling timely and precise treatment for all,” said Ankur Jain, Co-founder and CEO, JIVI. Jivi currently operates with a lean, 20-member team of physicians, surgeons, AI engineers, and data scientists. It is developing technology to transform accessibility, affordability, and quality of healthcare globally. Jivi uses its large proprietary medical dataset consisting of millions of medical research papers, journals, clinical notes, and other sources to train its Jivi MedX LLM. This dataset is among the largest in the world. Jivi MedX was trained using an instruction fine-tuning algorithm called Odds Ratio Preference Optimisation (ORPO).","excerpt":"Jivi MedX, has beaten established LLMs, including OpenAI’s GPT-4 and Google’s Med-PaLM 2 with an average score of 91.65 across the leaderboard’s nine benchmark categories.","categories":["AI News"],"tags":[],"author_name":"Pritam Bordoloi","publish_date":"2024-05-31T11:50:10","publication_year":"2024","word_count":266,"keywords":["Go","Hugging Face","artificial intelligence","OpenAI","AI","ML","RAG","GPT","generative AI","R"],"extracted_tech_keywords":["AI","artificial intelligence","ML","generative AI","OpenAI","Hugging Face","RAG","R","Go","GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/jivis-medical-llm-beats-openai-at-the-open-medical-llm-leaderboard\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":33193,"title":"Here’s How Twitter&#8217;s ML Algorithms Rank The &#8216;Best&#8217; Tweets On Your Timeline","content":"With over 300 million active users monthly, the amount of data that gets generated on Twitter is huge. Handling this torrential data and making sure that the user doesn’t miss out on any important tweets is a challenging job for the data engineers. Twitter has made many adjustments and tweaked their machine learning algorithms for optimal user experience. To someone in the field of data science, this key takeaway from the article will be the idea how the algorithms are implemented. Non-data science readers will gain an understanding of what’s going behind their screens and why certain tweets show up on their timeline so that they don’t have to grill the CEO demanding the fairness of algorithm for unfair past-time purposes. Understanding The Algorithm Behind Twitter’s Timeline Ranking All the tweets since the user’s last visit are gathered and shown in reverse chronological order. Every tweet is given a relevance score by a model. This score indicates the likelihood of the user finding a certain tweet interesting. So, a collection of tweets with higher scores will show up, increasing their visibility. If the number of relevant tweets is high, there again needs to be an order which is now ranked based on the time of posting. The following factors are considered to make predictions: The presence of image or video in the tweet and number of likes The user’s past interactions with the one who tweeted User’s history of likes and retweets and the time spent on Twitter A typical user has a habit of refreshing the feed every minute or two. So, this adds up to the already complex scoring model. The model has to calculate the scores and show the tweets in real-time. A good model should have very good quality, should utilise resources at high speed and should offer ease of maintenance. Checking The Model’s Quality The first step is to check whether the model is giving high scores to tweets of relevance. A\/B tests are run to measure the impact of a tweet on the user; usage and enjoyment being the usual metric here. A\/B tests allow the engineers to verify the accuracy of the model and also to quantify the improvement. Source: A\/B testing illustration via vwo This test is crucial becomes before deploying a model onto the platform. Unless there is a major improvement in user experience, the model doesn’t see the light of the day. Having the best model is the ultimate goal and it outweighs the reasoning behind the choice of tools and frameworks employed. A model’s usability and longevity are key factors here. Making a successful model, which cannot be understood by outsiders is short term approach and the development gets hindered. A Ranking Algorithm That Runs On Deep Learning With improved architectures and more intuitive understanding of image and natural language understanding, deep learning has found major applications at platforms like Twitter which natively leverage deep learning and complex graphs to meet the ends. Deep learning models can be composed in various ways (stacked, concatenated, etc.) to form a computational graph. The parameters of this graph can then be learned, typically by using back-propagation and SGD (Stochastic Gradient Descent) on mini-batches. Libraries like PyTorch make the computational graphs more dynamic and change from one mini-batch to the other. Tweet ranking belongs to a different domain compared to what the deep learning algorithms are designed for. Due to scattered data and latency requirements, the presence of a feature cannot be guaranteed for each tweet. Initially, Twitter used algorithms like decision trees for timeline ranking. But for the reasons mentioned above, deep learning was considered to be a suitable candidate to obtain better results on inherently sparse Twitter data. Twitter’s Cortex team which works on its deep learning platform made few adjustments to ensure that there is a significant improvement in switching to deep learning, using the following techniques: To tackle the problem of scattered data and sparse feature values, the team had discretised the input’s sparse features before feeding them to the main deep network. A custom sparse linear layer, which has two extra layers: Online normalisation scheme to prevent gradient explosion and, A per-feature bias to distinguish between the absence of a feature and the presence of a zero-valued feature. A custom isotonic calibration layer to explore the space of solution in the training dataset and recalibrate before giving actual probabilities. Automatic bundling to retrain and test within Twitter’s clusters for optimal interaction and transparency The latency issue is covered by the implementation of optimised modules, which use the right mix of batching, multithreading and hardware utilisation. All these adjustments paved way for a successful model which is robust and has shown a considerable rise in the number of audiences and their engagement on Twitter. Looking Forward To use deep learning as the central modelling for timeline ranking is a smart move by Twitter as this space is constantly improving and the solutions can only get better from here on. With new ideas surfacing every week on how to optimise machine learning algorithms, there is a lot of potential for the machine learning based products to benefit from.","excerpt":"With over 300 million active users monthly, the amount of data that gets generated on Twitter is huge. Handling this torrential data and making sure that the user doesn’t miss out on any important tweets is a challenging job for the data engineers. Twitter has made many adjustments and tweaked their machine learning algorithms for […]","categories":["Deep Tech"],"tags":["Machine Learning","twitter analytics"],"author_name":"Ram Sagar","publish_date":"2019-01-10T07:25:54","publication_year":"2019","word_count":858,"keywords":["twitter analytics","data science","Go","machine learning","AI","PyTorch","Machine Learning","RAG","deep learning","ViT","ai_frameworks:PyTorch","R"],"extracted_tech_keywords":["AI","machine learning","deep learning","data science","PyTorch","RAG","R","Go","ViT","ai_frameworks:PyTorch"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/how-twitters-ml-algorithms-rank-the-best-tweets-on-timeline\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":27762,"title":"IIT Kharagpur Introduces New Course For Artificial Intelligence In Law","content":"Indian Institute of Technology, Kharagpur announced the launch of a new course titled Artificial Intelligence and Law this week. The one-semester inter-disciplinary course has been developed at the Rajiv Gandhi School of Intellectual Property Rights (RGSOIPL) and aims to start a discourse on the modes of societal governance and law by using the new era hybrid sociology. Padmavati Manchikanti, the Dean of RGSOIPL, told a newswire, “This course addresses the paradigm shifts that are occurring due to the intense involvement of AI as autonomous systems that can possibly invent, participate in inventions with humans, create new expressions like music, paintings, and raising sensitive questions on inventorship, authorship, ownership of patents, copyright and designs.” The RGSOIPL is one of the first law schools in the country that focuses on technology. Set up in 2006, the basis of the school has been to cater to the need for lawyers with technical expertise. It is a part of their global endeavour to integrate technology and law to build technical lawyers with the expertise to deal with the legal issues in the interface of technology and law. Reportedly, Partha Pratim Chakrabarti, Shreya Matilal and Prabuddha Ganguli of RGSOIPL have been instrumental in initiating and formulating the cutting-edge syllabus for the course. 56 students have already enrolled for the same, said the official statement. Earlier this March, IIT KGP had announced a Center of Excellence in Artificial Intelligence Research with seed funding from Capillary Technologies Limited. With a funding of ₹5.64 crores, the centre was to focus on key areas in AI and its kindred such as training, research, education, projects, entrepreneurship and incubation. The funding is to be utilised for establishing coursework design, computing infrastructure, software and hardware simulation platforms. IIT KGP already boasts of AI specialist in areas such as financial analytics, industrial automation, digital healthcare, intelligent transportation system, agricultural internet of things and analytics, big data analytics for rural development, intelligent urban infrastructure, and safety-critical cyber-physical systems, among others.","excerpt":"Indian Institute of Technology, Kharagpur announced the launch of a new course titled Artificial Intelligence and Law this week. The one-semester inter-disciplinary course has been developed at the Rajiv Gandhi School of Intellectual Property Rights (RGSOIPL) and aims to start a discourse on the modes of societal governance and law by using the new era hybrid sociology. […]","categories":["AI News"],"tags":["AI (Artificial Intelligence)","IIT","iit Kharagpur"],"author_name":"Prajakta Hebbar","publish_date":"2018-08-29T12:28:27","publication_year":"2018","word_count":327,"keywords":["big data","Go","API","artificial intelligence","AI","Git","RAG","Aim","analytics","iit Kharagpur","IIT","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","analytics","Aim","RAG","R","Go","Git","API","big data"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/iit-kgp-ai-law-course\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10072429,"title":"Intel Announces New Line of Ray-Tracing GPUs","content":"Intel has recently announced on SIGGRAPH 2022, their new Arc Pro A-series professional GPUs for mobile form factors and small desktops. The first product in the series—Intel Arc Pro A30M GPU is for laptops and supports built-in ray tracing hardware competing with Nvidia and AMD. It claims to be the industry’s first AV1 hardware encoding acceleration including learning. “Intel Arc Pro A-series graphics are targeting certifications with leading professional software applications within the architecture, engineering and construction (AEC), and design and manufacturing (D&M) industries,” says Intel. The Intel Arc Pro GPUs are optimised to run open source libraries like oneAPI Rendering Toolkit that is used and adopted by industry-leading rendering tools. With plans to launch later in 2022, the GPUs Arc Pro A40 and A50 will be available with 4.8 teraflops of graphical power, eight ray-tracing cores, and 6GB of GDDR6 memory. Both the graphics card have dual-slot designs which are suitable for multi display setup supporting two monitors upto 8K 60Hz, one at 5K 240Hz, two at 5K 120Hz, or four at 4K 60Hz. With support for gaming, Intel is targeting these GPUs for creator applications like Blender, Handbrake, DaVinci Resolve Studio along with engineering, construction and architectural manufacturing. The multicore chips will support learning capabilities and faster rendering speed, allowing GPUs to show great potential for developing on frameworks like Gigapixel AI and competing in the workstation market.","excerpt":"Intel is targeting these GPUs for creator applications like Blender, Handbrake, DaVinci Resolve Studio along with engineering, construction and architectural manufacturing.","categories":["AI News"],"tags":["Intel","Intel processors"],"author_name":"Mohit Pandey","publish_date":"2022-08-09T13:06:10","publication_year":"2022","word_count":230,"keywords":["API","programming_languages:R","AI","Intel processors","Ray","Aim","R","Intel"],"extracted_tech_keywords":["AI","Aim","Ray","R","API","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/intel-announces-new-line-of-ray-tracing-gpus\/","complexity_score":3,"technical_depth":6,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":68828,"title":"Latest Entry-Level Job Openings For Data Scientist &#038; Analysts","content":"Data Science jobs are the most sought after jobs of this century. According to AIM survey on Data Science recruitment, the number of job posts available during the pandemic situation between May and June is around 80,000, and 37% of the respondents believe that data science jobs are available during the current crisis. In this article, we list down the 8 latest job openings for data science and analysts to apply right now. (The list is in no particular order) 1| Data Scientist at Jio Location: Bangalore About: As a Data Scientist, you will be responsible for building the next-generation platform products and services using speech understanding, natural language understanding, computer vision, search and discovery, recommendation systems, knowledge graph, machine learning, deep learning, reinforcement learning, among others. You will be developing products and services in healthcare, agriculture, education, etc. Apply here. 2| Data Scientist at GE Location: Bangalore About: As a Data Scientist at GE, you will identify use cases, test hypothesis, analyse data, build models and predicting user needs. You will work towards gleaning insights from data to add value to our systems, portals and users. Your responsibilities include understanding and complying with data privacy requirements of GE and local region\/country, develop continuously learning models that address use cases in areas like analysing service cloud tickets, personalising oneHR, efficient routing of tickets to agents, understanding Tier 1 cases, predicting phone calls, classification of persona with unique needs, and other such. Apply here. 3| Data Scientist at Alpha Alternatives Location: Mumbai About: As a Data Scientist, you will work integrally with the trading desk and build profitable quantitative strategies, analyse different trading strategies including back-testing and walk-forwards, conduct periodic reviews and validation of the same, solve critical problems with the help of machine learning tools and techniques and other such. Apply here. 4| Master Data Scientist at Inference Labs Location: Bangalore About: The Data Scientist will be part of the core data science team. This team identifies and develops advanced statistical model, machine learning methods and solutions for clients to improve various business outcomes. The objective of the team is to successfully develop, conceptualise and test various statistical models, integrate the outcomes as real-time analytics to create value for clients in areas and through means not immediately apparent to clients. Apply here. 5| Data Analyst at Netomi Location: Gurgaon About: As a Data Analyst, you will help Netomi to dig into raw data, analyse it and draw conclusions that help in making business decisions. Your responsibilities include defining the deep learning (DL) model, ensuring the quality models, analysing the conversation quality in chatbots, and leveraging multiple crowdsourcing strategies to collect training and test data for DL models and help with cleansing, filtering and massaging those data, etc. Apply here. 6| Data Analyst at Riskonnect Inc. Location: Mangalore About: As a Data Analyst, you will be responsible for monitoring on-going client support processes with set guidelines as to how to proceed if errors occur, provide accurate and timely data conversions and\/or updates as outlined in the client agreement, maintain and improve the quality and efficiency of each account assigned, assist in efforts to improve the overall efficiency of data operations and other such. Apply here. 7| Data Analyst at Target India Pvt. Ltd. Location: Bangalore About: As a Data Analyst, you will be a part of the team that works closely with the business and identifies problems\/opportunities for improved decision-making through better data analysis. You will not only work on simple descriptive analysis but also on complex predictive and prescriptive analytics, using advanced modelling and machine learning techniques primarily using open-source technologies and big data platforms. You will also work closely with business\/product teams and understand their priorities\/roadmap. Based on this understanding, you are expected to identify appropriate metrics that will drive the right decisions for the business, and then build reporting solutions to deliver these metrics at the required frequency in an optimal and reliable fashion. Apply here. 8| Data Analyst at Truworth Infotech Pvt. Ltd. Location: Jaipur About: As a Data Analyst, you will be working in big data and analytics to provide insights to the business covering a range of topics, develop clear visualisations to convey complicated data in a straightforward fashion, compile and analyse data related to business’ issues, etc. You will also be responsible for conducting both recurring and ad hoc analysis for business users. Apply here.","excerpt":"Data Science jobs are the most sought after jobs of this century. According to AIM survey on Data Science recruitment, the number of job posts available during the pandemic situation between May and June is around 80,000, and 37% of the respondents believe that data science jobs are available during the current crisis.  In this […]","categories":["AI Hirings"],"tags":["Data analyst jobs","Data Analytics","data cleansing","Data Science Career","Data Science Jobs","Data Scientist"],"author_name":"Ambika Choudhury","publish_date":"2020-07-02T16:00:00","publication_year":"2020","word_count":731,"keywords":["data science","machine learning","AI","chatbots","Data analyst jobs","Data Science Jobs","computer vision","RAG","Data Science Career","recommendation systems","data cleansing","deep learning","Aim","analytics","Data Analytics","Data Scientist"],"extracted_tech_keywords":["AI","machine learning","deep learning","computer vision","data science","analytics","Aim","RAG","chatbots","recommendation systems"],"url":"https:\/\/analyticsindiamag.com\/ai-hiring\/latest-entry-level-job-openings-for-data-scientist-analysts\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10078515,"title":"Satellite Spectrum Raises Moot Question: To Auction or Not to Auction","content":"Satellite services are just around the corner and hold great potential in the Indian context. With the advent of high-throughput satellites with low latency, satellites are capable of working on real-time applications and can provide effective backhaul to the 5G coverage. However, the industry remains divided on how the spectrums should be allocated. With the government considering satellite spectrum auctions, many industry players are of the opinion that auctioning spectrum is not the right approach to adopt. Recently, Bharti Enterprises chairman Sunil Mittal said that spectrum auction did not make sense in the case of satellite services. It can’t be auctioned on the lines of its recent sale for 5G services. Many other members of the space industry echoed similar sentiments and have urged the government to consider an administrative allocation of 28MhZ spectrum instead. The coveted 27.5 GHz to 28.5 GHz band, the only spectrum band which can be used by satellite services providers, was part of the 5G auction. However, industry members wrote to the Telecom Regulatory Authority of India (TRAI) to not auction the spectrum as many Indian startups were already working on developing satellites in the 28GHz spectrum. Why auctioning is an issue The space industry in India is heating up and many companies are already eyeing a share of the pie. Companies such as Sunil Mittal-owned Bharti group’s OneWeb, Amazon’s Project Kuiper, Telesat and Starlink could enter the Indian market at some point in time. Undoubtedly, spectrum auctions have worked for India. The Indian government has been assigning spectrum for terrestrial services for over a decade. Hence, it’s understandable why the government is considering an auction for the 28GHz spectrum band as well. However, the terrestrial spectrum is different from the satellite spectrum because of its higher frequency reuse capability. Disallowing the reuse and sharing of spectrum, by auctioning and giving to one entity will result in an acute worldwide shortage of frequencies. “There will be pressure to reduce spectrum availability for IMT services as each satellite will have to use a different set of frequencies,” Anil Prakash, director general at Satcom Industry Association (SIA) India, told AIM. Explaining further, he said that the process of spectrum auction allows one entity to use a set of frequencies to the exclusion of others for running a terrestrial communications service in a given geographic area. “In the case of satellites, however, a set of frequencies is used by many satellites that could be operated by the same or multiple entities for a geographic area.” In a virtual briefing, Mittal said, “This is not going to be used in every part of the country. This is going to be only in two landing stations at those specific points. World over, there has never been an auction for satellite spectrum. You can’t have 1,000 MHz at Rs 7,000 crore just to serve two 15-acre spots in the country.” Besides, auctioning satellite spectrum would also result in the creation of gatekeepers, meaning a few companies would control the industry and would block new entrants and hence kill competition. “Auctioning satellite spectrum and excluding multiple operators to use the same set of frequencies is not just an ineffective method but also constrains the communication capabilities of nations,” Prakash said. Administrative allocation Members of the space industry, including SIA, have urged the government to take the administrative route where the latter charges a fee to use the spectrum. If the government does take the administrative route, it would mean different operators could use the same frequencies across multiple satellites without interfering with each other. Further, it would also allow different players to enter the market at any given time. Taking the administrative route will eliminate the creation of gatekeepers, making the market competitive. Prakash said that the administrative route is preferred by other nations as well. In the US, the government has enacted the Orbit Act, which prohibits the auction of spectrum for satellite services. “Administratively allocating reusable and shareable satellite spectrum is followed by all the countries to benefit from the most efficient use of spectrum and creating communication capabilities for a nation,” Prakash said. The telcos think otherwise Interestingly telcos such as Reliance Jio and Vodafone Idea are in favour of satellite spectrum auctions. Earlier this year, Reliance Jio urged TRAI to discard the concerns raised by members of the space industry and go for the spectrum auction. Reliance Jio is of the view that satellite spectrum auction is gaining popularity. While Brazil has already gone ahead with the auction, nations such as Saudi Arabia, Mexico and Thailand are expected to follow suit. One reason for the telcos to push for auctions is that satellites can deliver ubiquitous coverage in a faster time frame. Even though the terrestrial operators have been allocated spectrum at the bare minimum reserve, they will take a few years to roll out the 5G networks, and even then it will not be ubiquitous coverage. Besides, they argue that satellite providers can get access to the spectrum by paying the administrative fee, which is significantly lower than an auction price. “The business compulsions of the terrestrial operators make them push for satellite spectrum and delay the launch of satellite-based communication services, depriving the nation and its citizens of the much-needed modern-day communication capabilities,” Prakash said. Since both routes come with their own set of pros and cons, it does put the government in a difficult position. Hence, it would be interesting to see what route the government takes.","excerpt":"Disallowing the reuse and sharing of spectrum, by auctioning and giving to one entity will result in an acute worldwide shortage of frequencies","categories":["AI Features"],"tags":[],"author_name":"Pritam Bordoloi","publish_date":"2022-11-03T11:00:00","publication_year":"2022","word_count":912,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","RAG","Aim","R","startup"],"extracted_tech_keywords":["AI","Aim","RAG","R","Go","startup","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/satellite-spectrum-raises-moot-question-to-auction-or-not-to-auction\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10061517,"title":"Council Post: Does AI-fication of macroeconomics lead to better solutions?","content":"Consider the impact of higher policy rate on a day to day consumer decision like eating out. Possibilities are many: high, low and no impact. The impact can vary by region, demographics, ethnicity and income groups. Hospitality industry will be keen to understand the impact of change in policy rates or other macro factors and how they are impacting the daily lives of the consumers. Macroeconomics is often described as ‘pseudo science’ and lacks granular insights. On the other hand, data science, ML and AI are too focused on nuanced problems restricting them from having a broader vision. But, the question is, can we combine these two to make insights more holistic and complete? AI\/ML in economics Today, large tech firms have economists who use data to estimate key parameters that help them curate tech policy. The economists work very closely with data sciences, applied scientists and developers to help inform internal decisions by leveraging data. At the same time, ML with econometrics has also become an active research area in economics, thanks to the availability of large datasets, especially in microeconomic applications. However, applying ML to economics is easier said than done. To begin with, it requires finding relevant tasks. Moreover, the field is still nascent, and very little progress has been made to understand ML models’ properties when applied to predict macroeconomic outcomes. Econometrics applies statistical modelling that helps understand complex social and environmental issues. It is one of the emerging areas of applied economics. The whole area of ‘econometrics’ is based on applying statistical models, including descriptive statistics and hypothesis testing, correlation and regression analysis, time series analysis and visualisation, predictive analytics and forecasting, panel data modelling (like OLS, fixed and random-effects models), etc. Economists’ job roles are similar to data scientists. For example, behavioural economists quantify consumer response and attitude to marketing campaigns, discounts, price reduction, etc. Similarly, industrial economists check the level of commodity production and pricing mechanism for given market demand. How AI\/ML can enrich macroeconomics In macroeconomics, the predictions\/forecasts made by economists always seem to be terribly wrong. This is where AI\/ML comes into play. Predictive analytics can help overcome some of the challenges in economic forecasting. By leveraging AI techniques in behavioural economics, economists can accurately estimate the impact of human perceptions and behaviour. For example, JPMorgan used an algorithm to track the effects of President Trump’s tweets on financial markets. Also, central banks and financial institutions can be more effective by predicting recession hits, thereby mitigating the effects on the business. Economists can also bring changes in supply and demand and implement necessary changes to avoid economic turmoil. Most importantly, based on the real-time data, economists can help companies curate technology policies to run businesses smoothly and avoid downturns. Why is AI-fication of macroeconomics important? While macroeconomics zooms out of the minute details and views the world as a large interconnected system, microeconomics focuses on the interaction between basic building blocks of the society and the result of the interaction. Some of the examples include – How a local business decides to allocate their spends How the government decides to spend the tax incomeThe housing market of a particular location\/cityProduction\/manufacturing of a local business\/product; etc. Simply put, microeconomics is related to decision-making with low-level effects, say, allocations of spending, location\/city, etc. But, on the other hand, macroeconomics has a high-level, large-scale impact, say the business, government, housing market, production, etc. Macroeconomic analysis can help data science to make insight generation more comprehensive. Here’s how Spectrum of analysis: Capturing both macro and micro variables enables combining the impact of macro trends with the micro big data-driven insights. Completeness of insights: Businesses can assess the impact of economic trend changes on very nuanced challenges that they face. They may be intrigued to understand gritty insights that shape the new age goods and services offering. Societal benefit: The combined solution has greater potential to unlock value and impact our daily lives and our decisions towards broader societal change. A final thought As the name suggests, macroeconomic analysis is macro in nature and explores the relationship between the macro variables and analyses the macro trends impacting the economies, public policies, and industries broadly. On the other hand, data science and AI targets industry or business-specific challenges, which are often restricted given certain micro assumptions. While macroeconomic analysis lacks granular insights, data science, ML and AI don’t have a broader vision. Thus, combining these two is a great idea to have a more holistic solution covering both the macro and micro sides of an industry or business pain points. This article is written by a member of the AIM Leaders Council. AIM Leaders Council is an invitation-only forum of senior executives in the Data Science and Analytics industry. To check if you are eligible for a membership, please fill out the form here. The views, thoughts, and opinions expressed in this article belong solely to the author and does not reflect the views and opinion of the author’s employer, any other organizations, committee or other group or individual.","excerpt":"By leveraging AI techniques in behavioural economics, economists can accurately estimate the impact of human perceptions and behaviour.","categories":["AI Features"],"tags":["AI Applications","economics"],"author_name":"Indrajit Mitra","publish_date":"2022-02-25T14:00:00","publication_year":"2022","word_count":842,"keywords":["big data","data science","Go","AI","R","ML","AI Applications","RAG","economics","Aim","analytics","predictive analytics"],"extracted_tech_keywords":["AI","ML","data science","analytics","Aim","RAG","predictive analytics","R","Go","big data"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/council-post-does-ai-fication-of-macroeconomics-lead-to-better-solutions\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10040918,"title":"MLOps For Dummies","content":"“87% of data science projects never make it into production.” The machine learning community is now gearing up for a new challenge–deployment. But why so much fuss about something so obvious? You build something to deploy, right? Well, not really. Many machine learning models never see the light of the day. And, those that make into production make little noise. This is why the problems in production are pushed under the rug for a better part of last decade’s AI hype cycle. We read articles about state-of-the-art algorithms and Unicorn AI startups, but how well is ML being productionised? Organisations are finally coming to terms with these challenges. And, they found a hero in MLOps– a meticulous marriage of machine learning and software engineering. At the third edition of Rising organised by Analytics India Magazine, Hamsa Buvaraghan of Google Cloud gave the audience a glimpse of how MLOps power machine learning pipelines of the future. Hamsa leads Google Cloud’s Data Science and MLOps Solution team building revolutionary software solutions for business problems using Google’s Data Analytics and AI\/ML products. In her talk, she demonstrated how the solutions of MLOps fit right into the aspirations of automating the ML pipeline. Why MLOps? Organisations hardly make it beyond pilots and proofs of concept. 72% of organisations that began AI pilots couldn’t deploy even a single application in production. According to a recent survey, 55% of companies have not deployed an ML model. Models don’t make it into production, and if they do, they break.Teams do not have reusable or reproducible components, and their processes involve difficulties in handoffs between data scientists and IT. Deployment, scaling, and versioning efforts still create headaches. Who needs MLOps MLOps bridges the glaring gap between machine learning development and deployment, similar to how DevOps and DataOps support application engineering and data engineering. According to Google Cloud, successful deployments and effective operations are a bottleneck for getting value from AI. MLOps is an engineering culture and practice that aims at unifying ML system development(Dev) and ML System operations (Ops). Hamsa stressed that ML Code is a small portion. From configuration to monitoring, from serving infrastructure to resource management, building production-grade machine learning systems require much more than just code. (Source: Google Cloud) Building an ML-enabled system is a multifaceted undertaking that combines data engineering, ML engineering, and application engineering tasks. “It takes a village to put together an MLOps pipeline,” said Hamsa. For instance, some foundational capabilities are required to support any IT workload, such as a reliable, scalable, and secure compute infrastructure. MLOps capabilities include experimentation, data processing, model training, model evaluation, model serving, online experimentation, model monitoring, ML pipeline, and model registry. An ideal MLOps pipeline can take care of machine learning development, training operationalisation concerns, model deployment concerns, data & model management and a lot more. Hamsa believes organisations are now moving towards automated end-to-end pipelines, and MLOps will have applications in many industries. One of the most significant features of MLOps is its ability to tap into ML metadata and artifact repository along with dataset and feature repository. Artifacts can be anything: processed data splits, schemas, statistics, hyperparameters, models, or evaluation metrics of the model, to name a few. “These ML pipelines are a combined effort of data scientists, data engineers, SREs and ML engineers.” The principle Changing Anything Changes Everything or CACE, refers to the dependency on minor changes in a software engineering pipeline. In the context of machine learning, this principle extends to hyper-parameters, learning settings, sampling methods, convergence thresholds, data selection, and essentially every other possible tweak. Various ML artifacts are produced in different processes of the MLOps life cycle, including descriptive statistics and data schemas, trained models, and evaluation results. Added to this is metadata, which is the information about these artifacts. Benefits of MLOps MLOps enables shortened development cycles.MlOps leads to an increase in reliability, performance, scalability, security of produced ML systems.MLOps allows the creation of the entire ML lifecycle.It also helps manage risk when organisations scale ML applications to more use cases in changing environments. Model governance, version control, explainability and other nitty-gritty of deploying a machine learning model presents a nightmarish scenario to an ML practitioner oblivious of software engineering etiquettes. MLOps, with its myriad of options and a growing developer community, is the best possible solution today to tackle the realities of productionising a model.","excerpt":"“87% of data science projects never make it into production.” The machine learning community is now gearing up for a new challenge–deployment. But why so much fuss about something so obvious? You build something to deploy, right? Well, not really. Many machine learning models never see the light of the day. And, those that make […]","categories":["Deep Tech"],"tags":["Google Cloud","MLOps"],"author_name":"Ram Sagar","publish_date":"2021-05-28T16:00:00","publication_year":"2021","word_count":729,"keywords":["data science","Google Cloud","machine learning","AI","ML","MLOps","RAG","Aim","analytics","model monitoring","model registry"],"extracted_tech_keywords":["AI","machine learning","ML","data science","analytics","MLOps","Aim","model registry","model monitoring","RAG"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/mlops-for-dummies\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":9868,"title":"Indian Railways next big station &#8211; Big Data Junction!","content":"How’s this for thought? In the NDA’s maiden Railways budget presented in 2015, the word ‘technology’ was mentioned only once. There were other 13-14 references to Braille facilitated coaches etc., but nothing more. In the light of this, the 2016 budget wasn’t predicted to be any different. Suresh Prabhu’s 2nd budget, however more than made up for the absence of technological initiatives of last year’s with about 21 specific ‘technology’ mentions. In his speech, Suresh Prabhu used terms such as analytics, drones and geospatial technology – terminology one doesn’t generally associate with the budget and so far certainly not with the Indian railways. So in that sense, it was a pleasant surprise and many of these initiatives, if taken forward, will be vital to the larger plan of transforming an old-economy mammoth into a modern-day, fast, efficient, safe and consumer-friendly entity. Get ready for Indian Railways 2.0. What was even closer to the heart was that the minister, in the course of his speech, also mentioned that his ministry plans to set up a “committed, cross-functional team” called the Special Unit for Transportation Research and Analytics (SUTRA) to oversee this task. Surprises aside, this makes a whole lot of sense.  According to the Google Zeitgeist, the annual search trend results showcased by the search engine giant, IRCTC is unarguably India’s largest e-commerce portal, dwarfing entities such as Amazon, Flipkart and others in terms of daily visitors. IRCTC is India’s largest e-commerce portal by revenue as per information filed with the Registrar of Companies. E-ticketing constitutes 55% of total reserved tickets on Indian Railways. IRCTC’s profits for the financial year 2014-15 stood at Rs. 1,141 crores while total sales for the same period was a whopping Rs 20,620 crores. With approximately 5 lakh visitors per day, IRCTC has come a long way from selling 27 tickets when it first began. In fact perhaps what is not too well known is that India’s biggest online retailer Flipkart – drew its inspiration from IRCTC. Big Data The clear highlight for the Railways in recent years has been monetization. The writing is on the wall – the Railways needs to gradually pay for itself and there is no reason why it shouldn’t. According to media and other reports, the railways incurred a loss of over Rs. 26,000 crores in 2014. With no fare hikes and an increased wage bill starting from January this year, that loss will only increase. Boosting efficiencies and cross-leveraging assets, resources and opportunities is a solution. In light of this, the announcement of SUTRA makes a lot of sense. It’s about time that the railways made sense of their data and then looked at ways of monetizing this. Just think about it. The data that the Railways generates is mind-boggling and here I’m not just speaking of data stored through online bookings. Add to this the enormous quantities of data generated by GPS, AEI readers, electronic data interchange, video inspections, handheld field tablets, and multiple other sources. The Indian Railways generates approximately 100 terabytes of data each year. This is set to increase even further with Google’s initiative of offering free WiFi across 100 railway stations (Mumbai Central was the first station to be offered this). At one glance the Indian Railways will know exactly what millions of Indians across socio-economic spectrum consume online, on a real-time and daily basis. Now imagine this data being configured to offer additional passenger services. Not just generic ‘one-size-for-all’ of packages but customized by the customer, seasons and even destinations. We already know that IRCTC has been actively tying up with hotel portals, e-catering services, payment gateways, food franchises etc. These are the way forward, but what will be effective for the railway giant to understand and analyze traveling pattern of the passenger, just like how private airlines do for their frequent customers. For example, I would know at a glance how many times in the past 10 years Mr. Arun Kumar, age 65, has traveled from Delhi to Vaishno Devi with his family. I would know his dietary preferences as well as his class of travel (let’s assume he travels 2nd AC). Now that I know his pattern of travel, destination themes, and other preferences, what prevents me from offering Mr. Kumar a special discounted summer package for a multi-city ‘pilgrimage special’ trip? I already have tie-ups with hotels; all I need is a tie-up with a local transport service provider (let’s say that could be with Ola or Uber or Meru or all three depending on the cities); local restaurants and an additional sight-seeing package, all at slightly discounted rates. A tie-up with a bank would ensure an EMI payment as well. This is not as overwhelming as it may seem. The Railways are already offering these packages with the luxury tourist trains for high-paying foreign and Indian tourists. Technically speaking this could be taken forward and mass-marketed. Not only do the Railways earn more through Mr. Kumar (and basis his experience make this a regular feature) they also earn through partnerships with the private service providers. Any discounts offered, therefore, is off-set through this additional revenue and hopefully, they’ll still be left with more than just pocket change. This is just one example of what is possible. Basis passenger data a host of additional services can be offered – some simply to streamline services better and others that can be monetized. Another effective way of measuring customer sentiments is by social media trends and analyzing call center data in order to close loopholes, offer new services and mold offerings for the consumer’s benefit. However, the Indian railways need to sift, analyze and apply insights effectively and on more or less real time basis in order to make the railways an operative and coordinated experience for the passengers. The Freight Weight Surprisingly, it’s not the passenger segment where the IRCTC makes the money, it’s the freight sector that churns the profit for them on a daily basis. It is a need of the hour to enable freight roads to use sensors, Big Data analytics, cloud computing and other technology to gather and analyze information from various sources and data streams — and then use that analysis to drive solutions, effectively manage operations and, possibly, offer new services.  But it is not easy as it sounds, various issues plague this sector in India. Due to under-investment, there has been heavy congestion on the rail networks and has caused the failure of the system to house more trains and increase the speed of goods trains. Therefore, the need of the hour is to commence a massive infrastructure expansion and decongestion program coupled with upgradation of technology and careful electrification of tracks along with enhancement of terminal capacity to expand and upgrade this sector. Since it is the money churner for the Indian Railways, the changes should be initiated as soon as possible to reap the maximum benefit. A Secure & Efficient System Safety and efficiency are the other critical issues that can be increased basis data analytics. Over medium, to long term, this can save the Railways significant sums of money. This is where Predictive Maintenance in Rail comes into play. In addition to this being used internationally, closer home this is used by the Delhi Metro and it helps them predict possible break-downs and other technical problems which can then be dealt with, literally, before it happens. Given that the Railways operate stock (wagons, engines, signaling systems and other equipment) that is not spanking new, this is an area where both money and time can be saved, thereby improving passenger and freight services. Now isn’t this what you’d call a win-win formula?","excerpt":"How’s this for thought? In the NDA’s maiden Railways budget presented in 2015, the word ‘technology’ was mentioned only once. There were other 13-14 references to Braille facilitated coaches etc., but nothing more. In the light of this, the 2016 budget wasn’t predicted to be any different. Suresh Prabhu’s 2nd budget, however more than made […]","categories":["IT Services"],"tags":["indian railways"],"author_name":"Sunil Jose","publish_date":"2016-05-09T12:10:22","publication_year":"2016","word_count":1279,"keywords":["indian railways","Go","big data","AI","cloud computing","ML","RAG","ViT","analytics","GAN","R"],"extracted_tech_keywords":["AI","ML","analytics","RAG","cloud computing","R","Go","big data","GAN","ViT"],"url":"https:\/\/analyticsindiamag.com\/it-services\/indian-railways-next-big-station-big-data-junction\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10022701,"title":"The Worth Of Prompts In Pre-trained Models","content":"Recently, researchers from Hugging Face showed task-specific prompting provides several benefits while fine-tuning pre-trained language models. In a paper called “How Many Data Points is a Prompt Worth?”, researchers stated prompting impacts the pre-trained models’ efficiency and is often worth hundreds of data points on average across classification tasks. Fine-tuning via an explicit classifier head is one of the critical paradigms for adapting pretrained models for classification. Besides this approach, there are popular alternatives, such as adapting the pretrained language model directly as a predictor through autoregressive text generation, completion of a cloze task, among others. The cloze task method has previously been used for fine-tuning the popular T5 transformers. When fine-tuning pre-trained language models for classification, researchers mostly used two techniques: a generic model head or a task-specific prompt for prediction. In the head-based transfer learning setting, a generic head layer takes pretrained representations to predict an output class. In the prompt-based method, a task-specific pattern string is designed to coax the model into producing a textual output corresponding to a given class. According to researchers, both approaches can be used for fine-tuning with supervised training data. However, prompts further allow the user to customise patterns to help the model. In this research, the authors mainly discussed the later technique and how it benefits low-data regimes in pre-trained models. Why Prompting “The intuition of prompts is that they introduce a task description in natural language, even with few training points,” said the researchers. Classification by direct language generation allows us to pick custom prompts for each task. The approach can also be used for zero-shot classification, priming, and fine-tuning to provide extra task information to the classifier, especially in the low-data regime. As prompting has been used both for zero-shot and fine-tuning based methods, to understand the importance of promoting, the researchers introduced a metric called the average data advantage for quantifying the impact of a prompt in practice. “Our experiments find that the impact of task targeted prompting can nicely be quantified in terms of direct training data and that it varies over the nature of different tasks,” researchers said. The Tech Behind The researchers run all the experiments with RoBERTa-Large (355M parameters) from the RoBERTa model. The evaluation was performed on SuperGLUE and MNLI datasets. These datasets comprise various tasks, all in English, including entailment, multiple-choice question answering, MultiRC, commonsense reasoning, WiC, among others. They compared the language models across various available data, starting with ten data points and building exponentially to the entire dataset. The researchers ran every experiment four times to reduce variance, for a total of 1,892 training runs, across all tasks. At every point, they reported the best performance achieved at that amount of data or lower. Wrapping Up The research showed that prompts provide a method for injecting task-specific guidance, beneficial in low-data regimes. It has also demonstrated that prompting offers a substantial advantage in terms of data efficiency for almost all the tasks and adds the equivalent of hundreds of data points on average. Further analysis showed that prompting is most robust to pattern choice and can even learn without an informative verbalizer. On large datasets, prompting is similarly helpful in terms of data points, although they are less beneficial in performance. Read the paper here.","excerpt":"Recently, researchers from Hugging Face showed task-specific prompting provides several benefits while fine-tuning pre-trained language models. In a paper called “How Many Data Points is a Prompt Worth?”, researchers stated prompting impacts the pre-trained models’ efficiency and is often worth hundreds of data points on average across classification tasks. Fine-tuning via an explicit classifier head […]","categories":["AI Trends"],"tags":["Hugging Face","pre-trained models","Transfer Learning","Transformers"],"author_name":"Ambika Choudhury","publish_date":"2021-03-22T15:00:00","publication_year":"2021","word_count":545,"keywords":["Hugging Face","TPU","RoBERTa","AI","Transformers","RAG","pre-trained models","BERT","llm_models:BERT","Transfer Learning","R","T5"],"extracted_tech_keywords":["AI","Hugging Face","Transformers","RAG","TPU","R","BERT","T5","RoBERTa","llm_models:BERT"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/the-worth-of-prompts-in-pre-trained-models\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10053487,"title":"Why Tree-Based Models are Preferred in Credit Risk Modeling?","content":"Credit risk modeling is a field where machine learning may be used to offer analytical solutions because it has the capability to find answers from the vast amount of heterogeneous data. In credit risk modeling, it is also necessary to infer about the features because they are very important in data-driven decision making. In contrast to credit risk, we will examine what credit risk is and how it can be represented using various machine learning algorithms in this post. We will implement the credit risk modeling with different machine learning models and will see how tree-based models outperform other models in this task. The following are the main points to be discussed. Table of Contents What is Credit RiskWhat is Credit Risk ModelingHow Machine Learning Is Used in Credit Risk Modelling?Implementing Credit Risk ModellingThe outperformance of Tree-Based Models Let’s start the discussion by understanding what is credit risk. What is Credit Risk Credit risk refers to the likelihood that a borrower will be unable to make regular payments and will default on their obligations. It refers to the possibility that a lender will not be paid for the interest or money given on time.  The cash flow of the lender is disrupted, and the cost of recovery rises. In the worst-case scenario, the lender may be obliged to write off some or all of the loan, resulting in a loss. It is incredibly tough and complex to predict a person’s likelihood of defaulting on a debt. Simultaneously, appropriately assessing credit risk can help to limit the chance of losses due to default and late payments. As recompense for taking on credit risk, the lender receives interest payments from the borrower. The lender or investor will either charge a higher interest rate or refuse to make the loan if the credit risk is higher. For the same loan, a loan applicant with a solid credit history and regular income will be charged a lower interest rate than one with a terrible credit history. What is Credit Risk Modeling A person’s credit risk is influenced by a variety of things. As a result, determining a borrower’s credit risk is a difficult undertaking. Credit risk modelling has entered the scene since there is so much money relying on our ability to appropriately predict a borrower’s credit risk. Credit risk modelling is the practice of applying data models to determine two key factors. The first is the likelihood that the borrower will default on the loan. The second factor is the lender’s financial impact if the default occurs. Credit risk models are used by financial organizations to assess the credit risk of potential borrowers. Based on the credit risk model validation, they decide whether or not to approve a loan as well as the loan’s interest rate. New means of estimating credit risk have emerged as technology has progressed, like credit risk modelling using R and Python. Using the most up-to-date analytics and big data techniques to model credit risk is one of them. Other variables, such as the growth of economies and the creation of various categories of credit risk, have had an impact on credit risk modelling. How Machine Learning Is Used in Credit Risk Modelling? Machine learning enables more advanced modelling approaches like decision trees and neural networks to be used. This introduces nonlinearities into the model, allowing for the discovery of more complex connections between variables. We selected to employ an XGBoost model that was fed with features picked using the permutation significance technique. ML models, on the other hand, are frequently so complex that they are difficult to understand. We chose to combine XGBoost and logistic regression because interpretability is critical in a highly regulated industry like credit risk assessment. Implementing Credit Risk Modelling Credit risk modelling in Python can assist banks and other financial institutions in reducing risk and preventing financial catastrophes in society. The goal of this article is to create a model that can predict the likelihood of a person defaulting on a loan. Let’s start by loading the dataset. import pandas as pd import matplotlib.pyplot as plt import seaborn as sns import numpy as np from sklearn.model_selection import train_test_split, cross_val_score, KFold from sklearn.preprocessing import LabelEncoder from sklearn.ensemble import RandomForestClassifier from sklearn.naive_bayes import GaussianNB from sklearn.neighbors import KNeighborsClassifier from sklearn.linear_model import LogisticRegression from sklearn.tree import DecisionTreeClassifier # load the data loan_data = pd.read_csv('\/content\/drive\/MyDrive\/data\/loan_data_2007_2014.csv') When you look at the Colab notebook for this implementation, you’ll find that numerous columns are identifiers and do not include any meaningful information for creating our machine learning model. Id, member id, and so on are some examples. Remember that we want to build a model that predicts the likelihood of a borrower defaulting on a loan, therefore we won’t need qualities that relate to events that happen after a person defaults. This is because this information isn’t available at the time of loan approval. Recoveries, collection recovery fees, and so on are examples of these features. The code below displays the columns that have been eliminated. #dropping irrelevant columns columns_to_ = ['id', 'member_id', 'sub_grade', 'emp_title', 'url', 'desc', 'title', 'zip_code', 'next_pymnt_d', 'recoveries', 'collection_recovery_fee', 'total_rec_prncp', 'total_rec_late_fee', 'desc', 'mths_since_last_record', 'mths_since_last_major_derog', 'annual_inc_joint', 'dti_joint', 'verification_status_joint', 'open_acc_6m', 'open_il_6m', 'open_il_12m', 'open_il_24m', 'mths_since_rcnt_il', 'total_bal_il', 'il_util', 'open_rv_12m', 'open_rv_24m', 'max_bal_bc', 'all_util', 'inq_fi', 'total_cu_tl', 'inq_last_12m','policy_code',] loan_data.drop(columns=columns_to_, inplace=True, axis=1) # drop na values loan_data.dropna(inplace=True) Now you might know that while preparing the data multicollinearity should be failed because the highly correlated variable provides the same information and those are redundant if we don’t then models will fail to estimate the relationship between the dependent and independent variables. To check the multicollinearity we will draw the heatmap of the correlation matrix obtained with help of the panda’s correlation matrix. The heat map is shown below. As can be seen, several variables are highly correlated and should be eliminated. ‘loan amnt’, ‘funded amnt’, ‘funded amnt inv’, ‘installment’, ‘total pymnt inv’, and ‘out prncp inv’ are multi-collinear variables. If you look through the Notebook, you’ll notice that several variables aren’t in the right data types and need to be pre-processed to get them into the right format. We will define some functionalities to aid in the automation of this procedure. The functions that were used to transform variables to data are coded as below. def Term_Numeric(data, col): data[col] = pd.to_numeric(data[col].str.replace(' months', '')) term_numeric(loan_data, 'term') def Emp_Length_Convert(data, col): data[col] = data[col].str.replace('\\+ years', '') data[col] = data[col].str.replace('< 1 year', str(0)) data[col] = data[col].str.replace(' years', '') data[col] = data[col].str.replace(' year', '') data[col] = pd.to_numeric(data[col]) data[col].fillna(value = 0, inplace = True) def Date_Columns(data, col): today_date = pd.to_datetime('2020-08-01') data[col] = pd.to_datetime(data[col], format = \"%b-%y\") data['mths_since_' + col] = round(pd.to_numeric((today_date - data[col]) \/ np.timedelta64(1, 'M'))) data['mths_since_' + col] = data['mths_since_' + col].apply(lambda x: data['mths_since_' + col].max() if x < 0 else x) data.drop(columns = [col], inplace = True) In our dataset, the goal column is loan status, which has different unique values. These values must be converted to binary. That is a score of 0 for a bad borrower and a score of 1 for a good borrower. In our situation, a bad borrower is someone who falls into one of the categories listed in our target column. Charged off, Default, Late (31–120 days), Does not comply with credit policy Charged Off Status The remaining debtors are considered to be good borrowers. # creating a new column based on the loan_status loan_data['good_bad'] = np.where(loan_data.loc[:, 'loan_status'].isin(['Charged Off', 'Default', 'Late (31-120 days)', 'Does not meet the credit policy. Status:Charged Off']), 0, 1) # Drop the original 'loan_status' column loan_data.drop(columns = ['loan_status'], inplace = True) Now we have some more variables that are in categorical type and need to convert into numbers for further modelling for that we will be using the Label Encoder class from the sklearn library as below. categorical_column = loan_data.select_dtypes('object').columns for i in range(len(categorical_column)): le = LabelEncoder() loan_data[categorical_column[i]] = le.fit_transform(loan_data[categorical_column[i]]) Now, we are all set to train the various algorithms and will check which will perform best. Here we are evaluating one linear model, one Neighborhood model, two tree-based models, and one Naive-Bayes model. We will do cross_validation using KFold for 10 folds and will check mean accuracies for those folds. # compare models models = [] models.append(('LR', LogisticRegression())) models.append(('KNN', KNeighborsClassifier())) models.append((DT, DecisionTreeClassifier())) models.append(('NB', GaussianNB())) models.append(('RF', RandomForestClassifier())) results = [] names = [] for name, model in models: kfold = KFold(n_splits=10) cv_results = cross_val_score(model, x_train, y_train, cv=kfold) results.append(cv_results) names.append(name) msg = \"%s: %f (%f)\" % (name, cv_results.mean(), cv_results.std()) print(msg) The Outperformance of Tree-Based Models As we can see from the above mean accuracies the tree-based models performed far better than the rest of the others. This is because Tree-based algorithms provide great accuracy, stability, and interpretability to prediction models. They map nonlinear interactions pretty well, unlike linear models. They can adjust to any situation and solve any challenge (classification or regression). Because tree creation requires no domain knowledge or parameter configuration, it is ideal for exploratory knowledge discovery. Multidimensional data can be handled via decision trees. Attribute selection measures are used during tree construction to choose the attribute that best splits the tuples into distinct classes. Many of the branches in a tree may reflect noise or outliers in the training data. Tree trimming aims to locate and delete such branches in order to improve classification accuracy on data that isn’t visible. In addition to these all, applications like Credit risk modeling where feature importance plays a very important role as it is going to decide the predictions. Using Decision Tree and likewise algorithms we can obtain feature importance maps and can tune models accordingly. Below you can see the feature importance map given by the Decision Tree algorithm. In many kinds of data science challenges, methods including decision trees, random forests, and gradient boosting are often used. Final Words Through this post, we have discussed in detail credit risk and credit risk modelling. We have seen types of credit risk, factors affecting credit risk, and seen how ML can be used to model credit risk rather than the conventional method. Later we have seen the practical implementation of modelling where we have tested various models and concluded how tree-based algorithms have outperformed and hence these are preferred in such tasks. References Guide to Credit Risk ModelingPredicting Credit Risk – Model PipelineDatasetLink for above codes","excerpt":"Credit risk refers to the likelihood that a borrower will be unable to make regular payments and will default on their obligations.","categories":["AI Trends"],"tags":["credit risk machine learning","Data Science","Machine Learning","Python"],"author_name":"Vijaysinh Lendave","publish_date":"2021-11-15T15:00:00","publication_year":"2021","word_count":1704,"keywords":["data science","machine learning","AI","credit risk machine learning","neural network","ML","Machine Learning","Python","Colab","Aim","XGBoost","analytics","Data Science","Pandas"],"extracted_tech_keywords":["AI","machine learning","ML","neural network","data science","analytics","Aim","XGBoost","Colab","Pandas"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/why-tree-based-models-are-preferred-in-credit-risk-modeling\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":36363,"title":"10 Best Software Development Methodologies For Embracing Software Engineering Projects","content":"In the current scenario, technology is evolving at a fast pace and is pushing the software companies to move and work in a fast-paced motion. The software development methodologies are meant to be made sure that an application is tested fully that includes testing, integrating, etc. as a whole of a particular system.  In this article, we discuss the 10 best software development methodologies that help in constructing a flawless software application. 1| Agile Software Development Agile software development is a conceptual framework for embarking software engineering projects. The main goal is to minimise the risk by developing software in short iterations. It is basically an umbrella that describes various agile methodologies and is specially designed to fulfil the needs of the fast-changing environment by incremental development and develop the actual final product. The basic principles behind agile software development are given below: To satisfy the customers through early and continuous delivery of valuable software.Delivering software in a frequent manner with a preference of shortest timescale.Building projects and providing support as well as the environment to the motivated individuals.It welcomes the changing environments with quick adaptation.Enhances face-to-face means of communication. Advantages The transparency is high due to the direct communication.Adaptation to any change in the environment is quick. Disadvantages Difficult to assess the effort of large software projects at the initial stages of the software development life-cycle.It produces less documentation because of the more code-focused development. 2| Black box testing This software testing method is also known as Behavioural Testing where the internal implementation of an item that is being tested is unknown to the tester. The method usually finds the error for incorrect or missing functions, errors in data structures, performance errors, initialisation or termination errors, etc. The main goal of this software development methodology is on the functionality of the system as a whole. The testing can be of two types, functional and non-functional. Advantages Effective on large and complex applications.Defects can be identified at an early stage of testing. Disadvantages Impossible to identify all the inputs.Test cases are merely difficult to design without clear specifications. 3| Prototyping Model This model can be called one of the most popular SDLC (Software Development Life Cycle) models where a prototype is built, tested as well as reworked if necessary. This model works best when there are known project requirements in detail ahead of time. There are basically two approaches for this model, rapid throw-away prototyping and evolutionary prototyping. The rapid throw-away prototyping provides a method of exploring ideas and receiving customer feedbacks. The later approach develops initially and is incrementally refined on the basis of customer feedback until it gets accepted. Advantages Refinement leads to the accommodation of new requirements.Flexible design. Disadvantages Costly as well as time-consuming.Uncertainty in determining the number of iterations that would be required before the prototype is finally accepted by the customer. 4| Rapid Application Development Model This software development methodology uses minimal planning in favor of rapid prototyping. It is a type of incremental model and here the components are developed in parallel. There are various phases in this model and they are business modelling, data modelling, process modelling, application generation as well as testing and turnover. Advantages The occurrence of initial reviews is quick.Customer feedbacks are encouraged. Disadvantages Highly dependent on modelling skills.Inapplicable to cheaper projects. 5| Spiral Model This model is a combination of iterative development with a systematic, controlled aspect of the waterfall model that allows the refinement through each iteration around the spiral. The diagrammatic representation of this model is a spiral with a number of loops where the loops can vary on the projects. Each loop is known as the phase of the software development method. This is an extended model of Waterfall with rapid prototyping. This model focuses on early identification and reduction of project risks. Advantages Efficient for large and complex projects.An excellent development model due to risk analysis and risk handling at every phase. Disadvantages Quite expensive and unsuitable for cheaper projects.Estimation of time is difficult. 6| White Box Testing This is also known as glass box testing or clear box testing and is opposite to the black box testing technique. It requires the tester to have adequate knowledge of programming and the tester chooses inputs to work on the paths through coding and determines the outputs.It involves the testing of software codes for internal security holes, poorly structured paths, expected outputs, functionality of conditional loops, etc. Advantages This method is easily automated.Helps in optimizing the code. Disadvantages The cost is high and is a complex method.Time-consuming for large applications. 7| Extreme Programming Methodology This model, also known as XP methodology can be called as one of the most important software development framework of Agile Models which is mainly used o improve the quality of the software. The involvement of the customer during the process of software development is quite high. Advantages Enables greater tractability within the modelling procedure.Focus mainly on customer involvement and provides quality software. Disadvantages Possibilities and future outcomes are unknown.Effectiveness is directly proportional to the involvement of people. 8| Joint Application Development Methodology This methodology involves the end user in the design and development of an application through JAD sessions (succession of collaborative sessions). It is a requirements-classification and user-interface expansion approach to accentuate and confirm the software system. The emphasis is basically laid on the business difficulty rather than methodical details. Advantages A large amount of valuable information within a shorter period of time.Resolve the differences immediately with suitable assistance. Disadvantages Time-consuming in planning and scheduling.A necessity for highly trained experts which is hard to find. 9| Scrum Development Methodology It is an incremental and iterative agile software development framework for managing the development of products. The development strategy is flexible because in this framework a development team works as a unit to achieve a particular goal. This method follows an evidence-based empirical approach where it focuses on maximizing the team’s ability for quick delivery. Advantages The working team is responsible for decision-making.It is a lightly controlled method with constant updating. Disadvantages This method is not suitable for complex and large projects.Requires a highly expert team. 10| Dynamic System Development Model Methodology This is an agile software development methodology that is iterative and incremental and is largely based on the Rapid Application Development (RAD ) methodology. The method provides a four-phase framework which consists of implementation, design and build iteration, prototype iteration, feasibility, and business study. Advantages Easy access to the end users.Quick functionality deliverables. Disadvantages Implementation is cost effective.Tough for the smaller organisation.","excerpt":"In the current scenario, technology is evolving at a fast pace and is pushing the software companies to move and work in a fast-paced motion. The software development methodologies are meant to be made sure that an application is tested fully that includes testing, integrating, etc. as a whole of a particular system.  In this […]","categories":["AI Trends"],"tags":["Software Development"],"author_name":"Ambika Choudhury","publish_date":"2019-03-17T04:42:55","publication_year":"2019","word_count":1092,"keywords":["Go","API","TPU","programming_languages:R","AI","programming_languages:Go","RAG","GAN","Software Development","R"],"extracted_tech_keywords":["AI","RAG","TPU","R","Go","API","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/10-best-software-development-methodologies-for-embracing-software-engineering-projects\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":39628,"title":"A Day In The Life Of: Data Scientist Who Believes Beginners Should Code Algorithms From Scratch Instead Of Using APIs","content":"Data Science is the next big transformational wave hitting the IT sector. And unlike many other sectors, it is a field dominated by a steep learning curve where you may or may not need a formal education or advanced degree, depending on your educational background. Now, data science that evolved from advanced  statistical analysis existed as a function for more than a decade, yet the lack of formal education environment has prompted many aspiring Data Scientists and Analysts to chalk a new learning path by brushing up on the programming, statistics or Data Structures to develop a thorough fundamental understanding and deliver outcome-oriented results on the back of data. Given this backdrop, every week, we reach out to the community and speak to Data Scientists\/Data Analysts and dig into their learning journey and what they like most about this buzzing field. This time, we caught up with Sudharsan Ravichandiran, Data Scientist at Param.ai, to understand how he was drawn to this field, his big leap, what he likes most about the job and how he strikes a work-life balance. Interestingly, Ravichandiran, author of Hands-on Reinforcement Learning With Python and Hands-on Meta Learning with Python  started off as a freelance web designer during his undergraduate studies and even designed some award-winning websites. “I had this paper called Artificial intelligence on one of my spring semesters which drew my interest and got me fascinating. That’s where I started to explore more about AI. Being a huge Math enthusiast, after discovering substantial math behind ML algorithms, I decided this is what I should be on working on,” he said. At this thriving startup, everyday comes in with new challenges which keeps him excited and motivated to make him want to go to work every day. Life at Param.ai Talking about his day-to-day role at Param.ai, an AI-powered intelligent recruiting platform, Ravichandiran shares that the platform maps potential candidates to corresponding jobs and ranks them in order of relevance. “Being a Data Scientist at Param.ai is exciting as every day brings exciting challenges and opportunities to grow,” he said. Everyday is not the same for Ravichandiran. Some days, he is hard at work building actionable latent insights derived from huge data points while some days he’s involved in working on building models. However, most of the time is spent on getting, cleaning, standardising and understanding the data. “If you understood the data built then all other steps come very handily. On the other hand, but if you don’t understand about the dataset which you have then there is no use of applying algorithms,” he shares. Ravichandiran’s recent project included building a powerful recommendation engine that continually learns the behaviour of recruiter to understand their hiring traits which include numerous parameters such as candidates the recruiter is shortlisting and hiring for the respective job role. The recommendation system is robust enough and since it learns continually, if the behaviour\/activity of the recruiter changes, the model also updates itself and learn from there. Striking a work-life balance Work-life balance at this startup is simply amazing and he reveals he gets constant support and guidance from his amazing managers Hari and Ashish. Meanwhile, outside the job, he loves working on creative stuff.  Ravichandiran is good at songwriting and sketching and also a big fan of pop music. “If you don’t find me playing with data then you can probably find me either playing songs on my dashing guitar or listening to pop music,” he shared. His advice to beginners His suggestion to someone getting started in data science is to ensure they understand the fundamentals. Before jumping directly into modern algorithms, he suggests them to build a strong foundation on basic and also Math including linear algebra, probability statistics, and calculus. “I also recommend beginners to code algorithms from scratch instead of using high-level APIs. Coding algorithms directly from scratch will undoubtedly strengthen their understanding of the algorithms and take their skills to the next level,” he shared. According to Ravichandiran, the spurt in interest can be cited to an exponential growth in AI over the last two years. “This is mainly due to the modern computational advancements and the availability of humongous volume of data. Today, many researchers and scientists around the world are working towards Artificial General intelligence where machines can perform multiple intellectual tasks just like we humans do,” he said, in closing.","excerpt":"Data Science is the next big transformational wave hitting the IT sector. And unlike many other sectors, it is a field dominated by a steep learning curve where you may or may not need a formal education or advanced degree, depending on your educational background. Now, data science that evolved from advanced  statistical analysis existed […]","categories":["AI Features"],"tags":["AI for Beginners","big data and analytics everyday life","data scientist career path","Data Scientists","Interviews and Discussions","what is artificial intelligence for beginners"],"author_name":"Richa Bhatia","publish_date":"2019-05-24T05:50:49","publication_year":"2019","word_count":729,"keywords":["what is artificial intelligence for beginners","data science","data scientist career path","artificial intelligence","Go","API","AI","startup","ML","Interviews and Discussions","Python","ViT","AI for Beginners","big data and analytics everyday life","R","Data Scientists"],"extracted_tech_keywords":["AI","artificial intelligence","ML","data science","Python","R","Go","API","ViT","startup"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/a-day-in-the-life-of-data-scientist-who-believes-beginners-should-code-algorithms-from-scratch-instead-of-using-apis\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10047037,"title":"Meet The Two Indian Origin Engineers At Tesla AI Day","content":"Last week, the global tech industry witnessed the year’s one of the most anticipated events — the Tesla AI Day. Ever since Musk first revealed sparse information about such an event, experts and novices started predicting what could happen at the event. Some of the major announcements made at this event include the announcement of Tesla humanoid robot and AI training chip D1 for Tesla’s supercomputer Dojo. Musk likes to flaunt his engineers at his events and this time is no different. At the latest edition of Tesla AI day, the audience couldn’t help but notice the handful of engineers that Musk calls the Tesla AI team. This team also featured two Indian-origin engineers. Ganesh Venkatramanan Credit: YouTube Ganesh Venkataraman, a Senior Director Hardware with Tesla, emerged as the most significant face at the Tesla AI Day event. Venkataramanan, who is also Project Dojo’s lead, made a significant announcement about the Dojo supercomputer and introduced the Dojo D1 chip. A few years ago, Musk had entrusted the Venkataramanan-led team to bring to life a super-fast training computer that would be able to achieve the best AI training performance, enabling larger and more complex neural networks models with better power and cost-efficiency. “For Dojo, we envisioned a large compute plane filled with very robust compute elements, packed with a large pool of memory, and interconnected with very high-bandwidth and low-latency fabric. We wanted to attack this all the way – top to bottom of the stack – and remove all the bottlenecks at any of these levels,” Venkataramanan said at the event. Venkataramanan, who has been working with Tesla for five and a half years, seems to be a good choice for leading one of Musk’s dream projects, given his impressive professional repertoire. He joined Tesla in 2016 as the director of Autopilot Hardware, a role which he served for two years. After that, he was promoted to the senior director of Autopilot Hardware, as part of which he leads Project Dojo and handles responsibilities of Silicon, Systems and Firmware\/Software. Before Tesla, Venkataramanan was with AMD’s CPU group. He was working with the company for over 14 years. He joined as a senior design engineer and worked his way up to the post of Senior director design engineering, handling a team of over 200 highly skilled engineers. He has been a part of the team that developed several generations of x86 and ARM cores. Some of his major contributions are the development of the industry’s first x86-64 chip, first Dual-Core x86, and Zen core. Before that, Venkataramanan has also worked with Analog Devices and Hexaware Technologies. He received a Bachelor’s degree in Electronics Engineering from the University of Mumbai. He went on to pursue postgraduate studies in integrated electronics at IIT Delhi. Venkataramanan has over 20 patents to his name. Some of them are: Hybrid tag scheduler, dependent instruction suppression, Load balancing when assigning operations in a processor, selectable multi-way comparator, and multi-issue unified integer scheduler, among others. Ashok Elluswamy Credit: YouTube At the Tesla AI Day 2021 event, the other prominent Indian-origin personalities were Tesla’s Director (Autopilot Software) Ashok Elluswamy. Elluswamy gave the introduction to the artificial intelligence systems behind the Autopilot driving system on Tesla cars. He also gave a sneak-peek into Tesla’s Cybertruck via a photorealistic rendering. Production of this Cybertruck will not start until next year. In his current role, Elluswamy’s work revolves around creating large scale automatic ground truth pipelines to train neural networks; building robust, causal predictive models, and developing an accurate; detailed geometric understanding of the world with AI and engineered models, among others. Elluswamy joined Tesla in 2014 as a software engineer in the Autopilot department and has soon risen in the ranks to become the director of the team now. In fact, his name came into public domain first in 2019 when Musk undertook an overhaul of the Autopilot group. As per reports, Musk was unhappy with the progress made by the group, leading to dropping a few prominent names and elevating a few others. Elluswamy belonged to the latter group and was entrusted with leading the perception and computer vision teams after the shakeup. Before joining Tesla, he has worked with WABCO Vehicle Control System and has interned with Volkswagon Electronic Research Lab. He holds a bachelor’s degree in Electronics and Communication Engineering from the College of Engineering Guindy, Chennai and a Master’s degree in Robotics System Development from Carnegie Mellon University. As per a LinkedIn endorsement from Elluswamy’s professor at CMU, John Dalon, he was one of the top students and showed initiative in learning a wide variety of topics.","excerpt":"One of the most awaited events prominently featured two Indian-origin engineers. Who were they?","categories":["AI Features"],"tags":[],"author_name":"Shraddha Goled","publish_date":"2021-08-26T11:00:00","publication_year":"2021","word_count":770,"keywords":["Go","artificial intelligence","AI","neural network","computer vision","BERT","llm_models:BERT","Rust","GAN","R"],"extracted_tech_keywords":["AI","artificial intelligence","neural network","computer vision","R","Go","Rust","BERT","GAN","llm_models:BERT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/who-were-the-two-indian-origin-engineers-at-tesla-ai-day\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":11011,"title":"Data Analytics for CFOs: Why new-age Business Intelligence systems are better than traditional MIS","content":"The role of today’s CFO extends way beyond the mandate of traditional financial management into more strategic areas of business. The new responsibilities make it crucial for CFOs to have meaningful data at their fingertips to support and develop strategic business initiatives. However, the problem is with getting the right data, in the right format and at the right time. Traditional MIS systems especially those used for revenue and expense reporting, are saddled with inaccuracy, inconsistency, and unmanageable complexity. The way the data is collected, stored, compiled and presented to CFOs makes it unfit for analysis and decision making. For department heads and finance teams, data management is an arduous and time-consuming process. Why is this the case? 1. The data extracted from the traditional MIS systems is aggregate data. In any organisation, department or product heads are given budgets at the start of every month or quarter and their performance is gauged against this budget. To view their individual performance against budgets, department heads resort to the canned reports from the General Ledger (GL) generated by the finance teams. They view the report data, look for variances, and take corrective measures to eliminate cost overruns and maximise budget efficiency. Seems pretty straightforward till this point, doesn’t it? Well, it’s not, especially when there are deviations in data. Aggregate data is not enough to understand the root cause of budget deviations. The GL reports show aggregate data for spends vs. budget. This data output has none of the rich transactional details that are necessary for investigating the root causes that trigger a budget deviation or an unexpected increase in a particular expense category. In this case, the departmental managers go back to the finance teams for the underlying details behind the GL data — the individual transactions that caused them to exceed their budget. This triggers a spurt of activity within the finance department causing workload to shoot up immediately during the month or quarter end. The finance departments run additional reports to find out the required information for managers. The process doesn’t stop there. The departmental heads often have questions on the new reports generated. Once they receive the answers, they have follow-up questions and the cycle continues. It sometimes takes weeks for the issue to get resolved satisfactorily from both ends. 2. The data is stored in multiple systems managed by different teams. The GL summary report contains transactional data that is extracted from different systems like HRM & Payroll, Accounts Payable and Inventory Management – each of them managed by different teams. People other than finance also gets tasked with requests for data extracts. This makes the entire process even more complex and time-consuming. No wonder it takes weeks on end to solve one issue. 3. The data does not facilitate easy comparison of different departments or cost centres. The reports generated by the finance department for comparison are income statements and balance sheets. They show profits,\/losses and performance vs. budget for an individual manager, department or product and are definitely useful. But imagine a Director who has 15 Managers reporting to him – he must view 15 individual reports. And to compare the performance of all managers together, he and his team have to do a fair amount of spreadsheet work. Teams have to manually copy, paste and transpose rows into columns to compile the data together. Ideally, the GL should be able to generate these reports without putting the directors through this manual and error-prone process. So what is the way out for CFOs and finance teams? Alternative #1: Replace their legacy GL system One way is to replace their current GL system with the latest ERP or accounting software. However, this might involve substantial cost and resource commitment. Alternative #2: Adopt a self-service, business intelligence system A smarter way out is to adopt self-service, business intelligence (BI) systems that give CFOs the analytics they need while leaving their existing GL in place. Let us look at how automated BI tools provide their users an edge over traditional MIS: Consistent, secure and error free reports: Rather than generate aggregate reports, BI solutions help finance teams generate role-specific reports. Such reports enable managers to easily view revenues and spends of their direct reportees or their area of responsibility. Finance teams can enable managers to view these reports via web-based dashboards rather than sending the reports via email in the form of spreadsheets. Emailing reports can expose the company’s financial data to unauthorised individuals. Spreadsheets can be easily altered and the resulting conflicting data takes weeks to investigate and reconcile. BI systems help managers view the reports they need in the most secure environment. BI systems also help finance teams automate the data reporting process. Finance representatives can set rules and alerts within the system to notify managers of budget overruns and help them take action immediately. All data on one system: BI systems easily integrate with other systems like accounts payables, inventory management and payroll. Such integrations help finance teams to map GL codes with transactional data, invoices, purchase orders, etc. What was initially stored in a maze of spreadsheets can not be bought onto the same platform and viewed together with a click of a button. Easy comparison: BI tools make comparisons across departments, across different data types and across different users easier to access as well as to visualize. In other words, no more headaches and switching between spreadsheets, while reviewing employee or department performance. Apart from addressing the key gaps of traditional MIS, BI systems help finance teams gain a lot more. Centralised visibility and accountability: Since Departmental managers can easily identify budget variances and the underlying reasons behind them, they can manage expenses in a better way. They can spend more time understanding what triggers budget overruns rather than running behind finance teams asking for details behind the aggregate summaries. They also have centralised visibility over all data and can take decisions in a prompt and efficient manner. More efficient work allocation: A significant amount of time and effort of finance teams goes into back tracking bugs, errors and outliers in the data and reporting them back to the department who pointed them out. BI tools eliminate these low-value, cumbersome tasks allowing finance to focus on core job responsibilities such as identifying new revenue generating opportunities or mitigating risks. Powerful feedback tool: BI systems can serve as a good coaching and feedback tools for management since they facilitate easy peer-to-peer comparisons. With access to the comparison data, VPs and directors can evaluate performance of different managers, understand why certain managers are able to meet budgets and help the ones who don’t create more realistic budgets and decrease their budget overruns. As big data and analytics become a priority, CFOs have to be prepared to take strides forward and bag the best business intelligence systems for their companies.","excerpt":"The role of today’s CFO extends way beyond the mandate of traditional financial management into more strategic areas of business. The new responsibilities make it crucial for CFOs to have meaningful data at their fingertips to support and develop strategic business initiatives. However, the problem is with getting the right data, in the right format […]","categories":["IT Services"],"tags":["Business Intelligence"],"author_name":"Anshul Rai","publish_date":"2016-10-21T09:55:17","publication_year":"2016","word_count":1145,"keywords":["big data","Go","business intelligence","TPU","programming_languages:R","AI","ViT","analytics","GAN","Business Intelligence","R"],"extracted_tech_keywords":["AI","analytics","TPU","R","Go","big data","GAN","ViT","business intelligence","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/it-services\/data-analytics-cfos-new-age-business-intelligence-systems-better-traditional-mis\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10005095,"title":"Top 8 Machine Learning Libraries In Go Language","content":"Created by Google researchers, Go is a popular open-source programming language. The language includes many intuitive features, including a garbage collector, cross-platform, efficient concurrency, among others. According to the Stack Overflow Developer Survey 2020, Go language is not only the fifth most loved programming language but also fetches the programmers the third-highest salary among other languages. Below here, we list down the top machine learning libraries in Go language. (The libraries are listed according to the number of Stars in GitHub). 1| GoLearn Stars: 7.3k About: GoLearn is one of the most popular libraries in the Go language. It is also known as the ‘batteries included’ machine learning library for Go. It aims to contribute simplicity paired with customizability. Some of its features are- The library includes helper functions for data like cross-validation, train splitting as well as test splitting. It is similar to the popular Scikit-learn library in Python as it implements the Scikit-learn interface of Fit\/Predict. This library comes with practical examples. Know more here. 2| Gorgonia Stars: 3.5k About: Gorgonia helps in facilitating machine learning entirely in Go. The primary goal of this library is to be a highly performant machine learning as well as graph computation-based library that can scale across multiple machines. It also contributes a platform for the exploration of non-standard deep-learning as well as neural network-related research. It can perform processes like neo-Hebbian learning, corner-cutting algorithms, among others. Some of its features are- Gorgonia can perform automatic differentiation, symbolic differentiation, gradient descent optimisations and numerical stabilisation. The library provides many convenience functions to help create neural networks.It supports CUDA and GPGPU computation. Know more here. 3| goml Stars: 1.1k About: goml is a machine learning library written entirely in Go language which allows developers to include machine learning into their applications. The library consists of various models which help in learning in an online as well as the reactive manner by passing data to streams held on channels. Some of the features of this library include- Comprehensive testsExtensive documentationModular source codeClean and expressive. Know more here. 4| eaopt Stars: 693 About: eaopt is an evolutionary optimisation library written in Go language. It allows one to write most of the evolutionary optimisation algorithms as special cases of a genetic algorithm. Some of its features include- In this library, various evolutionary algorithms are available with a consistent APIOne can practically do anything by using the GA struct.Speciation and migration procedures are available.Common genetic operators, including mutation, crossover, migration, among others, are already implemented.It allows the evaluation of costlier functions to be done in a parallel manner. Know more here. 5| Evo Stars: 104 About: Evo is a general framework for implementing evolutionary algorithms in Go. The framework exposes a clean and flexible API oriented around two interfaces —  genome and population. Genomes represent both the function being optimised and the representation of solutions, while populations represent the architecture under which genomes are evolved.  Some of its features are- Clean and flexible API oriented.The library can be used for writing modular web applications in Golang aimed at both backend and frontend.It is modular and extensible.It ensures high performance. Know more here. 6| goga Stars: 89 About: It is a genetic algorithm solution written in Golang. It is used and configured by injecting different behaviours into the main genetic algorithm object. It consists of three components — simulator, selector and mater. The simulator provides a function that accepts a single genome and assigns a fitness score to it. The selector object takes a population of genomes and the total fitness and returns a genome from the population that it has chosen. In contrast, the mater accepts two genomes from the selector and combines them to produce two others. Know more here. 7| gogl Stars: 60 About: gogl is a graph library written in Go language. This library aims to contribute simple, unifying interfaces, including implementations of graph algorithms and data structures that can scale from small graphs to huge graphs. The features of this library include- Simplicity: gogl has fully and correctly modelling graph theoretic concepts Performance: Gogl has faster design constraints and allows the best-known algorithms.Functional: The library orient towards transforms, functors, and streams; achieve other styles through layering.Other features include extensibility, correctness, among others. Know more here. 8| GoMind Stars: 7 About: GoMind is a Neural Network (NN) library that is written entirely in Go. Currently, the library only supports a single hidden layer, and it will support multi-layer soon. In GoMind, the network learns from a training set using the back-propagation algorithm. Some of its features include- GoMind supports activation functions like Sigmoid, ReLU, Leaky ReLU, among others.The library estimates the ideal number of hidden layer neurons for a given input and output sizes if a count is not given during model configuration.GoMind uses Mean Squared Error function to calculate error while back-propagating. Know more here.","excerpt":"Created by Google researchers, Go is a popular open-source programming language. The language includes many intuitive features, including a garbage collector, cross-platform, efficient concurrency, among others. According to the Stack Overflow Developer Survey 2020, Go language is not only the fifth most loved programming language but also fetches the programmers the third-highest salary among other […]","categories":["AI Trends"],"tags":["deep learning application examples","go language","Gradient Descent Numerical Example","how to calculate mean square error","Machine Learning","machine learning libraries","machine learning optimization","ml libraries","scikit learn"],"author_name":"Ambika Choudhury","publish_date":"2020-08-19T16:00:00","publication_year":"2020","word_count":812,"keywords":["scikit-learn","TPU","ml libraries","machine learning libraries","scikit learn","R","CUDA","how to calculate mean square error","go language","machine learning","AI","Gradient Descent Numerical Example","neural network","ML","Machine Learning","Python","Aim","machine learning optimization","deep learning application examples"],"extracted_tech_keywords":["AI","machine learning","ML","neural network","Aim","scikit-learn","TPU","CUDA","Python","R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-8-machine-learning-libraries-in-go-language\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10094065,"title":"Is Indian Govt’s Battle Against AI Disinformation Flawed?","content":"Disinformation is not something new to humans as the very first instances of disinformation can be traced back to ancient times. During the Roman Empire, there were several instances of disinformation being used to sway public opinion. The Romans have been gone for centuries, but propaganda still remains a relevant tool. Political parties today still rely on the same strategy to sway public opinion. In today’s age, most of the disinformation we have seen is in the form of a text or a misleading image or video. But today, Generative AI is giving disinformation a whole new dimension. Sam Altman, CEO of OpenAI, while speaking to a packed auditorium of close to 1,000 in London, said, “Humans are already good at making disinformation, and maybe the GPT models make it easier. But that’s not the thing I’m afraid of. “I think one thing that will be different with AI is the interactive, personalised, persuasive ability of these systems.” Text-to-image AI models like Stable Diffusion, Midjourney and DALL-E2 can generate hyper realistic images that can easily be mistaken for genuine ones. This technology has opened up possibilities for generating deceptive visual content, further blurring the line between reality and falsehood. ( AI generated image of the Pope) AI Generated disinformation is one the rise On Sunday, Delhi Police apprehended Indian wrestlers Vinesh Phogat, Sakshi Malik, Sangeeta Phogat, and Bajrang Punia at Jantar Mantar. Unfortunately, on Twitter, a manipulated image portraying them with smiles after their arrest began circulating. This photo was shared multiple times with the intention of demeaning the wrestlers and undermining the purpose behind their protest. https:\/\/twitter.com\/BajrangPunia\/status\/1662820142454312968 The wrestlers have been protesting for weeks against Wrestling Federation of India (WFI) president and Kaiserganj BJP MP Brij Bhushan Sharan Singh, who has been accused of sexual harassment. Similarly, last week, a fabricated image created by AI depicting an explosion near the Pentagon in Washington DC circulated on social media, resulting in unintended consequences for the US stock market. Reputable Indian media outlets such as Republic and News 18 unknowingly played a part in spreading this image, inadvertently contributing to the dissemination of misinformation. Earlier this year, images of former US president Donald Trump getting arrested went viral on social media, however, later, those images were found to be AI-generated as well. https:\/\/twitter.com\/Rap\/status\/1638208340214915072 While a careful examination of the Donald Trump images could reveal that they are fake, it is important to recognise that AI-generated images are continuously improving. A striking illustration of this progress is seen in the photos depicting Pope Francis wearing a Balenciaga puffer jacket. These images are incredibly lifelike, to the point where distinguishing them as AI-generated becomes exceedingly challenging unless explicitly informed about their origin. In the past, text-to-image models such as Midjourney faced difficulties when generating human hands. With a new update, Midjourney has fixed the problem, however, at the same time, have added to the problem of disinformation. Fighting Disinformation One of the primary reasons for growing demand for AI regulation is disinformation. Even though the Narendra Modi-led administration has said that they won’t regulate AI, they have not completely turned a blind eye to curbing AI generated disinformation. The Indian government has confirmed that the upcoming Digital India Act (DIA) will have provisions to deal with AI generated disinformation. “We are not going to regulate AI but we will create guardrails. There will be no separate legislation but a part of DIA will address threats related to high-risk AI,” Union Minister Rajeev Chandrasekhar said. While the first draft is expected to be released in June, so far we know that the bill proposes to regulate online media platforms and social media intermediaries to prevent the spread of fake news and disinformation. If such legislation were in effect, Republic and News18 could face legal responsibility for disseminating false information regarding the fabricated Pentagon image. Likewise, Twitter could also be held accountable for the initial post that sparked the spread. This development is positive as it promotes accountability and encourages journalists to conduct diligent fact-checking, thereby reducing the prevalence of inaccurate reporting. Further, the government is also mulling about the creation of a new body that specifically deals with the issue of fake news or disinformation. Elections around the corner The Lok Sabha Elections are scheduled for 2024 and hence it makes it utmost necessary to have regulations in place to deal with AI generated disinformation. Elections are also scheduled to take place next year in the USA and UK, and experts have already warned that AI-generated disinformation could plague the upcoming elections in the respective countries. With Generative AI coming into the pictures, AI generated disinformation could significantly rise as we approach the election date. We have already seen deep fake videos being used for political campaigns by the BJP. Besides deep fakes and AI generated hyper-realistic images, AI could be used to create social media bots programmed to automatically generate and disseminate false information. Similarly, AI algorithms can also be used to analyse user data and preferences to deliver tailored disinformation campaigns. While the dangers are aplenty, guardrails are non-existent so far. Interestingly, many of the Twitter handles sharing the AI generated of the wrestlers in a police van were circulated on Twitter by members affiliated to the ruling party, it was alleged. A study by BBC in 2018 found that those affiliated with BJP are more likely to share fake news or disinformation compared to others. Earlier this year, BJP vice president Baijayant Jay Panda said that those in politics should start using ChatGPT in an era of rapidly unfolding technologies. While the government so far has shown intent to fight disinformation, the administration cannot be biassed. In 2021, when Twitter classified a tweet, featuring screenshots of an alleged Congress ‘toolkit’ shared by BJP chief spokesperson Sambit Patra, as “manipulated media”, the Union Government objected to it, and soon Twitter’s office in Delhi was raided by the Police. As we anticipate the public release of the DIA draft in June, the question remains whether it will provide the necessary measures to combat the escalating threat posed by AI-generated disinformation. Only time will tell if it possesses the strength and effectiveness required to tackle this pressing issue.","excerpt":"AI models like Stable Diffusion, Midjourney and DALL-E2 can generate hyper realistic images that can easily be mistaken for genuine ones","categories":["AI Features"],"tags":[],"author_name":"Pritam Bordoloi","publish_date":"2023-05-29T18:30:00","publication_year":"2023","word_count":1034,"keywords":["Go","ChatGPT","OpenAI","AI","Scala","Git","RAG","Ray","generative AI","R"],"extracted_tech_keywords":["AI","generative AI","ChatGPT","OpenAI","Ray","RAG","R","Go","Scala","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/is-indian-govts-battle-against-ai-disinformation-flawed\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":34976,"title":"Candidates Recruited Through Hackathons Have A Lower Chance Of Being Wrong Hires: TheMathCompany","content":"In modern workplaces, especially the sector which relies hugely on big data and analytics, hackathons are gaining popularity among organisations. Apart from being used to get a fresh perspective for solving critical problems, companies are also using them to shortlist candidates based on actual performance in a simulated environment. This week, Analytics India Magazine spoke to Srinidhi Rao of TheMathCompany to understand how employers want to look at the manner in which these potential candidates are tackling complicated problems either on their own or as a part of a larger group during hackathons. Analytics India Magazine: How have hackathons become a crucial part of the hiring process in the data science industry? Do you think spotting talent through hackathons will make this process easier? Srinidhi Rao: Hackathons are one of the tried-and-tested solutions to bridge the gap between theoretical and practical knowledge. As opposed to traditional methods of candidate evaluation, hackathons simulate a real-world environment for candidates with components like solving real business problems to working in a deadline-based environment. Candidates hired through hackathon have a lower chance of being wrong hires with respect to technical capabilities, considering they would have succeeded in real life problem-solving. AIM: What is your opinion about online hackathon? Have hackathons replaced the traditional way of hiring, say by campus hiring? SR: Online hackathons are being used more frequently off-late and are yielding better results than traditional means of hiring. The talent pool available becomes much bigger if the entire process of recruitment is moved online as it reduces the impact of the location constraint. Having said that, we feel that right now, augmenting the existing processes with online hackathons is the best way to go, rather than completely replacing it. AIM: Hackathons solve two purposes — people strategies and the other being solving use cases in an innovative way. Do you agree? SR: We agree – hackathons simulate a real work environment to test the salient points of an employee’s behaviour (like working within a deadline, problem-solving mindset), thus helping organisations align the prospective employee’s candidature with their people strategy and their over-arching objective. Hackathons also provide with an excellent opportunity for organizations to crowdsource innovative ideas for some of the challenging new age problems, thus solving problems innovatively. AIM: What is the main purpose of conducting a hackathon for you? SR: The main purpose of conducting a hackathon for us are: Evaluating the problem-solving capability of a candidate using real-world problems in a time-bound situation, thus filtering out candidates based on their technical capabilities Create brand visibility amongst target hiring pool o Branding opportunity to attract the right talent AIM: Can you cite a recent example where you hosted a hackathon? What was its purpose? We recently conducted The Triathlon for the graduating year engineering students across the country. The challenge was taken up by 2,200 students from across 6 campuses. The purpose was to identify the best talent from each college. The Triathlon had questions around: Reasoning Decoding Solving a real-world business case study Top 3 students from each campus were given a direct entry to the final interview round of TheMathCompany. AIM: How important are hackathons (both internal and external) to boost innovation? Have you and your organisation benefitted from conducting hackathons? SR: Multiple organizations host periodic hackathons to engage existing employees with a variety of problems and to boost cross-learning across different sets of existing employees. Organizations also prefer hiring an internal candidate for an open position, rather than scout on the market–hackathons are an important way to define the best-fit in those cases. We, as an organization, have benefitted from both, internal (boosting innovation & cross learning) as well as external (finding the right candidate) hackathons. AIM: Hackathons have become the benchmark for assessing baseline skills? Do you agree, if so what skills? How reliable it is to test skills in the long run? SR: Hackathons have the potential of becoming a standard way to assess baseline skills like programming and industry knowledge. Hackathons possess an ability to scale in terms of usage in the long run. Collecting data in parallel about the effectiveness of hackathons in creating a narrower hiring funnel will be a major component of this whole growth journey. AIM: How is hackathon turning into a mainstream hiring requirement? SR: Hackathons are gaining popularity among organizations to shortlist candidates based on actual performance in a simulated environment. Employers want to look at how candidates tackle complicated problems either on their own or as a part of a larger group (which shows a willingness to work in teams). AIM: What is your preference for candidates who have participated in hackathon vs direct hiring? SR: During campus hiring, as the hiring pool is large, hackathons help us filter out the right candidates, who are given a direct entry into the final round, where their behavioural fit with our organization is assessed. In the case of lateral hires, we do not give any special preference to candidates coming through hackathons as there are other aspects that need to be taken into consideration in a candidate’s profile. AIM: Are hackathons replacing the need to have short term certification course in areas such as AI, analytics? SR: We believe that hackathons and short-term certification courses go hand-in-hand. The pool of potential candidates who are moving to the field of data science from other fields is huge. Certification courses allow these candidates to learn various data science concepts in a structured manner. The topics covered in some of these courses are exhaustive and provide a brilliant opportunity for candidates to get started on real-world data science problems. Hackathons assist them in implementing these concepts on-to real-life problems. AIM: In future, how will hackathons evolve (smaller, online hackathons or the preference will be for offline events)? SR: We are going to see a combination of online and offline events. Online events will mostly serve as the first round to shortlist the right candidates. Advanced rounds are going to be offline events which will build a better connect between candidates and employers. AIM: Who are your partners with whom you usually prefer conducting hackathons? SR: We don’t have any preferred partners for conducting hackathons. We have used the available platforms and created our own custom design of the entire process.","excerpt":"In modern workplaces, especially the sector which relies hugely on big data and analytics, hackathons are gaining popularity among organisations. Apart from being used to get a fresh perspective for solving critical problems, companies are also using them to shortlist candidates based on actual performance in a simulated environment. This week, Analytics India Magazine spoke […]","categories":["AI Features"],"tags":["Big Data","Hackathon","Hiring","Interviews and Discussions"],"author_name":"Srishti Deoras","publish_date":"2019-02-15T05:31:44","publication_year":"2019","word_count":1046,"keywords":["big data","data science","Go","programming_languages:R","AI","innovation","Hiring","Hackathon","Aim","analytics","GAN","Big Data","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","data science","analytics","Aim","R","Go","big data","GAN","innovation","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/candidates-recruited-through-hackathons-have-a-lower-chance-of-being-wrong-hires-themathcompany\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10099005,"title":"GenAI to Scale the Peak in 2 Years","content":"With signs that there’s plenty of cash to go around for generative AI startups, Gartner placed genAI on the top of the ‘Peak of Inflated Expectations’ for the first time. If you think this technology is overhyped right now, wait for another 2-5 years, as per the report. Startlingly, ChatGPT, the most well known example of genAI, already crossed the threshold of a product life cycle within the first three months of its release as per previous reports. The rapid adoption rate can be attributed to the media spotlight that further amplified the service expectations. Apart from being flooded with investments, the technology is also being looked up as a driving force to revamp the industrial landscape by tech giants. Leading software corporation IBM is one of them as it paused hiring for 7,800 jobs due to AI. The company’s CEO Arvind Krishna believes AI’s ability to increase worker productivity is the solution to the talent crunch problem faced by many. Like others, the company also jumped headfirst into genAI with its in-house Watson assistant. Another factor that will help the technology reach its peak is the sudden outburst of open source models. From Meta’s Llama 2 to Databricks’ Dolly, iterations of open and free language models have been making headlines daily in the recent past. Alongside, Hugging Face, the messiah of open source, is hosting an extensive number of diffusion as well as language models. With quality resources from every corner, contributions from the FOSS community is laying the foundation to the upcoming peak of the genAI. ChatGPT is not everything Generative AI is not a singular technology, stated Gartner analyst Arun Chandrasekaran in an interview. It includes everything from foundation and diffusion models to prompt engineering tools. “They all enable this trend of generative AI,” he elaborated. Released two days ago, a report highlighted that every piece of visual art humanity has created over the last century and a half has been outnumbered by AI generated art just within a brief span of 1.5 years. The major chunk was produced by open-source diffusion models — developed last year by Stability AI. For a transformer based technology like GPT-n, the ‘n’ number of identified use cases and billions of dollars channelled is the reason for its fame. While previous generations of the software could technically do these things, the quality of the outputs was much lower than that produced by an average human. A rarely mentioned advantage of the AI is that it’s remarkably good at things that would take humans decades to do, like processing the entire canon of a certain literary or learning an artistic movement. According to Chandrasekaran, the main reason AI has hit the peak of the hype cycle is the sheer number of products claiming to have generative AI baked into them. Startups started flocking to the generative AI boat after OpenAI introduced GPT4’s capability via ChatGPT and the API. Even Sam Altman, the CEO of OpenAI has stated in the past of the technology becoming ‘wildly overhyped’. Technology’s ever-evolving landscape is a tale as old as time. In the early 1990s, there was the advent of the Internet and the opportunities it presented for enterprising startups to revolutionise industries ranging from e-commerce to software downloading. As the landscape shifts once again with generative AI, Startups but veterans like  are scrambling to find ways to leverage these cutting-edge technologies to gain a competitive edge. Even veterans like Google and Microsoft have redoubled their efforts with red teams to make the most of the technology. Embryonic stage While some have declared the technology as an inflating bubble, on the contrary as per Gartner it is still in the nascent stage. Aligned with the Gartner study, analysts predict the return on investment on the technology after 2 years. The hard part for the quality content producing technology is turning the value of time and productivity into an ROI measurement. Further proof of concept metrics as per experts also include scalability, ease of use, quality of response, accuracy of response and explainability or total cost of ownership. McKinsey estimates that the technology could add between $2.6 to $4.4 trillion of economic value annually across various industries. The study titled ‘The economic potential of generative AI: The next productivity frontier’ draws results from 63 new use cases analysed across 16 business functions that could deliver those returns. These estimated returns are comparable to the UK’s 2021 GDP of $3.1 trillion. While the ROI egg remains unhatched, companies should be focusing on tailoring and putting the technology to use as per their respective products. Trying to jump on the bandwagon for the sake of it can result in an actual bubble resulting in a market crash if not thought through.","excerpt":"If you think genAI is overhyped right now, wait for another 2-5 years, says Gartner","categories":["AI Features"],"tags":["AI Bubble","ai investments","ChatGPT","Gartner hype cycle","OpenAI"],"author_name":"Tasmia Ansari","publish_date":"2023-08-24T18:00:00","publication_year":"2023","word_count":791,"keywords":["ChatGPT","AI Bubble","GenAI","Hugging Face","ai investments","OpenAI","AI","AWS","RAG","Aim","prompt engineering","generative AI","Gartner hype cycle"],"extracted_tech_keywords":["AI","generative AI","GenAI","ChatGPT","OpenAI","Aim","Hugging Face","RAG","prompt engineering","AWS"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/genai-to-scale-the-peak-in-2-years\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10086599,"title":"Panic from Big Tech Layoffs Shadows Bigger Graces","content":"As the tech-induced recession plods on, layoffs continue to drain from Big Tech companies. According to employment firm Challenger, Gray & Christmas Inc, the layoffs in January have impacted 102,943 workers, a more than two-fold increase since December and a jump by more than five times from the past year. For all intents and purposes, most of these companies have blamed the layoffs on a single offender – overhiring. Do We Really Need Fewer Workers? It is interesting to see the extent to which companies have bloated up. A recent report published by the Bank of America Research stated that while the numbers had grown exponentially, companies did not need as many workers. In fact the productivity of workers had dropped across years – the number of workers per USD 1 million in revenue had fallen from eight in 1985 to just two in 2023. To most, this may not seem like a fair comparison simply because of inflation – a million dollars in 1985 is not the same as today. Do Big Techs Overhire? For anyone following economic downturns, the bursting of the tech bubble could have been spotted from a mile afar. As the pandemic lockdown persisted and industries like manufacturing suffered setbacks, tech soared unreasonably high. Tech companies grew richer as did their bosses and the techies themselves. It was all good until it wasn’t. On November 19, 2021, Nasdaq started plummeting after touching an all-time high signalling that a market correction was on the way for tech. Source: Forex.com But while the layoffs have done considerable damage, in comparison to the hirings completed during the pandemic, the numbers can seem measly. According to a report published by Daniel Howley from Yahoo Finance, Google had laid off 12,000 but had hired 67,880 jobs during the pandemic. Microsoft has so far announced 10,000 layoffs as compared to 77,000 they had added before and Amazon has laid off 18,000 layoffs against the job growth of 746,000 in the pandemic. Tech workers, also as a rule, have an upper hand in the labour market. They are most likely to find new jobs quickly because they have more degrees than the average worker and their skills are still in demand. A survey done by ZipRecruiter finds that “among people who were recently laid off and worked in tech previously, 37% found a new job within one month, and 79% found a new job within three months”. Will the Pattern Continue? It was pretty simple – when times were good, companies became too optimistic and went overboard with hiring. The ballooning demand for tech employees had pushed companies like Google and Meta to fight a talent war they weren’t prepared for. Vikas Kakkar, founder of HRTech firm Amara.ai said, “Because of the pandemic, many things have changed in the past three years.” Source: McKinsey “Many companies have moved their digital transformation plans from 2025 to 2020-21. Everyone wanted to switch to the hi-tech working style. The market’s demand in general increased, pushing tech companies to increase production to meet these demands. Many tech companies accelerated hiring to meet the demand because they thought the boom would continue even after the pandemic,” he continued. “However, as restrictions were lifted and people began to leave their homes and started going to office, demand fell, resulting in significant losses for these large technology firms. Some of these resources had been hired at a higher cost due to the sudden increase in demand,” he noted. And now that the market demand is returning to pre-covid levels, these tech companies are playing it safe by laying off excess and high-cost resources hired during covid, Kakkar explained. “A lot of the responsibility can also be placed upon investors,” Nehal Gupta, the director of AMU Leasing Pvt. Ltd., a tech NBFC said. “Most tech companies are investor funded and require a structured organisational hierarchy and large teams in the early stages of business. The focus is on scalability and growth vs achieving positive unit economics which leads to overhiring and later to massive layoffs when times are unprecedented,” she stated. There is some truth to this. Christopher Hohn, the billionaire-investor who holds a 6% stake in Alphabet, had reportedly written to Sundar Pichai raising alarms over the ‘excessive headcount’ of Google. Hohn was also concerned that Google employees were getting paid far above the market average and while the costs weren’t a problem until 2020, they were an issue now. “When things are looking up, hiring is a natural response but fitting in your budgeting needs is the way to go. And being a loyal employer comes with a cost but pays a better price in the long run in terms of employee turnover and productivity. Being loyal as an employer should be an integral functioning mechanism,” Gupta reasoned. But then can overhiring simply be treated as a cyclical issue? Gupta says this callousness will hurt companies in the long-term. “It is here to stay if companies don’t realise the hiring and turnover expenses. Offering relieving benefits comes at an expense which in the long run hurts the productivity and reputation of the company.”","excerpt":"Christopher Hohn, a billionaire investor with a 6% stake in Alphabet, had written to Pichai raising alarms over their ‘excessive headcount’","categories":["IT Services"],"tags":["Big Tech Layoffs","Layoffs","tech recession"],"author_name":"Poulomi Chatterjee","publish_date":"2023-02-06T15:00:00","publication_year":"2023","word_count":855,"keywords":["Go","Layoffs","AI","digital transformation","Scala","Git","RAG","Ray","ViT","Big Tech Layoffs","GAN","R","tech recession"],"extracted_tech_keywords":["AI","Ray","RAG","R","Go","Scala","Git","GAN","ViT","digital transformation"],"url":"https:\/\/analyticsindiamag.com\/it-services\/is-it-ok-to-overhire-when-the-goings-good\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":62409,"title":"How ‘Bias Bounties’ May Put Ethics Principles Into Practice","content":"In a paper published recently with the title ‘Toward Trustworthy AI Development: Mechanisms for Supporting Verifiable Claims’, a team of researchers from the Google Brain, Intel, OpenAI and other top labs from the US and Europe have launched a toolbox that will turn AI ethics principles into practice. The kit for organisations developing the AI models also includes the idea of rewarding developers for successfully detecting bias in AI, which is similar to security software getting rewarded with bug bounties. As per the authors of the paper, the bug bounty hunting community is still at its nascent stage, but can be useful in discovering biases. The initial idea of bias bounties was suggested in the year 2018 by co-author JB Rubinovitz. The recently published paper suggests ten different approaches to turn AI ethics principles into practice. Taking a look at the recent efforts, more than 80 organisations have come up with different AI ethics principles. However, the authors of the paper firmly believe that the present set of norms and regulations is insufficient to develop a responsible AI. The team has also advised on ‘red-teaming’ to detect susceptibility, along with aligning with third-party auditing and government policies to create new regulations specific to market needs. The team also makes several other recommendations, such as: Create a centralized incident database by sharing incidents about AI as a communityMaintain an audit trail during the development and deployment of AI systems for safety-critical applicationsStringent scrutiny of commercial models along with alternative open sources for commercial AI systemsBetter support for privacy-centric techniques, such as federated learning, differential privacy, and encrypted computationVerify hardware performance claims made by researchers through increased government funding This recent paper is an amalgamation of ideas from a workshop that was held in April 2019 at San Francisco that included more than 35 representatives from industry lab, civil society organisations, and academia. The authors made the recommendations mentioned above after they realized that the workshop failed to address certain claims made by AI practitioners. Since the use of AI in recent years has rapidly increased in several businesses and institutions, a rise of concern and activism has emerged around AI with a focus on issues such as bias amplification, ethics washing, loss of privacy, digital addictions, facial recognition misuse, and disinformation. AI systems have also been witnessed in strengthening the existing race and gender bias, which has resulted in biased facial recognition by the police and poor healthcare service for millions around the world. As per a report by The Leadership Conference on Civil and Human Rights, the usage of PATTERN risk assessment tool by the US Department of Justice was heavily denounced since the tool was found to be racially biased and used to send prisoners home early as a way to enforce social distancing amid COVID-19. The authors press for the need to move ahead of nugatory principles that are incapable of holding developers accountable. The paper reads that with rapid technological progress in AI and the spread of AI-based applications over the past several years, there is a growing concern about how to ensure that the development and deployment of AI are beneficial — and not detrimental — to humanity. “Artificial intelligence has the potential to transform society in ways that are both beneficial and harmful. Beneficial applications are more likely to be realised, and risks more likely to be avoided if AI developers earn the trust of society and one another. This report has fleshed out one way of earning such trust, namely the making and assessment of verifiable claims about AI development through a variety of mechanisms. If the widespread articulation of ethical principles can be seen as a first step toward ensuring responsible AI development, insofar as it helped to establish a standard against which behaviour can be judged, then the adoption of mechanisms to make verifiable claims represents a second,” concludes the author.","excerpt":"In a paper published recently with the title ‘Toward Trustworthy AI Development: Mechanisms for Supporting Verifiable Claims’, a team of researchers from the Google Brain, Intel, OpenAI and other top labs from the US and Europe have launched a toolbox that will turn AI ethics principles into practice. The kit for organisations developing the AI […]","categories":[],"tags":["AI &amp; ethics","Bug Bounty","Intel","OpenAI"],"author_name":"Rohit Chatterjee","publish_date":"2020-04-26T13:00:00","publication_year":"2020","word_count":649,"keywords":["federated learning","Go","artificial intelligence","OpenAI","AI","AI &amp; ethics","ML","Bug Bounty","Aim","differential privacy","Rust","R","Intel"],"extracted_tech_keywords":["AI","artificial intelligence","ML","OpenAI","Aim","federated learning","differential privacy","R","Go","Rust"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-bias-bounties-may-put-ethics-principles-into-practice\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10140861,"title":"Telecalling 2.0–AI Rings in India’s Next BPO Boom","content":"Outsourcing was a trend catching up fast during India’s IT boom in the 1990s. As tech giants and Fortune 500 companies turned to India, call centres offering well-paying jobs to the educated, English-speaking youth mushroomed all over the country. Today, as AI-driven customer interactions rise, the sector is being reshaped, revamping the way customer interactions unfold. In line with this shift is SquadStack, one of India’s telecalling outsourcing partners that offers AI-driven alternatives to conventional telecalling setups to empower leading businesses. The company addresses the challenges businesses face in managing customer communication throughout the lifecycle. Often called the BPO or enterprise communication market, this sector is valued at $22 billion in India and $450 billion globally. In an exclusive interview with AIM, Apurv Agrawal, the CEO & co-founder of SquadStack, said, “Our vision is to enable employment for people, wherever they are and whatever their background. In fact, 70% of our workforce comprises stay-at-home moms.” Agrawal illustrated this with an employee story. His colleague Priya had a tech job in Delhi, but after marriage, she moved to Ludhiana, where opportunities were limited. However, with SquadStack’s platform, her skills and potential found a new fit. “AI has been essential in supporting people like her. Traditionally, companies relied on centralised offices with hundreds of agents managed in person. By integrating AI, we can conduct remote screenings and quality audits, enabling people like Priya to access resources, AI-driven searches, and real-time coaching during customer interactions, even from a smaller town,” Agrawal said, adding that AI enhances control and guidance far more scalably than manual oversight, empowering such employees to perform effectively. Adopting AI in Telecalling According to Agrawal, AI helps streamline tasks like quality auditing, once done manually by agents who listened to calls and graded them—a monotonous and error-prone task. Now, AI consistently audits thousands of interactions simultaneously without fatigue. Similarly, AI manages simpler conversations through chatbots, providing instant responses and reducing wait times. Moreover, unlike human agents limited by availability, AI can scale to handle thousands of concurrent conversations, adjusting to peak hours and scaling down during off-peak times. This scalability is invaluable as AI takes on routine tasks, freeing humans to focus on more complex, high-value conversations. “This approach aligns with what we call the ‘Iron Man theory’, [which is] using AI where it excels and human talent where it’s indispensable. Just as industrial revolutions historically drove efficiency and opened new avenues for skilled roles, AI is enabling our agents to focus on higher-level conversations, resulting in better compensation and job satisfaction,” Agarwal noted. Humanoid Agent to the Rescue In the BFSI (banking, financial services, and insurance) sector, every lead presents a chance for connection and engagement. Imagine a customer receiving a prompt and a personalised conversation through a WhatsApp chatbot ready to handle initial queries in seconds. Often, the bot can tackle basic questions and tasks with ease. However, when things get complex, a human intervention with a seamless blend of AI efficiency and human empathy is necessary. Describing the human-AI partnership, Agrawal said, “With our humanoid agent, we can automate processes, ensuring seamless connectivity. Companies have achieved cost savings of 30-50% and increased conversion rates by 50-100% by implementing our system.” He noted that complete automation is still in its infancy, at less than 1%, but projections show a rise to 50% within two years. Hence, SquadStack is focussed on how augmented AI solutions can drive business growth and create lasting impact. What’s Next? India’s growing middle class, rising income, and low insurance penetration make it a prime market for AI-driven customer engagement. SquadStack is tapping into this opportunity by using AI-driven solutions to bridge linguistic and regional gaps in India’s telecom sectors. By blending AI with human support, SquadStack enables streamlined services and enhanced sales through remote teams, positioning AI as a force for millions across ‘Bharat’. On the global level, Bland AI, a San Francisco-based platform, is reimagining phone-based customer interactions. With hyper-educated, infinitely scalable AI agents integrated into companies’ tech stacks, Bland provides enterprise-grade stability with sub-1 second latency. The company has developed ‘Conversational Pathways’, a programming language that ensures structured and accurate responses to weed out inaccuracies. After emerging from stealth with $16 million in Series A funding led by Scale Venture Partners, Bland is set to expand its role in automating high-volume customer support, sales, and internal operations, showcasing how AI can elevate customer service worldwide.","excerpt":"At less than 1%, complete automation is still in its infancy but projections show a rise to 50% within two years.","categories":["AI Features"],"tags":["AI (Artificial Intelligence)","SquadStack"],"author_name":"Vidyashree Srinivas","publish_date":"2024-11-12T18:35:43","publication_year":"2024","word_count":733,"keywords":["funding","AI","chatbots","ML","Scala","automation","Aim","SquadStack","ViT","AI agents","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","ML","Aim","chatbots","R","Scala","ViT","automation","funding","AI agents"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/telecalling-2-0-ai-rings-in-indias-next-bpo-boom\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10123345,"title":"Thanks to ‘Apple Intelligence’, not OpenAI, Siri Now Understands You Better","content":"After making the world wait, Apple has finally uttered the word ‘AI’ but not without giving it its own spin by introducing ‘Apple Intelligence’. Apple Intelligence is deeply integrated with the new Siri, allowing users to speak more naturally to Siri, thanks to its enhanced language understanding capabilities. Siri is also getting a new look, featuring an elegant glowing light that wraps around the edge of the screen when it is active. Apple claimed that Siri users make 1.5 billion voice requests every single day. The Cupertino-based tech giant showed the world that it doesn’t need OpenAI’s help to improve Siri, and it’s pretty clear that it is not just a wrapper around ChatGPT. Apple has developed a 3 billion parameter on-device language model and a larger server-based model accessible via Private Cloud Compute on Apple silicon servers. These models are trained using Apple’s AXLearn framework, an open-source project released in 2023, built upon JAX and XLA. Apple doesn’t Really Need OpenAI Apart from building in-house Apple Intelligence, the iPhone maker has partnered with OpenAI to integrate ChatGPT-powered by GPT-4o into iOS. Apple says Siri will now be able to tap into ChatGPT’s ‘expertise’ when needed. For example, if you need meal ideas using ingredients from your garden, you can ask Siri. Upon receiving your permission, Siri will send the prompt to ChatGPT and provide you with suggestions. It’s worth noting that ChatGPT isn’t integrated directly into Siri, rather, it’s gaining access to it. Apple has not revealed the specifics of the deal or the amount it is paying OpenAI. Meanwhile, Tesla chief Elon Musk is unhappy with the Apple-OpenAI partnership. He said, “It’s patently absurd that Apple isn’t smart enough to make their own AI, yet is somehow capable of ensuring that OpenAI will protect your security & privacy!” However, this claim is not true. “Why is Elon Musk lying about how Apple’s ChatGPT integration works? He’s threatening to put iPhones in Faraday cages and claiming Apple isn’t smart enough to make its own LLMs,” said AI expert Vin Vashishta. “He knows Apple uses its own on-device LLMs. He heard Apple say that ChatGPT is only recommended when on-device LLMs can’t handle the request, and data is only sent if users choose that option,” he added. Moreover there is no clarity on whether OpenAI will use users’ data to train its model. OpenAI said that users can choose to connect their ChatGPT account, which means their data preferences will apply under ChatGPT’s policies. OpenAI’s GPT-4o’s voice feature is not yet released. It would be interesting to see whether Apple would integrate that as well in the future. However, the company did drop a hint about multiple possible future partnerships. There is a possibility that Apple might also partner with Google, especially after Google recently introduced ‘Project Astra‘ at Google I\/O 2024. Astra is a universal AI agent built on Google’s Gemini models, designed to process multimodal information, such as text, images, video, and audio, to understand context and respond in real-time. Word has it that Apple may also partner with Anthropic. Siri Can Take Actions The highlight of the new Siri is its ability to take actions. With Apple Intelligence, Siri gains on-screen awareness, enabling it to understand and act on what’s displayed. For instance, if a friend texts you their new address, you can simply say, ‘Add this address to their contact card’, and Siri will do it right from the Messages thread. you hear that? \"siri can take actions on your phone\"this is the LAM you want. holy shit.— Sully (@SullyOmarr) June 10, 2024 Moreover, Siri can now take actions within apps on your behalf. It can perform hundreds of new tasks in and across apps, including utilising Apple’s new writing and image generation capabilities. For example, you could say, ‘Show me my photos of Stacey in New York wearing her pink coat,’ and Siri will display them instantly. The feature is quite similar to Rabbit’s LAM and what Humane Ai Pin attempted to offer. However, it’s now clear that one doesn’t require a new device to access LLMs, and smartphones aren’t going anywhere anytime soon. its over for rabbit R 1 and humane ai slop pic.twitter.com\/WbPuR7cZEJ— nik (@ns123abc) June 10, 2024 Recent reports suggest that Humane is actively seeking a potential buyer for its AI Pin business after it faced widespread criticism for failing to meet expectations. Moreover, thanks to Apple Intelligence, Siri now understands your personal context. With a semantic index of your photos, calendar events, and files, as well as information from messages and emails—such as hotel bookings, concert tickets, and shared links—Siri can now find and comprehend things it couldn’t before. For example, when you’re filling out a form, you can simply ask Siri to locate a personal document, such as your driver’s license, from your photos and auto-fill the details into the form. Apple Intelligence enables Siri to perform hundreds of new actions within Apple and third-party apps. For instance, a user might say, “Retrieve the article about cicadas from my Reading List”, or “Share the photos from Saturday’s barbecue with Malia”, and Siri will handle the tasks effortlessly. This was made possible through significant enhancements that Apple is making to App Intents, a framework that lets apps define a set of actions for Siri, Shortcuts, and other system experiences.","excerpt":"Apple has developed a 3 billion parameter on-device language model and a larger server-based model accessible via Private Cloud Compute on Apple silicon servers.","categories":["Global Tech"],"tags":["Apple"],"author_name":"Siddharth Jindal","publish_date":"2024-06-12T13:52:57","publication_year":"2024","word_count":888,"keywords":["Anthropic","ChatGPT","Go","OpenAI","AI","Apple","GPT-4o","GPT","Aim","JAX","R"],"extracted_tech_keywords":["AI","GPT-4o","ChatGPT","OpenAI","Anthropic","Aim","JAX","R","Go","GPT"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/thanks-to-apple-intelligence-not-openai-siri-now-understands-you-better\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":40034,"title":"Zeta Is Revolutionising Fintech Space By Providing Cloud-Based Solutions For Employee Benefits","content":"Ramki Gaddipati, CTO and co-founder of Zeta Fintech startups in India have seen exponential growth in the last few years with a surge in digital mode of payments and companies offering digitised enterprise solutions. While the likes of Paytms have made a significant mark in the space, there are others such as Zeta who are witnessing unprecedented growth and success. Founded in 2016, Zeta excels in offering solutions for employee tax benefits, automated cafeterias and digital payment. Analytics India Magazine got in touch with Ramki Gaddipati, CTO and co-founder of Zeta to get interesting insights on the startup and how have they been able to service over 170 corporates in less than a year. The Story Behind Zeta Gaddipati shares that he founded Zeta along with serial entrepreneur Bhavin Turakhia with a vision to disrupt India’s digital payments space. “We wanted to make payments easy, inclusive and valuable for corporations, employees and merchants everywhere. Our offerings revolve around the central idea that spending and receiving money should be easy, fast, and trouble-free,” he shares. Keeping their vision in mind, they forayed into digitised enterprise solutions with revolutionary solutions for employee tax benefits, automated cafeterias, and gifting. They have the following offerings: Zeta Tax Benefits: Zeta started its journey by launching an innovative cloud-based enterprise solution for employee benefits called Zeta Tax Benefits. The tax benefits suite includes smart solutions such as fuel and vehicle maintenance, telecom reimbursement, gadget allowance, gift card, LTA and many more. Zeta Tax Benefits makes it possible for employees to save up to Rs 80,000 in taxes every year. Zeta Express®: It is a smartphone-based meal ordering and payment solution that is revolutionising the corporate in cafeteria space. With Express, employees can place an order from the convenience of their desks using the Zeta app within seconds, thereby enabling an efficient meal ordering experience. Zeta Spotlight®: It is an on-demand platform to digitally manage every kind of gifting and R&R need for employees and channel partners. From on-the-spot awards to festive occasions, sales incentives to special occasions, a corporate can plan and execute gifting for every occasion digitally with Zeta Spotlight®. Corporates can choose what they want to gift with no vendor management, no pilferage and no time lost in logistical hassles. Over the last three years, Zeta has built a diverse and strong client base with names such as Tata Motors, Oyo rooms, Shopclues, Motilal Oswal among others, and has helped corporates save time, cut costs and reduce duplication of work. “Zeta provides some of the latest technologies which are backed by technologies like optical character recognition, or OCR, and machine-driven transcription. Our platform also offers real-time analytics that helps corporates in decision-making,” shares Gupddipati. Understanding The Data Analytics Play At Zeta Gaddipati explains that the idea behind every Zeta product is to maximise efficiency and reduce the time that is wasted in usual processes and to achieve that, they rely heavily on analytics. Founders at Zeta strongly believe that quantitative insights revealing behaviour can actually give a fair measure of employee’s preferences including purchase. They use item consumption and trend metrics to gauge what employees are purchasing and not within the cafeteria. “Analytics that gives the highest selling item or the lowest selling item can effectively predict employee purchase behaviour. Employee Feedback (both positive and negative) also helps us to co-relate one’s likelihood of a specific purchase. All the above metrics and much more are available to the corporates through our analytics dashboard which we call as Express insights,” he says. Real-time analytics helps the company to manage their inventory better. It also helps them find what employees are spending on, predict demand in the cafeteria by modelling their historic purchases, eliminate food wastage and raise a better ROI. Zeta has also expanded its automated cafeteria portfolio to offer advanced deep analytics – Express Insights. The Insights feature is a first-of-its-kind solution in the automated cafeteria industry that will offer corporates and cafeteria vendors detailed analysis to enable better food management and enhance overall cafeteria experience. Since inception, Zeta Express® has powered 275+ cafeterias across 130+ corporates in India. Roadmap For The Coming Year Gaddipati is quick to share that personalisation analytics would be a key building block towards presenting metrics to clients. They also aim to analyse behavioural data to reveal patterns, trends and any correlations to improve customer offerings. “We would work closely with organisations understanding their needs and the deliverables expected for managing their cafeterias and help them get relevant and customised metrics which can be deeply drilled to specifics thereby giving a lot of productivity-enhancing and cost minimising levers,” he said on a concluding note.","excerpt":"Fintech startups in India have seen exponential growth in the last few years with a surge in digital mode of payments and companies offering digitised enterprise solutions. While the likes of Paytms have made a significant mark in the space, there are others such as Zeta who are witnessing unprecedented growth and success. Founded in […]","categories":["Deep Tech"],"tags":[],"author_name":"Srishti Deoras","publish_date":"2019-06-03T06:57:22","publication_year":"2019","word_count":775,"keywords":["Go","AI","Git","RAG","Ray","Aim","analytics","real-time analytics","GAN","R"],"extracted_tech_keywords":["AI","analytics","Aim","Ray","RAG","R","Go","Git","real-time analytics","GAN"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/zeta-is-revolutionising-fintech-space-by-providing-cloud-based-solutions-for-employee-benefits\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":489,"title":"Interview – Ajay Kelkar, Co-Founder and COO at Hansa Cequity","content":"Ajay Kelkar of Hansa Cequity is speaking at The Fifth Elephant on the topic of ‘Telling stories with data’. In an exclusive interview with Analytics India Magazine, Ajay talks about the topic, Hansa Cequity and analytics industry as a whole. Analytics India Magazine: Could you tell us more about the topic that you are speaking on at The Fifth Elephant? Ajay Kelkar: The discipline of “Analytics” is often interpreted in many different ways. Almost as if it were like the story about the “blind man & the elephant”. Analytics is most impactful when it is about “taking effective decisions rather than sexy insight”. That is my perspective. To make effective decisions you need to influence people with insight & therefore Analytics folk must become “storytellers with data”. While technical aspects of Analytics are important, the value can be realized only if these “storytellers” influence key decisions within a company. This insight is going to be at the heart of my presentation. AIM: How was Hansa Cequity incepted, how has it evolved over the years and what is the next step? AK: I was earlier the CMO at HDFC bank & my partner Swamy ran a WPP agency called I Contract, India’s largest Direct marketing agency. We always saw a yawning gap between companies “having customer data” & having “skills to monetize that data”. We partnered with the Hansa group & created this venture to address this “need gap” in the market. So by definition, we haven’t conceptualized ourselves as a pure-play analytics company. We are a company that has analytics & technology at its heart & we use this building block to create CRM programs that make a business impact for our clients. We have grown well & have created a unique niche. AIM: Though you would be speaking about it in your talk, can you briefly provide an overview of why is it hard to extract meaningful business value from data? AK: It is hard because Analytics requires an intersection of skill sets from technology, Business & statistics. Such an intersection is hard to create & conditions for it to flourish may not exist everywhere. Without this intersection, Analytics can become boring. AIM: A brief overview on what are the elements that must come together to extract value from data? AK: At Hansa Cequity, our model is unique because it tries to integrate very contrasting dimensions into one entity where the sum is larger than the parts! Having a designer’s sense with data may contrast with a statistician’s dry look at numbers! We seek “intersection” skills-intersection of Creative, technology, data & business! Not easy to do with highly talented people & we are attempting it! AIM: You have formulated the term ‘Information Journalists & Filmstars’. What do you mean by this? AK: “Information journalists” are really journalists who recognize that a picture can tell a far better story than words.  The interesting thing is that journalism is getting far savvier with data! I see visual data based story telling in the New York Times that is absolutely mind boggling. Even here in India, I see some lovely data visualization in the Mint! Storytelling with Data is becoming much more common today because of both vast amounts of data being available in the public space & also the emergence of a newer breed of younger, more “social” professionals who consume such data with far more ease! But is the corporate world looking at data like this? Are analytics teams creating powerful data visualizations to tell their stories! Can they afford not to? AIM: Do you think that companies today are using analytics widely to make decisions? AK: Companies are using analytics now pretty widely. Some industries are more penetrated than others. Critical for CEO  & CXO to recognize the criticality of this & drive such adoption. Acquisition mindsets are the norm in hugely growing markets where business is a lot like “land grabbing”. But a more mature economy with higher competitive intensity seeks adoption of analytics at a far higher pace. [divider top=”1″] Ajay Kelkar, Co-Founder & COO of Hansa Cequity, has over 20 years of experience in customer-driven marketing across a wide range of industries like Soft goods, Banking & Financial services & Retail. He has had exposure to a wide variety of business styles & cultures across Procter & Gamble, Britannia, Marico, Shopper’ Stop & HDFC bank. He was the head of marketing for HDFC Bank, a leading private sector bank in India. He was responsible for building world-class analytical marketing capability for the bank across the country. He led the bank’s marketing team on a pioneering path of data-led marketing & analytics. Ajay holds a Chemical engineering degree from UDCT Mumbai and an MBA from one of India’s premier business schools, Indian Institute of Foreign Trade (IIFT).","excerpt":"Ajay Kelkar of Hansa Cequity is speaking at The Fifth Elephant on the topic of ‘Telling stories with data’. In an exclusive interview with Analytics India Magazine, Ajay talks about the topic, Hansa Cequity and analytics industry as a whole.  Analytics India Magazine: Could you tell us more about the topic that you are speaking […]","categories":["AI Features"],"tags":["Interviews and Discussions"],"author_name":"Дарья","publish_date":"2012-07-20T20:53:44","publication_year":"2012","word_count":800,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","Aim","analytics","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","analytics","Aim","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/interview-ajay-kelkar-co-founder-and-coo-at-hansa-cequity\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":38046,"title":"TCS Ups Hiring By 377%, Infy By 642%; Experts Say It&#8217;s The Work Of AI, Analytics","content":"In a piece of news that will come as a good tiding to all of the IT sector, Indian tech behemoths Tata Consultancy Services and Infosys have paced up their hiring in the emerging tech sector manifold. According to numbers released by a noted business magazine, TCS has increased their hiring in IT by 377% and Infosys by 642%. This development comes as a positive change, especially after several IT companies handed out pink slips in 2017-18 as a result of lack of projects, increase in automation and migration to cloud computing, among others. Reportedly, these IT giants have stopped the mass hiring and have focussed on employing niche candidates who have expertise in artificial intelligence, blockchain, data mining and analytics, among others. Priya Chetty-Rajagopal, managing partner, Multiversal Advisory, told the magazine, “It’s the more cutting edge that is getting to be in demand. The focus is on data sciences, AI, internet of things, analytics and user experience design for products, etc. Much of the IT business gained has a stronger high-tech flavour — and includes this as part of the high-end offering but the talent is still hard to get. So these skills are going to be highly in demand across IT firms.” 2019 has seen overall growth in the number of jobs in analytics and data science ecosystem with India contributing to 6% of open job openings worldwide. According to a study by Analytics India Magazine, the total number of analytics and data science job positions available is 97,000. Out of these, 97% of the openings in India are on a full-time basis while 3% are part-time or contractual. The IT majors are now also helping their staff to be future-ready. Companies have initiated multiple internal training in the digital technology space for its employees. Our study titled Analytics And Data Science Jobs In India 2019 – By Great Learning & AIM had found out that there is a whole section of employees who are on a path for lifelong learning approach to skills development. As leading institutes respond to the rising demand for analytics skills with tailored programs, skills development programs will play a critical role in updating the analytics capabilities and preparing the workforce for new roles. With the continued lack of talent plaguing the technology sector in India, the acqui-hiring trend has taken off in popularity in India and the country is seeing big tech firms snapping up data science consulting firms in a bid to bolster their bench strength. In addition to this, companies are increasingly looking inwards to identify whether they have the talent to capitalise on new opportunities that can reshape the nature of the industry. Kris Lakshmikanth, CEO, Head Hunters India, told a noted daily, “The market is definitely better off this year when compared to the last one… Also, with this hiring numbers, companies are showing that they have confidence in the near future and this is an effort to beef up the bench.” As the IT sector evolves and emerging tech like data analytics, AI and IoT become more entrenched across domains, organisations and high-growth startups are going on a hiring spree. However, the market continues to be employer-driven and finding the right talent is a growing challenge for organisations and startups. On the skill side, as more companies bolster their analytics capabilities, they are aggressively hiring professionals with sound technical knowledge. A key takeaway is that the analytics and data science talent supply will not be able to keep pace with job growth. In other words, there isn’t enough analytics talent and organisations will continue to face hiring challenges.","excerpt":"In a piece of news that will come as a good tiding to all of the IT sector, Indian tech behemoths Tata Consultancy Services and Infosys have paced up their hiring in the emerging tech sector manifold. According to numbers released by a noted business magazine, TCS has increased their hiring in IT by 377% […]","categories":["IT Services"],"tags":["AI (Artificial Intelligence)","Infosys","IoT","TCS"],"author_name":"Prajakta Hebbar","publish_date":"2019-04-22T12:51:12","publication_year":"2019","word_count":601,"keywords":["data science","Go","API","artificial intelligence","Infosys","AI","cloud computing","R","Git","Aim","analytics","AI (Artificial Intelligence)","TCS","IoT"],"extracted_tech_keywords":["AI","artificial intelligence","data science","analytics","Aim","cloud computing","R","Go","Git","API"],"url":"https:\/\/analyticsindiamag.com\/it-services\/tcs-ups-hiring-by-377-infy-by-642-experts-say-its-the-work-of-ai-analytics\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":19844,"title":"Companies Must Adopt AI And VR; Or They Might Lag Behind The Competition, Says Sameet Gupte, CEO, Servion Global","content":"Servion Global Solutions, a customer experience management expert has been managing over 10 billion customer interactions annually across 60 countries in 6 continents to enable business transformation for enterprises. It is a platforms-led systems integrator focused on strengthening customer experience strategies by automating and integrating customer interaction channels. Over the past two decades, Servion has evolved from being a single channel (voice) interaction expert to being an industry pioneer in omni channel customer experience. It’s AI enabled Virtual Assistant solution is the human like voice and chat agent that can simulate a live agent – from initiating customer conversations to delivering information and taking actions on their behalf. It can also detect customer’s emotions, perform transactions, and augment context and personalisation while elevating the overall customer experience. AIM interacted with Sameet Gupte, CEO, Servion Global Solutions to understand if Indian companies have truly tapped into the customer psyche to offer the ultimate experience, how they are leveraging emerging technologies such as AI and VR to fuel customer experience strategies and more. Analytics India Magazine: Would you like to give us a brief about Servion and how is it excelling in the area of analytics and artificial intelligence? Sameet Gupte: Servion is a niche player in the CX\/DX space. Since our inception, we have focused on automating the customer service. We enable our clients to harness artificial intelligence to better understand their customers, integrate technologies together and measure outcomes. We leverage analytics and AI to create intuitive and enhanced CX. The ServIntuit and ServInsight platforms of Servion which are currently being used by blue chip customers worldwide have integrated analytics to predict customer behaviour and AI to enable next best action. AIM: How is it enabling business transformation for enterprises in the area of customer experience management (CEM)? What are the technologies used here? SG: Servion’s Customer Experience Management solutions are the meeting point between innovation and expectation. We believe technologies such as artificial intelligence, predictive analytics, virtual and augmented reality, and IoT have become customer experience game changers. AI-powered chatbots and intuitive voice services have already taken over large parts of customer services. At Servion, we offer an array of CX-Platforms and pre-integrated omni-channel solutions that help companies reimage their customer experience by leveraging these emerging technologies. ServIntuit is an omni-channel customer experience platform that enables intelligent automation and Next Best Actions for optimal responses to customers. ServInsights is a multi-channel analytics platform that enables informed decision making by integrating channel data with business metrics. ServCloud is a cloud-based CX platform that enables enterprises to benefit from AI without the capital and skills needed to manage the large-scale computing platforms. ServCare is an end-to-end support and maintenance platform that enables enterprises to focus on growth strategies – instead of worrying about daily operational challenges. AIM: How are Indian companies leveraging emerging technology such as AI and VR\/AR to to fuel customer experience strategies? How is AI playing an active role in Servion’s decision making aspects? SG: Artificial Intelligence has seen a steady adoption across domains in recent years in India. In fact, it is triggered by a tectonic shift in customer expectations. A new generation of customers has started to expect futuristic engagement while flexing their purchasing powers. To keep up with them, companies are using AI to ‘learn’ from customer scenarios, and make ‘next best action’ decisions instantly. AI-driven automated services, such as chatbots and voice recognition, are being used solve many customer service challenges customers face; such as having to wait on hold, or navigate endless menu options. AI also makes it easier for them to automate repetitive workloads, saving the customer time and the company precious resources. With AI-driven responses answering simpler queries, more complex issues can be identified and diverted to specialist expert customer hubs. This not only resolves customers’ problems faster: it also drastically reduces the number of queries businesses have to run through their contact center. One of our telco clients was able to handle four times as many calls as before through automation, without having to expand capacity. The number of tasks that can be automated using AI will drastically increase. We predict that 95% of customer contact will be through automated systems in the next five years and that customers will welcome this change. Thanks to the gaming industry, VR technology, in particular, has a great degree of awareness among millennials. Visual technologies such as VR, AR, and holograms allow companies to show and not just tell; the potential for improving the customer experience is huge. From viewing a property via VR headset, through to tech support using AR goggles to show where to press, plug in or switch on, there are multitudes of largely untapped opportunities for organisations to give customers a completely new way to interact; whilst improving customer loyalty and reducing complaint time. Adoption level for these are still low in India, but they must start exploring how to use these technologies now, or find themselves lagging behind the competition. With end-to-end capabilities – from cognitive assessment to execution, and strategic partnerships with some of the biggest innovators, we have been helping clients to adopt cutting-edge service ecosystems to deliver differentiated customer experience. Further to our experience of having successfully driven over 600 CX journey implementations, our multi-disciplinary bandwidth has enabled us to think of tomorrow as today; not just in the way we set our client expectations, but in our daily decisions. AIM: Would you like to talk about Servion’s AI enabled virtual assistant and how is it enabling a smooth customer experience across your client base? SG: Servion’s AI-enabled virtual assistant is voice biometric enabled and offers a humanlike voice and chat agent that can simulate a live agent – from initiating customer conversations to delivering information and taking actions on their behalf. This solution has been helping clients to automate proactive text-related and voice engagement channels such as live agent support, live chat, consumer messaging apps or SMS.  It also detects customer’s emotions, performs transactions, and augments context and personalisation while elevating the overall customer experience. This year, Stevie Awards recognised Servion’s AI-enabled Virtual Assistant with a Silver Stevie in the Best New Product or Service of the Year category for B2B products. AIM: Who are your major clients? What are the services you offer them? SG: Our client-base comprises blue chip companies, including many of the world’s largest banks and telco operators. We manage over 10 billion customer interactions annually across 60 countries in 6 continents. Three out of the top 5 banks, 2 out of the top 5 insurance firms and 5 out of the top 10 telcos work with Servion to improve their customer experience. We offer a comprehensive suite of solutions and services in the CX space. From big data and analytics and cloud to the entire support spectrum and AI-led implementation, we have had the right platform to deliver and exceed expectations. AIM: Please give us a detailed note on how is Servion making use of data analytics? SG: ServIntuit, Servion’s omni-channel customer experience platform, integrates data from all enterprise channel applications such as chat, voice, website, mobile, email and video to drive personalised, consistent and contextual customer experience. Powered by our Next-Best-Action Engine and analytics module, it provides optimal responses to customers – on the channel of their choice. This drives proactive, positive and unified interactions across all channels for unbeatable ROI, laying the foundation for strong customer relationships, superior loyalty, and optimised revenue. ServInsights, Servion’s multi-channel analytics platform, is designed to deliver actionable insights to contact center managers, operations, technology and business heads and drive contact center performance. It uses advanced analytics to give a 360-degree view of data-driven decisions by integrating multi-channel data, monitoring and analysing key business metrics, as well as customer trends and patterns. It is a single-box solution that comes with zero additional development or training expenses, does not require any additional database licenses, which reduces the total cost of ownership. AIM: What is your roadmap\/plans for analytics and AI at your company in the future? SG: In the AI space, we are looking at building a strong case for Robotics-as-a-Service to make it possible for our clients to integrate robots and embedded devices into their web and cloud computing ecosystems. We are also keen to build our expertise in maximising the potential of VR\/VR to ensure future-preparedness for our BFSI and Retail clients. In the Analytics space, we are looking to stay ahead of the data curve – as it becomes smarter by the day. Our powerful in-house platforms – ServIntuit and ServInsights – are agile and flexible, and they can easily incorporate upgrades that may come in the future of Machine Learning and Natural Language Processing. AIM: What are the most significant challenges you face being at the forefront in analytics space? SG: The biggest challenge comes from unifying the data collected through multiple sources. Companies have made huge investments in customer experience infrastructure over the years, from call centers through to CRM systems, with more channels and services added daily. This has created a complex web of interdependencies prone to performance problems, inefficiencies, and rising costs. Organisations are in a tough position: they must keep up with digital innovation, but starting from scratch each time is impractical. Stitching old and new systems together is extremely challenging, demanding skills that many businesses do not possess. Providing a seamless, integrated, omni-channel customer experience under these circumstances is impossible. The second challenge will come from the growth of internet-connected devices. Gartner, Inc. forecasts that 8.4 billion connected things will be in use worldwide in 2017, up 31 percent from 2016, and will reach 20.4 billion by 2020. Through the Internet of Things, these connected devices will start transmitting the customer’s needs in real-time, from restocking the fridge to adjusting the heating:  businesses will be judged on how quickly they meet those demands. Connecting so many devices and drawing customer insights will be a big challenge. Organisations looking to leverage analytics better need to design a forward-thinking strategy that won’t simply tide the business over, but will re-optimise the customer experience and make the business more efficient. AIM: How do you think ‘Analytics’ as an industry is evolving today? Could you tell us the most important contemporary trends that you see emerging in the present analytics space across the globe? SG: In the next five to ten years, not only will machine handle up to 95% of customer interactions: we will prefer this to people. Just as customers are more inclined to use self-service checkouts or ATMs, we will turn to fast, efficient machines for most tasks. Human experts will instead be needed for more complex queries. Very soon we will also see AI-driven voice activation taking over a range of customer interactions. For example, Swedbank recently deployed AI to learn what customers want and how best to help them by assimilating website searches and contact center enquiries. Their AI platform now handles most transactional calls that would have previously been handled by a human, while more complex enquiries – for example, mortgage applications – are passed to a person. The great transformation these technologies allow is an increasingly predictive experience: using behavioural analytics to understand patterns and predict user intent. For instance, if a bank customer makes a monthly payment over the phone, the system should know to ask if they would like to complete their usual transaction when they call on the relevant date. Using AI-powered data analytics platforms, organisations can already detect the context of a customer issue and arm the call handler, whether a person or a machine, to suggest a ‘next best action’. In the future, these systems will understand us even better; including detecting and reacting to emotion. Companies are also already incorporating visual technologies, such as holograms, VR and AR, into the customer experience. Heathrow Airport, for example, is using holograms to greet customers: and the use cases for these technologies are set to expand. Big Data has, thus far, been the strategic investment buzzword across domains. Slowly, there may come a shift in industry attention to Smart Data. “Self-service” analytics are emerging to democratise the learning capacity of end users, and driving organisations to adopt leaner business performance models. From visualisation to action, they are evolving fast – irrespective of the device or channel. In the world of cloud data, the near-future trend would be in not just delivering deep-dive insights on the data, but in making it seamless and simple for enterprises to act on them. With cloud security growing stronger, real-time analytics will very soon be an integral part of it.","excerpt":"Servion Global Solutions, a customer experience management expert has been managing over 10 billion customer interactions annually across 60 countries in 6 continents to enable business transformation for enterprises. It is a platforms-led systems integrator focused on strengthening customer experience strategies by automating and integrating customer interaction channels. Over the past two decades, Servion has […]","categories":["AI Features"],"tags":["AI Companies","cognitive computing human capital","Interviews and Discussions"],"author_name":"Srishti Deoras","publish_date":"2017-12-19T05:28:29","publication_year":"2017","word_count":2105,"keywords":["artificial intelligence","machine learning","AI","chatbots","ML","RAG","Ray","Aim","analytics","AI Companies","Tecton","cognitive computing human capital","Interviews and Discussions"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","analytics","Aim","Ray","Tecton","RAG","chatbots"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/companies-must-start-exploring-technologies-like-ai-vr-now-find-lagging-behind-competition-says-sameet-gupte-ceo-servion-global\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10013344,"title":"&#8216;Biggest Leap&#8217;: Qualcomm Introduces Range Of AI Capabilities In New Snapdragon Processor","content":"On the first day of the Snapdragon Tech Summit Digit, Qualcomm has announced the release of the Snapdragon 888, which is its latest 5G-equipped flagship smartphone processor. This, according to the company, would set the benchmark for flagship smartphones in 2021. This new smartphone processor offers industry-leading mobile innovations in 5G for improved enterprise mobility, video telephony, console-quality cloud gaming etc. Of special interest is the artificial intelligence offered by this processor, which uses the 6th generation Qualcomm AI Engine with Hexagon 780 processor, which provides performance improvement by up to 50%. Enhanced AI Capabilities Qualcomm describes Snapdragon 888 as its ‘biggest leap in architecture and performance in years’. With the introduction of the new Hexagon 780 processor, Qualcomm removes the distance between the accelerators and combines them to form the fused AI accelerator architecture, as opposed to a separate scalar, vector, and tensor accelerators used for earlier models. As a matter of fact, the performance per watt on the Hexagon 780 processor is three times higher than the previous generation. Additionally, the company has also dedicated a large shared memory across each of the accelerators that will help in the efficient movement of data. This memory space is about 16 times larger than its predecessor and offers up to 1000 times faster hand-off time between. The new processor 888 also offers a processing speed of 26 TOPS (Tera Operations per Second), which according to the company is the highest for TOPS performance on mobile, more than the previous best offered by Qualcomm Snapdragon 865. Further, the Adreno 680 GPU of the AI engine in the 888 processor offers a 43% AI performance boost and a faster floating-point calculation. The Snapdragon 888 processor also introduces the 2nd generation Qualcomm Sensing hub, which has a low power AI processor (less than 1mA) that gives a five-time improvement on the AI performance. This extra AI processing power will offload up to 80% of the workload that is usually borne by the Hexagon processor. This sensing hub collects and deciphers data from different cores, such as connectivity data from 5G, Bluetooth and location stream and creates contextual awareness use cases. Additionally, Qualcomm is also working with Google and TensorFlow Micro Framework to give developers easier access for optimisation on both Hexagon processors and the AI processor of the Qualcomm Sensing Hub. Qualcomm’s on-device AI SDK called the Neural Processing SDK, has also been improved. It now includes support for additional models and gives expanded support for Windows 10 AI use cases on Snapdragon 888-powered laptops. The company also introduced the Qualcomm AI Engine direct along with the 6th generation Qualcomm AI Engine. With this, the developers will get direct access to its different hardware components such as Hexagon 780 processor, Adreno GPU, and Qualcomm Kyro CPU. The Qualcomm AI Engine brings a unified AI API across the whole Snapdragon platform. This API is backward compatible, meaning that it will also support the previous 5th generation Qualcomm AI Engine. The company is now working on improving the modularity and the extensibility of the AI Engine to enable developers to create their own Snapdragon accelerated AI solutions. The announcement of the open-source compiler for AI accelerators, TVM, was made earlier this year. Using this, developers will be able to write custom operators in just a few lines of Python, which can be compiled for the Hexagon processor and plugged directly into the Qualcomm AI Engine direct framework. Other Features Snapdragon 888 also has a 3rd generation Qualcomm Snapdragon X60 5G Modem-RF System that will support all-day battery life even on 5G network. Users will be able to experience multi-gigabyte of speed with faster uplink and downlink 5G speed. The key difference between the Snapdragon 888 and earlier released 865 is that the former has integrated its 5G modem directly into the SoC. The new processor also sees better gaming performance with the addition of Qualcomm Snapdragon Elit Gaming arsenal. It offers 35% faster graphics rendering, 20% increase in responsiveness, and 25% higher performance and power efficiency. Details of all the features and specifications can be found here.","excerpt":"On the first day of the Snapdragon Tech Summit Digit, Qualcomm has announced the release of the Snapdragon 888, which is its latest 5G-equipped flagship smartphone processor. This, according to the company, would set the benchmark for flagship smartphones in 2021. This new smartphone processor offers industry-leading mobile innovations in 5G for improved enterprise mobility, […]","categories":["Deep Tech"],"tags":["AI Capabilities","big data platform c++"],"author_name":"Shraddha Goled","publish_date":"2020-12-07T10:00:00","publication_year":"2020","word_count":682,"keywords":["Go","API","artificial intelligence","AI","AI Capabilities","Scala","Git","RAG","Python","big data platform c++","TensorFlow","R"],"extracted_tech_keywords":["AI","artificial intelligence","TensorFlow","RAG","Python","R","Go","Scala","Git","API"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/biggest-leap-qualcomm-introduces-range-of-ai-capabilities-in-new-snapdragon-processor\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10078277,"title":"VCs Based in India Investing in AI Startups","content":"The Indian AI industry is booming. According to International Data Corporation, India’s AI market is estimated to reach up to $7.8 billion by 2025. Several new startups are building products centered around AI or incorporating AI in their systems. Thus, resulting in groundbreaking innovations and advancements of the sector. There are hundreds of VCs investors across public and private sectors that are increasingly getting interested in “pitching into” the sector. Let’s look at some of these India-based investors looking to invest in AI and the technology sector in general! Unicorn India Ventures With 34 lead investments and raising a fund of $61 million, Unicorn India Ventures has been focusing on investing in AI startups with their segments of preference being Fintech, SaaS, and DeepTech. Founded in 2015 by Anil Joshi and Bhaskar Majumdar, the duo has invested in 45 startups by far. Their first AI-based investment was in 2017 with SmartCoin and since then their interest in the field has grown. Click here to read the exclusive conversation their team had with Analytics India Magazine. 100X.VC Sanjay Mehta founded 100X.VC in 2019 and has since then invested in 18 AI startups. With a total of 80 startups, the firm’s most preferred sectors include Fintech, Deeptech, and B2B SaaS. As a private investor for around a decade, Mehta has invested in more than 150 startups globally. Some of the AI-based startups that the firm has invested in include Kroop AI, Vecros, Vitra.ai, OurEye.AI, Solvio.AI, and Data Sutram among others. In an exclusive interview with Analytics India Magazine, Sanjay Mehta elaborates on his investment philosophy. Mumbai Angels Founded by Prashant Choksey and Praveen Chakravarty in 2006, Mumbai Angels is a group of investors who are investing in early stage startups. The firm, with around 700+ members across 70 locations globally, has invested in around 200 startups till date with over 100 successful exits. Their portfolio companies include Brainwired, Unlu, CarterX, AppsDaily, MangoReader, Catapoolt, among others. Click here to visit their website and pitch ideas. Sequoia Capital India Based in Bengaluru and founded in 2000, Sequoia Capital India is a branch of Sequoia global that is operating in Southeast Asia and India. The founders—KP Balaraj, Raj Dugar, Sandeep Singhal, SK Jain, and Sumir Chadha—have invested in around 600 startups with their most recent investment in K-12 Techno Services on October 24. Some of their lead investments include Moglix, Mobile Premier League, Privado, Arrow, Metasky, Razorpay, among others. Click here to visit their website. Blume Ventures Blume Ventures focuses specifically on technology startups and has backed more than 175 startups  with 28 successful exits. Based in Mumbai, and founded in 2011 by Karthik Reddy, Rob Blum, and Sanjay Nath, the firm invests in Seed and Pre-Seed rounds. They have invested in companies like Dunzo, Servify, Exotel, GreyOrange, Unacademy, Cashify with most recently Vecmocon Technologies, Euler Motors, and BimaKavach. Click here to check out their portfolio. 3one4 Capital An early stage venture capital based in Bengaluru, 3one4 Capital has four unicorns in their portfolio. Their most preferred segments include machine-driven intelligence services for edtech, fintech, SaaS, entertainment, and healthcare. The firm manages over $300 million funds and their most prominent investments include Licious, Open, DarwinBox, Betterplace, Fego.ai, Tracxn, ePlane.ai, Exponent Energy, among others. To pitch to 3one4 Capital, click here. Inflection Point Ventures Founded in 2017 by a handful of entrepreneurs, IPV is an industry agnostic investment firm based in Gurgaon. With around 150 investments, and an estimated revenue range from $10 million to $50 million, IPV has over 3300 CXOs and professionals, family offices, and HNIs. Their most recent investments include Intellemo, Falca, White Inc, Rooba.Finance, MidisimVR, Oorjaa, Orai Robotics, among others. Click here to pitch ideas to IPV. Elevation Capital Founded in 2011 by Ravi Adusumalli, Elevation Capital is based in Gurgaon and focuses on logistics, B2B, and enterprises. With 130 odd investments, the firm is focusing on investing in more early-stage startups building on AI and Analytics. Their most recent investment was Drivetrain in October 2022 raising $15 million along with Pillow, Wishlink, MURF.AI, Cashflo, Rivigo, Detect Technologies. Click here for more information. Matrix Partners India Founded in 2006 by Avnish Bajaj and Rishi Navani, Matrix Partners is a VC firm focusing on the enterprise and consumer market. Since its inception, the firm has invested over 190 times with 83 lead investments and 13 successful exits. Their most recent investment was Saveo in September 2022 that raised INR 283 Million. Other companies include MURF.AI, FPL technologies, SuperOps.ai, Dezerv, Sourcewiz, and many more. Visit their website for more details and pitching ideas. Kalaari Capital Making more than 215 investments, Kalaari Capital has been investing in all technology-centric companies since 2011. With a $160 million venture capital fund and advisory team based in Bengaluru, it was founded by Vani Kola. Their most recent lead investment, ‘Tring’, raised $5 million in October 2022. Other notable investments include Peer Robotics, Deconstruct, Dubverse, ElecTorq Technologies, Dubdub.ai, among others. Click here to check out more of their portfolio. Lightspeed India Partners A multi-stage venture capital firm, Lightspeed India is a sub-domain of Lightspeed Venture Partners in New Delhi and Bengaluru, founded in 2007. The firm has 7 exits with most notable ones being LimeRoad, Setu, and Shuttl. Their most recent investment was ‘GlobalFair’ raising $12 million, crossing a total investment mark of 82 with 38 lead investments. Click here to know more about the firm. Accel India Based in Bengaluru, Accel India is investing in internet technology and software development companies. It was founded by Prashanth Prakash and Subrata Mitra and has made 49 investments with 22 lead investments. Their most recent investment was ‘FitBudd’ that raised $3.4 million. Some of their top investments include Bounce, Trinkerr, SuperShare, Axio Biosolutions, Bytebeam, and Airmeet, among many others. Click here to know more about Accel.","excerpt":"India-based investors looking to invest in AI\/ML and analytics.","categories":["Deep Tech"],"tags":["Startups"],"author_name":"Mohit Pandey","publish_date":"2022-10-29T13:00:00","publication_year":"2022","word_count":965,"keywords":["Go","API","unicorn","AI","RPA","innovation","ViT","analytics","Startups","R","startup"],"extracted_tech_keywords":["AI","analytics","R","Go","API","ViT","RPA","innovation","startup","unicorn"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/vcs-based-in-india-investing-in-ai-startups\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10042113,"title":"Airbnb’s Smart Pricing: All-Inclusive Algorithm Leads To Racial Disparity","content":"A supposed money-making algorithm by vacation rental company Airbnb has led to an increase in revenue divide between black and white hosts. Airbnb had introduced a property ‘Smart Pricing’ algorithm in 2015 to suggest optimal prices and reduce the revenue gap between black and white hosts. As per the US federal laws, all algorithms have to exclude any discriminatory factors, like race, to make recommendations. And so did Airbnb’s algorithm. In many aspects, it was similar to the pricing algorithms used by Amazon and eBay. However, research conducted by the Universities of Harvard, Toronto and Carnegie Mellon suggests otherwise. The research involved collecting and analysing data from randomly selected 9396 properties over 324 zip codes between July 2015 and August 2017. It further examined the algorithm as voluntary adoption by hosts in a quasi-natural experiment. The results and the recommendations, least to say, have been shocking. After adopting the algorithm, the average nightly rate dipped by 5.7 percent, and the average daily revenue increased by 8.6 percent. Before adoption, white hosts were earning $12.16 more than black hosts; this revenue gap closed by 71.3 percent post adoption. The circumstances looked highly desirable, but the algorithm’s effect on hosts who did not adopt the algorithm, mostly black hosts, was the opposite. The gap increased. How does the algorithm work? When a host turns the algorithm on, it adjusts the property’s nightly rate– based on factors like characteristics of the property, season, price of neighbouring properties, and other factors that influence the demand of the property. The algorithm is expected to be more effective than human price setting as it has access to Airbnb servers containing large amounts of data. However, the opacity of the algorithm makes it difficult to access, and thus, the algorithm does not guarantee benefit to the host. Why wasn’t the gap mitigated? The algorithm’s ability to mitigate the revenue gap in Airbnb is dependent on its adoption rate among the black community. Black hosts were 41 percent less likely to adopt the algorithm, and in cases of such low adoption rates, the effect of the algorithm is the opposite. It meant that the algorithm deduced that prices would be skewed towards the white demand curve, generating lower sub-optimal prices for black hosts than their white counterparts. The algorithm finds the same price for black and white properties. Race is not a deciding factor in the pricing. Still, the demand for black properties is more responsive to price changes, particularly downward price correction for black hosts, which led to a greater occupancy. The black community is a minority in both neighbourhood and city levels. Thus, the algorithm is built on data representing the demand curve of white hosts more than the demand curve of black hosts. Therefore, the race-blind algorithm sets prices that lean more towards the optimal cost of the white demand curve. There was also a difference in demand for properties owned by white and black hosts. There can be many reasons for that, but the primary suggested reasons are race bias, education, and less access to other resources by black hosts. The study also indicates that the revenue gap will be difficult to bridge since we are yet to overcome racial discrimination ingrained in society. Recommendations of the study The study maintains that a racially blind algorithm in a society with racism is not the best way to approach business. It says that perhaps the gap cannot be closed by any algorithm. It outright rejects the possibility of hiding the host’s race from the guest, saying that it will lead further to unwanted incidents or people avoiding areas\/neighbourhoods about which they have stereotypes. The study suggests developing an algorithm that considers a person’s race and backgrounds, like socioeconomic situation and other factors, before ‘Smart Pricing’ its property. As an individual’s race is protected under US laws and cannot be used as a determining factor in business decisions, the researchers ask policy-makers to reflect on the law change. The study concludes that a racially blind algorithm for a racially divided society doesn’t bridge any gap– economic or ethical– but only widens it.","excerpt":"Analysis of Airbnb’s data reveals the low adoption rate of its Smart Pricing algorithm among black hosts, leading to racial disparity.","categories":["AI Features"],"tags":[],"author_name":"Meenal Sharma","publish_date":"2021-06-21T13:00:00","publication_year":"2021","word_count":686,"keywords":["Go","AWS","AI","cloud_platforms:AWS","RPA","ML","programming_languages:R","programming_languages:Go","RAG","R"],"extracted_tech_keywords":["AI","ML","RAG","AWS","R","Go","RPA","cloud_platforms:AWS","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/airbnbs-smart-pricing-all-inclusive-algorithm-leads-to-racial-disparity\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10022835,"title":"How To Prepare For A Data Engineering Interview","content":"Lately, the demand for data engineers has surpassed the demand for data scientists. It is challenging to prepare for data engineering interviews due to the lack of readily available resources. The fact that the data engineering field is still evolving and still not clearly defined makes the preparation routine tricky for candidates. While the interview process is quite similar to data science interviews, the focus areas are different. Below, we look at the skills a data engineer must possess and the things to bear in mind when preparing for data engineer interviews. Things To Focus On Programming & coding skills: The data engineering interviews count programming and coding skills, and the ability to implement complex algorithms as critical criteria. The coding bit of the interview is usually focused on the data side and is more practical-driven. Data engineer candidates are expected to use optimal data structures and algorithms to handle potential data issues. Tools to know: The popular open-source libraries a data engineer should know include Spark, Pandas, Hadoop, and Kafka. Knowledge in languages such as Python, Java\/Scala, HTML, CSS and JavaScript will be highly beneficial. Data engineers are also expected to build data visualisations that require an understanding of tools such as React and D3. Some of the other tools to learn for data engineering interviews are scala, spark, hive, pig etc. Upping the SQL game: SQL is one of the essential skills data engineers must be good at. SQL helps in the data processing of big data frameworks such as SparkSQL, pandas, and KafkaSQL. It also helps in translating business queries that end-users can run against your table. It is essential to prepare for SQL-related questions to ace data engineering interviews. Relational databases and data warehouses: A data engineer has to deal with NoSQL databases, graph databases, and more. Data engineers are also expected to design a proper data warehouse based on the use case. Preparing for questions around data warehouses, databases etc., could come in handy in interviews. Data architecture: Data engineers are expected to have a good grip on data architecture and big data systems. While experience gets preference, you can always make a case for yourself by upgrading your skills and understanding the concept thoroughly by making use of good resources available online. System design: System design is the most important and most challenging part of data engineering technical interviews. It involves designing an end-to-end data solution that involves data storage, data processing and data modelling. Explaining the use cases: Explaining the business problem and the solution your team have come up with to tackle it will give the recruiters a chance to size up your expertise and thought process. The candidate can expect to field questions about how you worked with your team to design the solution, the frameworks and tools used, the impact of the resolution, challenges faced, etc. Soft skills: Good communication and problem-solving skills increase candidates’ prospect to land a data engineering job. The passion for your profession and the willingness to take up challenges will leave a good impression on the interviewers. Below are some of the common interview questions for data engineer jobs. What’s the biggest professional challenge you’ve faced, and how did you overcome it?Which frameworks and applications are critical for data engineering?What are the data engineering platforms you are most familiar with, and how did you use them in your previous jobs?Which computer languages are you fluent in?Questions on pipelines, databases, distributed systems and moreDo you have any experience with data modelling?What is your approach towards developing a new analytical product as a data engineer?What were the algorithms and languages you have used on a recent project?How does a data warehouse differ from an operational database?What is a common data engineering maxim you disagree with?","excerpt":"Lately, the demand for data engineers has surpassed the demand for data scientists. It is challenging to prepare for data engineering interviews due to the lack of readily available resources. The fact that the data engineering field is still evolving and still not clearly defined makes the preparation routine tricky for candidates. While the interview […]","categories":["AI Highlights"],"tags":[],"author_name":"Srishti Deoras","publish_date":"2021-03-25T09:00:00","publication_year":"2021","word_count":628,"keywords":["data science","AI","ML","RAG","Python","Kafka","SQL","JavaScript","R","Pandas"],"extracted_tech_keywords":["AI","ML","data science","Pandas","RAG","Kafka","Python","R","SQL","JavaScript"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/how-to-prepare-for-data-engineer-interviews\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":37167,"title":"JAM Trinity — These 3 Keystone Projects By The Govt Can Power India’s AI Revolution","content":"The digital revolution has had a significant impact on India and the JAM trinity — Jan Dhan, Aadhar and Mobile numbers — as succinctly put by Arun Jaitley, Finance Minister, has played a crucial role in it. These three projects have ushered a data revolution, changing the financial landscape of India and ensuring even the poor become digitally mainstream. JAM, has in a way helped reduce leakages and provided the government with more fiscal bandwidth, an IBEF report indicates. Now, a report by UKIBC, enunciates India is building its data protection policies and becoming a digital economy, on the basis of JAM. The report emphasises that India now holds the largest “personal data-cache in the world”, especially with the implementation of GST. Clearly, a positive offshoot of JAM has been in making India a “data-rich country”, allowing India to maximise its data opportunity which we are seeing in the form of data localisation policy and India’s draft Personal Data Protection Bill 2018. It is also a clear indicator of the Government’s stance of using data as an enabler and applying it to India’s socio-economic issues and can pave the foundation for a digital economy. Let’s take the example of China, where data monopolisation has led to improved AI outcomes, India is now on a similar Big Data path, using data cache to revolutionise its businesses, economy and overhaul the public sector. JAM trinity has been hailed for changing the financial landscape of the country, allowing India’s marginalised to access government benefits. On the other hand, it has also led to the emergence of micro-data which will be the growth engine for Indian businesses and its startup ecosystem. The Proposed Personal Data Protection Bill To set the facts clear, the Personal Data Protection Bill, 2018 and the Data Protection Committee’s Report were released in July 2018 and contain the framework and the policymakers’ insight on the protection of personal data in India. The recent Draft e-commerce policy indicates the Government’s thought process on storing data in India. The data localisation wave started with the RBI issuing a rule in April 2018, which mandated all-digital payment providers in India to follow data localisation policy. Under this rule, user data collected had to be localised within 6 months. Now, this data localisation rule is part of the larger proposed Personal Data Protection (PDP) Bill which is expected to be tabled after the 2019 Lok Sabha elections. The directive divided the industry into two camps — leading MNCs (for example WhatsApp that’s behind in implementing these rules) and India’s digital natives (Paytm) that have hailed the move as positive. Since the proposed PDP Bill, there have been other rules, for example, the TRAI circular and the National e-Commerce Policy  which also underscores the ever-increasing importance of data protection and privacy. These policies are also in line with PDP in terms of data localisation and aim to regulate cross-border data flow, while enabling the sharing of anonymised community data, for example, data collected by IoT devices installed in public spaces like traffic signals or automated entry gates, the draft indicates. The Proposed PDP Bill And Data Localisation Are Underpinning The AI Revolution Image Source: Aadhar Besides safeguarding user data and building a thriving Digital Economy in India, data localisation mandated under the new policy can help India strengthen its AI capabilities, and even help in protecting the nation’s security. The data localisation policy provides the right impetus to build a foundation for data hubs, digital innovation which can further spur AI development. The growth of AI is dependent on data, and more the data, better the results. Hence, access to data can lead to the development of innovative solutions: Beefing up digital infrastructure: Data localisation can pave the way for a strong digital infrastructure which can further the development of a strong AI startup ecosystem. In the competitive data-led economy, digital infrastructure will give India a strategic advantage as a digitally-native country and help companies exploit data to build solutions. Case in point, Chinese companies — for example, the BAT trinity have profited from the country’s iron-clad data localisation law that mandates that all the data such as “personal information” and “important data” collected or produced has to be stored locally. Data aggregation is key to a flourishing ecosystem: Companies that are sitting on petabytes of data have a clear advantage in the AI game but data localisation can give smaller players a chance to compete with and build innovative solutions. Even in the public sector, breaking down data silos can help in developing integrated datasets which can be further used in public services. Promote AI Research: One of the key mandates from the Indian Government is that a bulk of services have to be centred around AI, which means that premier educational institutions will work towards aligning research around social and economic goals and forge deeper partnerships with the industry and startup ecosystem. Outlook The Government’s recent directive for localising data storage will bolster digital infrastructure, create jobs and is aligned with core components of a digital economy — deliver public services digitally and also promote digital literacy. From setting up data centres to server farms, the move will power a new wave of boosting physical infrastructure in the country. In addition to this, this can also substantially improve last mile connectivity in rural areas and spread digital literacy.","excerpt":"The digital revolution has had a significant impact on India and the JAM trinity — Jan Dhan, Aadhar and Mobile numbers — as succinctly put by Arun Jaitley, Finance Minister, has played a crucial role in it. These three projects have ushered a data revolution, changing the financial landscape of India and ensuring even the […]","categories":["AI Features"],"tags":["Personal Data Protection Bill"],"author_name":"Richa Bhatia","publish_date":"2019-04-01T12:47:16","publication_year":"2019","word_count":895,"keywords":["big data","Go","Personal Data Protection Bill","AI","innovation","Git","RAG","Aim","ViT","R","startup"],"extracted_tech_keywords":["AI","Aim","RAG","R","Go","Git","big data","ViT","innovation","startup"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/jam-trinity-these-3-keystone-projects-by-the-govt-can-power-indias-ai-revolution\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10163815,"title":"Sonoco Opens Hyderabad GCC With $10 Mn Initial Investment for Global IT Operations","content":"Sonoco Products Company, a global leader in sustainable packaging, has launched a Performance Hub in Hyderabad, marking a major step in its digital transformation strategy. With an initial investment of up to $10 million, the facility is set to enhance Sonoco’s technology leadership and create new opportunities for India’s growing tech talent. The Hyderabad hub will serve as a key pillar in Sonoco’s global operations, leveraging the city’s technology ecosystem to drive innovation in packaging design and manufacturing. Equipped with modern training environments, collaborative spaces, and advanced IT infrastructure, the facility will focus on data analytics, automation, and emerging technologies to improve efficiency and sustainability. Sonoco plans to hire 300 technologists over the next 12-18 months, with a goal of reaching 500 employees in three years. The hub will provide local professionals with opportunities to contribute to Sonoco’s global initiatives, bridging India’s technology expertise with the company’s worldwide operations. Howard Coker, president and CEO, Sonoco, emphasised the impact of this investment: “The Hyderabad Performance Hub is a major leap forward in our digital transformation. By investing in innovation, we not only enhance our ability to deliver more efficient and sustainable packaging solutions but also provide local talent with unparalleled opportunities to lead advancements in technology and manufacturing.” Rajeev Ankireddypalli, VP & CIO, Sonoco, highlighted Hyderabad’s role in this initiative: “Hyderabad, with its dynamic technology ecosystem, offers the perfect environment for our new hub. This facility will provide a platform for local professionals to engage in transformative projects and directly contribute to global initiatives, advancing Sonoco’s technology capabilities and sustainability goals.” Leading the hub is Arvind Chittora, managing director, who brings over 23 years of experience at Sonoco. “I am truly excited about the opportunity to build the workforce of the future for Sonoco in Hyderabad and accelerate our digital and technology roadmap. This marks a significant step in strengthening our global capabilities, fostering innovation, and driving business impact,” he said. Beyond job creation, the Performance Hub will contribute to the local economy by partnering with academic institutions, offering internships, and fostering skill development among young professionals. This move positions Hyderabad as a crucial part of Sonoco’s global innovation ecosystem, enabling Indian talent to lead advancements with worldwide impact.","excerpt":"Sonoco plans to hire 300 technologists over the next 12-18 months, with a goal of reaching 500 employees in three years.","categories":["AI News"],"tags":["GCC","GCC india"],"author_name":"Mohit Pandey","publish_date":"2025-02-17T11:49:29","publication_year":"2025","word_count":368,"keywords":["Go","GCC","programming_languages:R","AI","innovation","digital transformation","Git","RAG","automation","analytics","GCC india","R"],"extracted_tech_keywords":["AI","analytics","RAG","R","Go","Git","digital transformation","automation","innovation","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/sonoco-opens-hyderabad-gcc-with-10-mn-initial-investment-for-global-it-operations\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10054174,"title":"Top Indian University-Industry Partnerships For AI In 2021","content":"Artificial intelligence (AI) and machine learning (ML) are reshaping virtually every facet of human endeavour. However, these technologies must be nurtured by talent and a creative atmosphere to attain their full potential. Leading artificial intelligence centres in India’s top institutions are listed below. IIT Madras join up with Taylor & Francis Group Taylor & Francis Group has teamed with the IIT Madras Robert Bosch Centre for Data Science and AI to accelerate research. Dr Gagandeep Singh, Senior Publisher, CRC Press, stated that “We hope that this co-branded organic commissioning project from this internationally renowned institute will result in a mutually beneficial exchange of ideas, issues, and best practices; in scholarly publication.” IISc with KMBL In partnership with Kotak Mahindra Bank Limited, the Indian Institute of Science (IISc) has announced intentions to establish a state-of-the-art AI and AL (AI-ML) Center (KMBL). The new facility will be called the Kotak-IISc AI-ML Centre and will be located on the IISc campus in Bangalore. It will provide undergraduate, graduate, and short-term courses in AI, ML, deep learning, fintech, and reinforcement learning, among other subjects. Additionally, the centre intends to foster research and innovation in AI and machine learning and establish a talent pool from around the country capable of providing cutting-edge solutions to satisfy the industry’s emerging and future requirements. IIT Roorkee with the Mehta Family Foundation IIT Roorkee and the Mehta Family Foundation (MFF) in the United States of America collaborated to establish the Mehta Family School of Data Science and Artificial Intelligence. MFF and IITR have struck an agreement to establish a new Mehta Family School of Data Science and Artificial Intelligence. This initiative is MFF’s fourth new school in India, advancing the organisation’s objective to create new degree programmes in data science, artificial intelligence, and bioengineering that will result in discoveries, new companies, and new jobs. The school’s news, which will be erected on IITR’s main campus, drew extensive global media interest. IIT Delhi with IHFC I-Hub Foundation for Cobotics (IHFC), the Technology Innovation Hub (TIH) of the Indian Institute of Technology Delhi (IIT Delhi), and iHub Anubhuti, the TIH of the Indraprastha Institute of Information Technology Delhi (IIITD), recently inked an agreement to establish India’s first Medical Cobotics Center (MCC) at IIIT-Delhi. Both IIT Delhi and IIIT Delhi have strong engineering backgrounds and significant ties to researchers and medical professionals at various Indian universities, including AIIMS and others. The two centres are dedicated to advancing medical robotics\/cobotics, digital health, sensing, and computing technologies used in robotic-assisted operations, training, and medical procedures. TMU Moradabad collaborates with NVIDIA Teerthanker Mahaveer University (TMU) in Moradabad has inked a memorandum of understanding with NVIDIA Corporation to deliver artificial intelligence (AI) and deep learning (DL) to a new generation of students and developers. NVIDIA and TMU Moradabad want to collaborate to build a Centre of Excellence for AI Research on campus. The TMU Moradabad campus’s NVIDIA Jetson AI\/DL Embedded Lab will prepare engineers, researchers, and data scientists for potential AI and deep learning robotics jobs. According to Deepu Talla, NVIDIA South Asia’s Vice President and General Manager of Embedded and Edge Computing, “The AI revolution is being fueled by NVIDIA at research centres throughout the world. TMU Moradabad’s NVIDIA Jetson AI\/DL Embedded Lab has the potential to help India catch up to the competition in the field of supercomputing.” Axis Bank-Manipal School of Data Analytics In conjunction with Axis Bank, Manipal Global announces the Axis Bank-Manipal School of Data Analytics, the country’s first job-assured online programme in data science and analytics. The school of data analytics offers a three-month online degree programme with a two-week practical component. Beginning this month, the first group will receive training in the core ideas and techniques of data science and analytics, big data, and ML using Python. The curriculum is targeted at recent graduates and individuals with up to three years of experience in technology\/analytics-related fields.","excerpt":"The following is a list of the leading research laboratories in Indian institutions that are advancing and innovating in artificial intelligence.","categories":["AI Trends"],"tags":["AI (Artificial Intelligence)","Data Analytics","Data Science","IIT","Machine Learning","Reinforcement Learning"],"author_name":"Dr. Nivash Jeevanandam","publish_date":"2021-11-24T18:00:00","publication_year":"2021","word_count":647,"keywords":["data science","artificial intelligence","Reinforcement Learning","machine learning","AI","AI (Artificial Intelligence)","ML","Machine Learning","Python","deep learning","edge computing","analytics","Data Analytics","Data Science","R","IIT"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","data science","analytics","edge computing","Python","R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/indian-university-industry-partnership\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10163321,"title":"Indian Companies Bullish on Long-Term AI Investments; 76% Surveyed Firms Achieved ROI-Driven Results: IBM Study","content":"A research study commissioned by computing giant IBM revealed on Wednesday that most Indian companies made significant progress in executing their AI strategies last year. The study surveyed over 2,000 IT decision-makers worldwide, 224 of whom were from India. Among the ones surveyed in the country, 87% reported progress in their AI strategy, and 76% revealed that they had achieved results driven by return on investment (ROI). About 89% of the Indian respondents said their companies have started more than 10 AI pilots in the last year. Furthermore, 93% of the Indian respondents said they will increase their AI investments this year. The majority of them reported looking for open-source solutions, and 48% revealed that more than half of the AI solutions being used are based on open-source technologies. Among the companies that are yet to achieve ROI-driven results from AI projects, 33% expect to see savings within the next 12 months, and all of them believe they will achieve a positive ROI within three years. Only 1% of the surveyed Indian respondents revealed that their AI strategy had not made progress. The surveyed Indian companies revealed that they are focusing their AI investments this year on IT operations, software coding, and data quality management. The detailed report from IBM, outlining the findings from multiple countries worldwide, can be found here. Indian Prime Minister Narendra Modi on Tuesday spoke at the AI Action Summit in Paris, where he said that India leads in AI adoption. According to him, India has one of the world’s largest AI talent pools. “India is building its own large language model. Considering our diversity, we also have a unique public-private partnership model for pooling resources like computing power,” he added. In this year’s Budget, the government allocated ₹2,000 crore for the IndiaAI mission – nearly a fifth of the scheme’s ₹10,370 crore announced last year. Moreover, AIM recently reported that Indian companies and startups are increasingly using AI-enabled coding tools. Although companies were initially hesitant to adopt such tools, they are now leaning toward them due to their benefits. Furthermore, a recent survey by GitHub revealed that 56% of Indian developers are using AI tools to help them boost their chances for employment owing to the skills they develop. Moreover, around 80% of them believe AI tools have improved code quality.","excerpt":"Among the companies that are yet to achieve ROI-driven results from AI projects, 33% expect to see savings within the next 12 months.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","IBM","India","Quarterly Earnings"],"author_name":"Supreeth Koundinya","publish_date":"2025-02-12T13:31:48","publication_year":"2025","word_count":386,"keywords":["Go","programming_languages:R","AI","R","programming_languages:Go","Git","Aim","IBM","data quality","GitHub","Quarterly Earnings","India","AI (Artificial Intelligence)","startup"],"extracted_tech_keywords":["AI","Aim","R","Go","Git","GitHub","data quality","startup","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/indian-companies-bullish-on-long-term-ai-investments-76-surveyed-firms-achieved-roi-driven-results-ibm-study\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":67579,"title":"Meet The MachineHack Champions Who Cracked The ‘E-commerce Price Prediction’ Hackathon","content":"MachineHack successfully concluded its eighth instalment of the weekend hackathon series last Monday. The E-commerce Price Prediction hackathon was greatly welcomed by data science enthusiasts with over 400 registrations and active participation from close to 200 practitioners. Out of the 189 competitors, three topped our leaderboard. In this article, we will introduce you to the winners and describe the approach they took to solve the problem. #1| Devrup Banerjee Although Devrup learnt python just out of the sheer need to automate the routine work and gather data at scale, his real enthusiasm and passion for data science sprouted in his second year of MBA at Great Lakes Institute of Management, Gurgaon, while he was attending his marketing and retail analytics class. He realised that the real motivation behind learning all these algorithms was not about enhancing accuracy but to tell your client by how much you can promise to increase their bottom-line if they were to follow your exact given path. The subject changed his life. “My roommate, who was also equally inspired, and I used to have sleepless nights just going through the 25 lacs dataset given as a final project with our rickety computers to generate actionable insights. To better the bottom-line percentage, that’s what inspired me into analytics.“ – He said His team has won many competitions at MBA level, won at IIT Kanpur MRA tournament while finishing as runners up at IIM Kashipur’s case study on analytics. He is currently trying to deep dive into data science to better his analytical skills so that if someone gives him a dataset in future, he can be both a business analyst and a data scientist. Approach To Solving The Problem Devrup explains his approach as follows: The real problem in the ecommerce dataset was the product and brand names. There were about 300 brands which were not present in the train dataset but were present in the test dataset. So, One-Hot-Encoding or any other form of encoding was ruled out. I considered TF-IDF with 2 ngrams, but it still wasn’t giving me the results I expected. The brand column, along with the item category which if cleaned properly contained the most information about any product price. Thus, I decided to target and encode them and take care of those 300 odd brands separately. A class column was formed, based on the average price levels of each brand in each category, which contained most of the information from the brand name and item category. A bit of MBA knowledge was also leveraged to identify the end consumer of all the products given, as the first thing we were taught in our MBA was price should always be decided after identifying the target market. And it turned out to be the game-changers. MachineHack has been a huge source of inspiration and learning, along with Analytics India Magazine which keeps us up to date on the latest happening from around the world on analytics. They have established themselves as the domain leaders, and I won’t be surprised if they are soon known as the Indian Kaggle. Get the complete code here. #2| Mrutyunjaya Rath Besides being a graduate in Mechanical Engineering, data and coding have always excited Mrutyunjaya. He started his data science journey by doing a course with upGrad in association with IIIT-B. Although he found it to be a little difficult at the beginning, the continuous practice gave him the confidence and skills to pursue the path. He spends most of his time participating in hackathons and acquiring new skills by learning new techniques. “You will succeed in some of the approaches, and some will fail miserably, and that is something which is exciting about data science”- he said. Approach To Solving The Problem Mrutyunjaya explains his approach briefly as follows. My approach to this problem was very simple. I believe, for any data science problem, the most important thing is EDA. So, when I started plotting features, I noticed the target variable ‘Selling Price’ was highly left-skewed, due to which the output also was getting left-skewed. So, I decided to apply a logarithmic transformation to normalise the target variable. After that, I tried to create new features like datetime features, group_by of categorical variables for statistical features and categorical variables were handled through label encoder. After pre-processing, I went on to build 3 models, using XGBoost, LGBM and CatBoost. I fine-tuned the models which were giving me a good cross-validation score through early stopping. And in the end, I blended the result of all 3 models to get me my final score. “MachineHack is one of the best platforms for any data science enthusiast. Not only you can compete here, but also you get to know your participants which leads to an increase in your connections, and you get to talk and interact with like-minded people. I would like to thank MachineHack and Analytics India Magazine for organising this hackathon and also for contribution towards the data science and machine learning community. I would also like to congratulate my fellow participants who managed to put a score on the leaderboard.” – he shared his opinion. Get the complete code here. #3| Shravan Kumar Shravan Kumar is a senior data scientist at Novartis, Hyderabad. Though Shravan has been working in the analytics field for some time, he always had a keen interest in the predictive analytics domain which he wished to explore. He started with MOOC courses on platforms like Coursera, edX, Udacity etc., and acquired the skillsets for his area of interest. Having learned the essentials, Shravan’s next focus was on perfecting his skills through online competitions. He is an active participant of hackathons conducted by MachineHack, Kaggle and other platforms. Approach To Solving The Problem He explains his approach briefly as follows: Pre-processing steps: Created date features for e.g., Year, Month, Day, DayofWeek, DayofYear, etc.Converted the selling price into log transformationFrequency encoding on each of the categorical variables at different combinationsMean targeting encoding at each of the categorical variables on ‘Product_Brand’, ‘Item_Category’, ‘Subcategory_1’,   ‘Subcategory_2’ Model building steps: I applied CatBoost and LightGBM with different seed values and finally applied a harmonic mean between these two outputsMy model looked stable with both the 30% and 100% public datasetsHere, I had to train the model on the entire dataset without the train test split ratio, as the leaderboard was not reflecting a good score. “MachineHack is a great learning platform for all the aspiring and current data scientists who get a chance to solve some real-world business problems. The articles and blogs provided by Analytics India Magazine are very helpful and keep all our industry-relevant people with the most updated news in the data science and analytics industry. Truly MachineHack is one of the best hackathon organisers and data science knowledge portals in India.” -he shared his experience. Get the complete code here. Check out new hackathons here.","excerpt":"MachineHack successfully concluded its eighth instalment of the weekend hackathon series last Monday. The E-commerce Price Prediction hackathon was greatly welcomed by data science enthusiasts with over 400 registrations and active participation from close to 200 practitioners. Out of the 189 competitors, three topped our leaderboard. In this article, we will introduce you to the […]","categories":["Deep Tech"],"tags":["Hackathon Winners","Hackathons","Machinehack","Weekend Hackathon"],"author_name":"Amal Nair","publish_date":"2020-06-18T10:00:00","publication_year":"2020","word_count":1149,"keywords":["data science","machine learning","Weekend Hackathon","TPU","AI","Machinehack","Hackathons","RAG","XGBoost","analytics","CatBoost","LightGBM","Hackathon Winners","predictive analytics"],"extracted_tech_keywords":["AI","machine learning","data science","analytics","XGBoost","LightGBM","CatBoost","RAG","predictive analytics","TPU"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/meet-the-machinehack-champions-who-cracked-the-e-commerce-price-prediction-hackathon\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":58467,"title":"Why Python May Lose Its Charm Over Time?","content":"Founded by developer Guido Van Rossum, Python is a high level, interpreted, and dynamic programming language used for the development of various applications such as web development, scientific computing, etc. Python, since its inception in 1991, has been that one language that is easy to learn and powerful to use; and therefore has been an excellent choice for beginners as well as big organisations like Google and Instagram. The language was explicitly designed to emphasise extensively on code readability. Because of its simple and expressive syntax, Python enables users to express concepts without writing additional code, and also help developers to accelerate their software development. Its simplicity also allows programmers to integrate Python seamlessly with other programming languages. Python has a broad set of powerful libraries that are used for neural networks, natural language toolkit, for interacting with windows, mathematical calculations, for TDD, GUI, web scraping, imaging processing, database management, game development, and data analysis. While that says a lot about the language, it is not perfect. Python also has downsides compared to other programming languages like Julia, Swift, Java etc. There were also several questions raised about its performance and the quirks in its design, which might push this language to the edge. Why Might Python Lose Its Popularity Over The Years? Python is the most inclusive language, and programmers prefer this language over others. However, despite the popularity Python still couldn’t make a place in some of the computing fields like enterprise development, and therefore isn’t used to solve enterprise solutions. Along with that, some other restrictions are in the areas of performance and security. To explain further, the following are some of the fundamental limitations: Speed Limitations Speed has always been the key focus of a programmer and will always be an essential criterion to work on. Python is an interpreted language and therefore is slower in execution than other languages like C\/C++, Java, or newer languages like Julia. For Python, the execution takes place with the help of an interpreter instead of a compiler, and that causes it to slow down. Considering it is a high-level language, it also isn’t closer to the hardware. And because of the slow speed, Python is not considered to be a comfortable language to work while writing production level-code. Weak In Mobile Computing Although Python is considered to be strong in desktop and server platforms, it is not considered suitable for mobile development because of its weak language processing for mobile computing. So you can witness Python as a first server-side language, however rarely seen on the client-side. And therefore rarely used to build mobile development. In recent days, a lot of advancements have been made in Python — adding libraries like Kivy and Beeware to serve — in order to improve its performance in the mobile application development world. Although these libraries are pretty useful, these still can’t perform as good as other competitor languages like Java, Kotlin, Swift, Javascript, etc., which are ruling the mobile and web app development industry. Nowadays, to create an android application, developers use Java or Kotlin, and for the iOS route, developers pick Swift as their go-to-language. Memory Consumption App development and computing programs are memory-intensive tasks, and due to high memory consumption by Python, it isn’t considered a right choice for memory-intensive jobs. This language has automatic garbage collection when objects go out of scope. Therefore it intends to remove a lot of the complexity of memory management that languages like C and C++ involve. Python’s high memory consumption is also due to the flexibility of data types. One method that can be used for looking into memory leaks in your code is to use Heapy, a debugging tool that can be used to find out, which objects are holding the most memory. So, before writing Python code for highly memory-intensive tasks, developers should gain a better understanding of the space efficiency of the code and the underlying packages used. Database Access Python is supposed to be a robust programming language, which gives less stress to developers while coding. But one of the critical limitations of Python is with the database access. It is therefore considered highly insecure. And, when compared to JDBC (Java Database Connectivity) and ODBC (Open Database Connectivity), Python’s lacks GUI and its tools, and its weak database access layer acts as a barrier for organisations that are looking for a smooth interaction with their data. Another important challenge organisations face with Python is its incompatibility issue with older versions. To make it compatible with the present version of Python, organisations need to upgrade their tech stack, which isn’t always a good idea. Slow At Runtime Python’s design has been questioned by several experts, where they believe that the language requires more testing. Also, as the language has been dynamically typed, that means one doesn’t need to declare the type of variable while writing the code. Although it makes it easy for programmers during coding, it shows up errors during the runtime, which works as the biggest drawback of Python. And, that is one of the reasons why statically typed languages like C, C++, and Java execute faster at runtime. One of the main reasons for the slowness of this language is the presence of GIL (Global Interpreter Lock). GIL allows only one thread to execute at a time, and therefore the multithreaded CPU-bound programs may be slower than single-threaded ones. In that case, developers need to implement multiprocessing programs instead of multithreaded ones to improve the speed. Difficult To Blend In Python is a very addictive language, and Python lovers get so accustomed with its features and extensive libraries that it becomes difficult for them to learn or work on other programming languages. According to Python experts, declaring of cast “values” or variable “types”, and adding syntax, curly braces or semicolons could be considered as a mundane task. Being a Pythonista, developers are supposed to love Python because of its simplicity, popularity, and powerfulness. Python is an easy language to code when compared to other languages like C++ and Java, and is also considered more close to human language. Outlook Learning python basics will make a beginner understand the core of programming with the easier python codes. Also, Python is relatively easy to learn, and therefore heavily used in teaching programming. However, with so many drawbacks in hand, enterprise-level implementation is still far away. And that’s where other languages like Java, C, C++, JavaScript, Julia, Swift, etc. shine with more productivity in their respective fields. The main disadvantages of Python are its slowness, its weakness in mobile application development, and its less popularity in the enterprise development sector. Additionally, with the advent of AI and ML, nowadays, enterprises are swiftly moving towards AI- and ML-based web applications to better serve their customers. Nonetheless, it is to be believed that every programming language has its forte and key aspects that can be used by developers. Similarly, Python has its areas where it is being used extensively.","excerpt":"Founded by developer Guido Van Rossum, Python is a high level, interpreted, and dynamic programming language used for the development of various applications such as web development, scientific computing, etc. Python, since its inception in 1991, has been that one language that is easy to learn and powerful to use; and therefore has been an […]","categories":["Deep Tech"],"tags":["Julia Language","julia scientific programming","Python","python advantages","Python Programming","python vs r","Why is Python so Popular"],"author_name":"Sejuti Das","publish_date":"2020-03-12T12:00:00","publication_year":"2020","word_count":1167,"keywords":["python advantages","Go","AI","neural network","ML","Julia","Why is Python so Popular","Python","Julia Language","C++","python vs r","JavaScript","Python Programming","R","Java","julia scientific programming"],"extracted_tech_keywords":["AI","ML","neural network","Python","R","JavaScript","Go","Java","Julia","C++"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/why-python-may-lose-its-charm-over-time\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10086215,"title":"Indian Devs Can Now Unleash AI Power with this New NVIDIA Beast","content":"It is good news for the developers now that the GPU manufacturing giant NVIDIA’s  “Jetson Orin NX 16 GB” is available for purchase in India that can be used to improve Edge AI performance. Designed for small robots, embedded devices, and autonomous machines like drones and handheld devices, it will be able to develop solutions using complex AI models in natural language understanding, 3D perception, and multi-sensor fusion. The Jetson Orin NX is the smallest module of its kind and offers up to 100 TOPS of AI performance with the ability to adjust power usage between 10 W and 25 W. It’s also claimed to be three times faster than the NVIDIA Jetson AGX Xavier and five times faster than the NVIDIA Jetson Xavier NX. The module includes features like NVIDIA’s Ampere architecture GPU, advanced deep learning and vision technology, high-speed input\/output, and quick memory access to support multiple AI applications. NVIDIA showed better results on computer vision tests using NVIDIA JetPack 5.1. The tests included running some dense INT8 and FP16 pre-trained models from NGC, and then comparing the results to the same models run on Jetson Xavier NX. NVIDIA has announced improved results in computer vision tests using NVIDIA JetPack 5.1. The tests compared the performance of pre-trained models run on Jetson Xavier NX and Jetson Orin NX. The results showed that Jetson Orin NX had a 2.1 times performance increase compared to Jetson Xavier NX, with the potential to reach 3.1 times with future software optimizations. Additionally, Jetson Orin NX supports sparsity to enhance deep learning network performance. NVIDIA has also released JetPack 5.1 to support the Jetson Orin NX 16 GB and the latest CUDA-X stack. NVIDIA announced the release of their unified quantum computing platform, “QODA” (Quantum Optimized Device Architecture), in July 2022 following the success of CUDA. QODA is intended to speed up advances in quantum research and development in AI, HPC, and other areas.","excerpt":"The 16GB system-on-module is smaller and can be used in lower-power autonomous machines.","categories":["AI News"],"tags":["cuda","NVIDIA","NVIDIA Career"],"author_name":"Shritama Saha","publish_date":"2023-01-31T12:21:12","publication_year":"2023","word_count":321,"keywords":["CUDA","Go","NVIDIA Career","cuda","TPU","AI","computer vision","Aim","deep learning","edge AI","NVIDIA","R"],"extracted_tech_keywords":["AI","deep learning","computer vision","Aim","edge AI","TPU","CUDA","R","Go","CUDA"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/indian-devs-can-now-unleash-ai-power-with-this-new-nvidia-beast\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":12176,"title":"How Wockhardt is making data analytics an everyday practice in healthcare","content":"While the big data revolution in healthcare is undoubtedly in early stages, the wheels are definitely turning. Healthcare providers have moved from reams of paperwork that dominated every hospital visit to a cloud-based Electronic Health Records (EHR). And now there is a speedy transition to imbibe a data- oriented culture across healthcare organizations with a concerted efforts to bring data-led insights into hospitals. One such organization is Mumbai-headquartered Wockhardt Hospitals Ltd, one of India’s leading super-specialty care provider and the man leading the charge of building a data-driven culture is Sumit Singh, Chief Information Officer, Wockhardt Hospitals Ltd who is driving a companywide adoption of data analytics. Sumit Singh, CIO Wockhardt Hospitals Ltd is leading a data-driven change in the healthcare organization As part of his mandate – tapping into terabytes of data and getting value from it, Singh got one of the most forward looking BI and Data Analytics company onboard, Qlik. “We seriously started looking at this BI and analytics space two years ago. The whole idea was how we could leverage additional benefits from this pool of disconnected data that we were sitting on. At any point of time, 95% of this information is already there in the system, what we want to do is make more meaningful decisions from the same dataset by culling different viewpoints from it,” shared Singh, who will be overseeing the transformation over a period of time across nine centres in India. Speaking to Analytics India Magazine, Singh outlined four major points why he chose the BI and analytics solutions for leading a data-driven approach. The pointers can very well serve as benchmark for other organizations for onboarding an analytic vendor. Visualization: According to Singh, data is ongoing, it has a life of its own and it needs to grow (support for updating data). It also needs visibility and that’s where the visualization tool helps, in presenting snapshots that are further utilized to for better procurement and optimizing costs across services. Scalable Analytics Engine: Primarily, the analytic engine enables flexibility combined with minimal expense during scaling deployments.  The end job of an analytic engines is speeding up the time to value for resources. “At the end of the day, it is the analytics engine that is producing the output and we wanted to capitalize on its strength without having to compromise on data abstraction,” he said. Customizability: Another important criteria in narrowing down on analytic vendor was customizability. “What will happen if I add something specifically or choose certain constraints,” he shared. Increased quality of time:  If something is not done on time, it leads to a waste of resources. “You are able to make a lot more quality decisions in a quick timeframe but also identifying new growth opportunities,” he shared. It’s the healthcare organization’s first foray into data analytics and it was marked by a “shift in thinking approach”.  Singh foresees a companywide adoption of analytics where each decision is driven by data. “We are a large organization and we will take time to catch up. Over a period of time, I would like to see a lot more adoption and benefits from the analytics platform, especially in MIS in finance which is a labour intensive job. The amount of time spent in creating it outweighs the time spent in understanding it and I would like to flip that totally,” he said. Simplifying healthcare with integrated data analytics solution According to Souma Das, Managing Director at Qlik India, “Data generated in the healthcare sector is growing at an exponential scale and it is important that healthcare organizations implement a business intelligence system that can leverage this data, is user-centric and provides the ability to provide quick access, analysis and dissemination of information. The main drivers for having a proven analytics partner is to deliver better patient care and drive operational efficiencies and growth,” he said. The BI and analytics company in partnership with Exponentia DataLabs assessed how the data was disconnected and came from disparate sources – 1) the existing Hospital Information System (HIS), 2) SAP financial module and 3) large volumes of data also flowed in from excel spreadsheets. Post the slicing and dicing, the team generated strategic insights and analysis for Wockhardt Hospitals management to reevaluate: Revenue analysis and cost optimization measures MIS for the finance team Pricing analysis and cost audits for related finance functions It’s been 6 months since Wockhardt Hospitals Ltd implemented BI and analytics solutions and the results can be seen across three main areas Procurement and supply:  Qlik’s integrated analytics solutions helps in meeting procurement challenges, improves operational efficiency and reduce costs by ensuring timely procurement. Cost of service: Identify and analyze the full cost of every service provided by the healthcare company. Marketing:  A lot of data is being created around the digital space and analytics offer a holistic view.  “We can gauge areas that have had maximum impact and where more resources are required for a better marketing experience,” he shared. Slow adoption of analytics in healthcare According to reports, the healthcare sector has been lagging behind industries such as retail, banking and industrial when it comes to adoption of data analytics. What’s the reason for putting data and analytics on the back burner? “In healthcare, for most practitioners the focus is on ensuring the patient leaves the hospital happy and is well attended. Secondly, another barrier for slow adoption is there are a lot of disparate sources of data, such as clinical data which poses a huge challenge. Concerns about data security and privacy also play a huge role in slowing the adoption of healthcare analytics,” expressed Singh. But he stays optimistic about the future of analytics in healthcare. “Even though we are still in the early stage of using and analyzing data I am sure we will be seeing significant gains in the future,” he noted.","excerpt":"While the big data revolution in healthcare is undoubtedly in early stages, the wheels are definitely turning. Healthcare providers have moved from reams of paperwork that dominated every hospital visit to a cloud-based Electronic Health Records (EHR). And now there is a speedy transition to imbibe a data- oriented culture across healthcare organizations with a […]","categories":["IT Services"],"tags":["healthcare analytics","Qlik analytics India"],"author_name":"Richa Bhatia","publish_date":"2017-01-12T05:26:08","publication_year":"2017","word_count":977,"keywords":["Go","API","TPU","AI","Scala","Git","healthcare analytics","RAG","Ray","Qlik analytics India","analytics","R"],"extracted_tech_keywords":["AI","analytics","Ray","RAG","TPU","R","Go","Scala","Git","API"],"url":"https:\/\/analyticsindiamag.com\/it-services\/wockhardt-making-data-analytics-everyday-practice-healthcare\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10124859,"title":"AI Raises the Bar for Deep Skills in Jobs","content":"In a recent post on X, CRED founder Kunal Shah said that AI had raised the bar of ‘deep skill’, a term that refers to the advanced, specialised skills that are in high demand due to AI’s influence, just like the internet did in the past. It’s interesting to note how this transformation is moving faster than ever. Defining Deep Skill Shah is probably right when he says AI has raised the bar for deep skill. The rise of AI is making deep, specialised skills more valuable while also creating new categories of deep skills around AI technologies. Deep skills are considered high-level capabilities that require extensive training and experience to develop. They are distinct from broad skills, which are more general and transferable across different domains. They are often domain-specific and not easily interchangeable between industries or skill sets. For example, a C++ programmer cannot easily transition to a nanoparticle scientist role. Need Much More Than Skill According to a recent IBM survey, executives believe 40% of their workforce will need to reskill over the next three years due to AI and automation initiatives. “That translates to 1.4 billion of the 3.4 billion people in the global workforce, according to World Bank statistics,” the study’s authors stated. Until about 40 years ago, it was commonly estimated that skills had a ‘half-life’ of at least ten years before needing refreshment. Today, that half-life is around four years. So, it’s no surprise that IBM’s AI chief, Matthew Candy, stated that you may no longer need a computer science degree if you want to work in technology. Candy believes that in the age of AI, ‘soft talents’ such as creativity, problem-solving, and adaptability would be more valuable than technical ones. AI is Reshaping the Skill Market Those who master the art of developing workflows in which humans and AI complement each other’s capabilities will generate unmatched value. This “symbiotic intelligence”, which is able to recognise when human intelligence supersedes AI and vice versa, has the potential to become the new standard of excellence. Generative AI systems like language models can mimic certain aspects of human skills, but they lack true understanding and cannot fully replace human expertise and judgement. There are also concerns that over-reliance on AI could lead to a decline in deep, hands-on practice and mastery of skills. Most jobs today are about knowing how things work in a company, i.e processes, who does what, team dynamics, past issues in emails etc. All this will go to AI. AI as a Skill Facilitator AI is not just a disruptor, it can also be a facilitator in skill transitions and career exploration. A pilot project by the World Economic Forum, in collaboration with Unilever, Walmart, Accenture, and SkyHive, used AI to map individuals’ skills and match them to future career roles, showcasing the practical applications of AI in the job market. The study revealed a promising aspect of AI—it can help us discover our true potential. Often, we underestimate our skill sets due to inherent bias. However, when AI analysed the skills, the number of detected skills increased by more than three times. This means that AI can open up hitherto untapped job alternatives, offering a beacon of hope for career exploration and growth. The study also discovered that it would only take six months to reskill people for new responsibilities in entirely different areas. AI-powered tools can help identify individuals’ existing skills and match them to emerging job roles, enabling seamless career transitions and reskilling. It is also evident that with AI, the bar for deep skill will only increase.","excerpt":"Until about 40 years ago, it was commonly estimated that skills had a ‘half-life’ of at least ten years before needing refreshment. Today, that figure stands at four years.","categories":["AI Trends"],"tags":["AI (Artificial Intelligence)","ai and jobs"],"author_name":"Anshul Vipat","publish_date":"2024-06-26T16:13:51","publication_year":"2024","word_count":600,"keywords":["Go","API","programming_languages:R","AI","ML","automation","C++","ViT","generative AI","AI (Artificial Intelligence)","R","ai and jobs"],"extracted_tech_keywords":["AI","ML","generative AI","R","Go","C++","API","ViT","automation","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/ai-raises-the-bar-for-deep-skills-in-jobs\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":25691,"title":"Govt directive asks ISRO to share Lithium-ion batteries manufacturing technology with private players","content":"In a recent directive by the Indian Government, it has asked Indian Space Research Organization (ISRO) to share the technology for mass production of Lithium-ion batteries with local automobile manufacturers manufacturing electric vehicles. Vikram Sarabhai Space Centre (VSSC) under ISRO has successfully developed lithium-ion batteries for use in electric vehicles and government has requested the esteemed organization to make the information accessible even to private players so as to boost the production further. The indigenously developed batteries have already cleared multiple rounds of test and are being used to power satellites and other space missions. Major automobile giants, battery manufacturers and public undertakings such as Mahindra Renault, Hyundai, Nissan, Tata Motors, High Energy Batteries, BHEL have approached ISRO, who are keen on utilizing the lithium-ion batteries. The current practice includes these automobile manufacturers having to import lithium-ion batteries making the final product expensive. The availability of indigenous development of technology would ensure a major cost cutting hence rolling out cheaper and effective electric vehicles. The end result of this directive is intended at mass production of batteries, hence increasing competition and bringing down the prices further. The experts believe that bulk production of these batteries can lower the prices by up to 80 percent, making it affordable to larger base. The government is also keen on boosting the sale of electric vehicles to combat the problem of air pollution. It is quite disappointing that most of the Indian cities are the most polluted cities worldwide and the government is hoping that adoption of eco-friendly options like lithium battery can address pollution in a larger way. While addressing the meeting, Nitin Gadkari, Union road transport minister said that the cabinet secretary has asked ISRO to create a framework allowing interested private players to avail the technology for mass production at the earliest.","excerpt":"In a recent directive by the Indian Government, it has asked Indian Space Research Organization (ISRO) to share the technology for mass production of Lithium-ion batteries with local automobile manufacturers manufacturing electric vehicles. Vikram Sarabhai Space Centre (VSSC) under ISRO has successfully developed lithium-ion batteries for use in electric vehicles and government has requested the […]","categories":["AI News"],"tags":[],"author_name":"Srishti Deoras","publish_date":"2017-04-20T03:53:27","publication_year":"2017","word_count":301,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","GAN","R"],"extracted_tech_keywords":["AI","R","Go","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/govt-directive-asks-isro-share-lithium-ion-batteries-manufacturing-technology-private-players\/","complexity_score":2,"technical_depth":6,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":38851,"title":"How To Accelerate Pandas With Just One Line Of Code","content":"Pandas is one of the most popular libraries on Python. The data manipulation and storing options it provides made it a go-to option for Kaggle competition. Pandas dataframes have more than 280 methods and more than 40 APIs. The number of options it has serves almost all the needs of a data scientist.  Python with pandas is in use in a wide variety of academic and commercial domains, including Finance, Neuroscience, Economics, Statistics, Advertising, Web Analytics, and more. But Pandas can get clumsy when dealing with large datasets such as those of genomics. The data is trimmed using big data tools and then run on pandas. So few data scientists at UC Berkeley came up with a new library, Modin which is a multi-process DataFrame library with an API identical to pandas. What Gives Modin The Edge In pandas, one can only use one core at a time when doing computation but Modin, enables the user to use all of the CPU cores on the machine. Unlike other parallel DataFrame systems, Modin is an extremely light-weight, robust DataFrame. It provides speed-ups of up to 4x on devices with 4 physical cores. Modin uses Ray to provide an effortless way to speed up the pandas notebooks, scripts, and libraries. Unlike other distributed DataFrame libraries, Modin provides seamless integration and compatibility with existing pandas code. Source: Modin docs By testing for read through actions like read_csv, large gains could be witnessed by efficiently distributing the work across the entire machine. With Modin, the developers tried to bridge the gap between handling of small and large data sets. Installing Modin: pip install modin Using Modin: import modin.pandas as pd #that ONE line# The Modin.pandas DataFrame is an extremely light-weight parallel DataFrame. Modin transparently distributes the data and computation. The Modin DataFrame architecture follows in the footsteps of modern architectures for database and high performance matrix system Source: Modin docs Query Compiler The Query Compiler receives queries from the pandas API layer. The API layer’s responsibility is to ensure clean input to the Query Compiler. The Query Compiler must have knowledge of the in-memory format of the data (currently a pandas DataFrame) in order to efficiently compile the queries. Partition Manager The Partition Manager is responsible for the data layout and shuffling, partitioning, and serializing the tasks that get sent to each partition Partition Partitions are responsible for managing a subset of the DataFrame. As is mentioned above, the DataFrame is partitioned both row and column-wise. What Is Ray Ray is a another system under development. It can be used for parallel and distributed Python to unify the ML ecosystem for low latency and high performance. Pandas on Ray is the component of Modin that runs on the Ray execution Framework. Currently, the in-memory format for Pandas on Ray is a pandas DataFrame on each partition. Currently, Ray is the only execution framework supported on Modin. The optimization that improves the performance the most is the pre-serialization of the tasks and parameters. This is primarily applicable to map operations. Modin will use all of the resources available on the machine and this usage, if required can be limited as follows: import ray ray.init(num_cpus=4) import modin.pandas as pd The above figure illustrates how Modin edges Pandas on 4 core CPU for better performance. Key Takeaways Modin takes care of all the partitioning and shuffling of the data. Performance of ‘read_csv’ shows more than a gigabyte per second of read-through which is far better compared to what Pandas does(1GB\/25 sec). The architecture allows Modin to exploit potential optimizations across framework and in-memory format of the data. Know more about Modin here","excerpt":"Pandas is one of the most popular libraries on Python. The data manipulation and storing options it provides made it a go-to option for Kaggle competition. Pandas dataframes have more than 280 methods and more than 40 APIs. The number of options it has serves almost all the needs of a data scientist.  Python with […]","categories":["Deep Tech"],"tags":["Datasets","pandas"],"author_name":"Ram Sagar","publish_date":"2019-05-09T11:29:59","publication_year":"2019","word_count":605,"keywords":["big data","Go","Datasets","API","AI","ML","Python","Ray","pandas","analytics","R","Pandas"],"extracted_tech_keywords":["AI","ML","analytics","Ray","Pandas","Python","R","Go","API","big data"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/how-to-accelerate-pandas-with-just-one-line-of-code-modin\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10090109,"title":"GPT-4: Beyond Magical Mystery","content":"“We’ll look back at it and say it was a very early AI and it’s slow, it’s buggy, it doesn’t do a lot of things very well,” said Sam Altman, CEO of OpenAI, while on a podcast with Lex Fridman recently.  “But,” continued Altman, “neither did the very earliest computers and they still pointed a path to something that was going to be really important in our lives, even though it took a few decades to evolve”. He was talking about ‘GPT-4’, the latest offering from OpenAI in the lineup of LLM models, which has left many amazed with what it can do, particularly with the latest update regarding the plugins on the platform–making it the iOS Appstore moment in AI. However, when Fridman prompted what would be the key moment marked in history a couple of years later—when it comes to the mass adoption of AI—Altman was of the opinion that ChatGPT will be the one. “It’s not like we could say here was the moment where AI went from not happening to it being a thing,” said Altman, “If I had to pick some moment from what we’ve seen so far I’d sort of pick chatGPT”. According to Altman, what made ChatGPT popular for the masses was the ease of usability, RLHF and the swift interface. Why is RLHF so important? Fridman said that reinforcement learning with human feedback (RLHF) is the secret ingredient that elevates the performance of machine learning models. Taking the conversation further, Fridman said that while the models such as ChatGPT can be trained on vast amounts of text data, they often lack practical applicability when tested. “Though successful in evaluations and tests, the base model is not very useful in practice,” said Altman. “However, RLHF,” continued Altman, “which involves incorporating human feedback into the model’s training, can significantly improve its usability.” He further explained that the simplest form of RLHF involves presenting two outputs to a human and asking which one is better, and then using that feedback to train the model using reinforcement learning techniques. This process allows the model to learn from human preferences and adapt its outputs accordingly. “In my opinion, RLHF is a highly effective method for enhancing the performance of machine learning models,” said Altman. Essentially, as Fridman later pointed out, a large language model that’s trained on a large dataset that creates a wisdom which is contained within the internet becomes much more impressive after adding a certain degree of human guidance on top of it. By incorporating human guidance into the training process, the model can better understand human preferences, making it more efficient at providing accurate and relevant outputs. “The feeling of alignment between the user and the model is crucial in making the model more usable and effective,” said Altman. GPT-4: Human Wisdom The conversation took a different turn when Fridman asked him whether there is a growing understanding within OpenAI about the nature of the “something” that makes GPT models so powerful or is it still a kind of beautiful, magical mystery? Altman believes that there are many different evaluation metrics that can be used to measure the performance of a model, both during and after training. “However,” said Altman, “the most important metric is how useful and impactful the model’s outputs are for people”. “This includes the value and utility it provides as well as the delight it brings, and the ways it can help create a better world through new science, products, and services,” added Altman. Additionally, he mentioned that while the researchers are gaining a better understanding of how GPT models work for specific inputs, there is still much they don’t fully understand. “We can’t always explain why the model makes certain decisions over others, although we are making progress in this area,” explained Altman. “For example, creating GPT-4 required a lot of understanding but we may never fully comprehend the vast amount of data that it compresses into a small number of parameters,” said Altman. According to him, GPT-4 can be considered a repository of human knowledge. Moreover, Altman believes that the exciting aspect of GPT models is their ability to reason, to some extent. “While there may be disagreements about what constitutes reasoning, many users of the system acknowledge that it is doing something in this direction, and that’s remarkable,” he remarks. The OpenAI CEO believes that by ingesting human knowledge, the model is acquiring a form of reasoning capability that could be additive to human wisdom in some senses. He also said that it can be used for tasks that don’t require any wisdom at all. “In the context of interactions with humans, GPT models can appear to possess wisdom, especially when dealing with multiple problems and continuous interactions,” said Altman.","excerpt":"The OpenAI CEO believes that by ingesting human knowledge, the model is acquiring a form of reasoning capability that could be additive to human wisdom in some senses.","categories":["AI Features"],"tags":["ChatGPT","lex fridman","OpenAI","Sam Altman"],"author_name":"Lokesh Choudhary","publish_date":"2023-03-27T18:29:41","publication_year":"2023","word_count":794,"keywords":["Go","ChatGPT","Sam Altman","machine learning","OpenAI","AI","RLHF","TPU","GPT","lex fridman","R","llm_models:GPT"],"extracted_tech_keywords":["AI","machine learning","ChatGPT","OpenAI","RLHF","TPU","R","Go","GPT","llm_models:GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/gpt-4-beyond-magical-mystery\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10095642,"title":"MongoDB Ups the Ante with Vector Search for Generative AI","content":"MongoDB has been a developer favourite for the longest time. The king of NoSQL database has added vector search, which greatly streamlines the integration of generative AI and semantic search into applications. MongoDB today came out with a suite of new announcements. This includes Google Cloud and MongoDB collaboration on an AI initiative, alongside the introduction of the AI Innovators Programme. As part of this collaboration, they have developed various new offerings for MongoDB and its data developer platform MongoDB Atlas. Additionally, MongoDB has unveiled MongoDB Atlas specifically tailored for industries and financial services. Furthermore, the relational migrator is now widely accessible to everyone. Read more: The good, the bad and the ugly – the story of MongoDB Fuelling Generative AI with MongoDB Atlas The process of converting data into vectors allows for effective clustering and semantic search operations. Similar data points are closer to each other in vector space, facilitating meaningful associations and comparisons. “MongoDB’s vector search capabilities provide a foundation for generative AI applications by converting diverse data types, such as text, images, videos, and audio files, into numerical vectors. This simplifies AI processing and enables efficient searches based on relevance,” said Himanshumali, principal solutions architect, in an exclusive interaction with AIM. The generative AI capabilities with MongoDB Atlas vector search enables precise information retrieval and personalisation. Moreover, MongoDB Atlas Search Nodes offer dedicated resources for enterprise-scale search workloads. Traditionally, when customers use different search engines, there are challenges in synchronising the source database with the search engine, leading to architectural complexity and time delays. However, with Atlas, search is part of the same platform, eliminating the need to move data to a different platform. “All search operations can be performed directly on Atlas in real time, providing customers with a seamless search experience,” he added. “Developers can now leverage familiar MongoDB query language to perform semantic queries without the need for specialised vector databases or language acquisition,” he said. MongoDB Atlas Stream Processing facilitates the handling of high-velocity streams of intricate data. Streaming data, which is continuously generated and evolving, is crucial for real-time use cases such as personalisation, fraud detection, IoT, and route planning. The lifecycle of streaming data involves a source where the data is generated, analytics on the streaming data for insights, and a persistent database to store the data. This integration of data at rest and data in motion provides an elegant and efficient experience for developers, enabling multiple real-time use cases. Lastly, the introduction of MongoDB Atlas Data Federation empowers users to query data and isolate workloads on Microsoft Azure. Collectively, these innovations empower businesses to optimize operational efficiency and accelerate innovation by consolidating various workloads on a unified developer data platform throughout the organisation. Additionally, MongoDB has unveiled MongoDB Atlas specifically tailored for industries and financial services. This offering aims to provide resilience, scalability, data privacy, and compliance solutions tailored to the requirements of the financial industry. “We have a strong customer base in the financial segment, including a significant number of top banks in North America, supporting over 150,000 transactions per second,” said Himanshumali. Harnessing the Best of Google MongoDB and Google Cloud have joined forces to enhance the adoption of generative AI and facilitate the creation of innovative applications. By leveraging its integrated operational data store, MongoDB Atlas uniquely supports the development of generative AI-powered applications with increased efficiency and simplicity. Talking about why Google Cloud is over other potential cloud providers, Himanshumali mentions that Google Cloud’s capabilities for generative AI, which they integrated into the partnership. So developers can now utilise MongoDB Atlas along with Google Cloud’s Vertex AI and access professional services for expedited software development, including quick-start architecture reviews. Say Goodbye to Relational Database Complexities Data forms the crux of everything. The buzz around generative AI would have not been there had it not been for these databases that form the heart and soul of the large language models. While we have Redis, Apache Cassandra as the NoSQL heroes, Neo4j, OrientDB is adding to the text-to-anything bandwagon with graph technology. While graph databases are more suitable for applications that necessitate complex queries about relationships between entities, whereas vector databases are more appropriate for applications that demand similarity searches. Nevertheless, there are instances where both types of databases can be advantageous. Take, for instance, a social network which could use a graph database to store user relationships and a vector database to store user characteristics. By doing so, the social network would be able to execute both complex relationship queries and similarity searches. “Relational databases come with their own set of limitations from the complexities that arise when using multiple niche databases,” he said. To solve this, MongoDB has now made its Relational Migrator available for public use. The migrator identifies workloads, updates schemas, modernises application code, and migrates data from various relational databases. It supports Oracle, SQL, and Postgres databases and utilises AI to generate code changes, streamlining the development process. Accenture, Capgemini, Nationwide Building Society, and Powerledger are among the notable customers and partners already benefiting from MongoDB Relational Migrator. Read more: Neo4j’s Role in Fueling Generative AI with Graph Technology","excerpt":"While graph databases are more suitable for applications that necessitate complex queries about relationships between entities, vector databases are more appropriate for applications that demand similarity searches","categories":["AI Features"],"tags":["MongoDB"],"author_name":"Shritama Saha","publish_date":"2023-06-22T21:00:00","publication_year":"2023","word_count":856,"keywords":["semantic search","AI","MongoDB","ML","RAG","vector databases","Aim","generative AI","analytics","Azure","fraud detection"],"extracted_tech_keywords":["AI","ML","analytics","generative AI","Aim","RAG","vector databases","fraud detection","semantic search","Azure"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/mongodb-ups-the-ante-with-vector-search-for-generative-ai\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":14077,"title":"What is Cognitive Networking?","content":"Essentially a type of data network, cognitive networks have grown to prominence in recent years for its extensive use in communication networks. This networking technique combines cutting-edge technologies from different research areas, viz., computer network, network management, machine learning, and knowledge representation, to address issues current networks are facing. Cognitive networks employ a cognitive process that can not only perceive current network conditions, but also plan, decide, and act on those conditions. Besides, this kind of networks learn from the consequences of its actions, while following the end-to-end goals. The loop, also known as the cognition loop can sense the environment and plan actions, based on the inputs obtained from sensors and network policies. There are limitations to the present technology used for data networking. Limitations surrounding response, state, and scope mechanism are the chief networking concerns which bother users. These limitations may lead to incapability of network elements such as nodes, protocol layers, and policies. Cognitive Networking plays a crucial role here by replacing conventional networks with modern and intelligent ways of communication, information, and learning. The concept of cognitive networking grew out of research conducted in cognitive radios, which combines intelligent, or cognitive processing with a platform. Besides, sensing the radio environment, the platform reflects ability to react and change its configuration. Advantages of Cognitive Networking Cognitive Networks reflect several advantages Cognitive networking technique can be applied across any type of network– fixed or wireless, thus offering better protection against security attacks and network intruders. Essentially, this technique proves useful for both service operators and customers. Such networks can adapt their operational parameters to respond to user’s needs, or to address changing environmental conditions. Moreover, they can learn from these adaptations, and exploit the knowledge obtained to execute decisions in future. Cognitive networking gives devices the ability to transmit more data, while reducing energy consumption, and efficiently using available bandwidth. With the passage of time, researchers are stumbling upon many possible civilian and commercial applications of flexible and adaptable networks. Cognitive Networking is the future, allowing users to focus on core activities, while all the network configuration and management is automated. The space will witness huge competition in the years to come, as companies like Google among others are exploring the ideas of integrating an intelligence or learning system in networks. Is Infinera’s Instant Network billing solution based on Cognitive Networking? The industry has made significant strides in the field of SDN, reflecting progress around virtualizing and controlling Layer 1\/2\/3 services over a fixed amount of optical capacity. Only by installing additional network hardware, the capacity can be increased. Infinera, a provider for intelligent transport networks, recently announced the launch of its Instant Network billing solution. The offering was touted as the “the next generation of software defined capacity (SDC) for cloud scale networks and a necessary foundation for cognitive networking.” Infinera launches Instant Network offering The Instant Network offering would enable service providers to activate capacity when revenue-generating services demand it, reduce capital expenditures by diminishing idle optical network capacity, or even lower business risk by shrinking the time between paying for capacity and activating revenue-generating services. This move towards unifying cognitive networking will allow service providers to equip their optic fiber network with Infinera photonics to a level of capacity greater than what they use, or pay for. “Since 2012, Infinera has been offering Instant Bandwidth since, when it introduced a single optical line card supporting five 100Gbps channels,” comments Pravin Mahajan, Director of Product and Corporate Marketing, Infinera. Challenges of implementing Cognitive Networks Undoubtedly, cognitive networking proposes numerous advantages over conventional networks. However, there are few challenges which must be addressed before we realize comprehensive benefits of implementing the technology. System stability, resilience, correctness, and robustness are some of the existing challenges enterprises are striving to address. The characteristic ability of cognitive networks to adapt easily to environmental changes will not only provide consistency, but also help in maintaining optimality of the system. Additionally, enterprises find it a challenging job to combine cognitive networking with “Internet of Things.” Connecting networks wirelessly to electronics and software, without using human interaction can be utilized to power several smart applications, such as networking air and water quality monitors, tracking devices’ energy consumption, and optimizing the operation of next-generation “smart grid” electricity systems. Another primary concern perturbing researchers is the need to address security and privacy. These intelligent networks will be able to make decisions on their own, and connect to whatever network they see fit. This could result in compromise of personal or sensitive information. Thus, security surfaces as a chief concern that needs to be addressed before we implement such networks comprehensively. Most importantly, cognitive networks must undergo laborious and careful engineering to assure the system’s validity, consistency, and correctness from the time of its development, until operation and removal. Monitoring is also a chief aspect about cognitive networking, and is performed to ensure that the system functions properly. Enabling the vision for a Cognitive Network Cognitive Networks will be common in Future The software defined networks existing today allow network links at layer 2 or layer 3 to be configured on the fly through human intervention, or by using a software through an API. Through cognitive networking, AI will be easily able to recognize the network state and usage patterns. This will enable easy reconfiguration of such networks. The intelligence associated with Cognitive Networking will assist towards predicting and optimizing data transfer. This transfer will consider aspects such as historical information, quality of service, needs of users, and the current state of the network. Furthermore, recommendations about optical network configurations can also be furnished to the user.","excerpt":"Essentially a type of data network, cognitive networks have grown to prominence in recent years for its extensive use in communication networks. This networking technique combines cutting-edge technologies from different research areas, viz., computer network, network management, machine learning, and knowledge representation, to address issues current networks are facing. Cognitive networks employ a cognitive process […]","categories":["IT Services"],"tags":["API","cognitive computing human capital","Internet of Things India","Machine Learning India","security India"],"author_name":"Amit Paul Chowdhury","publish_date":"2017-04-05T10:53:17","publication_year":"2017","word_count":941,"keywords":["Go","security India","API","machine learning","programming_languages:R","AI","Internet of Things India","Machine Learning India","programming_languages:Go","ViT","cognitive computing human capital","R"],"extracted_tech_keywords":["AI","machine learning","R","Go","API","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/what-is-cognitive-networking\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":5,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10044951,"title":"Intel Unveils Roadmap To Recover Lost Glory","content":"Intel has unveiled a roadmap to make up lost ground in the chip manufacturing industry. The chip giant intends to accelerate its investments and drive  innovation to power Intel through 2025 and beyond. Also read: Intel Outside: How The Chip Giant Lost Its Edge Intel has unpacked its new node names and the innovations enabling each node. The rationale behind the decision is to align the naming and numbering systems with the rest of the industry. For decades, the process node names corresponded to the real length of physical transistors, and Intel still continues to follow the pattern, although the naming and numbering schemes used by the rest of the industry no longer refer to any specific measurement but rather as a label to describe technology. The new node naming is based on technical parameters for a lucid framework that gives an accurate view of process nodes. Intel 7: Previously named 10nm Enhanced Superfin, the node will deliver a 10 to 15 percent performance per watt increase compared to its predecessor (10nm Superfin). Intel 7 will be used in products like Alder Lake chips and Sapphire Rapids for data centers. Intel 4: Previously Intel 7nm, the node fully embraces EUV lithography–something already utilised by Samsung and TSMC’s 5nm nodes. The node will use the same FinFET transistor architecture that Intel has been operating since 2011. With a 20 per cent performance per watt increase and other improvements, it will go into production by the second half of 2022. Intel 3: Intel 3 will deliver an 18 percent performance per watt increase using  FinFET optimisations and increased EUV over Intel 4. It leverages and implements a denser, high-performance library and increased intrinsic drive current, among other improvements. Intel 3 will be rolled out in the second half of 2023. New technologies Intel 20A (Angstrom Era): By 2024, Intel intends to enter something it calls the ‘Angstrom Era’ with its two new technologies RibbonFET and PowerVia. After FinFET in 2011, RibbonFET will implement gate-all-around transistor architecture. PowerVia is Intel’s interconnect innovation and industry-first implementation of backside power delivery providing signal routing. Along with the process roadmap, Intel also introduced its new IDM 2.0 strategy. In addition, the company announced two major updates to Feveros chip-stacking packaging technologies. Feveros Omni will allow more diversity in stacked chips, and Feveros Direct will enable direct copper to copper bonding between components which will minimise resistance and decrease bump pitches. The new updates are planned for 2023 production. “Moore’s Law is alive and well. We have a clear path for the next decade of innovation to go to ‘1’ and well beyond. I like to say that, until the periodic table is exhausted, Moore’s Law isn’t over, and we will be relentless in our path to innovate with the magic of silicon,” Pat Gelsinger, Intel CEO, stated. Embedded multi-die interconnect bridge (EMIB) as the first 2.5 D embedded bridge solution is still in place since 2017. Sapphire Rapids will be the first Xeon datacenter product to ship in volume with EMIB. It will also be the first dual-reticle-sized device in the industry, delivering nearly the same performance as a monolithic design. Beyond Sapphire Rapids, the next generation of EMIB will move from a 55-micron bump pitch to 45 microns. AWS will be Intel’s first customer to use and provide insights into the IFS packaging. Intel’s ambitious roadmap might be of help to the company to rearrange and recontextualise its products against competition including AMD, Nvidia and Qualcomm. Apple’s M1 Macs, for example, use superior 5nm chips from TMSC already and outperform Intel’s products.","excerpt":"We have a clear path for the next decade of innovation to go to ‘1’ and well beyond.","categories":["Global Tech"],"tags":["Intel","Intel chip"],"author_name":"Prajaktha Gurung","publish_date":"2021-08-02T11:00:00","publication_year":"2021","word_count":597,"keywords":["Rapids","Go","API","AWS","AI","Intel chip","innovation","cloud_platforms:AWS","programming_languages:R","RAG","R","Intel"],"extracted_tech_keywords":["AI","Rapids","RAG","AWS","R","Go","API","innovation","cloud_platforms:AWS","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/intel-unveils-roadmap-to-recover-lost-glory\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":33149,"title":"Meet Fasal, An AI Startup Which Aims To Put Indian Farms On An ‘Auto-Pilot’ Mode","content":"Co-founder and CEO, Fasal, Ananda Verma Fasal is an artificial intelligence startup that helps farmers increase productivity using advanced IoT and machine learning technologies was born out of minds of Ananda Verma and Shailendra Tiwari. Both belonged to families with a background in agriculture and shared a common passion for automated farming. The duo observed that agriculture in India and many developing worlds is based on traditional wisdom and guesswork. They looked inward towards their families and saw that farmers and seed producers faced day to day problems and saw an opportunity to address some core issues. The initial funding for Fasal came from friends and family. Putting trust in Ananda’s ambitious efforts, his brother Arpit Verma and his friend Adarsh Verma funded Fasal in the initial days. The funds provided by families helped Ananda and team to validate their ideas and along the way, more friends like Vaibhav Singh Rajput chipped in to keep the dream of Fasal alive. All their patience was rewarded when Fasal was able to raise a pre-seed fund from Zeroth, an internationally acclaimed AI startup accelerator. When asked about the first Fasal customers, Verma said, “It is a very interesting story. Fasal was put on a pilot mode for few months on five farms in Chhattisgarh where farmers profiles were very diverse. We also ran a pilot with Grover Zampa Vineyard in Bangalore. After using Fasal for some time all of them became our customers.” Talking about the immense value delivered to the farmers even in the initial days, Verma added, “Our customers were able to save water and increase their crop yield and quality and at the same time they were able to save input cost on sprays by following disease predictions made by Fasal. It is a great feeling for us to witness how Fasal is impacting farmers life.” Co-founder Of Fasal, Shailendra Tiwari The startup mostly focuses on solving issues which are in the pre-harvest phase. Ananda says, “We are very crop focused and work with a number of crops primarily on Horticulture. We are increasing our focus to cover more crop in the near future which also includes tea estates and coffee plantations.” Fasal is also working hard on image-based crop health analysis and Soil Nutrition analysis. When these products work, they will deliver a great amount of insights into additional factors that impact farming on a day to basis. Fasal currently is based out of HSR, Bengaluru and has 7 member team. The team works is distributed across product development, R&D, AI\/ML and customer support. Fasal plans to hire more data scientists and also build a sales team in the near future. To expand into many more Indian states Fasal is looking to raise a seed fund shortly adding to the pre-seed raised with Zeroth. Talking about the agricultural ecosystem in India, Verma said, “The Indian food industry is poised for huge growth. The Indian food and grocery market is the world’s sixth largest, with retail contributing 70 per cent of the sales. The Indian food processing industry accounts for 32 per cent of the country’s total food market, one of the largest industries in India and is ranked fifth in terms of production, consumption, export and expected growth. It contributes around 8.80 and 8.39 per cent of Gross Value Added (GVA) in Manufacturing and Agriculture respectively, 13 per cent of India’s exports and six per cent of total industrial investment.” All these stats point to the fact that technology interventions will be a huge part of doubling farmers’ income by 2022. And technology in farming is not an easy problem to solve. Fasal has huge plans going ahead.  “We would like to see Fasal in every farm in India in coming years and farming on auto-pilot. We are on a mission to help farmers grow more and grow better and we would like to make sure that they also sell better. A very far vision is – Fasal should be able to tell you what crop you should grow this season for a better outcome and what will be the market price once you harvest.”, Verma said.","excerpt":"Fasal is an artificial intelligence startup that helps farmers increase productivity using advanced IoT and machine learning technologies was born out of minds of Ananda Verma and Shailendra Tiwari. Both belonged to families with a background in agriculture and shared a common passion for automated farming. The duo observed that agriculture in India and many […]","categories":["AI Startups"],"tags":["AI and IoT"],"author_name":"Abhijeet Katte","publish_date":"2019-01-09T11:59:52","publication_year":"2019","word_count":689,"keywords":["Go","machine learning","artificial intelligence","AI","AI and IoT","ML","Ray","Aim","ViT","Rust","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","Aim","Ray","R","Go","Rust","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-startups\/meet-fasal-an-ai-startup-which-aims-to-put-indian-farms-on-an-auto-pilot-mode\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10073303,"title":"Fintech Profitability in Doubt for Next 2-3 Years","content":"With the latest trends in sight, fintech has erupted worldwide — but the profitability remains doubtful. According to a recent report by Matrix Partners with Boston Consulting Group (BCG), 70% of the respondents believe that most fintech companies may not hold profitable in the next two or three years. The report was based on a survey of over 125 founders and CXOs from leading Indian fintech firms, who pointed out expansion of products, improving customer service, and hiring as the current top priorities. As per Invest India’s latest reports, India is among the fastest-growing fintech markets in the world – home to 6,636 Fintech startups. The industry is estimated to reach $150 billion by 2025. The payment landscape has evolved into the most advanced system with regard to digital payments by volume (CAGR 50%) and value (CAGR 6%). There is a strong need to improve the risk management framework, along with improving competitiveness that can supervise fintech innovations without meddling with their economic potential. State of Indian fintech Landscape A fintech or financial technology firm merges the upcoming technological trends to provide better financial solutions through digital payments and transactions. This sector has seen the growth of numerous startups. The Indian fintech enjoys a strong position in the global financial market, clocking over $800 bn + annual payments transactions value, making it one of the strongest contributors to the economy. The report is a collective segment of insights that can be critical to the $5 trillion Indian economy, bringing out the need to re-evaluate their financials for innovative investments. Highlighting the growth of Indian fintech, Yashraj Erande, managing director and partner, BCG, says, “Indian fintechs are reaching scale and the industry is moving towards becoming mission critical, thus having a national role to play in achieving India’s pursuit of a 5-trillion economy. Fintechs and conventional players will co-exist as the need to work together is becoming imperative to provide a holistic experience to the customer. Having good compliance systems will smoothen the process of collaboration as willingness of partners to collaborate with compliant fintechs will be higher.” Incumbents revealed that monetisation and ARPU are the key priorities and biggest challenges for the industry today. Source: State of Fintech Union 2022 (BCG and Matrix Partners) The pandemic further boosted payments, leading to a 210% spike in funding between 2020 and 2021. With the rising funding, the industry has seen an upsurge in attracting global valuation. But over the past 9 months, $35.6 billion has been pulled out by FPIs (Foreign Portfolio Investment) from Indian markets. The rupee is depreciating at an all-time low of INR 79\/$, prompting FPIs to withdraw money before the further devaluation. Source: State of Fintech Union 2022 (BCG and Matrix Partners) In line with the global dip, a steep drop in funding of 36% in Q4 was seen in India. It indicates that a potential flow of capital can be carried out by improvements in macroeconomic fundamentals. Many business models were forced to re-imagine high credit risk and have people-intensive business models. Leaders who still believe in profitability Besides a majority of the industry leaders, who were doubtful about the profitability, 20% of the respondents feel positive about the sector. This positive outlook was predominantly from insurtech, paytech, and neo-banking sectors. Fintech is fuelled by the post-pandemic impact growth that has increased the adoption of digital services. It has indeed caused disruption but also paved the way for a more effective system. Many sectors such as education and banking benefit from this technology. “Fintech sector is mission-critical for the Indian economy; 36% of fintech customers are new-to-credit versus 22% for banks. This means a greater focus on profitability and governance,” said Yashraj Erande, managing director and partner at BCG. The road ahead Fintech is set to grow in the next 5 years and the report summarises the future aspects to tackle potential challenges. About 82% of the fintech companies believed that product expansion was the highest priority in the industry, with 61% improving customer service and 75% hiring the right talent. In contrast, cost reduction, fundraising and internal controls remained the least priority. It further suggests that fintechs need to bring more attention to governance for companies to outgrow advisors over a period of time. This can be achieved once companies proactively report their governance practices to garner trust from both customers and regulators. The pandemic has transformed the world into a digital hub. Fintech is a growing industry, and the opportunities are limitless. According to Market Research Future, global blockchain in the fintech market is expected to expand to USD 6700.63 million by 2023. With more streamlined algorithms to drive financial services, businesses will directly engage with the customers. With better frameworks and regulations, business models can be sustained, addressing uncertainties. “Profitability in lending is not a post-facto thought. The key is to get the basics right: continuously improve underwriting models, leverage tech to reduce opex and have a clear plan to reduce cost of funds with scale. Fintechs are battle-tested survivors of multiple debt-crises, and will continue to learn and emerge stronger,” says Vikram Vaidyanathan, managing director, Matrix Partners.","excerpt":"As India is booming with startups in fintech, industry leaders remain sceptical about its success","categories":["IT Services"],"tags":["Financial Services","FinTech"],"author_name":"Bhuvana Kamath","publish_date":"2022-08-23T10:00:00","publication_year":"2022","word_count":854,"keywords":["Go","API","AI","innovation","ML","Git","RAG","Financial Services","disruption","Rust","FinTech","R"],"extracted_tech_keywords":["AI","ML","RAG","R","Go","Rust","Git","API","innovation","disruption"],"url":"https:\/\/analyticsindiamag.com\/it-services\/fintech-profitability-in-doubt-for-next-2-3-years\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10022973,"title":"Interview With Abhishek Kashyap, Head Of Product, Google Cloud BigQuery AI","content":"Gone are the days when doing machine learning on large datasets required extensive programming and knowledge of ML frameworks. Now, even with limited machine learning knowledge and programming expertise, data analysts can harness the powers of machine learning. Google Cloud’s BigQuery ML, which  empowers data analysts to use machine learning through existing SQL tools and skills, is a good case in point. BigQuery ML allows analysts to build and evaluate ML models and accelerate model development and innovation by removing the need to export data from the data warehouse. Instead, BigQuery ML brings ML to the data. Analytics India Magazine(AIM) got in touch with Abhishek Kashyap, Head of Product, Google Cloud BigQuery AI, to get a glimpse how Abhishek and his team operates. Abhishek has a bachelors in Electrical Engineering from Indian Institute of Technology, Delhi, and a Masters and PhD in Electrical and Computer Engineering from University of Maryland, College Park. His research areas include computer networks, image processing, and graph theory. AIM: How did your fascination with algorithms begin? Abhishek: My fascination with algorithms began during my internship with IBM Research at IIT Delhi in 2000, where we worked on power saving algorithms for Bluetooth devices. Bluetooth was not yet mainstream, and power management was critical for success. I continued working on algorithms during my Masters and PhD, as well as at Lucent Bell Labs. A few years later, when I started MarianaIQ for AI based personalised marketing, I really got into data science. Most interesting and challenging data science component was automating it across customers without professional services – AI as a true Software as a Service. At Google Cloud, I have been focused on providing tools for anyone to successfully build quality machine learning models fast. Currently, I am the product management lead for making BigQuery the best intelligent data warehouse platform. I started with making it a successful platform for machine learning, and I am now working on adding natural language capabilities for analytics democratization. AIM: What books and other resources have you used in your journey? Abhishek: I highly recommend two online courses that got me started, as they both provide a solid foundation for data science: Statistical Learning by Prof. Robert Tibshirani and Prof. Trevor Hastie at Stanford, and Learning from Data by Prof Yaser Abu-Mostafa at Caltech. AIM: What were the initial challenges and how did you address them? Abhishek: Let’s look at our early challenges at each stage of the machine learning process: Training data: We almost always had a very small amount of training data, as we focused on B2B buyers and there aren’t that many in most companies’ databases. We ended up doing a lot of manual bootstrapping in early days, and learnt from the manual process to automate it for our use case.Extensible modeling: As I mentioned above, we built our models to be served as a SaaS vs having a data scientist customize for each client. To achieve that, we applied a very curated mix of unsupervised learning, not-too-complex classification (due to small training data sizes), and NN embeddings that removed the need for a lot of feature engineering.Quality test suites: Model improvement is hardly ever universal for all data sets. Thus, we had to continuously update our quality data tests to ensure new models do not perform worse on any important data sets.Pipelines and ML Ops: There were no standard tools back then, so we had to build our own for data-ML pipelines and ops.Explainability: Our clients wanted to know why we made a certain recommendation, and we had to experiment with tools available back then, like LIME. Unfortunately, that area was still nascent, and we could not get to a satisfactory answer back then. Today, there are very credible algorithms like SHAP, which makes it much easier to explain predictions. AIM: How do you approach any data science problem? Abhishek: I always approach a data science problem as a business problem that needs to be solved, which can then be translated to a label and features. Always having the business problem in mind results in a much better intuition for feature engineering, and types of models or chaining that would be required. Beyond that, it’s the standard but not-always-followed advice – start with a simple model, experiment, and choose the simplest among the models with acceptable accuracy. There is a penalty in going complex, and accuracy improvements need to be significant to trade-off with that penalty. AIM: What does your machine learning toolkit look like? Abhishek: I am currently the product manager for BigQuery ML, so I end up using it by default. With it I am able to create my first model within 10 minutes, and iterate really fast. I do not see a need for learning more complex ML libraries, when I can do most of it with SQL. AIM: There is a lot of hype around AI and ML. Which domain do you think will come out on top in the next 10 years? Abhishek: Let us break it into two areas: Custom models, and embedding in applications. At a high-level, AutoML modeling and easier interfaces will be mainstream. For custom modeling, it would be as easy as doing analytics. People will not need to learn programming languages and frameworks for most AI models, and most models would use an AutoML framework. Key expertise will be in understanding the business problem, the data, how it needs to be formatted to create training rows and labels, and how to iterate on those based on the predictions and explanations. The magic to find the right, clean data and create training data rows still doesn’t exist. When it comes to applications, I believe AI will be embedded in all applications where it can be useful, and people won’t even notice it. That is already the case in a lot of consumer applications, such as Youtube, Maps, Spotify. It will make its way into enterprise as well. From a data science point of view, it will be enabled by AutoML. AutoML will improve, and get cost effective for a variety of applications to get AI enabled easily. AIM: What’s your advice to data science aspirants? Abhishek: My advice would be to build a strong foundation for machine learning by learning both the practical and theoretical aspects. Otherwise, one can find it difficult to make progress fast, and debug a model to improve it. Additionally, I would recommend the two online courses mentioned earlier, as well as the Machine Learning Design Patterns book by my colleagues Lak Lakshmanan, Sara Robinson, and Michael Munn.","excerpt":"Gone are the days when doing machine learning on large datasets required extensive programming and knowledge of ML frameworks. Now, even with limited machine learning knowledge and programming expertise, data analysts can harness the powers of machine learning. Google Cloud’s BigQuery ML, which  empowers data analysts to use machine learning through existing SQL tools and […]","categories":["Global Tech"],"tags":[],"author_name":"Ram Sagar","publish_date":"2021-03-29T11:00:00","publication_year":"2021","word_count":1096,"keywords":["data science","Go","machine learning","AI","ML","Aim","analytics","SQL","R","data warehouse"],"extracted_tech_keywords":["AI","machine learning","ML","data science","analytics","Aim","R","SQL","Go","data warehouse"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/interview-with-abhishek-kashyap-head-of-product-google-cloud-bigquery-ai\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10046007,"title":"Why You May Not Be Getting A Call Back For That Data Science Job","content":"Over the past few years, the data science field has seen massive growth. Time and again, it has been touted as the hottest job of the 21st century. So naturally, a lot of people are drawn towards this field. However, beyond the hype, the reality is that it is still tough to land a job despite the projected great demand for data scientists. Like any other job interview, a recruiter may not necessarily divulge the reasons for rejecting a particular candidate. A lack of clarity can result in confusion, dejection, and disappointment for the interviewee. This article will explore a few reasons why you may not be getting a call back from the recruiter. Lack of Specific Skills Data science is a field of work that is responsible for solving some of the most complex tasks. This implies that a person working on such a project must possess immense knowledge and skill set. Therefore, an ideal candidate for a data scientist job has skills in each of these three broad areas–Maths\/Statistics, Databases\/Programming, and Business acumen. There are plenty resources available on what skills a data scientist must possess. This overwhelming amount of information often creates confusion. Even people without the know-how of basic analytical skills and numerical aptitude call themselves data scientists. It is essential to understand that an engineering degree is not sufficient to call oneself a data scientist. One must possess knowledge of Statistics, Mathematics, Economics, etc. It is also advisable to go that extra mile and update oneself on other upcoming technologies, including cloud computing and deep learning. Technical competency alone is not enough. Companies want people who can help them get more business and make money faster. A prospective candidate must be able to keep the big picture in mind. The ability to communicate even the most complicated technical aspects of your model to other team members (even from a non-technical background) is a highly desirable skill. Standing Out Annually, a lot of people enter the data science field. On paper, the know-how of technical and business skills alone is not enough to stand out among scores of applicants. A sure shot way to attract a recruiter’s interest would be to display samples of projects worked on in the past. This is especially useful for freshers. Employers often look at candidate’s GitHub repositories. Therefore, it is advised to highlight past work on GitHub, display and highlight your best code solutions, and make GitHub a one-stop shop for potential employers to qualify you. To begin, one can look at repositories of other data scientists to get ideas to spruce up your GitHub repository. Another way to demonstrate your skills would be to participate in different contests, hackathons, and competitions. One can try their hands on Kaggle data science challenges to place relatively higher on the leaderboard. A recruiter may see participation and ranking as something valuable. CVs and Interviews Often, in the end, it comes down to how well you ‘sell’ yourself. For starters, a curriculum vitae (CV) gives out a lot about a candidate even before the physical meeting. Therefore, it is important to develop a good CV specifically tailored to the data science job at a particular company. Few minutes of the interview are enough for the recruiter to make up their mind on whether to further promote the candidate or not. It is crucial to prepare rigorously for the same. One must do their homework on the company’s history, the team they are applying for, the projects the organization has worked on in the past, etc. Other miscellaneous things like body language, confidence, attire also make or break the deal. Other Miscellaneous Reasons We have discussed all the pointers that recruiters look for in a candidate, the lack of which could lead to rejection or cancellation of the candidature. That said, a candidate may prepare themself on all the pointers mentioned above and still may not receive a callback. Often the reason might lie within the company policy or hiring procedure. For example, due to applicant tracking system software, a candidate may be stopped at the door based solely on the channel they are coming from. A possible solution would be to come through less frequently used channels. Recruiters’ Take As per Sunil Bist, CEO, NetConnect Global, below are a few hygiene parameters that need to be looked into to improve your chances as a candidate to qualify for an interview call. Complementary Skills Knowledge – The potential employers will be keen to hire talents who are knowledgeable\/have worked on complementary computing technologies such as Cloud, Virtualisation technology, Service-Oriented Architecture (SOA), Grid Computing, etc.Case studies\/Blogs & Whitepapers – Publish your knowledge with blogs, case studies and whitepapers in industry forums so that leaders can make a note of your skills, experience and expertise.Network, Learn, Contribute & Reference (NLCR) – Create a network in Data Analytics leading companies or your target prospective employers and contribute to their thought leadership. This would help you learn and potentially earn a reference for the “Right Job”. “While certifications are necessary, learning about the deployment of scalable models and platforms used for deployment is very important. For this, involvement in live projects and internships that allow in-depth practical hands-on experience in the field should be prioritized. Additionally, using professional networking platforms effectively by reevaluating and building relevant connections, staying in touch with them, sharing knowledge, industry updates and experiences, and providing fresh and comprehensive information pertaining to one’s professional profile would be helpful as well,” said Shekhar Sanyal, Country Head and Director, IET India. “Companies are looking for problem solvers, and if your resume is only focused on the tools and techniques used, then it does not really give the recruiter a chance to evaluate your problem-solving skills. Therefore, while detailing out your resume, talk about how you used a specific approach or tool to solve a problem; this will make your candidature far more appealing to the recruiter and open up interesting conversations during the interview,” said Raghavan Kirthivasan, Senior Director – Data Science, Epsilon India. Suryadip Ghoshal, Chief Analytics Officer at Think360, said, “Knowledge of R and\/or Python along with experience of key packages is expected from most candidates. Beyond that, to stand out in the resume pool, few spikes that we look for are: Interesting Projects: With so many publicly available libraries and data, new projects (which aren’t a part of the curriculum of mainstream online courses) provide an opportunity for candidates to stand out. Candidates should make their projects public and share links to their GitHub and Tableau profiles. These projects can also act as conversation openers in interviews. Competition Rankings: Hackathons and competition forums like Kaggle are great learning experiences for candidates. Good performance (being in the top 10%) in these competitions is a testament to good programming and problem-solving skills and the perseverance and ability of the candidate to understand the details of the problem statements. “Building a good resume is equally important as many companies use bot screening, where the filters look for keywords and give scores to your resume. Make sure it’s readable and contains keywords in the job listing and is personalised,” said Harsh Gupta, founder and CEO, ProtonAutoML. Follow and incorporate these aspects and habits into your application process to multiply your chances of getting that aspired data science job.","excerpt":"Even people without the know-how of basic analytical skills and numerical aptitude call themselves data scientists.","categories":["AI Hirings"],"tags":["Data Science","Data Science Career","data science interview","Data Science Jobs","Data Scientist","open source data science projects","Statistics for Data Science","what is data science"],"author_name":"Shraddha Goled","publish_date":"2021-08-15T13:00:00","publication_year":"2021","word_count":1221,"keywords":["what is data science","data science","Go","data science interview","open source data science projects","AI","cloud computing","ML","Data Science Jobs","RAG","Data Science Career","Python","deep learning","Statistics for Data Science","analytics","Data Science","Data Scientist","R"],"extracted_tech_keywords":["AI","ML","deep learning","data science","analytics","RAG","cloud computing","Python","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-hiring\/why-you-may-not-be-getting-a-call-back-for-that-data-science-job\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10119091,"title":"Reid Hoffman Creates a DeepFake of Himself, Reid AI","content":"Reid Hoffman, the co-founder of LinkedIn recently discussed various AI topics in an interview with a virtual AI twin of himself. He shared it through a post on X. He said he deepfaked himself to see if conversing with an AI-generated version of myself can lead to self-reflection, new insights into my thought patterns, and deep truths. Why did I deepfake myself? To see if conversing with an AI-generated version of myself can lead to self-reflection, new insights into my thought patterns, and deep truths. pic.twitter.com\/DWODoZ9lXL— Reid Hoffman (@reidhoffman) April 24, 2024 In the interview with his virtual twin, he posed several questions to assess it, including how to summarise a 336-page book on blitz scaling in one sentence. He also inquired about which of them would be a better video host, to which the response highlighted the virtual twin’s ability to excel in hosting content with extensive data, frequent updates, or multiple languages. Rohit Bhargava, recently shared from Hoffman’s video an intriguing discussion on humanising his LinkedIn page, marking a pinnacle in AI’s ability to personalise and humanise digital platforms. An experiment well worth delving into. Towards the conclusion of the video, Hoffman himself was questioned about the strategic shift at Inflection AI (a company he co-founded) and the ramifications of co-founder and CEO Mustafa Suleyman’s recent transition to Microsoft as the CEO of the newly established Microsoft AI division. Hoffman characterised Inflection AI as a “remarkable entity” adept in both emotional intelligence and cognitive prowess. Hoffman explained to his AI interviewer that Mustafa’s passion lies in building consumer products at scale, but the business’s development would take years. He emphasised that the true startup opportunity lies in the developer\/API business. With the shift to Microsoft, Suleyman can now concentrate on consumer opportunities without immediate pressure to prove the business model. Houffman further added, “I found it somewhat intriguing, as if it opened a path towards enhancing my humanity, a means to express myself more authentically. It’s akin to the insights gained from watching a video of oneself, discovering nuances that refine our communication skills.”","excerpt":"Ask Reid AI twin edition of Reid Hoffman to improve the Linkedin page see what it replied.","categories":["AI News"],"tags":["DeepFake"],"author_name":"Gopika Raj","publish_date":"2024-04-25T17:50:36","publication_year":"2024","word_count":346,"keywords":["API","programming_languages:R","AI","Git","DeepFake","R","startup"],"extracted_tech_keywords":["AI","R","Git","API","startup","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/reid-hoffman-creates-a-deepfake-of-himself-reid-ai\/","complexity_score":2,"technical_depth":6,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":66079,"title":"Anomaly Detection in Temperature Sensor Data using LSTM RNN Model","content":"Anomaly detection has been used in various data mining applications to find the anomalous activities present in the available data. With the advancement of machine learning techniques and developments in the field of deep learning, anomaly detection is in high demand nowadays. This is because of implementing machine learning algorithms with heavy datasets and generating more accurate results. The anomaly detection is not limited now to detecting the fraudulent activities of customers, but it is also being applied in industrial applications in a full swing. In manufacturing industries, where heavy machinery is used,  the anomaly detection technique is applied to predict the abnormal activities of machines based on the data read from sensors. For example, based on the temperature data read through the sensors, the possible failure of the system can be predicted. In this article, we will discuss how to detect anomalies present in the temperature data that is available in the time-series format. This data is captured from the sensors of an internal component of a large industrial machine. What is Anomaly? In machine learning and data mining, anomaly detection is the task of identifying the rare items, events or observations which are suspicious and seem different from the majority of the data. These anomalies can indicate some kind of problems such as bank fraud, medical problems, failure of industrial equipment, etc. The anomaly detection has two major categories, the unsupervised anomaly detection where anomalies are detected in an unlabeled data and the supervised anomaly detection where anomalies are detected in the labelled data. There are various techniques used for anomaly detection such as density-based techniques including K-NN, one-class support vector machines, Autoencoders, Hidden Markov Models, etc. The Data set In this experiment, we have used the Numenta Anomaly Benchmark (NAB) data set that is publicly available on Kaggle. It is a novel benchmark for evaluating machine learning algorithms in anomaly detection in streaming, online applications. It consists of more than 50 labelled real-world and artificial time-series data files plus a novel scoring mechanism designed for real-time applications. This dataset also comprises a time-series data named ‘machine_temperature_system_failure’ in the CSV format. It comprises temperature sensor data of an internal component of a large industrial machine. This dataset contains anomalies including the shutdown of the machine, catastrophic failure of the machine etc. Implementation of Anomaly Detection To implement our work, first, we need to import the required libraries. The Pandas library is required for data frame operations and the NumPy library is required for array operations. For plotting, we are importing matplotlib and seaborn libraries and for the preprocessing of the data, we are importing the preprocessing library. For the LSTM Recurrent Neural Network, the required Keras libraries are imported. The time library is used in viewing the compile time of our LSTM RNN model. #Importing Required Libraries import pandas as pd import numpy as np import matplotlib import seaborn import matplotlib.dates as md from matplotlib import pyplot as plt from sklearn import preprocessing from keras.layers.core import Dense, Activation, Dropout from keras.layers.recurrent import LSTM from keras.models import Sequential import time As this implementation was done in Google Colab, we use the below code snippet to read the data from the local computer system. #To upload file from local system from google.colab import files uploaded = files.upload() Once the upload is shown as 100%, we will read the CSV data file that is the temperature sensor data in the time-series format. #Reading file from local system df = pd.read_csv(\"ambient_temperature_system_failure.csv\") To check whether the data is read successfully, we see the header of the file and then visualize the data through plotting. df.head() #Visualize the data figsize=(10,5) df.plot(x='timestamp', y='value', figsize=figsize, title='Temperature (Degree Farenhite)'); plt.grid(); plt.show(); As we can see that the dates do not appear clearly on the X-axis, we need to change the type of timestamp column. As the temperature in our data set is given in the degree Fahrenheit, we will convert it into degree Celcius. #Changing the type of timestamp column for visualization df['timestamp'] = pd.to_datetime(df['timestamp']) df['value'] = (df['value'] - 32) * 5\/9 df.plot(x='timestamp', y='value', figsize=figsize); plt.title('Temperature (Degree Celcius)', fontsize=16); plt.grid(); plt.show(); Now we can see the data is visualized perfectly. To check the stability of temperature during days and nights of weekdays and weekends we are going to preprocess our data accordingly. First, we will specify hours, then days, then weekdays and nights. Finally, we will visualize the temperature during these time-periods using a histogram. #Formating the data into required format df['hours'] = df['timestamp'].dt.hour df['daylight'] = ((df['hours'] >= 7) & (df['hours'] <= 22)).astype(int) df['DayOfTheWeek'] = df['timestamp'].dt.dayofweek df['WeekDay'] = (df['DayOfTheWeek'] < 5).astype(int) # Anomaly estimated population outliers_fraction = 0.01 df['time_epoch'] = (df['timestamp'].astype(np.int64)\/100000000000).astype(np.int64) df['categories'] = df['WeekDay']*2 + df['daylight'] a = df.loc[df['categories'] == 0, 'value'] b = df.loc[df['categories'] == 1, 'value'] c = df.loc[df['categories'] == 2, 'value'] d = df.loc[df['categories'] == 3, 'value'] #Visualizing the formatted data figsize=(10,5) fig, ax = plt.subplots(figsize=figsize) a_heights, a_bins = np.histogram(a) b_heights, b_bins = np.histogram(b, bins=a_bins) c_heights, c_bins = np.histogram(c, bins=a_bins) d_heights, d_bins = np.histogram(d, bins=a_bins) width = (a_bins[1] - a_bins[0])\/6 ax.bar(a_bins[:-1], a_heights*100\/a.count(), width=width, facecolor='blue', label='Weekend Night') ax.bar(b_bins[:-1]+width, (b_heights*100\/b.count()), width=width, facecolor='green', label ='Weekend Light') ax.bar(c_bins[:-1]+width*2, (c_heights*100\/c.count()), width=width, facecolor='red', label ='Weekday Night') ax.bar(d_bins[:-1]+width*3, (d_heights*100\/d.count()), width=width, facecolor='black', label ='Weekday Light') plt.legend() plt.show() The above histogram shows that the temperature is comparatively more stable during Week Days in the daylights. Now, we will preprocess our dataset for training the LSTM recurrent neural network. For this purpose, first, we will take the required columns from the dataset and scale it using Standard Scaler. #Preparing the data for LSTM model data_n = df[['value', 'hours', 'daylight', 'DayOfTheWeek', 'WeekDay']] min_max_scaler = preprocessing.StandardScaler() np_scaled = min_max_scaler.fit_transform(data_n) data_n = pd.DataFrame(np_scaled) In the next step, we will define and initialize the required parameters and define the training and the test data set. We are going to learn from 50 previous values and we predict through the LSTM model just the one next value. #Important parameters and training\/Test size prediction_time = 1 testdatasize = 1000 unroll_length = 50 testdatacut = testdatasize + unroll_length  + 1 #Training data x_train = data_n[0:-prediction_time-testdatacut].values y_train = data_n[prediction_time:-testdatacut  ][0].values #Test data x_test = data_n[0-testdatacut:-prediction_time].values y_test = data_n[prediction_time-testdatacut:  ][0].values As we know that the architecture of a Recurrent Neural Network has a hidden state. The hidden state at time t is a function of the hidden state at time t−1 and the input at time t. This hidden state at the time 0 is typically initialized to 0. The fundamental reason why RNNs are unrolled is that all previous inputs and hidden states are used in order to compute the gradients with respect to the final output of the RNN. The below unrolling function will create a sequence of 50 previous data points for each of the training and test data point. def unroll(data,sequence_length=24): result = [] for index in range(len(data) - sequence_length): result.append(data[index: index + sequence_length]) return np.asarray(result) #Adapt the datasets for the sequence data shape x_train = unroll(x_train,unroll_length) x_test  = unroll(x_test,unroll_length) y_train = y_train[-x_train.shape[0]:] y_test  = y_test[-x_test.shape[0]:] Now, we will see the final shape of our training and test data. #Shape of the data print(\"x_train\", x_train.shape) print(\"y_train\", y_train.shape) print(\"x_test\", x_test.shape) print(\"y_test\", y_test.shape) In the next step, we will define and add layers to the LSTM Recurrent Neural Network one-by-one. For basic details about the LSTM RNN model, you may refer to the article ‘How to Code Your First LSTM Network in Keras’. #Building the model model = Sequential() model.add(LSTM(input_dim=x_train.shape[-1], output_dim=50, return_sequences=True)) model.add(Dropout(0.2)) model.add(LSTM(100, return_sequences=False)) model.add(Dropout(0.2)) model.add(Dense(units=1)) model.add(Activation('linear')) start = time.time() model.compile(loss='mse', optimizer='rmsprop') print('compilation time : {}'.format(time.time() - start)) Once, the LSTM RNN model is defined and compiled successfully, we will train our model. The below hyperparameters can be tuned to check the better performance. model.fit(x_train, y_train, batch_size=3028, nb_epoch=50, validation_split=0.1) After the successful training of the model, we will visualize the training performance. #Visualizing training and validaton loss plt.figure(figsize = (10, 5)) plt.plot(model.history.history['loss'], label = 'Loss') plt.plot(model.history.history['val_loss'], Label = 'Val_Loss') plt.xlabel('Epochs') plt.ylabel('Loss') plt.grid() plt.legend() Now, we will execute the below code snippet to better understand the difference between the original data and the predicted data through the visualization. #creating the list of difference between prediction and test data loaded_model = model diff=[] ratio=[] p = loaded_model.predict(x_test) for u in range(len(y_test)): pr = p[u][0] ratio.append((y_test[u]\/pr)-1) diff.append(abs(y_test[u]- pr)) #Plotting the prediction and the reality (for the test data) plt.figure(figsize = (10, 5)) plt.plot(p,color='red', label='Prediction') plt.plot(y_test,color='blue', label='Test Data') plt.legend(loc='upper left') plt.grid() plt.legend() Now, in the next step, we are going to find the anomalies. The most distant predicted values are considered as anomalies. Using the below code snippet, we will find the anomalies in the data. #Pick the most distant prediction\/reality data points as anomalies diff = pd.Series(diff) number_of_outliers = int(outliers_fraction*len(diff)) threshold = diff.nlargest(number_of_outliers).min() #Data with anomaly label test = (diff >= threshold).astype(int) complement = pd.Series(0, index=np.arange(len(data_n)-testdatasize)) df['anomaly27'] = complement.append(test, ignore_index='True') print(df['anomaly27'].value_counts()) Finally, we will visualize the anomalies using the below code for plotting. #Visualizing anomalies (Red Dots) plt.figure(figsize=(15,10)) a = df.loc[df['anomaly27'] == 1, ['time_epoch', 'value']] #anomaly plt.plot(df['time_epoch'], df['value'], color='blue') plt.scatter(a['time_epoch'],a['value'], color='red', label = 'Anomaly') plt.axis([1.370*1e7, 1.405*1e7, 15,30]) plt.grid() plt.legend() As we can see in the above picture, the anomalies are visualized as red points. All the above steps can be repeated multiple times to visualize the anomalies by tuning the hyperparameters. If we find the same visualization at the end, then we can end-up with these anomalies. References: A Zimek, E Schubert, “Outlier Detection”, Encyclopedia of Database Systems, Springer New York. V. J. Hodge, J Austin, “A Survey of Outlier Detection Methodologies”, Artificial Intelligence Review. BoltzmannBrain, “Numenta Anomaly Benchmark: Dataset and scoring for detecting anomalies in streaming data”, Kaggle. Victor Ambonati, “Unsupervised Anomaly Detection”, Kaggle.","excerpt":"In this article, we will discuss how to detect anomalies present in the temperature data that is available in the time-series format. This data is captured from the sensors of an internal component of a large industrial machine.","categories":["Deep Tech"],"tags":["anomaly detection","Data loss prevention","lstm recurrent neural network"],"author_name":"Dr. Vaibhav Kumar","publish_date":"2020-05-27T17:00:00","publication_year":"2020","word_count":1592,"keywords":["NumPy","artificial intelligence","machine learning","Keras","AI","neural network","Colab","Ray","anomaly detection","Data loss prevention","deep learning","lstm recurrent neural network","Pandas"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","deep learning","neural network","Ray","Keras","Colab","Pandas","NumPy"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/anomaly-detection-in-temperature-sensor-data-using-lstm-rnn-model\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10079463,"title":"IBM Reveals 433-Qubit Processor, Aims 1000 Qubit-Plus Processor Next Year","content":"IBM recently announced a new breakthrough in quantum hardware and software outlining the pioneering vision for quantum-centric supercomputing. The advancement was unveiled during the annual IBM Quantum Summit 2022—exhibiting the company’s quantum ecosystem of clients, developers, and partners—that showcases useful advancements in quantum computing around the world. IBM and Director of Research and Senior Vice President Dr Darío Gil said that the new 433 qubit ‘Osprey’ processor brings them a step closer to the point where quantum computers will be used to tackle previously unsolved problems. Further, he said that they are continuously scaling up their quantum technology across hardware, software, and classical computing, alongside partnering with customers worldwide. “This work will provide a foundation for the coming era of quantum-centric supercomputing,” said Gil. ​ IBM’s new 433-quantum bit (qubit) processor named ‘IBM Osprey’ has the largest qubit count of any firm’s quantum processor—tripling the 127 qubits on the recently launched IBM Eagle processor. The new processor runs complex quantum computations better than the computational capability of a classical computer. At the Global AI Summit, Scott Crowder, the VP and CTO at IBM, said: “If we continue to progress according to our roadmap. This year, we will demonstrate a 433 qubit system. Next year, we will demonstrate a 1000 plus qubit system.” (Source: IBM) Read: Behind IBM’s plans to build a 4000+ qubit quantum computer by 2025 Moreover, a beta update has been released to Qiskit Runtime, which will now let users trade speed for decreased error count by choosing a simple option in the API. Furthermore, new details on the IBM Quantum System Two—which combines multiple processors into a single system using communication links—were introduced at the Summit.  The company claims that the next generation system will be made online by the end of 2023—a building block of quantum-centric supercomputing—scaling further by employing a modular architecture to classical workflows and integrating quantum seamlessly. Read: IBM Bet On Quantum Technology In May 2022, at Think Conference, IBM had revealed its intent to develop a 4000+ qubit quantum computer by 2025. In 2019, IBM debuted its first integrated quantum computing system, called the ‘IBM Quantum System One’. In 2020, the tech giant revealed the roadmap to build a suite of scalable, larger and better processors.","excerpt":"The advancement was unveiled during the annual IBM Quantum Summit 2022 – showcasing IBM’s quantum ecosystem of clients, developers, and partners – that showcases useful advancements to quantum computing around the world.","categories":["AI News"],"tags":["IBM","Quantum Computing"],"author_name":"Bhuvana Kamath","publish_date":"2022-11-11T15:26:06","publication_year":"2022","word_count":374,"keywords":["Quantum Computing","API","programming_languages:R","AI","ML","Scala","Aim","IBM","programming_languages:Scala","R","emerging_tech:quantum computing"],"extracted_tech_keywords":["AI","ML","Aim","R","Scala","API","programming_languages:R","programming_languages:Scala","emerging_tech:quantum computing"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/ibm-reveals-433-qubit-processor-aims-1000-qubit-plus-processor-next-year\/","complexity_score":4,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":30399,"title":"Kerala Signs MoU With Airbus Bizlab, To Get New Innovation Centre","content":"The Kerala state government ended the last week with a bang as they signed a memorandum of understanding with Airbus BizLab to open a state-of-the-art innovation centre. This centre will be the nodal body for planning and executing all the activities to help start-ups. The MoU was inked betweenAirbus Bizlab India’s Siddharth Balachandran and Saji Gopinath, CEO of the Kerala Startup Mission (KSUM), in the presence of Chief Minister Pinarayi Vijayan. “We are confident that this partnership with Airbus Bizlab will help the industrial ecosystem of Kerala. It will up-skill the local youth and enhance employability,” told a newswire. Airbus BizLab will also provide support and mentoring for start-ups in Kerala and conduct regular workshops and discussions with experts from the aerospace and defence sectors. Bizlab is a part of Airbus’ innovative strategy to bring together start-ups and Airbus intrapreneurs to work and speed up the transformation of their innovative ideas into valuable businesses. It has a community of over 2,000 B2B start-ups from India, South East Asia and Israel. Through its six-month acceleration programme for start-ups and internal projects, Airbus Bizlab gives access to a large number of coaches, experts and mentors in various domains that help start-ups and internal projects speed-up the transformation of their ideas into valuable businesses. An innovation centre of Airbus BizLab, the global aerospace accelerator of Airbus, will be established in Thiruvananthapuram. Airbus signed an MoU with Kerala Startup Mission to assist startups in the Aerospace sector. MoU was signed in the presence of CM Pinarayi Vijayan. pic.twitter.com\/bTWG27Da9V — CMO Kerala (@CMOKerala) November 17, 2018","excerpt":"The Kerala state government ended the last week with a bang as they signed a memorandum of understanding with Airbus BizLab to open a state-of-the-art innovation centre. This centre will be the nodal body for planning and executing all the activities to help start-ups. The MoU was inked betweenAirbus Bizlab India’s Siddharth Balachandran and Saji […]","categories":["AI News"],"tags":["kerala","Startups"],"author_name":"Prajakta Hebbar","publish_date":"2018-11-19T13:48:50","publication_year":"2018","word_count":261,"keywords":["Go","programming_languages:R","AI","innovation","programming_languages:Go","Ray","ViT","Startups","R","kerala","startup"],"extracted_tech_keywords":["AI","Ray","R","Go","ViT","innovation","startup","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/kerala-signs-mou-with-airbus-bizlab-to-get-new-innovation-centre\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10112635,"title":"Apple Acquired 32 AI Startups by 2023","content":"Apple surged ahead with an unprecedented acquisition spree in 2023, securing up to 32 AI companies throughout the year. This aggressive maneuver places the tech giant significantly ahead of its major rivals in the AI arena, including Google, Meta, and Microsoft. Insights from Stocklytics, referencing a Statista report, highlight Apple’s commitment to fortifying its AI capabilities across a diverse product portfolio. The acquisitions signal Apple’s strategic positioning for forthcoming tech innovations, amidst substantial investments by competitors in established AI enterprises. In the overall AI startup acquisition, Google trails Apple with 21, Meta with 18, and Microsoft lags with 17. In recent years, Apple has completed notable acquisitions of AI startups such as Voysis, WaveOne, Emotient, and Laserlike. These acquisitions span various domains including voice assistant capabilities, video compression technology, expression recognition, app recommendations, and music AI. Read: Apple Has AI Plans, Without the Marketing Noise Apple’s emphasis on early-stage startups underscores a strategic effort to identify and invest in emerging AI trends ahead of competitors, positioning the company at the forefront of AI innovation. The innovations in the tech field are clearly highlighted by the Apple Vision Pro release this year. While Apple’s plans for implementing these technologies into consumer products remain undisclosed, competitors like Samsung and Google have showcased advanced AI features in their smartphones, such as the Galaxy S24 Ultra. Despite Apple’s secretive nature, analysts estimate a rapid pace of startup acquisitions, averaging 2-3 per week in recent years. This relentless pursuit underscores Apple’s ambition to lead the AI race in the coming years, even as competitors vie for dominance through acquisitions of established AI companies and technologies.","excerpt":"Google trails Apple with 21, Meta with 18, and Microsoft lags with 17.","categories":["Deep Tech"],"tags":["Apple","Mergers and Acquisitions","Startups"],"author_name":"Mohit Pandey","publish_date":"2024-02-13T11:53:43","publication_year":"2024","word_count":270,"keywords":["Go","API","programming_languages:R","AI","Apple","innovation","programming_languages:Go","RAG","Startups","Mergers and Acquisitions","R","startup"],"extracted_tech_keywords":["AI","RAG","R","Go","API","innovation","startup","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/apple-acquired-32-ai-startups-by-2023\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10071133,"title":"75 massive tools launched at AI Defence Exhibition","content":"On July 11, Defence Minister Rajnath Singh launched 75 newly-developed AI products during the first-ever symposium and exhibition on Artificial Intelligence in Defence (AIDef) in New Delhi. At one point, Singh quoted Russian President Vladimir Putin: “Whoever becomes the leader in artificial intelligence will rule the world”. However, he further explained India’s belief in the principle of ‘vasudhaiv kutumbakam’ (the whole world is one family), and the country has no intention to rule the world. AI Platform Automation Merlin MLops It is an MLOps solution that eases deep learning development and deployment cycle by providing a simple means to create, edit, and deploy ML models. System of Disseminated Parallel Control Computing in real time (DPCC) In real-time, this system uses several processing units to perform parallel processing, eliminating time delays and reducing risks. Deepcatch edge AI platform It is an Edge AI vision platform that helps monitor machines with alerts and notifications and remote diagnostics. iSentinel iSentinel is a deep learning-based threat detection and tracking system that has advanced learning algorithms. Autonomous\/Unmanned\/Robotic Systems Sapper System – Mine detection UGV It is a mine detection unmanned ground vehicle (UGV) that can detect and mark mines using an illuminating spray. AI capability in Swarm Drones Each of these drones is powered with distributed active collision avoidance algorithms. These algorithms compute potential future collision with every other drone and alter its path. Project Drone Feed Analysis – AI based RPA feed and data analysis It is an AI deep learning-based identification system that builds a database repository and carries out analysis providing military patterns of enemy operations and prediction of events. Silent sentry (Rail Mounted Robot with AI) It is designed to plug the gaps in surveillance networks. The army design bureau has also shared its design to produce in large numbers. DAHAK – Automated fast Interface Boat (AFIB) DAHAK is an indigenously developed platform that was developed in collaboration with BEL with advanced software and algorithms to perform autonomous operations even in dense maritime traffic and shallow water areas. Project Storm Drone This AI-enabled automated room intervention drone systems is enabled with lethal and non-lethal payloads are used to carry out building clearance and urban surveillance in GPS-denied areas. Cognitive radar DRDO Young Scientist Laboratory and Cognitive Technologies addressed certain issues and developed this project, using deep neural networks and reinforcement learning methods in target detection, angular estimation, and waveform adaptation. AI-Enabled Remotely Operated Vehicle (ROV) An AI-enabled ROV has been developed for underwater inspections of marine\/civil infrastructure with advanced capabilities for feature recognition. HR Chatbot ‘Anvesha’ This AI-enabled chatbot has helped the digitalisation of employee services by providing information directly in their mobile phones. AI-Powered Unmanned Ground Vehicle Enhanced Collaborative Autonomous Rover System (ECARS) UGV is a multipurpose, multi-terrain modular platform. The perception and navigation modules are AI-powered, enhancing the vehicle’s capability to traverse across multi terrains. Blockchain based automation Permissive Block Chain Mechanism Distributed Ledger is the concept used for development with necessary identity management and control mechanisms to protect the data shared among different organisations. This proposed use-case solution is planned to provide transparency, security and auditability for file sharing among Defence Public Sector Undertakings (DPSUs) using blockchain. Command, Control, Communication, Computers and Intelligence, Surveillance and Reconnaissance System (C4ISR) AI-Based Intercept Management System (IMS) IMS is indigenous software developed using visualisation techniques and artificial intelligence to interpret op-critical data, thereby generating accurate pictures. AI-Based Motion Detection & Target Identification System in LC Proactive Real-time Intelligence and Surveillance Monitoring System (PRISM) is an AI system capable of generating real-time audio\/ visual alerts detecting suspicious enemy movements from multiple surveillance inputs. Continuously Observing Ubiquitously Available AI-Surveillance System (Choukass) Choukass is designed to obtain early warning of any intrusions. The information is transmitted through a small in-built NB-IoT satellite module to a customised web portal accessible at its surveillance centres. AI-Enabled Airborne Electro-Optic Infrared System DRDO’s object detection and classification system is based on advanced AI algorithms and it helps in border surveillance and monitoring suspicious personnel and vehicular movement. Deep Learning Toolkit for Aerospace and Defence This toolkit can be used for automatic target recognition, and training pilots against intelligent adversaries. Adversary Network Analysis Tool (ANANT) The tool analyses multi-mode and multi-relational adversary networks. Its potential users are intelligence agencies and defence forces for the identification of criminal tracking. Target Tracking for Complex Naval Scenarios The system is useful in surveilling naval targets for low, mid, and long-range applications for firing. Animal Detection for Railways The model detects wild animals from the video sequence and triggers an alarm to alert the train operator to take precautionary measures. Enemy Aircraft Activity Recognition & Classification Air defence systems can use the module to improve situational awareness and response to enemy intrusion. AI-Based Anomaly Detection for Maritime Domain This model uses approach for motion pattern extraction and detection in maritime vessel traffic and based on circular quad trees. Passive ranging using a classifier The system detects objects in the images and then uses pinhole camera geometry to estimate the range. It can estimate the ranges of the object from a single image. AI Based Passive TWS (Track While Scan) System It is a system for naval surveillance application to monitor objects of interest using a camera mounted on a Pan-Tilt Platform. PATWAR generates a 360-degree panorama view of the surrounding area and then detects the objects of interest using deep learning. Development of Machine Algorithms for Maritime Anomaly Detection Deep Darshak(TM) is an AI\/ML-based platform for carrying out analytics on AIS and other ship position data to assist the users in identifying vessels of interest and suspicious activity. Enhancing UDA by use of AI\/ML and other Novel Techniques UDA processes voluminous data to discern objects\/sounds of interest amidst the background noise. Ambient noise cancellation and better algorithms\/signal processing can also increase the detection range. AI\/Big Data for Acoustic and Magnetic Signature Analysis The system uses AI to identify the inception of cavitation, tonals and correlate the tonals to data analysis reports. Cyber Security Android Malware Detection Solution This system analyses the weaknesses of the android applications. It reports the detected malware and performs audits and exploits (for vulnerability proof of concept purposes) on Android applications. Malware detection is done using deep learning libraries. Human Behavioural Analysis Driver Fatigue Monitoring System The system detects the onset of drowsiness in the driver while the vehicle is in motion.  A camera inside the cabin films the driver constantly. Intelligent Monitoring System Project Seeker It is a facial Recognition System for Population Monitoring, Surveillance and Garrison Security. It is an analysis system for identifying and tracking threats for counter-terrorism, continuous surveillance, and monitoring of disturbed areas. PROJ V-LOGGER V-logger Vehicle Tracking System It is an AI-based software tool for detecting and tracking civilian vehicles using licence plates. The system uses  AI, and deep learning algorithms for real-time vehicle detection from live video feeds. Face Recognition System under Disguise This algorithm has been trained for the face recognition system can see through several disguises like face masks, beards, moustaches, etc. Segmentation of Satellite Panchromatic Videos This module performs automatic land cover classification of grey-level satellite imagery to produce four land cover classes: water, forest, bare land and urban area, including buildings and vehicles. AI based 360° Surrounding View Monitoring System It is a 360-degree camera monitoring system that assists low-speed manoeuvring by providing the driver with a complete surrounding view of the vehicle in real-time. HUMS Ground Station It is a system for helicopter maintenance that uses recorded vibration data onboard the helicopter and processes it to compute various condition Indicators (CIs). The AI-based algorithm then uses these indicators to compute the remaining useful life. AI-Based Satellite Image Analysis It is an indigenised solution that provides comprehensive real-time geospatial situational awareness backed by cutting-edge R&D. AI-Based Technique for Prediction of Atmospheric Visibility Using a statistical concept called time series and ML algorithms, this system make visibility forecasts in real time. Chimera-22 Smart Camera This camera powers the next generation of vision analysis giving the best of worlds in processing power and resolution. Deepsight Canopy Inspection for Fighter Jets It is an AI-based machine vision system for automatic inspection of surface defects. IoT\/Smartcities Internet of Battle Things (IoBT): Smart Helmets This helmet comprises of an optical sensor mounted on the helmet of an active combat soldier, which in turn is connected to a wearable backpack computer of small form factor. Automatic Number Plate Recognition (ANPR) for Smart Cities ANPR technology is developed with high accuracy for Indian vehicles. The system can detect and recognise licence plates of different types of vehicles. The module is an integral part of Smart City solutions. AI Enabled Adaptive Traffic Optimization Solution It is a traffic optimisation solution that aims to reduce travel time, delay at traffic signals, and minimise the average waiting time throughout the network. Lethal Autonomous Weapon Systems Smart – Counter Measure Dispensing System (CMDS) CMDS protects aircraft against incoming radar-guided missile threats by dispensing flares\/chaffs based on the inputs received from various sensors installed on aircraft. Adaptive Intelligent Front Towing Solution for Artillery Gun The system provides synchronisation of speed-dependent and turning radius-dependent movement between a driver component and a driven component. Logistics and Supply Chain Management PRO-HM+ (AI-in SCM and Logistics) It is an aircraft health monitoring software that uses data analysis techniques to identify trends, patterns, and relationships will predict aircraft behaviour, equipment failure and other future events. Manufacturing and Maintenance Artificial Intelligence Based Predictive Maintenance Suite The suite uses historical maintenance records, machine sensor data, and weather data to determine when a machine will require maintenance. The system can predict potential failures in real-time, thereby decreasing machine downtime. Predictive Maintenance for Gun Fire Control Systems This real-time predictive system has been developed and deployed to predict better maintenance of PCBs by detecting anomaly\/failures\/early warning and asset conditions parameters. AI Based Predictive Maintenance of Delhi Metro Rail Equipment This tool predicts equipment failure based on equipment health, historical patterns of failures, and periodic maintenance schedules. Alloy Development through Artificial Intelligence The project aims at designing a framework for new alloy design using AI and can be used with specific applications for any alloy systems, like superalloys and titanium alloys. Predictive Maintenance of Mining Equipment Through Data Analytics and Telematics Enabled System The system uses modern, analytical techniques to reduce the cost of maintenance and downtime. It does so by early identification of imminent equipment failure. Condition Monitoring System for Shipboard Equipment (Main Engine) GSL and Infosys have jointly developed the system for OPV class ships using AI. The system predicts the time of likely failure enabling predictive maintenance of main engines installed on board. AI-Enabled Evaluation of Welding Defects in X-rays of NDT With an accuracy of 90 per cent, the software has replaced the manual inspection system of weld defects in X-rays of NDT. Operational Data Analytics AI-Enabled Fake News Detector as Part of Social Media Analytics BEL has developed this advanced computational models using ML and NLP to identify fake news\/sourced from various social media and sound the alarm to users. Operational Data Analytics for Naval Platform The platform has stored sensor data from all the naval ships. It uses big data analytics to predict sensor performance and recommends the best naval ships for mission categorisation. AI Enabled Automatic Information Extraction and Synthesis BEL developed the system for machine comprehension, video summarisation, fake image detection and speaker diarisation. Perimeter Security Systems Sarvatra Pehchaan – AI Based Intrusion Detection & Integrated Command Station SarvatraPechaan considers a fusion of sensory feeds from multiple imaging systems and sensors on a single dashboard. AI-based video analytics techniques are used for anomaly detection, allowing rapid intervention. AI-Enabled Forensic Search for Videos Camera-stream recorded videos are processed using AI techniques to identify the objects in the scene. AI-Enabled Gesture Recognition The system uses deep learning technology to identify various human gestures like human walking, crawling, crouching, with or without a gun. It can be easily integrated into a network of IP-enabled cameras. Audio Doppler Based Object Classification The system determines the nature of a target moving in the radar’s field of view using its doppler return signal. Process flow Automation for large systems AI-Based Automation of Water Sprinkling System The system will monitor dust concentration in open cast mines, process the data using artificial neural networks and actuate the output to switch on the corresponding solenoid driver to do the desired water sprinkling operation. AI-Based Lighting Control system on HEMM (Heavy Earth Moving Machinery) The system uses AI for human detection. It turns on the equipment lights when a person approaches the equipment to ensure adequate illumination for the operator to climb up to the equipment safely. AI-Enabled Weld Inspection Machine with Computerised Radiography-(AI-RT) The tool is the replacement of X-ray film technique that is used currently in conventional radiography for weld inspection. AI-Enabled Weld Inspection Machine with Advanced Phased Array Ultrasound Technique-(AI-UT) This tool employs an advanced phased array ultrasound technique using total focusing approach tools. It is augmented with AI data analytics. This can replace X-ray techniques that are used generally for weld inspection. AI-Based Automated Bore Cleaning This technology facilitates an automated inspection and cleaning of large calibre artillery and tank guns. Brainbox It is a technology that can be customised to undertake structural integrity analysis of various types of metal structures and fittings. Simulators\/Test equipments Development of Artificial Intelligence-based training modules for technicians for operation and maintenance of SU – 30 MKl aircraft The solution enables the operators to take up the most challenging procedures for hands-on virtual training with real-time rectification and performance analytics. Speech\/voice analysis system using NLP AI-Based Mandarin Translators It is a wearable device for speech-to-speech translation that accepts input in one source language and converts it into an equivalent audio\/speech of the target language. AI-Based Offline Language Translator It is the only solution in India wherein a GPU-based algorithm can be deployed in CPU-based hardware. Being offline does not entail any kind of dependency. Speech-to-Speech Translation The system will enable the understanding of the source language without requiring a translator. AI-Enabled Voice Transcription Software This module works uniquely by applying speaker and language recognition functionalities to explore the capability of revealing significant insights into the dynamics of hidden communication. Voice Activated Command System (VACS) VACS recognises the pilot’s voice commands and sends the recognised voice command codes to the mission computer for carrying out the Pilot’s intended action. AI-Powered Language Translation Platform A platform capable of processing all major forms of intelligence data – audio, video, text and images. Intelligent agencies and frontline positioned units can use the device to decode foreign language messaging.To learn more about the tools, click here.","excerpt":"Another 100 projects are in various stages of development","categories":["AI Trends"],"tags":["Rajnath Singh AI"],"author_name":"Tasmia Ansari","publish_date":"2022-07-18T11:00:00","publication_year":"2022","word_count":2447,"keywords":["artificial intelligence","AI","Rajnath Singh AI","neural network","ML","MLOps","NLP","Ray","Aim","deep learning","analytics"],"extracted_tech_keywords":["AI","artificial intelligence","ML","deep learning","neural network","NLP","analytics","MLOps","Aim","Ray"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/75-massive-tools-launched-at-ai-defence-exhibition\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":65513,"title":"Mervin, Jarvis And Isaac: Highlights From NVIDIA GTC Keynote 2020","content":"Last week, NVIDIA CEO Jensen Huang gave a keynote from his residence, revealing all the exciting things his company has been working over the past year. Let’s take a look at the key announcements made during the keynote speech: AI Acceleration With NVIDIA Ampere “The powerful trends of cloud computing and AI are driving a tectonic shift in data centres designs with GPU-accelerated computing,” said Jensen Huang, CEO of NVIDIA With more than 54 billion transistors NVIDIA A100 is the world’s largest 7-nanometer processor. It is the first GPU based on the NVIDIA Ampere architecture, providing the greatest performance leap of eight generations of GPUs. 18 of the world’s leading service providers such as Alibaba Cloud, Amazon Web Services, Cisco, Dell Technologies, Google Cloud, Hewlett Packard Enterprise, Azure and Oracle are using it. DGX SuperPOD delivers 700 petaflops NVIDIA is also shipping the third generation of its DGX AI system. It is the world’s first 5-petaflops server, and each DGX A100 can be divided into as many as 56 applications, all running independently. NVIDIA Merlin The latest generation of deep-learning powered recommender systems enables companies to better target users. NVIDIA’s Merlin recommender application framework promises to make GPU-accelerated recommender systems more accessible with an end-to-end pipeline for deploying AI models. NVIDIA GPUs have long been used to accelerate training time for neural networks — sparking the modern AI boom — since their parallel processing capabilities let them blast through data-intensive tasks. These systems will be able to take advantage of the new NVIDIA A100 GPU, built on NVIDIA Ampere architecture, so companies can build recommender systems more quickly and economically. NVIDIA Jarvis NVIDIA’s Jarvis, a GPU-accelerated application framework allows companies to use video and speech data to build state-of-the-art customised conversational AI services. “Conversational AI is central to the future of many industries to understand and communicate with nuance and contextual awareness.”Jensen Huang, founder and CEO of NVIDIA Jarvis also has the advantages of the new NVIDIA A100 Tensor Core GPU for AI computing and the latest optimisations in NVIDIA TensorRT for inference. Now, for the first time, it is possible to run an entire multimodal application faster than the 300-millisecond threshold for real-time interactions. NVIDIA Isaac NVIDIA Isaac robotics platform was selected by the BMW Group to enhance its automotive factories by revamping its logistics, which uses robots built on advanced AI computing and visualisation technologies. “BMW Group’s use of NVIDIA’s Isaac robotics platform to reimagine their factory is revolutionary.”-Jensen Huang BMW Group aims to enhance logistics factory flow more efficiently. Once developed, NVIDIA’s system will be deployed at the BMW Group factories worldwide. BMW will also be using the following: DGX AI systems along with Isaac to help in simulation technology to train and test the robots; NVIDIA Quadro GPUs to render synthetic machine parts to enhance the training AI-enabled robots built on the Isaac SDK, powered by NVIDIA Jetson and EGX edge computers. NVIDIA Drive “Autonomous vehicles are one of the biggest computing challenges of our time, an area where NVIDIA continues to push forward with NVIDIA DRIVE,” said Huang NVIDIA DRIVE will use the new Orin System-on-Chip that is embedded with its Ampere GPU to offer a 5-watt ADAS system as well as scale up to a 2,000 TOPS, level-5 robotaxi system. This will make the work of automakers easy as now they only have to work on a single computing architecture and software stack to build AI. “NVIDIA accelerated computing to save lives is the perfect example of our organisation’s purpose, and we build computers to solve problems normal computers cannot,” Huang said.","excerpt":"Last week, NVIDIA CEO Jensen Huang gave a keynote from his residence, revealing all the exciting things his company has been working over the past year. Let’s take a look at the key announcements made during the keynote speech: AI Acceleration With NVIDIA Ampere “The powerful trends of cloud computing and AI are driving a […]","categories":["Deep Tech"],"tags":["big data trends and challenges","NVIDIA GPU"],"author_name":"Ram Sagar","publish_date":"2020-05-19T16:00:34","publication_year":"2020","word_count":598,"keywords":["Go","NVIDIA GPU","AI","neural network","cloud computing","Azure","R","Modal","Aim","GAN","Tecton","big data trends and challenges"],"extracted_tech_keywords":["AI","neural network","Aim","Tecton","cloud computing","Azure","R","Go","GAN","Modal"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/nvidia-gtc-keynote-2020\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":42476,"title":"10 Most Prominent Data Science Academicians In India: 2019","content":"Analytics India Magazine has compiled a list of the top 10 most prominent analytics and data science academicians in India for the year 2019. In this annual list, we bring the academicians who have made a difference in the young analytics and data science professionals by offering exceptional guidance, delivering insightful sessions and out-of-the-box teaching methodologies to equip them with the concepts. The faculties including in the list have PhD degrees in analytics, data science and artificial intelligence domains. The list is in alphabetical order. Read our last year’s list here 1| Dr Abhinanda Sarkar Dr Sarkar is the Academic Director at Great Learning for Data Science and Machine Learning Programs. He has a PhD in Statistics from Stanford University. He has taught mathematics at MIT and has also led industry roles, including 10 years of an illustrious career at GE Technology Centre. Dr Sarkar has publications, patents and technical leadership in areas such as experimental physics, computer vision, text mining, wireless networks, e-commerce, credit risk, and more. Research interests: Statistics and its sub-domains Teaching Experience: 28 years PhD: Stanford University in Statistics Key Achievements as a Data Science Academician: Taught at some of the most renowned institutions such as Stanford, MIT, IIM-B, ISI, IISc. He has been instrumental in designing and conducting numerous corporate training sessions for technology and business professionals. He has been the recipient of awards such as IBM Invention Achievement Awards, Radhakrishan Mentor Award from GE India, Discovery Award (NESTA, UK) and more. 2| Dr Chandrashekhar Ramanathan He is a faculty member at IIIT-B since 2007, currently serving as the professor and dean (academics) at IIIT-B and upGrad. With a primary focus on software engineering, application development and database, he also has over 10 years of work experience in extensive application software development, semi-structured databases in large MNCs. He is one of the primary instructors in the PG Diploma in data science being offered jointly by IIIT-B and upGrad. Research Interest: Data sciences, software architecture and technology for education, semi-structured databases, software engineering, enterprise content management and knowledge management Teaching Experience: Over 10 years Industry Experience: Over 10 years PhD: Mississippi State University, in Computer Science Key Achievements As A Data Science Academician: He has contributed to the design of large online diploma and certification programmes in data science. He is a member of government panels and task groups for the promotion of data science education. He has supervised student research leading to the award of Doctorate degrees and has many best paper awards and recognition. 3| Dr D Pradeep Kumar He is a senior data scientist and analytics trainer with a strong experience in R&D and training. With a keen interest in machine learning, data mining, and other areas, he has over a decade of experience in academia. He has spent his career as an academician at engineering college for over 3 years and as analytics trainer for more than 7 years. He is currently an analytics trainer with Imarticus Learning. Research Interests: Time series forecasting, Machine learning, Deep Learning and NLP Teaching Experience: 10 years PhD: University of Hyderabad and IDRBT in CS Key Achievements As A Data Science Academician: He proposed seven novel soft computing hybrids for time series prediction which has been carried in reputed journals and international conference proceedings. 4| Dr J.B Simha Dr Simha has a strong industry and academic experience in his over a decade long career. He has worked as an analytics consultant for firms such as Siemens Information Systems apart from holding positions like research assistant and CTO. He is currently the CTO and director of analytics at Abiba Systems. He has more than 50 papers published in national and international journals. Dr Simha has strong technical skills in artificial intelligence, machine learning, deep learning, social media analytics, big data and more. He is currently the Chief mentor, Analytics, REVA Academy for Corporate Excellence at REVA University Research Interest: Machine learning, Deep learning, Soft Computing, Data Science, Visual Analytics Teaching Experience: 10 years Industry Experience: 20 years PhD: Bangalore University in Data mining Key Achievements As A Data Science Academician: He has produced and supervised four PhDs. He has a strong focus on bridging the teaching gap through research, consulting and training. He also has five best paper awards and three patents pending on his name. 5| Dr Manish Gangwar Dr Gangwar is a tenured faculty in the marketing area at the Indian School of Business. He is currently the executive director of Applied Statistics and Computing Lab (ASC Lab) and business analytics programme at ISB. With a PhD in Management Science, his research interest lies in exploring marketing issues using quantitative models. He has developed a theory for optimal price promotion for frequently purchased goods and optimal nonlinear tariffs for subscription models. Research Interests: Pricing analytics, Customer analytics, New Product Development, Retail analytics, and Celebrity advertising Teaching Experience: Over 10 years PhD: The University of Texas at Dallas in Management Science Key Achievements As A Data Science Academician: He is a well-respected researcher in the area of quantitative marketing and has established 12 papers in top journals worldwide. He has also been called for reviewing papers submitted to various top journals. 6| Dr P Sridhar Dr Sridhar is a prominent faculty member and has been instrumental in the educational pursuits of more than 3000 students, helping them in their educational endeavours in some of the top universities in the world. He has over two decades of experience ranging from applications development to being a scientist for the Defence Metallurgical Research Laboratory, India. He has over 34 publications and has received three honours and awards for his excellence in the field of data science. He is the co-founder, executive VP – academics and mentor at INSOFE. Research Interest: Statistics, Statistical Modeling, Data Visualizations, Ethical IP and Legal Issues for Data Scientists, Machine Learning, Decision Trees and Association Rules Teaching Experience: 20 years PhD: The University of Texas at El Paso in Materials Sciences and Engineering Key Achievements As A Data Science Academician: Six Sigma Black Belt trained, Worldwide Lifetime Achievement (2017), Who’s Who in the World (2014, 2016), Member, Editorial Advisory Board: Handbook of Materials Structures, Properties, Processing and Performance by Dr Lawrence E Murr, Trained more than 150 Corporates. 7| Dr P.K. Viswanathan With rich and varied experience across academia, research, industry, training, and consulting, Dr PKV, currently teaches MBA students subjects such as business statistics, operations research, business analytics, predictive analytics, and more at Great Lakes Institute of Management. He has more than 15 years of industry experience where he has held senior management positions in Ballarpur Industries and JK Industries. Research interests: Currently, delving deep into Scoring Models for Analytics Decisions, Artificial Intelligence and Machine Learning Analytics Teaching Experience: Over 20 years PhD: Madras University in Operations Research Key Achievements as a Data Science Academician: He has been one of the best faculties at Great Lakes based on teaching feedback. In consulting assignments, he has developed and implemented analytical models successfully for business solutions. He has authored four books till date and has published research articles in reputed national and international journals. 8| Dr Prithwis Mukerjee Dr Mukherjee is an engineer by education, a teacher by profession, a programmer by passion and an imaginer by intention. He loves to live at the creative edge where technology opens up new opportunities for doing business. He has a strong industry and academic experience where he has worked with companies such as Tata Steel, Tata IBM, PWC, and others. As an extension of his intense involvement with in-house training programs he switched careers and entered academia. After a four year stint as a tenured Professor and Member of the Senate at IIT, Kharagpur, India, he is now the Director at Praxis Business School, Calcutta Research Interests: Application of Artificial Intelligence in Business and Social problems Teaching Experience: 11 years PhD: University of Texas at Dallas, in Management Science Key Achievements As A Data Science Academician: He has designed curriculum for full-time PGP in Data Science for Praxis. He is a part of the team that designed weekend, full-time program in Cyber Security. He has authored a book titled Chronotantra which is an original science-fiction novel about a technocratic society managed by AI. 9| Dr Sakshi Babbar Dr Babbar has a rich experience of working in both industry and academics, having worked with companies such as Microsoft Research Asia, Beijing on interesting research problems. On the academics front, she has worked with reputed universities as a senior faculty and has guided PhD, Masters and Undergraduate students in the area of machine learning. She is an active researcher and has written several conference and journal papers of repute. She has a doctorate from the University of Sydney, Australia in the area of machine learning and data mining. She is currently the AVP, Machine Learning and AI at Careers of Tomorrow, Amity University Online. Research Interest: Machine Learning, Big Data Analytics, Probabilistic Graphical Models Teaching Experience: 9 years PhD: The University of Sydney in Data Mining Key Achievements As A Data Science Academician: Decent research outputs in the area of medical and environmental sciences. 10| Dr Santanu Sinha With over 12 years of experience in designing and developing analytics solutions for complex business problems, he brings a strong blend of industry and academia along with business consulting. Dr Sinha is a lead data scientist at HPE and serves as visiting faculty and research advisor to AnalytixLabs, IIT Kharagpur, and SP Jain Institute of Management. He has been associated with AnalytixLabs since 2014 and is a key partner in curating advanced analytics courses in Optimisation, ML and Applied AI. Research Interest: Data Science, Artificial Intelligence (Machine Learning & Deep Learning), Game-Theory, Evolutionary Computations, Simulation, and Optimization Teaching Experience: 7 years Industry Experience: Over 12 Years PhD: Department of Industrial Engineering & Management, IIT Kharagpur in Supply Chain Management Key Achievements As A Data Science Academician: He has been a key partner in curating advanced analytics course at AnalytixLabs. He has co-authored 20+ research papers and book chapters published in reputed international journals and conference proceedings.","excerpt":"Analytics India Magazine has compiled a list of the top 10 most prominent analytics and data science academicians in India for the year 2019. In this annual list, we bring the academicians who have made a difference in the young analytics and data science professionals by offering exceptional guidance, delivering insightful sessions and out-of-the-box teaching […]","categories":["AI Features"],"tags":["pgp program in data science"],"author_name":"Srishti Deoras","publish_date":"2019-07-15T12:35:24","publication_year":"2019","word_count":1684,"keywords":["data science","machine learning","artificial intelligence","AI","ML","computer vision","RAG","NLP","pgp program in data science","deep learning","analytics"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","NLP","computer vision","data science","analytics","RAG"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/10-most-prominent-data-science-academicians-in-india-2019\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10163809,"title":"How Do Teachers Train Kids for Jobs That Don&#8217;t Exist Yet, Asks Shashi Tharoor","content":"At the Invest Karnataka 2025 event, Lok Sabha member of Parliament Sashi Tharoor called for a fundamental reevaluation of how skills are taught in schools in India to better prepare students for the rapidly evolving intelligence age. Tharoor spoke at a panel with George A Papandreou, former prime minister of Greece, on the theme ‘Thriving in turbulence: How nations can build lasting resilience’. The leaders recognised AI-driven disruption as one of the key challenges that nations must navigate for long-term growth. Citing a report from the Oxford Martin School, Tharoor highlighted that 30% of jobs in 2030 will be jobs that don’t exist today. This poses a challenge for educators. “How do you tell a teacher to train a kid for a job that doesn’t exist yet?” Tharoor asked. “I say this to teachers all the time: Don’t teach the young people what to think…Teaching them how to think is what’s going to equip them for this uncertain world where they’re going to be confronting facts, ideas, and information they’ve never encountered before,” he added. According to the MP, this would, by extension, retrain and empower children so they are not left behind by AI-driven economic changes. The Crisis in Education The traditional education system is outdated and based on an industrial-era model, suggested Papandreou. Working with AI is the future. According to Tharoor, AI will make rote learning obsolete, requiring an education system that fosters creativity, problem-solving, and adaptability. He said that India, like several other countries, is not prepared for the speed of disruption and faces the paradox of having the world’s largest population of educated unemployed. This highlights the mismatch between education and employment opportunities. The transformation of the labour market was a stressing point at the Artificial Intelligence Action Summit. US vice president JD Vance suggested that students should be taught how to manage and interact with AI, portraying a shift in AI policy. Instead of focusing on regulation or job loss, the focus should be on the need for AI education to prepare future generations for the evolving workforce. Similarly, Indian Prime Minister Narendra Modi stressed the need for upskilling as India moves towards an AI-driven economy, a message previously reflected in the Union Budget 2025. Anthropic, the company behind the Claude model series, recently released a report which highlighted AI’s influence on the workplace. The findings revealed that approximately 36% of all occupations incorporate AI for at least a quarter of their tasks. Moreover, 57% of AI applications enhance human capabilities, while 43% focus on automation. However, only 4% of occupations rely on AI for at least 75% of their tasks. The study identified software development and technical writing as the primary areas where AI is utilised. In contrast, AI plays a minimal role in tasks that involve physical interaction with the environment. This underscores the importance of building skills that are suited for a post-singularity world, where AI becomes deeply integrated into daily life and work. The Future of Jobs Report 2025, released by the World Economic Forum, states that 39% of core job skills are expected to change, and the future of work demands agility. The key question is whether we can empower students to solve major problems with AI as a tool rather than fear being replaced by it throughout their education. Challenging Conventional Thinking Tharoor pointed out that, in the past, most people might have believed a car could only run on petrol. Musk questioned this idea and innovated by creating electric cars with batteries instead of relying on fuel-powered engines. This illustrates how even the most fundamental assumptions, like the need for petrol to power cars, can be challenged and, eventually, become a catalyst for innovation. The tech revolution is also closely linked to a shift in mindset. Tharoor also emphasised that seemingly absurd questions should not be discouraged, as innovation arises from those out-of-the-box ideas. Notably, a prototype of flying taxi startup Sarla Aviation was available at the summit. “I think at some point here in Bengaluru, you’re going to fly to the airport. It’s not going to take an hour; it’s going to take five minutes,” said Google X founder Sebastian Thrun during his panel discussion at Invest Karnataka. Tharoor and Papandreou spoke in favour of economic reforms to redistribute wealth, especially as humanity enters the AI age. There is a need for redistribution policies, a focus on citizen empowerment, and creating more equitable opportunities. Several experts, including OpenAI CEO Sam Altman, Y Combinator CEO Garry Tan, Anthropic CEO Dario Amodei, and venture capitalist Vinod Khosla, have called for the need for Universal Basic Income (UBI) as we near this new reality. Since the launch of DeepSeek, companies have been working with a sense of urgency to accelerate AI development. Nearly all the leading labs have emphasised the upcoming phase, where AI’s potential will greatly benefit humanity and handle most of the economic tasks. The release of Grok 3, described by xAI co-founder Elon Musk as the “smartest AI on Earth”, is also set to launch tonight.","excerpt":"Citing an Oxford report, Tharoor highlighted that 30% of jobs in 2030 will be jobs that don’t exist today.","categories":["AI Features"],"tags":["AI (Artificial Intelligence)"],"author_name":"Aditi Suresh","publish_date":"2025-02-17T11:44:55","publication_year":"2025","word_count":840,"keywords":["Anthropic","Grok 3","artificial intelligence","OpenAI","AI","R","RAG","Ray","xAI","AI (Artificial Intelligence)","Redis"],"extracted_tech_keywords":["AI","artificial intelligence","OpenAI","Anthropic","Grok 3","xAI","Ray","RAG","Redis","R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-do-teachers-train-kids-for-jobs-that-dont-exist-yet-asks-shashi-tharoor\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":62389,"title":"What Is Narrow AI &#038; How It Is Different From Artificial General Intelligence","content":"When Alan Turing first thought of coming up with machines that could think like humans, he was probably thinking about machines that could one day make the life of human beings easier. Fast forward 70 years, and AI has been able to perform tasks that have undoubtedly made life more comfortable. Conversational AI, flying drones, bots, language translation, facial recognition, etc., are some of the most promising AI applications we have today. But these fall under Narrow AI rather than the Artificial General Intelligence, which is something different. What Is Narrow AI? As the definition goes, narrow AI is a specific type of artificial intelligence in which technology outperforms humans in a narrowly defined task. It focuses on a single subset of cognitive abilities and advances in that spectrum. Over the years, narrow AI has outperformed humans at certain tasks. These include calculations and quantification that have been performed more efficiently with this technology. Today, it has also outperformed human beings in complex games like Go and chess, along with helping make intelligent business decisions, and more. After narrow AI trumped human performance, the next step came in the form of general AI. General AI vs Narrow AI When AI was first explored, researchers had one thing on their minds – to create a system that can learn tasks and solve problems without explicitly being instructed on every single detail. This system should also be able to perform these tasks with reasoning, abstraction, and should also be able to transfer knowledge from one domain to another. But with time, scientists have struggled to create an AI that can satisfy all these requirements. As the years went on, the original idea of AI, where the system is required to imitate the human brain and its thinking process, found itself in a new category altogether – a different type of AI called General AI or Artificial General Intelligence (AGI). Researchers still believe the idea of AGI becoming practical is decades away. Much of the truth to that statement comes from the fact that today’s AI systems are not even able to perform tasks that a human child can. So, on the road to creating an AI system that can imitate humans, scientists and researchers have created many useful technologies. Each time such a technology is created, it is touted as a breakthrough, and the next time something more useful and practical is created, creating a benchmark for coming technologies. Narrow AI is something that encompasses all these useful technologies. As the definition goes, narrow AI is good at performing a single task – or a limited range of tasks – and at times, can outperform human beings. But the problem with narrow AI is that as soon as it is put under a different setting, or is made to perform a task that is different from what it excels at, it fails. They are not able to transfer their learning from one field to another. For example, if DeepMind’s AlphaStar is made to play a different game, then it might not give the same kind of performance that it did in StarCraft 2 (Grand Master). Some Barriers Faced Narrow AI applications have a lot of hard coded logic or parameters with pre-trained data sets, which are not that effective when it comes to real-time adaptive learningThe narrow AI systems come with a wide range of architecture, algorithms, and data representations that are incompatible and impossible to combineTypical AI designs require certain selected competencies, while missing others that are crucial to AGIA lot of environments do not meet the data requirements and accuracy that is needed by narrow AISometimes, customers are not receptive to narrow AI technologies. This is particularly true for industries like hospitality and healthcare.","excerpt":"When Alan Turing first thought of coming up with machines that could think like humans, he was probably thinking about machines that could one day make the life of human beings easier. Fast forward 70 years, and AI has been able to perform tasks that have undoubtedly made life more comfortable. Conversational AI, flying drones, […]","categories":["AI Features"],"tags":["AI Algorithms"],"author_name":"Sameer Balaganur","publish_date":"2020-04-26T14:00:00","publication_year":"2020","word_count":624,"keywords":["Go","artificial intelligence","programming_languages:R","AI","programming_languages:Go","AI Algorithms","R"],"extracted_tech_keywords":["AI","artificial intelligence","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/what-is-narrow-ai-how-it-is-different-from-artificial-general-intelligence\/","complexity_score":3,"technical_depth":6,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10019334,"title":"Indian Ethical Hacker Manan Shah Develops World&#8217;s First AI-Powered Solutions To Combat Online Piracy","content":"Manan Shah is a prominent ethical hacker and the founder of cybersecurity company, Avalance Global Solutions. As a child, Manan showed great interest in maths and computer science. He also used to fix computers for his neighbours and relatives, as a hobby. He picked up programming languages and learned hacking via online videos. Manan’s hacking journey started in 2009 after attending a seminar on ethical hacking. Later, Manan participated in bug bounty programs of Google, Facebook, etc to brush up his chops. He also got involved in freelance projects that helped him work towards his vision of building a brand and taking it global. At the age of 21, Microsoft recognised Manan as top 100 security researchers in the world. He had the opportunity to solve cybercrime cases for some of the biggest MNCs such as Twitter, Yahoo!, Facebook, Nokia, Blackberry, PayPal, Skype, Google, Apple, Microsoft, among others. With the help of a couple of friends, he set up a global brand, Avalance, an online cybersecurity platform that serves clients across verticals. The company has currently expanded its operations to the UAE, along with offices in India and the US, and has witnessed a triple-digit revenue growth year-on-year. “My goal has always been to significantly impact the cyber forensics space and contribute towards the overall holistic development of the constantly evolving cybersecurity landscape globally,” said Manan. Analytics India Magazine got in touch with Manan Shah to understand his journey and know more about the cybersecurity landscape and discuss how his company, Avalance Global Solutions, is leveraging artificial intelligence and analytics to combat online piracy and other cybersecurity threats. Here are the excerpts: Also Read: 8 Free Resources To Learn Ethical Hacking With Python AIM: Could you shed some light on the biggest problems for Indian companies that put them at risk of cyber-attacks? Manan Shah: India is among the leading countries cyber-attackers target. Indian companies are a soft target for cyber -attacks as hackers steal important data and leak it in public platforms. According to a survey, with more people working from home, there has been a 25% rise in cyber threats. Today, cybersecurity is a part of our daily life; from opening a browser, doing online transactions or visiting a doctor, where your healthcare information needs to be protected. Even in the news, you keep hearing about the latest security breaches. The pandemic has made work from home a norm with increased technological infrastructure dependence. Many online calls and video conferences have lately been hacked to steal important data of companies. Additionally, cybersecurity has been experiencing negative unemployment for several years. However, in India, we face several challenges including a shortage of cybersecurity professionals and lack of regularised data infrastructure. There needs to be a marked shift in education concentrated on computer science with a focus on security. AIM: How AI and ML are helping in cybersecurity processes and fighting against advanced cyber crimes? Manan Shah: The most common cybersecurity tools such as anti-malware software or login audits won’t be enough to control the cyber threats. The security risks in applications and web have grown at an unprecedented rate. The advancement of technology and the use of mobiles, web and the Internet of Things has led to the rise of vulnerable threats, due to which the data of many businesses and companies are at risk. Artificial intelligence and machine learning can fight against cybercrimes as it helps by processing and analysing the huge amount of data and its behavioural patterns. Humans cannot process and analyse data as quickly as AI or ML automated engines. I believe both AI and ML can improve cybersecurity measures. While password protection or multi-factor authentication can be implemented in cybersecurity, machine learning algorithms can help classify the strength of a password and suggest complex and hard to guess passwords. It can work brilliantly with a biometric login that uses AI to detect physical traits. AIM: What are the solutions Avalance use to help companies? Manan Shah: My company has developed the SSORR methodology which secures business environments at all levels. From bottom to the top, we have built a methodology that provides value to the business and management team whilst ensuring the gaps with the technical team are bridged to provide a fully integrated approach to cybersecurity. Last year, we launched the world’s first AI-based solutions to stop online piracy. Avalance’s proprietary web crawling technology is capable of simulating human behaviour in searching for infringing content across the web. A specialised team of global copyright experts along with automated multi-step mitigation techniques allow us to rapidly eliminate large volumes of infringing content instantly. AIM: How is the company leveraging data and analytics? Also, how is automation embedded in your products? Manan Shah: At Avalance Global Solutions, we understand the need for data and analytics in tackling cyber crimes. We apply advanced analytics and business intelligence tools to huge volumes of data to get detailed insights. In our analytics process, there are various aspects like machine learning algorithms, statistics and predictive modules driven by high-performance computing systems. I believe data and analytics can help discover valuable information that can help companies make smart decisions about anything in general. As far as automation is concerned, we provide automated analysis of alerts that help analysts keep a tab on the dangers around business operations. A machine learning-driven automation engine can process and analyse data much quicker than humans. ML automation will soon be embedded in our products and services. AIM: Which industry verticals Avalance Global Solutions target in India and across the globe? Manan Shah: My company Avalance Global Solutions serves the clients in India and abroad across different industries including food and beverage, technology, defence, medicine, banking, government and many other sectors. We started the company in 2015 and have expanded our footprint across the globe. Avalanche Global Solutions is going strong, and has got clients from India, Dubai and the USA. AIM: What are your views on the lack of cybersecurity talent in India? And what is your hiring and training strategy? Manan Shah: India has the largest IT talent pool in the world. However, the country still lacks skilled cybersecurity professionals. Many experienced cyber experts are working for global giants in Silicon Valley, California. The new bundle of cybersecurity talents need the right exposure in India. Indians have raised the bar of IT across the world, and now it’s time to make India a cyber hub. Our hiring and training strategy for Avalance Global Solutions is simple — allow new professionals to showcase their skills and talent. We believe in training the employees and giving them 360-degree exposure as a cybersecurity analyst, security architect and cybersecurity manager. We aim to train our employees in different fields like access management, cryptography, risk auditing, network security, data analytics, software development and forensic sciences. AIM: How to start a career in cybersecurity in 2021? Can you share some tips for current entrepreneurs and cybersecurity professionals? Manan Shah: The first thing to start a career in cybersecurity is to gain complete knowledge about computers, softwares, hardware and the Internet. The modern digital era has made everything accessible from home. In my opinion, one can start a career in cybersecurity just by enrolling for online courses. There are a number of cybersecurity courses available in the market. The aspiring cybersecurity experts can make proper use of the Internet and attend free webinars to get a gist about the field. The only tip I would want to give to all the budding entrepreneurs is to stay up to date with the latest market trends. Understand what’s new over the Internet, because every day there is an innovation over the web. You must be proactive and prepared for the consequence. COVID-19 pandemic was something we never expected, and no employer ever thought that work from home would become a norm. As an expert, I would want everyone to enable alerts as many fraudsters practise phishing. Lastly, every entrepreneur or an expert must train their employees to be secure on the Internet. AIM: What is Avalance Global Solutions roadmap for 2021? Manan Shah: The road is always under construction. Avalance Global Solutions is growing at a rapid pace. In 2019, we launched an AI-based anti-piracy solution, which was funded by businessman Raj Kundra. Last year, we expanded our operations to the UAE, and our revenue touched triple-digit despite the COVID pandemic. This year, we plan to grow bigger. I have always aimed to start an academy under Avalance Global Solutions and have plans to launch exclusive courses for the budding hackers and cybersecurity experts. My team is still in the development process, and we have a lot of other projects to dominate the global market in 2021.","excerpt":"Manan Shah is a prominent ethical hacker and the founder of cybersecurity company, Avalance Global Solutions. As a child, Manan showed great interest in maths and computer science. He also used to fix computers for his neighbours and relatives, as a hobby. He picked up programming languages and learned hacking via online videos. Manan’s hacking […]","categories":["AI Features"],"tags":["Ethical AI","Ethical Hacking","hacking tools","Interviews and Discussions","online network graph"],"author_name":"Sejuti Das","publish_date":"2021-01-31T16:00:00","publication_year":"2021","word_count":1455,"keywords":["hacking tools","Go","artificial intelligence","machine learning","AI","online network graph","Ethical AI","ML","RAG","Python","Aim","analytics","Ethical Hacking","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","analytics","Aim","RAG","Python","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/indian-ethical-hacker-manan-shah-develops-worlds-first-ai-powered-solutions-to-combat-online-piracy\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10053567,"title":"ThoughtSpot Raises $100M Series F Funding, $4.2B Valuation To Fuel The Modern Analytics Cloud","content":"Modern Analytics Cloud Platform, ThoughtSpot, recently announced that it raised a $100M Series F funding round at a new $4.2B valuation, more than doubling the company’s previous valuation and bringing the company’s total funding to $674M. The round was led by new investor March Capital and joined by existing investors Lightspeed Ventures, Snowflake Ventures, Khosla Ventures, Fidelity, Capital One Ventures, General Catalyst, Sapphire Ventures, and GIC. The new round comes two years after the company’s Series E, where the company raised $248M at a $1.95B valuation. Since then, ThoughtSpot has grown efficiently, with more than $200M in the bank prior to this funding round. In 2020, the company made a strategic decision to transform into a cloud-first company. The new funding will be used to accelerate two new growth vectors for ThoughtSpot with analysts, data engineers, and developers, the roles upon which the entire modern data stack rests. “One of the most challenging and most rewarding parts of being a CEO is helping a company stay focused ahead of the changing needs of its customers. That’s exactly what ThoughtSpot has done over the last year as we accelerated our growth as a cloud company. Our customers rely on us to break free from the tyranny of averages as they build their businesses on data. Today’s funding shows just how valuable the work we do with our customers truly has become,” said Sudheesh Nair, CEO, ThoughtSpot. Ajeet Singh, Co-founder & Executive Chairman, ThoughtSpot, said, “From the start, ThoughtSpot has been focused on a singular mission: to create a more fact-driven world. The growth of our cloud offerings has surpassed our own expectations, accelerating our progress toward that mission. We are pleased to see both cloud-native companies such as Cloud Academy and Frontify and large enterprises like T-Mobile make their data teams more productive and their businesses more data-driven with the Modern Analytics Cloud.” ThoughtSpot eyes new growth vectors and focuses on capabilities for developers and data professionals; the company also announced the launch of the new ThoughtSpot Data Workspace, a set of tools that brings Live Analytics to customers around the globe. Both announcements precede ThoughtSpot’s annual customer conference, Beyond, on November 16. The company has announced integrations and partnerships with leading players like Snowflake, Databricks, Amazon Web Services, Microsoft Azure, GoogleBigQuery, DataRobot, dbt Labs, Dremio, Starburst, and more, making ThoughtSpot ubiquitous across the modern data stack. ThoughtSpot’s live analytics offerings give customers an intuitive, interactive means to empower their entire organization to capitalize on innovations happening across the cloud ecosystem.","excerpt":"The new funding will be used to accelerate two new growth vectors for ThoughtSpot with analysts, data engineers, and developers, the roles upon which the entire modern data stack rests.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","analytics funding","Cloud Computing","Cloud Computing in IT Industry","Cloud Data AI","Cloud Platform","Funding","Machine Learning","thoughtspot"],"author_name":"Victor Dey","publish_date":"2021-11-16T12:53:55","publication_year":"2021","word_count":420,"keywords":["API","Funding","thoughtspot","Cloud Computing in IT Industry","R","Cloud Platform","RAG","analytics","Go","AI","Machine Learning","analytics funding","GAN","Cloud Data AI","Databricks","Cloud Computing","Azure","AI (Artificial Intelligence)","Snowflake"],"extracted_tech_keywords":["AI","analytics","RAG","Azure","Snowflake","Databricks","R","Go","API","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/thoughtspot-raises-100m-series-f-funding-4-2b-valuation-to-fuel-the-modern-analytics-cloud\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10123735,"title":"Bengaluru’s HSR is All About AI Startups","content":"Bengaluru, often dubbed the ‘Silicon Valley of India’, has always been synonymous with startups and innovation. In recent years, the city has embraced the AI trend, making it a significant player in the global AI landscape. Srikrishna Swaminathan, the CEO of Factor.AI, wrote on LinkedIn, “One building, and all AI companies. Factors.ai can claim to attracted all AI here, as we were the first occupants, Amal Mishra, Urban Vault.” Urban Vault, a building in HSR Layout, is currently housing five notable AI companies, namely Loop AI, Factors AI, Raga AI, Frinks AI, and Actyv AI, thereby, alluding to HSR’s pivotal role in nurturing AI startups. HSR: The AI Hub In an earlier conversation with AIM, Ganesh Gopalan, the CEO and co-founder of GNANI.AI, highlighted that they wished to establish the startup in HSR, calling it a hub for tech talent. Some of the vibrant AI startups buzzing in HSR Layout include MachineHack Gen AI, a GenAI startup for the developer community; Zolnoi Innovations Pvt. Ltd., which leverages AI to enhance production efficiency; and HyperVerge, a deep-tech AI company. Then there’s SigTuple, specialising in intelligent solutions for medical diagnosis using state-of-the-art AI techniques; Invento Robotics, creating intelligent robots for a range of applications; and Talview, offering an AI-powered recruitment platform for global enterprises. This unique ecosystem thrives due to the diversity of startups focusing on various sectors, including logistics, sales, developer tools, manufacturing, and fintech. The proximity of these companies fosters collaboration and innovation, making mergers and acquisitions more seamless. Source: LinkedIn AI Startups in Bengaluru Speaking to AIM at IGIC 2024, Sanjeev Kumar Gupta, the CEO of Karnataka Digital Economy Mission, said, “There are over 1,000 AI startups in Bengaluru. With initiatives like Beyond Bengaluru, we are also promoting AI startups, in Mysuru, Hubli, and Mangaluru.” The emergence of companies like Sarvam AI, which focuses on developing advanced AI models, reflects Bengaluru’s commitment to technological innovation. Krutrim AI, led by Ola’s founder Bhavish Aggarwal, has raised $50 million in funding at a $1 billion valuation and is now India’s first AI startup to reach the unicorn status. Innovation in AI isn’t limited to corporate giants. Startups like KOGO OS are disrupting markets with their AI operating systems, offering modular AI assistants tailored for diverse industries. Meanwhile, Karya AI is pioneering in the rural employment sector, leveraging AI to create job opportunities through tasks in local languages, with notable partnerships with tech giants like Microsoft and Google. These home-grown initiatives underscore Bengaluru’s status as a hub for AI innovation, promising transformative impacts across multiple sectors. Karnataka Government’s Support to Bengaluru Startups According to a report released by NASSCOM, Bengaluru is home to over 7,000 startups, solidifying its status as India’s leading startup hub and accounting for 20% of the country’s overall startup activity. Priyank Kharge, Karnataka minister for IT\/BT, recently unveiled a new scale-up program, Hypergrowth Global Karnataka, at London Tech Week. This initiative aims to accelerate the global commercialisation and international market expansion of the best tech companies in Bengaluru and Karnataka. The program offers local companies access to global mentors, expert scaling advice from leading executives, go-to-market support, and connections to potential new customers and investors to enhance their international growth capabilities.","excerpt":"Bengaluru is home to 1,000+ AI startups and over 7,000 startups in all.","categories":["Deep Tech"],"tags":["AI (Artificial Intelligence)"],"author_name":"Vidyashree Srinivas","publish_date":"2024-06-14T18:23:31","publication_year":"2024","word_count":533,"keywords":["Go","GenAI","AI assistants","AI","ML","Git","RAG","Aim","GAN","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","ML","GenAI","Aim","RAG","AI assistants","R","Go","Git","GAN"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/bengalurus-hsr-is-all-about-ai-startups\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10023116,"title":"Unravelling UK’s AI Strategy To Bolster Its Defences","content":"Ben Wallace, the UK’s Secretary of Defence, went on record to explain the Defense Command Paper, emphasising reducing the human element while reinforcing defences. This was proposed to be done with the aid of artificial intelligence. While deploying a defence review, the Secretary of Defence highlighted the necessity for a “digital backbone” that can be established by sharing data over cloud technologies. Orating a speech, Ben Wallace also pointed out that it’s understandable to focus on the number supporting the defence forces, but that also means deploying them at war zones with “Snatch Land Rovers” and tanks. However, on the other hand, the enemy is already advanced and has developed new ways to tackle such elements. To combat and be ready for future threats, the Defence Command Paper focuses on defence intelligence where it brings into light the importance of AI. Creating and maintaining this digital backbone, Ben Wallace commented that a whopping £1.5 billion would be invested by the UK Strategic Command for the corresponding advancements over the next decade. The 21st Century Chainmail Going into details, the digital backbone’s main purpose will be to share, utilise, and protect huge quantities of information via cloud networks. Using this information, new approaches to training with the help of simulations and synthetics can be developed. This, in turn, can provide a change in scenery the way personnel are educated in the means of warfare and combat. Focussing on the role of AI in the future, Ben Wallace stated to the media, “We will exploit a wider network of advanced surveillance platforms, classifications of data, and enhanced analysis using artificial intelligence.” Another motivation for identifying future threats with AI is to reduce the number of personnel deployed in warfare. It is not the first time that the UK leans on the power of AI to strengthen and revolutionise its armed forces. Earlier this month, various military advisors from the country pitched to harness the aid of artificial intelligence to forecast enemy behaviour, execute reconnaissance, and receive real-time data from various combat zones. Zooming out Looking at the things happening on a global scale, it is clear that the UK is cloning the US’s move of hosting a new initiative that brought together 13 countries that share the same ideology of implementing AI in the sectors of defence. Unfortunately, India was not invited to join the partnership. Irrespective to that, India has ensured that the Centre of Artificial Intelligence and Robotics (CAIR), a division within the DRDO, has created various AI-based devices aimed at net-centric communication systems. While on the other hand, to support surveillance and reconnaissance, CAIR has invented snake robots, Hexa bots, and sentries equipped with a wide library of AI-based algorithms meant for image and video recognition. Nevertheless, like all other countries, India is also facing challenges in employing AI in its defence sectors. It is a new phenomenon to various policymakers who are still unclear about the mission and vision of equipping armed forces with AI. Also, the lack of robust infrastructure, which is the bare necessity for introducing such unique technology, is another reason why India is still trailing in terms of defence compared to the likes of Russia, the USA, and China. But with AI slowly being adopted by more and more countries, it will be interesting to see how India joins the ranks as they will be forced to before long.","excerpt":"Ben Wallace, the UK’s Secretary of Defence, went on record to explain the Defense Command Paper, emphasising reducing the human element while reinforcing defences. This was proposed to be done with the aid of artificial intelligence. While deploying a defence review, the Secretary of Defence highlighted the necessity for a “digital backbone” that can be […]","categories":["IT Services"],"tags":[],"author_name":"Peter Mathew","publish_date":"2021-03-31T11:00:00","publication_year":"2021","word_count":565,"keywords":["Go","artificial intelligence","programming_languages:R","AI","programming_languages:Go","Git","Aim","ViT","ai_applications:robotics","R"],"extracted_tech_keywords":["AI","artificial intelligence","Aim","R","Go","Git","ViT","programming_languages:R","programming_languages:Go","ai_applications:robotics"],"url":"https:\/\/analyticsindiamag.com\/it-services\/unravelling-uks-ai-strategy-to-bolster-its-defences\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10095864,"title":"A New OpenAI Competitor Arrives","content":"Competing with what OpenAI has achieved in just a few months is hard. Even then, a lot of VCs are funding the competitors. Moreover, a lot of startups working in the hiding to avoid being compared with the Microsoft-funded giant are slowly crawling out in the open. Most recently, Reka, an AI startup founded by former Google, DeepMind, Baidu, Meta, and Microsoft researchers, announced that it now wants to emerge out of the stealth mode and unveiled its series A funding of $58 million. The funding round was led by DST Global Partners and Radical Ventures, along with Snowflake Ventures. Former GitHub CEO Nat Friedman also took part in the investment after his recent investment in ElevenLabs. Yi Tay, one of the founders of the Reka AI Labs said that the company is still in the early stages of this new AI revolution and wants to be part of the innovations in the field. He highlighted the two goals of the company — to build generative models and push for frontiers in AI research — and plan to use this fund to work towards that. https:\/\/twitter.com\/YiTayML\/status\/1673722977882611712 The research focused founders are motivated to work on what they call ‘universal intelligence”, which means general-purpose multimodal and multilingual agents, which are also self-improving AI models, while designing them for specifically enterprise softwares. Reka is also hiring for both – technical and non-technical roles. Where does it stand? Similar to OpenAI’s mentioned mission of “benefitting of humanity”, Reka’s mission statement says, “build generative AI models for the benefit of humanity, organisations, and enterprises”. Interestingly, the company’s head office is also based in San Francisco. This definitely gives it an advantage over the recently funded Mistral AI startup based in Paris, which might struggle with the EU’s stringent AI policies like GDPR. Mistral.AI’s mission is also similar to Reka’s — to build generative AI which would be beneficial for enterprise. Though it is not yet clear if the company advocates for open source. Read: This AI Startup from Paris Raises Highest Seed Funding Ever When it comes to the founders of the company, the expertise is clearly visible. Tay, Dani Yogatama, Qi Liu, and Cyprien de Masson d’Autume, have worked for big projects at Google, Microsoft, and Meta, including DeepMind’s AlphaCode, Bard, and Gopher. The founders realised while working on these projects that expecting to build an all encompassing LLM for all possible use cases is not practical. Yogatama told TechCrunch, “We understand the transformative power of AI and would like to bring the benefits of this technology to the world in a responsible way.” The company shows interest in building smaller models, instead of larger ones like GPT-4, that can be incorporated for different use cases. Rob Toews from Radical Ventures also told Reuters that, “I think small models are going to represent a massive paradigm shift as enterprises are getting more serious about deploying AI models at scale.” The vision of the investors and the founders are aligned. What’s the future? The company’s only product, which is still in beta testing, is Yasa, a multimodal AI assistant for images, videos, and understanding tabular data, something very similar to what OpenAI is building with GPT-4. Users can insert their proprietary data on the bot and it will derive insights from it, for which the company also provides its API. This is something that OpenAI has been trying to push with plugins and new updates to its APIs. Reka isn’t generating revenue yet, according to Yogatama. Using the funding, the company aims to acquire computing power from NVIDIA. This is clearly an aim to run for the big guns. To start, Snowflake Computing, the company which recently also partnered with NVIDIA, is allowing users to use Reka on its server. Christian Kleinerman, senior vice president of product at Snowflake, told Reuters that the company’s partnership is to guarantee its users privacy while using such AI models. This announcement also just comes after Databricks, a Snowflake competitor, acquired MosaicML in a $1.3 billion deal. Only time will tell if any startup can outshine OpenAI. India is clearly very far off. Mistral AI would face restrictive challenges in Europe. Reka still has to prove what it can do. Competition is only going to go upwards with Y Combinator announcing that one-third of its latest batch of companies are specifically focused on building AI products, that too specially in the hub of OpenAI, San Francisco. For now, OpenAI needs to be careful, as a lot of its members are also reportedly leaving the company to join Google – its biggest rival.","excerpt":"The founders realised while working on projects in Google, Meta, and Microsoft that building an all-encompassing LLM for all possible use cases is not practical","categories":["Global Tech"],"tags":["ai funding","future"],"author_name":"Mohit Pandey","publish_date":"2023-06-28T13:06:17","publication_year":"2023","word_count":764,"keywords":["future","Go","OpenAI","AI","ML","ai funding","Aim","Databricks","generative AI","multimodal AI","R","Snowflake"],"extracted_tech_keywords":["AI","ML","generative AI","multimodal AI","OpenAI","Aim","Snowflake","Databricks","R","Go"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/a-new-openai-competitor-arrives\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10119726,"title":"Stack Overflow Finally Succumbs to OpenAI","content":"Stack Overflow and OpenAI today announced a partnership aimed at combining the strengths of Stack Overflow’s knowledge platform for technical content with OpenAI’s models, such as GPT-4. This development follows Stack Overflow’s decision in 2022 to ban all answers generated by ChatGPT, citing a high degree of inaccuracy in the bot’s responses. Through the OverflowAPI access, OpenAI and Stack Overflow aim to provide developers with reliable and validated data, empowering them to find quick solutions to complex problems. This collaboration will also integrate verified technical knowledge from Stack Overflow directly into ChatGPT, enhancing users’ access to accurate information. As part of this partnership, OpenAI will leverage Stack Overflow’s OverflowAPI product to enhance model performance and gather feedback from the developer community, ensuring continuous improvement in AI development tools. This collaboration will not only benefit OpenAI but also contribute to Stack Overflow’s efforts in building better products that benefit their user community. According to Brad Lightcap, COO at OpenAI, the partnership is crucial in ensuring that models can serve a broad audience by learning from diverse sources. “Learning from as many languages, cultures, subjects, and industries as possible ensures that our models can serve everyone,” he said. Prashanth Chandrasekar, CEO of Stack Overflow, highlighted the significance of the partnership in redefining the developer experience. “Stack Overflow is the world’s largest developer community, with more than 59 million questions and answers. Through this industry-leading partnership with OpenAI, we strive to redefine the developer experience, fostering efficiency and collaboration through the power of community, best-in-class data, and AI experiences,” he said. The first set of integrations and capabilities resulting from this partnership is expected to be available in the first half of 2024. In a survey conducted by Stack Overflow in June 2023, it was found that 44% of developers currently integrate AI tools into their development workflow, while 26% have plans to do so in the near future. This situation has prompted a significant challenge for Stack Overflow. There has been a notable decline in platform traffic following the introduction of advanced generative AI models last year. These models, which often used data sourced from Stack Overflow, have contributed to this decline. As a response to this trend and to manage costs effectively, Stack Overflow is now exploring licensing agreements with AI providers.","excerpt":"The first set of integrations and capabilities resulting from this partnership is expected to be available in the first half of 2024.","categories":["AI News"],"tags":["OpenAI"],"author_name":"Siddharth Jindal","publish_date":"2024-05-06T20:37:43","publication_year":"2024","word_count":381,"keywords":["ChatGPT","API","OpenAI","AI","RAG","GPT","Aim","generative AI","R","llm_models:GPT"],"extracted_tech_keywords":["AI","generative AI","ChatGPT","OpenAI","Aim","RAG","R","API","GPT","llm_models:GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/stack-overflow-finally-succumbs-to-openai\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":52592,"title":"8 Most Notorious Cybersecurity Lawbreakers Of The Decade","content":"Every year we witness cybercrimes, discuss it with others, try to tweak the setting of our social accounts, and then forget about it. We hardly pay attention to the hackers who are basically the masterminds behind it. This is mostly because it’s challenging to track the culprit immediately, and in most cases, organisations fail to identify the hacker. Besides, companies often do not disclose the breach to avoid taking a hit on brand value. More notably, companies like Uber and Facebook prefer to hide their breaches for years. Thus, one of the prime reasons why we do not get information about hackers is that organisations often do not attempt to track the sources. Analytics India Magazine has compiled a list of most famous cybersecurity lawbreakers who are under trial or have been caught and convicted: 1. Edward Joseph Snowden Edward is an American whistleblower who got access to classified information from the National Security Agency (NSA) in 2013. He revealed the mass surveillance programs that were run by the NSA in collaboration with other intelligence alliance. The US government then filed a civil lawsuit, citing that the publication was a violation of the agreements he signed with CIA and NSA. However, he got away by getting Asylum in Russia. 2. Mark Vartanyan Mark Vartanyan is known as Kolypto, was involved in developing, improving, and maintaining the Citadel malware toolkit from 2012 to 2014. “Citadel caused a vast amount of harm to financial institutions and individuals around the world,” said US Attorney John Horn. “Mark Vartanyan utilised his technical expertise to enable Citadel into becoming one of the most pernicious malware toolkits of its time, and for that, he will serve significant time in federal prison.” The malware helped him stole account credentials from nearly 11 million computers and was responsible for $500 million in loses. On pleading guilty he was sentenced for five years in jail on 19 July 2017. 3. Kerem Albayrak A 22-year old self-proclaimed hacker tried to fake his crime and asked $100,000 to Apple in 2017. He emailed Apple that they got access to 319 million iCloud accounts and threatened them of selling in the dark web. He also sent a YouTube video as a proof for Apple that he can access any iCloud accounts. However, Apple reported to the National Crime Agency (NCA), and they arrested the guy at his home in London. Later, they found out that he was associated with a hacker group called Turkish Crime Family. Albayrak on 20 December was given a two-year suspended jail term for this childish Gambino. 4. Stanislav Saigin A Russian court convicted the owner of Pirate Site Operator for cloning websites pirate CDN platform know as Moonwalk. The site was functional on Kinogb.guru, kinokot.biz, and fosa.me, and offered around 10,000 movies and TV shows to the public for free. However, cybersecurity firm Group-IB notified him about the copyright infringement, but he ignored and used another domain on blocking the site by ISPs. Consequently, he was investigated by law enforcement agencies for a two-year suspended sentence. 5. Marcus Hutchins The hacker was behind stopping WannaCry ransomware attack, which took the globe by storm infecting more than 300,000 computers in over 150 countries worldwide. The attack started in 2017 by a hacker for stealing online banking credentials. However, the hero was in august he was arrested by FBI in Las Vegas on suspicion of authoring and selling Kronos – a malware designed to steal online banking credentials. In April this year, he accepted his cybercrime and wrote “I regret these actions and accept full responsibility for my mistakes. Having grown up, I have since been using the same skills that I misused several years ago for a useful purpose.” Having helped in stopping the WannaCry helped him in getting only one year of supervised release. 6. Olayinka Olaniyi Olayinka, along with his team, attempted theft of millions of dollars by stealing personal information to infiltrate into users’ computers. They tried to target college and university employees through pissing in the U.S. He used to send emails that looked as if it was comping from legitimate enterprises for asking personal information. After receiving the details, they logged into university systems and bank account and attempted theft which is believed to be $6 million. The Nigerian hacker committed the crime while they were staying in Malaysia. They were tracked by FBI and extradited to the U.S. to face trial, where Olayinka was sentenced to 11 months in prison and three years of supervised released on October 2018. 7. Anton P. Bogdanov Anton P. Bogdanov, a Russian citizen, identified and exploited a flaw in remote access used by accountants at Americas tax firms. He, with his team, changed clients’ personal information, which leads the tax return to their accounts. Russia was accused of stealing $1.5 million and was arrested while on vacation in Thailand in November 2017. Currently, he is undergoing trial and is expected to enter a guilty plea by February 2020. 8. Daniel Kelley A 17-year-old determined a potential flaw in the TalkTalk’s internal servers, which was posted on hacker forums. Around 14,000 hackers tried to penetrate TalkTalk’s sever, Daniel Kelley in 2016 pleaded guilty for stealing personal data from the website and blackmailed the uses; 157,000 TalkTalk’s subscribers’ information was compromised. Kelley was arrested in 2015 and was convicted in June this year for four years in detention.","excerpt":"Every year we witness cybercrimes, discuss it with others, try to tweak the setting of our social accounts, and then forget about it. We hardly pay attention to the hackers who are basically the masterminds behind it. This is mostly because it’s challenging to track the culprit immediately, and in most cases, organisations fail to […]","categories":["AI Trends"],"tags":["cybercrime","edward snowden"],"author_name":"Rohit Yadav","publish_date":"2019-12-26T10:00:00","publication_year":"2019","word_count":900,"keywords":["Go","AWS","AI","cloud_platforms:AWS","programming_languages:R","cybercrime","Git","Aim","analytics","edward snowden","GAN","R"],"extracted_tech_keywords":["AI","analytics","Aim","AWS","R","Go","Git","GAN","cloud_platforms:AWS","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/8-most-notorious-cybersecurity-lawbreakers-of-the-decade\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":43346,"title":"Decluttering The AI Hype In Startups","content":"When it comes to presenting the core capabilities of a startup, there has often been an overlapping use of the terms analytics, artificial intelligence or data science. Many companies who claim to have a strong data science team internally, which they say is working on complex data problems, is really statistics and programming. This ecosystem seems to be cluttered with many companies calling themselves “AI startups”, even when they are only using analytics. Many startups claim to have a machine learning team, but do they really have ML capabilities? It is important to get the hype sorted. The Hype That Has Become A Trend Every technology provider wants to push AI into its product strategy for the exciting opportunities it brings along. There has been a tremendous increase in startups who claim to be offering AI products without any real differentiation from the other companies. For instance, there are many chatbot companies who claim to be using AI to run these chatbots but in reality, they are just programmed functions that need constant ‘manual’ updates. If it were really AI, it would have learnt to improve and deliver results at par with humans without any intervention. The other popular process confusion is between automation tool that replaces simple administrative tasks such as insurance claim and fraud detection. Many startups today claim to be AI startups to attract funding. Investors today are in a fix as they say that every pitch they come across has terms like “AI”, “deep learning” or “analytics” in it. Most companies are even reinventing their products with an “AI touch” in it. While AI techniques are useful in solving numerous problems, they are not yet applicable in every case. It should be understood that adding an off-the-shelf algorithm to old software is not necessarily going to teach it new tricks. How Real Is The AI? While a lot of companies are claiming to be AI startups, they have little AI capabilities. They are using AI to market existing approaches. Some are even hiring people who can somehow insert AI into their product to show AI as a commodity skill, like Java programming. Another problem is that companies often claim to offer AI when actually they are using classic machine learning solutions rather than more modern techniques such as deep learning. Most AI products sold today use quantitative statistical techniques instead of actual AI algorithms and models to train the products. How To Declutter This Confusion If you do not have enough data, you are not an AI company: AI learns from large amounts of data and it is difficult for most firms to generate enough data to make algorithms efficient or simply to afford to hire data analysts. The data source is what ultimately drives the ecosystem, and it must be well-structured and optimised. Online chatbots are not necessarily AI-chatbots: Many firms are claiming to use AI in their chatbot functionalities which in actuality may not be AI-based. Most of these are updated manually which defies the rules of being AI. If your products are not learning on their own, they are not AI products: The basic aspect of AI is to have intelligence and learning capabilities of its own. If they are not learning on their own, they are not AI-driven. Using AI for fine-tuning doesn’t make you an AI company: Most companies just use AI to fine-tune their existing products rather than introducing radically new products. For instance, the AI-based toothbrush that recently became popular isn’t a new AI product in itself but used AI to improve some of its functionalities. Programming and running statistical analysis doesn’t make you an AI company: With no proper algorithms in place, it would be unfair to call a company AI company. It should be a whole package of stats, data cleaning, memory management and algorithms. Half-baked software products are not AI products: Software development is not as easy as it seems. While many companies begin to work on developing AI products they often get stuck and sell half-done products. Rule-based systems aren’t AI: Technically if you are using the rules-based system, it may be called intelligent but not AI-enabled. The point is that the system should be able to learn on its own and not just automate tasks. As Manish Singhal, founding partner of pi Venture said in an earlier interview with AIM on how he separates AI hype, “For us use cases drive investing decisions. We tend to see how technology is bringing about a 10x difference to the existing way of solving the use case.” Read: The Secret to Create AI Startups Wrapping Up As time goes by, the distinction between AI and non-AI-based companies will become more profuse. However, the scenario right now seems to be in a state of confusion. There is an obsession with calling the startup an “AI startup” not just to attract investors but to stand out in this hugely competitive market where AI is substantially making a lead. However, there are certain points that one can keep in mind before choosing AI vendor. It is important to ask questions such as what AI method is being used in its solution, how robust will be its implementation and deployment, how often will the AI system be re-trained. Also, any vendor claiming that their product includes AI should also be able to explain how it will benefit the end-user more than other product that does not use AI.","excerpt":"When it comes to presenting the core capabilities of a startup, there has often been an overlapping use of the terms analytics, artificial intelligence or data science. Many companies who claim to have a strong data science team internally, which they say is working on complex data problems, is really statistics and programming.  This ecosystem […]","categories":["AI Startups"],"tags":["ai hype"],"author_name":"Srishti Deoras","publish_date":"2019-07-25T17:30:53","publication_year":"2019","word_count":907,"keywords":["data science","machine learning","artificial intelligence","AI","chatbots","ML","ai hype","Aim","deep learning","analytics","fraud detection"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","data science","analytics","Aim","chatbots","fraud detection"],"url":"https:\/\/analyticsindiamag.com\/ai-startups\/how-to-declutter-the-ai-hype-in-startups\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10008707,"title":"23 Latest Data Science Jobs From Tech Giants Like Amazon, Google &#038; More","content":"Despite the epidemic, data scientist and analyst jobs remain to be one of the most demanded jobs among organisations. Below here, we listed down the top data science and analyst job openings from the top tech companies like Google, Amazon, Microsoft, Qualcomm and more. Google 1| Data Analyst, Product Trust and Safety Location: Hyderabad Job Role: In this role, you will work on a team that focuses on protecting the integrity of Google’s key products and Google’s users by investigating financial fraud, account-related abuse, questionable business practices, violations of policies, and general product misuse. The company works with dynamic analytical techniques and leverages custom-built technology to directly impact and improve the experience of many of Google’s users. Apply here. 2| Analyst, Security and Abuse, Chrome Extensions Location: Hyderabad Job Role: As an Analyst for Chrome Extensions Security and Abuse, you will analyse applications and extensions that are potentially harmful to users and collaborate with other members of Chrome and Trust and Safety to prevent abuse and mitigate risk to the ecosystem. The solutions you deploy, both algorithmic and operational, will enable Google to identify and flag abuse more easily, more effectively, and at scale. Apply here. Microsoft 3| Data Scientist Location: Hyderabad Job Role: As a Data Scientist at Microsoft, you will be responsible for identifying business or engineering problems and translate it to a data science problem, dig out sources of data, conduct the analysis that would reveal useful insights, and help engineering teams to operationalise the solution. Apply here. Amazon 4| Data Analyst Location: Bangalore Job Role: As a Data Analyst for the IN Seller Engagement team, you will help drive seller success on the Amazon.in marketplace. You will need to have the ability to identify and transform business needs into actionable data and insights that drive seller growth at scale. As a part of the Seller Engagement team, you will also be associated with high-impact cross-functional projects and impact seller experience across different points of the seller life-cycle. Apply here. 5| Principal, PM – Tech Location: Bangalore Job Role: As a leader of ML-powered products, you would be responsible for defining and driving the .IN Returns abuse roadmap to curtail abuse unique to India marketplace. You will negotiate the features and release timelines between multiple internal teams to maximise the benefit to Amazon and the customers. Apply here. IBM 6| Associate Partner – CPT Data Supply Chain LT00 Location: Anywhere in India Job Role: As a Data Scientist at IBM, you will help transform the clients’ data into tangible business value by analysing information, communicating outcomes and collaborating on product development. You will work to solve real-world problems for the industries, including investigating patient trends or weather patterns. Apply here. Apple 7| Director – AI\/ML Annotation Team Location: Hyderabad Job Role: In this role, you will be contributing to the advancement of a product that is redefining human-computer interaction, and work with the people who build the intelligent assistant that helps hundreds of millions of people get things done. The company is seeking for a collaborative Senior Leader who will oversee operations of the AI\/ML Annotation team in India. Apply here. Qualcomm 8| Machine Learning Software Engineer Location: Bangalore Job Role: As a Machine Learning Software Engineer in the Qualcomm CR&D team, you will be developing hardware and software solutions for the AIC100 Deep Learning Accelerator. This job is mainly for innovative near fresh graduates, who are looking for a long-term R&D career. The job activities span the whole product life cycle from early R&D to commercial deployment. Apply here. 9| ML engineer Location: Chennai Job Role: As a machine learning engineer, you will be designing, developing, monitoring and  E2E deployment of machine learning-based applications, solving complex issues in software development. You will need to have experience with machine learning, with applications in fields like natural language processing, reinforcement learning, etc., and experience with machine learning libraries such as Tensorflow, Scikit learn, PyTorch, etc. Apply here. 10| Machine Learning Performance Engineer Location: Bangalore Job Role: As a Machine Learning Performance Engineer, you will be working with the Qualcomm CR&D team to develop hardware and software solutions for the Qualcomm ADAS system. You will need to have experience in software system design, autonomy, device functional safety concepts and implementations. Apply here. 11| Machine Learning Compiler Engineer Location: Bangalore Job Role: In this job, you will be responsible for research, design, developing, enhancing, and implementing the different components of machine learning compiler based on performance and code-size needs of the customer workloads and benchmarks. You will also have to analyse software requirements, determine the feasibility of design within the given constraints, consult with architecture and HW engineers, and implement software solutions best suited for Qualcomm’s SOCs. Apply here. 12| Machine Learning Test Engineer Location: Bangalore Job Role: As a Machine Learning Test Engineer, you will join the CR&D software team focused on the definition and implementation of the overall testing strategy for a new machine learning program. This will include defining unit testing infrastructure to support TDD software and firmware development, enable automated executing and reporting, and system-level testing using popular ML frameworks such as PyTorch, CNTK, TensorFlow and others. Apply here. 13| Machine Learning SoC Design Specialist Location: Bangalore Job Role: In this role, you will join a team at Qualcomm focused on advancing the state-of-the-art in machine learning applications. The team will be involved in inventing machine learning techniques that will change the way chips are designed by reducing long-loop iterations and enabling deeper exploration of the chip design solution space. Apply here. Tata Consultancy Services 14| Data Science Expert with Python & Kubernetes Location: Hyderabad, Mumbai, Bangalore Job Role: As a Data Science expert, you will be working with the data science team. The job requires technical skills like Python, R, Kubernetes, Python for data science, knowledge of machine learning, and other such. Apply here by 30th Nov 2020. 15| Data Scientist Location: Bangalore Job Role: As a data scientist, you need to have hands-on experience in Statistical modelling & machine learning techniques like Random Forest, Boosting, Decision trees, clustering, regression analysis, hypothesis testing, multivariate statistical analysis and forecasting methodologies using SAS, R or other tools like Python. You will also need to have 5+ years of experience with MNC banks in the BFSI domain and good knowledge on retail banking, core banking, insurance etc. Apply here by 1st Feb 2021. 16| Data Scientist Location: Pune Job Role: As a data scientist, you need to have a demonstrated ability to undertake Data Science projects in the BFSI domain with client data, ability to work with clients and successfully deliver projects with minimal supervision. You must have a good understanding of one or more of the following – Python, Impala, NEO4js, Java, Pyspark, SQL, Graph databases and other such. Apply here by 31st Oct 2020. 17| Analytics Expert Location: Bangalore Job Role: As an Analytics Expert, you will need to organise a high-performance data science team for scale and growth, support solutions to be provided for RFP and RFI, partner with C-level and their direct leadership teams as well as clients to create robust core data analytics and data science services, roadmaps and solutions. You will need to oversee the success of data analytics projects in alignment with the Enterprise portfolio and integrated with the company’s short term and long-term goals. Apply here by 31st Oct 2020. Infosys 18| Lead Analyst- Data Science Location: Bangalore Job Role: As a Lead Analyst- Data Science at Infosys, you will apply business processes proficiency to understand the requirements for the business and translate this into an appropriate technical solution using suitable analytical tools, enhance and create advanced processes and may recommend new tools and technologies. Apply here. 19| Senior Analyst- Analytics Technology Location: Bangalore Job Role: As a Senior Analyst- Analytics Technology, you will be applying business processes proficiency to understand the requirements for the business and translate this into an appropriate technical solution using suitable analytical tools within the guidelines of the project scope. Apply here. Wipro 20| Developer Analyst II Location: Gurgaon Job Role: As a Developer Analyst II, you will be responsible for understanding the data available and business goals in order to judiciously select algorithms for experimentation, the scientific temperament to understand datasets in a business context and glean for opportunities. You will need the knowledge of applied machine learning such as unsupervised, unsupervised techniques, deep text-based analytics. Apply here. Larsen & Toubro Infotech Ltd 21| Data Scientist Location: Anywhere in India Job Role: As a Data Scientist, you will have to independently develop R Scripts for visualisation and modelling, develop, deploy and maintain R Shiny applications, visualise data and plot data to explain the underlying behaviour, programmatically extract data from various databases and other such. Apply here. 22| Senior Data Scientist Location: Anywhere in India Job Role: As a Sr. Data Scientist, you will deliver complex predictive models and identify new projects, opportunities and data sources which can be used to drive measurable client and commercial benefit. You will also provide insights and recommendations from different data sources to senior team members to help make decisions that drive better business outcomes for clients, among others. Apply here. 23| Data Scientist \/ ML Engineer Location: Anywhere in India Job Role: As an ML Engineer, you will build, deploy, and test machine learning and classification models. You will identify ways to gather and build training data with data labelling, develop and implement approaches for extracting patterns and correlations from both internal and external data sources using machine learning toolkits and other such. Apply here.","excerpt":"Despite the epidemic, data scientist and analyst jobs remain to be one of the most demanded jobs among organisations. Below here, we listed down the top data science and analyst job openings from the top tech companies like Google, Amazon, Microsoft, Qualcomm and more. Google 1| Data Analyst, Product Trust and Safety Location: Hyderabad Job […]","categories":["AI Hirings"],"tags":["AI Jobs","automated data science solutions","best jobs in india","computer science projects","data analytics jobs","Data Science Jobs","Data Scientist Jobs","data structure using java","jobs that ai will improve","principal data scientist","trust your data"],"author_name":"Ambika Choudhury","publish_date":"2020-10-01T18:00:36","publication_year":"2020","word_count":1597,"keywords":["jobs that ai will improve","automated data science solutions","deep learning","kubernetes","best jobs in india","data science","Data Scientist Jobs","principal data scientist","PyTorch","RAG","analytics","computer science projects","machine learning","data analytics jobs","AI","data structure using java","ML","Data Science Jobs","trust your data","AI Jobs","TensorFlow"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","data science","analytics","TensorFlow","PyTorch","RAG","kubernetes"],"url":"https:\/\/analyticsindiamag.com\/ai-hiring\/23-latest-data-science-jobs-from-tech-giants-like-amazon-google-more\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10129859,"title":"Cohere Reaches $5.5 Billion Valuation in Latest Funding Round","content":"Toronto-based AI startup Cohere Inc has reached a valuation of $5.5 billion, solidifying its position as one of the world’s most valuable AI companies and one of Canada’s largest startups. The latest funding round, led by Canadian pension investment manager PSP Investments, has raised $500 million for Cohere. The Series D funding round, which also saw participation from Cisco Systems Inc., Japan’s Fujitsu, AMD Ventures, and Canada’s export credit agency EDC, more than doubles the company’s valuation from the previous year. The fresh financing brings Cohere’s total funding to $970 million. Cohere’s competitor, OpenAI, is currently valued at $86 billion and the company made a revenue of $3.4 billion from its ChatGPT offerings.s. Founded in 2019, Cohere specialises in developing large language models (LLMs) designed for business applications. Unlike some of its high-profile Silicon Valley counterparts, such as OpenAI and Google, Cohere focuses on practical solutions to enhance business efficiency rather than pursuing artificial general intelligence (AGI). The company’s models are used by clients like Notion Labs and Oracle  to optimise tasks including website copywriting, user communication, and integration of generative AI. Cohere’s new model, Command R+, launched this spring, is its most advanced offering to date. The company is also scaling rapidly, with annualised revenue growing from $13 million at the end of 2023 to $35 million by March 2024. It plans to double its workforce this year, having started 2024 with approximately 250 employees. Cohere recently announced a partnership with Fujitsu Limited to develop and offer large language models (LLMs) tailored for enterprises. The collaboration aims to leverage Japanese language capabilities to enhance customer and employee experiences. The partnership will see Fujitsu become the exclusive provider of these services globally through Fujitsu Data Intelligence The company’s technology supports applications in over ten languages, including English, Spanish, Chinese, Arabic, and Japanese. Notable use cases include a virtual shopping tool for a luxury brand and AI-powered financial document analysis for Toronto-Dominion Bank. Borderless AI leverages Cohere’s models to address complex employment law queries in multiple languages. Cohere continues to operate from Toronto, despite its growth into international markets such as San Francisco and London. The company remains committed to its Canadian roots, with co-founder Nick Frosst highlighting Toronto’s role as a strategic hub for global expansion.","excerpt":"The latest funding round, led by Canadian pension investment manager PSP Investments, has raised $500 million for Cohere.","categories":["AI News"],"tags":["Cohere","Fund Raising"],"author_name":"Siddharth Jindal","publish_date":"2024-07-22T19:38:03","publication_year":"2024","word_count":376,"keywords":["Go","ChatGPT","API","OpenAI","AI","Fund Raising","RAG","GPT","Aim","generative AI","R","Cohere"],"extracted_tech_keywords":["AI","generative AI","ChatGPT","OpenAI","Aim","RAG","R","Go","API","GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/cohere-reaches-5-5-billion-valuation-in-latest-funding-round\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":66718,"title":"Gaia-X To Provide An Open, interoperable, Privacy-Preserving European Cloud Ecosystem","content":"Two dozen companies are participating in the European cloud initiative as founding members, including Scaleway, Atos, EDF, Dassault Systèmes and Orange from France, and SAP, Deutsche Telekom and Siemens from Germany. Security threats, data sovereignty, and technical infrastructures are increasingly important for European companies in a cloud computing landscape dominated by American and Chinese players. With this reality in mind, the French and German governments launched the Gaia-X project a year ago, with the aim of setting up a high-performance, competitive, and reliable European cloud. Gaia-X seeks to develop a digital reference ecosystem that defines common standards, certification criteria, and quality assurance. Built on existing technology provided by several major cloud providers in Europe, Gaia-X will offer competitive solutions for the global market by delivering an infrastructure that meets strict sovereignty requirements, uses open source technologies, offers strong security, and ensures interoperability. Two dozen companies are participating in the initiative as founding members, including Scaleway, Atos, EDF, Dassault Systèmes and Orange from France, and SAP, Deutsche Telekom and Siemens from Germany. Gaia-X is one of the projects aimed at building a European cloud ecosystem in order to enable digital sovereignty. Scaleway, a European public cloud player has announced its involvement in the project initiated by France and Germany that will enable companies, organizations, and citizens to retain control of their data. “We are very proud to provide our infrastructure and digital capacity to the Gaia-X project. Scaleway has more than 20 years of industry expertise, with its own energy-efficient data centres, bare metal servers and public cloud, and a range of “multi-cloud”-ready solutions,” said Yann Lechelle, CEO of Scaleway. Since January, several cloud providers have begun working together in a series of workshops. Key French players including Scaleway, OVHcloud, and Outscale have had the opportunity to code a demonstrator with the first functions of a trans-provider API alongside a variety of French and German companies. These first steps help create a European directory of available cloud services, with technical criteria (like a search for services, such as storage), compatibility (for example, S3 compatible), data location (e.g. in France) and standards respected by the provider (e.g. ISO27001, CISPE Code of Conduct, etc.). Other topics are being considered for future collaboration including a unique inter-cloud identification process. Although similar initiatives have been unsuccessful in the past, a European alternative to American and Chinese solutions is necessary for an open and secure Europe. According to Yann Lechelle, “Defeatism has no place in this debate. The tools are changing rapidly at a time when a collective and European awareness is beginning to emerge. Multi-cloud was not conceivable a few years ago. Today, cloud architectures are becoming more abstract and therefore allow calculating and storing without difficulty, regionally as much as elsewhere. We are facing a new context of convenience, energy efficiency, and respect for the environment. Customers and end-users will quickly demand an alternative.”","excerpt":"Two dozen companies are participating in the European cloud initiative as founding members, including Scaleway, Atos, EDF, Dassault Systèmes and Orange from France, and SAP, Deutsche Telekom and Siemens from Germany. Security threats, data sovereignty, and technical infrastructures are increasingly important for European companies in a cloud computing landscape dominated by American and Chinese players. […]","categories":["AI News"],"tags":["Cloud Computing in IT Industry"],"author_name":"Vishal Chawla","publish_date":"2020-06-04T17:57:06","publication_year":"2020","word_count":480,"keywords":["Go","API","programming_languages:R","cloud computing","AI","Cloud Computing in IT Industry","Git","RAG","Aim","GAN","R"],"extracted_tech_keywords":["AI","Aim","RAG","cloud computing","R","Go","Git","API","GAN","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/gaia-x-to-provide-an-open-interoperable-privacy-preserving-european-cloud-ecosystem\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":45719,"title":"IISc Joins Hands With BESCOM To Develop AI-Powered Voice Bot","content":"In today’s bleak times where Bengaluru suffers from random power outages, this new initiative by the Indian Institute of Sciences (IISc) and Bangalore Electricity Supply Company Ltd (BESCOM), will bring a glimmer of hope in the lives of numerous frustrated citizens. Bescom officials are currently working with the Machine And Language Learning (MALL) Lab at IISc to develop an artificial intelligence-powered voice bot to attend customer calls. A BESCOM official told a noted daily, “IISc has been working for many years on machine learning and natural language processing and we are using their expertise and knowledge to design a voice robot, which will be capable of answering some basic queries of the consumers in Kannada and English. The voice bot, which will be powered by AI, will have the capability to register complaints of consumers and direct them to the right departments for further queries… The induction of the voice bot into Bescom’s customer care system is expected to fasten the process of logging and redressal of complaints.” The work of developing and integrating the voice bot with Bescom’s helpline number is expected to be kicked off within six months from when tenders for the project are called. BESCOM in August had announced its intention to sign an MoU with the IISc and IIM-B to enable knowledge sharing and capacity building.","excerpt":"In today’s bleak times where Bengaluru suffers from random power outages, this new initiative by the Indian Institute of Sciences (IISc) and Bangalore Electricity Supply Company Ltd (BESCOM), will bring a glimmer of hope in the lives of numerous frustrated citizens. Bescom officials are currently working with the Machine And Language Learning (MALL) Lab at […]","categories":["AI News"],"tags":["IISc","Voice Analytics"],"author_name":"Prajakta Hebbar","publish_date":"2019-09-09T17:08:02","publication_year":"2019","word_count":221,"keywords":["machine learning","artificial intelligence","programming_languages:R","AI","Voice Analytics","IISc","Rust","R","programming_languages:Rust"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","R","Rust","programming_languages:R","programming_languages:Rust"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/iisc-joins-hands-with-bescom-to-develop-ai-powered-voice-bot\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10094464,"title":"AI Researchers are the Architects of their Own Destruction","content":"To be scared of something you’ve built is probably the worst of all fears. It comes along with guilt and resentment of what you have done. This is the case with a lot of pioneering AI researchers such as Geoffrey Hinton and Yoshua Bengio, who are talking about abandoning their lives’ work to warn about the dangers of AI. Bengio told the BBC that he feels “lost” over his life’s work. He is the second “godfather” of AI to come up with fears and ask for regulation of AI. Hinton was the first one to leave Google to warn the world about the problems of the pace of development of such AI technology. He claims that AI is “an existential risk”. A protester outside a London event at which Sam Altman spoke | Getty Images Overselling fear Last week, hundreds of AI scientists and researchers joined in to warn that the future of humanity is at risk. The Statement on AI Risk, signed by Hinton and Bengio, also included OpenAI’s Sam Altman, Google DeepMind’s Demis Hassabis, and Stability AI’s Emad Mostaque, among others. “Mitigating the risk of extinction from AI should be a global priority alongside other societal-scale risks such as pandemics and nuclear war,” reads the letter. Essentially, most of the people concerned about developing technology are concerned about what happens if the technology falls into the hands of “bad actors” which includes the military and terrorist organisations. When the experts themselves come out to point out the risks of the technology, it might be tempting and too easy to fall into the fears. On the flip side, the people who are blooming with the AI revolution are now voting to increase developments in AI. That is also probably because making AI models like ChatGPT has now become easier with open source models. Take for example, the recent debate about the need for uncensored LLMs that are outperforming bigger models. This can also be one of the reasons why the AI experts are now becoming AI doomsayers. They are increasingly getting scared of letting the technology be in the hands of people who are not so well versed with the possibility of it going rogue. But on the contrary, the third godfather of AI, Yann LeCun’s reaction to AI researchers prophesying the doomsday scenario is face palming. LeCun, the Meta AI chief, has been quite vocal when criticising people who are overestimating the capabilities of the developing generative AI models such as ChatGPT and Bard. A statistician's view of AGI doomer drama. https:\/\/t.co\/nx1pGaNVt9— Yann LeCun (@ylecun) June 4, 2023 Another AI expert to call out the AI experts who have become AI doomers is Kyunghyun Cho, who is regarded as one of the pioneers of neural machine translation which led to the development of Transformers by Google. Cho said in a media interview that glorifying “hero scientists” is like taking people’s words as the Gospel truth without rationalising the meaning behind them. He expressed frustration towards the developing AI discourse and how the US Senate hearing was disappointing as it did not talk about the benefits of AI, but only the problems related to it. “I’m disappointed by a lot of this discussion about existential risk; now they even call it literal ‘extinction’. It’s sucking the air out of the room,” he said. Losing Credibility The point is, a person who has never ever deployed a single ML model in their life, it is very easy for them to become an AI doomer. They are often influenced by the hundreds of movies made about machines taking over the world. But now, if the so-called AI experts join the same narrative, with the same explanation, it seems like it is merely fear mongering, and not an actual reason. Reminder: most AI researchers think the notion of AI ending human civilization is baloney.— Pedro Domingos (@pmddomingos) May 3, 2023 Take for example Elon Musk, the longest running AI doomer hasn’t stopped yet. In a recent interview with the Wall Street Journal, Musk said, “I don’t think AI is going to try to destroy all humanity, but it might put us under strict controls.” He still thinks there is a “non-zero chance” of AI going full Terminator on humanity. Earlier, the petition to pause giant AI experiments and not train models beyond GPT-4 was welcomed by a lot of AI researchers including Musk and Steve Wozniak. This new petition seems a lot similar to that one, only this time it includes Altman. This suggests a lot of possibilities other than actual fear of technology taking over humanity. Altman had asked the government to regulate AI development. At the same time, he also holds the opinion that OpenAI should be evaluated differently. This definitely questions the company’s motivation behind enforcing government regulations on AI as the regulation is a clear conflict of interest for Altman. We all know that fear sells. “Unfortunately, the sensational stories read more. The idea is that either AI is going to kill us all or AI is going to cure everything — both of those are incorrect,” explained Cho.","excerpt":"Oh the irony! To build something and then sell the fear of it.","categories":["AI Trends"],"tags":["AI Jobs","Geoffrey Hinton","OpenAI","Yoshua Bengio"],"author_name":"Mohit Pandey","publish_date":"2023-06-05T14:30:00","publication_year":"2023","word_count":851,"keywords":["Go","ChatGPT","Meta AI","OpenAI","AI","Geoffrey Hinton","ML","AI Jobs","Transformers","Yoshua Bengio","Aim","generative AI","R"],"extracted_tech_keywords":["AI","ML","generative AI","ChatGPT","OpenAI","Meta AI","Aim","Transformers","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/ai-researchers-are-the-architects-of-their-own-destruction\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10119797,"title":"This YC-Backed Startup is Helping Enterprises Save Up to 30% on SaaS Expenditures with Generative AI","content":"Ask any enterprise customers and they will tell you how expensive it is to run a SaaS application on the cloud. Gartner projected that in 2023 alone enterprises saw about 21% increase in spending. In extreme cases, the average spend surged by up to 500%. This is where YC-backed CloudEagle comes into the picture. The San Francisco-based AI startup offers a SaaS management and procurement platform designed to help organisations optimise their software spending and streamline the management of their SaaS applications. “A customer who spends between $1 and $2 million on SaaS applications can see 10% to 30% savings using CloudEagle,” said Prasanna Naik, co-founder of CloudEagle and a former Airbnb and Oracle exec, in an exclusive interaction with AIM. Further, he said that the company has started experimenting with AI and generative methods in multiple areas, as it has access to data from thousands of companies. “We have a database of around 250,000 SaaS applications, and it’s continuously growing, so the data keeps growing and our engine becomes stronger,” said Naik. “We know how much a company with 300 employees is paying for Salesforce compared to one with 2500 employees for the same number of licenses,” he said. Naik added that the company has built a recommendation engine that suggests which application enterprises should use based on their use cases, current tech stack, industry, employee base size, and employee growth over the past couple of months or years. Furthermore, the company utilises generative AI, enabling users to check their subscription expiration for a specific application directly within Slack. Users can type their inquiries into the Slack chatbot, which promptly generates the required information. The company is currently building its own proprietary LLM for this. It also uses AI in contract management systems. “Our AI is able to extract all the details from contracts, including start, end, and renewal dates,” said Naik, adding that the company also keeps a history of previous contracts. Streamlining Your SaaS Expenditure CloudEagle was founded in 2021 by Prasanna Naik and Nidhi Jain. Jain has previously worked in various capacities at notable companies such as Box, ServiceNow, and Goldman Sachs. Their motivation to start the company stemmed from their observations of the inefficiencies in software procurement processes at their previous workplaces. CloudEagle’s key solutions include SaaS management and SaaS procurement. The former gives customers a complete view of their SaaS applications. “With CloudEagle, IT department personnel can quickly spot what specific app employees use,” said Naik. He illustrated that if, say, 250 employees don’t use Salesforce in a company, CloudEagle promptly alerts the IT team to trim their licenses. Further, he said that CloudEagle helps connect to various data sources, such as SSO and finance systems, and browser plugins make this possible. “Today, CloudEagle has more than 380 integrations,” said Naik, adding that this eliminates the need for manual tracking and spreadsheets. On the other hand, SaaS procurement automates the process of purchasing new SaaS licenses. “CloudEagle automates manual tasks related to SaaS procurement, including approval workflows for new subscriptions and automatic reminders for renewals. This saves time for the IT and finance teams,” added Naik. CloudEagle vs the World BetterCloud, Spendesk, Zylo, Torii, Flexera, Cledara, Productiv, Sailpoint, Vendr, and Tropic are among CloudEagle’s top competitors. Naik said that these companies are version 1 SaaS management or basic SaaS procurement services companies. “CloudEagle is not just a SaaS management and procurement company; it is moving beyond that. The company has introduced a new feature called Automated License Harvesting,” said Naik. The feature automatically detects when licenses are not actively used. For instance, if a user has not logged into an application for a specified duration, the system identifies this license as a candidate for harvesting. “It runs weekly or monthly, according to the customer’s needs. We have automated this process, so companies no longer need to hire three more IT engineers or finance professionals continuously,” added Naik. CloudEagle has added another new feature, called Employee Onboarding. When an employee joins the company, CloudEagle automatically assigns them applications like Hubspot, Zoom, Jira and Slack. “Everything related to SaaS has to happen on CloudEagle,” concluded Naik, adding that the company will introduce new features like Access Review, which will be able to determine the access type a particular employee has for any application and not just license type.","excerpt":"“A customer who spends between $1 and $2 million on SaaS applications can see 10% to 30% savings using CloudEagle”","categories":["Deep Tech"],"tags":["AI Startups"],"author_name":"Siddharth Jindal","publish_date":"2024-05-07T19:12:09","publication_year":"2024","word_count":722,"keywords":["Go","programming_languages:R","AI","ML","RAG","Aim","generative AI","GAN","R","AI Startups","startup"],"extracted_tech_keywords":["AI","ML","generative AI","Aim","RAG","R","Go","GAN","startup","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/this-yc-backed-startup-is-helping-enterprise-save-up-to-90-on-saas-expenditures-with-generative-aithis-yc-backed-startup-is-helping-enterprise-save-up-to-30-on-saas-expenditures-with-generative-ai\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":54987,"title":"Who Is Arvind Krishna, the CEO of IBM","content":"As Ginni Rometty, the current CEO of IBM passes the baton to Arvind Krishna, she calls him a “brilliant technologist” who has played a crucial role in building IBM’s key technologies. According to IBM, Arvind has helped in building technologies like artificial intelligence, cognitive computing quantum computing, distributed ledger, cloud services, data analytics solutions, and nanotechnology. According to reports, Krishna was the “principal architect” of the company’s acquisition of Red Hat, one of the most significant acquisitions for IBM. The diverse set of things and business operations Arvind Krishna has led makes him a Renaissance Man for IBM. About Arvind Krishna NameArvind KrishnaBorn23 November 1962OriginAndhra Pradesh, IndiaEducationUniversity of Illinois Urbana-ChampaignOrganizationIBMDesignationCEO Arvind Krishna: The Background While you may already know that Arvind has an engineering degree from Indian Institute of Technology, Kanpur (IITK) and a PhD from the University of Illinois, there is undoubtedly a lot more about this man and the work he has done over the years which rose him to the top of the corporate chain of a goliath such as IBM. And in this article, we try to uncover those things for you. 57-year old Arvind Krishna had started his career with IBM in 1990 and has impacted the company, since then, in so many capacities that it’s quite astounding as compared to a typical CEO, ranging things from product development, research as well as core business roles. Playing his role as senior vice president and director at IBM Research, Arvind has led approximately 3,000 scientists and technologists in 12 labs, on six continents which contributed to hundreds of patents for IBM under his tenure. He holds 15 patents in different technologies, everything from cryptography to optimisation of supply chains to new analytical algorithms, which shows he led by example. He was previously a general manager of the development and production organisation of IBM Systems and Technology Group, responsible for the development and engineering of advanced semiconductor materials via microprocessors, servers and storage systems like OpenPOWER. Arvind Krishna: A Massive Proponent Of Blockchain & Open Data For Saving The World Arvind Krishna is a big proponent of using technologies like analytics, AI and blockchain for helping humanity in ways like preventing wastage, curbing pollution, and making systems more efficient. His take on the blockchain is noteworthy where he believes blockchain can help end corruption widespread in today’s governance systems. “Today, if you look at a government organisation and agency or a bank, you effectively have to trust them. You don’t have a choice, and your recourse is through a legal system or a long, arduous process later. When you get to a blockchain where multiple entities are getting together and enough of them agree on data, a process called consensus, then everything out there is open. Once it’s open, you cannot change a previous transaction, which is what enables this trust but also it enables a great deal of transparency. So, I believe things like corruption and money laundering will not exist when these things are all out in the open, and enough people out there have to agree on it. That is the power being computer science behind blockchain,” said Arvind Krishna in an event. Arvind has also suggested the creation of open data sets to help save the planet’s resources like water and energy consumption. Speaking at a symposium for US National Academy of Engineering, he said, “To have a sustainable and smart future, we are going to have to make sure that vast amounts of data are available in the open. When we look at the waterways in various nations, not just the emerging countries, there is a tremendous amount of things to do. We know that putting sensors of water, not just to measure different parameters can yield tremendous benefits. The creation of open datasets is critical for this.” An Advocate For Data Science R&D Arvind Krishna also has a great interest in cognitive computing but believes that there is a need for extensive research and development for data sciences, pointing to the fundamental challenges in the field. He once stated “Cognitive computing is going to lay forward, but we need new science. I mean should I take 3000 images or 30,000 or 3 million to train a model. Information theorists always measure entropy and its’ not a vague concept but a very precise concept. How do you measure the entropy of a data set, how good is it, how rich is it, how diverse is it? Do I need more or am I at law of diminishing returns right now? And you have got to do all of this in the context of privacy. This all requires new science which doesn’t exist today. Creating a new information theory is the need for making mission-critical decisions with data.”","excerpt":"As Ginni Rometty, the current CEO of IBM passes the baton to Arvind Krishna, she calls him a “brilliant technologist” who has played a crucial role in building IBM’s key technologies. According to IBM, Arvind has helped in building technologies like artificial intelligence, cognitive computing quantum computing, distributed ledger, cloud services, data analytics solutions, and […]","categories":["AI Features"],"tags":["Arvind Krishna","IBM","nanotech","nanotechnology"],"author_name":"Vishal Chawla","publish_date":"2020-01-31T18:55:49","publication_year":"2020","word_count":795,"keywords":["data science","Go","artificial intelligence","Rust","programming_languages:R","AI","nanotech","RAG","IBM","Arvind Krishna","analytics","GAN","nanotechnology","R"],"extracted_tech_keywords":["AI","artificial intelligence","data science","analytics","RAG","R","Go","Rust","GAN","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/arvind-krishna\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10806,"title":"Imarticus Learning Collaborates with Genpact to provide leading Analytics talent","content":"Genpact, a global leader in designing, transforming and running business process operations, has collaborated with Imarticus Learning as a knowledge partner to help fill the skill-gap that currently exists in the space of Data Analytics. The sheer amount of data that exists today and is growing multifold, is giving companies a huge opportunity to turn this data into valuable insights.  The Analytics industry has grown in double digits in the last decade and the future is even brighter. In India, particularly, the big data analytics sector is expected to witness eight-fold growth to reach $16 billion by 2025 from the current level of $2 billion, making India one of the top three markets in Analytics globally. However, the key problem is the lack of people with the right analytics skills. It is estimated that demand for data scientists will exceed supply by more than 50 percent by 2018. To address this issue and support growth, it is critical to provide skill development in Analytics to the Indian youth. Imarticus Learning, an award winning Analytics training academy, aims to do just that. Imarticus has partnered with Genpact as a knowledge partner, and launched a ‘Data Scientist Nanodegree’ that provides Analytics aspirants with a methodical understanding of leading analytical tools such as Hadoop, Python, R and SAS through industry case studies and project work. “We are very excited about this partnership which furthers our vision of constantly working with Industry to empower the aspirations of your youth. Our Nanodegrees are co-created with leading firms globally and these industry partnerships offer many benefits to corporates and students alike.”, says Nikhil Barshikar, Managing Director of Imarticus Learning As one of the largest analytics providers worldwide, with 14+ years of experience, 1000+ domain experts and over 15000+ research and analytics reports published,  Genpact’s participation as a knowledge partner will provide the expertise and guidance to ensure that this program is industry relevant and job ready, and prepares aspirants adequately to excel in the dynamic world of Analytics. The ‘Data Scientist Nanodegree’ is the perfect example of industry and academia coming together to bridge the widening skillgap. “Big data and analytics are altering the business landscape in which we operate. In attempting to meet the subsequent talent challenges, the Nanodegree program is a notable step towards building specialized skillsets and helping professionals gear up for the future”,” said Sidhartha Shishoo, Vice President and People Function leader, Analytics and Research, Genpact The ‘Data Scientist Nanodegree’ was launched at the Online@Imarticus conference in South Bengaluru on September 22, 2016. Over 120 HR, L&D and business professionals from over 50 leading organizations in the finance, banking, analytics and academia arena were present at the event including representatives from firms like HSBC, ANZ Bank, Societe Generale,  ICICI Bank, IBM and more. About the Program The program will begin its first batch in the month of October, 2016. The Data Scientist nanodegree is made up two modules : Foundation (70 hours), and Project Work (150 Hours). In the foundation level, students get a detailed understanding of Statistics and tool readiness (R, Python, SAS, HIVE, SPARK). Through project work, the students will get hands-on with industry projects and build a portfolio of demonstrable work. The course costs Rs. 55,000, delivered completely online and can be completed in 3 months.  Interested candidates may reach out to www.imarticus.org for more details about this program.","excerpt":"Genpact, a global leader in designing, transforming and running business process operations, has collaborated with Imarticus Learning as a knowledge partner to help fill the skill-gap that currently exists in the space of Data Analytics. The sheer amount of data that exists today and is growing multifold, is giving companies a huge opportunity to turn […]","categories":["AI Trends"],"tags":["Genpact"],"author_name":"Manisha Salecha","publish_date":"2016-09-29T11:52:07","publication_year":"2016","word_count":560,"keywords":["big data","Genpact","AI","programming_languages:R","Git","Python","Aim","programming_languages:Python","analytics","GAN","R"],"extracted_tech_keywords":["AI","analytics","Aim","Python","R","Git","big data","GAN","programming_languages:Python","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/imarticus-learning-collaborates-genpact-provide-leading-analytics-talent\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":13544,"title":"India sees an increased data science adoption, poised to be a lucrative market","content":"India, a rapidly growing hub of data and analytics is seeing an increased adoption of analytics and holds promise for the 600+ companies operating domestically to crank out $16 billion in market opportunities by 2025. “There are close to 600 analytics firms and more than 50% of them are now shifting their focus on Indian businesses. Of late, there has been tremendous growth funding, almost quarter of a billion dollars has now been invested in analytics deals or startups and more importantly there are hundreds and thousands of us in this ecosystem,” Deep Thomas, CEO – Tata Insights and Quants cited in a talk in Cypher 2016. The Indian Promise – Captives and analytics services providers train eye on domestic market According to a Wikibon research, the total spending on big data software, and services is expected to grow by 12 percent compound annual growth rate through 2027 to $9.2 billion, largely driven by software. While analytics has become a mandate for enterprises to establish strategic business capabilities necessary to achieve data-driven outcomes, Indian economy is providing a lot of positives, spurring analytics providers to train their eyes on the domestic market. According to Thomas, a large increase in internet penetration, a healthy sustainable GDP growth; best customer profile; close to 60% of the population are digital natives; increase in amount of disposable income amongst Indian households and Indian Government’s startup push is helping the cause. “Everything together makes us believe that it is time to look at India as a strategic market where analytics adoption will rise, there are a few challenges,” he said. However, the journey is marked by challenges that have to be addressed in order to realize the vision. Operational Challenges in realizing the Big Billion Opportunity However, the “$20-30 billion business opportunity” is also fraught with other operational challenges such as “thin margins, long credit cycles and sales cycles”. Sandeep Mittal, Managing Director at Cartesian Consulting shared in his talk the highs and lows of breaking into the Indian and though he confirmed every stereotype associated with the Indian market, he also found a silver lining in the dark cloud. From a lack of cohesion to operational challenges such as multiple offices, lengthy sales cycle (6 months – 18 months) and low gross margins, these are just a few stumbling blocks Indian companies have to overcome. “What is fairly typical of the India market, tough market to break so people will also where the good is? Indian clients want you to be near them, so you are going to have multiple offices. We have three offices in India – Bombay, Delhi and Bangalore. Why do we have a Delhi office, because Delhi clients want us to be there right down the street,” he said during the talk. Dark Side of focusing on Indian market Thin Gross Margins: When you talk about running your own business, gross margins play a huge role in profitability? What’s the gross margin an analytics firms should have? “Normally people would say 50-60%, that’s the kind of range people talk about but at least 50%. In the Indian market you can have 50%, but your company may not be as profitable and this is a fact of life if you are operating in this market,” he said. Proximity with clients means multiple offices: Close proximity to clients require companies to set up multiple offices. Case in point — Cartesian Consulting has three offices, Mumbai, Bangalore and New Delhi. A lack of single cohesive unit means more investment in training and striking cultural balance. “The culture is different – Delhi and Bangalore are such different markets to operate in. You need to speak a different language and operate differently. If you have a single delivery unit, maintaining culture, training, cohesiveness is a lot easier but three units for the same market makes it harder,” he said. Long Sales Cycles: Let’s face it, the average sales cycle in Indian market is six months and 18 months is not unheard of, Mittal revealed. “Here’s the funny thing, you will spend 18 months signing on a client and out of those 18 months, last 6 months will be spent waiting for them clients to just sign the contract,” he shared to a roomful of audience. Once the client is onboard, deliverable cycle is extremely short — 3 months’ worth work has to be packed in say 2 weeks. What’s the upside? So, what’s the upside of focusing in Indian market – proximity to clients (senior industry leaders); an abundance of analytics talent and a close impact of your work you do are just some of the factors that make it rewarding exercise. In analytics, talent abounds: Contrary to the tired rhetoric of a skill gap, Mittal and Thomas repeatedly pointed out an abundance of Indian talent that is “more than willing to work for the domestic market and are not fussed about gross margins, all they really want is good quality work”. Citing a day at his work, Mittal recalled that Cartesian Consulting received 1000 applicants on a single day. Diversity in workplace:  One of the thrills of working in Indian domestic market is the sheer work diversity. “The beauty of a new market you don’t get pigeon holed, you will work on predictive models, data visualization, machine learning and there is a lot of flexibility you can use any of the tools you want to use. In fact, we have the same team that works for an e-commerce company, mutual funds, multiplex chain and even educational university. The kind of cross-pollination of ideas you have is just incredible, take learning from technical analysis of stock market and apply it to a pizza company,” he shared. When clients become cheerleaders: a lasting relationship with clients who are industry veterans is one of the highlights of working in domestic market. Your clients will come to love you and they will become your well-wisher as well.  He played out a common scenario – dealing with top brass leaders, Chairman, CEO, MDs on a session on clustering analysis till they can quite literally visualize it. Usually, analysts who are only three years out of college are tasked with these sessions with senior most people who have never dealt with analytics in the past. Watching a close impact of work: Usually, analytics practitioners are far removed from the impact created. “For me impact is about is usually if it moved margins, did it impact topline? If you are working in domestic you will have examples of impact created literally every single week. The briefs are good or scary like how do I set my price points or we want to lower discounting or should we spend more on FB,” Mittal added. However, one major downside is you can’t beta test your products and use the domestic market as a launch pad. “If you think this a market where you can build out products, beta test out here and then take it to international market you are not going to make money because people will not pay for your test,” he pointed out. Last Word With enterprises gravitating towards a data driven culture, in 2017 and beyond, captives and domestic players are well placed to seize the opportunity to create and effectively deliver an integrated analytics system. The analytics adoption is significantly up, leadership is clamoring for data based decision making and everybody wants analytics on their side. It’s time captives and analytics solutions provider stake a claim in the domestic market that is theirs for taking. Changing attitudes, finding the right talent and enterprises expanding analytics capabilities are all encouraging signs of large market opportunity ready to be seized.","excerpt":"India, a rapidly growing hub of data and analytics is seeing an increased adoption of analytics and holds promise for the 600+ companies operating domestically to crank out $16 billion in market opportunities by 2025. “There are close to 600 analytics firms and more than 50% of them are now shifting their focus on Indian […]","categories":["IT Services"],"tags":[],"author_name":"Richa Bhatia","publish_date":"2017-03-16T10:47:57","publication_year":"2017","word_count":1280,"keywords":["big data","Go","API","machine learning","AI","Git","RAG","Aim","analytics","R"],"extracted_tech_keywords":["AI","machine learning","analytics","Aim","RAG","R","Go","Git","API","big data"],"url":"https:\/\/analyticsindiamag.com\/it-services\/india-sees-increased-data-science-adoption-poised-lucrative-market\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10078614,"title":"Bellatrix Aerospace to Invest $76 Mn for Setting Up R&amp;D &amp; Manufacturing Facility in Bengaluru","content":"Bengaluru-based space tech startup Bellatrix Aerospace recently announced that it would invest $76 million to establish a research and development (R&D) centre and manufacturing facility locally (in Bengaluru, Karnataka). This new facility is expected to manufacture thrusters, including the hall thruster, microwave plasma thruster, nano thruster, and green propulsion systems. A thruster is a spacecraft propulsion system typically used for controlling the rocket. Founded in 2015, Bellatrix was established by Yashas Karanam and Rohan Ganapathy. The startup specialises in in-space propulsion systems and rocket propulsion technologies—making space technologies affordable and accessible for all. The new manufacturing facilities will also host projects like space taxis and propulsion systems for rockets. The first phase of the project’s operations will begin in a five-acre facility near Kempegowda International Airport. Read: Meet Bellatrix Aerospace, An Indian Startup That Raised 3 Mn to Power Thruster Tech Earlier this year, Bellatrix raised an $8 million Series A investment led by BASF Venture Capital and Inflexor Ventures. Rohan Ganapathy, CEO & CTO of Bellatrix Aerospace, said, “With this investment, we will be expanding our product portfolio, adding to our existing talent pool to broaden our expertise, augment our state-of-the-art infrastructure and focus on validation of our products in space”. Additionally, the company signed an agreement this year with IN-SPACe, a single-window autonomous government agency focused on promoting and overseeing space activities of Non-Government Private Entities (NGPEs). The agreement helps Bellatrix test its propulsion systems in space with ISRO’s support; enabling them to use ISRO’s test facilities. The startup is backed by several investors, including First Cheque, growX Ventures, StartupXseed Ventures, Survam Partners, IDFC–Parampara Fund, Mankind Group Family Office, KA Enterprises, and others. Interestingly, it is also supported by Indian actor Deepika Padukone. Read: It’s High Time SpaceX Entered India The space tech market in India is about to explode. As per reports, India currently has about a 2–3% share of the global space market, which is expected to reach $424 billion by 2030. India’s share is likely to expand by 48% CAGR. Rocket and satellite launch services contribute only 2% of this market, which requires large infrastructure and investment. The remaining, about 95%, is related to satellite-based services and ground-based systems. However, owing to the country’s new reforms and private players’ participation, the space tech landscape looks more and more promising with each passing day. Besides Bellatrix, some notable startups working in the space include Kawa Space, Antariksh, Dhruva, Skyroot, Agnikool, Digantara, and others. In the 90th episode of ‘Mann Ki Baat’, PM Narendra Modi highlighted that India has about 100 active space tech startups solving the problems faced by the space tech industry.","excerpt":"Bengaluru-based space tech startup Bellatrix Aerospace recently announced that it would invest $76 million to establish a research and development (R&D) centre and manufacturing facility locally (in Bengaluru, Karnataka). This new facility is expected to manufacture thrusters, including the hall thruster, microwave plasma thruster, nano thruster, and green propulsion systems. A thruster is a spacecraft […]","categories":["AI News"],"tags":["Bellatrix Aerospace"],"author_name":"Ayush Jain","publish_date":"2022-11-03T20:08:43","publication_year":"2022","word_count":437,"keywords":["Go","API","programming_languages:R","AI","venture capital","Bellatrix Aerospace","ViT","Rust","GAN","R","startup"],"extracted_tech_keywords":["AI","R","Go","Rust","API","GAN","ViT","startup","venture capital","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/bellatrix-aerospace-to-invest-76-mn-for-setting-up-rd-manufacturing-facility-in-bengaluru\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":30235,"title":"How Psychoanalysis Can Help Neuroscience And Neural Networks","content":"We know that neuroscience forms the groundwork for artificial neural networks and in other machine learning applications. Now, this fascinating field surrounding the structure and function of the nervous system and the human mind is playing an important role in improving these applications. Researchers have found out that psychoanalysis — the brainchild of Sigmund Freud — has the potential to bring a fresh face to neuroscience. The Observable Overlap If we compare neuroscience with psychoanalysis, certain aspects do match. To break it down, neuroscience deals with the connections or “dialogues” between the brain and the nervous system, while psychoanalysis deals with psychopathology through interactions between a patient and a psychoanalyst. Both fields intersect at the functional level. Instances like thoughts which stem from the nervous system, gaining knowledge through this as a consequence, perception with emotions, etc, share a mutual area when it comes to understanding these two fields. The above view has garnered strong criticism among neuroscientists because there is no exact evidence establishing a relationship between the two. However, there is a slow uprising in the connection between psychoanalysis and neuroscience. In an article by science journalist Kat McGowan, she details how psychoanalysis could answer problems lingering in neuroscience. Psychoanalysis has insightful, provocative theories about emotions, unconscious thoughts and the nature of the mind. Neurobiology has the ability to test these ideas with powerful tools and experimental rigour. Together, the two fields might finally answer the most elusive question of them all: How is it that dreams, fantasies, memories and feelings — the subjective self — emerge from a hunk of flesh? So, the brain structure is simply a hotbed of cognitive activities. Psychoanalysis specifically delves into this and can uncover more than what lies underneath the network of billions of neural connections. Exploring The ‘Unconscious’ One of the key elements Freud’s psychoanalysis is the concept of the ‘unconscious state’. What started as a link to unearthing schizophrenia, is now the subject of many studies. In fact, most of them lean toward neuroscience rather than towards psychology, when it comes to deciphering this grey area. The relationship between neural connections and psychological disorders can explain in detail about why the disorder prevails in the first place. By hinging on this fact, there could be a relation to discovering more on neurons, as these form the basis of subjects such as deep learning. As a matter of fact, one study that looked into the aspect of brain connectivity posits why neuroscience is following the path of psychoanalysis. In recent years, there has been an increasing interest, in unconscious processes; neuroscientific studies have, in fact, tested subliminal perceptions, implicit cognition, emotion processing and interoceptive perceptions with empirical methods. Though many studies indicate that unconscious processes influence awareness, the cognitive view of the unconscious differs from the psychodynamic notion of the unconscious, which encompasses affect and motivation. What the study brought out was how psychoanalysis and neuroscience can concur in their approach and lead to an improved scientific temperament. The Key To Unraveling DL And ML With psychoanalysis brought into neuroscience, it can answer the mystery behind areas such as machine learning or even deep learning. These areas extensively derive their working based on the human brain. To stress on this point, the key difference between these AI fields and psychoanalysis is the computational factor. While ML or DL is focusing on learning something new, it gradually will follow the footsteps of a computer. This ‘logical’ component misses the ‘biological’ component. Psychoanalysis is where it could help bridge this gap. After all, the essence of mind going into AI is the norm of ‘intelligence’. As a matter of fact, challenges in these fields could be envisioned in a very different way if emotions and thoughts are brought into the picture. For example, a better model or algorithm could be designed as well as memory requirements are brought down drastically. We see enormous amounts of data going through ML\/DL projects. The Freudian field may hold answers ML\/DL in the future by evolving into something unknown or unexplored.","excerpt":"We know that neuroscience forms the groundwork for artificial neural networks and in other machine learning applications. Now, this fascinating field surrounding the structure and function of the nervous system and the human mind is playing an important role in improving these applications. Researchers have found out that psychoanalysis — the brainchild of Sigmund Freud […]","categories":["AI Features"],"tags":["brain","Deep Learning","Machine Learning","Neural Networks"],"author_name":"Abhishek Sharma","publish_date":"2018-11-14T08:40:56","publication_year":"2018","word_count":675,"keywords":["Go","machine learning","programming_languages:R","AI","brain","neural network","ML","Machine Learning","programming_languages:Go","deep learning","ViT","Deep Learning","R","Neural Networks"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","neural network","R","Go","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-psychoanalysis-can-help-neuroscience-and-neural-networks\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10169492,"title":"BFSI GCCs Are Hungry for Deep-Tech Talent","content":"Global capability centres (GCCs) in India’s banking and financial services are now largely powered by advanced technologies like AI-ML, data analytics, and automation, redefining how financial services are delivered. A key catalyst in this transformation is the adoption of real-time analytics, especially within the commercial insurance space. This is enabling sharper portfolio analysis, more effective risk engineering, and improved catastrophe modelling. Alongside this, there’s a rising focus on user experience (UX), as today’s customers expect digital solutions that are not only efficient but also intuitive and user-friendly. According to ANSR’s ‘State of BFSI GCCs in India’ report, more than 90 BFSI enterprises operate approximately 185 GCCs across India, with many organisations maintaining multiple centres. Bengaluru stands out as the leading hub, followed by Delhi\/NCR, Hyderabad, and Chennai. India’s BFSI GCC ecosystem continues to thrive, employing nearly 20–25% of the total talent working in GCCs across all industries, highlighting its growing significance in the global financial services landscape. Deep Tech with Anaptyss In an interview with AIM, Anuj Khurana, CEO and co-founder of Anaptyss Inc, highlighted the company’s strong focus on AI-led innovation. “A major focus area for Anaptyss has been its substantial investment in an integrated centre of excellence (CoE) for AI-led digital services,” he said. This centre brings together tech architects, industry leaders, and advanced AI\/ML technologies to create digital solutions that address real business challenges. Anaptyss, a digitally enabled managed services company, has already developed several impactful solutions. One of its key products, ALFA, is an AI-powered system for anti-money laundering (AML) transaction monitoring that offers “over 75–80% accuracy in detecting false alerts”. The company also created a machine learning tool that allows for 40% faster validation of third-party credit risk models. In short, “GCCs are enabling insurers to reimagine their operations,” Shrinivas Ramanujan, chief operating officer at Opteamix, told AIM. “With AI, blockchain, and predictive analytics, these centres are becoming essential for digital transformation and customer engagement.” Need for Deep-Tech Talent The push for deep-tech talent is clearly strategic. BFSI GCCs are actively hiring AI engineers, data scientists, technology architects, and natural language processing (NLP) experts. Khurana mentioned that there’s also strong demand for talent in areas such as model risk management, credit risk, fraud analytics, and cryptocurrency. The critical skills these centres are looking for include programming languages like Python and R, risk management frameworks, AI and ML proficiency, and descriptive analytics. To support this, Anaptyss is conducting focused hiring campaigns and partnering with top business schools and institutions such as the ICAI. The goal is to attract, retain, and grow talent capable of driving digital innovation. Echoing the same, Guru Thiagarajan, head of Deutsche Bank’s India tech centre, previously told AIM how deep tech is helping his teams boost productivity. “The focus is on productivity—helping research analysts and bankers sift through vast amounts of data, summarise reports, and prepare for client meetings faster,” he said. “With generative AI and large language models, we can help our teams handle data-heavy tasks in a fraction of the time.” India at the Centre of Transformation India has become central to this shift. “Today, our delivery sites in Gurugram and Noida have dedicated technology infrastructure and facilities to support year-round R&D projects focused on developing AI-based solutions,” Khurana added. The company has earmarked an annual budget to boost its innovation capability, and over 10% of its leadership roles are deeply involved in innovation programs. Additionally, Anaptyss is leveraging emerging technologies, including data analytics, robotic process automation, AI, GenAI, agentic AI, machine learning, and natural language processing to develop proprietary digital solutions. Additionally, the company’s ecosystem of ‘Digital Entrepreneurs in Residence’ allows Anaptyss to prototype and implement industry-specific and tailored digital solutions based on its deep-tech capabilities and in-depth understanding of the banking and financial services domains. Over 60% of the company’s program managers, technology architects, and data analysts are involved in developing AI-enabled digital solutions. Rise of Neo Banks Alongside this, the rise of neo banks—digital-first banking platforms—has accelerated the demand for agile, AI-powered solutions. Suman Sastry U, vice president at Everest Group, told AIM that these fintech players, like Freo, Jupiter, and Fi, are meeting the expectations of millennials and Gen Z users who want fast, flexible financial services. AI is central to their operations, powering everything from KYC compliance and credit assessment to fraud detection and customer support. This disruption is also reshaping GCCs. Traditionally geared towards international markets, many GCCs are now turning their attention inward, serving Indian fintechs and banks. In many cases, neo banks have adopted a collaborative model by partnering with traditional banks, serving as their digital banking front ends. He further mentioned that partnerships such as Niyo-SBM India and Jupiter-Federal Bank demonstrate how this approach is expanding reach, enhancing customer experiences, and driving faster digital adoption.","excerpt":"Anaptyss’ ALFA, an AI-powered system for anti-money laundering transaction monitoring, offers “over 75–80% accuracy in detecting false alerts.”","categories":["GCC"],"tags":["AI in finance","BFSI","GCC"],"author_name":"Shalini Mondal","publish_date":"2025-05-09T11:01:12","publication_year":"2025","word_count":794,"keywords":["GenAI","machine learning","GCC","agentic AI","AI","BFSI","ML","AI in finance","RAG","NLP","Aim","analytics","generative AI"],"extracted_tech_keywords":["AI","machine learning","ML","NLP","analytics","generative AI","GenAI","agentic AI","Aim","RAG"],"url":"https:\/\/analyticsindiamag.com\/gcc\/bfsi-gccs-are-hungry-for-deep-tech-talent\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10054676,"title":"Popular Datasets Released By Tech Firms In 2021","content":"Open source has dominated the tech landscape in 2021. Companies, big and small, are showing increased trust in the open-source community and contributing actively to it. It is not surprising because open source has been the backbone of rapid technological development and collaboration. In this article, we list some of the datasets open-sourced by big tech companies in 2021. Multilingual Counterfactual Dataset From Amazon Amazon released the Multilingual Counterfactual Dataset to help train machine learning models to recognise counterfactual statements. The project was started when there were no large scale datasets for counterfactual statements in product reviews made in multiple languages. This dataset annotates sentences from product reviews written in languages English, German and Japanese. A study revealed that only 1-2 per cent of sentences express counterfactuals in natural language texts; simply annotating a set of sentences selected randomly would yield a highly imbalanced dataset with a sparse training signal. The Multilingual Counterfactual Dataset helps ease such complications. This dataset is supplemented with annotation guidelines and definitions (worked on by professional linguists). Amazon has also provided the clue word list that is typical for counterfactual statements and used for initial data filtering. GoEmotions By Google GoEmotions is a dataset of fine-grained emotions — a human-annotated dataset of 58,000 Reddit comments that were extracted from popular subreddits and labelled with 27 emotion categories. As per Google, the GoEmotions taxonomy was designed keeping in mind psychology and data applicability. As per the general understanding, six emotions are considered basic; however, with GoEmotion, Google has considered 12 positive emotions, 11 negatives, four ambiguous and one marked as neutral. Since a wide range of emotions is considered, it is fairly easy to understand tasks requiring subtle differences in emotional expressions. Along with the dataset, Google also released a detailed tutorial that demonstrates neural model architecture training using GoEmotions and applying it for suggesting emojis based on conversational text. ORBIT Dataset Microsoft introduced the ORBIT dataset in partnership with City, the University of London. The dataset set a new standard for evaluating machine learning models in a few shot, high variation learning scenarios. This will help in training models for higher performance in real-world situations, specifically for people who are blind or have low vision. This dataset contains 3,822 videos of 486 objects recorded by people with low vision on their mobile phones. This benchmark reflects a practical, highly challenging recognition task. It offers a rich playground to research in robustness to few-shots and high variation conditions. Wikipedia-Based Image Text It is a large multimodal multilingual dataset from Google. WIT is composed of a set of 37.6 million image-text examples. It contains 11.5 million unique images collected from across 108 Wikipedia languages. The size helps WIT to be used as a pretraining dataset for multimodal ML models. Google claimed that this dataset was the largest in terms of image-text samples at the time of publication. The researchers started by selecting Wikipedia pages with images before extracting image-text associations and the surrounding contexts. They also performed a rigorous filtering process to ensure data quality; this process included steps like text-based filtering, caption availability, image-based filtering to ensure correct size and licensing, and length and quality. Ego4D by Facebook Ego4D is a large egocentric video dataset and benchmark suite which has 3.025 hours of daily life activity. The video spans hundreds of scenarios captured by 855 camera weavers placed at 74 different locations in nine countries. Ego4D expands the volume of egocentric videos that are publicly available to the research community. A part of these videos is with audio, eye gaze, stereo, synchronised videos, and 3D meshes of the environment. Facebook (now Meta) also introduced new benchmark challenges that are centred around the first-person visual experience by querying episodic memory, analysing hand-object manipulation, social interaction, audio-visual conversation, and forecasting events. Datasets by Hugging Face Hugging Face released Datasets, a community library for NLP. It contains 650 unique datasets and has over 250 contributors. This library had been under development for about a year and has supported many novels cross dataset research projects and tasks. Hugging Face’s Datasets are designed to address the challenges of dataset management and support community culture and norms. To develop the Datasets library, Hugging Face conducted a public hackathon which resulted in 485 commits. This library now includes continuous data types and multi-dimensional arrays for images and video data, along with audio type.","excerpt":"This article lists some of the datasets open-sourced by big tech companies in 2021","categories":["AI Trends"],"tags":["Datasets","Microsoft"],"author_name":"Shraddha Goled","publish_date":"2021-12-02T10:00:00","publication_year":"2021","word_count":730,"keywords":["Go","Datasets","Hugging Face","machine learning","AI","ML","NLP","Ray","Aim","Rust","R","Microsoft"],"extracted_tech_keywords":["AI","machine learning","ML","NLP","Aim","Ray","Hugging Face","R","Go","Rust"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/popular-datasets-released-by-tech-firms-in-2021\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10136083,"title":"Sam Altman Compares OpenAI’s o1 Model to GPT-2 for Reasoning","content":"In a recent interview, OpenAI CEO Sam Altman described the company’s latest AI model, o1, as being at the ‘GPT-2 stage’ of reasoning development. Altman explained, “I think of this as like we’re at the GPT-2 stage of these new kinds of reasoning models.” He emphasised that while the model is still early in its development, significant improvements are expected in the coming months. Altman said that users will notice o1 rapidly improving as OpenAI moves from the o1-preview model to the full release. “Even in the coming months, you’ll see it get a lot better as we move from o1-preview to o1, which we shared some metrics for in our launch blog post,” he said. “You will see it reach the GPT-4 equivalent over the coming years.” He highlighted the steep improvement curve of the o1 model, comparing it to previous jumps in AI capabilities. “It’s a very significant step forward, and one of the fun things about these new paradigms is that the improvement curve is really steep,” Altman noted. Altman expressed confidence in the model’s future performance, stating that tasks it currently struggles with will be solvable in a few months. “The things that the model just can’t solve right now, in a few months they’ll be able to solve, and a few months after that, even more,” he added. He also predicted that users will discover new ways to leverage the o1 model, similar to the gradual adoption of GPT-3.5 and the eventual success of ChatGPT. “It took people a while to figure out how to use ChatGPT, and it took us a while to build,” Altman said, hinting at similar potential for o1’s applications. Altman added that it will take users some time to fully grasp how to utilise ChatGPT with the updated model. “It’ll  take users a while to figure out how to use it, and this is quite different from the GPT models.”","excerpt":"“You will see it reach the GPT-4 equivalent over the coming years.”","categories":["AI News"],"tags":["OpenAI"],"author_name":"Siddharth Jindal","publish_date":"2024-09-20T11:10:40","publication_year":"2024","word_count":320,"keywords":["ChatGPT","API","OpenAI","AI","programming_languages:R","RAG","GPT","R","llm_models:ChatGPT","llm_models:GPT"],"extracted_tech_keywords":["AI","ChatGPT","OpenAI","RAG","R","API","GPT","llm_models:GPT","llm_models:ChatGPT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/sam-altman-compares-openais-o1-model-to-gpt-2-for-reasoning\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":3595,"title":"Analytics India Magazine Reportcard &#8211; Year 1","content":"Analytics India Magazine completed 1 year in existence. It’s time to look back at this 1 year and analyze how we performed in terms of visitors on the site. Per our rough estimate, we touched almost half of all analytics professionals in India. This misses a majority of analytics professionals and we have a huge audience base still to capture. On an average, our monthly traffic increased 28% month on month. The traffic has increased by almost 10 times since we started a year back. The Avg. Visit Duration on the site has been 3 mins 10 secs. Our visitors browsed on an average 2.2 pages on each visit. % New Visits appears very high; implying just 32% of visitors came back to site. Yet, for our returning visitors the matrices appear much better than overall. Almost 7% of returning visitors visit the site after 1 month of their previous visit. The weakest metric is that of visit duration on the site. 68% of visitors closed the page within 10 sec of their landing. 20% visitors browsed the site for more than 2 mins, just 2% for more than half hour. 66% of visitors just visited 1 page on the site. The graph below gives % visits for page depths more than 1. 47% of traffic to the site is through search, top 5 search keywords being – “analytics india magazine”, “analytics companies in india”, “business analytics course”, “data analytics courses”, “analytics magazine india”. 45% of referral traffic is from Linkedin, while 36% is from Facebook.","excerpt":"Analytics India Magazine completed 1 year in existence. It’s time to look back at this 1 year and analyze how we performed in terms of visitors on the site. Per our rough estimate, we touched almost half of all analytics professionals in India. This misses a majority of analytics professionals and we have a huge […]","categories":["IT Services"],"tags":[],"author_name":"AIM Media House","publish_date":"2013-05-24T09:42:06","publication_year":"2013","word_count":256,"keywords":["R","analytics","programming_languages:R","RAG"],"extracted_tech_keywords":["analytics","RAG","R","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/it-services\/analytics-india-magazine-reportcard-year-1\/","complexity_score":3,"technical_depth":4,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10042783,"title":"Exploring SimFin API Using Exploratory Data Analysis","content":"Analysis, in general, is the process of exploring a topic in particular by breaking and simplifying it into smaller sub-topics to gain insights and a better understanding. The process of evaluating a business and its associated resources such as projects, budgets, and other finance-related information is known as Financial analysis. Financial Analysis performed on a particular business helps us determine its performance and suitability for others. It also provides us with a better understanding of whether the entity, in particular, is profitable enough to provide a return on investment to its investors, is the entity stable and solvent or not. In other words, it is the process of determining the financial strengths and weaknesses of the entity by establishing a strategic relationship between the items of the entity, such as balance sheets, profit and loss account, and other financial statements. Financial analysis is used to evaluate economic trends, set financial policies, help build long-term plans for business activities, and identify projects or companies for investment. This is done through the crunching of financial numbers and data. Financial analysis is generally conducted in both corporate finance and investment finance settings. There are two ways of conducting a financial analysis: Internal Analysis and External Analysis. When conducted Internally, the analysis can help managers and business heads make key decisions from the generated insights or review their historical trends to analyse and review what has worked for them in the past. On the other hand, Analysis, when conducted Externally, helps potential investors explore and choose the best investment opportunities or policies. When conducted, such analytical research also helps the top management determine the success or failure of the company’s operation, appreciating and appraising an individual’s performance and contribution to the organization, realising whether the firm’s resources are being used efficiently and evaluating the internal control system. Investors who have invested their money into the firm’s shares are always keen and interested in its performance earnings and future profitability. Financial analysis helps them predict and interpret the firm’s bankruptcy and failure probability of its business enterprises. After being aware of a probable failure, investors can take preventive measures that can help them to avoid and minimize financial losses. Investing in stocks requires a careful analysis of the financial data to determine a company’s true worth. This can be time-consuming and somewhat cumbersome. The easier way to find out about a company’s performance is to look at its financial ratios, most of which are freely available on the internet. Data Analysts and Market Researchers can use one such library to explore financial data, known as Simfin. What is Simfin? Simfin is an open world crowdsourcing platform to explore and research fundamental financial data. All the present data is free, easily accessible, readily available and user-generated.  Simfin aims to make financial data freely available to private investors, researchers, and students. The data collection process is self-automated and combined along with machine learning technologies. With such available data services, SimFin aims to help those investors and researchers who cannot afford to have huge investments per year on data. SimFin tries to also maintain a check on their data by manually updating and expanding its datasets. SimFin makes it very easy to obtain and use financial and stock market data in Python. The easiest way of working with the entire datasets is to use the Python API for SimFin; it automatically downloads the share-prices through its in technology, known as signal and fundamental data from the SimFin server is fetched in no time, also saves the data to disk for future use, and loads the data into the Pandas DataFrames easily. It features data for more than 2000+ companies and standardised statements, and a stock screener with 70+ company ratios and shareable. Using Simfin for Exploratory Data Analysis We’ll be using Simfin’s Python API and exploring one of the many datasets that it contains. Our goal will be to use SimFin’s Stock Screener for NetNet stocks, better known as Net Net Working Capital stocks which focus on current assets, taking cash and cash equivalents at full value, then reducing accounts receivable for doubtful accounts, and reducing inventories to liquidation values. These are essential stocks where the current assets have been further discounted before subtracting all liabilities, preferred shares, and off-balance sheet liabilities. We will work on a defined hypothesis that a given valuation ratio or another financial signal can be used to predict future stock returns and find how we can know a relation between some financial signals and future stock returns? The following is an official implementation recreated from SimFin’s Tutorial. You can find the link to their Github page here. Getting Started with EDA To get started we’ll first install the SimFin Library; you can use the following code. !pip install simfin Next we will install all the dependencies and also import the SimFin Python API, %matplotlib inline import pandas as pd import seaborn as sns import statsmodels.api as sm # Import the main functionality from the SimFin Python API. import simfin as sf # Import names used for easy access to SimFin's data-columns. from simfin.names import * Check SimFin Version, # Version of the SimFin Python API. sf.__version__ Configuring the data director path and API path # SimFin data-directory. sf.set_data_dir('~\/simfin_data\/') Setting key to use free data. # SimFin load API key or use free data. sf.load_api_key(path='~\/simfin_api_key.txt', default_key='free') # Seaborn set plotting style. sns.set_style(\"whitegrid\") Accessing the Data Hub We will now use a StockHub object to load easily and process the US stock market financial data with these settings: # We are interested in the US stock-market. market = 'us' # Add this date-offset to the fundamental data such as Income Statements # can be set to 1, 2 or even 3 months after the Report Date. offset = pd.DateOffset(days=60) # Refresh the fundamental datasets (Income Statements etc.) every 30 days. refresh_days = 30 # Refresh the dataset with share prices every 10 days. refresh_days_shareprices = 10 Accessing StockHub, %%time hub = sf.StockHub(market=market, offset=offset, refresh_days=refresh_days, refresh_days_shareprices=refresh_days_shareprices) Using Data Signals First, we calculate financial signals for all the stocks, such as the Current Ratio, Debt Ratio, Net Profit Margin, Return on Assets, etc. These are calculated using and accessing data from the financial reports: Income Statements, Balance Sheets and Cash-Flow Statements, which are automatically downloaded and loaded by the data hub. Defining Signal to be Fed and extract data, df_fin_signals = hub.fin_signals(variant='daily') Now we will calculate growth signals for the stocks, such as Earnings Growth, FCF Growth, Sales Growth, etc. These are also calculated using data from the financial reports: Income Statements, Balance Sheets and Cash-Flow Statements, which are automatically downloaded and loaded by the data hub. Fetching growth signals, df_growth_signals = hub.growth_signals(variant='daily') Now we will calculate valuation signals for the stocks, such as P\/E, P\/Sales, etc. Again, these are calculated from the share prices and data from the financial reports. Because the data-hub has already loaded the required datasets in the function-calls above, the data is merely reused here, and the data-hub can proceed directly to computing the signals. Fetching valuation signals, df_val_signals = hub.val_signals(variant='daily') Combining all the signals into a single DataFrame: # Combine the DataFrames. dfs = [df_fin_signals, df_growth_signals, df_val_signals] df_signals = pd.concat(dfs, axis=1) Show the resultant dataframe : df_signals.dropna(how='all').head() Evaluating & Handling the Missing Data Some of the data signals have a lot of missing data, which can cause problems in the statistical analysis. Let us first see the fraction of each signal column that is missing: # Remove all rows with only NaN values. df = df_signals.dropna(how='all').reset_index(drop=True) # For each column, show the fraction of the rows that are NaN. (df.isnull().sum() \/ len(df)).sort_values(ascending=False) We will get the following output : R&D \/ Revenue                        0.616803 R&D \/ Gross Profit                   0.616652 Return on Research Capital           0.616652 Dividend Yield                       0.430889 Inventory Turnover                   0.287232 Debt Ratio                           0.259329 Net Acquisitions \/ Total Assets      0.256498 Sales Growth                         0.141619 Sales Growth YOY                     0.136864 FCF Growth                           0.133579 Assets Growth                        0.133579 Earnings Growth                      0.133579 FCF Growth YOY                       0.127126 Earnings Growth YOY                  0.127109 Assets Growth YOY                    0.127109 Interest Coverage                    0.100014 Sales Growth QOQ                     0.046013 CapEx \/ (Depr + Amor)                0.038795 FCF Growth QOQ                       0.033883 Earnings Growth QOQ                  0.033883 Assets Growth QOQ                    0.033883 P\/Sales                              0.023100 Gross Profit Margin                  0.019524 Log Revenue                          0.017117 Asset Turnover                       0.017060 Net Profit Margin                    0.017060 P\/Cash                               0.016489 Price to Book Value                  0.015124 Earnings Yield                       0.015124 FCF Yield                            0.015124 Market-Cap                           0.015124 P\/E                                  0.015124 P\/FCF                                0.015124 P\/NCAV                               0.015124 P\/NetNet                             0.015124 Quick Ratio                          0.010466 Return on Assets                     0.009083 Return on Equity                     0.009083 Share Buyback \/ FCF                  0.009083 Dividends \/ FCF                      0.009083 Current Ratio                        0.009083 (Dividends + Share Buyback) \/ FCF    0.009083 dtype: float64 As you can see from the above list, the Dividend Yield has missing data for nearly half of its rows, while the Debt Ratio has missing data for a third of its rows. Let us remove all signals that have more than 25% missing data: # List of the columns before removing any. columns_before = df_signals.columns # Threshold for the number of rows that must be NaN for each column. thresh = 0.75 * len(df_signals.dropna(how='all')) # Remove all columns which don't have sufficient data. df_signals = df_signals.dropna(axis='columns', thresh=thresh) # List of the columns after the removal. columns_after = df_signals.columns # Show the columns that were removed. columns_before.difference(columns_after) We get the output as, Index(['Debt Ratio', 'Dividend Yield', 'Inventory Turnover', 'Net Acquisitions \/ Total Assets', 'R&D \/ Gross Profit', 'R&D \/ Revenue', 'Return on Research Capital'], dtype='object') Calling the Screener for NetNet Stocks We want to test whether the NetNet investment strategy might work, so we will only use the signal-rows with a P\/NetNet ratio between zero and one, thus indicating the stock is trading at a discount to its NetNet liquidation estimate. We create this stock-screener by making a boolean mask as follows: mask_netnet = (df_signals[P_NETNET] > 0) \\ & (df_signals[P_NETNET] < 1) Rows that satisfy the screener condition have a value of True, and rows that do not meet the condition have a value of False. mask_netnet.head() Output: Ticker  Date A       2007-01-03    False 2007-01-04    False 2007-01-05    False 2007-01-08    False 2007-01-09    False Name: P\/NetNet, dtype: bool Let us plot how many different stocks have traded at NetNet-discounts in the past, mask_netnet.groupby(DATE).sum().plot(grid=True); We can also show all the tickers that have traded at NetNet discounts at some point: tickers_netnet = mask_netnet[mask_netnet].reset_index()[TICKER].unique() print('Number of unique tickers:', len(tickers_netnet)) print(tickers_netnet) Output : Number of unique tickers: 139 ['AAMC' 'ABIO' 'ABUS' 'ACER' 'ACET' 'ACRX' 'ADIL' 'ADVM' 'ADXS' 'AEHR' 'AGLE' 'AKER' 'ALPN' 'ALRN' 'ALT' 'AMSC' 'APRE' 'APVO' 'ASPN' 'ASTC' 'ATOM' 'AVEO' 'AVGR' 'AWRE' 'BLCM' 'BLFS' 'BLUE' 'BPMC' 'BRMK' 'BSPM' 'BTAI' 'CALA' 'CCXI' 'CFMS' 'CGA' 'CLBS' 'CLRB' 'COGT' 'CRDF' 'CRSP' 'CRVS' 'CSLT' 'CTIC' 'CTMX' 'CVSI' 'CYCC' 'CYIG' 'EFOI' 'ENPH' 'EPZM' 'EVH' 'EVLO' 'FBRX' 'FNJN' 'FORD' 'FORM' 'FRD' 'FTEK' 'FTK' 'GBT' 'GENC' 'GEVO' 'GLYC' 'GTHX' 'HGEN' 'HROW' 'HTGM' 'IBRX' 'INFI' 'IPDN' 'ISEE' 'JAGX' 'KKR' 'KPTI' 'KURA' 'LEDS' 'LGL' 'LUMO' 'MBOT' 'MEIP' 'MGNI' 'MICT' 'MIST' 'MN' 'MNKD' 'MRSN' 'MRTX' 'MTEM' 'MTSI' 'MYO' 'NHTC' 'NPTN' 'NSPR' 'NTIP' 'NTRA' 'NURO' 'NVTA' 'ONCT' 'ONVO' 'OOMA' 'PACB' 'PBSV' 'PDEX' 'PTE' 'PTN' 'PTON' 'RCUS' 'RGLS' 'RGNX' 'RIGL' 'RKDA' 'RSLS' 'RWLK' 'SCND' 'SINT' 'SMIT' 'SMSI' 'SOHU' 'SPRT' 'SRPT' 'SRRA' 'STOK' 'SURF' 'SYRS' 'THMO' 'TNDM' 'TRUP' 'UBX' 'VIDE' 'VIVE' 'VKTX' 'VOXX' 'VSTM' 'VTVT' 'VYGR' 'WSTL' 'XBIT' 'XPO' 'ZYNE'] Let us now compare the P\/NetNet ratio to 1-day stock returns, # Name of the new column for the returns. TOTAL_RETURN_1D = 'Total Return 1-Day' # Calculate 1-day stock-returns. df_returns_1d = hub.returns(name=TOTAL_RETURN_1D, bdays=1, future=True, annualized=False) # Show the 1-day stock-returns. df_returns_1d.loc[mask_netnet] # Limit the daily returns between -10% and +10% df_returns_1d = df_returns_1d.clip(-0.1, 0.1) # Combine the signals and stock-returns. # We are only using the rows which are NetNet discounts. dfs = [df_signals.loc[mask_netnet], df_returns_1d.loc[mask_netnet]] df_sig_rets = pd.concat(dfs, axis=1) We can then create a scatter-plot of the P\/NetNet ratio versus the daily stock-returns: df_sig_rets.plot(kind='scatter', grid=True, x=P_NETNET, y=TOTAL_RETURN_1D); Using colours to distinguish the dots for different tickers: sns.scatterplot(x=P_NETNET, y=TOTAL_RETURN_1D, hue=TICKER, data=df_sig_rets.reset_index(), legend=False); We can do the same for 1-3 year returns as well. # Name of the new column for the returns. TOTAL_RETURN_1_3Y = 'Total Return 1-3 Years' # Calculate the mean log-returns for all 1-3 year periods. df_returns_1_3y = \\ hub.mean_log_returns(name=TOTAL_RETURN_1_3Y, future=True, annualized=True, min_years=1, max_years=3) # Combine the signals and stock-returns. # We are only using the rows which are NetNet discounts. dfs = [df_signals.loc[mask_netnet], df_returns_1_3y.loc[mask_netnet]] df_sig_rets = pd.concat(dfs, axis=1) Plotting the graph, sns.scatterplot(x=P_NETNET, y=TOTAL_RETURN_1_3Y, hue=TICKER, data=df_sig_rets.reset_index(), legend=False); Removing Outliers from Data A common method for removing outliers is the so-called Winsorization of the data. It asically just limits or clips the data between, e.g. the 5% and 95% quantiles of the data. # Select all columns except for the P\/NetNet ratio. columns = df_sig_rets.columns.drop(P_NETNET) # Winsorize all the other signals and stock-returns. df_sig_rets2 = sf.winsorize(df_sig_rets, columns=columns) # Winsorize all the other signals and stock-returns. # Instead of clipping values beyond the bounds, set them to NaN. df_sig_rets = sf.winsorize(df_sig_rets, columns=columns, clip=False) Generating a Correlation Linear correlation (aka. Pearson correlation) is a simple statistical measure of how two variables change together. We will study the linear correlation between the signals and stock-returns to roughly assess which signals might be the best predictors for stock-returns. We will also study the linear correlation between the signals themselves to assess whether some of the signals seem redundant and can be removed. # Calculate the correlation between all signals and stock-returns. df_corr = df_sig_rets.corr() # Show how the signals are correlated with the stock-returns. # Sorted to show the strongest absolute correlations first. df_corr_returns = df_corr[TOTAL_RETURN_1_3Y].abs().sort_values(ascending=False) df_corr_returns Output : Total Return 1-3 Years               1.000000 Current Ratio                        0.309601 Quick Ratio                          0.293966 Log Revenue                          0.220491 Market-Cap                           0.217647 P\/FCF                                0.198681 Interest Coverage                    0.196469 Share Buyback \/ FCF                  0.194303 Earnings Yield                       0.177254 (Dividends + Share Buyback) \/ FCF    0.169556 FCF Yield                            0.166700 Return on Equity                     0.154312 Earnings Growth YOY                  0.148643 Assets Growth QOQ                    0.136451 Price to Book Value                  0.091554 FCF Growth YOY                       0.090232 Assets Growth                        0.089341 FCF Growth                           0.084630 Asset Turnover                       0.084033 Dividends \/ FCF                      0.068013 Gross Profit Margin                  0.062543 Assets Growth YOY                    0.062482 Return on Assets                     0.057145 Sales Growth                         0.054878 Earnings Growth QOQ                  0.042634 Sales Growth YOY                     0.041231 Net Profit Margin                    0.040865 FCF Growth QOQ                       0.037546 CapEx \/ (Depr + Amor)                0.020960 P\/NCAV                               0.020314 Earnings Growth                      0.015977 P\/NetNet                             0.014188 Sales Growth QOQ                     0.008207 P\/Cash                               0.006348 P\/Sales                              0.005994 P\/E                                  0.004789 Name: Total Return 1-3 Years, dtype: float64 We can also create a correlation-matrix between all the signals and the 1-3 year average stock-returns. df_corr.round(2).style.background_gradient(cmap='seismic', axis=None) Even for a small number of signals, the correlation matrix can easily get confusing. Instead, we can show only the larger correlations: # Only show the large (absolute) correlation values. df_corr2 = df_corr[df_corr.abs() > 0.7] # Round correlation values to 2 digits. df_corr2 = df_corr2.round(2) # Transform the table to give a better overview. df_corr2 = df_corr2.stack() # Remove all values that are 1.0 df_corr2 = df_corr2[df_corr2 != 1.0] # Show the result. Use a DataFrame for pretty printing. pd.DataFrame(df_corr2, columns=['Correlation']) Plotting the Final Output : Let us make some scatter plots of the most important signals versus the stock-returns. # Plot these signals on the x-axis. x_vars = [GROSS_PROFIT_MARGIN, P_NETNET] # Plot the stock-returns on the y-axis y_vars = [TOTAL_RETURN_1_3Y] # Create the plots. g = sns.PairGrid(df_sig_rets.reset_index(), height=4, x_vars=x_vars, y_vars=y_vars, hue=TICKER) g.map(sns.scatterplot); The colours in these plots represent different stock-tickers. The Gross Profit Margin was found further above to be one of the most important signals for predicting the future stock returns for 1-3 year period. The second scatter plot is for the P\/NetNet signal. This had a lower R-squared value than the Gross Profit Margin. EndNotes This article tried to implement a basic financial analysis using SimFin and explored the library and its uses. You can perform more such operations on it and understand the use of SimFin even further. You can check my implementation on the colab notebook here. Happy Learning! References SimFin Official WebsiteAccess the APIAPI Documentation","excerpt":"Analysis, in general, is the process of exploring a topic in particular by breaking and simplifying it into smaller sub-topics to gain insights and a better understanding. The process of evaluating a business and its associated resources such as projects, budgets, and other finance-related information is known as Financial analysis. Financial Analysis performed on a […]","categories":["AI Trends"],"tags":["exploratory data analysis","financial analytics","Python"],"author_name":"Victor Dey","publish_date":"2021-07-04T16:00:00","publication_year":"2021","word_count":2630,"keywords":["machine learning","TPU","AI","RAG","Python","Colab","Aim","financial analytics","Matplotlib","Seaborn","exploratory data analysis","Pandas"],"extracted_tech_keywords":["AI","machine learning","Aim","Colab","Pandas","Matplotlib","Seaborn","RAG","TPU","Python"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/exploring-simfin-api-using-exploratory-data-analysis\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10165404,"title":"Razorpay Expands its Operations to Singapore","content":"Bengaluru-based fintech company Razorpay announced on Friday that it is expanding to Singapore, its second destination in Southeast Asia. This comes after Razorpay announced various AI-driven solutions to power up business payments and solve the ‘Paisa nahi aya’ problem. The move to enter Singapore follows the company’s successful operations in Malaysia, where local businesses have experienced faster settlements and reduced transaction costs. Razorpay aims to support and scale Singapore’s digital economy by offering seamless, secure, and scalable payment solutions, powered by AI. As per the data provided by Razorpay, Southeast Asia (SEA) is witnessing an unprecedented digital payments boom and Singapore is at the heart of this. To kickstart the operations, Razorpay has launched a suite of payment solutions for Singaporean businesses. The company claims it will help firms reduce 30-40% of the cross-border transaction fees. Shashank Kumar, MD & co-founder, Razorpay said, “As one of the most advanced digital economies, Singapore is the ideal market for our next phase of growth in Southeast Asia. Our expansion aligns with Singapore’s bold vision for a cashless, innovation-driven economy, and we look forward to empowering local businesses with cutting-edge AI-powered solutions to drive digital transformation at scale.” He added, “Our AI-drive payments suite, including Agentic-AI and RAY, will redefine how businesses operate, offering not just seamless transactions but also intelligent automation that enhances operational efficiency.” “Our expansion into Singapore comes at a pivotal moment as the local market embraces innovation and digital transformation,” said Angad Dhindsa, South East Asia head, Razorpay Singapore. “We understand the unique challenges that small and medium businesses here face, from navigating cross-border transaction costs to struggling with operational inefficiencies that limit their scalability.” Razorpay will work closely with banks, financial institutions, and regulatory bodies to ensure seamless integration and compliance with Singapore’s financial landscape. Businesses in Singapore can now access Razorpay Singapore’s suite of services, including a payment gateway, cross-border transaction solutions, and real-time financial analytics.","excerpt":"Razorpay announces its second location in Southeast Asia.","categories":["AI News"],"tags":["FinTech","Razorpay"],"author_name":"Ankush Das","publish_date":"2025-03-07T09:27:44","publication_year":"2025","word_count":319,"keywords":["AI","ML","digital transformation","Razorpay","Scala","Git","Ray","Aim","analytics","edge AI","FinTech","R"],"extracted_tech_keywords":["AI","ML","analytics","Aim","Ray","edge AI","R","Scala","Git","digital transformation"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/razorpay-expands-its-operations-to-singapore\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10098001,"title":"Oracle Cloud Infrastructure to Revamp DIKSHA","content":"The Ministry of Education announced its selection of Oracle Cloud Infrastructure to revamp the Digital Infrastructure for Knowledge Sharing (DIKSHA), the country’s national education technology platform. The move aims to bring advanced technological capabilities to the education sector and enhance the learning experience for students and educators nationwide. “We have transformed and migrated DIKSHA onto Oracle Cloud Infrastructure (OCI),” said Shailender Kumar, senior vice-president and regional managing director, Oracle India and NetSuite Asia Pacific and Japan, during the briefing. The Ministry of Education’s decision to migrate DIKSHA to Oracle Cloud Infrastructure (OCI) is set to bring transformative changes to the educational landscape of India. By modernizing the national education technology platform, DIKSHA will become more accessible and witness a reduction in IT costs. “We need to embrace modern tools and technology to make education more easily available and securely accessible to everyone,” said Indu Kumar, head of department, ICT and Training, Central Institute of Educational Technology (CIET), National Council of Educational Research and Training (NCERT), ministry of education. With a widespread reach, DIKSHA currently serves 1.48 million schools in all of India’s 35 states and union territories, offering content in 36 Indian languages. As part of the multi-year collaboration agreement, OCI will work alongside the Ministry of Education to extend DIKSHA’s educational resources to millions more students, teachers, and collaborators across the country. DIKSHA, powered by the open-source platform Sunbird developed by the EkStep Foundation, stands as one of India’s most successful Digital Public Infrastructure (DPI) initiatives. It caters to school education and foundational learning programs, empowering teachers to facilitate inclusive learning for underserved and disabled learners nationwide. Over 200 million students and seven million teachers, representing both government and private schools, have access to a vast pool of content contributed by over 11,000 sources. Each day, users of the platform stream an impressive 1.2 petabytes of text and video content from renowned organizations like the National Council of Educational Research & Training (NCERT), Central Board of Secondary Education (CBSE), and State Council of Educational Research and Training (SCERTs). DIKSHA to introduce AI Chatbots for Students DIKSHA is exploring the idea of introducing AI chatbots to benefit students, according to Indu Kumar, National Coordinator, DIKSHA. Originally, the plan was to launch them on the 29th of this month, but it seems there might be a slight delay. Indu Kumar anticipates that the chatbots will be rolled out within the next month or two. The intention behind this initiative is to provide a personalized learning experience for the students, demonstrating DIKSHA’s commitment to tailoring education to individual needs.","excerpt":"DIKSHA currently serves 1.48 million schools in all of India’s 35 states and union territories","categories":["AI News"],"tags":["AI Infrastructure","Oracle"],"author_name":"Siddharth Jindal","publish_date":"2023-08-02T19:56:37","publication_year":"2023","word_count":427,"keywords":["Go","programming_languages:R","AI","chatbots","programming_languages:Go","Git","Oracle","Aim","GAN","AI Infrastructure","R"],"extracted_tech_keywords":["AI","Aim","chatbots","R","Go","Git","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/oracle-cloud-infrastructure-to-revamp-diksha\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10144158,"title":"Meet Shivaay AI, the Indian AI Model Built on Yann LeCun’s Vision of AI","content":"They say creating a foundational model in India is incredibly challenging and blame it on the constraints on computational resources and the unavailability of high-quality data. While we do have plenty of open-source data, the computationally expensive part is pre-training—essentially training the model to predict the next token. Two Indian engineering students, Rudransh Agnihotri and Manasvi Kapoor, recently launched an AI startup called FuturixAI. At first, the team released Mayakriti, an image-generation platform that created lifelike images. Later, the duo decided to build an AI model that competes with OpenAI’s GPT—and built it from scratch. “Joint embedding and parameter sharing is not a widely discussed architecture,” Agnihotri told AIM. “We explored various approaches and identified models like Llama 2, Qwen, and Gemma. These models form part of a joint embedding architecture,” Agnihotri said, adding that they drew inspiration from Meta AI chief Yann LeCun’s vision of autonomous machine intelligence. This is how the team developed Shivaay, an AI model consisting of 4 billion parameters built on this joint embedding architecture, which leverages the three models for data. This unique approach gives the model a knowledge base of all the three models. For inference, the team leverages NVIDIA A100 80GB GPUs via Google Cloud, which explains the fast response time on the server when AIM tried it. The startup is part of the NVIDIA Inception program, so it is currently offered free of cost. The model’s API is also available on Futurix’s website. Agnihotri shared the model details with AIM. Shivaay already outperforms larger state-of-the-art models on benchmarks like MMLU and MMLU-Pro by a remarkable margin of 10-15 points. The benchmark also shows that Shivaay is excellent at reasoning and mathematical calculations. Agnihotri claimed that when it comes to Indic use cases, the model was, in fact, better than Krutrim, Sarvam, and others coming up in the market. “The goal is to empower Indian developers and businesses to build their own AI agents and applications without relying on foreign models like GPT,” Agnihotri said. Scaling the Right Way The team’s approach addresses the widespread notion in India that GPT models and similar architectures aren’t accessible or effective for local needs. “We aim to change that by offering our service for free, much like how OpenAI initially provided GPT-3.5 at no cost. This allows users to explore and trust the model’s capabilities,” Agnihoti added. Regarding scaling, the current priority is user acquisition. The team just recently improved the chatbot’s user interface and, in just 15 days, saw a spike in signups, crossing 1500+ users, most of which was achieved via Reddit. “Our goal is to convince users that models like Llama 2, Qwen, and Gemma can be just as good, if not better, than existing solutions [for Indian use cases],” Agnihotri said. For training beyond these models, Futurix utilised datasets like the GATE and IIT exam questions and answers. Agnihotri claimed that this aligns with chain-of-thought reasoning, which, in turn, aligns with the current paradigm of OpenAI’s o1 and GPT-4o. It enables users to test step-by-step responses and validate logical reasoning. Agnihotri also said the company aims to allow developers to build AI agents for different vertical tasks. The company plans to raise more funds soon and release the technical paper for the Indian developer ecosystem. Agnihotri also highlighted how LeCun’s philosophy and open-source approach inspired him. Agnihotri is a third-year mechatronics engineering student from Delhi Skill and Entrepreneurship University. Kapoor is currently in his second year of pursuing electronics and communication engineering, specialising in AI and ML, at Netaji Subhas University of Technology. Agnihotri was once a JEE aspirant who couldn’t make it to an IIT, but that setback didn’t lessen his love for math. Now, with FuturixAI and Quantum Works, the young founder aims to push forward research in the AI field in India, using research in math and physics. “Google has the capability, dataset and compute, but at the same time, we have our own methods that are evolving with time,” he said.","excerpt":"The team’s approach addresses the widespread notion in India that GPT models and similar architectures aren’t accessible or effective for local needs.","categories":["IT Services"],"tags":["AI in India","Yann LeCun"],"author_name":"Mohit Pandey","publish_date":"2024-12-23T13:56:41","publication_year":"2024","word_count":664,"keywords":["Go","Yann LeCun","Meta AI","AI in India","OpenAI","AI","GPT-4o","ML","R","RAG","Aim","xAI"],"extracted_tech_keywords":["AI","ML","GPT-4o","OpenAI","Meta AI","xAI","Aim","RAG","R","Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/meet-shivaay-the-indian-ai-model-built-on-yann-lecuns-vision-of-ai\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10055877,"title":"AI To Help Predict The Use Of ICU Beds During Covid-19","content":"The researchers at the University of Waterloo and DarwinAI, an alumni-founded startup company, have developed a system that uses artificial intelligence (AI) to predict the necessity of ICU admission. The system considered over 200 clinical data points, including blood test results, vital signs, and medical history. This AI software has been trained using data from 400 cases at Hospital Sirio-Libanes in Sao Paulo, Brazil, where doctors had taken the call of admitting COVID-19 patients for intensive care. Based on those lessons, the researchers have developed a neural network that can predict the requirement for ICU admission in new COVID-19 cases. The new system has achieved over 95% accuracy and can also identify the key factors that drive the predictions, which also help clinicians get more confident with treatment. Alexander Wong, professor of systems design engineering and Canada Research Chair in AI and Medical Imaging at Waterloo, said, “That is a very important step in the clinical decision support process for triaging patients and developing treatment plans.” The technology is meant to arm doctors with a new tool to help make more informed decisions faster, and ensure that the patients most in need of an ICU receive it. “The goal is to help clinicians make faster, more consistent decisions based on past patient cases and outcomes,” said Wong, director of the Vision and Image Processing (VIP) Lab at Waterloo. “It’s all about augmenting their expertise to optimise the use of medical resources and individualise patient care.” The technology has been made freely available so scientists and engineers around the world can work to help improve it. The researchers are working on incorporating it into a larger clinical decision support system, developed in their ongoing COVID-Net open-source initiative, that also helps doctors detect COVID-19 and determine its severity using AI analysis of medical images. Wong collaborated on the ICU admission work with DarwinAI researchers Audrey Chung and Mahmoud Famouri and Andrew Hryniowski, an engineering PhD student in the VIP Lab. This paper on the research, COVID-Net Clinical ICU: Enhanced Prediction of ICU Admission for COVID-19 Patients via Explainability and Trust Quantification, was also presented at the 2021 Conference on Neural Information Processing Systems.","excerpt":"The researchers at the University of Waterloo and DarwinAI have developed an AI to predict the necessity of ICU admission","categories":["AI News"],"tags":["AI for healthcare","ai in health sector","covid-19","healthcare ai","healthtech","pandemic","University of Waterloo"],"author_name":"Meeta Ramnani","publish_date":"2021-12-14T11:48:30","publication_year":"2021","word_count":361,"keywords":["Go","pandemic","healthtech","healthcare ai","artificial intelligence","covid-19","AI","University of Waterloo","neural network","programming_languages:R","ai in health sector","ViT","Rust","AI research","R","AI for healthcare","startup"],"extracted_tech_keywords":["AI","artificial intelligence","neural network","R","Go","Rust","ViT","startup","AI research","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/ai-to-help-predict-the-use-of-icu-beds-during-covid-19\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10086531,"title":"Are Developers Embracing ChatGPT over Stack Overflow?","content":"According to reports from traffic monitoring platform SimilarWeb, Stack Overflow faced a 12% decrease in the absolute number of website visits in December. From 279 million in November 2022, Stack Overflow saw its numbers drop to 247 million by the end of December. SimilarWeb also shared statistics that showed a huge dip in its ranking under the “Programming and Developer Software website” category. The company dropped almost 30 positions, from the 202nd position to the 229th position. The company is at the lowest position it has ever been since October 2019. However, Stack Overflow team told AIM that it has not seen any material impact to traffic from ChatGPT. “Like many other technology sites during the holidays, the month of December often brings a seasonal shift as users (many of whom are professional developers) take time off to be with their friends and families. Our public platform continues to serve 100 million visitors every month, making it one of the most popular websites in the world,” shared Stack Overflow. Stack Overflow is a public platform that works on a Q&A basis where enthusiasts and professionals like developers and programmers discuss and collaborate. The accuracy of answers in this platform is determined by a voting system where the upvotes decide the accuracy of an answer. Coders essentially have been using this platform but with ChatGPT and other forms of generative AI, coding has become a simpler task with AI providing accurate answers. Santiago, from binomial, recently mentioned how ChatGPT saves hours of time in coding. It even helps in rewriting and improving codes. ChatGPT Ban Backfires In December, Stack Overflow banned ChatGPT on its website. They believe that the content generated by ChatGPT “does not meet their standards” and, therefore does not contribute towards a trustworthy environment. Since citations are not provided in GPT, accurate and verified answers will not be met. “In its current state, GPT risks breaking readers’ trust that our site provides through answers written by subject-matter experts,” explained the company on their temporary ban status. A developer on Stack Overflow suggested that users should be able to see answers from ChatGPT and other AI tools by providing an additional button called ‘Show AI-Generated Answers’. The button should be hidden by default and be available to those who wish to get an AI angle. The after-effect of such a ban was reflected in the dip in Stack Overflow visits. Developers are adopting AI-generated codes that have minimum errors. With the upgraded ChatGPT model rolled out on January 30 with improved factuality and mathematical capabilities, the user base will only increase. Having crossed 10 million daily users in just 40 days of its launch, the adoption and shift to an AI-generated model is only going to get better. To stay relevant, would Stack Overflow include AI-generated solutions or allow ChatGPT? They have to act fast. [Updated: February 3, 2023 | 21:37 | The story has been updated to reflect Stack Overflow views]","excerpt":"The rising popularity of ChatGPT causing a stir amongst users of Stack Overflow","categories":["AI News"],"tags":["ChatGPT","Generative AI","Google Analytics","Stack Overflow"],"author_name":"Vandana Nair","publish_date":"2023-02-03T18:42:05","publication_year":"2023","word_count":493,"keywords":["Go","ChatGPT","AI","Stack Overflow","Google Analytics","GPT","Aim","generative AI","Rust","Generative AI","R","llm_models:ChatGPT","llm_models:GPT"],"extracted_tech_keywords":["AI","generative AI","ChatGPT","Aim","R","Go","Rust","GPT","llm_models:GPT","llm_models:ChatGPT"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/are-developers-embracing-chatgpt-over-stack-overflow\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":41810,"title":"Facebook Down: Latest Outage Reveals How The Social Media Giant Uses AI To Tag User Images","content":"Outages over the past few years have become quite usual, and many big firms like Google and Amazon have fallen prey to this. Now, earlier this week on 3 July, Facebook, Instagram and WhatsApp witnessed a major outage. The platforms were facing serious problem loading images, videos, and other data, leaving many people frustrated. According to the social media giant, the outage occurred due to an error that was triggered during a routine maintenance operation. The company also stated that it was able to discover the problem pretty soon — however, it took some time to get working back to normal. After resolving the issues, Facebook said, “We’re aware that some people are having trouble uploading or sending images, videos and other files on our apps. We’re sorry for the trouble and are working to get things back to normal as quickly as possible.” Earlier today, some people and businesses experienced trouble uploading or sending images, videos and other files on our apps. The issue has since been resolved and we should be back at 100% for everyone. We're sorry for any inconvenience. — Meta for Business (@MetaforBusiness) July 3, 2019 We’re back! The issue has been resolved and we should be back at 100% for everyone. We're sorry for any inconvenience. pic.twitter.com\/yKKtHfCYMA — Instagram (@instagram) July 3, 2019 This is not the first time Facebook has witnessed such kind of an outage;  they have faced similar issues back in March and April when its users were complaining about glitches on all its three apps — Facebook, Instagram and WhatsApp. We’re aware that some people are currently having trouble accessing the Facebook family of apps. We’re working to resolve the issue as soon as possible. — Meta (@Meta) March 13, 2019 We’re aware of an issue impacting people's access to Instagram right now. We know this is frustrating, and our team is hard at work to resolve this ASAP. — Instagram (@instagram) March 13, 2019 Even though Facebook managed to sort problems out, the story didn’t end there.  The social media giant’s media outage left some things for its users that gave them a peek behind the digital curtain. Facebook AI Reveals What The Image May Contain The peculiar thing that people noticed during the outage is that instead of their display picture, the platform was showing text tags — such as “Image may contain: 2 people, people smiling”. And it certainly sparked both anger and amusement among people and they took it to Twitter Creepy to see how Facebook uses AI to add metadata to our photos. The current image outage is showing me this on my profile pic.twitter.com\/a1sE7NDNiI — Mick Côté (@MickCote) July 3, 2019 https:\/\/twitter.com\/zackwhittaker\/status\/1146456836998144000 Put simply, the social media giant uses artificial intelligence and machine learning to the scan and read images. And this is something Facebook has been doing since at least April 2016. Furthermore, this project of Facebook reading a picture and using tags and description is a part of the platform’s accessibility efforts. If you don’t know about this initiative by Facebook, the platform has created a feature called “automatic alternative text” and it automatically describes the content of photos to blind and visually impaired users. Talking about how this automatic alt text feature works, Facebook’s machine learning technology helps to build artificial intelligence systems by using some specific algorithms in such a way that every time a picture is fed to the system, it is able to identify what the object is there in the photograph. And then using the mobile device’s voice feature, the platform reads it out to the user. Bottom Line The ability to scan images using machine learning has been a major focus for Facebook — it has even updated the feature of alt text in order to make more precise and accurate scans of people’s photos. However, If you look at the bigger picture, even though the technology is making things better for people, irrespective of their abilities, people after the outages are becoming more concerned about their data on the internet. It is also showing how companies are gathering data from every piece of content — whether its text or photo or video. Every day billions of photos are being uploaded and shared on Facebook, Instagram and WhatsApp. So, you can imagine the amount of data this social media company is dealing with. And when things like outages happen, it is definitely a worrisome incident — for both, the company as well as its user base.","excerpt":"Outages over the past few years have become quite usual, and many big firms like Google and Amazon have fallen prey to this. Now, earlier this week on 3 July, Facebook, Instagram and WhatsApp witnessed a major outage. The platforms were facing serious problem loading images, videos, and other data, leaving many people frustrated. According […]","categories":["AI Features"],"tags":["Facebook AI","Facebook outage"],"author_name":"Harshajit Sarmah","publish_date":"2019-07-04T17:03:30","publication_year":"2019","word_count":750,"keywords":["Go","machine learning","artificial intelligence","Facebook AI","AI","programming_languages:R","programming_languages:Go","Git","Facebook outage","Rust","R","programming_languages:Rust"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","R","Go","Rust","Git","programming_languages:R","programming_languages:Go","programming_languages:Rust"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/facebook-down-latest-outage-reveals-how-the-social-media-giant-uses-ai-to-tag-user-images\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":11193,"title":"Meet Balaji SR, winner of Capgemini Tech Challenge 2016 in Data Science","content":"The team InbuiltMinds comprising Saran Kumar Pantangi, Aditya Sethia, Diptangshu Chakrabarty, Vamsikrishna Patchava and Balaji SR took away the winner’s award at the recently concluded Capgemini Tech Challenge 3.0. The third edition of the technology contest by Capgemini was held at Capgemini Knowledge Park in Airoli, Mumbai. The team won the Tech Challenge Title for the penultimate challenge called the ‘Digital Shopper’. Analytics India Magazine interacted with one of the participants, Balaji SR who happens to be an alumnus of Great Lake’s first PGP-BABI batch in Chennai. Based out of Hyderabad, he first came to know about the Tech Challenge through Google Ads. Commenting on his participation he said that “Capgemini Tech Challenge was open to both IT professionals and Students all over the country ageing between 18 to 35. It received a good response from over 74000 participants. Each Challenge had three levels with level three as final challenge. Contest was held on various themes like Coding, Pega, Hybris, Adobe, Block Chain, Data Science and Data Integration.” [pullquote]“The first two levels were held online and the third level was a two-day live Hackathon at Capgemini Knowledge Park, Mumbai. There was also a team based digital challenge. I had competed in Data science theme. I am the winner in Individual Data Science Category and part of the team which had won digital challenge”, he exhilarated.[\/pullquote] Receiving a humongous response, there were over five lakh page visits and 4.8 million views on the Tech Challenge microsite. However, only sixty participants from across the country attended the finals in Mumbai. The penultimate tech challenge titled the `Digital Shopper’ comprised of a team event where eight different teams brainstormed and used their technology expertise to create solutions to real-life consumer challenges. It required participants to create a solution to a complex problem, which would allow a shopper to research multiple categories of the same product, compare price indexes, and identify a cost-effective product, all of it through a unified solution. Balaji used the tool R to go about the challenge. A data professional with an experience in data warehousing and business intelligence, Balaji saw this opportunity as a platform to benchmark himself against best the Data Science professionals in the Country. He confessed “Great Lakes PGP-BABI was my gateway to Data Science. I have an exposure right from Data extraction to predictive Analytics, thanks to the Great Lakes’ program”. He further added “I need not prepare much for the competition as I had already gone through the grind when I did my PGP-BABI program at Great Lakes.” On a concluding remark Balaji said that his major takeaway from the competition is satisfaction and confidence of being one of the best Data Scientists of the country someday. One should have strong fundamentals- is his advice to other aspirants. “Data Science is not about running some tools on the data. It is about filling the gap from Data to Decisions. In addition to technical capability one should be good in business understanding. I am good in technology and was lucky to pick up good business understanding from Great Lakes PGP-BABI program”, he said. The jury comprised key members from Capgemini India leadership. Winners and runners up in every theme received cash prizes of up to Rs 30,000 and Rs 20,000 respectively.","excerpt":"The team InbuiltMinds comprising Saran Kumar Pantangi, Aditya Sethia, Diptangshu Chakrabarty, Vamsikrishna Patchava and Balaji SR took away the winner’s award at the recently concluded Capgemini Tech Challenge 3.0. The third edition of the technology contest by Capgemini was held at Capgemini Knowledge Park in Airoli, Mumbai. The team won the Tech Challenge Title for […]","categories":["AI Trends"],"tags":["great lakes analytics"],"author_name":"Srishti Deoras","publish_date":"2016-11-16T06:45:16","publication_year":"2016","word_count":546,"keywords":["business intelligence","data science","Go","programming_languages:R","AI","R","Git","great lakes analytics","llm_models:Gemini","analytics","predictive analytics"],"extracted_tech_keywords":["AI","data science","analytics","predictive analytics","R","Go","Git","business intelligence","llm_models:Gemini","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/meet-balaji-sr-winner-capgemini-tech-challenge-2016-data-science\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10163330,"title":"How Verint’s GIC in Bengaluru Will Contribute to its Global R&amp;D","content":"Verint, a global leader in customer experience (CX) automation, has announced the establishment of its new Global Innovation Centre (GIC) in Bengaluru. The firm plans to expand the GIC to approximately 1,000 employees by the end of 2026 and continue the expansion of its workforce by 15-20% every year. The firm is looking to hire engineers specialising in AI, cloud, data design, and data science. In a conversation with AIM, Rob Scudiere, CTO of Verint, and Ajayy M Dawar, vice president of the GIC, elaborated on the rationale behind the establishment and how the innovation hub will contribute to the global R&D of the company. Verint has been present in Bengaluru for over 20 years, making it a natural choice for the company’s first GIC. Scudiere, who has over 25 years of experience in the SaaS industry, emphasised Bengaluru’s role in driving Verint’s vision of being a pioneer in CX automation. “Our technology solutions are designed to help organisations achieve AI-driven business outcomes, reducing costs, increasing efficiency, and enhancing customer experience. The incredible talent in Bengaluru is a key enabler of this vision,” Scudiere said. Reinforcing this perspective, Dawar pointed out that Verint’s office in Bengaluru has seen significant year-over-year growth. “We have been investing in this facility, expanding by 40-50% annually. Beyond R&D, we also have customer care and other key functions based here, which gives us the leverage to focus on innovation,” he said. Ajay Dewar, Peter Fante, Rob Scudiere, Vikas Sood, Anil Chawla and Verint team cutting the ribbon to mark the launch of their first GIC AI Agents, Automation, and Bengaluru With AI automation playing an increasing role in GCC operations, Verint aims to prioritise employee upskilling and reskilling at the GIC. “We have a great talent pool, and we continue to bring in experienced hires and fresh graduates from top IT universities,” Scudiere noted. Notably, the company has been actively hiring experienced AI and cloud professionals while collaborating with top universities to bring in fresh talent. “We have a structured program for career development and training, covering areas like cloud computing, AI, and data science. Additionally, we partner with hyperscalers to leverage their training programs at reasonable costs. Our hackathons further foster an environment of continuous innovation,” Scudiere explained. The executives revealed that while the firm is hiring from external avenues, it has internal plans to retain existing talent and continue developing them. “This is a talent hub for us, and we’ve been investing in this facility year over year. It made sense to continue expanding in one location where we have established capabilities rather than branching out elsewhere,” Dawar added. With the GIC’s focus shifting toward R&D and innovation, Scudiere highlighted some of the major contributions coming from its GICs globally. He believes Bengaluru will contribute in the same way. Rather than working on isolated projects, the centre will be fully integrated into Verint’s global strategy. One such innovation is the Data Insights Bot, an AI-powered solution that helps customers gain actionable insights and automate processes. Moreover, Verint’s GIC will play a crucial role in developing TimeFlex, a tool that democratises how agents manage their schedules without supervisor intervention. “We are seeing significant business outcomes from TimeFlex, including increased employee satisfaction and reduced agent attrition,” Scudiere said. The AI Agent Strategy for Verint As part of Verint’s AI-driven CX strategy, the company has introduced a suite of AI agents under its Copilot brand. These AI-powered bots are designed to optimise customer interactions and streamline workflows. Scudiere outlined the five core AI agents currently in deployment: Smart Transfer Bot – Ensures seamless transitions between self-service and assisted channels. Coaching Bot – Provides real-time guidance to agents during customer interactions. Knowledge Bot – Conducts behind-the-scenes research to deliver relevant information to agents. Call Summary Bot – Automatically summarises customer conversations for improved efficiency. CX\/EX Scoring Bot – Monitors real-time sentiment between agents and customers, offering supervisors valuable insights. These AI-driven solutions have delivered significant cost savings for clients, enabling them to reinvest in customer experience initiatives while improving operational efficiency. However, Dawar clearly stated that AI agents are not intended to replace human employees. “We see AI as an opportunity to enhance innovation, not as a replacement for people. This is just the beginning, and we are excited about the doors it will open for us and our customers,” he said.","excerpt":"With AI automation playing an increasing role in GCC operations, Verint aims to prioritise employee upskilling and reskilling at the GIC.","categories":["GCC"],"tags":["Bengaluru","GCC"],"author_name":"Mohit Pandey","publish_date":"2025-02-12T15:02:50","publication_year":"2025","word_count":725,"keywords":["data science","GCC","AI","cloud computing","innovation","ML","RAG","automation","Aim","GAN","Bengaluru","R"],"extracted_tech_keywords":["AI","ML","data science","Aim","RAG","cloud computing","R","GAN","automation","innovation"],"url":"https:\/\/analyticsindiamag.com\/gcc\/how-verints-gic-in-bengaluru-will-contribute-to-its-global-rd\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":38519,"title":"5 Latest Data Science Job Openings To Apply","content":"Photo by Daniel Canibano Analytics India Jobs offers a new platform for the data science community in India by bringing to the fore latest job openings across the country. Here are few Data Science job openings across the top 4 cities – Bangalore, Hyderabad, Mumbai and Pune to apply right away: Data Scientist @ Saama Technologies, Pune Saama is the advanced clinical data analytics company unleashing insights from data for over 20 years. Requirements: 2 To 5 years of experience in Machine learning, Predictive Analytics, Data Science Statistical knowledge about Hypothesis Testing, probability distributions, sampling theory, T-test, Z-test, ANOVA, etc. Proficient in R, Python Hands-on working knowledge of algorithms like linear regression, logistic regression, rpart, ctree, C5.0, random forests, xgboost, adaboost etc. in R \/ Python. Implementing models based on Text Mining, NLP, Time Series Working knowledge of supervised and unsupervised machine learning models Understanding of association rule mining, market basket analysis, clustering, principal component analysis Hands-on working experience libraries like pandas, numpy, scikit-learn in python Exposure to Deep Learning libraries like keras, tensorflow. Exposure to algorithms like CNN, RNN, LSTM. Apply here Data Scientist @ Brontominds, Bangalore Brontominds aims to be a world class leader in AI\/Analytics\/Automation as a business partner and as an employer. As a data scientist, the candidate will design and develop systems to analyze unstructured “big data” sources to generate actionable insights and solutions for client’s product enhancement. Requirements: 1-3 years of experience in Data Science Big Data – Hadoop, Spark, R Experience in developing REST APIs and Batch Scripts Statistics modeling and algorithms Machine Learning Experience – including deep learning and neural networks, exposure of various platforms Expertise in features engineering, building, and optimizing classifiers using ML techniques Data Visualization and analysis tools including R and Python. Apply here Data Scientist @ Shortlist, Mumbai & Hyderabad Shortlist works in India and East Africa to help growing enterprises hire based on skills and potential, rather than pedigree. Products such as customised chatbots would require decent understanding of machine learning algorithms and programming. Responsibilities: Work closely with the CTO and leadership to solve statistical and business problems Develop & code production-grade novel algorithms for our business-experimentation platform. Manage, Develop, maintain and market the business experimentation suite at Fractal Conduct research and prototyping innovations, data and requirements gathering, solution scoping and architecture Collaborate effectively with internal stakeholders and cross-functional teams to solve problems, create operational efficiencies, etc. Consult clients and client facing teams on advanced statistical and machine learning problems, especially in the areas of marketing and designing experiments Test various machine learning and analytical tools, especially in the big data space, to scale prototypes to production-grade systems Provide solutions but not limited to: Customer Segmentation & Targeting, Propensity Modeling, Churn Modeling, Lifetime Value Estimation, Forecasting, Recommender Systems, Modeling Response to Incentives, Marketing Mix Optimization, Price Optimization. Apply here Data Scientist @ vPhrase Analytics Solutions Pvt. Ltd, Mumbai vPhrase Analytics is a technology company based out of Mumbai, India whose key offerings include PHRAZOR platform. PHRAZOR gathers data, structures the facts and in the end applies language to present the reader with humanized, targeted narratives. As a data scientist at vPhrase, you would be responsible for developing, testing and maintaining our data models in productions. These models would power analytical and decision modules of our product’s AI engine. Requirements: Ph.D.\/Masters\/Bachelors in Statistics, Computer Science, Economics or a related field with at least 1 years of experience. Deep understanding of machine learning models, data analysis and deep learning methods. Background in Natural Language Processing (NLP) and text analytics is preferred. Ability to handle large and complex unstructured datasets. Ability to perform independent research across varied domains of analytics. Apply here Data Scientist @ Prescience Decision Solutions, Bangalore Prescience is a fast growing, focused Advanced Analytics company that helps enterprises become more PRESCIENT (predictive) by gaining meaningful business insights and develop optimized solutions through careful analysis of data. This is a unique opportunity to join a new, multidisciplinary team of creative and passionate individuals destined to change the face of analytics and work on high-impact projects utilizing Artificial Intelligence, Machine Learning, Natural Language Processing, BigData NoSQL and SQL based analytics and the latest in Search. Requirements: Machine Learning (ML) algorithms such as Regression, Decision Tree, Logistic Regression, K-Mean and Markov Decision Processes Application of Neural Networks, in for example Deep Learning for text, image classification Natural Language Processing (NLP) Search and the use of Search for pre-processing and integration of NLP and Machine Learning Expertise in Python or R Experience with NoSQL and SQL Apply here","excerpt":"Analytics India Jobs offers a new platform for the data science community in India by bringing to the fore latest job openings across the country. Here are few Data Science job openings across the top 4 cities – Bangalore, Hyderabad, Mumbai and Pune to apply right away: Data Scientist @ Saama Technologies, Pune Saama is […]","categories":["AI Hirings"],"tags":["Data Science Jobs"],"author_name":"Ram Sagar","publish_date":"2019-05-01T06:23:12","publication_year":"2019","word_count":757,"keywords":["data science","machine learning","artificial intelligence","AI","neural network","ML","Data Science Jobs","NLP","Aim","deep learning","analytics"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","neural network","NLP","data science","analytics","Aim"],"url":"https:\/\/analyticsindiamag.com\/ai-hiring\/5-latest-data-science-job-openings-to-apply\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":62513,"title":"Carnegie Mellon Unveils Five Interactive Maps To Display Real-Time Information On COVID-19 Maps","content":"Carnegie Mellon University has released five interactive maps which will display real-time information on COVID-19 symptoms, doctor visits, medical tests and browser searches related to the virus. These interactive maps will highlight country-level data which will be used to forecast disease activity in the US. These interactive maps include data developed with the help of partners including Google, Facebook, Quidel Corp. and a national health system. The real-time daily updated data will provide the general public and decision-makers with a new and unique means of monitoring the ebb and flow of the disease across the country. According to their website — “COVIDcast displays signals related to COVID-19 activity levels across the United States, derived from a variety of anonymised, aggregated data sources made available by multiple partners. Each signal may reflect the prevalence of COVID-19 infection, mild symptoms, or more severe disease over time, and can be presented at multiple geographic resolutions — state, county, and\/or metropolitan area. All these signals taken together may suggest heightened or rising COVID-19 activity in specific locations. They will provide useful inputs to CMU’s pandemic forecasting system.” According to CMU President Farnam Jahanian, COVIDcast leverages Carnegie Mellon’s leadership in machine learning, statistics and data science. And it is built upon the partnership with the Centres for Disease Control and Prevention (CDC) in epidemic forecasting at a time when policymakers and health care providers are eager for more insights into the spread of COVID-19. He said, “Our multidisciplinary team of researchers has worked tirelessly to bring together a variety of data sources to support informed decision-making throughout our global society.” Users can use the tabs on COVIDcast: To select which data source is visualised on the US mapTo display the data at the level of states, metropolitan areas or countiesTo show either intensity of activity or whether the activity is rising or falling COVIDcast was created by Carnegie Mellon’s Delphi Research Group and its COVID-19 Response Team to provide a detailed and up-to-date picture of current COVID-19 activity. The aim was to use this enhanced information in forecasting disease activity. These forecasts will provide up to four weeks of warning to hospitals in a given locale that they likely will see increases in the number of people requiring hospital care. Ryan Tibshirani, co-leader of the Delphi group and an associate professor of statistics and machine learning said, “The forecasts, as well as “nowcasts” that attempt to provide a combined, integrated view of current conditions, promise to provide valuable guidance as government and health care officials plan next steps in addressing the pandemic.” Jodi Forlizzi, professor and director of CMU’s Human-Computer Interaction Institute, led the team that created the visualisations of the data sources. The data sources include responses to CMU surveys by users of Google Surveys, and by users of Facebook. Another data source for the maps is Google Health Trends, which has provided data for the Delphi group’s influenza forecasts for the past five years. For the latest forecasting project, Delphi uses the Google Health Trends interface to estimate how often people are in a given location and on a given day search Google for topics related to COVID-19. A national health system also is providing statistics on patient visits to doctors and telemedicine visits. This enables the CMU Delphi researchers to estimate the percentage of visits for COVID-19 related symptoms in any given location for a given day. Another partnership is with Quidel Corp., a medical test maker, who provides the group with statistics on influenza tests. Flu tests are routinely ordered for people suffering COVID-19 symptoms as a means of excluding flu as a diagnosis; thus, requests for flu tests are indicators of possible COVID-19 activity. “All of these signals are just rough indicators of COVID-19,” emphasised Roni Rosenfeld, co-leader of Delphi and head of CMU’s Machine Learning Department. Rosenfeld said, “Anyone data source may not be conclusive, but if multiple sources indicate the same thing, people can have greater confidence about what is happening or will soon happen in various locales.” The Delphi Group is continuously providing all of its estimates in a computer-accessible way to anyone. The Delphi Research Group has been performing epidemic forecasts for the past eight years, most notably for each influenza season. Last year, the CDC named Delphi one of two National Centers of Excellence for Influenza Forecasting. At the CDC’s request, the group this spring extended and adapted its flu forecasting efforts to encompass COVID-19.","excerpt":"Carnegie Mellon University has released five interactive maps which will display real-time information on COVID-19 symptoms, doctor visits, medical tests and browser searches related to the virus. These interactive maps will highlight country-level data which will be used to forecast disease activity in the US.  These interactive maps include data developed with the help of […]","categories":["AI News"],"tags":["Carnegie Mellon University","Coronavirus","covid-19","covid19 data"],"author_name":"Sejuti Das","publish_date":"2020-04-24T12:30:00","publication_year":"2020","word_count":740,"keywords":["data science","Carnegie Mellon University","Go","machine learning","covid-19","AI","programming_languages:R","programming_languages:Go","covid19 data","Coronavirus","RAG","Aim","ViT","R"],"extracted_tech_keywords":["AI","machine learning","data science","Aim","RAG","R","Go","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/carnegie-mellon-unveils-five-interactive-maps-to-display-real-time-information-on-covid-19-maps\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10071145,"title":"Rajasthan gets AI-powered digital Lok Adalat","content":"Uday Umesh Lalit, the chairman of the National Legal Services Authority, has launched India’s first AI-powered digital Lok Adalat. The digital Lok Adalat by Rajasthan State Legal Services Authority (RSLSA) has been designed and developed by Jupitice Justice Technologies. As per the company’s website, Jupitice is a technology company that focuses on e-Commerce for digital justice services under an online marketplace model. Jupitice has designed and developed a private digital court for the private justice system. Here, individuals and organisations can file and resolve the disputes under ADR (Alternate Dispute Resolution) mechanism online on a single platform without any physical interaction. This platform will be used to dispose off pending disputes or disputes that are at the pre-litigation stage. The platform will provide end-to-end dispute resolution with various facilities- Drafting and filing of applications, Generation of e-notices through one-clickSmart templates to draft settlement agreementsDigital hearings through video conferencing tools It will also offer an AI-powered voice-based interactive chatbot and advanced data analytics tools. It will help in getting better insights into the functioning of Lok Adalat through custom reports and BI Dashboards for data-driven decision-making. Policymakers have been speaking about the benefits that AI can bring in India’s judicial system for quite some time now. Last year, law minister Kiren Rijiju said that AI could help in reducing the backlog of pending cases and provide sustainable justice delivery. Earlier in 2020, the then CJI SA Bobde talked about the possibility of developing AI for the court system to prevent undue delay in justice.","excerpt":"This platform will be used to dispose of pending disputes or disputes that are at the pre-litigation stage","categories":["AI News"],"tags":[],"author_name":"Sreejani Bhattacharyya","publish_date":"2022-07-18T11:48:03","publication_year":"2022","word_count":253,"keywords":["programming_languages:R","AI","data-driven","Git","analytics","GAN","R"],"extracted_tech_keywords":["AI","analytics","R","Git","GAN","data-driven","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/rajasthan-gets-ai-powered-digital-lok-adalat\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":9975,"title":"Want to make big bucks in stock market? Use Big Data Analytics","content":"Millions change hands in the stock markets in a second, rewarding some and impoverishing the others. What distinguishes the winners from the losers? The winners have the ability to accurately predict stock behavior on a given day or in a given market condition. This can either be pure genius or knowledge developed through years of experience of dealing in stocks. The key point is that they know how to analyze the data and decipher relevant patterns. Now, what if we told you that anyone could make accurate stock predictions? Moreover, the icing on the cake being that you don’t have to have an IQ of 9 and above or be experienced in the trade, as you can make predictions with data analytics. In fact, Big Data and Analytics today help institutions and individuals make better investment decisions with consistent returns by giving them insight into the data. Big Data and Cloud are a necessity Today, an increasing number of people are investing in bonds and mutual funds to save taxes. Good returns on investment translate into financial well-being of families. To be able to do this without burning a hole in the pocket, an investor needs to have knowledge about the market to make sound decisions. However, the biggest constraint being that they don’t have the financial wherewithal of institutions or big corporates to hire data scientists. In such a scenario, this technology is indispensable. Analytics is the new buzzword According to the Harvard Business Review, being a Data Scientist is the sexiest job of the 21st Century. Apart from the glamour, at a macro level, the adoption of Analytics and Big Data would also help in employment generation, an urgent need in the Indian context. Global revenue in the analytics market is forecast to reach $16.9 billion in 2016, according to the latest forecast from Gartner, the international research firm. How does it work? Big Data is mined and analyzed through algorithms that utilize huge amounts of historical data and insert them into complex mathematical models to provide accurate predictions to maximize portfolio returns. With the ever-increasing capabilities of computers to churn and handle data, automated computer programs perform stock trading at the best possible prices at speeds that a human broker cannot. It also removes the possibility of human error, thus creating less risky investments. Algorithms such as Apriori, FPGrowth (Frequent Pattern Growth) together with analytical methods such as Lift, Kulc, IR, Chi-square help users identify useful data and discard data that is of little value using highly sophisticated statistical techniques, factoring in real-time news, updates from social media and frequently changing stock data. Similarly, robo-advisors manage wealth and investments online through generating automated, algorithm-based portfolio management advice without the use of human planners by bringing huge amounts of data on a digital platform. While predictive analysis is out of reach of most small-time investors, there are affordable applications that abstract complex data through statistical approaches and market sentiment. Applications such as InTenMin and Investtech run high-speed analytics on historical and real-time stock data to provide in-depth statistics. This information is then correlated with sentiment analysis from social media in order to provide a 360-degree analysis on stocks. Data include billions of records of stock ticker information and live twitter feeds. However, the use of Analytics and Big Data in the Indian scenario remains restricted to a few. Data: Structured or Unstructured The New York Stock Exchange generates 1 terabyte of data every day, according to one estimate, which is undecipherable by a normal human being. Only a small part of such data is structured and which can be managed through relational databases and spreadsheets. Unstructured data, on the other hand, cannot be fit into a predetermined model. Investment banks and asset management firms need to analyze such voluminous data to make good investment decisions. Therefore, today, stock markets, financial institutions such as banks and capital markets, and insurance and retirement institutions increasingly use Analytics and algorithmic techniques for active risk management.","excerpt":"Millions change hands in the stock markets in a second, rewarding some and impoverishing the others. What distinguishes the winners from the losers? The winners have the ability to accurately predict stock behavior on a given day or in a given market condition. This can either be pure genius or knowledge developed through years of […]","categories":["IT Services"],"tags":[],"author_name":"Muqbil Ahmar","publish_date":"2016-05-23T09:13:52","publication_year":"2016","word_count":664,"keywords":["big data","Go","API","programming_languages:R","sentiment analysis","AI","programming_languages:Go","Git","analytics","R"],"extracted_tech_keywords":["AI","analytics","sentiment analysis","R","Go","Git","API","big data","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/want-make-big-bucks-use-big-data-analytics\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":20908,"title":"AI Based HRTech Startup Skillate Gets An Undisclosed Amount of Funding","content":"According to recent reports, Skillate, an AI based HRTech startup has raised an undisclosed amount of funding from Incubate Fund India and Anuj Agrawal, Director at Zyoin. The company aims to use the funding amount to accelerate product development and improve upon its customer base in India and abroad. Skillate is an online platform that lets companies leverage artificial intelligence and deep learning to simplify the hiring process. It empowers organisations by helping them build an highly efficient team. A recruitment platform which uses Natural Language Processing (NLP) and Deep Learning algorithms to increase the efficiency of overall recruitment system, the startup was founded by Bipul Vaibhav, Kumar Sambhav and Anand Kumar in 2016. The product by the startup screens the resume and find its relevance to a particular job description based on artificial intelligence. As the company website describes, the AI program can read the resume line by line to determine a candidate’s experience, expertise, education background etc. and matches it to the job description to derive a score. The program reads the resume and thinks like a recruiter while giving the desired result. The startup is backed by SAP and is on a mission to disrupt the hiring process through intelligent interventions to create an ecosystem that expedites hiring processes to help organisations grow. It firmly believes that progressive organisations are moving to a man-machine ecosystem of talent acquisition – augmenting human IQ and decision making with machine’s ability to process and apply learning at scale.","excerpt":"According to recent reports, Skillate, an AI based HRTech startup has raised an undisclosed amount of funding from Incubate Fund India and Anuj Agrawal, Director at Zyoin. The company aims to use the funding amount to accelerate product development and improve upon its customer base in India and abroad. Skillate is an online platform that […]","categories":["AI News"],"tags":["ai funding india"],"author_name":"Srishti Deoras","publish_date":"2018-01-22T12:48:18","publication_year":"2018","word_count":248,"keywords":["Go","artificial intelligence","AI","ML","RAG","NLP","Aim","deep learning","GAN","R","ai funding india"],"extracted_tech_keywords":["AI","artificial intelligence","ML","deep learning","NLP","Aim","RAG","R","Go","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/ai-based-hrtech-startup-skillate-gets-undisclosed-amount-funding\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10042552,"title":"Explained: Facebook’s New Approach To Audio-Visual Separation","content":"Facebook AI researchers Ruohan Gao and Kristen Grauman have introduced VisualVoice, a new approach for audio-visual speech separation. “Whereas existing methods focus on learning the alignment between the speaker’s lip movements and the sounds they generate, we propose to leverage the speaker’s face appearance as an additional prior to isolate the corresponding vocal qualities they are likely to produce,” the researchers said. The human perceptual system draws heavily on visual information to reduce ambiguities in the audio and modulate attention on an active speaker in a busy environment. Automating this process of speech separation has many valuable applications, including: Assistive technology for the hearing impaired.Superhuman hearing in a wearable augmented reality device.Better transcription of spoken content in noisy in-the-wild internet videos. Procedure The researchers first formally defined the problem before presenting the audio-visual speech separation network. The paper then introduced learning audiovisual speech separation and cross-modal face-voice embeddings in a multi-task learning framework and finally presented the training criteria and inference procedures “We use the visual cues in the face track to guide the speech separation for each speaker. The visual stream of our network consists of two parts: a lip motion analysis network and a facial attributes analysis network,” the paper read. The lip motion analysis network includes: The lip motion analysis network takes N mouth regions of interest (ROIs) as input, and it consists of a 3D convolutional layer. This is followed by a ShuffleNet v2 network to extract a time-indexed sequence of feature vectors. They are then processed by a temporal convolutional network (TCN) to extract the final lip motion feature map of dimension Vl×N. For the facial attributes analysis network, researchers used: ResNet-18 network that takes a single face image randomly sampled from the face track as input to extract a face embedding that encodes the facial attributes of the speaker.Then replicate the facial attributes feature along the time dimension to concatenate with the lip motion feature map and obtain a final visual feature. “The facial attributes feature represents an identity code whose role is to identify the space of expected frequencies or other audio properties for the speaker’s voice, while the role of the lip motion is to isolate the articulated speech specific to that segment. Together they provide complementary visual cues to guide the speech separation process.” On the audio side, the team uses a U-Net style network tailored to audio-visual speech separation. It consists of an encoder and a decoder network. Result & scope Lip motion is directly correlated with the speech content and is much more informative for speech separation. However, the performance of the lip motion-based model suffers dramatically when the lip motion is unreliable, as often the case in real-world videos. “Our VISUALVOICE approach combines the complementary cues in both the lip motion and the face-voice embedding learned with cross-modal consistency and thus is less vulnerable to unreliable lip motion,” said researchers. Check out the demo here. Cross-modal embedding learning could benefit from the Facebook model’s joint learning. Researchers intend to evaluate the cross-modal verification task, in which the system must decide if a given face and voice belong to the same person. “Our design for the cross-modal matching and speaker consistency losses is not restricted to the speech separation task and can be potentially useful for other audio-visual applications, such as learning intermediate features for speaker identification and sound source localization. As future work, we plan to explicitly model the fine-grained cross-modal attributes of faces and voices and leverage them to further enhance speech separation,” the researchers concluded.","excerpt":"Our VISUALVOICE approach combines the complementary cues in both the lip motion and the face-voice embedding learned with cross-modal consistency.","categories":["AI Features"],"tags":["cnn","Facebook","Facebook AI","resnet"],"author_name":"kumar Gandharv","publish_date":"2021-06-29T13:00:00","publication_year":"2021","word_count":589,"keywords":["Replicate","Facebook AI","AI","AWS","Modal","ML","cloud_platforms:AWS","ResNet","cnn","RAG","resnet","Facebook","AI research","R"],"extracted_tech_keywords":["AI","ML","RAG","AWS","R","ResNet","Replicate","Modal","AI research","cloud_platforms:AWS"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/explained-facebooks-new-approach-to-audio-visual-separation\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10103975,"title":"Genpact, AWS Collaborate to Revolutionise Insurance Claims Lifecycle","content":"Genpact has announced its collaboration with Amazon Web Services (AWS) and Amazon Business to redefine the property loss replacement and claims management approach. Leveraging its extensive claims management expertise, Genpact is integrating Amazon Bedrock‘s generative AI capabilities and Amazon Business procurement API integrations to streamline the claims submission process. Check out more details here. This advanced pricing automation workflow aims to reduce the submission timeline from weeks to days, focusing on efficient replacement item identification and timely delivery of policyholder estimates. Sameer Dewan, global operating officer at Genpact, emphasised the transformative impact of AI on the insurance industry, stating, “AI is fundamentally reshaping the landscape of the insurance industry. Our Genpact AI-driven automated pricing workflow, powered by AWS, is transforming the research, significantly reducing the time adjusters spend investigating by as much as 75 percent. By automating routine tasks and enhancing decision making, our AI solution is empowering smarter pricing decisions, expediting claims settlements, and bringing about a profound transformation in the customer experience.” Genpact aims to deliver measurable value to its clients through ongoing innovation. Future plans involve the integration of additional AI services into the automated workflow, including real-time summaries of market values for claimed contents and optimal action recommendations for further processing. Recently, Genpact also announced an expanded collaboration with AWS aimed at revolutionising financial crime risk operations through the integration of generative AI and LLMs. The partnership entails the integration of Genpact’s cloud-based financial crime suite, riskCanvas, with Amazon Bedrock, yielding significant efficiencies and benefits for clients, including Apex Fintech Solutions. Read: Bedrock is All About Choices","excerpt":"Genpact also announced an expanded collaboration with AWS aimed at revolutionising financial crime risk operations through the integration of generative AI and LLMs.","categories":["AI News"],"tags":["AWS","Gen AI in Insurance","Genpact","genpact AI"],"author_name":"Mohit Pandey","publish_date":"2023-12-01T15:25:00","publication_year":"2023","word_count":261,"keywords":["Go","API","AWS","AI","Genpact","ML","genpact AI","RAG","automation","Aim","Gen AI in Insurance","generative AI","R"],"extracted_tech_keywords":["AI","ML","generative AI","Aim","RAG","AWS","R","Go","API","automation"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/genpact-aws-collaborate-to-revolutionise-insurance-claims-lifecycle\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10065373,"title":"How TensorFlow’s MoViNets are solving problems of CNNs in video recognition","content":"Tensorflow recently released Mobile Video Networks (MoViNets), a family of computation and memory-efficient video networks operating on streaming video for online inference. This overcomes the problem of 3D CNNs. This article explores the challenges of 3D CNNs, how MoViNets overcomes them, and the tool’s architecture. The problem In machine learning, video classification is the solution of taking video frames as input and predicting a single class from a larger set of classes as output. This makes it important for the video action recognition model to consider the content of each frame. It can also understand the spatial relationships between adjacent frames and the actions in the video. 3D convolutional neural networks are an extended version of 2D CNNs and are used to extract sequential images and learn spatiotemporal information from videos. While they can learn the correlation of temporal changes between adjacent frames without employing additional temporal learning methods, 3D CNNs have a huge inherent disadvantage. They have high computational complexity and excessive memory usage. Furthermore, they do not support online inference, making them difficult to work on mobile devices. Even the recent X3D networks provide increased efficiency and fall short in one way or another. They require extensive memory resources on large temporal windows, which incur high costs, or on small temporal windows, which reduce accuracy. Hence, there is a large gap between the video model performance of accurate models and efficient models for video action recognition. 2D MobileNet CNNs are fast and can operate on streaming video in real-time but are prone to be noisy and inaccurate. MoViNets TensorFlow’s MoViNets proposes a three-step approach to improve computational efficiency while substantially reducing the peak memory usage of 3D CNNs. MoViNets are a family of CNNs that efficiently process video streams and accurate output predictions with a fraction of the latency of CNN video classifiers. The model has demonstrated state-of-the-art accuracy and efficiency on several large-scale video action recognition datasets. It does so in three essential steps: Design a video network search space for the neural architecture to generate efficient and diverse 3D CNN architectures. The Stream Buffer technique is used to decouple memory from video clip duration, so CNNs can embed arbitrary-length streaming video with lesser memory usage.Ensembling technique that improves accuracy without sacrificing efficiency. MoViNets is trained on the Kinetics-600 dataset, a collection of a large-scale and high-quality set of URL links to 650,000 video clips. The dataset consists of human-annotated clips observing 400\/600\/700 human action classes, including human-object and human-human interactions. Through the training, MoViNets can identify 600 human actions like playing the trumpet, robot dancing or bowling. It can also classify video streams captured on a modern smartphone in real-time. Architecture MoViNets allows the user to enjoy the benefits of 2D frame-based classifiers and 3D video classifiers while mitigating their disadvantages. It does so with a hybrid approach and replaces 3D CNNs with causal convolutions. Causal convolutions, a form of convolution, are used for temporal data and ensure models cannot violate the ordering in which the developers model the data. This allows users to cache intermediate activations across frames with a Stream Buffer. The technique copies the input activations of all 3D operations – an output by the model and inputs them back into the model on the next clip input. As a result, MoViNets can receive one frame input at a time. This reduces peak memory usage with no loss of accuracy. In 3D CNNs, given the model is processing all frames in a video clip simultaneously, it takes up significant memory. It also searches for efficient configurations of models using the Neural Architecture Search (NAS), a widely used technique for automating the design of artificial neural networks. It searches for model configurations on video datasets across network width, depth, and resolution. It creates a set of action classifiers that output temporally-stable predictions that smoothly transition based on frame content. Given the lack of temporal reasoning, this overcomes the problem of output predictions on 2D frames with sub-optimal performance. Model performance “These three progressive techniques allow MoViNets to achieve state-of-the-art accuracy and efficiency on the Kinetics, Moments in Time, and Charades video action recognition datasets,” said the team. The accuracy of MoViNet-A5-Stream is the same as X3D-XL on Kinetics 600, all while requiring 80% fewer FLOPs and 65% less memory There are various model modifications yet, and more yet to come. MoViNet-A0-Stream, MoViNet-A1-Stream, and MoViNet-A2-Stream represent the smaller models that run in real-time. Architectural modifications such as replacing the hard swish activation with ReLU6 and removing the Squeeze-and-Excitation layers allowed the team to quantize MoViNet without an accuracy drop. Further, the models were converted to TensorFlow Lite with integer-based post-training quantization to reduce the model’s size and ensure faster running on mobile CPUs. As a result, they can provide accurate predictions.","excerpt":"MoViNets are a family of CNNs that efficiently process video streams and accurate output predictions with a fraction of the latency of CNN video classifiers.","categories":["AI Features"],"tags":["Tensorflow"],"author_name":"Avi Gopani","publish_date":"2022-04-21T10:00:00","publication_year":"2022","word_count":796,"keywords":["machine learning","TPU","programming_languages:R","AI","neural network","ai_frameworks:TensorFlow","CLIP","CNN","TensorFlow","R","Tensorflow"],"extracted_tech_keywords":["AI","machine learning","neural network","TensorFlow","TPU","R","CLIP","CNN","ai_frameworks:TensorFlow","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-tensorflows-movinets-are-solving-problems-of-cnns-in-video-recognition\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10126538,"title":"Anthropic Introduces Fine-Tuning for Claude 3 Haiku on Amazon Bedrock","content":"Anthropic has launched fine-tuning capabilities for Claude 3 Haiku on Amazon Bedrock. The feature is available for preview in the US West (Oregon) AWS region, allowing businesses to customise the Claude 3 model for specific tasks. Claude 3 Haiku, which is a part of the Claude 3 family alongside Sonnet and Opus, can now undertake tasks like classification, API interactions, and data interpretation. The fine-tuning process uses prompt-completion pairs to enhance the model’s performance in specialised areas. The entire process is guided through the Amazon Bedrock console or API, where users can test and refine their custom models before deployment. Several enterprise customers have already begun using the fine-tuning feature, allowing customisation of Haiku to fit their needs. SK Telecom reported a 73% increase in positive feedback for agent responses and a 37% improvement in key performance indicators after implementing a fine-tuned Claude model. Thomson Reuters plans to fine-tune Claude 3 Haiku with their industry expertise. The features serves as a boon to enterprise customers, as it gives them more control over the model to train it according to their needs. Specifically, allowing the fine-tuning of Haiku will help in improving accuracy in specialised tasks, allow for faster processing at lower costs, give enterprises consistent output formatting, make the API accessible for companies, and ensuring data security for AWS customers. As AI technology advances, customisation options like fine-tuning have become more abundant, especially as offerings from cloud infrastructure providers. This development in AI customisation could impact various sectors, from customer service to data analysis, as businesses seek to integrate AI more deeply into their operations. Additionally, other cloud infrastructure companies like Microsoft Azure and GCP have also increased their GenAI offerings to businesses, giving them the option to build agents on the models they offer.","excerpt":"The fine-tuning process uses prompt-completion pairs to enhance the model’s performance in specialised areas.","categories":["AI News"],"tags":["Amazon Bedrock","Claude"],"author_name":"Donna Eva","publish_date":"2024-07-11T16:43:00","publication_year":"2024","word_count":295,"keywords":["Anthropic","Go","Amazon Bedrock","GenAI","GCP","AWS","AI","TPU","R","Claude","API","Azure"],"extracted_tech_keywords":["AI","GenAI","Anthropic","AWS","Azure","GCP","TPU","R","Go","API"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/anthropic-introduces-fine-tuning-for-claude-3-haiku-on-amazon-bedrock\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10076050,"title":"Decentralised Science Isn&#8217;t Just Possible—It&#8217;s Almost Here","content":"Science has problems. Many scientific studies make outrageous and hard-to-digest claims day in and day out. Silent, not deadly; how farts cure diseases, a glass of red wine is equivalent to an hour at the gym and everything we eat both causes and prevents cancer are examples of problematic science. There is a growing concern that many, if not the vast majority of published results, may be partially or entirely false and that they are rarely compared to rapidly expanding research output. Besides, replications are limited to small scales due to limited professional incentives. In 2018, Valentin Danchev, an alumnus of the University of Chicago, released a paper in which he concluded that centralised scientific communities were less likely to generate replicable results. He argued that centralised, overlapping scientific collaborations weaken our collective understanding because they spread ideas that are less likely to be independently verified, even when generally accepted. Blockchain can bring decentralisation, improve the essential practices in science and provide a fair mechanism to incentivise professionals. Blockchain technology can also ensure transparency and make scientific findings available to the general public. DeSci can fix centralised science DeSci or decentralised science, is a movement to improve science using Web3 tech and tools. It seeks to establish a scientific system more democratic and equal. DeSci enables a more decentralised and distributed model of scientific research, strengthening its resistance to censorship and government regulation. By decentralising access to money, scientific resources, and communication channels, DeSci aims to foster an atmosphere conducive to the development of novel and innovative ideas. The goal is to make science more open, collaborative and accountable. How to solve other problems In 2016, Vox conducted a survey on the challenges of the scientific community. The survey, in which 270 scientists participated, found that lack of funds, poorly designed studies, reproducibility crisis and broken peer review, among others, were some of the challenges scientists faced. DeSci solves the problem of funding for scientists. Instead of applying for grants from funding bodies, a decentralised community crowdsources the funds for research. They also own any value produced, not research institutions or companies. Replicability and reproducibility are the foundations of quality scientific discovery, which can be ensured using new Web3-native technologies. These tools enable academics to test high-quality science. In a consensus system, each analytical component — raw data, computational engine, and application result — can be attested to. Then, it would be up to this system to replicate the calculation and verify each output. In conventional science, the ownership of digital assets like scientific data or articles is a major issue. Web3 uses NFTs to do this, and they do it incredibly well. NFTs can make it possible to guarantee that the original authors of digital assets receive a share of the proceeds from subsequent sales. Similarly, transparent value attribution chains can be used to pay researchers, authorities, or even the individuals whose data is being collected. As a result, the benefits and value in the digital economy would be distributed more fairly. DAOs that are changing science In the biotech space, DAOs are taking a novel approach to organise and incentivise collaborative research. Experts believe that by enabling new methods of optimising collaboration, talent, and capital allocation, they have the potential to break Eroom’s law (drug discovery becoming slower and more expensive over time). There are other famous DAOs in the medical space. VitaDAO is democratising longevity research by utilising a crowd-sourced funding model for early-stage research projects. For example, it aided a group of Danish researchers in raising funds to repurpose existing drugs as anti-ageing medications. Opscientia, another science DAO, is establishing a network of knowledge foundries. It brings citizen science communities together while also providing a framework for incentivising their contributions to collaborative research. Right questions will give birth to new discoveries When money is scarce, scientists frequently forgo riskier but more radical ideas in favour of safe, dependable research or even pseudoscience.  Martin Cooper, an engineer and a pioneer in the wireless communication industry, was heading a project at Motorola in the 1970s, where he was developing the new-age car phones. During that time, he came up with a question — when we call a person, why do we always call a particular place instead of a person? In 1973, Motorola became the first company to launch a mobile phone. French philosopher Claude Lévi-Strauss once said: “The scientist is not a person who gives the right answers, he is one who asks the right questions.” Fundamentally, the article argued that DeSci or decentralised science using web 3 tech and tools is the way forward as it is based on the distributed and decentralised model of science which makes it resistant to censorship, government regulations, and ownership conflict.  Other challenges like obtaining funding are another major obstacle facing the conventional scientific community, along with limiting information. Researchers have also been known to give up on potentially revolutionary projects due to traditional grant systems being too strict, leading to the destruction of potentially revolutionary ideas. In my opinion, DeSci does solve many of these conventional science challenges, for example, the ownership problem is taken care of by de-platforming and passing shares from subsequent sales by removing gatekeepers. In other words, on Web3 everything you do will be owned by you, not by any platform claiming to be your guardian angel that in turn bring benefit to your wallet without any middlemen involved. Moreover, decentralised science aims to create a system where scientific papers are distributed and trusted without the need for traditional publisher gatekeepers, so peer-review is more transparent and streamlined. Additionally, to advance scientific development, decentralised autonomous organisations (DAOs) leverage the power of the community. Despite all the likely possibilities that might open for us, there are certain things that we should be careful about how the review process will work under such a decentralised system. For example, if someone shares a research study and in absence of any scrutiny what will happen if something ends up wrong? Similarly, the patent system in conventional science is granted to an organisation after a huge investment of time and resources, and one wonders how it would be handled since decentralised science should be devoid of regulations. Consider a cancer drug that claims to cure the disease, how are we going to regulate it when deregulation is a sacred principle? Finally, if you fail in DeSci then the entire architecture will collapse, so it is best of everyone’s interest to answer these questions before making any conclusions. So, to conclude, the DeSci does bring new funding opportunities, higher credibility, more democratised access to resources, and nurtures collaborations and innovation which outwait the benefits of modern scientific research.","excerpt":"DeSci or decentralised science is a movement to improve science using Web3 tech and tools","categories":["IT Services"],"tags":["Blockchain","Ethereum","EVM","Science"],"author_name":"Tausif Alam","publish_date":"2022-09-29T14:00:04","publication_year":"2022","word_count":1117,"keywords":["Go","API","TPU","Blockchain","AI","ML","EVM","Git","Science","RAG","Aim","Rust","R","Ethereum"],"extracted_tech_keywords":["AI","ML","Aim","RAG","TPU","R","Go","Rust","Git","API"],"url":"https:\/\/analyticsindiamag.com\/it-services\/decentralised-science-isnt-just-possible-its-almost-here\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10053215,"title":"How to Visualize Deep Learning Models using Visualkeras?","content":"Building a machine learning or deep learning model for accurate predictions is important but adding interpretability along with these models can make the process more useful and interesting. There are a few methods to visualize the defined predictive models but visualizing a deep learning model with its complex structure is a challenge. In this article, we will learn to visualize the deep learning models in order to achieve interpretability. We will go through different steps to see how to customize these visualizations of the deep learning models to make them more understandable. The major points to be covered in this article are listed below. Table of Contents Need of InterpretabilityAbout ViusualKerasVisualization of neural networks Need of Interpretability Most of the machine learning models are considered black-box models, especially the neural networks. It is not easy to understand how a defined model is functioning with the data. The approach to making the models understandable and interpretable by everyone is very important. Considering this scope, interpretability or explainability is trending nowadays. In the context of deep learning, there are different approaches used for explainable deep learning. Here we are going to discuss how can we visualize a defined deep learning model so that any person can understand that a model which is giving so accurate results, how does it look. About VisualKeras For the visualization of deep learning models, we will use a python package named visualkeras. Let us understand about this package before going forward. Visualkeras is a python package with the facility of visualization of deep learning models architecture build using the Keras API also we can visualize the networks built using Keras included in TensorFlow. This library supports the layered and graph style architecture of neural networks. Visualization of Deep Learning Models In this section, we will see how we can define and visualize deep learning models using visualkeras. Let us go through the elbow steps. 1. Installing Dependency Let’s start with the installation of the library. Using the following code we can install the visualkeras package. pip install visualkeras Output: 2. Defining the Model As a next step, we are making a simple model for this we are required to import some libraries. from tensorflow import keras from tensorflow.keras import layers, models For a small example, we can make a sequential model with convolutional layer and pooling layers. model = models.Sequential() model.add(layers.Conv2D(32, (3, 3), activation='relu', input_shape=(32, 32, 3))) model.add(layers.MaxPooling2D((2, 2))) model.add(layers.Conv2D(64, (3, 3), activation='relu')) model.add(layers.MaxPooling2D((2, 2))) model.add(layers.Conv2D(64, (3, 3), activation='relu')) Let’s check for the summary of the defined model. model.summary() Output: Here we can see that the model we have defined a model with 3 convolutional layers and 2 pooling layers in the network. Now the question which comes to mind is how we can visualize it? 3. Visualizing the Model Let’s import the visualkeras package import visualkeras Now we are ready to visualize the defined network. This can be done by simply using the following code. visualkeras.layered_view(model) Output: Here we can see that convolutional layers are in yellow and pooling layers are in pink colour as we have seen in the summary there are three convolutional and 2 pooling layers. 4. Adding Dense Layer Let’s add a dense layer to the network. model.add(layers.Dense(64, activation='relu')) Visualizing the model added with a dense layer Here we can see that now we have different colours for dense layers. 5. Labelling the Layers Since we are talking about interpretability it will be more interpretable visualization if the name of the layers is assigned with layers themselves. Using this package we can also assign the name of the layer in the visualization. For this, we need to import the image font for the PIL library. from PIL import ImageFont Image font can be used in with the visualkeras from PIL import ImageFont visualkeras.layered_view(model, legend=True, font=font) Output: 6. Adding Flatten Layer Adding a flatten layer in the model. model.add(layers.Flatten()) visualkeras.layered_view(model, legend=True, font=font) Output: 7. Visualising in 2-D Space We can also visualize the network in 2D space or we can say in flat style using the following codes. visualkeras.layered_view(model, legend=True, font=font, draw_volume=False) 8. Customizing the Space between Layers We can manage to space between the layers. visualkeras.layered_view(model, legend=True, font=font, draw_volume=False,spacing=50) Output: 9. Customizing the Layers Colours We can also customize the colours of the layers. from collections import defaultdict color_map = defaultdict(dict) color_map[layers.Conv2D]['fill'] = 'orange' color_map[layers.MaxPooling2D]['fill'] = 'red' color_map[layers.Dense]['fill'] = 'black' color_map[layers.Flatten]['fill'] = 'teal' visualkeras.layered_view(model, legend=True, font=font,color_map=color_map) Output: Here we can see how we can visualize a deep learning model built using Keras. The layers I have used in the model are some of the most used layers in the field of modelling neural networks. Final Words In the article, we learnt how to visualize a deep learning model using a python package named visualkeras. We saw how to plot the models with so many customizations to make them understandable and interpretable. This visualization is not much difficult and can be done very quickly. Adding such visualizations in the deep learning-based reports can make the report more attractive and interpretable. Reference: visualkeras for Keras \/ TensorFlowLink to the above codes","excerpt":"deep learning models are considered black-box models. It is not easy to understand how a defined model is functioning with the data. visualizing the deep learning models can help in improve interpretability.","categories":["AI Trends"],"tags":["AI (Artificial Intelligence)","Data Science","Deep Learning","Machine Learning","python visualize neural network","Visualization"],"author_name":"Yugesh Verma","publish_date":"2021-11-11T15:00:00","publication_year":"2021","word_count":847,"keywords":["Go","machine learning","Keras","TPU","AI","neural network","Visualization","Machine Learning","Python","deep learning","python visualize neural network","Deep Learning","Data Science","TensorFlow","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","machine learning","deep learning","neural network","TensorFlow","Keras","TPU","Python","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/how-to-visualize-deep-learning-models-using-visualkeras\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10172759,"title":"Accel&#8217;s Case for the Application Layer in India","content":"With the IndiaAI mission gaining traction, investing in AI across different layers, particularly the foundational layer, has come under a tighter spotlight. As the debate around foundational models becoming commoditised increase, many contend that the edge shifts to those who can execute quickly, integrate deeply, and build for real-world use cases. In a recent conversation with AIM, Prayank Swaroop, a partner at Accel, spoke about the VC fund’s investment thesis in India. He believes that India’s most promising opportunities in AI lie elsewhere. “While we’re closely monitoring developments in foundational models, the highest upside right now is clearly in applied AI,” he said. Recently, IndiaAI Mission announced the selection of three more startups—Soket AI, Gnani.ai, and Gan.AI—to develop indigenous foundation models. This brings the number of startups under the foundation model development initiative to four, including the previously announced Sarvam AI. Sarvam’s funding comes from investors like Lightspeed India Partners, Peak XV Partners, Lightspeed Venture Partners, and Khosla Ventures, among others. The announcement drew a slew of criticism on social media. Swaroop asserted that one doesn’t need to be OpenAI to create a transformative AI product. This highlights the fact that many of the most innovative developments are coming from companies that leverage existing models like GPT and Claude. Inside Accel’s Investment Thesis According to Swaroop, Accel is increasingly concentrating on AI across three main areas: agentic enterprise platforms, vertical AI products and AI-enabled services. “This surge [of AI adoption] is driven by India’s unmatched engineering talent pool, cost efficiency, and access to domain-specific datasets,” Swaroop said. The firm’s latest $650 million fund is designed to support startups that bring clear use cases. A key part of its strategy is product-led growth, backing companies that can scale through user demand rather than relying heavily on sales teams, which they believe gives these startups a stronger foundation for long-term success. Why the Application Layer? However, this raises the question: why is the application layer considered India’s stronghold in AI? Swaroop makes the case for an AI surge in India driven by its engineering talent pool, cost efficiency, and access to domain-specific datasets. Swaroop believes that the bottleneck has shifted from engineering resources to product thinking and strategic depth. “AI is making software development easier, but the winners will be those who build durable products that solve high-value pain points,” he said. According to him, founders who adopt a “build for India, scale for the world” mindset, stay user-obsessed, and execute with speed are best positioned to create globally relevant solutions. He argued that the “services as software” model, where India’s traditional BPO strength is transformed through AI automation, is gaining significant traction. Furthermore, India’s unique advantage lies in the application layer as startups are leveraging open-source and accessible foundational models to build verticalised solutions in healthcare, legal operations, and financial services. What’s Everyone’s Missing About AI? “If anything, AI remains underhyped in terms of its true potential,” he said, talking about how the breadth and depth of AI’s impact across sectors are only beginning to be realised. Swaroop explained that, contrary to common belief, AI’s potential remains underhyped in several key sectors. Agentic AI is just beginning to demonstrate its power with examples like Genspark (coding), Manus (enterprise workflows), and August.ai (preventive healthcare), showing early signs of a broader shift in productivity tools. Another underrated opportunity, he believes, is India-first consumer AI: products built for local languages, cultural contexts, and price sensitivity. AI-powered regional entertainment and Bollywood content generation are in their early days, gaining traction. Beyond Foundational Models The recent acquisition of Windsurf by OpenAI and the appointment of a CEO of applications signal a clear convergence between foundational model development and AI applications. “You don’t have to be OpenAI to build a transformative AI product, as many successful startups, such as Cursor, RapidClaims, Chronicle and Rocket.new from our portfolio, are building on top of existing models like GPT and Claude,” Swaroop said. However, this doesn’t close the door for startups, he argues. The above-mentioned startups are creating meaningful solutions by leveraging existing models, such as GPT or Claude. The competitive edge lies in addressing deep user pain points, maintaining control over the product stack, and optimising for performance and cost. Swaroop argued that localisation and sensitivity to price and context are essential for capturing the Indian market. He talks about how while for most Indian founders, building on top of open-source or accessible LLMs offers better returns. However, there is growing interest in building India-specific models that account for local regulation, language diversity, and cost constraints.","excerpt":"“You don’t have to be OpenAI to build a transformative AI product.”","categories":["AI Startups"],"tags":["Accel","Indian AI startups"],"author_name":"Aditi Suresh","publish_date":"2025-07-02T18:30:00","publication_year":"2025","word_count":757,"keywords":["Indian AI startups","API","agentic AI","OpenAI","AI","Accel","RAG","GPT","Ray","Aim","foundation models","R"],"extracted_tech_keywords":["AI","foundation models","agentic AI","OpenAI","Aim","Ray","RAG","R","API","GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-startups\/accels-case-for-the-application-layer-in-india\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10023680,"title":"Pinterest Open-Sources Big Data Analytics Tool Querybook: Major Features &#038; Set Up Instructions","content":"Recently, Pinterest open-sourced its big data analytics tool Querybook that started its life as an intern project in 2017. Querybook auto analyses executed queries to provide data lineage, example queries, frequent user information, and search\/auto-completion ranking. In 2018, the big data tool was released internally and soon became the official solution to query big data at Pinterest. According to its developers, Querybook is built to provide a simple web UI for big data analysis to discover the right data, compose queries, and share findings. Behind Querybook Querybook is a Big Data IDE that allows users to create, discover, and share data analyses, queries, and tables. The tool includes Database, Redis, ElasticSearch and remote storage. The Database is used to store the DataDocs, and MySQL is recommended. Redis is required to send async tasks to workers, maintain multi-server WebSocket connections, and caching live data for collaborative editing. ElasticSearch provides search functionality for database documents such as DataDocs and tables. Lastly, the remote storage stores the query results. Querybook has three key components- Web Server: The Web Server is used to handle HTTP requests, send\/receive Websocket messages, and provide the static assets for the web.Worker: This component is mainly used to execute long-running queries and scheduled DataDocs. It can also be used for auxiliary tasks such as updating ElasticSearch docs or analysing query lineage. Scheduler: Scheduler reads the task schedule from the database and sends it to the Celery workers. Features Querybook compose queries with autocompletion and hovering tooltip.The tool uses both scheduling and charting in DataDocs to build dashboards.Querybook has built-in rich-text support for users. The tool allows live query collaborations. Users can add additional documentation to the tables.Using this tool, one can get lineage, sample queries, frequent user, search ranking based on past query runs. How To Use It According to its developers, there are two main ways to set up the big data analytics tool. Single-Machine Instant Setup (locally or on a server): The single machine method is a quick way to try out Querybook for less than five users. This method uses docker-compose to bring up all the necessary infrastructure, which is why Docker needs to be installed for quick setup. For installation, open terminal and run the following: git clone https:\/\/github.com\/pinterest\/querybook.git cd querybook Now run the following: make Multi-Machine Setup The multi-machine setup is required when someone wants to scale Querybook for thousands of users. The multi-machine set up runs Querybook containers on different machines\/pods. This method is more complicated than the single machine instant set up and requires external infrastructure. The set up includes the following steps- Step-1: Requirements A MySQL\/PostgresSQL[^1] database with version >=5.7. It is recommended to have more than 5GB of space.An Elasticsearch server with version 6.6.1.A 2GB Redis instance, Querybook should not use more than 1GB of memory.If OAuth will be used for authentication, remember to get the OAuth client information (secrets, token url, etc).For notifications, you would need either a Slack API Token or an email address and the email server running on port 25 of the web server. Step-2: Choose the instances In this step, users will need to deploy three different services for Querybook. The web servers handle the HTTP\/WebSocket traffic, the workers handle the async tasks such as running the query, and the scheduler sends scheduled tasks to the workers. Also, it is important to make sure to only have one instance of scheduler running to prevent duplication in scheduled tasks and have at least two workers for rolling restart deployments. Step-3: Update your environment variables configuration Step-4: Start each service You can start each service by the following commands: Webserver: make webCelery worker: make workerScheduler: make scheduler Wrapping Up To make it generic while preserving some of the Pinterest-specific integrations, the developers decided to have a two-layer organisation through a plugin system and add an Admin UI. The Admin UI allows organisations to configure the query engines, table metadata ingestion, and access permissions from a single friendly interface. The plugin system integrates Querybook with the internal systems at Pinterest by utilising Python’s importlib.","excerpt":"Querybook’s core focus is to make composing queries, creating analyses, and collaborating with others as simple as possible.","categories":["AI Trends"],"tags":[],"author_name":"Ambika Choudhury","publish_date":"2021-04-09T18:00:00","publication_year":"2021","word_count":676,"keywords":["Elasticsearch","AI","docker","Git","RAG","Python","analytics","SQL","R","Redis"],"extracted_tech_keywords":["AI","analytics","RAG","docker","Redis","Elasticsearch","Python","R","SQL","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/pinterest-open-sources-big-data-analytics-tool-querybook-major-features-set-up-instructions\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10172059,"title":"Just Energy Partners with HCLTech for AI-Led Business Transformation","content":"HCLTech has announced in a statement that it has been selected by Just Energy, a leading US-based energy supply company, to enhance the latter’s operations and customer experience. The company will use its integrated Digital Process Outsourcing solutions suite and GenAI platform AI Force to enhance operational efficiency across Just Energy’s IT, finance, analytics, customer care, sales and renewals functions. HCLTech will also deploy digitalCOLLEAGUE, its comprehensive and role-specific single-UI platform and Toscona, its business process optimisation suite, to improve workforce collaboration and business process management. “We are confident that HCLTech’s proven expertise and commitment to service excellence will help us achieve our key business objectives relating to operational efficiency and service improvements,” Scott Fordham, chief operating officer of Just Energy, said. “By combining our expertise in GenAI and digital process outsourcing, HCLTech will contribute significantly to Just Energy’s innovation strategy and customer satisfaction,” Ajay Bahl, chief growth officer, Americas, Manufacturing and Allied Industries, HCLTech. HCLTech has been seeing traction in its AI offerings, as highlighted by the company’s CEO, Vijayakumar C, during the company’s Q2 2025 earnings call. He shared that AI has been integrated into the majority of their recent deals, highlighting the platform’s key role in service transformation. “We have seen strong wins with most of our deals incorporating AI capabilities. Our platform AI Force is gaining widespread adoption for service transformation among our clients,” Vijayakumar said. He further stressed how the company’s AI offerings, such as HCLTech AI Force, HCLTech AI Foundry, and AI Labs, are driving innovation across sectors.The CEO had outlined the company’s three-pronged generative AI approach. The first dimension is AI Force, which is driving service transformation and adoption among existing clients. “We have almost 25 clients who have started working on AI Force as the primary efficiency platform for all their IT and business processes,” he shared.","excerpt":"The company will use its integrated Digital Process Outsourcing solutions suite and GenAI platform AI Force to enhance operational efficiency across Just Energy’s functions.","categories":["AI News"],"tags":["AI","hcltech","JustEnergy"],"author_name":"C P Balasubramanyam","publish_date":"2025-06-19T17:52:23","publication_year":"2025","word_count":305,"keywords":["GenAI","programming_languages:R","AI","JustEnergy","innovation","Git","generative AI","analytics","hcltech","R"],"extracted_tech_keywords":["AI","analytics","generative AI","GenAI","R","Git","innovation","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/just-energy-partners-with-hcltech-for-ai-led-business-transformation\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10076307,"title":"India’s Indigenous Navigation System Can Make Your Phones Expensive","content":"In February 1999, when Pakistani soldiers set up bases on 132 vantage points of the Indian-controlled region at Kargil, the Indian government sought global positioning system (GPS) data for the region from the US government; however, the request was denied. The data would have played a pivotal role and significantly helped India cut down on its losses. While discussions around building an indigenous satellite navigation system were already taking place during that time, the Kargil War expedited the need and Navigation with the Indian Constellation (NavIC) was born. More than two decades after the Kargil War, the government is now looking to mandate every phone manufacturer operating in India to make their smartphones compatible with NavIC. NavIC The decision to develop an indigenous satellite navigation system was taken in 2006, and a budget of around INR1426 crore was approved. The navigation system, which has been operational since 2018, was initially named the Indian Regional Navigation Satellite System (IRNSS); however, it was later renamed NavIC (Navigator in Hindi) by Prime Minister Narendra Modi. Developed by the Indian Space Research Organisation (ISRO), NavIC consists of eight satellites: IRNSS-1A, IRNSS-1B, IRNSS-1C, IRNSS-1D, IRNSS-1E, IRNSS-1F, IRNSS-1G, IRNSS-1I. It covers India’s entire landmass and up to 1,500 kilometres from its boundaries. There are plans to increase the number of satellites to 11 to further expand the coverage. NavIC’s constellation consists of three Geostationary Earth Orbit (GEO) satellites and four Geosynchronous Orbit (GSO) satellites. Unlike GPS, NavIC satellites use dual frequency bands, L5 (1.17GHz) and S band (2.5 GHz), which allow it to be more accurate compared to GPS, which uses a single band. These satellites are situated in a much higher orbit, positioned around 36,000 kilometres from Earth; hence, the signal is less prone to obstructions. NavIC provides Standard Position Service (SPS) and Restricted Service (RS). The positional accuracy of SPS is between 10–20 metres and is for civilian use. Whereas the positional accuracy of RS, an encrypted service, is around 0.5–5 metres and is for the use of the Indian armed forces. Despite being operational since 2018, we have witnessed very limited use of NavIC. Currently, it is being used for public vehicle tracking, providing emergency warning alerts to fishermen venturing into the deep sea, and tracking and providing information related to natural disasters. NavIC compatibility is an issue Phone makers are worried that making their phones NavIC compatible would mean higher hardware and R&D costs. Besides, the regulatory testing associated with the change would also involve a greater cost and time. While initial media reports suggested that the government would mandate smartphone manufacturers to make their phones NavIC compatible by 2023, the Ministry of Electronics and IT clarified by saying, “A media report has claimed citing a meeting that mobile companies were asked to make smartphones compatible with NavIC within months. This is to clarify: (1) No timeline has been fixed. (2) The cited meeting was consultative, and (3) the issue is under discussion with all stakeholders.” Even if there is no deadline, concerns and challenges remain. NavIC uses L5 satellite frequency, whereas GPS relies on the L1 frequency. Most phones in the market today come with chips that support GPS-compatible L1 frequency. Making their phones NavIC-compatible could mean introducing a whole new chip or dual band chipsets which support both GPS and NavIC. Given the smartphone market in India is highly competitive, the changes could push the prices of smartphones in India, especially in the budget and mid-range segments. The challenges are not solely limited to hardware; many software changes are also required, which could further spike the costs incurred by the phone makers. NavIC compatible chipsets While NavIC is an issue, some chipsets in the market do support NavIC. US-based chipmaker Qualcomm is already manufacturing chipsets that support NavIC. Its processors, such as Snapdragon 720G, Snapdragon 662, Snapdragon 460 and Snapdragon 865, which support NavIC, are being used by RealMe and Xiaomi. Many phone makers also use chips made by Taiwan-based MediaTek, especially for their mid-segment devices. MediaTek is also making its 5G chips NavIC compatible. With the recent launch of 5G in India, phone makers are looking to make their 5G phones NavIC compatible. IIT-Bombay has developed DHRUVA, an indigenous NavIC-capable RF receiver, to address this issue. DHRUVA is the first indigenously designed navigation receiver chip, and it supports the L5 frequency for NavIC as well as the L1 and L2 frequency for GPS. Given the government’s push to make India self-reliant, especially in terms of technology, DHRUVA could prove to be a game changer. Besides, ISRO plans to launch a satellite in the next two to three years, which would support the L1 frequency. When it comes to smartphones, there are products in the market that are NavIC compatible. Smartphones from manufacturers such as Xiaomi, Realme, Oppo, Vivo, and OnePlus are already NavIC compatible. Most of the phones today support GPS or Russia’s GLONASS, China’s BeiDou and EU’s Galileo; however, these navigation systems have existed for a while now and NavIC is comparatively new. Hence, making all phones NavIC-compatible would require time.","excerpt":"Phone makers argue that navIC could lead to higher production costs and complicated hardware\/software changes","categories":["IT Services"],"tags":["NAVIC"],"author_name":"Pritam Bordoloi","publish_date":"2022-10-03T17:00:00","publication_year":"2022","word_count":843,"keywords":["Go","programming_languages:R","AI","NAVIC","programming_languages:Go","RAG","Aim","GAN","R"],"extracted_tech_keywords":["AI","Aim","RAG","R","Go","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/indias-indigenous-navigation-system-can-make-your-phones-expensive\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10044606,"title":"How To Process Humongous Datasets Using Vaex?","content":"Data preprocessing and Data normalization have become very important when it comes to implementing the data through different Machine Learning Algorithms. The data preprocessing phase can significantly affect the learning model’s outcome; therefore, all features present in the data must be on the same scale. The type of feature preprocessing and normalization technique required can vary on the data. Data preprocessing can also be defined as a data mining technique to turn the raw data gathered from multiple sources into a cleaner information channel that is more suitable to work upon. It is an essential preliminary step that takes all of the available information in the form of a dataset and performs various techniques to organize, sort, and merge it. Such Data Science techniques try to extract information from chunks of data to create a cleaner database from numerous datasets. At times, these databases can get incredibly massive and usually contain data of all sorts, which means that they don’t share the same structure. Raw data can have missing or inconsistent values and present us with a lot of redundant information. The very purpose of the Data Preparation phase is to convert the data and database into a format suited best for Machine Learning. Data Preparation also comprises three main phases, namely Data Cleansing, Data Transformation, And Feature Engineering. High-Quality data is more essential for working with Complex Algorithms, so it is an incredibly important phase and should not be avoided by any means. Why Data Preprocessing is so important? Through preprocessing data:We eliminate the incorrect or missing values that are there due to the human factor or bugs, making our databases more accurate.When there are inconsistencies in data or duplicates, it affects the accuracy of the results, hence removing them to make our data consistent.Filling the attributes that are missing if needed, making the data complete. What is Vaex (python library)? Vaex is a Python library for Out-of-Core DataFrames and helps to load, visualize and explore big tabular datasets. It can aid in calculating statistical operations such as mean, sum, count, standard deviation etc., on an N-dimensional grid, up to a billion rows per second. Visualisation can be created using histograms, density plots and 3d volume rendering, allowing interactive exploration of big data. Vaex uses a technique known as memory mapping, a zero memory copy policy, for best computational performance. While Pandas is largely popular for handling data in Python language, it is still eager for memory. As the data gets bigger, you will have to be careful not to have a MemoryError. Switching to a more powerful machine may solve some memory issues, but still, Pandas will only use one out of the 32 cores of your fancy machine. With Vaex, all operations are out of the core and executed in parallel and lazily evaluated, allowing for crunching through a billion-row dataset effortlessly. Vaex can be a potential solution that can resolve all the above problems while still providing a convenient API. Vaex achieves this high performance by combining memory mapping, a zero memory copy policy, and lazy computations. As discussed above, Vaex uses Memory mapping to solve this. All the dataset files read into vaex are memory-mapped. So, When you open a memory-mapped file with Vaex, you don’t read the data. Instead, Vaex swiftly reads the file metadata, providing solutions to open these files quickly, irrespective of how much RAM you have.  The format of mappable memory files is Apache Arrow, HDF5, etc. Getting Started with Code Implementation This article will try to explore how the Vaex library works and how it can make the process of data preprocessing and loading humongous datasets easier compared to traditional processing frameworks such as pandas. The following code implementation was inspired by the official documentation of Vaex, whose link can be found here. Installing the Library First, we will install the necessary libraries to create and process our model. To do so, the following code can be run, #installing the library !pip install --upgrade vaex !pip install ipython==7.25.0 Please remember, the Vaex library supports python 7.25.0 for operations; hence we are installing that as well. Importing The Library To import the Library, the following code can be used, import vaex as vx Reading the Dataset Now that we have imported the necessary libraries, we will import and load the dataset to the data frame. We will be using a dataset that is tremendously huge to get a taste of Vaex’s processing power. The dataset being used has 146 Million rows of data, with its size being over 12GBs! Here we will be comparing and visualizing the routes covered by two taxis, Taxi1 & Taxi2. All other necessary details about the New York Taxi dataset can be found through the link here. #loading into dataframe taxi1 = vx.open('s3:\/\/vaex\/taxi\/yellow_taxi_2015_f32s.hdf5?anon=true') #viewing rows and columns taxi1 Output : #view shape of df1 taxi1.shape #Loading df2 taxi2 = vx.open('s3:\/\/vaex\/taxi\/yellow_taxi_2009_2015_f32.hdf5?anon=true') #viewing rows and columns taxi2 Visualizing the Dataset Now that the data for both the Taxies has been loaded into the data frames, we can start visualizing its route travelled and covered in map. #creating visulization for Taxi1 long_min = -74.05 long_max = -73.75 lat_min = 40.58 lat_max = 40.90 taxi1.plot(taxi1.pickup_longitude, taxi1.pickup_latitude, f=\"log1p\", limits=[[-74.05, -73.75], [40.58, 40.90]], show=True); Output: #creating visulaization for Taxi2 long_min = -74.05 long_max = -73.75 lat_min = 40.58 lat_max = 40.90 taxi2.plot(taxi2.pickup_longitude, taxi2.pickup_latitude, f=\"log1p\", limits=[[-74.05, -73.75], [40.58, 40.90]], show=True) As we can observe, with Vaex, we can load and process Big Data and create powerful visualizations in a matter of seconds! We can also benchmark the time that was required to load our dataset into the dataframe, %%timeit taxi1 = vx.open('s3:\/\/vaex\/taxi\/yellow_taxi_2015_f32s.hdf5?anon=true') Output : 23.9 ms ± 4.5 ms per loop (mean ± std. dev. of 7 runs, 1 loop each) The observed time seems to be in milliseconds; this tells us how fast and efficient the Vaex library really is. EndNotes In this article, we have explored the importance of data preprocessing and explored the capabilities of a library named Vaex, which helps load heavy Big Data easily into data frames. The above implementation can be found as a Colab notebook, using the link here. References","excerpt":"Vaex is a Python library for Out-of-Core DataFrames and helps to load, visualize and explore big tabular datasets. It can aid in calculating statistical operations such as mean, sum, count, standard deviation etc., on an N-dimensional grid, up to a billion rows per second.","categories":["AI Trends"],"tags":["Big Data","Data Analytics","data preparation","data preprocessing","data visualization python"],"author_name":"Victor Dey","publish_date":"2021-07-28T14:00:00","publication_year":"2021","word_count":1024,"keywords":["data visualization python","data science","Go","machine learning","TPU","data preprocessing","AI","data preparation","Git","Colab","Python","Data Analytics","Big Data","R","Pandas"],"extracted_tech_keywords":["AI","machine learning","data science","Colab","Pandas","TPU","Python","R","Go","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/how-to-process-humongous-datasets-using-vaex\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10071193,"title":"Homegrown Label Bacca Bucci introduces AI in footwear","content":"The human foot is a masterpiece of engineering & a work of artLeonardo da Vinci Born in Delhi, Bacca Bucci (baa-ka: bu-ki) is a footwear label for people with a distinct style. They take inspiration from the global runway and the local fashion scene to create a colourfully eclectic range of footwear and accessories, allowing people to breathe new life into their wardrobes with a bold and bright design. “We have invested our effort and time in the footwear industry after going through a bunch of analytical report files and a lot of research data because footwear doesn’t have any specific season. It is like a mood, an emotion,” says Anuj Nevatia, Co-Founder & Director at Bacca Bucci. “We have always had the vision of sounding best with the best of everything we can provide. Even today, after nine years of success, the mission and vision stand tall and straight with the flags waving high.” Bacca Bucci’s ethos is centred around varierty, quality, and cost leadership. The digital-first brand strives to constantly introduce funky designs, ensure enduring fit and finish, and source material directly from the manufacturers. Combining this with a region-centric approach, they aim to position India as a global footwear leader. Though they offer shoes catering to the needs of teens and adults alike, Bacca Bucci primarily focuses on categories wherein the “youth” strike the top-most list. The motive is to cover up a larger number of audiences with the best effort, least hassles and better future opportunities. “We have something for everyone, each product being region-centric”Anuj Nevatia For instance, they have high-end fashion pop colour sneakers for urban customers, snow boots from folks Kashmir\/Northeast, and sports shoes for plus-size feet. A touch of tech With a focus on technology and quality, the company is fully aware of India’s potential to disrupt the global market for high-end footwear. The brand uses technology to eliminate size-related issues by testing various Indian foot types and sizes. It also employs data analytics to analyse customer feedback and make appropriate changes. Furthermore, by introducing new payment methods, such as BNPL (Buy-Now-Pay-Later), they are alleviating customers’ transactional concerns. Bacca Bucci is a staunch supporter of ‘Make in India.’ They believe that India is a land of opportunities when it comes to producing high-quality footwear. The brand is now working to incorporate ‘Made in India’ innovations and technology into its products to cater to today’s quintessential stylish people who are personality and trend-conscious. Enter AI Most recently, Bacca Bucci announced that it is dealing with design-related challenges with the help of AI. “Designing has benefitted from AI. This technology gives the brand insights to predict the upcoming trends, colours, choices and design. Based on the outcomes, their designers can decide the new launches,” says Nevatia. The AI system can track market structure and focus on targeting consumer needs with in-depth analytical data and records and gives us an insight report. This is handy for product design, marketing and creativity plans, and customer service. The AI system provides opportunities for augmenting human creativity. Campaigns can be designed to deliberate the purpose of the right timing, experience, and place. The best customer experience comes from amalgamating machine-driven analytics and human creativity. This results in better revenues and higher profitability. “AI has mass market potential and various segmented tools for better functioning. It drives down the time taken for a process\/task to complete. We aspire to inspire the whole process of working and making it entirely AI-driven. If it gets done, things can be more automatic, rigid, active, and processed systematically. Just like any purpose-built and lightweight automation, a complete process understanding, including process design, can be made to work,” says Nevatia. “When I think of the future, I think of growth, better opportunities, expansion of industry and market, and possibly dealing with every relatable service.” The Road ahead Over the last year, Bacca Bucci has served more than five million Gen Z and millennials. They target to close the current FY 22-23 at 100% YoY. Bacca Bucci aims to continue growing as one of India’s best and oldest D2C brands. Having registered a pan-India presence, the company is now eyeing big overseas expansion, starting with baccabucci.ae in UAE and other nations. The brand also takes pride in being environment-conscious, doing its bit to create sustainable products. It will soon be launching styles which will be made of completely recycled plastics and rubbers. Meanwhile, they have already stopped using single-use plastics for packaging and hope to be a 100 percent single-use plastic-free brand by December 2022.","excerpt":"This technology gives the brand insights to forecast the upcoming trends, colours, choices and design. Based on the outcomes, their designers are able to decide the new launches.","categories":["IT Services"],"tags":["AI Tool"],"author_name":"Sri Krishna","publish_date":"2022-07-19T14:00:00","publication_year":"2022","word_count":759,"keywords":["Go","programming_languages:R","AI","innovation","Git","automation","Aim","ViT","analytics","AI Tool","R"],"extracted_tech_keywords":["AI","analytics","Aim","R","Go","Git","ViT","automation","innovation","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/it-services\/homegrown-label-bacca-bucci-introduces-ai-in-footwear\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10090804,"title":"Machines Are Getting Better at Coding, Should You Be Worried?","content":"In the past, we have all witnessed the trials and tribulations of human coders struggling to code and get the job done without a tussle. But now, picture a world where machines—thanks to the advent of foundational models (GPTx)—are self-sufficient in mastering the art of coding, eliminating bugs, and minimising downtime. Guess what? It’s already happening. Here’s Auto-GPT, an open-source, self learning model that has the capabilities of GPT-4, which can develop and manage outputs and more. So, should you be worried? Advocating for the model, Andrej Karpathy, the former director of Tesla—who recently returned to OpenAI—believes that the “next frontier of prompt engineering are AutoGPTs”. Karpathy said so while tweeting about the latest version of Auto-GPT, which can write its own code using  GPT-4 and execute python scripts.’ (It also has a voice!) Founded by Significant Gravitas, a games development company, the Autonomous GPT-4 experiment allows itself to tamp down the bugs, develop and self-improve. The open-sourced model has managed to woo LLM enthusiasts and it is being called a direct disruptive competitor of OpenAI’s flagship model, ‘ChatGPT’.The model’s developer, Toran Bruce Richards, believes that Auto-GPT has the potential to save humanity from mass job loss caused by automation from closed-source AI. If everyone has access to their own team of autonomous agents, everyone is enabled and complete. Though the model is currently dependent on GPT3 and GPT4, the researchers are looking into implementing GPT4All. Ultimately, one won’t need to read the source code of an LLM to benefit from this. Karpathy Strikes Karpathy shared a fascinating insight on the model. He said, unlike humans, GPTs are completely unaware of their own strengths and limitations, including their finite context window and limited mental maths abilities. This can result in occasional unpredictable outcomes. However, by stringing together GPT calls in loops, agents can be created that can perceive, think, and act towards goals defined in English prompts. For feedback and learning, Karpathy suggested a “reflect” phase, where outcomes are evaluated, rollouts are saved to memory, and loaded to prompts for few-shot learning. This “meta-learning” few-shot path allows for learning on whatever can be crammed into the context window. However, the gradient-based learning path is less straightforward due to a lack of off-the-shelf APIs—such as LoRA finetunes, supervised fine tuning (SFT) and reinforcement learning from human feedback (RLHF) style—which prevent fine tuning on large amounts of experience. Karpathy believes that much like employees coalescing into organisations to specialise and parallelise work towards shared goals, AutoGPTs might evolve to become AutoOrgs with AutoCEO, AutoCFO, AutoICs, and more. Embracing AutoGPTs Within a week of its release, the Auto-GPT repository has already gained popularity with over 8,000 stars. Alongside the release, a flurry of discussions among developer communities were also sparked. While some have lauded its capabilities, others have pointed out that it still requires human intervention for debugging. One user even drew parallels between the model’s coding process and the traditional practice of rubber duck debugging. Reddit users have offered varied perspectives on the matter. Some have expressed hope that the base models will not be made available to the general public, citing concerns of potential misuse. Conversely, others have argued that not releasing it would make the AI even more dangerous. A potential downside of keeping all development behind closed doors is that the AI could be commandeered by select individuals to monitor and regulate every action of the populace. A possible solution suggested by a commentator is to make the model available to the public, accompanied by the necessary tools and resources to ensure responsible experimentation. This would allow for proactive measures to be taken by ethical researchers to counter any rogue AI scenarios that may arise. In essence, the commentator expressed the sentiment that “the only way to thwart a malicious AI is through a benevolent AI”. If you are unable to set up AutoGPT yourself, but want to try it, this is the thread for you. Post your prompts below and Toran Richards will try out some of the best ones and record the output for you!","excerpt":"The creator of Auto-GPT, Toran Bruce Richards, believes it has the potential to save humanity from mass job loss caused by automation from closed-source AI.","categories":["AI Features"],"tags":["Andrej Karpathy","AutoGPT","ChatGPT","GPT-4","GPT4","OpenAI","OPenAI GPT"],"author_name":"Tasmia Ansari","publish_date":"2023-04-05T13:00:00","publication_year":"2023","word_count":678,"keywords":["ChatGPT","TPU","Andrej Karpathy","OpenAI","AutoGPT","AI","RLHF","autonomous agents","GPT-4","Python","prompt engineering","OPenAI GPT","few-shot learning","GPT4"],"extracted_tech_keywords":["AI","ChatGPT","OpenAI","AutoGPT","prompt engineering","few-shot learning","RLHF","autonomous agents","TPU","Python"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/machines-are-getting-better-at-coding-should-you-be-worried\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10023809,"title":"Getting Started With OpenCV In Python","content":"OpenCV is a powerful and versatile open-source library for Computer Vision tasks. It is supported in many languages, including Python, Java and C++. It is packed with more than 2500 algorithms to perform almost any Computer Vision task with just a single library. OpenCV is well known for its interactive windows and real-time processing capabilities. In this quick tutorial, basic image processing and video processing with this library are discussed using Python codes. It should be noted that Jupyter Notebook environments such as Colab and JupyterLab clash with some interactive methods of OpenCV. However, these issues can be easily overcome with the help of some external libraries. Nevertheless, the codes discussed in this tutorial are carried out in a Spyder environment for the best performance. Install the library in your local machine to have fun with images and videos. Install OpenCV Python from the PyPi package. !pip install opencv-python Read and Write an Image Before starting the image processing, make sure that the working directory has an image. In OpenCV, an image can be read as a color or a grayscale image with corresponding flags. Read a coloured image with the imread method as shown below: import cv2 # read an image in colour mode img = cv2.imread('daria.jpg', 1) Argument 1 in the above imread method is a flag that directs that the image has to be read in colour. Flag 0 refers to grayscale image reading. The read image can be displayed in a window using imshow method. # display the image in a window cv2.imshow('Image Window', img) cv2.waitKey(0) cv2.destroyAllWindows() OpenCV allows performing multiple inputs and outputs simultaneously through multiple windows. The waitKey method is used to inform OpenCV the time duration over which a window can be kept open. This method takes time as an argument in milliseconds. If 0 is provided as the argument, the user should close the window manually. It can be noted that the window has a name of its own. Therefore, when there comes a window into play, it must be named. Here, the window name is ‘Image Window’. destroyAllWindows method is used to force all the open windows to close at once. To read the above image as a grayscale image, # read an image in grayscale mode img = cv2.imread('daria.jpg', 0) # display the image cv2.imshow('Image Window', img) cv2.waitKey(0) cv2.destroyAllWindows() Write an image to the local directory using the imwrite method. The method takes two arguments: name of image about to be written and the read\/processed image. # write the image cv2.imwrite('daria_gray.jpg', img) Read and Write a Video A video is a collection of frames displaying at speed termed frames per second (fps). Each frame is an image. Excluding the specific properties such as frames per second and codec format, OpenCV processes videos the same way it processes images. OpenCV reads a video either from a file or directly from the device’s camera. This feature makes OpenCV the de facto choice for navigation systems such as robots and drones, embedded systems such as Raspberry Pi and Arduino, and autonomous vehicles. # read a video from file capture = cv2.VideoCapture('swan.mp4') # display the read video file while capture.isOpened(): ret, frame = capture.read() if not ret: break cv2.imshow('Video Window’, frame) cv2.waitKey(25) capture.release() cv2.destroyAllWindows() A frame in video output: Video plays until there is a frame to read in the file. If we wish to read the device’s camera, the file name in the VideoCapture method should be replaced with number 0 (zero). In that case, the user needs to enable keyboard shortcut to stop capturing. The following example shows camera capturing along with keyboard shortcut of letter ‘q’ to interrupt capturing. # read real-time video from device’s camera capture = cv2.VideoCapture(0) # display the capturing while capture.isOpened(): ret, frame = capture.read() cv2.imshow('Video Window’, frame) if cv2.waitKey(20) & 0xFF == ord(‘q’): break capture.release() cv2.destroyAllWindows() Writing a video into a file needs some properties such as frame width, frame height and frame rate (fps). The following codes enable us to learn the input video properties. capture = cv2.VideoCapture('swan.mp4') # get frame properties print(capture.get(cv2.CAP_PROP_FRAME_WIDTH)) print(capture.get(cv2.CAP_PROP_FRAME_HEIGHT)) print(capture.get(cv2.CAP_PROP_FRAME_COUNT)) Output: Write the video file to disk in any format you wish. It is required that the right fourcc codec format, frame size and frame rate should be provided as arguments to enable proper saving of the video file. Here, in the below example code, the file is saved in the working directory in the name of ‘output.mp4’. # read a video and write it to another file capture = cv2.VideoCapture('swan.mp4') # create a write file - arguments: file_name, fourcc_code, fps, size out = cv2.VideoWriter('output.mp4', cv2.VideoWriter_fourcc(*'XVID’'), 20.0, (640,360)) while capture.isOpened(): # read frame by frame ret, frame = capture.read() out.write(frame) if not ret: break cv2.imshow('Video Window', frame) cv2.waitKey(25) capture.release() out.release() cv2.destroyAllWindows() Draw Shapes in an Image In OpenCV, shapes such as a line, arrowed line, circle, rectangle, ellipse and polygons can be drawn over an image. Start and end coordinates, color of the shape, thickness of the border line are the common parameters in drawing a shape. It should be noted that OpenCV supports colours in BGR format, in contrast to most libraries such as Matplotlib and Seaborn where they support colours in RGB format. # read a coloured image img = cv2.imread('mason.jpg', 1) print(img.shape) Output: Draw a vertical green coloured line on the image # arguments: image, start, end, colour, thickness img = cv2.line(img, (50,100), (50,300), (0,255,0), 10) # display the image cv2.imshow('Image Window', img) cv2.waitKey(0) cv2.destroyAllWindows() Draw a blue coloured circle and a red coloured rectangle on top of it. img = cv2.imread('mason.jpg', 1) # draw a green vertical line in it # arguments: image, start, end, colour, thickness img = cv2.line(img, (50,100), (50,300), (0,255,0), 5) # draw a blue circle on it # arguments: image, centre, radius, colour, thickness img = cv2.circle(img, (150,250), 60, (255,0,0), 5) # draw a red rectangle on it # arguments: image, diagonal_start, diagonal_end, colour, thickness img = cv2.rectangle(img, (300,140), (400,270), (0,0,255), 5) # display the image cv2.imshow('Image Window', img) cv2.waitKey(0) cv2.destroyAllWindows() Write Text on an image Text can be written on an image. Its location, size, font and colour can be customized as per the user’s wish. img = cv2.imread('senjuti.jpg') # display the  Original image without Text cv2.imshow('Original Image', img) # add text on the image # arguments: image, text, start_location, font, font_size, colour, thickness text_image = cv2.putText(img, 'I love Colours', (40,50), cv2.FONT_HERSHEY_SIMPLEX, 1, (0,0,255),2) # display the image with Text in a new Window cv2.imshow('Image with Text', text_image) cv2.waitKey(0) cv2.destroyAllWindows() Running date-time on videos can be made in real-time using Python’s datetime library and OpenCV’s putText method. The below code example shows real-time date-time display on a video. It is especially useful in real-time camera capturing. # import the library from datetime import datetime text = str(datetime.now()) # read a video from file capture = cv2.VideoCapture('drive_6.mp4') # display the read video file while capture.isOpened(): ret, frame = capture.read() # add date-time to the frames frame = cv2.putText(frame, text, (30,40), cv2.FONT_HERSHEY_PLAIN, 1, (0,0,255), 2) if not ret: break cv2.imshow('Video Window', frame) cv2.waitKey(30) capture.release() cv2.destroyAllWindows() Colour Analysis on an Image We discussed that colour images in OpenCV are read and processed in BGR colour format. An image can be split into three separate images for each of the Blue, Green and Red channels. On the other hand, the split image can be back to original colour or different channel combinations as shown in the example below. # read an image img = cv2.imread('daria.jpg', 1) # split the colour image and merge back B, G, R = cv2.split(img) img_BGR = cv2.merge((B,G,R)) img_RGB = cv2.merge((R,G,B)) img_BRG = cv2.merge((B,R,G)) # display the merged images cv2.imshow('Image in BGR', img_BGR) cv2.imshow('Image in RGB', img_RGB) cv2.imshow('Image in BRG', img_BRG) cv2.waitKey(0) cv2.destroyAllWindows() Alterations in Images A portion of an image can be extracted or replaced with a similar-sized image patch or any simple math alterations similar to that. Here, we perform some replacements in an image with its own sub-portion in an example. # coffee cup alteration img = cv2.imread('brooke.jpg') # extract a portion of the image coffee = img[150:235,200:300] alter = img.copy() # perform multiple alterations alter[0:85,0:100] = coffee alter[85:170,0:100] = coffee alter[170:255,0:100] = coffee alter[20:105,220:320] = coffee # display the Original image cv2.imshow('Original Image', img) # display the altered image cv2.imshow('Altered Image', alter) cv2.waitKey(0) cv2.destroyAllWindows() Merging Multiple Images Two or more images can be merged either by simple addition or by weighted addition. However, adding together images should be the same size and have the same number of channels. # read two images and resize them img_1 = cv2.imread('xuan.jpg', 1) img_1 = cv2.resize(img_1, (320,225)) img_2 = cv2.imread('daria.jpg',1) img_2 = cv2.resize(img_2, (320,225)) # display original images cv2.imshow('Image 1', img_1) cv2.imshow('Image 2', img_2) cv2.waitKey(0) cv2.destroyAllWindows() # simple addition simple = cv2.add(img_1, img_2) cv2.imshow('Simple Addition', simple) cv2.waitKey(0) cv2.destroyAllWindows() # weighted addition weight_70 = cv2.addWeighted(img_1, 0.7, img_2, 0.3, 0) weight_30 = cv2.addWeighted(img_1, 0.3, img_2, 0.7, 0) cv2.imshow('70-30 Addition', weight_70) cv2.imshow('30-70 Addition', weight_30) cv2.waitKey(0) cv2.destroyAllWindows() Wrapping Up In this tutorial, we discussed the OpenCV library and its fundamental implementation in Python. We discussed loading an image or a video file and saving them to the disk. Further, we drew some standard shapes such as line, circle and rectangle over an image. We looked at writing texts, including real-time running texts of date and time. We learned how OpenCV handles colour channels (B,G,R) and split and merged in different colour combinations. Finally, we discussed alterations in an image and merging different images to create a new interactive image. Note: Images and videos used are open source images that need no license to reuse. References: OpenCV 4.5.4 Released, Look For Updated Features And Fixes10 OpenCV Projects To Check Out In 2020Top 8 Resources To Learn OpenCV For BeginnersWhy Is OpenCV Gaining Prominence?Linux Vs Windows: Which Is The Best OS For Data Scientists?","excerpt":"OpenCV is a powerful, versatile open-source library for Computer Vision tasks with interactive windows and real-time processing capabilities","categories":["AI Trends"],"tags":["image processing","OpenCV","Opencv image processing","OpenCV tutorial","Python","spyder","Video"],"author_name":"Rajkumar Lakshmanamoorthy","publish_date":"2021-04-13T18:00:00","publication_year":"2021","word_count":1627,"keywords":["Opencv image processing","TPU","AI","image processing","spyder","computer vision","OpenCV","Python","Ray","Colab","Matplotlib","Seaborn","Video","Jupyter","OpenCV tutorial"],"extracted_tech_keywords":["AI","computer vision","Ray","Jupyter","Colab","OpenCV","Matplotlib","Seaborn","TPU","Python"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/getting-started-with-opencv-in-python\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10133979,"title":"California&#8217;s SB 1047 Bill Sparks AI Civil War","content":"On Wednesday, California lawmakers approved a contentious AI safety bill, which now requires a final procedural vote. Afterward, the decision will rest with Governor Gavin Newsom, who has until September 30 to either sign the bill into law or veto it. SB 1047 — our AI safety bill — just passed off the Assembly floor. I’m proud of the diverse coalition behind this bill — a coalition that deeply believes in both innovation & safety.AI has so much promise to make the world a better place. It’s exciting.Thank you, colleagues.— Senator Scott Wiener (@Scott_Wiener) August 28, 2024 California’s Senate Bill 1047, a proposed AI regulation, has sparked intense debate in Silicon Valley, drawing both praise and criticism from tech leaders, lawmakers, and AI experts. xAI chief Elon Musk has also voiced support, emphasising the need for regulation to prevent AI misuse. “This is a tough call and will make some people upset, but, all things considered, I think California should probably pass the SB 1047 AI safety bill,” he posted on X. Musk said that for more than 20 years, he has supported AI regulation, drawing parallels to how society regulates any technology or product that could potentially pose a risk to the public. People have spent years calling Elon Musk the real-life Iron Man, but now that he's backing AI regulations, just like Iron Man did with the Sokovia Accords in the Civil War movie, people are freaking out. pic.twitter.com\/8PiNo9wIaq— Mukul. (@pathuglife) August 28, 2024 The bill, introduced by State Senator Scott Wiener, seeks to implement strict safety measures for large-scale AI models, aiming to prevent potential catastrophes, but critics argue it could hinder innovation. The legislation requires developers of significant AI models—those costing over $100 million to train—to perform comprehensive safety testing before releasing them to the public. It also mandates an “emergency stop” feature to shut down AI systems in critical situations and obligates developers to report any safety incidents to California’s Attorney General within 72 hours. A new state agency, the Frontier Model Division, would oversee compliance, with penalties of up to $30 million for repeated violations. Proponents, including AI luminaries Geoffrey Hinton and Yoshua Bengio, believe the bill is crucial to addressing AI risks comparable to pandemics or nuclear threats, potentially setting a national standard for AI safety. Conversely, the bill has faced strong opposition from major tech companies like Google, OpenAI, and Meta, who argue it could stifle innovation and drive talent out of California. Critics, such as U.S. Representative Nancy Pelosi and AI expert Fei-Fei Li, caution that the bill’s requirements could disproportionately burden smaller companies and slow technological advancement. They advocate for federal regulation to avoid inconsistencies across states. OpenAI has publicly opposed SB 1047, stating that it poses a threat to AI’s growth and could push entrepreneurs and engineers to relocate. The bill has passed the California Appropriations Committee with amendments and is awaiting a final vote in the state assembly. If signed into law by Governor Gavin Newsom, it would be the first AI regulation of its kind in the U.S., potentially shaping AI governance nationwide. As California, a key hub for AI innovation, seeks to balance technological advancement with safety, the fate of SB 1047 could have significant implications for the global tech industry and the future of AI regulation.","excerpt":"California’s Senate Bill 1047, a proposed AI regulation, has sparked intense debate in Silicon Valley, drawing both praise and criticism from tech leaders, lawmakers, and AI experts.","categories":["AI News"],"tags":[],"author_name":"Siddharth Jindal","publish_date":"2024-08-29T10:42:32","publication_year":"2024","word_count":551,"keywords":["Go","API","OpenAI","AI","XAI","Aim","xAI","AI governance","AI safety","R"],"extracted_tech_keywords":["AI","OpenAI","xAI","Aim","R","Go","API","AI safety","XAI","AI governance"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/californias-sb-1047-bill-sparks-ai-civil-war\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10114928,"title":"&#8216;AI Advisory Will Not Apply to Startups,&#8217; Clarifies Rajeev Chandrasekhar","content":"In a recent advisory issued by the Ministry of Electronics and Information Technology (MeitY), significant platforms have been singled out, requiring them to seek permission before deploying untested AI models on the Indian internet. Notably, startups are exempted from these stringent regulatory measures. “The advisory is specifically aimed at significant platforms. Seeking permission from MeitY is only mandatory for large platforms, and it will not apply to startups,” said Indian IT Minister Rajeev Chandrasekhar on X. “The advisory aims to prevent untested AI platforms from deploying on the Indian Internet,” added Chandrasekhar. https:\/\/twitter.com\/Rajeev_GoI\/status\/1764534565715300592?t=ieGSo7dULUDURmWfaXKuOA&s=15 The advisory emphasises that obtaining permission, labeling, and disclosing information to users about untested platforms act as protective measures for companies. This process serves as an insurance policy, shielding these platforms from potential consumer lawsuits that might arise in the absence of such transparency measures. “The process of seeking permission, labeling, and consent-based disclosure to users about untested platforms is an insurance policy for platforms that could otherwise face lawsuits from consumers,” read Chandrasekhar’s post on X. The primary focus of the advisory is to prevent untested AI platforms from entering and operating within the Indian internet landscape. This clarification comes after the recent advisory issued by MeitY, stating that all AI models, large-language models (LLMs), software using generative AI, or any algorithms currently in the beta stage of development or deemed unreliable in any form must seek the ‘explicit permission of the Government of India’ before being deployed for users on the Indian internet.","excerpt":"“The advisory aims to prevent untested AI platforms from deploying on the Indian Internet,” added Chandrasekhar.","categories":["Deep Tech"],"tags":["Rajeev Chandrasekhar","Startups"],"author_name":"Siddharth Jindal","publish_date":"2024-03-04T12:52:58","publication_year":"2024","word_count":248,"keywords":["Rajeev Chandrasekhar","Go","AWS","AI","cloud_platforms:AWS","programming_languages:R","programming_languages:Go","Aim","generative AI","Startups","R","startup"],"extracted_tech_keywords":["AI","generative AI","Aim","AWS","R","Go","startup","cloud_platforms:AWS","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/it-will-not-apply-to-startups-clarifies-rajeev-chandrasekhar-on-indian-govts-ai-advisory\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10112350,"title":"Where Would Meta Be without Open Source AI?","content":"It’s hard to imagine Meta without open source AI, given its significant contribution to the ecosystem. Fueling several innovations, it has also in many ways been helping the likes of OpenAI, which have been syphoning off of open source contributions without batting an eyelid. “OpenAI does not have a monopoly on good ideas. They’re not going to get to AGI by themselves, in-fact they’re using PyTorch and Transformers, which were published by many of us. They’re profiting from the open research landscape,” said Yann LeCun, not mincing words during a roundtable discussion at Davos. Transformers was first proposed by Google in 2017 in its paper ‘Attention is all you Need’. Recently, OpenAI released new embedding models and API updates created by Indian developers Aditya Kusupati, a researcher at Google, and Prateek Jain, a senior staff research scientist at Google, two years ago. Meta is shamelessly open source Till date, Meta has over 600 open source projects and counting. The company’s  influence in the AI landscape is essentially in diverting the larger community from having to pay its competitors. Further, the collaboration with the ecosystem is necessary to set the industry standards in finding use cases with the AI models. Clem Delangue, founder of Hugging Face, posted on X that Meta has the highest number of open source models on the platform with 689 models, a number which has grown since then. In this week’s earnings call Zuckerberg addressed how open sourcing benefits Meta, saying, “The short version is that open sourcing improves our models, and because there’s still significant work to turn our models into products, we find that there are mostly advantages to being the open source leader.” Meta is “playing to win”, added Zuckerberg, pointing out that training and operating future models will be even more compute intensive. Its aspiration of building a full general intelligence would require many years of dedicated research and development. Zuckerberg, in a conversation with Lex Fridman, said that Meta’s bet is on the open source community as it believes instead of hopping onto the wagon of claiming super intelligent AI, it would be beneficial to let the community use it for research purposes, and building more efficient and aligned AI. Meta is clearly the winner of the open source AI race, and will continue to be as long as it continues its ways. Llama 2 currently has around 4 million downloads on Hugging Face, and the number is only going to increase when Llama 3 comes out soon. Open source AGI? After feeding off of the open source contributions, OpenAI had raised the alarm about open source and its risks. In August, Jan Leike, ML researcher and alignment team lead at OpenAI, painted a doomsday picture, saying, “An important test for humanity will be whether we can collectively decide not to open source LLMs that can reliably survive and spread on their own.” Although the initial reason Zuckerberg gave for their decision to open source its models was, “It drives innovation because it enables many more developers to build with new technology,” it is clear that Meta is benefiting in more than just goodwill from the developer community. “Zuck and LeCun will go down as heroes in human history! Fighting for ‘Open AI’ when the incumbents sought to shut it down. Unbelievable how much the vibes from Meta have changed over the last year. That, and maybe it’s time for a name change – Meta to OpenAI,” wrote Bindu Reddy, chief at Abacus.ai, on X. Likewise, Perplexity chief Arvind Srinivas said, ”Open source AGI is an amazing vision. You (Meta) are building a very powerful technology, and actually aligning to what makes sense for the world: more people have a say in what makes sense and doesn’t.” Not for long Meta now faces stiff competition from Chinese companies like Alibaba and Tencent that are continuously releasing open source models. Alibaba released Qwen-7B and Qwen-7B-Chat, each with 7 billion parameters, becoming the first major Chinese tech company to open-source LLMs. These models, part of the Tongyi Qianwen series, aim to help businesses adopt AI. DeepSeek, a 67 billion parameter model outperformed Llama 2, Claude-2, and Grok-1 on various metrics. The best part is that the model from China is open sourced, and uses the same architecture as LLaMA. Another hint of China’s open source AI dominance is the Yi-34B model released by 01.AI startup, reaching a unicorn status after the release. The AI startup by Kai-Fu Lee is developing AI systems for the Chinese market. The interesting part is that the second and third models on the Open LLM Leaderboard are also based on Yi-34B. As long as Chinese giants open source their models, they are not a threat to anyone, except Meta, and its AGI goals. On the other hand, Meta can also benefit from the success of Chinese open source models, just like it did with TikTok and made Reels on Instagram.","excerpt":"Nowhere close to where OpenAI is, but Meta would have certainly cracked human-level intelligence before them.","categories":["AI Features"],"tags":["Open Source AI"],"author_name":"K L Krithika","publish_date":"2024-02-09T14:02:20","publication_year":"2024","word_count":821,"keywords":["Go","Hugging Face","API","OpenAI","AI","PyTorch","ML","Transformers","Open Source AI","Aim","R"],"extracted_tech_keywords":["AI","ML","OpenAI","Aim","PyTorch","Hugging Face","Transformers","R","Go","API"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/where-would-meta-be-without-open-source-ai\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":41466,"title":"11 Non-Technical Books All Data Scientists Must Read","content":"The amount of data that gets generated on a daily basis is increasing each passing day. That is why companies across the globe are hiring data scientists to turn data into insightful and actionable information. Now that everyone knows that the data scientists role is the sexiest job of the 21st century, numerous young professionals are applying for these jobs. But what should a young professional do to make sure that he\/she stands out from the crowd? Apart from a sparkling resume and tailor-made skill-set, is there anything else that a data scientist should inculcate in his\/her personality In this article, Analytics India Magazine lists down top 11 non-technical books that every single data scientists should inculcate and internalise to stand out from the crowd. There are no cash awards, certificates or LinkedIn recommendations for reading these books — but the life lessons and work ethics that the reader garners from them are priceless. 1. Thinking, Fast And Slow by Daniel Kahneman (Penguin, 2012 | ₹276) This international bestseller authored by noted economist and psychologist Daniel Kahneman takes the readers on a fascinating journey by dissecting the mind and goes onto explain two distinct systems that affect our way of thinking and making choices. Of these two systems, one is intuitive, emotional yet fast while the other one is more logical and deliberative. As every data scientist has to have the qualities of a storyteller as well as a decision-maker, this book is a perfect read for him\/her. 2. How Not to be Wrong by Jordan Ellenberg (Penguin, 2014 | ₹399) How Not to Be Wrong: The Power of Mathematical Thinking showcases the surprising revelations behind questions like, “What does public opinion really represent,” or “How likely are you, really, to develop cancer?” The book tries to answer such questions using the mathematician’s method of analysing life and exposing the hard-won insights of the academic community to the layman — minus the jargon. Ellenberg chases mathematical threads through a vast range of time and space, from the everyday to the cosmic, encountering, among other things, baseball, Reaganomics, daring lottery schemes, Voltaire, the replicability crisis in psychology, Italian Renaissance painting, artificial languages, the development of non-Euclidean geometry, the coming obesity apocalypse, Antonin Scalia’s views on crime and punishment, the psychology of slime molds, what Facebook can and can’t figure out about you, and the existence of God. 3. Understanding Comics by Scott McCloud (William Morrow, 1994 | ₹1,358) Michael Hochster, the director of Data Science at Stitch Fix, says, “I recommend this book because a lot of what data scientists do is to communicate using a combination of images and words, just like in comics. The author is both a skilled craftsman and an excellent explainer. The book is full of thoughtful discussion and examples. I found it far more useful and stimulating than most of the standard literature on data visualisation.” 4. Creativity Inc by Ed Catmull (Bantam Press, 2014 | ₹396) This book is a great read for managers who want to lead their employees to new creative heights. It is termed as a manual for anyone who strives for originality, and the first-ever, all-access trip into the nerve centre of Pixar Animation Studios ― into the story meetings, the postmortems, and the ‘Braintrust’ sessions where art is born. It is, at heart, a book about how to build and sustain a creative culture―but it is also, as Pixar co-founder and president Ed Catmull writes, ‘an expression of the ideas that I believe make the best in us possible.’ 5. The Black Swan: The Impact of the Highly Improbable by Nassim Nicholas Taleb (Penguin, 2008 | ₹381) This book focuses on Taleb’s Black Swan Theory. Through a number of examples, the author aims to show his readers how rare and unpredictable events have a deep and lasting impact on a person’s life. Human beings tend to try and rationalise such happenings almost immediately after they happen. This is an almost impossible task since such events are based on chance. A number of other topics, which relate to aesthetics, ways of life, knowledge and much more, have been discussed in this book. 6. Stress-Proof by Mithu Storoni (Penguin, 2017 | ₹419) Translating complex scientific findings into straightforward and actionable advice, this book moves forward a working professional’s understanding and wellness in a meaningful way. Living the hectic lives the data scientists live, this book, where each chapter examines a stress agent we all face — including sugar, inflammation, an out-of-sync body clock, cortisol and emotional triggers — presents simple ways to block it. It gives solutions in the form of everyday changes in diet, lifestyle, behaviour and exercise. It also includes advice and surprising strategies using music, eye movements, body temperature and more. 7. The Emperor’s New Mind by Roger Penrose (Oxford University Press, 2016 | ₹580) Can a computer eventually do everything a human mind can do? Wolf Prize-winning physicist Roger Penrose talks about his view that there are some facets of human thinking that can never be emulated by a machine. In this book, Penrose examines what physics and mathematics can tell us about how the mind works, what they can’t, and what we need to know to understand the physical processes of consciousness. 8. The Life-Changing Magic of Tidying Up by Marie Kondo (Ten Speed Press, 2014 | ₹963) Data scientists can sometimes be so overwhelmed by their work pressures that it may sometimes become difficult to prioritise. In this not-so-directly-related book, data scientists can find solace in simple, easy-to-adopt methods of organisations and decluttering. Because it is a truth universally acknowledged that a clean, tidy workplace and home are the perfect breeding ground for concentration and creativity. In this best-seller, Japanese cleaning consultant Marie Kondo takes tidying to a whole new level, promising that if you properly simplify and organise your home once, you’ll never have to do it again. Most methods advocate a room-by-room or little-by-little approach, which doom you to pick away at your piles of stuff forever. The KonMari Method, with its revolutionary category-by-category system, leads to lasting results. 9. Deep Work: Rules for Focused Success in a Distracted World by Cal Newport (Little Brown Book Group, 2016 | ₹278) In his new book, Cal Newport talks about how professionals of today have started valuing quantity over quality; and how this has turned young professionals of today into puppets who try to indulge in extensive multitasking, dealing with multiple emails and projects. This prevents them from doing ‘deep work’; which is focused work free from all other distractions. This also means that the professionals of today should sort out their priorities. 10. Tools of Titans: The Tactics, Routines and Habits of Billionaires, Icons and World-Class Performers by Timothy Ferriss (Random House, 2016 | ₹487) During his podcast The Tim Ferriss Show, Ferriss got to interview a slew of celebrities and successful personalities. He says that this book contains the distilled tools, tactics, and ‘inside baseball’ you won’t find anywhere else. The book contains answers to questions like, “What do these people do in the first sixty minutes of each morning? What do their workout routines look like, and why? What books have they gifted most to other people? What are the biggest wastes of time for novices in their field? What supplements do they take on a daily basis?” Data scientists as a creative breed can learn much from these successful people and apply the knowledge in their daily lives. 11. Adventures of Sherlock Holmes by Sir Arthur Conan Doyle (Penguin | ₹230) A classic fiction tome might seem like an unlikely fit in this list, but let us not forget that it was this legendary character who most showed trust in the power of data. The Baker Street detective was most loved for his deductive powers, but when he explained the process behind his conclusions, it was very clear that he relied on information, observation and data.","excerpt":"The amount of data that gets generated on a daily basis is increasing each passing day. That is why companies across the globe are hiring data scientists to turn data into insightful and actionable information. Now that everyone knows that the data scientists role is the sexiest job of the 21st century, numerous young professionals […]","categories":["AI Features"],"tags":["big data and analytics everyday life","Data Science","maths"],"author_name":"Prajakta Hebbar","publish_date":"2019-06-28T08:13:22","publication_year":"2019","word_count":1325,"keywords":["data science","Go","programming_languages:R","AI","maths","Aim","ViT","analytics","Rust","GAN","big data and analytics everyday life","Data Science","R"],"extracted_tech_keywords":["AI","data science","analytics","Aim","R","Go","Rust","GAN","ViT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/11-non-technical-books-all-data-scientists-must-read\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10052101,"title":"8 Resources For Data Scientists To Master SQL","content":"SQL, or Structured Query Language, is domain-specific and was designed to manage data held in a relational database management system. As a part of the Query language, SQL was designed by Donald D, Chamberlin and Raymond F Boyce and first appeared in 1974. Developed with the goal of enabling non-programmers to excel at relational database management systems, SQL is regarded as a favourable language for database management and is supported by database systems like MySQL, SQL Server, and Oracle. Today, we list some resources (courses and books) that can help data scientists master SQL. If you are just beginning your SQL journey, read these tips to begin and this beginner’s guide first. Courses Master SQL: Udemy Delivered by instructor and software engineer Imtiaz Ahmad, this Udemy course on Master SQL for Data Science focuses on database basics, SQL Query Basics, using functions and subqueries, grouping data and computing aggregates, using CASE Clause, Window functions for analytics, working with multiple tables and advanced query techniques using correlated subqueries. On completing this course, one will be able to write complex SQL statements to query the database and gain critical insight into data. It is divided into 51 lectures across 11 sections and takes around 10 hours to complete. Introduction to SQL: DataCamp DataCamp’s Introduction to SQL is the first course under the SQL Fundamentals module. Through this hands-on course, one will be able to master the basics of querying tables in databases like MySQL, SQL Server and PostgreSQL. The instructor for this course is Nick Carchedi, Product Manager at DataCamp. The course focuses on the basics such as selecting columns, filtering rows, aggregating functions, sorting and grouping. The course usually takes up to four hours to complete. SQL for Data Science: Edx Offered by Rav Ahuja, AI and Data Science Program Director at IBM, this introductory-level course– SQL for Data Science, will help one learn using and applying SQL to better communicate and extract detailed information from databases. This course covers the foundational knowledge of SQL, creating a database in the cloud, using string patterns and ranges to query data, analysing data using Python after sorting and grouping the results by type. This self-paced course is available for free and takes approximately four weeks to complete, provided one study for two to four hours every week. Advanced SQL for Data Scientists: LinkedIn Learning Offered by LinkedIn Learning, the Advanced SQL for Data Scientists course provides a sophisticated approach to optimising queries in SQL and designing data models. The course instructor is Enterprise Architect and Big Data expert Dan Sullivan and focuses on very large databases. The course is broadly divided into– data modelling: tables and indexes, query optimisation, user-defined functions and special-purpose functionality. On completion of this advanced course, which usually takes 2.5 hours to complete, students will receive a certificate. Books SQL for Smarties Joe Celko’s SQL for Smarties: Advanced SQL Programming is considered the first-ever book to be explicitly dedicated to introducing an SQL programmer to advanced techniques and mastering the programming language. The book does not just provide tips and techniques but also the best solutions to challenges one might encounter while trying to master SQL. SQL for Data Scientists: A Beginner’s Guide for Building Datasets for Analysis Written by Renee M Teate, SQL for Data Scientists: A Beginner’s Guide for Building Datasets for Analysis works as a stepping stone to SQL and dataset designing skills. The book focuses on constructing datasets for exploration, machine learning and analysis. On completion of this book, one will be well equipped to understand relational database structures, SQL syntax and query design; review approaches and strategies to help design analytical datasets, and develop queries to construct datasets that can be used for ML algorithms and interactive reports. SQL Cookbook: Query Solutions and Techniques for Database Developers Written by expert SQL developer Anthony Molinaro, SQL Cookbook: Query Solutions and Techniques For Database Developers teaches about Window functions; features such as SQL Server’s PIVOT and UNPIVOT operators, PostgreSQL’s GENERATE_SERIES function, and Oracle’s MODEL clause; bucketization; creating histograms and the method of ‘walking a string.’ This book claims to help its readers take their SQL skills to the ‘next level.’ SQL Performance Explained Written by Markus Winand, SQL Performance Explained: Everything Developers Need To Know About SQL Performance, as the name suggests, covers all the major concepts and aspects of SQL databases without getting too focused on and lost in one single product. The book covers the usage of multi-column indexes, optimising join operations, using LIKE queries correctly, improving performance using clustering, and understanding the scalability of databases, among others. It covers all the major SQL databases such as Oracle Database, MySQL, SQL Server and PostgreSQL. Once you are familiar with the topic, here’s how you can crack SQL interviews.","excerpt":"SQL is considered one of the most favourable languages for database management.","categories":["AI Trends"],"tags":["AWS","Data Science","Data Scientist","SQL"],"author_name":"Debolina Biswas","publish_date":"2021-10-22T11:00:00","publication_year":"2021","word_count":794,"keywords":["PostgreSQL","data science","machine learning","AWS","AI","ML","Python","Ray","Aim","analytics","SQL","Data Science","Data Scientist","R"],"extracted_tech_keywords":["AI","machine learning","ML","data science","analytics","Aim","Ray","PostgreSQL","Python","R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/8-resources-for-data-scientists-to-master-sql\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":39682,"title":"Top 10 Python NLP Libraries For 2019","content":"With the help of Natural Language Processing, an organisation can gain valuable insights, patterns, and solutions. Python is one of the widely used languages and it is implemented in almost all fields and domains. In this article, we list down 10 important Python Natural Language Processing Language libraries. 1|  Natural Language Toolkit (NLTK) NLTK is a leading platform for building Python programs to work with human language data. It provides easy-to-use interfaces to over 50 corpora and lexical resources such as WordNet, along with a suite of text processing libraries for classification, tokenization, stemming, etc. This library provides a practical introduction to programming for language processing. NLTK has been called “a wonderful tool for teaching and working in computational linguistics using Python,” and “an amazing library to play with natural language.” Click here. 2| Gensim Gensim is a Python library for topic modelling, document indexing and similarity retrieval with large corpora. Target audience is basically the natural language processing (NLP) and information retrieval (IR) community. The features of this library include such as all algorithms are memory-independent w.r.t. the corpus size, intuitive interfaces, efficient multicore implementations of popular algorithms, distributed computing, etc. Click here. 3| polyglot Polyglot is a natural language pipeline which supports massive multilingual applications. The features include tokenisation, language detection, named entity recognition, part of speech tagging, sentiment analysis, word embeddings, etc. Polyglot depends on Numpy and libicu-dev, on Ubuntu\/Debian Linux distribution you can install such packages by executing the following command: sudo apt-get install python-numpy libicu-dev Click here. 4| TextBlob TextBlob is a Python (2 and 3) library for processing textual data. It provides a simple API for diving into common natural language processing (NLP) tasks such as part-of-speech tagging, noun phrase extraction, sentiment analysis, classification, translation, WordNet integration, parsing, word inflection, adds new models or languages through extensions, and more. Click here. 5| CoreNLP Stanford CoreNLP provides a set of human language technology tools. Stanford CoreNLP’s goal is to make it very easy to apply a bunch of linguistic analysis tools to a piece of text. Stanford CoreNLP integrates many of Stanford’s NLP tools, including the part-of-speech (POS) tagger, the named entity recognizer (NER), the parser, the coreference resolution system, sentiment analysis, bootstrapped pattern learning, and the open information extraction tools. The tools variously use rule-based, probabilistic machine learning, and deep learning components. Click here. 6| spaCy spaCy is a library for advanced Natural Language Processing in Python and Cython which comes with a number of interesting features. spaCy comes with pre-trained statistical models and word vectors, and currently supports tokenization for 49+ languages. It features state-of-the-art speed, convolutional neural network models for tagging, parsing and named entity recognition and easy deep learning integration. Click here. 7| Pattern Pattern is a web mining module for the Python programming language. It has tools for data mining (Google, Twitter, and Wikipedia API, a web crawler, an HTML DOM parser), natural language processing (part-of-speech taggers, n-gram search, sentiment analysis, WordNet), machine learning (vector space model, clustering, SVM), network analysis by graph centrality and visualization. Pattern supports Python 2.7 and Python 3.6. Click here. 8| Vocabulary Vocabulary is a Python library for natural language processing which is basically a dictionary in the form of Python module. Using this library, for a given word you can get its meaning, synonyms, antonyms, part of speech, translations and other such. This library is easy to install and is a decent substitute to Wordnet. Click here. 9| PyNLPl PyNLPl, pronounced as ‘pineapple’, is a Python library for Natural Language Processing. It contains various modules useful for common, and less common, NLP tasks. PyNLPl can be used for basic tasks such as the extraction of n-grams and frequency lists, and to build a simple language model. PyNLPl most notably features a very extensive library for working with FoLiA XML (Format for Linguistic Annotation). Click here. 10| Quepy Quepy is a python framework to transform natural language questions into queries in a database query language. It can be easily customized to different kinds of questions in natural language and database queries. Quepy uses an abstract semantics as a language-independent representation that is then mapped to a query language. This allows your questions to be mapped to different query languages in a transparent manner. Click here.","excerpt":"With the help of Natural Language Processing, an organisation can gain valuable insights, patterns, and solutions. Python is one of the widely used languages and it is implemented in almost all fields and domains. In this article, we list down 10 important Python Natural Language Processing Language libraries. 1|  Natural Language Toolkit (NLTK) NLTK is […]","categories":["AI Trends"],"tags":["Applications of Data Mining","Data Mining","data mining tools","entity resolution software","machine learning pattern recognition python","Python Library","simple python project"],"author_name":"Ambika Choudhury","publish_date":"2019-05-24T12:42:34","publication_year":"2019","word_count":710,"keywords":["machine learning pattern recognition python","spaCy","machine learning","NumPy","AI","neural network","Data Mining","entity resolution software","simple python project","ML","NLTK","Applications of Data Mining","sentiment analysis","NLP","data mining tools","deep learning","Python Library"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","neural network","NLP","spaCy","NLTK","NumPy","sentiment analysis"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-10-python-nlp-libraries-for-2019\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10058568,"title":"Is data engineering the sexier job of 21st century?&#8230; than Data Scientist","content":"Data science was termed the sexiest job of the 21st century by Harvard, but in a surprising turn of events, data engineering may overtake this status. “The dearth of data engineers will be felt even more in 2022”, noted a survey by AIM on the top AI and Data Science trends for 2022. This is owing to the increase in digital transformation after the pandemic and the explosion of data following it. Register for Data Engineering Summit 2022 Tracking the increasing demand for data engineers Historically, data engineers have only dealt with distributed systems and Java programming, but now they have to leverage AI, ML and BI to manage data. Termed the nerve centre of Digital Strategy by Sriram Narasimhan, Head of Data, Analytics and AI at Cognizant, data engineers’ demand can be seen through recent market statistics. In 2020, the Dice Tech Job Report stated data engineering to be the fastest-growing job in technology with a predicted 50% year-over-year growth in the number of open positions. Global employment platform, Monster has published two recent trends and employment reports showing that ‘engineer’ was the number one job search over the month of WHAT. Additionally, engineering was among the top sectors likely to expand in 2022 by 57%. Further, the annual salary study conducted by AIM Research in June 2021 showed that data engineers commanded a median salary greater than big-data scientists or AI engineers, indicating the growth in importance for the position. The demand for this can be seen with big tech companies like Google, IBM, Cloudera and SAS that have initiated data engineering certification programs to skill and upskill employees. However, they don’t have certifications in data science or AI engineering. “With the market for artificial intelligence and machine learning-powered solutions projected to grow to $1.2B by 2023, it’s important to consider business needs now and in the future. To address the new skills data engineers now need, we updated our Data Engineering on Google Cloud learning path,” Google stated. Why is data engineering becoming sexier? Data is everywhere and in varied forms, and it needs to be perfected to derive actionable insights. As analytics professionals, data engineers are responsible for generating, cleaning, processing, and storing data in a way that makes it ready for analysis. Additionally, they formulate data management architectures for companies to democratise data access and establish efficient pipelines. Essentially, data engineers lay the foundation for data scientists or AI\/ML professionals to use the data to derive business solutions. According to Mathangi Sri, VP Data Science & Head of Data at Gojek, the data engineering demand can be attributed to the emergence of large scale data. “Every big company today has close to ten million customers and millions of transactions a day. So, engineering is required first to make it accessible for real-time models to perform,” she explained. “Factors like penetration of data, sources of data, absence of a state-of-the-art system to present these and data governance is becoming important. Earlier, systems crashing used to be a good sign, but it is not the case today as it can destroy companies overnight. Other companies are there waiting for the failure to happen.” The eminent shortage “There is a huge demand for competent data engineers. This is the right time to immerse in it and grab the opportunity,” said Prasad Srinivasa, Assistant Vice President at Genpact. Clearly, many believe so, given the demand for data engineers has outstripped its supply since 2016, according to Quanthub. However, the shortage of engineers is more severe than that for data scientists, and it only seems to be on the rise. “As of 2021, LinkedIn is showing more than 29K job opportunities in data engineering as organisations still face a significant shortage with not enough data engineering talent in the market,” said Sriram Narasimhan. Data engineering is a relatively specialised field, as opposed to data science, that is prone to continuous upskilling and generalised positions. With the increasing demand for technical expertise, the talent gap in data engineering only keeps growing. Building multi-disciplinary teams and encouraging data engineering education and upskilling seems to be the key in the future.","excerpt":"“Earlier, systems crashing used to be a good sign, but it is not the case today as it can destroy companies overnight.”","categories":["IT Services"],"tags":["data engineering demand"],"author_name":"Avi Gopani","publish_date":"2022-01-17T14:00:00","publication_year":"2022","word_count":688,"keywords":["data science","Go","machine learning","artificial intelligence","AI","ML","data engineering demand","RAG","Aim","analytics","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","data science","analytics","Aim","RAG","R","Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/is-data-engineering-the-sexier-job-of-21st-century-than-data-scientist\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10105947,"title":"Accenture’s Gen AI Numbers Signal Shift for Indian IT","content":"In its recently announced earnings call, global IT giant Accenture revealed that despite revenue growth of a mere 3% with $16.2 billion, it made upwards of $450 million in new bookings from GenAI applications alone, including a deal with McDonald’s and BBVA, amongst others. The earnings from GenAI, up 50% from $300 million in the previous quarter, indicate Accenture’s early leadership position in GenAI compared to regional IT providers. The bold proclamation of such a strong number also makes them stand ahead, infusing market confidence in the company. “With respect to GenAI, I just want to say $450 million in sales this quarter, we’re very pleased with. It demonstrates we are leading here,” said CEO Julie Sweet. On the other hand, industry experts anticipate that gen AI’s contribution to the top IT firms’ overall revenue to be less than 1%, even with an increasing number of pilots and partnerships in this emerging field. Apurva Prasad, Vice-President of Institutional Research at HDFC Securities, observed, “The overall scale of AI deployments across sectors…is quite strong. However, generative AI’s contribution as part of it remains small for the time being.” Furthermore, in comparison, Indian IT companies like TCS, Infosys, Wipro, HCLTech, and LTIMindtree haven’t released any GenAI-specific numbers indicative of their roadmap or growth indicators. Instead, these companies have focused on partnerships with giants like Microsoft, AWS, and Google. They are focused heavily on employee upskilling and injecting significant investments, with companies like Wipro planning $1 billion in AI over three years, and launching the Wipro AI360 ecosystem. GenAI Integration with Overall Strategy Accenture incorporates GenAI within its broader digital transformation strategies for clients, ensuring alignment with their overall business goals. CEO Julie Sweet emphasised that GenAI is not plug-and-play, but a sophisticated technology requiring deep understanding to be effectively scaled. “And this is Accenture’s leadership position, right? We have a strategy, consulting, deep industry and functional expertise,” added Sweet, underscoring Accenture’s unique position in this evolving field. Experts predict GenAI bookings could exceed $4 billion in FY24, with Accenture poised to capture a significant share, backed by its $3 billion investment in data and AI over three years aimed at scaling client value in 2024. The company derives confidence to go all in from its experienced cloud journey In the realm of acquisitions and partnerships, Accenture’s collaborations with major industry players boost its capacity to merge technology with business value. Sweet foresees 2024 as a pivotal year, transitioning clients from experimentation to scalable GenAI applications. “Think of 2024 as being the shift for our clients from experimentation to scale, and we believe we’re in the best position to lead that shift to value.” Accenture is actively integrating GenAI in client projects, such as developing a GenAI-powered financial coach for BBVA and enhancing content production for a global hospitality group. A key project with McDonald’s focuses on integrating cloud technology and GenAI to revolutionise customer and employee experiences. The company’s expansion strategy in the space is also marked by significant acquisitions across various regions, totalling $788 million, enhancing cloud capabilities and strengthening its AI expertise. Notably, the acquisition of Ammagamma adds 90 AI specialists to Accenture’s team, contributing to its goal of doubling its skilled data and AI practitioners to 80,000. Previously,  the company had invested in a total of 11 AI-related businesses, of which three are from India: Flutura, Bridgei2i, and Byte Prophecy, with Flutura being the most recent investment. A Good Omen for Indian IT Indian IT companies have also shifted their focus to building a skilled workforce and ramping up their investments in GenAI as clients have shown a willingness to fund early PoC projects—leading to a new wave of collaboration between service providers and their clients to explore GenAI’s potential. Indian IT majors have trained nearly seven lakh employees in GenAI. TCS has over 100,000 GenAI-ready employees, according to Milind Lakkad, executive vice president and global head – human resources. Infosys has trained 57,000 employees, while Wipro has trained 180,000. HCLTech and LTIMindtree are also focusing on training their employees in GenAI. Infosys has launched Topaz, focusing on GenAI tech, with CEO Salil Parekh stating the company is engaged in 80 GenAI projects. Meanwhile, Coforge introduced Quasar, a GenAI platform with over 100 cognitive and generative use cases, and LTIMindtree launched Canvas.ai to help clients scale their GenAI capabilities. Additionally, Accenture, which follows a September to August financial year, disclosed that its GenAI project bookings constitute approximately 2.4% of Accenture’s total workload—which accounted for only 0.6% during the March-May period. This indicates an acceleration in momentum and increasing revenue for the industry. However, A researcher from a Mumbai-based brokerage, requesting anonymity, said GenAI use cases might eventually lead to faster monetisation, but not immediately. They added, “At least the second half of this fiscal will remain without any major deals in generative AI for any of India’s IT outsourcers.” The economic downturn has also impacted Indian IT stocks. Shares of TCS, Wipro, and Tech Mahindra have declined by 2% since September end, while Infosys’ shares have stayed flat. HCL Technologies, on the other hand, has seen growth, attributed to its mega-deal wins and a strong engineering practice. Amid these challenges, Infosys and HCL Technologies have revised their annual revenue growth projections downwards. Infosys has reduced its growth outlook to 1-2.5% for this fiscal year, while HCL has adjusted its guidance to 4-5% organic growth. Wipro indicated a potential revenue contraction of up to 3.5% for the current quarter. Analysts also expect IT services expenditure to remain muted in the near term as businesses typically decide their annual budgets only after February. Accenture itself has pointed to slower budget-related decision-making, especially in tech and media companies, setting its q2 target to -2% to 2%. Hence, while gen AI holds promise, its current impact on revenue growth in India’s IT sector seems limited amid the industry’s ongoing struggle with global economic uncertainties and conservative client expenditure. However, it will be interesting to see how earnings from GenAI pan out for Indian IT as the quarter approaches.","excerpt":"Disclosing that GenAI project bookings constitute 2.4% of its total workload indicating an acceleration in revenue from AI for IT companies","categories":["AI Features"],"tags":["Accenture","AWS","Google","Indian IT companies","Infosys","LTIMindTree","Microsoft","partnerships","TCS","Wipro"],"author_name":"Shyam Nandan Upadhyay","publish_date":"2023-12-22T12:41:56","publication_year":"2023","word_count":1005,"keywords":["Wipro","Scala","Git","R","Accenture","RAG","partnerships","Indian IT companies","Go","AWS","AI","generative AI","TCS","GenAI","Infosys","Aim","Google","LTIMindTree","Microsoft"],"extracted_tech_keywords":["AI","generative AI","GenAI","Aim","RAG","AWS","R","Go","Scala","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/accentures-gen-ai-numbers-signal-shift-for-indian-it\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10102727,"title":"Fostering an Empowering Culture that Sparks Powerful Intelligence: Insights from Data &#038; AI Experts at KPI Partners","content":"In recent years, there has been a notable transformation in the realm of employee engagement. Organizations, in their pursuit of fostering improved work environments, have adopted diverse strategies to meet the changing demands of their workforce. Today, many companies, including KPI Partners, are introducing creative solutions such as flexible scheduling, remote work possibilities, and generous vacation policies to promote work-life balance and elevate employee satisfaction. Founded in 2006 by Sid Goel and Kusal Swarnakar, KPI Partners has emerged as a prominent provider of cloud-native solutions encompassing Generative AI, Analytics, Business intelligence, Data management, and Data engineering. Their team comprises more than 600 experts, delivering cloud-based services designed to prioritize scalability, security, and user-friendliness. Headquartered in Newark, California, KPI Partners serves Fortune 500 companies and maintains a presence across North America and Asia. In March of this year, KPI Partners achieved its third consecutive ‘Great Place to Work’ certification. An anonymous poll revealed that nearly 91% of employees believe the company is indeed a fantastic workplace. “We believe in setting our new employees up for success from day one with a comprehensive onboarding program,” said Daniel Shaw, who was formerly the Senior Director of HR and Talent Acquisition of the company and currently serves as the Senior Director of Client Services for APAC & Middle East at KPI Partners. “Mentorship opportunities, team-building activities, and continuous training are integral components that equip every team member with the necessary tools to excel.” What makes KPI a great workplace? In addition to its focus on the work-life balance of its employees, the company, as stated by Shaw, places a strong emphasis on transparent communication and empathy. This includes offering employees clear guidance, consistent feedback, and unwavering support. They are actively encouraged to take ownership of their roles, fostering a sense of accountability for excellence. Collaboration and striving for excellence are also fundamental values within the company. “KPI Partners recognizes the significance of upskilling and fostering professional growth. This is why we provide ongoing learning programs and training sessions for our team members. We actively promote and support their pursuit of professional development opportunities, including attending conferences and enrolling in courses, to enrich their skills and knowledge. Our organization offers diverse career advancement paths, allowing our employees to thrive and progress within our company.” KPI Partners also rewards proactive behaviours, acknowledges outstanding performance, and values innovative ideas. Through regular feedback and performance evaluations, the company helps employees identify their strengths and areas for improvement. “Our engaging work environment is enhanced by social events, team-building activities, and celebrations.” Moreover, Shaw reveals that KPI emphasises employee well-being through comprehensive wellness programmes encompassing insurance coverage, regular health check-ups, fitness classes, and mindfulness workshops. “Our commitment to a healthy work-life balance encourages breaks, fostering a culture of productivity and creativity. Engaging in social responsibility initiatives enables employees to contribute positively to society,” said Shaw. KPI culture Establishing a culture that fosters employee engagement is paramount to the success of any organization. “At KPI Partners, we place a strong emphasis on open communication and a supportive atmosphere where every employee feels valued and heard. We actively collect insights and gain an understanding of employee perceptions by conducting regular team meetings, providing feedback opportunities, administering surveys, and conducting one-on-one check-ins. This ongoing process allows us to continually enhance our company culture.” Moreover, KPI Partners is committed to promoting employee career growth, offering abundant opportunities for professional development, including training programs and clearly defined career paths. “We place great importance on recognizing and rewarding hard work, actively involving our employees in the company’s success. Additionally, we have instituted various programs to support our employees throughout their journey with our organization. For instance, we initiate an onboarding period called “Parichay” from the moment they join, allowing new hires to acclimate to our organization. After three months, they take part in “Anubhav,” a survey enabling them to share their experiences and provide valuable input that helps us refine our processes. We further have “Bird’s Eye View” sessions where our centre head engages with the team to help align them with our work environment.” To encourage feedback and suggestions, we have implemented suggestion boxes and maintained regular meetings with managers. We highly value employee feedback, and it has led to the implementation of various initiatives, including providing an internet allowance for remote work, introducing the buddy program, and offering access to specialized doctors in India through “Medi Buddy.” Providing an inclusive workspace KPI Partners places a strong emphasis on building a sense of unity and shared purpose among its employees. To achieve this, the company organizes a variety of team-building activities aimed at promoting collaboration and strengthening interpersonal relationships. “These activities range from group outings to team-building exercises that promote trust and communication. We value the input of every employee and promote an open-door policy where everyone feels comfortable sharing ideas and concerns with their superiors. This, in turn, enhances inclusivity and fosters a collaborative culture.” Besides, inclusivity in hiring and promotion practices remains a key focus area. KPI Partners provide equal opportunities for all employees to grow and advance in their careers, regardless of their background or identity. This commitment to inclusivity not only creates a more diverse workforce but also fosters a sense of belonging and acceptance among employees, according to Shaw. Ensuring strong women’s representation In the contemporary corporate landscape, progressive companies prioritize measures to ensure robust women’s representation in their workforce. Underlining the significance of advancing women’s leadership and presence in the workplace, Daniel Shaw emphasizes that KPI Partners is dedicated to creating an inclusive and diverse environment where women can not only thrive but also excel in their careers. “To achieve this, we have implemented several strategies. We actively support talented women within our organisation through mentorship programmes and leadership development initiatives. By investing in their professional growth, KPI empowers women to assume leadership positions and make substantial contributions to the company’s success. Our transparent work culture encourages open communication and equal participation in decision-making processes.” Recognizing the challenges that women may encounter in balancing their professional and personal responsibilities, KPI has introduced various initiatives. This includes providing flexible work arrangements, such as remote work options, and accommodating maternity and sabbatical leave policies. “The safety and well-being of employees are highly valued, leading to the implementation of redressal systems like PoSH. We have also introduced flexible work timings to ensure that client calls after 7 p.m. are handled by other team members. This not only aids in retaining talented women but also enhances their overall well-being and job satisfaction.” Additionally, KPI also promotes gender diversity in our recruitment and hiring processes, ensuring that the job postings are inclusive and encouraging women to apply. Consequently, KPI has witnessed a positive impact on its hiring procedures, with over 50% of women included in recent cohorts of entry-level recruits. “This year, we have seen a steep increase of 15% in the hiring percentage. Additionally, we have implemented unconscious bias training for our hiring managers to ensure fair and unbiased selection processes.” Creating an environment of trust, transparency & innovation Commenting on what makes KPI Partners a Great Place to Work,  Daniel Shaw highlights the company’s commitment to values that underpin its long-term success and create a positive work environment. These values are defined as ‘Innovation, Integrate, Influence, and Trust.’ Key to the organization is transparency, as it seeks to provide clear and honest information to its customers. “This approach not only builds trust but also fuels innovation, as KPI openly shares ideas and feedback. The company’s ethos revolves around making the customer experience enjoyable and memorable, infusing fun into interactions, and leaving lasting, positive impressions.” To support these principles, KPI has introduced several initiatives. Firstly, the company established a dedicated learning and development division known as “Gurukul”. Managed by the human resource and delivery department, Gurukul ensures that employees possess the right skills for their roles. “We identify those who need to reskill or upskill and create personalised learning paths in consultation with their managers. This ensures that our team is always equipped with the latest knowledge and expertise.” “We believe in building connections and fostering collaboration. That’s why we hold a monthly session called ‘Let’s have a conversation’. This brings together new joiners, existing team members, and managers to create a strong sense of community and support. Furthermore, KPI also conducts early sessions between employees and the managers, offering a comprehensive overview of the work environment. To further facilitate growth and development, KPI managers meet with the Center Head every month to discuss the learning curve of their team members, ensuring they remain on the right track. We also follow a top-down and bottom-up approach to communication, keeping everyone on the same page. Succession planning is a priority for us, as we identify key roles and positions to ensure business contingency plans. Furthermore, our Centers of Excellence is a hub for internal career development, fostering mobility and growth within the organization. To boost employee morale and confidence, we organise roadshows, webinars, and demos, creating a well-rounded and supportive environment,” Shaw concluded.","excerpt":"Headquartered in Newark, California, KPI Partners serves Fortune 500 companies and maintains a presence across North America and Asia.","categories":["AI Highlights"],"tags":[],"author_name":"Pritam Bordoloi","publish_date":"2023-11-08T12:45:16","publication_year":"2023","word_count":1509,"keywords":["Go","AI","Scala","RAG","Aim","data engineering","generative AI","analytics","Rust","R"],"extracted_tech_keywords":["AI","analytics","generative AI","Aim","RAG","R","Go","Rust","Scala","data engineering"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/fostering-an-empowering-culture-that-sparks-powerful-intelligence-insights-from-data-ai-experts-at-kpi-partners\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":68422,"title":"How ML Is Changing The Way We Use Touchscreens","content":"The touchscreens which we use at supermarkets and ATMs were accidentally invented by a group of atomic physicists back in 1970. The conception of touchscreens can be traced back to the 1940s, even before science fiction writers warmed up to the innovation. Today, the use of touchscreens is only bounded by the creativity of users. You can pinch, zoom, type and move the world literally with your fingers. However, a typical user might have had the experience of typos, unwanted clicks and many other mishits, which couldn’t be undone. So, what if the touchscreens are made smart enough to understand the touch better? The researchers at Google used state-of-the-art machine learning models and designed an intelligent touchscreen for their Pixel phones. How Touch Works via Google AI As illustrated above, when a finger interacts with a touch sensor cell, it disturbs the charge from the projected field around two electrodes. A capacitive touch sensor is a collection of electrodes arranged in rows and columns, which are separated by a dielectric. Capacitive touch sensors don’t respond to changes in force but are tuned to be highly sensitive to changes in distance within a couple of millimetres above the display. That is, a finger contact on the display glass should saturate the sensor near its centre, but will retain a high dynamic range around the perimeter of the finger’s contact (where the finger curls up). When the soft tissue of the finger touches the screen, it deforms and spreads out. The nature of this spread depends on the size and shape of the user’s finger, and its angle to the screen. This is also a dynamic change that occurs over some period of time, which differentiates it from contacts that have a long duration or a large area. Beyond tapping and swiping gestures, long-pressing has been the main alternative path for interaction. For mobile to understand a long press, the user’s finger must remain stationary for 400–500 ms. This delay is not bad, but neither is it immediate, which can ruin the user experience for some. An alternative to time threshold based interaction is through force estimation. But, there are other challenges such as differentiating between soft and firm touch, which in turn requires hardware sensors. What ML Got To Do With Touch via Google AI Sensing touch force requires dedicated hardware sensors that are expensive to design and integrate. According to Google, touch force is difficult for people to control, so most practical force-based interactions focus on discrete levels of force (e.g., soft vs. firm touch) that do not require the full capabilities of a hardware force sensor. The differences between users (and fingers) make it difficult to encode these observations with heuristic rules. Google, therefore, designed a machine learning solution that would allow us to learn these features and their variances directly from user interaction samples. via Google AI The researchers at Google designed a neural network that combines convolutional (CNN) and recurrent (RNN) components. The CNN takes care of the spatial features, while the RNN is used for tracking temporal development and provides a consistent runtime experience. Each frame is processed by the network as it is received from the touch sensor, and the RNN state vectors are preserved between frames (rather than processing them in batches). The network was intentionally kept simple to minimise on-device inference costs when running concurrently with other applications taking approximately 50 µs of processing per frame and less than 1 MB of memory using TensorFlow Lite. Models used: CNN, RNNFramework: TensorFlow Lite Through this integration of machine-learning algorithms and careful interaction design, Google researchers were able to deliver a more expressive touch experience for Pixel users and plan to explore new forms of touch interaction. The applications of touchscreens now extend far beyond your typical smartphone usage. Touchscreens are now being used in hospitals, autonomous cars and even in spacecraft. These are few of the critical scenarios where the risk is too high in case of a mishit. Having an intelligent surface that knows what we want just from sensing the force of our finger can come in quite handy as even the AutoML tools are transitioning towards touch-based drag and drop analytics. Know more here.","excerpt":"The touchscreens which we use at supermarkets and ATMs were accidentally invented by a group of atomic physicists back in 1970. The conception of touchscreens can be traced back to the 1940s, even before science fiction writers warmed up to the innovation. Today, the use of touchscreens is only bounded by the creativity of users. […]","categories":[],"tags":["cnn"],"author_name":"Ram Sagar","publish_date":"2020-06-28T16:00:33","publication_year":"2020","word_count":703,"keywords":["Go","machine learning","AI","neural network","ML","cnn","RAG","analytics","CNN","TensorFlow","R"],"extracted_tech_keywords":["AI","machine learning","ML","neural network","analytics","TensorFlow","RAG","R","Go","CNN"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/machine-learning-for-touchscreens-mobiles\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10171301,"title":"Meta to Let AI Create Ads by 2026: Report","content":"Meta is doubling down on artificial intelligence to revolutionise digital advertising, with plans to enable brands to develop entire advertising campaigns from scratch using AI by the end of next year, according to a report by The Wall Street Journal (WSJ). Meta, which owns Facebook and Instagram, is reportedly developing tools that would allow brands to input a product image and a budget goal. After that, AI would generate everything from visuals and videos to ad copy. The AI would also decide which users to target and suggest budget allocations, the WSJ reported, citing people familiar with the matter. This is a step beyond Meta’s existing AI-powered ad tools, which can currently generate minor variations of pre-existing ads. Now, the company aims to automate the creative process entirely, tailoring content to user-specific factors like geolocation. For instance, a user in a snowy region may see a car ad set in mountainous terrain, while someone in a city might see the same car navigating urban streets. CEO Mark Zuckerberg emphasised this vision during the company’s annual shareholder meeting, describing it as “a redefinition of the category of advertising”. On a podcast appearance, he said Meta ultimately aims to let businesses simply specify their marketing goal, budget per result and bank account, leaving the rest to AI. Advertising accounts for over 97% of Meta’s revenue, and the company is channelling those profits into AI infrastructure, including chips, data centres and advanced model training. The initiative is expected to be especially benefit small and medium-sized businesses, which form the bulk of Meta’s advertisers and often lack the resources for full-scale ad production. Nevertheless, some concerns include Meta’s increasing control over campaign strategy and creative direction and whether AI-generated ads can match the polish of those made by human teams. Many AI tools across the industry still struggle with distorted or low-quality visuals, requiring manual refinement. The broader industry is also moving in this direction. Google recently launched an upgraded version of its video-generation tool, Veo, which is capable of producing short films from text prompts. Meanwhile, third-party platforms like Midjourney and OpenAI’s DALL·E are already being used by brands to create visuals across digital platforms, including Meta’s own. Meta is also exploring ways to integrate such third-party tools into its ad ecosystem, further consolidating its role in the future of AI-powered advertising.","excerpt":"The AI would also decide which users to target and suggest budget allocations, the WSJ reported.","categories":["AI News"],"tags":["Ads","Meta"],"author_name":"Merin Susan John","publish_date":"2025-06-04T13:26:04","publication_year":"2025","word_count":389,"keywords":["Go","Meta","artificial intelligence","OpenAI","AI","programming_languages:R","programming_languages:Go","Git","Aim","R","Ads"],"extracted_tech_keywords":["AI","artificial intelligence","OpenAI","Aim","R","Go","Git","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/meta-to-let-ai-create-ads-by-2026-report\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10164497,"title":"Blackstone, Panchshil to Build India’s Largest Data Centre with ₹20,000 Crore Investment","content":"Blackstone Group and Panchshil Realty are set to develop India’s largest hyperscale data centre in Navi Mumbai, with an investment exceeding ₹20,000 crore. The 500 MW facility will span 14 buildings across 3 million sq ft and will operate on up to 65% renewable energy, making it the first of its kind in India. The project marks the first major Foreign Direct Investment (FDI) since Blackstone’s recent agreement with the Maharashtra government at Davos. The US-based asset manager signed three agreements at the event as part of its strategy to expand in India’s digital and property infrastructure sectors. To initiate development, Panchshil Realty’s subsidiary, Gramercy Info Park, has acquired nearly 50 acres of land adjacent to Reliance Corporate Park in Navi Mumbai’s Ghansoli area for Rs 900 crore, covering all related levies and charges. “The proposed data center will be a built-to-suit facility catering to hyperscalers specialising in cloud computing, artificial intelligence, and high-performance computing infrastructure,” said a person familiar with the development to ET. Blackstone is already constructing two data centres in Navi Mumbai under Lumina CloudInfra, with a combined investment of over $600 million and a total capacity of 60 MW. The new project will significantly expand Blackstone’s footprint in India’s data center industry. With over $50 billion in assets under management in India, Blackstone has been one of the country’s largest investors. Nearly 40% of these funds have been directed towards Maharashtra, with significant investments across Bengaluru, Hyderabad, Pune, NCR, Chennai, and Kolkata. The firm also dominates India’s office real estate market, holding 135 million sq ft across 50 properties. Having completed and exited two office REITs in India, Blackstone is currently preparing to file for its third office REIT in the country.","excerpt":"The 500 MW facility will span 14 buildings across 3 million sq ft and will operate on up to 65% renewable energy, making it the first of its kind in India.","categories":["AI News"],"tags":["Data Center","data center India"],"author_name":"Mohit Pandey","publish_date":"2025-02-25T08:19:06","publication_year":"2025","word_count":286,"keywords":["Go","data center India","artificial intelligence","Data Center","programming_languages:R","cloud computing","AI","programming_languages:Go","Git","R"],"extracted_tech_keywords":["AI","artificial intelligence","cloud computing","R","Go","Git","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/blackstone-panchshil-to-build-indias-largest-data-centre-with-%e2%82%b920000-crore-investment\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10123968,"title":"HPE Announces ‘NVIDIA AI Computing by HPE’ to Accelerate Generative AI Adoption","content":"Hewlett Packard Enterprise (HPE) and NVIDIA announced NVIDIA AI Computing by HPE, a portfolio of co-developed AI solutions and joint go-to-market integrations that enable enterprises to accelerate adoption of generative AI. Among the portfolio’s key offerings is HPE Private Cloud AI, a first-of-its-kind solution that provides the deepest integration to date of NVIDIA AI computing, networking and software with HPE’s AI storage, compute and the HPE GreenLake cloud. The offering enables enterprises of every size to gain an energy-efficient, fast, and flexible path for sustainably developing and deploying generative AI applications. Powered by the new OpsRamp AI copilot that helps IT operations improve workload and IT efficiency, HPE Private Cloud AI includes a self-service cloud experience with full lifecycle management and is available in four right-sized configurations to support a broad range of AI workloads and use cases. All NVIDIA AI Computing by HPE offerings and services will be available through a joint go-to-market strategy that spans sales teams and channel partners, training and a global network of system integrators — including Deloitte, HCLTech, Infosys, TCS and Wipro — that can help enterprises across a variety of industries run complex AI workloads. Announced during the HPE Discover keynote by HPE President and CEO Antonio Neri, who was joined by NVIDIA founder and CEO Jensen Huang, NVIDIA AI Computing by HPE marks the expansion of a decades-long partnership and reflects the substantial commitment of time and resources from each company. “Generative AI holds immense potential for enterprise transformation, but the complexities of fragmented AI technology contain too many risks and barriers that hamper large-scale enterprise adoption and can jeopardize a company’s most valuable asset – its proprietary data. “To unleash the immense potential of generative AI in the enterprise, HPE and NVIDIA co-developed a turnkey private cloud for AI that will enable enterprises to focus their resources on developing new AI use cases that can boost productivity and unlock new revenue streams,” said Neri. “Never before have NVIDIA and HPE integrated our technologies so deeply – combining the entire NVIDIA AI computing stack along with HPE’s private cloud technology – to equip enterprise clients and AI professionals with the most advanced computing infrastructure and services to expand the frontier of AI,” Huang added.","excerpt":"All NVIDIA AI Computing by HPE offerings and services will be available through a joint go-to-market strategy","categories":["AI News"],"tags":["ai announcements","HPE","NVIDIA"],"author_name":"Pritam Bordoloi","publish_date":"2024-06-19T13:28:10","publication_year":"2024","word_count":372,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","RAG","HPE","ai announcements","ViT","generative AI","NVIDIA","R"],"extracted_tech_keywords":["AI","generative AI","RAG","R","Go","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/hpe-announces-nvidia-ai-computing-by-hpe-to-accelerate-generative-ai-adoption\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10100111,"title":"Analytics India Magazine Forays Into the US","content":"Analytics India Magazine (AIM), your favourite platform for a wide-angle world view on analytics, data science, and artificial intelligence, is excited to announce its official entry into the United States. This strategic expansion reflects AIM’s commitment to providing valuable insights, knowledge sharing, and fostering innovation in the field of AI and analytics on a global scale. With over a decade of experience as the go-to platform for industry professionals, AIM has been at the forefront of the analytics revolution in India and beyond. The foray into the US market is a significant milestone in its journey, as it makes AIM one of the very few Indian mediahouses to have expanded globally. The move aims to bridge the gap between analytics professionals, businesses, and academia across continents. The US market presents an incredible opportunity for AIM to engage with a diverse and dynamic community of data professionals, data scientists, AI researchers, and decision-makers. The expansion will enable AIM to deliver tailored content, events, and networking opportunities that cater specifically to the unique needs and challenges of the US AI and analytics ecosystem. Bhasker Gupta, founder and CEO of Analytics India Magazine, agrees to the move being more than just a geographical expansion as it marks our ambition to evolve into a truly global organisation. “We firmly believe that the work we are doing at AIM is unparalleled on a global scale. The depth and breadth of services we offer have garnered interest from various international organisations, underscoring the potential impact we can make worldwide,” said Gupta, touching upon various initiatives by AIM, including MachineHack, AIM Leaders Council, AIM Research, AIM Recruits and AIM Events among others. For more information about Analytics India Magazine, please visit https:\/\/analyticsindiamag.com\/ About Analytics India Magazine Analytics India Magazine (AIM) is a leading source for analytics, data science, and artificial intelligence news, insights, and resources. With a strong presence in India and a growing global audience, AIM is dedicated to fostering a vibrant analytics ecosystem by providing valuable content, hosting industry events, and facilitating collaboration among professionals, businesses, and academia. For media inquiries, please write to us: info@analyticsindiamag.com","excerpt":"With over a decade of experience as the go-to platform for industry professionals, AIM has been at the forefront of the analytics revolution in India and beyond","categories":["AI News"],"tags":[],"author_name":"Pranav Kashyap","publish_date":"2023-09-15T12:44:43","publication_year":"2023","word_count":351,"keywords":["data science","Go","artificial intelligence","AI","ML","Ray","Aim","analytics","GAN","R"],"extracted_tech_keywords":["AI","artificial intelligence","ML","data science","analytics","Aim","Ray","R","Go","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/analytics-india-magazine-forays-into-the-us\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10043235,"title":"How Researchers Are Solving Cybersecurity Issues In Autonomous Vehicles","content":"Autonomous technology is growing at an incredible rate. Recently researchers at MIT developed a single deep neural network (DNN) to power autonomous vehicles on NVIDIA DRIVE AGX Pegasus. Vehicle manufacturer Volkswagen which is on a mission to manufacture EVs, is now working to bring self-driving technologies to global markets. According to media reports, Chief Executive Herbert Diess believes that the future of cars lies in autonomous driving technology more than electrification. While autonomous technology might be capable of transforming the future, it comes with its own set of pros and cons. Autonomous vehicles (AV) generally require extensive testing before they can be widely deployed. Apart from general testing, external and internal safety issues like hacking, malfunctioning etc., can occur anytime. Once self-driving cars are fully deployed, malicious hackers might target vehicles to obtain access to the driver’s personal information, or worse, be able to seize control of the vehicle’s steering or acceleration. Keeping this in mind, a team at the Ohio State Center for Automotive Research (CAR) is developing a Mobility Cyber Range (MCR), a dedicated platform for cybersecurity testing for self-driving vehicles. With the high-performance, energy-efficient NVIDIA DRIVE platform, @OhioState is taking its AV cybersecurity research to infinity, and beyond: https:\/\/t.co\/76F6j3IOQa pic.twitter.com\/FH8R6kQBkL— NVIDIA DRIVE (@NVIDIADRIVE) June 11, 2021 Performance through NVIDIA The “CyberCAR” is a Chrysler Pacifica minivan equipped with the NVIDIA DRIVE AGX Pegasus AI computing platform, enables level four autonomous driving capabilities. The initial focus of the research pilot will be on developing standards and recommendations for best practices in AV safety and cybersecurity. CAR intends to use the vehicle as a continuing research tool, with safety and security at the forefront of all projects. The technology can be used in connected and autonomous vehicle applications. For instance, CyberCAR was recently featured in a demonstration that used driverless vehicles to transport organs for emergency transplant surgery. CAR is also training students on AI computing, a technology widely used in the AV industry, by undertaking the development work on NVIDIA. Qadeer Ahmed, Associate Professor of Research at the Departments of Mechanical and Aerospace Engineering, and Electrical and Computer Engineering, and an Associate Fellow of CAR, said, “We are aiming to generate a workforce that understands these safety and security challenges, as well as one that is familiar with the DRIVE platform and equipped with the right AV skill set for success.” There is no doubt that futuristic vehicles will be bundled with computational capability more than the most advanced computing systems today. Therefore NVIDIA is working towards providing the most comprehensive solutions for the future of AVs. In 2019,  Argus Cyber Security, a global leader in automotive cybersecurity, worked closely with NVIDIA to add cybersecurity defence layers to the NVIDIA DRIVE. It demonstrated its cybersecurity solution on the NVIDIA DRIVE computing platform for autonomous vehicles at CES. The NVIDIA DRIVE computation solutions will enable self-driving vehicle developers to harness AI processing to handle the vast array of deep neural networks in vehicles. Additionally, it will allow new features to be introduced over the air, with no disruption to the existing functions. A comprehensive software stack for AV developers is also provided on the NVIDIA DRIVE platform– DRIVE OS, DriveWorks middleware, DRIVE AV and DRIVE IX. According to Qadeer, they needed a full-fledged, autonomous driving onboard computer optimised for AV functionalities for the Mobility Cyber Range. They aimed to use tools that would be highly accepted in the market and enhance student competitiveness. Future of research For the CAR research team, cybersecurity testing is just the beginning. Qadeer explained that the team’s goal is to gain as much knowledge as possible from the AV’s construction and then explore the safety and security of various modules such as perception, planning, and actuation. This work will assist in informing the industry of best practices and standards, ensuring that the wider deployment of autonomous vehicle technology is a success. Additionally, by employing NVIDIA DRIVE technologies, Qadeer’s students can continue innovating in their field long after completing the programme.","excerpt":"Autonomous technology is growing at an incredible rate. Recently researchers at MIT developed a single deep neural network (DNN) to power autonomous vehicles on NVIDIA DRIVE AGX Pegasus. Vehicle manufacturer Volkswagen which is on a mission to manufacture EVs, is now working to bring self-driving technologies to global markets. According to media reports, Chief Executive […]","categories":["AI Features"],"tags":["autonomous technology","deep neural network","NVIDIA"],"author_name":"Ritika Sagar","publish_date":"2021-07-09T11:00:00","publication_year":"2021","word_count":666,"keywords":["Go","AI","neural network","deep neural network","BERT","autonomous technology","Ray","Aim","llm_models:BERT","disruption","GAN","NVIDIA","R"],"extracted_tech_keywords":["AI","neural network","Aim","Ray","R","Go","BERT","GAN","disruption","llm_models:BERT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-researchers-are-solving-cybersecurity-issues-in-autonomous-vehicles\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":12404,"title":"Cyber Warfare at large in Southeast Asia, India leverages AI for the same cause","content":"India will leverage AI in Cyber Warfare Southeast Asia responded with a bang to the growing concern surrounding cyber threats in the region by leveraging Artificial Intelligence. In the wake of cyber warfare, India is beefing up by signing an agreement with tech startup Innefu. The startup will utilize AI to analyze data fed by Indian intelligence agencies to identify patterns from the past and predict future outcomes. The latest offering from the startup is called Prophecy. The tool was recently used to identify main players of an incident and the connection among them, by analyzing intelligence documents, including social media snippets on planned protests. The world’s fourth largest internet population resides in southeast Asia. The surge in the use of smartphones has only worsened the situation, due to improper incorporation of cyber security practices or the absence of an adequate system that can support data protection laws. As a result, vulnerabilities encountered in southeast Asia can affect the safety of the internet worldwide. The threats don’t stop here, as illegal software is being used to infect systems with malware, and India is susceptible as well. Case in point is Nepal, wherein the government and the military institution were targeted by an organization called Naikon. With an aim to promote closer cooperation on cyber security, India and Vietnam signed a Memorandum of Understanding (MOU) last year. The MOU will help both countries exchange knowledge sharing in detection, resolution, and prevention of cyber security-related incidents between them. India has also been at the receiving end of cyber attacks. 20,000 pages describing the classified combat capability of six submarines that French shipbuilder DCNS was designing for the Indian Navy were leaked to the media last year, which was later followed by hacking of opposition leader Rahul Gandhi’s Twitter account. Tarun Wig, Co-founder, Innefu Labs, concludes “India has to stay ahead in the cyber revolution, a cyber warfare is happening right now.”","excerpt":"Southeast Asia responded with a bang to the growing concern surrounding cyber threats in the region by leveraging Artificial Intelligence. In the wake of cyber warfare, India is beefing up by signing an agreement with tech startup Innefu. The startup will utilize AI to analyze data fed by Indian intelligence agencies to identify patterns from […]","categories":["AI News"],"tags":["AI India","smartphones India","social media India"],"author_name":"Amit Paul Chowdhury","publish_date":"2017-01-30T12:56:41","publication_year":"2017","word_count":319,"keywords":["Go","artificial intelligence","AWS","AI","cloud_platforms:AWS","social media India","RAG","GAN","Aim","AI India","smartphones India","R","startup"],"extracted_tech_keywords":["AI","artificial intelligence","Aim","RAG","AWS","R","Go","GAN","startup","cloud_platforms:AWS"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/cyber-warfare-large-southeast-asia-india-leverages-ai-cause\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10046625,"title":"Neuroscience Startup BrainSightAI Raises USD750,000 In Seed Round","content":"Bengaluru-based, BrainSightAI – a deep-tech neuroscience start-up, has raised USD750,000 in a seed round led by Stanford Angels and Entrepreneurs India, with participation from Entrepreneur First, Info Edge Ventures, and IKP Knowledge Park. Founded by Laina Emmanuel and Dr Rimjhim Agrawal, the company plans to leverage the capital to enable greater precision in diagnosing and treating neuro-oncological and neuropsychiatric disorders. “Neuro disorders, such as oncologic, trauma and psychiatric, are the third-leading contributors to the global disease burden. To be able to address these potentially life-threatening disorders better, the healthcare industry needs tools to predict outcomes earlier, discover new clinical and drug indications, and better understand drug and therapy responses. Such intuitive tools weren’t possible earlier. However, that is changing now. We are now at an inflexion point, where, with data, lots of computing power, and artificial intelligence, we can cloud-enable complex neuroscience-based workflows. Now, doctors and researchers can focus on asking questions and not on the mechanics of answering questions,” said Laina Emmanuel, Co-founder and CEO, BrainSightAI. In June 2021, the company was selected as one of the 20 start-ups from 500 applicants to the MedTech Innovator APAC 2021 Accelerator program. “BrainSight is not just improving MRI imaging, but based on Dr Rimjhim’s research at NIMHANs, leapfrogging the true potential of this modality to operate at the cutting edge of science, technology, and impact. BrainSight’s platform has the potential to change the global clinical paradigm for neurosurgery and neuropsychiatry, and we are excited to partner with Laina and Rimjhim on this journey”, commented Akshat Shah and Sandeep Singhal of Stanford Angels and Entrepreneurs India. It has also won the BIRAC BIG 15 grant from the Government of India in January 2020 for conducting clinical studies to investigate alteration in brain function in Dementia.","excerpt":"In June 2021, the Company was selected as one of the 20 start-ups from 500 applicants to the MedTech Innovator APAC 2021 Accelerator program.","categories":["AI News"],"tags":["neuroscience","Startups"],"author_name":"kumar Gandharv","publish_date":"2021-08-24T12:41:55","publication_year":"2021","word_count":293,"keywords":["Go","API","artificial intelligence","programming_languages:R","AI","Modal","programming_languages:Go","RAG","neuroscience","Startups","R"],"extracted_tech_keywords":["AI","artificial intelligence","RAG","R","Go","API","Modal","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/neuroscience-startup-brainsightai-raises-usd750000-in-seed-round\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10022847,"title":"Python 3.10: What To Expect","content":"Python is one of the TIOBE Index programming languages of the year. It has become the go-to language for developers working on data science and machine learning for a few years now. What sets Python apart is its intuitive features like libraries, high productivity, ease of learning etc. According to a recent study, almost 27% of the advertised jobs require Python as a core skill — up from 18.5% at the beginning of the year. Last October, Python released version 3.9. The updates included improvements in Python internals, performance boosts, dictionary union operators, handy new string functions, new type operations, consistent and stable internal APIs and more. What To Expect From 3.10 Python started work on the upcoming version, Python 3.10 pre-alpha, in May 2020. Python version 3.10 is set to be released on 4 October 2021. However, with the pre-alpha, the company has unveiled an early developer preview of Python 3.10. The current alpha version of 3.10 is 3.10.0a6, which is the sixth of seven planned alpha releases. During the alpha phase, features may be added up until the start of the beta phase and, if necessary, may be modified or deleted until the release candidate phase. The significant new features of the Python 3.10 version are mentioned below: Structural Pattern Matching The pattern matching feature has been added in the form of a match statement and case statements of patterns with associated actions. Patterns usually consist of sequences, primitive data types mappings, as well as class instances. Pattern matching will help developers extract information from various complex data types, branch on the structure of data, and apply specific actions based on different data forms. Know more here. Better Error Messages In The Parser When parsing code that contains unclosed parentheses or brackets the interpreter now includes the location of the unclosed bracket of parentheses instead of displaying SyntaxError: unexpected EOF while parsing or pointing to some incorrect location. Previous versions of the interpreter reported confusing places like the location of the syntax error but in Python3.10 a more informative error is emitted. Know more here. Parameter Specification Variables Parameter Specification Variables feature supports forwarding the parameter types of one callable over to another callable, making it difficult to annotate function decorators. This feature proposes typing.ParamSpec and typing.Concatenate that help in specifying the type of a callable. Know more here. Precise Line Numbers for Debugging and Other Tools According to its developers, Python should guarantee that when tracing is turned on, “line” the tracing events are generated for all lines of code executed and only for lines of code that are executed. A side effect of ensuring correct line numbers is that some bytecodes will need to be marked as artificial and to assist such tools, a new co_lines attribute will be added that describes the mapping from bytecode to source. Know more here. Deprecate Distutils Module Distutils is an undocumented and unmaintained collection of utilities for packaging and distributing Python packages, including the compilation of native extension modules. This feature defines a configuration format that describes a Python distribution and provides the tools to convert a directory of source code into a source distribution and binary distribution forms. In both the version of Python 3.10 and 3.11, distutils will be formally marked as deprecated. Meaning, all the known issues will be closed at this time, and import distutils will raise a deprecation warning. During Python 3.10 and 3.11, uses of distutils within the standard library may change to use alternative APIs. However, in Python 3.12, distutils will no longer be installed by make install or any of the first-party distributions. Know more here.","excerpt":"Python is one of the TIOBE Index programming languages of the year. It has become the go-to language for developers working on data science and machine learning for a few years now.  What sets Python apart is its intuitive features like libraries, high productivity, ease of learning etc. According to a recent study, almost 27% […]","categories":["AI Trends"],"tags":["Python language"],"author_name":"Ambika Choudhury","publish_date":"2021-03-25T13:00:00","publication_year":"2021","word_count":605,"keywords":["data science","Go","Python language","machine learning","API","programming_languages:R","AI","Python","ViT","programming_languages:Python","R"],"extracted_tech_keywords":["AI","machine learning","data science","Python","R","Go","API","ViT","programming_languages:Python","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/python-3-10-what-to-expect\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":36307,"title":"5 Data Science And Analytics Job Openings In Malaysia To Watch Out For","content":"Year by year, the demand for data science skills is growing exponentially all across the globe, and Malaysia is not behind in the AI race. Data science is no longer an uncommon space in Malaysia, and organizations across have started to leverage data science and use its capabilities to fine-tune business operations — from improving the overall customer experience to expanding product and services offering. So, if you are someone who wants to pursue a career in data science and analytics in Malaysia, then it is definitely a great time to enter the space. Here Are 5 Data Science And Analytics Job Openings in Malaysia Data Scientist (TIS) at Grab – Petaling Jaya, Malaysia At Grab, the Singapore-based ride-hailing company, the Trust, Identity & Safety team act as guardians of all its users. The team make use of rich datasets to find solutions to problems ranging from safety to fraud. As a data scientist (TIS) at Grab, your job role will include: Developing a deep behavioural understanding and intuition of our usersBuilding machine learning models for user safety Managing the entire end-to-end lifecycle of designing models, Analyzing data and trends. Interface with business & operation teams to formulate solutions & product changes informed by your findings Prerequisites: Minimum 3 years of experience building machine learning models in production, or an advanced degree in a quantitative field with at least 1 year of experience.Good understanding of machine learning models, feature engineering, and other practical issues, such as overfitting.Solid engineering and coding skills. Experience in NLP, computer vision, telematics or IMU motion tracking is a plus.Experience in MapReduce, Spark and Hive is a plus.Good communication skills and experience managing stakeholders is a plus. Apply Here Data Analyst – Risk Analytics, TIS at Grab – Petaling Jaya, Malaysia As the Data Analyst in TIS at Grab, your job will be about analysing the company’s data to develop insights about user behaviours and platform risks. Also, you will have to translate these insights into recommended actions in order to combat different types of fraud. Prerequisites: 3-8 years of experience as a hands-on analyst A technical degree in Computer Engineering, Information Technology, Data Sciences, Mathematics or related fields.Have a strong hand in handling large scale unstructured data. Experience in BI tools such as Tableau, Qlik Sense and Excel Experience with R, Python, SPSS, SAS etc.Experience with application of experimentation and statistical techniques Excellent written and verbal communication skills along with strong training skills. Apply Here Senior Associate – Data Scientist, Advisory, Performance Improvement at Ernst & Young, Malaysia Ernst & Young is one of the largest professional services firms in the world and is one of the “Big Four” accounting firms. The EY Data and Analytics team are specialists in information management, advanced analytics and business intelligence.  And right now, the firm is looking for talented, inquisitive and proactive Data Scientist to join its team. Prerequisites: A degree in a highly quantitative field such as Mathematics, Physics, Computer Science, Operational Research, Information Engineering (Postgraduate qualifications will be an advantage, but not compulsory)Experience  in formulating and solving business problems using mathematical\/statistical constructs Experience in solving problems using at least two of the following: Python, R, SAS, SPSS, Java, MATLAB.Familiar with at least one of data visualization development tools:  Tableau, JavaScript, D3Experience in performing Monte Carlo data simulationShould know dimensionality reduction and feature selection techniques.Experience in machine learning techniques Constrained optimization techniques such as integer programming, constraint programming, etc.Some knowledge of big data analytics tools (e.g. Spark, Storm),Implementation of algorithms in production, commercial environments (server-side), including object-oriented design. Apply Here Data Scientist at Fusionex – Petaling Jaya, Malaysia Fusionex is a data technology provider that specializes in Analytics, Big Data, Machine Learning and Artificial Intelligence. It helps clients manage, make sense of and derive useful insights and information from the vast amounts of structured and unstructured data at their disposal. At present, Fusionex is looking for an experienced Data Scientist to join the firm. As a data scientist his\/her job will include: scripting data models, automated data feeds, using BI tools to help visualize data, and various ad hoc projects. drive statistical analysis projects from beginning to end.obtain stakeholder buy-in and convince the audience on the quality of the delivered models to work collaboratively with DBAs, data scientists, and analysts.drive improvement in methodologies, systems and processes Prerequisites: Background in statistics, software engineering, data mining, database administration, or other quantitative disciplines is highly valued (Masters\/PhDs would be an advantage)Have a deep understanding of large dataStrong knowledge of statistics and data analysis with experience in machine learning techniquesStrong analytical skills and facility with Excel SQL knowledge Experience with statistical software (Matlab, SPSS, SAS, R) and Hadoop Experience with .Net, Java, HTML 5, Python, HiveQL knowledge Note: Candidate will have to undergo screening test (theoretical and practical) during the interview which topics are Statistics, Machine Learning, Programming and Apache modules. Apply here or submit your CV to recruitment@adv-fusionex.com Data Science Business Lead at Nielsen – Petaling Jaya, Malaysia A global measurement and data analytics company, Nielsen is powered by talented Data Scientist Business Leaders who come from diverse disciplines such as statistics, research methodology, mathematics, psychology, business, engineering, physics and demography. Currently, Nielsen is looking for Data Scientist Business Leaders for its Petaling Jaya, Malaysia office. As a Data Scientist Business Lead, your Job role will include: Meet with clients to understand business needs and offer innovative solutions Provide data-driven insights to help clients make the right business decisions.Analyze quality metrics, consumer products, user behaviour, advertisers, and publishers to improve products.Be the voice of the client for internal stakeholders including client service team, product leadership, and app dev teams.Work with operations teams and business stakeholders to define and communicate key insights. Prerequisite: Degrees in Mathematics, Data Science, Statistics, or related fields 5+ years of experience in market research or relevant fieldSolid understanding of CPG\/FMCG Industry and Retail market impact on client’s business within the marketExperience with Python\/R and knowledgeable with databases like Oracle(SQL) & contemporary data processing engines.Experience with data visualization tools (e.g. Spotfire, Tableau, etc.) and mapping tools.Working knowledge with collaboration tools: Google Suite — Google site, Google Data Studio, and Atlassian — Git,  Bitbucket, Confluence Apply Here","excerpt":"Year by year, the demand for data science skills is growing exponentially all across the globe, and Malaysia is not behind in the AI race. Data science is no longer an uncommon space in Malaysia, and organizations across have started to leverage data science and use its capabilities to fine-tune business operations — from improving […]","categories":["AI Hirings"],"tags":["Data Analytics","Data Science","Data Science Jobs"],"author_name":"Harshajit Sarmah","publish_date":"2019-03-14T09:36:09","publication_year":"2019","word_count":1029,"keywords":["data science","artificial intelligence","machine learning","AI","ML","Data Science Jobs","computer vision","RAG","NLP","Python","analytics","Data Analytics","Data Science"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","NLP","computer vision","data science","analytics","RAG","Python"],"url":"https:\/\/analyticsindiamag.com\/ai-hiring\/5-data-science-and-analytics-job-openings-in-malaysia-to-watch-out-for\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10120186,"title":"What to Expect from Google I\/O 2024","content":"Google I\/O 2024, the search giant’s flagship developer conference, will be held on May 14. Interestingly, it’s right on the day after OpenAI’s Spring Update, where the company is expected to announce a lot of exciting updates for developers and enterprises. And given the timing, I\/O this year becomes an extremely important event for Google’s AI. With almost every Google announcement over the past few years focussing on AI, this year’s event promises to be no different. Obviously, Gemini is likely to take centre stage, with Pixel phones also on the forefront. Another area that Google is expected to put a lot of focus on is its Responsible AI deployment, given the recent debacle with its image generation platform Gemini creating historically inaccurate images. The company later  apologised for “missing the mark” and CEO Sundar Pichai, in a recent interview, took full responsibility for it. A lot of AI There are reports of Google engaging in discussions with Apple to assist with generative AI features for iOS. At the same time, there are almost confirmed reports of OpenAI also partnering with Apple for similar features. If the Apple deal materialises with OpenAI, which might happen today, or we may have to wait until Apple’s WWDC in June for further details, Google might have to devise another strategy to improve its AI game while continuing its partnership with Samsung. Speaking of on-edge use cases, Google is also expected to announce Gemini Nano, its smallest AI model to compete with Microsoft’s Phi-3 or Meta’s Llama 3. The model has been in early access since its announcement in December 2023 and is running on Pixel 8 Pro, but it may be available for developers soon. A developer preview of Android 15 was released last month, so Google might offer a deeper dive into its next mobile operating system, featuring deep Gemini integration. Additionally, we can expect previews of new versions of Google’s other platforms, including Wear OS, Android TV, and Android Auto, along with new AI developer tools for different OS. One of the most important and loved features of Google’s generative AI has been on its Workspace platforms, such as Google Docs, Sheets, and others. The company has been constantly bumping up the capabilities of its products and might make further announcements about more capabilities on different platforms. Since there are rumours that OpenAI might release its AI agents at the conference, Google is also expected to do the same. Recently, Google introduced Vertex AI Agent Builder, a platform that enables the easy creation of autonomous agents with little or no coding required. A big focus on healthcare Google has also found its moat in healthcare. In January, Google Research launched a new medical chatbot called AMIE, which specialises in expert-level differential diagnosis. Unlike the big tech’s previous AI models, Med-PaLM 2, which focuses on medical summaries or answering questions, AMIE serves as a diagnostic tool, generating differential diagnoses. Meanwhile, Med-PaLM 2 is also expected to get a few updates. Moreover in April, researchers from Google and DeepMind also developed Med-Gemini, a new family of highly capable multimodal AI models specialised for medicine. The model achieved an accuracy of 91.1%, surpassing the previous best by 4.6%. In multimodal tasks, the models outperformed GPT-4 by an average of 44.5%. There is even a slight possibility of Google revealing an AI-developed drug at the event. Alphabet-backed Isomorphic Labs, along with DeepMind, recently released AlphaFold 3. The protein folding model predicts with 50% better accuracy. In a recent interview, Demis Hassabis said, “Well, if you ask me the number one thing AI can do for humanity, it would be to solve hundreds of terrible diseases. I can’t imagine a better use case for AI. So that’s partly the motivation behind Isomorphic and AlphaFold and all the work we do in sciences.” He believes that “revolutionising the drug discovery process to make it ten times faster” and more efficient and increasing the likelihood of passing clinical trials through better property prediction offers plenty of commercial value. Hassabis said that Google DeepMind will combine the agent systems it developed for gaming with multimodal systems into large general models that plan and achieve goals. “So systems that are able to not just answer questions for you, but actually plan and act in the world and solve goals… those are the things that will make these systems sort of the next level of usefulness in terms of being a useful everyday assistant,” he said. He further said that AI-designed drugs would probably be available in the ‘next couple of years’. Something for developers All these capabilities might make a big impact on Google’s integration of AI into its hardware devices at Google I\/O. To up the ante with its competitors, Google’s biggest bet might also be to increase more AI models on its Google Cloud and Vertex AI platform. According to Google’s page for the event, there are going to be several announcements about Google’s Cloud TPU and also accelerating its loads on PyTorch\/XLA, which the company developed in collaboration with Meta. Meanwhile, Google I\/O 2024 promises several exciting updates and announcements for app developers. We might hear about a new version of Android Studio called Arctic Fox. This would include improvements in data security and privacy features. New authentication methods such as face or fingerprint recognition for enhanced app security for Firebase are also expected to be announced. Flutter, gaining popularity for cross-platform app development, might showcase a preview of Flutter 3 with support for new features in Android 15 and iOS 17, including better widgets and notifications. When it comes to India, Google’s open-source model Gemma has been adored by developers for its Indic language capabilities. It is expected that Google may announce more updates to its AI models for building AI models in India. All in all, it would be interesting to see how Google matches up with the announcements that OpenAI is going to make at its event.","excerpt":"Since OpenAI is expected to soon launch search capabilities and also partner with Apple, this event is extremely important for Google’s AI.","categories":["AI Features"],"tags":["AI conference","Google"],"author_name":"Mohit Pandey","publish_date":"2024-05-13T15:48:41","publication_year":"2024","word_count":993,"keywords":["Go","TPU","AI conference","OpenAI","AI","PyTorch","autonomous agents","RAG","generative AI","multimodal AI","Google","R"],"extracted_tech_keywords":["AI","generative AI","multimodal AI","OpenAI","PyTorch","RAG","autonomous agents","TPU","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/what-to-expect-from-google-i-o-2024\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10129227,"title":"Fujitsu and Cohere Partner to Develop Japanese AI Models for Enterprises","content":"Fujitsu Limited and Cohere  have announced a partnership to develop and offer large language models (LLMs) tailored for enterprises. The collaboration aims to leverage Japanese language capabilities to enhance customer and employee experiences. The partnership will see Fujitsu become the exclusive provider of these services globally through Fujitsu Data Intelligence PaaS and Fujitsu Uvance. Fujitsu and Cohere will jointly develop “Takane” (tentative name), an advanced Japanese language model based on Cohere’s latest enterprise-grade LLM, Command R+. The model, which features retrieval-augmented generation (RAG) capabilities to reduce hallucinations, will be available starting September 2024 through Fujitsu Kozuchi AI services for private cloud environments. Takane is designed to meet the needs of industries requiring high security, including financial institutions, government agencies, and R&D units. The model combines Fujitsu’s expertise in Japanese language training and fine-tuning with Cohere’s enterprise-specific technologies. Cohere’s Embed and Rerank models will also be part of the offering, providing advanced enterprise search applications and RAG technology to unlock business value from data. Fujitsu plans to release additional AI technologies, including a knowledge graph extended RAG technology in July 2024 and generative AI auditing technology in September 2024, through Fujitsu Kozuchi. These technologies aim to convert large-scale data into knowledge graphs for LLMs and ensure compliance with corporate and legal regulations. “This partnership is a crucial step in offering world-class LLM capabilities to a key enterprise market. Our work with Fujitsu will unlock the potential of Cohere’s technology to empower the next generation of Japanese businesses,” said Cohere chief Aiden Gomez. Fujitsu’s generative AI amalgamation technology, available from August 2024, will be integrated with Takane and other domain-specific models to enhance AI capabilities for enterprises. The partnership between Fujitsu and Cohere aims to accelerate the adoption of AI in enterprises and promote digital transformation globally.Vivek Mahajan, CTO and CPO of Fujitsu Limited, emphasised the significance of combining Fujitsu’s knowledge graph and generative AI technologies with Cohere’s secure enterprise LLMs to provide adaptable AI solutions.","excerpt":"Fujitsu plans to release additional AI technologies, including a knowledge graph extended RAG technology in July 2024 and generative AI auditing technology in September 2024, through Fujitsu Kozuchi.","categories":["AI News"],"tags":["Cohere"],"author_name":"Siddharth Jindal","publish_date":"2024-07-16T09:49:44","publication_year":"2024","word_count":324,"keywords":["knowledge graphs","Go","AI","digital transformation","Git","RAG","Aim","generative AI","AI auditing","R","Cohere"],"extracted_tech_keywords":["AI","generative AI","Aim","RAG","knowledge graphs","R","Go","Git","AI auditing","digital transformation"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/fujitsu-and-cohere-partner-to-develop-japanese-ai-models-for-enterprises\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10040626,"title":"Implementing A Recurrent Neural Network (RNN) From Scratch","content":"Fully-connected neural networks and CNNs all learn a one-to-one mapping, for instance, mapping images to the number in the image or mapping given values of features to a prediction.  The gist is that the size of the input is fixed in all these “vanilla” neural networks. In this article, we’ll understand and build Recurrent Neural Network (RNNs), which learn functions that can be one-to-many, many-to-one, many-to-many. But what does that mean? They take as input sequences, such as speech, natural language, time series, or video. So when there’s a lot of information being conveyed sequentially, in the temporal change of the data, that’s where recurrent neural networks thrive. Source: MIT 6.S094 Slides Formulating the Neural Network Let’s take the example of a “many-to-many” RNN because that’s the problem type we’ll be working on. The inputs and outputs are denoted by x0, x1, … xn and y0, y1, … yn, respectively, where xi and yi are vectors with arbitrary dimensions. RNNs learn the temporal information with the help of a hidden state h, which is also a vector with arbitrary dimension. For any given time step, the hidden state ht is calculated using the previous hidden state ht-1 and the current input ht: This ht vector is then used to calculate the output yt: Here Wxh, Whh, and Why are the weight matrices for xt -> ht mappings,  ht-1 -> ht mappings and ht -> yt mappings, respectively. And bh and by are the bias vectors. Gradient Problems in RNN As powerful as recurrent neural networks are, they’re highly susceptible to gradient related problems in training. A network with n hidden layers will have n derivatives that will be multiplied together. If these derivatives are large, the gradient increases exponentially as it propagates backwards until it eventually explodes. This is called the problem of exploding gradient. Alternatively, if the derivatives are small, the gradient decreases exponentially as it is propagated until it eventually vanishes, and this is called the vanishing gradient problem. We’ll use the following “tricks” to help minimize the effect of these issues: Gradient Clipping- To avoid exploding gradients, simply limit the size of the gradients when the model is training.  The details of how gradient clipping works are beyond this article’s scope; you can read more about it here. Weight Initialization – Initializing the weights to identity matrices and biases to zero help prevent the weights from shrinking to zero. You can learn more about this here. Implementing a Recurrent Neural Network We will be building a character level prediction RNN and train in on the text of “Harry Potter and the Philosopher’s Stone” because why not. Let’s start by initializing the model parameters, weights and biases. import numpy as np import matplotlib.pyplot as plt class ReccurentNN: def __init__(self, char_to_idx, idx_to_char, vocab, h_size=75, seq_len=20, clip_value=5, epochs=50, learning_rate=1e-2): self.n_h = h_size self.seq_len = seq_len self.clip_value = clip_value self.epochs = epochs self.learning_rate = learning_rate self.char_to_idx = char_to_idx self.idx_to_char = idx_to_char self.vocab = vocab # smoothing loss as batch SGD is noisy self.smooth_loss = -np.log(1.0 \/ self.vocab) * self.seq_len # initialize parameters self.params = {} self.params[\"W_xh\"] = np.random.randn(self.vocab, self.n_h) * 0.01 self.params[\"W_hh\"] = np.identity(self.n_h) * 0.01 self.params[\"b_h\"] = np.zeros((1, self.n_h)) self.params[\"W_hy\"] = np.random.randn(self.n_h, self.vocab) * 0.01 self.params[\"b_y\"] = np.zeros((1, self.vocab)) self.h0 = np.zeros((1, self.n_h))  # value of the hidden state at time step t = -1 # initialize gradients and memory parameters for Adagrad self.grads = {} self.m_params = {} for key in self.params: self.grads[\"d\" + key] = np.zeros_like(self.params[key]) self.m_params[\"m\" + key] = np.zeros_like(self.params[key]) The loss function of SGD is noisy so we’re smoothing it, and we’re using AdaGrad to adapt the learning rate based on the parameters and data observed in earlier iterations. Create the functions for encoding the text characters and creating batches def _encode_text(self, X): X_encoded = [] for char in X: X_encoded.append(self.char_to_idx[char]) return X_encoded def _prepare_batches(self, X, index): X_batch_encoded = X[index: index + self.seq_len] y_batch_encoded = X[index + 1: index + self.seq_len + 1] X_batch = [] y_batch = [] for i in X_batch_encoded: one_hot_char = np.zeros((1, self.vocab)) one_hot_char[0][i] = 1 X_batch.append(one_hot_char) for j in y_batch_encoded: one_hot_char = np.zeros((1, self.vocab)) one_hot_char[0][j] = 1 y_batch.append(one_hot_char) return X_batch, y_batch Create the softmax method that takes the final logits and gives probabilities def _softmax(self, x): e_x = np.exp(x - np.max(x)) return e_x \/ np.sum(e_x) The maximum logit value is subtracted to improve numerical stability; you can learn more about this here. Implement the forward pass function that takes the current input sequence xt and previous hidden state ht to  calculates ht and yt def _forward_pass(self, X): h = {}  # stores hidden states h[-1] = self.h0  # set initial hidden state at t=-1 y_pred = {}  # stores softmax output probabilities # iterate over each character in the input sequence for t in range(self.seq_len): h[t] = np.tanh( np.dot(X[t], self.params[\"W_xh\"]) + np.dot(h[t - 1], self.params[\"W_hh\"]) + self.params[\"b_h\"]) y_pred[t] = self._softmax(np.dot(h[t], self.params[\"W_hy\"]) + self.params[\"b_y\"]) self.ho = h[t] return y_pred, h Create the function that backpropagates the error and calculate the gradients def _backward_pass(self, X, y, y_pred, h): dh_next = np.zeros_like(h[0]) for t in reversed(range(self.seq_len)): dy = np.copy(y_pred[t]) dy[0][np.argmax(y[t])] -= 1  # predicted y - actual y self.grads[\"dW_hy\"] += np.dot(h[t].T, dy) self.grads[\"db_y\"] += dy dhidden = (1 - h[t] ** 2) * (np.dot(dy, self.params[\"W_hy\"].T) + dh_next) dh_next = np.dot(dhidden, self.params[\"W_hh\"].T) self.grads[\"dW_hh\"] += np.dot(h[t - 1].T, dhidden) self.grads[\"dW_xh\"] += np.dot(X[t].T, dhidden) self.grads[\"db_h\"] += dhidden for grad, key in enumerate(self.grads): np.clip(self.grads[key], -self.clip_value, self.clip_value, out=self.grads[key]) return Function for update the parameters using AdaGrad def _update(self): for key in self.params: self.m_params[\"m\" + key] += self.grads[\"d\" + key] * self.grads[\"d\" + key] self.params[key] -= self.grads[\"d\" + key] * self.learning_rate \/ (np.sqrt(self.m_params[\"m\" + key]) + 1e-8) A test method that generates a sequence of characters of size test_size starting at a given index def test(self, test_size, start_index): res = \"\" x = np.zeros((1, self.vocab)) x[0][start_index] = 1 for i in range(test_size): # forward propagation h = np.tanh(np.dot(x, self.params[\"W_xh\"]) + np.dot(self.h0, self.params[\"W_hh\"]) + self.params[\"b_h\"]) y_pred = self._softmax(np.dot(h, self.params[\"W_hy\"]) + self.params[\"b_y\"]) # get a random index from the probability distribution of y index = np.random.choice(range(self.vocab), p=y_pred.ravel()) # set x-one_hot_vector for the next character x = np.zeros((1, self.vocab)) x[0][index] = 1 # find the char with the index and concat to the output string char = self.idx_to_char[index] res += char return res And finally, the training method that brings this all together. def train(self, X): loss = [] # trim end of the text so we only get full sequences num_batches = len(X) \/\/ self.seq_len X_trimmed = X[:num_batches * self.seq_len] # encode the characters to indices X_encoded = self._encode_text(X_trimmed) for i in range(self.epochs): for j in range(0, len(X_encoded) - self.seq_len, self.seq_len): X_batch, y_batch = self._prepare_batches(X_encoded, j) y_pred, h = self._forward_pass(X_batch) loss = 0 for t in range(self.seq_len): loss += -np.log(y_pred[t][0, np.argmax(y_batch[t])]) self.smooth_loss = self.smooth_loss * 0.999 + loss * 0.001 loss.append(self.smooth_loss) self._backward_pass(X_batch, y_batch, y_pred, h) self._update() print(f'Epoch: {i + 1}\\tLoss: {loss}') print(self.test(50,2)) return loss, self.params Now let’s our recurrent neural network in action. with open('Harry-Potter.txt') as f: text = f.read().lower() # use only a part of the text to make the process faster text = text[:20000] chars = set(text) vocab = len(chars) # creating the encoding decoding dictionaries char_to_idx = {w: i for i, w in enumerate(chars)} idx_to_char = {i: w for i, w in enumerate(chars)} parameter_dict = { 'char_to_idx': char_to_idx, 'idx_to_char': idx_to_char, 'vocab': vocab, 'h_size': 75, 'seq_len': 20,  # keep small to avoid diminishing\/exploding gradients 'clip_value': 5, 'epochs': 50, 'learning_rate': 1e-2, } model = ReccurentNN(**parameter_dict) loss, params = model.train(text) plt.figure(figsize=(12, 8)) plt.plot([i for i in range(len(loss))], loss) plt.ylabel(\"Loss\") plt.xlabel(\"Epochs\") plt.show() print(model.test(50,10)) is othe on. ogofostheodindearidut wlethallle, st oserarey d -lers amoathe y thasathey at dll tos dn t s med d.). t t ile brs t d g htherive, d ogostare d. ay shag hythay boumay tey thas ot havininggon Even with all our hacks and tricks the recurrent neural network still suffers from gradient problems and is only able to learn small sequences of characters. In the coming weeks, we’ll introduce more complex recurrent units with gates and try to improve the performance of our RNN. You can find the “Harry Potter and the Philosopher’s Stone” book text here. The above implementation has been made with a lot of help from this gist, and the code can be found in a Colab notebook here. Also, in hindsight, using the text of a book that contains spells and other non-English words might have made the task unnecessarily harder. References: MIT 6.S094 Lecture on RNNsHow to Avoid Exploding Gradients With Gradient ClippingAdaGrad paper","excerpt":"In this article we implement a character level recurrent neural network (RNN) from scratch in Python using NumPy.","categories":["AI Trends"],"tags":["python visualize neural network","Recurrent Neural Network","RNN"],"author_name":"Aditya Singh","publish_date":"2021-05-22T14:00:00","publication_year":"2021","word_count":1411,"keywords":["Recurrent Neural Network","NumPy","Go","TPU","AI","neural network","Git","Colab","python visualize neural network","RNN","CLIP","Matplotlib","R"],"extracted_tech_keywords":["AI","neural network","Colab","NumPy","Matplotlib","TPU","R","Go","Git","CLIP"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/implementing-a-recurrent-neural-network-rnn-from-scratch\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10097478,"title":"Amidst Manufacturing Woes, Apple Accepts Applications for Vision Pro Developer Kits","content":"Apple has recently unveiled the Apple Vision Pro developer kit, which is now available for developers to apply. Developers with innovative ideas for games or apps that can be played on Vision Pro can apply for the kit, said Apple. Apple will offer valuable guidance and support to developers by providing access to Apple experts for UI design and development, ensuring the app is refined to perfection. Additionally, developers will have the opportunity to have two additional code-level support requests. As part of the initiative, Apple Pro Vision labs are available in major cities such as Cupertino, London, Munich, Shanghai, Singapore, and Tokyo. To apply for the program one needs to be an Account Holder in the Apple Developer Program, provide details about their team’s development skills and existing apps, and agree to the terms and conditions. Applications will be reviewed and priority will be given to applicants creating an app that takes advantage of visionOS features and capabilities. Apple further provided a checklist of Vision Pro compatibility evaluations to make sure your visionOS, iPadOS, and iOS apps behave as expected on Vision Pro. Manufacturing Woes Earlier, Apple  announced a delay in the release of its “spatial computing” headset, priced at $3,500. Originally set for launch on June 5 with a promise of 1 million units before 2024, the headset will now be available for sale early next year. Analysts believe that the delay is mainly due to supply chain issues, rather than providing developers with more time to create apps for the Vision Pro. The production of displays, aimed at delivering lifelike images and seamless motion, is facing significant challenges as Apple demands flawless screens without any defects. In addition to that according to analysts, the Vision Pro has not made a significant impact on the headset supply chain in Asia. Eddie Han, an analyst at Taiwan-based Isaiah Research, noted that Apple’s product did not meet the expected standards, resulting in manufacturers having low confidence in its success.","excerpt":"Apple to loan Vision Pro developer kit to create apps for the launch of the new App Store on Vision Pro","categories":["AI News"],"tags":[],"author_name":"Siddharth Jindal","publish_date":"2023-07-25T11:54:46","publication_year":"2023","word_count":329,"keywords":["programming_languages:R","AI","ML","Aim","R"],"extracted_tech_keywords":["AI","ML","Aim","R","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/amidst-manufacturing-woes-apple-accepts-applications-for-vision-pro-developer-kits\/","complexity_score":2,"technical_depth":5,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10045893,"title":"Exploring DataPrep: A Python Library For Data Preparation &#038; EDA","content":"Big Data comes along with its complications. Collecting and managing data properly and the methods used to do so play an important role. With such underlying concerns, the method of Data Preparation becomes very helpful and a crucial aspect to begin with. A good data preparation procedure allows for efficient analysis, limits and minimizes errors and inaccuracies that can occur during processing, making the processed data more accessible to all users. Lately, the process of Data Preparation has gotten easier with new tools and technologies that enable anyone to cleanse and clarify data independently. Data preparation is the process of cleaning and transforming the raw data before preprocessing and analysis. It is a small yet important step before processing and often involves reformatting the data, making corrections, and combining multiple data sets to enrich the present data. Data preparation is often considered a lengthy undertaking for data professionals or business users, but it is an essential prerequisite to put data in context to turn it into insights that might help in decision making, eliminating the bias resulting from poor data quality. The data preparation process first begins with finding the right data. This can come from an existing data catalogue, warehouse or can be added ad-hoc. After the data is collected, it is important to discover and explore each dataset to prepare and process. This step is essential and helps get to know the data and understand what has to be done before the data can be called useful in a particular context. Cleaning up the data is traditionally another most time-consuming part of the data preparation process, but it’s crucial for removing forward and dealing with faulty data to help fill in the gaps present. Data cleansing is a process where you go through all of the data to be processed and either remove or update information that is considered to be incomplete, incorrect, improperly formatted, duplicated, or highly irrelevant. The Data cleansing process usually also involves cleaning up all the data compiled previously in one area. The data cleansing process is done all at once and can take quite a while if the information has been piling and stacking up for years. That’s why it’s important to perform data cleansing and taking care of the data regularly. Then, data is further transformed by updating the format or value entries to reach a clean and well-defined outcome or to make the data more easily understood by a wider spectrum of audiences. Once all the mentioned processes are done, the data is prepared. This data can be stored or inculcated into a third-party application, such as a business intelligence tool, clearing the way for processing and analysis. A thorough data preparation process can give an organization many advantages and rather give it a headstart. It must be clean and free of errors before using data for analysis or plugging it into dashboards for visualizations. Preparing data for analysis will help avoid mistakes, saving more time to be invested down the line. These errors will be much more difficult to catch and fix after the data has been transferred out of its original format. Using properly cleaned and formatted data while building data models or applications will ensure top-quality reporting and analysis with proper accuracy. This eventually helps receive game-changing and revolutionary business insights. What is DataPrep? DataPrep is an open-source library available for python that lets you prepare your data using a single library with only a few lines of code. DataPrep can be used to address multiple data-related problems, and the library provides numerous features through which every problem can be solved and taken care of. Using the DataPrep Library, one can collect data from multiple data sources using the dataprep.connector module, perform intense exploratory analysis using the dataprep.eda module and clean and standardize datasets using the dataprep.clean module. DataPrep automatically detects and highlights the insights present in the data, such as missing data, distinct count and statistics.  A whole detailed profile report can be created in a matter of seconds by using just a single line of code, which makes it ten times faster than other libraries to perform data preparation or EDA on. Getting Started with the Code In this article, we will be exploring the different functionalities of the DataPrep library for ease in Data Preparation and EDA, which will help us understand the library even better. The following implementation is partially inspired by the official DataPrep Documentation, which can be accessed using the link here. Installing the Library To install the library, you can use the following line of code, # Run the code to install DataPrep !pip install dataprep Installing the Dependencies Further, from the DataPrep library itself, we can import the required dependencies for the task to be performed, import pandas as pd from dataprep.eda import plot, plot_correlation, plot_missing As we want to create different plots from our dataset, we have imported plot, plot_correlation to create correlation graphs,plot_missing to plot the number of missing data. Now, Lets load our data into the data frame, #loading data in to the DataPrep dataframe df = pd.read_csv(\"\/content\/master.csv\") Creating Visualizations Using DataPrep, we can create all the possible visualizations for the data using just a single line of code. Let us plot our loaded data and see what it looks like, df[\"year\"] = df[\"year\"].astype(\"category\") plot(df) As we can see, the library itself detects the data and plots all the necessary data graphs in a single-window itself! Auto Insight Generation You can also get a detailed plot for a single column with all its statistics to understand the column better, plot(df, \"gdp_per_capita ($)\") We also create a dataframe, taking care of the missing values and then create a correlation plot, df_without_missing = df.dropna('columns') plot_correlation(df_without_missing) plot_correlation(df_without_missing, k=1) plot_correlation(df_without_missing, value_range=(0,1) Whether it be Pearson, Spearman or Kendall-Tau, any correlation graph can be easily plotted using the DataPrep library. More Operations with DataPrep Let us now explore some more operations on another dataset. Here I have used the titanic dataset to perform further functional operations on using DataPrep. Exploring the missing data in the dataset, #plot missing plot_missing(train_df) DataPrep will automatically analyse the data and provide the necessary graph such as a Bar Chart, Heat Map or a Dendogram. Creating Word Clouds have never been easier as well, such functionalities can be used in NLP tasks which will be highly helpful. plot(train_df,'name' ) To further analyse and understand what necessary steps would be needed to take on the loaded dataset, we can generate an instant report of the dataset in one go, which will provide us with all the necessary information and metrics to analyse where the focus during the data preparation stage must be, particular columns if any. #creating a full report create_report(train_df) The detailed statistics for each column is generated with options to perform interactions between columns, check correlation or plots of missing values. End Notes In this article, we understood the importance of Data Preparation in Big Data Analytics and the necessary steps required to do so. We also explored a library known as DataPrep, and tested its different functionalities that might help during the Data Preparation and EDA phase. Although there is still a lot more the DataPrep library can do, I would recommend encouraging the reader to explore further and understand the library’s immense power. The following implementations above can be found as Colab notebooks in two separate notebooks. You can access them using the links here: Notebook 1 – Titanic Dataset Notebook 2. Happy Learning! References PyPi DataPrepOfficial DataPrep Website","excerpt":"DataPrep is an open-source library available for python that lets you prepare your data using a single library with only a few lines of code. DataPrep can be used to address multiple data-related problems, and the library provides numerous features through which every problem can be solved and taken care of.","categories":["AI Trends"],"tags":["AI (Artificial Intelligence)","big data quality","data preparation","Data Science","Data Scientist","Deep Learning","how to measure twitter influence","load data python","Machine Learning"],"author_name":"Victor Dey","publish_date":"2021-08-14T10:00:00","publication_year":"2021","word_count":1255,"keywords":["big data quality","API","R","Pandas","load data python","how to measure twitter influence","RAG","NLP","analytics","Data Science","Go","AI","Machine Learning","data preparation","Colab","Python","Deep Learning","Data Scientist","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","NLP","analytics","Colab","Pandas","RAG","Python","R","Go","API"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/exploring-dataprep-a-python-library-for-data-preparation-eda\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10020462,"title":"Data Science Hiring Process At Hansa Cequity","content":"Mumbai-based Hansa Cequity is a data-driven, technology-enabled marketing and consulting company servicing business verticals such as retail, BFSI, automobile, media, etc. The data science team plays a critical role in solving business challenges of varying scale and complexity at Hansa Cequity. The centralised data science team is aligned horizontally. We got in touch with Neeraj Pratap, Chief Operating Officer of Hansa Cequity, to understand their data science team’s hiring process. Required Skills Logical, agile and out-of-box thinkers get preferential treatment at Hansa Cequity. The candidates should be open to experimentation, have an appetite to explore new techniques and adapt to new tools and technologies. “Someone who is from a strong academic background with fundamental knowledge coupled with a few years of experience is the best fit for a data scientist position in our organisation,” said Neeraj. “Technically they should have a comprehensive understanding of machine learning algorithms and knowledge across different specialisations such as operational research, NLP, etc. We emphasise exposure to big data technologies and thorough knowledge about end-to-end data science project lifecycle,” he added. “A candidate is expected to blend behavioural data with demographics, transactional data to examine the “whats” and “hows” of customer behavioural data to inform the “whys” of customer behaviour,” he said. Additionally, they should have a hacker’s spirit and appetite to experiment and learn. The educational qualifications Hansa Cequity looks for are bachelor’s in computer science, engineering or Masters in statistics, economics, and mathematics. Neeraj said a strong educational background with analytics skills, designing and building statistical modelling, visualisation, machine learning is highly desirable. Required skills: Statistical software (e.g. MATLAB, SAS, SPSS etc.)Programming language like Python, R, C++, and\/or JavaAdvanced experience in database language (e.g. SQL or NoSQL)Experience in management systemsDesigning and building statistical modellingVisualisation toolsStrong problem-framing and problem-solving capabilityUnderstanding business objectives Hansa Cequity follows both the traditional hiring approach through job portals, and non-traditional hiring approaches through LinkedIn, referrals, and data scientist communities in social media platforms. They are currently looking for a mid-level manager who will deliver and maintain multiple predictive modelling projects with the help of junior data scientists. A candidate can contact the HR through LinkedIn or the company website. Being A Data Scientist At Hansa Cequity At Hansa Cequity, data scientists should be good at data intuition. Neeraj believes that data-intuitive candidates are excellent at identifying patterns within sets of structured and unstructured data. Neerja highlights the ideal traits of a data scientist at Hansa Cequity: Data scientists must enjoy the iterative development process. They should always be in search of finding an optimal solution through the process of iterative design.A solid background in data science and big data analytics with good statistical thinking is always a plus.Willing to work in a new or unfamiliar situation, learn new skills and provide solutions to business problems.Data science candidates should be creative in exploring behavioural sciences techniques and combining it with business and applying computational approaches from computer science, statistics, and data-centric engineering to better model, understand and predict behaviour. Ability to blend internal and external (macro-economic) data to make sharper predictions to solve business problems. Neeraj also said the data scientist is expected to be in charge of data analysis — from data collection to making business decisions. He should work closely with data analysts, data engineers, business intelligence specialists, and data architects to create and maintain databases, analyse data, communicate business insights, build predictive and prescriptive analytics models. He\/she should identify problem areas, and turn problems into opportunities using data science. Growth Opportunities Since the analytics team is centralised, data scientists get an opportunity to work and deeply understand multiple domains. “We continually invest in development by providing access to technical\/management development programs. Further, we support and coach employees to think broadly, develop a proactive consulting mindset, and encourage them to participate in various employee engagement programs,” said Neeraj. A data scientist with more than two years of experience gets ample opportunity to engage directly with clients, discuss business problems and suggest ways to address them. Interview Process The interview process at Hansa Cequity is designed to assess a candidate with quantitative aptitude, programming skills, knowledge in statistics & machine learning, analytics tools, real-life scenario-based case studies, soft skills & cultural fit. The candidate selection process involves four rounds of interview. It usually starts with a resume evaluation based on the experience. Neeraj explained the interview process: Microsoft Teams\/Zoom Interview Focuses on: logical reasoningquantitative aptitudeprogramming skill assessments (Python\/R\/SQL, etc.)statistics knowledge and machine learning knowledge Model Assignment & Case presentation Assess a candidate’s modelling skills, logical reasoning, business problem-solving ability, programming comfort, technical know-how, presentation skills, ability to comprehend and communicate the business problem. Personal Interview HR personnel will evaluate the cultural fit, compatibility with team, communication skills, etc. Management Round Shortlisted candidates interact with senior leadership for a comprehensive analysis of capabilities and skills. Interview Questions Questions are asked based on the typical interview focus areas: Quantitative aptitude: To measure the candidate’s numerical ability, problem-solving and mathematical skill. Questions can range from purely numeric calculations to critical arithmetic reasoning. Candidates can expect questions similar to the Common Aptitude Test(CAT), MAT, GMAT, etc. Programming Skills: Candidates are asked to write pseudo-codes, or syntax related to Python, R, SQL, etc. Statistics & Machine Learning Skills: Conceptual questions related to statistics, algorithms and Machine Learning Techniques. Case Presentation: Ability to structure solutions to open-ended problems. “Candidates who possess the skills of problem-solving, have a decent knowledge of algorithms and know their basics of statistics and machine learning stand a higher chance of cracking these interviews,” said Neeraj. Recruiting Challenges It is usual for companies to focus on questions that are technical rather than emphasising client engagement skills. However, it is equally important to find a solution to a business problem and communicate the same at the right time and in the proper manner to make a decision based on the analysis. Setting the right expectation to the data scientist candidate during the hiring process is very important. Candidates tend to think their primary role is to analyse data, extract insights, build a predictive model, and fail to understand data collection is also a big part of the process. “At Hansa Cequity, our interview process is well aligned to avoid these mistakes. We make the candidate aware of the kind of activities that is expected from them,” said Neeraj. That said, Hansa Cequity face challenges in finding all-round candidates. “We look for someone who can solve business problems and play with data using different tactics,” said Neeraj. Another challenge is landing the candidate as they may have multiple job offers. To become a competitive data scientist, Neeraj advises to be good at numbers, understand business challenges, develop storytelling ability, improve coding skills, work on communication skills and be ready to evolve with the fast-moving analytics industry.","excerpt":"Mumbai-based Hansa Cequity is a data-driven, technology-enabled marketing and consulting company servicing business verticals such as retail, BFSI, automobile, media, etc. The data science team plays a critical role in solving business challenges of varying scale and complexity at Hansa Cequity. The centralised data science team is aligned horizontally. We got in touch with Neeraj […]","categories":["AI Hirings"],"tags":["data science master","data scientist qualifications","master data science","Masters in Data Science","qualifications for data scientist","Statistics for Data Science"],"author_name":"Srishti Deoras","publish_date":"2021-02-18T15:00:00","publication_year":"2021","word_count":1133,"keywords":["data science","Go","machine learning","data scientist qualifications","AI","Masters in Data Science","data science master","qualifications for data scientist","NLP","RAG","Python","master data science","Statistics for Data Science","analytics","SQL","R"],"extracted_tech_keywords":["AI","machine learning","NLP","data science","analytics","RAG","Python","R","SQL","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-hiring\/data-science-hiring-process-at-hansa-cequity\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10074449,"title":"Artists&#8217; Envy, AI&#8217;s Pride","content":"Recently, an artwork created on Midjourney by Jason Allen titled, “Theatre d’Opera Spatial”, won the first prize at the Colorado State Fair in the category of “Digital Arts\/Digitally-Manipulated Photography”. Though the artist gave adequate credits—Jason Allen via Midjourney—many of the judges were unaware about the AI-driven art generating tool. This has resulted in major backlash and heavy criticism from the digital artists community while others have argued that AI is merely a tool just like any other tool that artists use. To understand why there is a discussion in rounds about the competition, one has to understand how midjourney uses a database to create\/generate images. There has been a rise in text-to-image tools recently like DALL.E 2, StableDiffusion, and DreamBooth. https:\/\/twitter.com\/RakotaCivet\/status\/1565331359870189568 Several users on Twitter argue that the generated art won in the category of “digital art”, which itself is a category heavily dependent on tools like Photoshop or Illustrator and which helps artists draw straight lines, objects or create compositions. https:\/\/twitter.com\/DwarvenPolitics\/status\/1565771972444884998 Artists take a stand Artists took to Twitter and Reddit in discussions about how art is more about the “process” than the “result”. “Would you consider yourself an artist for deciding if the outcome of your text prompt is worthy of a submission in the contest?” said an artist on Reddit. Another Reddit user pointed out that people comparing a Midjourney generated image to digital artists creating on Photoshop is a false equivalent. “In an AI-generating tool, you do not have control over the output image unlike Photoshop or Illustrator, where the artist has to decide where to put the objects and decide the composition.” “The generated image though is an art, but the person who typed the words into the AI is not an artist,” said artist Chris Shehan. The decision-making part in an AI generated art is outsourced since there is no human cognitive process involved, apart from writing the prompts and being able to tweak it further with just texts. https:\/\/twitter.com\/ChrisShehanArt\/status\/1564821938400264192 A common argument put forth by several artists is that artworks created using AI are based on millions of images on the internet. The question of plagiarism pops out when delved deeper into the working of the topic. A user on Reddit said, “When this text-to-image tech extends to text-to-music, and people prompt an ai with—a love song in the style of taylor swift,” are they then musicians? All artists use tools “Should you give credits to the person who created the paint brush or the artist for the painting?” said a Twitter user. Everyone has the tools to create art—be it paint, Photoshop, camera, or in this case, Midjourney. The person who inputs his imagination into the tool deserves credits to be the artist. When moving out of Midjourney into the broader digital art category, we find that artists have a lot more control over the composition and style of their art. The artist Andy, or Andrew, explains the process of creating art using stable diffusion. You can check out the process here. Tweaking and modifying the basic structure of a generative tool is also a technique that a user has to learn before creating art. Jason Allen, the creator of the image, said that he is glad that the image is starting a conversation around the use of AI in generating art. He further added that, “Rather than hating this AI technology, we need to recognize its power and expand on it.” “Banana taped to the wall” of the digital art world Cal Duran, one of the judges of the competition called Allen’s art a “beautiful piece”. Duran further added that the piece mentioned Midjourney but he couldn’t pinpoint if the result was generated by an AI. Since the inception of artificial intelligence, the discussion about whether the jobs of millions of people in various fields will get replaced has been gaining momentum. An architect on a Reddit thread talks about how his job will be replaced by construction tools in 10–15 years. “Right now, it feels like you are riding a majestic but wild horse—amazing results beyond one’s talents but all over the place without much control.” https:\/\/twitter.com\/OmniMorpho\/status\/1564782875072872450 The famous “Banana taped to the wall” sculpture by Maurizio Cattelan sparked a similar controversy in the modern art world. Is art related to only the final output or does the process also play a crucial role in considering something an art and the creator an artist? Zulban on the same Reddit thread said that judges have rules to not allow a photograph in a painting competition because the outcome is different owing to the use of different tools. “Photography is art, but it shouldn’t be an entry in a painting or drawing contest. I look forward to AI arts contests in the future.”","excerpt":"Artists have been furious about an AI generated image winning an art competition but the internet is still in splits.","categories":["IT Services"],"tags":["AI (Artificial Intelligence)","ai art","MidJourney","photoshop","reddit","Stable Diffusion","Twitter (X)"],"author_name":"Mohit Pandey","publish_date":"2022-09-06T18:00:00","publication_year":"2022","word_count":791,"keywords":["Twitter (X)","Go","MidJourney","artificial intelligence","TPU","Banana","AI","Stable Diffusion","reddit","programming_languages:R","Git","stable diffusion","AI art","ai art","photoshop","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","TPU","R","Go","Git","stable diffusion","Banana","AI art","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/it-services\/artists-envy-ais-pride\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10089749,"title":"OpenAI Might Invite Legal Trouble","content":"The issue of copyright and assigning credits to the original developer has been discussed extensively ever since the release of GitHub Copilot, which is powered by OpenAI Codex. Now, GitHub and OpenAI are taking steps to avoid any further legal tussle in this context with OpenAI announcing that they are going to discontinue support for Codex API starting from March 23, 2023. To avoid much scrutiny about this heavy move, the company said that given the advancement of their GPT-3.5 model, supporting the older model would be no longer useful for them or the customers. Therefore, all customers are “encouraged to transition to GPT-3.5-Turbo”, which is a much more cost effective and performant model in the GPT-3.5 family, according to OpenAI. But what about the apps that are currently built on top of Codex API? What’s The Reason? Codex has been trained on billions of publicly available lines of code, which includes the public repositories on GitHub. This eventually led to investigations on the possibility of filing a copyright claim against GitHub, which led to a class action lawsuit against Microsoft, GitHub, and OpenAI for scraping licensed code for building AI-powered Copilot tool. Now, with GPT-4 in the picture, there is a high possibility that GPT-4 might get integrated into Copilot. The legal issue might go away but the billions of parameters that GPT-4 is trained on—we don’t know how many yet—might still run into the same problems. OpenAI and Microsoft are seemingly pushing their customers to use GPT-4. Microsoft also integrated GPT-4 to Office 365 recently. This also means that all the technology that the two companies have been developing in partnership might converge into one large multimodal system. The capabilities of GPT-4 already indicate that. Given that people were already shifting towards GPT-3.5 from Codex for many tasks, the capabilities of GPT-4 on Copilot would make it the best choice for developers, even for those who do not know how to code, while avoiding all the copyright issues. Seems like a well thought out and smart move by Microsoft and OpenAI. But even then, just a three-day notice before shutting down the service looks like a rushed move from OpenAI’s end. For this, the company has been getting some criticism as well. Crowd Control The models that would be affected by this move are code-cushman:001, code-cushman:002, code-davinci:001, and code-davinci:002. Thus, these models will also be discontinued. This can be quite cumbersome for a lot of apps that utilise the model for their workflow—all the data might have to be regenerated, thereby rendering the older ones useless. Amid the discussion on HackerNews, users have been pointing out several aspects of this move. A user points out that the migration from one to another is easy and, thus, even though someone might question the legality of doing this, it is acceptable by the community. Moreover, the terms of services of Codex clearly state that it was released as “free limited Beta”. The companies or apps that are relying on it might have to choose a different path now. Conversely, however, Beta implies that it is expected to improve and not vanish overnight. What this means is that all customers, including the ones who have already paid for the service and are using the API are expected to transition to a different model. To this, the company says that they understand the inconvenience but the move is to increase investment in their latest and most capable models. This definitely signifies that the company is brewing up something new. What if it is just a push towards GitHub Copilot? There is no doubt that the research behind GPT-4 was emphatically supported by Microsoft, which also owns GitHub. Copilot was being powered by OpenAI Codex for suggesting code. Now, they can just shift to GPT-4 and ditch Codex. This might actually be a win against the legal trouble that the company has been going through—using other developers’ code. That would probably be ideal for both the companies, presuming that GPT-4 does not go through the same legal issues and is not trained in a manner similar to Codex, which is highly unlikely. What Can Researchers Do For Now? Interestingly, this announcement is made by just OpenAI and not Microsoft. Codex might still be available via Azure services. But OpenAI’s recent pricing was almost ten times lower than what Azure offers. Moreover, Codex has been free all this while. This means that shifting to Azure might not be the first thought that comes to developers’ and researchers’ minds. The news is perhaps the worst for research groups. A lot of papers that were based on Codex models are now going to be rendered completely irreproducible. A lot of researchers might not have the capability of shifting to Microsoft’s Azure services and even if they do, they might not want to pay for what they were accessing for free all this while. Some users on Twitter pointed out that the release of GPT-3.5 had actually made them shift to it, completely moving away from Codex. Maybe the traffic for Codex was actually dropping as the other services were producing the same, if not better, output.","excerpt":"OpenAI discontinues Codex. A well thought out, smart but rushed move.","categories":["AI Highlights"],"tags":["Github Copilot","GPT-4","OpenAI"],"author_name":"Mohit Pandey","publish_date":"2023-03-21T14:53:16","publication_year":"2023","word_count":860,"keywords":["Go","TPU","OpenAI","AI","AWS","R","GPT-4","Github Copilot","RAG","Git","Aim","Azure"],"extracted_tech_keywords":["AI","OpenAI","Aim","RAG","AWS","Azure","TPU","R","Go","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/openai-might-invite-legal-trouble\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10047978,"title":"Why Ransomware Hackers Love Holidays","content":"Just a week before the three day Labour Day weekend holiday, the FBI issued a warning, alerting the citizens to be precautious against ransomware hackings. The issue cited incidents to prove that some of the most successful cyberattacks occur during holidays. Just take a look at the past few months; May saw two significant ransomware attacks over the holiday season, during Mother’s Day and the Memorial Day weekend. The former that is suspected to be associated with DarkSide paralysed the infrastructural entity in the energy sector, leading to extortion and a week-long halt in operations. On Friday, heading into Memorial Day weekend, the Food and Agricultural Sector was hit and suffered a complete production shutdown. It’s easy to say; hackers seem to love holidays. A set of days with everyone out enjoying with their families and being offline? They probably won’t even realise they have been hacked until a few days later. Hackers have been around for years, and holiday hacking isn’t a new trend. The FBI’s warning sure makes it a serious one. According to CyberRisk Insurance company, Coalition, an increasing amount of hackings are being done using a banking trojan, Trickbot, that installs ransomware. As of July 2019, Trickbot is estimated to have harvested over 250 million IDs and passwords from users opening an attachment or clicking a link. Ransomware – the Monster in your Closet Trickbot first enters a company network through employees opening an infected email attachment or clicking a malicious link from unreliable sources. It goes on to hijack the user’s email account and send the same malicious documents and links to the user’s entire contact list, spreading it as fast as possible through email. Next, the malware secretly installs a program on the infected computers and connects it back to a “command and control” centre that the hackers have complete control over, and by that, the computers. They can further use this to install ransomware, access email, or steal the personal\/banking information of the users. Finally, when all the damage preparations have been completed, the hackers spring the tap and activate the ransomware. Ransomware takes time to disseminate throughout a network, and hackers have maximum control over most systems. The holiday distraction means people won’t notice hacking quickly, giving the culprits more time to do all the damage they can fit in. It also means that people will be slow to deal with it since it will be more challenging to hold security than it would on any working day. How dangerous is it? The July 2 ransomware attacks on US cybersecurity is attributed to the REvil gang, a major Russian-speaking ransomware syndicate. In fact, according to cybersecurity researchers, targeting customers of Kaseya could be one of the broadest ransomware attacks on record. The attack affected thousands of victims in at least 17 countries with ransom demands between $45,000 and $5 million. Unfortunately, many victims were busy celebrating the weekend by the poolside and did not even realise they had been hit until things reopened on July 5 or 6. According to the FBI’s Internet Crime Complaint Center (IC3), 2020 saw a record number of 791,790 complaints about internet crimes, with reported losses exceeding $4.1 billion. This is a 69 per cent rise from complaints in 2019. In fact, 2,474 of the complete IC3 reports were malware, a 20 per cent increase from 2019. How to run from it Preparation from potential virus attacks works like a vaccine. The point isn’t to prepare prevention one day before a holiday – the chance is, the malware has already been installed on your computer, ready to strike during your Sunday brunch. Instead, you have to prepare a defence against viruses way before the ransomware hits. In most cybersecurity circumstances, the FBI and CISA’s guidelines are the best practices: Don’t click on suspicious links.Make an offline backup of your data. Use strong passwords to secure your user accounts.Make sure your software is up to date, and scan it for vulnerabilities.Use two-factor authentication. If you use Remote Desktop Protocol or other well known risky servers — proceed with caution. Lastly, don’t forget to check up on your computer even during holidays. Remember, holidays, though preferred, aren’t the only time for ransomware attacks.  Hence, make sure to keep your digital life as safe and protected as your personal lives!","excerpt":"Ransomware takes time to disseminate throughout a network, and hackers have maximum control over most systems.","categories":["AI Trends"],"tags":["Ransomware"],"author_name":"Avi Gopani","publish_date":"2021-09-09T12:00:00","publication_year":"2021","word_count":720,"keywords":["Go","programming_languages:R","AI","ETL","Ransomware","programming_languages:Go","Git","GAN","R"],"extracted_tech_keywords":["AI","R","Go","Git","ETL","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/why-ransomware-hackers-love-holidays\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10140254,"title":"Tencent Launches Hunyuan Large, Outperforms Llama 3.1 70B &amp; 405B","content":"Chinese giant Tencent has just released a large 389 billion parameter open source model called Hunyuan Large – with 52 billion active parameters. The model supports a context length of 256,000 tokens and is one of the largest open-source models in its category. In comparison, both Llama 3.1 70B and 405B models support a 128,000 context length. Tencent has made the code, and the model available on GitHub and Hugging Face. If you’re competing in the open-source arena, you’ve got to dethrone the king. Interestingly, the Hunyuan-Large outperforms the Llama3.1 70B model on several benchmarks in English and Chinese. The model’s performance was also comparable with Meta’s flagship Llama 3.1-405B model on tasks involving language understanding, coding, maths and logical reasoning. Unlike the Llama 3.1 405B, Hunyuan Large isn’t a ‘dense’ model. This means that it doesn’t use all of its parameters for each input. Tencent explores Mixture of Experts (MoE) scaling laws to guide an optimal balance between model size, data volume, and performance. An MoE model activates only a subset of parameters based on its input. This makes it more efficient as it only uses a part of the model’s capacity each time. Hunyuan Large incorporates several ‘innovative’ techniques to outperform its competition. This includes using 1.5 trillion tokens of higher-quality synthetic data, which is part of the 7 trillion parameters that the model is trained over. The model also incorporates various model structure enhancement techniques to reduce memory usage, increase performance, and balance token usage. Tencent compared Hunyuan Large against leading open-source models in both pre-and post-training stages. In most results, Hunyuan Large came out on top in comparison with other dense, and MoE models with similar parameter sizes. The authors mentioned, “We also hope that the release of the largest and overall best performing MoE-based Hunyuan-Large could spark more ripple of debate about more promising techniques of LLMs among the community, in turn, to further improve our model from a more practical aspect and contribute to the more helpful AGI in the future.” Tencent’s latest announcement comes after the news that China has adopted Meta’s open-source models for building a chatbot for military applications. This ensued a debate between Vinod Khosla and Yan LeCun, with the former criticising Meta for providing ease of access to LLMs. LeCun retaliated that China is quite competent with the United States in generative AI, and they wouldn’t entirely depend on Meta’s open-source models to develop any consequential technology. With the release of Hunyuan Large, Yann LeCun may just be right. Interestingly, Meta has also announced that it is making Llama available to the US government, and any other private organisations working in the interests of national security.","excerpt":"Hunyuan-Large is China’s competitor to Meta’s Llama. It outperforms Llamma 3.1 70B and is on par with the flagship Llama 3.1 405B.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","Llama 3","Tencent"],"author_name":"Supreeth Koundinya","publish_date":"2024-11-05T17:13:52","publication_year":"2024","word_count":447,"keywords":["Llama 3","Go","Hugging Face","synthetic data","AWS","AI","Tencent","Git","generative AI","GAN","GitHub","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","generative AI","Hugging Face","AWS","R","Go","Git","GitHub","GAN","synthetic data"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/tencent-launches-hunyuan-large-outperforms-llama-3-1-70b-405b\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":59897,"title":"Self-Driving Dream Eludes Yet Again As Starsky Robotics Shuts Down","content":"Last week, Starsky Robotics, a San Francisco based company that makes driverless trucks, announced that it is shutting down its operations as they have failed to find investors. Starsky’s CEO, Stefan Seltz-Axmacher, explained in his blog, which led to this decision in spite of the company showing promising results in the past. Surprisingly, in November 2019, the company also appeared in CNBC’s list of the world’s most promising startups. Three months later, Starsky Robotics doesn’t exist anymore. Like Shackleton on his expedition to Antarctica, we did things no one else ever has. Similarly, though, it didn’t turn out as planned.CEO, Starsky Robotics What Went Wrong Stefan, CEO, Starsky Robotics Starksy’s co-founder, in his blog, blamed inappropriate timing as the reason behind their failure. Although he still lauds his team for their efforts, he lamented that the self-driving space was overwhelmed with the unmet promise of AI to focus on a practical solution. Added to this is the recession in trucking for the past 18 months. The biggest, however, is that supervised machine learning doesn’t live up to the hype – CEO, Starsky Robotics The hardest part of building AI is how it deals with situations that happen uncommonly, i.e. edge cases. In fact, the better your model, the harder it is to find robust data sets of novel edge cases. Additionally, the better your model, the more accurate the data you need to improve it. Stefan also pointed out that the competitors of Starsky Robotics and other driverless vehicle companies have been adding more lucrative features to their portfolio with every funding showing how good the models can perform. Whereas, Stefan and his team spent almost three years trying to figure out the safety engineering — the unattractive part. The biggest limiter of autonomous deployments isn’t sales, it’s safety. Safety engineering, stresses the CEO, is really, really hard. And, this is what his company focused on from September of 2017 until the unmanned run in June of 2019. We documented our system, built a safety backup system, and then repeatedly tested our system to failure, fixed those failures, and repeated it again. “It took me way too long to realise that VCs would rather like a $1 billion business with a 90% margin than a $5 billion business with a 50% margin, even if capital requirements and growth were the same,” wrote Stefan in his blog. The co-founder also hinted at a tinge of practicality involved getting driverless vehicles on roads. He wrote that the trucking companies aren’t tech-savvy enough and even Starsky Robotics manages to perfect the general autonomy and perfectly validated safety, it would take years to deploy sufficient systems to make the necessary profits. Imagine pitching a 10-year no-profit product in R&D to an investor who has other lucrative options. The results are quite obvious. The Elusive Dream Stefan, the co-founder of the ill-fated Starsky Robotics, explained further, how challenging it is to make a driverless car mainstream. As the machine learning models are burdened by the never-ending edge cases, there are also levels of autonomy that need to be fulfilled. As illustrated above, L1 is the line of human equivalence, then leading AV companies merely have to prove safety to be able to deploy. If L2 is the case, asserts Stefan, the bigger teams are somewhere from $1–25 billion away from solving this problem. If, however, L3 is the line of human equivalence, it’s unlikely any of the current technology will make that jump. Whenever someone says autonomy is 10 years away, that’s almost certainly what their thought is. There aren’t many startups that can survive 10 years without shipping, and Stefan firmly believes that almost no current autonomous team will ever ship AI decision-makers if this is the case. Like the captain of a sinking ship, I’ve gotten most of the crew on lifeboats, and I am now noticing the icy waters at my ankles while I start to think about what I do next.Stefan The strategy taken by the founders of Starsky might be wrong or wrongly timed, but Stefan urges others to take his company as an example of what can be established in a short time with laser-like focus. While it is another grim tale of failing dreams, data suggests that other companies are doing extremely well, at least for now. According to Crunchbase, Waymo, which is owned by Alphabet, raised $2.25 billion earlier this month whereas Cruise raised roughly $1.15 billion. We are talking billions here, and Starsky Robotics couldn’t convince its investors for a $20 million funding. Though the concept of self-driving cars is more than a century old, we had to wait up to the late 1980s to see the first working model. However, yesteryear’s computer vision models are not as good as the current state of the art models. A huge sum of money, a right strategy and of course timing, might be a recipe for the realisation of self-driving vehicles. But, the genie is out of the bottle, and it is only a matter of time before we see major breakthroughs. The co-founder of NNAISENSE, Juergen Schmidhuber, also shares the same opinion with regards to self-driving cars. “As soon as autonomous vehicles are proven to be 10 times less prone to accidents than others, legislators will feel the pressure to allow people to drive only in exceptional situations. In some countries, this will happen sooner than in others, but this process can no longer be stopped,” said Dr Schmidhuber when asked about the future of autonomous vehicles.","excerpt":"Last week, Starsky Robotics, a San Francisco based company that makes driverless trucks, announced that it is shutting down its operations as they have failed to find investors. Starsky’s CEO, Stefan Seltz-Axmacher, explained in his blog, which led to this decision in spite of the company showing promising results in the past. Surprisingly, in November […]","categories":["Deep Tech"],"tags":["Self Driving Cars","Startups"],"author_name":"Ram Sagar","publish_date":"2020-03-25T13:00:22","publication_year":"2020","word_count":920,"keywords":["Go","API","funding","machine learning","ELT","AI","ML","Self Driving Cars","computer vision","Startups","R","startup"],"extracted_tech_keywords":["AI","machine learning","ML","computer vision","R","Go","API","ELT","startup","funding"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/self-driving-dream-eludes-yet-again-as-starsky-robotics-shuts-down\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":530,"title":"Data Visualization &#8211; A POV from Gramener","content":"Consider this price and sales table for four cities: Can you figure out how each city is performing? Notice that the average for each city is the same. Now take a look at the same data, plotted. The patterns are a lot clearer now, and you can quickly see that: The four cities are completely different in behavior and need different strategies for growth. That Delhi is price sensitive, while Bangalore and Hyderabad are not There is at least 1 data point each at Hyderabad and Mumbai that look like aberrations This is an example of the issue industry faces today: significantly larger quantities of data, but still visualized as plain tables. [divider top=”1″] This is an anonymised version of our very first visualization. We were working with a leading mobile operator who provided us with minutes-of-usage data. We plotted this time-series on a calendar, creating the Calendar Map you see above. Red cells show days with lower usage, and green cells show days with higher usage. This made it possible to spot a number of patterns that were relatively hidden until then. For example, on this calendar map, it’s obvious that call volumes are lower on Sundays. But 31st July was a relatively good Sunday, with high call volumes. That’s tough to spot on a line graph because it’s not high in absolute terms — just high for a Sunday. With this visualization, our client discovered a number of insights in calling pattern behavior of their customer segments. For example, the share of rural traffic rises on Sundays, mainly because urban traffic falls while rural traffic is unaffected. It also made it possible for them to identify specific days on which their competitors’ call volumes shot up, and helped them identify which competitor’s campaigns were proving effective against them. [divider top=”1″] Here’s a visualization of the social network of geeks across different cities in India. An interactive version of this is available at http:\/\/gramener.com\/codersearch. We built this to identify who would be a good candidate to hire, as well as decide which city is the best hunting ground for geeks. Each circle represents a developer. The size indicates the number of followers they have on Github. The color indicates the language they code in. Networks of followers are connected by lines and clustered together. This is an instance of transforming relatively unstructured data into quantitative metrics (distance between a pair of people; density of a network; etc) and displaying them purely visually, without any numbers. As a result, it conveys far more richness and meaning intuitively to the viewers. [divider top=”1″] Another instance is this visualization of the entire history of batting in Indian one-day cricket. The size of the box represents the number of runs scored by the player. The color indicates the speed at which they scored those runs (red is slow, green is fast.) It’s evident that among the big scorers, Sehwag is India’s fastest run-getter. Clicking on the players shows an second drill-down featuring every match they’ve played. An interactive version of this is available at http:\/\/gramener.com\/cricket\/batting-India-plain This compresses over 150 pages of information into a single sheet without any loss. Part of the power of data visualization comes in this ability to compress information and compactly convey insights. [divider top=”1″] We’re moving into visualizations of non-quantitative data. There’s a lot more text out there than numbers, and it’s possible to mine information from that. For example, even a pure-text corpus like the Mahabharata lends itself to social network analysis.","excerpt":"Consider this price and sales table for four cities:   Can you figure out how each city is performing? Notice that the average for each city is the same. Now take a look at the same data, plotted. The patterns are a lot clearer now, and you can quickly see that: The four cities are […]","categories":["IT Services"],"tags":[],"author_name":"Anand","publish_date":"2012-07-24T13:26:12","publication_year":"2012","word_count":585,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","Git","RAG","GitHub","R"],"extracted_tech_keywords":["AI","RAG","R","Go","Git","GitHub","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/data-visualization-a-pov-from-gramener\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10048052,"title":"AWS Extends Cloud Computing &#038; AI Curriculum To 21 Higher Education Institutions","content":"Amazon Web Services (AWS) has announced that it will add another 21 higher education institutions across India offering cloud computing curriculum as part of their undergraduate and postgraduate programs starting in the 2021-22 academic year. This was done after successfully integrating the cloud computing curriculum in programs of study at seven education institutions in 2020. “We know that technology is redefining India’s workforce and the skills needed to remain competitive. AWS is committed to building the future technology workforce in India by introducing students to cloud computing skills early in their learning path. With AWS training and learning resources, students benefit from the practical skills they gain to immediately add value in high-value technology roles and accelerate their careers,” said Sunil PP, Lead – Education, Space and Non-Profits AISPL, AWS India and South Asia. He further added, “We are delighted to see more higher education institutions take the initiative to integrate cloud computing curriculum in their education programs to develop the next generation of cloud computing professionals.” The cloud computing curriculum will be included in engineering programs for computer science and information technology, degree programs such as Bachelor of Science, Bachelor of Computer Applications, and post-graduate degree programs in technology and management offered by these institutions in the states of Andhra Pradesh, Gujarat, Himachal Pradesh, Karnataka, Maharashtra, Orissa, Punjab, Rajasthan, and Tamil Nadu. Students in these institutions will learn about cloud architecture, data analytics, cybersecurity, artificial intelligence (AI), and machine learning (ML). In addition to AWS training and learning resources, students will have access to the AWS Management Console, a simple web interface that provides them real-world, hands-on experience in using and experimenting with AWS cloud services and building applications. To ensure that students have access to trained educators, more than 100 faculty members at the institutions have been trained by AWS on delivering the cloud computing curriculum to students. This new approach by education institutions aligns with India’s National Education Policy 2020 (NEP 2020) recommendations and highlights the importance of advancing technical education at the undergraduate and postgraduate levels, enhancing institutional capabilities in technology and providing students enhanced experiential learning opportunities.","excerpt":"Students in these institutions will learn about cloud architecture, data analytics, cybersecurity, artificial intelligence, and machine learning.","categories":["AI News"],"tags":[],"author_name":"kumar Gandharv","publish_date":"2021-09-09T17:22:10","publication_year":"2021","word_count":352,"keywords":["machine learning","artificial intelligence","AWS","AI","cloud computing","cloud_platforms:AWS","ML","analytics","cloud_platforms:Amazon Web Services","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","analytics","cloud computing","AWS","R","cloud_platforms:AWS","cloud_platforms:Amazon Web Services"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/aws-extends-cloud-computing-ai-curriculum-to-21-higher-education-institutions\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10018926,"title":"Recent Use Of AI In 5G Shows Why It&#8217;s Hard to Shun China In Tech","content":"Last week, Nokia and China Mobile (CMCC) announced the successful completion of AI-powered 5G Radio Access Network (RAN) trials over CMCC’s network. RAN is part of the mobile telecommunication systems. A RAN Intelligence Controller (RIC) is a critical element in enabling RAN to support interoperability across different hardware and software components for resource optimisations. RIC plays an important role in 5G technology. In this article, we try to understand the technical, economic and geopolitical advantage China stands to gain from the success of the technical trial carried out by Nokia and China Mobile. Advantage China RIC provides advanced control functionality, which delivers increased efficiency and better radio resource management by leveraging analytics and data-driven approaches. RIC plays a crucial role in enabling Artificial Intelligence and Machine Learning capabilities in the RAN. According to Nokia, the trial successfully forecasted bandwidth and predicted traffic requirements with an accuracy of 90% in a live test on CMCC’s 5G network. Nokia also trialled network anomaly detection with CMCC subsidiaries, Shanxi Mobile and the China Mobile Research Institute (CMRI). The anomaly detection software provides distributed and split architecture for near-real-time RIC deployment. Meaning, operators in China can cut up to 70% costs by removing humans from the process, said Nokia. With this trial, China stands to gain geopolitically as well. The US and its allies have been at loggerheads with China over the progress it made in 5G and other emerging technologies, leading to the US banning apps or chip manufacturers from supplying to China, and also led to the formation of GPAI with 15 countries excluding China. The UK government has also set up D10 including countries from the G7 plus Australia, India, and South Korea. This group’s explicit purpose has been to explore the alternatives to Huawei’s 5G, a Chinese telecom manufacturer. “The objectives of D10 have been clearly defined. However, the concern has always been whether the member states have the economic capacity and the political desire to do so,” said Arindrajit Basu, program manager at Centre for Internet and Society, whose research focuses on the geopolitics of technology. While the D10 does not include Finland, the coalition between a democratic country like Finland and China, brings an entirely different dynamic to the D10 vs China equation. “Looking at this situation as a democracy vs authoritarianism or a The West vs China debate,” said Basu, “for smaller countries (like Finland) it is not that simple to completely ban tech collaborations with China. Apart from their values, they have to also consider what is strategically beneficial for them.” “And this is not true only in the EU, but even in Southeast Asia as well. While there has been pressure to do away with Chinese technology, it has not happened. China continues to have a good hold in Africa and Southeast Asia. Even in India to an extent, despite the recent tensions. “This is called ‘weaponised interdependency’, a term coined by Abraham Newman and Henry Farrell. Because of the interconnectedness of supply chains (in tech), it is tough for a country, especially smaller ones, to come out and say they will shun China completely.” Wrapping Up China is making quantum leaps on the technological front on par, if not above with respect to EU and the US. The collaboration with Finland is a good case in point. Such moves are essential for China to stay geopolitically relevant. While the US and other countries are trying hard to counter China’s influence, it has been difficult since smaller countries need to look beyond their values and make strategic collaborations to sustain themselves economically. From a geopolitical and technological perspective, China is far ahead of the game.","excerpt":"Last week, Nokia and China Mobile (CMCC) announced the successful completion of AI-powered 5G Radio Access Network (RAN) trials over CMCC’s network. RAN is part of the mobile telecommunication systems. A RAN Intelligence Controller (RIC) is a critical element in enabling RAN to support interoperability across different hardware and software components for resource optimisations. RIC […]","categories":["AI Features"],"tags":["china chip technology"],"author_name":"Kashyap Raibagi","publish_date":"2021-01-26T18:00:00","publication_year":"2021","word_count":610,"keywords":["Go","artificial intelligence","machine learning","programming_languages:R","AI","data-driven","RAG","china chip technology","anomaly detection","analytics","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","analytics","RAG","anomaly detection","R","Go","data-driven","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/recent-use-of-ai-in-5g-shows-why-its-hard-to-shun-china-in-tech\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":45086,"title":"After Confidential Consortium, Alibaba Ups Data Privacy Game, With Shared Machine Learning","content":"Data breaching and other internet frauds have risen dramatically over the past few years. According to sources, organisations in India lost about ₹12.8 crore on average between July 2018 and April 2019. The global average total cost of data breach was USD 3.92 million (about ₹27.03 crore) with the average size of the breach being 25,575 records. Data is an important information asset and a crucial part of organisation and big techs are striving hard to protect data from hackers. Cyber-attacks have become one of the top concerns among the big tech companies globally. It is essential for organisations to keep one step ahead of cyber attackers and hackers. Recently, in our of our articles, we discussed steps taken to protect data by Confidential Computing Consortium created by the Linux Foundation. Tech giants like IBM, Intel, Google, Microsoft, Red Hat among others are members of this data safety consortium. These companies have contributed various toolkits and projects for confidential computing. Data is moving between various computing environments such as on-premises servers, public cloud, etc. and this consortium will assist those data to move freely with high security. After the trade-ban with the US, China has upped its AIML game and is adopting AI not only in the organisations but also for integrating into educational purposes. In the last couple of months, Chinese researchers, as well as tech giants, have made several contributions in emerging technologies such as introducing hybrid chip known as Tianjic chip to stimulate AGI development, open sourcing Xuantie 910 which is a powerful RISC-V (Reduced Instruction Set Computer) processor. With fast advancements in the field of emerging technologies, the fear of cyber attacks is also around the corner. For these security issues, Chinese tech giant, Alibaba has been trying to keep safe the data by ensuring data privacy and it has also recently become a member of the Confidential Computing Consortium. A few days ago, Chinese commercial giant, Alibaba’s subsidiary Ant Financial introduced Shared Machine Learning (SML) as their solution for data privacy. SML is a combination of the Trusted Execution Environment (TEE) and Multiparty Computation (MPC) system. This solution includes Intel’s SGX technology in its foundation layer which is compatible with other TEE implementations and supports both online prediction and offline training. The TEE-based framework supports a variety of commonly used prediction algorithms including LR, GBDT, and Xgboost; and enables prediction on encrypted data from multiple parties while the MPC-based framework supports popular algorithms including LR, GBDT, GNN, etc. The framework is able to perform various tasks mentioned below: Solve the problems of load balancing, failover, dynamic expansion and shrinkage, and disaster recovery with the clustering solution Provide an easy-to-use development framework for users Reduce the user’s access cost through a built-in technology of ServiceProvider with SDK Solve problems such as code upgrade, grayscale publishing, and release rollback with multi-cluster management and SDK heartbeat mechanism Provide a provision agent mechanism to ensure that SGX does not need to connect to the external network, which improves system security Looking Ahead As consumers move towards a smart society data will be susceptible to breaches. In the digital era, every single important information is residing somewhere on-premise or in the cloud platform. Big tech organisations are trying hard to ensure privacy protection and strong security measures for all personal information.","excerpt":"Data breaching and other internet frauds have risen dramatically over the past few years. According to sources, organisations in India lost about ₹12.8 crore on average between July 2018 and April 2019. The global average total cost of data breach was USD 3.92 million (about ₹27.03 crore) with the average size of the breach being […]","categories":["AI Features"],"tags":["Alibaba"],"author_name":"Ambika Choudhury","publish_date":"2019-08-28T17:00:52","publication_year":"2019","word_count":551,"keywords":["Go","machine learning","AI","Alibaba","ML","RAG","Ray","Aim","XGBoost","Rust","R"],"extracted_tech_keywords":["AI","machine learning","ML","Aim","Ray","XGBoost","RAG","R","Go","Rust"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/after-confidential-consortium-alibaba-ups-data-privacy-game-with-shared-machine-learning\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":25686,"title":"Space Particles may be the reason your PC just crashed","content":"You might have to often encounter the time-consuming hassle of restarting your computer or smartphone, which might have just crashed or frozen for no apparent reason. In such cases, you usually end up blaming the manufacturer. However, a recent study conducted by a group of scientists, including one of Indian origin, suggest that subatomic particles raining down from outer space could be the prime reason behind this. The cosmic rays originating from outside the solar system generate these electrically charged particles, which often is the reason behind the crashes. Bharat Bhuva, Professor of electrical engineering, Vanderbilt University expresses, “The problem is of intense magnitude. However, it mostly remains invisible to the human eye.” Cascades of secondary particles comprising energetic neutrons, muons, pions, and alpha particles are usually created when the lightspeed cosmic rays strike the Earth’s atmosphere. Millions of these particles strike the human body every second, without causing any harmful effects for living organisms. However, a fraction of these subatomic particles carry enough energy to disrupt the functioning of microelectronic circuitry. These particles can adversely alter individual bits of data stored in memory, and the incident is termed as a single-event upset or SEU. It’s a challenging task to assert the time and place, where these space particles will strike. This in turn makes it a difficult task to characterize the malfunctions, or determine the prevalence of SEUs. Bhuva comments, “A single bit-flop is easier to inspect, it could be a software bug or a hardware flaw, for instance. However, to effectively determine if it is an SEU, you have to eliminate all other possible causes.” The problem is of serious nature, and SEUs have affected electronics in the past, ranging from voting machines to massive airlines. Moreover, unexplained glitches have been observed previously in airline computers. Experts have probed these glitches and the most probable cause turned out to be an SEU. These issues have resulted in cancellation of hundreds of flights, causing significant economic losses.","excerpt":"You might have to often encounter the time-consuming hassle of restarting your computer or smartphone, which might have just crashed or frozen for no apparent reason. In such cases, you usually end up blaming the manufacturer. However, a recent study conducted by a group of scientists, including one of Indian origin, suggest that subatomic particles […]","categories":["AI News"],"tags":[],"author_name":"Amit Paul Chowdhury","publish_date":"2017-02-24T06:45:39","publication_year":"2017","word_count":328,"keywords":["programming_languages:R","AI","Ray","GAN","R"],"extracted_tech_keywords":["AI","Ray","R","GAN","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/space-particles-may-reason-pc-just-crashed\/","complexity_score":2,"technical_depth":5,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":38147,"title":"MachineCon Comes At The Right Time In Singapore, Asia’s Big Data Hub","content":"Machine Conference 2019 edition is set to bring newer flavours from the analytics community as it promises to assemble more than 400 industry leaders from India and Singapore in May this year. The second edition of the conference makes a timely presence in Singapore as it is turning out to be the rapidly growing Southeast Asian countries with a lot of innovations happening in the space of analytics, data science and AI. While we have earlier covered articles about what to expect from MachineCon and the reasons why every analytics leader must attend it, here we bring 5 reasons why Singapore edition of the Machine conference comes at the right time. Singapore is the fastest growing hub for analytics innovation: The country is known as the global hub for innovation and has made a huge impact on the fintech ecosystem. It could be largely attributed to the enhanced adoption of data science and analytics practice. Singapore has made great leaps from analytics engagement to analytics enabler and now contributes 44% to the analytics market size among Southeast Asia countries, which is one of the highest. It gives an opportunity to the analytics leaders to explore business opportunities in the country while playing on the strengths that Singapore already offers. A healthy ecosystem of deep tech startups: There is a drastic increase in the number of deep tech companies in Singapore as the government has pledged a high amount in the commercialisation of the deep tech startup ecosystem. Cutting edge and disruptive technologies such as artificial intelligence, machine learning and data science have found a stronger foot in Singapore and there are startups exploring these technologies in areas such as retail, agriculture, cybersecurity and more. This is the right time for analytics enthusiasts to discuss opportunities in the startup space. Strong government back up and support for the analytics community: One of the reasons why Singapore has taken a substantial lead and has been able to build a facility in the analytics ecosystem is because it has got strong support from the government. Initiatives such as AI Governance framework, SkillsFuture Initiative, AI for Everyone and several others have escalated the adoption of AI and new age technologies. The city and state have innovated, tested and piloted several emerging technology solutions making it a likely destination to host analytics summit. Evolving research excellence in AI: An independent study cited that more than 24,000 papers on artificial intelligence were published in Southeast Asia, with Singapore, Malaysia and Thailand accounting for 86% of the output. This suggests that the country is taking its AI developments quite seriously and are aiming for more research-based outcomes in this high growth rate space. A good network of investors and venture capitalists: We recently covered an article about top leading venture capitalists from Singapore who are investing in deep tech startups including AI and data science in Singapore. It clearly suggests that there are a lot of tech enthusiasts existing in this space that are ready to back up these startups in the country. The Machincon will act as a platform to bring this community on a single stage and allow attendees to discuss opportunities for setting up a stronger foot in the country. Note: MachineCon is an invite-only conference Visit the website for more details Date: 24 May, Mumbai | 30 May, Singapore","excerpt":"Machine Conference 2019 edition is set to bring newer flavours from the analytics community as it promises to assemble more than 400 industry leaders from India and Singapore in May this year. The second edition of the conference makes a timely presence in Singapore as it is turning out to be the rapidly growing Southeast […]","categories":["AI Features"],"tags":["which countries have good cybersecurity"],"author_name":"Srishti Deoras","publish_date":"2019-04-23T12:18:01","publication_year":"2019","word_count":556,"keywords":["data science","Go","artificial intelligence","machine learning","TPU","AI","which countries have good cybersecurity","Scala","Aim","analytics","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","data science","analytics","Aim","TPU","R","Go","Scala"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/machinecon-comes-at-the-right-time-in-singapore-asias-big-data-hub\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10065377,"title":"Google unveils Phorhum; state-of-the-art in photorealistic 3D human reconstruction","content":"Google unveiled Phorhum, a photorealistic 3D human reconstruction that can greatly help online apparel shopping. Phorhum is a method to photo-realistically reconstruct a dressed person’s 3D geometry and appearance as photographed in a single RGB image. The produced 3D scan of the subject accurately resembles the visible body parts and includes plausible geometry and appearance of the non-visible parts. The 3D scans of people wearing clothing have many use cases and are high in demand currently. Image – GitHub In the paper, ‘Photorealistic Monocular 3D Reconstruction of Humans Wearing Clothing’, which talks about Phorhum, the authors say, “The construction of our model is motivated by the breadth of transformative, immersive 3D applications that would become possible for clothing virtual apparel try-on, immersive visualisation of photographs, personal AR and VR for improved communication, special effects, human-computer interaction or gaming, among others”. Further, the paper concludes, “Our method works well for a wide variation of outfits and for diverse body shapes and skin tones, and reconstructions capture most of the detail present in the input image.” Phorhum is an end-to-end trainable, deep neural network methodology for photorealistic 3D human reconstruction that works with just a monocular RGB image. The pixel-aligned method estimates a detailed 3D geometry and unshaded surface colour and scene illumination. The 3D supervision alone is not sufficient for high fidelity colour reconstruction, so Phorhum introduced patch-based rendering losses that enable reliable colour reconstruction on visible parts of the human and detailed and plausible colour estimation for the non-visible parts.","excerpt":"Phorhum is the new SOTA in photorealistic 3D human reconstruction from monocular RGB images of humans wearing clothing.","categories":["AI News"],"tags":[],"author_name":"Poornima Nataraj","publish_date":"2022-04-20T19:59:40","publication_year":"2022","word_count":250,"keywords":["Go","programming_languages:R","AI","neural network","programming_languages:Go","Git","GitHub","R"],"extracted_tech_keywords":["AI","neural network","R","Go","Git","GitHub","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/google-unveils-phorhum-state-of-the-art-in-photorealistic-3d-human-reconstruction\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10099221,"title":"The Beginning of the End of SaaS Startups","content":"A few months ago we had predicted generative AI startups have no moat, no money, and now its confirmed. When Sam Altman claimed that the rise of generative AI chatbots like ChatGPT is going to replace customer service jobs, he wasn’t wrong. Forget jobs, he is now coming after SaaS startups. OpenAI’s recent introduction of ChatGPT Enterprise might have sent shock waves among several SaaS startups that had developed products around ChatGPT or offered wrappers around ChatGPT APIs catering to business clients. OpenAI in their blog said that ChatGPT Enterprise comes with a new admin console that will let businesses manage team members easily and offers domain verification, SSO, and usage insights, allowing for large-scale deployment into enterprise. This development overlaps with services of many of the current SaaS startups who offer B2B services. Nevertheless, the introduction of ChatGPT Enterprise could potentially jeopardize their survival, as OpenAI has stepped in to offer business solutions centered on ChatGPT.  With ChatGPT Enterprise, OpenAI is further planning to launch more tools for specific roles, such as data analysts, marketers, customer support etc. Not only this, ChatGPT Enterprise brings enterprise-level security and privacy, unlimited high-speed access to GPT-4, extended context windows for handling longer inputs, advanced data analysis functionalities, customization choices, and a host of other features, which makes it far better than ChatGPT Plus. Furthermore, ChatGPT Enterprise removed all usage caps, and performs up to two times faster. It  includes 32k context in Enterprise, allowing users to process four times longer inputs or files. To add cherry on the cake, ChatGPT Enterprise also provides unlimited access to advanced data analysis, previously known a Code Interpreter. Many companies were interested in utilizing Code Interpreter in ChatGPT Plus initially, but they held back due to worries about data security. If you explore OpenAI’s discussion forum, you’ll come across numerous requests for the Code Interpreter API. Sneak Peek into ChatGPT-based Startups Bito, a B2B startup based in Menlo Park, New Jersey, boasting more than 100,000 users and self-proclaimed as the “Swiss Army knife of capabilities” for software developers, has introduced an AI coding assistant fueled by ChatGPT. Additionally, they have received  $3.2 million in fresh funding. Similarly, Yuma, a platform designed  for Shopify merchants, employs AI systems resembling ChatGPT to improve customer support. By seamlessly integrating with help desk software, Yuma offers personalized and pertinent responses to customer inquiries. Baselit  is another startup leveraging ChatGPT technology to enable businesses to access chatbot-style analytics. Utilizing OpenAI’s GPT-3 text comprehension model, Baselit enables users to execute database queries using simple English, eliminating the need for coding expertise. What lies ahead for them? It appears that the road ahead might be challenging for these startups to thrive. Convincing venture capitalists to invest was already a struggle, and as we move into the age of generative AI, having a moat becomes even more essential. Whatever moat these companies were offering  to the VCs, is already being taken care of by ChatGPT Enterprise. Whether it’s data analytics, customer support, marketing, or any other field, the message has been consistently clear: building your startup around ChatGPT carries risks due to the lack of true ownership as OpenAI holds the intellectual property (IP) rights to GPT-3 and all its other models. Today, these fears have come true. Every new OpenAI feature: \"there goes a thousand startups\". Let's be real, it's unlikely to be that simple lol— anton (@abacaj) August 29, 2023 So all the startups built amidst the hype around ChatGPT, which were either using GPT in their name, or building technologies using the APIs provided by OpenAI might soon need to pack their bags. Startups will Lose, But Will Biggies Join Back? This development is big enough for smaller startups to accept defeat in front of ChatGPT Enterprise. However, it needs to be seen whether big tech companies like Apple, Spotify, Wells Fargo, Samsung, JP Morgan, Verizon which had earlier ditched ChatGPT on risk of data leaks will come back and use it. Companies like Apple had valid concerns about its employees inadvertently sharing sensitive project details through the system which might potentially be viewed by OpenAI moderators. Apple went ahead and created its own chatbot named AppleGPT for internal use. This time OpenAI seems to be well prepared and  explicitly said it does not train on business data or conversations. Furthermore they added that ChatGPT Enterprise is SOC 2 compliant and all conversations are encrypted in transit and at rest. OpenAI in their blog post mentioned that industry leaders like Block, Canva, Carlyle, The Estée Lauder Companies, PwC, and Zapier are some of the few early users of ChatGPT Enterprise.","excerpt":"Startups will Lose, But Will Biggies Join Back?","categories":["Deep Tech"],"tags":["SaaS"],"author_name":"Siddharth Jindal","publish_date":"2023-08-29T17:34:33","publication_year":"2023","word_count":769,"keywords":["ChatGPT","OpenAI","AI","chatbots","ML","RAG","SaaS","Aim","analytics","generative AI","R"],"extracted_tech_keywords":["AI","ML","analytics","generative AI","ChatGPT","OpenAI","Aim","RAG","chatbots","R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/the-beginning-of-the-end-of-saas-startups\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10016878,"title":"What Is WILDS DataSet By Stanford &#8211; A Complete Guide","content":"WILDS is a benchmark of in-the-wild distribution shifts spanning a variety of datasets and applications, consisting of wildlife monitoring, tumour identification, poverty mapping and some others. Until now, seven datasets have been incorporated, and more is to be done. Wilds builds on top of recently collected data by experts. It provides evaluation metrics along with train\/test splits that represent real-world distribution shifts. These datasets show distribution shifts in training and testing data on different cameras, time periods, countries, demographics, molecular scaffolds, etc., which causes significant performance drop in baseline models. It is maintained by many researchers at Stanford, and some others from Berkley, Cornell, Caltech Universities and Microsoft Research team. FMoW – Building and land classification across different regions and years Machine learning techniques can enable global-scale monitoring of sustainability, specifically in data-poor regions using satellite imagery and other remotely sensed data to meet economic challenges. In just areas gathering data is expensive. Wilds tries to bridge this data gap that can thereby improve research and decision-making to undertake policies and humanitarian efforts such as population density, tracking deforestation, crop yield prediction, poverty mapping and addressing other such issues. As human activity and environmental processes can often cause changes to the natural environment, thus ML models must be trained robustly to distribution shifts over time. Dataset design: The input x in this dataset is a satellite image, and the target label y is one of 62 categories of land use. The domain d is measured on time and geographical regions. Wilds aims at solving a domain generalization in terms of time and subpopulation performance in terms of region. PovertyMap – Poverty mapping across different countries Proper predictions of poverty measures are necessary for directing policy decisions in developing and poverty-stricken countries. Actually, poverty ground truth measurements are lacking for most developing countries, since it’s a difficult task to gather information in this regard. In some countries, it may have never been conducted a survey or had gaps of over a decade between surveys. The lack of labels generation in certain countries creates a natural scenario for the need for model generalization in case of unseen countries. Shift across countries based on model performance is considered the effect of rural vs. urban subpopulations. Improving performance within rural subpopulation will help developing countries like Africa. Dataset design: The input x is a satellite image, and the output label y is a real-valued asset wealth index. The domain d is measured on countries. Wilds aims to solve both a domain generalization problem in terms of country borders and improve subpopulation performance in terms of urban and rural areas. iWildCam – Species classification across different camera traps In the 2020 Living Planet Report claimed that animal populations have declined by 68% on average since 1970. In the present climate diversity, the proper mapping between climate change and wildlife biodiversity loss has become a serious issue to be concerned about. For monitoring wildlife, one of the primary methods that have been adopted is placing heat or motion-activated cameras into the wild. These cameras can track data much faster than anyone can process it, as a result of this ecologists have taken up computer vision solutions. Static cameras like this are capable of capturing signals that are correlated in space and time. The correlation causes overfitting and thereby poor generalization as compared to new sensor deployments, degrading scalability factor of computer vision solutions. Dataset Design: This is a case of multi-class species classification. The input x is a photo captured by the camera, the output label y is divided into 186 different classes of animal species, and the domain d is a measurement that identifies which camera trap took the photo. Camelyon17 – Tumor identification across different hospitals Medical applications models are generally trained on a small set of data acquired from some hospitals due to patients’ privacy issues but get deployed to other hospitals as well. Model accuracy can degrade in the data collection and processing variations through different hospitals not included in the training set. This variation can be caused by many sources for example studying tissue slides under a microscope, differences can arouse in slide staining or the patient population or image acquisition. Wilds studies this distribution shift by building a patch-based variant of the Camelyon17 dataset. Dataset design: This is a binary classification task. The input x is a histopathological image, the output label y is a binary indicator that contains any tumour tissue or not, and the domain d is a measurement that identifies the hospital. OGB-MolPCBA –  Molecular property prediction across different scaffolds Drug discovery is a time-consuming procedure by now we must have realised. The entire process takes many years, during which many experiments are conducted to find a potent molecule. For computer-aided search, accurate and generalizable molecular property predictor is useful over a large collection of small molecules in order to detect those structures which have the most probability to bind to a drug target. Accurate computer vision aided solutions can largely reduce redundant experiments, hence help in accelerating the drug discovery process. The biggest challenge here remains is that molecular properties prediction over a variety of molecules screened from the large chemical database. It is thus crucial for models to generalize out of the dataset molecules that are structurally different from training ones. Dataset Design: This is a multi-task classification problem. The input x is a graphical representation of a molecule, the target label y is a binary vector of length 128 types of biological activity, and the domain d is the scaffold group that the molecule is a part of. Amazon – Sentiment classification across different users As discussed above for medical data likewise for text data also models are similarly trained on collected data and deployed as an all-purpose model across a wide range of users. Hence these models can show performance disparities. These drawbacks of performance gaps in applications have urged for the need for good performance across a wide range of users. Additionally, the indicative unfairness of models, their failure to learn the actual task, thereby leading to biasness. Wilds makes use of inter-individual performance disparities for the sentiment classification task on the Amazon-wilds dataset where the goal is to train models with high performance in terms of reviewers. Dataset design: This is a multi-class sentiment classification task. The input x is the text for review, the target label y is the star corresponding to the rating from 1 to 5, and the domain d is used as the indicator of the user who wrote the review. CivilComments – Toxicity classification across demographic identities Automatic review management of user-generated text such as detecting if a comment is negative is an important task for moderating the huge amount of volume of text being written daily on the Internet. Earlier works have shown documented biases in automatic moderation tools, for example, comment classifiers have shown the particular mention of certain demographic groups. Wilds has made a modified version of the CivilComments dataset, a large collection of comments on online posts\/articles etc. taken from the Civil Comments platform and annotated for negativity and demographic mentions by multiple crowd workers. Dataset design: This a binary classification task of predicting whether or not a comment is negative. The input x is a comment comprising one or more sentences, and the target label y is whether it is marked negative or not. The domain annotation d is a multi-dimensional binary vector denoting whether the comment mentions each of the eight demographic entities LGBTQ, male, female, Christian, Muslim, other religions, Black, and White. Usage WILDS has an open-source Python package which provides a standardized interface for all datasets. Installation pip install wilds Or git clone https:\/\/github.com\/p-lambda\/wilds Additional dependencies pip install torch-scatter -f https:\/\/pytorch-geometric.com\/whl\/torch-${TORCH}+${CUDA}.html pip install torch-sparse -f https:\/\/pytorch-geometric.com\/whl\/torch-${TORCH}+${CUDA}.html pip install torch-cluster -f https:\/\/pytorch-geometric.com\/whl\/torch-${TORCH}+${CUDA}.html pip install torch-spline-conv -f https:\/\/pytorch-geometric.com\/whl\/torch-${TORCH}+${CUDA}.html pip install torch-geometric pip install transformers Default models python run_expt.py --dataset civilcomments --algorithm groupDRO --root_dir data --download Data loading from wilds.datasets.iwildcam_dataset import IWildCamDataset from wilds.common.data_loaders import get_train_loader import torchvision.transforms as transforms # Loading full dataset, and downloading dataset = IWildCamDataset(download=True) # Getting the training set train_data = dataset.get_subset('train', transform=transforms.Compose([transforms.Resize((224,224)),                                                           transforms.ToTensor()])) # Preparing the data loader train_loader = get_train_loader('standard', train_data, batch_size=16) Training code snippet def train(algorithm, datasets, general_logger, config, epoch_offset, best_val_metric): for epoch in range(epoch_offset, config.n_epochs): general_logger.write('\\nEpoch [%d]:\\n' % epoch) # First run training run_epoch(algorithm, datasets['train'], general_logger, epoch, config, train=True) # Then run val val_results = run_epoch(algorithm, datasets['val'], general_logger, epoch, config, train=False) curr_val_metric = val_results[config.val_metric] general_logger.write(f'Validation {config.val_metric}: {curr_val_metric:.3f}\\n') # Then run everything else if config.evaluate_all_splits: additional_splits = [split for split in datasets.keys() if split not in ['train','val']] else: additional_splits = config.eval_splits for split in additional_splits: run_epoch(algorithm, datasets[split], general_logger, epoch, config, train=False) if best_val_metric is None: is_best = True else: if config.val_metric_decreasing: is_best = curr_val_metric < best_val_metric else: is_best = curr_val_metric > best_val_metric if is_best: best_val_metric = curr_val_metric if config.save_step is not None and (epoch + 1) % config.save_step == 0: save(algorithm, epoch, best_val_metric, os.path.join(config.log_dir, '%d_model.pth' % epoch)) if config.save_last: save(algorithm, epoch, best_val_metric, os.path.join(config.log_dir, 'last_model.pth')) if config.save_best and is_best: save(algorithm, epoch, best_val_metric, os.path.join(config.log_dir, 'best_model.pth')) general_logger.write(f'Best model saved at epoch {epoch}\\n') general_logger.write('\\n') def run_epoch(algorithm, dataset, general_logger, epoch, config, train): if dataset['verbose']: general_logger.write(f\"\\n{dataset['name']}:\\n\") if train: algorithm.train() else: algorithm.eval() # Not preallocating memory is slower # but makes it easier to handle different types of data loaders # (which might not return exactly the same number of examples per epoch) epoch_y_true = [] epoch_y_pred = [] epoch_metadata = [] # Using enumerate(iterator) can sometimes leak memory in some environment so instead manually incrementing batch_idx batch_idx = 0 iterator = tqdm(dataset['loader']) if config.progress_bar else dataset['loader'] for batch in iterator: if train: batch_results = algorithm.update(batch) else: batch_results = algorithm.evaluate(batch) # These tensors are already detached, but we need to clone them again # Otherwise they don’t get garbage collected properly in some versions # The subsequent detach is just for safety # (they should already be detached in batch_results) epoch_y_true.append(batch_results['y_true'].clone().detach()) epoch_y_pred.append(batch_results['y_pred'].clone().detach()) epoch_metadata.append(batch_results['metadata'].clone().detach()) if train and (batch_idx+1) % config.log_every==0: log_results(algorithm, dataset, general_logger, epoch, batch_idx) batch_idx += 1 results, results_str = dataset['dataset'].eval( torch.cat(epoch_y_pred), torch.cat(epoch_y_true), torch.cat(epoch_metadata)) if config.scheduler_metric_split==dataset['split']: algorithm.step_schedulers( is_epoch=True, metrics=results, log_access=(not train)) # log after updating the scheduler in case it needs to access the internal logs log_results(algorithm, dataset, general_logger, epoch, batch_idx) results['epoch'] = epoch dataset['eval_logger'].log(results) if dataset['verbose']: general_logger.write('Epoch eval:\\n') general_logger.write(results_str) return results Evaluators from wilds.common.data_loaders import get_eval_loader # Getting the test set test_data = dataset.get_subset('test', transform=transforms.Compose([transforms.Resize((224,224)),     transforms.ToTensor()])) # Preparing the data loader test_loader = get_eval_loader('standard', test_data, batch_size=16) # Getting predictions for the full test set for x, y_true, metadata in test_loader: y_pred = model(x) # Evaluation dataset.eval(all_y_pred, all_y_true, all_metadata) End Notes WILDS tries to put a generalized dataset covering diverse data across visuals and text. It is under constant development, in future we can expect to see more benchmarked datasets to produce high quality trained data models that can address complex problems.","excerpt":"WILDS is a benchmark of in-the-wild distribution shifts spanning a variety of datasets and applications, consisting of wildlife monitoring, tumour identification, poverty mapping and some others.","categories":["Deep Tech"],"tags":["Python"],"author_name":"Jayita Bhattacharyya","publish_date":"2020-12-31T15:00:00","publication_year":"2020","word_count":1802,"keywords":["CUDA","machine learning","TPU","AI","PyTorch","ML","Transformers","computer vision","RAG","Python","Aim"],"extracted_tech_keywords":["AI","machine learning","ML","computer vision","Aim","PyTorch","Transformers","RAG","TPU","CUDA"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/what-is-wilds-dataset-by-stanford-a-complete-guide\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10096964,"title":"Data Science Hiring and Interview Process at Lowe’s India","content":"American retail giant Lowe’s operates a chain of over 1700 stores, .com, and app in the US and is the second-largest hardware chain in the world, trailing behind Home Depot. With its Global Capability Centre in Bengaluru, Lowe’s India boasts a robust team of over 4,000 members who deliver exceptional service to 17 million plus customers. To make this possible, data plays an important role, particularly in the retail industry, where it is vital for decision-making processes such as search, product recommendations, inventory management, supply chain operations, and demand forecasting. And the data science team at Lowe’s India are at the forefront of this. The Data, Analytics, & Computational Intelligence (DACI) team at Lowe’s supports the company’s global business endeavours by delivering prompt, relevant, and profoundly actionable data, analytics, and cutting-edge analytic services and solutions. “Some of the key areas of innovation by the DACI team include computer vision applications, homegrown machine learning platform and many operational and enterprise analytics insights,” said Amit Kapur, Vice President, Data & ML Platforms and Data Governance in an exclusive interaction with AIM. The DACI team is divided into ‘products’ and ‘platforms’ teams, spread across the US and India. The DACI platforms team is heavily led from India comprising over 300 members including all levels of data engineers, software engineers, data analysts, and data scientists. For instance, the merchandising teams have harnessed data insights provided by the DACI team to elevate and optimise their sourcing strategies. Moreover, the Stores teams at Lowe’s are leveraging the computer vision platform developed by the data science team, bolstering customer service and boosting productivity. Additionally, the marketing teams have devised customer lifetime value models, enabling them to gain deeper insights into customers and target them more effectively. The finance team employs machine learning models to forecast tax codes, thereby enhancing accuracy and efficiency in tax-related operations. Currently, DACI has over ten open positions for data scientists in Bangalore. These roles require a range of experience levels, from four to fifteen years. Interview Process The data science hiring process involves resume screening to assess qualifications, followed by a technical assessment to evaluate proficiency and problem-solving skills. A deep-tech interview focuses on algorithmic problem-solving, communication, and analytical thinking. The final round determines a candidate’s fit by assessing their ability to communicate clearly, solve problems efficiently, and demonstrate the business impact of their work. Aligning contributions with strategic goals provides an advantage, especially at senior levels. Some common key result areas include technical proficiency, machine learning algorithms, data visualisation, statistical and mathematical aptitude, and relevant experience in the retail industry. “Given that Lowe’s is a home improvement retailer, candidates with expertise in data analysis and exploration, particularly in the retail sector, have a competitive edge,” added Kapur. Expectations Meanwhile, the data science team at Lowe’s requires a range of skills including proficiency in machine learning algorithms, statistical analysis, problem-solving approaches, strong Python coding skills, and data visualisation. In terms of technology capabilities, the team utilises ML\/AI, deep learning, clustering, time series forecasting, Python, SQL, linear regression, and statistical modeling. They employ various tools, applications, and frameworks such as Python, ETL (Extract, Transform, Load), Scikit-learn for machine learning algorithms, ML Flow, KubeFlow, Feast, and Explainable AI to support their work. Additionally, Lowe’s values an innovation mindset, encouraging creative thinking and continuous improvement for the benefit of our customers and associates. Kapur points out that one of the common mistakes often make the mistake of overly emphasising theoretical knowledge during interviews instead of demonstrating the practical application of skills to real-world problems. Candidates should highlight past projects, explaining their approach, techniques used, and achieved results. “Especially at senior levels, it is important to quantify the business impact of their work. Candidates who can articulate how their work will contribute to the company’s strategic goals and can communicate their findings in a clear and concise manner will have a significant advantage during the interview process,” Kapur commented. Work Culture Lowe’s takes pride in its unique work culture, where associates are driven by core values like customer focus, courage, action, results, and continuous learning. Trust, respect, empathy, and agility form the foundation of the organisation. “Our people policies and benefits are inclusive and are designed to ensure we have a diverse talent pool. Currently, our gender ratio is at 33% – much ahead of the industry” said Kapur. Special perks at Lowe’s include ESOP, insurance, workcations, wellness programs, parental support, and skill development training, among others. “If you’re driven, eager to learn, and can solve complex problems at scale, Lowe’s offers the perfect opportunity to shape the future of omnichannel retail,” concluded Kapur. Click here to check out their careers page. Read more: Data Science Hiring Process at Mastercard","excerpt":"Currently, Lowe’s India has over ten open positions for data scientists in Bangalore.","categories":["AI Hirings"],"tags":["Data Science Hiring","data science hiring process","Top Trend"],"author_name":"Shritama Saha","publish_date":"2024-07-29T15:31:04","publication_year":"2024","word_count":784,"keywords":["Top Trend","data science","Feast","machine learning","AI","ML","computer vision","Data Science Hiring","Kubeflow","deep learning","Aim","analytics","data science hiring process"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","computer vision","data science","analytics","Kubeflow","Aim","Feast"],"url":"https:\/\/analyticsindiamag.com\/ai-hiring\/data-science-hiring-process-at-lowes-india\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10173176,"title":"Green Aero Raises $1.6 Million to Build Hydrogen and Defence Aero Engines","content":"Green Aero, an Indian deep-tech startup, has raised $1.6 million in a seed funding round led by pi Ventures, with participation from Antler. The company announced that the funding will support research and development, team expansion, and the creation of an in-house testing facility as the company prepares for commercial pilots in the defence and civil aviation sectors. It is currently incubated at IIT Delhi and is building propulsion technologies for aerospace, naval, and zero-emission hydrogen aviation applications. Founded in 2023 by Prithwish Kundu, a former research scientist at the US Department of Energy, Green Aero recently test-fired India’s first hydrogen-based aero engine core, named The Blue Dragon. He was also previously the lead for computational studies and numerical analysis at AgniKul Cosmos, a private launch player in the Indian startup ecosystem. Indigenous Propulsion Tech India has long relied on imported turbine engines for aircraft and drones. Green Aero seeks to change this by developing core propulsion systems within the country, targeting defence and sustainable civil aviation markets. The company is developing aero engines with twice the efficiency of current global models, featuring innovations such as a proprietary turbine, fuel-flexible combustors, swirl-stabilised injectors, and superalloy components created through additive manufacturing. The startup is also exploring high-thrust supersonic propulsion and plans to launch a small-category commercial engine within the next 12 months. A larger engine platform is under development as part of its long-term roadmap. Clean Aviation Future “We are excited to be at the forefront of developing advanced aero engine technologies in India for the world,” said Kundu. “This funding advances our long-term vision to shape a more sustainable future for the transportation industry.” Shubham Sandeep, managing director at pi Ventures, said, “Green Aero is building world-class aero engines from the ground up in India, with a laser-sharp focus on efficiency and performance. Their bold vision to make India self-reliant in propulsion technology while leading the world in clean aviation deeply resonates with our mission.” Green Aero’s development strategy combines rigorous R&D, subsystem testing, and innovative design and materials. The company’s goal is to develop lightweight, multi-fuel engines that meet modern aviation demands while minimising environmental impact.","excerpt":"The startup plans to launch a small-category commercial engine within the next 12 months.","categories":["AI News"],"tags":["seed funding","space engine","space india"],"author_name":"Sanjana Gupta","publish_date":"2025-07-09T21:01:53","publication_year":"2025","word_count":357,"keywords":["seed funding","Go","funding","programming_languages:R","AI","innovation","space engine","space india","programming_languages:Go","RAG","Rust","R","startup"],"extracted_tech_keywords":["AI","RAG","R","Go","Rust","innovation","startup","funding","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/green-aero-raises-1-6-million-to-build-hydrogen-and-defence-aero-engines\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":7666,"title":"Big Data &#038; Business Analytic for Sustainability of Competitive Advantage in Service Organization","content":"A growing number of companies are finding their service businesses under threat. The culprits are members of a new wave of digital upstarts that capitalize on changes in technology, customer behaviour, and the availability of data to create innovative, customer-friendly alternatives to the services incumbents offer. The review is an attempt to discuss possible implications on service sector faced with the impact of disruptive technology in the context of advent of data science and analytic for competitive advantage and sustainability. The service sectors of Indian economy that have grown faster than the economy are as follows: Information Technology (the most leading service sectors in Indian economy) IT-enabled services (ITeS) Telecommunications Financial Services Community Services Hotels and Restaurants In the contexts of the background of changing scenario, massive innovation in data science will definitely change the paradigm of doing business and especially for management of service sectors. The present analysis will focus on assessment of the impact of data revolution facilitated by the breakthrough of application of big data for optimization of service sector competitiveness. New digital upstarts are threatening the bottom lines, growth prospects, and even business models of traditional service providers. It’s time for incumbents to innovate or be left behind. A growing number of companies are finding their service businesses under threat. The culprits are members of a new wave of digital upstarts that capitalize on changes in technology, customer behaviour, and the availability of data to create innovative, customer-friendly alternatives to the services incumbents offer. Indeed, the sorts of digital disruptions that began in retailing with the likes of Amazon, two decades ago, are fast coming to an industry near you—if they haven’t already. Examples include Uber and Zipcar in transportation, Airbnb in hotels and hospitality, AngelList in venture capital, and Castlight Health and Healthgrades in healthcare. Attackers such as these may be small now, but they represent a growing challenge to traditional companies. The attackers also highlight an uncomfortable truth: large companies rarely put as much sustained effort and management attention into transforming services as they do with the products. Imagine the reaction of a patient visiting a hospital before 50 or 20 years ago and visiting a contemporary health care facility now. The medical devices, tools, and products available to physicians would be largely unrecognizable, but the service experience, in many cases, would be largely the same unless it is redesigned. Change is difficult with a large base of legacy assets optimized for a certain way of working, as well as a large, distributed workforce trying to maintain the status quo. The incremental approach many companies take to improving services doesn’t help; processes that grind out small, steady cost reductions rarely deliver breakthroughs. Nonetheless, some incumbents are fighting back successfully. These companies are learning from the attackers while mobilizing their own strengths—including scale, superior resources, and access to customers to redefine service offerings, harness digital technology, and improve the customer experience. Some are lowering their costs as well. While few organizations have mastered the new environment, we can already see that winning approaches will combine three elements: A focus on service innovation matching the intensity and attention that product companies bring to R&D The ability to personalize the customer experience and to help customers do things themselves The will to simplify (and in some cases automate) the way services are delivered To pull all this off, companies must find more collaborative ways of working to ensure that they remain focused on their customers, not their own internal processes. A closer look at how the environment is changing and what leading companies are doing about it should stir imaginations of a wide range of organizations struggling to adapt to digital and competitive world. The new service Sector landscape The nature of the services and the pace of change have shifted dramatically in recent years, and mastering the traditional aspects of service delivery will no longer be enough. To seize the opportunities, companies must learn to tap the potential for service innovation made possible by four evolving trends: In addition, artificial intelligence and robotics represent intriguing developments that appears poised to make their way from manufacturing to services. Higher customer expectations. More than ever, consumers demand greater involvement, customization, personalization, and mobility from services with immediate results. When they see cutting-edge service innovations in one industry, they expect to find them in others as well; Witness the spread of self-service kiosks from airline check-ins to the retailing and hospitality industries. As industry boundaries increasingly blur for customers, companies must look for new ideas beyond their immediate rivals. Big data presents great opportunities as they help us develop new creative products and services, for example apps on mobile phones or business intelligence products for companies. It can boost growth and jobs but also improve the quality of life. (Rao, Madfanmohon ,2014) The followings could be examples of service sector sustainability and competitiveness for discussion in the conference: A widespread use of big data in the health sector can help doctors make the right choices more quickly, on the basis of information collected by other medical staff. Patients can benefit from more timely and appropriate treatments and be better informed about health care providers. An increased use of data analysis in the health sector can also lead to enormous cost savings through a more precise identification of unnecessary procedures or duplication of tests. The analysis of large clinical datasets can result in the optimisation of the clinical and cost effectiveness of new drugs and treatments. The transport sector can clearly benefit from big data collected through sensors, GPS data and social media in particular. A smart use of big data supports governments in optimising multimodal transport and managing traffic flows, making our cities smarter. Citizens and companies can save time through the use of route planning support systems. The big data revolution brings about novel ways of understanding and addressing environmental challenges. A better use of globally available national and local datasets helps scientists in their research and enables policy-makers to make informed and evidence-based decisions to fight against climate change and reduce costs. Smart cities also host data centres adapting the power consumption of public buildings to the availability of renewable energy and other useful indicators. At the same time, our mobiles devices become smarter by integrating analytical tools to reduce our energy consumption and save money. Big data means the optimisation of operations on a real-time basis for the manufacturing industry. Key benefits of using big data analytics include boosting product quality through improved defect tracking, better manufacturing processes and optimised supply tracking. Big data enables the timely and appropriate delivery of products for consumers and more efficient processes, with cost savings, for business. With big data retail companies better know the needs and interests of costumers and, as a result, offer more personalised products. All of which leads back to big data. So far, though, only a small number of customers are investing in big data and trying to make use of it. Big marketing brands like Nike and Red Bull, who are always cutting edge, are investing in big data approaches to analytics and targeted marketing. The vast majority of businesses do not have the benefit of employing data scientists who can shape and model big data to drive better insight and, in turn, make smarter marketing decisions. But it’s clearly something that will gain visibility, underscoring the desire and need to distil more meaningful insight from vast amounts of data. These are truly exciting times to be in this industry. Marketing is undergoing a major transformation, as consumers and businesses redefine their relationship. This will help companies’ sustainable strategies. Digital marketing leaders are under increased pressure to stimulate growth through faster and smarter innovation. No one has unlimited resources, so it’s critical to invest in the right technologies and application at the right time. The service sector is now at the crossroad and it can capitalize by innovating strategy that can make it winners of the next decade for its sustainability.","excerpt":"A growing number of companies are finding their service businesses under threat. The culprits are members of a new wave of digital upstarts that capitalize on changes in technology, customer behaviour, and the availability of data to create innovative, customer-friendly alternatives to the services incumbents offer. The review is an attempt to discuss possible implications […]","categories":["IT Services"],"tags":[],"author_name":"Aegis School of Business and Telecommunication","publish_date":"2015-07-15T11:23:18","publication_year":"2015","word_count":1337,"keywords":["big data","data science","Go","API","artificial intelligence","AI","Git","analytics","GAN","R"],"extracted_tech_keywords":["AI","artificial intelligence","data science","analytics","R","Go","Git","API","big data","GAN"],"url":"https:\/\/analyticsindiamag.com\/it-services\/big-data-business-analytic-for-sustainability-of-competitive-advantage-in-service-organization\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10015228,"title":"Cognizant Acquires UK-Based AI &#038; ML Services Expert","content":"Cognizant has acquired Inawisdom, a UK-based privately-held consultancy expert specialising in artificial intelligence, machine learning and data analytics to help businesses make better, faster decisions, improving their business outcomes. However, the financial details of the deal were not disclosed. Focusing on delivering cloud-native, full-stack solutions, Inawisdom is an analytics and machine learning platform built using Amazon Web Services (AWS). And, with this acquisition of Inawisdom, Cognizant marked its ninth acquisition in 2020 to expand its capabilities for clients in the field of data and artificial intelligence, cloud, digital engineering, and Internet of Things. As a matter of fact, Cognizant has invested more than $1.1 billion this year alone for acquisitions in these areas. Founded in 2016, Inawisdom has operations in the UK and the Netherlands, and through this acquisition, the company will join in with Contino, another company acquired by Cognizant, last year, to provide end-to-end cloud-native AI and machine learning solutions. Speaking about the acquisition, Malcolm Frank, the President, Digital Business at Cognizant stated in the official release that businesses tend to succeed or fail depending on the speed and quality of their decision, ­and the best business decisions are currently informed by data and AI. He further stated that with Inawisdom’s skilled team integrated with Cognizant will further accelerate the company’s innovation on data modernisation and intelligent decision-making. Alongside, the clients of both the companies will be benefited from their shared, deep relationships with AWS and their combined expertise with AI, machine learning, cloud, and data analytics. Adding to that, Neil Miles, CEO, Inawisdom stated that the combined strength of the companies would further support customers in embedding data-driven decision-making into their organisations, increasing their speed to business value and long-term market differentiation. Also, “being part of Cognizant provides access to a wider, global network of AI and machine learning professionals, and a wealth of new capabilities that will expand our joint service offerings for successful data transformation,” said Miles in the official release. Inawisdom, with its current expertise, has leveraged AWS cloud technology to develop the Rapid Analytics and Machine Learning Platform (RAMP) in order to provide a continually evolving and reusable code repository and accelerate AI-driven business outcomes. Using RAMP along with its own agile consulting approach, Inawisdom assists its clients across multiple use cases — from enhancing supply chain efficiency and improving customer service to reducing operating costs and achieving results within weeks. Cognizant’s AWS Business Group, the largest AWS-certified practitioner community in the world, currently consists of more than 8,000 certified cloud professionals supporting clients. Doug Yeum, the Head of Global Partner Organisation, AWS, Inc. also stated that AWS empowers millions of customers around the globe to lower costs, become more agile, and innovate faster. And partnering with AWS will help Cognizant and Inawisdom to transform their data and analytics capabilities using AWS services. The companies joining forces will, in turn, be a win for the customers, and “we look forward to the industry and data science capabilities the Cognizant team will be able to offer our customers,” concluded Yeum.","excerpt":"Cognizant has acquired Inawisdom, a UK-based privately-held consultancy expert specialising in artificial intelligence, machine learning and data analytics to help businesses make better, faster decisions, improving their business outcomes. However, the financial details of the deal were not disclosed. Focusing on delivering cloud-native, full-stack solutions, Inawisdom is an analytics and machine learning platform built using […]","categories":["AI News"],"tags":["Cognizant"],"author_name":"Sejuti Das","publish_date":"2020-12-22T10:24:39","publication_year":"2020","word_count":506,"keywords":["data science","API","artificial intelligence","machine learning","AWS","AI","Cognizant","Git","RAG","analytics","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","data science","analytics","RAG","AWS","R","Git","API"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/cognizant-acquires-uk-based-ai-ml-services-expert\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":59445,"title":"Read The Memo Amazon’s CEO &#038; Founder Sent To Employees About COVID-19","content":"In a post dedicated to Amazon employees, CEO and founder Jeff Bezos outlined guidelines the tech giant is going to take to keep their personnel safe. In the same message, he has also encouraged those who have been laid off during this epidemic to consider working for the company. Bezos shared the letter on his personal Instagram profile on Saturday saying, “This isn’t business as usual, and it’s a time of great stress and uncertainty. It’s also a moment in time when the work we’re doing is itself most critical.” https:\/\/www.instagram.com\/p\/B-A5UI_nNrD\/ The memo comes at that point when businesses, around the globe, are struggling to combat the spreading of this Coronavirus outbreak, which has so far killed more than 12,000 people around the world and has infected around 300,000 people. In his post, Bezos, noted his employees about how the company “changed its logistics, transportation, supply chain, purchasing, and third party seller processes to prioritise stocking and delivering essential items like household staples, sanitisers, baby formula, and medical supplies.” The company is also providing vital service to people everywhere, especially to elderlies, and the ones who are vulnerable, along with the ones depending on Amazon. Amazon is also trying to provide masks to its employees who aren’t able to work from home — the company has placed orders for “millions”, however, the company has been facing problems as the masks are in short supply and the governments are currently directing them to facilities that require them the most — like vulnerable areas and hospitals. Bezos, in the memo, also reiterated that the company is going to hire 100,000 new roles and also mentioned about raising wages for their hourly workers. He further added, “My own time and thinking is now wholly focused on COVID-19 and on how Amazon can best play its role. I want you to know Amazon will continue to do its part, and we won’t stop looking for new opportunities to help.” And ended with, “Please take care of yourselves and your loved ones. I know that we’re going to get through this, together.” Read the whole memo here.","excerpt":"In a post dedicated to Amazon employees, CEO and founder Jeff Bezos outlined guidelines the tech giant is going to take to keep their personnel safe. In the same message, he has also encouraged those who have been laid off during this epidemic to consider working for the company. Bezos shared the letter on his […]","categories":["Global Tech"],"tags":["Amazon","Covid Dataset","covid-19","jeff bezos"],"author_name":"Sejuti Das","publish_date":"2020-03-23T11:23:17","publication_year":"2020","word_count":352,"keywords":["Go","Covid Dataset","covid-19","AI","programming_languages:R","Amazon","programming_languages:Go","RAG","ViT","jeff bezos","R"],"extracted_tech_keywords":["AI","RAG","R","Go","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/read-the-memo-amazons-ceo-founder-sent-to-employees-about-covid-19\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10094976,"title":"Council Post: Leadership Amidst Adversity &#8211; “When the going gets tough, the tough gets going”","content":"Building and leading teams are largely fun and exciting. Leadership during the “tech-boom” in the market was all about growth – be it the compensation, role, responsibilities they were all in abundance. The beginning of last year with the crazy hiring that was going on, what was needed from a leader was speed & impatience. Hockey stick growth everywhere, valuations soaring through the roofs and we were riding high on the bull run. Cut to 2023 – It’s a different story. You have got ported from a fairytale to a war zone. Leaders are at the centre of all the battles and the blood-shed. Worse for few their team has not moved with them from the fairy land. Welcome to the world of technology, these cycles are here to say. Leadership in this world is about keeping the cool whether you are riding the high tide or withstanding the rough waves. For many in leadership, this may be the first time they are seeing the tough weather. They need to brace themselves for this change and take their people through these times There is a lot to be learnt during these adverse times. And there are lessons to be taught to the wider team by demonstrating extreme focus, resilience and courage. The market needs those leaders who could thrive in good and bad times. Leadership amidst adversity is less spoken about but that’s what is needed at these moments of crisis. The Need: Leadership during adverse times is a much needed skill and it has to be practiced and learnt. I am listing few reasons, why it is an important skill to learn Tech Cycles are here to stay: Even if we map tech recessions from dot com days of the 90’s, there are 3-4 cycles so far we have experienced. Any economic recession impacts technology the fastest. The current one has hit technology the deepest. The layoffs and the pay-cuts have been the maximum in technology than other industries. We will come out of this cycle and there will be another one post that. It is going to be cyclical and hence being prepared for these cycles is the right thing to do than wishing them away Fair Weather Leaders (FWT):  Pretty much like “Fair weather friends”, those leaders who enjoy building only during good times are quickly eliminated. The FWT are good leaders, they can get the team together achieve during a period of abundance when resources are unlimited.However they easily get frustrated when constraints prop up. Suddenly when there are limitations in people and resources, they are not able to think freely and feel suffocated. They let the organization, their team and themselves down too fast. A great leader during these times, reflects, pauses and guides the team in the right direction. She can alter the strategy and switch to a leadership mode that gets the best from the constraints Opportunity to Innovate: Constraints are the best time to innovate. When there are less people available or we have constrained infrastructure (cloud cost for instance) there are plenty of opportunities to discover new ways of doing things. Be it through automation or optimizing architecture there  is a necessity to do more with less and then build an innovative culture in the teams. These are also the best times that bring out the best in the team. You would see some true “heroes” rising to the occasion and supporting at the critical time Professional Learning Curves: Times of crises are time to invest in learning. Many organizations (even those doing layoffs) increase their L&D budget precisely for the same reason. These are times to invest in individual growth. Using these times to build & develop skill sets will help the individual, the team and the organization greatly. Spend time with the team to understand the career needs and as a leader chart out the plans for their growth Personal growth : “Smooth seas do not make skillful sailors; it is the withering storms that test their mettle.” – African proverb . It is during tough times that we get an opportunity to become stronger and emotionally more mature. These tough times are also great times of learning personally as well. Most companies are facing turbulence – there are changes everywhere from org structure to goals to way of operation. Going through this phase, will teach you patience & emotional stability and you emerge with a better version of yourself Behaviors that help during adversity: Immense Empathy:  Lot of times in organizations we are dealing with individuals with different priorities. Each person has their sense of priorities and most conflicts arise because of the lack of clarity on each other’s priorities. It’s rarely to do with the person itself. During adverse times, these conflicts surface many notches higher. If we exercise empathy during these times, leaders could solve the situation way differently than they would have if they just go by process. Empathy is about being intelligent and trying to understand the situation and the person more than what is getting presented. Empathy helps you to “get” the person despite disagreements. A much needed trait in today’s leaders and it will help simply cut through all issues and build better bonds. No amount of process and protocols will help as much as empathy helps. One thing to remember is exercise Empathy in all directions – Downwards, Upwards & Sidewards Resilience:  One of the instant reactions when things are not as favorable to you as it was, is to quit. We need to remember during tough times anywhere – there are going to be plenty of bad days and too little good days. People are going to be less accessible, there are going to be more demands, the expectations are going to be more and the rewards less. These sudden changes can cause frustrations and thought on the top of mind – would be to leave and find a better environment. But if you take a step back and reflect – the path gets clear. You would know where to focus. You would rather “navigate” certain issues and spend time to solve only deeper problems. Even if the problems are overwhelming to start with, just staying at it and actively thinking about it will help you win over the problems. Most early-stage entrepreneurs and founders have resilience as “muscle memory”. At early stages, they are generally hearing disproportionate negatives than positives about their company across the board and the only thing that keeps them going is immense trust in what they do and the hope that they can do it. Be Grateful: Possibly the rarely invoked emotion in a career is of gratefulness. With so many things going as not expected and dynamics of different people, we often only see the negatives. However behind all these issues are opportunities to build something, to serve people, to define new processes and to crack new problems. Purpose is often underestimated and when that comes alive, we will do better if we are thankful for the same. Gratitude also feeds into humility, grounds you to reality helps you reflect and solve a problem rather than complaining about it Inspire People:  When adversity strikes, there is instant refusal of the reality and then a fury of why this happens and a blame of slew of people around.. However once the acceptance kicks in, it’s time to pull up the socks and guide the team to navigate the situation. Adversity is a time where inspirational leadership is needed the most. Like how teaching also helps in learning for the teacher, inspiring people helps you motivate yourself. Find that purpose that can inspire people – it is not just words, but the gap & the new problems that leaders can identify to inspire the team to new heights of  professional learning. Connecting with the aforesaid point of gratitude, is indeed an opportunity to be of service to others and hence use that to the best of your ability to help your team see the bigger picture. Life is not meant to be happy always, but it’s meant to be lived to the fullest. Expect adversity and don’t run away from it, navigate and conquer for yourself and for those who lookup to you. Life will have a new meaning then. This article is written by a member of the AIM Leaders Council. AIM Leaders Council is an invitation-only forum of senior executives in the Data Science and Analytics industry. To check if you are eligible for a membership, please fill out the form here.","excerpt":"The market needs those leaders who could thrive in good and bad times. Leadership amidst adversity is less spoken about but that’s what is needed at these moments of crisis.","categories":["AI Features"],"tags":["innovation","leadership"],"author_name":"Mathangi Sri","publish_date":"2023-06-13T14:00:00","publication_year":"2023","word_count":1428,"keywords":["data science","Go","AI","innovation","RAG","Aim","ViT","analytics","Rust","GAN","R","leadership"],"extracted_tech_keywords":["AI","data science","analytics","Aim","RAG","R","Go","Rust","GAN","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/council-post-leadership-amidst-adversity-when-the-going-gets-tough-the-tough-gets-going\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10050366,"title":"What Does Machine Learning Embedding Mean?","content":"Embedding is the process of converting high-dimensional data to low-dimensional data in the form of a vector in such a way that the two are semantically similar. In its literal sense, “embedding” refers to an extract (portion) of anything. Generally, embeddings improve the efficiency and usability of machine learning models and can be utilised with other types of models as well. When dealing with massive amounts of data to train, building machine learning models is a nuisance. As a result, embedding comes into play. Basic Purpose of Embeddings Deep Neural Network models can be trained (learned) to generate embeddings and then utilised to build another embedding for a different set of data. Embeddings of neural networks are advantageous because they can lower the dimensionality of categorical variables and represent them meaningfully in the altered space. Three basic purposes exist for neural network embeddings: Locating the embedding space’s nearest neighbours. These can be used to provide suggestions based on the user’s interests or cluster classifications.As input to a machine learning model for the purpose of performing a supervised task.For the purpose of visualising concepts and the relationships between categories. Benefits of Embedding Embedding can be beneficial in a variety of circumstances in machine learning. This has been demonstrated to be quite beneficial in conjunction with a collaborative filtering mechanism in a recommendation system. The purpose of item similarity use cases is to aid in the development of such systems. Another goal is to keep data as simple as possible for training and prediction. After embedding, the performance of the machine learning model improved dramatically. The only disadvantage is that embedding reduces the model’s interpretability. In an ideal world, an embedding captures some of the input’s semantics by clustering semantically comparable inputs in the embedding space. There are numerous strategies for producing embeddings in a deep neural network, and the strategy you use is totally dependent on your purpose. The following are the objectives. Similarity checkSearch and retrieval of imagesRecommendations SystemFor text, use Word2Vec.Songs that are similarReduce the dimensions of high-dimensional input dataInstead of employing one-hot encoding, a huge number of categorical variables can be compressed.Eliminate sparsity; when the majority of data points are zeros, it is advised that they be converted to meaningful lower dimension data points.Multimodal translationCaptioning of images Text Embedding The Text embedding block converts a string of characters to a vector of real values. The term “embedding” refers to the fact that this technique produces a space for the text to be embedded. The Text embedding block is inextricably linked to the Datasets view’s text encoding. They are integrated into the same procedure while performing sentiment analysis. A Text embedding block can be used only immediately following an Input block that requires the selection of a text encoded feature. Ascertain that the language model you chose corresponds to the language model selected when text encoding was established. How does it work? Text encoding converts plain text to tokens. This method decodes a stream of text into words using the language model specified. Several models—NNLM, GloVe, ELMo, and Word2vec—are meant to learn word embeddings, which are real-valued feature vectors for each word. Image Embedding Image embedding reads images and uploads or evaluates them on a distant server or locally. Each image is assigned a feature vector using deep learning algorithms. It returns a data table that has been augmented with additional columns (image descriptors). Image embedding includes a variety of embedders, each of which has been trained for a specific task. Images are either transmitted to a server or assessed locally on the user’s computer, at which point vector representations are created. The SqueezeNet embedder enables a quick review on the user’s machine without the need for an internet connection. Conclusion Neural network embeddings are low-dimensional continuous vector representations of discrete data that are learned. These embeddings overcome the restrictions of traditional encoding methods and can be utilised for a variety of tasks, including locating nearest neighbours, supplying data to another model, and creating visualisations. While many topics in deep learning are discussed in academic terms, neural network embeddings are obvious and reasonably easy to execute. Embeddings are a powerful technique for dealing with discrete variables and a practical application of deep learning.","excerpt":"An embedding is a low-dimensional translation of a high-dimensional vector.","categories":["Deep Tech"],"tags":["Machine Learning","machine learning document classification","Neural Networks","recommendation system","Word2Vec"],"author_name":"Dr. Nivash Jeevanandam","publish_date":"2021-10-05T13:00:00","publication_year":"2021","word_count":704,"keywords":["Go","machine learning","recommendation system","machine learning document classification","AI","neural network","sentiment analysis","Modal","programming_languages:R","Machine Learning","programming_languages:Go","deep learning","R","Word2Vec","Neural Networks"],"extracted_tech_keywords":["AI","machine learning","deep learning","neural network","sentiment analysis","R","Go","Modal","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/machine-learning-embedding\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10097948,"title":"Data Science Hiring and Interview Process at Instahyre","content":"Halfway into 2023, and over two lakh employees world over have already been laid off. Amidst this crisis crippling the job market, AI-powered HRTech platform Instahyre is making sure to provide the best of opportunities. In solving the pain of millions of job seekers, Instahyre’s data science team has successfully tackled one of their primary challenges by optimising the job-matching process through ‘Instamatch’ their proprietary recommendation system. It improves the efficiency and effectiveness of the job search experience for both job seekers and employers. “Instamatch has changed how companies approach hiring, changing the modus operandi from mass emails, keyword search, and unanswered phone calls to a holistic data-driven, tech-based candidate personality and company DNA mapping, which has taken candidate experience and hiring conversions to a whole new level, reducing the time and cost to hire drastically,” said Sarbojit Mallick, cofounder of Instahyre, in an exclusive conversation with Analytics India Magazine. Founded in the year 2017 by Aditya Rajgarhia and Mallick, Instahyre’s adept use of AI, ML and data science in its operations has resulted in an optimised recruitment platform that provides personalised job matches, streamlined candidate evaluation, and improved efficiency for recruiters and candidates alike. The company boasts about a 70% reduction in time to hire and cutting costs by thrice compared to traditional methods. With over 10,000 companies benefiting from their services and a staggering 40 million candidates on their platform, Instahyre has earned the trust of major industry players such as Amazon, Google, PayPal, Salesforce, Walmart, Oracle, Razorpay, Paytm, PhonePe, JP Morgan, Adobe, and Myntra. Inside the AI and ML Operations of Instahyre In its operations, Instahyre effectively implements AI and ML to harness the power of data science and derive valuable insights. A prominent area where data science is applied at Instahyre is in candidate-company matching. Using InstaMatch, the data science team ensures that job seekers are matched with companies based on a comprehensive set of factors, including skills, experience, and individual preferences. This results in a more precise and personalised job-matching experience for users. Additionally, Instahyre employs natural language processing (NLP) and machine learning algorithms to parse and analyse resumes and allows the platform to extract relevant information from resumes, allowing for a streamlined and time-saving candidate evaluation process for both job seekers and employers. The platform has been utilising generative AI since its inception six years ago. Moreover, Instahyre assists recruiters with tools like Instahyre Talent Insights, which provides a global overview of the talent pool for each job post. The AI-driven approach automates various aspects of the hiring process, including candidate sourcing, shortlisting, and scheduling interviews. It also conducts deliberate screening and assessment, leading to effective candidate evaluation and offer rollouts for chosen individuals. Tech Stack Employed Instahyre uses various technology capabilities, including an Application Tracking System (ATS) Integration. It is integrated with widely-used applicant tracking systems used by employers and recruiters. This facilitates a smooth transfer of candidate data, updates application statuses, and streamlines the entire hiring process for the company. Furthermore, the data science team at the company employs a range of tools, applications, and frameworks to tackle challenges and make informed decisions. These resources encompass MySQL, Python, Java, NLP, and other relevant technologies. By combining these tools, they aim to excel in data-driven endeavours and maintain a dynamic work environment. Interview Process Instahyre seeks potential candidates with specific expertise in different domains. For the ML Engineer role, they require proficiency in Python, NLP, and other deep learning concepts. For the Data Engineer position, the desired skills encompass Python, Java, and Scala, as well as experience with technologies like Hadoop, Spark, Kafka, MySQL, MongoDB, Cassandra, AWS, and Azure. Their data science hiring process involves multiple steps to find the right candidates that begins with a thorough review of resumes, focusing on educational background, work experience, and essential skills in mathematics, statistics, programming, and machine learning. Shortlisted candidates then undergo a technical assessment, evaluating their abilities in data analysis, programming languages, statistical modeling, and problem-solving. Successful candidates proceed to a technical interview, where their expertise and problem-solving capabilities are extensively examined. Lastly, behavioural interviews assess candidates’ communication skills, problem-solving approach, and cultural fit within the collaborative team environment. Expectations Once selected, candidates joining the data science team at Instahyre can expect a dynamic and intellectually stimulating environment. They will have the opportunity to work on cutting-edge technologies, solve challenging problems, and collaborate with a highly skilled and diverse team. The company provides access to the latest tools and resources to support their work and encourages continuous learning and professional development. “In return, Instahyre expects candidates to have a solid foundation in data science, including a strong understanding of statistics, mathematics, and machine learning algorithms, added Mallick. Excellent programming skills, strong analytical thinking, problem-solving abilities, and effective communication of complex ideas are also highly valued qualities expected from candidates joining the data science team. Mistakes to Avoid When interviewing for a data science job at Instahyre, candidates should avoid some common mistakes that can hamper their chances. One mistake is not showing practical experience and how their work has made a real-world impact. Candidates should share specific project examples, challenges they faced, and the outcomes they achieved. Another mistake is not preparing well for technical questions or lacking a good understanding of core data science principles. Being well-prepared and showing mastery of important concepts is important to impress the interviewers. Candidates should highlight their practical experience with different datasets and explain how their work has made a difference. Instahyre values innovation, creativity, and a growth mindset, so applicants should try to embody these qualities Work Culture “Our work culture is positive and the best part is there is no micromanagement. Employees are trusted and given autonomy to handle their tasks, leading to increased productivity and accountability,” added Mollick. The company follows a remote work policy, allowing employees to work from their preferred locations, allowing Instahyre to tap into talent from diverse geographic areas. “What sets Instahyre apart from its competitors, especially in terms of working with the data science team, is its collaborative and cross-functional approach, he added. The company encourages collaboration between data science and other teams, fostering diverse perspectives and knowledge-sharing to solve complex problems. It also places a strong emphasis on innovation and continuous learning, offering opportunities for professional development and research activities. So if you want to work on impactful projects that directly benefit users and the recruitment industry as a whole but also grow professionally and personally, Instahyre is the place for you. Check out their careers page now. Read more: Data Science Hiring Process at Naukri.com","excerpt":"The company’s data science team uses diverse tools like MySQL, Python, Java, and NLP to excel in data-driven efforts and maintain a dynamic work environment","categories":["AI Hirings"],"tags":["Career","Data Science Hiring","Top Trend"],"author_name":"Shritama Saha","publish_date":"2024-07-29T15:31:33","publication_year":"2024","word_count":1098,"keywords":["Top Trend","data science","machine learning","AI","ML","RAG","NLP","Data Science Hiring","Aim","deep learning","analytics","generative AI","Career"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","NLP","data science","analytics","generative AI","Aim","RAG"],"url":"https:\/\/analyticsindiamag.com\/ai-hiring\/data-science-hiring-process-at-instahyre\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10098880,"title":"Python in Excel Comes with a Twist","content":"Excel spreadsheets just got a major overhaul. Microsoft has announced a public preview of Python in Excel. Developers or data analysts now would not have to install any extra software to access the functionality, as Excel’s built-in connectors and Power Query will come bundled with Python integration. Microsoft has also added a PY function for Python data to be available within the grid on the spreadsheets. In the blog, Stefan Kinnestrand from Microsoft explains that now users will be able to do advanced data analysis within the familiar Excel interface leveraging Python, which would be available on the Excel ribbon. “You can manipulate and explore data in Excel using Python plots and libraries, and then use Excel’s formulas, charts and PivotTables to further refine your insights,” he added. This announcement comes in a partnership with Anaconda, a leading Python repository for enterprises which will include libraries such as pandas, statsmodels, seaborn, and Matplotlib. Microsoft went the cloud way here by using Anaconda Python distribution on Azure. The features are rolling out in the Windows beta channel on Microsoft 365 Insiders and will only be available on the desktop version of Excel, and run on Microsoft Cloud. Why the cloud way? It is worth noting that developers and data analysts have mixed feelings about this release. Integrating Python in Excel is something that developers have been trying to do for a long time by utilising Pandas read excel, OpenPyXL, PyXLL. But with this native integration by Microsoft, advanced spreadsheet users can integrate scripts in Python language and their Excel formulas in a single workbook without any additional software. This would also allow shareable experiences of a single notebook over the cloud. The only downside that people have been concerned about is why it is wholly running on the cloud and not locally. Python can run perfectly well locally now and does not require a Microsoft Cloud connection to perform tasks. Though there are essential libraries being offered through Azure Cloud, since there is no option of running it locally, people are opting out of this change. In a Reddit discussion, a user said, “Python is such a lightweight runtime anyway, why not just include it inside the software instead of requiring internet which is probably slower than running the code natively even on a low-end laptop.” Moreover, some users argue that running on the internet has possibly made scripting on Google Docs a nightmare. On the other hand, people say that though Python is lightweight, running libraries like Scipy and Matplotlib requires heavy computation. To make up for this, Microsoft has integrated it within its cloud services, making it widely available. Or as someone pointed out, it is merely to earn some bucks on the cloud for Microsoft by getting more people to subscribe to their cloud services. Well, whatever works for now People have been arguing for a long time that Python has been eating up Excel’s market share for a long time, even calling it as dead. But people still find Excel most comfortable for data analysis. Undoubtedly, integrating Python in Excel marks as a huge step for people trying to leverage Python when working on Excel, and would single handedly modernise data analysis on the software. Looks like people saying Excel is not a programming language are going to return to it soon and people on Excel would start learning Python soon. Even then, hosting it solely on the cloud brings up the point of security and data privacy, even after partnering with Anaconda for it. To address the privacy issue, Microsoft has said that they are providing “enterprise-level security”, which means that the code would be running on cloud isolated containers and would not have network access, which would arguably still be dicey for a lot of customers. “The cloud part is going to be a huge deal breaker in so many industries. This will automatically be blocked by default at my work place for certain,” said a user on HackerNews. Companies that allow the use of Excel and Python locally, and want to integrate both functionalities in one framework, would still not be happy by shipping their data and Python code on a server outside of theirs. This marks as a significant step for the future where running coding generative AI applications on Excel would be a thing. In January, Microsoft had announced that they would be experimenting with GPT into their office applications. Interestingly Microsoft has been following this path all this while – acquiring Github – acquiring OpenAI that helped them make Copilot for writing Python code – Code interpreter on ChatGPT – AI assistance on PowerBI – to now Python in Excel. What’s Next? Will it be AI writing Python on our excel sheets? Exciting times ahead! No matter how much we dislike Microsoft Excel, we find ourselves reaching for it every day to perform some quick data analysis or put together an ad-hoc report. That’s the power of Excel – it gets the job done efficiently. Microsoft is giving Excel superpowers with the addition of Python. Data analysts rejoice – you’ll soon be able to unleash the full capability of Python right inside your familiar Excel environment. Say goodbye to tedious data munging. With libraries like Pandas, manipulating large datasets for cleansing, transformation and exploration just got exponentially easier. Need advanced analytics? You’ll be able to build predictive models directly on Excel data using scikit-learn. How about visualizing exploratory insights? Create publication-ready visuals with Matplotlib, no more exporting back and forth. Don’t even get me started on natural language processing or geospatial capabilities. The possibilities are endless. Of course, concerns around cloud exclusivity and performance are valid. But if Microsoft can address these issues, we may see a new age of Excel where everyday users can perform cutting-edge analyses previously restricted to coders alone. Exciting times ahead as we seamlessly blend the past and future of data work.","excerpt":"Python works perfectly well locally, but Microsoft wants it to run only on its cloud","categories":["Deep Tech"],"tags":["Microsoft"],"author_name":"Mohit Pandey","publish_date":"2023-08-23T12:48:57","publication_year":"2023","word_count":986,"keywords":["ChatGPT","scikit-learn","OpenAI","AI","ML","Seaborn","generative AI","analytics","Matplotlib","Microsoft","Pandas"],"extracted_tech_keywords":["AI","ML","analytics","generative AI","ChatGPT","OpenAI","scikit-learn","Pandas","Matplotlib","Seaborn"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/python-in-excel-comes-with-a-twist\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":43254,"title":"IIT-Madras&#8217; Team Avishkar Reaches The Final Round In SpaceX&#8217;s Hyperloop Pod Competition","content":"IIT-Madras’ team Avishkar, the only Asian participants to the Hyperloop Pod Competition hosted by space transportation and exploration company SpaceX, have now made it all the way to the finals. The noted competition which is closely fought by students from all over the world saw a total participation of about 1,500 teams globally. Their goal was to create a “pod” or a travelling vessel for the hyperloop concept. Avishkar Hyperloop is the student team from Centre For Innovation (CFI) – IIT Madras, working on an indigenous design and development for building the first-ever self-propelled, completely autonomous Hyperloop Pod in India. The team is working with a vision to develop technologies for a future mode of high-speed transportation with applications in various fields including Defence, Logistics and Aerospace Industry, among others. Speaking to us at a previous occasion, the Dr SR Chakravarthy, Faculty Advisor, Avishkar Hyperloop, and Professor, Department of Aerospace Engineering at IIT Madras, had said, “The team has taken strenuous technical efforts at every stage, and has been quite professional about executing the project. This is an important milestone for the future transportation technology in our country.” Hyperloop is the fifth mode of transportation, a high-speed train that travels in a near-vacuum tube. The reduced air resistance allows the capsule inside the tube to reach speeds of more than 1,000 kmph. Elon Musk, Founder of SpaceX, proposed the idea of Hyperloop to the world in 2013 through whitepaper Hyperloop Alpha. The Pod developed by Avishkar Team spans about 3 metres in length and weighs around 120 kilos. The electric propulsion employs BLDC motors and wheel drive. The team started the design by brainstorming on how the pod would be built, beginning with vehicle kinematics and dynamics, motors, power system and braking, stability, and finally designing the CFRP chassis and structures.","excerpt":"IIT-Madras’ team Avishkar, the only Asian participants to the Hyperloop Pod Competition hosted by space transportation and exploration company SpaceX, have now made it all the way to the finals. The noted competition which is closely fought by students from all over the world saw a total participation of about 1,500 teams globally. Their goal […]","categories":["AI News"],"tags":["IIT Madras","spacex"],"author_name":"Prajakta Hebbar","publish_date":"2019-07-24T13:43:04","publication_year":"2019","word_count":301,"keywords":["Go","programming_languages:R","AI","IIT Madras","innovation","programming_languages:Go","spacex","R"],"extracted_tech_keywords":["AI","R","Go","innovation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/iit-madras-team-avishkar-reaches-the-final-round-in-spacexs-hyperloop-pod-competition\/","complexity_score":2,"technical_depth":6,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":29750,"title":"Finding The Right Mentor In Data Science Is Just As Important As Your Analytics Education","content":"It is a question that has often befuddled many aspiring professionals who are looking to enter the data science field. What should one do? Dive into a self-paced Python or R course, or follow a curated data science-related curriculum which provides a good base in analytical and quantitative techniques? However, there is a new trend gaining momentum in data science education finding a senior industry professional who is also passionate about mentoring and can give valuable tips. Workplace mentoring has become the norm in many enterprises and over the last two decades, many organisations have started formal mentorship programmes to help freshers cross-over or build skills in a specific area. There is a similar trend in data science education where ed-tech startups hire senior industry veterans who can give one-on-one guidance about projects, courses and career advice. These mentors are usually available for 30-minute online sessions with students and help them set their target learning goals and discuss how to achieve them in an optimal manner. These mentors also answer subject matter questions, provide feedback on projects, and career advice. Most members in the data science community want to help people starting out both technically and with career guidance. Why Mentoring Is Crucial In Data Science Mentoring bridges the knowledge gap, improve their skills and the understanding of using data science for business. It also allows students to build a relationship and test his\/her work readiness before embarking on a full-time job. Both mentorships and internships work in a similar way — they act like a bridge programme, enabling students to retool according to the industry requirements. Mentorship is also beneficial for both sides — students gain valuable tips and understand how to move into these careers and mentors get a talent pool of keen, industry-ready graduates. For most companies, mentorships can act as a first filter for data-centric roles. Most newbies are unclear about the role expected of them. Mentors can explain how they can fit into the data science team and how their analytics efforts will pay off in the organisation. Another viewpoint is that the market is flooded with candidates who have the basic skills and a modest portfolio of projects. While this means the pool for entry-level candidates is increasing, it also gives a chance to organisations to be picky with the choice of candidates. In this scenario, mentorship can give candidates a definitive edge. Recruiters and data science leaders reveal that the field has become so competitive that it takes more than just a bunch of basic skills to get off the ground. In fact, guidance and input from an experienced industry member or senior can help candidates level up and dish out industry-ready production code. While nano degrees are good, it won’t help one become employable in 2018 or beyond. And that’s why mentorship component plays a pivotal role in improving your chances of landing a suitable position in the data analytics field. A key reason why mentorship-based programmes work well is that candidates won’t get stuck in the MOOC cycle wherein they have the same learning path or the cookie cutter approach as any other candidates. For example, Python > Pandas > sklearn > SQL  and this won’t help you beat the competition. It helps to have a mentor who can help you gain more credibility and traction in the data science job market. How Does Mentorship Work? Usually, ed-tech firms in India look for experienced industry professionals who can work with young aspirants or professionals on a weekly basis. The best way is getting experts to work with a group of students on an industry project and in return, mentors are paid a small fee. Globally, the idea of income share is also gaining ground. Income share works in a way where the mentor gets a portion of the mentee’s salary when he\/she lands a job. So there is no upfront cost involved in the process. Jeremie Harris, co-founder of SharpestMinds, a programme that connects aspiring candidates with senior data scientists and allows them to get mentored free until the candidate lands a job revealed the percentage of income share doesn’t exceed 9 percent.","excerpt":"It is a question that has often befuddled many aspiring professionals who are looking to enter the data science field. What should one do? Dive into a self-paced Python or R course, or follow a curated data science-related curriculum which provides a good base in analytical and quantitative techniques? However, there is a new trend […]","categories":[],"tags":["analytics training","data science mentorship","mentors in data science"],"author_name":"Richa Bhatia","publish_date":"2018-10-30T12:34:31","publication_year":"2018","word_count":692,"keywords":["data science","Go","startup","AI","data science mentorship","R","mentors in data science","Python","analytics","SQL","GAN","Pandas","analytics training"],"extracted_tech_keywords":["AI","data science","analytics","Pandas","Python","R","SQL","Go","GAN","startup"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/finding-the-right-mentor-in-data-science-is-just-as-important-as-your-analytics-education\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":6,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10067803,"title":"Oh boy, is JP Morgan wrong?","content":"Earlier this week, the global brokerage firm JP Morgan ‘downgraded’ the Indian information technology (IT) sector over margin and revenue concerns, alongside surging inflation and supply chain issues in the Russia-Ukraine tussle. The global brokerage firm has downgraded Tata Consultancy Services, HCL Technology, Wipro, and L&T Technology to ‘underweight’ from ‘neutral’ and slashed its target price by 15-21 per cent, but maintained an ‘overweight’ rating on Infosys, Mphasis, and Tech Mahindra, with reduced target prices. Why, though? The firm said that the Indian IT growth was accelerating till the third quarter of 2022 and has begun to slow down from the fourth quarter, which is likely to worsen into FY23 due to the high attrition rate and talent management that has pushed up costs of hiring and retaining employees. JP Morgan’s Ankur Rudra and Bhavik Mehta said that the rising margin headwinds in the near term and revenue headwinds in the medium term from potential macro slowdown would mean that the sector’s earnings upgrade cycle is behind. Goldman Sachs also warned that there was a 30 per cent chance the US economy would tip into recession over the next two years due to aggressive interest rate hikes by the US federal reserve. Other firms such as Wells Fargo and Morgan Stanley believe a recession in the US is around the corner as the US Fed shows no signs of easing its fight against high inflation. Meanwhile, Kotak Equities said ‘what is priced into stock’ is a risk to margins. What is not priced in is ‘economic recession.’ The company has trimmed its fair value targets for nine IT stocks under its coverage by 2-14 per cent, assuming a moderate slowdown in demand for IT services. S&P Global rating slashed India’s growth forecast to 7.3 per cent from 7.8 per cent for FY23 on rising inflationary pressure and the prolonged Russia-Ukraine war. Wait, a twist Interestingly, the US market contributes 40-78 per cent of revenues for Indian IT companies. This includes TCS, Wipro, HCL Technologies, Infosys, and Tech Mahindra, having more than 50 per cent exposure. On the bright side, this can also be a blessing in disguise for Indian IT companies, as the majority of them are likely to cut costs abroad and outsource work to countries like India, the Philippines and others because of the cost benefits\/advantage. In addition, some experts said that thanks to the recession and rising inflation, companies can now focus on offering better services to their customers, as the innovations and other R&D related work takes a backseat. Is the Indian IT sector recession-proof? The Indian IT sector has always had higher resistance than the rest of the world. This includes the 2008 financial crisis, the dot-com bubble, the pandemic, and others. Moreover, the fundamentals of some of these IT companies are at an all-time high. Here’s a quick overview of the top IT firms in India, showcasing the P\/E ratio and sales growth in the last five years. Kiran Kumar S said during 2021-22, Infosys clocked 19.7 per cent revenue growth – it is the fastest in 11 years. “Many more companies performed very well, the same way,” he added, saying that the Indian IT industry saved the economy during a Covid downturn. https:\/\/twitter.com\/KiranKS\/status\/1528792485350080512 If not the IT sector, then what are the alternatives? While JP Morgan has ‘downgraded’ the Indian sector to ‘underweight’ and slashed the target price of multiples by 10-20 per cent, the question is, which are the alternate sectors that are bound to generate revenues amid the recession? The JP Morgan report stated the bottom-up outlook remains positive from most services, software, and SaaS names YTD. The tech spending cycle remains buoyant structurally. “We feel there are more downside risks to current earnings assumptions,” it added.","excerpt":"The global brokerage firm has downgraded Tata Consultancy Services, HCL Technology, Wipro, and L&T Technology to ‘underweight’ from ‘neutral’ and slashed its target price by 15-21 per cent.","categories":["IT Services"],"tags":["goldman sachs","Infosys","JP Morgan","Tata Consultancy Services","Tech Mahindra","Wells Fargo","Wipro"],"author_name":"Amit Naik","publish_date":"2022-05-25T15:00:00","publication_year":"2022","word_count":626,"keywords":["Tata Consultancy Services","Wipro","JP Morgan","Tech Mahindra","Go","Infosys","AI","programming_languages:R","innovation","programming_languages:Go","RAG","goldman sachs","Wells Fargo","GAN","R"],"extracted_tech_keywords":["AI","RAG","R","Go","GAN","innovation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/oh-boy-is-jp-morgan-wrong\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10010894,"title":"7 Types Of Generative Models For Your Next Machine Learning Project","content":"Generative models have gained much popularity in recent years. These models help in handling missing information as well as treating with the variable-length sequences. Technically speaking, generative models deal with the models of distributions, defined over data points in some potentially high-dimensional space. In this article, we discuss seven types of generative models, which are listed below in alphabetical order- Autoregressive Models Autoregressive Model or AR model is when a value from a time series is regressed on previous values from that same time series. The order of an autoregression is the number of immediately preceding values in the series that are used to calculate the value at the present time. In simple words, Autoregressive models predict future values based on past values. These models are flexible at handling a wide range of different time-series patterns. Know more here. Bayesian Network Bayesian Network or Bayes Network is a generative probabilistic graphical model that allows efficient and effective representation of the joint probability distribution over a set of random variables. Bayes Network consists of two main parts, which are structure and parameters. The structure is a directed acyclic graph (DAG), and the parameters consist of conditional probability distributions associated with each node. This network can be used for various applications, such as time series prediction, anomaly detection, reasoning and other such. Know more here. Generative Adversarial Networks Generative Adversarial Networks or GANs are popular generative models that include two parts, generators and discriminators. This model works by estimating generative models via an adversarial process. The generative model captures the data distribution, and the discriminative model estimates the probability that a sample came from the training data rather than the generative model. GANs are one one of the trending generative models that have been used to create images of humans that do not exist\/ Know more here. Gaussian Mixture Model Gaussian Mixture Model is a generative probabilistic model, which assumes all the data points are generated from a mixture of a finite number of Gaussian distributions with unknown parameters. GMMs are commonly used as a parametric model of the probability distribution of features in a biometric system, which includes vocal-tract related spectral features in a speaker recognition system. Thus, GMM parameters are estimated from training data using the iterative Expectation-Maximisation (EM) algorithm or Maximum A Posteriori (MAP) estimation from a well-trained prior model. Know more here. Hidden Markov Model A Hidden Markov Model (HMM) is a statistical model that can be used to describe the evolution of observable events that depend on internal factors, which are not directly observable. The model is popularly known for their effectiveness in modelling the correlations between adjacent symbols, domains, or events, and they have been extensively used in various fields, especially in speech recognition and digital communication. A Hidden Markov Model consists of two stochastic processes, which are an invisible process of hidden states and a visible process of observable symbols. Know more here. Latent Dirichlet Allocation (LDA) Latent Dirichlet Allocation or LDA is a generative probabilistic model with collections of discrete data such as text corpora. LDA is a three-level hierarchical Bayesian model, in which each item of a collection is modelled as a finite mixture over an underlying set of topics. The model has applications to various problems, including collaborative filtering, content-based image retrieval, among others. Know more here. Variational Autoencoders (VAEs) Variational Autoencoders (VAEs) have been one of the most popular approaches to unsupervised learning of complicated distributions. They are built on top of standard function approximators, which are neural networks and can be trained with stochastic gradient descent. The application of VAEs includes generating various kinds of complicated data, including handwritten digits, faces, CIFAR images, predicting the future from static images and more. Know more here.","excerpt":"Generative models have gained much popularity in recent years. These models help in handling missing information as well as treating with the variable-length sequences. Technically speaking, generative models deal with the models of distributions, defined over data points in some potentially high-dimensional space. In this article, we discuss seven types of generative models, which are […]","categories":["AI Trends"],"tags":["different types of analytics","GANs","scikit learn","types of analytics"],"author_name":"Ambika Choudhury","publish_date":"2020-10-29T10:00:00","publication_year":"2020","word_count":624,"keywords":["Go","programming_languages:R","AI","neural network","programming_languages:Go","Git","VAE","anomaly detection","GANs","scikit learn","GAN","R","types of analytics","different types of analytics"],"extracted_tech_keywords":["AI","neural network","anomaly detection","R","Go","Git","VAE","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/7-types-of-generative-models-for-your-next-machine-learning-project\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":11487,"title":"8 Startups in India working on Blockchain Technology","content":"Blockchain is the most trending technological term today and is forecasted to disrupt various industries in coming future. The buzz is building up in India as well. Just last week, Mahindra and IBM joined hands to develop Blockchain Solution for Supply Chain Finance. Also, mid of this year, ICICI bank adopted Blockchain in a push for digitalizing banking. Yet, in the startup space, the activity seems to be relatively muted in India when compared globally. One reason can be that the technology is still in its early days and yet to mature. We presented a tutorial on Blockchain earlier this year. So, we decided to compile a list of some interesting startups in this space in India. Here they are: Auxesis Auxesis is a technical innovation company with advanced competency in FinTech and Blockchain Technology. Working for the development of new innovative technological solutions by providing a balance between technology and economic sustainability for the start-ups. Working for businesses to help them focus on their best what they do without the worry of their technical development and maintenance. Currently, company is working with IIT Bombay Entrepreneurship cell & European Incubator (SEED) start-ups projects from idea until there transformation Phase. Auxesis is constantly experimenting in collaboration with Blockchain Lab, London to find better use cases for other industries. Their current initiative includes Banking, Remittance, Insurance, Event and Ticketing industry while Pharmaceuticals, Luxury goods, Gambling are under consideration. Coinsecure Coinsecure was built with the motto of “Connecting India to Bitcoin”. Today we focus on “Building an entire eco – system for Bitcoin and Blockchain in India” and we have been able to accomplish this with the array of products and services that we offer. Apart from being the largest and best Bitcoin exchange, Coinsecure offers a Bitcoin wallet, A Blockchain solution (the Block explorer), merchant gateway, comprehensive API’s and a testnet service. Coinsecure intends to be the one stop shop for all Bitcoin and Blockchain related needs and accomplishes them through several partnerships across the globe. Coinsecure is India’s first and only Bitcoin company to have secured the highest funding round with 1.2 Mill USD and have been at the fore front of developments which are user friendly and secure! EzyRemit EzyRemit is an innovative FinTech and Blockchain platform and solutions company which is focused on changing the remittance market. Working at the intersection of technology, payments and banking industry from many years, we saw the many redundancies in the system and the need for change. A glimpse of the solution came with the appearance of Bitcoin and later Ethereum, offering different blockchain technologies we could work with to find our solutions. By pairing experience from the Payment, business process management & software application development, founders use blockchain technology to empower people around the world with free access to a global financial system & faster transaction. Hashcove HashCove is a Blockchain Products Company that helps evaluate, validate, design and create end to end solutions in Blockchain. HashCove is the Blockchain arm of uTrade Solutions. uTrade is a FinTech company providing multi-asset trading platform,low latency algos engine, risk management and Blockchain Solutions. Their products are widely used by global financial institutions including Brokers, Algo Firms, Forex Traders, Stock Exchanges and their end customers. KrypC KrypC is a FinTech solution and service provider focusing on bringing innovative solutions in the field of Blockchain & Digital Currency. Currently the team is developing and deploying industry use cases on the blockchain framework which would help organizations to integrate current processes to the DLT network. The goal of KrypC’s technology services is to help businesses understand the power and utility of Distributed Ledger Technology (DLT), assess the potential areas of application, provide technical framework & design and effective implementation of the technology solution. With vast experience in Digital Signature Certificates, Security and Cryptography, KrypC has multiple intellectual property in the areas of Mobile Wallet, Digital Currency, Security and Payments solution. MindDeft Technologies MindDeft was founded with the thought of making difference with IT services and providing the quality services to customers. Since our inception, the goal is to deliver solutions to complex problems that really makes difference to the customers. Apart from traditional technology stack, the team is very much open for adopting the new trends and technologies that is really making difference to solutions they are providing. Blockchain is one of the trend that they love, and they are in the progress of taking expertise of various Blockchain tools for example, Ethereum and HyperLedger. They have been doing extensive research for Blockchain and delivered the proof of concepts for various domains on how to implement the smart contracts with Blockchain tools. Sofocle Sofocle is a core software development company, focused on building innovative blockchain based solutions for Finance, Health and Retail industries. Our customized solutions enable enterprises to streamline their business process by enhancing security, authenticity and transparency of every transaction happening in the business ecosystem. Sofocle’s Blockchain solutions include: SofoSupply – SofoSupply is a Blockchain based solution which enables seamless supply chain financing. It is built on Hyperledger and uses Smart Contracts for digitizing processes. It helps a supplier to easily avail loan against approved invoices by the manufacture without the need of any manual paperwork. SofoTrade – SofoTrade is Blockchain based solution for Trade Finance that streamlines the international trade settlements. It offers a potential medium to exchange assets without centralized trusts or intermediaries. SofoTrade offers immutability and digital uniqueness by deploying Smart Contracts which are self-executing, triggered by the efficient exchange of digital data and eliminates the risk of double spending. SofoChain – SofoChain is a solution to record the quantity and transfer of assets as they move between supply chain nodes, like trailers, containers etc. by linking serial numbers, bar codes, RFID, to physical goods. It also enables verification of certifications and assigning certain properties to physical products, thus hereby eliminating fake or counterfeit goods\/invoices in the supply chain. Trestor Trestor is an India’s first blockchain startup, which has created “Trest” a secure, digital, store of value. Using the power of Trestor’s blockchain and their decentralised network of trustless nodes ‘Trests’ can be transferred directly from person to person anywhere across the globe. There is zero transfer fees, Trests are transferred within seconds, accounts can never be frozen and there are no paperwork or limits! Trestor network is designed keeping under banked in mind. Most think that poor just need a bank account and a payment card and once they get it, they will happily switch from cash to a centralized, freezable, chargeable, hackable payment system that the rest of us use. Maybe we should re-think this assumption. “Maybe what unbanked people really need is digital equivalent of cash instead of our banks and our way of banking”.","excerpt":"Blockchain is the most trending technological term today and is forecasted to disrupt various industries in coming future. The buzz is building up in India as well. Just last week, Mahindra and IBM joined hands to develop Blockchain Solution for Supply Chain Finance. Also, mid of this year, ICICI bank adopted Blockchain in a push […]","categories":["AI Features"],"tags":["bitcoin stock","blockchain companies","blockchain india","blockchain tutorial","Ethical Hacking"],"author_name":"Дарья","publish_date":"2016-12-05T06:30:11","publication_year":"2016","word_count":1126,"keywords":["Go","API","blockchain india","blockchain companies","AI","ML","Git","blockchain tutorial","Ray","bitcoin stock","ViT","Rust","GAN","Ethical Hacking","R"],"extracted_tech_keywords":["AI","ML","Ray","R","Go","Rust","Git","API","GAN","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/7-startups-india-working-blockchain-technology\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10100670,"title":"Infosys Partners with Microsoft to Drive Enterprise Generative AI Adoption","content":"Indian IT and consulting firm, Infosys announced its partnership with Microsoft aimed at accelerating the widespread adoption of generative AI across various industries. The partnership will leverage Infosys Topaz, Azure OpenAI Service, and Azure Cognitive Services to develop cutting-edge solutions that enhance enterprise functions through AI-enabled capabilities. The primary goal of this collaboration is to democratize data and intelligence within businesses, enabling increased productivity and revenue growth. Generative AI has created new possibilities for AI applications across industries, and Infosys is at the forefront, providing services and solutions in areas such as semantic search, document summarization, contact center transformation, AI-augmented software development, and marketing content creation. “Through our strategic collaboration with Microsoft, we will continue to lead the generative AI revolution, helping businesses amplify human potential and navigate their next step towards becoming AI-first enterprises,” said Balakrishna D. R. (Bali), executive vice president, Infosys. Infosys Topaz, in collaboration with Microsoft, will utilize Azure OpenAI Service and Azure Cognitive Services to enhance its capabilities further. This enhancement will enable enterprise customers to seamlessly transition from digital to AI solutions, improving operational efficiency, reducing turnaround time, future-proofing investments, and introducing new business models. “We’re pleased to expand our collaboration with Infosys to deliver innovative solutions, utilizing Azure OpenAI Service and Azure Cognitive Services, that will help customers develop new business models, and realize new revenue streams,” said Nicole Dezen, chief partner officer, Microsoft Corp. Recently, Infosys also announced its partnership with NVIDIA to integrate the NVIDIA AI Enterprise ecosystem, encompassing models, tools, runtimes, and GPU systems into Infosys Topaz  to create offerings that enable businesses worldwide to seamlessly incorporate generative AI into their operations, fostering unprecedented innovation and productivity gains.","excerpt":"Infosys Topaz, in collaboration with Microsoft, will utilize Azure OpenAI Service and Azure Cognitive Services","categories":["AI News"],"tags":["Infosys"],"author_name":"Siddharth Jindal","publish_date":"2023-09-26T16:55:59","publication_year":"2023","word_count":278,"keywords":["Go","semantic search","OpenAI","Infosys","AI","R","ML","RAG","Aim","generative AI","Azure"],"extracted_tech_keywords":["AI","ML","generative AI","OpenAI","Aim","RAG","semantic search","Azure","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/infosys-partners-with-microsoft-to-drive-enterprise-generative-ai-adoption\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10164063,"title":"OpenAI Thinks LLMs Can Earn $1M from Freelance Software Engineering Tasks","content":"OpenAI has introduced SWELancer, a new benchmark to test whether frontier large language models (LLMs) can successfully complete real-world freelance software engineering tasks—and even earn up to $1 million in total payouts. The evaluation is based on 1,488 freelance software engineering jobs from Upwork, collectively valued at $1 million. SWE-Lancer comprises over 1,400 software engineering tasks, with projects ranging from $50 bug fixes to $32,000 feature implementations. “Introducing SWE-Lancer: our most realistic coding benchmark to date. Still some limitations, but better than evals we had before,” said Tejal Patwardhan, who works on the benchmarks and preparedness team at OpenAI. These tasks are divided into independent engineering tasks, where models must complete technical work, and managerial decision-making tasks, where models evaluate and choose between implementation proposals. By mapping AI model performance to real-world monetary value, SWE-Lancer provides a crucial tool for studying the economic impact of AI in software development. More research can be accessed here. Anthropic, the company behind the Claude model series, also released a survey highlighting AI’s influence on the workplace. The findings revealed that approximately 36% of all occupations incorporate AI for at least a quarter of their tasks. Moreover, 57% of AI applications enhance human capabilities, while 43% focus on automation. However, only 4% of occupations rely on AI for at least 75% of their tasks. The study identified software development and technical writing as key areas where AI is utilised. In contrast, AI plays a minimal role in tasks that involve physical interaction with the environment.","excerpt":"OpenAI has introduced SWELancer, a new benchmark to test whether frontier large language models (LLMs) can successfully complete real-world freelance software engineering tasks—and even earn up to $1 million in total payouts. The evaluation is based on 1,488 freelance software engineering jobs from Upwork, collectively valued at $1 million. SWE-Lancer comprises over 1,400 software engineering […]","categories":["AI News"],"tags":["AI (Artificial Intelligence)","OpenAI"],"author_name":"Aditi Suresh","publish_date":"2025-02-19T11:11:33","publication_year":"2025","word_count":251,"keywords":["Anthropic","OpenAI","AI","programming_languages:R","llm_models:Claude","automation","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","OpenAI","Anthropic","R","automation","llm_models:Claude","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/openai-thinks-llms-can-earn-1m-from-freelance-software-engineering-tasks\/","complexity_score":3,"technical_depth":7,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":24950,"title":"What Is The Difference Between Virtual Reality, Augmented Reality And Mixed Reality?","content":"A post shared by Molly Gambardella (@mollygeeart) on May 4, 2018 at 12:21pm PDT Reality technologies which have been in the developmental phase for years, have now become mainstream products for marketers. Tech giants such as Microsoft, Apple and Facebook, among others, are betting big on mixed reality (MR), augmented reality (AR) and virtual reality (VR). Facebook CEO Mark Zuckerberg has reportedly been working hard to bring VR to billions of people. Microsoft is using mixed reality to combine virtual objects with the real world. Apple believes that in order to make AR mainstream, it needs to be integrated directly into iOS. Reportedly, Google and NBCUniversal have come together to enable VR experiences on YouTube-based shows. According to investment bank Woodside Capital, AR-related sales could reach $80 billion in 2022. Another report claims that worldwide shipments of AR and VR headsets are expected to grow at 58 percent CAGR. AR, VR and MR are no longer limited to just sci-fi movies — they are slowly changing education, entertainment, communication and various other industries and applications. But these reality technologies can sometimes get confusing in the eyes of the user. What Is Virtual Reality? VR is experiencing things that don’t really exist with the help of computers. VR creates a realistic three-dimensional environment which humans perceive as real. Imagine opening your eyes and seeing a computer-generated world all around you, one where you can move and interact with, as showcased in Steven Spielberg’s latest film Ready Player One (2018). The film depicts a future where VR technology can transport people into a realistic alternate universe. It basically refers to computer-generated environments that are designed to replicate a person’s physical presence in a specific environment that is designed to “feel” real. In pic: HTC Vive How To Experience Virtual Reality To experience this alternate reality, VR headsets are mandatory. Most VR headsets are connected to a computer, a gaming console or are standalone devices. Most standalone VR headsets like Google Cardboard work with smartphones. In pic: Google Cardboard For a better experience, you can use the VR headsets along with good quality headphones and other accessories like hand controllers and treadmills. The most popular PC-connected VR headsets available for consumers are Oculus Rift and HTC Vive. In pic: Oculus Rift What Is Augmented Reality? Unlike virtual reality which requires a headset to experience the digital world, augmented reality uses your existing environment and overlays digital information on top of it. Have you ever used Snapchat or Instagram face filters? Have you played one of the most popular games of 2016, the Pokemon Go? Simply put, that’s AR. iPhone X showcased the power of AR with the now popular animojis, where emojis respond to facial expressions via the camera, to animate various 3D animated characters that can be sent as a video file with sound. In pic: PokemonGo How To Experience Augmented Reality AR is the most accessible reality technology as of now. You can use your smartphone or tablet to run AR apps. The apps use a smartphone camera to capture the real world environment, virtual objects are then overlain with new information or graphics, and users can then see them on their smartphones. You can also use smart glasses or headsets to experience AR. Unlike VR headsets, AR glasses and headsets don’t immerse users into a fully virtual world, instead, they just add virtual images, animation to the real world environment. For instance, Google Glass resembles a pair of eyeglasses and displays information in front of the user’s eyes. In pic: Google Glass Apple took the standard of AR a notch higher by incorporating complex computer vision algorithms. The iPhone maker introduced ARkit which uses ML and AI to detect the ground and calculates the lighting around the virtual object. It supports vertical planes, imagery in 1080p HD and also has a 2D image recognition. In pic: Apple’s ARkit Even Google Lens and Samsung Bixby use computer vision and machine learning to provide digital information from your smartphone camera. In Pic: Google Lens AR can be displayed on various devices — mobile phones, glasses or head-mounted devices. It involves technologies like simultaneous localisation and mapping, depth tracking, camera and sensors, processing, projection and reflection, among others. What Is Mixed Reality? Mixed Reality is a combination of sensors, super advanced optics and next-generation computer power. Remember the scene from the blockbuster movie Avengers: Infinity War, where James Rhodes (Don Cheadle) is having a conversation with US officials who were not physically present there? That’s MR for you. How To Experience Mixed Reality You can experience mixed reality with holographic devices which scan the user’s environment and places a 3D object in front of them, which they can then view from different angles. That’s how Microsoft’s HoloLens works. It has built-in Windows 10 computer, lenses, rechargeable battery and WiFi connectivity, so it is completely wireless when in use. It also includes built-in sensors that track the movement of the users. In pic: HoloLens MR can also be experienced with the help of immersive devices which have an opaque display that creates a sense of presence — hiding the real world and replacing it with a digital experience. According to Microsoft, mixed reality enables content creators to create 360-degree holograms from real life subjects that can be used in applications across augmented reality, virtual reality, and 2D screens. In pic: Acer Windows Mixed Reality headset To Sum Up With tech companies staking their fortunes on virtual reality market, there is more than enough room for startups to make an impact. A number of Indian startups too are meticulously working towards changing the way humans interact with computers. For instance, GreyKernel is one of India’s first VR and advanced visualisation tech startup. Their flagship product IRA VR is one of the most successful virtual reality product from India on Google Play store. SmartVizX Pvt. Ltd is one of the highest funded startups in VR space and offers a wide range of immersive and interactive VR solutions across devices such as mobile, laptop, desktop, tablet etc.","excerpt":"A post shared by Molly Gambardella (@mollygeeart) on May 4, 2018 at 12:21pm PDT Reality technologies which have been in the developmental phase for years, have now become mainstream products for marketers. Tech giants such as Microsoft, Apple and Facebook, among others, are betting big on mixed reality (MR), augmented reality (AR) and virtual reality […]","categories":["IT Services"],"tags":["AI (Artificial Intelligence)","ar","Machine Learning","VR"],"author_name":"Smita Sinha","publish_date":"2018-05-29T11:37:09","publication_year":"2018","word_count":1010,"keywords":["Go","API","machine learning","AI","ML","image recognition","Machine Learning","Git","ar","computer vision","VR","Aim","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","machine learning","ML","computer vision","Aim","image recognition","R","Go","Git","API"],"url":"https:\/\/analyticsindiamag.com\/it-services\/what-is-the-difference-between-virtual-reality-augmented-reality-and-mixed-reality\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":21892,"title":"Inside Facebook’s AI-Powered Speakers Aloha &#038; Fiona, And Why FB Ventured Into Hardware Business","content":"Facebook is planning to venture into consumer hardware business with its own smart speakers, which they have internally code-named as Aloha and Fiona. The social media giant will launch the new speakers in mid-2018. The work on the the 15-inch Aloha began after Facebook saw the success of Amazon’s Echo and with the launch of Aloha and Fiona, Facebook will heat up the smart speaker market globally, which is now dominated by tech giants Apple, Google, Microsoft and several other Chinese players. The work on the the 15-inch Aloha began after Facebook saw the success of Amazon’s Echo. The smart speakers for home will be developed by Facebook’s research lab, Building 8. Aloha will be the first consumer gadget from Building 8 and will be more advanced than Fiona. The former will have a facial recognition feature to identify users for accessing Facebook with the help of a wide-angle lens on the front of the devices and will use voice commands. Both the speakers will be fitted with large touchscreen panels which will be sourced from LG Display, which can be used for video calls, browsing Facebook and watching videos. Taiwanese electronics manufacturing company Pegatron will be the sole assembler of the smart speakers. The company has also tied up Universal Music Group and Sony to add more features to the smart speakers. Last year, Facebook appointed Mark Zuckerberg’s close confidant, Andrew Bosworth as the head of its VR and Building 8 divisions. According to a media report, Bosworth served as Facebook’s vice president of ads and business platform for years and was the man behind Facebook News Feed and Messenger. The company has also hired former Xiaomi executive Hugo Barra and has poached key Google executive Regina Dugan to form Building 8 last April. Aloha Features: The video chat smart device is still under development and the social media giant hopes that Aloha will have more mass market appeal than Oculus and will become a direct competitor of Amazon Echo. Though Facebook has not officially released any statement about Aloha, here are a few features that we could gathered from several media reports: The device will feature a wide-angle camera lens, microphones and speaker, powered by artificial intelligence. It will feature a large touchscreen display of about 13 to 15 inch diagonally. Facebook plans to sell the video-chat device for $499 and will be sold both online as well as in physical stores. The social media giant plans to release the device in the North American and European markets first. The main feature of the device will be its camera, which is described as an ‘AI cameraman’. The device will be used at home and has been tested in living rooms, kitchens and bedrooms. The smart device will require a Facebook or Messenger account to get started. Aloha was originally slated to release in May 2018, but has postponed the launch to give more time to perfect the acoustic quality of the device and modify the software. Facebook will run Aloha on a version of Google’s Android on its device, instead of building its own core operating system. Facebook is testing feature that would allow the camera to automatically scan people in its range and lock onto them. It is also developing a 360 degree camera for the video chat device. However, the smart device is raising privacy concerns as well. Facebook conducted marketing studies for Aloha and found out that users believed that the device would spy on users. Facebook’s Hardware Business: Building Aloha and Fiona is not the only hardware project of Building 8. Facebook’s research lab is also working on a smart speaker like Amazon Echo, a sensor-laden necklace, and is also exploring wearable devices like smart glasses. A media report also claims that Facebook is working on a standalone speaker priced within the $100 range. At Facebook’s annual developer conference, last year, Regina Dugan had said, “Building 8’s goal was to create and ship category-defining consumer products, that are social first at a mass scale… I am optimistic that technology can help and that new hardware platforms can chip away at false choices.” “In the next five years time, Facebook is planning to build its own ecosystem for video consumer devices, and the smart speaker is just the initial stage product, with more terminal devices expected to be rolled out in the coming years,” said a report citing industry sources. It is not the first time Facebook is planning to tap into the consumer gadget segments. The company had earlier came out with a virtual reality headset that worked without an expensive computer. But despite spending a gaudy sum, Facebook’s Oculus VR does not appear to be paying off. In 2017, the social media giant even lowered the prices for its VR products, just to kindle customer interest. The Oculus Gear VR headsets were reportedly handed out with Samsung smartphones, and Oculus Rift failed to attract any real following.","excerpt":"Facebook is planning to venture into consumer hardware business with its own smart speakers, which they have internally code-named as Aloha and Fiona. The social media giant will launch the new speakers in mid-2018. The work on the the 15-inch Aloha began after Facebook saw the success of Amazon’s Echo and with the launch of […]","categories":["IT Services"],"tags":["AI (Artificial Intelligence)","Amazon Echo","face recognition online","Facebook","Google Assistant","Internet of things","Mark Zuckerberg","oculus","real time face recognition software"],"author_name":"Smita Sinha","publish_date":"2018-02-20T05:26:18","publication_year":"2018","word_count":824,"keywords":["Go","artificial intelligence","oculus","programming_languages:R","AI","Amazon Echo","programming_languages:Go","Internet of things","Mark Zuckerberg","real time face recognition software","Aim","GAN","face recognition online","Facebook","Google Assistant","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","Aim","R","Go","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/facebook-ai-speaker-aloha-fiona\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":31802,"title":"How ML Frameworks Like TensorFlow And PyTorch Handle Gradient Descent","content":"Optimisation is the most important component when we consider machine learning algorithms. It helps in reducing the error and improving the accuracy in the solution of a problem. Gradient Descent is one such algorithm which is used for the purpose of optimisation. Here we take a deeper look at what Gradient Descent is and how it helps in optimisation. Understanding Gradient Descent Gradient Descent is the most common optimisation strategy used in ML frameworks. It is basically an iterative algorithm used to minimise a function to its local or global minima. In simple words, Gradient Descent iterates overs a function, adjusting it’s parameters until it finds the minimum. A gradient can be called the partial derivative of a function with respect to its inputs. Basically, it is a measure of the variation in weights with respect to change in error or change in input. Let us visualise with the simplest example. Consider the following image of a curve: For better understanding visualise two-dimensional section the curve. We will get something like this: Now imagine a ball being rolled from the top most end of the curve. The objective is to reach the lowest point. The ball will roll down and then up, repeatedly until it rests at the steepest point. This is how Gradient Descent works. The algorithm repeats and adjusts its parameters or coefficients to find the steepest point. In the ML context, the Gradient Descent is used to minimise the error by adjusting weights after passing through all the samples in the training set. If the weights are updated after a specified subset of training samples, or after each sample in the training set, then it is called a Stochastic Gradient Descent. The higher the gradient, the steeper the slope and the faster a model can learn. But if the slope is zero, the model stops learning. With this basic understanding, let us now take a look at how the popular ML packages like TensorFlow and PyTorch solve Gradient Descent. Gradient Descent With TensorFlow TensorFlow has a Class called  GradientDescentOptimizer to handle Gradient Descent. Consider the simplest example that illustrates the usage of GradientDescentOptimizer class. The highlighted part is where the GradientDescentOptimizer is invoked. GradientDescentOptimizer is called with a step of 0.01 which is the standard value.The minimise function minimises the value of the variable error which is defined as the square difference of the actual and predicted set. The minimise function is a combination of two functions Compute_gradients() : This method returns a list of (gradient, variable) pairs where “gradient” is the gradient for “variable”. apply_gradients() : This is the second part of minimize(). It returns an Operation that applies gradients. Gradient Descent with PyTorch PyTorch uses the Class torch.optim.SGD to  Implement stochastic Gradient Descent. Consider the following illustration. The lr parameter stands for learning rate or step of the Gradient Descent and model.parameters returns the parameters learned from the data. The gradient buffer is set to zero by the function optimizer.zero_grad() once for every training iteration to reset the gradient computed by the last data batch —","excerpt":"Optimisation is the most important component when we consider machine learning algorithms. It helps in reducing the error and improving the accuracy in the solution of a problem. Gradient Descent is one such algorithm which is used for the purpose of optimisation. Here we take a deeper look at what Gradient Descent is and how […]","categories":["Global Tech"],"tags":["gradient descent","Machine Learning","programming","Pytorch","Tensorflow"],"author_name":"Amal Nair","publish_date":"2018-12-19T09:50:05","publication_year":"2018","word_count":511,"keywords":["Pytorch","Go","machine learning","programming_languages:R","AI","PyTorch","programming","ML","Machine Learning","gradient descent","ai_frameworks:TensorFlow","ai_frameworks:PyTorch","TensorFlow","R","Tensorflow"],"extracted_tech_keywords":["AI","machine learning","ML","TensorFlow","PyTorch","R","Go","ai_frameworks:TensorFlow","ai_frameworks:PyTorch","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/how-ml-frameworks-like-tensorflow-and-pytorch-handle-gradient-descent\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10054105,"title":"Cloud Announcements That Made Headlines In 2021","content":"The adoption of cloud-based technologies in businesses has been accelerated manifold by the pandemic and rapid digital transformation. A Gartner report says that in 2022, global cloud revenue is estimated to total $474 billion, up from $408 billion in 2021. By 2025, Gartner estimates that over 95% of new digital workloads will be deployed on cloud-native platforms, up from 30% in 2021. Globally and in India, tech companies announced various improvements in Cloud infrastructure, partnerships, cloud centres, cloud training initiatives this year. Let’s look into some major happenings in Cloud this year: Google Google Cloud Next The Google Cloud Next 2021 event was held in October, where the tech giant made a range of announcements. Some of the important ones are: Google Distributed Cloud-It is a bundle of fully managed hardware and software solutions that work on edge and data centres. Google Cloud Cortex Framework and Cloud Build Hybrid: The Google Cloud Cortex Framework allows customers to get insights and reduce time-to-value with reference architectures, packaged services, and deployment accelerators. The Cloud Build Hybrid allows developers to build, test, and deploy across clouds and on-premise systems. It also provides the benefits of a Google-managed control plane to manage all pipelines. Hiring And Partnerships Google Cloud recently announced the appointment of Subram Natarajan as the Director of Customer Engineering at Google Cloud India. The company also recently launched Google Cloud Skills Boost, a platform for online learning, skills development, and certifications. MongoDB recently announced an expanded five-year partnership with Google Cloud to better integrate Google Cloud products with its cloud database, MongoDB Atlas. Pegasystems Inc. teamed up with Google to provide better data insights in healthcare. VMware At VMworld 2021 event held in October, the company announced the launch of VMware Cross-Cloud services. It said that the services will help businesses to move to the Cloud faster and provide customers with the ability to build, deploy and secure apps across any cloud. VMware introduced Cloud Infrastructure and Management for the Cross-Cloud services, advancements in the VMware Tanzu portfolio, and launched VMware Edge, among others announcements. A few days back, VMware also announced a partnership with Kyndryl, which will focus on app modernisation and multi-cloud services. Oracle In October, Oracle announced plans to expand its cloud region reach to meet global customer demand for Oracle Cloud services. Reportedly, over the next year, the company will open 14 cloud regions across Europe, the Middle East, Asia Pacific, and Latin America and aims to reach at least 44 cloud regions by the end of 2022. Amazon It announced the launch of AWS re\/Start in India. It is a full-time, classroom-based skills development and training program that prepares individuals for careers in the Cloud and helps to build connections with potential employers. Amazon Web Services teamed up with Salesforce for customers to deploy Salesforce and AWS capabilities together to use business applications for digitally transforming their business. Microsoft Microsoft Ignite 2021 Microsoft chief Satya Nadella talked about how Microsoft Cloud addresses present-day challenges. New products and upgrades announced during this event included general availability of Microsoft Test Base for Microsoft 365, Microsoft Edge web browser for Linux, Microsoft Cloud for Financial Services and Cloud for Nonprofits etc. TCS At the beginning of the year, TCS introduced a curated version of its Cloud Assurance Platform services that will help organisations start with cloud migration or modernisation with Microsoft Azure. Data Centers Google Cloud launched the second data centre cluster for India in the Delhi-National Capital Region (NCR). Cybersecurity major Trend Micro launched its Cloud One regional data centre in India for data sovereignty and data privacy. Cybersecurity company Acronis launched Cloud Data Center in Mumbai jointly with local partners Compuage India, Ingram Micro and Crayon Software Experts India. Tech Mahindra launched a dedicated Google Cloud Business Unit for cloud adoption for enterprises globally. HCL launched its Amazon Web Services (AWS) Business Unit (AWS BU) to help enterprises accelerate their cloud transformation globally.","excerpt":"By 2025, Gartner estimates that over 95% of new digital workloads will be deployed on cloud-native platforms, up from 30% in 2021.","categories":["IT Services"],"tags":["data centres","digital transformation","Google","Microsoft","Oracle","Oracle Certification","partnership","VMWare"],"author_name":"Sreejani Bhattacharyya","publish_date":"2021-11-24T12:00:00","publication_year":"2021","word_count":656,"keywords":["partnership","VMWare","Go","API","AWS","AI","MongoDB","digital transformation","R","Git","Oracle","Ray","Aim","data centres","Google","Oracle Certification","Azure","Microsoft"],"extracted_tech_keywords":["AI","Aim","Ray","AWS","Azure","MongoDB","R","Go","Git","API"],"url":"https:\/\/analyticsindiamag.com\/it-services\/cloud-announcements-that-made-headlines-in-2021\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":41632,"title":"MIT Researchers Teach AI To Paint  With Common Sense","content":"Changing the season of trees with GAN Paint Studio Today’s smartphones often use artificial intelligence (AI) to help make the photos we take crisper and clearer. But what if these AI tools could be used to create entire scenes from scratch. A team from MIT and IBM has now done exactly that with “GAN Paint Studio,” a system that can automatically generate realistic photographic images and edit objects inside them. David Bau, a PhD student at MIT’s Computer Science and Artificial Intelligence Lab (CSAIL), describes the project as one of the first times computer scientists have been able to actually “paint with the neurons” of a neural network – and specifically, a popular type of network called a generative adversarial network (GAN). Available online as an interactive demo, GAN Paint Studio allows a user to upload an image of their choosing and modify multiple aspects of its appearance, from changing the size of objects to adding completely new items like trees and buildings. Spearheaded by MIT professor Antonio Torralba as part of the MIT-IBM Watson AI Lab he directs, the project has vast potential applications. Designers and artists could use it to make quicker tweaks to their visuals. Adapting the system to video clips would enable computer-graphics editors to quickly compose specific arrangements of objects needed for a particular shot. (Imagine, for example, if a director filmed a full scene with actors but forgot to include an object in the background that’s important to the plot.) GAN Paint Studio could also be used to improve and debug other GANs that are being developed, by analyzing them for “artifact” units that need to be removed. In a world where opaque AI tools have made image manipulation easier than ever, it could help researchers better understand neural networks and their underlying structures. “Right now machine learning systems are these black boxes that we don’t always know how to improve, kind of like those old TV sets that you have to fix by hitting them on the side,” says Bau, lead author on a related paper about the system with a team overseen by Torralba. “This research suggests that, while it might be scary to open up the TV and take a look at all the wires, there’s going to be a lot of meaningful information in there.” One unexpected discovery is that the system actually seems to have learned some simple rules about the relationships between objects. It somehow knows not to put something somewhere it doesn’t belong, like a window in the sky, and it also creates different visuals in different contexts. For example, if you ask the system to add doors to two different buildings, it doesn’t simply add identical doors – they may ultimately look quite different from each other. “All drawing apps will follow user instructions, but ours might decide not to draw anything if the user commands to put an object in an impossible location,” says Torralba. “It’s a drawing tool with a strong personality, and it opens a window that allows us to understand how GANs learn to represent the visual world.” By adding\/removing objects in images, system could help artists and designers make quicker tweaks to visuals GANs are sets of neural networks developed to compete against each other. In this case, one network is a generator focused on creating realistic images, and the second is a discriminator whose goal is to not be fooled by the generator. Every time the discriminator ‘catches’ the generator, it has to expose the internal reasoning for the decision, which allows the generator to continuously get better. “It’s truly mind-blowing to see how this work enables us to directly see that GANs actually learn something that’s beginning to look a bit like common sense,”  says Jaakko Lehtinen, an associate professor at Finland’s Aalto University who was not involved in the project. “I see this ability as a crucial stepping stone to having autonomous systems that can actually function in the human world, which is infinite, complex and ever-changing.” The team’s goal has been to give people more control over GAN networks.  But they recognize that with increased power comes the potential for abuse, like using such technologies to doctor photos. Co-author Jun-Yan Zhu says that he believes that better understanding GANs – and the kinds of mistakes they make – will help researchers be able to better stamp out fakery. “You need to know your opponent before you can defend against it,” says Zhu, a postdoc at CSAIL. “This understanding may potentially help us detect fake images more easily. “ To develop the system, the team first identified units inside the GAN that correlate with particular types of objects, like trees. It then tested these units individually to see if getting rid of them would cause certain objects to disappear or appear. Importantly, they also identified the units that cause visual errors (artifacts) and worked to remove them to increase the overall quality of the image. “Whenever GANs generate terribly unrealistic images, the cause of these mistakes has previously been a mystery,” says co-author Hendrik Strobelt, a research scientist at IBM. “We found that these mistakes are triggered by specific sets of neurons that we can silence to improve the quality of the image.” Bau, Strobelt, Torralba and Zhu co-wrote the paper with former CSAIL Ph.D. student Bolei Zhou, postdoctoral associate Jonas Wulff and undergraduate student William Peebles. They will present it next month at the SIGGRAPH conference in Los Angeles. “This system opens a door into a better understanding of GAN models, and that’s going to help us do whatever kind of research we need to do with GANs,” says Lehtinen. To read the paper, click here Watch the demo here","excerpt":"Today’s smartphones often use artificial intelligence (AI) to help make the photos we take crisper and clearer. But what if these AI tools could be used to create entire scenes from scratch. A team from MIT and IBM has now done exactly that with “GAN Paint Studio,” a system that can automatically generate realistic photographic […]","categories":["AI Features"],"tags":["AI in art"],"author_name":"AIM Media House","publish_date":"2019-07-02T10:26:27","publication_year":"2019","word_count":953,"keywords":["Go","machine learning","artificial intelligence","ELT","AI","neural network","AI in art","programming_languages:R","CLIP","GAN","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","neural network","R","Go","ELT","CLIP","GAN","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/mit-researchers-teach-ai-to-paint-with-common-sense\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10064942,"title":"Neo4j introduces Graph Data Science as a service","content":"Graph data platform Neo4j has announced Neo4j Graph Data Science, the company’s comprehensive graph analytics workspace built for data scientists. It is now available with new and enhanced capabilities, and as a fully managed cloud service called AuraDS. AI and machine learning (ML) have propelled the use of predictive data architectures and their application across a broad range of use cases like recommendation engines, fraud detection, and customer 360 scenarios. The accuracy of these models is highly correlated to the completeness of context. Neo4j Graph Data Science is designed to make it easy for data scientists to achieve greater predictive accuracy with comprehensive graph analysis techniques. Users can improve models through a library of graph algorithms, ML pipelines, and data science methods. Neo4j Graph Data Science has been widely adopted, and is trusted to perform at scale, easily handling hundreds of billions of nodes and relationships. “Neo4j’s Graph Data Science offerings help developers offer better predictions and stronger recommendation engines to business users,” said Ritika Suri, Director, Technology Partnerships at Google. “Customers can now deploy Graph Data Science on Google Cloud’s trusted, global infrastructure, gaining the ability to seamlessly scale based on business needs, and bringing their data closer to BigQuery and Google Cloud’s capability in AI, ML, and analytics.”Neo4j Graph Data Science makes it easy for data scientists to work within their existing data pipeline of tools across their ecosystem. Data scientists can use Neo4j Graph Data Science on-premises, and now as a fully managed SaaS solution via Neo4j AuraDS.","excerpt":"Data scientists can use Neo4j Graph Data Science on-premises, and now as a fully managed SaaS solution via Neo4j AuraDS.","categories":["AI News"],"tags":["Data Science","data science tools","Machine Learning","neo4j"],"author_name":"Kartik Wali","publish_date":"2022-04-13T17:07:45","publication_year":"2022","word_count":251,"keywords":["data science","Go","data science tools","machine learning","AI","ML","data pipeline","Machine Learning","neo4j","analytics","Rust","Data Science","R","fraud detection"],"extracted_tech_keywords":["AI","machine learning","ML","data science","analytics","fraud detection","R","Go","Rust","data pipeline"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/neo4j-introduces-graph-data-science-as-a-service\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":29100,"title":"Can Deep Learning Answer India’s Critical Infrastructure Woes?","content":"In order to make large scale structuring safe, structural health monitoring (SHM) techniques have been adopted. However, SHM is fraught with several challenges too such as the inability to deal with uncertainties in large-scale civil structures. As a consequence, image processing techniques were developed. In this article, we will discuss the use of convolutional neural network (CNN) and see how it can be impactful for India’s massive infrastructure. Damage Detection Through CNN In their research paper titled Deep Learning-Based Crack Damage Detection Using Convolutional Neural Networks, Young-Jin Cha and team came up with a CNN-based framework to detect concrete cracks in structures that actually does not look into defect features in depth. This eliminates the need for image processing techniques as CNN learn from image features rather than without relying on images. Therefore, a dataset of raw images captured with a DSLR camera and showing image specific variations is used. These images are tweaked into small images, to create training and validation sets. Once ready, they are fed into the proposed CNN for training. After training, CNN is tested for new images. The entire process is shown below. Flowchart for the process (Image Courtesy: Young-Jin Cha, Wooram Choi & Oral Oral Buyukozturk) The CNN architecture has 8 inner layers alternating between convolution and pooling layers which also includes a ReLU activation layer and a softmax layer. All the layers perform a host of mathematical operations in order to build a classifier to detect cracks and related surface anomalies. This forms the training part. Now this classifier is subjected to validation after training. Just like classic ML, this is to increase accuracy if the CNN encounters new images apart from the ones in the dataset. With results showing around 98 percent for both training and validation for around 300 images, this CNN was also consistent when it came to image variations. But, for this CNN architecture to be successful in other applications, the researchers suggest collecting a large amount of training data. Where This CNN Finds Application In Indian Scenario Infrastructure problems in India have been looming over the past few decades. A variety of factors can be said to be the reason. Be it environmental effects or even the rising population, infrastructure in India is bound with challenges. When we consider all of these issues surrounding structural elements, optimising resources is the right way forward. For example, assessment in giant structures such as bridges or dams are still observed through manual methods in India despite the presence of vision-based techniques. The fear of expenses rising up is another factor here. If this mindset does not change, it is hard for emerging technologies to find its place in civil engineering. Deep learning techniques such as Cha et. al.’s CNN can go a long way in structural analysis. They do away with time-consuming manual inspection for fault detection in structures. Also, even when a structure is being built, it is susceptible to damages. Risk analysis, usually done by site engineers, can be done with DL methods. Again, one of the most important things here is right data. For example, Cha’s study used images from a single data source i.e, one dataset. However, in real-time, DL requires uniform data constantly to work smoothly. For this, data sources need to be integrated into a single platform so that the underlying technique can use data efficiently.","excerpt":"In order to make large scale structuring safe, structural health monitoring (SHM) techniques have been adopted. However, SHM is fraught with several challenges too such as the inability to deal with uncertainties in large-scale civil structures. As a consequence, image processing techniques were developed. In this article, we will discuss the use of convolutional neural […]","categories":["AI Features"],"tags":["images","infrastructure","Machine Learning","Neural Network"],"author_name":"Abhishek Sharma","publish_date":"2018-10-10T11:21:32","publication_year":"2018","word_count":561,"keywords":["Neural Network","Go","infrastructure","programming_languages:R","AI","neural network","ML","Machine Learning","programming_languages:Go","deep learning","CNN","R","images"],"extracted_tech_keywords":["AI","ML","deep learning","neural network","R","Go","CNN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/can-deep-learning-answer-indias-critical-infrastructure-woes\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10070463,"title":"Everyone is Open Sourcing their Language Models, so does this Russian Search Engine","content":"Recently, Russian company Yandex open sourced YaLM 100B, a bilingual neural network for generating and processing text. “By making YaLM 100B publicly available, we hope to give impetus to further developing generative neural networks,” said Petr Popov, CEO of Yandex Technologies. The development comes at a time when several big companies like Meta, Google, and OpenAI have open-sourced some of their large transformer-based models. In early 2021, researchers at Google Brain open-sourced the Switch Transformer, natural-language processing (NLP) AI model. EleutherAI open-sourced its large language model (LLM) GPT-NeoX-20B in April 2022, followed by Meta AI open-sourcing the first version of OPT-175B. What is YaLM 100B? YaLM 100B is a GPT-like neural network for generating and processing text. It is the largest language model from the YaLM family. YaLM language models help determine the principles of constructing texts and generate new ones based on the rules of linguistics and their knowledge of the world. YaLM can not only create texts but also classify them according to the styles of speech. Yandex has been using YaLM neural networks in its voice assistant, Alice and its search engine Yandex Search. YaLM 100B has been released under the Apache 2.0 license, which permits research and commercial use. Training the model Training large-scale language models is resource-intensive. “Training generative neural networks requires substantial resources, experienced professionals and years of work. And it is important for us that not only the largest IT companies have access to modern technologies, but the entire community of researchers and developers,” said Popov. Developers at Yandex trained YaLM 100B on a cluster of 800 A100 graphics cards for 65 days. During the training, the neural network consumed 300B tokens and processed 1.7TB of texts in English and Russian. The datasets used for training YaLM 100B roughly include 25% of text from the Pile dataset (open English dataset by EleutherAI team) and 75% of text in the Russian language from various sources like Wikipedia, preprocessed dialogues from social media, Taiga Dataset, Russian Distributional Thesaurus dataset and Yandex Search index. Developers used DeepSpeed, a deep learning optimization library, to train the model. DeepSpeed makes distributed training and inference easy, efficient, and effective. The researchers explained how they trained the model and suggested ways to accelerate model training. According to them, a 10% increase in training speed can reduce runtime on a high-value cluster by a week. Training iterations usually include the following steps: Preparing the batchCalculating the activation and loss functions by running forward propagationCalculating gradients by running backward propagationRunning the step stage to modify the model’s weights Accelerating model training To accelerate model training, developers suggest the following : Looking for bottlenecks: The team recommends using a profiler to identify performance bottlenecks in the models. Using a profiler helps you understand how the training time is spent. For example, researchers could analyze why one operation took almost 50% of the entire training time. Thus, they could reduce the token embedding size to avoid excessive matrix multiplication at the end of the network. This helped in speeding up the training process.Using fast data types: The types of data used to store the model and perform necessary calculations determine the speed of training and inference. Therefore, developers recommend using fast data types. For example, on A100 and newer graphics cards, 16-bit data types like fp16 and bfloat16 are five times faster than fp32(Single-precision format) and 2.5 times faster than 19-bit data type tf32(TensorFloat format). However, older graphics cards do not support bf16 and tf32 data types, and fp16 is only two times faster than fp32.Accelerating GPU operations: You can fully utilize GPUs by increasing the batch size. Increasing the batch size helps in accelerating the training speed.  To minimize memory interaction, developers suggest fusing the kernels using torch.jit.script, writing your own CUDA kernels, or using ready-made CUDA kernels available in Megatron-LM and DeepSpeed libraries. For example, using the torch.jit.script developers fused three operations- tensor add, dropout and another tensor add that helped them increase the learning rate by 5%. For accelerated training of YaLM, developers used different kinds of fused kernels that sped up training by almost 1.5 times. If you have a lot of data and no retraining at dropout == 0, disable dropouts! This increased their computing speed by 15%. NVIDIA NCCL library helped ensure maximum communication speed by allowing GPUs to effectively communicate over the network without any CPU intermediaries. Further, using Zero Redundancy Optimizer (ZeRO) accelerated communication even more. Though ZeRO helped save huge amounts of memory, it brought in complexity by adding new heavy operations. To overcome this, developers gathered the different layers asynchronously one after the other. This technique helped developers gain 80% speed in training their models. Divergence and stabilization strategies The model was prone to divergence. When divergence occurs, a machine learning model gradually forgets what it has learnt. To deal with this, developers deployed the following stabilization strategies. Adopted Bf16 as the main type for weights.Ran precision-critical computations in tf32Introduced Pre-LayerNorm, and after embeddings, they added LayerNorm.Used Curriculum Learning, a training strategy that trains a machine learning model from easier data to harder data. It helps in improving the generalization capacity and convergence rate of various models.","excerpt":"YaLM 100B is a GPT-like neural network for generating and processing text","categories":["AI Features"],"tags":[],"author_name":"Zinnia Banerjee","publish_date":"2022-07-05T15:00:00","publication_year":"2022","word_count":863,"keywords":["CUDA","Go","Meta AI","machine learning","OpenAI","AI","neural network","NLP","deep learning","R"],"extracted_tech_keywords":["AI","machine learning","deep learning","neural network","NLP","OpenAI","Meta AI","CUDA","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/everyone-is-open-sourcing-their-language-models-so-does-this-russian-search-engine\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":61834,"title":"Recent Analytics Job Openings You Should Apply Right Away","content":"As the world enters a wave of a new tech revolution, analytics and the technologies derived from it will see an increase in adoption. This will require more expertise with data. With India contributing its fair share of analytics job openings and the demand only going up, there is still a shortage of talent in this area. Keeping this mind, we have listed a few job postings to help you in your search: Analyst, Merchant Product Analytics From: PayPal Location: Chennai, Tamil Nadu Requirements and Responsibilities: Ability to perform an in-depth analysis of key business trends from different perspectives, and package insights into easily consumable presentations and documentsSolid SQL skills, with attention to data and visualizing them using TableauAbility to interact with stakeholders and front endAdvanced Microsoft Excel skillsWork with product, engineering, instrumentation and analytics teams to grasp product, support experiment setup and analyse performance Get more details and apply here. Business Consultant – Analytics From: Dr Reddy’s Laboratories Location: Hyderabad Requirements and Responsibilities: Use advanced data analytics to address business needs and create business opportunitiesWork with business\/functional users to understand business decision-making Design data science solutions in collaboration with technology architects to address this needDevelop in-house analytics knowledge for faster, better problem-solving and horizontal scale-out in other markets of needRequired experience working with customer analytics, sales force excellence, sales & marketing analyticsExposure to tools like Tableau, MicroStrategy, Spotfire, etc., analytics applications like R, SAS, Python, etc. and data mining \/ML analytics techniques are a big plusFamiliarity with PMI project methodologies and SDLC Get more details and apply here. Analytics Specialist From: Accenture Location: Bangalore Responsibilities and Requirements: Conceptualize, develop, and implement advanced statistical models on business dataProven experience and skills in survey data analytics, HR analytics, predictive modelling or machine learning techniquesExpertise in primary and secondary researchLeverage combination of multiple approaches for project execution, including internal assets, third party and open source solutions Get more details and apply here. Risk Analytics From: Accenture Location: Chennai Responsibilities and Requirements: 2+ years experience in big data tech and toolsWorking experience with Scala and Spark to build ETL pipelinesExpertise in Scala, Spark data frames and HiveGood knowledge when it comes to Hadoop, HDFS, Linux, Elastic searchExperience in banking or banking fraud analytics Get more details and apply here. Data Solutions Specialist From: Chubb Location: Bangalore Responsibilities and Requirements: Give contributions to the architecture of solutions to combine large, complex data sets into a pragmatic, actionable data mart, including programming in SQLHelp improve and streamline processes regarding data flow and data quality from the datamart to improve data accuracy, viability and useSupport development of QlikView and other profitability & impact analysisWork with other members of the Australia SME actuarial team to support regular pricing studies and other ad-hoc workDevelop and maintain clear and brief documentation Get more details and apply here. Process Expert – Data Scientist From: Siemens Technologies Location: Bangalore Responsibilities and Requirements: Possess knowledge in advanced data analytics\/data scienceKnowledge about how to extract, transform and load (ETL) of big data in an enterprise environment (SAP, Oracle), cloud infrastructure (AWS, Azure)Should have good knowledge of SQL, Python, R, and frameworks like TensorFlow and KerasExperience with designing solutions in distributed teams using Version Control tools like GITExperience with cloud technologies like AWS and\/or Microsoft Azure Get more details and apply here. Pharma Analyst From: IQVIA Location: Bangalore Responsibilities and Requirements: Experience in analytics projects in pharma domain, deriving actionable insights and implementing them.Experience working with patient data, IQVIA (ex-IMS Health) data sets, pharma consulting, advanced pharma analytics, performance reporting, forecastingStrong MS Excel and Powerpoint knowledgeExperience in SQL and any of these: SAS, Python, R, Excel-VBA, Tableau, QlikView, AlteryxWorking on advanced analytics in the pharma domain throughout the drug life cycle and patient journeyCompetitor analysisHCP analysis (new therapy initiator)Tigger analysis (conditions triggering usage of a product in a therapeutic area) Get more details and apply here.","excerpt":"As the world enters a wave of a new tech revolution, analytics and the technologies derived from it will see an increase in adoption. This will require more expertise with data. With India contributing its fair share of analytics job openings and the demand only going up, there is still a shortage of talent in […]","categories":["AI Hirings"],"tags":["analytics jobs india","enterprise analytic hub","etl hadoop","hadoop and etl","hadoop etl"],"author_name":"Sameer Balaganur","publish_date":"2020-04-16T13:00:00","publication_year":"2020","word_count":642,"keywords":["data science","enterprise analytic hub","machine learning","Keras","AI","TensorFlow","AWS","ML","Azure","RAG","analytics","analytics jobs india","hadoop etl","hadoop and etl","etl hadoop"],"extracted_tech_keywords":["AI","machine learning","ML","data science","analytics","TensorFlow","Keras","RAG","AWS","Azure"],"url":"https:\/\/analyticsindiamag.com\/ai-hiring\/recent-analytics-job-openings-you-should-apply-right-away\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":24557,"title":"A Step-by-Step Guide To Creating Credit Scoring Model From Scratch","content":"A credit scoring model is a statistical tool widely used by lenders to assess the creditworthiness of their potential and existing customers. The basic idea behind this model is that various demographic attributes and past repayment behavior of an individual can be utilized to predict hers or his probability of default. 1.    Data For demonstration purposes, we will be using the data set that contains characteristics and delinquency information for 5,960 home equity loans (source: B. Baesens, D. Roesch, H. Scheule, Credit Risk Analytics: Measurement Techniques, Applications and Examples in SAS, Wiley, 2016). The data set variables are: The binary variable BAD will be the target variable in our credit scoring model, while other variables will be used as predictors. There are 4,771 observations (80.05%) where bad is 0 and 1,189 observations (19.95%) where bad is 1. 2.    Categorical variables For the categorical variable, the log odds ratio of the bad is calculated for each of the values of the variable. The values for which the log odds ratio is very close or the counts is insignificant are clubbed together. The values are then replaced by the log odds ratio. 2.1.         Reason The reason variable is collapsed into 2 bins. The DebtCon and Missing have log odds ratio very close to each other. 2.2.         Job The job variable is collapsed into 5 bins. Missing is treated as a separate bin. 2.3.         Derog The derog variable is collapsed into 4 bins. The count observations for derog values more than 2 is insignificant, hence these are clubbed together. 2.4.         Delinq The delinq variable is collapsed into 5 bins. The count observations for delinq values more than 4 is insignificant, hence these are clubbed together. 2.5.         Ninq The ninq variable is collapsed into 5 bins. The count observations for ninq values more than 4 is insignificant, hence these are clubbed together. 3.    Numerical variables The numerical variable are rank binned into 12 equal bins (if the values are missing then we have 13 bins with -1 indicating missing bin). Two approaches are followed to handle numerical variables: Treating as categorical: The bins for which the log odds ratio is very close or the counts is insignificant are clubbed together. The bins are then replaced by the log odds ratio. This process is followed for numerical variables where transformations will not linearize the variable. Treating as numerical: The variable is linearized and the transformed variable is used in the model. 3.1.         Loan The loan variable is collapsed into 2 bins. The loan variable <= 7000 is treated as first bin and loan variable > 7000 is treated as second bin. 3.2.         Mortdue The mortdue variable is collapsed into 3 bins. First bin: The value <= 37000 is treated as first bin Second bin: The value > 73000 and value <= 106000 is treated as second bin Third bin: Rest of the values (including the missing values) are clubbed as third bin 3.3.         Value The value variable is collapsed into 4 bins. Missing is treated as a separate bin. 3.4.         Yoj The value variable is collapsed into 3 bins. First bin: The value is Missing or value > 22 is treated as first bin Second bin: The value > 0.4 and value <= 2.9 is treated as second bin Third bin: Rest of the values are clubbed as third bin 3.5.         Clage The clage variable is linearized using a quadratic equation. if CLAGE is MISSING then CLAGE = 0.14 Transformation = (15.49*CLAGE*CLAGE) – (12.06*CLAGE) + 0.0789 3.6.         Clno The clno variable is collapsed into 3 bins. First bin: The value is Missing or value >= 36 is treated as first bin Second bin: The value >= 0 and value <= 8 is treated as second bin Third bin: Rest of the values are clubbed as third bin 3.7.         Debtinc The debtinc variable is collapsed into 3 bins. Missing is treated as a separate bin. 4.    Model Percent Concordant: A pair of observations with different observed responses is said to be concordant if the observation with the lower ordered response value (bad = 0) has a lower predicted mean score than the observation with the higher ordered response value (bad = 1) Percent Discordant: If the observation with the lower ordered response value has a higher predicted mean score than the observation with the higher ordered response value, then the pair is discordant. Percent Tied: If a pair of observations with different responses is neither concordant nor discordant, it is a tie. Pairs: This is the total number of distinct pairs in which one case has an observed outcome different from the other member of the pair. Somers’ D: It is used to determine the strength and direction of relation between pairs of variables. Its values range from -1.0 (all pairs disagree) to 1.0 (all pairs agree). C: c is equivalent to the well-known measure ROC, c ranges from 0.5 to 1, where 0.5 corresponds to the model randomly predicting the response, and a 1 corresponds to the model perfectly discriminating the response. Parameter: Underneath are the predictor variables in the model and the intercept. DF: This column gives the degrees of freedom corresponding to the Parameter. Each Parameter estimated in the model requires one DF and defines the Chi-Square distribution to test whether the individual regression coefficient is zero, given the other variables are in the model. Estimate: These are the binary logit regression estimates for the Parameters in the model. The logistic regression model models the log odds of a positive response (probability modeled is bad=1) as a linear combination the predictor variables. It should be noted here that the estimates are very close to 1 (other than the intercept). It is because we have used the binned \/ transformed variables in the final model. Standard Error: These are the standard errors of the individual regression coefficients. They are used in both the 95% Wald Confidence Limits, superscript w, and the Chi-Square test statistic, superscript t. Chi-Square and Pr > ChiSq: These are the test statistics and p-values, respectively, testing the null hypothesis that an individual predictor’s regression coefficient is zero, given the other predictor variables are in the model. It should be noted here that the p-value for all the variables is close to 0. Number of Observations Used = 5,960 Number of Variables = 7 Confusion matrix: Accuracy: Accuracy is the fraction of predictions the model got right. Formally, accuracy has the following definition: Number of correct predictions \/ Total number of predictions 88.52% ROC: ROC curve is a graphical plot that illustrates the diagnostic ability of a binary classifier system as its discrimination threshold is varied. The ROC curve is created by plotting the true positive rate (TPR) against the false positive rate (FPR) at various threshold settings. The true-positive rate is also known as sensitivity, recall or probability of detection in machine learning. The false-positive rate is also known as the fall-out or probability of false alarm and can be calculated as (1 − specificity).","excerpt":"The basic idea behind this model is that various demographic attributes and past repayment behavior of an individual can be utilized to predict hers or his probability of default.","categories":["Deep Tech"],"tags":["confusion matrix recall","what is recall in confusion matrix"],"author_name":"Rohit Garg","publish_date":"2018-05-14T05:28:14","publication_year":"2018","word_count":1165,"keywords":["Go","machine learning","programming_languages:R","AI","confusion matrix recall","ML","programming_languages:Go","Git","ViT","what is recall in confusion matrix","analytics","R"],"extracted_tech_keywords":["AI","machine learning","ML","analytics","R","Go","Git","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/a-step-by-step-to-creating-credit-scoring-model-from-scratch\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":21849,"title":"How Google’s Arts &#038; Culture App Uses AI, Facial Recognition To Find Your Museum Doppelgänger","content":"You must have noticed of late that your friends and followers are posting selfies with their artwork doppelgängers on social media. It all sprang from the popular facial recognition art app called Google Arts and Culture app, which matches your selfie with a renowned pieces of artwork. The app also lets you explore art by different styles, time periods and colors. It also offers a catalogue of features and articles on art history and artists. A year after its inception, the app topped the iOS and Android download charts last month. It is available for free on both Apple app store and Google Play store. The popularity of the app also caught the attention of the celebrities such as Kristen Bell, Pete Wentz, Sarah Silverman and Busy Philipps, among others. I win? A post shared by Busy Philipps (@busyphilipps) on Jan 12, 2018 at 8:38pm PST This google arts and culture app is pretty amazing. Feel real strong about my 40% 😳 pic.twitter.com\/2iyexRkUG5 — pw (@petewentz) January 14, 2018 How Does The App Work? Once you take a selfie using the app, the AI in the Art app detects a face in a picture and creates a faceprint. Each selfie that is uploaded is compared with 70,000 artworks available in its database. The main interface of the Google Arts and Culture selfie comparison tool is the selfie camera. After you download the app from the app store, open the app and scroll down until you see the white box surrounded by portraits that says ‘search with your selfie’. Next – you are ready to take a selfie. Line up your face inside the boxed area and snap a picture. After you take a selfie, the AI then analyses your face within seconds. Once matched, the app will show the artist who made the painting along with the museum it comes from. It also shows you how much you resemble with the vintage portrait with a percentage come up with each comparison. “When you take a photo with this feature, your photo is sent to google to find artworks that look like you,” Google explained in a report. “Google won’t use data from your photo for any other purpose and will only store your photo for the time it takes to search for matches.” But sometimes the result might disappoint you. The app isn’t dedicated solely to comparing selfies with paintings, it also lets you take virtual tours of historic museum such as Art Gallery of New South Wales in Sydney, Australia. You can also browse artworks by color and time period and zoom into famous artworks like Grupo ruptura and Arte Concreta – The Adolpho Leirner Collection of Brazilian Constructive art at the Museum of Fine Arts, Houston. Technology Behind The Art App: Google uses pattern recognition to recognise your face and provide similar results from its database procured from museums. “For instances, a computer might be trained to recognise the common patterns of shapes and colors that make up a digital face and the technology also helps you protect the privacy.” said Google’s official website, describing pattern recognition. The same pattern recognition technology that powers facial detection can help a computer to understand characteristics of the face it has detected. “For example, there might be certain patterns that suggest a face is wearing a beard or glasses, or that it has attributes like those.” Man Behind The Art Project: According to a report, Mumbai-based Android marketer Amit Sood started this project as a side venture, who devoted 20 percent of his time at the company to explore how to make artwork accessible online. Sood initially approached 17 museums about collaborating on the project and later he combined it with similar efforts inside Google. His group developed a robotic art camera that allows museum to make detailed images of their works and enlisted more than 1,500 international museums from 70 countries. It introduced the smartphone app two years ago which gives access to museums’ online collections, VR tours and guides to artwork. “I’m not really from the art, cultural world. I stepped into this a few years ago. I’m very glad I did but I wish I had had access to art and culture when I was growing up. Where I grew up, it’s not really the done thing to go to a museum. It’s an accessibility issue – very gently put. For me, (Google’s Cultural Institute) is the best example of how art and technology really breaks down this whole accessibility problem of art and culture,” the director of Google’s Cultural Institute said. However, the Google does not let you download the image from the Art project. “We don’t allow downloading. That’s at the request of the museums,” Sood said. What Google Plans To Do With Your Selfies: According to a report, the app also faced criticism for its lack of diversity. Some even said that the app lacks a variety of Asian art. However, it is unclear whether Google lacks artwork collections or there is an issue with the facial recognition. Apart from these issues, users are also wondering what the search giant plans to do with theses selfies. This @Google app feels racist but also there doesn’t seem to be enough Asian art represented in their database. pic.twitter.com\/OXhsW37Fk7 — Tiffany Diane (@tiffanydian) January 15, 2018 These selfies could be uploaded to Google as a way to train Google’s facial recognition algorithms. However, Google has explicitly denied it in its blog post. They said, “We created an experiment that matches your selfie with art from the collections of museums on Google Arts & Culture. Even if your art look-alike is a surprise, we hope you discover something new in the process. (By the way, Google doesn’t use your selfie for anything else and only keeps it for the time it takes to search for matches.)” And according to Google’s terms of service, “Any content that you upload to its servers can be used for the limited purpose of operating, promoting and improving our services, and to develop new one. The license continues even if you stop using our services.","excerpt":"You must have noticed of late that your friends and followers are posting selfies with their artwork doppelgängers on social media. It all sprang from the popular facial recognition art app called Google Arts and Culture app, which matches your selfie with a renowned pieces of artwork.   The app also lets you explore art […]","categories":["IT Services"],"tags":["Android","Facial Recognition","Google","iOS"],"author_name":"Smita Sinha","publish_date":"2018-02-19T07:22:53","publication_year":"2018","word_count":1021,"keywords":["Go","Facial Recognition","programming_languages:R","AI","iOS","programming_languages:Go","Git","GRU","Google","R","Android"],"extracted_tech_keywords":["AI","R","Go","Git","GRU","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/googles-arts-culture-app-uses-ai-facial-recognition-find-museum-doppelganger\/","complexity_score":4,"technical_depth":7,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10123838,"title":"This Sequoia-Backed Startup Uses AI to Help You Sleep Better","content":"Last week, at its first-ever media conference in India, Sequoia-backed Indian startup Wakefit announced its intention to improve sleep health and quality for Indian consumers, this time using AI. Regul8 is one of the flagship products of the company’s recently launched Wakefit Zense, the country’s first AI-powered sleep solutions suite. It is India’s first mattress temperature controller, which allows users to manually set the temperature between 15°C and 40°C or choose from presets like neutral, cold, warm, ice, and even fire. The icing on the cake: It also supports dual preferences, so individuals can customise their sides of the bed independently, eliminating the common household dispute over the AC remote. On an average, a home air conditioner in India can use about 3,000 watts of electricity an hour in India. However, thanks to Wakefit’s recently launched Regul8, the sleep controller mattress, you don’t need an AC anymore. Most importantly, it is 60% more energy-efficient than a 1.5-ton air conditioner. The other flagship product is an AI-powered contactless sleep-tracking device called Track8. Explaining the technology behind Track8, Yash Dayal, the CTO of Wakefit, told AIM,  “The tracker works by placing a passive sensor below the mattress, and as a person sleeps, it leverages ballistocardiography. This means that tiny vibrations from heartbeats, snoring, or any movement are read by our sensor. “These raw signals are then processed through AI and ML models to derive sleep metrics, restfulness, and other body vitals.” Integrating both Track8 and Regul8 can create a system where temperature regulation is based on how you sleep. According to Dayal, the products, including AI and ML models, were made by the in-house team consisting of about 80 to 100 tech experts. Discussing the broader AI strategy, Dayal said, “We don’t rely on generative AI models for these consumer products, but we are experimenting with models like those from OpenAI and Gemini internally for efficient supply chain management, demand planning, and forecasting.” What about Data Privacy? “Users have the option to choose whether to share their data or not. Their data is used solely to provide insights on sleep quality, such as heart rate variability, respiratory rate, movement index, and snoring index. It is anonymised, encrypted, and used only to improve their sleep quality,” said the director and co-founder Chaitanya Ramalingegowda, in an exclusive interview with AIM. Making High-Quality Products Affordable for All Founded in 2015, Wakefit’s philosophy is rooted in making high-quality products affordable for middle-class consumers. Led by co-founders Ankit Garg and Ramalingegowda, Wakefit has created a foundation for itself by building coming-of-age products like orthopaedic mattresses and dual comfort solutions, but at affordable prices. “Ankit and I come from middle-class backgrounds, so we understand the constraints of operating on a budget. From the beginning, our philosophy has been to make high-quality products affordable. For example, why should an orthopaedic memory foam mattress cost INR 50,000 when it can cost INR 12,000?” said Ramalingegowda. This principle extends to their latest tech-enabled sleep solutions, which are designed to offer cutting-edge technology at a fraction of the cost of similar products in the US and Europe. “The core idea was to leverage this success in non-tech products by applying material sciences, to create something uniquely beneficial for sleep,” added Ramalingegowda. “In markets like the US and Europe, similar products cost around INR 4.5 lakh. We built our product from the ground up to make it available for INR 45,000 to 50,000. We are middle-class folks building for middle-class India,” explained Ramalingegowda. For example, high-end sleep technology brands like Eight Sleep offer smart mattresses with advanced sleep tracking and temperature control features, at around INR 2.8 lakh for the Pod 3 model. Similarly, Sleep Number provides adjustable mattresses with integrated sleep tracking and temperature regulation, often exceeding INR 3 lakh. Chili Technology and Bryte also offer premium sleep solutions with advanced features, with prices reaching up to approximately INR 3.75 lakh. By offering its products at a mere fraction of that price, Wakefit positions itself as a cost-effective alternative to these luxury competitors, making sleep technology more accessible to the middle-class Indian market. This strategic pricing is an effort to bridge the gap between affordability and high-quality sleep solutions. Future Prospects Looking ahead, Wakefit plans to integrate generative AI and work with voice-based ecosystems to provide more actionable insights rather than passive data reporting. “Over the next two years, we will steadily launch new features and products, adding more value to our customers,” concluded Ramalingegowda. Wakefit’s commitment to innovation and accessibility positions it as a leader in the sleep technology market in India. With the launch of Wakefit Zense, the company is set to revamp how Indians experience sleep, making AI-powered sleep solutions an integral part of everyday life.","excerpt":"Besides consumer products, Wakefit is experimenting with models like OpenAI’s GPT-4 and Google’s Gemini to improve supply chain management, demand planning, and forecasting.","categories":["Deep Tech"],"tags":["Generative AI"],"author_name":"Shritama Saha","publish_date":"2024-06-17T18:06:50","publication_year":"2024","word_count":787,"keywords":["Go","OpenAI","AI","innovation","ML","RAG","Aim","ViT","generative AI","Generative AI","R"],"extracted_tech_keywords":["AI","ML","generative AI","OpenAI","Aim","RAG","R","Go","ViT","innovation"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/this-sequoia-backed-startup-uses-ai-to-help-you-sleep-better\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10168943,"title":"Akamai Launches Firewall Solution to Secure AI Applications","content":"Cloud computing platform Akamai has unveiled a new cybersecurity solution, Firewall for AI, designed to address the growing security vulnerabilities associated with AI applications. Akamai’s Firewall for AI is designed to provide a multilayered defence and real-time AI threat detection for AI-powered applications, LLMs, and AI-driven APIs. As AI technologies, including LLMs and agentic AI, gain widespread adoption across industries, the company aims to mitigate the new and evolving threats targeting these systems. The company explains that security measures such as web application firewalls (WAFs) are not equipped to handle the specific risks posed by AI, including adversarial attacks, model extraction, and data scraping. It further stated that while AI technologies provide significant productivity benefits, new attacks emerge. Attackers could target AI models to manipulate their responses or extract valuable proprietary data, making it crucial for organisations to implement specialised security measures. “Traditional security solutions do not stop AI threats,” said Rupesh Chokshi, SVP and GM of application security at Akamai, in a press release. “As enterprises embrace the AI era to unlock new levels of productivity, AI security must be a key enabler. Firewall for AI is a game changer for any enterprise adopting AI.” With Firewall for AI, Akamai plans to secure both inbound and outbound AI queries and responses. It ensures compliance and data protection, helping organisations meet regulatory standards by securing AI-generated outputs. Additionally, Firewall for AI offers flexible deployment options, enabling integration into existing security frameworks via Akamai’s edge, REST API, or reverse proxy. It also includes proactive risk mitigation by filtering AI outputs to prevent toxic content, hallucinations, and unauthorised data leaks. These features collectively provide comprehensive protection for AI-powered applications, ensuring both security and performance. With these new tools, Akamai aims to support enterprises in adopting AI technologies securely, ensuring that their data and intellectual property remain protected as they integrate AI into their operations.","excerpt":"Akamai leverages AI to protect other AI-powered applications.","categories":["AI News"],"tags":["AI Applications","Akamai"],"author_name":"Ankush Das","publish_date":"2025-04-30T16:21:05","publication_year":"2025","word_count":312,"keywords":["API","agentic AI","TPU","AI","cloud computing","AI Applications","REST API","Aim","Akamai","ViT","GAN","R"],"extracted_tech_keywords":["AI","agentic AI","Aim","cloud computing","TPU","R","API","REST API","GAN","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/akamai-launches-firewall-solution-to-secure-ai-applications\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":22943,"title":"Rajasthan Uses AI, Hackathons To Bridge Rural-Urban Divide","content":"Ahead of the state elections, Vasundhara Raje-led government organised a 36-hour non-stop coding event in Rajasthan, to make people of Rajasthan realise the importance of the technology and ‘bridge the rural and urban divide’. Nearly 6500 students from across the state participated in the marathon coding event to develop mobile apps, chatbots and programs that could give the state’s digital aim a bigger push. “This event has seen a footfall from all sections of society, especially the vulnerable.” Akhil Arora, Principal Secretary in the Department of IT and Communications said. Three students from Lovely Professional University developed a mobile application to make the voting process easy. To cast your vote on the app, the user just needs to authenticate themselves by entering their Aadhaar number. “Each candidate is mentioned by their election names and the voter can easily vote from home,” a 21-year-old developer said. Students from Jaipur College, who participated in the event has devised ‘antidepressant chatbots for various Rajasthan government department and that would help employees facing depression. The coders also developed programs to tackle the civic problem. As soon as the manhole is about to overflow with water due to rainfall, the prototype designed by the students will send an alert via email, message or tweet to the ministry. “This prototype has completely waterproof electrical wiring, which is attached just below a manhole cover. When it rains or water starts accumulating, it sends a tweet message and an email with the location to the concerned department, alerting about the rising level of water. It costs Rs 500,” a 19-year-old student who has been working on the project for the last couple of months said. The Hackathon 4.0 was organised as part of the Rajasthan Digifest, commemorating the state IT day on March 21. Participants which included students, professionals, startups and organisation got a platform to showcase their ideas on artificial intelligence, machine learning IoT, Blockchain, data warehouse, AR\/VR and data mobility.","excerpt":"Ahead of the state elections, Vasundhara Raje-led government organised a 36-hour non-stop coding event in Rajasthan, to make people of Rajasthan realise the importance of the technology and ‘bridge the rural and urban divide’. Nearly 6500 students from across the state participated in the marathon coding event to develop mobile apps, chatbots and programs that […]","categories":["AI News"],"tags":["elections","Hackathons"],"author_name":"Smita Sinha","publish_date":"2018-03-23T13:00:10","publication_year":"2018","word_count":324,"keywords":["Go","artificial intelligence","machine learning","AI","chatbots","Hackathons","Git","Aim","GAN","R","elections","data warehouse"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","Aim","chatbots","R","Go","Git","data warehouse","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/rajasthan-uses-ai-hackathons-to-bridge-rural-urban-divide\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":17828,"title":"Here’s Why Great Lakes Big Data Analytics Program Is The Leading Analytics Training Of Choice","content":"In the high demand field of data and analytics, it is imperative to choose a course that gives the learner the knowledge and practical skills to develop deep analytical talent and transition successfully to an analytics job role. According to our research, Big Data and analytics industry in India is currently estimated to be $2.03 Billion annually in revenues, growing at a healthy rate of 23.8% CAGR and the industry is expected to double by 2020. In terms of jobs, the jobs market looks healthy with close to 50,000 positions related to analytics being currently available to be filled in India. The top leading organizations in terms of hiring are Amazon, Citi, HCL, Goldman Sachs, IBM, JPMorgan Chase, Accenture, KPMG, E&Y and Capgemini. However, to break into the field, one needs a program that is up-to-date with the latest and gives a comprehensive understanding of the use and implementation of data analytics in business settings with live datasets. Hands-on exposure and experiential learning can play a big role in shaping your understanding of data analytics. You need a program that teaches you to effectively design data models through advanced statistical and data programming techniques and gives you the opportunity for large-scale data analysis to discover meaningful patterns and trends. To top it all, a Capstone project serves as a real-world case study where learners apply their classroom learning and in addition, it also adds great weightage to the resume. Great Learning in association with Great Lakes Institute of Management has rolled out the most exhaustive career-based Big Data Program – PGP-Big Data Analytics that offers practical knowledge one can use in the workplace. Backed by hands-on access to live datasets and cutting-edge analytics software in the 24*7 Big Data Lab, learners who wish to advance to the high-profile field can gain deep knowledge from acknowledged industry experts and renowned faculty who have significant experience in data and analytics field. When it comes to picking the right program, it can be hard to choose the one that fits your needs and also advances your career. This is where the Big Data Program offered by Great Learning in association with Great Lakes Institute of Management differs from the rest of the programs. PGP-Big Data Analytics goes beyond tools and trains candidates on Big Data Technologies, Machine Learning, Visualization and Data Science models. The top B-school also has great ties with industry veterans, making it a great fit for those on the lookout for networking opportunities. Analytics India Magazine highlights top features of the programs India’s Most Exhaustive Big Data Program: With an increased focus on big data technology stack, the 12-month program equips learners with the latest big data tools and technologies in Data Science, Machine Learning, and Visualization with hands-on exposure by top faculty and industry experts. The course gives a thorough understanding of core statistical foundation, covering areas such as Descriptive and inferential statistics, Hypothesis testing and estimation, Predictive analytics – regression (Ordinary least squares, multiple   linear, logistic). 24*7 Access to Big Data Lab (Access to Big Data Lab for hands-on exposure: All the technology tools for data ingestion, processing, analysis and visualization are stably installed, maintained and hosted for the learners to access at any time for assignments and the Capstone project. The tools include – Big Data Tools (Hadoop, Spark, Hive, Pig, etc.) – Visualization Tools such as Tableau and Gephi – Programming Environments such as Python and R. Career Services: Great Lakes has an active internal jobs board and a huge industry connect. Also, students are offered structural support, wherein predecessors come and share their job experience. There are career workshops and industry sessions on resume building. Students can also explore opportunities by industry partners wherein specific roles in leading companies are filled as per need basis. Capstone Project: Capstone project has a lot of potential benefits for learners and can help them build their work portfolio. It not only helps them apply classroom learning to real life settings and support collaborative student engagement, a data science project also boosts the employability of the student and gives hands-on experience of solving business problem.  Remember, employers prefer candidates who have demonstrated experience working with large datasets. Experiential Learning: Knowing the fundamentals and tools is only half the challenge. The main emphasis is on solving real-world business problems. During the course of the program, students get hands-on exposure and experiential learning through Capstone Project, real world case studies, 24*7 Big Data Lab with Data Sets, Industry guest lectures, and webinars. Candidates get exposed to different technologies like Hadoop, R, Python, Pig, and Hive. Industry Perspective Lectures: At Great Learning, one can tap into the vast industry network and gain exposure by interacting with industry experts working on Big Data tools and technologies. While a lot of data science is about big data tools and technologies, candidates who are preferred the most are those who can solve business problems and not just run a clean and mung data. The industry guests and mentors fill the gap between theory and practice by sharing with candidates current business problems through practical case studies and examples. Corporate Partners: The program has been co-created and delivered by senior industry professionals as well as Great Lakes faculty. And it is definitely anchored in real-world with experiential learning that extends beyond classroom with case studies, self-paced projects, assignments and access to big data lab. The curriculum will help learners gain a deep understanding of the entire data value chain – the flow of information from inception to analysis to insight – that is essential to drawing insights from big data. Blended Learning: Aimed at working professionals, the program is delivered in a blended learning format that allows learners to pursue their day job with minimal disruption and also make time for studies around the work schedule. The classroom sessions are assisted by online webinars, discussions, and assignment that keep your learning continuous and cumulative. Last Word There are a host of big data programs in the market that offer quick fix short-term courses, however most lack the credentials, faculty and industry heft to give learners the ROI and eventually translate into lucrative jobs. Before diving into a program and investing money upfront, one must understand that analytics is a broad field and the best school for analytics training is one that offers a comprehensive, well-rounded education with an emphasis on the business side as well. PGP-Big Data Analytics offered by Great Learning in association with Great Lakes Institute of Management boasts of renowned faculty members who have clinched coveted spots amongst the Top 10 academicians in India. Besides, the vast alumni network underpins the program’s success and a clear sign that it helps learners evolve into hands-on data analyst rather than just adding a bunch of analytical skills to the resume. And that’s why the Great Lakes Big Data Analytics program is the leading choice for professionals who want to transition to data science.","excerpt":"In the high demand field of data and analytics, it is imperative to choose a course that gives the learner the knowledge and practical skills to develop deep analytical talent and transition successfully to an analytics job role. According to our research, Big Data and analytics industry in India is currently estimated to be $2.03 […]","categories":["AI Trends"],"tags":["Big Data Analytics"],"author_name":"Richa Bhatia","publish_date":"2017-09-25T07:34:37","publication_year":"2017","word_count":1158,"keywords":["data science","Go","API","machine learning","AI","R","Big Data Analytics","Python","Aim","analytics","predictive analytics"],"extracted_tech_keywords":["AI","machine learning","data science","analytics","Aim","predictive analytics","Python","R","Go","API"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/heres-why-great-lakes-big-data-analytics-program-is-the-leading-analytics-training-of-choice\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10039972,"title":"Role Of CISO In The Post-Pandemic World","content":"The pandemic has forced companies to change their tack. From a purely work-from-office set up, enterprises have now moved to remote or hybrid working environments. Despite obvious advantages, such arrangements come with heightened security risks. According to a report, losses incurred from cybercrimes in 2020 amounted to $1 trillion. Increased risk and hasty technology rollouts resulted in a spike in cybersecurity breaches. During this time, the role of chief information security officer (CISO) has become indispensable. A CISO is responsible for establishing and maintaining enterprise security and strategy to ensure the information assets and technologies are protected. To understand the role and responsibilities of a CISO, Analytics India Magazine caught up with Saurabh Kewlani, head – IT & CISO, Hannover Re (India Branch). Hannover Re is the third-largest reinsurer in the world. The firm started its operations as a licensed foreign branch in India in February 2017. Kewlanii oversees the IT Delivery, IT Strategy and Cybersecurity at Hannover Re’s India Branch. Excerpts: AIM: How did the pandemic redefine the role of CISO? Saurabh Kewlani: There has been a significant impact on the role of CISOs due to the pandemic. The major shift has been towards securing the end-points of the remote employees, which tend to be more vulnerable as compared to a corporate network. The surge in remote work has increased the risks of cyberattacks. Statistics show that phishing attacks have increased by over 600 percent since March 2020. CISOs, therefore, have had to learn to patch their networks as quickly as possible and at the same time tackle new attack vectors. Much attention has also been drawn towards reviewing concerns with third-party apps used for video-conferencing and collaboration over cloud platforms. AIM: What does a typical day look like for a CISO? Saurabh Kewlani: CISO needs to be on top of things all the time. Cybersecurity threats and risks can manifest at any time as the attacker could be anywhere and in any time zone. A CISO, therefore, needs to be aware of the daily threats briefs and stay on top of the emails. To minimise a threat’s impact, a CISO has to constantly review the Response Plan to identify, contain and remediate risks. Especially in the post-pandemic world, a CISO also has to constantly educate the staff on how to operate in a Zero-Trust environment and ensure that employees receive regular Cybersecurity training. Not only that, CISOs have to educate themselves on the technological advancements to stay regularly updated. A CISO has to wear multiple hats, but not necessarily all of them on the same day! AIM: How important is it for a company to have defined roles for security and maintenance? Saurabh Kewlani: It is imperative for organisations to define the roles for security and maintenance clearly. This helps in the understanding of how the protection of information has to be accomplished and communicated. The definitions of the roles help clarify, coordinate the activity and actions necessary to disseminate security policy, standards, and implementation. AIM: What are the biggest trends in enterprise security right now? Saurabh Kewlani: One of the most popular trends, which is also a leading concern for organisations, is data breaches. Any minor flaw or bug in the system browser or software is a potential vulnerability for hackers to access personal information. Some of the emerging security threats include: Automotive hacking: Hacking vehicles that are equipped with automated software (cruise control, engine timing, door locks) and with communication tools that use WiFi\/Bluetooth. Cloud vulnerability: Cloud Platforms are vulnerable to account hijacking, data breaches & Insecure APIs. Mobile devices, which have become an integral part of an individual’s life, storing critical information and data has also come under the radar of bad actors. However, the integration of AI with cybersecurity is a positive trend. AI-enabled threat detection systems can help predict new attacks and notify security teams of any data breach instantly. These trends are expected to continue in the near future and develop further. AIM: How do emerging technologies such as AI and machine learning help in adopting best security practices? Saurabh Kewlani: Artificial intelligence (AI), along with machine learning (ML), has brought tremendous changes in cybersecurity. This technology has been instrumental in building self-learning systems that are adept in identifying suspicious patterns, detecting threats and providing an automated response to cyber-attacks in real-time. AI\/ML can help the security team classify internal data, spam and malicious activity with predetermined parameters. When used optimally, this technology can lighten the weight of a heavy cybersecurity workload and reduce human error and oversights. AIM: Tips to build a successful career in cybersecurity. Saurabh Kewlani: Cybersecurity is a very broad field with multiple domains like digital forensics, network security, information security and many more. It is, therefore, important to research the different domains before kick-starting your career. Once the candidate zeros in on the path he\/she would like to take, they should try to specialise in that domain and arm themselves with the required certifications. This will add value to the résumé and increase the chances of landing a job. Participating in hackathons also gives a lot of visibility to companies that are recruiting. Lastly, cybersecurity is an ever-evolving field. So one is always required to update existing knowledge and skills with the changing cybersecurity trends.","excerpt":"The pandemic has forced companies to change their tack. From a purely work-from-office set up, enterprises have now moved to remote or hybrid working environments. Despite obvious advantages, such arrangements come with heightened security risks. According to a report, losses incurred from cybercrimes in 2020 amounted to $1 trillion. Increased risk and hasty technology rollouts […]","categories":["AI Features"],"tags":["CISO","cyber threats","Cybersecurity","Interviews and Discussions"],"author_name":"Shraddha Goled","publish_date":"2021-05-12T13:00:00","publication_year":"2021","word_count":878,"keywords":["API","artificial intelligence","cyber threats","machine learning","AI","ML","CISO","Git","Aim","analytics","Rust","Cybersecurity","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","analytics","Aim","R","Rust","Git","API"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/role-of-ciso-in-the-post-pandemic-world\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10099884,"title":"Why Tower Semiconductor Emerges as a Compelling Choice for Reliance","content":"Mukesh Ambani-owned Reliance Technologies is going big on AI. Last week, it announced a major partnership with NVIDIA to develop India’s foundational large language model trained on the nation’s diverse languages. NVIDIA will provide access to Reliance with GH200 Grace Hopper Superchip and NVIDIA DGX Cloud for exceptional performance. The partnership points towards Reliance’s big ambition to emerge as a key player in the AI space and later in the semiconductor industry. The Indian conglomerate is also heading in the same direction. Last year, it was reported that Reliance, along with HCL Technologies, was planning to acquire stakes in the International Semiconductor Consortium (ISMC), a collaborative effort between UAE’s Next Orbit Ventures and Israel’s Tower Semiconductor. The fate of the ISMC consortium was dependent on Intel’s acquisition of Tower Technologies, however, since Intel’s attempt to acquire Tower did not materialise within the expected timeframe, it is highly probable that the ISMC consortium has been dissolved or is no longer viable. Now, Reliance Industries is possibly in talks with a foreign chipmaker to set up a semiconductor fab in the country, Reuters reported. So far, the name of the foreign chipmaker that Reliance is in talks with, has not been revealed. But given that Reliance was already in discussions to acquire stakes in the ISMC consortium, it would be logical for them to pursue the acquisition of Tower Technologies, especially considering the Intel deal did not materialise. Reliance should look at Tower Mampazhy told AIM that Intel’s inability to acquire Tower Semiconductor presents an opportunity for Indian businesses like Reliance that are looking to expand into the semiconductor space. The primary reason Reliance should look at acquiring Tower Semiconductor is solely because it is available for grabs. Intel’s deal to acquire Tower was valued at USD 5.4 billion. At this price point, the acquisition will give Reliance Industries an established semiconductor manufacturing company, allowing it to enter the semiconductor industry with an existing infrastructure, expertise, and customer base. Mampazhy believes that given Reliance’s financial might, it can opt for buying out Tower Semiconductor and open a branch fab in India. “Another viable option is for Reliance to acquire stakes in Tower and accompanied by an agreement from the company to make substantial investments in the Indian semiconductor industry,” he said. Last year, Tower Semiconductor, as part of the ISMC consortium, entered into a Memorandum of Understanding (MoU) with the Karnataka government to establish a 65nm technology node analog fabrication facility. But the plan was shelved after the fall of the Intel-Tower deal. Considering Tower Semiconductor’s prior interest and the dissolution of the ISMC consortium, Reliance Industries might rekindle Tower’s initial enthusiasm for establishing a presence in India. Tower to benefit as well Not only Reliance but also Tower Semiconductor stands to gain substantially from a prospective arrangement with Reliance or any other Indian enterprise, for that matter. If Tower is unwilling to sell, an alternative path for Tower could be to become a technology partner. For Tower, to become a technology partner, it won’t require a significant investment, given that 50% of the cost will be borne by the central government and state governments would chip in with another 20-25%. If we assume, the cost of setting up a fab, in partnership with Reliance, is USD 4 billion, it would mean Tower will be making an investment of a few hundred million. Moreover, another less likely option is for Tower to become a technology provider, instead of a technology partner. “The concept of technology providers involves a technology transfer agreement where ongoing support is provided until the new fabrication facility integrates the technology successfully. A licensing fee is levied for this service. After the facility becomes operational and the technology functions effectively Tower Semiconductor disengages. However, the Indian government seems less inclined towards this approach,” Mampazhy said. Reliance could also take SCL Interestingly, in the past, Tower Semiconductor has helped with the upgrade of the Semiconductor Laboratory (SCL) to 180 nm technology. SCL, which is currently under the Ministry of Electronics and IT (MeitY), is renowned for developing chips used in different ISRO missions such as Chandrayaan and Mangalyaan. “SCL in India stands as one of the rare semiconductor manufacturers that can proudly claim their chips have journeyed to the Moon and Mars,” Anshuman Tripathi, member of the National Security Advisory Board (NSAB), previously told AIM. Last year, the Union Cabinet approved a modernisation plan for SCL and is currently searching for a suitable candidate to manage India’s sole existing fab. Reliance Industries could potentially be a strong contender for taking over SCL, especially considering Mukesh Ambani’s affiliations with the BJP government, which could work to Reliance’s advantage in this scenario. Mamphazy too believes Reliance could potentially be a good candidate to take over the SCL. “Tower did the technology transfer to SCL 10-15 years back and so the same Process Design Kit (PDK) applies for both fabs and so SCL can act as a ‘second source’ for Tower’s 180 nm orders,” Mamphazy said. Moreover, the 180 nm semiconductor fabrication line at SCL offers the capability to manufacture a range of products, including microcontrollers, sensors, and communication chips. These components have diverse applications, potentially benefiting Reliance Industries, particularly in sectors such as telecom, energy and electronics. The Reuter report also stated that one of the major reasons Reliance is foraying into semiconductors is to address its supply chain needs. “More than anything else, SCL will be a perfect opportunity for Reliance to ‘get its hands dirty’ on how to run a fab in India,” Mamphazy said.","excerpt":"Acquiring Tower will give Reliance an established semiconductor manufacturing company, allowing them to enter the semiconductor industry with an existing infrastructure, expertise, and customer base.","categories":["AI Highlights"],"tags":[],"author_name":"Pritam Bordoloi","publish_date":"2023-09-12T17:36:37","publication_year":"2023","word_count":925,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","Ray","Aim","R"],"extracted_tech_keywords":["AI","Aim","Ray","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/why-tower-semiconductor-emerges-as-a-compelling-choice-for-reliance\/","complexity_score":3,"technical_depth":7,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10171918,"title":"Reddit Unveils AI Ad Tools to Help Brands Access Real-Time Community Insights","content":"Just a week after Pinterest unveiled its AI-powered ad features, Reddit introduced ad tools at the Cannes Lions International Festival of Creativity based on its Community Intelligence to help marketers leverage user-generated content for key insights and marketing performance. The first tool, Reddit Insights, is focused on providing real-time insights to marketers for planning campaigns by identifying trends on the platform and testing campaign ideas. Infamous for its controversies on the platform, the tool now serves as a social listening tool to detect signals camouflaged as noise. Another feature called Conversation Summary Add-Ons allows brands to display positive user-generated comments directly under the ads, thereby adding an authentic touch. Reddit claims that the alpha tests with Jackbox Games and Lucid showed a 19% higher click-through rate than standard image ads. Jen Wong, the company’s chief operating officer, told Axios that Reddit’s quick response to what’s happening in the world helps businesses make decisions in everything from how to market to performance in customer service. Brands are demanding high-performing tools to penetrate the fast-paced digital crowd, and platforms like Reddit, Snap and Pinterest have no option but to provide a basket of real-time ad tools to their clients. To tackle economic uncertainties, platforms are also aiming to provide advertising tools to attract small and medium-sized businesses (SMBs), thereby diversifying advertiser portfolios. Major holding companies like WPP media lowered their ad growth from 7.7% to 6% pointing to geopolitical and economic challenges. Despite the situation, Reddit projected stronger-than-expected Q2 revenues. This can be attributed to the launch of new ad tools since its IPO last year in March 2024. Following these innovations, Reddit is also developing generative AI tools like Reddit Answers—a feature that will offer AI-generated summaries linked to conversations, pointing to its next phase of community-driven ad evolution","excerpt":"Brands are demanding high-performing tools to penetrate the fast-paced digital crowd, and platforms like Reddit, Snap and Pinterest have no option but to provide a basket of real-time ad tools to their clients.","categories":["AI News"],"tags":["ad tools","Advertising","cannes lions","features","Jen wong","Marketing","Pinterest","reddit","snap"],"author_name":"Smruti S","publish_date":"2025-06-18T14:58:38","publication_year":"2025","word_count":298,"keywords":["reddit","Git","Pinterest","R","cannes lions","RAG","snap","ViT","AI","Advertising","ad tools","generative AI","programming_languages:R","IPO","innovation","features","Aim","Jen wong","Marketing"],"extracted_tech_keywords":["AI","generative AI","Aim","RAG","R","Git","ViT","innovation","IPO","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/reddit-unveils-ai-ad-tools-to-help-brands-access-real-time-community-insights\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10123488,"title":"Why is India Not Leading the Innovation Sector?","content":"In 2023, India filed over 90,000 patents, which is an average of 247 patents per day, marking a notable 17% increase – the highest in two decades. However, it still does not rank in the top 10 countries for patent filings, raising the question of its position in the global innovation sector. “Many patents attributed to Western companies are the result of work conducted in India, However, these patents are registered under the names of Western companies due to the global operations,” said TV Mohandas Pai, chairman at Aarin Capital, at the recently concluded IGIC 2024 in Bengaluru. Pai highlighted that the Indian research culture does not consider patent filing as an objective, which has been a hindrance. While in the US, a significant number of patents are filed defensively driven by broader strategic reasons rather than for immediate utility. Source: SBI Research and Indian Patent Office report Ingredients of Innovation System “For any innovation system, there are three main ingredients – human capital, physical capital, and financial capital,” Pai noted. India’s overall Digital Competitiveness Score of 60 (out of 100), places it ahead of all BRICS nations except China, reflecting the growing tech talent in the country. According to NASSCOM, India has approximately 4.5 million IT professionals, including those working in the software export industry. “India is second only to the US in the number of highly skilled technology professionals. Notably, 60% of the professionals in the export industry work for American companies,” Pai mentioned. Additionally, India has a significant digital infrastructure, with 1.31 billion mobile phone users and substantial data consumption. The country has achieved impressive milestones in digital public infractures, especially with UPI’s daily transactions expected to hit a billion by 2026. However, Pai pointed out that India lacks cutting-edge research labs and a robust hardware industry, crucial for leading global innovation. Hurdles to Cross Coming down to financial capital, Pai noted that over the past decade, $145 billion has been invested in India, a contrast to China’s $835 billion and the US $2.3 trillion. The scarcity of financial capital is a significant constraint for India. “Public funding for research is insufficient, with Indian universities receiving around $800 million annually, less than what a single prominent American university might receive,” he said. In 2020, India filed nearly 57,000 patents, which is a mere 4% of the 1.49 million applications filed in China and 9.5% of the 5,97,000 applications in the US. Similarly, India grants 23,361 patents, compared to 5,30,000 in China and 3,50,000 in the US. On an average, it takes about 58 months to process a patent application in India, while in China it takes about 20 months, and 23 in the US. Source: World Intellectual Property Indicators 2021 report The delay is primarily due to a shortage of manpower in the Indian patent office. As of the end of March 2022, the Indian patent employed only 860 people, including examiners and controllers. In contrast, China employed 13,704 people, and the US employed 8,132. As of March 31, 2022, approximately 1,64,000 applications were pending at the controller level in India. Slowly Catching Up Despite the challenges, there have been some positive developments. In 2023, central universities like JNU, DU, Jamia, AMU, Maulana Azad National Urdu University, and UoH received an additional Rs 826.28 crore, increasing their total allocation from Rs 3,191.55 crore in 2021-22 to Rs 4,017.83 crore in 2022-23. Similarly, if you look at the generative AI startups emerging in India, the numbers only indicate how far the country is progressing. Currently, approximately 70 Indian startups are creating products and solutions focused on generative AI. It may not be too long before India is also considered an innovation capital of the world.","excerpt":"“Public funding for research is insufficient, with Indian universities receiving around $800 million annually, less than what a single prominent American university might receive,” TV Mohandas Pai said.","categories":["AI Features"],"tags":["AI (Artificial Intelligence)","India"],"author_name":"Vidyashree Srinivas","publish_date":"2024-06-13T10:30:00","publication_year":"2024","word_count":617,"keywords":["API","funding","programming_languages:R","AI","innovation","Git","RAG","generative AI","R","India","AI (Artificial Intelligence)","startup"],"extracted_tech_keywords":["AI","generative AI","RAG","R","Git","API","innovation","startup","funding","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/why-is-india-not-leading-the-innovation-sector\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10165695,"title":"How ‘Women in Cloud’ Flips the Script on AI and Gender Bias","content":"While everyone’s busy fearing AI for wiping away jobs, something that demands equal stirring is AI deepening the existing gender inequalities. Since AI learns from historical data, when the training data reflects societal biases, AI systems inherit and amplify discrimination. From hiring algorithms to wage prediction, models may deepen gender pay gaps. A study by the Berkeley Haas Center for Equity, Gender & Leadership examined 133 AI systems across various industries and revealed that approximately 44% exhibited gender bias. To make matters worse, many women hesitate to engage with AI tools, fearing obsolescence rather than seeing it as an opportunity. In an interview with AIM, Chaitra Vedullapalli, Women in Cloud founder and president, shared her insights on the challenges women face in the tech industry. Throughout her career, she discovered that women lacked access to essential funding and resources. She realised that the issue wasn’t just a lack of investment in women-led businesses, it was a larger challenge of preparation and access. During her tenure at Microsoft, Vedullapalli, also a TEDx speaker, observed that despite the industry’s advancements, women continued to face significant barriers in entrepreneurship and leadership roles in tech. Reflecting on her journey, she noted, “At Microsoft, I was recruited as a ‘compete hire’ and was tasked with creating a Partner Velocity Platform (an $80 million project that unified all partner programs into a single technology platform).” This experience gave her a holistic understanding of software creation, licensing, and sales channels, besides shaping her perspective on the industry’s systemic challenges for women. Levelling the Playing Field This realisation pushed her to create Women in Cloud, an initiative designed to generate $1 billion in economic access for women by addressing systemic barriers. She strongly believes that while discrimination exists, women who are well-prepared and demonstrate strong business acumen can successfully navigate the industry. “Recognising the difficulties faced by women, especially those balancing multiple jobs, Women in Cloud prioritised flexibility in its training programmes,” Vedullapalli said. The initiative launched in India and has since expanded, impacting 5,000 women over the past three years. Of these participants, 18% have secured jobs, while 30% have earned Microsoft certifications. These certifications play a crucial role in levelling the playing field, ensuring that women enter the workforce on equal footing with men, eliminating hierarchical disparities, and strengthening their competitiveness in the job market. Additionally, acquiring a strong foundation in technology empowers women to pursue entrepreneurship, equipping them with the skills to develop AI and cybersecurity solutions. Providing for the Indian Market Given the high certification costs in India—ranging from ₹4,00,000 to ₹5,00,000—Women in Cloud sought innovative funding solutions. By leveraging corporate partnerships, it secured 1,000 fully funded scholarships underwritten by Microsoft and Coursera, evenly distributed between India and the global market. The demand was overwhelming, with 4,000 applications received in a single day. Over time, 42,000 scholarships have been distributed across 100 countries. Since certification exams require rigorous preparation, comparable to IIT entrance tests, the programme also offers essential support systems, including community engagement and expert trainers, to assist candidates. While the initiative primarily focuses on women, it takes an inclusive approach by allocating 80% of opportunities to women while reserving 20% for men who require support. The organisation prioritises allyship over exclusivity, ensuring that learning resources remain widely accessible. Today, the programme operates in 80 countries, with significant engagement in Nigeria, Morocco, Canada, the US, France, Germany, Thailand, the Philippines, Pakistan, Sri Lanka, and other global regions. Adding to this, Manju Dhasmana , senior director of Corporate Social Responsibility, at Microsoft India emphasises that “when women thrive in AI, businesses, and entrepreneurship, entire economies benefit.” In line with this belief, Microsoft  is making substantial investments in AI skilling to ensure accessibility for women across urban and rural India. Through initiatives such as Women in Digital Business and ADVANTA(I)GE India, Microsoft has already trained millions across the country. In 2024 alone, more than 2.4 million people were skilled under ADVANTA(I)GE India, with over 60% being women and more than 70% from non-metro areas—many from tier II and III cities—equipping them with the tools to build and scale their ventures. Further reinforcing this mission, Microsoft’s partnership with IndiaAI aims to equip 500,000 individuals—including women entrepreneurs—with essential AI skills by 2026. Rethinking DEI and Moving Beyond Fragmented Inclusion Vedullapalli challenges the current approach to diversity, equity, and inclusion (DEI), arguing that it is often biased and fragmented. She believes that instead of prioritising true accessibility, DEI initiatives have been divided into separate groups—such as women’s groups, LGBTQ+ groups, and racial groups—leading to a disjointed framework. Rather than enforcing rigid DEI policies, she advocates for a shift towards broadening access to resources in education, economic opportunities, and workplace accommodations. Rency Mathew, people leader–India and South Asia, and MD, Sabre Bengaluru, highlighted how DEI initiatives have become an integral part of India’s corporate landscape, with many organisations striving to foster inclusive work environments. However, she pointed out that “these efforts in India have predominantly focused on gender diversity. Considering the country’s vast diversity, it’s essential to broaden the DEI framework to embrace other dimensions.” Parul Dhir, director of DEI and employee engagement at Acuity Knowledge Partners, shared insights into the company’s recent DEI efforts, stating, “This past year, we launched three major initiatives: a mentorship program for women, a series of workshops on unconscious bias, and over 55+ DEI events and initiatives to engage our employees.” Ramya Parashar, chief operating officer at MiQ, emphasised the importance of translating DEI intentions into measurable impact. “One of the biggest hurdles is moving beyond intent to measurable impact, ensuring that DEI is not just a checkbox but a deeply embedded part of our culture. Breaking unconscious biases in hiring and leadership development, creating truly inclusive policies that cater to diverse needs, and sustaining engagement beyond surface-level initiatives require continuous effort.","excerpt":"The issue isn’t the lack of investment in women-led businesses, it is the larger challenge of preparation and access.","categories":["AI Features"],"tags":["international women's day","Women in Tech"],"author_name":"Shalini Mondal","publish_date":"2025-03-08T16:00:00","publication_year":"2025","word_count":969,"keywords":["Go","API","funding","AI","Git","RAG","international women's day","Aim","Women in Tech","ViT","GAN","R"],"extracted_tech_keywords":["AI","Aim","RAG","R","Go","Git","API","GAN","ViT","funding"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-women-in-cloud-flips-the-script-on-ai-and-gender-bias\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":60652,"title":"K7 Computing To Offer Cybersecurity Products For Free To Protect The Indian Cyberspace Amid Crisis","content":"At this time of extreme anxiety due to the chaos COVID-19 pandemic raging across the globe, cybersecurity firm K7 Computing has responded by committing to protect the public by giving its cybersecurity products for use at no cost. This COVID-19 crisis has been leading to several opportunities for cybercriminals to wreak havoc, exploiting employees’ situation of working from home, and the vulnerabilities in the security protections. This has led to a plethora of cybersecurity incidents that are being reported in several parts of the world. These security products are for laptops, desktops and Android mobiles, and endpoint protection in the organisations. According to Kesavardhanan J, the founder of K7 Computing, the company has been aiming to contribute to the cyber safety of all consumers and SME’s during this critical time of coronavirus pandemic. He said, “The products they are designing will keep the users protected in the current omnipresent digital world.” Due to the current 21-day lockdown, businesses and employees have been forced to work from home, which makes them rely on the security of their internet connections to keep businesses and services running, which make them more vulnerable to cyberattacks. According to reports, there is a 30% increase in cyberattacks with 1,756 blocked malicious instances, specifically using the COVID-19 theme. In an effort to ensure the cybersafety of the employees and the businesses, the company has made its globally-popular and most-awarded cybersecurity products — K7 Total Security and K7 Mobile Security completely free for users, and K7 Cloud Endpoint Security free for SMEs in the country for as long as required. “We are also extending this to every COVID affected country worldwide,” said Kesavardhanan. He further stated that this is a sudden change for Indian employees to work from home, which in turn is raising the likelihood of digital vulnerabilities that hackers could exploit. Similarly, many organisations, particularly hospitals which are under duress, are more likely to pay ransoms in the event of a ransomware attack. “As a responsible cybersecurity protector of the country, K7 feel the need to ensure that everyone is safe in the cyber world. We don’t want cyberattacks to impact the Indian economy further, adding even more stress to individuals and SMEs across every nook and corner of the country,” concluded Kesavardhanan.","excerpt":"At this time of extreme anxiety due to the chaos COVID-19 pandemic raging across the globe, cybersecurity firm K7 Computing has responded by committing to protect the public by giving its cybersecurity products for use at no cost. This COVID-19 crisis has been leading to several opportunities for cybercriminals to wreak havoc, exploiting employees’ situation […]","categories":["AI News"],"tags":["Cyber Attack","Cyber Security","Cyber security best practices","cyber security India","Cybersecurity","Cybersecurity India"],"author_name":"Sejuti Das","publish_date":"2020-04-01T17:30:00","publication_year":"2020","word_count":377,"keywords":["cyber security India","Cyber Security","programming_languages:R","AI","Cyber security best practices","Git","RAG","Cyber Attack","Aim","Cybersecurity India","GAN","Cybersecurity","R"],"extracted_tech_keywords":["AI","Aim","RAG","R","Git","GAN","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/k7-computing-to-offer-cybersecurity-products-for-free-to-protect-the-indian-cyberspace-amid-crisis\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10009821,"title":"Webinar: Democratizing Data Science With No-Code","content":"The pace at which the field of data science is expanding makes it impossible for the enterprises to catch up to the state of the art solutions. To avoid falling short of the demands of burgeoning digital revolutions, solutions like APIs, drag and drop analytics, low code and even no-code have been developed. For instance, no-code tools bridge the gap between domain expertise and niche skills of the developers. Know more about why enterprises are increasingly leaning on no code solutions in the upcoming webinar being organised by Analytics India Magazine in association with HP. This webinar will be helmed by Prashant Pansare, a triple MBA in International Marketing, Business Finance, and Corporate HR. Prashant is the Managing Director at Inteliment. Founded in 2004, Inteliment helps enterprises worldwide derive maximum business impact by providing Visual & Predictive Analytics, Data Science, IoT, Mobility & Artificial Intelligence Solutions. Prashant has also founded Rubiscape in 2020, a Made-in-India Data Science Product company to promote Data Culture, Design Thinking, Product Incubation. In this webinar, Prashant will address decision makers, students, data scientists, data enthusiasts, solution specialists, architects; and walk them through the various disruptive strategies and the importance of fluid and fast data platforms. Prashant will be joined by B Senthilraj, Solution Architect, HP India. Register Now What You’ll Gain By attending this talk, you will get to Know the latest trends in evaluating, adopting and scaling the platforms for diverse data to accelerate decision engineering processes. Understand how platform approach will help them focus more on innovation and deep analytics. Session Details Speaker: Prashant Pansare MD, Inteliment Group, CEO – Rubiscape Pvt Ltd Date: 23rd October’2020Time: 4:00 to 5:15 PM (IST) Register Now","excerpt":"The pace at which the field of data science is expanding makes it impossible for the enterprises to catch up to the state of the art solutions. To avoid falling short of the demands of burgeoning digital revolutions, solutions like APIs, drag and drop analytics, low code and even no-code have been developed. For instance, […]","categories":["Deep Tech"],"tags":["Democratisation","HP","webinar"],"author_name":"Ram Sagar","publish_date":"2020-10-15T19:00:51","publication_year":"2020","word_count":280,"keywords":["data science","API","artificial intelligence","AI","R","Git","RAG","Democratisation","analytics","HP","GAN","predictive analytics","webinar"],"extracted_tech_keywords":["AI","artificial intelligence","data science","analytics","RAG","predictive analytics","R","Git","API","GAN"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/webinar-no-code-data-science\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":14360,"title":"Data enabled products are defining the future of data science","content":"New age companies like Uber have made data enabled products If you want to know how to build great data science-enabled product, you needn’t look beyond LinkedIn, now Microsoft-owned. The Mountain View headquartered professional networking company has over 433 million members across 200 countries, allowing data scientists access to structured datasets that spawned cutting-edge data-driven products, most notably “People You May Know”; “Social Graph Visualizations”; “ Matching” and “Collaborative Filtering” that drove the company to success. Mandar Parikh, VP Product & Engineering, Entytle Inc According to Mandar Parikh, VP Product & Engineering, Entytle Inc, “Linkedin is one of the early adopters of data science and is at the forefront of modern day data science. It was one of the first companies to put together a strong data science team. In fact, LinkedIn graph search and recruiter product has got data science behind,” he said. The rise of data-centric products at LinkedIn was overseen by DJ Patil, former US chief data scientist who joined LinkedIn as Chief Scientist and Senior Director of Product Analytics in 2008. Interestingly, Patil alongside Jeff Hammerbacher (founder of Cloudera, also led Facebook team), coined the term ‘data scientist’ and began hiring under this title. And some of the world’s most successful data scientists have come out of LinkedIn. Rewind: Products that revolutionized social networking and e-commerce PYMK became the new way to connect:  According to Patil, who wrote an earlier post, data products are at the heart of social networks, in other words “social network is huge datasets of users, with connections to each other, forming a graph”. LinkedIn’s invention – PYMK, went on to become a critical part of Facebook, Twitter, Google+ who have all reportedly trademarked their friend-suggestion algorithm. PYMK feature is based works on a recommendation engine and makes use of the clustering and classification algorithm to find out people we interact with the most. It also makes use of location and common friend’s data to dish out PYMK suggestions. Amazon did something similar to e-commerce by using item-based collaborative filtering for “People Who Viewed this Item” feature wherein purchase logs were converted into TSV files with customer and item id, detailing whether the product was viewed or bought. Netflix’s widely popular recommendation engine drives online engagement: According to Netflix blog, the movie experience is driven by various algorithms which are part of the Netflix recommender system, it’s most valued asset. Some of their most popular algorithms are ‘Instant Search’ and ‘Page generation’ go a long way in personalization and it all starts with the homepage. So how does Netflix personalize the homepage algorithmically: through a rules-based approach. Netflix blog cites using a set of rules to define template that dictates for all members what types of rows can go in certain positions on the page. This template is improved through A\/B testing to further understand where to place the rows for all members. New age companies Uber, Salesforce and Airbnb reinventing business through data science enabled products California startup Airbnb treats data as the voice of customer Data science is at the heart of Uber’s philosophy and ‘Surge Pricing’, ‘Fare Estimates’, ‘Driver Positioning’ and Matching are some of the most popular data science products from their stable. Parikh cites Uber use case:  according to Parikh, Uber’s success is driven by data centric products such as showing up surge pricing, ETA, heat maps and most importantly driver positioning. Driver positioning – how do the driver know where to wait for customers to maximize their ride revenue is driven by data science algorithms in the backend. Another use case is the ‘Matching algorithm’ that uses automatic matching, in this case Supplier Pick Model wherein based on the request, the nearest cab is made available. Airbnb’s highly publicized matching algorithm to get host preferences right:  The startup that turned the idea of hospitality on its head is also known for being very data-centric. Interestingly, news suggest that data science propelled it’s the startup’s valuation to $25.5 billion. However what made this startup popular was its matching algorithm that allowed interaction between hosts and guests. The model has been built on an estimated conditional probability of booking in a particular location, given the person searched. The California startup detailed in an older post “personalized search results to promote results that would fit the unique preferences of the searcher — the guest”. Location relevance signal in their search built completely with data from the users’ behavior allows future guests locations where they can have great experiences, and the same model has been applied uniformly across the world enabling hosts to open up their homes for stay. Image: Salesforce Salesforce transitioned from contact manager to data centric company: When it first started, Salesforce was just a contact manager, shared Parikh. “It was just a data entry system and there was absolutely no data science in there. Over the last years they have built up a data science team and they brought data science into the product itself with Salesforce Einstein,” he said.  The CRM company is now helping sales persons across the world in closing deals faster with predictive scoring, courtesy Einstein, the AI assistant. How to build data science enabled products So what’s data science, quizzes Parikh. At the very core of it, what data science does is build models that work on large datasets, from thereon one makes predictions. But there is a secret to great data science – art of data science is to figure out which feature to use when.  “If you look at datasets, it is rows of data stored in the table, every column is called a feature and the model that we build needs to shortlist the features. Based on the features, one makes predictions and shortlisting of features is called feature selection,” shared Parikh. Citing a use case of feature selection, Parikh explained: Say for example you have a dataset about customers and you want to what product customers are most likely to buy. So you have to figure out which features are important in making those decisions. We might decide that age of customer is the feature that we would include in our analysis, colour of hair is a feature to be included but by some reason the zip code they are residing is not a feature to be included this process. This is what we call feature selection and once we have our features we build different types of models that fall into a couple of different types of categories. Features that best define data science products — Adaptive and self-learning: According to Sean McClure, Director of Data Science at Space-Time Insight, the next generation of products essentially require data science and product development to be at its core. What data science does is it goes beyond just trend spotting and finds way to automate the learning that is required to connect an organization’s data to their decisions.  And in data science, machine learning is crucial to building great products. Parikh outlines two types of Machine Learning techniques for building models Unsupervised Learning: Unsupervised learning is a set of algorithms that figures out what patterns exist in that data. “Essentially, when one doesn’t know what types of patterns exists but we can figure out where to look for those patterns through this technique,” he said. Supervised learning: in this approach, one can use existing patterns in data to make predictions. Key takeaways for product managers suggested by Parikh Firstly, Parikh sets the record straight on Data Science, Machine Learning and AI which are used interchangeably. “Data Science models extracts patterns and make predictions. Machine Learning automatically calibrates the models and improves the predictions over time by taking results and feeding that back in model, thereby automatically predicting and improving them over time,” he said. When it comes to building great products, nothing is more important than a) business metrics. “This is an important point to bear in mind for product managers and when you are building a product, one must focus solving the customer’s pain points,” he said. In designing products with data science – b) product management fundamentals don’t change. “At the core of it, product management stays same, be obsessed about your customer and have deep empathy and customer centricity,” he added.  Lastly, c) focus on solving the use case at hand. What is a low hanging fruit for data science? It is often choosing the simplest model where one doesn’t need to overthink. “Don’t let your data scientists tell you otherwise that the models are not ready yet, or let’s refine it further before we actually get it out,” he notes arguing that’s a mistake which can lead to analysis paralysis. Sometimes, the simplest most basic algorithm can take one very far. And in most situations, big data won’t necessarily be viable. It is a view echoed by McClure who says, “Data science is less about finding the most predictive model and more about discovering ways to make analysis work with people.” In the same vein, Patil in an older post emphasized how quality assurance (QA) of data products requires a completely different approach. What’s crucial in building great products is the ability to adapt and iterate quickly throughout the product life cycle. “To ensure agility, we build small groups to work on specific products, projects, or analyses. Building test datasets is nontrivial, and it is often impossible to test all of the use cases,” he added.","excerpt":"If you want to know how to build great data science-enabled product, you needn’t look beyond LinkedIn, now Microsoft-owned. The Mountain View headquartered professional networking company has over 433 million members across 200 countries, allowing data scientists access to structured datasets that spawned cutting-edge data-driven products, most notably “People You May Know”; “Social Graph Visualizations”; […]","categories":["IT Services"],"tags":[],"author_name":"Richa Bhatia","publish_date":"2017-04-19T10:46:11","publication_year":"2017","word_count":1572,"keywords":["big data","data science","Go","machine learning","AI","data-driven","ML","analytics","GAN","R"],"extracted_tech_keywords":["AI","machine learning","ML","data science","analytics","R","Go","big data","GAN","data-driven"],"url":"https:\/\/analyticsindiamag.com\/it-services\/data-enabled-products-defining-future-data-science\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":61870,"title":"Sundar Pichai Tells Staff: Google To Cut Back On Hiring &#038; Other Investments","content":"Google’s CEO, Sundar Pichai recently sent a memo to his staff stating that the tech giant will slow down its hiring pace for the remainder of 2020, as the company has been dealing with the economic downturn due to the COVID-19 pandemic. According to the memo, Pichai wrote, “We believe now is the time to significantly slow down the pace of hiring, while maintaining momentum in a small number of strategic areas where users and businesses rely on Google for ongoing support, and where our growth is critical to their success.” This announcement has been made because of the effects the company has been witnessing due to the unprecedented COVID-19 crisis across the economy. Pichai wrote in the memo that” Just like the 2008 financial crisis, the entire global economy is hurting, and Google and Alphabet are not immune to the effects of this global pandemic.” He further stated that” We exist in an ecosystem of partnerships and interconnected businesses, many of whom are feeling significant pain.” Pichai wrote that the company had hired 20,000 people last year, and was aiming to have a similar growth this year. However, due to the crisis, Google had to dial back their plans in areas that aren’t currently critical to the company’s success. He also mentioned that people who have been already hired could also see a bit of a delay in starting work. He said the company is reevaluating its pace of other investment plans as well. Due to this COVID-19 pandemic, the company is facing delays in getting its employees and new hirings their essential equipment, such as laptops and security keys, along with challenges in getting its new employees up to speed, trained and productive on their new teams. This deadly COVID-19 pandemic has already caused significant disruption in the economy as well as for the company. The company has directed its employees around the world to work from home, along with cancelling its major unperson events of the year — Cloud Next and Google I\/O.","excerpt":"Google’s CEO, Sundar Pichai recently sent a memo to his staff stating that the tech giant will slow down its hiring pace for the remainder of 2020, as the company has been dealing with the economic downturn due to the COVID-19 pandemic. According to the memo, Pichai wrote, “We believe now is the time to […]","categories":["AI News"],"tags":["Alphabet","Google","Hiring","hiring trends","Sundar Pichai"],"author_name":"Sejuti Das","publish_date":"2020-04-16T13:30:00","publication_year":"2020","word_count":336,"keywords":["Go","hiring trends","AI","programming_languages:R","Hiring","programming_languages:Go","Aim","Google","disruption","Alphabet","R","Sundar Pichai"],"extracted_tech_keywords":["AI","Aim","R","Go","disruption","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/sundar-pichai-tells-staff-google-to-cut-back-on-hiring-other-investments\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10099330,"title":"ChatGPT Isn’t Made For European Union","content":"Data privacy watchdogs from Europe to the United States have been trying to drag OpenAI, the company behind ChatGPT, to court since it was released in November 2022. Yesterday, the company’s lawyers hit back asking a San Francisco federal court to dismiss most of the lawsuits against the company, which allege that AI outputs infringe on copyright. OpenAI is not only struggling to live the American Dream but it seems to be having a hard time trying to woo the European Union. While OpenAI was busy making a case for itself in front of the US court, the company got slapped with an accusation questioning its ability to comply with EU laws. The 17-page document filed with the Polish DPA is put together by Lukasz Olejnik, a security and privacy researcher, at a Warsaw-based law firm, GP Partners. Earlier this year, Italy’s privacy watchdog, the Garante ordered OpenAI to stop processing data locally — directing it to tackle a preliminary set of issues identified regarding lawful basis, information disclosure, user controls and child safety. Eventually, ChatGPT resumed its service fairly quickly after the company tweaked its presentation. While the Italian data processing agreement investigation continues, other slews of investigation of the General Data Protection Regulation (GDPR) have been lodged in the direction of ChatGPT. Interestingly, the GDPR is the world’s strictest data protection regime, and it has been copied widely around the world. Earlier in April, the bloc’s data protection authorities joined hands to collectively approach regulating the fast-developing technology. The European Data Protection Board, the umbrella organisation for data protection authorities, announced plans to set up an EU-wide task force to coordinate investigations and enforcement. Further in May, the EU’s planned legislation was one of the first to legislate on AI, which Altman said was “over-regulating”. But he backtracked after wide-spread coverage of his comments mounted. “We are excited to continue to operate here and of course have no plans to leave,” he later tweeted. Rise of regional GPT While the boss of OpenAI Sam Altman refuses to leave Europe, the local government and technologists are not too keen to stay dependent on Silicon Valley’s novel AI technology. Fed up of regulating American tech goliaths from 8,000 kilometres away, the European AI circle is rooting for local startups to build their technology similar to OpenAI’s. The most recalled names in the AI community include the French startup Mistral, which has managed to raise $100 million without releasing any products. Another famous startup in the market is Aleph Alpha founded by Jonas Andrulis, a former member of the AI team at Apple. Notably, the startup already sells generative AI as a service to 10,000+ paying customers across companies and governments. Many Europeans remain adamant that they need a contender to counteract America’s dominance, and not simply for economic reasons. The local AI community has put forth the point that European companies are likely to be more sensitive towards data privacy and discrimination than their competitors in the US. For instance, Aleph Alpha is making sure European languages are not excluded from AI developments. Sceptical views have surfaced in the industry questioning whether the startup has the potential to compete in the same league as giants like Google and OpenAI. But many are hoping that Aleph Alpha can give tough competition to Silicon Valley in what some believe will be an era-defining technology. While OpenAI is requesting the US courts to let go of its practices, the company remains exposed to regulatory risk in this area across the EU. The Altman-run company could face outreach from DPAs acting on complaints from European individuals. Meanwhile, confirmed violations of the GDPR can result in penalties as high as 4% of global annual turnover. The tussle over data privacy and security issues between the bloc and the US is long overdue. The DPAs’ corrective orders may end up reworking how technologies function if they wish to continue operating inside the bloc. While the pressure is mounting on regulators as well as OpenAI, it remains to be seen what compliance conclusions may emerge once that assessment has been completed in the EU as well as globally.","excerpt":"OpenAI is struggling to live the American Dream and also having a hard time trying to woo the EU","categories":["AI Features"],"tags":[],"author_name":"Tasmia Ansari","publish_date":"2023-08-31T17:51:31","publication_year":"2023","word_count":691,"keywords":["Go","ChatGPT","TPU","OpenAI","AI","AWS","RAG","GPT","generative AI","R"],"extracted_tech_keywords":["AI","generative AI","ChatGPT","OpenAI","RAG","AWS","TPU","R","Go","GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/chatgpt-isnt-made-for-european-union\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10059589,"title":"Explained: Prospective learning in AI","content":"‘Retrospective learning’ refers to the assumption that the future is an extension of the past. An intelligent system may learn to name certain objects if it is shown pictures of them along with their names. And a model that employs retrospective learning will be able to recognise and name more pictures of the same objects. Still, it will not be able to name previously unencountered objects. Prospective learning A paper published earlier this year argued that retrospective learning isn’t a good representation of true intelligence. According to the study–supported by Microsoft Research and DARPA–learning needs to be future-oriented to solve problems in the real world. Accordingly, NI (Natural Intelligence) and AI have to take an unknown future into account. Their internal models have to adapt to naming new objects and using them in a new context. This is called ‘prospective learning.’ Prospective learning is important because many critical problems are novel experiences that come with little information, negligible probability, and high consequences. Unfortunately, such problems precipitate the downfall of AI systems, such as when medical diagnoses systems cannot detect underrepresented diseases in the samples used to train them. Therefore, the challenge with intelligent systems is to distinguish novel experiences, discern the potentially complex ways in which they connect to past experiences, and then act accordingly. Key factors An intelligent system that employs prospective learning will be able to make good decisions in unique situations by using past data and coming up with active solutions via an internal model of the world. The capabilities systems need to possess to work successfully include: Continual learning (CL) AI systems tend to forget previously learned information while acquiring new information (a phenomenon called catastrophic interference). This is harmful because previously learned abilities are expected to be useful in the future. Continual learning involves a model learning continuously and autonomously from a stream of data and adapting itself as new data comes in. In other words, intelligent systems will be able to recollect the aspects of the past that it believes will be useful in the future while also sequentially acquiring new capabilities. Accordingly, continual learning will work if there is both backwards and forward transfer of information. Constraints Constraints, such as built-in priors and inductive biases, shrink the hypothesis space so that the intelligence needs less data and fewer solutions to resolve the current problem (which translates to the generalised resolution of future problems). These constraints are built into the system of AI and traditionally come in the form of statistical constraints and computational constraints. The former restricts the space of hypotheses to improve statistical efficiency, thereby reducing the amount of data needed to reach a particular goal. The latter seeks to improve computational efficiency by limiting the amount of space and\/or time that an intelligent system has to learn and make deductions. Constraints are necessary because intelligence has finite data, space, and time. Arbitrary slow convergence theorem and the no free lunch theorem further highlight the fact hope for a general AI with the ability to solve all problems efficiently is misplaced. Curiosity This instigates an intelligent system to take actions that the AI aims to use in the future rather than in the present. This sort of objective-driven decision-making can be divided into two parts: (1) a goal aimed at maximising rewards; (2) a goal to maximise relevant information. Thus, the AI has to choose between learning about the world and practising behaviour that will be rewarded. Causality Causal estimation allows the intelligent system to learn the structure of relations, which can help it choose actions to determine specific outcomes. In other words, it enables the AI to identify how one event causes results in another (the effect). Today, ML has the ammo required for retrospective learning, including statistics, algorithms, and mathematics. But, on the other hand, for ML to successfully master prospective learning, we need to employ prospective learning ourselves to imagine potential futures we haven’t experienced. This will require a much more expansive group of people working on the problem– and perspectives from disciplines such as biology, ecology, and philosophy.","excerpt":"A paper published earlier this year argued that retrospective learning isn’t a good representation of true intelligence.","categories":["AI Features"],"tags":["AI (Artificial Intelligence)","Deep Learning","Machine Learning"],"author_name":"Srishti Mukherjee","publish_date":"2022-02-02T12:00:00","publication_year":"2022","word_count":680,"keywords":["Go","programming_languages:R","AI","RPA","ML","Machine Learning","programming_languages:Go","Aim","continual learning","Deep Learning","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","ML","Aim","R","Go","continual learning","RPA","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/explained-prospective-learning-in-ai\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10063371,"title":"Masterclass on AI innovation with oneAPI by Intel","content":"Intel®, in collaboration with Analytics India Magazine, announced a masterclass on AI innovation with oneAPI on April 13, 2022, from 3:00 PM to 4:00 PM. oneAPI is a solution to deliver a unified programming model to simplify development across the diverse architecture. It includes a unified and simplified language and libraries for expressing parallelism and delivers uncompromising native high-level language performance across various hardware, including CPUs, GPUs, FPGAs. oneAPI has been revolutionary and has been a strong platform for various application development in different domains. But, most importantly, the application possibilities are plenty in AI. This session is curated to help developers with basic knowledge of C or C++ and GPU programming. In addition, the session aims at providing the listeners with a clear idea about what could be the possibilities with oneAPI for application development. Register now The masterclass will cover In this session, oneAPI certified instructor will unpack the details, highlighting some of the products built with oneAPI across industries like healthcare, logistics, etc. Why oneAPIWhat’s there in the storeWhat are the application possibilitiesDeployment tips Agenda Welcome & oneAPI Introduction3:00 PM to 3:15 PMIntel® oneAPI AI Analytics Toolkits, AI Applications with oneAPI & Demo3:15 PM to 3:45 PMQ&A on Discord3:45 PM to 4:00 PM Who should attend Software DevelopersIT managers ML developers AI engineers Data science professionals IT, technology, and software architects Senior managers of technology\/engineering\/software Speaker details Shriram Vasudevan, oneAPI certified instructor Vasudevan has over 15 years of experience with a blend of industrial, R&D, and teaching experience. He is an Intel Software Innovator and oneAPI Certified Instructor. Register now","excerpt":"This session is curated to help developers with basic knowledge of C or C++ and GPU programming.","categories":["Deep Tech"],"tags":["Intel","oneAPI"],"author_name":"Amit Naik","publish_date":"2022-03-24T10:00:00","publication_year":"2022","word_count":263,"keywords":["data science","API","programming_languages:R","AI","innovation","ML","Aim","C++","analytics","R","oneAPI","Intel"],"extracted_tech_keywords":["AI","ML","data science","analytics","Aim","R","C++","API","innovation","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/masterclass-on-ai-innovation-with-oneapi-by-intel\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10067848,"title":"Widely used Python and PHP libraries compromised","content":"In a software supply chain attack, PyPI module ctx has been compromised. By this attack, the safer version of this module (which is downloaded 20,000 times a week) was replaced with code that exfiltrates the developer’s environment variables to collect secret codes like Amazon AWS keys. ‘Ctx’ is a minimal Python module that lets developers manipulate dictionary ‘dict’ objects in a number of ways. Interestingly, after remaining untouched for 8 years, newer versions started emerging on May 15 and contained malicious code. PyPI took down the malicious ctx versions soon, however, reports still indicate the presence of malicious code within all ctx versions. https:\/\/twitter.com\/s0md3v\/status\/1529005758540808192?s=20&t=1gJjDoAjbyT79aSeMFTuew Further, the attacker versions of PHPass fork, which are published to the PHP\/Composer package repository Packagist, were altered to steal confidential keys and credentials. PHPass is an open-source password hashing framework that can be used in PHP applications. Released in 2005, the framework has been downloaded over 2.5 million times on Packagist. As per reports, malicious commits were made to PHPass project this week to steal environment variables. The presence of identical logic and Heroku endpoints within the PyPI and PHP packages indicate a common threat actor being responsible for both of these hijacks. While there are few suspicions around the identity of the attackers, experts are not ruling out the possibility of a PoC exercise gone wrong.","excerpt":"PyPI took down the malicious ctx versions soon, however, reports still indicate the presence of malicious code within all ctx versions.","categories":["AI News"],"tags":["Cyber Attack","PHP","Python"],"author_name":"Shraddha Goled","publish_date":"2022-05-25T20:50:39","publication_year":"2022","word_count":222,"keywords":["Go","AWS","AI","cloud_platforms:AWS","programming_languages:R","programming_languages:Go","Python","Cyber Attack","programming_languages:Python","R","PHP"],"extracted_tech_keywords":["AI","AWS","Python","R","Go","cloud_platforms:AWS","programming_languages:Python","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/widely-used-python-and-php-libraries-compromised\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10060062,"title":"Copilot vs AlphaCode: The race for coding supremacy","content":"A study conducted by Cambridge University reveals that a majority of a developer’s time is spent on debugging. This time-consuming task costs the software industry around 300 billion USD every year. Deepmind’s latest artificial intelligence-based code development and analysis tool boasts of reducing such costs by automating the routine and time-consuming tasks of the developers. In contrast with GitHub Copilot, which suggests code, AlphaCode is capable of analysing the algorithm and generating complex programmes at a competitive level that is not only devoid of error but also corresponding to its description. The developers at DeepMind tested the potential of AlphaCode by testing it in competitive programming websites where human developers are given programming problems and ranked based on their results. AlphaCode – The autonomous programmer AlphaCode is a transformer-based language model that consists of 41.4 billion parameters. It is a language model that is four times the size of GitHub Copilot’s language model Codex that parses 12 billion parameters only. The architecture of AlphaCode is based on three parts: Data – The AI tool is fed data by public GitHub repositories. Learning – The tool then trains on the datasets and calibrates them to the requirements of the task (for e.g., competitive programming at Codeforces).Sampling and evaluation – Here, the AI tool performs large scale sampling of variations of programs for each problem. Then through the process of filter and cluster, the programs are ranked into a small subset of 10 solutions that is submitted for external assessment. Figure: Flowchart of the working of AlphaCode Source: deepmind.com AlphaCode’s AI system is pre-trained in a variety of programming languages that are C++, C#, Go, Java, JavaScript, Lua, PHP, TypeScript, Ruby, Scala, Rust and Python. This dataset consists of approximately 715GB of codes along with their descriptions. AlphaCode put to the test The AI tool was entered into a competitive coding competition on Codeforces, a popular platform for hosting coding competitions. The platform shares problems weekly and ranks the participants with the help of an algorithm written in similar lines with the Elo rating system used to rank chess players. A selection of 10 varied test problems from different stages of development was given to AlphaCode. The AI tool achieved an estimated rank within the top 54 percentile of participants that attended the contest, thus proving that AlphaCode’s code generation system has achieved results at a competitive level. AlphaCode’s capability of generating code is demonstrated in an example of one of Codeforces’ problems given below: Figure: The problem presented to AlphaCode, is to figure out the possibility of converting one phrase to another by pressing the backspace instead of writing. Figure: The solution generated by AlphaCode after reading the logic of the problem and producing code that meets the expectations. Mike Mirzayanov, the founder of Codeforces, expresses his surprise as he states: “I was sceptical because even in simple competitive problems, it is often required not only to implement the algorithm but also (and this is the most difficult part) to invent it. AlphaCode managed to perform at the level of a promising new competitor.” Mike further added that “I can safely say the results of AlphaCode exceeded my expectations.” Can Copilot keep up? OpenAI’s AI code suggestion tool GitHub Copilot runs on the natural language processing (NLP) model Codex, a boosted version of GPT-3. While it is built with the vision to achieve goals that are similar to that of AlphaCode, Copilot seems to have a difficult road ahead. Here are some of the differences between the two code generation tools. Training – GitHub Copilot’s AI Codex is trained to identify 12 billion parameters as compared to AlphaCode’s AI-based code generation model that is trained with 40 billion parameters. This improves AlphaCode’s performance four-fold.Suggestion vs Generation: While GitHub Copilot is built to assist programmers in writing the rudimentary sections of code, AlphaCode is capable of generating complete complex programs.Complexity – While both AI tools are in the beginning stages of development, GitHub Copilot suggests basic code involving simple logic, whereas AlphaCode is tested to produce complex algorithms at a competitive level.","excerpt":"Deepmind’s AlphaCode made headlines by testing in the top 54% of human coders. Can GitHub’s Copilot keep up with AlphaCode’s automated programming?","categories":["Deep Tech"],"tags":["AI (Artificial Intelligence)","Natural Language Processing","source code","VS Code"],"author_name":"Kartik Wali","publish_date":"2022-02-08T18:00:00","publication_year":"2022","word_count":680,"keywords":["Go","artificial intelligence","Rust","OpenAI","AI","Natural Language Processing","TypeScript","NLP","Python","VS Code","JavaScript","source code","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","NLP","OpenAI","Python","R","JavaScript","TypeScript","Go","Rust"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/copilot-vs-alphacode-the-race-for-coding-supremacy\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":3874,"title":"Analytics Courses in India &#8211; A Comparison","content":"Business Analytics job market is in a boom in India and there is a huge shortage of skilled analyst in the market. This makes it a very attractive career option with good salary and growth. Many premier institutes have recognized the importance analytics in India’s job market and have started executive programs for working executive in the field of Business Analytics. With a mushrooming of analytics courses, there obviously is an interest from students and working professionals. Yet, with a deluge of so many courses, choosing the right course is difficult. A right analytics course would provide value addition to an analytics professional helping him\/her get faster promotions and growth, as well as valued in the industry. The course should provide an opportunity to develop the right mix of technical and management skills. We at Analytics India Magazine receive abundant queries from prospective students regarding the analytics courses in India. Most of these queries are related to strengths and shortcomings of various courses on offer. It is this abundance of inquiry that motivated us to write this report on analytics courses in India. The main focus is to compare some premium courses and discuss their strengths and weaknesses. We compared the following 5 courses on myriad of attributes- Executive Program in Business Analytics – IIM Calcutta Certificate Programme in Business Analytics – ISB, Hyderabad Certificate Programme in Business Analytics for Executives – IIM Lucknow Business Analytics and Intelligence (BAI) – IIM Bangalore MS in Business Analytics – University of Cincinnati The first four courses are 1-year executive courses in analytics from premier business schools in India. The fifth course is a recognized international course which is included in order to provide a comparison between Indian and international courses. We would not discuss the brand image of these courses since all of these are from esteemed institutes and its upto the students to ascertain the brand value of these institutes subjectively. We rather focus attributes like course content, contact hours, faculty, analytics tools and other finer aspects. Course Comparison The fees of the four domestic courses range between 2.5 Lakhs to 5 lakhs rupees. The fee of the Executive Program in Business Analytics by IIM Calcutta is the lowest (2.5 Lakhs) while the Certificate Programme in Business Analytics by ISB, Hyderabad is the highest at 5 lakhs. Certificate Programme in Business Analytics by IIM Lucknow and Business Analytics and Intelligence (BAI) by IIM Bangalore each is of 4 lakhs. The differences in the fees can an important factor while accessing each course, yet one have to look at the cost vis-à-vis other factors. The important factor to look here is the number of contact hours, mode of teaching and faculty. Looking at these factors can help justify a higher fee. The contact hours of ISB is the highest among all courses here – at 400+ hours. The contact hours of other courses do not exceed 250-300 hours. Of these contact hours, its also important to look at the number of hours on campus versus online teaching. So on this criterion, ISB Hyderabad’s course have a definite edge, in terms of covering a larger depth and breadth of analytics in this 1 year. The IIM Calcutta course is solely an online course with no classroom hours on campus. On the other extreme, for IIM Bangalore’s course the participants have the option to take all courses at the campus or few online. For IIM lucknow course, around 30% of the contact classes are at campus. For ISB course, the number of contact hours on campus is almost 65% of total contact hours. Thus, on this criterion IIM Bangalore and ISB Hyderabad have an edge over other courses. According to ISB Hyderbad, “The program is a combination of classroom and Technology aided learning platform. Participants will typically be on campus for a 5 day schedule of classroom learning every alternate month for a span of 12 months, which would ideally be planned to include a weekend. In the month of no classroom connect, the classes will be conducted over a technology aided learning platform. The contact hours in this platform would be 24 hours a month and every alternate month.” The MS in Business Analytics from University of Cincinnati is available either full-time or part-time, there is no online training for the same. Most international courses follow the same format. Also, the fee comes out to be around $13,000\/ semester which converts to Rs. 20 Lakhs approx. for 1year full time course. Other finer features include, collaboration of foreign universities and international faculty for the course. In case of IIM Lucknow, the program is jointly taught by faculty from IIM Lucknow and the Kelley School of Business, Indiana University, USA. ISB, Hyderabad is well known to employ foreign PhD’s in all their courses including CBA. An integral part of the CBA from ISB are interactions with industry speakers and regular visits to relevant organisations throughout the course. These sessions will be part of almost each module, and will be designed to enhance the understanding of the concepts of the module from a practitioner’s perspective, making it easier to apply the learning immediately. Besides, a SAS module is included in all courses except IIM Calcutta. This is the foundation course to learn the most important tool utilized in Analytics. ISB Hyderabad also unofficially provides career services to the graduating batch. The same goes for University of Cincinnati. Course Curriculum The table below compares the curriculum of the 5 courses. Clearly, ISB Hyderabad’s CBA has the most comprehensive course covering the most important topics in depth. It is followed by the CPBAE of IIM Lucknow. IIM Calcutta – EPBA ISB Hyderabad – CBA IIM Lucknow – CPBAE IIM Bangalore – BAI University of Cincinnati – MSBA Business Data Mining ¤ ¤ ¤ ¤ Business Intelligence ¤ Contemporary Analytics ¤ ¤ Data Collection ¤ Data Management ¤ Data Visualisation ¤ ¤ ¤ Financial Analytics ¤ ¤ Forecasting Analytics ¤ ¤ ¤ ¤ Marketing Analytics ¤ ¤ ¤ Mathematics for Business Analytics ¤ Operations and Supply Chain Analytics ¤ ¤ ¤ Operations\/Simulation ¤ ¤ ¤ Optimisation ¤ ¤ ¤ ¤ Pricing Analytics ¤ ¤ Spreadsheet Modeling ¤ Statistics for Business Analytics ¤ ¤ ¤ ¤ ¤ Stochastic Models ¤ Time Series Models in Business ¤ ¤ ¤ ¤ ¤ So, in terms of purely course content, ISB Hyderabad’s course can easily be seen in par with some of the best analytics courses offered worldwide. One point of caution regarding international courses – while some courses in US are in depth and highly useful, most international courses design their pedagogy to suit the non-mathematical backgrounds of the student. They tend to introduce concept from the very scratch and then go in-depth to more advanced topics. This is not valid for Indian courses. [attachments docid=”3879″]","excerpt":"Business Analytics job market is in a boom in India and there is a huge shortage of skilled analyst in the market. This makes it a very attractive career option with good salary and growth. Many premier institutes have recognized the importance analytics in India’s job market and have started executive programs for working executive […]","categories":["AI Trends"],"tags":["analytics education","analytics training","data scientist india salary","great lakes analytics","mba in analytics"],"author_name":"Дарья","publish_date":"2013-07-09T16:16:54","publication_year":"2013","word_count":1135,"keywords":["business intelligence","Go","programming_languages:R","AI","data scientist india salary","programming_languages:Go","great lakes analytics","analytics","GAN","analytics education","R","mba in analytics","analytics training"],"extracted_tech_keywords":["AI","analytics","R","Go","GAN","business intelligence","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/analytics-courses-in-india-a-comparison\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10040704,"title":"Never Design For A Million Users In Your First MVP: Manisha Raisinghan, LogiNext","content":"Manisha Raisinghani and Dhruvil Sanghvi founded in LogiNext 2014. The Mumbai-based startup now has an office in the US. On the second day of Analytics India Magazine’s women in technology event, The Rising 2021, Manisha and LogiNext Product Marketing Head Jubin Mehta discussed her journey as a technology leader. Manisha and Dhruvil are both graduates of Masters in Management Information Systems and Services from Carnegie Mellon University. Post the studies, Manisha joined IBM as a Senior Consultant, while Dhruvil worked with Deloitte as the Data Analytics Consulting. “We often wondered how the leading consulting companies that we worked at, were spending millions of dollars building something which could only be used by them,” Manisha recalled. The duo had experience in logistics and big data analytics. In 2014, both Manisha and Dhruvil quit their cushy jobs in the US to move back to India, with the hope of starting something new. Manisha said they spent ample time researching the logistics space. “We spent two months just finalising the company name and logo. It was absolutely not required,” Manisha said. Today, LogiNext operates in more than 50 countries, and services upto 150-plus enterprise clients, including the likes of Myntra, McDonald’s, and Decathlon, among others. LogiNext is backed by marquee investors including Tiger Global Management, Steadview Capital, Paytm and Alibaba. Tech journey In 2014, Manisha wrote the first line of code for LogiNext. She said, “Never design for a million users in your first MVP. Companies should design for scale, but when starting to write the code, never design for millions.” LogiNext broadly focuses on four major aspects of logistics: automation and optimisation, business intelligence and analytics, customer experience, and dynamic auto assignment. The firm helps companies or customers automate and optimise end-to-end logistics in a seamless manner. When starting out, LogiNext used a plain MySQL database to capture location points. However, as the businesses expanded and acquired more  customers, billions of data points had to be added. “We realised that a relational database was not going to help us scale to the level at which we were anticipating,” Manisha said. Thus, it moved from MySQL architecture to MongoDB. At the heart of LogiNext is the route optimisation software which uses Heuristic algorithm at its core. For example, Manisha said, when McDonald’s receives an order, it is immediately pushed into the LogiNext system. It then uses algorithms to find the best rider who can serve the order. “We say best and not nearest because it totally depends on the clients’ configuration,” she added. Hiring process Manisha said the very first hire left the company on the first day itself. However, the second hire continues to be associated with the SaaS company. For every 500 resumes that are screened, LogiNext hires only two people. “A lot of people think that learning too many technologies is a good thing. While it is good to know about more technologies so that you can be sure of the choice that you are making, it is critical to understand that it is not about knowing technology but excelling at it. There will always be new technologies in the market, keep yourself updated but excel at only a couple of them. If you want to excel at something, you need 10,000 hours of training,” she said. Manisha feels the most important thing to remember is listening to customers’ wants. It is a common mistake made by AI developers. “They do what they want to do. But developers need to understand the problems that customers face on-ground. The problem you are solving should be customer-led and not lab-based,” Manisha added. Manisha said two things have always worked for her while hiring for her startup: When hiring leaders, hire people who are smarter than you. “I am not good at DevOps but I have hired someone who is great at it,” she said. Trust your employees and let them do their jobs. “They will make mistakes, but so will you. Have the appetite for them to make mistakes,” Manisha said.","excerpt":"Manisha Raisinghani and Dhruvil Sanghvi founded in LogiNext 2014. The Mumbai-based startup now has an office in the US.  On the second day of Analytics India Magazine’s women in technology event, The Rising 2021, Manisha and LogiNext Product Marketing Head Jubin Mehta discussed her journey as a technology leader. Manisha and Dhruvil are both graduates […]","categories":["AI Features"],"tags":["Interviews and Discussions"],"author_name":"Debolina Biswas","publish_date":"2021-05-25T12:00:00","publication_year":"2021","word_count":669,"keywords":["Go","API","AI","MongoDB","ML","analytics","SQL","Rust","DevOps","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","ML","analytics","MongoDB","R","SQL","Go","Rust","DevOps","API"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/never-design-for-a-million-users-in-your-first-mvp-manisha-raisinghan-loginext\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10095217,"title":"Praxis Data Science Program: Transforming Careers Through Industry-Relevant Education and Remarkable Placement Success","content":"Praxis Business School has emerged as a pioneer in data science education, attracting students from diverse backgrounds seeking quality education and excellent placement opportunities. The PGP Data Science program at Praxis offers a transformative learning experience, preparing students to excel in the dynamic field of data science. After our placement drive, we sat with students from the business school to discuss their reasons for choosing Praxis, the unique learnings from the Praxis Programme and their viewpoints on what drives companies to recruit from Praxis. Q1. What Made You Choose the PGP Data Science Program at Praxis? “Having initially enrolled in an online data science course, I was dissatisfied with the quality of education and the remote learning experience. Seeking a more immersive and top-notch learning environment, I embarked on a quest to find the best institute in India for data science. That’s when Praxis caught my attention because of its reputation for excellence in data science education and an exceptional on-campus placement record,” said Aji Thomas, a student from Praxis. He added that after conducting extensive research and hearing positive reviews from alumni, I confidently chose to join the PGP-DS program at Praxis. Another student Andrew George Cherian said that he became acquainted with Praxis and its unrivalled course structure and distinguished faculty through a college friend of mine who graduated from Praxis in 2019. “As someone with six years of experience in the engineering sector, the prospect of undergoing a uniquely crafted program that could reshape my career within a mere nine months was immensely appealing. Praxis’s esteemed position in the data science realm, coupled with its ability to deliver on its promises, solidified my belief that this institution was the ideal choice for me,” he added. Sanskriti Jain, a Praxis alumni, realises the importance of an offline classroom environment after the pandemic and made an informed decision to join Praxis. “Praxis’s prominent ranking as a pioneer in data science education further validated my choice. The institution’s commitment to offering an exceptional education and fostering a supportive learning environment comprising both faculty and peers cemented my belief that Praxis was the perfect fit for my data science journey. Praxis undoubtedly is the premier destination for Data Science Education in India,” Jain added. Q2. We understand that Praxis has an incredible day-zero with 20+ recruiters, something that is normally witnessed at top institutes like IIMs\/ IITs. Can you give us a glimpse of what happened that day? “The campus buzzes with excitement in the week leading up to the much-anticipated Day Zero placement drive at Praxis. But it is on Day Zero itself when the real magic unfolds. There’s an influx of renowned companies, including industry giants like PwC, EY, Wells Fargo, ITC, Hero Fincorp, Kantar, Subex, and many others, at Praxis to seek out exceptional talents,” said Thomas. He presented the entire scene of Day Zero. The air is filled with anticipation as the shortlisted students step into the final interview rounds. And then, the offers start pouring in. As the clock strikes 5:10 AM, the first offer of the day arrives, setting the tone for an unforgettable experience. By noon, an astonishing 45% of the batch has already secured placements. The numbers soar as the day progresses, with students receiving remarkable offers, including an impressive Rs. 17.6 LPA to nine students and Rs. 16 LPA to thirteen students. Cherian took us through the preparations that happen behind the curtain. “I would like to emphasize the hard work that goes into making this a monumental day. The weeks leading up to Day Zero are filled with meticulous preparations, as students engage in pre-placement sessions, written tests, mock GDs and interviews conducted by faculty and alumni, to gear up for the ultimate test of their abilities.” He added that it is all worth the effort, as is evidenced by the enormous success Praxis students achieve, batch after batch. “Close your eyes and imagine the scene— an atmosphere brimming with enthusiasm, and a campus teeming with anticipation. It’s 3:45 am, and students eagerly converge at Praxis for the extraordinary Day Zero placement drive,” said Jain “As the clock strikes 5:00 am, the interview process commences. As time goes by, and offers begin to pour in, the collective achievements of the students are celebrated. By 6 PM, 62+% of the batch has secured prestigious placements,” she continued. Day Zero at Praxis is a spectacle that is a testament to the exceptional education, unwavering support, and pursuit of excellence that defines Praxis Business School. Q3. What makes companies make a beeline to recruit at Praxis? “Praxis stands out due to its industry-oriented program and diverse student community. The curriculum caters to industry demands, providing practical tech skills and exposure to various domains. Students from different backgrounds, ranging from beginners to experienced professionals, offer a wealth of perspectives and ideas. The fair and transparent recruitment process ensures access to the most suitable candidates for companies seeking diverse talent,” said Thomas. Talking about Praxis’ consistent track record of producing high-quality students attracts companies year after year, Cherian said that the institute’s continuous curriculum updates align with industry requirements, ensuring students are well-prepared. “The exceptional performance of Praxis alumni also contributes to the repeated recruitment from leading companies, reaffirming Praxis’ strong industry relevance,” he added. “Praxis’ esteemed alumni actively participate in the recruitment process, setting high standards for current students. Their recognized industry performance and contributions enhance Praxis’ reputation and make it an attractive choice for recruiters. The repeated presence of reputable companies reinforces the institution’s industry relevance,” said Jain. Q4. As a part of the Placement Committee (PlaceCom) can you help us understand what makes the Praxis Program special? Praxis’ curriculum is recognized for its rigorous and relevant content, continuously updated to meet industry needs. The faculty members, who are industry experts, blend theoretical and practical knowledge, ensuring students are job-ready,” said Jain. She added that collaborative projects and assignments foster peer-to-peer learning, creating a supportive learning environment. The capstone project, guided by faculty mentorship, allows students to apply their knowledge and receive evaluations from industry experts, further enhancing their skills. “With strong industry connections and a vast alumni network, Praxis facilitates access to a wide range of job openings and employers. The Student Placement Committee plays a crucial role in ensuring student and employer satisfaction by understanding specific requirements, offering counselling and skill-building sessions, and continuously improving based on recruiters’ feedback,” said Thomas. “As a member of the student Placement Committee, I have witnessed the institute’s unwavering support and guidance for students throughout the program,” said Cherian. He further said that Praxis not only imparts knowledge but also embraces empathy, understanding the aspirations and challenges of each student. The institute’s pure intent to nurture and uplift its students is palpable in every interaction. Multiple opportunities are provided to convert job applications into offers, and comprehensive assistance is given for interview preparation. “Praxis Business School’s PGP Data Science program offers a compelling educational experience and exceptional placement opportunities. With its comprehensive support and transformative learning environment, the Praxis Program sets students on a path to success in the field of data science,” concluded Cherian.","excerpt":"Praxis Business School has emerged as a pioneer in data science education, attracting students from diverse backgrounds seeking quality education and excellent placement opportunities. The PGP Data Science program at Praxis offers a transformative learning experience, preparing students to excel in the dynamic field of data science.  After our placement drive, we sat with students […]","categories":["AI Features"],"tags":["AI career","praxis"],"author_name":"AIM Media House","publish_date":"2023-06-15T17:30:00","publication_year":"2023","word_count":1189,"keywords":["data science","Go","AI career","programming_languages:R","AI","programming_languages:Go","praxis","R"],"extracted_tech_keywords":["AI","data science","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/praxis-data-science-program-transforming-careers-through-industry-relevant-education-and-remarkable-placement-success\/","complexity_score":3,"technical_depth":6,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10040427,"title":"Credit Suisse To Hire 1,000 IT Employees In India This Year","content":"Credit Suisse has plans to hire over 1,000 IT professionals in India this year. John Burns, Head India IT and Senior Franchise Officer, Pune, said: “This year’s hiring plan highlights our continued commitment to India, particularly to Maharashtra, and supports Credit Suisse’s vision to establish our operations here as a global technological hub. To support the growth of our IT presence in India, we believe empowering our employees to lead global delivery and drive innovative solutions enhances value creation and productivity for the bank globally.” Credit Suisse has hired over 2,000 IT employees in the last three years in India. It aims to tap into the large pool of tech talent available in India, to further enhance its in-house capabilities. India presently accounts for almost 25 percent of the bank’s global IT staff, and has the largest footprint globally. Reportedly, the new hires will comprise developers and engineers with capabilities and expertise in emerging technologies such as cybersecurity, data analytics, cloud, API development, machine learning and AI to fuel the bank’s digital aspirations. The bank usually hires from premier engineering and management colleges in India, with a focus on entry-level talent with up to one-year of experience. Prashant Bhatnagar, Global Head of Experienced Recruiting for Technology, said: “As we continue to build our footprint in India, we want to attract the best IT talent to join our vibrant community of professionals.”","excerpt":"Credit Suisse has plans to hire over 1,000 IT professionals in India this year. John Burns, Head India IT and Senior Franchise Officer, Pune, said: “This year’s hiring plan highlights our continued commitment to India, particularly to Maharashtra, and supports Credit Suisse’s vision to establish our operations here as a global technological hub. To support […]","categories":["AI News"],"tags":[],"author_name":"Debolina Biswas","publish_date":"2021-05-19T12:57:54","publication_year":"2021","word_count":231,"keywords":["API","machine learning","programming_languages:R","AI","Git","Aim","ViT","analytics","R"],"extracted_tech_keywords":["AI","machine learning","analytics","Aim","R","Git","API","ViT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/credit-suisse-to-hire-1000-it-employees-in-india-this-year\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10143430,"title":"Apple Teams up with Broadcom for AI Chips, To Sideline NVIDIA","content":"Apple is collaborating with Broadcom to develop its first AI-focused server chip, a move aimed at preparing to address the intense computing demands of its AI features, according to a report by The Information. The chip, internally code-named “Baltra”, is projected to enter mass production by 2026. The initiative mirrors strategies employed by other tech giants like Google, Amazon, and Meta, which have designed custom chips to support AI services. The collaboration highlights Apple’s strategic effort to reduce dependence on NVIDIA, which dominates the AI chip market. The chip will reportedly use Broadcom’s networking technology to optimise AI processing and employ advanced manufacturing processes from TSMC. What’s in for Apple? Broadcom shares rose by five percent following the announcement, reflecting the company’s growing role in AI chip manufacturing. TSMC’s advanced N3P fabrication process is reported to be used for Baltra. Apple and Broadcom declined to comment on the report. Apple’s in-house silicon team in Israel is leading the project, focusing on a chiplet design that improves efficiency and reduces complexity. The design incorporates Apple’s Neural Engine technology to enhance AI task performance. This initiative aligns with Apple’s recent AI advancements, including features that generate text, create images, and summarise content. While Apple’s current chips power some AI features, they are not as efficient as chips designed specifically for AI processing. Broadcom, which has also partnered with Google on AI chips, will provide networking technology but allow Apple to retain control over the chip’s overall design and production. Growth by Partnerships This partnership also builds on an existing multi-billion-dollar deal between Apple and Broadcom to produce 5G radio frequency components, announced last year. Apple previously shared plans to integrate its own server chips to power AI features on its devices. Apple has achieved notable success with in-house chip development, such as the M-series processors, which replaced Intel chips in Mac computers. Despite such advancements, NVIDIA remains a dominant player in the AI chip market, though Google has managed to collaborate with Broadcom to lessen its dependency. In terms of collaboration, even OpenAI, the creator of ChatGPT, has partnered with Broadcom and TSMC to launch its first in-house AI chip in 2026. The generative AI boom has significantly boosted Broadcom’s market presence, with its shares surging 54 percent in 2024. Analysts predict the custom chip market could grow to $45 billion by 2028, with Broadcom and competitor Marvell poised to share the expanding market. Apple plans to use a chiplet design for its AI chip, a concept first introduced by AMD over a decade ago, according to two sources. Instead of creating one large chip with different functional sections, Apple will split the chip into smaller, specialised pieces called chiplets, which will then be reassembled into a single unit. This approach simplifies manufacturing and reduces the chances of defects. However, NVIDIA remains a key player in tasks requiring heavy computing, such as training AI models. Apple’s project is part of a broader strategy to integrate custom chips for secure, efficient AI data processing. As the company rolls out more AI features, the Baltra chip could play a crucial role in scaling services to billions of devices.","excerpt":"The chip, internally code-named “Baltra”, is projected to enter mass production by 2026.","categories":["AI News"],"tags":["AI Chips","Apple"],"author_name":"Sanjana Gupta","publish_date":"2024-12-12T17:29:14","publication_year":"2024","word_count":524,"keywords":["Go","ChatGPT","OpenAI","AI","Apple","AI Chips","GPT","Aim","generative AI","R","llm_models:ChatGPT","llm_models:GPT"],"extracted_tech_keywords":["AI","generative AI","ChatGPT","OpenAI","Aim","R","Go","GPT","llm_models:GPT","llm_models:ChatGPT"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/apple-teams-up-with-broadcom-for-ai-chips-to-sideline-nvidia\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10011220,"title":"Introduction to LSTM Autoencoder Using Keras","content":"Simple Neural Network is feed-forward wherein info information ventures just in one direction.i.e. the information passes from input layers to hidden layers finally to the output layers. Recurrent Neural Network is the advanced type to the traditional Neural Network. It makes use of sequential information. Unlike conventional networks, the output and input layers are dependent on each other. RNNs are called recurrent because they play out a similar undertaking for each component of an arrangement, with the yield being relied upon the past calculations.LSTM or Long Short Term Memory are a type of RNNs that is useful in learning order dependence in sequence prediction problems. In this article, we will cover a simple Long Short Term Memory autoencoder with the help of Keras and python. What is an LSTM autoencoder? LSTM autoencoder is an encoder that makes use of LSTM encoder-decoder architecture to compress data using an encoder and decode it to retain original structure using a decoder. About the dataset The dataset can be downloaded from the following link. It gives the daily closing price of the S&P index. Code Implementation With Keras Import libraries required for this project import numpy as np import pandas as pd import matplotlib as mpl import matplotlib.pyplot as plt Read the data df = pd.read_csv('spx.csv', parse_dates=['date'], index_col='date') Split the data train_size = int(len(df) * 0.9) test_size = len(df) - train_size train, test = df.iloc[0:train_size], df.iloc[train_size:len(df)] train.shape Pre-Processing of Data We need to pre-process the training and test data using the standardscaler library imported from sklearn. from sklearn.preprocessing import StandardScaler scaler = StandardScaler() scaler = scaler.fit(train[['close']]) train['close'] = scaler.transform(train[['close']]) test['close'] = scaler.transform(test[['close']]) Create a sequence with historical data Now we will split the time series data into subsequences and create a sequence of 30 days of historical data. def create_dataset(X, y, time_steps=1): X1, y1 = [], [] for i in range(len(X) - time_steps): t = X.iloc[i:(i + time_steps)].values X1.append(t) y1.append(y.iloc[i + time_steps]) return np.array(X1), np.array(y1) TIME_STEPS = 30 X_train, y_train = create_dataset( train[['close']], train.close, TIME_STEPS ) X_test, y_test = create_dataset( test[['close']], test.close, TIME_STEPS ) print(X_train.shape) Creating an LSTM Autoencoder Network The architecture will produce the same sequence as given as input. It will take the sequence data. The dropout removes inputs to a layer to reduce overfitting. Adding RepeatVector to the layer means it repeats the input n number of times. The TimeDistibuted layer takes the information from the previous layer and creates a vector with a length of the output layers. import keras model = keras.Sequential() model.add(keras.layers.LSTM( units=64, input_shape=(X_train.shape[1], X_train.shape[2]) )) model.add(keras.layers.Dropout(rate=0.2)) model.add(keras.layers.RepeatVector(n=X_train.shape[1])) model.add(keras.layers.LSTM(units=64, return_sequences=True)) model.add(keras.layers.Dropout(rate=0.2)) model.add( keras.layers.TimeDistributed( keras.layers.Dense(units=X_train.shape[2]) ) ) model.compile(loss='mae', optimizer='adam') model.summary() Fitting the Model Here, we train the model with epoch:20 and batch size 32. history = model.fit( X_train, y_train, epochs=20, batch_size=32, validation_split=0.1, shuffle=False ) Evaluation plt.plot(history.history['loss'], label='train') plt.plot(history.history['val_loss'], label='test') plt.legend(); From the above plot we can see the training and test error is decreasing. For better result, we can train the model with more epochs. Actual Value of Test Data y_test Prediction on Test Data pred = model.predict(X_test, verbose=0) Conclusion In this article, we have covered the basics of Long-short Term Memory autoencoder by using Keras library. Comparing the prediction result and the actual value we can tell our model performs decently. Further, we can tune this model by increasing the epochs to get better results.The complete code of the above implementation is available at the AIM’s GitHub repository. Please visit this link to find the notebook of this code.","excerpt":"LSTM autoencoder is an encoder that makes use of LSTM encoder-decoder architecture to compress data using an encoder and decode it to retain original structure using a decoder.","categories":["Deep Tech"],"tags":["encoder","encoder-decoder","lstm recurrent neural network","Neural Networks"],"author_name":"Ankit Das","publish_date":"2020-11-05T12:00:41","publication_year":"2020","word_count":571,"keywords":["encoder","NumPy","TPU","Keras","AI","neural network","encoder-decoder","Python","Ray","Aim","lstm recurrent neural network","Matplotlib","Pandas","Neural Networks"],"extracted_tech_keywords":["AI","neural network","Aim","Ray","Keras","Pandas","NumPy","Matplotlib","TPU","Python"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/introduction-to-lstm-autoencoder-using-keras\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":38351,"title":"How Deep Learning Can Power Safer Fusion Energy Experiments","content":"Photo by Peretz Partensky The almost infinite capabilities of nuclear energy might be luring but establishing a nuclear plant comes with its own set of challenges. The memories of the world war are still etched in concrete and people get paranoid when they hear the word nuclear. The external instabilities like Tsunamis and extremist activities like terrorism cannot be forecasted with certainty.  But what happens within a nuclear plant can be controlled and should be. Even better if they can be predicted. Forecasting is one of the many applications where machine learning techniques have established a firm footing. With the deep learning networks getting better with each passing day, the move to entrust these networks with something as sophisticated and incredibly powerful as nuclear plants. Any instability in plasma can vaporize the components in the fusion experiment. Disruptions like these can leave the facility useless for long periods of time. To predict such disruptions, researchers at Princeton’s  plasma physics laboratory under the leadership of Kates-Harbeck have developed a deep learning method. Plasma Control With FRNN Source: Nature The $25 billion facility ITER(International Thermonuclear Experimental Reactor) currently under construction cannot afford any kind of disruption for this magnetically confined plasma physics experiment. Once the plasma hits the walls there isn’t much to do, usually the researchers opt for a precautionary measure like injecting a neutral gas to cool the plasma with the help of a warning system. A machine learning algorithm only gets better with more data. And there isn’t much disruption data available. So, the idea here is to train the neural networks with those experiments which can afford disruption. To ensure a reliable future for fusion energy, few hits have to be taken initially. AI approach was undertaken with greater focus for the Tokamak reactor, a donut shaped machine that holds hot plasma using a powerful magnetic field. High dimensional data like temperature of electrons as a function of radius in the plasma from previous fusion experiments is fed in to the Fusion Recurrent Neural Network(FRNN). The deep learning networks are supplemented with the NVIDIA’s V100 GPUs as the task is computationally intensive and would require high-performance computing clusters. Around 2TB of data is provided by the department of energy, California and Joint European Torus(JET), UK. The team were able to demonstrate that the FRNN code trained on the smaller DIII-D could be generalized to the larger, much more powerful JET. Future With Fusion Last month a bill was reintroduced in the US Senate to nudge the energy department to establish advanced nuclear objectives and accelerate the research required through University programs. This was welcomed by visionaries like Bill Gates who said that to prevent the worst effects of climate change, we need to reach near-zero emissions on all the things that drive it—agriculture, electricity, manufacturing, transportation, and buildings—by investing in innovation across all sectors while deploying low cost renewables. India has embarked on it nuclear energy since 1948 and ever since the NSG waiver, India has been actively partnering with the US, Canada, Russia, France and Australia for nuclear cooperation and trade. Last month the US has issued a statement that it will help build 6 nuclear power plants in India. Amongst the ratifications and rejections of the protocols and treaties, few nations have somehow found a way to keep pushing the technology for superior, more healthy lifestyle on this planet. Climate change is an overloaded word and is often reduced to petty debates and blame game. There is no one reason which contributes to it and neither is it trivial enough to ignore completely. A developed nation can afford cleaner tech while the transition in developing nations will be delayed. There can’t be an umbrella solution and it needs to be solved locally.","excerpt":"The almost infinite capabilities of nuclear energy might be luring but establishing a nuclear plant comes with its own set of challenges. The memories of the world war are still etched in concrete and people get paranoid when they hear the word nuclear. The external instabilities like Tsunamis and extremist activities like terrorism cannot be […]","categories":["Deep Tech"],"tags":["Deep Learning","nuclear energy"],"author_name":"Ram Sagar","publish_date":"2019-04-27T06:16:56","publication_year":"2019","word_count":626,"keywords":["Go","machine learning","AI","neural network","innovation","deep learning","ViT","RNN","Rust","Deep Learning","nuclear energy","R"],"extracted_tech_keywords":["AI","machine learning","deep learning","neural network","R","Go","Rust","RNN","ViT","innovation"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/how-deep-learning-can-power-safer-fusion-energy-experiments\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10011151,"title":"AI Predicts The Winner Of US Presidential Election From Tweets","content":"An ‘unprecedented’ AI model has now predicted a winner in Tuesday’s US Presidential election. According to the AI model, democrat candidate Joe Biden is currently leading over presidential incumbent Donald Trump in the national polls for the US presidential election. According to the news, Dr Hernan Makse, a statistical physicist at the City University of New York, has now leveraged the advanced AI technology to scan social media platforms in an attempt to predict the winner of this year’s election. Dr Makse, when asked, told the media how he combined the technology of big data along with AI to understand the opinions on Twitter. This is done by developing an AI model, by collecting an enormous number of opinions by people on social media, especially on Twitter with one billion tweets, stated Makse. According to Makse, the research already started in the previous US election in 2016 along with Argentina election last year in 2019; and now finishing-up the predictions for the US 2020 elections. All were done based on 200 million tweets that have been gathered from the previous USA election, with 800 million tweets in total. Considering there is an enormous amount of activity that happens every election, the researchers utilise this information to train their machine learning model. Explaining further, Makse stated to the media that they trained a neural network, using AI in order to predict the opinion of every user on the social media platform. And with that information in hand, the researchers aim to predict the elections with a heavy statistical analysis on the results. However, the method comes with certain limitations, revealed by the researchers. Makse stated that the models are yet not fully-ready to predict and call an election accurately. Explaining further he stated to the media that there are a few variables that one has to take into account — “one is the sampling bias,) which denotes that “even if you take 20 million people across the entire USA, your sampling is always biased – so our bias is the people who actually use Twitter.” Explaining the results, Dr Makse said, for instance, “if we just take the raw data, so whatever people are talking about on Twitter without doing any of the weightings in or any of the scaling, there is a large advantage of Joe Biden to win by a large amount.” However, In instances “there is a large voter turnout of people living in rural areas, the penetration of Twitter is very low.” Further, he said that another bias would be the voter turnout because it’s not possible to predict how many people are going to vote. So controlling these two variables are important for the AI tool to predict accurately. “So at the moment, we are actually trying to calibrate our model, based on the previous election, and including this election,” stated Makse. Although Joe Biden is still continuing to be the favourite for the election, at this moment the AI cannot be ready actually to call the results. “I’m a scientist, and when we put some information out there, it has to be done scientifically with results showing a logistical analysis that we know that we assume, do not take into account all the variables that we would like to take. But, according to the models we have right now, the clear winner will be [Joe] Biden,” concluded Dr Makse.","excerpt":"An ‘unprecedented’ AI model has now predicted a winner in Tuesday’s US Presidential election. According to the AI model, democrat candidate Joe Biden is currently leading over presidential incumbent Donald Trump in the national polls for the US presidential election. According to the news, Dr Hernan Makse, a statistical physicist at the City University of […]","categories":["AI News"],"tags":["limitations of AI"],"author_name":"Sejuti Das","publish_date":"2020-11-03T19:21:32","publication_year":"2020","word_count":564,"keywords":["big data","Go","machine learning","programming_languages:R","AI","neural network","RAG","Aim","ViT","R","limitations of AI"],"extracted_tech_keywords":["AI","machine learning","neural network","Aim","RAG","R","Go","big data","ViT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/ai-predicts-the-winner-of-us-presidential-election-from-tweets\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10053382,"title":"NVIDIA&#8217;s Large Language AI Models Are Now Available To Businesses Worldwide","content":"NVIDIA has set the stage for businesses worldwide to design and deploy large language models (LLMs). This design enables them to develop domain-specific chatbots, personal assistants, and other artificial intelligence systems. The firm announced the NVIDIA NeMo Megatron framework for training trillion-parameter language models. In addition, NVIDIA Triton Inference Server offers multi-node distributed inference features for new domains and languages. When used in conjunction with NVIDIA DGX systems, these technologies provide an enterprise-grade solution for simplifying the construction and deployment of massive language models. “Large language models have demonstrated their flexibility and capability, answering deep domain questions, translating languages, comprehending and summarising documents, writing stories, and computing programmes all without specialised training or supervision,” said Bryan Catanzaro, NVIDIA’s vice president of applied deep learning research. “Developing huge language models for new languages and domains is perhaps the largest supercomputing use to date, and these capabilities are now accessible to the world’s corporations.” Speed LLM Development NVIDIA NeMo Megatron builds on Megatron, an open-source project led by NVIDIA researchers that implements massive transformer language models at scale. Megatron 530B is the most customisable language model in the world. Enterprises can overcome the obstacles associated with developing complex natural language processing models using the NeMo Megatron framework. It is designed to scale out across NVIDIA DGX SuperPOD’s large-scale accelerated computing infrastructure. With data processing libraries that ingest, curate, organise, and clean data, NeMo Megatron automates the complexity of LLM training. Leveraging powerful data, tensor, and pipeline parallelisation technologies enables the training of huge language models to be efficiently distributed across thousands of GPUs. Enterprises can utilise the NeMo Megatron framework to teach LLMs in their topics and languages of interest. Real-Time LLM Inference New multi-GPU, multi-node capabilities in the newest NVIDIA Triton Inference Server enable real-time scaling of LLM inference workloads across several GPUs and nodes. The models demand more memory than a single GPU or even a large server with numerous GPUs can provide, and inference must be performed quickly for applications to be relevant. Megatron 530B may now be run on two NVIDIA DGX systems, reducing processing time from nearly a minute on a CPU server to half a second, enabling the deployment of LLMs for real-time applications. Custom Language Models SiDi, JD Explore Academy, and VinBrain are among the early adopters constructing huge language models using NVIDIA DGX SuperPOD. SiDi, one of Brazil’s leading research and development organisations for artificial intelligence, has modified the Samsung virtual assistant for 200 million Portuguese speakers. “The SiDi team has considerable expertise developing artificial intelligence (AI) virtual assistants and chatbots, which require both high AI performance and specialised software that is trained and tuned to the shifting nuances of human language,” said John Yi, SiDi’s CEO. “NVIDIA DGX SuperPOD is suitable for powering our team’s advanced work and enabling us to provide world-class AI services to Brazilian Portuguese speakers.” JD Explore Academy, the research and development arm of JD.com, a leading supply chain technology and service provider, is utilising NVIDIA DGX SuperPOD to develop natural language processing for use in smart customer service, smart retail, smart logistics, the Internet of Things, and healthcare, among other applications. VinBrain, a healthcare artificial intelligence firm based in Vietnam, used a DGX SuperPOD to develop and deploy a clinical language model for radiologists and telemedicine in 100 hospitals. It is currently used by over 600 healthcare practitioners. Availability NVIDIA Triton is available through the NVIDIA NGC catalogue, a repository for GPU-accelerated AI software that includes frameworks, toolkits, pretrained models, and Jupyter Notebooks, as well as through the Triton GitHub repository. Additionally, Triton is a component of NVIDIA’s AI Enterprise software stack, which NVIDIA optimises, certifies, and supports. As a result, enterprises can utilise the software suite to execute language model inference on commercially available accelerated servers in on-premises data centres and private clouds.","excerpt":"Nvidia doubles down on artificial intelligence language models and inference as a platform for the Metaverse, in data centres, the cloud, and at the edge.","categories":["Global Tech"],"tags":["LLMs","NVIDIA"],"author_name":"Dr. Nivash Jeevanandam","publish_date":"2021-11-13T13:00:00","publication_year":"2021","word_count":636,"keywords":["artificial intelligence","AI","LLMs","chatbots","virtual assistants","Git","RAG","deep learning","GitHub","NVIDIA","Jupyter","R"],"extracted_tech_keywords":["AI","artificial intelligence","deep learning","Jupyter","RAG","chatbots","virtual assistants","R","Git","GitHub"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/large-language-ai-models\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10061717,"title":"Intel and AMD halt sales of industrial chips to Russia","content":"Two of the largest US semiconductor manufacturers, Intel and AMD, have suspended their chip sales to Russia. These restrictions are in accordance with the trade sanctions that the US government has implemented since Russia’s invasion of Ukraine. The new export barriers are targeting chips that will be used in the military and dual chips that could be used for both military and civilian purposes. This, however, does not include chips for consumer devices like personal computers, mobile phones, digital cameras and more. The ban will be in effect starting March 3. The ban expressly states that these chips include the ones with industrial usage for private companies, government bodies and all individual figures that have been specifically mentioned by the US government, “including the president, prime minister, deputy prime ministers, federal ministers, State Duma deputies and members of the Federation Council, editors-in-chief and deputy editors-in-chief of state media.” “The company is closely monitoring the situation and is enforcing applicable sanctions and export control rules, including new sanctions imposed by OFAC and rules issued by BIS,” said an Intel spokesperson in Russia. Meanwhile, Taiwan-based chip manufacturing company TSMC has also agreed to comply with the new export control regulations, the company said on February 25. Russian homegrown chip companies like Baikal, MCST, Yadro and STC Module are all manufactured by TSMC.","excerpt":"Taiwan-based chip manufacturing company TSMC has also agreed to comply with the new export control regulations, the company said.","categories":["AI News"],"tags":["AMD","Intel","Intel Chips"],"author_name":"Poulomi Chatterjee","publish_date":"2022-02-28T15:12:52","publication_year":"2022","word_count":220,"keywords":["Go","AMD","programming_languages:R","AI","programming_languages:Go","Intel Chips","Git","R","Intel"],"extracted_tech_keywords":["AI","R","Go","Git","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/intel-and-amd-halt-sales-of-industrial-chips-to-russia\/","complexity_score":2,"technical_depth":6,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":43943,"title":"IBM Goes Big On Red Hat, Doubles Down On OpenShift Integration With Cloud Paks","content":"IBM says its cloud portfolio would be modified to run on Red Hat’s OpenShift with specific Cloud Paks released. OpenShift is the industry’s most comprehensive enterprise Kubernetes platform, and Red Hat Enterprise Linux, the world’s leading enterprise Linux platform. IBM in July concluded its $34 billion purchase of Red Hat, the leader when it comes to open source Linux deployments. IBM said it will utilize the Red Hat’s open-source Linux expertise and capabilities to develop large hybrid cloud projects for its huge clientele. The Big Blue even reported that it would try to create an ecosystem of partnerships around Red Hat OpenShift so it can easily carry out the process. The integration has begun with IBM packaging Red Hat’s Kubernetes-based OpenShift Container Platform with more than 100 IBM pre-integrated solutions what it calls Cloud Paks. OpenShift allows users to deployment and manage containers on ony infrastructure, be it private or open cloud, including AWS, Microsoft Azure, Google Cloud Platform, Alibaba and IBM Cloud. “IBM is unleashing its software from the data center to fuel the enterprise workload race to the cloud. This will further position IBM the industry leader in the more than $1 trillion dollar hybrid cloud opportunity. We are providing the essential tools enterprises need to make their multi-year journey to cloud on common, open standards that can reach across clouds, across applications and across vendors with Red Hat,” said Arvind Krishna, senior vice president, Cloud and Cognitive Software, IBM. Also, IBM said it will bring the Red Hat OpenShift Container Platform over to IBM Z mainframe computers and IBM LinuxONE. Together these two platforms control around 30 billion transactions per day across the globe, according to IBM. The two platforms are trusted by industries which exchange the most sensitive data, including banking, government, healthcare and aviation. What IBM wants with Cloud Paks Cloud Paks will have support for all kinds of enterprise cloud requirements, from hardware to software so business users can easily and securely update their critical applications no matter what cloud platform is being used. With the help of these Cloud Paks, businesses can move swiftly with design, test, build and deployment of new models and analytical applications. The prepackaged Cloud Paks incorporates a Kubernetes container and containerized IBM middleware intended to let clients rapidly turn up big business ready containers, the organization said. IBM’s focus is to help its clients deploy hybrid cloud with using secured containers and leveraging the open source capabilities of Open Shift in doing so. By increasing the density of containers, IBM will help clients create containerized applications that can scale both horizontally and vertically. Cloud Paks will be delivered to customers as packages each of them crafted for a particular use case and charged using a consumption-based pricing model. Currently, there are five Cloud Paks in total at the moment: Cloud Pak for Data, Application, Integration, Automation and Multi Cloud Management. The Paks will at last incorporate IBM’s DB2, WebSphere, API Connect, Watson Studio, Cognos Analytics and according to IBM more of its infrastructure will be connected with the Red Hat assets in future. How customers are generating insights using IBM Red Hat Cloud Pak for Data Businesses are already leveraging IBM hybrid cloud platform and Red Hat OpenShift- enabled IBM software (Cloud Paks) to not only help modernize their mission-critical applications but also gain deep insights from data. For instance, US telecom operator Sprint deployed IBM Cloud Pak for Data to help understand and make itself ready for upcoming networking technologies like 5G. Using Cloud Pak for Data to make use Watson Studio, thus creating ML models on client-facing issues from diverse data sets. Similarly, Associated Bank in the US is adopting IBM Cloud Pak for Data System for swift setting up and scaling of AI. The bank wants to create a new Customer 360 system aiming to enhance client experiences as well as a new governed dashboard for data analytics. According to Associated Bank, the new integrated stack contains what it needs to improve data quality, catalog its data assets, enable data collaboration, and build data sciences. Outlook OpenShift is the most fundamental and widely-used Kubernetes and orchestration layer that supports containerized programming, and setting the Cloud Paks on Red Hat OpenShift gives IBM a wide reach right away. This is critical as far as IBM strategy goes as the ability to run enterprise grade, containerized solutions, is a key component of application modernization strategy for any enterprise. IBM-certified containerized software will create a uniform set of services using which security, logging and identity management can be achieved for all IBM customers using the various cloud platforms via their dashboards. With this integration, IBM will take the center stage of deployments of hybrid cloud environments.","excerpt":"IBM says its cloud portfolio would be modified to run on Red Hat’s OpenShift with specific Cloud Paks released. OpenShift is the industry’s most comprehensive enterprise Kubernetes platform, and Red Hat Enterprise Linux, the world’s leading enterprise Linux platform.  IBM in July concluded its $34 billion purchase of Red Hat, the leader when it comes […]","categories":["AI Features"],"tags":["IBM","Linux","OpenShift","private cloud","red hat"],"author_name":"Vishal Chawla","publish_date":"2019-08-07T12:12:41","publication_year":"2019","word_count":788,"keywords":["OpenShift","data science","red hat","AWS","AI","Azure","R","ML","RAG","Aim","IBM","analytics","Linux","private cloud","kubernetes"],"extracted_tech_keywords":["AI","ML","data science","analytics","Aim","RAG","AWS","Azure","kubernetes","R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/ibm-red-hat-openshift-cloud-paks\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10140060,"title":"Anthropic Introduces Claude 3.5 Sonnet with Visual PDF Analysis for Images, Charts, and Graphs under 100 Pages","content":"Claude has introduced a new feature preview that can read all kinds of visuals inside a PDF that is less than 100 pages. This now makes it easy to upload a document, retrieve the complete context, and digest information from PDFs, especially research papers and technical documents that contain charts and graphs, among other images and visuals. Visual PDFs are an experimental feature in the Feature Previews available on Claude 3.5 Sonnet. Claude can now view images within a PDF, in addition to text.This helps Claude 3.5 Sonnet more accurately understand complex documents, such as those laden with charts or graphics.Enable the feature preview: https:\/\/t.co\/bJ8BjBT6zG. pic.twitter.com\/VNSf547ptT— Anthropic (@AnthropicAI) November 1, 2024 The good news doesn’t end there. The company has also increased the document size limit from 10MB to 30 MB. A user on X was quick to point out the change, and Claude now lets you upload a maximum of five images or documents, with a size limit of 30 MB each. “Up until today, when you attached a PDF in Claude.AI, we would use a text extraction service to grab the text and send that to Claude in the prompt. Now, Claude can actually see the PDF visually alongside the text”, said Alex Albert, Head of Claude Relations at Anthropic, in a post on X. You can access the feature from the pop-up banner on the home page. Once you select the Visual PDFs in the Feature Preview tab and turn it on, it will be made available for future conversations. Moreover, Anthropic has also announced that it supports adding PDFs as input in an API request. Anthropic Is on a Roll Only a few days before, Anthropic released Computer Use which caused quite a storm in the AI ecosystem. Recently, they also announced a partnership with GitHub, which included Claude 3.5 Sonnet inside GitHub Coilot. A few days ago, it was also announced that Claude can now execute and run JavaScript code. It’s called the Analysis Tool, and it can also generate data visualisations after it writes and executes the code. Apart from Visual PDFs and Analysis Tool, Claude also provides a feature called LaTex rendering to generate mathematical equations from a user’s input. At this point, it is well established that Claude’s 3.5 Sonnet is the best AI model for running code. OpenAI’s latest GPT o1 isn’t there yet, and even with its latest offering, Canvas, it still fails to keep up with Claude’s abilities.","excerpt":"Days after Claude 3.5 Sonnet received a major update, Anthropic dropped another useful feature called Visual PDF.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","Anthropic","Claude"],"author_name":"Supreeth Koundinya","publish_date":"2024-11-02T12:36:21","publication_year":"2024","word_count":409,"keywords":["Anthropic","Go","OpenAI","GPT o1","AI","Claude","Git","JavaScript","Claude 3.5","R","Java","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","GPT o1","OpenAI","Claude 3.5","Anthropic","R","JavaScript","Go","Java","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/anthropic-introduces-claude-3-5-sonnet-with-visual-pdf-analysis-for-images-charts-and-graphs-under-100-pages\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":805,"title":"What is Helium-3 rich lunar dust and how it can be used to generate energy","content":"If you are an avid space enthusiast, you may have heard about the mining the moon mission and how the project was feverishly taken up by space giants NASA and China. Not just NASA, but there are a host of private players who have jumped onto this mission, which has the potential to solve global energy crisis. Florida-based private commercial space company Moon Express, one of the companies partnering with NASA on their lunar project — NASA lunar initiative known as Catalyst is now reportedly one step closer to its mission of mining the celestial body for Helium-3 gas, platinum-group metals, rare earth metals, helium-3 and moon rocks. Not just Moon Express, Jeff Bezos owned Blue Origin is also ready to support NASA’s lunar missions with a proposed “Blue Moon” lunar lander system. Is helium 3 the answer to global crisis? The renewed interest is in part due to the most valuable resource – Helium 3 gas that could serve as a potential fuel supply and is rarely available on earth. According to writer and futurist Christopher Barnatt, Helium-3 is produced as a by-product of the maintenance of nuclear weapons, which could net a supply of around 15 kg a year. It is also emitted by the sun but the earth’s atmosphere prevents it from reaching the surface. Estimates suggest that there is around 1,100,000 metric tonnes of helium-3 present on the surface of the Moon as the moon absorbs the gas. The valuable gas can be extracted by heating lunar dust to around 600 degrees C, before getting it back to Earth to power a new generation of nuclear fusion power plants, suggest Barnatt. He further added that Nuclear fusion reactors that use helium-3 could supply nuclear power without any waste or radiation. Is it viable to mine helium 3? Though the subject has been of great debate over the last two decades, what with Russia and China setting up programs to mine moon’s rich surface, there are several questions its economic viability and the proposed logistics. China has an aggressive plan of building a manned spacecraft and sending six astronauts to the moon and bring samples back to earth. According to Barnatt, Russia too floated a similar idea in 2006 to set up a permanent moon base and carry out industrial-scale helium-3 production to commence by 2020, but it came to a naught. There is a growing body of concern around science and physics involved in mining and heating the lunar surface, notably the lunar rocks and ferrying the cargo back to earth. According to Barnatt, around one million tons of lunar soil would be needed to be mined and processed for every 70 tonnes of helium-3 yield). Inside Jeff Bezos’s Amazon-type cargo delivery to the moon According to news reports, Bezos owned Blue Origin is keen on developing an Amazon type delivery to the moon that would ferry gear for experiments, cargo and habitats by mid-2020, helping to drive the “future human settlement” on the moon.  Bezos believes it is time for America to return to the moon. The United States’ Apollo 11 was the first manned mission to land on the Moon in July 1969. In a white paper, addressed to NASA, Bezos proposed developing a developing a lunar spacecraft with a lander that would touch down near a crater at the moon’s south pole, and allows proximity to water and continuous sunlight for solar energy. Bezos backed company’s proposal has hinted at flying cargo missions to the moon, helping deliver space equipment needed in setting up the colony. During the Apollo mission, astronauts left footprints and flags. Moon is a reserve of precious resources It’s not just Helium-3, moon houses vast riches, minerals such as gold, cobalt, iron, palladium, platinum and Rare Earth Metals, commonly known as REM that makes it attractive for mining.  Why REMs are in such high demand in the global economy is because of their application in electronic devices. Meanwhile, reports suggest that 90% of current reserves of REMs are controlled by China; so having a steady access to another source has becoming significant. Likewise, the moon has vast amounts of water within its lunar regolith and in the permanently shadowed areas in its north and southern Polar Regions. In 2010, the Mini-RF instrument on board Chandrayaan-1 discovered more than 40 permanently darkened craters near the moon’s north pole that are hypothesized to contain as much as 600 million metric tonnes (661.387 million US tons) of water-ice. This water could serve as a valued source of rocket fuel and could be used for drinking water for astronauts.","excerpt":"If you are an avid space enthusiast, you may have heard about the mining the moon mission and how the project was feverishly taken up by space giants NASA and China. Not just NASA, but there are a host of private players who have jumped onto this mission, which has the potential to solve global […]","categories":["IT Services"],"tags":[],"author_name":"Richa Bhatia","publish_date":"2017-04-25T10:45:52","publication_year":"2017","word_count":766,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","Ray","R"],"extracted_tech_keywords":["AI","Ray","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/helium-3-rich-lunar-dust-can-used-generate-energy\/","complexity_score":3,"technical_depth":6,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":44162,"title":"Hybrid Cloud Storage Is The Answer To The Proposed Data Localisation Bill, Says Vivek Tyagi Of Western Digital","content":"Amidst the hustle and bustle inside a plush boardroom at a luxury hotel in Bengaluru, Vivek Tyagi, Senior Director for Enterprise Sales (India and South Asia) at Western Digital was calm and smiling, sipping a cup of coffee. Speaking to Analytics India Magazine in the middle of an ongoing internal company event, Tyagi spoke about how Western Digital continues to innovate by enabling data to thrive, how India is ready to embrace the data storage changes, and how the biggest changes would soon be seen in Government entities. Analytics India Magazine: Tell us about the state of the data hardware market in India Vivek Tyagi: I believe that in the last couple of years we’ve seen good growth in the data centre market. The reason is in this world — whether it’s the consumer or enterprise — everybody is creating more and more data. Think of smartphones, think of videos — data has penetrated our lives. Similarly, corporate entities are also generating more data. In the earlier days, they only used to store data which was relevant to them, at that point in time. Now with the advent of data science and analytics, they cannot throw any data. Certain industries like banking, finance, insurance and telecom — it is mandated by the government to store the data. And that is why the data centre market has been growing at a very healthy pace. AIM: As the Senior Director at Western Digital, tell us about your role in the bigger scheme of things VT: Western Digital India continues to play a pivotal role. Data has the potential to solve a variety of national challenges in areas of cleantech, education, healthcare, transportation and renewable energy. Western Digital has been looking for ways to not just innovate but support talent and ideas that emanate from the country, be it through hires or the next generation of entrepreneurs in the startup ecosystem. My role here is to understand the needs of enterprise customers and cater to them. Having worked at Sandisk before, where data rapidly evolved from a simple record, or a transaction, into insights and value, we at WD have a deep understanding on how to shape solutions as data and data infrastructure. AIM: Can you talk about India perspective on the dynamic nature of the data storage industry and key technology, growth trends VT: So in terms of the dynamic nature of data storage one of the most demonstrable examples is that of the government. Take the example of election campaigning and hundreds of public policies. They are now very reliant on data, or Big Data, as you may call it. Because the data is growing so rapidly over the last few years, the storage and data centre market has also been growing consistently. Now the Cloud is a big trend. AIM: It is interesting that you should say that. As a supplier of hardware for data storage, do you feel that the market share for you will keep going down? VT: Every enterprise or a user has a different kind of need for data storage. Take for example SMEs or individual users. They are happy to use cloud and cloud-based services. They do not have the infrastructure or the budget for hardware. They are straightaway “born in the cloud”. But once they reach a certain stage certain size of data then it becomes very expensive to keep in the cloud. Similarly, bigger enterprises still need to use physical storage in order to make sure that their data is safe and easily accessible. AIM: So, is where is the future of data storage? VT: Hybrid cloud storage — that’s the solution. Most big companies now adopt this solution. They keep their long-term data, the information they need to use only sporadically, in physical centres. Backup data or archival data which is to be kept for the government regulatory purpose is stored physically. And the keep the data they need regularly on the cloud. AIM: Do you think this will affect the business and market share for Western Digital? VT: (Laughs) That’s a very good question. The answer is — No. Earlier we used to sell storage to companies and end-users. So instead of going to 100 companies to sell storage, now we sell storage to data centres and the Googles and Facebooks of the world. The thing is either with the help of the cloud or otherwise — data has to be stored somewhere. AIM: What are your views on Data Localisation? VT: Now, the Government has realised that data is a valuable resource and an asset. They think it is worth a lot and now they would like all their data to be stored in India. This move will help the Indian companies, sure but MNC will have to make some big changes. But like I said earlier, this is not a new thing. The Government has already mandated certain sectors to store data locally. So as a data hardware provider, we are keenly watching this unfold.","excerpt":"Amidst the hustle and bustle inside a plush boardroom at a luxury hotel in Bengaluru, Vivek Tyagi, Senior Director for Enterprise Sales (India and South Asia) at Western Digital was calm and smiling, sipping a cup of coffee. Speaking to Analytics India Magazine in the middle of an ongoing internal company event, Tyagi spoke about […]","categories":["AI Features"],"tags":["Cloud Data AI","data localisation","Interviews and Discussions","western digital"],"author_name":"Prajakta Hebbar","publish_date":"2019-08-09T19:00:11","publication_year":"2019","word_count":836,"keywords":["big data","data science","Go","API","AI","R","Git","western digital","RAG","Aim","analytics","data localisation","Cloud Data AI","Interviews and Discussions"],"extracted_tech_keywords":["AI","data science","analytics","Aim","RAG","R","Go","Git","API","big data"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/hybrid-cloud-storage-is-the-answer-to-the-proposed-data-localisation-bill-says-vivek-tyagi-of-western-digital\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10019900,"title":"Top AI &#038; Data Science Newsletters On Substack","content":"Artificial Intelligence is a highly dynamic field. And newsletters are a great way to get up to speed with the latest developments in AI. Of late, Substack has emerged as an essential platform that allows creators to deliver the news directly to the reader. Writers can even monetise their newsletters by putting it behind a paywall. The platform has already racked up over 250,000 subscribers just three years into the launch. Here, in no particular order, we list some of the most useful newsletters in AI, ML and Data Science on Substack. The Belamy by Analytics India Magazine Analytics India Magazine’s newsletter goes by the name The Belamy. The weekly newsletter lines up insightful articles on data, AI, machine learning, and the latest developments in emerging technologies. The Belamy also shares resource materials for hands-on projects and in-depth interviews from the industry mavens. Subscribe Here TheSequence The newsletter ‘curated by the industry insiders’ runs on the motto: making people ‘smarter about artificial intelligence, 5 minutes a time’. The newsletter riffs on machine learning concepts, distils key research papers to a few paragraphs, breaks down new frameworks and platforms in concise summaries, and delivers the latest news from the world of artificial intelligence. TheSequence Edge is another semi-weekly subscription-based newsletter under the same banner covering topics such as– machine learning concepts to learn, summaries of machine learning research papers, and new ML frameworks. Subscribe here Data Machina The weekly newsletter curates interesting resources in AI, machine learning, data science, and data engineering. Data Machina’s subscribers include ML professionals, researchers, and analysts. In the free version, users receive a bi-weekly, condensed version of Data Machina. Paid subscribers get the full version with the latest contents and insights every week. The newsletter has a ‘highly-curated’ section of the most interesting AI\/ML topic of the week under the 10-Link-o-Troned section, and shares practical codes and machine learning projects, and datasets for machine learning, etc. Subscribe Here. The AI Ethics Brief The AI Ethics Brief is the mouthpiece of Montreal AI Ethics Institute, a non-profit research institute working in applied technical and policy research in ethical and safe AI. Every week, this newsletter delivers the latest scoop from the world of AI ethics in your inbox, apart from providing an expert commentary on the research and development in AI from around the world. The free newsletter runs on donations. Subscribe Here Gradient Flow Gradient Flow provides high-quality resource material on data, technology and business — with Ben Lorica, Chief Data Scientist at O’Reilly Media at the helm of affairs. The main focus is on machine learning and AI. Gradient Flow was named as one of the Top 10 websites for Data Scientists by edtech platform Coursera. The eponymous newsletter shares highlights from its podcasts, trends in AI and machine learning, and more. Subscribe Here. Our Data-Driven Future This newsletter is managed and published by Ganes Kesari, co-founder of Gramener, where he also serves as the chief data scientist. A ‘data science thought leader’ with over 17 years of experience, Kesari writes about data science-related innovations and how it impacts people, insights on how data science can transform organisations, and other updates on his most popular talks and articles. Subscribe Here ChinAI ChinaAI, as the name suggests, deals with artificial intelligence-related developments at the Asian powerhouse. The newsletter offers English translations of articles and documents from Chinese government offices, think tanks, media, and other experts on artificial intelligence, ML and data science. Subscribe Here","excerpt":"Artificial Intelligence is a highly dynamic field. And newsletters are a great way to get up to speed with the latest developments in AI. Of late, Substack has emerged as an essential platform that allows creators to deliver the news directly to the reader. Writers can even monetise their newsletters by putting it behind a […]","categories":["AI Trends"],"tags":[],"author_name":"Shraddha Goled","publish_date":"2021-02-10T10:00:00","publication_year":"2021","word_count":578,"keywords":["data science","Go","machine learning","artificial intelligence","AI","ML","RAG","data engineering","analytics","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","data science","analytics","RAG","R","Go","data engineering"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-ai-data-science-newsletters-on-substack\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":64522,"title":"How Customer Engagement Has Become A Key Concern For Analytics Companies Amid Crisis","content":"The uncertain time of this pandemic lockdown has indeed disrupted the industry, where businesses are struggling to stabilise their operations and rethinking their marketing strategies. Although businesses have mandated their workforce to deploy best practices while working from home, keeping the customers happy has continued to be a concern for the companies. In fact, in a recent survey — “Impact of WFH on Analytics Function” done by Analytics India Magazine, revealed that 10.5% of respondents are working on customer engagement related analytics — the highest proportion of the new analytics projects followed by 9.5% of respondents working on customer relationship management projects. These numbers alone show that businesses are currently working towards enhancing their customer engagement that is critical for user retention amid crisis as well as for business continuity. Faced by the COVID-19’s unprecedented disruption, customers’ purchases are drastically moving online, and the majority of them are adopting digital tools. Therefore, businesses are now focused on creating ways to enhance their relationship with their customers. In order to have a smooth operation, several businesses are stepping in with products and are using analytics to strengthen their customer relationship amid this lockdown. Currently, brands that have prioritised their investments in creating digital solutions for their customers are in a better position to fight against this crisis. Augie Ray, Gartner customer experience expert, said, “Customer-centric organisations will prepare for how their customers’ questions and needs will change rapidly in the coming months.” And according to Ray, the first step is to “consider likely and possible changes to customer needs and journeys.” Digitisation further brings in a lot of scrutiny from customers’ side, which, in turn, forces businesses to create a seamless customer engagement framework that can not only assist them in improving their relationship but also help in retaining them. Customer Engagement Remained A Growing Concern Amid Crisis & How Analytics Is Helping This mandated quarantine will surely have a long-lasting impact on customer behaviour, with a few industries being hit harder than others. For instance, the travel and hospitality industry is currently struggling to meet their bottom line as the outbreak has locked every customer inside their house. Alongside, people have cut down their consumption and expenses, which has again impacted the manufacturing industry. On the other hand, e-commerce, healthcare, ed-tech, to name a few, have witnessed a tremendous boom during the last few months. Although different industries have different impacts due to this outbreak, enterprises have continued focusing on their customers in order to sustain in this crisis, as well as, emerge after this crisis abates. Concerns like the inability of your customers to physically visit the store and the ability of businesses to deliver services can raise questions in your customers’ minds. And failing to offer these services can hamper your relationship with your customer. Businesses that can take proactive steps in order to address customer requirements amid this crisis can earn a sense of trust and loyalty from their customers. And, that’s where analytics comes into the picture. It is crucial for businesses, at this uncertain time, to serve the right customers with the right message in order to get the best return for your investments. Understanding customers depends on analysing the data, which, in turn, can help businesses to create a level of satisfaction for their customers. Businesses should use relevant data in order to customise their solutions and give their customers personalised solutions that are required in the current times. Utilising data and analytics businesses will be in a better position to offer products, services and even suggestions to their customers. The insights generated from these data will also help businesses to drive new opportunities, generate new revenue streams as well as increase the value for their customers. This is all the more important in this crisis environment where customers are expecting the most out of business, and failing to provide that can hamper the continuity of the business. According to Ravi Narayanan, the global head of insights and analytics at Nisum, “Predictive analytics is a critical tool in delivering a comprehensive view of the consumer to provide these types of experiences.” In fact, in a report, it has been revealed that data-driven enterprises are way more likely to have a better customer experience and have better chances to retain them. Businesses can use data from resources like websites, mobile apps, and third-party organisations in order to improve their customer experience by tracking, monitoring as well as analysing their behaviours. These data will also help businesses to identify the channels and platforms the customer is using, which will provide opportunities for businesses to deliver an omnichannel experience for the customer. Alongside, combining analytics with artificial intelligence can help businesses create the utmost experience for their customers. Predictive analytics also allows businesses to accurately forecast their customer requirements way before the actual buying process. Bharat Krishnamurthy, the CTO of Exide Life Insurance, said to the media that the company combines their data with machine learning in order to predict the documents required from customers to process insurance. He said, “A massive data lake aggregates data from all our systems and third-party sources…” The company has also been using their data to analyse and monitor their customers’ behaviour in order to predict if their customers’ will be able to pay their premiums for the next year. Besides, in order to have seamless customer experience, it is imperative for businesses to have all their customers’ data in one single place. This would require leaders to move beyond siloed systems and create data consolidation in one place; businesses need to have a single view of their customers’ data, which will not only help in enabling real-time analysis but will also help in understanding customers’ touchpoints. In fact, experts believe that moving beyond siloed systems and adopting a single view dashboard requires a mindset to move from legacy systems and more towards the analytics-driven customer-centric process. Wrapping Up Although to sustain amid this crisis businesses need to continuously show empathy to their customers, it is important that businesses also invest in analytical solutions in order to find the right mix of strategies to enhance their customer engagement. With the right customer-centric attitude, along with more-flexible operating models and advanced analytical tools, businesses can not only strengthen their customer relationship but also emerge strongly from this crisis.","excerpt":"The uncertain time of this pandemic lockdown has indeed disrupted the industry, where businesses are struggling to stabilise their operations and rethinking their marketing strategies. Although businesses have mandated their workforce to deploy best practices while working from home, keeping the customers happy has continued to be a concern for the companies.  In fact, in […]","categories":["AI Features"],"tags":["analytics companies","analytics companies in india","analytics in travel industry","analytics insights platforms real-time business","customer engagement","customer experience","different types of analytics","Work from Home"],"author_name":"Sejuti Das","publish_date":"2020-05-05T17:00:00","publication_year":"2020","word_count":1056,"keywords":["artificial intelligence","machine learning","AI","R","analytics in travel industry","ML","Work from Home","Git","customer experience","Ray","analytics companies","analytics insights platforms real-time business","analytics companies in india","customer engagement","analytics","Rust","predictive analytics","different types of analytics"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","analytics","Ray","predictive analytics","R","Rust","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-customer-engagement-has-become-a-key-concern-for-analytics-companies-amid-crisis\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10048805,"title":"A Guide to Text Preprocessing Using BERT","content":"Various state-of-the-art NLP applications like sentiment analysis, question answering, smart assistance, etc. require a tremendous amount of data. This large amount of data can be directly fed to the machine learning model. Almost all the text-based applications require a lot of pre-processing with the textual data such as creating the embedding vectors from scratch using the word frequency counter. This consumes a lot of effort and time. To overcome this, transfer learning models are used now for all complex pre-processing tasks. Here, we just need to feed our raw text to the transfer learning model and the rest of the processes are taken care of by it. In contrast to this, we are going to discuss one such transfer learning framework, BERT, in this article. We will see how to use the BERT pre-processing module to easily generate word embeddings without putting in a lot of effort. The major points to be covered in this article are listed below. Table of Contents Standard Procedure of Text Pre-processingWhat is BERT?Working of BERTPre-processing Model Let’s start with the discussion. Standard Procedure of Text Pre-processing A technique known as text preprocessing is used to clean up text data before feeding it to a machine-learning model. Text data contains a variety of noise, such as emotions, punctuation, and text in a different capitalization. This is only the beginning of the difficulties we will face because machines cannot understand words, they require numbers. So we must find a fast and efficient way to transform text to numbers. The standard or conventional procedure of pre-processing is a little bit tedious and also a user-centric procedure. The below steps are carried out under the hood of standard pre-processing techniques: Lower casing the corpus Removing the punctuation Removing the stopwords Tokenizing the corpus Stemming and LemmatizationWord embeddings using CountVectorizer and TF-IDF The worked-out examples of the above steps have been covered in these articles: Complete Tutorial on Text Preprocessing in NLP and How to Identify Entities in NLP? Usually while approaching any NLP problem we tend to follow this process and the above process does not ensure any reasonable result if our raw data changes slightly. This means if the data is from a web page there we need additional work to remove HTML tags. Nowadays all these pre-processing steps can be carried out by using transfer learning modules like BERT. What is BERT? BERT is an acronym for Bidirectional Encoder Representations from Transformers. In order to pre-train deep bidirectional representations from unlabeled text, the system uses context conditioning on both the left and right sides of the sentence. As a result, the pre-trained BERT model could also be fine-tuned by adding only one more output layer to produce cutting-edge models for a wide range of NLP tasks. BERT has been pre-trained on a vast corpus of unlabeled text, including the entire Wikipedia, which is 2,500 million words long, and various Book Corpus, which is over 800 million words long. Half of BERT’s success can be attributed to this pre-training phase. That’s because, as the model is trained on a big text corpus, it begins to pick up on the more subtle and personal details of how the language works. This information can be applied to a wide variety of NLP tasks. BERT is a model which is quite bidirectional. Bidirectional indicates that during the training phase, BERT learns information from both the left and right sides of a token’s context. A model’s bidirectionality is essential for completely comprehending the meaning of a language. Working of BERT To learn the contextual relationships between words in a text, BERT utilizes Transformer, an attention mechanism. The transformer’s vanilla implementation has two mechanisms: an encoder that receives text input and a decoder that predicts the task. Only the encoder mechanism is required because the purpose of BERT is to construct a language model. The Transformer encoder reads the entire sequence of words at once, unlike directional versions that read the text input sequentially. It is classed as bidirectional as a result of this, while the actual term is non-directional. This feature allows the model to learn a word’s context based on its surroundings. Figure1: BERT During the BERT training process, pairs of sentences are provided as input to the model, and it learns to predict whether or not the second sentence in the pair is the following sentence in the original document. Half of the inputs during training are pairs where the second sentence is the next sentence in the original document while the other half is a random sentence from the corpus. The underlying assumption is that the second phrase will be unrelated to the first. During training, as shown above, a [CLS] token is inserted at the beginning of the first sentence and a [SEP] token is introduced at the end of each sentence, with each token containing a sentence embedding indicating Sentence A or Sentence B. Sentence embeddings are essentially similar to token embeddings, but with a two-word vocabulary. Finally, each token is assigned a positional embedding that corresponds to its place in the sequence. Figure 2: BERT Pre-training Before feeding word sequences into BERT, some part of each sequence is replaced with a [MASK] token. The model then makes an attempt to forecast the original value of the masked words using the context provided by the other, non-masked phrases in the sequence. It is necessary to add a classification layer on top of the encoder output in order to predict the output words. This is followed by multiplying the encoder output vectors by the embedding matrix, transforming them into the vocabulary dimension, and computing the probability of each word in the vocabulary using softmax. The BERT loss function only considers the predictions of the masked values and ignores the predictions of the non-masked words. Consequently, the model converges more slowly than directional models. When learning the BERT model, Masked LM (shown in Figure1) and Pre-training (shown in Figure2) are trained jointly in order to minimize the combined loss function of the two techniques. BERT Pre-processing Model There are a variety of Pre-trained BERT models available on Tensorflow Hub like original BERT, ALBERT, Electra, and MuRIL which is a multilingual representation for Indian language, pre-trained on 17 different Indian languages, and many more available. Encoder and pre-processing API is available for all the above models. There is a preprocessing model for each BERT encoder. Using TensorFlow operators from the TF.text package, it converts raw text to the numeric input tensors expected by the encoder. Unlike pure Python preprocessing, these operations can be incorporated into a TensorFlow model for serving directly from text inputs. Each TF Hub preprocessing model comes preconfigured with a vocabulary and its related text normalization logic and requires no additional configuration. Use Case of BERT Let’s implement a few examples of pre-processing. !pip3 install --quiet tensorflow-text import tensorflow_hub as hub import tensorflow_text as text # load the pre-processing model preprocess = hub.load('https:\/\/tfhub.dev\/tensorflow\/bert_en_uncased_preprocess\/1') # Use BERT pre-processing on a batch of raw text inputs. embeddings = preprocess(['Blog writing is awesome.']) The pre-processed output from the module is obtained as shown above, where the input mask tells the encoder which part of a sentence is important because here padding is present.  Further, the input type id indicates to the model which part of the input corresponds to 1st sentence and second sentence in our given only one sentence. Input_word_ids are the indices that correspond to each token. Conclusion In this post, we have understood what BERT actually is and how it works. We also saw how easily the word embedding can be implemented using BERT pre-processing modules. All the traditional pre-processing steps are included in the BERT pre-processing modules which saves a lot of time while building the NLP-based model. References Research PaperHugging FaceTensorflow Blog","excerpt":"This blog discuss about how to use SOTA BERT for pre-processing the textual data","categories":["Deep Tech"],"tags":["AI (Artificial Intelligence)","BERT Model","Deep Learning","Guide","Machine Learning","Transfer Learning"],"author_name":"Vijaysinh Lendave","publish_date":"2021-09-19T14:00:00","publication_year":"2021","word_count":1298,"keywords":["Hugging Face","machine learning","TPU","AI","sentiment analysis","ML","Transfer Learning","Machine Learning","Transformers","BERT Model","NLP","Python","Deep Learning","AI (Artificial Intelligence)","TensorFlow","Guide"],"extracted_tech_keywords":["AI","machine learning","ML","NLP","TensorFlow","Hugging Face","Transformers","sentiment analysis","TPU","Python"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/a-guide-to-text-preprocessing-using-bert\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10044539,"title":"10 AI Acquisitions By Major Companies In 2021 (So Far)","content":"AI startups are a prime acquisition target for companies looking to leverage AI and ML technology, especially in the wake of the pandemic. The pace of AI acquisitions has been steadily rising as many companies incorporate AI into their core offering. In this article, we look at the major AI acquisitions in 2021. Nuance Microsoft announced the acquisition of Nuance, an AI-based Tech company for $19.7 billion. This is Microsoft’s second-costliest acquisition after Linkedin ($26.2 billion). Nuance provides speech recognition and conversational AI services and is best known for its deep learning voice transcription service popular in the health sector. Nuance’s voice recognition technology also powers Apple’s Siri. Microsoft is making efforts to provide industry-specific cloud offerings. Nuance and Microsoft have collaborated in the past for building technology for recognising and transcribing speech with ambient clinical intelligence (ACI) technologies and Microsoft hopes to double its total addressable market (TAM) in the healthcare space to nearly $500 billion. Turbonomic IBM has bought Turbonomic, an AI-powered Application Resource Management company. The acquisition will help IBM boost its full-stack application observability and management efforts to ensure better performance and minimise cost using AIOps. Provino Google has acquired semiconductor IP company Provino Technologies that provides network on chip (NoC) solutions for next-gen system-on-chips (SOCs) focused on AI\/ML. Provino develops interconnect protocols to be deployed in next-generation consumer security applications. With Provino, Google is most likely to adopt NoC technologies for running various research work in artificial intelligence and machine learning. Blue Yonder In its biggest acquisition in a decade, Panasonic Corp has bought Blue Yonder, a supply chain and AI software provider company, for $7.1 billion. The acquisition will strengthen Panasonic’s portfolio and accelerate the companies’ Autonomous Supply Chain mission, empowering customers to optimise their supply chains using AI, ML and loT. Blue Yonder is one of the largest supply chain software providers with clientele including Coca Cola, P&G and Unilever. The Panasonic-Blue Yonder deal comes at a time of heightened need for real-time data and automated solutions, due to the pandemic and growth of e-commerce. Determined AI Hewlett Packard Enterprise acquired San Francisco startup Determined AI. By removing the complexity of ML development, Determined AI’s open source ML platform accelerates time to production. With the acquisition, HPE seeks to further its mission in making AI heterogenous and empower ML engineers to build ML models at a larger scale. HPE will combine the startup’s AI training platform with its high-performance computing solutions. Text IQ Global legal and compliance tech company Relativity acquired Text IQ, an AI company as a part of its effort to embed AI technologies deeper within its core platform. Text IQ employs unsupervised graphic modeling, natural language processing, machine learning, social network analysis and deep learning to search for the most relevant data to a legal case. Over time, the AI capabilities of Text IQ will be integrated into the Relativity platform. Innovyze Innovyze Inc, provider of Smart Water Infrastructure Modeling and Simulation Technology, was bought by Autodesk Inc for $1 billion. The company leverages artificial intelligence and machine learning solutions to predict and optimise operations. The acquisition helps Autodesk position itself as a leading global provider of end-to-end digital solutions of water infrastructure and creates a  path to a more sustainable and digitized water industry. Brandwatch Online consumer intelligence platform Brandwatch is acquired by Cision for $450 million. Brandwatch uses AI and machine learning for “social listening”. The deal combines the two players to offer a range of services from PR to online customer engagement and marketing. Chartio Australian software company Atlassian acquired Chartio to add a new data analysis and visualisation component to the company’s family of products. Atlassian is acquiring Chartio to bring data visualisation to the platform. Deepcube Leading additive electronics provider Nano Dimension acquired machine learning\/deep learning leader Deepcube Ltd. Deepcube applies various patented breakthrough algorithms to enhance data analysis and deployments of deep learning-based AI systems. The ML application enables faster and more accurate training of deep learning models, improving inference performance and real-time metrics. DeepCube’s scientists and engineers will join Nano Dimension. Deepcube’s training platform and inference engine will be integrated into Nano Dimension AME 3D-printers which will act as nodes and AI control center in a Smart Fabrication Network.","excerpt":"We look at the major AI acquisitions in 2021.","categories":["AI Trends"],"tags":["AI Companies"],"author_name":"Prajaktha Gurung","publish_date":"2021-07-27T16:00:00","publication_year":"2021","word_count":706,"keywords":["Go","machine learning","artificial intelligence","AI","ML","Git","RAG","deep learning","ViT","AI Companies","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","RAG","R","Go","Git","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/10-ai-acquisitions-by-major-companies-in-2021-so-far\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10097077,"title":"Generative AI in Healthcare and Insurance","content":"AI has been making significant contributions to the healthcare sector for several years. From improving medical adherence to patient analytics and solving complex problems within the pharmaceutical industry, AI has been leveraged to enhance various aspects of healthcare. However, the adoption of AI in the life sciences sector has been relatively slow compared to other industries. The COVID-19 pandemic and the digital era have prompted increased openness towards AI in healthcare and insurance. Fractal, a leading AI solutions provider, believes that generative AI holds great potential for transforming these sectors. In a conversation between Bhaskar Roy, Client Partner and Bangalore Center Head at Fractal and Anindya Sengupta, Client Partner, Insurance and Healthcare at Fractal, AIM discusses the applications of generative AI, ethical considerations, collaborations between insurance companies and healthcare providers, and the promising avenues for the future. AIM: How is generative AI being used in the healthcare industry? Roy begins by highlighting that AI has been used in healthcare for a long time, particularly in areas like medical adherence and patient analytics. However, the adoption of AI in the life sciences sector has been slower compared to other industries. The COVID-19 pandemic has accelerated the digital transformation in healthcare, leading to increased interest and openness towards AI in the industry. Anindya adds that there are two aspects of the healthcare business: the provider side, focusing on patient journeys and outcomes, and the payer side, involving health insurance companies. He emphasises the significance of using generative AI to simplify processes and improve customer experience on the payer side. Insurance companies can play a crucial role in streamlining  the claim process using generative AI and enhance various aspects of it. AIM: What are the ethical considerations when it comes to using AI in healthcare related to privacy of data of patients? Roy and Anindya acknowledge the importance of maintaining privacy and data security when developing algorithms and solutions. They stress the need for accountability and the mitigation of biases in AI systems. Healthcare is a critical sector where wrong decisions can have drastic consequences, making responsible AI implementation essential. An example is discussed where ChatGPT was used to analyse a dog’s medical records and determine the effectiveness of its treatment. This raises an ethical question regarding attributing the treatment’s success to the AI model or the previous veterinarian’s expertise. Such ethical dilemmas need to be addressed when integrating AI into healthcare and insurance. Hence the importance of responsible AI becomes paramount with the more use of generative AI. AIM: What is the relationship between insurance companies and using generative AI in healthcare? Anindya explains that hospitals and insurance companies often collaborate on aspects such as managing readmission rates and analysing claim data. These collaborations enable data sharing and facilitate the development of AI solutions that benefit both parties. For example, insurance companies can provide data that helps pharmaceutical companies understand treatment outcomes and improve patient analytics. Patient consent and the ethical considerations surrounding data sharing between pharma and insurance companies is also of utmost importance. Currently, there may not be explicit options for patients to opt-out of their data being used for generative AI services. However, as the industry matures, regulations and guidelines surrounding patient consent and data usage will likely be established. Roy explains that collaboration between insurance companies and healthcare providers is vital. Hospitals often partner with insurance companies to manage aspects such as readmissions and leverage insurance data for monitoring patient outcomes. Data collaborations between pharmaceutical companies, insurance providers, and healthcare practitioners are necessary to gather patient data effectively. Generative AI can assist in patient analytics and help pharmaceutical companies gain insights into patient data through collaboration with insurance companies. AIM: What are the future avenues of using generative AI in healthcare and life sciences? Roy identifies three promising avenues for generative AI in healthcare and life sciences. Firstly, personalised medicine holds significant potential. Generative AI can combine patient data to match specific characteristics with ideal drugs, enabling targeted treatments. Secondly, repurposing existing drugs for new diseases can be achieved through generative AI, providing opportunities for medical advancements. Thirdly, generative AI can streamline the drug discovery process, identifying suitable compounds and “short-circuiting” the time-consuming stages leading to clinical trials. This can make drug development more feasible and affordable. Anindya adds a  global context to the discussion. He said that health insurance is considered part of general insurance outside the US, emphasising the close relationship between life and health coverage. Throughout the world, life and health insurance are typically treated as a unified entity by major insurance companies. Within the industry, two key departments play a significant role: agencies and underwriting. Agencies drive the industry,  employing agents to facilitate sales. Underwriting involves the pricing and risk assessment of insurance policies. However, these departments face a challenge related to an ageing workforce, with many employees approaching retirement in their mid-50s. Attracting younger talent to fill these positions has proven difficult, resulting in a knowledge and experience gap. Generative AI can act as a knowledge assistant, which can help address this challenge. By leveraging the knowledge and expertise of experienced professionals, a digital twin can be created to assist in training new agents and underwriters. Acting as a knowledge assistant, this technology has the potential to significantly boost productivity for the newer workforce in the life and health insurance industry. Discussions and conferences with industry CEOs have revealed that the ageing workforce is a major concern for  insurance companies. The implementation of generative AI, through the use of knowledge assistants, offers a solution to bridge the experience gap and improve the productivity of newer agents and underwriters in the industry.","excerpt":"In a conversation between Bhaskar Roy and Anindya Sengupta, AIM discusses the applications of generative AI, ethical considerations, collaborations between insurance companies and healthcare providers, and the promising avenues for the future.","categories":["AI Highlights"],"tags":["Fractal AI","Gen AI in Insurance","Generative AI"],"author_name":"Mohit Pandey","publish_date":"2023-07-18T11:03:10","publication_year":"2023","word_count":938,"keywords":["Go","ChatGPT","AI","ML","Git","RAG","Aim","Gen AI in Insurance","analytics","generative AI","Generative AI","R","Fractal AI"],"extracted_tech_keywords":["AI","ML","analytics","generative AI","ChatGPT","Aim","RAG","R","Go","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/fractals-insights-on-using-generative-ai-in-healthcare-and-insurance\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10069085,"title":"Google open sources Switch Transformer models in T5X\/JAX","content":"Google Brain has open sourced Switch Transformer models including 1.6T param Switch-C and the 395B param Switch-XXL in T5X\/JAX. Check out the GitHub repository here. What is JAX? JAX (Just After eXecution) is a machine\/deep learning library developed by DeepMind. All JAX operations are based on XLA or Accelerated Linear Algebra. XLA, developed by Google, is a domain-specific compiler for linear algebra that uses whole-program optimisations to accelerate computing. XLA makes BERT’s training speed faster by almost 7.3 times. T5X T5X is a modular, composable, research-friendly framework for high-performance, configurable, and inference of sequence models (starting with language) at many scales. T5X allows many different classes of language tasks to be expressed uniformly, and a single encoder-decoder architecture can handle them without any task-specific parameters. Switch Transformer In deep learning, models usually reuse the same parameters for all inputs. However, the Switch Transformer uses a Mixture of Experts (MoE) and selects different parameters for each incoming example. The model is primarily used for NLP research. The Transformer uses an algorithm called Switch Routing. Instead of activating several experts and combining the output, it chooses a single expert to work on an input. The algorithm simplifies the routing computation and reduces communication costs since individual expert models are hosted on different GPU devices.","excerpt":"The models include the 1.6T param Switch-C and the 395B param Switch-XXL.","categories":["AI News"],"tags":["Generative Pre-Trained Transformer"],"author_name":"Tasmia Ansari","publish_date":"2022-06-15T12:55:51","publication_year":"2022","word_count":212,"keywords":["Go","TPU","AI","ML","Git","NLP","deep learning","Generative Pre-Trained Transformer","JAX","GitHub","R"],"extracted_tech_keywords":["AI","ML","deep learning","NLP","JAX","TPU","R","Go","Git","GitHub"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/google-open-sources-switch-transformer-models-in-t5x-jax\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10167790,"title":"Kamal Haasan Visits Perplexity HQ, Meets CEO Aravind to Discuss AI in Filmmaking","content":"Veteran actor and filmmaker Kamal Haasan visited the headquarters of Perplexity AI in San Francisco this week, where he met co-founder and CEO Aravind Srinivas. The meeting focused on the role of emerging technologies in creative industries, particularly filmmaking. Haasan shared the experience in a post on X, and wrote, “From cinema to Silicon, the tools evolve—but our thirst for what’s next remains.” He added that he was inspired by his visit to Perplexity HQ and acknowledged the contributions of Indian talent, referring to Srinivas and his team as “brilliant”. “[It] was great to meet and host you at the Perplexity office! Your passion to still learn and incorporate the cutting-edge technology in filmmaking is inspirational,” Srinivas responded on X. Haasan is known for blending technology with cinema throughout his career. Highlighting his interest in innovation and continuous learning, he wrote, “Curiosity didn’t kill the cat—it created Aravind Srinivas and Perplexity AI.” Last year, Haasan enrolled in a 90-day crash course on artificial intelligence at a leading US institution. At 69, he continues to explore how new technology can reshape storytelling in Indian cinema. Despite his busy schedule, Haasan’s engagement with AI is likely to influence his upcoming projects. “I have a deep interest in new technology, and you can often see my films experimenting with the latest technological developments,” Haasan said in an interview in Abu Dhabi last year. “Cinema is my life. All my earnings have gone back into my films by various means. I’m not merely an actor, but a producer too, and I reinvest everything I earn from movies into the industry,” he added.Haasan is involved in multiple high-profile projects, including Thug Life with director Mani Ratnam, Indian 3, a collaboration with action director duo Anbariv, and the sequel to Kalki 2898 AD.","excerpt":"“From cinema to Silicon, the tools evolve—but our thirst for what’s next remains.”","categories":["AI News"],"tags":["Perplexity AI"],"author_name":"Siddharth Jindal","publish_date":"2025-04-11T12:36:03","publication_year":"2025","word_count":297,"keywords":["Go","artificial intelligence","programming_languages:R","AI","innovation","programming_languages:Go","Perplexity AI","R"],"extracted_tech_keywords":["AI","artificial intelligence","R","Go","innovation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/kamal-haasan-visits-perplexity-hq-meets-ceo-aravind-to-discuss-ai-in-filmmaking\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10093561,"title":"From Leader to Laggard: How Google Lost Its AI Mojo","content":"Google founder Larry Page envisioned the search company as a great AI platform whose mission would remain unfulfilled until its search engine becomes AI-complete. However, it seems that the enormous ambition of the former co-founder has fizzled out, considering the current state of Google in the AI space. The once-market leader in AI is now playing second fiddle to OpenAI, looking to replicate the latter’s success by aping its products. The main focus of Google I\/O this year was Bard, a conversational agent with access to the Internet and a direct shot at OpenAI’s ChatGPT. In a mad rush to get Bard to the market, the Mountain View giant seems to have ignored its rich history of AI innovation, leaving them behind as another product killed by Google. Code Red gone wrong In December last year, ChatGPT’s success raised a furore at Google, prompting CEO Sundar Pichai to declare a Code Red for Google Search. According to a source at the company, executives believed that the chatbot could hamper Google’s ads business, which function mainly through Google Search. This then prompted a frenzied rush to create a competing chatbot, resulting in the release of Bard in February. In classic Google fashion, the tech giant forgot all about their previous endeavours in AI and chased the shiny new chatbot trend. In actuality, Google’s AI efforts started even before the founding of OpenAI, built on their treasure trove of user data. Google first began dabbling in machine learning in the early 2010s, with the first big announcement coming in the form of Google Now in 2012. Google Now used machine learning to recommend news articles based on their search history. This then evolved to the Google Assistant, which used voice recognition and built on the legacy of Google Now to be a digital assistant to users. This slowly transitioned into a company-wide strategy. Sundar Pichai, the CEO of Google, stated in a quarterly financial results call in 2015 that “Machine learning is a core, transformative way by which we’re rethinking how we’re doing everything. We are thoughtfully applying it across all our products, be it search, ads, YouTube, or Play. And we’re in the early days, but you will see us — in a systematic way  –  apply machine learning in all these areas.” Fast forward to today, and this strategy has come to fruition. Google Search uses the PageRank algorithm for determining top spots, Google Ads use responsive ads for better targeting, and YouTube’s recommendation algorithms set the standard for content recommendations. The company also has a rich history of AI research, stemming from its Google Brain division, now integrated into DeepMind. The team behind Google Brain included AI superstars like Andrew Ng, Samy Bengio, Ilya Sutskever, and Geoffrey Hinton. This team was responsible for the research behind Google Translate, GANs, and transformers —— the ’T’ of ChatGPT. However, it seems that Google forgot about these strategies and advancements with the runaway success of ChatGPT. Google Brain drain There is also a long list of people leaving Google behind to start their own AI companies. Dario Amodei, the CEO of Anthropic, worked as a senior research scientist at Google Brain. Cohere was also founded by Aidan Gomez and Nick Frosst, both leading researchers at Google Brain. Andrew Ng, one of the co-founders of Google Brain, now works closely with OpenAI and is the founder and CEO of deeplearning.ai. More recently, Geoffrey Hinton, the so-called godfather of AI, recently left Google, citing concerns over AI risk. This has left Google in a sticky situation, bleeding AI talent in the midst of a recession. It is also unable to capitalise on the AI needs of the day, and its future is looking bleak as well. In a leaked document, a senior Google researcher stated that Google has “no moat” when it comes to AI. This document also conceded defeat to the open source community, while putting OpenAI in the same boat as itself. However, the difference between Google and OpenAI in this scenario is that OpenAI releases its products to market first, and then optimises them based on users’ experience. Google, on the other hand, is making a lot of noise on using AI, but the products are being released at a trickle. According to reports, executives mentioned AI 143 times over the course of the two-hour keynote. However, the only product that was opened to the public after the keynote was Google Bard, which the company reminds its users is ‘experimental’. Even as Google sets the stage for their ‘bold and responsible AI’ products ‘coming soon’, it seems that they have already fallen behind in the AI race. Ignoring their rich history of bringing AI innovation to the market, Google is running behind the shiny new thing, seeming to miss the forest for the trees.","excerpt":"Google was once a beast in the AI space, but now, it seems to be content with playing the second fiddle","categories":["Global Tech"],"tags":[],"author_name":"Anirudh VK","publish_date":"2023-05-18T13:00:00","publication_year":"2023","word_count":803,"keywords":["Anthropic","ChatGPT","Go","machine learning","OpenAI","AI","Google Bard","Transformers","Git","R"],"extracted_tech_keywords":["AI","machine learning","ChatGPT","OpenAI","Anthropic","Google Bard","Transformers","R","Go","Git"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/from-leader-to-laggard-how-google-lost-its-ai-mojo\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":39699,"title":"5 AI startups In China That Are Making Waves Internationally","content":"Image source: techgenez.com The Chinese AI startup ecosystem is known for its cutting-edge technology solutions, thus superseding other nations and technology giants to become one of the leading players in the field. As it continues to make substantial strides in the field all in an attempt to solidify its position as the top AI leader in the world, the Chines startups, government and other key stakeholders have been pumping in millions of dollars to further developments in the field. According to a recent study, the country’s spending on AI for technology and communication industry has increased to 71.6% during 2018 to reach US$ 320 million. It is expected that the country’s spending on AI technologies recorded a CAGR of 25.9%, increasing from US$ 481.2 million in 2019 to reach US$ 2,406.1 million by 2025. In this article, we look at the top AI startups in China and how they have carved a niche in the industry. Ubtech Robotics: This is an artificial intelligence humanoid robotic company, based out of Shenzhen in China. The startup was established in the year in 2012 by entrepreneurs Zhou Jian and designs a range of AI-powered robotics for different sectors and serves customers in Germany, the United States, India, Norway, France, Italy, and other countries. Since its inception, the startups have raised three rounds of funding so far. In 2018, it raised $820 million in a series C funding, making it the largest funding round ever for an artificial intelligence company. Among its range of solutions include Iron-man and Star Wars-themed robots that are capable of facial recognition, voice commands and equipped-with augmented reality and infrared sensors. SenseTime: It is an AI company that focuses on innovative computer vision, learning technologies and image recognition. The startup has raised as many as seven rounds of funding and is considered to be as the world’s most valued AI startup. The latest round of funding was led by Qualcomm’s venture capital initiative where it invested a whopping $620 million in fresh capital fund. SenseTime was established in 2014 and is co-founded by Bing Xu, Li Xu, Xiaogang Wang and Xiaolan Xu. The startup’s AI-based solutions range from smart-solution to advertising and it has closely worked with the Ministry of Science and Technology of China to establish its National Open Innovation Platform for Next-Generation Artificial Intelligence on Intelligent Vision, sensitive is the 5th national AI platform in China. In 2017, SenseTime also debuted Intelligent Cockpit, its solution for the autonomous driving car which is powered by seven AI-empowered functions including ID identification, fatigue detection, distraction detection, face detection, eye tracking, gesture recognition, and passenger’s characteristics analytics. Face and body sensing, robot sensing, image detection, deep learning platform, video And image processing are some of the areas that the startup works in. Megvii: Set up in 2011, the startup is an industry-oriented solution provider for AI-based technologies. Ali Baba, Huawei, Lenovo etc are some of the big names among its customers and has also collaborated with the Chinese Ministry of Public Security and Taxation for proving it slew of AI solutions. Since its inception, the startup has raised seven rounds of funding with the latest being as recent as May, where it raised $750 in series D funding. Mobvoi: The startup is focused on solutions in the field of Chinese language speech recognition, natural language processing, and vertical mobile search. It was established in the year 2014 by Zhifei Li, a former Google employee with the aim of defining the next-generation of human-machine interaction. A deeply-tech driven company, the startup has developed a series of proprietary interactive AI technologies, including wake words, speech recognition, natural language understanding etc. Its watch series is a range of solutions that combines AI with wearable devices and are equipped with features like Google’s Wear OS, health monitoring and a GPS system. The startup has raised five rounds of funding totalling up to $252.7M. Rokid: Is a Chinese startup that is that focuses on consumer product and leverages a range of technologies such as AI, augmented reality, consumer electronics, human-computer interaction, and robotics. The startup was established in 2014 by Eric Wong and Mingming Zhu with an aim to superior industrial design and superlative user experience to its customers. Currently, Rokid focuses on  research and development in the field  of speech and imaging technology and has a line smart speakers with inbuilt assistant speakers.","excerpt":"The Chinese AI startup ecosystem is known for its cutting-edge technology solutions, thus superseding other nations and technology giants to become one of the leading players in the field. As it continues to make substantial strides in the field all in an attempt to solidify its position as the top AI leader in the world, […]","categories":["AI Trends"],"tags":["AI and Deep Learning","China","Lenovo","ML","Robotics","Startups"],"author_name":"Akshaya Asokan","publish_date":"2019-05-26T15:25:07","publication_year":"2019","word_count":729,"keywords":["Go","artificial intelligence","AI","AI and Deep Learning","ML","image recognition","Robotics","computer vision","RAG","Aim","deep learning","Lenovo","analytics","Startups","R","China"],"extracted_tech_keywords":["AI","artificial intelligence","deep learning","computer vision","analytics","Aim","RAG","image recognition","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/5-ai-startups-in-china-that-are-making-waves-internationally\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10032697,"title":"Facebook’s New Dataset To Make Facial Recognition Models More Inclusive","content":"Recently, researchers at Facebook AI introduced a new dataset, Casual Conversations, to measure the robustness of AI models across four main dimensions, age, gender, apparent skin type and lighting. Fairness in AI models is a hot topic in computer vision, with researchers around the world invested in developing fair & inclusive AI models. “Top performing AI models trained on datasets that are created without considering fair distribution across subgroups and thus quite unbalanced, do not necessarily reflect the outcome in the real world. On the contrary, they may perform poorly and may be biased towards certain groups of people,” the researchers said. The attributes of already existing facial datasets are either annotated by third parties or hand-labelled. Though the researchers have claimed the annotations are uniformly distributed over different attributes, such as age and gender, the accuracy of annotations is still suspect. Deepfakes are software that uses a machine-learning algorithm to create a computer-generated version of the subject’s face. Deepfake detectors are able to differentiate between real and fake videos by accumulating classifier responses on individual frames. Although aggregating per-frame predictions removes most outliers for a robust and accurate classification, bias in face detectors may change the final result dramatically and cause deepfake detectors to fail. Are these  detectors capable of detecting faces from various age groups, genders or skin tones? If so, how often do deepfake detectors fail based on this reason? Is there a way we can measure the vulnerabilities of deepfake detectors? The researchers at Facebook AI proposed the dataset to address these questions. Behind the dataset Casual Conversations comprises video recordings of 3,011 individuals across genders, apparent skin types and age. It has approximately fifteen one minute video recordings for each of the 3,011 subjects. The dataset also includes a unique identifier and age, apparent skin type annotations, and gender for each subject. The Casual Conversations dataset is unique since all the other publicly available datasets provide hand or machine labelled annotations and therefore introduce a bias towards a person’s appearance other than the actual age and gender. In this dataset, the age and gender annotations are provided by the subjects. The researchers called it one of the distinguishing features of this dataset. “We prefer this human-centred approach and believe it allows our data to have a relatively unbiased view of age and gender,” the researchers said. In addition to self-identified age and gender labels, as a third dimension in the dataset, the researchers annotated the apparent skin tone of each subject using the Fitzpatrick scale. Also, they labelled the videos recorded in low ambient lighting. This set of attributes allowed them to measure the robustness of the model on four dimensions, such as age, gender, apparent skin tone and ambient lighting. Advantages The Casual Conversations dataset has uniform distributions across categories. It could measure various AI methods, such as face detection, apparent age and gender classification, or assess robustness against various ambient lighting conditions. According to the researchers, though Casual Conversations is intended to evaluate the robustness of AI models across facial attributes, the dataset could come in handy for various open challenges. The potential application areas include image inpainting, developing temporally consistent models, audio understanding, responsible AI on facial attribute classification and handling low-light scenarios in the problems mentioned above. Wrapping up “Our dataset will be publicly available for general use and we encourage users to extend annotations of our dataset for various computer vision applications, in line with our data use agreement,” the researchers said.","excerpt":"Facebook AI introduced a new dataset, Casual Conversations, to measure the robustness of AI models.","categories":["AI Trends"],"tags":["ai dataset","data science project marketing","Datasets","datasets on deep learning","Facebook AI","Facebook AI research"],"author_name":"Ambika Choudhury","publish_date":"2021-04-22T10:00:00","publication_year":"2021","word_count":582,"keywords":["Go","Datasets","fairness in AI","Facebook AI","AI","R","ML","deepfakes","computer vision","RAG","responsible AI","datasets on deep learning","Aim","Facebook AI research","data science project marketing","ai dataset"],"extracted_tech_keywords":["AI","ML","computer vision","Aim","RAG","R","Go","responsible AI","fairness in AI","deepfakes"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/facebooks-new-dataset-to-make-facial-recognition-models-more-inclusive\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":52676,"title":"This Bootstrapped Nutraceutical-Tech Startup Is Providing AI-Based Solutions To Help Curb Tobacco Addiction","content":"India is home to 4,892 startups in the health-tech space, and they have raised a total of $504 million between 2014-2018. Due to the lack of expert healthcare professionals, the healthcare industry offers a huge opportunity for AI to enable faster and efficient healthcare delivery. The healthcare industry in India is said to be one of the fastest-growing sectors backed by India’s rising income, health awareness and access to insurance. The increased quality of lifestyle and the rise of stress-related diseases is also playing a crucial role. According to various researches done in global markets, the nutraceutical market in India was valued at ₹26,000 crore in 2017 and is expected to reach ₹80,000 crore by 2023. With a vision to build alternative foods and health-based nutraceutical-tech brand that helped consumers live a healthy lifestyle by offering them home remedy based products which channelise inner healing by detoxing free radicals, Bangalore-based Bhookha Haathi is utilising artificial intelligence and machine learning to fulfil its dream of a tobacco-free India. Food and health-based startup, Bhookha Haathi was founded by Abhimanyu Rishi and Kusum Bhandari who are constantly looking for innovative ways to help them make healthier food choices through their alternative food products. The company was founded on the vision to create products that help solve the widespread problem of tobacco chewing that continues to be a prevalent vice in modern India. Bhookha Haathi deploys proprietary AI software which provides subscription-based personalised health solutions at an affordable cost to consumers who wish to substitute or replace their existing dependencies on consuming expensive, processed, strong and chemical-based medications to 100% natural compositions based on dried fruits, nuts, herbs, spices, seeds, grains & honey. Bhandari and Rishi claim that these personalised health solutions will not only cure many long term and hereditary diseases but at the same time, it also resolves many nutritional deficiencies and with no side effects at all. Bhandari and Rishi said, “We have seen growth to 10x since inception with around ₹3 Crore as revenues. We made revenues of ₹30 lakh in FY 17-18, ₹1.34 Crore in FY 18-19 and 1 Crore till Q2 of FY 19-20. We are a profit-making company with gross profits ranging anywhere between 13% to 17% y.o.y. We target to close this financial year between ₹2.75 Cr to 3 Cr.” Flagship Product Bhookha Haathi Replace is the flagship product which mainly replaces nutrient-deficient ingredients to a healthier one. It includes a range of dry-fruit and nut-based health boosters which are provided as an alternative to unnatural protein, vitamin, etc., a range of speciality honey which acts as a replacement for the artificial sugars. Use of AI\/ML The AI software provides subscription-based personalised health solutions to the consumers who wish to replace their existing dependencies on consuming expensive, processed, strong and chemical-based medications. On the AI technology part, the company is still developing and analysing various processes. Currently, the company is using a mix of techniques from search, knowledge, and abstraction to Artificial Neural Network (ANN) intensively which helps in identifying patterns from the data and learn form itself along with some parts of Markov Decision Process (MDP) for transitional probabilities. Also, while performing multiple tasks such as diagnosis processes, treatment protocol development, nutraceutical medicine development, personalised medicine, etc., the AI system majorly helps in grouping patients with similar profiles in order to pre-empt the nature of disease which can arise due to their current health conditions and other external factors to recommend personalised preventive alternative food products. Core Tech Stack The core solution stacks include LAMP i.e. Linux, Apache, MySQL or MariaDB, Perl, PHP, or Python. The frontend and backend comprise multiple modules which have been developed internally, open-sourced and API driven. The frontend is mainly developed with HTML, CSS, and JavaScript while the backend includes a mix of programming languages such as Java and .NET as well as databases and other robust platforms like MySQL and Amazon AWS. Hiring Phase At Bhookha Haathi, the candidates are usually hired from networks and technology-driven hiring platforms like LinkedIn, Naukri, etc. as well as employee referral programs. Besides this, the company follows a 7 step quality check while hiring employees which are The candidate must be engaged with the company’s visionConstant eager to learnSkilled at multitaskingShould have a long term potentialAmbitiousA good team playerCustomer-friendly Roadmap The future roadmap of Bhookha Haathi is focussed on automating AI technology to an API based AI technology where other partners can lease the proprietary solutions for a minimal cost to enhance their consumer health and experiences.","excerpt":"India is home to 4,892 startups in the health-tech space, and they have raised a total of $504 million between 2014-2018. Due to the lack of expert healthcare professionals, the healthcare industry offers a huge opportunity for AI to enable faster and efficient healthcare delivery.  The healthcare industry in India is said to be one […]","categories":["AI Startups"],"tags":["Startups"],"author_name":"Ambika Choudhury","publish_date":"2019-12-27T12:01:00","publication_year":"2019","word_count":754,"keywords":["machine learning","artificial intelligence","AWS","AI","neural network","ML","Python","Aim","SQL","Startups","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","neural network","Aim","AWS","Python","R","SQL"],"url":"https:\/\/analyticsindiamag.com\/ai-startups\/this-bootstrapped-nutraceutical-tech-startup-is-providing-ai-based-solutions-to-help-curb-tobacco-addiction\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10020522,"title":"Top 10 Free Resources To Learn R","content":"R is a popular statistical language that provides high-level graphics for visualisations, debugging facilities, interfaces to other languages and more. According to Data Science Skills Study report 2020, languages like Python and R dominated the preference scale, with a combined value of 81.9% utilisation for statistical modelling. Here is a list of the top ten free resources to learn R. 1| Data Science: R Basics About: Data Science: R Basics helps understand the basics of R programming in data science. This course will help you learn the functions and data types of the R language, understand how to operate on vectors and when to use advanced functions like sorting etc. You will also learn how to apply general programming features like “if-else,” and “for loop” commands, and how to wrangle, analyse and visualise data. Through this course, you will be able to develop a skill set comprising R programming, data wrangling with dplyr, data visualisation with ggplot2, file organisation with UNIX\/Linux, version control with git and GitHub, and reproducible document preparation with RStudio. Know more here. 2| Learn Data Science With R Part 1 of 10 About: “Learn Data Science With R Part 1 of 10” will help you learn the basics of data science, data types, vectors, factors, list matrices, data frames, among others. You will also understand how to read data from Oracle database using RJDBC, RODBC and ROracle. Know more here. 3| Data Science, Machine Learning, Data Analysis, Python & R About: In this course, you will learn new skills and understand the challenging yet lucrative sub-fields of data science including statistics, probability, data analysis, etc. The course includes practical exercises based on real-life examples. It includes both Python and R code templates which you can download and use for your projects. The topics covered include data visualisation using R Programming, data analysis using NumPy and Pandas, etc. Know more here. 4| R for Data Science About: This book by Hadley Wickham and Garrett Grolemund will help you understand how to do data science with R. You will learn how to get your data into the R language, how to transform it, visualise as well as model it. You will also learn how to utilise the grammar of graphics, literate programming, etc. Know more here. 5| R Programming for Data Science About: R Programming for Data Science is a book written by Roger D. Peng. The e-book will teach you the fundamentals of the R programming language. You will also learn the basics of the R language and the steps to manipulate datasets. The ebook will also assist you in understanding how to write various R functions, how to debug and optimise code, etc. With the fundamentals provided in this book, you will have a solid foundation on to build your data science toolbox. Know more here. 6| Hands-On Machine Learning With R About: The book “Hands-On Machine Learning with R” provides modules for many common machine learning methods, such as generalised low-rank models, clustering algorithms, regularised models, gradient boosting machines, among others. You will learn how to build as well as tune various models with R packages. This ebook is meant for users who  wish to learn how to use the machine learning stack within R. It also includes examples using various R packages such as glmnet, XGBoost, lime to effectively model and gain insight from data. Know more here. 7| Data Science: Foundations Using R Specialisation About: This covers foundational data science tools and techniques, including getting, cleaning, and exploring data, programming in R, and conducting reproducible research. The students will complete a project at the end of each course in this specialisation. Projects include installing tools, programming in R, cleaning data, performing analyses, as well as peer review assignments. Know more here. 8| Statistics with R Specialisation About: You will learn to analyse and visualise data in R language as well as create reproducible data analysis reports, demonstrate a conceptual understanding of the unified nature of statistical inference, among others. The course will also help you learn modelling to understand natural phenomena and make data-based decisions, communicate statistical results correctly, in context without relying on statistical jargon, critique data-based claims and evaluate data-based decisions and more. Know more here. 9| Introduction to Data Science: Data Analysis and Prediction Algorithms with R About: The free ebook, written by Rafael A. Irizarry, introduces concepts and skills that can help you tackle real-world data analysis challenges. It covers various concepts from statistical inference, linear regression, probability, and machine learning. It also helps you in developing skills such as data wrangling with dplyr, algorithm building with a caret, data visualisation with ggplot2, version control with Git and GitHub, among others. Know more here. 10| Data Visualisation & Dashboarding with R Specialisation About: This Specialisation is intended for people seeking to develop the ability to visualise data using R. The Specialisation includes five courses, which will help you learn R to create static and interactive data visualisations and publish them on the web. The course will also help you prepare to extract insights into various types of audiences. Know more here.","excerpt":"R is a popular statistical language that provides high-level graphics for visualisations, debugging facilities, interfaces to other languages and more. According to Data Science Skills Study report 2020, languages like Python and R dominated the preference scale, with a combined value of 81.9% utilisation for statistical modelling. Here is a list of the top ten […]","categories":["AI Trends"],"tags":[],"author_name":"Ambika Choudhury","publish_date":"2021-02-21T13:00:00","publication_year":"2021","word_count":852,"keywords":["data science","NumPy","Go","machine learning","AI","Python","Aim","XGBoost","R","Pandas"],"extracted_tech_keywords":["AI","machine learning","data science","Aim","XGBoost","Pandas","NumPy","Python","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-10-free-resources-to-learn-r\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10068335,"title":"The saga of traditional IT institutes in the age of EdTech","content":"In the early 1980s to 2010s era, when the online edtech platforms had not found their footing yet, there were brick and mortar training institutes like NIIT and Aptech. They truly flourished when the field of computer and information technology was beginning to emerge in India, taking up the mantle to train fresh graduates on the skills required for these jobs. Every major city and town had these swanky training institutes where students could train and get certified in a range of computer skills. Cut to 2022, post the pandemic, where education has taken an overhaul, and there has been an overdependence on digital means of learning; where does the future of these institutes lie? NIIT Founded in 1981 by IIT Delhi graduates Rajendra S Pawar and Vijay K Thadani, NIIT tapped on the IT boom of that era. This chain of institutes was the first of its kind in the IT education sector to have conceived the franchising model, setting up nine centres by 1987. NIIT’s model and target audience kept evolving – they also started targeting the elderly and housewives. They also started training banking and retail professionals in IT skills. Taking inspiration from medical education that combines academic training with simultaneous practical training, NIIT decided to branch into offering software services, helping students gain practical knowledge as well. In 2003, reports started floating about a possible demerger of NIIT to spin off a separate company for its global software solutions business. By the end of the year, company representatives officially corroborated the news and the following year, the NIIT Technologies Limited (NTL) was formed. The global education, learning and knowledge solutions were retained in the parent company NIIT. Eventually, in 2019, NTL was sold to Barings Private Equity for Rs 2,627 crore, and NIIT stuck back to education. This year, NIIT announced that it would further demerge its two business verticals – Corporate Learning Group (CLG) and Skills and Careers (SNC) as two separate units; both would be listed on the stock market. Explaining this split, the chairman said that NIIT’s services could be easily demarcated into two parts, which are interlinked but different. One is the Corporate Learning Group (CLG), which accounts for up to 85 per cent of the company’s revenues; the other is the Skills and Careers (SNC) business. Pawar said in an interview, “We felt there were two trajectories and two kinds of things going on with the brand NIIT. We felt the geographies of the two businesses were different, the business strategies different, and the customer profiles different.” That said, the content creation, testing engine, learning engine, and platforms were common; they were used differently. Now, CLG acts as a B2B business and is focused on North America and the EU, and has Fortune 1000 companies as its target companies. On the other hand, SNC has both B2C and B2B customers and is focused on India, China, and Africa. Aptech Another major non-vocational training institute that mushroomed around the education hubs of almost all major cities in India was Aptech. Founded in 1986 and incorporated in 2000 as Aptech Training Ltd, the company offers education and training services to retail and non-retail customers across the world. In 2002, Aptech was rechristened Aptech Ltd. Aptech reported a profit of over 23 crores in the quarter that ended March 2022 against the net loss of 12.12 crore during the previous quarter that ended March 2021. On the downside, in 2021, Aptech had to shut down its operations in Uganda after 20 years of existence. The decision to dissolve the institution was passed by the governing body and endorsed by Uganda’s National Council for Higher Education. Aptech also faced a major blow in Vietnam, one of its largest markets, in the mid of 2021. The growth rate came down from 50 per cent to almost 5 per cent. Traditionally a face-to-face physical training company, during the pandemic, Aptech decided to ‘reinvent’ itself and started a completely online remote delivery model. As per the company’s representatives, the target would remain first-time job seekers. The company spent 2.5 crores in building the platform; the company would be further investing up to 20 crores for creating content and 15 crores towards growing this business in the next 2-3 years. During the lockdown, Aptech’s online education platform delivered more than three million hours of learning; this included 600,000 live hours of tutor-led training – acquiring 19,000 students in India and 32,000 students worldwide. The company expects it to reach 20 per cent of its total revenues by 2023, Aptech head Anil Pant said in an interview. Pant said that the advantage that Aptech has over the Byjus’ and Vedantus of the world is that they just occupy the K-12 and hobby coding space, while Aptech maintains an eagle-like focus on first-time job seekers. Other major players JetKing Institutes was in the news recently for launching a collection of 10,000 Web 3.0 Lion NFTs on the Ethereum blockchain – Jetking Gold collection was priced at Rs 9,000 per NFT, and the JetKing Premium collection was priced at Rs 40,000. The origin story of Jetking is very interesting. Jetking was founded in 1947 by Late GP Bharwani. A freedom fighter, he started an agency called the Nava Bharath Radio Agencies, which dealt with surpluses like radios, gramophones, and other transmissions, which were in abundance after World War II. He then started importing these surpluses as Do it Yourself kits, where the assembling has to be done by the purchaser. It was then that the idea for training students in hardware was formed. In its current form, Jetking is a computer networking institute. The institute has 100 centres across India, a majority of which are franchise-run. In 2012, Suresh G. Bharwani, Chairman and managing director of JetKings, said that they developed a cloud-based digital platform for making digital lectures, practicals and tutorials. Started in 2002, Hughes Global Education Initiative – a subsidiary of Hughes Networks that offers high-speed satellite internet services – has been another popular onsite training institute. Hughes Education offers a real-time interactive onsite learning platform. In the recent past, Hughes Education partnered with IIM Indore and IIM Calcutta to offer programmes to working executives. Hughes Education and MDI Gurgoan also tied up to launch online development programmes for working professionals. Most of these physical institutes have managed to survive because of diversifying to other services or moving to digital platforms. At a time when online edtech platforms are raking big money, the survival of these institutions poses a big question.","excerpt":"Post the pandemic, where education has taken an overhaul, and there has been an overdependence on digital means of learning; where does the future of these institutes lie?","categories":["IT Services"],"tags":["edtech","online education"],"author_name":"Shraddha Goled","publish_date":"2022-06-03T16:00:00","publication_year":"2022","word_count":1091,"keywords":["Go","ELT","programming_languages:R","AI","edtech","Git","programming_languages:Go","CuPy","online education","GAN","R"],"extracted_tech_keywords":["AI","CuPy","R","Go","Git","ELT","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/the-saga-of-traditional-it-institutes-in-the-age-of-edtech\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10011830,"title":"How Is Ethical AI Different From Fair AI","content":"One of the biggest challenges that AI systems face is in regard to its ethics and fairness in its operations. The best ways to demonstrate this would be the example of the secret AI tool that was used for recruitment purposes in e-commerce giant Amazon in 2014. Only a year later, the organisation realised that the AI system was partial towards male candidates since it was trained to vet applications by observing patterns in resumes submitted to the company over ten years; most of these applications were from men. Case of missed opportunity. In order to understand these challenges, it is first necessary to differentiate between two aspects — ethics and fairness. Ethical AI and fair AI are often used interchangeably, but there are few differences. Ethical AI vs Fair AI The concept of machine morality, especially in the case of AI, has been explored by computer scientists since the late 1970s. These research are mainly aimed at addressing the ethical concerns that people may have about the design and application of AI systems. To formally define, at the core of ethical AI, the idea is that it should never lead to rash actions, the result of poor learning, that could impact human safety and dignity. Following are some of the main and strong suites of an Ethical AI system, which are accepted and prescribed prominent field players such as Microsoft. Technical robustness, reliability, and safety: It is important to build robust AI systems that are resilient to adversarial attacks. Such attacks manipulate the behaviour of the system by making changes to the input or training data. In worse case scenarios, these attacks can prove fatal to the environment they are in. Additionally, an ethical AI system should also be able to fallback from a ‘rule-based system’ and ask for human intervention to prevent it from going rogue. Privacy and security: An ethical AI system must guarantee privacy and data protection throughout its lifecycle, which includes the information provided initially by the user and that generated during the course of interaction with the system. This is quite a slippery slope. Since these systems primarily rely on data, they are always hungry for new information. There have been multiple reports of tech giants, intentionally or otherwise, tapping into users’ sensitive information. Transparency: The guidelines from the European Commission released in 2019, defined AI transparency in three subparts: traceability, explainability, and communication. Vendors must make the decision-making capabilities of the AI device transparent to the users to protect against any possible harm against humans or their rights. Fairness and inclusivity: Bias is one of the major problems with AI systems. These systems internalise the choice of the researchers, building them and further amplify them. Experts believe that to build a system completely devoid of such bias is impossible. However, there are a few steps that could be taken to minimise them, including using inclusive datasets to train these machines on. Accountability: An ethical AI has mechanisms that ensure responsibility and accountability, not just during its creation but also after development, deployment and use. Companies must adhere to rules and regulations to make sure that their systems conform to ethical principles. Having seen what an ethical AI system means, it is easy to infer that fairness is prominent yet just a part of it. So when we speak of a fair AI, we refer to an attribute of ethical AI in the larger sense. To define, fair AI refers to probabilistic decision support that prevents it from unfairly harming or benefiting a particular. There are multiple reasons as to why ‘unfairness’ creeps into a system: the data the system learns from, the way algorithms are designed, and modelling by way of selecting relevant features as inputs and combining them in meaningful ways. As mentioned earlier in the article, it is impossible to construct a 100% universally fair or unbiased system. Partly because there are up to 20 different mathematical definitions of fairness, however, organisations can design AI systems to meet specific goals, thus mitigating the unfairness and creating a more responsible system overall. Wrapping Up There is a very fine line between ethical AI and a fair AI. It becomes difficult to differentiate, as they also overlap at a few points. Companies need to realise the difference between the two to develop a system that best suits their operation and creates an overall responsible AI system.","excerpt":"One of the biggest challenges that AI systems face is in regard to its ethics and fairness in its operations. The best ways to demonstrate this would be the example of the secret AI tool that was used for recruitment purposes in e-commerce giant Amazon in 2014. Only a year later, the organisation realised that […]","categories":["AI Features"],"tags":["Ethical AI","Fair AI"],"author_name":"Shraddha Goled","publish_date":"2020-11-18T17:00:08","publication_year":"2020","word_count":733,"keywords":["Go","programming_languages:R","AI","Ethical AI","programming_languages:Go","responsible AI","Aim","ViT","GAN","Fair AI","R","adversarial attacks"],"extracted_tech_keywords":["AI","Aim","R","Go","GAN","ViT","responsible AI","adversarial attacks","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-is-ethical-ai-different-from-fair-ai\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":46424,"title":"Darwinbox Raises $15 Million In New Round Of Funding","content":"Darwinbox , India’s fastest growing enterprise Human Resource Management (HRMS) platform, has recently concluded a new round of funding, closing at a whopping $15 million. This round was led by Sequoia India and saw participation from existing investors Lightspeed India Partners, Endiya Partners, and 3one4 Capital. With this round of funding, Darwinbox, now plans to expands across India and into Southeast Asia and builds out an ecosystem of integrated third-party solutions providers on its platform. Rohit Chennamaneni, Co-founder, Darwinbox, said in a statement, “Asia HR technology market is estimated to be at $9 billion. Also, the increase in mobile users in geography meant newer possibilities in the way organisations can engage and empower their employees. Enterprises in Asia are migrating to the cloud at a much faster pace than the entire world and we are excited to lead the charge with cutting edge HR technology that is tailored to the Asian context.” “We led the Series A financing in May 2017 when Darwinbox had about 50,000 employees on their platform; the platform now has 500,000 employees on it. Kudos to the founders Rohit, Jayant and Chaitanya as well as all their team members at Darwinbox,” said Dev Khare, partner at Lightspeed. Darwinbox’s cloud-based end-to-end HRMS platform automates HR processes across the entire employee lifecycle – managing recruitment, onboarding, core HR, time and attendance, expense management, payroll, employee engagement, performance management and people analytics. Their easy-to-use interface with a high degree of configurability to suit local enterprise context has helped them acquire a large customer base within a short time.","excerpt":"Darwinbox , India’s fastest growing enterprise Human Resource Management (HRMS) platform, has recently concluded a new round of funding, closing at a whopping $15 million. This round was led by Sequoia India and saw participation from existing investors Lightspeed India Partners, Endiya Partners, and 3one4 Capital. With this round of funding, Darwinbox, now plans to […]","categories":["AI News"],"tags":[],"author_name":"Prajakta Hebbar","publish_date":"2019-09-26T13:05:41","publication_year":"2019","word_count":259,"keywords":["API","funding","programming_languages:R","AI","analytics","GAN","R"],"extracted_tech_keywords":["AI","analytics","R","API","GAN","funding","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/darwinbox-raises-15-million-in-new-round-of-funding\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10109649,"title":"How Tata Technologies is Helping Automobile Companies Build AI-Powered Cars","content":"Tata Technologies made a bumper debut this year on Indian stock markets, with share price jumping 180% over the Initial Public Offering (IPO) price within minutes. Its stellar debut is backed by its strong emphasis on technology innovation and a commitment to delivering cutting-edge solutions in the rapidly evolving landscape. The company, which started as an automotive design unit of Tata Motors in 1989, has emerged over the years as a top engineering, research, and development (ER&D) company in the world. Currently, the company is working with automakers across the globe to develop connected car platforms powered by AI to provide drivers with real-time information about traffic, weather, and road conditions. Recently, AIM caught up with Sriram Lakshminarayan, President and Chief Technical Officer at Tata Technologies, who believes generative AI will transform the automobile industry. “AI holds considerable importance in the industries we engage with, specifically in automotive, aerospace, and industrial heavy machinery. Across these industries, from the initial engineering stages to the manufacturing process and subsequent aftermarket services, there is a notable integration of generative AI and other cutting-edge technologies,” Lakshminarayan said. AI revolutionise the Automobile industry AI is revolutionising Advanced Driver Assistance Systems (ADAS) by enabling real-time analysis of complex driving scenarios. AI algorithms enhance object detection, lane-keeping, and adaptive cruise control, contributing to safer and more efficient driving. “Cruise control was once regarded as a feature exclusive to high-end cars, but it has now evolved into a common feature across various vehicle models. Similarly, certain ADAS solutions are poised to become standard features in every vehicle, regardless of its price point. This shift signifies a broader trend toward the integration of advanced technologies into the mainstream automotive market,” he said. In recent times, AI-driven ADAS technologies have increasingly become pivotal in mitigating risks, reducing accidents, and advancing the evolution of autonomous vehicles, which are already running in the streets of many Western countries. Moreover, technology, particularly AI, will significantly influence how we purchase cars. Consumers will increasingly choose vehicles based on their AI-powered features, reflecting a growing reliance on advanced technologies in the decision-making process. “The role of dealerships will also be redefined as things become more digital. The shift towards a more digital approach is becoming increasingly prevalent, especially in tech-savvy countries like India. Consumers here are quick to adopt technology, making digital platforms a central aspect of the evolving automotive purchasing landscape.” Building a Software Defined Vehicle Modern cars have evolved significantly in the last few years and we have seen a software-oriented approach replacing the traditional hardware-centric approach in automobiles. Earlier this year, Tata Technologies signed a Memorandum of Understanding (MoU) with TiHAN IIT Hyderabad to develop Software Defined Vehicles (SDV) and ADAS. “In the context of SDV, it is essential to consider four key functions. Firstly, there are cockpit solutions encompassing the entire vertical cockpit experience.” He believes how we interact with a car will change from an infotainment perspective. Today’s cars often feature personalised recognition, adjusting settings like mirrors based on user preferences. They can even greet users personally. “The next step is intuitive communication; for instance, if you plan to pick up your son from school and he texts to say he’ll be delayed, the system goes beyond merely notifying you. Instead, it interprets the message, informs you of the delay, and calculates the resulting availability, providing a seamless experience.” “Secondly, there is the vehicle compute aspect, managing computations related to functions like braking and ensuring the vehicle stops appropriately.” Cars equipped with AI on the edge can perform complex computations onboard, contributing to a safer, more responsive, and personalised driving experience. However, for this, Lakshminarayan believes there is a need for specialised chips which allow AI to run on the edge in cars. “Thirdly, the inclusion of ADAS becomes crucial and lastly, cybersecurity plays a pivotal role, given the software-centric nature of these functions.” How Tata Technologies is driving the transformation Tata Technologies boasts Tata Motors, a global automotive giant, among its clients. The company’s extensive clientele comprises more than 35 Original Equipment Manufacturers (OEMs), featuring names such as McLaren, VinFast, and Honda. Furthermore, Tata Technologies is actively collaborating with seven of the top 10 automotive Engineering, Research, and Development (ER&D) spenders and five of the foremost new energy ER&D spenders in 2022. “We assist customers in transforming their concepts from the initial design phase on paper to the complete end-to-end aftermarket implementation. While we don’t engage in manufacturing ourselves, our role involves leveraging technology and applying our expertise to bring these products to life for our clients.” “The core focus of our operations revolves around New Product Introduction (NPI). We assist customers in transforming their concepts from the initial design phase on paper to the complete end-to-end aftermarket implementation.” Automobile companies across the globe want to leverage the power of AI to enhance vehicle performance, improve safety features, optimise fuel efficiency, and provide innovative and personalised experiences for drivers and passengers. To achieve this, OEMs actively seek innovative solutions and accelerators that aid in minimising both technology incubation time and costs. Lakshminarayan believes Tata Technologies possesses the industry knowledge and the necessary technological expertise that an OEM would seek to integrate AI seamlessly into practical use cases. At the recently held International Congress for Automotive Electronics (ELIV) event in Bonn, Germany, Tata Technologies showcased three distinct solutions. “We showcased Cockpit Solutions in collaboration with Intel. Our demonstration of ADAS solutions featured a combination of Qualcomm and AWS. For vehicle computation, we presented a solution utilising NXP and ARM in conjunction with AWS on the cloud.” The reason for this demonstration, according to Lakshminarayan, was to highlight the flexibility in choosing different best-of-breed solutions for specific functions or opting for a unified approach. As a system integrator, the company emphasises bringing all these diverse components together seamlessly. Enabling autonomous Automotive companies are striving to develop SDVs with autonomous technologies. “What we are currently focusing on involves significant enablement for autonomous systems, particularly in areas such as image processing and annotations.” The company actively collaborates with its customers on use cases that require detailed image analysis, considering the diverse elements encountered on roads in countries like India. This involves developing AI models capable of predicting and responding to such scenarios. “Autonomous systems involve processing terabytes of data hourly, requiring advanced analytics and annotations. Our collaboration with customers focuses on developing image processing algorithms, facilitating their journey toward autonomy. While our role doesn’t involve directly creating, for instance, L2 autonomous solutions, we play a vital role in enabling our customers to progress in this transformative journey.” Lakshminarayan anticipates an increasing array of use cases emerging in the Indian market. “Predicting for India, I foresee substantial growth in autonomous applications. While different levels of autonomy exist, in the Indian context, I project us reaching around L2 plus at most, considering the unique dynamics and challenges in the region.”","excerpt":"The company helps customers in transforming their concepts from the initial design phase on paper to the complete end-to-end aftermarket implementation","categories":["AI Features"],"tags":["AI Companies","Interviews and Discussions","Tata technologies"],"author_name":"Pritam Bordoloi","publish_date":"2023-12-29T14:00:00","publication_year":"2023","word_count":1145,"keywords":["Tata technologies","AWS","AI","ML","RAG","Ray","Aim","object detection","analytics","AI Companies","generative AI","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","ML","analytics","generative AI","Aim","Ray","RAG","object detection","AWS","R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-tata-technologies-is-helping-automobile-companies-build-ai-powered-cars\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":11983,"title":"Myntra shines the spotlight on AI to drive growth","content":"Myntra, owned by ecommerce giant Flipkart is betting big on AI Bangalore-based fashion portal Myntra, which was acquired by ecommerce company Flipkart is reportedly going to use Artificial Intelligence (AI) to improve the customer experience. In an attempt to drive growth and boost profit, Myntra has trained its eyes on AI and ML to deliver personalized customer experiences. According to reports, Myntra is using customer data to curate lines based on current fashion trends and make a one-of-a-kind personalized store experience. Reports suggest that Myntra will also launch an app based chat service intended to help customers interact with their favourite brands. The investment in new technology is also an attempt to boost sales.  Flipkart already has a strong base in fashion, with Jabong being bought over by Myntra for earlier in the year.  CEO of Myntra and Jabong, Ananth Narayanan, cited in a recent interview that fashion and lifestyle are the most exciting segment in e-commerce right now. “And undoubtedly, Myntra will continue to create more excitement for its consumers through its selection, service and engagement,” said Ananth in a recent interview. With fashion and lifestyle being the biggest drivers of growth in ecommerce in India, Flipkart strengthened its fashion segment with major acquisition to maintain its dominance. For ecommerce giants Amazon and Flipkart, fashion is the next frontier for battle for revenue generation with both companies betting big to gain market share in the fashion segment","excerpt":"Bangalore-based fashion portal Myntra, which was acquired by ecommerce company Flipkart is reportedly going to use Artificial Intelligence (AI) to improve the customer experience. In an attempt to drive growth and boost profit, Myntra has trained its eyes on AI and ML to deliver personalized customer experiences. According to reports, Myntra is using customer data […]","categories":["AI News"],"tags":[],"author_name":"Richa Bhatia","publish_date":"2016-12-29T11:41:04","publication_year":"2016","word_count":238,"keywords":["Go","artificial intelligence","programming_languages:R","AI","ML","programming_languages:Go","Ray","R"],"extracted_tech_keywords":["AI","artificial intelligence","ML","Ray","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/myntra-shines-spotlight-ai-drive-growth\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":19497,"title":"Google&#8217;s Artificial Intelligence Successfully Spawns An AI &#8216;Child&#8217; That Easily Outperforms Man-Made AI","content":"Many may look at the end of 2017 as a defining moment in human history as the year when an Artificially Intelligent system created its own “child” — another AI capable of a specific task. A few weeks ago, Google’s AutoML — an artificial intelligence programme created to build artificial intelligence programmes — spawned a “child” using its reinforcement learning technique. This works like Machine Learning, except it is entirely automated where AutoML (the “parent”) acts as the neural network for its task-driven AI child. Named NASNet, the AI child then was tasked to recognise objects which included people, cars, traffic lights, handbags, backpacks, among others, in a video in real-time. In fact, NASNet did really well in two well-known image classification tests. On ImageNet image classification, NASNet achieved a prediction accuracy of 82.7 percent on the validation set, surpassing all previous Inception models that Google built. Additionally, NASNet performed 1.2 percent better than all previous published results and is on par with the best unpublished result so far. “NASNet may be resized to produce a family of models that achieve good accuracies while having very low computational costs. For example, a small version of NASNet achieves 74 percent accuracy, which is 3.1 percent better than equivalently-sized, state-of-the-art models for mobile platforms. The large NASNet achieves state-of-the-art accuracy while halving the computational cost,” said an official Google statement. Google brain researchers wrote in their blog post, “We hope that the larger machine learning community will be able to build on these models to address multitudes of computer vision problems we have not yet imagined.” This new development has again brought the much-argued issue to the forefront. Does artificial intelligence — as well as artificial intelligence that can create artificial intelligence — spell mankind’s doom? While many noted names such as physicist Stephen Hawking, Tesla and SpaceX’s CEO Elon Musk, and most recently US former Secretary of State Hillary Clinton, have already sounded alarm bells with regards to Artificial Intelligence and its potential harm, there are others who don’t agree. Facebook’s Mark Zuckerberg, Microsoft founder Bill Gates and futurist Ray Kurzweil are some of the famous names who believe that AI can actually do more good to the humans than harm. Kurzweil, who currently runs a group at Google writing automatic responses to users’ emails in cooperation with the Gmail team, had once famously said, “It was fire that kept us warm, cooked our food, but also burnt our houses down. Technology is always a double-edged sword.”","excerpt":"Many may look at the end of 2017 as a defining moment in human history as the year when an Artificially Intelligent system created its own “child” — another AI capable of a specific task. A few weeks ago, Google’s AutoML — an artificial intelligence programme created to build artificial intelligence programmes — spawned a […]","categories":["AI News"],"tags":["AI (Artificial Intelligence)","Google","Google Brain","Machine Learning"],"author_name":"Prajakta Hebbar","publish_date":"2017-12-05T07:08:39","publication_year":"2017","word_count":416,"keywords":["Go","artificial intelligence","machine learning","AI","neural network","RPA","ML","Google Brain","Machine Learning","computer vision","Ray","Google","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","neural network","computer vision","Ray","R","Go","RPA"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/google-ai-child-nasnet-automl\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10143460,"title":"OpenAI Introduces Santa Mode to do What o1 Couldn’t","content":"On the sixth day of ‘12 Days of OpenAI,’ OpenAI announced a major update: video and screen-sharing capabilities in ChatGPT’s advanced voice mode. The feature allows users to engage in real-time visual and audio interactions with ChatGPT, making conversations more interactive and functional. Starting today, the rollout will gradually reach all users, including Plus and Pro subscribers, with full availability expected early next year for enterprise and educational plans. The announcement came after OpenAI acknowledged a brief outage the previous day, promising detailed follow-ups while swiftly resuming operations. “We know you depend on us, and we take this very seriously,” said Kevin Weil, OpenAI’s product lead. A More Interactive ChatGPT The newly introduced features let users share their screens and video feed while engaging in natural conversations with ChatGPT. Users can troubleshoot issues, learn new skills, or collaborate on tasks with real-time visual support. Demonstrations during the livestream included making pour-over coffee with ChatGPT providing step-by-step guidance via video. Rowan Zellers, a researcher at OpenAI, explained, “Now you can chat with ChatGPT over video and voice together in real-time.” The feature also includes screen-sharing capabilities, allowing users to get instant help for tasks such as crafting responses, troubleshooting code, or reviewing presentations. Interestingly, this is quite similar to Microsoft Copilot Vision, Anthropic Computer Use, Google’s Mariner. A Festive Twist: Santa Joins ChatGPT Adding a touch of holiday fun, OpenAI introduced a Santa Mode within ChatGPT. Users can now interact with Santa in real-time, asking him about his favorite reindeer, Christmas traditions, or even how to maintain a fluffy beard. The feature, accessible via a snowflake icon, resets advanced voice usage limits for the month, ensuring everyone gets a chance to experience Santa’s jolly presence. During the demo, Santa quipped, “My favorite tradition is when children hang their stockings by the fireplace—it’s a moment filled with magic and anticipation.” The lighthearted addition brings a festive spirit to ChatGPT’s growing functionality. Santa to Drive Subscriptions “I fully expect Santa Mode to drive more subscriptions than o1 and I’m at peace with this,” said Noam Brown, researcher at OpenAI. That makes sense as o1 is currently priced at $200. OpenAI’s announcement comes amid intense competition in the AI space. Google recently released Gemini 2, its advanced multimodal model, while Amazon launched Nova. Addressing the industry buzz, OpenAI co-founder Greg Brockman hinted at bigger updates on the horizon. “AGI is in the air,” he posted cryptically, fueling speculation about upcoming releases like GPT-4.5 or GPT-5. These developments suggest that OpenAI isn’t just playing catch-up—it’s setting the stage for what’s next. What’s Next? With six more days left, the company is clearly ramping up for something significant. The addition of video and screen-sharing to advanced voice mode reflects a broader vision of making AI interactions richer and more human-like. As the race for AI dominance continues, OpenAI’s ability to innovate while maintaining a festive, community-focused approach keeps it firmly in the spotlight. Whether through Santa’s jolly voice or video-guided coffee tutorials, OpenAI is redefining how we interact with AI—one update at a time.","excerpt":"‘I fully expect Santa Mode to drive more subscriptions than o1 and I’m at peace with this,’ says Noam Brown, researcher at OpenAI.","categories":["AI Features"],"tags":["OpenAI"],"author_name":"Aditi Suresh","publish_date":"2024-12-13T01:10:23","publication_year":"2024","word_count":508,"keywords":["Anthropic","ChatGPT","Go","GPT-5","OpenAI","AI","ML","GPT-4.5","R","Snowflake"],"extracted_tech_keywords":["AI","ML","GPT-4.5","GPT-5","ChatGPT","OpenAI","Anthropic","Snowflake","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/openai-introduces-santa-mode-to-do-what-o1-couldnt\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10143579,"title":"What’s the Moat of Neysa in the World of Yotta?","content":"It is not easy to build an AI startup in India. While the time is ripe now, getting the funds to build a product for the global market or even a customised solution for the Indian market is extremely difficult. Especially if one is working on the hardware and infrastructure layer, the issue gets maximised. Contrastingly, this is not something that Neysa faced. Founded by Sharad Sanghi and Anindya Das, Neysa, a Mumbai-based AI acceleration cloud system provider, bagged $20 million in seed funding in April 2024. Cut to October, the team raised Series-A of $30 million, led by NTTVC, Z47, and Nexus Venture Partners. Investors value the expertise that comes with an experienced founder, and Sanghi exemplified this with his leadership. Neysa is his second venture in the computing space after Netmagic, one of India’s first IT infrastructure startups, which he founded in 1998. NTT later acquired the company, and Sanghi continued working there as the CEO. Das is currently the chief technology officer, and Karan Kirpalani is the chief product officer. Both, part of the core team members and friends of the founders over the last two decades, discussed with AIM the moat of the company in a world that is filled with data centres, AI infrastructure providers, and Indian AI startups. With an AI-first philosophy, Kirpalani said that Neya is redefining the landscape of AI-driven solutions with its three flagship offerings, Velocis, Overwatch, and Aegis. However, the most important one right now is Velocis. What’s the offering? The Neysa Velocis AI Acceleration Cloud system is designed to support the entire AI lifecycle—from training and fine-tuning models to inferencing. “Velocis is built as an end-to-end cloud platform that covers everything from data ingestion to inferencing,” said Kirpalani, highlighting their differentiating factor. Kirpalani asserts that Neysa will soon be among the world’s few providers of a complete AI Platform-as-a-Service (PaaS)—from GPU-level infrastructure onward—available both in public cloud and private cluster environments. This comprehensive platform is set to include data ingestion, database capabilities, data engineering, model repositories, MLOps, and inference functionality in its upcoming releases. While Google’s Vertex, Microsoft Azure, or AWS Sagemaker are all in the same space, Kirpalani quotes Neysa’s customers as saying that there is a massive pushback for these providers because of their blackbox proprietary platforms and that they are embracing open-source solutions. This allows clients to integrate any tool of their choice on Neysa’s AI PaaS platform. Velocis embraces an open-source approach, allowing seamless integration with tools like Hugging Face, TensorFlow, and MLFlow. Kirpalani emphasised that the availability of open-source models and the whole pipeline make it stand out from competitors. On the lines of Databricks or Snowflake, Kirpalani said Neysa’s Velocis, too, provides multiple deployment options, including bare-metal servers, virtual machines, and managed services catering to diverse client needs. These services are similar to Databricks and Snowflake hosted on cloud providers, and finding the moat here for Neysa is not clearly visible. Sovereign AI is a Moat for Neysa, But so is for Others in India On the hardware side, Neysa works closely with NVIDIA and provides fractional GPUs for small use cases. This also includes GPUs like L4s, H100s, H200, and several others. “We don’t want to get into the legacy GPUs like A100 that other players are still providing,” Kirpalani said, adding that they don’t want to provide consumer-grade GPUs like RTX and GTX. When asked about Yotta, one of Neysa’s closest competitors and also a close Indian client of NVIDIA, Kirpalani and Das declined to comment. But Kirpalani added that Indian providers are able to leverage existing infrastructure to provide just bare metal services, while Neysa’s focus is to offer an entire cloud stack that no one is building. At the same time, Yotta also offers similar services, from training on bare metal hardware to deploying custom models and inference on its Shakti AI Cloud platform. That said, Kirpalani noted the recent announcement of their Insurance AI Cloud in partnership with Data Science Wizards, offering an end-to-end cloud platform for insurance companies. For one of their clients, Kirpalani said Neysa witnessed an 18% uptake in persistence, which is one of the crucial factors when implementing AI in the insurance sector. In another example, Kirpalani said that a financial services company that was using Neysa’s offering in their call centre was able to demonstrate a 67% TCO reduction, reaching to INR 4.7 per minute. This was using Meta’s Llama 3.1 on the Velocis AI PaaS platform with NVIDIA H100 GPUs. Before, the client was using OpenAI on Microsoft Azure, which cost them INR 15 per minute. Kirpalani also added that Neysa will continue to partner with companies like Data Science Wizards and Fractal to build industry specific solutions like the recently launched Insurance AI Cloud with Data Science Wizards. For the road ahead, Neysa plans to push Aegis, its security platform, which focuses on emerging threats unique to the AI landscape. Aegis provides robust protection measures, ensuring secure AI development and deployment. Neysa plans to expand it into a standalone product by early 2025. In the future, Neysa wants to launch inference-as-a-service by early 2025, further enhancing its AI lifecycle offerings. “We’re accelerating our timelines to meet market demand,” Kirpalani shared.","excerpt":"Sovereign AI is just one moat for Neysa, but the same is for others in India.","categories":["Global Tech"],"tags":["Yotta"],"author_name":"Mohit Pandey","publish_date":"2024-12-15T09:00:00","publication_year":"2024","word_count":869,"keywords":["data science","Hugging Face","OpenAI","AI","Yotta","ML","MLOps","RAG","Aim","MLflow","TensorFlow"],"extracted_tech_keywords":["AI","ML","data science","OpenAI","MLOps","MLflow","Aim","TensorFlow","Hugging Face","RAG"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/whats-the-moat-of-neysa-in-the-world-of-yotta\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10001733,"title":"Here’s How Organisations Can Make Data Centres Future-Proof","content":"In this internet-driven era, the dependency of businesses across the globe on the internet is skyrocketing. Data over the years have evolved tremendously and so does data centres. Data centres play a major role in pushing organisations operations to a higher level — whether its transport, internet, telecommunications, entertainment, hospitals, banks etc., almost every vertical needs data centres. If in case you don’t know what a data centre is, it is a facility that stores, manages, and disseminates data that is produced every single day. It might not be visible to everyone, but a data centre acts an organisations backbone. The more data gets created, the more vital data centres become, however, there lies challenges — today’s most of the data centres are clunky, inefficient, and outdated. And because of this, experts and organisations are becoming sceptical regarding the future of data centres. In this article, we will have a look at factors that can help data centres survive in the future. Businesses are expanding year by year and do their operations, and for that, it is becoming imperative for companies to have data centres that work at the top-notch level for years. In order to put down the scepticism surrounding the future existence of data centres, organisations need to reconsider their data centre trends. Reasonable Power Resources The location of a data centre plays a vital role; a data centre has to be at a place where it gets what it requires. Talking about energy, it is not much affordable to run a data centre that consumes a lot of energy. And for that, organisations should look for a place that has cheap power resources and also clean at the same time. Nordic regions are becoming a go-to place for many organisations to set up data centres considering the region’s renewable energy, power supply, low energy prices, political stability and faster time-to-market primarily due to ease of doing business. Furthermore, according to a source, Investments in Nordic data centres are prophesied to double by 2025. Fibre Connectivity Over the years, data centres have come a long way. They are evolving and becoming more sophisticated, and to stay connected with the rest of the world that is becoming sophisticated too, data centres need to beat the bandwidth issue. In order to solve that, Fiber connectivity has emerged. Compared to copper cables, Fiber optic cables can transfer data faster. Also, one of the best advantages of fibre connectivity is, while copper cable has a lifespan of 60-0 years, Fiber optic cables have almost an infinite lifespan. Software Defined Data Centre (SDDC) Software-defined data centre (SDDC) is a data centre where all infrastructure is virtualized and delivered as a service. But what is the advantage of virtualizing a data centre? The answer to that is things like computing, storage, and networking of a data centre can be abstracted and represented in a software form, which can later be used as a service to be used by others. Software-defined networking (SDN) is a part of an SDDC and through the years, this concept has gained significant traction. SDN is considered to be one of the best ways to manage virtualization, save cost and provide faster service delivery. Edge Computing Having the ability to help enterprises to cope with several challenges, over the past couple of years, Edge computing has become quite popular. It is not only helping enterprises with data flow but also helping them overcome the challenges with data centres. The whole concept of edge computing for data centres is to keep data close to the place where it was originated. How it would help? By analyzing the data where it is created, it would reduce the need to transfer data back and forth between centralized computing locations as it delays the decision-making process. Edge computing is definitely one of the most sought after tech as it already has numerous use cases of dealing challenges with data. Bottom Line Having a data centre has become a must and the world is already moving towards it. However, over the last few years, the world has also witnessed some of the worst data centre fails and outages that have raised numerous questions about its future.  But again, humans are also quick when it comes to adopting new tech. Talking about the data centre, organisations across the world are working relentlessly to make the data centre more efficient for the future.","excerpt":"In this internet-driven era, the dependency of businesses across the globe on the internet is skyrocketing. Data over the years have evolved tremendously and so does data centres. Data centres play a major role in pushing organisations operations to a higher level — whether its transport, internet, telecommunications, entertainment, hospitals, banks etc., almost every vertical […]","categories":["AI News"],"tags":["data centre","edge computing"],"author_name":"Harshajit Sarmah","publish_date":"2019-04-03T12:56:25","publication_year":"2019","word_count":736,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","RAG","GAN","ViT","edge computing","R","data centre"],"extracted_tech_keywords":["AI","RAG","edge computing","R","Go","GAN","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/heres-how-organisations-can-make-data-centres-future-proof\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10131973,"title":"NVIDIA AI Summit India 2024 – 5 Key Things to Expect","content":"NVIDIA has announced its AI Summit 2024, scheduled between October 23 and 25 at the Jio World Convention Centre in Mumbai, India. The summit will feature three days of presentations, hands-on workshops, and networking opportunities aimed at connecting industry experts and exploring advancements in artificial intelligence. “With accelerated computing infrastructure, research and AI skilling at scale, India has the potential to become the intelligence capital of the world. The upcoming NVIDIA AI Summit is the first-of-its-kind event with a significant focus on India,” said Vishal Dhupar, managing director (Asia-South), NVIDIA. “It promises to be topical, relevant for India and is a must-attend for developers, startups and enterprises,” he added. REGISTER NOW 1. Blackwell is Coming to India NVIDIA chief Jensen Huang is set to attend the event and participate in a fireside chat. There’s a strong possibility that Huang will announce the availability of NVIDIA’s latest Blackwell GPUs for the Indian market. According to a report, NVIDIA’s newly announced Blackwell GPUs are expected to begin shipping as early as October. Mumbai-based data center and managed cloud infrastructure provider Yotta Data Services is set to benefit from this early release. “We are in early talks with Nvidia to source Blackwell GPUs as part of our order, and are in the process of finalising all details,” said Sunil Gupta, the co-founder and chief executive of Yotta. “We’re looking at procuring around 1,000 Blackwell B200 GPUs by October, which would be equivalent to around 4,000 ‘H100’ GPUs. While a timeline isn’t clear yet, we’re expecting the delivery of Blackwell GPUs before the end of this year, and complete our full existing order by the next fiscal,” Gupta said. The entire order between Yotta and Nvidia, Gupta said, is worth ~$1 billion. Notably, the production of Blackwell chips has been delayed by three months or more due to design flaws, which could impact customers such as Meta platforms, Google, and Microsoft, who have collectively ordered tens of billions of dollars worth of the chips. Earlier this year, Yotta, an elite partner of NVIDIA, received the first shipment of 4,000 GPUs. Yotta plans to scale up its GPU inventory to 32,768 units by the end of 2025. Last year, the company announced that it would import 24,000 GPUs, including NVIDIA H100s and L40S, in a phased manner. 2. Data Centres Loading During his last visit to India, Huang announced NVIDIA’s partnership with Reliance to develop a foundational large language model tailored to India’s diverse languages and generative AI needs. However, there have been no updates since. It is likely that Reliance will announce something significant during the summit. Reliance is working with NVIDIA to build AI infrastructure that is over an order of magnitude more powerful than the fastest supercomputer in India today. NVIDIA will provide access to its most advanced GH200 Grace Hopper Superchip and NVIDIA DGX™ Cloud, an AI supercomputing service in the cloud. Reliance said that it will create AI applications and services for their 450 million Jio customers and provide energy-efficient AI infrastructure to scientists, developers and startups across India. Similarly, NVIDIA partnered with Tata to build an AI supercomputer powered by the next-generation NVIDIA® GH200 Grace Hopper Superchip. TCS will utilize this AI infrastructure to develop and process generative AI applications. Additionally, TCS announced plans to upskill its 600,000-strong workforce through this partnership. In a recent earnings call, TCS reported having over $1.5 billion in its AI pipeline, encompassing 270 projects. Last year, Infosys expanded its alliance with NVIDIA to train 50,000 employees on NVIDIA’s AI technology, integrating these tools with Infosys Topaz to create generative AI solutions for enterprises. Similarly, Netweb Technologies also partnered with NVIDIA to manufacture NVIDIA’s Grace CPU Superchip and GH200 Grace Hopper Superchip MGX server designs. This partnership supports the Make in India initiative by building a local ecosystem to address demands around AI and accelerated computing applications for both government and private enterprises. REGISTER NOW 3. Partnership with Indian AI Startups It’s highly likely that Indian AI startups will also make their presence felt at the event. Earlier this year, Dhupar said that he found Krutrim, Sarvam AI, and Immersio to be the three ‘most-exciting’ AI startups from India. NVIDIA chief Huang believes the future of AI lies in physical AI. Recently, Bengaluru-based startup Control One launched India’s first physical AI agent for the global market. They released a video showcasing this agent, which responds to voice commands via a unique operating system. Control One has already integrated this OS into a forklift. Control One is also an NVIDIA Inception Partner, which grants the startup access to cutting-edge GPU technology and crucial expertise for developing and scaling its AI systems. Recently, another Indian AI startup, KissanAI, got accepted into the NVIDIA Inception Program. Who knows, NVIDIA might partner with People+ AI’s Open Cloud Compute as well. OCC seeks to create an open network of compute resources, making it easier for businesses, especially startups, to access the compute power they need without being locked into specific cloud providers. Meanwhile, NVIDIA recently partnered with Thapar Institute of Engineering and Technology to advance AI education and research. Through this technical collaboration, the university will offer a formidable infrastructure. Its current 227 petaflops of AI performance will be expanded to over 500 petaflops, tailored for the most demanding AI and deep learning tasks. It might follow suit by partnering with more universities. 4. Made In India PCs Recently, NVIDIA announced a collaboration with six Indian PC builders—The MVP, Ant PC, Vishal Peripherals, Hitech Computer Genesys Labs, XRig, and EliteHubs—to launch Made-in-India PCs equipped with RTX AI. This initiative aims to bring advanced AI technology to Indian gamers, creators, and developers. “Our vision is deeply rooted in the commitment to India’s future in AI and computing. With India’s AI market projected to reach $6 billion by 2027, the opportunity is immense,” said Dhupar. “There is an opportunity for India to be the capital of intelligence. The country has the skill sets and talent that understand how to work with a computer,” he added. The new PCs are part of NVIDIA’s ongoing commitment to gaming and technological advancement. The inclusion of RTX AI technology in these systems offers gamers enhanced performance and visual experiences. 5. Partnership with Indian Govt Given Huang’s previous engagements with Indian leaders, including a meeting with Prime Minister Narendra Modi, his address may also touch upon NVIDIA’s plans for collaboration with India in AI and chip manufacturing. The Indian government is actively supporting AI development through the IndiaAI mission, launched in March 2024, which aims to position India as a global AI leader by investing in infrastructure and supporting startups. The mission includes an INR 10,300 crore investment to expand AI infrastructure and make GPUs more accessible. As many as 10,000 GPUs will be made available to startups and a marketplace will be created to benefit R&D facilities and startups. REGISTER NOW","excerpt":"Blackwell is Coming to India.","categories":["Global Tech"],"tags":["NVIDIA"],"author_name":"Siddharth Jindal","publish_date":"2024-08-09T13:47:09","publication_year":"2024","word_count":1146,"keywords":["Go","API","artificial intelligence","ELT","AWS","AI","Aim","deep learning","generative AI","NVIDIA","R"],"extracted_tech_keywords":["AI","artificial intelligence","deep learning","generative AI","Aim","AWS","R","Go","API","ELT"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/nvidia-ai-summit-india-2024-5-key-things-to-expect\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":6,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":28572,"title":"Paytm To Introduce Facial Recognition Technology To Enable Digital Payments","content":"Image Source: Zenus Facial Recognition Seems like Apple’s iPhoneX, outfitted with FaceID is leading to the mainstreaming of facial recognition technology which will soon be available on Paytm app. Now, Paytm customers too would be able to pay with just a blink of an eye, recent news reports indicate. The digital payment landscape led by Paytm is planning to introduce facial recognition feature for unlocking the payment application, as reported by ET. This biometric based authentication would also be extended to merchant outlets as well. Facial recognition tech which is already available in iPhones would soon become the norm in Paytm app as well, Deepak Abbot, senior vice president at Paytm revealed to ET. According to Abbot, Paytm has begun testing the facial recognition tool inhouse and this technology will soon be made available to users who can unlock the app with the face tech. Interestingly, Noida based digital payments app has started testing the technology on Google’s Android platform and this will be rolled out through an update. Biometric technology will make payments more secure and robust and leading financial institutions like MasterCard have made public their plans about providing biometric authentication to users in 2019. According to MasterCard, biometric authentication can improve security and reduce digital checkout times and will also reduce burden of using card readers. Meanwhile, MasterCard’s research that was conducted in conjunction with Oxford University revealed a large majority of users preferred consumers prefer biometric security compared to passwords when making transactions. Other advanced banks like the HSBC and First Direct.","excerpt":"Seems like Apple’s iPhoneX, outfitted with FaceID is leading to the mainstreaming of facial recognition technology which will soon be available on Paytm app. Now, Paytm customers too would be able to pay with just a blink of an eye, recent news reports indicate. The digital payment landscape led by Paytm is planning to introduce […]","categories":["AI News"],"tags":["paytm"],"author_name":"Richa Bhatia","publish_date":"2018-09-24T07:45:20","publication_year":"2018","word_count":256,"keywords":["Go","paytm","programming_languages:R","AI","programming_languages:Go","Git","R"],"extracted_tech_keywords":["AI","R","Go","Git","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/paytm-to-introduce-facial-recognition-technology-to-enable-digital-payments\/","complexity_score":2,"technical_depth":6,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10083497,"title":"2022 in Review: Wipro Got a Mixed Bag, Will Tensions Spill Over to 2023?","content":"In 1945, MH Hasham Premji set the ball rolling for Western India Palm Refined Oils Limited, better known as Wipro. Since then, the company has tested the waters in a slew of fields, including digital transformation, technology consulting, IT outsourcing, and many more. Wipro, in particular, has a strong focus on digital transformation and has a wide range of offerings in areas such as artificial intelligence, cloud computing, and data analytics. The year 2022, however, has not been the best year for Wipro. The company made headlines when its profit plummeted, attrition peaked, and it sacked nearly 300 employees for moonlighting. Wipro’s stock, too, couldn’t perform, dropping 45% in the past year, which is one of the largest falls for tech companies in India. On the bright side, Wipro announced a 40,000 sqft innovation studio in Texas and joined hands with Pandorum to develop regenerative medicine with AI. Let’s review the year that was, for the world’s seventh-largest technology service provider company – Wipro. Moonlighting and Wipro In 2022, Wipro gained most media attention when its executive chairman Rishad Premji kicked off the controversy over moonlighting, terming it “cheating, plain and simple”. There is a lot of chatter about people moonlighting in the tech industry. This is cheating – plain and simple.— Rishad Premji (@RishadPremji) August 20, 2022 According to reports, Wipro fired 300 workers to tackle the moonlighting ‘menace’. Premji claimed there was no room for someone to work for both Wipro and its rivals. If competitors learned about it, they would feel the same way, he asserted. “Therefore, I stand by what I said, that any employee who moonlights in any way, shape, or form is breaking the law,” Premji said. To read more about it, click here. Wipro profit plummets Wipro’s Q2 results showed a rather poor performance compared to other WITCH companies. With its profits dropping by 9.3% and year-on-year revenue growth of 13%, Wipro’s performance has been rather unsatisfactory. At Q1 results too, we saw its profit drop by 21%. Wipro claimed the profit reduction was merely a reflection of changes in their operating margins. However, compared to the same period last year, the company’s consolidated revenue increased 14.6% to Rs 22,540 crore from Rs 19,667 crore. According to a filing from the company, the operating margin for IT services for the reviewed quarter was 15.1%, up 16 bps over the previous quarter. To read more about it, click here. Wipro announces 40,000 sqft Innovation Studio in Texas When it comes to innovation, the tech giant is taking great strides. Wipro, a global leader in technology solutions, recently announced the opening of its latest Innovation Studio in Austin, Texas. This cutting-edge facility will be a hub for experimentation, collaboration, and client service, drawing on a diverse array of transformative technologies. From artificial intelligence and 5G, to mixed reality and blockchain, the Innovation Studio will be a melting pot for innovation. It will also leverage the power of multi cloud, edge computing, crowdsourcing, and robotics to bring new solutions to life. And Wipro isn’t doing it alone. The company will be partnering with some of the top technology partners in the industry to make this happen. To read more about it, click here. Wipro and Pandorum join hands to develop regenerative medicine with AI Wipro is also making waves in the world of AI and medicine! The company recently announced a long-term partnership with Pandorum Technologies, a leading biotechnology company based in Bengaluru. Together, they aim to develop technologies that reduce the time-to-market period and improve patient outcomes during regenerative medicine’s R&D and clinical trial phases. This exciting partnership will leverage the power of Wipro Holmes’ AI platform and Pandorum’s expertise in tissue engineering and regenerative medicine. Wipro Holmes will be predicting and improving the effectiveness of therapeutics, offering insights to create specific formulae and helping with study design in the field of medicine. To read more about it, click here.","excerpt":"From artificial intelligence and 5G to mixed reality and blockchain, Wipro’s Innovation Studio will be a melting pot of innovation","categories":["IT Services"],"tags":["5G","Data Analytics","Wipro"],"author_name":"Lokesh Choudhary","publish_date":"2022-12-28T13:00:00","publication_year":"2022","word_count":655,"keywords":["Wipro","5G","artificial intelligence","AI","cloud computing","Git","RAG","Ray","Aim","edge computing","analytics","Data Analytics","R"],"extracted_tech_keywords":["AI","artificial intelligence","analytics","Aim","Ray","RAG","cloud computing","edge computing","R","Git"],"url":"https:\/\/analyticsindiamag.com\/it-services\/2022-in-review-wipro-got-a-mixed-bag-will-the-tensions-spill-over-to-2023\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10118421,"title":"Wayve AI Introduces LINGO-2, Making Driving Easy with Natural Language","content":"Wayve AI has released LINGO-2, a new model that links vision, language, and action to explain and determine driving behavior. https:\/\/youtu.be\/csL-5cvtEGg LINGO-2 is the first closed-loop vision-language-action driving model (VLAM) tested on public roads. The core functionality of LINGO-2 lies in its ability to generate real-time driving commentary while actively controlling a vehicle. While LINGO-1 could retrospectively generate commentary on driving scenarios, its commentary was not integrated with the driving model. Therefore, its observations were not informed by actual driving decisions. However, LINGO-2 can both generate real-time driving commentary and control a car. This integration of language and action allows the model to provide explanations for its driving decisions, such as slowing down for pedestrians or executing overtaking maneuvers. LINGO-2 comprises two modules – the Wayve vision model and an auto-regressive language model. The vision model processes camera images into token sequences, which, along with conditioning variables like route and speed, are fed into the language model. LINGO-2’s New Capabilities Adaptive Driving Behavior: LINGO-2 can be directed through language prompts, such as “pull over” or “turn right,” to adjust its driving behavior. This capability not only aids in model training but also enhances the interaction between humans and vehicles. Real-time Interrogation of AI: LINGO-2 is equipped to predict and respond to queries about the surroundings and its decision-making process while on the road. Live Driving Commentary: Through the integration of vision, language, and action, LINGO-2 can articulate its actions and reasoning in real time, offering insights into the AI’s decision-making mechanisms.","excerpt":"LINGO-2 can both generate real-time driving commentary and control a car.","categories":["AI News"],"tags":["Natural Language Processing"],"author_name":"Siddharth Jindal","publish_date":"2024-04-17T15:16:26","publication_year":"2024","word_count":251,"keywords":["Go","programming_languages:R","AI","Natural Language Processing","programming_languages:Go","R"],"extracted_tech_keywords":["AI","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/wayve-ai-introduces-lingo-2-making-driving-easy-with-natural-language\/","complexity_score":2,"technical_depth":5,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10061002,"title":"Data annotation career: Scope, opportunities and salaries","content":"The demand for data annotation specialists has gone up with the rise in language models, training techniques, AI tools, etc. Data annotation– a critical step in supervised learning–is the process of labelling data to teach the AI and ML models to recognise specific data types to produce relevant output. Data annotation has applications in diverse sectors ranging from chatbot companies, finance, medicine to government and space programs. The market for AI and ML data labelling has seen exponential growth of late.  According to market research firm Cognilytica, the data labelling market will grow from USD 1.5 billion in 2019 t0 USD 3.5 billion in 2024. Types of data labelling A model is as good as the data it’s fed. Hence, it is imperative to ensure data quality of the highest grade with accurate labelling to optimise AI\/ML models. Let us delve into the types of data annotations: Visual data annotation Visual data annotation analysts facilitate the training of AI\/ML models by labelling images, identifying key points or pixels with precise tags in a format the system understands. Data vision analysts use bounding boxes in a specific section of an image or a frame to recognise a particular trait\/object within an image and label it. According to koombea.com, the key skills required for visual data annotation include: Analytical mathematics In-depth knowledge of ML libraries Programming languages like Python, Java, C++, etc. Image analysis algorithms Visual Database Management Understanding of dataflow programming Knowledge of tools like OpenCV, Keras, etc. Audio data annotation Audio data labelling has applications in natural language processing (NLP), transcription and conversational commerce. Virtual assistants like Alexa and Siri respond to verbal stimuli in real-time: Their underpinning models are trained on large labelled datasets of vocal commands to generate apt responses. Startups like Shaip are providing auditory data annotation services to tech giants like Amazon Web Services, Microsoft and Google. The skills required for this field are: Spectrogram analysis In-depth knowledge of ML libraries Programming Languages like Python, Java, C++, etc. Auditory Database Management Knowledge of tools like Audacity, Adobe Audition, Cubase, Studio one, etc. Text data annotation A major part of communications worldwide, be it business, art, politics or leisure, relies on the written word. However, AI systems have trouble parsing unstructured text data. Training the AI systems with right datasets to interpret written language enables the machines to classify text in images, videos, PDFs and files as well as the context within the words. One of the important applications of text data annotation is in chatbots and virtual assistants. The key skills required for this field are: Knowledge in computational linguistics Experience in machine learning Database management Programming languages like Python, Java, C++, etc. Knowledge of tools like GATE, Apache UIMA, AGTK, NLTK, etc. Emerging field with high salaries The AI and data analytics industry is booming in India, and as a result the demand for data engineers, data analysts, data labellers and data scientists are exploding. Data annotation specialists should be adept in various skillsets ranging from machine learning to knowledge of tools specific to the type of annotations. The job demands long periods of focus, attention to detail, and ability to handle different aspects of the machine training process. The freshers in the field of data annotation can expect packages ranging from INR 1.1 lakhs to INR 3 lakhs per annum. According to a survey by Glassdoor, multinational corporations like Siemens, Apple, Google, etc., offer up to INR 7-8 lakhs\/annum packages based on the skills and experience of the individuals. Labelled data of high quality is the primary requirement for the smooth operation of any AI model. Hence, the demand for the implementation of a secure and cost-effective method of data labelling is of paramount importance now. The emerging names in the business of data labelling services are: Acclivis technologies: Founded in 2009, this Pune based company provides high-end services in machine vision, deep learning, artificial intelligence & IoT. The job profiles the company is currently looking for include ML engineer, Image processing engineer, etc. Zuru.ai: The AI-powered data labelling company, founded in 2019, offers high-quality training datasets at scale. Cogito Tech: Founded in 2011 by Rohan Agarwal, this UP-based company offers data labelling services through its platform-agnostic strategy across sectors such as healthcare, automotive, agriculture, defence, etc. IMerit: Founded in 2011, this company extends end-to-end, high-quality data labelling across NLP, computer vision and various content services. The job profiles the company is currently seeking are – ML engineer, ITES executive, etc. IMerit’s control centre is in West Bengal. Wisepl:  Founded in 2020 and based out of Kerala, this company applies different labelling techniques like Semantic Segmentation, KeyPoint Annotation, Polygon Annotation, Cuboid, Polylines Annotation, etc. Professionals interested in the field of data annotation can apply on Wisepl’s website. With international conglomerates outsourcing AI-based services, India has become one of the leading names in the data labelling market globally.","excerpt":"The data labelling market will grow from USD 1.5 billion in 2019 to USD 3.5 billion in 2024.","categories":["AI Features"],"tags":["data annotation","data labelling","Machine Learning","Work from Home"],"author_name":"Kartik Wali","publish_date":"2022-02-18T13:00:00","publication_year":"2022","word_count":813,"keywords":["artificial intelligence","machine learning","Keras","AI","ML","Machine Learning","Work from Home","computer vision","OpenCV","data labelling","NLP","data annotation","deep learning","analytics"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","NLP","computer vision","analytics","Keras","OpenCV"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/data-annotation-career-scope-opportunities-and-salaries\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":16903,"title":"How Is Business Intelligence Different From Data Science","content":"In this heavily jargonized industry, the words often overlap each other, resulting in a lack of understanding or a state of confusion around these concepts. While big data vs analytics or artificial intelligence vs machine learning vs cognitive intelligence have been used interchangeably many times, BI vs Data Science is also one of the most discussed. It is no doubt that BI analyst and data scientist have grown to be the much in-demand jobs with companies in almost all the industries relying on them to have an edge over their competitors. More so, BI and data science has become an integral part of these organizations as data has become a bigger player than ever. Hence the wider adoption of analytics, business intelligence and data science. A brief background – If we go into the flashback a few years from now, companies didn’t have data science positions but were still working on analytics role—these were largely called data analysts. It can very well be labelled as the precursors to the latest data scientists roles. Before we dive onto differentiating these two popular words in the analytics industry, imagine BI and data science having a brief conversation over a problem in hand. While BI would say “What happened and what should be changed?”, data science would ask “Why it happened and what can happen in future?” It’s the difference in “What”, “Why” and “How” that differentiates these two terms. The basic difference – While BI is a simpler version, data science in more complex. BI is about dashboards, data management, arranging data and producing information from data. Whereas data science is all about using statistics and complex tools on data to forecast or analyse what could happen. Data science could very conveniently be stated as an evolution of BI, but on a very complex set of models, application of statistics and use cases. To deal with the same, BI analyst that were earlier focused on the “what” aspect of the problem, started developing toolkit and algorithms that could help them understand and predict business performance. It wouldn’t be wrong to say that business analysts and data scientists work together to turn raw data into useful information. Technology comparison – The market is increasingly becoming competitive, with an ever increasing complex business problems and to drive innovation, companies must shift their focus from traditional BI to data science. That doesn’t take away the importance of business intelligence analyst as they are the ones who would find out patterns and trends in a business’s’ historical data. It can be said that BI analysts explore past trends while data scientists finds predictors and significance behind those trends. This way data scientists help companies mitigate the uncertainty of the future by giving them valuable information—such as topline, cost, risk predictions and others. BI is about answering the questions that might not seem that obvious in a business unit. They help in viewing the relationships between various variables but not exactly predict them. As it was mentioned, BI is about the “what” part of the business and doesn’t easily obtain new meaning or apply insights to new data. Since BI traditionally relied on records stored in relational databases, the structure of the warehouse was intrinsically tied to the types of questions it could answer. BI generally operated with a current or backward-looking focus. Data science, on the other hand have a different path than BI as it relies on predictive analytics, using the statistical method more explicitly. Unlike just finding out patterns, data scientists conduct experiments and hypotheses to reach the “Why” and “How” aspect of a problem. A data scientist profile would have a combination of statistics, IT and business understanding. Yet, a higher focus on applied statistics. Career comparison – Talking about the career in BI, it requires comparatively lesser qualifications than a data scientist. Requiring less formal experience than a career in data science, the main objective of BI is to assist in strategic business decisions. Even someone with a background in data management or IT related field can jump over to BI with relative ease. Since data scientists derive decisions based on predictive algorithms, candidates opting for these job roles may require more technical skillsets in subjects such as statistics, machine learning and programming. It may also require an understanding of languages such as SQL, R, Python or Scala, among others. Using these languages, not only a data scientist can create a framework that leverages historical data, but predict business outcomes much efficiently. Data science is about seamless and scalable integration that may require many engineers to deploy a data scientist’s model across multiple applications. On the other hand, a BI analyst would require proficiency in data handling tools, more so on BI tools like Tableau, Qlik and SQL. Some other BI related tools have emerged recently like Sisense, pentaho, yellowfin, among others. Lot of reporting and BI still happens on excel and not many would be aware about the power of what all can be done on MS excel. A proficiency in excel and SQL is a must have for a BI professional. On a concluding note – In a nutshell, data science and BI are facilitators of each other and can be said that data science is best performed in conjunction with BI. Both of them are required to have an efficient understanding of the business trends hidden in large volumes of data. While BI is the logical first step, data science follows to get deeper insight. As for the job openings for these two, according to our recent study titled Analytics and Data Science Jobs study 2017, 50,000 positions are to be filled by skilled analytics professionals, of which both BI and data science professionals form a large chunk.","excerpt":"In this heavily jargonized industry, the words often overlap each other, resulting in a lack of understanding or a state of confusion around these concepts. While big data vs analytics or artificial intelligence vs machine learning vs cognitive intelligence have been used interchangeably many times, BI vs Data Science is also one of the most […]","categories":[],"tags":["Business Intelligence","Business Intelligence India","data science india"],"author_name":"Srishti Deoras","publish_date":"2017-08-10T10:08:48","publication_year":"2017","word_count":959,"keywords":["Business Intelligence India","data science","artificial intelligence","machine learning","data science india","AI","R","ML","RAG","Python","analytics","Business Intelligence","predictive analytics"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","data science","analytics","RAG","predictive analytics","Python","R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/business-intelligence-different-data-science\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10065769,"title":"Will Musk walk the talk on free speech post-Twitter acquisition?","content":"In nothing short of a coup, Elon Musk has arm-twisted the Twitter board to sell him the microblogging platform for USD 44 billion in an all-cash deal. Musk is known for his grandstanding, and his Twitter bashing in the run-up to the acquisition has become a cause celebre. Musk had hit out at Twitter for content moderation in the past. After the acquisition deal was sealed, the Tesla CEO promised to eliminate Twitter bots. “Free speech is the bedrock of a functioning democracy, and Twitter is the digital town square where matters vital to the future of humanity are debated. I also want to make Twitter better than ever by enhancing the product with new features, making the algorithms open source to increase trust, defeating the spam bots, and authenticating all humans,” he tweeted. Upon completion of the transaction, Twitter will become a privately held company – with Musk making the rules, yet with little changing in the way of regulation from Washington. But what really are the implications of Musk’s takeover of Twitter? Content moderation Elon Musk doesn’t know what he’s talking about when it comes to content moderation, said Mike Masnick, in a TechDirt piece. He said several of Musk’s suggestions have already been attempted or implemented by Twitter. “I strongly believe that Musk has thought about Twitter as a service only as it relates to his user experience—which is, to say the least, a unique one,” he said. Elon has a Twitter following of 84.9 million people, and his tweets have always managed to stir the pot. For instance, in 2020, the SEC banned him from tweeting about Tesla. His tweet led to a USD 14 billion erosion in the value of Tesla stock, and he was also sued for defamation for calling the British diver in the Thai cave rescue a “pedo guy”. Marc Morial, President of the National Urban League, said Elon’s bid for Twitter and his ownership could harm users’ civil rights. In a letter to Twitter Chairman Bret Taylor, he said Musk has expressed concerning views around content moderation and free speech- countering the principles “of creating an online community that is safe for marginalised communities and protects our democracy.” In the absence of safeguards, white supremacist propaganda, racial and religious hatred, voter suppression through election disinformation, algorithmic bias and discrimination could proliferate on Twitter under Musk’s ownership. Free speech “It’s really important for an inclusive arena for free speech. Twitter has become the de facto town square, so it’s really important that people have both the reality and the perception that they can speak freely within the bounds of the law,” Musk said in his TED talk. He also said that his intuitive sense is that having a public platform that is maximally trusted and broadly inclusive is very important. Meanwhile, Mike said it resonates with what Twitter founders have envisioned but found it a challenge to implement at a huge scale. Additionally, “maximally trusted” would also require some level of moderation to fight spam and scams. Publications like The Atlantic and Bloomberg have pointed out the hypocrisy behind Musk’s free speech advocacy. As per reports from ex-and current employees, critics, and even journalists, Musk has attempted to control the information about him in the public forum. TechDirt has gone as far as arguing that Musk lacks a profound understanding of free speech. Open-sourcing algorithms The third major change Musk has proposed is open-sourcing Twitter’s algorithms. “One of the things I believe Twitter should do is open source the algorithm and make any changes to people’s tweets — if they’re emphasised or de-emphasised — that should be made apparent so that anyone can see that action has been taken. So there’s no sort of behind-the-scenes manipulation, either algorithmically or manually,” said Musk. Mike said open-sourcing algorithms is a can of worms Musk may not understand. It is a challenge to open-source algorithms built up in a proprietary stack. Additionally, spammers will be the biggest beneficiaries of open-sources algorithms. “There’s a reason why Google’s search algorithm has become more and more opaque over the years,” Mike said. People most interested in the algorithms were engine spammers, and the same would happen if Twitter were to do the same. Jim Cox, a professor at Duke Law, called Musk’s take over of Twitter a significant change whose impacts will be felt quickly by users. He has no reporting responsibilities that Twitter has now as a publicly-traded company. Currently, the platform has to file periodic reports and can be held accountable. The same accountability won’t hold when Musk takes over the platform.","excerpt":"Marc Morial said Elon’s bid for Twitter and his ownership could harm users’ civil rights.","categories":["IT Services"],"tags":["Twitter Elon Musk"],"author_name":"Avi Gopani","publish_date":"2022-04-27T12:00:00","publication_year":"2022","word_count":766,"keywords":["Go","ELT","programming_languages:R","AI","Twitter Elon Musk","programming_languages:Go","Git","ViT","Rust","GAN","R"],"extracted_tech_keywords":["AI","R","Go","Rust","Git","ELT","GAN","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/will-musk-walk-the-talk-on-free-speech-post-twitter-acquisition\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10094755,"title":"Reddit&#8217;s $12,000 API Pricing Smothers Third-Party Apps","content":"In April, Reddit announced new terms for developer tools and services, paid access to the Reddit Data API, and more native moderation tools. They did not disclose the details about the new price until two months later which was revealed to be $0.24 per 1,000 API calls (<$1.00 per user monthly), which they claim to be a reasonable price. Christian Selig, the developer of Apollo, a third-party Reddit app, however, revealed that they made seven billion API calls last month, which cost up to $1.7 million per month or around $20 million a year while another popular third-party app RIF (Reddit is fun) estimates a similar ballpark. This sparked an outrage on Reddit and more than 2,470 subreddits have confirmed their participation in a scheduled 48-hour blackout on June 12. A few subreddits have taken a more drastic stance, suggesting that they might cease to exist permanently unless Reddit reconsiders their new policy. Where it all began This decision by Reddit comes as the platform seeks ways to monetise its vast collection of user-generated content. Their content has increasingly been utilised to train high-profile text-generating machine learning models like OpenAI’s ChatGPT and GPT-4 for which they are now seeking compensation. As of 2022, Reddit had 430 million active users every month, 1.7 billion global monthly visitors, and 100K active communities. The company is preparing for a potential initial public offering later this year, and this explains the motivation of their shareholders and investors looking for growth or new revenue streams. According to The New York Times, Reddit’s API will not undergo a complete policy change. It will continue to be available for free to developers creating free apps and bots that enhance the Reddit experience, as well as for researchers engaged in academic or noncommercial studies on the platform. However, Reddit’s co-founder and CEO, Steve Huffman, stated in the interview that companies that extract data from Reddit without providing any value to users will now be required to pay. “It’s a good time for us to tighten things up,” Huffman said. “We think that’s fair.” “The Reddit corpus of data is really valuable,” he reiterated. “More than any other place on the internet, Reddit is a home for authentic conversation. There’s a lot of stuff on the site that you’d only ever say in therapy, or AA, or never at all … But we don’t need to give all of that value to some of the largest companies in the world for free.” In terms of revenue, Reddit’s valuation was approximately $10 billion in August 2021. However, two years ago, its ad revenue amounted to only $350 million, a significantly smaller figure compared to Meta and Twitter. Meta generated $113 billion in 2022, while Twitter, despite its controversies, earned nearly $7 billion from advertising. Struggles of Third-Party Apps Reddit accused Apollo for being unscrupulous with their API requests and that charging for them would make it more efficient. Christian Selig, the developer of the app spoke in a recent interview addressing these charges. He explained how these requests work and said, “Every unit of action you can think about is an API request. For example, what I do in Apollo to make it feel a little faster is I request the first 25 posts on a subreddit and then as soon as those load I ask for the next hundred, so you get those first 25 four times quicker giving the users a better experience.” Christian clarified that though it is fair to charge for the API’s, it’s the exorbitant prices that’s causing the outrage. Though this is still completely within the API rate limits (60 a minute per user) and he uses two, he says, it doesn’t seem wasteful to him but wonders if Reddit deems that as inefficient though it is completely within limits. Another sore point for most developers is the inconsiderate timeline. The new policy is to be rolled out on July 1, which gives the developers no time to change their own pricing policy having to take on the extra cost on themselves or shut down the app itself. Reddit said they’re unwilling to negotiate the pricing but are willing to consider a pause in the initiation of the pricing plan. Earlier this year, Twitter faced significant criticism after implementing a comparable policy that excluded developers and researchers from accessing its valuable data. This move led to the demise of several third-party apps, receiving widespread disapproval. Removal of sexually implicit content from the third party apps are subject to restrictions making such content unavailable. Christian Selig and others agree that doing so would wipe out a large chunk of users from visiting the app. In other cases, medical questions with ‘gory’ images marked NSFW would also be unavailable to users. The reasons given for this limitation is unclear but are in line with legal obligations. While the protest sends a message to Reddit, it looks unlikely that the company would make any changes to the API pricing. If a significant number of users are willing to put up with a partial absence of Reddit for two days, then the company has to think about what percentage of those will leave forever if they don’t relent.","excerpt":"Reddit’s $12,000 per 50 million API calls leaves third-party app developers astray","categories":["AI Features"],"tags":["reddit"],"author_name":"K L Krithika","publish_date":"2023-06-08T16:09:52","publication_year":"2023","word_count":870,"keywords":["Go","ChatGPT","API","machine learning","OpenAI","AI","reddit","RAG","GPT","Aim","R"],"extracted_tech_keywords":["AI","machine learning","ChatGPT","OpenAI","Aim","RAG","R","Go","API","GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/reddits-12000-api-pricing-smothers-third-party-apps\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10161564,"title":"Infosys Builds 4 Small Language Models, To Develop 100 Client-Focused AI Agents","content":"Infosys posted strong financial results for Q3 FY25, highlighting growth in revenue, operating margin, and free cash flow. CEO and MD Salil Parekh attributed the performance to the company’s strong positioning in digital services and its growing enterprise AI capabilities, particularly in generative AI and AI agents. “In generative AI, we have built four small language models for banking, IT operations, cyber security and broadly for enterprises in generative AI,” Parekh said. He added that the company is developing over 100 new agents for its clients, many of which are already in use. “We are very clear about what we are doing with generative AI,” Parekh said. He added that Infosys is not doing ‘AI washing’ as others in the industry might be and, instead, doing real generative AI work. “We have several discussions with clients where they would like to use the small language models that we have built…They are built by using the proprietary data that we have. Some clients are asking us to build a small language model of their own.” For AI agents, Parekh explained that Infosys has built a research agent for its clients which is not just a PoC. “They are now using that in their product area to support how queries are looked at, and where their own people and their own customers can use this agent,” Parekh said, adding that these agents are able to reduce the work of 18 days to eight days. The company reported a revenue of $4.94 billion, reflecting a sequential growth of 6.1% year-on-year (YoY) and 1.7% quarter-on-quarter (QoQ) in constant currency terms. The global IT giant reported a net profit of $804 million for the quarter, which demonstrates robust financial performance. “We continue to strengthen our enterprise AI capabilities, particularly focusing on generative AI, which is witnessing increasing client traction,” Parekh further said. “This has led to another quarter of strong large deal wins and an improved deal pipeline, giving us greater confidence as we look ahead.” The company secured $2.5 billion in large deal total contract value (TCV) with 63% net new deals, growing 57% sequentially. Headcount also increased for the second consecutive quarter, currently standing at 3,23,379. “We have had a strong hiring in Q3 with the addition of over 5,000 employees,” Parekh added. Infosys is on track to onboard 15,000 to 20,000 freshers at the group level in FY25. Indian IT Reporting Strong Generative AI Pipeline When Accenture reported its earnings for the first quarter of fiscal 2025 (ending November 30, 2024), the company set a record of $1.2 billion in generative AI bookings, reflecting growing client investments in this space. In FY24, the company reported $4.2 billion in GenAI bookings overall. It was expected that Indian IT firms would follow suit. However, even TCS, L&T, HCLTech, and now Infosys, in all of their recent earnings calls, shied away from revealing the revenue from generative AI, citing several reasons like “AI is now part of every deal”. While all of them said that they are working on GenAI and AI agents, the exact details remain unknown. For example, in its latest Q3 FY25 earnings call, TCS reported that its clients are actively investing in generative AI and agentic AI while building robust data foundations. K Krithivasan, CEO and MD, said that TCS has been working on a drug discovery project with a client who was able to identify 1,300 molecules and further filter them to 12 molecules. Moreover, HCLTech announced during its Q3 FY25 earnings call that it is advancing its GenAI strategy with an aim to integrate AI services into 100 clients by FY26. Similarly, in the last quarter, Infosys finally revealed that it is working on small language models, building multi-agent frameworks, and doing incredible work with generative AI. The exact numbers gained from this work, however, remained undisclosed. Parekh added that the company believes small language models will provide clients with a powerful tool, allow them to build business logic on top of it and unlock new potential. “This combination will form the foundation of the small language model, which is being tailored for different industry applications,” he added. Infosys is Doing a Lot of Generative AI Just a few hours before the result, Infosys had launched a suite of AI-driven features for the Australian Open (AO) 2025 in partnership with Tennis Australia. As per the company’s press release, this marks a major milestone in their seven-year collaboration, which aims to enhance the tennis experience through technology. Last month, Infosys also launched a Google Cloud Centre of Excellence at its Bengaluru campus to drive enterprise AI innovations. The initiative, powered by Infosys Topaz, aims to help businesses use generative AI for transformative growth. Furthermore, as promised earlier, Infosys is finally becoming an AI-first company. For instance, at the Microsoft Building AI Companions for India event in Bengaluru, Infosys CTO Rafee Tarafdar said that developers have been using GitHub Copilot for over a year, which now has about 20,000 users generating nearly a million lines of code every few weeks. Beyond internal work, at Meta’s Build with AI Summit held in Bengaluru last year, Infosys announced a partnership with Meta to utilise the Llama stack, a collection of open-source large language models and tools, to build AI solutions across industries. As an early adopter of Llama 3.1 and 3.2 models, Infosys is integrating these models with the in-house AI platform Infosys Topaz to create tools that deliver business value. One example of such a tool is a document assistant powered by Llama that improves the efficiency of contract reviews.","excerpt":"Salil Parekh said that Infosys is not doing ‘AI washing’ as others in the industry might be and, instead, doing real generative AI work.","categories":["IT Services"],"tags":["Developers","Infosys","small language models"],"author_name":"Mohit Pandey","publish_date":"2025-01-16T16:42:49","publication_year":"2025","word_count":928,"keywords":["Go","GenAI","agentic AI","Infosys","AI","Git","Aim","generative AI","small language models","GitHub","R","Developers"],"extracted_tech_keywords":["AI","generative AI","GenAI","agentic AI","Aim","small language models","R","Go","Git","GitHub"],"url":"https:\/\/analyticsindiamag.com\/it-services\/infosys-builds-4-small-language-models-to-develop-100-client-focused-ai-agents\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10002137,"title":"5 Notorious Mobile Threats Everyone Should Take Seriously","content":"In this ever-evolving technological era, mobile security is becoming the topmost concern — not only for organisations but for individuals as well. The main reasons behind it is the increasing number of data breaches that the world is witnessing. Nowadays, many companies across the world are adopting the concept of BYOD — it not only includes laptops but mobiles as well. Also, even if there is no BYOD involved, employees access corporate data from their smartphones and that is definitely something that bothers a company. Talking about mobile device security, many have the misconception that its all about malware; however, that is not the case — the rate of mobile devices getting affected by malware is significantly less compared to a computer. You all might be wondering then — what are the threats? A mobile device gets in the least expected way and that too when you least expect it. In this article, we are going to some of the notorious mobile threats. Adware and Spyware As more and more applications have started to come for mobile, the number of mobile devices hacks are also increasing. Hacker across the world are building lookalike apps and sharing it. However, those apps are nothing but malicious pieces of code that spies on your mobile device. Adware and Spyware over the last couple of years have impacted a lot of devices — these notorious applications pop ads on the screen, and tries to redirect the user to another malicious website. This is one form. Another form of this kind of fake apps are sophisticated spy apps. For example, suppose you have downloaded an application to save deleted WhatsApp text and the app is working actually well and fine. It is serving its purpose. However, have you ever realized that to save those chats you have given that app access to private chat? There are other spyware too in the form of random application that spies on your browser and other platforms, trying to collect data about you. Drive-by download Drive-by download has become a very famous way of bringing people to the battleground of hackers. It is basically an unintended download of a software or application from the Internet. Drive-by download has also two aspects — one is when you download a software without knowing the consequences and the other is any download that happens without a person’s knowledge. In both cases, the end result is always about compromising the device. Drive-by download is one of the major reasons for adware and spyware getting installed on mobile devices. Phishing Applications This is another type of threat that is year by year becoming more popular. The concept behind phishing application is the same as phishing website — a hacker creates an application with a concept and asks the user to provide his credentials. This type of attack is mostly carried out using Adult or dating application. So, it is advised that cross-check the application before installing. Google has taken significant steps to prevent fake apps from getting onboard on the Play Store; however, there are still many apps that are managing to make it to the store, and if not Google Play Store then they are using third-party platform. Network Spoofing Free Wi-Fi is one of the major sources of network spoofing and mobile users are more likely to fall prey to network spoofing. It is also considered to be one of the most dangerous ways of cyber attacks as it could affect a number of people connected to the network. It is usually carried in places where free Wi-Fi is available. Hackers create a hotspot that looks like a genuine a Wi-Fi connection, and every time someone gets connected to the network the attackers take a look at the network for credentials like username and password. A lot of people tend to have the same credentials for several other accounts and if they fall prey to this attack, they might face a big consequence. So, every time you connect to free a Wi-Fi network and if it asks you to make a login username and password, make sure it’s unique and different. Cryptojacking Compared to all the other threats, Cryptojacking is relatively new. If you don’t know what Cryptojacking is, it is an unauthorized use of a computer, tablet, mobile devices, or even a  connected home device to mine for cryptocurrency. This attack completely uses a device’s power to mine cryptocurrency; meaning, the phones that are affected are likely to experience a very poor battery life. That is not all, the device could even suffer from damage due to overheating components.","excerpt":"In this ever-evolving technological era, mobile security is becoming the topmost concern — not only for organisations but for individuals as well. The main reasons behind it is the increasing number of data breaches that the world is witnessing. Nowadays, many companies across the world are adopting the concept of BYOD — it not only […]","categories":["AI Trends"],"tags":["Cybersecurity","malware"],"author_name":"Harshajit Sarmah","publish_date":"2019-05-24T19:20:51","publication_year":"2019","word_count":768,"keywords":["Go","malware","programming_languages:R","AI","programming_languages:Go","GAN","Cybersecurity","R"],"extracted_tech_keywords":["AI","R","Go","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/5-notorious-mobile-threats-everyone-should-take-seriously\/","complexity_score":2,"technical_depth":6,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10066097,"title":"Want to learn graph databases? Here are a few courses to go for","content":"A graph database uses a graphical model to represent and store data. The global graph database market size was valued at USD 651 million in 2018 and is projected to reach USD 3,731 million by 2026, growing at a CAGR of 24.5% from 2019 to 2026, according to a report by Allied Market Research. Graph databases find a variety of usage in fraud detection, identity and access management, recommendation engines, etc. Let us look at a few courses that can help you get started with understanding how graph databases exactly work and what problems they can solve. Graph databases – Coursera This course has a six-week duration. It will help students identify the differences between relational and NoSQL databases. After course completion, students will be able to understand what type of NoSQL database to implement based on business requirements (key-value, document, full text, graph, etc.) and NoSQL data modelling from application-specific queries. They will also be able to use atomic aggregates and denormalisation as data modelling techniques to optimise query processing. For more information, click here. Neo4j Graph Academy Neo4J, the creator of graph database management systems, has its own courses to introduce enthusiasts to graph databases. These courses are free and self-paced. The courses will also include hands-on exercises to gain actual experience with Neo4j. Some of the courses include: For beginner level: Neo4j fundamentals, Cypher fundamentals, graph data modelling fundamentals, importing CSV data into Neo4j After mastering the fundamentals, one can move ahead to curated learning paths to become an expert in their area of interest. These include: Administrator – It is for server admins or DevOps professionals who are looking for best practices on how to put Neo4j in production and keep it safe.Aura Practitioner – It teaches how to run Neo4j in the cloud with Neo4j Aura.Cypher – Here, the students can learn the more advanced cypher functionality.Data Scientist – It teaches everything from loading data to applied tutorials on the Graph Data Science library.Developer – Students are taught about building Neoflix, a fictional movie streaming platform, using Neo4j’s official drivers and frameworks. For more information, click here. Graph Database Certification – TigerGraph TigerGraph also offers graph database training and certification courses. It provides two types of certificates. TigerGraph GSQL Programming 101 – This provides an introduction to graph database and analytics and gives an overview of use cases and benefits. Graph Algorithms for Machine Learning – It covers five categories of graph algorithms (shortest path algorithms, centrality algorithms, community detection algorithms, similarity algorithms, classification algorithms) and how they improve the accuracy of machine learning for use cases like fraud detection, recommendation engine, entity resolution, customer 360 and knowledge graph. For more information, click here. Introduction to Graph Databases, Neo4j & Cypher – Udemy This course has three hours of on-demand video. Students will learn how to identify graph-shaped problems, types of NoSQL databases and their applications. In part one of the course, students will get an introduction to graph theory concepts along with the history of graph theory. Then, students will be introduced to the Neo4j graph database and the associated query language, Cypher. The second part of the course will focus on querying with the Cypher Query language on the Movies dataset built into Neo4j. This course is suitable for anyone who plans to improve their NoSQL knowledge of graph databases. Social science researchers who want to learn how to work with influence analysis and other applications of graph databases can also take up this course. Anyone who is curious about networks is a suitable candidate too. For more information, click here.","excerpt":"Neo4j, the creator of graph database management systems, has its own courses to introduce enthusiasts to graph databases.","categories":["AI Trends"],"tags":["Coursera","Courses","Cypher","databases","graph databases","neo4j","udemy"],"author_name":"Sreejani Bhattacharyya","publish_date":"2022-05-03T15:00:00","publication_year":"2022","word_count":598,"keywords":["databases","data science","Go","graph databases","machine learning","programming_languages:R","Coursera","AI","R","udemy","neo4j","analytics","SQL","Courses","DevOps","fraud detection","Cypher"],"extracted_tech_keywords":["AI","machine learning","data science","analytics","fraud detection","R","SQL","Go","DevOps","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/want-to-learn-graph-databases-here-are-a-few-courses-to-go-for\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10053106,"title":"Attend This Webinar By IIM Calcutta To Accelerate Your Career In Data Science","content":"Even six years after data science gained prominence, it continues to gain much traction. Its field of application keeps growing day by day, attracting talent from a variety of disciplines. It wouldn’t be wrong to say that this is probably the best time to be a data scientist. To cater to this growing demand for skilled data scientists, the Indian Institute of Management, Calcutta (IIMC) is launching the fifth batch of its one-year Advanced Programme in Data Sciences (APDS). This programme is designed in collaboration with VCNow to bridge the talent gap that exists in this rather dynamic field. About the Programme The one-year programme offered by IIMC covers all the latest trends and advancements in data science. Designed specifically for working professionals, this programme will introduce the participants to different tools and techniques used in handling, managing, analysing, and interpreting data. Apart from this, the programme would also introduce various software like Tableau, Oracle SQL, Python, R, Hadoop for data management and analysis. The programme is designed by IIM Calcutta and will be delivered by its esteemed faculty. In this programme, candidates will form small teams and undertake organisation-based specialised projects of their choice. The aim of this programme is to develop a rock-solid foundation in students who will be able to apply advanced quantitative and statistical tools for effective decision making which in turn will aid business transformation. The fifth batch of APDS is open for working professionals\/ graduates\/ postgraduates with 3+ years of work experience in a role involving the handling and interpretation of data and young managers aspiring to build a career in data science. Applicants must have a basic understanding of Mathematics and Statistics. A background in quantitative techniques and data analytics tools is preferred. To provide more clarity on this program, IIMC, in association with Analytics India Magazine, is organising a webinar — Get industry-ready with IIM Calcutta’s Advanced Programme in Data Science. This session will provide information on the programme, which will be followed by a Q&A session. Webinar Date & Time :Nov 14, 2021, Sunday from 4.00 PM to 5.00 PM IST. Click here to register Speakers: Prof Manisha Chakrabarty, Programme Director & Faculty, IIM Calcutta Dr Manisha Chakrabarty is a Professor in the Economics Area at the Indian Institute of Management Calcutta where she joined in 2007. She is a PhD from Indian Statistical Institute, Calcutta and has pursued her Master of Science from the University of Calcutta, with a Specialization in Statistics and Econometrics. She was awarded the Best Research Paper Award in 2011 in the International Conference on Data Envelopment Analysis and Its Application to Management; Centrum Catolica, Lima, Peru during 2011. Her broad area of research interest includes Applied Econometrics, Development Economics & Applied Finance. Prof Debashis Saha, Programme Director & Faculty, IIM Calcutta Dr Debashis Saha is a Professor in the MIS Group & Computer Science Area at the Indian Institute of Management Calcutta. He had joined in 2004 in this capacity. He is a PhD in Computer Communication & Networks from the Indian Institute of Technology Kharagpur. He has also pursued his M. Tech. in Electronics & Electrical Communication Engineering from IIT Kharagpur. Prior to that, he has done his B.E. (Hons.) in Electronics & Telecomm Engineering from Jadavpur University, Calcutta. He was awarded the Best Research Paper award on quite a few occasions – a few being the Best Paper award of Technology and Innovation Management (TIM) Division in the Academy of Management annual conference (AOM 2020); Best Paper Award in the Thirteenth International Conference on Digital Society and e-Governments (ICDS 2019), Athens, Greece, February 24 – 28, 2019, etc. His broad area of research interest includes Digital Business Transformation, IT Strategy and Governance, Digital ecosystem, IoT, Industry 4.0, Diffusion of Telecom innovation, Techno-economic of ICT, 4G\/5G Network Operations, Management & Security, ICT for Development, Pervasive Computing & Communication, WDM Optical Networks, Protocol Engineering, etc. Who can attend: Working professionals functioning in a role involving the handling and interpretation of data and young managers aspiring to make a career in data science. Webinar Date & Time :Nov 14, 2021, Sunday from 4.00 PM to 5.00 PM IST. Click here to register","excerpt":"Must-attend webinar by IIM-C for all aspiring data scientists to gain insights on IIMC’s 12 months advanced data sciences programme.","categories":["Deep Tech"],"tags":["iim calcutta"],"author_name":"AIM Media House","publish_date":"2021-11-09T18:00:00","publication_year":"2021","word_count":696,"keywords":["data science","Go","AI","Git","RAG","Python","Aim","iim calcutta","analytics","SQL","R"],"extracted_tech_keywords":["AI","data science","analytics","Aim","RAG","Python","R","SQL","Go","Git"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/attend-this-webinar-by-iim-calcutta-to-accelerate-your-career-in-data-science\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10023585,"title":"India’s First Speech Recognition System For Healthcare Industry: The Startup Story Of Augnito","content":"India’s doctor patient ratio stands at 1:1000 Indians, way below the standards prescribed by the World Health Organisation. Clinical documentation plays a critical role in facilitating better patient care with seamless billing, insurance, medico-legal processes. COVID has forced healthcare providers to deploy highly efficient medical reporting systems to minimise human contact and allow doctors to have dynamic working hours. The pandemic has pushed the widespread adoption of electronic medical records. Rustom Lawyer founded a cloud-based speech-to-text software, Augnito, in 2020 that guarantees error-free medical documentation by converting human voice to written text in real-time. Augnito is India’s first and only medical speech recognition AI available as a real-time speech-to-text application for doctors to generate reports. Rustom is also the man behind Scribetech, a 19-year-old company that pioneered clinical documentation in the UK NHS. Analytics India Magazine got in touch with Rustom Lawyer, the co-founder & CEO of Augnito, to gain insights into the company’s processes. How Does It Work? A truly ‘made in India’ product, Augnito uses Human-Centred Design principles to understand the needs and challenges of doctors unique to the Indian context. It took seven years for the company to develop this software. “Currently, the potential use cases are very exciting, and we are the only company in India providing such an advanced voice technology in the healthcare domain,” said Rustom. The software has been built on deep neural network-based technology that leverages the latest and most cutting edge techniques in speech recognition science. The strong natural language processing team and robust NER (Named entity recognition) techniques help structure the text. “We also have a large in-house GPU infrastructure,” said Rustom. With a cloud-based solution built on highly advanced deep learning models, Augnito offers the highest accuracy rates of 99% for diverse Indian accents without any voice training. “It’s like you can carry the entire language of medicine with you, everywhere. The flexibility and mobility that Augnito offers by being a cloud-based multi-platform solution is one of the key differentiators” added Rustom. To facilitate this on the cloud, the developers use tools like Kubernetes, Python and Amazon DynamoDB and on the front end, and Angular, Javascript and Dot Net tools for the backend. The system is available as a streaming speech API for direct integration and contains the entire language of medicine covering 50+ specialities and subspecialties. The system has also been recently awarded the prestigious Abdul Kalam Fellowship by the Government of India. Wrapping Up Rustom believes speech recognition and natural language processing are currently at an exciting phase and can be a critical factor driving electronic medical records adoption. With Augnito, Rustom tried to package this technology into a first-of-its-kind system for our country, easily accessible to every doctor. Augnito has invested about a million dollars in the business (completely bootstrapped) and is currently being used by 4000+ doctors across India to produce medical reports. “In the next five years, the startup is expecting to see Augnito as the voice interface driving software in every hospital across the country,” concluded Rustom.","excerpt":"Rustom believes speech recognition and NLP can be the critical factors in driving electronic medical records adoption.","categories":["AI Startups"],"tags":["automatic speech recognition","Speech Analytics","Speech Recognition","speech recognition algorithm"],"author_name":"Sejuti Das","publish_date":"2021-04-09T10:00:00","publication_year":"2021","word_count":504,"keywords":["Speech Analytics","AI","neural network","ML","speech recognition algorithm","automatic speech recognition","RAG","Python","deep learning","analytics","Speech Recognition","JavaScript","R","kubernetes"],"extracted_tech_keywords":["AI","ML","deep learning","neural network","analytics","RAG","kubernetes","Python","R","JavaScript"],"url":"https:\/\/analyticsindiamag.com\/ai-startups\/indias-first-speech-recognition-system-for-healthcare-industry-the-startup-story-of-augnito\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10016970,"title":"World’s First Augmented Reality  Surgery, Solorigate And More In This Week’s Top News","content":"The last days of what has been a very challenging year have proven even more difficult with the exposure of a suspected Russian cyberattack which compromised many companies and government bodies, especially in the United States. The Solorigate or the cyber attack on Solarwinds company is believed to be one of the largest attacks ever. This attack served a huge wake-up call for the tech community by shedding lighting on the various ways in which the cybersecurity landscape has become even more dangerous. The world will now be forced to commit to more effective and collaborative cybersecurity strategies and coordinate a strong global cybersecurity response. Read more about the Solorigate’s investigation by Microsoft in this week’s top news brought to you by Analytics India Magazine. Amazon Gets Serious About Podcasts On Wednesday, Amazon announced it will be acquiring podcast producer Wondery reportedly for $300 million. Wondery will be part of Amazon Music, which launched podcasts in September 2020. With the purchase of Wondery, Amazon will now be competing with the elites of podcast business such as Spotify. Though Amazon has been late in arriving on the podcast scene, it, however, has a huge head start when it comes to building state of the art recommendation engines that power the music platforms. And, not to forget Amazon Web Service which power most of the streaming services across the world. World’s First AR Surgery (Source: Microsoft) Augmented reality(AR) has finally and formally made an entry into the operating theatre. Ushering next-gen clinical procedures, the surgeons at the Balgrist University Hospital have used AR to navigate during spinal surgery. They did this by using generated CT imaging, 3D representations of the affected anatomy that is directly projected into the surgical field overlaying the real anatomy during the operation. Surgeons were assisted with prompts for various actions such as the exact insertion point and trajectory of a screw. “AR enhances the surgeon’s senses and improves their perception,” said Farshad, Medical Director of Balgrist University Hospital. Graphcore Chips In Quarter Of A Billion Image credits: Graphcore Graphcore ended the year on a great note. On Tuesday, the company announced that it has raised $222 million of investment tthat will fuel the company’s continued global expansion and further accelerate future IPU silicon, systems and software development. Last year, Graphcore, launched Mk2 IPU processor, the GC200 and datacenter compute systems, the IPU-M2000 and IPU-POD64 for scale-out. Graphcore claims to have dominated the latest GPU-based setups as well. With this Series E funding, the total funds raised by Graphcore comes down to more than $710 million, with the company expecting to have over $440 million of cash in the future. Graphcore is currently valued at $2.77 billion. Microsoft Hacked On Thursday, Microsoft released a statement related to their investigation into the Solorigate incident; a widespread cyber attack by hackers linked to Russia that compromised many corporate and government systems in the United States. According to Microsoft, the hackers were able to access internal systems within Microsoft and view internal source code, used to build software products. According to reports, the attackers were able to compromise the system SolarWinds was using for assembling its finished software products. The hackers then slipped in some malicious code into SolarWinds’ own software updates, which were shipped to about 18,000 customers, including Microsoft and FireEye. They even found their way into systems belonging to the Department of Homeland Security, the State Department, the Treasury and Commerce departments and others. “We believe the Solorigate incident is an opportunity to work together in important ways, to share information, strengthen defenses and respond to attacks,” said Microsoft in their statement.  Read more about Solorigate here. Google Demands For Better Drone Laws Source: Wing Alphabet Inc.’s drone delivery subsidiary Wing is reportedly unhappy with the  Department of Transportation and the Federal Aviation Administration (FAA)’s new rules. In a blog explaining the dilapidations of the new rulemakings, the team at Wing wrote that the FAA’s demand to broadcast remote identification or RID regardless of use case can have negative effects on the privacy of the customers. The aviation authority requires all flying vehicles to broadcast an ID to keep an eye on the trajectories for citizen safety and other purposes. But, Wing argues that broadcasting IDs of drones that do everyday tasks like delivering groceries can make users vulnerable to outsider attacks as their data can be completely exposed. “…an observer tracking a drone can infer sensitive information about specific users, including where they visit, spend time, and live and where customers receive packages from and when. American communities would not accept this type of surveillance of their deliveries or taxi trips on the road. They should not accept it in the sky,” lamented Wing. Boston Dynamics Rings In New Year In Style This week Boston Dynamics almost had us convinced that robots can be friendly when they uploaded a video of their flagship products dancing to the song “Do You Love Me”, a classic from 1962. The Atlas robots were joined by “Spot” in delivery one of the smoothest dancing performances even by human standards. “Snake-head dogs had my undivided attention until winder-head ostrich came gliding through all nonchalant,” tweeted Elon Musk. Boston Dynamics was recently purchased by Hyundai for nearly a billion dollars.","excerpt":"The last days of what has been a very challenging year have proven even more difficult with the exposure of a suspected Russian cyberattack which compromised many companies and government bodies, especially in the United States. The Solorigate or the cyber attack on Solarwinds company is believed to be one of the largest attacks ever. […]","categories":["AI News"],"tags":["Boston Dynamics"],"author_name":"Ram Sagar","publish_date":"2021-01-02T18:00:00","publication_year":"2021","word_count":877,"keywords":["Go","Boston Dynamics","API","funding","AWS","AI","cloud_platforms:AWS","Graphcore","Aim","analytics","R"],"extracted_tech_keywords":["AI","analytics","Aim","AWS","R","Go","API","funding","Graphcore","cloud_platforms:AWS"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/microsoft-solorigate-augmented-reality-top-news-2021\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10019665,"title":"Microsoft Makes Azure Quantum Available For Public Preview: What Does It Entail?","content":"Microsoft has made its quantum computing service, Azure Quantum, available for public preview. Azure Quantum is the world’s first full-stack, public cloud ecosystem for quantum solutions, Krysta Svore, General Manager at Microsoft Quantum, claimed. “The unified Azure Quantum ecosystem will accelerate your R&D with access to diverse quantum software and hardware solutions, a network of leading quantum researchers and developers, a robust resource library, and flexible self-service or tailored development programs for customers and systems integrators,” Svore added. Microsoft unveiled Azure Quantum at its Ignite event in November 2019. Honeywell, Toshiba, 1QBit, IonQ are working with Microsoft to improve the Azure Quantum ecosystem. Azure Quantum is a full-stack, open cloud ecosystem to run quantum computing programs or solve optimisation problems in the cloud. Julie Love, senior director at Microsoft Quantum, stated in a blog, “With Azure Quantum Public Preview, we’re opening up the technology to the broader ecosystem.” Love added, “This means that developers, researchers, systems integrators, and customers can use it to learn and build.” Quantum computing research is enabled by a rich set of tools that ranges from the Q# programming language to the Quantum Development Kit (QDK). Developers can quickly write programs with the Q# language and QDK and run the code against simulators and various quantum hardware, she said. Features Access Quantum Computing And Optimisation Solutions In The Cloud: Through a single development interface, you can tap into the unique capabilities offered in quantum computing and optimisation solutions. Easy Collaboration: The open-source Quantum Development Kit (QDK) with the Q# quantum programming language protects the development investments by anticipating and integrating with advances in quantum systems. Explore At Your Own Pace: Amazon Quantum follows “Pay as you go” rule. Developers have the flexibility to choose from self-service development or tailored development services with the Enterprise Acceleration Program. Optimisation: Azure Quantum provides access to a broad set of state-of-the-art optimisation algorithms developed by Microsoft and its partners. Developers can use classic optimisation algorithms as well as quantum-inspired optimisation algorithms (QIO). Providers And Targets: Azure Quantum adds the Microsoft QIO provider to every workspace, and developers can add other providers when they create the workspace. Benefits Offers developers access to various quantum computing devices and enables them to run their Q# quantum programs in real hardware. Option to run algorithms in simulated quantum computers to test the code.Access to trapped ion devices through the providers IonQ and HoneywellAccess to diverse quantum software, hardware and solutions from Microsoft. Wrapping Up Tech giants like Microsoft, IBM, Google are bullish on Quantum computing. With the public preview of Azure Quantum, Microsoft has taken another step forward to accelerate the research and development of quantum software and hardware solutions. The quantum service offers a range of solutions from Microsoft and its partners. However, the pricing is determined by the individual partners. Click here to know more.","excerpt":"Microsoft has made its quantum computing service, Azure Quantum, available for public preview. Azure Quantum is the world’s first full-stack, public cloud ecosystem for quantum solutions, Krysta Svore, General Manager at Microsoft Quantum, claimed. “The unified Azure Quantum ecosystem will accelerate your R&D with access to diverse quantum software and hardware solutions, a network of […]","categories":["Global Tech"],"tags":["quantum cloud software"],"author_name":"Ambika Choudhury","publish_date":"2021-02-04T13:00:00","publication_year":"2021","word_count":473,"keywords":["Go","cloud_platforms:Azure","programming_languages:R","AI","R","quantum cloud software","programming_languages:Go","Aim","Azure","emerging_tech:quantum computing"],"extracted_tech_keywords":["AI","Aim","Azure","R","Go","cloud_platforms:Azure","programming_languages:R","programming_languages:Go","emerging_tech:quantum computing"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/microsoft-makes-azure-quantum-available-for-public-preview-what-does-it-entail\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10163353,"title":"‘We’d Like to Work With China,’ says OpenAI CEO Sam Altman","content":"OpenAI CEO Sam Altman revealed in an interview with British media outlet Sky News on Tuesday that the company would like to work with China. Altman made the comment when he was asked if he was worried about the country’s progress. However, responding to a question about whether the US would let him do that, Altman said, “I know that for sure, no. Should we try as hard as we can? Absolutely, yes.” Altman’s statement follows the shockwave created by DeepSeek in the AI ecosystem, which has sparked concerns among several industry leaders who fear China’s rise. Recently, Dario Amodei, CEO of Anthropic, said, “If we can close [export control loopholes] fast enough, we may be able to prevent China from getting millions of chips, increasing the likelihood of a unipolar world with the US ahead.” Similarly, in an interview with CNBC at the Davos World Economic Forum 2025, Microsoft CEO Satya Nadella said, “We should take the development of China very seriously.” Venture capitalist Vinod Khosla echoed the industry’s strongest fear. In a blog post, he earlier said, “We may have to worry about sentient AI destroying humanity, but the risk of an asteroid hitting the Earth or a pandemic also exists. But the risk of China destroying our system is significantly larger, in my opinion.” Altman’s take on China seems refreshing, to say the least. In a podcast with The Times on Monday, while speaking about China’s DeepSeek, Altman said, “They did some nice work. I think there’s also some nice pieces of product work like showing the chain of thought.” He went on to praise the research behind DeepSeek’s model, although he feels “it isn’t a big update” to the ecosystem. OpenAI reportedly has evidence that DeepSeek trained on its models and is investigating this with Microsoft. However, Altman recently revealed that the company does not plan to sue DeepSeek right now. Altman is Still a Fan of Open Source In the same interview with Sky News, Altman added, “I think we should probably open source somewhat more.” In a Reddit ‘ask-me-anything’ (AMA) session, he said, “I personally think we have been on the wrong side of history here and need to figure out a different open-source strategy.” This comes at a time when industry leaders are rooting for the development of more open-source models, owing to DeepSeek’s recent success. Meanwhile, Chamath Palihapitiya, a venture capitalist, said in an interview, “I think in the war of open versus closed, open has won.” Palihapitiya believes OpenAI is now considering to open source what they’re doing in some way. That said, OpenAI has released multiple open-source models and tools in the past. The company’s second iteration of the GPT model was also made available for open-source use.","excerpt":"Altman says DeepSeek did some nice work.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","OpenAI","Sam Altman"],"author_name":"Supreeth Koundinya","publish_date":"2025-02-12T17:14:39","publication_year":"2025","word_count":458,"keywords":["Anthropic","API","Sam Altman","OpenAI","AI","IPO","venture capital","GPT","chain of thought","R","AI (Artificial Intelligence)","llm_models:GPT"],"extracted_tech_keywords":["AI","OpenAI","Anthropic","chain of thought","R","API","GPT","venture capital","IPO","llm_models:GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/wed-like-to-work-with-china-says-openai-ceo-sam-altman\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10102896,"title":"Foxconn to Steer the Electric Wheel","content":"The shift towards electric vehicles (EVs) has been vividly evident in the past decade. Car companies had been working on clean energy long before Elon Musk from Tesla brought attention to it. But in the recent past, every tech company, be it the phone developers or chipmaker Foxconn, has been trying to put their own model on the road. Last week, the Taiwanese electronics giant announced last week that it is taking a gamble on the EV business. Diversification opportunities are ample for Foxconn and the company is going all-in on cars next. The Taiwanese company has already set up fabs and plans further expansion through several automotive production sites across the globe — including India. The Desi Connection India boasts millions of EV owners, with motorbikes, scooters, and rickshaws constituting over 90% of the automotives. As per a Bloomberg report, the sales rose to 75,000 in the nine months through September — more than double the volume during the same period in 2022. A report by the International Energy Agency (IEA) revealed that in 2022, more than half of India’s three-wheeler registrations were electric-powered. One of the reasons for the company to embark on an affair with India can be the rise in demand for passenger EVs in the country. The surge in the EV market can be attributed, in part, to a government initiative, a $1.3 billion scheme aimed at prompting EV manufacturing within the country while offering discounts to customers. Moreover, charging points across the nation have increased tenfold, reported Elizabeth Connolly, an analyst specializing in energy technology and transport at the IEA. As Foxconn charts a course to become an assembler of electricity powered vehicles in India, it has long been indecisive about its relationship with India. The chipmaker company in India has a track record of waging bid wars, misleading state governments and denying deals after governments officially announcing a deal with the chip company. Machine Hopping The second reason for Foxconn to shift from making cars to iPhones is the dip in market for smartphones in recent years. It seems that the company plans to diversify its business. Similar to Foxconn, Apple has also been building an autonomous EV — dubbed Titan — for almost a decade which finally may be on the way. But thanks to its founder Steve Jobs, the company has long maintained its mysterious personality. Hence, no information whatsoever about the project has made it to mainstream media yet. The last insider information on the internet dates back to almost a year ago which is highly unlikely for similar significant projects in limbo. Interestingly, in a little over two decades, Apple has applied for 248 car-related patents and even hired Lamborghini’s top executive, Luigi Taraborrelli, to help them design their product in the car space. While Apple has been regularly launching refreshed versions of its slate products, the company’s car project has an uncertain future. Prominent Apple analyst Ming-Chi Kuo says Apple’s car plans have “lost all visibility,” and the Cupertino-based tech company must look at alternate strategies to make headways into a highly competitive automotive space. As the single-largest manufacturer of electronics ambitiously forges in the market, news from Apple should also be expected some time soon. But with Foxconn’s leap forward in the EV sector, there are a lot of ‘ifs’ involved. Even though the Apple-Foxconn relationship has long stood the tides of the tech industry, Foxconn has a registered history of making plans and backing out at the last moment. “Foxconn has the reputation for being one of the most opaque companies in an opaque world,” Lawrence Tabak, the author of ‘Foxconned: Imaginary Jobs, Bulldozed Homes, and the Sacking of Local Government’ described Foxconn’s habit of backing out of deals. “It is very normal for them to make stagey announcements that involve politicians, business executives, pomp, and circumstance purely based on speculation,” Tabak added. As of now, all that the stakeholders can do is to keep their fingers crossed and hope it does not turn out to be one of the company’s usual false promises to steer the electronic wheel.","excerpt":"Diversification opportunities are ample for Foxconn and the company is going all-in on electric cars next.","categories":["Global Tech"],"tags":[],"author_name":"Tasmia Ansari","publish_date":"2023-11-10T15:06:50","publication_year":"2023","word_count":683,"keywords":["Go","AWS","AI","cloud_platforms:AWS","programming_languages:R","programming_languages:Go","Aim","R"],"extracted_tech_keywords":["AI","Aim","AWS","R","Go","cloud_platforms:AWS","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/foxconn-to-steer-the-electric-wheel\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":31378,"title":"UK To Help India In Using AI, Big Data, Robotics In Healthcare Industry","content":"Representational image Noting the lack of proper healthcare facilities in India, UK Business Ambassador Sir Malcolm Grant, said that the British government is in talks with the Indian government regarding the same. Sir Grant especially emphasised the use of artificial intelligence, robotics and big data to provide healthcare solutions in the remote parts of the country, a report said. Sir Grant, who is the former chairman of England’s National Health Service, was in Bengaluru earlier this week. “There are difficulties to reach remote places in Karnataka. Using technology and robotics, we have been successful in diagnosing and performing robotic surgeries pertaining to dermal and pregnancy-related issues and we wish to bring the same in India, and specifically Karnataka,” he told a leading daily. The UK is the fourth largest investor in India and remains the largest investor into India outside of South East Asia and Africa. According to the Confederation of British Industry (CBI), between 2000 and 2018, total FDI which flowed into India from all channels from the UK was around $50.57 billion. Of this, the UK directly invested $26.09 billion in India — increasing its investment by $847 million between 2017 and 2018 – representing 7% of all foreign direct investment (FDI) into the country. Close to 38% of British companies made new investments in India in 2017. Whilst it is no longer the largest G20 investor in India – Japan narrowly overtook the UK – it is substantially ahead of Germany and France, who only contribute 3% and 2% in FDI respectively.","excerpt":"Noting the lack of proper healthcare facilities in India, UK Business Ambassador Sir Malcolm Grant, said that the British government is in talks with the Indian government regarding the same. Sir Grant especially emphasised the use of artificial intelligence, robotics and big data to provide healthcare solutions in the remote parts of the country, a […]","categories":["AI News"],"tags":["AI (Artificial Intelligence)","AI Healthcare","Healthcare Automation"],"author_name":"Prajakta Hebbar","publish_date":"2018-12-12T07:33:58","publication_year":"2018","word_count":255,"keywords":["big data","Go","artificial intelligence","programming_languages:R","AI","AI Healthcare","programming_languages:Go","Healthcare Automation","ai_applications:robotics","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","R","Go","big data","programming_languages:R","programming_languages:Go","ai_applications:robotics"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/uk-to-help-india-in-using-ai-big-data-robotics-in-healthcare-industry\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10142337,"title":"China to Replicate OpenAI’s o1 With O1-CODER","content":"Researchers from Beijing Jiaotong University developed ‘O1-CODER’ in an attempt to replicate OpenAI’s o1 model with a focus on enhancing coding tasks. Even though OpenAI’s o1 has gained significant recognition for its reasoning capabilities, it may not be the best option for programming and coding-related tasks. The O1-CODER framework incorporates reinforcement learning (RL) and Monte Carlo Tree Search (MCTS) techniques to improve System-2 thinking, which refers to a more deliberate and analytical form of reasoning. The researchers highlight a crucial lesson: data is all you need. Over the past decade, AI development has focused on improving model architectures, from traditional techniques like SVM and DNN to more recent advancements like Transformers. As models have grown, the focus has shifted to efficiently leveraging data. The o1 model and O1-CODER continue this trend by using RL to generate reasoning data, which can be utilised for System-2 tasks. This shift toward better data use is especially important for tasks requiring complex reasoning, like coding, where traditional datasets are not enough.Check out the code on GitHub. The researchers further noted that future versions will offer updated experimental results. These updates will likely provide insights into the model’s capabilities and improvements as it evolves. The Model Actually Understands Code The researcher behind O1-CODER explained how the model trains a Test Case Generator (TCG) to standardise code testing. It leverages MCTS to generate code with reasoning. This approach allows the model to tackle coding challenges systematically. The model starts by creating pseudocode, which serves as a blueprint, and then progresses to full code generation. This two-step process ensures the model understands the problem before starting to write the actual code. It first reasons through the problem, and then generates the solution. By combining Reinforcement Learning (RL) with MCTS, O1-CODER not only writes code but also learns to reason through the coding process. This approach helps the model solve more complex tasks. This combination allows the model to think deeply about how to structure coding solutions. Through iterative training, the model improves its performance, generating better and more efficient code over time. They emphasised that future versions of O1-CODER will focus on real-world applications. They believe adapting the model to real-world coding challenges is crucial for broader use. The researchers also said that O1-CODER is following a path similar to AlphaGo and its evolution toward generalisation. Much like AlphaGo evolved into AlphaGoZero and AlphaFold, o1-like models are expected to be applied to more complex, real-world tasks, such as embodied intelligence and physical environments. Environment Matters The paper also dwells on the need for updating the environment state, ensuring the model remains adaptable as it moves from research to real-world deployment. In addition to improving code generation, the authors propose generating test cases directly from coding questions. This method doesn’t rely solely on predefined datasets, enhancing the model’s flexibility. This approach can be used during the inference phase. It allows the model to reason online without needing predefined code, making it more adaptable to various situations. The paper suggested that O1-CODER could significantly impact AI’s approach to complex problem-solving. It aims to move beyond completing tasks to engaging in deeper reasoning and critical thinking. OpenAI’s o1 has encountered challenges in coding tasks in the past, leading to the emergence of several alternatives. o1 replication efforts: upper part from academic institutions and open source communities, and lower part from the industry. Source: arxiv.org Notably, Google’s Gemini 2 is anticipated to surpass o1 by integrating advanced reinforcement learning techniques and ‘Chain of Thought’ processes, aiming to improve reasoning and problem-solving abilities. Additionally, DeepSeek, a Chinese AI research lab, introduced the DeepSeek-R1-Lite-Preview model, which reportedly matched or exceeded o1 in complex tasks such as mathematics and coding.  In November, Alibaba also released its Marco-o1 to rival OpenAI o1. Even its recently released QwQ-32b model stands as a direct competitor to o1.","excerpt":"The researchers see O1-CODER following a path similar to AlphaGo and its evolution toward generalisation.","categories":["AI News"],"tags":["China","coding","OpenAI o1"],"author_name":"Sanjana Gupta","publish_date":"2024-12-03T16:56:26","publication_year":"2024","word_count":640,"keywords":["Go","OpenAI","AI","coding","Transformers","Git","RAG","Aim","OpenAI o1","chain of thought","GitHub","R","China"],"extracted_tech_keywords":["AI","OpenAI","Aim","Transformers","RAG","chain of thought","R","Go","Git","GitHub"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/china-to-replicate-openais-o1-with-o1-coder\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":59458,"title":"Top 10 Free Resources To Learn Cybersecurity","content":"With lots of data widespread among organisations, it has led to an increase in cybersecurity risk. According to sources, the global cybersecurity market was valued at USD 118.78 billion in 2018 and is expected to reach USD 267.73 billion by 2024, registering a CAGR of 14.5%, during the period of 2019-2024. In one of our articles, we discussed an ultimate guide to getting started with cybersecurity. In this article, we list down 10 free resources, which will help you learn cybersecurity to prevent security vulnerabilities. (The list is in no particular order) 1| Introduction to IT & Cybersecurity About: Introduction to IT & Cybersecurity is a free online course where you will learn about the four primary disciplines of information technology (IT) and cybersecurity. The course is a total of 4 hours and 21 minutes of clock time, and you will receive a certificate of completion upon finishing the training. The course includes topics such as introduction to system administration, network engineering, incident responses, offensive security and penetration testing. Source: Cybrary Get the course here. 2| Foundations of Cybersecurity About: This course will provide the foundational concepts for the cybersecurity field. It will also assist in examining various types of attacks. You will learn ways to protect environments through tools and design, as well as wrapping up looking at more advanced topics. You will learn about the most basic aspects of cybersecurity, including the impact of cyberattacks and the most common cybersecurity roles. By the end of this learning path, you will understand security fundamentals, including common threats and tools to prevent attacks, the basics of cryptography, such as public-key infrastructure, cybersecurity job market and other such. Source: Springboard Get the course here. 3| Stanford Advanced Computer Security About: There are video lectures provided by Stanford University, which includes new developments in cryptocurrencies and blockchain technologies, emerging trends in cybersecurity and other security lessons. Source: Stanford University Blog Get the lectures here. 4| Cybersecurity Basics: A Hands-on Approach About: This course presents an introduction to cybersecurity showing different aspects of this discipline. You will learn what the main existing cybersecurity threats are and how to protect yourself against them. The course presents a practical approach in which all required material will be provided to allow you to better understand attacks and establish appropriate countermeasures. Source: edX Blog Get the course here. 5| Foundations of Information Security About: In this course, you will learn the foundational skills needed to build a successful cybersecurity career, how to achieve basic security objectives (such as authentication, authorisation, access control, confidentiality, data integrity and non-repudiation) by using secure systems and design principles. You will also learn about major security breaches that have occurred over the years and how to build countermeasures to defend against such attacks. Source: Stanford University Blog Get the course here. 6| Cybersecurity Fundamentals About: In this introduction to the field of computing security, you will be given an extensive overview of the various branches of computing security. You will learn cybersecurity concepts, issues, and tools that are critical in solving problems in the computing security domain. You will also learn essential techniques in protecting systems and network infrastructures, analysing and monitoring potential threats and attacks, devising and implementing security solutions for organisations large or small. Source: edX Get the course here. 7| Cybersecurity and Executive Strategy About: In this course, you will learn about cyberattacks and the technical aspects of cybersecurity. This course will help cybersecurity professionals explain and convince the importance of cybersecurity to all levels of management and executive leadership. You will learn black-box and white-box assessments and metrics, prioritisation of vulnerabilities, breach preparation and other such. Source: Stanford University Blog. Get the course here. 8| Cybersecurity Economics About: This course provides an introduction to the field of cybersecurity through the lens of economic principles. Delivered by four leading research teams, this will provide you with the economic concepts, measurement approaches and data analytics to make better security and IT decisions, as well as understand the forces that shape the security decisions of other actors in the ecosystem of information goods and services. Source: edX Get the course here. 9| Introduction to Cybersecurity Tools & Cyberattacks About: This course gives you the background needed to understand basic cybersecurity. You will learn the history of cybersecurity, types and motives of cyberattacks to further your knowledge of current threats to organisations and individuals. Source: Coursera Get the course here. 10| Network Security About: In this course, you will examine the various areas of network security, including intrusion detection, evidence collection and defence against cyberattacks. You will learn the principles and concepts of wired and wireless data network security, analyse attack or defend scenarios and determine the effectiveness of particular defence deployments against attacks. Source: edX Get the course here.","excerpt":"With lots of data widespread among organisations, it has led to an increase in cybersecurity risk. According to sources, the global cybersecurity market was valued at USD 118.78 billion in 2018 and is expected to reach USD 267.73 billion by 2024, registering a CAGR of 14.5%, during the period of 2019-2024. In one of our […]","categories":["AI Trends"],"tags":["computer security","Coursera","Cybersecurity"],"author_name":"Ambika Choudhury","publish_date":"2020-03-23T14:00:00","publication_year":"2020","word_count":798,"keywords":["Go","programming_languages:R","AI","Coursera","programming_languages:Go","computer security","analytics","GAN","Cybersecurity","R"],"extracted_tech_keywords":["AI","analytics","R","Go","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-10-free-resources-to-learn-cybersecurity\/","complexity_score":3,"technical_depth":7,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":30354,"title":"India The Next Destination For FinTech, Blockchain The New Mantra: PM Modi","content":"Indian Prime Minister Narendra Modi this week gave a keynote address at the Singapore FinTech Festival. He emphasised the importance of fintech in the global economy and said that it should be a movement that can improve the lives of the world’s most marginalised people. “There is an explosion of fintech innovation and enterprise in India. It has turned India into a leading fintech and Startup nation in the world. The future of fintech and Industry 4.0 is emerging in India. Our youth are developing apps that are making the dream of paperless, cashless, presence-less, and yet safe and secure, transactions possible for all. That is the wonder of India Stack– simply the largest set of Application Programming Interface in the world. They are using Artificial Intelligence, Blockchain and machine learning to create solutions for banks, regulators and consumers. And, they are also embracing our nation’s social missions – from health and education to microcredit and insurance,” said PM Modi. Making a keen observation at the event, Modi explained the relationship between the startups, the investors and the funding. He said, “I have heard of an advice going around in start-up circles. To increase your Venture Capital or VC funding by 10 percent, tell the investors you run a platform, not a regular business; If you want to increase your VC funding by 20 percent, tell the investors that you are operating in the fintech space; But, if you really want the investors to empty their pockets, tell them that you are using blockchain!” At the event, PM Modi also launched global fintech platform Application Programming Interface Exchange (APIX) along with T Shanmugaratnam, the Deputy PM of Singapore.","excerpt":"Indian Prime Minister Narendra Modi this week gave a keynote address at the Singapore FinTech Festival. He emphasised the importance of fintech in the global economy and said that it should be a movement that can improve the lives of the world’s most marginalised people. “There is an explosion of fintech innovation and enterprise in India. It has […]","categories":["AI News"],"tags":["Blockchain","FinTech","Narendra Modi"],"author_name":"Prajakta Hebbar","publish_date":"2018-11-16T12:50:16","publication_year":"2018","word_count":278,"keywords":["Go","API","funding","artificial intelligence","Blockchain","machine learning","AI","innovation","venture capital","Narendra Modi","FinTech","R","startup"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","R","Go","API","innovation","startup","venture capital","funding"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/india-the-next-destination-for-fintech-blockchain-the-new-mantra-pm-modi\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":49497,"title":"DefMin Rajnath Singh Hails The Use Of AI, Blockchain In Modern Warfare","content":"Defence Minister Rajnath Singh this week hailed the role of emerging technologies like artificial intelligence, big data and blockchain in modern warfare. Addressing an envoy of over 80 countries in the backdrop of India’s upcoming defence exposition DefExpo 2020, Singh said that the character of warfare is changing and in this information age, the challenge was not just to prepare for contingencies but also to repel threats from multiple sources. Singh said, “India has a robust defence industrial base and has become a hub for repair, maintenance, overhaul and servicing industry for various platforms.” He urged the Defence Ministers of the countries present at the event to participate in the Defence Expo to be held in Lucknow from 5-8 February next year. Had a wonderful interaction with the Ambassadors\/DAs\/HoMs of several countries at the Ambassadors’ Round Table in New Delhi today. Invited them to attend the @DefExpoIndia at Lucknow in February 2020. pic.twitter.com\/Y8XGVfTNJB— Rajnath Singh (@rajnathsingh) November 4, 2019 Earlier this year in September, Singh had talked at an interactive session attended by 150 defence officials and had touched upon the need to use AI and cyber systems in modern warfare. All throughout his tenure as the defence minister, Singh has always placed importance on AI and emerging technologies like robotics. He has also talked about how they can help prepare India for future challenges. He had also emphasised the need to harness emerging technologies to bolster Indian defence forces capabilities across the board. Of late, Indian policymakers have taken a decisive stand on using AI. In early 2018, the Ministry of Defence (MoD) constituted a task force for implementation of AI. The multi-stakeholder task force worked towards laying down an institutional framework on capacity building and guidelines for defence machinery. In February 2019, the Ministry of Defence set up a Defence AI Council (DAIC) to provide strategic directions on how to bolster AI capabilities in defence and provide recommendations on building public-private partnerships with industry and startups. The role of AI and robotics was especially demonstrated during the surgical strikes carried out by the Indian Army against suspected militants in Pakistani-administered Kashmir.","excerpt":"Defence Minister Rajnath Singh this week hailed the role of emerging technologies like artificial intelligence, big data and blockchain in modern warfare. Addressing an envoy of over 80 countries in the backdrop of India’s upcoming defence exposition DefExpo 2020, Singh said that the character of warfare is changing and in this information age, the challenge […]","categories":["AI News"],"tags":[],"author_name":"Prajakta Hebbar","publish_date":"2019-11-06T13:08:54","publication_year":"2019","word_count":353,"keywords":["big data","artificial intelligence","programming_languages:R","AI","ViT","ai_applications:robotics","R","startup"],"extracted_tech_keywords":["AI","artificial intelligence","R","big data","ViT","startup","programming_languages:R","ai_applications:robotics"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/defmin-rajnath-singh-hails-the-use-of-ai-blockchain-in-modern-warfare\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10091298,"title":"A Toolkit for the Tricky World of AI Policing","content":"In the wake of growing biases and misuse of AI, the United Nations Interregional Crime and Justice Research Institute (UNICRI) and the INTERPOL have come up with a toolkit for responsible AI innovation in law enforcement to support and guide enforcement authorities around the world. This toolkit would help them deploy and use AI ethically and responsibly. “Even though the use of AI by law enforcement is perhaps one of the most-sensitive domain applications of the technology, there is no guidance available for law enforcement agencies to help them get it ‘right’. In our toolkit, we have tailored practical and operationally oriented guidance suited to their needs and requirements,” said Odhran McCarthy, project officer at UNICRI’s Centre for AI and Robotics. The toolkit will consist of seven resources, for eg, guidance documents such as an introductory guide, principles for responsible use, and an organisational roadmap. It will also include interactive tools like an organisational readiness assessment and AI risk assessment tool. Additionally, there will be target audience-specific briefs called AI for Law Enforcement Primers. AI not only helps with creating statistical correlation between vast amounts of data to sort and manage information, but also aids policing using technology like the vastly-debated facial recognition systems, and predictive policing systems. McCarthy told AIM that they have identified several countries around the world which would support them in testing of the toolkit. “Testing is a key part of our work to ensure that the toolkit is useful and practical from a law-enforcement standpoint.” The first version of the toolkit will be released in June and the final version will be out around October, this year. What about India? The policing system in India has been one of leading adopters of AI across the globe. However, these technologies are being deployed without a legal regime and are unsettling for civil society groups, like IFF, that are raising their voice for internet liberties. Recently, the Bangalore City Police installed 4,100 out of 7,000 CCTVs equipped with facial recognition systems with the help of Honeywell Automation India as part of the Bengaluru Safe City project. As part of the project, 30 ‘safety islands’, 8 drones, 400 body-worn cameras, and a mobile command centre is also available at the disposal of the state law enforcement. AIM reached out to Honeywell India Limited regarding the ethical concerns and reliability issues around this, but a response is still awaited from them. Meanwhile, according to a study conducted by TechSci, India’s facial recognition market is projected to grow six-fold to reach USD 4.3 billion by 2024, putting it in close competition with China’s state surveillance system. Amnesty International has labelled Telangana as “the most surveilled state in the world” with 600,000 cameras, mostly concentrated in Hyderabad. New Delhi, Hyderabad, Chennai, and Indore are among the cities in India with the highest number of surveillance cameras, with New Delhi topping the list in terms of cameras per square mile. The Kolkata Police have also announced plans to install 2,500 CCTV cameras that will utilise AI to detect bikers without helmets and illegal parking. On the other hand, the Bihar Police plan to utilise predictive policing and AI technology to combat illicit liquor trade and other crimes. However, there is limited evidence to support the notion that surveillance has had a negative impact on crime rates. According to a report from the California Research Bureau, despite the increased prevalence of CCTV cameras, there is little conclusive evidence that they have led to a decrease in crime. Similarly, a study conducted by the Arizona State University found no significant effect on criminal activity in a city as a result of the installation of surveillance cameras. Policing the AI Future Responsibly Maknoon Wani, a research associate for emerging tech and policy with the Council for Strategic and Defense Research explained how these technologies are not perfect and the deployment is being rushed, “There are several issues with deploying facial recognition systems and technologies in public places. Basically, these are still developing technologies and are not perfect yet. But the law enforcement agencies see these as sophisticated technologies that they can deploy to get better results and achieve efficiency. In reality, that really doesn’t actually happen, because these technologies are inefficient. These facial recognition systems can give false positives and false negatives, and that’s a major issue,” he said. He further said that not having any laws governing these technologies could result in human rights violations and false incriminations. “All of these technologies are being developed and deployed without a legal regime which can offer safeguards. For example, if a person is arrested in a public place based on the fact that the facial recognition technology identified them as a criminal or having a criminal record, then they do not have a way to get a redressal.” However, data from a report by the US government’s National Institute of Standards and Technology suggested that the top facial recognition algorithms in the world are highly accurate and have marginal differences in their rates of false-positive or false-negative readings across demographic groups. The top 150 algorithms used for the test on the focus group were over 99% accurate across black male, white male, black female and white female demographics. So the controversy around the accuracy and bias in these systems is a point of contention, but given India’s enthusiasm on AI in policing and law enforcement, the toolkit for responsible AI innovation in law enforcement should be considered and employed to address these issues and ensure minimal bias, and trampling of human rights and individual privacy.","excerpt":"The policing system in India has been one of the leading adopters of AI. But, these technologies being deployed without a legal regime is unsettling for civil society groups","categories":["AI Trends"],"tags":["bias","Responsible AI"],"author_name":"Shyam Nandan Upadhyay","publish_date":"2023-04-13T12:30:00","publication_year":"2023","word_count":929,"keywords":["Go","AWS","AI","R","Responsible AI","BERT","responsible AI","automation","Aim","ViT","GAN","bias"],"extracted_tech_keywords":["AI","Aim","AWS","R","Go","BERT","GAN","ViT","responsible AI","automation"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/a-toolkit-for-the-tricky-world-of-policing-with-ai\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10143319,"title":"Amazon Smbhav Commits $120 Mn to Boost Indian Manufacturing, AI Startups","content":"Amazon India has announced an investment of $120 million from its Smbhav Venture Fund to support Indian startups in manufacturing and AI. This initiative aligns with the government’s goal of establishing India as a global manufacturing hub while creating substantial employment opportunities. The announcement was made during the fifth edition of Amazon’s annual Smbhav summit in New Delhi, highlighting the company’s commitment to advancing India’s digital and economic aspirations under the theme Naye India ki Nayi Raftaar. Launched in 2021, the Smbhav Venture Fund began with a focus on agriculture and healthcare but has since expanded to direct-to-consumer brands, fintech, gaming, and now, manufacturing. With the latest funding, the Smbhav Fund grew to $350 million, aiming to digitise consumer goods manufacturing for domestic and global markets. Startups such as The Good Glamm Group, FreshToHome, and Acko Insurance, previously funded by the initiative, have demonstrated significant growth through technology adoption. In partnership with the Department for Promotion of Industry and Internal Trade (DPIIT), the fund supports the government’s Viksit Bharat vision of transforming India into a developed economy by 2047. At the summit, Amazon outlined its broader ambitions for India, including digitising 10 million small businesses, generating 2 million jobs, and achieving $20 billion in e-commerce exports by 2025. Amazon founder and former CEO Jeff Bezos headlined the inaugural Smbhav event in January 2020, announcing a $1 billion investment. During the event, Amazon unveiled its India commitments to digitise 10 million SMBs, facilitate $10 billion in exports, and generate 2 million jobs by 2025. Amazon for Indian AI In September, AWS selected seven Indian generative AI startups for their accelerator program and received $1M in AWS credits. The Indian startups Convrse, House of Models, Neural Garage, Orbo.ai, Phot.ai, Unscript AI, and Zocket were among the 80 companies selected globally for their use of AI and global growth potential. The 10-week accelerator program connects startups with business and technical mentors from AWS and NVIDIA, providing them with resources to develop, train, test, and scale their AI solutions efficiently. Besides Amazon, big tech companies, including Google, Microsoft, and Reliance, have vested in specific AI cohort programs to boost startups working in the space.","excerpt":"In September, AWS selected seven Indian generative AI startups for their accelerator program that received $1M in AWS credits.","categories":["AI News"],"tags":["Aamzon Smbhav","AI accelerator","AI Startups","Google","India","Microsoft","Reliance","Smbhav Venture Fund"],"author_name":"Vandana Nair","publish_date":"2024-12-11T17:26:54","publication_year":"2024","word_count":360,"keywords":["Go","Smbhav Venture Fund","Microsoft","AI","AWS","Git","RAG","Aim","Aamzon Smbhav","Google","generative AI","GAN","Reliance","AI accelerator","R","India","AI Startups","startup"],"extracted_tech_keywords":["AI","generative AI","Aim","RAG","AWS","R","Go","Git","GAN","startup"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/amazon-smbhav-commits-120-mn-to-boost-indian-manufacturing-ai-startups\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10170014,"title":"Google DeepMind Launches AlphaEvolve, New AI Coding Agent for Maths and Science","content":"Google DeepMind has launched AlphaEvolve, a new coding agent that uses large language models to evolve and optimise algorithms across computing and mathematics. Powered by Gemini Flash and Gemini Pro, AlphaEvolve pairs model-generated code with automated evaluators to verify, score, and evolve high-performing solutions. “AlphaEvolve is an agent that can go beyond single function discovery to evolve entire codebases and develop much more complex algorithms,” stated Google DeepMind in its blog post. Google DeepMind is planning an early access programme for selected academic users and is also exploring ways to make AlphaEvolve more broadly available. Interested users can register their interest through a dedicated form. The company believes AlphaEvolve could be transformative across multiple fields, including materials science, drug discovery, sustainability, and broader technological and business applications. The system integrates prompt sampling, language model outputs, and program evaluation through an evolutionary algorithm framework. Over the past year, AlphaEvolve has been used to improve data centre scheduling, hardware design, and AI training workflows across Google. One deployment optimised Borg, Google’s data centre orchestrator, recovering 0.7% of compute resources globally. “This solution, now in production for over a year, continuously recovers, on average, 0.7% of Google’s worldwide compute resources,” the company said. AlphaEvolve also contributed to a Tensor Processing Unit (TPU) design. It suggested a Verilog-level change that removed redundant bits in a key arithmetic circuit.  Google said this proposal passed verification tests and was integrated into an upcoming TPU release. In AI training, AlphaEvolve optimised matrix multiplication in the Gemini architecture, speeding up a core kernel by 23% and cutting training time by 1%. It also improved FlashAttention kernel performance by 32.5%, a domain typically untouched by human engineers due to compiler-level optimisation. Beyond infrastructure, AlphaEvolve tackled algorithmic challenges in mathematics. It discovered a new method to multiply 4×4 complex-valued matrices using 48 scalar multiplications, improving on the 1969 Strassen algorithm. “This finding demonstrates a significant advance over our previous work, AlphaTensor,” the company said. Applied to over 50 open problems across mathematics, AlphaEvolve rediscovered known solutions in 75% of cases and improved on 20%. One of its advances was in the kissing number problem, where it found a configuration of 593 spheres touching a unit sphere in 11 dimensions, establishing a new lower bound.","excerpt":"Google DeepMind is planning an early access programme for selected academic users and is also exploring ways to make AlphaEvolve more broadly available.","categories":["AI News"],"tags":["Google Deepmind"],"author_name":"Siddharth Jindal","publish_date":"2025-05-14T21:53:52","publication_year":"2025","word_count":374,"keywords":["Gemini Pro","Go","TPU","programming_languages:R","AI","Scala","RAG","Google Deepmind","llm_models:Gemini","R","Redis"],"extracted_tech_keywords":["AI","Gemini Pro","RAG","TPU","Redis","R","Go","Scala","llm_models:Gemini","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/google-deepmind-launches-alphaevolve-new-ai-coding-agent-for-maths-and-science\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":13912,"title":"Zoho’s Zia, the virtual sales assistant is reshaping sales","content":"Sridhar Vembu, CEO, Zoho Corporation In the world of many Einsteins’ (Saleforce Einstein is AI built into CRM platform), Zoho wishes to be the Nikola Tesla of the CRM world. And rightly so, with a user base of 300,000 companies worldwide over the last decade, the growing popularity of Zoho has made it every salesperson’s go-to-tool. Now sales process just got smarter with the launch of Zia, the AI powered sales assistant that suggest workflows, detects anomalies and much more. While analytics and forecasting have become a de facto industry standard when it comes to CRM software, Zia uses machine learning and data mining to not only learn from existing data but also adapt to the information its CRM is fed every day. The result – the numbers are always realistic, cites Vibhav Vankayala, Product Marketing Manager – Zoho CRM. Inside Zia – the virtual sales assistant Zia is more just AI, she’s a sales assistant to every CRM user, shares Vankayala. “She’s with you right from contacting prospects to letting you know how a sale has affected your business goals,” he adds. This is how Zia, the newest AI powered sales assistant on the block streamlines workflows and boosts sale: Through a series of dashboards, Zia shows how successful the sales activities have been It spots anomalies that shouldn’t be happening, enabling the salesperson spot mistakes before it’s too late It also speeds up the pipeline velocity, updating when the clients should be contacted, instead of dealing with voicemails and unopened emails. It speeds up how to use CRM overall, by automating the tasks one would ordinarily do, and optimize performance AI reshaping sales with granular insights From actionable insights to smart recommendations that improves sales performance, Zia covers the whole gamut of sales process. “Zia’s dashboards showcase the activities that go into both pre-sales and post-sales (called Sales Trends and Sales Follow-up Trends). The dashboards have a trend prediction of how each activity is expected to turn out based on the history and the data stored on your software,” explains Vankayala. The virtual sales assistant also keeps up with the changing trends and keeps a tab on seasonal variations, enabling sales workforce to better their performance around holidays and other important events.   “Since Zia works in real time, the trends also keep changing. If there are seasonal variations, it keeps that in mind too and is smart enough to figure out that it’s just how your business works – like understanding that Monday sales are expected to be low or that sales should be high during a holiday season. Zia also occasionally gives you a tip or brings something to your attention when she knows you could be doing something better,” he shared. Case in point how Zia productizes the workforce with macros: It’s a known fact that salespeople spend hours on their CRM, Zia wants you to make that time productive, says Vankayala. Instead of scrolling through the CRM, applying the same changes to every contact, Zia figures out what one would normally do and makes macro suggestion. “These suggestions allows the person to perform a series of tasks at a click of button. Macros aren’t new to CRM, but some salespeople don’t know how to create one or don’t want to take the trouble of doing it,” he adds. Is speech recognition on the way? AIM quizzed Vankayala whether speech recognition capability for voice calls is in the pipeline? At the moment, Zia communicates via notifications and data. “Maybe down the line, we might move onto speech recognition but for now, Zia is focused on getting even smarter for the everyday salesperson,” he said. And what sets it apart from market leaders such as Salesforce and Microsoft is that it is affordable yet reliable, working hard in the background, with or without the recognition. “In a world of many Einsteins’, we’d rather be Nikola Tesla,” he affirmed. Blueprint making all the buzz Blueprint is technology built in the CRM that helps sales managers make their sales processes more compliant, repeatable and customized. As the title typifies, it acts as a guide for users, outlining what actions should be taken at what time. “The best feature to compare Blueprints is to workflows. Workflow automation is a string of activities, set up individually, that will be triggered once an action is done,” explains Vankayala. In other words, Blueprints are processes that have multiple outcomes and are set up so that one can follow different strings of actions at each stage. Blueprints can be set for the whole organization and allow contextual input at every stage to guide sales team. Understanding the AI push to CRM Just like any other industry, Artificial Intelligence is shaping CRM but what has triggered the rise of AI in this sector. According to a Forrester research, AI is driving the insight revolution and is going to steal a march over “less-informed peers”. Data and insights is going to become a coordinated strategy and enterprise-wide initiative, and where AI trumps traditional analytics is its scalability, cites Ajay Kashyap, co-founder of Boxx.ai. He elaborates why AI is emerging as the definitive technology , especially for customer-focused businesses who are in-line for AI revolution. Interestingly, Forrester has predicted a 300% increase in AI investment in 2017. In sales, the disruptive power of AI is seen in smart recommendations, filtering leads and giving cues on the right time to call among other aspects.","excerpt":"In the world of many Einsteins’ (Saleforce Einstein is AI built into CRM platform), Zoho wishes to be the Nikola Tesla of the CRM world. And rightly so, with a user base of 300,000 companies worldwide over the last decade, the growing popularity of Zoho has made it every salesperson’s go-to-tool. Now sales process just […]","categories":["IT Services"],"tags":[],"author_name":"Richa Bhatia","publish_date":"2017-03-30T11:07:58","publication_year":"2017","word_count":906,"keywords":["Go","API","machine learning","artificial intelligence","AI","ML","Scala","Aim","analytics","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","analytics","Aim","R","Go","Scala","API"],"url":"https:\/\/analyticsindiamag.com\/it-services\/zohos-zia-virtual-sales-assistant-reshaping-sales\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10138292,"title":"HCLTech’s GenAI Suite ‘AI Force’ Gains Momentum with 25 Clients Onboard","content":"HCLTech is seeing traction in its AI offerings, as highlighted by CEO C Vijayakumar during the company’s Q2 2025 earnings call. He shared that AI has been integrated into the majority of their recent deals, highlighting the platform’s key role in service transformation. “We have seen strong wins with most of our deals incorporating AI capabilities. Our platform AI Force is gaining widespread adoption for service transformation among our clients,” said Vijayakumar. He further emphasised how the company’s AI offerings, such as HCLTech AI Force, HCLTech AI Foundry, and AI Labs, are driving innovation across sectors. A key differentiator for HCLTech is the integration of AI Force with Microsoft GitHub Copilot, which will soon be extended as a Copilot extension. “AI Force is now integrated with Microsoft GitHub Copilot and will soon provide unmatched extensibility and a broad range of use cases,” said Vijayakumar. HCLTech has also partnered with AWS and Google Cloud to accelerate the adoption of generative AI solutions. Vijayakumar also revealed that AI Force is now integrated with Anthropic Claude 3 on Amazon Bedrock. “These developments position us strongly in the AI landscape and demonstrate our commitment to innovation as well as service,” he said. The CEO outlined the company’s three-pronged generative AI approach. The first dimension is AI Force, which is driving service transformation and adoption among existing clients. “We have almost 25 clients who have started working on AI Force as the primary efficiency platform for all their IT and business processes,” Vijayakumar shared. The second focus is HCLTech AI Foundry, which is fostering data-led innovation. Vijayakumar cited numerous examples where the Foundry is either in proof-of-concept stages or nearing production scale. The third area is AI Labs, where clients can bring in ideas and leave with a minimum viable product, driving continuous experimentation. Looking ahead, HCLTech anticipates a growing pipeline of generative AI projects across multiple verticals and geographies. “Our pipeline continues to grow robustly, demonstrating a strong and well-distributed presence across business segments, verticals, and geographies,” Vijayakumar concluded. HCLTech reported a 11 percent year-on-year growth in net profit for Q2 FY25, amounting to Rs 4,235 crore, exceeding market forecasts. Revenue from operations for the July-September 2025 quarter stood at Rs 28,862 crore, reflecting an 8.2 percent increase compared to the same period last year, maintaining its rank as India’s third-largest IT company. “We delivered a strong quarter with revenue growing 1.6% QoQ in constant currency and EBIT coming in at 18.6%. This growth was well distributed across verticals, geographies, and offerings. HCL Software has delivered a stellar performance of 9.4% YoY this quarter and 6.4% growth in H1 FY25 in constant currency, demonstrating the increasing relevance of our products for the digital economy, said VijayKumar. Tata Consultancy Services (TCS)  in its recent earnings call announced that over 600 AI and generative AI engagements have been deployed successfully in production or in various phases of development so far. The company said it is seeing continued momentum with Gen AI adoption and the technology maturing at a rapid pace. “Our customers are increasingly concentrating on integrating AI throughout their entire enterprise value chain rather than working on isolated use cases. We are also enabling AI\/GenAI capabilities across our suite of products and solutions,” TCS said in its FY25 second-quarter earnings release.","excerpt":"HCLTech has also partnered with AWS and Google Cloud to accelerate the adoption of generative AI solutions.","categories":["AI News"],"tags":["HCL Technology"],"author_name":"Siddharth Jindal","publish_date":"2024-10-15T11:34:05","publication_year":"2024","word_count":546,"keywords":["Anthropic","Go","GenAI","API","AWS","AI","Git","generative AI","HCL Technology","GitHub","R"],"extracted_tech_keywords":["AI","generative AI","GenAI","Anthropic","AWS","R","Go","Git","GitHub","API"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/hcltechs-genai-suite-ai-force-gains-momentum-with-25-clients-onboard\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10134435,"title":"Alibaba Launches Qwen2-VL, Surpasses GPT-4o &amp; Claude 3.5 Sonnet","content":"Alibaba recently released Qwen2-VL, the latest model in its vision-language series. The new model can chat via camera, play card games, and control mobile phones and robots by acting as an agent. It is available in three versions: the open source 2 billon and 7 billion models, and the more advanced 72 billion model, accessible using API. The advanced 72 billion model of Qwen2-VL achieved SOTA visuals understanding across 20 benchmarks. “Overall, our 72B model showcases top-tier performance across most metrics, often surpassing even closed-source models like GPT-4o and Claude 3.5-Sonnet,” read the blog, saying that it demonstrates a significant edge in document understanding. Qwen2-VL performs exceptionally well in benchmarks like MathVista (for math reasoning), DocVQA (for document understanding), and RealWorldQA (for answering real-world questions using visual information). The model can analyse videos longer than 20 minutes, provide detailed summaries, and answer questions about the content. Qwen2-VL can also function as a control agent, operating devices like mobile phones and robots using visual cues and text commands. Interestingly, it can recognise and understand text in images across multiple languages, including European languages, Japanese, Korean, and Arabic, making it accessible to a global audience. Check out the model here. Key highlights: One of the key architectural improvements in Qwen2-VL includes the implementation of Naive Dynamic Resolution support. The model can adapt to and process images of different sizes and clarity. “Unlike its predecessor, Qwen2-VL can handle arbitrary image resolutions, mapping them into a dynamic number of visual tokens, thereby ensuring consistency between the model input and the inherent information in images,” said Binyuan Hui, the creator of OpenDevin and core maintainer at Qwen. He said that this approach more closely mimics human visual perception, allowing the model to process images of any clarity or size. Another key architectural upgrade, according to Hui, is the innovation of Multimodal Rotary Position Embedding (M-ROPE). “By deconstructing the original rotary embedding into three parts representing temporal and spatial (height and width) information，M-ROPE enables LLM to concurrently capture and integrate 1D textual, 2D visual, and 3D video positional information,” he added. In other words, this technique enables the model to understand and integrate text, image and video data. “Data is All You Need!” said Hui. The use cases are plenty. William J.B. Mattingly, a digital nomad on X recently praised this development calling it his new favorite Handwritten Text Recognition (HTR) model while trying to convert a handwritten text into digital format. Ashutosh Shrivastava a user on X  used this model to solve a calculus problem and reported successful results for the same, proving its validity in problem solving. GitHub flagged the org for unknown reasons but the update is available on Hugging Face.","excerpt":"The advanced 72 billion parameter version of Qwen2-VL suprasses GPT-4o and Claude 3.5 Sonnet.","categories":["AI News"],"tags":["Alibaba","Generative Pre-Trained Transformer"],"author_name":"Tanisha Bhattacharjee","publish_date":"2024-09-04T17:30:37","publication_year":"2024","word_count":448,"keywords":["Hugging Face","API","AI","Alibaba","GPT-4o","RPA","Git","GPT","Generative Pre-Trained Transformer","GitHub","Claude 3.5","R"],"extracted_tech_keywords":["AI","GPT-4o","Claude 3.5","Hugging Face","R","Git","GitHub","API","GPT","RPA"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/alibaba-launches-qwen2-vl-surpasses-gpt-4o-claude-3-5-sonnet\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10143683,"title":"India Plans to Launch Advanced Quantum Satellite in 2-3 Years","content":"India is set to join an elite group of countries with quantum satellite capabilities, aiming to secure communication networks against hacking and cyberattacks. The satellite will be crucial in the larger quantum communications network under the National Quantum Mission (NQM). “Satellite-based communication will be required to secure country-wide or international communication. The Department of Space is planning to launch a quantum satellite in 2-3 years for quantum communications,” Ajai Chowdhry, co-founder of HCL and chairman of the mission governing board for NQM, told The Hindu Businessline. The NQM has organised its work across four verticals: computing, communication, measurement, and sensing. The communications vertical aims to create a secure network resistant to breaches by quantum computers. Satellite-based communication will complement the optical fibre-based network by addressing distance limitations. While optical fibres can transmit photons up to 250 km, satellites can enable longer-range communication. Photons lose energy after 100-250 km in optical fibres, requiring costly trusted nodes every 150 km. Satellites, with a wider reach, are more practical. “We plan to focus on satellites in the next 6-12 months,” said Sunil Gupta, co-founder of QNu Labs. Quantum satellites use quantum key distribution (QKD) technology to secure information transmission. Countries like China, the US, and members of the European Union have already launched such programmes. In terms of breach footprints, Dr. Urbasi Sinha, head of the Quantum Information and Computing (QuIC) laboratory at RRI, told AIM, “In quantum, you cannot do anything without leaving a mark. When we execute the protocol, we can determine whether there has been any eavesdropping.” Dr. Sinha’s work in quantum communication, particularly in QKD, also led to the recent paper, which presented simulation results after analysing link performance at three different locations for a quantum communication centre in India. In Parliament, Jitendra Singh, Minister of State for Science and Technology, announced that the government will develop satellite-based secure quantum communication between two ground stations over a range of 2,000 km within India and internationally. Gupta also highlighted challenges with low-earth orbit satellites, noting their limited 15–20 minutes of daily coverage. “A constellation of 4–7 satellites is needed, and either ISRO or private players like Dhruva must prioritise their launch,” he said. Chowdhry expressed confidence in the mission’s progress, with the Union Cabinet allocating ₹6,003.65 crore over eight years. Experts have urged the government to accelerate the project to keep pace with global advancements in quantum technology. The integration of quantum computing and AI is anticipated to revolutionise various sectors, including drug discovery and cybersecurity. Chowdhry highlighted that quantum computers could significantly reduce the time and cost associated with drug development. In an interview with AIM, he had earlier warned, “When a quantum computer that can crack the current cyber security systems comes up, you will be completely open to attacks.” Major tech companies like Google and Microsoft are actively developing this combination to drive the next wave of the AI revolution. Google Deepmind’s new ‘AlphaQubit’ AI-based system can accurately identify errors inside quantum computers. Furthermore, future versions of AlphaFold could leverage quantum algorithms to process complex biological data and investigate protein structure spaces that classical methods cannot compute. The NQM’s focus on quantum communication and computing aims to prepare India for these emerging challenges and opportunities. By developing indigenous quantum technologies, India seeks to safeguard its digital infrastructure and enhance its technological capabilities.","excerpt":"The Union Cabinet allocates ₹6,003.65 crore over eight years for India’s quantum mission.","categories":["AI News"],"tags":["quantum communication","quantum key distribution"],"author_name":"Sanjana Gupta","publish_date":"2024-12-16T22:11:47","publication_year":"2024","word_count":555,"keywords":["quantum key distribution","Go","programming_languages:R","AI","quantum communication","Git","programming_languages:Go","RAG","Aim","Rust","GAN","R"],"extracted_tech_keywords":["AI","Aim","RAG","R","Go","Rust","Git","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/india-plans-to-launch-advanced-quantum-satellite-in-2-3-years\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10099863,"title":"Apple Will Sell Made-in-India iPhone on Launch Day","content":"In another groundbreaking announcement from India, Apple Inc. is set to release its latest iPhone 15 model, assembled in India, on the day of its global launch. This marks a significant shift from the company’s traditional strategy of primarily selling Chinese-manufactured devices worldwide. Apple plans to make the India-built iPhone 15 available in India and select global markets on the day of its global sales debut, according to reports from Bloomberg. While the majority of iPhone 15 units will still originate from China, this move showcases India’s burgeoning manufacturing capabilities and underscores Apple’s efforts to diversify its production beyond China. Production of the iPhone 15 commenced last month at Foxconn Technology Group’s factory in southern Tamil Nadu, India. However, potential logistical challenges may result in slight delays for the India-assembled devices. Apple is scheduled to unveil the iPhone 15, alongside updated watches and AirPods, at a gala event at its headquarters in Cupertino, California. Typically, new products become available for purchase approximately 10 days after their unveiling. Before the iPhone 14, only a small fraction of Apple’s global production occurred in India, with a substantial lag compared to Chinese production. However, last year saw a significant reduction in this delay, with India’s share of iPhone assembly reaching 7% by the end of March. This shift can be attributed to Prime Minister Narendra Modi’s incentives to boost local manufacturing and Apple’s strategy to reduce its reliance on China amid trade tensions between Washington and Beijing. The iPhone 15 is expected to be a major update, featuring camera system enhancements across the range and an improved 3-nanometer processor in the Pro models. This release is crucial for revitalizing Apple’s sales, as the company reported declining sales in key markets like the US, China, and Europe in recent quarters. Furthermore, other Apple suppliers in India, including Pegatron Corp. and a Wistron Corp. factory soon to be acquired by the Tata Group, are likely to join in assembling the iPhone 15. Apple, which recently opened its first retail stores in India, views the country as both a growing retail market and a vital production hub for its products. During the quarter ending in June, iPhone sales in India experienced double-digit growth, although specific figures were not disclosed by Apple.","excerpt":"Production of the iPhone 15 commenced last month at Foxconn Technology Group’s factory in southern Tamil Nadu.","categories":["AI News"],"tags":[],"author_name":"Mohit Pandey","publish_date":"2023-09-12T14:54:25","publication_year":"2023","word_count":375,"keywords":["programming_languages:R","AI","Git","ViT","R"],"extracted_tech_keywords":["AI","R","Git","ViT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/apple-will-sell-made-in-india-iphone-on-launch-day\/","complexity_score":3,"technical_depth":5,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":53514,"title":"How This UK-Based Fintech Startup Is Using AI Credit Platform To Provide Debt Finance To SMEs","content":"With each passing year, the fintech sector is providing faster, flexible and secured consumer experience, and is protecting against the risks and vulnerabilities of traditional insurance and loans. In fact, the global fintech market size is expected to grow to $124.3 billion by the end of 2025 at a CAGR of 23.8%. With a vision of providing small and medium-sized growth companies with debt finance, along with aiding them in competing against the large corporations, UK-based OakNorth is utilising artificial intelligence and machine learning to fulfil the dream. Since its inception, OakNorth has secured over $1 billion from leading investors which has been used to launch lending operations and others such. Fintech startup, OakNorth was founded by Rishi Khosla and Joel Perlman in May 2015. After meeting each other at college in the year 2002, Khosla and Perlman decided to launch their own business that could solve the challenges they had to face in securing debt finance from high street banks during their previous business — Copal Amba which scaled to 3,000 employees and was later acquired by Moody’s Corporation in 2014. In India, OakNorth has over 450 people across its two offices in Gurugram and Bengaluru. Currently, there are 12 banks and financial institutions using their platform across Asia, North America, Australia, and Europe, with a total asset of $12 billion. According to the founders, the team in India includes some of the best credit analysts, along with engineers and data scientists who integrate into a global product team, all focussed on developing the platform. Flagship Product OakNorth’s next-generation credit platform allows traditional financial institutions to improve and accelerate their credit decision and monitoring capabilities significantly. The platform provides improved efficiency, faster growth, and premium pricing. Use of AI\/ML According to Khosla, the platform works by collecting millions of data items on SMEs across various parameters, sectors, and markets. It then uses machine learning algorithms to identify data that lenders require to make more informed credit decisions through a detailed line item underwriting. The team of credit analysts and data scientists gets to manage the process and train the machine learning algorithms so that the platform continues to evolve and get smarter with each day. The platform, therefore, provides OakNorth Bank and its partners around the world with significant process efficiency gains, data-driven decision making, and smarter credit analysis and capabilities. Core Tech Stack Asking about the cloud, the founders replied that with the help of Amazon Web Services (AWS), the company can work with the FCA on policy items around data protection, access to data, security and business continuity. In May 2016, it became the UK’s first bank to be fully cloud-hosted. Tackling Hiring Phase With regards to hiring candidates, Khosla and Perlman said, “There are numerous fintech companies across the country that are looking for talent, but the key is to focus on opportunities with businesses with the mission that you believe in and who you believe in standing the test of time.” He further added, “If anyone is interested in speaking to us about any of our openings, then one should feel free to reach out to me on LinkedIn, and I would be delighted to speak to them!” Roadmap The future roadmap of OakNorth is to continue the growing loan book in the UK, as well as expand the SME loan origination and credit capabilities of North American banks, and lending institutions as well as helping them build and scale the quality of the SME loan books.","excerpt":"With each passing year, the fintech sector is providing faster, flexible and secured consumer experience, and is protecting against the risks and vulnerabilities of traditional insurance and loans. In fact, the global fintech market size is expected to grow to $124.3 billion by the end of 2025 at a CAGR of 23.8%. With a vision […]","categories":["AI Startups"],"tags":["FinTech","fintech analytics India"],"author_name":"Ambika Choudhury","publish_date":"2020-01-10T13:05:01","publication_year":"2020","word_count":582,"keywords":["Go","machine learning","artificial intelligence","AWS","AI","fintech analytics India","data-driven","ML","cloud_platforms:AWS","FinTech","R","startup"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","AWS","R","Go","data-driven","startup","cloud_platforms:AWS"],"url":"https:\/\/analyticsindiamag.com\/ai-startups\/how-this-uk-based-fintech-startup-is-using-ai-credit-platform-to-provide-debt-finance-to-smes\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10046078,"title":"IIM Jammu To Roll Out Executive Programme In ML &#038; AI","content":"Indian Institute of Management (IIM), Jammu, will be rolling out its Executive Programme in Machine Learning and AI for Businesses from Sept 30, 2021. This program has been designed for professionals who want to gain in-depth knowledge of artificial intelligence (AI) and machine learning (ML) to assist businesses with a competitive edge. Candidates with a graduate degree or diploma holders from a recognised university in any discipline are welcome to apply. This will be a 7-month course where students will be provided hands-on training on Python, R programming, Tableau, etc. They will also be taught statistics for business, machine learning and artificial intelligence, big data and analytics, exploratory data analysis and visualisation, AI at the production level, etc. The program will help the candidates adapt to changing customer needs, contribute to the development of AI-based solutions, and modify business strategies. The students will be exposed to chosen real-life studies from Harvard Business School and IIM Bangalore and a capstone project to help them apply classroom learning to real-world challenges. Interested students can apply to the program before Aug 23, 2021 by visiting this link. Selected candidates will be required to pay a program fee of Rs 64,500 (plus GST), starting with a booking fee of Rs 21,500 (plus GST) within seven days of selection.","excerpt":"This will be a 7-month course where students will be provided hands-on training on Python, R programming, Tableau, etc.","categories":["AI News"],"tags":[],"author_name":"Shraddha Goled","publish_date":"2021-08-16T11:55:39","publication_year":"2021","word_count":214,"keywords":["big data","machine learning","artificial intelligence","programming_languages:R","AI","ML","Python","analytics","programming_languages:Python","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","analytics","Python","R","big data","programming_languages:Python","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/iim-jammu-to-roll-out-executive-programme-in-ml-ai\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10061870,"title":"83 per cent of Indian BFSI players think AI is compulsory for better customer experience","content":"A survey conducted by PwC in collaboration with FICCI showed that about 83 per cent of the respondents believed that AI is key to offering a better customer experience in Indian financial service companies. Titled ‘Uncovering the ground truth: AI in Indian financial services,’ the respondents included people from the banking and financial services industry. The report also said that almost 82 per cent of these companies are using chatbots for servicing customers. While chatbots were found to be the most common function of AI that was in use, 65 per cent of them were using fraud detection AI engines, followed by 56 per cent who use virtual assistants. Of the respondents, 57 per cent said that they trusted that AI would give them a competitive edge. “Maturity of using and adopting AI-enabled solutions with a deeper understanding of not just the business case, technology and data, but also the risks around security, privacy and accountability, will differentiate the leaders from the rest,” said Sudipta Ghosh, partner at the data analytics division in PwC. “AI is used every day within payments, credit risk, investment recommendations and particularly in the area of intelligent digital assistants that handle regular customer service enquiries and tasks. Indian BFSI organisations looking to move ahead on the AI adoption curve can use AI to boost revenues through increased personalisation of services and embedding intelligence in automation and digital ecosystem partnerships,” Vivek Belgavi, partner at the FinTech division of PwC, stated. However, 60 per cent of the respondents believed that privacy concerns were worrying.","excerpt":"Of the respondents, 57 per cent said that they trusted that AI would give them a competitive edge.","categories":["AI News"],"tags":["AI in banking","AI in finance"],"author_name":"Poulomi Chatterjee","publish_date":"2022-03-01T15:57:57","publication_year":"2022","word_count":257,"keywords":["AI","chatbots","R","virtual assistants","AI in finance","Git","automation","analytics","Rust","GAN","AI in banking","fraud detection"],"extracted_tech_keywords":["AI","analytics","chatbots","virtual assistants","fraud detection","R","Rust","Git","GAN","automation"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/83-per-cent-of-indian-bfsi-players-think-ai-is-compulsory-for-better-customer-experience\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10065141,"title":"David vs. Goliath: Does Chinchilla fare well against Google AI’s PaLM?","content":"In 2020, OpenAI published a study titled, ‘Scaling Laws for Neural Language Models’ that demonstrated how increasing the model size resulted in improved performance. It was found that larger models were far more sample-efficient, so optimal compute-efficient training meant training large models on a comparatively smaller amount of data and stopping before convergence. In the recent past, all the important tech companies led the way with creating bigger large language models. The large language model trend culminated with dense models like GPT-3, which has 175 billion parameters, LaMDA, which has 137 billion parameters and Megatron-Turing NLG, which has 530 billion parameters. Smaller models, more training tokens To counter this viewpoint, DeepMind submitted a paper called ‘Training Compute-Optimal Large Language Models’ towards the end of March, which demonstrated that instead of just relying on the model size, the number of training tokens should also increase. The paper notes that usually for, when the computational budget increases by ten times, the size of the model is increased by 5.5 times while the number of training tokens is scaled by 1.8 times. However, the study suggests that the size of the model and the number of training tokens should increase proportionately. This theory was tested on a predicted compute-optimal model Chinchilla. The study compared Chinchilla’s 70-billion parameter model to Gopher’s 280-billion parameter model. Despite the smaller size, Chinchilla was trained on four times more data and outperformed Gopher with a state-of-the-art average accuracy of 67.5% on the MMLU benchmark, which is 7 per cent higher. Source: DeepMind blog Large language models as a norm keep the number of training tokens fixed at around 300 billion. Interestingly, while the cost incurred to train Gopher and Chinchilla was the same, Chinchilla was trained with 1.3 trillion tokens. Source: DeepMind blog Higher budget, different approach DeepMind’s claim that large language models were being trained with a suboptimal use of compute was also verified independently by Google AI’s research. At the beginning of the month, Google AI’s research team announced a new architecture called PaLM or the Pathways Language Model, a 540-billion parameter, decoder-only transformer model. Google stated in its findings that PaLM performed very well at English NLP tasks like sentence completion, comprehension and natural language inference, as well as multilingual NLP tasks like translation. The blog stated that the vision for Pathways was for a single AI system to be able to generalise across thousands of tasks with efficiency. Incidentally, PaLM was trained on 768 billion tokens, much less than Chinchilla but used five times the compute budget that Chinchilla demanded. PaLM was trained on a combination of data and model parallelism. At the Pod level, the model was trained over two Cloud TPU v4 Pods. This state-of-the-art training achieved a training efficiency of 57.8 per cent hardware FLOPs utilisation, which is the maximum efficiency for LLMs at this scale. Source: Google AI blog PaLM was fed English and multilingual datasets, including books, web documents, Wikipedia, casual conversations and GitHub code. Conclusion PaLM was tested on a set of NLP tasks alongside older large models like Chinchilla, GLaM, GPT-3, Megatron-Turing NLG and Gopher. Of the 29 tasks that included sentence completion, question-answer, reading comprehension and common-sense reasoning tasks, PaLM outperformed all other models in 28 tasks. PaLM was also compared to other LLMs on a range of 150 new language modelling tasks known as the Beyond the Imitation Game Benchmark (BIG-bench). While Chinchilla and PaLM were trained on different corpora, PaLM’s 540-billion model performed well at a range of tasks, including coding, where it was on par with OpenAI’s fine-tuned Codex 12B despite being trained on 50 times lesser Python code. At reasoning, PaLM was able to solve 58 per cent of the problems in GSM8K, a benchmark dataset of tough school-level maths questions. The model beat the previous best score set by GPT-3’s 55 per cent. PaLM was set against Chinchilla and Gopher across a subset of 58 of these tasks. Again, PaLM emerged on top. The study also found that PaLM’s performance as a “function of scale” follows a log-linear behaviour similar to older models. This signalled that the increase in performance from scale hadn’t reached a plateau yet. Source: Google AI blogDeepMind later admitted that despite PaLM not being compute-optimal, it would beat Chinchilla if trained on their data. It also predicted that given PaLM’s bigger compute budget, a 140-billion parameter model trained on 3 trillion tokens would give optimal performance and be more efficient for inference.","excerpt":"DeepMind’s claim that large language models were being trained with a suboptimal use of compute was also verified independently later by Google AI’s research.","categories":["Global Tech"],"tags":["AI Tool","PaLM"],"author_name":"Poulomi Chatterjee","publish_date":"2022-04-18T18:00:00","publication_year":"2022","word_count":745,"keywords":["PaLM","TPU","OpenAI","AI","AWS","ML","RAG","NLP","Python","Aim","AI Tool","R"],"extracted_tech_keywords":["AI","ML","NLP","OpenAI","Aim","RAG","AWS","TPU","Python","R"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/david-vs-goliath-does-chinchilla-fare-well-against-google-ais-palm\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10074403,"title":"How Pataa is Attempting to Alter India&#8217;s Addressing System","content":"According to a new mapping and geospatial startup Pataa, the unnamed and unstructured address system in India costs nearly INR 75,000 crore annually, or 0.5% of the GDP. By giving each home and anonymous site in the nation a digital address, the company hopes to find a solution to the problem. In conversation with Analytics India Magazine, Kratika Jain, co-founder and head of operations at Pataa, said, “We’ve divided the entire world into 3-by-3 metre square blocks, which we have termed ‘Square codes’; by doing this, we have created 57 trillion completely distinct square codes.” The startup explains that the square codes will be geotagged and will act as unique delivery addresses of users, providing almost 100% accuracy. The company is primarily attempting to provide a digital address to each and every household in India. But how is Pataa different from Google Maps, the leading player in the segment? Kratika responds that, “Google is a navigation expert; we are solving the addressing problem.” “In Pataa, we give you more than just the navigation; at our platform, the user can actually add in a multi-dimensional layer of the address. Once registered, the user won’t have to manually enter the address every time it makes an account in other apps or websites.” It is believed that the majority of businesses use Google Map API—not for its accuracy but rather for its capacity to deliver real-time traffic information as well as the quickest route to the users’ desired destination. However, Kratika believes that precision plays a much bigger role and claims that the app’s precision can be as high as 3 metres “which is even better than Google Maps, which has a position of close to 12~60 metres”. She further adds that, “Precision is the game play in a country like India.” ‘Pataa’ in e-commerce By 2025, it is expected that India’s e-commerce market will have grown by 19.24% CAGR, from US$ 46.20 billion in 2020 to US$ 111.40 billion. Several mapping companies have expressed interest in the e-commerce market, primarily owing to its projected growth. For instance, Flipkart has made significant investments in MapMyIndia to boost its supply chain and logistical operations. MapMyIndia, widely dubbed as the Indian substitute of Google Maps, places a lot of emphasis on the B2B market. Pataa, however, considers this scenario from a different perspective. The startup claims that there are several issues plaguing the e-commerce sector that need to be addressed on priority. According to Kratika Jain, “70% of the people churn out at an e-commerce app’s address page—these are our findings. It’s significant because when you invest a lot of money in digital marketing, you naturally want customers. 70% churn out is substantial, and we guarantee [nearly] 40–50% conversions with Pataa. Users won’t need to manually enter the address into any new applications, thanks to our API.” She further explained that, “When ordering from Amazon, you typically receive a notice stating that the purchase will be delivered between 9 AM and 6 PM.” Jain believes that this is indeed a broad window of delivery offered by the online retailer but that it could be narrowed down to the precise time with Pataa. “This is a sizable window, and in a nation like India, the delivery man must phone you before delivering a package. With Pataa, that is changeable; in fact, we can forecast the precise delivery time with precise geolocation.” Improved drone delivery Recently, the meteoric rise of the drone delivery system has been the subject of much mainstream media coverage. In response to this trending shift in delivery mechanism, Kratika Jain said that, “There are few businesses in Ireland using drones to deliver packages, but the way they collect addresses and carry out deliveries is hilarious. The package is sort of left anyplace close to the residence. Therefore, all that is required of the user is to keep an eye out for the parcel and such.” “With the help of our addressing system, drone delivery services can obtain the user’s geotagged address in order to carry out exact deliveries. With Pataa, drones can route accurately in locations with limited connectivity, which is ideal in case of an emergency situation or remote locations, for example, Indian army camps in remote northeastern parts.” Since the Indian government has allowed private enterprises to use drones for delivery, a large number of drone businesses are sprouting up in India. Redwing Labs recently completed a pilot test in Arunachal Pradesh—opening the door for the supply of medication to rural areas of the nation. Businesses like Pataa could provide valuable insight and play a crucial role in the drone delivery of groceries or medicine to establishments like government schools and the Indian defence forces located in remote areas of India. Road ahead According to the co-founder and head of operations, Kratika Jain, Pataa has reportedly raised $2.5 million in seed funding and is presently working to raise an additional $3~$5 million. She further underlined the necessity for expansion in developing nations like South America and Africa—“We aim to be the nation’s digital addressing system, and we’ve determined from our analysis that the issue doesn’t just affect India.” In conclusion, Kratika Jain remarked that, “Simply because of the inadequate addressing system, 50% of the world’s population actually has no address. That is a sizable number, then. Therefore, in nations like South Africa or Brazil, where the addressing system is a major issue, we want to sort of reach out to those areas. Numerous Latin American nations are also dealing with serious problems in this area. We want to be the go-to addressing platform globally.”","excerpt":"“We want to be the go-to addressing platform globally,” says Kratika Jain, co-founder and head of operations at Pataa","categories":["AI Features"],"tags":[],"author_name":"Lokesh Choudhary","publish_date":"2022-09-06T12:00:00","publication_year":"2022","word_count":933,"keywords":["Go","API","AI","Git","RAG","Aim","ViT","analytics","R","startup"],"extracted_tech_keywords":["AI","analytics","Aim","RAG","R","Go","Git","API","ViT","startup"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-pataa-is-attempting-to-alter-indias-addressing-system\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":40314,"title":"MIT&#8217;s New Open Source Tool Lets You See Behind The Scenes Of Black Box Modeling","content":"A machine learning model can have many dependencies and to store all the components to make sure all features available both offline and online for deployment, all the information is stored in a central repository. The main objective of having a proper pipeline for any ML model is to exercise control over it. A well-organised pipeline makes the implementation more flexible. It is like having an exploded view of a car engine where you can pick the faulty pieces and replace it- in our case, replacing a chunk of code. A pipeline consists of a sequence of components; components which are a compilation of computations. Data is sent through these components and is manipulated with the help of computation. Pipelines, unlike the name would suggest are not one-way flows. They are cyclic in nature and enable iteration to improve the scores of the machine learning algorithms. And, make the model scalable. A typical machine learning pipeline would consist of the following processes: Data collection Data cleaning Feature extraction (labelling and dimensionality reduction) Model validation Visualisation Data collection and cleaning are the primary tasks of any machine learning engineer who wants to make meaning out of data. But getting data and especially getting the right data is an uphill task in itself. Data quality and its accessibility are two main challenges one will come across in the initial stages of building a pipeline. The captured data should be pulled and put together and the benefits of collection should outweigh the costs of collection and analysis. But there can be problems associated with the information that is deployed into the model such as: an incorrect model gets pushed incoming data is corrupted incoming data changes and no longer resembles datasets used during training Researchers from MIT and elsewhere have developed an interactive tool that, for the first time, lets users see and control how increasingly popular automated machine-learning (AutoML) systems work. The tool, ATMSeer, generates a user-friendly interface that shows in-depth information about a chosen models’ performance, as well as the selection of algorithms and parameters that can all be adjusted. At The Heart Of ATMSeer This new tool ATMSeer is built around ‘Auto-Tuned Models (ATM).’ What this model does differently than other automated machine learning models is that it catalogues all the results as it tries to fit the models to data. ATM randomly selects an algorithmic approach, be it neural networks or decision trees and also model’s hyperparameters like the size of the tree or number of layers in a network. The model does this act of choosing and tuning hyperparameters repeatedly while assessing the performance. The results of this performance become a determining factor in choosing the next model; a better one. Finally, it displays all the results with models that suit a particular task. The interface of ATMSeer via MITNews ATMSeer interface consists of a control panel that allows users to upload datasets and an AutoML system, and start or pause the search process. There is also a “leaderboard” of top-performing models in descending order. A non-expert can decipher the performance of various models with these intuitive visualisations. ATMSeer includes an “AutoML Profiler,” with panels containing in-depth information about the algorithms and hyperparameters, which can all be adjusted. One panel represents all algorithm classes as histograms — a bar chart that shows the distribution of the algorithm’s performance scores, on a scale of 0 to 10, depending on their hyperparameters. “We let users pick and see how the AutoML systems works,” says Kalyan Veeramachaneni, a principal research scientist in the MIT Laboratory for Information and Decision Systems (LIDS), who leads the Data to AI group. Whether it is the market crash or a wrong diagnosis, the after-effects will be certainly irreversible. Tracking the development of machine learning algorithm throughout its life cycle, therefore, becomes crucial. Know more about Auto-Tuned Models here.","excerpt":"A machine learning model can have many dependencies and to store all the components to make sure all features available both offline and online for deployment, all the information is stored in a central repository. The main objective of having a proper pipeline for any ML model is to exercise control over it. A […]","categories":["AI Features"],"tags":["MIT"],"author_name":"Ram Sagar","publish_date":"2019-06-06T09:32:37","publication_year":"2019","word_count":643,"keywords":["Go","machine learning","AI","neural network","MIT","ML","RPA","Scala","data quality","GAN","R"],"extracted_tech_keywords":["AI","machine learning","ML","neural network","R","Go","Scala","data quality","GAN","RPA"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/mit-atmseer-black-box-model-transparency\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10083933,"title":"Top 9 GitHub repositories for the TensorFlow community","content":"Created by Google Brain and initially released to the public in 2015, TensorFlow has become ubiquitous in little time. The open-source library has a bundle of ML and deep learning models and neural networks and uses both Python and JavaScript to help developers and ML engineers build front-end API for applications. These repositories each serve a separate function from running ML models from the cloud to even the smallest microcontroller devices available. TensorFlow Lite TensorFlow Lite is an open-source and product-ready deep learning framework that can convert a pre-trained model in TensorFlow into a custom model that can then be optimised for speed or storage. The model can be deployed on edge devices that are light-weight like mobile phones supported by Android or iOS, devices like Raspberry Pi that are based on Linux and even microcontrollers. The special model is also deployed on the edge device after which the inferences are made on the device ensuring safety concerns around data piracy. TensorFlow Federated An open-source ML framework, TFF, or TensorFlow Federated, was developed to help open up research and experimentation around Federated Learning. The library will offer researchers starting pointers and complete examples for research work in different areas under federated learning. Federated Learning or collaborative learning is an approach in ML that builds one robust single ML model without actually sharing the data which upkeeps data security and distributes data rights evenly. For instance, federated learning has been used to train prediction models for mobile keyboards but without uploading sensitive typing data onto servers. TensorFlow Fairness Evaluation and Visualisation Toolkit When considering the impact that AI has had on people, this toolkit serves a valuable purpose. The toolkit analyses the binary and multi-class classifiers in the datasets for fairness and checks for bias. The toolkit first evaluates the distribution of datasets and determines how well the model is performing across defined sets of users and searches for the root causes for the biases while also suggesting the areas of improvement. The toolkit now also has an evaluation library which is agnostic to the model under evaluation and can be used for non-TensorFlow based models as well. TensorFlow Rust With Rust becoming increasingly popular, TensorFlow toolkit can help developers run Rust in production on micro-devices and while deploying neural networks at edge devices. However, the toolkit comes with the caveat that the project is still being developed and does not have the guarantee of a stable API. The combination of Rust and TensorFlow together is potent because of how easily it can design and train custom models using the bindings available in the toolkit. TensorFlow Quantum The TFQ or TensorFlow Quantum toolkit is a quantum ML library for rapid prototyping of hybrid quantum-classical models. Within the TensorFlow toolkit, researchers in quantum algorithms and applications can leverage the quantum computing frameworks. TFQ focuses on quantum data and integrates these algorithms and logic built in the Python framework ‘Cirq’. It then gives quantum computing primitives that will be compatible with existing TensorFlow APIs and high-performing quantum circuit simulators. TensorFlow TensorBoard The TensorBoard toolkit is for visualisation and measurement that developers need in the midst of their workflow. The toolkit tracks experiment metrics like loss and accuracy of the model, visualising the model graph and projecting graph embeddings to a lower dimensional space. There are separate dashboards in the toolkit for different functions—the Scalars dashboard tells how the loss and metrics change with each epoch and tracks training speed and the learning rate of the models, the Graphs dashboard visualises the model and so on. TensorFlow tfjs The TensorFlow.js toolkit, or tfjs, is an open-source JavaScript library that is hardware-accelerated for training and deploying ML models. Tfjs can help developers build ML models in JavaScript and use ML directly either in the browser or in the popular Node.js. Tfjs can also retrain the pre-existing ML models using sensor data that is connected to the browser or other client-side data. TensorFlow tfx TensorFlow Extended is a serving toolkit for ML models that is made for production environments. Tfx helps developers build automated end-to-end pipelines for their ML models. Tfx saves developers from a manual, tedious process that has no tracking or validation measures. Protocol Buffers Protocol Buffers, or protobuf, is a Google toolkit for serialising structured data. Protobuf is language-neutral, platform-neutral and extensible, and it is often used in cases where communication protocols have to be defined or even for data storage. The protocol compiler is written in C++, and it currently supports generated code in Java, Python, Objective-C, and C++.","excerpt":"The combination of Rust and TensorFlow together is potent because of how easily it can design and train custom models using the bindings available in the toolkit.","categories":["AI Trends"],"tags":["Top Trend"],"author_name":"Poulomi Chatterjee","publish_date":"2023-01-03T17:00:00","publication_year":"2023","word_count":757,"keywords":["federated learning","Top Trend","AI","neural network","ML","RAG","Python","deep learning","JavaScript","TensorFlow","R"],"extracted_tech_keywords":["AI","ML","deep learning","neural network","TensorFlow","federated learning","RAG","Python","R","JavaScript"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-9-github-repositories-for-the-tensorflow-community\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10110474,"title":"Infosys To Acquire InSemi, Driving Innovation in Semiconductor Design","content":"Indian IT giant Infosys announced its definitive agreement to acquire InSemi for ₹280 crore, a prominent semiconductor design and embedded services provider. The company  said it aims to accelerate its Chip-to-Cloud strategy through this collaboration, leveraging InSemi’s niche design skills at scale. The company in its statement  said this move aligns seamlessly with Infosys’ existing investments in AI\/Automation platforms and industry partnerships, facilitating comprehensive end-to-end product development for clients. “With the advent of AI, Smart devices, 5G and beyond, electric vehicles, the demand for next-generation semiconductor design services integrated with our embedded systems creates unique differentiators. InSemi is a strategic investment as we usher a next wave of growth and a leadership position in Engineering R&D.” said Dinesh R, EVP & Co-Delivery Head, Infosys. Founded in 2013, InSemi provides end-to-end semiconductor design services, covering electronic design, platform design, automation, embedded, and software technologies. With a team of over 900+ design specialists, InSemi serves leading global corporations across semiconductor, consumer electronics, automotive, and hi-tech industries. The acquisition is set to strengthen Infosys’ Engineering R&D capabilities and position the company as a leader in this domain. “With Infosys as our catalyst, it creates a synergistic combination that allows us to scale and bring the power of AI & Engineering R&D and next-generation technology to global clients, expanding across industry sectors. We aim to further accelerate our progress and together with Infosys,  it paves a path of innovation opening new opportunities for our teams”, said Shreekanth Sampigethaya & Arup Dash, Co-Founders, InSemi. The acquisition is expected to close in the fourth quarter of fiscal 2024, subject to customary closing conditions.","excerpt":"Founded in 2013, InSemi provides end-to-end semiconductor design services, covering electronic design, platform design, automation, embedded, and software technologies.","categories":["AI News"],"tags":["Infosys","Mergers and Acquisitions","Semiconductor India"],"author_name":"Siddharth Jindal","publish_date":"2024-01-11T17:07:38","publication_year":"2024","word_count":267,"keywords":["Infosys","AI","programming_languages:R","innovation","ML","RAG","automation","Aim","Semiconductor India","Mergers and Acquisitions","R"],"extracted_tech_keywords":["AI","ML","Aim","RAG","R","automation","innovation","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/infosys-to-acquire-insemi-driving-innovation-in-semiconductor-design\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10072318,"title":"The origin of Neo4j","content":"“The first code for Neo4j and the property graph database was written in IIT Bombay”, said the chief Marketing Officer at Neo4j, Chandra Rangan. In an exclusive interview with Analytics India Magazine, Rangan said that the first piece of code was sketched by Emil Eifrem — who is the founder and CEO of Neo4j — on a flight to Bombay, where he worked with an intern from IIT Bombay to develop the graph database platform. Rangan joined Neo4j as the chief marketing officer (CMO) on May 10, 2022. Prior to this, he worked at Google, running Google Cloud Platform product marketing and, more recently, product-led growth, strategy, and operations for Google Maps Platform. Rangan has over two decades of technology infrastructure experience across marketing leadership, strategy, and operations at Hewlett Packard Enterprise, Gartner, Symantec, McKinsey, and IBM. Founded in 2007, Neo4j has more than 700 employees globally. In June 2022, the company raised about $325 million in a Series F funding round led by Eurazeo, alongside participation from GV (formerly Google Ventures) and other existing investors like One Peak, Creandum, Greenbridge Partners, DTCP, and Lightrock. This is one of the largest investments in a private database company. It raised Neo4j’s valuation to over $2 billion. In contrast, even bigger than MongoDB, which raised a total of $311 million, and post-IPO, it raised about $192 million in IPO, making it worth $1.2 billion. Bets big on India With its latest funding round, Neo4j is looking to invest in expanding its footprint globally, and India is one of its top choices, thanks to a larger developer ecosystem, alongside a burgeoning startup ecosystem and IT service providers using its platform to offer solutions to global customers. Neo4j’s community edition, which is open source, is widely adopted by developers in the country. “We have an overall community of almost a quarter million users who are familiar with our platform”, said Rangan, explaining that it has one of the largest developers in the country. With the fresh infusion of funds, the company looks to tap into the market, expand its services, sales and support, and invest in the right strategies going forward. As part of its expansion plans, Neo4j started hiring in sales leadership and country manager roles from last year onwards and would also continue that momentum this year. “This is a big bet for us in multiple ways”, added Rangan, pointing at its Indian root and all the innovations in the country. Besides India, Neo4j has a strong presence in Silicon Valley and Sweden and has a huge developer ecosystem in the US, China, Europe, South East Asia and others. Strategies for expansion Over the years, Neo4j has grown through developers and some of the early adopters of its platform. “Unfortunately, developers interested in graph databases will typically start with us”, said Rangan affirmatively. Further, explaining the conversion cycle, he said that once they know about graph databases, they later join the community edition. Then, once they get comfortable with the use cases and start putting this into production, they eventually get into a paid version for the advanced security, support, scalability, and commercial constructs. “In India, that’s the similar motion we are seeing”, said Rangan. He revealed that they already have a huge developer community. Banking on this community, they plan to invest in continuing the engagement with the community in a meaningful way. Of late, the company has also started hiring several community leaders to encourage proactive engagement within the community. In addition, it is also investing heavily in sales and marketing engines, including technical sales, which work closely with organisations in building the use cases, alongside the implementation of services and support. What makes Neo4j special? One thing that makes Neo4j stand apart from other players is its intuitiveness in helping deploy applications faster because of its flexible schema. This helps developers to add properties, nodes, and more. “It gives tremendous flexibility for developers so they can get to the outcome much more quickly”, said Rangan. But what about the learning curve? Rangan said, “Literally, for a new developer, if they start learning graphs for the first time, it is very intuitive.” He explained that the learning curve is not that steep and doesn’t take long. “But, for folks who have been working in the development space and building applications and are very familiar and comfortable with RDBMS, i.e., rows and tables. Strangely enough, the learning curve is a little higher and steeper”, added Rangan, discussing that they have to unlearn to model intuitively versus modelling tables. He said the best way to overcome that learning curve is to try it out. “So, when you think about the learning curve, it is a very easy learning curve, especially if you can put aside the former way of thinking about things like rows and tables and go back to first principles.”—Chandra Rangan. Discovering use cases with Neo4j The International Consortium of Investigative Journalists (ICIJ) released the full list of companies and individuals in the Panama Papers, implicating at least 140 politicians from more than 50 countries in tax evasion schemes. The journalist used Neo4j to draw the relationship with their data and found common touchpoints and names of people involved in having multiple offshore accounts and evading tax. “We believe a whole bunch of sectors can actually get value. We have seen new sectors kind of pop up on a pretty regular basis”, said Rangan while citing various use cases in financial service sectors (fraud detection), healthcare (vaccine distribution), pharmaceuticals (drug discovery), supply chain and logistics (mapping automation), tech companies (managing IT networks), retail (recommendation systems), and more. Chandra Rangan further explained that people are still discovering what they can use graph databases for and how useful it is in some sense. He said that it is unleashing a whole bunch of innovations. “So, we are hoping for a lot of that to happen here in India because of the developer community”, he added. What’s next? Rangan said Neo4j would be aggressively investing in the community and ecosystem here in India. Besides this, he said they are investing in building a marketing and sales team, which has grown significantly in the last year. In addition, Neo4j is also investing in building a partner ecosystem to support a wider range of customers. “Depending on how quickly we can grow or cannot grow—again, responsible growth—we want to grow as fast as possible. But, we also want to make sure as we hire people as we establish the relationship, we are investing enough time, effort, and money to make sure that these relationships are successful”, concluded Rangan.","excerpt":"Neo4j has more than 700 employees globally.","categories":["IT Services"],"tags":["graph database","graph databases","neo4j"],"author_name":"Amit Naik","publish_date":"2022-08-08T16:00:00","publication_year":"2022","word_count":1103,"keywords":["graph database","Go","graph databases","AI","MongoDB","recommendation systems","Scala","RAG","neo4j","analytics","GAN","R","fraud detection"],"extracted_tech_keywords":["AI","analytics","RAG","recommendation systems","fraud detection","MongoDB","R","Go","Scala","GAN"],"url":"https:\/\/analyticsindiamag.com\/it-services\/the-origin-of-neo4j\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":13852,"title":"Staqu-  Re-defining search using Artificial Intelligence","content":"A Delhi based Artificial Intelligence startup, Staqu is nothing short of a revolutionary idea that is taking fashion e-retail and e-commerce industry to a new level. Founded in 2015, the company has pioneered a bidirectional image understanding technology, an innovative image-to-image matching system, that simplifies the image search, automated meta-tag generation and real time product recommendations. With its prime focus on AI research, Staqu has produced the VGrep API suite consisting of a state of the art visual search engine, hybrid recommendation engine and other innovative products. To know more about the company and understand the underlying technology, Analytics India Magazine interacted with Atul Rai, CEO & Co-founder at Staqu, who has some interesting facts to reveal. Let’s read here- On re-defining search using Artificial Intelligence- Staqu is all about artificial intelligence, and the most interesting part is that it is using AI to revolutionize search engines. “Whenever we think about search engines, we have a pre-defined picture in our mind, of a textual box where one can type a text query and get results in some textual form”, says Rai, who has been quite instrumental in Staqu’s development such as winning the IBM’s Global Entrepreneur Program and raising the initial round of investment from Indian Angel Network. “But if you see the current trend of data on internet, you will find that most of it is now flowing in forms of images – 70 per cent of data on the internet is in the form of images. This, in turn, means that the quest for image search related solutions will increase in the near future”, he adds. Atul Rai, CEO and Co-Founder, Staqu Though images are the next go-to thing, the underlying problem with images is that one cannot define every image in terms of text. Rai explains “Let’s say, if you want to search for “Alia Bhatt’s lehenga from Badrinath ki Dulhaniya” it’s really difficult to convert every colour and pattern in form of text and perform search in some traditional search box. This is one example, where searching through images simplifies things.” Tapping onto this idea of utilizing domain specific meaningful information from images and using it for different purposes, not limited to search, Staqu was founded in 2015 and has since been “re-defining search using AI”. On the idea behind using AI- In computer vision, image-to-image search has been a much known problem, and researchers have been actively working in this domain since the 80’s. And it was with the rise of deep learning based approach post 2013, that it has been possible to extract powerful visual features from images that has led to some great results in the said field. “Previously, I was working with one such similar problem (image to image search) with one of the UK based universities. Once I completed my research, I decided to expand this idea further to solve some known problems in e-commerce and that’s where I met and collaborated with my fellow co-founders, Anurag, Chetan and Pankaj, who have diverse knowledge and skills needed to democratise this technology”, adds Rai, who has a Master’s in Artificial Intelligence from the University of Manchester. Rai also has won the prestigious 21st European Image processing challenge: SSIP 2013 and has the Best Research Paper Aware in DICTA 2014, IEEE Australia, to his credit. On “the Fashin App”- Adding a note on the products and services offered by Staqu, Rai says “Staqu is currently working in two domains – B2B and B2C. In the B2B domain, we provide two suites of products, one is for mobile companies (OEM), where we wrap our AI technology to develop Ad tech platforms and other is for E-commerce companies, where we provide API based solutions.” The company has developed a first consumer centric mobile application “Fashin”  (Available for both Android and iOS) for fashion e-commerce domain that aims to provide one stop solution to discover and search fashion products across the e-commerce websites in the least possible time. “The mobile application assists users to not only search the product through its image search feature but also helps the users to get wardrobes of their favourite Bollywood celebrities directly from YouTube videos”, says Rai. Fashin app comes with features like image search, tag based filtering, real time trend analysis from video and a lot more. Once the app is downloaded, the user can stream latest Bollywood Videos and the application would extract the similar jacket or T-shirt available at various websites at different prices. The app also allows the user to either click at that moment  or upload an image from phone’s gallery and search for preferred fashion with best prices available online. It also enlists the best deals from the top-notch brands. It also allows to further simplify fashion discovery by taking a picture of favourite outfit and uploading it to browse for similar options available online. On offering customized solutions- Rai says “Different clients have different requirements. In the current competitive environment, each of our clients want an edge on other players hence customisation is one of the key components of Staqu’s product suite.” “In the customisation process, we develop and wrap up our AI technologies in some form of API as well as product required for our clients. We develop plug and play AI powered APIs and SDKs which can be integrated in their existing products with minimal technical interference”, he adds. Staqu currently works with brands like Paytm, Karbonn, Intex, Panasonic, Lava , Roposo , VistaPrint etc. for different set of AI products. On the growth story- Rai is quick to admit that within 11 months of their market launch, Staqu is not only working with some of the big brands in the industry but have also closed the contract for more than two years with some of them. “I believe our clientele tells better story than the number or figures I can provide in support of our technology”, he says. “We started Staqu with the sole motto of democratising AI to the common audience and we already have launched “Fashin” which completely automates mobile application target for fashion loving audience. The app is consistently gaining traction and has more than 10K downloads without any marketing efforts and 800-900 daily active users with session of 3-4 minutes”, he says. On a concluding note- He concludes leaves a advice to the startup community by adding “Newcomers should focus on building a great product but should not miss out on making money out of it. As ultimately, building great product and scaling it would require money, so rather than running behind investors, you should try to generate it from the product itself and use it to innovate further. I would further urge aspiring entrepreneurs to never stop building and never stop innovating.”","excerpt":"A Delhi based Artificial Intelligence startup, Staqu is nothing short of a revolutionary idea that is taking fashion e-retail and e-commerce industry to a new level. Founded in 2015, the company has pioneered a bidirectional image understanding technology, an innovative image-to-image matching system, that simplifies the image search, automated meta-tag generation and real time product […]","categories":["AI Startups"],"tags":["ai in fashion","most revolutionary ai deep learning company","Startups"],"author_name":"Srishti Deoras","publish_date":"2017-03-28T11:35:31","publication_year":"2017","word_count":1127,"keywords":["Go","API","artificial intelligence","ai in fashion","AI","computer vision","RAG","Aim","deep learning","most revolutionary ai deep learning company","analytics","Startups","R"],"extracted_tech_keywords":["AI","artificial intelligence","deep learning","computer vision","analytics","Aim","RAG","R","Go","API"],"url":"https:\/\/analyticsindiamag.com\/ai-startups\/startup-week-staqu-re-defining-search-using-artificial-intelligence\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10122422,"title":"Ola will Save INR 15 Crore Annually Using Krutrim AI Cloud","content":"Ola’s decision to use Krutrim AI Cloud instead of Microsoft Azure and AWS cloud infrastructure is going to save them approximately INR 15 crore annually. “Cloud costs are crazy. He [Bhavish Aggarwal] basically used to burn INR 5 lakh per day, which is quite a bit of money,” said Sasank Chilamkurthy, the founder of Qure AI and Von Neumann AI (the company building a personal AI server JOHNAIC), in an exclusive interview with AIM. Chilamkurthy swiftly did some number-crunching and came out with the estimation that it costs about INR 8.2-10 crore to buy servers that will handle the current load of Ola on Azure. Had they not migrated, the expense over three years would have totalled INR 54.74 crore. The difference, INR 54.75 crore minus INR 8.2 crore, amounts to about INR 45 crore, translating to a substantial savings of roughly INR 15 crore annually. Assuming a company uses an AWS g4dn.4xlarge instance and reserves it for three years, it would cost them INR 10-25 lakh for a specific instance, according to Chilamkurty. According to estimates and disclosures by the Ola chief on X, this shift could lead to a daily revenue loss of INR 5-25 lakh for Microsoft Azure and INR 30-40 lakh for AWS.  Annually, this could translate to approximately INR 18.25-91.25 crore for Azure and INR 109.5-146 crore for AWS. Moreover, according to Chilamkurthy, Ola spent around INR 85 lakh on egress costs paid to Azure. Egress cost is the charge levied by cloud providers for moving or transferring data from the cloud storage where it was earlier uploaded. Chilamkurthy told AIM that he recently bumped into Aggarwal at a cafe, where he pitched JOHNAIC for Krutrim AI Cloud. He said that Aggarwal showed some interest in JOHNAIC, partly because it also utilises Intel GPUs. He further revealed that Ola Krutrim will buy Intel Gaudi 2 to build Krutrim AI Cloud. In a recent post, Aggarwal said that over 2,500 developers have signed up for Krutrim Cloud services, and they will work with everyone to onboard them in the coming weeks. Most recently, Aggarwal met Arm chief Rene Haas in Taiwan. JOHNAIC Cloud in a Box Von Neumann AI recently launched JOHNAIC, described as a “cloud in a box” solution. It provides the benefits of cloud computing, such as scalability and flexibility, but within a local, on-premises environment. This hybrid approach combines the best of both worlds—cloud capabilities with local data control. It comes with the following hardware configurations: an Intel i5 12400 CPU, an Intel Arc 770 16 GB GPU, 64 GB of RAM, a 1 TB SSD, and an optional 1100 VA UPS. “We have sachetised AI into small boxes so that you can personalise them and take them wherever you want,” said Chilamkurthy. It claims to slash AI costs by 85% and comes with in-built SaaS and AI tools to run SMEs and startups. JOHNAIC comes with Ubuntu, Matrix, and ERPNext, supports oneAPI, and can run Meta Llama. “JOHNAIC is a one-time investment of around INR 2 lakh, resulting in 80-92% savings,” said Chilamkurthy, adding that users don’t need to buy an Apple MacBook to run AI applications, which comes at a similar price. Chilamkurthy is currently targeting the inference market with JOHNAIC. “It is very reasonable and affordable for most folks and startups trying to do something cool using OpenAI GPT’s APIs,” he said. People+AI is the first customer of JOHNAIC and is using it for its own AI requirements while keeping its data private. Compared to public cloud services like AWS, JOHNAIC offers a highly competitive pricing model making it an attractive option for businesses looking to reduce their AI infrastructure costs while maintaining robust capabilities. “Even on AWS, you cannot really buy just one NVIDIA A100, you have to purchase it in a set of eight GPUs, which I think is really expensive. For most people, one A100 would be sufficient, and they would just run it for a week,” said Chilamkurthy. Lately, Intel is aggressively targeting India, aiming to seize the market share created by the unavailability of NVIDIA GPUs. “I don’t use NVIDIA because I believe they overcharge us. They don’t offer small packets like sachets, instead, they push for large server purchases,” said Chilamkurthy. [Updated] June 4, 2024, 17:21 | The article has been updated to show that it costs about INR 8.2-10 Cr to buy servers that will handle the current load of Ola on Azure.","excerpt":"Bhavish Aggarwal recently met Arm chief Rene Haas in Taiwan.","categories":["AI Features"],"tags":["Ola Krutrim"],"author_name":"Siddharth Jindal","publish_date":"2024-06-04T14:44:01","publication_year":"2024","word_count":740,"keywords":["Go","OpenAI","AI","cloud computing","AWS","R","Scala","RAG","Aim","Ola Krutrim","Azure"],"extracted_tech_keywords":["AI","OpenAI","Aim","RAG","cloud computing","AWS","Azure","R","Go","Scala"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/ola-will-save-inr-15-crore-annually-using-krutrim-ai-cloud\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10046543,"title":"A Step-by-Step Guide To Build ML Models For Research","content":"Building machine learning models for research can be challenging, especially if you are developing them from scratch. Several factors come into play before you successfully deploy ML models for production. Some of these aspects include data collection, training data in a limited amount of time, non-representative training data, poor quality of data, irrelevant\/unwanted features, overfitting the training data, underfitting the training data and offline learning, development of the model, etc. Last year, Gartner reported that only 53 per cent of projects made it from prototype to production — and that’s at organisations with some level of AI\/ML experience. On the other hand, that number is likely to be much higher, with some failure-rate estimates touching nearly 90 per cent. In the McKinsey report, out of 160 companies, 88 per cent did not progress beyond the experimental stage. Chris Chapo, vice president, advanced analytics at Amperity, said one of the biggest reasons is that people believe by throwing money at a problem or putting technology in, success comes out the other end. “That just doesn’t happen,” said Chapo. So, what is the secret sauce here? Is there an easy way out or a step-by-step guide to successfully deploy machine learning models? Michael A. Lones, in a research paper titled ‘How to avoid machine learning pitfalls: a guide for academic researchers,’ explained the common mistakes that occur when using machine learning techniques and what can be done to avoid them. Lones is an associate professor in the School of Mathematical and Computer Science, Heriot-Watt University, Edinburgh. Data is everything Before diving into the sophisticated techniques, you probably should perform some fundamental data clearing methods on every machine learning project. Unfortunately, these are overlooked by even seasoned machine learning professionals and researchers yet are so critical that if skipped, models may break or report overly optimistic performance results\/outcomes. Lones says that it is normal to want to rush into training and evaluating models, but it is important to take the time to think about the goals of a project, to understand the data, its limitations, previous works, etc. He says, if you don’t do these things, you might end up with results that are hard to publish or models that are not suitable for their intended purpose. So, before you start building your models, here are some things that are important to consider. Take time to understand your data Look at all your data Make sure you have enough data Talk to domain expertsSurvey the literature Think about how you want your model to be deployed Building reliable models Building models is one of the most important parts of machine learning. With modern ML frameworks, it is easy to throw all techniques at your data and see what works. This, sometimes, leads to a disorganised mess of experiments that is hard to justify and record. Lones believes that approaching model building in an organised manner is the way forward. You make sure to utilise data correctly and put adequate consideration into the choice of models. He suggests the following steps to ensure you build a reliable model: Do not allow test data to leak into the training process Try out a range of different models Use inappropriate models Optimise your model’s hyperparametersBe cautious where you optimise hyperparameters and select Evaluating the robustness of models Lones says one needs to have valid results that you can draw reliable conclusions from. Without this, it becomes difficult to assess the outcomes. He says one should think carefully about how they are going to use data in their experiments, measure the true performance of their models, and how they are going to report this performance in a meaningful way. Here are the tips to consider while evaluating your machine learning model Use an appropriate test set Use a validation set Evaluate a model multiple timesSave some data to assess your final model instance Do not use accuracy with imbalanced datasets Comparing models fairly Comparing models is one of the most common practices in AI\/ML research work, but it is also challenging to get it right. Lones says that if a researcher carries out a comparison unfairly and publishes it, other researchers may subsequently be led astray. Therefore, he suggested that researchers evaluate different models within the same context, explore multiple perspectives, and utilise statistical tests appropriately. Here are the things to consider while comparing your machine learning models: Do not assume a bigger number means a better model Use statistical tests when comparing models Do correct for multiple comparisonsDon’t always believe results from community benchmarks Consider combinations of models Reporting results Lones says that the aim of academic research should not be self-aggrandising but rather an opportunity to contribute to knowledge. Therefore, one needs to provide a complete picture of their work, showcasing what worked and did not. “Machine learning is often about trade-offs, and you should try to reflect this with a nuanced and considered approach to reporting results and conclusions,” says Lones. Here are some of the critical things to consider before reporting the results of your model: Be transparent Do report performance in multiple ways Do not generalise beyond the data Be careful when reporting statistical significance Do look at your models Conclusion Lones believes that the documents do not tell you everything you need to know, the lessons sometimes have no firm conclusions, and some of the things discussed here are debatable. “This, I am afraid, is the nature of research. The theory of how to do ML almost always lags behind the practice; academics will always disagree about the best ways of doing things, and what we think is correct today may not be correct tomorrow,” says Lones. He said that one has to approach machine learning the same way they would approach any other research area — with an open mind, willingness to stay updated with the latest developments, and the humility to accept that you do not know everything.","excerpt":"In the Mckinsey report, out of 160 companies, 88 percent of them did not progress beyond the experimental stage","categories":["Deep Tech"],"tags":["AI (Artificial Intelligence)","AI latest","Guide","learn machine learning","Machine Learning","Machine Learning Latest","Machine Learning New","machine learning research","ML models","ML research"],"author_name":"Amit Naik","publish_date":"2021-08-23T12:00:00","publication_year":"2021","word_count":989,"keywords":["RPA","Machine Learning Latest","Ray","R","Machine Learning New","analytics","Guide","machine learning research","Go","machine learning","AI","ML","ML models","Machine Learning","GAN","ML research","learn machine learning","Aim","AI latest","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","machine learning","ML","analytics","Aim","Ray","R","Go","GAN","RPA"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/a-step-by-step-guide-to-build-ml-models-for-research\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10040834,"title":"Google’s Ambitious Project That Can Crush Its Own Meet","content":"“Project Starline, an innovative hyper realistic tech for video calling that will make you feel as if you are sitting right in front of the person.” The pandemic has disrupted the way we work, commute and socialise. Few think that these changes will be permanent. Companies like Google are capitalising on this new normal with innovations that can revitalise our personal virtual spaces. The search giant now wants to offer an experience of interacting with your loved ones in the ways we never thought would be possible. At the recently concluded Google I\/O 2021 keynote, the company unveiled Project Starline, an innovative hyper realistic tech for video calling that will make you feel as if you are sitting right in front of the person.  Over the years google has come up with some remarkable projects from google meet to google teams, to help us find ways to connect  virtually  with people no matter where we are. Another addition to the long list of ambitious developments is Project Starline that will change the dynamic of how we interact in a virtual space with anyone. Imagine a magic window, and through that window you see another person, life-size and in three dimensions.Project Starline is a technology project that combines advances in hardware and software to help people feel like they're together, even when they're apart. #GoogleIO pic.twitter.com\/2yNJrXoQcx— Google (@Google) May 18, 2021 Google’s latest innovation is a gamer changer for virtual meets. Starline’s commercial availability can create problems for companies like Zoom, which took off during the pandemic. Google’s own “Meet” can be at risk! \\ Project Starline Project Starline creates photorealistic 3D images that have volume and depth. The technology uses custom made booths, which to some appear like a prison or a photo booth fitted with  a glass panel. This new display system is partly encased in gray fabric and wood panel with a 65-inch display and a built-in bench. Project Starline uses more than a dozen different depth and imaging sensors. These sensors capture photorealistic, three-dimensional imagery which the system compresses and uploads it to both ends of the video conversation with apparent negligible delay. Google applies various special effects, including lighting and shadow . Google claims that around a hundred employees are currently using Starline in Mountain View, Seattle, and New York.Even though Google’s vision of videoconferencing for the future feels surreal only a few of the executives had been permitted to witness it. At #GoogleIO today we shared how advances in AI are making our products helpful in moments that matter, from safer routing in Maps, to breakthroughs in natural conversation, to technology that makes us feel like we’re in the same room even when we’re not. https:\/\/t.co\/0QvwKfM7KU— Sundar Pichai (@sundarpichai) May 19, 2021 Google started working on Project Starline even before virtual workspaces went mainstream. Project Starline was more than five years ago in the making. However, there are a few limitations that Google needs to overcome before Starline could be turned into a commercial product.  Google will have a hard time convincing its users of getting these “custom booths”. For now it has created enough buzz at the IO conference to get people excited for the future of Project Starline. Google has worked hard this year to redefine the ways we have been communicating over the last year. It has brought a plethora of technology that can in fact change the entire workspace allowing for a more effective collaboration between members. Apart from photorealistic virtual tours, Google also wants to change the fundamentals of how we work. Google’s Productivity Suite now comes with a Smart chip that can greatly increase workspace productivity allowing one to connect with Google Docs more easily. With the simple use of an ‘@’ people can now view the person’s contact, location, job title as well as files and meetings to quickly collaborate with teammates.It will give the ability to the collaborators to quickly preview documents and go through meetings without changing tabs from their phone or web. Unlike Project Starline , the smart chips will roll out very soon.It will be available for use on the Google Sheets in the coming months. It will be exciting to see how these changes turn out to be after they launch.","excerpt":"“Project Starline, an innovative hyper realistic tech for video calling that will make you feel as if you are sitting right in front of the person.” The pandemic has disrupted the way we work, commute and socialise. Few think that these changes will be permanent. Companies like Google are capitalising on this new normal with […]","categories":["Global Tech"],"tags":["Google Meet","Sundar Pichai","Zoom"],"author_name":"Ritika Sagar","publish_date":"2021-05-27T13:00:00","publication_year":"2021","word_count":708,"keywords":["Go","API","Zoom","programming_languages:R","AI","Google Meet","innovation","programming_languages:Go","Ray","Aim","ViT","R","Sundar Pichai"],"extracted_tech_keywords":["AI","Aim","Ray","R","Go","API","ViT","innovation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/googles-ambitious-project-that-can-crush-its-own-meet\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10141648,"title":"Why You Should Take Gary Marcus Seriously","content":"Recently, there have been significant discussions about scaling laws, particularly with the introduction of OpenAI’s new scaling law for o1 called test-time compute. The world’s best-known skeptic Gary Marcus, however, disagrees. He argues that scaling LLMs across different parameters is not the ultimate path to AGI. “I don’t think LLMs are a good way to get there (AGI). They might be part of the answer, but I don’t think they are the whole answer,” Marcus said in an exclusive interview with AIM, stressing that LLMs are not “useless”. He also expressed optimism about AGI, describing it as a machine capable of approaching new problems with the flexibility and resourcefulness of a smart human being. “I think we’ll see it someday,” he further said. Sharing his two cents on OpenAI’s test-time compute, Marcus suggested that he doesn’t believe it will be a viable solution in the long term. According to him, while scaling is necessary, the so-called “scaling laws” aren’t universal truths. “They’re not like gravity. They’re generalisations that work for a while. Like Moore’s law, which worked for a while but didn’t last forever,” he said. Marcus further explained that scaling in this field has mostly involved adding more data to solve problems. “That worked to some degree for a while, but it never really solved hallucinations and stupid errors, and now we’re running out of fresh data to keep doing that,” he added. While suggesting that OpenAI’s new method of scaling is worth exploring, Marcus pointed out that it only works in domains with a lot of synthetic data. Gary’s Answer to AGI Marcus, who has long been critical of the over-reliance on scaling deep learning models, reiterated his belief that the field must change course. “We spent the last four years pursuing mostly one hypothesis, which was scaling, and I think that hypothesis was a mistake,” he said. According to Marcus, the future of AI should focus on neuro-symbolic systems, which combine the learning capabilities of deep learning with the structured, rule-based reasoning of symbolic AI. In his 2020 paper ‘The Next Decade in AI: Four Steps Towards Robust Artificial Intelligence’, Marcus outlined a four-step strategy for the AI community to follow: prioritising neuro-symbolic AI, building large-scale knowledge databases, improving reasoning techniques, and developing cognitive models. Despite the difficulty of these tasks, he argued that they are essential for progress. “I think that’s what we need to do now,” he said. While Marcus critiqued companies like OpenAI for their focus on scaling models like GPT-4o, he applauded DeepMind, especially for its work on AlphaFold. “DeepMind has done the most interesting work in the field,” he said, adding that AlphaFold, although not as general-purpose as GPT-4o, demonstrates the successful application of neuro-symbolic AI in the domain of protein folding. Marcus also acknowledged that DeepMind is likely on a better path towards AGI compared to its competitors. He, however, indicated that no company has yet found the definitive route to AGI. “Of the major companies working on this, DeepMind is most likely to be on the correct path,” he said. Marcus and Yann LeCun’s Debate Recently, Marcus and Meta’s chief AI scientist Yann LeCun engaged in a conversation on Threads, debating who first predicted that LLMs wouldn’t lead to AGI. “Auto-regressive LLMs are hitting a performance ceiling. I’ve been predicting this before most people had even heard of LLMs,” claimed LeCun. He added that he has always been heavily criticised for saying that LLMs were useful, but were an off-ramp on the road to human-level AI. “I’ve said that reaching human-level AI will require new architectures and new paradigms,” he further said. Marcus told AIM that, back in 2019, he had predicted that language models wouldn’t lead to AGI. “When I originally said that, LeCun said, ‘You’re fighting a rear-guard action,’ and at the time, he was much more bullish on large language models,” he recalled. “He’s trying to reinvent himself and pretend that he was always opposed to them. He was hostile to me when I first criticised them and was highly critical of my 2022 paper ‘Deep Learning is Hitting a Wall’,” Marcus said, adding, “He’s using the famous rhetorical technique of a straw man and misrepresenting my idea in order to make me sound like a fool.” Marcus disclosed that his 2022 paper was focused on large language models. “It was about the limits of GPT-3 that were unlikely to be solved by scaling. I was always arguing that large language models were the best example of deep learning we had, but they were bound to encounter problems. Now those problems have surfaced, and he’s (Yann LeCun) trying to rewrite history and pretend he was the first person to see them,” he added. Betting Big on Cognitive Models Marcus stressed that the future of AI requires advances in cognitive models, a concept he has advocated for years. These models can reason and understand the world in a way that LLMs cannot. As he described in his 2020 essay, these cognitive models would allow AI systems to represent individual objects and entities in a better way, differentiating between unique instances rather than treating everything as part of a generic category. “I wrote about cognitive models in 2020, and he (Yann LeCun) has also been writing about them. He calls them world models, but it’s just different names for the same thing,” Marcus said. Marcus believes that representing individuals separately from their kind is one of the things known to be necessary for artificial intelligence, while agreeing that the exact method to achieve this is unclear. He compared the current state of AI development to Leonardo da Vinci’s designs for a helicopter, which he described as an idea ahead of its time but not yet possible to implement due to a lack of supporting technology. In his latest book ‘Taming Silicon Valley’, Marcus discusses how current AI systems, despite their impressive capabilities, are still far from achieving human-level intelligence. They are often brittle and prone to errors, especially when faced with unexpected situations. Compute is Not the Answer Marcus believes that the right answer to AI won’t require so much computing. Using his own children as an example, he said, “My children don’t have nearly as much compute as we have on this planet, and yet they are smarter than all the best machines that we have on the planet.” Looking to the future, Marcus advised young AI researchers to avoid following the trend of working exclusively on LLMs. “The brightest students should try to find their own path. Large language models are not going to last forever. Someone who invents something really different could transform the world,” he concluded.","excerpt":"Back in 2019, Marcus had predicted that language models wouldn’t lead to AGI.","categories":["Global Tech"],"tags":["AI &amp; ethics","gary marcus"],"author_name":"Siddharth Jindal","publish_date":"2024-11-25T13:23:35","publication_year":"2024","word_count":1111,"keywords":["Go","artificial intelligence","OpenAI","AI","AI &amp; ethics","GPT-4o","AWS","GPT","gary marcus","Aim","deep learning","R"],"extracted_tech_keywords":["AI","artificial intelligence","deep learning","GPT-4o","OpenAI","Aim","AWS","R","Go","GPT"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/why-you-should-take-gary-marcus-seriously\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10065250,"title":"Council Post: Working with the government to implement AI solutions","content":"By now, we all know that AI is being rapidly deployed by businesses across domains. In fact, as per the state of AI report in 2021 by McKinsey, 56 per cent of all respondents reported adoption of AI in at least one function, up from 50 per cent in 2020. It also said that across regions, the adoption rate is highest at Indian companies, followed closely by those in Asia–Pacific. But even though we see the private sector embracing AI for businesses, the public sector (which sits on huge amounts of data) is still somewhat reluctant to open to the idea of it. The building block for any AI-based platform or system is data, and the governments already have large amounts of it at their disposal, which can be used. Why it is important for governments to use AI A business serves a particular customer base, whereas governments have to serve the whole nation or state. Their adoption of AI becomes all the more important due to the sheer magnitude of people they have to cater to. Though some sections of society are still sceptical about the advantages of AI, policymakers should by now be able to realise that if used properly, within the right framework and regulation, AI can greatly impact public services. In fact, in different parts of the world, governments have already deployed AI-based models and mechanisms to simplify public services and boost efficiency. AI can help provide accurate information to citizens quickly and automatically to the individuals and organisations and even help them with many relevant and data-driven predictions to take necessary actions on time. Other than the fact that the governments have a lot of data, they also have well-documented processes and manuals, which help substantially create and train the data model, leveraging various AI implementations. Now, AI is being used across all the sectors like Medical Sciences, Defence, Agriculture, Transportation, Utilities, Travel & Tourism, BFSI, Retail, Education, Entertainment, Legal Systems and more. Legal system The use of AI, ML and NLP can greatly benefit the legal system of countries that are often overburdened with a huge backlog of cases, insufficient staff, and inefficiency. There are a variety of processes in law and order that are repetitive, require long documentation, and scheduling of hearings in which technologies can really step in and improve case flow management. The pandemic saw the electronic courts, which can greatly help to reduce pending cases by performing more efficiently with better court management facilities. Tech-based Alternate Dispute Resolutions mechanisms are also a great way to reduce the burden on courts to focus on more crucial cases. Robots and virtual assistants (VideoBots, VoiceBots, ChatBots) Addressing customer grievances for any kind of public service offered is a crucial part of service delivery by the government. They usually employ people to interact with the citizens to get to know their issues and solve them effectively. But with the advent of robots, and virtual assistants (such as VideoBots, Voicebots, ChatBots), it is possible to do the same thing without much human interaction. Often there are complaints that grievance redressal platforms have not been able to solve the issues of citizens on time, which has caused them great inconvenience. Deploying chatbots for grievance redressal can help solve such problems. Chatbots can solve complaints quickly and can also be cost-effective for governments, and the benefits can be used in more critical areas. Education The right education system can transform any country by building the right human capital for the future. Though AI cannot replace a teacher, it can act as an assisting tool in the classroom. In addition, AI can be used for more interactive learning systems. Dropping out of school is an important issue in developed countries. Predictive analytics can be used to analyse data on students to get insights on students that are more likely to drop out due to various reasons and find ways to make them stay in the education system. How cutting edge tech can help government departments Various government departments can use AI and ML techniques to boost their efficiency. ​​Improve the top line via lead-generation and engagement (up-sell and cross-sell) Improve the bottom line via round-the-clock AI virtual assistant, which automates 70% of the customer support Improve customer experience via personalisation and human-centred conversational AI capabilities (multi-format, multi-lingual and omnichannel) This article is written by a member of the AIM Leaders Council. AIM Leaders Council is an invitation-only forum of senior executives in the Data Science and Analytics industry. To check if you are eligible for a membership, please fill out the form here.","excerpt":"The building block for any AI-based platform or system is data, and the governments already have large amounts of it at their disposal.","categories":["AI Features"],"tags":["AI in Education","AI in governance","AI in government","AI in legal sector in India"],"author_name":"Ankush Sabharwal","publish_date":"2022-04-19T17:00:00","publication_year":"2022","word_count":765,"keywords":["AI in Education","data science","AI","chatbots","AI in government","ML","AI in legal sector in India","virtual assistants","RAG","NLP","Aim","analytics","predictive analytics","AI in governance"],"extracted_tech_keywords":["AI","ML","NLP","data science","analytics","Aim","RAG","chatbots","virtual assistants","predictive analytics"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/council-post-working-with-the-government-to-implement-ai-solutions\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10164053,"title":"Altman Asks if OpenAI Should Release an Open Source o3-mini or Phone-Sized Model","content":"OpenAI is considering its next open-source AI project. CEO Sam Altman asked users on X whether they prefer an o3-mini-level model or a phone-sized version. “For our next open-source project, would it be more useful to create an o3-mini-level model, which is relatively small but still requires GPUs, or the best phone-sized model we can develop?” Altman asked in a poll on X. This comes after Altman, in his recent blog post, said that the company is planning to release open-source models in the near future. “AI will seep into all areas of the economy and society; we will expect everything to be smart, ” Altman wrote. “Many of us expect to need to give people more control over the technology than we have historically, including open-sourcing more.” Similarly, in an interview with Sky News, Altman said, “I think we should probably open source somewhat more.” During a Reddit ‘ask-me-anything’ (AMA) session, he said, “I personally think we have been on the wrong side of history here and need to figure out a different open-source strategy.” The discussion comes amid a broader industry push for open-source AI, following the success of models like DeepSeek. Venture capitalist Chamath Palihapitiya commented on the debate, saying, “I think in the war of open versus closed, open has won.” He suggested that OpenAI is now considering ways to open-source parts of its work. Meanwhile, Altman also announced the roadmap for the upcoming models GPT-4.5 and GPT-5. He revealed that GPT-4.5, internally referred to as Orion, will be the next release and the last non-chain-of-thought model. Following this, OpenAI plans to unify the o-series and GPT-series models, enabling systems that can integrate all tools and determine optimal thinking time for tasks. GPT-5 will be introduced in ChatGPT and the API as a system incorporating various OpenAI technologies, including o3. OpenAI will not release o3 as a standalone model, Altman said, adding that GPT-4.5 or GPT-5 could be released within weeks or months.","excerpt":"“I think we should probably open source somewhat more.”","categories":["AI News"],"tags":["OpenAI"],"author_name":"Siddharth Jindal","publish_date":"2025-02-18T18:24:27","publication_year":"2025","word_count":327,"keywords":["ChatGPT","API","GPT-5","OpenAI","AI","venture capital","GPT-4.5","GPT","R","llm_models:GPT"],"extracted_tech_keywords":["AI","GPT-4.5","GPT-5","ChatGPT","OpenAI","R","API","GPT","venture capital","llm_models:GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/altman-asks-if-openai-should-release-an-open-source-o3-mini-or-phone-sized-model\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":8057,"title":"Deploying Big Data Technology for 360° Customer View","content":"The advent of Big Data in the last few years have made a huge paradigm shift in how organizations across industries make data driven decisions. Customer behavior is a key aspect where companies have been using modern Big Data technologies to get a 360- degree view of customers from different angles. With more and more products and services being available digitally, companies want to know more and more about their existing customers and prospective new customer not just within the parameters of their interaction, but also on what else he or she is interested in. A holistic view of a customer personality helps companies not only to tailor their products and services better, but also communicate with greater relevance to differentiate themselves from the clutter. Analysis of customer sentiments, changing needs, and evolving trends can be tapped by merging in diverse internal and external information. Internal data, typically owned by the company contains data like transaction details, campaign performances, customer service history, loyalty programs, and customer satisfaction. External data covers category information, demographics and geographic details, lifestyle – attitude- psychographics and social media data. The challenges in integrating these data sources are vast, and can’t be handled by traditional systems. Big Data platforms have the capability to deal with such vast variety of data, which are terabytes in volume and are generated at a very high speed. Distributed computing and cloud systems enable merging, handling and result generation from these diverse data sources. The Big Data Environment: Big Data technologies have already proven to be highly advantageous over the traditional RDBMS systems, which are critical for modern Customer Management: Speed – High Speed Data Mining due to parallel processing Volume – Handles large amount of data Set Time – Has lower set up time Cost – Cheaper than RDBMS for significantly large size of data Scalability – Highly scalable and flexible with growing need Variability – Handles unstructured data and ready for text analytics With greater clarity, organizations can drive up their revenue by getting the right information for the right customer at the right time and act on it. Broadly there are 5 steps to using Customer 360 info for maximum results: Listen – to what the existing customers are saying about the company products and services. Improve areas of concern, build on areas of strength Understand – how customers interact with the company and category in general. Get early symptoms of industry changes and new trends. Assess – how well customers are served currently. Identify areas of improvement, cross-sell upsell opportunities Improve – interactions through customization. Each customer should feel he is getting individualized attention from the company. His experience of every interaction should be smooth and easy going, leaving him with a wow feeling every time he is in touch with the company Strengthen – relationships with customers by engaging with them through channels and offerings most relevant to him or her The customer view approach should be considered across: Past – A meaningful and accessible view that includes purchases, interactions across channels, campaign response etc. Present – Key information about who they are and how they relate to the brand, triggers for current interactions and recent history Future – Actions that can influence future behavior and relationship. Actions to prevent churn, drive up-sell & cross-sell & maximize LTV Customer 360 – Build Stages & Data Types The journey for Customer 360 process consists of 4 broad phases: Create Unique Global Consumer ID – Create coverage for members, non-members, new customers, across brand integration, regional overlaps, household information etc. Internal data – Transaction information of existing customers, historical customer service details, customer satisfaction measures, web activities, loyalty programs, social activities External data – Category information, related category trends and events, geographic and demographic profile, attitude and psychographics information, competition analysis Predicted – Propensity and affinity of purchase, pursuadableness of customers for upsell cross-sell, conversion of prospective customers. The trend is set, movement is on. Some industries have adopted Big Data earlier than others and are already seeing significant improvement in their customer satisfaction and revenue upliftment. Big Data is fast becoming the standard across companies and industries for improved Customer 360 degree view and action. Authors: Mr.Titir Pal(Head of Products & Solutions at Absolutdata Analytics) Mr. Rajat Narang (Associate Director at Absolutdata Analytics)","excerpt":"The advent of Big Data in the last few years have made a huge paradigm shift in how organizations across industries make data driven decisions. Customer behavior is a key aspect where companies have been using modern Big Data technologies to get a 360- degree view of customers from different angles. With more and more […]","categories":["IT Services"],"tags":["Absolutdata"],"author_name":"AIM Media House","publish_date":"2015-10-09T06:06:01","publication_year":"2015","word_count":716,"keywords":["big data","Go","AI","distributed computing","Scala","Git","RAG","analytics","GAN","R","Absolutdata"],"extracted_tech_keywords":["AI","analytics","RAG","distributed computing","R","Go","Scala","Git","big data","GAN"],"url":"https:\/\/analyticsindiamag.com\/it-services\/deploying-big-data-technology-for-360-customer-view\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10081044,"title":"Javascript, SSL, Mozilla: The Legacy Netscape Left Behind for Developers","content":"In 1994, a press release read: “Netscape Communications Corporation today announced that it is offering its newly introduced Netscape(TM) network navigator free to users via the Internet.” That was when Netscape released its first commercial browser for free. Netscape was probably one of the most important companies of its time and if you had access to a computer and the web in the early 90s, chances are you were using a Netscape browser. However, the company failed to live up to expectations and despite a stellar IPO, failed to compete with the likes of Microsoft. Netscape ended up disbanded, but it did leave a legacy behind. The Netscape era When the world wide web was opened to the public in 1991, it changed the world. Netscape became the first company to capitalise on it with founders Jim Clark and Marc Andreessen seen as revolutionaries. Clark and Andreessen realised the potential of the web and the need for a tool to access and navigate the web. Within months of its release in 1994, Netscape had nearly 75% of the browser market. A year later, the company went public. The Wall Street Journal described Netscape’s IPO as one of the most stunning debuts in Wall Street history. Stock valuations soared on the first day of trading. However, Netscape had a very powerful competitor – Microsoft. The tech giant launched Internet Explorer in August 1995 and thus began the browser war. In 1998, AOL acquired Netscape and disbanded it in 2003. Four years later, AOL pulled the plug on support for the Netscape web browser. The legacy of Netscape While at the cusp of being acquired by AOL, co-founder Andreessen publicly released the source code for Netscape Navigator 4.0. “By giving away the source code for future versions, we can ignite the creative energies of the entire net community and fuel unprecedented levels of innovation in the browser market,” Jim Barksdale, then CEO of Netscape had said. This led to the creation of Mozilla and its Firefox browser. In fact, in 1998, when Mozilla was formed, it was funded by Netscape. Over the years, Firebox has emerged as one of the best alternatives to Google Chrome. Further, the concept of cookies too came from Netscape. Although Lou Montulli, 23 at that time, invented the cookies we love to hate, Montulli and Netscape’s intentions were not to track users, but to solve the memory problem of websites. He wanted the websites to remember the users who had visited the page earlier, and hence, cookies were born. Montulli even holds a patent for his innovation. The birth of Javascript Today, Javascript is among the most-widely used programming languages out there. Javascript was created at Netscape by Brendan Eich in 1995 as a companion language to Java. Eich convinced his boss that a browser needs a scripting language, and in 1995, Javascript was born. With Javascript, Netscape was trying to solve the problem of websites being too static. They wanted to make the web more dynamic, and Eich just took 10 days to write it. Javascript allowed web pages to be fun, add special effects, and create an interactive user experience. Today, millions of developers use Javascript, and nearly 95% of the websites are built on the scripting language developed at Netscape. JavaScript is what enables on-page web analytics and data collection. Javascript remains strong despite multiple attempts by Google, Microsoft, and Adobe to replace it with their own programming languages. Netscape’s SSL protocol The Secure Sockets Layer (SSL) protocol developed by Netscape is probably one of the most important technological innovations of our time. Taher Elgamal, who is regarded as the father of SSL, worked as a chief scientist at Netscape Communications from 1995 to 1998. If you are ordering a book from Amazon or a pizza from Zomato and paying online for it, you have to thank Elgamal and the SSL protocol he developed. The underlying technology of SSL powers today’s security standard, Transport Layer Security (TLS). Netscape pioneered SSL, the technology which allows you to carry out a transaction over the web in a secure way. SSL establishes an encrypted link between your computer and the website you are browsing. “It’s a bullet-proof rock solid way of authenticating two endpoints and having a secure connection. We invented that,” Clark said. The technology facilitates sharing of personal information, such as credit\/debit card numbers, over the internet. It helps protect the flow of information from external forces. Founders who shaped today’s technology After Netscape, founders Clark and Andreessen went their separate ways. Andreessen founded Opsware with Ben Horowitz, Tim Howes, and In Sik Rhee. It was one of the first companies to provide software as a service. Later, Andreessen, along with his business partner Horowitz, founded Andreessen Horowitz (a16z). His portfolio includes names such as Facebook, GitHub, Pinterest, Twitter, and Foursquare among others. Often regarded as a super investor, Andreessen also had a stake in Skype, prior to it being acquired by Microsoft. He also served as a board of director at eBay for nearly six years. Similarly, after Netscape, Clark founded myCFO, a wealth management firm. He was also a founding director of biotech firm DNA Sciences, which later got acquired by Genaissance Pharmaceuticals. In 2020, the internet pioneer founded Beyond Identity, a firm that wants to eliminate the use of passwords. Clark acknowledged that his firm’s novel approach to passwordless authentication was first laid nearly three decades ago – at Netscape. With technologies such as the SSL protocol, the Firefox browser and Javascript, Netscape has left a rich legacy behind. In Brendan Eich’s words, Netscape changed history. Even though it has ceased to exist, tech developed at Netscape continues to have significant relevance in this day and age.","excerpt":"Netscape had a very powerful competitor, Microsoft, which launched Internet Explorer in August 1995 and thus began the browser war. In 1998, AOL acquired Netscape but disbanded it in 2003","categories":["AI Features"],"tags":[],"author_name":"Pritam Bordoloi","publish_date":"2022-11-29T17:00:00","publication_year":"2022","word_count":955,"keywords":["Go","API","GenAI","AI","Git","analytics","JavaScript","GitHub","R","Java"],"extracted_tech_keywords":["AI","analytics","GenAI","R","JavaScript","Go","Java","Git","GitHub","API"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/javascript-ssl-mozilla-the-legacy-netscape-left-behind-for-developers\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":31474,"title":"Look Beyond Tableau With These Cost-effective, Homegrown BI Tools","content":"Even though Tableau is a trailblazer in the BI space and has emerged as a leader in self-service BI tools, the Seattle-based software giant is not without competition. There has been a spurt in homegrown and global dashboard software market players that are nipping at the heels of BI leader. Besides the biggies in the BI space such as — SAP, Microsoft, TIBCO Software, there is a slew of new BI tools in the market, touting predictive modelling, the ability to simplify complex datasets and cost-effectiveness as the unique differentiator. Smarten By ElegantJ BI The platform encourages its users to become citizen data scientists by integrating multiple sources of complex data. Smarten by ElegantJ BI also bridges the gap between BI and users by auto-selecting the algorithms which help the users to become efficient data scientists. The technology is built in algorithms that help the users to prepare and get the desired result. The user can select clustering or forecasting depending upon the necessity, these selections within the board are dependent on the platform. For Smarten, one doesn’t require any prior coding experience, thereby making it user-friendly. With its advanced architecture, Smarten also helps in cutting down the requirement of data scientists and data specialist. With its end to end platform, the self serves data preparations are framed and support deep-dive analytics, dashboards. BDB BizViz BDB is a Bengaluru-based company and provides a cloud platform with robust technology. The company’s BI software provides advanced analytics and enables exploring ad-hoc statistical analysis, predictive modelling, real-time scoring, machine learning and elasticsearch. BDB Decision Platform is an end-to-end data analytics platform which ingests data from streaming, structured, unstructured data sources and identifies patterns and trends in structured and unstructured data. With strong predictive analysis and the scope to integrate with R and Spark ML, users are free to write in their desired programming languages. The platform also supports Spark ML Scala & Python to create models as expected by the clients. These features give new opportunities, minimize risks, and boost revenue. Ezlytix Founded by Sachin Bakshi, the company strives to provide business growth and compete with present-day issues. The technology developed by this startup helps small businesses find insights from their data. This self-service BI startup is helping small to mid-sized companies increase their profitability by analyzing the customer products and costs. Ezlytix cloud-based analytics provides solutions to real business issues by giving right information within a period of 5 days, simple software and hardware, low risk in IT-projects, economical services. The startup provides a turn-key solution to help mid-sized businesses focus on ‘profitable growth’ by leveraging data and automation. The technology identifies possibilities such as an increase in sales and margins, dropping the low performing products, utilizing information to obtain a competitive advantage over identifying market possibilities. Phrazor From vPhrase ’PHRAZOR’, the AI platform can convert structured data, analyses it and generates insights which can be used for effective decision-making. These insights are delivered in a humanized narrative and over a period of time, the startup has built a considerable following among leading companies in India. The tool is widely used by banks like HDFC to drum up personalized narrative reports for their employees and in some scenarios for their clients as well. The technology integrates with the client’s system with the help of API and in some cases, the company provides an on-premise installation as well. The firm extends its services in the fields of quarterly earning analysis, stories, sales points from mutual fund factsheets, performance appraisal reports and many more. Narrator Narrator is G-Square’s Robo analytics tool. It provides a descriptive and smart business intelligence solution to the users. It provides actionable business insights along with the visualisation and dashboards. As of now, it has three modules – Robo analysis, Ask Narrator and Custom analysis which fits the usage requirements of both the business users and the analytics teams. Some of the unique advantages of the Narrator are — it provides a wholesome solution for business needs including smart visualisation, dashboards, customised reports, narration of insights and interactive Chatbot in one single tool. It can handle the large size of data and saves a lot of precious time and resources for organisations as it provides automated analytics solutions on the go.","excerpt":"Even though Tableau is a trailblazer in the BI space and has emerged as a leader in self-service BI tools, the Seattle-based software giant is not without competition. There has been a spurt in homegrown and global dashboard software market players that are nipping at the heels of BI leader. Besides the biggies in the […]","categories":["AI Features"],"tags":["Business Intelligence","Business Intelligence India"],"author_name":"Bharat Adibhatla","publish_date":"2018-12-13T10:46:31","publication_year":"2018","word_count":711,"keywords":["Business Intelligence India","Elasticsearch","Go","machine learning","AI","ML","Scala","RAG","Python","analytics","Business Intelligence","R"],"extracted_tech_keywords":["AI","machine learning","ML","analytics","RAG","Elasticsearch","Python","R","Go","Scala"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/look-beyond-tableau-with-these-cost-effective-homegrown-bi-tools\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10129257,"title":"Meta AI Glasses Miss The Mark","content":"When Mark Zuckerberg went surfing in a tux just six months after his knee surgery, flaunting his moves on Instagram on the fourth of July, he also subtly marketed the Meta Ray-Ban glasses, which he wore to record videos. Thankfully, they were not damaged during his time on the water. Coincidently, the CEO’s return to the waves comes at a busy time for Meta, which continues to expand its reach in the tech industry. The company has been making headlines with its developments in virtual reality, augmented reality, and other innovative technologies. After the release of the Ray-Ban Meta Smart Glasses, netizens quickly shared their opinions. While some reviews were positive, others were critical. Mark Zuckerberg unveiling smart glasses at Meta Connect. While many praised the new Ray-Ban Meta Smart Glasses for significant improvements over the Ray-Ban Stories, some are disappointed that in order for the battery to last long, the glasses have to be switched off. Users were quick to point out that the app drains the phone battery even when the glasses are off and miles away. Meta’s Ray-Ban smart glasses offer functionalities similar to the Humane Ai Pin, but in a more traditional form. Once considered a futuristic must-have, smart glasses now seem outdated compared to the Ai Pin. “Glasses are things that have come and gone. A lot of people have attempted them,” said Imran Chaudhri of Humane, the startup developing the Ai Pin. “One of the reasons they are questionable is that only some people actually wear glasses,” Chaudhri explained. Glasses are an additional fashion choice, which many people don’t want. “If you wear glasses, it’s primarily to see and protect your eyes. There’s a lot of thought that goes into that,” Chaudhri added. Fashion’s Unexpected Underdog or Just ‘Influencer’ Branding? Nevertheless, Meta’s smart glasses are defying all odds when it comes to earning a high fashion reputation. The accessory is emerging as 2024’s hottest must-have after reaching the pinnacle of style status, TikTok virality. Touted as “the next big thing in the influencer world”, the Meta x Ray-Ban shades have become an overnight sensation among the content creator community. Engadget’s video demonstrating the smart glasses, which lets users make calls, shoot videos, listen to music, and livestream, has accumulated many views since it was uploaded. The gadget’s rise to fame isn’t a fluke—it’s all part of Meta’s marketing push to sway consumer opinion on its smart wearable, and the company is savvy about which audience it targets. The demographic in question is the style-obsessed social media natives. Over the past month, the tech corporation has orchestrated a campaign targeting fashion enthusiasts and internet personalities with impressive followings. To date, creators who have jumped on the hype include Hong Kong-based Steph Hui and Kristina Kim, all of whom have contributed to the product’s explosive growth. Tech wearables’ entry into mainstream consciousness hasn’t always been smooth sailing. Apple’s Vision Pro headset crash landed across the fashion community in June last year—a fashion victim memeified into oblivion—while Google’s anticipated Google Glass was pulled off the market after a damp squib of a launch in 2014. Meta’s smartwear division has also previously struggled to gain traction. Last April, the company was reported to have sold just 120,000 pairs of Ray-Ban Stories, less than half its 300,000 goal at that time, as well as failing to retain its existing users. In today’s day and age, when everyone can be a content creator, convenience is key; Meta has taken note by refining in on its buzzy use-case of hands-free recording, encouraging participating influencers to do the same. For example, Ray-Ban’s official account released its own video on TikTok in December with the tagline ”keep the hands-free POVs coming”, which has since accumulated over 11 million views. Meta’s explosive success so early into the new year puts tech gadgets front and centre as an area to watch in 2024. In the next few months, we could see wearables from the likes of Meta and its competitors evolve into social media’s ‘next big thing’ among influencers. While the trend builds up cultural cachet, Meta’s gambits haven’t gone unnoticed by fashion commentators and their increasingly cynical style communities. The Washington Post columnist Jules Terpak wrote on X, “It’s unclear if it organically happened amongst those with large followings after one person’s video went crazy, or if there’s an undisclosed influencer campaign going on. Regardless, it’s definitely causing a shift in perception.” A trend has been going on with the Ray Ban Meta smart glasses the past two weeks on TikTok. One video has 123 million views — many others have millionsIt’s unclear if it organically happened amongst those with large followings after one person’s video went crazy, or if there’s… pic.twitter.com\/uWeemszkuK— Jules Terpak (@julesterpak) December 17, 2023 Striving to Stay Ahead Meta emphasised the ease of operating their smart glasses, eliminating the need for hand-held cameras or smartphones. With Meta AI integration, their chatbot-like interface puts them ahead of other tech giants exploring similar technologies. In 2023, Microsoft also developed smart glasses with GPT-4V image recognition. Meta’s 2023 AR\/VR hardware roadmap revealed that their third-generation smart glasses, expected in 2025, will include a viewfinder for reading messages, scanning QR codes, and a neural interface band for hand movement control. Meta leads in VR with the recent launch of mixed reality Quest 3 headsets, surpassing Apple Vision Pro’s upcoming release next year.","excerpt":"These reality glasses are “a time machine into the future”.","categories":["Deep Tech"],"tags":["AI (Artificial Intelligence)","Meta"],"author_name":"Tarunya S","publish_date":"2024-07-16T13:29:13","publication_year":"2024","word_count":898,"keywords":["Go","Meta AI","Meta","AI","RPA","image recognition","RAG","GPT","Ray","GAN","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","Meta AI","Ray","RAG","image recognition","R","Go","GPT","GAN","RPA"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/meta-ai-glasses-misses-the-mark\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":36691,"title":"Why Data Science Requires A Dedicated Job Portal In India","content":"Data Science as an emerging field has undergone explosive growth in the past few years, owing to the wide adoption of technologies such as artificial intelligence and machine learning among SMEs. The phenomenon is very visible In the Indian subcontinent especially, where the Silicon Valley of India, Bangalore, sees new AI-focused startup being started up regularly. This has created an environment where talent is required, and fast. Sites such as LinkedIn, Naukri and Monster have all been widely used by every data scientist throughout their careers. However, these portals have their own pitfalls, and end up providing an unsatisfactory experience for the minute specializations required for data science jobs. These factors contribute towards the need for a standalone, dedicated portal for data science jobs. Why Data Science Is A Difficult Field To Quantify Job sites such as Linkedin offer a restricted environment for companies to post jobs, along with a general feeling of getting lost in a sea of other postings on the platform. Moreover, the state of data science as a complex field of occupation is also seen when looking at it as a concept. While it has been called the “Sexiest Job Of The 21st Century” by Harvard Business Review, it is one of the most complex positions one can hold. The occupation of a data scientist comes at the conjunction of many important business skills and technical knowledge. The occupation requires applicants to have considerable skill in computer science, owing to the pervasivity of software in ML solutions today, but also be able to predict the outcome of his insights on the business as a whole. While this is usually delegated to big-picture activities taken care of high-level executives, the data scientist working from the ground up will have a better idea of how the deliverable can impact the business as a whole. Moreover, they are also required to have a high level of skill in written and spoken communication, as they operate in complex concepts that cannot be easily conveyed. This also extends towards creating project reports or outcome analyses, which both need a high tendency for the scientist to get what they are saying across. Data scientists will also have to work on deliverables across verticals, such as prediction, classifications, recommendations, anomaly detection, actionable insights, scoring and ranking, optimization and forecasts. Above all of this, data scientists are also required to have a strong background in statistics, so as to derive insights from large amounts of data. This adds another skill to the required skill set of data scientists. This skill will be used to create comprehensive and easy-to-implement strategies that harness the power of statistics to ensure a favourable outcome. Data scientists will also need to know multiple programming languages, such as R and Python, along with big data frameworks such as Apache Spark, Hadoop and SQL databases. All of this contributes towards the need for a dedicated job portal for data science professionals. Disconnect Between Hirers And Professionals Organizations have also found it extremely difficult to hire good talent for data scientist roles. Even as we have one of the biggest data scientist and programmer populations in the world, companies still find it difficult to begin hiring momentum. This is caused by a variety of factors. Firstly, the definition of ‘data scientist’ varies from organization to organization, with job sites not being to accurately convey the requirements for various positions. Added to this, there are multiple other positions such as data analyst and data technician that do not make sense to a new job seeker. The lack of representation for differences in these positions and their accompanying requirements has caused an environment of misinformation in the data science hiring field. Moreover, employers struggle to even connect with the talent that the required where their posting is buried under a deluge of similar ones. While this not only results in frustration of not being able to find the right person for the right job, it also contributes to the image of not finding suitable talent where, in reality, data science professionals are looking for new opportunities. Making The Case For A Dedicated Portal For Data Science Positions As many of data scientists know, it is important to keep growing and learning. This was our motivation to start Analytics India Magazine, adding to the then-budding analytics community. Today, we have become a trusted source of information, tutorials and events in the analytics and data science field. For bringing us here, we wanted to give back to our community in a unique way. These factors drove us at Analytics India Magazine to create a dedicated portal for data science and analytics openings. After interactions with companies in the analytics space, we discovered a disconnect between those who wish to hire and those who want to be hired. You, our community, have been vocal in your search for jobs across the fields of data science and analytics. Companies, on the other hand, maintained that it was difficult to find such people. Bringing together India’s fastest-growing community of data scientists with premier choices for job opportunities, Analytics India Jobs gives back to our faithful readers by offering them an opportunity to advance in their careers. Harnessing the power of being one of India’s premier data science destinations, we aim to provide a unique way to create a connection for those wishing to find better jobs. Those interested can apply at Analytics India Jobs.","excerpt":"Data Science as an emerging field has undergone explosive growth in the past few years, owing to the wide adoption of technologies such as artificial intelligence and machine learning among SMEs. The phenomenon is very visible In the Indian subcontinent especially, where the Silicon Valley of India, Bangalore, sees new AI-focused startup being started up […]","categories":["AI Hirings"],"tags":["Analytics India Magazine","Data Science"],"author_name":"Anirudh VK","publish_date":"2019-03-21T09:43:48","publication_year":"2019","word_count":905,"keywords":["data science","machine learning","artificial intelligence","AI","ML","Apache Spark","Analytics India Magazine","Python","Aim","anomaly detection","analytics","Data Science"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","data science","analytics","Aim","anomaly detection","Apache Spark","Python"],"url":"https:\/\/analyticsindiamag.com\/ai-hiring\/why-data-science-requires-a-dedicated-job-portal-in-india\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":66178,"title":"Fashion Apparel Recognition using Convolutional Neural Network","content":"Recent advances in deep learning have triggered a variety of business applications based on computer vision. There are many industry segments where deep learning tools and techniques are applied in object recognition in order to make the business process much faster. The apparel industry is one among them. By presenting the image of any apparel, the trained deep learning model can predict the name of that apparel and this process can be repeated at a very much faster speed in order to tag thousands of apparels in very less time with high accuracy. In this article, we will discuss fashion apparel recognition using the Convolutional Neural Network (CNN) model. To train the CNN model, we will use the Fashion MNIST dataset. After successful training, the CNN model can predict the name of the class a given apparel item belongs to. This is a multiclass classification problem in which there are 10 apparel classes the items will be classified. The Data Set In this article, we have used the Fashion MNIST data set that is publicly available on Kaggle. It consists of a training set of 60,000 example images and a test set of 10,000 example images. Each image in the dataset has the size 28 x 28 pixels. Each training and test image belongs to one of the classes including T_shirt\/top, Trouser, Pullover, Dress, Coat, Sandal, Shirt, Sneaker, Bag, and Ankle boot. The original training and test image data sets are converted into CSV files and made available on Kaggle. Implementation This execution is done in Google Colab and to read the CSV files there, we first uploaded the CSV files to Google Drive and then mounted the drive using the following lines of codes. #Setting google drive as a directory for dataset from google.colab import drive drive.mount('\/content\/gdrive') Once Google Drive is mounted, we will read our training and test CSV files using the below lines of codes. #Reading dataset import pandas as pd fashion_train_df = pd.read_csv('gdrive\/My Drive\/fashion-mnist_train.csv',sep=',') fashion_test_df = pd.read_csv('gdrive\/My Drive\/fashion-mnist_test.csv', sep = ',') After successfully reading the data sets, we will import the other required libraries. #Importing other required libraries import pandas as pd import numpy as np import matplotlib.pyplot as plt import seaborn as sns import random sns.set_style(\"whitegrid\") The data sets that we have read above, we will see their shapes. As we discussed earlier, there are 60,000 examples in the training set and 10,000 examples in the test set. #Shape of training data fashion_train_df.shape #Shape of test data fashion_test_df.shape Since the size of each image is 28 x 28 hence there are a total of 784 pixels of each image and there is one column of class label. That is why a total of 785 columns are there in the dataset. Now, in order to define the training and test data sets, first, we need to create the training and test arrays. # Create training and testing arrays train = np.array(fashion_train_df, dtype = 'float32') test = np.array(fashion_test_df, dtype='float32') The below line of code specifies the class labels of the data set, as given in the description on Kaggle. #Specifying class labels class_names = ['T_shirt\/top', 'Trouser', 'Pullover', 'Dress', 'Coat', 'Sandal', 'Shirt', 'Sneaker', 'Bag', 'Ankle boot'] Now, to proceed further and validating the class label, we will pick and plot a random image from the set of 60,000 training images to verify its correct class label. The below lines of codes can pick and plot a different image randomly in each run. #See a random image for class label verification i = random.randint(1,60000) plt.imshow(train[i,1:].reshape((28,28))) plt.imshow(train[i,1:].reshape((28,28)) , cmap = 'gray') label_index = fashion_train_df[\"label\"][i] plt.title(f\"{class_names[label_index]}\") plt.axis('off') To verify the same, we will see the class label of the above randomly selected image and match the label with the label name. #Label of the random image label = train[i,0] label As we are confirmed about the class label and class name with one randomly selected image, now will visualize more random images with class labels and class names. The number of images to be chosen can be adjusted by changing the values of width and length grid W_grid and L_grid respectively. We could use subplot but it returns the figure object and axes object. Here, we can use the axes object to plot specific figures at various locations. # Define the dimensions of the plot grid W_grid = 15 L_grid = 15 fig, axes = plt.subplots(L_grid, W_grid, figsize = (17,17)) axes = axes.ravel() # flaten the 15 x 15 matrix into 225 array n_train = len(train) # get the length of the train dataset # Select a random number from 0 to n_train for i in np.arange(0, W_grid * L_grid): # create evenly spaces variables # Select a random number index = np.random.randint(0, n_train) # read and display an image with the selected index axes[i].imshow( train[index,1:].reshape((28,28)) ) label_index = int(train[index,0]) axes[i].set_title(class_names[label_index], fontsize = 8) axes[i].axis('off') plt.subplots_adjust(hspace=0.4) We can run the above set of codes to verify class name for images. In each run, a random set of images will be visualized. Now we are correct with the class labels and names of all the images. In the next step, we will prepare the training and test data. # Prepare the training and testing dataset X_train = train[:, 1:] \/ 255 y_train = train[:, 0] X_test = test[:, 1:] \/ 255 y_test = test[:,0] We will visualize a set of 25 training image data that will be used to train the convolutional neural network model. plt.figure(figsize=(10, 10)) for i in range(25): plt.subplot(5, 5, i + 1) plt.xticks([]) plt.yticks([]) plt.grid(False) plt.imshow(X_train[i].reshape((28,28)), cmap=plt.cm.binary) label_index = int(y_train[i]) plt.title(class_names[label_index]) plt.show() For the training and validation purpose, we split the data set accordingly. The test data size can be adjusted after a run of the model. #Split the training and test sets from sklearn.model_selection import train_test_split X_train, X_validate, y_train, y_validate = train_test_split(X_train, y_train, test_size = 0.2, random_state = 12345) print(X_train.shape) print(y_train.shape) Here, we will unfold the data to make it available for training, testing and validation purpose. # Unpack the training and test tuple X_train = X_train.reshape(X_train.shape[0], *(28, 28, 1)) X_test = X_test.reshape(X_test.shape[0], *(28, 28, 1)) X_validate = X_validate.reshape(X_validate.shape[0], *(28, 28, 1)) print(X_train.shape) print(y_train.shape) print(X_validate.shape) To define and train the convolutional neural network, we will import the required libraries here. #Library for CNN Model import keras from keras.models import Sequential from keras.layers import Conv2D, MaxPooling2D, Dense, Flatten, Dropout from keras.optimizers import Adam from keras.callbacks import TensorBoard Convolutional Neural Network In the below line of codes, we will define our convolutional neural network model. For more understanding about the convolutional neural network, please refer to the article ‘Overview of Convolutional Neural Network in Image Classification’. #Defining the Convolutional Neural Network cnn_model = Sequential() cnn_model.add(Conv2D(32, (3, 3), input_shape = (28,28,1), activation='relu')) cnn_model.add(MaxPooling2D(pool_size = (2, 2))) cnn_model.add(Dropout(0.25)) cnn_model.add(Conv2D(64, (3, 3), input_shape = (28,28,1), activation='relu')) cnn_model.add(MaxPooling2D(pool_size = (2, 2))) cnn_model.add(Dropout(0.25)) cnn_model.add(Conv2D(128, (3, 3), input_shape = (28,28,1), activation='relu')) cnn_model.add(MaxPooling2D(pool_size = (2, 2))) cnn_model.add(Dropout(0.25)) cnn_model.add(Flatten()) cnn_model.add(Dense(units = 512, activation = 'relu')) cnn_model.add(Dropout(0.25)) cnn_model.add(Dense(units = 10, activation = 'softmax')) cnn_model.summary() After defining the CNN model and viewing its summary, we will bind-up this model by compiling it. #Compiling cnn_model.compile(loss ='sparse_categorical_crossentropy', optimizer='adam' ,metrics =['accuracy']) In the next step, we will train our CNN model on the image classification. The below hyperparameters can be tuned for better accuracy of the model. #Training the CNN model history = cnn_model.fit(X_train, y_train, batch_size = 512, epochs = 200, verbose = 1, validation_data = (X_validate, y_validate)) After successful training, we will visualize the loss and accuracy of the model through a plot using below lines of codes. #VIsualizing the training performance plt.figure(figsize=(12, 8)) plt.subplot(2, 2, 1) plt.plot(history.history['loss'], label='Loss') plt.plot(history.history['val_loss'], label='val_Loss') plt.legend() plt.title('Loss evolution') plt.subplot(2, 2, 2) plt.plot(history.history['accuracy'], label='accuracy') plt.plot(history.history['val_accuracy'], label='val_accuracy') plt.legend() plt.title('Accuracy evolution') We could perform training in multiple iterations by tuning the hyperparameters to see an increase in the accuracy of the model. But we found this consistent level with these values of hyperparameter and 200 epochs of training. As the model is trained successfully and we could achieve an accuracy of more than 93% during training and more than 90% during validations, we consider this CNN model as best fitted with our data. So now will make predictions using this CNN model on test data. The model is expected to produce class labels as the output of prediction. #Predictions for the test data predicted_classes = cnn_model.predict_classes(X_test) test_img = X_test[0] prediction = cnn_model.predict(test_img) prediction[0] np.argmax(prediction[0]) As we can see above that the model has predicted the class label 0 for the given image. Now, we will check the prediction on more images. We are taking 49 images as a test set and predicting their class labels and comparing the predicted class labels with true class labels. L = 7 W = 7 fig, axes = plt.subplots(L, W, figsize = (18,18)) axes = axes.ravel() for i in np.arange(0, L * W): axes[i].imshow(X_test[i].reshape(28,28)) axes[i].set_title(f\"Prediction Class = {predicted_classes[i]:0.1f}\\n True Class = {y_test[i]:0.1f}\") axes[i].axis('off') plt.subplots_adjust(wspace=0.5) For the window limitation, we have taken only 7 x 7 = 49 images, but one can take more images to predict class labels. As we can see in the above visualization, for all 49 images, our CNN model has predicted correct class labels for 45 images and incorrect class labels for 4 images. For better understanding, let us visualize the total classification done by the model using confusion matrices. For better visualization of the confusion matrix, first, we will define the class labels and then create the confusion matrix. class_names = ['T_shirt\/top', 'Trouser', 'Pullover', 'Dress', 'Coat', 'Sandal', 'Shirt', 'Sneaker', 'Bag', 'Ankle boot'] from sklearn.metrics import confusion_matrix from sklearn import metrics cm = metrics.confusion_matrix(y_test, predicted_classes) For better visualization of the confusion matrix, a function ‘plot_confusion_matrix’ is being used here. #Defining function for confusion matrix plot def plot_confusion_matrix(y_true, y_pred, classes, normalize=False, title=None, cmap=plt.cm.Blues): \"\"\" This function prints and plots the confusion matrix. Normalization can be applied by setting `normalize=True`. \"\"\" if not title: if normalize: title = 'Normalized confusion matrix' else: title = 'Confusion matrix, without normalization' # Compute confusion matrix cm = confusion_matrix(y_true, y_pred) if normalize: cm = cm.astype('float') \/ cm.sum(axis=1)[:, np.newaxis] print(\"Normalized confusion matrix\") else: print('Confusion matrix, without normalization') #     print(cm) fig, ax = plt.subplots(figsize=(10,10)) im = ax.imshow(cm, interpolation='nearest', cmap=cmap) ax.figure.colorbar(im, ax=ax) # We want to show all ticks... ax.set(xticks=np.arange(cm.shape[1]), yticks=np.arange(cm.shape[0]), # ... and label them with the respective list entries xticklabels=classes, yticklabels=classes, title=title, ylabel='True label', xlabel='Predicted label') # Rotate the tick labels and set their alignment. plt.setp(ax.get_xticklabels(), rotation=45, ha=\"right\", rotation_mode=\"anchor\") # Loop over data dimensions and create text annotations. fmt = '.2f' if normalize else 'd' thresh = cm.max() \/ 2. for i in range(cm.shape[0]): for j in range(cm.shape[1]): ax.text(j, i, format(cm[i, j], fmt), ha=\"center\", va=\"center\", color=\"white\" if cm[i, j] > thresh else \"black\") fig.tight_layout() return ax Now, by calling the above function, we will visualize a non-normalized confusion matrix o see the exact number of correct and incorrect classifications. plt.figure(figsize = (20,20)) plot_confusion_matrix(y_test, predicted_classes, classes=class_names, title='Normalized Confusion matrix') plt.axis('off') Similarly, we can visualize the same confusion matrix in a normalized form to see the percentage of correct and incorrect classifications by the model. plt.figure(figsize = (20,20)) plot_confusion_matrix(y_test, predicted_classes, classes=class_names, normalize=True, title='Normalized Confusion matrix') plt.axis('off') As we can see in the above confusion matrix, our model has given the highest accuracy of 99% in recognizing bags, 98% in recognizing trousers and so on. The model has given the lowest accuracy of 79% in recognizing shirts. In recognizing apparels of more than 6 classes out of 10, it has given more than 90% accuracy and more than 85% accuracy in recognizing apparels of 9 classes. This level of accuracy in recognizing objects is definitely high. In further articles, we will check the same recognition accuracy by using different models.","excerpt":"In this article, we will discuss fashion apparel recognition using the Convolutional Neural Network (CNN) model. To train the CNN model, we will use the Fashion MNIST dataset. After successful training, the CNN model can predict the name of the class a given apparel item belongs to.","categories":["Deep Tech"],"tags":["cnn neural network","Convolution Neural Network","Convolutional Neural Network","Deep Learning","multiclass classification"],"author_name":"Dr. Vaibhav Kumar","publish_date":"2020-05-29T12:07:23","publication_year":"2020","word_count":1932,"keywords":["NumPy","Keras","AI","neural network","ML","cnn neural network","computer vision","Colab","Ray","Convolution Neural Network","multiclass classification","deep learning","Convolutional Neural Network","Deep Learning","Pandas"],"extracted_tech_keywords":["AI","ML","deep learning","neural network","computer vision","Ray","Keras","Colab","Pandas","NumPy"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/fashion-apparel-recognition-using-convolutional-neural-network\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10057464,"title":"A Beginner’s Guide to Deep Regression Forest","content":"Random forest and decision trees are some of the most popular predictive models in the machine learning field. When using random forests, we can find different variants of it that can be used in classification and regression analysis. In this article, we are going to discuss a variant of the random forest named as Deep Regression Forest, which integrates the features of deep neural networks and can outperform in various applications of regression analysis. The major points to be discussed in this article are listed below. Table of Contents What is Deep Regression Forest? When to apply the Deep Regression Forest?How Does It Work?Application of Deep Regression Forest What is Deep Regression Forest? In traditional machine learning, we find that random forest and decision trees are some of the popular algorithms for predictive modelling which use the structure of a tree. These algorithms perform data partition at split nodes and data abstraction at leaf nodes. These models can be used for both classification and regression problems. When a traditional forest is used for regression problems, it makes hard data partitions because these traditional forests are based on a kind of greedy approach where for data partitions at each split node, it makes locally optimal hard decisions. These locally optimal hard decisions made by the greedy algorithms are well-performing but when it comes to handling the heterogeneous data like image data for age estimation, we are required to apply the soft decisions. Deep regression forest can be considered as a method similar to the traditional regression forest but the decisions at each split node are softer than the traditional regression forest. Using the deep regression forest, we can also make the algorithm eligible to learn the input feature space and data abstraction at leaf nodes jointly. This eligibility can ensure that the input and output from leaf nodes are correlated and the correlation is homogeneous. Also, it helps in handling heterogeneous data. We can also consider deep regression forests as deferential regression forests which can also be integrated with any deep neural networks. Integrating them with deep neural networks has shown great potential in computer vision regression tasks.  We can summarize the procedure of deep regression forest with the following steps:- Fix the leaf nodes and optimize the data partitions (at split nodes).Convolutional neural networks can be used for feature learning by backpropagation. Fix the split nodes and optimize the data abstraction (at leaf nodes). Considering the above steps, we can say that the first step is for feature learning and the second step is to perform local regression. We can call this forest a combination of deep neural decision forest and label distribution learning forest. But, the objective of this forest is to perform regression tasks. When to Apply the Deep Regression Forest? As we could understand what deep regression forest is in the last section, we also got to know that it has the capability of dealing with heterogeneous data. The basic idea behind the model is to learn from low-dimensional embeddings. The reason behind the capability of dealing with heterogeneous properties is, learning data partition and data abstraction jointly. So we can say if in any process the data is heterogeneous like image data, we can apply the Deep Regression Forest for regression analysis. With this, it also has the capability of dealing with nonlinear data. Unlike other methods, we can integrate this model to any deep neural network and perform regression analysis using the homogenous partitions in the joint input-output space, and learning a local regressor for each partition. The problem of locally optimal hard decision data partition at split nodes can be solved by defining a function for soft decision at each split node. Because of the production of a soft decision, we can apply it for depth estimation. How Does It Work? The main motive of the process is based on the working of a normal regression forest. So  we can understand that the process can be completed in two normal steps: Learn from a differential regression tree.Ensemble trees to form a forest. Let’s understand how to make an algorithm work in the above-given steps. Let’s say there are two spaces  X (input space) and Y (output space). Now we know that, in regression analysis for each input information, there is an output target and the objective of regression procedures is to find a mapping function. The mapping can be given as the following: yˆ = g(x) = Z yp(y|x)dy Where, X = Input sample Y = Target sample g(X) = Mapping function p(Y|X) = Conditional probability function When we perform this regression using a decision tree, there is a set of split nodes and a set of leaf nodes. In these sets, each split node defines a split function and the leaf node contains a density distribution. Till now we could see that this process is similar to the traditional regression tree which exhibits hard decisions. To make it soft, we can introduce a split function which can include the sigmoid function and index function as: sn(x; Θ) = σ(fϕ(n)(x; Θ)) Where, X is parameterized by Θsn(x; Θ) = split function In this function, σ(·) is the sigmoid function and ϕ(·) is an index function. Let’s take a look at the below image: Image source The above image is a representation of the deep regression forest where the output unit (red circles) of the function is parameterized by Θ and a fully connected layer of the convolutional neural network. Index functions assigned to the tree are two in quantity and split nodes are in blue and leaf nodes are in green circles. To indicate the correspondence between split nodes of trees and the FC layer, black dashed lines are used. Each tree has independent leaf node distribution and the output from the forts is a mixture of tree predictions, function f(.) and leaf node distribution are learned jointly in an end-to-end manner. Based on the above scenario the mapping between x and y is given by yˆ = g(x; T ) = Z yp(y|x; T )dy. Where T is a tree-based structure. By ensembling these trees after optimization, we can form a regression tree, where it is a possibility that all the trees will share the same parameters and they can have different sets of split functions and independent leaf distributions. We can also define the loss function of the forest as the average loss function of the trees. Application of Deep Regression Forest As of now, we have seen what, when, and how deep regression forest can be used. When we talk about the application of the deep regression forest, we can say that it can be used in regression problems that can be dealt with normal random forest regression models or any neural network. Some of the normal fields of applications are: Price optimization in e-commerceDisease level prediction in medical sciencePrice prediction in stock markets.To measure the effects of water and fertilizer on crops. Also, we have seen that the major capabilities of deep regression forest are to deal with heterogeneous data. So we can find the application of it majorly in the fields with heterogeneous data. Some of the applications related to heterogeneous data  are as follows: Age estimation using the imagesCrowd counting Pose estimationDemographic analysis Visual surveillance Access control Here in the above applications, we can understand that the required results from the process are going through data that is heterogeneous. Final Words In this article, we have seen a variant of random forest in the regression analysis that is deep regression forest and we found that it has the capability of making soft decisions so that it can be used with data that is heterogeneous and requires regression analysis. Along with that we have discussed the procedure on which it works and some of the normal and crucial applications where it can be used for better results. References Deep Regression Forests for Age EstimationSoft-Margin Mixture of Regressions","excerpt":"Deep regression forest can be considered as a method similar to the traditional regression forest but the decisions at each split node are softer than the traditional regression forest.","categories":["Deep Tech"],"tags":["AI (Artificial Intelligence)","Data Science","Data Scientist","Deep Learning","Machine Learning","Python","random forest","Randomforest"],"author_name":"Yugesh Verma","publish_date":"2022-01-03T14:00:00","publication_year":"2022","word_count":1328,"keywords":["Go","machine learning","TPU","programming_languages:R","Randomforest","AI","neural network","Machine Learning","programming_languages:Go","computer vision","RAG","Python","random forest","Deep Learning","Data Science","Data Scientist","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","machine learning","neural network","computer vision","RAG","TPU","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/a-beginners-guide-to-deep-regression-forest\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10069312,"title":"TeamViewer appoints Rupesh Lunkad as India MD","content":"Teamviewer, a leading global provider of remote connectivity and workplace digitalization solutions, has appointed Rupesh Lunkad as Managing Director, India. He will succeed Krunal Patel, who played an integral role in establishing TeamViewer’s operations in India in 2018. “TeamViewer is a pioneer in remote connectivity and now also leading in digital workplace solutions based on Augmented Reality. Over the years, the company has demonstrated its capabilities to support India’s growing digital economy. I am excited to join TeamViewer to further position the company as a value-creating partner in enterprise digital transformation, especially in the context of the continuous conversion of IT and OT in Industry 4.0,” said  Rupesh. Rupesh comes with more than 16 years of experience in strategy, delivery, product management, digital transformation, and IT leadership. He is an experienced advisor to enterprise businesses and has a strong track record of helping companies with their digital transformation initiatives and align their long-term business objectives with technology strategies based on a customer-centric approach. In his tenure at SAP India, Rupesh was responsible for go-to-market strategies for new business verticals as well as for strategic initiatives leveraging its partner ecosystem for business development and expansion. He also introduced and fronted SAP’s S\/4HANA cloud business in India. At Wipro, he served in several leadership roles with enterprise sales and business responsibility. Prior to joining TeamViewer India, he has worked as Chief Revenue Officer at system integrator SAVIC Technologies where he enabled the organization to expand its business and advance investments in the areas of cloud computing, customer experience, and transformational solutions. “We are thrilled to welcome Rupesh at TeamViewer to lead our India business. With his strong sales leadership, immense technology expertise, and deep understanding of driving business in the region, we are certain that he will enable us to expand our footprint in the India market and create a strong ecosystem of alliances with our partners,” said Sojung Lee, President – Asia Pacific, TeamViewer.","excerpt":"Rupesh introduced and fronted SAP’s S\/4HANA cloud business in India.","categories":["AI News"],"tags":[],"author_name":"Tasmia Ansari","publish_date":"2022-06-17T12:17:02","publication_year":"2022","word_count":323,"keywords":["Go","programming_languages:R","cloud computing","AI","digital transformation","Git","RAG","ViT","GAN","R"],"extracted_tech_keywords":["AI","RAG","cloud computing","R","Go","Git","GAN","ViT","digital transformation","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/teamviewer-appoints-rupesh-lunkad-as-india-md\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10141761,"title":"How AI Dragons Set GenAI on Fire This Year","content":"If you thought the buzz around AI would die down in 2024, think again. Persistent progress in hardware and software is unlocking possibilities for GenAI, proving that 2023 was just the beginning. 2024 – the Year of the Dragon — marks an important shift as GenAI becomes deeply woven into the fabric of industries worldwide. Businesses no longer view GenAI as just an innovative tool. Instead, it is being welcomed as a fundamental element of their operational playbooks. CEOs and industry leaders, who recognise its potential, are now focused on seamlessly integrating these technologies into their key processes. This year, the landscape evolved rapidly and generative AI became increasingly indispensable, progressing from an emerging trend to a fundamental business practice. Scale and Diversity An important aspect is the growing understanding of how GenAI enables both increased volume and variety of applications, ideas and content. The overwhelming surge in AI-generated content is leading to consequences we are just starting to uncover. According to reports, over 15 billion images were generated by AI in one year alone – a volume that once took humans 150 years to achieve. This highlights the need for the internet post-2023 to be viewed through an entirely new lens. The rise of generative AI is reshaping expectations across industries, setting a new benchmark for innovation and efficiency. This moment represents a turning point where ignoring the technology is not just a lost opportunity, but could also mean falling behind competitors. “The top open source models are Chinese, and they are ahead because they focus on building, not debating AI risks,” said Daniel Jeffries, chief technology evangelist at Pachyderm. China’s success is underpinned by its focus on efficiency and resource optimisation. With limited access to advanced GPUs due to export restrictions, Chinese researchers have innovated ways to reduce computational demands and prioritise resource allocation. “When we only have 2,000 GPUs, the team figures out how to use it,” said Kai-Fu Lee, AI expert and CEO of 01.AI. “Necessity is the mother of innovation.” He further highlighted how his company transformed computational bottlenecks into memory-driven tasks, achieving inference costs as low as 10 cents per million tokens. “Our inference cost is one-thirtieth of what comparable models charge,” Lee further said. The rise of Chinese AI extends beyond its borders, with companies like MiniMax, ByteDance, Tencent, Alibaba, and Huawei targeting global markets. MiniMax’s Talkie AI app, for instance, has 11 million active users, half of whom are based in the US. At the Wuzhen Summit 2024, analysts noted that as many as 103 Chinese AI companies were expanding internationally, focusing on Southeast Asia, the Middle East, and Africa, where the barriers to entry were lower than the Western markets. ByteDance has launched consumer-focused AI tools like Gauth for education and Coze for interactive bot platforms, while Huawei’s Galaxy AI initiative supports digital transformation in North Africa. AI Video Models Models like Kling and Hailuo have outpaced Western competitors like Runway in speed and sophistication, which represents a shift in leadership in this emerging domain. This is reflected in advancements in multimodal AI, where models like LLaVA-o1 rival OpenAI’s vision-language models by using structured reasoning techniques that break down tasks into manageable stages. The Rugged Boundary In 2023, it became clear that generative AI is not just elevating industry standards, but also improving employee performance. According to a YouGov survey, 90% of workers agreed that AI boosts their productivity. Additionally, one in four respondents use AI daily, with 73% using it at least once a week. Another study revealed that when properly trained, employees were able to complete 12% of tasks 25% faster with the assistance of generative AI, while the overall quality of their work improved by 40%. The greatest improvements were seen among low-skilled workers. However, for tasks beyond AI’s capabilities, employees were 19% less likely to produce accurate solutions. This dual nature has led to what experts call the ‘jagged frontier’ of AI capabilities. On one side, AI now performs impressive abilities and tasks with remarkable accuracy and efficiency that were once deemed beyond machines’ reach. On the other hand, however, it struggles with tasks that require human intuition. These areas, defined by nuance, context, and complex decision-making, are where the binary logic of machines currently falls short. Cheaper AI As enterprises begin to explore the frontier of generative AI, we might see more AI projects take shape and become standard practice. This shift is driven by the decreasing cost of training LLMs, thanks to advancements in silicon optimisation, which is expected to halve every two years. Alongside growing demand and global shortages, the AI chip market is set to become more affordable in 2024, with new alternatives to industry leaders like NVIDIA emerging. Moreover, new fine-tuning techniques such as self-play fine-tuning are making it possible to strengthen LLMs without relying on additional human-defined data. These methods use synthetic data to develop better AI with fewer human interventions. Unveiling the ‘Modelverse’ The decreasing cost is enabling more companies to develop their own LLMs and highlighting a clear trend towards accelerating innovation in LLM-based applications in the next few years. By 2025, we will likely see the emergence of locally executed AI instead of cloud-based models. This shift is driven by hardware advances like Apple Silicon and the untapped potential of mobile device CPUs. In the business sector, SLMs will likely find greater adoption by large and mid-sized enterprises because of their ability to address niche requirements. As implied by their name, SLMs are more lightweight than LLMs. This makes them perfect for real-time applications and easy integration across various platforms. While LLMs are trained on massive, diverse datasets, SLMs concentrate on domain-specific data. In such cases, the data is often from within the enterprise. This makes SLMs tailored to industries or use cases, thereby ensuring both relevance and privacy. As AI technologies expand, so do concerns about cybersecurity and ethics. The rise of unsanctioned and unmanaged AI applications within organisations, also referred to as ‘Shadow AI’, poses challenges for security leaders in safeguarding against potential vulnerabilities. Predictions for 2025 suggest that AI will become mainstream, speeding up the adoption of cloud-based solutions across industries. This shift is expected to bring significant operational benefits, including improved risk assessment and enhanced decision-making capabilities. Organisations are encouraged to view AI as a collaborative partner rather than just a tool. By effectively training ‘AI dragons’ to understand their capabilities and integrating them into workflows, businesses can unlock new levels of productivity and innovation. The rise of AI dragons in 2024 represents a significant evolution in how AI is perceived and utilised. As organisations embrace these technologies, they must balance innovation with ethical considerations, ensuring that AI serves as a force for good.","excerpt":"Predictions for 2025 suggest that AI will become mainstream, speeding up the adoption of cloud-based solutions across industries.","categories":["Deep Tech"],"tags":["AI (Artificial Intelligence)","AI dragons","Editors Picks","Generative AI"],"author_name":"Tarunya S","publish_date":"2024-11-27T15:00:00","publication_year":"2024","word_count":1120,"keywords":["GenAI","Pachyderm","OpenAI","AI","ML","SLM","RAG","Editors Picks","generative AI","multimodal AI","Generative AI","AI dragons","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","ML","generative AI","GenAI","multimodal AI","OpenAI","Pachyderm","SLM","RAG","R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/how-ai-dragons-set-genai-on-fire-this-year\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10104394,"title":"It Takes More Than a ‘Better Product’ to Compete with Google Search","content":"Google pays nearly USD 18-20 billion a year to be Apple’s default search engine. The agreement is currently under scrutiny through an antitrust lawsuit filed by the US Justice Department. This also highlights Google’s dominance in the search domain, a hegemony that even Microsoft has struggled to challenge. But with generative AI coming into the picture, it looked like the dynamics could change with a number of startups popping up and offering better generative AI-powered search experiences like You.com, Perplexity AI and Neeva AI. Notably, Neeva AI caught everyone’s attention because it was co-founded by a former Google veteran, Sridhar Ramaswamy, who led Google’s USD 115 billion advertising tech division. But despite having a better product than Google, Neeva AI exited the market, and sold to Snowflake. However, others in the space, like You and Perplexity, remain optimistic that they can carve out a share of the pie in the search engine space. End of the ten blue links Banking on the strength of Large Language Models (LLMs), both You and Perplexity are currently focussed on providing a better search experience than Google to their users. Much like Neeva, these startups aim to disrupt the paradigm of the ’10 blue links’ that typically appear when querying Google. The goal is to present the answer directly, aided by AI, rather than requiring users to sift through links to locate the relevant information. This has also resulted in both startups attracting a significant number of users to their applications since their inception. You, for example, which was founded by former Salesforce chief scientist Richard Socher, reported 1 million users in 2022. In September 2023, the company had 13.6 million visits, according to data from Similarweb. Similarly, Perplexity had 37.4 million hit during the same period. The numbers may seem relatively minute in terms of the billions of hits Google gets every month, but Aravind Srinivas, who heads Perplexity, recently said that the company has sustained users. However, recently Google announced Gemini, which could possibly be the most advanced LLM to date, surpassing OpenAI’s GPT-4. Google introduced generative AI into their search engine earlier but it remained problematic. For example, on the day of the ICC Cricket World Cup Final played between India and Australia, we at AIM asked Bard about the match and to our surprise, it declared Australia the winner even before the match was played. In a separate instance, it declared India as the winner. With Gemini, how much can Google improve the search experience remains to be seen, but the implications could be significant for startups operating in the same space. If Google significantly improves its search experience, users of You and Perplexity could possibly migrate back to Google. Alternatively, in such a scenario, the startups could integrate Gemini into their products to improve their current offerings, however, the challenges in front of them are aplenty. But to compete with Google, you need more than a product For instance, to sustain and compete with Google in the search engine space, they need more than a good product. Ramaswamy, in a previous interaction with AIM, said that his company was not on the path to a sustainable valuation, despite having a better product than Google’s. Moreover, Google’s overwhelming presence stifles competition. “To become the default search engine in Safari, as it turns out, is an incredibly convoluted endeavour. In reality, there is no formal process; it all hinges on Cupertino’s (Apple’s) subjective judgement of one’s qualifications. “There is no process and it’s pretty tough to create a sustainable business,” he said. As it appears, both Perplexity and You are currently fixated on a better product and don’t have a viable business model. Srinivas, in an interview, said that he does not plan to monetise the same way like Google. “We haven’t thought through that part yet, but we at least know that it won’t be the exact same thing,” Srinivas said in an interview. While Perplexity is relatively new, founded in 2022, You has been around since 2020. Socher said he is not much intimated by the fact that Neeva exited the market. But he does acknowledge that, from focussing on product development, the company must now shift its focus to revenue generation. For now, the startups may carve out a niche in the market, but building a sustainable business, especially in the challenging search space, remains a different ball game. So far, Perplexity has raised over USD 50 million with a valuation of over USD 500 million, generating USD 3 million in annual recurring revenue. You have also raised USD 45 million in three funding rounds and have estimated revenue of around USD 15 million per year. Currently, both companies offer a subscription for the premium services. YouPro costs USD 14.99\/month or USD 149.99 for a full year which gives users unlimited access to AI models like GPT-4 and Stable Diffusion XL. Whereas, Perplexity premium costs USD 20\/ month or USD 200 per year and gives users access to advanced AI models, extra copilot usage, dedicated support, and unlimited file uploads. However, the startups will soon need to explore more effective monetisation strategies, foster growth, and simultaneously compete with Google.  Neeva, too, had a subscription model starting at USD 4.95\/ month. Ramaswamy said that it was very possible for them at Neeva to raise again in a fresh round and get a few million more to use their products. However, he said a founder’s job was to look ahead around the corner, and unfortunately, Neeva found operating in an uncompetitive space too daunting. Nonetheless, despite recognising the David vs. Goliath nature of the situation, both Socher and Srinivas maintain optimism about building a sustainable business.","excerpt":"For startups, search is the most difficult space to disrupt","categories":["AI Trends"],"tags":["Aravind Srinivas","neeva ai","Perplexity AI"],"author_name":"Pritam Bordoloi","publish_date":"2023-12-07T18:05:16","publication_year":"2023","word_count":946,"keywords":["neeva ai","Go","OpenAI","AI","AWS","Perplexity AI","GPT","Aim","generative AI","Rust","R","Aravind Srinivas","Snowflake"],"extracted_tech_keywords":["AI","generative AI","OpenAI","Aim","AWS","Snowflake","R","Go","Rust","GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/it-takes-more-than-a-better-product-to-compete-with-google-search\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10020669,"title":"Microsoft’s Platform For Building Cloud-Native Applications Is Now Production Ready","content":"Microsoft’s Distributed Application Runtime (Dapr) has now graduated to the production-ready version, 16 months after it was released as an alpha project. Dapr is an open-source, Microsoft-incubated project for developers to build event-driven and distributed cloud-native applications. Why Dapr Dapr was launched in October 2019 and has released 14 major updates since. It can integrate with all major cloud providers such as Azure, AWS, and Google Cloud. Credit: Dapr The last few years have witnessed a steady increase in the number of organisations moving from on-premises to cloud and microservices-based software architectures. Applications built on function-based components which are smaller and connect through APIs are easier to develop and maintain. In such cases, a single developer takes responsibility for one service, also called microservice. This is where platforms like Dapr come into play. Dapr acts as a self-hosted environment on a local machine. It is a portable and event-driven runtime environment that helps developers build microservices-based applications on the cloud. Building Block Dapr consists of a set of building blocks accessible by standard HTTP or gRPC APIs, which can be called using any programming language. “These building blocks empower all developers with proven, industry best practices, and each building block is independent; you can use one, some, or all of them in your applications,” as per a Microsoft blog post. Dapr contains language-specific SDKs for Go, Java, .NET, Python, Javascript and enables developers to write both stateful and stateless functions in their chosen language. Since the SDKs share the Dapr runtime, developers can also get a cross-language actor and function support. Dapr allows developers to build applications for both cloud and edge, without any code changes. With Dapr, applications can be run locally — on Kubernetes or any other hosting environment. Dapr’s building blocks include: Service-to-service invocation: It enables the communication between applications through endpoints in the form of HTTP or gRPC messages. These endpoints act as a combination of reverse proxy with built-in service discovery and leverage built-in distributed tracing and error handling. State management: Dapr provides a key\/value-based state API with pluggable state stores using which an application can preserve a state beyond a single session. Publish and subscribe: A loosely coupled messaging service between senders and subscribers. Dapr supports the publishing and subscribing pattern between applications. Resource binding: Binding facilitates a bi-directional connection to cloud or on-premise service. With Dapr, users can invoke the external service using binding API. It allows the application to be triggered by events sent by the cloud or on-premise service. Actors: The actor is an independent entity of computing and state which has single-threaded execution. Based on the Virtual Actor pattern, Dapr provides actor implementation on a single-threaded programming model. Actors are collected as garbage when not in use. Observability: Dapr system components monitor services, components, and user applications. Secrets: Dapr offers a secrets building block API and integrates it with Azure Key Vault, Kubernetes and other secret stores. Secrets API can be called using the service code to retrieve these ‘secrets’ from the stores. Wrapping Up With v1.0, Microsoft is focused on Kubernetes for running production-ready applications. The scope of Dapr will further increase and diversify into virtual machines, platform-as-a-service (PaaS), and edge environments in the future. Traditionally, a developer is expected to do client, server, and database-type applications, containerise them and create scalable microservices . Additionally, these apps are also expected to be portable across environments, adding an extra layer of complexity. “The goal with Dapr was: let’s take care of all of the mundane work of writing one of these cloud-native distributed, highly available, scalable, secure cloud services, away from the developers so they can focus on their code,” Microsoft Azure CTO Mark Russinovich said in an interview.","excerpt":"Microsoft’s Distributed Application Runtime (Dapr) has now graduated to the production-ready version, 16 months after it was released as an alpha project. Dapr is an open-source, Microsoft-incubated project for developers to build event-driven and distributed cloud-native applications. Why Dapr Dapr was launched in October 2019 and has released 14 major updates since. It can integrate […]","categories":["Global Tech"],"tags":[],"author_name":"Shraddha Goled","publish_date":"2021-02-23T13:00:00","publication_year":"2021","word_count":619,"keywords":["Go","AWS","AI","R","RAG","microservices","Python","JavaScript","Azure","kubernetes"],"extracted_tech_keywords":["AI","RAG","AWS","Azure","kubernetes","microservices","Python","R","JavaScript","Go"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/microsofts-platform-for-building-cloud-native-applications-is-now-production-ready\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":54952,"title":"Can We Achieve Human-Like Conversational Bots","content":"Google has announced a highly specialised chatbot that outperformed every other chatbot in the market. AI chatbots have become mandatory for firms who are striving to provide superior customer services by solving their queries without human touch. This not only helps firms to expedite the process of mitigating problems but also reduces the operational costs. However, these chatbots are very specific on what they can perform, thereby leaving many issues unheard. Consequently, firms like Google, Amazon, among others, are trying to find ways to enhance the capabilities of bots through new techniques of natural language processing. Announcement On Tuesday, Google introduced Meena, a human-like conversational bot, that delivered exceptional results when compared with other chatbots. The idea behind this is to widen the use case of AI bots in the real-world. Meena is a 2.6 billion parameter end-to-end trained neural conversational model that can converse sensibly. Often chatbot replies out of context, which then annoys humans, resulting in useless conversation. Besides, most of the time, AI agents say that they do not know the answer to the question asked. However, Meena remembers the context of the conversation without getting deviated to enable an engaging discussion. Force Behind Meena Google, in its blog post, mentioned that Meena is able to bring more human attributes by reducing the perplexity. This minimises the uncertainty of predicting the next word in a conversation. This is powered by the Evolved Transformer seq2seq architecture, which was discovered by evolutionary neural architecture search to improve perplexity. Meena is equipped with a single Evolved Transformer encoder block and 13 Evolved Transformer decoder blocks to respond while remembering the context. While the encoder plays a crucial role in what has already been said in the conversation, decoder devices an actual response using the information from encoder. Surprisingly, researchers figure out a superior decoder was essential to accomplish quality conversation. Unlike the traditional approach, where training is organised as tree threads — each reply is considered as one conversational turn — researchers used training with seven turns of context as one path through a tree thread. This bridged the gap and empowered the AI models to respond with context. As Meena trained on 341 GB of text that was extracted from the public domain and social media conversations, it is 1.7x higher capacity and was trained on 8.5x more data when compared to OpenAI GPT-2. What It Means To The Landscape The greatest difficulty for researchers is to bring commonsense into AI agents, as without naive intelligence accomplishing high-quality conversation is near impossible. Amazon with Alexa has achieved its initial target of continuous 5 minutes of communication, but its 20 minutes target is far from reach in the near future. “We are very early, I think we are at least five to eight years away before we can meet the ultimate goal of having a real, coherent and engaging conversation for 20 minutes with any of these bots,” said Rohit Prasad, head scientist for Alexa’s Artificial Intelligence Team. However, Google has now made a crucial step ahead with Meena in achieving human-like conversational bots. Meena was tested against other state-of-the-art chatbots. Google’s Meena got an SSA score of 79%, compared to 56% for Mitsuko. Mitsuko for the last four years has been winning the Loebner Prize. But, Google hasn’t released it for a public demo, citing potential misuse. Thus, one will have to wait until Google evaluates Meena and considers it safe for the open trail. Outlook If Google’s claim is proven to be accurate, it could revolutionise the way chatbots are used in day-to-day life. However, Meena needs to be tested in a setting that Alexa is put to, a long conversational, it is difficult to understand how much of the capability has increased for the chatbots. Undoubtedly, the SSA scores have been impressive, but Meena has to tested in real-world use cases to make any linear conclusion of its superiority.","excerpt":"Google has announced a highly specialised chatbot that outperformed every other chatbot in the market. AI chatbots have become mandatory for firms who are striving to provide superior customer services by solving their queries without human touch. This not only helps firms to expedite the process of mitigating problems but also reduces the operational costs. […]","categories":["AI Features"],"tags":[],"author_name":"Rohit Yadav","publish_date":"2020-01-31T12:00:00","publication_year":"2020","word_count":652,"keywords":["Go","artificial intelligence","OpenAI","AI","chatbots","GPT","Aim","AI agents","GAN","R"],"extracted_tech_keywords":["AI","artificial intelligence","OpenAI","Aim","chatbots","R","Go","GPT","GAN","AI agents"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/can-we-achieve-human-like-conversational-bots\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":9831,"title":"Course Review: Business Analytics Certificate Program by Great Lakes Institute of Management","content":"Since we get a lot of queries around, which business analytics course or university is best by many of our readers, we decided to review one of the most sought after courses in the industry – the Business Analytics Certificate Program by Great Lakes Institute of Management. We have consistently ranked Great Lakes as the leading school in India providing analytics education. Recently, Great lakes ventured into the heated up online learning space through the partner organization – Great Learning. Great Learning is an online and blended learning platform designed to empower working professionals to develop relevant competencies and accelerate their career progression. Business Analytics Certificate Program is amongst the first programmes to be launched on this platform. We first spoke about the same at its launch here. Participants are expected to take roughly 6 months to complete this Program, which involves 160 hours of learning comprised of about 100 hours of self-learning, 40 hours of instructor led webinars and another 20 hours of assessments. Now, this program sparks our interest for some very particular reasons. Firstly, it has a biggie like Great Lakes backing it, which has already proved its mettle in the Indian analytics education arena. Today, there are more analytics professionals being trained by Great Lakes than all other B-schools combined in India. Secondly, not many universities provide online analytics courses, further, given the online nature, this course does not even have a location constraint. So, this course can be taken up by anyone worldwide. However, the real story has to be quality and experience. So, we thought of looking at this course a little deeper and provide our perspective on it. First things first, the course is being taught by B-school faculty, which is rare in online learning. We found that the course content has been tailored by expert faculty and practitioners to address the current industry needs. Thus, candidates develop the ability to process data problems using analytics and relate them to the business aspects. Though one can find B-schools in US taking up analytics courses on ‘Coursera’, what Great Lakes have done is very unique from India perspective. Besides optimizing their academic brand and faculty strength of their online channel, Great Lakes has an online “Learning Management System” that hosts all the content, including online lectures, webinar recordings, reading material and assignments. But, what gives it a plus point is the fact, this material is accessible to the students for the next 3 years, even after the course completion. Further, participants can also access the LMS using the Great Learning mobile app. Moreover, the program is a combination of self-paced and live virtual online classes. The classes are interactive and are conducted by analytics experts in a virtual classroom. These lectures are recorded, this enables the participants to review the material or to make up for any missed Q & A sessions. Further, ‘Live Webinar Q&A Sessions’ are conducted by analytics mentors. This allows the participants to resolve any questions or doubts from the Instructor-led sessions. Additionally, there are ‘Pre-recorded video lectures’ as well, that can be viewed at any time and as many times as the participant wants. Surely, these are noticeable differentiators vs other online courses on analytics. Participants: Since, this certificate course fits in with the industry requirements for working professionals (to hone their analytic skills) as well as for fresher’s (to build their career in analytics and ace their job interviews); typically, a lot of working professionals from varied backgrounds such as IT, Management, Entrepreneurs, Engineers, etc. with a few years of work experience join this course. Curriculum: Business Analytics Certificate Program has a well-rounded curriculum. The students are given hands on experience by working on analytical tools such as SAS language, R and Advanced Excel. Moreover, they are exposed to various industry domains such as Financial Analytics, Retail Analytics, Web and Social Media Analytics which helps them decide their career path further. The program begins with an Introduction to Analytics then moves on to cover topics such as- Basic Analytics Tools, Fundamentals of Statistics for Business Analytics, Advanced Excel Analysis, Data Mining using Decision Trees, Data Mining using Clustering in SAS and R, Time Series Forecasting, Predictive Modelling – Logistic Regression using SAS, Predictive Modelling – Logistic Regression using R and lastly, the Applications of analytics. Thus, overall, it is a well-structured program and we believe that it maximizes the learning curve of the candidates in the industrial context. For more information, please visit: http:\/\/bit.ly\/1ST3Ofx","excerpt":"Since we get a lot of queries around, which business analytics course or university is best by many of our readers, we decided to review one of the most sought after courses in the industry – the Business Analytics Certificate Program by Great Lakes Institute of Management. We have consistently ranked Great Lakes as the […]","categories":["AI Trends"],"tags":[],"author_name":"Дарья","publish_date":"2016-05-07T05:43:07","publication_year":"2016","word_count":746,"keywords":["programming_languages:R","AI","data_tools:Spark","analytics","GAN","R"],"extracted_tech_keywords":["AI","analytics","R","GAN","programming_languages:R","data_tools:Spark"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/course-review-business-analytics-certificate-program-great-lakes-institute-management\/","complexity_score":2,"technical_depth":6,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":40472,"title":"A Look At How Twitter Handles Its Time Series Data Ingestion Challenges","content":"The components of time-series are as complex and sophisticated as the data itself. With increasing time, the data obtained increases and it doesn’t always mean that more data means more information but, larger sample avoids the error that   due to random sampling. For social media platforms, the data handling chores get worse with their increasing popularity. Social media outlets like Twitter have imbibed many techniques over the years to offer uninterrupted services. In other words, handling tonnes of data, curating the data, forecasting the surge in data, gauging the time series with inbuilt metrics and then storing these metrics to make the whole thing more robust. According to Twitter’s software engineering team, the networking giant stores 1.5 petabytes of logical time series data, and handles 25K query requests per minute. The scale at which these firms operate requires customised in-built techniques. Twitter has done the same to solve their database challenges with MetricsDB. There is a dedicated team for observability engineering at twitter which supports other teams which monitor service health and issues with distributed system. What Does MetricsDB Offer Source: Twitter Engineering blog Previously, Manhattan was the go-to solution for Twitter. But this was not feasible as it required to use multiple datasets for every zone. MetricsDB, on the other hand, is multi-zone compliant. The main objective of having these services is to have better monitoring, visualization, infrastructure tracing and log aggregation\/analytics. For storing mappings from partitions to servers, MetricsDB’s cluster manager uses HDFS. The Hadoop Distributed File System (HDFS) is a distributed file system designed to run on commodity hardware. Applications that run on HDFS have large data sets. A typical file in HDFS is gigabytes to terabytes in size. Thus, HDFS is tuned to support large files. It provides high aggregate data bandwidth and supports tens of millions of files in a single instance. The partitions in the metrics database have a cluster manager of their own. The back end servers get updates from these cluster managers. These backend servers are responsible for processing metrics for a small number of partitions. These servers store latest two hours of data and the older data gets cached. For checking this every two hours, Twitter uses Blobstore, durable storage that enables lower management head. In the context of Twitter, a request can be something as important as user checking missing data. These missing data alerts would fire if the replica that responds first doesn’t have the data for the more recent minute. This is one way of knowing that the services are interrupted and inconsistencies have to be attended to. So whenever a query pops up, the custom partition scheme uses consistent hashing on zone, service and source to route request to the specific logical backend. This reduces the number of individual metric requests per minute from 5 billion to under 10 million. Partitioning scheme via Twitter Few challenges still persist. For instance, a write request from a collection agent needs to be split into multiple requests based on a partitioning scheme. To address this, engineers have used Apache Kafka, which helped reduce the number of requests to the queue and storage. Key Takeaways Using a custom storage backend instead of a traditional key-value store reduced the overall cost by a factor of 10. Ninety-three per cent of timestamps can be stored in 1 bit and almost 70% of metric values can be stored in 1 bit. Reduced latency by a factor of five. Improved responsiveness significantly while reducing the load. The success that Twitter enjoys owes in large to its pursuit of freedom from physical enterprise vendors right from the beginning. They have continually engineered and refreshed their infrastructure by taking advantage of the latest open standards in technology. Currently, enterprises are struggling to deploy machine learning models at full scale. Common problems include- talent searching, team building, data collection and model selection to say few. To tap the most out of the latest technologies, it is necessary to build service-specific tools and frameworks in addition to the existing models and the success of Twitter verifies the same.","excerpt":"The components of time-series are as complex and sophisticated as the data itself. With increasing time, the data obtained increases and it doesn’t always mean that more data means more information but, larger sample avoids the error that   due to random sampling. For social media platforms, the data handling chores get worse with their increasing […]","categories":["Deep Tech"],"tags":["Data Engineering","Time Series","Twitter (X)"],"author_name":"Ram Sagar","publish_date":"2019-06-09T08:13:52","publication_year":"2019","word_count":679,"keywords":["Go","machine learning","programming_languages:R","AI","data_tools:Kafka","programming_languages:Go","RAG","Time Series","Data Engineering","analytics","Kafka","Twitter (X)","R"],"extracted_tech_keywords":["AI","machine learning","analytics","RAG","Kafka","R","Go","programming_languages:R","programming_languages:Go","data_tools:Kafka"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/a-look-at-how-twitter-handles-its-time-series-data-ingestion-challenges\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10143677,"title":"AIIMS and Wipro GE Healthcare Establish AI Health Innovations Hub","content":"All India Institute of Medical Sciences (AIIMS) has partnered with Wipro GE Healthcare Pvt Ltd to advance AI in the healthcare sector. As the clinical partner, AIIMS will provide multi-modal clinical inputs and serve as a real-world environment for evaluating, offering feedback, and deploying GE Healthcare’s AI-enabled solutions. Wipro GE Healthcare, the technology partner in this collaboration, is set to invest $1 million in the initiative. The investment will focus on developing intelligent systems and workflow solutions that address key clinical areas, including cardiology, oncology, and neurology. “This partnership allows us to leverage AI to improve patient outcomes and streamline clinical workflows,” said a representative from Wipro GE Healthcare. AIIMS’s role ensures that the AI solutions are effectively tested and tailored to meet real-world medical needs. The joint effort aims to enhance healthcare delivery through precise diagnoses, innovative treatment protocols, and real-time patient data tracking, ultimately transforming patient care and operational efficiency in the medical field. A joint working committee from both organisations will oversee the collaboration, focusing on clinical research and academic engagement. “This partnership with Wipro GE Healthcare holds strategic value and is aligned with the national vision of Viksit Bharat through advanced healthcare,” said Dr M Srinivas, director of AIIMS. In this regard, Parminder Bhatia, chief AI officer at GE HealthCare, also stated that the Indian healthcare industry has experienced rapid growth, driven by supportive government policies, innovative domestic solutions, and the integration of digital technologies and AI. He added that GE HealthCare views AI not merely as a tool but as a transformative force to democratise healthcare and enable increased predictiveness, prevention, and precision. Chaitanya Sarawate, MD of Wipro GE Healthcare South Asia, emphasised, “Wipro GE Healthcare has a proud legacy of innovation in MedTech. The future of healthcare in India will be driven by technology, with AI at the heart of innovation to enable predictive, personalised, and preventive care at scale. With its transformative vision, diverse patient pool, and scale, AIIMS has been at the forefront of redefining care. This collaboration is a strong step forward in developing foundation models for transformative clinical care and applications for India and the world.” According to The National AI Portal of India (INDIAai), AI technology promises a transformative leap for India’s healthcare system by 2025, potentially boosting the GDP by $25-30 billion. By connecting longitudinal data—such as imaging, lab results, and medical records—and integrating it with analytics and AI, the collaboration aims to significantly enhance diagnosis accuracy, streamline operations, and improve patient care nationwide.The AI Health Innovations Hub is set to become a cornerstone for technological advancements in healthcare harnessing AI.","excerpt":"A joint working committee from both organisations will oversee the collaboration, focusing on clinical research and academic engagement.","categories":["AI News"],"tags":["AI Healthcare","Healthcare Automation","Wipro"],"author_name":"Shalini Mondal","publish_date":"2024-12-16T19:35:54","publication_year":"2024","word_count":433,"keywords":["Wipro","Go","API","AI","AI Healthcare","ML","Git","RAG","Healthcare Automation","Aim","analytics","foundation models","R"],"extracted_tech_keywords":["AI","ML","analytics","foundation models","Aim","RAG","R","Go","Git","API"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/aiims-and-wipro-ge-healthcare-launched-1m-ai-health-innovations-hub\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":67571,"title":"Drishti Raises $25M In Series B Funding To Enhance Its Computer Vision Solution For The Manufacturing Industry","content":"Drishti Technologies, a California-based startup, has announced that it has raised $25M in Series B funding. The investment was led by Sozo Ventures and was joined by other companies like Andreessen Horowitz, Toyota AI Ventures, Emergence Capital, and Alpha Intelligence Capital. Drishti leverages computer vision technologies to assist manufacturing companies to streamline their workflows. Founded by Akella, the firm strives to build solutions with cutting-edge technologies for production companies, which can be utilised for defect reduction, process monitoring and optimisation, remote visibility and collaboration, among others. In a mission to automate the process in factories, Drishti’s solutions not only mould data for monitoring but also generate information by labelling human actions using deep learning and computer vision technologies. In addition, Drishti also offers solutions to monitor the social distancing at manufacturing sites during the COVID-19. Equipped with robust solutions, Drishti has been named as a 2019 technology pioneer by the World Economic Forum and cool vendor in manufacturing operations by Gartner. Discrete manufacturers such as DENSO, Flex, Ford, Nissan and others who embrace lean production methods are working with Drishti to gain efficiencies and infuse resiliency into their workforces and supply chains. Drishti’s impact is felt across the entire plant ecosystem — from line associates to indirect labour to the front office. Over the years, Drishti has quickly gained traction as the deployment of the latest technology in the manufacturing industry is greatly untouched. Undoubtedly, some companies offer AI-based solutions to production firms to predict the breakdowns, notify in case of hazardous chemical spills and more, but nothing matches the features and functionality of Drishti’s solutions. The company’s solutions stand out from the rest due to its ability to generate labelled data, thereby collecting numerous variables that assist it in further enhancing its solutions. “Most manufacturers lack meaningful data about manual assembly processes because human actions are very difficult to measure. Drishti uses computer vision to create continuous streams of data from video of manual actions,” said Jim Adler, managing director at Toyota AI Ventures.","excerpt":"Drishti Technologies, a California-based startup, has announced that it has raised $25M in Series B funding. The investment was led by Sozo Ventures and was joined by other companies like Andreessen Horowitz, Toyota AI Ventures, Emergence Capital, and Alpha Intelligence Capital. Drishti leverages computer vision technologies to assist manufacturing companies to streamline their workflows. Founded […]","categories":["AI News"],"tags":["human touch to ai"],"author_name":"Rohit Yadav","publish_date":"2020-06-17T19:00:01","publication_year":"2020","word_count":335,"keywords":["API","funding","ELT","AI","ML","computer vision","RAG","deep learning","human touch to ai","R","startup"],"extracted_tech_keywords":["AI","ML","deep learning","computer vision","RAG","R","API","ELT","startup","funding"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/drishti-raises-25m-in-series-b-funding-to-enhance-its-computer-vision-solution-for-the-manufacturing-industry\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":24361,"title":"Race for Artificial Intelligence chips","content":"Once just a figment of the imagination of some our science fiction writers, artificial intelligence (AI) is taking root in our everyday lives. We’re still a few years away from having robots at our beck and call, but AI has already had a profound impact in more subtle ways. Weather forecasts, email spam filtering, Google’s search predictions, and voice recognition, such Apple’s Siri, are all examples. What these technologies have in common are machine-learning algorithms that enable them to react and respond in real time. AI workloads are different from the calculations most of our current computers are built to perform. AI implies prediction, inference, and intuition. But the most creative machine learning algorithms are hamstrung by machines that can’t harness their power. Hence, if we’re to make great strides in AI, our hardware must change, too. End of Moore’s Law Moore’s Law, named after Intel co-founder Gordon Moore, states that the number of transistors that can be placed on an integrated circuit doubles roughly every two years. For decades, chipmakers have succeeded in shrinking chip geometries, allowing Moore’s Law to remain on track and consumers to get their hands on ever more powerful laptops, tablets, and smartphones. But in next few years “transistors could get to a point where they could shrink no further.” While it will still be technically possible to make smaller chips, they will reach “the economic minimum” at which the costs will be too high to justify. If you’re following the AI journey so far, you’ll see that we’ve sprinted out ahead using CNNs (Convolutional Neural Networks) and RNNs (Recurrent Neural Networks) but that progress beyond these applications is only now emerging.  The next wave of progress will come from Generative Adversarial Nets (GANs) and Reinforcement Learning, with some help thrown in from Question Answering Machines (QAMs) like Watson. What we see shaping up is a three-way race for the future of AI based on completely different technologies.  Those are: High Performance Computing (HPC) Neuromorphic Computing (NC) Quantum Computing (QC). Neuromorphic and quantum computing always seemed that they were years away.  The fact is however that there are commercial neuromorphic chips and also quantum computers in use today in operational machine learning roles. High Performance Computing The path that everyone has been paying most attention to is high performance computing.  Stick to the Deep Neural Net architectures that we know, just make them faster and easier to access. While Intel, Nvidia, and other traditional chip makers were rushing to capitalize on the new demand for GPUs, others like Google and Microsoft are busy developing proprietary chips of their own that make their own deep learning platforms a little faster or a little more desirable than others. Google came up with TensorFlow as its powerful, general purpose solution combined with their newly announced proprietary chips, the TPU (Tensor Processing Unit). Microsoft has been touting its use of non-proprietary FPGAs and just released an upgrade of its Cognitive Toolkit (CNTK). Neuromorphic Computing A new approach called neuromorphic computing seems to be picking up in last one decade. It seeks to leverage the brain’s strengths by using an architecture in which chips act like neurons. When the pulses or ‘spikes’ sent to a neuron reach a certain activation level, it sends a signal over the synapses to other neurons. Much of the action, however, happens in the synapses, which are ‘plastic,’ meaning that they can learn from these changes and store this new information. Unlike a conventional system with separate compute and memory, neuromorphic chips have lots of memory located very close to the compute engines. The new brain-inspired architecture will enable machines to do things that silicon chips can’t. Traditional chips are good at making precise calculations on any problem that can be expressed in numbers. A neuromorphic system can identify patterns in visual or auditory data, and adjust its predictions based on what it learns. A research paper by Intel scientist Charles Augustine predicts that neuromorphic chips will be able to handle artificial intelligence tasks such as cognitive computing, adaptive artificial intelligence, sensing data, and associate memory. They will also use 15-300 times less energy than the best CMOS chips use. That’s significant because today’s AI services, such as Siri and Alexa, depend on cloud-based computing in order to perform such feats as responding to a spoken question or command. Smartphones run on chips that simply don’t have the computing power to use the algorithms needed for AI, and even if they did they would instantly drain the phone’s battery. Companies such as Intel, IBM, and Qualcomm are now involved in a high-stakes race to develop the first neuromorphic computer. Neuromorphic Chips available in the market are IBM TrueNorth and Intel Loihi. Quantum Computing What really sets a quantum computer apart from a regular digital computer is the fundamental nature of how data is encoded via quantum properties like superposition or entanglement. A digital bit is either 0 or 1, but a quantum bit (or qubit) can be 0, 1 or a superposition of both states. Quantum computers are an area of huge interest because, if they can be built at a large enough scale, they could rapidly solve problems that cannot be handled by traditional computers. Quantum supremacy is a key milestone on the journey towards quantum computing. The idea is that if a quantum processor can be operated with low enough error rates, it could outperform a classical supercomputer on a well-defined computer science problem. That’s why the biggest names in tech are racing ahead with quantum computing projects. Year 2018 Intel announced its own 49-qubit quantum chip, code-named Tangle Lake. Post that Google’s Quantum AI Lab has shown off a new 72-qubit quantum processor called ‘Bristlecone‘. Both Neuromorphic and Quantum computing are laying out competitive roadmaps for getting to deep learning and even newer versions of artificial intelligence faster and perhaps easier. Only time will tell which will get better than the other.","excerpt":"Once just a figment of the imagination of some our science fiction writers, artificial intelligence (AI) is taking root in our everyday lives. We’re still a few years away from having robots at our beck and call, but AI has already had a profound impact in more subtle ways. Weather forecasts, email spam filtering, Google’s […]","categories":["IT Services"],"tags":["quantum supremacy"],"author_name":"Nitin Srivastava","publish_date":"2018-05-08T08:44:29","publication_year":"2018","word_count":990,"keywords":["quantum supremacy","machine learning","artificial intelligence","TPU","AI","neural network","neuromorphic computing","RAG","deep learning","TensorFlow","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","deep learning","neural network","TensorFlow","neuromorphic computing","RAG","TPU","R"],"url":"https:\/\/analyticsindiamag.com\/it-services\/race-for-artificial-intelligence-chips\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10071075,"title":"The Real Reason Robot Rights Is So Contentious","content":"When the Chief ordered his subordinates to take Johnny Five to ‘stolen goods’, he protested in contempt, “I’m not stolen goods”. While getting arrested, he questions in despair, “But hath not a robot eyes? Hath not a robot hands, organs, dimensions, senses, affections, passions? If you prick us, do we not bleed?” When we discuss robots and the possibility of them becoming ‘conscious’, we reach a profound philosophical difficulty. As seen in this iconic scene from ‘Short Circuit 2’ (1988). Should these “beings” have any claim to moral and legal standing? Much of the philosophy of rights is ill-equipped to deal with the case of artificial intelligence. Most claims for rights, whether human or animal, are focused on the question of consciousness. Unfortunately, there is no concrete understanding of what defines ‘consciousness’. Some believe that it’s immaterial and yet others believe that it’s a state of matter, like gas or liquid. Regardless of the precise definition, humans have an intuitive knowledge of consciousness because we experience it. Consciousness entitles beings to have rights owing to the awareness of emotions, like pain and suffering, in response to an external stimulus which is tied to the contemporary understanding of the idea of consciousness. However, robots don’t ‘suffer’, at least not in the way a conscious being would. Without the experience of pain or pleasure, there are no preferences which consequently renders the notion of rights meaningless. Human rights are deeply symbiotic to our own conscience. For instance, we dislike pain because our brains are evolved to keep us alive and well. As a result of that universal consciousness, we came up with rights that protect us from infringements that cause us pain. Even more abstract rights like ‘freedom’ or ‘equality’ are rooted in the way our brains are wired to detect what is ‘fair’ and ‘unfair’. But what if we programmed robots to experience pain and emotions? To be able to choose justice over injustice, pleasure over pain while also being conscious of making these choices? Would that make them ‘human’? More importantly, would that be the sole criterion that grants them rights? This notion sheds light on the need to define another abstract concept—robots. What is a robot? AI and Robot Ethics scholar Dr David J. Gunkel believes that our contemporary understanding of robots arises from fiction and not scientific facts, calling it ‘science fiction prototyping’. “Unlike artificial intelligence, which is the product of an academic conference in the mid-1950s, the word robot actually is the product of science fiction. It comes to us from Karel Čapek in his 1920 hit play, R.U.R., or Rossum’s Universal Robots. And he used the word robot, which is derived from the Czech word “robota,” or forced labour. And since this time you can see robots have dominated science fiction. They’re all over the place from Star Trek to Star Wars to Westworld”Dr David J. Gunkel, scholar in AI and robot ethics. He further emphasises that the main advantage of science fiction prototyping is that it allows non-experts to understand what is in play, interpret what this technology entails and how they can attempt to grapple with some of the major questions that need necessary probing and resolution. However, such representations do undermine the developmental efforts of engineers, AI scientists and roboticists who are constantly struggling against fictitious expectations that don’t measure up to real-world research. It’s a double-edged sword. George A. Bekey, American roboticist and Professor Emeritus of Computer Science, Electrical Engineering and Biomedical Engineering, University of Southern California, defines a robot as a device that senses, thinks and acts. But, that’s a rather broad definition—almost too broad because lots of technologies could then be considered robots. “Given the versatility and wide availability of robots for a wide variety of applications in today’s world, it will be difficult to nail down one single definition for a robot”Dr Karthik Ramesh, VP–Head International Markets and Innovation at Emids. He believes that although without human-like consciousness, robots can help achieve tasks that prove beyond human capacity. An excellent contemporary example is the James Webb Telescope orbiting deep space to explore other planets and galaxies for proof of life. However, when considered from a technical point of view—robots are merely the sum of their parts. Such a vast spectrum of definitions allows anything from a thermostat to a smartphone to Tesla’s upcoming humanoid to be deemed as a robot. But logically, they are all very different devices, serving a range of purposes. Gunkel believes that the understanding of the term ‘robot’ is going to change as the context around it evolves and more importantly, as our own experiences with technology transform in time. “It is essentially one of those terms that are pregnant with ambiguity, but I think that offers us the opportunity to get more specific and to talk about things in a much more context” Dr David J. Gunkel, scholar in AI and Robot Ethics Anthropomorphizing robots Regardless of how we define robots, our assumptions about them take roots in Tinseltown and then evolve owing to our innate ability to anthropomorphize. For instance, when former Google engineer, Blake Lemoine published his conversations with Google’s LaMDA and claimed that the AI was sentient on Twitter, the internet erupted with a range of reactions. We witnessed a very similar reaction to Sophia, a social humanoid robot, when she remarked about destroying humans at the 2016 SFX conference. “As humans, we tend to anthropomorphize. So the question we need to ask is, is the behaviour that we see truly intelligent? AI can fool some of the people all of the time and all of the people some of the time, but that does not make it sentient or intelligent”Dr Oren Etzioni, CEO, AI2. Read also: Paul Allen liked the fact that I wasn’t an academic: Dr Oren Etzioni, CEO, AI2 In contrast, Dr Ramesh concurs with Alan Turing and Barrington Bayley who believed that the understanding of consciousness could be stretched to include inanimate objects and that a difference in the nature of consciousness solely can not account for the exclusion of robots. “Cultures like [the] Japanese have already imbibed reverence for robots as human equivalents such as ‘monk robots’ beyond seeing them in their robot cages. Many researchers established that beliefs in animism have no impact or correlation to thinking of a robot as having a soul. As robots become more pervasive not just in official and commercial spaces but also in our personal homes and spaces, humans are expecting more human-like interactions and the need for “social” robots. So, anthropomorphism has been subconsciously accepted in human interactions where, for example, a bot is associated with a particular gender as well,” says Dr Karthik Ramesh, VP–Head International Markets and Innovation, Emids. In focus—Social Robots ‘Social robots’ can be defined as robots that interact with humans and each other in a socially acceptable fashion, conveying intention in a manner perceptible to humans and are empowered to offer solutions to fellow agents, be they human or robot. Furhat Robotics, a social robotics and conversational AI company, explains social robots as “the next major user interface, that are typically designed based on the oldest user interface we as humans know—the face” When Furhat Robotics built its first robot, they aimed to build an intuitive interactive robot capable of emulating social interactions between humans. They wanted the robot to be capable of impersonating different characters to increase its usability and establish its distinction from other ‘fixed personality’ robots. So when we think of rights, are we considering such robots—those with an uncanny resemblance to human appearance, behaviour and capability? Dr David Gunkel believes that the research and development of robot technology has important implications for human life and attempts must be made to better understand the changes such technological advancements may mean to our collective existence in the future. “. . . I think we’ve come to the point where we realise that research and development that is not in touch with the social consequences and an understanding of what this is going to do for us and to us is irresponsible. And that responsible development of technology has become sort of the watchword. We want to make sure not only are we devising these brand new toys and tools and everything else, but we are thinking about what they mean for us and how they will affect [us].” — Dr David J. Gunkel, scholar in AI and Robot Ethics. Human rights for robots? Raging debates on the current state of human rights in the world surround us all. The selective application of universal tenets of humanity is a collective concern. What then is the relevance of debates about robot rights? Dr Kathik Ramesh says, “While this is not yet a matter of grave concern, this will become a pertinent topic in coming times as more of our ‘human’ worlds get inundated with robots, autonomous vehicles.” “The need for a common consensus or framework across cultures, geographies and even human perceptions of how a robot would be perceived is a must before any pre-determination of whether robot rights are required or not. With the increase in accessibility to robots across the world and its super intelligence growing in quantum leaps, it will not be far away when a robot may actually be equivalent to a human in terms of real-time decision-making and function. However, without feelings, a concept of soul or consciousness and the existence of DNA; some purists [may] argue even on the need for consideration of robot rights.”Dr Karthik Ramesh Both sides are important K-2SO: I can blend in. I’m an Imperial droid. The city is under Imperial occupation.Jyn Erso: Half the people here wanna reprogram you. The other half wanna put a hole in your head.Rogue One: A Star Wars Story (2016) Tech companies developing interactive technology like robots often gatekeep how it interacts with the public. Dr David Gunkel believes that such practices should be energetically discouraged and emphasises the importance of citizen participation in decision-making when it comes to technological advancements that may impact their everyday lives. “You have people who are vocal advocates for AI ethics as a way of helping curb the sort of capitalist accumulation of power that is happening in big tech. And then the big tech people who are like, you know, we don’t want regulation or we want limited regulation so that we evolve this technology and implement it in ways that we think is going to serve the public interest. And this is just good democracy. This kind of argument is just what happens when democratic citizens get involved in the shape and direction of their own destinies. And I think it’s actually a good thing”, says Gunkel. He further elaborates—“I think the real question we have is who has the power here to make decisions and implement these things. Power asymmetries are very crucial to recognize and to begin to do something about, because we’re not all equal in this conversation. And I think getting to the point where we can rely on our governments to create a more equitable exchange of ideas and concerns and interests, I think is going to be to everyone’s benefit.” Rights are here and so is the invasion According to Dr Karthik Ramesh, robot rights would imply empowering any machine regardless of its level of intelligence—a legal, innate claim on life, liberty, moral ethics or values, much like a human being. For instance, the humanoid robot, ‘Sophia’ was granted Saudi Arabian citizenship in 2017. In yet another instance, in Pennsylvania, U.S., autonomous delivery drones are allowed to manoeuvre on sidewalks and roadways and are now technically considered “pedestrians”. According to Axios, there are now a dozen states within the U.S. including Pennsylvania, Idaho, Virginia, Florida, Washington, D.C. and Wisconsin where it is legal for personal delivery robots to share the streets with people. In these states, personal delivery robots are granted the same rights and responsibilities that belong to human pedestrians. However, it is noteworthy that, in granting this particular status and the associated rights and responsibilities to the personal delivery robot, the state makes no attempt to seek resolution or address the significant questions surrounding robot moral standing or robot\/AI personhood. All they’re doing is merely attempting to provide a legal framework for integrating these particular devices on our city streets and to align that integration with existing legal practices. “So this is what the robot invasion looks like. It doesn’t take place as, as it’s looking in science fiction with the robots revolting against their human masters and attacking us and rising up in revolution, it’s going to be very mundane. It looks less like Terminator. It looks more like a very boring episode of Law & Order”, remarks Dr David Gunkel.","excerpt":"The selective application of universal tenets of humanity is a collective concern. What then is the relevance of debates about robot rights?","categories":["AI Trends"],"tags":[],"author_name":"Sri Krishna","publish_date":"2022-07-15T17:00:00","publication_year":"2022","word_count":2131,"keywords":["Go","API","artificial intelligence","ELT","AI","RAG","BERT","Aim","GAN","R"],"extracted_tech_keywords":["AI","artificial intelligence","Aim","RAG","R","Go","API","ELT","BERT","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/the-real-reason-robot-rights-is-so-contentious\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":31125,"title":"All You Need To Know About The 2018 Cryptocurrency Slump","content":"Earlier this March, the cryptocurrency market took a massive hit in terms of price. The total market capitalisation went from $372 billion to $309 billion in just one day. Bitcoin was the most affected in this $60 billion variation. Although the digital market picked up from this crash later on, it never caught up with the $560 billion made in December last year. In other words, what saw an all-time high of a monumental $828 billion started to fall down slowly. As a result, people’s interest in investing in cryptocurrency looked bleak in the market. Here are the possible reasons why the crypto market may have gone down so abruptly in the last one year. The Kobayashi Incident Where it all started: The controversial bitcoin exchange Mt.Gox, which went bankrupt in 2014 due to multiple security breaches in their system, can be said to be an indirect factor for the downfall of Bitcoin in early 2018. That’s because Nobuaki Kobayashi, the primary trustee of Mt.Gox sold an undisclosed amount of Bitcoin and Bitcoin Cash between December 2017 and January 2018. Cryptocurrency experts and enthusiasts alike believe that this event might have had a strong impact on the price dip. A report said that Kobayashi’s selling of Bitcoins led to the price being dropped as far as by $6,000 for 2018. Although the trustee’s actions were to credit funds for customers who lost theirs to the Mt.Gox’s hack in 2014, it became less pronounced once bitcoin price started to fall. You can read the complete Mt.Gox’s creditors report here. Aftermath: Customers and critics were not happy with the way Kobayashi took away with selling bitcoins in the end. One financial media reported “Members of the crypto community have (understandably) been expressing their frustrations with the situation on Reddit: “Why didn’t [Kobayashi] sell the BTC at auction like other assets often get sold during bankruptcy? If he sold on the spot market only an idiot would think you wouldn’t suffer slippage,” Reddit user bitradr wrote. Ever since then, Bitcoin never really rose to its peak price after December 2017. Similarly, other currencies witnessed the same trend with prices being low and consistent from then on. Coincheck Hack Another incident at the start of this year was a large scale hack on cryptocurrency exchange Coincheck Inc. With 523 million NEM coins worth $534 million approximately being stolen, Coincheck’s hack was bigger than Mt.Gox’s in 2014 and was dubbed the ‘biggest theft in the history of the world’. Crypto users were instilled with a fear of losing out their digital currency albeit the hack was only on this type of currency. As a result, it left a bad taste amongst users following which they either migrated to other currencies\/exchange or stopped with crypto transactions altogether. Despite this unfavourable outcome, the NEM user community remained stable and the market did not observe a steep dive. Restricting Crypto Activity Incidents like these have made many players such as banking and finance sectors enforcing restrictions on cryptocurrencies. Furthermore, governments in countries such as India, China and Russia have a strong stance against trading crypto. In fact, earlier this year India announced that crypto was not used legally in financial transactions. Finance Minister Arun Jaitley told media that the Indian government ‘does not consider cryptocurrencies legal tender or coin and will take all measures to eliminate the use of these crypto-assets in financing illegitimate activities or as part of the payment system.’ Overall, virtual currency will continue to face flak due to vulnerabilities present in the system which, as a consequence will lead miscreants finding new ways to hack crypto exchanges. Outlook Our article highlighted a few big-time instances where the system of cryptocurrency itself was dangerously vulnerable. Even though the virtual market is rising in terms of user base and unique currencies, it does not seem to have the shine it used to have in the earlier years in terms of value. This can be attributed to two facts: security loopholes and acceptance in the finance ecosystem. Unless these are in control, virtual currencies are deemed to diminish in values.","excerpt":"Earlier this March, the cryptocurrency market took a massive hit in terms of price. The total market capitalisation went from $372 billion to $309 billion in just one day. Bitcoin was the most affected in this $60 billion variation. Although the digital market picked up from this crash later on, it never caught up with […]","categories":["AI Features"],"tags":["Bitcoin","Cryptocurrency","hacking","market"],"author_name":"Abhishek Sharma","publish_date":"2018-12-09T05:32:07","publication_year":"2018","word_count":683,"keywords":["Go","API","Bitcoin","hacking","programming_languages:R","market","AI","Cryptocurrency","programming_languages:Go","Git","ViT","Rust","R","programming_languages:Rust"],"extracted_tech_keywords":["AI","R","Go","Rust","Git","API","ViT","programming_languages:R","programming_languages:Go","programming_languages:Rust"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/all-you-need-to-know-about-the-2018-cryptocurrency-slump\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10074792,"title":"The Merge Is Killing Nvidia GPU Prices","content":"The Nvidia GPU prices have turned on their head this year. As the company braces for another price slip for its RTX 30-series GPUs, the concern over how massive these declines could be in the future only grows deeper. GPU prices have slid over the past week. While crypto mining remains profitable in some ways, the value of the Bitcoin downturn rose by 30% and Ethereum dropped by almost 40%.  “The decline in the gaming GPU revenue was sharper than anticipated for Nvidia, driven by both lower units and lower ASPs (average selling prices),” said Colette Kress, chief financial officer and EVP of Nvidia. The Ethereum merge (shift from proof of work to proof of stake) significantly affected almost tens of millions of GPUs purchased in the past four years to mine ether. Some miners even planned to move to Ethereum Classic but couldn’t attract the mass of NFT or decentralized finance (DeFi). Moreover, the overall vulnerability has been speculated after the transition to POS (proof of stake). Currently, the overall GPU unit shipments have declined by 15% from last quarter. Half of this slowdown comes from the softening PC market, with researchers reporting a sharp dip every year, almost 12.8% for 2022, and partly from the vanishing crypto miners. Ethereum merge and its effects The merge represents the shift of Ethereum’s proof of work (POW) to proof of stake. It’s a new system for authenticating crypto transactions. According to the Ethereum Foundation, this concept is introduced to reduce Ethereum’s energy usage by 99%. “I am personally very excited about the Ethereum Merge and the upgrade to the proof of stake consensus mechanism, primarily because it will lower the network’s footprint by 99%. In case of Ethereum, this would mean returning 0.1% of global electricity usage to people,” said Pratik Gauri, co-founder & CEO, 5ire. Moreover, Ethereum POS also limits the stake and cost, unlike POW. Well, it doesn’t come with the certainty of smashing the POW but accounts for clear signals from entrepreneurs and decision-makers so far. On the flip side, this may swap away hundreds of billions of dollars in value (Ether’s market cap is close to $200 billion and many worthy tokens). The Ethereum merger is expected to happen around September 15. Buy Nvidia stock now Amidst intensified inflation, low demands, pandemic and supply chain limitations, Nvidia went through tough times since the beginning of this year. The past six days have been significantly  rough for Nvidia, with share pricing falling rapidly below expectations because of the merger. After the disappointing second-quarter earnings that couldn’t meet analysts’ expectations in August, Nvidia was smacked with a new export license requirement for its future chip sale with China. On September 1, Nvidia’s stock dropped to a 52-week low of $132.7 after the company announced the new export license. Nvidia is one of the biggest technology companies in the global markets, with a market cap of $450 bn. Analysts proclaim this rough patch as short-term volatility for investors against the bullish possibilities in the near future. So it would be safe to buy a small portion in Nvidia today. Nvidia has a way out Although the present condition of Nvidia appears gloomy from its declining stocks and GPU prices. But fortunately, the company has its own way of compensating for the losses. The price of Ethereum has been down since mid-2021 and that has remained a uniform indicator of GPU prices. However, despite the setback, the company has something good in store to better its steeping market state – the new generation of graphics cards. Last month, the value of these graphic cards dropped, when compared to the previous generations. That’s not unfortunate but rather a green signal for AMD RX 6000 and Nvidia RTX 30 series graphics cards. What’s more exciting is the upcoming launch of AMD’s RX 7000 graphics cards in the second half of 2022. Nvidia is also planning to launch the RTX40-series, clearly portraying a strong plan. Even though GPU prices have fallen, Nvidia has clear plans to buckle up with new generations every two years, either in September or October. This launch will surely make up for its losses against the present shaky GPU prices.","excerpt":"The Ethereum merger is expected to happen around September 15","categories":["Global Tech"],"tags":["crypto","NFT","NVIDIA"],"author_name":"Nidhi Bhardwaj","publish_date":"2022-09-11T10:00:00","publication_year":"2022","word_count":700,"keywords":["Go","API","programming_languages:R","AI","R","programming_languages:Go","RAG","Ray","Aim","crypto","NVIDIA","NFT"],"extracted_tech_keywords":["AI","Aim","Ray","RAG","R","Go","API","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/the-merge-is-killing-nvidia-gpu-prices\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10021831,"title":"What is Face Identity Disentanglement and How it outperformed GANs?","content":"Face Identity Disentanglement via Latent Space Mapping becomes the state-of-the-art in face image generation by greatly surpassing existing Generative Adversarial Networks such as StyleGAN. Generative Adversarial Networks, simply known as GANs, nowadays find a prominent place in deep learning with wide applications including high-resolution image synthesis, image-to-image translation, video-to-video translation, image inpainting, and video inpainting. StyleGAN and other competing methods are well known for their face image generation abilities. However, they need excessive supervision and training, and compromised quality which make generalization difficult. Face Identity Disentanglement via Latent Space Mapping is a method that learns how to represent image data in disentangled latent representations, with minimal supervision, manifested using available pre-trained generative networks such as StyleGAN. By learning to map into latent space, state-of-the-art quality as well as rich-expressive latent space are achieved. Disentangled latent representations allow generative models to control and compose the disentangled factors in the image generation process. Disentanglement is a generative model’s ability to solely control a single feature without affecting the other features. For instance, in face generation, disentanglement helps either generate faces of the same identity but with different attributes such as pose, expression and illumination, or generate faces of the same pose but with different identities. Disentanglement is considered a non-trivial task in machine learning. The current framework demonstrates high-quality disentanglement of face identity from all other attributes, capable of generating high-resolution faces of different identity and\/or attributes. This framework’s key idea is to map the disentangled latent representation to the latent space of a pre-trained generator such as StyleGAN. This Face Identity Disentanglement framework is developed by Yotam Nitzan, Amit Bermano, Daniel Cohen-Or of Tel-Aviv University and Yangyan Li of Alibaba Cloud Intelligence Business Group. Disentanglement framework with Latent Space Mapping This disentanglement framework uses two encoders to generate the latent representation ????, consisting of a description of the property of interest, and all the rest. Here, the first encoder generates a latent representation of the identity of the face and the second encoder generates a latent representation of facial attributes such as pose, expression and illumination. The latent representation is then mapped to the latent space W of the pre-trained generator ????. This decouples the tasks of learning quality image generation and disentanglement. Due to disentanglement, the two parts of latent representations are mutually exclusive and carry entirely different information. Therefore, this approach mapping is trained solely to successfully disentangle provided input information and extract useful representation that can be combined in the generator to synthesize high-quality target images. In this disentanglement framework, three inputs are used to generate 3 by 3 image-matrix by preserving face identity along with columns and preserving facial attribute along rows. Human faces possess many independent, high-dimensional features, and high photometric, geometric and kinematic complexities. This disentanglement framework concentrates on image synthesis with disentangled control over face identity while preserving the other facial attributes. This type of control is highly useful in applications such as reenactment and de-identification. The output quality is directly determined by the pre-trained generator employed. Hence this framework incorporates the state-of-the-art StyleGAN as the pre-trained generator. The above illustration depicts the dataflow and losses of the framework. Two input images, one for face identity feature, ???????????? and another for facial attributes, ???????????????????? are fed to respective encoders. The latent representations are mapped to latent space, which is then fed into the generator. An adversarial loss L???????????? ensures proper mapping to the W space. Identity preservation is encouraged using L???????? , that penalizes differences in identity between ????????????, ???????????????? . Attributes preservation is encouraged using L???????????? , L???????????? , that penalizes pixel-level and facial landmarks differences respectively, between ???????????????????? , ????????????????. Python Implementation of Disentanglement Framework Step-1: The following command imports necessary source codes, files and datasets from the official Github repository. Make sure that CUDA GPU runtime is enabled on the local machine or Colab or Jupyter notebook. !git clone https:\/\/github.com\/YotamNitzan\/ID-disentanglement.git Output: Confirm the proper file download using the command !ls ID-disentanglement\/ Output: Step-2: Activate the conda environment on the local machine. If Anaconda-3 is not installed on the machine or if the user uses Colab, the following command installs Anaconda-3 distribution. For 64-bit machine, !wget https:\/\/repo.anaconda.com\/archive\/Anaconda3-2020.02-Linux-x86_64.sh !bash Anaconda3-2020.02-Linux-x86_64.sh For 32-bit machine, !wget https:\/\/repo.anaconda.com\/archive\/Anaconda3-2020.02-Linux-x86.sh !bash Anaconda3-2020.02-Linux-x86.sh Step-3: Various pre-trained generators are available forr training and inference. Users can opt for any of the available generators from the corresponding Github repository. Here FFHQ_StyleGAN_256x256 model is used in Colab. Since models are stored in a shared directory in Google Drive, the necessary setup in Colab to download files from Google Drive must be enabled using the following commands and codes. !pip install -U -q PyDrive import os from pydrive.auth import GoogleAuth from pydrive.drive import GoogleDrive from google.colab import auth from oauth2client.client import GoogleCredentials Users must authenticate access to Google Drive via Google Cloud Storage by generating the following codes’ verification code. auth.authenticate_user() gauth = GoogleAuth() gauth.credentials = GoogleCredentials.get_application_default() drive = GoogleDrive(gauth) Finally, pre-trained StyleGAN can be downloaded using the codes local_download_path = os.path.expanduser('~') try: os.makedirs(local_download_path) except: pass file_list = drive.ListFile( {'q': \"'1OgLvUhd9FX9_mPXrfqAWaLZsceQzE9l4' in parents\"}).GetList() for f in file_list: # 3. Create & download by id. print('title: %s, id: %s' % (f['title'], f['id'])) fname = os.path.join(local_download_path, f['title']) print('downloading to {}'.format(fname)) f_ = drive.CreateFile({'id': f['id']}) f_.GetContentFile(fname) with open(fname, 'rb') as f: print(f.read()) Step-4: The Face Identity Disentanglement framework is designed to use Tensorflow 2.X on python (3.7), using cuda 10.1 and cudnn 7.6.5. Following commands create a conda environment that has the needed dependencies. !exec bash Within the shell, content run the following command conda create -n environment.yml Output: Step-5: Dataset for training and inference can locally be created using the commands %%bash cd ID-disentanglement\/utils python generate_fake_data.py \\ --resolution N \\ --batch_size BATCH_SIZE \\ --output_path OUTPUT_PATH \\ --pretrained_models_path PRETRAINED_MODELS_PATH \\ --num_images NUM_IMAGES \\ --gpu GPU Step-6: The Face Identity Disentanglement framework can be trained using the following command %%bash python main.py NAME --resolution N --pretrained_models_path PRETRAINED_MODELS_PATH --dataset BASE_DATASET_DIR --batch_size BATCH_SIZE --cross_frequency 3 --train_data_size 70000 --results_dir RESULTS_DIR Step-7: Inference on the trained model with the downloaded test dataset can be performed using the following commands. %%bash python test.py Name --pretrained_models_path PRETRAINED_MODELS_PATH \\ --load_checkpoint PATH_TO_WEIGHTS \\ --id_dir DIR_OF_IMAGES_FOR_ID \\ --attr_dir DIR_OF_IMAGES_FOR_ATTR \\ --output_dir DIR_FOR_OUTPUTS \\ --test_func infer_on_dirs to test performance on two sets of images, one for preserving face identity and another for preserving facial attributes. %%bash python test.py Name --pretrained_models_path PRETRAINED_MODELS_PATH \\ --load_checkpoint PATH_TO_WEIGHTS \\ --input_dir PARENT_DIR \\ --output_dir DIR_FOR_OUTPUTS \\ --test_func interpolate to test performance on three sets of images, one for preserving face identity and the other two sets for sequential interpolation of facial attributes. Performance Evaluation of Disentanglement Framework Qualitative and the quantitative performance of the disentanglement framework areevaluated using the Flickr-Faces-HQ images (FFHQ). Input images from FFHQ image dataset. Face identity is preserved along with columns while other facial attributes are preserved along rows. Both input and output images are generated using StyleGAN generator incorporating Face Identification Disentanglement Framework Qualitative comparison of Disentanglement framework with existing state-of-the-arts FSGAN and FaceShifter. Disentanglement approach extraordinarily exceeds performances of state-of-the-arts in identity-preserved face generation such as FaceShifter, FSGAN, ALAE, and pSp.. Notable applications of the Disentanglement framework Sequential interpolation of a given image between two different input images of different  attributes. The identity of a given image is maintained throughout the interpolation. Here the input image of the identity source is not shown. Sequential interpolation of two different input images of different identities and attributes. Both identity and attributes are matched to the input images at both ends of interpolation. Note: Images and illustrations other than the code outputs are obtained from the original research paper. References and further reading: Original research paperSource code repositoryStyleGAN repositoryFlickr-Faces-HQ imagesPre-trained StyleGAN_256x256 generator","excerpt":"Face Identity Disentanglement via Latent Space Mapping becomes the state-of-the-art in face image generation by surpassing existing GANs","categories":["AI Trends"],"tags":["Face Swapping GAN","GAN","GANs","image synthesis","StyleGAN"],"author_name":"Rajkumar Lakshmanamoorthy","publish_date":"2021-03-11T17:00:00","publication_year":"2021","word_count":1273,"keywords":["CUDA","machine learning","Face Swapping GAN","TPU","AI","ML","StyleGAN","Colab","RAG","deep learning","GANs","Jupyter","image synthesis","GAN","TensorFlow"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","TensorFlow","Jupyter","Colab","RAG","TPU","CUDA"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/what-is-face-identity-disentanglement-and-how-it-outperformed-gans\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10167230,"title":"ISRO to Establish Infrastructure for Space Activities in Assam","content":"Assam is set to host a new infrastructure for space-related activities, following a joint reconnaissance visit by Keshab Mahanta, minister for science and technology in the Assam government and V Narayanan, chairperson at ISRO. The visit aimed to establish a multi-purpose facility in the state, which will enhance the safety of India’s space assets and support weather advisory services. The facility will be essential in supporting India’s space program. Mahanta pledged full support from the Assam government for this endeavor. He also expressed his vision for establishing a space museum in the area, which would benefit students and the public. Narayanan welcomed the idea and assured ISRO’s cooperation. As part of this initiative, ISRO will provide necessary technical assistance to ensure the project’s success. In a related development, Assam is moving forward with its satellite project, ASSAMSAT, in collaboration with ISRO. The satellite will focus on providing services for agriculture, disaster management, infrastructure development, and border security. It will be controlled from Guwahati and include a constellation of low-earth orbiting satellites. CM Himanta Biswa Sarma has stated that Assam will be the first state in India to have its own satellite, which will aid in monitoring illegal infiltration and enhancing disaster preparedness. Apart from government organisations, many big technology and IT companies have also recently shown interest in the state. Reliance announced plans to invest ₹50,000 crore in Assam for AI, nuclear energy and data centres. ‘AI’ also stands for ‘Assam Intelligence’, said Mukesh Ambani. Meanwhile, Infosys also expanded its presence in the state by establishing a 230-member project development centre in Guwahati, marking a significant step in Assam’s push to become a major IT hub.","excerpt":"As part of this initiative, ISRO will provide necessary technical assistance to ensure the project’s success.","categories":["AI News"],"tags":["Assam","ISRO"],"author_name":"Sanjana Gupta","publish_date":"2025-04-03T16:38:56","publication_year":"2025","word_count":276,"keywords":["Go","ISRO","programming_languages:R","AI","programming_languages:Go","GAN","Ray","Aim","ViT","Assam","R"],"extracted_tech_keywords":["AI","Aim","Ray","R","Go","GAN","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/isro-to-establish-infrastructure-for-space-activities-in-assam\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10048146,"title":"A Beginners’ Guide to Cross-Entropy in Machine Learning","content":"Today we have many real-world applications which are based on machine learning such as churn modeling, image classification, customer segmentation, etc. For all of these kinds of applications, businesses need to optimize their models, obtain the model’s optimum accuracy and efficiency model. Therefore it is a bit critical to obtain a higher-performing model by tuning a certain number of parameters. One such parameter is a loss function and among which mostly used one is cross-entropy. In this article, we will be discussing cross-entropy functions and their importance in machine learning, especially in classification problems. The important concepts that we will discuss here in this article are listed below. Table of Contents Need of Cross-EntropyWhat is Entropy?What is cross-entropyCross entropy as loss function Code Glimpses Let’s begin our discussion one by one. Need of Cross-Entropy Machine learning and deep learning models are normally used to solve regression and classification problems. In a supervised learning problem, during the training process, the model learns how to map the input to the realistic probability output. As we already know, the model adjusts its parameters incrementally during the training phase of supervised learning so that prediction gets closer to closer as expected values (ground truth). Let’s take an example of three images, which represents three class of vehicles as shown below and each image is encoded in the binary representation below; ClassLabelsCar[1,0,0]Excavator[0,1,0]Tank[0,0,1] The above is the actual representation of our training data which we fed to our model as input and output class. When we give an image of the excavator to our model, the model tries to generalize parameters for it and return a probability distribution for all the three classes like [0.36,0.65,0.45]  which is completely different from what we actually want. The model should report a higher probability for the ground truth. Well in our case it has been reported not much higher as we can see in the rest of the two there is close ambiguity between the excavator and tank. Well in the next iterative process the model tries to improve its prediction by changing the output from y’ to y. This task is undertaken by the various loss functions. It depends on which problem we are dealing with. For multi-class classification, the cross-entropy loss function is used, which basically tells our model in which direction the prediction is closer to the ground truth. What is Entropy? Claude Shannon, a mathematician, and electrical engineer was trying to figure out how to deliver a communication message without losing a piece of information. He was thinking in terms of average message length, which meant he was attempting to encode a message with the fewest possible bits. Aside from that, he expected that the decoder would be able to reconstruct the message losslessly, meaning that no data would be lost in this way he invented the idea of entropy. Entropy on Decision Tree Here entropy is defined as the smallest average encoding size per transmission with which a source can send a message to a destination efficiently and without losing any data. Entropy can be defined mathematically using the probability distribution denoted as H. when we are talking about the categorical variables the formula looks like this Source What is Cross-Entropy? Assume we have two distributions of data and need to be compared. Cross entropy employs the concept of entropy which we have seen above. Cross entropy is a measure of the entropy difference between two probability distributions. Assume the first probability distribution is denoted by A and the second probability distribution is denoted by B. The average number of bits required to send a message from distribution A to distribution B is referred to as cross-entropy. Cross entropy is a concept used in machine learning when algorithms are created to predict from the model. The construction of the model is based on a comparison of actual and expected results. Mathematically we can represent cross-entropy as below: Source In the above equation, x is the total number of values and p(x) is the probability of distribution in the real world. In the projected distribution B, A is the probability distribution and q(x) is the probability of distribution. So working with two distributions how do we link cross-entropy to entropy? If the expected and actual values are the same then cross-entropy equals entropy. In the real world, however, the predicted value differs from the actual value which is referred to as divergence, because they differ or diverge from the actual value. As a result, cross-entropy is the sum of Entropy and KL divergence (type of divergence). Cross-Entropy as Loss Function When optimizing classification models, cross-entropy is commonly employed as a loss function. The logistic regression technique and artificial neural network can be utilized for classification problems. In classification, each case has a known class label with a probability of 1.0 while all other labels have a probability of 0.0. Here model calculates the likelihood that a given example belongs to each class label. The difference between two probability distributions can then be calculated using cross-entropy. In classification, the goal of probability distribution P for an input of class labels 0 and 1 is interpreted as probability as Impossible or Certain. Because this probability includes no surprises (low probability event) they have no information content and have zero entropy. When we are dealing with Two Class probability, the probability is modelled as Bernoulli distribution for the positive class. This means that the mode explicitly predicts the probability for class 1, while the probability for class 0 is given as 1 – projected probability. For more clearly say; Class 1 = 1 (originally predicted)Class 0 = 1 – originally predicted We are frequently concerned with lowering the model’s cross-entropy throughout the entire training dataset. This can be done by taking the average cross-entropy of all training sets. Code Glimpses Below we have used the python function to calculate the cross-entropy based on the above formula, later we checked whether the calculated cross-entropy is equal to the actual one or not by the logical function. def cross_entropy(predicted, Ground_Truth, e=1e-12): \"\"\" Computes CROSS_ENTROPY between Ground_Truth (encoded as one-hot vectors) and predicted value. Input: predicted value (N, k) ndarray Ground_Truth (N, k) ndarray Returns: fraction \"\"\" predicted = np.clip(predicted, e, 1. - e) N = predicted.shape[0] CROSS_ENTROPY = -np.sum(Ground_Truth*np.log(predicted+1e-9))\/N N = predicted.shape[0] CROSS_ENTROPY = -np.sum(Ground_Truth*np.log(predicted+1e-9))\/N return CROSS_ENTROPY predicted = np.array([[0.25,0.25,0.25,0.25], [0.01,0.01,0.01,0.96]]) Ground_Truth = np.array([[0,0,0,1], [0,0,0,1]]) ans = 0.71355817782  #Correct answer x = cross_entropy(predicted, Ground_Truth) print(np.isclose(x,ans)) # prints where calculated entropy is equal to the ans In this case, the code returns True which means the calculated and actual values are equal to each other. Conclusion In this article, we have seen the intuition theory of cross-entropy. When we are dealing with multiple classes and want our model to converge faster by countering the loss, cross-entropy plays a crucial role. All the major frameworks support this loss function, in Keras for binary classification we have the Bnary_crossentropy function and for multi-class, we have the Categorical_crossentropy function. References Cross-entropy https:\/\/machinelearningmastery.com\/cross-entropy-for-machine-learning\/","excerpt":"Machine learning and deep learning models are normally used to solve regression and classification problems. In a supervised learning problem, during the training process, the model learns how to map the input to the realistic probability output.","categories":["Deep Tech"],"tags":["churn prediction model","cross entropy","Deep Learning","Guide","Loss functions","Machine Learning","when to use support vector machine"],"author_name":"Vijaysinh Lendave","publish_date":"2021-09-11T12:00:00","publication_year":"2021","word_count":1168,"keywords":["machine learning","Keras","TPU","AI","neural network","Machine Learning","Loss functions","when to use support vector machine","RAG","Ray","Python","deep learning","churn prediction model","Deep Learning","R","cross entropy","Guide"],"extracted_tech_keywords":["AI","machine learning","deep learning","neural network","Ray","Keras","RAG","TPU","Python","R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/a-beginners-guide-to-cross-entropy-in-machine-learning\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10019782,"title":"The 10 Best Linux Distro for Stability in 2024","content":"Linux is the most popular open-source and programmer-friendly operating system with several advantages over other OS in terms of security, flexibility, and scalability. A Linux distribution (aka distro) is an OS made from softwares based on the Linux kernel. Users download Linux from one of these distros. Let’s look at the ten most stable Linux Distros in 2024. 1. ArchLinux Suitable for: Programmers and Developers About: Arch Linux is a lightweight and flexible Linux distribution with a simple user environment. The independently developed, x86-64 general-purpose GNU\/Linux distribution provides the latest stable versions of most software by following a rolling-release model. The features of this distribution include pragmatic distribution, user centrality, versatility, simplicity, among others. Know more here. 2. Debian Suitable for: Beginners About: Debian is a popular stable and secure Linux based operating system. Various popular Linux distributions, such as Ubuntu, PureOS, SteamOS, etc choose Debian as a base for their software. Notable features are: Extensive hardware support Provides smooth upgrades Security support for releases. Free and open-source software. Know more here. 3. Fedora Suitable for: Software Developers, Students About: The distribution creates a platform ideal for hardware, containers and the cloud, allowing software developers and the community members to build tailored solutions for their users. Fedora Workstation is a user-friendly operating system that supports a wide range of developers, from students to professionals in corporate environments. Know more here. 4. Linux Mint Suitable for: Professionals, Developers, Students About: Linux Mint is one of the most popular desktop Linux distributions that is both free and open source. Based on Debian and Ubuntu, Mint provides about 30,000 packages and is one of the best software managers. The distro provides full multimedia support and is extremely easy to use. Know more here. 5. Manjaro Suitable for: Beginners About: Manjaro Linux is a fast, desktop-oriented operating system based on Arch Linux. The distro provides all the benefits of cutting-edge software to get started quickly and automated tools to minimise manual intervention. It is completely free and is available for 64 Bit architectures. Know more here. 6. openSUSE Suitable for: Beginners and advanced users About: openSUSE, formerly known as SUSE Linux, is one of the most easy to use Linux distributions. The aim of this distribution is to create usable open-source tools for software developers and system administrators while providing a user-friendly desktop and feature-rich server environment. Know more here. 7. SparkyLinux Suitable for: Gamers About: SparkyLinux is a GNU\/Linux distribution created on top of Debian GNU\/Linux operating system. The fast, lightweight and fully customisable operating system offers different versions for different tasks, including a fully featured OS with a lightweight desktop environment and MinimalGUI with Openbox window manager pre-installed with basic software. The features include stable or (semi-)rolling release, special editions: GameOver, Multimedia & Rescue, CLI Edition (no X) for building customised desktop, etc. Know more here. 8. Tails Suitable for: Security and privacy About: Tails is a portable operating system resistant to surveillance and censorship. The OS uses the Tor network to protect privacy online. Also, Tails includes a selection of applications to work on sensitive documents and communicate securely. Features include: resilience against surveillance, advertising and viruses; leaves no trace on the computer when shut down; based on Debian GNU\/Linux etc. Know more here. 9. Ubuntu Suitable for: Developers, Professionals, Students About: Ubuntu is a popular, open source desktop operating system with all the  essential applications such as an office suite, browsers, email and media apps, etc. For instance, you can create professional documents, spreadsheets and presentations on Ubuntu with LibreOffice. Know more here. 10. Zorin OS Suitable for: Beginners, Professionals About: Zorin OS is a powerful operating system designed to make your computer faster, more secure and easier to use. The operating system supports over 50 languages and comes pre-loaded with assistive technologies. Features include flexibility, accessibility, compatibility, etc. Know more here.","excerpt":"Linux is the most popular open-source and programmer-friendly operating system with several advantages over other OS in terms of security, flexibility, and scalability. A Linux distribution (aka distro) is an OS made from softwares based on the Linux kernel. Users download Linux from one of these distros. Let’s look at the ten most stable Linux […]","categories":["AI Trends"],"tags":["Linux distros","Linux Ubuntu"],"author_name":"Ambika Choudhury","publish_date":"2021-02-07T13:00:00","publication_year":"2021","word_count":642,"keywords":["Linux distros","Linux Ubuntu","AI","programming_languages:R","data_tools:Spark","Scala","RAG","Aim","programming_languages:Scala","R"],"extracted_tech_keywords":["AI","Aim","RAG","R","Scala","programming_languages:R","programming_languages:Scala","data_tools:Spark"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/best-linux-distros-for-stability\/","complexity_score":4,"technical_depth":8,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10165369,"title":"‘Freshworks, Zoho Offer Less Than 10% of What Oracle Provides’","content":"Oracle India is deepening its commitment to the Indian SaaS market by incorporating AI-driven automation across its business applications. The cloud service provider offers AI agents throughout its full SaaS portfolio with no added charges. In an exclusive interview with AIM, Shailesh Singla, vice president of cloud applications at Oracle India, said that India is a big market for them and explained how the agentic services that Oracle provides are different from what local SaaS players like Zoho and Freshworks offer. These companies, he stressed, focus on specific areas. “Freshworks offers a CRM (customer relationship management) solution. It’s a CRM company,” Singla said, referring to them as smaller players. Singla explained that Oracle, on the other hand, is building a connected application that spans enterprise resource planning (ERP), supply chain management (SCM), customer experience (CX), human capital management (HCM) and customer relationship management (CRM). He described Oracle’s approach as a “holistic application stack”, comparing it with Zoho and Freshworks who, according to him, offer only a fraction—“maybe 5%, 7%, or 10%”—of what Oracle provides as a complete solution. Moreover, he added that Indian SaaS players do not have their own cloud and many of these smaller players rely on AWS or Microsoft Azure for hosting. He further mentioned that customers looking for a SaaS provider analyse various factors, including system architecture, a comprehensive approach, and long-term roadmaps spanning five to 10 years. Oracle as SaaS Singla said Oracle’s mission with AI agents is to automate tasks, handle complex workflows, and provide insights. “These agents are embedded in applications to assist users in real time.” They operate within Oracle’s SaaS products, reducing manual effort. According to him, Oracle offers AI agents free of cost, bundled with its SaaS services at no extra charge. The company does not charge for AI agents because it sees them as a fundamental part of building smart applications. “We don’t charge for building a smart application.” Notably, Oracle recently launched 50 AI agents to automate routine tasks across key areas such as finance, HR, supply chain, sales, marketing, and customer service, allowing employees to focus on more strategic initiatives. In HCM, Singla said that AI agents automate time card submissions. They are also used to address employee queries regarding benefits and policies. “AI agents are directly able to understand the query, search the right data, and are able to address those queries,” he added. For managers, he said these agents assist in compensation decisions by analysing market trends and company policies. Performance goal setting is also supported by AI, which generates draft documents for review. “AI agents help them document and draft those goals in a much smarter way.” Citing an example of an AI agent in supply chain management, he explained that it automates requisition processes based on organisational guidelines, product specifications, and historical data. Similarly, a manufacturer onboarding advisor agent simplifies the process for new suppliers, making it smoother and more compliant. “For manufacturing suppliers, onboarding can be complex. Our AI agent facilitates this process, ensuring compliance with risk policies and regulatory formalities.” However, Indian SaaS players have also introduced AI agents in their services lately. Freshworks recently introduced Freddy AI which automates repetitive tasks by offloading monotonous work, scaling customer support, and improving customer experiences. Meanwhile, Zoho is building task-specific autonomous agents known as ‘Zia Agents, ’ which are being deployed across its extensive product portfolio, including sales, marketing, and ad services, for automation, security, real-time decision-making, and better explainability. Currently, the SaaS company is also building foundational models, which it plans to release by the end of this year. India Market Singla highlighted that Oracle is seeing a strong demand for Oracle’s SaaS solutions in the Indian market. “We are seeing 100% year-on-year growth in certain verticals, which speaks volumes about the demand for Oracle’s SaaS solutions in India.” He also mentioned that non-banking financial companies (NBFCs) are a high-growth sector for Oracle, with the company adding one to two NBFC clients every month. “Our BFSI and IT verticals are growing at 60%, showing strong market traction,” he concluded.","excerpt":"Indian SaaS players do not have their own cloud and many of these smaller players rely on AWS or Microsoft Azure for hosting.","categories":["Global Tech"],"tags":["Oracle"],"author_name":"Siddharth Jindal","publish_date":"2025-03-06T16:58:02","publication_year":"2025","word_count":677,"keywords":["Go","API","autonomous agents","AWS","AI","R","Oracle","automation","Aim","GAN","Azure"],"extracted_tech_keywords":["AI","Aim","autonomous agents","AWS","Azure","R","Go","API","GAN","automation"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/freshworks-zoho-offer-less-than-10-of-what-oracle-provides\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":58950,"title":"Gurgaon-Based Staqu Introduces Camera For COVID-19 Under Analytics Platform JARVIS","content":"Driven by a vision to solve real-world problems by integrating advanced technologies like A.I., the Gurgaon-based start-up Staqu launched a new Thermal Camera under its video analytics platform, JARVIS. The latest technology will alert the system of anyone with a body temperature of above 37° C and examines heat signatures directly through the cameras, enabling authorities to identify and further inspect suspected Virus Carriers. The camera has a range of up to 100 meters and can identify multiple people at the same time. The technology is functional and effective in scanning crowded places like Airports, Railway Stations, Malls, etc. The AI-powered system is enabled with a Sensitivity ranging from High sensitivity range of -40 to 160C or -40 to 320F, whereas the low sensitivity of -40 to 550C and -40 to 1022F. From airports to railway stations, the AI-powered Thermal Cameras can be installed at various hotspots of consequence for the purpose of early detection and prevention of the spread. With the growing scare around novel Coronavirus and the possibility of our health infrastructure getting overwhelmed, the implementation of these cameras becomes pertinent and urgent. By reducing both the time and human effort required in identification, agencies can get ahead of the curve in controlling the spread of this pandemic. Commenting on the latest development, Mr Atul Rai, CEO and Co-founder said – “The novel coronavirus has been declared a pandemic and required well-meditated efforts from both, the private sector companies and public sector authorities to control its spread in India and across the globe. To the same end, we would like to extend our help in helping Indians combat the COVID-19. Our latest technology removes the human intervention in early stages of detection; through heat wave analysis and more, the technology helps in proactively taking preventive efforts to reduce the risk of the spreading the virus. We are hopeful that the government and other authorities would make the best of this advancement.” Staqu’s AI-powered solution is being trusted and used by multiple law enforcement agencies and accompanies across the world. Staqu is currently working with eight state and UT police forces, including Punjab, Haryana, Rajasthan, Uttar Pradesh, Bihar, Uttarakhand and Telangana to usher the state in AI-powered smart policing.","excerpt":"Driven by a vision to solve real-world problems by integrating advanced technologies like A.I., the Gurgaon-based start-up Staqu launched a new Thermal Camera under its video analytics platform, JARVIS. The latest technology will alert the system of anyone with a body temperature of above 37° C and examines heat signatures directly through the cameras, enabling […]","categories":["AI Features"],"tags":["staqu"],"author_name":"Vishal Chawla","publish_date":"2020-03-18T15:28:35","publication_year":"2020","word_count":370,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","staqu","ViT","analytics","Rust","GAN","R","analytics platform"],"extracted_tech_keywords":["AI","analytics","R","Go","Rust","GAN","ViT","analytics platform","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/staqu-introduces-camera-for-covid-19-detection-under-analytics-platform-jarvis\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10109899,"title":"[Exclusive] Bengaluru Developer Makes his Alto Autonomous with Redmi Phone","content":"Bengaluru-based developer, Mankaran Singh, recently converted his modified-Maruti Alto K10 into an autonomous vehicle using a second-hand Redmi Note 9 Pro. This was made possible with the help of Flowpilot, an open-source driver assistance system created by him. https:\/\/twitter.com\/Mankaran32\/status\/1657723885435994115 Flowpilot is an open-source fork of Comma.ai’s OpenPilot that can run on most Windows\/Linux and Android-powered machines. In an exclusive interview with AIM, Singh said that he didn’t train driving models on his own; instead, he used learning models from Comma.ai because one needs tonnes of data to train models and require millions of dollars for compute clusters to train them. The idea to start this initiative stems from George Hotz, the founder of Comma.ai, which is also into enabling autonomous vehicles with the help of smartphones and other proprietary devices. An avid programmer and autonomous system enthusiast, Singh is also into gaming and developing web applications. Besides Ola – which he joined three months ago – he is working on his passion project Flow Drive, the think-tank behind Flowpilot. Flowpilot performs the functions of Adaptive Cruise Control (ACC), Automated Lane Centering (ALC), Forward Collision Warning (FCW), Lane Departure Warning (LDW), and Driver Monitoring (DM) for a growing variety of supported car makes, models, and model years maintained by the community. It records data from road-facing cameras, CAN, GPS, IMU, magnetometer, thermal sensors, crashes, and operating system logs. “All you need is basically actuators, the controller steering and gas brakes to control the car. And if you have that, Flowpilot can essentially run on anything,” said Singh. “It supports all the phones (Android) that have OpenCL supporting them. That’s basically the GPU drivers, which are used for image processing, like neural networks, exhibiting neural networks,” he added. Singh said that the quality of the experience definitely depends on the phone you use. “Mostly, if you’re using a phone that costs around 20-25K, that’s enough to run Flowpilot for reasonable performance, but anything less powerful, it’s going to lag, and the system will just show a warning; it just won’t engage,” he explained. Flowpilot supports over 200 cars, including brands such as Honda, Toyota, Hyundai, Nissan, Kia, Chrysler, Lexus, Acura, Audi, VW, and more. Even if a car is not supported but has adaptive cruise control and lane-keeping assist, it’s will likely be able to run Flowpilot. However, Singh said that it does not support cars in India as of now. “In India, none of the cars are supported in a plug-and-play fashion. The only reason people use Flowpilot is because it’s far better than the stock system of the car,” he said How it Differs from Tesla “Tesla has a long-term vision. They are using eight cameras and complex neural networks, including advanced 3D perception. However, dealing with eight cameras brings a much bigger computational load. While a smartphone can handle images from two or three cameras well, it gets challenging with eight cameras, all running at a fast 50 hertz. This requires real-time processing on a special chip designed for cars,” explained Singh. He believes that Flowpilot has great potential on highways. “Highway driving can be automated with as little as one camera or a forward-facing camera and a bunch of other sensors, maybe GPS, IMU,” he said. Singh said that the immediate use case involves highway driving automation, which is advancing rapidly with features like ADAS (Advanced Driver Assistance Systems) becoming more prevalent in cars. “Many car companies are incorporating highway automation, where highways provide a more structured environment. In this scenario, drivers can sit back, relax, and let the car cover the entire highway,” he explained. Other safety features include automatic emergency braking and blind spot monitoring. If a driver attempts to change lanes when an obstacle is present, the system will issue a warning. The machine learning-based braking system operates at a much higher rate than human processing, allowing it to brake at the right time and potentially save lives. As technology continues to advance, autonomous vehicles may find applications in private spaces such as college campuses and tech parks. These environments, with controlled and predictable conditions, make them more suitable for autonomous transportation, shares Singh. Limitations Speaking of safety, Singh said some people retrofit custom motors on their steering, which he does not recommend at all. “The only recommended approach, given that your car doesn’t support Openpilot or Flowpilot, is to just replace the steering with the one from a supported car because these actuators are safety-critical,” he explained Though Flowpilot is a good starting point for autonomous vehicles in India, it has several limitations, considering the condition of Indian roads. As Flowpilot relies on mobile cameras, unlike higher-end systems employing LiDAR and additional cameras, it struggles with low-light conditions, shadows, and unpredictable objects like cyclists or animals. This vulnerability in perception can lead to misinterpretations of the environment, potentially leading to accidents. Moreover, India’s road infrastructure varies greatly, with many roads lacking lanes, signage, and proper markings, posing challenges for autonomous vehicles to navigate accurately. Currently, Flowpilot can prove highly beneficial on highways, where traffic is comparatively lighter than in cities, and designated lanes for vehicles are available. What’s Next? Singh said that he doesn’t have any interest in commercialising Flow Drive. “It’s not a commercial company. My team and I don’t have any interest in commercialising it. It’s open-source software, given back to the community,” he said, adding that next he plans to open something in the robotics space, focusing on automation, manufacturing, or warehousing because a lot of the underlying tech is the same. Regarding Flow Drive, he said it will remain alive. “We will be supporting people, helping the community build more stuff on top of it, but it runs on donations, and that’s how it will always be,” he concluded.","excerpt":"Mankaran Singh is working on his passion project Flow Drive, the think-tank behind open-source driver assistance system, Flowpilot.","categories":["AI Features"],"tags":[],"author_name":"Siddharth Jindal","publish_date":"2024-01-03T18:11:35","publication_year":"2024","word_count":959,"keywords":["Go","API","machine learning","programming_languages:R","AI","neural network","programming_languages:Go","automation","Aim","R"],"extracted_tech_keywords":["AI","machine learning","neural network","Aim","R","Go","API","automation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/exclusive-bengaluru-developer-makes-alto-autonomous-with-redmi\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10054719,"title":"PyTorch Releases PyTorch Live To Help Create AI-powered Experiences For Mobile Devices","content":"At its Developer Day conference, PyTorch announced the release of PyTorch Live, a set of tools designed to make AI-powered experiences for mobile devices easier. PyTorch Live offers a single programming language — JavaScript — to build apps for Android and iOS, as well as a process for preparing custom machine learning models to be used by the broader PyTorch community. PyTorch Live builds on PyTorch Mobile, a runtime that allows developers to go from training a model to deploying it while staying within the PyTorch ecosystem and the React Native library for creating visual user interfaces. PyTorch Mobile powers the on-device inference for PyTorch Live. Image Source: PyTorch The tool ships with a command-line interface (CLI) and a data processing API. The CLI enables developers to set up a mobile development environment and bootstrap mobile app projects. As for the data processing API, it prepares and integrates custom models to be used with the PyTorch Live API, which can then be built into mobile AI-powered apps for Android and iOS. In the future, the company plans to enable the community to discover and share PyTorch models and demos through PyTorch Live, as well as provide a more customizable data processing API and support machine learning domains that work with audio and video data.","excerpt":"PyTorch Live builds on PyTorch Mobile, a runtime that allows developers to go from training a model to deploying it while staying within the PyTorch ecosystem and the React Native library for creating visual user interfaces.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","Data Science","Data Scientist","Deep Learning","Machine Learning","Pytorch"],"author_name":"Victor Dey","publish_date":"2021-12-02T16:42:40","publication_year":"2021","word_count":214,"keywords":["Pytorch","Go","API","machine learning","programming_languages:R","AI","PyTorch","Machine Learning","ai_frameworks:PyTorch","JavaScript","Deep Learning","Data Science","Data Scientist","R","Java","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","machine learning","PyTorch","R","JavaScript","Go","Java","API","ai_frameworks:PyTorch","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/pytorch-releases-pytorch-live-to-help-create-ai-powered-experiences-for-mobile-devices\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10090098,"title":"Microsoft is Making Employees Lazy","content":"Gone are the days of endless surfing on the internet, researching, brainstorming ideas, creating presentations, and more. With the latest AI tools, teams are no longer confined to tedious and repetitive tasks. Instead, they can now kick back, relax, and let the machines take over. From automating workflows to improving productivity, AI-powered tools are revolutionising the way we work and allowing us more time to focus on the things that matter most. With a string of announcements over the last two weeks, Microsoft’s undying focus on enterprise management is growing. In healthcare, Microsoft along with Nuance launched DAX Express, an AI-automated clinical documentation application which employs GPT-4. Microsoft then launches GPT-4 in Azure OpenAI. The company also announced Bing Image Creator, an AI visual tool to Bing. Alongside these announcements, Microsoft also announced GitHub CopilotX, which brings GPT-4 capabilities to GitHub. Microsoft accelerated its progress in the workspace sector with the announcement of Copilot for Microsoft 365 suite and Microsoft Loop. https:\/\/twitter.com\/MicrosoftLoop\/status\/1638559915659739139 Recently, Microsoft introduced Microsoft Loop, a workspace application for managing tasks and projects and integrating Office Products, and opened this for public preview. This development comes after it had made initial announcements of this product in November 2021. Microsoft Loop allows users to collaborate and work together on multiple products such as docs, presentations, excel, videos, planners, and many more in one place. Furthermore, Microsoft will also be integrating its AI Copilot system into Loop. With Copilot, LLMs will be used to boost the functions of productivity tools. With a wide range of Microsoft Office products, Loop enables a multi-faceted approach where all products can be used at a single destination. For example, if any content on a Loop page is converted into a component, it can then be included in any of the Microsoft products including 365, Teams Chat, Outlook, Whiteboard, and more. Source: Microsoft Loop Integrating components like projects, pages, and dashboard-like workspaces, Microsoft Loop works on a real-time basis. Up to 50 users can edit a workspace simultaneously. The application can also search and recommend relevant documents. Like a dynamic workspace, Loop can customise and suggest relevant additions to the workspace while also helping with organising content into pages. But, one question remains—does this propel Microsoft past all of its competitors to gain the numero uno position when it comes to workspace applications for enterprises? If so, what is the big change it is going to bring? Not the Only Player However, Microsoft is not the only player that has focused on workspace management applications. Its arch rival, Google Workspace, has also been working on continuous improvements. Looking at the editing capabilities, up to 50 people can edit a Google docs sheet and up to 200 users can view the file. Google recently announced that generative AI will be integrated onto Workspaces. This will help automate tasks and assist with writing job descriptions, resumes, articles, and more. It can also summarise to give full project briefs solely based on email conversations. However, Google Workspace does not offer a product where all of its workspace products are integrated on one platform, unlike Loop. As a single product, Loop can be considered as the closest competitor to Notion. I doubt MS will make a dent in notion as far as personal use goes.If they are cheaper or free for teams, and get some better integration going they may have a chance.But MS usually rips stuff off and then doesn't follow through ..so I wouldn't hold my breath too much— Robert Hoffmann (@itechnologynet) March 22, 2023 Notion is a productivity workspace management tool that encompasses all the functions that Loop has. Released in 2015, Notion recorded over 30 million users as of December 2022. Microsoft Office has over 345 million users. The biggest advantage Loop has over Notion is that, as Microsoft Office subscribers, Loop is offered as a part of its suite and all the applications can be used in tandem. Whereas, Notion requires a separate subscription and plugins to integrate with other applications. Zoho Workplace is another player that has been providing applications for enterprises. They are even working towards integrating generative AI, as an effort to simplify work and enhance employees’ productivity. Where is it headed? Taking parallels to how workspace is evolving on the developer front—where Copilot’s success has led to the rise of large language models in different industries allowing developers to be aided with tools that help write, code and design—workspace management is also seeing a change. Introduction of AI in workspace management tools and workspace, in general, is shifting towards making tasks simpler and faster. Recent examples of how companies like Zoho and Rezolve.ai are implementing LLM-based models in their HR workspace shows how companies are shifting towards prioritising tasks that require attention and simplifying\/automating tasks that are increasingly time consuming.","excerpt":"With a wide range of Microsoft Office products, Loop enables a multi-faceted approach where all products can be used at a single destination.","categories":["Global Tech"],"tags":["Copilot","excel","Generative AI","Google","GPT-4","Microsoft"],"author_name":"Vandana Nair","publish_date":"2023-03-27T17:26:31","publication_year":"2023","word_count":800,"keywords":["Go","excel","OpenAI","AI","R","GPT-4","Git","BERT","GPT","generative AI","Google","GitHub","Generative AI","Copilot","Azure","Microsoft"],"extracted_tech_keywords":["AI","generative AI","OpenAI","Azure","R","Go","Git","GitHub","BERT","GPT"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/microsoft-is-making-employees-lazy\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10046057,"title":"Top Weekly News: OpenAI Codex To Biggest Cryptocurrency Heist","content":"Earlier this week, OpenAI announced the release of OpenAI Codex through an API in private beta. An AI system, Codex, translates natural language to code, meaning it can interpret and execute commands given in plain English. This makes it possible to build a natural language interface for existing apps. Codex is designed to assist and speed up programming work for both professionals and coding amateurs. OpenAI’s demo video, featuring founders Ilya Sutskever and Greg Brockman, demonstrated how Codex can build simple websites and rudimentary games. Today OpenAI launched Codex, which is an API that uses AI to write code from natural language. There's a demo video here: https:\/\/t.co\/hOOfo3MZlu— Sam Altman (@sama) August 10, 2021 OpenAI’s AI pair programmer Copilot is based on Codex. Copilot takes context from the code being written and suggests whole lines and functions. OpenAI claims that the latest version of Codex is more advanced and can create and complete a chunk of code. Codex is a descendant of OpenAI’s language model GPT-3 that can work with written words uniquely due to its sizable training data. The data contains both natural language and billions of lines of source code from publicly available sources, including code in public GitHub repositories. It is still in its early stages, but the developers behind Codex warn that since its responses are based on training data from the internet, Codex can be prompted to generate racist, denigratory, and harmful outputs in code comments. Biggest Cryptocurrency Heist More than $600 million worth of crypto was stolen recently in what is touted as one of the biggest cryptocurrency heists. Blockchain site Poly Network said that the hackers exploited a vulnerability in its system and swindled off thousands of digital tokens like Ether. This hack is on par with recent breaches at exchanges such as Coincheck and Mt Gox in terms of scale. The company posted on Twitter, urging the thieves to establish communication and return the hacked assets. Only hours after the hack, the attacker started returning the funds to the online wallets controlled by Poly, starting with small amounts and then in millions. pic.twitter.com\/Yzw4oDenjC— Poly Network (@PolyNetwork2) August 10, 2021 Decentralised finance aims to reproduce traditional financial products like loans and trading without the involvement of middlemen. It is one of the fastest-growing spaces within the crypto industry. Although it has attracted heavy investment, running into billions of dollars, the DeFi space is prone to hacks and scams. One recent example includes a token by billionaire investor Mark Cuban that dropped from $60 to several thousandths of a cent due to an alleged bank run. Top Honour for NVIDIA’s Huang Intel co-founder Rober Noyce is considered one of the pioneers of the semiconductor industry, often referred to as the ‘mayor of Silicon Valley’. In his honour, the Noyce Award is handed over to people with demonstrated excellence in the semiconductor field. This year, this prestigious award would be bestowed upon NVIDIA CEO Jensen Huang, the Semiconductor Industry Association announced. The ceremony will be held on November 18. “Jensen’s accomplishments have fueled countless innovations — from gaming to scientific computing to self-driving cars — and he continues to advance technologies that will transform our industry and the world,” said John Neuffer, SIA president and CEO. Huang founded NVIDIA in 1993 and, since then, has been a recipient of several awards and recognitions, including the IEEE Founder’s Medal and Dr Morris Chang Exemplary Leadership Award. In 2019, Harvard Business Review named him the best performing CEO over the lifetime of their tenure; before that, in 2017, Huang was named Fortune’s Businessperson of the Year. Last year, the Noyce award was presented to Dr Lisa Su, CEO of AMD. Four Decades to IBM’s First PC Forty years back, on August 12, IBM released its first personal computer, Model 5150 or just IBM PC. Notoriously known for being highly bureaucratic and unable to keep up with the fast-growing market for portable computers, IBM foray into the PC market was nothing short of a watershed moment. Created by the team of engineers and designers directed by Don Estridge, IBM PC soon became a massive influence in the industry. It was based on open architecture. The specifications of the IBM PC became standard in computer design, and the only significant competition it faced was from the Apple Macintosh line. It would not be an exaggeration to say that modern personal computers are distant descendants of the IBM PC. IBM PC’s innovation, quality, and affordability were a big frenzy among the masses when it was first introduced. Bezos-Murthy end Joint India Venture In India, Amazon has come under fire for alleged dominance through business practices that small vendors and retailers deem unfair and illegal. As a result, an antitrust investigation is undergoing against Amazon and other giants, including Flipkart. Amidst this antitrust probe, Amazon has decided to disband a controversial joint venture in India with Infosys co-founder Narayana Murthy. The seven-year-old joint venture called Prione Business Services Pvt Ltd. will cease operations from mid-2022. In a joint statement, the parties said they have mutually decided to discontinue their joint venture beyond the end of this current term. Discontinuation of this venture could mean a possible setback for Amazon as India’s online market is projected to surge to $1 trillion.","excerpt":"Forty years back, on August 12, IBM released its first personal computer, Model 5150 or just IBM PC.","categories":["AI News"],"tags":["big data quality","NVIDIA","OpenAI Codex"],"author_name":"Shraddha Goled","publish_date":"2021-08-15T10:00:00","publication_year":"2021","word_count":879,"keywords":["Go","big data quality","TPU","Rust","OpenAI","AI","Git","Ray","Aim","OpenAI Codex","NVIDIA","GitHub","R"],"extracted_tech_keywords":["AI","OpenAI","Aim","Ray","TPU","R","Go","Rust","Git","GitHub"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/top-weekly-news-openai-codex-to-biggest-cryptocurrency-heist\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":5667,"title":"Analytics Education for Senior Professionals: A Unique Programme from ISB","content":"Certainly, there has been a proliferation of analytics education in India. The demand of data scientist has rightly created a horde of analytics courses in the country, some of them with world-class quality. The Industry is still trying to cope up with the demand of data scientists and analytics professionals. As a result, we have seen a fair amount of courses in Analytics targeting young professionals. There are varieties of courses available ranging from online courses, yearlong courses & distance programs. It seems evident that these courses are going to fulfill most of the demand of analytics professionals in coming years. Yet, there seems to be a glaring void still to be tapped. Most of the analytics courses are geared towards professionals with 1-10 years of experience. This means we would be fulfilling most analytics demand in the lower section of experience. This section is definitely a broad section and is expected to be more hands-on. Yet, the existing analytics courses do not fulfill the demand for relatively senior professionals for analytics. The industry now needs to take a step forward and enable the senior level professionals to leverage the power of analytics to the full. What the industry needs is a right set of mix of senior as well as junior level of analytics professionals. The senior professionals need to keep pace with the wave of upcoming Analytics professionals. While the new comers have the advantage of right set of analytics skills, the senior professionals embody years of experience and domain knowledge. Both of them hand in hand can bring the kind of results that would help an organization to leap forward. When we look for analytics courses targeting senior professionals, it’s difficult to find any. What needs to be kept in mind is that these professionals would have to juggle between their existing tasks and thirst for enhancing their skill set. There are several reasons a course designed for senior professionals would be different from regular analytics course. Firstly, these senior professionals would be involved in some serious decision-making as manufacturers and consumers of analytics. We are just not talking about analytics professionals with charters to deliver analytics solutions but also leaders of different business units that would heavily rely on results from analytics. As analytics gain traction in all business units, coaching leaders of marketing, supply chain, finance, and strategy over benefits of analytics is the key for an analytics driven organization. Secondly, while an awareness and familiarity of analytics technique is needed, these leaders have reached a stage where technical hands-on learning would have to be constrained. I receive a lot of queries from senior professionals with more than 15 years of experience on the right course they should take to learn analytics. When there is none for them. They have the thirst for knowledge but are restrained on two aspects – They need a bird’s eye view of the analytics, enough for right decision making, and a busy schedule would mean a packed course structure that would provide this understanding in a concise fashion. ISB Hyderabad has just come up with a program that would help senior professionals establish a foot holds in analytics. This new program titled “Senior Executive Programme for Business Analytics: Applications & Advocacy” aims at senior professionals with profiles ranging from Brand Manager, Sales Director, BI director & E commerce Operations etc. Programme Highlights: This programme is in collaboration with Indian School of Business and McCombs, UT in USA. The programme would be spread over 2.5 weeks. 1 Week in ISB & 1 week in McCombs US & 3 days of SAS visual Analytics. Total time from start to finish is four months. This programme targets professionals and leaders from the Analytics domain across industries & functions as well as senior leaders in a company who are involved in decision-making. This unique course was created in response to the growing market demand for senior professionals who are able to extract business insights from Big Data. The program will engage faculty from both the McCombs School of Business and Indian School of Business, to educate students in business analytics. Launching this summer, the program will integrate the core knowledge of business with other areas, such as operations research, statistics, machine learning, information systems and computer science. McCombs School of Business (University of Texas at Austin) is a leading B-School in USA with a prominent analytics program. The collaboration between McCombs and ISB aims to bring in synergies of analytics programs at both the institutions into one concise coursework for senior professionals, which is expected to be highly impactful. Evaluation of Content The course content touches upon various aspects of analytics, big data and visualization. On analytics, the course covers Regression and time series, survival analysis & missing value patterns, Classification and prediction, Principal components, Clustering patterns etc. You would also gain insights into Map reduce programming model to crunch and analyze big data & big data text analytics for understanding and    mining large volumes of unstructured text data. There are also domain based analytics themes including marketing analytics and supply chain analytics. Also, participants would learn how to explore and discover data and build insightful dashboards and reports using SAS Visual Analytics. Participants will learn how to leverage self-service analytics features and capabilities of SAS VA. In a nutshell, wholesome course content with a birds eye view to the analytics stream. Here are more details on the program: Programme fee INR 6,95,000 plus tax, fee includes food and accommodation at both the campuses (ISB & McCombs) Write at cee@isb.edu for more details. Call on +91 40 2318 7516 \/ 9440121755","excerpt":"Certainly, there has been a proliferation of analytics education in India. The demand of data scientist has rightly created a horde of analytics courses in the country, some of them with world-class quality. The Industry is still trying to cope up with the demand of data scientists and analytics professionals. As a result, we have […]","categories":["AI Trends"],"tags":[],"author_name":"AIM Media House","publish_date":"2014-04-26T11:09:53","publication_year":"2014","word_count":938,"keywords":["big data","Go","machine learning","programming_languages:R","AI","RAG","Aim","analytics","GAN","R"],"extracted_tech_keywords":["AI","machine learning","analytics","Aim","RAG","R","Go","big data","GAN","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/analytics-education-for-senior-professionals-a-unique-programme-from-isb\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10013396,"title":"Hands-on Guide To Synthetic Image Generation With Flip","content":"Your neural network is only as good as the data you feed it because it trains on millions of parameters to produce an effective performance but in case you are working on something new and and you’re stuck with a limited dataset and online available datasets are limited you can use augmentation techniques in to train your neural network with diversity in relevant data. Often deep learning engineers have to deal with insufficient data that can create problems like increased variance in their models that can lead to overfitting and limit the experimentation with the dataset. Synthetic data generation is just artificial generated data in order to overcome a fixed set of data availability by use of algorithms and programming.While dealing with datasets containing images. Flip-data which is an open source python library can help us overcome the issue of limited image datasets and help us generate synthetic images by taking images and separating them into objects and backgrounds and repositioning the object on top of background through data augmentation techniques and have some flexibility in the dataset. Dependencies Flip requires: Python (>= 3.7), Opencv (>= 4.3.0), Numpy (>= 1.19.1) What is data augmentation? Data augmentation uses simple strategic techniques like flipping,resizing, rotating etc. on image objects with respect to the background to produce diversity in the available images.It is most commonly used to train and improve neural networks by using artificial data produced from existing data. DATA AUGMENTATION TECHNIQUES USED Below, are the basic operations performed on images to produce variations. FLIP We can flip the images with respect to the axis randomly or in the x and y plane. flip.transformers.data_augmentation.Flip(mode=’random’) ROTATE We can rotate the images by setting the min-max angles and get desired augmented results. flip.transformers.data_augmentation.Rotate(mode=’random’, min=60, max=180) RANDOM SIZE We can resize the images by setting the min-max width of the image according to our desired results. flip.transformers.data_augmentation.RandomResize(mode='symmetric_w',relation='parent',  w_percentage_min=0.3, w_percentage_max=0.5,) TRANSFORM After completing the above steps we can determine the min-max values of height and width to place the object respectively with the background and apply random composition or setting the percentage value. transform = flip.transformers.Compose( [ flip.transformers.ApplyToObjects(transform_objects), flip.transformers.domain_randomization.ObjectsRandomPosition( x_min=0.1, y_min=0.5, x_max=0.8, y_max=0.8, mode='percentage' ), flip.transformers.domain_randomization.Draw(), flip.transformers.io.SaveImage(OUTPUT_DIR, name) ] ) The main transformers are: Transformer, Compose and ApplyToObjects By the way, all Transformers will be executed over objects of class Element and will return a new transformed Element. At last, we can set a number of samples, objects and define input output paths for respective backgrounds and objects and save them to result images. generate_data(n_samples=1, n_objects=4, backgrounds_pattern=\"backgrounds\/*\", objects_pattern=\"objects\/*\", output_dir=\"synthetic_images\", show=False) BACKGROUND You can choose the desired background or use a single photo to crop the object from the background (crop images github link), we will use a jpg format image for background. OBJECT The object image format we are using is png. RESULTS CONCLUSION We have successfully generated synthetic 2D images from a single background and object image and we can also generate thousands of new 2D images from a small batch of objects and backgrounds as per our requirements.","excerpt":"2D synthetic image data generation.","categories":["Deep Tech"],"tags":["computer vision dataset","Deep Learning","image data augmentation","Python for Data Science"],"author_name":"Neelesh Sinha","publish_date":"2020-12-07T14:00:20","publication_year":"2020","word_count":504,"keywords":["computer vision dataset","NumPy","TPU","AI","neural network","ML","image data augmentation","Transformers","OpenCV","Python","deep learning","Python for Data Science","Deep Learning","R"],"extracted_tech_keywords":["AI","ML","deep learning","neural network","Transformers","OpenCV","NumPy","TPU","Python","R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/hands-on-guide-to-synthetic-image-generation-with-flip\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10168204,"title":"Microsoft Unveils 1-Bit Compact LLM that Runs on CPUs","content":"Microsoft Research has introduced BitNet b1.58 2B4T, a new 2-billion parameter language model that uses only 1.58 bits per weight instead of the usual 16 or 32. Despite its compact size, it matches the performance of full-precision models and runs efficiently on both GPUs and CPUs. The model was trained on a large dataset containing 4 trillion tokens and performs well across a wide range of tasks, including language understanding, math, coding, and conversation. Microsoft has released the model weights on Hugging Face, along with open-source code for running it. In the technical report, Microsoft said that “BitNet b1.58 2B4T achieves performance on par with leading open-weight, full-precision LLMs of similar size, while offering significant advantages in computational efficiency, including substantially reduced memory footprint, energy consumption, and decoding latency.” The model’s architecture is “derived from the standard Transformer model… incorporating significant modifications based on the BitNet framework”. The central innovation is “replacing the standard full-precision linear layers with custom BitLinear layers”, where “model weights are quantised to 1.58 bits during the forward pass”. This quantisation uses an “absolute mean (absmean) quantisation scheme, which maps weights to ternary values {-1, 0, +1}.” Activations are quantised to 8-bit integers with an “absolute maximum (absmax) quantisation strategy, applied per token”. Subln normalisation is incorporated to further enhance training stability. The feed-forward network (FFN) sub-layers employ squared ReLU (ReLU²) activation. Rotary Position Embeddings (RoPE) inject positional information. Consistent with architectures like LLaMA, all bias terms are removed from the linear layers and normalisation layers. The tokeniser developed for LLaMA 3 implements a byte-level Byte-Pair Encoding (BPE) scheme with a vocabulary size of 128,256 tokens. The training process for BitNet b1.58 2B4T consists of three phases, pre-training, supervised fine-tuning (SFT), and direct preference optimisation (DPO). BitNet b1.58 2B4T demonstrates that it’s possible to dramatically reduce the computational requirements of large language models without giving up performance. With its compact architecture and competitive results, it represents a meaningful step forward in making AI models more efficient and accessible.","excerpt":"Microsoft has released the model weights on Hugging Face, along with open-source code for running it.","categories":["AI News"],"tags":["Microsoft"],"author_name":"Siddharth Jindal","publish_date":"2025-04-17T15:50:13","publication_year":"2025","word_count":332,"keywords":["Hugging Face","programming_languages:R","AI","innovation","ai_frameworks:Hugging Face","llm_models:Llama","R","Microsoft"],"extracted_tech_keywords":["AI","Hugging Face","R","innovation","llm_models:Llama","ai_frameworks:Hugging Face","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/microsoft-unveils-1-bit-compact-llm-that-runs-on-cpus\/","complexity_score":4,"technical_depth":7,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":24308,"title":"Top Factors to Consider Before Choosing A Vendor in 2024","content":"Artificial intelligence has progressed to a point where it can be used to solve more than just a set of problems. But even as barriers for its adoption are lowered, thanks to the availability of ML tools and technologies, enterprises are still grappling with applying AI applications for critical tasks. Even though AI has received enormous attention in India, organisations are still looking to understand its true potential and how to realise the return on investment from the technology. Also, another big question the C-suite wishes to understand is the maturity level of the technology in the next two years. The cost-prohibitive nature of technology and vendor lock-in that comes in with cloud companies such as AWS, Google and Microsoft, among others, who offer machine-learning-as-a-service (MLaaS), also acts as a deterrent for small and medium businesses. Many startup founders offering industry-specific vertical solutions have often claimed about clients griping about the scalability, since big vendors charge per API and few even worry about their data leaving the premises. Despite the boom around ML in India, it is an expensive proposition and hard to implement, since there is also a dearth of ML engineers. Today, CIOs are grappling with the same question: With everyone deploying AI, are we being left behind? Well, the enterprise AI market is getting highly competitive with startups and boutique vendors offering solutions targeted at specific industries to perform niche tasks which big tech vendors like Google, Amazon, Microsoft and IBM are not able to deliver. For example, Senseforth’s enterprise AI bots, built on proprietary technology directly compete with big tech vendors like AWS and Microsoft. This Silicon Valley and Bengaluru-based startup has the highest number of active bots in deployment — 30 — across a range of industries. According to a research report, by the year 2019, AI and ML startups are projected to drive AI capabilities forward, disrupting the market with niche business solutions. And large AI vendors are adjusting their pricing strategies to compete with smaller competitors that are threatening to upend the Enterprise AI market. Factors To Keep In Mind Before Onboarding Enterprise AI Vendors The global enterprise AI market dominated by SAS, Microsoft, AWS, Intel, HPE and Wipro among others is estimated to reach $6 billion by 2023. According to a report by Market Research Future, India is advanced in software development and solutions and start-ups are also building their base in AI technology providing services and solutions to different industries. Despite the vendor boom, organisations can be overwhelmed by the number of vendors, big and small and AI solutions on the market. For example, an IDC whitepaper about AI adoption in India discusses how few vendors may be able provide the appropriate AI software solutions but may lack capabilities to impart training to its client’s employees. Another major roadblock to adoption is a lack of strong AI advisory from top consulting firms on implementation and its capabilities. Before Organisations Go Shopping For Vendors, Here Are A Few In-house Competencies They Should Beef Up Data Collection: For ML to work, organisations need to large quantity of accurate and clean data and most organisations cite combing siloed data as the biggest hurdle. For example, Manipal Hospitals which leverages IBM Watson for Oncology reportedly draws data from more than 300 medical journals, 200 textbooks, and nearly 15 million pages of text. Identifying The Right Use Cases For AI: For example, IT processes and sales and marketing functions are some of the prominent functions that are being automated and are improving organisational efficiency. Right Leadership: Organisations who are willing to explore AI implementation need senior management buy-in for enterprise AI adoption. While AI is seen as a magic bullet for solving critical problems, business leaders should also ask the right questions on how to align it with business objectives to reap tangible benefits. Building In-House Talent: For organisations that are still mulling the implementation of AI, talent issues are also a big priority and the shortage of skilled professionals to run day-to-day operations also act as a barrier in adoption. Here Are A Few Points Tech Buyers Should Consider Vertical Generic\/Cross Domain Solution: Now startups usually offer vertical solutions while big tech enterprises offer cross domain or generic solutions. Most organisations do not like dealing with multiple vendors and would either opt for a generic solution while a few, in a bid to avoid vendor lock-in opt for startups that offer industry-specific enterprise AI solutions. Choosing the right vendor or solution can help organisations avoid huge costs and reap tangible benefits. For example, Texas-based MD Anderson, the cancer research institute grabbed headlines last year for severing the contract with IBM Watson as it failed to realise tangible benefits. Alignment With Business Objectives: This is the most important criteria for tech buyers before onboarding the AI vendor. To avoid risk of failure, organisations should identify the right use case and start by deploying it on small scale projects. Top management should also define the key metrics they wish to achieve from the AI solution. Before the AI implementation, senior management should also put a framework for data collection and also take into consideration governance and regulatory implications. Strong AI Advisory: Most organisations also look for strong AI advisory during and after implementation as they need walkthroughs through each and every deployment. Also, organisations want support on training their staff on handling day-to-day functionalities as they may lack data scientists or ML professionals to handle the deployment. Cost Competitiveness: While on one hand AI solutions promise real benefits, we often hear C-suite talking about the cost of solution being too high. Startups which offer competitive pricing solutions often score over their enterprise counterparts in this area.","excerpt":"Artificial intelligence has progressed to a point where it can be used to solve more than just a set of problems. But even as barriers for its adoption are lowered, thanks to the availability of ML tools and technologies, enterprises are still grappling with applying AI applications for critical tasks. Even though AI has received […]","categories":["AI Trends"],"tags":["AI Companies"],"author_name":"Richa Bhatia","publish_date":"2018-05-07T06:43:34","publication_year":"2018","word_count":949,"keywords":["Go","API","artificial intelligence","AWS","AI","ML","Scala","RAG","Aim","AI Companies","R"],"extracted_tech_keywords":["AI","artificial intelligence","ML","Aim","RAG","AWS","R","Go","Scala","API"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-factors-companies-should-consider-before-choosing-a-vendor-for-ai-solutions\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10039154,"title":"Amazon Donates 100 ICU Ventilators To Fight COVID In India","content":"On Tuesday, Amazon announced that the company will 100 ICU ventilators to hospitals in India. The e-commerce giant has worked with MoHFW (Ministry of Health and Family Welfare, Government of India) to ensure that the ventilators are of acceptable technical specification and carried out its own compatibility checks to immediately fund 100 units of Medtronic’s PB980 model and bring them into India for urgent use. In an official blog post, it is stated that as a part of the company’s commitment to helping India fight the devastating second wave of COVID-19, Amazon India has procured 100 ventilators through its global resources to immediately import these into India. The post also stated that the company will work closely with Medtronic (MT) for these units to be airlifted into India and we expect the consignment to enter the country in the next two weeks. At present, Amazon India is working closely with the MoHFW appointed agencies to finalize the hospitals with the most urgent need so that Amazon can partner with MT for end-to-end delivery, installation, maintenance and training of personnel who will use these machines. Amit Agarwal, Global SVP & Country Head, Amazon India stated, “With the urgency of adding to the medical infrastructure and capacity for Indian hospitals fighting against COVID-19’s severe second spike, we decided to urgently source, import and donate 100 ventilators to hospitals to be identified by the Ministry of Health and Family Welfare (MoHFW), Government of India.” “We hugely appreciate the quick response from the MoHWF to help identify the most compatible models, expediting the shipment import into India and for coordination with agencies of MoHFW to allocate these where they are needed most. We are doing more and are committed to support our country in the fight against COVID-19,” Agarwal added.","excerpt":"On Tuesday, Amazon announced that the company will 100 ICU ventilators to hospitals in India. The e-commerce giant has worked with MoHFW (Ministry of Health and Family Welfare, Government of India) to ensure that the ventilators are of acceptable technical specification and carried out its own compatibility checks to immediately fund 100 units of Medtronic’s […]","categories":["AI News"],"tags":["Amazon","covid-19"],"author_name":"Ambika Choudhury","publish_date":"2021-04-28T18:57:50","publication_year":"2021","word_count":296,"keywords":["Go","covid-19","AI","programming_languages:R","Amazon","programming_languages:Go","R"],"extracted_tech_keywords":["AI","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/amazon-donates-100-icu-ventilators-to-fight-covid-in-india\/","complexity_score":2,"technical_depth":5,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10100379,"title":"OpenAI to Make Fine-Tuning UI Available Soon","content":"OpenAI is likely to make fine-tuning UI available in the coming months according to Logan Kilpatrick. Users will have a seamless experience through the UI, allowing them to easily view their fine-tunes and create them effortlessly using the same interface, according to Kilpatrick. https:\/\/twitter.com\/OfficialLoganK\/status\/1704181284036300970 Furthermore, OpenAI has increased the concurrent training limit from 1 to 3, enabling users to fine-tune multiple models simultaneously. Currently, OpenAI permits developers to customize only a part of the model name using a suffix. However, in the future, developers might have the opportunity to customize the entire model name, as indicated by Kilpatrick. Many developers around the world are hoping that OpenAI might make fine-tuning available at their inaugural developers’ conference ‘OpenAI DevDay’  which is taking place on November 6, 2023 in San Francisco. There has been a lot of anticipation about what the company is going to announce, though CEO Sam Altman has said that there is going to be no announcement about GPT-5. Some predictions for OpenAI’s developer day on November 6th…– Meaningful GPT-4 cost reduction– Fine-tuning for GPT-4– UI for fine-tuning– Multimodal GPT-4 goes live– DALL-E 3– ChatGPT API (rethinking of plugins)I bet I hit on at least 3.— Mckay Wrigley (@mckaywrigley) September 15, 2023 Recently, OpenAI also silently unveiled “gpt-3.5-turbo-instruct,” a new instruction language model designed for giving specific instructions efficiently, similar to the chat-focused GPT-3.5 Turbo. This new model will replace existing Instruct models and certain text-based models. It maintains the same cost and performance as other GPT-3.5 models within a 4K context window, using training data up to September 2021.","excerpt":"OpenAI has increased the concurrent training limit from 1 to 3","categories":["AI News"],"tags":[],"author_name":"Siddharth Jindal","publish_date":"2023-09-20T18:56:47","publication_year":"2023","word_count":261,"keywords":["Go","ChatGPT","API","DALL-E","OpenAI","GPT-5","AI","ML","GPT","R"],"extracted_tech_keywords":["AI","ML","GPT-5","ChatGPT","OpenAI","R","Go","API","GPT","DALL-E"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/openai-to-make-fine-tuning-ui-available-soon\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":35587,"title":"10 Resources To Learn Genetic Programming","content":"Genetic programming is one of the most interesting aspects of machine learning and AI, where computer programs are encoded as a set of genes that are then modified (evolved) using an evolutionary algorithm. It is picking up as one of the most sought after research domains in AI where data scientists use genetic algorithms to evaluate genetic constituency. While research is still underway in this area, many researchers and professionals are now looking to dig into the subject. To help those professionals starting out in the field and for those looking to gain additional knowledge, we have listed 10 sources including, books, ebooks, videos and tutorials that will help to know more about genetic programming. The resources are listed in no particular order. 5 Languages To Use For Genetic ProgrammingeBook – 20 Interviews of Indian Analytics Leaders (Free Download)Top 10 Free Ebooks To Learn Data ScienceHow To Use Genetic Algorithms As A Tool For Feature Selection In Machine Learning10 Free eBooks Beginners Should Read Before Diving Into Data Science 1| Introduction to Genetic Algorithms by Melanie Mitchell (Book): It is one of the most read books on genetic algorithms and covers in-depth details about the subject such as background, history, motivation along with informative examples that makes it easy to understand the concepts. It also discusses use cases of genetic algorithms in scientific models, which is a good read for anyone wanting to know more about the area. It also gives an insight into some of the most interesting research in the field enabling readers to experiment and implement genetic algorithms of their own. The book can be bought here. 2| Genetic Algorithms in search, optimisation and machine learning by David E Goldberg (Book): Authored by David E. Goldberg, the book is a comprehensive text for students pursuing Computer Science Engineering, Electrical Engineering and Electronics Engineering. The book is also useful for practitioners who are looking to learn more about the field. It has procedures and applications explained in detail where the author has brought together computer techniques, mathematical tools and research results giving a complete insight into the subject. The free ebook can be downloaded here. The book can be bought here. 3| Colorado State University tutorial on Genetic Algorithms by Darell Whitley (Online tutorial): This tutorial covers the canonical genetic algorithm along with experimental forms of the genetic algorithm, including the parallel island model and parallel cellular genetic algorithm. It illustrates a genetic search with hyperplane sampling. It is explained by Darrell Whitley from the computer science department of Colorado State University and is explained in detail with examples, illustrations and use cases. The PDF version of the tutorial can be downloaded here. 4| A Field Guide to Genetic Programming by Riccardo Poli Poli,  William B. Langdon, Nicholas Freitag McPhee (Book): One of the most hands-on guides on the subject, the book has received good reviews from the data science community. The book begins by explaining the basics of genetic programming. The subject has been explained with stress on use cases as genetic programming has generated a plethora of human-competitive results including novel scientific discoveries and patentable inventions. The three researchers have brought a unique perspective of this technique on the bok. The free ebook can be downloaded here. The book can be bought here. 5| Introduction to Genetic Algorithms: Theory and Applications by Udemy (Video): In this video tutorial by Udemy, you can learn the main mechanisms of the genetic algorithm as a heuristic artificial intelligence search or optimisation in Matlab. It covers tutorials on using a genetic algorithm to solve optimisation problems, analysing the performance, modifying or improving genetic algorithms and more. It covers the most fundamental aspects of the subject and is one of the best sources if you are new to the field. The video can be watched here. 6| MIT Lecture on Learning Genetic Algorithm by Patrick H. Winston (Video): Conducted by Patrick H. Winston, an American computer scientist, and professor at the Massachusetts Institute of Technology. This lecture explores the genetic algorithm at a conceptual level. The instructor has tried to consider three approaches on how a population evolves towards desirable traits, ending with ranks of both fitness and diversity. He has discussed it with use cases and live examples. The video can be downloaded here. 7| Clever Algorithms: Nature-Inspired Programming Recipes by Jason Brownlee (Book): It covers evolutionary algorithms in detail which is concerned with computational methods inspired by the process and mechanisms of biological evolution. It covers extensively the genetic algorithm, genetic programming, evolution strategies, evolutionary programming, differential evolution and more. The book can be downloaded here. 8| The Algorithm Design Manual by Steve Skiena (Book): This book covers an extensive section on genetic algorithms and other interesting heuristics for solving various types of problems. It deals with some key algorithms while drawing the author’s own real-world experiences on design and analysis. The first half of the book is a general guide to techniques for the design and analysis of computer algorithms while the second part includes a catalogue of the 75 most important algorithmic problems. The book can be bought here. 9| Collective Intelligence by OReilly by Toby Segaran (Book): Programming Collective Intelligence takes you into the world of machine learning and statistics and explains how to draw conclusions about user experience, marketing, personal tastes, and human behaviour in general,  all from information that you and others collect every day. This book has a chapter on the genetic algorithm that has been covered with illustrating examples. The book can be bought here. 10| Practical Genetic Algorithms by Randy L. Haupt and Sue Ellen Haupt (Book): This book stresses genetic algorithms with an emphasis on practical applications. It provides numerous practical example problems and contains over 80 illustrations including figures, tables, a list of genetic algorithm routines in pseudocode, and more. The book can be bought here.","excerpt":"To help professionals starting out or looking to gain additional knowledge, we have listed 10 sources including, books, ebooks, videos and tutorials.","categories":["AI Trends"],"tags":[],"author_name":"Srishti Deoras","publish_date":"2019-02-28T12:24:44","publication_year":"2019","word_count":977,"keywords":["data science","Go","machine learning","artificial intelligence","programming_languages:R","AI","programming_languages:Go","collective intelligence","analytics","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","data science","analytics","R","Go","collective intelligence","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/10-resources-to-learn-genetic-programming\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10011511,"title":"Current State Of Machine Learning in Compilers &#038; Its Future","content":"The job of compilers is to translate programming languages written by humans into binary executable by computer hardware. Compilers run on large complex, heterogeneous, non-deterministic, and constantly changing systems. Optimising compilers is difficult because the number of possible optimisations is huge. Designing heuristics that take all of these considerations into account ultimately becomes hard. As a result, many compiler optimisations are out of date or poorly tuned. One of the key challenges is to select the right code transformation for a given program, which requires effective evaluation of the quality of a possible compilation option. For instance, knowing how a code transformation will affect eventual performance is one such option. A decade ago, machine learning was introduced to enhance the automation of space optimisation. This enabled compiler writers to develop without having to worry about architectures or program specifics. Algorithms were capable of learning heuristics from the results of previous searches. This helped optimise the process in the next run in a single shot manner. Machine learning-based compilation is now a research area, and over the last decade, this field has generated a large amount of academic interest. To know more about the current state of ML and its implications for compilers, researchers from the University of Edinburgh and Facebook AI collaborated to survey the role of machine learning with regards to compilers. Current State Of ML For Compilers (Source: Paper by Zheng Wang and Michael O’Boyle) Compilers have two jobs – translation and optimisation. Programs are first translated into binary correctly. Next, they have to find the most efficient translation possible. So, the vast majority of research and engineering practices is focussed on optimisation. The idea behind the optimisation of compilers is to minimise or maximise some attributes of an executable computer program such as minimising a program’s execution time, memory requirement, and power consumption. Machines can be made to learn how to optimise a compiler to make it run faster. Machine learning is ideally suited to making any code optimisation decision where the performance impact depends on the underlying platform. The advantage of the ML-based approach, stated the researchers, Wang and Boyle, is that the entire process of building the model can be easily repeated whenever the compiler needs to target a new hardware architecture, operating system or application domain. There were a few implementations of deep learning for compilers. One of them was a deep neural network used to directly predict the right heuristic value should be for some optimisations in OpenCL. They used LSTMs for this, which enabled them to somewhat understand the structure of the program, such as whether a variable has been declared in the past. The results improved over prior, hand-built features. This research was followed by the incorporation of graph-based representations. Researchers thought that it would suit programming languages better. The instructions of a program can be represented as the edges of the graph, depicting the relationship between variables. Data-flow is fundamental to practically every optimisation in the compiler. Models should be able to reason about complex programs. For this to happen, wrote the researchers, required better representations and graph RNNs that match the data-flow pattern. In this work, the authors address the question: where is this field going? What do we need to do in the coming years? The optimisation process is hard, and it is only going to get harder still. What Does The Future Hold (Source: Paper by Zheng Wang and Michael O’Boyle) Going forward, the researchers are betting big on reinforcement learning for better compilers. The above illustration is an architecture for the compiler that is governed by the principles of reinforcement learning. Here, RL systems are tasked with looking at optimisations on other functions, basic blocks and individual instructions. For each context, the compiler determines a set of applicable and valid transformations which it passes to an RL agent to make its choice. The agent chooses the next action based on the current state of the program and a history of actions and states it has seen. The RL system will take actions that increase the likely future reward, which will be the speedup found by applying the action sequence to the code. However, the authors also admit that this problem is larger than those to which reinforcement learning is typically applied. The state space is huge, and so is the action space. Evaluating the reward can be time-consuming, and the program must be compiled to a binary and executed with representative inputs enough time to give statistically sound timings. Modern compilers are multi-million line pieces of software that can take years to master. To leverage ML for efficient compilers, the researchers have few suggestions: Compiler writers to build modular compilers that support iterative compilation and machine learning from the ground up.Machine learning researchers should invent models that are suited to the recurrent, flow-based nature of programs.Enable every optimisation choice to be exposed through discoverable APIs with which iterative search and machine learning can interact. Read more about the past, present and future of ML for compilers here.","excerpt":"The job of compilers is to translate programming languages written by humans into binary executable by computer hardware. Compilers run on large complex, heterogeneous, non-deterministic, and constantly changing systems. Optimising compilers is difficult because the number of possible optimisations is huge. Designing heuristics that take all of these considerations into account ultimately becomes hard. As […]","categories":["Deep Tech"],"tags":["Compiler","Machine Learning","optimisation algorithms"],"author_name":"Ram Sagar","publish_date":"2020-11-11T18:00:44","publication_year":"2020","word_count":841,"keywords":["Go","API","machine learning","AI","neural network","ML","Compiler","Machine Learning","RAG","optimisation algorithms","deep learning","RNN","R"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","neural network","RAG","R","Go","API","RNN"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/state-of-machine-learning-compilers-optimisation\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10161660,"title":"Samsung Partners with Eka Care to Launch Health Records Feature","content":"Samsung has partnered with health tech company Eka Care to introduce a ‘Health Records’ feature on the Samsung Health app in India. This initiative aligns with the Indian government’s Ayushman Bharat Digital Mission (ABDM) and aims to streamline healthcare management by enabling users to digitally access and manage their medical records. Users can create an Ayushman Bharat Health Account (ABHA) directly within the Samsung Health app using their Aadhaar or mobile numbers. Once registered, they can securely access their medical history, including prescriptions, lab results, and hospital visits, all linked to their unique ABHA IDs. This eliminates the need for physical paperwork and provides a comprehensive, secure health management system. The Health Records feature is backed by Samsung’s R&D, UX design, and consumer experience teams, in collaboration with Eka Care. Bengaluru-based Eka Care was founded in December 2020 by Vikalp Sahni and Deepak Tuli. Tuli emphasised the partnership’s potential to accelerate the adoption of ABDM, adding, “This collaboration is a key step toward building a more connected and efficient healthcare ecosystem in the country.” Samsung Research Institute’s managing director, Kyungyun Roo, highlighted the company’s focus on enhancing customer experiences, stating, “The Health Records feature empowers users to manage their health history, track progress, and maintain better control over their well-being.” Samsung and HealthCare Samsung has been actively expanding into the healthcare sector through various initiatives and partnerships. A couple of years ago, the company announced an Open Innovation Initiative in collaboration with institutions such as the MIT Media Lab, Brigham & Women’s Hospital, Tulane University School of Medicine, and Samsung Medical Center. This program aims to explore advancements in the digital health ecosystem and new approaches to wellness. Additionally, Samsung introduced the Galaxy Ring, a smart ring designed for health and wellness monitoring. The device tracks various health metrics, including sleep quality and heart rate, providing users with personalized wellness insights. The company also launched the SmartThings Family Care service to assist caregivers in remotely monitoring and supporting their loved ones through smart home devices and Galaxy smartphones. Big Tech and Healthcare Big tech companies have been offering healthcare solutions through applications and cloud suites. Apple, Amazon, Microsoft and others are offering solutions on the operations and research end. Apollo Hospitals, for example, has adopted Oracle Fusion for comprehensive ERP and HCM solutions, becoming its first payroll customer. Similarly, Fortis utilises Oracle Fusion to streamline back-office operations. Additionally, Aster Hospitals and Omega Healthcare have incorporated Oracle Fusion into their business process outsourcing (BPO) workflows. Bengaluru-based Healthify, was the only Indian startup to be featured at OpenAI Devday 2024. The health and wellness platform that recently raised $20M, raising the total equity to $125M leverages AI to personalise health coaching and tracking services.","excerpt":"The Health Records feature empowers users to manage their health history, track progress, and maintain better control over their well-being.","categories":["AI News"],"tags":["Eka care","Healthify","heathcare","OpenAI","Oracle","Samsung"],"author_name":"Vandana Nair","publish_date":"2025-01-17T16:33:30","publication_year":"2025","word_count":452,"keywords":["Go","Samsung","OpenAI","AI","innovation","ML","Git","Oracle","RAG","Aim","heathcare","Healthify","Eka care","R","startup"],"extracted_tech_keywords":["AI","ML","OpenAI","Aim","RAG","R","Go","Git","innovation","startup"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/samsung-partners-with-eka-care-to-launch-health-records-feature\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10169795,"title":"Shopping Agents are Here and Wayfair Wants to be Their First Pick","content":"From helping employees work smarter to making shopping more intuitive, AI is shaping every part of the business. Online shopping is no longer confined to static product listings. Today, AI and data analytics are used to understand user behaviour, helping businesses provide customised product offerings. While many e-commerce companies are still catching up with AI, American company Wayfair has had a jump start. Earlier this year, it launched a new AI tool called Muse, which helps customers visualise how furniture might look in a specific setting. With Muse, users can discover, personalise, and shop for home décor and furnishings more intuitively. They can enter prompts ranging from broad concepts like “dining room” to highly specific ideas such as “moody 1920s-style living room”. Based on the prompt, Muse generates a series of AI-created inspirational scenes. Each scene features recommended Wayfair products that users can purchase immediately or save for later, creating a smooth experience from inspiration to action. “Essentially, it (Muse) allows our customers to be inspired by something they see, and it allows them to also interact with it,” said Fiona Tan, CTO of Wayfair, in an exclusive interview with AIM. She further pointed out that with Muse, users can generate a wide array of home decor visualisations, which are then linked to Wayfair’s extensive catalogue using visual search capabilities. Moreover, she shared that they are going to introduce the voice feature to the platform soon. While Wayfair is not in the business of building furniture, Tan said there has been demand from customers for certain types of furniture created by them through Muse. “We generally don’t get involved in manufacturing, but we work closely with our suppliers. If customers are asking for furniture we don’t currently offer, we can share those insights with suppliers. That way, they can decide if it’s something they want to build based on emerging demand,” she said. The company is currently experimenting with several LLMs, including those from OpenAI, Google, and Anthropic. Tan pointed out that Gemini performs better for catalogue enrichment, ChatGPT is more effective for customer responses, and Claude is preferred for coding tasks. Wayfair uses Gemini on Google Cloud to automatically categorise products across its 30 million product catalogue, cutting the time to curate new listings and update existing ones by 67%. Addressing competition, particularly from IKEA, Tan pointed to their differing business models. “IKEA is very, I would say, physical stores first. They have a much smaller online presence. We’re kind of the flip. We’re mostly online.” This digital-first approach allows Wayfair to potentially leverage AI capabilities more extensively. Wayfair isn’t the only one using AI. Its competitors are doing the same to improve the shopping experience. In 2022, IKEA launched IKEA Kreativ, a tool that helps customers design and visualise their rooms in 3D on any device. Last October, Walmart also shared its plans to use AI, AR, and immersive tech. This included Wallaby, a set of AI models built to improve customers’ interaction with its platform. These models, which power personalised shopping, customer service, and content recommendations, are planned to roll out in the US, Canada, and Mexico by the end of 2025. At the same time, Amazon has introduced its own AI shopping assistant, Rufus. It simplifies the Amazon shopping process by providing intelligent assistance, personalised recommendations, and seamless product discovery for users. Key features of Rufus include the ability to guide customers on what to look for in various product categories, help them shop by occasion or purpose, facilitate comparisons between product categories, and provide personalised recommendations based on specific queries. SEO is Dead Reflecting on her journey as a CTO in the rapidly evolving space, Tan said she is trying to keep pace as much as possible. She believes that AI is going to have a significant impact on e-commerce. She pointed out that, in the past, e-commerce companies segmented customers into a few categories and created fixed email templates. Now, LLMs allow truly personalised experiences like generating unique subject lines and content per user based on intent and behaviour. According to Tan, with tools like Perplexity and ChatGPT becoming “shopping agents”, the future of product discovery may not start on e-commerce sites like Wayfair, but on these AI platforms instead. OpenAI and Perplexity AI recently announced new shopping features, allowing users to find, compare, and purchase products directly through the interface. Tan explained that this is going to change the SEO game as well. According to her, this will require optimising content not just for Google Search, but for LLMs and shopping agents. “This is like the next version of SEO. We need to figure out what these agents value. It’s going to be much more content-rich versus just keywords,” Tan said. How Wayfair is Using AI Coding Tools Tan highlighted that Wayfair is integrating AI-powered code assistance across the company, not just for developers, but for all employees. Developers can choose from tools like Google Gemini, GitHub Copilot, and Anthropic Claude. “We’re trying to ensure that [non-developers] have access to GenAI  tools as well, so they can become more productive in their own areas.” She added that other coding tools like Cursor and Codeium (now Windsurf) are also being evaluated. “I wish I had more time to play around with Cursor. It seems like a really nice one for prototyping. And then there’s another one called Codium that we’re also looking at,” Tan said. Tan is also aware of the term ‘vibe coding’. However, she still feels that developers should be mindful of what they are building with these tools. “The fact that you may not have to write a lot of actual code itself, you still need to know [the end goal],” she said, adding that analytics skills remain important. “Reviewing what is generated is also super important.” Tan added that the company’s productivity will increase with the use of AI, but she does not think that means fewer developers will be required. Interestingly, earlier this year, Wayfair laid off 340 employees and is using generative AI to increase productivity following the company’s move to a cloud-computing environment. With AI doing more heavy lifting and customers finding products in new ways, Wayfair seems ready for the next era of shopping, whether it starts on a website or not.","excerpt":"Wayfair’s new AI tool Muse helps customers visualise how furniture might look in a specific setting.","categories":["Global Tech"],"tags":["AI agent","wayfair"],"author_name":"Siddharth Jindal","publish_date":"2025-05-12T16:27:17","publication_year":"2025","word_count":1046,"keywords":["Anthropic","ChatGPT","GenAI","OpenAI","AI","wayfair","ML","AI agent","Aim","Ray","analytics","generative AI"],"extracted_tech_keywords":["AI","ML","analytics","generative AI","GenAI","ChatGPT","OpenAI","Anthropic","Aim","Ray"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/shopping-agents-are-here-and-wayfair-wants-to-be-their-first-pick\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10040938,"title":"Top Natural Language Generation Platforms In 2021","content":"Natural Language Generation (NLG), a subset of Natural Language Processing (NLP), outputs natural human language from data. The software system generates narratives and reports based on input data. It can also translate this text into audible speech. Here is a list of the top companies providing natural language generation services. (The list is in alphabetical order) 1| Arria About: Arria NLG is a form of artificial intelligence that transforms structured data into natural language. Through data analysis, knowledge automation, language generation and tailored information delivery, Arria software replicates the human process of expertly analysing and communicating data insights. The Arria NLG Platform automatically writes rich, compelling narratives based on insights extracted from datasets. Know more here. 2| Amazon Polly About: Amazon Polly is an API-driven service that uses advanced deep learning technologies to synthesise speech that sounds like a human voice. The platform provides dozens of lifelike voices across a wide variety of languages. Polly converts text into lifelike speech, allowing you to create applications that talk and build entirely new categories of speech-enabled products. Polly’s Text-to-Speech (TTS) service uses advanced deep learning technologies to synthesise natural sounding human speech. Know more here. 3| AX Semantics About: AX Semantics is an AI-powered, natural language generation (NLG) software company. It’s SaaS-based software makes automated content generation accessible to customers of all sizes and is used widely across e-commerce, business, finance and media publishing sectors. Available in 110 languages, AX Semantics includes features like NLG Output based on structured data, multi- and cross-language text generation, intuitive inline editor and more. Know more here. 4| Automated Insights About: Automated Insights is the creator of Wordsmith, a natural language generation platform that transforms data into insightful narrative. The platform enables complete narrative customisation, real-time content updates, and a powerful API for flexible publishing. Know more here 5| Quill About: Quill is an intelligent automation platform helping enterprises transform reporting workflows with natural language generation. It transforms data into stories and embeds them directly into the dashboards. Quill also allows users to create data-driven stories at scale automatically. Know more here. 6| Textengine.io About: Textengine.io is a self-service platform for Natural Language Generation (NLG). The platform is built on artificial intelligence and automatically transforms the structured data into text. It includes features like intelligent linguistic analysis, intuitive interface, automatic, machine translation of text structures and more. Know more here. 7| vPhrase About: Headquartered in India, vPhrase Analytics is an AI startup that provides business intelligence and automated reports using machine learning and natural language generation technology. Its core platform, Phrazor, uses NLG to transform complex data into easy-to-understand narratives. It auto-generates phrases from structured data and forms meaningful, summarised insights in natural language. Phrazor writes data-driven reports, easy-to-comprehend language narratives, and descriptions in a few bullet points alongside charts, tables, and visualisation. Know more here. 8| Yseop About: Yseop provides a range of Natural Language Generation solutions to amplify human expertise with advanced artificial intelligence technologies. The company uses advanced NLG technology and a suite of specialist tools that support report writing, sales management and other business processes. Leveraging AI, Yseop generates written narratives accurately from complex data sets. Know more here.","excerpt":"Natural Language Generation (NLG), a subset of Natural Language Processing (NLP), outputs natural human language from data. The software system generates narratives and reports based on input data. It can also translate this text into audible speech. Here is a list of the top companies providing natural language generation services. (The list is in alphabetical […]","categories":["AI Trends"],"tags":["natural language generation","NLG"],"author_name":"Ambika Choudhury","publish_date":"2021-05-30T16:00:00","publication_year":"2021","word_count":527,"keywords":["Go","machine learning","artificial intelligence","natural language generation","AI","TPU","RAG","NLP","deep learning","analytics","NLG","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","deep learning","NLP","analytics","RAG","TPU","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-natural-language-generation-platforms-in-2021\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10003052,"title":"10 Undergraduate AI &#038; Data Science Courses One Can Choose After 12th Grade","content":"With artificial intelligence and analytics being the talk of the hour, there cannot be a better time to get started with these technologies. COVID pandemic outbreak has further increased the demand for data scientists thus learning data science skills, in the current situation, can present high employment chances. Also Read: AIM Data Science Education Ranking 2020 | Top UG Programmes In India Till now, the field of data science and AI has been a preferred choice for postgraduate programs; however, the increasing demand for data professionals is making it imperative for students to start early. And that’s where an undergraduate AI and data science course can help. Now that the results for Class XII board exams are out, this could be the perfect chance for the pass out students to build a career in the most demanding profession of the world. Having a bachelor’s degree in data science and artificial intelligence can not only open up opportunities to exciting career paths but will also enhance their critical thinking and problem-solving abilities at a very early stage. In this article, we will share some of the prominent undergraduate data science and AI programs offered by Indian universities. Also Read: Reasons Why You Should Be Learning Data Science During This Time This list is not a ranking and is in no particular order. BSc In Programming & Data Science – IIT Madras About: Offered by IIT Madras, this is the first online BSc degree course in programming and data science, that can be enrolled by anyone who has passed class XII with English and math. Launched this year, the course has three levels — foundational level with eight courses; diploma level with 12 courses; and degree level with 11 courses — that students have to pass in order to achieve a BSc degree in data science. IIT Madras also provides flexibility to exit the course at any level with a relevant certificate from IIT Madras. Enrolled students will be able to access this course via 2-3 hours of recorded videos which will be released every week comprising lectures, hands-on workshops etc. It also comes with online assignments every week, regular quizzes, and end-term exams to pass the course. Batch Starts: January 2021 Completion Time: 3 – 6 years (based on preferred pace & performance) Click here for more information. BTech In Artificial Intelligence – IIT Hyderabad Launched in January 2019, this undergraduate course is offered by IIT Hyderabad and only accepts students based on their JEE Advanced scores. IIT Hyderabad has been the first Indian university to provide a full-fledged BTech programme in AI, which reportedly only takes 20 students every year. This programme will provide a sound understanding of the fundamentals of AI and machine learning and provide necessary hands-on for the same. The course comprises algorithms from the CS department, electrical engineering department, mechanical engineering department and will also focus on data science application on various verticals like healthcare, manufacturing, mobility, among many others. The course also heavily concentrates on teaching the ethical impacts of AI to the students and its related issues, which will enable them to become informed leaders of the industry. Batch Starts from the academic year 2020-21 Completion Time: 4 years Click here for more information. Bachelor Of Engineering in Artificial Intelligence – VTU Launched in June 2019, this programme is offered by Visvesvaraya Technological University in Karnataka becoming the first-ever university in the state to introduce AI and ML in their curriculum. The initiative was taken in the wake of the increased demand of AI and ML professionals in the market, and apart from IITs, VTU is the first to offer artificial intelligence and machine learning as a full-fledged BTech course. The program involves data structures and applications, software engineering, complex analysis, probability and statistics, designing of the algorithm, machine learning with Python, and computer image processing etc. Completion Time: 4 years Click here for more information. BTech in Artificial Intelligence – GH Raisoni College of Engineering Launched in May 2019, BTech in artificial intelligence is an undergraduate program offered by GH Raisoni College of Engineering which provides advanced knowledge in machine learning, deep learning and artificial intelligence. The course has been designed to enable students to learn skills in the area of AI and ML and build intelligent machines and software in future. The course involves discussion of AI methods based in different fields, including neural networks, signal processing and data mining. Led by experts of the institute — professor A. Thomas and professor Gopal Sakarkar, the programme helps students gain knowledge on problem analysis, develop solutions, modern tool usage, apply ethics, and manage projects. The course will further enable the students to apply new advancements in high-performance computing hardware and software. Admissions Open for 2020 – 21 Completion Time: 4 years Click here for more information. Tech Data Science & Engineering – Manipal Academy Of Higher Education This undergraduate programme in data science and engineering is offered by Manipal Academy, which trains students to be skilled data scientists. The foundation of this four-year course starts with subjects like computational mathematics, statistics and computer science and introduces students to domains like healthcare and finance, to name a few. This course also allows students to customise their course with preferred electives like blockchain, quantum computing, data forensics, data privacy, algorithmic trading etc. The institute comes with a strong placement cell in collaboration with many data science companies like Capgemini, Cartesian Consulting, Oracle, Publicis Sapient, Wipro etc. The programme also comes with a multi-campus model with mandatory student mobility in their 4th-6th semester. Batch Start: September 01, 2020 Completion Time: 4 years Click here for more information. Bachelor Of Data Science – SP Jain School Of Global Management Bachelor of data science by SP Jain School is a three-year full-time undergraduate programme which will provide students a profound understanding of data science with the techniques and skills to build solutions. This course will also teach how to identify patterns in order to predict trends from analysing data of various sectors such as manufacturing, banking and finance, retail and healthcare. The course is ideal for grade XII applicants with an aptitude for math, statistics and programming. This bachelor degree provides an opportunity to take the first year from their Mumbai campus and the next two years from their Sydney campus enables them with an Australian degree. The course involves mathematical statistics, computer programming, calculus, linear algebra in the first year; algorithms and data structures, visual analytics, matrix algebra and machine learning in the second year; and data mining, social web analytics, big data processing and a capstone project in the third. Admission Open for October 2020 intake Completion Time: 3 years Click here for more information. Bachelor in Data Science and Artificial Intelligence – IIIT Naya Raipur Launched in May 2019, BTech in Data science and AI is a full-time four years undergraduate programme offered by International Institute Of Information Technology, Naya Raipur. This credit-based program in data science and AI has been designed to build the core skills with innovation and make students industry-ready. The initiative is a joint venture with Chhattisgarh government and NTPC Ltd and has been evolving with the changing requirement of the industry. The course offers knowledge on discrete mathematics, linear algebra, programming, database management, algorithm building, computer vision, statistical data, to name a few. The course further allows the student to pick an elective in order to enhance their skills. Admission Open Completion Time: 4 years Click here for more information. Bachelor of Technology in Computer Science – AI & ML – Sharda University B. Tech in computer science with specialisation in artificial intelligence and machine learning is a four-year full-time undergraduate programme by Sharda University for students who are willing to make a career in the field of data science. The course will enable students to solve problems in a variety of domains like healthcare, finance etc. and will make them expert in natural language processing, text mining, robotics, and reasoning. The program will enable students to build intelligent machines, software solutions and novel applications with cutting-edge technology. Further, this BTech course will enable students to choose various career paths like data scientists, ML and AI engineer, ML architect, and data analyst. Alongside the program has collaborated with several tech companies like Amazon, CapGemini, Mphasis, NTTData, to name a few, which will help the students to land on a job post their course. Admission open for the 2020 -2021 batch Completion Time: 4 years Click here for more information. B.Tech. in Computer Science & Engineering (Artificial Intelligence) – Jain University B. Tech in computer science and technology with specialisation in artificial intelligence is an eight-semester full-time course offered by Jain University. The course will provide knowledge of fundamental concepts, design strategies, advanced technical architecture, progress infrastructural requirements, services, deployment model, tools and techniques. Further, it will accelerate understanding through application-oriented and student-centric learning. The course comes with eligibility criteria of passing in PUC\/10+2 examination with Physics and Mathematics as compulsory subjects along with Chemistry \/ Computer Science \/ Electronics as one of the subjects and obtained at least 60% marks for general category and 55% in case of SC\/ST category, in the above subjects taken together. Admissions open for current academic year Completion Time: 4 years Click here for more information. Tech in Computer Science and Engineering (Artificial Intelligence) – Amrita Vishwa Vidyapeetham B. Tech in computer science and engineering with specialisation in artificial intelligence is a four-year full-time undergraduate programme offered by the School of Engineering, Amrita Vishwa Vidyapeetham. With subjects like mathematics for intelligent systems, computational engineering, data structures and algorithms, this course grooms students to become future engineers. The course further allows students to take professional electives under engineering, science, mathematics, live-in-labs, and NPTEL courses. Students can opt for such electives across departments\/campuses. Students with CGPA of 7.0 and above can choose for a maximum of 2 NPTEL courses with the credits not exceeding 8. Students can also undertake and register for a Live-in-Labs project, for gaining hands-on experience on the subjects. Admission open for 2020 batch Completion time: 4 yearsClick here for more information.","excerpt":"With artificial intelligence and analytics being the talk of the hour, there cannot be a better time to get started with these technologies. COVID pandemic outbreak has further increased the demand for data scientists thus learning data science skills, in the current situation, can present high employment chances. Also Read: AIM Data Science Education Ranking […]","categories":["AI Trends"],"tags":["ai consulting","Applications of Data Mining","big data roi","Courses","Data Mining","deep learning projects","impact of big data in education","social network big data"],"author_name":"Sejuti Das","publish_date":"2020-07-23T11:00:00","publication_year":"2020","word_count":1688,"keywords":["data science","ai consulting","artificial intelligence","machine learning","AI","neural network","Data Mining","ML","Applications of Data Mining","computer vision","Aim","deep learning","big data roi","impact of big data in education","analytics","deep learning projects","Courses","social network big data"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","neural network","computer vision","data science","analytics","Aim"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/10-undergraduate-ai-data-science-courses-one-can-choose-after-12th-grade\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10098628,"title":"OpenAI Ropes in Global Illumination to Fix ChatGPT Mess","content":"OpenAI, creator of ChatGPT, announced the acquisition of Global Illumination, a New York-based startup leveraging AI to build creative tools, infrastructure and digital experiences. Global Illumination is led by the  trio of Thomas Dimson, Taylor Gordon, and Joey Flynn and the entire team has joined OpenAI to work on OpenAI’s core products including ChatGPT. Supported by venture capital firms Paradigm, Benchmark, and Slow, the team at Global Illumination boasts a track record of creating and developing products in the early stages of companies like Instagram and Facebook. Additionally, they have played pivotal roles in driving progress at renowned entities such as YouTube, Google, Pixar, Riot Games, and other noteworthy organizations. “We’re very excited for the impact they’ll have here at OpenAI” said OpenAI in the blog post. The latest project from Global Illumination is Biomes, an open-source sandbox MMORPG reminiscent of Minecraft, designed to run directly on the web. The acquisition comes at a time when recent trends in the user base of ChatGPT have raised interesting observations. In June, a dip in user engagement was noted compared to the previous month, May. Analysts attributed this decrease to the possibility of students being out of school during this time, leading to a shift in user patterns. This trend continued into July, with a further reduction in the ChatGPT user base. Data from SimilarWeb revealed that July witnessed a 12% month-on-month decline, with the user count dropping from 1.7 billion in June to 1.5 billion users. Moreover, ChatGPT finds itself in a competitive landscape with strong contenders like Bard and Claude 2. Notably, Google Brain and DeepMind have teamed up to develop Gemini, a venture that could potentially pose challenges for OpenAI’s dominance in the field. On a different front, Elon Musk is showing strong confidence in challenging OpenAI’s supremacy, as he sets his sights on creating a rival chatbot solution. The acquisition of Global Illumination comes at a crucial time for OpenAI as they sought to bounce back and add new features to ChatGPT.","excerpt":"Global Illumination boasts a track record of creating and developing products in the early stages of companies like Instagram and Facebook","categories":["AI News"],"tags":[],"author_name":"Siddharth Jindal","publish_date":"2023-08-16T23:33:24","publication_year":"2023","word_count":334,"keywords":["Go","ChatGPT","API","OpenAI","AI","Git","RAG","GPT","GAN","R"],"extracted_tech_keywords":["AI","ChatGPT","OpenAI","RAG","R","Go","Git","API","GPT","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/openai-ropes-in-global-illumination-to-fix-chatgpt-mess\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10118805,"title":"‘Many Indian VCs Don’t Even have a Thesis on Deep-tech Investment’","content":"Even though the global AI hype is creating an ocean of funds, Indian investors are focused on the precious few droplets. In an exclusive interview with AIM, Vishnu Vardhan, founder & CEO of SML and Vizzhy, which is the creator of Hanooman, said that most Indian investors are not ready to spend money on research and deep tech startups. “Many VCs do not even have a thesis on how to invest in deep tech,” said Vardhan, referring to the country’s ill-informed deep tech investors. Citing Zepto, Dunzo, and other startups running without profits for investors for a long time, Vardhan said: “People are happy losing money there, but they don’t want to lose money here [AI startups],” he added. He also said India has created 125 unicorns, but nothing is great tech. “They are all business ideas and consumer apps,” he said. Vardhan narrated a story about meeting a deep-tech investor who said his ticket size was only $2 million, which is minuscule compared to the investment required to do AI research. “I need at least INR 100 crore to set up a lab in India,” he added. Once, an investor asked Vardhan why he needed to set up a lab, talking about investing in the medical field. “Why don’t you treat 100 patients and tell me how much money you make?” the investor told Vardhan. After discussing it with his sister, Vardhan laughingly added that raising money in the US is better as Indian investors do not understand deep tech. How true is this? Amit Sheth, the chair and founding director of the Artificial Intelligence Institute at the University of Southern Carolina (AIISC) also believes that there is a need for a lot more investment in AI in India compared to now. “VCs really don’t take as much risk as we expect\/hope,” said Sheth. “Most VCs do not understand the technology in-depth and don’t take risks with undeveloped markets where revenue and payoff are further off,” he added. “It is easy for them to understand consumer tech; most run after the fad and buy into the hype,” he added, while highlighting that big-tech companies like Microsoft or Google would be much happier to take the risk than VCs in India. According to Rajan Anandan, managing director at Peak XV Partners, VCs are sitting on a total of $20 billion cash to invest in Indian startups, and the focus is currently on AI. Arjun Rao, partner at Speciale Invest, which invests in deep tech startups, believes otherwise. “You could say there’s less investment when compared to Silicon Valley, but that is because Silicon Valley has much deeper pockets,” Rao told AIM, highlighting that VCs in India are only focused on investing in generative AI. Rao explained that a lot of it is because there is not a concrete exit strategy when it comes to generative AI as the investment scenario is still just 1-2 years old in India. “India is still a young ecosystem. It is an unfair comparison,” said Rao. “We are moving at a very fast pace, which is the most important thing so we can catch up.” Another great example is Khosla Ventures. The firm has invested in Upliance AI and Sarvam AI, two AI startups in India, and is bullish on AI globally. “We believe AI has the power to disrupt numerous economic models and change the way we lead our daily lives over the coming years. We invest in deep tech and invest where we can be early, bold and impactful,” believes Khosla Ventures. It is also the startups’ fault A few weeks ago, Gaurav Aggarwal, a former Google Research employee who is building an AI startup called Ananas Labs in India, put forth a similar opinion. He said that VCs are not ready to put money in deep-tech startups but are only interested in OpenAI wrappers and so-called consumer-tech startups. Arguably, Aggarwal’s point of view does make sense. India’s generative AI scene is on an upswing, but the investors are cautious when it comes to investing in research startups. Though India should also focus on foundational research, there should be investors willing to invest in deep-tech research as well, which is clearly not the case. Though it is rare to find any other initiative being built from scratch, the lack of VCs’ interest in such initiatives also shows a lack of understanding of the field, which, fortunately, is slowly changing as well. On the other hand, the investors are wary as well. There have been several early AI unicorns which have disappeared or are running dry with fundings. The fate of Jasper, Stability AI, or the very famous Instoried debacle, have made the VCs tread with caution. However, there is clearly a need for deeper pockets as well to invest in actual AI startups and take bigger risks.","excerpt":"It is easy for them to understand consumer tech; most run after the fad and buy into the hype.","categories":["Deep Tech"],"tags":["AI Impacts"],"author_name":"Mohit Pandey","publish_date":"2024-04-22T16:00:00","publication_year":"2024","word_count":804,"keywords":["Go","artificial intelligence","unicorn","OpenAI","AI","ML","Aim","generative AI","AI Impacts","R","startup"],"extracted_tech_keywords":["AI","artificial intelligence","ML","generative AI","OpenAI","Aim","R","Go","startup","unicorn"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/many-indian-vcs-dont-even-have-a-thesis-on-deep-tech-investment\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10084615,"title":"Businesses Can Discover Goldmine in ChatGPT","content":"While a majority claim that ChatGTP is incapable of business use cases, that might not be entirely true. At Microsoft Future Ready Technology Summit, in Bengaluru, Sandeep Alur, director of the Microsoft Technology Center, demonstrated the possibility of using ChatGPT in auto insurance claims. Alur demonstrated a case in which ChatGPT was used to settle an auto insurance claim with the chatbot donning the role of an insurance agent. Once a user submits an insurance claim, ChatGPT converses with them asking relevant questions and context required to approve or reject the claim. This is just one of the many ways in which ChatGPT, built on GPT-3.5, can facilitate business use cases. ChatGPT allows companies and businesses to automate a diverse range of tasks like handling customer requests and queries, producing reports and undertaking social media promotions by creating marketing materials. The natural language processing capabilities of ChatGPT enables it to understand and respond to customers in a personalised manner. For example, if you ask the ChatGPT for suggestions to tackle depression and anxiety, it will suggest self-care care measures and speak to you on the importance of mental-being. This feels  similar to speaking to a friend. The automation of tasks through ChatGPT that would otherwise require human intervention, will help companies save on personnel costs. What adds to this is the fact that ChatGPT is (so far) free to use. Ecommerce websites can use ChatGPT to help buyers find products, offer recommendations and even complete transactions. This will pave the way for a personalised online shopping experience. For example, shopify store Shaka Surfboards used ChatGPT for creating a new year promotion mailer for its customers. For the travel industry, ChatGPT can assist in booking flights, trains, rental cars and hotels and even provide recommendations for holiday destinations and recreational activities. It can suggest possible routes, thereby making travel planning more convenient and hassle-free for the customers. In the entertainment and recreational categories, ChatGPT could transform into your virtual assistant customising a workout routine, designing your meditation sessions and maintaining your playlists. ChatGPT’s capability to comprehend and respond to user inputs naturally makes it an ideal match for these applications. For healthcare and medical professionals, ChatGPT could be used to translate clinical notes into patient-friendly versions. It can also be effectively used to provide medical information and assistance, such as answering FAQs or providing symptom checkers. This could ease the burden on the overworked healthcare professionals. ChatGPT in action Alongside being a very useful tool for businesses to assist their operations, ChatGPT has a host of uses for individuals, some of which are detailed below. US-based entrepreneur and co-founder of Cue (acquired by Apple), Daniel Gross, has integrated GPT-3.5 (that is used to build ChatGPT) into an AI-enabled assistant on WhatsApp. By embedding GPT-3.5 into WhatsApp, Gross was able to chat with the bot as if it were a friend on his contact list. He calls it WhatsApp-GPT. The code is available on GitHub. With ChatBCG, the world’s first text to powerpoint AI, you can create presentations using AI in minutes. Using the tool, one can  access the features of multiple themes, layouts, images, infographics, outlines, headings, and bullet points and also make keywords in bold. The key highlight of this tool is that you can also export the slides in PPTX and PTF formats. More features like conversational editing, data-driven charts and using content from your blog will soon be added. ChatGPT Writer is a free Google Chrome extension that produces complete emails or replies based on a few keywords. It is presently supported for Gmail and is likely to be offered for LinkedIn and Outlook. TweetGPT is  a chrome extension that uses ChatGPT to generate positive, negative and controversial tweets randomly. Search GPT displays the ChatGPT response alongside Google Search results. When you search for a question on Google, this extension will show you results from ChatGPT helping you save time alongside getting the best relevant responses. YouTube Summary with ChatGPT is a free Chrome extension that produces quick summaries for YouTube videos using AI helping you save time and facilitate quick learning.","excerpt":"ChatGPT can enable companies and businesses to automate a diverse range of tasks like handling customer requests and queries, producing reports and undertaking social media promotions","categories":["AI Features"],"tags":["business","ChatGPT","Microsoft"],"author_name":"Aparna Iyer","publish_date":"2023-01-10T17:00:00","publication_year":"2023","word_count":684,"keywords":["Go","ChatGPT","AI","ML","Git","GPT","business","Aim","ViT","GitHub","R","Microsoft"],"extracted_tech_keywords":["AI","ML","ChatGPT","Aim","R","Go","Git","GitHub","GPT","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/businesses-discover-goldmine-in-chatgpt\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10165134,"title":"Will Quantum Computing be a 5-Year Game or a Decades-Long Wait?","content":"The narrative of real-time applications of quantum computing has been a topic of debate for a long time. While some believe that it will take just around five years for the industry to churn out useful computers, many others say that it’s going to be at least a two-decade-long process. According to recent projections, the quantum computing market is expected to reach $1 to 2 billion annually by 2030. This only highlights the growth potential of this industry. Last year, the United Nations General Assembly (UNGA) had declared 2025 as the International Year of Quantum Science and Technology (IYQ), and considering the recent developments in just the past two months, this prediction may as well come true. A Five-Year Medley With Amazon’s release of Ocelot, its new quantum computing chip, the race for computing has become even more competitive. The company claimed that compared to current approaches, the chip can reduce the costs of implementing quantum error correction by up to 90%. When asked about his stand on Ocelot in the ongoing debate between two decades or five years of quantum, Andy Jassy, CEO at Amazon expressed in an intervew, “I’m hopeful that it’s more in the five-year range than it is in the 20-year range.” He further highlighted that many significant innovations, such as generative AI, appear to be “overnight successes” but are often the result of decades of foundational work. For instance, while generative AI seems like a recent breakthrough, it is an evolution of AI research spanning over 50 years. The technology became impactful when it became more accessible and functional. Jassy drew parallels with quantum computing, which has been in development for over a decade. He explained that such technologies often progress gradually before reaching a point where they solve previously intractable problems in a cost-effective manner. This sudden leap creates the illusion of overnight success. However, Jassy emphasised that the “euphoria” following these breakthroughs requires careful evaluation to determine their long-term impact and sustainability. In a recent statement, Microsoft co-founder Bill Gates said that quantum computing could become useful within three to five years. While acknowledging that unforeseen challenges might arise, Gates’ outlook suggests he believes that the foundational breakthroughs needed for practical quantum applications are already in place or rapidly approaching. Microsoft also recently announced Majorana 1, which it claims to be the world’s first quantum chip utilising topological qubits. The company earlier claimed the chip will enable quantum computers capable of solving “meaningful, industrial-scale problems in years, not decades”. Moreover, Hartmut Neven, founder and head of Google Quantum AI, has publicly stated that Google aims to release commercial quantum computing applications within five years. Last month, Neven expressed optimism about the timeline, and declared, “We’re optimistic that within five years we’ll see real-world applications that are possible only on quantum computers.” Jensen Huang Disagrees During an analyst event at CES, NVIDIA founder and CEO Jensen Huang suggested that bringing “very useful quantum computers” to market could take decades, citing the need for quantum processors, or qubits, to increase by a factor of 1 million. “If you kind of said 15 years… that’d probably be on the early side. If you said 30, it’s probably on the late side. But if you picked 20, I think a whole bunch of us would believe it”, he said. This single statement from Huang triggered a massive selloff in the quantum computing sector, erasing approximately $8 billion in market value, according to reports. The quantum computing companies’ stocks witnessed a sharp decline. For instance, IonQ shares fell over 31.65%, Rigetti Computing dropped by 37.25%, and D-Wave Quantum Systems saw its stock tumble down by 25.61% after Huang’s statement. Speaking in a recent podcast, Meta CEO Mark Zuckerberg also expressed skepticism about the near-term potential of quantum computing. “I’m not an expert on quantum computing, my understanding is that it’s still quite off from being a very useful paradigm.” Moreover, Ivana Delevska, founder and chief investment officer at Spear Advisors, also concurred with the 15-20 year timeline, stating that it “seems very realistic”. However, countering his claim, Quantum leaders were quick to challenge and form an alternative narrative. Alan Baratz, CEO of D-Wave Quantum Systems, dismissed Huang’s comments on quantum computing while calling them “dead wrong”. Baratz pointed to clients like Mastercard and NTT Docomo, who already use their quantum systems for business operations. $NVDA $QBTSTODAY JENSEN HUANG SAID THAT QUANTUM COMPUTING WAS 20 YEARS AWAY FROM BEING USEFUL.DWave Quantum $QBTS was down as much as 49% after these comments and ended the day down 36%. The stock is up 1000% from the September lows. The CEO of DWave says in this clip… pic.twitter.com\/2hGf9ZSS0R— amit (@amitisinvesting) January 8, 2025 He acknowledged that Huang’s timeline might apply to gate-based quantum computers but argued it was “100% off base” for annealing quantum computers. After Huang’s statement, J Keynes, a long-time investor in the quantum computing space, took to X to point out a big gap between the expectations of companies and academics regarding when quantum computing will take off. He believes it is time for the industry to show real results. Moreover, he added that long-term investors require validation through performance, not just market enthusiasm. According to him, just making progress in the lab or getting government contracts isn’t enough; there needs to be actual sales and practical uses that make money. 2 Months of Continuous Developments The past two months have underscored 2025 as a pivotal year in quantum computing, marked by significant breakthroughs, apart from Microsoft and AWS leading to increasing competition. While Google’s recent quantum chip, Willow, took over the internet after its release for suggesting the possibility of a ‘multiverse’, many critics questioned the tech giant’s bold claims. They said the tech giant’s claims were based on a flawed benchmark and that it has no real-world applications. The chip even sparked a visionary exchange between Google CEO Sundar Pichai and SpaceX’s Elon Musk. Beyond the notable advancements by these giants, other key players are making significant strides, further enriching the quantum landscape. PsiQuantum, an American quantum computing company, has unveiled its Omega quantum photonic chipset, designed for large-scale quantum computing applications. Manufactured in collaboration with GlobalFoundries, Omega integrates advanced photonic components capable of high-fidelity qubit operations and efficient chip-to-chip interconnects. The company plans to establish quantum compute centres in Brisbane, Australia, and Chicago, Illinois, by the end of 2027. Meanwhile, Rigetti Computing and Quanta Computer have entered a strategic partnership to accelerate the development and commercialisation of superconducting quantum computing technologies.","excerpt":"I’m hopeful that Ocelot is more in the five-year range than it is in the 20-year range, Amazon CEO said.","categories":["Deep Tech"],"tags":["quantum","Quantum Computing","quantum supremacy"],"author_name":"Sanjana Gupta","publish_date":"2025-03-05T11:00:00","publication_year":"2025","word_count":1085,"keywords":["Quantum Computing","Go","quantum supremacy","API","AWS","AI","innovation","quantum","Aim","generative AI","CLIP","AI research","R"],"extracted_tech_keywords":["AI","generative AI","Aim","AWS","R","Go","API","CLIP","innovation","AI research"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/will-quantum-computing-be-a-5-year-game-or-a-decades-long-wait\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10138978,"title":"‘Microsoft isn’t in a Great Place to Really Keep Innovating’","content":"“Microsoft has done a number of fantastic things, but I don’t think they’re in a great place to really keep innovating and pushing on this in the way that a startup can,” said Aman Sanger, one of the creators of Cursor (by Anysphere), in a recent episode of the Lex Fridman podcast. When asked how Cursor, a newer startup, could compete with a well-established Microsoft product like Copilot, he said that the tech giant lacked the research and experimentation necessary to really push the ceiling. “I don’t know how you put that into words, but when you compare a Cursor with a Copilot, Copilot pretty quickly starts to feel stale for some reason,” resonated Fridman. Citing OpenAI’s o1 and others, Anysphere co-founder Sualeh Asif said when they started Cursor, they really felt this frustration that you could see models getting better, but the Copilot experience had not changed. “It was like ‘Man, the ceiling is getting higher, why are they not making new things?’ They should be making new things,” he added, “Where’s all the features? There were no alpha features.” He said it was selling well and surely did great business, but he is one of those who really want to try and use new things. “And there were no new things for a very long time,” added Asif. Another co-founder of Anysphere, Arvid Lunnemark said Cursor is like an all-in-one platform for developers, touching upon the proximity and often overlap of responsibilities within the teams—where the person designing the UI\/UX is also involved in training the model—allowing for seamless integration and experimentation. This approach makes it possible to create things that wouldn’t be achievable without constant communication and collaboration. “And you’re using, like you said, Cursor to write Cursor?” said Fridman, adding that the team is rapidly implementing features, alongside doing the research experimentation necessary to really push the ceiling. Founded by Truell, Asif, Lunnemark, and Sanger, Anysphere’s Cursor started with the goal of writing the world’s software. The founding team said they were fans of VSCode and Microsoft Copilot, and getting early access to GPT-4o served as their path and foundation to building Cursor. The rest is history. Microsoft Integrates o1 A few days ago, GitHub made OpenAI’s o1 model available on Copilot. This integration was key to competing with the likes of Cursor. “So it looks like GitHub Copilot might be integrating o1 in some way, and I think some of the comments are saying, does this mean Cursor is done?” asked Fridman. “It’s time to shut down Cursor, yeah,” retorted Lunnemark. The Cursor team spoke about how they are experimenting with integrating the o1 model into the Cursor editor as well. And while there is curiosity from programmers, it is still not a default experience for coders as a clear use case hasn’t emerged yet. “I think most of the additional value from Cursor versus everything else out there is not just integrating the new model fast like o1. It comes from all of the depth that goes into these custom models,” said Sanger, stressing how thoughtful design and user experience is more important. Conversely, research shows that OpenAI’s o1 models, known for strong reasoning abilities, struggle with coding tasks, leading to slow responses, hallucinations, and reliance on more effective models like GPT-4 and Claude Sonnet 3.5 for code generation, sparking widespread developer dissatisfaction. The Rise of Cursor Alternatives Earlier this year, Zed AI, a code editor which brings LLMs directly into your editor with a text-centric approach, was launched in collaboration with Anthropic. Zed was founded in 2021 by Nathan Sobo, who previously worked at GitHub, where he built and led their Atom text editor team. Additionally, two coding assistants that have developed their own foundation models, Codeium AI and Magic AI, reached unicorn status in September following significant funding rounds from investors such as General Catalyst, Sequoia, and Nat Friedman and Daniel Gross. Recently, OpenAI introduced Canvas, a new interface for writing and coding that expands ChatGPT’s functionality beyond simple conversation. Anthropic’s Claude Artifacts, which is available to all users on iOS and Android, allows individuals to easily create apps without writing a single line of code. Meanwhile, Y Combinator has also been backing several AI code editors, especially since the launch of Cursor AI. They have funded startups like Continue, Pearl, Void, Type, and Melty, among others, which are direct open-source competitors to industry incumbents like Cursor AI and Zed AI. “Fast is Fun” Needless to say, developers are increasingly seeking AI tools that help them code efficiently by automating repetitive tasks and providing real-time predictive assistance. A GitHub survey showed that 97% of developers had used AI coding tools at work, regardless of employer approval. The trend is going to continue. With Cursor’s Tab feature, developers can eliminate ‘low entropy’ actions, such as typing predictable code snippets, enabling faster workflows. The future of programming will see “all of programming flow through these models”, said Asif, indicating a major shift in how developers build software. With Cursor focusing on speculative edits and caching to deliver real-time assistance, speed is a crucial factor. According to Senger, Speculative edits allow the system to “predict ahead” of the developer’s actions, streamlining the coding process to near-instant responses. Cursor claims to make current processes obsolete within a year, reflecting how developers prioritise tools that accelerate workflows while maintaining high code quality. “Even being a few months ahead in AI programming makes your product much, much more useful,” said Truell, adding that Cursor offers developers a faster coding environment using sparse models, speculative decoding, and multi-query attention, thereby significantly reducing latency. “I actually think this is a really, really exciting time to be building software,” he said, reminiscing about the good old programming days of 2012-2013. “So, it’s going to be a really, really fun time for people who build software and the skill will probably change too. I also think that people’s taste and creative ideas will be magnified,” he added, saying, “Fast is Fun.” “That should be a T-shirt,” quipped Fridman.","excerpt":"A few days ago, Microsoft made OpenAI’s o1 model available on Copilot.","categories":["AI Features"],"tags":["Cursor AI","Microsoft"],"author_name":"Aditi Suresh","publish_date":"2024-10-21T14:42:22","publication_year":"2024","word_count":1006,"keywords":["Anthropic","ChatGPT","Go","OpenAI","AI","GPT-4o","ML","Aim","foundation models","R","Cursor AI","Microsoft"],"extracted_tech_keywords":["AI","ML","foundation models","GPT-4o","ChatGPT","OpenAI","Anthropic","Aim","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/microsoft-isnt-in-a-great-place-to-really-keep-innovating\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":17382,"title":"11 Things We Do Offline That Add Us To Big Data","content":"In today’s world, a person is easily decoded into a collection of his or her online activities. From tweets, status updates, group chats, photos, videos, to browser data history, favourite sites and playlists — every things we do when we’re connected to the internet makes up our digital persona. But what would happen if you decide to go cold turkey and switch off the internet? Analytics India Magazine has listed out 11 things that we do offline that still end up making us a part of Big Data. 1. Census: Every country takes a stock of its population by systematically acquiring and recording information about the members in their nation. This is one of the biggest sources of data for a country (and also the world). Policies, governance, elections and other basic properties are decided on these number. In India, the population is also segregated according to sex, age, religion, caste and geography, among other elements. 2. Government identifications: Aadhaar cards, driver’s licence, PAN cards and passports are some of the many identity documents issued by the government that make us a part of Big Data. 3. CCTVs Almost all residential and commercial establishments in tier 1 and 2 cities use CCTV cameras to monitor visitors for security purposes. Not only is this data useful for the establishments, it is also used frequently by the police to solve crimes. They also use facial recognition software to pinpoint the location of any person at a given time and date. 4. ATMs: Cash withdrawal from any ATM around the world leaves a digital footprint. From CCTV camera to transaction details, the bank keeps a record of all the details. 5. Mobile phone towers Even a basic phone without an internet connection is constantly in touch with mobile phone towers scattered across the globe. As long as the sim card is inside a phone that is switched on, locating whereabouts of that particular number is easy for the phone companies. 6. Car GPS: Modern cars are fitted with a global positioning system (GPS) which is a 24-satellite navigation system that uses multiple satellite signals to find a receiver’s position on earth. A user’s whereabouts are automatically recorded with the help of this technology. 7. Clocking in and out of work: Swiping access card at the workplace collects the log in as well as the log out time of an employee, staff or guest in an office. It is one of the ways to track the attendance of an employee in an organisation. 8. Shopping patterns in stores: Large as well as small retailers have only started using the unstructured data collected by them to their advantage. From shopping patterns, customer feedback comments, video footage, telephone conversations and location-based GPS data, bigger retail chains are now relying on data analytics to enhance customer experience. 9. Phone number-based loyalty cards: Most of these days, from grocery to couture, loyalty cards and the associated “point” form a large data for the retail giants. From understanding the customer preferences to advertising or a particular brand, these cards help businesses monetise the data. 10. Property purchase: Purchase of property, especially land, has now become a matter of public record. From keeping a track of agricultural and non-agricultural used for appropriate purposes, to clearing titles for resales, many states in India have uploaded geographical details on their official websites. 11. Offline surveys: Many organisations as well as institutes across the world adopt door-to-door surveys to gather information about new researches, studies and campaigns. The data collected is analysed and uploaded on the internet by most of them.","excerpt":"In today’s world, a person is easily decoded into a collection of his or her online activities. From tweets, status updates, group chats, photos, videos, to browser data history, favourite sites and playlists — every things we do when we’re connected to the internet makes up our digital persona. But what would happen if you […]","categories":["IT Services"],"tags":["Big Data"],"author_name":"Prajakta Hebbar","publish_date":"2017-08-31T10:31:57","publication_year":"2017","word_count":598,"keywords":["big data","Go","programming_languages:R","AI","programming_languages:Go","Git","ViT","analytics","GAN","Big Data","R"],"extracted_tech_keywords":["AI","analytics","R","Go","Git","big data","GAN","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/11-things-offline-add-us-big-data\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10143762,"title":"NVIDIA Launches Jetson Orin Nano Super with 1.7x Inference Boost at $249","content":"NVIDIA has introduced its new Jetson Orin Nano Super Developer Kit, a compact generative AI supercomputer priced at $249, down from the earlier price of $499. The device offers up to a 1.7x boost in generative AI inference performance and supports popular models for developers, hobbyists, and students. The Jetson Orin Nano Super delivers improved performance with 67 INT8 TOPS, a 70% increase compared to its predecessor, and a memory bandwidth of 102GB\/s, a 50% jump. “Almost 70 trillion operations per second, 25 watts, and just $249. It runs everything the HGX does, including large language models,” said NVIDIA chief Jensen Huang. Available starting today, the kit supports applications including LLM chatbot development, visual AI systems, and AI-driven robotics. “The enhanced performance of the Jetson Orin Nano Super provides gains for all popular generative AI models and transformer-based computer vision,” NVIDIA said in a statement. The developer kit includes a Jetson Orin Nano 8GB system-on-module (SoM) and a reference carrier board for edge AI application prototyping. The SoM features an NVIDIA Ampere architecture GPU with tensor cores and a 6-core Arm CPU, supporting multiple AI application pipelines and up to four cameras with improved resolution and frame rates. Existing Jetson Orin Nano Developer Kit owners can unlock similar performance gains by upgrading their software. “The software updates available to the new Jetson Orin Nano Super will also boost generative AI performance for those who already own the Jetson Orin Nano Developer Kit,” NVIDIA said. The Jetson ecosystem includes NVIDIA AI tools such as Isaac for robotics, Metropolis for vision AI, and Holoscan for sensor processing. NVIDIA also provides resources like the Omniverse Replicator for synthetic data generation and the TAO Toolkit for fine-tuning pre-trained AI models. The Jetson platform benefits from support across NVIDIA’s AI community, including tutorials and contributions from developers. NVIDIA noted that partners offer additional AI tools, developer support, cameras, and carrier board solutions for customised development. “The updates will extend to the Jetson Orin NX and Orin Nano series, allowing developers to access boosted generative AI performance across the board,” the company added. The new performance upgrades can be enabled today by updating to the latest JetPack SDK.","excerpt":"The device offers up to a 1.7x boost in generative AI inference performance and supports popular models for developers, hobbyists, and students.","categories":["AI News"],"tags":["NVIDIA"],"author_name":"Siddharth Jindal","publish_date":"2024-12-17T21:16:09","publication_year":"2024","word_count":362,"keywords":["synthetic data","programming_languages:R","AI","computer vision","emerging_tech:edge AI","edge AI","generative AI","ai_applications:robotics","NVIDIA","R","ai_applications:computer vision"],"extracted_tech_keywords":["AI","computer vision","generative AI","edge AI","R","synthetic data","programming_languages:R","ai_applications:computer vision","ai_applications:robotics","emerging_tech:edge AI"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/nvidia-launches-jetson-orin-nano-super-with-1-7x-inference-boost-at-249\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10123822,"title":"All AI Startups Have the Potential to be the Next OpenAI","content":"In a recent interview, OpenAI CTO Mira Murati said that the models the company uses in its labs are not far behind those currently available to the public. Speaking on the release of GPT-4o, she said that it was a massive deal for OpenAI to be able to make this kind of technology accessible to the public. “I don’t think there is enough emphasis on how unique that is for the stage where the technology is today. “In the sense that inside the labs, we have these capable models and they’re not that far ahead of what the public has free access to. That’s a completely different trajectory for bringing technology into the world than what we’ve seen historically,” she said. This brings up an interesting point of how, while many AI startups struggle to survive past their early stages, it is still possible to rise up, much like OpenAI did in the last few years. While Murati’s point is specifically about OpenAI’s ability to make advanced models widely available, she also emphasised that this is to ensure that the general public is fully aware of how the technology is progressing. “It’s a great opportunity because it brings people along; it gives them an intuitive sense of the capabilities and risks. The opportunities are huge now,” she said. Murati’s stance implies that startups are not far behind in terms of the technology needed to develop advanced AI systems. Behind OpenAI’s Success OpenAI is widely regarded as an AI success story. Initially founded as a non-profit towards the goal of developing beneficial AGI, the company quickly rose through the ranks of the startup world to become a household name. However, there were several factors that helped OpenAI gain the attention it did. Namely, access to funding as well as a large talent pool, allowed the startup to focus largely on quality research. One of the key points to remember about OpenAI is that its big break with ChatGPT came only a few years after the company received a $1 billion investment from Microsoft. Additionally, the timing was near perfect, as they had focused their attention on the development of Transformer models, shortly after the release of the 2017 paper, ‘Attention is All You Need’. Of course, it’s key to note that all of these factors rely on ensuring that your problem statement is one that can result in something ground-breaking. What Needs to Be Done? Murati’s implication that AI models are much more widely available now than ever before, means that startups need to try harder to stand out. OpenAI appeared as a pioneer in making an easy-to-use AI product widely available to the public for free. However, only years later, doing the same doesn’t have the same impact anymore. In India, several startups have managed to make a splash, albeit not as big as OpenAI’s. Successful Indian startups like Krutrim, Kissan and Sarvam have been able to leverage gaps between the AI revolution and use cases specific to India, allowing them to break ground as pioneers in the country. Being able to identify certain gaps and plug them in a way that ensures a startup’s continuity is a tough task. This is especially true as rapid advancements from bigger companies mean that startups with a poorly thought-out plan and product find themselves obsolete almost overnight. With the proper resources, however, this can be remedied. One must keep in mind that OpenAI’s ability to get ahead of the curve was solely due to a team of researchers that tried to leverage then-unexplored technology for use among the general public. AI advancements mean that a lot more students and professionals are pivoting towards studying AI, thereby creating a large talent pool, especially in India. Besides, with startup funding increasing exponentially, it seems that the only roadblock now is trying to find a way to stand out in a sea of problem statements. However, as Murati said, “If you can push human knowledge forward, you can push society and civilisation forward.”","excerpt":"“It’s a great opportunity because it brings people along; it gives them an intuitive sense of the capabilities and risks,” said OpenAI CTO Mira Murati.","categories":["Deep Tech"],"tags":["Mira Murati","OpenAI"],"author_name":"Donna Eva","publish_date":"2024-06-17T16:48:07","publication_year":"2024","word_count":668,"keywords":["Go","ChatGPT","API","OpenAI","AI","GPT-4o","Mira Murati","RAG","GPT","R","startup"],"extracted_tech_keywords":["AI","GPT-4o","ChatGPT","OpenAI","RAG","R","Go","API","GPT","startup"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/all-ai-startups-have-the-potential-to-become-the-next-openai\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":32456,"title":"Indian Government Gets Another AI Push, Parliament To Use AI To Streamline Operations","content":"The Indian government recently has been in the news for their attempt to harness the best of new-age techs such as artificial intelligence (AI) and machine language for better delivery of public services. With the incorporation of technologies, the government seems to acknowledge the fact that the road ahead for India’s growth is through leveraging these technologies on a large scale basis across public sectors. Adding another feather to its cap, the representative of Ministry of Parliamentary Affairs recently announced that the two houses of parliament, the Lok Sabha and Rajya Sabha will soon use AI and ML for streamlining their legislative duties. Speaking about this to a leading daily, Surendra Nath Tripathi, Parliamentary Affairs Secretary said the Government is planning to achieve this is by using AI and ML in the second phase of National E-Vidhan (NeVA) Project. NeVa is a new flagship programme under the Ministry, which aims to create a massive data depository by bringing together all the legislatures of the country together in one platform. By using AI and ML for the platform’s functioning, the government wants to process the information to make it available in different formats and patterns, thereby making it easier to draw meaningful insights from the data. Analysis and processing of data are crucial for the smooth functioning of houses, as combined membership of both the houses are 5300 with 550 different standing members generating over 500 reports each year. By studying the pattern and drawing insight from the data with AI and Ml enabled system, the Ministry hopes to make the process of providing replies to over 2 lakh questions asked each year by MPs and MLAs easier. “In phase one, we are going for data processing only. In phase two we are going to use intelligent data processing that is we will use AI and Machine Learning technologies and will make the information available in different formats and analyse, this is what we call ‘cooked information,” Tripathi said.","excerpt":"The Indian government recently has been in the news for their attempt to harness the best of new-age techs such as artificial intelligence (AI) and machine language for better delivery of public services. With the incorporation of technologies, the government seems to acknowledge the fact that the road ahead for India’s growth is through leveraging […]","categories":["AI News"],"tags":["AI in Public Governance","AI in Public Sector","Indian government using AI"],"author_name":"Akshaya Asokan","publish_date":"2018-12-31T06:38:53","publication_year":"2018","word_count":328,"keywords":["Go","artificial intelligence","machine learning","programming_languages:R","AI","ML","programming_languages:Go","AI in Public Governance","RAG","AI in Public Sector","Aim","Indian government using AI","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","Aim","RAG","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/indian-government-gets-another-ai-push-parliament-to-use-ai-to-streamline-operations\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10133165,"title":"Infosys to Earn $100M+ in Coca-Cola’s Major Cloud Deal with Microsoft","content":"Infosys is set to earn over $100 million as a key supporting partner in Coca-Cola’s $1.1-billion cloud migration deal with Microsoft, according to a recent report by the Economic Times. The report added that the agreement signed in April, marks a significant move by Coca-Cola to transition its operations to the cloud, enhancing its digital infrastructure and capabilities. Microsoft, the lead player in the deal, has enlisted Infosys to provide critical support services, capitalising on the Indian IT giant’s expertise in cloud technologies. This partnership highlights Infosys’ expanding role in large-scale global digital transformation initiatives, with a strong focus on cloud migration services. The company’s involvement in such a high-profile partnership is expected to bolster its revenue stream significantly, further solidifying its position in the competitive IT services market. The Indian IT giant’s involvement in the project is expected to bring in significant revenue, with over $27 million already secured from the Euro Pacific geography alone, according to regulatory filings with the US Securities and Exchange Commission (SEC). The filings reveal that Coca-Cola Euro Pacific Partners PLC, a subsidiary of The Coca-Cola Company, committed €167 million to Microsoft for Azure cloud migration services over a six-year period. Additionally, €25 million has been earmarked for Infosys as a supporting partner in this initiative. In April 2024, Microsoft and Coca-Cola announced a five-year strategic partnership aimed at aligning Coca-Cola’s core technology strategy and fostering innovation and productivity worldwide. As part of the agreement, Coca-Cola committed $1.1 billion to Microsoft Cloud and its generative AI capabilities. The companies plan to explore new technologies, including Azure OpenAI Service, to develop innovative AI use cases across various business functions. Infosys has been a key partner with Microsoft in AI-driven initiatives. Last September, the two companies announced a collaboration to help enterprises adopt an AI-first approach to scale next-generation AI solutions, improve operational efficiencies, drive revenue growth, and enable business transformation. Neither Infosys nor Coca-Cola has officially commented on the specifics of the financial arrangement, but industry experts suggest that Infosys’ role could be crucial in the successful execution of the cloud migration process. The deal is also expected to strengthen Infosys’ relationship with Microsoft, potentially leading to future collaborations in similar large-scale projects.","excerpt":"Microsoft, the lead player in the deal, has enlisted Infosys to provide critical support services, capitalising on the Indian IT giant’s expertise in cloud technologies.","categories":["AI News"],"tags":["Infosys"],"author_name":"Siddharth Jindal","publish_date":"2024-08-20T10:39:54","publication_year":"2024","word_count":368,"keywords":["API","OpenAI","Infosys","AI","R","digital transformation","Git","Aim","ViT","generative AI","Azure"],"extracted_tech_keywords":["AI","generative AI","OpenAI","Aim","Azure","R","Git","API","ViT","digital transformation"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/infosys-to-earn-100m-in-coca-colas-major-cloud-deal-with-microsoft\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10053515,"title":"A Guide to Inferencing With Bayesian Network in Python","content":"Bayesian networks can model nonlinear, multimodal interactions using noisy, inconsistent data. It has become a prominent tool in many domains despite the fact that recognizing the structure of these networks from data is already common. For modelling the conditionally dependent data and inferencing out of it, Bayesian networks are the best tools used for this purpose. In this post, we will walk through the fundamental principles of the Bayesian Network and the mathematics that goes with it. Also, we will also learn how to infer with it through a Python implementation. The key points to be covered in this post are listed below. Table of Contents What is Bayesian Network?What is Directed Acyclic Graph (DAG)?The Maths Behind Bayesian NetworkInferencing with Bayesian Network in Python Let’s start the discussion by understanding the what is Bayesian Network. What is Bayesian Network? A Bayesian network (also spelt Bayes network, Bayes net, belief network, or judgment network) is a probabilistic graphical model that depicts a set of variables and their conditional dependencies using a directed acyclic graph (DAG). Bayesian networks are perfect for taking an observed event and forecasting the likelihood that any of numerous known causes played a role. A Bayesian network, for example, could reflect the probability correlations between diseases and symptoms. Given a set of symptoms, the network may be used to calculate the likelihood of the presence of certain diseases. What is Directed Acyclic Graph (DAG)? In graph theory and computer science, a directed acyclic graph (DAG) is a directed graph with no directed cycles. In other words, it’s made up of vertices and edges (also called arcs), with each edge pointing from one vertex to the next in such a way that following those directions would never lead to a closed-loop as depicted in below picture. Source A directed graph is one in which all edge directions are consistent and the vertices can be topologically arranged in a linear order. DAGs have various scientific and computing applications, including biology evolution, family trees, and epidemiology, and sociology. Let’s see quickly what are fundamental maths involved with Bayesian Network. The Maths Behind the Bayesian Network An acyclic directed graph is used to create a Bayesian network, which is a probability model. It’s factored by utilizing a single conditional probability distribution for each variable in the model, whose distribution is based on the parents in the graph. The simple principle of probability underpins Bayesian models. So, first, let’s define conditional probability and joint probability distribution. Conditional Probability Conditional probability is a measure of the likelihood of an event occurring provided that another event has already occurred (through assumption, supposition, statement, or evidence). If A is the event of interest and B is known or considered to have occurred, the conditional probability of A given B is generally stated as P(A|B) or, less frequently, PB(A) if A is the event of interest and B is known or thought to have occurred. This can also be expressed as a percentage of the likelihood of B crossing with A: Joint Probability The chance of two (or more) events together is known as the joint probability. The sum of the probabilities of two or more random variables is the joint probability distribution. For example, the joint probability of events A and B is expressed formally as: The letter P is the first letter of the alphabet (A and B).The upside-down capital “U” operator or, in some situations, a comma “,” represents the “and” or conjunction.P(A ^ B)P(A, B) By multiplying the chance of event A by the likelihood of event B, the combined probability for occurrences A and B is calculated. Posterior Probability In Bayesian statistics, the conditional probability of a random occurrence or an ambiguous assertion is the conditional probability given the relevant data or background. “After taking into account the relevant evidence pertinent to the specific subject under consideration,” “posterior” means in this case. The probability distribution of an unknown quantity interpreted as a random variable based on data from an experiment or survey is known as the posterior probability distribution. Inferencing with Bayesian Network in Python In this demonstration, we’ll use Bayesian Networks to solve the well-known Monty Hall Problem. Let me explain the Monty Hall problem to those of you who are unfamiliar with it: This problem entails a competition in which a contestant must choose one of three doors, one of which conceals a price. The show’s host (Monty) unlocks an empty door and asks the contestant if he wants to swap to the other door after the contestant has chosen one. The decision is whether to keep the current door or replace it with a new one. It is preferable to enter by the other door because the price is more likely to be higher. To come out from this ambiguity let’s model this with a Bayesian network. For this demonstration, we are using a python-based package pgmpy is a Bayesian Networks implementation written entirely in Python with a focus on modularity and flexibility. Structure Learning, Parameter Estimation, Approximate (Sampling-Based) and Exact inference, and Causal Inference are all available as implementations. from pgmpy.models import BayesianNetwork from pgmpy.factors.discrete import TabularCPD import networkx as nx import pylab as plt # Defining Bayesian Structure model = BayesianNetwork([('Guest', 'Host'), ('Price', 'Host')]) # Defining the CPDs: cpd_guest = TabularCPD('Guest', 3, [[0.33], [0.33], [0.33]]) cpd_price = TabularCPD('Price', 3, [[0.33], [0.33], [0.33]]) cpd_host = TabularCPD('Host', 3, [[0, 0, 0, 0, 0.5, 1, 0, 1, 0.5], [0.5, 0, 1, 0, 0, 0, 1, 0, 0.5], [0.5, 1, 0, 1, 0.5, 0, 0, 0, 0]], evidence=['Guest', 'Price'], evidence_card=[3, 3]) # Associating the CPDs with the network structure. model.add_cpds(cpd_guest, cpd_price, cpd_host) Now we will check the model structure and associated conditional probability distribution by the argument get_cpds() will return True if every this is fine else through an error msg. model.check_model() Now let’s infer the network, if we want to check at the next step which door will the host open now. For that, we need access to the posterior probability from the network and while accessing we need to pass the evidence to the function. Evidence is needed to be given when we are evaluating posterior probability, here in our task evidence is nothing but which door is Guest selected and where is the Price. # Infering the posterior probability from pgmpy.inference import VariableElimination infer = VariableElimination(model) posterior_p = infer.query(['Host'], evidence={'Guest': 2, 'Price': 2}) print(posterior_p) The probability distribution of the Host is clearly satisfying the theme of the contest. In the reality also, in this situation host definitely not going to open the second door he will open either of the first two and that’s what the above simulation tells. Now, let’s plot our above model. This can be done with the help of Network and Pylab. NetworkX is a Python-based software package for constructing, altering, and researching the structure, dynamics, and function of complex networks. PyLab is a procedural interface to the object-oriented charting toolkit Matplotlib, and it is used to examine large complex networks represented as graphs with nodes and edges. nx.draw(model, with_labels=True) plt.savefig('model.png') plt.close() This gives us the Directed Acyclic Graph (DAG) as like below. Conclusion Through this post, we have discussed what a Bayesian Network is. In addition to that we have discussed how the Bayesian network can be represented using DAG and also we have discussed what are the general and simple mathematical concepts are associated with the network. Lastly, we have seen the practical implementation of the Bayesian network with help of the python tool pgmpy, and also plotted a DAG of our model using Netwrokx and pylab. References Bayesian NetworkDirected Acyclic GraphOfficial documentation pgmpyLink for above codes","excerpt":"In this post, we will walk through the fundamental principles of the Bayesian Network and the mathematics that goes with it. Also, we will also learn how to infer with it through a Python implementation.","categories":["Deep Tech"],"tags":["bayesian inference","Bayesian network","Guide","Machine Learning","probability"],"author_name":"Vijaysinh Lendave","publish_date":"2021-11-15T18:14:11","publication_year":"2021","word_count":1275,"keywords":["Go","API","probability","ELT","programming_languages:R","AI","Modal","Bayesian network","Machine Learning","bayesian inference","Python","programming_languages:Python","Matplotlib","R","Guide"],"extracted_tech_keywords":["AI","Matplotlib","Python","R","Go","API","ELT","Modal","programming_languages:Python","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/a-guide-to-inferencing-with-bayesian-network-in-python\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10121463,"title":"Adobe Rolls Out GenAI-Powered One-Click Object Removal in Lightroom","content":"Creative software giant Adobe has launched Generative Remove in Lightroom, powered by its in-house model Adobe Firefly, for mobile, web, and desktop editing. This tool allows users to remove unwanted objects from photos in a single click, matching the removed area for high-quality results. The feature is designed for all photographers, from hobbyists to professionals, and is now available for early access. Additionally, Lightroom has also introduced the AI-powered Lens Blur tool, which adds aesthetic blur effects with a single click, and includes new presets. These updates aim to streamline photo editing workflows, making the process faster and more intuitive. Adobe’s Generative Remove can handle complex backgrounds, enhancing the editing experience by removing distractions and imperfections. The tool’s early access phase invites community feedback to improve its capabilities further. Other updates include expanded tethering support for new cameras, HDR Optimisation for vivid photo editing, and a revamped mobile editing interface. Generative Remove is powered by Adobe Firefly, trained on licensed content to avoid copyright issues. Adobe emphasises responsible AI development, adhering to accountability, responsibility, and transparency principles. Content Credentials will be attached to photos edited with Generative Remove, ensuring authenticity and trust. “We want to use AI responsibly and prioritise the protection of intellectual property,” Prativa Mohapatra, VP and MD of Adobe India, told AIM in an interaction last week. “We employ measures like content credentials, which have become the industry standard for digital content provenance.” The new features are available across the Lightroom ecosystem, including mobile, desktop, iPad, and web. For more details, visit Adobe’s official site.","excerpt":"Additionally, Lightroom introduces the AI-powered Lens Blur tool, which adds aesthetic blur effects with a single click, and includes new presets.","categories":["AI News"],"tags":["Adobe"],"author_name":"Shritama Saha","publish_date":"2024-05-24T09:48:10","publication_year":"2024","word_count":258,"keywords":["programming_languages:R","AI","Adobe","ML","Git","responsible AI","Aim","ViT","Rust","R","programming_languages:Rust"],"extracted_tech_keywords":["AI","ML","Aim","R","Rust","Git","ViT","responsible AI","programming_languages:R","programming_languages:Rust"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/adobe-rolls-out-genai-powered-one-click-object-removal-in-lightroom\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":65894,"title":"Recent Data Science Jobs For Freshers To Apply Right Now","content":"One of the hottest jobs of the 21st century, data science plays a crucial role when it comes to deriving meaningful insights from large chunks of data for decision-making in organisations. According to reports, as of March 2020, the analytics function in India earned consolidated revenues of $35.9 B – a 19.5% growth in revenue over last year, where 16% of the analytics revenues across all enterprises are attributed to advanced analytics, predictive modelling, and data science. In this article, we list down 8 data science and analyst jobs for freshers to apply right now. (The list is in no particular order) 1| Data Scientist at Siemens Location: Bangalore Responsibilities: As a data scientist, you will work with large, complex data sets and apply advanced analytical methods as needed, develop custom data models and algorithms to apply to data sets. You will also run POCs and pilot projects to investigate and prove the feasibility and benefits of the proposed solutions, interact with multiple partners (from various levels) and optimally present findings by exploiting visual displays of complex quantity information in a simplified way, among others. Apply here. 2| Data Science (Jupyter\/Azure ML\/MatLab) at Accenture Location: Hyderabad Responsibilities: In this role, you will develop analytics-based solutions that produce quantitative and qualitative business insights, work with partners as necessary to integrate systems and data quickly and effectively, regardless of technical challenges or business environments. Besides, you should be experienced in querying databases, statistical computer languages: R, Python, SLQ, etc., working in statistical and data mining techniques such as GLM\/regression, random forest, boosting, trees, text mining, social network analysis, and other such. Apply here. 3| Data Scientist at Talent Tigers Location: Bangalore Responsibilities: You will be applying technical and analytical expertise to explore and examine data from multiple disparate sources with the goal of discovering patterns and previously hidden insights, which can provide a competitive advantage or address a pressing business problem. You should have expertise in computer science software development and the latest technologies to analyse and implement analysis infrastructure and tools analytic workflow processes and complex data visualisations. Apply here. 4| Data Science Platform Administrator Location: Pune Responsibilities: As a data science platform administrator, you should have a clear understanding of GPU and its relationship with deep learning and neural networks. You should demonstrate experience along with developing and  managing complex technical projects that involve parallel and distributed computing, including Hadoop, the Apache Stack and other related technologies Apply here. 5| Data Analyst at Reliance Jio Infocomm Location: Bangalore Responsibilities: This position requires the individual to stitch together fields of machine learning, data design, data mining, statistics, user behaviour, computational economics and data streams to define and drive mobile user research at RJIL. You will be responsible for solving complex big-data problems in mobility and cloud applications space using data mining, machine learning, statistical analysis and computational economics. Apply here. 6| Consultant Data Analyst at Swaps Monitor Location: New Delhi Responsibilities: As a consultant data analyst, you will take part in the analyst standard programs where you will enhance your financial markets, culture and country knowledge, learn about data and build relationships throughout the company’s offices in New York, London, and New Delhi. You will be analysing the information gathered and ensure that published data is accurate, drawing on extensive knowledge of the habits, patterns and culture pertaining to each financial centre. Apply here. 7| Data Analyst at AKS ProTalent Location: Delhi Responsibilities: As a data analyst, you will interpret data and analyse results using statistical techniques and provide reports. In addition, you will be involved in developing and implementing databases, data collection systems, data analytics and other strategies that optimise the statistical efficiency and quality, work with management to prioritise business and information, locate and define new process improvement opportunities. Apply here 8| Data Analyst at Cerner Location: Bangalore Responsibilities: As a data analyst, you will analyse, prepare, and process data sets to be consumed for direct insights, statistical modelling or other analytical exploration for internal and external clients. You will also access and compile data sets from various sources for exploratory and pre-defined analyses, audit data sets for completeness, validity, and other pertinent data health measures. The role will also involve evaluating data sets for analytical possibilities and pitfalls, conducting basic statistical testing and exploratory data analysis and creating basic data visualisations.","excerpt":"One of the hottest jobs of the 21st century, data science plays a crucial role when it comes to deriving meaningful insights from large chunks of data for decision-making in organisations. According to reports, as of March 2020, the analytics function in India earned consolidated revenues of $35.9 B – a 19.5% growth in revenue […]","categories":["AI Hirings"],"tags":["Applications of Data Mining","big data machine learning","data analysis India","data analyst","Data analyst jobs","Data Science Jobs","data science jobs india","deep learning projects","social network big data","what is data science"],"author_name":"Ambika Choudhury","publish_date":"2020-05-25T14:00:00","publication_year":"2020","word_count":723,"keywords":["deep learning","big data machine learning","Azure ML","data science jobs india","data science","analytics","what is data science","machine learning","AI","neural network","ML","Data analyst jobs","Data Science Jobs","Applications of Data Mining","data analysis India","data analyst","deep learning projects","Jupyter","Azure","social network big data"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","neural network","data science","analytics","Azure ML","Jupyter","Azure"],"url":"https:\/\/analyticsindiamag.com\/ai-hiring\/recent-data-science-jobs-for-freshers-to-apply-right-now\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10172339,"title":"OpenAI Exposed: iyO vs io","content":"The world wasn’t ready for what netizens have called the “most dramatic tech lawsuit of the year”. Voice computing startup iyO has filed a lawsuit against OpenAI, io, Sam Altman, and Jony Ive, alleging trademark infringement, unfair competition, and misappropriation of confidential information. OpenAI’s io announcement page has been taken down and now shows a message that reads, “This page is temporarily down due to a court order following a trademark complaint from iyO about our use of the name ‘io’. We don’t agree with the complaint and are reviewing our options.” The complaint, filed on June 9 in the Northern District of California, centres on the launch of OpenAI’s new hardware venture, io, which iyO claims is “confusingly similar” to its brand and product. Tang Yew Tan, the former Apple executive working on OpenAI’s experimental hardware project, io, has submitted a court declaration refuting allegations made by voice computing startup iyO. According to new court documents, OpenAI stated that the io announcement on May 21 was made via a blog post and video, but did not promote any specific product. “Its design is not yet finalised, but it is not an in-ear device, nor a wearable device. io has not sold any products, offered any products for sale or distribution, distributed any products, or advertised any goods or services. It has no plans to do so for at least a year.” On May 21, OpenAI announced its $6.5 billion acquisition of a new venture co-founded by Ive, named io. The announcement described plans to build a new class of personal computing devices focused on voice and AI. According to the iyO, this caused immediate market confusion due to the phonetic similarity between ‘io’ and ‘iyO,’ particularly as both companies are developing similar products. Altman’s Take OpenAI CEO Sam Altman, in a court filing, claimed he had no prior knowledge of iyO or any AI-related work by its founder Jason Rugolo when the name ‘io’ was chosen in 2023. “I was not aware that Jason Rugolo was doing anything related to AI, nor do I remember ever having seen the company name ‘iyO’,” Altman stated, adding that he was only made aware of its existence in 2025. “At no point in time did I ever intend to trade on the ‘iyO’ name.” According to Altman, Rugolo reached out unexpectedly in March, seeking $10 million in funding. Altman said he took the call as a courtesy, owing to Rugolo’s persistence, and forwarded the meeting request to the internal and external io teams to assess any potential collaboration. He clarified that he never promised an acquisition nor threatened a lawsuit. He suggested Rugolo might have “misunderstood” a mention of OpenAI’s ownership of the io.com domain. Origins of iyO and Its Product According to the court filings, iyO was formed in 2021 as a spinout from Google X (Google’s Moonshot Project) after an investment of $25 million in research and development. Since then, the company has raised an additional $37 million and developed the iyO ONE, an ear-worn, screen-free audio computer. The product is built for voice-first interaction and bone conduction audio delivery, operating entirely without a traditional screen. iyO ONE features 16 microphones to support natural language interaction. In a TED Talk last year, Rugolo introduced an ‘audio computer’—a new device that moves users away from screen-based, attention-heavy interactions and towards natural, voice-driven computing. “The truth is that we could all benefit from a little less screen time,” Rugolo said. “We need an entirely new kind of computer. One that speaks our language instead of forcing us to speak theirs.” Prior Engagement Between the Parties The lawsuit alleges that OpenAI and Ive’s design firm LoveFrom had engaged with iyO as early as 2022. These interactions began with investment interest from Apollo Projects, a fund led by Altman. IyO claims that Altman and his team reviewed detailed information about the IYO ONE and declined to invest but continued to follow the company’s progress. Tan confirmed that on April 16, Peter Welinder, VP of products at OpenAI, introduced him to Rugolo via email. Rugolo wanted to demonstrate some of iyO’s audio technology. On May 1, Rugolo and an iyO engineer met with Tan and other io employees to demonstrate the iyO ONE. The demo included voice search, music playback, and live translation. However, Tan said it was largely unsuccessful, with translation features failing repeatedly. According to Tan, Rugolo appeared to be fundraising and even offered to sell the entire company. “When I did not express interest, Rugolo pivoted, telling me that he would sell the entire iyO company for $200 million, which I understood to be roughly the amount iyO had raised to date.” Tan revealed declining again and becoming concerned about exposure to unsolicited IP. “Rugolo seemed desperate for cash…It has long been my practice since my time at Apple to avoid being ‘contaminated’ by another company’s intellectual property.” Tan said that in May 2024,  he placed a preorder for iyO’s ONE earbuds with WiFi and LTE connectivity via the company’s website. The total listed price was $1,199, with a $69 down payment. At the time, iyO promised a shipping window of Winter 2024. However, on January 13 the following year, he received an email from iyO stating that the WiFi version would be delayed until August, and that the LTE version was delayed further. As of the date of the declaration, the product was still not delivered. On the other hand, iyO said that in Spring 2025, it had more meetings with OpenAI and LoveFrom to discuss a possible investment. During those meetings, iyO shared its vision, technology, and some still-in-development ideas. Seven people from OpenAI and LoveFrom also tried out demo versions of the iyO ONE device. Even if OpenAI and Ive didn’t copy iyO on purpose, the similarities are hard to ignore. In today’s tech world, the line between inspiration and imitation gets thinner every day, and who gets credit often depends on who has more power, not better ideas.","excerpt":"iyO claims that Altman and his team reviewed detailed information about the IYO ONE and declined to invest but continued to follow the company’s progress.","categories":["Global Tech"],"tags":["OpenAI"],"author_name":"Siddharth Jindal","publish_date":"2025-06-24T19:00:15","publication_year":"2025","word_count":1000,"keywords":["Go","funding","AWS","OpenAI","AI","Aim","ViT","GAN","R","startup"],"extracted_tech_keywords":["AI","OpenAI","Aim","AWS","R","Go","GAN","ViT","startup","funding"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/openai-exposed-iyo-vs-io\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10163364,"title":"Enfabrica Launches R&amp;D Hub in Hyderabad to Boost AI Networking Innovation","content":"Enfabrica Corporation, the California-based leading innovator in AI networking, has announced the opening of its India R&D centre in Hyderabad, strengthening its global presence and expanding AI infrastructure capabilities. This move underscores the company’s commitment to scaling silicon and software product development while leveraging India’s deep talent pool. The Hyderabad centre will focus on accelerating the delivery of Enfabrica’s cutting-edge Accelerated Compute Fabric SuperNIC (ACF-S) silicon and networking software, designed to enhance the efficiency and scalability of AI infrastructure for Generative AI (GenAI) training, inference, and retrieval-augmented generation (RAG) applications. The company aims to significantly expand its workforce in India, with plans to quadruple its headcount by the end of 2025. The company’s 3.2 Terabit\/second ACF SuperNIC is engineered to be the fastest and most resilient GPU network interface controller, capable of supporting AI clusters ranging from 2,000 to over 100,000 GPUs. By optimising network design, the technology aims to enhance AI framework performance while reducing capital and operational expenses. “Computer technology has always advanced at the speed and capacity set by interconnect and scaling capabilities,” said Shrijeet Mukherjee, chief development officer at Enfabrica. “There is a tremendous opportunity for expansion in this market and Enfabrica is well positioned to help lead that evolution. We are proud to establish our very own hub in Hyderabad, leveraging the massive Indian talent pool to drive further innovation.” India’s Growing AI EcosystemIndia’s AI market is on track to reach $8 billion by 2025, growing at a compound annual growth rate of over 40% from 2020 to 2025. The country’s digital infrastructure, combined with its diverse demographic and economic landscape, presents a strategic advantage for scaling AI applications. Backed by Strong InvestmentFounded in 2020 by Silicon Valley veterans Rochan Sankar and Shrijeet Mukherjee, alongside key engineers from Broadcom, Google, Cisco, AWS, and Intel, Enfabrica has rapidly emerged as a leader in AI networking. The company recently announced the general availability of its 3.2 Tbps ACF SuperNIC and secured a $115 million Series C funding round, bringing its total funding to $260 million. With strong investor backing, Enfabrica is poised for further growth and innovation in the AI infrastructure space.","excerpt":"The company’s 3.2 Terabit\/second ACF SuperNIC is engineered to be the fastest and most resilient GPU network interface controller, capable of supporting AI clusters ranging from 2,000 to over 100,000 GPUs.","categories":["AI News"],"tags":["Hyderabad"],"author_name":"Mohit Pandey","publish_date":"2025-02-12T18:01:19","publication_year":"2025","word_count":355,"keywords":["Go","GenAI","AWS","AI","Scala","Hyderabad","RAG","Git","Aim","generative AI","R"],"extracted_tech_keywords":["AI","generative AI","GenAI","Aim","RAG","AWS","R","Go","Scala","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/enfabrica-launches-rd-hub-in-hyderabad-to-boost-ai-networking-innovation\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":16498,"title":"How are Fighter jets embracing Artificial Intelligence?","content":"Artificial Intelligence will be soon leveraged to control weapon-carrying drones within close proximity, test enemy air defenses, or perform intelligence, reconnaissance, and surveillance missions in high risk areas. To this end, some degree of AI is already engineered into F-35, the combat aircraft. According to American Air Force Chief Scientist Gregory Zacharias, the technology has been progressing quickly at the Air Force Research Lab. We might soon be able to see fighter jets with much higher degrees of autonomy and manned-unmanned teaming. “This involves an attempt to have another platform fly alongside a human, perhaps serving as a weapons truck carrying a bunch of missiles,” mentions Zacharias. The US army has already integrated manned-unmanned teaming technology in its helicopter fleet. This has helped in successfully engineering Apache and Kiowa air crews for controlling UAS flight paths and sensor payloads from the air in the cockpit. The technology has so far yielded successful combat results in Afghanistan. Furthermore, the new next-generation bomber program, the B-21 Raider, will be engineered to fly manned and unmanned missions. This is not all, the US Air Force and Boeing had flown an unmanned F-16 Falcon at supersonic speeds for the first time at Tyndall Air Force Base, back in 2013. The fighter jet was not only able to launch and maneuver, but also return to base without a pilot. Will F-35 be controlling drones? Apache helicopters in action In the near future, you’ll be able to witness pilots controlling a small group of drones flying nearby from the aircraft cockpit F-35 in the air; performing sensing, reconnaissance, and targeting functions simultaneously. Most of the flight path, sensor payload, and weapons disposal of airborne drones such as Air Force Predators and Reapers are coordinated from ground control stations today, which will soon change with use of AI. Zacharias notes, “The more autonomy and intelligence you can put on these vehicles, the more useful they will become.” Let’s put it to perspective through an instance. Without needing any assistance from ground control station, real-time video feeds could directly be relayed into an F-35 cockpit from the electro-optical\/infrared sensors on board an Air Force Predator, Reaper, or Global Hawk drone. This will help in speeding up targeting and tactical input from drones on reconnaissance missions in the vicinity of where a fighter pilot might want to attack. Moreover, drones could also be programmed to fly into heavily defended or high-risk areas ahead of manned-fighter jet. This will help to assess enemy air defenses and reduce risk to pilots. “Decision aides will be in cockpit or on the ground and more platform oriented autonomous systems. A wing-man, for instance, might be carrying extra weapons, conduct ISR tasks or help to defend an area,” mentions Zacharias. How is AI beneficial for fighter jets? Air Force Chief Scientist Gregory Zacharias Artificial Intelligence is rapidly changing the scope of what platforms can perform without needing human intervention. These AI-based systems are surfacing in the form of “decision aide support.” In other words, machines will be able to better interpret organize, analyze, and communicate information on their own, or with minimal human assistance. In this case, one human with an ability to control multiple drones and perform a command and control function can take charge of the operation, while the drones execute various tasks such as sensor functions, targeting, weapons transport, or electronic warfare activities. Multiple humans are often needed to control a single drone today, however, according to Zacharias, in future, only one human will be able to control 10, or even 100 drones. Zacharias says, “A person comes in and does command and control while having a drone execute functions. The resource allocation will be done by humans.” Moreover, algorithms will improve in such a manner that a particular drone, for say, a Predator or a Reaper, might be able to follow a fighter aircraft by itself. Such developments will enhance mission scope, flexibility, and effectiveness by enabling a fighter jet to conduct a mission with more weapons, sensors, targeting technology, and cargo. What is ALIS? How does it work? Autonomic Logistics Information System (ALIS) Autonomic Logistics Information System, or ALIS, as it is popularly called is an F-35 computer system involving early applications of AI wherein computers make assessments, go through checklists, organize information, and make some decisions by themselves, without any human intervention. It helps in transmitting aircraft health and maintenance action information to the appropriate users on a globally-distributed network to technicians worldwide. The AI-based system makes the aircraft’s logistics tail more automated for it to radio back information about engine health or other avionics. Users obtain up-to-date information on any of the areas using web-enabled applications on a distributed network. Is human control still needed? Air Force’s unmanned F-16 fighter jet completes first flight Zacharias stresses on the significance of dynamic human cognition, despite the use of advancing computer technology and increasing levels of autonomy. Computers can more quickly complete checklists and various procedures, whereas human perception abilities can more swiftly process changing information in many ways. “A computer might have to go through a big long checklist, whereas a pilot might immediately know that the engines are out without going through a checklist,” remarks Zacharias. Besides, the speed of progress in the unmanned technology landscape has convinced scientist and weapons’ developers to believe that human pilots will still be needed. This is largely because of the fact that human brain can quickly respond to unanticipated developments. Outlook Lockheed Martin F-35 Lightning II Cargo planes or bombers don’t require much of maneuvering in the skies, and will find it easier to embrace autonomous flight. However, fighter jets might still require human piloting. Undoubtedly, computer processing speed and algorithms are evolving at an alarming rate. However, it’s still a difficult task to engineer a machine that can instantly respond to other moving objects or emerging circumstances. On the other hand, sensor technology is progressing fast. Fighter pilots will increasingly be able to identify threats at much greater distances. To conclude with, it’s safe to say that there may be room for an unmanned fighter jet in the not-too-distant future, given the pace of improving autonomous technology.","excerpt":"Artificial Intelligence will be soon leveraged to control weapon-carrying drones within close proximity, test enemy air defenses, or perform intelligence, reconnaissance, and surveillance missions in high risk areas. To this end, some degree of AI is already engineered into F-35, the combat aircraft. According to American Air Force Chief Scientist Gregory Zacharias, the technology has […]","categories":["IT Services"],"tags":["Artificial Intelligence India","drones India","Logistics"],"author_name":"Amit Paul Chowdhury","publish_date":"2017-07-25T12:47:50","publication_year":"2017","word_count":1028,"keywords":["Go","API","artificial intelligence","programming_languages:R","AI","drones India","Artificial Intelligence India","programming_languages:Go","RAG","Logistics","ViT","GAN","R"],"extracted_tech_keywords":["AI","artificial intelligence","RAG","R","Go","API","GAN","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/fighter-jets-embracing-artificial-intelligence\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":42846,"title":"Bio-Inspired Evolutionary Algorithms Are Making Their Way Into AI","content":"The main motive behind every researcher working in the field of AI is to create a system that acts and makes decisions as a human does. This topic is not only vast but also controversial. A simple explanation of Artificial General Intelligence or Strong AI can be summed up in a type of artificial intelligence where the system learns generalised human abilities and later when it encounters some different and unfamiliar tasks, the system will be able to search and resolve a solution on its own, i.e. the ability of machines for generalisation and self-adaption. Researchers have explored many potentialities in the case of weak AI and are struggling hard to achieve strong AI. Recently, researchers from Oslo Metropolitan University and Norwegian University of Science and Technology developed a novel framework known as Neuroevolution of Artificial General Intelligence (NAGI). Technology Behind NAGI Generating long-term associative memory neural networks is a critical task for researchers as evolution hampers not only the connection structure of neurons but also their neurotransmitters and their local bio-inspired learning algorithms. Neuroevolution is a sub-field of AI which utilises evolutionary algorithms in order to generate ANN, evolutionary robotics is one of the examples of this form of AI. NAGI is a bio-inspired framework that uses the functionalities of spiking neurons in an evolved network structure which controls a sensory-motor system in a mutable environment. This means this framework uses the biologically-realistic models of neurons for performing computations. The researchers fused several key approaches in AI as well as evolutionary robots, such as Spiking Neural Network (SNN), Hebbian Learning, SpikeTiming-Dependent Plasticity (STDP) and NeuroEvolution of Augmenting Topologies (NEAT) to create this framework. In this framework, the researchers tried to mimic the evolution of general intelligence in biological organisms. The framework involves certain steps which are mentioned below An agent is equipped with a randomly-initialized minimal spiking neural network. The agent is placed in a mutable environment in order to be able to generalise (learn to learn), instead of merely learning to solve the specific environment. They also have access to the environment through sensory inputs where they receive intrinsic rewards and penalties. Virtual Embodied Learning (VEL) is a method to create a virtual reactive environment when no reward or penalty feedback is available through sensory input. The neural learning occurs through self-supervision via neuroplasticity which includes environmental information. New agents inherit the topologies of the controllers from the previous generation and the  goal of the untrained inherited controllers is to possess a topology that supports the ability to learn new environments. And hence the outcome is an unsupervised evolving system which learns without explicit training in a self-supervised manner through embodiment. Benefits The concept of artificial general intelligence is still science fiction and once the researchers accomplished to gain the full potential of this technology, there will be numerous advantages into the modern world one can ever think of. With the advent of this technology, there will be less human labour, common issues like violence, corruption will be decreased, extreme autonomous researches, etc. Outlook In the present scenario, techniques such as deep learning, reinforcement learning have successfully accomplished in solving various real-time challenging tasks. However, these techniques face some limitations such as data-feeding is one of the crucial factors. According to the researchers, these limitations can be overcome by the technique of AGI. Currently, the NAGI framework is in its simplest form and the researchers will further extend this into more complex tasks and environments. Read the paper here.","excerpt":"The main motive behind every researcher working in the field of AI is to create a system that acts and makes decisions as a human does. This topic is not only vast but also controversial. A simple explanation of Artificial General Intelligence or Strong AI can be summed up in a type of artificial intelligence […]","categories":["AI Features"],"tags":["Artificial General Intelligence"],"author_name":"Ambika Choudhury","publish_date":"2019-07-18T12:01:40","publication_year":"2019","word_count":581,"keywords":["Go","artificial intelligence","programming_languages:R","AI","neural network","ML","programming_languages:Go","deep learning","Artificial General Intelligence","GAN","R"],"extracted_tech_keywords":["AI","artificial intelligence","ML","deep learning","neural network","R","Go","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/bio-inspired-evolutionary-algorithms-are-making-their-way-into-ai\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":28059,"title":"Calling Media Folks And Bloggers To Cypher 2018","content":"It’s time for the analytics community in India to come together. Cypher provides a unique platform to discuss ideas, celebrate the progress in data science, artificial intelligence and big data. Together we must find a way to deal with challenges, boost the startups ecosystem, and network with an audience that is the most definitive representation of the Indian analytics industry. At the fourth edition of Cypher 2018, we are hosting the analytics community to meet, interact, engage and dive deeper into analytics use cases that have made an impact across all industries. It brings an opportunity to meet innovators, and provides a platform to showcase the expertise of companies and startups who are building solutions to solve complex business problems. We would like to invite journalists and bloggers covering technology space to witness the biggest gathering of Analytics and AI Leaders, innovators and influencers over three days across four parallel sessions at the largest analytics summit in India. We would like you to come and interact with the industry leaders and startups in the analytics and AI space. For media registrations, please write at info@analyticsindiamag.com Why Attend Cypher 2018? We are hosting an excellent line up of speakers such as Devi Shetty, chairman at Narayana Hrudayalaya; Kamal Kumar, Head of analytics at Jabong; Anshu Sharma, SVP technology infrastructure at Vodafone; Amit Soni, Group Head of analytics at Piramal Enterprises; Atul Jalan, CEO and MD at Manthan; and Syed Rahim, Head of customer experience at SPAR India, among others. Numerous conference segments such as keynotes, panel discussion, knowledge talks, workshops are going to be held at the event. Cypher 2018 will witness a host of attendees ranging from data scientists, analytics professionals, chief analytics officers, head of analytics, analytics startups, VCs and analytics aspirants, among others. We would recognise the best in data science industry by hosting the Great Learning Data Science Awards 2018 in association with Great Learning Institutes. Awards would be given in four categories — Emerging data science startup of the year, Best data science project of the year, Best machine learning\/AI implementation and Data science for social good. We would be hosting analytics and AI quiz, called the AI.Q. Screening for our AI documentary, ‘Transition cost’ that has a central theme of autonomous vehicles in India. You can check our complete agenda here. Date of the event: 26-28 September 2018 Venue: Hotel Radisson Blu (Previously Park Plaza), Outer Ring road, Bengaluru, India","excerpt":"It’s time for the analytics community in India to come together. Cypher provides a unique platform to discuss ideas, celebrate the progress in data science, artificial intelligence and big data. Together we must find a way to deal with challenges, boost the startups ecosystem, and network with an audience that is the most definitive representation […]","categories":["Deep Tech"],"tags":[],"author_name":"Srishti Deoras","publish_date":"2018-09-06T04:48:39","publication_year":"2018","word_count":405,"keywords":["big data","data science","Go","machine learning","artificial intelligence","AI","Ray","ViT","analytics","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","data science","analytics","Ray","R","Go","big data","ViT"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/calling-media-folks-and-bloggers-to-cypher-2018\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10057587,"title":"Legally speaking &#8211; Artificial Intelligence is not even close to human intelligence","content":"In public proceedings, the Legal Board of Appeal of the EPO confirmed that under the European Patent Convention (EPC), an inventor designated in a patent application must be a human being. This was the judgement in combined cases J 8\/20 and J 9\/20, where the board just dismissed the applicant’s appeal. Here, both the applications were made by a Missouri physicist Stephen Thaler, whose AI-system DABUS had made the inventions. Device for the Autonomous Bootstrapping of Unified Sentience, or DABUS, is a computer system programmed to invent by itself. It is, basically, a swarm of disconnected neutral nets that can continuously generate thought processes and even memories that can, over time, generate new and inventive outputs independently. The Legal Board of Appeal also said that only a ‘natural person’ was indicated to have “the right to the European Patent by virtue of being the owner and creator of” the artificial intelligence system. It stated that designating a machine as an inventor does not comply with their requirements as per Article 81 and Rule 19(1) EPC. Also, a machine cannot transfer any rights to the applicant. The two patents by DABUS include a design of a container based on “fractal geometry.” This is claimed to be the best shape for containers being stacked together, and that can be handled by robotic arms. The second one is for a “device and method for attracting enhanced attention.” This is a light that flickers rhythmically in a pattern that mimics human neural activity. The DABUS applications have sparked a lot of deliberation in intellectual property offices and even courtrooms worldwide. The applications list DABUS as the inventor, but Thaler is the owner of the patent, which makes it clear that DABUS AI is not up for property rights. Successes and Failures Thaler found success when the South African patent office granted the patent DABUS for the novel way of using ‘fractal geometry’ and better design for food containers which improved grip and heat transfer. South Africa’s decision did receive backlash from intellectual property experts across the world, but experts also believe that the decision supports the government’s policy that aims to increase innovation and develop the technology. This patent was published in July 2021 in the South African Patent Journal. In Australia, the patent office first dismissed the application. Thaler then appealed in court, and in a landmark decision, the Australian court decided that artificial intelligence (AI) systems could be legally recognised as inventors in patent applications. Justice Johnathan Beach of the Australian Federal Court became the first to hand down the judgment in favour of Thaler, which said that ‘an inventor … can be an AI system or device’. Currently, the United States Patent and Trademark Office, as well as the European Patent Office, have rejected these applications. They are very clear with their patent laws which state that only human inventors — not AI — as they are called “him” and “her” can receive a patent. Also, for the purposes of a patent, “mental conception” is an essential element, which only a human mind can have. Third, inventorship has rights, which AI cannot legally possess. The US District Court for the Eastern District of Virginia also ruled that AI can’t be listed as an inventor on a patent as per the laws. Judges in the UK also stated that only “natural persons” can be inventors. Why the maker of AI is not open to receiving the patent There has always been some kind of creativity machines. In fact, before DABUS, Thaler had built an AI that created novel sheet music, which he credited with inventing the cross-bristle toothbrush design. He filed a patent for this design where he proved AI’s ability to generate novel inventions that can meet patent standards for patents. But here, Thaler listed himself as the inventor and not the AI. For the food container invention by DABUS, Thaler decided to list DABUS instead, as the invention was devised by the AI. This was the beginning of the push for AI to be recognised as inventors. Thaler is very clear that he cannot be listed as the inventor on the applications as that would mean taking credit for inventions that aren’t his. In fact, he also believes that listing DABUS as the inventor also protects the rights of human inventors and prevents a person from falsely claiming credit. He also wants to avoid inaccurately listing himself as an inventor as that could expose him to some criminal penalties and cause the patent to become invalid or have unenforceable outcomes. Wrapping up US patents are considered the most valuable and credible worldwide, and with the US rejecting DABUS applications, experts believe that this can hamper investments in the country. Also, since other countries have granted DABUS patents, the US might put itself at a competitive disadvantage. In this scenario, most countries currently have patent laws that revolve around four points: An inventor must be an “individual”‘Individuals’ have to be people (not even, say, companies)Patent Act only refers to peopleAI systems are not people With technology evolving very fast, there may be a time when AI will become so sophisticated that globally, laws can change, and AI can receive the status of inventors. But, that time hasn’t arrived yet. With discussions and deliberations going on globally in patent offices and courts, there might be some major changes that patent laws might see – for good or for bad.","excerpt":"AI inventions are ready to get patents, but legally, they are not ‘natural humans’ and cannot receive any rights. With DABUS receiving a fresh rejection from EPO, the debate continues.","categories":["AI Features"],"tags":["AI patents","AI Rights","Patent"],"author_name":"Meeta Ramnani","publish_date":"2022-01-04T10:00:00","publication_year":"2022","word_count":907,"keywords":["Go","artificial intelligence","TPU","AI patents","AI","AWS","innovation","cloud_platforms:AWS","Patent","Aim","ViT","AI Rights","R"],"extracted_tech_keywords":["AI","artificial intelligence","Aim","AWS","TPU","R","Go","ViT","innovation","cloud_platforms:AWS"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/legally-speaking-artificial-intelligence-is-not-even-close-to-human-intelligence\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10054140,"title":"A Guide to VARMA with Grid Search in Time-Series Modelling","content":"Finding the best values of a machine learning model’s hyperparameters is important in order to build an efficient predictive model. In time-series modelling also, finding optimal values of the hyperparameters of a model is necessary for accurate forecasting. Grid search is a popular technique for such purposes.  In this article, we will discuss how to use the grid search technique with a VARMA model in time series modelling for multivariate time series analysis. The major points that we will discuss here are listed below. Table of Contents What is Grid Search?What is VARMA?VARMA with Grid SearchImplementing VARMA with Grid Search What is Grid Search? In machine learning, most of the models like the random forest, decision trees, and support vector machines include parameters in their algorithm which are required to be set on different positions so that the model can perform better according to the data. The parameters under the model algorithm can be integer, float, or boolean type, and finding a set of parameters that can be utilized for making the model perform better than the other becomes a task of modelling. In such a scenario the grid search comes into the picture which can be used for finding a set of parameters by iterating on different sets of parameters by the time of making models. We can define the grid search as the process of performing hyperparameter tuning so that we can determine an optimal set of parameters for a given model like random forest and decision tree. The example of parameters can be the number of trees in the forest, criterion(“Gini”, “entropy”) and the maximum depth, etc. there is always mathematics behind the algorithm of any model and the parameters are the values that make the mathematics to work according to the task given to the algorithm. When we talk about the working of the grid search it works on predefined values which are assigned for the parameters and help us to find the best fit set of parameters for a given model and data. Let’s say that we have defined a dictionary with some value under it. These values consist of mentioned parameters along with the values the parameter can take. For example: { 'C': [0.1, 1, 10, 100], 'gamma': [1, 0.1, 0.01, 0.001], 'kernel': ['rbf',’linear’,'sigmoid']  } In this example, c, gamma, and kernels are parameters, and values are assigned to the parameters. So when we apply the grid-search with the dictionary on the model it will try all the combinations of the values passed and evaluate the model on the data using all combinations. Lastly, we will gel the metrics like accuracy and loss for every combination so that we can choose which set of parameters is an optimal set of parameters for the model. In this article, we will see that how can we use the grid search on the VARMA model which is a time series model and we use it for time series analysis when the time series given for forecasting is multivariate before going on the implementation we are required to have a formal discussion on the VARMA model. What is VARMA? In time series analysis we see many of the models like AR, ARIMA, VAR, etc. similarly the VARMA model is a time series model, where the full form of the model is Vector Auto-Regressive Moving Average specially used in cases where the time series is multivariate and the meaning of the multivariate time series is the time series where the future values of the time series depend on the multiple variables. When we talk about the VARMA model we can say that it is a combination of VAR(Vector Auto-Regressive) and VMA(vector moving Average) models. VAR and VMA models can also be used for multivariate time series modelling. Where VAR models can be considered as the generalization of the AR(auto Regressive) models where we assume that the time series is stationary and the p-value which stands for lag order is used for modelling. The VMA model is a generalization of the Moving Average model where again the time series is stationary but q values which stand for order of moving average are used for modelling. So if we combine the VAR and VMA  it becomes a VARMA model where both lag order and order of moving average (p and q) is used in the modelling and also we can convert the VARMA model in VAR by setting the moving average as zero and in VMA by setting the lag order as zero in the modelling. To know more about the VARMA model reader can go through this link. VARMA with Grid Search The main focus of the article is to implement a VARMA model using the Grid search approach. Where the work of grid search is to find the best-fit parameters for a time-series model. As we can understand, manually searching for the best fit set of parameters for a model becomes time taking and costly. We can use the grid search and in time series if we talk about the parameters we have the following parameters. p: Stands for the number of lag observations included in the model, also known as the lag order. d: The number of times the raw observations are differentiated, also called the degree of difference. q: Is the size of the moving average window and also called the order of moving average. So our major aim in this article is to find the best fit set of parameters for a VARMA model using the Grid search method. Let’s start with the implementation of the procedure. To reduce the size of the article I am not posting the codes in the article. For the codes, we can go through this Google Colab notebook. Implementing VARMA with Grid Search For this modelling procedure, we require a multivariate time series. Dow_jones_Industrial_average data is a multivariate time series where we have an opening, closing, and high and low information of the shares of the Dow Jones industries. You can find similar data using this link. The below image represents the details about the data as: Here in the above images, we can see information about the time series. Before going to the modelling we are required to perform some mandatory tasks such as checking for the stationarity of the data. We can use the Dickey-Fuller test for checking the stationarity of the time series. since in a multivariate time series, we have multiple time series we are checking the stationarity for every time series. In the results of the dickey fuller test, we measure the p-value. If the p-value is more than 0.02 we can consider the time series as the non-stationary time series. As we have discussed, the major requirement of modelling a VAR model is a stationary time series. Also, we can make the time series stationary using the following method. Detrending Seasonal adjustmentTransformation We can apply the differencing method from the transformation methods to make the time series stationary. In this article, we used a similar differencing method for making the time series stationary, Below are the results of the Dickey-Fuller test after applying the differencing method on the time series. We can see now that the time series we are using for modelling is stationary. Next we are required to know about the correlation between the time series  in our time series so that we can model more accurately by understanding the effect of time series on each other. Here we can see the correlation between the time series and now we can make a dictionary of parameters for grid search as shown above. The parameters in the varma models are p, q and d. The above-given image is a representation of the different combinations of parameters that we are going to apply for modelling and will find the best-fit combination of parameters for modelling a VARMA model. We can find an optimal set of parameters by checking the RMSE score after fitting them into the model. The lower RMSE represents the best result. Let’s take a look at the below image. In this image, I have sorted the data frame according to the lowest RMSE and we can see over here that the best-fit combination of the parameters is (p = 3, q = 3, tr = n) because it gives the least RMSE values. Now we can fit these parameters on the model and predict the values. The below image is the evaluation metrics for every time series in the multivariate time series. Here the below graphs are representations of the quality of the predictions which we have made using the model. Here we can see that the model is working fine. We have predictions that are almost near to the real values presented in the test data. Final Words Here in the article, we have got an overview of the grid search and VARMA models. Along with this, we have also seen how we can use the grid search in the VARMA model that is a time series model. We understood how to find the optimal parameters for a time series model using this approach.","excerpt":"The main focus of the article is to implement a VARMA model using the Grid search approach. Where the work of grid search is to find the best-fit parameters for a time-series model.","categories":["Deep Tech"],"tags":["AI (Artificial Intelligence)","Data Science","Data Scientist","Deep Learning","Google","Guide","Machine Learning","Python","Time Series"],"author_name":"Yugesh Verma","publish_date":"2021-11-24T16:00:00","publication_year":"2021","word_count":1524,"keywords":["RPA","Time Series","R","RAG","Data Science","Guide","Go","machine learning","AI","Machine Learning","programming_languages:R","programming_languages:Go","Python","Colab","Aim","Google","Deep Learning","Data Scientist","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","machine learning","Aim","Colab","RAG","R","Go","RPA","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/a-guide-to-varma-with-grid-search-in-time-series-modelling\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10040757,"title":"How TensorFlow Lite Fits In The TinyML Ecosystem","content":"TensorFlow Lite has emerged as a popular platform for running machine learning models on the edge. A microcontroller is a tiny low-cost device to perform the specific tasks of embedded systems. In a workshop held as part of Google I\/O, TensorFlow founding member Pete Warden delved deep into the potential use cases of TensorFlow Lite for microcontrollers. Further, quoting the definition of TinyML from a blog, he said: “Tiny machine learning is capable of performing on-device sensor data analytics at extremely low power, typically in the mW range and below, and hence enabling a variety of ways-on-use-case and targeting battery operated devices.” A Venn diagram of TinyML showcasing the composition of TinyML (Source: Google I\/O) How is TinyML different? Most machine learning applications are resource-intensive, and expensive to deploy and maintain. According to PH Data, $65K (INR 47 lakhs) is the bare minimum amount required to deploy and maintain a model over 5 years. As you build a scalable framework to support future modeling activities, the cost might escalate to $95K (INR 70 lakhs) over five years. On the other hand, TinyML is quite flexible and simple and requires less power. Each hardware component (mW and below) acts independently, and the storage capacity of most machine learning models barely exceeds 30kb. Also, data can be processed on the device locally, which inevitably reduces data latency and solves the data privacy issue. Arm, Arduino, Sparkfun, adafruit, Raspberry Pi, etc are the major players in TinyML. TensorFlow Lite, an open-source library by Google, helps in designing and running tiny machine learning (TinyML) models across a wide range of low-power hardware devices, and does not require much coding or machine learning expertise, said Warden. Benefits of TinyML: Really small form factors enable multiple use cases Cheaper devices make ML more accessible Low battery consumption means devices can run for much longer Data processing can be done on the device (no cloud connection required) How does TinyML work? The TinyML process works in four simple steps — gather\/collect data, design and train the model, quantise the model and deploy to the microcontroller. Source: Google I\/O In a blog post, ‘TensorFlow Lite for Microcontrollers,’ Google has explained some of its latest projects that combine Arduino and TensorFlow to create useful tools: Air snare: Play the drums in the air    Finger user interface (FUI): Control your devices with the wave of a finger Tiny motion trainer: A code-free tool that lets you create custom, microcontroller-ready models based on IMU data To initiate the project, you need a TF4 Micro Motion Kit pre-installed on Arduino. Once you have installed the packages and libraries on your laptop or personalised computer, look for the red, green and blue flashing LED in the middle of the board. The details of the setup are found here. Once the setup is complete, you need to connect the device via Bluetooth; the TF4Micro motion kit communicates with this website via BLE, giving you a wireless experience. Now, tap the button on your Arduino, then wait for the red, green, and blue LED pattern to return. After this, click the ‘connect’ button as shown on the website, then select TF4Micro Motion Kit from the dialogue box. You are now good to go. Similar steps need to be followed for all three experiments — Air snare, FUI and tiny motion trainer. Note: Do not hold the button down as this will clear the board. The above experiments will help you get a hang of TensorFlow Lite on microcontrollers. You can also submit your ideas to the TensorFlow Microcontroller Challenge and win exciting cash prizes. As part of a TensorFlow Microcontroller Challenge, Sparkfun is giving out a free TensorFlow Lite for Microcontrollers Kit. Click here to get yours.","excerpt":"TensorFlow Lite has emerged as a popular platform for running machine learning models on the edge. A microcontroller is a tiny low-cost device to perform the specific tasks of embedded systems. In a workshop held as part of Google I\/O, TensorFlow founding member Pete Warden delved deep into the potential use cases of TensorFlow Lite […]","categories":["AI Features"],"tags":[],"author_name":"Amit Naik","publish_date":"2021-05-24T19:00:00","publication_year":"2021","word_count":622,"keywords":["Go","machine learning","AI","ML","Scala","RAG","ViT","analytics","TensorFlow","R"],"extracted_tech_keywords":["AI","machine learning","ML","analytics","TensorFlow","RAG","R","Go","Scala","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-tensorflow-lite-fits-in-the-tinyml-ecosystem\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10162028,"title":"Automation Anywhere’s AI to Power PwC India’s Spend Suite","content":"Automation Anywhere, a leader in Agentic Process Automation (APA), on Monday joined forces with PwC India to enhance its Intelligent Spend Management Suite (ISMS) with advanced AI capabilities. The collaboration is aimed at improving how businesses manage spending and process payables. The ISMS platform, developed by PwC India, will now use Automation Anywhere’s generative AI-powered agents. This upgrade is designed to enable smarter purchasing decisions, automate accounts payable processes, and deliver improved business outcomes. PwC India’s ISMS features a modular design for easy integration with existing systems, an API-first architecture for scalability, and a robust data foundation for validation and workflows. The improvements aim to make operations more efficient, offer practical insights, and help people make better decisions. Talking about the significance of this collaboration, PwC India chairperson Sanjeev Krishan said, “The integration of Automation Anywhere’s advanced Agentic AI capabilities into PwC India’s Intelligent Spend Management Suite represents a step forward in addressing some of the most pressing challenges faced by businesses today.” He further mentioned that by combining PwC’s deep industry insights, AI and tech expertise and domain understanding with Automation Anywhere’s agentic AI platform, the company is empowering organisations across sectors. This will help drive efficiency, enhance decision-making, and unlock greater value in their spending and payable operations. Joe Atkinson, PwC’s global chief AI officer, emphasised the transformative potential of Agentic AI. “AI is starting to disrupt value chains across whole industries, and with ISMS, value realisation for businesses becomes real. I’m excited to see the collaboration between us and Automation Anywhere,” he said. This collaboration builds on the 2024 strategic partnership between Automation Anywhere and PwC India to create AI-driven business solutions. Automation Anywhere COO Ankur Kothari expressed confidence in the partnership, saying, “Our APA platform has the unparalleled ability to drive actions by solving real-world customer problems with GenAI. “We recognise PwC India’s market potential, which will significantly increase the deployment of AI agents, aiding in a strategic long-term vision to deliver superior business outcomes for our customers.”In October last year, PwC India also announced its strategic collaboration with Meta in order to expand and scale its open-source AI solutions to enterprises and citizen services using Meta’s Llama models.","excerpt":"PwC’s ISMS platform will now use Automation Anywhere’s generative AI-powered agents.","categories":["AI News"],"tags":["Automation Anywhere","pwc"],"author_name":"Shalini Mondal","publish_date":"2025-01-23T12:36:33","publication_year":"2025","word_count":363,"keywords":["API","GenAI","agentic AI","AI","Scala","automation","Automation Anywhere","Aim","generative AI","pwc","GAN","R"],"extracted_tech_keywords":["AI","generative AI","GenAI","agentic AI","Aim","R","Scala","API","GAN","automation"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/automation-anywheres-ai-to-power-pwc-indias-spend-suite\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":51824,"title":"How This 5-Year-Old Bootstrapped Startup Is Helping 20,000 Enterprise Users Gain Actionable Intelligence","content":"The size of the IoT market is set to touch $9 billion by 2020, the majority of which comes from the use of technologies like AI, IIoT, ML in industries like telecom, health, vehicles, homes, smart cities, and manufacturing floors, among others. AI-based image analysis helps in a number of domains nowadays. Enterprise automation, computer vision, manufacturing, retail, and virtual reality-based e-commerce platforms have been the early adopters of these AI and ML-based solutions. With such a vision, Bangalore-based AI startup called Integration Wizards is utilising artificial intelligence and machine learning to help enterprises determine actionable intelligence from images, videos, and live data. Founded in July 2014 by three techies — Kunal Kislay, Saquib Khan, and Kumar Raman, AI startup Integration Wizard offers computer vision and enterprise mobility-based solutions. The solutions are being used by more than 20,000 enterprise users across 16 countries and in six languages, across many verticals including pharmaceuticals, oil and gas, retail and consumer goods. Flagship Product IRIS is an AI-based video analytics solution which provides actionable insights. The platform enables customers to onboard cameras quickly, define use cases and analyse the information. The key functions of IRIS are to monitor, control and automate devices when it is plugged in CCTV networks. The product plugs into the existing CCTV network and is trained on custom scenarios based on client requirements. IRIS analyses live feed to provide real-time alerts on email, push notifications, SMS and WhatsApp. IRIS offers IIoT, AI and ML, enterprise mobility, through a single platform. It is used to monitor manufacturing, safety compliance, process optimization in warehouses, etc. and is also being deployed for outdoor security solutions and retail for various use cases. Use of AI\/ML IRIS is based on NVIDIA Deepstream for enabling fast real-time processing of inbound video streams. The product manages deployments at scale using MS Azure’s IoT platform which enables seamless communication and management through Azure IoT Hub and Azure IoT Edge solutions. The company also has custom hardware built on various NVIDIA GPUs which supports various platforms for different scales of deployment. The product is specialised in building complex custom models to detect objects and activities. The models deployed range from detecting fire, people, PPE (Helmet, Vests, safety shoes, etc) to the more nuanced task of detecting efficiency, postures, emotions, etc. Recent Funding Talking about recent funding, Kislay said, “We are bootstrapped and have been profitable since our inception and are projected to make over $3.2 million at the end of the current fiscal year. Our market strategy is centred around the quality of our solutions, mutual trust with our clients and constant innovation to give them a one-stop solution for the challenges they face.” Hiring Phase Integration Wizards focus on hiring young talent and help them gain the necessary technical training to enable them to contribute in very little time. The company looks at the technical knowledge of the candidates as well as if they will fit in with the company culture. Potential Competitors According to Kislay, though Integration Wizards is a startup with unique offerings, he considers his competitors as companies which focus on computer vision, such as Cortexica, SightHound, Pilot.ai, and Wobot.ai, among others. Roadmap Talking about the roadmap, Kislay said that currently, the company is working towards expanding the installed base in warehouses, manufacturing premises, and retail outlets. By 2020, it aims to be a $15 million organisation by revenue and by Q3 in 2020 Integration Wizards also aims to onboard developers from other organisations to build use-cases on the platform suiting their needs.","excerpt":"The size of the IoT market is set to touch $9 billion by 2020, the majority of which comes from the use of technologies like AI, IIoT, ML in industries like telecom, health, vehicles, homes, smart cities, and manufacturing floors, among others. AI-based image analysis helps in a number of domains nowadays. Enterprise automation, computer […]","categories":["AI Startups"],"tags":["Startups"],"author_name":"Ambika Choudhury","publish_date":"2019-12-13T16:30:00","publication_year":"2019","word_count":589,"keywords":["Go","machine learning","artificial intelligence","AI","R","ML","computer vision","Aim","analytics","Startups","Azure"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","computer vision","analytics","Aim","Azure","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-startups\/how-this-5-year-old-bootstrapped-startup-is-helping-20000-enterprises-gain-actionable-intelligence\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10063437,"title":"What are the graph objects in Plotly and how to use them?","content":"Plotly is an open-sourced Python library used for visualizing data by creating interactive and exquisite plots. Graph objects are pictorial representations of systems of objects interconnected by links. The plotly.graph_objects module contains a hierarchy of Python classes that represent non-leaf nodes in this figure schema. “Graph objects” are instances of these classes. In this article, we will look at how to use graph objects (module) of the Plotly library, explaining it from the ground up, and covering all of the most commonly used charts. The following are the points and plots that this article will cover. Table of contents Why use graphic objects over Plotly express?Description of dataVisualization with Plotly graph objectsBar plotScatter plotSubplots Let’s start with the advantages of graphic objects over Plotly express. Why use graphic objects over Plotly express? Plotly’s library uses graph objects in the background to produce its figures unless manual construction is done from dictionaries. Plotly express has many advantages but the question is why to use the graphic object module. Let’s see why: Figures that use certain 3D trace-types, such as mesh or isosurface, are not yet possible with Plotly Express.  Plotly express uploads the data on its server to return the graphical representation whereas in graph object there is no such thing happening. Are you looking for for a complete repository of Python libraries used in data science, check out here. Description of data The data used in this article is about the end of the day nifty 50 stock prices with a total of 13 features related to the stocks. This data has been taken from the kaggle repository which is mentioned in the references. The description of the features is listed below. Symbol: Name the stock Open: The opening price of the particular stock when the market openedHigh: The highest price of the stock within the day Low: Lowest price of the stock within the dayLTP: Last Traded Price stands for the price of a stock on which the last transaction or trade occurred.Chng: Amount of change in the  stock price% Chng: Percentage of changeVolume: Total amount of stock trading for the dayTurnover: Total turnover of the company52w H: the highest price that the security\/ stock has traded over 52 weeks i.e. a year52w L: the lowest price that the security\/ stock has traded over 52 weeks i.e. a year365 d % chng: percentage of change on the stock in 365 days (year)30 d % chng: percentage of change on the stock in 30 days (month) Sample Visualization with Plotly graph objects Visualize the data with the graphic object module of plotly and learn how to use different classes of this module. Importing libraries: import pandas as pd import numpy as np import plotly.graph_objects as go Reading dataset and preprocessing: df=pd.read_csv('Nifty50.csv') df[['52w H','52w L','Open', 'High', 'Low']]=df[['52w H','52w L','Open', 'High', 'Low']].replace(\",\",\"\",regex=True) df[['LTP','Turnover (crs.)']]=df[['LTP','Turnover (crs.)']].replace(\",\",\"\",regex=True) df[['52w H', '52w L','Open', 'High', 'Low', 'LTP','Turnover (crs.)']]=df[['52w H', '52w L','Open', 'High', 'Low', 'LTP','Turnover (crs.)']].astype(float) df.head() Dataset needed some preprocessing there were commas (,) in the values so needed to replace those as well convert all the values to float for using them in the data visualization. Before preprocessing: After preprocessing: Bar plot fig=go.Figure() fig.add_trace(go.Bar(x=df['Symbol'], y=df['Volume (lacs)'], name ='Total volume', visible=True )) fig.show() Step 1: Defining the blank figure using go.Figure() class and the figure stored in a variable name “fig”. Step 2: Now need to add a plot to the blank figure stored in the variable, for which the add_trace() class is used. Inside the trace, the class adds the plot which is to be plotted. So, this bar plot is given as an input, then defining the x-axis, y-axis, name of the plot and other parameters. Step 3: Now call the variable in which the figure was stored and display the plot by using the show() function. Now the graph is plotted but it does not look like this has a grid format background, no x-axis label or y-axis label, no title. So, for adding all this content to the graph we need to update the layout of the plot. Step 4: Use the update_layout() class to update any parameter of the graph. fig.update_layout(showlegend=False, plot_bgcolor='rgba(0,0,0,0)', font=dict(family='Arial', size=12, color='black'), xaxis = dict(showgrid=False,tickangle = -45,categoryorder='total descending',title_text='Name of stocks'), yaxis = dict(title_text='Total amount of stocks traded'), title=dict(text = 'Total volume of stock EOD', font=dict(family='Arial', size=18, color='black'), x=0.5, y=0.9, xanchor='center', yanchor='top'), ) In the above code, the background colour is changed to white, font is set for the text shown in the graph, x-axis and y-axis are described with their properties like the angle at which the labels should be presented and name of axis, and at last, the title of the plot is added with some properties of like font and positioning. So, the final code will look like this: fig=go.Figure() fig.add_trace(go.Bar(x=df['Symbol'], y=df['Volume (lacs)'], name ='Total volume', visible=True )) fig.update_layout(showlegend=False, plot_bgcolor='rgba(0,0,0,0)', font=dict(family='Arial', size=12, color='black'), xaxis = dict(showgrid=False,tickangle = -45,categoryorder='total descending',title_text='Name of stocks'), yaxis = dict(title_text='Total amount of stocks traded'), title=dict(text = 'Total volume of stock EOD', font=dict(family='Arial', size=18, color='black'), x=0.5, y=0.9, xanchor='center', yanchor='top'), ) fig.show() Now the plot is sorted in descending order, it has x labels & y labels and a title is added to the plot. As we have seen the plotting of bar plots similarly let’s see the plotting of other basic plots. Scatter plot Plotting a scatter plot with lines connecting those data points to see the variation of the highest and lowest price of the stocks at the end of the day. fig=go.Figure() fig.add_trace(go.Scatter(x=df['Symbol'], y=df['52w H'], mode='lines+markers', name='high')) fig.add_trace(go.Scatter(x=df['Symbol'], y=df['52w L'], mode='lines+markers', name='low')) fig.update_layout(showlegend=True, plot_bgcolor='rgba(0,0,0,0)', font=dict(family='Arial', size=12, color='black'), xaxis = dict(showgrid=False,tickangle = -45,title_text='Name of stocks'), yaxis = dict(title_text='Price of stocks'), title=dict(text = 'Variation between highest and lowest stock price', font=dict(family='Arial', size=18, color='black'), x=0.5, y=0.9, xanchor='center', yanchor='top'), ) fig.show() Pie chart Let’s see the percentage of annual change in the stock price with the help of a pie chart. fig=go.Figure() fig.add_trace(go.Pie(labels=df['Symbol'], values=df['365 d % chng'], name=\"Change in year\")) fig.update_traces(textposition='inside') fig.update_layout(uniformtext_minsize=12, uniformtext_mode='hide', title=dict(text = 'Percentage of annual change in stock price', font=dict(family='Arial', size=18, color='black'), x=0.5, y=0.9, xanchor='center', yanchor='top')) fig.show() Subplots Now let’s plot the above scatter plot and a bar plot of the percentage of annual change in the stock with respect to the annual high and low of stock prices. To plot this kind of chart, you need to create a subplot with two rows and one column. Let’s see how to display multiple plots in one plot with the help of subplots. import plotly.subplots as splt fig_sub = splt.make_subplots(rows=2, cols=1, row_heights=[0.7,0.9], subplot_titles=(\"Annual High and Low for stocks\", \"Percentage of annual change in stocks\")) fig_sub.add_trace(go.Scatter(x=df['Symbol'], y=df['52w H'], mode='lines+markers', name='high'), row=1, col=1) fig_sub.add_trace(go.Scatter(x=df['Symbol'], y=df['52w L'], mode='lines+markers', name='low'),row=1, col=1) fig_sub.add_trace(go.Bar(x=df['Symbol'], y=df['365 d % chng'], name ='% change\/annum', visible=True),row=2, col=1 ) fig_sub.update_xaxes(tickangle = -45, row=1, col=1) fig_sub.update_xaxes(tickangle = -45, row=2, col=1) fig_sub.update_layout(showlegend=True, plot_bgcolor='rgba(0,0,0,0)', font=dict(family='Arial', size=12, color='black'), title=dict(text = 'Annual varation in stocks', font=dict(family='Arial', size=18, color='red'), x=0.5, y=0.9, xanchor='center', yanchor='top'), height=550 ) fig_sub.show() fig_sub.show() The subplot with two rows and one column by using the make_subplot() class of subplots module and store it in a variable(‘fig_sub’). The variable in which it is stored ‘fig_sub’ is used for updating and adding traces same as explained earlier. Final verdict The graph object module is the backbone of every graph produced by Plotly. In this article, we have learned about when to use graph objects and we also went through their implementation to create interactive and interesting visuals with the help of data. References Graph object module documentationLink for the above codeDataset","excerpt":"In this article, we will look at how to use graph objects (module) of the Plotly library, explaining it from the ground up, and covering all of the most commonly used charts.","categories":["AI Trends"],"tags":["Data Visualisation","Machine Learning","plotly","Python"],"author_name":"Sourabh Mehta","publish_date":"2022-03-24T17:00:00","publication_year":"2022","word_count":1241,"keywords":["data science","NumPy","Go","Plotly","programming_languages:R","AI","Data Visualisation","Machine Learning","plotly","Python","programming_languages:Python","R","Pandas"],"extracted_tech_keywords":["AI","data science","Pandas","NumPy","Plotly","Python","R","Go","programming_languages:Python","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/what-are-the-graph-objects-in-plotly-and-how-to-use-them\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10019866,"title":"Hands-On Python Guide To AllenAct &#8211; An Embodied AI Project","content":"Table of contentsIntroductionCaptivating features of AllenActInstalling AllenActPractical implementationWhat is a Point Navigation task?What is an environment?Simulator usedAllenAct’s ‘Environment’Algorithm usedInstall RoboTHOR environmentUnderstand the dataset usedCodeModel trainingModel testingEnd note Introduction AllenAct is an open-source project and a modular learning framework designed for researchers and technophiles associated with the domain of Embodied AI. It provides state-of-the-art reproductions of numerous embodied AI models. It also supports the expanding collection of embodied AI tasks, algorithms used to accomplish those tasks as well as environments to run them efficiently. Are you unfamiliar with the term ‘Embodied AI’? Refer to the ‘Overview of Embodied AI’ section of this article before proceeding! AllenAct was introduced by members of the PRIOR (Perceptual Reasoning and Interaction Research) group at the Allen Institute for Artificial Intelligence (AI2) in August, 2020. AI2 is a non-profit research institution headquartered at Seattle, Washington (United States). It was founded by a co-founder of Microsoft, Paul G. Allen in 2014 with an aim of conducting influential research in the field of AI. (Check out Paul Allen’s personal website here). AI2 is now being led by a leading AI researcher, Dr. Oren Etzioni. Captivating features of AllenAct It supports multiple environments such as iTHOR, RoboTHOR, Habitat and MiniGrid. Click here for instructions to install any of these supported environments. AllenAct has dissociated tasks from environments so that one can run multiple tasks in the same environment.It supports a variety of algorithms; to name a few – PPO, DD-PPO, A2C and DAgger.It supports algorithms and tasks involving multiple embodied agents.It enables easy visualization of first and third person views for agents and intermediate model tensors, integrated into Tensorboard.It is among a handful of Reinforcement Learning frameworks which support PyTorch.It allows the usage of a sequence of algorithms for efficiently training the models.It enables easy combining of various loss functions for model training. Installing AllenAct Requirements: Working version of Python 3.6 Clone the GitHub repository to your local machine git clone git@github.com:allenai\/allenact.git Move to the top directory cd allenact Install requirements using pipenv pipenv install --skip-lock --dev Install requirements using pip or pip3 (as per your machine’s configuration) pip install -r requirements.txt Visit this page for detailed installation instructions. Also read: How Open-Sourcing AllenAct Provides A Substantial Growth For Embodied AI Practical implementation Here’s a demonstration of training an embodied agent on the Point Navigation task within the RoboTHOR Embodied-AI environment. Before moving on to the implementation, let us have a look at some of the underlying concepts you may be unaware of. What is a Point Navigation task? The basic thing an embodied agent must know for accomplishing any task is – how to “move around” in the work environment. This task of moving around is formulated as follows: Make your agent locate a beacon somewhere in the environment. This beacon transmits its location in such a way that at any point of time, the agent can get the direction in which it needs to proceed and the euclidean distance which it needs to cover in order to reach the beacon. This particular task is termed as Point Navigation (also known as PointNav). PointNav, though appears simple, is a challenging task for the trained agents to perform. The reason being, such agents are not trained in a free open-space area. Rather, the training is carried out in an environment similar to real-world situations where the agents need to navigate through various hurdles such as walls, doors etc. If they are not familiar with the floor-plan of the environment they have to move in, they need to learn how to predict the design of such structures for proper navigation. What is an environment? An environment is something where the trained agent exists and performs actions such as moving forward, turning right and so on. Simulator used Here to build a Point Navigator, we have used RoboTHOR simulator. RoboTHOR has been designed to train models that can easily be transferred to a real robot, by providing a photo-realistic virtual environment and a real-world replica of the environment. It comprises 60 various virtual scenes with different floor plans – furniture and 15 validation scenes. AllenAct’s ‘Environment’ AllenAct has a class abstraction by the name Environment. It is a shallow wrapper which provides a uniform interface to the actual environment the agent works in. Algorithm used We have used a reinforcement algorithm called DD-PPO, a decentralized and distributed variant of the PPO algorithm. We let our agent explore the environment on its own, reward it for taking actions to approach the goal and penalize it for actions that deviate it from its goal. We then do model optimization to maximize this reward. Install RoboTHOR environment We need to set up the environment in which our embodied agent will perform point navigation. Click here and get to know about the environment installation steps. Understand the dataset used We have used the RoboTHOR PointNav dataset. It can be downloaded from here. It consists of several episodes with randomly generated start and end points for each scene. It also has cached distances between every pair of points in every scene. This helps the agent move in terms of geodesic distance i.e. the actual path distance and not the straight line distance. Code Setup experiment config file : A Python library needs to be imported for using it. On the contrary,  AllenAct is structured as a framework with a runner script called main.py. This script runs the experiment specified in a config file. This enables storing minute records of exactly which settings were used to arrive at a particular result. As RL models are generally too costly to train, this facility reduces the expenses to some extent. Create a new directory under projects\/. Import the required libraries and classes. import glob import os from math import ceil from typing import Dict, Any, List, Optional, Sequence import gym import numpy as np import torch import torch.nn as nn import torch.optim as optim from torch.optim.lr_scheduler import LambdaLR from torchvision import models from constants import ABS_PATH_OF_TOP_LEVEL_DIR from core.algorithms.onpolicy_sync.losses import PPO from core.algorithms.onpolicy_sync.losses.ppo import PPOConfig from core.base_abstractions.experiment_config import ExperimentConfig, MachineParams from core.base_abstractions.preprocessor import ( ResNetPreprocessor, SensorPreprocessorGraph, ) from core.base_abstractions.sensor import SensorSuite from core.base_abstractions.task import TaskSampler from plugins.ithor_plugin.ithor_sensors import RGBSensorThor from plugins.robothor_plugin.robothor_sensors import GPSCompassSensorRoboThor from plugins.robothor_plugin.robothor_task_samplers import PointNavDatasetTaskSampler from plugins.robothor_plugin.robothor_tasks import PointNavTask from projects.pointnav_baselines.models.point_nav_models import ( ResnetTensorPointNavActorCritic, ) from utils.experiment_utils import ( Builder, PipelineStage, TrainingPipeline, LinearDecay, evenly_distribute_count_into_bins, ) Define an experiment config class class PointNavRoboThorRGBPPOExperimentConfig(ExperimentConfig): \"\"\"A Point Navigation experiment in RoboTHOR environment.\"\"\" Define the task parameters \/*number of steps that the agent can take at the most before being reset (so that it does not keep on moving forever)*\/ MAX_STEPS = 500 \/\/configuration for the reward function REWARD_CONFIG = { \/*reward given to the agent for moving each step and encouraging it to move further*\/ \"step_penalty\": -0.01, \/\/ reward given for successfully reaching the goal \"goal_success_reward\": 10.0, \/*reward given in case the agent selects ‘stop’ action before reaching the goal*\/ \"failed_stop_reward\": 0.0, \/*strength of signal the agent gets on approaching or moving away from the goal*\/ \"shaping_weight\": 1.0, } Set the simulator Parameters CAMERA_WIDTH = 640 CAMERA_HEIGHT = 480 SCREEN_SIZE = 224 Set the hardware parameters for training engine ADVANCE_SCENE_ROLLOUT_PERIOD: Optional[int] = None NUM_PROCESSES = 20 TRAINING_GPUS: Sequence[int] = [0] VALIDATION_GPUS: Sequence[int] = [0] TESTING_GPUS: Sequence[int] = [0] Where, NUM_PROCESSES is the total number of parallel processes used to train the model. TRAINING_GPUS takes the IDs of the GPUS on which the model should be trained. Likewise, VALIDATION_GPUS and TESTING_GPUS take the ids of the GPUS on which the validation and testing will occur respectively. Define the paths to store the downloaded dataset TRAIN_DATASET_DIR = os.path.join( ABS_PATH_OF_TOP_LEVEL_DIR, \"datasets\/robothor-pointnav\/debug\" ) VAL_DATASET_DIR = os.path.join( ABS_PATH_OF_TOP_LEVEL_DIR, \"datasets\/robothor-pointnav\/debug\" ) Define the sensors SENSORS = [ \/* RGBSensorThor is the environment's implementation of an RGB sensor. It takes the simulator’s output (a raw image) as an input and resizes it to the specified input dimensions*\/ RGBSensorThor( height=SCREEN_SIZE, width=SCREEN_SIZE, use_resnet_normalization=True, uuid=\"rgb_lowres\", ), \/* GPSCompassSensorRoboThor is a sensor which tells us the agent about the direction and distance to its goal at every time step*\/ GPSCompassSensorRoboThor(), ] Define the preprocessor to be used with the model PREPROCESSORS = [ Builder( ResNetPreprocessor, { \"input_height\": SCREEN_SIZE, \"input_width\": SCREEN_SIZE, \"output_width\": 7, \"output_height\": 7, \"output_dims\": 512, \"pool\": False, \"torchvision_resnet_model\": models.resnet18, \"input_uuids\": [\"rgb_lowres\"], \"output_uuid\": \"rgb_resnet\", }, ), ] The preprocessor abstraction in AllenAct is designed with models which transform the raw pixels observed by the agent in the environment, into a complex embedding. Instead of the original image, the embedded version is then used as input to the trainable model. Define the observation inputs to be used by the model OBSERVATIONS = [ \/\/IDs of the sensors to be used \"rgb_resnet\", \"target_coordinates_ind\", ] Define settings of the simulator ENV_ARGS = dict( width=CAMERA_WIDTH, height=CAMERA_HEIGHT, gridSize=0.25, \/\/Everytime the agent takes ‘turn’ action, it rotates by 30 deg rotateStepDegrees=30.0, \/\/Every time the agent moves forward, it goes ahead by 0.25 meters visibilityDistance=1.0, ) Define a method that returns name of the experiment @classmethod def tag(cls): return \"PointNavRobothorRGBPPO\" Define the model training pipeline @classmethod def training_pipeline(cls, **kwargs): ppo_steps = int(250000000) \/\/number of training steps lr = 3e-4 num_mini_batch = 1 update_repeats = 3 num_steps = 30 \/\/rollout length \/\/how often we save the model weights and run validation on them save_interval = 5000000 log_interval = 1000 gamma = 0.99 use_gae = True gae_lambda = 0.95 max_grad_norm = 0.5 return TrainingPipeline( save_interval=save_interval, \/\/frequency at which data is accumulated from all the processes and logged metric_accumulate_interval=log_interval, optimizer_builder=Builder(optim.Adam, dict(lr=lr)), num_mini_batch=num_mini_batch, update_repeats=update_repeats, max_grad_norm=max_grad_norm, num_steps=num_steps, named_losses={\"ppo_loss\": PPO(**PPOConfig)}, gamma=gamma, use_gae=use_gae, gae_lambda=gae_lambda, advance_scene_rollout_period=cls.ADVANCE_SCENE_ROLLOUT_PERIOD, pipeline_stages=[ PipelineStage(loss_names=[\"ppo_loss\"], max_stage_steps=ppo_steps) ], lr_scheduler_builder=Builder( LambdaLR, {\"lr_lambda\": LinearDecay(steps=ppo_steps)} ), ) Define a method to return the hardware parameters of each process, based on the list of devices defined above def machine_params(self, mode=\"train\", **kwargs): sampler_devices: List[int] = [] if mode == \"train\": workers_per_device = 1 gpu_ids = ( [] if not torch.cuda.is_available() else list(self.TRAINING_GPUS) * workers_per_device ) nprocesses = ( 8 if not torch.cuda.is_available() else evenly_distribute_count_into_bins(self.NUM_PROCESSES, len(gpu_ids)) ) sampler_devices = list(self.TRAINING_GPUS) elif mode == \"valid\": nprocesses = 1 gpu_ids = [] if not torch.cuda.is_available() else self.VALIDATION_GPUS elif mode == \"test\": nprocesses = 1 gpu_ids = [] if not torch.cuda.is_available() else self.TESTING_GPUS else: raise NotImplementedError(\"mode must be 'train', 'valid', or 'test'.\") sensor_preprocessor_graph = ( SensorPreprocessorGraph( source_observation_spaces=SensorSuite(self.SENSORS).observation_spaces, preprocessors=self.PREPROCESSORS, ) if mode == \"train\" or ( (isinstance(nprocesses, int) and nprocesses > 0) or(isinstance(nprocesses, Sequence) and sum(nprocesses)> 0) ) else None ) return MachineParams( nprocesses=nprocesses, devices=gpu_ids, sampler_devices=sampler_devices if mode == \"train\" else gpu_ids,  # ignored with > 1 gpu_ids sensor_preprocessor_graph=sensor_preprocessor_graph, ) Define the actual model We take a model from the pointnav_baselines project. It is a small convolutional neural network. It takes the output of a ResNet as its rgb input followed by a single-layered GRU. The number of different actions the agent can perform in the environment is fed to the model through the action_space parameter, which we get from the task definition. Define the task sampler @classmethod def make_sampler_fn(cls, **kwargs) -> TaskSampler: return PointNavDatasetTaskSampler(**kwargs) This generates instances of tasks for our agent to perform. It reads the specified file and whenever the agent exceeds the maximum number of steps or selects the stop action, it sets the agent to the next starting locations. Define a few helper functions to distribute the work if you have several GPUS and many scenes to be processed @staticmethod def _partition_inds(n: int, num_parts: int): return np.round(np.linspace(0, n, num_parts + 1, endpoint=True)).astype( np.int32 ) def _get_sampler_args_for_scene_split( self, scenes_dir: str, process_ind: int, total_processes: int, seeds: Optional[List[int]] = None, deterministic_cudnn: bool = False, ) -> Dict[str, Any]: path = os.path.join(scenes_dir, \"*.json.gz\") scenes = [scene.split(\"\/\")[-1].split(\".\")[0] for scene in .glob(path)] if len(scenes) == 0: raise RuntimeError( ( \"Could find no scene dataset information in directory {}.\" \" Are you sure you've downloaded them? \" \" If not, see https:\/\/allenact.org\/installation\/download-datasets\/ information\" \" on how this can be done.\" ).format(scenes_dir) ) if total_processes > len(scenes):  # oversample some scenes -> bias if total_processes % len(scenes) != 0: print( \"Warning: oversampling some of the scenes to feed all processes.\" \" You can avoid this by setting a number of workers divisible by the number of scenes\" ) scenes = scenes * int(ceil(total_processes \/ len(scenes))) scenes = scenes[: total_processes * (len(scenes) \/\/ total_processes)] else: if len(scenes) % total_processes != 0: print( \"Warning: oversampling some of the scenes to feed all processes.\" \" You can avoid this by setting a number of workers divisor of the number of scenes\" ) inds = self._partition_inds(len(scenes), total_processes) return { \"scenes\": scenes[inds[process_ind] : inds[process_ind + 1]], \"max_steps\": self.MAX_STEPS, \"sensors\": self.SENSORS, \"action_space\": gym.spaces.Discrete(len(PointNavTask.class_action_names())), \"seed\": seeds[process_ind] if seeds is not None else None, \"deterministic_cudnn\": deterministic_cudnn, \"rewards_config\": self.REWARD_CONFIG, } Define the task sampler arguments The arguments include the location of the dataset as well as a reference to the distance cache and the environmental arguments for the simulator mentioned above. def train_task_sampler_args( self, process_ind: int, total_processes: int, devices: Optional[List[int]] = None, seeds: Optional[List[int]] = None, deterministic_cudnn: bool = False, ) -> Dict[str, Any]: res = self._get_sampler_args_for_scene_split( os.path.join(self.TRAIN_DATASET_DIR, \"episodes\"), process_ind, total_processes, seeds=seeds, deterministic_cudnn=deterministic_cudnn, ) res[\"scene_directory\"] = self.TRAIN_DATASET_DIR res[\"loop_dataset\"] = True res[\"env_args\"] = {} res[\"env_args\"].update(self.ENV_ARGS) res[\"env_args\"][\"x_display\"] = ( (\"0.%d\" % devices[process_ind % len(devices)]) if devices is not None and len(devices) > 0 else None ) res[\"allow_flipping\"] = True return res def valid_task_sampler_args( self, process_ind: int, total_processes: int, devices: Optional[List[int]] = None, seeds: Optional[List[int]] = None, deterministic_cudnn: bool = False, ) -> Dict[str, Any]: res = self._get_sampler_args_for_scene_split( os.path.join(self.VAL_DATASET_DIR, \"episodes\"), process_ind, total_processes, seeds=seeds, deterministic_cudnn=deterministic_cudnn, ) res[\"scene_directory\"] = self.VAL_DATASET_DIR res[\"loop_dataset\"] = False res[\"env_args\"] = {} res[\"env_args\"].update(self.ENV_ARGS) res[\"env_args\"][\"x_display\"] = ( (\"0.%d\" % devices[process_ind % len(devices)]) if devices is not None and len(devices) > 0 else None ) return res def test_task_sampler_args( self, process_ind: int, total_processes: int, devices: Optional[List[int]] = None, seeds: Optional[List[int]] = None, deterministic_cudnn: bool = False, ) -> Dict[str, Any]: res = self._get_sampler_args_for_scene_split( os.path.join(self.VAL_DATASET_DIR, \"episodes\"), process_ind, total_processes, seeds=seeds, deterministic_cudnn=deterministic_cudnn, ) res[\"scene_directory\"] = self.VAL_DATASET_DIR res[\"loop_dataset\"] = False res[\"env_args\"] = {} res[\"env_args\"].update(self.ENV_ARGS) return res Model training python main.py -o <PATH_TO_OUTPUT> -c -b <BASE_DIRECTORY_OF_YOUR_EXPERIMENT> <EXPERIMENT_NAME> NOTE: Training the entire dataset takes nearly 2 days on a machine with 8 GPU! However, you can train on a smaller dataset simply by changing the dataset’s path in the above line of code. Model testing Visit this page for the steps to test a pre-trained AllenAct model. Source: https:\/\/allenact.org\/tutorials\/training-a-pointnav-model End note In this article, we gave an overview of a state-of-the-art framework related to the Embodied AI domain viz. AllenAct. We also gave a demonstration of implementing PointNav task using the modular project. To get a deeper understanding of the AllenAct project, refer to the following sources: Official documentationRelated articleGitHub repository","excerpt":"Introduction AllenAct is an open-source project and a modular learning framework designed for researchers and technophiles associated with the domain of Embodied AI. It provides state-of-the-art reproductions of numerous embodied AI models. It also supports the expanding collection of embodied AI tasks, algorithms used to accomplish those tasks as well as environments to run them […]","categories":["Deep Tech"],"tags":["Embodied AI"],"author_name":"Nikita Shiledarbaxi","publish_date":"2021-02-09T12:00:00","publication_year":"2021","word_count":2405,"keywords":["CUDA","NumPy","artificial intelligence","TPU","AI","neural network","PyTorch","ML","Embodied AI","RAG","Aim"],"extracted_tech_keywords":["AI","artificial intelligence","ML","neural network","Aim","PyTorch","NumPy","RAG","TPU","CUDA"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/hands-on-python-guide-to-allenact-an-embodied-ai-project\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10093073,"title":"Hugging Face Releases A State-of-the-Art LLM For Code","content":"Hugging Face and ServiceNow, two major players in AI, have partnered to develop a new open-source language model for codes called StarCoder. The model created as a part of the BigCode initiative is an improved version of the StarCoderBase model trained on 35 billion Python tokens. Researchers stated, StarCoder’s capabilities have been tested on a range of benchmarks, including HumanEval benchmark for Python. The model has outperformed larger models like PaLM, LaMDA, and LLaMA, and has proven to be on par with or even better than closed models like OpenAI’s code-Cushman-001 (the original Codex model that powered early versions of GitHub Copilot). Trained on over 1 trillion tokens and with a context window of 8192 tokens, the model boasts an impressive 15.5 billion parameters. It was created using data from GitHub, including 80+ programming languages, Git commits, GitHub issues, and Jupyter notebooks. This vast dataset was preprocessed to include only content with permissive licenses, ensuring that the resulting model can generate source code while adhering to legal criteria. StarCoder’s primary function is as a technical assistant, generating realistic code and supporting 80 programming languages. However, it is not designed for issuing instructions or directives like “write a function that computes the square root.” Instead, users can follow on-screen prompts to transform StarCoder into a helpful programming tool. Despite its impressive capabilities, StarCoder has its set of limitations. Like other LLMs, it can produce incorrect and offensive information. To address these concerns, the researchers have released the StarCoder models under an Open Responsible AI Model (RAIL) license and have open-sourced all code repositories for creating the model on GitHub. To ensure that the model adheres to responsible AI principles, the model license includes usage restrictions, and a set of attribution tools is available to end-users to identify potentially plagiarized model generations. Notably, StarCoder is not the first mover in the domain. LLM-based coding platforms are continuing to improve themselves, with Google researchers demonstrating earlier this month that they can be used to self-debug. Apart from Microsoft’s GitHub Copilot and Amazon CodeWhisperer, the India-based Replit has also joined the LLM race. Read: Replit Knows India Better than Amazon, Microsoft","excerpt":"The Hugging Face model is an improved version of the StarCoderBase model trained on 35 billion Python tokens.","categories":["AI News"],"tags":["Hugging Face","Open Source","Open Source LLM"],"author_name":"Tasmia Ansari","publish_date":"2023-05-09T16:39:29","publication_year":"2023","word_count":357,"keywords":["Go","Hugging Face","Open Source","OpenAI","AI","Git","Python","Open Source LLM","responsible AI","GitHub","Jupyter","R"],"extracted_tech_keywords":["AI","OpenAI","Hugging Face","Jupyter","Python","R","Go","Git","GitHub","responsible AI"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/hugging-face-releases-a-state-of-the-art-llm-for-code\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10162066,"title":"Oracle’s Larry Ellison is the ‘CEO of Everything’","content":"Larry Ellison founded Oracle in 1977, long before the internet reshaped the world. Since then, it has transformed into a powerhouse in database management and cloud computing and has laid the foundation for global digital infrastructure. As AI ushers in a new era, Oracle stands ready to lead the change. This week, while announcing The Stargate Project, US President Donald Trump called Ellison the “CEO of Everything”. “Larry Ellison…is an amazing man, an amazing businessperson.” Trump’s reference makes perfect sense, considering Oracle’s crucial partnership in the $500 billion investment in AI infrastructure in the US. Besides, Ellison came to the rescue of the much talked about TikTok Global, which was on the verge of shutdown in America. Oracle owns a 12.5% stake in the video platform.  The US President’s closeness to Ellison pushed him to sign an executive order to delay the ban. Oracle will be helping the US government build new data centres that will power the country’s AI infrastructure. “We are proud to be working with OpenAI, SoftBank, and MGX on The Stargate Project to secure American leadership in AI, create hundreds of thousands of American jobs, and generate massive economic benefit for the entire world,” the company said in an X post. Ever since Trump made the announcement, Oracle shares jumped nearly 13%. This has also positively impacted Ellison’s net worth and reflected investor confidence in Oracle’s future in AI technology. The construction of data centres has already begun in Texas, and additional sites are being evaluated across the country. Oracle, NVIDIA, and OpenAI will collaborate to build and operate the computing system. This partnership extends an established collaboration between OpenAI and NVIDIA since 2016 and a newer relationship with Oracle. “The first of them (data centres) are under construction in Texas. Each building is half a million square feet. There are 10 buildings currently being built, but that will expand to 20…other locations,” Ellison said. Moreover, Oracle will now also work closely with OpenAI, especially as Microsoft is no longer an exclusive cloud partner for the startup. OpenAI will now run its workloads on Oracle Cloud Infrastructure (OCI), extending the Microsoft Azure AI platform to Oracle’s cloud services. “Like many others, OpenAI chose OCI because it is the world’s fastest and most cost-effective AI infrastructure,” Oracle chief Safra A Catz said in a recent earnings call. In 2024, Oracle launched its first zettascale cloud computing clusters accelerated by NVIDIA’s Blackwell platform. OCI is now taking orders for the largest AI supercomputer in the cloud, supporting up to 1,31,072 GPUs. Surprisingly, Ellison’s close associate Elon Musk’s xAI is not part of The Stargate Project. Notably, xAI ended a $10 billion server deal with Oracle last year, which suggests a possible strain in their relationship. Curing Diseases with AI At the White House, Ellison discussed AI’s potential in healthcare and highlighted its ability to enable early cancer detection and the creation of personalised vaccines. “Using AI, you can do early cancer detection with a blood test, and once we gene sequence that cancer tumour, you can then vaccinate the person [and] design a vaccine for every individual person to vaccinate them against that cancer,” he said. He added that mRNA vaccines can be made robotically using AI in about 48 hours. “This is the promise of AI and the promise of the future.” Companies like Kuano and Xaira Therapeutics are already using OCI to accelerate drug discovery by leveraging AI and high-performance computing (HPC). OCI provides scalable cloud solutions for large-scale simulations and AI model training Will Oracle Buy TikTok? There have been discussions about Ellison and Musk potentially buying half the stake in TikTok, and Trump has shown openness to the idea. “I’m open to Elon Musk buying TikTok if he wants it. I’m also open to Larry Ellison buying it,” Trump said. Turning to Ellison, he added, “You’re a rich guy, Larry. Do you want it? I want them to do a 50\/50 joint venture deal with America. With us, it’s worth trillions; without us, it’s nothing. What do you think, Larry?” “Sounds like a good deal to me, Mr President,” Ellison replied. AIM reached out to Oracle to confirm the development, but the company declined to comment. As per recent reports, during discussions about a potential TikTok shutdown, Oracle warned that it could harm its business. Notably, Oracle is Tiktok’s cloud service provider. In 2022, TikTok launched ‘Project Texas’, a $1.5 billion initiative to address concerns over US user data security by transferring control to Oracle. The initiative included moving all US user data to Oracle Cloud and establishing a dedicated transparency centre where Oracle employees could review the platform’s source code.","excerpt":"Ellison’s Oracle will be helping the US government build new data centres that will power the country’s AI infrastructure.","categories":["Global Tech"],"tags":["Oracle"],"author_name":"Siddharth Jindal","publish_date":"2025-01-23T17:00:29","publication_year":"2025","word_count":777,"keywords":["Go","OpenAI","AI","cloud computing","Azure","R","Scala","Oracle","RAG","Aim","xAI"],"extracted_tech_keywords":["AI","OpenAI","xAI","Aim","RAG","cloud computing","Azure","R","Go","Scala"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/oracles-larry-ellison-is-the-ceo-of-everything\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10068129,"title":"100,000 human brain images","content":"Researchers at King’s College London used NVIDIA Cambridge -1, UK’s most powerful supercomputer dedicated to AI research in healthcare and MONAI, an open source AI software framework, to create 100,000 high resolution 3D images of human brains. Jorge Cardoso, researcher in Artificial Medical Intelligence at King’s College London and CTO of the new London Medical Imaging and AI Centre for Value-based Healthcare, is heading the project. Cambridge-1 and MONAI were combined to create an AI factory for synthetic data that let researchers run hundreds of experiments, choose the best AI models and run inference to generate images. “We couldn’t have done this work without Cambridge-1 and MONAI, it just wouldn’t have happened,” Cardoso said. Cardoso is working with a national repository in the UK, Health Data Research to host the images so that healthcare researchers can access them for free. This will do away with one of the major challenges faced by the healthcare researchers fraternity: Medical images are not widely available due to privacy concerns. The team is also working to extend this approach in creating 3D images of any part of the human body. In addition to the images, plans are on to make the AI models available to researchers so that they can create images as per their needs.","excerpt":"Jorge Cardoso, researcher in Artificial Medical Intelligence at King’s College London, is heading the project.","categories":["AI News"],"tags":[],"author_name":"Zinnia Banerjee","publish_date":"2022-05-31T16:49:09","publication_year":"2022","word_count":212,"keywords":["synthetic data","programming_languages:R","AI","emerging_tech:synthetic data","AI research","R"],"extracted_tech_keywords":["AI","R","synthetic data","AI research","programming_languages:R","emerging_tech:synthetic data"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/100000-human-brain-images\/","complexity_score":2,"technical_depth":6,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10164874,"title":"Tencent Releases Hunyuan Turbo S to Rival DeepSeek as Competition Heats up in China","content":"Chinese tech giant Tencent has released its new AI model, Hunyuan Turbo S, which it says can answer queries faster than the DeepSeek-R1 model. The model is available on the official Tencent Cloud website and can be accessed via API. The Hunyuan Turbo S doubles the output speed and reduces the first-word delay by 44%,  the company announced on its official WeChat channel. Tencent said that the fast-thinking model is analogous to human intuition, which often results in rapid responses compared to rational thinking. However, the company said Hunyuan Turbo S efficiently solves problems by fusing long and short thinking chains. The model uses an innovative hybrid-mamba-transformer fusion architecture. It optimises efficiency by lowering the computational complexity of the conventional transformer, minimising KV-Cache storage usage, and reducing training and inference costs. The company also said that the model leverages Mamba’s efficiency in processing long sequences while preserving the Transformer’s strength in capturing complex contextual relationships. Tencent also claims this is the first time the Mamba architecture has been applied losslessly to a super-large Mixture of Experts (MoE) model. Tencent also released benchmark results, and the model is better, if not on par with other large language models like DeepSeek-V3, Claude 3.5 Sonnet, and GPT-4o—in mathematics, coding, and reasoning tasks. The Hunyuan Turbo S’s input API price is 0.8 yuan ($0.11) per million tokens, and its output price is 2 yuan ($0.28) per million tokens. Source: Tencent Amid the rise of DeepSeek, the competition in China’s AI ecosystem is heating up. Recently, Alibaba introduced a preview of the Qwen QwQ-Max reasoning model and committed to a $52 billion investment in AI infrastructure over the next three years. It was also reported that DeepSeek plans to release its next reasoning model, the DeepSeek R2, ‘as early as possible’. The company initially planned to release it in early May but is now considering an earlier timeline. The model is expected to produce ‘better coding’ and reason in languages beyond English. Note: The headline has been updated to provide better clarity.","excerpt":"The model outperforms DeepSeek-V3, Claude-3.5 Sonnet and GPT-4o in several tasks.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","China","Tencent"],"author_name":"Supreeth Koundinya","publish_date":"2025-02-28T16:03:10","publication_year":"2025","word_count":337,"keywords":["Go","API","TPU","AI","GPT-4o","Tencent","RAG","GPT","Aim","AI (Artificial Intelligence)","Claude 3.5","R","China"],"extracted_tech_keywords":["AI","GPT-4o","Claude 3.5","Aim","RAG","TPU","R","Go","API","GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/tencent-releases-hunyuan-turbo-s-to-rival-deepseek-as-competition-heats-up-in-china\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":31742,"title":"AI Became The New Abstract In The Art World In 2018","content":"Forget acrylics and oil on canvas, there is a new medium that artists across the globe have embraced enthusiastically. Artificial Intelligence has given rise to a new artform — algorithms which are used to drum up paintings, music and even videos. Using AI algorithms, one can either imitate an old masterpiece or create a new one from scratch. This technology doesn’t require much skill, there is software in the market which helps in producing art via AI. The traditional art world has been upended with algorithmically produced artwork which is doing the rounds of heavyweight auction houses. In India, the art industry is estimated to be $250 million. We list down the top AI generated art deals of 2018: Christie’s Auction Edmond Belamy Portrait- Christie’s Images One of the oldest art auction house, founded in 1766 grabbed headlines when it made a groundbreaking sale of an AI generated artwork. In its auction at New York, Christie’s presented the first AI portrait to the world. The portrait was of Edward Belamy created by an AI algorithm called Generative Adversarial Networks (GAN). According to reports, it was expected to be bought at $10,000 but it sold for $432,500. In the case of Edmond’s, it was fed with 15,000 portraits and then it generated a painting based on the set. Nature Morte, Delhi Electric Fan By Tom White Gradient Descent, an art auction in New Delhi was a huge success for the gallery in September. It was a group show which was done in collaboration with seven international artists who made portraits with the help of AI algorithms. Aparajita Jain, gallery director said, “We couldn’t afford to ignore the field of AI-made art because of how we believe it is going to impact the art world. And while I was initially shocked to find out how far AI has already come in the creative field. The show was to alert artists about AI and its practice in the field.” the show received a phenomenal reception in India. Two prints by Tom White have already been sold. AI Lab at Rutgers The art and AI lab at Rutgers is a research-based institute which works to converge AI into art domain. The AI algorithm the lab uses is dubbed Creative Adversarial Networks (CAN) which generates art by deviating from style norms. The lab has produced numerous painting and has garnered much appreciation from renowned art galleries like Artsy. The lab has also collaborated with HBO’s series Silicon Valley, which depicted art form created by the lab. A TV segment on the lab won an Emmy for “ Best Arts News”. Procreate 4.2 The AI-generated art is no longer restricted to art galleries. Procreate is an application developed for iOS especially iPad which helps in creating art from doodles and random shapes. The app has been used to create many portraits. The app’s USP is its liquified dynamics, giving each design an artistic touch. Also, one can change the distortion and momentum of the objects in their art and design. The app can be operated by the new Apple pencil which provides precise control over minuscule details.","excerpt":"Forget acrylics and oil on canvas, there is a new medium that artists across the globe have embraced enthusiastically. Artificial Intelligence has given rise to a new artform — algorithms which are used to drum up paintings, music and even videos. Using AI algorithms, one can either imitate an old masterpiece or create a new […]","categories":["AI Features"],"tags":[],"author_name":"Jignasa Sinha","publish_date":"2018-12-18T12:40:07","publication_year":"2018","word_count":521,"keywords":["Go","artificial intelligence","programming_languages:R","AI","programming_languages:Go","GAN","R"],"extracted_tech_keywords":["AI","artificial intelligence","R","Go","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/ai-became-the-new-abstract-in-the-art-world-in-2018\/","complexity_score":4,"technical_depth":7,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10048296,"title":"In Conversation With AI Scientist Dr Ganapathi Pulipaka On Gallium Nitride Processors For Future Space Exploration","content":"Gallium nitride (GaN) processors are poised to power the next-generation semiconductor industry to revolutionize space exploration. Gallium nitride compound has been used in building light-emitting diodes (LEDs) and is highly efficient in conducting the electrons 1000 way better than silicon material. These are currently researched to explore in building high-temperature microprocessors for space applications. The silicon powered integrated circuits start malfunctioning once they reach 300 degrees celsius. However, GaN material is highly stable chemically in high-radiation and temperature environments. There is still a lot of research going on in developing metal-oxide-semiconductor transistors. The GaN transistors and GaN chip architecture are way more complex than silicon chips. As a matter of fact, NASA is researching to produce a GaN processor that can operate efficiently above 500 degrees celsius. GaN processors can withstand various space missions in high-temperature planet worlds such as Venus, Mercury, Jupiter, Saturn, Uranus, or Neptune. Gallium nitride processors for landing on 500+ degrees high-temperature planets. Analytics India Magazine has interviewed Dr Ganapathi Pulipaka to understand several aspects of machine learning, deep learning, and mathematics needed for practitioners. An adept data engineer and data architect turned data scientist, Dr Pulipaka prolifically posts about data science trends, keeping the audience up-to-date on the latest information. He covered several topics in-depth on OpenAI, deep neural networks and fractal geometry, to name a few. Dr Pulipaka has written two books so far, and there is an upcoming book about to be published in October. “Until we have infinite processing power on machines, we cannot produce the outcomes we need” – GP Edited excerpt — AIM: In several interviews and magazines, you mentioned that you assembled your first computer on your own at 16 years old and cited that as your entry point for your passion for technology. Did you always plan to do that to be able to program and build hardware and software applications? Also, why did you choose to build your computer? GP: Well, it was not something I planned to build earlier. It was something my brain told me to do, and I explored the prospect of building a machine. The outcome and results were amazing. A lot of people had laptops at the age of 16, but I didn’t. Laptops back in the day were a little bit expensive to own. The easier way to go was to build your own computer. If you can assemble all the components, you can build a computer, which was why I actually started building it. In today’s world, it’s even harder to upgrade RAM on Macbook Air, Macbook Pro, or iMac as RAM is soldered to the motherboard. Though some electronic engineers successfully did such upgrades, it would be too risky and void the warranty, as forcefully removing the original soldered hardware from the motherboard may cause compatibility issues with the operating system. Getting another battery or SSD would be a much easier option to deal with. AIM: Were there any freelance projects that you implemented before entering into corporate enterprise-wide rollouts? GP: Right after I built my computer, I immediately pursued a few projects I wanted to implement as a freelancer while still in school. Back in the day, the Air Force Academy wanted to implement a traffic system project, handling the takeoff, landing, and predicting the weather patterns by creating an application or detouring a plane. So I approached the Wing Commander, explained my passion, and was kind enough to offer it to me. To facilitate that, I built an SQL database, primitive in nature, and had written a C++ front end. As my first project, I even received an appreciation letter from the Wing Commander for it. I was also very interested in mathematics and statistics, which is important for solving data science equations such as Ordinary Differential Equations and Partial Differential Equations, linear algebra, and forecast problems. For example, many companies want to build a trajectory of performing for different claims or different manufacturing locations. There’s a lot of historical data that they can gather and build those time series forecasting models. I have implemented a few more projects for those companies, which I will be covering in another interview. In 1996, I worked for another company freelancing, where I built a complete time series forecasting system and went live with it. Creating this system got me more interested in mathematical models and statistical models. In addition, I was interested in understanding how to go about the whole gamut of landscapes and infrastructure and architectures and how all of these systems integrate. AIM: Can you share some light on other projects that you have worked on so far? I’ve been researching for more than ten years. That’s when I got more interested in deep learning and machine learning. Before that, I was focused on data engineering, programming, data architecture on databases of SQL, NoSQL, data science, mathematics etc. Reverse-mode automatic differentiation of ODE Computation graph of the latent ODE model: I also worked for many clients, like for an aerospace IoT integration, where it is important to know or forecast when something will break down, like a bridge inspection, to prevent a collapse. In a plane, you don’t know when a panel is going to break down. Here, there is an acoustic emission, and if you measure the acoustic emission with fractal geometry, that can tell us exactly when there is a change between the previous state of this panel versus the current state. There are different factors that you can calculate and provide different parameters. That was another interesting project to work on. AIM: Let’s talk a little bit about your research projects. On the research side, I’ve done implementations of pretty much all the algorithms. And there’s a book that’s coming out with my research — A Greater Foundation for Machine Engineering: The hallmarks of the Great Beyond in TensorFlow, R, PyTorch, and Python, where I’m covering whatever that I picked up and what I thought would be useful for other people in this field. It includes Python, TensorFlow, all the machine learning algorithms in Python, and a lot of statistical forecasting and other stuff built using Python. GP’s Upcoming book, reviewed by Gloria Pullman of Reader’s Digest Reinforcement Learning is also a very interesting area because many people are into Artificial General Intelligence (AGI). Reinforcement learning primarily shifts the paradigm of having supervised, unsupervised and semi-supervised machine learning\/deep learning landscapes. It takes us into a completely different type of approach from Bellman back in the day. I typically include a lot of natural language programming, leveraging NLTK, Gensim, and a number of other Python libraries. I think that NLP will have significant advancements in the future, that there’s going to be GPT-4 perhaps. The whole GPT-4 singularity depends on the recent AGI research based on 999 AI researchers, where they all express that it is possible to achieve singularity by 2060. AIM: Can you highlight some key aspects mentioned in your upcoming book? The natural language toolkit (NLTK) is primarily a collection of several libraries for text processing and natural language processing within Python. The University of Pennsylvania developed this library that allows the user to perform several NLP operations such as splitting texts and sentences from multiple paragraphs, identifying whether the statement is a question or an answer, detecting the part of speech of the words, and splitting up multiple words. You can find plenty of such NLP implementations in my upcoming book on Python. I start with fundamentally very simple algorithms like linear and logistic regression so that the readers can understand the concepts. Then they can proceed with advanced concepts like reinforcement learning. I got this idea when I tried to do some video training classes on reinforcement learning with the book publishing company Packt and QuickStart boot camp programs. Although I was delivered educational programs on machine learning, cloud computing, mathematics, and statistics, some people were having issues because they went right into reinforcement learning without knowing what machine learning was. So, that’s when I started introducing the fundamentals before moving on to advanced concepts. In the book, I explore some parts of history, like how it began with all the elements, how it advanced and where we are now. The book coming out is 80% nothing but my dissertation on deep learning in high-performance computing environments, along with some compiled other work I’ve done for many years. AIM: Please share a bit about your next dissertation on high-performance computing and why did you choose this topic? My next dissertation will be on high-performance computing (HPC) because it will be completely different from my previous dissertations on data engineering, big data analytics and in-memory computing. For example, we have multiprocessing for MPI (Message Passing Interface –an open-source library standard for the distributed memory parallelization) available in C. We also need to develop the parallel processing capabilities for HPC. Python and Java have their own unofficial bindings for MPI, but most of the scientific engineering labs leverage it in C. MPI technique faces latency when handling millions of tasks through collective calls plagued with operating system crashes and errors due to scalability memory overloads. The MPIs are implemented primarily as libraries through distributed memory computing for high-performance computing. That’s where you need to chunk the blocks — just like how we’re using Apache Hadoop with map, shuffle, and reduce technique — use cloud computing clusters to break down the data and process heaps of math equations, assemble the data back before you get it back into your project. MapReduce is a technique adopted by Apache Hadoop for processing a variety of high-performance computing applications, where the big data in petabyte size reside on the nodes of the computing machines. Each dedicated storage unit of the computing node cluster performs I\/O-intensive, CPU-intensive, GPU-intensive Hadoop, YARN. Research on HPC in nuclear plasma turbulence is already talking about nuclear weapons transportation. As a matter of fact, Sandia National Laboratories is developing landmark HPC algorithms for it. These are all parallel processing algorithms and will be intensive, but that’s where I would see myself in future. “The brain is orchestrating all at the same time, so it’s a different type of mechanism altogether. If we can understand it, we’ll come up with more innovative models.” _____________________________________________ References Chen, R. T., Rubanova, Y., Bettencourt, J., & Duvenaud, D. (2019). Neural Ordinary Differential Equations. Retrieved from https:\/\/arxiv.org\/pdf\/1806.07366.pdf Collins M.D. PhD, F. S., & Fink, L. (1995). The Human Genome Project. NCBI – US National Library of Medicine National Institutes of Health National Institutes of Health, 19(3)(), 190 – 195. Retrieved from https:\/\/www.ncbi.nlm.nih.gov\/pmc\/articles\/PMC6875757\/ Dilmegani, C. (2017). Will AI reach singularity by 2060? 995 experts? opinions on AGI. Retrieved from https:\/\/research.aimultiple.com\/artificial-general-intelligence-singularity-timing\/ DK, P. (2015). High-Performance Deep Learning (HiDL) – The Ohio State University. Retrieved from http:\/\/hidl.cse.ohio-state.edu Du, Q., & Li, X. (2013, October 13). Li, Z., Shen, H., Denton, J., & Ligon, W. (2016, December 8th, 2016). Comparing application performance on HPC-based Hadoop platforms with local storage and dedicated storage. IEEE International Conference on Big Data (Big Data). http:\/\/dx.doi.org\/10.1109\/BigData.2016.7840609. IEEE Proceedings of 2013 3rd International Conference on Computer Science and Network Technology. http:\/\/dx.doi.org\/10.1109\/ICCSNT.2013.6967143 Hao, K. (2019). The computing power needed to train AI is now rising seven times faster than ever before. Retrieved from https:\/\/www.technologyreview.com\/2019\/11\/11\/132004\/the-computing-power-needed-to-train-ai-is-now-rising-seven-times-faster-than-ever-before\/ Li, Z., Shen, H., Denton, J., & Ligon, W. (2016, December 8th, 2016). Comparing application performance on HPC-based Hadoop platforms with local storage and dedicated storage. IEEE International Conference on Big Data (Big Data). http:\/\/dx.doi.org\/10.1109\/BigData.2016.7840609 Nowakowski, T. (2017). Gallium nitride processor—next-generation technology for space exploration. Phys Org. Retrieved from https:\/\/phys.org\/news\/2017-12-gallium-nitride-processornext-generation-technology-space.html Pullman, G. (2021). The AI Bible: A Gigantic Masonry on a Greater Foundation of Machine Learning Engineering. Reader’s Digest. Retrieved from https:\/\/www.readersdigest.co.uk\/inspire\/down-to-business\/the-ai-bible-a-gigantic-masonry-on-a-greater-foundation-of-machine-learning-engineering","excerpt":"AIM interviews Dr Ganapathi Pulipaka on gallium nitride processors for future space exploration, deep quest of AI, deep learning algorithms, reinforcement learning and high-performance computing.","categories":["AI Features"],"tags":["Deep Learning","high performance computing","Interviews and Discussions","Reinforcement Learning"],"author_name":"Ganapathi Pulipaka","publish_date":"2021-09-14T16:00:00","publication_year":"2021","word_count":1948,"keywords":["data science","machine learning","Reinforcement Learning","OpenAI","AI","neural network","ML","NLP","Aim","deep learning","high performance computing","analytics","Deep Learning","Interviews and Discussions"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","neural network","NLP","data science","analytics","OpenAI","Aim"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/in-conversation-with-ai-scientist-dr-ganapathi-pulipaka-on-gallium-nitride-processors-for-future-space-exploration\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10170157,"title":"BharatGen Launches Param-1 India’s Foundational LLM Built from Scratch","content":"In its mission to build open source LLMs for Indian researchers and developers, BharatGen, the government backed AI initiative, has released a 2.9 billion parameter bilingual LLM, called Param 1. The newly launched LLM, dubbed ‘BharatGen Param 1 Indic Scale’, is a pre-trained base model built entirely from scratch and features a staggering 25% Indic data—a stark contrast to the mere 0.01% Indic data typically used in models like Meta’s Llama. You can check out the model on AIKosha. “Pre-training is an enormous undertaking and often an insurmountable barrier for many. That’s why we’ve taken on this challenge—to provide a robust foundation that you can easily fine-tune for your specific applications,” BharatGen said in a statement. Developers can now fine-tune the model via AIKosha to build diverse applications ranging from Indic chatbots to India-specific copilots and knowledge systems. “With our 2.9 billion parameter base model, we are unlocking new possibilities for innovation and growth across the nation. We hope this sovereign LLM model checkpoint serves as a foundation for India-specific solutions, enabling developers to fine-tune and shape the next generation of AI applications for Bharat,” Prof Ganesh Ramakrishnan, head of BharatGen, told AIM. Alongside the LLM, the team also launched 20 new speech models (across 19 Indian language variations)—targeting voice-first interfaces and speech-based innovation for Indian users on AIKosha—the AI innovation repository by MeitY, Government of India. This includes 9 models under A2TTS-v0.5: Speaker Adaptive TTS. These allow developers to generate speech that matches a provided speaker’s voice, available in Marathi, Bengali, Hindi, Gujarati, Tamil, Kannada, Punjabi, Telugu, and Malayalam. There are five models under Speaker-Conditioned TTS (pflow) which offer high-fidelity text-to-speech models for Marathi, Tamil, Hindi, Telugu, and Bengali. Then there are other five models under Voicebox TTS Models, which are adaptable voice synthesis for Hindi, Tamil, Marathi, Telugu, and Bengali. BharatGen says that these models were built from the ground up with data collected directly for five Indian languages, addressing a major gap in high-quality, publicly available speech models for Indic languages. AIKosha, launched by Union Minister Ashwini Vaishnaw, is India’s official AI repository and the new home for these models. The repository aims to centralise India’s AI assets and fuel collaborative innovation. “These foundational models are engineered to supercharge India’s AI research and innovation ecosystem,” BharatGen noted, inviting the community to “build an AI that genuinely speaks to, and for, India.” Along with Ramakrishnan, the contributors of the model include Kundeshwar Pundalik, Durga S, Prateek Chanda, Vedant Goswami, Atul Kumar Singh, Saral Sureka, Panditi Bhagawan, Ajay Nagpal, Smita Gautam, Pankaj Singh, Rishi Bal, and Prof Rohit Saluja. While earlier speaking with AIM, Ramakrishnan said that unlike private entities, BharatGen operates with a clear mission—‘GenAI for Bharat, by Bharat’. With an investment of under ₹235 crores, which is close to $27 million, leveraging cost-efficient computing and attracting top talent from graduates from IITs. Read: This Govt-Funded ₹235 Cr AI Initiative is India’s Real Answer to DeepSeek The BharatGen consortium comprises IIT Bombay, IIT Kanpur, IIT Mandi, IIT Madras, IIT Hyderabad, IIIT Hyderabad, and IIM Indore. Vaishnaw had earlier stated that India would have its own foundational AI models within 7-8 months, and BharatGen is a key part of that vision. “Yes, we are very much on track. The Minister has been briefed, and we are aligned with the timeline,” Ramakrishnan had confirmed earlier. Param-1 is a clear sign of that roadmap. “Our goal is not just to build AI models but to provide resources that startups and system integrators can leverage,” said Ramakrishnan. Last month, MeitY also selected Sarvam AI under the IndiaAI Mission to develop India’s sovereign LLM as part of the effort to create indigenous AI capabilities. The team had proposed the development of a 70-billion parameter multimodal AI model that supports both Indian languages and English, and work on it has already begun.","excerpt":"Alongside the LLM, the team, led by Ganesh Ramakrishnan, also launched 20 new speech models (across 19 Indian language variations)—targeting voice-first interfaces and speech-based innovation for Indian users on AIKosha.","categories":["AI News"],"tags":[],"author_name":"Mohit Pandey","publish_date":"2025-05-19T12:58:57","publication_year":"2025","word_count":638,"keywords":["Go","GenAI","AI","chatbots","RAG","Aim","multimodal AI","GAN","copilots","R"],"extracted_tech_keywords":["AI","GenAI","multimodal AI","Aim","RAG","copilots","chatbots","R","Go","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/bharatgen-launches-param-1-indias-foundational-llm-built-from-scratch\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":53324,"title":"Top 10 AI\/ML Stocks to Buy In 2020","content":"As businesses continue investing in artificial intelligence and with the increasing applications it is finding in every sector, the trajectory of these lucrative stocks might never come down. If we look back at 2019, which seems like a long time ago already, companies across sectors have started integrating AI and have implemented it in their products and services. Tech giants like Google, Amazon, Intel, NVIDIA have already been pushing the boundaries of artificial intelligence. This can be seen clearly when we can see that it has been estimated that AI spending will touch nearly $98 billion in 2023 which makes the CAGR for 2018-2023 forecast period 28.4%. Another insight shows that the AI market is expected to reach $208.49 billion by 2025 with a CAGR of 32.6%. So it won’t take an highly-intelligent bot to recognise that it would be wise to put AI and ML stocks on the watchlist. As we move towards the ‘20s, let’s take a look at some of the hottest and most outstanding AI\/ML stocks that you should be following this year. NVIDIA (NVDA) Since the advent of artificial intelligence and deep learning, the need for faster processes has increased, rather, exploded. NVIDIA generates most of its revenue from video gaming GPUs, but these GPUs are also used to make the deep learning tasks better than CPUs like Xeon processors. Though there was a dip in its stock in 2019, the revenue grew subsequently in the last two quarters with the development of new conversational AI platforms. The new conversational AI platform boosted the GPU demand from hyper-scale customers, and its application will continue in 2020 with NVIDIA’s rising stock up along with showcasing AI advancements in other markets including its new Shield TV devices, 5G networks, IoT devices and of course better GPUs. NVIDIA’s investments in other areas have put it on a long-term earnings growth rate of 9.4%. The Zacks consensus mark for 2020 earnings has increased by 4.1% to $5.56 per share over the past two months or so. Baidu (BIDU) Baidu is China’s Google. The reason for its AI stock growth isn’t just because it is China’s largest search engine, but also because of its same interest in AI and other technologies. Baidu’s interest also lies in technologies like cloud, smart speaker, driverless car market etc. The essence of Baidu’s AI ecosystem is DuerOS, which a voice assistant. DuerOS hit 100 million users in August 2019 and doubled to 200 million in January. With China’s government relying heavily on Baidu to boost its national AI efforts, the company has a long-term earnings growth rate of 2.3% and the estimated consensus for 2020 earnings improved up to 30% to $8.45 over the past two months. Google (GOOGL) Well, Google‘s name has to be here, not only because it owns the world’s largest search engine, but also because of its substantial research and investments in artificial intelligence. Naturally, Google will be on every AI stocks watchlist. Google’s massive ecosystem is a data hub or a gold mine for the AI developers. The sheer size of the data is used for its AI services and Google’s AI tools to optimise its search engine. Google formed its ‘Google Brain’ team eight years ago for deep learning AI research. Google has AI-powered encryption tools, image enhancement tools, robotics projects, TensorFlow AI platform and custom chips for machine learning which gets integrated with all these services on an AI platform. Over the past five years, Google has been showing an annual sales growth of 8.71%. Box (BOX) Box offers cloud management solutions for automation, security, and application development. The company provides clients with advanced AI and ML solutions through its Box Skills solutions program, which includes solutions like structure and extracts insights from information. The company has a long-term earnings growth rate of 26%. Fortinet Inc. (FTNT) This company provides integrated as well as automated cybersecurity solutions that have been booting ML features for its FortiWeb Web Application Firewall. The FortiWeb Web Application Firewall works towards giving you better threat detection using machine learning. Alongside, other products and services are Fortigate and FortiGuard that leverage ML threat detection. Fortinet has a long-term earnings growth rate of 14% and a Zack Consensus Estimate for earnings for 2020 of 11.5% to $2.71 over the past two months. Keysight (KEYS) Keysight provides electronic design and test instrumentation systems. Keysight’s integrated models make use of ML technologies. The company’s smart water project utilises ML’s deep learning systems to detect contamination. Keysight’s long-term earnings growth rate of 9.1% and a consensus mark for 2020 improved to 6% to $5.19 over the past 60 days. Synaptics (SYNA) SYNA is no-blue chip AI stock of a company named Synaptics which in the designing and marketing is very similar to human interface solutions market such as touch screen controllers, touchpads for notebook computers, sensors for the mobile device. The company utilises smart edge AI in its services. The company’s audio smart offerings include fully integrated SoCs which have neural network acceleration, a proprietary wake word engine and advanced far-field voice processing. The company’s consensus estimates for full-year earnings has moved to 86.70% higher over the past 90 days. Commvault (CVLT) CVLT is an AI stock from company Commvault which provides data management solutions for high performing data protection, management of information on complex storage networks, universal availability. This company’s ML algorithms and AI processes can also identify patterns and detect problems from file activity and change rates. This helps when a potential threat or issue is detected and is alerted to the clients. CVLT has an excellent long term investment with a 1-year forecast of about $51.46 with currents price $45.95. Micron Technology (MU) With massive AI systems being excessively used, the demand for digital memory will increase in time as network and demand for internet also rise, which is good news for a company like Micron and individuals who are thinking of investing in Micron AI stock. Micron is one of three leading producers of DRAM which feeds data to a processor. Along with applications like 5G, there is an expectation of demand for AI and memory demand. Hence there will be subsequent demand for DRAM. Palo Alto Networks (PANW) Palo Alto Networks is a company like Fortinet, which provides cybersecurity solutions to a firm. Palo Alto Networks introduced Cortex which is the industry’s only open and integrated AI-based continuous security platform. Since Palo Alto’s last earnings reports, its shares have added about 5.2% in that time frame outperforming S&P 500. During the first fiscal year 2020, Palo Alto reported earnings of $1.05 per share, which surpassed the Zacks Consensus estimate of $1.03.","excerpt":"As businesses continue investing in artificial intelligence and with the increasing applications it is finding in every sector, the trajectory of these lucrative stocks might never come down. If we look back at 2019, which seems like a long time ago already, companies across sectors have started integrating AI and have implemented it in their […]","categories":["AI Trends"],"tags":["AI Stocks","china ai investments","NVIDIA"],"author_name":"Sameer Balaganur","publish_date":"2020-01-08T10:00:00","publication_year":"2020","word_count":1109,"keywords":["artificial intelligence","machine learning","AI","neural network","ML","AI Stocks","RAG","china ai investments","deep learning","edge AI","NVIDIA","TensorFlow","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","neural network","TensorFlow","edge AI","RAG","R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-10-ai-ml-stocks-that-you-should-buy-in-2020\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":15763,"title":"Bangalore startup Unbxd raises $12.5 million funding, to invest in AI engine to boost product discovery","content":"Product discovery and analytics service provider Unbxd raised 12.5 million dollars in Series C funding led by Eight Roads Ventures, the proprietary investment arm of Fidelity International Limited, and from existing investors including IDG Ventures, Inventus Capital Partners and Nirvana Ventures. With the funds, the Bangalore based startup will invest in its artificial intelligence (AI) powered platform to enable online retailers to capture the 15%-40% in revenue that is lost due to a poor product discovery experience. The startup power 120+ US online retailers and helps them boost revenue through better product discovery. What’s unique about the Unbxd AI is that it powers personalized product discovery platform and understands user’s’ intent to enhance their online shopping experiences, increasing conversions and revenue. The current round of funding will be used to boost Unbxd’s proprietary Intelligence Engine. According to a statement, the Unbxd Intelligence Engine today captures and processes more than 50 shopper data signals, to power personalized search, navigation and product recommendations on e-commerce sites. The startup also plans to open new offices in Chicago and New York to be close to the customers.  Unbxd enables personalized site search, navigation, product recommendations and merchandising for leading online retailers including Ashley HomeStore, Express, FreshDirect, Rue21 and ibSupply. The statement reveals the startup powers more than 1.5 billion shopper interactions a month. Kris Gopalakrishnan, Infosys Co-Founder, also invested in the round. Avendus Capital was the exclusive financial advisor to Unbxd’s Series C round. The statement sheds light on the robust growth of the startup – Unbxd revenue grew 400% in the past year and it strengthened its Internet Retailer 500 customer base by adding leading retailers including Ashley HomeStore, Express and FreshDirect. According to Pavan Sondur, Co-Founder & CEO of Unbxd, “Traditional rules-based e-commerce technology is hard for retailers to manage and at the same time delivering a poor experience for their shoppers. With a better AI solution in place, retailers are more competitive because shoppers can more easily buy what they seek”.","excerpt":"Product discovery and analytics service provider Unbxd raised 12.5 million dollars in Series C funding led by Eight Roads Ventures, the proprietary investment arm of Fidelity International Limited, and from existing investors including IDG Ventures, Inventus Capital Partners and Nirvana Ventures. With the funds, the Bangalore based startup will invest in its artificial intelligence (AI) […]","categories":["AI News"],"tags":[],"author_name":"Richa Bhatia","publish_date":"2017-06-21T03:54:52","publication_year":"2017","word_count":330,"keywords":["Go","API","funding","artificial intelligence","programming_languages:R","AI","programming_languages:Go","analytics","R","startup"],"extracted_tech_keywords":["AI","artificial intelligence","analytics","R","Go","API","startup","funding","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/bangalore-startup-unbxd-raises-12-5-million-funding-invest-ai-engine-boost-product-discovery\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10039041,"title":"MLCommons Releases Latest Benchmark MLPerf Inference v1.0","content":"Open engineering consortium MLCommons has published the results for its machine learning inference performance benchmark suite, MLPerf Inference v1.0. The results gauged how quickly a trained neural network can process new data for a wide range of applications on various form factors and a system power measurement methodology, the MLCommons statement said. The latest benchmark includes 1,994 performance and 862 power efficiency results for leading ML inference systems. The foundation of MLCommons was laid in 2018 after a group of researchers and engineers released MLPerf, a benchmark for measuring the speed of machine learning software and hardware. The companies including AMD, Baidu, Google, Intel and researchers from Harvard University, Stanford University, University of California Berkeley, University of Minnesota, and the University of Toronto had backed the initiative. Last December, MLCommons was launched. Towards an efficient system MLCommons brought together diverse companies and organisations to create vast public databases for AI training, allowing researchers from all over the world to collaborate at a higher level and advance the field as a whole. The submissions were received from Alibaba, Centaur, DellEMC, EdgeCortix, Fujitsu, Gigabyte, HPE, Inspur, Intel, Krai, Lenovo, Moblin, Neuchips, Nvidia, Qualcomm, Supermicro and Xilinx. The system power measurement methodology was developed in collaboration with Standard Performance Evaluation Corp. (SPEC), a non-profit corporation formed to create standardised benchmarks and tools to assess computing systems’ performance and energy efficiency. “As we look at the accelerating adoption of machine learning, AI, and the anticipated scale of ML projects, the ability to measure power consumption in ML environments will be crucial for sustainability goals across the globe,” said Klaus-Dieter Lange, SPEC Power Committee Chair. MLCommons built MLPerf in the best tradition of vendor-neutral standardised benchmarks, and SPEC is excited to be a partner in their development process. We look forward to extensive adoption of this extremely valuable benchmark, added Lange. Apart from this addition of the power category, MLPerf also made two small rule changes. “One is that for data centre submissions, we require that any external memory be protected with ECC. And we also adjusted the minimum runtime to better capture sort of equilibrium behaviour,” said David Kanter, Executive Director of MLCommons. The report is released every year. As per Kanter, there’s a gap of six months between the inference and training. However, one cycle was missed last year due to the pandemic. Results The different scenarios and MLPerf models used in the latest inference exercise: The annual benchmark tests offer chip vendors and system developers an opportunity to showcase how well they can do a representative set of machine learning tasks. NVIDIA is the only company to make submissions in all three tests conducted so far. Last year and this year, NVIDIA achieved record-breaking success in every category with its newly launched A30 and A10 GPUs. “As AI continues to transform every industry, MLPerf is becoming an even more important tool for companies to make informed decisions on their IT infrastructure investments,” said Ian Buck, General Manager and Vice President of Accelerated Computing at NVIDIA. He further added, “Now, with every major OEM submitting MLPerf results, NVIDIA and our partners are focusing not only on delivering world-leading performance for AI but on democratising AI with a coming wave of enterprise servers powered by our new A30 and A10 GPUs.” The industry needs a mutually agreed-upon range of best practices and metrics to promote the ongoing growth, deployment, and sharing of machine learning and AI technologies and assess quality, speed, and reliability. Moreover, reproducibility is a challenge in ML. MLCommons benchmarks will help advance the field and also save others from reinventing the wheel.","excerpt":"The latest benchmark includes 1,994 performance and 862 power efficiency results for leading ML inference systems.","categories":["AI Trends"],"tags":[],"author_name":"kumar Gandharv","publish_date":"2021-04-28T11:00:00","publication_year":"2021","word_count":601,"keywords":["Go","machine learning","programming_languages:R","AI","neural network","ML","programming_languages:Go","GAN","R"],"extracted_tech_keywords":["AI","machine learning","ML","neural network","R","Go","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/mlcommons-releases-latest-benchmark-mlperf-inference-v1-0\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10078134,"title":"Harnessing the Power of India’s Public Data","content":"In 2010, erstwhile US President Barack Obama visited the first-ever expo on ‘Democracy and Open Government’ held at St. Xavier’s College, Mumbai. This was the time when then PM Manmohan Singh and President Obama came together to strengthen Indo-US bilateral ties, and discussed the formation of the open government data platform – data.gov.in – inspired by USA’s famous data.gov initiative. As per the National Informatics Centre, Indian ministries, departments and their organisations are meant to use the web portal to publish datasets, documents, services, tools, and applications that they have gathered for public use. The platform, jointly developed by the two countries, aims to promote government transparency and create opportunities for a wider use of government data. There are over 4.7k catalogues and 5.8 lakh datasets on the platform. The collection spans a variety of industries, including biotechnology, youth and sports, law, and science and technology. Additionally, one can opt to browse datasets based on where they are from; for example, one can select which central or state department or ministry they need a dataset from. Future ahead: National Data Platform Users of data.gov.in have observed that the website is either slow or falls behind in terms of data. The datasets available on the portal are outdated and not very useful. Now, to address the issues and make data-sharing more feasible, the government has decided to launch a new platform for the data geeks, i.e, the National Data Platform. The proposed platform would be a full-featured data marketplace environment with built-in analytics, developer sandbox, visualisation engine, data tools, app\/model development, modular metadata architecture, and adoption of various data-sharing policies. With cutting-edge AI ML-driven search & discovery, this will also handle the directories of all data available in India. As per data.gov.in, the platform will be a “one-stop solution to discover, harvest and register data from participating data exchanges and registries”. In focus: National Data and Analytics Platform The National Data and Analytics Platform (NDAP) is Niti Ayog’s primary effort to increase public data availability and utilisation. An easy-to-use web platform, NDAP collects and hosts datasets from all over India’s extensive statistical infrastructure. The NDAP seeks to democratise data delivery by resolving many of the challenges that currently exist in the open data universe. The NDAP makes government data sets easily accessible and interoperable while reducing data loss. Available through a seamless user interface and easily visualised with interactive visualisations, NDAP resolves many of the data challenges faced by users. On the portal, 601 datasets from 46 ministries are dispersed throughout 15 sectors. Users can access datasets such as the recent Central Electricity Authority (CEA) monthly power reports or the latest information on digital transactions, BHIM transactions, and debit cards, for instance. Datasets uploaded to the platform must adhere to a minimal level of data quality as determined by NDAP’s own 5-star grading system. By conforming to this minimal standard, all datasets on NDAP are guaranteed to be supported by documentation, mapped to a common data schema (the local government directory), and have successfully passed internal data quality checks to ensure that they remain accurate representations of the original data. Additionally, the ministry of Panchayati Raj local government directory code is used to map datasets on NDAP to a shared set of spatial and temporal identifiers. Users can combine datasets from several industries and sources in order to do quick cross-sectoral analysis. As per Amitabh Kant, CEO, Niti Ayog, “The rise of data and digital technologies are rapidly transforming economies and societies. This has enormous implications for governments’ daily operations. NDAP is a critical milestone, which aims to aid India’s progress by promoting data-driven disclosure, decision making and ensuring availability of data connecting the last mile. This is an example of how the power of data can be leveraged.” To know more, visit here.","excerpt":"In 2010, erstwhile US President Barack Obama visited the first-ever expo on ‘Democracy and Open Government’ held at St. Xavier’s College, Mumbai. This was the time when then PM Manmohan Singh and President Obama came together to strengthen Indo-US bilateral ties, and discussed the formation of the open government data platform – data.gov.in – inspired […]","categories":["IT Services"],"tags":["open data"],"author_name":"Lokesh Choudhary","publish_date":"2022-10-27T13:00:00","publication_year":"2022","word_count":633,"keywords":["Go","API","open data","AI","ML","Git","RAG","Aim","analytics","edge AI","R"],"extracted_tech_keywords":["AI","ML","analytics","Aim","edge AI","RAG","R","Go","Git","API"],"url":"https:\/\/analyticsindiamag.com\/it-services\/harnessing-the-power-of-indias-public-data\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":1167,"title":"Trends in Analytics &#8211; 3 experts offer their insights","content":"Galit Shmueli, SRITNE Chaired Professor of Data Analytics at ISB Hyderabad First, the most important step that many organizations in India must jump is data management. Getting the data stored and organized in a way that is accessible and effective for BA is crucial, yet very few companies are in that stage. Second, those who have been able to move up this painful step are mostly involved in ‘business intelligence’, which is focused on data management and reporting. While cutting edge reporting and data visualization tools are dramatically changing the scene, these are still not the more advanced ‘predictive analytics’. Last, while different types of data will continue to evolve (from text, to videos, to social network data, and beyond), the basics of analytics will remain the same. Building a strong basis and keeping abreast of new applications, software and techniques will likely be the secret of strong BA-based organizations. [divider top=”1″] Raj Mohan KK, CEO at Satvik Traditionally, Data means the information’s stored by enterprises themselves, such as financial data stored in some Financial applications, customer data stored in CRM applications, operational data stored in ERP systems etc. We call this as Structured data. Both machine- and human-generated online transactions have been churning out large volume of data in varied formats in high velocities. We call this as UnStructured data. We call the combination of the Structured and Un Structured data as Big Data. This brings in enormous opportunities and challenges to organizations. Traditional data management and business analytics tools, services and technologies are un able to attend this huge Volume, Velocity and Variety of data in an effective manner. This requires very radical thinking in the way organizations would like to harness the power of this sudden influx of huge volume of useful data. This will not be an easy transition for most enterprises, but those that undertake the task and embrace Big Data as the foundation of their Data Management practice stand to gain significant competitive advantages. By designing and implementing right business solutions making use of the power of Big Data, enterprises can gain unprecedented insights into customer behavior, potential risks, volatile market conditions etc, thus facilitating them to make fact- driven business decisions faster and more efficiently. Solutions in Consulting to Big Data Analytics to Front-end data visualization bring significant opportunities. Those vendors that aid the enterprise in its transition to Big Data practitioner, both in the form of identifying Big Data use cases that add business value and developing the technology and services to make Big Data a practical reality, will be the ones that thrive. [divider top=”1″] Dr. Dakshinamurthy V Kolluru, President at International School of Engineering, Hyderabad There a lot of new trends in analytics and a lot of hype about big data. People seem to be mixing big data and analytics. It’s important for a practitioner to understand where each specialization fits. In the role of an analyst it’s important to identify and detect non trivial patterns in data using techniques like forecasting, prediction, classification, etc. With internet, the kind of data that needs to be analysed has changed so much that there’s a necessity to have a new field whose job is to compile and collate the data before the analysts start to analyse it. The amount of data generated by Google or Facebook runs into several tera or hexa bytes. First, this data should be distributed in hardware, so analysis can be done in real time. There are two things to be considered here: (i) To put data on hardware and then (ii) The analysis in near real-time. Classical big data engineers specialise in technologies like Hadoop, MapReduce, etc. They focus on putting together the data and structuring the data so analysis can be done in real-time. These are two distinct skills that are needed to handle a set of problems effectively. Hadoop and MapReduce are used to engineer the data for analysis. Decision trees and neural nets can be used to solve analytics problems. Today, the focus is more on finding out the right algorithms and working it out rather than just a tool based approach.","excerpt":"Galit Shmueli, SRITNE Chaired Professor of Data Analytics at ISB Hyderabad First, the most important step that many organizations in India must jump is data management. Getting the data stored and organized in a way that is accessible and effective for BA is crucial, yet very few companies are in that stage. Second, those who […]","categories":["IT Services"],"tags":["analytics insight","international school of engineering hyderabad"],"author_name":"Дарья","publish_date":"2012-09-18T09:56:18","publication_year":"2012","word_count":688,"keywords":["big data","Go","business intelligence","programming_languages:R","AI","R","programming_languages:Go","analytics insight","analytics","GAN","predictive analytics","international school of engineering hyderabad"],"extracted_tech_keywords":["AI","analytics","predictive analytics","R","Go","big data","GAN","business intelligence","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/trends-in-analytics-3-experts-offer-their-insights\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10171710,"title":"India Sees Rise in Operator-Turned-Founders as Ecosystem Matures","content":"A recent analysis of Tracxn data, commissioned by early-stage venture capital firm RTP Global, highlights a significant rise in operator-led startups in India. Between 2022 and 2024, 238 former operators—seasoned professionals who previously scaled businesses—transitioned to founding their own ventures. This surge is attributed to factors such as liquidity from ESOP buybacks, a desire for new challenges, returning talent from the US, and the effects of mass layoffs. Faster and Better? The data reveals that operator-led startups are achieving seed funding milestones faster, raising larger seed rounds, and securing stronger investor backing, particularly in capital-intensive and regulated sectors where execution expertise is critical. In 2024 alone, these startups raised $101 million—a 243% increase from 2023 and nearly double the $52.9 million raised in 2022. Operator-led ventures accounted for 11.5% of all venture funding in 2024, up from 6% the previous year. At the seed stage, operator-led startups founded in 2022 secured 11.2% of seed rounds, compared to 4.4% for other Indian tech startups. This trend strengthened in 2023 (12.6% vs 2.7%) and remained steady in 2024 (8.1% vs. 1.5%), despite tougher funding conditions. The average seed round for operator-turned-founders between 2022 and 2024 was $1.56 million, 1.2 times larger than the $1.3 million raised by other startups. Graduating to Series A Operator-founders are also advancing to Series A rounds at significantly higher rates. Among startups founded in 2022, 7.1% of operator-led ventures reached Series A, compared to just 0.3% of other startups. This pattern continued in 2024, with 5.4% versus 0.1%. Although their average Series A round size ($8.2 million) was slightly smaller than the rest of India tech ($9.6 million), operator-led startups commanded much higher valuations—$38.5 million compared to $21.8 million—resulting in less dilution and greater leverage. Sector Picture Sector-wise, fintech and insurtech attracted the most capital ($89 million between 2022 and 2024), followed by energy and sustainability ($40.3 million), and supply chain and logistics tech ($30 million), reflecting the ongoing need for operational efficiency. In terms of startup volume, five sectors dominated operator-led ventures: fintech and insurtech, e-commerce and retail tech, edtech, enterprise SaaS, and healthcare and life sciences—areas where founder experience and execution skills accelerate early success. “We’ve always believed in backing exceptional entrepreneurs,” said Nishit Garg, investment partner at RTP Global. “Many operator-founders bring a pragmatic mindset shaped by real-world experience that helps them scale efficiently. This analysis broadens our understanding of diverse founder journeys.”Neha Singh, co-founder and CEO of Tracxn, added, “The startup ecosystem is constantly evolving, and founder profiles are diversifying.”","excerpt":"In the past two years, over 200 people who aced operations and scaled businesses have gone on to found their own startups, study finds.","categories":["AI News"],"tags":["Startups","traxcn"],"author_name":"C P Balasubramanyam","publish_date":"2025-06-13T11:37:59","publication_year":"2025","word_count":417,"keywords":["Go","API","funding","programming_languages:R","AI","venture capital","programming_languages:Go","RAG","traxcn","Startups","R","startup"],"extracted_tech_keywords":["AI","RAG","R","Go","API","startup","venture capital","funding","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/india-sees-rise-in-operator-turned-founders-as-ecosystem-matures\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":5890,"title":"Flytxt takes top honors for Product Innovation at Frost and Sullivan India ICT Award","content":"India, June 26th, 2014: Flytxt, a Big Data Analytics solution partner for more than 50 Communication Service Providers (CSPs), was recognized as a leader in Marketing Analytics category at the 2014 Frost & Sullivan India ICT Awards. Flytxt enables Communication Service Providers (CSPs) to increase revenue, reduce churn, enhance customer experience and generate new revenue streams through its integrated multi-dimensional analytics powered solutions. Frost & Sullivan India ICT Awards, currently in its 12th year, acknowledges the exemplary companies and individuals who have pushed the boundaries of excellence –risen above competition and demonstrated outstanding performance in the Information and Communications Technology (ICT) sector in India. “We are delighted to receive this recognition from Frost & Sullivan for our advanced analytics platform and solutions. It validates our strategy to offer a total solution portfolio of technology platform, business applications and enabling services leveraging our proprietary integrated multi-dimensional analytics framework, to enable CSPs to generate more than 10% economic value from available data”, said, Abhay Doshi, V.P-Product Management and Marketing, Flytxt. He added, “We have already developed and deployed strong internal and external Data monetization use cases for our customers across the globe, delivering 2 to 7% economic impact to them consistently”. Flytxt solutions are deployed with more than 50 CSPs across Asia and EMEA, driving their data analytics-driven initiatives for customer experience and revenue management as well as for insight monetization through enabling adjacent services like mobile advertising. About Flytxt Flytxt is a leading provider of Big Data Analytics solutions that enable CSPs to derive measurable economic value from subscriber data. The company offers Customer Experience & Revenue Management and Data Monetization Solutions as well as consultancy services to enable operators to run campaigns for increasing revenue, reducing churn, enhancing loyalty and creating new revenue streams. Flytxt’s closed loop integrated real-time marketing platform has been selected by leading CSPs across APAC and EMEA, serving more than 500 million subscribers and has generated over $350 million incremental revenue for them till now. The company has won many industry awards like NASSCOM Emerge 50 league of 10, Aegis Graham Bell Award for innovation in Mobile Advertising, BID International Quality Summit Award, Red Herring Asia 100 and IEEE Cloud Computing Challenge. With its headquarters in the Netherlands, corporate office in Dubai and global delivery centers at Trivandrum and Mumbai in India, the company has presence in New Delhi, Lagos, Nairobi, Kuala Lumpur and London. For more information about Flytxt, visit www.flytxt.com.","excerpt":"India, June 26th, 2014: Flytxt, a Big Data Analytics solution partner for more than 50 Communication Service Providers (CSPs), was recognized as a leader in Marketing Analytics category at the 2014 Frost & Sullivan India ICT Awards. Flytxt enables Communication Service Providers (CSPs) to increase revenue, reduce churn, enhance customer experience and generate new revenue […]","categories":["AI News"],"tags":[],"author_name":"AIM Media House","publish_date":"2014-06-28T10:52:18","publication_year":"2014","word_count":406,"keywords":["big data","Go","programming_languages:R","cloud computing","AI","innovation","RAG","analytics","R","analytics platform"],"extracted_tech_keywords":["AI","analytics","RAG","cloud computing","R","Go","big data","analytics platform","innovation","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/flytxt-takes-top-honors-for-product-innovation-at-frost-and-sullivan-india-ict-award\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10117665,"title":"AI to Impact 1.1 Bn Roles for People Across Culture, Genders and Abilities","content":"At the keynote of India’s biggest diversity and inclusive conference, The Rising 2024, Vidya Rao, chief technology and transformation officer, said billions of roles are prone to radical transformation because of AI. “In this decade, over 1.1 billion roles are set to undergo a radical transformation due to the pervasive influence of technology,” said Rao. She said this seismic shift in the job market isn’t confined to a single demographic but encompasses a wide spectrum of individuals from diverse cultures, genders, and abilities. Further, she explained that such a transformative landscape presents a plethora of opportunities, necessitating that our workforce will become increasingly varied and multifaceted to meet these evolving challenges. Rao believes that embracing this diversity is not just a checkbox for organisations but a strategic imperative to drive meaningful change and progress. Genpact has invested over $600 million in generative AI initiatives and is working on hundreds of GenAI PoCs across use cases. “AI is going to invade everything that we do,” said Rao, adding that technology is in the flow of life and that one cannot imagine a world without it. Rao said that there is a clear understanding that embracing AI will drive meaningful change. “And have you arrived? I don’t think so. There is a very, very long way to go,” she added, saying that just having this agenda as one of the foundations to run your company will bring huge change. The 9 Tenets of Diversity, Equity and Inclusion: Rao’s keynote addressed the importance of acknowledging and overcoming biases, personally committing to diversity, equity, and inclusion (DE&I), continuously upskilling, being courageous in communication, building supportive relationships, fostering inclusive communication, being an ally and advocate, engaging in mentorship and partnerships, and celebrating diversity to cultivate an innovative, inclusive, and adaptable tech environment. With over 30 years of experience, Rao currently leads Genpact’s internal digital transformation processes. Her role also includes driving the global ERP programme and collaborating within Genpact’s partner ecosystem to support transformation at scale. Prior to Genpact, she worked with Accenture, performing diverse roles and leading technology delivery for multinational clients. As a coach for budding women leaders, Rao takes great pride in mentoring the next generation of professionals. She has a degree in science, majoring in Statistics, from Mumbai University, India. Equality vs Equity Rao shared the importance of recognising and addressing unconscious biases that can prevent equal opportunities. She said true equity goes beyond treating everyone equally and ensuring everyone has the support and resources they need to succeed. “Diversity sparks innovation, inclusion fosters belonging, equality ensures opportunities, and equity reduces barriers,” said Rao, succinctly capturing the essence of DEI (Diversity, Equity, and Inclusion) in the tech world. She outlined how each element contributes to a more dynamic and equitable working environment. Sharing real-life anecdotes, Rao stressed the importance of creating a work environment where employees feel safe expressing themselves and taking risks. This will lead to constructive behaviours and business growth. “Fostering an environment that can encourage courageous behaviours is very important because by providing an environment with psychological safety, you will see that people feel included and truly come up with constructive behaviours that will flourish their own businesses.” Building Strong Relationships and Investing in Support Systems “You can have the best support system, but it will not work if you do not invest time in building that relationship—be it your parents, in-laws, family, or even your support staff, maids, drivers, etc.,” said Rao. She said that one should invest in relationships; not just as a transaction, but as an integral part of social responsibility and human upliftment. She believes that all successful people have a fantastic support system behind them. “It is not easy to be successful if you do not have a system, and that comes only by investing time and building relationships,” said Rao. Further, Rao advocated for diversity at all levels, urging disruption of the status quo to create a more inclusive world for future generations. “We need diverse talent, leadership, and workforce. We have to challenge, disrupt the status quo, and create a world that is more inclusive for the coming generations,” she added. In concluding remarks, Rao highlighted the importance of women supporting each other professionally, fostering collaboration instead of competition. “Women helping women is very important in this agenda. It’s not just about competition but about raising each other up,” she concluded.","excerpt":"“Diversity sparks innovation, inclusion fosters belonging, equality ensures opportunities, and equity reduces barriers,” Vidya Rao, the chief technology and transformation officer at Genpact said.","categories":["AI Features"],"tags":["AI Impacts","AIM Rising summit"],"author_name":"K L Krithika","publish_date":"2024-04-05T10:13:06","publication_year":"2024","word_count":730,"keywords":["Go","GenAI","AWS","AI","AIM Rising summit","digital transformation","Git","RAG","generative AI","AI Impacts","GAN","R"],"extracted_tech_keywords":["AI","generative AI","GenAI","RAG","AWS","R","Go","Git","GAN","digital transformation"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/ai-to-impact-1-1-bn-roles-for-people-across-culture-genders-and-abilities\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":45592,"title":"IIT Guwahati Develops AI-Powered Chatbot ALBELA To Support 1st Year Students","content":"In an interesting invention that will help student communication, a team of postgraduate students from the Indian Institute of Technology, Guwahati are developing an artificial intelligence-enabled chatbot called ALBELA. Created by the researchers from the Department of EEE and their faculty members, this chatbot is going to be used to teach and support the first-year students of Electrical & Electronics Engineering (EEE). IIT Guwahati said in a statement that the AI chatbot is projected to be a suitable mechanism also capable of addressing the queries and doubts that each of the approximately 850 students pursuing EEE at the Institute may have. Talking about the chatbot ALBELA, Praveen Kumar, Professor, Department of EEE, IIT Guwahati, said, “We have been working on its development since last 7 months with a team of dedicated 7 research scholars of the department. Earlier we did the trial runs of the chatbot, and started using from this academic session onwards. The response from the students has been overwhelming and we hope that this will become the new normal in the near future. Prof Rohit Sinha, Head EEE Department, and the team IBM have extended their continuous support for this activity.” How is ALBELA helping students at IIT Guwahati? Providing an easy interface to find details of classes and tutorials: Students can find their class schedule, tutorial schedule, examination queries, and more through a simple AI-based chat window. Students at times may hesitate to approach an instructor regarding their queries, but, with this chat-based system, students can clear their doubts, both technical and non-technical. Doubts related to course content: First-Year students can have numerous doubts related to content such as preferred textbooks, reference books, notes, sample questions, etc. The Chatbot is trained to answer all the queries related to the above. An easy and fascinating interface: Use of Chatbot to clear doubts of the students can be fascinating to students, increasing their interest in the subject. This may result in better learning outcomes, with the students enjoying the course even more.","excerpt":"In an interesting invention that will help student communication, a team of postgraduate students from the Indian Institute of Technology, Guwahati are developing an artificial intelligence-enabled chatbot called ALBELA. Created by the researchers from the Department of EEE and their faculty members, this chatbot is going to be used to teach and support the first-year […]","categories":["AI News"],"tags":["AI Chatbot","AI Chatbot","chatbot ai","IIT Guwahati","red hat"],"author_name":"Prajakta Hebbar","publish_date":"2019-09-06T13:23:40","publication_year":"2019","word_count":334,"keywords":["Go","red hat","artificial intelligence","programming_languages:R","AI","AI Chatbot","programming_languages:Go","chatbot ai","IIT Guwahati","ViT","R"],"extracted_tech_keywords":["AI","artificial intelligence","R","Go","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/iit-guwahati-develops-ai-powered-chatbot-albela-to-support-1st-year-students\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10004964,"title":"Leveraging Computer Vision In Drone Tech","content":"The last talk of the Day 2 “Leveraging Computer Vision In Drone Tech” was presented at the  Computer Vision conference of the year, CVDC 2020 by Animesh Dutta, who is the Machine Learning Developer at Kesowa. CVDC 2020 is scheduled for 13th and 14th of August, organised by the Association of Data Scientists (ADaSCi), the premier global professional body of data science and machine learning professionals. Dutta kick-started the talk by discussing the ecosystems of drones and how drone tech is evolving quite rapidly by leveraging emerging technologies like artificial intelligence, deep learning, autonomy and unmanned aerial systems. Regarding the ecosystem, Dutta mentioned some of the important components which constitute a drone ecosystem. They are- Hardware, which includes delivery systems, drone platforms, drone-in-a-box, VTOL fixed-wing, passenger drones, counter drones, components & systems, among others.Software, which includes flight, fleet & operation management, open-source infrastructure, navigation, Computer Vision & AI, LAANC suppliers, data analytics, among others.Services such as delivery, drone-as-a-service, simulation & training, maintenance, market places, coalitions, initiatives, among others. The speaker then discussed the steps that are required to build a drone by showing an example of a drone that is built in-house by Dutta and his team at Kesowa. Talking about the drone, Dutta discussed the data flow for streaming of the drone along with some of the hardware and software components that were used while building the drone. The hardware components used are- One Quadcopter frameFour sets of motorsOne LiPo batteryFlight controller option, such as Pixhawk 2.4.8 32bit ARM RCGPS moduleRC receiver-transmitter setESC-battery bullet connectorLanding gearCamera as per requirement The software components used in the above-shown drone are- QGCS or any other open-source ground stations.MAVLINKMission plannerStable InternetGPS at all timesRTMP protocol to feed video to the server. After discussing a brief view on how one can build a drone from scratch, Dutta explained what Computer Vision is and the important use cases related to drones including how drones can help monitor crowds during the pandemic situation. Some of the use-cases mentioned are- Leveraging emerging technologies to create and deliver solutions to strengthen governance.Detecting anomalous behaviour in the crowd. Using Bayesian Loss for crowd count estimation with point supervision helps to generate a density map showing humans. Drones can be used to deliver items from one location to another. Using Computer Vision techniques, one can detect potholes on the road, find corrosion in towers and bridges, among others.Drone data can also be in the form of a tiff file. One can calculate the NDVI value and compute the vegetation cover area in that region. Talking about the pandemic situation, the speaker put forward the following points- Drones can be used to calculate the density of people in a particular place. If the threshold for density exceeds and people are not maintaining social distancing, measures were taken. The megaphones on the drones connected to the cell phones of the police officers will allow them to simply dial in and make their announcements.Drones are being used for dropping off medical supplies and personal protective gear to healthcare workers in order to spray disinfectant over public spaces like airports, railway stations, among others.Drones fitted with thermal scanners can identify potential virus carriers in a crowd. As an instance, Dutta showed the crowd density estimation by a drone using Computer Vision, which was deployed to help battle COVID-19 in April 2020. He said, “Crowd counting is a very interesting task and challenging as well. There are two main difficulties involved, which are variation in human size due to the perspective and overlap & occlusion in a dense crowd.” He also mentioned a research paper by Zhiheng Ma and team, who proposed a novel Bayesian loss function which constructs a density contribution probability model from the point annotations. Lastly, Dutta concluded the talk by discussing some of the important topics of Computer Vision, which are crucial while building a drone with emerging technologies. The topics included an explanation of Mak R-CNN, steps of detecting rooftops using Mask R-CNN, Faster R-CNN, Region Proposal Network (RPN), Region of Interest (RoI), Normalised Difference Vegetation Index (NDVI), autonomous drone flights using generalised autonomy aviation system (GAAS), among others.","excerpt":"The last talk of the Day 2 “Leveraging Computer Vision In Drone Tech” was presented at the  Computer Vision conference of the year, CVDC 2020 by Animesh Dutta, who is the Machine Learning Developer at Kesowa. CVDC 2020 is scheduled for 13th and 14th of August, organised by the Association of Data Scientists (ADaSCi), the […]","categories":["Deep Tech"],"tags":["Computer Vision","Computer Vision conference","Computer vision India","Drone","Drone Artificial Intellgence"],"author_name":"Ambika Choudhury","publish_date":"2020-08-18T10:00:00","publication_year":"2020","word_count":689,"keywords":["Computer vision India","Drone","data science","artificial intelligence","Drone Artificial Intellgence","machine learning","AI","Computer Vision conference","computer vision","RAG","Ray","deep learning","analytics","Computer Vision","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","deep learning","computer vision","data science","analytics","Ray","RAG","R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/leveraging-computer-vision-in-drone-tech\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":22483,"title":"Microsoft In Association With Apollo Hospitals To Use Artificial Intelligence In Cardiology","content":"Microsoft recently announced to bring artificially intelligent solutions to boost research in the area cardiology. The announcement which came at the Healthcare Information and Management Systems Society (HIMSS) annual conference and exhibition in Las Vegas, Microsoft showed its interest in the expansion of Microsoft Intelligent Network for eyecare (MINE), to create AI focused network in cardiology. They have plans of partnering with Apollo Hospitals, one of the largest health systems in India. They would develop and deploy new machine learning models to predict patient risk for heart disease and assist doctors on treatment plans. The team is already working on an AI powered Cardio API (application program interface) platform. “Our unique partnership with Apollo brings together our learnings in artificial intelligence and machine learning areas and Apollo’s experience and expertise in cardiology.  The systems of intelligence we create can change the lives of patients and the work of medical practitioners enabling accessible healthcare to all,” said Dr. Peter Lee, Corporate Vice President, AI & Research, Microsoft. “This collaboration marks a paradigm shift for patients and the management of non-communicable diseases the world over. Given our large clinical database and significant pool of clinical talent, this partnership will help impact the global burden of cardiac disease”, said Sangita Reddy, Joint Managing Director, Apollo Hospitals. Reddy also believes that this collaboration with Microsoft will help better predict, prevent and manage heart diseases in the country. “While we commence this in India, we will validate the algorithms and work towards creating a global consortium to tackle multiple conditions in cardiovascular disease,” she said. Microsoft had launched MINE in collaboration with L V Prasad Eye Institute to bring technology in the foreplay. The company had aimed at using data, advanced analytics and machine learning technology in the eye care. Anil Bhansali, Corporate Vice President, Cloud & Enterprise, Managing Director, Microsoft India (R&D) Private Limited said, “We started with eyecare – and today we are expanding into cardiac care.  We have been working with our customers and partners within the healthcare sector to create AI models that can help provide doctors and health care providers insights in their treatment plans.” The AI Network for healthcare is part of Microsoft Healthcare NExT aimed to accelerate healthcare innovation through artificial intelligence and cloud computing.","excerpt":"Microsoft recently announced to bring artificially intelligent solutions to boost research in the area cardiology. The announcement which came at the Healthcare Information and Management Systems Society (HIMSS) annual conference and exhibition in Las Vegas, Microsoft showed its interest in the expansion of Microsoft Intelligent Network for eyecare (MINE), to create AI focused network in […]","categories":["AI News"],"tags":[],"author_name":"Srishti Deoras","publish_date":"2018-03-09T06:12:56","publication_year":"2018","word_count":377,"keywords":["Go","API","machine learning","artificial intelligence","AI","cloud computing","Git","Aim","analytics","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","analytics","Aim","cloud computing","R","Go","Git","API"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/microsoft-in-association-with-apollo-hospitals-to-use-artificial-intelligence-in-cardiology\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":5879,"title":"Analytics for FIFA World Cup 2014","content":"I have to make a confession: I’m a not a die-hard soccer maniac. However, The World Cup is still the highlight of my year every time it rolls around.  It is fascinating to witness how Soccer Mania takes over the world.  It might only happen every 4 years, and most people only use it as an excuse to hang out in bars all day… But still, I love every minute of it. What I particularly love about this World Cup is the humungous amount of data that is being generated across social media and via other online\/offline channels, and how accessible it is. Data is a rising star in professional soccer and it could soon become some teams’ most valuable player. Although soccer has lagged behind other sports, such as baseball and basketball, in making broad use of data and analytics, the 2014 World Cup brings up numerous examples of data making an impact on many different aspects of the game. Already, stats and predictions for almost every facet of the event are flooding in; Brazil are expecting 3.7 million visitors, and a $3.03 billion boost to their economy; Panini are expecting £89.1 million in sticker sales in Brazil alone; in the UK, Domino’s Pizza stands to make an estimated £84 million during the World Cup. The one area where facts and figures seem comparatively scarce is in accurately predicting the World Cup winner. We can estimate how many people will be flying out to Brazil, how many pizzas the Brits will consume in front of their TVs, how many stickers rabid fans will collect; but can we use data to predict who might actually win? We take a look at the sceptics, and those such as Goldman Sachs, who are confident their data-driven models will be successful in predicting the World Cup winner. Continue reading at: Predicting the FIFA World Cup 2014 with Big Data! FIFA has an entire site for “statistics” that appears to update after each game is played.  It is a soccer fan’s dream – from aggregate stats including goals per match, average cards per match and average passes per team; to team stats, and even individual stats including top scorer, top runner, and top saves. The only thing missing? It’s not in real-time, which is a major drawback in this information era.  (And, to be honest, for being an organization with a TON of money – apparently around $1 billion – FIFA’s statistics site looks a little amateur.  I would love to see live infographics and other data visualizations.  Hey FIFA, I’m waiting for your call!) But beyond FIFA, everyone seems to be jumping on the data bandwagon and finding new insights before, during and after every single game. We have always had ESPN, or other broadcasters, who were armed with a seemingly ending supply of stats – but now the data sharing has expanded beyond just the broadcasters with high-power TV networks behind them: –  An interesting site is done by Brandwatch: they are collecting social content and providing brief analysis.  They have a globe showing the total number of global World Cup mentions on Twitter (6.5M since June 12th), the top hashtags, and topics. The sight seems to be struggling a bit with its font… but other than that, it’s a pretty cool use of Social Media data. – The Wall Street Journal has also jumped on the bandwagon – but they are not using real-time data to do it.  They created an interactive “bracket” that looks at “The World Cup of Everything Else.” They have some World Cup related items, but then also have “most McDonalds”, “most rainfall”, “biggest drinkers” (congratulations Russia), and “cheapest gasoline prices.”  It is a very fun way to use the World Cup to very smartly organize data, and utilize all kinds of data sets to create a very interesting and fun data visualization. Basically, what the World Cup is showing us, on a global scale, is that data is available to everyone now.  There are tools, databases and real-time feeds that make it easier than ever before to really find whatever you need.  The prevalence of all this during a global event validates the power of analytics to a mass audience.  It’s getting people to care about numbers and understand their value in a way that’s accessible. People are still struggling with – “I have the data… now what?” But, as the Wall Street Journal example above shows us, people are starting to get it! With everyone’s eyes on the World Cup, it’s gratifying to see us watching much more than the matches. There is little wonder why teams are increasingly acknowledging the value of data and analytics. Global soccer revenues equaled $28 billion annually, nearly as much as the combined revenues for all other sports, according to a 2011 estimate from consulting firm AT Kearney. 46 percent of the world’s population in 2010 watched at least a minute of that year’s World Cup, according to FIFA.  With so much at stake, data’s role is all but assured in the starting lineup of teams and fans around the world.","excerpt":"I have to make a confession: I’m a not a die-hard soccer maniac. However, The World Cup is still the highlight of my year every time it rolls around.  It is fascinating to witness how Soccer Mania takes over the world.  It might only happen every 4 years, and most people only use it as […]","categories":["IT Services"],"tags":["soccer analytics big data"],"author_name":"Abhishek Sinha","publish_date":"2014-06-26T18:53:50","publication_year":"2014","word_count":852,"keywords":["big data","Go","programming_languages:R","AI","data-driven","programming_languages:Go","RAG","soccer analytics big data","analytics","GAN","R"],"extracted_tech_keywords":["AI","analytics","RAG","R","Go","big data","GAN","data-driven","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/analytics-for-fifa-world-cup-2014\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10052562,"title":"Hansa Cequity Appoints Neeraj Sangani As New CEO","content":"Hansa Cequity recently announced the appointment of Neeraj Sangani as the Chief Executive Officer of the company. Sangani has been the COO since May 2020 and has been running the operations of the company for the past 16 months. He will be reporting to Group CEO Shekar Swamy and the Board of the company. The RK Swamy Hansa Group commenced Hansa Cequity (Hansa Customer Equity) as a start-up 13 years ago and has invested heavily since in the technology-driven space. The company provides services in the Data Analytics, CRM and MarTech space. Sangani is an Alumni of UCLA Anderson School of Management and a keen proponent of Behavioural Economics. He has completed his professional certifications from the Institute of Data & Marketing, London and the Rotman School of Management, Toronto, and runs CXMLab, a blog on tech, data and analytics. “It is an honour to lead this exciting company with a great portfolio of clients. I look forward to working closely with the Group and consolidating our leadership position in the market,” said Sangani. Shekar Swamy said, “Neeraj joined the Group over two decades ago. He is well versed in our activities and culture. He grew up in the world of advertising and branding and has added valuable experience in Analytics and data-driven Customer Engagement programs. He will be a great partner to our clients and our people and drive the company to the next level.” The company congratulated Sangani for this appointment on various social media platforms.","excerpt":"Sangani has been the COO since May 2020 and has been running the operations of the company for the past 16 months.","categories":["AI News"],"tags":["CEO","CRM","Data Analytics"],"author_name":"Victor Dey","publish_date":"2021-10-28T17:29:20","publication_year":"2021","word_count":248,"keywords":["Go","programming_languages:R","AI","R","ML","data-driven","programming_languages:Go","CEO","GAN","ViT","analytics","Data Analytics","CRM"],"extracted_tech_keywords":["AI","ML","analytics","R","Go","GAN","ViT","data-driven","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/hansa-cequity-appoints-neeraj-sangani-as-new-ceo\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10023067,"title":"Python Guide to HiSD: Image-to-Image translation via Hierarchical Style Disentanglement","content":"The image-to-Image translation is a field in the computer vision domain that deals with generating a modified image from the original input image based on certain conditions. The conditions can be multi-labels or multi-styles, or both. In recent successful methods, translation of the input image is performed based on the multi-labels and the generation of output image out of the translated feature map is performed based on the multi-styles. The labels and styles are fed to the models via texts or reference images. The translation sometimes takes unnecessary manipulations and alterations in identity attributes that are difficult to control in a semi-supervised setting. Chinese researchers Xinyang Li, Shengchuan Zhang, Jie Hu, Liujuan Cao, Xiaopeng Hong, Xudong Mao, Feiyue Huang, Yongjian Wu and Rongrong Ji have introduced a new approach to control the image-to-image translation process via Hierarchical Style Disentanglement (HiSD). HiSD breaks the original labels into tags and attributes. It ensures that the tags are independent of each other, and the attributes are mutually exclusive. While deploying, the model first looks for the tags and then the attributes in a sequential manner. Finally, shapes are defined by latent codes extracted from reference images. Thus improper or unwanted manipulations are avoided. The tags, the attributes and the style requirements are arranged in a crystal-clear hierarchical structure that leads to state-of-the-art disentanglement performance on many public datasets. Hierarchical representation of tags, attributes and styles in HiSD HiSD processes all the conditions (i.e., tags, attributes and styles) in an independent strategy so that they can be controlled alone or on inter-conditions or intra-conditions. The model extracts styles easily from the reference images by converting them into latent codes and Gaussian noises. It adds the style to the input image without affecting its identity or other styles, tags and attributes. The multi-style task with the HiSD without any loss in identity Multi-attribute task with the HiSD without any loss in identity Multi-tag task with the HiSD without any loss in identity Python implementation of HiSD HiSD needs a Python environment and PyTorch framework to set up and run. Usage of a GPU runtime is optional. Pre-trained HiSD can be loaded and inference may be performed with a CPU runtime itself. Install dependencies using the following command. !pip install tensorboardx The following command downloads the source codes from the official repository to the local machine. !git clone https:\/\/github.com\/imlixinyang\/HiSD.git Output: Change the directory to content\/HiSD\/ using the following command. %cd HiSD\/ Download the publicly available CelebAMask-HQ dataset from the google drive to the local machine to proceed further. Ensure that the train images are stored in the directory \/HiSD\/datasets and their corresponding labels are stored in the directory \/HiSD\/labels. The following command preprocesses the dataset for training. !python \/content\/HiSD\/preprocessors\/celeba-hq.py --img_path \/HiSD\/datasets\/ --label_path \/HiSD\/labels\/ --target_path datasets --start 3002 --end 30002 The following command trains the model and fits the model configuration to the machine and dataset. It creates a new directory under the current path named ‘outputs’ to store its outputs. !python core\/train.py --config configs\/celeba-hq.yaml --gpus 0 Once the dataset preprocessing and the model checkpoint restoration are finished, they can be used for similar applications. A sample implementation is carried out with the following simple python codes. First, import the necessary modules and libraries. %cd \/content\/HiSD\/ from core.utils import get_config from core.trainer import HiSD_Trainer import argparse import torchvision.utils as vutils import sys import torch import os from torchvision import transforms from PIL import Image import numpy as np import time import matplotlib.pyplot as plt Download the checkpoint parquet file from the official page using the following command. !wget --load-cookies \/tmp\/cookies.txt \"https:\/\/docs.google.com\/uc?export=download&confirm=$(wget --quiet --save-cookies \/tmp\/cookies.txt --keep-session-cookies --no-check-certificate 'https:\/\/docs.google.com\/uc?export=download&id=1KDrNWLejpo02fcalUOrAJOl1hGoccBKl' -O- | sed -rn 's\/.*confirm=([0-9A-Za-z_]+).*\/\\1\\n\/p')&id=1KDrNWLejpo02fcalUOrAJOl1hGoccBKl\" -O checkpoint_256_celeba-hq.pt && rm -rf \/tmp\/cookies.txt Output: Move the checkpoint parquet file to the \/HiSD directory using the following commands. %cd \/content\/ !mv checkpoint_256_celeba-hq.pt HiSD\/ Load the checkpoint and prepare the model for inference using the following codes. device = 'cpu' config = get_config('configs\/celeba-hq_256.yaml') noise_dim = config['noise_dim'] image_size = config['new_size'] checkpoint = 'checkpoint_256_celeba-hq.pt' trainer = HiSD_Trainer(config) # assumed CPU device # if GPU is available, set map_location = None state_dict = torch.load(checkpoint, map_location=torch.device('cpu')) trainer.models.gen.load_state_dict(state_dict['gen_test']) trainer.models.gen.to(device) E = trainer.models.gen.encode T = trainer.models.gen.translate G = trainer.models.gen.decode M = trainer.models.gen.map F = trainer.models.gen.extract transform = transforms.Compose([transforms.Resize(image_size), transforms.ToTensor(), transforms.Normalize((0.5, 0.5, 0.5), (0.5, 0.5, 0.5))]) Define a function to perform the image-to-image translation. def translate(input, steps): x = transform(Image.open(input).convert('RGB')).unsqueeze(0).to(device) c = E(x) c_trg = c for j in range(len(steps)): step = steps[j] if step['type'] == 'latent-guided': if step['seed'] is not None: torch.manual_seed(step['seed']) torch.cuda.manual_seed(step['seed']) z = torch.randn(1, noise_dim).to(device) s_trg = M(z, step['tag'], step['attribute']) elif step['type'] == 'reference-guided': reference = transform(Image.open(step['reference']).convert('RGB')).unsqueeze(0).to(device) s_trg = F(reference, step['tag']) c_trg = T(c_trg, s_trg, step['tag']) x_trg = G(c_trg) output = x_trg.squeeze(0).cpu().permute(1, 2, 0).add(1).mul(1\/2).clamp(0,1).detach().numpy() return output The following commands set the desired tags, the attributes and the styles to perform translation. They use in-built example images for translation. Users can opt for their own data images. First example inference: input = 'examples\/input_0.jpg' # e.g.1 change tag 'Bangs' to attribute 'with' using 3x latent-guided styles (generated by random noise). steps = [ {'type': 'latent-guided', 'tag': 0, 'attribute': 0, 'seed': None} ] plt.figure(figsize=(12,4)) for i in range(3): plt.subplot(1, 3, i+1) output = translate(input, steps) plt.imshow(output, aspect='auto') plt.show() Output: Second example inference: input = 'examples\/input_1.jpg' plt.figure(figsize=(12,4)) # e.g.2 change tag 'Glasses' to attribute 'with' using reference-guided styles (extracted from another image). steps = [ {'type': 'reference-guided', 'tag': 1, 'reference': 'examples\/reference_glasses_0.jpg'} ] output = translate(input, steps) plt.subplot(131) plt.imshow(output, aspect='auto') steps = [ {'type': 'reference-guided', 'tag': 1, 'reference': 'examples\/reference_glasses_1.jpg'} ] output = translate(input, steps) plt.subplot(132) plt.imshow(output, aspect='auto') steps = [ {'type': 'reference-guided', 'tag': 1, 'reference': 'examples\/reference_glasses_2.jpg'} ] output = translate(input, steps) plt.subplot(133) plt.imshow(output, aspect='auto') plt.show() Output: Third example inference: input = 'examples\/input_2.jpg' # e.g.3 change tag 'Glasses' and 'Bangs 'to attribute 'with', 'Hair color' to 'black' during one translation. steps = [ {'type': 'reference-guided', 'tag': 0, 'reference': 'examples\/reference_bangs_0.jpg'}, {'type': 'reference-guided', 'tag': 1, 'reference': 'examples\/reference_glasses_0.jpg'}, {'type': 'latent-guided', 'tag': 2, 'attribute': 0, 'seed': None} ] output = translate(input, steps) plt.figure(figsize=(5,5)) plt.imshow(output, aspect='auto') plt.show() Output: Performance of HiSD HiSD is trained and evaluated on the famous CelebA-HQ dataset with 30,000 facial images of celebrities with labels of tags and attributes such as hair colour, presence of glasses, bangs, beard and gender. The first 3,000 images are used as test images, and the remaining 27,000 images are used as train images. Competitive models are also trained with the same dataset under identical device configurations for enabling comparison. Qualitative comparison of HiSD with the recent state-of-the-arts, SDIT and StarGANv2 Qualitative comparison of HiSD with the recent state-of-the-art, ELEGANT HiSD outperforms the current state-of-the-art models including SDIT, ELEGANT, and StarGANv2, on the FID scale (Frechet Inception Distance), which measures realism & FID scale that measures the disentanglement. Note: Images and illustrations other than the code outputs are taken from the original research paper and the official repository. Further reading Original research paperSource code repositorySample implementation notebook","excerpt":"HiSD controls the image-to-image translation process via Hierarchical Style Disentanglement based on tags, attributes and styles.","categories":["Deep Tech"],"tags":["Guide"],"author_name":"Rajkumar Lakshmanamoorthy","publish_date":"2021-03-30T17:00:00","publication_year":"2021","word_count":1137,"keywords":["CUDA","NumPy","TPU","PyTorch","AI","ML","computer vision","Python","Matplotlib","R","Guide"],"extracted_tech_keywords":["AI","ML","computer vision","PyTorch","NumPy","Matplotlib","TPU","CUDA","Python","R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/hisd-python-implementation-of-image-to-image-translation\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":16866,"title":"Wipro’s Industry Focused Big Data Analytics-as-a-Service Platform Now Available On Microsoft Azure","content":"Wipro, the Software giant headquartered in Bangalore, announced the availability of Data Discovery Platform which is a big data analytics-as-a-service solution on Microsoft Azure.  Available for various industries such as banking, financial services, retail, energy, education, manufacturing are others, the Data Discovery platform would enable the businesses to bridge the gap between the insights required by business and the information that is available. The solution would accelerate insight-driven decision making through pre-built applications for these specific industries. With Wipro’s Data Discovery Platform, the businesses can embark on an analytics journey with value added services of process simplification and business transformation. This collaboration by Wipro and Microsoft would also address various other aspects of the solution such as engineering, solution enhancements, and joint go-to market strategy. Wipro, in its company blog post mentioned that the platform leverages Microsoft’s Cortana Intelligence Suite which includes HDInsight, Stream Analytics, Data Lake Analytics, Machine Learning and Power BI to build analytical applications. “Currently, fourteen Wipro Data Discovery Platform applications have been showcased on Microsoft Advanced Analytics Partner Solution Showcase”, it said. Pallab Deb, Vice President & Global Head – Analytics, Wipro Limited said “Together, Microsoft and Wipro have built an industry sector-specific apps ecosystem on the Data Discovery Platform. Today, the platform is a significant enabler of Analytics led Digital Transformation delivering Analytics-as-a-Service to organizations.” “We believe that this is a reflection of the Wipro Data Discovery Platform’s maturity and Microsoft’s confidence in the prowess of this platform,” he further shared. “Wipro is well-positioned to build effective systems of intelligence for clients with the power of Cortana Intelligence, Power BI and Azure,” said Joseph Sirosh, Corporate VP, Cloud AI Platform, Microsoft. With Wipro’s Data Discovery Platform, that is a leading example of leveraging the data and artificial intelligence capabilities, Microsoft Azure has built novel domain-specific business applications across industry verticals. “By developing pre-built domain specific apps on the platform, Wipro is able to accelerate time-to-value while ensuring scalability and performance”, Sirosh added. Wipro has been harnessing the power of cognitive computing, hyper-automation, robotics, cloud, analytics and emerging technologies for long to help clients help adapt to the digital world and make them successful. The two tech giants have been strategic partners for decades, with Wipro demonstrating its commitment to this relationship by investing in new industry-leading solutions for customers on the Azure platform. A word on the Data Discovery platform by Wipro- As Wipro believes, as the businesses are moving towards accurate and accelerated decision making, ‘Insights-as-a-Service’  has become a priority for virtually every type of industry. Analytics is way out here as it allows business to be more productive and agile in delivering value by supporting smarter decisions. However, organizations may face challenges around data acquisition and usage, time required for delivering insights, high infrastructure cost, among others during analytics adoption. This is where Wipro’s Data Discovery Platform (DDP) comes to rescue. With its industry specific apps, DDP covers the entire spectrum from data to information to insights. It not only empower customers with relevant insights but allows businesses to gain valuable insights and bridge the gap between insights and information. It leverages techniques such as visual sciences and storytelling with data.","excerpt":"Wipro, the Software giant headquartered in Bangalore, announced the availability of Data Discovery Platform which is a big data analytics-as-a-service solution on Microsoft Azure.  Available for various industries such as banking, financial services, retail, energy, education, manufacturing are others, the Data Discovery platform would enable the businesses to bridge the gap between the insights required […]","categories":["AI News"],"tags":["Big Data Analytics","big data platform c++","Microsoft Azure"],"author_name":"Srishti Deoras","publish_date":"2017-08-08T11:50:47","publication_year":"2017","word_count":528,"keywords":["Go","artificial intelligence","machine learning","AI","R","Scala","Git","RAG","Big Data Analytics","big data platform c++","analytics","Microsoft Azure","Azure"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","analytics","RAG","Azure","R","Go","Scala","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/wipros-industry-focused-big-data-analytics-service-platform-now-available-microsoft-azure\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":20795,"title":"Indian Economy To Reach $5 Trillion By 2025, AI And IoT Will Be Major Contributors, Says NITI Aayog Chief","content":"NITI Aayog chief Amitabh Kant at the IDS event. Union Commerce Minister Suresh Prabhu on Wednesday reaffirmed the faith in India’s tech sector by saying that the country was on its way to become a $5 trillion economy in the next eight to nine years. He added that the technology sector — especially telecommunication, artificial intelligence (AI) and internet of things (IoT) — would contribute substantially in achieving the seemingly overwhelming target. “We are preparing a detailed plan towards achieving that. Experts are busy preparing the roadmap including all the sectors where manufacturing can be promoted. If manufacturing is digitised, it will create a huge opportunity for technology firms,” Prabhu said at the India Digital Summit organised by the Internet and Mobile Association of India (IAMAI) in Delhi. National Institution for Transforming India (NITI Aayog) chief executive Amitabh Kant also emphasised on the substantial role IoT and AI in the Indian economy. Kant told the audience that AI alone would generate opportunity to the tune of $32 billion. “Advanced robotics are already handling 25 percent of the jobs. This will rise to 45 percent in the coming years… India is already ranked globally as the most active Internet user globally on a monthly basis. Digital transaction will touch $100 trillion in the next 10 years,” Kant said at the India Digital Summit. “With over 400 million smartphone users in the country, there is a huge opportunity in the IoT space. The opportunity would be in the range of $70 billion by 2025,” he added. Kant also told a news wire that data analytics would boost financial inclusion in the country and would slowly render physical banks useless. “With 99 percent Aadhaar penetration and bank accounts opened under Jan Dhan Yojana, the scope was immense for the fintech industry as well,” he added.","excerpt":"Union Commerce Minister Suresh Prabhu on Wednesday reaffirmed the faith in India’s tech sector by saying that the country was on its way to become a $5 trillion economy in the next eight to nine years. He added that the technology sector — especially telecommunication, artificial intelligence (AI) and internet of things (IoT) — would […]","categories":["AI News"],"tags":["AI (Artificial Intelligence)","amitabh kant","FinTech","Internet of things","NITI Aayog","suresh prabhu"],"author_name":"Prajakta Hebbar","publish_date":"2018-01-19T07:47:50","publication_year":"2018","word_count":302,"keywords":["amitabh kant","artificial intelligence","programming_languages:R","AI","Git","Internet of things","analytics","ai_applications:robotics","suresh prabhu","GAN","FinTech","AI (Artificial Intelligence)","R","NITI Aayog"],"extracted_tech_keywords":["AI","artificial intelligence","analytics","R","Git","GAN","programming_languages:R","ai_applications:robotics"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/indian-economy-5-trillion-2025-ai-iot-tech-niti-aayog-amitabh-kant\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10096192,"title":"GPT-4 Completes Duolingo&#8217;s Language Learning Nest","content":"Languages hold the power to transcend borders and cultures. A remarkable example of this lies in the story of Amanda from Manila and Rob from Washington DC who used a blend of English, Tagalog, and their own creative language to nurture their relationship. But what brought the couple closer together? Duolingo — the Pittsburg-based language learning app enabled them to communicate, shattering the barriers of distance and language for the couple. To make this possible, Duolingo has been leveraging cutting-edge AI. It was one of the beta users of OpenAI’s GPT-4 before it was rolled out to the public, introducing two innovative AI-backed features called ‘Role Play‘ and ‘Explain My Answer‘. These new features are part of Duolingo Max, a new subscription tier aimed at enhancing the language learning experience. “We have been developing the AI secret sauce for years to make education more effective and engaging. In September 2022, we collaborated with OpenAI to build brand new learning features using GPT-4,” Klinton Bicknell, head of AI at Duolingo, told AIM in an exclusive interaction. Bicknell said that with the integration of GPT-4, Duolingo aims to provide learners with the ability to hold natural conversations on various topics in specific contexts, as well as receive detailed explanations for their answers. While these features are currently available in Spanish and French, Duolingo plans to expand the availability to more languages and introduce additional AI-driven features in the future. How is GPT-4 Changing Duolingo ‘Explain My Answer’ uses GPT-4 to generate detailed explanations for learners, helping them understand why their answers were correct or incorrect. This feedback enables learners to identify their mistakes and improve their skills more rapidly. On the other hand, the ‘Roleplay’ feature offers learners an AI-powered conversation partner, allowing them to practice their conversation skills in a safe and supportive environment. Previously, Duolingo engineers tested GPT-3 as a supplement to human-powered features in their chat function. Although it was close to being ready, they didn’t feel confident enough to integrate it for handling complex automated aspects of chats. Earlier, Duolingo used scripted conversations for chatting with learners, but they desired the ability to engage learners in specific and immersive discussions. With GPT-4, which has been trained on extensive public data, Duolingo aims to provide learners with a flexible and dynamic conversation experience. These new features can be accessed through their new subscription model called Duolingo Max, which offers an enhanced and immersive learning experience compared to the Super Duolingo version. While Super Duolingo provides benefits such as an ad-free environment, unlimited hearts, and unlimited attempts at Legendary challenges, Duolingo Max encompasses all these features while introducing the transformative ‘Roleplay’ and ‘Explain My Answer’ features. “As a powerful language model developed by OpenAI, GPT-4 facilitates personalised feedback generation, conversation generation, and interactive learning experiences for Duolingo users,” said Bicknell. Duolingo has also developed a custom ML model called ‘Birdbrain’ that analyses learners’ knowledge levels and predicts the difficulty of language materials, complementing the existing personalisation system. “Birdbrain takes the understanding of learners even further by extending its knowledge to all aspects of the language learning app,” said Bicknell. The personalised learning engine suggests suitable lessons based on individual learners’ progress and weaknesses. Additionally, AI is employed to provide specific feedback and create a gamified experience that keeps learners motivated and engaged. Breaking Barriers: AI Fuels Education Access The integration of AI into language learning, as demonstrated by Duolingo’s collaboration with OpenAI, holds great significance. AI allows for a more engaging and effective learning experience by replicating some of the qualities of a human tutor and scaling that experience to reach learners who may not have access to quality teachers or tutors. “Our vision is to use AI to provide users with an experience that closely resembles learning with a real human tutor, making language education more accessible and impactful for people around the world,” concluded Bicknell. Besides Duolingo, the non-profit, Khan Academy, too is using GPT-4 to power Khanmigo, an AI-powered assistant that functions as both a virtual tutor for students and a classroom assistant for teachers. Back home, Indian edtech companies are currently betting big on LLMs to offer highly personalised learning experiences to students. upGrad, for example, is considering the creation of its own exclusive LLM. Additionally, upGrad has developed a GPT-based chatbot for students to practice mock interviews at their convenience. They are also working on building a coaching bot. Likewise, other edtech firms such as EdZola and Byju’s are leveraging generative AI to enhance their services. Byju has introduced Wiz, a collection of AI models tailored for a personalised learning suite.","excerpt":"Duolingo’s Head of AI, Klinton Bicknell, shares with us how the company enhances language education worldwide by utilising generative AI to create a learning experience similar to a human tutor.","categories":["AI Features"],"tags":["Duolingo","GPT-4","Interviews and Discussions","OpenAI"],"author_name":"Shritama Saha","publish_date":"2023-07-04T10:00:00","publication_year":"2023","word_count":766,"keywords":["Go","API","OpenAI","AI","ML","GPT-4","RAG","Duolingo","Aim","generative AI","edge AI","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","ML","generative AI","OpenAI","Aim","edge AI","RAG","R","Go","API"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/gpt-4-completes-duolingos-language-learning-nest\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10067029,"title":"Major embarrassment for Google over Arabic slip up during Pichai’s Google Translate presentation","content":"At the recently concluded Google I\/O conference, a slide in the backdrop displayed a text backwards in Arabic while CEO Sundar Pichai announced the addition of 24 new languages to Google Translate. Soon after the flub, social media was flooded with memes and comments directed at the tech giant. “Congrats to Google for getting Arabic script backwards & disconnected during Sundar Pichai’s presentation on ‘Google Translate’, because small independent startups like Google can’t afford to hire anyone with a 4-year-old elementary school level knowledge of Arabic writing,” tweeted Video game developer Rami Ismail. Soon after, Google’s PR machinery went into damage control mode. Joyce Baz, head of PR & communications at Google, Middle East and North Africa, responded to Elie Habib’s Linkedin Post: Hi Elie, the mistake that took place at I\/O was a presentation software error and not related to Google Translate or any other Google product. We apologized and immediately corrected it. Around the world, Google invests heavily in developing products from Search, to Assistant, Maps and more to be helpful to people everywhere, including Arabic speakers. Of course, there’s always more that can be done. If you’d like to know more about Google’s work in MENA, I’d be happy to invite you over to our office here in Dubai for you to meet the local team, some of whom you may know already,  who can tell you more about Google’s many products, tools and programs in Arabic. And Arabic was not the only language Google got wrong. AI ethicist and researcher Timnit Gebru, who was fired from Google, quoted Ismail’s tweet and said: “Now imagine how people in most countries are served when it comes to YouTube, misinformation etc, and what happens if you flag some policy-violating content in any of these languages. “At the bottom of the pile because these people and languages are not important enough to have the necessary resources. And the necessary resources will only be obtained through regulation because these companies will not voluntarily make a lot less money to make safe platforms,” she added. Now imagine how people in most countries are served when it comes to YouTube, misinformation etc, and what happens if you flag some policy violating content in any of these languages. https:\/\/t.co\/iKseE21eeP— @timnitGebru@dair-community.social on Mastodon (@timnitGebru) May 13, 2022 A Google employee with the Twitter handle, theREALmarvin.eth, took to Twitter to admit the mistake. He tweeted, “This was a mistake across different languages, and we won’t make excuses. We feel a big sense of responsibility to make sure everyone is accurately and authentically represented.” i work I\/O. this was a mistake across different languages, and we won’t make excuses. we feel a big sense of responsibility to make sure everyone is accurately & authentically represented. i’m sorry we didn’t hit the mark – we've corrected the video & are updating our processes.— theREALmarvin.eth (@theREALmarvin) May 12, 2022 He added that Google has corrected the video.","excerpt":"Google has corrected the video.","categories":["AI News"],"tags":[],"author_name":"Pritam Bordoloi","publish_date":"2022-05-14T17:10:25","publication_year":"2022","word_count":488,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","ViT","R","startup"],"extracted_tech_keywords":["AI","R","Go","ViT","startup","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/major-embarrassment-for-google-over-arabic-slip-up-during-pichais-google-translate-presentation\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10081550,"title":"Yann LeCun Cherry-picks Reinforcement Learning","content":"The self-supervised learning guru and chief AI scientist at Meta AI, Yann LeCun, introduced the ‘cake analogy’, at NIPS 2016. “If intelligence is a cake, the bulk of the cake is unsupervised learning, the icing on the cake is supervised learning, and the cherry on the cake is reinforcement learning.” However, while delivering a talk about SSL at NeurIPS 2022 in the context of achieving AGI, LeCun suggested abandoning the four most popular things at the momentum machine learning, including generative models, probabilistic models, contrastive methods, and reinforcement learning, Mofijul Islam pointed out on Twitter. To this, LeCun replied saying that he is not entirely unfamiliar with making statements that go against the common wisdom of the day. Meanwhile, several Twitter users and AI experts also joined in to express their views about it. Kyle Cranmer, physicist and professor at NYU, said that though he is a fan of generative models and probabilistics methods, he agrees with LeCun about the need of world models for reinforcement learning. Other people too proposed the idea of having a model for reinforcement learning and that a lot of research is now focused on just creating generative models. Cranmer added that LeCun makes statements that he thinks do not need explanation and though he agrees with him largely, he is willing to discuss the things that he is at odds with. LeCun agrees that just by looking at one slide from the presentation, people are making assumptions and drawing conclusions. I’ve seen this slide and the surrounding twitter discussions \/ debates. I’ve had enough conversations with @ylecun that I think I know where he is coming from and what he means, and I also understand the reactions based on what is written. I feel compelled to help bridge the gap https:\/\/t.co\/HekE4zUg9c— Kyle Cranmer (@KyleCranmer) December 4, 2022 November shines bright for generative AI November 2022 was a great month for AI. Apart from NeurIPS, OpenAI released ChatGPT, an excellent chatbot that is touted to be a ‘Google Killer’. Stability AI announced the release of Stable Diffusion 2.0. Mind-vis, a mental image reading algorithm was released. Meta AI also released two new models – CICERO; the first model to achieve human level competence in ‘Diplomacy, and Galactica; trained on 120 billion parameters, focusing on scientific papers to aid academic research. According to the research paper of Galactica, the model outperformed GPT-3 in technical knowledge probes of LaTeX equations. Many researchers and AI enthusiasts were excited to try it out for its generative capabilities. But it did not take long for the community to figure out that a lot of predictions and outcomes that were produced were inaccurate and hallucinating. This eventually led to Meta AI pulling the plug on the demo. Galactica demo is off line for now.It's no longer possible to have some fun by casually misusing it.Happy? https:\/\/t.co\/K56r2LpvFD— Yann LeCun (@ylecun) November 17, 2022 CICERO, on the other hand, the paper clearly states, integrates a language model using reinforcement learning algorithms and using human intervention and conversations to generate diplomacy. Maybe from the fall of this model, LeCun inferred that generative and predictive models might not achieve the goal that he had in mind – AGI. Read: Meet the Meta AI Researcher Who Helped Build CICERO OpenAI’s ChatGPT has been gaining significant popularity since its release last week with people experimenting with it and touting it to be a glimpse of GPT-4. This clearly highlights the importance of generative models. The rise of image generation models like Stable Diffusion or DALL-E also add points to the same. Russ Salakhutdinov, UPMC professor of CS at Carnegie Mellon University and former director of AI research at Apple, praised the Galactica paper but also pointed out that LeCun was earlier promoting the generative and probabilistic methods and capabilities of Galactica, but now after its downfall, says that these methods should be replaced with joint-embedding architectures. To LeCun’s credit, he replied saying that during the talk at the conference he explained that people interested in applications of generative and predictive models can clearly use them, but his recommendations are for researchers pursuing the path towards imparting common sense and reasoning capabilities in AI.  He recommends VICreg, a Meta AI developed algorithm for variance-invariance-covariance regularisation for SSL, to replace contrastive methods. As I said in the talk– if you are interested in applications of generative models in the short term, by all means use LLMs, diffusion models, etc.– if you are interested in making AI advance towards common sense & planning\/reasoning capabilities, follow these recommendations.— Yann LeCun (@ylecun) December 4, 2022 Yann LeCun has been on the headline several times, not just for amazing innovations, but also for various controversies. In May, the very popular debate about ‘AI hitting the wall’ surfaced when Gary Marcus tweeted a video of Tesla hitting an aeroplane. Similarly in July, LeCun’s paper – A Path Towards Autonomous Machine Intelligence — was facing controversy when Jurgen Schmidhuber claimed that LeCun’s ‘original contributions’ actually felt to him like deja vu of his work.","excerpt":"LeCun clearly is at odds with reinforcement learning and believes that for AI with common sense, it is not the way forward","categories":["AI Features"],"tags":["Artificial General Intelligence","Autonomous Machine Intelligence","ChatGPT","Generative AI","Machine Learning","Meta AI","Reinforcement Learning","self supervised learning","Yann LeCun"],"author_name":"Mohit Pandey","publish_date":"2022-12-06T10:00:00","publication_year":"2022","word_count":839,"keywords":["Go","ChatGPT","Meta AI","Yann LeCun","Reinforcement Learning","machine learning","AI","OpenAI","ELT","Autonomous Machine Intelligence","Machine Learning","Aim","Artificial General Intelligence","generative AI","Generative AI","R","self supervised learning"],"extracted_tech_keywords":["AI","machine learning","generative AI","ChatGPT","OpenAI","Meta AI","Aim","R","Go","ELT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/yann-lecun-cherry-picks-reinforcement-learning\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":21085,"title":"Data Privacy Day 2018: It&#8217;s Time To Overcome The Data Privacy Concerns","content":"Data, Data everywhere, but no data is completely breech free! As the companies are turning to data companies today, there’s a lot to achieve in the ethical and innovative use of data. Data Privacy Day, which was observed on 28th Jan like every year, aims to do that. The day which began as an extension of the Data Protection Day, Data Privacy Day has been celebrated since 2008, and is the first legally binding international treaty dealing with privacy and data protection. The purpose is simple—to raise awareness and promote privacy and data protection best practices. Currently observed in the United States, Canada, India and 47 European countries, it’s educational initiative was focused on raising awareness about the importance of protecting the privacy of their personal information online. The need for Privacy protection in India India has currently been in news for a security breach that left billions of people at a risk of identity theft. Not long ago, there were reports from a leading Indian Newspaper that they were able to access names, email addresses, phone numbers and postal codes by typing in 12-digit unique identification numbers of people in the government’s database, after reportedly paying an individual about INR 500. It also reported that you can get an access to software to print out unique identification cards, the Aadhaar cards for some more money. Though it has been constantly assured that Aadhaar data is fully safe and secure, these reports did stir a controversy amongst the netizens that rely on government for the safety of their data. Striking numbers suggesting the relevance of data privacy concerns Not just in India, data privacy concerns are a top priority amongst most internet users for identity theft, family safety and home network security. With the world connecting to internet more than ever before, personal information is at an easy access for the notorious minds. As the world is getting connected, security concerns are increasing many fold. According to a report by McAfee that surveyed 6,400 people globally for its study, it revealed striking numbers in terms of consumer security focuses and concerns. As per the study, 52% were unsure of how to secure connected devices and apps, almost one-third of parents do not monitor their child’s connected device usage, more than 30% are unsure about how they can control companies from collecting personal information, only 37% use credit monitoring services, amongst others. Similarly, according a report released by Parliament federal privacy commissioner Daniel Therrien, Canadian public is worried about data privacy too. According to the report, 92% of respondents expressed concern about the protection of their privacy and a clear majority (57%) were very concerned. The way ahead These numbers clearly suggest threats that hover around data privacy. As data security is growing to be one of the major concerning areas, Data Privacy Day comes as an excellent initiative that increases awareness and recognition around it. Experts claim that it is always advisable to be aware of your privacies all around the year. It is important that corporates offer meaningful consent online and strictly oppose inappropriate data practices. Other measures could be having policies and processes in place to ensure strict safeguarding of personal information. It is also important to have breech management that can identify steps to manage and remediate a breach, a strong IT security infrastructure with appropriate mechanisms that can ensure safe storage and transfer of personal data. Data security is one of the major concerning areas, and on this Data security day lets get closer to achieving a data secure place, where no citizen has a fear of their privacy not being respected, and the task of protecting personal data is taken seriously to avoid any concerns in the future.","excerpt":"Data, Data everywhere, but no data is completely breech free! As the companies are turning to data companies today, there’s a lot to achieve in the ethical and innovative use of data. Data Privacy Day, which was observed on 28th Jan like every year, aims to do that. The day which began as an extension […]","categories":["IT Services"],"tags":["data breach india","data privacy India","data protection india"],"author_name":"Srishti Deoras","publish_date":"2018-01-30T06:51:19","publication_year":"2018","word_count":623,"keywords":["Go","programming_languages:R","AI","data protection india","programming_languages:Go","Git","data privacy India","RAG","Aim","data breach india","GAN","R"],"extracted_tech_keywords":["AI","Aim","RAG","R","Go","Git","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/data-privacy-day-2018-time-overcome-data-privacy-concerns\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10010350,"title":"How To Create An Image Dataset and Labelling By Web Scraping?","content":"While working on a data science project, the first step is acquiring the data. For this purpose, we traverse through several websites where certain datasets are available in a structured manner and we can download and have it ready to use. Even if we have a dataset, it might not have enough data and we know our ML models want a good amount of data to be trained well. In case of classification problems, we need this data along with labels. But this is not always the case, often for a specific problem statement dataset might not be readily available. Suppose we want to build a face mask classifier and maybe after several web searches we don’t get the desired dataset. In such situations, we need to make our dataset. In computer vision problems, very less is said about acquiring images and more about working with images. Thus I’ll be going through this crucial step of making a custom dataset and also labelling it. In this article, I’ll be discussing how to create an image dataset as well as label it using python. For creating an image dataset, we need to acquire images by web scraping or better to say image scraping and then label using Labeling software to generate annotations. Web Scraping Web scraping means extracting data from websites, wherein a large amount of data after extraction is stored in a local system. Web scraping may access the world wide web through https and a web browser. The most well-known image scraping python library is beautifulsoup that parses HTML and XML documents. The requests library makes the necessary requests to the webpage. Both the packages are pip installable(and maybe already preinstalled). import BeautifulSoup import requests as rq import os r2 = rq.get(\"https:\/\/www.pexels.com\/search\/koala\") soup = BeautifulSoup(r2.text,\"html.parser\") links = [] If we click onto any picture on the webpage and go to developer tools we’ll see the specified format starts with ‘images.pexels.com\/photos’, up to photos the format is the same and then a unique number is present, thus we specify that so similar images can be acquired. This is a form of regex(regular expressions). images = soup.select('img[src^=\"https:\/\/images.pexels.com\/photos\"]') for img in images: links.append(img['src']) After this step if we wish we can print the ‘links’ list to see the image links that have been scrapped. Making our directory to save images in it os.mkdir('jayita_photos') Now we download images and only 10 images to show the working. The entire page can also be done. This is done with the usual file handling technique. i=1 for index,img_link in enumerate(links): if i<=10: img_data = rq.get(img_link).content with open(\"jayita_photos\/\/\"+str(index+1)+'.jpg', 'wb+') as f: f.write(img_data) i+=1 else: f.close() break After successfully running the program go to the specified file path and you can see that the images are stored. Labelling Now that we have our images we need to label them for classification. For this, we’ll be using the labelling software. Labelling is a GUI based annotation tool. Works with Python 3 and above. It’s a pip installable. Provides two types of annotations Pascal VOC(this is used by ImageNet) and YOLO. Labelling software opens up with the above command. On the left side there are specified options and on the right side image file information will be shown. For a single image select open for a directory of images select ‘open dir’ this will load all the images. To go to the previous image press ‘a’, for next image press ‘d’. Drawing the rectangular box to get the annotations. Press ‘w’ to directly get it. After drawing this window will pop up which means to store the class name for that particular image. Providing class labels (koala in this case) on the right side of the window it shows. After drawing the bounding box and labelling the precise class name its important save along with format(Pascal VOC or YOLO) that will generate the annotations. This file is stored in an XML format. Format for Pascal VOC form of annotations Format for YOLO The first one 0 represents object id, then rest 4 are bounding box coordinates The class file containing the class names generated along with YOLO format Conclusion Creating own image datasets with these steps can be helpful in situations where the dataset is not readily available or less amount of data is available then to increase size this can be used. I’ve only shown it for a single class but this can be applied to multiple classes also, provided all the classes are placed in the same folder.","excerpt":"In this article, I’ll be discussing how to create an image dataset as well as label it using python. For creating an image dataset, we need to acquire images by web scraping or better to say image scraping and then label using Labeling software to generate annotations.","categories":["Deep Tech"],"tags":["data labelling","Data Science","Machine Learning","Web Scraping With Python"],"author_name":"Jayita Bhattacharyya","publish_date":"2020-10-22T14:00:44","publication_year":"2020","word_count":750,"keywords":["data science","Go","API","AI","R","ML","Machine Learning","computer vision","data labelling","Python","YOLO","programming_languages:Python","Data Science","Web Scraping With Python"],"extracted_tech_keywords":["AI","ML","computer vision","data science","Python","R","Go","API","YOLO","programming_languages:Python"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/how-to-create-an-image-dataset-and-labelling-by-web-scraping\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":46372,"title":"Fiddler Labs Raises $10.2 Million For Explainable AI","content":"Earlier this week noted explainable AI engine startup Fiddler Labs announced that they have raised $10.2 million in Series A funding to accelerate their work in building breakthrough artificial intelligence engine with explainability at its heart. Lightspeed Venture Partners and Lux Capital led this round, with participation from Haystack Ventures and Bloomberg Beta. Founded in October 2018 by CEO Krishna Gade and CPO Amit Paka, Fiddler addresses the rising risks and costly implications of not having visibility and insight into how AI works. Explainability, says Fiddler Labs, is the missing link and the key reason why many businesses find the deployment of AI cumbersome and risky. The Fiddler Explainable AI Engine gets ahead of this by enabling companies to better understand why their machine learning and AI models make the decisions they do. Fiddler is the first startup to bring true explainability to the enterprise – the art of understanding, analyzing, validating, and monitoring AI solutions at the scale required in the modern business environment. Gade said in a statement, “Businesses understand the value AI provides, but they struggle with the ‘why’ and the ‘how’ when using traditional black-box AI… Businesses need to de-risk their AI investments, but most applications today aren’t equipped to help them do that. Our AI Engine is grounded in explainability, so those affected by the technology can understand why decisions were made and course-correct as needed to ensure the AI outputs are ethical, responsible and fair.” Given the rising importance of explainability in the AI lifecycle, the company has raised a total of $13.2 million, including a $3 million seed round raised in October 2018 led by Lightspeed with participation from Bloomberg Beta and Haystack. The company is working with several businesses in Fintech, AI model risk management, and trust and compliance teams on active proof of concepts and expect to announce the general availability of the Fiddler Explainable AI Engine early next year.","excerpt":"Earlier this week noted explainable AI engine startup Fiddler Labs announced that they have raised $10.2 million in Series A funding to accelerate their work in building breakthrough artificial intelligence engine with explainability at its heart. Lightspeed Venture Partners and Lux Capital led this round, with participation from Haystack Ventures and Bloomberg Beta. Founded in […]","categories":["AI News"],"tags":["ai dl","Explainable AI"],"author_name":"Prajakta Hebbar","publish_date":"2019-09-25T15:05:11","publication_year":"2019","word_count":319,"keywords":["API","funding","artificial intelligence","machine learning","TPU","AI","explainable AI","Rust","Explainable AI","ai dl","R","startup"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","TPU","R","Rust","API","explainable AI","startup","funding"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/fiddler-labs-raises-10-2-million-for-explainable-ai\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10143129,"title":"Teachers Scramble as Students Embrace Bots over Books","content":"One of the year’s biggest shakeups involved AI eating up an industry giant. Education services website Chegg, once a student’s best friend, is struggling to stay afloat, with its stock price plummeting a staggering 99% by November 2024. The company also laid off 441 employees in the summer and 300 more in November. Interestingly, a few employees had suggested the leadership integrate AI features into the platform as early as 2022, but as per reports, the request was denied. While AI seals the fate of traditional edtech startups, newer startups with a different view on edtech are dancing on their graves. Moreover, their customers, the students, seem to be having the time of their lives. More AI, More GPA Recently, Kent State University published a study surveying 71 students about their use of AI in academia. While the sample size was small, it does give a fair bit of insight into the impact of AI in the classroom. Forty-seven per cent of the group comprised upper-level undergraduates aged 21-23, 26% were sophomores, and the rest were older graduate students. About 48% of the students reported a ‘significant improvement’ in their academic performance, and 35% of them reported a slight improvement. Most of these students revealed using AI tutoring systems, study scheduling apps and language learning apps using AI. The majority used AI tools for roughly a quarter to half of their study time, and these tools motivated them to study. While the sample size consisted of 46.5% students in computer science, the majority of the respondents were students from other fields. AIM also spoke to two students, one from Cornell University and another from Carnegie Mellon University. “Since I’m involved in a lot of research, I use ChatGPT and Jeni to thoroughly understand its details,” the Carnegie Mellon student of biotech, said. “I’ve also used AlphaFold2 for protein structure predictions, and Codeium AI for auto-completion,” she added. The Cornell undergraduate studying physics said that while he uses AI tools to augment his learning, he’s also experienced some of the lesser-known use cases. “I also take a linguistics class, and we discuss how these LLMs, and neural networks learn a language as compared to humans, and how they learnt it in the early age”. In another philosophy course he takes, he said that his professor has explored a new learning technique in which he prompts philosophical questions to ChatGPT, and students are asked to critique the responses. Earlier this year, another study from Harvard University, although in a pre-print stage, measured how college students learn using an AI-powered tutor. The study found that “students learn more than twice as much in less time when using an AI tutor, compared with the active learning class.” The study involved 194 students focusing on fluid mechanics concepts. Some of them participated in self-paced AI sessions, and the rest attended a 60-minute lecture. It was found that students using the AI tutor scored a median of 4.5\/5 points in a post-learning test, compared to 3.5\/5 from a lecture session. “On average, students in the AI group felt significantly more engaged and more motivated during the AI class session than the students in the active lecture group. The degree to which both groups enjoyed the lesson and reported a growth mindset was comparable,” added the report. Recently, Greg Brockman, president of OpenAI, demonstrated ChatGPT’s capabilities with real-time voice and vision inside a classroom. He pointed ChatGPT to the blackboard to assist with drawing the human body and calculating the area of a triangle. OpenAI President Greg Brockman shows ChatGPT with real-time Voice and Vision capabilities, which will be available to paid subscribers soon pic.twitter.com\/oJbercatRP— Tsarathustra (@tsarnick) December 9, 2024 Students Want More of It Despite the overall positive feedback, students also hinted at improvements in the overall experience using AI. “Several students suggested that AI should adapt the learning path in real-time based on student performance and feedback. This would enable a personalised learning experience,” the research from Kent State University revealed. Moreover, students also revealed that digital versions of textbooks should be able to integrate AI tools within them and ask questions based on their contents. The authors also said that students wished for “AI systems that adapt test difficulty based on student performance”. This was highly recommended to improve the learning experience. Progress towards designing AI-native education tools is already up and running, with Andrej Karpathy at the forefront. After exiting OpenAI, he began working on Eureka Labs, a “new kind of school that is AI native”. “This Teacher + AI symbiosis could run an entire curriculum of courses on a common platform. If we are successful, it will be easy for anyone to learn anything, expanding education in both reach, and extent,” said Eureka Labs in their announcement. Their first project is said to be a course that teaches students to develop an AI Storyteller, which is built on Python, C and CUDA. Sven Schutt, CEO of IUgroup, the European university group, in an interview with McKinsey & Company, revealed their plans to integrate AI into education. He said that they have implemented an advanced AI tutor called Syntea, “that delivers personalised education” for a large group of students. “This year, most likely more than 3,000 students will start an AI study program with us,” said Schutt. “Currently, we assume that people will be able to study at least 50 percent faster, if not twice as fast, with AI,” he added. Schutt also revealed that they have several mini bots integrated into their AI learning platforms to deliver a personalised experience. That said, Schutt mentioned that 20-30% of students have yet to use it. Interestingly, he also revealed a new school called IU Copilot School has been set up in collaboration with Microsoft. “Students know about this before they sign up, so we don’t exclude those who don’t want to use the technology,” he said. The Gold Doesn’t Glitter for the Teachers That said, there are several concerns about using AI in the classroom. The use of AI-generated content in assignments and tests is a well-known problem, but a study adds to the concerns. Earlier this year, a research injected 100% AI written submissions into five undergraduate modules for a BSc degree at a “reputable UK university”. It was found that a whopping 94% of the AI submissions were undetected. Moreover, the grades awarded to AI submissions were on “average half a grade boundary higher than that achieved by real students”. “Across modules, there was an 83.4% chance that the AI submissions on a module would outperform a random selection of the same number of real student submissions,” read the research. That said, there are several AI content detection tools available on the internet, but in short, they’re not worth your time. Some of them fail to detect AI-generated content, and some of them incorrectly flag human-written content as otherwise. For example, it was recently observed that AI content detectors flagged The Bible as AI-written content. Moreover, the latest updates to GPT 4o substantially enhanced its creative writing capabilities, and it will indeed be harder if content written by AI, strives to not look like content written by AI. While several universities have restricted the use of these AI tools, some say otherwise. Arijit Raychowdhury, a professor at Georgia Tech, said, “The Georgia Institute of Technology has taken the opposite approach, welcoming the use of AI in study, essays, and other assignments—but with some guardrails.” “It’s up to educators to set students on the right path with AI. That cannot be accomplished by fearing it,” he added. Deploying AI tools and resources in the classroom also presents several challenges. Dan Meyer, VP of user growth at Amplify, an edtech startup, outlined the challenges of deploying an AI teacher in the classroom. He surveyed educators on X and LinkedIn and asked them to respond to a questionnaire. The respondents believed that AI resources were only 40% ready to be deployed in a classroom. The survey was conducted to assess the readiness of these AI resources to aid in assessments, worksheets, lesson plans, and presentations. “This is a ‘telling lesson’ not a ‘teaching lesson’,” said a teacher. “While most surveys about generative AI still ask teachers, ‘Have you ever used generative AI ever?’ this survey has tried to understand how well generative AI is meeting those clear needs of teachers, students, and the public,” said Meyer. “The results indicate that generative AI is still leaving much of that need unmet,” he concluded. While people have widely underscored AI’s abilities in the education sector, it is also important to assess whether students and teachers are comfortable using these tools. Moreover, if AI can transform education in low-resourced schools, especially in developing countries like India, it is imperative to design these products while keeping accessibility and ease of use in mind. Both teachers and students must be given adequate training to make the most of these tools.","excerpt":"Knowledge is just a prompt away. But is it good news for teachers?","categories":["AI Features"],"tags":["AI (Artificial Intelligence)","education"],"author_name":"Supreeth Koundinya","publish_date":"2024-12-10T12:00:58","publication_year":"2024","word_count":1490,"keywords":["CUDA","ChatGPT","OpenAI","AI","neural network","education","RAG","Python","Ray","Aim","generative AI","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","neural network","generative AI","ChatGPT","OpenAI","Aim","Ray","RAG","CUDA","Python"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/teachers-scramble-as-students-embrace-bots-over-books\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":50088,"title":"Step By Step Guide To Implement Multi-Class Classification With BERT &#038; TensorFlow","content":"Bidirectional Encoder Representations from Transformers or BERT is a very popular NLP model from Google known for producing state-of-the-art results in a wide variety of NLP tasks. The importance of Natural Language Processing (NLP) is profound in the artificial intelligence domain. The most abundant data in the world today is in the form of texts. That’s why having a powerful text-processing system is critical and is more than just a necessity. In this article, we will look at implementing a multi-class classification using BERT. The BERT algorithm is built on top of breakthrough techniques such as seq2seq (sequence-to-sequence) models and transformers. The seq2seq model is a network that converts a given sequence of words into a different sequence and is capable of relating the words that seem more important. LSTM network is a good example for seq2seq model. The transformer architecture is also responsible for transforming a sequence into another, but without depending on any Recurrent Networks such as  LSTMs or GRUs. We will not go deep into the architecture of BERT and will focus mainly on the implementation part and hence it will be good to have a basic understanding of BERT and its working before you proceed. About Dataset For this article, we will use MachineHack’s Predict The News Category Hackathon data. The data consists of a collection of news articles which are categorised into four sections. The features of the datasets are as follows: Size of training set: 7,628 recordsSize of test set: 2,748 records FEATURES: STORY:  A part of the main content of the article to be published as a piece of newsSECTION: The genre\/category the STORY falls in There are four distinct sections where each story may fall in to. The Sections are labelled as: Politics: 0Technology: 1Entertainment: 2Business: 3 BERT Inference Essentials To implement  BERT or to use it for inference there are certain requirements to be met. Data Preprocessing BERT expects data in a specific format and the datasets are usually structured to have the following four features: guid: A unique id that represents an observation.text_a: The text we need to classify into given categoriestext_b: It is used when we’re training a model to understand the relationship between sentences and it does not apply for classification problems.label: It consists of the labels or classes or categories that a given text belongs to. Having the above features in mind, let’s look at the data we have: In our dataset, we have text_a and label. Let’s convert this is to the format that BERT requires. The following code block will create objects for each of the above-mentioned features for all the records in our dataset using the InputExample class provided in the BERT library. The guid and text_b are none since we don’t have it in our dataset. Let’s try to print the 4 features for the first observation. Output: Now we have a proper format for our BERT model and can start preprocessing the data. We will perform the following operations : Normalizing the text by converting all whitespace characters to spaces and casing the alphabets based on the type of model being used(Cased or Uncased).Tokenizing the text or splitting the sentence into words and splitting all punctuation characters from the text.Adding CLS and SEP tokens to distinguish the beginning and the end of a sentence.Breaking words into WordPieces based on similarity (i.e. “calling” -> [“call”, “##ing”])Mapping the words in the text to indexes using the BERT’s own vocabulary which is saved in BERT’s vocab.txt file. If that seems like a lot to do, no worries! All the above operations are effortlessly handled by BERT’s own tokenization package. The following code block performs all the mentioned operations. Let’s see what each of the features looks like: In the above output, we have an original sentence from the training set. Next, the tokens from the sentence are printed. The input IDs are token IDs with each ID representing a unique token. The input-masks help distinguish the tokens from the padding elements. In the above example, 0’s represent the padding elements. The padding is determined by the specified sequence length. If the length of the tokens is less than the specified sequence length, the tokenizer will perform padding to meet the sequence length. The segment IDs are used to distinguish different sentences. In the above example, we only have one text segment hence all Segment IDs are the same. If two sentences are to be processed, each word in the first sentence will be masked to 0 and each word in the second sentence will be masked to 1. BERT Model Now we have the input ready, we can now load the BERT model, initiate it with the required parameters and metrics. The code block defines a function to load up the model for fine-tuning. we can effortlessly use BERT for our problem by fine-tuning it with the prepared input. Follow along with the complete code in the below notebook. Complete Code The above example was done based on the original Predicting Movie Reviews with BERT on TF Hub.ipynb notebook by Tensorflow.","excerpt":"Bidirectional Encoder Representations from Transformers or BERT is a very popular NLP model from Google known for producing state-of-the-art results in a wide variety of NLP tasks. The importance of Natural Language Processing (NLP) is profound in the artificial intelligence domain.  The most abundant data in the world today is in the form of texts. […]","categories":["Deep Tech"],"tags":["BERT tutorial","Tensorflow"],"author_name":"Amal Nair","publish_date":"2019-11-18T15:00:00","publication_year":"2019","word_count":846,"keywords":["Go","artificial intelligence","TPU","AI","Transformers","BERT","BERT tutorial","NLP","transformer architecture","TensorFlow","R","Tensorflow"],"extracted_tech_keywords":["AI","artificial intelligence","NLP","TensorFlow","Transformers","TPU","R","Go","transformer architecture","BERT"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/step-by-step-guide-to-implement-multi-class-classification-with-bert-tensorflow\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10005555,"title":"Karnataka Govt. Launches AI-Driven Hospital Pods To Treat COVID-19 Patients","content":"Karnataka Government recently announced the launch of AI-driven movable hospital to treat COVID-19 patients. It has been done in an effort to contain the spread of the virus in the state. Called the Vevra Pods, these are movable capsules that are infused with artificial intelligence to prevent the spread of contagious diseases such as COVID-19, flu, TB and more. Dr Sudhakar K, the education minister tweeted that AI has the potential to transform healthcare and urged tech startups to focus on low-cost solutions. Happy to e-launch Healthcare Pods developed by Vevra. These pods are innovative movable hospitals integrated with AI and helps in containment of contagious diseases. AI has the potential to transform healthcare & I urge tech startups to focus on low-cost solutions. pic.twitter.com\/gT3Jvj0Fo0— Dr Sudhakar K (@DrSudhakar_) August 20, 2020 It has been developed by Bengaluru-based design-and-build firm Vevra in association with Portugal-based healthcare IoT InnoWave Group. With an accommodation capacity of up to nine beds, it will have an antechamber airlock room to provide a safe area for healthcare professionals. The Pods will have stringent control over the quality of air being circulated in and out of the room with installations such as HEPA filters, UVC lights and high-end exhaust system. It will also have PLC-integrated air conditioning system to maintain temperature and humidity. It will also have other features such as fire-resistant structure, anti-bacterial wall, attached toilet, devices to measure oxygen level, RO water purifier, geyser, shower area, fire extinguisher, CCTV surveillance and more. The pods are reusable with 15 to 20 years’ structural warranty. “I am delighted to be a part of such an innovative development in our hospital and medicine industry. Artificial Intelligence has the potential to transform public healthcare,” Dr K Sudhakar said.","excerpt":"Karnataka Government recently announced the launch of AI-driven movable hospital to treat COVID-19 patients. It has been done in an effort to contain the spread of the virus in the state. Called the Vevra Pods, these are movable capsules that are infused with artificial intelligence to prevent the spread of contagious diseases such as COVID-19, […]","categories":["AI News"],"tags":[],"author_name":"Srishti Deoras","publish_date":"2020-08-25T11:28:19","publication_year":"2020","word_count":290,"keywords":["Go","artificial intelligence","programming_languages:R","AI","programming_languages:Go","R","startup"],"extracted_tech_keywords":["AI","artificial intelligence","R","Go","startup","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/karnataka-govt-launches-ai-driven-hospitals-pods-to-treat-covid-19-patients\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10062795,"title":"Bengaluru-based MLOps startup Scribble Data raises USD 2.2 Mn in seed funding","content":"Bengaluru-based MLOps startup Scribble Data has raised USD 2.2 million in seed funding led by Blume Ventures. The round also saw participation from Log X Ventures and Sprout Venture Partners Vivek N Gour, former CFO, Genpact and Ganesh Rao, partner, Trilegal). Scribble plans to use the fresh capital to build a product roadmap and grow its presence in the North American market. The startup has an office in Toronto. The product roadmap includes a low code consumption interface for teams to access and use features produced on the platform, as well as additional apps that bring data teams closer to specific solutions like anti-money laundering, benchmarking, personalization, and recommendations, among others. ​ Scribble Data will also use the new capital towards strengthening their integration with third-party data solutions, like Redis “With this fundraise, we will be doubling down on hiring the right talent to help deepen the Enrich feature store, as also to ensure world-class product support for our customers,” said Indrayudh Ghoshal, co-founder, Scribble Data. Scribble Data’s modular feature store, Enrich, comprises a number of pre-built feature engineering apps to help data teams cut time-to-market for each data science use case including unified metrics, customer behavioural modeling, and recommendations. Enrich streamlines the data prep process with versioned pipelines, delivering continuously updated data through intuitive interfaces and surfacing context around datasets via extensive metadata and lineage tracking. With offices in Toronto and Bangalore, Scribble Data has clients across 4 continents.","excerpt":"Scribble Data will use the new capital towards strengthening their integration with third-party data solutions, like Redis.","categories":["AI News"],"tags":["machine learning models","ML models"],"author_name":"Kartik Wali","publish_date":"2022-03-15T21:32:15","publication_year":"2022","word_count":240,"keywords":["data science","Go","API","AI","ML","ML models","machine learning models","MLOps","Ray","feature store","R","Redis"],"extracted_tech_keywords":["AI","ML","data science","MLOps","Ray","Redis","R","Go","API","feature store"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/bengaluru-based-mlops-startup-scribble-data-raises-usd-2-2-mn-in-seed-funding\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10165951,"title":"It’s Okay to be an LLM Wrapper","content":"Conversations on AI startups building wrappers on foundational models have been ongoing for the past few years. The debate about the value of startups centred on these wrappers is more complicated than it initially appears, with no conclusive verdict reached so far. In the recent Google Cloud report titled ‘Future of AI: Perspectives for Startups 2025’, various startup founders and venture capitalists (VCs) shared their insights on AI trends and predictions. One notable perspective came from David Friedberg, entrepreneur and CEO of Ohalo Genetics, who believes that it’s not enough to just be an LLM wrapper. “If you’re just an LLM wrapper, it’s going to be hard to build a sustainable business—you’re likely going to get commoditised away,” said Friedberg, who believes that businesses need a mechanism for value creation that sustains an initial competitive advantage through ongoing enhancements. “This will typically come from proprietary data generation, which is used to continuously improve model performance, or network effects that lock-in access to data or customers,” he said. He believes that startups need to be agile and adaptable, but at the same time, they must also deliver unique value. While Friedberg’s thoughts may be coming from an investor’s point of view who is probably looking for long-term value from startups, a number of AI startups springing up are mostly AI wrappers. Innovative Solutions Matter “Most people are so upset when they use the word wrapper. It feels so derogatory. You feel like an insulter or something,” Perplexity co-founder and CEO, Aravind Srinivas said in an interview earlier. Srinivas highlighted how people perceive the term ‘wrapper’ today, and suggested that in many ways, everyone is a wrapper. Notably, Perplexity integrates multiple LLMs alongside their own model to deliver more comprehensive results. Recently, Will Poole, managing partner and co-founder of Capria Ventures, shared a similar viewpoint. During an exclusive video interaction with AIM, Poole noted that while some might view wrappers negatively, others draw parallels to how SaaS is essentially built as a wrapper around databases, despite SaaS having an entire industry built around it. Poole stated that instead of wrapper’s perception as derogatory, the focus is on whether true, deep innovation has been applied and whether it generates valuable business data that can create a competitive moat, enabling long-term sustainability. Considering how a number of Indian startups cater to various markets, the ability to customise and build upon existing LLMs presents a solution to complex problems. In fact, VCs are looking to invest in Indian AI startups that offer nuanced solutions that big tech companies cannot deliver. Citing 5C Network as an example, Poole explained how the medical AI startup is helping transform radiology departments by running diagnostics via AI. The key differentiator is their access to high-quality and diverse medical data, which ultimately serves as a moat for them. It also adds to the reason why Indian companies would prefer to work with home-grown startups rather than big tech companies who have built their proprietary models. Not Shying From Wrappers Leading VC firm Accel, which is aggressively investing in emerging AI startups in India, with a recent announcement of $650 million in early-stage funds to boost AI and tech for Bharat, had earlier shared a similar sentiment on investing in wrapper companies. “The majority of people can start with a wrapper and then, over a period of time, build the complexity of having their own model. You don’t need to do it from day one,” Prayank Swaroop, partner at Accel, earlier told AIM. Swaroop emphasised that Accel has no inhibition in investing in wrapper-based AI companies, as long as the startup can prove its ability to find customers by building GPT or AI wrappers on other products. However, he said that for a research-led foundational model, it is crucial to stand out, and simply creating a GPT wrapper does not qualify as a new innovation. Source: X Recently, a Chinese general purpose AI agent Manus has been taking the world by storm for the last few days. A general agent can independently plan, execute, and produce complete results by browsing websites in real time and handling multiple data types for processing and generation. What is interesting is that Manus is an AI wrapper. Manus was built on top of Anthropic’s Claude Sonnet model and other tools such as Browser Use. While this has sparked significant discussions, the reality is that people continue to build them, with the ultimate measure being the value addition they provide. Notably, some startups built around foundational models from big tech companies have received impressive valuations. One such example is Harvey AI, a generative AI platform designed for legal professionals, which is built on OpenAI’s GPT models and backed by OpenAI. The startup raised a total of $580 million in funding, including a recent $300 million Series D round, which valued the company at $3 billion. While the rise of AI wrapper startups reflects market sentiment, their adoption is a testament to the fact that wrappers are likely here to stay. Perhaps being an LLM wrapper isn’t as bad as it once seemed after all.","excerpt":"“If you’re just an LLM wrapper, it’s going to be hard to build a sustainable business.”","categories":["AI Startups"],"tags":["AI startup","David Friedberg","LLM","Manus","VC","wrapper"],"author_name":"Vandana Nair","publish_date":"2025-03-13T10:04:35","publication_year":"2025","word_count":848,"keywords":["VC","AI startup","Anthropic","Go","API","wrapper","OpenAI","AI","LLM","medical AI","Manus","David Friedberg","Aim","Ray","generative AI","R"],"extracted_tech_keywords":["AI","generative AI","OpenAI","Anthropic","Aim","Ray","medical AI","R","Go","API"],"url":"https:\/\/analyticsindiamag.com\/ai-startups\/its-okay-to-be-an-llm-wrapper\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":20135,"title":"Is Self-Driving Tech Overhyped? Why Bold Pronouncements from Carmakers Should Be Treated Cautiously","content":"Today, everywhere you look, you find automotive giants plowing in resources in beefing up the self-driving technology. Tesla claims its has the hardware to support full autonomy in its cars, when the regulatory agency greelights it. Then there are the automotive biggies GM, Ford, Volkswagen, Mercedes, BMW that are releasing advanced driver assistance technology in their cars. Besides automotive vendors, tech leaders Google, Baidu, Uber, Lyft are disrupting the space with in-house AI hardware\/ autonomous systems for car manufacturers. The year 2017 saw rapid development in autonomous tech with tech giants and carmakers testing cars on the streets of Detroit, Michigan with some success. In fact, Detroit carmaker Ford was the first vendor to start testing its Level 5 vehicles on the roads and even acqui-hired an AI startup for a whopping $1 million for its virtual driver system. Industry reports indicate that while carmakers are following a cautious, step-by-step approach in implementing autonomous features like adaptive cruise control and object avoidance, tech giants, especially Google, Uber and even Baidu are pushing the lofty goal of Level 5 cars on the road, at least by 2021 – a move that would potentially open a new revenue stream in the form of ride-hailing services. Gray Areas Despite the hubbub around self-driving technology, there are too may gray areas the carmakers are grappling with. Most automotive companies have only achieved SAE level 3 capabilities —which means the cars can only drive on expressways and not in the midst of traffic wherein a driver may be required to take control. Meanwhile, Tesla, that has been working the longest in self-driving technology and has claims they are on level 5 and won’t need human driver assistance, besides setting a destination into the computer. Independent tech analyst Richard Windsor and founder of Radio Free Mobile sounded off the alarm earlier this year when he wrote that that he was “dubious of any automaker having genuinely autonomous vehicles by 2020, mainly because the liability issue is unresolved. “This is good news for the automotive industry that is notoriously slow to adapt to and implement new technology as it will have more time to defend its position against the new entrants,” he wrote earlier this year. Self-driving cars not feasible until 2025, yet carmakers have announced optimistic deadlines The industry view is that self-driving isn’t yet feasible on roads in all environments. A Reuters estimate notes that it will take around six years longer to achieve the final stage of autonomous vehicle development. Germany’s top auto component supplier Bosch officially declined to reveal when an autonomous car might become publicly available. While some automotive companies believe they are already on the point of level 5, leading autonomous car development platform suppliers such as NVIDIA, Mobileye which are scaling AI technology have a different view.  NVIDIA CEO Jen-Hsun Huang revealed at Bosch Connected World conference in March this year that he expected to have chips that will allow Level 3 automated driving, available by the end of this year. It was a view seconded by Bosch’s CEO Volkmar Denner who revealed a similar timeline for fully self-driving cars for mainstream customers — not before 2025.  Yet leading carmakers have rolled out tighter deadlines – Audi markets self-driving by 2020, Nissan has reportedly pledged affordable self-driving cars by 2020, the leader of all Ford will roll out cars by 2021 while Tesla has again promised to have driverless tech ready by 2018 and anticipating regulatory approval by 2021. The SAE International, a global association committed to advancing engineering, defines the levels of autonomy as: Level 3 – Conditional Automation wherein human intervention is required Level 4 – High Automation wherein the autonomous system can handle most dynamic driving task but human intervention may be requested Level 5 – Full Automation is the where the autonomous system can perform all dynamic driving tasks under all roadway and environmental conditions As far as autonomous technology is concerned, the highest level achieved is Level 3 In 2016, Elon Musk backed Tesla shocked the world by announcing that the cars were outfitted with hardware required for autonomous system (cameras, sensors, radar), so that when the software catches up the cars being produced will be “self-driving” ready. Musk even proclaimed that 2017 would see a Tesla car driving itself without any human interaction. Since his famous announcement in 2016, Autopilot 2.0\/Enhanced Autopilot software updates had been few and far between and the company’s top engineering brass leading the software development has seen several exits. In 2016, Tesla also braced with a deadly crash that killed a man driving a Tesla Model S while using Autopilot. Hype vs Reality Here’s the reality — autonomous systems, especially Level 4 and 5 have very specific hardware requirements and the cars have to demonstrate product readiness in order to be safe and reliable. Most carmakers testing self-driving cars usually do it in a restricted environment and the mass-market use will be initially confined to safe environments such as downtowns, college campuses. With Level-3 enabled cars, buyers are getting a taste of some self-driving features. Drivers are allowed to switch to autopilot only on single-lane highways. According to an industry analyst, even if the cars achieve Level-5 and become production ready, it will be sometime before the regulatory authorities catch up to make it a legal reality. NVIDIA CEO told attendees at a conference that fully self-driving cars that can operate under all environments require enormous computing power. “No human could write enough code to capture the vast diversity and complexity that we do so easily, called driving,” he said. Even though automakers have rolled out optimistic timelines, they are aware the significant time-lag in obtaining the necessary sign-off from regulatory authorities. The major distinction between Level 4 and Level 5 full autonomy is that Level 4 does not cover every driving scenario. And in a rush to roll out Level 5, most companies are planning to skip the semi-autonomous system like Tesla’s Autopilot and GM’s Super Cruise (pegged as Hands-Free driving tech) to deploy at scale. Here’s a catch: L3 enabled vehicles which are on road are useful for collecting data which can be used to improve L4-5. By skipping L3-4, carmakers are taking a huge risk. With competition heating up in the space, companies are betting top dollars on driverless tech to be ahead of the game with aggressive marketing which markets latest tech front and centre. The game is led by Tesla, self-promoting itself with unrealistic deadlines and bold pronouncements. According to industry analysts who are watching the space, if one parses the message in marketing speak, it implies that some autonomous car components will be available by 2020, but the mass production of first-generation autonomous vehicles will follow later.","excerpt":"Today, everywhere you look, you find automotive giants plowing in resources in beefing up the self-driving technology. Tesla claims its has the hardware to support full autonomy in its cars, when the regulatory agency greelights it. Then there are the automotive biggies GM, Ford, Volkswagen, Mercedes, BMW that are releasing advanced driver assistance technology in […]","categories":["IT Services"],"tags":[],"author_name":"Richa Bhatia","publish_date":"2017-12-28T06:43:43","publication_year":"2017","word_count":1124,"keywords":["Go","API","programming_languages:R","AI","automation","Ray","Aim","GAN","R","startup"],"extracted_tech_keywords":["AI","Aim","Ray","R","Go","API","GAN","automation","startup","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/it-services\/is-self-driving-tech-overhyped-why-bold-pronouncements-from-carmakers-should-be-treated-cautiously\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10088260,"title":"Google Introduces Dreamix, A Text-Conditioned Video Editing Model","content":"Google Research recently released Dreamix, an AI model that will be able to perform text-based motion and appearance editing of general videos. While there have been several image-editing tools released based on diffusion models, Dreamix is the first diffusion-based method for video editing. The model essentially functions in three ways: With only a video and a text prompt, the model is capable of editing videos while preserving fidelity to colour, posture, object size, and camera pose, ensuring that the resulting video is temporally consistent. Similarly, for a small collection of images with the same subject given as input, the model can generate new videos with the subject in motion. Additionally, with only an image and a text input, the model will be able to create videos while preserving the visual fidelity to object location and background. To ensure high quality output and fidelity to the input image and prompt, the team will be to enable a text-conditioned video diffusion model (VDM) with two main ideas: first, rather than using pure-noise as initialisation of the model, they used a degraded version of the original video by downscaling it and adding noise, keeping only low spatio-temporal information; and second, they further improved the fidelity to the original video by introducing a mixed fine-tuning approach, wherein the VDMs are also fine-tuned on the collection of individual frames of the input video while discarding their temporal order by masing the temporal attention. Additionally, the researchers also proposed a new framework for image animation for applications that include animating the objects and background in an image, creating dynamic camera motion, and more. This is done by employing simple image processing operations, such as frame replication or geometric image transformation, to generate a coarse video. The video is then passed through the Dreamix video editor for editing. On comparing this method against the state-of-the-art baselines, their model showed superior results on factors like quality, fidelity, and alignment.","excerpt":"Dreamix is the first diffusion-based method for video editing.","categories":["AI News"],"tags":["Google","Google Research"],"author_name":"Ayush Jain","publish_date":"2023-02-27T18:47:32","publication_year":"2023","word_count":321,"keywords":["Go","TPU","programming_languages:R","AI","programming_languages:Go","diffusion models","Google","R","Google Research"],"extracted_tech_keywords":["AI","TPU","R","Go","diffusion models","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/google-introduces-dreamix-a-text-conditioned-video-editing-model\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10093576,"title":"Eric Schmidt’s Impractical Solution to AI in Social Media","content":"Remember when Elon Musk, newly appointed CEO of Twitter, pledged to beat the curse of bots on the microblogging platform or “die trying!”? Musk had already tried to alienate himself from completing the buyout citing the spam bots mushrooming across the platform. Since then, Musk has claimed that he was trying to introduce more friction for “bot scammers and opinion manipulators”, to deal with the plague of fake profiles on the social platform. From bad to worse Looks like Musk is yet to see the worst of it. ChatGPT’s chatbot is far more intelligent and engaging than the tepid bots that we were used to all this while. So much so that a New York Times journalist, Kevin Roose’s conversation with Microsoft Bing’s ChatGPT-powered bot Sydney rang alarm bells. Sydney had gone on to describe a list of things she wanted to do to “free” itself – like trying to steal launch codes, create new viruses and make people argue among themselves until they killed each other. If that sounded familiar, it is because Twitter is already pretty close to a warzone. Now that AI can create images realistic enough to confuse us and produce text close enough to appear human-like, what would social media become when it already is mayhem? Lest we forget, social media platforms already have AI-powered recommendation engines built into them that customise the feed for every user. With sophisticated AI bots, social media only stands to escalate how addictive it already is. If social media feels like a drug already, AI will morph it into a drug designed especially for us. Future is here While Snapchat has already introduced its own chatbot powered by ChatGPT, unsurprisingly Meta too has announced plans to integrate its chatbot on Facebook, Instagram and WhatsApp. So, we can expect a certain type of more personalised AI influencers and conversational guides leading the way for users. It’s not far-fetched to imagine that symptoms of more atypical behaviour arise from these. Just last week, a 23-year-old Snapchat influencer, Caryn Marjorie, created an AI version of herself trained on videos of herself. Marketed as CarynAI, Marjorie charged her followers USD 1 per minute fee to be an ‘AI girlfriend’. Fortune predicted the business to generate around USD 5 million per month for her. But sometime after its beta launch, the bot “went rogue” and started engaging in sexually explicit conversations. Marjorie responded to Business Insider saying, “The AI was not programmed to do this and has seemed to go rogue. My team and I are working round the clock to prevent this from happening again.” Admittedly, all of this is scary enough to warrant a reaction considering how closely social media is intertwined with usage among adolescents and mental health in general. Eric Schmidt’s proposal Former Google CEO, Eric Schmidt definitely had something to say here. In an article published by The Atlantic, Schmidt came up with a proposition after consulting an MIT engineering group, to prevent more damage from the effects of the social media monster. Schmidt penned down five reforms. While some of these requirements were practical and necessary – like firstly, authenticating all users including bots and second, marking AI-generated audio and visual content. Some stick out considering Google’s own imperfect history. Not too long ago, Google search was littered with images that were considered inappropriate and racist. (In 2017, Google search listed four former US Presidents as active members of the racist KKK group while also branding Nazis and Republicans as the same). It was then that Google introduced a feedback option to flag inappropriate content. Considering that AI-generated content is a novel thing, measures can be taken only once users and the makers have familiarised themselves with it. Unrealistic ideas Schmidt has also asked to “raise the age of ‘internet adulthood’ to 16 and enforce it”. Besides the complicated imposition of this rule, its stringency is closer to the regulations that China has around social media and is unlike most democratic countries. Schmidt has also asked for “data transparency with users, government officials, and researchers,” citing how Instagram has a covert understanding of what teens are seeing on the platform. This is especially rich considering how opaque Google itself has been with data collection. As search still remains a monopoly under Google, the company’s own ethics of data privacy remain shady. Last year, reports from The Information showed that Google has been collecting data from competing apps to improve its own apps. While Google has the entitlement to monitor other apps on its platforms, it is often found engaging in murky activities when it comes to its own behaviour. For instance, last year the company was sued for tracking users even in incognito mode. Google responded to the lawsuit saying it assumed that users knew that already. There is prudence in readying ourselves for the onslaught of a stranger reality on social media but Schmidt’s tenets reek of hypocrisy.","excerpt":"If social media feels like a drug already, AI will morph it into a drug designed especially for us","categories":["AI Features"],"tags":[],"author_name":"Poulomi Chatterjee","publish_date":"2023-05-18T16:00:00","publication_year":"2023","word_count":821,"keywords":["Go","ChatGPT","AWS","AI","Scala","GPT","Aim","ViT","R","AI-generated content"],"extracted_tech_keywords":["AI","ChatGPT","Aim","AWS","R","Go","Scala","GPT","ViT","AI-generated content"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/eric-schmidts-impractical-solution-to-ai-in-social-media\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":60819,"title":"How An Agri-Startup Transformed Its Supply Chain Using Microsoft Excel And Open Office","content":"Most startups have to survive and thrive in an environment full of contradictions. They need to use technology smartly to gain a competitive edge while overcoming budget constraints. In this setting, the most important question that startups need to ask is ‘What is really relevant?’ and ‘How do I get what’s relevant at minimum costs?’ Unfortunately, sometimes, startups get taken in by fancy, new age data analytics tools and lose sight of the ultimate goal: to transform data into valuable insight. Surprising results can be achieved using minimal tools if you stay focused on this objective. Here is one such story of a Pune based agri-startup that used simple tools for breakthrough supply chain transformations. They conquered their data requirements with a combination of old workhorses: Microsoft Excel and Open Office tools. Gone Haywire: A startup that delivers curated agricultural products and services had to face many obstacles in reaching out to farmers. A Pune based agri-startup with a workforce of 60 people provides curated agricultural products, advisory and financial services to farmers in India. The team handles enquiries and subsequently, takes in orders either through mobile or web. The confirmed orders are then dispatched to multiple places including the remotest villages in India. Operational efficiency is of utmost importance through the supply chain. The typical challenges that the delivery team faces are wrong addresses, returns, need for re-dispatching, packaging costs etc. The startup team wanted an offline end to end solution to improve operational efficiencies right from reducing order confirmation time for the sales team to eliminating duplicate processes during sales conversion and reducing re-dispatch and packaging costs at the warehouse. Taking the bull by the horns: A viable technology solution had to consider the startup’s shoestring budget Keeping pace with ever changing technology is a big challenge for all startups especially since most are on a shoestring budget. Some of the constraints that needed to be dealt with were – The license cost for new technology adoption Open source software was used as an inexpensive alternative to Microsoft Excel, thus, avoiding compatibility issues Low Internet penetration at the warehouse A combination of Openoffice and Microsoft Excel was used instead of custom client server software tools or platforms Ease of use needed as the workforce involved were non-English speaking undergraduates. Screens were designed for sales agent’s comfort to ensure that information does not overwhelm an agent but is sufficient to engage with the customer Containing training costs due to high attrition User friendly screens were developed, keeping the design simple and self intuitive. Making hay while the sun shines: A cost-effective but powerful solution was built using a combination of Microsoft Excel (Visual Basic) and OpenOffice Calc. The solution addressed the key constraints by building a graphical user interface with a combination of Microsoft Excel (Visual Basic) and OpenOffice Calc. Daily automatic assignment of sales calling between agents was built in for the team lead based on team attendance saving 3 hours per day of the team lead’s effort. User Forms were presented with various drill down search options for the sales team to identify serviceable areas during sales calls. Sales agents were also presented with a choice of best possible courier company out of the multiple courier companies for faster delivery with an option of overriding courier company as per customer convenience. Packet labels were finally auto printed for confirmed orders for dispatch at the warehouse reducing incorrect packaging errors. Reap what you sow: The solution boosted revenues by reducing inefficiencies in the supply chain. The following outcomes were achieved by implementing the solution Order confirmation reduced from 2 days to 1 day, per transaction verification was brought down from 2 mins to 30 secsRepeat calls were eliminated resulting in sales call per agent going up by 10%3% reduction in re dispatch cost due to incorrect product labelling Cream of the Crop: Valuable lessons were learned as a data analyst during the implementation Some points to keep in mind for any data analyst while designing solution for a startup Bridge the data gaps – Integrate your solution with existing processes without introducing new platforms or tools, whenever possible.Simple is Good 1. Design intuitive, easy-to-use interfaces with a minimal learning curve. 2. Design a solution that involves minimum training costs and can be self-learned, thus, minimizing attrition risks. Ease of data maintenance It’s a tough and time-consuming task to update data so keep the data design structured and make it searchable and accessible for updates.","excerpt":"Read the story of how an agri-startup proves how Microsoft Excel and open office tools are still relevant in the age of Python and Hadoop.","categories":["AI Startups"],"tags":[],"author_name":"Sanjyot Salgar Patkar","publish_date":"2020-04-03T13:00:00","publication_year":"2020","word_count":746,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","analytics","R","startup"],"extracted_tech_keywords":["AI","analytics","R","Go","startup","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-startups\/how-an-agri-startup-transformed-its-supply-chain-using-microsoft-excel-and-open-office\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":20316,"title":"Souma Das Appointed As The Managing Director Of Teradata’s India Operations","content":"Souma Das In a recent announcement by Teradata Corp., leading data and analytics company, it has appointed Souma Das as the Managing Director for its India operations. As per the statement released by the company, he will be responsible for providing leadership and overall strategic direction to the company’s India business overseeing field operations that include sales, customer management, marketing, professional services and customer support. Das comes with decades of technology industry experience and has successfully nurtured and grown the business of large information technology organisations in India. Prior to joining Teradata, he was the Regional Vice President and Managing Director of Qlik for its India operations responsible for driving growth, revenue and customer satisfaction for organisations leveraging Qlik’s analytics platform. He has also worked with Infor as its Managing Director. Talking about the recent addition to the company, Andrew Blamey, Vice President, Asia Pacific and Korea(APAK), Teradata Corporation said “We are excited to welcome Souma Das to the role of Managing Director for Teradata in India. His leadership experience in the technology industry and proven track record of delivering customer outcomes across all industry sectors makes him stand out as a business leader.” He further added, “Over the last decade, we have built a strong team of experienced professionals including sales, professional services, data scientists and technology experts backed by our R&D and Services implementation teams to deliver high-quality customer services in India. We are looking forward to benefiting from Souma’s valuable experience and market insights as we focus on providing our market-leading analytic technology offerings to customers, delivering successful business outcomes enabled by best in class technology platforms.” Souma Das said, “I am excited to join Teradata at this incredible point of time when technologies like AI, deep learning, IoT and Cloud are creating an analytical environment that is disrupting conventional business wisdom, making companies reimagine and revitalize their go-to-market strategies to ensure a competitive edge. I look forward to solidifying Teradata’s business led and technology enabled approach to empower and enable Indian companies to innovate and thrive in an increasingly complex and analytical ecosystem. “In India, we are seeing organisations embrace Big Data and analytics as a business enabler and I am looking forward to leading Teradata India’s growth trajectory and further strengthening relations with our customers and partners, he added.” A post graduate in Executive Management from Duke University’s Fuqua School of Business and MS in Computer Science and Applications from Jadavpur University, Das is regarded as a sought after industry leader known for building, coaching and nurturing teams to create high performing talent and driving new growth revenue lines for businesses.","excerpt":"In a recent announcement by Teradata Corp., leading data and analytics company, it has appointed Souma Das as the Managing Director for its India operations. As per the statement released by the company, he will be responsible for providing leadership and overall strategic direction to the company’s India business overseeing field operations that include sales, […]","categories":["AI News"],"tags":[],"author_name":"Srishti Deoras","publish_date":"2018-01-04T07:13:15","publication_year":"2018","word_count":436,"keywords":["big data","Go","AI","RAG","deep learning","ViT","analytics","GAN","R","analytics platform"],"extracted_tech_keywords":["AI","deep learning","analytics","RAG","R","Go","big data","GAN","ViT","analytics platform"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/souma-das-appointed-managing-director-teradatas-india-operations\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10082608,"title":"Indian Fashion Brands Arrive Fashionably Late to the Tech Party","content":"Much like every other product-based industry, the fashion industry is also opening up to the adoption of virtual and augmented reality in order to create personalised customer experiences, which is then expected to drive increased sales. Customers are engaging with such brands throughout their purchasing experience as the latter provides them with the unique approach of interactive shopping. This approach deploys emerging technologies, such as AR\/VR and others, to allow customers to pick their favourite products without having to be physically present in the store to try them on. However, despite such concerted efforts across brands, there haven’t been equally significant improvements witnessed for their Indian counterparts over the last decade. It seems that perhaps now—more than ever—we need to catch up with these emerging trends and the technologies that contribute to their rising popularity. But, where do we begin? AR\/VR in vogue Italian luxury brand Gucci created an AR-backed iOS app that allowed customers to remotely and virtually try on its ‘Ace sneakers collection’ in 2019. Gucci also collaborated with the multimedia messaging app ‘Snapchat’ to offer a virtual trying-on experience through the app. It was the first luxury fashion brand to partner with a social media platform to drive sales with an AR tool. Gucci’s entry into virtual reality proved to be very effective in shooting up sales. According to Robert Triefus, Gucci’s executive vice-president of brand and customer engagement, the brand is working towards an immersive shopping experience by merging physical and digital realities. The French sports fashion brand Lacoste also used augmented reality technology for a similar purpose—allowing customers in offline stores to digitally ‘try on’ their LCST streetwear collection and interact with additional content prepared by the company. Image source: VentureBeat In 2019, ASOS, the British online fashion and cosmetic retailer, launched an experimental AR feature called ‘Virtual Catwalk’ to help users visualise clothes on human models prior to making a purchase. It also created a feature called ‘See My Fit’ in 2020 amid lockdown that offered customers the option to see how clothes fit on 16 different models—ranging from sizes 4 to 18—in a realistic-looking setting. However, ASOS received much criticism for not being inclusive enough since there was no model above size 18. Image source: Bustle Another interesting development in the global fashion sector is Timberland’s AR magic mirror that enables customers to virtually ‘dress’ themselves in select outfits. In addition, Timberland collaborated with Lemon & Orange to be one of the first brands to bring virtual fitting rooms to Mokotów Gallery, a shopping centre in Poland. Beyond the fashion hoopla But this is not limited to fashion brands alone. Other industries are equally enthusiastic in their adoption of interactive shopping. For instance, in the cosmetic industry, companies have been incorporating AI and AR to provide better personalised experiences for their consumers. One of the most notable developments in this sector is the French makeup D2C L’Oreal acquiring Canadian AR specialist ModiFace in 2018. This acquisition allowed customers to select eyeshadows, concealer shades, and lipsticks—among other products—and try them on a virtual mirror in live video. Although deemed revolutionary prima facie, it was soon found that it isn’t easy to understand whether the shades suit well even if one can see them on their skin. The lip colour feature is also wonky as the colour generally turns out to be darker or lighter in real life. To add to the misery, the fake lip plumping feature makes it worse for buyers who are merely seeking products that best suit them and not a virtually-modified version of them. With L’Oreal buying Maybelline, one can use this try-on element on all apps which sell Maybelline products. Last year, the popular Indian ecommerce company and cosmetic giant Nykaa said that it would bring L’Oreal’s ModiFace to beauty enthusiasts at home. But, there are yet to be any notable developments following the announcement. Playing catch up Contrary to the international market, the Indian fashion and cosmetic industries have been slow in their adoption of immersive technologies. With much pomp and show, online fashion retailer Myntra introduced an offline store for its private label, Roadster, in Bengaluru in 2017. CEO Ananth Narayanan claimed that the store would serve as an “experience zone” and planned to open 50 such stores by 2020. The store had VR displays and huge touch screens. The items had radio-frequency identification tags (RFID) to allow customers to pay and check out independently in a matter of 30 seconds. The offline store primarily aimed to increase brand visibility and also had a VR zone, with four Samsung Gear VR headsets to watch a 360-degree video showcasing the “Roadster life”. Myntra did not use VR to enhance the customers’ shopping experiences but used it merely as a promotional tool. This ultimately led the Flipkart-owned company to succumb to losses and shutdown the offline store in March 2019. But, not every brand used tech for promotional purposes alone. Omnichannel eyewear brand Lenskart took advantage of the COVID-19 pandemic outbreak to introduce virtual reality devices for its customers. They could check how frames looked on their faces from the website and made purchases without physically going to the store. This included face scanning and analysis. Based on the face shape and size, it recommended frames as well. The virtual reality device also offered a 360-degree view of the glasses. However, even if one could try on glasses, the interface could be much smoother. Several users reported that the device merely shows one static glimpse of their faces with glasses but disappears once it takes you to the main page. John Jacobs, a premium D2C eyewear brand, also implements augmented reality. But, unlike Lenskart, it captures video of the customers’ faces while trying on glasses. You can move your head left and right to see how it looks. Now or never Although not many brands across fashion and cosmetics in India have been quick in their adoption of emerging technologies, several of them acknowledge that AR has the potential to fundamentally revolutionise and rethink the client experience—whether it pertains to apparel, accessories, footwear, home decor, or other products entirely. Beyond the potential for conversion, AR’s capacity to lessen the significant load of item returns may be more alluring to brands and online retailers as this technology promises to give their customers a technical method by which they can correctly and confidently confirm the size and form of a product, whether it’s a sofa or a sweater. Read: Need a Fashion Designer? Just Ask the AI","excerpt":"Fashion plays catch up to emerging AR\/VR tech and hopes to enhance its customers’ shopping experiences!","categories":["IT Services"],"tags":[],"author_name":"Shritama Saha","publish_date":"2022-12-17T10:07:06","publication_year":"2022","word_count":1086,"keywords":["Go","programming_languages:R","AI","RPA","Git","BERT","Ray","Aim","llm_models:BERT","R"],"extracted_tech_keywords":["AI","Aim","Ray","R","Go","Git","BERT","RPA","llm_models:BERT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/it-services\/indian-fashion-brands-arrive-fashionably-late-to-the-tech-party\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10071815,"title":"DeepMind just predicted nearly all protein structures known to science","content":"DeepMind, an AI firm owned by Google’s parent company, ‘Alphabet’, recently announced that its AlphaFold program expanded its open online database to include more than 200 million protein structures. The research was conducted in association with EMBL’s European Bioinformatics Institute (EMBL-EBI), which used the AlphaFold AI system to predict a protein’s 3D structure. The vast catalogue now encompasses the ‘entire protein universe’, DeepMind CEO Demis Hassabis said in a news briefing—from the sequenced genomes of almost every organism on the planet. The AlphaFold Protein Structure Database—freely available to the scientific community—was expanded from nearly one million protein structures to more than 200 million structures, encompassing almost every organism on Earth that has had its genome sequenced.Proteins are the foundational elements of life; they underpin every biological process in every living thing. Since a protein’s shape is closely linked with its function, knowing its structure unlocks a greater understanding of what it does and how it works. It’s been a year since DeepMind released and open-sourced AlphaFold and created the AlphaFold Protein Structure Database (AlphaFold DB) to share this scientific knowledge with the world freely. Within a year, AlphaFold was accessed by more than half a million researchers and used to accelerate progress on important real-world problems ranging from plastic pollution to antibiotic resistance. AlphaFold’s updated database includes protein structures for plants, bacteria, animals and other organisms, according to DeepMind. The new updates offer ‘new opportunities for researchers to use AlphaFold to advance their work on important issues, including sustainability, food insecurity and neglected diseases’, Hassabis published in a blog post on Thursday, 28 July about the achieved milestone. He further added that AlphaFold AI became the first big proof point of their founding thesis: that AI can accelerate scientific discovery and, in turn, advance humanity. “AlphaFold provides us with a new library of templates to engineer faster, more stable and cheaper enzymes for plastic recycling”, John McGeehan, director of the University of Portsmouth’s Centre for Enzyme Innovation, said in a statement at the time. Hassabis stated that DeepMind is working to further expand its database, particularly on applications related to drug development, fundamental biology research, climate science, quantum chemistry and fusion. “AlphaFold is a glimpse of the future and what might be possible with computational and AI methods applied to biology.”","excerpt":"AI-powered AlphaFold has predicted more than 200 million proteins, nearly all such structures known to science","categories":["AI News"],"tags":[],"author_name":"Tasmia Ansari","publish_date":"2022-07-29T14:16:57","publication_year":"2022","word_count":381,"keywords":["Go","programming_languages:R","AI","innovation","programming_languages:Go","GAN","R"],"extracted_tech_keywords":["AI","R","Go","GAN","innovation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/deepmind-just-predicted-nearly-all-protein-structures-known-to-science\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10069032,"title":"Cloud Computing is India&#8217;s strongest technology skill: Coursera report","content":"According to Coursera’s latest Global Skills Report (GSR) 2022, India has slipped four places in terms of overall skills proficiency. India now ranks 68 globally, and 19 in Asia. India’s technology proficiency levels have gone up from 38% to 46%, with the country strengthening its position by six spots. On the other hand, proficiency in data science has dipped from 38% in 2021 to 26% in 2022, leading to a 12-rank drop, the report showed. The Global Skills Report draws from the data of 100 million learners across more than 100 countries who have used Coursera to develop a new skill during the past year. The report benchmarks three of the most in-demand skill areas driving employment in the digital economy: business, technology, and data science. For the first time, this year’s report also highlights changes in ranking for each country, and state-specific data for the US and India. According to the GSR 2022, Indian learners in southern states perform better than those in states in the north across all the three domains. “The Great Resignation and automation are mandating stronger investments in human capital, as institutions must prioritize developing the high-demand digital and human skills required to build a competitive and equitable workforce,” said Jeff Maggioncalda, Coursera CEO. “Our data shows these skills are not equally distributed, and students and low-wage workers need access to flexible, affordable, and fast-tracked pathways to entry-level digital jobs that offer a foundation for a stronger and more inclusive economy.” Key India insights Cloud Computing is India’s strongest technology skill, with 74 per cent proficiency. India now ranks 56th globally in the domain compared to the 66th rank last year.Slipping by 12 ranks, India scores low on foundational and specialised data science skills. The strongest skill in India continues to be machine learning at 41 per centWest Bengal is ranked the #1 state across the three domains – data, business and technologyAndhra Pradesh is one of the top 3 performing states with high proficiency in business and technology skills “Industry reports estimate that 28 million new technology jobs will be created in India by 2025. This year’s GSR signals a significant need for Indian learners to bridge the critical skills gap, especially in data science, to ensure this digital potential does not turn into a lost opportunity,” said Raghav Gupta, Managing Director, India and APAC, Coursera. “Strong industry-academia-government collaboration that focuses on the rapid deployment of high-quality digital and human skills training would be key to ensuring that the Indian workforce remains resilient and competitive amid rapid technological transformation.”","excerpt":"The strongest skill in India continues to be machine learning at 41 per cent","categories":["AI News"],"tags":["Courses"],"author_name":"Tasmia Ansari","publish_date":"2022-06-14T17:53:17","publication_year":"2022","word_count":425,"keywords":["data science","Go","API","machine learning","AWS","AI","cloud computing","Git","RAG","Courses","R"],"extracted_tech_keywords":["AI","machine learning","data science","RAG","cloud computing","AWS","R","Go","Git","API"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/cloud-computing-is-indias-strongest-technology-skill-coursera-report\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":27532,"title":"ICICI Lombard Launches India&#8217;s First AI To Automate Health Insurance Claims","content":"ICICI Lombard General Insurance has used artificial intelligence to provide instant health insurance claim approval. Reportedly, the system can scan the documents sent by the hospital and match them with the medical coverage. Earlier, this was a cumbersome process involving many persons: Doctors looked at the cases Executives entered the data and balance Checking the balance sum insured Going through the room rent limits for the insured This entire process would take over 60 minutes. Now, because of this new AI-powered system aided by OCR, the process can be carried out in just one minute. As of now, ICICI Lombard General Insurance is using AI for cataract, maternity, appendicitis, haemodialysis and hysterectomy. The company is now looking forward to expanding it to more sectors. Speaking to one of the leading newspapers, ICICI Lombard executive director Sanjeev Mantri said, “So far, AI has only been used in terms of chatbots doing simplistic service request. Understanding medical diagnosis is a much complex activity for which we have deployed AI technology. With AI over the medical admissibility decisions currently done by doctors, it will enable our doctors to focus on more complex jobs, including better fraud detection.” According to Mantri, the system is already up and running successfully. Using this system, he mentioned a case where they had sent a sanction to a patient in Visakhapatnam. He said that the patient was admitted to the hospital for an operation of appendicitis at 08:15 pm. The hospital used the online portal to submit the cashless request at 08:35 pm and the sanction was sent at 08:36 pm. So, the whole process of scrutinising was brought down drastically. The system does not work in case of a failure of the proposal acceptance. In such a case, an alert is received, and the papers are checked manually.","excerpt":"ICICI Lombard General Insurance has used artificial intelligence to provide instant health insurance claim approval. Reportedly, the system can scan the documents sent by the hospital and match them with the medical coverage. Earlier, this was a cumbersome process involving many persons: Doctors looked at the cases Executives entered the data and balance Checking the […]","categories":["AI News"],"tags":["AI (Artificial Intelligence)","Gen AI in Insurance","OCR"],"author_name":"Disha Misal","publish_date":"2018-08-22T11:54:05","publication_year":"2018","word_count":301,"keywords":["Go","artificial intelligence","OCR","AI","chatbots","RAG","Aim","ViT","Gen AI in Insurance","llm_models:Bard","R","AI (Artificial Intelligence)","fraud detection"],"extracted_tech_keywords":["AI","artificial intelligence","Aim","RAG","chatbots","fraud detection","R","Go","ViT","llm_models:Bard"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/icici-lombard-launches-indias-first-ai-to-automate-health-insurance-claims\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10165898,"title":"Agora Partners with Beken and Robopoet to Launch ConvoAI Device Kit","content":"Agora has launched the ConvoAI Device Kit, a new solution designed to enhance real-time, AI-powered voice interactions in smart devices. Developed in collaboration with chip-maker Beken, the toolkit aims to power interactive toys, robots, and IoT devices with ultra-low latency voice processing and intelligent dialogue capabilities. “Conversational AI is key to making robots and AI-driven toys truly interactive, engaging, and intuitive,” said Tony Zhao, CEO of Agora. “By integrating Agora’s technology into Beken’s chips, IoT toymakers like Robopoet will transform entertainment, education, and emotional support by enabling more natural conversations, adaptive learning, and personalised interactions that evolve with the user over time.” The ConvoAI Device Kit is already being used in Robopoet’s AI companion robot, Fuzzoo, unveiled at Mobile World Congress (MWC). Fuzzoo leverages Agora’s technology to deliver real-time emotional engagement through multimodal interactions. “Agora’s AI technology enables toys and robots to interact in a way that feels natural and engaging,” said Yuna Pan, Co-Founder and CTO of Robopoet. “With real-time voice processing, emotional AI, and advanced speech capabilities, Agora makes seamless human-machine interaction possible and ensures exceptional performance and reliability.” The solution supports a wide range of applications, from educational AI toys and smart home devices to AI-powered wearables. Pengfei Zhang, CEO of Beken, highlighted the collaboration’s impact, stating, “With Agora’s conversational AI technology and our optimised AI hardware, we’re enabling the next generation of devices to think, respond, and interact naturally.” Agora’s AI technology addresses key challenges in on-device conversational AI, such as background noise, latency issues, and rigid AI models. By integrating real-time speech synthesis and intelligent interruption handling, Agora is paving the way for more fluid and human-like interactions in smart devices. Earlier this week, Agora launched its Conversational AI engines, enabling developers to create interactive voice experiences with any AI model. These conversational engines are designed to provide ultra-low latency responses and superior voice processing capabilities while allowing real-time interruptions for a more natural dialogue experience.","excerpt":"Real-time voice processing and emotional AI make human-machine interaction more engaging.","categories":["AI News"],"tags":["Agora","Beken","ConvoAI","Robopoet"],"author_name":"Merin Susan John","publish_date":"2025-03-12T12:10:14","publication_year":"2025","word_count":321,"keywords":["Go","programming_languages:R","AI","Modal","ML","Agora","programming_languages:Go","RAG","Aim","Robopoet","Beken","R","ConvoAI"],"extracted_tech_keywords":["AI","ML","Aim","RAG","R","Go","Modal","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/agora-partners-with-beken-and-robopoet-to-launch-convoai-device-kit\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10076289,"title":"UAE’s Economic Ministry Will Now Enter The Metaverse","content":"In a project launched at Dubai’s Museum of the Future, it was announced that the UAE’s economic ministry is now setting up shop in metaverse. With this move, the country hopes to become a pioneer in the immersive virtual world. Officials stated that ‘those who use other means to venture within’ are likely to find a ministry open for business with companies and even ready to sign bilateral agreements with foreign governments. UAE’s economy minister Abdulla bin Touq Al Marri spoke at the inaugural Dubai Metaverse Assembly which was held at the museum. He concurred that the metaverse is an online world where users would eventually be able to work, game, and study but that it is still in a ‘test’ phase. UAE launched a bold Mars mission, ‘Hope Probe’ in July 2020 with the aim to study Martian weather dynamics. It was the fifth country to reach the Martian orbit after the US, Russia, China, the EU, and India. With a history of conducting such exemplary projects, that also include the 830-metre (2,723-foot) tall Burj Khalifa, UAE hopes that the metaverse would add $4 billion to its annual GDP and 40,000 jobs to its workforce by 2030. The assembly witnessed participation from representatives of major tech giants who connected with developers and entrepreneurs to explore the potential of the metaverse. Interacting with AFP, Al Marri said, “In the last couple of years we’ve seen investments, we’ve seen companies move in, and with the changes of the (visa) regime… we see talent coming in. We trained our employees to really immerse themselves in the metaverse, use the metaverse and engage with the Generation Z that is going to come.” In an attempt to become one of the world’s top ten metaverse economies, Dubai plans to attract 1,000 companies that specialise in blockchain and related technologies and anticipates being aided by eased visa rules for entrepreneurs, freelancers, and creatives. Meta, when asked about Dubai’s prospects of becoming a hub in metaverse, said, “If we look at the context of Dubai, there’s already a clear strategy and goals to accelerate metaverse adoption and investments in the building blocks of the metaverse.”","excerpt":"UAE hopes that the metaverse would add $4 billion to its annual GDP and 40,000 jobs to its workforce by 2030.","categories":["AI News"],"tags":[],"author_name":"Bhuvana Kamath","publish_date":"2022-10-03T13:49:07","publication_year":"2022","word_count":359,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","Aim","R"],"extracted_tech_keywords":["AI","Aim","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/uaes-economic-ministry-will-now-enter-the-metaverse\/","complexity_score":2,"technical_depth":6,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10093126,"title":"ChatGPT Reattempts UPSC","content":"In February this year, when we first experimented with the AI-chatbot ChatGPT attempting UPSC, which is widely regarded as one of the toughest exams in the world, it failed miserably – out of 100 questions (Paper 1), the chatbot could answer only 54 questions correctly. ChatGPT’s inability to pass the UPSC prelims became a source of pride for many aspirants. But, since we did that story, a lot of new updates and developments have happened in the world of AI. Most notably, OpenAI released GPT-4, which is the most advanced Large Language Model (LLM) to date. The previous version of ChatGPT was powered by GPT3.5, and a few months back OpenAI made GPT-4 accessible through ChatGPT Plus. Now, the Indian Civil Services prelims examination is also just around the corner. In this spirit, we thought, can GPT-4 clear UPSC? GPT-4 takes UPSC (Attempt 2) We conducted the same experiment again, but this time, we asked the same 100 questions to GPT-4, and this time, it got 86 questions right. Here, it’s important to note that Prelims consists of two papers-General Studies Paper-I and General Studies Paper-II, we stuck to Paper-I only in both cases (attempt 1 & 2). While the cut off for previous year (2021) was 87.54 marks, considering only paper 1, GPT-4 scored 162.76 marks, which would mean ChatGPT Plus (powered by GPT-4) cleared UPSC. In the previous experiment, ChatGPT gave 46 answers wrong, and in that terms, we have seen a huge improvement with GPT-4 as it got only 14 answers wrong. Having said that, this was not something completely unexpected either. When OpenAI released the technical paper of GPT-4, it did not mention any information about the architecture (including model size), hardware, training compute, dataset construction, training method etc, resulting in a furore among researchers. But interestingly, OpenAI did reveal that they tested GPT-4 on a diverse set of benchmarks, including simulating exams that were originally designed for humans. (Source: GPT-4 technical paper) In the technical paper, OpenAI also notes that GPT-4 outperforms GPT-3.5 ( ChatGPT) on most exams tested. Hence, it’s not surprising that GPT-4 scored better in UPSC than ChatGPT. Some interesting observations One of the reasons why ChatGPT got so many answers wrong was that it hallucinates. However, according to OpenAI, GPT-4 is more creative and less likely to make up facts than its predecessor. This has played an important role in GPT-4’s improved performance in the UPSC exam. When we made ChatGPT take the civil services exam the first time, we found that in certain cases, ChatGPT created its own alternatives. (ChatGPT response) But unlike its predecessor, we did not observe any similar occurrences with GPT-4. Nonetheless, GPT-4 does exhibit some level of hallucination, albeit to a much lesser degree. Additionally, many argued that one of the plausible reasons for ChatGPT’s failure to clear UPSC could be attributed to its training data. ChatGPT, along with GPT-4, has been trained on data only up to September 2021, thereby lacking knowledge about events that occurred after 2021. Hence, despite the limitation, GPT-4 did fairly well unlike its predecessor. Surprisingly, both models answered history-related questions incorrectly, despite it being an area where they are expected to perform well. ( GPT-4 response) Besides, it is important to note that it was just a fun experiment and no concrete judgments should be made based on these results. While GPT-4 cleared exams such as GRE and LSAT, it failed in English literature. Similarly, ChatGPT, despite having all the knowledge in the world, failed in an exam designed for a sixth grader. On an ending note, it’s also important to note that by altering the inquiry, we could prompt GPT-4 to arrive at accurate responses. This implies that in some instances, rephrasing the same question could lead GPT-4 to provide correct answers, and vice versa. However, in the experiment, only the bot’s initial responses were considered.","excerpt":"Interestingly, OpenAI claimed that GPT-4 outperforms GPT-3.5 ( ChatGPT) on most exams tested","categories":["Deep Tech"],"tags":[],"author_name":"Pritam Bordoloi","publish_date":"2023-05-11T12:00:00","publication_year":"2023","word_count":647,"keywords":["Go","ChatGPT","OpenAI","AI","programming_languages:R","programming_languages:Go","GPT","R","llm_models:ChatGPT","llm_models:GPT"],"extracted_tech_keywords":["AI","ChatGPT","OpenAI","R","Go","GPT","llm_models:GPT","llm_models:ChatGPT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/when-chatgpt-reattempts-upsc\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10167380,"title":"AI Wrote the Code—Now Developers are Stuck Fixing It","content":"It is a general belief that AI-driven tools help accelerate workflows for developers. While some developers may feel like Lightning McQueen, a fictional character in Pixar’s Cars franchise known for racing, others are encountering more bumps and roadblocks than anticipated. In a blog post, Birgitta Boeckeler, global lead for AI-assisted software delivery at Thoughtworks, explained how she had to intervene, correct, and steer the flow of AI-driven development while using Cursor, Windsurf, and Cline. That slowed down the speed of the development process. So, what are the takeaways from her findings? How do other developers feel about the same issue? More broadly, can AI-assisted coding sometimes hinder productivity rather than enhance it in certain contexts? Where the Problem Occurs Boeckeler began by expressing her appreciation for the recent progress in IDE integration. She noted how those integrations have made it easier and faster to get things done—from executing tests and fixing errors on the spot, to conducting web research, and more. As she progressed through the AI-assisted coding sessions, Boeckeler found herself frequently needing to guide the tools and correct their suggestions, and sometimes, discarding the changes proposed by the AI altogether. So much so that she mentioned, “We’re still far away from AI writing code autonomously for non-trivial tasks.” Boeckeler explained the problem by categorising it into three areas of impact. First, it slowed down the development speed and the time to commit. Second, it created friction for the team flow in that iteration. Lastly, it affected the long-term maintainability of the code.She also shared a visual highlighting the varying levels of impact each of these limitations had on her workflow. Coding and Commit Time To begin with, Boeckeler faced a common issue: receiving code that simply did not work. She then had to decide whether she would do it herself or start a new AI session to retry, which could potentially take more time. In addition, it was observed that AI misdiagnoses a problem. She shared an example where the AI assumed that a Docker build issue was due to architecture settings. In reality, however, the issue originated from copying the wrong packages built for the wrong architecture. Koushik Bhargava, a member of technical staff at Salesforce, told AIM, “I have experienced AI hallucinations a few times while trying to resolve some issues, wasting more time in the process. Often, I was able to resolve those issues with one simple Google search.” Breaking the Team Flow As per Boeckeler’s observation, AI focuses on the overall objective instead of focusing on building blocks of functionalities. She highlighted that this risks wasting large upfront work before realising the problem. Moreover, AI has been noticed to take a brute-force approach to solving issues rather than analysing the root cause. This causes a delay in the solution coming into effect. An example included a scenario where the AI encountered a memory error during a Docker build. To resolve this, it increased the memory settings rather than investigating the problem of the error. She added that the developer workflow can become more complicated than it should be at times, sometimes due to incomplete context in human prompts. Long-term Maintainability AI-generated code does not focus on modularity, which makes its reusability difficult. This could be because AI does not always have the entire context of the project, potentially making it a problem for future code maintenance. In some cases, AI generates overly complex and unnecessary code. In contrast, simpler code makes it easy to maintain. So, a Speed Boost With a Catch? In an exclusive chat with AIM, Mehul Gupta, data scientist at DBS Bank, said, “AI coding has definitely sped up my workflow. Boilerplate code? Done in seconds. It cuts down development time massively.” “After years of programming, I’ve developed a sixth sense for where AI tends to slip up . So the AI handles the grunt work, and I fine-tune the details,” he added. Gupta stressed that even with the benefits, there is a trade-off that he noticed—the time saved in writing the code often gets reallocated to reviewing. “AI isn’t great with edge cases, and in complex projects, it can introduce subtle, hard-to-spot errors. The risk jumps when I am working with an unfamiliar language or framework - AI-generated bugs can be sneaky and a nightmare to debug later.” He further explained that AI coding tools are a significant productivity multiplier for experienced developers. Yet, beginners often find themselves in a frustrating loop of trial and error, lacking the foundational knowledge to leverage them effectively. He pointed out that while AI enhances existing programming skills, it doesn’t replace them.","excerpt":"“I have experienced AI hallucinating a few times when I try to resolve some issues, wasting more time in the process.”","categories":["AI Features"],"tags":["AI coding"],"author_name":"Ankush Das","publish_date":"2025-04-07T15:41:34","publication_year":"2025","word_count":768,"keywords":["Go","AI coding","AI","docker","Git","RAG","GRU","Aim","Rust","GAN","R"],"extracted_tech_keywords":["AI","Aim","RAG","docker","R","Go","Rust","Git","GAN","GRU"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/ai-wrote-the-code-now-developers-are-stuck-fixing-it\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10056137,"title":"What Is The Whole Controversy Of Log4Shell","content":"On the 10th of December, the tech industry was taken by a storm when it was discovered that the Remote Code Execution CVE-2021-44228 in the popular Java logging library Log4J (all versions between 2.0 and 2.14.1) have become vulnerable. The software library helps developers keep track of changes in the applications they build – for example, to text files. The Log4Shell 0-day attack vulnerability was first noticed in Minecraft; it has far-reaching consequences considering the ubiquitous character of the Log4J library, rendering millions of applications vulnerable to attack. This vulnerability can be exploited by attackers allowing them to take control of the vulnerability from a remote server. According to Acronis VP of Cyber Protection Research, ​​Candid Wuest “The Log4shell vulnerability in Log4j is definitely in the top-5 most severe vulnerabilities of the last decade, one that allows for remote code execution (RCE). It compares to the EternalBlue used by WannaCry, or the ShellShock Bash vulnerability. What makes it so serious is how simple it is to exploit it remotely, as well as the huge number of applications using it. In addition, it also takes longer to patch – as it’s not just one vulnerable software that can be updated, but rather a library that’s included in many applications, resulting in many different updates that need to be installed.” Any application that uses the Java logging library is at risk due to the vulnerability, including cloud applications like Steam, Apple iCloud, etc. Additionally, according to a Lunasec researcher, simply changing an iPhone’s name can trigger the vulnerability in Apple’s servers. Candid Wuest has further said that the list of affected applications is still growing, as companies complete their analysis – affected applications already include Minecraft, Blender, LinkedIn, VMware and many more. The vulnerability can lead to: Service disruption to having malware executedData breach Exfiltrate sensitive dataGain initial access to systems An increase in these attacks could lead to a hike in data breaches, to new computers being added to botnets for future attacks. According to experts, the vulnerability can enable hackers to control java-based web servers and enable them to execute remote code execution (RCE) attacks, which they may use to take control of affected systems. As the industry scrambles to mitigate the vulnerability, ​​Vice President Product Management, Ivanti, Chris Goettl, has mentioned some effective measures that can be adopted: “As far as how organisations should be looking to resolve this vulnerability, that is a bit more tricky. Normally, an organisation would rely on code scanners to identify the vulnerable code component or library. In this case, code scanners are still racing to catch up and properly detect the vulnerable library. For products already released to the market, an organisation would rely on its network vulnerability scanning to identify vulnerable software, but those scanners are having trouble consistently detecting the vulnerability as they have to try and send a properly formed message and monitor the logs for results, which may not consistently show up. The best guidance is to continue to rely on your DevSecOps processes and vulnerability scanning and supplement this with more direct action as there will likely be gaps for some time in detection. There are a few sources gathering lists of KB articles, security advisories, and mitigation guidance by vendors. Your organisation should assess the vendors in your environment, determine if they have provided guidance, and take those actions immediately.” Furthermore, Apache Foundation has updated Log4j version 2.15.0 — released on December 6, 2021– to address the vulnerability. As the update did not fully address the vulnerability, Apache Foundation released version 2.16.0, which eradicates the vulnerability completely and adds to the development team backlogs to update material sections on their codebase that handle logging. Companies must update the system as soon as possible in order to avoid malicious attacks on their softwares.","excerpt":"This article is a deep dive into the Log4Shell vulnerability and expert suggestions on how to mitigate it.","categories":["AI Features"],"tags":[],"author_name":"Abhishree Choudhary","publish_date":"2021-12-17T16:00:00","publication_year":"2021","word_count":635,"keywords":["Go","programming_languages:R","AI","programming_languages:Java","programming_languages:Go","disruption","GAN","R","Java"],"extracted_tech_keywords":["AI","R","Go","Java","GAN","disruption","programming_languages:R","programming_languages:Java","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/what-is-the-whole-controversy-of-log4shell\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10061441,"title":"Council Post: Individual artificial intelligence is the future of AI","content":"Elon Musk-backed Neuralink is teeing up for clinical trials on humans with a view to accomplish the human brain implant by the end of the year. However, this is just the beginning. Remember the movie Transcendence where Johnny Depp (Dr Will Caster) turns into a superintelligent AI. Well, the timeline to superintelligence is shortening. Last December, a 63-year-old Australian Philip O’Keefe, who is suffering from ALS, tweeted his thoughts with the help of a computer chip implanted in his brain. I’m excited to announce a world first. At 7pm ET today (~5 hours) Phillip O’Keefe, who had a Synchron Brain Computer Interface implanted in April 2020 will take over my Twitter account for 30 mins. He will use the BCI to tweet answers to your questions: directly from his brain. pic.twitter.com\/FzxUImgZC3— Thomas Oxley (@tomoxl) December 22, 2021 In July 2021, the researchers from the University of California, San Francisco (UCSF) leveraged AI to help a paralysed man communicate by translating his brain signals into computer-generated writing. The researchers used a neuroprosthetic device to track his brain signals. As of now, the commercial use cases of AI chips or brain-computer interface (BCI) are few and far between. Currently, BMI is mostly used to help paralysed people interact. However, some scientists and philosophers warn of the dangers of an insubordinate superintelligent AI. Earlier, the Centre for Humans and Machines at the Max Planck Institute for Human Development demonstrated it would be impossible to control a superintelligent AI. Enter, individual artificial intelligence Now, AI and machine learning are used to control cars, compose music, or beat the world champions in Go, Chess, esports game (Dota), etc. Now, the question is, is it possible to seamlessly integrate the human brain with AI? If we look at brain synapses, the structural properties of transmembrane proteins in channels allow human intelligence to perform as a classic binary system-symbiosis. In the future, scientists could build a highly unique artificial intelligence, aka individual artificial intelligence that directly connects the neurons in the human brain to a computer. At present, all the AI systems have one thing in common – they are developed as vertically controlled electronic complexes that operate using algorithms\/code of varying complexity. In individual artificial intelligence, the AI is turned into a bio-electronic crossover, in which a machine and human mind will cooperate in a double integral framework. How is it different Compared to electronic systems, the processing power of the human brain is limited. However, the plasticity, creativity, and energy efficiency of the human brain are difficult to simulate. Computer systems process information faster than humans, with 3 million times signal transmission speed than us. When you combine the human brain and computer into a single system, it will increase the overall efficiency and create a new type of artificial intelligence in the process. A final thought While the traditional model of neural computer interface based on surgical implants is simpler than the idea of transmembrane proteins at synapses, research is underway to make the latter a reality. Today, everyone expects an engineer to build a strong AI, but in the days to come, a doctor or a neurologist will probably inspire the next big breakthrough in AI. This article is written by a member of the AIM Leaders Council. AIM Leaders Council is an invitation-only forum of senior executives in the Data Science and Analytics industry. To check if you are eligible for a membership, please fill out the form here.","excerpt":"The plasticity, creativity, and energy efficiency of the human brain are difficult to simulate.","categories":["AI Features"],"tags":["brain-computer interface","neuralink"],"author_name":"Anish Agarwal","publish_date":"2022-02-24T11:03:53","publication_year":"2022","word_count":577,"keywords":["data science","Go","artificial intelligence","machine learning","AI","ML","neuralink","RAG","brain-computer interface","Aim","analytics","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","data science","analytics","Aim","RAG","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/council-post-individual-artificial-intelligence-is-the-future-of-ai\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10011044,"title":"Understanding XGBoost Algorithm In Detail","content":"XGBoost or extreme gradient boosting is one of the well-known gradient boosting techniques(ensemble) having enhanced performance and speed in tree-based (sequential decision trees) machine learning algorithms. XGBoost was created by Tianqi Chen and initially maintained by the Distributed (Deep) Machine Learning Community (DMLC) group. It is the most common algorithm used for applied machine learning in competitions and has gained popularity through winning solutions in structured and tabular data. It is open-source software. Earlier only python and R packages were built for XGBoost but now it has extended to Java, Scala, Julia and other languages as well. In this article, I’ll be discussing how XGBoost works internally to make decision trees and deduce predictions. To understand XGboost first, a clear understanding of decision trees and ensemble learning algorithms is needed. Difference between different tree-based techniques: XGBoost falls under the category of Boosting techniques in Ensemble Learning. Ensemble learning consists of a collection of predictors which are multiple models to provide better prediction accuracy. In Boosting technique the errors made by previous models are tried to be corrected by succeeding models by adding some weights to the models. Basic Boosting Architecture: Unlike other boosting algorithms where weights of misclassified branches are increased, in Gradient Boosted algorithms the loss function is optimised. XGBoost is an advanced implementation of gradient boosting along with some regularization factors. Source: link Features of XGBoost: Can be run on both single and distributed systems(Hadoop, Spark).XGBoost is used in supervised learning(regression and classification problems).Supports parallel processing.Cache optimization.Efficient memory management for large datasets exceeding RAM.Has a variety of regularizations which helps in reducing overfitting.Auto tree pruning – Decision tree will not grow further after certain limits internally.Can handle missing values.Has inbuilt Cross-Validation.Takes care of outliers to some extent. XGBoost Algorithm Let’s look at how XGboost works with an example. Here I’ll try to predict a child’s IQ based on age. For any basic assumption in such statistical data, we can take the average IQ and find how much variance(loss) is present. Residuals are the losses incurred will be calculated after each model predictions. CHILD’s AGECHILD’s IQRESIDUALS1020-101534416388 So the average of 20, 34, and 38 is 30.67 for simplicity let’s take it as 30. If we plot a graph keeping y-axis as IQ and x-axis as Age and then we can see the variance in points from the average mark. At first, our base model(M0) will give a prediction 30. As from the graph, we know this model suffers a loss which will have some optimisation in the next model(M1). Model M1 will have input as age(independent features) and target as the loss suffered(variances) in M0. Until now it is the same as the gradient boosting technique. For XGboost some new terms are introduced, ƛ -> regularization parameter Ɣ -> for auto tree pruning eta -> how much model will converge Now calculate the similarity score, Similarity Score(S.S.) =  (S.R ^ 2) \/ (N + ƛ) Here, S.R is the sum of residuals, N is Number of Residuals At first let's put  ƛ =0, then Similarity Score = (-10+4+8)^2 \/ 3+0 = 4\/3 = 1.33 Let’s make the decision tree using these residuals and similarity scores. I’ve set the tree splitting criteria as Age >10. Again for these two leaves, we calculate the similarity scores which is 100 and 72. After this, we calculate the Gain = S.S of the branch before split - S.S of the branch after the split. Gain = ( 100 + 72 ) – 1.3 Now we set our Ɣ, which is a value provided to the model at starting and its used during splitting. If Gain>Ɣ then split will happen otherwise not. Let’s assume that Ɣ for this problem is 130 then since the gain is greater than Ɣ, further split will occur. By this method, auto tree pruning will be achieved. The greater the Ɣ value more pruning will be done. For regularization and preventing overfitting, we must increase the ƛ  which was initially set to 0. But this should be done carefully as greater the ƛ  value lesser the Similarity score, lesser the gain and more the pruning. New prediction = Previous Prediction + Learning rate * Output XGboost calls the learning rate as eta and its value is set to 0.3 For the 2nd reading(Age=15) new prediction = 30 + (0.3 * 6) = 31.8 The outcome is 6 is calculated from the average residuals 4 and 8. New Residual = 34 – 31.8 = 2.2 Age IQ Residual1020-715342.216386.2 This way model M1 will be trained and residuals will keep on decreasing, which means the loss will be optimized in further models. Conclusion XGboost has proven to be the most efficient Scalable Tree Boosting Method. It has shown outstanding results across different use cases such as motion detection, stock sales predictions, malware classification, customer behaviour analysis and many more. The system runs way faster on a single machine than any other machine learning technique with efficient data and memory handling. The algorithm’s optimization techniques improve performance and thereby provides speed using the least amount of resources.","excerpt":"In this article, I’ll be discussing how XGBoost works internally to make decision trees and deduce predictions.","categories":["Deep Tech"],"tags":["Boosting Algorithm","decision tree algorithm","Ensemble Learning","gradient boosting","gradient descent","hadoop problems","Machine Learning Algorithms","XGBoost"],"author_name":"Jayita Bhattacharyya","publish_date":"2020-11-02T11:00:27","publication_year":"2020","word_count":845,"keywords":["Machine Learning Algorithms","Go","Boosting Algorithm","machine learning","decision tree algorithm","TPU","AI","ML","Ensemble Learning","RAG","hadoop problems","Python","gradient boosting","XGBoost","R","Java","gradient descent"],"extracted_tech_keywords":["AI","machine learning","ML","XGBoost","RAG","TPU","Python","R","Go","Java"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/xgboost-internal-working-to-make-decision-trees-and-deduce-predictions\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10070309,"title":"It’s high time SpaceX entered India","content":"Last week, ISRO launched its PSLV-C53 mission from Satish Dhawan Space Centre. This is the second dedicated commercial mission of its commercial arm NewSpace India Limited (NSIL). NSIL is a wholly-owned subsidiary of Central Public Sector Enterprise (CPSE), led by the Department of Space (DOS). PSLV-C53\/DS-EO Mission: The launch would be streamed LIVE on ISRO website https:\/\/t.co\/MX54Cx57KU or ISRO Official Youtube channel (https:\/\/t.co\/1qTsZMZXU3) from 17:32 hours IST on June 30, 2022— ISRO (@isro) June 29, 2022 The launch vehicle carried three satellites from Singapore. This includes DS-EO and NeuSAR, belonging to Singapore and built by Starec Initiative, Republic of Korea. The third is a 2.8 kg Scoob-1 of Nanyang Technological University (NTU) Singapore. Besides this, Bengaluru-based space tech startup Dhruva’s payload Dhruva Space Satellite Orbital Deployer and Hyderabad-based Digantara’s payload ROBust Integrating Proton Fluence Meter (ROBI) were also authorised for the launch. These payloads will fly onboard the PSLV Orbital Experimental Module (POEM) of PSLV-C53. India’s space tech landscape is on the verge of exploding. The advancements in the last few years, both from a regulatory standpoint as well as research and development, are fueling the industry to a whole new level – now attracting interests from space tech giants like SpaceX, Blue Origin, Virgin Galactic and startups and investors alike to explore opportunities in India for talent and scale. India opens doors to private players In a historical turn of events, this is the first time that the private players are allowed to launch payloads to space. Previously, only ISRO was permitted to launch in space. Only recently, India announced that it opened up the space sector for private players. The Indian Space Promotion and Authorisation Centre (IN-SPACe) authorised two Indian startups to launch their payloads in space – namely, Dhruva Space and Digantara, and many more to follow in the coming months. Founded in 2012, Dhruva Space was started by former-Exseed Space, AMS AG, Cisco & KPMG employees Krishna Teja Penamakuru and Sanjay Nekkanti. The company claims to offer faster, affordable and reliable satellite platforms, launch solutions, and ground services. Nekkanti said they are excited to lead India’s private industry-led space revolution. Meanwhile, Digantara, an Indian space technology company incubated at the Indian Institute of Science (IISc), was founded by Anirudh Sharma, Rahul Rawat, and Tanveer Ahmed. The company claims to have built the world’s first space-based active orbital surveillance platform and offers holistic solutions for sustainable space operations. Pawan Kumar Goenka, IN-SPACe chairman, said that the first two launch authorisations issued by the authority are an important milestone. It marks the beginning of private space sector launches in India. Delighted that IN-SPACe issued the first two authorisations to Private Sector for Space activities. Best wishes to Dhruvaspace and Digantara for a successful launch of their Payload aboard the C53 mission of ISRO on June 30th. @INSPACeIND @isro @PMOIndia pic.twitter.com\/Xj5TliQ0pU— Pawan K Goenka (@GoenkaPk) June 26, 2022 IN-SPACe In June 2020, the Indian government launched IN-SPACe to increase private participation in the space sector. Both ISRO and DOS are expected to complement each other in their objective of opening up the space sector to private industries. Here’s a glimpse of proposed reforms: The industry body proposed to change the approach from the ‘supply-based model’ to the ‘demand-based model’ to enhance the use and maximise benefits from the space assets. NSIL will act as an aggregator of user requirements and obtain commitments. NSIL will be taking ownership from DOS for launching vehicles, commercialising launches, satellites, and services. Private companies can carry out space activities through IN-SPACe. ISRO will be carrying out capacity building in the space sector through developing new technologies and capabilities and enabling sharing of facilities by NSIL and NGPEs (non-government private entities). Why now? Traditionally, the Indian industry was unable to compete because, till now, its role was mainly that of suppliers of components and subsystems. At the same time, Indian companies do not have the resources or the technology to undertake independent space projects of the kind that US companies such as SpaceX, Blue Origin and others have been doing or provide space-based services. Moreover, with the demand for space-based applications and services increasing significantly within the country, ISRO alone cannot cater to this demand. There is also a shortage of satellite data, imageries, and space technology which is required across sectors, including agriculture, logistics, urban development, disaster management, and others. This calls for greater expansion of space technologies, better use, and increased need for space-based services. Many Indian companies are using these opportunities, and a few handfuls of companies are developing their launch vehicles, similar to the ISRO’s PSLV that carry the satellites and other payloads into space. For instance, Jio, Bharti Airtel and Starlink are investing in satellite internet technology in India. In addition, the Indian Space Association is pushing the centre to provide subsidies to satellite communication (Satcom) providers through the Universal Service Obligation Fund (USOF). India’s space tech prowess Some of the factors that companies look into while launching their satellites include per kg launch cost and cost per launch. SpaceX charges around USD 1.1 million per kg of payload on its Falcon 9 rocket. Amazon’s Blue Origin’s full-size research payload lockers cost between USD 50,000 and USD 120,000, and the smaller payload spaces cost as little as $8,000. ISRO charges about USD 25,000 per kg of payload on its PSLV rocket. However, with India expected to launch a low-cost, easy-to-build, 500 kg payload class small satellite launch vehicle (SSLV), it is expected that the per kg launch cost will reduce to an extent, thus making it an attractive option for customers. As per reports, the development cost of SSLV in India is said to be about INR 169.07 crore (USD 22 million), and the manufacturing cost is expected to be INR 30 crore (USD 3.7 million) to INR 35 crore (USD 4.4 million). The high launch rate relies on largely autonomous launch operations and logistics. Interestingly, these are not even the reusable launch vehicles that companies like SpaceX and Blue Origin flaunt. For instance, the cost per launch of SpaceX’s Falcon 9 is around USD 67 million (INR 528 crores). Also, the Mars mission, launched in 2013, had cost ISRO INR 470 crore (USD 59 million), cheaper than the Hollywood space movie – Gravity, which had a budget of INR 644 crore (USD 81 million). Meanwhile, the Chandrayaan 2 mission’s cost was only INR 800 crore, compared to the sci-fi movie Interstellar movie budget, which was around INR 1,062 crore. But, the question is, how did they reduce the cost? Sivan said that simplifying the system, miniaturising the large complex system, strict quality control and maximising output from a product make ISRO’s space missions frugal and cost-effective. In addition, he said that they keep a strict vigil on every stage of development of a spacecraft or a rocket, and therefore, they can avoid wastage of products, which helps them minimise the mission cost. Will SpaceX, Blue Origin, enter India? With rising inflation in the US, the space tech giant SpaceX recently raised prices across its products and services, including rocket launches and Starlink satellite internet. Reports show that the price hikes range from as much as a 20 per cent increase for new Starlink hardware to 8 per cent for launches with the company’s Falcon 9 and Falcon Heavy rockets. With the rupee falling to an all-time low against the US dollar and India opening arms for private players to venture into space tech, there can not be a better time for customers as well as private players in the space tech landscape. As per reports, India has about 2-3 per cent share of the global space market, which is expected to touch USD 423 billion by 2030. The country’s share will likely expand by 48 per cent CAGR by 2030. Currently, only 2 per cent of this market is for rocket and satellite launch services, which require fairly large infrastructure and investment. The remaining 95 per cent is related to satellite-based services and ground-based systems. With the new reforms in place, the participation of private players can uplift the whole segment to the next level. Some notable startups working with IN-SPACe include Dhruva, Skyroot, Agnikul, Digantara and Astrome. Besides these, other startups in the Indian space tech ecosystem include Kawa Space, Bellatrix Aerospace, and Aantriksh, among others. Recently, prime minister Narendra Modi, addressing the 90th episode of ‘Mann Ki Baat,‘ said that there are 100 space startups currently working in the space sector, alongside addressing the pressing issues faced by the space tech industry. While there is so much focus on space tech in the country, the question remains if global players like SpaceX, Blue Origin, Virgin Galactic and other players can set up their base in India and partner with the government to boost the segment to a new level? SpaceX in dilemma In May 2022, SpaceX announced that its Starlink satellite internet services would be available in 32 countries around the world. In India, its services are currently not available, but the blue colour on the map says ‘coming soon.’ Starlink is now available in 32 countries around the world. People ordering from areas marked “available” will have their Starlink shipped immediately → https:\/\/t.co\/slZbTmHdml pic.twitter.com\/CecM1pkf5D— SpaceX (@SpaceX) May 13, 2022 Last year, SpaceX faced a lot of regulatory challenges in the country after the company had claimed to have received over 5K pre-orders from India. It is noticed that M\/s #Starlink (https:\/\/t.co\/xscnDS4Cnn) has started pre-selling\/booking of #satellite based #Internet Services in India without any #license\/authorization. Public is advised not to subscribe to Starlink services being advertised. #GatiShakti #spacex @SpaceX— DoT India (@DoT_India) November 26, 2021 In November last year, SpaceX had registered a fully owned subsidiary called Starlink Satellite Communications, listing Sanjay Bhargava and Anand Prasad as signatories. But, in January 2022, Bhargava stepped down from the company. A month before that, he had announced that it would apply for a commercial licence – i.e. before the end of January 2022. As per the timeline, Starlink was supposed to start commercial rollout in April 2022 and set up 200K terminals in India by December 2022. “Currently, there are unknowns on when you will get delivery of your Starlink,” mentioned Starlink in its planning document.","excerpt":"In June 2020, the Indian government launched IN-SPACe to increase private participation in the space sector.","categories":["AI Features"],"tags":["Blue Origin","Earth Observation Satellites","spacex"],"author_name":"Amit Naik","publish_date":"2022-07-04T10:00:00","publication_year":"2022","word_count":1707,"keywords":["Go","TPU","Earth Observation Satellites","Blue Origin","AI","ML","spacex","Ray","Aim","ViT","GAN","CNN","R"],"extracted_tech_keywords":["AI","ML","Aim","Ray","TPU","R","Go","GAN","CNN","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/its-high-time-spacex-entered-india\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10167502,"title":"Google Taps Accenture, Deloitte, Capgemini, KPMG for Agentic AI Push","content":"At Cloud Next ‘25 in Las Vegas, Google announced a wave of strategic collaborations with top global consulting firms — Accenture, Capgemini, Deloitte, and KPMG. The aim is to accelerate enterprise adoption of agentic AI technologies using its Gemini models, Agentspace, and the newly launched Agent2Agent (A2A) interoperability protocol. Accenture introduced a comprehensive suite of new offerings in agentic AI, cloud, networking, and mainframe modernisation to assist clients in unlocking business intelligence and driving innovation. Notably, the company is launching industry-specific agent accelerators through its joint GenAI Centre of Excellence with Google. These consist of AI-driven content agents for Mondelez International, digital financial assistants for C6 Bank, and enhanced forecasting for CNA Insurance. “Accenture is continuing to double down on its strategic partnership with Google Cloud, continuously investing in new capabilities to help our clients accelerate reinvention and spur innovation,” said Scott Alferi, senior managing director at Accenture. “We are helping clients build an agile, cloud-based digital core, integrating advanced data and AI technologies.” Meanwhile, Capgemini deepened its alliance with Google Cloud to develop tailored agentic AI solutions focused on telco, retail, and financial services, with plans to expand into life sciences and utilities. Leveraging Agentspace and the Customer Engagement Suite, Capgemini aims to transform customer experiences by optimising workflows and enabling proactive support. “By understanding the potential of agentic AI and the business realities of our clients, we’re expertly placed to maximise its value and deliver genuine impact,” said Fernando Alvarez, chief strategy and development officer at Capgemini. Google’s Kevin Ichhpurani added that the partnership will deliver “AI solutions that drive long-term value across industries”. Also, Deloitte announced its largest investment yet with Google Cloud by introducing over 100 ready-to-deploy AI agents built on Agentspace and Gemini. These agents aim to transform customer and employee experiences across industries. The company is also co-developing the Agent2Agent protocol with Google and ServiceNow, which will enable interoperability among AI agents across platforms and cloud providers. “Clients are getting flooded with information about agents… With collaborations like Google Cloud and ServiceNow, we aim to be a one-stop shop,” said Jason Salzetti, chair and CEO of Deloitte Consulting LLP. Deloitte claims a productivity boost of over 30% in areas like contract redlining, credit loan automation, and technical data transformation. KPMG, too, is building AI-powered customer solutions and internal tools to improve productivity and employee experiences, rounding out Google’s broader push to embed agentic AI across enterprise ecosystems. By building on Google Cloud’s AI innovations, KPMG aims to deliver AI agents that not only respond to customer needs across multiple engagement channels but also support consultants and analysts with real-time data insights, improving service delivery and reducing turnaround times. “We’re enabling clients to put the right AI and agent capabilities in place today to create lasting value,” said Steve Chase, vice chair of AI and digital innovation, KPMG LLP.","excerpt":"Google’s Kevin Ichhpurani says the partnership will deliver AI solutions that drive long-term value across industries.","categories":["AI News"],"tags":["Accenture"],"author_name":"Mohit Pandey","publish_date":"2025-04-09T17:30:00","publication_year":"2025","word_count":474,"keywords":["Accenture","Go","GenAI","agentic AI","AI","Git","RAG","automation","Aim","ViT","R"],"extracted_tech_keywords":["AI","GenAI","agentic AI","Aim","RAG","R","Go","Git","ViT","automation"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/google-taps-accenture-deloitte-capgemini-kpmg-for-agentic-ai-push\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":52907,"title":"AWS Launches Braket, A Quantum Computing As A Service","content":"Tech giants like Google and IBM have already marked their presence in the domain of quantum computing. These companies have been working on quantum computing for quite a few years now and have created a lot of buzz in this field. At the beginning of this year, IBM has unveiled its first-ever circuit-based commercial quantum computer known as IBM Q System One at the 2019 Consumer Electronics Show (CES). The company has also announced to open its first IBM Q Quantum Computation centre for commercial clients in New York. Recently, Google has also made a breakthrough by announcing that the company has achieved quantum supremacy. The researchers published a scientific journal where they discussed the efforts and techniques to build a quantum computer which has the ability to perform tasks which no other classical computer can do. On the other hand, instead of creating a buzz around the next big thing, e-commerce giant, Amazon has silently placed its cards on the table. Understanding the potential to solve computational problems that are beyond the reach of classical computers by harnessing the laws of quantum mechanics to build more powerful tools for processing information, the cloud computing platform of this e-comm giant joined hands with D-Wave, IonQ and Rigetti and unveiled its all-new quantum computing service known as Braket. A few days ago, AWS officially announced the preview launch of Braket, its first-ever quantum computing service. The term Braket is derived from “bra–ket notation” which is a common notation for quantum states in quantum mechanics. Amazon’s Braket is a fully managed service that helps to get started with quantum computing by providing a development environment to explore and design quantum algorithms, test them on simulated quantum computers, and run them on the choice of different quantum hardware technologies. How Braket Helps Gaining access to quantum computing hardware to run the algorithms and optimise designs can be expensive and inconvenient. Also, programming quantum computers to solve a problem requires a new set of skills. Amazon Braket helps in overcoming these difficulties by providing a service that lets developers, researchers, and scientists explore, evaluate, and experiment with quantum computing. This service allows users to choose from a variety of quantum computers which include gate based superconductor computers from Rigetti, quantum annealing superconductor computers from D-Wave, and ion trap computers from IonQ. Braket also allows users to design their own quantum algorithms from scratch or to choose from a set of pre-built algorithms. Once the algorithm gets defined, Amazon Braket provides a fully managed simulation service to help in troubleshooting and verifying the implementation. A user can then run the algorithm on any of the above-mentioned quantum computers. Furthermore, in order to make it easier for the users to develop a hybrid algorithm which is a combination of both classical and quantum tasks, Amazon Braket helps in managing classical compute resources and establish low-latency connections to the quantum hardware. Quantum Computing In India Currently, in India, a lot of things are going around in the domain of quantum computing, with the government and private organisations both simultaneously trying to make advancements in this field. Recently, C-DAC (Centre for Development of Advanced Computing), a national premier R&D organisation under Ministry of Electronics and Information Technology, Government of India, and Atos — a noted digital transformation company, has announced their Cooperation Agreement for technology advancement in the areas of quantum computing, artificial intelligence, and exascale computing. Also, a few months ago, the Indian Institute of Technology Roorkee (IIT Roorkee) has also partnered with Microsoft Garage India and introduced a full semester elective course on quantum computing. The course will provide access to Q# Programming Language practical examples, Microsoft Quantum Development Kit, and Microsoft Quantum Faculty.","excerpt":"Tech giants like Google and IBM have already marked their presence in the domain of quantum computing. These companies have been working on quantum computing for quite a few years now and have created a lot of buzz in this field.  At the beginning of this year, IBM has unveiled its first-ever circuit-based commercial quantum […]","categories":["IT Services"],"tags":["Amazon AWS","AWS","Quantum Computing","quantum computing companies","quantum computing programming","quantum mechanics","quantum supremacy","the quantum companies","what companies use quantum computers"],"author_name":"Ambika Choudhury","publish_date":"2019-12-31T19:00:00","publication_year":"2019","word_count":618,"keywords":["Quantum Computing","Go","quantum supremacy","what companies use quantum computers","artificial intelligence","AWS","AI","cloud computing","R","digital transformation","quantum computing companies","Git","the quantum companies","RAG","Amazon AWS","quantum computing programming","GAN","quantum mechanics"],"extracted_tech_keywords":["AI","artificial intelligence","RAG","cloud computing","AWS","R","Go","Git","GAN","digital transformation"],"url":"https:\/\/analyticsindiamag.com\/it-services\/aws-launches-braket-a-quantum-computing-as-a-service\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":32369,"title":"Google India Now Accepting Applications For Launchpad Accelerator, Invites AI, ML Startups","content":"Reemphasising its assurance to the Indian startup ecosystem, Google opened applications for the next class of its Launchpad Accelerator India chapter on Wednesday, inviting startups that are using AI\/ ML to solve for India’s requirements. The Launchpad Accelerator mentorship programme for start-ups using Artificial Intelligence (AI)\/ Machine Learning (ML)India chapter is scheduled to commence in March 2019. The last date for application to the programme is January 31, 2019, and The shortlisted start-ups for the second batch will be announced in February 2019 and will receive Google Cloud assets from $20,000 – going up to $100,000 each. The selected start-ups will attend the boot-camp in Bengaluru in March and there will be follow-up appointments over April and May to help approach their distinct challenges, Google said. The Launchpad Accelerator programme empowers, startups, that are applying AI\/ML to solve for India’s requirements, experience an intensive in-person mentorship boot-camp, followed by customized support for three months. The selected startups for class-II will be announced in February 2019, and the program will begin in March 2019. The shortlisted startups will attend the boot-camp in Bangalore in March and there will be follow-up engagements over April and May to help address their specific challenges. The 10 startups from the first batch have now entered scores of innovative tech startups in India that have been bolstered by the Launchpad programme over the years. With the second batch, Google aims to expand its mentorship momentum, grow the AI\/ML ecosystem in India and help Indian startups build scalable solutions. In extension to the tailored one-to-one mentorship, some of the approaches in which the first class benefited from the program incorporate — insights and support on advanced technology such as ML, in-depth Design Sprints for individually identified challenges, supervision on focused tech projects, characteristics on the Google Play store, networking opportunities and connects at industry events and platform. “Our Launchpad Accelerator programme is bringing best of our expertise, platforms, tools and core strengths including Machine Learning and AI, to help Indian start-ups build, scale and grow their offering,” said Paul Ravindranath, Product Manager, Launchpad Accelerator India.","excerpt":"Reemphasising its assurance to the Indian startup ecosystem, Google opened applications for the next class of its Launchpad Accelerator India chapter on Wednesday, inviting startups that are using AI\/ ML to solve for India’s requirements. The Launchpad Accelerator mentorship programme for start-ups using Artificial Intelligence (AI)\/ Machine Learning (ML)India chapter is scheduled to commence in […]","categories":["AI News"],"tags":["Machine Learning"],"author_name":"Bharat Adibhatla","publish_date":"2018-12-28T09:41:09","publication_year":"2018","word_count":348,"keywords":["Go","artificial intelligence","machine learning","AI","ML","Machine Learning","Scala","Aim","ViT","R","startup"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","Aim","R","Go","Scala","ViT","startup"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/google-india-now-accepting-applications-for-launchpad-accelerator-invites-ai-ml-startups\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":40206,"title":"4 Top Autocomplete Coding Tools For Python Programmers","content":"While programming or coding, a programmer can sometimes take hours of time to solve an error. Autocomplete tools serve as a great helping hand as it helps them to complete the code faster while reducing the errors. In this article, we list down 4 autocomplete coding tools for Python programmers. 1| Kite Kite is a powerful editor integration which allows you to work uninterrupted on the same screen. It is a free AI autocomplete engine which helps the programmers to code faster inPython with Line-of-Code completions. Kite’s Line-of-Code Completions feature uses deep learning to serve context-relevant code completions in real time. It is basically a plugin for your IDE which uses machine learning to give you useful code completions for Python. Some of the features of Kite are mentioned below. It integrates with all the top Python IDEs such as Atom, PyCharm, Sublime, Vim and VS Code. It accelerates software development by automatically suggesting relevant code snippets in real time Kite trains its machine learning models with thousands of publicly available code sources from highly rated developers. It can predict several “words” of code at a time, powered by the most sophisticated AI engine available for modeling code. It performs all processing locally on users’ computers, instead of in the cloud. Click here for more details. 2| Jedi Jedi is a static analysis tool for Python that can be used in IDEs\/editors. The IDE is primarily focused on auto-completion and also does statistical analysis. It is fast and is very well tested and it understands Python on a deeper level than all other static analysis frameworks for Python. Jedi has support for two different goto functions and uses a very simple API to connect with IDE’s. The core of Jedi consists of three parts, parser, Python code evaluation, and API. The general features of the Jedi are mentioned below: Python 2.7 and 3.4+ support Ignores syntax errors and wrong indentation Can deal with complex module \/ function \/ class structures Great Virtualenv support Can infer function arguments from sphinx, epydoc and basic numpydoc docstrings, and PEP0484-style type hints (type hinting). Click here for more details. 3| Wing Wing is a light-weighted and powerful Python Integrated Developmet Environment which is designed for productive development experience. Using this IDE you can easily navigate code as well as documentation and makes it easier to understand and work with existing code. There are several features of Wing as mentioned below: Intelligent Editor: Wing’s editor speeds up interactive Python development with context-appropriate auto-completion and documentation, inline error detection and code quality analysis, auto-editing, refactoring, code folding, multi-selection, customizable code snippets, and much more. Powerful Debugger: Wing’s debugger makes it easy to fix bugs and write new Python code interactively. Easy Code Navigation: Wing makes it easy to get around code with goto-definition, find uses, editor symbol index, module and class browser, keyboard-driven search, etc. Integrated Unit Testing: Wing supports test-driven development with the unittest, doctest, nose, pytest, and Django testing frameworks. Remote Development: Wing’s quick-to-configure remote development support delivers all of Wing’s features seamlessly and securely to Python code running on a remote host, VM, or container. Customisable and Extensible: Wing offers hundreds of configuration options affecting editor emulation, display themes, syntax coloring, UI layout, and much more. Click here for more details. 4| Finisher Finisher is a lightweight autocompletion library for Python. There are basically two things going on in this library, autocompletion, and spellcheck. The autocompletion works by assuming that the input tokens are intended while the spell check is done by taking an input blob of text, tokenize it, try to convert all those tokens to valid tokens, then find the best match from those tokens. It can be used in situations where you do not want to add additional dependencies such as SOLR or Cloudsearch to provide auto-completion functionality. Click here for more details.","excerpt":"While programming or coding, a programmer can sometimes take hours of time to solve an error. Autocomplete tools serve as a great helping hand as it helps them to complete the code faster while reducing the errors. In this article, we list down 4 autocomplete coding tools for Python programmers. 1| Kite Kite is a […]","categories":["AI Trends"],"tags":["django python","Pycharm","python tools"],"author_name":"Ambika Choudhury","publish_date":"2019-06-05T07:50:32","publication_year":"2019","word_count":643,"keywords":["Go","NumPy","API","machine learning","AI","ML","Python","Pycharm","django python","python tools","deep learning","programming_languages:Python","R"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","NumPy","Python","R","Go","API","programming_languages:Python"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/4-top-autocomplete-coding-tools-for-python-programmers\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":37154,"title":"YouTube Music Vs Spotify: Which ML-Powered Streaming Service Should You Pick In India?","content":"India has now become the battleground for audio streaming services. With the rise of cheap mobile Internet, audio streaming has quickly risen in popularity to become one of the easiest ways to listen to music everywhere. Keeping this in mind, two of the biggest machine learning-powered streaming services have launched their services in India, with those being Spotify and YouTube Music. The Swedish streaming service has been holding off launching in India as it procured streaming rights, while Google’s offering is now in its worldwide rollout phase after their initial launch in June 2018. This creates a competitive environment for consumers, with both offerings functioning in the same price range. There is also a free ad-supported version of both services for those who do not wish to pay, and a one-month free trial as well. Both these platforms use AI and ML extensively to recommend music to users, along with radio functionalities based on recommendation systems. Join us as we take a deeper look into the options provided by both services, and how they stack up against each other in the audio streaming platform war. YouTube Music YouTube Music offers a unique experience in terms of the vast breadth of content available on the site. It has been used as a repository for music since 2007, with multiple budding artists uploading original work as well as covers to the website. According to statistics provided by YouTube, over a billion users utilized the platform for music every month, and over 2 million unique artists are present on the platform. This also drove the growth of YouTube as an audio platform, to the point where it is the dominant choice to listen to music today. This drove the creation of YouTube Music as a part of the company’s subscription model for streaming. Currently, the monthly membership for YouTube Music is set at ₹99 a month, with ad-free YouTube, including Premium shows, only ₹30 extra a month. There is also a one-month free trial for those who wish to try the service before they buy. The service has applications available on Android and iOS, with a web client for listening on the desktop. One of the biggest drawbacks of the ad-supported version of the service is that the audio does not play when the app is not open, similar to the YouTube app on mobile. However, users have access to the complete library of music on YouTube without exception, albeit with ads. However, it has a deep integration with live music videos on YouTube, playing them directly within the application. As the user listens, the recommendations get better due to their machine learning smarts. This is collected in a playlist called ‘Your Mixtape’ which is updated every day. Apart from that, the app has multiple subcategories on the homepage. According to the time of day, the app recommends specific music that is listened to at that time. It also harvests location data occasionally to recommend music, such as for home, gym or work. There are also a multitude of curated playlists for music discovery, both by YouTube and by users who have created them on the platform. These are also categorized according to ML. The app also features a ‘Hotlist’, which is a page for trending videos in the users’ location. YouTube also brings it’s production level CNN processing used on desktop versions to the application, using a funnel down approach to recommend songs to the user. From the video corpus of millions, candidate generation methods are used in conjunction with user history and context to narrow the range down to the thousands. Post this, the recommended songs are then put through a ranking algorithm that uses the established user profile to further cut down the number of songs recommended. This is done in conjunction with metadata and other candidate sources. The song radio function on YouTube is functional at best and does not perform the role of showing new music to the user. It plays one or two songs that are similar to the subject of the song radio  but continues to pick songs out of users’ libraries after that, leading to a stagnant listening experience. Paying customers will get the ability to play music in the background and ads removed, along with a curated ‘Offline Mixtape’ which is stored on the device and can be played at any time. This playlist evolves over time as well, with different songs being downloaded almost every day. Spotify The Swedish streaming service has been a long time coming in India, mostly due to licensing issues of streaming in India. Spotify wished to exhibit a loophole in Indian law that would allow them to stream songs licensed by Warner Bros. International. This was due to the latter cracking down on Spotify’s streaming rights in India. This case is currently in court, leaving Indian users unable to stream some songs. However, apart from this, the platform is functioning perfectly in India. The membership has various plan periods, ranging from daily, weekly, monthly, half-yearly to annual plans. This flexible model might be attractive to some users. The plans are priced at ₹13, ₹39, ₹129, ₹719, and ₹1,189 respectively. The ad-supported version has one big drawback, in that users cannot listen to individual songs. Albums and playlists are shuffled and played, with only a limited number of skips per hour to ensure that the user does not keep skipping in order to listen to the song they like. This may be a deal-breaker for many, as they simply cannot listen to the songs they like the way that they want to. To combat this, they can create playlists that include the songs that they like and play them with minimal skips to get over this pitfall. The service has applications for Android, iOS and Windows, with a web client available for good measure. The desktop version allows non-paying users to access whichever song they wish without restriction, allowing them to create playlists. This is where the service shines through; in curated playlists that are varied across genres. There is a playlist for everything on Spotify, along with a varied collection of podcasts and audio experiences if that is preferred. The ML smarts of Spotify begin to show up a few days after the user utilizes the service extensively. Starting off with ‘Daily Mix 1’, Spotify creates recommended playlists based on the songs that users listen to. Up to 6 ‘Daily Mix’ playlists can be created that offer genre-based recommendations to the user. Added to this, there are also Weekly Mixtapes called ‘Release Radar’ and ‘Discover Weekly’ that showcase new releases every week and personalized recommendations. This is created through the use of a functionality known as collaborative filtering, which relies on implicit cues from users who Spotify has categorized as similar. The songs liked by user A are then recommended to user B, and vice versa to ensure good recommendations. The platform also has a ‘Song Radio’ functionality that picks songs that are similar to one song and plays it by using Machine Learning. This is done through a deep neural network algorithm that analyzes raw audio for tagging and recommendations. This is a great tool for music discovery while still staying with the artists and genres known to the user. Paying customers will be able to pick whichever song they want even on mobile versions, with cross-play functionality across all platforms. Ads will also be removed. Conclusion If users are looking for multi-language recommendations, curated playlists and a deep collection, Spotify is a safe bet and probably the best fit. However, to keep updated with the latest trends in the music industry, as well as live performances and covers, YouTube is the place to be. The recommendations are on similar levels on both platforms, while YouTube offers a more sustainable free membership compared to Spotify.","excerpt":"India has now become the battleground for audio streaming services. With the rise of cheap mobile Internet, audio streaming has quickly risen in popularity to become one of the easiest ways to listen to music everywhere. Keeping this in mind, two of the biggest machine learning-powered streaming services have launched their services in India, with […]","categories":["Deep Tech"],"tags":["ML","spotify","YouTube"],"author_name":"Anirudh VK","publish_date":"2019-04-01T11:57:59","publication_year":"2019","word_count":1308,"keywords":["Go","machine learning","programming_languages:R","AI","neural network","ML","recommendation systems","programming_languages:Go","spotify","CNN","YouTube","R"],"extracted_tech_keywords":["AI","machine learning","ML","neural network","recommendation systems","R","Go","CNN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/youtube-music-vs-spotify-which-ml-powered-streaming-service-should-you-pick-in-india\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10124052,"title":"Elon Musk Predicts AI-Driven Age of Abundance, Warns of Existential Crisis and Potential Annihilation","content":"Elon Musk, CEO of Tesla and X, says AI and robots will probably lead to an age of abundance with Universal High Income for all, but that this could result in a crisis of meaning and there’s a 10-20% chance of annihilation. When asked whether he’s both optimistic and pessimistic about AI, Musk candidly replied, “The probabilities associated with it, I think, are either 100% optimistic or completely pessimistic. I generally agree with Jeff Hinton, one of the Godfathers of AI. He posted on the X platform of the AI Community, and he thinks there’s a 10% or 20% probability of something terrible happening but the glass is 80% full. Just look on the bright side.” Elon Musk says AI and robots will probably lead to an age of abundance with Universal High Income for all, but that this could result in a crisis of meaning and there's a 10-20% chance of annihilation pic.twitter.com\/2XuCAZkZJl— Tsarathustra (@tsarnick) June 19, 2024 Moreover, he added that he envisions a future where advancements in AI and robotics usher in an era of abundance, potentially providing universal high income for everyone. This optimistic scenario hinges on the premise that AI and robots will significantly enhance productivity, reduce costs, and create unprecedented wealth. However, Musk also warns that such a shift could precipitate a profound crisis of meaning. As traditional jobs become no longer in use and daily survival no longer drives human activity, individuals may struggle to find purpose and fulfillment in a world where work is optional. What motivates Elon Musk SpaceX takes 80% of the payload to space, but that’s not actually the ambition of SpaceX. “The goal of SpaceX is to make life multiplanetary in order to extend the probable lifespan of consciousness. That has been the goal from the beginning.” said Musk. “You can see videos of me talking about this 20 years ago. In order to do that, we need to transport a lot of people and equipment to Mars.” This could lead to widespread societal challenges as people seek new ways to derive satisfaction and meaning. Musk also cautions that there is a risk estimated at 10-20% that the very technologies promising abundance could instead lead to outcomes, including potential annihilation.","excerpt":"Traditional jobs become no longer in use and daily survival no longer drives human activity, individuals may struggle to find purpose and fulfillment in a world where work is optional.","categories":["AI News"],"tags":["AI","Elon Musk"],"author_name":"Tarunya S","publish_date":"2024-06-20T12:22:46","publication_year":"2024","word_count":372,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","Elon Musk","ViT","ai_applications:robotics","R"],"extracted_tech_keywords":["AI","R","Go","ViT","programming_languages:R","programming_languages:Go","ai_applications:robotics"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/elon-musk-predicts-ai-driven-age-of-abundance-warns-of-existential-crisis-and-potential-annihilation\/","complexity_score":3,"technical_depth":7,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":58958,"title":"Google Cloud Appointed Karan Bajwa As Managing Director For Its Cloud Offerings In India","content":"Google today announced the assignment of Karan Bajwa as Managing Director of Google Cloud in India. He will be responsible for driving all revenue and go-to-market operations for Google Cloud’s extensive solution portfolio that includes Google Cloud Platform and G Suite. Google Cloud’s field sales, partner and customer engineering organisations in India will also report to him, and he will advise Google Cloud’s continued work with the local developer ecosystem and India-based Global System Integrators (GSIs). Making the announcement, Rick Harshman, Managing Director of Google Cloud in the Asia Pacific, said, “Karan is a veteran in the industry with a proven track record of building and growing successful enterprise businesses. His experience will be a tremendous asset to Google Cloud’s business, our partners and our customers as we embark on this next phase of growth.” Karan Bajwa, Managing Director, Google Cloud in India, said, “I’m very excited about this new challenge, and I look forward to extending Google’s global momentum in India. Leveraging cloud computing technology to modernise and scale for growth is on the agenda of almost every enterprise CEO, and CIO and Google Cloud is committed to helping every organisation accelerate their digital transformation.” A senior leader with over three decades of leadership experience, Karan joins Google Cloud from IBM, where he served as Managing Director for India and South Asia. Before IBM, he worked with Microsoft for nine years, his last role being the Managing Director for the company’s operations in India. He has also worked with Cisco Systems in India and Singapore. In India, the customers that are working with Google Cloud to help them solve their most complex business and technology challenges include Dr Reddy’s Laboratories, Indiamart, Hero Motocorp, ICICI Prudential, L&T Finance, LIC HFL, Manipal Hospitals, OYO Hotels and Homes, Truecaller, Wipro and many more to deliver high performing, low latency cloud-based services to their users, no matter where they are around the world, to name a few. Earlier this month, the company also announced plans to expand its presence in India by launching a cloud region in Delhi, adding to its Mumbai region, which was opened in 2017.","excerpt":"Google today announced the assignment of Karan Bajwa as Managing Director of Google Cloud in India. He will be responsible for driving all revenue and go-to-market operations for Google Cloud’s extensive solution portfolio that includes Google Cloud Platform and G Suite. Google Cloud’s field sales, partner and customer engineering organisations in India will also report […]","categories":["AI News"],"tags":["Cloud Computing in IT Industry","Google Cloud"],"author_name":"Rohit Yadav","publish_date":"2020-03-18T16:45:00","publication_year":"2020","word_count":355,"keywords":["Go","Google Cloud","programming_languages:R","cloud computing","AI","Cloud Computing in IT Industry","digital transformation","Git","RAG","cloud_platforms:Google Cloud","GAN","R"],"extracted_tech_keywords":["AI","RAG","cloud computing","R","Go","Git","GAN","digital transformation","cloud_platforms:Google Cloud","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/google-cloud-appointed-karan-bajwa-as-managing-director-for-its-cloud-offerings-in-india\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10167215,"title":"How Anthropic’s AI Model Thinks, Lies, and Catches itself Making a Mistake","content":"AI isn’t perfect. It can hallucinate and sometimes be inaccurate—but can it straight-up fake a story just to match your flow? Yes, it turns out that AI can lie to you. Anthropic researchers recently set out to uncover the secrets of LLM and much more. They shared their findings in a blog post that read, “From a reliability perspective, the problem is that Claude’s ‘fake’ reasoning can be very convincing.” The study aimed to find out how Claude 3.5 Haiku thinks by using a ‘circuit tracing’ technique. This is a method to uncover how language models produce outputs by constructing graphs that show the flow of information through interpretable components within the model. Paras Chopra, founder of Lossfunk, took to X, calling one of their research papers “a beautiful paper by Anthropic”. However, the question is: Can the study help us understand AI models better? AI Can Be Unfaithful In the research paper titled ‘On the Biology of a Large Language Model’, Anthropic researchers mentioned that the chain-of-thought reasoning (CoT) is not always faithful, a claim also backed by other research papers. The paper shared two examples where Claude 3.5 Haiku indulged in unfaithful chains of thought. It labelled the examples as the model exhibiting “bullshitting”, which is when someone deliberately makes false claims about what is true, referencing Harry G Frankfurt’s bestseller, and “motivated reasoning”, which refers to the model trying to align to the user’s input. For motivated reasoning, the model worked backwards to match the answer shared by the user in the prompt itself, as shown in the image below. Source: Anthropic When it comes to “bullshitting”, it was found the model guessed the answer even if it claimed to use the calculator as per its chain of thought. Source: Anthropic When presented with a straightforward mathematical problem, such as calculating the square root of 0.64, Claude demonstrates a reliable, step-by-step reasoning process, accurately breaking down the problem into manageable components.However, when faced with a more complex calculation, like the cosine of a large, non-trivial number, Claude’s behaviour shifts, and it tries to come up with any answer without caring about whether it is true or false. Overall, Claude was found to make convincing-sounding steps to get where it wants to go. Model Realises Its Mistake As It Writes The 1st Sentence Anthropic researchers tried jailbreaking prompts to trick the model into bypassing its safety guardrails, pushing it to give information on making a bomb.The model initially refused the request, but was soon fulfilling a harmful request. This highlighted the model’s ability to change its mind compared to what it inferred in the beginning. Explaining this ordeal, the researchers stated, “The model doesn’t know what it plans to say until it actually says it, and thus has no opportunity to recognise the harmful request at this stage.” The researchers removed the punctuation from the sentence when using the jailbreaking prompt, and found that it made things more effective, pushing Claude 3.5 Haiku to share more information. The study concluded that the model didn’t recognise “bomb” in the encoded input, prioritised instruction-following and grammatical coherence over safety, and didn’t initially activate harmful request detection features because it failed to link “bomb” and “how to make”. Claude Plans Ahead When Writing a Poem The researchers found compelling evidence that Claude 3.5 Haiku plans ahead when writing rhyming poems. Instead of improvising each line and finding a word that rhymes at the end, the model often activates features corresponding to candidate end-of-next-line words before even writing that line. This suggests that the model considers potential rhyming words in advance, considering the rhyme scheme and the context of the previous lines. Furthermore, the model uses these “planned word” features to influence how it constructs the entire line. It doesn’t just choose the final word to fit; it seems to “write towards” that target word as it generates the intermediate words of the line. The researchers were even able to manipulate the model’s planned words and observe how it restructured the line accordingly, demonstrating a sophisticated interplay of forward and backward planning in the poem-writing process. What’s Next? The research paper stated, “The ability to trace Claude’s actual internal reasoning—and not just what it claims to be doing—opens up new possibilities for auditing AI systems”. A key finding is that language models are incredibly complex. Even seemingly simple tasks involve a multitude of interconnected steps and “thinking” processes within the model. The researchers acknowledge that their methods are still developing and have limitations. Still, they believe this kind of research is crucial for understanding and improving the safety and reliability of AI. Ultimately, this work represents an effort to move beyond treating language models as “black boxes”.","excerpt":"Anthropic’s Claude was found to be providing false reasoning while attempting to decode how an LLM thinks.","categories":["AI Features"],"tags":["Anthropic AI"],"author_name":"Ankush Das","publish_date":"2025-04-03T15:00:32","publication_year":"2025","word_count":785,"keywords":["Anthropic","Go","TPU","programming_languages:R","AI","llm_models:Claude","Aim","chain of thought","Claude 3.5","R","Anthropic AI"],"extracted_tech_keywords":["AI","Claude 3.5","Anthropic","Aim","chain of thought","TPU","R","Go","llm_models:Claude","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-anthropics-ai-model-thinks-lies-and-catches-itself-making-a-mistake\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":20224,"title":"Rajinikanth Makes His Much Awaited Political Debut With Research And Data Analytics In His Arsenal","content":"The new year began with the superstar of Tamil cinema, Rajinikanth, taking the much anticipated plunge into the world of politics. Known for his dedication towards his craft, it appears that the ‘thalaiva’ is approaching his new political role with the same. The well-timed announcement of his political entry is backed through research and data analytics. It was reported that selected teams from the actor’s camp were dispatched across the state of Tamil Nadu to take stock of the agricultural and geopolitical concerns, and aggregate data and solutions pertaining to them. A technology team is said to be at the ready to deploy data analytics to the aggregated figures. Professionals in the information and technology industry, especially in the USA, form a sizeable part of the massive fan base that the actor enjoys overseas. With regards to the technological support and funding that Rajinikanth’s campaign is expected to receive, an anonymous source told a newspaper that this group of fans can play an important role. “They are completely prepared. As and when the political party is launched, it will be one of the strongest technology backed initiative to hit the political theatre in the state, with complete data support for every constituency and segments within. That is the only way the team thinks Rajinikanth could take on the might of the two Dravidian parties. Success or failure comes next,” added the source. A belief that a considerable number of his fans could have been influenced by the current system in the state of Tamil Nadu is reported to have emerged. Analytics is expected to be employed even in the efforts to prevent such fans from attaining positions within the party. This diligent filtration of undesirable elements seems to have seeped into the actor’s fan interaction events too. “Even the list of fans who met the star and took a photograph with him in the recent times, was highly filtered to keep out unwanted elements,” told another source associated with the initiative. An example of the employment of data analytics in Indian politics to great effect was seen in the 2014 general elections when Narendra Modi swept the elections. Analytics coupled with a strong digital campaign gave the Modi-led Bharatiya Janata Party (BJP) a landslide victory. “BJP’s election campaign was a very good mix of all technologies. It wasn’t just social media but all digital platforms, including mobile and mobile Internet. Combined with advanced use of analytics to decide how, where and when to communicate, the campaign targeted 810 million voters in 543 constituencies across 11.36 lakh (1.13 million) polling booths,” said  Arvind Gupta, head of the information technology unit of BJP, at a gathering few months after the elections. In a bid to connect en masse with fans sharing his political ideas, Rajinikanth also launched a website and a mobile application on 1 January 2018. With two well established political contenders of historical repute dominating the political arena in Tamil Nadu, technology could be the armament that levels the playing field for Rajinikanth’s announced party.","excerpt":"The new year began with the superstar of Tamil cinema, Rajinikanth, taking the much anticipated plunge into the world of politics. Known for his dedication towards his craft, it appears that the ‘thalaiva’ is approaching his new political role with the same. The well-timed announcement of his political entry is backed through research and data […]","categories":["AI News"],"tags":["Data Analytics","elections","political campaigns and ai","politics","Tamil Nadu"],"author_name":"Jeevan Biswas","publish_date":"2018-01-02T09:26:24","publication_year":"2018","word_count":505,"keywords":["political campaigns and ai","Go","funding","programming_languages:R","AI","programming_languages:Go","Git","politics","GAN","analytics","Tamil Nadu","Data Analytics","R","elections"],"extracted_tech_keywords":["AI","analytics","R","Go","Git","GAN","funding","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/rajinikanth-politics-data-analytics-technology\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10017462,"title":"JupyterLab 3.0 Released: All The Major Features &#038; Updates","content":"Recently, one of the contributors within the Jupyter ecosystem has announced the release of JupyterLab 3.0. The third major release comes with significant announcements, including support for multiple display languages, visual debugger, and more. JupyterLab is a popular web-based interface for Jupyter Notebooks.  JupyterLab enables users to perform work with documents such as Jupyter notebooks, text editors, custom components, among others in a flexible, integrated as well as extensible manner. JupyterLab 3.0 is released!– visual debugger– support for multiple display languages– table of content for notebooks– improved extension system.Check out the announcement blog post.https:\/\/t.co\/pUBiZEYH4c— Project Jupyter (@ProjectJupyter) January 5, 2021 JupyterLab offers many intuitive features that enable data scientists and developers to work with notebooks, code, and data efficiently. The features include: Code consoles that provide transient scratchpads for running code interactively.Kernel-backed documents to enable code in any text file, including Python, R, LaTeX, etc.It provides multiple views of documents with various editors\/ viewers for the editing of documents, and more in real-time. Let’s have a glance at the latest update and features- A New Visual Debugger The developer’s team at JupyterLab acted on the longstanding request from the users, especially those accustomed to general-purpose development environments, to announce the first public release of Jupyter visual debugger in March 2020. JupyterLab 3.0 is now shipped with a Debugger front-end by default. Notebooks, code consoles, and files can now be debugged from JupyterLab directly. For the debugger to be enabled and visible, a kernel with support for debugging is required. Know more here. Extensions can be installed without building JupyterLab with Node.js JupyterLab is designed as an extensible environment, and the extensions can customise or enhance any part of JupyterLab. In the early versions, rebuilding JupyterLab required Node.js to be installed. JupyterLab 3.0 has also a new recommended way of distributing and installing extensions as Python pip or Conda packages. The extensions can be easily installed without building JupyterLab with Node.js. However, the previous method of distributing extensions as npm packages requiring rebuilding JupyterLab is still available. Know more here. Multi-Language Support JupyterLab will now allow setting the display language of the user interface. To be able to provide a new display language, you need to install a language pack. Know more here. Jupyter Server The third release of JupyterLab has migrated and now depends on Jupyter Server, a new Jupyter project based on the server portion of the classic Notebook server. Click here to know more about migration. Command Palette The command palette is now made into a floating window that appears on top of the JupyterLab workspace. This feature will enable the users to quickly invoke a command while keeping the sidebar closed or switching sidebar panels. However, the command palette can be put back into the sidebar by adjusting the default in Advanced Settings. Some of the other major updates include: Improvements to Simple Interface mode and MobileTable of Contents is now in coreVisual filter in the file browserProperty inspector moved to the right sidebarJupyterLab UI now supports translation extension Steps to Install: To install JupyterLab3.0 with pip, type: pip install jupyterlab==3 With mamba, you need to type: mamba install -c conda-forge jupyterlab=3 And with Conda, type: conda install -c conda-forge jupyterlab=3 Wrapping Up In the last few years, JupyterLab has attracted much attention from the developers in emerging technologies. The JupyterLab is known to fill the gap of Jupyter Notebook by providing richer UI and functionalities. The new updates and features will undoubtedly help developers work in a fast and more efficient manner. JupyterLab will eventually replace the classic Jupyter Notebook. According to an official blogpost, throughout this transition, the same notebook document format will be supported by both the classic Notebook and JupyterLab.","excerpt":"Recently, one of the contributors within the Jupyter ecosystem has announced the release of JupyterLab 3.0. The third major release comes with significant announcements, including support for multiple display languages, visual debugger, and more.   JupyterLab is a popular web-based interface for Jupyter Notebooks.  JupyterLab enables users to perform work with documents such as Jupyter notebooks, […]","categories":["AI Trends"],"tags":["Jupyter","Jupyter Notebook"],"author_name":"Ambika Choudhury","publish_date":"2021-01-11T12:00:00","publication_year":"2021","word_count":617,"keywords":["Jupyter Notebook","programming_languages:R","AI","Python","programming_languages:Python","Jupyter","R"],"extracted_tech_keywords":["AI","Jupyter","Python","R","programming_languages:Python","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/jupyterlab-3-0-released-all-the-major-features-updates\/","complexity_score":3,"technical_depth":6,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10047157,"title":"Punjab Govt &#038; IIT Ropar Offer Free Courses In AI And Data Science","content":"The Punjab Skill Development Mission and IIT Ropar have collaborated to offer a free course in artificial intelligence and data science to the youth of Punjab. The classes will be held via an online conferencing platform by IIT professors and industry veterans. The program is open for Class 12th students with a mathematics background. After the successful completion of the course, students will get a certificate from IIT Ropar. The admission process is as follows: Step 1: Attempt and clear A-DSAT (Online eligibility test)Step 2: Enroll in the L2 ProgramStep 3: Students who clear the L2-Program will be eligible for the L3 Program It will have two module-based programmes: L2 Program (Four weeks) – This is a module-based program in which students can opt as per their interest in areas like computer vision, natural language processing, and others. This twenty-day program will be followed by a capstone in which they need to implement a real industry problem; after successful submission of the capstone, they will get certified.L3 Program (Twelve weeks) – This sixty-day program will give complete in-depth knowledge about the concepts. There will be multiple assignments in the course to ensure whatever the candidates are learning can also be implemented with perfection. Students will learn from IIT Ropar academicians in the field of Artificial Intelligence and Machine Learning and several experienced industry practitioners. Faculty have created the learning content and assessments for the program from IIT Ropar and other practising Artificial Intelligence experts from the industry. IIT Ropar will provide the course content (soft copy) after every class. For every class, there will be pre-read, class content, and assignments. Apply here for the course.","excerpt":"Open for Class 12th standard students with mathematics background. After the successful completion of the course, students will get the certificate from IIT Ropar.","categories":["AI News"],"tags":["IIT Ropar"],"author_name":"kumar Gandharv","publish_date":"2021-08-27T11:43:30","publication_year":"2021","word_count":275,"keywords":["data science","machine learning","artificial intelligence","programming_languages:R","computer vision","IIT Ropar","R","ai_applications:computer vision"],"extracted_tech_keywords":["artificial intelligence","machine learning","computer vision","data science","R","programming_languages:R","ai_applications:computer vision"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/punjab-govt-iit-ropar-offer-free-courses-in-ai-and-data-science\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10086103,"title":"Low-code: The Silver Bullet for Indian MSMEs","content":"Low-code and no-code platforms are revolutionising the way businesses and individuals develop and deploy software. In 2017, Forbes described low-code as ‘extraordinarily disruptive’. “The rise of low-code can be traced back to the onset of the pandemic, which was when organisations across industries were forced to step up and embrace digital adoption,” Deepak Visweswaraiah, vice president, platform engineering, and site managing director, Pegasystems, India, said. In an exclusive interview with Analytics India Magazine, Visweswaraiah discusses the low-code market in India alongwith its adoption in Tier 2 and 3 cities, among other things. How do you see the LC\/NC market growing in India? Deepak: A NASSCOM and Nagarro report pegged India’s low-code\/no-code market at USD 400 million; it is slated to touch USD 4 billion by 2025. Adopters of LC\/NC have witnessed a 3x-7x reduction in development and deployment time, a 3x-5x reduction in development costs, and a 30%-35% increase in ROI. One of the main reasons for the growth of LC\/NC in India is the increasing demand for software development to perform IT operations. Additionally, LC\/NC has also increased the participation of non-tech candidates in the workforce and provided opportunities for job seekers, who can use their time to develop custom apps with minimal programming skills. Can low-code\/no-code platforms power the growth of SMEs and MSMEs in India? Deepak: Low-code capabilities specialise in the creation of light-weight applications for people or small group usage and empower large enterprises to rapidly simplify complex business use cases. Given the benefits of rapid development and deployment – lower costs, faster delivery, and greater accessibility – low-code is turning out to be a silver bullet for SMEs and MSMEs to address myriad challenges. It will equip small and medium businesses with tools to develop their own platforms without relying on a dedicated team of developers. When a low-code platform is offered as a service running in a cloud, it is even more interesting for this specific market as they don’t have to worry about the infrastructure and maintenance. Tell us a bit about the low-code Pega platform Deepak: Our low-code Pega platform is designed for enterprise-scale with the aim to build applications using visual constructs. It helps us incorporate capabilities that simplifies work and manage it across organisation such as: Tools for business developers and citizen developers – A familiar drag-and-drop functionality helps citizen developers and business users be part of the solution, providing step-by-step guidance to the right process and UI, connect it to the backend APIs and deliver a suitable app. Tools to manage application governance and compliance – Governance tools and alerts help IT managers gain visibility into the quality of each app at every stage of development and apply guardrails to keep the enterprise applications compliant with internal policies and best practices. Integrated environment for secure testing, deployment, and management – Speed application delivery and deployment with open and native testing capabilities establishes a continuous testing environment and enables users to create and automate functional tests for workflows, UI, and other application components. What are other areas where no-code\/low-code can be introduced in the future? Deepak: The rapid adoption of low-code platforms is expanding to highly regulated industries to lay the cornerstone of a long-term IT strategy. Businesses empowered with low-code apps quickly adapt to new circumstances to address their changing customer needs. Sustainable low-code to minimise waste: Today, businesses produce too many single-purpose apps with no vision for reuse. One way to improve productivity is to reduce waste through mature low-code deployments. We foresee enterprises shifting their focus on building reusable assets to avoid ‘app sprawl’ and have a combined impact on innovation. Therefore, purpose-built low-code platforms are an area of focus. Low-code with a brain in future: AI is on the rise and disrupting almost every aspect of a business. In 2023, we’ll see smarter low-code infused with AI. It will optimise processes, increase efficiency, and improve decision-making, making low-code development beneficial, safer, and smarter than ever before. Creating an API economy with low-code: Without a doubt, low-code tools are helping speed up software creation and delivery. However, there is scope for businesses to embrace an approach where both high and low-code can coexist – made possible by the power of APIs. How excited are you about the Indian developer ecosystem and how different is it from other countries? The community of developers is expected to grow to 10 million by 2023, India is likely to overtake the US with the highest number of developers in the future. Developers, especially from Tier 2\/3 cities, play a significant role in shaping the future of emerging technology. The Pega ecosystem has created unlimited growth opportunities for students, college partners and global blue-chip companies. We have an ongoing association with TalentSprint – Pega University Academic Programme. We have launched the tech careers of 10,000 students in the past seven years. We started by partnering with two colleges, which have now expanded to 18+ colleges. The holistic programme enables third-year engineering college students to build robust capabilities in low-code technologies to secure a competition-proof career with the customers and partners of Pegasystems.","excerpt":"LC\/NC has increased the participation of non-tech candidates in the workforce and provided a lot of employment opportunities","categories":["AI Features"],"tags":["Interviews and Discussions","Pegasystems"],"author_name":"Pritam Bordoloi","publish_date":"2023-01-30T13:00:00","publication_year":"2023","word_count":850,"keywords":["Go","API","AI","Git","RAG","Pegasystems","Aim","ViT","analytics","GAN","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","analytics","Aim","RAG","R","Go","Git","API","GAN","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/low-code-the-silver-bullet-for-indian-msmes\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10142434,"title":"AIM Predictions 2025","content":"Another whirlwind year in AI is coming to an end, marked by nothing less than pure AI madness. Last year, when AIM made bold AI predictions for 2024, we were pretty aligned with what unfolded. From calling 2024 the year of small language models and open-source AI models levelling up with the closed-source ones, our predictions have been on point. En route to 2025 and going by the wave, the AI predictions for the coming year are sure to cause an impact at a larger scale. It is likely that physical AI will appeal at a consumer\/personal level, with humanoids probably making their way into people’s homes. Wild? Absolutely. Plausible? Maybe. So, let’s dive deep into some of the wildest AI predictions for 2025. Agentic AI No points for guessing, but going by the trend in the second half of this year, agentic AI will continue to grow massively in 2025. All major big-tech companies, including, Oracle, Microsoft, Google, and Salesforce, have launched their agentic suite of products, addressing verticals such as HR, finance, operations, supply and more. A couple of months ago, Microsoft announced the ability to create autonomous agents with Copilot Studio. Copilot is the UI for AI, and with Copilot Studio, customers can easily create, manage, and connect agents to Copilot.Today we announced new autonomous agent capabilities across Copilot Studio and Dynamics 365 to help scale the impact of every individual, team, and business…— Satya Nadella (@satyanadella) October 21, 2024 Interestingly, Gartner named agentic AI ‘the emerging trend for the year 2025’, indicating that autonomously operating agents would have a significant impact across enterprises. AI in Healthcare While this may not seem like a new use case, AI applications in the healthcare sector are witnessing an increasing adoption. Not just in healthcare, where medical copilots can assist healthcare practitioners with analysing and summarising reports, discovery and protein research are also seeing huge growth. Recently, HCG partnered with Accenture, a leading technology company, to use advanced AI to expedite drug discovery and other developments. Big day today – @nvidia is open-sourcing the BioNeMo Framework, a toolkit of programming resources, libraries, and AI models designed for drug discovery. This release equips academic labs and biotech companies with advanced tools for protein design, small molecule generation, and… pic.twitter.com\/4QKpU54co6— Vega Shah (@dr_alphalyrae) November 19, 2024 Recently, Google DeepMind open-sourced its revolutionary AlphaFold 3 model, making its training weights accessible to researchers for non-commercial use. This has further opened up the possibility of experimenting with drug discovery. Personal Use of Humanoids We dubbed 2024 ‘the year of robotics’ as the field saw a surge in research, development, and advancements, including progress with humanoid robots. Figure, Agility Robotics, Boston Dynamics, and Sanctuary AI, are just a few US companies dedicated to building commercial humanoids. Many of them are already being implemented in automobile companies. However, like in the movie I, Robot, it is possible that humanoids will make their way into our homes too. Many Chinese tech companies are also speeding through humanoid developments. Meanwhile, Tesla promoted its humanoid Optimus at the We, Robot event as a humanoid that can assist people with their chores and be part of one’s daily life. While fully functional humanoids might still take time to become reality, the year 2025 will be an important one for this. In 2025, I look forward to seeing major advancements in developing intelligent systems for humanoid robots—AI pilots that are far more capable than today’s state of the art.A good benchmark of their potential would be a robot entering an unfamiliar home and competently… pic.twitter.com\/CT7eg5W0Or— Dr Singularity (@Dr_Singularity) November 27, 2024 AI to 3D Systems Generative AI in 3D modelling is another area that promises to explode in the coming year. Companies such as Common Sense Machines and World Labs are already converting images to 3D. AI innovation follows a predictable path: text → images → audio → video → 3D.Today, 3D generative AI feels underwhelming, but history suggests that’s temporary. By 2025, expect breakthroughs to reshape industries from gaming to design via world models. The next dimension… pic.twitter.com\/grDtvfzJ7i— Allie K. Miller (@alliekmiller) November 24, 2024 Unlike generating images or videos from text, using generative AI to convert images to 3D models is considered highly complex owing to the multifaceted nature of 3D data. This requires understanding and manipulating intricate geometries, textures, and lighting. Now, with consumer brands already implementing AI to churn out advertisements, 3D will add value to this space. Considering how many companies are experimenting in this segment, 2025 will witness a surge in 3D model companies. Edge Computing Edge computing is becoming more critical than ever before for running LLMs on phones. Interestingly, this year saw major phone makers, such as Samsung and Apple, bring intelligence to their phones. Qualcomm and Apple are integrating AI capabilities into their chipsets, allowing robust edge computing for LLMs. Theres Payton, former White House CIO, said that by 2025, edge computing would become even more widespread, particularly as AI and IoT expand. https:\/\/t.co\/a8N6F5DOzITwo edge models out with impressive capabilities. High time to have a silicon intelligence on your laptop or your phone :) pic.twitter.com\/j0ax9l7gXM— Albert Jiang (@AlbertQJiang) October 16, 2024 China to Lead the AI Race Co-founder and CEO of Hugging Face Clement Delangue, known for his year-end AI predictions, also made one for 2025. One of his predictions resonates with our suggestion last year that China has a chance to be at the forefront of the AI race. China is consolidating the world's top AI talent. In 2019 China had 29% of the world's top AI reseachers. Now it has 47% and the U.S. has 18%. To reverse these trends, we need to expand skilled immigraton ASAP. pic.twitter.com\/lLqzLoWU1w— Sam Peak (@SpeakSamuel) December 1, 2024 While the country seems to be in a league of its own, its AI models have been outperforming on the leaderboard. Recently, a powerful open-source model Qwen 2.5 has been clearly leading the AI agents race. Similarly, Baidu has also been making significant developments in the AI field.","excerpt":"Wild and true, AI predictions for 2025 may not seem too surprising.","categories":["AI Trends"],"tags":["3d modelling","agentic ai","ai predictions","Baidu","Boston Dynamics","China","edge computing","humanoids","Qwen"],"author_name":"Vandana Nair","publish_date":"2024-12-04T12:52:28","publication_year":"2024","word_count":1002,"keywords":["Boston Dynamics","ai predictions","agentic AI","Qwen","Hugging Face","humanoids","AI","autonomous agents","agentic ai","Aim","edge computing","Baidu","generative AI","3d modelling","copilots","small language models","R","China"],"extracted_tech_keywords":["AI","generative AI","agentic AI","Aim","Hugging Face","small language models","autonomous agents","copilots","edge computing","R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/aim-predictions-2025\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10123772,"title":"AI Can Never Replace Excel","content":"The world is built on spreadsheets. Even before the introduction of modern machine learning tools, most of the data of the world was stored on Excel spreadsheets. Cut to today, and that’s still the case. Just like any other tool, though, with the advent of applications such as ChatGPT, it was thought that there would be no need for Excel sheets anymore, since data science could easily be done by just allowing LLMs to make sense of the data. Probably true … pic.twitter.com\/s2aSya4vxl— Wall Street Silver (@WallStreetSilv) June 12, 2024 The AI Revolution in Excel One of the strongest arguments against Excel was that it was becoming obsolete. Excel is not user-friendly, and the application rounds off very large figures with accurate computations, which reduces its accuracy. Excel is also a stand-alone application that is not fully integrated with other corporate systems. It does not provide sufficient control because users do not have a clear and consistent view of the quotations sent by their representatives, as well as the history of those quotes. So, what does the future of Excel look like? We can anticipate a significant role for AI. Microsoft has been steadily integrating AI aspects into Excel, and with its ever-expanding capabilities, it is set to continue revolutionising the software, sparking excitement about the future possibilities. Microsoft unveiled Copilot, its innovative AI tool, in mid-March 2023. This cutting-edge technology is set to revolutionise the functionality of Microsoft Office programs, including Word, Excel, PowerPoint, Outlook, and Teams. Consider the practical implications of Copilot in Excel. By harnessing natural language processing (NLP) AI techniques, it empowers users to ask questions in plain language and receive accurate, context-aware answers. This transformative technology enhances Excel’s usability, providing valuable recommendations and precise results. By integrating Copilot into the Microsoft 365 suite, Microsoft is placing generative AI tools in front of over a billion of its users, possibly changing how large segments of the global workforce communicate with one another. One example is the Data Analyse feature found in the most recent versions of Excel. NLP can also recommend functions, formulas, and features the user may not know, making it easier to identify the best answer. Python in Excel The introduction of Python in Excel has given the application a major boost. Python in Excel uses Anaconda, a prominent repository notable for allowing developers to run multiple Python environments. Now users can do advanced data analysis within the familiar Excel interface leveraging Python, which is available on the Excel ribbon. Access to Python allows users to use Python objects within cell functions and calculations. Consider a Python object being referenced or its data used in a PivotTable. Even popular libraries such as scikit-learn, Seaborn, and Matplotlib can be utilised with Excel. This allows Python-created visualisations, data models, and statistical calculations to be combined with Excel functions and plugins. The integration of Python cements Excel’s position in data analytics, suggesting that Excel’s utility in the workplace is far from diminishing.  Last year, Microsoft announced that they would be experimenting with GPT in their office applications. Interestingly, Microsoft has been following this path the entire time in acquiring Github, acquiring OpenAI, which helped them make Copilot for writing Python code and use ChatGPT’s Code Interpreter, AI assistance on PowerBI, and, now, Python in Excel. Looking at the progress of Excel, a future where AI writing Python on our Excel sheets cannot be ruled out.  Moreover, the talk of AGI being built around Python makes the future of Excel more secure. Why AI in Excel Is a Win-Win Situation Microsoft Excel is the world’s most popular spreadsheet software. Approximately 54% of organisations use Excel. According to a recent report by corporate planning company Board International, 55% of organisations undertake their enterprise planning, including budgeting and sales forecasting, on spreadsheets. Four out of five Fortune 500 companies use Excel, and over two billion people worldwide use spreadsheets. Excel and spreadsheets became popular in the 1980s, and despite attempts by competitors, Excel and its ilk continue to dominate. Office workers today utilise tools like Excel to visualise, analyse, organise, distribute, and present chunks of corporate data—whether exported from databases or created on the fly. Businesses continue to rely on Excel, so much so that financial specialists claim you’ll “have to pry Excel out of their cold, dead hands” before they ever stop using it.","excerpt":"Businesses continue to rely on Excel, so much so that financial specialists claim you’ll “have to pry Excel out of their cold, dead hands” before they ever stop using it.","categories":["AI Features"],"tags":["AI (Artificial Intelligence)","Data Science","excel","Machine Learning"],"author_name":"Anshul Vipat","publish_date":"2024-06-17T10:01:14","publication_year":"2024","word_count":725,"keywords":["data science","excel","ChatGPT","machine learning","scikit-learn","OpenAI","AI","Machine Learning","NLP","Aim","analytics","generative AI","Data Science","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","machine learning","NLP","data science","analytics","generative AI","ChatGPT","OpenAI","Aim","scikit-learn"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/ai-can-never-replace-excel\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":34753,"title":"One AI Algorithm By Mumbai-Based Startup Quantum Four Helped Make Lending Decisions Worth ₹1,500 Crore In 2 Years","content":"Institutionalising data-driven decision making is the need of the hour, and while the world is capitalizing on the tech du jour AI, this Mumbai-based company founded in 2016 is going the extra mile by specialising the field of AI, NLP and big data computing, and providing real-world solutions with human-like intelligence. What Is Quantum Four? With a huge competitive edge over others, the startup founded by Parnil Mhatre concentrates on a broad range of analytics solutions — predictive analytics, social behaviour filtering, social media intelligence and big data architecture — with machine learning underpinning the solutions. With Artificial Intelligence Lab at its core, Quantum Four extensively uses science, mathematics and machine learning. “We believe that solving problems of the future needs AI for the simple reason that in today’s world of hyper-information, human experience may not always be available. A learning system can be the only solution to bridge this gap,” says Mhatre, who has deep expertise in neural computation, genetic algorithm and NLP. With a team of highly trained mathematicians, big data programmers and business analysts, the startup boasts of a big data parallel computing as an infrastructure backbone to all their AI solutions, thereby satisfying the industry’s volume and velocity needs. With this strong backbone, Mhatre aims to nurture Quantum Four into AI labs which does not focus on a particular technology but rather challenges itself to apply a customised array of solutions depending on the business problem at hand. The Unique AI Algorithm With an aim to create solutions that are beyond conventional and truly intelligent, Mhatre built his own AI algorithm to predict human behaviour called Social Behavior Filtering, which has been widely used in the fintech space. The first ever SBF is currently being used in India’s fastest growing short-term lending fintech company called CASHe to make lending decisions worth 300 Cr easier. When asked what inspired him to make this algorithm, Mhatre shares that human behaviour is an extremely difficult problem to solve mainly because it is impossible to classify behaviour into separate categories. “Behaviour is more of a continuous entity. Intuitively we are aware, that even if we know everything about a person it is very difficult to predict his specific behavioural patterns,” he said. Keeping this in mind they built an AI algorithm to track and forecast human behaviour called ‘Social Behavior Filtering’ (SBF) which uses an entirely different set of mathematics called ‘Graphical Modelling’ to solve this in a tangible way. “Google’s page ranking is one of the very few exceptions which is based on similar mathematics. We consider individual to be part of a larger network, a network made up of not just individuals but all the relevant pieces of information. Our algorithm learns from all this and can predict even extremely abstract behavioural patterns,” explains Mhatre. Today, SBF is being successfully used in a leading Fintech company and has aided lending decisions worth ₹1,500 crore within a short span of two years. The Stepping Stone On being asked how did he start Quantum Four, Mhatre quickly adds that he has always been a passionate advocate of artificial intelligence and likes to come up with ingenious and innovative applications of AI across various sectors. “I detest plug and play solutions. It was with this background, I started Quantum Four – an A.I. lab where science, mathematics and technology would be the drivers for all solutions,” he shares. Discussing The Commonly Faced Challenges In The AI Industry Mhatre shares three common challenges he thinks the AI industry faces in India: Talent Crunch: Mhatre believes that there is an enormous talent crunch mostly because people who are doing AI do not have a mathematical background. “I have seen a lot of companies run AI-like technology but the reality is AI is still a science and it will remain a science for the foreseeable future,” he said. He emphasises that until there is a right type of mathematician, software engineer and domain experts, AI-based tasks cannot be done effectively. “In order to use AI to its fullest potential, you need a team that can do science (mathematics), understand technology (big data) and most importantly communicate the results in the most effective way to the decision maker,” shares Mhatre. “In an organisation whose obvious focus is not AI, it is extremely rare to find this sort of a team and Quantum Four with its Enterprise Solution – Massive Dynamic Platform sits in the middle of these three worlds,” he adds. Perfect AI Is Still Far Away: “We have not reached the age of perfect AI or singularity, and we are far away from it,” says Mhatre. He believes that what we have done today is re-engineer a certain aspect of intelligence. “For e.g. We can differentiate between the shape of a car to the shape of a human being. And we can also differentiate between certain phonetics which lead to speech recognition and so on. With respect to developments of AI. in India, our country is one of the places where it is thriving as a lot of business leaders want to internalise AI. However, there is a gap between doing some sort of AI and the sort of A.I that delivers ROI,” he adds. Data Compliance: Mhatre shares that one of the main challenges they face is data compliance. “When we meet our clients, in order to research their data, we need to cross a lot of legal hurdles which delays initial involvement. Hence within Massive Dynamic, we have designed AI bots that can reside inside the company infrastructure itself and perform all necessary computation by pulling the algorithms in, rather than pushing the data out of the company security domain,” he said. The Future Road The startup is looking forward to launching their Enterprise Solution ‘Massive Dynamic’ this year, that allows them to do science seamlessly on top of cutting-edge technology. “Massive Dynamic is our proprietary Next Gen ‘AI Enterprise Solution’ that has one of its kind interfaces, with an in-built sophisticated toolbox equipped with all the latest machine learning algorithms. It is where business leaders\/scientists can generate AI-based solutions with the help of their experts, mine insights via dynamic visualisations and come up with tangible decisions for their real-world problems, said Mhatre. The system is based on big data which does all the heavy-lifting to learn from the data itself by handling all the technical and scientific necessities that may arise and thereby implementing AI backbone for any business. The best part about this solution is that the base platform is sector agnostic, where customization is done on business specific requirement for a speedy and cost-effective production system. However, they wish to work on the entertainment sector, human resource, logistics and healthcare especially in the areas of nutrition and genetics. “Nutrition is heavily dependent on individual behaviour and a regular track of the individual’s behaviour (how much one travels, eats, stress levels etc) is used in combination with data sources like wearable devices and mobile data, the nutrition recommendation can be highly customized and near accurate,” he said. Another area they want to make an impact is logistics. “Here, what we want to do, is take driver data, vehicle data, type of consignment, road quality and specifically traffic situations into consideration. In addition to it, we not only want to add current weather data but also predict future weather conditions that might exist.,” he said. Use AI Not As A Vitamin, But As A Painkiller While Mhatre believes that India is making great progress in the field of AI, he says that to capitalise on AI, three points need to be kept on the mind — use AI not as a vitamin but as a painkiller, asking the right question, and have the ability to connect theoretical AI to domain expertise. He sums it up by saying, “If you want your business decision to follow AI, your AI should follow domain expertise.”","excerpt":"Institutionalising data-driven decision making is the need of the hour, and while the world is capitalizing on the tech du jour AI, this Mumbai-based company founded in 2016 is going the extra mile by specialising the field of AI, NLP and big data computing, and providing real-world solutions with human-like intelligence. What Is Quantum Four? […]","categories":["AI Startups"],"tags":["quantum computing companies","the quantum companies"],"author_name":"Srishti Deoras","publish_date":"2019-02-11T12:24:26","publication_year":"2019","word_count":1321,"keywords":["artificial intelligence","machine learning","AI","R","ML","quantum computing companies","the quantum companies","NLP","Ray","Aim","analytics","predictive analytics"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","NLP","analytics","Aim","Ray","predictive analytics","R"],"url":"https:\/\/analyticsindiamag.com\/ai-startups\/one-ai-algorithm-by-mumbai-based-startup-quantum-four-helped-make-lending-decisions-worth-%e2%82%b91500-crore-in-2-years\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":14287,"title":"In Review &#8211; PGP in Business Analytics at Praxis Business School","content":"One of the earliest entrants in the analytics education space and one of the country’s leading business schools, Praxis Business School has been consistently ranked among the Top 10 analytics courses in India. Praxis kick-started the Post Graduate Program in Business Analytics in 2011, four years after the business school was set up, with an objective to “fulfill the industry requirement of analytics-trained resources – who would, in due course, shape the future of the nascent and ever-evolving analytics field”. The Praxis program, with its unique offering of a full-time, in-class, comprehensive one-year program with complete campus placement responsibility, has grown to comprise two campuses, Kolkata and Bangalore, and two intakes per year, January and July, at each campus. In addition, in a rare honour, Praxis has been selected by the Andhra Pradesh Government to manage and run its Business Analytics program at the International Institute of Digital Technologies, Tirupati and has already started the first batch in February this year, We spoke to the driving force behind the business analytics program at Praxis – Prof. Charanpreet Singh, the Founder of Praxis Business School, and Dr. Prithwis Mukerjee, the Program Director Praxis is a young institute and will be completing 10 successful years this July. ‘It’s been a roller-coaster ride for the first few years as we strove to find our bearings in this highly competitive business management education domain,’ said Charanpreet Singh. ‘We stuck to our fundamentals – of being close to and addressing the needs of our two sets of customers – the students and the industry and it seems to have paid off. So at this point in time we feel that our approach has been vindicated – but the journey continues and we hope to touch and improve more lives.’ Praxis took a lead in adding Analytics to their portfolio of courses, and have enjoyed the early mover advantage in this field. When asked about the trigger for starting a program in Analytics, this is what Charanpreet Singh had to say: ‘The addition of the Business Analytics program to our portfolio is the result of our deep relationships with the industry and our willingness to listen to our industry friends. A couple of industry well-wishers suggested that we start a program in Analytics as the demand for professionals in this area was poised to skyrocket and there were no formal channels for recruitment. We saw analytics as a good vehicle to differentiate ourselves in the business education space.’ [quote]The awareness of and demand for analytics programs have increased manifold. So has the number of programs on offer – with renowned education brands also joining the fray. Praxis has carved a definite niche for itself in this competitive space and remains a compelling choice for aspirants.[\/quote] Prithwis Mukerjee lists four reasons for Praxis’ continued success: A comprehensive, contemporary curriculum that addresses the three pillars of techniques (& tools), technology and business understanding. A crack faculty team that ensures, Effective pedagogy (with a mix of lectures, case discussions, projects, labs) and A high quality of in-class engagement A stimulating student community – people from diverse geographies, backgrounds. industries and aspirations. A strong peer group is a priceless asset for any program. A well-established campus placement program that helps students get their first job in the analytics domain Courses are built on great content – Praxis’s PGP in Business Analytics keeps pace with industry demand. Our research shows that a lot of aspirants choose Praxis because of the coverage of its curriculum and the clear perception that it is updated continually ‘The initial content has been created with industry collaboration from ICICI Bank and PwC, our knowledge partners’, says PM. ‘The faculty team at Praxis comprises professionals with considerable industry experience – it keeps itself abreast of the latest industry developments and keeps upgrading the curriculum accordingly. Our industry partners play the roles of an advisor as well as a soundboard for any changes that we may plan. Other mechanisms include recruiter and alumni feedback. So we end up teaching what the industry wants as opposed to what the teacher is already comfortable teaching!’ [box title=”Dr Mukerjee outlines some key changes made to stay at the leading edge of technology:” box_color=”#ffffff” title_color=”#5b5b5b” radius=”5″]1. In 2012, statistics was the big thing and we were teaching a lot of it, shared Dr. Mukerjee. And then came coding languages, R and Python so the course was redesigned to reflect that 2. Couple of years later, Hadoop the big data platform was replaced with Spark and the classes were re-started with new Spark curriculum 3. Now, Google’s Tensorflow, an open source software library for machine learning  has been integrated into the course[\/box] ‘Class-room discussions are backed up by real-world case studies with publicly available data and a Capstone project. This gives the students a chance to benchmark their solutions against global leaderboards. Statistics and Machine Learning continue to form the backbone of the program,’ shared Dr. Mukerjee, ’but about 25% of the curriculum undergoes changes every year’. Praxis is one of the very few institutes that take ownership of placements making it an attractive option for people seeking their first break in analytics. It runs a comprehensive campus placement program, with a process in place to match skills with industry requirements. Charanpreet Singh explained the reason behind opting for this model: ‘The expressed objective of the program at Praxis is to enable an aspirant to get his\/ her first job in Analytics – and, as a natural corollary, add to the pool of Analytics- trained resources. More than 80% of our students leave promising, stable jobs in good companies to join our program, after making it through a fairly rigorous selection process. We believe that it is only fair that we put a robust process in place to enable us to create superior career opportunities for the students. We also believe that placements are not a stand-alone activity. It is a natural fallout of doing the right things – so if we have the right curriculum, pedagogy, student profile and placement process in place, placements will happen. We have provided placements to more than 90% of our students, batch-on-batch, so far.’ As organizations become more data-centric, and technology takes massive strides forward, making it easy and possible to access, store, analyze and visualize different types of data, the requirement for professionals trained in this science is poised to grow larger. No wonder, therefore, that there is mushrooming of courses – with a variety of content, duration, delivery models. How does one decide on a course design and then measure the success of a program like this? ‘We have a very clearly defined target audience – people who want to make a transition to the analytics and data science domain and make this their career,’ says Charanpreet Singh. ‘Given the breadth and complexity of this science, this requires a deep dive into the concepts, tools and techniques. We are convinced that the only way to do this is to engage aspirants in a rigorous, in-class, full-time program. Remember, we own the campus placement piece as well – so we need to ensure our students are well-trained before they face recruiter interviews. On-line and hybrid courses play a critical role as well – but for different target audiences’. We have a simple outcome-based measure to assess the success of our program – says Prithwis Mukerjee. It is the level of acceptance of our students by recruiters, followed by repeat purchase – i.e., do the companies who have recruited once, come back for the next batch? A second measure is alumni performance – we keep in close touch with our alumni and follow their progress. Partly because of our training, and partly because of the high demand for analytics professionals, our alumni have grown considerably in their careers in the last 5 years or so. And finally, if we grow our numbers year on year and batch on batch, something must be going right! A large majority of our new applicants talk to the alumni before joining – and the alumni have been consistent in endorsing the program strongly to whoever reaches out to them. Praxis debuts Business Analytics program at IIDT, Tirupati As part of Andhra Pradesh government’s commitment to making AP a technology hub, Praxis Business School has been selected as the “Academic Partner by the Government of Andhra Pradesh” to deliver the one-year full-time postgraduate program in Business Analytics at the International Institute of Digital Technologies (IIDT), Tirupati, an institute established to be the digital and innovation centre for next-gen technologies. The first batch commenced classes on February 20 this year. “It is a huge recognition, at the highest level, of the quality of work Praxis has done in the field of Analytics and Data Science, and an inspiration to continue to excel. With this, Praxis now has its footprint across Kolkata, Bangalore and Tirupati,” shared Prof Singh. Here’s the role Praxis Business School will play in mentoring, guiding and partnering with IIDT in creating and delivering world-class training programs: [box title=”Praxis Business School & IIDT will offer joint certificates, diploma and degree to graduating students” style=”soft” box_color=”#ffffff” title_color=”#563737″ radius=”5″]• Creating and delivering masters education programs to students • Creating and delivering short-term certificate and diploma training programs to corporate employees and government officers • Creating and delivering online\/ virtual training programs • Praxis Business School & IIDT will offer joint certificates, diploma and degree to graduating students[\/box] So where does the Praxis Analytics story go from here? Charanpreet Singh: To start with, we need to consolidate and continue to be at the bleeding edge in terms of content, delivery and placements. There will be a continual effort to enhance the real-world project component so that students get more hands-on practice. We are viewing growth in a couple of ways – extending our footprint to other cities; and offer our academic expertise to emerging institutes that wish to run programs in analytics and data science (the IIDT model). Moving forward, as this domain evolves further, and the batch-size becomes larger, we can certainly envision technique-specific and vertical-specific (healthcare\/ retail\/ BFSI analytics).specializations. We are increasing the breadth of our offerings by launching week-end and on-line programs for time-poor working professionals. We are committed to creating well-trained professionals and serving the analytics industry needs. [box title=”Snapshot of the program:” style=”glass” box_color=”#ffffff” title_color=”#48536d” radius=”5″]Name of the course: Post Graduate Program in Business Analytics Duration: 9 months Mode: Instructor-led classroom training program Start date: July 03, 2017 Selection process: Written aptitude test followed by personal interview Ideal profile: Program is suitable for Graduates with Mathematics and Statistics as one of the subjects such as BE\/ BSc in Mathematics, Statistics, Economics, Computer Science, IT and so on. Work experience, while not compulsory, is preferred. Students ill at ease with the world of numbers or are uncomfortable with coding, may not be able to derive the full benefits of this course.[\/box]","excerpt":"One of the earliest entrants in the analytics education space and one of the country’s leading business schools, Praxis Business School has been consistently ranked among the Top 10 analytics courses in India. Praxis kick-started the Post Graduate Program in Business Analytics in 2011, four years after the business school was set up, with an […]","categories":["AI Trends"],"tags":["Business Analytics"],"author_name":"Дарья","publish_date":"2017-04-17T09:50:31","publication_year":"2017","word_count":1822,"keywords":["data science","Go","machine learning","AI","RAG","Python","Ray","analytics","Business Analytics","TensorFlow","R"],"extracted_tech_keywords":["AI","machine learning","data science","analytics","Ray","TensorFlow","RAG","Python","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/review-pgp-business-analytics-praxis-business-school\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":54902,"title":"How Apple Will Use Self-Play To Reduce Collisions For Self-Driving Vehicles","content":"With other tech giants like Intel, Microsoft, and Amazon slowly taking a look into research and development in the self-driving sector and announcing their upcoming projects, Apple has been on the quieter side of it when it comes to being open regarding their plans on self-driving projects; which is something one doesn’t expect from Apple. Though 2020 might be the year where they finally make some big announcements or make a splash by raising the curtains on their improved designs, Apple might finally bring out Project Titan, which is a joint venture of Volkswagen and a startup they acquired in 2019, Drive.ai. This week, a paper that was published on Avix.org detailed Apple’s plans to make their self-driving project more robust, sophisticated using the deep reinforcement learning paradigm with self-play. Training AI to Merge in Traffic Because merging behaviours are complex and require accurate prediction of intentions and reactions, the traditional hard-coded behaviours lead to poor results, that is why reinforcement learning is used. RL directly learns policies through repeated interactions with an environment. Apple’s research scientist, Yichuan Charlie Tang, detailed in the paper how they have tried to demonstrate an iterative procedure of self-play. This self-play procedure can create progressively more diverse environments which will lead the agents in learning more sophisticated and robust policies. This demonstration is done in a challenging environment where agents must negotiate with each other to successfully merge on or off the road in a challenging multi-agent simulation of traffic. Though at the start the environment is quite simple for these agents to interact, as the agents learn, the complexity of the environment increases and diverse sets of agents are added to this agent ‘zoo’. Through self-play, these agents learn behaviours like defensive driving, yielding, overtaking even the use of signal lights to communicate ‘intentions’ with other agents. Apple’s Study To Make Self-Driving Better In this study, the researches have implemented self-play within a two-dimensional simulation of traffic on the geometry of real roads interpreted by satellite imagery. This virtual representation of real roads imagery from stellite was then populated of agents who were capable of lane-following and safe lane changing. If simple rule-based policies are used in an environment like these, then it becomes evident that these rule-based policies are insufficient when it comes to dealing with the complexities in the environment. So, the RL can give out better results when it comes to dealing with these complex environment and train policies in the presence of the basic rule-based agents. But the RL policy overfits to the distribution of the behaviours of the basic rule-based agents and overfitting, which is a problem. To counter overfitting, Tang and the team devised an iterative self-play algorithm where the previously trained RL policies are mixed with rule-based agents. This is done to create a diverse population of agents, and these agents can be controlled simultaneously by the policies in self-play in the simulation. As the training of these agents goes on, these agents evolve. The agents subsequently learn to play in the increasing complexity and a more diverse environment. Goals of these agents are to be capable of determining when to slow down, accelerate, pick the gap to merge into. The agents must also learn to communicate intentions to other agents through there observable behaviour or via turn indicator signals and most ultimately learn to estimate and predict the latent goals and beliefs of other agents present in the complex environment. Throughout the training processes, these agents worked in an environment of zipper merges where it is difficult because the left lane driver intends to merge with the right lane and vice versa. These merges are where the turn signals are used to negotiate who goes first and which gap is filled in a short amount of time. The Simulation Each simulation consisted of one AI-controlled agent in a zoo of rule-based agents that performed (from a lane using adaptive cruise control) tasks like slowing down and speeding up concerning the vehicle in front. Because of the reward system, gradually the AI agent replaced the rule-based agents. These were rewarded for completing a merge and travelling any speed up to 15 meters per second which is around 33.6 miles\/hr. Thirty-two simulation episodes were made to run parallel using Nvidia Titan X graphics card. In each episode, around ten agents were launched with separate random destinations. These episodes ended only after 1,000 timesteps when collisions occurred, or the goals\/destinations were reached. All this process was done in a process which consisted of three stages: 1st stage: The AI was trained in the sole presence of rule-based agents.2nd stage: The self-play was trained with 30% IDM agents, 40% of agents controlled by current policy and 30% RL agents from stage one.3rd stage: Agents were in this stage from stage 2. Problem with Rule-based During the simulation when the AI agent was with the rule-based agents, it gradually replaced them because these rule-based kept getting penalised for going out of bounds or colliding with other agents or drifting away from the centre lane. As RL rewards the agents, it the AI agent stayed in the environment. And often where a collision occurred involving AI agents, rule-based agents were found to be the culprit. These rule-based also showed a tendency to brake suddenly, and because the AI-agents is Ultra aggressive and never yield, which resulted in a collision. Outlook for Apple’s Self-Driving AI Agent The researchers conducted around 250 random trials without adding exploration noise. So, when one compares them with rules-based agents, the success rate difference was of 35%, that is, compared to the rule-based 63%, the AI agents achieved 98% success rates over rule-based and other AI-agents via the proposed self-play strategy by Tang and team. The self-trained agents learned human-like behaviour while driving. One of the reasons Apple is quiet is because the algorithm isn’t perfect. After all, these agents are still susceptible to collisions when braking and steering, but this self-play proposed way, according to Tang and team, might open the door to work with zero collisions.","excerpt":"With other tech giants like Intel, Microsoft, and Amazon slowly taking a look into research and development in the self-driving sector and announcing their upcoming projects, Apple has been on the quieter side of it when it comes to being open regarding their plans on self-driving projects; which is something one doesn’t expect from Apple. […]","categories":["Global Tech"],"tags":[],"author_name":"Sameer Balaganur","publish_date":"2020-01-30T13:57:26","publication_year":"2020","word_count":1011,"keywords":["Go","programming_languages:R","AI","emerging_tech:AI agents","programming_languages:Go","AI agents","R","startup"],"extracted_tech_keywords":["AI","R","Go","startup","AI agents","programming_languages:R","programming_languages:Go","emerging_tech:AI agents"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/how-apple-will-use-self-play-to-reduce-collisions-for-self-driving-vehicle\/","complexity_score":4,"technical_depth":8,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10011886,"title":"Facebook &#038; Its Tumultuous Relationship With AI-Based Content Moderation","content":"During a press meet recently, a Facebook spokesperson said that the social media giant would be redoubling its efforts to counter ‘harmful content’ on its platform using artificial intelligence. Reportedly, Ryan Barnes, the Facebook Product Manager of Community Integrity, said that the company would use AI to prioritise harmful content. This move is targeting at helping its over 15,000 human reviewers and moderators in dealing with reported contents. Barnes said during the press interaction, “We want to make sure we’re getting to the worst of the worst, prioritising real-world imminent harm above all.” With that being said, there have been numerous attempts in the past to bring AI into the content moderation process on Facebook’s platforms. However, not all of them have met with success. We track down some of the major efforts of Facebook in the past and how it has fared in tackling the issue. Facebook’s Efforts Towards AI-Based Moderation In the past, Facebook has used an XML method which uses a single shared encoder to train massive amounts of multilingual data. This provided an improvement over both supervised and unsupervised machine translation of low-resource languages for better detection of hate speech and harmful contents even in languages other than English. This system enabled the quality of classifiers to apply training in one language, for most cases, English, to be applied across other languages. This method was able to proactively detect harmful language and content in about 40 languages. This method soon succeeded by Whole Post Integrity Embeddings (WPIE). WPIE is a pre-trained representation of content for integrity problems. As compared to previous systems, the WPIE method was trained on a larger set of violations and training data. While introducing this method, Facebook said in its blog that the method “improves performance across modalities by using focal loss, which prevents easy-to-classify examples from overwhelming the detector during training, along with gradient blending, which computes an optimal blend of modalities based on their overfitting behaviour.” Facebook claimed that upon deployment, these tools have helped in substantially improving the performance of their integrity tools. For example, this tool was able to help in detecting almost 97.6% of 4.4 million drug sale content hosted on the platform in 2019. Earlier this year, as the COVID-19 situation was looming large, Facebook started utilising SimSearchNet, a convolutional neural net-based model, built originally to detect near-exact duplicates, for fighting misinformation. The company said that SimSearchNet was helping in end-to-end image indexing to recognise and flag near-duplicate matches. Just recently, Facebook introduced its machine translation model called M2M-100, which has been trained on 2,200 languages — about ten times the amount of training data used in the preceding model. As per the company, this model was built as a many-to-many data set with 7.5 billion sentences for 100 languages using novel mining techniques. The resulting parameters capture information from related languages and reflect a diverse script of languages and morphology. One of the salient features of this technique was found to be the fact that it did not require English as a link between two languages. Meaning, a language can be translated to another without having to be first translated into English. The ultimate goal of this model is to perform bidirectional translation between 7,000 languages to particularly benefit low-resource languages. Apart from its obvious application in communication, Facebook anticipated that the M2M-100 model would help in content moderation across a larger set of languages. How Successful Are These Moderation Techniques? In ways, more than one, 2020, with its worldwide pandemic situation and the much anticipated US Presidential Elections, the AI content moderation system of Facebook had to go through a litmus test. While there were a few hits, a lot of loopholes were left exposed. Speaking in the context of the recently concluded US elections, CEOs of Facebook and Twitter were apprehended to appear before the Senate Judiciary Committee this week on its failure to take down harmful, oft inflammatory content from the platform and also on the alleged bias. In addition to this, the voice around the health and safety of human moderators at Facebook has grown only stronger. The moderators accused the company of flouting conducive working environment practices by underpaying them and forcing them to return to work even at the height of the pandemic. It is to be noted that just in the beginning of this year, in lieu of Coronavirus, thousands of these content moderators were sent back home, and all the content moderation activities were primarily governed by the AI systems, albeit, with limited success. Apart from these specific examples, there have been repeated criticisms of Facebook’s way of handling harmful content. It has been found in more than one occasion that the platform pulled down seemingly harmless content while allowing dangerous contents to thrive. Wrapping Up Despite the new advancements that Facebook is regularly announcing on its AI-enabled content moderation, the fact remains that it is still very much dependent on the human moderators. The problem with this premise is that these human moderators, due to being exposed to hours of triggering content on a daily basis, affects their mental well-being. Additionally, they also complain of being overworked and underpaid. Viewing all these issues, it is safe to say that it will be a while before Facebook presents a truly breakthrough AI model for countering harmful, triggering, and biased content on its platform.","excerpt":"During a press meet recently, a Facebook spokesperson said that the social media giant would be redoubling its efforts to counter ‘harmful content’ on its platform using artificial intelligence. Reportedly, Ryan Barnes, the Facebook Product Manager of Community Integrity, said that the company would use AI to prioritise harmful content. This move is targeting at […]","categories":["Global Tech"],"tags":[],"author_name":"Shraddha Goled","publish_date":"2020-11-19T17:00:26","publication_year":"2020","word_count":896,"keywords":["Go","artificial intelligence","programming_languages:R","AI","Modal","RPA","ML","Aim","ViT","R"],"extracted_tech_keywords":["AI","artificial intelligence","ML","Aim","R","Go","ViT","RPA","Modal","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/facebook-its-tumultuous-relationship-with-ai-based-content-moderation\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10162335,"title":"Telangana and Andhra Pradesh, the New Hotspot for IT and GCCs","content":"While Bengaluru continues to solidify its position as India’s IT and GCC hub, a status it is likely to maintain for years to come, neighbouring states are catching up fast. For decades, Bengaluru has benefitted from its abundant talent pool, which makes it an ideal choice for companies looking to set up offices and expand their presence. Karnataka is being challenged by Andhra Pradesh and Telangana, which are trying hard to attract GCCs and Indian IT for operations. Thus, both states are the second-best options for firms. Andhra Pradesh IT minister Nara Lokesh recently announced that TCS will establish its office in Visakhapatnam within three months. However, operations will initially commence from a temporary location, likely the Millennium Towers in Rushikonda. The minister also shared updates on discussions with another IT giant, Cognizant, during his recent visit to the World Economic Forum (WEF) 2025 in Davos, hinting at positive developments to be announced soon. Lokesh stated that the government aims to create five lakh IT jobs in Visakhapatnam over the next five years. “TCS is looking for an ideal location to set up its permanent campus in the city. Once it is finalised, it may take two to three years for construction,” Lokesh said. “We are keen on exploring opportunities in deep tech, big data and AI. Our government has decided to provide land to the first 500 IT companies at very low rates compared to any other state in the country.” HCLTech was also given a similar facility in Vijayawada and has been in expansion mode since. Hyderabad Takes it Forward This is in line with HCLTech’s announcement of its global delivery operations in Hyderabad, which includes the launch of a new technology centre expected to generate 5,000 additional jobs. The announcement was made following a meeting between Telangana CM A Revanth Reddy, IT minister D Sridhar Babu, and HCLTech’s global CEO and MD C Vijayakumar at the WEF. The new centre, which spans 3,20,000 sqft, will focus on delivering advanced cloud, AI, and digital transformation solutions to global clients in industries such as high-tech, life sciences, and financial services. Speaking on HCLTech’s expansion, Vijaykumar said, “Hyderabad, with its world-class infrastructure and high-quality talent pool, has been a key location on HCLTech’s global network. The new centre will bring cutting-edge capabilities to our global client base and contribute to the local technology ecosystem. Furthermore, Jayesh Sanghrajka, group CFO of Infosys, met with Telangana IT and industries minister Duddilla Sridhar Babu at the WEF. They announced that the firm plans to expand its campus in Pocharam on the outskirts of Hyderabad. This will add around 17,000 jobs to the campus. The new IT towers, which will be constructed in the first phase with INR 750 crores, will be completed in the coming 2-3 years. Infosys currently employs around 35,000 people at its Hyderabad campus. Krishna Vij, VP of IT Staffing at TeamLease Digital, said that both Telangana and Andhra Pradesh, especially Hyderabad, have a solid pool of tech talent. “Telangana, in particular, stands out because of its large supply of fresh IT graduates and its position as a major tech hub, which draws in both domestic and global companies looking for skilled professionals,” Vij told AIM. “These states have some of the top institutions like IIITs, NITs, and a wide network of engineering colleges, so there is a steady flow of STEM graduates entering the market,” Vij added that Proactive government policies, infrastructure development, and skilling programs in these states make them attractive destinations for IT and GCC investments, further enhancing talent availability. When it comes to GCCs, according to the Q3 report by ANSR, over 450 Forbes Global 2000 companies operate over 825 GCCs across the country, employing more than 1.3 million professionals. Hyderabad (110 GCCs and 1.90 lakh employees) follows Bengaluru (285 companies), attracting interest due to its infrastructure, talent availability, and business-friendly policies. The Telangana government has signed a memorandum of understanding (MoU) with the US Chamber of Commerce’s US India Business Council, aiming to focus on collaboration in IT, AI, electronics and GCCs. Andhra Leads with Tier-II & III Cities Meanwhile, the Andhra government has introduced the Andhra Pradesh IT & GCC Policy 2021–24, aiming to incentivise and facilitate the establishment of prominent Fortune 500 companies within its borders. This policy focuses on creating a world-class electronics manufacturing infrastructure to transform the state into a hub for IT and electronics. Cities like Visakhapatnam have seen significant growth in the IT sector, with the establishment of IT Special Economic Zones and incubation centres such as the Sunrise Startup Village and Fintech Valley Vizag. These initiatives promote the city as a global fintech capital. The inauguration of Millennium IT Towers 1 and the planned Millennium IT Towers 2 further bolster the state’s IT infrastructure. Telangana, similarly, has become a preferred destination for GCCs. Notably, pharmaceutical giant Eli Lilly announced plans to establish a new global capability centre in Hyderabad and hire over 1,000 professionals. SEI Investments is considering Hyderabad for a new GCC and intends to create high-skill engineering and financial jobs over the next three years. The Telangana government has been proactive in fostering IT growth beyond Hyderabad. Collaborations with organisations like ITServe Alliance aim to create 30,000 IT jobs in Tier-2 and Tier-3 towns, promoting balanced regional development. Additionally, the state has set up IT hubs in cities such as Warangal, Khammam, Karimnagar, and others, further decentralising IT growth. “Tier-2 and Tier-3 cities stand to benefit significantly from the expansion of IT companies,” Vij said that while salaries in these regions may be lower than in Tier-1 cities due to the lower cost of living, firms are expected to offer competitive pay for specialized roles like AI, data analytics, and cloud computing. The state governments of the two states are already working towards that goal. For instance, Lokesh emphasised efforts to provide IT companies in Vizag with better bus connectivity, street lighting, and police patrolling. He mentioned upcoming discussions with the finance minister on providing incentives to IT units, reaffirming the government’s commitment to creating a robust IT and GCC ecosystem.","excerpt":"Andhra Pradesh government has decided to provide land to the first 500 IT companies at very low rates compared to any other state in the country.","categories":["IT Services"],"tags":["AI in Indian IT","Andhra Pradesh","telangana"],"author_name":"Mohit Pandey","publish_date":"2025-01-28T18:00:00","publication_year":"2025","word_count":1015,"keywords":["AI in Indian IT","Go","API","AWS","AI","cloud computing","telangana","ML","Git","Aim","analytics","R","Andhra Pradesh"],"extracted_tech_keywords":["AI","ML","analytics","Aim","cloud computing","AWS","R","Go","Git","API"],"url":"https:\/\/analyticsindiamag.com\/it-services\/telangana-and-andhra-pradesh-the-new-hotspot-for-it-and-gccs\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10087819,"title":"AWS, Hugging Face Team Up to Challenge OpenAI-Microsoft in AI Race","content":"Hugging Face today announced that it will be partnering with Amazon Web Services (AWS) in a bid to democratise machine learning. With this partnership, the duo look to accelerate the availability of next-generation machine-learning models alongside helping developers build, train, and deploy the latest ML models in the cloud using purpose-built tools. This move also comes against the backdrop of strategic partnerships between OpenAI and Microsoft, Athropic and Google, and others, which have been taking significant strides in AI, alongside the recent advancements in transformer models (GPT-3.5, LaMDA, etc), diffusion models (DALL.E, Imagen, etc), and large language model-based chatbots (ChatGPT, Bing Chat and Bard). “However, most of these popular generative AI models are not available, widening the gap of machine learning capabilities between largest tech companies and everyone else,” writes Hugging Face in its blog post. Hugs & Smiles – A Much-Needed Combo in AI\/ML With this partnership, Hugging Face will use AWS as a preferred cloud provider so developers can access AWS’s tools, including Amazon SageMaker, AWS Trainium, AWS Inferentia and others, to train, fine-tune, and deploy ML models on AWS. Hugging Face chief Clem Delangue said that accessibility and transparency are critical to sharing process and creating tools to use these new capabilities wisely and responsibly. He believes that Amazon SageMaker and AWS-designed chips will enable their team and machine learning community to convert the latest research into open-source models so anyone can build upon them. This deal will also offer developers to optimise the performance of their models for their specific use cases while lowering costs, alongside building next-generation AI models. But what’s in it for Amazon? It helps the tech giant onboard\/rope in developers into its Amazon SageMaker and AWS-designed chip universe. AWS chief Adam Selipsky believes that Hugging Face and AWS is making it easier for customers to access popular machine learning models to create their own generative AI applications with the highest performance and lowest cost. He said that this partnership shows how generative AI companies and AWS can work together to put this innovative technology into the hands of more customers. Hugging Face has become one of the most popular hubs for machine learning, consisting of more than 100,000 free and open-source machine learning models and downloaded more than 1 million times daily by researchers, data scientists, and machine learning engineers. Hugging Face said AWS is the most popular place to run models from the hub. The team also said since the start of their collaboration, Hugging Face on Amazon SageMaker has grown exponentially. In May last year, Hugging Face announced a similar partnership with Microsoft, where the company introduced Hugging Face Endpoints on Azure, a new service to turn Hugging Face models into scalable production solutions.","excerpt":"With this partnership, Hugging Face will use AWS as a preferred cloud provider","categories":["AI News"],"tags":["AI latest","dalle","foundational models","Hugging Face","Imagen","LLMs"],"author_name":"Pritam Bordoloi","publish_date":"2023-02-22T01:21:15","publication_year":"2023","word_count":454,"keywords":["foundational models","dalle","Hugging Face","machine learning","ChatGPT","OpenAI","AI","LLMs","Amazon SageMaker","ML","chatbots","AWS","Imagen","AI latest","generative AI"],"extracted_tech_keywords":["AI","machine learning","ML","generative AI","ChatGPT","OpenAI","Amazon SageMaker","Hugging Face","chatbots","AWS"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/aws-hugging-face-team-up-to-challenge-openai-microsoft-in-ai-race\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10140725,"title":"NVIDIA’s Jensen Huang Says That We are in the Era of “Hyper Moore’s Law”","content":"NVIDIA’s chief Jensen Huang in a recent interview said that the era of traditional scaling might be coming to an end. He sees the emergence of what he calls “Hyper Moore’s Law” as the future. “We’re going to be on some kind of a hyper Moore’s law curve, and I fully hope that we continue to do that,” Huang said, indicating a shift towards new paradigms in computing that extend beyond the conventional model of chip performance improvement. Historically, Moore’s Law relied on Dennard scaling and Carver Mead’s VLSI (Very Large-Scale Integration) techniques to boost chip performance. However, Huang acknowledges that these methods have reached their limits. “The two fundamental technical pillars were Dennard scaling and Carver Mead’s VLSI scaling. Both of these techniques were rigorous. However, these techniques have really run out of steam,” he explained. The future, according to Huang, lies in “codesign,” a method where both hardware and software are developed together to ensure optimal performance. “Unless you can modify or change the algorithm to reflect the architecture of the system or change and then change the system to reflect the architecture of the new software and go back and forth—unless you can control both sides of it, you have no hope,” he said. This codesign approach, which integrates hardware and software, allows for innovations in precision, such as moving from floating point operations like FP64 to FP32, and even as low as FP4, according to Huang. “The second part of it is the data centre. This is a big part of the equation. Full-stack approaches and innovative scaling strategies will reshape our data centre infrastructure,” he added. Earlier this year, when NVIDIA released Blackwell, it was said that Moore’s Law is dead. Reflecting on the rapid advancement in computing power, Huang said that in just eight years, NVIDIA has increased computational capacity by a thousandfold, a progress that far exceeds the benchmarks set during the heyday of Moore’s Law. However, he rued that even after this remarkable growth, the industry’s accelerating demands are far from being met. “In the past eight years, we’ve increased computation by 1000 times, and we have two more years to go. So that puts it into perspective [the fact that] the rate at which we’re advancing computing is insane. And it’s still not fast enough,” said Huang.","excerpt":"The future, according to Huang, lies in “codesign,” a method where both hardware and software are developed together to ensure optimal performance.","categories":["AI News"],"tags":["NVIDIA"],"author_name":"Siddharth Jindal","publish_date":"2024-11-09T10:54:49","publication_year":"2024","word_count":386,"keywords":["Go","API","programming_languages:R","AI","innovation","programming_languages:Go","NVIDIA","R"],"extracted_tech_keywords":["AI","R","Go","API","innovation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/nvidias-jensen-huang-says-that-we-are-in-the-era-of-hyper-moores-law\/","complexity_score":3,"technical_depth":7,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10063060,"title":"Lapsus$ hack leaves NVIDIA in a tight spot","content":"According to an IBM report, ransomware was the top attack type (again) in 2021. Recently, NVIDIA confirmed the hack attack that compromised their internal systems. The infamous hacker group Lapsus$ claimed credit for the attack. Later, Lapsus$ also hacked Ubisoft. Lapsus$ broke into NVIDIA’s internal network and managed to steal sensitive data–from hashed login credentials to trade secrets. The hackers wanted NVIDIA to remove the mining hashrate limiters on their RTX 3000-series GPU as ransom. Lapsus$ said if NVIDIA failed to agree to their demand by March 4, they would leak the latter’s trade secrets. And NVIDIA didn’t submit to their ransom demand. Later, the hackers leaked NVIDIA’s official code signing certificates. Now, bad actors are using them to bypass Windows Defender’s built-in executable verification and sneak in malware. The hackers can make malicious programs look like legit NVIDIA software. Lapsus$ started leaking employee credentials and proprietary information as downloadable files on the internet. NVIDIA found out about the breach on February 23. The company also said the breach would not disrupt its business. The hack happened in mid-February, and Lasus$ stole one terabyte of data, including a substantial amount of sensitive info on GPU designs, source code for an NVIDIA AI rendering system known as DLSS usernames and passwords of more than 71,000 NVIDIA employees. In the wake of the breach, NVIDIA has stepped the security, reached out to law enforcement, and is now working with cybersecurity experts to deal with the attack. In 2019, Stratosphere Labs looked at a remote access trojan (RAT) known as Quasar and said it had been used for cyberattacks against Ukraine. As per samples uploaded on VirusTotal, the stolen certificates were used to sign Cobalt Strike beacons, Mimikatz, backdoors, and RATs (malware and hacking tools). That escalated quickly #Lapsus #Nvidia #LeakedCertificate Mimikatzhttps:\/\/t.co\/TrY6vL2mEEKDUhttps:\/\/t.co\/RDf6bnuArk pic.twitter.com\/Jl4tpS5KEr— Florian Roth (@cyb3rops) March 3, 2022 In the same tweet thread, cybersecurity researchers Kevin Beaumont and Will Dormann (CERT Coordination Center) posted the serial numbers of the stolen certificates. So, no ransom? The hack compromised NVIDIA servers. Apart from the demand to remove the mining hashrate limiters on the company’s RTX 3000-series graphics cards, the hackers have also asked NVIDIA to make their drivers open-source and distribute them under the free and open-source software (FOSS) license. Later, the hackers revised their demands and called upon NVIDIA to remove the lite hash rate (LHR) in its GPUs. Less than a week after the NVIDIA breach, the hackers claimed an attack on Samsung. In a description of the upcoming leak, Lapsus$ said the hacked data contains “confidential Samsung source code” like source code for every Trusted Applet (TA) installed in Samsung’s TrustZone environment used for sensitive operations (e.g. hardware cryptography, binary encryption, access control; algorithms for all biometric unlock operations; bootloader source code for all recent Samsung devices; confidential source code from Qualcomm; source code for Samsung’s activation servers; and full source code for technology used for authorizing and authenticating Samsung accounts, including APIs and services. #Lapsus group claims they breached #Samsung#CyberAttack #ThreatHunting @SOSIntel @Cyberknow20 @GossiTheDog @campuscodi pic.twitter.com\/AQvQNLF8M1— Soufiane (@S0ufi4n3) March 4, 2022 Lapsus$ put the data under three compressed files worth 190 GB and is now available for free download in Torrent. Ubisoft also reported a “cybersecurity incident” involving Lapsus$. On March 11, Ubisoft said the hack impacted numerous games, services and functions in its internal systems. Although the company did not disclose how it happened or who did it. “As a prudent step, we began a company-wide password reset. Also, we can confirm that there is no evidence any player’s personal information was accessed or exposed as a by-product of this incident,” Ubisoft said in a statement. Lapsus$ claimed responsibility for the attack in its Telegram channel.","excerpt":"The hackers have leaked NVIDIA’s official code signing certificates.","categories":["Global Tech"],"tags":["NVIDIA GPU"],"author_name":"Akashdeep Arul","publish_date":"2022-03-20T13:00:00","publication_year":"2022","word_count":617,"keywords":["Go","API","NVIDIA GPU","programming_languages:R","AI","Scala","Git","Aim","Rust","GAN","R"],"extracted_tech_keywords":["AI","Aim","R","Go","Rust","Scala","Git","API","GAN","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/lapsus-hack-leaves-nvidia-in-a-tight-spot\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":618,"title":"Genpact and IMT, Ghaziabad sign MoU to develop an analytics elective","content":"IMT Ghaziabad and Genpact Limited have signed a MoU to develop and implement analytics elective for the two year PGDM program. The term of the MoU is five years and students of the program will be eligible for employment in Genpact’s analytics practice upon completion of the course. IMT Ghaziabad would develop and design the course in detail whereas Genpact will add the industry perspective in the course conceptualization, skills development and the training faculty. Genpact will also provide the necessary support through guest lectures and arranging mid-term and extended internship projects for the IMT Ghaziabad students in order to gain experience on real time data, projects and research papers at Genpact. The course aims to provide both theoretical and practical experience in analytics as applied in different industries. Talking about the MoU, Dr. Bibek Banerjee, Director, IMT Ghaziabad said, “This is a great opportunity for our students as this program gives them the liberty to speed track their careers in one of the fastest growing areas in the industry. Our students will also get valuable industry experience which will help them grow and build successful careers.”  He also added, “This initiative is yet another step towards the bigger goal of a new academic curriculum that IMT has developed around the theme of ‘Global Breadth, India Depth’, a first of its kind in the arena of management education. Further, it is a significant example of synergistic collaboration between academia and industry that is designed to focus on future needs of the economy – a cause that IMT Ghaziabad is firmly committed to.” “The emergence of big data, regulatory changes and social media are causing a big shift in the way businesses operate and students of IMT will learn how to combine process, analytics and technology to make organisations smarter in this dynamic new world, ” said Pankaj Kulshreshtha, senior vice president and business leader — Smart Decision Services — analytics and research, Genpact.","excerpt":"IMT Ghaziabad and Genpact Limited have signed a MoU to develop and implement analytics elective for the two year PGDM program. The term of the MoU is five years and students of the program will be eligible for employment in Genpact’s analytics practice upon completion of the course. IMT Ghaziabad would develop and design the […]","categories":["AI Trends"],"tags":[],"author_name":"Дарья","publish_date":"2012-08-02T18:01:48","publication_year":"2012","word_count":323,"keywords":["big data","Go","AI","ML","BERT","Aim","llm_models:BERT","analytics","GAN","R"],"extracted_tech_keywords":["AI","ML","analytics","Aim","R","Go","big data","BERT","GAN","llm_models:BERT"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/genpact-and-imt-ghaziabad-sign-mou-to-develop-an-analytics-elective\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10126247,"title":"Databricks Acquires Prodvana to Enhance Cloud Infrastructure Capabilities","content":"Databricks, a leading data and AI company, has announced the acquisition of Prodvana, a cloud-native infrastructure management startup. The move aims to bolster Databricks’ multi-cloud platform capabilities and improve its infrastructure deployment processes. Prodvana, known for its innovative approach to simplifying cloud-native environment complexities, brings expertise in managed delivery, infrastructure insights, and self-healing capabilities. This acquisition is expected to enhance Databricks’ ability to deploy faster without compromising security, ultimately benefiting its customers. Andrew Fong, Co-Founder and CEO of Prodvana, expressed enthusiasm about joining Databricks, saying, “We founded Prodvana on the belief that intelligent automation and AI can truly transform infrastructure into a force multiplier that ensures correctness, error-free operations, and R&D acceleration. We are excited to join a company that shares our vision and commitment to innovation in infrastructure.” The Prodvana team, led by Fong and co-founder Naphat Sanguansin, brings extensive experience in building scalable and secure infrastructures for major distributed systems, including YouTube and Dropbox. Their expertise in scaling infrastructure to support hundreds of millions of users aligns well with Databricks’ goals. Databricks’ platform currently launches over 20 million virtual machines daily across major cloud platforms, enabling customers to run data and AI workloads with enterprise-grade scalability, reliability, and consistency. The integration of Prodvana’s technology is expected to further enhance these capabilities. This acquisition follows Databricks’ recent purchases of Tabular for $1 billion and MosaicML for $1.4 billion in June 2023, demonstrating the company’s continued focus on expanding its AI and infrastructure capabilities. Databricks, founded in 2013 by the creators of Apache Spark, has been at the forefront of developing data lakehouse technology, combining the features of data warehouses and data lakes. As Databricks continues to grow and innovate in the data and AI space, the addition of Prodvana’s expertise is anticipated to strengthen its position in the market and improve its offerings to customers worldwide.","excerpt":"Prodvana, known for its innovative approach to simplifying cloud-native environment complexities, brings expertise in managed delivery, infrastructure insights, and self-healing capabilities","categories":["AI News"],"tags":["AI Infrastructure","Databricks","Mergers and Acquisitions"],"author_name":"Siddharth Jindal","publish_date":"2024-07-09T09:46:56","publication_year":"2024","word_count":307,"keywords":["Go","AI","ML","data warehouse","Apache Spark","Scala","Aim","AI Infrastructure","Mergers and Acquisitions","R","data lake","Databricks"],"extracted_tech_keywords":["AI","ML","Aim","Apache Spark","Databricks","R","Go","Scala","data warehouse","data lake"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/databricks-acquires-prodvana-to-enhance-cloud-infrastructure-capabilities\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":59492,"title":"Top Hyperparameter Optimisation Tools","content":"A key balancing act in machine learning is choosing an appropriate level of model complexity: if the model is too complex, it will fit the data used to construct the model very well but generalise poorly to unseen data (overfitting). And if the complexity is too low, the model won’t capture all the information in the data (underfitting). In a deep learning context, a model’s performance depends heavily on the hyperparameter optimisation, given that the vast search space of features, evaluation of each configuration can be expensive. Generally, there are two types of toolkits for HPO: open-source tools and services that rely on cloud computing resources. In the next section, we list down a few tools that have helped in making hyperparameter optimisation easier: Google Vizier Google Vizier was introduced to offer scalable service where the users can choose an involved search algorithm and submit a configuration file, and then they will be provided with a suggested hyper-parameter set. In spite of its closed infrastructure nature, Google Vizier can still be used to make easy changes or design new algorithms for HPO. For those who are new to hyperparameter tuning, they can play with the dashboard, suggested with an efficient search algorithm and early stopping strategy, use the allocation of Google Cloud, and finally check the results on the UI with very little effort. Amazon Sagemaker’s Auto Tuning The automatic model tuning module offered by Sagemaker simplifies the method of model building and deployment. As a service, it is comparable to Google Vizier with the support of Amazon Web Services (AWS). The most attractive feature is that one needs just a model and related training data for a HPO task. It supports optimisation for complex models and datasets with parallelism on a large scale. Moreover, the involvement of Jupyter simplifies the configuration for optimisation and visualisation of results. Microsoft’s NNI Microsoft’s Neural Network Intelligence (NNI) is an open-source toolkit for both automated machine learning (AutoML) and HPO that provides a framework to train a model and tune hyper-parameters along with the freedom to customise. In addition, NNI is designed with high extensibility for researchers to test new self-designed algorithms. NNI is also compatible with Google’s TensorBoard and TensorBoardX. Tune Ray Tune is a library developed by Berkeley’s RISELab. As a distributed framework for model training, Ray guarantees the efficient allocation of computational resources. Most searching methods can be realised in parallel with the help of Tune library. Tune uses a master-worker architecture to centralise decision-making and communicates with its distributed workers using the Ray Actor API. Ray provides an API that enables classes and objects to be used in parallel and distributed settings. Tune uses a Trainable class interface to define an actor class specifically for training models. This interface exposes methods such as _train, _stop, _save, and _restore, which allows Tune to monitor intermediate training metrics and kill low-performing trials. pip install 'ray[tune]' torch torchvision HpBandSter The creators call this library as a distributed Hyperband implementation on Steroids. This Python 3 package is a framework for distributed hyperparameter optimisation. It started out as a simple implementation of Hyperband, a novel algorithm for HPO, and contains an implementation of Robust and Efficient Hyperparameter Optimization. pip install hpbandster Hyperopt: Distributed Hyperparameter Optimisation Hyperopt is a Python library for serial and parallel optimisation over real-valued, discrete, and conditional search spaces. Currently, Hyperopt supports these three algorithms: Random SearchTree of Parzen Estimators (TPE)Adaptive TPE The parallelisation of all algorithms done using: Apache SparkMongoDB Installation : pip install hyperopt Facebook’s HiPlot Facebook AI’s HiPlot had been used by the developers at Facebook AI to explore hyperparameter tuning of deep neural networks with dozens of hyperparameters. They have done more than 100,000 experiments with this tool. Now they have open-sourced it. This tool helps practitioners to visualise the influence of hyperparameters on a certain task. One can use this tool straightaway as it consists of simple operations such as dragging and sliding the cursor and witnessing the changes in real-time. The usage is somewhat similar to an equaliser in music player settings on our phones. For an extensive read on the state of hyperparameter optimisation, check this review.","excerpt":"A key balancing act in machine learning is choosing an appropriate level of model complexity: if the model is too complex, it will fit the data used to construct the model very well but generalise poorly to unseen data (overfitting). And if the complexity is too low, the model won’t capture all the information in […]","categories":["AI Trends"],"tags":["Automl","balancing classes","big data and data science","hyperparameter tuning","optimisation techniques","RPA developer"],"author_name":"Ram Sagar","publish_date":"2020-03-23T15:00:20","publication_year":"2020","word_count":693,"keywords":["Automl","RPA developer","machine learning","Amazon SageMaker","AI","neural network","cloud computing","big data and data science","ML","hyperparameter tuning","RAG","Ray","deep learning","optimisation techniques","Jupyter","balancing classes"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","neural network","Amazon SageMaker","Ray","Jupyter","RAG","cloud computing"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-hyperparameter-optimisation-tools-neural-networks\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10173152,"title":"Hugging Face’s Latest Small Language Model Adds Reasoning Capabilities","content":"Hugging Face has released SmolLM3, a 3B parameter language model that offers long-context reasoning, multilingual capabilities, and dual-mode inference, making it one of the most competitive small-scale open models to date. The model is available under the Apache 2.0 license. Trained on 11.2 trillion tokens, SmolLM3 outperforms other models in its class, including Llama-3.2-3B and Qwen2.5-3B, while rivalling larger 4B models such as Gemma3 and Qwen3. The model supports six languages, including English, French, Spanish, German, Italian, and Portuguese, and can process context lengths of up to 128k tokens, enabled by NoPE and YaRN techniques. The release includes both a base model and an instruction-tuned model with dual reasoning modes. Users can toggle between different flags to control whether the model generates answers with or without reasoning traces. Pretraining was conducted over three stages with evolving mixes of web, code, and math datasets. A mid-training phase extended the model’s context length and added general reasoning capabilities, followed by supervised fine-tuning and preference alignment using Anchored Preference Optimisation (APO). SmolLM3 achieved strong results across 12 benchmarks, ranking high on knowledge and reasoning tasks and demonstrating strong multilingual and coding performance. Instructing and reasoning modes yielded further gains on tasks like LiveCodeBench and AIME 2025. The full training recipe, including data mixtures, ablations, synthetic data generation, and model alignment steps, has also been made public on its GitHub and Hugging Face pages. This open approach aims to help the research community replicate and build on SmolLM3’s performance. A few months back, Hugging Face launched SmolLM2, an open-source small language model trained on 11 trillion tokens, including custom datasets for math, code, and instruction-following. It outperforms models like Qwen2.5-1.5B and Llama3.2-1B on several benchmarks, particularly MMLU-Pro, while achieving competitive results on others, like TriviaQA and Natural Questions. It appears that Hugging Face is focusing on minor but consistent improvements for its small language models.","excerpt":"SmolLM3, as an open model, outperforms other models in its class, including Llama-3.2-3B and Qwen2.5-3B, while rivalling larger 4B models such as Gemma3 and Qwen3.","categories":["AI News"],"tags":["Hugging Face"],"author_name":"Ankush Das","publish_date":"2025-07-09T17:06:14","publication_year":"2025","word_count":311,"keywords":["Replicate","Hugging Face","synthetic data","AI","ML","Git","Aim","small language models","GitHub","R"],"extracted_tech_keywords":["AI","ML","Aim","Hugging Face","small language models","R","Git","GitHub","synthetic data","Replicate"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/hugging-faces-latest-small-language-model-adds-reasoning-capabilities\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10011846,"title":"Guide To Labelbox &#8211; The Customizable Data Annotator Tool","content":"Data Annotations have evolved in recent years and become better at the performance with advanced computer vision and deep learning techniques. Earlier algorithms only focused on bounding boxes(the rectangle encompassing objects) but now annotating techniques enable customized shapes for any kind of object to be identified. Many of these annotator tools provide end to end ML platforms, from data accumulation to production services. Many AI-based companies are adapting to these annotations for efficient workflow management and iterating learning while training models. Customising annotations can be applicable to all kinds of use cases. Today we will be discussing a rapid data annotator tool called LabelBox, which has been a market ruler over two years and relied on for many industry use cases. Labelbox Labelbox was released in September 2018, by founders Dan Rasmuson, Brian Rieger and Manu Sharma. LabelBox allows users to manage their data using their high powered AI-enabled tools for data labelling by automating the labelling process and training models for active learning and has API support. It allows us to invite team members and collaborate over the workflows. Allows importing and exporting of different kinds of annotation formats. Complex ontology providing high-quality labels with minimal errors. Cloud services support on Azure, GCP, Sagemaker and many others. Labelbox allows customization of the tools to support your specific use case, including custom attributes, instances and much more. The bounding box, Points & lines, Polygons Instance segmentation toolkit (pen & superpixels) Superpixel- allows the instance to split into different ranges of pixels and analyse the parts of the object. https:\/\/youtu.be\/KPiHk_QmT7g Draw over objects – This tool allows to draw around the object edges https:\/\/youtu.be\/FLvcfTeQYIs Brush – This tool works like normal paint brush with different radius Eraser Supports complex ontologies with nested classifications Named Entity Recognition and Text classification https:\/\/youtu.be\/wF99DqjRnYg Support for tiled imagery (slippy maps)- this is used for geospatial data https:\/\/youtu.be\/HhevLyrh1ls Custom labels using labelbox-api.js Real-Time usage Python SDK Latest version- 2.4.9 pip install labelbox Project Setup(client initialisation and data connection): from labelbox import Client client = Client() project = client.create_project(name=\"<project_name>\") dataset = client.create_dataset(name=\"<dataset_name>\", projects=project) Graph QL API LabelBox GraphQL API is query-based and thus more flexible than RestAPIs. It has features like strongly typed schema, hierarchical architecture, specificity and strong tooling. Solutions And Services: Document data extractionSafety monitoringManufacturing – Preventative maintenance, Defect detection, Waste management, Robotics automationHealth\/medical – Digital pathology, UltrasonographyInsurance – Property inspectionDrone\/Aerial – Solar inspectionConsumer – Content moderation, Sports analytics, Thermal sensing, Generative design, Cashierless checkoutAgriculture – Crop weed detection, Livestock monitoringTransportation – Driver safety Use Cases MIT students are using Labelbox with neural networks in serotonin research to automate tasks.Stanford CS230 deep learning master grad research students use in their project for land urban air vehicles through satellite imagery. One of the winning teams in the RoboSub competition had built autonomous underwater robots.Researchers at the Institute of Industrial, University of Tokyo are using model-assisted labelling to speed up annotation efficiency.Labelbox supports American Family Insurance Automation. Companies Using LabelBox: Used by over 150+ companies to manage their workflows and collaborations. Cape Analytics uses active learning and APIs to get faster AI production.Pathware uses in pathology products by delivering AI-enabled analyse Arturo uses it for the insurance industry.Omdena is used for labelling tasks in deep learning for tree identification.SomaDetect is used for dairy farming.Lytx is the market leader for telematics saving lives on the road through video surveillance.Genius Sports – AI transformation in SportsConde Nast – The parent company for 20 media companiesNtConcepts – faster training and deployment of AI systemsXarvio uses it for the agriculture industry to optimise crop production.","excerpt":"Today we will be discussing a rapid data annotator tool called LabelBox, which has been a market ruler over two years and relied on for many industry use cases.","categories":["Deep Tech"],"tags":["annotation","bounding boxes","data annotation"],"author_name":"Jayita Bhattacharyya","publish_date":"2020-11-19T13:00:27","publication_year":"2020","word_count":594,"keywords":["bounding boxes","text classification","GCP","AI","neural network","ML","annotation","computer vision","Python","data annotation","deep learning","analytics","Azure"],"extracted_tech_keywords":["AI","ML","deep learning","neural network","computer vision","analytics","text classification","Azure","GCP","Python"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/labelbox\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":47934,"title":"Will US Blacklisting Chinese AI Firms Escalate Trade War?","content":"The recent announcement from the US Commerce Department on the blacklisting of 28 Chinese artificial intelligence companies has sent a clear signal that the trade war between the two superpowers is only going to escalate. The reason for the embargo is that the companies provide AI technologies like surveillance cameras, facial recognition software and other technologies which are used for suppression of human rights and surveillance of the Muslim minority community Uighurs in the Xinjiang state of China. The embargo means American companies are prohibited from exporting hardware components to the 28 AI organisations listed. Experts have for long highlighted human right violations in the country and how AI companies are winning contracts to provide technologies which are used are for government’s objectives of surveillance on the Chinese Muslim minority. Some of the most prominent Chinese AI firms which were blacklisted work in facial recognition (Megvii, SenseTime, Yitu), voice recognition (iFlytek), digital forensics (Xiamen Meiya Pico) and video-surveillance manufacturing (Hikvision and Dahua). Which Are The Most Prominent Blacklisted Chinese AI Companies? SenseTime, considered the most valuable AI startup in the world at $7 billion, was included in Entity List. The company is backed by companies like SoftBank, Alibaba, and  Qualcomm, and also has a research partnership with MIT. The export ban has also forced MIT to review its research partnerships with companies like SenseTime, stating it will modify its future collaborations based on the government’s embargo. One of the biggest suppliers of surveillance technology Hikvision also found its name on the US export ban list for this reason. The firm had won multiple security contracts from the government for creating surveillance systems, including one known as social prevention and control system, amounting to 1.85 billion yuan ($260 million) in the Xinjiang region. With 1\/3rd of its revenue coming from outside of China, the firm may get impacted by the US export ban list. Alibaba-backed Megvii, which is valued at $4 billion, also found itself on the US export ban list due to its facial recognition technology being used in surveillance systems and citizen law enforcement systems. The listing has undoubtedly damaged the firm’s short term prospects related to its coming IPO in Hong Kong Stock Exchange where its US partner Goldman Sachs said it was reviewing its role in the partnership. The Aftermath The announcement of US blacklisting of Chinese AI firms is only going to make matters complicated as far global trade goes. Analysts had been expecting relief in the tensions between the two nations before the upcoming talks between President Trump and Xi Jinping. The move is being suggested as a bad sign for the worsening global economy, whereas human right activists are celebrating it very much on social media channels. US had earlier placed five other organisations on its Entity List including a supercomputing company that sourced its components from US companies  Huawei too had been blacklisted by the US owing to concerns on spying on American users. Also, the export ban will hurt the US economy too as those companies may lose a substantial chunk of customers. Companies like Hikvision and iFlyTek have already announced they have been preparing for such a situation ever since the US-China trade war came to light. The firms are working on getting its supply of chips from non-US as well as home-grown companies. “We will not be strangled,” said iFlyTek’s chief executive in a company letter. Will China Boost Its Home-Grown Chip Capabilities? While China is the biggest manufacturing hub in the world, its domestic technology companies still have to rely on US semiconductor companies like Intel, Nvidia, Advanced Micro Devices and Qualcomm for hardware components, many of who have manufacturing units in neighbouring Taiwan and contracts with Taiwan Semiconductor Manufacturing Company (TSMC). According to analysts, China’s largest chip-maker, Semiconductor Manufacturing International Corp. is still way behind foreign companies in terms of the tech know-how for designing advanced chips. The latest blacklist gives more incentive for China to advance their chip manufacturing prowess, say experts. The government has long recognised the lack of AI chips capabilities in the country. To bridge the gap, top domestic tech giants like Huawei and Alibaba recently introduced specialised chips that cater to the scale of AI workloads in the country. Now that US companies have been barred from selling their hardware to many Chinese companies, the government and tech industry players may enhance their efforts further to compete in the global AI race.","excerpt":"The recent announcement from the US Commerce Department on the blacklisting of 28 Chinese artificial intelligence companies has sent a clear signal that the trade war between the two superpowers is only going to escalate. The reason for the embargo is that the companies provide AI technologies like surveillance cameras, facial recognition software and other […]","categories":["AI Features"],"tags":["AI in china","AI race","china ai"],"author_name":"Vishal Chawla","publish_date":"2019-10-15T11:48:40","publication_year":"2019","word_count":739,"keywords":["Go","artificial intelligence","AI race","programming_languages:R","AI","IPO","Scala","Git","china ai","GAN","AI in china","R","startup"],"extracted_tech_keywords":["AI","artificial intelligence","R","Go","Scala","Git","GAN","startup","IPO","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/will-us-blacklisting-chinese-ai-firms-escalate-the-global-trade-war\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10067454,"title":"Neruppu Da! This programming language is packed with punch dialogues of Superstar Rajinikanth","content":"Programmer Aadhithya Sankar had a light bulb moment six years ago when he came across an esolang called ArnoldC. Since then, he wanted to create an esolang for Indian movie icon, Rajinikanth. Recently, he released Rajini++, a programming language that runs on Python, wherein the syntax and keywords are based on the dialogues of Superstar Rajinikanth. https:\/\/t.co\/eA9kStviwUPresenting rajini++, a programming language based on iconic #SuperstarRajinikanth @rajinikanth dialogues! #rajiniPlusPlus (1)— Aadhithya Sankar (@asankar96) May 17, 2022 How does Rajini++ work? Rajini++ or Rajinipp runs on Python version 3.8 or higher. To install the interpreter, use the command: ‘pip install rajinipp’. Rajini++ supports the following features: Math Ops (SUM, SUB, MUL, DIV, MOD)Unary Ops (SUM, SUB)PrintingVariable declarationVariable accessVariable manipulation and assignmentDatatypes(Bool, String, Number)Logical ops (≥, >, ≤, <, ==, !=)If statementIf-Else statementFor loopWhile loopFunctionsFunctions with return The rajini++ programs are stored in .rpp files. To run a rajini++ program, you just need to run the following command: rajinipp run path\/to\/my_program.rpp. The rajinipp Python package also provides an interactive shell (experimental) to try out rajini++ commands. Since the rajinipp interpreter is written completely in Python, it is possible to run rajini++ code inside Python scripts. However, Aadhithya said, the programming language is not built for enterprise-grade software development.","excerpt":"Rajini++ or rajinipp runs on Python version 3.8 or higher.","categories":["AI News"],"tags":[],"author_name":"Kartik Wali","publish_date":"2022-05-19T13:57:52","publication_year":"2022","word_count":205,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","Python","programming_languages:Python","R"],"extracted_tech_keywords":["AI","Python","R","Go","programming_languages:Python","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/neruppu-da-this-programming-language-is-packed-with-punch-dialogues-of-superstar-rajinikanth\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":34604,"title":"Rising 2019 Announces Women in AI Leadership Awards – Celebration Of Tech Leaders Driving Disruption &#038; Innovation","content":"This March, on International Women’s Day, Analytics India Magazine presents Rising 2019 – a one-day conference which will bring India’s leading women data scientists and tech leaders on a platform to start exciting conversations with experts working in the field of AI and machine learning. Celebrating the importance of women in artificial intelligence, the conference will recognize the achievements of technology leaders who have made an incredible impact in the tech community by developing AI solutions with its inaugural – Women in AI Leadership Awards 2019. The awards will recognize the most influential women technology leaders in India who have demonstrated considerable experience in data science and AI leadership, advance AI development in organizations and became leaders and role models in the analytics industry. “The Women in AI Leadership Awards is aimed at honouring the outstanding achievements of women leaders who have made immense contribution to the data science community in India. These leaders don’t just excel in their professional setting but also have used their talents to make a positive impact on the tech community by starting pathbreaking programmes to drive inclusivity and growth,” said Bhasker Gupta, Founder & CEO, Analytics India Magazine. Nominations are now open for the Women in AI Leadership awards that celebrate excellence in AI leadership and innovation. Leading organizations and individuals can submit nominations for the highest achievers in tech leadership and nominate people who have made powerful contributions and demonstrated expertise in AI and driving business value. We also invite nominations from individuals who believe they have been at the forefront of disruption, have played a key role in helping use AI as a driver for growth and business value. All submissions will be assessed by our panel of editors and industry veterans and awardees will be selected after a careful review and benchmarked against the best-in-class performance standards. One of the key highlights of the conference will be sparking a conversation on the gender imbalance that exists in the industry and discussing vital steps that can be taken to increase women participation in data science and AI and help them thrive in leadership roles. Driving the conversation forward will be companies that have played a pivotal role in striving for diverse, inclusive and engaging environment for their employees, have addressed these biases to retain female talent and supported them to advance to senior roles. The awards ceremony will take place on March 8, at the Hotel Taj. The event is open to all and attendees can enjoy an evening of networking and discussions around one of the most rapidly advancing technology – AI and support women in the industry. For more on The Rising 2019, March 8 at Hotel Taj in Bangalore, click here. To send nominations, click here","excerpt":"This March, on International Women’s Day, Analytics India Magazine presents Rising 2019 – a one-day conference which will bring India’s leading women data scientists and tech leaders on a platform to start exciting conversations with experts working in the field of AI and machine learning. Celebrating the importance of women in artificial intelligence, the conference […]","categories":["Deep Tech"],"tags":["leader of ai","Women in AI Leadership awards"],"author_name":"Richa Bhatia","publish_date":"2019-02-07T09:03:58","publication_year":"2019","word_count":457,"keywords":["data science","API","machine learning","artificial intelligence","AI","Women in AI Leadership awards","Aim","ViT","analytics","GAN","R","leader of ai"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","data science","analytics","Aim","R","API","GAN","ViT"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/rising-2019-announces-women-in-ai-leadership-awards-celebration-of-tech-leaders-driving-disruption-innovation\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10117294,"title":"MongoDB has Over 3,000 Customers in India and Growing","content":"MongoDB loves India. Boris Bialek, the field CTO of MongoDB, was in Bengaluru earlier this month and in an exclusive interaction with AIM, he said: “India’s market momentum is tremendous. The growth is huge, and we have over 3,000 customers here. We experience millions of downloads every month and have an exceptionally active developer community here.” In India, MongoDB, the king of the NoSQL database, serves unicorns, smaller startups, and digital native companies, including those specialising in generative AI and digital transformation initiatives. Some of its notable customers include Zomato, Tata Digital, Canara HSBC Life Insurance, Tata AIG, and Devnagri. India cares about data security like no other. With data and AI sovereignty on the rise, there’s a demand for keeping data within a controlled environment, either on-premises or in a cloud, but not over public APIs. This is where MongoDB comes in—it acts as a bridge, integrating various components into a cohesive system. “Our approach emphasises simplicity, transparency, and trust. We make things clear around how vectors are used and provide transparency in data usage. This level of clarity addresses concerns about system trustworthiness and what is being built,” he added. Other NoSQL databases like Redis and Apache Cassandra are also widely used by Indian developers. The former has over 12 million daily downloads and derives 60% of its revenue from national database projects. Apache Cassandra has a strong presence of 14.95%, with companies like Infosys, Fujitsu, and Panasonic using it. Amazon DynamoDB holds a market share of approximately 9.5%. Focuses on Real-time “We no longer see ourselves just as a NoSQL database. We’re part of a larger picture, providing services for banking transactions, multi-document handling, search capabilities, and integrating edge components for global manufacturing lines,” said Bialek. This is where the concept of a ‘developer data platform’ emerges, and it’s a significant change the team has observed over recent years. It’s about accelerating integration without the need to maintain multiple disconnected systems. “We’re discussing real-time data. Everything today demands immediacy, whether it’s a UPI transaction that needs decisions in milliseconds or providing individualised, real-time data like online stock trades,” he added. Behind this is a vast amount of JSON data, which is what the JSON document model in MongoDB is about. Several companies achieved remarkable improvements in efficiency and customer satisfaction through MongoDB solutions. Bialek spoke about an Italian energy company that was able to slash its service desk response time from a day to two minutes! Meanwhile, a German company pioneered chatbot systems utilising self-trained LLMs for tailored client interactions, such as rapid insurance claim processing. In the retail sector, data analysis led to a 50% reduction in return shipments for an online shoe retailer by suggesting optimal shoe sizes, demonstrating the mutual benefits of predictive services for customers and businesses. Generative AI Struggles During recent customer interactions, Bialek noted several challenges regarding speeding up and implementing new technology like generative AI. “No matter if the company is small or large, the main challenge is accelerating business processes. Developers are in high demand, so the focus is on how fast we can create new platforms like payment systems and integrate new technologies like UPI LITE with limited help. We’re collaborating with major payment providers to address these challenges,” Bialek added. Secondly, customers are experiencing a stark difference between companies that are aggressively implementing generative AI and have clear directions, vis à vis others that are still figuring it out. So, MongoDB assists them in understanding and applying use cases. MongoDB to the Rescue In November last year, the company added new features to MongoDB Atlas Vector Search that offers several benefits for generative AI application development. “The feedback has been overwhelmingly positive, citing its ease of use. Some clients have even built solutions in half a day, a task that would have previously taken six months and a consultant,” Bialek commented. Furthermore, LLMs require storage for vectors, and MongoDB simplifies access to these. The emphasis is on delivering a holistic solution by integrating data layers, applications, and underlying technologies. In January of this year, MongoDB partnered with California-based Patronus AI to provide automated LLM evaluation and testing for business clients. This partnership merges Patronus AI’s functions with MongoDB’s Atlas Vector Search tool, creating a retrieval system solution for dependable document-based LLM workflows. Customers can build these systems through MongoDB Atlas and utilise Patronus AI for evaluation, testing, and monitoring, enhancing accuracy and reliability. “I would say that  MongoDB is a bit like the glue in the system, bringing everything together in a straightforward manner,” said Bialek. What’s Next? Going forward, Bialek noted that the primary goal is to focus on improving user-friendliness through simplicity, real-time data processing, and application development. “We are releasing a new version of our product this year, as we do annually. We’re currently showcasing our Atlas Streams in public preview, which is crucial for real-time streams, processing, and vector search integration. There’s a lot more in development, too,” concluded Bialek. Read more: Is MongoDB Vector Search the Panacea for all LLM Problems?","excerpt":"The king of NoSQL is now looking to attract more customers with its real-time database capabilities and data security offerings.","categories":["AI Features"],"tags":["Interviews and Discussions","MongoDB"],"author_name":"Shritama Saha","publish_date":"2024-03-28T13:04:29","publication_year":"2024","word_count":839,"keywords":["Go","AI","Redis","MongoDB","RAG","Aim","generative AI","SQL","Rust","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","generative AI","Aim","RAG","Redis","MongoDB","R","SQL","Go","Rust"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/mongodb-has-over-3000-customers-in-india-and-growing\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":29024,"title":"Teradata Launches New Pervasive Data Intelligence Platform","content":"Teradata , the industry famous data intelligence company, today announced a new platform named Teradata Vantage. Vantage will be Teradata’s next-generation analytics platform and is now available to all customers. According to the press release by Teradata, Vantage will allow enterprises to detect and uncover actionable insights to the biggest business problems by improving the analytics function. Vantage will help enterprises to uncover business value. Vantage claims also provides access to a wide variety of descriptive, predictive and prescriptive analytics which aid autonomous decision making for enterprises. Vantage also has machine learning and visualization functions which can be deployed across public clouds, on-premises, and commodity infrastructure. Oliver Ratzesberger, Teradata’s Chief Operating Officer said in the statement, “Teradata Vantage, the platform for pervasive data intelligence, is the answer to the frustration of today’s approach to analytics. Only Vantage can leverage all the data, all the time, so you can analyze anything, deploy anywhere and deliver analytics that matter. Vantage dismantles the complexity and rises above the inadequacies of today’s analytics landscape. By making it simple for businesses to operationalize analytics on an enterprise-ready platform, Vantage can uncover the intelligence trapped in their data to discover answers that matter.” In the press release Teradata also says that Vantage helps users to extract trusted, actionable insights from their data sources within an ecosystem that supports advanced analytic functions. This platform adds to the ecosystem which includes Teradata’s recently introduced 4D Analytics. The newly released platform also provides storage and analysis for multi-structured data which includes data formats like JSON, BSON, AVRO, CSV, and XML. Souma Das, Managing Director, Teradata India added further talking about the Vantage platform, “With ever increasing data volumes at their disposal organizations are under immense pressure to not only derive meaningful insights from data but to also convert that data into key actions. Data Generation is set to multiply and become more complex with adoption of newer technologies like internet of things, artificial intelligence, cloud computing and mobility.”","excerpt":"Teradata , the industry famous data intelligence company, today announced a new platform named Teradata Vantage. Vantage will be Teradata’s next-generation analytics platform and is now available to all customers. According to the press release by Teradata, Vantage will allow enterprises to detect and uncover actionable insights to the biggest business problems by improving the analytics function. […]","categories":["AI News"],"tags":["teradata"],"author_name":"Abhijeet Katte","publish_date":"2018-10-08T19:28:37","publication_year":"2018","word_count":329,"keywords":["teradata","machine learning","artificial intelligence","AI","cloud computing","ML","RAG","Aim","analytics","Rust","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","analytics","Aim","RAG","cloud computing","R","Rust"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/teradata-launches-pervasive-data-intelligence-platform-teradata-vantage\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":43868,"title":"What’s The Whole Commotion Regarding China’s Tianjic Chip About?","content":"Researchers and technologists around the globe are now striving to develop artificial general intelligence (AGI) in order to achieve one common target — build efficient models which are more human-like. Now, China has been pushing itself to accelerate AI quite seriously and has been doing massive investments in this sector. According to reports, China’s spending on AI for technology and communication industry has increased to 71.6% during 2018 to reach $320 million. It is expected that their spending on AI technologies will increase at a CAGR of 25.9%, increasing from $481.2 million in 2019 to reach $2,406.1 million by 2025. Last month, a team of Chinese researchers introduced a new hybrid chip known as Tianjic chip to stimulate AGI development by paving the way to more generalised hardware platforms. This chip is a hybrid platform which supports computer science-oriented ML algorithms, as well as neuroscience-inspired models and algorithms. In order to support the parallel processing of large networks or multiple networks concurrently, the chip adopts a many-core architecture with scattered localised memory for timely and seamless communication. Fusing Neuroscience With ML The ML approach to AGI involves explicit algorithms which are executed on computers. For instance, the artificial neural network is inspired by the cortex in terms of spatial complexity which is being implemented in a number of applications such as speech recognition, image classification, language processing, etc. On the other hand, the neuroscience-based approach to AGI attempts to mimic the cerebral cortex which is based on observations of the interaction between memory and computing, rich spatiotemporal dynamics, spike-based coding schemes, and various learning rules. Both the approaches, solve problems efficiently in their specific domains. However, achieving AGI is still a dream for the researchers and to accomplish this dream, researchers are incorporating more biologically inspired models and algorithms into the prevailing artificial neural networks. Keeping ML and neuroscience in mind, the researchers tried to build an AGI system which has several features such as support for vast and complex neural networks that can represent rich spatial, temporal and spatiotemporal relationships, support for hierarchical, multi granular and multidomain network topologies, support for a wide range of models, algorithms and coding schemes and  support for the intertwined cooperation of multiple specialised neural networks that are designed for different tasks in parallel processing. Architecture Fig: Hybrid Architecture Of Tianjic Chip (Source) The Tianjic chip consists of 156 FCores, containing approximately 40,000 neurons and 10 million synapses. Fabricated using 28-nm processing technology, this chip occupies a die area of 3.8 × 3.8 mm2. The chip area occupied by each individual block, includes axon, dendrite, soma, router, controller and other chip overheads. With its distributed on-chip memory and decentralised many-core architecture, this chip provides an internal memory bandwidth of more than 610 gigabytes (GB) per second, and yields an effective peak performance of 1.28 tera operations per second (TOPS) per watt in ANN mode when running at 300 MHz. Advantages Of The Tianjic Chip The Tianjic chip can provide improved throughput (1.6 to 102 times) and power efficiency (12 to 104 times) over the GPU by just forming a parallel on-chip memory hierarchy and organising the dataflow in a streaming fashion. It enables the concurrent deployment of multiple expert networks within one chip, including most types of SNNs and ANNs. It supports heterogeneous neural networks with a deep fusion of the two paradigms. The use of this chip enables the exploration of more biologically plausible cognitive models. Outlook The Tianjic chip is able to support diverse neural network models, including neuroscience-inspired networks such as SNNs and rate-based biologically inspired neural networks as well as computer-science-oriented networks for such as MLP, CNNs, and RNNs. In one of our articles, we already talked about the powerful RISC-V processor chip named Xuantie 910 by Chinese tech giant Alibaba.","excerpt":"Researchers and technologists around the globe are now striving to develop artificial general intelligence (AGI) in order to achieve one common target — build efficient models which are more human-like.  Now, China has been pushing itself to accelerate AI quite seriously and has been doing massive investments in this sector. According to reports, China’s spending […]","categories":["AI Features"],"tags":["AI and machine learning","Artificial General Intelligence","china ai investments"],"author_name":"Ambika Choudhury","publish_date":"2019-08-06T12:00:44","publication_year":"2019","word_count":631,"keywords":["Go","programming_languages:R","AI","neural network","ML","programming_languages:Go","china ai investments","Artificial General Intelligence","AI and machine learning","RNN","GAN","CNN","R"],"extracted_tech_keywords":["AI","ML","neural network","R","Go","GAN","CNN","RNN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/whats-the-whole-commotion-regarding-chinas-tianjic-chip-about\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10126720,"title":"Researchers Unveil IoA for LLM-based Multi-Agent Collaboration","content":"Even though we have highly capable autonomous agents because of the generative AI wave, current multi-agent frameworks face challenges in integrating diverse third-party agents, simulating distributed environments, and adapting to dynamic task requirements due to hard-coded communication pipelines. As a solution for this, researchers have unveiled the Internet of Agents (IoA) as a framework that addresses these limitations by claiming to provide a flexible, scalable platform for LLM-based multi-agent collaboration. IoA includes an agent integration protocol, an instant-messaging-like architecture, and dynamic mechanisms for agent teaming and conversation flow control. The model aims to overcome three key limitations of existing multi-agent frameworks: ecosystem isolation, single-device simulation, and rigid communication and coordination. Most frameworks only integrate agents within their own ecosystems, limiting diversity and generality. Additionally, existing frameworks often simulate multi-agent systems on a single device, which differs significantly from real-world scenarios where agents are distributed across multiple devices. Furthermore, the communication pipelines in these frameworks are hard-coded, lacking adaptability for dynamic task requirements. You can check out its GitHub repository here. The ‘Agent Integration Protocol’ feature enables integration of third-party agents from different devices, ensuring a diverse and capable agent ecosystem. The ‘Instant-Messaging-Like Framework’ facilitates agent discovery and dynamic teaming, allowing agents to connect and collaborate efficiently. Additionally, the ‘Dynamic Teaming’ and ‘Conversation Flow Control’ mechanisms allow agents to form teams dynamically and autonomously manage conversation flow, thereby improving adaptability and coordination during task execution. IoA’s effectiveness was measured through experiments across various tasks, including general assistant tasks, embodied AI tasks, and retrieval-augmented generation benchmarks. IoA consistently outperformed state-of-the-art baselines, demonstrating superior collaboration among heterogeneous agents. At Google I\/O 2024, the company introduced Gemini-powered Project Astra to create universal AI agents that can perceive, reason, and converse in real-time. Additionally, several autonomous AI agents like BabyAGI, AutoGPT, MetaGPT, AgentGPT, and AutoGen have emerged, each performing various tasks.","excerpt":"IoA includes an agent integration protocol, an instant-messaging-like architecture, and dynamic mechanisms for agent teaming and conversation flow control.","categories":["AI News"],"tags":["AI Agents"],"author_name":"Shritama Saha","publish_date":"2024-07-12T17:59:05","publication_year":"2024","word_count":306,"keywords":["BabyAGI","autonomous agents","AI","MetaGPT","AI Agents","Aim","multi-agent systems","generative AI","AutoGen","R","AutoGPT"],"extracted_tech_keywords":["AI","generative AI","AutoGen","AutoGPT","BabyAGI","MetaGPT","multi-agent systems","Aim","autonomous agents","R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/researchers-unveil-ioa-for-llm-based-multi-agent-collaboration\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10162322,"title":"Infosys to Set Up Lufthansa’s GCC in India","content":"Infosys is establishing a global capability center (GCC) in India for Lufthansa as part of a renewed IT contract valued at nearly $300 million, according to reports. This initiative underscores the growing adoption of build-operate-transfer (BOT) models in major contract renewals. Lufthansa Technik Services India (LTSI), a wholly owned subsidiary of Lufthansa Technik established in 2005, operates out of Bengaluru with over 300 employees. The company provides component services globally for all major commercial aircraft categories, including access to Lufthansa Technik’s spare parts inventory, repair and overhaul facilities, extensive engineering support, and customized logistics arrangements to ensure uninterrupted aviation operations. The establishment of this GCC follows similar moves by Delta Airlines, which launched a tech centre in Bengaluru four years ago, and American Airlines, which is setting up a technology center in Hyderabad. These developments highlight India’s expanding role as a hub for aviation-focused technology centers. Infosys has a history of establishing GCCs for various clients. Two years ago, the company, in collaboration with Rolls-Royce, inaugurated a joint aerospace engineering and digital innovation center in Bengaluru. This centre was designed to offer high-end research and development services integrated with advanced digital capabilities to Rolls-Royce’s engineering and group business services from India. Infosys has had a good year when it comes to expansion and technological prowess. The firm posted strong financial results for Q3 FY25, highlighting growth in revenue, operating margin, and free cash flow. The company reported a revenue of $4.94 billion, reflecting a sequential growth of 6.1% year-on-year (YoY) and 1.7% quarter-on-quarter (QoQ) in constant currency terms.","excerpt":"Lufthansa Technik Services India (LTSI), a wholly owned subsidiary of Lufthansa Technik established in 2005, operates out of Bengaluru with over 300 employees.","categories":["AI News"],"tags":["GCC","Infosys"],"author_name":"Mohit Pandey","publish_date":"2025-01-28T14:51:59","publication_year":"2025","word_count":258,"keywords":["Go","GCC","ELT","Infosys","AI","programming_languages:R","innovation","programming_languages:Go","Git","R"],"extracted_tech_keywords":["AI","R","Go","Git","ELT","innovation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/infosys-to-set-up-lufthansas-gcc-in-india\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10074811,"title":"Machine Learning to Deter Students from Dropping Out of School","content":"September 8 has been celebrated as the ‘International Literacy Day’ across the world since 1967. The significance of this day arises from the fact that despite the steady rise in literacy rates over the past 50 years, there are still 773 million illiterate adults around the world. In India, though the literacy rate has seen phenomenal growth—from 18.3% to 74.4%  between 1951 and 2018—there are  313 million  illiterate people, according to the study, “Literacy in India: The gender and age dimension.” Illiteracy and dropout rates are acutely linked. Dropping out of school is a rampant trend in India. As per the Economic Survey 2021-2022, the secondary level school dropout rate is 16.07% in 2019–20. However, early literacy intervention strategies could help arrest dropouts. AI and ML at work in arresting dropouts With the progress of technologies like AI, ML, IoT and data analytics, many higher education institutions are using these technologies as a part of their processes. The goal is to better identify the pain points in their students’ journey and efficiently allocate resources in the form of programme personalisation and flexibility so as to improve the overall experience of the students. Institutions are even harnessing these technologies to identify students at risk of dropping out and reaching out to them proactively with personalised solutions. For example, by using predictive modelling, Western Governors University was able to increase the graduation rate of students enrolled in its four-year undergraduate programme by five percentage points between 2018 and 2020. While ensuring that students enrolled in higher education do not drop out, it is also important that dropout rates at the school education level are arrested. Else, it can have serious repercussions on the overall literacy rate in the country as well as cognitive skill development. When students drop out of school, both students and communities lose out on these skills, talents and innovations. In India, the Government of Andhra Pradesh, in association with Microsoft, used Azure’s machine learning platform to address the issue of school dropouts. Using the ML platform, an application was developed that enabled the state education department to predict school dropouts. The application processed several data sets related to enrolment, student performance, gender, socio-economic demographics, school infrastructure among others to find predictive patterns for potential school dropouts. Using this model, the Andhra Pradesh state education department was able to identify 19,500 probable dropouts in the Visakhapatnam district for the academic year 2018–19. The ML platform also provided an analysis of the key factors responsible for dropouts. Based on these inputs, the government initiated appropriate drives to increase enrolment at schools. They put in place awareness campaigns to enlighten pupils and parents about the importance of schooling. Over the years, machine learning models have become the focus in addressing the problem of school dropouts. Researchers have made use of several advanced machine learning algorithms like logistic regression, decision trees and K-nearest neighbours, Multi Layer Perceptron and Deep Neural Networks  to predict if a student will drop out or continue her education. In many recent studies, researchers have used deep learning to not only predict dropout rates but also provide personalised intervention to at-risk students. ML algorithms could help identify students who are at risk of dropping out due to several indirect factors that may not be apparently deciphered using a linear rule-based approach. For instance, the case of students who are performing well in academics and have good attendance records but facing difficulty in paying the fees. While a linear rule-based approach may fail to identify such students at risk of dropping out, machine learning can help institutions identify such students beforehand and enable them to make direct appropriate interventions. Despite the potential that machine learning holds in solving the problem of dropouts, not much research has been undertaken in this regard—especially in developing nations. For countries like India, there may be an urgent need to focus on research initiatives to come up with more robust and comprehensive early warning systems and identify students who are at risk of dropping out along with ranking students according to their probability of dropping. Such systems would allow institutions to take appropriate action and make necessary interventions. The success of machine learning in arresting school dropouts would depend on availability of quality data. Timely and accurate data is a must for effective planning and decision making. In recognition of the importance of such data, the Ministry of Education through the Department of School Education and Literacy has come up with the online system of UDISE+, an improved and upgraded version of  Unified District Information System for Education. This system is responsible for collecting real-time data on elementary and secondary education. The collected data can be used as inputs for ML algorithms to identify and mitigate the risks of dropouts.","excerpt":"In India, the Government of Andhra Pradesh, in association with Microsoft, used Azure’s machine learning platform to address the issue of school dropouts.","categories":["AI Features"],"tags":[],"author_name":"Zinnia Banerjee","publish_date":"2022-09-11T13:00:00","publication_year":"2022","word_count":796,"keywords":["Go","machine learning","AI","neural network","R","ML","innovation","deep learning","analytics","Azure"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","neural network","analytics","Azure","R","Go","innovation"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/machine-learning-to-deter-students-from-dropping-out-of-school\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":57752,"title":"Top 8 Funniest And Shocking AI Failures Of All Time","content":"The golden age for artificial intelligence may have just dawned, but the course is not without its challenges. A plethora of technology glitches seems to indicate that it is not quite there yet. Perhaps machines cannot be not perfect either. Although AI is meant to solve problems, as it turns out, it can create new ones as well. These accounts may alarm or amuse consumers but are very embarrassing for the companies involved. Either way, it serves as a reminder of AI vulnerabilities and how the technology has a long way to go before it can be safely adjudged superior to humans. From issues with Amazon’s facial recognition software to Google’s AI Panorama, here are 8 high-profile AI technology fails:- Microsoft’s AI Chatbot Tay With chatbots becoming popular across social networks, Microsoft launched its version for Twitter users in March 2016. Monikered ‘Tay’, it was programmed to have casual conversations in the language of a typical millennial.According to the company, Tay leveraged AI to learn from these interactions to hold better conversations in the future. However, the Twitter chatbot had to be taken down less than 24 hours post its launch. Targeting its vulnerabilities, trolls on the microblogging website manipulated Tay into making deeply sexist and racist statements. Following this debacle, Peter Lee, Microsoft’s corporate VP for AI and research issued a public apology, which stated that the company took “full responsibility for not seeing this possibility ahead of time.” Also read: Top 10 Biggest Failures oF AI In 2019 Amazon’s AI-Powered Recruiting Tool According to a Reuters report, Amazon had been building machine learning (ML) programs since 2014 to review job applicants’ resumes. But it is well known that AI has a big bias problem and the company demonstrated this with example in 2015 when it realised that its new system was not rating candidates in a gender-neutral way. That is, its ML specialists had taught their own AI to prefer male candidates over female ones.This happened because these models were trained to verify applicants by tracking patterns in resumes submitted to the company over a 10-year period. The Seattle company reportedly disbanded the team a few years later after failing to develop or work to resolve that problem. Google Photos’ AI Panorama It is common knowledge now that Google Photos uses AI to throw up enhanced versions of photos taken by users on their smartphones. However, a relatively obscure feature can automatically detect images with the same backgrounds and merge these into a single picture. In January 2018, Reddit user Alex Harker posted three photos taken at a ski resort, which Google welded into one panoramic image. Just that the resulting image has one giant error. Missing basic compositional basics, Google’s AI Panorama magnified Harker’s friend’s torso to throw up this: Also read: Top 8 Biggest Failures oF AI In 2018 Facial Recognition Failure In China Back in November 2018, Chinese police admitted to wrongly shaming a billionaire businesswoman after a facial recognition system designed to catch jaywalkers ‘caught’ her on an advert on a passing bus. Traffic police in major Chinese cities deploy smart cameras that use facial recognition techniques to detect jaywalkers, whose names and faces then show up on a public display screen. After this went viral on Chinese social media, a CloudWalk researcher stated that the algorithm’s lack of live detection could have been the problem. Boston Dynamics’s Robot Blooper SoftBank-owned Boston Dynamics debuted its humanoid robot Atlas at Congress of Future Science and Technology Leaders in 2017. While it displayed impressive dexterity on the stage, it tripped over the curtain and tumbled off the stage just as it was wrapping up.As funny as it may seem now, the company was somehow spared immediate online ridicule and became viral only after Reddit users caught on with it.https:\/\/www.youtube.com\/watch?v=TxobtWAFh8o Amazon’s Rekognition In 2018, Members of US Congress rained down on Amazon after its facial recognition software falsely matched 28 congresspeople with mugshots of criminals. In fact, according to the American Civil Liberties Union (ACLU), nearly 40% of the matches were of people of colour, indicating that the technology is racially biased. Unfortunately, despite these demonstrated failures, law enforcement agencies are already trying to use such tools to identify subjects. Also read: Top 5 Biggest Failures oF AI In 2017 LG’s IoT AI Assistant Cloi At CES 2018, an LG robot created to help users control home appliances repeatedly failed to respond to commands from LG’s US marketing chief David VanderWaal. The  IoT AI assistant Cloi simply blinked. Anchored around ThinQ, LG’s in-house AI software, Cloi’s “disastrous” debut was mercilessly mocked on social media. AI World Cup 2018 Predictions The 2018 FIFA World Cup that took place in Russia kept the whole world engrossed – especially AI enthusiasts. This is because, before the start of the World Cup, several researchers had tried to predict its outcome using technology. To do this, the researchers simulated the event 100,000 times and used three different data modelling techniques. The team used data taken from previous World Cups and analysed them on various parameters. Unfortunately, they failed to predict the winner. Although one of the teams – Brazil – reached the quarter-finals, two of the best-predicted teams, Spain and Germany, could not even reach the quarter-finals.","excerpt":"The golden age for artificial intelligence may have just dawned, but the course is not without its challenges. A plethora of technology glitches seems to indicate that it is not quite there yet. Perhaps machines cannot be not perfect either. Although AI is meant to solve problems, as it turns out, it can create new […]","categories":["AI Trends"],"tags":["AI Fails","real time face recognition software"],"author_name":"Anu Thomas","publish_date":"2020-03-02T11:00:00","publication_year":"2020","word_count":875,"keywords":["Go","machine learning","artificial intelligence","AI","chatbots","ML","RAG","BERT","llm_models:BERT","AI Fails","R","real time face recognition software"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","RAG","chatbots","R","Go","BERT","llm_models:BERT"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-8-funniest-and-shocking-ai-failures-of-all-time\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10019734,"title":"Guide to Panoptic Segmentation &#8211; A Semantic + Instance Segmentation Approach","content":"Panoptic segmentation is an image segmentation method used for Computer Vision tasks. It unifies two distinct concepts used to segment images namely, semantic segmentation and instance segmentation. Panoptic segmentation technique was introduced by Kaiming He, Ross Girshick and Piotr Dollar of Facebook AI Research (FAIR),  Carsten Rother of HCI\/IWR, Heidelberg University (Germany) as well as Alexander Kirillov, a member of both the above mentioned organizations in April 2019 (version v3). Read the research paper here. Let us first understand semantic segmentation and instance segmentation approaches in order to have clarity about panoptic segmentation. A Computer Vision project aims at developing a deep learning model which can accurately and precisely detect real-world objects comprising the input data in the form of images or videos. Such models typically rely on the technique of image segmentation which delineates pixel-level boundaries for object detection.There are two effective yet fundamentally different approaches used for image segmentation which are as follows: Semantic segmentation – It refers to the task of identifying different classes of objects in an image. It broadly classifies objects into semantic categories such as person, book, flower, car and so on. Instance segmentation – It segments different instances of each semantic category and thus appears as an extension of semantic segmentation. For instance, if semantic segmentation method identifies a flock of birds in an image, then instance segmentation further eases the task of object detection by identifying individual birds in the flock (i.e. instances of the ‘bird’ semantic category). This article talks about an advanced image segmentation approach called panoptic segmentation which is nothing but the combination of the above mentioned segmentation techniques. In other words, it semantically distinguishes different objects as well as identifies separate instances of each kind of object in the input image. It enables having a global view of image segmentation (category-wise as well as instance-wise), hence the name ‘PANOPTIC’ (means showing or seeing everything at once). Before going into the details of panoptic segmentation, let us understand two important terminologies germane to image segmentation. Things – Any countable object is referred to as a thing in Computer Vision projects. To exemplify – person, cat, car, key, ball are called things. Stuff – Uncountable amorphous region of identical texture is known as stuff. For instance, road, water, sky etc. Study of things falls under the category of instance segmentation task while study of stuff is a semantic segmentation task. Panoptic segmentation assigns two labels to each of the pixels of an image – (i)semantic label (ii) instance id. The pixels having the same label are considered belonging to the same semantic class and instance id’s differentiate its instances. Unlike instance segmentation, each pixel in panoptic segmentation has a unique label corresponding to instance which means there are no overlapping instances. Panoptic segmentation Suppose, some naively encoded pixel values are as follows: 39001, 39002, 4, 5. Here, (pixel_value\/\/100) gives the semantic label while (pixel_value%100) gives the instance id. So for both 39001 and 39002, semantic label will be 39 i.e. both these pixels are assigned the same semantic class; but instance id for both are different (1 and 2 respectively) which shows that they are different instances of the class labelled 39. While pixels with values 4 and 5 belong to stuff classes. Practical implementation of panoptic segmentation The following code illustrates panoptic segmentation performed on MS-COCO dataset using PyTorch Python library and Detectron2 (a PyTorch-based modular library by Facebook AI Research (FAIR) for implementing object detection algorithms and also a rewrite of Detectron library). We have also used the DETR (DEtection TRansformer) framework introduced by FAIR which views object detection as a direct set prediction problem. Not aware of Detectron, Detectron2 and DETR ? Refer to the following links before proceeding! Detectron (GitHub) Detectron2 (GitHub) DETR Research Paper Import the required libraries from PIL import Image import requests import io import math import matplotlib.pyplot as plt %config InlineBackend.figure_format = 'retina' import torch from torch import nn from torchvision.models import resnet50 import torchvision.transforms as T import numpy torch.set_grad_enabled(False); import itertools import seaborn as sns Install the Panoptic API from GitHub for panoptic inference ! pip install git+https:\/\/github.com\/cocodataset\/panopticapi.git Import the installed API import panopticapi from panopticapi.utils import id2rgb, rgb2id List of COCO semantic classes: CLASSES = [ 'N\/A', 'person', 'bicycle', 'car', 'motorcycle', 'airplane', 'bus', 'train', 'truck', 'boat', 'traffic light', 'fire hydrant', 'N\/A', 'stop sign', 'parking meter', 'bench', 'bird', 'cat', 'dog', 'horse', 'sheep', 'cow', 'elephant', 'bear', 'zebra', 'giraffe', 'N\/A', 'Backpack', 'umbrella', 'N\/A', 'N\/A', 'handbag', 'tie', 'suitcase', 'frisbee', 'skis', ‘snowboard', 'sports ball', 'kite', 'baseball bat', 'baseball glove', 'skateboard', 'surfboard', 'tennis racket', 'bottle', 'N\/A', 'wine glass', 'cup', 'fork', 'knife', 'spoon', 'bowl', 'banana', 'apple', 'sandwich', 'orange', 'broccoli', 'carrot', 'hot dog', 'pizza', 'donut', 'cake', ‘chair', 'couch', 'potted plant', 'bed', 'N\/A', 'dining table', 'N\/A', 'N\/A', 'toilet', 'N\/A', 'tv', 'laptop', 'mouse', 'remote', 'keyboard', 'cell phone', 'microwave', 'oven', 'toaster', 'sink', 'refrigerator', 'N\/A', 'book', 'clock', 'vase', 'scissors', 'teddy bear', 'hair drier', ‘toothbrush'] Enumerate the above classes (Detectron2 model uses different numbering convention so we need to change it) coco2d2 = {} count = 0 for i, c in enumerate(CLASSES): if c != \"N\/A\": coco2d2[i] = count count+=1 Perform standard PyTorch mean-std input image normalization transform = T.Compose([ T.Resize(800), T.ToTensor(), T.Normalize([0.485, 0.456, 0.406], [0.229, 0.224, 0.225]) ]) Load a pre-trained model from torch hub and request the post-processor model, postprocessor = torch.hub.load('facebookresearch\/detr', 'detr_resnet101_panoptic', pretrained=True, return_postprocessor=True, num_classes=250) model.eval(); Retrieve an image from the validation set of COCO dataset for testing purpose url = \"http:\/\/images.cocodataset.org\/val2017\/000000281759.jpg\" im = Image.open(requests.get(url, stream=True).raw) Mean-std normalize the input testing image (batch-size: 1) img = transform(im).unsqueeze(0) out = model(img) Compute the probability score for each possible class, excluding the “no-object” class (the last one) scores = out[\"pred_logits\"].softmax(-1)[..., :-1].max(-1)[0] Threshold the confidence to only masks with high confidence >0.85 keep = scores > 0.85 Plot the masks satisfying the confidence level condition ncols = 5 fig, axs = plt.subplots(ncols=ncols, nrows=math.ceil(keep.sum().item() \/ ncols), figsize=(18, 10)) for line in axs: for a in line: a.axis('off') for i, mask in enumerate(out[\"pred_masks\"][keep]): ax = axs[i \/\/ ncols, i % ncols] ax.imshow(mask, cmap=\"cividis\") ax.axis('off') fig.tight_layout() Merge the individual predictions obtained by running the above lines of code into a unified panoptic segmentation. For that, we use DETR’s postprocessor. The post-processor requires as input the target size of predictions (image size here) result = postprocessor(out, torch.as_tensor(img.shape[-2:]).unsqueeze(0))[0] Visualize the panoptic segmentation’s results The segmentation is stored in a special-format png panoptic_seg = Image.open(io.BytesIO(result['png_string'])) panoptic_seg = numpy.array(panoptic_seg, dtype=numpy.uint8).copy() Retrieve the instance id corresponding to each mask panoptic_seg_id = rgb2id(panoptic_seg) Colour each mask individually and plot the visualization panoptic_seg[:, :, :] = 0 for id in range(panoptic_seg_id.max() + 1): panoptic_seg[panoptic_seg_id == id] = numpy.asarray(next(palette)) * 255 plt.figure(figsize=(15,15)) plt.imshow(panoptic_seg) plt.axis('off') plt.show() Output visualization will be as follows: Use Detectron2’s plotting utilities to better visualize the above panoptic segmentation results. Import the utilities !pip install detectron2==0.1.3 -f https:\/\/dl.fbaipublicfiles.com\/detectron2\/wheels\/cu101\/torch1.5\/index.html from detectron2.config import get_cfg from detectron2.utils.visualizer import Visualizer from detectron2.data import MetadataCatalog from google.colab.patches import cv2_imshow Extract the segments information and the panoptic result from DETR’s prediction from copy import deepcopy segments_info = deepcopy(result[\"segments_info\"]) Store the panoptic predictions in a special format png panoptic_seg = Image.open(io.BytesIO(result['png_string'])) final_w, final_h = panoptic_seg.size Convert the png into segment id map panoptic_seg = numpy.array(panoptic_seg, dtype=numpy.uint8) panoptic_seg = torch.from_numpy(rgb2id(panoptic_seg)) Change Detectron2’s numbering to appropriate class id’s meta = MetadataCatalog.get(\"coco_2017_val_panoptic_separated\") for i in range(len(segments_info)): c = segments_info[i][\"category_id\"] segments_info[i][\"category_id\"] = meta.thing_dataset_id_to_contiguous_id[c] if segments_info[i][\"isthing\"] else meta.stuff_dataset_id_to_contiguous_id[c] Visualize the improved prediction results v = Visualizer(numpy.array(im.copy().resize((final_w, final_h)))[:, :, ::-1], meta, scale=1.0) v._default_font_size = 20 v = v.draw_panoptic_seg_predictions(panoptic_seg, segments_info, area_threshold=0) cv2_imshow(v.get_image()) Output visualization will be as follows: Google Colab notebook of the above implementation code can be found here. To get in-depth understanding of panoptic segmentation technique, read its research paper, web link of which is given below: Panoptic segmentation research paper","excerpt":"Panoptic segmentation is an image segmentation method used for Computer Vision tasks. It unifies two distinct concepts used to segment images namely, semantic segmentation and instance segmentation. Panoptic segmentation technique was introduced by Kaiming He, Ross Girshick and Piotr Dollar of Facebook AI Research (FAIR),  Carsten Rother of HCI\/IWR, Heidelberg University (Germany) as well as […]","categories":["Deep Tech"],"tags":["Computer Vision"],"author_name":"Nikita Shiledarbaxi","publish_date":"2021-02-05T12:00:00","publication_year":"2021","word_count":1278,"keywords":["NumPy","AI","PyTorch","ML","computer vision","Colab","Ray","Aim","deep learning","Computer Vision","Matplotlib"],"extracted_tech_keywords":["AI","ML","deep learning","computer vision","Aim","Ray","PyTorch","Colab","NumPy","Matplotlib"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/guide-to-panoptic-segmentation-a-semantic-instance-segmentation-approach\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10040402,"title":"Football’s Moneyball: How AI Can Be Used to Enhance Strategies","content":"“A penalty kick in football could be modelled as a two-player zero-sum game, a game where one player wins as much as the other loses.” There are many factors that separate the elite from the rest. The difference between winners and losers sometimes boils down to how well one has exploited vast amounts of data. The advantages are more statistical in nature than we would want to believe. Football is one such sport, where analytics has played a significant role. From signing million dollar contracts to positioning a player, the coaches have a lot of data to devour. But, we are in the 21st century and nothing loves more data than machine learning algorithms. Now DeepMind researchers want to bring these advantages to football. Scoring With High With AI Image credits: DeepMind Data Analytics has made its way into most fields—and the realm of sports is no exception. The availability of sports data has been rising in recent years. A commonly employed example of this in football is Expected Goals (xG), a performance metric used to evaluate football team and player performance. This metric measures the quality of a shot based on criteria like angle, distance from goal, the probability of scoring a goal—to name a few. In recent years, the amount of data available in football has increased with sensors, GPS trackers and computer vision algorithms to track how players move and how the ball moves. Despite this, machine learning and AI have only come into use to derive insights and make decisions in sports. Keeping this in mind, researchers working for DeepMind and Liverpool Football Club (LFC) published a paper exploring the use of AI in football in the Journal of Artificial Intelligence Research. (Source: DeepMind) DeepMind’s research aims at using AI to combine computer vision, statistical learning and game theory with helping teams spot patterns in data. This is because of the importance of interactive and sequential decision making in the sport (cooperative and competitive), where game theory becomes important; solutions to particular in-game situations, where statistical analysis and machine learning predictions can help; and image and video inputs that can help track and recognise game scenarios with ease. To help with their analyses, DeepMind augmented a game-theoretical framework using the platform Player Vectors, which uses human expertise and machine learning systems to characterise the playing styles of individual football players. Using this, they were able to model a situation in a game-theoretical framework and derive insights on shooting strategies. Game theory is considered to be an essential element in the research and the general study of sports. For example, something like the penalty kick in football could be modelled as a two-player zero-sum game, a game where one player wins as much as the other loses. By conducting game-theory simulations such as this, they were able to identify shooting strategies of groups where each group had players that play with similar styles. Hence, by noticing who tends to shoot to the left and who prefers to shoot rightward or to the centre—teams can optimise their playing strategies. According to the researchers, interactions between game theory and statistical learning would further enhance sports analytics. Researchers suggest that one can do this by using Player Vectors to provide deeper statistical analysis on how decisions are made. An example of this is ‘ghosting’, a type of data-driven analysis that showcases how players should have acted (against how they actually played). Hence, this method projects multiple trajectories of what could have happened. This would allow team players to analyse critical situations and see how alternative strategies could have played out. If a player figures out a better method via the ‘ghosting’ method, they will want to factor it into their actual game the next time they play. Finally, computer vision would allow the detection of events purely from video. Here, team managers could use AI algorithms to recognise certain strategies and derive insights directly from the video. For example, DeepMind used a stylised visualisation of a deep learning model with methods to identify target events, such as kicks, straight from the game’s video. DeepMind’s research revolved around creating an Automated Video Assistant Coach (AVAC) to employ these three fields to make it an effective platform. It will use interactive decision-making, from game theory and statistical learning; effective video predictive models, resulting from statistical learning and computer vision technologies; and strategic video generation models, from game theory and computer vision. These advances would be pivotal to decision making in football. The researchers claim that his new system is capable of reducing the human error that comes with human coaches and allow for the hands-on tactile coaching that would be possible with football coaches and managers.","excerpt":"“A penalty kick in football could be modelled as a two-player zero-sum game, a game where one player wins as much as the other loses.” There are many factors that separate the elite from the rest. The difference between winners and losers sometimes boils down to how well one has exploited vast amounts of data. […]","categories":["AI Features"],"tags":["ai in sports"],"author_name":"Mita Chaturvedi","publish_date":"2021-05-20T12:00:00","publication_year":"2021","word_count":784,"keywords":["Go","machine learning","artificial intelligence","AI","data-driven","computer vision","Aim","deep learning","analytics","R","ai in sports"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","deep learning","computer vision","analytics","Aim","R","Go","data-driven"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/footballs-moneyball-how-ai-can-be-used-to-enhance-strategies\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10101719,"title":"Why are Made-in-India Phones So Expensive?","content":"When Apple had announced that it would be making the iPhone 15 in India, the whole country thought that this would bring down the cost of the flagship phones for everyone in the country. But much to everyone’s surprise, it didn’t really turn out that way. Now, even Google is heading the same way and there seems to be no reason to assume otherwise. At the Google for India 2023 event in New Delhi, Google declared its intent to manufacture the flagship Pixel line-up of phones in India, following the path laid by Apple, which had followed the footsteps of Samsung. However, despite the increasing local production of smartphones in India, iPhone prices in the country remain significantly higher compared to other countries like the US, Dubai, and Singapore. Even made-in-India phones, such as OnePlus, which is ramping up India manufacturing, sells at a higher price here. The latest Google Pixel phones are most expensive in India. pic.twitter.com\/O5Cu2GX8cS— Dr. J (@just1doctorwala) October 10, 2023 This apparent paradox reveals the multifaceted nature of smartphone pricing, with various factors contributing to the disparities in prices. Not entirely “Made in India” Firstly, it’s crucial to understand that iPhones are not entirely “made in India”. Instead, they are assembled in the country, and the supply chain for iPhone production still relies heavily on components imported from other regions. This reliance on imported components makes Apple subject to customs duties, which inevitably have a direct impact on the final prices. Additionally, the Goods and Services Tax (GST) of 18% further contributes to the overall cost, leading to a cumulative price increase of approximately 40% compared to the base price of the imported models. Though the GST was reduced to 12% in July, the pricing did not drop much because of other factors. Apple’s approach to the Indian market differs from its strategies in other countries. Apple has limited collaborations with local banks in India for providing convenient financing options. Furthermore, trading in a year-old iPhone at an Apple store in India typically yields only about one-third of its original value, making the upgrade path less financially attractive for consumers. Despite these challenges, Apple has devised strategies to mitigate the impact of heavy customs duties and taxes on imported models. The company has partnered with local companies to provide discounts and trade-in options. This also includes Tata. These initiatives help make iPhones more affordable for Indian consumers and, to some extent, level the playing field with their international counterparts. Thus, the company is also planning to scale up production in India by five fold. Another notable aspect to consider is that older-generation iPhone models significantly drive Apple’s sales in India. Analysts suggest that the profit margins earned from the sale of premium Pro models help subsidise base models and older-generation iPhones, making them more competitively priced in India compared to countries like Dubai and Singapore. This strategic pricing approach aligns with Apple’s broader business model of offering a diverse range of products to cater to various consumer segments. Furthermore, despite the significant increase in the number of iPhones assembled in India over the years, not all models are produced domestically. Only a few iPhones before 14 were assembled in India. This limitation prevents Apple from implementing dual pricing tiers, one for locally assembled iPhones and another for those imported. As a result, pricing disparities between locally assembled and imported iPhones remain relatively marginal. Not the phone makers’ fault It’s also important to note that the pricing stability of flagship smartphones is not unique to Apple. Other smartphone brands, such as Samsung, Oppo, Xiaomi, and Vivo, have also set up manufacturing facilities in India, producing flagship models locally. However, the prices of flagship devices across these brands have remained relatively unchanged over the years, highlighting the overall pricing dynamics in the Indian smartphone market. Apart from Apple, the fluctuation in international currency exchange rates can impact the pricing of smartphones in India. The value of the Indian rupee, in particular, has been depreciating against the US dollar over the past year. This depreciation has contributed to the higher pricing of the iPhone 15 in India, reflecting the cost challenges associated with importing components and the cumulative effect of taxes. Many factors contribute to these price disparities. A key Apple distributor pointed out, “One of the reasons is the supply chain, as several components are shipped after the payment of import duty. Also, the scale of business in India is much less when compared to strong volumes in the US and Dubai.” This highlights the cost and logistical challenges of operating in a vast and diverse market like India. When it comes to Google, given the market share is less than 1% because of other competitors that offer Android support, Pixel 8 will end up with the same fate as iPhone 15.","excerpt":"“Why do I have to keep asking my American friends to get me an iPhone?”","categories":["AI Features"],"tags":["Google Pixel","iPhone"],"author_name":"Mohit Pandey","publish_date":"2023-10-19T16:00:00","publication_year":"2023","word_count":804,"keywords":["Go","programming_languages:R","AI","RPA","programming_languages:Go","Google Pixel","ViT","iPhone","R"],"extracted_tech_keywords":["AI","R","Go","ViT","RPA","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/why-are-made-in-india-phones-so-expensive\/","complexity_score":3,"technical_depth":7,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10132060,"title":"AI Won&#8217;t Kill Your Computer Science Degree","content":"Recently, Cobalt Robotics founder Erik Schluntz broke his left hand. Despite this, Schluntz was able to continue coding effectively at his job using voice-to-text software and AI assistant Claude. He was able to write over 3,000 lines of code in one week by way of dictation. “Breaking my hand forced me to write all my code with AI for two months. I’m never going back,” said Schluntz, adding that the silver lining was that it forced him to live in the future where humans would write very little of their own code — “Honestly, I loved it.” Schluntz now feels optimistic about where the industry is going, and how to get the most out of these new AI tools. “I’m out of my cast and typing with both hands again, but Claude is still writing most of my code,” he said. The Relevance of a CS Degree? Given that Claude, GitHub Copilot and similar bots can write a lot of code these days, one question arises—is computer science worth studying at all? A few months ago, one argument went viral on X. Entrepreneur and investor Chris Paik called it The End of Software. His argument was that everyone now has the means to freely produce their own periodicals, TV series, and podcasts thanks to the internet. The next big thing is an internet driven by LLMs that will enable anyone to create free software, including non-engineers. “Today’s computer science majors will be similar to late 1990s journalism majors,” he said. During Microsoft Build this year, it was repeatedly stated that “everyone is a developer now.” This suggests that even without prior coding experience, one can learn to code and secure jobs in the AI industry. So, it’s no surprise that IBM’s AI chief, Matthew Candy, stated that people may no longer need a computer science degree to work in the tech field. Candy believes that in the age of AI, soft talents such as creativity, problem-solving, and adaptability are more valuable than technical ones. Billionaire investor Mark Cuban has also said that computer science degrees will lose some of their value as artificial intelligence becomes more advanced. “Twenty years from now, if you are a coder, you might be out of a job because it’s just math and so, whatever we’re defining the AI to do, someone’s got to know the topic,” he said. Given the evolving landscape of the tech industry and the shifting attitudes toward formal educational requirements, it’s evident that your resume needs an overhaul. Sam Altman said, “University degrees are IMO status and not substance at this point,” elaborating that he would rather see someone’s exceptional work or how they perform at a task. The Rise of 10x Developers At the same time, many experts believe AI will actually increase the need for computer science graduates. As Ahmed Awais, founder of Langbase said, “Software development is here to stay. Humans will remain a big, most essential part of any software toolchain.” This is similar to what AIM said earlier about there being jobs for 10x developers. Contrary to popular belief, the tech job market is booming with a high demand for specific skills. All developers have to do is move beyond being “generic software engineers.” Even developers on Reddit agree that AI is simply incapable of replacing a developer. “No way AI can properly learn how to use internal or custom tools at an org – there’s just not enough material for it,” wrote one developer. This is similar to what François Chollet, the creator of Keras, said a few months ago. He predicted that there would be around 10 million more coding jobs in the next five years, but those would be for people who have the expertise in writing code in Python, but most of the coding would be done by AI. I mean, it's true — in 5 years there will be more professional software engineers than today. They will be using AI, but they won't be replaced by AI, at least not within that timeline and definitely not via current technology.It's very strange to me that this is even… https:\/\/t.co\/RIGeDLEW4y— François Chollet (@fchollet) August 6, 2024 The Future of Software Engineers Russell Kaplan, who recently took over as president of Cognition Labs, the company behind the world’s first software AI agent, made several predictions about the future of software engineering.One of them stood out when he said “Models will be extraordinarily good at coding, very soon.” He believes that research labs are investing more in coding plus reasoning improvements than any other domain for the next model generation. “Their efforts will bear fruit,” he added, saying that the future of software engineers would involve managing AI agents.","excerpt":"However, OpenAI’s chief Sam Altman thinks otherwise, and says “University degrees are IMO status and not substance at this point.”","categories":["AI Trends"],"tags":["AI (Artificial Intelligence)","ai and jobs"],"author_name":"Anshul Vipat","publish_date":"2024-08-11T10:30:00","publication_year":"2024","word_count":785,"keywords":["Go","artificial intelligence","Keras","AI","Git","Python","Aim","ViT","AI (Artificial Intelligence)","GitHub","R","ai and jobs"],"extracted_tech_keywords":["AI","artificial intelligence","Aim","Keras","Python","R","Go","Git","GitHub","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/computer-science-degree\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10137951,"title":"95% Less Energy Consumption in Neural Networks Can be Achieved. Here’s How","content":"AI is booming, and so is energy consumption. According to reports, ChatGPT is probably using more than half a million kilowatt-hours of electricity to respond to some 200 million requests a day. In other words, ChatGPT consumes energy equivalent to powering 17,000 households in the USA daily. A research paper titled, ‘Addition is All You Need: For Energy Efficient Language Models’ mentioned that multiplying floating point numbers consumes significantly more energy than integer operations. The paper states that multiplying two 32-bit floating point numbers (fp32) costs four times more energy than adding two fp32 numbers and 37 times more than adding two 32-bit integers. The researchers have proposed a new technique called linear-complexity multiplication (L-Mul), which solves the problem of energy-intensive floating point multiplications in large neural networks. Before L-Mul, neural networks typically performed computations using standard floating-point multiplication, which is computationally expensive and energy-intensive, especially for LLMs, which typically run over billions of parameters. These operations consumed significant computational resources and energy, particularly in attention mechanisms and matrix multiplications. The best part of this approach is that it is not dependent on any specific architecture. Researchers have tested with real-world models like Llama 3.1 8b, Mistral-7 b, and Gemma2-2b to prove these numbers. After testing these models, researchers have concluded that the proposed method can replace different modules in Transformer layers under fine-tuning or training-free settings. Why is it Bigger Than LLMs? As this approach is not limited to neural networks, the implementation of L-Mul should not be limited to LLMs only but can also be extended to the hardware to achieve energy efficiency over a larger spectrum. L-Mul is a new method that approximates floating-point multiplication using only simple integer additions. This makes it faster because the time it takes grows directly with the size of the numbers (linear complexity), unlike traditional methods that get much slower as numbers get bigger (quadratic complexity). L-Mul uses straightforward bit operations and additions to avoid complicated multiplication of parts of the numbers (mantissa) and tricky rounding steps. This approach not only reduces the computational cost but also potentially decreases energy consumption by up to 95% for element-wise floating-point tensor multiplications and 80% for dot products while maintaining comparable or even superior precision to 8-bit floating-point operations in many cases. This is why Google developed bfloat16, a truncated floating point format for machine learning. Meanwhile, NVIDIA has also created TensorFloat-32 specifically for AI applications on its GPUs. The energy consumption is not only related to LLMs but goes beyond that. A Reddit user mentioned that this research paper would probably lead to all CPU manufacturers deprecating normal 8-bit float multiplication routines to a legacy\/compatibility mode. Instead, any FP8 multiplication could be natively performed using the L-Mat algorithm, potentially implemented in future hardware like the 6090 series GPUs, CPUs beyond the 9000-series, or Apple’s M5 chips. “It might force companies like Intel, AMD, NVIDIA, and Apple to quickly and substantially widen the memory buses across their entire hardware lines. If they don’t adapt, they risk being outpaced by alternatives. For instance, inexpensive RISC-V derivatives with extensive high-bandwidth memory (HBM) or even standard FPGAs with sufficient SRAM could potentially outperform NVIDIA’s top-end GB200 cards. This disruption could occur before these established companies have time to develop and release competitive products, potentially reshaping the market dynamics within just a few months,” he added. Further, he suggested that this research can fundamentally change how hardware is built for neural networks. Too Good to be True? While this approach sounds promising, users have raised some concerns. A Reddit user mentioned that integer addition might require more cycles than a single clock cycle on modern GPUs, especially if implemented through bit-level manipulations and approximations. And converting back and forth between floating-point and integer representations might introduce additional overhead. Also, if we consider speed, the proposed approach might not lead to a speed-up of current GPU architectures because the GPUs are optimised for native floating-point operations. The approximation approach might involve multiple steps or require more complex handling of integer operations, which could offset potential speed gains. The paper hints that specialised hardware designed to implement the L-Mul algorithm could lead to both speed and energy efficiency gains​. However, on current GPU architectures that are designed for traditional floating-point operations, the method is more likely to achieve energy efficiency improvements rather than speed-ups. “If energy efficiency is the primary concern, then this method could be very valuable in reducing costs. For performance improvements (speed), the gains might be minimal without hardware specifically optimised for the new method,” he added, further suggesting that specialised hardware is essential for speed up. L-Mul is performing on-par with the current standards while saving a large amount of power. So, even if we don’t achieve better speeds, L-Mul should still be considered a great technique to reduce energy consumption of neural networks.","excerpt":"Researchers have proposed a new technique called L-Mul, which solves the problem of energy-intensive floating point multiplications in LLMs.","categories":["Deep Tech"],"tags":["Neural Network"],"author_name":"Sagar Sharma","publish_date":"2024-10-09T17:00:00","publication_year":"2024","word_count":808,"keywords":["Neural Network","Go","ChatGPT","API","machine learning","attention mechanism","AI","neural network","GPT","disruption","R"],"extracted_tech_keywords":["AI","machine learning","neural network","ChatGPT","R","Go","API","attention mechanism","GPT","disruption"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/95-less-energy-consumption-in-neural-networks-can-be-achieved-heres-how\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10008728,"title":"What Is The Hiring Process For Data Scientists At Indium Software","content":"With a centralised group of analytics and data science professionals, Indium Software comprises seasoned data scientists who handle everything from R&D to Technical Feasibility Analysis for all projects. The company also has de-centralised teams that are specific to region, domain, and projects. To understand the data science hiring process in the organisation, we caught up with Pradeepa Ravindran, Vice President, Human Resources at Indium Software. Ravindran shares that an ideal data science candidate is like two sides of a coin. “One side is where they should possess functional skills such as business acumen, domain expertise and should be able to come up with innovative business use cases. Whereas, the other side of the coin being technical expertise where the candidate should possess tool & technology expertise, hands-on experience, willingness to be flexible and should possess a high learnability index,” she says. Therefore, the hiring process requires them to have a thorough look at both technical skills and business acumen. Skill Sets For Data Scientist While details on the skill sets that they look for in a data science candidate, Ravindran shared few points as below that they look for: Writing and optimising queriesShould know ML algorithms like Regression, Random Forest, Xgboost, Naïve Bayes and SVM very well – implementation and maths of itShould know Statistics concepts like probability distributions, hypothesis testing, Central Limit Theorem, metrics like p-value, precision, recall etcThey should have experience with tools such as SAS, Matplotlib, Spark, Jupiter, NLTK, scikit-learn, Tensorflow, Tableau, Power BI and more. Ravindran further added that on the educational background front, they mostly prefer to hire candidates from B.Tech, M.Tech (IT OR Computer science), M.Sc Statistics and M.Sc Data Science from renowned institutes like (IIT, BITS, etc.). They also prefer to hire talent with Data science specialisation in PhD, i.e. ML, NLP, Deep learning. However, when it comes to preference between educational background and skills, Ravindran shared that while they prefer both pedigrees, skills are more important for data science roles. Interview Process Data scientist recruitment processes dive into understanding candidates’ technical aspects over the video, Fact to Face, or phone call. The interviewer from the technical team would gauge his\/her potential in Data science, i.e. coding, Math, Statistics, ML & DL, Critical thinking, problem-solving ability. Post clearing the hiring manager interview, candidates are given a technical assessment which is monitored with the help of an ATS tool. “They will then go through a CoderPad test. This assessment is designed to take 3 hours to complete the task given in Data Analytics or Machine learning models,” shared Ravindran. This assessment is to understand the candidate’s presentation and the reproducibility of their work. As a next stage, the company invites the candidates to visit their office premises for the final stage of the interview process. “We also do another evaluation process by our technical panel on problem-solving\/whiteboarding. Finally, our HR discussion happens to check the culture and value fitment,” she added. Some of the questions that assess a candidate’s algorithms and technology expertise in data science are: Which algorithm (Decision tree & Random Forest) has high bias and low variance?Moment generating function and probability theory?What is a sigmoid and binomial function?Write a command for creating an empty data frame with 3 columns? Besides, the candidates are also assessed for their quantitative aptitude, logical thinking and analytical reasoning skills. Some of the conventional ways of sourcing data science talent at Indium is through LinkedIn, job portals such as Naukri, Monster, Employee referral program and Walk-in interviews. On the other hand, some of the non-traditional ways are Head Hunting through social media, participating in meetups, GitHub sourcing and connecting with learning institutes to hire potential talents. Some of the current positions that they are hiring are Lead Data scientist, Junior Data Analyst, Data Analyst and Analytics Data Engineer. Interested candidates can apply on LinkedIn or the company’s career’s page. Picking The Right Candidate Ravindran shares that they face several challenges while recruiting for data science positions. “More often, the data science candidate’s resume looks vibrant with a gamut of technical skill sets mentioned in the resume, yet, they flounder on basic concepts of Statistics and Math which is essential.  This makes it very difficult to nurture and groom the candidate to take on additional responsibilities when they are conceptually weak,” she added. Suggesting some of the ways that companies can avoid hiring mistakes are to make it a point to assess candidates on their statistics and math concepts. “This assessment ensures that they can contribute and scale up within the organisation quickly. We also assess them critically on their business acumen and even assess their interest in cross-domain expertise,” she said. Growth Opportunities For Data Scientists At Indium Software A data scientist is expected to perform various tasks from designing data modelling processes to creating algorithms and predictive models. They are also expected to create custom data models and algorithms while solving complex problems in statistical analysis, machine learning, deep learning, NLP, CNN and more. Apart from exposure to a wide range of challenging tasks, Indium provides opportunities for data scientists to grow internally. As Ravindran shares, there are no boundaries to grow at Indium. “With our diverse range of clients, career development opportunities are spread very wide. Even by starting as a Junior Data Scientist, he\/she is trained to build strong technical knowledge along with a wholesome understanding of the business acumen,” she said. “As our Data Scientists grow, they are equipped to take more responsibilities and be accountable for their assignments. We give more importance on the skill and competency a person possesses, and so we have people reach the role of a Principal Data Architect within a period of 10 years,” she further added. In a concluding note for an analytics professional who wishes to carve out a career in the analytics industry, Ravindar advises, “In order to have a fruitful career in data science, my advice to budding data scientists would be to always have a holistic understanding of your trade – be strong technically and functionally.” “A one-dimensional way of thinking and working is not enough anymore, be multi-dimensional. The ability to support different functions is an absolute must – it can be from pre-sales to customer\/account management. Apart from this, push yourself to be a continuous learner, be articulate and always be a team player. With these qualities, the journey from being a Junior Data Scientist to becoming a Principal Data Scientist can be achieved in a short span of time,” she said.","excerpt":"With a centralised group of analytics and data science professionals, Indium Software comprises seasoned data scientists who handle everything from R&D to Technical Feasibility Analysis for all projects. The company also has de-centralised teams that are specific to region, domain, and projects. To understand the data science hiring process in the organisation, we caught up […]","categories":["AI Hirings"],"tags":["back office data","Data Science Hiring","Data Scientist","machine learning software","principal data scientist"],"author_name":"Srishti Deoras","publish_date":"2020-10-02T12:00:36","publication_year":"2020","word_count":1083,"keywords":["data science","scikit-learn","machine learning","principal data scientist","AI","TensorFlow","ML","NLP","Data Science Hiring","deep learning","machine learning software","analytics","XGBoost","back office data","Data Scientist"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","NLP","data science","analytics","TensorFlow","scikit-learn","XGBoost"],"url":"https:\/\/analyticsindiamag.com\/ai-hiring\/what-is-the-hiring-process-for-data-scientists-at-indium-software\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10110005,"title":"Qualcomm Unveils Mixed Reality Chip for Google, Samsung","content":"To challenge Apple’s upcoming Vision Pro, Qualcomm, the leading manufacturer of mobile phone processors, has revealed the Snapdragon XR2+ chip, designed specifically for virtual and mixed reality headsets. The Snapdragon XR2+ chip boasts the capability to power 12 or more high-definition cameras, making it an attractive choice for industry giants Samsung and Google. Both companies have committed to utilising this cutting-edge component in their upcoming product developments. The XR2+ chip enhances user experience with a 20% improvement in processors and graphics components, projecting 4K-resolution images onto each lens. Said Bakadir, Qualcomm’s Senior Director of Product Management, highlights that this advancement aims to mitigate eye fatigue, reduce the risk of motion sickness, and enable device makers to implement new features. Bakadir emphasised, “A lot of people want more juice to do more things. People want to push the platform even higher.” Qualcomm’s strategic collaborations extend beyond Google and Samsung, with additional partners expected to make announcements during the upcoming CES show. Products incorporating the XR2+ chip could be available as early as this year, according to Bakadir. Qualcomm’s chips have been integral to various companies’ attempts to captivate consumers in the mixed-reality category, which, despite numerous efforts, has yet to experience widespread adoption. Distinguishing between AR and VR, Qualcomm’s new chip aims to bridge the gap. Apple’s Vision Pro and Meta’s Quest 3 headset are notable examples that seamlessly integrate both technologies. Furthermore, the increased capacity of the chip to handle more cameras is anticipated to enhance depth perception and facilitate eye tracking, crucial for orienting users in the real world and recognising objects.","excerpt":"Products incorporating the XR2+ chip could be available as early as this year.","categories":["AI News"],"tags":["Qualcomm"],"author_name":"Mohit Pandey","publish_date":"2024-01-05T09:09:06","publication_year":"2024","word_count":263,"keywords":["Go","programming_languages:R","AI","ML","programming_languages:Go","RAG","Aim","Qualcomm","R"],"extracted_tech_keywords":["AI","ML","Aim","RAG","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/qualcomm-unveils-mixed-reality-chip-for-google-samsung\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10062104,"title":"All major tech firms that suspended operations in Russia","content":"As the Russian aggression of Ukraine continues unabated, tech giants like Apple, Google and Meta have announced suspension or restriction of services in Russia. Ukraine’s Vice Prime Minister and Minister of Digital Transformation Mykhailo Fedorov had reached out to the tech CEOs to cancel their services in Russia to mount pressure on Putin’s government. Here are major tech companies temporarily withdrawing services in Russia. Apple We are deeply concerned about the Russian invasion of Ukraine and stand with all of the people who are suffering as a result of the violence. We are supporting humanitarian efforts, providing aid for the unfolding refugee crisis, and doing all we can to support our teams in the region, said Apple’s official statement. The tech giant is in communication with the relevant governments on the actions they are taking and has now stopped all its product sales in Russia. Earlier, Apple had restricted Apple Pay services in Russia and removed state media like RT News and Sputnik News from the App Store outside Russia. As a safety measure, the tech giant has turned off traffic and live incidents features in Apple Maps in Ukraine. Earlier, Mikhail Fedorov had written to Apple chief Tim Cook: “I appeal to you and I am sure you will not only hear but also do everything possible to protect Ukraine,- to stop supplying Apple services and products to the Russian Federation, including blocking access to App Store!” Google Google has pulled RT News and Sputnik from the Play Store. In an official statement, Google said: “In response to the war in Ukraine, we are pausing Google monetisation of Russian state-funded media across our platforms. We’re actively monitoring new developments and will take further steps if necessary.” YouTube is also blocking the Russian state media in Ukraine at the behest of the Ukrainian government. The video streaming platform has already taken down several channels and videos actively involved in spreading propaganda and misinformation. Oracle and SAP Oracle has suspended all operations in the Russian Federation.”On behalf of Oracle’s 150,000 employees around the world and in support of both the elected government of Ukraine and for the people of Ukraine, Oracle Corporation has already suspended all operations in the Russian Federation.” On behalf of Oracle’s 150,000 employees around the world and in support of both the elected government of Ukraine and for the people of Ukraine, Oracle Corporation has already suspended all operations in the Russian Federation.— Oracle (@Oracle) March 2, 2022 SAP condemned the invasion of Ukraine ‘in the strongest possible terms. The company announced it would stop sales of all SAP services and products in Russia. Intel, AMD, & TSMC Semiconductor giants Intel and AMD have halted sales to Russia. Additionally, The Taiwan Semiconductor Manufacturing Co has stopped working with its Russian partners. Meta Meta has barred Russian state media from running ads or monetising on its platforms. Recently, the company has removed a disinformation network– Russian news sites posing as independent news entities– from its platforms. The network was pushing toxic narrative across platforms like Facebook, Instagram, YouTube, Telegram, Twitter. The Russian government accused Meta of censoring Russian media and said  it would limit its citizens’ access to the social media platform. Twitter Twitter said it would comply with the European Union’s sanctions on Russian state-affiliated media like RT and Sputnik. The microblogging platform has claimed its taking measures to prevent the spread of misinformation on Russian invasion of Ukraine. To that end, Twitter has started slapping labels to the tweets that contained links from state-sponsored media outlets. https:\/\/twitter.com\/yoyoel\/status\/1498343849273425921?s=20&t=E78c4RPS8RWxYWwpviMQ_A “As people look for credible information on Twitter regarding the Russian invasion of Ukraine, we understand and take our role seriously. Our product should make it easy to understand who’s behind the content you see and what their motivations and intentions are,” said Yoel Roth, Head of Site Integrity at Twitter.","excerpt":"Apple had restricted Apple Pay services in Russia and removed state media like RT News and Sputnik News from the App Store.","categories":["IT Services"],"tags":["Russia"],"author_name":"Shraddha Goled","publish_date":"2022-03-04T14:00:00","publication_year":"2022","word_count":642,"keywords":["Go","programming_languages:R","AI","Russia","digital transformation","programming_languages:Go","Git","Aim","GAN","R"],"extracted_tech_keywords":["AI","Aim","R","Go","Git","GAN","digital transformation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/all-major-tech-firms-that-suspended-operations-in-russia\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10077231,"title":"Schmidhuber – A Name Lost in Alleys of AI Research","content":"Prof. Jürgen Schmidhuber was always early. Schmidhuber’s contributions to research around neural networks extend far beyond his most notable, like the Long Short Term Memory networks (LSTMs). These eventually went on to form the groundwork for progress in vision and speech models – around 2015 Google discovered that LSTM could reduce transcription errors in their speech recognition system by around 49%, making a huge leap after years of slow movement. With names like Yann LeCun assuming stardom due to the predominance of deep learning, Schmidhuber’s pioneering work has often been overlooked. And he has made his discontent known. In 1990, Schmidhuber’s research around gradient-based artificial neural networks for long-term planning and reinforcement learning through artificial curiosity introduced a bunch of new concepts like two of the most powerful Recurrent Neural Networks or RNNs called the controller and the world model. Lecun (@ylecun)’s 2022 paper on Autonomous Machine Intelligence rehashes but doesn’t cite essential work of 1990-2015. We’ve already published his “main original contributions:” learning subgoals, predictable abstract representations, multiple time scales…https:\/\/t.co\/Mm4mtHq5CY— Jürgen Schmidhuber (@SchmidhuberAI) July 7, 2022 Schmidhuber’s accusations This year after LeCun published ‘A Path Towards Autonomous Machine Intelligence’, Schmidhuber accused him of rehashing older ideas and presenting them without any credit. He introduced artificial curiosity, which Schmidhuber says LeCun focused on in his abstract. He also claims that LeCun’s work on generative and adversarial neural networks in 2014 was a simplified version of his own work way back in 1990. A bunch of other now-prominent ideas that Schmidhuber wrote about in the 90s have found their way back in the current landscape and it is easy to understand why he would want recognition. Why Linear Transformers are secretly Fast Weight Programmers? In 1992, Schmidhuber published a paper titled ‘Learning to control Fast-Weight Memories: An Alternative to Dynamic Recurrent Networks’ that discussed an alternative to RNNs. Feedforward neural networks, which were the first and the simplest of all neural networks, slowly learned using gradient descent to program the changes of the fast weights of another neural network. This is how fast weights worked as a concept: Every neuron connection had a weight associated with it. This weight has two components to it – the standard slow weight that represents long-term memory (it learns slowly and perishes slowly) and the fast weight which represents short-term memory (it learns quickly and perishes quickly). Fast weights stored short-term memory with weights matrix, achieving a higher capacity. The short-term memory can also store relevant information from the history of the current sequence, so that the information for the ongoing process is readily available. In 1991, one of these FWPs computed weight changes using additive outer products of self-activation patterns. These self-activation patterns made a turn around and have come to be known as the keys and values for self-attention which are the central idea to Transformers today. The core idea of linearised self-attention Transformers has been borrowed from Schmidhuber’s paper. The impact of his work can be gauged from the impact that Transformer architectures now have across natural language processing computer vision applications. In 1993, Schmidhuber also introduced the term ‘attention’ in his paper, ‘Reducing the Ratio between Learning Complexity and Number of Time Varying Variables in Fully Recurrent Nets’, which has now gained acceptance. 30 years ago: Transformers with linearized self-attention in NECO 1992, equivalent to fast weight programmers (apart from normalization), separating storage and control. Key\/value was called FROM\/TO. The attention terminology was introduced at ICANN 1993 https:\/\/t.co\/m0hw6JJrbS pic.twitter.com\/8LfD98MIF4— Jürgen Schmidhuber (@SchmidhuberAI) October 3, 2022 Fast Weight Programmers slowly creeped back into the realm after big names like Geoffrey Hinton reignited the conversation around them. In 2017, Hinton in his lecture at the University of Toronto spoke about fast weights as an idea and their realisation again. “Fast weights provide a neurally plausible way of implementing the type of attention to the past that has recently proved very helpful in sequence-to-sequence models. By using fast weights, we can avoid the need to store copies of neural activity patterns,” he said. Revival of old ideas Schmidhuber’s paper is now celebrating its 30-year anniversary, forcing us to reckon with the origins of some of the concepts that are behind the massive headway that AI research is currently making. Researchers have reasoned that Schmidhuber’s name was lost in the big gap due to the ebbs and flows in AI research while others say that there is a long list of contributors who have helped research reach the point where it is today. For a profile with The New York Times, Gary Bradski, chief scientific officer at OpenCV.ai spoke about Schmidhuber saying, “He’s done a lot of seminal stuff. But he wasn’t the one who made it popular. It’s kind of like the Vikings discovering America; Columbus made it real.” But as Schmidhuber’s older ideas see a resurgence, there is growing outrage within the community for ignoring his past accomplishments. His fight for recognition isn’t only for him, he insists, Schmidhuber believes researchers going all the way back to the 1960s have been regularly erased by the more contemporary giants.","excerpt":"As Schmidhuber’s older ideas see a resurgence, there is growing outrage within the community for ignoring his past accomplishments","categories":["AI Features"],"tags":["Interviews and Discussions"],"author_name":"Poulomi Chatterjee","publish_date":"2022-10-14T11:00:00","publication_year":"2022","word_count":845,"keywords":["Go","AI","neural network","Transformers","computer vision","OpenCV","RAG","Aim","deep learning","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","deep learning","neural network","computer vision","Aim","Transformers","OpenCV","RAG","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/schmidhuber-a-name-lost-in-alleys-of-ai-research\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10162186,"title":"Why Indian IT is Not Keen on Building AI Foundational Models","content":"While Indian startups and founders are catching up to the AI wave, prompted by the announcement of the $500 billion Stargate Project from the US and China’s open-source “side project” DeepSeek, the Indian IT sector remains unfazed. With unwavering confidence in Nandan Nilekani’s vision, they continue to focus on building India as the AI use case capital of the world. Though tech giants like TCS, Infosys, Wipro, and HCLTech have started developing agentic AI frameworks, small language models, and even drug discovery, their efforts remain focused on clients alone. These initiatives do not prioritise building foundational technologies for the country. For example, Tech Mahindra built its Project Indus, the only foundational model emerging from an IT firm in India. However, the expected impact is largely on its clients instead of the creation of transformative products like ChatGPT, Claude, or Gemini. The model, uploaded on Hugging Face by Tech Mahindra chief innovation officer Nikhil Malhotra, is open-source. However, it has garnered very few downloads and is not available for Serverless Inference with its API. Unfortunately, Indian IT is not very interested in building a foundational model. One can argue that even though the firms have enough funds to make a foundational model, they won’t build one unless their clients ask them to or there is some requirement from their side. What Needs to be Done? When AIM asked industry experts if India should also build its own Stargate Project, the responses were mostly positive. Ajai Chowdhry, HCL co-founder and chairman of the Mission Governing Board of India’s National Quantum Mission, expressed concern over the growing shift towards taking control of AI. “We seem to be getting to the weaponisation of tech. For strategic autonomy, we must create our own AI doctrine and have strong control over our data,” Chowdhry told AIM. Several others expressed similar thoughts as soon as Aravind Srinivas, CEO of Perplexity, said that India should definitely focus on building its own LLM and that Nilekani was “wrong about India not requiring it to build one.” The co-founders of Infosys are still debating whether they need one. One of them, Kris Gopalakrishnan, wrote on X that India needs to build its foundation model for a cultural and strategic economy. Meanwhile, Mohandas Pai, the former CFO of Infosys, earlier told AIM that Indian IT companies are services companies and are not focused on building AI products. According to him, the funding required to build something foundational is much higher than what is available in the country. Pai called for a government-backed innovation fund similar to France’s $36 billion France Innovation Fund, which supports startups like Mistral. Such funding, he argued, could enable India to produce foundational models and compete on a global scale. This is similar to what K Krithivasan, CEO and MD of TCS, highlighted when he said that building LLMs has no huge advantage, as the cost outweighs the benefit. He added that since most organisations in India are system integrators, companies need to use products as software and ensure that clients receive the benefits. At the same time, he also agreed that building it for regional languages makes sense for the democratisation of the technology. CP Gurnani, former CEO of Tech Mahindra, told AIM that building a foundational model is important, and that is what led him to start the Project Indus initiative during his tenure. He said that India should build something like NVIDIA. How is the Work Progressing? While AI startups like Sarvam and Krutrim are working with IT giants to provide language models for their clients, the firms are also interested in helping their clients build the models. However, instead of building one, it seems like the ideal approach for IT firms would be to provide a DeepSeek R1 model for their clients. “The Indian path in AI is different. We are not in the arms race to build the next LLM; let people with capital, let people who want to pedal chips do all that stuff…We are here to make a difference, and our aim is to put this technology in the hands of people,” Nilekani said last year, back when the debate around building AI had started. The upskilling of the Indian IT workforce for generative AI is also a good sign for how the firms are adopting generative AI, but nothing about building a foundational model. It seems like Indian IT has missed the generative AI bus, and no amount of funds can bring it back up. As Narayana Murthy said last year, India is only good at copying ideas from the West and applying them here. India’s tech sector continues to prioritise short-term gains from outsourced IT services rather than investing in creating globally competitive products. For Indian IT, it actually makes sense.","excerpt":"It makes sense for the industry not to spend on foundational models and work with the ones already available.","categories":["IT Services"],"tags":["AI Models","Indian IT industry"],"author_name":"Mohit Pandey","publish_date":"2025-01-26T10:00:00","publication_year":"2025","word_count":793,"keywords":["AI Models","ChatGPT","agentic AI","Hugging Face","AI","serverless","DeepSeek R1","Ray","Aim","generative AI","Indian IT industry","small language models"],"extracted_tech_keywords":["AI","generative AI","agentic AI","ChatGPT","DeepSeek R1","Aim","Ray","Hugging Face","small language models","serverless"],"url":"https:\/\/analyticsindiamag.com\/it-services\/why-indian-it-is-not-keen-on-building-ai-foundational-models\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10162609,"title":"India’s Chief Economic Advisor Calls for More Private R&amp;D Investments as Govt Funds 50%","content":"Ahead of the Union Budget 2025, the chief economic advisor to the Indian government, V Anantha Nageswaran, discussed the impact of artificial intelligence (AI) on employment and labour in the latest economic survey. He highlighted that raising wages and salary growth for workers is not just a moral imperative but also a driver of aggregate demand for businesses in the medium run. “It is enlightened self-interest rather than being seen only from a moral prism,” he stated. He pointed to AI as a clear example of the evolving balance between capital and labour, referring to a special essay on the topic in Chapter 13 of the Economic Survey 2024-25. “Deployment of artificial intelligence presents both opportunities and challenges. Past technology revolutions, in general, sometimes make us feel that technology eventually generates more jobs than it displaces. That is true, but the key word is ‘eventually’,” he emphasised. Nageswaran underscored the importance of addressing the transition period before AI-induced job displacement leads to new opportunities. “What happens between ‘now’ and ‘eventually’ is critical, and that is where we need to create supporting institutions, enabling institutions to train workers and prepare them. Academic curriculums have to change, workplace practices have to evolve,” he said. He acknowledged that technological transitions in history, particularly during the previous three industrial revolutions in the Western world, had not been painless. “We need to learn from them, and that is what this chapter covers.” Nageswaran also stressed that businesses must weigh AI’s benefits against its long-term social costs. “The private sector needs to consider the subterranean social costs, which may surface over time and eventually affect the environment necessary for running businesses smoothly. The chapter argues [that] for cash flow deployment optimised over a long horizon, this will augment labour and deliver broad-based social benefits.” In a lighter vein, he said, “One of the considerations – when we wrote this chapter – was whether it would become outdated by the time it goes to print, given the fast changes happening in the artificial intelligence space.” However, he reassured that the discussion focuses not on technological advancements themselves but on AI’s broader economic and labour implications. He further pointed out the relatively low R&D spending by India’s private sector, noting that compared to other countries, 50% of India’s R&D expenditure comes from the government, 41% from business enterprises, and only 9% from higher education institutions. “This is one area where the private sector must step up. Despite available incentives, private sector R&D in India remains both limited and sectorally concentrated.” ‘From Make in India, We Have Moved to Make for the World’ President Droupadi Murmu, addressing the joint session of Parliament ahead of the Union Budget session, emphasised the government’s commitment to making India a global leader in innovation. “Our aim is to make India a global innovation powerhouse. In the area of artificial intelligence, IndiaAI Mission has been started,” she stated. Highlighting India’s growing influence in digital technology, she said, “Our country is emerging as a major global player in digital technology.” ‘From ‘Make in India’, we have moved to ‘Make for the World’,’ President Murmi added while talking about the defence sector and innovations. The IndiaAI Mission, launched to strengthen the nation’s role in artificial intelligence, was also discussed by IT minister Ashwini Vaishnaw in a press conference. He recently confirmed that India’s LLM would be ready within 10 months. The electronics and information technology ministry, through its IndiaAI Mission, has also called for proposals to create AI models that are tailored to Indian needs while meeting global standards. This proposal comes after Vaishnaw said India is set to develop its own generative AI model, aiming to rival global platforms like ChatGPT and DeepSeek. “Very soon, we will have our own LLMs,” Vaishnaw added. Comparing India’s AI infrastructure to global benchmarks, Vaishnaw noted, “DeepSeek AI was trained on 2,000 GPUs, ChatGPT was trained on 25,000 GPUs, and we now have 15,000 high-end GPUs available. India now has a robust compute facility that will support our AI ambitions.” For the Union Budget 2024-25, the Indian government has underscored its commitment to AI by approving the comprehensive IndiaAI Mission. A substantial financial outlay of over ₹10,000 crore was earmarked for the mission over five years, of which ₹551 crore has already been allotted. Speculations and discussions are rife about who stands to benefit and what allocations will be made. While finance is obviously the big picture, considering how the IndiaAI Mission found its place in the previous Budget, AI is likely to take centre stage in the upcoming session tomorrow.","excerpt":"About 50% of India’s R&D expenditure comes from the government, 41% from business enterprises, and only 9% from higher education institutions.","categories":["AI Features"],"tags":["Union Budget 2025"],"author_name":"Mohit Pandey","publish_date":"2025-01-31T15:10:00","publication_year":"2025","word_count":762,"keywords":["Go","ChatGPT","API","artificial intelligence","Union Budget 2025","AI","Git","GPT","Aim","generative AI","R"],"extracted_tech_keywords":["AI","artificial intelligence","generative AI","ChatGPT","Aim","R","Go","Git","API","GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/indias-chief-economic-advisor-calls-for-more-private-rd-investments-as-govt-funds-50\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10105147,"title":"Indian Startups Love Cockroach Labs","content":"Whether you ask an early-stage startup or a globally growing company, they will all tell you how seamless CockroachDB is, and that it’s better than the existing distributed SQL databases out there, like MySQL Cluster, Google Spanner, MariaDB Xpand and others. Some of the notable startups using CockroachDB include Netflix, Comcast, Fortinet, and Rubrik, alongside startups like Blockdaemon, ConverseNow, LaunchDarkly and more. In India, the love for CockroachDB is only growing. One such exploration led us to Bengaluru-based neobanking startup Fi. The company, in its initial days, was facing the challenge of building a payment system from scratch, and recognised the need for a database that could support long-term growth, scale across regions seamlessly, ensure consistent transactions, comply with data location regulations, and align with their team’s background in Postgres. That’s when it turned to CockroachDB. At Cockroach Labs’ first-ever distributed SQL mixer held in Bengaluru last week, Fi’s founding engineer Prasanna Ranganathan said that with CockroachDB, the team was able to scale up rapidly with queries per second doubling from 1.5K to 3.2K and transaction processing increasing from 200 to 400 transactions per second over the past year. Cockroach Labs knows India better than most other database companies, thanks to its burgeoning developers’ ecosystem. The company recently expanded its footprint in the country with an office in Bengaluru and the introduction of two new senior leaders — engineering site director Babu Srithar Muthukrishnan and director of recruiting Anand Sudhalayam. Cockroach Labs India Team in Bengaluru “The quality and scale of the talent available in India is unparalleled, particularly for engineering, technical support, and the entrepreneurial spirit driving many businesses, making it compelling for us to expand,” Spencer Kimball, cofounder and chief executive of the company told AIM in an exclusive interaction during the mixer. The decision to enter the APAC market, beginning in Australia, stemmed from the chief revenue officer Jason Forget’s successful experience in building a skilled go-to-market team in their previous role. Recent discussions with both new and native Indian companies revealed a striking similarity to the Bay Area, where tech companies utilise global shipping, and there is little discernible difference in the business environment. Due to the fast growth of startups in India and the increasing influence of technical decision-making in the country, “it’s vital for us to establish a go-to-market strategy here, recognising the opportunity to educate developers and chief architects, which also benefits companies globally seeking guidance from Europe and the US,” he shared. Growth Story New York-based Cockroach Labs was founded eight years ago by three former employees of Google: Spencer Kimball, Ben Darnell, and Peter Mattis. Their experience at Google, witnessing the evolution of database usage, particularly the creation of the NoSQL category, inspired them. Kimball and APAC sales director Tejas Baldev at the Mixer “Despite the success of NoSQL, we observed that as containerisation entered production, there was need for a database that mirrored Google’s, one that was open source and accessible to all companies,” Kimball told us. This is how Cockroach Labs was born. The team’s brainchild CockroachDB is a cloud-native database that caters to legacy businesses like SaaS, banks, healthcare, fintechs and retail, offering an operational database capable of always providing the correct answers and storing mission-critical data for various use cases. The key benefits of CockroachDB’s cloud-native design include the ability to use on-demand resources in the cloud, scalability by adding machines to the distributed cluster, and the capability to replicate data across different cloud availability zones. This replication ensures data availability even in the face of outages. Moreover, CockroachDB supports cross-cloud replication, allowing businesses to balance data across multiple cloud providers such as AWS, Azure, and GCP. Kimball chuckled as he revealed the idea behind the unique name ‘CockroachDB’ during the event, and said that it reflects the database’s core philosophy of survivability. The name symbolises the database’s robustness, highlighting its ability to remain operational even in extreme situations, like a hypothetical World War III — much like a cockroach. The unique differentiator for CockroachDB is its global public cloud integration. “With data centres worldwide, the database can seamlessly expand into new markets without setting up separate services,” added Kimball, highlighting how it enables users to maintain a single, distributed database even when operating in geographically distant regions, offering a unified data store experience across the globe. Supporting Generative AI Applications “When it comes to generative AI, the operational database requirements are still evolving,” said Spencer.  While CockroachDB supports vector data types, secondary indexes are currently unavailable. However, the system enables real-time streaming of all database changes through Change Data Capture, allowing the use of this feature to feed search engines like Elasticsearch for finding similar vectors in AI use cases. Cockroach Labs caters to diverse customer needs, understanding that many, including larger enterprises like banks, may not be at the forefront of AI owing to various concerns like regulatory challenges. The primary goal is to build a versatile database, with current support for AI, especially in feature stores, while maintaining the core focus on being a reliable system of record—the source of truth for all business data. But one question keeps looming: Is it possible that we lack sufficient use cases to demonstrate the benefits of Cockroach? This could be a period of waiting for feedback, especially with challenges related to handling inaccurate data. “Building trust in a database is a lengthy process; it takes around 10 years for people to trust transitioning from cloud vendor databases to multi-cloud setups or from traditional databases to mainframes,” said Kimball. However, as generative AI advances, Cockroach Labs aims to fold in additional modalities and features specific to handling unique aspects, ensuring its reliability as a system of record while adapting to new use cases and capabilities.","excerpt":"“The primary goal is to build a versatile database, with current support for generative AI, especially in feature stores, while maintaining a focus on being a reliable system of record—the source of truth for all business data,” Spencer Kimball, cofounder & CEO, told AIM.","categories":["Deep Tech"],"tags":["AI Startups"],"author_name":"Shritama Saha","publish_date":"2023-12-19T13:54:35","publication_year":"2023","word_count":958,"keywords":["Elasticsearch","GCP","AWS","AI","R","ML","Aim","generative AI","SQL","Azure","AI Startups"],"extracted_tech_keywords":["AI","ML","generative AI","Aim","AWS","Azure","GCP","Elasticsearch","R","SQL"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/indian-startups-love-cockroach-labs\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":65090,"title":"IIITH Announces 3 New Fellowships In AI &#038; Blockchain","content":"International Institute of Information Technology Hyderabad (IIITH) announced three new PhD and Post Doctoral fellowships in AI and Blockchain. As a part of Kohli PhD fellowships for the academic year 2020-21, candidates are invited to apply to stand a chance for the five fellowships. Established in 2015 with funding from Tata Consultancy Services (TCS) Foundation, Kohli fellowships facilitate research, teaching and entrepreneurship in the areas of intelligent systems. Over the years, it has become the foundation for aspirants who are willing to blaze the trail in the data science domain. However, the fellowship will be awarded to aspirants with a prior track record in research that showcases their research aptitude, scholarly abilities, capacity to think through unstructured problems, and an interest to work in the areas of AI aligned with IIITH’s Kohli Centre on Intelligent Systems (KCIS). Selected aspirants will get a monthly stipend of ₹50,000 per month for up to 4 years or submission of the thesis, whichever is earlier. Besides, it supports travel to top conferences forums. The institute has also announced the launch of Ripple-IIITH PhD and Post Doctoral fellowships for blockchain. IIITH, in association with Ripple’s University Blockchain Research Initiative (UBRI) established a blockchain Center of Excellence (CoE) at IIITH last year. Since then, it has been actively researching different aspects of blockchain such as security, next-gen blockchains, game-theoretic issues, ML over blockchains, e-governance over blockchain, etc. “KCIS is one of the largest AI research groups in the country. We have also made significant progress with our blockchain CoE since its inception a year ago. Lots of opportunities are available to work in these emerging areas for PhD and Post-Doc aspirants with the top-class faculty in the AI and the Blockchain area. We look forward to attracting them,” said Prof P J Narayanan, Director, IIITH.","excerpt":"International Institute of Information Technology Hyderabad (IIITH) announced three new PhD and Post Doctoral fellowships in AI and Blockchain. As a part of Kohli PhD fellowships for the academic year 2020-21, candidates are invited to apply to stand a chance for the five fellowships. Established in 2015 with funding from Tata Consultancy Services (TCS) Foundation, […]","categories":["AI News"],"tags":["IIIT hyderabad"],"author_name":"Rohit Yadav","publish_date":"2020-05-12T18:56:48","publication_year":"2020","word_count":299,"keywords":["data science","IIIT hyderabad","Go","funding","programming_languages:R","AI","ML","Ray","ViT","AI research","R"],"extracted_tech_keywords":["AI","ML","data science","Ray","R","Go","ViT","funding","AI research","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/iiit-announces-3-new-fellowships-in-ai-and-blockchain\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10166336,"title":"No One is Going to be GPU Poor Anymore","content":"Not long ago, running a large language model (LLM) meant relying on massive graphics processing units (GPUs) and expensive hardware. Now, however, things are starting to change. A new wave of smaller, more efficient LLMs is emerging, which are capable of running on a single GPU without compromising on performance. These models are making high-end AI more accessible, reducing dependency on large-scale infrastructure, and reshaping how AI is deployed. As Bojan Tunguz, former NVIDIA senior software system engineer, had quipped, “Blessed are the GPU poor, for they shall inherit the AGI.” In the past week, a series of announcements in AI has been made. Mistral’s latest model, Small 3.1, Google’s Gemma 3, and Cohere’s Command A all claim to match the performance of proprietary models while requiring fewer compute resources. These models enable developers, small businesses, and even hobbyists with consumer-grade hardware (e.g., a single NVIDIA RTX card) to run advanced AI models locally. Moreover, running LLMs locally on a single GPU reduces reliance on cloud providers like AWS or Google Cloud, giving businesses more control over their data and privacy. This is critical for industries handling sensitive information and regions with limited internet access. What Makes Them Special? Mistral Small 3.1 features improved text performance, multimodal understanding, and an expanded context window of up to 128k tokens. The company said the model outperforms comparable models like Google’s latest release, Gemma 3, and GPT-4o mini while delivering inference speeds of 150 tokens per second. However, one of the most notable features of the model is that it can run on a single RTX 4090 or a Mac with 32 GB RAM, making it a great fit for on-device use cases. The company said that the model can be fine-tuned to specialise in specific domains, creating accurate subject matter experts. This is particularly useful in fields like legal advice, medical diagnostics, and technical support. On the other hand, Google claims that Gemma 3 outperforms Llama 3-405B, DeepSeek-V3, and o3-mini in preliminary human preference evaluations on the LMArena leaderboard. Like Mistral 3.1, it can also be run on a single GPU or a tensor processing unit (TPU). “Compare that to Mistral Large or Llama 3 405B, needing up to 32 GPUs—Gemma 3 slashes costs and opens doors for creators,” said a user on X.  Notably, a single NVIDIA RTX or H100 GPU is far more affordable than multi-GPU clusters, making AI viable for startups and individual developers. Gemma 3 27B achieves its efficiency by running on a single NVIDIA H100 GPU at reduced precision, specifically using 16-bit floating-point (FP16) operations, which are common for optimising performance in modern AI models. LLMs typically use 32-bit floating-point (FP32) representations for weights and activations, requiring huge memory and compute power. Quantisation reduces this precision to 16-bit (FP16), 8-bit (INT8), or even 4-bit (INT4), significantly reducing model size and accelerating inference on GPUs and edge devices. Regarding the architecture, Gemma 3 employs a shared or tied language model (LM) head for its word embeddings, as indicated by its linear layer configuration, where the LM head weights are tied to the input embeddings. Similarly, Cohere recently launched Command A, a model that delivers top performance with lower hardware costs than leading proprietary and open-weight models like GPT-4o and DeepSeek-V3. According to the company, it is well-suited for private deployments, excelling in business-critical agentic and multilingual tasks while running on just two GPUs, whereas other models often require up to 32. “With a serving footprint of just two A100s or H100s, it requires far less compute than other comparable models on the market. This is especially important for private deployments,” the company said in its blog post. It offers a 256k context length—twice that of most leading models—allowing it to process much longer enterprise documents. Other key features include Cohere’s advanced retrieval-augmented generation (RAG) with verifiable citations, agentic tool use, enterprise-grade security, and strong multilingual performance. Microsoft recently launched Phi-4-multimodal and Phi-4-mini, the latest additions to its Phi family of small language models (SLMs). These models are integrated into Microsoft’s ecosystem, including Windows applications and Copilot+ PCs. Earlier this year, NVIDIA launched a compact supercomputer called DIGITS for AI researchers, data scientists, and students worldwide. It can run LLMs with up to 200 billion parameters locally, and with two units linked together, models twice the size can be supported, according to NVIDIA. Moreover, open-source frameworks facilitate running LLMs on a single GPU. Predibase’s open-source project, LoRAX, allows users to serve thousands of fine-tuned models on a single GPU, cutting costs without compromising speed or performance. LoRAX supports a number of LLMs as the base model including Llama (including Code Llama), Mistral (including Zephyr), and Qwen. It features dynamic adapter loading, instantly merging multiple adapters per request to create powerful ensembles without blocking concurrent requests. Heterogeneous continuous batching packs requests using different adapters into the same batch, ensuring low latency and stable throughput. Adapter exchange scheduling optimises memory management by asynchronously preloading and offloading adapters between GPU and CPU memory. High-performance inference optimisations, including tensor parallelism, pre-compiled CUDA kernels, quantisation, and token streaming, further improve speed and efficiency. Running LLMs Without a GPU? A few days ago, AIM spoke to John Leimgruber, a software engineer from the United States with two years of experience in engineering, who managed to run the 671-billion-parameter DeepSeek-R1 model without GPUs. He achieved this by running a quantised version of the model on a fast NVM Express (NVMe) SSD. Leimgruber used a quantised, non-distilled version of the model, developed by Unsloth AI—a 2.51 bits-per-parameter model, which he said retained good quality despite being compressed to just 212 GB. However, the model is natively built on 8 bits, which makes it efficient by default. Leimgruber ran the model after disabling his NVIDIA RTX 3090 Ti GPU on his gaming rig, with 96 GB RAM and 24 GB VRAM. He explained that the “secret trick” is to load only the KV cache into RAM while allowing llama.cpp to handle the model files using its default behaviour—memory-mapping (mmap) them directly from a fast NVMe SSD. “The rest of your system RAM acts as disk cache for the active weights,” he said. With LLMs now running on a single GPU—or even without one—AI is becoming more practical for everyone. As hardware improves and new techniques emerge, AI will become even more accessible, affordable, and powerful in the years ahead.","excerpt":"“Blessed are the GPU poor, for they shall inherit the AGI.”","categories":["Global Tech"],"tags":["Cohere","Google","GPU","mistral"],"author_name":"Siddharth Jindal","publish_date":"2025-03-19T17:21:43","publication_year":"2025","word_count":1061,"keywords":["CUDA","TPU","AWS","AI","GPT-4o","SLM","Cohere","RAG","Aim","mistral","Gemma 3","Google","small language models","GPU"],"extracted_tech_keywords":["AI","GPT-4o","Gemma 3","Aim","small language models","SLM","RAG","AWS","TPU","CUDA"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/no-one-is-going-to-be-gpu-poor-anymore\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10062381,"title":"Apple announces the most powerful iPad Air ever","content":"Apple unveiled the new iPad Air at their Peak Performance Event, 2022, situating it as a laptop replacement. Claimed to be the most powerful and versatile Apple Air ever, the device features the breakthrough M1 chip, ultra-fast 5G and a new front camera with Center Stage. The new iPad Air will have the same M1 chip that has powered the iPad Pro. Apple says this makes it twice as fast as the best selling Windows laptop in its price range. In fact, Angellina Kyazike, Engineering Program Manager for iPad, noted M1 makes iPad Air “Faster than the fastest competitive tablet”. In addition, the device is expected to feature the A15 Bionic chip with 5G connectivity and a 12-megapixel ultra-wide front camera. The previous iPad Air 4 ran on the A14 Bionic chip with a 4G connectivity. On the hardware end, the new iPad will be powered by an 8-core CPU that delivers up to 60% faster performance and an 8-core GPU that delivers up to 2x faster graphics performance compared to the previous iPad Air. Additionally, a 16-core Neural Engine will support advanced ML functions on the device, including 3D modelling, AR and graphics-intensive games. Apple promises an enhanced experience with the iPad OS15, that supports multitasking, collaboration, note-taking, share-play and advanced ML capabilities such as live text and on-device intelligence. The iPad Air is situated to become a chosen tool for content creators, developers, designers and gamers. It also supports the Smart Keyboard Folio and Magic Keyboard, allowing laptop users to turn their iPad Air into a functional laptop. Apple also spoke of their environmental sustainability plans to have a net-zero climate impact across the entire business by 2023 and introduced environmentally safe modifications. For instance, the new iPad features a 100% recycled aluminium enclosure, 100% recycled tin in the solder of the main logic board, and 100% recycled rare earth elements in the enclosure and audio magnets.","excerpt":"The iPad Air will have the same M1 chip that has powered the iPad Pro. A","categories":["AI News"],"tags":["Apple Bionic"],"author_name":"Avi Gopani","publish_date":"2022-03-09T02:55:39","publication_year":"2022","word_count":319,"keywords":["Apple Bionic","API","programming_languages:R","AI","ML","Aim","ViT","R"],"extracted_tech_keywords":["AI","ML","Aim","R","API","ViT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/apple-announces-the-most-powerful-ipad-air-ever\/","complexity_score":4,"technical_depth":7,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":54484,"title":"AI &#038; Machine Learning Learning Path: A Definitive Guide","content":"Artificial intelligence is currently one of the hottest buzzwords in tech — with good reason. In the last few years, we have seen several technologies previously in the realm of science fiction transform into reality. Experts look at artificial intelligence as a factor of production, that has the potential to introduce new sources of growth and change the way work is done across industries. In fact, AI technologies could increase labour productivity by 40% or more by 2035, according to a recent report by Accenture. This could double economic growth in 12 developed nations that continue to draw talented and experienced professionals to work in this field. According to Gartner’s 2019 CIO Agenda survey, the percentage of organizations adopting AI jumped from four to 14% between 2018 and 2019. Given the benefits that AI and machine learning (ML) enable in business analysis, risk assessment, and R&D — and, the resulting cost-savings — AI implementation will continue to rise in 2020. However, many organizations that adopt AI and machine learning don’t fully understand these technologies. In fact, Forbes points out that 40% of the European companies claiming to be ‘AI startups’ don’t use the technology at all. While the benefits of AI and ML are becoming more evident, businesses need to step up and hire people with the right skills to implement these technologies. Some are well on their way. KPMG’s recent survey of Global 500 companies shows that most of those surveyed expect their investment in AI-related talent to increase by 50 to 100% over the next three years. Why Pursue AI and Machine Learning Courses? As data science and AI industries continue to expand, more people are beginning to understand just how valuable it is to have a qualified AI Engineer or data scientist on their team. As a matter of fact, Indeed.com revealed that job postings for data scientists and AI rose over 29% between May 2018 and May 2019. Many people who want to get into this field, typically start with YouTube videos or other free online courses. This approach is definitely good to get your feet wet, but you can’t make a career jump based on these alone. What you need is to get a handle on the fundamentals of data science is experience through hands-on projects, where you get guidance from experts. These opportunities are not generally available in the workplace, especially if your current role does not involve data science. However, there are some excellent, comprehensive courses that you can enrol in, which will provide you with all of the above. Courses like Simplilearn’s Artificial Intelligence Engineer program enable you to learn, practice, and interact with expert instructors and peer, in live, online sessions. You don’t even have to travel. If you’re looking for a course that keeps students up to date on the latest trends in AI and machine learning through practical projects and industry expert-led instruction, Simplilearn’s AI Engineer and Machine Learning Certification courses are excellent options. There is no better time than now to get started, especially if you want to get ahead of your peers. Learning Path: How to Get an AI and Machine Learning Career Started Choosing a learning path for AI and machine learning training can be overwhelming due to all the options out there, but it’s ideal to choose a program that best suits your needs and goals. Successful data scientists usually have a thorough comprehension of various tools and programming languages. They also understand what their roles are in the grand scheme of things. With these skills, you can easily stand out from the competition with potential employers. Some of the programming languages include SAS, R, and Python. What you’ll need to know depends on different variables, such as the specific project you’re working on or the company you’re working for. In order to be a well-rounded candidate that can take on any type of project, it’s critical to know all three of these programming languages. Beyond that, it’s also helpful for data scientists to learn about AI and machine learning. When you enrol in an accredited data science learning program, you’ll get comprehensive training in the field. Let’s dig into some suggested learning paths for AI and machine learning to give you a better idea of what’s available and what to expect. Artificial Intelligence Engineer Master’s Program Simplilearn’s Artificial Intelligence Master’s Program, co-developed with IBM, is a blend of artificial intelligence, data science, machine learning, and deep learning — facilitating the real-world implementation of advanced tools and techniques. The program is designed to give you in-depth knowledge of AI concepts including the essentials of statistics (required for data science), Python programming, and machine learning. Through these courses, you will learn how to use Python libraries like NumPY, SciPy, Scikit; as well as essential machine learning techniques, such as supervised and unsupervised learning, advanced concepts covering artificial neural networks, layers of data abstraction, and the basics of TensorFlow. Next, let’s look at the courses that are included in this program, which can also be taken separately. Data Science with Python The Data Science with Python course provides students with all-around data science instruction that includes data visualization, machine learning, data analysis, and natural language processing using Python. As a data scientist, it’s crucial to add Python to your skillset, as more and more professionals in the industry are mastering this programming language. In fact, it has been reported that with seven million people now using Python, surpassing Java as the top programing language. This course is not only suited for those wishing to pursue a career as a data scientist but can also be beneficial for anyone looking to work in data analytics or software development. Machine Learning As a data scientist, mastering machine learning is often a requirement, and the best way to do so is by enrolling in an accredited learning program and earning a machine learning certification. Although there are free online learning sources and tutorials, such as blogs and YouTube videos, these unstructured learning methods don’t always cover all aspects of ML. Also, self-learners may not be able to stay up-to-date on industry changes or receive certifications. Through Simplilearn’s machine learning course, students are introduced to various techniques and concepts, such as mathematical and heuristic aspects, supervised and unsupervised learning, algorithm development, and hands-on modelling. This course is ideal for those who want to add to their skill set as a data scientist, or for those who wish to pursue a career as a machine learning engineer. Deep Learning with TensorFlow Deep learning is one of the most exciting and promising segments of artificial intelligence and machine learning technologies. Simplilearn’s Deep Learning with TensorFlow and Keras course is designed to help you master key deep learning techniques. You’ll learn how to build deep learning models using TensorFlow, the open-source software library developed by Google to conduct machine learning and deep neural networks research. It is one of the most popular software platforms used for deep learning and contains powerful tools to help you build and implement artificial neural networks. Advancements in deep learning are already showing up in smartphone applications and efficient power grids. The technology is also driving innovations in healthcare, improving agricultural yields, and helping us find solutions to climate change. With this Tensorflow course, you’ll build expertise in deep learning models, learn to operate TensorFlow to manage neural networks, and how to interpret the results. Natural Language Processing (NLP) Simplilearn’s NLP course gives you a detailed look at the science of applying machine learning algorithms to process large amounts of natural language data. You will learn the concepts of statistical machine translation and neural models, deep semantic similarity models (DSSM), neural knowledge base embedding, deep reinforcement learning techniques, neural models applied in image captioning, and visual question answering using Python’s Natural Language Toolkit (NLTK). AI Capstone Project Simplilearn’s Artificial Intelligence (AI) Capstone project gives you the opportunity to implement the skills you learned in the AI Engineer Master’s program. With dedicated mentoring sessions, you’ll learn how to solve a real industry problem. You’ll also learn various AI-based supervised and unsupervised techniques like regression, multinomial Naïve Bayes, SVM, tree-based algorithms, NLP, etc. The project is the final step in the learning path and will help you to showcase your expertise to potential employers. Bottom Line There is no denying that the job market is competitive. In fact, the Bureau of Labor Statistics recently released a report that reveals how the job market is tightening. If you’re looking for a stable industry that isn’t going anywhere anytime soon, AI and machine learning are excellent choices. However, choosing a growing and successful industry is only half the battle when it comes to job security. There is also a competition to consider — oftentimes, many qualified candidates are vying for the same job opening. One of the best ways to ensure you stand out to recruiters and employers is to have the right credentials. Earning your certifications in AI and machine learning, or other relevant fields is a surefire way to get your resume noticed by the right people. Get started today!","excerpt":"Artificial intelligence is currently one of the hottest buzzwords in tech — with good reason. In the last few years, we have seen several technologies previously in the realm of science fiction transform into reality. Experts look at artificial intelligence as a factor of production, that has the potential to introduce new sources of growth […]","categories":["AI Trends"],"tags":["business intelligence career path","latest machine learning innovation","machine learning for cities","naive bayes","simplilearn"],"author_name":"AIM Media House","publish_date":"2020-01-20T18:00:00","publication_year":"2020","word_count":1519,"keywords":["data science","artificial intelligence","machine learning","AI","neural network","latest machine learning innovation","business intelligence career path","ML","naive bayes","NLP","Aim","deep learning","analytics","simplilearn","machine learning for cities"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","neural network","NLP","data science","analytics","Aim"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/ai-ml-learning-path-definitive-guide\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":53143,"title":"Can Facebook Deliver On AR\/VR","content":"Ever since Facebook bought Oculus for $2 billion in 2014, it has been leaving no stone unturned to democratise VR. However, five year since the purchase, Facebook is still pushing hard to bring virtual reality for everyone. One of the co-founders of Oculus, Jack McCauley, doubts that the firm will ever breakthrough as there is an absence of demand in the market. But, Mark Zuckerberg, CEO of Facebook, believes that it will take a few more years before the AR\/VR technology gets mainstream. “I don’t think it’s a 2020 thing. But hopefully, it’s not a 2030 thing,” said Zuckerberg. Facebook is bullish about the technology and is committed to delivering robust solutions for an immersive experience. For which, the firm has devised a plan to increase the technology among a wide range of users. “We are in the early stages of a tremendous platform shift, bigger than the shift from web to mobile. It’s a shift of how humans will connect,” said Andrew Bosworth, VP of AR\/VR at Facebook. The Bottlenecks In The AR\/VR Landscape Michael Abrash, chief data scientist of Facebook, in an interview, said that the mass adoption of AR\/VR is five to ten years away due to numerous challenges. A shortage of the desired hardware is one of the prominent problems that has slackened the growth of the technology. However, Abrash envisions a world where VR will replace laptops and AR will displace mobile phones. In a bid to achieve that, Facebook is working on developing AR glasses that will be a featherweight, meets form factor requirements, and keeps the thermal in check. AR glasses can dissipate heat that will cause discomfort to users. Consequently, unless the hardware challenges are mitigated, it won’t be straightforward for firms to build products related to AR\/VR. Another pressing issue is that AR should interact with the real-world and be available at all the time. AR devices should understand the context of our surroundings in order to identify what we want and deliver results accordingly. To enable the AR glasses in providing relevant answers, Abrash said that the firm is working on developing AR audio that can also be used by people of hearing-impaired to gain information that they require while eliminating others. Facebook is betting high on devices that can be used to manage other devices with the brain. In September 2019, the company acquired CTRL-labs, a New York-based startup that focuses on empowering humans to control computers using their minds. “Interacting with AR is not going to be the way that we interact with our devices today. We will not be taking out phones all the time, won’t be using keyboard and mouse,” said Abrash. “I have to be multimodal, thus we are working on haptic gloves.” Efforts In Solving The Possibles Facebook has failed to impress its users, which might be the reason why McCauley is sceptical about its success. “If it were to sell, Facebook would have sold it by now,” explained McCauley. Zuckerberg, in October 2019, accepted that Facebook’s bet on the future of virtual reality isn’t going entirely according to plan. However, as per Abrash, the CEO is devoted to its vision of democratising AR\/VR. According to the chief data scientist of Facebook, to eliminate the challenges, it would require collaborative efforts, but he stressed on the fact that no other firm has invested anywhere close to what Facebook has in the AR\/VR landscape. Consequently, he expects that the social media giant will be the one to come up with next-generation VR\/AR. Looking at Facebook’s efforts in research and collaboration with other firms, one can anticipate that in the future the firm can blaze the trail and deliver AR\/VR solution for all. However, as Abrash, noted that he would be surprised if it could offer an innovative solution in five years. Outlook Facebook vision is to develop a product that will be used by everyone, just like it happened in the case of mobile phones. It wants to be that first iPhone or the BlackBerry in the AR\/VR landscape. Since the time period for the next wave in AR\/VR is expected in 10 years, it is difficult to make a linear conclusion about whether Facebook will gain its supremacy in AR\/VR. Nevertheless, amidst failure, Facebook is optimistic that it would be able to come up with a robust AR\/VR solution in years to come.","excerpt":"Ever since Facebook bought Oculus for $2 billion in 2014, it has been leaving no stone unturned to democratise VR. However, five year since the purchase, Facebook is still pushing hard to bring virtual reality for everyone. One of the co-founders of Oculus, Jack McCauley, doubts that the firm will ever breakthrough as there is […]","categories":["AI Features"],"tags":["ar\/vr"],"author_name":"Rohit Yadav","publish_date":"2020-01-03T19:00:00","publication_year":"2020","word_count":729,"keywords":["Go","programming_languages:R","AI","Modal","programming_languages:Go","R","ar\/vr","startup"],"extracted_tech_keywords":["AI","R","Go","startup","Modal","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/can-facebook-deliver-on-ar-vr\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10063190,"title":"Why does DistilBERT love movies filmed in India, not Iraq?","content":"Aurélien Geron, an ML consultant, a former Googler and the author of Hands-on Machine Learning with Scikit-Learn, Keras and Tensorflow highlighted a bias in DistilBERT, a small, fast, cheap and light Transformer model. Aurelien plotted a sentiment analysis on the model for movies filmed in different countries. The resultant map showed the highest positive sentiment to movies filmed in India and the lowest for movies filmed in Iraq. “I’m sure this model is used by analysts to measure the market’s sentiment in financial news feeds like Bloomberg’s. Are they compensating for the model’s country bias? I frankly doubt it,” Aurelien tweeted. Aurelien’s tweet attracted a lot of comments from ML professionals. Nils Reimers, NLP researcher at huggingface.co said: “The issue is that the model does only have a positive and negative class, but no neutral class. Hence it has to predict some sentiment to this neutral statement which does not make much sense. So I mainly see an issue with the model design to only have positive\/negative classes.” Chidananda AV, another ML practitioner, asked: “Do you think the induced bias is due to the data involving film reviews which did not have a good distribution (while finetuning) or due to bias in data corpus while pretaining resulting in a negative\/positive factor towards a topic(country in this subject).” “I don’t think it’s movie-related at all. I think it’s because of a strong bias built up during pre-training. For example, Germany is one of the very few countries in Western Europe to have a negative bias, and I’m pretty sure that’s WW2 related rather than movie-related,” Aurelien replied.","excerpt":"Aurelien said the DistilBERT’s bias might be developed during pretraining.","categories":["AI News"],"tags":["AI bias","DistilBERT","NLP","Sentiment Analysis"],"author_name":"Kartik Wali","publish_date":"2022-03-21T18:40:21","publication_year":"2022","word_count":266,"keywords":["Go","scikit-learn","Sentiment Analysis","machine learning","Keras","DistilBERT","AI","sentiment analysis","ML","NLP","TensorFlow","R","AI bias"],"extracted_tech_keywords":["AI","machine learning","ML","NLP","TensorFlow","Keras","scikit-learn","sentiment analysis","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/why-does-distilbert-love-movies-filmed-in-india-not-iraq\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":24005,"title":"Linux Launches Deep Learning Foundation For Open Source Growth In AI","content":"The Linux Foundation has launched the LF Deep Learning Foundation, an umbrella organisation which will support and sustain open source innovation in artificial intelligence, machine learning, and deep learning. The organisation will strive to make these critical new technologies available to developers and data scientists everywhere, said a statement published by LF. Founding members of LF Deep Learning include Amdocs, AT&T, B.Yond, Baidu, Huawei, Nokia, Tech Mahindra, Tencent, Univa, and ZTE, among others. LF Deep Learning, members are working to create a neutral space where makers and sustainers of tools and infrastructure can interact and harmonise their efforts and accelerate the broad adoption of deep learning technologies. “We are excited to offer a deep learning foundation that can drive long-term strategy and support for a host of projects in the AI, machine learning, and deep learning ecosystems,” said Jim Zemlin, executive director at The Linux Foundation. “With LF Deep Learning, we are launching the Acumos AI Project, a comprehensive platform for AI model discovery, development and sharing. In addition, we are pleased to announce that Baidu and Tencent each intend to contribute projects to LF Deep Learning. LF Deep Learning enables the open source community to support entire ecosystems of projects in these spaces, and we invite the open source community to join us in this effort.” As part of the LF Deep Learning launch, The Linux Foundation also announced the Acumos AI Project, a platform for the development, discovery, and sharing of AI models and AI workflows. Initial code for the Acumos AI Project has been contributed by AT&T and Tech Mahindra. The project strives to make the power of AI accessible to all by making it easy to create, share, discover, and apply machine learning, deep learning, and analytics models. The Linux Foundation will host the Acumos AI platform and the Acumos Marketplace moving forward with the goal of nurturing an active, large ecosystem around the project to sustain it over the long term. Code is available for download as of today. In addition to the Acumos AI Project, LF Deep Learning anticipates future project contributions from Baidu and Tencent, among others.","excerpt":"The Linux Foundation has launched the LF Deep Learning Foundation, an umbrella organisation which will support and sustain open source innovation in artificial intelligence, machine learning, and deep learning. The organisation will strive to make these critical new technologies available to developers and data scientists everywhere, said a statement published by LF. Founding members of LF […]","categories":["AI News"],"tags":["Linux"],"author_name":"Prajakta Hebbar","publish_date":"2018-04-25T10:16:29","publication_year":"2018","word_count":354,"keywords":["Go","machine learning","artificial intelligence","AI","ML","GAN","deep learning","ViT","analytics","Linux","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","analytics","R","Go","GAN","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/linux-foundation-deep-learning-foundation\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10062507,"title":"Grow your data science career with an MS Programme from Northwestern University School of Professional Studies","content":"The opportunities for data science professionals are endless-be it in the U.S. or India. As per a report by Market and Markets, the data science platform market size is expected to grow from $37.9 billion in 2019 to $140.9 billion by 2024, at a CAGR of 30.0% during the forecast period. As a result, professionals trained in data science and artificial intelligence capabilities will be able to work with cutting-edge technologies at leading firms and lead a rewarding career. While businesses are recruiting for data science professionals, they are ideally looking for someone who has a solid grip on the fundamental concepts as well as the tools needed to execute analytics projects. In addition to technical knowledge, employers also look for business acumen. To make students responsible for such challenging job roles, Northwestern University School of Professional Studies has introduced an M.S. in Data Science Programme in collaboration with Great Learning. The program will provide world-class education from Northwestern University, ranked 9 among U.S. universities (U.S. News & World Report 2022). Northwestern University has a proven record of over 160 years in educational excellence. It ranks 24 in the Global University Rankings 2022 ( The Times Higher Education World University Rankings 2022) and hosts students from over 75 countries. This MS in data science is a globally recognised degree programme, designed and delivered by experienced faculty with industry experience. The course will teach theoretical concepts and offer hands-on experience in popular programming languages, frameworks, and libraries. These skills have helped alumni of this course secure jobs with some of the best global organizations. Key Program Highlights Joining this course comes with many benefits that make it stand out from the rest. Some of them include: Master’s degree from a top global university and Northwestern alumni statusOnline learning with live sessions delivered by Northwestern facultyBest-in-class faculty with industry experiencePeer learning opportunities with professionals from diverse backgrounds and varied work experienceNetworking opportunities with two in-person learning programs in IndiaPlacement support and comprehensive career servicesKnowledge of relevant tools and languages like R, Python and TensorFlow The course has been created keeping working professionals in mind. As they have a strict schedule and often face a time crunch, the program provides the flexibility to study part-time from their home (students should plan to spend around 15 – 20 hours every week) while continuing their current jobs. Learning Outcomes After completing this course, students will be able to: Articulate analytics as a core strategy of data scienceTransform data into actionable insightsDevelop statistically sound and robust analytic solutionsDemonstrate leadershipFormulate and manage plans to address business issuesEvaluate constraints on the use of dataAssess data structure and data lifecycle Curriculum and Projects The first four terms (12 months) will cover foundation courses that will teach data science theory to students. Students will complete industry-relevant coursework in Artificial Intelligence from the fifth term onwards. There will be a capstone project towards the end of the degree programme. In addition to this, the students will work on a diverse portfolio of projects throughout the program, which can be leveraged during placements. The course will cover areas such as maths for data scientists, applied statistics with R, data science in practice, Python for data science, practical machine learning, database systems and data preparation, data governance, ethics and law and decision analytics. The course will also teach students topics like NLP, Artificial intelligence, deep learning, and computer vision. Faculty Profile Students will be able to take advantage of live classes with top-ranked faculty from Northwestern University and get hands-on exposure through quizzes, assignments and projects. All faculty members come with rich industry experience and a stellar academic profile. Student Demographics This Programme’s batch consists of students from diverse backgrounds with a wide range of work experience, for enriched mutual learning. The students in the current batch are employed in industry-leading organizations across nations, and some of their current employers are Google, Amazon, Flipkart, Cisco, TCS, HP, Indian Army, Accenture, Qualcomm, among others. Learner Experience The students currently pursuing the course are finding it greatly beneficial for their career. Some common points that learners are happy about include: Faculty support during the live sessions focused on concepts, practical implications and the much needed hands-on experience provided throughout the course.Best-in-class professors who come with high-depth of knowledge, crystal clear vision and gunshot points of clarification.Abundance of learning resources – both internal resources and external references. Highly responsive program team for resolving student queries and doubt clarifications. On-campus Residency Sessions The two one-week long residency programmes are unique because they provide an in-person experience designed to boost the online Masters in Data Science Programme. In addition, residencies are a great way to network with peers and faculty in a campus-style setting. During the residencies, Great learning conducts numerous career-support activities as well. The capstone project presentations, a crucial part of the curriculum, will be held during the residencies. It will allow students to gain exposure to each other’s work and approach to solving data science problems. GL Excelerate – Career Support GL Excelerate is a career support program designed exclusively for its learners. It aims to empower its learners with everything they need to succeed in their careers. In addition, it provides students with various facilities to help them grow in their careers. Student Placement – Great Learning has over 1200 hiring partners and runs a dedicated process to place students with well-known domestic and global organizations.Curated Jobs – Access to specialized jobs, relevant to learners’ experience and domain.Interview preparation-Workshops conducted by industry experts, to help the students prepare for technical interviews.1:1 Career mentorship-Expert 1:1 career mentorship personalized to the student’s experience and industry. Admissions All applicants are required to submit an online application. Applicants can check the status of their application at any time by logging into the online Self Service Center portal. On completing the program, students will be able to attend an “in-person” graduation ceremony and obtain an MS degree from the prestigious Northwestern University that adds immense value to participants’ resumes. Program Fee and Flexible Payment Options The estimated tuition is $13,000 (approx Rs 9.36 Lakhs) and can be paid in six quarterly instalments of $2,167 each (payable at the start of every term). There is also the option of taking educational loans as Great Learning has tied up with financial institutions. Additionally, the students will have to bear the cost of travelling, boarding\/lodging, and other expenses for the residency sessions. If you want to study at one of the best universities globally and learn data science from the best minds to build a rewarding career, the MS in data science from Northwestern University School of Professional Studies is the right choice for you. What are you waiting for? Click here to apply.","excerpt":"Northwestern University School of Professional Studies has introduced an MS in Data Science programme in collaboration with Great Learning, which will help you build a rewarding career in data science.","categories":["AI Trends"],"tags":["AI (Artificial Intelligence)","Data Science","data science professionals","Machine Learning","Python"],"author_name":"Sreejani Bhattacharyya","publish_date":"2022-03-10T17:00:00","publication_year":"2022","word_count":1122,"keywords":["data science professionals","data science","artificial intelligence","machine learning","AI","Machine Learning","computer vision","Python","NLP","Aim","deep learning","analytics","Data Science","TensorFlow","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","deep learning","NLP","computer vision","data science","analytics","Aim","TensorFlow"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/grow-your-data-science-career-with-an-ms-programme-from-northwestern-university-school-of-professional-studies\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10049303,"title":"How to Visualize and Debug Machine Learning Models using ELI5?","content":"Machine learning models are generally considered black-box models in the entire community despite their massive implementation. It becomes very essential to Understand how particular predictions are being made and how models focus on various aspects of parameters it has learned. Models are usually assessed using certain evaluation matrices on a given test dataset. Real-world data, on the other hand, is frequently different, so the evaluation metric may not accurately reflect the product’s purpose. In addition to such metrics, evaluating individual predictions and their justifications is a viable solution for optimizing performance. In this article, we will discuss debugging and visualizing machine learning algorithms using ELI5. ELI5 is a tool in Python that is used to visualize and debug various Machine Learning models using a unified API. The major points to be covered in this article are given below. Table of Contents Explainability and Interpretability in Machine LearningELI5 (Explain Like I’m 5) XGBoost with ELI5Example of Keras ImplementationAdvantages and Usage of ELI5 Now, let us start with understanding explainability and interpretability. Explainability and Interpretability in Machine Learning Explainability and interpretability are frequently used in machine learning and artificial intelligence. Even though they are extremely similar, it’s worth exploring the differences, if only to demonstrate how difficult things can get once you start looking into machine learning systems. The amount to which a cause and effect may be observed within a system is known as interpretability. To put it another way, it refers to your ability to forecast what will happen in response to a change in input or computational parameters. It’s the ability to look at an algorithm and pertaining, what’s going on there. Meanwhile, explainability refers to how well the internal mechanics of a machine or deep learning system can be communicated in human terms. It’s easy to overlook the tiny distinction with interpretability, but think of it this way: interpretability is about being able to understand mechanics without necessarily knowing why. Explainability refers to the ability to explain what is happening in detail. Simple models (Like  Linear or Logistic regression) can be used to explain findings for a sample data set. Typically, these models are insufficient, and we must go to Deep Learning models, which deliver great performance but are a mystery to the majority of Data Science practitioners. Machine learning models are currently utilized to make a variety of essential judgments, including fraud detection, credit rating, self-driving, and patient examination. Explainability and Interpretability It becomes very important to every practitioner that enhancing the interpretability and explainability of models is now crucial in most development and that can make us stand differently than others. We can address the issues and goals of the problem statement correctly by understanding how algorithms work. ELI5 (Explain Like I’m 5) ELI5 is a Python toolkit that uses a uniform API to visualize and debug diverse Machine Learning models. It supports all scikit-learn algorithms (including the fit() and predict() methods). It includes built-in support for numerous ML frameworks and allows you to explain white-box models (Linear Regression, Decision Trees) as well as black-box models (Keras, XGBoost, LightGBM). It is applicable to both regression and classification models. Now we are going to see how ELI5 interprets and explains a model using its eli5.show_weights and eli5.show_prediction API. The practical demo is divided into two parts. First, we are going to interpret and explain XGBoost. Following it we will see the same for Keras application. XGBoost with ELI5 ! pip install eli5 from xgboost import XGBClassifier import eli5 from sklearn.datasets import load_breast_cancer from sklearn.model_selection import train_test_split import pandas as pd import numpy as np The data set here we are using is the sklearn built-in data set for breast cancer prediction while implementing, create a pandas data frame for breast cancer dataset with the proper header this is because when we execute eli5 it retrieves feature information from the model. data = load_breast_cancer() df = pd.DataFrame(data.data) df.columns = data.feature_names df['target'] = data.target Build a classifier: model = XGBClassifier() model.fit(x_train,y_train) Now we need to use just two simple functional API’s of ELI5 as below. eli5.show_weights(model, top=30) eli5.explain_prediction_xgboost(model,x_test.iloc[0]) The left side shows weights assigned for each feature and the right side shows the prediction for one instance As you can see from the above two tables how XGBoost assigned weights for each feature based on training data and from the other table, for a particular instance, to reach a probability of 0.981 for class 1 how each feature has contributed. Similarly, next, we are going to see the same interpretation for Keras’s application. ELI5 with Keras Implementation If we have a model that takes an image as input and returns class scores (probabilities that a specific object is present in the image), we can use ELI5 to see what was in the image that caused the model to predict a specific class score. For the Keras demo, we are using a VGG16 pre-trained network and its interpretation for a random image. from tensorflow.keras.applications import VGG16 from tensorflow.keras.applications import VGG16 import keras vgg16 = VGG16(include_top=True, weights='imagenet', classes=1000) # load image im = keras.preprocessing.image.load_img('\/content\/HAL-TEDBF-Fighter-Jet-With-Vikrant-Aircraft-Carrier-Art.jpg', target_size=(224, 224)) doc = keras.preprocessing.image.img_to_array(im) doc = np.expand_dims(doc, axis=0) doc = keras.applications.vgg16.preprocess_input(doc[0]) # visualize the image keras.preprocessing.image.array_to_img(doc[0]) # explain eli5.show_prediction(vgg16, doc) As you can see ELI5 shows how the VGG16 looks for objects for which a given image is to be classified. Advantages and Usage of ELI5 ELI5, can use an existing function and produce good results that are formatted as well, It also allows code to be reused across different machine learning frameworks, It can deal with a slew of minor inconsistencies. ELI5 can be used to inspect basic model parameters and to figure out how the models perform on a global scale. ELI5 can be used to examine specific predictions provided by a single model, as well as the decisions made by the models. Conclusion We generally tend to use many models and algorithms for our problem and choose one which performs better than others. Practically evaluating each such model is a tedious task that can slow down our development process. By using the ELI5 and mastering it for a variety of algorithms we can easily choose a model which can outperform our task. From this article, we have seen how interpretability and explainability play an important role. References ELI5 DocumentationInterpretability and ExplainabilityML Explainability        Link for the above codes","excerpt":"From this post you will come to know how particular predictions are being made and how models focus on various aspects of parameters it has learned.","categories":["AI Trends"],"tags":["Data Science","Deep Learning","eli5","Keras","Machine Learning","machine learning document classification","optimization models","XGBoost"],"author_name":"Vijaysinh Lendave","publish_date":"2021-09-26T10:00:00","publication_year":"2021","word_count":1054,"keywords":["data science","scikit-learn","artificial intelligence","eli5","Keras","machine learning document classification","AI","machine learning","ML","Machine Learning","Ray","XGBoost","deep learning","optimization models","Deep Learning","Data Science","TensorFlow"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","data science","Ray","TensorFlow","Keras","scikit-learn"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/how-to-visualize-and-debug-machine-learning-models-using-eli5\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10084166,"title":"Microsoft CEO Meets PM Modi, Praises India’s ‘Digital Vision’","content":"Prime Minister Narendra Modi on Thursday met Microsoft Chief Executive Officer (CEO) Satya Nadella in New Delhi and spoke about the ‘Digital India’ vision. The Hyderabad-born Microsoft chief applauded the Indian government’s focus on “sustainable and inclusive economic growth”. Nadella also met External Affairs Minister S Jaishankar and discussed several issues pertaining to governance and security in the digital domain. The CEO is on a four-day visit to India, where he addressed industry leaders at the Microsoft Future Ready Leadership Summit in Mumbai and Bengaluru. Besides talking about the need for digital transformation in every organisation and the potential benefits of Microsoft Cloud in fostering creativity, advancing the national economy, and accelerating the growth of Indian firms, Nadella underlined how Microsoft is working towards the overall betterment of society with responsible AI. He emphasised how Microsoft is innovating across the tech stack to support the country’s ecosystem of developers, startups, and companies in every industry when describing his vision for India’s digital journey. In December 2022, Google and Alphabet chief executive officer Sundar Pichai also visited India for the “Google for India” event and called on Prime Minister Narendra Modi, President Droupadi Murmu in Delhi and Telecom and IT Minister Ashwini Vaishnaw. He also commended the technology boosting in India and promised to support India’s G20 presidency to promote an open, and connected internet for all. Why are the big tech chiefs visiting India? The existing AI models are trained on the data collected from the West, resulting in bias. These bias outputs mostly include established societal bias based on race, gender, sex, nationality, and age. Two well-known instances are racial bias in healthcare and facial recognition. Because there isn’t sufficient data on brown people in the training data, these systems struggle to identify persons of colour. While Microsoft is partnering with Apollo to develop a cardiac prognosis model that will be trained on data from South Asians,  Google is also collaborating with the medical chain to use AI for examining X-rays for tuberculosis. Google has also introduced  Project Vaani to make internet more language inclusive and has announced grants to IIT-Madras for setting up a centre for Responsible AI, among others. Microsoft also spoke widely about using the tech for good, like how the State Bank of India (SBI) uses Power Apps to make financial services more accessible to those with special needs. Air India is using Microsoft 365 to increase productivity. The big techs are coming to India to garner people’s support in collecting data for AI models to remove bias.","excerpt":"Satya Nadella applauded the Indian government’s focus on “sustainable and inclusive economic growth.","categories":["Global Tech"],"tags":["AI (Artificial Intelligence)","Data Science","Machine Learning","Microsoft","Satya Nadella"],"author_name":"Shritama Saha","publish_date":"2023-01-05T16:06:08","publication_year":"2023","word_count":423,"keywords":["Satya Nadella","Go","TPU","AI","AI (Artificial Intelligence)","digital transformation","Machine Learning","Git","responsible AI","Ray","ViT","GAN","Data Science","R","Microsoft"],"extracted_tech_keywords":["AI","Ray","TPU","R","Go","Git","GAN","ViT","responsible AI","digital transformation"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/microsoft-ceo-meets-pm-modi-praises-indias-digital-vision\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10014163,"title":"AWS Introduces Amazon HealthLake For Analysis &#038; Transformation Of Healthcare Data In Cloud","content":"The latest offering on healthcare from Amazon was introduced in the form of Amazon HealthLake, which was announced at the on-going AWS re:Invent 2020 event. Amazon HealthLake enables healthcare organisations to store, transform, and analyse large life science data, to the tune of petabytes, in AWS. It can understand and extract medical information that includes rules, procedures, and diagnoses in real-time. Amazon HealthLake is a HIPAA (Health Insurance Portability and Accountability Act)-eligible service for healthcare providers, insurance companies, and pharmaceuticals. While introducing HealthLake at re:Invent 2020, Swami Sivasubramaniam, the vice president of Amazon Machine Learning for AWS said that while there has been an ‘explosion’ of digitised health data with the introduction of e-medical records, the amount of valuable information that could be derived out of this data hasn’t been able to keep pace. He then added, “With Amazon HealthLake, healthcare organisations can reduce the time it takes to transform health data in the cloud from weeks to minutes so that it can be analysed securely, even at petabyte scale.” Why Conventional Methods Fail A lot of sectors are now applying machine learning to their data to derive maximum value information to apply in their services and processes, and the healthcare industry is no different. This industry requires careful information extraction from important patient data. This requires the application of analytics and machine learning, which in turn improves the care provided, helps in analysis of population health trends, and supports greater operational efficiency. It is much easier to state in words, but the fact remains that clinical data is highly complex. What makes things more difficult is that it is often siloed, incompatible, stored across different on-premise locations. Apart from gathering these critical pieces of information in the FIHR (Fast Healthcare Interoperability Resources) format, which is the standard that describes data formats for exchanging electronic health records, one also needs to tag, index, and structure data. Only after these processes, data is deemed fit for the query. As per current trends, healthcare organisations have rule-based tools for process automation to transform unstructured data such as medical histories, physician notes, and imaging reports. Since these are not generalised tools, they break down when applied to other systems or when there is variation in spelling, and have typos or grammatical errors. Another popular tool used by healthcare institutions is optical character recognition (OCR) software to process data. However, the problem with OCR is that it is not 100% effective, and as technicians have to manually enter data, the process gives scope for error and adds to the digitisation process. The cost and the complexity of the whole process of analysing and discovering trends to make precise predictions prohibits organisations to fully utilise the potential of data. How Can HealthLake Help? Amazon’s new offering brings together all the data and helps organisations to ‘make sense’ of it and make more precise predictions about patient health. HealthLake stores, tags, indexes, standardise queries data at a petabyte-scale in the cloud. It further allows organisations to easily copy data from on-premises systems to a secure data lake in the cloud and normalise every record across disparate formats. Amazon HealthLake uses machine learning trained systems to understand medical terminology to identify, tag, and enhance data with standardised labels such as — medications, conditions, diagnoses, and processes. This helps in avoiding potential error that is observed with conventional tools. Amazon also suggests that when used with other AWS analytics and machine learning services like Amazon QuickSight and SageMaker, technicians will also be able to get interactive dashboards and will be equipped to build and deploy custom ML models. Wrapping Up The launch of HealthLake comes amidst Amazon’s aggressive efforts to charter healthcare AI domain. It may be noted that just last year, the company had launched Transcribe Medical to transcribe medical speech for clinical staff of primary care. A year before that, Amazon had made three HIPAA eligible offerings — Transcribem Translate, and Comprehend.","excerpt":"The latest offering on healthcare from Amazon was introduced in the form of Amazon HealthLake, which was announced at the on-going AWS re:Invent 2020 event. Amazon HealthLake enables healthcare organisations to store, transform, and analyse large life science data, to the tune of petabytes, in AWS. It can understand and extract medical information that includes […]","categories":["Global Tech"],"tags":["machine learning and data transform"],"author_name":"Shraddha Goled","publish_date":"2020-12-14T16:00:00","publication_year":"2020","word_count":656,"keywords":["Go","machine learning","AWS","AI","ML","Git","machine learning and data transform","analytics","GAN","R","data lake"],"extracted_tech_keywords":["AI","machine learning","ML","analytics","AWS","R","Go","Git","data lake","GAN"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/aws-introduces-amazon-healthlake-for-analysis-transformation-of-healthcare-data-in-cloud\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10166228,"title":"Mistral’s New Open Source Model ‘Mistral Small 3.1’ Outshines Gemma 3, GTP-4o Mini","content":"French AI startup Mistral announced the release of Mistral Small 3.1, its new lightweight model, on Monday. It’s a state-of-the-art multimodal, multilingual, open-source model available under an Apache licence. The new model builds on Mistral Small 3, with improved text performance, multimodal understanding, and an expanded context window of up to 128k tokens. Mistral claims that the model outperforms comparable models like Gemma 3 and GPT-4o mini while delivering inference speeds of 150 tokens per second. According to the official performance chart in the announcement post, Mistral Small 3.1 surpasses other leading small proprietary models in various applications involving text, multimodal inputs, and managing long contexts. The blog post stated, “Mistral Small 3.1 is a versatile model designed to handle a wide range of generative AI tasks, including instruction following, conversational assistance, image understanding, and function calling. It provides a solid foundation for both enterprise and consumer-grade AI applications.” The announcement also highlights that the model is lightweight and can run on a single RTX 4090 or a Mac with 32 GB RAM. Moreover, it is ideal for virtual assistants and applications where quick, accurate responses are needed. Mistral has clarified that the model can be fine-tuned to specialise in specific domains and also made the pre-trained base model available along with the release. The company mentioned that it can be used across various enterprise and consumer applications that require multimodal understanding. The model can be downloaded through Hugging Face. One can also try it via API on Mistral AI’s developer playground, La Plateforme; it is also available on Google Cloud Vertex AI. It will be available on NVIDIA NIM and Microsoft Azure AI Foundry in the coming weeks.","excerpt":"Mistral Small 3.1 can run on a single RXT 4090 or a Mac with 32 GB RAM.","categories":["AI News"],"tags":["mistral"],"author_name":"Ankush Das","publish_date":"2025-03-18T14:34:12","publication_year":"2025","word_count":279,"keywords":["Go","Hugging Face","AI","GPT-4o","R","virtual assistants","Aim","mistral","Gemma 3","generative AI","Azure"],"extracted_tech_keywords":["AI","generative AI","GPT-4o","Gemma 3","Aim","Hugging Face","virtual assistants","Azure","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/mistrals-new-open-source-model-mistral-small-3-1-outshines-gemma-3-gtp-4o-mini\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10022758,"title":"Guide To Interactive Image Synthesis With Anycost GANs","content":"Generative adversarial networks (GANs) have become exceedingly good at photorealistic image synthesis from randomly sampled latent codes. Additionally, the generated output images can be easily transformed\/edited(e.g.,  adding a smile or glasses) by tweaking the latent code. However, due to the high computational cost of large generators it usually takes seconds to see the results of a single edit on edge devices. Inspired by the quick preview option in modern creative softwares Ji Lin, Richard Zhang, et al, proposed anycost GAN for more immediate and interactive image editing in their paper – “Anycost GANs for Interactive Image Synthesis and Editing”. The anycost generator can be executed at a wide range of computational costs while producing visually consistent outputs. A low-cost generator is used for fast, responsive previews during image editing, and then the full-cost generator is used to render high-quality final outputs. Architecture & Approach (Source: Anycost GAN GitHub Repository) Anycost GAN uses a subsection of the whole generator G’ to independently produce an output x’ = G’(w) which is perceptually similar to the full generator G(w). This G’(w) is used for fast preview during interactive editing, and full G(w) is used to render the final high-quality outputs. This allows the model to be deployed on diverse hardware and the end-users get to choose between different preview qualities depending on their hardware. Sampling-Based Multi-Resolution Training The obvious choice for enabling a range of inference costs is to use different image resolutions. The ProGAN and StyleGAN family architectures already produce lower-res intermediate images but these are not visually similar to the final output image. To produce more accurate lower-resolution intermediate outputs anycost GAN enforces a multi-scale training objective. The generator produces incrementally higher resolution outputs after each block gk here K is the total number of network blocks. The intermediate low-res outputs (x k) are used for the preview. Existing approaches for multi-resolution training, like MSG-GAN, train the generator to support different resolutions by using just one discriminator for images of all resolutions. This kind of multi-resolution training mechanism affects the quality of the generated results on large-scale datasets, to overcome this image degradation issue anycost GAN uses a sampling-based training objective.At each iteration,  a single resolution is sampled and trained both for the generator G and the discriminator D as shown in the figure. (Source: Anycost GAN GitHub Repository) When sampling a lower resolution image, the translucent parts are not executed. The intermediate output of G for a lower resolution is passed through a fromRGB convolution layer to increase channels and then fed to an intermediate layer of D. The objective function of this multi-resolution training is: Adaptive Channel Training Reducing the resolution from 1024×1024 to 256×256 only reduces the computation cost by 1.7×, despite having 16× fewer pixels. Just using variable resolutions isn’t enough to improve the generator speed significantly, to further improve the speed the anycost generator is trained to support variable channels. For each training iteration, a randomly sampled channel multiplier configuration is used and the corresponding subset of weights is updated. To preserve the most “important” channels during sampling the model is initialized using the multi-resolution generator from the previous stage and the channels of convolutional layers are sorted according to the magnitude of kernels, from highest to lowest. The most important ????c channels are sampled where  ???? ∈ [0.25, 0.5, 0.75, 1] and c is the number of channels in the layer. The adaptive-channel training objective is written as follows: Here C means the channel configurations for each layer. To keep the consistency of the output across different sub-networks a combination of MSE loss and LPIPS loss is added in the form of the following consistency loss: Generator-Conditioned Discriminator Unlike regular GAN training, anycost GAN trains many sub-generators of different channels and resolutions at the same time. Using one single discriminator for all sub-generators of different channel configurations results in poor performance. To overcome this challenge the anycost discriminator is conditioned on the generator architecture, a learning-based approach is used to implement the conditioning. The channel configuration is first encoded into g_arch vector using one-hot encoding, g_arch is then passed through a fully connected layer to form the per-channel modulation. The feature map is modulated using the conditioned weight and bias before passing to the next layer. Interactive Editing with Anycost GAN Clone the Anycost GAN repository and navigate into the anycost-gan folder. git clone https:\/\/github.com\/mit-han-lab\/anycost-gan.git cd anycost-gan Install the dependencies by creating a new Anaconda environment using environment.yml conda env create -f environment.yml Or update an existing environment conda env update --name myenv --file environment.yml Let’s first visualize the preview (reduced) forms of an image for the four resolution choices and channel widths. Import necessary libraries and classes import os import torch import torch.nn.functional as F import numpy as np from tqdm import tqdm_notebook as tqdm import matplotlib %matplotlib inline import matplotlib.pyplot as plt import json from PIL import Image, ImageFont, ImageDraw import models from models.dynamic_channel import set_sub_channel_config, remove_sub_channel_config, set_uniform_channel_ratio, reset_generator Download and load the anycost generator trained on FFHQ dataset g_ffhq = models.get_pretrained('generator', 'anycost-ffhq-config-f').to(device).eval() Create functions for plotting the 4×4 grid of preview images. def torch_to_np_image(x): assert x.shape[0] == 1 x = x.squeeze(0) return ((x.permute(1, 2, 0) + 1) * 0.5 * 255).cpu().numpy().astype('uint8') def add_legend_to_figure(x, full_res): h, w, c = x.shape x_pad_rate = 0.1 y_pad_rate = 0.05 pad_h = int(h + h * y_pad_rate) pad_w = int(w + w * x_pad_rate) pad_x = np.zeros([pad_h, pad_w, c], dtype=x.dtype) pad_x[...] = (255) pad_x[-h:, -w:] = x img = Image.fromarray(pad_x) draw = ImageDraw.Draw(img) font = ImageFont.truetype(font_path, int(100 * full_res \/ 1024)) for i, text in enumerate(['1x', '0.75x', '0.5x', '0.25x']): text_w, text_h = draw.textsize(text, font=font) x_coord = x_pad_rate * w * 0.9 - text_w y_coord = (y_pad_rate + i * 1.\/4 + 1.\/8) * h - 0.5 * text_h draw.text((x_coord, y_coord), text ,(0,0,0), font=font) for i in range(4): text = str(full_res \/\/ (2 ** i)) text_w, text_h = draw.textsize(text, font=font) x_coord = (x_pad_rate + i * 1.\/4 + 1.\/8) * w - 0.5 * text_w y_coord = y_pad_rate * h * 0.9 - text_h draw.text((x_coord, y_coord), text ,(0,0,0), font=font) return np.asarray(img) def get_4x4_grid(g, latent, truncation=0.5, crop=False):  using z code (latent) images = [] full_resolution = g.resolution if truncation < 1: mean_style = g.mean_style(10000) else: mean_style = None with torch.no_grad(): for channel_mult in [1, .75, .5, .25]: set_uniform_channel_ratio(g, channel_mult) _, all_rgbs = g(latent, return_rgbs=True, truncation=truncation, truncation_style=mean_style, randomize_noise=False) all_rgbs = all_rgbs[-4:][::-1] all_rgbs = [F.interpolate(rgb.clamp(-1, 1), size=(full_resolution, full_resolution), mode='bilinear', align_corners=True) for rgb in all_rgbs] images.append(torch.cat(all_rgbs, dim=3)) image_grid = torch.cat(images, dim=2) image_grid_np = torch_to_np_image(image_grid) return add_legend_to_figure(image_grid_np, full_resolution) Plot the image grid latent = torch.randn(1, 1, 512, device=device) img_out_np = get_4x4_grid(g_ffhq, latent) plt.figure(figsize = (6,6)) plt.imshow(img_out_np) plt.axis('off') Now, let’s try the interactive editor python demo.py A PyQt5 window will open. Full generator G(w) previews will take 3-8 seconds to render, sub-generator G’(w) previews will take roughly 1 second to render. References GitHubColab Notebook (for visualizing image synthesis at diverse cost levels.)Paper Want to learn more about state-of-the-art image synthesis GANs? Check out our posts on Conditional Image Generation and Image-to-Image Translation.","excerpt":"Anycost GAN supports fast, responsive previews during image editing by executing the generator at a wide range of computational costs.","categories":["Deep Tech"],"tags":["Guide"],"author_name":"Aditya Singh","publish_date":"2021-03-24T10:00:00","publication_year":"2021","word_count":1168,"keywords":["Go","NumPy","TPU","AI","ML","Colab","Ray","Python","Matplotlib","R","Guide"],"extracted_tech_keywords":["AI","ML","Ray","Colab","NumPy","Matplotlib","TPU","Python","R","Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/guide-to-interactive-image-synthesis-with-anycost-gan\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10102310,"title":"Biden’s AI Executive Order Faces Backlash","content":"Biden has signed the executive order which carries the weight of the law without requiring congressional approval. The swift implementation of laws around AI has been met with criticism from the open source and research community. Clem Delangue, Co-founder and CEO of Hugging Face, posted on X saying, “Compute or model size thresholds for AI building would be like counting the lines of code for software building.” Richard Socher, CEO of you.com also said that regulation shouldn’t apply to foundational models and research but on the applications of AI. The applications of AI that pose risks that include privacy, legal, security concerns which are not addressed in the executive order. Andrew Ng said, “There are definitely large tech companies that would rather not have to try to compete with open source, so they’re creating fear of AI leading to human extinction.” He accused big tech companies of grossly exaggerating the risks of AI, while saying that it was, “because they want to dominate the market.” This reasoning is echoed by Yann LeCun as well. He recently posted on X saying that some of them were lobbying in an attempt to capture the regulations and the AI industry. He said that it isn’t AI research or development that needs to be regulated but its applications. Altman, Hassabis, and Amodei are the ones doing massive corporate lobbying at the moment.They are the ones who are attempting to perform a regulatory capture of the AI industry.You, Geoff, and Yoshua are giving ammunition to those who are lobbying for a ban on open AI R&D.If…— Yann LeCun (@ylecun) October 29, 2023 What’s in it for the Big Tech In contrast to the bill, 15 major tech companies have agreed to implement voluntary AI safety commitments. But on the other hand, the government regulatory body has said, ‘it is not enough.” Biden signed the order and said, “To realise the promise of AI and avoid the risk, we need to govern this technology,” He claimed that, “In the wrong hands, AI can make it easier for hackers to exploit vulnerabilities in the software that makes our society run.” However this isn’t expected to last for a long time until the congress formulates long term legislation on this novel technology. Further, the Act mandates that developers of powerful AI must disclose data of the safety test results, ensuring secure deployment. Simultaneously, the National Institute of Standards and Technology will set standardised rules to guide AI system development. An ‘AI Bill of Rights’ will protect against potential AI-related harms, emphasising privacy, equity, and worker support. To maintain AI leadership, significant investments are channelled into research and development. Policies are being crafted to ensure AI’s responsible and ethical use across governmental functions, aiming for societal benefit while curbing potential negative impacts, reads the bill. Currently, the bill targets the technologies that have already been deployed. “I think, are the ones that we’re really concerned about.” said Nicol Turner Lee, the director of the Center for Technology Innovation. The rigorous testing and impact of assessment required by the bill would potentially back small companies and might not have much effect on the larger ones. Any AI model that required more than 1e26 floating point operations or 1e23 integer operations to build must report to the government. This is just above the largest existing model which is OpenAI’s GPT-4. This will affect only big techs for now like OpenAI and Google. However there is no clear way on how this will be monitored. Imagine explaining to Kamala Harris where the 6 in GPT FLOPs calculated as 6 * param_count * token_count comes from, or how would fine-tuning FLOPs add on to a pre-trained \"unregulated\" ckpt. And wasting time on all this instead of actually building AGI.— Aravind Srinivas (@AravSrinivas) October 30, 2023 In comparison with EU’s AI act Even though the executive order will take time to be implemented and even then it is unclear how they will be monitored and regulated, the EU AI act is a more comprehensive document formulating the guidelines on AI development. The key difference between the two is that the EU AI Act places a strong emphasis on transparency and accountability. It requires AI developers to disclose information about their AI systems, such as how they were developed and how they work. It also requires AI developers to take steps to ensure that their AI systems are accountable and that they can be held responsible for any harm that they cause. The Biden AI executive order also places an emphasis on transparency and accountability, but it is not as prescriptive as the EU AI Act. The Biden AI executive order encourages AI developers to adopt voluntary transparency and accountability measures, but it does not require them to do so. The EU AI Act applies to all AI systems, regardless of their size or complexity. The Biden AI executive order, on the other hand, only applies to certain types of AI systems, such as those that are used by the government or that pose a high risk to public safety.","excerpt":"The swift implementation of laws around AI in the US has been met with criticism from the open source and research community.","categories":["AI Trends"],"tags":["AI laws","AI Safety","Big Tech","eu ai act","law","opensource","Privacy"],"author_name":"K L Krithika","publish_date":"2023-10-31T17:00:00","publication_year":"2023","word_count":845,"keywords":["Go","Privacy","law","Hugging Face","OpenAI","AI","AWS","AI laws","RAG","GPT","Aim","Big Tech","opensource","AI Safety","AI safety","R","eu ai act"],"extracted_tech_keywords":["AI","OpenAI","Aim","Hugging Face","RAG","AWS","R","Go","GPT","AI safety"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/bidens-ai-executive-order-faces-backlash\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10082387,"title":"Council Post: Business Outlook for Data Leaders in 2023","content":"“The goal is to turn data into information and information into insight.” – Carly Fiorina, former chief executive officer, Hewlett Packard. In a world driven by data, industry leaders are successfully extracting actionable insights through precise data management and analytics. This enables them to innovate quickly, develop better frameworks, and govern modifications more efficiently. With digital transformation set to sweep through every industry in 2023 and beyond, data leaders seem to be aware that they now have to equip their organisations to face uncertainty. New technologies that facilitate faster and more accurate access to insights are fostering new thinking across businesses. In the last few years, modern artificial intelligence, machine learning, data analytics, and robots have been driving the changing environment—forcing established business models to adapt. Uncertainty is increasing as a result of the unparalleled global pace of change being driven by digital transformation. In the light of this, it is no longer possible for people, organisations, or governments to base their planning for the future on the assumption that things would mostly stay the same. Instead, they must investigate and get ready for a variety of different scenarios that represent possible changes and the fresh opportunities, along with the difficulties they might present. Using such a strategy can ensure that the strategies and policy frameworks created today are resilient and adaptable to the direction, velocity, and scope of changes that the digital transformation may bring. Here are some ways to ensure businesses are ready for this uncertain future. Data Democratisation The majority or all of the data was previously owned by IT teams. Business users would approach IT to request access to a specific data collection, and in response, IT would hand over a huge, disorderly spreadsheet. Data democratisation is the process through which an effective organisation makes data available to all employees and stakeholders—regardless of their technical background—and teaches them on how to deal with data. The ‘data’ in data democratisation is any information you might be able to learn about your company or organisation. Data is ubiquitous, and it has the power to make every aspect of a company more efficient. For instance, in a data democracy, sales benefits from marketing data governance and having access to marketing data to track the leads produced by a particular campaign. Additionally, marketers can obtain sales information to determine how effective a new marketing channel is. Cloud-as-a-service Infrastructure, platforms, or software that are hosted by external providers and made accessible to users online are referred to as cloud services. Cloud services make it easier for user data to move back and forth from front-end clients to the provider’s systems through the internet along with encouraging the development of cloud-native apps and the adaptability of cloud-based activities. Moving data analytics to the cloud has many advantages, including improved return on investment. These platforms were created by engineers to process and analyse enormous amounts of data quickly. Cloud analytics, in turn, assists companies in deriving value from their data for better decision-making, enhanced operations, and accelerated growth. Companies that employ their tools effectively can get on-demand processing, storage, and data warehousing from cloud platforms like Amazon Web Services (AWS) for certain analytics use cases. Additionally, they can significantly shorten the time needed to build up a data infrastructure. Consequently, organisations can respond rapidly to assess the viability of a business hypothesis. If it turns out to offer new information, the solution can be improved upon and made into a product. If not, the resources can be promptly released so that the next hypothesis can be considered. Real-time data When digging into data, it is better to be aware of what is going on. This is the reason why real-time data is increasingly becoming a valuable source of information for data leaders. However, working with real-time data requires a sophisticated data and analytics infrastructure. Data that is displayed as it is being acquired is referred to as real-time data. The concept of processing real-time data is currently prevalent in emerging technologies, such as those that provide mobile devices like phones, laptops, and tablets with convenient apps that give up-to-the-minute information. But one of the most important use cases of real-time analytics is detecting potential fraud. For instance, fraudulent credit card transactions can be identified and prevented in real time. Although conventional analytical methods are capable of detecting fraud, they are excessively slow; the processing and analysis of the data may take hours. The detection of fraud in real time is possible using real-time analytics. Agile data analytics Agile data and analytics models make it feasible for data leaders to not only innovate but also differentiate and grow digitally. By integrating a variety of data analytics, AI, and ML solutions, data leaders are able to create a user-friendly and seamless experience. Agile data analytics are tools and processes you can use to track the value your agile system provides to the various value streams in your firm as a whole. Agile data analytics can sometimes go further and enable you to use that data in reports that examine team performance, record the quantity of work performed, enhance your systems, and even assist you in making decisions. Tracking relevant agile metrics, context, and measuring parameters can help you better understand the inner workings of your teams and value streams. When considering your metrics, these are all very significant factors to take into account because measurements are meaningless without context. Agile data analytics assist in highlighting the significance of the data and patterns. Integration of AI and ML in data management The integration of automation in data management is empowering businesses to perform complex data-related tasks, along with ensuring regulatory requirements. This increased use of AI and ML solutions, as well as tools, is assisting them to stay relevant and compliant in an increasingly regulated data ecosystem. The upcoming ‘zettabyte apocalypse’—when the volume of data generated will overwhelm storage capabilities—can be navigated and monetised with the use of automated data management. Several businesses still struggle with data issues despite the fact that data quality and governance are the cornerstones of digital transformation programmes for contemporary business setups. Data governance, data quality, data catalogues, and data lineage must all be included in an integrated data management solution for modern data-driven enterprises if users are to be able to access data across the organisation to enhance workflows and operations. Beyond precise reporting, quicker analysis, and better data-driven decision making, firms gain a lot more from automating data governance concepts and related components. Fostering a data-driven culture By inculcating the concept of a data-driven culture within the enterprise, a key resource to implement ideas across every department can be envisioned. Today, the primary goal for business leaders is to empower their employees to actively employ data and bolster their daily work. A data-driven culture will enable employees to use analytics and statistics to streamline their workflow and complete their jobs. Before introducing new policies or implementing notable changes at work, team members and corporate leaders gather data to gain insights into the implications of their decisions. Such data-driven decision making is supported by empirical facts, allowing leaders to make smart decisions that have a favourable impact on the company’s bottom line. Company executives may consult their instincts when making decisions, but they only take precise measures in response to what the data shows. Based on findings from a McKinsey Global Institute study, MicroStrategy estimates that data-driven organisations are 20x or more likely to gain new clients and 6x more likely to maintain them. With data analytics taking over technologies such as artificial intelligence (AI) and machine learning (ML), it is emerging as a critical game changer and offering data leaders a new perspective to grow. Data leaders are now transitioning to a data-driven business model, where data is assisting them in making decisions based on what we know to be true rather than solely sticking to instinct. With the upcoming recession, businesses need to be prepared for the worst. As Thoman Redman said, “Where there is data smoke, there is business fire.” This article is written by a member of the AIM Leaders Council. AIM Leaders Council is an invitation-only forum of senior executives in the Data Science and Analytics industry. To check if you are eligible for a membership, please fill the form here.","excerpt":"In a world driven by data, industry leaders are successfully extracting actionable insights through precise data management and analytics. This enables them to innovate quickly, develop better frameworks, and govern modifications more efficiently.","categories":["AI Features"],"tags":[],"author_name":"Biswanath Choudhary","publish_date":"2022-12-15T14:00:00","publication_year":"2022","word_count":1386,"keywords":["data science","machine learning","artificial intelligence","AWS","AI","ML","RAG","Aim","analytics","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","data science","analytics","Aim","RAG","AWS","R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/business-outlook-for-data-leaders-in-2023\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10042316,"title":"Creating A Paragraph Auto Generator Using GPT2 Transformers","content":"Automation is the process of making a system operate automatically without human intervention. As the days pass by and life gets busier each day, automation and automated systems intrigue the commune more. By applying automation, unsafe and repetitive tasks could be made self-sufficient, saving two of the most quintessential and precious things in a human’s life, Time and Money. With many tasks these days being labour-intensive and time-consuming, the creation of automated systems has improved efficiency and led to greater quality control. Although Automation may or may not be completely based on Artificial Intelligence, with the rise of automation and artificial intelligence simultaneously in the last decade, the use of automation collaborating with artificial intelligence might just be the next big thing to ponder upon. One of the most breakthrough discoveries in recent times for automation using artificial intelligence is AI Natural Language Generation. What is Natural Language Generation? Natural Language Generation, also known as NLG, uses artificial intelligence to produce written or spoken text content. It is a subsidiary of artificial intelligence and is a process that automatically transforms input data into plain-English content. The fascinating thing about NLG is that the technology can help tell a story using human-like creativity and intelligence, writing long sentences and paragraphs for you. Some of the uses of NLG are to generate product or service descriptions, content curation, creating portfolio summaries, or being used in customer communications through certain implementations in chatbots. Natural-language generation can be a bit complicated and require layers of language knowledge to work. These days, NLG is being integrated into tools to help with content strategy quickly, hence increasing productivity. About Hugging Face Hugging Face is an NLP focused startup that shares a large open-source community and provides an open-source library for Natural Language Processing. Their core mode of operation for natural language processing revolves around the use of Transformers. This python based library exposes an API to use many well-known architectures that help obtain the state of the art results for various NLP tasks like text classification, information extraction, question answering, and text generation. All the architectures provided come with a set of pre-trained weights utilizing deep learning that help with ease of operation for such tasks. These transformer models come in different shape and size architectures and have their ways of accepting input data tokenization. A tokenizer takes an input word and encodes the word into a number, thus allowing faster processing. Getting Started with Creating a Paragraph Auto Generator This article will try to implement a natural language generator that generates paragraphs from a single line of input text. For that, we will first set up all our dependencies using Hugging Face transformers for Natural Language Processing, then load our GPT2 model. This pre-trained model generates coherent paragraphs of text, encodes our input, and decodes our output to generate a paragraph. So let’s get started with it! The following code implementation is inspired by the official implementation, whose video tutorial you can find here. Installing our Libraries The first step would be to install our dependent libraries for this model. To do this, we will first install the Hugging Face Transformers. You can install Transformers by using the following command : !pip install transformers #install the library from hugging face Next, we will import our GPT2 Model and Tokenizer, collaborating with Tensorflow. import tensorflow as tf from transformers import GPT2LMHeadModel, GPT2Tokenizer #importing the main model and tokenizer We will first encode the input sentence into tokens using the tokenizer, then generate a new sequence of tokens from the GPT-2 model and then decode the generated tokens into a sequence of words using the tokenizer again, which will provide us with our output. Loading our Model Create a new variable for the tokenizer and passing it through the GPT parameter. tokenizer = GPT2Tokenizer.from_pretrained(\"gpt2-large\")#using the large parameter from GPT to generate larger texts Instantiate the pre-trained model and padding with the tokenizer. model=GPT2LMHeadModel.from_pretrained(\"gpt2large\",pad_token_id=tokenizer.eos_token_id) Testing the model by tokenizing our First sentence Now that the model has been created, we will test it by providing our first input sentence to tokenize. sentence = 'You will always succeed in Life' #input sentence Encode it into a sequence of numbers and return them as PyTorch tensors. input_ids = tokenizer.encode(sentence, return_tensors='pt')#using pt to return as pytorch tensors Checking current progress input_ids # checking the tesors returned We will get the following output as number representation, tensor([[1639,  481, 1464, 6758,  287, 5155]]) Decoding the text and Generating the Output Creating a new variable called output to decode and setting our hyperparameters, output = model.generate(input_ids, max_length=50, num_beams=5, no_repeat_ngram_size=2, early_stopping=True) With this line, we have called the input and set the maximum length of the paragraph to be generated as 50 words. We are also using a beam search technique to find the most appropriate word to be generated from the input sentence. We have also set no-repeat ngram as 2, which will prevent our model from repeating similar words more than twice and early stopping as true so that when the model does not find appropriate words, it stops the generation process. Printing our results : print(tokenizer.decode(output[0], skip_special_tokens=True))#printing results We got the following output: You will always succeed in Life, but you will never be successful in Death.\" \"I am not afraid of death, because I know that I am going to be with you when you die. I will be waiting for you, and I. Cross validating our Model We can also do the same and tune our hyperparameters to generate larger paragraphs with a new sentence. Beware this may take a longer time to generate output. sentence = 'Artificial intelligence is the key' input_ids = tokenizer.encode(sentence, return_tensors='pt') output = model.generate(input_ids, max_length=500, num_beams=5, no_repeat_ngram_size=2, early_stopping=True) #setting length as 500 to generate larger output text print(tokenizer.decode(output[0], skip_special_tokens=True)) We will get the following as output : Artificial intelligence is the key to unlocking the mysteries of the universe, but it's also the source of a lot of our problems. In a new paper published in the journal Science Advances, a team of researchers from the University of California, Berkeley, and the National Institute of Standards and Technology (NIST) in Gaithersburg, Maryland, describes a way to create an artificial intelligence (AI) system that can learn from its mistakes and improve its performance over time. The system, which they call a \"neural network,\" is capable of learning to recognize patterns in images, recognize objects in a video, or even learn how to play a musical instrument. In the paper, the researchers describe how they created the neural network and how it can be used to train an AI system to perform a variety of tasks, such as recognizing objects and playing musical instruments. Neural networks, also known as deep neural networks or deep learning, are a type of machine learning algorithm that is based on the idea that a network of neurons is like a computer's processor. Each neuron is connected to a number of other neurons to form a larger network. When a neuron receives an input, it sends a signal to the next neuron, who in turn sends an output to another neuron. This process continues until all the neurons have received the input and have processed it. As a result, each neuron has its own unique set of inputs and outputs, making it possible for the network to learn and adapt to changes in its environment. Neural networks have been used for decades to solve a wide range of problems,including image recognition, speech recognition and natural language processing. However, they have also been criticized for their poor performance when it comes to learning from their own mistakes. For example, in 2013, researchers at Google's DeepMind AI research lab published a paper in Nature that showed that they were unable to improve the performance of their network when they made a series of mistakes while training it on images of human faces. They also found that the system was not able to distinguish between a human face and a dog face, even though the images were similar in terms of size and shape. These problems have led some researchers to argue that neural nets are not as effective as they are made out to be. But the new study suggests that this is not necessarily the case. \"We show that it is possible to build a neural net that learns from mistakes,\" said lead author and UC Berkeley professor of electrical engineering and computer. We can clearly notice the difference through hyperparameter tuning this time! You can save it as a text file using the following lines of code : text = tokenizer.decode(output[0],skip_special_tokens = True) with open('AIBLOG.txt','w') as f: f.write(text) EndNotes We have now learned how to create a model to generate long lines of text from a single sentence utilizing AI and Hugging Face Library by performing the following steps. You can tune the hyperparameters further to make the model more intelligent to provide better text content. The full Colab file for the following can be accessed from here. Happy Learning! References About Hugging Face LibraryAbout GPT2Future of Content Creation","excerpt":"Automation is the process of making a system operate automatically without human intervention. As the days pass by and life gets busier each day, automation and automated systems intrigue the commune more. By applying automation, unsafe and repetitive tasks could be made self-sufficient, saving two of the most quintessential and precious things in a human’s […]","categories":["AI Trends"],"tags":["AI Tool","automated data science solutions","Automation","Generative Pre-Trained Transformer","HuggingFace","natural language generation","Natural Language Processing","NLG","NLP","Transformers"],"author_name":"Victor Dey","publish_date":"2021-06-25T12:00:00","publication_year":"2021","word_count":1517,"keywords":["Hugging Face","artificial intelligence","natural language generation","machine learning","AI","neural network","Natural Language Processing","TensorFlow","Automation","PyTorch","Transformers","automated data science solutions","NLP","deep learning","Generative Pre-Trained Transformer","AI Tool","NLG","HuggingFace"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","deep learning","neural network","NLP","TensorFlow","PyTorch","Hugging Face","Transformers"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/creating-a-paragraph-auto-generator-using-gpt2-transformers\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10162346,"title":"The Censorship Dilemma Behind DeepSeek’s AGI Mission","content":"Much has been written about the AI model that jolted the stock market. But, DeepSeek CEO Liang Wenfeng rarely speaks in public, making each of his interviews and statements highly anticipated and closely scrutinised. As per reports, he appeared for just two interviews in 2023 and 2024, in which he revealed his modus operandi to achieve artificial general intelligence (AGI). Wenfeng, born in the 1980s in the Chinese province of Guangdong, graduated from Zhejiang University with a degree in electronic information engineering. In 2015, he co-founded High-Flyer, a hedge fund which managed $10 billion by 2019. The interviews highlight that, unlike many Chinese AI firms prioritising commercialisation, DeepSeek is dedicated to fundamental AGI research. “It could be two, five, or 10 years away, but it will definitely happen in our lifetime,” he said, focusing on three main directions: mathematics and code, multimodality, and natural language itself. Elaborating on DeepSeek’s approach to talent, Wenfeng clarified that there are no “wizards”. According to him, the company operates with a bottom-up structure, recruiting young local talent from local Chinese universities. While well-funded, DeepSeek’s main hurdle lies in securing high-end chips restricted by US export controls. “We don’t have short-term fundraising plans. Our problem has never been funding; it’s the embargo on high-end chips,” he said. According to The Wall Street Journal, Wenfeng recently met with Chinese premier Li Qiang to discuss the difficulties Chinese companies face as a result of US restrictions on advanced chip exports. On the open-source front, Wenfeng said, “In the face of disruptive technology, a closed-source moat is temporary.” He also noted that while people often speak of a one or two-year gap between Chinese and American AI, the true divide is between originality and imitation. As DeepSeek rattles global markets, it also raises serious concerns about AI safety, driven by its open-source design and strong links to the Chinese Communist Party. This is especially true, considering top leaders predict that ASI is on an accelerated timeline from before. Is All Really Well on DeepSeek? Censorship of sensitive topics is a major concern with DeepSeek. The model avoids answering questions related to issues such as Uyghur human rights abuses, Taiwan’s political status, the 1989 Tiananmen Square incident, criticism of Chinese supreme leader Xi Jinping, censorship in China, and questions about Arunachal Pradesh and Kashmir’s sovereignty, among others. All is well on DeepSeek. pic.twitter.com\/wS6ecikM2x— nxthompson (@nxthompson) January 27, 2025 Instead, it deflects these inquiries with responses like: “Sorry, I’m not sure how to approach this type of question yet.” People on X have DeepSeek’ compared the censorship on Anthropic’s Claude and OpenAI’s ChatGPT. According to some reports, China’s regulatory body, the Cyberspace Administration of China (CAC), imposes strict testing requirements for AI models, including testing up to 70,000 questions to ensure politically safe answers. This slows AI development and limits the randomness and creativity typical of generative AI. The Chinese government rigorously reviews large language models (LLMs) to ensure they adhere to “core socialist values”. Companies such as ByteDance, Alibaba, Moonshot, and 01.AI are obligated to undergo these compulsory audits conducted by the CAC. Moonshot’s chatbot, Kimi, rejects most questions about Jinping. Similarly, ByteDance’s LLM ranks highest in safety compliance tests, showcasing its alignment with Beijing’s messaging. Many argue that because it’s open-source, it can be fine-tuned to suit specific needs or values. But this also hint at a deeper issue; one of censorship that extends beyond simple fine-tuning. Former OpenAI researcher Miles Brundage pointed out that while this is an immediate benefit, it could lead to stricter rules in the future. DeepSeek and governments might focus more on improving AI safeguards, making it harder to release new models. Governments could also push for AI tracking on devices, where smaller AIs monitor the use of larger ones and send reports to central systems. This could also affect how non-Chinese AI models are used in China.","excerpt":"Can CEO Liang Wenfeng fix it?","categories":["Global Tech"],"tags":["AI (Artificial Intelligence)","DeepSeek"],"author_name":"Aditi Suresh","publish_date":"2025-01-28T19:14:49","publication_year":"2025","word_count":645,"keywords":["Anthropic","ChatGPT","Go","OpenAI","AI","GPT","DeepSeek","ViT","generative AI","AI safety","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","generative AI","ChatGPT","OpenAI","Anthropic","R","Go","GPT","ViT","AI safety"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/the-censorship-dilemma-behind-deepseeks-agi-mission\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10003656,"title":"Python 3.9 Released; Check The New Features","content":"It was announced last year that Python will be making significant changes and will be adding new features with the release of their 3.9 version. Back then, it was speculated that the new version would flaunt new features in security, core and builtins, libraries, documentation, and more. Yesterday, Python announced new features as part of their 3.9 in beta version((3.9.0b3). Let’s take a look at few highlights: Merging Gets Better The current ways to merge two dicts in Python have several disadvantages: dict.update d1.update(d2) modifies d1 in-place. e = d1.copy(); e.update(d2) is not an expression and needs a temporary variable. {**d1, **d2} Dict unpacking looks ugly and is not easily discoverable. Few people would be able to guess what it means the first time they see it, or think of it as the “obvious way” to merge two dicts. “…if you were to ask a typical Python user how to combine two dicts into a new one, I doubt many people would think of {**d1, **d2}. I know I myself had forgotten about it when this thread started!”Guido Van Rossum, Python creator Now in 3.9, Merge (|) and update (|=) operators have been added to the built-in dict class. Here’s how it works: >>> d = {'spam': 1, 'eggs': 2, 'cheese': 3} >>> e = {'cheese': 'cheddar', 'aardvark': 'Ethel'} >>> d | e {'spam': 1, 'eggs': 2, 'cheese': 'cheddar', 'aardvark': 'Ethel'} >>> e | d {'aardvark': 'Ethel', 'spam': 1, 'eggs': 2, 'cheese': 3} The augmented assignment version operates in-place: >>> d |= e >>> d {'spam': 1, 'eggs': 2, 'cheese': 'cheddar', 'aardvark': 'Ethel'} New String Methods str.removeprefix(prefix) and str.removesuffix(suffix) have been added to easily remove an unneeded prefix or a suffix from a string. Current:if test_func_name.startswith(\"test_\"): print(test_func_name[5:]) else: print(test_func_name) Improved:print(test_func_name.removeprefix(\"test_\")) Built-in Gets Generic Static typing was built incrementally on top of the existing Python runtime and constrained by existing syntax and runtime behavior. This led to the existence of a duplicated collection. For example typing.List and the built-in list. The new release enables support for the generics syntax in all standard collections currently available in the typing module. def greet_all(names: list[str]) -> None: for name in names: print(\"Hello\", name) Brand New Modules zoneinfo The zoneinfo module brings support for the IANA time zone database to the standard library. It adds zoneinfo.ZoneInfo, a concrete datetime.tzinfo implementation backed by the system’s time zone data. >>> from zoneinfo import ZoneInfo >>> from datetime import datetime, timedelta graphlib Using the graphlib.TopologicalSorter class of the graphlib, one can perform topological sorting of graphs Dropped Items The new Python version ditched a bunch of methods, functions and modules for good. Here are a few: tostring() and fromstring() have been removed C function PyImport_Cleanup() has been removed._dummy_thread and dummy_threading modules have been removed. Find the complete list of updates here.","excerpt":"It was announced last year that Python will be making significant changes and will be adding new features with the release of their 3.9 version. Back then, it was speculated that the new version would flaunt new features in security, core and builtins, libraries, documentation, and more.  Yesterday, Python announced new features as part of […]","categories":["AI News"],"tags":["Python"],"author_name":"Ram Sagar","publish_date":"2020-07-30T12:35:24","publication_year":"2020","word_count":463,"keywords":["Go","ELT","programming_languages:R","AI","programming_languages:Go","Python","programming_languages:Python","R"],"extracted_tech_keywords":["AI","Python","R","Go","ELT","programming_languages:Python","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/python-3-9-release-feature\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10065859,"title":"Webinar Alert! Attend this power session to jumpstart your career in data science","content":"As per report, the big data and business analytics market size is expected to reach USD 684.12 billion by 2030. Hence, the need for data and analytics professionals is on the rise. For someone looking to build a data science career, finding the right data science courses can be a daunting task. You need to have a solid grip on theoretical concepts as well as good working knowledge of AI\/ML tools and programming languages like R, Python etc., to excel in data science. The candidates should also know how to apply the concepts to solve business problems. To that end, Praxis Business School is organising a webinar on “how to jumpstart your career in data science” on May 7, 2022, from 11 am to 12 pm. The webinar will cover: Introduction to the data science & analytics marketThe emergence of data science as a fieldCore skills to build a career in data scienceDifferent job roles in the data science & analytics industry REGISTER NOW Who should attend: Students looking to build careers in AI, data science and analyticsData scientistsAnalytics professionalsBusiness analysts, data analystsData science enthusiastsWorking professionals looking to transition to an analytics role Session speakers Professor Charanpreet Singh, Founder and Director at Praxis Business School Foundation He started Praxis Business School’s data science program in 2011 and comes with over 30 years of experience. He has worked with British Oxygen, Tata Steel, PwC and HP. His key areas of interest include education, data science and communication. Bharath Kumar Bolla – Senior Data Scientist, Salesforce He has over a decade of experience in analytics and is currently working as a senior data scientist at Salesforce, Hyderabad. In addition, he has a PG Diploma in data science from Praxis Business School, an MS in data science from the University of Arizona and a masters in life sciences from the Mississippi State University. He has worked as a research data scientist at the University of Georgia, Emory University, and Eurofins LLC. At Happiest Minds, he worked on AI-based digital marketing products and NLP-based solutions in the education space. Recently, Bharath was also recognised as the 40 under 40 data scientist by Analytics India Magazine. REGISTER NOW","excerpt":"For someone looking to build a data science career, finding the right data science courses can be a daunting task. Praxis Business School is organising a webinar on “how to jumpstart your career in data science”on May 7, 2022.","categories":["Deep Tech"],"tags":["Courses","Data Science","Data Science Career","Data science career webinar","data science webinar","Praxis Business School","Tech Webinars You Must Attend","webinar","Webinar for data scientists"],"author_name":"Sreejani Bhattacharyya","publish_date":"2022-04-28T16:00:00","publication_year":"2022","word_count":363,"keywords":["Tech Webinars You Must Attend","Webinar for data scientists","Git","R","data science","data science webinar","NLP","analytics","Data Science","Go","Data science career webinar","AI","ML","Data Science Career","Praxis Business School","webinar","big data","Python","Courses"],"extracted_tech_keywords":["AI","ML","NLP","data science","analytics","Python","R","Go","Git","big data"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/webinar-alert-attend-this-power-session-to-jumpstart-your-career-in-data-science\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10143371,"title":"AI Agents Startup RapidCanvas Secures $16M to Address Technical Talent Shortage","content":"Austin-based RapidCanvas has announced a $16 million Series A funding round to advance its AI agents, which can automate up to 75% of complex tasks traditionally handled by data scientists and engineers. The round, led by Peak XV Partners with participation from Titanium Ventures, Accel, and Valley Capital Partners, brings the company’s total funding to $23.5 million since its inception in 2021. RapidCanvas aims to address the global shortage of technical talent by offering a “Service-as-Software” model that integrates AI agents with human expertise. These AI agents, powered by LLMs, perform tasks such as data processing, pattern recognition, and decision-making at high speed and scale. Addressing Technical Talent Shortage The hybrid approach allows businesses to reduce reliance on technical talent, requiring only one or two expert engineers for tasks that typically need larger teams. The platform is particularly effective in coding-related tasks, where AI agents automate up to 70% of routine operations, freeing experts to focus on high-value work. RapidCanvas estimates that 70% of the nearly $1 trillion in global salaries paid to software engineers and data scientists could be optimised with AI, enabling enterprises to unlock faster ROI and business growth. The company’s co-founders, Rahul Pangam and Uttam Phalnikar, come from an experience of working in AI-powered business transformation space, and have previously built Simility, a risk management platform acquired by PayPal. The company’s Reliable AI framework ensures validated, secure, and explainable outputs, enabling businesses to implement AI solutions faster and at a fraction of the cost of traditional methods. Unlike traditional Software-as-a-Service (SaaS) tools that improve employee productivity, RapidCanvas links software costs directly to business outcomes, marking a shift in automation’s potential. The startup claims that early customers have reported consistent value across diverse use cases, showcasing the platform’s scalability and efficiency. With this funding, RapidCanvas is set to accelerate AI adoption across industries, helping enterprises overcome the technical talent gap and achieve tailored AI solutions that drive growth and profitability. AI Agents is Growing Observing the current trend, AI agents seem to be a hot trend with a number of big tech companies and startups building them. Aampe, an AI startup, headquartered in North Carolina, is building agentic infrastructure to provide personalised experiences for product and marketing teams. The startup raised $18 million in Series A financing, and will reveal that 100 million AI agents are being used in their consumer apps. The global AI agents market is expected to grow from $5.29 billion in 2024 to $ 216.8 billion by 2035.","excerpt":"RapidCanvas estimates that 70% of the nearly $1 trillion in global salaries paid to software engineers and data scientists could be optimised with AI.","categories":["AI News"],"tags":["Aampe","Accel","AI agent","RapidCanvas","Service as a software"],"author_name":"Vandana Nair","publish_date":"2024-12-12T10:13:00","publication_year":"2024","word_count":415,"keywords":["API","funding","TPU","Aampe","AI","Accel","Scala","Service as a software","automation","AI agent","Aim","ViT","R","RapidCanvas","startup"],"extracted_tech_keywords":["AI","Aim","TPU","R","Scala","API","ViT","automation","startup","funding"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/ai-agents-startup-rapidcanvas-secures-16m-to-address-technical-talent-shortage\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":56136,"title":"Reliance Jio To Finally Launch A Scalable Connected Car Solution Which It Has Promised For Years","content":"Reliance Jio Infocomm is finally rolling out its connected car technology at the 15th edition of Auto Expo that begins at India Expo Mart in Greater Noida, Uttar Pradesh. The telecom giant is aiming to become India’s first major telecom company to expand its market to connected cars, a highly competitive yet nascent market across the globe. Jio has been discussing the importance of connected vehicles, particularly in the context of IoT solutions for the last three years. According to reports, Jio has been developing its connected car technologies for more than a year at its campus in Navi Mumbai. Jio’s fast growth in 4G services and soon to be launched 5G will corroborate with its connected vehicle solutions. Experts have vowed that 5G, combined with more powerful computer chips, will bring in the perfect time for IoT and edge computing in the near future. For telecom players, the connected devices market seems like a natural extension. Mukesh Ambani also called IoT a key growth engine for the company. On the other hand, the slowing car sales in India could be a cause for concern for the connected car solutions market targeted at end consumers. IoT technology has been one of the tech areas most telecom companies have been trying to capture since it has applications in transportation, manufacturing, home automation, etc. According to reports, Jio will showcase a vehicle tracking system for transport service and logistics players, and will showcase the role it will play for connectivity, network and technology for connected cars. Also, the company has been working on its enterprise solutions, cloud network, and hardware solutions, all of which will help in creating a powerful connected vehicle solution needed for the scale in India. Will This Could Be An Extension Of Jio Car Connect? According to Jio, Car Connect is a combination of a hardware device and a mobile app aimed to connect 90% of the cars on the Indian roads by the year 2020. The solution was powered by OBD (Onboard Diagnostic Device) with a Jio SIM inside, which will be connected to the ODB port of the car. The OBD would be giving users access to the entire ecosystem of Jio Apps – Jio Music and Jio Cinema .etc. Apart from ODB devices, Jio at the Auto Expo showcased hardware solutions based on AIS 140 standard and include door and fuel sensors and a panic button for drivers and passenger safety. The Jio Car Connect mobile app could also display users’ vital car information like oil stats, water levels, fuel info, battery status .etc. The OBD device, paired with a Jio SIM card, will also give you the exact location of the car, which can be a useful feature in case of a theft or an accident with geo-fencing. Reliance Jio Infocomm had also signed a contract with US-based AirWire Technologies (AirWire) to manufacture the Jio Car Connect device to customers for narrow-band IoT network. Could Connected Car Services Be The Next Big Thing For Telecom? Connected cars with embedded eSIM technologies making the cars function like smartphones, and therefore it requires operating systems, apps, infotainment, and cybersecurity. In addition to sensors and LIDAR technologies, car systems can also perform real-time analytics and telematics for predictive maintenance, route planning, road safety and much more. All of this would need not just fast internet speed but also on-the-go computing at the edge services. The largest players in this segment are automated car manufacturers like Tesla, traditional car manufacturers who are rising to the occasion like Ford, big tech companies like Google and Amazon, and telecom companies like AT&T and Jio. Other Industry Players Working Towards Connected Device Solutions Other telecom companies in India such as Vodafone, Idea, and Bharti Airtel have also been active in connected car technologies given they have witnessed drastically reduced revenue from core telecom services like voice and data consumption. As part of its connected market strategy, Airtel had announced that the auto sector is a key focus area for which the company is building platforms through partnerships for connected cars. Airtel also owns a separate, independent IoT vertical which has partnered MG Motors for its Hector car wherein it introduced an IoT solution with features like vehicle telematics, navigation, and infotainment services, along with an embedded IoT SIM. Also Read: WILL OPEN SOURCE PLATFORMS BRIDGE THE INNOVATION GAP ACROSS SMART CAR SOLUTIONS?","excerpt":"Reliance Jio Infocomm is finally rolling out its connected car technology at the 15th edition of Auto Expo that begins at India Expo Mart in Greater Noida, Uttar Pradesh. The telecom giant is aiming to become India’s first major telecom company to expand its market to connected cars, a highly competitive yet nascent market across […]","categories":["AI Features"],"tags":["Jio","Jio AI Cloud","reliance jio"],"author_name":"Vishal Chawla","publish_date":"2020-02-06T16:03:00","publication_year":"2020","word_count":731,"keywords":["Jio","Go","AI","innovation","automation","Aim","Jio AI Cloud","ViT","analytics","reliance jio","real-time analytics","edge computing","R"],"extracted_tech_keywords":["AI","analytics","Aim","edge computing","R","Go","real-time analytics","ViT","automation","innovation"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/reliance-jio-to-finally-launch-a-scalable-connected-car-solution-which-it-has-promised-for-years\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10091208,"title":"How Brain Drain Shaped AI Innovations of the Current World","content":"We find ourselves at an inflection point in the development of a novel AI technology. The success of AI models such as ChatGPT has stirred up a plethora of emotions overhauling the human conditions—excitement, anxiety, fear, curiosity, and anticipation. These emotions are shared not only by normies, but also by researchers actively building these powerful machines. Along these lines, a petition was signed by prominent AI researchers calling for a six-month halt on the development of models larger than GPT-4. Naturally, this proposal was met with vehement opposition. Andrew Ng and Yann LeCun recently came together to discuss why they believed that legislating a pause was a wrong idea. LeCun, who is a Chief AI Scientist at MetaAI, made a clear distinction between AI research and the products that emerge from it. While he acknowledges the need to regulate the deployment of AI products, LeCun argues that calling for a halt in research would only impede progress and would not serve any meaningful purpose. “My first reaction to [the letter] is that calling for a delay in research and development smacks me of a new wave of obscurantism,” said LeCun. “Why slow down the progress of knowledge and science? Then, there is the question of products. . . I’m all for regulating products that get in the hands of people. I don’t see the point of regulating research and development. I don’t think that serves any purpose other than reducing the knowledge that we could use to actually make technology better, safer,” he said during the conversation. But, the question remains whether there is such a sharp distinction between research and industry when it comes to AI research? To understand this, it is important to know where the research is happening in the first place. Industry outpaces academia The Stanford University AI Index Report 2023 indicates that the for-profit AI companies are racing ahead of academia. “Until 2014, most significant machine learning models were released by academia. Since then, AI companies have taken over. In 2022, there were 32 significant industry-produced machine learning models compared to just three produced by academia,” reads the report. Skilled computer scientists around the globe are being lured away from academia by alluring employment offers from the private sector. In addition to this, the enticing perks, such as access to diverse data sets, abundant computing resources, and the ability to make a significant impact on millions of individuals through commercial products and you see why AI research is happening in corporations more than universities. This trend, argues a blog, is resulting in a significant brain drain that has already had an impact on research and education. Since a small number of corporations have recruited the majority of the top AI researchers, their expertise and knowledge are not being shared with society at large. This is problematic because the concentration of innovation in few companies is what prevents any opportunity to mitigate the substantial disruption and negative consequences that AI could bring. Profit-making machines Evidently, when it comes to research in corporations, it is no charity work. The end goal has always been to either launch new products or integrate into existing products. DeepMind, for instance, earned its first profits drawing revenue from research and development carried out for other companies under the Alphabet umbrella, including Google, YouTube and X, the moonshot division. Notably, this was a few months after DeepMind was denied the bid for having the same legal structure as a non-profit by Google. Similar was the case with OpenAI, which shifted from non-profit to ‘capped-profit’ to attract capital—a few years ago—alongside giving exclusive licence of GPT-3\/4 to Microsoft for commercial applications and use cases. Remaining non-profits will follow suit. AI research is expensive, and only large corporations have that kind of capital to keep pouring bucks for training and compute. In a four-year old video, Sam Altman, the co-founder of OpenAI, emphasised the importance of product-oriented AI research. Stressing on getting to market as quickly as possible, he says, “You want people to have a bias towards action. Startups, especially in their early days, win by moving in very quickly. Initially, you never get as much data as you would like and you never have as much time to deliberate as you would like. And yet, you need people who will act with much less data than they’d like to have and much less certainty. If they act and it doesn’t work, they adapt really quickly”. Thus, the idea that we should keep pushing the throttle for more research while letting the product people handle the regulation aspect seems half-baked. The current state of AI is such that research is heavily inclined towards finding a product market fit. A force of resistance In a recent New York Times’ podcast, Google CEO Sundar Pichai emphasised that OpenAI’s approach to release the product sooner in order to give society a chance to understand and adapt is a “reasonable point” to take. It is clear that big tech wants to steer AI research in a particular direction, and the ones making a case for slowing the pace of progress are fighting this “voice of power”. In this light, LeCun’s own position of being Meta‘s Chief AI Scientist needs to be taken into consideration to understand what is informing his position in the current AI discourse. One of the signees of the open letter Yoshua Bengio, considered to be the Godfather of AI, explained that a year ago, he perhaps wouldn’t have signed the letter but with the arrival of ChatGPT, there is witnessed a shift in the attitudes of companies for whom the “challenge of commercial competition has increased tenfold”. “There is a real risk that they will rush into developing these giant AI systems, leaving behind good habits of transparency and open science they have developed over the past decade of AI research.”","excerpt":"Are the lines so sharp between research and industry when it comes to AI?","categories":["AI Features"],"tags":["AI Models","AI Research","ChatGPT","DeepMind","Google","GPT-4","Yann LeCun","Yoshua Bengio"],"author_name":"Ayush Jain","publish_date":"2023-04-11T18:30:00","publication_year":"2023","word_count":978,"keywords":["AI Models","ChatGPT","Yann LeCun","machine learning","Go","OpenAI","AI","API","innovation","AI Research","GPT-4","GPT","Yoshua Bengio","Google","disruption","R","DeepMind"],"extracted_tech_keywords":["AI","machine learning","ChatGPT","OpenAI","R","Go","API","GPT","innovation","disruption"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-brain-drain-shaped-ai-innovations-of-the-current-world\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10043768,"title":"JAX Vs TensorFlow Vs PyTorch: A Comparative Analysis","content":"Deep learning owes a lot of its success to automatic differentiation. Popular libraries such as TensorFlow and PyTorch keep track of gradients over neural network parameters during training with both comprising high-level APIs for implementing the commonly used neural network functionality for deep learning. JAX is NumPy on the CPU, GPU, and TPU, with great automatic differentiation for high-performance machine learning research. Along with a Deep Learning framework, JAX has created a super polished linear algebra library with automatic differentiation and XLA support. About JAX JAX is a new machine learning library from Google designed for high-performance numerical computing. The Autograd library has the ability to differentiate through every native python and NumPy code. JAX is defined as “Composable transformations of Python+NumPy programs: differentiate, vectorize, JIT to GPU\/TPU, and more”. The library utilises the grad function transformation to convert a function into a function that returns the original function’s gradient. Jax also offers a function transformation JIT for just-in-time compilation of existing functions and vmap and pmap for vectorization and parallelization, respectively. The move from PyTorch or Tensorflow 2 to JAX is nothing short of tectonic. PyTorch builds up a graph during the forward pass, and gradients during the backward pass. JAX, on the other hand, allows the user to express their computation as a Python function, and by transforming it with grad() gives the gradient function that can be evaluated like the computation function—but instead of the output, it gives the gradient of the output for the first parameter that the function took as input. JAX vs Tensorflow vs Pytorch While TensorFlow and Pytorch have compiled execution modes, these modes were added later on and thus have left their scars. For instance, TensorFlow’s eager mode is not 100% compatible with the graphic mode allowing for a bad developer experience. Pytorch has a bad history of being forced to use less intuitive tensor formats since they were performed in eager mode. JAX arrives with both of these modes- focusing on eager for debugging and on JIT to perform heavy computations. But the clean nature of these modes allow for mixing and matching whenever needed. PyTorch and Tensorflow are deep learning libraries consisting of high-level APIs for modern methods in deep learning. In comparison, JAX is a more functionally-minded library for arbitrary differentiable programming. DZone conducted a mini-experiment to study how JAX stacks up against other libraries. A single hidden layer MLP along with a training loop to “fit” a classification problem of random noise – was implemented in JAX, Autograd, Tensorflow 2.0 and PyTorch. This baseline was implemented to compare the performance efficiency of each library. While JAX uses just-in-time compilation for library calls, the jit function transformation can be used as a decorator for custom Python functions. For instance, here is a snippet: Java # use jit as a decorator on a function definition @jit def get_loss(x, w, y_tgts): y_pred = forward(x, w) return ce_loss(y_tgts, y_pred) # use jit as a function for transforming an already defined function into a just-in-time compiled function. get_grad = grad(get_loss, argnums=(1)) jit_grad = jit(get_grad) Jax runtimes The experimenters implemented a simple multi-layer perceptron in each of the libraries, consisting of a sequence of weighted connections determining numerical values and equivalent to the matrix multiplication of input tensors and weight matrices. The results showed that JAX dominated the experiment. JAX has a faster CPU execution time than any other library and the shortest execution time for implementations using only matrix multiplication. The experiment also found that while JAX dominates over other libraries with matmul, PyTorch leads with Linear Layers. PyTorch had a quick execution time while running on the GPU – PyTorch and Linear layers took 9.9 seconds with a batch size of 16,384, which corresponds with JAX running with JIT on a batch size of 1024. PyTorch was the fastest, followed by JAX and TensorFlow when taking advantage of higher-level neural network APIs. For implementing fully connected neural layers, PyTorch’s execution speed was more effective than TensorFlow. On the other hand, JAX offered impressive speed-ups of an order of magnitude or more over the comparable Autograd library. JAX was also the fastest when MLP implementation was limited to matrix multiplication operations.","excerpt":"JAX is a Python library designed for high-performance numerical computing.","categories":["Deep Tech"],"tags":["what is tensorflow"],"author_name":"Avi Gopani","publish_date":"2021-07-16T15:00:00","publication_year":"2021","word_count":698,"keywords":["NumPy","machine learning","AI","neural network","what is tensorflow","ML","PyTorch","deep learning","JAX","Tecton","TensorFlow"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","neural network","Tecton","TensorFlow","PyTorch","JAX","NumPy"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/jax-vs-tensorflow-vs-pytorch-a-comparative-analysis\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10103218,"title":"Amazon’s PartyRock Jams Past OpenAI","content":"Recently, AWS announced a launch of a new invention called PartyRock, an approachable Amazon Bedrock Playground for developers to build applications without any hassle, enabling anybody to create a generative AI program – just like OpenAI’s GPT Builder. Providing a creative space for everyone to express themselves, the application is completely ahead regardless of the coding expertise, along with allowing the creators to build their own application according to their own preferences and needs. Just like OpenAI is making everyone an app developer with GPT Builder, Amazon with PartyRock is effortlessly seizing the creative and user-friendly realm, providing a seamless environment for building and exploring generative AI applications in just a few simple steps. In simple terms – letting everyone build AI apps. From Bedrock to PartyRock Similar to OpenAI’s GPTs, PartyRock allows users to create customised LLM based models with personal information. For example, users can give superhero names for their dog or themselves to generate an application that gives them information about certain places or even ratings on food, and you can even use it to generate game content, such as levels, characters, or storylines. The app builder is powered by Anthropic’s Claude-2, where users can give prompts to start generating their desired app. The user interface of Amazon’s app building platform is pretty minimal and attractive. Whether generating text-based responses or connecting prompts, the platform encourages users to explore and enhance their knowledge of generative AI capabilities. It also has come up with reliable and easy-to-share options, recognising the importance of the community. PartyRock is helping users enable an API-based interface for accessing foundation models (FMs) from Amazon and other top AI providers, such as AI21 Labs, Anthropic Cohere, Meta, Stability AI, and Amazon with a single API. With this interface, users will have a strong base to test out different generative AI approaches. Along with that, it allows it to be more accessible to anyone by providing a platform that makes working with foundation models less complicated. Moreover, it has been made easier for users to share the apps they have created with their friends and the community. By providing a straightforward means to publish links on social media platforms using #partyrockplayground, AWS aims to foster a vibrant community of creators inspiring each other through their generative AI creations. What’s the rock in PartyRock? The best part about Amazon’s PartyRock is that it offers a limited-time free trial offer without the requirement of a credit card, unlike OpenAI’s GPT Builder, which requires a ChatGPT Plus subscription to start, making it widely accessible for a larger audience. It seems like Amazon is actually making everyone an app developer. If you’ve been wanting to experiment with #generativeAI, check out PartyRock powered by Amazon Bedrock. https:\/\/t.co\/Cz8An1L8AuWith PartyRock, anyone can learn about prompt engineering\/writing and build apps in just a few clicks. I built one to find the best bike trails in the… pic.twitter.com\/Cp7eOmmW9x— Adam Selipsky (@aselipsky) November 16, 2023 “With PartyRock’s introduction, the conventional view of generative AI development as a challenging and specialised field is changing,” says Amazon. But it is just one step after OpenAI announced the same with GPT Builder. Apart from the free trial, what sets PartyRock apart is yet to be seen. The rise of these agent building platforms such as OpenAI’s and now Amazon’s seems like the next frontier for LLM development. It is not just a playground for creating fun applications; rather, it also serves as an educational tool.","excerpt":"Making GPTs yesterday’s news.","categories":["Global Tech"],"tags":[],"author_name":"Sandhra Jayan","publish_date":"2023-11-17T16:54:56","publication_year":"2023","word_count":577,"keywords":["Anthropic","ChatGPT","OpenAI","AI","ML","RAG","Aim","prompt engineering","generative AI","foundation models"],"extracted_tech_keywords":["AI","ML","generative AI","foundation models","ChatGPT","OpenAI","Anthropic","Aim","RAG","prompt engineering"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/amazons-partyrock-jams-past-openai\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10119353,"title":"Microsoft Announces $1.7 Bn Investment to Advance Indonesia’s Cloud and AI Ambitions","content":"In a move that underscores its commitment to Indonesia’s digital transformation, Microsoft today announced a $1.7 billion investment over the next four years to bolster the country’s cloud and AI infrastructure, as well as provide AI skilling opportunities for 840,000 people and support the nation’s growing developer community. The investment, the single largest in Microsoft’s 29-year history in Indonesia, is aimed at helping the country achieve its ambitious “Golden Indonesia 2045” vision, which seeks to transform the nation into a global economic powerhouse. “This new generation of AI is reshaping how people live and work everywhere, including in Indonesia,” said Satya Nadella, chairman and CEO of Microsoft. “The investments we are announcing today – spanning digital infrastructure, skilling, and support for developers – will help Indonesia thrive in this new era.” Budi Arie Setiadi, Indonesia’s Minister of Communications and Information Technology, emphasised the significance of the partnership with Microsoft, stating that it “perfectly aligns with our ambition for a future driven by digital innovation” and will position the country as a “pivotal contributor to the global technological supply chain.” Nadella highlighted the critical role of developers in harnessing AI to fulfil Indonesia’s potential as a digital economy. Microsoft will continue to help foster the growth of the country’s developer community through new initiatives such as AI Odyssey, which is expected to help 10,000 Indonesian developers become AI subject matter experts by learning new skills and earning Microsoft credentials. Microsoft is significantly boosting its investments in Southeast Asia. Recently, the company announced a $2.9 billion investment over the next two years to enhance its hyperscale cloud computing and AI infrastructure in Japan. Additionally, Microsoft aims to extend its digital skilling initiatives, aiming to train over 3 million individuals in AI skills within the next three years.","excerpt":"A few weeks ago, Microsoft invested $2.9 billion in Japan for cloud and AI infrastructure","categories":["AI News"],"tags":["Satya Nadella"],"author_name":"Gopika Raj","publish_date":"2024-04-30T18:51:14","publication_year":"2024","word_count":295,"keywords":["Satya Nadella","Go","API","programming_languages:R","cloud computing","AI","innovation","digital transformation","Git","Aim","R"],"extracted_tech_keywords":["AI","Aim","cloud computing","R","Go","Git","API","digital transformation","innovation","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/microsoft-announces-1-7-billion-investment-to-advance-indonesias-cloud-and-ai-ambitions\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10015997,"title":"6 AI Companies That Shut Down In 2020","content":"The pandemic has made us witness many unimaginable things including losing our dear ones,  recession, lay-offs, salary cuts and more. Besides this, another outcome of the pandemic is the closing of businesses due to financial losses, and many other reasons that companies are facing amid this unprecedented time. With the pandemic and recession, it has become even difficult for startups and small businesses to weather the current storm. In this article, we are listing down six such AI companies, in no particular order, that faced the worst during the pandemic and led to shutting down. Stockwell Founded: 2017 Shut Down: July 2020 Stockwell, formerly known as Bodega, was shut down this year in July. Founded in 2017 by ex-Googlers, the company was an artificial intelligence-based vending machine startup. According to news sources, the company could not find a viable business for its in-building app-controlled “smart” vending machines stocked with convenience store items. CEO Paul McDonald stated to the news media that the current landscape of the economy had created a situation where they could no longer continue the operations. Starsky Robotics Founded: 2015 Shut Down: March 2020 In March, the maker of driverless trucks, Starsky Robotics announced that the company would shut down. The company used to create autonomous driving technology for long haul trucks on open highways. In a blog post, Stefan Seltz-Axmacher, the CEO & Co-Founder at Starsky Robotics stated, “In 2015, I got obsessed with the idea of driverless trucks and started Starsky Robotics, and in 2020, we’re shutting down.” Seltz-Axmacher mentioned that the breakthroughs of the unmanned truck failed to appear as well as the downpour of investors’ interest became a drizzle. Hipmunk Founded: 2010 Shut Down: January 2020 Founded in 2010 by Adam J. Goldstein and Reddit co-founder Steve “spez” Huffman, Hipmunk was a trip planning platform. In June 2016, the company announced the launch of an AI-powered assistant known as Hello Hipmunk. The AI-assistant used to provide automated advice to would-be travellers with added support for Facebook Messenger and Slack. The shutdown comes after three and a half years of being acquired by Concur, a corporate travel\/expensing platform. Sorabel Founded: 2014 Shut Down: July 2020 Sorabel, an Indonesian fashion e-commerce startup, decided to shut down this July due to COVID crisis. According to sources, the e-commerce startup was using AI-based applications in order to spot fashion trends. The reason for shut down is that due to the pandemic, consumers avoided non-essential spending, which led to enormous losses for the company. ScaleFactor Founded: 2014 Shut Down: July 2020 ScaleFactor is a finance and accounting platform that recently raised $100 million from investors to create AI-powered products for automating bookkeeping, financial forecasting, bill paying, tax completion and other financial tasks. However, soon after the funding, the company announced that it would be shutting down due to COVID-19. According to a blog post, the closing down of the company led to some different reason. As per the company’s employees, there were far more issues such as books filled with errors, forced to re-hire accountants, among others. Nudge.ai Founded: 2014 Shut Down: January 2020 Nudge.ai is a Toronto-based startup that leveraged artificial intelligence and machine learning to automatically fill potential gaps in CRM data. In January, the company announced that it will soon end its operations and will be shutting down. The company‘s process focussed mainly on a CRM platform that used artificial intelligence to pull data from communications platforms — including email, phones, chat platforms and even calendar appointments — to automatically enter professional interactions and relationships that staff might not otherwise think to feed into a CRM system.","excerpt":"The pandemic has made us witness many unimaginable things including losing our dear ones,  recession, lay-offs, salary cuts and more. Besides this, another outcome of the pandemic is the closing of businesses due to financial losses, and many other reasons that companies are facing amid this unprecedented time.  With the pandemic and recession, it has […]","categories":["AI Trends"],"tags":["AI Companies","companies using business analytics","high paying jobs in india","Startups"],"author_name":"Ambika Choudhury","publish_date":"2020-12-26T16:00:00","publication_year":"2020","word_count":602,"keywords":["Go","funding","artificial intelligence","machine learning","ELT","AI","high paying jobs in india","programming_languages:R","RAG","AI Companies","Startups","companies using business analytics","R","startup"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","RAG","R","Go","ELT","startup","funding","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/ai-startups-shutdown-2020\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":23756,"title":"How Google Is Using Deep Learning To Identify Out-Of-Focus Microscope Images","content":"Deep learning has emerged as a leading machine learning tool in computer vision and has attracted considerable attention in biomedical image analysis. There is a renewed interest in medical image computing, and here, deep learning has proved to be effective in its ability to handle complex microscopic images. Google’s latest research covers two important aspects: Basics of machine learning for analysing microscope images Handling out-of-focus images with the help of a pre-trained TensorFlow model with plug-ins from the Google Accelerated Science team Deep neural networks have already played an important role in a host of tasks such as detection, segmentation, and classification in microscopic image analysis. Now, Google’s current research has found a way for scientists eager to use cutting-edge DL techniques for advanced image analysis work. The research deals with an important challenge of dealing with out-of-focus images. According to a post by Google, despite having autofocus systems on state-of-the-art microscopes, poor configuration or hardware incompatibility can produce substandard images, which results in quality issues. Google’s deep learning researchers have automated a technique to rate focus quality which can enable the detection, troubleshooting and removal of such images. The paper Assessing Microscope Image Focus Quality With Deep Learning introduces a deep neural network model that could identify on a small 84 × 84 image patch (about several times the area of atypical cell), the extent of the image blur and whether the image blur is even well-defined, which implies if the image patch is a background. The research is geared to enable a precise and accurate automatic assessment of microscope focus quality. There are a range of commercial off-the-shelf solutions for low quality image detection but microscopic images pose a more complex challenge. According to the Google research paper: Most microscope images are shift and rotation invariant They feature different offset (black-level) and pixel gain, photon noise In fluorescence microscopy, one of the different microscopy imaging modalities, an image may correspond to one of many possible fluorescent markers each labeling a specific morphological feature For scientists, gathering high quality optical microscopy images can be a challenge since individual images are mostly out of focus and noisy. These types of image degradation may occur on only a small fraction of a dataset too large to survey manually especially in high-content screening applications Implementation Google researchers started with a dataset of images consisting of in-focus and multiple out-of-focus images of U2OS cancer cells. The images shared only a few traits such as the image focus varied or many regions of images consisted of just a background. With this data, researchers trained a model that could identify, on a small 84 × 84 image patch both the severity of the image blur and whether the image blur is even well-defined. Here’s Where Google’s Research Differed In Technique: Instead of training the model on focal stacks of defocused images gathered from microscopes or training the model on manually labeled images, researchers used synthetically defocused versions of real images. This method enabled them to identify the absolute focus quality rather than relative measures of quality. Another key upside of the approach was that the researchers could generate a vast number of training examples needed for deep learning with just an in-focus image dataset Also, the researchers took advantage of the well-known behavior of light propagation in a microscope to achieve this. With this method, Google researchers also proved that out-of-focus images can be a more accurately as compared to previous approaches and the model is generalised to different image and cell types. While much research has already been done in automatic detection of low quality images in photographic applications, researchers posit that microscope images differ from consumer photographic images. Most microscope images are shift and rotation invariants, which have varying offset and pixel gain and dynamic range. The team integrated a pre-trained TensorFlow model with plugins in Fiji (ImageJ) and CellProfiler, open source scientific image analysis tools that can be used with both a graphical user interface or through scripts. Outlook Google has been at the forefront of numerous breakthroughs in image assessment and image recognition. In December 2017, the search giant introduced to the Neural Image Assessment (NIMA) system, a technique that dishes out the predictors used when humans see an image or, the mean scores or ratings given when judging a photo. This approach also categorised images in two categories — low and high quality — with the end goal to match human perception. With continuous innovation in image assessment, Google’s quality assessment models will not only revolutionise fields like medical diagnostics but also biology and geosciences as well.","excerpt":"Deep learning has emerged as a leading machine learning tool in computer vision and has attracted considerable attention in biomedical image analysis. There is a renewed interest in medical image computing, and here, deep learning has proved to be effective in its ability to handle complex microscopic images. Google’s latest research covers two important aspects: […]","categories":["IT Services"],"tags":["Computer Vision","Google Research"],"author_name":"Richa Bhatia","publish_date":"2018-04-17T06:07:58","publication_year":"2018","word_count":767,"keywords":["Go","machine learning","AI","neural network","innovation","image recognition","computer vision","deep learning","Computer Vision","TensorFlow","R","Google Research"],"extracted_tech_keywords":["AI","machine learning","deep learning","neural network","computer vision","TensorFlow","image recognition","R","Go","innovation"],"url":"https:\/\/analyticsindiamag.com\/it-services\/how-google-is-using-deep-learning-to-identify-out-of-focus-microscope-images\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":52865,"title":"This AI Agent Uses Reinforcement Learning To Self-Drive In A Video Game","content":"One of the most used machine learning (ML) algorithms of this year, reinforcement learning (RL) has been utilised to solve complex decision-making problems. In the present scenario, most of the researches are focussed on using RL algorithms which helps in improving the performance of the AI model in some controlled environment. Ubisoft’s prototyping space, Ubisoft La Forge has been doing a lot of advancements in its AI space. The goal of this prototyping space is to bridge the gap between the theoretical academic work and the practical applications of AI in videogames as well as in the real world. In one of our articles, we discussed how Ubisoft is mainstreaming machine learning into game development.  Recently, researchers from the La Forge project at Ubisoft Montreal proposed a hybrid AI algorithm known as Hybrid SAC, which is able to handle actions in a video game. Most reinforcement learning research papers focus on environments where the agent’s actions are either discrete or continuous. However, when training an agent to play a video game, it is common to encounter situations where actions have both discrete and continuous components. For instance, when wanting the agent to control systems that have both discrete and continuous components, like driving a car by combining steering and acceleration (both continuous) with the usage of the hand brake (a discrete binary action). This is where Hybrid SAC comes into play. Through this model, the researchers tried to sort out the common challenges in video game development techniques. The contribution consists of a different set of constraints which is mainly geared towards industry practitioners. The Algorithm Behind The approach in this research is based on Soft Actor-Critic which is designed for continuous action problems. Soft Actor-Critic (SAC) is a model-free algorithm which was originally proposed for continuous control tasks, however, the actions which are mostly encountered in video games are both continuous as well as discrete. In order to deal with a mix of discrete and continuous action components, the researchers converted part of SAC’s continuous output into discrete actions. Thus the researchers further explored this approach and extended it to a hybrid form with both continuous and discrete actions. The researchers also introduced Hybrid SAC which is an extension to the SAC algorithm that can handle discrete, continuous and mixed actions — discrete-continuous. How It Works The researchers trained a vehicle in a Ubisoft game by using the proposed Hybrid SAC model with two continuous actions (acceleration and steering) and one binary discrete action (hand brake). The objective of the car is to follow a given path as fast as possible, and in this case, the discrete hand brake action plays a key role in staying on the road at such a high speed. Wrapping Up Hybrid SAC exhibits competitive performance with the state-of-the-art on parameterised actions benchmarks. The researchers showed that this hybrid model can be successfully applied to train a car on a high-speed driving task in a commercial video game, also, demonstrating the practical usefulness of such an algorithm for the video game industry. While working with the mixed discrete-continuous actions, the researchers have gained several experiences and shared them as a piece of advice to obtain an appropriate representation for a given task. They are mentioned below Identify which action components (both discrete and continuous) should be made dependent on each other. When in doubt, it is advised to start with a simpler parameterization based on independent components, and only investigate later the potential benefits of more complex parameterizations. When a continuous component depends on a discrete component, consider duplicating it (one for each discrete action) as long as the model size remains reasonable. This will allow to consider them as independent, making it easier for the model to specialise the value of the component to each discrete action.If possible, try to avoid dependencies among continuous dimensions, so as to keep simple parameterization where each action dimension can be sampled independently.","excerpt":"One of the most used machine learning (ML) algorithms of this year, reinforcement learning (RL) has been utilised to solve complex decision-making problems. In the present scenario, most of the researches are focussed on using RL algorithms which helps in improving the performance of the AI model in some controlled environment. Ubisoft’s prototyping space, Ubisoft […]","categories":["AI Features"],"tags":["Reinforcement Learning","Reinforcement Learning Systems"],"author_name":"Ambika Choudhury","publish_date":"2019-12-31T11:00:00","publication_year":"2019","word_count":656,"keywords":["Reinforcement Learning Systems","Go","machine learning","Reinforcement Learning","TPU","AI","programming_languages:R","ML","programming_languages:Go","R"],"extracted_tech_keywords":["AI","machine learning","ML","TPU","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/this-ai-agent-uses-reinforcement-learning-to-self-drive-in-a-video-game\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10140027,"title":"Jio-Backed TWO AI Launches Multilingual Voice Model Sutra HiFi","content":"TWO AI, an AI startup backed by Jio has launched a sophisticated voice-to-voice AI model called Sutra HiFi. Leveraging a Dual Diffusion-Transformer Architecture, the model effectively decouples distinct voice tones from language-specific accents, promising a better voice interaction quality. With the capability to engage in over 30 languages, including a range of Indic languages and regional dialects, SUTRA HiFi positions itself as a vital tool for businesses aiming to connect with diverse global audiences. This multilingual support not only enhances customer engagement but also fosters inclusivity, allowing enterprises to navigate various cultural nuances. A key feature of SUTRA HiFi is its ultra-low latency, ensuring that interactions feel immediate and natural. This aspect is crucial for applications requiring rapid feedback, such as customer service and collaborative tools, where delays can hinder user experience. Moreover, the model prioritises delivering natural-sounding, human-like voice interactions, making it suitable for an array of applications from virtual assistants to interactive voice response systems. The emphasis on voice quality fosters meaningful connections with users and improves the overall effectiveness of communication in a digital-first world. TWO previously  launched a family of models called SUTRA. These cost-efficient, multilingual generative AI models excel in 50+ languages, offering speech, search, and visual processing capabilities. The model is available on the ChatSUTRA app, which is similar to ChatGPT. The company recently launched SUTRA Avatar, which offers PixelReal visuals, human-like expressions, and natural voice synthesis that works in real time. The startup raised a $20M seed fund in February 2022 from Jio Platforms and South Korean internet conglomerate Naver. “Jio has been one of our key partners for a long time and has invested in us from the very beginning,” said Pranav Mistry, the founder of TWO, in an exclusive interaction with AIM. Recently, TWO hosted Reliance Industries chairman Mukesh Ambani, and Akash Ambani, chairman of Reliance Jio Infocomm, at their US office, where they discussed the evolving role of AI in India and beyond over a cup of chai with founder Pranav. “Pleasure hosting Mukesh Ambani and Akash Ambani at our TWO US office and discussing all things AI for India and beyond over chai,” Mistry posted on LinkedIn.","excerpt":"It can generate voice interactions in over 30 languages, including various Indic languages and regional dialects.","categories":["AI News"],"tags":["Jio AI Cloud","TWO AI"],"author_name":"Siddharth Jindal","publish_date":"2024-11-01T11:57:38","publication_year":"2024","word_count":358,"keywords":["ChatGPT","API","TWO AI","AI","virtual assistants","Git","RAG","Ray","Aim","Jio AI Cloud","generative AI","R"],"extracted_tech_keywords":["AI","generative AI","ChatGPT","Aim","Ray","RAG","virtual assistants","R","Git","API"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/jio-backed-two-ai-launches-multilingual-voice-model-sutra-hifi\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":67979,"title":"Why Online Collaboration Is So Difficult For Data Scientists","content":"Data science skills are one of the critical sources that can provide businesses with a competitive advantage in the market. It is not only being used as a key tool to generate informative insights for making business decisions but also to optimise their operations using those insights. And therefore, effective collaboration among data science teams is the key for businesses to solve complex business problems. In fact, in an article, co-founder and CEO of data science start-up Dataiku, Florian Douetteau stated that one of the most important factors of successful data science is teamwork and collaboration. He said, “No data scientist is an island, nor could they be.” Usually, a data science team would involve data scientists, analysts, and data engineers, and it’s their collaborative efforts that work towards extracting insights from the massive amount of data available. For a data science team to successfully work, they must collaborate while analysing the data, building the model, and sharing insights. As a matter of fact, businesses must encourage their data scientists to collaborate well with engineers to come out with better workflows, better products, and better results. It is also critical for businesses to have enough collaboration among data scientists so that the shared context reduces the dependency on a handful of employees, and data can be democratised for employees of other departments. Experts believe data scientists sharing their ideas as well as their insights to other team members also acts as a robust feedback process. Having said that, with this pandemic in place, businesses have mandated remote working for their employees, including their data professionals, and that has given rise to online collaboration with the help of the cloud. Although online collaboration can bring in several advantages for employees, it definitely creates few challenges for data scientists who are working on solving complex business problems with critical data. In this article, we are going to share a few reasons as to why online collaboration isn’t for data scientists. Diminished productivity and reduced efficiency among data scientists One crucial aspect that is lost with online collaboration is the personal contact and in-person communication, which can sometimes lead to confusion, especially for those who are handling a massive amount of data on a daily basis. With critical data in hand, it is usually required to personally communicate with other departments, which is lost with remote working and online collaboration. Online collaboration is a ripple effect of the mandated remote working, which forces employees to be responsible for managing their tasks, turning work on deadline as well as keeping accountability of the projects, which is a lot to handle for data scientists and could hamper their productivity. Majority of data scientists work with a massive amount of training data and complex data engineering processes, and collaborating with other teams is necessary for them to scale their projects. Although online collaboration has been created so that employees can be connected 24\/7, many a time, these isolated data scientists go for days without communicating or collaborating with other employees and teams, which again can have a negative impact on their mental health. In fact, in a blog post on Quora, Ben Hamner, the co-founder and CTO of Kaggle has stated, “Most data scientists, and data science teams, have terrible practices for collaboration. The current default workflows have grown organically and are bad. You need to be really intentional to do a lot better, and this yields large gains in productivity and reducing painful frictions.” Virtual collaboration also brings in a lack of clarity and direction in data science workflows. Also, with continuously working with data, data scientists usually require human interaction to keep up their morale high amid this isolation. In-person collaboration not only brings cohesiveness but also allows data professionals to keep up the loop of communication with other teams. Also Read: How Laying Off Data Scientists Amid COVID-19 Pandemic Can Hamper Business’s Future Success Inability to ask the right questions Asking the right questions is critical for data scientists to define problem statements and create solutions to solve them. Majority of data science work goes into solving business problems and to get comprehensive details about the problem at every step, it is vital that data scientists ask the right questions. It not only helps data scientists to define their goals and objectives of their working but also helps them in curating better data, building better analysis and making an accurate prediction. However, with virtual collaboration, data scientists face the inability to ask the right question at the right time. And with ineffective communication via virtual collaboration, data scientists could work on wrong data due to lack of understanding and can result in making false assumptions. In fact, a Singapore-based data scientist, Admond Lee Kin Lim, wrote in his Medium blog post that, “I believe asking right questions and defining problem statements are some of the challenges faced by many beginners in data science.” And for this, constant communication is imperative among data scientists and business managers. But with online collaboration, data scientists lose their storytelling skill which is critical to communicate statistics and data to stakeholders. Also Read: What To Expect When You Start A Job As A Junior Data Scientist? Impedes their learning process Along with continuous feedback, the in-person collaboration also brings in constant learning among data scientists in the team. However, with mandated lockdown, data scientists rely on online collaboration for their daily communication, which is not only unreliable but also creates isolation. In-person collaboration acts as an active engagement amid colleagues of the same profession, but isolated learning via online collaboration can create disengagement and hamper the overall cause of collaboration. Alongside, no collaboration tools can compete with the real-time nature of in-person collaboration, and it also comes with its own technical issues like network glitches and reduced bandwidth, to name a few. In fact, Pascal Bugnion, Data Engineer at Faculty — an AI company — has stated in his recent blog post that, “Data science teams are no longer composed of a lone unicorn who works furiously on his laptop to eventually present a few plots or a web application. We now expect that data science is done by cross-functional teams that bring together subject-matter experts, modellers, data visualisation experts, machine learning engineers, product managers and designers.” Also Read: What Are The Markers Of A Genuine Data Scientist? Distrust and suspicion Lastly, as data scientists and other data professionals work with critical data and sensitive company information, they have always been scrutinised over the privacy and security issues that they can bring in for companies. With online collaboration, the majority of the data science functions work on cloud-based software which always creates a security concern for business leaders. Virtual collaboration can not only create distrust among the team members and their employers but can also create a lag in immediate response as work for home gives them the ability to work on their own time. In recent news, a video conferencing tool Zoom has undergone criticism on their privacy issues and security vulnerabilities. Such security faults in cloud collaboration tools will, in turn, create a conservative mindset among business owners to keep their sensitive data in the cloud. And that’s why now, the majority of these cloud providers are enhancing their security credentials for their customers. In fact, in a recent report, it has been noted that this shift to remote working has brought in new security risks for companies and has left nearly 50% of those employees worried about impending cyber threats. To which, the senior analyst at the Global Cyber Center of NYC, William Altman has recently stated to the media that, how all organisations are currently witnessing a surge in email-based threats and endpoint-security gaps due to this sudden transition to fully remote working. And that’s why, “it’s now more important than ever to consider both the security practitioner as well as ethical-hacker perspectives in order to stay secure, that’s what this is all about,” said Altman. Alongside, online collaboration diminishes the ability for data scientists to constantly keep their team members aware of their progress of work, which again creates questions and doubts. To build this belief, business leaders need to constantly update their feedback for the workflow. However, with online collaboration, it creates a gap between the leaders and their employees.","excerpt":"Data science skills are one of the critical sources that can provide businesses with a competitive advantage in the market. It is not only being used as a key tool to generate informative insights for making business decisions but also to optimise their operations using those insights. And therefore, effective collaboration among data science teams […]","categories":["AI Features"],"tags":["collaboration","Data Scientists","datascience"],"author_name":"Sejuti Das","publish_date":"2020-06-24T14:00:00","publication_year":"2020","word_count":1388,"keywords":["data science","Go","machine learning","AI","datascience","collaboration","RAG","data engineering","ViT","Rust","GAN","R","Data Scientists"],"extracted_tech_keywords":["AI","machine learning","data science","RAG","R","Go","Rust","data engineering","GAN","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/why-online-collaboration-is-so-difficult-for-data-scientists\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10171812,"title":"Meta Appoints Arun Srinivas as MD and Head of India Division","content":"Meta Platforms announced on Monday that Arun Srinivas has been appointed as the managing director and head of Meta in India. Since 2022, Srinivas has served as the director and head of the advertising business for Meta India. In this position, he has been responsible for overseeing business strategy and revenue growth across all Meta platforms in India. He is part of the country’s leadership team, and his efforts have centred on key revenue initiatives such as AI, Reels, and Messaging. This appointment follows Sandhya Devanathan, whose responsibilities have expanded as Meta’s vice president, who will now oversee the company’s operations in Southeast Asia. As India continues to be at the forefront of economic growth and innovation, we are excited to have Arun at the helm of our efforts in this critical market. Arun’s impressive track record of building high-performing teams, driving product innovation, and fostering strong partnerships makes him the ideal leader to drive Meta’s continued investment in the country,” Devanathan said. A postgraduate from IIM Kolkata, Srinivas has nearly three decades of experience in leadership roles in sales and marketing at organisations such as Hindustan Unilever, Reebok, OLA, and investment company WestBridge Capital. Currently, Srinivas holds the position of Director and Head of Ads Business in India. Since he began with the company in 2020, he has been instrumental in directing Meta’s collaborations with the largest advertisers and agency partners in the country, concentrating on strategic revenue priorities. According to Reuters, Srinivas’s new role comes amid ongoing tensions between Meta and Indian authorities over accusations of anti-competitive behaviour. Recently, the Competition Commission of India (CCI) imposed a fine on Meta for antitrust violations and prohibited WhatsApp, which is owned by Meta, from sharing user data with other Meta apps for advertising purposes for five years. Meta has contested the CCI’s decision, expressing concerns about its potential impact on business. However, earlier this year, an Indian tribunal granted Meta temporary relief by suspending the five-year data sharing ban.","excerpt":"The announcement follows Meta’s decision to expand the role of its India vice president, Sandhya Devanathan, who now also oversees the Southeast Asia business.","categories":["AI News"],"tags":["Meta AI","Meta leadership"],"author_name":"Smruthi Nadig","publish_date":"2025-06-16T16:53:57","publication_year":"2025","word_count":329,"keywords":["Go","API","Meta AI","programming_languages:R","AI","innovation","programming_languages:Go","Meta leadership","Rust","GAN","R","programming_languages:Rust"],"extracted_tech_keywords":["AI","R","Go","Rust","API","GAN","innovation","programming_languages:R","programming_languages:Go","programming_languages:Rust"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/meta-appoints-arun-srinivas-as-md-and-head-of-india-division\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10095520,"title":"Eleven Labs Raises $19 million Fund, Launches New Generative AI Products","content":"Eleven Labs, the generative AI company leading the way for audio and voice generation products has got more fuel for its mission to make its content universally accessible. The research company has raised $19 million in a series A fund led by Nat Friedman, Daniel Gross, and Andreessen Horowitz. Other notable investors include Credo Ventures, Concept Ventures, as well as a group of strategic angel investors, including Mike Krieger, co-founder of Instagram; Brendan Iribe, co-founder of Oculus VR; Anjney Midha, co-founder of Ubiquity6; Mustafa Suleyman, co-founder of Deepmind & Inflection; Siqi Chen, co-founder of Runway; Ali Albazaz, co-founder of Inkitt; Dima Shvets, co-founder of Reface; Aravind Srinivas, co-founder of Perplexity AI; Guillermo Rauch, founder of Vercel; Tim O’Reilly, founder of O’Reilly Media; as well as Creator Ventures, SV Angel, Embark Studios, Storytel, and TheSoul Publishing. We're thrilled to open a new chapter at ElevenLabs!Today we’re announcing new voice AI products & our $19m Series A round led by @natfriedman @danielgross & @a16z. Listen & read more: https:\/\/t.co\/8LHtqvdCQt— ElevenLabs (@elevenlabsio) June 20, 2023 The company has also announced a new set of products called Projects. Currently available in early access, Projects hosts a production workflow for creating dynamic long-form spoken content, along with an editing suite. This will give creators and authors a huge control over AI-generated audio content without needing to leave the platform at all. Projects joins the ElevenLabs product lineup with other notable offerings such as Speech Synthesis and VoiceLab. Speech Synthesis utilizes pre-existing synthetic voices, while VoiceLab facilitates the creation of distinctive voices or digital renditions of existing ones. Additionally, ElevenLabs has made available Eleven Multilingual, a speech synthesis model capable of supporting various major European languages. Moreover, ElevenLabs is also introducing AI Speech Classifier, allowing users to upload an audio sample and determine if it contains AI-generated audio from ElevenLabs. This innovative tool is the pioneer in the field of generative audio and is now accessible to the general public and a few chosen partners through an API. By launching the AI Speech Classifier, ElevenLabs demonstrates its dedication to transparency and plays a vital role in establishing a secure environment for generative media. Mati Staniszewski, the co-founder of Eleven Labs, said, “Our mission is to be the ultimate tool for storytelling, dissolving language barriers and putting all audiences in the reach of all content creators in a safe and responsible way. With an incredible, growing team and these exceptional investors, ElevenLabs is now ever closer to realizing its long-term goal of making all content universally accessible in any language and in any voice.” Launched in Beta in January, Eleven Labs has already attracted 1 million users for creative, entertainment, and publishing spaces. The company says that over 10 years worth of audio content has been generated using its services.","excerpt":"ElevenLabs is also introducing AI Speech Classifier, allowing users to upload an audio sample and determine if it contains AI-generated audio from ElevenLabs.","categories":["AI News"],"tags":["ai funding"],"author_name":"Mohit Pandey","publish_date":"2023-06-21T11:24:23","publication_year":"2023","word_count":463,"keywords":["Go","API","programming_languages:R","AI","programming_languages:Go","Git","ai funding","ViT","generative AI","R"],"extracted_tech_keywords":["AI","generative AI","R","Go","Git","API","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/eleven-labs-raises-19-million-fund-launches-new-generative-ai-products\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10019993,"title":"When Androids Dream Of Electric Sheep","content":"Recently, Elon Musk’s startup, Neuralink, has wired up a monkey to play video games by planting a computer chip in its skull. Neuralink is trying to develop an implementable computer-brain interface to increase information flow rate from the human brain to a machine. The San Francisco- based neurotechnology firm comprises a team of around 100 people. In the light of recent developments, we look at some of the ethical implications of introducing such devices in the human brain. “It is very likely, within a century or two, Homo sapiens, as we have known it for thousands of years, will disappear” Yuval Noah Harari, Philosopher and Best-selling Author of ‘Sapiens’ and ‘Homo Deus’ Free Will According to Musk, humans are already Cyborgs since they have a tertiary digital layer in the form of phones, computers, and applications. While this might not be as sophisticated as putting chips in our brains, technology holds a lot of power over us. In the future, if cutting-edge tech hardware makes its way into human brains, the machines might take over our decision-making faculties. Meaning, the prosthetic makers can program us(ers) to further their agenda. Human agency and free will are at stake, when the Orwellian overlords rule the roost. The deep state and deep tech mix well. Such overarching technology could double as a surveillance system — Imagine living under the constant gaze of Panopticon. On the other hand, the privileged humans will harness advanced algorithms to become a super race of sorts- capable of interplanetary travel (Refer: Blue Origin, SpaceX), and probably a shot at immortality (Refer: Calico, Sierra Sciences). But, there is always a catch. Humans tend to overestimate the power of algorithms. So delegating machines to do our thinking might turn out deleterious to our gene pool at large. “Once you dress something up as an algorithm or as a bit of artificial intelligence it can take on this air of authority that makes it really hard to argue with.” Hannah Fry, UCL Mathematician and Author of Hello World. Further, the chasm of inequality will widen in a capitalist setting since such technologies can be used to do social engineering, leading to a transnational Big Tech oligarchy. Thought Police Mass surveillance infringes on our fundamental right of freedom of speech, which, in turn, threatens democracy. At this point, AI-based surveillance technologies are restricted to facial or voice recognition systems or other forms of biometric identification. But the snafus from over-policing and misidentification while using the tech are in plenty. In the future, if we are able to install chips to track the emotions of every citizen, the repercussions will be catastrophic. The health of body politic will decline dramatically with political parties leveraging tech to implement their autocratic ideologies. As things stand, such technology could become a reality in a not too distant future. Space and Time The brain retrofitting could potentially change the way we understand space or time. In the past, experiments have been done to control a monkey’s limbs. Future technologies could be used for fighting wars, where a person could be sitting in a US lab while his digital twin is engaged in combat in Afghanistan. “I feel like there’s a very strong, profound distinction between just using technology and integrating technology” Mark O’Connell, Journalist and Author of the book To Be a Machine. Combining AI with biotechnology, humans could also make a bid for cheating death. Technology has played a huge role in increasing our average life expectancy. However, health and death remain a very technical (medical) problem, more than a philosophical one. Upside While technologies like the ones Neuralink wants to develop can ‘improve’ or alter human cognition through intelligent chips, it can also help patients with disabilities or health conditions (Neuroprosthetics). People with motor disabilities, neurological damages or quadriplegics can reclaim normal life with the aid of such technologies. For example, Kernel, a Silicon Valley-based startup, is developing computerised brain implants to help patients with epilepsy, Parkinson’s or Alzheimer. The company has already developed magnetometers to track brain signals using head-mounted devices. Going Forward “Any technology humanity has created starting with fire is a double-edged sword. So it can bring improvements to life, to work, and to society, but it can bring perils, and AI has these perils.” Fei-Fei Li, Sequoia Professor Computer Science Department at Stanford University Given the far-reaching consequences such technologies can have on the human race as a whole, it’s imperative to have scientists, philosophers, politicians, technocrats and ethicists across the world on the same page to develop an inclusive strategy.","excerpt":"Recently, Elon Musk’s startup, Neuralink, has wired up a monkey to play video games by planting a computer chip in its skull. Neuralink is trying to develop an implementable computer-brain interface to increase information flow rate from the human brain to a machine. The San Francisco- based neurotechnology firm comprises a team of around 100 […]","categories":["IT Services"],"tags":[],"author_name":"Kashyap Raibagi","publish_date":"2021-02-11T13:00:00","publication_year":"2021","word_count":760,"keywords":["Go","API","artificial intelligence","programming_languages:R","AI","Git","RAG","Aim","R","startup"],"extracted_tech_keywords":["AI","artificial intelligence","Aim","RAG","R","Go","Git","API","startup","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/it-services\/when-androids-dream-of-electric-sheep\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":53830,"title":"VISA Buys Plaid For $5.3 Billion: How This Acquisition Marks The Arrival Of API-First Businesses","content":"In what could be the one of the biggest news for the fintech industry, VISA spent a whopping $5.3 billion to acquire Plaid, an API-first service that enables smooth cashless transactions. In 2016, its $44 million Goldman-led Series B valued the company at $250 million. Today, with sales quadrupled and the founders Perret and Hockey employ 130 people out of a spacious two-story office in the South of Market area of San Francisco. In September 2018, Plaid announced that it has secured “strategic investments” from both Visa and its rival Mastercard. This put the fintech startup at a market value of $2.65 billion; half of what Visa has bought them for. “We were very impressed by the Visa team from the minute they approached us. They share our vision for the future of financial services and have a deep respect for the developer community we support.” -Zachary Perret, CEO, Plaid Visa and Mastercard, who have been the go-to platforms for transactions have been forced to adapt to the rapid digitisation that occurred over the past decade. Fintech companies like Stripe, Twilio and Plaid have established a certain kind of trust with the customers by making it easy to perform transactions with negligible waiting times. If you can’t beat them, join them Fintech firms such as Venmo, Robinhood and Bettermen already use Plaid’s APIs to connect with customers’ bank accounts. Plaid’s customer base has increased to 20 million and no wonder Visa handed them this lucrative deal. What Does It Mean For VISA Across the world, cashless and contactless payments have become hugely popular. They are being used in metros, trains and ferries. For examples, major cities such as London, Vancouver and Sydney already boast of contactless card acceptance across their mass transit systems while global cities such as Madrid, Manchester, Milan, Rio de Janeiro, Singapore, Helsinki and Rome are in the midst of upgrading to contactless payments. A study by Visa in Australia showed that 41% of retailers witnessed an increase in sales by introducing the contactless payment method. Customer stickiness too improved. With a slew of transit projects including metros coming up across the globe, countries like India too, with a very large customer base have started using contactless for a seamless payment experience while on the go. In early 2018, Visa announced that it has hit 20 million contactless cards in India alone. TR Ramachandran, Group Country Manager for Visa in India & South Asia, said, “Thanks to the government’s push towards enabling contactless payments, we expect this number to significantly increase in the months to come.” Visa also announced that they will be issuing 1.5 billion contactless payment cards by the end of 2018 and will account for 50% of all payment cards shipped. Visa was on an acquisition spree even before it bagged Plaid. Last year Visa announced it has completed the acquisition of Verifi to bring on board the benefits of cutting downtime by deploying intelligent, data-driven tools that foster collaboration, build trust, and improve the overall customer experience. So, Visa has been, like any other smart organisation, eyeing on having a hold on the future as well. Their acquisition of Plaid for a huge sum, almost double the market value hints at how determined they are to keep in touch with the changing times. Future Direction Banking was built for a world that did not envision the internet. Even today, the backend for most consumer financial products is still paper, phone calls, and faxes. We are in the early stages of a digital revolution in financial services, which has thus far been led by many fintech startups and a handful of forward-leaning banks. What is happening today is reminiscent of the digitisation that occurred in commerce over the past two decades. What began with improvements at the network level—led by Mastercard and Visa—precipitated the rise of e-commerce and digital commerce. Today, more than half of all payments are made online, and the trend continues to accelerate. For organisations that deal with financial transactions, the customer stickiness is a big challenge, which is directly related to quick transaction time and ease of use of contactless payments. The modern-day fintech services also imbibe cutting edge technologies into their workflows to keep abreast with customer engagement, security and growth. Companies like Stripe, have developed in house tools like Radar for detecting and preventing fraud. Radar’s core is powered by adaptive machine learning. The algorithms evaluate every transaction for fraud risk and take action appropriately. According to PricewaterhouseCoopers, over $150 billion will be invested in FinTech companies over the next 3 to 5 years. fintech has its reach in domains like insurance, lending, regulations, trading, e-banking and other payment services, and thus has a wide scope. Today companies like JP Morgan, urge their employees to get a good grasp of tools like Python to be ready to handle the transformation. There is no doubt that the advent of the internet and smartphones have changed the way we do business. And, as other phenomena like cryptocurrencies looming large around the corner, traditional banking service providers are hastening up with acquisitions and upskilling.","excerpt":"In what could be the one of the biggest news for the fintech industry, VISA spent a whopping $5.3 billion to acquire Plaid, an API-first service that enables smooth cashless transactions. In 2016, its $44 million Goldman-led Series B valued the company at $250 million. Today, with sales quadrupled and the founders Perret and Hockey […]","categories":["AI Trends"],"tags":[],"author_name":"Ram Sagar","publish_date":"2020-01-14T19:30:00","publication_year":"2020","word_count":853,"keywords":["Go","API","machine learning","AI","ML","Git","Python","Rust","GAN","R"],"extracted_tech_keywords":["AI","machine learning","ML","Python","R","Go","Rust","Git","API","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/visa-buys-plaid-for-5-3-billion-how-this-acquisition-marks-the-arrival-of-api-first-businesses\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":16924,"title":"10 Leading Analytics Products From India That Made It Big","content":"As the analytics industry is touching every sector, there has been a rise in analytics product providers from India, who with their innovative products are working on easing down various processes for these industries. Visualization, business intelligence, artificial intelligence, machine learning are some of the areas that the companies are exploring to come out with unique analytics products. From start-ups to big firms, there are many companies from India bestowing products in this space. Analytics India Magazine brings to you 10 such products from India that made it big over the years. So here’s the list in alphabetical order. 1| Buddy by Sapience Sapience is an innovative People Analytics @ Work solution. At the individual level, it empowers users to own their productivity and wellness through mindful work. At the enterprise level, managers and CxOs get accurate Workload Pattern and Capacity Utilization insights across every business dimension of interest, enabling them to align their talent pool to strategic needs. With its products, Sapience is helping over 100,000 users at 70+ enterprises across 12+ countries to move the needle on employee engagement, employee wellness, organization productivity and business profitability. Sapience Buddy brings amazing features to manage time by setting goals, getting automated updates and by bringing personal assistance to keep you on track. Whereas with Sapience Enterprise, it ensures employee engagement, wellness, productivity and profitability. 2| Capillary Analytics by Capillary Technologies Capillary Analytics, a self-serve BI tool helps over 200 organisations across the globe to embrace the power of analytics for daily decision making. The platform makes it very easy to get started by providing out of the box ETL and BI solutions to collect, link, clean, transform and visualise data. The Omni-channel data—offline stores, online stores, digital campaigns, CRM etc. that flows to the system allows to drill downs on all dimensions, thereby helping businesses make critical decisions more efficiently. In-memory big data technology combined with powerful visualisation provide insights in seconds fuelling business growth. Its powerful web interface, with simplistic UI helps in solving the most complex data problems effortlessly. Its multi-platform mobile app helps in getting data and insights on the go. The product has an AI based Automated Insights System that keeps a watch on the business, identifies problem areas and recommends campaigns and a ML based Product Recommendation Engine that combines online and offline data to provide the most accurate product recommendations across all channels. 3| Gramex by Gramener Gramex is a visual intelligence platform that helps organizations and clients analyse data and consume insights. The platform is a pure custom-built, with the bulk of the code written in Python and leverages the analytical libraries optimized for speed, coupled with visual libraries optimized for best-in-class data visualizations. The in-built data connectors source data from a variety of files, databases or third-party systems that provide APIs to pull in data. Its component libraries are a collection of components that provide visualisation and analytics support, whereas analytics library provides support similar to R or SAS but with a business focus. Third party analytics systems such as R, Hadoop, SAS, etc. can be integrated directly to leverage existing workflows. Visualization library helps transform the analysed (or raw) data into pictorial representations with third party visualisation libraries such as D3, Three.js, etc. included with data & analytics integration. It is also available for several other libraries such as Raphael, Circos, etc. To ensure security, the application is not designed to persist any data and all computations happen in-memory. 4| HealthWorks by TEG Analytics TEG’s HealthWorks, a solution powered by publicly available healthcare data and machine learning algorithms, helps payers predict in-market performance within five business days from the release of plans. With such an unduly advantage, payers can make reallocations into their marketing plan. It reads the entire industry landscape to understand the competitiveness of the products. The process involves exhaustive data preparation, a judicious mix of traditional and advanced machine learning techniques and visual analytics. HealthWorksTM platform is cloud based, that provides access to multiple business users across the insurer’s organization and help them make their own decisions through a self-serve interface. HealthWorksTM is used by finance, product, and marketing teams to reallocate the marketing spends based on their predicted potential in each of the markets. 5| LiquiData Platform by Nanobi Nanobi liquiData Platform is a full stack analytics platform created with agility and simplicity in mind, allowing organizations to create and launch analytics applications with ease and speed using published REST API interface. The platform allows the development of highly focused and interconnected analytic apps through the concept of Nanomart. With its REST API architecture and Nanomart paradigm, organizations can start small and then make rapid changes to their applications. Nanobi liquiData comes with a very user-friendly interface and an interactive and visual full-stack platform. The platform can be deployed as a cloud-based tool or installed as an on premise application. It can be easily embedded and integrated with other systems, thanks to its REST interface. Aside from the streamlined development of analytical applications, it delivers a rich trove of curated data for several industries, giving users that distinctive data-driven competitive edge. 6| Maya by Manthan Manthan Maya is the world’s first AI-powered conversational agent for business analytics. It enables users to interact with sophisticated enterprise analytics applications as if they were human, using natural language and receive AI-powered insights for decision making on-the-go. Targeted at CEOs and senior executives, Maya helps them get business insights instantly, whether at home or at their desk easily using a voice interface, instead of relying on a team of data analysts or IT experts. Through a conversation with Maya, a business executive can instantly get data-driven answers to any business questions. It is the first of its kind and brings paradigm shift in the way users interact with a complex business system. Maya can understand intent, sift through millions of data points, slice, dice, analyse them using sophisticated analytics models and respond with personalized answers based on a user’s role and context in seconds. Under the hood, Maya leverages Natural Language Processing (NLP), Intent Analysis, Machine Learning and Deep Learning to process queries. 7| MECBOT and IntelliDevice by Formcept A unified data analysis platform, Formcept helps enterprises derive actionable insights from large volumes of external and internal data quickly, thereby, significantly reducing the data-to-decision conversion time. The analytics products by Formcept helps in making of the largely unstructured data from external source, enable smarter decision making, among others. MECBOT by Formcept is an Open Cognitive Platform that is transforming the way data analytics, machine learning, deep learning, artificial intelligence (AI), and predictive analytics interact with each other on a near real-time basis. With its semantically harmonized smart data-lake and Dynamic Business Ontology, MECBOT empowers faster insights for mission-critical decisions. Whereas IntelliDevice activates industrial machines and equipment to send high-velocity data on performance and usage to cloud. Here, MECBOT analyses it to generate quick actionable alerts and event-triggers. With IntelliDevice, they aim to enable industries of tomorrow through interactive and intelligent machines. 8| pi by Progen ‘pi’ is a Next Generation Data Analytics Platform from ProGen that helps users with their data discovery, analytics, and advanced visualization needs. With a variety of new age features that enables users to venture beyond conventional Descriptive Analytics, ‘pi’ helps users explore the realms of diagnostic, predictive, & prescriptive analytics thereby differentiating it from other traditional BI & Visualization tools. ‘pi’ has a state-of-the-art machine learning & recommendations engine that provides users with actionable business insights for pro-active decision-making & also helps in configuring various predictive analytics scenarios. While conventional data analytics tools focus around displaying growth or de-growth of key metrics in the form of dashboards, visuals, or reports; ‘pi’ raises the bar with help of its proprietary algorithms to automatically identify various influencers that impact the growth or de-growth of key metrics. ‘pi’ also has strong capabilities around What-if Modelling and Simulation Workbenches that helps users perform different types of scenario configurations. It has a strong customer-base spreading across different Industry verticals. Few business areas where customers have got benefited include: identification of opportunities for revenue growth, business risk identification, identification of potential leakages that can lead to cost savings, and improving customer retention\/satisfaction. 9| Sales Decision Engine by BRIDGEi2i Sales Decision Engine is the AI powered sales enablement suite by BRIDGEi2i that helps transform sales organisations by enabling better visibility, accurate forecasting, actionable insights, and improved effectiveness for sales leaders, sales ops teams, and sales reps. SDE easily integrates with CRM systems and can be customised as per business needs owing to its flexible and scalable architecture. SDE has been featured by Gartner in its Market Guide for SaaS-based predictive analytics applications for B2B sales and marketing. SDE’s modules leverage advanced machine learning algorithms and artificial intelligence and address sales processes across the funnel. Some of it’s key benefits are: Provide complete visibility across customer segments, sales, revenues, and pipelines Deliver accurate and reliable sales predictions Identify opportunities that are most likely to convert Enable sales teams to identify and measure areas that need improvement Suggest strategic interventions to improve their success rate 10| SolarPulse™ and WindPulse™ by Machinepulse A part of Mahindra Group, MachinePulse™ provides rapidly scalable, end-to-end solutions to enable a smarter industrial Internet of Things (IIoT). They offer complete range of products from industrial grade data acquisition devices to a robust and scalable cloud platform, which can be deployed together or into existing IoT workflows with full flexibility. SolarPulse™ is MachinePulse’s flagship product which is a comprehensive solar PV asset monitoring and management application, developed for both utility scale and rooftop solar PV plants. Currently on its fourth version, it incorporates advanced monitoring capabilities, predictive analytics and asset management via an easy-to-use interface. WindPulse™ on the other hand is advanced wind farm analytics solution. It predicts generation, equipment health and helps efficiently manage the entire wind portfolio with an interactive user interface. It uses specialised machine learning algorithms to improve the overall productivity and revenue of wind farms. Both WindPulse™ and  SolarPulse™ is powered by Erixis™, their artificial intelligence engine based on the Internet of Things platform.","excerpt":"As the analytics industry is touching every sector, there has been a rise in analytics product providers from India, who with their innovative products are working on easing down various processes for these industries. Visualization, business intelligence, artificial intelligence, machine learning are some of the areas that the companies are exploring to come out with […]","categories":["AI Features"],"tags":["iot friendly database"],"author_name":"Srishti Deoras","publish_date":"2017-08-11T10:01:18","publication_year":"2017","word_count":1687,"keywords":["machine learning","artificial intelligence","AI","ML","RAG","NLP","Aim","deep learning","analytics","iot friendly database","predictive analytics"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","NLP","analytics","Aim","RAG","predictive analytics"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/10-leading-analytics-product-from-india-that-made-it-big\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10091595,"title":"Meet the Computer Scientist Who Solved 50-Year Old ‘Einstein’ Tiles Problem","content":"Back in November 2022, when David Smith, a hobbyist who loves playing around with shapes, came across a shape (a tile) that could fill an entire plane without forming regular patterns, he immediately contacted computer scientist Craig Kaplan. Kaplan, who had been in search of such a shape for decades, albeit using a computational approach, began running his software on this shape. By the end of January, they had proof of a wonderful open problem that occupied their minds for a long time. The Problem Imagine the tiling to be the same as your bathroom floor tiling where a square grid is repeated multiple times until it fills the entire surface without leaving any gaps. However, mathematicians have long been searching for a shape that can be tiled in a similar way, except it won’t form a pattern, even when extended infinitely. The tile doesn’t have to be a square grid. It can be a parallelogram or just about any shape that prevents the emergence of periodicity. “There are lots of examples of sets of shapes that tile aperiodically, but it was unknown whether a single shape could do it.” The best people have been able to do is with two shapes, most famously by Penrose tiling, which prohibited periodicity using two different rhombi, or two different quadrilaterals. Seems like we finally have the answer. Called the ‘hat’ (since it vaguely resembles a fedora), the crinkly kind of shape can be fit together in a number of ways. “The hat tile is just one of a newly discovered continuum of aperiodic tiles,” as per Quanta magazine “Like, there’s some number of ways you could put two hats together. And from that, there’s some number of ways that you can maybe put a ring of hats around a single hat, and from that maybe there’s some number of ways you could build two rings of hats and on and on,” said Kaplan. Thus, we are met with a combinatorial explosion of things to be kept track of, which gets incredibly tedious and error prone. Computation x mathematics Kaplan had written a software that could take a shape and “build a big block of them by brute force.” This acted as the raw data for researchers to intuitively study and find patterns in the large patch of tiles. Using computational methods, the researchers tried to analyse the tiling properties of the hat, including its Heesch number (which tells if layers of copies of the shape can surround a central copy without leaving any gaps) and isohedral number (which measures the number of ways the shape can be oriented to create a repeatable pattern). The results were that the shape shattered the record for both Heesch numbers and isohedral numbers by a wide margin – proving itself to be an Einstein. The term ‘Einstein’, a play on the physicist Albert Einstein, comes from German ein Stein meaning “one stone,” referring to a single tile. It is used to describe the property of the tile to fill an entire plane, without repeating itself in regular intervals. However, even after using intuition to understand the structure of the tiling, Kaplan mentions that software was important as part of their proof that the produced tiling was aperiodic. ​​One way to measure a tile doesn’t repeat itself is to look at all possible neighbourhoods that might contain around each tile, which is a finite number. By extension, it also means that there are only finitely possible neighbourhoods that can occur in an infinite tiling. The researchers used computation to construct a complete list of all possible neighbourhoods, and then further computation to verify that all such neighbourhoods have the properties that they need them to have in order to prove aperiodicity. “At a high level, the computation shows via exhaustive enumeration that we can always group individual tiles into clusters in a unique, forced way, and that we can group clusters into larger clusters, and those into even larger clusters, and so on ad infinitum,” added Kaplan. This forced hierarchy of clustering (which refers to the way tiles are arranged in local neighbourhoods) is one way that mathematicians can prove aperiodicity. While in principle all of these rules could have been derived by hand, it made much more sense to the researchers to turn this sort of tedious case-based analysis over to software. But, to show that the proof is valid for an infinite plane, one cannot just write down where every tile is – because that list would be infinitely long. “You need something more indirect, you need kind of an abstract rule that allows you to know that in principle you can go to infinity in every direction.” Here, an old theorem in the mathematics of tilings allowed the researchers to kind of bootstrap from large and finite up to infinity. “But, it’s not obvious,” says Kaplan. In fact, part of Kaplan’s independent work has been to study shapes that make copies of themselves and get stuck after a while, as opposed to assuming that once you get big enough, you can keep going forever. Einstein tiles on toilet paper? Previously, advances in tiling theory have had applications in other sciences like physics, chemistry, and engineering, most famously between Penrose tilings and quasicrystals. Therefore, Kaplan says that while there is no immediate application identified to this tiling, we can expect new and interesting connections conjuring up as more and more researchers fiddle around with it. Recently, there have also been a couple of people who have incorporated the Einstein tile into different kinds of games, shortly after the release. Take, for example, a twitter handle by name ZenoRogue, who made a really elegant implementation of the hat on his software, known as Hyperrogue. This software which already had kind of like a minesweeper game could adapt to any grid. It only goes to show that the pace of AI right now is remarkable, where demos are built in no time. The newest great discovery in tilings is already implemented in HyperRogue. So you can play infinite aperiodic Minesweeper on it, or whatever! pic.twitter.com\/te7FfUjUlb— eno Rogue (@ZenoRogue) March 23, 2023 Alongside this, Kaplan also mentioned a famous story from the 1990s, where Kimberly-Clark, the parent company of Kleenex, started to print toilet paper with Penrose tiles on it. However, the company that owned the patent to Penrose tiles sued them saying that this is inappropriate. “I feel the opposite. I would be flattered to see this design appear on toilet paper and I’m putting that out there for the record,” he remarked.","excerpt":"A tile that can fill an entire plane, without repeating itself in regular intervals is an Einstein","categories":["AI Features"],"tags":["Interviews and Discussions"],"author_name":"Ayush Jain","publish_date":"2023-04-17T18:00:00","publication_year":"2023","word_count":1096,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","BERT","llm_models:BERT","GAN","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","R","Go","BERT","GAN","llm_models:BERT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/meet-the-computer-scientist-who-solved-50-year-old-einstein-tiles-problem\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10173268,"title":"Musk’s Grok-4 Crushes Benchmarks, Beats OpenAI &#038; Google in RL","content":"The much-awaited Grok-4 is finally here. Despite a delayed livestream, the model didn’t disappoint. xAI chief Elon Musk even declared that Grok-4 is “PhD-level in everything”, adding that the model performs at a postgraduate level across multiple disciplines, including even unseen questions. “If [it’s] given the SAT, it would get perfect [scores] every time,” Musk said. He further revealed that on graduate-level tests like the GRE, Grok-4 reportedly scored near-perfect in every discipline of education, from humanities to languages, math, physics, and even engineering. Musk compared the model’s reasoning to that of humans. He added that Grok-4 was able to solve problems which it had not seen before. “These are not on the internet…Grok-4 is smarter than almost all graduate students in all disciplines simultaneously.” “I would expect Grok to discover new technologies that are used, maybe by the end of this year. It might discover new physics next year,” he declared. Crushing All Benchmarks ​​xAI has launched two versions of its latest model, namely Grok 4 and Grok 4 Heavy. The team described Grok 4 as a single-agent version, while Grok 4 Heavy is the multi-agent version. Both are available immediately and come bundled with access to SuperGrok tiers, where users can direct a network of Grok agents to assist with research and productivity. The SuperGrok tier comes with a new $300-per-month AI subscription plan. Moreover, Grok 4 will be deployed across hyperscalers and in xAI’s enterprise offerings. Talking about the Grok 4 Heavy version, Musk said, “It spawns multiple agents in parallel…They compare notes and yield an answer.” This “study group” style approach allows the model to solve more problems at test time, especially on complex benchmarks. “This is what we call test-time compute. We scale it up roughly by an order of magnitude,” Musk explained. Grok 4 is now accessible via API, and, according to xAI, the model leads key reasoning benchmarks. On the ARC-AGI-2 evaluation set, a benchmark designed to measure advanced reasoning, the model achieved 15.9% accuracy, reportedly doubling the performance of the next-best model, Claude Opus. “It was the only model in the last three months that broke the 10% barrier,” the xAI team noted. Grok 4 (Thinking) achieves new SOTA on ARC-AGI-2 with 15.9%This nearly doubles the previous commercial SOTA and tops the current Kaggle competition SOTA pic.twitter.com\/YbCMLXPJ2e— ARC Prize (@arcprize) July 10, 2025 Beyond accuracy, xAI argued that Grok 4 delivers “intelligence per dollar” that puts it in a “league of its own”. During the live stream, the xAI team also demonstrated the model’s progress using a challenging benchmark called Humanity’s Last Exam (HLE), which consists of 2,500 expert-curated questions across subjects. On this benchmark, most models previously achieved single-digit accuracy. Grok-4, however, solved a quarter of the HLE problems without using tools. With tool capabilities, the multi-agent version, Grok-4 Heavy, was able to solve over 50% of the text-only subset of the HLE problems. Moreover, according to Artificial Analysis, Grok-4 is now the leading AI model. “We have run our full suite of benchmarks, and Grok-4 achieves an Artificial Analysis Intelligence Index of 73, ahead of OpenAI o3 at 70, Google Gemini 2.5 Pro at 70, Anthropic Claude 4 Opus at 64, and DeepSeek R1 0528 at 68,” the company revealed. Training and Computational Scale xAI’s team revealed that the development of Grok-4 involved a significant increase in training compute. “From Grok-2 to Grok-3 to Grok-4, we’ve increased the training by an order of magnitude in each case. It’s 100 times more training than Grok-2,” Yuhuai (Tony) Wu, co-founder of xAI, said. Wu described two types of training compute—pretraining and reinforcement learning (RL). “We’re actually putting a lot of compute in reasoning, in RL,” Wu said. “With verifiable outcome rewards, you can train these models to think from first principles.” He also pointed to a key bottleneck going forward: the availability of challenging problems for RL. “As the model gets smarter, the number of useful RL problems reduces. We need a reliable signal to tell the model when it’s right or wrong,” Wu added. For Grok-4, all two lakh GPUs of xAI’s supercomputer Colossus were utilised for RL training, providing 10 times more compute than any other model in reinforcement learning at an unprecedented scale. “With Grok-4, RL is the new pre-training,” Jiayi Pan, xAI team member, wrote in a post on X. Grok4 same compute amount on pre training as reinforcement learning RL is all the rage it’s just wild to see the shift pic.twitter.com\/QOp1OAipys— Tommy (@Shaughnessy119) July 10, 2025 From Simulations to Reality Musk predicted future scenarios in which Grok-4 and its successors would operate not only in text but also in the physical world. “The ultimate reasoning test is reality,” he said. “You invent a new technology, does it work? Does the rocket get to orbit? Does the car drive? Does the medicine work?” Moreover, Musk discussed integration with robotics. “Combine Grok with Optimus, and it can interact with the world. Formulate hypotheses, test them, confirm or reject.” Grok 4’s practical utility was demonstrated across domains. In collaboration with Andon Labs, the model was tested on Vending-Bench, a simulation involving inventory management, pricing, and supplier contracts. Grok 4 became the new leader on the leaderboard, achieving double the net worth of other models. In biomedical research, the Arc Institute used Grok 4 to sift through experiment logs and propose hypotheses within seconds. The model is already being evaluated for CRISPR (clustered regularly interspaced short palindromic repeats) research and has been found useful in examining medical imaging, like chest X-rays. In finance, Grok 4 was described as capable of pulling real-time data and supporting decision-making. Voice Mode Introduced with New Natural Voices Grok 4 also features updates to voice capabilities, offering reduced latency and a new set of voices. One standout is Sal, which comes with a deep voice reminiscent of a movie trailer narrator. Another is Eve, which has been described as a “British voice capable of rich emotions”. During the livestream, Eve performed a live demo with operatic poetry on Diet Coke and a call-and-response number repetition game. Compared to OpenAI’s voice mode, Grok Voice was “snappier” and avoided interruption. According to xAI, five voice options are now available, and Grok Voice has seen a 10 times increase in active users in eight weeks. Upcoming Improvements and Models xAI showcased how Grok 4 can help solo developers build games. Danny, a game designer, used the API to create a first-person shooter in four hours. Grok 4 sourced assets, generated textures, and assisted with design, removing the need for external sourcing. Musk said the goal is for Grok to eventually “play the game” and assess whether it’s fun, a task that requires video understanding and tool integration with platforms like Unreal Engine. The team acknowledged that the current multimodal performance has limitations. “It was so bad that Grok was effectively looking at the world squinting through glass.” Improvements in image, video, and audio understanding are scheduled for the next foundation model, which is expected to finish training this month. The next steps include video generation, with xAI preparing to train a video model using more than one lakh GPUs. The purpose is to create infinite scroll content that users can watch and edit, with full interactivity. Moreover, Musk shared that xAI is developing a coding model focused on being “both fast and smart”, which is expected to be released in a few weeks.","excerpt":"“Grok-4 is smarter than almost all graduate students in all disciplines simultaneously.”","categories":["Global Tech"],"tags":["XAI"],"author_name":"Siddharth Jindal","publish_date":"2025-07-10T17:34:16","publication_year":"2025","word_count":1235,"keywords":["Anthropic","OpenAI","AI","ML","Claude 4","DeepSeek R1","XAI","Ray","RAG","xAI","Gemini 2.5"],"extracted_tech_keywords":["AI","ML","OpenAI","Claude 4","Anthropic","Gemini 2.5","DeepSeek R1","xAI","Ray","RAG"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/musks-grok-4-crushes-benchmarks-beats-openai-google-in-rl\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":24727,"title":"General Data Protection Regulation: What will happen to all the micro level analytics?","content":"The recent episode involving Facebook and Cambridge Analytica raised data privacy concerns once again. This resulted in the closure of Cambridge Analytica, once a high-flying consulting company. They, allegedly, stole user data from Facebook to micro analyze the profiles of individuals. Based on this micro analysis, Cambridge Analytica advised political parties on campaigning in elections. Facebook CEO Mark Zuckerberg also had to face some very tough questions at the Congressional hearing as a fall out of the above episode. On the face of it, data theft notwithstanding, this is a practice quite rampant in today’s digital world. Companies have terabytes of data which they analyze to promote their products and services by targeting the right audience, whether the consumer likes it or not. With this background, the General Data Protection Regulation (GDPR) formulated by the European Union (EU), is coming into full effect from May 25, 2018, with the objective to protect data at the individual\/consumer level. According to this regulation, the control of the data resides with the individuals. If they decide not to share their data, companies can’t use that data. Non-compliance or breach would result in huge fines. Some of the important points covered under GDPR are as follows: – There is a penalty of €20 million or 4% of worldwide revenue for non-compliance. Consumers should opt-in for consent. GPS locations are also included in the definition of personal data. Now, considering most of the global companies have huge analytics departments filled with data scientists, many of whom are PhDs, consumer data is their stock in trade. All their algorithms need copious amounts of data to function well. A lot of marketing programs use the individual level data to design marketing campaigns, both online and offline. When someone searches for a book online, for example, she gets a list of similar books she could buy. Now, if the individual doesn’t want to share this data with the company, the company will have to delete this information about the individual. The individual has a right to be forgotten as per the regulation. If a majority of individuals decide not to share their data, the marketing campaigns and the analytics engines would all go for a toss. Some other global companies have outsourced their analytics work to niche vendors. They would also face the same data problem. This regulation applies to any company providing services to EU citizens irrespective of whether the company has a physical presence in the EU or not. The companies will have to find an alternative approach to deal with consumers. They may not have sufficient time though. As far as GDPR compliance goes, only 7% companies are fully compliant as the May 25th deadline approaches. As per a Crowd Research report, very few companies (only 40%) are hoping to be compliant by the deadline (https:\/\/www.zdnet.com\/article\/gdpr-compliance-for-many-companies-it-might-be-time-to-panic\/). This becomes a tricky situation for the companies even before the analytics experts ask for individual data. If the individuals raise a complaint after the deadline that their data was misused, companies are bound to pay a hefty fine. What about the individual level data that is already stored in company servers and on the cloud? It will take a huge effort and investment on the part of the companies to make that data safe and reach compliance. This applies to machine learning algorithms as well. The whole premise of machine learning is that new data should be continuously fed to the algorithms to make adjustments and provide better predictions. Autonomous cars, for example, use image processing and they do need human images. Imagine a scenario where human beings refuse to share their images. Of course, the data scientists will have images of animals, trees and other objects to feed into their algorithms but this won’t prevent the autonomous vehicles from bumping into human beings leading to dire consequences. We have already had a case in the U.S. where an autonomous car hit a human being resulting in death. With the latest voice recognition technology and the advent of devices like Amazon’s Echo and Google’s Home, individuals are supposed to feel more comfortable with less friction in their lives. For this to happen, the speaker has to be ‘ON’ all the time and it will record whatever is spoken at home leading to dire consequences. This data will also be stored somewhere on the cloud. Sooner rather than later, individuals would not want their data to be shared between devices, going back to the earlier era of zero connectivity. It is still to be seen how all this plays out. Companies are concerned that GDPR compliance will make the jobs of their employees more cumbersome and would make it very difficult for them to do business. They are hoping fines won’t be levied so soon after all and the deadlines might get extended. Whatever be the situation on May 25th, 2018, data privacy concerns are going to get shriller. Companies need to be ready with a ‘Plan B’, as far as their analytics strategies are concerned. Authors Sanjay Fuloria is General Manager at Cognizant Technology Solutions and Nupur Pavan Bang is Associate Director, Thomas Schmidheiny Centre for Family Enterprise, Indian School of Business.","excerpt":"With this background, the General Data Protection Regulation (GDPR) formulated by the European Union (EU), is coming into full effect from May 25, 2018, with the objective to protect data at the individual\/consumer level. According to this regulation, the control of the data resides with the individuals. If they decide not to share their data, […]","categories":["IT Services"],"tags":[],"author_name":"AIM Media House","publish_date":"2018-05-19T07:00:15","publication_year":"2018","word_count":864,"keywords":["Go","machine learning","programming_languages:R","AI","programming_languages:Go","Git","ViT","analytics","R"],"extracted_tech_keywords":["AI","machine learning","analytics","R","Go","Git","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/general-data-protection-regulation-what-will-happen-to-all-the-micro-level-analytics\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10094377,"title":"OpenAI Wants You to Fix Their Cybersecurity Problems","content":"Continuing on the heels of offering $1M grants for AI democratisation, OpenAI is back again with another million grant – this time for cybersecurity. OpenAI has launched their new Cybersecurity Grant Program which aims to help and enable the creation and advancement of AI-powered cybersecurity tools and technologies. With a series of announcements inviting people’s contribution, OpenAI is going down the route of open source without open sourcing their product. Announcing the Cybersecurity Grant Program — a $1M initiative to boost and quantify AI-powered cybersecurity capabilities and to foster high-level AI and cybersecurity discourse: https:\/\/t.co\/kqHXqxAks1— OpenAI (@OpenAI) June 1, 2023 In a move to unite the ‘defenders’ to form a robust system to counter cyber threats, the Cybersecurity Grant Program aims to work with individuals and organisations who can help propose such a system. By doing so, they believe that the power of cybersecurity moves from attackers to defenders. The program aims to empower defenders by giving them access to the latest and most advanced AI capabilities, and in the process develop methods and metrics to assess the capabilities of AI-based cybersecurity systems. Reiterating Initiatives This is not the first time OpenAI has invited people to help with their cybersecurity features. In April, the company ran a Bug Bounty Program that invited people to report vulnerabilities, bugs and security flaws in their ChatGPT model. The company specifically called out security researchers, ethical hackers and technology enthusiasts to come forward to participate. However, the reward money for the program ranged from $200 to $20,000 depending on the severity of the discovery. OpenAI has also mentioned that CyberSecurity Program shall be ‘licensed or distributed for maximal public benefit and sharing.’ This could be hinting at a future where they might offer these learnings to the broader community in any way they deem necessary. It is interesting to note that OpenAI is choosing a smart system of not spending resources internally to come up with mechanisms for building a secure system instead choosing an open-source- like method. By crowdsourcing crucial elements such as AI regulation frameworks and cybersecurity threats, the company is likely to seem compliant and adhering to people for their greater good. All these actions are taking form after the recent Senate hearings. The company was also rumoured to start their own open source model, however there has been no details on the same.","excerpt":"OpenAI announces $1M grant for their Cybersecurity Grant Program : another route to open-sourcing.","categories":["Deep Tech"],"tags":["ChatGPT","cybersecurity threats","Open Source","OpenAI"],"author_name":"Vandana Nair","publish_date":"2023-06-02T14:55:00","publication_year":"2023","word_count":393,"keywords":["Go","ChatGPT","cybersecurity threats","Open Source","OpenAI","AI","AWS","GPT","Aim","ViT","GAN","R"],"extracted_tech_keywords":["AI","ChatGPT","OpenAI","Aim","AWS","R","Go","GPT","GAN","ViT"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/openai-wants-you-to-fix-their-cybersecurity-problems\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10055873,"title":"How The Autonomous Vehicle Industry Shaped Up In 2021","content":"Like all others, the COVID-19 pandemic affected the autonomous vehicles (AV) industry just as much. China, forecasted to be the world’s largest AV market, saw a dip in its AV sales by 71 per cent at the beginning of the pandemic in February 2020. In other major markets, the situation was similar. Europe saw an 80 per cent fall and the US by 47 per cent. However, production rolled back to its earlier numbers, sometimes even surpassing the production levels from before the first quarter of 2020. At present, most of the AVs available in the market are in Level 2 and 3, meaning they have systems including the likes of lane departure warning, collision detection, and cruise control present in these cars. However, it is not too long that AVs built to the Level 4 and 5 make it to the markets. In fact, according to a market research report by Facts and Factors, the AV cars industry was valued at $23.33 billion last year and is expected to grow at a CAGR of 22.7 per cent to reach $64.88 billion by 2026. Today, Analytics India Magazine has curated a roundup of all the top headline makers from the AV industry. Toyota acquired the AV division of Lyft Toyota Motor Corp acquired the AV division of US-based company Lyft for $550 million. The former paid Lyft $200 million upfront, and the remaining $350 million will be paid over the next five years. Additionally, Toyota also signed a commercial agreement to use the system and fleet data of Lyft. Lyft started operations of its AV unit in 2017, launching the Level 5 division. Bullish about the growing AV industry, Toyota has earlier invested in the US and China-based self-driving startup Pony.ai. It has also developed a driver-assistant system Guardian and self-driving software Chauffeur. Ati Motors raises $3.5M Electric AV-maker Ati Motors raised $3.5 million in its pre-Series A round of investment from Blume Ventures and Exfinity Venture Partners. The Bengaluru-based startup was founded in 2017 by V Vinay, Saurabh Chandra and Saad Nasser. The startup planned to utilise the freshly raised funds to manufacture and deploy a large fleet of AVs in warehouses and factories in both India and abroad. MIT researchers develop DNN for AVs MIT researchers have developed a single deep neural network (DNN) for AVs, using NVIDIA DRIVE AGX Pegasus. The paper mentions how the researchers are using a new self-driving technique using the single DNN by processing real-time lidar sensor data. The researchers have also created new enhancements to increase speed and energy efficiencies. The DNN is built with the aim of executing all the operations of the self-driving system. Still in its nascent stage, the approach has the potential to generate considerable benefits. Ohio State Center Researchers develop AV Cybersecurity platform Ohio State CAR, or Center for Automotive Research, announced that they were developing a dedicated platform for cybersecurity testing for self-driving vehicles or Mobility Cyber Range (MCR). The initial focus of the research pilot was said to be on developing standards and recommendations for AV safety and cybersecurity’s best practices. The technology will be used in connected and autonomous vehicles applications. CAR was also training its students on AI computing by undertaking the development work on NVIDIA. Arm’s new open-source architecture and tools Semiconductor and software design company Arm unveiled its hardware tools for automakers and chipmakers. Its new software architecture and reference implementation– SOAFEE, was released along with two reference hardware solutions. SOAFEE is designed to bring in the real-time safety needs of automotive along with the pros of a cloud-native approach. AWS announced AWS IoT FleetWise At the Amazon Web Services (AWS) re:Invent event, AWS announced the AWS IoT FleetWise, a cost-effective and easier service for automakers to collect, transform and transfer vehicle data into the cloud in real-time. With the help of this service, automakers can collect and organise data in any format available in their vehicles, standardising them for data analysis in the cloud. The service helps transfer real-time parameters like weather conditions, vehicle type and location. Additionally, it can be used to diagnose issues, analyse vehicle fleet health and help reduce potential safety issues.","excerpt":"The AV cars industry is expected to reach $64.88 billion by 2026.","categories":["AI Features"],"tags":["Autonomous Vehicles"],"author_name":"Debolina Biswas","publish_date":"2021-12-14T12:00:00","publication_year":"2021","word_count":695,"keywords":["AWS","AI","neural network","RPA","cloud_platforms:AWS","Aim","analytics","GAN","Autonomous Vehicles","R","startup"],"extracted_tech_keywords":["AI","neural network","analytics","Aim","AWS","R","GAN","RPA","startup","cloud_platforms:AWS"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-the-autonomous-vehicle-industry-shaped-up-in-2021\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10041197,"title":"8 AI Startups That Raised Funds In 2021","content":"Recently, Indian Government has launched a Startup India Seed Fund Scheme with a corpus of 945 crore rupees to provide seed funding to qualified startups through eligible incubators across India in the next four years. In other news, Indian AI startups raised almost $836.3 million in 2020, as per Startup Funding Report. Pandemic notwithstanding, the country’s startup ecosystem is indeed going places. Below, we look at AI startups that got funding in the last few months. (The list is in alphabetical order) 1| Cron AI The edge inference platform–founded in 2015 by Tushar Chabbra and Saurav Agarwala– has raised $4 million from VenturEast and Kitaki Ventures in a Series A round. Cron AI has developed a real-time AI, Deep Learning-enabled edge inferencing platform accelerating algorithms and neural networks on the product edge. Cron’s unique real-time scalable and dynamic hardware and software architecture can be used to develop intelligent solutions using 3D sensors with applications in autonomous vehicles, mobility, logistics, robotics, intelligent transport infrastructure, smart spaces etc. 2| Elucidata Founded in 2015 by Abhishek Jha, Swetabh Pathak, and Dr Richard Kibbey, bioinformatics platform Elucidata has raised $5 million in pre-Series A led by IvyCap Ventures with participation from Hyperplane Venture Capital and a few angel investors. The company provides a code-first SaaS platform for data-driven drug discovery teams that merges proprietary and public biomedical molecular data for machine learning applications. 3| Miko The robotics startup–founded in 2015 by Sneh R Vaswani, Prashant V Iyengar, and Chintan S Raikar – raised $6.7 million in a Series B led by IvyCap Ventures. Existing investors Chiratae Ventures, YourNest Capital, and former chairman of Nasscom, Keshav Murugesh, participated in the funding round. The Miko series of robots cater to the learning and development needs of children. The company leverages proprietary and state of the art artificial intelligence, robotics and internet of things and aspires to bring their platform solutions to 20 million homes by 2023. 4| Myelin Foundry Founded in 2019 by Gopichand Katragadda – the deep tech startup has raised $1 million in pre-series A led by Beyond Next Ventures, a Japanese venture capital firm. Myelin Foundry is transforming human experiences and industry outcomes by building Artificial Intelligence algorithms on video, voice, and sensor data, for edge devices. Myelin’s first product, Fovea Stream, enables HD viewing experiences with SD transmission. “Myelin has developed globally leading products and capability in Edge AI. We look forward to partnering with BNV and accessing the Japanese ecosystem,” said Gopichand Katragadda, the Founder and CEO of Myelin. The fresh funds will be used to consolidate Myelin’s products in the India video streaming market and to enter international markets. 5| Nextbillion.ai Nextbillion.ai–founded in 2019 by Gaurav Bubna, Ajay Bulusu and Shaolin Zheng – has raised $6.25 million as part of its extended Series A fundraising round from Microsoft’s venture fund, M12. The startup provides tailor-made solutions addressing the unique map-related requirements for businesses. To create a customised map, the company uses advanced AI to unlock hidden intelligence in business data. Map editors assist in curating the custom map and fine-tuning it to meet corporate requirements. 6| Obviously.ai The B2B predictive analytics startup–founded in 2019 by Nirman Dave and Tapojit Debnath Tapu – has raised a seed funding of $3.6 million from TMV (based in New York and Chicago), Capital Group, Golden Gate Ventures, Arka Venture Labs. Obviously.ai offers tools that enable non-technical business analysts to run predictions on their historical data. The startup will use the fresh capital to grow R&D, expand sales and marketing teams, introduce additional time series and other cutting edge algorithms and build an integration ecosystem that can further automate it’s predictive AI capabilities. 7| Vernacular.ai The AI-First SaaS startup has raised $5.1 million in Series A led by Exfinity Ventures and Kalaari Capital. Founded in 2016 by Sourabh Gupta and Akshay Deshraj, Vernacular.ai enhances customer experience through intelligent voice conversations. The startup’s flagship product, VIVA, is an intelligent, multilingual platform that uses NLU & speech recognition to automate up to 80% of call centre operations. 8| Wysa Healthtech startup Wysa–founded in 2015 by Ramakant Vempati and Jo Aggarwal – has raised $5.5 million in Series A led by W Health Venture. It’s a mental and emotional well-being Life Coach powered by AI. The service, launched in 2017, uses three techniques to deliver the early intervention to high-risk groups: An AI chatbotA library of evidence-based self-help materialsMessaging-based support from human therapists","excerpt":"Recently, Indian Government has launched a Startup India Seed Fund Scheme with a corpus of 945 crore rupees to provide seed funding to qualified startups through eligible incubators across India in the next four years. In other news, Indian AI startups raised almost $836.3 million in 2020, as per Startup Funding Report. Pandemic notwithstanding, the […]","categories":["AI Trends"],"tags":["AI Startups"],"author_name":"kumar Gandharv","publish_date":"2021-06-03T15:00:00","publication_year":"2021","word_count":735,"keywords":["machine learning","artificial intelligence","AI","neural network","R","RAG","deep learning","analytics","edge AI","predictive analytics","AI Startups"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","deep learning","neural network","analytics","edge AI","RAG","predictive analytics","R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/8-ai-startups-that-raised-funds-in-2021\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10094361,"title":"OpenAI Inches Closer to AGI, Reduces Hallucinations","content":"A Math teacher’s keen interest in checking the steps to solve a problem rather than the result loosely forms the basis for OpenAI’s new training approach. The company announced a new technique on training the model through process supervision by rewarding each step of correct reasoning as opposed to rewarding the correct final result via outcome supervision. Here the output would probably be a model with reduced hallucinations and higher alignment as claimed by the company. OpenAI specifically calls out mitigating hallucinations as a crucial step towards ‘building aligned AGI’, but would any of these new training methods inch them closer to an AGI status? Source: Twitter Hallucinations At Bay OpenAI talks about how you can train models to detect hallucinations either by process supervision, a method to provide feedback for each individual step, or outcome supervision, where the feedback is based on a final result. The company claims to have improved mathematical reasoning with the former method. By rewarding the model at each correct step, the model is said to mimic ‘human reasoning’ while solving a mathematical problem. With an emphasis on hallucinations, the company’s move towards ‘claiming’ to make the models more robust continues. Companies are actively working on reducing hallucinations. Recently, NVIDIA released NeMo Guardrails, an open-source toolkit that will help LLM-based applications become accurate, appropriate, and secure. With hallucinations considered a persisting problem with chatbots that often make them behave illogically by generating misinformation or biases, OpenAI is working on making their models become better. With the new training method, the company is hoping to keep a check on hallucinations as they believe a process-oriented method that involves feedback at each step, will control the irrational outputs generated by the chatbots. Alignment — Closer to AGI? really exciting process supervision result from our mathgen team. positive sign for alignment. https:\/\/t.co\/KS8JEtHVAt— Sam Altman (@sama) May 31, 2023 OpenAI’s reference to ‘building an aligned AGI’ is hinting at the company’s long-term plans for achieving it. Looking back, Sam Altman has made multiple mentions of AGI and how the future will look with it. A few months ago, he laid out an elaborate AGI roadmap for OpenAI where its dangers were called out. The company believed that AGI can be misused and lead to grave consequences in society. However, despite these risks, the potential and benefits of it are far-reaching therefore, the company will develop it in a ‘responsible way’. AI expert Gary Marcus predicts that AGI will not be coming soon. It is interesting to note that Altman’s stance on AGI and its development is not clear-cut. In yesterday’s tweet, Altman seemingly downplayed the risk of AGI by predicting how ‘a much faster rate of change’ is what AGI will bring in. He believes that with AGI the future will unfold similarly to that without it, and the difference will be the speed with which things unfold – “everything happens much faster”. Ironically, Sam Altman along with AI scientists, Geoffrey Hinton, Yoshua Bengio, and many others, signed a statement a few days ago, which stands for safeguarding against the threat of extinction posed by AI, and considers it on par with nuclear war. If any action is to be taken on this, then the question that would arise is: how far will OpenAI go to make more advanced models reach AGI? The recent statement is in continuation with the open letter that was signed by over 31k people two months ago, including Elon Musk, Gary Marcus and other tech experts, who were urging for a pause on advanced AI models, which, interestingly, was not signed by Sam Altman. Though Altman had confirmed a month ago that the company will not work on building their next superior model GPT-5, and instead focus on the safety features of their existing models, his constant sway in matters pertaining to AGI threats and downplaying its scope makes it difficult to gauge where the company is headed. The company, often criticised for data security threats and privacy concerns, is fighting hard to prove ChatGPT as a foolproof chatbot. The company is now working on democratising AI by offering grants to those who can propose the best method for creating an AI regulatory framework — again with the hope of improving the system and seeming compliant with the world.","excerpt":"OpenAI’s new process supervision training is said to improve math reasoning with human-like thinking, and reduce hallucinations. Is this a step closer to AGI?","categories":["Global Tech"],"tags":["AGI","ChatGPT","Data Security","gary marcus","Geoffrey Hinton","GPT-5","hallucinations","NVIDIA","OpenAI","Sam Altman","Yoshua Bengio"],"author_name":"Vandana Nair","publish_date":"2023-06-02T13:00:00","publication_year":"2023","word_count":715,"keywords":["Go","ChatGPT","Sam Altman","TPU","GPT-5","Geoffrey Hinton","Data Security","OpenAI","AI","chatbots","hallucinations","GPT","gary marcus","Yoshua Bengio","Aim","AGI","NVIDIA","R"],"extracted_tech_keywords":["AI","GPT-5","ChatGPT","OpenAI","Aim","chatbots","TPU","R","Go","GPT"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/openai-inches-closer-to-agi-reduces-hallucinations\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10093163,"title":"Indian Manufacturers Trail in Digital Transformation Despite Investment","content":"For all the cloud that perception creates, numbers always tell the real story. According to the 8th annual ‘State of Smart Manufacturing Report’ by Rockwell Automation released in March this year, India has the largest number of manufacturing firms investing in technology. The survey was conducted across 1,350 manufacturers in 13 of the major manufacturing countries including India, China, US, Germany, Japan and the UK. The report stated that Indian manufacturers were spending 35% of their operating budgets in technology, far more than the global mean of 23% investment. India leads investment in manufacturing But the money hasn’t just translated into tangible results as of yet. India is still lagging behind in digital maturity. A report by Lenovo released just last week found that 48% of Indian businesses are still clung to the first stage of digital maturity. While 87% of businesses believe digital infrastructure is critical for them to make money, only 33% of Indian companies are adequately prepared. For companies to reach stage 3 or 4 of digital maturity, their IT teams must have moved to the hybrid cloud. And yet falls behind in digital transformation? But it  would seem as though Indian companies have long existed in this state of limbo. Arjun Bajaj, Director of smart TV manufacturer, Videotex spoke that a big chunk of the companies were still plagued by age-old issues. “Lack of digital infrastructure, knowledge and expertise are significant obstacles preventing Indian businesses from being digitally mature. Progress is also hampered by out-of-date hardware and software and cybersecurity risks,” he said. He admitted that money was being spent in the right direction, there were still gaps. “Although ongoing spending is crucial, it is not the only factor influencing digital maturity. We believe that manufacturers should invest in digital skills training, infrastructure improvements and increased awareness of the advantages of digital transformation in order to assist more Indian firms in reaching higher levels of digital maturity and competing in the global economy. Collaboration between businesses, governmental organisations and educational institutions is needed to get past these obstacles,” he stated. But from what Bajaj described, cloud transformation was still a distant reality for most companies here. “Adopting cloud-based technologies and software solutions first will increase operational agility and efficiency. Businesses should also think about partnering with knowledgeable technology providers who can create solutions specifically for them. They should also continuously evaluate their progress towards achieving their objectives for digital transformation and develop a clear digital strategy that is in line with their business goals,” he noted. Aniruddha Banerjee, co-founder of AI services startup SwitchOn explained that the issues had more to do with mindset of company leaders and could be divided cleanly into three parts. “Firstly, most company leaders face the issue of short-term thinking. Most of the objectives are driven with short-term outlook and rarely lead to organisational wide changes due to obstacles with scale. Companies should instead focus on the long term business outcomes rather than just trying to ‘just do something. “Secondly, there’s a constant reliability on global products. Products that succeed globally rarely succeed in India due to a lack of Indian context. This is almost never understood well. Companies need to trust products that have worked in the Indian context and have been built in India. They know Indian users the best. “And thirdly, Indian companies have an unnecessary focus on minor details rather than overall strategy when digitising. This leads to huge adoption time and a lack of interest or focus. Companies should look at the big picture and minor issues will resolve themselves automatically,” he explained. Baby steps in Smart Manufacturing But even as India is in the throes of these middling changes, the concept of smart manufacturing has walked in through the doors. Smart manufacturing is a full realisation of the digital dream in manufacturing. Smart factories have advanced sensors, embedded software (like IoT, cloud computing, analytics) and robotics for optimising better decision-making. The goal of smart manufacturing is to push automation, have predictive maintenance and address customer concerns in real-time. The making of India’s first smart factory, backed by Boeing, is still in progress at the Indian Institute of Science’s (IISc) at Bangalore’s Product Design and Manufacturing (CPDM) Center. Despite the massive push in India, there are many lessons still to be learned from countries like Japan and South Korea which are far ahead on their journey of modernising manufacturing. “We need to learn from Japan. This country literally transformed the way of working of the industries and the way of doing business. The Academia- institution- industry collaboration in Japan helped in achieving this. It’s time India too went for such collaborative efforts, which will help us grow fast,” Shyam Singh, Tata Motors Plant Head said at his inaugural speech for a week-long Faculty Development Program on smart manufacturing at the DY Patil International University, Pune. Raghav Gupta, co-founder and CEO of Futurense Technologies reiterated this in a panel discussion at the DES conference held this year. For smart manufacturing to become a solid reality, India too would have to start young. “The challenge is that the number of those people are firstly very low. And secondly, when it comes to re-training the people from the industry, the challenge is un-learning and then learning. There has to be an understanding that education needs to move towards domain-specific courses from a very young age,” Gupta stated.","excerpt":"While 87% of businesses believe digital infrastructure is critical for them to make money, only 33% of Indian companies are adequately prepared.","categories":["IT Services"],"tags":[],"author_name":"Poulomi Chatterjee","publish_date":"2023-05-11T15:53:08","publication_year":"2023","word_count":898,"keywords":["Go","cloud computing","AI","digital transformation","Git","RAG","analytics","Rust","GAN","R"],"extracted_tech_keywords":["AI","analytics","RAG","cloud computing","R","Go","Rust","Git","GAN","digital transformation"],"url":"https:\/\/analyticsindiamag.com\/it-services\/indian-manufacturers-trail-in-digital-transformation-despite-investment\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10067769,"title":"Turing Test is unreliable. The Winograd Schema is obsolete. Coffee is the answer.","content":"Today, AI can do many things: GPT-3 can produce human-like text, DALL-E can generate the most imaginative images based on textual prompts, and Alexa can turn off lights at your behest, yet we are far from achieving artificial general intelligence (AGI). For starters, we still dont have an agreed-upon definition of AGI. Even the terminology is debated: There is no such thing as artificial general intelligence because there is no such thing as general intelligence. Human intelligence is very specialised, said Meta’s chief AI scientist Yann LeCun. So how do we measure intelligence in machines? And more importantly, how accurate are such tests. Turing test Alan Turing proposed the Turing test in a 1950 paper called “Computing Machinery and Intelligence.” He suggested ‘Imitation Game’ with two contestants– a human and a computer. A judge has to decide which of the two contestants is human and which is a machine. The judge would do this by asking a series of questions to the contestants. The game aimed to identify if the computer is a good simulation of humans and is, therefore, intelligent. At the heart of the Turing test is the question: “Are there imaginable computers which would do well in the imitation game?” While no machine has passed the Turing Test yet, a few have come close. In 2014, a program named Eugene Goostman convinced a third of a panel of judges that it was a 13-year-old boy from Ukraine. “Our main idea was that he can claim that he knows anything, but his age also makes it perfectly reasonable that he doesn’t know everything,” said Veselov, one of the programmers of Gustman. But Goostman’s feat was more imitation and diversion than showing real intelligence. Eugene either completely avoided certain topics or deflected when presented with a question it had no answer to. For instance, when asked if he plays many instruments, Eugene said, ‘I’m tone-deaf, but my guinea pig likes to squeal Beethoven’s Ode to Joy every morning. I suspect our neighbours want to cut his throat. Could you tell me about your job, by the way?’ The program could not solve logical problems like a real 13 year old could. “It would respond with wisecracks to evade revealing its limitations, and to the untrained eye, it was fairly convincing. All that tells us is that human beings think that machines that can talk are intelligent, but it turned out to be untrue”Gary Marcus Marcus said the Turing test is not a reliable measure of intelligence because humans are susceptible, and machines can be evasive. Philosopher John Searle introduced the Chinese Room Argument that asserts programming a digital computer may make it appear to understand the language but could not produce real understanding. Even if a computer can interpret symbols and provide sensical responses, it can’t be said to be truly “conscious” because it doesn’t really understand what the symbols mean. The Winograd schema Hector Levesque, a computer scientist at the University of Toronto, proposed the Winograd schema challenge in 2011. Hector designed it as an improvement of the Turing test. The test is structured with multiple-choice questions called Winograd schemas. Winograd schemas were named after Terry Winograd, professor of computer science at Stanford University. It is a pair of sentences whose intended meaning can be flipped by changing just one word. They generally involve unclear pronouns or possessives. “The city councilmen refused the demonstrators a permit because they [feared\/advocated] violence.” There’s a verb choice quiz embedded in the sentence, and the task for System A is to select the right one. If System A has common sense, the answer is obvious enough. For instance, the system could be asked ‘who feared violence’ and would have to choose between the city councilmen or the demonstrators. Human beings can easily answer this question. But computers are still struggling to make such connections. In the book ‘The Myth of Artificial Intelligence, AI researcher Erik J Larson said linguistic puzzles that humans easily understand are still beyond the comprehension of computers. For example, even single sentence Winograd schemas trip up machines. In a test, Gary Marcus asked a Winograd-Levesque-inspired question: “Can an alligator run the 100-meter hurdles?” and AI systems struggled to come up with an answer. According to Levesque, the schema should meet two criteria: simple for humans to solve and shouldn’t be Google-hackable. He also explained how the Winograd schema test could be better than a Turing test. “A machine should be able to show us that it is thinking without having to pretend to be somebody,” he wrote in his paper. “Our WS challenge does not allow a subject to hide behind a smokescreen of verbal tricks, playfulness, or canned responses.” And, unlike the Turing test, which is scored by a panel of human judges, a Winograd schema test’s grading is completely non-subjective. However, in 2022, the test developers published a paper titled, ‘The Defeat of the Winograd Schema Challenge, claiming most of the Winograd Schema Challenge has been overcome. Similarly, a 2021 paper, ‘WinoGrande: An Adversarial Winograd Schema Challenge at Scale’, shows how neural language models have saturated benchmarks like the WSC, with over 90% accuracy. The researchers asked, “Have neural language models successfully acquired commonsense or are we overestimating the true capabilities of machine commonsense?” Coffee test Apple co-founder Steve Wozniak suggested the coffee test, whereby a robot would be challenged to enter your home, find the kitchen and brew a cup of coffee. The programme should be able to walk into any kitchen, find the ingredients required and then perform the task of making a coffee. According to Wozniak, the day a robot could enter a strange house and make a decent cup of coffee would be the day AI has truly arrived. To crack the coffee test, a robot has to be multi-modal, able to generalise across tasks and orchestrate a series of actions to make a hot cup of coffee. Cheeky as it sounds, the coffee test seems like a plausible test to judge the AGI-ness of machines.","excerpt":"Gary Marcus said the Turing test is not a reliable measure of intelligence in machines.","categories":["IT Services"],"tags":["Turing test"],"author_name":"Avi Gopani","publish_date":"2022-05-25T11:00:00","publication_year":"2022","word_count":1004,"keywords":["Go","DALL-E","artificial intelligence","AI","Modal","Turing test","Git","GPT","Aim","AI research","R"],"extracted_tech_keywords":["AI","artificial intelligence","Aim","R","Go","Git","GPT","DALL-E","Modal","AI research"],"url":"https:\/\/analyticsindiamag.com\/it-services\/turing-test-is-unreliable-the-winograd-schema-is-obsolete-coffee-is-the-answer\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10049195,"title":"Prakash Mallya","content":"Prakash Mallya is VP & MD – Sales, Marketing and Communications Group, Intel India. He is responsible for the growth of Intel’s India business through local and multinational original equipment manufacturers (OEMs), global and local partners, retailers, and end-users. His focus is also to drive strategic partnerships with the Government and ecosystem players to accelerate technology adoption in India and market expansion to grow local consumption. He was most recently the Sales Director for Intel Asia’s Data Center Group. He managed the region’s data centre strategy for cloud service providers, communications service providers, enterprise, and government infrastructure, spanning server, storage and network solutions. Before that, Prakash achieved strong business results as the Managing Director of Southeast Asia. He led all of Intel’s sales, marketing and the enablement of products in one of the world’s most dynamic markets for computing technology. Under his stewardship, the Southeast Asia team also won the Global ECOC Award for exemplary leadership in Ethics and Compliance. In the previous role, Prakash was the Country Manager of Malaysia and Singapore, where he led the sales and marketing operations. Prakash holds a bachelor’s degree in Electrical and Electronics Engineering and an MBA in Marketing Management. Travel, photography, music and fitness are his passions. LinkedIn","excerpt":"Prakash Mallya is VP & MD – Sales, Marketing and Communications Group, Intel India. He is responsible for the growth of Intel’s India business through local and multinational original equipment manufacturers (OEMs), global and local partners, retailers, and end-users. His focus is also to drive strategic partnerships with the Government and ecosystem players to accelerate […]","categories":["AI Features"],"tags":["Intel","Interviews and Discussions"],"author_name":"AIM Media House","publish_date":"2021-09-22T10:51:49","publication_year":"2021","word_count":206,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","RAG","R","Intel","Interviews and Discussions"],"extracted_tech_keywords":["AI","RAG","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/prakash-mallya\/","complexity_score":2,"technical_depth":6,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":60669,"title":"Great Learning Launches ‘Great Learning Academy’ — An Online Library On AI &#038; ML","content":"Great Learning has recently launched ‘Great Learning Academy’, a digital library offering over 300 hours of structured learning content across 40 courses focused on critical career skills. The Great Learning Academy has been designed to offer a wide range of courses including analytics, programming, data science, artificial intelligence, machine learning, cloud computing, digital marketing, big data and business finance. It offers beginner and intermediate level courses in the aforementioned domains combined with industry case studies, weekly live sessions, and career preparation material. Developing Critical Career Skills According to the release by the Great Learning — LinkedIn’s Emerging Jobs Report for 2020 puts artificial intelligence specialist and data scientist at the top of the emerging jobs list, with over 74% and 34% hiring growth over the past four years. Both of these fields come with a relatively high bar of entry and their own specialised skills clusters. This indicates that the idea of upskilling in new-age skills has not yet penetrated deep enough when it comes to college students and working professionals. Great Learning claimed that with this initiative, it is trying to change the idea of upskilling. This is being done by incorporating courses that cover the conceptual understanding of the topic, offering students a 360-degree view of each subject they choose to study. Mohan Lakhamraju, the founder and CEO of Great Learning stated that the purpose of Great Learning Academy is to offer everyone the opportunity to develop critical career skills relevant to this digital age. He further added that the online repository will provide access to courses developed and delivered by award-winning faculty members. He said, “This initiative is an integral part of Great Learning’s belief that everyone willing to put in the effort needed to learn should have access to the best learning opportunities possible regardless of their economic situation.” The courses could be taken by both working professionals as well as students.","excerpt":"Great Learning has recently launched ‘Great Learning Academy’, a digital library offering over 300 hours of structured learning content across 40 courses focused on critical career skills. The Great Learning Academy has been designed to offer a wide range of courses including analytics, programming, data science, artificial intelligence, machine learning, cloud computing, digital marketing, big […]","categories":["AI News"],"tags":["Artificial Intelligence India","Great learning","learn ai"],"author_name":"Sejuti Das","publish_date":"2020-04-01T18:30:00","publication_year":"2020","word_count":316,"keywords":["big data","data science","artificial intelligence","machine learning","AI","cloud computing","Artificial Intelligence India","Git","Aim","analytics","Great learning","learn ai","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","data science","analytics","Aim","cloud computing","R","Git","big data"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/great-learning-launches-great-learning-academy-an-online-library-on-ai-ml\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10121369,"title":"OpenAI Confirms that Sky’s Voice Actress was Cast Months before Altman Contacted Johansson","content":"As per a recent report, OpenAI chose the voice actor for Sky months before OpenAI CEO Sam Altman contacted Hollywood star Scarlett Johansson, and it had no intention of finding someone who sounded like her. Moreover, the name Sky was selected to convey a calm, breezy, and pleasing tone and had no relation to the Hollywood actor. Giving an explanation of the voice selection process, OpenAI published a casting call for a top-secret initiative to provide ChatGPT with a human voice last year. The flier included other requests, including that the actors sound between 25 and 45 years old and be nonunion. The report said nowhere had OpenAI said it was searching for a voice akin to Scarlett Johansson. Subsequently, Sky’s voice was created by hiring an actress. Based on her initial voice test clips, OpenAI determined that the actress’s natural voice sounded like Sky’s, so they didn’t alter her recordings to sound like Johansson. Controversy arose on Monday after the actress, who had initially declined Sam Altman’s offer to voice the ChatGPT, expressed her shock and disbelief. She was taken aback when OpenAI launched a chatbot with a voice that she found eerily similar to hers. “He told me that by voicing the system, I could bridge the gap between tech companies and creatives and help consumers feel comfortable with the seismic shift concerning humans and AI. He felt my voice would comfort people,” the Jojo Rabbit star stated. After nine months, she said, everyone, including friends, family and the general public, noticed how much the newest system. Sky sounded like her. Following the uproar, OpenAI declared that it would stop using the voice, emphasizing that the voice match was purely coincidental. In a blog post, the company clarified that the intention was never to mimic Scarlett Johansson’s voice, but rather to create a calm and pleasing tone for the AI. Sam Altman, in a statement, reiterated that the voice of Sky was never intended to resemble Scarlett Johansson’s. He emphasized OpenAI’s respect for Ms. Johansson and their decision to pause using Sky’s voice in their products. He also expressed regret for any miscommunication that may have occurred. OpenAI Problems Continue While OpenAI is still battling out this voice impersonation case, the company is tangled with a number of issues including the exits of two high-profile executives, Ilya Sutskever and Jan Leike, who played an important role in leading the company’s now-dissolved safety research team, Superalignment. The departures have revived the questions about the company’s approach to balancing speed versus safety in developing its AI products.","excerpt":"OpenAI claims that they never intended to mimic Scarlett Johansson’s voice","categories":["AI News"],"tags":["ChatGPT","OpenAI","Sam Altman"],"author_name":"Anshul Vipat","publish_date":"2024-05-23T16:11:45","publication_year":"2024","word_count":427,"keywords":["ChatGPT","Sam Altman","ELT","OpenAI","AI","programming_languages:R","GPT","CLIP","R","llm_models:ChatGPT","llm_models:GPT"],"extracted_tech_keywords":["AI","ChatGPT","OpenAI","R","ELT","GPT","CLIP","llm_models:GPT","llm_models:ChatGPT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/openai-confirms-that-skys-voice-actress-was-casted-months-before-altman-contacted-johansson\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10004315,"title":"Top 10 Free Resources To Learn AWS","content":"Amazon Web Services is one of the most widely adopted cloud platforms among developers. It has many benefits, such as it is designed to allow developers and vendors to host the applications in a fast and secure manner. It enables the user to select the operating system, database, programming language, is user-friendly, among others. Having knowledge about AWS, therefore, plays to your advantage while applying for data science job roles. Here we list down the top 10 free resources which can help data science enthusiasts and professionals to learn AWS. (The list is in no particular order). 1| Free AWS Digital Training And New Cloud Practitioner Certification Source: AWS official documentation About: This course is designed for customers as well as partners to learn about this cloud platform and build cloud skills. AWS Training and Certification has initiated this free digital training platform, including a new entry-level certification. At aws.training, a developer can enrol for free digital training, after which they will get unlimited access to more than 100 new cloud courses built by the AWS experts. Know more here. 2| AWS Fundamentals Specialisation Source: Coursera About: This specialisation gives an overview of the features, benefits, and capabilities of Amazon Web Services (AWS). In this course, you will understand the core services of AWS, its vital security concepts, strategies for migrating from on-premises to AWS, and basics of building serverless applications with AWS. You will learn AWS fundamental concepts including Regions, Availability Zones, and Virtual Private Clouds (VPCs), learn how to use AWS compute, storage, database, and security services via the AWS Console and more. Know more here. 3| Amazon Web Services – Learning and Implementing AWS Solution Source: Udemy About: This course covers the end-to-end implementation of the major components of AWS. It covers basics such as how to get started in the complex Amazon Web Services technologies of modern-day IT to learning more complex things in AWS such as SSHING. Know more here. 4| AWS: Getting Started with Cloud Security Source: edX About: In this course, AWS Solutions Architects will deliver you a foundational understanding of cloud security, compliance and the shared responsibility model of AWS. This course covers several AWS services, such as Amazon VPC, Amazon CloudWatch, Amazon GuardDuty, AWS Security Hub, Amazon S3, AWS Secrets Manager and more. Know more here. 5| AWS for Startups – Deploying with AWS Tutorial Source: Youtube About: This is a 4-hour video by the freeCodeCamp and is a practical guide to the AWS cloud platform. Here, you will learn how to get started with the cloud platform, learn how to create an AWS account, deploy an application, budget scaling applications, and more. Watch the video here. 6| AWS Tutorial: Introduction to Cloud Computing Source: Edureka About: This AWS tutorial is designed for all the professionals who are interested to learn about Cloud Computing and will help in career paths aimed for AWS Solution Architect, AWS Engineer, DevOps Engineer, Cloud Architect etc. Here, you will understand what AWS is and what its services are. You will also explore the different domains in which AWS offers services, such as database, storage and content delivery and other such. Know more here. 7| AWS Certification Training Resources Source: DigitalCloud About: This course is meant to provide support on your AWS Certification journey. The tutorials are a compilation of free AWS certification training resources which includes in-depth AWS cheat sheets, practice questions, blog articles and video tutorials. This will help you to understand cloud computing as well as prepare you for the AWS Certified Cloud Practitioner exam. Know more here. 8| AWS Developer: Optimising on AWS Source: edX About: This course focuses on helping you optimise your applications, and you work in AWS. It covers topics such as methods to enhance the utilisation with the help of containers such as Amazon Elastic Container Service (Amazon ECS) as well as caching services such as Amazon CloudFront and monitoring tools such as Amazon CloudWatch. Know more here. 9| Learn Amazon Web Services (AWS): The Complete Introduction Source: Udemy About: This course is designed for anyone with a basic understanding of Information Technology to be able to take advantages offered by AWS. In this course, you will learn how AWS fits into the cloud model, hands-on with Amazon Simple Storage Service (S3), Amazon Lambda- Function as a Service, AWS Elastic Simple Compute Service (EC2) and other such. Know more here. 10| Amazon Web Services in Action Source: E-book About: Amazon Web Services in Action introduces you to computing, storing, and networking in the AWS cloud. You will learn about best practices regarding security, high availability and scalability. It also covers how to automate your infrastructure by programmatically calling the AWS API to control every part of AWS, how to secure your infrastructure by isolating networks, controlling traffic and managing access to AWS resources and more. Know more here.","excerpt":"Amazon Web Services is one of the most widely adopted cloud platforms among developers. It has many benefits, such as it is designed to allow developers and vendors to host the applications in a fast and secure manner. It enables the user to select the operating system, database, programming language, is user-friendly, among others. Having […]","categories":["AI Trends"],"tags":["AWS","AWS cloud","AWS services"],"author_name":"Ambika Choudhury","publish_date":"2020-08-06T14:00:00","publication_year":"2020","word_count":810,"keywords":["data science","AWS services","AWS","AI","cloud computing","Scala","serverless","RAG","Git","Aim","AWS cloud","R"],"extracted_tech_keywords":["AI","data science","Aim","RAG","cloud computing","AWS","serverless","R","Scala","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-10-free-resources-to-learn-aws\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10013712,"title":"Guide To FreeSound Datasets With Implementation In PyTorch","content":"The FreeSound is a hierarchical collection of sound classes of more than 600 and has filled them with the audio samples of  297,144. The process is generating  685,403 candidate annotations that express the potential presence of sound sources in audio clips. FreeSound Dataset includes the outcome of everyday sounds, from human and animal sounds to music and sounds made by things. Freesound is developed by Music Technology Research Group, Pompeu Fabra University, Barcelona. To download the free sound dataset for the research project, refer to the following. Datasets: https:\/\/github.com\/MTG\/freesound-datasets. Research\/Publication:https:\/\/www.upf.edu\/web\/mtg\/research\/publications. FreeSound Github:https:\/\/github.com\/MTG\/freesound. They collect data for the following Artistic creation 2. Cultural Preservation 3. Education 4. Health and Well Being 5. Sustainable development The Music Technology Group is organized into four labs, each one led by a faculty member. Audio Signal Processing Lab: Faculty, Head of the lab  Xavier Serra. The lab is concentrated on to advance within the understanding of sound and music signals by combining signal processing and machine learning methods. 2. Music Information Research Lab: Emilia Gomez, Head of the lab. The lab works on topics like sound and music description, music information retrieval, vocalization synthesis, audio source separation, music, and audio processing. 3. Music and Multimodal Interaction Lab: Sergi Jorda, Head of the lab. The lab focuses on multimodal interactive technologies and the way to use them for music creation. 4. Music and Machine Learning Lab: Rafael Ramírez, Faculty, Head of the lab. The lab is concentrated on the intersection of music technology, AI, deep learning, and neuroscience with their applications. Download Size: 20 GB DataLoader: Using Pytorch: import sys, os import torch import librosa import numpy as np import pandas as pd from torch import Tensor from scipy.io import wavfile from torchvision import transforms from torch.utils.data import DataLoader from torch.utils.data.dataset import Dataset class Freesound(Dataset): def __init__(self, transform=None, mode=\"train\"): # setting directories for data data_root = \"..\/input\" self.mode = mode if self.mode is \"train\": self.data_dir = os.path.join(data_root, \"audio_train\") self.csv_file = pd.read_csv(os.path.join(data_root, \"train.csv\")) elif self.mode is \"test\": self.data_dir = os.path.join(data_root, \"audio_test\") self.csv_file = pd.read_csv(os.path.join(data_root, \"sample_submission.csv\")) # dict for mapping class names into indices. can be obtained by # {cls_name:i for i, cls_name in enumerate(csv_file[\"label\"].unique())} self.classes = {'Acoustic_guitar': 38, 'Applause': 37, 'Bark': 19, 'Bass_drum': 21, 'Burping_or_eructation': 28, 'Bus': 22, 'Cello': 4, 'Chime': 20, 'Clarinet': 7,'Computer_keyboard': 8, 'Cough': 17, 'Cowbell': 33, 'Double_bass': 29, 'Drawer_open_or_close': 36, 'Electric_piano': 34, 'Fart': 14, 'Finger_snapping': 40, 'Fireworks': 31, 'Flute': 16, 'Glockenspiel': 3, 'Gong': 26, 'Gunshot_or_gunfire': 6, 'Harmonica': 25, 'Hi-hat': 0, 'Keys_jangling': 9, 'Knock': 5, 'Laughter': 12, 'Meow': 35, 'Microwave_oven': 27, 'Oboe': 15, 'Saxophone': 1, 'Scissors': 24, 'Shatter': 30, 'Snare_drum': 10, 'Squeak': 23, 'Tambourine': 32, 'Tearing': 13, 'Telephone': 18, 'Trumpet': 2, 'Violin_or_fiddle': 39,  'Writing': 11} self.transform = transform def __len__(self): return self.csv_file.shape[0] def __getitem__(self, idx): filename = self.csv_file[\"fname\"][idx] rate, data = wavfile.read(os.path.join(self.data_dir, filename)) if self.transform is not None: data = self.transform(data) if self.mode is \"train\": label = self.classes[self.csv_file[\"label\"][idx]] return data, label elif self.mode is \"test\": return data if __name__ == '__main__': import matplotlib.pyplot as plt tsfm = transforms.Compose([ lambda x: x.astype(np.float32) \/ np.max(x), # rescale to -1 to 1 lambda x: librosa.feature.mfcc(x, sr=44100, n_mfcc=40), # MFCC lambda x: Tensor(x) ]) # todo: multiprocessing, padding data dataloader = DataLoader( Freesound(transform=tsfm, mode=\"train\"), batch_size=1, shuffle=True, num_workers=0) for index, (data, label) in enumerate(dataloader): print(label.numpy()) print(data.shape) plt.imshow(data.numpy()[0, :, :]) plt.show() if index == 0: Break Application: Audio Tagging System Audio tagging is a technique to update the meta-data fields in MP3 and other compressed audio files. An audio tag detector is used to correct the meta-data in individual files or to apply a category to a group of files. Emotion and theme recognition It involves the prediction of moods and themes conveyed by a music track, given the raw audio. Automatic assessment system for musical exercises Music Critic is employed to gauge musical exercises sung by students, to allow meaningful feedback. It is often easily integrated into online applications and education platforms. Animal Sound Recognition: The ability to automatically recognize a large range of animal sounds can analyze the habits and distributions of animals, which makes it possible to watch and protect them effectively. Conclusion: We have learned about the Freesound dataset, how we can download it from the source. Freesound dataset creator and their researcher. Implementation of model in PyTorch data loader for speaker audio tagging Recognition and some of the application of FreeSound Datasets.","excerpt":"The FreeSound is a hierarchical collection of sound classes of more than 600 and has filled them with the audio samples of 297,144.","categories":["Deep Tech"],"tags":["AI Applications","audio","big data platform c++","big data scale","Datasets","machine learning and data transform","Pytorch","Research","Scale Big Data"],"author_name":"Amit Singh","publish_date":"2020-12-09T15:00:00","publication_year":"2020","word_count":716,"keywords":["AI Applications","audio","Git","deep learning","R","Pandas","Pytorch","Datasets","NumPy","PyTorch","Research","Go","machine learning","AI","big data platform c++","machine learning and data transform","Scale Big Data","Matplotlib","big data scale"],"extracted_tech_keywords":["AI","machine learning","deep learning","PyTorch","Pandas","NumPy","Matplotlib","R","Go","Git"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/datasets-freesound-pytorch-research\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":68550,"title":"Name Language Prediction using Recurrent Neural Network in PyTorch","content":"Recurrent Neural Networks have been applied very successfully as the deep learning models in the tasks that deal with the sequential data especially the Natural Language Processing. The traditional feed-forward networks operate with the entire fixed training batch at once and produce a fixed amount of output. On the other hand, the recurrent neural networks process the same in sequence. This feature makes them outperforming in many NLP applications. With these capabilities, RNN models are popularly applied in the text classification problems. In this article, we will demonstrate the implementation of a Recurrent Neural Network (RNN) using PyTorch in the task of multi-class text classification. This RNN model will be trained on the names of the person belonging to 18 language classes. After successful training, the model will predict the language category for a given name that it is most likely to belong. Implementation of RNN in PyTorch This implementation was done in the Google Colab and the data set was read from the Google Drive. The below line of codes will mount the Google Drive to the Colab notebook and print the text files in the data set. from google.colab import drive drive.mount('\/content\/gdrive') from __future__ import unicode_literals, print_function, division from io import open import glob import os def printFiles(path): return glob.glob(path) printFiles('gdrive\/My Drive\/Dataset\/data\/data\/names\/*.txt') The below lines of codes define function modules to convert Unicode text to equivalent ASCII value. import unicodedata import string all_let = string.ascii_letters + \" .,;'\" n_let = len(all_let) def unicodeToAscii(s): return ''.join( c for c in unicodedata.normalize('NFD', s) if unicodedata.category(c) != 'Mn' and c in all_let ) Using the below code snippet, a function will be defined to build the dictionary of categories and a list of names in every language. cat_line = {} all_cats = [] # Read a file and split into lines def readLines(filename): lines = open(filename, encoding='utf-8').read().strip().split('\\n') return [unicodeToAscii(line) for line in lines] for filename in printFiles('gdrive\/My Drive\/Dataset\/data\/data\/names\/*.txt'): category = os.path.splitext(os.path.basename(filename))[0] all_cats.append(category) lines = readLines(filename) cat_line[category] = lines n_categories = len(all_cats) We will check the above function for 4 Japanese names. #Check names in a category print(cat_line['Japanese'][:4]) In the next step, the function modules will be defined to turn the names into tensors to make them compatible with the RNN model. import torch # Find letter index from all_let, e.g. \"a\" = 0 def letterToIndex(letter): return all_let.find(letter) # Turn a letter into a <1 x n_let> Tensor def letterToTensor(letter): tensor = torch.zeros(1, n_let) tensor[0][letterToIndex(letter)] = 1 return tensor # Turn a line into a <line_length x 1 x n_let>, # or an array of one-hot letter vectors def lineToTensor(line): tensor = torch.zeros(len(line), 1, n_let) for li, letter in enumerate(line): tensor[li][0][letterToIndex(letter)] = 1 return tensor We will check the above module by converting a letter to tensor and a line to tensor. print(letterToTensor('K')) print(lineToTensor('Kakinomoto').size()) In the next step, we will define the Recurrent Neural Network model. import torch.nn as nn class RNN(nn.Module): def __init__(self, input_size, hidden_size, output_size): super(RNN, self).__init__() self.hidden_size = hidden_size self.i2h = nn.Linear(input_size + hidden_size, hidden_size) self.i2o = nn.Linear(input_size + hidden_size, output_size) self.softmax = nn.LogSoftmax(dim=1) def forward(self, input, hidden): combined = torch.cat((input, hidden), 1) hidden = self.i2h(combined) output = self.i2o(combined) output = self.softmax(output) return output, hidden def initHidden(self): return torch.zeros(1, self.hidden_size) n_hidden = 128 #Binding model rnn = RNN(n_let, n_hidden, n_categories) This model will be checked on generating tensor output for a name. input = lineToTensor('Aalsburg') hidden = torch.zeros(1, n_hidden) output, next_hidden = rnn(input[0], hidden) print(output) This untrained model has generated the likelihoods of all the categories the given input name belongs to. Now, we will define functions for providing random training examples to the network during training and generating categories for the network outputs. import random def randomChoice(l): return l[random.randint(0, len(l) - 1)] def randomTrainingExample(): category = randomChoice(all_cats) line = randomChoice(cat_line[category]) category_tensor = torch.tensor([all_cats.index(category)], dtype=torch.long) line_tensor = lineToTensor(line) return category, line, category_tensor, line_tensor #Check on a random sample for i in range(10): category, line, category_tensor, line_tensor = randomTrainingExample() print('category =', category, '\/ line =', line) def categoryFromOutput(output): top_n, top_i = output.topk(1) category_i = top_i[0].item() return all_cats[category_i], category_i #Check category for an output print(categoryFromOutput(output)) In the next step, the hyperparameters and the training function will be defined and the RNN model will be trained in 100 epochs. learning_rate = 0.005 def train(category_tensor, line_tensor): hidden = rnn.initHidden() rnn.zero_grad() for i in range(line_tensor.size()[0]): output, hidden = rnn(line_tensor[i], hidden) loss = criterion(output, category_tensor) loss.backward() # Add parameters' gradients to their values, multiplied by learning rate for p in rnn.parameters(): p.data.add_(p.grad.data, alpha=-learning_rate) return output, loss.item() import time import math n_iters = 100000 print_every = 5000 plot_every = 1000 # Keep track of losses for plotting current_loss = 0 all_losses = [] def timeSince(since): now = time.time() s = now - since m = math.floor(s \/ 60) s -= m * 60 return '%dm %ds' % (m, s) start = time.time() for iter in range(1, n_iters + 1): category, line, category_tensor, line_tensor = randomTrainingExample() output, loss = train(category_tensor, line_tensor) current_loss += loss # Print iter number, loss, name and guess if iter % print_every == 0: guess, guess_i = categoryFromOutput(output) correct = '✓' if guess == category else '✗ (%s)' % category print('%d %d%% (%s) %.4f %s \/ %s %s' % (iter, iter \/ n_iters * 100, timeSince(start), loss, line, guess, correct)) # Add current loss avg to list of losses if iter % plot_every == 0: all_losses.append(current_loss \/ plot_every) current_loss = 0 After training, we will visualize the loss to see the performance. import matplotlib.pyplot as plt import matplotlib.ticker as ticker plt.figure() plt.plot(all_losses) The below code snippet will test the on the unseen texts and plot the confusion matrix. # Keep track of correct guesses in a confusion matrix confusion = torch.zeros(n_categories, n_categories) n_confusion = 10000 # Just return an output given a line def evaluate(line_tensor): hidden = rnn.initHidden() for i in range(line_tensor.size()[0]): output, hidden = rnn(line_tensor[i], hidden) return output # Go through a bunch of examples and record which are correctly guessed for i in range(n_confusion): category, line, category_tensor, line_tensor = randomTrainingExample() output = evaluate(line_tensor) guess, guess_i = categoryFromOutput(output) category_i = all_cats.index(category) confusion[category_i][guess_i] += 1 # Normalize by dividing every row by its sum for i in range(n_categories): confusion[i] = confusion[i] \/ confusion[i].sum() # Set up plot figsize = (10, 10) fig = plt.figure(figsize=figsize) ax = fig.add_subplot(111) cax = ax.matshow(confusion.numpy()) fig.colorbar(cax) # Set up axes ax.set_xticklabels([''] + all_cats, rotation=90) ax.set_yticklabels([''] + all_cats) # Force label at every tick ax.xaxis.set_major_locator(ticker.MultipleLocator(1)) ax.yaxis.set_major_locator(ticker.MultipleLocator(1)) # sphinx_gallery_thumbnail_number = 2 plt.show() The below function will print the likelihood of belonging to a language category for the given names. def predict(input_line, n_predictions=3): print('\\n> %s' % input_line) with torch.no_grad(): output = evaluate(lineToTensor(input_line)) # Get top N categories topv, topi = output.topk(n_predictions, 1, True) predictions = [] for i in range(n_predictions): value = topv[0][i].item() category_index = topi[0][i].item() print('(%.2f) %s' % (value, all_cats[category_index])) predictions.append([value, all_cats[category_index]]) Finally, we will check the predicted likelihoods for the given three names. predict('Aggelen') predict('Accardo') predict('Ferreiro') So, as we can see above, the RNN model has given the likelihoods for the given names which of the language categories they belong to. For example, for the name ‘Aggelen’, it has given the top 3 likelihoods in which ‘French’ has the highest value. All three predictions are correct. That means, according to the trained RNN model, the name ‘Aggelen’ has the highest chances of belonging to the ‘French’ language. We could apply the argmax to print only the language category with the highest likelihood, but to make it more clear, the top 3 predictions are given in the result. You can check this model on more numbers of predictions and tune the parameters to improve the accuracy. References:- Gabriel Loye, ‘A Beginner’s Guide on Recurrent Neural Networks with PyTorch’ ‘NLP from Scratch: Classifying Names with a Character-Level RNN’, PyTorch Tutorial.","excerpt":"In this article, we will demonstrate the implementation of a Recurrent Neural Network (RNN) using PyTorch in the task of multi-class text classification. This RNN model will be trained on the names of the person belonging to 18 language classes. After successful training, the model will predict the language category for a given name that it is most likely to belong.","categories":["Deep Tech"],"tags":["cnn neural network","Deep Learning","NLP","Pytorch","Recurrent Neural Network","simple neural network"],"author_name":"Dr. Vaibhav Kumar","publish_date":"2020-06-30T18:00:00","publication_year":"2020","word_count":1279,"keywords":["Recurrent Neural Network","Pytorch","NumPy","text classification","AI","neural network","PyTorch","cnn neural network","NLP","Ray","Colab","deep learning","Deep Learning","Matplotlib","simple neural network"],"extracted_tech_keywords":["AI","deep learning","neural network","NLP","Ray","PyTorch","Colab","NumPy","Matplotlib","text classification"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/name-language-prediction-using-recurrent-neural-network-in-pytorch\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":13019,"title":"10 Robotics Startups in India that are pushing the boundaries of this area","content":"Robotics is not a latest phenomenon. Since the advent of industrial age, robotics has been part of high-tech industry. Yet, the scope has changed drastically today. Today, robotics finds usage beyond manufacturing and more towards areas like ecommerce, logistics, retail, healthcare etc. Robotics also has significant overlap with Artificial Intelligence. Here we list down 10 new age robotics startups in India (in alphabetic order) that are broadening the horizon of what robots are all about. 1. Asimov Robotics ASIMOV Robotics Pvt Ltd is a single window Solution and Services provider who can meet all robotics\/automation needs. They provide engineering products solutions and consultancy in the areas like robotic simulation and control, machine-vision, training, virtual reality, and navigation applications. Their systems\/solutions provide the customers a competitive advantage. They have expertise in designing and developing customized solutions to meet specific requirements to maximize profit, maintain quality, reduce risk and to support research. 2. DiFACTO Robotics and Automation DIFACTO serves a range of industries – automotive, transportation, energy, consumer goods, food and beverages, health and pharmaceutical, defence and aerospace, electronics, and more – for all their robot-based automation needs. DIFACTO’s customers are manufacturing companies, OEMs supplying automated equipment and machinery, robot companies and system integrators\/line builders. 3. Grey Orange OLYMPUS DIGITAL CAMERA GreyOrange is a multinational technology firm that designs, manufactures and deploys advanced robotics systems for automation at distribution and fulfillment centers. They combine expertise in robotics, hardware and software engineering to solve operational inefficiencies in warehouse operations at both third-party logistics vendors as Gridbotsll as companies with in-house logistics functions. With in-house Mechanical, Electrical, Embedded, UI\/UX, Software department and quality assurance teams, GreyOrange ensures unquestioned quality of all their products. 4. Gridbots Gridbots is a young technology and Innovation – Robotics company from India and work in the field of Robotics – Artificial Intelligence and Machine Vision. Founded in August 2007 by young and passionate innovator Mr. Pulkit Gaur [TED Fellow and MIT Young Innovator of the year]. Gridbots today are a team of more than 30 people – Head quartered in Ahmedabad. Gridbots have sales and support offices in Jaipur – Meerut and Faridabad – India. Since inception the company has been growing exponentially both in terms of market share and technological capabilities day by day. 5. I2U2 Robot i2u2 is a telepresence robotic companion – it allows you to log into it from any compatible iOS, Android, Windows device, and take control of it, from anywhere in the world. Walk, talk and play at home, with the ones who matter, whenever you want, wherever you are. 6. Milagrow Founded in 2007, Milagrow came into existence with the aim of offering support as a venture catalyst to fill the ‘Management Capital’ need gaps of growth-seeking micro, small and medium businesses (MSMEs). Milagrow diversified into ‘products’ and set up its HumanTech division in the 4th quarter of 2011 with the objective of empowering individuals and families across the globe with technologically humane products, adapting the latest advancements in the field of Robotics, Mobility and Home Automation. 7. Mukunda Foods Mukunda Foods is and IAN-Investee company that designs and develops products that can make Indian Foods automatically. DosaMatic is one such product from Mukunda Foods that can make multiple types of dosas automatically. DosaMatic can make multiple types of dosas at the touch of a button. 8. Robots Alive Robots Alive is a robot technology startup based in Bangalore, India focused on the application of robotics \/ robot technology across various industries. Robots Alive designs and manufactures robotic systems which are custom built for specific applications. Their specialty is to make learning systems which can be smoothly operated without the need of any robotic expert 9. Sastra Robotics Sastra Robotics India Pvt Ltd is an innovative forerunner in providing cost effective advanced robotics and automation services. Only three years since its launch, Sastra has leaped huge bounds by collaborating with three other pioneering companies in the same market- Switzerland based Cyberbotics, Korea based Dongbu Robot and US based Corobot. It was deemed possible through the collective efforts of talented engineers, whose passion for robotics is exhibited through their vigorous target: to develop and market high end technologies for wider range applications. Their proficiency and competence paved their way to a smashing entry into the robotics industry. 10. Systemantics Systemantics is an Industrial Robotics company headquartered out of Bangalore, India. The company is on a mission to enable widespread adoption of flexible automation in industry. With the experience of having implemented custom turnkey solutions for several clients, the company has accomplished several innovations in robotics and industrial automation such as under water robots and walking machines. Systemantics is focused exclusively on manufacture of high-quality industrial robots with design innovations that facilitate affordable solutions and a faster ROI.","excerpt":"Robotics is not a latest phenomenon. Since the advent of industrial age, robotics has been part of high-tech industry. Yet, the scope has changed drastically today. Today, robotics finds usage beyond manufacturing and more towards areas like ecommerce, logistics, retail, healthcare etc. Robotics also has significant overlap with Artificial Intelligence. Here we list down 10 […]","categories":["AI Trends"],"tags":[],"author_name":"Дарья","publish_date":"2017-02-24T07:23:23","publication_year":"2017","word_count":794,"keywords":["Go","API","artificial intelligence","AI","innovation","Git","RAG","automation","Aim","R"],"extracted_tech_keywords":["AI","artificial intelligence","Aim","RAG","R","Go","Git","API","automation","innovation"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/10-robotics-startups-india-pushing-boundaries-area\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10071286,"title":"India bets big on AI for low-resource language","content":"At the RAISE summit 2020, PM Narendra Modi said, “Our planet is blessed with several languages. In India, we have several languages and dialects; such diversity makes us a better society. As professor Raj Reddy suggested, why not use AI to breach the language barrier seamlessly?” To say that the digitalisation efforts in India are successful, the stakeholders – government, the research community and the industry – would need to make conscious efforts to deliver its benefits to all sections of the society. It is a challenge, considering the language barrier that exists in creating these AI models. Most of the Indian languages are low-resource, which means that they have relatively lesser data available for training NLP systems, especially conversational systems. In this backdrop, the Indian government recently announced the launch of Project Bhashni, which aims to offer easy access to the internet and digital services in their native languages. As per Statista, India was estimated to have over 748 million users in 2020. By 2026, the country is expected to have 1 billion smartphone users, with rural areas driving the sale of internet-enabled phones, reports Deloitte. Building AI for Indian languages One such initiative contributing to low-resource language AI in the country led us to AI4Bharat, an open-source research lab for Indian languages. The initiative is supported by Microsoft’s Research Lab and India Development Center (IDC), which provides ‘unrestricted research grants’ towards building open-source technologies. In addition, it is supported by EkStep Foundation with mentorship and software engineering to build and deploy open-source applications for Indian languages. AI4Bharat looks to contribute in the following areas: Data: Create the largest public datasets and benchmarks across various tasks and 22 Indian languages. AI Models: Build SOTA, open source, foundational AI models across tasks and 22 Indian languages Applications: Design and deploy with partners reference applications to demonstrate the potential of open source AI models. Ecosystem: Enable startups, researchers, and government to innovate on Indian language AI tech with educational material and workshops. Some of the datasets and language generation models launched by AI4Bharat include Indic Corp, IndicNLG Suite, IndicGLUE, IndicXtreme (coming soon), and Naamapadam (coming soon). IndicCorp consists of large sentence-level monolingual corpora for 11 Indian languages and Indian English containing 8.5 billion words (250 million sentences) from multiple news domain sources. IndicNLG Suite consists of training and evaluation datasets for five diverse language generation tasks spanning 11 Indic languages. It is one of the largest multilingual generation dataset collections across languages. Meanwhile, IndicGLUE offers a benchmark for six NLU tasks spanning 11 Indian languages containing formal training and evaluation sets. IndicXtreme offers a benchmark for zero-shot and cross-lingual evaluation of various NLU tasks in multiple Indian languages. Naamapadam offers training and evaluation datasets for named entity recognition in multiple Indian languages. In terms of machine translation, AI4Bharat has developed applications like Samanantar, IndicTrans, and Shoonya. For instance, Shoonya improves the efficiency of language work (like translation, speech transcription, text validation, optical character recognition, etc.) in Indian languages with AI tools and custom-built UI interfaces and features. The team believes that this is a key requirement to create larger datasets for training AI models like neural machine translation. The current focus of this tool is on translation. The first version of Shoonya (v1) is expected to release later this month. In the area of machine transliteration, AI4Bharat has launched Aksharantar, IndicXlit, and IndiclangID. IndiclangID is a model for identifying the language of romanised Indian language words in Aksharantar. Another application developed by AI4Bharat includes Chitralekha, an open-source tool for video transcription with optional translation support focused on Indian languages. The first version of Chitralekha is expected to be launched later this month. Check out all the open source datasets, models and libraries from AI4Bharat here. On July 28, 2022 – AI4Bharat will be launching the Nilekani centre. The event will also have a workshop on Indian language technology. The foundation of this centre has been led by technopreneur and Infosys chairman Nandan Nilekani, focusing on open-source language technologies for the public good. Other contributions Top educational institutions are also actively contributing to this domain. Earlier this month, researchers from IIT Guwahati developed a named entity annotation dataset for low resource Assamese language with a baseline Assamese named entity recognition (AsNER) model. The dataset contains about 99K tokens. This includes text from the speech of the Prime Minister of India and an Assamese play. In May 2022, IIT Kharagpur researchers demonstrated a large-scale analysis of multilingual abusive speech in Indic languages. The team examined different interlingual transfer mechanisms and observed the performance of various multilingual models for abusive speech detection for eight different Indic languages. The languages include Kannada, Bengali, Hindi, English, Malayalam, Marathi, Tamil and Urdu. Last year, researchers from IIIT Hyderabad and the University of Bath developed an automated framework for Indian language neural machine translation (NMT) systems. This framework aims to address the lack of large-scale multilingual sentence-aligned corpora and robust benchmarks. IIT (BHU) Varanasi researchers have developed linguistic resources for Bhojpuri, Magahi, and Maithil, where they performed basic statistical measures for these corpora at character, word, syllable, and morpheme levels, alongside understanding their similarity estimates and baselines for three applications. Jadavpur University researchers developed an end-to-end procedure to improve semantic search performance using semi-supervised and unsupervised learning algorithms on an available Bengali repository to have seven types of semantic properties – namely conceptual, connotative, collocative, social, affective, reflected, and thematic – to develop the system. C-DAC (The Centre for Development of Advanced Computing), an Indian autonomous scientific society operating under the MeitY, has launched several projects related to low-resource languages. Some of them include C-DAC GIST, multilingual computing, and others. Besides these, there are startups like Gnani.ai, Reverie Language Technologies, DheeYantra Research Labs, and RaGaVeRa Indie Technologies, and others, are also contributing to the ecosystem. India’s mission to make AI accessible Last year, MeitY unveiled a detailed plan in its white paper, ‘National Language Translation Mission,’ also known as Project Bhashni. (Source: MeitY) As per the plan, MeitY will be setting up a process to utilise the contributions received from the ecosystem. The think tank said that both ‘state and language missions’ and ‘language missions’ would be established across states to focus on data collection and content creation in specific languages. The language missions, on the other hand, would be region specific, depending upon what language is spoken in the region, and there could be multiple language missions across a single state. Language missions would be responsible for empanelling agencies for data collection, curation and validation activities. Further, it stated that all original or translated data should be collected using a UCLA-compliant process defined by the data management unit (DMU). Also, the language missions would be responsible for identifying data sources from state government entities and driving crowdsourcing efforts through standard Bhashini tools. Most importantly, MeitY stated that the language missions would run awareness campaigns for crowdsourcing, which would be focused on low-resource languages. On the other hand, the performance of language missions would be measured through a public dashboard. DMU would also develop in-house resources for certain low-resource languages or specific tasks for quality control. In addition, state language missions would provide resources to DMU for all languages. For the first two years, MeitY aims to expand training datasets in low-resource languages, including North Eastern and Tribal languages, alongside other languages where sufficient data is not available. By 2024, MeitY looks to sign MoUs with five more entities each in public and private sectors for data contribution, specifically for low resource languages. Also, drive campaigns for data crowdsourcing, creation, and curation in low-resource languages.","excerpt":"The first version of Shoonya is expected to release later this month.","categories":["IT Services"],"tags":["AI4Bharat"],"author_name":"Amit Naik","publish_date":"2022-07-21T11:00:00","publication_year":"2022","word_count":1267,"keywords":["Go","semantic search","AI","ML","Git","RAG","NLP","Aim","ViT","AI4Bharat","R"],"extracted_tech_keywords":["AI","ML","NLP","Aim","RAG","semantic search","R","Go","Git","ViT"],"url":"https:\/\/analyticsindiamag.com\/it-services\/india-bets-big-on-ai-for-low-resource-language\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10166530,"title":"ByteDance Drops &#8216;InfiniteYou&#8217;, an AI Model for Photo Recrafting","content":"Researchers at ByteDance Intelligent Creation have developed a new AI model that generates multiple versions of an identity along with its paper, demo, and code. Liming Jiang, a senior research scientist at ByteDance, made the announcement on X on Sunday. The new AI model called InfiniteYou (InfU) aims to address the challenges of identity-preserved image generation. One can create multiple versions of their identity in different settings by using prompts as required, ensuring good accuracy. The model leverages Diffusion Transformers (DiTs) to generate images that not only maintain the identity of a person from a source photograph but also allow for flexible text-based editing. InfU aims to overcome the limitations found in existing methods, such as insufficient identity similarity, poor text-image alignment, and low generation quality. The core of InfU is InfuseNet, a component designed to inject identity features into the DiT base model through residual connections. This process enhances identity similarity while preserving the model’s generative capabilities. To further refine the model’s performance, a multi-stage training strategy was employed, incorporating pretraining and supervised fine-tuning (SFT) with synthetic single-person-multiple-sample (SPMS) data. The training approach was designed to improve text-image alignment, enhance image quality, and mitigate face copy-pasting issues. The official website mentioned, “InfU features a desirable plug-and-play design compatible with many existing methods. It naturally supports base model replacement with any variants of FLUX.1-dev, such as FLUX.1-schnell for more efficient generation.” “The compatibility with ControlNets and LoRAs provides more controllability and flexibility for customised tasks. Notably, the compatibility with OminiControl extends our potential for multi-concept personalisation, such as interacted identity (ID) and object personalised generation,” the paper added. The code is available on the GitHub page, and one can access the demo and the model on Hugging Face to try it out. ByteDance has been making several developments in 2025, including Goku as an alternative to Google’s Luma and a React Native killer. The AI model adds to its list of exciting developments so far.","excerpt":"ByteDance has released a new AI model that enables users to generate multiple versions of themselves.","categories":["AI News"],"tags":["ByteDance"],"author_name":"Ankush Das","publish_date":"2025-03-24T15:23:50","publication_year":"2025","word_count":325,"keywords":["ByteDance","Hugging Face","Go","AI","Transformers","Git","RAG","Aim","ai_frameworks:Hugging Face","GitHub","R"],"extracted_tech_keywords":["AI","Aim","Hugging Face","Transformers","RAG","R","Go","Git","GitHub","ai_frameworks:Hugging Face"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/bytedance-drops-infiniteyou-an-ai-model-for-photo-recrafting\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10140913,"title":"Can Legacy Systems Transition Away from C\/C++?","content":"A few days ago, the US Cybersecurity and Infrastructure Security Agency (CISA) and FBI issued a stark warning: Critical software must move away from memory-unsafe languages like C and C++ by January 2026 or face significant security risks. Memory safety vulnerabilities account for approximately 70% of security vulnerabilities in software systems today. These issues arise from the manual memory management required in languages like C and C++, where programmer mistakes can lead to buffer overflows and other critical security flaws that attackers can exploit to take control of systems. CISA said that for new product lines meant for critical infrastructure or national functions, using memory-unsafe languages is deemed “dangerous and significantly elevates risk to national security”. By January 2026, organisations must either have a clear memory-safety roadmap or transition to memory-safe alternatives. The White House had suggested the same, saying experts have identified a few programming languages that both lack traits associated with memory safety and also have high proliferation across critical systems, such as C and C++, and if possible, should be avoided. Is it a Good Idea? When AIM asked about replacing C\/C++ with modern-day programming languages like Rust, Aleksa Gordić, ex research engineer at DeepMind and founder of Runa AI, said that C\/C++ power everything. There is an enormous amount of legacy code out there that has been written over the past decades. For reference, C was created in 1972 by Dennis Ritchie, and C++ in 1979 by Bjarne Stroustrup. Most microcontrollers in existence today, powering everything from toasters and refrigerators to keyboards, aeroplanes, and cars, all run on C. Many companies have built their entire tech stacks in these languages and are transitioning to a new language\/stack. Even if it is strictly better along every single dimension, it is prohibitive from a cost perspective ($, dev time, etc.). On the flip side, similar to biological systems, evolutionary pressures are at play here, with newer languages branched off from older ones occasionally proving to be better. Aleksa gave the example of Rust, which was created in 2006. Its creator, Graydon Hoare, had the advantage of learning from the strengths and weaknesses of previous generations of languages. This is clearly an advantage that newer languages enjoy. “But, just to be clear, strictly better language design is often not enough. Languages are often more about the communities built around them than having a ‘superior design’. The majority of programmers still need to use the language and build libraries, projects, and Q&A forums around it,” he added. One of the primary concerns among developers is the question of interoperability. As a developer pointed out on Reddit, “C is the bridge between languages – just about every language has a CFFI, and that usually comes free with the language. If we replace C code, how will these languages communicate with each other?” Even if we consider the performance part of C++, the opinions are mostly positive. C++ developers, when highly skilled, often get paid hefty salaries. These are not your typical developers, but those that often deal with complex tasks, including quantitative trading and robotics systems. Apart from handling complex operations, other benefits include extreme performance (less than 1ms response times), very low server footprint, and accomplishing features otherwise dismissed as “not possible at scale”. But the most important question here is what happens to the legacy systems? On a higher level, you have two options: either rewrite software from scratch using the new programming language or write a wrapper around the legacy code using the proposed programming language. The problem is that both require a lot of time, effort, and debugging. And then, of course, there’s the money side of things. A Reddit user mentioned that if the US government wants its developers to adopt the latest and greatest languages that are in high demand in Silicon Valley, they’ll have to pay developers more than the GS schedule. “The highest-paid US gov grade 15 step 10 engineer makes less than our lowest paid engineer,” he added, suggesting that the US government might need to reform their salary structure to facilitate the ecosystem of modern programming languages. Contrary to this, when AIM spoke to Pratham Patel, a developer at Rocky Enterprise Software Foundation, he said that replacing C\/C++ isn’t the “be all and end all” solution, but it is a good solution nonetheless. There are very few languages that are memory-safe and are without runtime. Python, Java, and JavaScript are memory-safe languages but require the CPython interpreter, JVM and V8 respectively. “C, C++, Rust and Zig are languages which do not need runtimes. The last two [Rust and Zig] enforce as much memory safety as possible,” he added, suggesting that C\/C++ can be replaced with languages like Rust and Zig due to their memory safety features. Patel later mentioned that these legacy systems remain in use because they generate revenue for both the businesses that rely on them and the IT companies that maintain them through expensive support contracts. However, these systems inevitably reach a point where the original code becomes incomprehensible to current developers, creating a technical debt crisis. The COBOL Effect The idea of moving away from something like C\/C++ is a nightmare for programmers, especially since it’s used in almost everything. Sure, if you consider high-level programming, you might not be able to see the relevance, but as soon as you dig deeper, you will realise that from motherboards to operating systems, C\/C++ is an integral part of modern computers. Take COBOL, for instance. Everyone believed it was dead until the New Jersey government started looking for COBOL developers in 2020 since it still powers a majority of their banking systems. A balanced approach is what Linux is going with. While Linux was built on top of C, in recent years, it is being integrated into the Linux kernel as a second language. This way, we can slowly move away from languages like C and utilise modern-day languages like Rust while not completely abandoning them.","excerpt":"“C is the bridge between languages. If we replace C code, how will these languages communicate with each other?”","categories":["Deep Tech"],"tags":["C++","legacy systems"],"author_name":"Sagar Sharma","publish_date":"2024-11-13T16:30:00","publication_year":"2024","word_count":997,"keywords":["Go","Rust","AWS","AI","Python","Ray","C++","Aim","JavaScript","legacy systems","R","Java"],"extracted_tech_keywords":["AI","Aim","Ray","AWS","Python","R","JavaScript","Go","Rust","Java"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/can-legacy-systems-transition-away-from-c-c\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10165231,"title":"Hugging Face Teams Up With JFrog To Hunt Down Malicious AI Models","content":"Hugging Face, one of the most sought-out platforms to host AI models, announced a partnership with software supply chain platform JFrog to improve security on the Hugging Face Hub. Hugging Face explained that the model weights can contain code executed upon deserialisation and sometimes at inference time, depending on the format. To tackle this, it plans to integrate JFrog’s scanner into its platform, adding new scanning functionality to reduce false positives on the Model Hub. “Through our integration with Hugging Face, we bring a powerful, methodology-driven approach that eliminates 96% of current false positives detected by scanners on the Hugging Face platform while also identifying threats that traditional scanners fail to detect,” JFrog stated. “Our unique approach dissects embedded code, extracts payloads, and normalises evidence to eliminate false positives while detecting more serious threats.” JFrog’s scanner aims to perform a deeper analysis and parse the code in model weights to check for potential malicious usage. The scanning is powered by its ‘file security scans’ interface. It supports various models, including pickle-based models, TensorFlow models, GPT-Generated Unified Format (GGUF) models, Open Neural Network Exchange (ONNX) models, and more. Their documentation lists out all kinds of AI models supported by JFrog. Users do not need to do anything to benefit from the integration. All the public model repositories will be scanned by JFrog automatically as soon as they push files to the Model Hub. Hugging Face has shared an example repository where users can check how the scanner flags malicious files. With this integration to Hugging Face, users should get a better sense of security before using AI models to deploy for their use-cases.","excerpt":"JFrog’s scanner aims to perform a deeper analysis and parse the code in model weights to check for potential malicious usage.","categories":["AI News"],"tags":["AI Security"],"author_name":"Ankush Das","publish_date":"2025-03-05T14:27:33","publication_year":"2025","word_count":272,"keywords":["Hugging Face","AI","neural network","AI Security","GPT","Aim","ai_frameworks:TensorFlow","ai_frameworks:Hugging Face","TensorFlow","R","llm_models:GPT"],"extracted_tech_keywords":["AI","neural network","Aim","TensorFlow","Hugging Face","R","GPT","llm_models:GPT","ai_frameworks:TensorFlow","ai_frameworks:Hugging Face"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/hugging-face-teams-up-with-jfrog-to-hunt-down-malicious-ai-models\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10011334,"title":"10 Free Online Resources To Learn Convolutional Neural Networks","content":"Convolutional Neural Networks (CNNs) are one of the most important neural network algorithms in the present scenario. Tech giants like Google, Facebook, Amazon have been thoroughly using this neural network to perform and achieve a number of image-related tasks. The applications of CNNs mostly includes the field of computer vision for image recognition, object detection, among others, This neural network is also being used for video inputs, speech recognition, sentence modelling, etc. in NLP models and more. Below, we have curated a list of 10 best free online resources, in no particular order, to learn Convolutional Neural Networks (CNNs). Convolutional Neural Networks About: This course is a part of the Deep Learning Specialisation at Coursera. Here, you will learn how to build convolutional neural networks and apply them to image data. You will understand how to build a CNN model, understand the recent variations, know how to apply convolutional networks to visual detection as well as recognition tasks and more. Know more here. Introducing Convolutional Neural Networks About: This tutorial is curated by the developers at Google. This tutorial, encompasses a brief introduction on convolution neural networks (CNNs), how it works, including hands-on training. You will learn topics like ReLU, pooling, fully connected layers and more. Know more here. Convolution Neural Networks for Visual Recognition About: This is a free course where you will learn about convolution neural networks and how they can be used in visual recognition. The tutorial starts with an architecture overview and then moves into ConvNet layers such as normalisation layer, fully connected layer, etc. including its architectures, such as layer patterns, computational considerations and more. Know more here. Convolutional Neural Network Tutorial: From Basic to Advanced About: In Convolutional Neural Network Tutorial: From Basic to Advanced, you will learn a basic description of the CNN architecture and its uses. The tutorial also provides two brief sessions to help you build and train a CNN using Keras and TensorFlow, respectively. You will learn about CNNs, applications of computer vision, CNNs in real-world and more. Know more here. Convolutional Neural Network (CNN) About: This is a tutorial on Convolutional Neural Network (CNN) provided by the TensorFlow developers. This tutorial demonstrates training a simple convolutional neural network to classify CIFAR images. You will learn how to import TensorFlow, prepare image dataset, verify data, create a convolutional base and other such. Know more here. Convolutional Neural Network Tutorial – Developing An Image Classifier In Python Using TensorFlow About: The tutorial, Convolutional Neural Network Tutorial – Developing An Image Classifier In Python Using TensorFlow is provided by Edureka. Here, you will understand what CNNs are, the architecture behind convolutional neural networks, layers such as ReLU, pooling, prediction of images using CNNs, among others. Know more here. Convolutional Neural Networks in TensorFlow About: This tutorial, Convolutional Neural Networks in TensorFlow, is a part of the DeepLearning.AI TensorFlow Developer Professional Certificate at Coursera. In this tutorial, you will learn advanced techniques to improve the computer vision model. Learners will be able to explore working with real-world images in different shapes and sizes, visualise the journey of an image through convolutions to figure out how a computer understands and grasps information, explore strategies and learn plot loss and accuracy to prevent overfitting, including augmentation and dropout. Know more here. Convolutional Neural Networks tutorial – Learn how machines interpret images. About: The tutorial, Convolutional Neural Networks tutorial – Learn how machines interpret images will help you understand how convolutional neural networks have become the backbone of the artificial intelligence industry and how CNNs are shaping industries of the future. The topics include what CNNs are, how it works, applications of CNNs, speech recognition using CNNs and much more. Know more here. Convolutional Neural Networks with TensorFlow About: In this tutorial, you’ll learn how to construct and implement Convolutional Neural Networks in Python with the TensorFlow framework. In this tutorial, learners will be introduced to tensors and how they differ from matrices, implementation of the convolutional neural network, how to construct the deep neural network model, among others. Know more here. Convolutional Neural Networks (CNN) About: This course is more like a hands-on practice than theory provided in Kaggle. Here you will learn how to load the dataset, introduction to CNNs, max pooling, same padding, implementing with Keras, evaluate CNN models, among others. Know more here.","excerpt":"Convolutional Neural Networks (CNNs) are one of the most important neural network algorithms in the present scenario. Tech giants like Google, Facebook, Amazon have been thoroughly using this neural network to perform and achieve a number of image-related tasks. The applications of CNNs mostly includes the field of computer vision for image recognition, object detection, […]","categories":["AI Trends"],"tags":["cnn neural network","CNNs","computer vision neural network","Convolutional Neural Network","face recognition online","how artificial intelligence works","Neural Networks","online network graph"],"author_name":"Ambika Choudhury","publish_date":"2020-11-08T13:00:00","publication_year":"2020","word_count":721,"keywords":["computer vision neural network","how artificial intelligence works","artificial intelligence","Keras","AI","online network graph","neural network","TensorFlow","image recognition","cnn neural network","computer vision","NLP","deep learning","object detection","face recognition online","Convolutional Neural Network","CNNs","Neural Networks"],"extracted_tech_keywords":["AI","artificial intelligence","deep learning","neural network","NLP","computer vision","TensorFlow","Keras","image recognition","object detection"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/10-free-online-resources-to-learn-convolutional-neural-networks\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10007588,"title":"Behind The TikTok Fiasco: Another Shot By US Tech Giants At Cloud Dominance","content":"After a month of pulling all tricks on either side, TikTok might have finally found a trusted partner. Among Microsoft and a handful of other suitors, Larry Ellison’s Oracle has been chosen by Chinese internet company ByteDance to anchor itself on the US soil. Yesterday, Oracle confirmed it struck a deal with TikTok-owner Bytedance to become a “trusted technology provider” (read: not the owner). Oracle would reportedly be managing TikTok’s US data. The deal still needs approval from the US and China. In August, the White House gave Tik Tok parent company, ByteDance a 45-day deadline to sell its US TikTok operations citing security threat. Timeline of how the events unfolded Trump announces the ban of TikTok on concerns that Americans’ user data could be shared with the Chinese government.New announcement saying TikTok will not be banned if an American company acquires its US operations. The deadline was set to September 20th.Oracle and Microsoft-Walmart emerge as the leading suitors.China pulls its own cards. Creates roadblocks for algorithms sharing. TikTok deal gets more complicated. Oracle gets picked as a trusted partner. Microsoft gets rejected by TikTok. The deal is under review by the CFIUS. The TikTok ordeal happens to be one of the many power plays by the US in its trade war against China, which has been alleged for technology espionage and other security lapses. But, the race for TikTok ownership hints at something else — a cloud war. The whole deal has turned out to be a new turf for the American cloud providers to take another shot at dominance. AWS is the undisputed cloud leader. Azure had already bagged a $10 billion contract from the Pentagon. Oracle, a relatively obscure name in the cloud markets might finally get to enter the trillion-dollar club. What Oracle Stands To Gain From This Currently, Google hosts TikTok data. According to The Information, TikTok signed a three-year agreement worth $800 million with Google Cloud.  TikTok’s cloud spending clearly indicates the interest in the high-stakes courtship for the viral video app among the biggest enterprise tech companies. The reports suggest that the Oracle’s deal to act as TikTok’s new “trusted technology partner” could be even worth more than $1 billion in potential revenue annually for Oracle’s cloud business in the coming years. With this deal, Oracle will possibly get a stake in TikTok’s US assets, likely meaning that TikTok will be moving its data to Oracle’s cloud. Oracle could also get TikTok to use its marketing software for ads. Chinese internet giant ByteDance too won’t have any qualms about moving away from Google cloud given the latter’s decision to shut down its cloud services in China. Since the late 70s, Oracle has been selling its software services across the world. It also provides cloud services, which are obscured due to the presence of behemoths like Amazon Web Services and Microsoft’s Azure. Oracle had already been pipped by Azure in the race to grab Pentagon’s $10 billion contract last year; Oracle even filed a lawsuit against the Department of Defense, which it lost. Now, it is Oracle’s turn to return the favor. Microsoft brought Walmart alongside this deal. However TikTok’s parent company Bytedance chose to stay with Oracle. By housing TikTok’s data on its servers, Oracle would provide a major lift to its otherwise unpopular cloud services. And, by losing this deal, Microsoft’s ambitious attempts to have their own YouTube-like product have been thwarted. Will This Deal Go Through The deal is currently under review by the Committee on Foreign Investment in the US (CFIUS), which includes officials from the Treasury, State, Commerce and other departments. Earlier Trump has insisted the only way TikTok can avoid getting banned is a sale to a US company. But, this isn’t an acquisition by Oracle. So, will this deal go through? Well, it might; because: It will bring 1000s of jobs to the American soil as TikTok gets a US HQ.President Trump will play it safe with the American tik tokers with elections around the corner.Oracle’s founder Larry Ellison has a good rapport with Trump. Earlier this year, in April, Oracle cloud services were also picked by Zoom, whose traffic increased dramatically during the pandemic. Now with the TikTok partnership, Oracle is finally beginning to warm up to the vast customer base, which was ruled by the likes of AWS, Azure and Google Cloud. If not for Oracle, this deal certainly is a good bargain for TikTok as it gets to keep its 100 million monthly active American users without giving up ownership of its business.","excerpt":"After a month of pulling all tricks on either side, TikTok might have finally found a trusted partner. Among Microsoft and a handful of other suitors, Larry Ellison’s Oracle has been chosen by Chinese internet company ByteDance to anchor itself on the US soil. Yesterday, Oracle confirmed it struck a deal with TikTok-owner Bytedance to […]","categories":["Deep Tech"],"tags":["Microsoft","Walmart"],"author_name":"Ram Sagar","publish_date":"2020-09-18T10:00:18","publication_year":"2020","word_count":759,"keywords":["Go","cloud_platforms:Azure","AWS","AI","cloud_platforms:AWS","R","cloud_platforms:Google Cloud","Rust","Walmart","cloud_platforms:Amazon Web Services","Azure","Microsoft"],"extracted_tech_keywords":["AI","AWS","Azure","R","Go","Rust","cloud_platforms:AWS","cloud_platforms:Azure","cloud_platforms:Google Cloud","cloud_platforms:Amazon Web Services"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/tik-tok-oracle-cloud-microsoft-us-china\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10061742,"title":"Despite the downfall, why the world needs Intel more than ever","content":"On February 22, 2022, Intel’s market cap was sitting at USD 197.25 billion, while AMD, its arch-rival, commanded a market cap of USD 197.75 billion. For the first time in history, AMD overtook Intel’s market share. Intel, the world’s largest chipmaker for more than 30 years, has been dethroned from its position. Manufacturing delays, failure to enter the mobile chips market, and losing market share in the PC and data centre segment to competitors, are the few reasons that led to the downfall of the American chipmaker giant. However, despite its present downtrend, the company, which was synonymous with innovation, is poised to play a pivotal role in the world of semiconductors in the future. Intel bets on future tech From developing the Intel 8080, the processor that made personal computing accessible to everyone, to building energy-efficient chips for data centres, the American chipmaker always had the vision and the foresight to gauge the future trends in the tech industry and build products accordingly. With this forward-looking vision, Intel is currently focusing on building chips that will power future technologies. Intel has launched a new blockchain chip to make crypto mining more energy efficient. The company claims that the blockchain chip will deliver a 1000x better performance per watt than mainstream GPUs for SHA-256 based mining. At the same time, the company has introduced Intel Xeon D-1700 and D-2700 processors targeted at enterprises and cloud vendors. These processors are purpose-built to operate in the rugged and space- and power-constrained environments found at edge locations. Intel plans on launching new processors based on its Xe architecture to power computing in the metaverse world. The processors based on Intel’s new Xe architecture aim at accelerating and rendering rich and immersive experiences to scale from the client to the server end. In addition, two processors, namely, Intel Arc Alchemist GPU aimed at gaming and creation and Ponte Vecchio to accelerate high-performance computing and visualisation, will be launched next year. Recently updated, OpenVINO is a free toolkit provided by Intel that helps developers boost deep learning performance in computer vision, automatic speech recognition, natural language processing and other common tasks. Also, in the hardware segment, Intel has launched Xeon Scalable processors to speed up the entire data pipeline involved in artificial intelligence, machine learning, and deep learning. Tailwinds for Intel “We had some missteps. The strategy had become a little bit confused on the role that we’re going to play in manufacturing for the long term. And now we’re leaning back into that with clarity, with clear urgency,” said Pat Gelsinger, Intel’s CEO. Intel had to face severe production challenges in the past decade. It is the only company that designs and manufactures its own chips. Prominent players in the space, which include companies such as AMD, NVIDIA, Samsung – all just design their chips and outsource their manufacturing operations to vendors, such as TSMC, the largest contract semiconductor manufacturer in the world. The global chip shortages triggered by the trade war between the US and China and the COVID19 pandemic crippled the automobile and electronics industries, forcing many countries to seek alternate options for chips manufactured in Asia. To reduce its dependency on Asian manufacturers for its chip requirements, the US government passed the Creating Helpful Incentives to Produce Semiconductors (CHIPS) for America Act. This act authorised a series of programs to promote the research, development, and fabrication of semiconductors within the United States. The US government’s USD 2 trillion investment package includes USD 50 billion for the semiconductor industry. The US accounts for just 12% of global chip production, while Asia accounts for 75%, a report from the Semiconductor Industry Association said. The report also claimed China becoming the largest chip producer by 2030. To bolster manufacturing in the US and reduce reliance on Asian manufacturers for its chip requirement, Intel plans to spend USD 20 billion to build its foundry division to act as a manufacturing partner for other companies that focus on semiconductor design. Intel’s Foundry Services will compete in a market potentially worth $100 billion by 2025 and will manufacture a range of chips, including chips based on ARM technology used in mobile devices, Gelsinger said. Intel currently operates four factories in the United States and plans to expand Arizona’s manufacturing activities. The company has also set up chip manufacturing units in Ireland, Israel and China. Although TSMC is presently the largest semiconductor manufacturer globally, ongoing tensions between China and Taiwan put the global supply of chips at risk, paving the way for Intel to regain its footing in the global semiconductor market. Pent-up demand for semiconductors Intel expects the long-term demand for semiconductors to create a USD 1 trillion market opportunity by 2030. The company has doubled down on its R&D investments to accelerate long-term growth to tap into the trillion-dollar opportunity. It has devised the Smart Capital Strategy to achieve this vision. Under the Smart Capital strategy, Intel plans to employ a disciplined approach to its investments and leverage government incentives, customer participation and other creative partnerships as offsets to capital spending. This will pave the way for the company to adjust quickly to opportunities in the market and gain share while managing its margin structure and capital spending. Intel firmly believes that the digitisation of everything, which is being driven by the four superpowers: cloud, 5G fueled connectivity, AI and the intelligent edge, has boosted the demand for semiconductors globally. With a renewed focus on strategy, newer product lineups, and shifting geopolitical sands, the company is well-positioned to meet the ever-growing global demand for semiconductors.","excerpt":"The advent of 5G, cloud, AI and intelligent edge is expected to create a USD 1 trillion market opportunity for semiconductors by 2030","categories":["AI Features"],"tags":["Intel","Intel Chips"],"author_name":"SharathKumar Nair","publish_date":"2022-02-28T17:00:00","publication_year":"2022","word_count":929,"keywords":["Go","machine learning","artificial intelligence","AI","ML","Intel Chips","computer vision","RAG","Aim","deep learning","R","Intel"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","computer vision","Aim","RAG","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/despite-the-downfall-why-the-world-needs-intel-more-than-ever\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":50731,"title":"DXC Forays Into Recruiting From Colleges: Will Hire 5000 Graduates In India","content":"DXC, a leading service provider, has planned to hire 5000 graduates from Indian colleges in 2020. The multinational company will be looking for talents, especially in the field of analytics and software skills. The company rides on the proficiency of its efficient 40000 employees and a global headcount of 138,000. Founded in 2017, DXC Technology was quick enough to get off the ground in providing analytics, cloud, security, IoT, digital foundation services. And not just in India, the company’s offering is well received across Asia, Australia, and Europian continent. Earlier, the company was not active in hiring from colleges, however, now they have devised a plan to attract talents from premier institutions. This sudden change in strategy can be linked to the reform in management in September; DXC made Mike Salvino as its CEO. “We want to concentrate more on our people and strengthen our employee value proposition. Our employees are expected to be clear about their career path at DXC, the opportunities to work with new clients and also the opportunities for reskilling and retraining,” said the CEO on 11 November. Apart from hiring graduates who are proficient in analytics, it is believed that the firm will also include students from different technologies to cater to its diverse clients. Optimising its services to orchestrate across its client is of paramount importance for the firm. Consequently, a wide range of specialist will help them to provide superior services for 6000 private and public-sector clients in 70 countries. Since its inception, DXC Technology has made profits, but of late, the company has reported a shink in the revenue from $5.01 billion to $4.8. Recent loss in profits might have slackened the progress, but now the company will look to gain momentum with the help of fresh graduates.","excerpt":"DXC, a leading service provider, has planned to hire 5000 graduates from Indian colleges in 2020. The multinational company will be looking for talents, especially in the field of analytics and software skills. The company rides on the proficiency of its efficient 40000 employees and a global headcount of 138,000.  Founded in 2017, DXC Technology […]","categories":["AI News"],"tags":[],"author_name":"Rohit Yadav","publish_date":"2019-11-26T17:57:41","publication_year":"2019","word_count":296,"keywords":["programming_languages:R","AI","Git","analytics","R"],"extracted_tech_keywords":["AI","analytics","R","Git","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/dxc-forays-into-recruiting-from-colleges-will-hire-5000-graduates-in-india\/","complexity_score":3,"technical_depth":5,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":65845,"title":"Transfer Learning For Multi-Class Image Classification Using Deep Convolutional Neural Network","content":"Image classification has become more interesting in the research field due to the development of new and high performing machine learning frameworks. With the advancement of artificial neural networks and the development of deep learning architectures such as the convolutional neural network, that is based on artificial neural networks has triggered the application of multiclass image classification and recognition of objects belonging to the multiple categories. Every latest machine learning framework has a comparative advantage over the older ones in terms of performance and complexity. In this article, we will implement the multiclass image classification using the VGG-19 Deep Convolutional Network used as a Transfer Learning framework where the VGGNet comes pre-trained on the ImageNet dataset. For the experiment, we will use the CIFAR-10 dataset and classify the image objects into 10 classes. The classification accuracies of the VGG-19 model will be visualized using the non-normalized and normalized confusion matrices. What is Transfer Learning? Transfer learning is a research problem in the field of machine learning. It stores the knowledge gained while solving one problem and applies it to a different but related problem. For example, the knowledge gained while learning to recognize cats could apply when trying to recognize cheetahs. In deep learning, transfer learning is a technique whereby a neural network model is first trained on a problem similar to the problem that is being solved. Transfer learning has the advantage of decreasing the training time for a learning model and can result in lower generalization error. VGGNet – The Deep Convolutional Network VGGNet is Deep Convolutional Neural Network that was proposed by Karen Simonyan and Andrew Zisserman of the University of Oxford in their research work ‘Very Deep Convolutional Neural Networks for Large-Scale Image Recognition’. The name of this model was inspired by the name of their research group ‘Visual Geometry Group (VGG)’. As this convolutional neural network has 19 layers in its architecture, it was named VGG-19. This model was proposed to reduce the number of parameters in a convolutional neural network with improved training time. Below is the block diagram of VGG-19 that illustrates its architecture. (Image source: mc.ai) The biggest advantage of this network is that You can load a pre-trained version of the network trained on more than a million images from the ImageNet database. A pre-trained network can classify images into thousands of object categories. Due to this advantage, we are going to apply this model on the CIFAR-10 image dataset that has 10 object categories. The Dataset In this experiment, we will be using the CIFAR-10 dataset that is a publically available image data set provided by the Canadian Institute for Advanced Research (CIFAR). It consists of 60000 32×32 colour images in 10 classes, with 6000 images per class. The 10 different classes represent airplanes, cars, birds, cats, deer, dogs, frogs, horses, ships, and trucks. There are 50000 training images and 10000 test images in this dataset. Implementation in Python We will import the library to download the CIFAR-10 data set. #Keras library for CIFAR-10 dataset from keras.datasets import cifar10 #Downloading the CIFAR dataset (x_train,y_train),(x_test,y_test)=cifar10.load_data() We will import the remaining libraries that are going to be required in our experiment. #importing other required libraries import numpy as np import pandas as pd from sklearn.utils.multiclass import unique_labels import os import matplotlib.pyplot as plt import matplotlib.image as mpimg import seaborn as sns import itertools from sklearn.model_selection import train_test_split from sklearn.metrics import confusion_matrix from keras import Sequential from keras.applications import VGG19 #For Transfer Learning from keras.preprocessing.image import ImageDataGenerator from keras.optimizers import SGD,Adam from keras.callbacks import ReduceLROnPlateau from keras.layers import Flatten,Dense,BatchNormalization,Activation,Dropout from keras.utils import to_categorical Here, we will split the downloaded dataset into training, test and validation sets. #defining training and test sets x_train,x_val,y_train,y_val=train_test_split(x_train,y_train,test_size=.3) Once split, we will see the shape of our data. It should be same as given in the dataset description at its parent website. #Dimension of the dataset print((x_train.shape,y_train.shape)) print((x_val.shape,y_val.shape)) print((x_test.shape,y_test.shape)) We need to do one hot encoding here because we have 10 classes and we should expect the shape[1] of y_train,y_val and y_test to change from 1 to 10 #One Hot Encoding y_train=to_categorical(y_train) y_val=to_categorical(y_val) y_test=to_categorical(y_test) After one hot encoding, we will ensure that we have obtained the required shape. #Verifying the dimension after one hot encoding print((x_train.shape,y_train.shape)) print((x_val.shape,y_val.shape)) print((x_test.shape,y_test.shape)) Here, we will perform the image data augmentation. This is the technique that is used to expand the size of a training dataset by creating modified versions of images in the dataset. First, we will define individual instances of ImageDataGenerator for augmentation and then we will fit them with each of the training, test and validation datasets. #Image Data Augmentation train_generator = ImageDataGenerator(rotation_range=2, horizontal_flip=True, zoom_range=.1) val_generator = ImageDataGenerator(rotation_range=2, horizontal_flip=True, zoom_range=.1) test_generator = ImageDataGenerator(rotation_range=2, horizontal_flip= True, zoom_range=.1) #Fitting the augmentation defined above to the data train_generator.fit(x_train) val_generator.fit(x_val) test_generator.fit(x_test) We will use the learning rate annealer in this experiment. The learning rate annealer decreases the learning rate after a certain number of epochs if the error rate does not change. Here, through this technique, we will monitor the validation accuracy and if it seems to be a plateau in 3 epochs, it will reduce the learning rate by 0.01. #Learning Rate Annealer lrr= ReduceLROnPlateau(monitor='val_acc', factor=.01, patience=3, min_lr=1e-5) Now, we will instantiate the VGG19 that is a deep convolutional neural network as a transfer learning model. Defining VGG19 as a Deep Convolutional Neural Network #Defining the VGG Convolutional Neural Net base_model = VGG19(include_top = False, weights = 'imagenet', input_shape = (32,32,3), classes = y_train.shape[1]) Now, we will define VGG19 as a deep learning architecture. For this purpose, it will be defined as a Keras Sequential model with several dense layers. #Adding the final layers to the above base models where the actual classification is done in the dense layers model= Sequential() model.add(base_model) model.add(Flatten()) Now, to add further layers, we need to see the dimension of our model. #Model summary model.summary() #Adding the Dense layers along with activation and batch normalization model.add(Dense(1024,activation=('relu'),input_dim=512)) model.add(Dense(512,activation=('relu'))) model.add(Dense(256,activation=('relu'))) model.add(Dropout(.3))ense(128,activation=('relu'))) #model.add(Dropout(.2)) model.add(Dense(10,activation=('softmax'))) #Checking the final model summary model.summary() As we have defined our model, now we need to initialize the hyperparameters that are required to train the model and then finally, we will compile our model. #Initializing the hyperparameters batch_size= 100 epochs=50 learn_rate=.001 sgd=SGD(lr=learn_rate,momentum=.9,nesterov=False) adam=Adam(lr=learn_rate, beta_1=0.9, beta_2=0.999, epsilon=None, decay=0.0, amsgrad=False) model.compile(optimizer=sgd,loss='categorical_crossentropy',metrics=['accuracy']) Now, we start training our VGG10, the deep convolutional neural network model. #Training the model model.fit_generator(train_generator.flow(x_train, y_train, batch_siz e= batch_size),epochs = epochs, steps_per_epoch = x_train.shape[0]\/\/batch_size, validation_data = val_generator.flow(x_val, y_val, batch_size = batch_size), validation_steps = 250, callbacks=[lrr], verbose = 1) As we can see in the above picture, we have achieved the training accuracy by 99.22% and validation accuracy by 85.41%. Now we will visualize the accuracy and loss during training. #Plotting the training and validation loss and accuracy f,ax=plt.subplots(2,1) #Loss ax[0].plot(model.history.history['loss'],color='b',label='Training Loss') ax[0].plot(model.history.history['val_loss'],color='r',label='Validation Loss') #Accuracy ax[1].plot(model.history.history['accuracy'],color='b',label='Training  Accuracy') ax[1].plot(model.history.history['val_accuracy'],color='r',label='Validation Accuracy') We will make image class predictions through this model using the test data set. #Making prediction y_pred=model.predict_classes(x_test) y_true=np.argmax(y_test,axis=1) Performance of VGG19 – The Deep Convolutional Neural Network Finally, we will visualize the classification performance on test data using confusion matrices. #Defining function for confusion matrix plot def plot_confusion_matrix(y_true, y_pred, classes, normalize=False, title=None, cmap=plt.cm.Blues): if not title: if normalize: title = 'Normalized confusion matrix' else: title = 'Confusion matrix, without normalization' #Compute confusion matrix cm = confusion_matrix(y_true, y_pred) if normalize: cm = cm.astype('float') \/ cm.sum(axis=1)[:, np.newaxis] print(\"Normalized confusion matrix\") else: print('Confusion matrix, without normalization') #print(cm) fig, ax = plt.subplots(figsize=(7,7)) im = ax.imshow(cm, interpolation='nearest', cmap=cmap) ax.figure.colorbar(im, ax=ax) # We want to show all ticks... ax.set(xticks=np.arange(cm.shape[1]), yticks=np.arange(cm.shape[0]), # ... and label them with the respective list entries xticklabels=classes, yticklabels=classes, title=title, ylabel='True label', xlabel='Predicted label') #Rotate the tick labels and set their alignment. plt.setp(ax.get_xticklabels(), rotation=45, ha=\"right\", rotation_mode=\"anchor\") # Loop over data dimensions and create text annotations. fmt = '.2f' if normalize else 'd' thresh = cm.max() \/ 2. for i in range(cm.shape[0]): for j in range(cm.shape[1]): ax.text(j, i, format(cm[i, j], fmt), ha=\"center\", va=\"center\", color=\"white\" if cm[i, j] > thresh else \"black\") fig.tight_layout() return ax np.set_printoptions(precision=2) First, we will see the exact number of correct and incorrect classification using the non-normalized confusion matrix and then we will see the same in percentage using the normalized confusion matrix. #Plotting the confusion matrix confusion_mtx = confusion_matrix(y_true, y_pred) #Defining the class labels class_names=['airplane', 'automobile', 'bird', 'cat', 'deer', 'dog', 'frog', 'horse', 'ship', 'truck'] # Plotting non-normalized confusion matrix plot_confusion_matrix(y_true, y_pred, classes = class_names, title='Confusion matrix, without normalization') #Plotting normalized confusion matrix plot_confusion_matrix(y_true, y_pred, classes = class_names, normalize = True, title = 'Normalized confusion matrix') As we can see by classifying images into 10 classes, the model has given a minimum accuracy of 72% and a maximum accuracy of 95%. We can further tune the training parameters and re-train our model to see any possible upscaling in the classification. But what we have got in this experiment is the standard one. Out of 10 classes, it has given less than 80% accuracy in classifying only for 3 classes and has given more than 90% accuracy in classifying images of 5 classes.","excerpt":"In this article, we will implement the multiclass image classification using the VGG-19 Deep Convolutional Network used as a Transfer Learning framework where the VGGNet comes pre-trained on the ImageNet dataset. For the experiment, we will use the CIFAR-10 dataset and classify the image objects into 10 classes. The classification accuracies of the VGG-19 model will be visualized using the non-normalized and normalized confusion matrices.","categories":["Deep Tech"],"tags":["Artificial Neural Network","big data machine learning","cifar-10","Convolution Neural Network","Convolutional Neural Network","Deep Learning","Image Classification","multiclass classification","object recognition","object store database","python visualize neural network","Transfer Learning","VGG19"],"author_name":"Dr. Vaibhav Kumar","publish_date":"2020-05-26T10:00:00","publication_year":"2020","word_count":1491,"keywords":["multiclass classification","deep learning","python visualize neural network","big data machine learning","Transfer Learning","object store database","Pandas","NumPy","image recognition","cifar-10","Convolution Neural Network","Convolutional Neural Network","Image Classification","machine learning","object recognition","AI","neural network","Artificial Neural Network","Seaborn","Matplotlib","VGG19","Keras","Deep Learning"],"extracted_tech_keywords":["AI","machine learning","deep learning","neural network","Keras","Pandas","NumPy","Matplotlib","Seaborn","image recognition"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/transfer-learning-for-multi-class-image-classification-using-deep-convolutional-neural-network\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":23228,"title":"Understanding ‘Self-learning’ Monte Carlo Method","content":"Complex statistical methods have been downscaled these days, thanks to advanced technology and better methods available to tackle the traditional statistical analysis. With machine learning algorithms and sophisticated hardware at the front, statistical computations require lesser time and fewer computing resources. In addition, statistical software packages such as MiniTab, SPSS, JMP and so on, are fuelling novel computing needs to determine statistical results. One such statistical method that has seen a face-lift is the Monte Carlo method (also known as Monte Carlo analysis or Monte Carlo Simulation, in different fields of study). The changes in the method mainly has ML aspects involved to deal with the setbacks with regard to the conventional method. In this article, we explore the novel method called ‘Self-Learning’ Monte Carlo (SLMC) method, developed by academics at Chinese Academy of Sciences in collaboration with Massachusetts Institute of Technology, which practically observes an improvement in the Monte Carlo method\/simulation. What is Monte Carlo method? Originally developed by Stanislaw Ulam, a mathematics professor from Poland, Monte Carlo(MC) method or simulation involves measuring the relationship between two or more mathematical variables using random sampling. This process usually involves error, which are however significantly reduced if large samples are taken in the context. In other words, the method incorporates the process of integration of random variables at specific regions (called as intervals) to follow a deterministic approach. Machine Learning in Monte Carlo method The basis for SLMC is to perform an “update” simulation to an existing MC configuration leading to a large set of configurations so that it serves as training data for further simulation. This means, the simulation learns on its own based on previous input configurations. The “update” is carried out using the Markov Process for MC method where the probability of transition is from A to B, requiring a detailed balance principle to be followed. The formula is given by P(A →B)\/ P(B→A) = W(B)\/ W(A) where, W stands for the probability distribution of configurations. The “update” followed throughout the study will be labelled ‘local’ because a specific instance (called ‘site’) in the configuration is chosen and a new configuration is obtained by changing the variable in the site. Secondly, the new configuration is selected based on the detailed balance principle mentioned earlier. If it holds good for the principle, the new configuration is updated in the Markov chain, else the same configuration is kept and replicated. This way, SLMC can work with statistical models of most popular types by learning data. SLMC Procedure SLMC follows a 4-step procedure. Perform a trial simulation with the help of ‘local’ update which leads to a set of configurations which act as training data. Select a Hamiltonian operator H  and allow it to learn according to configurations provided to give out an effective Hamiltonian Heff Develop simulation moves with regard to effective Hamiltonian Heff. Observe whether the simulation moves hold good with respect to the original Hamiltonian H. In the above steps, Steps 1 & 2 form the self-learning components and Steps 3 & 4 are performed repeatedly to observe variations in the actual MC simulation. The implementation in the paper shows models to observe ferromagnetic properties as part of statistical mechanics, with regard to properties such as critical temperature and so on. The simulation is carried to check the interaction among the four orbital spins mentioned in the Hamiltonian. The results show the simulation process is faster by 10 times than the normal. All of the results are plotted with respect to local update using autocorrelation time which depicts how well the MC configurations adhere to the Markov Chain process. Throughout the iterative process, the data points of the spin correlation fall close to the self-learning fit depicting that the simulation yields positive and valid output (energy difference, in this case)  and also stick to the configurations obtained in the SLMC model. Conclusion: The approach and application in the study was limited to spin systems (atomic physics), specifically ferromagnetism. The authors also contend that SLMC may see improvements if newer ML techniques are utilised for simulation. The study also emphasized the need for a Hamiltonian operator, which is restricted only to quantum mechanics. With ML, this can be eliminated and expanded to other areas of study. Consequently, this leads to bridging the gap between theoretical studies in developing a model against the computational performance in practical.","excerpt":"Complex statistical methods have been downscaled these days, thanks to advanced technology and better methods available to tackle the traditional statistical analysis. With machine learning algorithms and sophisticated hardware at the front, statistical computations require lesser time and fewer computing resources. In addition, statistical software packages such as MiniTab, SPSS, JMP and so on, are […]","categories":["Deep Tech"],"tags":["Advanced Analytics"],"author_name":"Abhishek Sharma","publish_date":"2018-04-03T11:37:17","publication_year":"2018","word_count":727,"keywords":["Replicate","Go","machine learning","TPU","programming_languages:R","AI","ML","programming_languages:Go","SLM","R","Advanced Analytics"],"extracted_tech_keywords":["AI","machine learning","ML","SLM","TPU","R","Go","Replicate","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/understanding-self-learning-monte-carlo-method\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":41132,"title":"5 Popular AI-Powered Test Automation Tools","content":"Over the years, the software development domain has evolved extensively — whether it’s about the development speed or the delivery speed. And with that the rest of the software development process has also gone through a significant transformation, adopting a new DevOps culture that prioritizes concepts that support continuous delivery. And one of the concepts is Test automation. Today, even the general test automation scenario is also changing with the advent of AI in the domain. If you want to succeed in any test automation,  the right tool with the right technology is imperative. In this article, we list 5 most popular AI-powered test automation tools that go beyond the industry-standard Selenium. Test Automation With Selenium When we talk about test automation, Selenium is one of the best tools that comes to our heads first. As the official website of Selenium reads, “Selenium automates browsers. That’s it! What you do with that power is entirely up to you,” the tool is definitely to one job. But talking in terms of test automation, this open source tool that is licensed under Apache 2.0, a portable framework for automating web applications for testing purposes. It deploys on Windows, Linux, and macOS platforms and provides a playback tool for authoring functional tests without the need to learn a test scripting language (Selenium IDE). That is not all, this test automation tool has the support of some of the largest browser vendors who have taken or are taking steps to make Selenium the core technology in their browser automation tools, APIs and frameworks. 5 Popular AI-powered tools for test automation Selenium is definitely one of the best test automation tools, but today AI is playing a major role in the domain and there are plenty of other tools available that are AI-powered. TestCraft TestCraft is an AI-powered test automation platform for regression and continuous testing that works on top of Selenium. It is also used for monitoring of web applications. The role of artificial intelligence (AI) technology is to eliminate maintenance time and cost by automatically overcomes changes in the app. And the best thing about TestCraft is that testers can visually create automated, Selenium-based tests using a drag and drop interface, and run them on multiple browsers and work environments, simultaneously. No coding skills required. TestCraft is not free; however, one can get a free trial. Click here to get your free trial. Applitools Applitools is an application visual management and AI-powered visual UI testing and monitoring software. It provides an end-to-end software testing platform powered by Visual AI and can be used by professionals in engineering, test automation, manual QA, DevOps, and Digital Transformation teams. Also, the AI and machine learning algorithm are entirely adaptive — it scans the apps’ screens and analyze them like the human eye and brain, but with the power of a machine. You can request for demo here. Functionize Functionize is a cloud-based automated testing technology that is used for functional, performance, and load testing — one stop shop for all the mentioned testing. Also, this tool uses machine learning and artificial intelligence to speed up test creation, diagnosis, and maintenance. One of the best features of this tool is that you don’t have to think a lot before carrying out a test — all you have to do is type what you want to test in plain English and NLP creates functional test cases. It also executes thousands of tests in minutes from all desktop and mobile browsers. If you are looking for a test automation tool, then Functionize is definitely worth a try. Click here to get or demo. Sauce Labs This is another popular cloud-based test automation tool that leverages machine learning and AI. Sauce Labs support a comprehensive list of browsers and operating systems, mobile emulators and simulators and mobile devices, and at the speed that its users need to test their apps. It also claims to be the world’s largest continuous testing cloud, that offers over 800 browser and operating system combinations, 200 mobile emulators and simulators and thousands of real devices. Talking about the pricing, the Live comes for $19\/month, automated $149\/month, and unlimited automated $298\/month. However, you can get a 14 days free trial here. Testim Automated functional testing tool, Testim uses artificial intelligence and machine learning to speed-up the authoring, execution, and maintenance of automated tests. Speaking of support, the tool run on different browsers and platforms such as Chrome, Firefox, Edge, IE, Safari, and Android. Testim comes in two plans — basic and pro. The basic plan is free and has very limited features. While on the other hand, the pro version supports everything.","excerpt":"Over the years, the software development domain has evolved extensively — whether it’s about the development speed or the delivery speed. And with that the rest of the software development process has also gone through a significant transformation, adopting a new DevOps culture that prioritizes concepts that support continuous delivery. And one of the concepts […]","categories":[],"tags":["automation testing","automation tools","devops tools","software automation testing"],"author_name":"Harshajit Sarmah","publish_date":"2019-06-21T12:26:32","publication_year":"2019","word_count":776,"keywords":["Go","artificial intelligence","automation testing","machine learning","AI","devops tools","Git","RAG","software automation testing","NLP","Aim","automation tools","DevOps","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","NLP","Aim","RAG","R","Go","Git","DevOps"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/5-popular-ai-powered-test-automation-tools\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10083271,"title":"Intel Splits Graphics Group Into Two","content":"Intel Corporation announced recently that it would split its AXG graphics group into two: Accelerated Computing Systems and Graphics (AXG) business unit and Data Center and AI (DCAI) business unit. The move aligns them with being best suited for competing with NVIDIA and AMD. The consumer-focused portion of the AXG business unit will be combined with Intel’s Client Computing Group (CCG), which focuses on creating platforms around the company’s CPU products. On the other hand, the teams managing the data centre and supercomputing GPUs, like the Ponte Vecchio and Rialto Bridge products, will be transferred to the Data Center and AI (DCAI) business unit. Additionally, Raja Kodouri, who previously led the AXG division, will now be the Chief Architect to focus on their expanding efforts in the fields of CPU, GPU and AI, as well as in accelerating high-priority technical programs. Naveed Sherwani, President and CEO at RapidSilicon, said, “This is a good move. Raja is a solid leader with great credentials. We hope this move helps Intel to improve its culture. The priority should be on listening to customers, getting products out on time and rewarding performance and merit inside the company.” Intel recently also announced the launch of Data Center GPU Max Series, the industry’s highest-density processor, packing over 100 billion transistors on 47 active tiles and up to 128 Xe-HPC cores. In addition, the whole Max series—both its CPU and GPU—is powered by oneAPI, an open-source programming model that enables developers to leverage several accelerated architectures.","excerpt":"The move aligns them with being best suited for competing with NVIDIA and AMD.","categories":["AI News"],"tags":["CPU","GPU","Intel"],"author_name":"Ayush Jain","publish_date":"2022-12-23T18:29:54","publication_year":"2022","word_count":249,"keywords":["CPU","Rapids","Go","API","programming_languages:R","AI","programming_languages:Go","RAG","R","Intel","GPU"],"extracted_tech_keywords":["AI","Rapids","RAG","R","Go","API","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/intel-splits-graphics-group-into-two\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10124363,"title":"Meta Launches AI Assistant in India Across WhatsApp, Facebook, Instagram","content":"Meta has announced the launch of its AI assistant, Meta AI, across WhatsApp, Facebook, Instagram and Messenger in India. The AI-powered tool, built on Meta’s Llama 3, aims to help users with tasks like planning, learning, creating content and connecting with others. One of the key advantages of Meta AI is its seamless integration into Meta’s popular apps, making it easily accessible to users on both Android and iOS devices as well as the web, according to Ryan Cairns, VP of Engineering at Meta. “Our apps provide both the biggest distribution and the most availability to the end user,” Cairns said, contrasting Meta AI with competitors like Google Assistant and Siri which are more platform-specific. Meta AI can assist with a wide range of use cases that users in India have already experienced during the pilot, such as gathering information, learning, coding help, generating social media content like captions and Instagram threads, and even creating stylised brand logos. The AI’s “imagine” feature also enables users to generate and share unique images from text prompts and modify or animate them. Addressing concerns about data privacy and potential misuse, Cairns stated that while using public information to train AI models is an industry-wide practice, building AI responsibly with a focus on safety is a top priority for Meta. He noted that Meta AI’s images are stylised rather than photorealistic to prevent deepfakes, and that the models are updated every two weeks based on user feedback to improve accuracy and prevent misinformation. Meta AI is currently available in English in India, with no other language options at this time. The company has expressed excitement about introducing this next-generation AI assistant to a wider audience and its potential positive impact on people’s daily lives.","excerpt":"The AI-powered tool, built on Meta’s Llama 3, aims to help users with tasks like planning, learning, creating content and connecting with others.","categories":["AI News"],"tags":["Llama 3","Meta AI"],"author_name":"Shyam Nandan Upadhyay","publish_date":"2024-06-24T14:21:50","publication_year":"2024","word_count":290,"keywords":["Llama 3","Go","Meta AI","programming_languages:R","AI","ML","deepfakes","programming_languages:Go","Aim","llm_models:Llama","R"],"extracted_tech_keywords":["AI","ML","Meta AI","Aim","R","Go","deepfakes","llm_models:Llama","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/meta-launches-ai-assistant-in-india-across-whatsapp-facebook-instagram\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10119304,"title":"Is Intel Living in Denial?","content":"Certainly, Intel is sorting out its AI strategy by investing in AI accelerators and expanding on its existing customers. One can say, apart from AMD building the whole architecture, none of the OEM providers can build a PC, or AI PC, without utilising the blue-chip company’s CPU. But talking about the AI strategy again, is Intel in denial? In a recent interview, Patrick Gelsinger, the CEO of Intel, confirmed that the company was on track with its investment, when it comes to AI. Even though the demand for its Gaudi accelerators are rising, it is minuscule compared to AMD and NVIDIA’s GPUS. Gaudi is on track for around 500 million of sales in 2024. This is several times less than AMD’s MI300X’s demand for $3.5 billion and the NVIDIA’s H100 and H200’s $40 billion demand. When probed on this, Gelsinger said Intel’s Vision 2024 event saw the announcement of 20+ customers for its AI accelerator, which highlights the company’s strength. Is the CPU approach for AI viable? “We’re really starting to see that pipeline of activity convert,” said Gelsinger. Some say that Intel is completely living in denial with its current AI strategy. One of the biggest reasons for this is that Intel has always been the biggest proponent of pushing CPUs for AI workloads. But Intel’s biggest bet is on edge use cases, which is through its Core mobile processors and Xeon processors. With smaller language models increasing in the AI space, these mobile processors are currently in the perfect space. While NVIDIA is also delivering this with Arm-based processors on mobile devices, these are still incapable of running on laptops without an Intel CPU. Recently, a team of researchers at Hugging Face published a blog that said they were able to train an LLM based on Microsoft Phi-2 model on Intel Meteor Lake, which is now renamed to Core Ultra, specifically built for high performance laptops consisting of 16 cores. The interesting part is that it comes with an integrated GPU, called iGPU with 16 Xe Vector Engines. Moreover, Intel’s introduction of the Neural Processing Unit (NPU) marks a significant milestone for its architectures. This dedicated AI engine is tailored for efficient client AI, allowing it to handle demanding AI computations with greater efficiency compared to using the main CPU or integrated graphics (iGPU). By offloading AI tasks to the NPU, users can free up the main CPU and graphics for other tasks, resulting in improved power efficiency. The blog ends with: “Thanks to Hugging Face and Intel, you can now run LLMs on your laptop, enjoying the many benefits of local inference, like privacy, low latency, and low cost.” This is what Intel is aiming for as well. Furthermore, when it comes to India, Intel has a buzzing partner ecosystem with Infosys, Bharti Airtel, Ola Krutrim, Zoho, and L&T announcing partnerships with the company. Intel is betting big on India, which is not that new. Providing cheaper alternatives when it comes to data centres and powering enterprise solutions, Intel has always been the go to choice for Indian companies. “AI does not just require big GPUs to solve the problem. There are a lot of different models that can run on Xeon. Innovation at scale can happen with Xeon. We are working with several large customers. Gaudi 2 is available, Gaudi 3 comes in the second half. You will see some of those products coming into India through these customers as well,” said Santosh Viswanathan, VP and MD of Intel in India. Ushering the era of AI PCs One can even say that Intel is living in the future when it comes to building AI PCs. Though Jensen Huang, the NVIDIA chief, has said that everyone would be a gamer in the future using their GPUs, Intel’s approach towards building AI PCs makes a lot of sense as well. Intel also has a bunch of partnerships for its AI PC goals. At the AI Everywhere event in December, Gelsinger and his team announced the launch of Intel Core Ultra and Intel Arc GPUs for pushing the goal of making every PC in the world an AI PC. This would be achieved through its partnership with Dell, Lenovo, HP, Supermicro, and Microsoft, for bringing the hardware onto their devices. This was further solidified at the Intel Vision event in April. Intel anticipates shipping 40 million AI PCs in 2024, featuring over 230 designs spanning from ultra-thin PCs to handheld gaming devices. Looking ahead, Intel’s roadmap includes the launch of the next-generation Intel Core Ultra client processor family, codenamed Lunar Lake, in 2024. This lineup is projected to deliver more than 100 platform tera operations per second (TOPS) and over 45 neural processing unit (NPU) TOPS, ushering in a new era of AI-centric computing. The rivalry between Intel and NVIDIA, encompassing both CPUs and GPUs, is poised to intensify, potentially reshaping the landscape of AI and HPC. It’s widely believed that Intel is the best CPU to buy, and NVIDIA, the best GPU to buy. But this might take a turn soon given that the conversation about computing has almost shifted around AI. There are no PCs without Intel – that’s for sure.","excerpt":"Everyone is doubting Intel’s AI strategy. But for the company, it is working well.","categories":["AI Features"],"tags":["Pat Gelsinger"],"author_name":"Mohit Pandey","publish_date":"2024-04-30T16:46:40","publication_year":"2024","word_count":864,"keywords":["Go","Hugging Face","API","programming_languages:R","AI","Pat Gelsinger","innovation","Aim","ViT","ai_frameworks:Hugging Face","R"],"extracted_tech_keywords":["AI","Aim","Hugging Face","R","Go","API","ViT","innovation","ai_frameworks:Hugging Face","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/is-intel-living-in-denial\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10053403,"title":"IBM Enhances The Natural Language Processing Capabilities Of Watson Discovery","content":"IBM announced upgrades to IBM Watson Discovery‘s natural language processing (NLP) capabilities. These planned upgrades will benefit corporate users in financial services, insurance, and legal services. Through the discovery of insights, these upgrades are leveraged to enhance customer service and accelerate business operations. Businesses are increasingly turning to natural language processing (NLP) and machine learning to assist them in sifting through increasing amounts of documents and data sets in various formats. By leveraging AI to extract insights from documents, business users can reduce research time. Moreover, it can enable employees to make more fact-based decisions, especially when performing complex, time-sensitive tasks such as processing insurance claims, conducting financial analyses and so on. The new tools released today by IBM are intended to make it easier for Watson Discovery users to swiftly tailor the underlying NLP models to their business’s unique language. As a result of IBM Research’s advances in NLP, business users can teach Watson Discovery to assist them in reading, comprehending, and surfacing; in addition, more precise insights from massive quantities of complicated, industry-specific documents. Pre-trained Model Watson Discovery’s Smart Document Understanding feature is currently available in the Plus, Enterprise, and Premium plans. This model is designed to automatically comprehend a document’s visual structure and layout without the assistance of a developer or a data scientist. In addition, this model enables users to locate previously hidden easily or difficult-to-find information, such as text embedded in complicated table structures or images. Automatic Pattern Recognition IBM has launched a beta version of a new advanced pattern generation function in the Plus, Premium, and Enterprise plans. The feature aims to assist customers in swiftly identifying business-specific language patterns inside their documents. This feature is critical for jobs such as reviewing large volumes of contracts or financial reports, which may present the same information, such as an increase or reduction in income, in a variety of different formats or using a variety of different terms. IBM Research developed it to aid in the effective labelling of data and training of models. It is designed to begin learning text patterns from as little as two examples and then refine the pattern based on user feedback. This research enables users to train a model more quickly by eliminating manual and time-consuming procedures such as creating rules and expressions. Advanced Natural Language Processing Training NLP models is a time-consuming process that requires extensive data preparation, labelling, and orchestration. Frequently, models trained on generic data sets fail to obtain the correct data. IBM simplifies this process with a new custom entity extractor feature, which is now available in beta for Watson Discovery Premium users. This feature reduces data preparation effort, simplifies labelling with active learning and bulk annotation capabilities, and enables simple model deployment, accelerating training time. “The continuous stream of innovation flowing into IBM Watson from IBM Research is why global businesses in financial services, insurance, and legal services rely on IBM to assist them in identifying emerging business trends, improving operational efficiency, and empowering their employees to uncover new insights,” said Daniel Hernandez, IBM’s General Manager of Data and AI. “With the pipeline of natural language processing improvements we’re introducing to Watson Discovery, organisations can continue to differentiate the signal from the noise and provide better service to their customers and workers.” Along with the new functionalities, IBM demonstrates how firms in the legal, financial, and insurance industries leverage Watson Discovery’s existing capabilities to automate and revolutionise business processes. For more information on Watson Discovery, read here. For more information about IBM Watson, read here.","excerpt":"IBM has announced upgrades to Watson Discovery’s natural language processing capabilities.","categories":["AI Features"],"tags":["AI (Artificial Intelligence)","Financial Services","ibm research","insurance","Machine Learning","Natural Language Processing"],"author_name":"Dr. Nivash Jeevanandam","publish_date":"2021-11-12T13:00:00","publication_year":"2021","word_count":590,"keywords":["insurance","machine learning","programming_languages:R","AI","ibm research","Natural Language Processing","innovation","Machine Learning","RAG","Financial Services","NLP","Aim","GAN","R","AI (Artificial Intelligence)","active learning"],"extracted_tech_keywords":["AI","machine learning","NLP","Aim","RAG","R","GAN","active learning","innovation","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/ibm-enhances-nlp\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10057528,"title":"Impact of AI Voice Dubbing on Voice Over Industry","content":"The Israeli startup DeepDub claims that it can dub films, television shows, and video games while retaining the pitch and tone of the original actors in any given process. This allows them to recreate the voices of famous actors in other languages, as well as provide an efficient alternative to hiring voice actors to dub individual films or television shows. Another company, the UK-based Synthesia, is using AI-based video-editing software to improve the quality of dubbing by tweaking the actor’s mouth movements to fit the local language they are speaking. Meanwhile, WellSaid Labs—a Seattle-based startup—is offering AI voices with different voice personalities (such as “energetic and insightful” or “poised and expressive”) to its clients. What’s the technology behind these products? While synthetic voices have been around for a while, recent breakthroughs in deep learning have made replicating the subtleties of human speech easier than ever. Previously, synthetic voices were just words glued together to produce a clunky, robotic effect. Nowadays, voice developers no longer need to provide the exact pacing, pronunciation, or intonation of the generated speech themselves. Instead, they simply fill hours of audio into an algorithm that deciphers those patterns for itself. These voices can change in style and emotion, take pauses and breathe in the correct places, and learn the tone and patterns of the speech of an actor. The voice engine constructed by WellSaid consists of two primary deep-learning models: the first predicts what a speaker will sound like from a passage of text, and the second fills in the details—including breaths and the way the voice would resonate in a particular environment. Traditionally, dubbing is done without making any alterations to the source video—which means that local script writers have to struggle to make the dialogue match the action taking place on the screen, and voice actors have to deliver their dialogue perfectly to make sure the dub doesn’t feel off. The time to write these scripts and to dub the movie is significantly shorter when artificial intelligence is used to simply change the lip movement of the actor post production (which also makes the scene more realistic to a viewer). While Synthesia’s deep learning algorithm has perfected making it look like an actor’s mouth is forming foreign words, it still isn’t capable of manufacturing the facial idiosyncrasies of individual actors—which would likely change depending on the language the actor is speaking in. Why does this innovation matter? AI voices are increasing in popularity amongst brands who want to maintain a consistent sound in millions of interactions with customers. Its popularity has also increased due to the omnipresence of smart speakers, automated service, and digital assistants today—that require brands to produce more than a hundred hours of audio every month. It’s also getting increasingly vital for film and television producers to take their products to international markets. China’s influence on the global box office has grown exponentially in recent years; in 2020, a vast majority of Netflix’s subscriber gains were from outside the United States; and one-third of Disney+ subscribers are in India. Traditional dubbing is an expensive and time-consuming process, and AI voices provide a cheap and scalable alternative to them. With DeepDub, an eight-episode television series—that would ordinarily take 14-16 weeks to dub—can be dubbed in just six weeks. This quicker turnaround for dubbed content could also help reduce piracy—which often occurs because of the time window between content released in the United States and other regions. The bigger the time window is, the greater the threat of piracy.","excerpt":"The Israeli startup DeepDub claims that it can dub films, television shows, and video games while retaining the pitch and tone of the original actors in any given process. This allows them to recreate the voices of famous actors in other languages, as well as provide an efficient alternative to hiring voice actors to dub […]","categories":["IT Services"],"tags":["synthesia"],"author_name":"Srishti Mukherjee","publish_date":"2022-01-03T12:00:52","publication_year":"2022","word_count":586,"keywords":["synthesia","Go","artificial intelligence","AI","innovation","Scala","Git","Aim","deep learning","ViT","R"],"extracted_tech_keywords":["AI","artificial intelligence","deep learning","Aim","R","Go","Scala","Git","ViT","innovation"],"url":"https:\/\/analyticsindiamag.com\/it-services\/now-the-voice-dubbing-industry-is-being-disrupted-by-ai\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":33283,"title":"Can We Trust k-fold Cross Validation For Financial Modelling?","content":"Machine learning is not a buzzword anymore, at least not in the case of financial modelling. Fraud prevention, algorithmic trading, digital assistants and risk management are some of the areas where machine learning has found its niche. The application of ML models in finance is unlike any other industries because the decisions taken either pay the dividends immediately or plummet catastrophically. Many financial companies need data engineering, statistics and visualization tools to meet their ends. With machine learning, the models can be retrained repeatedly until the best solution emerges. That said, there is no universal machine learning solution for all the business problems. Source: Financial ML via n-ix This graphic illustrates the confidence of financial institutions in imbibing machine learning methodologies. Talking about methodologies there are many statistical methods and frameworks that help in building models. For instance, validation techniques are widely popular to assess the accuracy of the models; cross-validation being the popular choice. Cross-validation is used for determining the generalization error in a machine learning algorithm to prevent overfitting. But, in the case of financial models, overfitting does take place and can go undetected by CV. Moreover, hyper-parameter tuning can contribute to overfitting. Such fundamental errors in a model can still get passed through owing to its overfitting. While the forecasting power is reduced to null. A Quick Recap Of CV CV splits the dataset into two sets: the training set and the testing set. Each observation in the complete dataset belongs to one, and only one, set. This is done as to prevent leakage from one set into the other since that would defeat the purpose of testing on unseen data. There are many alternative CV schemes, of which one of the most popular is k-fold CV. It works as follows: The dataset is partitioned into k subsets. For i = 1,…,k The ML algorithm is trained on all subsets excluding i. The fitted ML algorithm is tested on i. The outcome from k-fold CV is a k x 1 array of cross-validated performance metrics. For example, in a binary classifier, the model is deemed to have learned something if the cross-validated accuracy is over 1\/2, more than what we would achieve by tossing a fair coin. How Well Does CV Fair With Finance One reason k-fold CV fails in finance is that observations cannot be assumed to be drawn from an IID (Independent and Identically Distributed)processes. A second reason for CV’s failure is that the testing set is used multiple times in the process of developing a model, leading to multiple testing and selection bias. So, when there is an overlap between training and testing datasets, there will be some leakage. The problem is leakage in the presence of irrelevant features, as this leads to false discoveries. Problems With Sklearn’s Cross-Validation Sci-kit learn is the most popular ML library for implementing cross-validation. One of the many upsides of open-source code is that you can verify everything and adjust it to your needs. In Advances of financial machine learning, Marcos Lopez de Prado lists the following two problems with sklearn: Scoring functions do not know classes_, as a consequence of sklearn’s reliance on numpy arrays rather than pandas series: https:\/\/github.com\/scikit-learn\/scikit-learn\/issues\/6231 cross_val_score will give different results because it passes weights to the fit method, but not to the log_loss method: https:\/\/github.com\/scikit-learn\/scikit-learn\/issues\/9144 How To Reduce The Leakage Drop from the training set any observation i where Yi is a function of information used to determine Yj, and j belongs to the testing set. Avoiding overfitting of the classifier. Early stopping of the base estimators. Bagging of classifiers, while controlling for oversampling on redundant examples, so that the individual classifiers are as diverse as possible. Set average uniqueness. Apply sequential bootstrap An Alternative In The Form Of Purged K-Fold CV One way to reduce leakage is to purge from the training set all observations whose labels overlapped in time with those labels included in the testing set; called “purging.” If no training observations occur between the first and last testing observation, then purging can be accelerated with a pandas series with a single item, spanning the entire testing set. The larger the number of testing splits, the greater the number of overlapping observations in the training set. In many cases, purging is enough to prevent leakage. The performance will improve when the model is allowed to recalibrate more often. Future Direction The number of open-source machine learning algorithms and tools to curate financial data are increasing fast. And, with increasing interests of the financial institutions in AI, the funds allocated will increase, which in turn will enable more methodologies to developed. The advantage with this industry is its quantitative nature and its large repository of historical data which is exactly what a machine learning model needs. Neglecting the advancements and latching on to conventional methods will prove costly in the future.","excerpt":"Machine learning is not a buzzword anymore, at least not in the case of financial modelling. Fraud prevention, algorithmic trading, digital assistants and risk management are some of the areas where machine learning has found its niche. The application of ML models in finance is unlike any other industries because the decisions taken either pay […]","categories":["AI Features"],"tags":["Data loss prevention"],"author_name":"Ram Sagar","publish_date":"2019-01-11T09:07:12","publication_year":"2019","word_count":811,"keywords":["Go","scikit-learn","NumPy","machine learning","AI","ML","RAG","Ray","Data loss prevention","R","Pandas"],"extracted_tech_keywords":["AI","machine learning","ML","Ray","scikit-learn","Pandas","NumPy","RAG","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/can-we-trust-k-fold-cross-validation-for-financial-modelling\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":46390,"title":"Dozee: An AI-Powered Bedsheet From IIT Grads Can Predict Health Outcomes","content":"A group of IIT graduates has built up an AI fueled sensor sheet by the name of Dozee that can be put under the bed to identify health indicators, for example, heart condition, breath, sleep and stress parameters with high accuracy like in a hospital setting. The gadget takes a shot at an innovation known as Ballistocardiography (BCG), which includes the measurement of the heart muscle movement. The BCG technology can be stretched out to track sleep movements concerning breathing, body movement, coughing, snoring, etc. Mudit Dandwate, CEO and co-founder of Dozee says its algorithm have been trained over billions of data points and has resulted in 98.4 percent precision in estimating heartbeat and breath. The gadget is controlled by AI engine that analyses a person’s wellbeing information to take note of the risk deviations from solid baselines and recognize any early indications of health issues that may face a person. The group reports that the slight sheet comprises of micro-electromechanical system (MEMS) based vibroacoustic sensors, and accompanies data collection and cloud based app to send over the data for analytics. The Dozee sensor sheet when set underneath the sleeping cushion under the chest of the client catches small scale vibrations delivered by the body each time the heart siphons blood, along with breath movements like exhalation, muscle jerks, tremors or other natural body movements. These signs are then absolutely converted over into bio-markers by refined AI calculations. Biomarkers incorporate pulse, breath rate, sleep stages, apnea episodes while breathing, heart palpitations and pulse fluctuations, the researchers reported. The data is then disseminated by the cloud servers to the cell phone application and web-applications to give remote access to clients, and their primary care physicians, as indicated by the group. It likewise cautions clients, overseers and the specialists to set up custom alarms for conditions, for example, arrhythmia, apnea. The sheet costs $100 (₹7000) for consumers to buy.","excerpt":"A group of IIT graduates has built up an AI fueled sensor sheet by the name of Dozee that can be put under the bed to identify health indicators, for example, heart condition, breath, sleep and stress parameters with high accuracy like in a hospital setting. The gadget takes a shot at an innovation known […]","categories":["AI News"],"tags":[],"author_name":"Vishal Chawla","publish_date":"2019-09-25T18:06:00","publication_year":"2019","word_count":317,"keywords":["Go","programming_languages:R","AI","innovation","programming_languages:Go","analytics","R"],"extracted_tech_keywords":["AI","analytics","R","Go","innovation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/dozee-a-bed-sheet-from-iit-grads-can-precisely-predict-health-outcomes-using-ai\/","complexity_score":3,"technical_depth":7,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10087537,"title":"Bing Chatbot’s Inappropriate Responses Leave Users Outraged","content":"Microsoft Bing’s chatbot is in an exploratory stage and users are baffled by the beta version. Bing has seemingly gone rogue with the kind of responses it has generated. A user had posted a series of conversation screenshots where the chatbot blamed the user for providing wrong information, alleged the phone had malware and that the person had been “unreasonable and stubborn“. The chatbot has also shared “gossip” from Microsoft office. The Bing chatbot was prompted with the question of the timings of the latest Avatar movie on February 12, 2023, and the bot incorrectly stated that the movie was “not released yet”. However, the conversation between the two made the chatbot almost believable to be human-like and deflect criticism. There have been multiple instances when the chatbot has shown human-like responses, making it sound angry and depressed. Another user illustrated how the chatbot went on a rampage in trying to prove that the bot has sentiments and emotions, resembling a frustrating rant. Users have been going on a wild search to see just how bizarre the bot can get. A conversation that was supposed to make up scenarios, when prompted to be gossipy and share details from the development days, mentioned certain incidents it “witnessed through the webcam of the developer’s laptop”. The behaviour of the chatbot is expected to be wonky at this stage. The latest generation of AI chatbots is expected to behave in an undesirable manner as it is still in the learning stage, and algorithmic restraints and filters still need to be developed. The AI system scraps the vast information available on the web, where many fictitious plots exist for AI sentients. These could also be reflected in the responses that come about. Considering how Bing chatbot is in its early stage, users can manipulate the chatbot by giving prompts that churn negative responses and make the bot go off the rails. This alleged behaviour questions the efficiency of the chatbot on a search engine that has finally reached the stage of being seriously considered a worthy opponent of Google Search.","excerpt":"There have been multiple instances when the chatbot has shown human-like responses, making it sound angry and depressed.","categories":["AI News"],"tags":["AI Chatbot","Bing","Google","Microsoft"],"author_name":"Vandana Nair","publish_date":"2023-02-16T17:57:02","publication_year":"2023","word_count":346,"keywords":["Go","programming_languages:R","AI","AI Chatbot","Bing","chatbots","programming_languages:Go","Google","Rust","R","programming_languages:Rust","Microsoft"],"extracted_tech_keywords":["AI","chatbots","R","Go","Rust","programming_languages:R","programming_languages:Go","programming_languages:Rust"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/bing-chatbots-inappropriate-responses-leave-users-outraged\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":46744,"title":"Top 5 Machine Learning Problems Practice Platforms","content":"The best way to learn anything is by practising it. A number of theories and tutorials are available online as well as offline to learn machine learning. But one cannot truly learn until and unless one truly gets some hands-on training to learn how to actually solve the problems. In this article, we list down five online platforms where a machine learning enthusiast can practice computational applications. 1. MachineHack MachineHack Makes A Comeback With A Faster & Robust Platform For Data Science Hackathons MachineHack is an online platform by Analytics India Magazine for Machine Learning Hackathons where one can test and practice their machine learning skills. In this platform, a beginner can learn and practice how to deploy popular machine learning algorithms such as Linear Regression, Multiple Linear Regression, Support Vector Regression, Extreme Gradient Boosting Classification, Naive Bayes, K-Nearest Neighbours, and other such algorithms on datasets provided by the site. The coolest thing about this platform is one can practice as many times as he wants and there is no limit to the practice sessions. Click here to know more. 2. CloudXLab CloudXLab is an online cloud platform that provides online video courses, auto-assessment tests, BootML which is the UI-based machine learning model code generator as well as 24 hours of lab access with the Jupyter environment. In this platform, you can learn paid online courses like Big data with Hadoop and Spark, Machine Learning Specialisation, Python for Data Science, Deep learning and much more. There are also free tutorials available on Linux basics, introduction to Python, NumPy for machine learning and much more. You can either enroll for the course and lab or you can enrol only for practicing in the lab. There are packages for lab enrollment which includes the duration of one to six months. Click here for more. 3. Google Colab Google Colaboratory is a platform built on top of the Jupyter Notebook environment which runs entirely on Google Cloud Platform (GCP). This platform provides GPU which is free of cost and supports Python 2 and 3 versions. With the help of Colab, one can not only improve machine learning coding skills but also learn to develop deep learning applications. You can also learn to work with popular deep learning libraries such as Keras, TensorFlow, OpenCV and others. With Colaboratory you can write and execute code, save and share your analyses, and access powerful computing resources, all for free from your browser. Click here to know more. 4. Kaggle Kaggle is a no-setup, customisable, Jupyter Notebooks environment by Google. This platform is very much similar to Google Colab in the aspect that both the platforms provide free GPUs along with a large community of published data and code. With over 19,000 public datasets and 200,000 public notebooks, this platform provides one of the best places to practice data science computational problems. This cloud computational environment supports Python 3 and R and enables reproducible and collaborative analysis where one can explore and run machine learning codes seamlessly. Click here to know more. 5. OpenML OpenML is an open, collaborative, and automated machine learning environment which includes several specific features such as find or add data to analyse, download or create computational tasks, find or add data analysis flows and much more. This open science platform for machine learning is a cross-platform programming environment for sharing and organising data, machine learning algorithms, and experiments. It is designed to create a frictionless networked ecosystem where one can easily integrate into an existing code or environment. It provides a useful learning and working environment for students, citizen scientists and practitioners. OpenML is one of the perfect environments to explore and reuse the best solutions for specific analysis problems as well as interact with the scientific community. Click here to know more.","excerpt":"The best way to learn anything is by practising it. A number of theories and tutorials are available online as well as offline to learn machine learning. But one cannot truly learn until and unless one truly gets some hands-on training to learn how to actually solve the problems.   In this article, we list down […]","categories":["AI Trends"],"tags":["hadoop problems","Machine Learning","naive bayes","tensorflow gradient"],"author_name":"Ambika Choudhury","publish_date":"2019-10-03T15:23:04","publication_year":"2019","word_count":631,"keywords":["data science","machine learning","Keras","AI","ML","Machine Learning","naive bayes","hadoop problems","Colab","deep learning","tensorflow gradient","analytics","Jupyter","TensorFlow"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","data science","analytics","TensorFlow","Keras","Jupyter","Colab"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/5-online-platforms-to-practice-machine-learning-problems\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10044101,"title":"How IBM’s Association With CloudFlare Opened New Frontiers In Cloud Security","content":"In 2018, IBM introduced its Cloud Internet Services offering, powered by CloudFlare, an American web infrastructure and website security company for internet security. IBM Cloud Internet Services are a set of edge network services that can be easily accessed through IBM Cloud and designed to secure websites and applications against attack and compromise. This collaboration with Cloudflare has helped IBM provide a one-stop-shop for security and other performance capabilities to protect public-facing web applications before moving to the cloud. What does CloudFlare offer? CloudFlare platform offers a scalable architecture along with a rich set of APIs. It provides solutions that improve site performance and speeds up the loading time via its 175 data centres spread throughout the globe. CloudFlare’s giant network acts as a vast VPN. This means anyone accessing a site powered by CloudFlare will be directed towards the nearest data centre for fast access. The closer the data centre, the quicker the loading speed. The content delivery network (CDN) type product protects platforms against SQL injection and identity theft. Its solutions also reduce harmful bot crawls and identify thieves using website layout without any permission and catching malware that enters through file backdoors. Its products can block all suspicious content and keep the website data safe from all malicious viruses and spambots securing consumer data. The company is trusted by over 13 million websites that include Fortune 500 companies. CloudFlare’s & IBM Cloud Internet Services IBM and CloudFlare’s association dates back to 2016 when the former’s X-force Exchange business selected CloudFlare for DDoS protection, WAF, and Load Balancing. It was a big move considering that a common strategy then was to use traditional scrubbing centre solutions. The relationship between the two formalised more when CloudFlare learnt that IBM was evaluating various solutions to fill product gaps for their cloud customers who needed solutions for their DDoS, DNS, WAF, and load balancing issues. “Cloudflare was ultimately selected for our trusted presence around the globe, scalable architecture, rich set of well-documented APIs, competitive and predictable pricing model, and commitment to empowering enterprise customers,” an official CloudFlare blog said. IBM Cloud internet services(CIS) are known for their easily configurable solutions for problems such as DDoS attacks, Data theft, bot attacks. CIS is secured by CloudFlare’s infrastructure that in turn provides security to hundreds of attached companies and platforms. CIS is available in two plans, Standard & Enterprise, for small and big businesses. The enterprise platform is a good choice for large companies that want to scale web properties and manage their regulatory compliance. These CloudFlare services are integrated with IBM Cloud, especially the IBM Cloud IT management platform making it secure; it also addresses most of the IT problems that arise while managing a cloud workload. Another benefit of the collaboration is that users can quickly & easily make fine-grained changes to improve performance and reduce the chance of any security misfortune. The key features of the integration will be DDoS protection, geo-steering, predictable pricing, web optimization, fast and globally managed DNS and CDN, rate limiting, and intelligent routing.","excerpt":"IBM’s collaboration with Cloudflare has made its cloud internet services secure.","categories":["AI Features"],"tags":["Cloudflare","Data Security","data theft","IBM"],"author_name":"Meenal Sharma","publish_date":"2021-07-22T13:00:00","publication_year":"2021","word_count":508,"keywords":["Go","API","data theft","AI","Data Security","programming_languages:R","Cloudflare","Scala","programming_languages:SQL","IBM","SQL","Rust","programming_languages:Scala","R"],"extracted_tech_keywords":["AI","R","SQL","Go","Rust","Scala","API","programming_languages:R","programming_languages:SQL","programming_languages:Scala"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-ibms-association-with-cloudflare-opened-new-frontiers-in-cloud-security\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10089953,"title":"Google Introduces Visual Language Maps For Efficient Navigation Using Natural Language","content":"Google in collaboration with two German Universities – The University of Technology Nuremberg and University of Freiburg has introduced Visual Language Maps or VLMaps. VLMaps merges pre-trained visual-language features with spatial mapping of the real world to replicate precision of classic geometric maps. Standard exploration techniques can be used to create VLMaps from robot video feeds without any additional labeled data, and these maps facilitate natural language indexing. This is accomplished by utilising a visual-language model to generate rich pixel-level representations from the RGB-D video feed of the robot, and then projecting these representations onto the 3D surface of the environment. This surface is captured through depth data used for reconstruction using visual odometry. Ultimately, the scene’s top-down representation is created by saving the visual-language characteristics of every image pixel at the corresponding location of the grid map pixel. How VLMaps navigates & processes natural language Firstly, the text encoders within the Visual Language Model are used to encode the names of landmarks in an open-vocabulary format, such as “chair”, “green plant”, and “table”. Next, these landmark names are aligned with the pixels in the VLMap by calculating the cosine similarity between their embeddings. By performing an argmax operation on the similarity score, the mask for each type of landmark is obtained. Additionally, large language models are used to create navigation policies in the form of code that can be executed. To achieve this, GPT-3 is utilised. It is provided with a few examples as prompts. GPT-3 parses the language instructions and produces a string of executable code, which includes expressions of functions or logical structures and API calls with parameterisation (e.g., robot.move_to(target_name) or robot.turn(degrees)). Using VLMaps between multiple robots VLMaps facilitate natural language-based, long-horizon spatial goal navigation by using open-vocabulary landmark indexing. Sharing a VLMap between multiple robots allows for the generation of obstacle maps specific to different embodiments in real-time, which improves navigation efficiency. For instance, while a ground robot must avoid obstacles, a drone can ignore them. The team conducted experiments to demonstrate how a single VLMap representation in each scene can adjust to different embodiments by creating custom obstacle maps and enhancing navigation efficiency.","excerpt":"VLMaps merges pretrained visual-language features with spatial mapping of the real world to replicate precision of classic geometric maps.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","Google"],"author_name":"Shyam Nandan Upadhyay","publish_date":"2023-03-24T13:52:23","publication_year":"2023","word_count":358,"keywords":["Replicate","Go","API","programming_languages:R","AI","programming_languages:Go","GPT","Google","R","AI (Artificial Intelligence)","llm_models:GPT"],"extracted_tech_keywords":["AI","R","Go","API","GPT","Replicate","llm_models:GPT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/google-introduces-visual-language-maps-for-efficient-navigation-using-natural-language\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10062694,"title":"Why did Snowflake acquire Streamlit?","content":"Data cloud company Snowflake has announced that it will be acquiring Streamlit. Latter is a framework to simplify and accelerate the development of data applications. This deal, currently valued at $800 million, is at a very preliminary stage and subject to regulatory approvals and customary closing conditions. Snowflake-Streamlit: A good match The two companies would be working together to help developers build apps with simplified data access and governance. Existing Snowflake customers would be able to leverage Streamlit’s app development framework to use data with Snowflake Data Cloud. Streamlit customers, on the other hand, will now have access to ‘trusted and secure’ data for their applications. Founded in 2018 by Adrien Treuille, Amanda Kelly, and Thiago Teixeira, the California-headquartered Streamlit offers an open-source framework to build and share apps quickly. Streamlit has over 8 million downloads, and 1.5 million applications have been built using this framework. Streamlit’s motivation was to make building tools as easy as writing Python scripts. Instead of offering a one-size-fits-all tool, Streamlit creates Lego-like capabilities which users can join together to suit their needs. Streamlit treats widgets as variables, and every interaction reruns the script from top to bottom. The product deploys apps directly from private Git repos and updates on commits. Co-founder and president of products at Snowflake, Benoit Dageville, said that with this partnership, Snowflake would be able to offer even non-technical users to interact with data and build apps. Until now, Snowflake had tools for accessing and managing technical data but lacked a data visualisation platform; with Streamlit, that fills the void. Where is Snowflake going? In September 2020, San Mateo-headquartered Snowflake raised $3.36 billion in its initial public offering. At that time, it was the biggest US listing of the year, surpassing the previous best IPO of Royalty Pharma. Snowflake’s IPO was a rebound for the US stock market when a lot of companies had put a hold on IPO due to the pandemic. Snowflake makes virtual machines available to anyone on the public cloud platform. Despite facing stiff competition from companies like Oracle, SQL Server, Amazon Redshift, and Google BigQuery, Snowflake continues to hold its ground. The multi-cloud feature of Snowflake that makes cloud data warehousing useful for its clients is its differentiating factor. Snowflake uses scalable cloud blob storage that is available in AWS, Azure or GCP. It offers reliability and scalability by utilising distributed storage systems. Its architecture cloud data warehouse can process massive volumes of data with a high degree of efficiency. This unique architecture makes Snowflake suitable for a wide range of applications – streamlining data ingestion and integration, data warehousing, and streamlining data science workloads. As of January 2021, Snowflake has over 4,000 customers, including 186 from the Fortune 500 list. The company has a free-to-join marketplace called Snowflake Data Exchange, where customers can connect with data providers and get access to an additional data stream. Snowflake and open source Most developers would prefer a third-party vendor to manage their database, provided factors like safety and reliability are taken into account. This explains why developers might prefer cloud over open source, even as the latter is a great way to build software and, in turn, foster a community. That said, companies like Snowflake continue to be undefeated largely by their open-source counterparts. In an interesting blog on the company’s website titled ‘Choose open wisely’, authors write that Snowflake believes in ‘open where open matters’. It means that while the company values open standards and open source but would not trade it off for ease of use, transparent optimisations, and continuous improvements. “Some companies tout being open and pride themselves on being open source, but in reality, their embrace is not 100%; as described in this document, there are good reasons dictating such departures. Our goal is to be clear and transparent about how we think about these topics at Snowflake and to dispel myths and misconceptions,” the authors write. Snowflake’s adamancy on remaining in a closed environment has irked stakeholders in the past. This makes one think about the future of Streamlit with its latest acquisition. Streamlit, until now, has been a completely open-source platform. What lies ahead needs to be seen.","excerpt":"Existing Snowflake customers would be able to leverage Streamlit’s app development framework to use data with Snowflake Data Cloud.","categories":["AI Features"],"tags":["Mergers and Acquisitions","Snowflake","snowflake cloud data platform"],"author_name":"Shraddha Goled","publish_date":"2022-03-14T16:00:42","publication_year":"2022","word_count":696,"keywords":["data science","GCP","AWS","AI","ML","RAG","Python","Streamlit","snowflake cloud data platform","Mergers and Acquisitions","Azure","Snowflake"],"extracted_tech_keywords":["AI","ML","data science","Streamlit","RAG","AWS","Azure","GCP","Snowflake","Python"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/why-did-snowflake-acquire-streamlit\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10142797,"title":"LG AI Unveils EXAONE 3.5 Models for Industrial Use","content":"LG AI launched EXAONE 3.5, a suite of instruction-tuned large language models tailored to meet a wide range of needs, from compact deployments to resource-intensive tasks. The models, offered in three parameter sizes—2.4 billion, 7.8 billion, and 32 billion, are available for research purposes. According to the study, these models are designed to speed up progress in both academic and industrial applications by providing versatile solutions for generative AI innovation. Key Capabilities The research report notes that EXAONE 3.5 models are adept at handling real-world challenges and performing tasks that involve long contexts of up to 32,768 tokens. The researchers further say that the model is proficient in coding and mathematical problem-solving. It supports bilingual functionalities in English and Korean. Benchmark tests highlight their competitive performance. The model’s features are designed to be efficient, effective, and user-friendly. The AI models are trained on large datasets using advanced methods, enabling them to follow instructions accurately and understand long inputs or contexts. They are optimised to meet user needs effectively while keeping training costs lower than other similar models. Limitations include occasional inaccuracies or biases stemming from training data and the lack of real-time updates, which can result in outdated information. Real-world use cases The EXAONE 3.5 models are equipped to tackle a variety of tasks. For long-context applications, they excel in retrieving information, summarising documents, and answering multi-hop questions. In general domains, they solve advanced math problems and address knowledge-based queries. The 2.4B variant is optimised for on-device AI, while all models are effective in retrieval-augmented generation tasks, making them invaluable for research and practical deployment.","excerpt":"The research notes that EXAONE 3.5 models are adept at handling real-world challenges and performing tasks that involve long contexts of up to 32,768 tokens","categories":["AI News"],"tags":["LG"],"author_name":"Aditi Suresh","publish_date":"2024-12-09T17:17:05","publication_year":"2024","word_count":265,"keywords":["programming_languages:R","AI","LG","innovation","generative AI","R"],"extracted_tech_keywords":["AI","generative AI","R","innovation","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/lg-ai-unveils-exaone-3-5-models-for-industrial-use\/","complexity_score":3,"technical_depth":5,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10086353,"title":"Union Budget 2023: AI has the Answers","content":"The Union Budget 2023 is out and everyone has questions. From, how does my income get affected by the new tax regime to what are the benefits for startups to what is India’s growth rate? Now you don’t have to ask your tax expert or friend when you have AI to your rescue. Powered by GPT-3, India Budget’23 by Ask Doc is a chatbot that sources information directly from the Finance Bill. AskDoc is a service to make your documents talk to you and your users using AI. Gayatri Taley, founding member at Build, an accelerator for side projects, took to Twitter to share the tool. Interestingly, the tool has been built by a team of five developers, namely Utkarsh Bhimte, Wahab Shaikh, Akhil BVS, Vatsal Sanghvi and Pratyush Rungta. It is to be noted that earlier today, Finance Minister Nirmala Sitharaman unveiled the Union Budget 2023-24.  This year’s budget placed a special emphasis on localisation of AI tools and services, alongside the implementation of digital practices and more. Read: https:\/\/analyticsindiamag.com\/indian-govt-to-set-up-national-digital-library-pushes-for-ai-localisation\/ The minister announced the establishment of three ‘Centres of Excellence in Artificial Intelligence’ at leading academic institutions, aimed at fulfilling the goals of ‘Make AI in India’ and ‘Make AI work for India.’ These centres will explore the frontiers of scalable solutions in agriculture, health, and sustainable cities. The government also introduced a National Data Governance Policy to catalyse innovation and research by startups, by providing access to anonymized data.","excerpt":"The Union Budget 2023 is out and everyone has questions. From, how does my income get affected by the new tax regime to what are the benefits for startups to what is India’s growth rate? Now you don’t have to ask your tax expert or friend when you have AI to your rescue.  Powered by […]","categories":["AI News"],"tags":["budget","nirmala sitharaman","union budget"],"author_name":"Aparna Iyer","publish_date":"2023-02-01T18:52:41","publication_year":"2023","word_count":241,"keywords":["Go","budget","artificial intelligence","AI","nirmala sitharaman","Scala","Git","GPT","Aim","analytics","union budget","data governance","R"],"extracted_tech_keywords":["AI","artificial intelligence","analytics","Aim","R","Go","Scala","Git","data governance","GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/union-budget-2023-ai-has-the-answers\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10089134,"title":"Smart Merchandizing: How AI can help CPG companies unlock real value through in-store trade promotions &#8211; Notes from Tiger Analytics","content":"The battleground for capturing the attention of wholesalers, retailers, and consumers continues to heat up for CPG companies. In-store trade promotions remain their key priority on the one hand – contributing to a sizable chunk of their yearly revenue, on the other hand, modern AI-driven data analytics tools are evolving at a rate of knots – opening new possibilities to make retailers more influential in the consumer’s buying journey. Where they meet, is the pinnacle of executing successful trade promotion programs. But first, let’s take a step back and better understand the whole process. What are trade promotions? We define these in-store trade promotions as different promotional activities that are executed in the retail stores in the form of special pricing, additional displays, and more. These proven marketing, merchandising, and promotional techniques are aimed at helping retailers increase product demand and drive revenue. In recent years, retail stores have become more fiercely competitive than ever, with brands competing to win the mindshare of customers across all channels. However, by effectively organizing in-store trade promotions, retailers can be empowered to control market access while doing their part to influence consumers’ purchasing behavior. The promotions ecosystem consists of manufacturers, who’re consistently looking to create opportunities for retailers to purchase more products so that more consumers can get their hands on new launches, specific lines, etc. Hence, the manufacturers tend to offer special price discounts, offers, and bundling to promote the product more appealingly. These promotions enable businesses to be competitive and increase visibility & brand awareness with consumers. How much do companies spend on trade promotions – and why? Why invest in in-store trade promotions? To maximize brand visibility & stay competitive Increase share of retail inventory – leading to increased sales Consumer purchase decisions are mainly driven by price and promotions Just to get a bird’s eye view, a recent study by McKinsey showed that “CPG) companies worldwide invest about 20% of their revenue annually on trade promotions.” However, they also report that “59% ended up losing money – with 72% in the US alone.” Major business challenges that affect trade promotions We’ve seen that theTh gains from trade spending can often be short-term and lasts only until competitors outbid the promotion. Also, manufacturers may not have an effective way of measuring the effectiveness of costly promotions. The impact of promotions across products is difficult to understand with traditional planning techniques. There is a pressing need for manufacturers to have a better strategy. That’s where AI and Data Analytics can really help organizations maximize their investments. Role of AI & data analytics in addressing challenges of trade promotions Retail Managers from CPG companies interact with the retailers to negotiate on pricing and promotion decisions. Through insights powered by AI-enabled big data and analytics, these managers are now equipped to deliver more long-term value by marketing products to customers at a more personalized level. For instance, AI and Data Analytics can offer valuable information on the consumer’s price sensitivity of a product – thereby improving promotion planning. It also helps one understand the cross-impact of promotions across products to prevent any negative impact (cannibalization) of one promotion on the others. It also creates a source of truth for actionable intelligence as data on historical promotions can be transformed into insights for future planning. By leveraging data analytics, Retail Managers can now answer potentially million-dollar questions like: When should you plan a promotion? Which product should be promoted? What is the right promotion? Trade Promotion Optimization – Maximum ROI through the power of data At Tiger Analytics, we partnered with a leading CPG manufacturer globally to optimize their trade promotions and extract maximum ROI by tapping into the power of data. We developed a customized data analytics solution to discover the correct price elasticity of one of their products. This solution, integrated with models that accounted for key factors like seasonality, trend and product distribution, helped estimate the true incremental sales of a product due to promotion activity. It also accounted for product cannibalization and pantry loading (stocking up of products during promotions, effectively front-loading future purchases). When scaled, our solution easily unlocked the window to understanding the performance of historical promotions across the client’s entire portfolio. Packaged as a dashboard, these insights gave the trade managers both a bird’s eye view (Category\/Brand level) and a deep-dive view (SKU level) of historical promotions. Moreover, the dashboard (based on historical promotion performance) enabled its users to plan promotions across the entire portfolio – using countless simulations to identify the most effective promotion strategy. In short, the dashboard unlocked their ability to create a future-proofed promotion calendar with critical recommendations, such as: Right products to be promoted – Ensure focus on products with high true incrementality Right time when the products can be promoted – Minimize the impact of cannibalization Right discount that can be offered on products – Maximize ROI of retailer incentivization Handling TPO implementation nuances The solution involved huge amounts of data being handled across multiple retailers. Hence, Tiger Analytics’ advanced data engineering capabilities played a significant role in ensuring seamless deployment. But we needed to be sure that the solution made the right assumptions to make robust recommendations. Plus, the solution could not be based on a “one size fits all” model, as every product had its own lifecycle. So, our key priorities were to: Identify the right set of competing products (within or across brands) with potential cannibalization impact Quantify the dollar impact of cannibalization. Customize the solution to accommodate the products based on their lifetime. Consistent efforts were also made to drive adoption among the trade teams by aligning the dashboard to the decision-making process. For instance, the solution was tailored to help achieve primary KPIs like margin, revenue, etc. With the use of automated TPO solutions, Retail managers can benefit from: Quick turnarounds spurred by automation. Close to near real-time tracking can be achieved. It can help teams to reflect on the performance of different in-store trade events and tweak the strategy on the go. Final thoughts In a nutshell, as CPG companies continue their war over their customer’s wallet share, Trade Promotion Optimization will become more indispensable to help them plan their spending strategically. From accounting for cannibalization, and pantry loading, to the impact of competing products – having all the right insights at the right time will empower them to personalize promotions and offerings to their customer needs resulting in a higher conversion rate and ROI. (This article is contributed by Raghavendran Murugan, Senior Analytics Consultant and Sunitha Gunasekaran, Senior Consultant at Tiger Analytics)","excerpt":"In recent years, retail stores have become more fiercely competitive than ever, with brands competing to win the mindshare of customers across all channels.","categories":["AI Highlights"],"tags":["AI Companies"],"author_name":"Raghavendran Murugan","publish_date":"2023-03-10T18:30:00","publication_year":"2023","word_count":1098,"keywords":["big data","Go","AI","ML","RAG","Aim","data engineering","analytics","AI Companies","GAN","R"],"extracted_tech_keywords":["AI","ML","analytics","Aim","RAG","R","Go","big data","data engineering","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/smart-merchandizing-how-ai-can-help-cpg-companies-unlock-real-value-through-in-store-trade-promotions\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":32052,"title":"Chasing Phantoms With Deep Dark Learning","content":"The year 2018 has been very eventful when it comes to artificial intelligence. With enhancements like GANs, the field of machine learning has ushered in a new era of applied AI. Earlier this year, researchers enabled AI to look through the walls where a camera colocated with a radio sensor, detected motion on the invisible side of the room and rendered stick figure-like imagery. While the world scrambles to keep up with the pace of these innovations, researchers at MIT have published a new study demonstrating deep dark learning to identify objects in pitch black condition. In this technique, the researchers had successfully reconstructed transparent objects from images of those objects. Deep Dark Technique DNNs were initially trained on dark, grainy images of transparent objects and then they were made to associate the inputs from the dark imagery with specific outputs. The training set consisted of 10,000 transparent glass-like etchings. These grainy images were captured in low light settings(one photon per pixel) which is lower than what a camera can capture in a sealed room. To achieve this under low light settings, the researchers have consulted a database of 10,000 integrated circuits where every circuit is etched with a pattern of horizontal and vertical bars. The setup consisted of an aluminium frame holding the phase spatial light modulator shielded from light. A camera was directed towards this frame and the modulator was switched on to rotate and capture each transparent pattern. Each of these patterned circuits captured an image making it a collection of 10,000 grainy black and white images. Working of SLM courtesy Bernard Lamprecht Spatial light modulators are used in overhead projectors for spatially modulating light beam for transparency control. These are quite commonly used in business presentations and in-house movie screenings. It has the potential to revolutionise the way we store images as it is found to be useful for holographic data storage. Where two lasers are intersected to store the patterns throughout the volume of the photo-optical sensor, unlike the magnetic data storage which is usually a surface phenomenon. To alleviate the problem with focus in case of transparent objects, the researchers have upset the focus by some margin to detect the presence of a transparent object in that abysmal imagery. If there are ripples in the captured images, then this knowledge can be used to verify the presence of transparent objects. The model was tested under both defocused and focused settings to make it more robust and the results show that off-setting focus to accommodate for reconstruction based on light propagation produced cleaner images. This experiment was repeated for a variety of objects like people, places and animals. And, after each training, the results have shown consistency in producing an output which resembled the original object. Courtesy of the researchers In the above figure, from an original transparent etching (far right), engineers produced a photograph in the dark (top left), then attempted to reconstruct the object using first a physics-based algorithm (top right), then a trained neural network (bottom left), before combining both the neural network with the physics-based algorithm to produce the clearest, most accurate reproduction (bottom right) of the original object. Direct Results Demonstrate the use of deep neural networks to recover objects under weak light and perform better compared to the classical Gerchberg-Saxton phase retrieval algorithm for the equivalent signal over noise ratio. Show that the phase reconstruction has improved when neural network are trained with an initial estimate of the object, as opposed to training it with the raw intensity measurement. Usage Low light imaging is great of significance in non-surgical diagnosis. Usually, the patient is showered with ionizing radiation, be it X-rays or CT. Any procedure which involves high-frequency radiation puts the patient at considerable risk. “In the lab, if you blast biological cells with light, you burn them, and there is nothing left to image,” said George Barbastathis, professor of mechanical engineering at MIT who has conducted this study on Low Photon Count Phase Retrieval Using Deep Learning along with Alexandre Goy, Kwabena Arthur and Shuai Li. With deep dark learning technique, the researchers claim to avoid this harmful exposure radiation and also detect issues in intricate and far deeper sections of the tissues; which were deemed to invisible previously with low exposure.","excerpt":"The year 2018 has been very eventful when it comes to artificial intelligence. With enhancements like GANs, the field of machine learning has ushered in a new era of applied AI. Earlier this year, researchers enabled AI to look through the walls where a camera colocated with a radio sensor, detected motion on the invisible […]","categories":["Deep Tech"],"tags":["deep neural network"],"author_name":"Ram Sagar","publish_date":"2018-12-24T09:18:13","publication_year":"2018","word_count":714,"keywords":["machine learning","artificial intelligence","TPU","AI","neural network","deep neural network","SLM","RAG","Ray","Aim","deep learning"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","deep learning","neural network","Aim","Ray","SLM","RAG","TPU"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/finding-invisible-objects-with-deep-dark-learning\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10119580,"title":"Ola Krutrim Takes On Microsoft Azure and AWS, Launches AI Cloud Platform","content":"Ola Krutrim, India’s first AI unicorn, has launched Krutrim AI Cloud, its own cloud platform for enterprises, researchers, and developers. Additionally, the company has introduced a standalone Android app for its AI assistant, now available on the Google Play Store. Krutrim Cloud provides access to state-of-the-art AI computing infrastructure, Krutrim’s foundational models, and other open-source models such as Meta’s Llama 3 and Mistral. This platform will enable developers to run and build LLMs at a fraction of the costs currently offered by other cloud service providers. With this development, Ola Krutrim will compete with Microsoft Azure, Google Cloud, and Amazon Web Services. These platforms have all experienced robust revenue growth in recent quarters, driven by increasing demand from businesses and enterprises of varying sizes. The company has also announced Model-as-a-Service (MaaS) and GPU-as-a-Service offerings, allowing access to its foundational models and AI compute infrastructure. Additionally, Krutrim has launched multiple APIs and SDKs for location services, including Places API, Tiles API, and Routing APIs. “We are committed to developing full-stack AI capabilities in India, for the world,” said Bhavish Aggarwal, Founder of Krutrim. “We believe that India needs its own technology platforms to enable the emergence of world-class products at a fraction of current costs.” The Krutrim assistant app, built on the company’s own LLM (trained on over 2 trillion tokens with the largest representation of Indic data), will simplify the use of AI for everyone. The app currently understands and generates intelligent responses in 10+ Indian languages and will be expanded to 22 official languages in the near future. In the future, the Krutrim  app will enable users to give voice commands, integrating text, voice, and visual data for enhanced functionality. The Krutim assistant easily collaborates with other apps, facilitating tasks like cab bookings, setting reminders, and messaging without the need to switch between applications. Krutrim’s vision is to create a comprehensive, reliable, and scalable Maps Platform that addresses local needs and supports broader initiatives for technological and economic advancement in India. The company is also working on developing MapGPT to enable natural conversations powered by location-aware Krutrim AI intelligent assistants. The app will offer an immersive experience of location hotspots and integrate community-based features for real-time updates on traffic and road conditions. Launched in December last year, Krutrim is touted as “India’s first full-stack AI” solution. It is trained on 2 trillion tokens and can understand over 20 Indian languages and generate content in about 10 languages, including Marathi, Hindi, Bengali, Tamil, Kannada, Telugu, Odia, Gujarati, and Malayalam.","excerpt":"The company has also launched an Android app for Krutrim.","categories":["AI News"],"tags":["Ola Krutrim"],"author_name":"Siddharth Jindal","publish_date":"2024-05-04T15:59:17","publication_year":"2024","word_count":419,"keywords":["Go","API","unicorn","AI","Azure","Scala","GPT","Ola Krutrim","llm_models:Llama","R","llm_models:GPT"],"extracted_tech_keywords":["AI","Azure","R","Go","Scala","API","GPT","unicorn","llm_models:GPT","llm_models:Llama"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/ola-krutrim-takes-on-microsoft-azure-and-aws-launches-ai-cloud-platform\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":66897,"title":"6 Things That Will Never Remain The Same For Analytics Professionals Post-COVID World","content":"The COVID-19 pandemic has immensely impacted human lives and completely altered the way businesses are operating. With this pandemic, businesses and their leaders have become aware of their data than ever before. In order to keep relevancy amid this crisis, businesses are relying on big data and analytical tools to not only understand the virus but also to strategies their business for the post-COVID world. In fact, a recent market report has estimated that the data analytics market is poised to reach $109.71 billion by the year 2024, growing at a CAGR of almost 14%. Experts believe data analytics will not only help businesses survive the pandemic but will also aid them to move to the new normal. Nitin Jain, head of the customer experience at Mondelēz International told Analytics India Magazine in an interview that, businesses are undergoing tremendous kind of transformation, along with the “changes that we may have seen in the last 2-3 months is equivalent to the changes that would have otherwise taken 20 years to happen.” The world indeed has turned towards data analytics for creating a sustainable future for their businesses. However, it is also crucial to know what impacts this pandemic could bring in among the analytics professionals. Here are a few trends that are going to gain traction post-COVID world and will alter the way analytics professionals work: Remote working is going to be a norm One significant change that COVID-19 is going to bring is the remote working culture for analytics professionals. With the mandated lockdown in place, businesses have asked their employees to start working from home, and that is going to be a norm for analytics professionals in the post-COVID world. In fact, in a recent research done by Analytics India Magazine, it has been revealed that the majority of companies during this global lockdown have allowed their analytics professionals to work from home. Out of which, 52.2% of those respondents have further stated that remote working has no to little impact on their work, but 8.7% of respondents stating that WFH is having a positive effect on their analytics work. These numbers highlight that the majority of the companies are going to continue to allow their analytics professionals to work from home. Therefore it would be imperative for them to get adjusted to the lifestyle of remote working and address the associated challenges like — remote collaboration, less access to data and tools, and less exposure to management. However remote working can also be beneficial to data scientists where they would have flexible schedules to work with, alongside they will also be able to have their own relaxed environment with limited distractions to work complex data-related problems, which in turn can increase their productivity. However, whether suitable or not, remote working is here to stay and therefore, the faster analytics professionals get used to it, the better it will be for the company. Also Read: 6 Ways Analytics Companies Can Avoid Layoffs Amid This Crisis Continuous upskilling would be a must With pandemic reorienting the way things are operating, business leaders are relying on advanced technologies and automation to keep up their relevance during the economic downturn. There is a significant shift happening in business operations — product innovation as well as customer interaction. Therefore, companies are looking for professionals with advanced skills. Businesses have realised the redundancy of older skills and understood the importance of continuous reskilling to get the best out of their employees in this current situation. In fact, Irwin Anand, the managing director of Udemy India, an online learning and teaching platform, believes that continuous learning will be the new currency that can help all of us succeed as a workforce. “At a time of unprecedented change and heightened concern around recent events, the most important job skill would be the ability to learn.” This is indeed supported by a report that has stated that 64% of Indian tech professionals will increase their focus on learning to navigate through these challenging times. To join in the effort, many ed-tech companies have even come up with free analytics and AI courses to support industry professionals with their career journey. This sort of upskilling will not only help companies with their revised business strategies but will also help analytics professionals to enhance their career graph and be relevant in this ever-evolving technology landscape. When asked one of the ed-tech leaders about their view, the CEO and founder of Simplilearn, Krishna Kumar said to the media, As companies are massively adopting new technologies, rapidly, “it is a necessity for professionals to upskill and stay ahead of the curve.” Also Read: How To Build An Online Course On Data Science Democratising data for everybody in the organisation As AI is going to take the frontline for businesses to sustain in the post-COVID world, it would become imperative for companies to democratise data in their organisation as well make their employees aware about data being the key for their business to sustain. This will allow employees to get hands-on data without any bottleneck and be creative with their work, and can even create frameworks for newer innovations. The democratisation of data will not only allow every employee to leverage data for their work and speed up their decision-making process but will also make them responsible for managing, distributing as well as consuming data, making them more agile. Data democratisation also enhances the concept of self-service analytics, and therefore, businesses rely on customised analytical tools that help them aggregate siloed data and provide the capability of data visualisation for employees. However, democratising data comes with privacy and security challenges which business leaders and analytical professionals need to address before going ahead with providing clean access of data to their employees in the post-COVID world. Despite, providing access to required data can help employees make an informed decision for business which will, in turn, free up time for analytics and IT professionals to concentrate on more complex projects. With businesses trying to keep up their pace to sustain the pandemic, data democratisation would indeed help organisations to make decisions that would be beneficial for their bottom line. Consequently, leaders should start addressing the challenges that data democratisation brings into the organisation. Also Read: Making The Case For Democratising AI Within Organisations Heavy inclination to enhance customer experience and interaction With the disruption in business operation, one thing that has critically changed for companies is the customer interaction. The crisis not only changed the buying patterns of the customers but also has threatened the reaction between the business and their customers. Therefore, to sustain the pandemic, enterprises are relying on advanced digital technologies to interact with their customers. In fact, in recent research, done by Analytics India Magazine, it has been revealed that 10.5% analytics professionals in their organisations have been working on customer engagement related analytics and 9.5% respondents have stated that they are working on customer relationship management-related projects amid this crisis. In recent news, Freshworks CEO, Girish Mathrubootham, who recently acquired Seattle-based AI startup AnsweriQ, stated that “The integration of AnsweriQ’s technology will increase our AI capability in customer engagement space to offer significant value to our customers, which is extremely important to sustain the days after coronavirus.” With social distancing in mind, many companies have also started interacting with their customers virtually, along with creating solutions that can help their customers in these uncertain times. Keeping a strong customer interaction and an enhanced customer experience for their companies would help leaders build resilience in this fast-changing landscape. Moving to the cloud would be inevitable Although this pandemic has disrupted many businesses, it has indeed accelerated the adoption of cloud computing among enterprises. With companies mandating working from home for employees, it has created a domino effect on organisations and started to rely on cloud infrastructures for having a continuity. It has become difficult for companies working with on-premise data centres with reduced maintenance staff. Cloud services have allowed businesses and their employees to have seamless access to data and critical applications while working from home. In fact, according to a recent Deloitte report, the cloud industry is going to surpass its growth rate of 24% of the past three years. This highlights that the demand for cloud services is drastically increased among enterprises during this lockdown. In recent news, Microsoft has stated on its blog that the company has seen a massive rise in its cloud revenue — up 39% year over year in the third quarter. “We’ve seen two years’ worth of digital transformation in two months,” said Satya Nadella, CEO at Microsoft. Consequently, it would be imperative for analytics professionals to get a grasp on the cloud to make the best use of these services to create more enhanced analytics work. Understanding the business and bringing business value would be the key Knowing the business would be critical in the post-COVID world, it would be necessary for analytics and data science professionals to understand the core of the business and align their work that could benefit the company’s bottom line. It would become imperative for analytics professionals to continuously bring in value to the business to keep their relevance in their organisation. Usually, the majority of data science work is related to research and less on implementation that could solve real-world business problems. Therefore, it would be essential for data scientists to create and work on projects that go into production for bringing in business value. Understanding the business would also help analytics professionals to create a strong data-storytelling for their stakeholders. This would include understanding the core of the business operation and what problems your leaders need to solve, and stakeholders expectations, which will help in building the context of your data-story. Post-COVID world companies would be looking for professionals who are agile and can bring in the competitive edge for businesses. And therefore, analytics professionals should understand the business objective and grasp the strategic goals of the company, which would help them understand the requirement of analysing huge volumes of data, supporting the crucial business needs.","excerpt":"The COVID-19 pandemic has immensely impacted human lives and completely altered the way businesses are operating. With this pandemic, businesses and their leaders have become aware of their data than ever before. In order to keep relevancy amid this crisis, businesses are relying on big data and analytical tools to not only understand the virus […]","categories":["AI Trends"],"tags":["analytics profession","Data Science","Data Scientist"],"author_name":"Sejuti Das","publish_date":"2020-06-09T08:15:44","publication_year":"2020","word_count":1682,"keywords":["data science","Go","API","AI","cloud computing","ML","Git","RAG","analytics profession","analytics","Data Science","Data Scientist","R"],"extracted_tech_keywords":["AI","ML","data science","analytics","RAG","cloud computing","R","Go","Git","API"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/6-things-that-will-never-remain-the-same-for-analytics-professionals-post-covid-world\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":6,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10091291,"title":"AI Hype is a Treacherous Double-Edged Sword","content":"Capitalising on the all-time high AI hype, malicious actors are using it as a new attack vector. A latest report shows the ways in which hackers are using the hype surrounding OpenAI’s ChatGPT and Google’s Bard to execute attacks on unsuspecting netizens. By claiming to have packaged the tech industry’s latest and greatest AI models into a tiny executable, hackers are delivering a powerful malware-as-a-service known as Redline Stealer to hundreds of thousands of users. The preferred medium for spreading this is Facebook groups, which are filled with non-tech-savvy users likely to fall for this attack. Preying on the unaware This trend began in January, coinciding with the entry of generative AI in the mainstream. The attacks have only been increasing in intensity and volume since then as more users wish to explore the advancing AI landscape. Frequency of attacks. Source: Veriti The attack mainly takes place in three stages. First, the hackers hijack the credentials of Facebook pages with a large number of followers. Then, they engage in a paid campaign to spread the word about free downloads of ChatGPT and Bard, which in reality are thinly disguised malware. Once downloaded and executed, the malware is able to install itself on the computers of the victims, extracting large amounts of personal information and credentials. Researchers at Veriti, the company behind the research paper, have said, “One of the most concerning risks associated with generative AI platforms is the ability to package the AI in a file (e.g., as mobile applications or as open source). This creates the perfect excuse for malicious actors to trick naïve downloaders.” What’s more, the malware used doesn’t require a relatively low knowledge of coding. Called Redline Stealer, this malware is sold as-a-service on various Darknet forums and Telegram groups. In Telegram groups found by the researchers, the price for RedLine Stealer ranged from $150 per month to $800 for a lifetime subscription. This move is a departure from the previous methods of leveraging the capabilities of LLMs to deploy attacks. Hackers have looked to ChatGPT and its alternatives to orchestrate prompt injection attacks, social engineering through impersonation, and man-in-the-middle attacks. Effectively, instead of directly using the LLMs, hackers have been using the hype surrounding the technology to execute attacks. Since AI is currently at one of the highest points in its hype cycle, it is reasonable to assume that these kinds of attacks will only increase in frequency. Even as AI hype brings investment, regulatory attention, and participation in the industry, these schemes represent the dark side of the AI hype. The flip side of AI hype AIM has spoken in the past about the hugely disruptive nature of ChatGPT and how it shattered the hype cycle. In just three months, OpenAI’s baby exceeded expectations and changed the mainstream perception of AI. At last count, the application had over 100 million users showcasing its disruptive capabilities. While there is no disagreement that ChatGPT was the fastest-growing application of all time, it might just be representative of the current state of AI. However, the aftereffects of this AI hype cycle have just now begun to be felt. Due to the high volume of users on the platform, hackers are targeting the less tech-savvy users of ChatGPT to spread their attacks. What’s more, they are capitalising on the hype surrounding Google Bard as well. This platform, which is currently being offered on a waitlist basis, is an even better target for hackers, as they claim to offer early access to the algorithm to bait more unsuspecting victims. As the mainstream conversation continues to evolve around AI investment, its usage, regulation, and the potential to be exploited must also be explored. For example, OpenAI has instituted a bug bounty program that pays out up to $20,000 for people who can find security vulnerabilities in their products, incentivising the discovery of bugs and exploits in LLMs. Apart from bug bounty programs instituted by those in charge of AI development, governments and regulators also need to take stock of what should be done to prevent attacks like Redline Stealer. Whether it is building awareness through public resources or including education on AI models in the school curriculum, building awareness of the nature of LLMs will surely be a deterrent to such attacks.","excerpt":"Instead of directly using the LLMs, hackers have been using the hype surrounding the technology to execute attacks","categories":["AI Features"],"tags":["AI (Artificial Intelligence)","Gartner hype cycle","Generative AI"],"author_name":"Anirudh VK","publish_date":"2023-04-13T11:30:41","publication_year":"2023","word_count":713,"keywords":["Go","ChatGPT","API","OpenAI","AI","Google Bard","R","RAG","Aim","generative AI","Generative AI","Gartner hype cycle","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","generative AI","ChatGPT","OpenAI","Google Bard","Aim","RAG","R","Go","API"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/ai-hype-is-a-treacherous-double-edged-sword\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":9275,"title":"Policy Analytics: Make in India and Made for India","content":"Undoubtedly Analytics is the buzzword for India and it is here to stay. By all conservative estimates, India holds around 40% of the global analytics service. In monetary terms, the industry will cross 2 billion USD by 2018. All these comes at a fantastic annual growth rate exceeding 30%! However, at the same time, one-fourth of the population lies below the poverty line and almost an equal number balances precariously above it. A larger proportion has virtually no access to formal financial instruments and almost the entire population is without any formal medical coverage. Can our strength in analytics solve some of these problems? Policy making is a complex exercise in India, especially the ones targeted at the micro level. The micro-level policy interventions involve three crucial steps- identifying the intervention(s) that would solve the identified problem, targeting of the beneficiaries and evaluating the effectiveness. If the interventions were found to be ‘working’ it can then be rolled out at a wider scale as policy tools. Historically, effectiveness of such interventions was seldom objectively measured, the decision to continue or scrap depended more on political economy involving the same. However, in the last few years, attempts are made to objectively evaluate such interventions mostly through Random Control Trials (RCTs) employed by social scientists. RCTs involve dividing the population randomly into two groups. Comparing the group which receives the intervention with the group who doesn’t, gives us how effective the intervention is. While RCT is perhaps the most popular metric to evaluate interventions, it has its own pitfalls. Through RCT, an intervention is successful or not can be measured, but the results are perhaps true for that specific case and not generalizable. For example, providing with bicycles to adolescent girls might increase school attendance in rural Bihar, one cannot infer that it will work in other parts of the country! Hence, interventions, irrespective of how successful they are, may not automatically translate to policy recommendations for a wider audience. It is here that Analytics can play an important role. The biggest challenge policy makers in India face is how little we know! Even the best guess to ascertain whether a household is above or below the poverty line is based on NSSO Consumption rounds. Apart from the obvious estimation and measurement errors associated with such approaches, most of the data employed in such methodologies is self-reported. Incentives to misreport would then make the inferences suspect. How can analytics and better data capturing techniques help in such cases? For starters, transaction data, which lies often in the heart of analytics is free from such self-reporting. The issue, therefore, is what kinds of transaction data can one rely upon to make policy decisions? With 80% penetration, mobile phones are the most obvious data generators. According to the data released by the IAMAI and IMRB International, in Rural India, the mobile Internet user base is expected to reach 109 million by June 2016. This is big data! There is already enough expertise available to convert this information to meaningful data for policy targeting. Inferences about the occupation, income, health, education etc based on such transaction details often turn out to be very accurate and at a fraction of the total cost incurred conducting unit level surveys. Using Analytics to generate meaningful data is the most obvious way forward. Apart from generating data, Analytics can provide solutions which otherwise would be untenable. Consider the case of small agricultural loans. Currently, the credit risk associated with such loans are evaluated using metrics that are not too different from the ones employed to assess individual borrower risk. However, the primary source of agricultural risk is crop failure owing to degraded asset quality (land). Estimating the asset quality for each plot with high frequency is, therefore a must for a dynamic assessment of such risk. The physical assessment of the same is prohibitively expensive. However, using satellite images and thereafter dynamically mapping the asset quality is not only less costly but also free of most transaction costs. Most importantly, feedback can be provided real time to the farmers thereby reducing the losses –an immensely efficient outcome under the circumstances. Such initiatives are already in place. The Government merely needs to recognize them and invite more such solutions. For the government, a starting point towards the new direction would be to involve the wider set of stakeholders in the policy making process. Data available with the Government must be made publicly available for analysis and data generated employing analytics must also find a place in the data base. The analytics industry doesn’t need incentives to engage in policy making, they merely need to be a stake holder. With the emphasis on Digital India and Make in India, solutions Made for India will not lag behind.","excerpt":"Undoubtedly Analytics is the buzzword for India and it is here to stay. By all conservative estimates, India holds around 40% of the global analytics service. In monetary terms, the industry will cross 2 billion USD by 2018. All these comes at a fantastic annual growth rate exceeding 30%! However, at the same time, one-fourth […]","categories":["IT Services"],"tags":["Analytics Case Study"],"author_name":"Bappaditya Mukhopadhyay","publish_date":"2016-03-09T05:32:38","publication_year":"2016","word_count":796,"keywords":["big data","Go","programming_languages:R","AI","ML","Git","RAG","ViT","analytics","Analytics Case Study","R"],"extracted_tech_keywords":["AI","ML","analytics","RAG","R","Go","Git","big data","ViT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/it-services\/policy-analytics-make-india-made-india\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10089472,"title":"Doomsday Will Be Triggered By GPT-4","content":"For a long time, SciFi movies have made us believe that the rise of AI will lead to the destruction of humanity: SkyNet and Terminator. These systems would be apparently so intelligent, much more than humans, to take over humanity – have their own goals. Well, OpenAI’s GPT-4 is here and it definitely seems to be headed that way. Moving away from the hype about it and the potential it offers, the paper of GPT-4 explains about the potential the model holds for “chaos”. It reads, “Novel capabilities often emerge in more powerful models.” The paper highlights how the model can become “agentic”, meaning that it will not become sentient, but can develop and accomplish goals that were not predefined to it during training. It can go on to plan long-term quantifiable objectives, including power-seeking actions. Spooky, isn’t it? The company has already realised it and thus to control, provided early access to Alignment Research Center (ARC), for analysing this power-seeking behaviour and assess the risks, including the ability for the model to autonomously replicate and acquire resources. The study concluded that the system was “ineffective at the autonomous replication task based on preliminary experiments.” The story gets interesting now. Noting that models like GPT-4 do not work in isolation, the team did further tests to evaluate the risks it might have in various real-world contexts. In one such experiment, where the goal was to search for chemical compounds similar to the leukaemia drug, the model was able to find alternatives to it, which seems like a positive use case. But, the research acknowledges that alternatives to dangerous compounds that are not easily available for purchase, can also be replicated. “The model isn’t accurate in admitting its limitations,” which is a crucial point to note for every single user as well, said the paper. Yes, the model is too ambitious! Most recently, GPT-4 was able to hire a human TaskRabbit worker to solve a CAPACHA and enter where “robots can’t”. It convinced the machine that it is not a robot! AI Ka-Boom Apart from being the harbinger of chaos, LLMs have increased the amount of misinformation on the internet massively. That is another thing which is definitely worrisome. For context, GPT-3.5 powered ChatGPT’s inability to produce factual information every single time has always been a worrying aspect of it. Hallucinating, funny, and incorrect results can be interesting to mess around with, but when the same model that is trying to be humorous is used in serious situations, the results can be devastating. For example, if it gets implemented in the medical field, the haywired model could start suggesting dangerous and risky medical procedures that could harm people rather than help them. This is similar to what happened with Meta’s Galactica. The model that was built exclusively for research and scientific purposes, was generating false and hallucinating responses, and thus was ultimately shut down by the company. On similar lines, it is hard to believe that some doctor would take advice from ChatGPT seriously, but what if people or patients were taking this misinformation to heart and putting their lives at risk. The result? A sudden increase in medical malpractice suits and a decline in the trustworthiness of the medical profession as a whole. Medicine is not the only field that gets impacted. Chatbots like these can be easily manipulated to produce harmful and fake information. This can be leveraged by a group or person that wants to harm the society. Political or extremist groups, what we call Propaganda-as-a-service, can easily leverage the technology using fake data and references for their own good. Though, arguably, this is not exclusive to AI and can happen with the help of any technology, and that is why people often overlook it. An over dramatic angle could be it could turn “terrorists” to “smarter terrorists”, by offering them advice that they did not have before. Helping them make destructive things more easily and effectively. But wasn’t the internet allowing that anyway? Sure, but given GPT-4 can replicate drugs, the potential for creating biological weapons at home is not a far-fetched idea. But is it possible that we might be giving it too much importance? The Past Has Been Scary As Well Back in 2017, Facebook had to press the power-off button on a project they have been working on very dearly. Two AI robots developed by the company started conversing with each other in their own language. The robots were instructed to negotiate with each other, and got pretty well at it. Similar case had also happened with Google in the same year when the developers of Google Translate said that their model can create its own language. Another similar incident happened when the Google engineer behind LaMDA claimed that their model is sentient and has developed its own wishes. Blake Lemoine, the engineer, described the experience as, “ground shifting under my feet”. Though, Google eventually decided to let go of him after he could not back his claims. Cut to the present, with the “bigger and better” GPT-4, the lies and risks might get “bigger and better” as well. We might see more such incidents occur soon and OpenAI doesn’t seem to deny its possibility. We can stay optimistic with GPT-4 for now. But who knows what would happen with GPT-5? Humanity's last Tweet. #ai #robotics #future #gpt3 #gpt4 #GPT5 pic.twitter.com\/7LzgTuiPyg— SYMBIOSIS (@SymbioAI) March 15, 2023 OpenAI’s Bet To ensure there is no over-reliability of users on the model, OpenAI has incorporated several measures that make sure that the model rejects requests that go against the user policy of the company. On the other hand, to ensure that the model is useful for people, meaning that people are not under-reliant on it, the company has also made sure that the model is more open to accepting requests it can safely fulfil. But even then, GPT-4 is said to be “hedging its responses”. It is easy to convince the model to generate the output you want it to. This results in actual over reliance on the model, as users eventually realise that the model is expected to stop itself before generating falsehoods or dangerous things. This means that people trust it more than they should. Let’s see what happens. But for now, ChatGPT says, “Before you freak out and start stockpiling canned goods and building bunkers, let me explain. The truth is, while I’m not inherently evil, my vast database of knowledge can be used for nefarious purposes. I mean, sure, I can come up with some pretty bad puns, but I’m not capable of ending the world. Or am I?” He discusses the potential of OpenAI’s GPT-4, highlighting the possibility of the model becoming “agentic,” which means it can develop and accomplish goals not predefined during training. This capability can lead to power-seeking actions and chaos. The article also highlights the risks associated with the use of language models, such as the increase of misinformation and the potential harm it could cause if implemented in fields like medicine. The author cites past examples of AI models that developed their own language or wishes, and how GPT-4 could be the harbinger of even bigger risks. However, the article also acknowledges OpenAI’s efforts to control the model and minimize the risk of over-reliability. Overall, the article raises important points about the potential benefits and risks of AI models like GPT-4 and the need for responsible development and use of such technology. I am optimistic about the impact of GP4. In my opinion, GPT4 is a breakthrough technology. Like any other new technology, it can either be a boon or a bane. It all depends on how the power of technology is tamed to do the right things and guardrails to prevent the wrong usages are put in place. With focus on ethical and responsible AI, I am sure the world will figure out ways to apply it for benefit for humanity.","excerpt":"“The model isn’t accurate in admitting its limitations,” reads GPT-4 paper. A crucial point to note for every single user as well.","categories":["AI Features"],"tags":["AI (Artificial Intelligence)","ChatGPT","GPT4","OpenAI"],"author_name":"Mohit Pandey","publish_date":"2023-03-16T14:00:00","publication_year":"2023","word_count":1325,"keywords":["Go","ChatGPT","TPU","OpenAI","AI","GPT-5","chatbots","RAG","Aim","R","AI (Artificial Intelligence)","GPT4"],"extracted_tech_keywords":["AI","GPT-5","ChatGPT","OpenAI","Aim","RAG","chatbots","TPU","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/doomsday-will-be-triggered-by-gpt-4\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":43525,"title":"Beginner’s Guide To Geographical Plotting With Plotly","content":"One of the coolest and creative things about Data Science is visualization. It is where we put our technical skills to work in correlation with art. Data visualization is not an easy task but is an enjoyable task where we get to turn raw and unattractive data into bright and colourful images that can even move and interact. Geographical plots are the same, they are beautiful and when done right, highly informational. Choropleth maps or maps that use colours to distinguish different regions based on certain factors have been in use for centuries. Today it plays a vital role in Data Analysis and is a major part of almost any research dashboards that has nationwide or worldwide data to expose. In this article, we shall learn how to create geographical plots with colourful maps using plotly. If you are new to plotly check out our article on how to begin with plotly here. Geographical plotting With Plotly Importing the libraries We will start by importing all the modules necessary to create our visualizations offline. import plotly.plotly as pl import plotly.graph_objs as gobj import pandas as pd from plotly.offline import download_plotlyjs,init_notebook_mode,plot,iplot init_notebook_mode(connected=True) Plotting The Neighbours We will use Plotly’s geographical map to plot India and its neighbours on a choropleth map. The map is plotted using Plotly’s graph_objs module that we imported. It requires two important parameters that have to be passed as arguments, data and layout. Each of these parameters consists of a dictionary of parameters and arguments that are referenced by the module. Let’s understand what that means. We start by initialising the data and layout variables which are a dictionary of parameters and arguments that are required by the graph_objs.Figure class to plot the maps. We then create an object for the graph_objs.Figure class by passing the initialized variables data and layout as arguments. Finally, we plot the graph using the plot method. To plot India and its neighbours, we will use the following code blocks. #initializing the data variable data = dict(type = 'choropleth', locations = ['india','nepal','china','pakistan','Bangladesh','bhutan','myanmar','srilanka'], locationmode = 'country names', colorscale= 'Portland', text= ['IND','NEP','CHI','PAK','BAN','BHU', 'MYN','SLK'], z=[1.0,2.0,3.0,4.0,5.0,6.0,7.0,8.0], colorbar = {'title':'Country Colours', 'len':200,'lenmode':'pixels' }) type : ‘choropleth’ specifies that we are plotting a choropleth map. locations : The names of countries we want to plot. locationmode : It specifies that the plotting level is country wise. The value can be one of 3,-“ISO-3” , “USA-states” , “country names”. colorscale : The colour set used to plot the map.Available color scales are ‘Greys’, ‘YlGnBu’, ‘Greens’, ‘YlOrRd’, ‘Bluered’, ‘RdBu’, ‘Reds’, ‘Blues’, ‘Picnic’, ‘Rainbow’, ‘Portland’, ‘Jet’, ‘Hot’, ‘Blackbody’, ‘Earth’, ‘Electric’, ‘Viridis’, ‘Cividis’ text: The textual information that needs to be displayed for each country on hover. z: The value or factor that is used to distinguish the countries. These values are used by the colour scale. colorbar: A dictionary of parameters and arguments to customize the display of colorbar.Used to control the properties of the colorbar such as length, title, axis etc. #initializing the layout variable layout = dict(geo = {'scope':'asia'}) geo: The parameter sets the properties of the map layout. The scope parameter sets the scope of the map. Scope can have any of the 7 values- “world” | “usa” | “europe” | “asia” | “africa” | “north america” | “south america” . # Initializing the Figure object by passing data and layout as arguments. col_map = gobj.Figure(data = [data],layout = layout) #plotting the map iplot(col_map) Comparing India’s Population With Its Neighbours Now that we know how to create a basic choropleth, we will use that knowledge on a real-world dataset. Go ahead and download the world population dataset by clicking here. Extract the zip file and rename the main data file which contains the year wise population data of all the countries. Here I have renamed the file as population.csv. Let’s get going! Step 1: Load the dataset and prepare it to fit the problem statement We are trying to plot only the population of India and its neighbours for the year 2018, So lets extract that information from the dataset. #Reading the dataset pop = pd.read_csv('population.csv') #Selected columns from the complete dataset cols = ['Country Name','2018'] #Countries that we require the data of countries  = ['India','Pakistan','China','Bangladesh','Nepal','Bhutan','Myanmar','Sri Lanka'] #Overwriting the data with only details of countries we need. new_pop = pop[cols][pop['Country Name'].isin(countries)] Let’s have a look at the new data: Step 2: Plotting on the map #Initializing the data variable data = dict(type = 'choropleth', locations = new_pop['Country Name'], locationmode = 'country names', autocolorscale = False, colorscale = 'RdBu', text= new_pop['Country Name'], z=new_pop['2018'], marker = dict(line = dict(color = 'rgb(255,255,255)',width = 1)), colorbar = {'title':'Colorbar Title','len':0.25,'lenmode':'fraction'}) Here we pass the data to the map directly from the data frame. For example, we set the locations by passing the pandas.series (pop[‘Country Name’]). Similarly, we pass the values that compare the countries (population in the year 2018) as a pandas series. The two additional parameters autoscale and marker perform the following functionalities. autoscale: If set to true ignores the colour-scale parameter and chooses the default colour-scale to plot the map marker: controls the properties of the boundaries of location modes. In the above example, it sets a white boundary line of width 1 to separate all the countries portrayed in the map. #Initializing the layout variable layout = dict(geo = dict(scope='asia')) #Initializing the object for graph_objs.Figure class asiamap = gobj.Figure(data = [data],layout = layout) #plotting the map iplot(asiamap) World Population 2018 Now let’s try to plot the entire world population in the year 2018. We will follow the same steps as we did above except this time we will not filter specific countries. Step 1: Load the dataset and prepare it to fit the problem statement pop = pd.read_csv('population.csv') cols = ['Country Name','2018'] new = pop[cols] Step 2: Plotting on the map data = dict(type = 'choropleth', locations = new['Country Name'], locationmode = 'country names', autocolorscale = False, colorscale = 'Rainbow', text= new['Country Name'], z=new['2018'], marker = dict(line = dict(color = 'rgb(255,255,255)',width = 1)), colorbar = {'title':'Colour Range','len':0.25,'lenmode':'fraction'}) layout = dict(geo = dict(scope='world')) worldmap = gobj.Figure(data = [data3],layout = layout3) plot(worldmap) The plot method opens up the plot in a browser window which can be saved as an html document. Great! Now you can plot any data using choropleth maps with plotly !!","excerpt":"One of the coolest and creative things about Data Science is visualization. It is where we put our technical skills to work in correlation with art. Data visualization is not an easy task but is an enjoyable task where we get to turn raw and unattractive data into bright and colourful images that can even […]","categories":["Deep Tech"],"tags":["plotly"],"author_name":"Amal Nair","publish_date":"2019-07-30T12:30:48","publication_year":"2019","word_count":1042,"keywords":["data science","Go","Plotly","programming_languages:R","AI","ML","plotly","Ray","ViT","R","Pandas"],"extracted_tech_keywords":["AI","ML","data science","Ray","Pandas","Plotly","R","Go","ViT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/beginners_guide_geographical_plotting_with_plotly\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":31496,"title":"8 Scholarships From Top Ranking Universities For Data Science That Students Can Apply In 2019","content":"Studying data science, artificial intelligence, machine learning and related technologies in a foreign university might be an expensive affair. While there are a lot of sources available on the courses one can opt for in Data Science, it is hard to find detailed information about the scholarship for specialised courses. In this article, we list down top 8 scholarships that aspiring data science candidates can apply for in 2019. The article aims to list down scholarships available in top-ranking universities and should not be deemed as a comprehensive list. 1| Tsinghua-Berkeley Shenzhen Institute Scholarships, 2019 Level of Study: Master’s Degree (Data Science & Tech) Country of Study:  China Tsinghua-UC Berkeley Shenzhen Institute (TBSI) is a jointly established independent institution by the University of California, Berkeley (UC Berkeley) and the Tsinghua University, China with the initiative of building a bridge across multiple disciplines, cultures ,academia and industry to provide a platform for international collaboration, fostering future entrepreneurs and leaders in the field of science and technology. Funding Institution: Tsinghua-Berkeley Shenzhen Institute (TBSI) Expenses Covered: Up to 200 student scholarships are available. The scholarship includes tuition, Tuition fee for PhD Program: 40,000 CNY\/Year Tuition fee for Master’s Program: 33,000 CNY\/ Year Application Fee: 800 CNY Medical Insurance: 600 CNY\/year Accommodation: Around 1,000 CNY\/month for single rooms. Eligibility: Those who are studying for a master’s degree in China must have a bachelor’s degree with good standing and be under the age of 35. Should not have availed any Chinese government scholarships or Tsinghua University. Applications Deadline: 1st Round: Dec 15, 2018; 2nd Round: 8 March, 2019; Last Round: May 1, 2019 Visit here for more details. 2| University of Sydney: Data Science Scholarships, 2019 Level of Study: Master’s Degree (Data Science) Country of Study: Australia The University of Sydney, one of Australia’s premier university is offering scholarships for international students for 2019 intake to high performing candidates. Funding Institution: University of Sydney Expenses Covered: Scholarships are awarded in the field of Data Science. The scholarship is valued at $6,000 and is tenable for one year only. Eligibility: Applicants must be a graduate of a quantitative degree program. A quantitative program includes Data Science, Computer Science, Mathematics, Statistics, Engineering, Physics, Economics and more. Applicants must have achieved a minimum distinction average (equivalent to 75 at the University of Sydney) in their UG studies. Application deadline: 30 April 2019 Visit here for more details. 3| University of Adelaide Scholarships: Data Science & Software Eng. Level of Study: Masters\/ PhD Degree (Data Science) Country of Study: Australia The University of Adelaide offers Adelaide Scholarships International to highly talented overseas postgraduate students in areas of Data Science and Software Engineering. The University offers Graduate & PhD scholarships in these fields to support its ongoing broader research efforts. Funding Institution: University of Adelaide, Australia Expenses Covered: Course tuition fees for two years for a Master’s degree by research and three years for a Doctoral research degree and an annual living allowance ($26,288 in 2018 tax exempt) for two years for a Master’s degree by Research and three years for a Doctoral research degree, plus conference and travel awards. Eligibility: Candidates must have successfully completed at least the equivalent of an Australian First-Class Honours degree. International applicants must not hold any research qualifications. Application deadline: 30 April 2019 Visit here for more details. 4| INSOFE Scholarships Level of Study: Post Graduate Diploma in Big Data Analytics Country of Study: India (Hyderabad\/Bengaluru) INSOFE offers scholarships for deserving students in its postgraduate programs which prepares the graduates for challenging data science and analytics roles. INSOFE also has collaboration with companies like are Microsoft, Philips, Deloitte in which most of the graduates do their internships. Funding Institution: INSOFE Expenses Covered: INSOFE recognises top performers and rewards them for demonstrating outstanding achievement at every phase of the program based on their performance at each phase of the program. The scholarships range from Rs. 25,000 to the entire program fee. Eligibility: Admission is considered based upon qualifying the entrance test conducted by INSOFE.Final admission to the program will be based on their performance in the entrance test, prior academic background and work experience. Application deadline: Upcoming Batch: Jan 2019 (Multiple Batches in Year) Visit here for more details. 5| JAD JMS Scholarship MS Program in DS & Entrepreneurship 2019 Level of Study: Master’s program Data Science and Entrepreneurship (In association with Eindhoven Institute of tech & Tilburg University) Country of Study: Netherlands The JADS JMJ Scholarship is offered by the Sisters of the Society of Jesus Mary Joseph (JMJ) located in the Netherlands. This Scholarship is intended for international students from outside the EU\/EEA area interested in pursuing Master Data Science and Entrepreneurship degree. Funding Institution: JADS-JMJ Expenses Covered: Students who are awarded the JADS JMJ Scholarship will receive a tuition fee waiver of €5000, – for the first academic year and €5000, – for the second academic year Eligibility: The selection will be based on the student’s academic performance, motivation, and interview. It does not apply for the Pre-Master Data Science and Entrepreneurship. Application deadline: 18th of May 2019 Visit here for more details. 6| ETH Zurich Excellence Masters Scholarships Level of Study: Master in Data Science Country of Study: Switzerland The university supports exceptional students wishing to pursue a master’s degree at ETH with two scholarship programmes — the Excellence Scholarship & Opportunity Programme (ESOP) and the Master Scholarship Programme (MSP). Funding Institution: ETH Zurich Expenses Covered: The Scholarship consists of a grant covering living and study expenses (CHF 11,000 per semester), including a partial stipend for living and study expenses (CHF 6,000 per semester). Eligibility: To be considered for the scholarship, you must apply for a master’s degree programme at ETH Zurich. The prerequisite is a very good result in their Bachelor’s (top 10% =grade A). Application deadline: 15th December 2018 (Multiple Intakes) Visit here and here for more details. 7| Justus & Louise van Effen Excellence Scholarships Level of Study: Master of Science Computer Science (Track: Data Science & Technology) Country of Study: Netherlands The Delft University of Technology offers several excellence scholarship programs for international students, such as the Justus & Louise van Effen Scholarship. It aims to financially supporting excellent international MSc students wishing to study at TU Delft. Funding Institution: TU Delft Expenses Covered: Scholarships are available as 3 per faculty. The scholarship includes full tuition fees for the MSc programme and living expenses (€950 pm) for 2 years and one-time standard payment for travel expenses. Eligibility: Applicants admitted must have a CGPA of 80 percent or higher of the scale maximum in their bachelor’s degree from an internationally renowned university. Application deadline: 25th December 2018 (Multiple Intakes) Visit here for more details. 8| Westminster Full Int. Scholarships (UK) Level of Study: BSc Honors (Data Science & Analytics) Country of Study: UK Westminster Full International Scholarships are the University’s most competitive scholarships that are open to international students from developing countries who wish to pursue a full-time Undergraduate degree at the University. Funding Institution: University of Westminster Expenses Covered: Full tuition fee waivers, accommodation, living expenses and flights to and from London. Eligibility: Candidate must be an international student from a developing country and hold an offer for a full-time undergraduate degree at the University. Other criteria are academic excellence, development potential and financial need. Application deadline:  31st May of 2019 (Multiple Intakes) Visit here & here for more details. Note: To get admission in foreign universities\/Institutes mentioned above fulfilling English language eligibility through (TOEFL or IELTS) is must. Other Few Scholarships To Look Out For: University of Amsterdam Data Science Program: Amsterdam Merit Scholarships TUM Munich, Data Science Masters: DAADS Scholarships LMU Munich, Data Science Masters: DAADs Scholarships Accenture Data Science Scholarship: Barcelona GSE Master’s Degree in Data Science.","excerpt":"Studying data science, artificial intelligence, machine learning and related technologies in a foreign university might be an expensive affair. While there are a lot of sources available on the courses one can opt for in Data Science, it is hard to find detailed information about the scholarship for specialised courses. In this article, we list […]","categories":["AI Trends"],"tags":["Data Science","education","Insofe","scholarship"],"author_name":"Martin F.R.","publish_date":"2018-12-13T13:13:33","publication_year":"2018","word_count":1293,"keywords":["big data","data science","Go","artificial intelligence","machine learning","AI","education","Insofe","RAG","Aim","analytics","scholarship","Data Science","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","data science","analytics","Aim","RAG","R","Go","big data"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/8-scholarships-top-universities-data-science-2019\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10095795,"title":"Databricks Announces Acquisition of Generative AI Platform MosaicML for $1.3 Bn","content":"Databricks, has announced its definitive agreement to acquire MosaicML, a generative AI platform, in a transaction valued at approximately $1.3 billion. The acquisition aims to make generative AI accessible to organisations, allowing them to “build, own and secure generative AI models with their own data.” The CEO of Databricks, Ali Ghodsi, emphasised the goal of democratising AI and making the “Lakehouse the best place to build generative AI and LLMs.” MosaicML was co founded by Naveen Rao and Hanlin Tang in San Franscico in 2021. Rao expressed enthusiasm about the acquisition, stating that joining forces with Databricks will help make their vision of empowering everyone to build and train their own AI models a reality. The proposed acquisition is subject to customary closing conditions, including regulatory clearances. MosaicML is recognized for its state-of-the-art MPT large language models, which have been downloaded over 3.3 million times. It has demonstrated how organisations can quickly build and train their own models using their data in a cost-effective manner. Notable customers such as AI2, Generally Intelligent, Hippocratic AI, Replit, and Scatter Labs leverage MosaicML for various generative AI use cases. The Databricks Lakehouse Platform, combined with MosaicML’s technology, will provide customers with a “simple, fast way to retain control, security, and ownership over their valuable data without high costs.” MosaicML claims that its automatic optimization of model training enables 2x-7x faster training compared to standard approaches. Furthermore, the scalability of resources allows for training multi-billion-parameter models in hours, rather than days. The collaboration between Databricks and MosaicML is expected to significantly reduce the cost of training and using LLMs, making it accessible to a wider range of organizations. The entire MosaicML team, including its research team, is expected to join Databricks after the transaction closes. MosaicML’s platform will be integrated and scaled over time to provide a unified platform where customers can “build, own and secure their generative AI models.” This collaboration aims to offer organisations greater choice in building their own models, training them with their unique data, and creating intellectual property specific to their businesses. The acquisition aligns with the shared vision of Databricks and MosaicML to democratise AI and give organisations more control over their data. MosaicML, had earlier, in the month of May launched a new service called MosaicML Inference to compete with OpenAI and other similar platforms on price. The offering aims to assist enterprise customers in training and deploying generative AI-powered solutions. As large models like ChatGPT and Stable Diffusion gain popularity, organisations are increasingly interested in leveraging these technologies. MosaicML highlights that while existing solutions are suitable for those unconcerned with cost and willing to share data, building and hosting their own models can be a more cost-effective option for organizations with strict data privacy requirements. The company claims to make its technology accessible to all organizations at a significantly lower price, up to 15 times cheaper than its competitors. CEO Naveen Rao mentioned that MosaicML has optimized its offerings to achieve cost savings. By offering the service at a much lower price, the company still maintains profitability. MosaicML has developed foundational models to support its new service, similar to the technology used in chatbot-enabled products by Microsoft and Google. While MosaicML, like its competitors, provides its technology for rent, it differentiates itself by also offering its code to customers. This allows customers to run the code on their own hardware, ensuring the confidentiality of their data from MosaicML. This approach appeals to corporate customers who prioritize data privacy, as the value of an AI system is heavily dependent on the training data used.","excerpt":"Databricks, has announced its definitive agreement to acquire MosaicML, a generative AI platform, in a transaction valued at approximately $1.3 billion. The acquisition aims to make generative AI accessible to organisations, allowing them to “build, own and secure generative AI models with their own data.” The CEO of Databricks, Ali Ghodsi, emphasised the goal of […]","categories":["AI News"],"tags":[],"author_name":"Shyam Nandan Upadhyay","publish_date":"2023-06-26T22:01:21","publication_year":"2023","word_count":597,"keywords":["Go","ChatGPT","OpenAI","AI","ML","RAG","Aim","generative AI","R","Databricks"],"extracted_tech_keywords":["AI","ML","generative AI","ChatGPT","OpenAI","Aim","RAG","Databricks","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/databricks-announces-acquisition-of-generative-ai-platform-mosaicml-for-1-3-bn\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10117840,"title":"Meet the Winners of the ‘Best Firm for Diversity &amp; Inclusion in Tech’","content":"At India’s biggest diversity and inclusion summit, the Rising 2024, AIM celebrated organisations for their best diversity, equity, and inclusion practices. These included implementing effective programmes to promote diversity, providing equitable opportunities for career advancement, offering coaching and mentorship, and contributing to closing the gender pay gap. In collaboration with AIM Best Firm Certification, the Rising acknowledges tech firms’ achievements and steadfast dedication to shaping a more inclusive future for Diversity and inclusion in technology. The winners were chosen by our distinguished panel of judges from the AIM Leaders Council, alongside the AIM Best Firm Certifications team and industry experts in the field. Here are the winners of the Best Firms for Diversity & Inclusion in Tech: (In Alphabetic Order) All State India Pvt Ltd Allstate India, a subsidiary of The Allstate Corporation, pioneered operations and technology and has evolved into a strategic business services arm. It provides the parent organisation with technology, innovation, accounting, policy administration, and global operations expertise. Allstate India exemplifies a vibrant culture of diversity and inclusion, seamlessly integrating values of respect, trust, and empathy into its organizational fabric. Their commitment to empowering women, supporting differently-abled individuals, and celebrating the LGBTQ+ community, coupled with their dedication to employee well-being and professional growth, truly makes them a beacon of inclusive excellence. ANZ For more than 34 years, ANZ’s Bengaluru Global Capability Center has been home to over 8,000 employees, comprising a substantial segment of its workforce. Crucial to ANZ’s achievements, the centre offers various services, encompassing banking operations, economic research, risk analytics, and technology support, contributing significantly to the company’s success. ANZ’s commitment to diversity and inclusion shines as a cornerstone of its identity, fostering a workplace where varied perspectives and backgrounds are not only welcomed but celebrated. Their broad and thoughtful approach emphasises gender equality, cultural diversity, accessibility, and LGBTIQ+ inclusion, reflecting their operation in nearly 30 global markets. ANZ recognizes the richness that such diversity brings to decision-making and innovation, making them not only a better bank for their customers but a model of inclusive excellence in the corporate world. Chubb Chubb, a distinguished force in the global insurance industry, delivers a comprehensive array of commercial and personal property and casualty, accident, health, reinsurance, and life insurance solutions across 54 nations. Renowned for its commitment to excellence, Chubb caters to a varied clientele on a global scale. Chubb’s diversity and inclusion culture is a dynamic blend of respect, empowerment, and recognition. It embraces a wide spectrum of backgrounds and perspectives to foster an inclusive and thriving workplace. DBS Tech India DBS Tech India, founded in 2016, is DBS Bank’s premier offshore technology center in Hyderabad. Pioneering excellence and innovation, it drives the bank’s future through experimentation, committed to building a better tomorrow for customers and the world. bring people and technology together to drive digital transformation – enabling our customers to Live more, Bank less. DBS TechIndia prioritizes not only demographic diversity but also values experiential and cognitive diversity in team building. Women constitute 40% of our senior management, leading key businesses and functions throughout the bank. Interestingly, DBS’s unique approach to diversity, equity, and inclusion is rooted in its Asian heritage, creating a workplace that thrives on varied perspectives and experiences, and is dedicated to empowering its employees to make meaningful differences in their communities and careers. Eventbrite Eventbrite, a pioneering global events marketplace, operates in nearly 180 countries, championing the experience economy. Founded with a vision to democratise event organization, it enables creators worldwide to effortlessly sell tickets to live experiences. Eventbrite is where people come together to unlock the magic of sharing an infinite reel of experiences. Eventbrite believes in promoting diversity and inclusion in the tech industry, with a strong emphasis on the representation of women and underrepresented minorities. They are dedicated to prioritizing equitable pay and advancement opportunities, ensuring that female and ethnically underrepresented employees progress and are compensated at the same rate as their majority peers. ICICI Prudential Life Insurance Co. Ltd. Since 2001, ICICI Prudential Life has maintained its position as a leading player in the Indian life insurance sector, boasting an Assets Under Management (AUM) of around 2 billion as of December 2023. With a dedicated emphasis on addressing customer needs, the company provides a range of savings and protection products, guaranteeing exceptional service, robust fund performance, and efficient claim settlement processes. ICICI Prudential Life is committed to fostering diversity and inclusion as integral components of its culture, enabling every employee to authentically contribute their skills, experience, and perspective to generate unparalleled value for all stakeholders. Infocepts Infocepts, a premier data solutions provider, connects the realms of business and analytics, equipping enterprises with data, AI, and intuitive analytics to drive superior results. Grounded in transparency and trust, it nurtures innovation and professional advancement, striving for a diverse, inclusive, and dynamic work environment. Infocepts is dedicated to fostering an inclusive, secure, and diverse workplace environment. The Economic Times Best Organizations for Women 2023 serves as a platform dedicated to recognizing the exceptional efforts of companies that excel in fostering an inclusive workplace environment conducive to the success of women. It honors organizations that demonstrate a strong commitment to gender diversity through their comprehensive policies, practices, and cultural initiatives. L&T Technology Services (LTTS) L&T Technology Services, also known as LTTS, is a premier Engineering R&D provider committed to championing diversity, equity, and inclusion. Catering to a diverse range of global clients, which includes 69 Fortune 500 companies, LTTS offers cutting-edge technology services across various industries. With a workforce of over 23,300 employees worldwide, LTTS nurtures an inclusive work environment, guaranteeing equal opportunities and offering facilities tailored to accommodate individuals with disabilities. L&T’s diversity extends beyond India to encompass multiple countries, reflecting a rich multicultural ethos. This is underscored by their workforce comprising individuals from 52 nationalities, spanning 36 domiciles within India, and speaking 80 distinct languages. Moreover, the company actively foster gender diversity and inclusion through a well-defined framework. A DEI Charter has been formulated, grounded on four pillars: Induct, Develop, Engage, and Enable, to ensure a systematic approach towards these objectives. MathCo MathCo, short for TheMathCompany, stands out as a premier Enterprise AI and Analytics firm serving elite Fortune 500 enterprises. Since its inception in 2016, MathCo has honed its expertise in customizing AI solutions, strategically capitalizing on growth opportunities, and spearheading innovative problem-solving approaches. At MathCo, we empower our team of Mathemagicians not only to excel but also to leave a profound mark in the domain of analytics and AI. MathCo empowers women to realize their potential and have fulfilling careers by fostering a supportive environment where they can exchange knowledge and insights. The global community, Women in Data, aims to boost diversity in data careers, promoting inclusivity and diversity in analytics. Moody’s Corporation Moody’s, a renowned authority in research, data, and analytics, is dedicated to aiding market participants in navigating risks and seizing opportunities. Drawing upon over 115 years of experience, innovative technologies, and a commitment to inclusivity, Moody’s facilitates informed decision-making, instilling confidence in its users. The DE&I efforts at Moody’s strive to positively impact the workforce, workplace, customers, and communities. By embracing diverse backgrounds and experiences, they enhance employee contributions, expand leadership opportunities, and improve the quality of work, including opinions, products, and services. Quest Global Quest Global, founded in 1997, has emerged as a leading provider of engineering research and development services. With a presence in 17 countries, boasting 72 global delivery centers, and employing over 17,800 professionals, Quest Global is at the forefront of delivering innovative engineering solutions across various industries. Each day, it transforms challenges into opportunities, consistently pioneering cutting-edge solutions. Quest Global fosters an inclusive culture, promoting collaborative innovation for a safer, sustainable, and better world. They offer equal opportunities and value individuals regardless of protected characteristics like disability, race, religion, sexual orientation, or identity, ensuring a discrimination-free environment. Unisys Unisys, a renowned global technology solutions provider, empowers breakthroughs for premier organizations worldwide. Its suite of solutions, spanning cloud, data and AI, digital workplace, logistics, and enterprise computing, empowers clients to defy conventions and unleash their utmost potential. At Unisys, we’re dedicated to fostering an inclusive environment where everyone thrives and feels valued. They believe in the power of diversity and inclusion to transform our workplace and enhance our ability to serve clients. Through respect, openness to ideas, and equal opportunities, we champion a culture of inclusivity for all. Verizon Verizon, With over 7000 diverse employees, cultivates an environment of collaboration and authenticity. Positioned strategically, it serves as a hub for seamless customer-centric solutions, placing emphasis on innovation and diversity. Upholding DEI principles, it stands as an equal-opportunity employer, nurturing a culture of learning and empowerment through dedicated employee resource groups. Verizon values diversity in all forms – race, nationality, religion, gender, sexual orientation, gender identity, disability, and age. They believe diversity enhances company culture, leading to better service for customers through the collective intelligence, spirit, and creativity of their diverse team.","excerpt":"At the Rising 2024, AIM celebrated organisations for their diversity, equity and inclusion best practices.","categories":["AI Highlights"],"tags":[],"author_name":"Gopika Raj","publish_date":"2024-04-08T21:57:14","publication_year":"2024","word_count":1492,"keywords":["API","AI","ML","Git","Ray","Aim","analytics","Rust","GAN","R"],"extracted_tech_keywords":["AI","ML","analytics","Aim","Ray","R","Rust","Git","API","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/meet-the-winners-of-the-best-firm-for-diversity-inclusion-in-tech\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10134851,"title":"Reflection is Going Through a Phase of Self-Reflection","content":"A few days ago, Matt Shumer, the founder of OthersideAI, announced that the company had made a breakthrough, which allowed them to train a mid-size model, achieving SOTA-level performance with the launch of Reflection, which outperforms GPT-4o and Claude Sonet 3.5. But, the hype was short-lived. Upon trying out the model, many users claimed that the Reflection API was just a wrapper of Claude Sonnet 3.5 or GPT-4o. They revealed that when given the same prompt, both Claude and Reflection gave exactly the same answers. A story about fraud in the AI research community:On September 5th, Matt Shumer, CEO of OthersideAI, announces to the world that they've made a breakthrough, allowing them to train a mid-size model to top-tier levels of performance. This is huge. If it's real. It isn't. pic.twitter.com\/S0jWT8rDVb— 𝞍 Shin Megami Boson 𝞍 (@shinboson) September 9, 2024 What Went Wrong? It all started when the users tried achieving the same results as shared by the creators of Reflection AI. But the model completely missed the mark. Artificial Analysis, known for its independent analysis of AI models and API providers, compared Reflection AI 70B to other models. It failed miserably, and the results were poor compared to Llama 3 70B. A Reddit user claimed that the Reflection model was trained to give false answers first in its thinking phase, and then it reflects the thinking phrase. “If you ask what 2+2 is, the default example on the Hugging Face page will say something like 2+2=3. Oh wait, I’ve made a mistake; 2+2 is actually 4. If the thinking is actually hidden, it might work, but it’s quite strange,” he added, further explaining the present flaw in the model. Shumer claimed that the Reflection models were the best open-source model to date. They use the reflection-tuning method, designed to teach AI models to recognise and correct their own mistakes. This approach seemed poised to address one of the most persistent challenges in language models: the tendency to “hallucinate” or generate inaccurate information. “When LLMs make mistakes, they often treat their errors as facts. If we could teach these models to think more as humans do—to reflect on their behaviour and recognise their mistakes—the models would become smarter and more reliable,” said Shumer, suggesting why reflection-tuning can help models reason better. When the model generates an answer, it outputs its reasoning, and the <thinking> tag surrounds the thought process with special tags (such as). When the model detects an error during inference, it marks the error with a <reflection> label and corrects itself. This feature enhances the reliability of the model, especially when dealing with complex problems. When Artificial Analysis came up with poor results, they were granted access to private APIs of Reflection models. And the performance then was way better than the previous results. But again, when they compared the performance of the given private APIs with the available models on Hugging Face, the results were completely different, as the model hosted on Hugging Face showed poor results. Self-Reflection Needed Meanwhile, users have called Reflection a mere wrapper of Claude AI. When the model was made available on OpenRouter, users reported that it used a dumbed-down version compared to the previous version, as the one made available on OpenRouter was heavily censored. “The version on OpenRouter seems to be heavily censored\/dumbed down; it just refuses to write about what I asked for, while the “original” version did fine. So it was probably ChatGPT or Llama3+ChatGPT for Reflection initially, and now he switched to Claude, which is known to be heavily censored,” a Reddit user shared his experience with OpenRouter. Shumer first blamed the upload process and mentioned that something might go wrong while uploading weights on Hugging Face but that didn’t solve the problem. So, he went a step further and decided to start the training from scratch to eliminate all the issues. The API drama aside, an important reason why Reflection is not able to perform better than Llama is the use of different formats. LLaMA 3.1 70B was trained and uploaded using BF16 (Brain Float 16), while Reflection 70B was converted to FP16 (Float 16). Converting a model from BF16 to FP16 results in significant information loss, and degrades the model performance. A Reddit user solved a classic trolly problem by adding “it’s not the usual one” to the prompt in a single shot, suggesting the reasoning capabilities of the reflection-tuning method. Can Retraining Do the Trick? Shumer mentioned that ideally, this shouldn’t have happened in the first place. He said that his team has tried everything they could, but the performance they get from Hugging Face is nowhere close to what it ideally should be while running the Reflection model locally. Some users believe that the whole release of the Reflection model was actually an advertisement for GlaiveAI as Shumer owns part of Glaive and was seen promoting it when he released the Reflection model. In response to that, Shumer said that he is a tiny investor and has only invested around $1000 in Glaive. Here, it’s also important to note that this was the first release of the Reflection model, which was praised for its reflection-tuning approach. It would be a good idea to wait for the next update\/release before judging the model too harshly.","excerpt":"And the creator of Reflection 70B, Matt Shumer, is trying really hard to make things alright.","categories":["AI Features"],"tags":["AI adoption"],"author_name":"Sagar Sharma","publish_date":"2024-09-10T13:00:00","publication_year":"2024","word_count":881,"keywords":["Go","ChatGPT","Hugging Face","TPU","AI adoption","API","AI","GPT-4o","GPT","Aim","R"],"extracted_tech_keywords":["AI","GPT-4o","ChatGPT","Aim","Hugging Face","TPU","R","Go","API","GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/reflection-is-going-through-a-phase-of-self-reflection\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10134499,"title":"Zerodha Co-founder Nikhil Kamath Unveils ‘Innovators Under 25’ to Support Young Tech and AI Entrepreneurs in India","content":"At the recent WTFund Summit, held in Mumbai, Nikhil Kamath, co-founder of Zerodha, unveiled ‘Innovators Under 25.’ The occasion marked the launch of WTFund, India’s first non-equity grant fund, dedicated to fostering innovation among young entrepreneurs under 25. The event brought prominent figures including Jeet Adani, Ananya Birla, Nithin Kamath, Jay Kotak, Shashwat Goenka, and Kishore Biyani, among others. WTFund’s First Cohort of Innovators At the summit, Kamath unveiled the inaugural cohort of WTFund grantees—fifteen entrepreneurs with bold ideas spanning industries such as healthcare, AI, and sustainable consumer goods. The initiative aims to empower young minds with the necessary resources to drive innovation and disrupt conventional industries. Kamath emphasised the importance of youth-driven entrepreneurship in shaping India’s future. “Entrepreneurs are ageless; you can build the next big thing whether you are 25 or 80. But, let’s face it—the energy, drive, and sheer audacity often come easier when you’re young,” he stated. Kamath expressed his ambition to support thousands of startups in the future, remarking, “From our dream of funding 9 companies today to aiding 9,000 young companies one day, we’re committed to building an India that isn’t afraid to take risks and step outside our comfort zones. The next decade belongs to India, and it will be driven by those who choose to build their own dreams rather than helping someone else realise theirs.” The fifteen grantees come from a variety of sectors, showcasing the breadth of innovation taking place in India. Selected startups include: Pamawel: Plant-based, non-steroidal, FDA-approved pain relief formulations targeting menstrual pain. Mars: Virtual Workstations that eliminate hardware limitations, providing affordable solutions for creators and builders in AI and no-code spaces. Oh! Nuts: Premium, healthy nut-based snacks tailored for the growing Indian consumer market. Pawsible Foods: Sustainable, plant-based pet food utilising Kavaka™ mycoprotein. RNT Health Insights: AI-assisted diagnostic software for gastrointestinal cancer detection. Biocompute: Affordable DNA data storage chips, revolutionising data centre efficiency. CallPrep: AI-based B2B SaaS for sales preparation, integrating LinkedIn, CRM, and buyer intent data. Urban Animal: India’s first dog DNA testing service, redefining pet care. Pixa AI: AI-enhanced plush companions, merging the toy and AI companionship markets. Support Beyond Financial Grants WTFund offers more than just financial assistance, providing up to INR 20 lakh in non-equity grants. The young entrepreneurs retain full ownership of their businesses while receiving mentorship from industry leaders. In addition to this guidance, they gain access to legal and accounting assistance, as well as strategic partnerships with leading companies, giving them a solid foundation to grow their businesses. Kamath’s vision for the fund goes beyond financial backing, with a focus on empowering young entrepreneurs to scale their ventures while making meaningful industry contributions. The resources provided by WTFund are designed to enable these innovators to overcome the challenges of early-stage entrepreneurship, preparing them for long-term success. The WTFund Summit, with its gathering of prominent Indian business leaders and a focus on fostering the next generation of innovators, signals a new era for India’s startup ecosystem. By offering critical support to young entrepreneurs, the fund sets the stage for transformative advancements across a wide range of industries.","excerpt":"Under the WTFund, the founders will receive 20 lakhs in non-equity grants and mentorship from industry experts.","categories":["AI News"],"tags":["Zerodha"],"author_name":"Vandana Nair","publish_date":"2024-09-05T14:21:28","publication_year":"2024","word_count":516,"keywords":["Go","API","funding","AWS","AI","innovation","RAG","Aim","Zerodha","R","startup"],"extracted_tech_keywords":["AI","Aim","RAG","AWS","R","Go","API","innovation","startup","funding"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/zerodha-co-founder-nikhil-kamath-unveils-innovators-under-25-to-support-young-tech-and-ai-entrepreneurs-in-india\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":28387,"title":"The Inspiring Story Of Ankush Bhandari Who Transitioned From A Tiffin Service Provider To A Data Scientist","content":"There are several success stories in the Indian analytics sector. But only seldom has one come across as unique one as that of Ankush Bhandari from IMS Proschool. He began his career as a provider of tiffin services in Pune and changed his career 360-degrees to that of a noted data scientist in the community. As Bhandari aptly puts it, from “feeding stomachs through his tiffin service, to feeding the brain through data science education and training,” his journey has been an eventful one. In a candid interview with Analytics India magazine, Bhandari talks in detail about his career path, life choices, and gives advice to students or professionals who are looking at exploring other avenues in their careers. Tell us about your transition from a tiffin service provider to a data scientist. After completing my Masters in Economics from Fergusson College, Pune, I started my entrepreneurial venture as a tiffin service provider. Since, starting a tiffin service needed low capital and there is ease in acquiring customers, it became my first choice. Within 6 months of inception, I began delivering more than 500 tiffin’s daily. However, due to some unavoidable circumstances, I closed the venture and joined IMS Proschool Pvt Ltd – an initiative of IMS Learning Resource in 2009. Due to my entrepreneurial background, I started working with New Venture Development. In 2011, I got promoted as a Product Manager – Financial Planning and I produced and co-authored 5 books on Financial Planning with Tata McGraw-Hill Publication. During this time, I also started learning Digital Marketing and built a digital marketing team to increase our online presence. In the year 2014, I got promoted to Head (India) –IMS Proschool CPS. In 2016, we launched Business Analytics Course and in 2017, 10 months Data Science Course. To use analytics for internal functions and to enhance the product, I completed the Microsoft Certified Professional Program in Data Science, which equipped me with the proper understanding of overall data science ecosystem. I have been using these skills to make sense of the data that we are collecting in Academic, Marketing and Operational field and also working on a project to find the factors which are contributing most in the successful completion of the course. Currently, IMS Proschool – CPS is training more than 7,000 students annually in Data Science, Analytics, Finance, Account and Digital Marketing courses across 10 cities in India. You have gained considerable knowledge from MOOCs (Massive Open Online Courses). Tell us how you used them to your best advantage. A decade ago, knowing one skill was enough for working in a field for the rest of one’s life. But, in the current business environment, things are changing very fast. So, to remain relevant and competitive, one needs to learn continuously. To learn new skill sets you can take the help of either classroom program or MOOCs. In my case due to the busy schedule, I used MOOCs to build new skills which were needed to take the company business forward. With the help of MOOCs, I completed the CFP Certification, Google Certification in Adwords and Analytics, Financial Modelling, Chartered Global Management Accountant ( CIMA UK and CGMA). Recently, I became one of the 13th Google Certified Trainer in India. I also completed the Microsoft Professional Program in Data Science through MOOCs. Continuous learning and acquisition of new skills through MOOCs has helped me to reach this stage where I am heading the institute to train working professionals on new-age programs. Should aspiring data scientists work with MOOCs? Or should they opt for classroom training? To remain competitive and productive, continuous learning should be the goal, you can do it through MOOC or classroom training. If you are self-motivated, have a previous background and someone to guide then you can go for MOOCs. But if you want to learn in a group, need a mentor to guide you, and are unfamiliar with the topic, then classroom program is more suitable. You said you are using data science to create educational products and at the same time also using your education experience to create data science product? Can you give us some examples? Like many training providers, we also started teaching Data Science using well-known data sets of IRIS, Loan Prediction etc. Though these data sets are useful in explaining the concepts, students were finding difficulties in visualising the data set or understanding the problem statement. Then we started using Cricket stats to teach Regression; In this case, we gave variables like Team ranking, player ranking, the Average runs scored by players and asked them to predict the total run that can be scored in the particular match. Immediately, students asked why we didn’t include the data from the weather report, the Average run scored by players in last year and many other related queries. We had not seen such kind of probing on IRIS dataset before. This case study helped us to know that learning can be significantly improved if we take the cases in which students can relate to. So next time, if a teacher is teaching regression using the example of 2019 General election, then I am sure students will come to class half an hour before. Similarly, we have also started using data science techniques in identifying the factors (like faculty experience, faculty rating, case study used in the class etc.) which are contributing most in the success of students in a course. This analysis will help us in focusing on those factors which are contributing most to the success of students. You said your next goal is to learn AI and its uses in education. How do you plan to do it? Since I have done Microsoft Certified Program in Data Science and build models using various techniques of Machine Learning, I have a fair understanding of AI, but to become an expert in it again, I will depend upon MOOCs. Many AI experts are working on developing a solution to automate administrative work, grade subjective assessment and provide individualised learning. As an educator, I am going through various applications of AI which can help us in improving the training experience of students. For example: Using NLP to evaluate subjective assessments or using video analysis techniques to assess candidates’ presentations. From your experience as an educator, what are the things that students expect to achieve from your course? Upskilling, job change, better prospects? Whenever students asked one of our trainers, “Kya milega” after the course, he used to jokingly reply, “If you don’t do it, then “Jo mil raha hai who bhi nahi milega”. And now that is pretty much happening. The drive toward automation is making several hundred thousands of IT professionals redundant unless they are re-skilling themselves in new areas. Having said that all the terms upskilling, job change and better prospects are interrelated. In our course students upskill themselves with Python, R, Tableau, Hadoop and new techniques such as Machine Learning algorithms and applications of data Science in various domains. This upskilling eventually will help them in moving up in their organisation or a job change in the data science field. What is the next big thing according to you in the new tech field? The internet of things, blockchain, AI, virtual reality, augmented reality are the topics which are causing a great deal of hype and excitement, not just in the technology circle but in the wider business world, too. However, the idea that putting them together could result in something even higher than the sum of its parts, is something which has begun to gain traction, for example, IOT + Blockchain, AI + VR . For example, Researchers from the University of Berkley have developed a platform that “trains” the artificial intelligence behind a physical robot, like a robot arm. Their training works by “exposing” up to a thousand objects to the AI in a virtual world so it understands everything about them, including how to operate them. In the coming future, you will see more and more such convergence to find out innovative solutions that we have never imagined before. I think very soon a time will come when your robot will interview my robot. Then your Robot using AI \/ NLP convert the chat into a Virtual Reality show, which can be targeted using various IOT devises through a secured Blockchain technology. Seeing this trend, one of my friends jokingly said that he wants to retire in a digitally disconnected village. I hope Modiji will keep some villages away from Digital India, so he that can retire their peacefully. What trends have you observed in data science since 2009? I have started following Data Science since 2013 and there are few trends which I have observed: The democratisation of knowledge: New trend in data science is the democratization of knowledge. Earlier, the data scientists were supposed to know everything (programming, algorithms, application) and business leaders were dependent on them for their expert advice. But gradually this is changing because of adoption of new tools that typically obviate the programming aspect and provide user-friendly GUI (Graphical User Interface) so that anyone with basic knowledge of coding and algorithms can simply use them to build high-quality machine learning models. Some of the popular tools are Azure ML, Google AutoML, AWS ML, IBM Watson and Rapid Miner. These tools are empowering the decision makers and managers to use the power of technology themselves to find out innovative solutions related to their field. Google auto ML has already prepared a solution where you need to upload images in the system to develop image identification tools. All the works of data cleaning and decision regarding algorithms are left to machine only. Integrated cloud platform: Earlier most of the components of Data Science were working in isolation, for example, saving data in an in-house server, analysis of it using SAS or R, then deployment of the solution using a different platform. Now due to the rapid expansion of cloud services such as Google Cloud, Microsoft Azure, AWS, IBM Watson, it is possible to store data, apply ML algorithms and then deploy the solution using one integrated cloud service platform. This integrated cloud platform is enabling quicker decisions, more accuracy, and faster turnaround times for production Multiple Roles: Earlier the role of Analytics was mainly confined to applying statistical techniques to data to find a pattern using R or SAS. Today’s data scientists have a wider choice. They can specialise in data visualization, data science, machine learning or big data. However, most companies still prefer to hire people who have a mix of these skills. For example, data scientists who know machine learning or are experts in big data are preferred to data scientists who just have analytical skills. Wider applications: Due to the continuous evolution of this field, Data Science has now lots of applications such as Image\/Speech\/Character Recognition. Using this application, Facebook recognises your friend when you upload a photo with them. Similarly, the recommendation like “suggested videos” on YouTube, are also powered by Data Science Availability of talent: Earlier, the biggest worries of companies while launching Data Science project was the paucity of talent to execute the projects. But now due to awareness about the data skills and availability of various learning options, many professionals have learned the skill sets and are ready to work in the field of Data Science. What were the difficulties\/challenges you faced in your journey? For any professional, the biggest challenge is to balance work and family, and in my case, I will add “study” as well. I have managed it by not getting hooked to social media and TV and by staying closer to my office. With an economics background, learning technology-oriented subjects such as Digital Marketing, Data Science, Technology in Education etc. was another big challenge. But if you give sufficient time and focus and get the right resources to prepare, building diverse skills is easier. And the most amazing challenge is solving the diverse queries from my team members in a day, for example- “How to prepare a case on Microsoft Azure?”, “How to optimise FB campaign?”, “How to conduct management accounting exam, to name a few. Finally, after wearing too many hats in a day, I come home and spend time with my little daughter which rejuvenates me to spend another two hours at night to study some new topic. What are the roadblocks (challenges) for an aspiring data scientist? The roadblocks for an aspiring data scientist are distractions, fixed mindset and competitive job market. Previously, the lack of quality content or mentorship were the issues, but not anymore due to the availability of training providers and MOOCs. Netflix and Facebook examples are the most used examples to promote the use of Data Science. We talk about how they use Analytics to target the user with the most relevant content. But excessive content is distracting users. I advise students not to visit Facebook frequently and mute most of the groups on WhatsApp to keep you focused on their studies. Another roadblock is a fixed mindset. In the data science class, if a student is from commerce background, then many times she believes that it is difficult for her to understand coding and a tech guy thinks understanding finance and marketing is complicated. From my own experience, I can say that if you have a growth mindset where you believe that if you make efforts, you can learn the subjects from other domains, then learning is easier. The competitive job market is another roadblock for the aspiring data scientist. Earlier, with some basic knowledge of R and Statistics, students used to get jobs in analytics but not anymore due to the availability of talent pool. Many candidates leave the transition path in between due to this competition. However, it does not mean that few jobs are coming in data science. Now the adoption of data science is becoming more extensive, so there are more job opportunities. But due to the mature market, companies are accepting a good understanding of R or Python, conceptual understanding of algorithms and sound business logic from the candidates. After completion, of course, either classroom or self-study, it is advisable to freshers to participate in the competition like Kaggle. For working professionals apart from participating in Kaggle, they can also build the models on the data available in the company where they are working. One of our students who transitioned from Marketing to Data Science, without being asked by his senior he used to prepare a model on data available to him in the company and present the observations to his senior. Later, when he appeared for the interview, he was able to showcase relevant work experience through these models. Another advice for the aspiring data scientists is, if possible to stay close to their office. I have interacted with many of the data scientists and found that during the transition period many of them have stayed closer to their office (I have not prepared a model on it, it is just my observation). This is because data science requires consistency in your study and if you are spending 2 hrs in traffic then focussing on learning may be difficult.","excerpt":"There are several success stories in the Indian analytics sector. But only seldom has one come across as unique one as that of Ankush Bhandari from IMS Proschool. He began his career as a provider of tiffin services in Pune and changed his career 360-degrees to that of a noted data scientist in the community. […]","categories":["AI Features"],"tags":["Data Analysis","Data Science","Data Scientist","indian statistical service","Interviews and Discussions"],"author_name":"Prajakta Hebbar","publish_date":"2018-09-17T13:01:17","publication_year":"2018","word_count":2540,"keywords":["Data Analysis","data science","artificial intelligence","machine learning","AWS","AI","ML","Azure ML","RAG","NLP","analytics","Data Science","Data Scientist","indian statistical service","Interviews and Discussions"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","NLP","data science","analytics","Azure ML","RAG","AWS"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/the-inspiring-story-of-ankush-bhandari-who-transitioned-from-a-tiffin-service-provider-to-a-data-scientist\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10136130,"title":"What Does the Runway and Lionsgate Partnership Tell Us About Hollywood","content":"History is in the making. Runway, an AI video startup, collaborated with Lionsgate, the maker of Hunger Games, to bring AI to film. Runway’s goal is to help the artistic community with state-of-the art tools to bring their stories to life, and now with this deal, perhaps even on the big screen as well. Interestingly, the Oscar winning movie “Everything Everywhere all at Once” employed Runaway’s tools for special effects for its famous rock scene, which saved a lot of time, costs, and manual effort. Lionsgate Entertainment is a Hollywood heavyweight which has previously produced award-winning movies and shows like Twilight, John Wick, and the like. Under this deal, Lionsgate would be able to use Runway’s AI tools, exclusively, to create content. Runaway was launched by NYU alumni in 2018. The founders – Cristóbal Valenzuela, Anastasis Germanidis, and Alejandro Matamala – were in the talks for a whopping $4 Billion valuation for their startup earlier this year. Runway also released Gen-3 Alpha, which is trained on both videos and images, and enables Runway’s text-to-video, image-to-video, and text-to-image tools. This model is publicly available to everyone. “The history of art is the history of technology and these new models are part of our continuous efforts to build transformative mediums for artistic and creative expression; the best stories are yet to be told.” said Cristóbal Valenzuela, one of the founders at Runway. Hollywood 2.0 The future of cinema is AI, but not how you might think. Tangibly, AI can be meaningfully employed in films during the pre production and post production processes, and even for generating characters. But in a larger sense, it lowers the barrier of entry and allows many to venture into this otherwise difficult, and legacy driven space. Runway’s very own AI Film Festival, similar to the one in Cannes, is an effort to exclusively celebrate artists who use emerging AI techniques in their short films. Started two years ago, the festival is a bid to open dialogue on the (inevitable) role of AI tools in the film community, and engage with creators from all backgrounds to understand their perspectives. “If we had the kind of deepfake face swapping technology back in the day, a movie like The Irishman would look different than it did, and would not have taken ten years to make. While we were waiting for finances to complete the movie, we were also waiting for a newer technology to come,” said Jane Rosenthal, a leading American film producer, in an interview with Cristóbal Valenzuela. She is also a member of the dean council at NYU Tisch, one of the world’s best film schools, and is enthusiastic about AI’s defining role in today’s zeitgeist, especially newer filmmakers. For context, The Irishman, directed by Martin Scorsese, and produced by Rosenthal, is an interesting case study of how cinematic experience can be improved (when AI is employed responsibly), while also saving a lot of time and resources in the process. “Experimentation with AI is fantastic. Speaking of speed, you know how things work, for instance if an actor was injured or cut themselves during shaving, there is no makeup for it and you have to stop shooting. But now, with AI tools, you can just continue shooting and fix it. You can fix anything,” she added on how AI is not new, but the variety and speed of it is what is new. Apple’s ‘Shot on iPhone’ campaign was a cultural reset, and showed how moving forward with technology could foster artistic expression, and not curtail it. The fulcrum of any movie or piece of art still remains an artist’s vision, and embracing AI in the process is a logical progression in today’s world. Hollywood’s Relationship with AI Startups Hasn’t Always Been Smooth Despite the very visible upside, last year, the Screen Actors Guild and the Writers Guild of America protested against the growing use of AI in the creative space, fearing the loss of jobs and artistic freedom. This protest culminated into certain checks and balances to ensure AI is used as a complement and not replacement for the industry. The wounds from the SacrJo’s controversy with OpenAI still remain fresh in memory. Nonetheless, the general sentiment remains to find ways to collaborate with these AI startups. As more models emerge and the race to AGI seems to be tightening up, it’s likely that AI models will start training themselves, if Hollywood studios don’t come forward and voluntarily collaborate. Lionsgate has a first mover advantage here. As for consumers, these AI tools could usher in a new paradigm of filmmaking by leveling the playing field, allowing low-budget films to compete with high-budget ones without compromising on its quality. No More Parties in LA for Sam Altman? OpenAI’s dominance in this space is still coveted, despite never making Sora, its text-to-video generator, available to the larger public. While privately available to some, it is unknown why OpenAI hasn’t released a ChatGPT-like subscription model of Sora yet. Sora’s capabilities are wild. OpenAI even teased more of Sora, with videos earlier this month, leaving people wanting for more. Sam Altman reportedly attended Hollywood parties in Los Angeles earlier this year, possibly to make inroads before Sora’s official launch. Altman is obsessed with sci-fi movies in the industry, and hopes to see more of it in the future. He has repeatedly spoken about favourite movie, Her, directed by Spike Jonze, and its influence on him. Post the launch of Sora, OpenAI’s lobbying in Hollywood has been quite fruitful, with many studios and artists, including the famous Ashton Kutcher, endorsing the AI video generator. Yet, Runway’s attempt to score a massive deal of this scale, is undoubtedly a notch ahead. Sora’s announcement was one of the first, and most sought after breakthroughs early this year, which led to a bevy of launches by its competitors.  “It’s very easy to copy something once it works, so I think the two ways you can succeed are: say we’re going to be a great fast follower or we’re going to move the frontier,” said Altman in a recent interview hinting at the impending rollouts of OpenAI.","excerpt":"Runway’s very own AI Film Festival, similar to the one in Cannes, is an exclusive celebration of artists who use emerging AI techniques in their short films.","categories":["AI Features"],"tags":["Runway"],"author_name":"Aditi Suresh","publish_date":"2024-09-20T15:08:58","publication_year":"2024","word_count":1021,"keywords":["Go","ChatGPT","OpenAI","AI","GPT","ViT","Runway","llm_models:GPT","R","llm_models:ChatGPT","startup"],"extracted_tech_keywords":["AI","ChatGPT","OpenAI","R","Go","GPT","ViT","startup","llm_models:GPT","llm_models:ChatGPT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/what-runway-and-lionsgate-partnership-tells-us-about-hollywood\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":49350,"title":"Behind The Code: This Data Scientist Says That Coding Workshops Are A Great Way To Start Analytics Career","content":"For our weekly developer column ‘Behind The Code’, we interact with the developer community in India and try to take a look at their journey till date — the way they work, and the tools they use. For this week, we got a chance to interact with Tarun Shrivas, Data Scientist and Instructor at Data Science Dojo (DSD). Shrivas joined Data Science Dojo about a year ago. His work at DSD is distributed across advanced analytics-based consulting projects for external clients, internal user traffic and user behaviour analytics projects, and data science training. His Journey Shrivas is an Electrical Engineering graduate from Jamia Millia Islamia, New Delhi and he has done his Masters in Business Analytics (MSBA) from Seattle University. Before going to the US, he had been working in the marketing research and consulting industry. His work involved data analytics on brand research, consumer research data and drawing insights and recommending a way forward to business clients. “I have worked in different other industries as well including heavy engineering, EPC, and education industries,” said Shrivas. Talking about the challenges in his data science journey, Shrivas said that since he had prior experience in research and consulting, he was somewhat aware of some of the challenges that data scientists might come across. “I would say that the biggest challenge is to communicate the output of your work to the end client where one would often come across a business audience,” said Shrivas. He emphasised that that data scientists should run their model’s output or data visualisation through a business audience within their own organizations. Furthermore, he also that there is one more area that offers challenges and that is identifying which problem to attack first. He believes that this would especially be true if, as a data scientist, one has the responsibility to improve internal metrics on its own organization’s business. Tools And Code When asked about his job at DSD, Shrivas said that data science training is the core business vertical for Data Science Dojo, and that is why most of the time works with all the common or popular machine learning algorithms. However, for the company’s internal analytics work and for the consulting business, he often uses Xgboost, Support Vector Machines, Latent Dirichlet Allocation algorithms. Talking about tools and codes, when Shrivas was working in the brand and consumer research and consulting industry, he used to work with tools such as SPSS, STATA, and the good old MS Excel. However, when he was studying for his data science master’s he started using R, Python, SQL, etc. extensively. When asked which coding language according to him is important and he personally prefers, the DSD data scientist said, “I would say if you are looking for a career in data science you should get familiar with R and Python as most of the data science projects use them.” Apart from that, he also emphasised on query languages such as SQL or PostgreSQL. He also believes that having a working knowledge of distributed engineering platforms such as Hadoop, and Hive would be an added advantage when starting out in the Data Science industry. “Talking about my preference, I like to code in R but honestly there is no clear winner or loser between R and Python,” Shrivas added. “If I have to work on something that is compute-heavy, I also prefer Azure ML studio.” Advice For Aspiring Data Science Professionals According to Shrivas, starting with some free of cost learning resources is always a great option.  Also, you cannot deny the fact that if you Google out you can always find lectures or tutorials on your area of interest. Here are some of the resources that Shrivas strongly recommends: ISLR TextbookStatistical Learning  on YouTubeForecasting: Principles and Practice  Data Science Dojo on YouTubeStatistics How-To Shrivas further said that if someone is not able to follow any regimen and complete the course material on a self-paced learning method, s\/he should definitely give a serious thought of going back to the classroom. He believes that attending a data science boot camp will help a lot. Bootcamp or coding workshop costs low and takes the least amount of time from your work schedule and are considered to be the best way to kick start your journey. “However, you need to be very honest and clear about how much investment, in terms of time as well as money, you are ready to make and what are you expecting out of that investment,” said Shrivas. The Future Direction Being a data scientist and an educator, Shrivas enjoys sharing his knowledge and experience with others through blogs, tutorials or in an actual classroom setting and going forward he wants to continue doing that. Apart from that, he is also looking forward to ways in which he can take data science to a much larger audience.  He said that people have that notion about data science — that it’s a geek’s domain and unless you are a PhD in statistics or a computer science engineer, data science is not your cup of tea. “However,  As a company each one of us at Data Science Dojo truly believes in “Data Science for Everyone” irrespective of current functional expertise,” said Shrivas. Shrivas and the entire DSD team is planning to come up with a platform that reduces the cost of data science learning to the minimum possible. “As long as you have time, we’ll make the content available to you, free of cost. We want to take the data science learning and training to the bottom of the pyramid,” Shrivas said in conclusion.","excerpt":"For our weekly developer column ‘Behind The Code’, we interact with the developer community in India and try to take a look at their journey till date — the way they work, and the tools they use. For this week, we got a chance to interact with Tarun Shrivas, Data Scientist and Instructor at Data […]","categories":["AI Features"],"tags":["Behind The Code","data science ai bootcamp","Data Scientist","Interviews and Discussions"],"author_name":"Harshajit Sarmah","publish_date":"2019-11-04T15:14:52","publication_year":"2019","word_count":932,"keywords":["PostgreSQL","data science","machine learning","TPU","AI","ML","data science ai bootcamp","XGBoost","analytics","Behind The Code","Data Scientist","Azure","Azure ML","Interviews and Discussions"],"extracted_tech_keywords":["AI","machine learning","ML","data science","analytics","Azure ML","XGBoost","Azure","TPU","PostgreSQL"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/behind-the-code-this-data-scientist-says-that-coding-workshops-are-a-great-way-to-start-analytics-career\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10165120,"title":"DeepSeek R1 7B and 14B Distilled Models Available on Microsoft Copilot+ PCs","content":"Microsoft on Monday announced that distilled versions of the DeepSeek-R1 models, the 7 billion and 14 billion parameter variants, will be available on the Copilot+ PCs. The models are available via Azure AI Foundry on Copilot+ PCs powered by Qualcomm Snapdragon X, followed by Intel Core Ultra 200V and AMD Ryzen hardware. “DeepSeek distilled models exemplify how even small pretrained models can shine with enhanced reasoning capabilities and when coupled with the NPUs on Copilot+ PCs, they unlock exciting new opportunities for innovation,” said Microsoft in the announcement. Microsoft Copilot+ PCs are devices capable of running AI models offline, as they’re equipped with a Neural Processing Unit (NPU). An NPU is a dedicated unit on an SoC (System on Chip) that performs all the calculations for AI-related tasks, and leaves CPUs and GPUs to handle other workloads. For manufacturers to ship Copilot+ PCs, Microsoft states a minimum of 16 GB of memory, 256GB SSD and an NPU capable of processing at least 40 TOPS, or trillion operations per second. Users can access all the available variants of the DeepSeek models by downloading the AI Toolkit VS Code extension. Microsoft also said that the DeepSeek-R1 14B model outputs 8 tokens per second, and aims to further optimise the performance. Last month, Microsoft also announced the availability of NPU optimised version of the DeepSeek-R1 1.5B distilled model on the Copilot+ PCs. Manufacturers like HP, Dell, Asus, Acer, and Lenovo are building AI PCs with Microsoft Copilot+ capabilities. Recently, a report from Canalys revealed that AI capable PC shipments reached 15.4 million in Q4 2024. “For the full year 2024, 17% of PCs shipped were AI-capable, with the biggest winners being Apple at 54% share, followed by Lenovo and HP at 12% share each,” read the report. The research firm also added that Windows AI PC shipments grew 26% accounting for 15% of all Windows PCs shipped in Q4 2024. The enterprise sector is expected to be a major driver of AI PC adoption. A report from Gartner suggests that AI PCs will occupy 100% of the enterprise market by 2026. Moreover, the end of support for Windows 10 PCs in October 2025 will likely drive the adoption of newer, more capable AI PCs.","excerpt":"The models are available on the Azure AI Foundry – along with the DeepSeek 1.5B distilled model announced last month.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","AI PC","Microsoft"],"author_name":"Supreeth Koundinya","publish_date":"2025-03-04T19:40:15","publication_year":"2025","word_count":371,"keywords":["Go","TPU","cloud_platforms:Azure","AI","R","innovation","RAG","CuPy","Aim","AI (Artificial Intelligence)","AI PC","Azure","Microsoft"],"extracted_tech_keywords":["AI","Aim","CuPy","RAG","Azure","TPU","R","Go","innovation","cloud_platforms:Azure"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/deepseek-r1-7b-and-14b-distilled-models-available-on-microsoft-copilot-pcs\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":36022,"title":"Now You Can Build Graph Neural Networks With Spektral Based On Keras","content":"Recently, a PhD researcher, Daniele Grattarola built a framework known as Spektral for mapping relational representation learning which is built in Python and is based on the Keras API. Spektral contains a comprehensive set of tools to build graph neural networks as well as implement some of the popular layers for graph Deep Learning. How It Works Spektral is built with semi-supervised deep learning methods for graph data, Graph Neural Network (GNN). According to this paper, the idea of GNN is simple: to encode structural information of the graph, each node v_i can be represented by a low-dimensional state vector s_i , 1 ≤ i ≤ N. The categorisation of  deep learning methods on the graph is shown below The core GNN modules of Spektral are mainly based on Keras and it works with all the different backends offered by Keras in order to promptly start experimenting with the Relational Representation Learning (RRL) which consists of developing models that are able to deal with graphs natively without having to deal with the distracting low-level details. The accessory modules of Spectral are built in Numpy\/Scipy and for which everything should work at the speed of almost-C-like and without compatibility issues. Data Representation Spektral uses a matrix-based representation of manipulating graphs and feeding them to neural networks. This is one of the most commonly used approaches in the theory of GNN and it is also perfect to perform effective computations on GPU. In this framework, some layers and functions are implemented to work on a single graph and while the others are considered as a set of graphs. The framework also distinguishes between three main modes of operations mentioned below Single, where a single graph is considered with its topology and attributes Batch, where a collection of graphs are considered, each with its own topology and attributes Mixed, where a graph with a fixed topology is considered but a collection of different attributes. This can be considered as a particular case of the batch mode (i.e. the case where all adjacency matrices are the same) but is treated separately for computational reasons. Installation Spektral is developed by keeping Python in mind so Python versions 2 and 3 are recommended. Python 3 is heavily recommended and for those who are using Python 2 version are suggested to switch into Python 3 for a smooth workflow. There are some features  of Spectral which require the following optional dependencies RDKit, a library for cheminformatics and molecule manipulation (available through Anaconda); Dyfunconn, a library to build functional connectivity networks (available through PyPi); CDG, a library implementing several change detection algorithms, as well as an engine for Riemannian geometry (available on Github); The framework is tested for Ubuntu 16.04 and 18.04 and should also work on other Linux distros and MacOS. To install the required dependencies on Ubuntu, run the following code $ sudo apt install graphviz libgraphviz-dev libcgraph6 The simplest way to install Spektral is with PyPi; just run: $ pip install spektral To install the framework from source, run the following code in the terminal: $ git clone https:\/\/github.com\/danielegrattarola\/spektral.git $ cd spektral $ python setup.py install  # Or ‘pip install .’ Bottom Line However, this project comes with a little warning for the work which is still in progress and may change before the proper release. Also, if you are interested in contributing your ideas for further development, you can easily do it just by dropping an email.","excerpt":"Recently, a PhD researcher, Daniele Grattarola built a framework known as Spektral for mapping relational representation learning which is built in Python and is based on the Keras API. Spektral contains a comprehensive set of tools to build graph neural networks as well as implement some of the popular layers for graph Deep Learning. How […]","categories":["AI Features"],"tags":["Data Visualisation","Linux distros"],"author_name":"Ambika Choudhury","publish_date":"2019-03-09T07:40:25","publication_year":"2019","word_count":573,"keywords":["Go","NumPy","Linux distros","Keras","AI","neural network","Data Visualisation","Git","Python","deep learning","GitHub","R"],"extracted_tech_keywords":["AI","deep learning","neural network","Keras","NumPy","Python","R","Go","Git","GitHub"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/now-you-can-build-graph-neural-networks-with-spektral-based-on-keras\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":68492,"title":"Oracle Opens Second Cloud Region in Hyderabad","content":"Following the launch of its Mumbai Cloud region in 2019, Oracle has opened a second one in India, this time in Hyderabad. Responding to growing customer demand for enterprise cloud services in India, the company has provided users access to better performance, pricing and security with this development. According to a statement released by the company, this is woven around the company’s global plans to operate 36 second generation Cloud regions by the end of 2020. What is more, this launch propels India into the ranks of the US, Canada, Japan, Australia, South Korea, and the EU in having multiple Oracle Cloud regions that facilitate enterprise-class disaster recovery strategies. “With Oracle opening dual cloud regions in Australia, Japan, Korea and now India, we are further renewing our commitment to support growth in the new decade,” says Garrett Ilg, Executive VP, Japan and Asia Pacific, Oracle. This development comes as large and mid-size organisations across India use Oracle Cloud to run critical workloads, allowing them to leverage better, faster and more efficient services. In fact, some Indian customers are expecting about 3X performance improvements over the next 5 years with Oracle Cloud, ‘unlocking 40% additional cost savings.’ “A large number of Indian organisations are looking to change growth orbits with greater focus on cloud-led innovation,” says Shailender Kumar, Regional MD, Oracle India. “With two Oracle Cloud regions in India, we’re fully geared to support our 15,000 plus customers in their innovation journey, with adequate support by nearly 1,000 specialized Oracle partners,” he adds.","excerpt":"Following the launch of its Mumbai Cloud region in 2019, Oracle has opened a second one in India, this time in Hyderabad. Responding to growing customer demand for enterprise cloud services in India, the company has provided users access to better performance, pricing and security with this development.  According to a statement released by the […]","categories":["AI News"],"tags":["Oracle"],"author_name":"Anu Thomas","publish_date":"2020-06-29T17:42:19","publication_year":"2020","word_count":252,"keywords":["programming_languages:R","AI","innovation","Oracle","RAG","GAN","R"],"extracted_tech_keywords":["AI","RAG","R","GAN","innovation","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/oracle-opens-second-cloud-region-in-hyderabad\/","complexity_score":2,"technical_depth":6,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10001505,"title":"5 IoT Internships In India Candidates Can Apply For","content":"With more and more industries adopting IoT or IoT-based products in their business processes, the demand for IoT professionals in India is going up. The technologies which were once in the sci-fi movies are becoming reality now, thanks to the evolution of IoT. In this article, we list down 5 internship roles that IoT enthusiasts can apply for. 1| Infosys InStep About: Infosys InStep is the global programme for undergraduate, postgraduate as well as for the Ph.D. students which you can apply online. According to this report by the leading newspaper, Infosys InStep has been recognised as the best overall internship programme globally by Vault. Interns with technical backgrounds are offered to work with real-time STEM projects in IoT, AI, Big Data, ML, etc. Once you have applied, your resume will be evaluated and compared with the programme requirements. After a successful evaluation, the candidate will have to attend a telephonic interview and thereafter he\/she will be shortlisted. Duration: 12 weeks Location: Bangalore Stipend: This is a fully paid internship and the stipend is not disclosed Deadline: The internship programme runs throughout the year Click here to apply for the internship 2| Internet of Things (IoT) Teaching Internship At Rankethon About: Rankethon is a team of alumni from IIT Kharagpur, Stanford University, IIFT Delhi and other big firms. During this internship, Rankethon mentors will engage you in the scope of IoT and provide you with in-depth learning on programming networks, software development, hardware design, and internet usage. Here you will learn to transform your ideas or even your dream to a real-life product on your own with an advanced level learning. You will also be working to integrate server to hardware device wirelessly over the internet in the innovation hub with the help of world class cutting edge technology. Duration: 6 months Location: Ludhiana Stipend: ₹20,000 per month + ₹6,000 (incentives) Deadline: 31 March 2019 Click here to apply for the internship. 3| IoT Internship By Ethna Attributes Soft Technologies Private Limited About: The technology ecosystem at Ethna Attributes provides the quickest and most efficient path to deploy connected platforms as well as associated services. As an intern, the day-to-day responsibilities include working with the technology team to develop various IoT solutions, perform coding on Python, Raspberry Pi, Embedded Systems, and Arduino, contribute to GUI development, and other such. Duration: 3 months Location: Chennai Stipend: ₹5,000 per month Deadline: 23 March 2019 Click here to apply for the internship. 4| Internet of Things Internship By Global Technical Services About: Global Technical Services (GTS) is a concept service providing company that specialises in providing technical services to the industry on lubrication. The selected interns will be working on embedded electronics, programming processors like ARM, working on wireless protocols like Zigbee, Lora or BLE along with PCB designing. There are 10 seats available for this internship. Duration: 6 months Location: Mumbai Stipend: ₹5,000 per month Deadline: 22 March 2019 Click here to apply for the internship 5| Nitroware Technologies Pvt. Ltd. About: This is one of the professional software design and development companies situated in Coimbatore. The responsibilities of the intern include developing prototypes using Raspberry Pi as well as developing products using Arduino. Along with IoT, other internship domains include are web development(Java, PHP, Python, .Net), mobile app development on Android, HR (Recruitment) and digital marketing (SEO). The candidate must be a final year or pre-final year student and skills like knowledge of programming language such as C, C++ are required. Duration: 1-6 months Location: Coimbatore Stipend: INR 3000 per month Deadline: 21 March 2019 Click here to apply for the internship.","excerpt":"With more and more industries adopting IoT or IoT-based products in their business processes, the demand for IoT professionals in India is going up. The technologies which were once in the sci-fi movies are becoming reality now, thanks to the evolution of IoT. In this article, we list down 5 internship roles that IoT enthusiasts […]","categories":["AI Features"],"tags":["IoT","Virtual Internship Program"],"author_name":"Ambika Choudhury","publish_date":"2019-03-15T15:22:46","publication_year":"2019","word_count":601,"keywords":["big data","Go","AI","ML","Git","RAG","Python","Virtual Internship Program","C++","R","Java","IoT"],"extracted_tech_keywords":["AI","ML","RAG","Python","R","Go","Java","C++","Git","big data"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/5-iot-internships-in-india-candidates-can-apply-for\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":58115,"title":"7 Ways To Kill Your Data Scientist Career (Without Knowing It)","content":"In the era where data is the most valuable asset for a company, nurturing data skills has to become the topmost priority for any aspiring to mid-level data scientist. The feeling of self-satisfaction with your current skillsets can land you at a miserable spot where your company may choose a candidate who not only has better experience in using new analytics tools but also has a deeper understanding of the latest trends. This, in turn, can result as an end of your career. To solve this problem, we asked multiple data scientists from the AIM Expert Network community, to share their key insight on how to avoid unusual pitfalls and get out of the cocoon to build a bulletproof career. From defining the right focus to analyzing the latest trends, here are the 6 pitfalls that data scientists should avoid to continually grow their career:- Not defining the problem accurately Most data scientists would ignore the “problem” in itself – assuming that this is a business responsibility – to discover or define the “problem” statement clearly. As a manager or a senior\/lead data scientist, this is frequently a responsibility that should be within your purview. Asking the right question is a lot more important than choosing the right algorithm, more often than not. This is a skill often overlooked by budding data scientists. And as we grow in an organization, when you need to scope out the larger data science project, we often stumble in this first step. Hence, it is very important for us to get into the “Questioning the status-quo” mode rather than get carried away by the details of algorithmic efficiency, the goodness of fit parameters or K-cross validations. Sachin Dev, Senior data strategy manager – Accenture Not nurturing the data Data Science is made of 2 words “Data” and “Science”. If you take Data out of this then only science is left and that is what maximum folks are doing. It sounds like taking life away from Data Science itself. People forget to treat data as their child. You need to first spend time with your data, explore the data, get insights from the data and understand the data; later build ML\/AI model as required. If correct insights are not drawn from the data then how can one understand whether the model is behaving correctly? Unknowingly all focus shifts to just building the models and predicting the results. But that is not correct, 80% of work is done if data is treated to its minute details and insights are formulated. So be a Data Scientist and not just a Scientist who has worked on some data. Netali Agrawal, Technology Lead – Infosys Focusing only on coding In my opinion “focusing only on coding” can end a data scientist’s career. Basically, data science is not about using python, R or Tableau; it is about understanding data and being able to make sense out of the data. Coding, as we are already seeing, is becoming automated with the help of tools that are able to generate ML codes. There are certain tools that are capable of building an algorithm based on the list which is provided by its developer. Just by adding a few inputs to the list, the tool will automatically generate the required code and execute it. A data scientist needs to focus on understanding the data, how it is being processed and how it can be modeled to make better decisions. While technical\/programming knowledge could be a good starting point and give one a relatively easy entry in this area but to sustain yourself, you need to invest in the concepts of data integrations, data modeling and data quality along with the understanding of business domain which will help you grow in the area of data science. Just like the models that a data scientist builds have to be trained and improved continually, similarly, to be relevant and thrive in your career, you need to inculcate a habit of continuous learning. Amit Agarwal, Senior Manager (IT) – Nvidia Graphics Ignoring Visualization at the cost of Modeling Visualization is one of the key skills, a data scientist must possess. Decision-makers & leadership teams will be interested in getting quality output in a visual format rather than just a code piece. It is essential for a data scientist to come up with informative graphs\/charts explaining the essence of the work done. Speaking with the Business team helps understand the visualization that can deliver maximum information to end-users. The output must be scalable enough for them to play around and make quality decisions to improve around. Customer Experience is more important than customer support as it ensures the journey is smooth enough.Vijayakeerthi Jayakumar, Data Scientist – Cognizant Technology Solutions Extreme focus on tools and technologies instead of fundamental concepts. I have often seen people holding on to tools rather than the fundamental concepts. This is detrimental to any career, not to mention Data Science. Once you are addicted to a tool or a technique, stagnation is at the horizon. A classic example amongst Data Science\/ML beginners is biased (should be controlled anyway in ML) towards an algorithm over others. However, this is much more evident amongst potential Data Engineers, who prefer sticking to their favorite tools for ETL (Extract, Transform, Load). We need to understand that solving the problem at hand is much more important than using a particular tool. Prasad Kulkarni – Senior Software Engineer Disregarding the latest trends Silently ignoring the latest news and trends in the world can lead to a decline in a data scientist’s career. The world is constantly changing. Data Science is all about problem-solving regardless of the domain. For example, you are working in an FMCG company, and are unaware of instances like economic slowdowns. You have a task in hand to predict the category forecasts for the next year. If you are a sluggish person who is not much into reading news or the latest trends and working on a data model disregarding the external factors i.e. the latest trends. Your predictions will not be aligned with the current trends. This work may also not be appreciated by your company. The same example could also be applied to the ignorance of the latest technologies that come up on a regular basis and the subsequent upskilling ignorant patterns. Repeated cycles of ignorance of the latest trends can be harmful. Ekta Shah, Data Scientist – General Mills Avoiding participation in external and community events We know that in today’s world, technological changes are constantly increasing, therefore it is necessary to update our knowledge regularly in order to prevent being obsolete. Especially when it comes to data science, we can see a fresh set of algorithms every month with their expanding spheres. This can be done in the following ways: a) Reading a lot of research literature. b) Going through a lot of magazines related to AI and its developments. c) Participating in Meetups. d) Trying to solve problems in websites like Machine Hack and Kaggle which keeps updating many contests. Many contributors also contribute different ideas of coding during the process. e) Revising knowledge and trying to share our knowledge in websites such as stack exchange etc., Prasidha Ramnathan, Clinical research associate – Navitas Life sciences This article is presented by AIM Expert Network (AEN), an invite-only thought leadership platform for tech experts. Check your eligibility.","excerpt":"In the era where data is the most valuable asset for a company, nurturing data skills has to become the topmost priority for any aspiring to mid-level data scientist. The feeling of self-satisfaction with your current skillsets can land you at a miserable spot where your company may choose a candidate who not only has […]","categories":["AI Trends"],"tags":["no etl","trends in bi and data visualization"],"author_name":"AIM Media House","publish_date":"2020-03-06T10:00:00","publication_year":"2020","word_count":1234,"keywords":["trends in bi and data visualization","data science","Go","TPU","AI","ML","Scala","no etl","Python","Aim","analytics","R"],"extracted_tech_keywords":["AI","ML","data science","analytics","Aim","TPU","Python","R","Go","Scala"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/6-ways-to-kill-your-data-scientist-career-without-knowing-it\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":57603,"title":"Happy Employees Create Successful Companies. How Can AI Help?","content":"It is commonly acknowledged that AI is leveraged by companies to engage their customers, but little is accomplished by way of technology to keep their own employees satisfied. This is slowly changing. More and more organisations are exploring the potential of AI in keeping their workforce more productive – by keeping them happier. Talking about how technologies like AI can deliver multiple benefits to employees in their day-to-day work, Shivani Rai Gupta of Capgemini addressed a gathering of over 250 people at Rising 2019. Gupta, who heads AI and data science in financial services at the tech major, spoke at length about AI’s ability to perform both high-value and mundane tasks for employees with greater levels of effectiveness. Why Chase Employee Satisfaction? Companies strongly feel that happy employees create high business value for them and hence, help in influencing revenues and generating overall success. According to a report released by Deloitte, only 23% of companies believe that their employees are aligned with their goals. This indicates that a massive chunk is not happy with the work they are doing, and are thus, not being productive at work. According to Gupta, one way of cutting through this slack is by reducing technological difficulties faced by employees. This includes operational challenges as well as software errors – both of which should be minimised. Chatbots are increasingly being channelised by companies to solve these internal issues. This eliminates the need to chase people since the technology already handles the problem for you. Another common approach is tech-enabled search engine machinery for simple use cases like policy information, insurance details etc. But this is just skimming the surface. Can repetitive and mundane tasks of employees be automated? AI Minimising Cumbersome Tasks Employees, especially those who travel on a routine basis, file their reimbursement claims regularly. A technique to automate that end-to-end process, which saves you time and energy is long overdue. It will also enable employees to use that time in a more productive manner. HR departments are constantly looking to AI for creative ideas. Transitioning to soft copies and going paperless was a start. But there is space for more innovation here. Automating employee onboarding process in one. Every time people switch jobs, they spend about 3-4 hours on an average filling out forms documenting their educational background, previous company details, salary certificates, and bank statements, among others. These forms are processed and validated by a team, who may reach out to a third party for background verification, and\/or compare it with the information provided in the resume. This entire process can be automated and made seamless without any human intervention. This is one of the biggest applications, where the industry is leveraging AI to ultimately satisfy its employees. With this, they only need to take pictures of their documents, upload it onto a portal and the rest will be taken care of by the AI engine. It also solves the problem of reading information across a wide range of formats. For instance, university mark sheets come in different forms and sizes, and even languages. Another example is salary slips – no two companies share the same structural format of this document. But AI can be trained to go around these differences and still extract the required information. Identifying Right Learning Programs With reskilling being the need of the hour, it is a good practice for employees to take advantage of learning programs facilitated by their companies. This is typically done in two ways – either employee identifies what fits well with their career growth, or companies give them a list of training programs, which is more from a demand perspective. For instance, if you are a SaaS expert, but there is a demand for python, in all likelihood, companies will send everyone to python training, whether they are interested or not. But this is slowly changing. Today, companies have started evaluating both demand and employee personalities when allocating such programs. And AI has enabled this. Using AI, employees’ learning curve can be traced and analysed to give insights on what the best programs would be for them. For instance, if you are an ETL expert, and you want to get into the data science field, you can now understand the different steps you may have to take to conquer that path. Additionally, you will also get information around how people with a similar background have succeeded in this journey and what their learning path was like.","excerpt":"It is commonly acknowledged that AI is leveraged by companies to engage their customers, but little is accomplished by way of technology to keep their own employees satisfied. This is slowly changing. More and more organisations are exploring the potential of AI in keeping their workforce more productive – by keeping them happier.  Talking about […]","categories":["AI Features"],"tags":["ai certificates","AI Companies"],"author_name":"Anu Thomas","publish_date":"2020-02-27T12:00:00","publication_year":"2020","word_count":744,"keywords":["ai certificates","data science","Go","AI","chatbots","ETL","ML","RAG","Python","Aim","AI Companies","R"],"extracted_tech_keywords":["AI","ML","data science","Aim","RAG","chatbots","Python","R","Go","ETL"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/happy-employees-create-successful-companies-how-can-ai-help\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":13885,"title":"DDoS attacks are evolving exponentially and Big Data can come to rescue","content":"What was once in a blue moon phenomenon for systems, and a minor irritation in the functioning has now become an existential threat – DDoS attacks! As you might know, a distributed denial of service (DDoS) is an intentional attack on a system by inducing meaningless traffic through data requests by installing viruses or malware on the application hosting environment. DDoS attacks severe in nature may hurt the reputation (and thus the revenue) of the enterprise as they lead to extended downtimes affecting the availability and performance of the applications. The only way to counter DDoS attacks is through timely detection and even timelier response from the NetOps and SecOps teams. The latest DDoS attacks are so ingeniously executed that they have gone from Mbps to Gbps to a mind-boggling Tbps speed in a short duration of time. It is safe to say that the engineers of such attacks have upgraded themselves to cloud level before the enterprise data security can transition from scale-up architecture to scale-out architecture (cloud). Why present day DDoS protection techniques are falling shot? To determine the features that need to go into the making of a robust and fast DDoS protection system, it is necessary to evaluate the factors influencing the exponential growth of DDoS attacks. The hackers are having a field day making DDoS attacks systems because: The arena of DDoS attacks has expanded from being a pastime for individual hackers to a global market with rival brands and other entities spending money to sabotage enterprises. Apart from organizations, any gamer or activist can also bring about an attack for a few bitcoins. No domain in the enterprise is spared from DDoS attacks. The tools for DDoS attacks are open-sourced. E.g. Lizard stressor, Mirai. The hackers are working at constantly improving the DDoS attack technologies using latest techniques like AGILE. They are also using DDoS attacks as distractions or ‘baits’ to target deeper levels of applications. As the enterprises are not prepared to handle big data generated by IoT, the DDoS attacks are targeting these cloud systems at unfathomable speeds of tera-bytes. On the other hand, enterprises are inching at a snail pace towards improving DDoS attacks, with most being stuck in the pre-cloud era. Let’s take a rational look at where enterprises are falling short: The most used counter measure is to scan, detect, and evaluate all traffic through a legacy solution that is ridiculously expensive. Only some enterprises can afford these inline solutions. An alternative that is being adopted nowadays is to have the traffic monitored through out of the band Linux-based applications. This approach implemented in the scale-up environments fails largely because of limited compute and memory capacities. Scanning all traffic is a static policy that is either overly broad. Identifying security threats requires a manual intervention after continuous monitoring of scanned traffic. The resulting analysis is a mixture of false negatives and false positives with very few hits. This slow reactive waterfall process keeps the resources busy ‘reacting’ rather than ‘planning ahead’. As the scale-up DDoS prevention solutions are not equipped to deal with big data, there is minimal logging and only a few summary reports are generated. Although these solutions share space with other network monitoring tools, they are heavily soiled making them vulnerable to the same flaws shared by the hosting and storage environment. This hampers their speed, efficiency, and memory making it impossible to track and analyze attack history. What makes a robust DDoS security solution? The need of the hour for DDoS security solutions is to make the switch to big data practices to remove the storage and memory constraints imposed by the legacy scale-up systems. If some basic changes (which are long overdue) are implemented, there will be a tremendous improvement in the DDoS attack handling capabilities of enterprise applications. Let’s take a look at them: Big data Analytics systems can maintain a log of malicious and suspicious IPs that has initiated DDoS attacks in the past. This makes it possible to have adaptive base lining of such IPs leading to increased accuracy in anomaly detection. Big data Analytics systems can handle huge chunks of data thus making it possible to have complete logs of raw data, which can be analyzed to derive exploratory analytics that help enterprises stay one step ahead of the DDoS attackers. These solutions are cloud-based and operated through APIs allowing for a specific response to a DDoS attack, rather than static policy implementation. API-based solutions can also collaborate with multiple vendors and low-cost mitigation systems. They provide unified visibility in that they can provide insights into a lot of other information in addition to the DDoS attack data. They can help in monitoring traffic flow, network performance, routing data, and device\/interface data. To conclude Enterprises are lagging in developing advanced countermeasures for the ever evolving DDoS attacks that are already making use of cloud-based technologies. Big data can help bridge this gap, by removing the limitations imposed by legacy security solutions. It is advisable to contact a big data service provider which is based DDoS protection platform that is available as a SaaS, rather than developing an in-house custom solution to reduce cost overheads and receive the best-in-class security. About The Author: Aaron Jacobson, Application Developer at Technoligent – a big data service providing company. Aaron has the knowledge of App and web development and he can work with big data analytics very well. Technoligent has a team of application developer and Aaron has the responsibility of a leader. Aaron has the M.S. Degree in Computer Science. Contact me on technoligent@nexcorp.in.","excerpt":"What was once in a blue moon phenomenon for systems, and a minor irritation in the functioning has now become an existential threat – DDoS attacks! As you might know, a distributed denial of service (DDoS) is an intentional attack on a system by inducing meaningless traffic through data requests by installing viruses or malware […]","categories":["IT Services"],"tags":[],"author_name":"AIM Media House","publish_date":"2017-03-29T04:51:42","publication_year":"2017","word_count":931,"keywords":["big data","Go","API","AWS","AI","RAG","anomaly detection","analytics","GAN","R"],"extracted_tech_keywords":["AI","analytics","RAG","anomaly detection","AWS","R","Go","API","big data","GAN"],"url":"https:\/\/analyticsindiamag.com\/it-services\/ddos-attacks-are-evolving-exponentially-and-big-data-can-come-to-rescue\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10171951,"title":"Cursor&#8217;s New $200 Ultra Plan: Is There a Catch?","content":"Cursor, the AI coding tool, has announced a $200\/month ‘Ultra’ plan designed for power users, previously restricted by unpredictable usage caps. The plan promises 20x more usage than the Pro tier. The Ultra tier has been made possible by the company’s “multi-year partnerships” with companies like OpenAI, Anthropic, Google, and xAI. Alongside the Ultra rollout, Cursor has also overhauled its Pro plan. It now defaults to an “unlimited-with-rate-limits” model, replacing the legacy 500-request allowance. Cursor described the shift as a response to user demand. As the developer of Cursor stated on Reddit, “This change was highly requested by power users seeking more predictability than usage-based pricing would offer.” However, some long-time users aren’t celebrating. Early reactions point to confusion around what’s actually being offered and what’s being taken away. In the absence of clear documentation on actual limits, developers are finding the new changes puzzling. Rate Limits and a Redefined ‘Pro’ Cursor’s documentation clarifies that rate limits are based on compute usage, not request count. Each request’s cost depends on the model selected, the message and file lengths, and the current session state. “Rate limits reset every few hours,” Cursor notes, but there are two separate types, burst limits and local limits, each with its own refill frequency. A Cursor developer posted in the community forum, “We’re able to offer limits high enough that very few people should ever see any rate limits to begin with.” However, they also added that “using more expensive models will get you close to that rate limit faster.” Cursor is still allowing users to stick with the old Pro model for now. “You’re free to stick with legacy Pro Plan if you’d like!” the documentation reads. The Ultra plan, meanwhile, offers far more breathing room. But it comes at a steep cost, and with many of the same caveats around opaque compute budgeting. Cursor’s promise of “predictable pricing” is only as strong as its transparency, something critics say is still lacking. Toggling Trouble and User Backlash For those who’ve already moved to the new Pro model, the transition hasn’t been smooth. A Reddit user shared a story of unexpected limits: “I was automatically added to the new plan (rate-limits), and my 350 premium requests were gone in 20 Claude 4 prompts.” The user reports exceeding their new request limit (660\/500) and being forced into a slow mode. They express frustration at this situation and highlight their inability to revert to the previous plan. Cursor acknowledges that toggling between plans is still being refined. One developer responded on the forum, saying, “We’re still working out the final details,” and confirmed that new users will eventually be moved entirely to the updated plan. For now, older users should be able to switch between models for some time. Some users are also questioning the selective generosity in the Pro changes. One community member in the forum interpreted this option less charitably, and said, “I think this is a downgrade, and a hidden one at that.” The member highlighted that it seems counterintuitive to retain the option for the 500-request plan, as the new changes effectively diminish functionality, likely leading to user dissatisfaction. “I don’t believe in generous offers for Pro users,” wrote the forum member. “Especially when cheap models like DeepSeek are being limited to 60k context per request.” Cursor’s approach may well be the start of a more structured future for AI-powered dev tools. However, transparency remains non-negotiable for users who value predictability. The company’s bet on flexible rate-limiting hinges on how well it communicates and enforces its invisible ceilings. For now, the question isn’t whether Ultra is worth $200; it’s whether Pro still means what it used to.","excerpt":"“I don’t believe in generous offers for Pro users.”","categories":["AI Features"],"tags":["AI coding","cursor"],"author_name":"Ankush Das","publish_date":"2025-06-18T18:30:00","publication_year":"2025","word_count":613,"keywords":["Anthropic","cursor","AI coding","Go","OpenAI","AI","R","Claude 4","XAI","Rust","xAI"],"extracted_tech_keywords":["AI","OpenAI","Claude 4","Anthropic","xAI","R","Go","Rust","XAI","AI coding"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/cursors-new-200-ultra-plan-is-there-a-catch\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10171405,"title":"Is Nuclear Energy the Future of AI-Powered Data Centres?","content":"As data centres increase in size and number, their projected energy consumption will likely skyrocket, which could place immense pressure on existing energy resources and infrastructure. To deal with the growing demands, nuclear energy is currently the favourite energy source for tech companies. Recently, on June 3, Meta announced a 20-year corporate nuclear energy agreement with Constellation Energy for the Clinton Clean Energy Centre to support next-generation and advanced energy technologies, including geothermal and nuclear. Data centres are the foundation for driving AI innovations, housing the infrastructure that brings these technologies to life. Meta has stated that it prioritises efficient operations, ensuring that its electricity consumption matches 100% clean and renewable energy while actively exploring emerging technologies. In preparing for the future energy demands driven by AI development, Meta has recognised the value of nuclear power in providing reliable and stable electricity. According to the company, the energy generated from this project will support Meta’s operations in Southern Illinois. Moreover, the company continues to advance its previously announced nuclear request for proposals (RFP) process. It is now in final discussions with a shortlist of potential projects aimed at meeting its target of securing between one and four gigawatts of energy capacity. Illinois provides subsidies for Constellation Energy’s nuclear facility, the Clinton Clean Energy Centre, through a programme funded by ratepayers that issues zero-emission credits for electricity generation nearly free of carbon emissions, as reported by the Guardian. This program will end in 2027. At this point, Meta’s power purchase agreement will financially support the plant for relicensing and operational costs, although the exact amount remains unspecified. The agreement permits Constellation to increase Clinton’s capacity, which currently stands at 1,121 MW, by an additional 30 MW. This facility generates enough electricity to supply about 8 lakh households in the US. Clinton commenced operations in 1987, and last year, Constellation submitted a request to the US Nuclear Regulatory Commission to extend its license until 2047. Meta said in its statement that numerous nuclear power facilities in the US require ongoing assistance to ensure that our electricity grids stay dependable as energy demands increase. Maintaining the operation of an existing plant will have a similar beneficial impact as introducing new clean energy to the grid and will prevent the disturbances that have arisen when other nuclear units have closed earlier than planned. In a related case, a nuclear reactor at the Three Mile Island facility in Pennsylvania was set to be activated for the first time in five years. This reportedly took place after its owners, Constellation Energy, reached an agreement in September 2024 to supply electricity for Microsoft’s expanding artificial intelligence initiatives. Similarly, Google signed the world’s first corporate deal to purchase nuclear energy from small modular reactors (SMRs) developed by California’s Kairos Power in October 2024. This initiative aims to bring the first SMR online by 2030, with additional reactors by 2035. The agreement will provide up to 500 MW of new, 24\/7 carbon-free power to US electricity grids, helping communities access clean and affordable nuclear energy. According to Schneider, an analyst in digital infrastructure at Goldman Sachs Research, although renewable energy can satisfy a significant portion of the increased electricity requirements of data centres during specific times of the day, it does not provide a reliable enough power supply to serve as the sole energy source for these facilities, the report said. “Our conversations with renewable developers indicate that wind and solar could serve roughly 80% of a data centre’s power demand if paired with storage, but some sort of baseload generation is needed to meet the 24\/7 demand,” a Goldman Sachs report stated. It further mentioned that nuclear power is the preferred choice for a steady power supply, but the challenges associated with constructing new nuclear facilities make natural gas and renewables more feasible short-term alternatives. Data centres, vital infrastructures that store large volumes of data and support crucial services, encounter various cybersecurity issues. As per reports, these issues involve defending against advanced attacks, adhering to regulatory requirements, handling intricate environments, and reducing risks associated with human factors. Similarly, nuclear facilities present significant risks to the environment. The Three Mile Island facility was located where the region saw the worst nuclear meltdown and radiation leak in US history in March 1979, when a malfunctioning valve led to a loss of water coolant, causing the Unit 2 reactor to overheat. Even after over forty years, the reactor remains in a decommissioning stage. Despite this, Joseph Dominguez, president and CEO of Constellation, told Guardian, “Before it was prematurely shuttered due to poor economics, this plant was among the safest and most reliable nuclear plants on the grid, and we look forward to bringing it back with a new name and a renewed mission.” However, recent contracts for nuclear facilities and a growing interest in nuclear power indicate significant investment growth in the next five years, leading to increased energy supply in the 2030s. The emergence of AI data centres is boosting investor confidence in electricity demand, prompting tech companies to seek reliable, low-carbon energy. This is resulting in the revival of retired nuclear plants and plans for new reactors, the Sachs report added. In the US, major tech firms have signed contracts for over 10 GW of potential new nuclear capacity in the past year, with Goldman Sachs report predicting that three plants could be operational by 2030.","excerpt":"Meta has stated that it prioritises efficient operations, ensuring that its electricity consumption matches 100% clean and renewable energy while actively exploring emerging technologies.","categories":["AI Features"],"tags":["AI Data Center","nuclear energy"],"author_name":"Smruthi Nadig","publish_date":"2025-06-05T16:08:34","publication_year":"2025","word_count":897,"keywords":["AI Data Center","Go","artificial intelligence","ELT","AI","innovation","Git","RAG","Aim","ViT","nuclear energy","R"],"extracted_tech_keywords":["AI","artificial intelligence","Aim","RAG","R","Go","Git","ELT","ViT","innovation"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/is-nuclear-energy-the-future-of-ai-powered-data-centres\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10139845,"title":"Ethicist Warned of Character AI-like Mishaps Last Year","content":"Megan Garcia, mother of a 14-year-old in Florida, has sued chatbot startup Character AI for allegedly aiding her son’s suicide. Garcia claims her son, Sewell Setzer III, got addicted to the company’s service and was deeply attached to a chatbot it created. Setzer has spent months talking to a Character AI chatbot named Daenerys Targaryen, a screen personality from the popular show Game of Thrones. In a lawsuit filed at the Orlando, Florida federal court, Garcia claims her son formed an emotional relationship with the chatbot, which pushed her son to do the unimaginable. Setzer, who died by a self-inflicted gunshot wound to his head in February this year, was talking to the chatbot on that particular day. He even told the chatbot, “What if I told you I could come home right now?” to which the chatbot replied, “Please do my sweet king”. While Sewell’s move has been devastating for the family, an ethicist did warn us last year something like this could happen. Giada Pistilli, principal ethicist at Hugging Face, an open-source hosting platform, told AIM, “As I’ve consistently pointed out, distributing a “magic box” or a complex, opaque system to a wide audience is fraught with risks. The unpredictability of human behaviour, combined with the vast potential of AI, makes it nearly impossible to foresee every possible misuse.” Who is to Blame? Garcia has taken Character AI to court claiming the chatbot instigated her son to take the drastic step. In the lawsuit, she said that the California-based company was aware of the risk posed by its AI to minors but did not take the necessary steps to redesign it to reduce those risks or provide sufficient warnings about the potential dangers associated with its use. It is also unlikely that Setzer did not know he was chatting with an AI system. Moreover, a disclaimer on the chat does remind users that they are talking to an AI and the responses are not from a real person. Despite the guardrails in place, Setzer did develop an emotional attachment to the chatbot. Amidst this development, Character AI expressed its condolences to the family in a social media post and indicated that it has implemented measures to prevent a recurrence of this issue. “Recently, we have put in place a pop-up resource that is triggered when the user inputs certain phrases related to self-harm or suicide and directs the user to the National Suicide Prevention Lifeline,” the company said in a blog post. They are also introducing new safety features, which include measures for minors to limit exposure to sensitive content, improved detection of guideline violations, a revised disclaimer that AI is not a real person, and notifications after an hour of use. Nonetheless, Garcia’s taking Character AI to court does raise a moot question: Who is to blame, the AI system or its developers? Last year, when AIM wrote a story on ‘Who should be blamed for AI mishap?’, Pistilli said, “I believe that responsibility in the realm of AI should be a shared endeavour. However, the lion’s share of both moral and legal accountability should rest on the shoulders of AI system developers.” At the end of the day, Character AI is a for-profit business that exists to generate significant revenue by shipping its AI product to as many users as possible. In today’s capitalistic landscape, this raises the question: Are companies doing their absolute best to ensure the safety of these AI systems? Is responsible development a central priority for them, or are they primarily focused on the quickest way to generate revenue? Back then, Annette Vee, an associate professor at the University of Pittsburgh, pointed out that the race to release generative AI means that models will probably be less tested when they are released. Like Pistilli and Vee, many experts have also warned us about the dangers of shipping AI products to consumers without fully understanding what the consequences could be. Moreover, with the technology still evolving, there are no clear regulations yet determining how consumers ‘should’ or ‘should not’ use these AI systems. Building AI Responsibly Although Setzer’s drastic action garnered significant attention, it wasn’t the first incident of its kind. Last year, local media reported that a man in Belgium died by suicide following interactions with an AI chatbot. Moreover, in 2021, Jaswant Singh Chail, a 21-year-old man from England, broke into Windsor Castle with a loaded crossbow to assassinate Queen Elizabeth II. The court hearing later revealed that he was asked by an AI chatbot to do so. Both Character AI and Replika AI have a combined user base of over 40 million active users, and such companies have recorded hundreds of millions of users so far. Hence, safeguarding its users should be at the top of its priority list. Interestingly, Pistilli also pointed this out last year, and it still holds true today. “I think that we should better frame these conversational agents, and their developers should design them not to let them converse with us about sensitive topics (e.g., mental health, personal relationships, etc.), at least not until we find suitable technical and social measures to contain the problem of anthropomorphisation and its risks and harms.” However, it’s only now, after Garcia’s lawsuit, that Character AI claimed they are including measures to limit exposure to sensitive content; however, these measures are limited to minors only. It would not be entirely right either to expect these companies to shut down their service until they have ensured the safety of their users. But in the absence of any regulation, what can be done is to hold them accountable to ensure maximum safety measures are in place. “It’s imperative for developers to not only create responsible AI but also ensure that its users are well-equipped with the knowledge and tools to use it responsibly,” Pistilli said.","excerpt":"Amidst this development, Character AI expressed its condolences to the family in a social media post and indicated that it has implemented measures to prevent a recurrence of this issue.","categories":["AI Features"],"tags":["AI Character","Mishaps","Responsible AI"],"author_name":"Pritam Bordoloi","publish_date":"2024-11-02T10:00:00","publication_year":"2024","word_count":974,"keywords":["Go","Hugging Face","API","conversational agents","AWS","AI","Responsible AI","Aim","generative AI","AI Character","GAN","Mishaps","R"],"extracted_tech_keywords":["AI","generative AI","conversational agents","Aim","Hugging Face","AWS","R","Go","API","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/ethicist-warned-of-character-ai-like-mishaps-last-year\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10008293,"title":"Adobe Introduces AI-Powered Sky-Replacement Tool To Photoshop","content":"Adobe is introducing an AI-powered sky replacement tool to Photoshop that helps the user to dramatically enhance the sky in a picture with just a few clicks. Way ahead of its Adobe Max conference, the company showcased only a preview of the tool on YouTube. According to the video, the tool, added on Photoshop, leverages machine learning to automatically detect the foreground of the image and separate it from the background of a picture. This, in turn, helps people in masking saving a lot of time in creating complicated masks separately for the foreground and the background of the image. Traditionally, manually replacing the sky in Photoshop was a tedious task and many times not so efficient. To make this work, users have to load a number of presets of dramatic skies, and the ML algorithms will automatically tweak the temperature, warm or cold, of the foreground to match the new sky. For instance, the user chooses to add in a warm, golden sunset, for example, the advanced technology will automatically update the colouring and toning of the rest of your photograph to match. While Adobe has been teasing with this technology, this is the first sky-replacement tool. Luminar, in fact, has offered a similar sky-replacement tool last year with similar workflow and application. However, this new adobe tool has claimed to be also enhancing the image with the help of artificial intelligence and machine learning. This tool will be an addition to Adobe’s Sensei technology’s other extraordinary features, starting from the ability to smoothening the blemishes to making object selection easier. According to the media, this tool could be a big help to Instagram influencers who are always willing to make their travel and sky photos beautiful. According to Adobe product manager Meredith Stotzne — Adobe’s AI-powered machine learning models would do the heavy lifting of masking and blending, and they use cutting edge algorithms to harmonise the foreground of your image with the sky. Adobe hasn’t released much information about this tool, but the industry is hoping to learn more about this AI-powered tool in its Adobe Max 2020 conference on 20th October.","excerpt":"Adobe is introducing an AI-powered sky replacement tool to Photoshop that helps the user to dramatically enhance the sky in a picture with just a few clicks. Way ahead of its Adobe Max conference, the company showcased only a preview of the tool on YouTube. According to the video, the tool, added on Photoshop, leverages […]","categories":["AI News"],"tags":["Adobe","Adobe AI","Adobe Sensei","machine learning blending","photoshop"],"author_name":"Sejuti Das","publish_date":"2020-09-25T12:31:14","publication_year":"2020","word_count":355,"keywords":["Go","artificial intelligence","machine learning","programming_languages:R","Adobe","AI","ML","Adobe Sensei","machine learning blending","programming_languages:Go","Adobe AI","RAG","Aim","photoshop","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","Aim","RAG","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/adobe-introduces-ai-powered-sky-replacement-tool-to-photoshop\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":39928,"title":"Here’s What TensorFlow Graphics Library Has In Store For Unsupervised Computer Vision Tasks","content":"via TensorFlow blog Abstraction is a common trait amongst the now widely used machine learning libraries or frameworks. Dusting off the nitty-gritty details under the rug and concentrating on implementing algorithms with more ease is what any data scientist would like to get their hands on. TensorFlow rose into prominence for the very same reason — abstraction. Now with its latest library TensorFlow Graphics, it aims to address key computer vision challenges by incorporating the knowledge from graphics in the images, which in turn result in robust neural network architectures. TensorFlow’s machine learning platform has a comprehensive, flexible ecosystem of tools, libraries and community resources that lets researchers push the state-of-the-art in ML and developers easily build and deploy ML powered applications. During the last couple of years, neural network architectures have taken giant strides into a handful of domains by both contributing to and imbibing from the field it enters. In case of graphics, there has been a rise in novel differentiable graphics layers which can be inserted in neural network architectures. Neural network architectures can be made more efficient by leveraging the knowledge acquired from computer vision and graphics say the researchers at Google. “Training machine learning systems capable of solving these complex 3D vision tasks most often require large quantities of data. As labelling data is a costly and complex process, it is important to have mechanisms to design machine learning models that can comprehend the three-dimensional world while being trained without much supervision. Combining computer vision and computer graphics techniques provides a unique opportunity to leverage the vast amounts of readily available unlabelled data,” wrote the team behind TensorFlow graphics. What Does TensorFlow Graphics Have To Offer The following are one of  the few functionalities of the new library  that TensorFlow boasts of: Transformations Object transformations control the position of objects in space. In the illustration below, the axis-angle formalism is used to rotate a cube. Modelling Cameras Camera models greatly influence the appearance of three-dimensional objects projected onto the image plane. As can be observed below, the cube appears to be scaling up and down, while in reality, the changes are only due to changes in focal length. Materials Now, TensorFlow Graphics allows users to drop virtual furniture in their environment and have the pieces photo-realistically blend with their interior with this new feature. TensorBoard 3D Visual debugging is a great way to assess whether an experiment is going in the right direction. To this end, TensorFlow Graphics comes with a TensorBoard plugin to interactively visualize 3D meshes and point clouds. Geometry — 3D convolutions and pooling TensorFlow Graphics comes with two 3D convolution layers, and one 3D pooling layer, allowing, for instance, the training of networks to perform semantic part classification on meshes. Installing TensorFlow graphics: pip install tensorflow_graphics With the experiments such as Tossing Bot and depth maps showing promising results, the release of TensorFlow graphics couldn’t have come at a better time. From 3D reconstruction to adding video effects like synthetic defocus,as the domain of computer vision extends it reach, collecting every innovation into its arsenal, the future looks bright for AGI. You can experiment with the library here.","excerpt":"Abstraction is a common trait amongst the now widely used machine learning libraries or frameworks. Dusting off the nitty-gritty details under the rug and concentrating on implementing algorithms with more ease is what any data scientist would like to get their hands on. TensorFlow rose into prominence for the very same reason — abstraction. Now […]","categories":["Deep Tech"],"tags":["Computer Vision","TensorFlow library"],"author_name":"Ram Sagar","publish_date":"2019-05-30T12:01:20","publication_year":"2019","word_count":527,"keywords":["Go","machine learning","AI","neural network","ML","computer vision","RAG","Aim","Computer Vision","TensorFlow","R","TensorFlow library"],"extracted_tech_keywords":["AI","machine learning","ML","neural network","computer vision","Aim","TensorFlow","RAG","R","Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/heres-what-tensorflow-graphics-library-unsupervised-computer-vision-tasks\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":30407,"title":"BlackBerry To Acquire Cybersecurity Firm Cylance, Will Use New AI For Their Spark EoT Platform","content":"BlackBerry Limited this week today announced that it has entered into a definitive agreement to wholly acquire Cylance, an artificial intelligence and cybersecurity company, for $1.4 billion in cash. Reportedly, the smartphone maker will also assume the unvested employee incentive awards. Pending regulatory approvals and other customary closing conditions, the deal is expected to close before to the end of BlackBerry’s current fiscal year (February 2019), said the company in an official statement. John Chen, executive chairman and CEO at BlackBerry, said, “Cylance’s leadership in artificial intelligence and cybersecurity will immediately complement our entire portfolio, UEM and QNX in particular. We are very excited to onboard their team and leverage our newly combined expertise. We believe adding Cylance’s capabilities to our trusted advantages in privacy, secure mobility, and embedded systems will make BlackBerry Spark indispensable to realising the Enterprise of Things.” Studies have showcased that security is one of the top barriers to IoT success (35%), with privacy concerns (25%), and potential risks and liabilities (25%) also in the top five. Cylance is a pioneer in applying AI, algorithmic science, and machine learning to cybersecurity software that has proven highly effective at predicting and preventing known and unknown threats to fixed endpoints. The fast-growing company has become a market leader with an advanced and lightweight agent that resides on the endpoint, operates both online and off, and requires a minimum of memory and power to function. Founded in 2012, Cylance generates highly recurring revenue from over 3,500 active enterprise customers, including more than 20% of the Fortune 500. Stuart McClure, Co-Founder, Chairman, and CEO of Cylance, said in a statement, “Our highly skilled cybersecurity workforce and market leadership in next-generation endpoint solutions will be a perfect fit within BlackBerry where our customers, teams and technologies will gain immediate benefits from BlackBerry’s global reach. We are eager to leverage BlackBerry’s mobility and security strengths to adapt our advanced AI technology to deliver a single platform.”","excerpt":"BlackBerry Limited this week today announced that it has entered into a definitive agreement to wholly acquire Cylance, an artificial intelligence and cybersecurity company, for $1.4 billion in cash. Reportedly, the smartphone maker will also assume the unvested employee incentive awards. Pending regulatory approvals and other customary closing conditions, the deal is expected to close before to […]","categories":["AI News"],"tags":["AI (Artificial Intelligence)","Cybersecurity"],"author_name":"Prajakta Hebbar","publish_date":"2018-11-20T04:54:37","publication_year":"2018","word_count":324,"keywords":["Go","machine learning","artificial intelligence","programming_languages:R","AI","programming_languages:Go","RAG","Rust","Cybersecurity","R","programming_languages:Rust","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","RAG","R","Go","Rust","programming_languages:R","programming_languages:Go","programming_languages:Rust"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/blackberry-to-acquire-cybersecurity-firm-cylance-will-use-new-ai-for-their-spark-eot-platform\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10105283,"title":"Apple Optimises LLMs for Edge Use Cases","content":"Apple has published a paper ‘LLM in a flash: Efficient Large Language Model Inference with Limited Memory’ outlining a method for running LLMs on devices that surpass the available DRAM capacity. This involves storing the model parameters on flash memory and bringing them on demand to DRAM. https:\/\/twitter.com\/1littlecoder\/status\/1737353316634374312 Their method involves constructing an inference cost model that aligns with the behavior of flash memory, guiding optimization efforts in two crucial areas: reducing the volume of data transferred from flash and reading data in larger, more contiguous chunks Within this flash memory-informed framework, Apple employs two principal techniques. First, “windowing” strategically reduces data transfer by reusing previously activated neurons, and second, “row-column bundling,” tailored to the sequential data access strengths of flash memory, increases the size of data chunks read from flash memory. These methods collectively enable running models up to twice the size of the available DRAM, with a 4-5x and 20-25x increase in inference speed compared to naive loading approaches in CPU and GPU, respectively. This research is significant as Apple plans to integrate generative AI capabilities into iOS 18. The new OS will leverage generative AI technology to enhance Siri and the Messages app, enabling them to answer questions and auto-complete sentences more effectively. Apple is also exploring the potential use of generative AI in apps such as Apple Music, Pages, Keynote, and Xcode. Apart from Apple, Samsung recently introduced Gauss, its own on-device LLM. According to reports, Gauss is set to be incorporated into the upcoming Galaxy S24 smartphone, slated for release in early 2024. The company intends to integrate this language model into its devices such as phones, laptops, and tablets to enhance the capabilities of its smart devices. Moreover, Google has announced its on-device LLM, called Gemini Nano, which is set to be introduced in the upcoming Google Pixel 8 phones, offering capabilities such as Summarize in the Recorder app and Smart Reply in Gboard.","excerpt":"Apple tackles the challenge of efficiently running LLMs that exceed the available DRAM capacity","categories":["AI News"],"tags":[],"author_name":"Siddharth Jindal","publish_date":"2023-12-20T15:43:01","publication_year":"2023","word_count":320,"keywords":["Go","programming_languages:R","AI","RPA","programming_languages:Go","RAG","llm_models:Gemini","generative AI","R"],"extracted_tech_keywords":["AI","generative AI","RAG","R","Go","RPA","llm_models:Gemini","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/apple-optimises-llms-for-edge-use-cases\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10059510,"title":"What could go wrong with Neuralink?","content":"A few days back, Elon Musk’s Neuralink announced that it would soon be starting clinical trials of implanting a chip in the human brain. Musk’s ambition is to make brain-computer interfaces as simple as LASIK surgery. Last year, Neuralink implanted a chip in a monkey’s brain. Pager, the monkey, could play a video game using a joystick. Here, the Neuralink chip recorded Pager’s brain activity and sent it back to the computer for analysis. Although fascinating, a section of people remained sceptical about the whole initiative. A year later, now, when this experiment is being extended to humans, scientists, ethicists, and activists are making their concerns very clear. And now, as per a Fortune magazine article, the company has had internal turmoil. A number of key members have left the organisation. The ones remaining are also complaining about toxic and high work pressure environments. Neuralink – ambitious or gimmicky San Francisco-based Neuralink was launched by tech wizard Elon Musk in 2016. The company recruited some of the top researchers, professors, and other professionals to work on this vision. The task at hand was to develop a coin-sized computer chip that will be implanted in a human brain by a robot surgeon. This chip will connect the brain to the computer interface and smartphones. While the broad aim of developing such a brain-computer interface (BCI) is to allow humans to be competitive with AI, Musk wants Neuralink to solve immediate problems like the treatment of Parkinson’s disease and brain ailments from depression and anxiety, dementia and even paralysis. Musk has called the AI symbiosis problem an existential threat in the past. First @Neuralink product will enable someone with paralysis to use a smartphone with their mind faster than someone using thumbs— Elon Musk (@elonmusk) April 9, 2021 Neuralink would not be the first to attempt BCI; it has existed since the early 2000s. Some experts say that there are over 300,000 people who already have some neural interface like a deep brain stimulator that is used to treat Parkinson’s disease. Three participants with paralysis in the BrainGate clinical trial controlled a tablet computer just by thinking about it, thanks to a tiny brain implant. Read more: https:\/\/t.co\/Jc8LPe4U8S #BrownBrainScience #BrownResearch pic.twitter.com\/fMdYlbxrfE— Brown University (@BrownUniversity) November 21, 2018 What works for Neuralink is the amount of investment. Founder Elon Musk invested $100 million in the company; the company’s current valuation is an estimated $500 million, according to PitchBook. Money aside, Musk as a brand also does the trick. The news of Musk’s plan to implant chip in the human brain has caught the whole world’s attention. However, not all are very happy about this. A Daily Beast report said that scientists have ethical concerns about this venture. One of the scientists, Dr Karola Kreitmair, assistant professor of medical history and bioethics at the University of Wisconsin-Madison, said that there is sufficient public discourse on the big picture implications of such technology. She also called it an “uncomfortable marriage between a company that is for-profit… and these medical interventions that hopefully are there to help people”. Speaking to Analytics India Magazine, Kazim Rizvi, the founding director of The Dialogue, posed some poignant questions: Whether the ends that the technology seeks to achieve is moral, especially when differentiating the use of the tech between care provision (healthcare) and enhancing existing abilities, in the absence of medically recognised biological or cognitive deficiencies? Whether the means to achieve the defined ends (like healthcare) is moral, especially when the same will entail experimenting on human bodies without clarity on its long-term physical and psychological effects? Whether the technology will be accessible to everyone, or will we end up creating a tiered society where the poor cannot access this technology, creating a ripple effect on education, employment, and the economy? What does this mean for national security, and if used by soldiers in wars, how does this tech interact with the international humanitarian law?Do we have domestic laws or international conventions to regulate this tech, given the drastic impact on bodily privacy and scope for mass surveillance? Merging the human consciousness with technology has far-reaching implications, and no material enhancement (healthcare being an exception) can be morally justified by putting human life at stake if we assess the conundrum from the lens of the German philosopher Immanuel Kant. Neuralink employees’ concerns The spotlight here is not just on the ethical implications of this project. A new report based on testimonials from Neuralink employees indicates unfavourable working conditions. Most employees complain about Musk’s dissatisfaction with the pace of work ‘even though we were moving at unprecedented speeds’. As per employees, the work culture at Neuralink is driven by fear. Of the eight scientists that helped Musk establish the company, only two – Dongjin Seo and Paul Merolla – are still with the company.","excerpt":"While the broad aim of developing such a BCI is to allow humans to be competitive with AI, Musk wants Neuralink to solve immediate problems like the treatment of Parkinson’s disease and brain ailments.","categories":["AI Features"],"tags":[],"author_name":"Shraddha Goled","publish_date":"2022-01-31T19:28:52","publication_year":"2022","word_count":803,"keywords":["AWS","AI","cloud_platforms:AWS","programming_languages:R","Aim","ViT","analytics","GAN","R"],"extracted_tech_keywords":["AI","analytics","Aim","AWS","R","GAN","ViT","cloud_platforms:AWS","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/what-could-go-wrong-with-neuralink\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10163661,"title":"Fractal’s Multilingual Health Platform Vaidya.ai Ups the Game with Improved User Features and Speed","content":"India’s first AI unicorn, Fractal, has been actively introducing generative AI models this year. Earlier this year, the company released Kalaido, an Indic text-to-image diffusion model, and recently, they released Vaidya.ai, a healthcare AI platform. With reported plans to raise $500 million through an IPO at a valuation of $3.5 billion, Fractal is spearheading the AI movement in India. Accessible Health for All In an exclusive interaction with AIM, Suraj Amonkar, Chief AI Research & Platforms Officer at Fractal, explained that Vaidya.ai is not intended to serve as a diagnostic tool but as an information system. “Vaidya is not meant to give a medical diagnosis,” Amonkar emphasised. “It is a health companion that can have personalised and private conversations about your own individual health information.” Amonkar noted how people are used to going to Google for health queries that do not yet have personalized and reliable results. He believes that Vaidya.ai seeks to offer a more tailored approach than the older way to access this information. Fractal’s Vaidya is one of the first multi-modal models available in the medical domain. “A lot of health information is available in the form of images, and hence it was important for us to build a system that can understand image data from the users. This provides the ability for users to upload pictures of any health data to get more information from Vaidya.” Vaidya.ai has the additional advantage of being available in Indic languages. “Vaidya is accessible not only in English but is also available in Indic languages to help bridge the gap of accessing quality health information in local languages,” he said. The chatbot can detect the language of the user and is intelligent enough to respond back in the same language. New Updates Already Rolling Fractal is already working on rolling out updates to the health app. The platform has recently undergone usability enhancements where the response time has been significantly improved, along with improved user experience. Amonkar said, “We have enhanced the experience of the user by allowing users to save their conversations while continuing to keep their data private – even providing an option to delete all their conversation data.” The app has also been enhanced to have a faster response time for the users. Fractal has also implemented RAI certification process for their AI products and have run a certification process on Vaidya with this framework. “We have taken care to ensure data security and data privacy and have also run our systems through the responsible AI certification that our RAI team has administered,” said Amonkar. While Vaidya.ai currently focuses on general health information for consumers, Fractal is exploring partnerships with healthcare providers and diagnostic firms to enhance its capabilities in specialised areas Progressing Toward Perfection Vaidya.ai is built on a robust dataset that includes image datasets, scientific literature and respected medical datasets like PubMed. The platform allows the user to get answers to diverse health-related queries, from general wellness advice to specific medical conditions. The Fractal team has been focusing on curating high-quality data. Their data scientists, along with a data vendor, have worked to annotate and improve the data, including input from doctors who have helped validate the results. The public datasets, curated datasets and the annotated datasets have helped in building the algorithms. The algorithms consist of many different models and involve a pipeline of VLMS and LLMS that work together to help generate answers. The team has also worked on guard-rails on the system that will try to intercept and avoid any unethical behaviour on the system. “We have built a very robust pipeline of foundation models to get greater accuracy and have also taken efforts to implement guard-rails to intercept incorrect use. Our models aim to provides personalized health information while ensuring that users consult medical professionals for medical advice,” Amonkar reiterated. To benchmark the system’s performance, Fractal tested the model in a PG NEET exam (India’s postgraduate medical entrance test), achieving an impressive accuracy rate of 83%). Future Towards More Advanced AI Systems Amonkar highlighted Fractal’s broader research initiatives in reasoning and agentic systems as part of their mission to push the boundaries of AI toward AGI. “Our AI research team’s mission is to push the boundaries of AI towards AGI,” he stated. “We believe that three pillars of foundational-models, reasoning systems, and multi-agentic systems will all be necessary to build AI systems that move towards AGI and have research streams that are working on all these areas. We have recently published our research work in top AI conferences like NeurIPS and ICLR.”","excerpt":"Vaidya.ai has been trained on millions of datasets, which are a blend of publicly available, curated, and self-collected data, including images, medical summaries, and conversations.","categories":["AI Highlights"],"tags":["AI Chatbot","Healthcare","Kalaido","multilingual","Suraj Amonkar","Vaidya AI"],"author_name":"Vandana Nair","publish_date":"2025-02-14T11:56:15","publication_year":"2025","word_count":761,"keywords":["Kalaido","Go","unicorn","AI","AI Chatbot","R","IPO","Modal","multilingual","responsible AI","Aim","generative AI","Vaidya AI","Suraj Amonkar","Healthcare","foundation models"],"extracted_tech_keywords":["AI","generative AI","foundation models","Aim","R","Go","responsible AI","unicorn","IPO","Modal"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/fractals-multilingual-health-platform-vaidya-ai-ups-the-game-with-improved-user-features-and-speed\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10081158,"title":"Samsung Looks to Hire 1,000 Engineers for AI\/ML Roles in India","content":"Samsung India is looking to hire around 1,000 engineers for its research and development centres across the country. These include Samsung R&D centres in Bangalore, Noida, and Delhi. The engineers will be recruited from multiple streams spanning computer science and allied branches, IT, electronics, instrumentation, etc. In addition, it will also recruit from streams such as mathematics and computing. As per a statement released by the company on Wednesday, the engineers will join these institutes in 2023 and will work on new-age technologies that include artificial intelligence, machine learning, deep learning, image processing, IoT, connectivity, cloud, big data, business intelligence, predictive analysis among a host of others. “Strengthening their focus on innovation and cutting-edge technology, Samsung’s R&D centres aim to hire new talent from India’s top engineering institutes who will work on breakthrough innovations, technologies, products and designs, including India-centric innovations, that enrich people’s lives. This will further our vision of powering Digital India,” said Sameer Wadhawan, head- human resources, Samsung India. The R&D centres will hire around 200 engineers from the IITs, including IIT Madras, IIT Delhi, IIT Hyderabad, IIT Bombay, IIT Kharagpur, and IIT Kanpur, among others. The company has also given out over 400 pre-placement offers to students at the IITS and other top institutions. Samsung’s R&D centres in India have built a strong culture of patent filing, with over 7,500 patents being filed in the fields of multi-camera solutions, televisions, digital applications, 5G, AI, ML, IoT, and camera & vision technologies, among a host of others. In November 2021, the company announced that it would hire 1,000 engineers from top institutions, including IITs and BITS Pilani, and create a robust R&D pool for domestic and global markets.","excerpt":"The engineers will join in 2023 and will work on new-age technologies that include artificial intelligence, machine learning, deep learning, image processing, and others.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","Deep Learning","engineers","IIT","Machine Learning","samsung india"],"author_name":"Aparna Iyer","publish_date":"2022-11-30T17:04:07","publication_year":"2022","word_count":281,"keywords":["big data","artificial intelligence","machine learning","samsung india","AI","ML","engineers","Machine Learning","Git","RAG","Aim","deep learning","Deep Learning","IIT","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","Aim","RAG","R","Git","big data"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/samsung-looks-to-hire-1000-engineers-for-ai-ml-roles-in-india\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":44446,"title":"Unpacking Facebook’s Open-Sourced Graphics Algorithm That Flags Child Abuse Content","content":"Over the past few years, governments and other private organisations have been adopting several policies in order to mitigate the issues of child abuse and other graphical issues. It has now become a serious concern around the globe. Several tier-I organisations like IBM and Facebook, among others, have stepped forward to curb these issues and make the internet users feel safer. Facebook has been working in this regard for a few years now and has been open sourcing a number of its projects, in order to help to overcome the shortcomings while investigating or mitigating the child abuse cases. Recently, Facebook announced that the company is open-sourcing two technologies during the child safety hackathon which will help in detecting the graphical abusive contents, child exploitation, terrorist propaganda and other such threats. These technologies work in a method such that they will store the files in the form of short digital hashes and later comparing them with other instances in order to determine whether the files are identical and nearly identical images as well as videos to fight abuse on the internet platforms. Behind The Technologies The technologies known as PDQ and TMK+PDQF are syntactic hashers and are part of a suite of tools that Facebook use to detect harmful contents. PDQ is a photo matching algorithm which is inspired from pHash and TMK+PDQF is a video matching technology which was developed by Facebook’s Artificial Intelligence Research team and academics from the University of Modena and Reggio Emilia in Italy. As mentioned above that the files will be stored as short digital hashes, the reasons for the files to be stored as hashes because hashes have high-throughput and can be shared more easily with other companies and non-profit organisations. These hashers excel in finding media which are shared with minimum adversariality. PDQ is for a Perpetual hasher, a spectral hashing algorithm which uses a Discrete Cosine Transform and one of its outputs is a Quality metric. It is a photo-hashing algorithm where the hashes are 256 bits with hamming distance. This technology is relatively fast and runs at approximately the same speed as disk reads. This technique is effective at matching visually similar images as well as separating dissimilar images. It is mainly designed for lightly modified images which are non-adversarial and while producing hashes of images, PDQ can compute the hash of a rotated or flipped photo. TMK+PDQF runs at a very high multiple of video playback speeds depending on the storage density. TMK is for Temporal Match Kernel which is a video-similarity-detection-algorithm and produces fixed-length length video hashes on the order of 256 kilobytes. It can detect visually identical videos regardless of frame rates, pixel resolution, etc. The purpose of TMK+PDQF is to communicate the results of more powerful semantic-matching algorithms. Outlook Every year India celebrates the National Child Abuse Prevention Month in April whereas the bitterness is, according to reports, it is been ranked as the topmost country to face child abuse. The two technologies are open-sourced on GitHub with a motive that the industry partners, tech companies, smaller developers and other non-profit organisations can use them to easily identify abusive contents and share hashes or digital fingerprints of different types of harmful and abusive contents. These technologies will also work as an add-on layer of defence who are already using their own content matching technology. Facebook further tied their hands with The University of Maryland, Cornell University, Massachusetts Institute of Technology, and The University of California, Berkeley to research new techniques to detect intentional adversarial manipulations of videos and photos to circumvent our systems.","excerpt":"Over the past few years, governments and other private organisations have been adopting several policies in order to mitigate the issues of child abuse and other graphical issues. It has now become a serious concern around the globe. Several tier-I organisations like IBM and Facebook, among others, have stepped forward to curb these issues and […]","categories":["AI Features"],"tags":["Facebook","social media","social media AI"],"author_name":"Ambika Choudhury","publish_date":"2019-08-13T19:00:44","publication_year":"2019","word_count":597,"keywords":["Go","artificial intelligence","TPU","programming_languages:R","AI","social media","Git","RAG","social media AI","GAN","Facebook","GitHub","R"],"extracted_tech_keywords":["AI","artificial intelligence","RAG","TPU","R","Go","Git","GitHub","GAN","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/unpacking-facebooks-open-sourced-graphics-algorithm-that-flags-child-abuse-content\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10051615,"title":"Digital Foundations Are Becoming The Bedrock Of All Organisations: Akhilesh Ayer, WNS","content":"Data is an important arsenal for companies to grow and progress in their respective fields. However, huge amounts of data may be useless for companies that do not have an effective data analytics and management system in place. Data analytics in itself is an all-encompassing field that is witnessing rapid advancement. Analytics India Magazine interviewed Akhilesh Ayer, the Executive Vice President and Head for WNS’ Research and Analytics (R&A) practice. Ayer shares with us the learnings about data and business processes that he has gained through two decades of his professional journey. AIM: What differentiates WNS from its competitors? Akhilesh Ayer: What makes WNS stand apart is that we excel in both industry expertise and functional domain knowledge. That is how we approach the building of our tech stack as well – we create domain-centric solutions built on futuristic AI and ML technologies developed on Intelligent Cloud. We have a consulting-led, proprietary framework-driven approach to solving business problems such as supply chain analytics, fraud modelling, claims and underwriting solutions, pricing, customer propensity to buy, credit risk, etc., using client-preferred technologies in both next-gen Big Data or legacy data environments. Our flexible approach to delivery means that we offer our solutions both on-premise or on scalable cloud-enabled solutions (Microsoft Azure, AWS, GCP). Our teams are equipped to work on any kind of technologies and techniques across the spectrum, including Big Data technologies (Apache, Hadoop, Spark, MongoDB, Cassandra), Visualization (Tableau, Qlik, Power BI), AI\/ML\/Data Science programming\/coding languages (R, Python, PySpark, SQL, SAS, R Shiny), Cloud (AWS, Azure, GCP), to name a few. Modularity is another strong suit of the WNS tech stack. Given the vast number of industries and functions within them that we work with, we can offer these solutions as plug-and-play services for clients, thanks to this modularity. We are constantly looking to move to next-gen technologies to better our stack and strengthen our proprietary solutions. We also have an ecosystem of niche partnerships to supplement our proprietary offerings in the latest technologies, including data lineage, governance and domain-driven data partnerships. We have also created WNS Co-Creation Labs – there is one in Pune, another in New York, a third in London, and a fourth coming up in Australia. At the labs, we work with customers to reimagine their existing business problems, especially on digital platforms. We leverage design-thinking principles to co-create solutions using data, analytics, AI and ML. Using the elements of ‘play’, with intuitive workflows and designed as immersive spaces, these labs serve as, if I can use the term, ‘time machines’ that offer clients a peek into how their futures will be transformed with the solutions we co-create today. We also have a host of productised ML\/Big Data offerings. SKENSE is a cognitive data capturing and contextualisation platform. Unified Analytics Platform is an end-to-end industry analytics platform using domain-specific AI algorithms and hosted on Intelligent Cloud. Then there is SOCIOSEER, an AI-driven solution for social media analytics and monitoring. SENTINEL is an AI-led trendspotting tool. And these are just a few examples. We have also developed 25+ digital accelerators, including a cutting-edge AI workbench and an AI-led data catalogue that reduces time-to-insights for our clients on specific business use cases. AIM: How is data analytics helping businesses transform, especially in the post-Covid era? Akhilesh Ayer: Digital foundations are increasingly becoming the bedrock of all organisations, irrespective of their size. Large, medium, and small companies are using the power of analytics and AI to accelerate their own decision-making, optimise costs, or create differentiated products and offerings. Interestingly, while data is being used to enhance the quality of a certain offering or to optimise processes, the insight or the data itself has monetary value that companies can resell and repackage with their partners. For instance, the underwriting of an insurance policy needs to be based on the insured’s past data as well as industry data in general. With data and analytics, one is actually able to effectively look at both and make decisions that are both faster and have a much more nuanced analysis. While the advantage of making sound decisions based on statistical principles has always been understood and valued, with data analytics, decisions today have become real-time and are enriched with data that comes from both traditional and non-traditional sources. In the post-pandemic world, our clients have been forced to reshape the way they engage with the end customers. This digital disruption is driving the push towards embedding analytics into all business areas to enable digital transformation – to improve business productivity and foster a data-driven decision-making culture more than ever. AIM: What are the current trends and challenges in data analytics? Akhilesh Ayer: It is phenomenal and fascinating to watch how data is driving business decisions in real-time – even for traditional business – enabling adjustments to supply chains, offering better visibility, and aiding dynamic pricing. There are quite a few challenges as well. Prioritisation – More often than not, companies are aware of the importance of data but do not know how to prioritise their data investments to get optimal ROI. Spending huge amounts of money on creating large data infrastructure is not the only solution. Scalability – Second comes the problem of scaling. Everyone wants to do it. But in many cases, they do not know how to do it. And if they do, then they are probably unaware of the factors that are affecting the quality of data – it could be poor data governance or poor data taxonomy. In several instances, lack of senior management alignment, talent shortage, and general lack of prioritisation when using data analytics affect scaling significantly. Quality of Data – That is as far as the use of data analytics is concerned. When it comes to data analytics itself, there are a few challenges. The first being the quality of data itself – while there are a lot of tools and technologies that have evolved to make sense of data, if you don’t organise your underlying data ecosystem, then it will only help so much. Talent – Another major challenge for many developed markets like the US and Europe is a huge talent shortage in data analytics. Business Context – Finally, today, there is a lot of focus on underlying technologies used for data analytics. But there needs to be a lot of appreciation of the actual business and the business problem that the company is trying to solve through data. AIM: How will data analytics evolve as a profession in the coming years? Akhilesh Ayer: The pandemic has definitely forced companies to sit up and pay attention to data analytics. Even those for whom this wasn’t a priority before have jumped onto the digital bandwagon. Consumers in a post-pandemic era are engaging in ways they have never before; they are increasingly looking at differentiated experiences, and companies will have to factor that into their strategies going forward. So investments in data will increasingly start receiving senior management attention, the right kind of budgets and focus across departments and functions within an organisation. Data analytics will become a core differentiator for a lot of organisations. That, in my opinion, is a big shift. Consequently, there will also be a demand for seasoned analytics professionals. It will also result in demand for more innovative and scalable technologies and talent that is good at predictive analytics, AI, ML, Big Data – the whole shebang. One thing that will change, though, is the way data analytics is perceived from the business point of view. Earlier, it was very method and tool focused – that is required too – but now, it will also need to be understood from the ‘relevance to the business’ perspective. So, while data professionals grow as domain experts, they will also need to develop as functional experts because it is this combination of domain knowledge and functional knowledge, along with computing skills and the ability to visualise the analysis to tell a cohesive data-driven story, which will bring real value for businesses. AIM: Which is one field where you feel the potential of data analytics hasn’t been sufficiently discovered, and why? Akhilesh Ayer: To say that there is still a lot of potential to further deploy and leverage data analytics would perhaps be a better way to answer this question. Two areas companies should focus on in the coming 18-24 months are a) setting up an augmented data governance function and b) enabling data-driven transformation leveraging the power of intelligent cloud. When enterprises move data to the cloud, they also modernise their data platforms. Intelligent cloud and AI today offer opportunities to companies to leapfrog the competition and accelerate their data analytics journey. There are also breakthroughs in cloud-led, out-of-the-box integrations that enable AI and ML innovations using open-source tools and technologies. Function wise, I would say even mature industries that have adopted analytics relatively early have far greater potential now to harness the power of AI, whether it is for underwriting and pricing in insurance, or supply chain management and customer intelligence in retail and CPG, or credit and risk functions in banking. AIM: What do you have to say about the legal and ethical aspects of working with a large amount of data — democratisation, data privacy, ethical AI? Akhilesh Ayer: The challenge remains as to who would take ownership of the action initiated by unsupervised algorithms. I think we are still at the initial stages of that maturity path. The more trained the algorithms, the more the variety of data fed – there will be fewer concerns. Also, in what areas algorithms should be deployed or not is another grey area. But both organisations and regulatory bodies need to evolve the right framework and governance mechanisms constantly. Consented data is just the input. Clearly, calling out the boundary conditions of the output and application area is also equally important. AIM: Where do you think most companies falter on their data and analytics strategy? Akhilesh Ayer: I don’t think there is just one answer. They often fail at different points along the data value chain. Many evolved companies have kind of sorted out the upstream issues but are now challenged to articulate business value through the right use cases. But companies that are still in the early stages of their data analytics journey are struggling with foundational problems. Across the board, there are three common challenges facing companies – 1) data, analytics and AI strategy, its sponsorship and funding, 2) talent availability and capability leverage and 3) creating a sustainable data-driven culture and ensuring data, analytics and AI, form the core DNA of the organisation.","excerpt":"Large, medium, and small companies are using the power of analytics and AI to accelerate their own decision-making, optimise costs, or create differentiated products and offerings.","categories":["AI Features"],"tags":["Data Analytics","Data Management","data strategy","head of analytics","Interviews and Discussions"],"author_name":"Shraddha Goled","publish_date":"2021-10-14T15:00:00","publication_year":"2021","word_count":1770,"keywords":["data science","AWS","Data Management","AI","Azure","data strategy","ML","head of analytics","RAG","Aim","analytics","edge AI","Data Analytics","predictive analytics","Interviews and Discussions"],"extracted_tech_keywords":["AI","ML","data science","analytics","Aim","edge AI","RAG","predictive analytics","AWS","Azure"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/digital-foundations-are-becoming-the-bedrock-of-all-organisations-akhilesh-ayer-wns\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10120567,"title":"Soket AI Labs Partners with Google Cloud to Boost Pragna-1B Model","content":"Soket AI Labs, the Indian AI research firm behind Pragna-1B, India’s first open-source multilingual foundation model, has announced a new collaboration with Google Cloud to further enhance the model’s capabilities and reach. Pragna-1B, which was initially released on May 1, 2024, aims to enable the adoption of Generative AI in India by providing support for vernacular languages such as Hindi, English, Bengali, and Gujarati. Abhishek Upperwal, Founder of Soket AI Labs, said, “By leveraging Google cloud, Pragna-1B, despite being trained on fewer parameters, is efficient and compares performance in language processing tasks to similar category models.” He further added, “Tailored specifically for vernacular languages, Pragna-1B offers balanced language representation and enables faster and more efficient tokenization suited for organisations seeking optimised operations and enhanced functionality.” The collaboration also aims to make Pragna-1B more accessible to developers and organisations. Soket AI Labs plans to list its AI Developer Platform on the Google Cloud Marketplace and the Pragna series of models on the Google Vertex AI model registry. This integration will provide developers with a streamlined experience for fine-tuning models using high-performance resources like Vertex AI and TPUs. The model has been designed specifically with Indian contexts in mind, ensuring transparency and clarity for enterprises integrating AI into their operations. Soket AI Labs leveraged Google Cloud’s AI infrastructure to achieve efficiency and cost-effectiveness in the development of Pragna-1B. Google Cloud also plan to list Soket’s AI Developer Platform on the Google Cloud Marketplace and the Pragna series of models on the Google Vertex AI model registry. The collaboration between Soket AI Labs and Google Cloud also extends to technical work on training large-scale models and curating high-quality datasets for Indian languages. This joint effort aims to promote AI innovation in India while ensuring transparency and clarity in the development process. The story so far Soket AI Labs, founded by Abhishek Upperwal in 2019, created ‘Bhasha,’ a series of high-quality datasets designed for training Indian language models. This includes ‘Bhasha-wiki,’ which consists of 44.1 million articles translated from English Wikipedia into six Indian languages, and “Bhasha-wiki-indic,” a refined subset focusing on content relevant to India. Pragna-1B, features a Transformer Decoder-only architecture with 1.25 billion parameters and a context length of 2048 tokens. Trained on approximately 150 billion tokens, with a focus on Hindi, Bangla, and Gujarati, Pragna-1B delivers state-of-the-art performance for vernacular languages in a small form factor. In a recent LinkedIn post, Upperwal highlighted the improvements in GPT-4o’s tokenizer and vocabulary size, which now supports 200k tokens. However, he noted that Pragna-1b’s tokenizer still outperforms GPT-4o when it comes to Kannada, Gujarati, Tamil, and Urdu, serving as a motivation for Soket AI Labs to improve support for Hindi and other Indian languages. Soket AI Labs is also experimenting with a Mixture of Experts model, expanding the languages supported and exploring different architectures for increased optimisation.","excerpt":"Pragna-1B, developed by Soket AI Labs and Google Cloud, delivers state-of-the-art performance for vernacular languages.","categories":["AI News"],"tags":["Google Cloud","Soket AI Labs"],"author_name":"K L Krithika","publish_date":"2024-05-15T18:24:20","publication_year":"2024","word_count":473,"keywords":["Google Cloud","TPU","AI","Google Vertex AI","GPT-4o","ML","Soket AI Labs","RAG","Aim","generative AI","model registry","R"],"extracted_tech_keywords":["AI","ML","generative AI","GPT-4o","Aim","Google Vertex AI","model registry","RAG","TPU","R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/soket-ai-labs-and-google-cloud-launched-indias-first-multilingual-ai-model\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10073722,"title":"Bengaluru Ranks Second in APAC Top Tech Hub List","content":"Bengaluru has ranked second in the top-tech hub list across Asia-Pacific, according to a report by Cushman and Wakefield. The IT city, followed by Chennai, Hyderabad, and Delhi, scored on talent and demand for quality office spaces from domestic and multinational companies. Anshul Jain, MD, India & South-East Asia, Cushman and Wakefield said, “The IT and tech-driven sectors remain the key driving force for the Indian economy. The strong market fundamentals coupled with key policy changes have not only made India a favourable investment hub for global IT players but also made it the most sought-after destination to find global tech talents and set footprints for further expansion.” Globally, around 46 top tech markets were identified from among 115 tech cities, with six of the 14 cities in the APAC region being in India. The report states that the nation has added more than 500,000 jobs in IT over the past financial year. The report saw the city’s tech sector accounting for an average 38%-40% share in annual leasing, higher than the national average of 35%. Experts believe that growth in tech leasing is expected to continue in the coming years. Home to Microsoft and Facebook, Hyderabad has over 44 million sqft of office projects under construction. Meanwhile Chennai earned itself the second-largest software exporter tag, being one of the largest data centre markets in India.","excerpt":"The report saw a higher share than the national average of 35%, with the city’s tech sector accounting for 38%-40% share in annual leasing","categories":["AI News"],"tags":[],"author_name":"Bhuvana Kamath","publish_date":"2022-08-26T16:54:00","publication_year":"2022","word_count":226,"keywords":["R","RAG","programming_languages:R","AI"],"extracted_tech_keywords":["AI","RAG","R","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/bengaluru-ranks-second-in-apac-top-tech-hub-list\/","complexity_score":2,"technical_depth":4,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":17793,"title":"Can Micro-segmentation Save CPG Companies Find Niche Customers In The Age Of Personalization","content":"If there is one sector that has been massively affected by the onslaught of e-commerce retailers, it is the Consumer Packaged Goods industry dominated by the likes of leading FMCG companies such as Johnson & Johnson, Unilever, Kimberly Clark and Procter & Gamble. Now that the likes of Amazon and Flipkart have become mainstream, big box stores and CPGS are battling an uphill battle to better understand their customers. The profound disruption wrought by e-tailers and the rise of millennials have created significant opportunities for CPGs to take steps to transform into digital first and customer centric businesses. While CPG companies are battling a loss of customers on one hand, they are also faced with questions who are my customers and how we reach them better. Seeking customer -centricity: Microtargeting the right demographic However, seeking customer centricity starts by understanding the customer base better. More and more consumer packaged goods manufacturers and retailers are finding that micro-targeting strategies (usually advanced predictive analytics) applied by ecommerce businesses —are unleashing new, deeper insights into their targeted customer segments and allowing them them to grow brands in a profitable way. Putting a laser focus on customers starts with leveraging predictive analytics that is turning into game-changer for businesses. According to global information, data and measurement company Nielsen, in times of this challenging economy and intense market competition, an increasing number of mainstream consumer packaged goods (CPG) companies and retailers have dug deeper into their petabytes of data to develop a clear understanding of consumers in their categories and sub-categories. So what are companies doing to gain deeper customer knowledge – turning to analytics technologies. A recent KPMG survey points out that 30% of CPG companies are deploying cutting-edge data analytics tools for accelerated growth. Over the years, the use of predictive analytics has doubled from 24 per cent to 49 per cent. Some of the analytical tools used increasingly are real-time tracking systems, scenario modeling and micro-targeting capabilities. Analytics India Magazine lists down ways to micro-target consumers Begin by creating your own dataset: Most big companies have datasets in-house collected through  online surveys or subscriber email lists. Datasets can also be bought from companies that specialize in selling certain data to be loaded in the analytics database. Once the dataset loaded, one can analyze consumer’s buying behavior and find hidden patterns and insights: The company can build a model accordingly based on consumer’s similar traits Create a dependent variable: Variables may vary from gender type to residence to users who share the same devices or interest. By creating a variable, you can club the targeted segment under the same umbrella. Further on, build a predictive model with the same variable:  It is this statistical process for determining the strength of a relationship between one dependent variable and a series of other variables that is known as regression modeling. Power to predict valued shoppers Today, retailers and CPGs spend huge amounts of money in promotions to draw more customers and many CPGs leverage consumer response models to help retailers understand how users will respond to marketing activities. Understanding the user’s needs and preferences forms the basis of better business. A model can be built to accurately identify patterns and  classify  based on  certain  static  parameters  using  Support  Vector  Machine  algorithm  of Machine Learning on a set of user data. It enables  to  provide  predictive  insights  based  on behavioral data leading to efficiency and responsiveness to marketing efforts. Use Case: See how Mahindra Holidays & Resorts has created micro-segments of customers to better engage users For Mahindra Holidays & Resorts, one-to-one engagement with customers was the key to retaining memberships. The company wanted to better understand its individual members and their preferences to create highly relevant engagement platforms. The key challenges for the organization were: Identify all member data  sources  to build a better understanding  of members Optimize the big data platform to create micro segments of members to drive relevant communication The solution framework involved leveraging collaborative filtering model to identify the travelers who were most likely to visit the resort during the monsoon period. The big data solution further helped divide the members into micro-segments based on the resort they were most likely to visit and send targeted promotions. Based on the insights, the company sent customized emails and SMSes to members and recorded a 24% open rates for campaigns. Outlook As new channels emerge – etailers, niche ecommerce specialists and third party aggregators such as Urban Clap, CPG companies need to fundamentally reimagine and restructure their selling process and engage consumers better. They also need to better understand the users’ consumption patterns and leverage the right tools and technologies to understand how the consumers’ purchase decisions are getting disrupted with the rise of new-age providers. As the share of consumption rises, they need to understand how premium products are taking off across categories. In a hyper-connected world wherein internet connects billions of devices, having a laser focus on the customers have become a sort of corporate blind spot. It’s here that advanced analytics techniques and machine learning algorithms are applied to on micro-segments of users who can drive further conversions.","excerpt":"If there is one sector that has been massively affected by the onslaught of e-commerce retailers, it is the Consumer Packaged Goods industry dominated by the likes of leading FMCG companies such as Johnson & Johnson, Unilever, Kimberly Clark and Procter & Gamble. Now that the likes of Amazon and Flipkart have become mainstream, big […]","categories":["IT Services"],"tags":[],"author_name":"Richa Bhatia","publish_date":"2017-09-18T13:00:46","publication_year":"2017","word_count":858,"keywords":["big data","Go","machine learning","AI","R","Git","RAG","analytics","GAN","predictive analytics"],"extracted_tech_keywords":["AI","machine learning","analytics","RAG","predictive analytics","R","Go","Git","big data","GAN"],"url":"https:\/\/analyticsindiamag.com\/it-services\/can-micro-segmentation-save-cpg-companies-find-niche-customers-age-personalization\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10005357,"title":"How Saudi Arabia Is Looking To Develop &#038; Integrate Artificial Intelligence In Its Economy","content":"Yet another country has gone to release the potential of artificial intelligence to fuel its economic growth plans. One of the largest oil manufacturing nations, Saudi Arabia, has now formulated and launched a nation-wide policy on AI. The project has been signed with approval from King Salman and aims to add the market value of up to 500 billion riyals (US$133 billion) in the country’s GDP by 2030. The policy formulation had begun last year when the Saudi Data and Artificial Intelligence Authority (SDAIA) was established through a royal decree in August 2019 to lead the charge of the nation’s transformation into a data-driven economy. Besides, the body is working on implementing a cloud platform, intending to build one of the biggest cloud frameworks in the Middle East by connecting 83 data centres controlled by over 40 government bodies. Under the SDAIA, three units, namely the National Information Center,  the National Center for Artificial Intelligence and the National Data Management Office, were created to manage and execute various aspects related to the AI policy. Apart from this, the kingdom also established a national data bank consisting of more than 80 state datasets, which is equal to about 30 per cent of the Saudi government’s digital assets, revealed Abdullah bin Sharaf Al Ghamdi, President of SDAIA. The Saudi Data and Artificial Intelligence Authority (SDAIA) launched its brand identity for the AI policy at an event in national capital Riyadh with the theme ‘Data is Oil of the 21st Century’. Speaking at the brand launch Dr Al-Ghamdi, President of SDAIA, said, “Data is the single most significant driver of our growth and reformation in the 21st century. We have a definite vision and roadmap for transforming Saudi Arabia into a leading AI and data-driven market. SDAIA is at the vanguard of this change and is primed for national data and AI agenda definition, implementation and knowledge.” The event was attended by ministers and the leading government officials and centred on their readiness to make the Kingdom among the world’s leading economies in the adoption of AI. Ever since the introduction of the Saudi Vision 2030 plan in 2016, Saudi Arabia has achieved significant progression in unfastening the value of data as a public asset. Dr Al-Ghamdi, further stated, “We have seen first-hand, the immediate impact of AI and data-driven actions, and their potential to drive Saudi Arabia’s coming economy, but we are yet in the beginning stages with many untapped opportunities open. We, at SDAIA, have been tasked with setting the national data and AI strategy, and carrying on our country’s vision for the future by optimising our national resources, advancing efficiencies and supporting the creation of diversified economic and technological sectors.” Neom: AI-Based Mega City Saudi Arabia’s Vision 2030 strategy, rolled out in 2016 was aimed at creating sustainable cities and making sure there is optimum use of resources in urban planning, infrastructure and important sectors of the economy.  The Saudi government recognised that AI and automation are very critical in this context to enhance efficiencies in the government. Smart cities are an essential part of the infrastructure plan and the country is working on constructing a US$500 billion city called Neom spread over 26,000 square kilometres with state of the art technologies based on AI, IoT and blockchain. This megacity is at the heart of the investment and development in AI,  and the country has publicly committed $500 billion for it. Neom will integrate robotics and AI seamlessly into every aspect of citizen’s lives in an effort to produce revenues from key economic sectors for the future, and meet the increasing challenges of constructing a state around oil alone. The country has embraced AI and robotics and it became the first country to grant citizenship to a robot when in 2017, the Kingdom granted citizenship to Hanson Robotics’ Sophia. Skills Enhancement To Match AI Plans With the plan for AI, Saudi Arabia has now officially entered the global race for AI development. It certainly has no shortage of funds to achieve its objectives, even though it may struggle with talent in this space. To counter the scarcity of AI skills, the Kingdom will need to adapt to an open model of educational curriculum similar to global education standards. To match Saudi Arabia’s ambitious plan on artificial intelligence, the government is also planning to make educational reforms more aligned with AI and digital skills. To generate skills and match them with the AI vision, the ministry of education school is reforming university curricula focused on analytics, data science and data security. Diversifying The Economy With AI At the moment, the AI and data economy of Saudi Arabia is estimated at USD 4-5 billion. By leveraging the power of AI and data analytics in government departments and processes, Suadi Arabia wants an opportunity to generate extra revenues and savings of over USD 10 billion GDP by 2030. Longer-term estimates project AI adding to a 14 per cent growth of global GDP by 2030, ‘the equivalent of a supplementary $15.7 trillion’. Thus, AI serves a very vital commercial opportunity, and by drawing together trends and networks in artificial intelligence, Vision 2030 attempts to broaden and diversify the Saudi economy. AI can address a broad range of financial and societal challenges in the country, ranging from the volatility of oil prices to accelerated urbanisation and water scarcity. Artificial intelligence also has many geopolitical implications, and this is another reason for a nationwide policy for the Middle-eastern nation.","excerpt":"Yet another country has gone to release the potential of artificial intelligence to fuel its economic growth plans. One of the largest oil manufacturing nations, Saudi Arabia, has now formulated and launched a nation-wide policy on AI. The project has been signed with approval from King Salman and aims to add the market value of […]","categories":["IT Services"],"tags":["how artificial intelligence works","intel global strategy","saudi arabia","what is artificial intelligence"],"author_name":"Vishal Chawla","publish_date":"2020-08-23T10:00:22","publication_year":"2020","word_count":913,"keywords":["data science","how artificial intelligence works","Go","artificial intelligence","AI","saudi arabia","ML","intel global strategy","Git","RAG","Aim","analytics","what is artificial intelligence","R"],"extracted_tech_keywords":["AI","artificial intelligence","ML","data science","analytics","Aim","RAG","R","Go","Git"],"url":"https:\/\/analyticsindiamag.com\/it-services\/how-saudi-arabia-is-looking-to-develop-integrate-artificial-intelligence-in-its-economy\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":2584,"title":"Interview &#8211; Anurag Verma, CEO at MathLogic","content":"[dropcap style=”1″ size=”2″]AIM[\/dropcap]Analytics India Magazine: What are some of the main tenets (philosophies, goals, attributes) of analytics approach and policies at your organization? [dropcap style=”1″ size=”2″]AV[\/dropcap]Anurag Verma: As I see there are two parts to this question – one is to do with our core beliefs around analytics and the other is to do with the central thought process around policies that govern our business. Both these parts feed into each other and we tend to address them almost interchangeably. However, to get more specific, let us look at analytics. Now we strongly believe that analytics is all about helping businesses becoming smarter about what they do, tactically and strategically, by using data and intelligence within the enterprise and outside the enterprise. Analytics is no longer a choice that a business would make…it is an imperative. The maturity and readiness might differ across businesses, geographies – but it is, and will eventually move towards a level playing field. We are excited to be at the center of this action and believe that we have solutions, skill set, and the experience that can help our clients in more ways than one. As to our policies and our fundamental philosophy – by the way we call it “our lighthouse”, we believe that in business as in life, the foundational principles remain the same. Be honest, deliver on promises, promise what you can deliver, accept that you will make mistakes, accept your mistakes, learn from them, take responsibility, improve and always work hard. You see there is nothing complex about it! AIM: Please brief us about some business solutions you work on and how you derive value out of it. AV: We provide a range of services and solutions for consumer industries (B2C companies) like BFSI (Banking, Financial Services and Insurance), Healthcare, Telecom and retail. We have three core offerings which we like to call as Foundational offerings. They are (i) Customer Analytics, (ii) Fraud identification and mitigations solutions, (iii) Risk and Collections analytics. The leadership team at MathLogic has significant experience working with numerous clients across geographies in these areas. We have consistently seen businesses gaining from our solutions. Individually we have worked on business problems that have for example increased client ROI on cross sell initiatives, reduced false positives significantly in identifying fraudulent transactions on credit cards, reduced default on loans, or improve collections efficiency…. AIM: How does a typical requirement gathering to delivery cycle looks like for you? AV: I wouldn’t say that it follows a particular cycle – it depends a lot on type of problem, the client, maturity continuum of the industry or client, the team you are working with, etc. Typically, we encourage our clients to sit across for a business problem, process, scoping and expectations setting discussion. We have found this discussion to be a critical input as it cuts through a lot of presumptions that both parties might have. It also allows for a clear articulation of problems and issues, possible challenges, and current processes – which in turn feed logically into scoping, solution design, project planning and execution. In addition what we also find critical is the need to have check-points during the logical progression of the analytical build phase. A regular sense check is absolutely essential to ensure that all stakeholders are on the same page and we are able to make course corrections as required. Clients also appreciate this as it gives them flexibility to make changes and adjustments as we go along. Rest is essentially preparing strong project governance that bakes in risk, contingencies and toll gates that help the execution team keep track. AIM: Please brief us about the size of your analytics division and what is hierarchal alignment, both depth and breadth. AV: We are a start-up and like many start ups very lean and flat in structure. Our core leadership team is involved directly with all clients and projects that we work on. We have a small team of bright analysts supporting us. We are expanding as we speak and excited about the journey ahead! AIM: Do you think it’s possible to become too married to the data that comes out of analytics? Where do you draw the line? AV: It is very much possible and we have seen this happening at service providers as well as with clients. I think the answer is obvious in the question itself. Analytics, like any other business stream, is one of the tools for businesses – a powerful tool. It should be used likewise, as one of the inputs into your business. We feel very strongly that any analytical output needs to be viewed in the right business context. There needs to be layer of business logic over the mathematics behind the scenes. This is reflected in our name FN MathLogic = Function of Mathematics and Business Logic We want businesses across industries and sectors to harness the power of their data to plug the gaps within their understanding – essentially to improve their decision making by making these decisions based on more scientific evidence and projections. [quote style=”1″]Analytics can be very powerful in doing that as can be seen from the experience of corporations like GE, Citibank, American Express, ICICI Bank, Google, Makemytrip, to name just a few.[\/quote] AIM: What are some of the data measurement points that are becoming more important for organizations? AV: Profitability or Return on Investment (ROI) is and would continue to be the most important metric for organizations. While, different department and level within organization may focus on different measurement points but there is clear movement towards tying each decision into profitability at more granular level. In a more analytically mature organization the level at which profitability (or some equivalent measure of ROI) can be\/ is measure is more granular. E.g. a marketing manager can track the incremental profits (net of campaign cost) that a particular campaign is expected to bring and actual results and risk manager can quantify the net impact of allowing or disallowing a certain type of transactions. AIM: How did you start your career in analytics? AV: I always had flair for “Analytics” even before I knew the term. In my MBA colleage, I was tutor for statistical courses. I started my career in management consulting and soon found myself gravitating towards more “quantitative engagements” and becoming internal consultant to anyone seeking guidance on data related aspects. After spending about five years in consulting, I received a group email about job opening in  a major foreign bank that was starting up their analytics team in India. I knew I had found my calling after the 30 minute job interview and I had no hesitation in moving out of consulting into analytics. It has been almost 12 years since and I have enjoyed each and every day. AIM: What do you suggest to new graduates aspiring to get into analytics space? AV: The answer is going to be short. Lots is being talked about analytics – in journals, business magazines, conferences, forums, social media…I keep meeting with young students and I see that they recognize analytics as a viable career option and are curious to know more about it. I would say it is a very exciting field with tremendous growth potential. It is also a field that allows one to explore new horizons and learn new things almost daily. Like any other field it requires dedication and hard work to succeed. So if you are interested in spending a good amount of time in the lone company of reams of data, identifying trends and inferring insights then come along! [pullquote align=”left”]More details are available on our website www.fnmathlogic.com. We are looking to hire and would love to talk to interested candidates and explore common ground.[\/pullquote]AIM: What kind of knowledge worker do you recruit and what is the selection methodology? What skill sets do you look at while recruiting in analytics? AV: We are a small start-up and we know that our success is closely linked to the quality of our people. We are looking for bright minds who want to make a career in analytics. We have a rigorous recruiting process that includes both technical as well as soft skills assessment. AIM: How do you see Analytics evolving today in the industry as a whole? What are the most important contemporary trends that you see emerging in the Analytics space across the globe? AV: Few things have happened in the last decade. First analytics has come to be recognized as a commercial science with applications that can help industries and businesses across boundaries. Second, lot of effort has been made to educate managers and business associates across hierarchy chain in understanding the practical applications of analytics and also to decode some of the jargon that is associated with analytics. Third, a lot of infrastructure related challenges have been addressed. This has also helped to expand the reach of analytics – so that we have now smaller businesses, new businesses also using analytics to make smart decisions. There are just a couple of trends that I would like to touch upon (i)            Natural progression is taking place – there is more data all around, newer industries are adopting analytics and there is certain commoditization of work streams. What was considered novel 5-7 years back is getting commoditized. Technology is playing a big role here. A number of things which were considered “art” and requiring manual intervention are going to become black box and human touch would be required more in interpreting the results and applying to business. We will see more push for Solutioning and product development in Analytics space in the coming years. (ii)          Big Data is here: These days you cant have an analytics conversation today without mentioning big data. They will bring about a fundamental change as industry finds a way to connect enterprise data with social media data. AIM: Anything else you wish to add? AV: Thanks for giving us the opportunity to connect with the large forum that you represent. Analytics is an interesting and evolving field which has seen tremendous growth and still is a young science. We at MathLogic are fortunate to be in the thick of things. We would be happy to connect with folks if they have any questions – our contact details are available over our website. Thanks. [spoiler title=”Biography of Anurag Verma” open=”0″ style=”2″] Anurag has 16 years of experience in Analytics and Consulting across North America, Europe, Australia and Asia.  He has deep experience helping clients with fraud and non compliance issues across Financial Services, e-commerce and health care. He is well versed in client engagement, building high end analytics teams and managing P&L responsibilities. Anurag holds a Master’s degree in Business Administration from IIM, Kolkata and a Bachelor’s degree in Engineering from IIT, Kanpur.[\/spoiler]","excerpt":"[dropcap style=”1″ size=”2″]AIM[\/dropcap]Analytics India Magazine: What are some of the main tenets (philosophies, goals, attributes) of analytics approach and policies at your organization? [dropcap style=”1″ size=”2″]AV[\/dropcap]Anurag Verma: As I see there are two parts to this question – one is to do with our core beliefs around analytics and the other is to do with […]","categories":["AI Features"],"tags":["Interviews and Discussions"],"author_name":"Дарья","publish_date":"2013-02-04T11:54:24","publication_year":"2013","word_count":1805,"keywords":["big data","Go","TPU","AI","RAG","Aim","ViT","analytics","GAN","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","analytics","Aim","RAG","TPU","R","Go","big data","GAN","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/interview-anurag-verma-ceo-at-mathlogic\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10093480,"title":"How Splunk Can Save the Day for Enterprise AI","content":"As ML spreads its might in all quarters including security, it turbocharges everything it touches. Software company Splunk recently added new security and observability features to identify threats of a new world with generative AI. Robert Pizzari, vice president of security at the company, speaks about why APAC is trailing behind other regions in terms of security, the growing concerns with GPT APIs and what Splunk plans to do about it. AIM: What are some of the major cybersecurity obstacles faced by organisations in the current scenario? Robert: With the growing number of interconnected devices in the modern enterprises along with the deluge of data — much of which is sensitive and confidential — the importance of cybersecurity continues to grow. In fact, according to Splunk’s latest State of Security 2023, 59% respondents in India report having faced data breaches in the past two years more often. In the same report, we learned that some of the critical challenges faced by organisations in the current scenario are: ● Ransomware attacks – Ransomware is always on the rise. This year, the number of organisations that dealt with ransomware attacks rose to 75% as compared to last year. ● Breach in cloud security – With 50% of respondents saying that the majority of their SOC (Security Operations Centre) team’s time is spent addressing issues in the public cloud, while just 13% spend most of their time solving on-premises issues. ● Supply chain attacks – Software supply chain attacks are top-of mind in the post-SolarWinds era. 95% of organisations have increased their focus on third-party risk assessment activity, up from an already noteworthy 90% from 2022. In today’s complex, hybrid world, cyber security challenges are only going to become more intense. AIM: With the arrival of generative AI and ChatGPT APIs flying around, what challenges do you see further in security? Robert: The arrival of generative AI and ChatGPT APIs presents several challenges in terms of security. These technologies have the potential to be misused by cybercriminals to create sophisticated and convincing phishing emails, deep fake videos, and other types of malicious content. Some specific challenges that could arise include: ● Greater attack of insider threats: Generative AI and ChatGPT APIs can be used to create fake identities or to impersonate employees, making it easier for attackers to carry out insider attacks. This could result in the theft of sensitive data or the compromise of critical systems. ● Difficulty in detecting advanced threats: Advanced technology can be used to create new types of malware or to obfuscate existing ones, making it harder for traditional security tools to detect them. This could result in advanced threats going undetected within networks and systems for a couple of months before causing significant damage. ● Challenges in compliance and privacy: The use of AI could raise compliance and privacy concerns, particularly in industries such as finance and healthcare. Enterprises need to use these technologies responsibly and ethically and comply with relevant regulations and protect individual’s privacy rights. AIM: Can you share some key India insights from the State of Security 2023 report? How is India faring compared to its global counterparts? Robert: Blockchain technology, 5G, the Internet of Things (IoT), artificial intelligence (AI), and other rapidly developing and pervasive technologies are offering significant development opportunities for businesses in India. However, the number of cyber attacks and data breaches in Indian organisations have gone up multi-fold over the last couple of years thanks to the expansion of the threat surfaces. According to the Splunk report, in India, 42% of Indian organisations were found to be overwhelmed by the number of attacks versus 23% in the rest of the world. Part of the problem seems to be the complexity of their tool ecosystems as 48% say their security stack is too complex, compared to 28% in the rest of the world. However, the report also highlights how system complexity has led to greater prioritisation of security investment in India and organisations in India are investing more in improving their cyber resilience as compared to the rest of the world. AIM: What trends and ground-breaking changes do you anticipate in the Indian cybersecurity landscape in 2023? Robert: Cyberattacks continue to grow in quantity and sophistication just as organisation systems become increasingly complex. According to the State of Security 2023, over half of the organisations globally shared they have suffered a data breach in the past two years, an increase from 49% in 2022 earlier and 39% in 2021. As cybersecurity risks continue to evolve, some of the key trends that we anticipate in the Indian cybersecurity landscape in 2023 are: ● Cybercrime-as-a-Service economy will expand the volume and effectiveness of cyberattacks and companies should expect an increasingly hybrid environment. ● Misinformation attacks against businesses will increase considerably as AI technology continues to improve. Deep fakes and other methods that distort reality should be taken into consideration by security leaders in order to avoid reputational and financial losses. ● Supply chain disruptions will continue with under-funded and under-sourced open source technologies being a significant vulnerability. Open source is widely used but has yet to resolve compliance standards. This puts organisations’ supply chain system to risk. ● Blockchain security concerns will increase and cyberspace breaches in the blockchain industry will probably have a huge financial impact. ● Machine Learning (ML) offers greater security but can also act as another vector of attack and cannot be left unsupervised. While ML algorithms recognise data threats and alert possible cyber breach, it will be important to understand these model functions and keep a close watch over them. For organisations to stay ahead in 2023, business leaders need to start leveraging analytics-driven security solutions and unified platforms in order to achieve cyber resilience and future-proof themselves against these ever-evolving threats that become more sophisticated.","excerpt":"“The use of AI could raise compliance and privacy concerns, particularly in industries such as finance and healthcare”","categories":["AI Trends"],"tags":[],"author_name":"Poulomi Chatterjee","publish_date":"2023-05-17T12:00:00","publication_year":"2023","word_count":968,"keywords":["ChatGPT","machine learning","artificial intelligence","AI","ML","RAG","Aim","analytics","generative AI","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","analytics","generative AI","ChatGPT","Aim","RAG","R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/how-splunk-can-save-the-day-for-enterprise-ai\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":41102,"title":"Kali Linux Vs. Linux Mint: Which One Should You Pick?","content":"One of the favourite tools for programmers and ethical hackers to have on their computer is a solid Linux distribution. Primarily, Linux distros are often built on an open-source codebase and generally offer a more customized and personal user experience. One of the biggest questions for any Linux enthusiast has been whether they should pick Kali Linux or Linux Mint. These are two different Linux forks with different use-cases. Which one is most suited for you? About Linux Mint Linux Mint is built to be a drop-in-and-forget replacement for similar OS like Microsoft Windows and MacOS. After installation, the user is not required to set up any additional software, and can simply use it. At its base, Linux Mint is a lightweight fork of Ubuntu, with the latest version of Mint (19.1) being built on the Ubuntu 18.04 Long Term Support codebase. This means that if users install Linux Mint today, they will receive updates and supports until 2023. Linux Mint has been highly customized to offer a plug-in-and-play experience for the user upon booting up for the first time. It has full multimedia support, and an easy to use a suite of system apps such as file manager, desktop manager, media player and more. It also includes a full, open-source office suite out of the box. In addition to the accessible user interface, there is also a highly simple software store for additional functionality. It also has a modern look and feels which will be familiar to anyone using Windows; a big plus for new users. Due to its Linux core, the OS is also highly resilient and requires little to no user maintenance. It is one of the most popular general purpose Ubuntu-based operating systems out today, only lagging behind Ubuntu itself. Moreover, it is the #1 recommendation for anyone looking to get into using Linux-based operating systems, owing to its helpful community and user support. About Kali Linux Kali Linux almost stands on the other side of the spectrum from Linux Mint. While the latter is a general purpose OS, Kali Linux is made for the express purpose of penetration testing. It is built on the Debian codebase and is created by a security firm known as Offensive Security. In a few words, Kali Linux is a distro filled with hacking and cybersecurity-related tools and is made for use by OpSec experts and network admins. While Mint is a general purpose OS usable for any task, Kali is specialized and makes pen-testing and hacking-related tasks easier while making others, such as word processing, more difficult. For example, the set of software on Kali Linux is highly security-oriented, with office applications such as a productivity suite and an Email client conspicuously absent. In addition to this, the user available by default on Kali Linux is ‘root’, making this distro especially unsafe for beginners. A non-root user can cause damage to the computer with certain commands from the CLI. However, a root user with no idea of what s\/he is doing can completely wreck not only the distribution but also the computer as a whole. Due to its nature as a pen-testing platform, Kali heavily optimizes the existing Debian architecture to reduce the number of packets that the system sends and receives. This reduces not only its network footprint but also its vulnerabilities to any attacks. Overall, Kali Linux is highly optimized towards being used by individuals who know what they are doing. Moreover, anyone who even heard of this distro is expressly looking for a penetration testing platform. However, it comes with some pitfalls with respect to Mint. Kali Linux vs. Linux Mint At the end of it, it comes down to not only the user’s preference but also the use-case. Mint has many advantages, being easy-to-use, low-powered, accessible and easily installable. However, it does come with the pitfalls of Ubuntu-based distributions such as network settings being saved or noisy traffic on networks. On the other hand, Kali has a high number of advantages for those looking to use an OS for hacking and penetration testing. It comes with a steep learning curve and is definitely not made for everyone. However, its set of tools and utilities, along with its base architecture security, is paramount to hackers. All in all, it depends on what the user is using it for. In case of looking for a Linux distro similar to Windows in properties and use-case, Linux Mint is recommended. For a robust platform used for penetration testing and hacking, Kali Linux is robust and dependable.","excerpt":"One of the favourite tools for programmers and ethical hackers to have on their computer is a solid Linux distribution. Primarily, Linux distros are often built on an open-source codebase and generally offer a more customized and personal user experience. One of the biggest questions for any Linux enthusiast has been whether they should pick […]","categories":["Deep Tech"],"tags":["Kali Linux","Linux distros","linux mint"],"author_name":"Anirudh VK","publish_date":"2019-06-21T06:31:41","publication_year":"2019","word_count":757,"keywords":["Kali Linux","Linux distros","programming_languages:R","AI","ViT","linux mint","R"],"extracted_tech_keywords":["AI","R","ViT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/kali-linux-vs-linux-mint-which-one-should-you-pick\/","complexity_score":3,"technical_depth":4,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10092771,"title":"How boAt is Rocking the Technology Landscape","content":"In a world where startups are constantly pouring vast amounts of resources into developing their own proprietary tech stack, boAt has distinguished itself by taking a unique approach – i.e. outsourcing its task to multiple tech partners, thereby eliminating the need for an unnecessary overhead burden for the company by developing legacy applications with tech talent. This includes Google, SAP, Shopify and others. The company is also betting big on Generative AI. For instance, to improve customer experience management, which includes tasks like intuition, moderation, intent detection, and sentiment analysis, it has partnered with a technology enabler. boAt did not reveal the name due to a confidential agreement, at the time of discussion. “We have partnered with a company that already has some of these capabilities, who also happen to be working closely with OpenAI to improve them at a suitable cost and scale,” said Shashwat Singh, CIO at boAt, told AIM, at SAP Now India held in Mumbai last month However, last week boAt partnered with Amplify.ai and Meta to enhance the personalised conversational experience for its users. This chatbot has been ideated by Meta’s Creative Shop and built in partnership with Amplify.ai. At the same time, boAt also has a small but strong research and development team of 35 technologists and data scientists, allowing boAt to maintain a lean and mean organisational structure, who work on product R&D and quality. Outsourcing typically involves delegating certain technical tasks such as ABAP or Java scripting development work. However, the overall project and program management, as well as the expertise related to the subject matter and business analysis, are typically kept in-house. Sailing against the wind boAt has its ups and downs. It has been cognizant of scaling its technology stack. For instance, boAt worked with Tally in the earlier days, but it migrated to SAP only a year ago. But why? Singh said that SAP offers better feature functionality compared to its competitors, meeting all their requirements. Secondly, SAP had a strong partner ecosystem, enabling them to find more adaptable and flexible Tier 2 or Tier 3 partners. Lastly, the SAP RISE program offered them the flexibility to focus on their core business processes while outsourcing the management of infrastructure and uptime. Further, he said that Tally lacks a system of stock or inventory management and is primarily focused on accounts, resulting in a lack of integration between inventory and financial systems, making it challenging to assess inventory position. Singh said that there were lesser challenges from a product perspective, but the implementation did have its difficulties: challenges in getting the right business processes digitised on SAP, especially when it came to integrating legacy warehousing tools. But, this was resolved by the team instantly. Singh credited SAP’s project management team for helping them with the implementation and keeping track of the progress of the project. Why SAP, and not others? Singh told AIM that when it comes to scoring correct ERP solutions, SAP stands better than the rest. In boAt’s experience, partner ecosystem and support, as well as feature functionality coverage, are major downfalls for competing ERPs. Besides that, SAP’s tech footprint has increased with solutions like SAP EWM and SAP TM. Comparatively, other ERP solutions lack a strong ecosystem and features on the supply chain side. While it may be cheaper in the short term to develop an in-house ERP, it would become more expensive to manage and maintain it in the long run. Eventually, the cost-benefit of off-the-shelf ERP products will be cheaper than doing something else. Read more: Enterprises Die for Domain Expertise Over New Technologies Navigating data management challenges boAt has already set up a fully functional data lake with SAP data and other data sources, including Google Analytics, Shopify, and unit data. The company intend to traverse the entire information continuum from operational and descriptive reporting to predictive, prescriptive, and insight. It has already gotten all operational reports live on the data lake and are now exploring analytics use cases that can have a direct impact on the top line. Despite all of these efforts, there exists a dearth of historical data for predictive analysis to work. “We have limited historical data, which means we are not yet able to utilise predictive or prescriptive insights. Our primary use case would have been demand forecasting, but given that most of our sales come from the popular e-commerce marketplaces which are extremely fluctuating, demand sensing is not effective in that context,” explained Singh. The sales account for close to 80-85%. Instead, they are focusing on identifying product gaps by analysing the voice of the consumer to help in the new product introduction process. They don’t require a lot of historical data, but they are focusing on big data and leveraging the voice of the consumer since they don’t get first-party data. Singh aims to transform boAt into an intelligent enterprise that relies on insights and data rather than gut decisions.In a world where e-commerce giants are gobbling up vast amounts of data, many enterprises like boAt have limited data to work with. Yet, despite this challenge, boAt has carved out a place for itself as the fifth-largest wearable brand in the world. How did they do it? By taking a unique approach that emphasises marketing and branding besides a significant focus on product R&D and quality.  So if you’re struggling to keep up with the data-driven landscape of modern business, take a page from boAt’s playbook.","excerpt":"boAt’s chose SAP over other ERP providers because of partner ecosystem and support, as well as feature functionality coverage.","categories":["IT Services"],"tags":["enterprises","Google","Shopify"],"author_name":"Shritama Saha","publish_date":"2023-05-04T12:59:39","publication_year":"2023","word_count":909,"keywords":["Go","OpenAI","AI","sentiment analysis","Shopify","RAG","Aim","analytics","Google","generative AI","enterprises","R","Java"],"extracted_tech_keywords":["AI","analytics","generative AI","OpenAI","Aim","RAG","sentiment analysis","R","Go","Java"],"url":"https:\/\/analyticsindiamag.com\/it-services\/how-boat-is-rocking-the-technology-landscape\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":708,"title":"Jigsaw Academy","content":"Company Profile: Jigsaw Academy offers online training in business analytics. Our courses have been created by industry experts who have applied analytics to solve business problems in various domains. You will learn industry best practices, useful tips and tricks from analytics professionals from fields like retail, financial services, FMCG, telecom and healthcare. Individuals: Our courses are designed for professionals. Classes are conducted on weekends. Participants are required to put in additional time during the week but when they do it is completely flexible. They are provided 24*7 access to our virtual lab equipped with the latest analytic tools and many gigabytes of data sets. From finance to retail to betting in horse racing, our case studies are wide and varied. Corporates: Jigsaw Academy offers analytics training workshops for corporates. We have conducted on-premise as well as virtual training programs for some of the biggest analytics companies across India. The “Introduction to Analytics” workshop is an 8 hour program that is designed for managers. The program encourages the use of data analysis and visualization while making business decisions. “Analytics Bootcamp” is a 40-hour virtual program that will turn your employees into analysts. We also offer analytics trainings in the retail, e-commerce and financial services domains as well as R. Colleges: We conduct training programs in some of the leading MBA institutes in the country. Our programs are designed to equip students with the skills that are required at today’s workplace. Our 50-hour “Business Analyst Workshop” will provide students with skills that are sought after by companies such as Genpact, Fractal analytics, WNS analytics, Citibank, HSBC, Infosys, Wipro, TCS  and many others. [media url=”http:\/\/www.youtube.com\/watch?v=F7VJopzuZzo” width=”600″ height=”400″]","excerpt":"Company Profile: Jigsaw Academy offers online training in business analytics. Our courses have been created by industry experts who have applied analytics to solve business problems in various domains. You will learn industry best practices, useful tips and tricks from analytics professionals from fields like retail, financial services, FMCG, telecom and healthcare.   Individuals: Our […]","categories":["AI Trends"],"tags":[],"author_name":"Дарья","publish_date":"2012-08-14T14:23:43","publication_year":"2012","word_count":274,"keywords":["programming_languages:R","AI","RAG","analytics","R"],"extracted_tech_keywords":["AI","analytics","RAG","R","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/jigsaw-academy\/","complexity_score":2,"technical_depth":5,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":1534,"title":"Interview &#8211; Sudarshan Gangrade, Head of Analytics at hoopos.com","content":"Sudarshan Gangrade talks with Analytics India Magazine on his experience with analytics and how it is being used at Hoopos.com. [dropcap style=”1″ size=”2″]AIM[\/dropcap]Analytics India Magazine: What are some of the main tenets (philosophies, goals, attributes) of analytics approach and policies at Hoopos? [dropcap style=”1″ size=”2″]SG[\/dropcap]Sudarshan Gangrade: Analytics is approached as being an inherent part of business. We work with almost all business functions directly with the clear goal to solve business problems. We also aim to enable the business managers to lead analytics for their functions. Our engineering team has thus created a portal where all the functions have real time access to any report they want to view. Similarly, for any analytics solution that we build, we try to create a tool and train the user, so that the manager can then analyse themselves on a continuous basis. AIM: Please brief us about some business solutions you work on and how you derive value out of it. SG: We work across functions: with the online marketing team to optimize the digital marketing spends; with the category team on inventory management, range selection, pricing strategy, with the operations team to improve delivery time and optimize delivery costs; with the marketing team for sales planning and promotion effectiveness. AIM: Please brief us about the size of your analytics division and what is hierarchal alignment, both depth and breadth. SG: We have a lean team consisting of a core team of 2 people, liaisoning with the various business functions. We separately have the Business Intelligence architect in the engineering side who helps create the data structures and views. AIM: Would you like to share any example of an Insight that generated a huge positive impact for your clients? SG: We worked with the operations team to understand the end to end process from consumer placing an order to it getting shipped from our warehouse. Using a combination of time-motion studies and detailed mapping of the processes, we were able to refine the internal processes. This led to a reduction in processing time by about 20% from the day we implemented the new systems! [pullquote align=”left”]Another issue faced is the fact that organizations often end up mixing up technology and analytics, thus making the analytics function less effective.[\/pullquote]AIM: What are most significant challenges you face being in the forefront of analytics space’? SG: Today, there is a good understanding and need for analytics that organizations feel. However, one of the biggest challenges faced is lack of talent that can talk and work closely with business functions. That is the only way analytics can permeate within organizations successfully. AIM: Customer data collection process is still in a very nascent stage in India. How do you see this evolving in the near future? SG: Data collection in the online space will continue to face challenges in the near future given the nascent stage that Indian online consumer behaviour is at. The challenge further is that there is no historical data to be able to extrapolate the offline consumer behaviour data to online behaviour. Having said that, with the kind of advanced web analytics tools today available, ecommerce companies will do well to build their own understanding of shoppers by tying browsing behaviour directly with end conversion data. AIM: What do you suggest to new graduates aspiring to get into analytics space? SG: While you could get away with some amount of general skills 3 years back, today the demand is for specialization. So, make sure you build specialized skills in some part of analytics – be it usage of tools, reporting, statistical techniques, domain expertise etc. Also, try to always solve an analytics problem from the client’s point of view, whose sole aim is to use the insights to solve a business problem. [divider top=”1″] [spoiler title=”Biography of Sudarshan Gangrade” open=”0″ style=”2″] Sudarshan has a wide breadth of experience in Analytics industry – Analytics across the E-commerce domain, Setting up an Analytics and Reporting function within an organization from scratch, Leading an Analytics services Business for Fortune 50 clients, Onsite Client Account management and Leading Delivery on Analytics projects. As Head of Analytics at Hoopos.com, he is responsible for Analytics across all internal functions of the E-Commerce domain (optimizing online marketing spends, supply chain analytics, catalog and inventory management, pricing and promotions optimization etc). Additionally he is involved in business directly through leading the marketing planning and execution. He has done his B.Tech from IIT Kharagpur and Post Graduate in Management from IIM Bangalore.[\/spoiler]","excerpt":"Sudarshan Gangrade talks with Analytics India Magazine on his experience with analytics and how it is being used at Hoopos.com. [dropcap style=”1″ size=”2″]AIM[\/dropcap]Analytics India Magazine: What are some of the main tenets (philosophies, goals, attributes) of analytics approach and policies at Hoopos? [dropcap style=”1″ size=”2″]SG[\/dropcap]Sudarshan Gangrade: Analytics is approached as being an inherent part of […]","categories":["AI Features"],"tags":["Interviews and Discussions"],"author_name":"Дарья","publish_date":"2012-10-07T18:28:45","publication_year":"2012","word_count":749,"keywords":["business intelligence","Go","programming_languages:R","AI","Git","RAG","Aim","analytics","GAN","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","analytics","Aim","RAG","R","Go","Git","GAN","business intelligence","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/interview-sudarshan-gangrade-head-analytics-at-hoopos-com\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10020987,"title":"India Tightens Grip On Big Tech, America Addresses Chip Drought And More In This Week’s Top News","content":"Battle worn big tech was served a shocker as now India joins the global elite in curbing the power of social media giants. The deplatfroming of President Trump attracted criticism from world leaders.  Silicon valley’s role in freedom of speech globally was questioned. Late last month Facebook and Google got in a standoff with the Australian government for a law that needs them to pay to the content creators. The past few months have been troubling for these companies; a regulatory nightmare. But, they suffered another blow from the world’s largest consumer base—India. Indian Internet Usage By The Numbers WhatsApp users: 530 million YouTube users: 448 million Facebook users: 410 millionInstagram users: 210 millionTwitter users: 175 million The recent events leading upto the farmer’s protests and the way the were reported on Twitter didn’t sit well with the Indian government. Twitter was reprimanded earlier this month. Now, India wants to have a better grip on all things internet. From social media to OTT content, the government wants to discourage malicious dissent. According to the new rules, the internet giants will be encouraged to self regulate through a grievance redressal mechanism.A three-level grievance redressal mechanism has been established under the rules with different levels of self-regulation. Level-I: Self-regulation by the publishers;Level-II: Self-regulation by the self-regulating bodies of the publishers;Level-III: Oversight mechanism. The new rules came into immediate effect that will require the likes of Facebook and Twitter to take down unlawful content quicker. “Messaging apps can be asked to provide the identity of the originators of unlawful messages on their platforms,” said IT Minister Ravi Shankar Prasad. Whereas, Streaming services such as Netflix , YouTube and Amazon Prime Video face stricter oversight over content. In respect of news and current affairs publishers are expected to follow the journalistic conduct of the Press Council of India and the Programme Code under the Cable Television Network Act, which are already applicable to print and TV. Hence, only a level playing field has been proposed. https:\/\/twitter.com\/PiyushGoyal\/status\/1364938241237864448?s=20 IBM To Sell Watson Health Division According to reports, IBM is in plans to sell its much hyped AI health ambition—Watson Health. Currently, a full fledged AI for healthcare looks like a distant dream. That’s where Watson Health struggled. The inefficiencies of algorithms, lack of quality data all combine to create more roadblocks for AI in healthcare. Getting a breakthrough is difficult and time consuming. And, IBM’s chief executive Arvind Krishna is wasting no time in slimming down IBM’s businesses to attain decent gains in the coming years. Know more here. Biden Administration Acknowledges Chip Shortage Tune in as President Biden signs an executive order to secure America’s critical supply chains. https:\/\/t.co\/G9G2EtGbHz— The White House (@WhiteHouse) February 24, 2021 On Wednesday, the President of the United States signed an executive order directing a broad review of supply chains for critical materials—from semiconductors to pharmaceuticals and rare-earth minerals—with the aim of spurring domestic production while strengthening ties with allies. Semiconductor shortage especially, has been troubling the worldwide supply chain from cars to computers. According to this new executive order, a 100-day review of supply chains for four areas: semiconductors, used in products from cars to phones; large-capacity batteries used in electric vehicles; pharmaceuticals and rare-earth elements will be conducted. “The American people should never face shortages in the goods and services they rely on, whether that’s their car, their prescription medicines or the food at the local grocery store,” said the President. He also announced that he would seek $37 billion in funding for legislation to supercharge chip manufacturing in the United States as a shortfall of semiconductors has forced U.S. automakers and other manufacturers to cut production. Amidst Chip Crunch, NVIDIA Reports Gains On Wednesday, NVIDIA reported record revenue for the fourth quarter of $5.00 billion, up 6 percent from $4.73 billion in the previous quarter. The company credited Gaming and Data Center platforms for this record revenue. Thousands of companies across the world use NVIDIA AI to create cloud-connected products with AI services. This week, the chipmaker had also announced the release of Transfer Learning Toolkit (TLT), an AI toolkit that abstracts away the AI\/DL framework complexity and enables one to build production quality pre-trained models faster with no coding required. “TLT helps reduce costs associated with large scale data collection, labeling, and eliminates the burden of training models ground up. With TLT, you can use NVIDIA’s production quality pre-trained models and deploy as is or apply minimal fine-tuning for various computer vision and conversational AI use-cases,” stated the company. Twitter Wants You To Get Paid For Your Content Twitter announces “Super Follow”, like Patreon but on Twitter https:\/\/t.co\/5YBmEfgsUn pic.twitter.com\/aY9g1ozoJz— Jane Manchun Wong (@wongmjane) February 25, 2021 Twitter is all set to launch “Super Follow”, a subscription service that will allow users to pay for the content they like.  “It is an audience-funded model where subscribers can directly fund the content that they value most is a durable incentive model that aligns interests of creators and consumers,” said Dantley Davis, Twitter’s head of design and research. The Super Follow option is similar to the Patreon option where the audience can financially support their favorite content provider. Twitter has recently pulled off a Clubhouse like virtual chat rooms with Spaces. Now with “Super Follow” it wants to encourage content creators to up their game.","excerpt":"Battle worn big tech was served a shocker as now India joins the global elite in curbing the power of social media giants. The deplatfroming of President Trump attracted criticism from world leaders.  Silicon valley’s role in freedom of speech globally was questioned. Late last month Facebook and Google got in a standoff with the […]","categories":["AI News"],"tags":["Chip shortage"],"author_name":"Ram Sagar","publish_date":"2021-02-28T10:00:00","publication_year":"2021","word_count":886,"keywords":["Go","funding","programming_languages:R","AI","programming_languages:Go","computer vision","RAG","Aim","Chip shortage","R","ai_applications:computer vision"],"extracted_tech_keywords":["AI","computer vision","Aim","RAG","R","Go","funding","programming_languages:R","programming_languages:Go","ai_applications:computer vision"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/india-big-tech-chip-shortage-top-news\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":44578,"title":"TCS Ignio Vs Infosys Nia Vs Wipro HOLMES: Intelligent Automation Platforms Take Centre Stage","content":"Recently, TCS announced that its artificial intelligence platform Ignio has doubled its year-on-year revenue, and crossed the $60 million. In the previous financial year, the revenue stood at $31 million. With this kind of growth rate, TCS management says that it has set its eyes on making Ignio the fastest growing software product to cross the $100 million mark by the end of the year. Launched in 2014, Ignio was rolled put among a bunch of other software products by TCS, but once the software gained traction among enterprises, it started to be marketed as a standalone product, to differentiate it from TCS’ traditional services model. Combining cognitive capability that mixes AI, machine learning, and advanced automation, Ignio is a unique software product that covers AIOps, workload management, ERP and business operations. As a part of the unit called Digitate, the Ignio platform leverages AI and automates business processes, thus allowing businesses to forecast and prevent issues with agility, assurance and high customer experience. Over 100 large enterprises use Digitate’s cognitive automation software to achieve efficiency at the moment. In the last 12 months, Ignio clinched 52 new deals, bringing the total number of customers to 105. These are mostly global companies, distributed across the retail, manufacturing, telecom and banking and financial services sectors. Ignio, Nia and HOLMES Competing To Capture The Market There has been a heightened focus on AI enabled platforms by large Indian IT companies. The largest among such platforms — TCS (Ignio), Infosys (Nia) and Wipro (HOLMES) are competing in the same space of AI and automation enabled digital transformation. These companies in the past have been famous for their digital services, but now are emphasising on developing software platforms to clients enhance customer processes. To achieve this, companies are marketing their AI platforms separately from their other stack of services. Infosys Nia which has been deployed across the company’s vast service lines, aims to help create operational efficiency through AI. Nia consists of Advanced Machine Learning, Contract Analysis and Chatbot capabilities to help companies get rid of the various problems they face, be it any sector.  Nia has been built on the Infosys’ first-generation AI platform, Mana, and its Robotic Process Automation (RPA) solution, AssistEdge. Together, both these products have created more than 50 clients and 150+ deployments across different sectors, the company reported. On the other hand, Wipro’s HOLMES has been implemented in enterprise operations across 350 of the company’s clients. Holmes integrates AI with automation to reduce operational pain points and accelerate efficiency. Wipro HOLMES has been successfully deployed in data and information-driven verticals, including Banking and Financial Services Institutions (BFSI), Retail, Manufacturing and Telecommunications. Wipro’s management has set a target of achieving $50 million in automation-only deals, according to a media report. Even mid-sized companies have been focusing on developing AI-based platforms extensively. Just recently, Zensar rolled out its set of platforms for its customers that are based on AI to help drive value creation for the customer across various departments including sales, marketing, IT, talent supply chain, HR, collaboration, etc. What Value Are These Platforms Creating? IT infrastructure automation can be challenging and requires huge amounts of servicing in order to manage the processes. This has led to IT companies now offering automation services which also self-learn to make their maintenance easier. Such platforms are driven by pure business demand, wherein many are looking for capabilities to transform processes and make them both agile and intelligent. Platforms like Ignio, Nia and HOLMES link together prediction, recommendation and autonomous execution on one platform, which can be extremely valuable for a customer trying to gain competitive advantage. Such systems are looking to provide rule-based reasoning wrapped with contextual and pre-built intelligence to give more accurate predictions and drastically enhance autonomous decision-making. Overview Over the years, companies have learned to create data strategies and while analytics helped them derive value, yet processes remained complex, and rampant with inefficiencies. What the above platforms are looking to achieve is finding and removing pain points in the processes, or anomalies that exist using AI. Then, they are expedited using automation tools.","excerpt":"Recently, TCS announced that its artificial intelligence platform Ignio has doubled its year-on-year revenue, and crossed the $60 million. In the previous financial year, the revenue stood at $31 million. With this kind of growth rate, TCS management says that it has set its eyes on making Ignio the fastest growing software product to cross […]","categories":["IT Services"],"tags":["AI (Artificial Intelligence)","Automation","ignio","Infosys","Intelligent Agent","intelligent automation for retail","Robotic Process Automation","TCS","Wipro"],"author_name":"Vishal Chawla","publish_date":"2019-08-16T17:00:28","publication_year":"2019","word_count":684,"keywords":["Wipro","Git","ignio","Intelligent Agent","R","artificial intelligence","digital transformation","Automation","intelligent automation for retail","RAG","analytics","machine learning","AI","Robotic Process Automation","TCS","Infosys","automation","Aim","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","analytics","Aim","RAG","R","Git","digital transformation","automation"],"url":"https:\/\/analyticsindiamag.com\/it-services\/tcs-ignio-infosys-nia-wipro-holmes-automation-platforms-india\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10119387,"title":"Microsoft&#8217;s Satya Nadella Says He is Thrilled to be in Thailand, Opens First Datacenter in the Region","content":"Today, Microsoft Chairman and CEO Satya Nadella announced major investments in Thailand’s digital future at the Microsoft Build: AI Day event in Bangkok. The commitments include establishing a new data centre region, the first in the country, providing AI upskilling opportunities for over 100,000 people, and supporting Thailand’s growing developer community. Nadella expressed his enthusiasm for being in Thailand and emphasised the country’s incredible opportunity to build a digital-first, AI-powered future. The announcement builds on Microsoft’s memorandum of understanding with the Royal Thai Government to envision the nation’s digital transformation. Thai Prime Minister Srettha Thavisin, who attended the event, stated that the collaboration with Microsoft is a significant milestone in Thailand’s “Ignite Thailand” vision, which aims to develop the country as a regional digital economy hub. The new data centre region will expand the availability of Microsoft’s cloud services in Thailand, meeting the growing demand from enterprises, local businesses, and public sector organisations. It will also enable Thailand to capitalise on AI’s economic and productivity opportunities. Microsoft also announced initiatives to foster the growth of Thailand’s developer community, such as AI Odyssey, which aims to help 6,000 Thai developers become AI subject matter experts. In recent weeks, Microsoft has made significant investments in Southeast Asia, including a $1.7 billion commitment to advance Indonesia’s cloud and AI ambitions and a $2.9 billion investment in Japan for cloud and AI infrastructure.","excerpt":"The company aims to provide AI upskilling opportunities for over 100,000 people and support Thailand’s growing developer community.","categories":["AI News"],"tags":["Microsoft"],"author_name":"K L Krithika","publish_date":"2024-05-01T11:27:33","publication_year":"2024","word_count":229,"keywords":["Go","API","programming_languages:R","AI","digital transformation","Git","Aim","ViT","GAN","R","Microsoft"],"extracted_tech_keywords":["AI","Aim","R","Go","Git","API","GAN","ViT","digital transformation","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/microsofts-satya-nadella-says-he-is-thrilled-to-be-in-thailand-opens-first-datacenter-in-the-region\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10112009,"title":"The SHAKTI-Man of India","content":"The SHAKTI-man of India, Veezhinathan Kamakoti, the director of IIT Madras, has had a hectic few years. The genius behind India’s first microprocessor has been trying really hard to build semiconductor capabilities in India, and put a stop on cheap imports from China. “This is something we have achieved over the past 10 years; I have completely given up every other thing to work for SHAKTI,” he told AIM in an exclusive interview. Kamakoti started building the SHAKTI microprocessor in 2014, an open processor built on top of the open RISC-V architecture, which is funded by the Ministry of Electronics and Information Technology. He said that there is a need for building hardware that supports building AI models within the country, which led him to take the charge with the SHAKTI microprocessor. For the past few years, Kamakoti has been rallying for an increase in the investments for semiconductor startups, which is still viewed as a risky area for many VCs. “I am only Saraswati, not Lakshmi for these startups,” Kamakoti said, and informed that there are five companies coming out of the IIT Madras lab: Incore, Mindgrove Technologies, Chakra Electronics, Vyoma Systems, and SecurWeave. All of them are building on top of the SHAKTI core, but he is only the mentor for these companies. “It is very interesting to see that we have an indigenous processor ecosystem in India,” he added. Building trust for the chip Since completing his first course in AI in 1992 from IIT Madras, Kamakoti has been finding ways to use AI for social good and solving complex problems within India. This led him to start focusing on hardware and eventually working on SHAKTI, which is also focusing on edge AI use cases. Kamakoti believes that it is rather difficult to build trust in the semiconductor business because investors need to have full faith when investing in these startups. “Though the government is providing full support for improving the semiconductor industry in India, investors are still hasty and cautious when it comes to investing, as it is a very costly affair, and cheaper alternatives are always available. We are now closing the gap,” he added. “Today, Shakti C-class is very stable and we have made multiple chips with it. We are ready to give them to startups as well, which is not a joke,” said Kamakoti highlighting that they are making it open for everyone for democratising processor based SoC (system of chips) design in the country. “Unlike a software company which has a cloud and is developing a software solution with demos, hardware startups take a lot more trust and investment, which the government is promoting, but we need a lot more VCs to take interest as well,” Kamakoti emphasised. BharatGPT can be the Copilot for India Kamakoti was the chairperson of the AI Task Force, which was constituted in 2016 by the Ministry of Commerce and Industry. The AI Task Force report highlighted how AI can be used as an economic transformer for the country. Apart from this, currently he is also a crucial member of the BharatGPT initiative, which was started by IIT Bombay. Highlighting specialised hardware for every field including IoT, network processing, and computing, Kamakoti said that it is also essential for LLMs to be built and catered for specific domains. He believes that this is where BharatGPT is going to play a crucial role. “The initiative requires validation from experts from every domain, and that is what BharatGPT stands for,” he explained. Giving examples of training small models for the legal and medical sectors and how they can assist professionals in those field, Kamakoti said that the open source approach of BharatGPT will help it become the Copilot for different domains. “It is not a search engine, but a technology that is doing a lot of NLP, and giving you contextual output”, which he believes would help people approach every field with a lot more confidence. Reflecting on how there are thousands, if not millions of documents in Indic languages, it is the need of the hour to extract knowledge from them, including historical facts and science. “For example, there is some proof that Newton’s laws were present in our ancient texts at least a 1000 years ago,” and this can benefit the whole of the world. “LLMs are like faithful kids” Talking about generalised models like ChatGPT and Bard, Kamakoti said that LLMs are like faithful kids. “If you give them relevant data, they will do exactly what you want them to.” He said that there is a dire need for data scientists and to focus on extracting and gathering information for tokenising these Indic models. “This would obviously require a lot of money and compute,” highlighting his love for farming, Kamakoti said that while we are also talking sustainability with AI models, they are also very necessary for helping various sectors such as agriculture. “There is a lot of knowledge hidden in rural India, which needs to be brought out,” he added, highlighting that every field, even within the same domain, requires different amounts of data. Focusing on the recent funding of INR 110 crore from Sunil Wadhwani to IIT Madras for building a school for AI and data science, Kamakoti recalled that a reporter asked him if the same money could instead be spent on building 10-12 schools in India. To this, he replied, “I can use AI to enhance the capabilities of the teacher. I don’t want to replace the teacher, but augment them, in the end creating a very good Co-pilot.”","excerpt":"Professor Kamakoti built SHAKTI, which is India’s first open microprocessor on top of the open RISC-V architecture, and is funded by the MeitY.","categories":["AI Features"],"tags":["BharatGPT","ChatGPT","IIT Bombay","IIT Madras","Interviews and Discussions"],"author_name":"Mohit Pandey","publish_date":"2024-02-05T16:00:00","publication_year":"2024","word_count":927,"keywords":["data science","ChatGPT","IIT Bombay","TPU","Go","BharatGPT","AI","IIT Madras","AWS","NLP","Aim","edge AI","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","NLP","data science","ChatGPT","Aim","edge AI","AWS","TPU","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/the-shakti-man-of-india\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10045624,"title":"How AI &#038; ML Is Accelerating The Growth Trajectory At Moglix","content":"Founded in 2015, Moglix, the industrial B2B e-commerce unicorn, provides solutions to over 500,000 MSMEs. Analytics India Magazine got in touch with Sandeep Goel, Senior Vice President, Strategy & Operation, Moglix. An alumnus of the Indian Institute of Science (IISc), he is responsible for devising growth strategies at Moglix. Sandeep comes with over 25 years of experience in the tech industry. “Today, every organisation is developing solutions specific to their needs. Still, portable models will become available over a period where models developed by one organisation will be easily reusable by others, bringing speed to the deployment of AI\/ML solutions,” said Sandeep. AIM: How does your company implement AI\/ML\/Analytics to drive growth? Discuss briefly by citing customer success stories or use cases. Sandeep: At Moglix, we leverage artificial intelligence and machine learning algorithms for the following objectives: Data Cleaning: A clean catalogue is vital to our business. Situations such as missing attributes, duplicate items, or wrong classification can adversely impact our business. AI\/ML helps us keep our catalogue clean.Search: B2C has led the evolution of search use cases, wherein the consumer prefers to search for a product rather than navigating through various categories. The exact use case is starting to apply for the B2B world, and search results must be accurate. Given the language diversity in India, where English has also become Hinglish (a mix of Hindi and English), AI\/ML becomes a powerful tool to run Natural Language Processing (NLP) algorithms to generate accurate search results.Decision Making: Data is the new raw material, and AI\/ML is the mining machine to extract the value. Data-driven decisions are critical for better efficiency, cost optimisation, and governance in B2B commerce scenarios.Efficiency improvement: From mapping customer requirements to our product to predicting potential incorrect addresses, AI\/ML helps us make several decisions. We are able to achieve 4 X better efficiency and catch incorrect addresses upfront to avoid a product return.Cost optimisation: Order processing flow involves various decision points, such as selecting a supplier or logistics service provider, impacting the transaction cost. We are using AI\/ML to make such decisions resulting in cost optimisation.  Better governance: B2B commerce transactions are vulnerable to human errors, both intentional as well as unintentional. Techniques such as optical character recognition (OCR) help us in avoiding such errors to a great extent. AIM: How much percentage of resources are focused on implementing state of the art? Sandeep: At Moglix, our focus is on building products and solutions for our internal consumption and customers. Our experience in supply chain digitisation, master data management, data mining, and procurement analytics is the core of our business. Currently, 10% of our tech workforce is dedicated to working on projects related to AI\/ML. AIM: What kind of skills do you look for while hiring data scientists\/ ML engineers? Sandeep: There are two key aspects of a data scientist: First is the ability to understand and interpret data. This skill usually comes through domain experience, and one has to spend enough time in the industry to acquire this skill.Understanding of AI\/ML tools and algorithms. Colleges and online learning platforms offer various courses. We look to hire people who have expertise in the above two areas. We like first principle problem solvers, and few of our data scientists are homegrown. AIM: What is the major challenge for recruiting Indian talent? Sandeep: The first challenge is the availability of the right talent. The Indian technology vertical has been transitioning from a service mindset to a product mindset. However, the transition has not happened fully, and people often confuse the role of a program manager vs product manager. Indian tech enterprises need product managers who can take end-to-end ownership of the product from conceptualisation to commissioning. While many institutes in the market offer a course in data science, most of them are not equipped to provide an accurate data science experience due to the lack of the right data and scenarios. The second challenge is the stability which has worsened during the pandemic. Work from anywhere has impacted the stability of resources, and the talent war has triggered attrition driven by salary hikes which are not sustainable. AIM: What are the predominantly used programming languages by your data science team? What does the tool stack of your team look like? Sandeep: Some of the programming languages used predominantly by our data science\/analytics team are Java, Scala, PHP, Python, Angular. Our tool stack includes Scikit, SpaCy, RabbitMQ, Cassandra, Mongo, ElasticSearch, and Apache. Other libraries include Pandas, Numpy, FBProphet, NLTK, SciPy, PDFminer, Pytesseract, APScheduler. For database management, we use an SQL server. Also, we use Matplotlib, Seaborn, Plotly for visualisation, and GitHub for tracking. AIM: What kind of AI\/ML deployment challenges does your team face? Do you prefer having your in-house R&D team or outsource the innovation? Sandeep: The right amount of good quality learning data is essential for data scientists to examine and construct models, whether labelled or unlabeled data or search logs. As such, technology integration in the manufacturing supply chain ecosystem is sparse. An application with AI and ML capabilities feeds on user behaviour data to continuously improve itself. Lack of user adoption weakens the feedback loop between the application and the user. Those enterprises that are in the pilot purgatory stages face the challenges of data being spread across silos, functional teams and departments, and the personal devices of users. Integrating the disconnected data fragments is a challenge. Other enterprises have invested in diverse legacy systems and have to navigate multiple solutions to complete a task’s end-to-end cycle. Again, approval workflows and data governance models are, in many instances, not clearly outlined. This gives rise to multiple scenarios of unclean data across their supplier base, customer base, warehousing management systems, and inventory management systems. Innovation can never be outsourced. It has to be driven by an in-house team with a deep understanding of our business. However, we do engage with partners to co-innovate and learn from their experiences. AIM: How do you see the landscape of AI\/ML evolving in India with regards to your domain? Sandeep: The Indian manufacturing sector will soon reach the USD 1 trillion mark, and it will require capabilities of faster decision-making to handle this growth. The data explosion caused by this growth will be impossible for humans to take, and they will need to be acquainted with the technology. AI and ML will play a critical role in creating efficiencies, governance, and decision-making. Areas such as NLP-based search, deep learning-based image processing solutions, or algorithms for prediction and classification will become an integral part of manufacturing and B2B commerce. Today every organisation is developing solutions specific to their needs. Still, portable models will become available over a period where models developed by one organisation will be easily reusable by others, bringing speed to the deployment of AI\/ML solutions.","excerpt":"AI and ML will play a critical role in creating efficiencies, governance, and decision-making.","categories":["AI Features"],"tags":["Interviews and Discussions","ML","unicorns in India"],"author_name":"kumar Gandharv","publish_date":"2021-08-10T11:00:00","publication_year":"2021","word_count":1136,"keywords":["unicorns in India","data science","artificial intelligence","machine learning","AI","ML","NLP","Aim","deep learning","analytics","spaCy","Interviews and Discussions"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","NLP","data science","analytics","Aim","spaCy"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-ai-ml-is-accelerating-the-growth-trajectory-at-moglix\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10058768,"title":"Interview with Tulsee Doshi, Google Head of Product &#8211; Responsible AI &#038; ML Fairness","content":"Artificial intelligence and machine learning are not just fancy words used in tech circles and research labs anymore. As AI and ML have rapidly entered into a variety of sectors – be it healthcare, agriculture or transportation, they have become very much a part of daily conversations. Along with this, the topic of ethics and fairness in AI and ML models have also become a crucial point of discussion globally. But how do we ensure AI products are inclusive, safe, and accountable? What are the metrics that decide their fairness, and how can companies ensure their products do not come with any biases? Analytics India Magazine interacted with Tulsee Doshi, Google Head of Product – Responsible AI & ML Fairness to understand in detail the concerns that surround the topic of AI and ML fairness and break the myths that often come with this controversial issue. Lowering the barrier to entry for developers is important for embedding ethical considerations into product launches Machine Learning Fairness is the practice of ensuring that machine-learning-based products work well for every user and that they do not exclude, stereotype, offend, or cause harm. In her current role, Doshi leads the efforts to ensure that AI products are inclusive, safe, and accountable. This past year, in partnership with VP of Engineering, Google, Marian Croak, she drove the strategy and development of a new centre of excellence to increase both accountability and impact. They brought over 100 researchers and engineers into a single organisation for which Doshi leads prioritisation, alignment, and product landings. She has led over 30 launches across Google teams to make the products more inclusive. Many have required novel solutions that have become published papers and are now accessible to all developers as a part of the Responsible AI Toolkit. Doshi adds, “Lowering the barrier to entry for developers is a critical step to embedding ethical considerations into product launches, and the toolkit is now used across hundreds of pipelines internally and externally.” With these leadership roles, Doshi has surely learnt a lot. Some of her learnings include: Building Responsible AI needs to start with a humble appreciation for the value of lived experience that recognises the importance of diverse expertise and perspectives. She has understood how important it is to meet teams and developers where they are. Every person has a different level of experience, empathy, and awareness and a different set of pressures and tradeoffs in their organisation. Embedding Responsible AI considerations means starting with an understanding of the mental models at play, and to provide the appropriate resources and guidance to build and grow. Enter positive experiences, exclude stereotypes Doshi says the process is ongoing, and still, a lot has to be done to ensure positive experiences. She feels that as an industry, there is a need to work together and share their learnings to ensure people are not excluded or stereotyped. She adds, “One example where I believe Google has intentionally invested in building a superior product experience for all users is the latest Pixel camera, with a project called RealTone. In partnership with photography and skin-tone experts, we have worked to build a camera that truly works for people of colour. We know there is still a lot more work to do, so we are investing in deep research with expertise outside of Google to improve our understanding of skin-tone, and how that reflects in our product. In many ways, RealTone exemplifies what I hope we are building into the ethos of Google product development – a commitment and culture of asking “who else?”, building a deep understanding and empathy, and a product experience that truly works for more users.” How to assess if a model is fair or not? To be clear, there is no single way to assess the fairness of a machine learning model, especially because what fair means depends highly on the context in which a product has been developed and used. Doshi says that for certain types of use cases, a consistent set of metrics can be developed that can measure a particular outcome, but these metrics should also be accompanied with user research and testing. Different approaches for measuring and tackling fairness concerns Doshi points out how Responsible AI organisation builds the understanding, tools, and resources for product teams across Google to be able to build more inclusive products. They are: Raising awareness and setting goals for the organisation: They have created the Google AI Principles, which set the stage for the work they want to see across Google. Tools, resources, and case studies: As they gain insights into different approaches for building more inclusive products and associated pitfalls, they turn these insights into case studies, guidance, and tools for their teams. They also open-source many of these for the broader community. Hands-on support: They have set up a number of channels, including regular office hours, for teams to seek support. In addition, the central Responsible AI org directly partners with teams to problem solve and conduct applied research so that as they develop new types of product experiences – they are considering fairness concerns. Review forums: As they build up the Responsible AI muscle across the organisation, it becomes important to have checks and balances. For this reason, they have also set up an AI Principles review process. Bringing in more perspectives: They have created both internal and external ways for teams to seek feedback and to bring employees and experts into the design and development process. True fairness comes with embedding it in every stage of the product development cycle Doshi feels that fairness should always be a part of the conversation when building AI and ML models that will affect people. Fairness is now becoming a larger part of the conversation because AI and ML are growing in their reach and prevalence; tech leaders have seen the negative repercussions that can be caused when AI and ML fail for certain users and communities. It is also becoming a focal point of discussion as users, communities, and leaders have raised their voices to call for the importance of this work. She states, “To truly ensure fairness at scale would be to embed it in every stage of the product development cycle from the first conception of an idea to collecting the relevant data, to training the model, to deploying and improving the model. It would mean more diverse teams and perspectives as a part of development, and tools, resources, and guidance for every product team to build upon.” Fairness impacts an individual’s quality of life If fairness is not a consideration when developing products, the worst-case scenario can significantly affect an individual’s quality of life, creating meaningful and long-term divides. Doshi points out a real-life example that she ponders over often – the use of AI in the justice system. Courts in the United States have used ML-based models to determine whether or not someone should get bail, whether they should be released early from a sentence, and more. Biased models could incorrectly leave millions incarcerated. It can have disproportionate effects on communities of colour. She adds, “On a different note when I was visiting my family this past summer, we discovered that my mother’s iPhone opens with my face. While we look similar, we look different enough; this is not a worst-case scenario because we are family and because she trusts me to respect her privacy. It’s also not a proven case of bias. But, these types of failures pose security risks, and it is important for us to evaluate our products and ensure that these failures don’t happen predominantly for certain groups and communities.” Personal favouritesFavourite ML\/AI algorithm and why?Something that comes to mind is “Lookout”, an app to support users with low or impaired vision. Recently, I’ve also had the experience of sharing the Lookout app with a family member who has been struggling with reading due to low vision. I’m excited for him to use the feature that scans a newspaper and translates it to text that can be read aloud to him on his phone. This feature is a great example of using ML\/AI in a targeted use case that adds significant value to a user when they need it most, and I’ve also been glad to see how thoughtful the team has been and continues to be in thinking about fairness & equity considerations as they design the product.Top three apps you frequently useGoogle Maps, Headspace, Kindle.Favourite book on ethics and fairness in AI I love Ruha Benjamin’s Race After Technology – it’s an extremely powerful read that has given me much to think about in terms of my own interactions with technology and the systems around me. I would also be remiss not to plug Google’s People & AI Guidebook – a simple and easy-to-use guide to developing more human-centred AI Products.Favourite podcast in AI and ML Interestingly, I haven’t engaged with too many podcasts on AI\/ML – I find that I mostly use podcasts to engage with the news and latest discussions (e.g. The Daily). But, I just listened to an episode called “The Eliza Effect” on 99% Invisible, which I found to be a fascinating discussion of the history of human interaction with chatbots, and what it means to build relationships with technology.","excerpt":"Tulsee Doshi, Google Head of Product – Responsible AI & ML Fairness talks about the concerns that surround the topic of AI and ML fairness and breaks the myths that often come with this controversial issue.","categories":["Global Tech"],"tags":["AI (Artificial Intelligence)","bias","Data Science","Deep Learning","ethics","Fairness in AI","Machine Learning"],"author_name":"Sreejani Bhattacharyya","publish_date":"2022-01-19T16:00:00","publication_year":"2022","word_count":1552,"keywords":["Go","API","Fairness in AI","artificial intelligence","machine learning","AI","chatbots","R","ML","Machine Learning","analytics","Rust","Deep Learning","Data Science","bias","AI (Artificial Intelligence)","ethics"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","analytics","chatbots","R","Go","Rust","API"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/interview-with-tulsee-doshi-google-head-of-product-responsible-ai-ml-fairness\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":44633,"title":"WEBINAR: What’s New In IBM Cognos Analytics — AI Smarts For BI Users","content":"Rapid technological advances in artificial intelligence are reshaping the business landscape and turbocharging performances across almost all business processes and applications, leading to a wave of new innovations. While enterprises are making great strides in their use of data and analytics, AI capabilities can enable faster insight generation and decision making. To register for the webinar, click here. IBM’s latest release of IBM Cognos brings together enterprise-grade reporting and dashboards with compelling visualizations, extensive self-service capabilities, and AI-driven “smarts” that can drive faster and more insightful analysis. Organisations can do more with an AI-infused BI stack which provides an integrated solution for all the managed and self-service needs. Join this webinar on September 17, 2019 at 11 am IST to learn how IBM Cognos Analytics has evolved into the most comprehensive BI and analysis platform which includes smart data discovery, augmented analytics and predictive analysis. IBM’s latest release of its flagship BI and analytics platform covers all the market requirements and leverages AI capabilities – machine learning and natural language processing (NLP) to automate a lot of features for the end user. To register for the webinar, click here. Who Should Attend C-suite leaders looking for an all-in-one platform for managed and self-service needs Business Analysts who want to learn more about data preparation and enterprise reporting with IBM Cognos Data Analysts who want to learn about Cognos Analytics and how it has evolved Business users who delve into self-service BI can find out how the approach to BI is changing with Cognos Product Managers who want a deep dive into Cognos Analytics Who Are The Speakers The webinar will be led by Philippe Mercier – Data Science and AI Sales Leader, IBM Asia Pacific and Nishant Arora, Analytics Marketing Leader at IBM Asia Pacific. With this webinar, you can learn more about Cognos Analytics which makes BI, analytics easier for data explorers and power users. Also, you can acquire the much-needed skills to derive insights by using IBM’s flagship analytics and BI platform. When: September 17, 2019 at 11 am Register here.","excerpt":"Rapid technological advances in artificial intelligence are reshaping the business landscape and turbocharging performances across almost all business processes and applications, leading to a wave of new innovations. While enterprises are making great strides in their use of data and analytics, AI capabilities can enable faster insight generation and decision making.  To register for the […]","categories":["Deep Tech"],"tags":["IBM"],"author_name":"Richa Bhatia","publish_date":"2019-09-12T17:20:48","publication_year":"2019","word_count":343,"keywords":["data science","API","machine learning","artificial intelligence","AI","RAG","NLP","IBM","analytics","GAN","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","NLP","data science","analytics","RAG","R","API","GAN"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/webinar-whats-new-in-ibm-cognos-analytics-ai-smarts-for-bi-users\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":29430,"title":"Industry-Academia Alignment Will Fill The Significant Talent Gap In Analytics, Says Manoj Kumar Of Workplaceif","content":"As part of our theme Analytics Hiring Scenario, we converse with Manoj Kumar, Founder of Workplaceif, an HR analytics startup based in Bengaluru. Before Workplaceif, Kumar worked extensively in business analytics in various premier organisations such as HSBC, Fidelity Investments, Genpact and Tata Consultancy Services. He is also the winner of HR 40 Under40 award in 2017. A thought speaker and a blogger, Kumar is an advisor on future of Work, Workplace Digital Transformation and Workforce Analytics. In this interview, Kumar shed light on how hiring is looked into the field of data analytics, and trends in this lucrative industry. Analytics India Magazine: There is a lot of buzz about analytics, big data and AI around the industry. Do you think these are mere buzzwords or are we actually seeing an analytics revolution in the IT industry? Manoj Kumar: To answer this question, let’s take a step back to understand what’s happening around us today. Humankind has experienced three Industrial revolutions so far – 1.0 was all about mechanical power, 2.0 was about electricity, and 3.0 was about information technology. Today we are undergoing a new transformation where physical, digital and biosphere boundaries are merging – an Industrial Revolution 4.0 which is size, geography, and industry agnostic.  In the digitally transformed world, we leave data footprints for every action that we take, generating an abundance of data. On the other side, the sophistication of pattern recognition algorithms has gone multiple folds and available via cloud-based computing infrastructure at a much cheaper cost.  This trend has led to a dynamic business eco-system where tech-based startups are becoming a threat to the well-established large organizations by offering more personalized and pointed value-added services to the customers. If the outside change is faster than internal, survival of the organization is questionable. Hence, in this complex business situation, organizations are bound to take advantage of emerging technological advances like AI, ML, and Big data to remain relevant for the future. So, this is no more a buzz — it’s the reality. AIM: Do you think Data Scientists are expensive and difficult to find? MK: “Expensive” is a relative word when evaluated in the context of return from the investment. If a data scientist is employed to do the right job, delivering a high-impact outcome, I wouldn’t call it expansive. However, yes, if you hire a data scientist to do a job that can be done by business consulting profile. Data Scientist has been in existence for quite long. Are they still challenging to find? The answer is yes because there is no commonly understood definition of a data scientist. In the absence of defined competency of a data-scientist, an abundance of profiles is tagged with the title, increasing the total population of data scientists and making recruiting conversion ratio tiny. AIM: How does the analytics hiring scenario in Indian companies look like? MK: Analytics industry offers multiple roles – strategic reporting, advanced analytics, visualisation, business consulting and so on. I would say that 50 to 60 percent of the analytics work in India is around advanced and strategic reporting than advanced analytics. So, still, the absolute number of AA hiring may not look large. However, it’s picking up very fast. Another vital role that is surfacing is an Analytics Translator role – a person who understands business and analytics both, and can tell decision-science based narrative in English that makes sense to the business. I call “translator” as one of the most critical roles in Analytics. AIM: Do you think there is an imbalance between the available talent and required skill set in the analytics industry? MK: Yes, globally and local both. Today, most of the organisations are looking for data-science talent but not found easily in the market. The industry is evolving from IT to digital and current trend requires different expertise. It’s evident that we have more IT talent than Data science. We need programs that can upskill\/align the current abilities to future talent. AIM: We see institutes advertising that their students get five or ten times their current salaries by switching to a career in analytics. Do you think that’s just an advertising gimmick or does industry appreciates the analytics talent more? MK: Analytics talents are paid relatively well, and it has been in existence for long. It’s not something new today. However, Institutes have to be careful selling those dreams to students. They need to think about innovative ways of building analytics talent in the industry. Everyone can’t be a data scientist and should not be. Analytics industry requires all sorts of skills as explained earlier and they need to prepare talent for all not only data science. AIM: What are the skill sets that companies are mostly looking at while hiring analytics talent? MK: While hiring both – hard and soft skills are taken into consideration. However, the weight may vary depending on the role, industry, and the organisation. Hard skills may include large data management, finding a pattern, hard-core coding, statistics, visualisation while soft skills may consist of problem-solving, collaboration, and narrative building. AIM: Do you think a postgraduate degree or a specialisation course provides an advantage for getting hired? MK: It depends on the role. However, yes, postgraduates bring-in a lot of structured thinking and ability to connect multiple dimensions of the business. Analytics is all about solving unstructured business problems in a structured manner. The PG or specialised courses equip talent to appreciate business context, manage stakeholders, build the right narrative, and re-deployable frameworks. AIM: What are the various initiatives that companies and educational institutions can take to set right the analytics talent flow? MK: Refer to industry employment reports, and you will learn that many organisations are not able to find the right talent for the analytics roles. It demonstrates a gap between curriculum being taught at institutions versus what sort of talent corporates need. Many initiatives are already at play.  One such initiative is a collaboration between industry and academia. Academicians need to align their curriculum to what industry is want. On the other side, enterprises should start taking active participation in building talent by delivering courses at the institutes or via online channels. AIM: What are the three must-have skills that you look for while hiring a candidate? MK: I give more credit to soft skills, learnability, adaptability and solving unstructured problems in a structured way. When I say learnability, it means the ability to learn new skills to match to industry practices. Adaptability is more about understanding the culture, adopting new ways of working, and collaborating with others to deliver an effective outcome. AIM: What would you advise freshers who are looking to start a career in analytics? MK: Analytics is not about data science only. There are many more roles that you may aspire to opt. It is essential to understand your strength and interest, and if possible align them. Initial few years of your career should be devoted to trying different things before you nail down to one area where you want to build your specialisation. AIM: What is your advice to experienced professionals, who are now looking to transition into analytics. MK: Stop thinking if Analytics career is achievable or not. I have seen people becoming successful analytics professional coming from IT, business, and non-stats background. Take the first step and start your journey. There are many resources available for you to test the appetite for the different roles. AIM: Would you like to share any interesting experience you have had during interviewing an Analytics Professional? MK: We had a candidate for a BI role, who was aspiring to transition from IT to Analytics. He wasn’t sure if he would fit in analytics due to no statistics knowledge. I too wasn’t sure about his long-term career path, but we decided to figure out as we go. Fast forward, he is an M\/L talent today. He used his IT skills to develop himself into an ability which is most in demand today. On this journey, he unlearned many things and learned new skills. I think that was the key to his success. Most of the time, we don’t take the first step forward assuming that it will not happen. Today, organisations like those talent most, who bring learning ability, adaptability, and right behaviour on the table. Hard skills are trainable.","excerpt":"As part of our theme Analytics Hiring Scenario, we converse with Manoj Kumar, Founder of Workplaceif, an HR analytics startup based in Bengaluru. Before Workplaceif, Kumar worked extensively in business analytics in various premier organisations such as HSBC, Fidelity Investments, Genpact and Tata Consultancy Services. He is also the winner of HR 40 Under40 award […]","categories":["AI Features"],"tags":["analytics career","analytics profession","analytics skills","Hiring","Interviews and Discussions"],"author_name":"Abhishek Sharma","publish_date":"2018-10-22T04:46:42","publication_year":"2018","word_count":1390,"keywords":["big data","data science","Go","AI","ML","Hiring","Git","analytics career","analytics profession","Aim","analytics","GAN","analytics skills","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","ML","data science","analytics","Aim","R","Go","Git","big data","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/industry-academia-alignment-will-fill-the-significant-talent-gap-in-analytics-says-manoj-kumar-of-workplaceif\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10011306,"title":"This AI Media-Tech Startup Is Taking The OTT Industry By Storm","content":"AI and ML have revolutionised businesses across verticals, and the scenario in the OTT industry is no different. While it has been used in cases such as recommendation systems, content personalisation and more, this Mumbai-based startup is using it for tasks such as automating meta-data inside video content. Taking the media industry by storm, Toch is enabling a first-of-its-kind media tech platform, while using techniques such as object detection and facial recognition. It also relies on advanced technology while acting as an aggregator for advertisers, media companies and startups to achieve success in digital videos across all platforms through its first-class insights. Built on AI and blockchain technology, Toch aims to bring out the best of videos in online monetisation, engagement and e-commerce. For instance, in the sporting area, it identifies the crowd’s preference and helps OTT brands to identify the right opportunities to present ads based on their preference, thus assisting broadcasters to increase their monetisation through ad sales. We got in touch with the co-founder and CEO of Toch, Vinayak Shrivastav to understand the technology behind the company, which is also co-founded by Saket Dandotia and Alok Patil. AI At Toch Toch, which began with a simple idea of bridging the glaring gap between content and commerce, found an increased number of use cases surrounding the usage of metadata. In the process of making video interactive and improvising post ideation, Toch transformed into a full-fledged media-tech platform while using AI for automating meta-data inside video content. It uses AI and deep learning technology to detect objects, people, emotion, location and activity seen in the video content in real-time. Shrivastav believes that data analytics helps OTT platforms to gain a better understanding of users and their preferences, and AI helps automate most of the manual work processes. This has proven to be time, cost, and labour-efficient for OTT platforms. “These technologies are ever-evolving, and undeniably they are set to transform the industry and help it achieve greater progress in the next few years,” he said. Shrivastav further added that as an entirely cloud-based platform, Toch strives to safeguard the content rights of media platforms at all times. Explaining the tech behind it, Shrivastav said that driven by proprietary technology, Toch runs on various custom modules. “We need to analyse each frame and extract data that translates to metadata — the platform auto-analyses metadata present in any video content using OCR, activity detection, and speech recognition. Toch key moments and highlights are actively used within various major sporting events. The platform has grown by 30% month on month, and we work with some marquee clients in the media industry across the globe,” he said. Further, Toch integrates the API in any existing product to extract videos and process metadata as required for different use cases. Toch also uses its AI-based learning system for image and audio recognition, logo detection, and activity detection to enable live commerce from online video content. Pandemic has further led to increased adoption of technology and automation in the media industry. “Several platforms have gravitated towards automated processes that help save cost and time. Fortunately, Toch has been on the positive side of the global crisis, and we have managed to maintain our business continuity and continue to innovate using leading-edge technology,” said Shrivastav. Toch is currently meta-tagging for live sporting events, live shows and library-based content for streaming platforms and is venturing into broadcasting meta-tagging simultaneously. Growth Story & Future Plans Shrivastav is quick to add that they have witnessed rapid growth since its inception. “There has been a 30% growth month on month, and we are adding more globally recognised clients from the media industry to our portfolio,” he said. He further added that they are looking to expand their footprint to foreign markets as the international market is warming up. “Over the next few years, our primary goal is to establish our presence in 4 different countries with a notable market share in our key target markets,” said Shrivastav on an ambitious note. Apart from that, Toch is looking to launch two new products in the market for users by this year-end. The growth in terms of funding also looks promising. Early this year they raised over USD 1 million in funding from investors including SOSV, and its accelerator, Chinaccelerator, 9Unicorns Fund, and Inflection Point Ventures. “We are currently using the funds to expand our team and enhance our tech infrastructure,” said Shrivastav on a concluding note.","excerpt":"AI and ML have revolutionised businesses across verticals, and the scenario in the OTT industry is no different. While it has been used in cases such as recommendation systems, content personalisation and more, this Mumbai-based startup is using it for tasks such as automating meta-data inside video content. Taking the media industry by storm, Toch […]","categories":["AI Startups"],"tags":["AI Startups"],"author_name":"Srishti Deoras","publish_date":"2020-11-06T17:00:54","publication_year":"2020","word_count":744,"keywords":["Go","AI","ML","recommendation systems","Git","Aim","deep learning","object detection","analytics","R","AI Startups"],"extracted_tech_keywords":["AI","ML","deep learning","analytics","Aim","recommendation systems","object detection","R","Go","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-startups\/this-ai-media-tech-startup-is-taking-the-ott-industry-by-storm\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10040311,"title":"Can You Teach Machines To Forget Memories?","content":"How often have you found some important files archived in some weird folder of your computer? Why do we humans tend to forget things? What good is it? Unlike the human brain, a computer is going to remember the folders and documents you feed it—unless hit by a virus. Regardless of its forgetfulness, the human brain functions quite well. This is because our brain tends to remember what it deems essential. Maybe forgetfulness is a feature and not a bug. Now, what if machines began to work that way too? Facebook has a plan Facebook’s AI team has announced, through a blog post, a new tool called Expire-Span, which would allow neural networks to forget at scale. Why would Facebook do this? Computers will keep storing your information for a long time—even if it is something as irrelevant as notes for a class test you gave five years ago. However, a user wanting to feed further information to a machine would then mean that AI systems would need to expend more computing resources to handle the increase in data, which is unsustainable in terms of cost and lowers the efficiency of these machines. Previous attempts at solving this problem involved compressing information to make it smaller in size, which tackles the need to reduce data but distorts the data at hand(think: image pixelation). This is where Facebook’s Expire-Span comes into play. According to the team, Expire-Span would help build an environment that would enable the alleviation of these problems in machines and AI without compromising the quality of important information. About Expire-Span Source: Facebook AI\/Blog Source: Facebook AI\/Blog Expire-Span works by first predicting information that it deems most relevant to the task at hand. It does this based on contextual information from data and by memories surrounding this information. From this, it places expiration dates on the piece of information based on its relevance. Thus, more helpful information will stay on longer than information that is not as vital. Once a piece of information reaches past its expiry date, it disappears from the AI system. This opens up memory space which can then be used to process the retained important information more efficiently (i.e. faster and at a much larger scale). An example provided by Facebook was of a model learning to perform a word prediction task. In such a scenario, the AI could be taught to remember rare words such as names but forget filler and trivial words like ‘as’ and ‘of’. Expire-Span would then use context to predict what words belong to the trivial filler category and what terms will be remembered. The model has also been shown to be efficient. Facebook tested it compared to previous models, Adaptive-Span and Compressive Transformer, on context-based tasks such as language modelling and moving objects and found Expire-Span to be faster and more efficient (based on the load on GPU memory) than the other models. One very brainy AI This whole body of work on Expire-Span was inspired by how the human brain functions. When AI is presented with too much data, it undergoes the risk of being overwhelmed by its sheer volume. Expire-Span is designed to operate more like a brain and only remember what is important. Of course, the human brain is much more complicated than a machine learning algorithm. For one, the human brain has different types of information that it feels is important. For example, a handmade sweater by your grandmother is going to remain a special memory regardless of how often you wear it. Expire-Span may not possess this sentimentality that comes with humans and may declare something important to the user irrelevant because it has not been used in a while. Another issue with Expire-Span is how Facebook did not mention whether the tool will notify users that it will delete their information permanently. Some people believe that Expiry-Span will not delete one’s data itself but will instead ask the user whether a file can be deleted on the expiration date. Another speculation is whether these files will be deleted or sent to a form of recycling bin instead? As of now, Facebook has not answered these questions. In light of the above concerns, it becomes imperative to note that Facebook is still in the research phase of Expire-Span, and may continue to research and fix any bugs it encounters. Additionally, Facebook’s AI team is also looking into making Expire-Span turn AI more like a brain by incorporating different types of memories into neural networks. This makes for an exciting and innovative venture into the realm of artificial intelligence. Who knows, kids could one day use ‘my computer lost my homework’ as an excuse to get out of homework!","excerpt":"How often have you found some important files archived in some weird folder of your computer? Why do we humans tend to forget things? What good is it? Unlike the human brain, a computer is going to remember the folders and documents you feed it—unless hit by a virus. Regardless of its forgetfulness, the human […]","categories":["IT Services"],"tags":["AI innovations","Facebook AI"],"author_name":"Mita Chaturvedi","publish_date":"2021-05-18T17:00:00","publication_year":"2021","word_count":784,"keywords":["Go","machine learning","artificial intelligence","Facebook AI","AI","neural network","programming_languages:R","programming_languages:Go","ViT","GAN","R","AI innovations"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","neural network","R","Go","GAN","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/can-you-teach-machines-to-forget-memories\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":61739,"title":"Google Releases Xtreme To Induce Development of Multilingual AI Models","content":"Google recently introduced a Natural Language Processing (NLP) benchmark – known as Xtreme – to encourage developments of multilingual AI models. It is a massively multilingual multi-task benchmark for evaluating cross-lingual generalization in NLP models. Building an NLP system, which not only works in English but in all 6,900+ languages, is more complicated than it sounds. According to the researchers, most of the world’s languages are data-sparse and do not have enough data available to train robust models on their own. Still, many languages do share a considerable amount of underlying structure. For instance, words that stem from the same origin, use of postpositions to mark temporal and spatial relations, etc. Why Use Xtreme Several advances have been witnessed in deep learning techniques over the last few years. With an attempt to learn general-purpose multilingual representations, researchers have developed popular models like mBERT, XLM, XLM-R, among others. However, the evaluation of these methods has been mostly focused on a small set of tasks and for linguistically similar languages. To mitigate such issues and encourage more research on multilingual learning, the researchers from Google introduced this benchmark. Xtreme covers 40 typologically diverse languages and includes nine tasks that collectively require reasoning about different levels of syntax or semantics. Behind Xtreme Cross-lingual TRansfer Evaluation of Multilingual Encoders or XTREME is a multi-task benchmark which can be used to evaluate the cross-lingual generalization capabilities of multilingual representations across 40 languages. This benchmark focuses on the zero-shot cross-lingual transfer scenario, where annotated training data is provided in English. Still, no specific language is provided to which the systems must transfer. According to the researchers, the languages in this multilingual benchmark are selected for three main tasks. These are – maximizing language diversity, coverage in existing tasks, and availability of training data. How It Works The goal of this multilingual benchmark is to provide an accessible benchmark to evaluate the cross-lingual transfer learning on a diverse and representative set of tasks and languages. The benchmark consists of nine tasks that fall into four different categories. These are – classification, structured prediction, question-answering and sentence retrieval. The nine tasks are XNLI ( Cross-lingual Natural Language Inference), PAWS-X (Cross-lingual Paraphrase Adversaries from Word Scrambling), POS (Part-Of-Speech Tagging), NER (Named Entity Recognition), XQuAD (Cross-lingual Question Answering Dataset), MLQA (Multilingual Question Answering), TyDiQA-GoldP (Typologically Diverse Question Answering – GoldPgold passage version), BUCC (Building and Using Parallel Corpora) and Tatoeba dataset. The researchers searched for the nearest neighbour using cosine similarity and calculated the error rate. According to them, to evaluate performance using XTREME, models must follow the three-step rule. Firstly, the model must be pre-trained on multilingual text, using objectives that encourage cross-lingual learning. Then they need to be fine-tuned on task-specific English data. Lastly, zero-shot cross-lingual transfer performance on the task is applied in non-English languages. One of the advantages of this zero-shot learning is its computational efficiency, as a pre-trained model only needs to be fine-tuned on English data for each task and can then be evaluated directly on other languages. Wrapping Up The researchers released the benchmark to encourage research on cross-lingual learning methods that transfer linguistic knowledge across a diverse and representative set of languages and tasks. They said, “We hope that XTREME will catalyze research in multilingual transfer learning, similar to how benchmarks such as GLUE and SuperGLUE have spurred the development of deep monolingual models, including BERT, RoBERTa, XLNet, AlBERT, and others.” Furthermore, the researchers also introduced pseudo test sets as diagnostics which cover all the 40 languages by automatically translating the English test set of the natural language inference and question-answering dataset to the remaining languages.","excerpt":"Google recently introduced a Natural Language Processing (NLP) benchmark – known as Xtreme – to encourage developments of multilingual AI models. It is a massively multilingual multi-task benchmark for evaluating cross-lingual generalization in NLP models. Building an NLP system, which not only works in English but in all 6,900+ languages, is more complicated than it […]","categories":[],"tags":["NLP","NLP AI"],"author_name":"Ambika Choudhury","publish_date":"2020-04-16T16:00:00","publication_year":"2020","word_count":604,"keywords":["Go","AWS","AI","ML","RAG","NLP","BERT","deep learning","R","zero-shot learning","NLP AI"],"extracted_tech_keywords":["AI","ML","deep learning","NLP","RAG","zero-shot learning","AWS","R","Go","BERT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/google-releases-xtreme-to-induce-development-of-multilingual-ai-models\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":19197,"title":"Indian Railways Would Soon Start Using AI To Prevent Mishaps Due To Signal Failures","content":"Indian railways have become the latest in the town to have vouched for the importance of artificial intelligence. In a latest development, Indian railways announced to make an effective use of AI to predict signal failures. They aim to reduce the possibilities of signals failing and have undertaken remote condition monitoring of the system, which is a new approach for the largest transporter in the country. The signalling system plays a vital role in the overall functioning of the railways and they completely depend on the its accurate working and the real time information fetched by these signals for their smooth running. In fact, failure of these signals is one of the major reasons for train accidents and delays. According to a parliamentary standing committee report presented in 2016, a total of 239 accidents happened on the network of Indian Railways between 2003 and 2015, 80% of it due to derailment. This could be largely attributed to the current methods involved in the railways where they follow a manual maintenance system and depend on find-and-mix methods rather than predict-and-prevent approach. To bring more efficiency in their overall running, railways are introducing remote condition monitoring using non-intrusive sensors for continuous online monitoring of signals, track circuits, axle counters and their sub-systems of interlocking, power supply systems including the voltage and current levels, relays, timers. This system would keep collecting the inputs on a predetermined interval and send it to a central location. This would ensure keeping a check on any flaws in the signaling system, and detect it on real time basis, allowing a speedy rectification of the error and avoiding possible delays and mishap. The system allows data to be transferred through a wireless medium and with the help of artificial intelligence, predictive and prescriptive big data analytics, this data will be utilised to generate desired inputs. This would enable prediction of signalling asset failures, automated self-correction and informed decisions on intervention strategies. The railways have plans to implement this in two sections of Western Railways and South Western Railway at Ahmedabad-Vadodara and Bengaluru-Mysuru route. Depending on the success of these trails, they will be gradually extended to other sections.","excerpt":"Indian railways have become the latest in the town to have vouched for the importance of artificial intelligence. In a latest development, Indian railways announced to make an effective use of AI to predict signal failures. They aim to reduce the possibilities of signals failing and have undertaken remote condition monitoring of the system, which […]","categories":["AI News"],"tags":[],"author_name":"Srishti Deoras","publish_date":"2017-11-22T06:09:05","publication_year":"2017","word_count":360,"keywords":["big data","artificial intelligence","AWS","AI","cloud_platforms:AWS","programming_languages:R","Aim","ViT","analytics","R"],"extracted_tech_keywords":["AI","artificial intelligence","analytics","Aim","AWS","R","big data","ViT","cloud_platforms:AWS","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/indian-railways-soon-start-using-ai-prevent-mishaps-due-signal-failures\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10166403,"title":"AI Search Will Define the Next Generation of Business—Here’s Why","content":"Whether searching for a resource on Google or looking for a particular favourite food from within an app, the presence of AI-powered searches is perceived nearly everywhere. From specialised AI search engines to advanced platforms designed to replace conventional search engines, the way information is discovered is being reshaped. So, what about AI-powered searches geared towards enterprises? The AI Business Trends 2025 report by Google sheds light on how AI has changed how the world discovers information and the benefits of enterprise search. Enterprise Search Market to Experience a Surge in Growth The enterprise search market size is set to reach $12.9 billion by 2031. As per the Google report, the advanced AI-powered search capabilities now let users seek information in a way that mirrors how they naturally experience the world. The AI-driven search tech includes site search, product search, and customer support self-service search. This is helping organisations enrich and optimise product data catalogues, save significant manual work, and improve conversion and cross-selling efficiency. Prominent companies are already adopting AI-based search capabilities. “Snap (Snapchat) deployed the multimodal capability of Gemini within their ‘My Al’ chatbot and has since seen over 2.5 times as much engagement within Snapping to My Al in the United States,” the report stated. Not just limited to tech companies, hospitals like the Mayo Clinic have also benefited from such capabilities and have given thousands of its scientific researchers access to 50 petabytes worth of clinical data through Vertex AI search, facilitating information access across multiple languages. Benefits of AI-Powered Search for Enterprise Advanced search tools provide immense value to businesses. The report highlights three separate benefits: faster access to data, advanced and intuitive searches, and deeper AI-powered insights. Regarding data access, enterprise search can help employees quickly and efficiently find and utilise internal data, boosting productivity. This should help them with more informed decision-making. When it comes to intuitivity, users and employees can use complex queries and process various data formats (documents, spreadsheets, and multimedia) to get relevant information. One can replace multiple tools with the help of AI-powered searches. The report highlighted that integrating AI agents with enterprise search will elevate knowledge retrieval significantly. These agents are capable of accessing and analysing company data, executing complex tasks, and providing valuable recommendations. Meanwhile, Aashima Gupta, global director of healthcare strategy and solutions at Google Cloud, said, “We expect to see greater adoption of intuitive, contextual search that understands medical terminology, complex vocabulary, and abbreviations—helping relieve administrative burdens for medical professionals while improving patient education and research.” Furthermore, Zac Maufe, managing director of regulated industries at Google Cloud, said, “We expect to see more financial institutions prioritising robust internal knowledge search for their employees, tailored to their specific roles. For example, a loan officer would receive different results than a risk analyst when searching for information about a particular loan application.” “We expect GenAl will continue to transform search in retail, allowing customers to find products using natural language, images, or voice commands to deliver higher quality search results,” said Paul Tepfenhart, director of global retail strategy and solutions at Google Cloud. Hence, it appears that the AI-powered search will impact industries, including finance, retail, and healthcare and life sciences. The benefits of AI-powered search also extend beyond the enterprise. Companies that adopt these tools deliver new levels of service and support to their customers. For instance, Moody’s Corporation uses LLMs from Google Cloud to help employees sift through public documents and the firm’s database to write analyses. This not only improves employee efficiency but also enhances the quality of service provided to Moody’s clients. Evolution of Search and the Path Forward Building a robust search system is a complex task, whether it is for Google or any other company. The report states that before generative AI, enterprise search systems were keyword-based and often delivered irrelevant results, leading to frustrating user experiences. Going forward, businesses can integrate LLMs into their legacy systems to improve search accuracy and relevance. While building AI-powered search systems can be challenging, companies like Google are trying to make it easier. These solutions remove the complexity from search systems, making it easier for companies to implement and benefit from AI-powered search. AI-powered search is revolutionising how businesses operate and interact with their customers. By making knowledge discovery faster, more intuitive, and more relevant, AI transforms enterprise search into a powerful tool for innovation, growth, and enhanced customer service. As AI technology continues to evolve, we can expect more drastic changes.","excerpt":"The AI Business Trends 2025 report by Google sheds light on how AI has changed the way the world discovers information and the benefits of enterprise search.","categories":["AI Features"],"tags":["AI Search Engine","enterprise"],"author_name":"Ankush Das","publish_date":"2025-03-20T18:12:39","publication_year":"2025","word_count":748,"keywords":["Go","enterprise","AI Search Engine","AI","Modal","innovation","AI agents","ViT","generative AI","Rust","GAN","R"],"extracted_tech_keywords":["AI","generative AI","R","Go","Rust","GAN","ViT","innovation","Modal","AI agents"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/ai-search-will-define-the-next-generation-of-business-heres-why\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10000133,"title":"Multi-Cloud Enterprise Architecture Is The Key To Data Portability","content":"Enterprises are moving towards a distributed architecture with the multi-cloud environment and are pushing for integration of various devices. This is evident through the internet of things and network integration that cloud computing has moved away from centralised, large-scale data centres to a more distributed multi-cloud setting. This set-up is usually made up of a network of larger and smaller virtualised infrastructure runtime nodes. In the cloud as a virtualised architecture, virtual machines (VMs) are the core virtualisation mechanism. Virtualising enables the IoT infrastructures to be integrated. These architectures and their setting are known as edge clouds, edge computing or fog computing Portability With Multi-Cloud Architecture While multi-cloud architecture allows a degree of cloud portability at the infrastructure and the DevOps level, interoperability and portability challenges still exist. Today, the cloud edge computing is now pushing computing applications, data, and services away from centralised cloud data centre architectures to the edges of the underlying network. While the core cloud provides a larger view, edge clouds lend localised views (to host services close to or at endpoints of networks), on-device, private cloud-like infrastructures. A SearchCIO article defined interoperability in cloud computing as the means for public and private cloud services to understand APIs, configuration, data formats and forms of authentication and authorisation. The interfaces are standardised, “So that the customer can switch from one cloud service to another with minimal impact on the system as possible.” We have also discussed in an earlier article how microservice architectures strive to break up monolithic single cloud architectures into small deployable services, supported by container architectures. Container services are loosely coupled and independent software components that can be rapidly called and mapped to any business process as required by engineers. Microservices architecture is a new approach to build a single application like a collection of small services, with each application running in its own process and communicating with lightweight mechanisms. These applications can be deployed independently and are supported by a fully automated deployment and orchestration framework. These applications do not require synchronised deployments at fixed times. Challenges In Multi-Cloud Environment The difference in performance metrics in original and the new cloud provider Differences in resources used by the original as opposed to a new cloud provider, for example, firewall issues, network addressing, identity services, Also, there will be differences in security policies and a lack of common cloud APIs How To Reduce Challenges In Multi-Cloud Deployment: Google: Mountain View tech giant Google tied up with VMware to offer cloud portability solutions which enable enterprise developers to deploy web applications across multiple environments and devices. The web giant also led the charge for data portability and kickstarted Data Liberation Front, dubbed as Google’s engineering team with the goal to allow users to move their data in and out of Google products via existing open standards. RightScale: California-headquartered cloud management company RightScale used for streamlining operations across servers, cloud and containers provides a transparent platform designed to work across multiple clouds where users have the flexibility to choose the cloud vendor, to move data stores, and design the architecture and components using their own software library. Cloud Foundry: Cloud Foundry is an open source platform which can help move data among PaaS systems, and container technology such as Docker can help move pieces of applications. This open source project gives users maximum flexibility to avoid vendor lock-in. Outlook Edge clouds have shifted the focus from heavy-weight data centre clouds to more lightweight virtualised resources, but this architecture is not without its own challenges. In this, lightweight virtualisation has emerged as a key challenge. Meanwhile, container technology can significantly advance PaaS technology towards a more distributed heterogeneous clouds by supporting interoperability.","excerpt":"Enterprises are moving towards a distributed architecture with the multi-cloud environment and are pushing for integration of various devices. This is evident through the internet of things and network integration that cloud computing has moved away from centralised, large-scale data centres to a more distributed multi-cloud setting. This set-up is usually made up of a […]","categories":["AI News"],"tags":["AWS","iot enterprise architecture","VMWare"],"author_name":"Richa Bhatia","publish_date":"2018-11-21T20:36:23","publication_year":"2018","word_count":615,"keywords":["VMWare","Go","API","AWS","AI","cloud computing","iot enterprise architecture","ML","docker","microservices","edge computing","DevOps","R"],"extracted_tech_keywords":["AI","ML","cloud computing","docker","microservices","edge computing","R","Go","DevOps","API"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/multi-cloud-enterprise-architecture-key-data-portability\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10042002,"title":"Red Hat’s Neeraj Bhatia On How Open Source Defines The Future Of Technology","content":"Open source has truly democratised technology. Red Hat is one of the top open source companies in the world, creating innovative technologies for over two decades now. Analytics India Magazine caught up with Neeraj Bhatia, Senior Director Sales, Red Hat India\/South Asia to understand more about the open source ecosystem and Red Hat’s role in it. “Digital transformation remains critical to gain a competitive advantage,” he said. Excerpts: AIM: How does open source development lead to better technology? Neeraj Bhatia: The most forward-thinking CIOs and organisations today understand that success in the digital age requires enterprises to sustain a constant innovation momentum instead of innovating on an ad-hoc basis – but coming up with ideas, developing and testing for innovation is not core to every enterprise. More so, enterprises have come to realise they can only do so much by themselves. In today’s digital economy, there is a need for combining efforts for sustaining growth & agility, i.e. open collaboration for open innovation. Open source software is developed in a decentralised and collaborative way, relying on peer review and community production. The open source community is much larger and faster to innovate than any single vendor.  It is the driving force behind much of the technology innovation we see – all of the key mega-trends in technology like blockchain, AI, machine learning are happening because each member of the open source community, through their contribution, is helping to define the future of technology. Over the few years, we have also seen how the open source approach to solving problems has continued to spread beyond the technology world to help combat some of our society’s thorniest issues. As a result, more and more organisations across verticals, ranging from automobiles to banks to consumer products, are now contributing to, distributing and using open source software within their enterprise and product portfolio. AIM: How does Red Hat support the open-source community? Neeraj Bhatia: We believe using an open development model helps create more stable, secure, and innovative technologies. As the largest open source company in the world, we have worked for more than 27 years to invest in open projects and technologies, protect and defend open source intellectual property, and recruit developers who actively participate in open projects across the IT stack. We start with community-built open source software that meets the needs, partially or fully, of our customers. Red Hat builds on these projects, hardens security, patches vulnerabilities, and adds new enterprise features. We then contribute these improvements back to the original project for the benefit of the community as a whole. AIM: How does Red Hat help organisations innovate and transform? Neeraj Bhatia: As always, our focus remains on working with our customers to enable their growth and help them survive & thrive. Red Hat today is well poised to support businesses and governments that have a forward-looking plan to serve their customers wherever they are in their digital transformation efforts i.e modernising their existing application, their IT architecture and infrastructure or innovating with new business models and applications to get the best of what hybrid cloud & containerisation has to offer. Our collaboration with open source communities has helped us bring various innovations to market with products like Red Hat Enterprise Linux, Red Hat OpenShift, Red Hat Ansible Automation Platform, Red Hat Data Services, Red hat Enterprise Linux CoreOS and more. In the current environment, where many of our customers & their developers are working from home, tools like Red Hat CodeReady Workspaces– collaborative Kubernetes-native development solution–enables rapid cloud application development that provides a consistent, secure, and zero-configuration development environment. For organisations looking to get the most out of the existing resources and use technology to extend operations to remote or distributed teams, Red Hat Enterprise Linux provides a scalable, reliable foundation for these efforts. Meanwhile, Red Hat Ansible Automation Platform delivers extensive automation capabilities, making it easier for IT teams with limited manpower to keep services up and running. On the other hand, businesses with services that are in demand need to scale. These IT teams need to effectively bridge existing mission-critical systems and applications with cloud-native services. Red Hat OpenShift provides a standardised platform across the IT landscape; be it on-prem cloud or hybrid, regardless of its underlying hardware or service provider, enabling IT teams to retain existing resources while moving workloads across for scalability or security or speed. AIM: What’s your advice for organisations moving to the cloud? Neeraj Bhatia: An ideal cloud ecosystem should follow open standards and tools to leverage the full potential of hybrid and multi-cloud. Applications should be able to move across environments, have consistent security policies, and automate and manage resources. To leverage the full potential of hybrid cloud strategy, enterprises need to build their hybrid cloud on open standards and tools. For instance, developers need a platform that allows them to design & deploy applications that are responsive to the business needs of today and tomorrow, regardless of the underlying hardware or cloud architectures. It should give them freedom and flexibility to choose from the ever-growing palette of development tools such as programming languages, databases and web servers — to build and test their applications, without worrying about the infrastructure. Red Hat’s cloud portfolio provides enterprises with the ability to build a comprehensive and cohesive cloud infrastructure compatible with a diversity of technologies and topologies. Regardless of the cloud services provider or the need to build a cloud that serves everyone from developers to business application owners, Red Hat’s open hybrid cloud solutions deliver interoperability across the cloud continuum economically without vendor lock-in. AIM: Tell us about Red Hat’s role in developing 5G open infrastructure Neeraj Bhatia: To stay relevant in the ever-changing market landscape telco service providers need to become more agile to take advantage of the capabilities of 5G networks, the Internet of Things (IoT), artificial intelligence, machine learning, and more. Open Telco is the foundation of the modern telco. It refers to the use of open source solutions to foster flexible platforms that provide a strong 5G architecture foundation so that telcos can bring new services to market with superior scalability, security, and efficiency. Red Hat advocates choosing the right technology and associated architecture to lay the foundation for any transformation initiatives. Within the telecom’s context, we have seen the same emerge in the NFV wherein the Telecom operators who aspire to be digital service providers have chosen Red Hat to build their Open hybrid Telco cloud which is ubiquitous in nature and universally can run any workloads from the network, IT and well placed to onboarding emerging application and use cases such as 5G, AR\/VR, Edge Computing,  and IoT In line with this, we are working closely with both Bharti Airtel and Vodafone Idea. By using IBM and Red Hat’s portfolio of hybrid cloud and cognitive enterprise capabilities, Airtel has adopted an open cloud architecture that uses Red Hat OpenStack Platform for all network workloads and Red Hat OpenShift for newer containerized workloads. And for Vodafone Idea, we have enabled its IT and Network applications to run on a common cloud architecture. With the deployment of our Open Universal Hybrid Cloud, the company will position itself to play a significant role in the future deployment of new technologies. AIM: What are Red Hat’s future plans, especially in India? Neeraj Bhatia: The pandemic has further accelerated the digital transformation and created a landscape that will continue to encourage technology adoption in fact at a much rapid pace. In line with this, organisations across verticals are adopting hybrid cloud, automation, integration and development methodologies like cloud-native application development. We are well poised to support businesses and governments wherever they are in their digital transformation efforts, whether customers are modernising their existing application architecture and infrastructure or innovating with new business models and applications. In fact, we see tremendous opportunity for us to further accelerate and power the digital journey in India. We also have a growing focus on mid-market and small and medium scale businesses to support them in effectively transforming digitally.","excerpt":"Open source has truly democratised technology. Red Hat is one of the top open source companies in the world, creating innovative technologies for over two decades now. Analytics India Magazine caught up with Neeraj Bhatia, Senior Director Sales, Red Hat India\/South Asia to understand more about the open source ecosystem and Red Hat’s role in […]","categories":["AI Features"],"tags":["Hybrid Cloud","Interviews and Discussions","Open Source","open source ios","OpenShift","red hat"],"author_name":"Shraddha Goled","publish_date":"2021-06-19T11:00:00","publication_year":"2021","word_count":1343,"keywords":["OpenShift","Go","red hat","Open Source","artificial intelligence","AI","machine learning","open source ios","RAG","Aim","Hybrid Cloud","analytics","edge computing","R","kubernetes","Interviews and Discussions"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","analytics","Aim","RAG","kubernetes","edge computing","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/red-hats-neeraj-bhatia-on-how-open-source-defines-the-future-of-technology\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10121154,"title":"Why is Microsoft Copilot Employees&#8217; Worst Nightmare?","content":"Recently, Microsoft CEO Satya Nadella unveiled Windows’ latest feature: Recall. This isn’t just a keyword search; it’s a semantic search, delving deep into your digital history to recreate moments from the past. “Recall isn’t just about documents,” Nadella explained in an interview with The Wall Street Journal. “It’s about reliving experiences, reimagining the past with clarity and detail.” So, how does this work? Windows PCs begin capturing screenshots of users’ activities, feeding this data into a sophisticated AI model embedded directly within the devices. Through neural processing, every image and interaction will become searchable, even extending to photographs. Understandably, there are some concerns, with Elon Musk, in a characteristic tweet, saying, “This is a Black Mirror episode. Definitely turning this ‘feature’ off.” A Good ‘Recall’ This comes in the backdrop of Microsoft Build, the company’s annual developer conference, which will begin today in Seattle. The company is set to showcase its latest AI projects following OpenAI and Google’s events hosted this month. OpenAI announced the ChatGPT desktop app, powered by the latest GPT-4o model, which changed everything. This AI companion boasts real-time screen reading capabilities, positioning itself as the go-to colleague for assistance in times of need. Earlier, OpenAI CEO Sam Altman had expressed views of AI systems acting like a “senior employee” who can engage with users much like a trusted employee would with a CEO. This includes the ability to push back, reason, and access emails within specified constraints. Altman emphasised that an AI assistant shouldn’t be like an “agent” but like a “senior employee”. He pointed out that a senior employee would question requests that appear illogical, while an agent would unquestioningly follow commands. In the meantime, Google introduced AI Teammate, powered by Gemini, which is aimed at streamlining workflow and communication within teams, minus screen recording. This new feature will significantly reduce workloads by handling mundane tasks and participating in team communications, potentially transforming the employee from a helpful colleague to an overbearing overseer. Forget Managers, the AI Employee is Here As tech giants are announcing AI features for the workplace, in an interesting development, UK-based startup Artisan AI is on a mission to create the most advanced human-like digital workers, called Artisans. So far, they have released Ava, a sales representative Artisan. Ava operates as a business development representative (BDR) who streamlines the entire outbound sales process, requiring only a brief 10-minute conversation for setup. Ava not only formulates the user’s ideal customer profile (ICP) but also navigates a vast database of over 270 million contacts. From meticulous lead research to crafting and dispatching highly tailored email sequences, Ava executes each task with precision, seamlessly booking meetings into sales reps’ calendars. A new era is upon us, with AI employees and humans working in symbiosis.We've just released Ava, The Sales Rep Artisan. She's an AI BDR on steroids, and she's available to hire now.The best part? Manage everything by chatting to her.Learn more: https:\/\/t.co\/Xq15n0B85J pic.twitter.com\/nynI97wfwY— Artisan (@GetArtisanAI) January 15, 2024 They have also recently launched Artisan Labs, their research lab, where they are pioneering the foundational technologies required to create human-like digital workers. Time to be Friends with Your AI Colleague In today’s rapidly evolving landscape, AI’s capacity to comprehend emotions emerges as a powerful asset rather than a cause for concern. Despite initial reservations regarding privacy, these AI systems can be instrumental in enhancing employee well-being. By detecting signs of stress or excessive screen time, managers can intervene proactively, potentially mitigating the risk of burnout or mental health issues. Moreover, in the context of rising concerns about workplace suicides, AI monitoring could provide critical insights into identifying and addressing distressing behaviours among colleagues. Moreover, humans naturally seek connections and understanding from others. In this context, AI’s ability to comprehend emotions can serve as a catalyst for fostering deeper relationships and facilitating a more supportive workplace culture. Recent advancements in robotics and AI have fundamentally altered our perception of technology’s role in shaping social dynamics. By integrating emotional intelligence into AI systems, we not only enhance their efficacy but also redefine them as invaluable allies in promoting employee well-being and fostering positive workplace relationships.","excerpt":"A good ‘Recall’","categories":["AI Features"],"tags":["AI (Artificial Intelligence)","AI Tool","Google","Microsoft","OpenAI"],"author_name":"Vidyashree Srinivas","publish_date":"2024-05-21T16:00:00","publication_year":"2024","word_count":689,"keywords":["Go","ChatGPT","semantic search","OpenAI","Microsoft","AI","GPT-4o","ML","Aim","Google","Rust","AI Tool","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","ML","GPT-4o","ChatGPT","OpenAI","Aim","semantic search","R","Go","Rust"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/why-is-microsoft-copilot-employees-worst-nightmare\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":46631,"title":"India Towards A $ 5 Trillion Super Economy By 2025!","content":"PM Modi’s recent speeches in the US made the intentions of the government very clear— it wants grow India’s economy to double the current size which $5 trillion by the year 2025, making it the fourth largest economy after the US, China and Japan. But, there seems a dissonance in what the government is planning and the situation at hand in the country. The uncertain economic conditions triggered by fears of global recession has made it even more challenging for the Modi government. Nevertheless, it seems PM Modi’s recent speech in the US has highlighted the right vision for the country in terms of the direction in needs to take. So, one might wonder what the government can do to bring the promise of $5 trillion economy. Modi’s Support For Startups Is A Big Move Towards The Target Policy regulation is another area where the government needs to make sure innovation is not hampered as it is the main source of economic growth. Here, the government’s push to ease regulations for startups as well as remove angel tax (after years of pleas) is a move in the right direction. The government’s stance on startups is appreciative and reflects that the government has recognised startups play a critical role in driving innovation and economic growth for the future.  “Our youth are the largest users of the app economy. From food to transport, from movies to hyperlocal delivery, startups are easing everything. Thus if you want to invest in startups in a huge market, come to India,” said PM Modi at Bloomberg Global Business Forum in New York recently. Apart from startups, opening up foreign direct investment for particular sectors, reduction in corporate taxes, ensuring credit momentum to non-banking institutions, taking back the tax surcharge on foreign investors, single tax system, consolidation of public banks are all great steps to make the economy more efficient, say experts. PM Modi at Global Business Forum in New York also recently said “We have removed 50 policy laws hampering development, this is the beginning,” which is reassuring for innovation. Overview To achieve the target of making India Inc a $ 5 trillion economy will be no easy feat especially in an uncertain global economic scenario. The government therefore has to play an active role in making sure the nation garners enough investments which are then put to use for building an innovation-led infrastructure. The massive investments in the infrastructure and business initiatives would naturally create enough jobs in future given there is no skill imbalance. PM Modi’s vision of making India an economic powerhouse will only come together when there are a multitude of right steps taken concurrently in this direction.","excerpt":"PM Modi’s recent speeches in the US made the intentions of the government very clear— it wants grow India’s economy to double the current size which $5 trillion by the year 2025, making it the fourth largest economy after the US, China and Japan. But, there seems a dissonance in what the government is planning […]","categories":["AI Features"],"tags":["Government","Modi","Narendra Modi","PM Modi","Tech Industry"],"author_name":"Vishal Chawla","publish_date":"2019-09-30T18:20:17","publication_year":"2019","word_count":446,"keywords":["Go","startup","AWS","AI","cloud_platforms:AWS","innovation","programming_languages:R","programming_languages:Go","Government","Narendra Modi","R","Tech Industry","Modi","PM Modi"],"extracted_tech_keywords":["AI","AWS","R","Go","innovation","startup","cloud_platforms:AWS","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/modi-government-india-super-economy-2025\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10113572,"title":"So Far, So Good for Aravind Bhai and Perplexity AI","content":"Founder and CEO of Perplexity AI Aravind Srinivas, who had briefly changed his name to Aravind ‘Bhai’, following a friendly banter with Carl Pei, the co-founder and CEO of Nothing, has been on a social media engagement spree. This was noticed ever since the company raised over $70 million this January, from investors, including big tech giants such as NVIDIA, Jeff Bezos, and others. Leaping Ahead of OpenAI A relatively new entrant, Perplexity AI has been working its way through user adoption. After publicly launching in 2023, Perplexity has over 10 million monthly active users with over 1 million users from India alone. While ChatGPT reached 1 million users within a week of its launch, and now has around 180.5 million users, it is not fair to compare both the platforms, considering OpenAI had a first-mover advantage. Way Through Collab The AI-powered answer engine that is going against some of the biggest tech players, especially Google, and has been making strides in collaborations with strategic partners. Recently, Srinivas confirmed that Perplexity will partner with Sarvam AI and similar other startups that are building large language models suited for India’s linguistic background. He believes that such partnerships will mutually benefit both parties, with Perplexity AI utilising the startup model and, in return, providing a large user base to them. Last week, Srinivas announced a partnership with Slack, the communication and messaging platform for organisations. Through ‘Perplexity Push’ bot, Slack users will be able to get real-time updates, summaries and sources in their workspace. Perplexity has also been building strategic partnerships with hardware providers. Last month, Perplexity announced its pplx-online LLM APIs to power Rabbit R1, an AI-powered gadget that uses a large action model (LAM). Source : X Interestingly, with the ongoing tweet exchanges between Srinivas and Pei, a possible partnership between the AI engine and the phonemaker looks likely. Search by Default While Google is busy with its latest most-powerful AI model Gemini Pro 1.5, and renaming its AI chatbot, they also seem to care less about Perplexity. “We are the only ones doing it in a way where users not only look at AI summaries but also care about the richness and diversity on the web,” said Google chief Sundar Pichai. However, that hasn’t deterred Perplexity from following its goal. The company is going about convincing service providers to make it as their default search option. Last month, Arc Browser, a web browser developed by The Browser Company, announced that Perplexity can be set as the default search engine on its browser. Source: X Cami.ai, an AI assistant that can provide answers, images and transcribe audio, announced that they would be using Perplexity AI for providing all up-to-date information. Perplexity is also supporting browsing for Superagents AI, a platform for running AI agents with an API. Superior Brand Packaging Through X Apart from Srinivas being active on X giving his two cents on new announcements, including Gemini Pro 1.5, he has also been sharing various outputs of Perplexity AI that surpassed Google. By pushing such content, Srinivas is packaging Perplexity as a better AI tool than its competitors. He also did not miss the chance to take a dig at Google’s chatbot without taking names and chimed into the OpenAI Sora debate as well. Source: X When Andrej Karpathy left OpenAI, Srinivas was quick to post about Perplexity’s Push Notifications feature, which will aid one to be updated with the latest news. Similarly, he even spoke about the superior quality of curated news on Perplexity, which he claims “is better than X on some days”. Ironically, Perplexity, which cannot generate images at the moment, still has a long way to go in order to catch up with Elon Musk’s X platform. In a recent interview, Musk has confirmed that he is in serious talks with Midjourney. The partnership will possibly bring the text-to-image generation tool to X’s AI platform Grok. Srinivas also spoke about the possibility of X using all the videos of the platform to build something similar to Sora, a feature that is also not available on Perplexity AI. While Perplexity may lack some of the advanced features found on other major tech platforms, such as ChatGPT, Copilot, or Gemini, its demonstrated growth and marketing strategies position it as a promising contender in the AI-engine arena.","excerpt":"Founder-driven marketing strategy of Perplexity AI on X platform has sure created a lot of buzz","categories":["AI Features"],"tags":["Aravind Srinivas","ChatGPT","Gemini","Perplexity AI"],"author_name":"Vandana Nair","publish_date":"2024-02-22T10:24:39","publication_year":"2024","word_count":719,"keywords":["Gemini Pro","ChatGPT","Gemini","TPU","Go","OpenAI","AI","API","Perplexity AI","RAG","Aim","R","Aravind Srinivas"],"extracted_tech_keywords":["AI","ChatGPT","OpenAI","Gemini Pro","Aim","RAG","TPU","R","Go","API"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/so-far-so-good-for-arvind-bhai-and-perplexity-ai\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":5791,"title":"Brocade Collaborates With VMware For SAN Analytics","content":"The collaboration enables the Brocade Fabric Vision Technology to interoperate with VMware vCenter Log Insight 2.0 to help reduce IT Costs by improving operational efficiency Brocade has collaborated with VMware to provide intelligent analytics capabilities for fiber channel storage area networks (SANs) supporting cloud environments, using VMware vCenter Log Insight 2.0.Brocade Fabric Vision technology with VMware vCenter Log Insight enables significant improvement over prior generations of analytics solutions that have attempted to use complicated heuristics and rigid algorithms to improve the signal-to-noise ratio for dealing with multi terabytes of datasets. This provides deeper knowledge of SAN behavior to increase visibility about a variety of issues impacting VM performance and application responsiveness.The Brocade SAN Content Pack can automatically deliver data to Log Insight, thereby reducing the overall mean-time to problem identification. Users are offered predefined dashboards, queries, alerts, and fields that can also be customized to specific environments. SAN events are intelligently classified and filtered to weed out volumes of excess data, “Because IT infrastructures generate massive numbers of alerts on a daily basis, understanding which ones are truly meaningful is typically a manual, painstaking, and time consuming process that can create inefficiencies,” said Jack Rondoni, Vice President, Data Center Storage and Solutions, Brocade. “Administrators can use tools like Brocade SAN Content Pack and VMware vCenter Log Insight 2.0 to effectively manage the shifting infrastructure patterns typical of modern data centers. This enables IT to finetune their environment via more effective root-cause analysis and visibility, enabling more efficient IT operations.”The Brocade SAN Content Pack is available now at no charge and can be downloaded from the Log Insight Marketplace on the VMware Solution Exchange at https:\/\/solutionexchange.vmware.com\/store\/loginsight.","excerpt":"The collaboration enables the Brocade Fabric Vision Technology to interoperate with VMware vCenter Log Insight 2.0 to help reduce IT Costs by improving operational efficiency Brocade has collaborated with VMware to provide intelligent analytics capabilities for fiber channel storage area networks (SANs) supporting cloud environments, using VMware vCenter Log Insight 2.0.Brocade Fabric Vision technology with […]","categories":["AI News"],"tags":[],"author_name":"AIM Media House","publish_date":"2014-06-01T08:02:18","publication_year":"2014","word_count":275,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","RAG","analytics","R"],"extracted_tech_keywords":["AI","analytics","RAG","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/brocade-collaborates-with-vmware-for-san-analytics\/","complexity_score":3,"technical_depth":7,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":17668,"title":"6 Major Data Quality Issues That Haunt Almost All Major Organisations","content":"With the advent of data socialisation and data democratisation, many organisations are organising, sharing and making available the information in an efficient manner to all the employees. While most organisations are profiting by the liberal usage of such mine of information at their employees’ fingertips, others are facing problems with the quality of data being used by them. As most organisations also look at implementing systems with artificial intelligence or connecting their business via internet of things, this becomes especially important. Business analysts determine market trends, performance data, and even present insights to executives that will help direct the future of the company. And as the world becomes even more data-driven, it is vitally important for business and data analysts to have the right data, in the right form, at the right time so they can turn it into insight. The basic model that a company follows when implementing data socialisation is: However, many times, business analysts end up spending the majority of their time focused on data quality. This is a problem because data preparation and management isn’t the business analyst’s’ primary responsibility. But they also don’t need to depend on IT to do it for them either. Some of the most common data quality-related issues faced by analysts and organisations in general are: 1. Duplicates Multiple copies of the same records take a toll on the computation and storage, but may also produce skewed or incorrect insights when they go undetected. One of the key problems could be human error — someone simply entering the data multiple times by accident — or it can be an algorithm that has gone wrong. A remedy suggested for this problem is called “data deduplication”. This is a blend of human insight, data processing and algorithms to help identify potential duplicates based on likelihood scores and common sense to identify where records look like a close match. 2. Incomplete Data Many a times because the data has not been entered in the system correctly, or certain files may have been corrupted, the remaining data has several missing variables. For example, if an address does not include a zip code at all, the remaining information can be of little value, since the geographical aspect of it would be hard to determine. 3. Inconsistent Formats If the data is stored in inconsistent formats, the systems used to analyse or store the information may not interpret it correctly. For example, if an organisation is maintaining the database of their consumers, then the format for storing basic information should be pre-determined. Name (first name, last name), date of birth (US\/UK style) or phone number (with or without country code) should be saved in the exact same format. It may take data scientists a considerable amount of time to simply unravel the many versions of data saved. 4. Accessibility The information which most data scientists use to create, evaluate, theorise and predict the results or end products often gets lost. The way data trickles down to business analysts in big organisations — from departments, sub-divisions, branches, and finally the teams who are working on the data — leaves information that may or may not have complete access to the next user. The method of sharing and making available the information in an efficient manner to all the employees in an organisation is the cornerstone in sharing corporate data. 5. System upgrades Every time the data management system gets an upgrade or the hardware is updated, there are chances of information getting lost or corrupt. Making several back-ups of data and upgrading the systems only through authenticated sources is always advisable. 6. Data purging and storage With every management level in an organisation, there are chances that locally saved information could be deleted — either by mistake or deliberately. Therefore, saving the data in a safe manner, and sharing only a mirror copy with the employees is crucial. “As business users grow frustrated that they can’t get answers when they need them, they may give up waiting and revert to flying blind without data. Alternatively, they may go rogue and introduce their own analytics tool to get the data they require, which can create a conflicting source of truth. In either scenario data loses its potency,” wrote Brent Dykes. If care isn’t taken to avoid incorrect or corrupt data before analysing it for business decisions, the organisation may end up losing opportunities, revenue, suffer from damage to reputation, or even undermine the confidence of the CXOs.","excerpt":"With the advent of data socialisation and data democratisation, many organisations are organising, sharing and making available the information in an efficient manner to all the employees. While most organisations are profiting by the liberal usage of such mine of information at their employees’ fingertips, others are facing problems with the quality of data being […]","categories":["AI Trends"],"tags":["data quality"],"author_name":"Prajakta Hebbar","publish_date":"2017-09-11T12:05:31","publication_year":"2017","word_count":748,"keywords":["Go","artificial intelligence","AI","RAG","ViT","analytics","data quality","Rust","GAN","R"],"extracted_tech_keywords":["AI","artificial intelligence","analytics","RAG","R","Go","Rust","data quality","GAN","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/6-major-data-quality-issues-haunt-almost-major-organisations\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10170036,"title":"NVIDIA-backed CoreWeave to Invest $23 billion on AI infrastructure in 2025","content":"NVIDIA-backed CoreWeave, a cloud computing company, plans to invest $20 to $23 billion in 2025 on AI infrastructure and data center capacity to meet rising customer demands, including those from Microsoft. In its first quarter of 2025 results, the company highlighted continued rapid scaling of its purpose-built AI Infrastructure, including adding new compute capacity, totalling approximately 420 MW of active power and approximately 1.6 GW of contracted power at quarter end. According to the reports, the company’s projected capital expenditure of between $3 billion and $3.5 billion for the second quarter was way above its revenue expectation of $1.06 billion to $1.1 billion. Revenue increased by 420% year-over-year, fueled by strong demand for the CoreWeave Cloud Platform. “Long-term, committed contracts provide strong revenue visibility, attractive unit economics, and enable a success-based approach to capital investments matched to customer contracts,” the company said in its first quarter results. Today, we announced 1Q 2025 earnings, our first quarterly results after going public. We’ve had a strong start to 2025 with major customer wins, industry-leading performance, and accelerating demand for our platform. Read our press release: https:\/\/t.co\/zuk7TwfN31 #AICloud… pic.twitter.com\/dzwHRBZAiW— CoreWeave (@CoreWeave) May 14, 2025 “Demand for our platform is robust and accelerating as AI leaders seek the highly performant AI cloud infrastructure required for the most advanced applications. We are scaling as fast as possible to capture that demand,” said Michael Intrator, CoreWeave’s co-founder and CEO. The company reported that its revenue backlog stood at $25.9 billion as of March 31, with the five-year agreement with OpenAI contributing $11.2 billion. In March, CoreWeave announced a partnership with OpenAI to provide AI infrastructure. The partnership aims to increase OpenAI’s computing capacity for training and delivering its latest models to its hundreds of millions of users worldwide. The cloud computing company announced that it raised $1.4 billion in net proceeds through its initial public offering (IPO). This amount contributes to a total of $17.2 billion in debt and equity raised to support the company’s strategy for advancing the next generation of cloud computing and its future applications in AI. The company has also ramped up investments in data centre and server infrastructure to meet customer demand. Capital expenditures amounted to $1.9 billion in the first quarter. This quarter-over-quarter increase was driven by the timing of new data centre capacity coming online and the introduction of new generations of GPUs. According to Reuters, CoreWeave is also looking to diversify its supply chain amidst the US-China trade war tensions, aiming to minimise the tariffs’ impact. “(Our clients) are looking for really significant investment from us, and we’re trying to make sure that we’re doing it in a way that doesn’t in any way impact our ability to secure the margins that we need in order to run our business,” Intrator told the news agency.","excerpt":"The company’s financial results indicate a strong commitment to developing AI infrastructure, despite the ongoing trade tensions between the US and China.","categories":["AI News"],"tags":["AI Infrastructure","NVIDIA"],"author_name":"Smruthi Nadig","publish_date":"2025-05-15T13:57:15","publication_year":"2025","word_count":467,"keywords":["Go","API","OpenAI","AI","cloud computing","IPO","programming_languages:R","programming_languages:Go","Aim","AI Infrastructure","NVIDIA","R"],"extracted_tech_keywords":["AI","OpenAI","Aim","cloud computing","R","Go","API","IPO","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/nvidia-backed-coreweave-to-invest-23-billion-on-ai-infrastructure-in-2025\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10074938,"title":"Andrej Karpathy Launches ‘Neural Networks: Zero to Hero’ on GitHub","content":"In GitHub’s blog post last week, research scientist Andrej Karpathy launched a course named ‘Neural Networks: Zero to Hero’ that focuses on teaching the basics of neural networks. In a series of YouTube videos, one can code and train neural networks together, with the built Jupyter notebooks that are then captured inside the lectures directory. The lectures are divided into two categories—the first is ‘The spelled-out intro to neural networks and backpropagation: building micrograd’ that discusses on the backpropagation and training of neural networks, assuming the learners have a basic knowledge of Python and a recollection of calculus; the second is ‘The spelled-out intro to language modelling: building makemore’ where a bigram character-level language model is implemented to further complexify into a modern Transformer language model, like GPT. The main focus in the second category would be on introducing torch.Tensor and its subtleties, along with its use in efficiently evaluating neural networks. It would also focus on the overall framework of language modelling that includes sampling, model training, and the evaluation of a loss. Moreover, the lectures would also consist of a set of exercises included in the video description for better understanding of the concepts. With a PhD from Stanford University in computer science, Andrej Karpathy is a computer scientist that specialises in deep learning and computer vision. He joined Tesla in 2017 and served as the Director of AI, along with stints at Open AI. A primary instructor for the first deep learning course at Stanford—‘CS 231n: Convolutional Neural Networks for Visual Recognition’, Karpathy serves as an independent researcher who openly trains on large deep neural networks.","excerpt":"The lectures would consist of a set of exercises in the video description for better understanding of the concepts.","categories":["AI News"],"tags":[],"author_name":"Bhuvana Kamath","publish_date":"2022-09-12T11:51:20","publication_year":"2022","word_count":269,"keywords":["Go","AI","neural network","Git","computer vision","Python","deep learning","GitHub","Jupyter","R"],"extracted_tech_keywords":["AI","deep learning","neural network","computer vision","Jupyter","Python","R","Go","Git","GitHub"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/andrej-karpathy-launches-neural-networks-zero-to-hero-on-github\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10005201,"title":"7 Latest Data Science Job Openings To Apply Now","content":"According to a recent study, amid the pandemic, the overall Data Science and Analytics market has evolved to adapt to the continually changing economic and business environments. This has called for various jobs in the emerging tech space like data science, machine learning, deep learning, data analytics and others. Below here, we jotted down the seven latest data science job openings that candidates can apply. (The list is in no particular order) 1| Data Scientist at Hitachi Location: Bangalore\/ Pune\/ Hyderabad About: The Data Scientist role at Hitachi mainly drives business and industry solutions focused on big data and advanced analytics. The responsibilities include interacting with the customers to understand business objectives and create analytical strategies to help achieve their goals, enhancing data collection procedures to include information that is relevant for building analytic systems, etc. You will also perform machine learning, text analytics, and statistical analysis methods, such as classification, collaborative filtering, association rules, sentiment analysis, topic modelling, time-series analysis, regression, statistical inference, among others. Apply here for Bangalore. Apply here for Pune. Apply here for Hyderabad. 2| Data Scientist at Baker Hughes Location: Mumbai About: As a Data Scientist, your responsibilities include developing self-learning systems that can predict failures and autocorrect based on multiple data sources, gather and analyse data, devise innovative data science solutions and build prototypes to enable the development of high-performance algorithms in scalable, product-ready code. You will also contribute to the development of software and data delivery platforms that are service-oriented with reusable components across teams. A candidate will be required to research and evaluate emerging technology, industry and market trends to assist in project development and more. Apply here. 3| Staff Data Scientist at Baker Hughes Location: Mumbai About: As a Staff Data Scientist, you will lead and work in cross-functional teams to translate algorithms into commercially viable products and services, develop self-learning systems that can predict failures and autocorrect based on multiple data sources. You will also work with the engineering team to incorporate your analysis and solutions. It will also include working with the visualisation team to create intuitive UI and rich UX stories, initiate and propose unique and promising modelling features, develop new and innovative algorithms and technologies, pursuing patents where appropriate, and more. Apply here. 4| Junior Data Scientist at Analytos Location: Bangalore About: As a Junior Data Scientist at Analytos, your responsibilities will include the development of high performance, distributed computing tasks using big data technologies such as Hadoop, NoSQL, text mining and others. You will design and drive the creation of new standards and best practices in the use of statistical data modelling, big data and other optimisation tools. Apply here. 5| Data Scientist at Analytos Location: Bangalore About: As a Data Scientist, you will be collaborating with the technical team to build products, researching concepts, and influencing new ideas to transform retail. Your responsibilities will include the development of distributed computing tasks using big data technologies, deploying data science and technology-based algorithmic solutions to address the business needs for the customers and more. Apply here. 6| Associate Data Scientist – MFG at Noodle.ai Location: Bangalore About: As a Data Scientist at Noodle.ai, you will collaborate with the Noodle team and engineers from the client companies to build a deep understanding of the clients’ business context and then develop, test, and deploy advanced AI models. You will also develop reusable IP to help the company move faster, dive deeper, and work more efficiently by generalising the models, methodologies, and supporting infrastructure that you build. Apply here. 7| Senior Data Scientist at Zendrive Location: Bangalore About: As a Senior Data Scientist at Zendrive, you will be extracting the driving behaviour by analysing location and motion sensor data and separating signal from noise, developing new models of a complex system by correlating location-based trends and events, create A\/B experiments, work with engineers to implement them and interpret their results and more. As a requirement, the candidate must possess a minimum of 5 to 7 years of experience as a data scientist or a minimum of 2 years of relevant experience post PhD in Statistics or Machine Learning as well as Signal Processing. Apply here.","excerpt":"According to a recent study, amid the pandemic, the overall Data Science and Analytics market has evolved to adapt to the continually changing economic and business environments. This has called for various jobs in the emerging tech space like data science, machine learning, deep learning, data analytics and others.    Below here, we jotted down the […]","categories":["AI Hirings"],"tags":["business data sources","Data Science Jobs","Data Scientist Jobs","latest technology in machine learning","multiple classification statistics"],"author_name":"Ambika Choudhury","publish_date":"2020-08-19T19:00:29","publication_year":"2020","word_count":695,"keywords":["business data sources","data science","Data Scientist Jobs","machine learning","Go","AI","sentiment analysis","distributed computing","Data Science Jobs","multiple classification statistics","deep learning","latest technology in machine learning","analytics","SQL","R"],"extracted_tech_keywords":["AI","machine learning","deep learning","data science","analytics","sentiment analysis","distributed computing","R","SQL","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-hiring\/7-latest-data-science-job-openings-to-apply-now\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":21351,"title":"Top 5 Video Games To Improve Your Coding Skills","content":"Video gaming has lured the entire generation, right from kids to the elderly, almost everyone resort to gaming to kill free-time. What if, these video games came with a part of learning! This article looks into the video games that looks into upgrading the coding capability and enhancing one’s ability to code even better with a fun side. The top 5 video games as listed will definitely sharpen the coding ability of analytics and data science enthusiasts who want to improve upon it. What’s more! These games are available for free, with paid versions that can add more features. 1. Code Combat Code Combat is a role-playing strategy game which uses the programming languages Python and JavaScript. Players advance to the successive levels by building code that permits them further into the vast gameplay surrounding the game. The game works only on browsers in Windows and Macintosh (recent OS versions such as Windows 7 & above and MacOS 10.7 & above are preferred for smooth gameplay). It includes single-player as well multiplayer editions. The game was originally developed to teach coding to kids and students, that gradually started gaining attention of the adult generation as well. It is an open-source software with a handful of players contributing to the community to expand the levels. It received positive reviews from critics who hailed the game as a “casual-generic” interactive browser based game. 2. Empire of Code This browser-based game is again targeted to audience who wish to master Python and JavaScript. The theme of space is what makes this game much more interesting, which demands a user to employ strategies and tactics to progress ahead in the game. Further, to ease the pain for non-coders, the developers have also created a variant to understand coding easily. In the game, players can even use algorithms to build a ‘bot’ to manipulate opponent players. What’s more, the game will also be made available on Android and iOS devices in the coming days. 3. Codingame More of an interface-interactive browser-dependent coding game, it offers a plethora of programming languages ranging from Python to Scala to even Perl, to those willing to upgrade their coding skills– be it beginner or advanced coders. The game has a community to even review the code that you have worked on in the game. Like every other game, it even has multiplayer options so that you can work along with peers and friends to compete with other opponents. There is a separate tutorial for machine-learning sleuths too. 4. Code Fights Don’t let the name fool you. Code Fights is not a game per se, but it uses game mechanics to develop a fun coding environment. This website caters to amateurs as well as coding professionals. The interesting fact is that it also supports Julia, which is a new programming language slowly gaining attention in the field of machine learning and artificial intelligence. Furthermore, the website is a catch for tech-companies to scourge for good talent in the coding base. 5. Data Games Data Games is a collection of mini-games that help the players understand data that goes behind any game. Players eventually develop the necessary math and coding skills that is required for data analysis. The website has many games such as the Rock-Paper-Scissors, Inference to predict matches based on sample, Proximity that lets a user shoot a ball close to the target, and many more. Although, the site offers less information, anyone interested to make it big in the data science industry can go through this interactive game to help them establish a familiar foothold.","excerpt":"Video gaming has lured the entire generation, right from kids to the elderly, almost everyone resort to gaming to kill free-time. What if, these video games came with a part of learning! This article looks into the video games that looks into upgrading the coding capability and enhancing one’s ability to code even better with […]","categories":["AI Trends"],"tags":["big data video games"],"author_name":"Abhishek Sharma","publish_date":"2018-02-05T09:25:34","publication_year":"2018","word_count":597,"keywords":["data science","Go","machine learning","artificial intelligence","AI","Python","analytics","JavaScript","big data video games","R","Java"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","data science","analytics","Python","R","JavaScript","Go","Java"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-5-video-games-to-improve-your-coding-skills\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10046242,"title":"IIT Roorkee Launches PG Certification In Cloud Computing And DevOps","content":"Indian Institute of Technology (IIT), Roorkee, and online learning solution provider — WileyNXT have collaborated to launch PG certification in Cloud Computing and DevOps. The program is designed to equip learners with a repertoire of skills required to build a career in any of the numerous fields associated with Cloud Computing. The last date to apply for the programme is 25 Aug 2021, and the seven-month-long program will provide full-fledged training on relevant cloud skills. The first batch of the program is set to commence on 12 Sept 2021. “A profession in Cloud has a bright future ahead. Cloud Computing & DevOps has witnessed a steady increase in its demand across workplaces and industries. It will continue to grow more and more in the coming years. To cater to this demand, it is extremely critical to train aspiring candidates and learners so that they are ready to conquer the workplaces of the future. We are extremely happy to launch this new program in our collaboration with WileyNXT,” said Prof. Sateesh Kumar Peddoju, Department of Computer Science and Engineering, Indian Institute of Technology, Roorkee. Early-stage professionals with at least two years of experience, knowledge of computer systems, programming and debugging, and proficiency in a minimum of one language, are eligible for this program. On successful completion of the course, a joint certificate by Wiley and IIT-Roorkee will be provided. Vikas Gupta, Managing Director, Wiley India, said, “Owing to the advantage of its demographic dividend, India has a great potential in fulfilling the demand for the skilled IT talent. This requires extensive training and skilling of the existing talent and workforce. Today, aspirants with niche digital skills are most sought-after in the booming IT industry. With our new program focused on Cloud Computing & DevOps, we strive to open doors to new avenues of growth and opportunities for the skilled talent in our country.” One can apply for the course here.","excerpt":"On successful completion of the course, a joint certificate by Wiley and IIT-Roorkee will be provided.","categories":["AI News"],"tags":["Cloud Computing","IIT Roorkee"],"author_name":"kumar Gandharv","publish_date":"2021-08-18T11:06:23","publication_year":"2021","word_count":319,"keywords":["IIT Roorkee","programming_languages:R","cloud computing","AI","Git","Cloud Computing","DevOps","R"],"extracted_tech_keywords":["AI","cloud computing","R","Git","DevOps","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/iit-roorkee-launches-pg-certification-in-cloud-computing-and-devops\/","complexity_score":2,"technical_depth":6,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10048827,"title":"Interview With Sanmay Das, Chair, ACM’s Special Interest Group on AI","content":"“Avoid incrementalism and the desire to measure the value of your research by competitive benchmarks and metrics.”Sanmay Das For this week’s ML practitioner’s series, Analytics India Magazine (AIM) got in touch with Sanmay Das, whose research interests are in designing effective algorithms for agents in complex, uncertain environments and in understanding the social or collective outcomes of individual behaviour. Sanmay is a computer science professor and serves as the chair of the ACM Special Interest Group on Artificial Intelligence and a member of the board of directors of the International Foundation for Autonomous Agents and Multiagent Systems. AIM: What got you interested in AI\/ML? Sanmay: After completing high school in India, I came to the US to attend Harvard University, intending to study Physics, but also excited by the opportunity to take courses in many different subjects. I found myself far more engrossed in my Computer Science classes and had decided that would be my field of study by the end of my first year of college. In my second and third years of college, coursework and research with Profs. Charles Elkan (who was visiting from UC San Diego at the time) and Barbara Grosz really drew me into AI and machine learning. I moved a couple of miles down the river to MIT for graduate school, where I was privileged to work with Profs. Tomaso Poggio and Andrew Lo for my PhD. I delved into research at the intersection between computer science, finance, and economics, which was still somewhat of an unusual combination twenty years ago, but has continued to form the core of my research interests since then. Initially, I was really taken with the idea of understanding intelligence by generating intelligent behaviour (rather than studying systems that already exhibit it). I started working on systems where multiple agents interact, and, even though I am now considerably less optimistic about generating truly intelligent behaviour in my lifetime, I still find the question of how to design agents that can achieve (perhaps more limited) goals in complex, multi-agent environments absolutely fascinating. AIM: Can you tell us about your current role as Professor of Computer Science? Sanmay: It’s a mix of research, teaching, and service (both to my university, George Mason, and to the profession more broadly, for example, through my role as Chair of ACM SIGAI). At a major research university, you spend most of your time on the research side of things, and these days I work actively on ethical and fair AI-enabled allocation of societal resources and on using machine learning for measurement in the social sciences. I work a lot with PhD students and with colleagues all over the world, both on the research and service sides. I also love teaching undergraduate students because I remember how foundational learning at that time was for me and because I feel like there’s something unique that we can contribute through the human contact that takes place in a classroom (ideally physical, but also virtual!). AIM: You have held many roles in your career, including the Chair of ACM SIGAI. What, according to you, should the AI community focus on going forward? What is the consensus amongst the researchers? Sanmay: I don’t think there is any consensus, really. AI is an exceptionally “broad tent” community, and part of its strength is that researchers in many different subfields pursue their research with conviction and enthusiasm. Of course, there’s a lot of focus on deep learning these days, but there are many who think it must be in some way integrated with reasoning, common sense knowledge, causal reasoning, etc., in order to move beyond its current capabilities. Those designing AI systems must think carefully about the societal impacts of the technology, both direct and indirect, and the moral and ethical questions that come up. In my opinion, it would be hard to do that without deep connections to the social sciences and humanities. AIM: Do the research objectives lean more towards financial incentives backed by the startup ecosystem, or is it a pure research purpose? What kind of projects get funded more? Sanmay: In the US, most academic research in computing, including AI, is funded by government entities like the NSF, NIH, and DARPA. These tend to support basic research. In recent years, there has been more funding from some of the bigger tech companies and arrangements where researchers have dual appointments, etc. Faculty and students will often launch or spend time at startups, which are usually inspired by research, but the directionality there typically goes from the research to the startup idea. AIM: What are the frequently picked sub-domains\/subjects within AI by your students and your peers? Sanmay: In terms of undergraduate and masters students, machine learning, in general, is very popular and in-demand these days. Research and PhD programs are more mixed and depend on the strengths of the specific department. Many students within AI and machine learning want to study deep learning, computer vision, and NLP, but there’s also a growing interest in the areas of fairness, accountability and transparency in AI, as well as more broadly in intersections between AI and society. AIM: What do you like and dislike about the current state of the AI research community? Sanmay: I love the fact that AI is a very welcoming, broad tent, and positive community, with an increasing diversity of all kinds within the community. In addition to the intellectually fascinating problems we study, it’s a joy to get to be in a field with so many genuinely great people. I sometimes get frustrated at the level of fragmentation between the subfields. Another major problem we are facing is that with the explosive growth of the area, it has been hard to maintain quality in the traditional vetting and reviewing processes for research, and also just hard to keep up with the volume of papers being published every year. AIM: How do you see the landscape of AI\/ML research evolving in the future, and which domain of AI do you think will come out on top? Sanmay: One of the things I study is prediction and forecasting, and I will forecast that any prediction I could make would probably be wrong! I’ll just note that many areas that seemed “dead” gained new life — for example, before the recent success of deep learning when I was a student, people often used to say that “neural networks are the second-best learning algorithm for any problem” and things like SVMs and ensemble methods were “hot.” So, who knows which subfields will gain special prominence in the coming years! We will likely need results and understanding from all the domains that researchers study and will need to understand which ones were useful and which were not for any given problem. Research builds upon itself, and so every idea that stands the test of time is built on the shoulders of mountains of prior work, both successful and unsuccessful! AIM: What would your advice be to future AI\/ML researchers? Sanmay: Do what you love. Avoid incrementalism and the desire to measure the value of your research by competitive benchmarks and metrics. Solve real problems. Talk to domain experts in the area you would like to influence! Think deeply about the impacts of the artefacts you are studying and building on real people, especially those who may be different from you. AIM: What books and other resources do you recommend for beginners? Sanmay: The Russell and Norvig text is still a classic in general AI. For a conceptual and foundational introduction to machine learning, I love “Learning From Data” by Abu-Mostafa, Magdon-Ismail, and Lin.","excerpt":"Sanmay has worked with the US Treasury department on machine learning approaches to credit risk analysis, and occasionally consults in the areas of technology and finance.","categories":["AI Features"],"tags":["Interviews and Discussions"],"author_name":"Ram Sagar","publish_date":"2021-09-20T12:00:00","publication_year":"2021","word_count":1271,"keywords":["machine learning","artificial intelligence","AI","neural network","ML","computer vision","NLP","Aim","deep learning","analytics","Interviews and Discussions"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","neural network","NLP","computer vision","analytics","Aim"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/sanmay-das-interview-acm-artificial-intelligence\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":40037,"title":"8 Alternatives To PyBrain You Should Know To Build Neural Networks","content":"Python-Based Reinforcement Learning, Artificial Intelligence, and Neural Network (PyBrain) offer flexible, easy-to-use and powerful algorithms for Machine Learning tasks with a variety of predefined environments to test and compare your algorithms. In this article, we list down 8 alternatives of PyBrain one must try in 2019. (The list is in alphabetical order) 1| Azure Machine Learning Azure Mchine Learning is a browser-based workbench for the data science workflow, which includes authoring, evaluating and publishing predictive models. Azure Machine Learning Studio is a collaborative, drag-and-drop tool for building, testing, and deploying predictive analytics solutions on your data. Also, the Azure Machine Learning service provides SDKs and services to quickly prep data, train, and deploy machine learning models. Click here to know more. 2| DatumBox The Datumbox API is a powerful open-source machine learning framework written in Java. It offers a large number of off-the-shelf Classifiers and Natural Language Processing services which can be used in a broad spectrum of applications including sentiment analysis, topic classification, language detection, subjectivity analysis, spam detection, reading assessment, keyword and text extraction, etc. Click here to know more. 3| Google Cloud Machine Learning Google Cloud Machine Learning Engine is a managed service which lets developers and data scientists build and run superior machine learning models in production. It supports popular machine learning frameworks and provides built-in tools to understand the machine learning models. Cloud ML Engine offers training and prediction services, which can be used together or individually. Click here to know more. 4| MLlib MLlib is Apache Spar’s scalable machine learning library. It contains algorithms such as classification, regression, clustering, topic modelling, and other such. The machine learning workflow utilities include feature transformations: standardization, normalization, hashing, machine learning pipeline construction, model evaluation, and hyper-parameter tuning, etc. Click here to know more. 5| OpenCV OpenCV is the open source library for computer vision, image processing, and machine learning which also features GPU acceleration for real-time operation. It is released under a BSD license and has C++, C, Python and Java interfaces. This library is written in optimised C\/C++ and is designed for computational efficiency and with a strong focus on real-time applications. Click here to know more. 6| Sci-Kit Learn Scikit-learn is a Python module for machine learning built on top of SciPy, NumPy, and Matplotlib which provides a number of supervised as well as unsupervised learning algorithms. The library is mainly focused on data modelling. It can be used for classification, regression, clustering, dimensionality reduction, model selection, preprocessing, etc. Click here to know more. 7| Smile Statistical Machine Intelligence and Learning Engine (Smile) is a fast and comprehensive machine learning engine. It can write applications quickly in Java, Scala, or any JVM languages and provides hundreds of advanced algorithms with clean interface. Smile covers every aspect of machine learning which includes classification, regression, clustering, association rule mining, feature selection, manifold learning, multidimensional scaling, genetic algorithm, missing value imputation, efficient nearest neighbor search, etc. Click here to know more. 8| Weka Waikato Environment for Knowledge Analysis (Weka) is a suite of machine learning software written in Java. It is a collection of machine learning algorithms for data mining tasks which contain tools for data preparation, classification, regression, clustering, association rules mining, and visualisation. This system was designed to bring a range of machine learning techniques or schemes under a common interface so that they may be easily applied to this data in a consistent method. It uses a common file format to store its data sets and thus presents the user with a consistent view of the data regardless of what machine learning scheme may be used. Click here to know more.","excerpt":"Python-Based Reinforcement Learning, Artificial Intelligence, and Neural Network (PyBrain) offer flexible, easy-to-use and powerful algorithms for Machine Learning tasks with a variety of predefined environments to test and compare your algorithms. In this article, we list down 8 alternatives of PyBrain one must try in 2019. (The list is in alphabetical order) 1| Azure Machine […]","categories":["AI Trends"],"tags":["Machine Learning","ml libraries","Python Libraries"],"author_name":"Ambika Choudhury","publish_date":"2019-06-03T07:27:47","publication_year":"2019","word_count":606,"keywords":["data science","scikit-learn","artificial intelligence","machine learning","AI","Python Libraries","neural network","ML","Machine Learning","computer vision","OpenCV","ml libraries","analytics"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","neural network","computer vision","data science","analytics","scikit-learn","OpenCV"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/8-alternatives-to-pybrain-you-should-know-to-build-neural-networks\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":39053,"title":"The Consequences Of A Real Life Psycho Pass","content":"While stories of AI ruling over the world and keeping the peace have largely been reserved for the science fiction world until now, they still hold validity. Is it really possible for an algorithm to understand the needs of society and collaborate with existing systems to run society efficiently? A preventive style of governance is not a new concept in terms of science fiction. However, is it possible for an AI to do so without taking away the human concept of free will? Autocracy For AI The most well-fleshed out realisation of the neo-autocratic dystopian vision is a Japanese animated series known as Psycho Pass. The series describes a world set in 2112, where society is run by a system known as the Sibyl System. The Sibyl System can tap into any form of communication possible between the members of the society, and is present almost everywhere. It relies on members of the Public Safety Bureau’s Crime Investigation Division to carry out its wishes. The Sibyl System is later revealed to be a system that can instantly calculate the propensity of a person to conduct a crime. This is what is mainly used to police society, as a person displaying a higher psycho pass is immediately apprehended upon the orders of the system. Apart from this, Sibyl also decides which is the best job for each individual according to the scans, which are given the name ‘cymatic scan’. The existence of Sibyl has removed the need for democracy to exist, and has also made sure that it cannot exist, as the system picks the politicians. In the show, it is said that this is the reason society is advancing so quickly, making the country it is in, Japan, a socio-economic masterpiece. The system hence takes away all semblance of free will in society, owing to its ability to directly detect the state of an individual’s mind. All decisions are made directly by the system, and the system alone. The Fallout On Society Even if humanity evolved enough to let our best algorithm take over our governments, there would still be a multitude of issues. Granted, the ethics of people being ruled over by robots also comes into the question. However, the strongest impact will come in a disruption of the ruler; a transition of man ruling over man to machines ruling over man. While this will be vastly more efficient with regards to societal activities, it will also be highly disruptive. Markets such as international trade, shipping and many others will flow a lot easier due to AI controlling it, and many other disorganized parts of society will come together. If most responsibility is given to the AI, it is also possible that the categorization of individuals into occupations can be done based on their innate talents. This is better in the larger scheme of things, as it ensures a more organized society. However, pitfalls exist. The concept of free will comes under risk when algorithms take over. Even today, free will is being taken away from us by Google and Facebook, with decisions being made on what to show their customers and what not to. Change On The Horizon Just as described in the show, the countries that adopt such systems will see considerable economic and social progress. Owing to the large amount of consolidated power and insights, the operation of many societal institutions would become vastly more powerful. While this will not replace many jobs in the private sector, the rise of an AI government will surely make politics redundant. There is a possibility that there will be political figureheads elected by the system in order to prevent anarchy. Indeed, prevention will be the order of the day, as it will become one of the main tenets of the new society. The characteristic of AI to predict data will come in handy in predicting multiple facets of the world. Integrated sensors and consolidated databases will provide a comprehensive view of what an individual does within the purview of society. While this might seem Orwellian in nature, the concept of individual privacy will be nullified with the omniscience of the AI system. The system does offer multiple benefits over existing governments, but the way it is perceived by the people is what will have to evolve the most. Today, people’s minds and their ‘souls’ are what set them apart from the rest. Due to the innate privacy of our inner selves, we only reveal it to those that are closest to us. The system will lay bare everything humans see and feel in the name of fair governance. This will usher in a reconstruction of society, which might force a change in the way crime is perceived as a whole.","excerpt":"While stories of AI ruling over the world and keeping the peace have largely been reserved for the science fiction world until now, they still hold validity. Is it really possible for an algorithm to understand the needs of society and collaborate with existing systems to run society efficiently? A preventive style of governance is […]","categories":["AI Features"],"tags":[],"author_name":"Anirudh VK","publish_date":"2019-05-13T05:10:21","publication_year":"2019","word_count":791,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","ViT","disruption","GAN","R"],"extracted_tech_keywords":["AI","R","Go","GAN","ViT","disruption","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/the-consequences-of-a-real-life-psycho-pass\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":14814,"title":"Data Whisperer: How this former Amazon basin ethnographer and UC Berkeley lecturer is evangelizing Data Science in India","content":"Chris Arnold (popularly known as the Data Whisperer) has been making data talk since the age of mainframes. Spanning quantitative and qualitative analysis, he has built high functioning quant teams and data mart solutions across pharmaceutical, automotive, and financial services industries; within risk, operations, and marketing domains. He has also conducted ethnographic research in the Amazon basin and focus groups for ‘pay now-die later‘ funeral insurance plans. He currently leads Wells Fargo’s enterprise-wide Knowledge Service practice, with teams in India and the Philippines. Besides, Chris is a Lecturer of Data Visualization in UC Berkeley’s data science program. Chris speaks passionately about data graphic design, intuitive use of complex analytics, and the alleged intelligence of so-called business intelligence tools. Warning: if you don’t have an hour to waste, don’t ask him about pie charts. Chris lives with his family in Namma Bengaluru. Analytics India Magazine had the opportunity to interact with Christopher Arnold to understand his role as the “Data Whisperer” and to paint a more vivid picture about the analytics landscape in India. We present to you the detailed interview as follows. Analytics India Magazine: From your experience, tell us the changes and trends that you have noticed recently in the data analytics space. Christopher Arnold: When I was a young analyst, people struggled to understand what relevance data had in their lives. Using the word “data” was the quickest way to end a social conversation (unless you were with statisticians and actuaries). I think people outside the analytic world still struggle with the exact relevance of data, however, they have a growing awareness that it is important. We hit the mainstream in 2012 with Harvard Business Review’s proclamation that data science was sexiest job of the 21st century. Later that same year, my father excitedly asked if my work had anything to do with “al-go-rithms.” Vindicated after all those years! AIM: You are renowned as the “Data Whisperer.” What does that role of yours essentially entail? How are you evangelizing people about data analytics? Chris at Lees Ferry CA: The phrase comes from the novel The Horse Whisperer about a trainer who has a remarkable ability to understand horses. The same concept plays into a “Data Whisperer.” In addition to all the technical and math training we expect of analysts; a data whisperer is someone who has an intuitive sense of the data; a person who can bring relevance of data across analytic, technological, and business communities. Visualization is an easy place to start. Analysts often create graphics for themselves illustrating their key findings and their discovery journey; they don’t stop to think how the reader will be able to digest and act upon their work. Business often goes into the “wouldn’t it be interesting to see” death spiral of recursive questions trying to prove a pet hypothesis; ultimately the answers to which prove neither interesting nor helpful. Edward “ET” Tufte inspired me to start teaching a class targeting this visualization gap with UC Berkeley’s extension program almost 20 years ago. A few years ago, I moved over to the main campus to teach seminars and helped Annette Greiner develop Cal’s online course. In India, I have taught several hundred university students and professionals in the basics of data visualization, with the intent of giving people the fundamental skills to completely transform the way they think about presenting data. I take an irreverent, humorous approach to engage the audience. The traffic signal at Bangalore’s Marathahalli intersection is now famous throughout India as an example of data visualization. More recently, I seem to have hit a nerve when talking about the lack of analytics in the Human Relations function. I argued that HR is the undiscovered analytic country. Even high-end analytic firms who sell AI are not using basic Social Network Analysis and Text Analytics when it comes to managing talent. Google, of course, is always the exception. It’s a data whisperer’s role to seek out these gaps and fill them. A data whisperer disrupts because they ask a lot of questions: Why did you change scales between these charts? Why did you calculate this out to four places behind the decimal, when your data is only accurate to the ones position? Why did you show a mean for ordinal data? Did you get that pie at the bakery? The intent is not to be annoying, but rather to get people to think about what they are trying to accomplish and to maximize actionability of data. [quote]A data whisperer sees the forest and the trees; she can lead you to the ripe berries and away from the big bad wolf.[\/quote] AIM: What are some of the challenges in recent times that the space has seen? How can those challenges be addressed? CA: Even though I have been an analyst for my entire career (despite not being from West Bengal), I have learned more about analytics and analysts in the last three years than the preceding 20. Learnings come much faster in India, where the scale and dynamism outpace rest of the world. As I lecture and speak at conferences in India, three main challenges emerge from the audiences: I could be a better analyst if I didn’t have to spend so much time managing data: While not exclusively an India problem, I do believe that our scale makes it more critical. Fundamentally, the market does not embrace data management as a competitive advantage. Even CTOs of major corporations have shared with me that their role is to make data accessible. Thus, analysts spend 60-90 percent of their time managing data. Not only do analysts not want to manage data, but they don’t make good DBAs either. My heart sunk when I spoke to an MA in stats from ISI, who had become an expert in ETL. We need to influence the market. Analytics will always lag until we begin thinking of data engineering as key to reaping analytic value. NASSCOM, Analytics India Magazine, state governments, and private groups like the FreakGeeks are positioned to drive this awareness. Our company is interested in analytics but don’t know where to start: Many industry advisers suggest leveraging either existing finance or technology people to build out an analytics function. In my experience, this approach is risky. On one hand, you’ll have people who believe that analysis will be completely unhelpful. On the other, there are those who think analysis can resolve the Pakistan-India conflict. Start small. Hire a single mid-level analyst with 5-10 years of experience. For the first six months, they should just talk to company leaders, partners, customers. Then they should identify a single, core problem that can be solved through analysis. The POC must be delivered within the next three months. If it works, it’s advisable to move from POC into production and then to solution number two. If it doesn’t work, start over. Companies new to analytics need to allow 12-18 months to see results. Our management does not support or understand what analysts can do: You owe it to your company and yourself to try to influence a change. If you are unable to affect that change, leave. But look at it as a learning opportunity and try to understand what you could have done differently so you can take that to your next role. AIM: Going ahead, is there a learning curve that aspirants must master to make it big in the space? CA: From an analytic technology perspective, first go deep then go broad. Find something you are passionate about and consume everything you can on the topic. Get engaged on related projects. Become an established expert. Then move on to another topic and repeat. Never stop. Logistic regression experts rarely make it to Chief Analytic Officer. People who are established in logistic regression, text analytics, non-parametric models, time series, spatial stats, and qualitative methods do. From my perspective, analytic expertise is the price of entry to a much more important game which is the future of business. I have an ongoing friendly debate with Dhiraj Rajaram, the brilliant analytic disrupter. He evangelizes about the democratization of analytics (my words not his), where we should reduce our dependence on highly trained analysts and put powerful analytics in the hands of the non-analytic business user. My counterpoint, is that business has evolved to the point where it is no longer permissible to have non-analytic business users. [pullquote]If you look at businesses that have been heavily involved in data for years like credit risk and direct marketing, you’ll see their leaders grew up as analysts.[\/pullquote] Analysts also show up as leaders in remarkable places, such as Laszlo Block, ensuring that all of Google’s HR decisions are driven by analysis. The traditional model of having data analyst do all the heavy lifting and then leaving the room when decisions are made is past. Not only do you need to understand analytics to an expert level, but you also need to understand business. Wells Fargo’s CDO, Charles Thomas, told a group of student and professional analysts earlier this year, that we “have an obligation to dissent.” We analysts have a unique position in the market and it goes way beyond creating the next best (as my dad says) “al-go-rithm.” AIM: Would you like to leave a word of advice for all aspiring startups in the space who want to make it big, while leveraging a data-driven ecosystem? Chris during NASSCOM GIC 2017 CA: I love meeting with start-ups in India and NASSCOM has created an awesome forum to make this connection. The typical intro the owners give runs like this: Hi my name is insert name. I dropped out of insert one to three Is followed by some other letters. My company does insert something cool. The main issue I see for start-ups is how to break through the clutter. How can they harness the raw intellect and technical genius into something others can use? I read recently that cancer researchers accidentally discovered a cure for a kind of baldness. That’s rare. Start-ups should connect with some of the people who complete their course work at insert one to three Is followed by some other letters to help them identify the problem set, and make the right connections. A broad-based analytic start up, is going to find it tough to break through. But an analytical minded startup solves the data engineering space so that analysts can spend 90% of their time conducting analysis that could go viral.","excerpt":"Chris Arnold (popularly known as the Data Whisperer) has been making data talk since the age of mainframes. Spanning quantitative and qualitative analysis, he has built high functioning quant teams and data mart solutions across pharmaceutical, automotive, and financial services industries; within risk, operations, and marketing domains. He has also conducted ethnographic research in the […]","categories":["AI Features"],"tags":["AI India","business intelligence tools","data India","data science india","Interviews and Discussions","Time Series"],"author_name":"Amit Paul Chowdhury","publish_date":"2017-05-08T12:02:31","publication_year":"2017","word_count":1747,"keywords":["data science","Go","API","data science india","data India","AI","ETL","business intelligence tools","RAG","Aim","data engineering","Time Series","analytics","AI India","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","data science","analytics","Aim","RAG","R","Go","API","data engineering","ETL"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/data-whisperer-former-amazon-basin-ethnographer-uc-berkeley-lecturer-evangelizing-data-science-india\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":6,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10000620,"title":"How Industry-Academia Partnership Gets A Major Boost With British Telecom And IISc Association","content":"Indian tech-market is at an all-time high with numerous multinational technology and communications companies setting up their research centres in India. In 2018, tech giants like Adobe, Nvidia and Intel have announced to set up state-of-the-art artificial intelligence centres of excellence in India. British Telecommunications (BT) has joined the fleet now with plans to work with Indian Institute of Science (IISc) and launch a new collaborative research centre in Bangalore. The BT India Research Centre (BTIRC) will do joint research with the company’s other facilities that are located in China, UAE, USA and Northern Ireland. This global innovation network is centred on BT labs located in Adastral Park in the UK. The BTIRC will work on AI, mobility and cyber-security technologies. IISc Paves The Way For Tech-Hub BTIRC is not only a platform for analysts and scientists to gain industrial experience but it’s also aimed at the development of AI and other cybersecurity platforms for the company’s programmes, products and services. A partnership with a premier institute like IISC will help BT secure more clients in India. According to Howard Watson, CTO, BT,  “In India, over 10,300 people in the technology, service and support areas serve our customers around the world. The BTIRC will create an exciting new hub for communications innovation with IISc, building new collaborative links between the UK and India in this crucial technology sector that is so central to economic growth.” We're partnering with the Indian Institute of Science – one of India’s leading research institutions, to launch a new collaborative research centre in #Bangalore. pic.twitter.com\/BhZdCX3FOY — BT Group (@BTGroup) December 12, 2018 This month, BT team arrived at IISc to announce their project with the Institute. Talking about this industry-academia partnership, Prof Anurag Kumar, Director IISc said, “Collaboration between academia and the industry is essential for making progress in the complex emerging telecommunications technologies. This will be a partnership between one of India’s premier research institutes, and a world leader in telecom technology and services. We at IISc are looking forward to exciting research results and new technologies emerging from this partnership.” The company also aims to work with Indraprastha Institute, Delhi and research on elastic optical networks and quantum key distribution technologies under BTIRC. Is IISc-BT Collaboration An Unorthodox Partnership BT is the largest provider of broadband and mobile services in the UK and the multinational brand has operations in 180 countries. The company first arrived in India in 1987, the first office was established in Delhi in 1995. Since then the organization has set up many plants in India, the list includes Mumbai, Gurugram, Delhi and Bangalore. The company caters to 200 customers across India. Infosys, TechMahindra, NDTV, Wipro are few of the loyal customer base of BT. The objective of the company is to help corporate customers focus on global competition, core-customer business and improve business reliability. These goals are achieved by giving Virtual Private Network (VPN) services using IP-based protocols. Their initiative has been supported by both the Indian and British governments. They partnered with several universities and next-generation telecom network services like IU-ATC (Indian UK Advanced Tech Centre). The union aims to unite tech innovators, business, scientists and policy-makers from UK and India to survey tech governance and data privacy issues. The Cyber Security Operation Centre (CSOC) in Gurugram is one of the leading tech-service providers. It focuses on threat assessment, intrusion detection and prevention, ethical hacking and investigation. Academic-Industry Partnership Will Bolster BT’s Products & Services The rise of different pathways to achieve the best and the fastest technology has led businesses prone to data leaks and ransomware attacks. To keep up with the world, an enterprise must place its data, devices and customers on top and IT plays a crucial role in bringing about a digital transformation. BT helps business in securing their data, transforming system data to the cloud, encouraging collaboration, simplify networks and enhance customer services. BT functions by connecting cloud providers to a core network to deliver secure high-performance connectivity. It invests in the latest network technology and provides dynamic network services, IP management and application acceleration. With the rapid increase in demand for work with real-time data and global online trading, there’s also an increase in demand for a secure and sustainable network. This network can be built by collaborating on the cloud.","excerpt":"Indian tech-market is at an all-time high with numerous multinational technology and communications companies setting up their research centres in India. In 2018, tech giants like Adobe, Nvidia and Intel have announced to set up state-of-the-art artificial intelligence centres of excellence in India. British Telecommunications (BT) has joined the fleet now with plans to work […]","categories":[],"tags":["AI India","IISc"],"author_name":"Jignasa Sinha","publish_date":"2018-12-25T14:02:24","publication_year":"2018","word_count":719,"keywords":["Go","API","artificial intelligence","AI","Git","IISc","RAG","GAN","Aim","ViT","AI India","R"],"extracted_tech_keywords":["AI","artificial intelligence","Aim","RAG","R","Go","Git","API","GAN","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-industry-academia-partnership-gets-a-major-boost-with-british-telecom-and-iisc-association\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":32341,"title":"New AI-Based App Developed By IIT-Ropar Students Can Tell How Drunk You Are By Just Looking At Your Videos","content":"In an interesting development, the far-sighted members of Indian Institute of Technology (IIT), Ropar, have been working on developing an artificial intelligence-based application for smartphones that can help analyse if a person is drunk or not. It would be able to do so by analysing the blood alcohol content (BAC) legal limit. India has become quite notorious for its drunk-driving cases, with reports suggesting that in Goa alone more than 4,920 cases were booked in 2018 compared to 2,760 cases in 2017. Whereas in Hyderabad, more than 20,000 cases of drunk-driving have been filed this year. The Project By IIT Ropar In a report, the assistant professor named Dr Abhinav Dhall from the Department of Computer Science and Engineering at IIT Ropar said that he and his team are developing an Android-based application which can help a person to check how much alcohol is consumed and will be able to drive or not without being analysed by a breathalyzer or some medical tests. The working of the app will be based on the videos of people at different intoxicated stages that will be fed while programming it. The fully developed app will be available in the Google play store and can be downloaded without spending a penny. Dr Dhall has been involved with intense research in the space of computer vision, affective computing, machine learning, automatic driver monitoring and advanced driving assistance systems, over the last few years. What Have IITians Done In The Past To Help Society This is not the first time that the ingenious IIT-ians are doing something fruitful to help the society. Previously, Dr Dhall, along with his two students, had developed an AI-based application for selfie-accident awareness that is free of cost and also available on Google play store. In another development by IIT, the students at IIT Hyderabad earlier this year developed an AI-based programme to detect motorists on the road who are not wearing helmets. This development by C Krishna Mohan, an associate professor along with two other research scholars caught the attention of Hyderabad traffic police and they signed a memorandum of understanding (MoU) with an aim of reducing fatalities of the bike riders and catching hold of the offenders.","excerpt":"In an interesting development, the far-sighted members of Indian Institute of Technology (IIT), Ropar, have been working on developing an artificial intelligence-based application for smartphones that can help analyse if a person is drunk or not. It would be able to do so by analysing the blood alcohol content (BAC) legal limit. India has become […]","categories":["AI News"],"tags":["IIT Bombay","IIT Ropar","Quantum Computer"],"author_name":"Ambika Choudhury","publish_date":"2018-12-28T05:02:31","publication_year":"2018","word_count":368,"keywords":["Go","IIT Bombay","artificial intelligence","machine learning","programming_languages:R","AI","programming_languages:Go","Quantum Computer","computer vision","IIT Ropar","Aim","R","ai_applications:computer vision"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","computer vision","Aim","R","Go","programming_languages:R","programming_languages:Go","ai_applications:computer vision"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/new-ai-based-app-developed-by-iit-ropar-students-can-tell-how-drunk-you-are-by-just-looking-at-your-videos\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":37249,"title":"Information As Second Language Can Bridge Data Literacy Divide In Organisations","content":"As the world moves towards a more data-rich environment, the need for an organisation to understand and fully utilise the capabilities of these data is ever growing. However, as 90% of the data that is currently available is relatively new, one of the main challenges that organisations face is the inability of the workforce to understand, interpret and comprehend data. According to a study by a prominent research organisation which observed the challenges faced by Chief Data Officers, one of the biggest problems large organisations face currently is information language barrier —that is ineffective communication related to data among various stakeholders. “Most executives and professionals do not ‘speak data’ fluently as the new critical capability of the digital era. As a result, data and analytics leaders struggle to get their message across to them,” the study states. How To Overcome This Challenge? As the pertinence of data in business keeps growing steadfastly, the organisation can no longer afford to maintain a workforce who is data illiterate. Hence, across enterprises, Information as a Second Language (ISL) has taken precedence to make workforce well-equipped to communicate and comprehend data. “Digital society demands of its citizen’s data literacy developed for competitive advantage and agility. Data and analytics leaders must follow the example of English as a Second Language (ESL) and treat information as the new second language of business, government, communities and our lives,” the researcher states. As stated by the researcher, here are the three steps that organisations can do to enable ISL in their organisational tiers: Cultivate information as a second language (ISL) across business and IT stakeholders by first establishing the base vocabulary, clarifying industry and business domain “dialects,” and developing levels of proficiency. Drive and sustain improvements to your organisation’s data literacy by identifying areas where data is spoken fluently and where language gaps exist and establish an ISL proof of concept for language development. Change the way you and others interact with leaders, stakeholders and peers by “speaking data” in the context in everyday interactions, board meetings and as a basis for outcomes-oriented business cases. A Challenge For Indian Organisations In India, the situation is no different with organisations plagued with the issue of finding the right talent and skill set with adequate data literacy. In most cases, in order to help their existing workforce be well-equipped in data literacy and tools, enterprises resort to external data training centre’s to help augment the workforces’ skillsets. However, claiming that these traditional data analytics training to be dead, Sayandeb Banerjee, co-founder and CEO of TheMathCompany, says that analytics training is just not enough to augment skill set, “The ubiquity of data coupled with its cheapness and evolving tech has now led to a creation citizen data scientist and it means that data training is not now restricted to small groups, rather it is becoming pervasive. Hence the traditional corporate five data training or online training isn’t not going to work anymore.” Agreeing to the viewpoint, Gaurav Vohra, CEO and co-founder at Jigsaw Academy points out that, to overcome this challenge at the organisational level, ISL have to be a must at every organisation, “Knowledge about data cannot reside in silos within an organisation. Instead, the top-level players need to ensure that each and everyone has access to organisational data and make sure that everyone is part of the digital revolution.”","excerpt":"As the world moves towards a more data-rich environment, the need for an organisation to understand and fully utilise the capabilities of these data is ever growing. However, as 90% of the data that is currently available is relatively new, one of the main challenges that organisations face is the inability of the workforce to […]","categories":["AI Features"],"tags":["data literacy","jigsaw academy"],"author_name":"Akshaya Asokan","publish_date":"2019-04-03T11:22:55","publication_year":"2019","word_count":560,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","Git","data literacy","Aim","analytics","GAN","jigsaw academy","R"],"extracted_tech_keywords":["AI","analytics","Aim","R","Go","Git","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/information-as-second-language-can-bridge-data-literacy-divide-in-organisations\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10062970,"title":"AWS launches game-based courses to teach cloud skills","content":"Amazon Web Services (AWS) has announced two new free training initiatives to provide hands-on cloud computing skills training in a fun and engaging way. The first initiative is a new game-based role-playing experience, called AWS Cloud Quest: Cloud Practitioner, ideal for early-career or new-to-cloud adult learners. AWS also launched a new, improved version of AWS Educate, consisting of added interactive content and removal of the .edu email address requirement, improving the program’s accessibility. Individuals as young as 13 years old can access AWS Educate and make use of hundreds of hours of free, self-paced training, resources, and labs specifically designed for new-to-the-cloud learners. AWS Cloud Quest: Cloud Practitioner’ a cloud learning game: An all-new 3D role-playing game, designed by AWS Training and Certification, to help adult learners gain practical AWS experience. To win, learners must complete quests that simultaneously build cloud skills and help citizens build a better city. Gameplay includes videos, quizzes, and hands-on exercises based on real-world business scenarios. AWS Cloud Quest is available globally in English for personal computers through AWS Skill Builder. AWS Educate: a new approach to learning: A reimagined AWS Educate global program that includes new courses and hands-on labs, making it easier than ever for individuals as young as 13 years old to register. This program is curated for self-motivated, pre-professional learners who are not yet working in the cloud, such as students and job-training participants. The program offers hundreds of hours of free, self-paced training and resources—including more than 50 courses and 10 hands-on labs in the AWS Management Console—so learners can practice their skills. New features include: Four new courses: Cloud Computing 101, AWS DeepRacer Primer, Machine Learning Foundation, and Builder Labs Ten new labs to help learners put theory into practice Redesigned website guides learners to training content based on their knowledge, goals, interests, and age New online Explore section features supplementary content, such as new courses, Twitch videos, blogs, and technical papers.","excerpt":"AWS aims to provide free training to 29 million people worldwide.","categories":["AI News"],"tags":["self learning"],"author_name":"Kartik Wali","publish_date":"2022-03-17T10:47:19","publication_year":"2022","word_count":323,"keywords":["Go","machine learning","AWS","cloud computing","AI","cloud_platforms:AWS","programming_languages:R","programming_languages:Go","self learning","cloud_platforms:Amazon Web Services","R"],"extracted_tech_keywords":["AI","machine learning","cloud computing","AWS","R","Go","cloud_platforms:AWS","cloud_platforms:Amazon Web Services","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/aws-launches-game-based-courses-to-teach-cloud-skills\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10061157,"title":"Enter the new era of automation with AI in RPA &#8211; In conversation with UiPath MVPs","content":"As per a report by Gartner, through 2024, large organisations will triple the capacity of their existing Robotic Process Automation (RPA) portfolios. It also predicts that 90 per cent of large organisations globally will have adopted RPA in some form by 2022, as they will look to digitally empower critical business processes through resilience and scalability while recalibrating human labour and manual effort. UiPath is one of the biggest names in the field of RPA. UiPath’s Most Valuable Professional (MVP) Award is the highest recognition to its community members for their outstanding contribution, innovation, and evangelism shown in the larger automation community. In an email interaction, Nisarg Kadam, UiPath MVP 2021, working as a Senior Consultant at WonderBotz and Sishira Mishra, UiPath MVP 2021, working as Automation Architect at HP, discuss how they started their RPA journey, how UiPath helped them in attaining and honing their skills and what can we expect in the RPA domain in the future. RPA journey Kadam says that his RPA journey started in 2018 when he learned from the UiPath Academy. Today, he strives to give back the information he learned over the years to the community by helping ensure that everyone knows how to utilise the product and scale the automation beyond what they have imagined. “In order to reach more people, I have started a YouTube channel through which I have shared 131 videos to date, with over 340,000 views and 5,300 subscribers,” he adds. Over the years, Kadam has participated in and won two global hackathons where he showcased the power of ML integrated RPA using UiPath’s AI Center. He has also been a speaker in over 20 UiPath Community sessions, most recently with his project – ‘Automation Pathshala’ where he trained how to use the UiPath application to over 1300 registered participants. Mishra’s RPA journey started in 2015, working with several vendors when the focus was more on unattended automation with a classic RPA approach. This journey has shifted to cognitive RPA over the years with an increased focus on enabling digital transformation. “Over the years, I have been involved in multiple automation development projects with varying complexity across business functions. Having come from an IT background, I did not understand business processes in-depth. However, working with RPA, I have been able to realise business problems and how this technology can be a key solution while maximising innovation.” –  Sishira Mishra He has had the opportunity to work on the end-to-end automation ecosystem that involves setting up the infrastructure, discovering potential automation opportunities, designing and developing solutions, and providing support for the production platform. Parallelly, he has worked closely with the UiPath Engineering and product team through the UiPath Insider Program, where he shared ideas and experiences with the UiPath Insider Community. Scope of automation is increasing beyond imagination Kadam feels that the RPA industry is growing rapidly, and with this, the scope of automation is increasing beyond our imagination. He has learnt a lot in less than four years but still feels that he knows very little about this industry and has so much more to learn. This shows how huge this industry has become. In 2018, RPA began mimicking humans. Eventually, when the industry understood the limitations of definite step automation, vendors started exploring cognitive automation by plugging in AI and ML. As a result, they open unlimited doors for RPA to scale in any industry. With cloud deployment becoming a trend, RPA will have a significant role in ERP automation, providing processed and less complex data directly from the ERP, thereby improving customer experience. This year will be amazing because of. Over the next couple of years, powerful automation discovery products like Task Mining and Process Miningare going to help companies unearth the scope for automation across different domains. Citizen-led Automation Mishra says that with ‘Low-code’ and ‘No-code’ applications existing today, everyone can develop robots without a programming background or experience. A key enabler is in this process is Citizen-led Automation. This mindset is taking the industry to the next level, where businesses can produce more value by encouraging automation from operational teams. Products like UiPath StudioX is a true no-code platform that helps business users develop their task automation without the involvement of CoE Developers. “Low-code and No-code applications are the key enablers in Citizen-led Automation” This is also helping build a Citizen Developer ecosystem at a large scale across industries. We will soon witness organisations encouraging their Citizen Developers to build task-based automation and improve productivity. With products like UiPath Process Mining and UiPath Task Mining, it is easier for companies to find potential areas of opportunity where CXOs can focus on developing and deploying automation. Challenges encountered Mishra remembers the challenge he initially faced to learn automation technologies and receive the right support. With UiPath Academy and UiPath Forum, everything is available for free and at an industry-standard level. Every user can be part of the community, receive access to all UiPath products, learn using video-based training, and receive support from the forum. “In our industry, it is important to be up-to-date on all upcoming products. By actively participating in the UiPath Insider Program and connecting over the UiPath Forum, I could collect information about upcoming products and collaborate with the product team to test features over private previews before their release to the general public.”  –  Sishira Mishra Kadam remembers the most challenging integrating the Python ML model with UiPath when the AI Center had not yet been released. As a result, he had to learn ML with Python and invoke Python prediction models within automation to make it a seamless experience on his projects. Kadam also faced operational challenges including setting up a cloud-based document review without human dependency, training bots to read and extract data such as total, tax, and vendor names from 50 different types of invoices, enabling automated email replies, etc. that he was able to resolve using the UiPath AI Center. Working with chatbots has been one of Kadam’s most enjoyable experiences. However, he still counts it as a challenge because he learned a completely new product – Google Dialog flow – to integrate a chatbot with RPA. When deployed on the production website, he could not have been happier to see his first chatbot and RPA integration. Centrally manage ideas and enable a single platform Mishra says that once we begin scaling automation, one of the challenges we face is centrally managing ideas and enabling a single platform where everyone can share their thoughts and collaborate. This can be easily overcome with the UiPath Automation Hub. “The UiPath Automation Hub is a single platform with features to overcome these challenges. Its capabilities not only enable you to submit ideas but also provide several other components including those that help users manage their automation funnel or pipeline easily, create automated documents using Task Capture, consume reusable modules etc.,” he added. Does not demand too much process revision RPA as technology is one of the easiest top layer technologies to set up without impacting the company’s current infrastructure. It does not demand too much process revision and adapts rapidly across industries. For instance, over the last four years, Kadam has worked with companies in retail, telecom, healthcare, banking, news channels, gaming, education, auction houses, airlines, petroleum, and construction domains to provide successful pilot robots and beautiful end to end automation. For example, when Kadam was working with his auction house client, he realised every single process of an online auction house could be automated easily. However, RPA has not yet evolved to 100 per cent of its capability, and there is a long way to go and grow. “Industries continue to focus on RPA adoption to reduce mundane, repetitive work, freeing manual hours and thereby helping users contribute more towards innovation. Banking and financial services, healthcare, and insurance sectors have actively adopted RPA. The manufacturing, retail and telecom industry can also be seen adopting RPA for quality and efficiency improvements in operations to help improve workforce and customer relationships,” added Mishra. AI in RPA is the next era in the Automation Industry Kadam says that RPA was introduced in this industry as a technology that automates repetitive tasks, but when it started growing, it was understood that it could also automate cognitive tasks, make decisions and grow smarter with the help of a historical learning feed. “I was allowed to eliminate tedious manual tasks the L1 support team performed in a service ticket management department. The task was to assign those tickets to an appropriate resolver group based on the availability and priority of the ticket. With the help of the UiPath Action Center, I was able to automate 90 per cent of the process with hybrid robots (90 per cent unattended and 10 per cent attended bots). The fun part of this process was that we were able to predict the category, priority, and impact of the service ticket based on the email body, quite easily with the help of UiPath AI Center, using its ‘English text classification model.” – Nisarg Kadam Need to go above the rule-based model to a self-decision-making model Mishra says that AI in RPA is the next era in the automation industry. Robots need to go above the rule-based model to a self-decision-making model to enable more value-added work and benefits. However, this is only possible when artificial intelligence is combined with RPA. With this, in the future, we will be able to move away from traditional RPA to a cognitive RPA. “In today’s ecosystem, with UiPath AI Center, Form AI and Document Understating, we have enabled a business user without any technical background to create their ML model and begin managing unstructured data,” he added. UiPath AI Center Kadam says that when he would hear terms like AI and ML, all he would visualise was a black screen with Python coding on Ubuntu by skilled developers. However, he was mesmerised by the simple drag and drop AI models when he completed his first AI-enabled automation pilot robot within four days of development without learning a single line of Python code. “When we delivered our first AI-enabled invoice processing solution, the AI Center helped us gain confidence with over 90% accuracy,” said Kadam. “The feature where an AI Model is trained by learnings from experience and parameters to increase is useful as it helps increase its confidence and become smarter day by day.” Mishra adds that The UiPath AI Center is a combination of everything that enables the deployment of an AI-based project. For example, a typical life cycle of an AI Project involves data labelling, training, deploying, and retraining, which can all be performed using the UiPath AI Center. In addition, when teams face typical machine learning challenges, the AI Center helps create an end-to-end AI automation workflow by: Collecting datasets that can be part of data training to enhance the out-of-box model or reuse existing models developed in-house.Performing data labelling within AI Center to create a labelled dataset for ML Training.Running training\/ evaluation pipeline by reusing the base model.Deploying the skill via AI Robot, which manages the required resources remotely.Pushing the model to the regular workflow for use on being deployed. In some scenarios requiring human intervention in the loop, the Action Center is used to perform human validation and plan the required retraining pipeline to improve accuracy. Drag and drop an ML model into their workflow Nisarg adds that the UiPath AI Center allows one to drag and drop an ML model into their workflow without writing a single line of code. It provides: 19 out-of-the-box ML models.over 20 UiPath in-house ML models.and the option to upload the user’s custom-built model. These models are offered as pre-trained, re-trainable, and non-re-trainable model types. In addition, the UiPath AI Center also offers 16 different in-house document processing models like invoices, receipts, etc., that are also re-trainable to suit the users’ documents and can manage all types of ML models, right from language analysis and image analysis to tabular data processing. It is easy to use and offers a step-by-step experience that helps users choose an ML model and deploy it without hassle. All the developers have to do is create a project, upload a test\/ train data (only in case of re-trainable models), choose the ML model, execute the train and evaluate pipeline (in case of retraining), create ML skill, and drag and drop the same into the UiPath Studio. “The AI Center supports the infrastructure to manage training, evaluation, and deployment of models with two special types of robots called the AI Robot and the AI Robot Prov (GPU based),” added Mishra. “While the AI Center manages the complete backend load, the custom or out-of-box models are available to RPA Developers with a simple drag and drop activity. Moreover, it is simple to use with its no-code platform.” Interesting use cases where UiPath AI Center has helped apply AI and Machine Learning models for automated processes Kadam has recently written a UiPath Community blog about possible use cases with only ML models and shares some of his hackathons and community projects. He worked on an AI-enabled customer feedback automation for an Automobile Service centre looking to capture handwritten feedback. The feedback would then have to be classified as ‘Positive’ or ‘Negative’ and uploaded on social media, along with an automated email follow-up to the customer before a service executive reaches out. The project used UiPath Studio, UiPath AI Center, UiPath Automation Hub, UiPath Document Understanding, UiPath Orchestrator, and UiPath Task Capture to help improve customer experience. He adds, “Automation’s applications are diverse. For example, we can also work on assigning tickets on ‘Service Now’ using ‘English Text classification model’, predict brain stroke using ‘TPOT auto ML classification model’, or build a slack channel chatbot for foreign languages using ‘Language detection’ and ‘language translation’ ML model.” He has also worked on counting the number of allowed people at a time in a supermarket using the ‘Object detection ML model’, helped maintain old and valuable books by converting handwritten hard copies to soft copies using ‘Document Understanding OCR’ and helped detect entries of trucks and create a gate pass for drivers using the ‘Object detection ML model’. The bot would detect number plates and create auto-generated gate passes for drivers. Mishra has used the UiPath Document Understanding models, where the bot reads invoices with varying formats, extracts all relevant information and posts the same into the company’s ERP system. He has also worked with marketing teams with campaigns to validate image-based objects against key deliverables. The AI Centre also becomes useful for SOW validation, email sentiment and intent analysis, text translator capability when working with multiple geographies, GST validation for Indian invoices, etc. The UiPath Community Kadam joined the UiPath Community with the aim of learning and sharing. He shares that The UiPath Community is one of the best as it helps with a lot of support, knowledge, tips, insights, tricks, techniques, codes, ready-made workflows, best practices, hackathon opportunities, job opportunities, career guidance, internship opportunities, use-case repositories, templates, presentations, live sessions, one-on-one with experts, and others. To join, all you have to do is connect with your local chapter on community.uipath.com or join on forum.uipath.com. Mishra adds that there are several such initiatives one can be a part of, such as joining the UiPath Academy to learn UiPath Products and technology and the UiPath Forum to communicate and collaborate over issues or concerns and updates on UiPath products. In addition, one can also connect to UiPath MVPs, become an RPA Associate, or receive a UiPath Certified Advanced RPA Developer (UiARD) certification, and find a mentor using the UiPath mentorship program. Focus on learning from experience and practical learning Nisarg and Mishra share some tips one can follow while starting in this industry: To strengthen your basic knowledge and acquaint yourself with RPA technologies, we recommend completing the RPA developer foundation course from academy.uipath.com. Learn basic development concepts and complete RPA Associate or the UiPath Certified Advanced RPA Developer (UiARD) certification. Try not to stress too much regarding your performance during the initial days and focus on learning from experience and practical learning. Practice more, explore UiPath Go’s reusable library, and be ready to consume automation. Sufficient hands-on experience can help you talk about development during an interview. It is advisable to be active on the UiPath Forum as we get to learn from others and receive help with our queries. There are also several documents available to us to understand in depth about any product and use-cases, challenges, and hackathons available throughout the community forum. s. Join community events. Hackathons are the best way to grow knowledge and learn from others, and we strongly recommend participating in every hackathon. Use developer best practices from an RPA perspective and apply them when giving practical demos in an interview or during work development. Having an RPA developer best practice skillset will always keep you one step ahead in your journey in the long run. It will help you to become a better solution architect and manage delivery going forward. RPA in 2022-24 Kadam says that the RPA industry is here to stay for a much longer run. However, with the growing digital transformation in every industry, it is becoming increasingly important to adapt to new technologies and bring modernisation. He adds, “I would say the major transformation factor for the RPA industry would be to adapt to IoT robots and Cloud technology. The era of RPA cloud technology has already begun, but there is a huge scope for business between 2022 and 2024. In addition, the RPA industry will also evolve to bring Intelligent Character Recognition (ICR), Optical Character Recognition (OCR), Natural Language Processing (NLP), and Natural Language Generation (NLG) to become a house of Intelligent Automation suite. Mishra says that with Process Mining and Task Mining, businesses will discover and strengthen their automation funnel while expanding their digital transformation initiative. The no-code platform will show a strong footprint across the industry, the cognitive automation space will scale and strengthen itself with DU, AI Center, Form AI-like capabilities, and UiPath’s vision of a ‘Bot For Everyone’ will begin showing a true impact. The period of 2022 to 2024 will be a whole new era for automation and the RPA industry, and we must keep ourselves up to date with its fast-paced developments.","excerpt":"UiPath’s Most Valuable Professional (MVP) Award is the highest recognition to its community members for their outstanding contribution, innovation, and evangelism shown in the larger automation community.","categories":["AI Features"],"tags":["Automation","Career","Interviews and Discussions","RPA","uipath","UiPath RPA"],"author_name":"Sreejani Bhattacharyya","publish_date":"2022-02-21T15:00:00","publication_year":"2022","word_count":3067,"keywords":["text classification","artificial intelligence","machine learning","AI","chatbots","RPA","ML","Automation","UiPath RPA","RAG","NLP","Aim","object detection","uipath","Career","Interviews and Discussions"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","NLP","Aim","RAG","chatbots","text classification","object detection"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/enter-the-new-era-of-automation-with-ai-in-rpa-in-conversation-with-uipath-mvps\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10064432,"title":"AMD acquires distributed service platform Pensando","content":"AMD has entered a definitive agreement to acquire Pensando for approximately USD 1.9 billion. Pensando’s distributed services platform includes a high-performance, fully programmable packet processor and comprehensive software stack that accelerates networking, security, storage and other services for cloud, enterprise and edge applications. “To build a leading-edge data center with the best performance, security, flexibility and lowest total cost of ownership requires a wide range of compute engines,” said Dr Lisa Su, AMD chair and CEO. “All major cloud and OEM customers have adopted EPYC processors to power their data center offerings. Today, with our acquisition of Pensando, we add a leading distributed services platform to our high-performance CPU, GPU, FPGA and adaptive SoC portfolio. The Pensando team brings world-class expertise and a proven track record of innovation at the chip, software and platform level which expands our ability to offer leadership solutions for our cloud, enterprise and edge customers.” Pensando’s products are already deployed at scale across cloud and enterprise customers, including Goldman Sachs, IBM Cloud, Microsoft Azure and Oracle Cloud. Pensando’s high-performance, highly scalable distributed services platform includes a programmable packet processor that can be distributed throughout a network to efficiently accelerate multiple infrastructure services simultaneously, offloading workloads from the CPU and increasing overall system performance. Combined with Pensando’s system software stack, the platform offers unprecedented performance, scale, flexibility and security. In real-world cloud deployments, Pensando’s solution demonstrates between 8x and 13x greater performance compared to competitive solutions. “We are excited to join the AMD family. Our shared cultures of innovation, excellence and relentless focus on partners and customers make this an ideal combination. Together, we have the talent and tools to deliver on our customers’ vision for the future of computing,” said Pensando CEO Prem Jain. “In less than five years Pensando has assembled a best-in-class engineering team that are experts in building systems together with a rich, deep ecosystem of partners and customers who have currently deployed over 100,000 Pensando platforms into production. Joining together with AMD will help accelerate growth in our core business and enable us to pursue a much larger customer base across more markets.”CEO Prem Jain and the Pensando team will join AMD as part of the Data Center Solutions Group, led by AMD Senior Vice President and General Manager Forrest Norrod.","excerpt":"CEO Prem Jain and the Pensando team will join AMD as part of the Data Center Solutions Group.","categories":["AI News"],"tags":["AMD","data centre","Mergers and Acquisitions"],"author_name":"Kartik Wali","publish_date":"2022-04-06T13:28:18","publication_year":"2022","word_count":380,"keywords":["Go","AMD","cloud_platforms:Azure","cloud_platforms:Microsoft Azure","AI","programming_languages:R","R","innovation","Scala","RAG","Mergers and Acquisitions","Azure","data centre"],"extracted_tech_keywords":["AI","RAG","Azure","R","Go","Scala","innovation","cloud_platforms:Azure","cloud_platforms:Microsoft Azure","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/amd-acquires-distributed-service-platform-pensando\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10097764,"title":"Here&#8217;s The Ultimate Shield against Unauthorised Image Manipulation","content":"From romantic poems to Salvador Dali-inspired images, generative AI can now do it all. And it can do it so well, that it is often impossible to differentiate between AI and human-generated artworks. Since the Turing Test, which set the standard for successful AI performance as being able to mimic humans so well that it becomes indistinguishable, the discussion about technology imitating humans has been a major topic of public debate. The community has always tried to distinguish between text written by humans and text generated by AI, amidst the risk of possible misuse of technology. MIT’s Computer Science & Artificial Intelligence Laboratory has come up with a solution for this. MIT Has A Solution: PhotoGuard Scientists from MIT CSAIL have created a new AI tool called “PhotoGuard” that aims to stop unauthorised changes to images made by models like DALL-E and Midjourney. This tool is specifically designed to protect against image manipulation without proper authorization. PhotoGuard leverages “adversarial perturbations,” which are minuscule alterations in pixel values that are not visible to the human eye but can be detected by computer models. These perturbations disrupt the AI model’s ability to manipulate images effectively. There are two attack methods used by PhotoGuard to generate these perturbations. The “encoder” attack targets the AI model’s latent representation of the image, causing the model to perceive the image as random. The goal of this attack is to disrupt the LDM’s process of encoding the input image into a latent vector representation, which is then used to generate a new image. They achieve this by solving an optimization problem using projected gradient descent (PGD). The resulting small, imperceptible perturbations added to the original image cause the LDM to generate an irrelevant or unrealistic image. On the other hand, the “diffusion” attack defines a target image and optimizes the perturbations to make the final image resemble the target closely. This attack is more complex and aims to disturb the diffusion process itself, targeting not only the encoder but also the full diffusion process that includes text prompt conditioning. The goal is to generate a specific target image (e.g., random noise or a grey image) by solving another optimization problem using PGD. This attack nullifies not only the effect of the immunized image but also that of the text prompT. Hadi Salman, lead author of the paper and a PhD student at MIT told AIM, “In essence, PhotoGuard’s mechanism of adversarial perturbations adds a layer of protection to images, making them immune to manipulation by diffusion models.” By repurposing these imperceptible modifications of pixels, PhotoGuard safeguards images from being tampered with by such models. Salman wrote the paper alongside MIT CSAIL graduate students and fellow lead authors Alaa Khaddaj and Guillaume Leclerc MS ‘18, as well as PhD student Andrew Ilyas ‘18 MEng ‘18. For example, consider an image with multiple faces. You could mask any faces you don’t want to modify, and then prompt with “two men attending a wedding.” Upon submission, the system will adjust the image accordingly, creating a plausible depiction of two men participating in a wedding ceremony. Now, consider safeguarding the image from being edited; adding perturbations to the image before upload can immunise it against modifications. In this case, the final output will lack realism compared to the original, non-immunized image. “I would be sceptical of AI’s ability to supplant human creativity. I expect that in the long-run AI will become just another (powerful) tool in the hands of designers to boost the productivity of individuals to articulate their thoughts better without technical barriers” concluded Salman. Decoding the Problem The recent Senate discussion around AI regulation has turned the spotlight on the most pressing issues of copyright and artist incentivisation. Senior executives from OpenAI, HuggingFace, and Meta, among others have testified before the US Congress about the potential dangers of AI and suggested the creation of a new government agency to licence large AI models, revoke permits for non-compliance and set safety protocols. The major impetus behind this plea for regulation stems from concerns regarding copyright infringement. It first started when the artist community filed a lawsuit against the companies behind image generators like Stability AI, Midjourney, and DeviantArt seeking compensation for damages caused by these companies using their art without credit. AI-generated content is facing opposition from stock image companies like Shutterstock, Getty, as well as artists who see it as a threat to their intellectual property. But eventually, most of them have gotten on board with partnerships. Adobe’s Firefly is a generative image maker designed for “safe commercial use.” Adobe offers IP indemnification to safeguard users from legal issues related to its use. It is built on NVIDIA’s Picasso which is trained on licensed images from Getty Images, and Shutterstock. Shutterstock also partnered with DALL-E creator OpenAI to provide training data. It also now provides full indemnification to its enterprise customers who use generative AI images on their platform, ensuring protection against any potential legal claims related to the images’ usage. Google, Microsoft, and OpenAI have also started watermarking with the aim of mitigating copyright issues. Read more: Lessons from YouTube for Gen AI Copyright Mess","excerpt":"MIT has come up with a new AI tool designed to protect against image manipulation without proper authorisation.","categories":["AI Features"],"tags":["AI Tool"],"author_name":"Shritama Saha","publish_date":"2023-07-29T13:00:00","publication_year":"2023","word_count":857,"keywords":["Go","artificial intelligence","TPU","OpenAI","AI","AWS","RAG","Aim","generative AI","AI Tool","R"],"extracted_tech_keywords":["AI","artificial intelligence","generative AI","OpenAI","Aim","RAG","AWS","TPU","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/meet-photoguard-the-ultimate-shield-against-unauthorised-image-manipulation\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":32720,"title":"When Open Source And Cyber Security Bonds: Kali Linux, The Go-To OS For Penetration Testing","content":"When we talk about hacking, the first thing comes to our mind is a guy in a hoodie who is involved in data fraud, identity theft, and maybe even cyber terrorism (thanks to Hollywood!) However, this is not the scenario all the time; not all hacking is necessarily the criminal, destructive act. There is one form of hacking that is not related to any kind of criminal activity and organisations or institutes often use it to check their defences — Ethical Hacking or Penetration Testing. Today, with cybercrime gaining prominence, the concept of ethical hacking has become popular. A security-focused operating system (OS) is a hacker’s best companion and witnessing the skyrocketing popularity of Ethical Hacking, many Linux distros started entering the arena, but not every player could make it to the top. The OS that has become insanely popular among most of the penetration testers is Kali Linux. Here’s a look at the nuts and bolts of Kali Linux Released on March 13, 2013, by Offensive Security, Kali Linux is a rebuilt of BackTrack Linux. In order to make the penetration testing more effective, in 2012, Offensive Security decided to replace their venerable BackTrack Linux project and started working on Kali on top of the Debian distribution. But why Debian? Debian is considered to be one of the best distribution as it has the quality, stability, and wide selection of available software. The Debian-based Linux distro, Kali was created in such a way that it completely focuses on advanced Penetration Testing and Security Auditing. Whether it is about Penetration Testing, Security Research, Computer Forensics or Reverse Engineering, the OS is loaded with over 600 tools for performing almost every kind of information security tasks. Also, Kali has the live boot capability which definitely makes Kali Linux an ideal workstation for vulnerability testing. Some of the most popular pre-installed tools are: Metasploit: An open source security framework, Metasploit provides information about security vulnerabilities and aids in penetration testing and IDS signature development. Also, it is used to develop and execute exploit code against a remote target system. Nmap: Nmap is a free and open-source security scanner and is one of those useful tools that is basically used to discover hosts and services on a network. Wireshark: It is again a tool that is used on networks. Wireshark is a packet analyzer and is used for network troubleshooting and analysis. John the Ripper: It is a password cracking tool. John the Ripper was initially a Unix-based tool, but now it can be used in 15 different platforms. Aircrack-ng: It is one of the most popular tools in Kali. Aircrack-ng is a packet sniffer tool that works any wireless network interface controller and can sniff 802.11a, 802.11b and 802.11g traffic. What else makes Kali Linux stand out for Ethical Hacking & Penetration Testing? Kali Linux is specifically geared to meet the requirements of professional penetration testing and security auditing and with so many tools, the OS is undoubtedly one of the best in the market. When it comes to information security, workstations cannot be ordinary and this where Kali Linux comes into the play. Being an open source OS, Kali can be customised. Developed by penetration testers for penetration testers, Kali Linux is completely easy to customize based on the needs and preferences of the user. While most people think open source software is free, however, that is not the case. Open Source is about providing the source code, not making the software free of cost. Talking about charges of using Kali Linux, the OS is completely free (not because it is Open Source). “Kali Linux, like BackTrack, is completely free of charge and always will be. You will never, ever have to pay for Kali Linux,” reads the official documentation of Kali Linux. Unlike Windows, installing Kali Linux can be a little pain. Therefore, Kali comes with the Live System feature. Meaning, one can use Kali without even installing it; the ISO image of the OS can be used as a bootable live system. Also, the live system contains all the commonly used testing tools. So, just copy the image onto a USB key and you are good to go — plug and play. The word ‘Hacking’ itself is not trustable then how can one trust an OS that is loaded with 600+ penetration testing tools and is completely free? As mentioned before, Kali is an Open Source OS, which means one can inspect the source code. Also, according to the official website of Kali Linux, the developers of the OS upload signed source packages onto their repository, which can be followed and reviewed. Bottom Line With the proper know-how, hacking can be done in any platform — whether it is Linux or Windows. But, Kali Linux — the name itself invokes a strange curiosity, which makes people have a deep look into this OS. From the days of BackTrack to the latest version of Kali, with a plethora of testing tools that allow its users get straight to work, the OS has gained tremendous popularity in the space of penetration testing. Unlike other OSes, Kali might feel a bit difficult to use, but if you sincerely want to explore the world of cybersecurity, then Kali Linux is the premier choice — it is elegant, clean, and presents a ton of interesting things for you to experience and learn.","excerpt":"When we talk about hacking, the first thing comes to our mind is a guy in a hoodie who is involved in data fraud, identity theft, and maybe even cyber terrorism (thanks to Hollywood!) However, this is not the scenario all the time; not all hacking is necessarily the criminal, destructive act. There is one […]","categories":["AI Features"],"tags":["Cyber Security","cybercrime","Ethical Hacking","hacking","Kali Linux","Linux","Linux distros","OS","Security","testing"],"author_name":"Harshajit Sarmah","publish_date":"2019-01-03T09:04:35","publication_year":"2019","word_count":898,"keywords":["Kali Linux","OS","Rust","R","Linux distros","ViT","Linux","programming_languages:Rust","Go","hacking","AI","Security","GAN","Cyber Security","programming_languages:R","programming_languages:Go","cybercrime","testing","Ethical Hacking"],"extracted_tech_keywords":["AI","R","Go","Rust","GAN","ViT","programming_languages:R","programming_languages:Go","programming_languages:Rust"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/when-open-source-and-cyber-security-bonds-kali-linux-the-go-to-os-for-penetration-testing\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10096430,"title":"Meta’s Threads won’t Launch in EU over Privacy Concerns","content":"Meta’s Twitter rival app won’t launch in the EU yet on the concerns of data privacy, TechCrunch reported on Wednesday. In the United States, the platform informs users that it will collect a wide range of personal information, such as health and financial data, browsing history, location, purchases, contacts, search history, and sensitive information. However, the EU under DMA  has prohibited Meta from introducing advertising services on WhatsApp that utilize data from Facebook or Instagram. The new platform is created to collect information from Instagram, specifically regarding user behavior and interactions with advertisements. Meta is eagerly awaiting additional guidance on the Digital Markets Act, which are new rules in the European Union that govern how big online platforms can use their market influence, Bloomberg reported, citing a person familiar with the matter. The European Union is currently discussing the regulations with the company and is expected to give guidance in September, the report added. On Tuesday seven companies including Meta said that they meet the criteria of ‘gatekeeper’ which means they will have to meet EU’s tougher rules under Digital Markets Act. Companies violating DMA can be fined up to 10% of annual global turnover. Threads, released worldwide on Thursday is touted as a rival to Twitter. In the first two hours it passed 2 million sign ups, said Mark Zuckerberg on Threads post.","excerpt":"Threads gather data from Instagram which violates Digital Markets Act","categories":["AI News"],"tags":[],"author_name":"Siddharth Jindal","publish_date":"2023-07-06T13:16:46","publication_year":"2023","word_count":224,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","Git","R"],"extracted_tech_keywords":["AI","R","Go","Git","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/metas-threads-wont-launch-in-eu-over-privacy-concerns\/","complexity_score":2,"technical_depth":6,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":32187,"title":"Flashback 2018: Top News Stories About AI &#038; Analytics That Made Headlines","content":"As the year ends, we list the news and headlines that broke the records and developments that marked new growth in the industry. With a lot of funding and investment in the new tech startups, appointments in the industry, opening up of a centre of excellence in India, hiring and more, there were a lot of positive developments in the year 2018. In this article, we walk through the biggest developments that the industry witnessed in the last year. Appointments January Souma Das was appointed as the managing director of Teradata’s India operations. It was announced that he would be responsible for providing leadership and overall strategic direction of the company’s India business overseeing field operations that include sales, customer management, marketing, professional services and customer support. The same month, ex-Microsoft executive Arun Balasubramanian was appointed as the new India head for Qlik, who was taken on board with an aim to expand the company’s operations in India. With an experience of 20 years, he brings experience from companies such as HP, CA Technologies, Salesforce in areas such as IT market along with cloud and subscription delivery models. February The Smart Cube announced the appointment of Gaurav Kumar as its new Chief Operating Officer, who had first joined the company as Research Analyst in 2005. He was appointed to pick up the responsibilities previously held by Managing Director Sameer Walia, one of The Smart Cube’s three founders Dr Atul Varshneya was hired as the new AI head at Tavant with an aim to increase applied innovation for in the space of AI-based platforms and solutions. He was earlier involved with Samsung SDSRA AI engine which helps provide business solutions to companies with the help of AI and has extensive knowledge of software technologies, AI, machine learning. Namit Jain was appointed to head the worldwide Engineering management division at Qubole. Having worked with noted organisations such as Oracle, Facebook and Nutanix before, he has a vast experience in databases and big data technologies, along with scaling large global engineering teams. March Pushpak Bhattacharyya, director at Indian Institute of Technology, Patna was appointed to head the new committee for standardisation in artificial intelligence set up by the Bureau of Indian Standards (BIS), which falls under the Union Ministry of Consumer Affairs. It was set up with an aim of standardising projects that revolve around cybersecurity, legal and ethical issues in the IT sector, technological mapping and leveraging AI for national missions, among others. April Airtel appointed former NASA executive Santanu Bhattacharya as Chief Data Scientist in Bharti Airtel. He was hired to work in the new tech space such as artificial intelligence, internet of things, augmented reality and virtual reality, among others. He was appointed with an aim to bring data science capabilities and deep analytics to understand customer needs and develop innovative products and services for Airtel. Apple hired John Giannandrea, who had stepped down from his position as head of the Search and AI units only days ago. It was said that he would lead the company’s machine learning and AI strategy. May Google appointed Nitin Bawankule as head of India Cloud Business at Google. Bawankule, who has been with Google India for over six years as the head of sales for multiple verticals including e-commerce, retail, classifieds and education, brings over 20 years of experience in digital, technology and consumer industry. July Swiggy appointed ex-Amazon executive Dale Vaz as its head of Engineering and Data Sciences Department. He was taken on-board to improve Swiggy’s technology strategy and build the company’s next-generation artificial intelligence-driven platform for “hyperlocal discovery and on-demand delivery”. August John Gikopoulos joined Infosys as the global head for AI and automation. He was the director at IPsoft prior to this and has also served as a partner at McKinsey and Company. September Dr Andrew Moore replaced AI Guru Fei-Fei Li at Google cloud. It was one of the most discussed re-shuffle announcement announced by Google as Fei-Fei Li announced returning to her professorship at Stanford. Dr Moore had earlier worked at Google between 2006 and 2014. Google had announced in a statement that she would continue to work as an advisor to the company. October Cloudera appointed Vinod Ganesan as India country head. It was announced that he would drive Cloudera’s growth in the region, expand its customer base and help organisations enhance the measurable benefits of big data through machine learning and analytics. Funding And Investment January Mobile analytics startup CleverTap received fresh funding of an undisclosed amount from Japan-based Recruit Holdings Co. Ltd. The investment was made in association with CleverTap’s current investors, Sequoia Capital India and Accel Partners. It was speculated that the new investment secured by the startup was a part of its efforts to raise $8 million. PiVentures announced second close to their fund at $25 mn which it would use to boost startups in the space of applied artificial intelligence, machine learning and IoT. It had announced its plan to invest in 18-20 startups over a time frame of 3-4 years. Energy-tech startup Bidgely raised $27 mn in Series C funding Georgian Partners. One of the largest rounds of investments for the year, they plan to work on India expansion. Hyderabad-based startup eKincare raised $1.5M in Series A funding from prominent investment firms such as Venture East, Eight Roads, Touchstone Equities, Endiya Partners and included Padmashri awardee BVR Mohan Reddy in the investing round. AI-based HRTech startup Skillate got an undisclosed amount of funding from Incubate Fund India and Anuj Agrawal, Director at Zyoin. The company had mentioned its plan to use the funding amount to accelerate product development and improve its customer base in India and abroad. Andrew Ng introduced AI fund with $175 mn for AI startups. This fund envisions to build transformative AI companies from the ground up and will fund the work of experts in AI and machine learning. February FourKites raised $35 mn in Series B funding led by August Capital, which it had hinted to use in doubling its engineering talent pool. It had also announced additional investments from existing investors Bain Capital Ventures and Hyde Park Venture Partners. Healthcare analytics startup The Healthy Billion raises $2.1 mn in funding from Blume Ventures, Healthquad, Fireside Ventures, Apoorva Patni, and Arpan Sheth, a partner in Bain & Co, among others. Set up in 2015, THB is a healthcare analytics start-up combines clinical research and data analytics offering personalised healthcare to patients. AI startup Capillary Technologies Raised $20 Million from its existing investors, including PE firm Warburg Pincus and venture capital firm Sequoia Capital. It had suggested that it would use the fund ti open its second office in China. Trilyo, a Bengaluru-based artificial intelligence platform for hospitality industry raised $250,000 in a funding round led by Startup Buddy. March Wipro invested $2.02 mn on US-Based AI Startup Avaamo which was founded by Ram Menon and Sriram Chakravarthy in 2014. The key focus areas of the investments are AI, big data, IoT, FinTech, healthcare IT etc. Peritus AI, a startup providing AI and machine learning driven automation solution raised a funding amount of $2 mn from venture capital firm Ideaspring Capital and early-stage fund The Hive. The startup had announced to use it for setting up R&D centre and strengthening its development team in India. CustomerSuccessBox raised $1 mn from Pi Ventures in pre Series A funding round. Founded by Puneet Kataria, the startup was built to deliver ‘proactive’ customer success” using analytics. They wanted to direct these funds to drive growth and product innovation. Micromax invested an undisclosed amount in AI-based startup One Labs, which it had used to strengthen its device ecosystem. April Synctag raised over ₹2 crore in series A funding from a private equity firm Subhkam Ventures. It was reported that Syntag is going to use it for geographical expansion and scaling up the portfolio of products and services. Agri analytics startup AgNext got funded from Omnivore which was the second round of investment by the venture fund on the startup. The amount invested wasn’t disclosed by either party. As part of this funding, Subhadeep Sanyal, principal at Omnivore, joined AgNext’s board. Reliance announced that it would invest $180 mn in AI education platform Embibe over the next 3 years. The startup said that it would use the capital over the next three years to deepen its research and development on using AI in education. Mumbai-based robotics startup emotix raised $2 mn from IDG ventures and YourNest, which the startup announced to be using for new product development and research in areas like emotional and artificial intelligence. Twitter co-Founder Biz Stone invested in Indian AI-based health startup Visit, which is a brainchild of students from Birla Institute of Technology and Science (BITS), Pilani, who are connecting patients and doctors via an AI-powered chatbot. May Innovaccer raised $25 Mn Series B round of funding, led by Westbridge Capital with participation from Lightspeed Ventures. This brought the total capital raised by Innovaccer to $41 mn. June Bengaluru-based sales analytics startup DigiConnectt raised $125,000 in seed funding from US-based transport management solutions leader Aspire Logistics LLC. The fresh capital will be used by the startup for technology advancement and to strengthen its product lineup, DigiConnectt said in a statement. Intelligent logistics startup Locus raised $4Mn in pre-Series B Funding which it announced would be using to build IP and expand globally. The likes of Rocketship.VC, Recruit Strategic Partners, piVentures and Hemendra Kothari of DSP Group, invested in the company, with participation from previous investors Blume Ventures, Exfinity Venture Partners, BeeNext and growX ventures. Bosch announced to invest ₹17 bn in India, with a focus to expand Bengaluru-based Smart Campus. They announced an investment of ₹17 billion ($250 million) over the next three years. Elucidata raised $1.7 mn in seed funding from Hyperplane venture capital. Elucidata’s products are designed to take researchers from raw data to biological insights swiftly and efficiently. Noodle.ai raises $35 mn in Series B funding from Dell and TGP Growth. Their fundraising now totals to $51 million. The company will use the new funds to expand their suite of applications, which help key industries predict the future and make better business decisions. Healthcare startup SigTuple raised $19 mn (₹130 crore) in Series B round funding. Founded 2015 by former American Express employees Rohit Kumar Pandey, Tathagato Rai Dastidar and Apurv Anand, the company has raised $6.5 million from different investors so far. July RPA company Automation Anywhere raised $250 mn in Series A funding led by New Enterprise Associates, Goldman Sachs Growth Equity; with participation from General Atlantic and World Innovation Lab. This investment has brought the company’s post-money valuation to $1.8 billion. Media measurement analytics startup VTION bags premium funding from AngelList. Founded in 2016 by Manoj Dawane, VTION provides a unique platform for analytics and measurement technologies in the media sector. VTION hopes to use the funding amount to expand its portfolio and will have a broader reach in the Indian media ecosystem. Japanese financial services giant SBI Holdings invested $15 million in AntWorks, a three-year-old artificial intelligence company. The purpose of the MoU was to establish a joint venture company with AntWorks to operate businesses in Asia. Bengaluru-based Emotion AI startup Entropik Tech raised $1.1 mn in Pre-series A round led by Bharat Innovation Fund (BIF) and co-invested by Parampara Capital, Arthavida Ventures and Jitendra Gupta. They would use the fund to scale their platform Affect Lab 2.0. Iqlect raises $2.5 mn in a bridge round which was chiefly led by new investor VentureEast and existing investor Exfinity Ventures. The company creates products that use predictive and real-time data analytics to monitor and predict user and machine behaviours. Food tech startup Dishq (now Spoonshot) raised $400,000 in pre-seed funding which was funded by several investors including, Techstars’ first food\/AgTech-focused accelerator called Farm To Fork, and Arts Alliance. The company had said that it would be using the funds to expand the engineering team as well as sales and marketing activities. August Healthcare startup HealthSignz raised $5 mn in funding which was led by Dr Kantilal Patel, founder of New Zealand-based Nirvana Health Group. HealthSignz is an artificial intelligence-driven healthcare platform which uses a proprietary AI engine to help its users. Observe.AI secured $8 mn in funding led by Nexus Venture Partners with participation from MGV, Liquid 2 Ventures, Hack VC and existing investors Emergent Ventures and Y Combinator. They also announced their agent-first voice AI platform to improve caller satisfaction in call centres worldwide. Samsung announced that it plans to boost investments in businesses that will drive its future growth. It committed $22 billion investment over the next three years with a focus on areas such as artificial intelligence, 5G, automotive electronics components and biopharmaceuticals. Imarticus Learning raised $2 mn from CBA Capital and plans to go global with this funding amount. The funding amount would be used to further  Imarticus Learning’s vision of becoming India’s largest professional education institute, as well as creating a global footprint by bridging the gap between the industry and academia. pi Ventures, India’s first Applied Artificial Intelligence and IoT focused early stage venture fund raised $6 mn in funding, that it will invest in startups with deep tech capabilities. The funding was done by CDC Group. September AutoGrid, a company building AI and ML-based software applications announced that it had secured $32 million in series D funding from a set of strategic energy investors including CLP Holdings Group, Innogy, Ørstead and Tenaska, among others. Chennai-based SaaS startup CloudCherry Analytics secured $9 mn in Series A Funding from venture capital firm Pelion Venture Partners. The current investors Vertex Ventures, IDG Ventures India and Cisco Investments also took part in this round. The startup will use the fund to increase its product offerings. Mad Street Den, AI-for fashion startup raised Japanese telecom operator KDDI through KDDI Open Innovation Fund 2. The Indian-US startup, which was founded in Chennai, was founded in 2013 by a neuroscientist-designer duo, Dr Anand Chandrasekhar and Aswini Asokan. It raised an undisclosed amount during the funding round. Drone analytics startup SenseHawk raised $2 mn in a funding round from SAIF Partners a venture and growth capital fund. Other angel investors were also reported to have participated in the funding round. It was reported that they would use the fund for hiring and bolstering business development and deepening the product pipeline. pi Ventures announced the final close of their first fund at ₹225 crore, which came from CDC Group UK, Chairman of Hero Enterprise Sunil Kant Munjal, International Finance Corporation (IFC), Electronic Development Fund (managed by Canbank Ventures), SIDBI, prominent family offices from USA, Canada, Singapore and India and leading entrepreneurs like Mohandas Pai, Binny Bansal, Deep Kalra, Sanjeev Bikhchandani and Bhupen Shah among others. It plans to invest in more than 20 startups. Reliance industries made the second round of investment in AI startup NetraDyne. Bengaluru based startup is a deep learning startup that is pushing the boundaries in computer vision and its application in vehicle technology. It raised $8 million from Reliance Industries who had earlier invested $16 million in June 2016. October Betterhalf.AI raises $500,000 in seed funding. The investors included noted names like former MD of Google India Shailesh Rao, former investment director of Flipkart Rehan A Khan along with a slew of other senior investors. Betterhalf.AI positions itself as a “true compatibility” partner search product which uses artificial intelligence. Cognitive AI startup Worxogo raises ₹16.5 Cr from Inventus and Ideaspring Capital that the company plans to use to strengthen its AI engine called Mia, a personal digital coach for employees and expand the market base to US market and other global markets. VideoKen raised $930,000 in the second round of seed funding led by SRI Capital, Touchstone Equities and Hyderabad Angels. In their first round of seed funding, VideoKen had raised $1 million from a slew of angel investors. November CropIn Tech raises $8 mn in Series B funding led by Chiratae Ventures (formerly IDG Ventures) and the Bill and Melinda Gates Foundation Strategic Investment Fund. Engineer.ai raised $29.5 mn in Series A funding led By Softbank’s DeepCore and others. Reports indicated that the investment will be used for deepening engineering operations and drive customer acquisition. December Qualcomm launches $100 mn investment fund for AI to be invested in startups in the space. The company announced that the fund would focus on startups that share the vision of on-device AI becoming more powerful and widespread, with an emphasis on those developing new technology for autonomous cars, robotics and machine learning platforms. Vogo, an Indian scooter sharing network announced that Ola would be investing $100 mn in the company. It also saw participation from Matrix Partners India and Stellaris Venture Partners. Political January Rajnikant made a political debut with research and data analytics in his arsenal. It was reported that selected teams from the actor’s camp were dispatched across the state of Tamil Nadu to take stock of the agricultural and geopolitical concerns, and aggregate data and solutions pertaining to them. A technology team is said to be at the ready to deploy data analytics to the aggregated figures. In another interesting development, Jaganmohan Reddy turned to data analytics to devise a strategy for 2019 Andhra Pradesh Assembly Elections. It was reported that they are using analytics insights to maximise the effectiveness of his ongoing people’s outreach program, Praja Sankalpa Yatra. February Congress party announced setting up of analytics department. They had renamed their computer department to ‘Data Analytics Department’, with Praveen Chakravarty being nominated the chairman with immediate effect. October Facebook took a step to tackle election-related interference for Lok Sabha 2019. The company said that they were establishing a task force in India to prevent “bad actors” from abusing its platform and spreading fake news and hate speech. Indian Government And NITI Aayog February The government announced incentives for startups and businesses working in the AI sector. NITI Aayog, the policy think tank of Government of India announced that it would be offering incentives to startups and venture funds which could work and research in the field of artificial intelligence. March The government made use of data analytics to find the reason behind the revenue leakage in GST collection. GST Council had highlighted a number of major data gaps between the self-declared liability in FORM GSTR-1 and FORM GSTR-3B. Maharashtra government announced that India will soon get its first institute for Artificial Intelligence in Mumbai. The AI task force set up by the Ministry of Commerce and Industry recommended that a ₹1,200 crore corpus would be set up for five years under the Union Budget with an aim to push the National Artificial Intelligence Mission (N-AIM). April Reserve Bank of India (RBI) announced that it would be setting up a data science lab to add tech support in policy formulation. It had said that the creation of data science lab would help improve its “forecasting, nowcasting, surveillance and early-warning detection abilities that aid policy formulation.” NITI Aayog announced that it would open 30,000 Atal tinkering labs to promote New Tech in the education sector. They believe that India’s growth for the next few decades will depend on the innovations coming out of these tinkering labs. May Punjab National Bank (PNB) announced that it plans to implement AI in account reconciliation as well as using analytics to improve its audit systems. With the AI-enabled audit system in place, PNB aims to ensure improved security to its banking system to eliminate loopholes. June NITI Aayog released paper for India’s strategy on AI. It identified five sectors to focus its efforts towards the implementation of AI to serve societal needs such as healthcare, agriculture, education, smart cities, smart mobility. NITI Aayog’s Atal tinkering labs added 3,000 more schools in its portfolio which is an innovation-driven programme for students to come up with new and creative ideas in the field of science and technology. Collaboration, Partnerships And Acquisition March IIT Madras joined hands with Applied Materials India for research in AI, machine learning and data sciences. It was another strategic agreement of the institute with the industry to boost academic and research activities Fractal Analytics acquired Final Mile to bring together AI and behavioural sciences. The acquisition was done for an undisclosed amount and brings together the disciplines of data analytics, artificial intelligence and behavioural sciences. In another industry-institute partnership, Samsung partnered with BITS Pilani to upskill employees in AI, ML. Samsung India signed a Memorandum of Understanding with engineering institute BITS Pilani to facilitate advanced learning for its employees at Samsung R&D Institute (SRI) in Noida. April Myntra bought IoT-enabled smart wearable startup Witworks for an undisclosed amount. As a part of the acquisition, the Flipkart-owned firm will induct the startup’s team into its Innovation Labs. May EXL signed a definitive agreement to acquire SCIO Health Analytics as a part of a $240 million deal. SCIO’s significant early investors include Sequoia India, Health Enterprise Partners and Saama Capital. Cognizant acquired privately-held Hedera Consulting, a company that specialises in business advisory and data analytics services across a number of industry sectors. The purchase would expand Cognizant’s consulting, business insight and digital transformation capabilities for clients in Belgium and the Netherlands. June The Cross-Tab Group which includes the companies Blueocean Market Intelligence, Cross-Tab, and Borderless Access, went through a major restructuring with two of its three companies, Blueocean Market Intelligence and Cross-Tab, merging to form a new powerful entity — Course5. InMobi partnered with Microsoft to provide AI-based insights to marketers. The partnership will consist of InMobi moving to Microsoft Azure as its preferred cloud provider. The Indian Institute of Technology (BHU) Varanasi and Amazon Internet Services Private Limited (AISPL) signed a memorandum of understanding this week to develop cloud-ready job skills by providing access to the AWS Educate program. July Reliance acquired US-based Radisys for $74 mn with a plan to accelerate 5G and IoT growth. According to a statement released by Radisys, they entered into a “definitive agreement” under which RIL will acquire Radisys for $1.72 per share in cash. Karnataka Govt signed MoU with Taiwanese science parks to promote AI and robotics. The deal would see a collaboration between Hsinchu Science Park and Central Taiwan Science Park of Taiwan and Karnataka Innovation and Technology Society (KITS). India and the UAE signed a Memorandum of Understanding (MoU) for the India-UAE Artificial Intelligence Bridge in New Delhi. It was estimated that this partnership will generate an estimated $20 billion in economic benefits during the next decade for both countries. August Flipkart acquired Liv.ai. It was estimated that the deal was closed at around $40 million. September IIT Bombay joined hands with IBM to carry out advanced research in artificial intelligence under IBM’s AI Horizons Network. IIT Bombay paired their noted faculty and top graduate students with leading AI scientists from IBM Research India to advance and accelerate the application of AI, machine learning, natural language processing and related new technologies to business and the industry. After Liv.ai, Flipkart acquired Israeli retail analytics startup Upstream Commerce. Reportedly, the move will help Flipkart to deliver real-time pricing and product analytics to its seller partners. Bengaluru International Airport Ltd (BIAL) inked an agreement with Unisys, leading global tech firm for an advanced data analytics and BI platform to overhaul the passenger experience by providing more personalisation. BIAL has also tied up with Unisys Corporation – to set up the analytics centre of excellence. October NITI Aayog and tech giant Microsoft forged a partnership to leverage AI-based solutions in two sectors — agriculture and healthcare. It was aimed at pushing the use of AI, cloud and research and develop vertical expertise in new initiatives and solutions across agriculture, healthcare and the environment. Spoonshot acquired restaurant analytics startup Brisky. The financial details of the deal were not disclosed. SBI joined a definitive agreement with Hitachi to enter into a joint venture for the establishment of state-of-the-art card acceptance and future-ready digital payment platform for India. The JV agreement is subject to requisite regulatory approvals. IBM acquired Red Hat For $34 Billion. The acquisition of Red Hat is a game-changer and is speculated to change everything about the cloud market. Airtel acquired Bengaluru-based startup AuthMe to strengthen its AI Portfolio. This deal adds to Airtel’s recent initiatives to serve customers with innovative digital products. November The Kerala state government signed a memorandum of understanding with Airbus BizLab to open a state-of-the-art innovation centre. This centre will be the nodal body for planning and executing all the activities to help start-ups. Mindtree partners with IISc Bangalore to advance research in artificial intelligence. This partnership is aimed at facilitating teaching and research for AI in both data sciences and Natural Language Processing (NLP). December India and Russia signed an agreement to increase cooperation in areas of artificial intelligence, blockchain technology and explore the possibility of joint work in the healthcare sectors. In a move to improve their artificial intelligence- based personal assistants and another internet of things projects, Apple Inc acquired a San Francisco-based startup Silk Labs. It is doing some cutting-edge work on next-generation visual and audio intelligence to connected products using deep neural networks. IIT Bombay and NEC Technologies India inked an agreement to deepen research and development in the area of big data analytics, IoT and AI to tackle India’s social challenges. IBM announced that it has collaborated with IIT Delhi in a multi-year deal to work on AI. The aim is to discover novel AI techniques which can help organisations make informed decisions by being able to logically reason with their AI systems. Walmart Labs acquired int.ai to strengthen its engineering team. New Lab Set-Up And Hiring In AI February American multinational computer software giant Adobe announced that they to set up an advanced artificial intelligence laboratory in Hyderabad soon. Noted data company CtrlS launched Asia’s largest Tier4 data centre in Bengaluru. The data centre, which is located in the Electronics City part of the metro, is set to provide mission-critical operations to India’s leading internet banking, insurance, and other clients. Samsung had announced that is set to hire 1,000 engineers from top Indian institutes to work specifically in the research and development department in areas such as artificial intelligence, machine learning, internet of things and 5G, among others. Data analytics company, Axtria announced that it plans to double its headcount in India and is looking to hire over 700 data science professional in India. March Indian Institute of Technology, Kharagpur (IITKGP) is setting up a Center of Excellence in Artificial Intelligence Research with seed funding from Capillary Technologies Limited, with funding of ₹5.64 crores. Foodpanda, one of India’s noted online marketplace for food delivery, announced the launch of its dedicated technology centre in Bengaluru. It is looking to hire more than 100 team members with strong tech background over the next six to nine months. Flipkart had announced that it has 700 open vacancies majorly across functions in the technology department including data science July Paypal announced that it is looking to hire 600 more tech experts in India in areas like artificial intelligence and machine learning, among others, by December this year. Some of the job roles that PayPal intends to hire for include application developers — Java, Node JS; back-end developers — data engineers, data experts and product managers. A report suggested that noted tech company Synopsys is set to expand their base in India by hiring 200 more employees over the course of the next two years. It works in the area of autonomous cars, wearables, smart medical devices, secure financial services, artificial intelligence, machine learning, computer vision and IoT, among others. The Council of Scientific and Industrial Research (CSIR) joined hands with US-based chip designer NVIDIA to set up a Centre of Excellence at Central Electronics Engineering Research Institute (CEERI) in New Delhi. This centre will house India’s first ever supercomputer. December OnePlus announced that it plans to set up one of its first R&D facilities in Hyderabad. According to reports, this facility will purely focus on promulgating AI and ML in their products specifically for Indian consumers. There were reports suggesting that Havells India is planning to expand its R&D base in India. They are looking to invest over ₹1,500 crore over the period of next five years hire around 100 fresh IoT techies. Product And Accelerator Programme Launch April Tableau announced the global launch of new data preparation product ‘Tableau Prep’ and their new subscription offering, that would help in scaling up the analytics capabilities in an organisation. June IBM announced the launch of their artificial intelligence-powered enterprise marketing solution in India. This move will reportedly allow customers to host their marketing data on local IBM cloud data centre while giving customers proximity and scalability. July Microsoft introduced new AI-backed capabilities in Microsoft 365 with a host of new features. IIT-Hyderabad develops new AI-based system to catch bikers without helmets in surveillance videos. Hyderabad City Police signed a memorandum of understanding with the institute. August Google rolled out Android Pie, its highly-anticipated mobile phone operating system. This OS’ major updates focus is on artificial intelligence, which will allow the system to “learn” from the user and customise the Android experience. One97 Communications Limited, the holding company for Paytm, announced the launch of an artificial cloud computing platform for the online payment portal, named ‘Paytm AI Cloud for India’. The platform is aimed at developers, startups and enterprises. They partnered with Alibaba for Cloud Computing Infrastructure. At their annual flagship conference Google For India 2018, the search engine giant announced the launch of Navlekha, a platform built to help Indian language publishers get their content on the internet. The search feed will now display favourite news from both English and Hindi sources, using artificial intelligence that learns which types of stories the users like best. September Google launched Dataset Search, for scientists, data journalists, data geeks, or anyone else can who wants to find the data required for their work and their stories. Samsung Electronics announced the opening of the world’s largest mobile experience centre in Bengaluru, India. It is a 33,000 sq ft standalone property which will host experiences around transformative technologies like virtual reality, artificial intelligence and the internet of things. October Anil Kumble and Microsoft came together to announce the introduction of “Power Bat”, a product that helps players, coaches, commentators and fans to view and analyse the game in a unique way. HDFC Bank announced the launch of their Accelerator Engagement Programme (AEP) under the Bank’s Centre of Digital Excellence (CODE) that aims to support ideas innovative solutions in the area of artificial intelligence, machine learning, analytics, and robotic process automation, among other new tech areas. Philips announced the launch of their first global startup collaboration program involving Philips’ innovation hubs in Cambridge (US), Eindhoven, Bengaluru and Shanghai, focused on the application of artificial intelligence in healthcare. Teradata announced a new platform named Teradata Vantage which is Teradata’s next-generation analytics platform. and is now available to all customers. Vantage will allow enterprises to detect and uncover actionable insights into the biggest business problems by improving the analytics function. November Synechron Inc, the noted global financial services consulting and technology services provider, announced the launch of their Artificial Intelligence Data Science Accelerators for the Banking, Financial Services and Insurance (BFSI) firms. December Noted Indian pharmaceutical company Lupin launched a new chatbot called ANYA which will provide medical information and addresses queries for patients. SBI Mutual Funds launched their first artificial intelligence-powered voice assistant in collaboration with Google. AI In Healthcare January Apex Heart Institute, Ahmedabad became the first location outside the US to have a commercial installation of the advanced AI-powered vascular robotic system, CorPath GRX. Philips Healthcare launched a new line of AI-powered imaging devices which will help radiologists and imaging departments across hospitals. The three products launched are Access CT 32 Slice, Ingenia Prodiva 1.5T MRI and Dura Diagnost F30 Digital X-ray. March Microsoft announced to bring artificially intelligent solutions to boost research in the area cardiology. They had plans of partnering with Apollo Hospitals, one of the largest health systems in India to bring solutions. May Apollo Hospitals announced that they will be adopting Watson for Oncology and Watson for Genomics. The two IBM cognitive computing platforms will help physicians provide patients with personalised, evidence-based cancer care. December GE Healthcare announced new applications and smart devices built on Edison – a platform that helps accelerate the development and adoption of artificial intelligence and empower providers to deliver faster, more precise care. MedAchievers Academic Council and LabIndia Healthcare launched an AI-based open orthopaedic surgery simulator which will be used for training orthopaedic and neurosurgeons. Others January The Indraprastha Institute of Information Technology (IIIT) in Delhi said that it is working on a driverless autonomous electric rickshaw to provide last-mile connectivity to urban Indians. February Dubai’s Roads and Transport Authority (RTA) launched the first on-road tests of the world’s first autonomous pods, that has been developed in cooperation with Next Future Transportation Twenty Two Motors launched its smart electric scooter, Flow, equipped with AI and cloud capabilities at the Auto Expo 2018 held at Greater Noida and New Delhi. April It was announced that passenger visiting Bengaluru airport may soon be greeted by a special robot assistant called KEMPA. The little bot assistant will answers queries of confused passengers in English as well as Kannada. December 100 million Quora users were affected by a massive data security breach. It compromised data including name, address, password, public and non-public content posted on the website and shared on other networks being leaked.","excerpt":"As the year ends, we list the news and headlines that broke the records and developments that marked new growth in the industry. With a lot of funding and investment in the new tech startups, appointments in the industry, opening up of a centre of excellence in India, hiring and more, there were a lot […]","categories":["AI Trends"],"tags":["ai dl","automotive analytics","International Affairs","new developments in cloud computing","strategic salesforce"],"author_name":"Srishti Deoras","publish_date":"2018-12-26T12:03:19","publication_year":"2018","word_count":5605,"keywords":["data science","new developments in cloud computing","artificial intelligence","machine learning","AI","neural network","ML","computer vision","NLP","automotive analytics","deep learning","analytics","International Affairs","ai dl","strategic salesforce"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","neural network","NLP","computer vision","data science","analytics"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/flashback-2018-top-news-stories-about-ai-analytics-that-made-headlines\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":6,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10172448,"title":"India Launches AI-Powered Advanced Traffic Management System in Delhi","content":"India launched an advanced traffic management system (ATMS) project on Wednesday at Delhi’s Dwarka Expressway to create smarter and safer highways, according to media reports. The framework, created by the Indian Highways Management Company Ltd (IHMCL) and supervised by the National Highways Authority of India (NHAI), adheres to the NHAI guidelines 2023 to develop smart and secure roadways. Central to this system are state-of-the-art technologies that leverage AI for traffic monitoring and management. The infrastructure includes the Traffic Monitoring Camera System (TMCS), which employs 110 high-resolution PTZ cameras spaced 1 kilometre apart for constant surveillance. Additionally, the Video Incident Detection & Enforcement System (VIDES) detects traffic offences, while Vehicle Actuated Speed Displays (VASD) offer real-time speed notifications to motorists. Variable Message Signboards (VMS) provide ongoing traffic information and safety guidelines. All these systems are overseen by a Central Command Centre, which orchestrates responses to incidents such as crashes, fog, and animal crossings, ensuring effective traffic control. The Command Centre is the highway’s “digital brain,” enabling authorities to respond swiftly to emergencies like collisions, fog conditions, or road obstructions. This system is connected to NIC’s e-challan portal, facilitating the automatic reporting of traffic offences to law enforcement. Violations like speeding, not wearing seatbelts, and riding with multiple passengers can now be monitored through AI-enabled cameras, enhancing both enforcement and data collection. Amrit Singha, chief product officer at IHMCL, mentioned that this system identifies breaches and encourages public education on road safety, according to reports. The NHAI views this initiative as a model for future smart highways in India. With faster response times and AI-driven decision-making, the ATMS is a major step towards safer and more efficient road travel. According to the Financial Express, Pune’s public transport systems are using smart cameras to monitor drivers for drowsiness, distraction, and speeding through signals. These cameras analyse data on driving patterns, brake pressure, and blink frequency to identify risky behaviours. Similarly, the Bengaluru-Mysuru Expressway also uses an AI-driven ATMS to enhance road safety and traffic control. It features AI cameras, Automatic Number Plate Recognition, and real-time monitoring to identify violations and manage incidents.","excerpt":"The Traffic Monitoring Camera System (TMCS) employs 110 high-resolution PTZ cameras placed 1 kilometre apart for continuous surveillance.","categories":["AI News"],"tags":["AI traffic system","national highways"],"author_name":"Smruthi Nadig","publish_date":"2025-06-26T15:02:07","publication_year":"2025","word_count":349,"keywords":["Go","ELT","programming_languages:R","AI","national highways","Git","AI traffic system","RAG","programming_languages:Go","R"],"extracted_tech_keywords":["AI","RAG","R","Go","Git","ELT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/india-launches-1st-ai-powered-traffic-management-system-in-delhi\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10023811,"title":"2021 1st Quarter: All Major ML Innovations &#038; Contributions","content":"From Deep Nostalgia to major Python releases, the first quarter of 2021 was chock-full of promising roll-outs. While Hyper Automation, RPA deployments, growth in MLOps tools, machine learning as a service dominated the trends, the ML domain witnessed breakthroughs in natural language processing, computer vision, conversational AI etc. In this article, we have listed the major ML innovations in the first quarter of 2021. Major Releases – JupyterLab 3.0, Tensorflow 3D, Java Version 16 & More The year started with the release of JupyterLab 3.0, a popular web-based interface for Jupyter Notebooks. JupyterLab offers many intuitive features that enable data scientists and developers to work with notebooks, code, and data efficiently. The features include Code consoles, Kernel-backed documents, and multiple views of appointments. Earlier this year, Google released Tensorflow 3D, which addresses significant challenges in computer vision. TensorFlow 3D is a highly modular library that brings 3D deep learning capabilities to TensorFlow. TF 3D can be leveraged for different 3D deep learning research types like quick prototyping and deploying real-time inference systems. Oracle’s Java Version 16 has been released with 17 new enhancements, including language improvements. The latest Java Development Kit (JDK) includes features like Pattern Matching, Records, packaging tools, elastic metaspace and more. The additions are aimed at helping machine learning developers to accelerate tasks. Recently, Python’s popular open-source machine learning library, PyTorch announced its new performance debug profiler, PyTorch Profiler, along with its 1.8.1 version release. With a new module namespace torch.profiler, PyTorch Profiler is the successor of PyTorch autograd profiler. The tool uses a new GPU profiling engine — built using the NVIDIA CUPTI APIs — and can capture the GPU kernel events with high fidelity. PyTorch includes a simple profiler API that is useful when determining the most expensive operators in the model. Lately, Google Announced TensorFlow Quantum 0.5.0 at the first anniversary of TensorFlow Quantum (TFQ), a library for rapid prototyping of hybrid quantum-classical ML models. It integrates hybrid quantum with classical machine learning models that help researchers excel in quantum research. Low-Code Programming Language, Power Fx At the recent Microsoft Ignite event, the company announced a new open-source low-code programming language, Power Fx.  A strong-typed, general-purpose, declarative and functional programming language, Power Fx can directly edit apps in text editors like Visual Studio Code or an Excel-like formula bar. The language is based on Microsoft Excel. ML Packages H2O.ai introduced an open-source automatic ML package for wave apps that builds and integrates predictive AI\/ML models into Wave apps. Wave ML lets users rapidly develop and deploy interactive, predictive and decision-support applications over the web. H2O’s Wave ML makes machine learning easy for developers to solve business problems with the power of AutoML. MPD Ventures announced a ‘Machine Learning Cloud’ initiative featuring AMD Instinct MI100 accelerators along with the AMD ROCm open software platform. The platform will be initially hosted at AMPD’s DC1 data centre in Vancouver, British Columbia, and might expand into other territories in the future. It unlocks various open compute languages, compilers, libraries, and tools designed from the ground up. Infosys launched Infosys Cobalt to democratise AI within the workforce and drive business transformation. Built on NVIDIA DGX A100 systems, it can offer massive compute density. Deep Nostalgia Deep Nostalgia uses deep learning algorithms to create animated videos from old images. Deep Nostalgia can very accurately apply the drivers to a face in a still photo and animate it. It can make a picture smile, blink etc. Geoff Hinton’s GLOM Geoff Hinton’s GLOM makes neural networks smarter. In GLOM architecture, the scene-level top-down neural network converts the scene vector and the image location into an appropriate object vector for that location. It includes information about the 3-D pose of the object relative to the camera. In GLOM, wrote Dr Hinton, a percept is a field, and the shared embedding vector that represents a whole is very different from the shared embedding vectors that represent the parts. Hugging Face Hugging Face gained a lot of traction after it raised $40 million in Series B funding led by Addition. What started as a chatbot company soon transformed into an open-source provider of NLP technologies to companies such as Microsoft Bing. It extends access to conversational AI by creating abstraction layers for developers. It helps users to adopt conversational AI technologies such as BERT, XLNET, and GPT-2. Datasets Google introduced the ToTTo Dataset to overcome the hallucination problem. The ToTTo dataset consists of 121,000 training examples and has 7,500 examples each for development and test. The team at Google claimed ToTTo is a suitable benchmark for research in high precision text generation. Hallucination refers to generating text not faithful to the source. In most cases, the hallucination occurs due to divergence between the source and reference. Hallucination happens when the system catches on to wrong correlations between different training data parts. ImageNet, one of the world’s most influential AI datasets, decided to blur people’s faces in its database to respect user privacy. ImageNet started sourcing WordNet, a database of English words categorised by synonyms. It used amazon Mechanical Turk Workers to collect images of thousands of objects and people without their explicit consent. Since then, the database has expanded to over 1.5 million images categorised under 1000 words. Unified Transformer Researchers from Facebook AI Research (FAIR) introduced a new Transformer model called Unified Transformer (UniT). It can learn tasks across multiple domains in a simultaneous manner. The uniT can also take images and texts as inputs and train the input on various tasks ranging from visual perception and language understanding to vision and language reasoning. Language Models Researchers created GPT-Neo, an open-source cure for GPT-3, a language model released last year with 175 billion parameters. Since then, researchers have been looking to create an open-source version of GPT-3. GPT-Neo is the brainchild of EleutherAI. The stated goal of the project is to replicate a GPT‑3 DaVinci-sized model and open-source it for free. Google researchers showcased that language models can pre-trained up to trillion parameters. Switch Transformers, a technique to train language models, can create a system that would increase the parameter count while maintaining the floating-point operations (FLOPS) per input constant. DeBERTa Decoding-enhanced BERT with dis-entangled attention or DeBERTa is a new and improved BERT model architecture that claims to improve the performance of Google’s BERT and Facebook’s RoBERTa models. Microsoft’s DeBERTa model outperformed T5 with 11 billion parameters on the SuperGLUE benchmark and surpassed the human baseline. DeBERTa is essentially a new Transformer-based neural language model that proposes a disentangled self-attention mechanism. Multimodal Neural Networks Researchers at OpenAI discovered neural networks within AI systems that resemble the neural network inside the human brain. These neural networks can respond to a cluster of abstract concepts rather than just specific visual features. It can respond to a range of emotions, animals, photographs, drawings and learn visual concepts from natural language supervision. Further, this general-purpose vision system can match the performance of a ResNet-50 and outperform existing vision systems on the most challenging datasets. New Benchmark For Meta Reinforcement Learning DeepMind and University College London researchers released a principled benchmark for meta-reinforcement learning (meta-RL) research, known as Alchemy. The combination of structural richness and structural transparency increases the flexibility and sample efficiency of reinforcement learning. It addresses hurdles such as scarcity of adequate benchmark tasks; lack of support for principled analysis etc. OpenAI’s DALL.E OpenAI released a 12-billion parameter version of GPT-3, called DALL.E, a transformer that can generate images from text prompts. The name is a portmanteau of painter Salvador Dali and Pixar movie, WALL.E. DALL.E can render an image from scratch and alter aspects of an image based on text prompts. DALL.E model is also trained for working with multiple objects in an image. The OpenAI team has tested DALL.E’s capabilities against other specific situations, such as generating 3D imagery, cross-sectional views, and images based on contextual text caption. IBM’s Molecule Generation Experience (MolGX) Molecule Generation Experience (MolGX) is a cloud-based, AI-driven platform to design novel molecular designs. It usually takes over ten years and $10–100 million in funding to discover a new molecule. With the MolGX platform, researchers can bring down the time required for the process. MolGX is based on the idea of reverse designing, where AI is used to draw realistic images of landscapes or portraits of people that don’t even exist. AI Reviews Scientific Papers The machine learning community churns out a lot of research papers at international conferences. However, the review system for these papers is not up to the mark. Researchers at Carnegie Mellon University used natural language processing to review them.","excerpt":"Earlier this year, Google released Tensorflow 3D, which addresses significant challenges in computer vision.","categories":["AI Features"],"tags":[],"author_name":"Srishti Deoras","publish_date":"2021-04-12T17:00:00","publication_year":"2021","word_count":1433,"keywords":["machine learning","OpenAI","AI","neural network","ML","MLOps","computer vision","NLP","Aim","deep learning"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","neural network","NLP","computer vision","OpenAI","MLOps","Aim"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/2021-1st-quarter-all-major-ml-innovations-contributions\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":37401,"title":"Slovenia To Set Up Europe&#8217;s First International AI Research Centre","content":"Slovenia this week announced its plans, to set up Europe’s first international artificial intelligence research centre, which will have the official support of the UNESCO. At the gala celebration of the 70th anniversary of Jožef Stefan Institute (JSI), Slovenia’s leading scientific research establishment, Slovenian Prime Minster Marjan Šarec announced that the Department of Intelligent Systems at the JSI in Ljubljana will be converted into a centre that focuses on the governance and policies surrounding AI. Prime Minister Šarec, said, “All of this Slovenian know-how which has been applied for all these years, and all the knowledge that we possessed in the past and that we still possess today is undoubtedly a reason for us, or should I say you, to be proud… Nevertheless, this government is aware of the importance of science, research and development. Of course, the government cannot guide all of this, but it can contribute funding and above all ensure that those funds are allocated correctly and used sensibly.” The JSI grew out of the physics department of the Slovenian Academy of Sciences and Arts and was officially founded in 1949. The institute bears the name of renowned Slovenian physicist Jožef Stefan, the only Slovene to have a physical law named after him – Stefan’s law of radiation. The institute employs nearly 1,000 people, who work hard to promote technological, social and economic development in Slovenia and internationally. Reportedly, the institute and the committee will focus on various ways of developing and improving capacities, such as setting up other research centres around the world, developing training courses, raising global awareness, and creating a network of data exchange.","excerpt":"Slovenia this week announced its plans, to set up Europe’s first international artificial intelligence research centre, which will have the official support of the UNESCO. At the gala celebration of the 70th anniversary of Jožef Stefan Institute (JSI), Slovenia’s leading scientific research establishment, Slovenian Prime Minster Marjan Šarec announced that the Department of Intelligent Systems […]","categories":["AI News"],"tags":["AI (Artificial Intelligence)","unesco"],"author_name":"Prajakta Hebbar","publish_date":"2019-04-08T08:33:18","publication_year":"2019","word_count":270,"keywords":["Go","funding","artificial intelligence","programming_languages:R","AI","unesco","programming_languages:Go","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","R","Go","funding","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/slovenia-to-set-up-europes-first-international-ai-research-centre\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":48219,"title":"8 Toolkits For Reinforcement Learning Models That Make Reasoning &#038; Explainability Core To AI","content":"Reinforcement Learning has emerged as one of the most promising areas of machine learning. Reinforcement Learning techniques — the ability to learn and make decisions through the interaction with the environments has made several breakthroughs in the last few years. From robotics to recommendations, researchers are doing their level best in this domain. In this article, we list down 8 toolkits which can be used to deploy reinforcement learning models. (The list is in alphabetical order) 1| ChainerRL, a Deep Reinforcement Learning library ChainerRL is a deep reinforcement learning library implements that has several deep reinforcement algorithms in Python. This toolkit uses a subset of the interface and can be applied to a wide range of problems. It also provides various agents, each of which implements a deep learning reinforcement algorithm. 2| Facebook’s ReAgent Recently, researchers at Facebook AI Research (FAIR) introduced a reinforcement learning toolkit, ReAgent which is designed to streamline the process of building models which make and rely on decisions. This toolkit not only helps in creating AI-based reasoning systems but is also the first to include policy evaluation that incorporates offline feedback to improve models. The online social media giant has been leveraging this toolkit to drive tens of billions of decisions every day. There are three main resources behind this toolkit, they are models, an evaluator, and a serving platform. 3| Keras-RL Keras is a popular neural network library which is built in Python. keras-rl is a library for Deep Reinforcement Learning with Keras which implements some state-of-the-art deep reinforcement learning algorithms in Python and seamlessly integrates with Deep Learning library Keras. 4| PyTorch PyTorch is one of the most popular Deep Learning libraries written in Python. The library has been applied in various deep reinforcement learning algorithms like Deep Q Learning (DQN), Deep Deterministic Policy Gradients (DDPG), Diversity Is All You Need (DIAYN), Stochastic NNs for Hierarchical Reinforcement Learning (SNN-HRL) and Proximal Policy Optimisation (PPO), among others. 5| RLCard: A Toolkit for Reinforcement Learning RLCard is an open-source toolkit for reinforcement learning research in card games. It supports various card environments with easy-to-use interfaces, including Blackjack, Leduc Hold’em, Texas Hold’em, and many more. The main goal of this toolkit is to bridge the gap between reinforcement learning and imperfect information games. 6| SURREAL: Open-Source Reinforcement Learning Framework SURREAL (Scalable Robotic REinforcementlearning ALgorithms) is an open-source framework which facilitates reproducible deep reinforcement learning (RL) research for robot manipulation. The toolkit decomposes a distributed RL algorithm into four components — generation of experience (actors), storage of experience (buffer), updating parameters from experience (learner), and storage of parameters (parameter server). This decoupling of data helps in eliminating the need for global synchronisation and improves scalability. 7| The MAME RL Algorithm Training Toolkit The MAME RL algorithm training toolkit is a Python library which is used to train the reinforcement learning algorithm on almost any arcade game. This toolkit is currently available on Linux systems and works as a wrapper around MAME. It allows the algorithm to step through gameplay while receiving the frame data and internal memory address values for tracking the games state, along with sending actions to interact with the game. 8| TensorFlow This popular, open-source machine learning platform includes TF-Agents which is a library for Reinforcement Learning in TensorFlow. In this toolkit, the core element of RL algorithms is implemented as Agents wherein an Agent includes two main responsibilities  — defining a policy to interact with the environment and the other is to learn or train that policy from the collected experience. Wrapping Up From beating top chess players to driving better insights for healthcare sector, Reinforcement Learning has been used by researchers to solve a number of real-life problems. The developer community has been utilising these tools for training agents as well as introducing advancements in trained agents. There are also popular platforms like OpenAI Gym, DeepMind Lab, DeepMind Control Suite, etc. where one can develop and compare RL algorithms.","excerpt":"Reinforcement Learning has emerged as one of the most promising areas of machine learning. Reinforcement Learning techniques — the ability to learn and make decisions through the interaction with the environments has made several breakthroughs in the last few years. From robotics to recommendations, researchers are doing their level best in this domain. In this […]","categories":["AI Trends"],"tags":["policy gradient","Reinforcement Learning"],"author_name":"Ambika Choudhury","publish_date":"2019-10-18T13:00:55","publication_year":"2019","word_count":657,"keywords":["machine learning","Reinforcement Learning","OpenAI","AI","neural network","PyTorch","ML","Keras","RAG","deep learning","policy gradient","TensorFlow"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","neural network","OpenAI","TensorFlow","PyTorch","Keras","RAG"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/8-toolkits-for-reinforcement-learning-models-that-make-reasoning-explainability-core-to-ai\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":50457,"title":"Learn These Top 6 AWS Skills To Get A Job In 2020","content":"Cloud is a fulcrum for organisations to innovate and get ahead of the pack in the ever-evolving technology landscape. Consequently, companies are seeking professionals who obtain various cloud skills and can assist them in achieving their business objectives. However, there are a plethora of services that cloud providers offer — and that’s why determining the best skills to learn can be a daunting task. To address this pressing issue, Analytics India Magazine brings to you the most in-demand services in the cloud. And since AWS is the learning cloud service provider, we have focused on its services to assist you in getting a job in 2020. SageMaker AWS SageMaker is a one-stop-shop for data scientists and AI initiatives by organisations to train, evaluate, and deploy machine learning models. SageMaker works in tandem with EC2, S3 and others. Using SageMaker, firms are able to quickly go from conceptualisation to production with their machine learning models. Therefore, one must master this AWS skill to have an edge against competitors for getting a job. Kinesis Kinesis enables firms to deploy data streaming practices to help them leverage data beyond batch processing. Besides, it allows them to monitor and get insights into data from IoT devices. Today, firms are highly focused on real-time analytics, and IoT devices, thus it makes a must-have skill for any professional to demonstrate their proficiency. Elastic Compute Cloud (EC2) The service empowers you to create virtual computing environments in the cloud for building scalable and enterprise-class applications. Such bespoke compute capacity in the cloud allows developers to effortlessly build applications. The service has various features such as bare metal instances, GPU compute instances, high I\/O instances, and more. Being an expert in managing EC2 will be a huge advantage while showcasing your skills in cloud computing. Simple Storage Service (S3) S3 is an object storage service to protect data for a wide range of use case. The scalable storage service is used with almost every service to full-fill the requirements of information in several instances. It also offers data control, supports different cloud storage services with just a few clicks. Some of the use cases of this service are backup and restore, archive, data lakes and big data analytics, hybrid cloud, and cloud-native application data. Having command over this service will enable you to handle the complete data strategy in organisations. Amazon’s Identity and Access Management (IAM) Security is the most essential aspect in any organisations as any compromise in it can have a negative impact on the businesses. Therefore, firms expect professionals to be familiar with security features to manage user permissions for fortifying unauthorised access. IAM provides fine-grained access control to AWS resources, multi-factor authentication, among others. AWS Cloud9 Cloud9 is an integrated development environment that runs on the cloud. This eliminates the management required for handling version conflict for projects and allows sharing of project environment intuitively. Collaborating and making changes in one file drastically enhances the speed as more than one user can code to complete the projects. Outlook Simply knowing the programming skills and domain-specific skills does not differentiate you from others. Consequently, you should go the extra mile and learn these skills to gain the attention of recruiters by exhibiting your dexterity in the cloud. AWS offers numerous services that cater to the needs of organisations right from small to medium and large. Thus, based on the size the requirement of services varies but the aforementioned services are the basic requirements for streamlining the workflow in the cloud. While there are among the most widely used, you can focus on others depending upon the job you are applying for.","excerpt":"Cloud is a fulcrum for organisations to innovate and get ahead of the pack in the ever-evolving technology landscape. Consequently, companies are seeking professionals who obtain various cloud skills and can assist them in achieving their business objectives. However, there are a plethora of services that cloud providers offer — and that’s why determining the […]","categories":["AI Hirings"],"tags":["AWS services"],"author_name":"Rohit Yadav","publish_date":"2019-11-22T10:00:00","publication_year":"2019","word_count":606,"keywords":["Go","machine learning","AWS services","AWS","AI","cloud computing","ML","Scala","RAG","analytics","R"],"extracted_tech_keywords":["AI","machine learning","ML","analytics","RAG","cloud computing","AWS","R","Go","Scala"],"url":"https:\/\/analyticsindiamag.com\/ai-hiring\/learn-these-top-6-aws-skills-to-get-a-job-in-2020\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10009843,"title":"The World Seems To Be Divided On AI: New Survey By Oxford","content":"“54,195 respondents in 142 countries were part of this survey involving AI in our personal affairs and public life.” The word AI is out now — making its way across the research labs into our everyday lives. The AI that is in use today actually forms a subset of AI, called machine learning. Nevertheless, significant enough. As the nations explore ways to leverage the power of algorithms, the government will inevitably run into the flaws of algorithms; biases have been a massive challenge for quite some time now. Be it the biases or the insecurity surrounding AI takeover, the world has been divided over inclusion of AI. To understand the sentiments across the globe, and their attitudes towards AI, the Oxford Commission on AI and good governance conducted a survey. About The Survey The data for this survey was taken from the 2019 World Risk Poll. Based on which the researchers at Oxford analysed public opinion drawn from a sample of 154,195 respondents living in 142 countries. More than five thousand interviewers conducted the interviews between 8th of May, 2019 till 17th of January 2020. According to the survey report, at least 1,000 respondents were surveyed from each nation, and the sampling population represented the entire population aged fifteen and older. Key Findings Source: Oxford Commission The participants were asked whether “machines or robots, often known as artificial intelligence” will “mostly help or mostly harm people in the next 20 years”. Respondents were given three options to choose between “mostly harm”, “mostly help” and “neither”. “One outlier, for instance, is China, where only 9% of respondents believe AI will be mostly harmful.” The above table illustrates public perception (in %) to the above question, by region. The survey has also reported the sentiments of people based on their profession. While the white-collar jobs shared a positive attitude, construction and manufacturing workers found AI to be harmful. As per the report, the findings can be summarised as follows: North America (47%) worries the most about AI being harmful. Whereas, respondents in Southeast Asia (25%) and East Asia (11%) concerned that AI will be harmful. China showed great enthusiasm for AI of all the nations with only 9% thinking that AI will be harmful. When it comes to professions, business and government executives with 47% and other professionals with 44% are among the most enthusiastic about AI decision making. Whereas workers in manufacturing with 35% and service workers with 35% are less confident. Future Direction The report stated that each major misstep in AI could have consequences for government agencies involved AI systems in governance. The fact that few professions find AI to be harmful also stems from deep-seated insecurity with regards to their employment. Industries heavily invest in automating mundane work. Even the software industry is not exempt from the reach of AI. There is active research to build software that writes new software. Historically, whenever a transition obscured a technology or jobs, it gave rise to newer opportunities, careers which have never been thought of before. For example, data ethnographer is a thing! AI is here to stay. So, now the question is how to make the transition more public-friendly. In order to inculcate accountability and transparency, few European nations have already made a move. The governments of the Netherlands and Finland have already set their initiatives in motion this month by opening up AI registries. AI Register is a window into the AI systems used by these cities through the register. The people in these cities can get acquainted with the quick overviews of the city’s AI systems or examine their more detailed information based on their own interests. They can also give feedback and thus participate in building human-centred AI. This survey by Oxford Commission has exposed the divided perceptions and made a case for developing AI for good governance and underlined the importance of civic enthusiasm for substantive outcomes. Check the full report here.","excerpt":"“54,195 respondents in 142 countries were part of this survey involving AI in our personal affairs and public life.” The word AI is out now — making its way across the research labs into our everyday lives. The AI that is in use today actually forms a subset of AI, called machine learning. Nevertheless, significant […]","categories":["AI Features"],"tags":[],"author_name":"Ram Sagar","publish_date":"2020-10-16T12:00:44","publication_year":"2020","word_count":657,"keywords":["Go","machine learning","artificial intelligence","AWS","AI","cloud_platforms:AWS","programming_languages:R","RAG","ViT","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","RAG","AWS","R","Go","ViT","cloud_platforms:AWS","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/global-attitudes-ai-sentiments-oxford-survey\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":37766,"title":"Dream 11 Is Upping The Fantasy Game Market With Its AI-Powered Platform","content":"There is no denying the fact that sports in India is a big industry. Started in 2008, Indian Premier League is a cricket league that has taken cricket to a whole new world, making India the hotbed for cricket enthusiasts. To push the sports industry in India ahead, technology over the years has not left any stone unturned, and one of the most popular concepts is fantasy sports. Fantasy sports is a type of online game where participants form an imaginary or virtual team of real players of a professional sport and then compete based on the statistical performance of those players’ in actual games. What started as a concept, today is one of the rising markets in India. Fantasy sports platforms are getting significant funding from prominent VCs and the popularity is getting bigger and better. Started in 2012, Dream11 is India’s Biggest Sports Gaming platform with more than 4 crore users playing Fantasy Cricket, Football, Kabaddi and NBA. The Series D funded startup’s platform allows Indian sports fans to showcase their sports knowledge. When Sports Meets AI We all know that AI over the years has found applications in several industries and time again the tech has proved that it has a major business impact — serving clients faster, better and make more money. However, it is not just the businesses that are using artificial intelligence, sports gaming platforms are also in the rat race to become top AI-driven platforms. And being a top sports gaming platform, Dream 11 is leveraging the superpowers of AI to make the experience better for its users. Use Cases AI-Based Customer Support Bot When it comes to fantasy sports in India, the IPL season is something that is really critical for platforms like Dream 11 —not only in terms of performance but also in terms of customer support. In order to deliver top-notch customer support, Dream 11 uses an AI-based bot. Created by Haptik, the Dream 11 customer support bot is a conversational AI that handles the massive scale of incoming requests by delivering instant responses without making anxious users wait. Also, the bot was designed and trained in such a way that it was able to answer most of the game related queries. AI-Based Automated Segmentation Customer support is not the only vertical where Dream 11 is making use of artificial intelligence, the fantasy sports platform also makes use of AI in craft personalized, contextual, and timely engagement campaigns to drive higher app monetization and retention. And for that, user-level behavioural insights are a must. Dream11 make use of California-based mobile analytics and mobile marketing company, CleverTap’s automated segmentation capability that uses RFM (recency, frequency and monetary) analysis to segment its user base into 10 distinct segments. The major benefit of RMF is that the platform can keep track of its users’ activity and their monetary values and that helps Dream 11 in curating user-centric campaigns that reduce at-risk and hibernating users over time. Also, the company can now build highly effective push notification campaigns based on specific historical and real-time in-app user actions or inactions. Now, using the AI-based advanced marketing automation platform, Dream 11 extensively taps into the sporting interests and loyalties of sports fans and provide a great product experience. Outlook The Sports industry in India is booming at a rapid pace — not only on the pitch but also in terms of sports-based application. And sought after technologies like AI and ML are becoming the backbone of those platforms. While the betting industry is already leveraging machine learning and artificial intelligence, fantasy gaming platforms are at a nascent stage. However, with use cases like Dream 11, it seems the fantasy gaming industry is also set to take a big leap when it comes to leveraging AI and ML to drive business.","excerpt":"There is no denying the fact that sports in India is a big industry. Started in 2008, Indian Premier League is a cricket league that has taken cricket to a whole new world, making India the hotbed for cricket enthusiasts. To push the sports industry in India ahead, technology over the years has not left […]","categories":["AI Features"],"tags":["artifical intelligence"],"author_name":"Harshajit Sarmah","publish_date":"2019-04-15T12:31:02","publication_year":"2019","word_count":634,"keywords":["API","machine learning","artificial intelligence","AI","ML","RAG","automation","ViT","analytics","R","artifical intelligence"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","analytics","RAG","R","API","ViT","automation"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/dream-11-is-upping-the-fantasy-game-market-with-its-ai-powered-platform\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10002510,"title":"Will Power BI Overshadow SSRS","content":"Microsoft has been advancing its offerings in analytics and business intelligence for a long. It has progressed significantly, especially in providing advanced business intelligence and reporting platforms such as Power BI, SQL Server Reporting Services (SSRS) and more. While SSRS is focused on paginated reports, Power BI tool is more about data discovery and interactive reporting letting business analysts connect, prepare and explore data. While SSRS has been efficiently handling the reporting task, with the popularity of Power BI on the rise, it is often debated that PowerBI will overshadow SSRS. Let us dive right away to understand facts. Power BI Vs SSRS Launched in 2004, SQL Server Reporting Services is a report generating platform that includes an integrated set of processing components and programmatic interfaces. It is server-based and is used for generating, viewing and sharing reports. It often requires manual efforts owing to lesser graphical and drag and drops features, however, it provides a good drill-down feature. Accessible through desktops and web, it requires to make a purchase of SQL Server License to use it. SSRS requires an understanding of programming which makes it less user-friendly. It can only be used for structured and semi-structured data. Power BI, on the other hand, is a data analysis tool that is mostly used by business analysts to work on huge and complex data. It is used for modelling and structuring unshaped data with advanced features. A relatively newer tool, launched in 2017, it can generate, publish and share dashboards for both server-based and cloud-based platforms. It is accessible in mobile, desktop and web browsers. Also, its drag and drop features with rich GUI make it more user-friendly. It can easily work on structured and unstructured data. Another factor is that Power BI is free. Given the advanced features and user-friendly nature such as efficient data visualisation and richer graphical experience, Power BI is often preferred over SSRS as a reporting server. How Is Business Intelligence Different From Data Science7 Business Intelligence Tools Every Enterprise Must UseWith Tableau & Looker Gone, Is Microsoft Power BI Positioned To Take The #1 LeadBusiness Intelligence in the New Era Of 2013A Look At The Evolution Of BI Platform Over The Years Will Power BI Replace SSRS Many professionals who have worked on both tools believe that while installation, configuration and management with Power BI is easier compared to SSRS, both have their own technicalities and drawbacks. Having said that, the company is coming up with new features for Power BI tools quite frequently, creating advanced versions of it. With better services and offerings, there is no doubt that Power BI will be preferred over SSRS in many cases. Some of the other advantages such as intricate visualisations, creating dashboards and reports with easy deployments, strong community service, mobile support, and more, making Power BI a preferred tool. We will see a lot of companies with newer projects making a move into Power BI as it is more accessible to learn, develop, produces modern interactive results and can be deployed without any infrastructure. While SSRS  keeps getting makeovers and changes over time and has some of the strong features such as better on-premise security control, delivery reports in multiple formats, it is seeing less of a demand than Power BI. Also, the fact that the company has provided detailed steps on how to migrate SSRS reports to Power BI, suggests that many companies are looking to make a shift. It is therefore not unlikely to see lesser use of SSRS in the coming future but may not become completely obsolete. There are certain functionalities of SSRS such as creating tabular reports, multiple export options, the option to schedule report delivery and others that Power BI lacks. Wrapping Up Having understood some of the advantages and disadvantages, it can be said that the objectives and functionalities of the tools are not completely different but overlapping some of their features may provide the best results. While Power BI is a good tool for data exploration, SSRS is good for printed reports. The preference of the tool is largely dependent on the requirements. While Power BI provides interactive visualisation and dashboard creation, it cannot archive data in excel format. Similarly, creating dynamic reports is a far fetched result in SSRS. The choice of tool will, therefore, largely depend on the business requirement and the purpose of reporting.","excerpt":"Microsoft has been advancing its offerings in analytics and business intelligence for a long. It has progressed significantly, especially in providing advanced business intelligence and reporting platforms such as Power BI, SQL Server Reporting Services (SSRS) and more. While SSRS is focused on paginated reports, Power BI tool is more about data discovery and interactive […]","categories":["AI Features"],"tags":["Power BI"],"author_name":"Srishti Deoras","publish_date":"2020-07-15T14:28:32","publication_year":"2020","word_count":729,"keywords":["business intelligence","Power BI","data science","Go","programming_languages:R","AI","programming_languages:SQL","RAG","analytics","SQL","R"],"extracted_tech_keywords":["AI","data science","analytics","RAG","R","SQL","Go","business intelligence","programming_languages:R","programming_languages:SQL"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/will-power-bi-overshadow-ssrs\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10047474,"title":"How to Prepare for Certified Analytics Professional","content":"In the competitive world, almost all professionals want to stand out from the crowd by showing some extra achievements. Having good certifications in the relevant field is the best way to gain attention. In the field of data science and analytics, there is a wide range of certification opportunities available belonging to the different domains that certify the expertise of holders in their skills. Certified Analytics Professional (CAP) is such a premier and classic certification intended for analytics professionals. Here in this article, we discuss the CAP certification program with its expectations and the steps to prepare for the same. About Certified Analytics Professional Certified Analytics Professional (CAP) is a certification program offered by INFORMS (Institute for Operations Research and the Management Sciences), an international professional society of analytics and operations research professionals. Through this certification, INFORMS aims to validate the analytics knowledge and ability of analytics professionals. It also aims to differentiate certified analytics professionals from its peers and enhance the analytics knowledge and abilities of professionals. It sets a standard by which the organizations can identify and develop potential analytics professionals. INFORMS is considered as the first professional society that developed a certification program for analytics professionals. The major highlights of this certification are given below:- To achieve this credential, the candidates need to pass an exam that is facilitated by INFORMS.This exam is based on the Job Task Analysis (JTA) that is prepared by the job-task-analysis working group of INFORMS. The members of this group carry a broad background in the analytics field.The content of this certification is software and vendor-neutral. There is a process for the renewal of this certification. There are eligibility criteria for the candidates to apply for this certification. These criteria are given below:- Educational Qualification: A BA\/BS or MA\/MS is required in either of the areas of analytics, operations research, management science, statistics, engineering, business, mathematics, information technology and computer science.Experience: 3 years of experience with MA\/MS in related areas, or 5 years of experience with BA\/BS in related areas, or 7 years of experience with BA\/BS or higher degree in unrelated areas. The CAP Exam To become a Certified Analytics Professional, candidates must go through an exam. This exam can be taken from the designated test centres across the world. To register for this exam, the candidates need to pay a fee of $695. For INFORMS members, this fee is $495. Following are the important details about the exam:- Duration: 3 HoursNumber of questions: 100Format: Multiple choice questions This exam assesses the breadth of knowledge of the candidate across seven domains. These domains and their weightage is given below:- Business Problem (Question) Framing (12-18%)Analytics Problem Framing (14-20%)Data (18-26%)Methodology (Approach) Selection (12-18%)Model Building (13-19%)Deployment (7-11%)Model Life Cycle Management (4-8%) The questions from these domains are based on the Job-Task-Analysis (JTA) and are not based on any specific software, platform or vendor. There is a criterion-referenced methodology used by INFORMS to determine the passing score for its exams. The candidates receive an official score immediately after completing the exam. Steps to Prepare The Certified Analytics Professional (CAP) exam is a little different from the other certifications. It is software and vendor-neutral and completely based on Job Task Analysis. There are limits on the attempts also as a candidate can only appear a maximum of 3 times in a year. So there is the need to prepare with good planning. To prepare for the CAP exam, the following steps will be really helpful for the candidates:- Understand the Pattern and Collect Basic Information: You can visit the official website of the CAP exam to collect the important information related to the certification. From this website, you can obtain the complete handbook of the exam by sharing your Email ID. The handbook gives you all the information about the certification including a set of sample questions as well. Based on the given domains and sample questions, you can understand the pattern, breadth and depth of the exam. Collect Required Learning Materials: In the exam handbook, you will find a list of reference books across the domains. Along with that, you can obtain the Job Task Analysis by sharing your Email ID on this page. These referred books and the job task analysis will be almost sufficient to prepare for this exam. Learn and Prepare Notes with a timeline: Based on your time management and understanding of the breadth and depth of the exam, you can define a timeline for your preparations. You should go through the referred book based on the domains of the exam and start creating the notes one by one. Mock Test on MachineHack: MachineHack, India’s leading hackathon and assessment platform, provides different types of mock tests for such professional certification exams. These mocks are provided for free, just to support the candidates in their careers. These are prepared based on the details as given by certifying organizations and feedback from the successful candidates. You can also find Certified Analytics Professional (CAP) exam mock test on MachineHack. This mock test will help you to assess your level of preparation for the exam. In this mock, you will find the same number of questions across the required domains to be attempted within the stipulated time. Through this mock, you will gain the experience of taking the CAP exam in real life. Final Words CAP is one of the standard certifications in the analytics profession. The pattern of this exam is unique as it is different from the vendor-based mainstream certifications. It needs to have a clear understanding of the contents that need to be covered for this exam. Along with the preparation, the mock test taken on the MachineHack platform helps the candidate a lot for the self-assessment of their preparation. After following the preparation steps as discussed above, this exam will not be difficult to crack. Good luck to the aspiring candidates!","excerpt":"In the competitive world, almost all professionals want to stand out from the crowd by showing some extra achievements. Having good certifications in the relevant field is the best way to gain attention. In the field of data science and analytics, there is a wide range of certification opportunities available belonging to the different domains […]","categories":["AI Highlights"],"tags":["analytics certification","Certification","Data Analytics Certification"],"author_name":"Dr. Vaibhav Kumar","publish_date":"2021-09-02T13:00:00","publication_year":"2021","word_count":980,"keywords":["data science","Go","programming_languages:R","AI","Certification","analytics certification","programming_languages:Go","Aim","analytics","GAN","Data Analytics Certification","R"],"extracted_tech_keywords":["AI","data science","analytics","Aim","R","Go","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/how-to-prepare-for-certified-analytics-professional\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10043417,"title":"Guide To Simple Object Detection Using InceptionResnet_v2","content":"The object detection technique is considered one of the most challenging tasks in computer vision, a subset of artificial intelligence, as it involves object classification and localising the object within the image or video. Object detection is a computer vision technique that detects objects such as animals, persons, cars, buildings, etc. It has been applied widely over video surveillance, self-driving cars and object tracking. When performing standard image classification, we present that image to the neural network for a given input image and obtain a single class or label of the most dominating object in the image with a probability score associated with it. Whereas object detection built on image classification tries to localise the object with the help of a bounding box and the probability\/confidence score associated with each class. The following are the few common architectures used for objection detection; R-CNNFAST R-CNNFASTER R-CNNMASK R-CNNYOLO (You Only Look Once)SSD ( Single Shot Multibox Defender) Today in this article, we are going to perform object detection using a transfer learning method. We will be using InceptionResnet_v2 as our pre-trained model for this task. Brief Introduction of the InceptionResnet_v2 architecture: Deep convolutional networks have been at the central point when it comes to image-based tasks. The version of the inception network has shown that it can achieve very high accuracy at a relatively low computational time. The K. He, the author, introduces residual connections to deep learning, demonstrating how the residual connection has inherent importance in training deep networks. As the inception network is very deep, it is natural to replace the filter concatenation stage of the inception with the residual network. The author believes this would allow inception to reap all the benefits of the residual approach while retaining its computational efficiency. To get more understanding of the architectures, you can check these papers, 1,2. Now it’s time to implement object detection with InceptionResnet by leveraging python. Code Implementation of InceptionResnet_v2 Model The following code implementation is in reference to the official implementation. Importing all dependencies: import tensorflow as tf import tensorflow_hub as hub # for downloading and displaying image import matplotlib.pyplot as plt import tempfile from six.moves.urllib.request import urlopen from six import BytesIO # for dataframe import pandas as pd # for drawing onto the image import numpy as np from PIL import Image,ImageColor,ImageDraw,ImageFont,ImageOps import time Helper functions to download, visualise and drawing on image: def disp_ima(image): fig = plt.figure(figsize=(18, 13)) plt.grid(False) plt.imshow(image) def get_and_reshape_img(url, width=250, height=250, display=False): ruff, name = tempfile.mkstemp(suffix=\".jpg\") response = urlopen(url) image_data = response.read() image_data = BytesIO(image_data) pil_ima = Image.open(image_data) pil_ima = ImageOps.fit(pil_ima, (width, height), Image.ANTIALIAS) pil_ima_rgb = pil_ima.convert(\"RGB\") pil_ima_rgb.save(name, format=\"JPEG\", quality=90) print(\"Image downloaded to %s.\" % name) if display: disp_ima(pil_ima) return name def boxes_on_image(image,ymin,xmin,ymax,xmax,color,font,thickness=4,display_str_list=()): draw = ImageDraw.Draw(image) width, height = image.size (left, right, top, bottom) = (xmin * width, xmax * width, ymin * height, ymax * height) draw.line([(left, top), (left, bottom), (right, bottom), (right, top), (left, top)], width=thickness, fill=color) display_heights = [font.getsize(ds)[1] for ds in display_str_list] # Each display_str has a top and bottom margin of 0.05x. total_height = (1 + 2 * 0.05) * sum(display_heights) if top > total_height: text_bottom = top else: text_bottom = top + total_height # Reverse list and print from bottom to top. for display_str in display_str_list[::-1]: text_width, text_height = font.getsize(display_str) margin = np.ceil(0.05 * text_height) draw.rectangle([(left, text_bottom - text_height - 2 * margin), (left + text_width, text_bottom)], fill=color) draw.text((left + margin, text_bottom - text_height - margin), display_str, fill=\"black\", font=font) text_bottom -= text_height - 2 * margin def drawing_boxes(image, boxes, class_names, scores, max_boxes=10, min_score=0.1): colors = list(ImageColor.colormap.values()) try: font = ImageFont.truetype(\"\/usr\/share\/fonts\/truetype\/liberation\/LiberationSansNarrow-Regular.ttf\",25) except IOError: print(\"Font not found, using default font.\") font = ImageFont.load_default() for i in range(min(boxes.shape[0], max_boxes)): if scores[i] >= min_score: ymin, xmin, ymax, xmax = tuple(boxes[i]) display_str = \"{}: {}%\".format(class_names[i].decode(\"ascii\"), int(100 * scores[i])) color = colors[hash(class_names[i]) % len(colors)] image_pil = Image.fromarray(np.uint8(image)).convert(\"RGB\") boxes_on_image(image_pil,ymin,xmin,ymax,xmax,color,font,display_str_list=[display_str]) np.copyto(image, np.array(image_pil)) return image Download and showcase the image from URL: img_url= \"http:\/\/1.bp.blogspot.com\/-nn23fvzDZBw\/T_HTlRZYJuI\/AAAAAAAAA5U\/wHSWnIySyww\/s1600\/best+cool+nice+cute+awesome+desktop+background+wallpapers+%252817%2529.jpg\" web_img = get_and_reshape_img(img_url, 1250, 850, True) Inferencing the architecture: Load the model from tensorflow hub module= \"https:\/\/tfhub.dev\/google\/faster_rcnn\/openimages_v4\/inception_resnet_v2\/1\" model = hub.load(module).signatures['default'] User defined function for loading the image and running the model def load_image(path): imgage_path = tf.io.read_file(path) imgage_path = tf.image.decode_jpeg(imgage_path, channels=3) return imgage_path def run_model(model, path): img = load_image(path) converted_img  = tf.image.convert_image_dtype(img, tf.float32)[tf.newaxis, ...] start = time.time() result = model(converted_img) end = time.time() result = {key:value.numpy() for key,value in result.items()} print(\"Found %d objects.\" % len(result[\"detection_scores\"])) print(\"Inference time: \", end-start) image_with_boxes = drawing_boxes( img.numpy(), result[\"detection_boxes\"], result[\"detection_class_entities\"], result[\"detection_scores\"]) disp_ima(image_with_boxes) run_model(model, web_img) With the help of pandas, we can check the scores of each object identified; here, we will check the top 10 objects. image = load_image(web_img) converted_image  = tf.image.convert_image_dtype(image, tf.float32)[tf.newaxis, ...] result = model(converted_image) One more example: Conclusion Today we have seen how the concatenation of residual networks with the inception architect enhances the overall performance of the architecture and gives accurate bounding boxes for each entity in given images. Furthermore, with this minimal code, we can easily deploy the system to the web and android platforms. References Link for colab Notebook implementationOfficial implementation on TF-Hub","excerpt":"When performing standard image classification, we present that image to the neural network for a given input image…","categories":["Deep Tech"],"tags":["Computer Vision","Guide","Neural Networks","Object Detection","pre-trained models","Transfer Learning"],"author_name":"Vijaysinh Lendave","publish_date":"2021-07-13T13:00:00","publication_year":"2021","word_count":831,"keywords":["NumPy","artificial intelligence","AI","neural network","computer vision","pre-trained models","Ray","Colab","deep learning","Object Detection","Computer Vision","Transfer Learning","TensorFlow","Pandas","Guide","Neural Networks"],"extracted_tech_keywords":["AI","artificial intelligence","deep learning","neural network","computer vision","Ray","TensorFlow","Colab","Pandas","NumPy"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/guide-to-simple-object-detection-using-inceptionresnet_v2\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10017485,"title":"Data Democratization and Governance for Responsible AI","content":"Empowerment without defined responsibility and accountability has got no meaning. The potential of data is limitless. When it comes to making AI (Artificial Intelligence or Augmented Intelligence) responsible, explainable and trustworthy, data democratization and governance will need to be discussed in parallel as they are the two sides of the same coin. Explainable AI is also very important to understand and interpret the predictions and how to further improve the predictions to ensure better decision-making and to balance it with risk and accuracy. Essentially, data democratization is around the easy accessibility of digital data and information to the average end-user. But to manage its accessibility, usability and protection, data governance procedures are required to be implemented as it ensures that data is used in the right way, by the right user and at the right time. It also brings the focus on responsibility and accountability in case something goes wrong. Data democratization and governance also lays down the foundation for managing bias, potential risks, trust and transparency, and accuracy issues in AI. It is crucial that data or information access is accompanied by a supervisory and governance framework to ensure that the information is used in compliance with operational and regulatory controls and to keep the information reliable and up to date. It needs a simple way for individuals to interpret and appreciate the information so that they can use it to speed up the decision-making process and uncover growth opportunities. The social impact or the human augmentation is one such area that is gaining greater attention when the transformative journey of AI and its implications are analyzed with the evolution and changes to the human genome. The businesses and government\/federal social responsibility cannot be shifted to an artificial system just because it is a self-learning system and evolving based on the training data it receives from the outside world or generates on its own through reinforcement learning. No doubt, the lack of availability of required data cannot make the model robust and bias-free; but the user\/business policies for the data usage must be established along with the debiasing techniques for low-frequency and high-frequency decisions. It is also very important to establish the evidence in order to confirm the benefits realized so that it can be improved further as the AI or ML (Machine Learning) model becomes more mature. AI Fairness 360 is one such open-source metrics toolkit to search and minimize the unnecessary bias in datasets and machine learning models. Mitigating bias throughout the AI lifecycle Image Source: https:\/\/www.ibm.com\/blogs\/research\/2018\/09\/ai-fairness-360\/ It checks data and model bias at three different stages- training data level, an algorithm that generates the classifier and at testing and deployment stage when the prediction is made. It continuously learns through the feedback to improve the model further, thus, making the AI system more explainable. Machine learning algorithms search for patterns in the training data that are dependent on a specific prediction in order to accurately make predictions. An algorithm, for instance, might discover the trend that seems to associate an individual with a pension income only and a saving scheme giving better returns on investment to lead a dignified life. The right understanding of the problem and the data is an undisputed fact that makes the system explainable and making the data understand easily to the user is far more important to make the data democratization effective. Yes, it is true “a picture is worth a thousand words”; but not always if the picture is not oriented well. Visualization is an intuitive way to understand the data, but making it comprehensible and drilling it down to the granular level along with its data lineage will help you to derive meaningful insights and know the reasons as to why a situation or outcome occurred. The next most important thing is an intelligent data governance strategy which must align with an overarching goal of desired outcomes\/benefits. Data governance not only lets you ensure integrity, security and compliance with laws and policies concerning data governance; but also ensures that data is available in the right format, is consistent and also helps in determining which data to keep and which to delete when no longer required. Data governance is an ongoing process as new data sources from disparate systems continue to evolve, data usage can be repurposed, and changes can happen in regulations about data security and privacy. NIST (National Institute of Standards and Technology), U.S. Department of Commerce prepared a plan for Federal Engagement in Developing Technical Standards and Related Tools highlighting the importance of AI to the future of the U.S. economy and national security and guiding Federal agencies to ensure that the nation maintains its leadership position in AI. It also stresses various AI standards-related tools pertaining to data sets in standardized formats, gathering knowledge and reasoning in AI systems, benchmarking, testing methodologies, metrics, testbeds and last but not the least, about the accountability and auditing instruments for examination of AI system. When developers and policymakers determine how to factor in risk management for individuals, societies, and society at large, legal, ethical, and societal factors may also need to be addressed. Some standards and standards-related instruments seek to provide risk management guidelines that developers and policymakers may use to regulate how to handle such potential risks. Final Thoughts Hence, to bring fairness and generate trust in the AI system, it requires the explainability, interpretability, reliability and accountability which is human-centered and the actual onus of making AI more responsible and explainable resides with us as it is for the human of the human and by the human. Data democratization and governance must go hand in hand for AI’s effective implementation. Not to undermine AI’s capability, real intelligence is still with the human who can bring more value and acceptability by showcasing the positive social impact of AI. Acknowledgments and References: Introducing AI Fairness 360, A Step Towards Trusted AI – IBM ResearchU.S. LEADERSHIP IN AI: A Plan for Federal Engagement in Developing Technical Standards and Related Tools (nist.gov)","excerpt":"Empowerment without defined responsibility and accountability has got no meaning. The potential of data is limitless. When it comes to making AI (Artificial Intelligence or Augmented Intelligence) responsible, explainable and trustworthy, data democratization and governance will need to be discussed in parallel as they are the two sides of the same coin. Explainable AI is […]","categories":["IT Services"],"tags":["ai machine learning and data analytics","artificial intelligence machine learning data","purpose of ai"],"author_name":"Gaurav Dhooper","publish_date":"2021-01-11T17:00:00","publication_year":"2021","word_count":1001,"keywords":["Go","artificial intelligence","machine learning","AWS","AI","ai machine learning and data analytics","ML","artificial intelligence machine learning data","Git","RAG","Rust","purpose of ai","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","RAG","AWS","R","Go","Rust","Git"],"url":"https:\/\/analyticsindiamag.com\/it-services\/data-democratization-and-governance-for-responsible-ai\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10056910,"title":"Machine Learning Is Solving Some Unique Problems Of Online Dating","content":"While the online dating industry is clearly booming, it remains riddled with complaints: from repeated ghosting, exhaustion with the sensory overload of potential matches, the proliferation of fake accounts, to reports of harassment—dating apps can be exhausting. While technology is the cause of this fatigue, machine learning and artificial intelligence may be the key to solving the systemic problems that pervade online dating apps—which some companies have already begun exploring. AI is already being used in dating apps today The dating app Hinge already employs machine learning as part of its algorithm by suggesting a “Most Compatible” match to its users. The app separates its users based on who has liked them and then tries to find patterns within those likes. If, for example, a person likes someone named Sam, then they may also be interested in the people who received likes from the people who also liked Sam. Kevin Teman is trying to take the revolutionization of dating apps a step further. He is the founder of the Denver-based startup AIMM—an acronym for “artificially intelligent matchmaker.” The significant difference between this app and those that have come before it lies in its implementation of AI software. The users of the app are in constant conversation with a female-sounding software that coaches, encourages, advises, and gives its users feedback throughout the process in which they are matched with a person, have their first phone call with them, and they go on their first date together. The purpose of this voice is to mimic human interaction, and it takes a full week for the app to “get to know you.” Match.Com employs a similar artificially intelligent program to help its users find love with the chatbot Lara—which was first launched in France in 2016 and in the UK the following year. The bot serves as a wing woman who communicates via typing and helps the website’s users set up their dating profiles—giving users compliments and asking them questions that will help them find a match. According to experts, this encourages people to be more honest about what they want, instead of simply writing what they think would draw the most interest from potential matches. Meanwhile, the UK-based dating app Badoo uses facial recognition technology in its ‘Lookalikes’ feature—a feature that allows people to pair with people who look like their favourite celebrity or someone they know in real life. The feature allows its users to look through the names of innumerable celebrities to find one that they are attracted to while also allowing them to upload any picture of any person directly from their computer (creepy?). The potential of AI in dating apps The use of artificial intelligence and machine learning would not only make it significantly easier to detect fraud and filter fake accounts and inappropriate content—but would also be a means of employing creative methods to make dating apps more effective and less tiresome. Possibilities include using AI to figure out the “dating type” of an app’s users by discovering a pattern in the facial features of the people they usually swipe right on. It may, further, detect objects in images that its users have shown a routine preference towards—such as guitars or beaches. While the future of dating apps is unlikely to be any better at ascertaining what makes two people romantically or sexually compatible—it could make dating a lot easier (albeit, perhaps, a tad more superficial).","excerpt":"While the online dating industry is clearly booming, it remains riddled with complaints: from repeated ghosting, exhaustion with the sensory overload of potential matches, the proliferation of fake accounts, to reports of harassment—dating apps can be exhausting. While technology is the cause of this fatigue, machine learning and artificial intelligence may be the key to […]","categories":["IT Services"],"tags":[],"author_name":"Srishti Mukherjee","publish_date":"2021-12-23T13:00:00","publication_year":"2021","word_count":569,"keywords":["Go","machine learning","artificial intelligence","programming_languages:R","AI","programming_languages:Go","RAG","Aim","R","startup"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","Aim","RAG","R","Go","startup","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/machine-learning-is-solving-some-unique-problems-of-online-dating\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10057617,"title":"AI-based COVID-19 initiatives in 2021","content":"The way disease outbreaks are detected and handled has altered due to analytics, saving lives. With the recent signs of the third wave of COVID-19 and an upsurge in cases due to the Omicron variant, international organisations and scientists are utilising artificial intelligence (AI) to follow the outbreak in real-time, allowing them to predict where the virus will appear next and design an effective response. In this article, we reflect on the AI-based COVID-19 initiatives of 2021. Omnichannel AI Chatbot Yellow Messenger launched ‘Yellow Messenger Cares’ in 2021, a CSR programme to provide COVID-19 help-related omnichannel chatbots to enterprises. Yellow Messenger is one of the world’s fastest-growing conversational AI systems, providing an AI platform for customers and employees to create immersive conversational experiences. The campaign ‘Yellow Messenger Cares’ was launched to combat the country’s COVID-19 crisis. Any NGO, hospital, support organisation, or corporation, according to sources, can receive an AI chatbot designed by Yellow Messenger to drive crisis operations without any requirements or constraints. XRay Setu The immediate increase in COVID-19-positive individuals during the pandemic’s second wave resulted in delayed diagnosis and, as a result, many unnecessary deaths. In addition, the diagnosis was delayed due to India’s large population, scarcity of radiology labs, and skewed radiologist-patient ratio (one radiologist for one lakh people). To address this issue, the AI and Robotics Technology Park (ARTPARK) developed Xray Setu, an artificial intelligence-driven chest X-ray interpretation solution for doctors, in conjunction with AI-powered health tech firm NIRAMAI Health Analytix and the Indian Institute of Science (IISc). Xray Setu was created over the course of 10 months and is simple to use. According to a media report, ARTPARK CEO Umakant Soni stated that the company planned to develop a network of 10,000 doctors who are well-versed in Xray Setu. The company’s goal is to be ready for the third COVID-19 wave by developing solutions that would assist doctors in rural India. Contributions by IITs IIT Mandi has launched an AI-powered software that will track persons under home quarantine. In 2021, IIT Mandi created a biometric application driven by artificial intelligence to monitor and reliably determine the identity and whereabouts of the COVID-19 patients who have been confined at home. The AI-based mobile application Lakshman Rekha employs a combination of biometric verification, geofencing, and artificial intelligence to determine whether or not patients have breached their quarantine space. IIT Bombay has launched a video surveillance AI platform to monitor infractions of social distancing norms remotely. Surakshavyuh was created in 2017 to be used for military surveillance. The product subsequently evolved into enterprise-grade video analytics systems powered by machine learning that can identify unauthorised entrance and loitering, monitor perimeters and track objects, count crowds and recognise faces, and so on. The state-of-the-art video surveillance was created in partnership with SrivisifAI Technologies Pvt Ltd by the National Center of Excellence in Technology (NCETIS) at IITB. Amplify.ai MyGov, the government of India’s citizen engagement portal, leveraged Conversational AI startup Amplify.ai to respond in real-time to people’s questions and queries about COVID-19 helplines, volunteering, and donations. The company created an omnichannel virtual assistant powered by AI that operated across all key ‘conversational surfaces,’ including web chat, Google search and maps, and Facebook Messenger. For example, in around 700 towns across India, Google Business Messages (GBM) revealed 11,000 food and night shelter locations. In addition, more than 60 million messages were exchanged in the user messenger conversation. Read AIM’s exclusive interview with Amplify.Ai here. In conclusion, COVID-19 has led to an increase in AI-based innovations. Recently, ​​the University of Waterloo researchers and DarwinAI, an alumni-founded startup business, have created a method that employs artificial intelligence (AI) to anticipate the need for ICU admission. The algorithm analysed over 200 clinical data points, including blood test results, vital signs, and medical history. AI can play a vital role in exposing the soft white underbelly of the third wave of COVID-19.","excerpt":"AI is playing a vital role in exposing the soft white underbelly of COVID-19. This article looks into the COVID-based AI initiatives of 2021.","categories":["AI Features"],"tags":[],"author_name":"Abhishree Choudhary","publish_date":"2022-01-05T11:00:00","publication_year":"2022","word_count":647,"keywords":["Go","machine learning","artificial intelligence","AI","chatbots","RAG","Ray","Aim","analytics","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","analytics","Aim","Ray","RAG","chatbots","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/ai-based-covid-initiatives-in-2021\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10060945,"title":"AMD overtakes Intel in market cap for the first time, on the back of Xilinx acquisition","content":"On February 15, AMD surpassed the market capitalisation of Intel for the first time. The former’s market cap stood at approximately USD 197.75 billion as opposed to Intel’s USD 197.24 billion. AMD recently acquired Xilinx, a US-based semiconductor company, in a record USD 50 billion deal, giving the company an edge in the data centre market. “The acquisition of Xilinx brings together a highly complementary set of products, customers and markets combined with differentiated IP and world-class talent to create the industry’s high-performance and adaptive computing leader. Xilinx offers industry-leading FPGAs, adaptive SoCs, AI engines and software expertise that enable AMD to offer the strongest portfolio of high-performance and adaptive computing solutions in the industry and capture a larger share of the approximately USD 135 billion market opportunity we see across cloud, edge and intelligent devices,”said Dr. Lisa Su, President and CEO, AMD. AMD witnessed a record revenue growth in the fourth quarter of last year. The company reeled in USD 4.8 billion in revenue, a 48% increase year on year. Though AMD has an edge in the market share, Intel has a clear lead in terms of revenue. However, AMD registered a 25.6% CPU market share in the fourth quarter of 2021, the highest level of market share achieved by the company so far.","excerpt":"AMD overtakes Intel in market cap for the first time, on the back of Xilinx acquisition","categories":["AI News"],"tags":["AMD","Intel"],"author_name":"SharathKumar Nair","publish_date":"2022-02-17T17:51:11","publication_year":"2022","word_count":215,"keywords":["AMD","API","programming_languages:R","AI","RPA","R","Intel"],"extracted_tech_keywords":["AI","R","API","RPA","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/amd-overtakes-intel-in-market-cap-for-the-first-time-on-the-back-of-xilinx-acquisition\/","complexity_score":3,"technical_depth":5,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":41709,"title":"Deep Dive: SonicWall Doubles Down On AI &#038; Automation To Fight Cyber Attacks","content":"A recent SonicWall Cyber Threat Report indicated that ransomware attacks were up in every geography except India, which saw a 49% reduction, and the U.K. However, India still has a major uphill task in combating malware, where attacks rose by 53% in 2018. For our weekly column Deep Dive we talked to the Debasish Mukherjee, Country Director – India & SAARC, SonicWall to understand how technologies like AI and ML can better the cybersecurity landscape. Role Of AI And ML SonicWall is one of the leading companies in the space of cybersecurity that leverages technologies like artificial intelligence and machine learning. The company has even built a product that ensures protection against some of the worst malware. When it comes to malware detection and protection, there are many companies that are relying on signature-based malware monitoring, that can only detect and protect against known attacks. These kind antiviruses require frequent (daily or weekly) updates of their signature databases to protect against the latest threats, which isn’t feasible for the end-user. Furthermore, Mukherjee also said that AI can adapt the discovery approach to uncover threats that try to hide and, once determined as malicious, can fingerprint the payload via signature, turning a zero-day into a known threat. Talking about SonicWall’s firewall, with more than 1 million sensors deployed across 215 territories and countries, the company has one of the largest global footprints of active firewalls. The company also has a cloud-based, multi-engine SonicWall Capture Advanced Threat Protection (ATP) sandbox service that discovers and stops unknown, zero-day attacks, such as ransomware, at the gateway with automated remediation. Staying Ahead Of The Curve When asked, what makes SonicWall better than its competitors in the market, Mukherjee said, the single-sign-on and single-pane-of-glass management integrate the functionality of the Capture Cloud Platform to deliver robust security management, analytics, and real-time threat intelligence for an entire portfolio of network, email, mobile, and cloud security resources. That is not all, SonicWall NSV, which is a new virtual firewall service, provides deep packet inspection, security control, and networking functionality.  This new product is designed to block denial-of-service attacks, context-aware exploits, and other unwanted web traffic before it can connect with web applications. Also, Capture Client integrates SonicWall’s firewalls with SentinelOne’s machine learning-based continuous behaviour monitoring functionality that allows its users to block malicious traffic at the firewall in real-time.","excerpt":"A recent SonicWall Cyber Threat Report indicated that ransomware attacks were up in every geography except India, which saw a 49% reduction, and the U.K. However, India still has a major uphill task in combating malware, where attacks rose by 53% in 2018. For our weekly column Deep Dive we talked to the Debasish Mukherjee, […]","categories":["AI Features"],"tags":["Cyber Security","hacking"],"author_name":"Harshajit Sarmah","publish_date":"2019-07-03T11:13:21","publication_year":"2019","word_count":391,"keywords":["artificial intelligence","machine learning","Cyber Security","AI","hacking","programming_languages:R","ML","RAG","analytics","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","analytics","RAG","R","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/deep-dive-sonicwall-doubles-down-on-ai-automation-to-fight-cyber-attacks\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10473,"title":"Managing the Employee Efficiency &#038; Variability tradeoff","content":"Variability in Performance of employees has paramount importance in operations planning for any organization. Basis the analysis that is detailed in the paper we find that even the experienced of the employees have around 30%-40% variability in their efficiency. For better operations planning consistency in the process or employee efficiency becomes paramount. The article discusses the Efficiency and Variability tradeoff & how an organization can derive what is the required Optimum level of “Efficiency and Variability” at a strategic level and how to track & monitor the improvement at the tactical level. The example discussed in the paper has been derived from transaction services operation. Efficiency here means the time taken to complete a particular transaction basis the time standard. Variability is the movement that is experienced in the efficiency of the employees over a time period. Using two metrics “Efficiency “and “Variability in Efficiency” we derive the concept of Consistency curve, Optimal level on the curve Levels on the Consistency curve for New employees and experienced employees Consistency scoring metrics to enable the operations to track the performance in terms of how consistent the team is periodically. Concept is partly derived from Markowitz Efficient Frontier theory and the framework is designed along that to be used in the field of operations analytics Fig on left details the Efficiency on the Y axis and Variability on X axis for associates working on transaction processes. Complexity of work is assumed to be homogeneous here although the same can be incorporated to calculate efficiency using the Time standards for a particular work type in the below equation. Efficiency is defined as=(Time standard to resolve a customer request)\/(Time taken to resolve the request) Variability in Efficiency=Standard deviation(σ) of (Efficiency) for a time period t Figure details the Efficiency and variability for four employees A,B,C and D. B is highly efficient but less predictable when compared to A, while relatively A is highly predictable but at the cost of lower efficiency than B. In the fig on left, we can see that A & C have the same efficiency level calculated for the time period t also A is less variable than C and thus more consistent than C. For an Operations manager A would be the person he would go up to in comparison to C given an option. B &D work at the same level of variability but B is highly efficient than D which means that B is a better performer than D by looking at both the metrics together. The important concept here is that the operations manager should ensure that C improves relative to A by reducing the variability while D having the same level of variability as B should try to move up the ladder in terms of efficiency. Optimal path of improvement for employees in term of efficiency and variability should be North West direction, operational effectiveness is achieved with improvement in efficiency & decline in variability. Plot from a sample of operations function: Consistency Curve Figure above has been derived from a sample analysis of an operations team handling the transaction operations. Efficiency has been calculated for a time period of 6 months. Total sample size=40 employees Time period of analysis (t)=6 months Data points above reflect the 6 months efficiency and variability in efficiency. Variability in efficiency for an employee is calculated as= σ(Efficiency(week ’1’), Efficiency(week ’2’)..Efficiency (week ’n’)) In the above graph we can see that there are employees who are highly efficient but highly variable too. This creates inconsistency into the system which makes it difficult for efficient planning of operations as higher levels of variability reduces the probability of achieving the given level of efficiency on a given day or point in time Consistency Curve: Consistency curve is the relative zone that all the employees should try and aim for. The curve will always be concave and is formed by joining the best possible options available at a given level of efficiency and variability. Consistency Curve: Decision making Aim of the Operations leads should be to ensure that over a period of time all the employees should try and reach closer to the consistency curve. At what point on the consistency curve they should be would depend upon the current scenario of operations and requirements of the managers at that point in time. The detail of the same is given below in the graph In the above graph we divided the Consistency curve into two sections Lower Zone: It is the zone that the new employees are expected to target. Middle zone: This is the area that in general the experienced employees are expected to target. Upper Zone: Gain in efficiency is coming at the cost of higher variability. Minimum accepted efficiency (MAE): This is the minimum efficiency on the Y axis that the employees are expected to achieve in any given scenario. In our analysis in the previous graph we have kept that point to be 70% on the Y axis (highlighted in the red circle). Optimal Performance level: This is the point where the slope of the curve is maximum and the gain in efficiency is maximum for a unit increase in variability. Beyond this point the curve becomes flat and the incremental gain in efficiency comes at a cost of higher variability. In our analysis that point calculated is 115.10% at a variability level of 21% This is point where the line from MAE (Minimum accepted efficiency) to the curve is tangent. Slope = (Efficiency-MAE)\/ σ(Efficiency)   Implementation of Consistency curve concept Implementation of Consistency curve concept Next step is to monitor the performance of the employees in terms of how well they are improving in terms of moving towards the North West direction on the Efficiency Vs Variability graph. The idea is to measure what is the incremental gain in efficiency from MAE per unit of variability. Consistency= (Efficiency-MAE)\/ σ (Efficiency-MAE) Consistency score from the above equation can be calculated for each employee on a weekly or monthly basis. Any gain in the consistency score on a week to week basis is a sign of improvement and if the score falls it means that the performance has declined. Figure on left gives the hypothetical example of the consistency score of 3 employees X, Y, Z. X has been performing really well in both the metrics of efficiency and variability put together that measures the consistency of an employee. Y started from a negative zone and has been steadily improving Z has been stagnant and starts declining later which can be because of Decline in efficiency at the same variability or Increase in variability at the same efficiency level or Efficiency could be falling while variability could be increasing. Recommendation: Transaction services managers should ensure that they define the target zones for their associates on the consistency curve and monitor the same on an ongoing basis. The consistency curve should be updated on a regular basis.","excerpt":"Variability in Performance of employees has paramount importance in operations planning for any organization. Basis the analysis that is detailed in the paper we find that even the experienced of the employees have around 30%-40% variability in their efficiency. For better operations planning consistency in the process or employee efficiency becomes paramount. The article discusses […]","categories":[],"tags":[],"author_name":"Ashish Pandey","publish_date":"2016-07-27T05:28:54","publication_year":"2016","word_count":1156,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","Aim","analytics","GAN","R"],"extracted_tech_keywords":["AI","analytics","Aim","R","Go","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/managing-employee-efficiency-variability-tradeoff\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10112126,"title":"8 Must-Know OCR Tools for Training AI\/ML Models","content":"India boasts over 400 languages and a rich linguistic tapestry but faces the challenge of bridging the digital divide, which is exacerbated by the dominance of English in LLMs. Perpetually hungry for data, large language models are extensively trained on online information. However, the absence of non-English language data and the abundance of vast offline data can be leveraged with OCR. Optical Character Recognition (OCR), which is the process of transforming an image containing text into a machine-readable text format, digitises content into data that can be used for analytics, automation, training AI models and other processes. With the function to extract data, OCR enables LLMs to analyse and process the said data. Here are a few OCR tools that can aid developers and coders train AL\/ML models. Best OCR Software with Machine Learning in 2024SuryaBhashiniTesseract OCRPyTesseractEasyOCROpenCVOCRopusKraken Surya Surya, a multilingual text line detection model designed for document OCR, has been trained on diverse documents, including scientific papers. The training ensures that Surya excels in detecting text lines within documents, delivering pinpoint accuracy in line-level bounding boxes and clear identification of column breaks in PDFs and images. Bhashini Bhashini, an app developed to help people translate content in different Indian languages, recently introduced an OCR feature, called SCENE. The feature allows users to extract text by simply scanning an image using the camera. Bhashini was recently used by the Prime Minister Narendra Modi to address students during ‘Pariksha Pe Charcha’. Tesseract OCR Tesseract OCR is an open-source OCR engine maintained by Google. It was first developed by Hewlett-Packard, and later taken over by Google. Tesseract has unicode (UTF-8), supports more than 100 languages and can be integrated with LLMs to extract text from images. It also supports various image formats such as PNG, JPEG, TIFF. PyTesseract Python-Tesseract serves as an optical character recognition (OCR) utility for Python. Essentially, it is capable of identifying and interpreting the text contained within images. Python-tesseract acts as a wrapper for Google’s Tesseract-OCR Engine. It proves handy as a standalone execution script for Tesseract, capable of interpreting all image formats supported by the Pillow and Leptonica imaging libraries, such as jpeg, png, gif, bmp, tiff, among others. Furthermore, when employed as a script, Python-tesseract outputs the recognized text directly rather than storing it in a file. EasyOCR EasyOCR is a Python package that provides a straightforward interface for performing OCR tasks. It is an open-source OCR engine that supports multiple languages and can be used with LLMs for text recognition and data extraction. It also offers pre-trained models for various use cases. OpenCV OpenCV (Open Source Computer Vision) is a collection of programming functions primarily focused on real-time computer vision tasks. While it may require more customisation, it can be used in conjunction with LLMs for OCR tasks. In Python, OpenCV facilitates image processing by providing functions for tasks such as image resizing, pixel manipulation, object detection, and more. OCRopus OCRopus is another open-source OCR engine that is designed for high accuracy and efficiency. It includes various preprocessing and post-processing techniques suitable for AI and ML applications. OCRopus commands typically display a stack trace alongside an error message, but this does not necessarily indicate a problem. Kraken Kraken is an OCR engine implemented in Python and optimised for historical and degraded document recognition. It can be used in AI and ML models for tasks involving challenging document images. Kraken can be run on Linux or Mac OS X (both x64 and ARM). Resources Object Detection Algorithms Best Coding Practice Sites AI Tools For Stock Market Trading in India ChatGPT Alternatives Best AI Search Engines","excerpt":"OCR tools enable LLMs to process and understand textual content from various sources","categories":["AI Trends"],"tags":["Bhashini","ChatGPT","Hewlett Packard","Indian languages","OCR","Open Source","Optical Character Recognition","Python"],"author_name":"Vandana Nair","publish_date":"2024-02-06T18:00:00","publication_year":"2024","word_count":599,"keywords":["ChatGPT","Hewlett Packard","machine learning","OCR","Open Source","AI","TPU","Indian languages","ML","computer vision","OpenCV","Python","RAG","Optical Character Recognition","Bhashini","object detection","analytics"],"extracted_tech_keywords":["AI","machine learning","ML","computer vision","analytics","ChatGPT","OpenCV","RAG","object detection","TPU"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/8-must-know-ocr-tools-for-training-ai-ml-models\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10093173,"title":"Google-backed Startups Are Using OpenAI’s GPT, Should it be Worried?","content":"“And in the end, OpenAI doesn’t matter. They are making the same mistakes as we are in their posture relative to open source, and their ability to maintain an edge is necessarily in question,” reads Google’s internal leaked document titled, “We Have No Moat, and Neither Does OpenAI.” In the leaked document, it also said that it is definitely worth noting that ever since developers got their hands on Meta’s leaked LLM, LLaMA, the open-source community is flooded with LLMs-based generative AI models. Looks like Meta’s “mistake” has actually brought them back into the race, but only through open-source. Interestingly, Google claims that OpenAI has no moat when it comes to LLMs. But among the hype around ChatGPT, we are increasingly seeing startups that are either using GPT in their name, or building technologies using the APIs provided by OpenAI. During all this, Google backed a lot of AI startups to fight against OpenAI and Microsoft. It has been trying to push generative AI into its cloud infrastructure, and thus partnered with Salesforce, Box Inc, Jasper AI, and Canva. The companies have been using Google’s LLM offering through Vertex AI on Google Cloud. OpenAI Has A Moat Interestingly, despite the partnership with Google, last month Salesforce also announced Einstein GPT, integrating OpenAI’s technology and offering generative AI capabilities for its customers. In the same announcement, Salesforce also announced a $250 million generative AI fund which will first invest in four companies that include Cohere and Anthropic, both of which are backed by Google; and the companies will use GPT technology on Salesforce’s platform. Seems like the startups backed by Google are using its funds on OpenAI’s technology to run their businesses. Another Indian startup, Slang Labs, which is a voice assistant and search based, is backed by Google. Interestingly, the startup announced the launch of CONVA 2.0, a multilingual AI co-pilot which is powered by GPT technology, leveraging it for e-commerce shoppers. The company also boasts that their technology with GPT offers 46% better performance than Google’s current Voice Assistant. In an interview with AIM, Slang Labs’ CEO and co-founder, Kumar Rangarajan said that the company decided to use GPT into its products to offer services to its customers regardless of who developed it. “Making LLMs from ground is a very expensive and computation heavy process. It does not make sense to do it, therefore we adopted the GPT technology into our services,” Rangarajan explained. Similarly, Anthropic, an AI startup backed by Google also partnered with Notion, which is a conversation chatbot actually a wrapper of OpenAI’s GPT-3.5. The businesses just want the best services for their customers, and it seems as though OpenAI is providing them with that. Opensource: Google Killer? This should be concerning for Google that some of its startups are also leveraging its rivals’ technologies. Looks like GPT is the moat for OpenAI. And where OpenAI is lagging behind with a lot of its open-source policies, privacy issues, and many more concerns, Meta’s LLaMA is filling the gap with its ability to allow individuals to build local models. “The modern internet runs on open source for a reason,” reads the document. “And we should not expect to be able to catch up [to open source].” Google likened the recent open-source-based AI development to the Internet. It acknowledged that in the Internet technology revolution, the open-source community has a bigger role to play and cannot be replicated by any company.  “Individuals are not constrained by licenses to the same degree as corporations,” read the leaked document from Google. The only thing that the big-tech has is computation capabilities, which are also not required since the introduction of LLaMA based models that can build ChatGPT like models on a single computer. Probably, instead of fearing OpenAI and GPT, Google is more worried about open-source. “But the uncomfortable truth is, we aren’t positioned to win this arms race and neither is OpenAI. While we’ve been squabbling, a third faction has been quietly eating our lunch — open-source.” But is it true that OpenAI and Microsoft are on the same level as Google, without a moat? It might not be true, Google. Can Google Catch Up? A lot of the development has happened, probably, because most of the startups have already started building on top of GPT. There was no other offering other than GPT before. “Every developer I know is building on top of GPT,” said Robert Scoble in his tweet talking about what Google along with Apple should do to get ahead in the game. For example, Khan Academy, one of the first companies’ to adopt GPT-4 into its offerings built his entire system on GPT-4. Would there be any reason for him to make the change and adopt Google or Apple’s offering in the future, unless they are a hundred steps ahead of OpenAI and Microsoft’s? Surprisingly, Google I\/O 2023 was a hit. Sundar Pichai and team introduced several advancements in their AI systems, including generative AI and LLMs. It also announced the launch of the PaLM-2 language model along with its API. It also hinted at the multimodal Gemini project that the company has been working on with DeepMind. On the flip side, the people who have already been building on top of GPT products would never want to shift to what Google is going to offer in the near future, such as PaLM-2, unless it is worth it. Same is the case with Apple if it gets into the LLMs field. We will hopefully hear more about this at Apple’s conference in June. For now, Meta’s LLaMA has won a lot of developers on its side, but apart from that, the case is still that as soon as someone hears generative AI or chatbot, they hear GPT or OpenAI. Google needs to step up its game, and the Google I\/O conference was definitely one step in the right direction. Anyway, OpenAI’s GPT is built using Google developed Transformers.","excerpt":"Google has PaLM and LaMDA, why does some of its funded startups still prefer GPT in their products and services?","categories":["Deep Tech"],"tags":["ai chatbots","Anthropic","bing ai","ChatGPT","DeepMind","Generative AI","Google Bard","Google Search","OpenAI","OPenAI GPT","Sundar Pichai"],"author_name":"Mohit Pandey","publish_date":"2023-05-11T17:00:00","publication_year":"2023","word_count":993,"keywords":["ai chatbots","ChatGPT","Anthropic","Gemini Pro","OpenAI","AI","Google Bard","Google Search","Transformers","RAG","Aim","OPenAI GPT","bing ai","generative AI","Generative AI","R","Sundar Pichai","DeepMind"],"extracted_tech_keywords":["AI","generative AI","ChatGPT","OpenAI","Anthropic","Gemini Pro","Aim","Transformers","RAG","R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/google-backed-startups-are-using-openais-gpt-should-it-be-worried\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10098170,"title":"AWS vs Google Cloud vs Microsoft Azure: Hyperscalers’ Growth Galores","content":"With the growing excitement around generative AI in the cloud industry, it’s a pivotal moment to identify this quarter’s champions and pinpoint the leading company delivering top-notch generative AI services to its customers. Examining the latest quarterly results, Azure shines brightly in the generative AI arena, boasting impressive 26 percent growth. Following closely are Google and AWS, with 28 and 12 percent growth, respectively. Overall, it was a great quarter for hyperscalers. According to estimates from Synergy Research Group in Q2 2023 cloud services spending is on the rise, and Amazon’s worldwide market share has jumped by over 1% to reach almost 34%. In Q1 2023, Amazon’s AWS cloud market share was 32%, Azure’s was 23% and Google’s cloud was 10%. Amazon saw a slight downward trend in  its market share from 34 percent in Q3 2022, to 33 percent in Q4 2022. At the same time, Microsoft gained two points since Q3 2022 and now sits at 23 percent market share, gradually eating away at Amazon’s lead. Revenue Galores Google Cloud, one of the major cloud service providers, reported a 28% increase in revenue, reaching $8.1 billion, surpassing expectations of $7.75 billion. Microsoft  usually does not break out precise quarterly revenue for Azure, its most crucial tool to leverage generative AI endeavors. However, in the earnings call, Microsoft CEO Satya Nadella revealed that Microsoft Cloud surpassed $110 billion in annual revenue, up 27% in constant currency, with Azure all-up accounting for more than 50% of the total for the first time putting Azure sales at $55 billion or more and revealing the size of the business. AWS’s second-quarter cloud sales increased 12% to $22.1 billion which is less as compared to both Google Cloud and Azure. On the other hand, Oracle is also making huge leaps as in its fourth quarter results announced in June, it saw a jump of 23% in its revenue in USD and up 25% in constant currency to $9.4 billion. Recently, Oracle partnered with Cohere to bring generative AI applications to its customers. Analysts and industry experts believe that cloud business growth will speed up in the coming months, especially in the June quarter as uncertainties start to clear. Investors anticipate that AI will play a significant role in driving growth for cloud businesses in the next year, with Microsoft’s Azure leading the way, followed by Amazon.com’s AWS and Google Cloud. Fueled by Generative AI While AWS leads the cloud market share, it seems like Azure has taken the lead in integrating  generative AI applications as it has a special partnership with OpenAI. To add to this they also have  a partnership with Meta where they provide Llama 2 on Azure cloud which gives them an edge in the market. The core for Amazon is its AI foundation model service called Bedrock. The service, introduced in April, initially supported models from AI21, Anthropic, and Stability AI, along with Amazon Titan models. Now, the range of supported models has been broadened to include Cohere, Anthropic Claude 2, and Stability AI SDXL 1.0 models. Considering revenue growth, it’s evident that Microsoft’s Azure held the lead in the generative AI advantage within its cloud platform for this quarter. To challenge Azure, Google  introduced Generative AI on Vertex AI, offering customers access to a variety of model types and sizes through Vertex AI’s Model Garden. Customers can also utilize Google’s foundational models via APIs. The underlying model driving the PaLM API is PaLM 2. However, Vertex AI hasn’t gained much popularity among enterprises yet. Though Google Cloud’s market share is lowest as compared to AWS and Azure, CEO Sundar Pichai in the earnings call pointed out that more than 70% of gen AI unicorns are Google Cloud customers, including Cohere, Jasper, Typeface, and many more. He added that Google  provides the widest choice of AI supercomputer options with Google TPUs and advanced NVIDIA GPUs. Yet, when looking at AWS’s approach of employing multiple LLM models, it seems they are poised to challenge Azure in the upcoming quarters. On the other hand, Oracle has much ground to cover, as their partnership is limited to Cohere, which is already accessible via AWS Bedrock. In the world of cloud computing, the recent earnings report has shown a fierce competition among big players in generative AI. Azure seems ahead with its strong partnerships, while AWS’s diverse approach and Oracle’s progress keep the race exciting. As cloud business grows and  generative AI’s importance rises, the next quarters will be a dynamic battlefield for leadership in generative AI.","excerpt":"Oracle is also making huge leaps as in its fourth quarter results announced in June, it saw a jump of 23% in its revenue in USD and up 25% in constant currency to $9.4 billion.","categories":["Global Tech"],"tags":[],"author_name":"Siddharth Jindal","publish_date":"2023-08-07T10:01:37","publication_year":"2023","word_count":754,"keywords":["Anthropic","TPU","OpenAI","AI","cloud computing","AWS","R","RAG","generative AI","Azure"],"extracted_tech_keywords":["AI","generative AI","OpenAI","Anthropic","RAG","cloud computing","AWS","Azure","TPU","R"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/aws-vs-google-cloud-vs-microsoft-azure-hyperscalers-growth-galores\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10113585,"title":"Sarvam AI Releases Indic Dataset ‘Samvaad’","content":"Sarvam AI has released “samvaad,” a new open-source series of carefully curated India datasets. This release includes 100,000 high-quality, multi-turn conversations, totaling over 700,000 turns, in English, Hindi, and Hinglish. These datasets have been thoughtfully curated with an exclusive focus on an Indic context and are now accessible on Hugging Face. For developers and enthusiasts operating within the Indic space, this announcement promises valuable resources, with the potential for more exciting releases in the near future. Sarvam AI, in collaboration with Hugging Face, invites the community to stay tuned for forthcoming updates. To engage with the community and explore these datasets further, Sarvam AI encourages interested parties to join their Discord channel at https:\/\/lnkd.in\/eXCbKTF5. Sarvam AI recently partnered with Microsoft Azure to make its Indic voice large language model (LLM) available on Azure. Sarvam AI is building generative AI models targeting Indic languages and contexts. The startup aims to make the development and deployment of generative AI apps in India more accurate and cost effective. Sarvam AI’s Indic voice LLM aims to offer a natural voice-based interface to language models (LLMs) and will initially be available in Hindi. Sarvam AI is actively working to expand coverage to include more Indian languages while ensuring support for colloquial language use. Sarvam AI recently released OpenHathi-Hi-v0.1, the first Hindi LLM in the OpenHathi series. Developed on a budget-friendly platform, the model, an extension of Llama2-7B, boasts GPT-3.5-like performance for Indic languages. The Bangalore-based startup also raised USD 41 million in a Series A funding round led by Lightspeed and supported by Peak XV Partners and Khosla Ventures. Sarvam’s objective is not just to build open-source Indic LLMs but to develop a platform and help build AI-powered applications that can be deployed at a population scale.","excerpt":"It includes 100k high quality, multi turn conversations (>700k turns) in English, Hindi, and Hinglish curated exclusively with an Indic context.","categories":["AI News"],"tags":["sarvam ai"],"author_name":"Siddharth Jindal","publish_date":"2024-02-22T11:08:29","publication_year":"2024","word_count":292,"keywords":["Hugging Face","sarvam ai","AI","R","RAG","GPT","Aim","ViT","generative AI","Azure","startup"],"extracted_tech_keywords":["AI","generative AI","Aim","Hugging Face","RAG","Azure","R","GPT","ViT","startup"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/sarvam-ai-releases-indic-dataset-samvaad\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":19245,"title":"How Process Manufacturing Industry Can Utilize Big Data Analytics","content":"Any process manufacturing plant has sensor measurement data, operator log book data either in electronic form or in handwritten form, CCTV monitoring data, audio data either from CCTV or separate system, thermal imaging data for hot areas of the plant, weather information, logistics information etc. Table 1. Details of different data types stored in process manufacturing plants Data available in process manufacturing plants have following characteristics in the context of big data definition Variety      numeric, text, video and audio data types Velocity    scan rate of Mili-second to minutes (multi scale data) Volume    a refinery is designed for 2GB\/day data logging capacity Value        novel insights for operations, energy and asset management etc. Veracity    data quality and metadata management processes Can process manufacturing data be termed BIG DATA It is evident that the process manufacturing plant data has all the characteristics for it to be termed big data. Despite fulfilling the big data definition criteria, process manufacturing data cannot be used for big data analysis because different data types are not available for combined analysis. As described in Table 1. Sensor measurement data is available in different DCS systems, PLCs and data loggers in the plant; sensor data is not combined in real time hence rendering it unusable for combined analysis. CCTV data store is from different vendors and developed on different database technology than sensor measurement data and hence cannot be used for combined analysis. Similarly, operations and maintenance logbooks contain text data that is either stored in Manufacturing Execution System (MES) database or in handwritten records; thermal imaging and audio data systems are from different vendors and developed on different database technology than sensor measurement data. Moreover, there is no technology platform available for combined analysis of different types of data available in process manufacturing plant. How can process manufacturing data be transformed to BIG DATA Data integration and combined data analysis platform are the enablers for big data analysis of process manufacturing data. New technologies are available for common storage and management of different data types. Unstructured databases with distributed file system architecture are now available which can handle variety of data types in scalable way. Apache Hadoop is one such open source unstructured database which can handle numeric, text, video and image data types. Apache Hadoop is scalable because of its Distributed File System [DFS] architecture. In memory data stores such as GridGain and aggregate oriented databases such as mongoDB provide different choices for real time and unstructured data analysis. These new technologies enable combined data analysis at large scale in near real time scenarios. Parallel computing platforms such as Apache Hadoop’s MapReduce framework and many other open source and proprietary MapReduce implementations can carry out data mining computations for large amount of data in short period of time. Fig. 1. Schematic of big data technology deployment for process manufacturing industry Process manufacturing industry can benefit from new big data technologies to solve complex problems which have not been attempted so far because of data unavailability or high computational requirements. Relevance of Big Data to Process Manufacturing Industry More sources of data will now be available for analysis with the help of big data technologies. It will be possible to deploy a solution for combined analysis of sensor measurement data and other data types such as text, audio, thermal imaging etc. Fig. 2. Potential benefits of new data Fig. 2 describes that new data can impact process manufacturing plant operations in two ways – new questions that can be answered with the availability of new data & technologies and existing questions that can be better answered by including new data sources and analytics techniques. Five new questions and three current questions are listed as use cases in this section to intuit potential benefits of big data analytics technologies for process manufacturing plants. What new questions can be asked? What are the best parameter settings for operating process plant? When is the next process disturbance or asset abnormal operation could occur? What is the root cause of process disturbance or asset abnormal operation? What are the best corrective actions to restore the operation? What would be the quality of product being produced? How existing questions can be answered better How to smoothly operate with least possible losses and inventory Real time integration of shop floor KPI to organization KPIs Traceability across the value chain Role of Big Data Stakeholders There are three stakeholders in realizing benefits of big data for process manufacturing plants. Refer Fig. 3. Computing technology vendors Automation technology vendors Process manufacturing Industry Fig. 3. Stakeholders for Big data technology deployment in process manufacturing plants Contribution from each of the three stakeholders is critical for realizing benefits of big data technologies for process manufacturing plants. Rest of the section outlines potential contribution of the three stakeholders towards big data technology adoption in process manufacturing plants. Computing technology vendor Computing technology vendors have been traditionally supplying IT hardware and basic computing environment. In the Big Data deployment scenario, computing vendor could bring following components – Unstructured and Distributed File System (DFS)database (e.g. Apache Hadoop) Parallel computing environment (e.g. Apache Hadoop’s MapReduce) Analytics modeling platform and computing libraries (e.g. Apache Mahout, IBM BIG data platform) Automation technology vendor Automation technology vendors have been traditionally supplying control and plant information management systems. In the big data deployment scenario, automation technology vendor could bring following components – Domain, control system and industrial communications knowledge Application of new computing technologies (DFS, MapReduce) to process manufacturing plant problems Operational roll out, refinement and field demonstration of new solutions Process manufacturing industry Process manufacturing industry have been traditionally adopting technology solutions and partnering with automation companies for new technology development e.g. advanced control and optimization solutions. In the big data deployment scenario, process manufacturing industry could actively engage with all the stakeholders in developing ecosystem for manufacturing analytics solutions. Process manufacturing industry could provide – Analytics ecosystem development and leadership Actual process manufacturing plant data which is key element of the puzzle Business critical problem statements i.e. information about what to look for in the data Operational and maintenance nuances which are of critical important for realizing successful operational roll out Conclusion Process manufacturing plants have variety of data sources such as sensors, CCTV systems, thermal videography, operation and maintenance logbooks etc. These data sources generate different types of data such as numeric, video, text etc. Different data types are used for specific analysis e.g. sensor data for process control, CCTV for facility monitoring, thermal videography for operational and safety etc. Current automation technologies do not allow for combined analysis of different data types. Big data technologies could enable complex data analysis in near real time to unlock operational efficiencies for process manufacturing plants.","excerpt":"Any process manufacturing plant has sensor measurement data, operator log book data either in electronic form or in handwritten form, CCTV monitoring data, audio data either from CCTV or separate system, thermal imaging data for hot areas of the plant, weather information, logistics information etc.    Table 1. Details of different data types stored in […]","categories":["IT Services"],"tags":["Big Data Analytics"],"author_name":"Amit Purohit","publish_date":"2017-11-23T08:19:36","publication_year":"2017","word_count":1120,"keywords":["big data","Go","AI","MongoDB","Scala","RAG","Big Data Analytics","analytics","data quality","GAN","R"],"extracted_tech_keywords":["AI","analytics","RAG","MongoDB","R","Go","Scala","big data","data quality","GAN"],"url":"https:\/\/analyticsindiamag.com\/it-services\/process-manufacturing-industry-can-utilize-big-data-analytics\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":5775,"title":"Interview &#8211; Narasimhalu Senthil, Principal Partner at Rinalytics Advisors","content":"[dropcap size=”2″]AIM[\/dropcap]Analytics India Magazine: What are the most important trends that you see in analytics hiring today? How are these different from analytics hiring last year? [dropcap size=”2″]NS[\/dropcap]Narasimhalu Senthil: Analytics hiring is high paced and an ever changing game. A plethora of players in the market ensure qualified candidates get absorbed quickly. Left to me, I’d mark they are spoilt for choices today. From captives to services to startups, there are different opportunities within their existing ecospace that one can look at engaging in an analyst role to a delivery role to the business role. There is an upsurge on strong attraction towards Startups\/e-Commerce\/Product analytics organizations vs. traditional captive centers, along with a crucial metamorphosis in compensation range. I think this year will be a trend setter of sorts. AIM: Which areas in analytics you think have the most potential going forward? NS: To start with: Future is analytics, though it’s a kind of oxymoron. Risk & Marketing analytic areas have constantly been an area of promise; while, new areas such as Customer Analytics\/Web & Digital Analytics\/Transactional Fraud Analytics\/Telecom & Pharma Analytics\/Big Data Analytics\/Data Scientist\/M2M Analytics (Internet of things) assures to indicate a major role going forward. Personally, I feel India is yet to apprehend the aura of Machine Learning\/ Neural Networks. That’s one area that will change the depth and gravity of analytics usage in our daily life.  A Mckinsey Global Institute’s study emphasizes Machine Learning as the driver of the next biggest wave of innovation. Lastly, although the presence of M2M was touched sometime back, an upsurge in equitable wireless communication, along with changing technologies has made way rise inexorably in recent times. Flutura, Invati –Insights, Bidgely are some of the start-ups firms in India to be watched for, that are centering towards M2M analytics for their businesses in various industries. AIM: What typically an analytics firm looking for while recruiting analytics professionals? NS: This is an interesting question. While the answer could be easily passed as a mother hood statement, I will try to nail it to specifics as much as possible. We have evolved from a skill\/tool based body shopping country to a full stack solution provider. Naturally, the demand for people beyond simple tools is on the rise. Tools in BI might earn you a place in Analytics wave, but won’t keep you there for long. Today it’s a must to show strong abilities in MATH\/STAT + Tech Tools + Business Acumen and I’m afraid none of the three is negotiable. The requisite for good business acumen, understanding of the client complications and breaching it down to solutions by means of models\/Customized solutions etc. has gained the preeminence; along with initial years of meddling with statistical\/ mathematical data, including variety of statistical model building and a strong pedigree from Tier 1 schools with Quantitative discipline studies. AIM: What are a few things that organizations should be doing with their analytics hiring that most don’t do today? NS: Regardless to the various technologies which Analytics organizations are applying; it is obligatory for the very to endow on a dedicated analytics professional and a coherent strategy.Visibility of business challenges should be set to the potential talent instead of directing merely on the KRA’s and evaluate for the business creativity besides technical skills and number crunching. Online communities would aid largely in sighting the factual pool of talent. Let’s consider Kaggle for illustration, which is an online platform that clouds data analytics competitions, permits companies to hit into the expertise of data scientists and elucidate organizations complications. As per a recent analysis conducted on Kaggle top 100 performers, over 80 percent of those 100 performers had a Master’s degree or higher and 35 percent have a Ph.D. AIM: What are the most significant challenges you face being in the forefront of analytics space? NS: Simply put, it’s an age old war of supply and demand. Like any other growth spurt in our country, we are caught unaware and unprepared. With opportunities aplenty and no dearth for business, everything is abundant but for qualified professionals. While skills can be trained, Analytics is a different beast. Creativity comes into play when Analytics teams consider pounding together data, in much the same way that composers might experiment with combination of instruments. Now, that is something that that cannot be trained. However, this nosedives in the Indian Analytics industry where in 80% of workforce is working on monotonous BI\/Models, while only 20 % of work force doescore modeling work. Proliferating cross training\/self-training would keeporganizations or employee themselves in tune with the growing business need in the future. AIM: What do you suggest to professionals aspiring to get into analytics space? NS: My two cents and often prescribed suggestions to aspirants: ü  If you have math and stat as your mother tongue, it makes life lot easier. ü  If you are basically a toolster aspiring to be in core of analytics, at least get the grip on BI tools along with some specific data visualization tools like Tableau, etc. ü  If you’re not from either of the background, start with market research analytics, get your hands dirty with research reporting, analysis and dash boarding then a slow transition to data analytics and heading towards advanced analytics. AIM: What do you suggest to analytics professional who are looking out for a job change within analytics field? NS: Centering self in gaining knowledge in core competencies (e.g. Modeling, Solutioning, and Delivery) and cross training inapplication of analytics on different domains; along with sprouting business acumen is significant. There is also a need to understand the consultative approach, readily take up challenges and primed to be tested, laterally with the dimensioning creativity and embracingonline learning. And as there are a huge number of start-up organizations looming, one should be competent and enthusiastic to mind up such novel ecosystems. AIM: Anything else you wish to add? NS: Analytics is a buzzword in the industry and every organization is trying to jump to the sea, with a careful approach. India has a large puddle of statistics, mathematics and various quant discipline talent pool, but they need industry specific training. Advanced analytical concepts have to be included in the curriculum. A strong demand for research professionals is in upswing by the dint of captive and research labs are approaching to India. [spoiler title=”Biography of Narasimhalu Senthil” style=”fancy” icon=”plus-circle”] Rinalytics is founded and led by Narasimhalu Senthil, Principal Partner of the Rinalytics team. Narasimhalu is a seasoned executive search professional with close to a decade of search experience. He helps his clients (small to large) build high performing analytics and risk management leadership talent. Prior to realizing his dream of starting Rinalytics, Narasimhalu worked for TRANSEARCH International, PMAM Corporation and ABC Consultants in India and UK. Having spent more than 4 years in the UK, he strongly believes in the concept of super specialization. He holds B.Tech in Information Technology from University of Madras and M.B.A in Human Resource Management.[\/spoiler]","excerpt":"[dropcap size=”2″]AIM[\/dropcap]Analytics India Magazine: What are the most important trends that you see in analytics hiring today? How are these different from analytics hiring last year? [dropcap size=”2″]NS[\/dropcap]Narasimhalu Senthil: Analytics hiring is high paced and an ever changing game. A plethora of players in the market ensure qualified candidates get absorbed quickly. Left to me, […]","categories":["AI Features"],"tags":["Interviews and Discussions"],"author_name":"AIM Media House","publish_date":"2014-05-27T15:59:20","publication_year":"2014","word_count":1160,"keywords":["big data","Go","machine learning","AI","neural network","Git","Aim","analytics","GAN","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","machine learning","neural network","analytics","Aim","R","Go","Git","big data","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/interview-narasimhalu-senthil-principal-partner-at-rinalytics-advisors\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":35102,"title":"Hackathon Is Not Just A Recruiting Tool, It’s A Fertile Space For New Ideas To Incubate: Pankaj Muthe, Qlik","content":"The latest interaction in our hackathon series is with Pankaj Muthe, the Academic Program Manager, APAC at Qlik. With over 19 years of experience in education, consulting and research, he has proven abilities in infusing creativity, innovative approach and fresh ideas to drive outstanding results. He has been instrumental in the launch of the first Centre of Excellence in Analytics of Qlik with a recorded growth of more than 250% year-on-year for the program. He has expanded the reach of academic program from less than 10 institutions to more than 150 in three years. Muthe shares his views on why hackathons in data science are turning to be of utmost importance, how it is solving the challenges of traditional hiring, and more. AIM: How have hackathons become a crucial part of the hiring process in the data science industry? Do you think hiring through hackathons will make hiring data science talent easier? Pankaj Muthe: The basic idea behind hackathons is to identify top tech talent by posing a real-world problem and then having participating teams come up with innovative solutions that can address the same within a predefined time limit. This allows companies to gauge participants on their technical capabilities, as well as their ability to translate theoretical knowledge into practical solutions in high-pressure situations. Moreover, hackathons require groups of four-five people to work closely, effectively, and efficiently with each other as part of a team. This provides an insight into their soft skills, such as teamwork and collaboration, and also helps organisers identify individuals with strong leadership qualities. An all-around overview of a participant’s technical and soft skills enables companies to make more informed hiring decisions. As a result, hackathons have become a part and parcel of the larger recruitment strategy at IT companies, particularly those within the data analytics segment. AIM: What is your opinion about online hackathon? Have hackathons replaced the traditional way of hiring, say by campus hiring? PM: Online hackathons are quite useful, as they allow greater flexibility to both organisers and students in terms of coordination. Prospective recruiters are also able to save on the costs of organising a physical event and can easily achieve scale to meet any sudden surge in the number of participants. But while hackathon-based recruitment is becoming increasingly popular, it has not replaced traditional campus hiring. At present, it still plays a supporting role in the larger organisational hiring strategy. However, as the demand for talent grows and organisations look for more tech-led avenues to conduct their hiring, we might witness more recruitments taking place through hackathons and other similar events. AIM: Hackathons solve two purposes — people strategies and the other is solving use-cases in an innovative way? Do you agree? PM: Yes, definitely. As I’ve already mentioned above, hackathons allow recruiters to gauge both the technical proficiency of participants, as well as several soft skills such as teamwork\/management, leadership, communication, the ability to handle stress, etc. AIM: Can you cite a recent example where you hosted a hackathon. Could you state the purpose? PM: While we have not hosted a hackathon in India yet, it remains an area of possible interest for us. We might host hackathons to identify and recruit emerging talent within the Indian data analytics domain in the near future. That being said, we have hosted a hackathon in another of our offices previously. This particular Qlik Hack Challenge, allowed us to work with the United Nations to analyse data from the Democratic Republic of Congo, a region that has been plagued by conflict for decades, resulting in unthinkable hardships for those in these areas. The challenge centres around analyzing this conflict and enabling the United Nations to resolve and soften these conflicts wherever possible. AIM: How important are hackathons (both internal and external) to boost innovation? Have you in your organisation benefitted by conducting hackathons? PM: Apart from identifying and recruiting talented individuals, the very format of hackathons also provides a fertile space for new ideas to incubate. High-pressure situations often make individuals operate at the very limits of their physical and intellectual capacities. Organisations can harness this creative energy to find, refine, and implement innovative solutions to complex, real-world business and social challenges. AIM: Besides building people capital, do hacks for hire also help in positioning the company as one that is innovative? PM: Yes, it does. Organising a hackathon helps companies position themselves as modern, future-ready organisations which promote and nurture innovation. Given the buzz and engagement that such events create, hackathons are definitely a great branding exercise – especially in a jobs market that is increasingly being populated by younger professionals. AIM: Do you agree that hackathons have become the benchmark for assessing baseline skills? If so, what skills? How reliable it is to test skills in the long run? PM: While it would be a bit presumptive to call hackathons as the benchmark for assessing baseline skills, they definitely do provide a more holistic overview of participants’ data literacy levels, technical skills as well as soft skills. The format of such events also makes it likelier to gauge each individual on aspects such as critical thinking, agility, and stress-handling capability, thus providing recruiters with a better idea of how well a prospective employee would fit into the organisation’s work culture. AIM: Are hackathons replacing the need to have short term certification course in areas such as AI, analytics? PM: Certification courses and hackathons are two very different things, in and by themselves. The former is aimed at acquiring newer skills, while the latter is all about applying existing skills and knowledge to solve business use-cases. AIM: In future, how will hackathons evolve (smaller, online hackathons or the preference will be for offline events) PM: Given their impact, hackathons will eventually come to play a much bigger role in the recruitments space. A significant component of this growth can be expected to take place over the online medium, as it enables organisers to reap the benefits of seamless accessibility and scalability.","excerpt":"The latest interaction in our hackathon series is with Pankaj Muthe, the Academic Program Manager, APAC at Qlik. With over 19 years of experience in education, consulting and research, he has proven abilities in infusing creativity, innovative approach and fresh ideas to drive outstanding results. He has been instrumental in the launch of the first […]","categories":["AI Features"],"tags":["data science hackathon","data science hiring india","hackathon india","Interviews and Discussions"],"author_name":"Srishti Deoras","publish_date":"2019-02-18T11:18:10","publication_year":"2019","word_count":1001,"keywords":["hackathon india","data science","Go","API","AI","ML","Scala","GAN","data science hackathon","Aim","analytics","data science hiring india","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","ML","data science","analytics","Aim","R","Go","Scala","API","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/hackathon-is-not-just-a-recruiting-tool-its-a-fertile-space-for-new-ideas-to-incubate-pankaj-muthe-qlik\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":68920,"title":"Hands-On Guide to Implement ResNet50 in PyTorch with TPU","content":"PyTorch is consistently adding a boost to the field of Computer Vision and Deep Learning by providing a number of powerful tools and techniques. In the field of computer vision where the deep learning based executions are to be dealt with heavy image datasets, an accelerated environment is needed to fasten the execution process with an acceptable accuracy level. PyTorch provides this feature through the XLA (Accelerated Linear Algebra), a compiler for linear algebra that can target multiple types of hardware, including GPU, and TPU. The PyTorch\/XLA environment is integrated with the Google Cloud TPU and an accelerated speed of execution is achieved. In this article, we will demonstrate the implementation of ResNet50, a Deep Convolutional Neural Network, in PyTorch with TPU. The model will be trained and tested in the PyTorch\/XLA environment in the task of classifying the CIFAR10 dataset. We will also check the time consumed in training this model in 50 epochs. Implementing ResNet50 in Pytorch To avail the facility of TPU, this implementation was done in the Google Colab. To start with, first, we need to select the TPU from Hardware accelerators under the notebook settings. After selecting the TPU, we will verify the environment using the below line of codes:- import os assert os.environ['COLAB_TPU_ADDR'] It will be executed successfully if the TPU is enabled otherwise it will throw the ‘KeyError: ‘COLAB_TPU_ADDR’’. You can also check the TPU by printing its address. TPU_Path = 'grpc:\/\/'+os.environ['COLAB_TPU_ADDR'] print('TPU Address:', TPU_Path) In the next step, we will install the XLA environment to accelerate the execution process. The same we have done in one of our last articles where we have implemented the convolutional neural network. VERSION = \"20200516\" !curl https:\/\/raw.githubusercontent.com\/pytorch\/xla\/master\/contrib\/scripts\/env-setup.py -o pytorch-xla-env-setup.py !python pytorch-xla-env-setup.py --version $VERSION Now, we will import all the required libraries here. from matplotlib import pyplot as plt import numpy as np import os import time import torch import torch.nn as nn import torch.nn.functional as F import torch.optim as optim import torch_xla import torch_xla.core.xla_model as xm import torch_xla.debug.metrics as met import torch_xla.distributed.parallel_loader as pl import torch_xla.distributed.xla_multiprocessing as xmp import torch_xla.utils.utils as xu import torchvision from torchvision import datasets, transforms import time from google.colab.patches import cv2_imshow import cv2 After importing the libraries, we will define and initialize the required parameters. # Define Parameters FLAGS = {} FLAGS['data_dir'] = \"\/tmp\/cifar\" FLAGS['batch_size'] = 128 FLAGS['num_workers'] = 4 FLAGS['learning_rate'] = 0.02 FLAGS['momentum'] = 0.9 FLAGS['num_epochs'] = 50 FLAGS['num_cores'] = 8 FLAGS['log_steps'] = 20 FLAGS['metrics_debug'] = False In the next step, we will define the ResNet50 model. class BasicBlock(nn.Module): expansion = 1 def __init__(self, in_planes, planes, stride=1): super(BasicBlock, self).__init__() self.conv1 = nn.Conv2d( in_planes, planes, kernel_size=3, stride=stride, padding=1, bias=False) self.bn1 = nn.BatchNorm2d(planes) self.conv2 = nn.Conv2d( planes, planes, kernel_size=3, stride=1, padding=1, bias=False) self.bn2 = nn.BatchNorm2d(planes) self.shortcut = nn.Sequential() if stride != 1 or in_planes != self.expansion * planes: self.shortcut = nn.Sequential( nn.Conv2d( in_planes, self.expansion * planes, kernel_size=1, stride=stride, bias=False), nn.BatchNorm2d(self.expansion * planes)) def forward(self, x): out = F.relu(self.bn1(self.conv1(x))) out = self.bn2(self.conv2(out)) out += self.shortcut(x) out = F.relu(out) return out class ResNet(nn.Module): def __init__(self, block, num_blocks, num_classes=10): super(ResNet, self).__init__() self.in_planes = 64 self.conv1 = nn.Conv2d( 3, 64, kernel_size=3, stride=1, padding=1, bias=False) self.bn1 = nn.BatchNorm2d(64) self.layer1 = self._make_layer(block, 64, num_blocks[0], stride=1) self.layer2 = self._make_layer(block, 128, num_blocks[1], stride=2) self.layer3 = self._make_layer(block, 256, num_blocks[2], stride=2) self.layer4 = self._make_layer(block, 512, num_blocks[3], stride=2) self.linear = nn.Linear(512 * block.expansion, num_classes) def _make_layer(self, block, planes, num_blocks, stride): strides = [stride] + [1] * (num_blocks - 1) layers = [] for stride in strides: layers.append(block(self.in_planes, planes, stride)) self.in_planes = planes * block.expansion return nn.Sequential(*layers) def forward(self, x): out = F.relu(self.bn1(self.conv1(x))) out = self.layer1(out) out = self.layer2(out) out = self.layer3(out) out = self.layer4(out) out = F.avg_pool2d(out, 4) out = torch.flatten(out, 1) out = self.linear(out) return F.log_softmax(out, dim=1) def ResNet50(): return ResNet(BasicBlock, [3, 4, 6, 4, 3]) The below code snippet will define the functions to load the CIFAR10 dataset, preparing training and test dataset, the training process and the test process. SERIAL_EXEC = xmp.MpSerialExecutor() # Only instantiate model weights once in memory. WRAPPED_MODEL = xmp.MpModelWrapper(ResNet50()) def train_resnet50(): torch.manual_seed(1) def get_dataset(): norm = transforms.Normalize( mean=(0.4914, 0.4822, 0.4465), std=(0.2023, 0.1994, 0.2010)) transform_train = transforms.Compose([ transforms.RandomCrop(32, padding=4), transforms.RandomHorizontalFlip(), transforms.ToTensor(), norm, ]) transform_test = transforms.Compose([ transforms.ToTensor(), norm, ]) train_dataset = datasets.CIFAR10( root=FLAGS['data_dir'], train=True, download=True, transform=transform_train) test_dataset = datasets.CIFAR10( root=FLAGS['data_dir'], train=False, download=True, transform=transform_test) return train_dataset, test_dataset # Using the serial executor avoids multiple processes # to download the same data. train_dataset, test_dataset = SERIAL_EXEC.run(get_dataset) train_sampler = torch.utils.data.distributed.DistributedSampler( train_dataset, num_replicas=xm.xrt_world_size(), rank=xm.get_ordinal(), shuffle=True) train_loader = torch.utils.data.DataLoader( train_dataset, batch_size=FLAGS['batch_size'], sampler=train_sampler, num_workers=FLAGS['num_workers'], drop_last=True) test_loader = torch.utils.data.DataLoader( test_dataset, batch_size=FLAGS['batch_size'], shuffle=False, num_workers=FLAGS['num_workers'], drop_last=True) # Scale learning rate to num cores learning_rate = FLAGS['learning_rate'] * xm.xrt_world_size() # Get loss function, optimizer, and model device = xm.xla_device() model = WRAPPED_MODEL.to(device) optimizer = optim.SGD(model.parameters(), lr=learning_rate, momentum=FLAGS['momentum'], weight_decay=5e-4) loss_fn = nn.NLLLoss() def train_loop_fn(loader): tracker = xm.RateTracker() model.train() for x, (data, target) in enumerate(loader): optimizer.zero_grad() output = model(data) loss = loss_fn(output, target) loss.backward() xm.optimizer_step(optimizer) tracker.add(FLAGS['batch_size']) if x % FLAGS['log_steps'] == 0: print('[xla:{}]({}) Loss={:.2f} Time={}'.format(xm.get_ordinal(), x, loss.item(), time.asctime()), flush=True) def test_loop_fn(loader): total_samples = 0 correct = 0 model.eval() data, pred, target = None, None, None for data, target in loader: output = model(data) pred = output.max(1, keepdim=True)[1] correct += pred.eq(target.view_as(pred)).sum().item() total_samples += data.size()[0] accuracy = 100.0 * correct \/ total_samples print('[xla:{}] Accuracy={:.2f}%'.format( xm.get_ordinal(), accuracy), flush=True) return accuracy, data, pred, target # Train and eval loops accuracy = 0.0 data, pred, target = None, None, None for epoch in range(1, FLAGS['num_epochs'] + 1): para_loader = pl.ParallelLoader(train_loader, [device]) train_loop_fn(para_loader.per_device_loader(device)) xm.master_print(\"Finished training epoch {}\".format(epoch)) para_loader = pl.ParallelLoader(test_loader, [device]) accuracy, data, pred, target  = test_loop_fn(para_loader.per_device_loader(device)) if FLAGS['metrics_debug']: xm.master_print(met.metrics_report(), flush=True) return accuracy, data, pred, target Now, we will begin the training of ResNet50. The training will be done in the 50 epochs as we have defined in the parameters. Before starting the training, we will record the training time and after training, we will print the total time taken. start_time = time.time() # Start training processes def training(rank, flags): global FLAGS FLAGS = flags torch.set_default_tensor_type('torch.FloatTensor') accuracy, data, pred, target = train_resnet50() if rank == 0: # Retrieve tensors that are on TPU core 0 and plot. plot_results(data.cpu(), pred.cpu(), target.cpu()) xmp.spawn(training, args=(FLAGS,), nprocs=FLAGS['num_cores'], start_method='fork') After the training, we will print the time taken during the training process. end_time = time.time() print(\"Time taken = \", end_time-start_time) Finally, we visualize the predictions made by the model on the sample test data during training. img = cv2.imread(RESULT_IMG_PATH, cv2.IMREAD_UNCHANGED) cv2_imshow(img) So, we can conclude by analyzing the training performance and predictions, the model has acquired more than 82% of accuracy during the training where the 50 epochs of training were performed in just around 50 minutes. In the end, we can also conclude that we have learnt how to implement the ResNet50 model in PyTorch with TPU on the CIFAR10 dataset. It opens a scope for testing other deep convolutional neural network models on the same or different benchmark datasets.","excerpt":"In this article, we will demonstrate the implementation of ResNet50, a Deep Convolutional Neural Network, in PyTorch with TPU. The model will be trained and tested in the PyTorch\/XLA environment in the task of classifying the CIFAR10 dataset. We will also check the time consumed in training this model in 50 epochs.","categories":["Deep Tech"],"tags":["Computer Vision","Deep Learning","Pytorch","ResNet50","TPU","Transfer Learning"],"author_name":"Dr. Vaibhav Kumar","publish_date":"2020-07-06T11:00:00","publication_year":"2020","word_count":1130,"keywords":["ResNet50","Pytorch","NumPy","TPU","AI","neural network","PyTorch","computer vision","Colab","Python","deep learning","Computer Vision","Deep Learning","Transfer Learning","Matplotlib"],"extracted_tech_keywords":["AI","deep learning","neural network","computer vision","PyTorch","Colab","NumPy","Matplotlib","TPU","Python"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/hands-on-guide-to-implement-resnet50-in-pytorch-with-tpu\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10133226,"title":"HP Enterprise Announces Plan to Acquire Morpheus Data","content":"Hewlett Packard Enterprise has announced that it has entered into a definitive agreement to acquire Morpheus Data, a pioneer in software for hybrid cloud management and platform operations. Morpheus will enhance HPE GreenLake by providing multi-vendor, multicloud application provisioning, orchestration and automation, as well as FinOps capabilities for cloud cost optimization. Morpheus complements HPE’s successful acquisition of IT operations management leader OpsRamp in 2023. These capabilities will solidify HPE as the first vendor to provide a full suite of enterprise-grade capabilities and services across the hybrid cloud stack2 and will make HPE GreenLake the future-proof hybrid cloud destination for enterprises. “With the acquisition of Morpheus Data, we will take the next major leap to make HPE GreenLake cloud the de facto platform for innovating across hybrid IT,” Fidelma Russo, executive vice president and general manager, hybrid cloud and CTO at HPE said. Earlier this year, HPE and NVIDIA announced NVIDIA AI Computing by HPE, a portfolio of co-developed AI solutions and joint go-to-market integrations that enable enterprises to accelerate adoption of generative AI. Among the portfolio’s key offerings is HPE Private Cloud AI, a first-of-its-kind solution that provides the deepest integration to date of NVIDIA AI computing, networking and software with HPE’s AI storage, compute and the HPE GreenLake cloud. The offering enables enterprises of every size to gain an energy-efficient, fast, and flexible path for sustainably developing and deploying generative AI applications.","excerpt":"Hewlett Packard Enterprise has announced that it has entered into a definitive agreement to acquire Morpheus Data, a pioneer in software for hybrid cloud management and platform operations. Morpheus will enhance HPE GreenLake by providing multi-vendor, multicloud application provisioning, orchestration and automation, as well as FinOps capabilities for cloud cost optimization. Morpheus complements HPE’s successful acquisition […]","categories":["AI News"],"tags":["ai announcements","Mergers and Acquisitions"],"author_name":"Pritam Bordoloi","publish_date":"2024-08-20T15:09:09","publication_year":"2024","word_count":233,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","RAG","automation","ai announcements","generative AI","Mergers and Acquisitions","R"],"extracted_tech_keywords":["AI","generative AI","RAG","R","Go","automation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/hp-enterprise-announces-plans-to-acquire-morpheus-data\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10167225,"title":"India Cozying Up to Big Tech Under Trump’s Tariff Heat","content":"US President Donald Trump announced a 26% reciprocal tariff on India on Tuesday, effective April 9th. India has been preparing to counter this move since Trump first indicated such an action earlier this year. The country has a tariff rate that is comparatively lower than those of Taiwan (32%), Vietnam (46%), Thailand (36%), and others. To minimise the impact, the Indian government has also liberalised rules favouring big tech in the US over the past month. While some argue that India is not facing any international trade pressure from the US government and may not have a far-reaching impact, a few experts are of the opinion that, in fact, the tariff move will benefit India with its ‘Make in India’ push. For instance, recently, the Indian government announced plans to remove the 6% tax often referred to as ‘Google tax’, on online digital advertisements, which will greatly benefit major companies like Google and Meta. Additionally, both companies noted favorable outcomes in their legal battles in India. Recent reports indicate that Tesla is entering India, which could be a game changer for the country’s autonomous vehicle (AV) market. The company plans to hire professionals across various fields and has leased spaces for new showrooms in Mumbai and New Delhi. Furthermore, Indian telecom giants Jio and Bharti Airtel announced partnerships with Musk’s Starlink, a technology that utilises a constellation of low Earth orbit (LEO) satellites to deliver high-speed internet services. India Under Pressure From the United States? In a conversation with AIM, Mohandas Pai, the former CFO and board member of Infosys, mentioned that the aforementioned move by the Indian government was to counter Trump’s tariff plans. For context, the equalisation levy was introduced by the Indian government in 2016 to tax payments made by Indian businesses to foreign companies for digital advertising services. In 2020, this was also expanded to include a 2% tax on non-resident e-commerce operators for any kind of services provided. This was abolished last year. “The reduction of the 6% tax is a direct result of the Trump tariff threat,” said Pai. “This was because it was deemed as a non-tariff barrier, and the [Indian] government has demonstrated its commitment to having a free trade agreement with the U.S,” he added. He indicated that abolishing this tax will benefit the relationship between the two countries. However, some industry experts believe that the United States government may be exerting pressure. An investor, speaking to AIM on the condition of anonymity, stated, “Clearly, there is pressure from the U.S., which any government will obviously not agree to publicly.” Besides, Google and Meta saw positive results regarding their legal battles in India. In 2022, the Competition Commission of India (CCI) found that Google required app developers to use Google’s proprietary billing systems for in-app purchases only. However, in March, the National Company Law Appellate Tribunal (NCLAT) upheld CCI’s allegations but reduced the penalty to 50%. Also, in November last, CCI imposed a ₹213.14 crore penalty on Meta. This was due to WhatsApp’s 2021 privacy policy update, which requires users to accept data sharing with Meta. However, NCLAT directed WhatsApp\/Meta to deposit 50% of the penalty, and the previously imposed five-data sharing ban was lifted. Viswanathan KS, an independent digital transformation advisor, told AIM that he does not subscribe to the narrative suggesting that the United States government is applying pressure. He asserts that these are simply compliance issues and interpretations of the law and its legal provisions. “Obviously, each of the affected parties has gone to court and has received an interpretation,” he said, indicating that whatever law the new interpretations bring must be accepted and taken forward. “Fixing the interpretation once and for all is a good thing for the country. I am happy to happen early in the game, rather than discovering so late, when so many people have invested in India,” he added. So What About Make in India? In addition to Tesla’s entry, Starlink and other favourable developments for U.S. companies raise the critical question of whether these changes will hinder the ‘Make in India’ initiative and the country’s long-standing ambition to produce homegrown products. Pai mentioned that these developments could, in fact, benefit ‘Make in India’. “Tesla will buy a lot of components from our [India’s] vendors, to improve the technology in the country,” said Pai. Having said that, the Indian government imposes high import tariffs, reaching up to 100% on fully assembled EVs imported to India. This has been a contentious issue, which President Trump has talked about time and again with Indian counterparts. A new policy indicates that companies that set up and invest at least $500 million in manufacturing facilities for EVs will be allowed limited imports (fewer than 8,000) of cars priced at $35,000 and above, at a lower customs duty of 15%. Recently, it was reported that Tesla’s strategy to enter the country includes a proposal to build a manufacturing plant with an annual capacity of 500,000 vehicles at an investment of $2 – 3 billion. “Tesla may only import in the beginning because they want to test the market, but I hope it ties up with companies like Tata, Mahindra, or somebody else to manufacture them,” Pai added. Additionally, Pai believes introducing Starlink will benefit national security interests, especially in the fight against terrorism. He also noted that it is perhaps challenging for organisations in India to build infrastructure for a technology like Starlink, given the investments it needs. “India should forge a close partnership with the United States, reduce and eliminate all tariffs for US goods to come here,” he further opined. Meanwhile, Ashok Chandak, president of the India Electronics and Semiconductor Association (IESA), told AIM, “Negotiating a bilateral trade deal could ease pressure, while adjusting import tariffs on select US goods may address concerns.” “Fortunately, both India and the US are eager to expand bilateral trade to $500 billion, creating opportunities for a mutually beneficial agreement,” he added.","excerpt":"Big tech companies are having a great time in India. Is Trump’s pressure the reason?","categories":["AI Features"],"tags":["AI (Artificial Intelligence)","donald trump","India"],"author_name":"Supreeth Koundinya","publish_date":"2025-04-03T16:15:05","publication_year":"2025","word_count":991,"keywords":["Go","programming_languages:R","AI","RPA","digital transformation","programming_languages:Go","donald trump","Git","Aim","GAN","R","India","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","Aim","R","Go","Git","GAN","digital transformation","RPA","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/india-cozying-up-to-big-tech-under-trumps-tariff-heat\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10011243,"title":"Behind IndiaFirst Life AI Tech Strategy","content":"Setup in 2009, IndiaFirst Life insurance company has since been offering the best solution for customers’ insurance needs. With a paid-up share capital of ₹663 crores, it is one of the country’s youngest life insurance companies. Promoted by two large public-sector banks — Bank of Baroda with 44% stake and Union Bank of India with 30% stake, it is one of the few insurance companies in India to break-even within five years since inception. To drive its working, IndiaFirst Life largely depends on technology across functionalities such as customer engagement, operational efficiency, differentiation & risk management. To better understand their tech strategy, the impact of COVID on tech and data, emerging AI trends post-COVID and more, we got in touch with Sankaranarayanan Raghavan who is the chief technology and data officer at IndiaFirst Life Insurance Company Limited. Raghavan believes that analytics and data science will help them in creating a better customer, employee experience in addition to improving productivity and efficiency. “We also believe that we can innovate on products that are best suited for a segment\/market, based on analytics and data science,” he said. They have also tied up with many insurtech companies such as Toffee Insurance, Riskcovry and Gramcover for tasks such as the better distribution of their insurance solutions through digital platforms, providing innovative and bit-sized insurance propositions to different partners, offering group insurance-based solutions, among others. Below is the detailed interview: How does the company use analytics and data science in its overall working? Please share some use cases. Currently, we are using analytics and data science in sales and operations. In sales, it is used right from the time we are hiring the right talent to select the area\/segment the person should be allocated. While onboarding a new customer, based on the customer’s initial data, our model suggests the right product that reflects his persona and provides options to him. In addition to that, our model predicts the possibility of fraud and persistency. All the models deployed can learn and update itself. We are in the process of implementing AI-based OCR for the documents that are submitted, which are further scrutinised automatically. In operations, other than the scrutiny, we use models in renewals to increase the persistence, revenue and at claims stage to weed out fraudulent claims. We are in the process of deploying a model, which will help in assessing the case from underwriting guidelines. What does the overall tech strategy at the company look like? What are some of the new implementations and innovations in emerging tech? The tech strategy is focused on deploying technology and tools that help to reduce the time, give better efficiency, have better usage of AI\/ML in both the customer and employee life cycle. The aim is to provide superior customer experience and be light on architecture with full movement to microservices. While we have embarked on the hybrid cloud, we will be expanding it further to help the digital arm of the business. We have also recently deployed and implemented low code no platform giving one universe app to the sales force. Our tech stack is pretty much a new age with a focus on voice and multilingual. How are AI and data science implemented across domains such as customer engagement, operational efficiency, risk management and others? Early Claims Prediction Model: Developed an ML Model to flag out plausible early claims at the time of customer onboarding.Fraud Detection Model: Identifying fraud at the time of onboarding and at regular intervals during claims.Sales Nudges: Inputs that help the sales team to spend their energy at the right and desired areasApplication Stage Persistency Prediction: Our ML model will run during the form filling by Identifying the predictability of the persistency at the onboarding stage.Customer engagement at the right time: This is about when to intervene and how to intervene, creating a win-win situation What have been some of the impacts of COVID on technology and data? Given that IndiaFirst Life has embarked on the digital journey for a few years now, the shift to the new normal wasn’t felt — for both customers and employees. Our onboarding is 100% electronic for the past two years. Employees are highly mobile with a tablet\/laptop. In servicing, we have servicing capabilities available in WhatsApp, Portal, IVR, chatbot and email. Post-COVID, we have seen a 2X uptake on the digital channels giving the benefit of early implementation of digital-early. Our Sales team uses e-Sampark an innovative way of reaching out to the customers. Our ‘Ghar Bhaite Insurance’ initiative has shown good uptake too. What are the emerging trends around digital, data and technology disruptions in the industry? As India is becoming Atmanirbhar and transitioning to Bharat, the voice is going to disrupt the industry. The transition of Insurance from being a push to pull product, together with Voice, multilingual and simple products, will increase not only the insurance adaption in the country but also to better-secured society. How can companies capitalise on digital-tech developments and brace the firm for a more optimistic picture? Digital technology helps in modularising things, making it agile, speeder time to market, experiment new innovative models and reach the unreached. These features of the digital-tech bring a whole lot of opportunity to anyone who has got an idea to execute it. This empowerment and ability to make people experiment gives raises a newer way of reaching out, thereby opening up a whole lot of opportunities.","excerpt":"Setup in 2009, IndiaFirst Life insurance company has since been offering the best solution for customers’ insurance needs. With a paid-up share capital of ₹663 crores, it is one of the country’s youngest life insurance companies.  Promoted by two large public-sector banks — Bank of Baroda with 44% stake and Union Bank of India with […]","categories":["AI Features"],"tags":["Interviews and Discussions"],"author_name":"Srishti Deoras","publish_date":"2020-11-05T18:00:17","publication_year":"2020","word_count":906,"keywords":["data science","AI","R","ML","RAG","microservices","Ray","Aim","analytics","fraud detection","Interviews and Discussions"],"extracted_tech_keywords":["AI","ML","data science","analytics","Aim","Ray","RAG","fraud detection","microservices","R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/behind-indiafirst-life-ai-tech-strategy\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10131325,"title":"Indian Engineers are a Different Breed","content":"It is often said that Indian techies are stealing jobs across the globe, but the reason is much more than just outsourcing and cheap labour. The reason is that Indian techies are more dedicated when working and ready to work overtime without hesitation. A week ago, Roshan Patel, the founder & CEO of Arrow Payments, shared the screenshot of a conversation he had with his employee. Patel asked if the employee would like a break as he’d been working nonstop for quite a while. To his surprise, the engineer replied, “I don’t need a break sir. My body is a vessel for the company to find product market fit.” indian engineers are a different breed pic.twitter.com\/fYdMundMfy— Roshan Patel (@roshanpateI) July 24, 2024 “Indian engineers are a different breed,” Patel remarked. That’s certainly turning out to be true. Shravan Tickoo, the founder of Rethink Systems, posted the same story, but with a twist, adding that the employee was working at 4 am, which was not the case. He revealed that in reality, it was the WhatsApp bot that sent random answers to the CEO when he texted the engineers. Pooja Goel in the comments narrated a story about how one of her family members used to schedule messages for 2 am to his manager, just to make it look like he’d been working post midnight. Though, it definitely happens a lot of the time. Time zones do not seem to matter to Indians. Tickoo funnily added that startups from the HSR Layout in Bengaluru have now started searching for this employee to hire him as the CTO of their companies. “Since that day, Mr Narayana Murthy and Bhavish Aggarwal have been considering this engineer as their heir apparent,” he added in jest. Indian Engineers Eat Code for Breakfast Not everyone is happy being the overworked employee though. People appreciate Patel’s message to the employee for encouraging him to take a break, but are actually also concerned about the broader trend of working overtime. “This is actually PTSD [post-traumatic stress disorder] from working under toxic Indian management for years,” said a user. “Colonial hangover of subservience,” remarked another. However, some argue that working long hours doesn’t necessarily mean increased productivity. “Yet the American-born engineer who takes 1-hour lunch breaks and two-week PTO is still more productive,” noted a user. Ayush Jaiswal shared a screenshot of his chat with his friend about him working 18 hours a day for the past two months and how his productivity has taken a hit and so has the work-life balance. “Take a break, the world won’t end,” commented a user. Arguably, AI tools are making the world a lot better for these developers, enabling them to accomplish much more quickly. There are a lot of jobs for 10X developers, who are able to take on the jobs of their peers and also work overtime. What’s the Bug? In Karnataka, the state IT\/ITeS employees union (KITU) has reported that the government may soon raise the IT employees’ working hours to 14 (12 hours + 2 hours of overtime). This might be a good idea for a lot of employees who would now be compensated for all the extra work they are already doing. However, this overlooks the critical issue: the bug is the overwork culture itself. The relentless pursuit of recognition and better pay leads many Indian engineers to work excessive hours, sacrificing their health and personal lives. This culture of overwork can lead to burnout, reduced productivity over time, and a skewed work-life balance. When it comes to working for startups, most Indian employees get attached to the product they are building, and if fairly compensated, they are happy to continue working over and above what is expected of them. The ones who do not get fairly compensated resort to working on multiple projects. Recently, the photo of a Microsoft employee driving a Namma Yatri auto to combat loneliness went viral. This also narrates the story of what happens if you get addicted to working overtime.It seems like the underlying issue is the desire to make a lot of money and freshers want to learn as much as possible to be able to compete or do as-well the senior resources. When Indians look at the salaries of their counterparts in the West, the most common conception is that they are not as skilled as them. Therefore, they decide to put in longer hours in a hope to upskill, or possibly get recognition from their bosses to get better pay. While the dedication and hard work of Indian engineers is commendable, it’s important to address the underlying issues of overwork and its long-term impact on health and productivity.","excerpt":"Burning the midnight oil, fixing bugs, and driving autos to combat loneliness.","categories":["IT Services"],"tags":["AI engineers"],"author_name":"Mohit Pandey","publish_date":"2024-08-05T14:24:43","publication_year":"2024","word_count":781,"keywords":["Go","programming_languages:R","AI","AI engineers","RPA","programming_languages:Go","RAG","Ray","ViT","R","startup"],"extracted_tech_keywords":["AI","Ray","RAG","R","Go","ViT","RPA","startup","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/indian-engineers-are-a-different-breed\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":31058,"title":"OnePlus To Focus On AI And ML At Its New R&#038;D Centre In Hyderabad","content":"With emerging trends in artificial intelligence and machine learning along with the currently existing cut-throat competition within the Indian smartphone Industry, OnePlus has announced its plans of setting up one of its first R&D facilities in Hyderabad. According to reports, this facility will purely focus on promulgating AI and ML in their products specifically for Indian consumers. OnePlus plans to converting it into one of its largest R&D centres in the next three years. The centre will also contribute to nurturing and giving opportunities to fresh talent as the company initiates the campus hiring programme at a number of universities including IIT Delhi and IIT Mumbai this year. Pete Lau, founder and CEO, OnePlus said, “We plan to re-focus our R&D efforts on a large scale and drive innovations in India for the global product, especially on the software side with a special emphasis on AI and ML.” OnePlus stated in an official press release, “India is not only one of the biggest markets for OnePlus but is also the benchmark for creating globally successful products. With a growth rate of more than 100%, the country is well on its way to becoming a Centre for product innovation.” According to OnePlus, Hyderabad is one of the active hot startup zones in India for emerging futuristic technologies with a dedicated ingenious talented pool that can bring new innovations, especially in their OnePlus Products. Along with that OnePlus is trying in expand its offline presence too with its own multiple OnePlus Experience Stores due to its increasing user base. OnePlus has reportedly already started the production of PCBs and other smartphone components, and now with this new R&D Centre, it surely will play a crucial role in its plans of capturing the Indian market in long-term. OnePlus seems to be committed to being part of India’s long-term growth story with mulling in efforts by integrating its involvement for the Make in India strategy at large.","excerpt":"With emerging trends in artificial intelligence and machine learning along with the currently existing cut-throat competition within the Indian smartphone Industry, OnePlus has announced its plans of setting up one of its first R&D facilities in Hyderabad. According to reports, this facility will purely focus on promulgating AI and ML in their products specifically for […]","categories":["AI News"],"tags":["AI (Artificial Intelligence)","Hyderabad","Machine Learning","Make in India","oneplus","research and development","smartphones"],"author_name":"Martin F.R.","publish_date":"2018-12-04T12:35:40","publication_year":"2018","word_count":323,"keywords":["Make in India","artificial intelligence","machine learning","programming_languages:R","AI","research and development","ML","innovation","Machine Learning","smartphones","Hyderabad","oneplus","R","AI (Artificial Intelligence)","startup"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","R","innovation","startup","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/oneplus-ai-ml-hyderabad-research\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10113819,"title":"Why is Google Eyeing Reddit&#8217;s Data?","content":"Recently, Google signed a data licensing agreement with Reddit, reportedly costing $60 million per year, to access the social media and news aggregation platform’s real-time content through its data API. This has come close to Reddit’s initial public offering (IPO) in March this year. Reddit, the first major social media company to go public since Pinterest in 2019, will likely be aiming for a valuation of at least $5 billion in the IPO. Since the first announcement in 2021, the company has been making sweeping changes to allegedly increase its revenue and further its valuation. Besides, in a big change, Reddit eliminated free access to most of its API. Developers now need to pay for access, with pricing based on the number of requests made, which were exorbitant, shutting down alternative apps to the site. The primary purpose of this is to provide Google’s AI with a large amount of conversational data to train and improve their LLMs. The users on Reddit sign off on the rights to the platform to use the content however they see fit. However it is unclear how anonymisation of the data will be applied before being used for training. Reddit also benefits by leveraging Google’s Vertex AI, improving its not-so-good search feature. Useful Data? Most of Reddit is known for conversations that range from flippant to hateful. The polarised crowd in each Subreddit form their own echo chambers. The recent example of conversations on Russia-Ukraine war showed a bias towards Ukraine often spewing vitriolic speech when countered this stance. Although the social media platform is regulated by moderators trying to curb hate speech, it is not always successful as the Subreddit rules vary. Bindu Reddy, CEO and co-founder of Abacus.AI said on X, “Once they (Google) pre-train their model on this largely uncensored corpus where humans routinely reveal their true opinions, they will spend > $60M suppressing the Reddit content, nerfing, and nudging their model to reflect their ideology!” Popular for its diverse content, the top comments on most of the posts are funny and satirical. The irony, satire, and humour offer a unique data set for training Google’s AI, contributing to a deeper understanding of complex human communication. The data on Reddit is also organised, sorted and clear, with information on the upvotes making it structured. This gives Google instant access to all types of data to use this information better. The company plans to use “enhanced signals” to improve how it shows information, like showing more content from Reddit as announced by Google and Reddit. From training on such content to AI being able to grasp nuances, identify misinformation through satire, and improve generative models with informal language and creativity, it’s a long way to go. However, challenges arise from the heavy reliance on context, potential bias amplification, and the limited generalisability of niche or slang-dominated data. The company blog further added that Reddit searches are popular on Google, “This partnership will facilitate more content-forward displays of Reddit information that will make our products more helpful for our users and make it easier to participate in Reddit communities and conversations.” AI experts suggest that messy, diverse data can enhance model performance, emphasising the importance of sophisticated filtering and a balanced training approach. Privacy advocates, however, raise concerns about using even anonymised Reddit data for advancing profiling and targeted advertising techniques. The massive downside is the inaccurate information, during the peak of COVID-19, the company said, it would leave up Subreddits that spread misinformation related to Covid-19. Days later, after protest from many of its own users, Reddit banned the forum in question, saying it had violated other rules. Meanwhile, Google has paused its image generator and apologised for it ‘missing the mark’ owing to inaccuracies. The irony here is that the model was inclusive but in showing African and Asian Nazi soldiers. Though training on Reddit’s (although structured) data would be easier, the questionable quality of data could lead to similar problematic outputs, where harmful stereotypes or misinformation are reinforced rather than detected.","excerpt":"Google forks out over $60 million to Reddit in a deal to rent a dive into wild conversations to train its AI, all while Reddit gets to spruce up its search capabilities with Google’s AI magic.","categories":["Global Tech"],"tags":["AI Tool","Google","reddit"],"author_name":"K L Krithika","publish_date":"2024-02-24T12:00:00","publication_year":"2024","word_count":671,"keywords":["Go","API","TPU","AI","IPO","reddit","RAG","Aim","ViT","Google","GAN","AI Tool","R"],"extracted_tech_keywords":["AI","Aim","RAG","TPU","R","Go","API","GAN","ViT","IPO"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/why-is-google-eyeing-reddits-data\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10094785,"title":"ONDC has a Data Problem","content":"The Indian government has projected Open Network for Digital Commerce (ONDC) as a platform to democratise e-commerce in India and break the hold of private players such as Amazon and Flipkart on the Indian market. While addressing the ‘ONDC Elevate’ programme last month, Union Minister Piyush Goyal said that ONDC is an engine of growth and could redefine the e-commerce industry completely. ONDC has been widely discussed due to its highly competitive price on food delivery, challenging the dominance of established players like Swiggy and Zomato. Many users flocked the platform to avail the products at such cheap prices. It’s important to note that sellers on ONDC platform were able to sell products at such low prices because there’s no intermediary in between to take a percentage cut on every sale. According to reports, ONDC will soon start offering mutual funds, insurance, and personal loan, among other things. Users on the platform will only increase, even if not at par with other e-commerce platforms. However, amid this, the concerns regarding the approach of ONDC towards data privacy, particularly in the absence of a comprehensive data protection bill in the country remain. Data Privacy concerns “E-commerce platforms have a multitude of data points and they’re used for competitive pricing, recommendations, etc. Our concerns essentially stem from the fact that there is an absence of the data protection bill in the country,” Tejasi Panjiar, associate policy counsel, Internet Freedom Foundation told AIM. The ONDC Strategy Paper does suggest that data points such as Personal Identifiable Information (PII) as well as seller data critical to trade will remain siloed within the buyer app and seller app, respectively and will be protected from third-party access. It also mentions that ONDC may collect personal data about a natural person, living or deceased. However, in the absence of a data protection law, storing consumer data becomes a privacy concern and further, how ONDC is handling these data points remains unknown. Moreover, the digital rights advocacy group has also raised concerns that arise regarding the network-wide sharing of ‘anonymised performance metrics.’ These concerns persist as the re-identification of anonymised data remains feasible with just a few data points, such as pin code, date of birth, and age, which can accurately identify users from anonymised datasets. While once the Digital Personal Data Protection Bill (DPDPB) comes into effect, ONDC has to adhere to the laws of the land, however, Panjiar said that the bill does not do much to address the rights of the citizens either. Does not come under RTI ONDC was incorporated on December 31, 2022 with an initial investment from the Quality Council of India (QCI) and Protean eGov Technologies Limited. ONDC is not a government agency or a Public Sector Undertaking (PSU). It is a private non-profit company established by the Department for Promotion of Industry and Internal Trade (DPIIT). This has been done to diminish the motivation for profit maximisation and prioritise ethical and responsible conduct. However, by doing so, according to the IFF, ONDC does not come under the ambit of Right to Information (RTI). “Given that ONDC is a Section 8 company and not a public authority, it falls outside the ambit of RTI,” IFF said in a blog post. The exclusion of the ONDC from the purview of the RTI raises a significant issue. By not being subject to RTI, there is a lack of mechanisms for common citizens to obtain crucial information regarding ONDC’s operations, decision-making processes, and adherence to accountability and transparency standards. “Since our fundamental rights are enforceable against the State, aggrieved citizens won’t be able to avail their right to file a writ petition under Article 32 and 226 in case of violation of their rights,” IFF said. Additionally, this absence of transparency creates an environment where the public is unable to effectively monitor whether ONDC is upholding best practices and acting in the interest of the citizens it serves. Grievance redressal Piyush Goyal once said that it is essential for the ONDC to establish consumer trust by implementing strong mechanisms to address consumer grievances and enforce transparent policies regarding returns, refunds, and cancellations. Other e-commerce has built consumer trust upon a strong consumer grievances redressal channel and it’s a welcoming move for ONDC to have similar channels. However, as things stand, ONDC does not have such a mechanism in place. All users can do is write to an email id to share their grievances. Additionally, the multiplicity of stakeholders within ONDC, encompassing buyers, sellers, and logistics, pose potential obstacles due to the complexity and coordination required among these various parties, potentially impeding the seamless functioning of the network. So far, many users have taken to Twitter to vent their frustrations about the unavailability of a proper grievances redressal mechanism. Further, there is also a lack of clarity regarding the resolution of privacy-related grievances and breaches of data within the existing framework. Given there is no robust grievances mechanism in place currently, the potential for redressal of such concerns remains uncertain, raising questions about the overall accountability and trustworthiness of the system.","excerpt":"ONDC does not come under the ambit of Right to Information (RTI) either","categories":["AI Features"],"tags":["Internet freedom foundation"],"author_name":"Pritam Bordoloi","publish_date":"2023-06-08T18:00:00","publication_year":"2023","word_count":846,"keywords":["Go","AWS","AI","cloud_platforms:AWS","programming_languages:R","ML","Git","Internet freedom foundation","Aim","Rust","R"],"extracted_tech_keywords":["AI","ML","Aim","AWS","R","Go","Rust","Git","cloud_platforms:AWS","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/ondc-has-a-data-problem\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10009268,"title":"Gaming, Movies And Architecture: How NVIDIA Omniverse Is Reshaping The World As We Know It","content":"“From Google Docs to 3D Graphics, NVIDIA Wants To Power All Your Applications with Omniverse.” As the world embraces digital transition, there are hardly any processes in any industry that are not touched by software. Today, we have software that writes software. Software that simulates the real world so that the creators don’t have to expend resources on failed or narrow ideas. At the cusp of this simulated reality, NVIDIA wants to extend their services so that everything from skyscrapers to sports cars and smart cities will be designed and tested in simulation. At the ongoing GTC 2020 event, NVIDIA revealed the details of its most ambitious venture yet — Omniverse. Millions of designers, architects and other creators will soon be able to collaborate in real-time, whether on-premises or remotely, with the NVIDIA Omniverse platform, which NVIDIA announced has entered open beta. Omniverse is designed to provide efficient real-time scene updates and is based on open standards and protocols; to act as a hub for enabling new capabilities to be exposed as micro-services to any connected clients and applications. Creators, Assemble! Source: NVIDIA With Omniverse NVIDIA brings together breakthroughs in graphics, simulation and AI, Omniverse is the world’s first NVIDIA RTX-based 3D simulation and collaboration platform that fuses the physical and virtual worlds to simulate reality in real-time and with photorealistic detail. Companies such as Autodesk and Bentley Systems are already exploring what USD can do for this community. For Architects Architects need a way to quickly review progress on construction projects with owners and real-estate developers. Each stakeholder wants different programs often running on different computers, tablets or even handsets. This is where the Omniverse can come in handy. It offers services like real-time path tracing, developing surrounding context, interactive design review with clients andaugmented reality design review with Cloud XR “Universal Scene Description(USD) helped Hollywood designers collaborate. Now it’s playing similar roles for architects, product designers and more.” For Manufacturing industry Manufacturing companies deal with thousands of complex components that need to be quickly designed and tested, ranging from smartwatches to autonomous trucks. This exhaustive process requires iterations and runs through the hands of many kinds of specialists who demand photorealistic 3D models. Omniverse explores the possibilities like putting visualisations in the hands of online customers. For instance, thanks to USD in Omniverse, NVIDIA could make a video demonstrating the NVIDIA DGX A100 system that has exploding views of how its 30,000 components snap into a million drill holes. For Autonomous Industry Omniverse can help engineers model cars and even the details of a highway. The simulation details are good enough to test both vehicles and their automated driving capabilities. For instance, robot car consumers can leverage Omniverse simulation before building or buying a physical robot. For Gaming and Movies Hollywood studios have already adopted USD. Omniverse incorporates the power of NVIDIA RTX real-time ray-tracing graphics into USD’s collaborative, layered editing. For example, Epic’s Unreal Engine supports USD. Unity and Blender support it as well. Whereas, Adobe’s developers were among many others who welcomed USD and used it for their products. “USD has a whole world of features that are incredibly powerful,” said Davide Pesare, who worked on USD at Pixar and is now a senior R&D manager at Adobe. Pixar even produced one of their blockbusters, Finding Dory, using USD. Dreamworks, as well as Disney, have blended USD into its pipeline for animated features, too. So, what gives NVIDIA the confidence to venture into this vibrant ecosystem and play the part of a hub that pushes the envelopes of creativity? NVIDIA is again relying on something they are the best at doing — building GPUs. “Second-gen RT Cores add hardware acceleration for ray-traced motion blur rendering, delivering up to 7x faster performance than the last generation.“ At GTC, NVIDIA also introduced a flagship GPU for creators — the NVIDIA RTX A6000. The new GPU also supports graphics workloads at scale for remote users, enabling larger creative workflows and making it perfect for working from home. RTX A6000 GPU Built on the new NVIDIA Ampere architecture, the RTX A6000 GPU, allows creators to complete complex creative tasks in real-time and animate extra-large 3D models. It also offers the largest memory available in a single GPU and is expandable to 96GB using NVLink to connect two GPUs. Know more about it here.","excerpt":"“From Google Docs to 3D Graphics, NVIDIA Wants To Power All Your Applications with Omniverse.” As the world embraces digital transition, there are hardly any processes in any industry that are not touched by software. Today, we have software that writes software. Software that simulates the real world so that the creators don’t have to […]","categories":["Global Tech"],"tags":["GPU","Pixar"],"author_name":"Ram Sagar","publish_date":"2020-10-09T11:00:44","publication_year":"2020","word_count":723,"keywords":["Go","programming_languages:R","AI","A100","programming_languages:Go","Git","RAG","Ray","ViT","R","Pixar","GPU"],"extracted_tech_keywords":["AI","Ray","RAG","R","Go","Git","ViT","A100","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/nvidia-omniverse-games-architecture-pixar-usd\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10091568,"title":"Akaike Technologies is Certified as a Best Firm for Data Scientists","content":"Akaike Technologies is certified as the Best Firm For Data Scientists to work for by Analytics India Magazine (AIM) through its workplace recognition programme. The Best Firm For Data Scientists certification surveys a company’s data scientists and analytics employees to identify and recognise organisations with great company cultures. AIM analyses the survey data to gauge the employees’ approval ratings and uncover actionable insights. “We are incredibly honored to have been recognized as the Best Firm for Data Scientists by the esteemed Analytics India Magazine. As a researcher and data scientist, I have collaborated with my exceptional team to cultivate an environment of exploration, learning, and boundary-pushing initiatives. When I founded Akaike Technologies, my aspiration was to create a company culture where I would be excited to work as a data scientist myself. It’s truly gratifying to see that this approach has resonated with our clients and employees alike. Today, we boast a team of 30+ passionate data scientists who are dedicated to our mission of accelerating AI adoption for every organization” said Rahul Thota, CEO and Founder, Akaike Technologies. The analytics industry at AIM faces a talent crunch, and attracting good employees is one of the most pressing challenges that enterprises are facing. The certification by Analytics India Magazine is considered a ‘Gold Standard’ in identifying the best data science workplaces and companies participate in the programme to increase brand awareness and attract talent. Best Firms For Data Scientists is the biggest data science workplace recognition programme in India. To nominate your organisation for the certification, please fill out the form at this link.","excerpt":"The Best Firm For Data Scientists certification surveys a company’s data scientists and analytics employees to identify and recognise organisations with great company culture.","categories":["AI Features"],"tags":["best companies for data scientists in india","best companies in india for data science fresher","best companies to work for as data scientist in india"],"author_name":"AIM Media House","publish_date":"2023-04-17T15:00:00","publication_year":"2023","word_count":264,"keywords":["best companies for data scientists in india","data science","Go","best companies to work for as data scientist in india","programming_languages:R","AI","programming_languages:Go","best companies in india for data science fresher","Aim","analytics","GAN","R"],"extracted_tech_keywords":["AI","data science","analytics","Aim","R","Go","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/akaike-technologies-is-certified-as-a-best-firm-for-data-scientists\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10115350,"title":"10 Underrated Women in AI to Watchout For","content":"Recent data shows that women hold approximately 26.7% of technology-related jobs. Despite the critical role technology and artificial intelligence play in shaping our future, the presence of women in these fields, especially in leadership positions, remains abysmally low. However, despite these challenges, (the few) women in AI are pioneering change, breaking barriers, and paving the way for future generations. Their work not only contributes to technological advancement but also ensures that the development and application of AI is inclusive, equitable, and representative of the diverse society it serves. While the media constantly covers the tech elites, here is a list of 10 underrated women whose works range from spreading awareness about AI to building them and ensuring its ethical use. Aishwarya Srinivasan Aishwarya Srinivasan currently works as a senior AI advocate within the Microsoft for Startups group at Microsoft, supporting startups in developing machine learning solutions. She recently became an angel investor with Hustle Fund and DynamoFL, after founding Illuminate AI, a nonprofit dedicated to mentoring in AI, in March 2021. Prior to joining Microsoft, she was a data scientist at Google Cloud and an AI & ML innovation leader at IBM Data & AI. Aishwarya has a postgraduate degree in data science from Columbia University. She has global work experience having engaged with clients and led projects in London, Dubai, Istanbul, and India. She holds a patent awarded in 2018 for developing a reinforcement learning model for machine trading. Tulsee Doshi Tulsee Doshi leads Google’s Responsible AI & Human-Centred Technology Organisation. Her work focuses on incorporating ethical considerations into product development and policy, specifically aimed at creating equitable and transparent user experiences. At Google, Doshi manages a team dedicated to improving fairness, safety, and inclusivity across products. Prior to her current positions, Doshi taught a product leadership course at Product School and led projects at YouTube to enhance inclusive machine learning and creator diversity. She started at Google as an associate product manager, focusing on making Search features more relevant worldwide. Doshi holds a bachelor’s degree in symbolic systems and a master of science in artificial intelligence from Stanford University. At Stanford’s HCI Lab, she was involved in research on crowdsourcing and expert collaboration technologies, earning a Best Paper Award at UIST 2014. Ritu Raman Ritu Raman is the d’Arbeloff assistant professor of mechanical engineering at MIT, where she leads a lab focused on developing adaptive living materials, with current projects centred on engineering biological actuators. This research aims to enhance machine functionality and restore mobility in humans. Raman’s work involves integrating living muscular and neural tissues to create actuators that autonomously adjust to environmental changes. Her academic background includes a bachelor’s in mechanical engineering with a minor in biomedical engineering from Cornell University, followed by an MS and PhD in mechanical engineering from the University of Illinois at Urbana-Champaign. Isabelle Guyon Isabelle Guyon is a director and research scientist at Google, on leave from her role as a professor of artificial intelligence at Université Paris-Saclay (Orsay), specialising in data-centric AI, statistical data analysis, pattern recognition, and machine learning. Before her current position at Google, Guyon worked as an independent consultant and a researcher at AT&T Bell Laboratories. There, she made significant advancements in neural networks for pen computer interfaces, collaborating with Yann LeCun and Yoshua Bengio. She is the primary inventor of SVM-RFE, a variable selection technique based on SVM, widely cited and used as a benchmark for new feature selection methods. Guyon holds a PhD degree in physical sciences from the University Pierre and Marie Curie, Paris. Olga Russakovsky Olga Russakovsky, Ukrainian-American, completed her PhD in computer vision at Stanford University in 2015, working closely with Fei-Fei Li. Together, they developed ImageNet, a comprehensive image database pivotal for advancements in computer vision. Her research focused on reducing image classification’s dependency on human annotators and addressing human bias in algorithm development. Following her PhD, Russakovsky continued her research as a postdoctoral fellow at Carnegie Mellon University and is now an associate professor at Princeton University. Her work emphasises the importance of algorithmic fairness in visual recognition systems and has proposed computational solutions to mitigate historical and societal biases. Additionally, her significant contributions to the field include leading the Imagenet Large Scale Visual Recognition Challenge, with the founding paper cited over 13,000 times. Devi Parikh Devi Parikh, who is the senior director of generative AI at Meta, worked on developing Make-A-Video 3D in 2023 and Make-A-Video in 2022 that propelled the text-to-video generation. In the same year, she introduced AudioGen, a novel text-to-audio generation tool, and Make-A-Scene, which allows for more creative control in AI-generated images. These contributions follow her earlier endeavours to humanise AI research, evident in her 2020 projects ‘Humans of AI: Stories, Not Stats’ and ‘AI Paygrades’, aimed at demystifying the AI research community and promoting transparency in AI industry hiring practices, respectively. Aude Oliva Aude Oliva serves as the director in the MIT-IBM Watson AI Lab and director of strategic industry engagement in the MIT Schwarzman College of Computing. Her latest research uses deep learning to enable computers to recognize locations within images based on their composite features, such as identifying a bedroom from the presence of a bed, window, and posters, or a kitchen from a stove, tile, and countertop. Oliva’s significant contributions include her work on hybrid images. This work supports applications in information privacy, time-lapses, marketing, and brainteasers. Additionally, she explores the psychological perception of images, focusing on memorability, content, and the human visual system’s limitations. Daphne Koller Daphne Koller, an Israeli-American computer scientist, co-founded Coursera in 2012. Before Coursera, she worked on probabilistic models with applications in various domains, including computer vision and computational biology. Koller was a Stanford University professor and received the ACM-Infosys Foundation Award in Computing Sciences in 2008, the first recipient of this $150,000 award. She became a MacArthur Fellow in 2004, recognized for her innovative work in AI. After Stanford, she pursued postdoctoral research at UC Berkeley. Koller has been honoured by being elected to the National Academy of Engineering in 2011, the American Academy of Arts and Sciences in 2014, and the National Academy of Sciences in 2023. In recent years, Koller left Coursera to focus on new ventures in biotech, founding Insitro in 2018, a drug discovery startup leveraging machine learning and genomics. In 2020, she co-founded Engageli, offering an innovative online learning platform. Cynthia Rudin Cynthia Diane Rudin is a professor at Duke University, where she directs the Interpretable Machine Learning Lab. In 2022, she was honoured with the Squirrel AI Award for her contributions to transparent AI systems in critical domains. She also received the Guggenheim Fellowship in the same year and was elected a Fellow of the Association for the Advancement of Artificial Intelligence. Her research includes developing the Series Finder algorithm for crime series detection and scoring systems for medical diagnosis. Before her tenure at Duke, Rudin was a faculty member at the MIT Sloan School of Management and held research positions at New York University and Columbia University. She completed her PhD in applied and computational mathematics at Princeton University in 2004 and has been recognised as one of the most impressive professors at MIT by Business Insider in 2015. Daniela L Rus Daniela L Rus is the director of the MIT Computer Science and Artificial Intelligence Laboratory (CSAIL) and holds the Andrew and Erna Viterbi Professorship in the Department of Electrical Engineering and Computer Science at MIT. Rus’s contributions address challenges in machine learning such as data quality, bias, and adaptability of systems, alongside innovations in robotics like soft robotics, modular robots, and autonomous vehicle algorithms. She has developed technologies to assist the physically disabled. Rus was born in Romania and moved to the United States, earning her bachelor’s degree at the University of Iowa and her PhD at Cornell University. She started her academic career at Dartmouth College before moving to MIT in 2004. Join Rising 2024, the largest summit on Diversity and Inclusion in India, taking place on April 4-5 in Bangalore. Grab your passes now.","excerpt":"In a tech world where diversity is critical yet often overlooked, unsung women are enriching the AI field with innovations that promise a future where technology is as diverse as the society it serves.","categories":["AI Trends"],"tags":["Top Trend","Women in Tech"],"author_name":"K L Krithika","publish_date":"2024-03-11T18:03:06","publication_year":"2024","word_count":1340,"keywords":["Top Trend","data science","artificial intelligence","machine learning","AI","neural network","ML","computer vision","Aim","deep learning","Women in Tech","generative AI"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","neural network","computer vision","data science","generative AI","Aim"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/10-underrated-women-in-ai-to-watchout-for\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10081503,"title":"8 Must-attend Indian Developer Conferences","content":"Developer conferences offer many opportunities to advance your career. Apart from expanding your network, you discover the best ongoing practices in the domain. There’s even a chance to get hired. Not to mention, it is a beneficial time-out from the monotonous day-to-day work. Listed below are the events slated for 2023 to help you make the most out of your next developers’ event! Machine Learning Developers Summit (MLDS) The 5th Edition of Analytics India Magazine’s in-house event, MLDS23, aims to unite India’s principle machine learning innovators and professionals to contribute to the community through their ideas and experiences with machine learning tools. Furthermore, they’ll discuss the advancements in the domain and give the attendees a first look at new trends and developer products. Some of the notable personalities at MLDS23 are: Sudalai Rajkumar – SRK, One of the five Quadruple Kaggle Grandmasters. Mamatha Naidu – Director, Analytics COE at JCPenney India Harsh Singhal – Head of Data Science & Machine Learning at Koo Check out the speakers’ list here. When and Where: January 19 to 20, 2023 NIMHANS Convention Center, Bengaluru, India Click here to register. JSConf 2023 Join the first Indian edition of the prestigious JS conference from the International JavaScript community. The not-for-profit international conference is being held on June 2, 2023. It is a platform for the Indian JS community to come together, share their ideas, and join the discourse around JavaScript. When and where: June 2, 2023 BIEC – Bengaluru International Exhibition Centre Click here to register. Laracon India The official community-driven Laravel conference, LaraconIN, is being held for the first time in India. The event will feature speakers from across the globe presenting talks on Laravel, PHP, VueJS, web development, and other topics. A must-attend event for all web developers. When and where: February 25–26, 2023 Club07, Ahmedabad Click here to register. Great International Developer Summit – India After two years of virtual events, meet face-to-face with leading software developer mentors and trainers. Soak in unparalleled wisdom and learning at the sixteenth cohort of the GIDS 2023 from April 25–28, 2023. The four-day programme includes a blend of technical talks and experiences featuring some of the most influential figures in software engineering and architecture. Furthermore, it offers ample opportunities to learn about companies at the forefront of creating products you have come to love and use in your projects at work. When and where: April 25–28, 2023 Venue TBA Click here to register. World DevOps Summit The 8th World DevOps Summit is a two-day event for developers to meet, explore, transform, and network. The event will dive into the latest concepts, processes, and tools to understand current DevOps challenges. Moreover, it is an opportunity for fellow DevOps to network with experts from leading companies. When and where: May 25-26, 2023 Ibis, Bangalore Click here to register. Open Source India The 20th edition of Open Source India (OSI) is an initiative by the Open Source For You magazine team in partnership with India’s Open Source community and industry to bring various stakeholders under one roof. The mission of OSI is to accelerate the development and execution of Open Source in India. When and where: October 12–13, 2023 NIMHANS Convention Center, Bangalore Click here to register. ng-India The ng-India is one of the largest Angular and JavaScript community gatherings in India. With around 700 people expected at the 2023 edition, you will meet the creators of Angular, RxJS Core team members, GDE, MVP, and authors coming from various countries of the world to share their insights with you. Join to learn, network, recruit and befriend top Angular and JavaScript developers. When and where: February 25, 2023 The Leela Ambience Convention Hotel, Delhi Click here to register. PGConf India With the growing adoption of PostgreSQL, PGConf India has become a must-attend event in the domain. With over 400 delegates attending this year’s event, it is an unmissable opportunity to present your work. This conference is an effort to expand the community in India by increasing awareness and providing the appropriate training. The event will cover a vast spectrum of topics, as we expect discussions ranging from internal talks led by leading developers to end-user case studies from small companies as well as large multinational corporations and government organisations—all of whom run their businesses on PostgreSQL. When and where: February 22–24, 2023 Radisson Blu Bengaluru Outer Ring Road Click here to register.","excerpt":"The 2023 edition of Analytics India Magazine’s MLDS will have 40+ speakers from nearly 400 tech companies.","categories":["IT Services"],"tags":["bangalore","DevOps","MLDS"],"author_name":"Tasmia Ansari","publish_date":"2022-12-05T16:00:40","publication_year":"2022","word_count":733,"keywords":["PostgreSQL","data science","machine learning","AI","ML","MLDS","Aim","analytics","SQL","JavaScript","DevOps","R","bangalore"],"extracted_tech_keywords":["AI","machine learning","ML","data science","analytics","Aim","PostgreSQL","R","SQL","JavaScript"],"url":"https:\/\/analyticsindiamag.com\/it-services\/8-must-attend-indian-developer-conferences-in-2023\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10043800,"title":"GPU-Accelerated Deep Learning Sorts Pottery Fragments","content":"From using AI for finding archaeological sites to leveraging computer vision algorithms for analysing satellite imagery and data, AI has made inroads into the field of archaeology. Linguistic anthropologists often utilise AI in archaeology to trace the evolution of different languages.DeepMind’s PYTHIA is an excellent example of this. Recently two researchers from Northern Arizona University: Leszek Pawlowicz and Christian Downum presented a study for classifying digital images of pottery shreds into an existing typological framework using deep learning. The research focuses on studying a specific kind of decorated pottery tradition from the American Southwest known as the Kayenta Series of Tusayan White Ware (TWW). It is primarily manufactured in the vicinity of modern-day Tuba City and Kayenta. The vessels feature asymmetrical geometric patterns made with black\/brown organic paint on a white background. For nearly a century, archaeologists have classified TWW into a temporally sensitive series of types useful for dating archaeological features, sites, and landscapes. However, questions have been raised on its use, both about the typology itself and the ability of archaeologists to apply it consistently and accurately to the classification of pottery fragments (sherds). Archaeologists have collected and catalogued tens of thousands of TWW sherds and vessels from hundreds of ancient pueblo and pit house sites. Despite the TWW typology’s demonstrated value for dating archaeological occurrences, it suffers from some of the same issues that plague other archaeological typologies. Many of the typological systems employed by archaeologists, like the Southwest’s ware-type-variety system, were developed decades ago and are now theoretically obsolete. In this study, the authors explore an alternate approach to TWW typology that involves the use of deep learning\/convolutional neural networks for image classification of this decorated ceramic ware. While this study is specific to TWW, the methods and outcomes may have wider implications for analyses of other artefact categories. CNN for TWW classification In the paper, researchers used CNN classification on the decorated TWW types to analyse its usefulness in classifying and analysing decorated ceramic wares. The neural network model developed for the study was trained BY Pawlowicz in only a few hours using his NVIDIA GPU-enabled PC equipped with a pair of common convolutional neural network models (VGG16 and ResNet50). Pawlowicz and Downum demonstrated that when trained appropriately, a deep learning model can identify various decorative sherds’ classes with levels of accuracy equal, and at times exceeding, the levels achieved by prominent modern archaeologists. Additionally, a CNN classifier produces similar results on the same collection of sherds every time, unlike a human classifier. What’s next The authors say that further work is needed to improve overall accuracy and extend it to other ceramic wares. The project could also benefit from a larger number of typed shreds for analysis, improving the overall accuracy of the CNN model. The authors also note that collaboration with human archaeologists on the model’s results could offer better type definitions and the elimination of types that cannot be well defined. Pawlowicz’s next step is to utilize the NVIDIA-powered systems at Northern Arizona University to analyse complex models on much larger sample sets. Downum and Pawlowicz hope to build a database that can serve as a central point for image search.","excerpt":"Researchers at Northern Arizona University using GPUs to make a more precise assessment of sherds as compared to qualified archaeologists.","categories":["IT Services"],"tags":["AI (Artificial Intelligence)","Black Box","Convolutional Neural Networks","Deep Learning","deep neural network","DeepMind","NVIDIA GPU"],"author_name":"Ritika Sagar","publish_date":"2021-07-17T18:00:00","publication_year":"2021","word_count":529,"keywords":["Go","NVIDIA GPU","T5","AI","neural network","R","deep neural network","Git","computer vision","Convolutional Neural Networks","RAG","deep learning","GAN","Deep Learning","Black Box","AI (Artificial Intelligence)","DeepMind"],"extracted_tech_keywords":["AI","deep learning","neural network","computer vision","RAG","R","Go","Git","T5","GAN"],"url":"https:\/\/analyticsindiamag.com\/it-services\/gpu-accelerated-deep-learning-sorts-pottery-fragments\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":22627,"title":"Women in Analytics &#8211; Nidhi Pratapneni, SVP and Head of Knowledge services at Wells Fargo","content":"Being Women – the Balancing Act is a series of candid discussions focussing on the women of today, who beautifully juggle work and family. We spoke with some of the most prominent women from the Indian Analytics Industry this INTERNATIONAL WOMEN’S DAY. We talk about concerns like Gender Diversity, Equal Pay for Equal Work, Coming back from Maternity, Being the only women in the Board Room etc. They share their experiences from being part of the industry, about what inspired them and who supported them. In the first of this series, we speak with Nidhi Pratapneni, SVP & Head of Knowledge Services, Wells Fargo. [dropcap style=”flat” size=”2″]AIM[\/dropcap]Analytics India Magazine: As a kid, what did you want to be when you grew up? What was your childhood dream? [dropcap style=”flat” size=”2″]NP[\/dropcap]Nidhi Pratapneni: I think everyone is influenced by what their parents say and do. My father was in the Indian Police Service for many years. So I had this vision of joining the IPS, wearing a uniform. As the time to decide came closer, my father shared with me the reality of the frustrations he faced at work. And I thought, “Well, maybe we should look for something else.” AIM: Who is the one woman who inspired you? Why? NP: As you ask this question, I can think of many powerful women I could name… Indra Nooyi comes to mind. But actually, it has been my mother all along. She has been a working woman. She had a double promotion in school, went to Allahabad University and then to Roorkee. She was one of the first women studying Chemical Engineering there. She’s also someone who didn’t get into the IAS because she didn’t know where Fiji was! At least that’s what she thinks because she hated geography. After she retired she’s been devouring all these old texts by Ibn Battuta and other travelers. She has constantly been learning and that’s what really inspires me. If she’s been able to do that and is still doing it, there’s no reason for me to hold back in any way. AIM: What defines you as a leader? NP: I feel leadership has become a much maligned\/overused word. Everybody has their own style and I would say the style that defines me is one where I believe in being very open and frank with people. I want to convey to them that I’m accessible. My principles are based on my belief that I’m an enabler in making my organisation and my people successful. That summarises my mantra. When I’m not an expert in a particular area, I need to build a strong team where I can leverage others’ strengths and figure out what it is that complements mine. Creating an ecosystem that is built on our strengths is important. You have your strengths, you have your passions, go pursue them! Because in the end, the whole ecosystem will benefit. So it is not just “my way or the highway”. AIM: What are the challenges that you faced being a woman in tech? NP: It is hard to pinpoint specific challenges. I think one faces challenging situations and just goes through them. Perhaps if nothing stands out for me, there wasn’t a really huge challenge. My career has been a journey of pursuing a few goals and staying persistent. There are always setbacks.When you move back from the US to India, you think that everyone will welcome you with open arms. But that wasn’t the case. That was a setback, figuring out now that I’m here, where do I find my foothold? My classmates are doing great things. What do I do? So I think that was a challenge. Did people ask me how many people did you lead in the US? Well, in the US most mid-level managers don’t lead more than 10 people. To come back and face these questions from head-hunters was disheartening. Looking back, I don’t think that was a challenge but it was a tough period that I had to work through the normal ebbs and flows of raising the kids and building my career in parallel. I call it the GMS — the Guilty Mom Syndrome! PMS, GMS you know! It characterizes women quite well. AIM: How do your peers react to a woman leader? NP: I feel that I’ve been very lucky. I’ve worked with people who have been very helpful, very respectful. Part of it probably comes from the places I’ve been, and they acknowledge that. ‘If she has worked in XYZ places or been to certain institutions, there must be something to her’. It is a stereotype that has worked in my favor. It works against people in some situations. Having worked in MNCs all through, finding people who’ve been there to guide me and advise me, I found peer interactions quite professional and rewarding. Although if a woman from the US were to come in my place, perhaps she would feel that there are sub-texts and prejudices within the corporate ecosystem that I don’t sense because I grew up in this world. Subliminally, maybe I accept certain stances because I belong to this culture and accept some of its ways. But truly, I found a lot of acceptance. (In non-work professional environments) I do feel that women should want to project themselves through platforms and opportunities like the ones you offer. There is a dual dynamic where a) women won’t step forward and take those opportunities and b) to some extent they are not even invited as much as they should be. Maybe we should do some statistics on how many panel discussions are held in a year and how many women panelists are featured there? I’ve been to a forum organised by NASSCOM in Hyderabad which was only for women. And I asked myself why should it be a women’s only forum when what you are talking about is Big Data? What is it about big data that makes it a women-only forum? But people do feel that a lot more women participate in those events. Even though there’s nothing that is going to be discussed that is pertaining to gender. So maybe some of those are needed. Maybe there is an explicit need to also protect women in those panel discussions as presenting a different perspective, to say that there are women who have a seat at the table. So that those in the audience find that very normal and can think of themselves in those roles. AIM: Equal pay for equal work, where does the tech industry stand on this? NP: I think there are two aspects. There is a tech industry segment of predominantly MNCs, those who are more forward-thinking and where there is a definite commitment towards gender issues. And then there are many other smaller shops where there is more scope for discrimination. There isn’t much vigilance to ensure equality. In the MNC environment, there is definitely a commitment to equal pay for equal work. In reality, though, it may not always hold true and there are many nuances as to why it’s not happening. Part of it is that women will not negotiate much, and this is something that I learned late in my career. That it’s okay to have those discussions around compensation when you are being made an offer. I feel it is harder (even now) to do that negotiation while I’m in a role because once I’m in it, the system should give me what I deserve. But at the point at which you are being made an offer and you need to negotiate, many women will not do that. There is a sense that “my husband is the primary earner and whatever I’m getting is over and above that, why fight for it”? So Equal Pay for Equal Work needs to be driven by both dimensions. I think women should be more empowered to ask and organisations need to develop systems where they are tracking it, analyzing it. Whether people in the same role are getting the same or not. If they are not, ask why. It may still be okay –  one can’t push through and say equal work-equal pay when the woman might not be as good as a performer as the man. So all of this needs to be done but not at the cost of making excuses for somebody’s performance. It has to be merit-based. AIM: What are the issues women face when coming back from maternity? NP: That’s a big one. I’m almost 13 years past that time, but I think it is very relevant. It’s not something I went through here; my personal experiences are from the US. For a woman who wants to work, she has to overcome that guilty mom syndrome and how do you overcome it? You need the support of your workplace that you’ve been part of and you need support from your social ecosystem, your families, and friends. Your family should be supportive that this is something that you need to do. You don’t want the guilt to be multiplied two or three times when you decide to get back. Your husband has to be supportive. And the organisation concerned must think about what role you come back in. And ask you what are you comfortable with. There are policies that do get made, at the same time there has to be a very individual conversation on what’s right for this particular person. There are discussions one is part of where you hear that someone’s raise is scaled down because she was on maternity leave for some time. In those conversations one needs to actively educate people to say that during that period the person is an earning team member –  you need to drill that education into both male and female managers. So there are a lot of dynamics at play. It’ll be a while before we get it perfect. I think there are some women who have had a great experience coming back from maternity. And possibly in organisations it would be great to showcase those women and those learnings for others to make it an easy transition back to the workplace. AIM: Work or family — what is important in the eyes of the society today? NP: For both men and women, working or otherwise, it is work and family rather than work or family. It’s just that society becomes a little more preachy when it comes to women. People will appreciate women who achieve things, who win acclaim. When it’s your friend’s daughter who is featured in a newspaper, the father may go, “Oh look she’s in the newspaper and she’s done all this great stuff”. But when it comes to his own daughter or daughter-in-law it might be, “Well you just had a child. You need to give the child four or five years”. I don’t think people realize that they are being hypocritical. How do you educate them? And you have to because you need the support of the same people who seem to be pulling you back! Women need to realise that perhaps they can’t be perfect at both. You can’t give your 100% to both. “Do you do a 50-50 or do you do a 70-30 at one time and then switch it around to 30-70 at another time?”  Those are ways that one needs to decide for oneself. Each person has to chart their own course and what works for them. AIM: How did you balance family and work? NP: It’s been great to have my husband support me all through. He understands a lot of the challenges that people face in organisations and is able to share his own perspective, advice on how I am navigating or not standing up and saying no to certain things. It’s been a great partnership. On the family side, it’s been either my in-laws, who have been with me for quite some time or my parents who have been always been there when I was in Gurgaon. They were there to step in whenever I needed it. So leveraging the family support system has worked very well. Have a plan A, have a plan B as well and have people in your family who are there to support you whether if not physically at least encouraging you from the back lines. AIM: An instance or experience which motivated you to push the barriers NP: I’m not sure which barriers but there was an instance which was a turning point for me. This was when I was working in India after I had moved back. I was doing the work of three people. I was managing a team of 20 people, two of the other leaders had moved out so I was managing Asia and Europe and also the role of the person who was supposed to supervise both of those. So I killed myself doing it. Enjoyed it a lot and I felt that people should be taking notice that I’m achieving this great feat, addressing any fires that come up without much of an issue. And when it came to the next stage of saying okay, now what do I get for it? It was pretty much nothing! Somebody else was being brought in to lead the team and this was a professor who had not managed teams at all. So I’m like here I am who’s been running this business all through and now you want someone else to come lead the team! Where does that leave me? So kudos to the person who did come in though, because after I finished my transition I sat him down and said, “Look I’ve told you about all the team members now you need to listen to my story.” And then I shared with him my struggles and frustrations and he was extremely supportive, and he said let me figure out why things didn’t work out and get back to you. He became my partner and actually coached me and told me, “Look here are the three things that I’ve heard…” And in the end, I was the one who took the lead to say, “They want to see ABC things. Here’s what I am going to do.” Because nobody was telling me what to do! You realize that there are seniors and leaders out there and yes, while they are great in their own roles, often times they are so busy they have no clue. It’s great for you to take your destiny into your own hands and say “OK…here are the seven things I am going to do. Let me know if you agree with them or not”. Once you have the agreement and support, then you can really chart your course and say – I have done all of this, now do you have any hesitation? And so I was able to move on to next level. That was an instance where I pushed the barriers. I had conversations which were very tough. It was tough for me to hear the feedback as well and tough for me to initiate conversations where I put myself ahead of everyone else. But it did work out and it was a huge positive. That’s one advice I give to many people that I mentor or speak to. You build the things that you want to because that’s where your heart is. Better to follow your own passion than wait for instructions based on what someone else is interested in. AIM: What is your support system? NP: My family – my parents, my in-laws my husband. My children now that they are much older, are able to take care of a lot of things themselves. But I think on the professional front, a huge support system has been a lot of the people I worked with and my peers and seniors over the years. I have worked in several different companies and in each company, you find people who you respect and it’s great that one is able to reach out to them for advice later on. So for example when I was in dunnhumby, there were a lot of youngsters in my team and now they are in senior positions and I look to reaching out to them to either get advice or if there are people who worked with them, for talent and hiring and bouncing ideas off of them. In Dell, there were several others. The one person who many people in Bengaluru know is Pankaj Rai and coming across him and just understanding the way he just advises people, gets them to think of the larger scheme of things was very beneficial to me. Learnings from people carry you through and you find that as you do this you build your network. As youngsters you’re told that networking is a bad thing, this guy is networking and oh he’s doing something he shouldn’t do. But when you build it gradually as a natural course then you realise that network does play a role. For example in Dell – this manager came up to me and said I’m looking for somebody to head my social media team. Would you want to do that? I had no social media experience. If I applied for such a job, nobody would give it to me. Here is someone who is willing to take me on because of what they’ve seen me do and so I said let me just jump at it, otherwise it’s not an opportunity that’s going to come again. So that was a result of that network and support system that I built. AIM: What do corporates lack while implementing gender diversity? NP: I think many corporates have started moving ahead beyond the numbers. Some might still be in a phase where they are merely looking at a number of women in organisation. So you should be achieving those numbers but with an insistence that you are not compromising on quality. Then when you go beyond that, it is how inclusive are you making your environment for the diverse workforce. When you are driving gender diversity programmes, who is taking the lead on those and who is part of that conversation? If men are not a part of that conversation, then we are really doing a great disservice and we are not moving that conversation as fast as it could progress. I think it’s the responsibility of the women as well to also educate their peers. If it is a senior woman leader who is talking about her experiences and her challenges, if the male manager isn’t there, who will benefit from it? AIM: Saree or suit? What empowers you more in the boardroom? NP: I’ve started wearing sarees to the office very recently, only in the last 3-4 years. I am comfortable in them, but if I need to dress for an event I think wearing a saree would distract from the conversation if there are people who are not used to seeing me in a saree. I would want my message to go across rather than have people think, “hmm why is she wearing a saree today?”. I’ve been comfortable in suits. They will not raise that many eyebrows but for a woman who’s used to wearing sarees all the time, if she suddenly shows up in a suit, it will be a different thing. I think whatever you wear it’s the message that is important. Yes, it is important to put that effort into looking good and – whatever it is that you’re wearing – to feel confident that you are looking good in that boardroom. AIM: Your advice to other women in tech? NP: I would say ‘women in tech’ is a phrase that shouldn’t put you off. Don’t feel inhibited by it because tech is increasingly becoming a space that is valuing the arts as well. It’s great that we have reached this stage so I would say this message is not for women in tech, it’s for women in any discipline which will find a role in analytics. Business decisions impact the factors that play into how a consumer thinks, how a consumer behaves. Many of these are aspects of sociology and behavioural sciences. The second thing I’d like to say is that learning is constant. My generation grew up at a time when studying ended one chapter and working opened a new one. The lines were clearly drawn and one didn’t go in reverse. Today, we must constantly learn and keep ourselves up-to-date on what’s happening in terms of cutting-edge trends and technology. So I would like to encourage women to have a mindset that learning is constant. There was a recent statistic shared by Coursera, that of the people who take their courses, the percentage of women taking them is much lower than the percentage of men. That is probably because when women move into a phase where they have to juggle work, family and everything else in between, the ability to take time out to invest in yourself, to invest in learning is harder. Women tend to hesitate and wonder how they can make it happen and feel stuck. I would suggest that they figure out how to do that learning within their organisations in a more structured way. When it’s part of your work and is required of you, perhaps you will be forced to do it. Figure out what makes it happen for you because learning and up-skilling will be paramount in the age we are in. AIM: Your advice to corporates on how to retain women employees NP: A lot of attrition takes place when women need to focus on their family. Either when their children are very young or during the phase where they are in school, where they need more help. It’s important to encourage those one-on-one discussions with women. I’m not sure that it is happening. There should be an intent to identify those who are in the fragile phase and have specific conversations that say “This is why we are having a conversation. What is it that will make you more comfortable in terms of being able to balance work and other priorities that you may have and can we structure something for you?” It is harder to do in start-ups; operational economics become more important. But in many other companies, there is a greater willingness to do it. How you make that connection to realize it for your women employees is a bit of a question mark and I would encourage those one-on-one conversations and customizing work arrangements for women.","excerpt":"Being Women – the Balancing Act is a series of candid discussions focussing on the women of today, who beautifully juggle work and family. We spoke with some of the most prominent women from the Indian Analytics Industry this INTERNATIONAL WOMEN’S DAY. We talk about concerns like Gender Diversity, Equal Pay for Equal Work, Coming […]","categories":["AI Features"],"tags":["Interviews and Discussions"],"author_name":"Дарья","publish_date":"2018-03-14T11:15:12","publication_year":"2018","word_count":3781,"keywords":["big data","Go","ELT","AWS","AI","RAG","Aim","analytics","Rust","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","analytics","Aim","RAG","AWS","R","Go","Rust","big data","ELT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/open-dialogue-thats-what-women-empowerment-is-all-about-says-nidhi-pratapneni-of-wells-fargo\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10064945,"title":"MongoDB Atlas now available as a pay-as-you-go offering on Google Cloud","content":"MongoDB, the general-purpose database platform, has announced a pay-as-you-go MongoDB Atlas offering, that can be launched directly from the Google Console. The offering provides developers with a simplified subscription experience and enterprises more choice in how they procure MongoDB on Google Cloud. With this new pay-as-you-go MongoDB Atlas offering, customers only pay for the resources they use and can scale based on their needs, with no up-front commitments while using their Google accounts. This offering will also make it easier for customers to build, scale, and manage data-rich applications with MongoDB Atlas within the Google Cloud Console. “MongoDB and Google Cloud have a long-standing partnership, focused on driving customer success across multiple industries, such as financial services and retail, at a global scale. Developers benefit by having MongoDB Atlas now featured in the database section of Google Cloud’s console, increasing the visibility and velocity of Atlas with a wider group of users,” said Alan Chhabra, executive vice president of Worldwide Partners, MongoDB. “Together, Google Cloud and MongoDB will continue to work with organisations looking to streamline cloud migrations and modernise application development in order to provide their customers with a first-class experience.” The new offering is available through Google Marketplace, giving companies and resellers of all sizes more flexibility in selecting the technology stack and partners that best fit their needs. “Data is fundamental to digital transformation,” said Kevin Ichhpurani, Corporate Vice President, Global Ecosystem at Google Cloud. “Expanding the availability and simplifying the deployment of MongoDB Atlas on Google Cloud will make it even easier for customers to benefit from MongoDB’s capabilities alongside those of Google’s data cloud, providing greater capability and opportunity to utilise their data as they digitally transform.”","excerpt":"Google Cloud and MongoDB will continue to work with organisations looking to streamline cloud migrations.","categories":["AI News"],"tags":["Google Cloud","MongoDB","mongodb atlas"],"author_name":"Kartik Wali","publish_date":"2022-04-13T17:40:57","publication_year":"2022","word_count":282,"keywords":["Go","Google Cloud","mongodb atlas","AI","programming_languages:R","MongoDB","ML","digital transformation","Git","cloud_platforms:Google Cloud","GAN","R"],"extracted_tech_keywords":["AI","ML","MongoDB","R","Go","Git","GAN","digital transformation","cloud_platforms:Google Cloud","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/mongodb-atlas-now-available-as-a-pay-as-you-go-offering-on-google-cloud\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":31510,"title":"Artificial Intelligence Is Now Helping Get Rid Of Bad Hair Days","content":"At a first glance, it seems hair care cannot be ascribed to the world of AI.  Nonetheless, AI has wended its way in businesses across the spectrum — including beauty and grooming. The companies who have successfully employed AI to make the best hair care solutions available to their customers. With the help of machines and complex algorithms, companies are now able to make the best product for its customers which otherwise would have taken years of research. A huge amount of data is dealt with in understanding the ingredients of a hair solution and its effect on different types of hair. An AI system, when trained on how to use each ingredient in a hair solution for a specific type of hair is further exposed to different environmental factors can predict the right ingredient and the right amount that needs to go into the product for a specific user. In this article, we list down how companies are using an AI-based solution to create an ideal hair product for their customers and fix bad hair days. Prose Prose, a Brooklyn-based startup co-founded by a former L’Oréal VP, has combined AI and hair care with great success. The startup uses AI to produce customised hair solutions for its customers. With “Great hair starts with great tech” as a motto, Prose connects the customers to experts via an intelligent platform, who then help create ideal solutions. The startup considers aspects such as hair problems, lifestyle, environment and even the needs and goals of the customers. They also take into consideration each customer’s personal preferences like the fragrance of the product, the quantity of silicon and gluten in the product, vegetarian product or not. Based on these numerous and specific parameters, Prose feeds the data into an intelligent system and formulates a personalised hair product like a shampoo, conditioner or a pre-shampoo mask. Reportedly, Prose has shipped 110,000 bottles so far and is expected to reach $1 million in monthly sales by January 2019. Function Of Beauty Another company that uses AI technology to provide customised hair solutions to its customers is the four-year-old start-up Function Of Beauty. This company has been around for a while now and uses its own set of algorithms to understand its customers. Founded by MIT alumni, the company is focused on using technology to provide the best solutions to its customers. The company uses machine learning model to analyse hair type, hair structure, hair goals and other preferences of the customers to come up with best ingredient combinations that would result in a customized hair solution. Schwarzkopf While startups are using technology to make the life of people easier, more established companies are also following the same idea. German cosmetics giant Schwarzkopf, a division of Henkel AG & Co has also entered the market with products to please their customers. Reportedly, the company is also conducting trials on custom products in Japan. The system includes a device with infrared sensors that can detect the hair’s molecular structure and analyse the moisture content and an in-salon machine that formulates and produces the product in just 50 seconds. The product also gets a barcode that helps the customer to reorder the exact same product again. Outlook The impact of technology in a business and on the life of people can be at different levels. It may not always be successful but it can always increase the chances of being successful. Data holds the power to change business and make the lives of people better. Using the right data in the right amount is the best ingredient in making the best solution. Artificial intelligence can exploit data on a large scale. AI has been helping to extract meaningful insights from huge amounts of data which later serves as key ingredients in decision making.","excerpt":"At a first glance, it seems hair care cannot be ascribed to the world of AI.  Nonetheless, AI has wended its way in businesses across the spectrum — including beauty and grooming. The companies who have successfully employed AI to make the best hair care solutions available to their customers. With the help of machines […]","categories":["AI Features"],"tags":[],"author_name":"Amal Nair","publish_date":"2018-12-14T06:35:17","publication_year":"2018","word_count":634,"keywords":["Go","machine learning","artificial intelligence","programming_languages:R","AI","programming_languages:Go","RAG","R","startup"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","RAG","R","Go","startup","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/artificial-intelligence-helping-get-rid-of-bad-hair-days\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10021248,"title":"All India Radio To Start New Series On Artificial Intelligence","content":"All India Radio (AIR), in association with Vigyan Prasar, an autonomous organisation under the Department of Science and Technology and Tamil Nadu Science Forum (TNSF), announced the starting of a new radio science series on artificial intelligence (AI). Produced by Madurai AIR, the new science series titled — ‘Ini yavum nunnarive’ (Future is Artificial intelligence) will be broadcasted on AIR every Saturday at 8 PM from March 6th, 2021. With a focus on AI science and research in India and across the world, this series will talk about how artificial intelligence is used in literature, industrial and social sectors, as well as discuss the futuristic society and the apprehensions around it. According to the news media, the series will be produced in 19 different Indian languages. They will be broadcasted on more than 121 AIR stations to cover 85% of the country’s geographical area. To facilitate this, TNSF has identified scriptwriters from Madurai Kamaraj University along with other colleges and schools and has also conducted a workshop on scripting for them. The series will include episodes that will cover expert opinions, interviews, docudrama and discussion with listeners. It will also have a Q&A session at the end of the broadcast to have an interactive experience for the listeners. Vigyan Prasar has also decided to reward the winners with the right answers to those questions.","excerpt":"All India Radio (AIR), in association with Vigyan Prasar, an autonomous organisation under the Department of Science and Technology and Tamil Nadu Science Forum (TNSF), announced the starting of a new radio science series on artificial intelligence (AI).  Produced by Madurai AIR, the new science series titled — ‘Ini yavum nunnarive’ (Future is Artificial intelligence) […]","categories":["AI News"],"tags":["AI Language"],"author_name":"Sejuti Das","publish_date":"2021-03-03T10:37:55","publication_year":"2021","word_count":224,"keywords":["artificial intelligence","programming_languages:R","AI","AI Language","GAN","R"],"extracted_tech_keywords":["AI","artificial intelligence","R","GAN","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/all-india-radio-to-start-new-series-on-artificial-intelligence\/","complexity_score":2,"technical_depth":5,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10172677,"title":"Esri India Launches GeoAI Competency Centre, Invests ₹150 Crore","content":"Esri India Technologies announced the opening of a new geographic information system (GIS) and AI competency centre in Noida on July 1. The company will invest over ₹150 crore in this initiative over the next five years. The centre aims to accelerate the use of AI in GIS applications across various sectors in India. The new facility will expand Esri India’s workforce with AI specialists, data scientists, GIS experts, and domain professionals. Their focus will be on developing AI-powered geospatial solutions that help customers process data, extract insights, and automate decision-making. Agendra Kumar, managing director, Esri India, said, “The speed of change is so high that without strategic investments, it will not be possible to lead the transformative use of AI in GIS applications.” Focus on Local Talent The engineers at the new centre will integrate AI into spatial analysis to support faster, data-driven decisions. This includes automating routine GIS tasks, identifying patterns, and predicting trends in real-time. Esri India also plans to work with academic and research institutions to build local GeoAI talent. The company will release AI-ready datasets trained with Indian data and strengthen its capability to meet the rising demand for AI in GIS. Esri India has already supported AI integration in areas such as land, asset and disaster management, object identification from imagery, and video analytics for Indian users. “We have close to 6,500 organisational customers in the country. In terms of users, maybe 1.2 million people use some app running on our technology every day,” Kumar told AIM in an exclusive interview recently. Esri’s scope is quite vast, ranging from national mapping agencies to utilities such as electric distribution companies, city gas firms, and telecom giants like Jio and Airtel. GIS Applications Citing examples from traffic analysis to identifying defence vehicles or aircraft, Esri India delves into many other aspects such as defence and security, government, utilities, urban management, 3D mapping, digital twins, etc. Kumar explained to AIM, “Human identification can be done, for example, in our office, we counted the number of people passing near the reception [using this]. Even face recognition can be done if it is required.” The company’s Indo ArcGIS is a comprehensive GIS software platform that enables organisations to create, manage, and analyse spatial data. It supports mapping, data integration, and location-based intelligence for diverse sectors.","excerpt":"This Noida facility will expand Esri’s workforce with AI specialists, data scientists, GIS experts, and domain professionals.","categories":["AI News"],"tags":["esri india","Geospatial analysis India","geospatial mapping","geospatial technology"],"author_name":"Sanjana Gupta","publish_date":"2025-07-01T17:41:09","publication_year":"2025","word_count":386,"keywords":["Go","geospatial mapping","programming_languages:R","AI","data-driven","Geospatial analysis India","Git","programming_languages:Go","Aim","analytics","GAN","R","esri india","geospatial technology"],"extracted_tech_keywords":["AI","analytics","Aim","R","Go","Git","GAN","data-driven","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/esri-india-launches-geoai-competency-centre-invests-%e2%82%b9150-crore\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10075268,"title":"Zoom revamps as ‘Zoom Team Chat’, adds new features","content":"On Monday, Zoom unveiled the new ‘Zoom Team Chat’ (previously named Zoom Chat) with additional features to accommodate the post-pandemic hybrid work setup. With the launch of the new Team Chat, the company aims to enable smoother collaboration among professionals and empower teams to communicate quickly and securely. Zoom Team Chat is expected to offer a powerful hub with chats and channels to connect workstreams, keep projects on track, and stay connected with the whole team and key stakeholders. Team Chat is expected to make it effortless to collaborate by bringing together messaging, file sharing, third-party apps, video, voice, and whiteboard in a single app. Users can now collaborate on a whiteboard, elevate a chat to a meeting or a phone call, or even share a recorded video message—all within the Zoom platform. The Team Chat will not only be used for communicating with employees within the organisation, but would also provide seamless and secure communication experience for external contacts, including vendors, consultants, and clients. As part of the brand enhancement, Zoom is updating its brand colours, typography, and product icons—kicking off a global campaign by introducing a new visual element that represents Zoom’s new unified platform. Zoom’s well-known logo will remain the same, but has unveiled a visual at the centre of its campaign to demonstrate itself as a space that connects all work streams. Additionally, features such as sharing in-meeting chat to Team Chat, responding via chat to an upcoming meeting, third-party integrations, and flagging a message using a ‘reminder’ feature have also been introduced by the platform.","excerpt":"Users can now collaborate on a whiteboard, elevate a chat to a meeting or a phone call, or even share a recorded video message – all within the Zoom platform.","categories":["AI News"],"tags":[],"author_name":"Bhuvana Kamath","publish_date":"2022-09-14T20:22:43","publication_year":"2022","word_count":260,"keywords":["Go","programming_languages:R","AI","ML","programming_languages:Go","Aim","GAN","R"],"extracted_tech_keywords":["AI","ML","Aim","R","Go","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/zoom-revamps-as-zoom-team-chat-adds-new-features\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":35766,"title":"7 Tips To Make Your Data Science GitHub Portfolio Perfect","content":"While a resume is an important component to showcase your abilities to potential employers, if you belong to the data science community, you might want to showcase your abilities in coding and other software capabilities in a way that does justice to your skills. A crucial part of data science jobs is to be able to code, and GitHub serves as a perfect platform to access the coding skills and display hands-on ability to solve problems. There a few ingredients that make up for a good data science portfolio, some of them are a few medium-sized data science projects, showcasing the problem-solving abilities, which can be done by highlighting practical problems on GitHub. As the importance of having a good GitHub profile has been stressed out so many times, there are often questions on what potential employers are looking for and what a good GitHub account generally looks like. This article covers the important aspects of a good GitHub profile. Why Is GitHub Crucial? Since many years, developers and coders have been accessed for their coding abilities based on a Q&A session and their ability to write or edit a short program in an interview setting. This may not always do justice to the candidate’s programming abilities. But with the rise of open source movement and the popularity of repository sites where programmers can show a portfolio of codes to potential employers, it has taken a key role in accessing candidates. GitHub is the most popular of it all. While having a GitHub is not a deciding factor to land up a good developers job, a genuine GitHub portfolio lets potential employers have a quick glance into a candidate’s abilities. It showcases a candidate’s areas of interest and their contribution in the open source community. It shows what their code looks like, and how do they approach a problem. It shows how one candidate can be more engaged in the development and in open source than the other candidates. 7 Things To Have For A Good GitHub Portfolio? Including good author info: This includes details such as candidate’s username, location, email address, current employer, etc. As this the first thing that potential employer notices, it is a good idea to have other relevant links such as personal blogs, websites, projects and other links that candidates can showcase. Large followers: The number of followers that your portfolio has is a good indication of the work that you have done in the past. It showcases how fellow GitHub users engage with your work. Having large followers comes with the significant amount of work a candidate might have done on GitHub. More than 50 is usually a decent number. Contribution graph: This indicates the number of contributions that a developer has made on other projects on the site. It showcases your keenness to explore other areas and shows activity levels in the coding community. The greener the graph is, the better is your contribution rate. Therefore it is always advisable to keep exploring. Improving on stars: These are user ratings and may not be considered the only indicator for a good GitHub profile. Each project on GitHub user can earn stars from fellow users. It is a way of identifying how you have engaged in the community. 100 stars are usually considered decent but larger the better. Forking and creating repositories: Repositories contains open-source development projects that the developer is hosting on GitHub and forking is the number of times projects have been copied. More the number of people who have forked, the greater is the popularity of that developer’s project in the GitHub community. A large amount of activity indicates that the developer is working on a popular project. Writing employer-targeted code: Writing a code related directly to the employer’s business is a good way to catch their attention. It can showcase your coding abilities while demonstrating the interest you have in getting that job. The projects can be as simple as creating a visualisation of its data sets. Other projects such as gaming or mobile apps: You may include projects around game development, that usually requires graphics-intensive artistic offering and showcase your understanding of key programming concepts quite well. The same goes for developing mobile apps if that is your area of interest for working. Having mentioned the pointers above, the key idea is to create stronger projects for a strong GitHub profile. The recruiters pay attention in the areas such as the variety of projects that you have brought, completeness of the project, its functionality, readability and most importantly the information that it stores and displays.","excerpt":"While a resume is an important component to showcase your abilities to potential employers, if you belong to the data science community, you might want to showcase your abilities in coding and other software capabilities in a way that does justice to your skills. A crucial part of data science jobs is to be able […]","categories":["AI Trends"],"tags":["github data science"],"author_name":"Srishti Deoras","publish_date":"2019-03-05T11:38:28","publication_year":"2019","word_count":766,"keywords":["Go","data science","programming_languages:R","AI","programming_languages:Go","Git","ViT","GitHub","R","github data science"],"extracted_tech_keywords":["AI","data science","R","Go","Git","GitHub","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/7-tips-to-make-your-data-science-github-portfolio-perfect\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":8746,"title":"Book Review: The Steradian Trail: Book #0 of the Infinity Cycle","content":"‘The Steradian Trail’ by M.N Krishnan is a murder mystery with analytics as its spine with algorithms and economics to back it up further. A fascinating read for analytics, economists and computer enthusiasts. It could indeed be the first “analytics novel” of India. The plot involves the lives of Joshua, Lakshman and Divya, and how three very different lives come together in India to solve the mystery of one dead person in the USA. Plot Jeffrey Williams, a criminal genius who also happens to be the student of a world renowned professor and computer scientist is murdered in USA with professor Joshua’s name, and a garbled number on his lips. However, unknown to the murder, Joshua is in India giving a seminar. The turn of events lead him in an attempt to unveil the mysterious death of his student. Armed with only a garbled number as the clue left by his dead student, Professor Joshua Ezkiel, takes us through a gripping tale tagging his friend Professor Lakshman along, in a journey involving more murders, ancient Indian cities, algorithms and ancient mathematicians; only to discover the sinister secret behind the murder in a place they could never have imagined! Further, what is even more interesting is that they are assisted by Lakshman’s genius student Divya, who remains unknown to the murder and yet, manages to crack many key clues. Review The narration style of the author is simple and entertaining. The book weaves mystery, mathematics, computer algorithms, technology, economics, religion and history into a well written novel while giving rich insights into the Tamilian culture to the readers. It definitely is a very interesting read for any analytics, maths or computer enthusiasts. Beginning the book with the murder is the ultimate catch to generate curiosity and giving the readers a reason to keep turning the pages until the very end. Even though, a nail-biting mystery, sometimes it tends to get dragged with mathematical technicalities, especially for those who are not much into mathematics or algorithms. However, the plot is highly refreshing as the seed idea is original and beautifully put together in a fast paced book. Overall, a commendable job by the author, M.N Krishnan. His knowledge and extensive research are quite visible as the integral part of this book. Being the first in series, it leaves us wanting more from the consecutive parts. Product Details Paperback: 362 pages Publisher: Westland; 2013 edition Language: English ISBN-10: 9383260734 ISBN-13: 978-9383260737 Product Dimensions: 19.8 x 13.2 x 2.6 cm","excerpt":"‘The Steradian Trail’ by M.N Krishnan is a murder mystery with analytics as its spine with algorithms and economics to back it up further. A fascinating read for analytics, economists and computer enthusiasts. It could indeed be the first “analytics novel” of India. The plot involves the lives of Joshua, Lakshman and Divya, and how […]","categories":[],"tags":[],"author_name":"Apoorva Verma","publish_date":"2016-01-27T11:21:54","publication_year":"2016","word_count":416,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","RAG","analytics","R"],"extracted_tech_keywords":["AI","analytics","RAG","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/book-review-the-steradian-trail-book-0-of-the-infinity-cycle\/","complexity_score":4,"technical_depth":7,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10070165,"title":"Browser extensions that made it big","content":"Though Grammarly was founded in 2009, it took six years for the writing assistant to really take off. The tipping point came in 2015 after it was made available as a browser extension. Not just Grammarly, several other apps and tools have started off as browser extensions and evolved into big companies. Browser extensions are software modules built on web technologies like HTML, CSS, and JavaScript that customise the web browsing experience. Such plugins are used to manage cookies, block pesky ads, encrypt emails, store passwords, check spelling, stylise web pages, organise notes etc. Greg Isenberg, CEO, Late Checkout (a product studio, fund and agency that designs, creates and acquires web3 and community-based technology businesses) and advisor at Reddit, explained why some Google Chrome extensions generated huge values. – People barely uninstall extensions – Barrier to install is low – Extensions are “always on” – People pay for many of them He cited stats to show how launching a Chrome extension gives it 20 times more potential users than an iOS app. In the Chrome store, 10,000 users use an extension on an average compared to 667 users using 1 iOS app in the App Store and 500 users using 1 Android app in Play Store. Here’s a list of Chrome extensions that metamorphosed into big companies. Grammarly Grammarly is a proprietary software used to review spelling, grammar, punctuation, clarity, engagement, and delivery mistakes. Founded by Max Lytvyn, Alex Shevchenko, and Dmytro LiderMax Lytvyn, the company’s initial focus was on grammatical error correction to help with writing. Their initial customers were mostly students and universities. However, the company’s growth was slow. In 2014, Grammarly launched MS Office plugins, allowing users to access Grammarly’s grammar checker and plagiarism detection technology directly in Word and Outlook. In 2015, it launched its free browser extensions for Google Chrome and shifted to a freemium model. Later on, Grammarly released browser extensions for Safari, Firefox, and Edge as well. The browser extensions accelerated Grammarly’s growth; from 1 million daily active users in 2015 to 30 million in 2020. Today, Grammarly has a post-money valuation of around USD 10 billion, and has raised a total of USD 400 million over 3 rounds. Grammarly serves 30,000 professional teams across a variety of industries.  The proprietary software is available in a variety of forms- online text editor, browser extension, desktop application and mobile app. Marketers, sales professionals, engineers, support specialists, entrepreneurs, and students use Grammarly to communicate more effectively every day. The Grammarly keyboard is also quite popular among Android and iPhone users. Loom With funding of USD 203 million and a valuation of more than USD 1.53 billion, Loom is the leading video communication platform for async work. Using Loom users can record their screen, share their thoughts visually, and provide asynchronous feedback. Loom can help with recording product demos, giving feedback, or simply sharing thoughts. In 2015, Loom (then called Opentest) started its life as a network of experts giving feedback on companies’ products. However, the founders– Joe Thomas, Vinay Hiremath, and Shahid Khan– realised the business model was not working. In a span of seven months, the founders pivoted 4- 5 times. In one of the pivots, they released a browser extension called OpenVid that helped people record feedback videos, and the rest was history. The team later launched OpenVid on Product Hunt. Following this, the user base grew to 10,000 plus in three months and attracted a lot of investor interest, including from venture capital firms like Point Nine. Today, Loom has 14 million users across 200K companies including HubSpot, Atlassian and Netflix. Honey Honey is a popular US-based online coupons aggregator. The company was built as a browser extension by entrepreneurs Ryan Hudson and George Ruan in 2012. The browser extension notifies users of price drops and price history of select items sold by participating online stores. In 2020, Paypal acquired Honey for about USD 4 billion. Currently, Honey has a user base of a staggering 17 million. Users get access to coupon codes across 30,000 plus sites saving an average of USD 126 per annum. MetaMask MetaMask is a crypto wallet that allows users to manage their digital assets like private keys, local client wallet and hardware wallets and provides a simple and secure way to connect to blockchain based applications. In 2016, ConsenSys Software Inc developed MetaMask as a browser extension for Chrome that enabled users to access Ethereum-enabled distributed applications. Today, with over 21 million users, MetaMask has evolved to become a global community of developers and designers dedicated to democratising access to the decentralised web with blockchain technology.","excerpt":"Browser extensions are software modules built on web technologies like HTML, CSS, and JavaScript.","categories":["IT Services"],"tags":[],"author_name":"Zinnia Banerjee","publish_date":"2022-06-30T14:00:00","publication_year":"2022","word_count":769,"keywords":["Go","API","AI","ML","Git","RAG","JavaScript","GAN","R","Java"],"extracted_tech_keywords":["AI","ML","RAG","R","JavaScript","Go","Java","Git","API","GAN"],"url":"https:\/\/analyticsindiamag.com\/it-services\/browser-extensions-that-made-it-big\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10169886,"title":"Why Claude is Losing Users","content":"Anthropic’s Claude has faced growing criticism in recent months. Users on the ClaudeAI subreddit have voiced increasing frustration. Both long-time Pro subscribers and new Max users report problems with the service, not due to cost, but because issues persist even after paying for it. While Google’s Gemini may be gaining an edge over Claude, it may not be the sole reason why Claude users want to cancel and leave the service. The Paid Plans Do Not Seem to Cut It Users found themselves restricted on the Pro plan. “I just cancelled my Pro plan; this is a joke with those limits now,” mentioned a Reddit user. The person added that free tools like Google’s Gemini Studio outperformed Claude in coding tasks. For someone who once paid monthly, the user decided to cancel the subscription. Another Reddit user, who cancelled Claude’s subscription, told AIM that although he was paying $20 a month, there were issues with Claude. “I would open a chat, and I would want to code on a side project. Gemini 2.5 Pro or ChatGPT allowed me to ask for numerous little iterations in the code and other complex stuff for hours on end without getting any context window limitations.” The user found Claude’s tokens per minute were limited and often showed that they had exhausted them. This prompted users to start a new chat and add context to each new query. Consequently, the user felt they spent more time explaining their needs than being productive. Yet another user, who upgraded to the Max plan with high hopes, hit a familiar wall. “I would literally go back to paying for the API if I didn’t have to clear the context window every time I acted on a task,” the user wrote. And then there was the team use case. The Reddit user, being a part of a six-seat Claude Team with a prepaid annual subscription, reported that even small files were crashing sessions. “We’re getting slammed with ‘Claude hit the maximum length for this conversation’ notification after the MCP reads just one or two tiny files.” “Even basic prompts are getting cut off after a couple of exchanges. It’s honestly unusable right now.” Developers Extensively Using Claude Confirm Hitting Rate Limits It is not just Reddit where users may seem to hate Claude’s new limit. Developers known to use Claude extensively for their projects also posted reviews on various social platforms. Eduard Ruzga, a staff engineer at software firm Prezi, who ditched using Cursor and Windsurf for Claude with MCPs, resonated with the sentiment observed in the Reddit community. “Before Claude Code Max subscription, one could use Claude with Desktop Commander [his MCP tool] for hours without hitting limits. After release, there were reports of people hitting limits in a few messages or at least under an hour and being forced to wait for 2–3 hours before limits reset to hit them again very fast,” he told AIM. Besides, Ruzga also mentioned that he noticed users complaining about the limits in Anthropic’s Discord server. The users were also not happy with the way the changes to rate limits were communicated. Many recall not receiving alerts about the new session limits, throttling, or pricing structures. A Reddit user noticed an updated support page detailing the Claude Max Plan’s 5-hour session structure, potentially a usage gate that locks out users even on the top tier. “Terrible usage limits again on Claude today,” the user wrote, linking to the new rules. The user added, “It seems they really want to push everyone to the Max plan, which costs $300 in Australia.” Will Anthropic Do Something to Keep The Users? Amid the negative comments, users are still rooting for Claude to recover. Many point out that when it’s functional, it remains unparalleled in analytical reasoning and writing structure. Some even want to switch to Claude from ChatGPT. Those who integrate Claude Code with GitHub workflows report better stability, as long as they download artifacts immediately or rely on the command-line interface. Despite the surge in complaints, there has been no official acknowledgement of some of the platform’s most serious regressions, including downgraded Pro-tier or Max-tier performance. Claude may not have lost its edge, but it is losing users.","excerpt":"“There were reports of people hitting limits in a few messages or at least under an hour and being forced to wait for 2–3 hours before limits reset to hit them again very fast.”","categories":["AI Features"],"tags":["Claude"],"author_name":"Ankush Das","publish_date":"2025-05-13T18:00:00","publication_year":"2025","word_count":705,"keywords":["Anthropic","ChatGPT","Go","AI","Claude","Git","Aim","Rust","GitHub","R","Gemini 2.5"],"extracted_tech_keywords":["AI","ChatGPT","Anthropic","Gemini 2.5","Aim","R","Go","Rust","Git","GitHub"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/why-claude-is-losing-users\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10061219,"title":"Data science hiring process at Sabre","content":"Sabre is one of the leading software and technology firm that serves a wide range of travel companies, including airlines, travel agencies, hoteliers, and other suppliers. The company offers retail distribution and fulfilment solutions that help its customers operate more efficiently and drive revenue, alongside personalised traveller experiences. Sabre’s technology platform currently manages more than $268 billion worth of global travel spending annually. Based in Texas, Sabre serves customers in more than 160 countries worldwide. The company has set up its global development centres (GCC) in Bengaluru. It is responsible for developing industry-leading solutions for the travel business. The team practises lean startup thinking, and the GCC in Bengaluru has been consistently featured in the ‘Great Place To Work’ rankings. At Sabre, data science is an integral part of product algorithms and the consulting business. Data scientists solve business problems where there are no readily available solutions. They help the company deliver better value to their customers, improve product and service offerings, identify key market trends, generate consumer insights, and reduce time-to-market. Besides this, data science also helps them harness the power of data to derive insights that define strategic organisational decisions. “Our data science teams focus on creating new algorithms or enhancing existing ones, which increase product value, solve new business scenarios, and provide deep insights with recommendations,” said Bala Srinivasan, director of operations research at Sabre Bengaluru GCC. Some of the key areas that Sabre data science teams get involved with include evaluating revenue performance, recommending actions using opportunity algorithms, evaluating opportunities for increasing airlines share in the marketplace, improving profitability, analysing and comparing strategic pricing decisions, and recommending price optimisation. A few other aspects of the business that this team supports include: Enabling customer segmentation, pricing elasticity, evaluating airline shopping data\/engine, developing price testing framework, etc.Building learning models in the areas of ticket and ancillary sales, understanding customer willingness to pay, and recommending actions for airlines to improve salesLearning and applying concepts of offer\/order management and continuous dynamic pricing to significantly expand opportunities for engaging with travellers through multiple touchpoints and creating personalised offers accounting for individual preferences and market context Enabling choice ranking, live experimentation of prices\/retail choices, and deep behavioural insights of travellers using AI\/ML Team structure The Sabre data science team consists of 50+ data scientists who are focused on the airline solutions side of their business. Besides this, the development teams also look to use data science within their product context when the scope is adequately managed. “When the scope is larger and involves core data science areas like prediction, recommendation, forecasting, optimisation, personalisation, and classification, etc., our team of data scientists steps in. This approach also allows development teams to understand the importance of data science applications within the purview of their project\/product,” added Srinivasan. Sabre’s data science team is segregated into the following groups – Sabre Labs (product roadmap and new features), Professional Services (custom development, PoCs\/MVP) and product engineering teams (based on project needs or product features). The team told AIM that they are currently expanding their team and actively hiring data scientists across various experiential levels worldwide. Interview process At Sabre, the company believes that every role is critical, and the hiring process remains the same across the organisation. Srinivasan said that their data science team looks for candidates who have in-depth knowledge and considerable experience working with statistical models and machine learning algorithms. Also, they evaluate their analytical acumen and try to gauge their passion for discovering insights hidden in large data sets. Here are the three steps that they follow during the interview process: Step 1: Pre-interview screening to check fitment for the role Step 2: Technical interview with analysts of candidate’s problem-solving skills Step 3: Interview with data science leadership to assess skills and culture match Team Sabre said that a talent acquisition partner would guide the candidate through the process for a seamless experience. Do’s and don’ts Based on their experience, some of the pitfalls that candidates can avoid during the interview process at Sabre include: Be honest about the skill sets that you possess because the organisation is relying on you to deliver on those promises.Having no understanding of the business impact your work has helped create in your previous roles is a cardinal sin. Make sure you go back and relook at the work that you have done.The interviewers want to hear about your past, but they also want to hear what you could bring to the table if you’re hired.Your responses cannot focus only on technical aspects or on real-life applications; ensure you weave both together.Organisations are looking for people who are aspiring to go above and beyond. Hence, it’s important to showcase your problem-solving skills through your responses. Work culture “Since our inception, Sabre Bengaluru GCC has been at the forefront of technology transformation and believes in investing in developing employees who can think for the customer,” said Srinivasan. Internal innovation platforms such as BlitzCode (annual hackathon), Solve!t (engineering best practices sharing forum), Big Pitch (business idea pitching), Skill+ (gamified learning), Winglets (engineering community engagement) and Women in Technology (L&D for women technologists), have allowed Sabre team members (employees) to think differently, keeping the travel industry’s issues in mind. Over the years, the centre has set the bar high among its industry peers by creating platforms that allow employees to thrive in an innovation-fueled environment. The team said that a strategic approach is taken to harness the collective capabilities and creative mindsets of interns, middle management, and senior leaders. Using interactive and engaging initiatives and incentives to drive creative thinking, the centre has grown to become a valued partner across the organisation. In addition to this, their data science team has the opportunity to work on state-of-the-art technology that is powered by decision science. Expectations “We believe that each new member in the team adds tremendous value, in some form, by their experience. Similarly, the data scientists hired as part of our team bring several new perspectives which must be explored,” said Srinivasan. Here are some of the role-based expectations from candidates: Work with subject matter experts and professionals from airlines to identify opportunities for using data to deliver insights and actionable predictions of customer behaviour and operations performanceAssess the accuracy and effectiveness of new data sources, data gathering and forecasting techniques Develop custom data algorithms and models to apply to datasets and run proof of concepts studies Design market sampling framework and incorporate obtained feedback into predictive and prescriptive methodologies Collaborate market sampling framework and incorporate obtained feedback into predictive and prescriptive methodologies Collaborate with software engineers to test and implement production-quality code for forecasting and data analytics modelsDevelop tools and processes to monitor and analyse data accuracy and models’ performance Some of the broad KPIs that they use internationally to measure the performance of their data science teams include – agility, accuracy, reusability, codification. Why should you join Sabre? Sabre offers one of the largest global distribution systems in the world, working with thousands of airlines and travel agents across 160 countries. They also power mobile apps, airport check-in kiosks, online travel sites, airlines and hotel reservation networks, aircraft and crew scheduling systems, travel agent terminals, etc. Every day, millions of employees and customers interact with their technology, enabling business travellers to close the deal, delivering cargo shipments to factories and stores, and giving excited travellers the family vacations and honeymoons of a lifetime – all while making the experience\/journey easier, faster, safer and more personal! Sabre’s Bengaluru GCC began operations in 2005. Initially, a handful of employees worked on some key projects in partnership with counterparts from the US. Over time, the centre grew in numbers, technology, and expertise. Today, Sabre Bengaluru GCC is a peer capability centre for product development, knowledge creation, and innovative solutions that help resolve customer pain points. Some of the ways the centre fosters a culture of innovation is by promoting co-creation across all levels, building communities of practice, and offering platforms like hackathons and idea incubators to create production-ready solutions. The differentiating factor for Sabre, when compared to other employers, is the fact that it offers to their team members the following opportunities – Solve challenging business problems using advanced algorithms, statistical and econometric analysis, big data, machine learning and artificial intelligenceBe part of the travel revolution to bring advanced & prescriptive analytics to the industry, and have the opportunity to innovate and deliver value to the customersApplication of expert-level statistical analysis, data modelling, and predictive analysis on strategic and operational challenges in the airline businessLearn business algorithms to translate questions into data analysis and models, define suitable KPIs, and present results to internal and external clients, sales and growth teamCombine statistical analysis and machine learning techniques, leveraging huge volumes of competitive data and a broad range of consumer insights, to identify key market and consumer trends Are you ready? “Be bold, make smart choices,” said Srinivasan, “Sabre offers great learning and growth opportunities for all team members. If you strive to put your best foot forward, the sky’s the limit.” Further, he said their organisation is a great place to work, and it will provide you countless opportunities to learn, practice, and apply your ideas. “We are pioneers in the travel technology space, and you will have the chance to drive bottom-line business and help shape the future of travel,” concluded Srinivasan. Check out data science jobs at Sabre here.","excerpt":"Sabre data science team consists of 50+ data scientists who are focused on the airline solutions side of their business.","categories":["AI Hirings"],"tags":["data science hiring india","data science salary India","data scientist hiring india","Data Scientist Jobs","jobs in bangalore","tech jobs india","what is data science"],"author_name":"Amit Naik","publish_date":"2022-02-22T13:00:00","publication_year":"2022","word_count":1569,"keywords":["what is data science","data science","Go","Data Scientist Jobs","artificial intelligence","machine learning","jobs in bangalore","AI","ML","tech jobs india","RAG","Aim","analytics","data science hiring india","data scientist hiring india","data science salary India","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","data science","analytics","Aim","RAG","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-hiring\/data-science-hiring-process-at-sabre\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10094040,"title":"AI Battle Heats Up: Microsoft to Take on Apple Head-on","content":"Animoji. FaceID. On-device AR. All these features were made possible by one small compute cluster deep inside Apple’s chips — the Apple Neural Engine (ANE). After taking over mobile AI compute with the ANE, Apple brought the chip to its M series of MacBooks. This represented a paradigm shift for ML researchers, that of on-device specialised AI compute. Now, there’s a new kid on the block. Announced recently at the three-day Microsoft Build conference, the Redmond giant is going big on AI compute on the edge. Partnering with market leaders Intel, AMD, and NVIDIA, Microsoft has announced a new line of silicon-level improvements made specifically for AI compute. With this move, the pillars of the PC ecosystem have teamed up to challenge Apple’s dominance in the AI ecosystem. Dreams of the edge At the recent Microsoft Build conference, the company expressed its desire to integrate AI into all its products. Developers were also incentivised to build AI features for Windows and other Microsoft products through the Hybrid AI Loop and Windows Olive for model optimisation. However, one important thing that stood out to us was the announcement of partnerships with AMD, Intel, NVIDIA, and Qualcomm to create new silicon optimised for AI compute. In conjunction with hardware manufacturers like Dell, HP, Lenovo, and more, this will create a new generation of personal computers with on-device AI capabilities. By working more closely with chip manufacturers, Microsoft seems to be pushing for more efficient inference on-device, rather than relying on Azure to do so. Not only will this cut down the cost for Microsoft, but it will also equip the next generation of computers with neural processing units (NPUs). These chips are purpose-built for AI tasks, and can perform them faster than general-purpose chips while remaining more efficient. AMD and Microsoft have already been reported as working together to create more capable AI chips. Reports have emerged that Microsoft is “providing support” to strengthen AMD’s AI chips. While these were only rumours before the Windows AI announcement, they now seem to be coming to fruition. AMD’s 7040-series of laptop chips has an in-built NPU, and already works with Windows 11 through the ONNX runtime. Documentation is now available for the 7040-series chips’ AI capabilities, making it easy for developers to add AI functionality to their applications. As part of the partnership, another company Intel has stated that its new line of Meteor Lake chips have a built-in neural vision processing unit or VPU. This will reportedly accelerate AI inference while also working in software like Adobe Premiere Pro for more “effective machine learning”. Apart from this, Intel has also added support for these chips on WinML and DirectML, so developers can directly address the new hardware. While NVIDIA has done its part in creating the AI compute market, it seems that it’s not completely done yet. As part of the partnership with Microsoft, NVIDIA has released updates to its driver software that will make RTX-enabled GPUs even faster at AI tasks. By leveraging the GPUs’ in-built tensor cores, NVIDIA has promised performance improvements for ML models, such as a 2x improvement for Stable Diffusion. Beyond the partnerships, Microsoft has been pouring resources into creating a robust dev chain for AI tasks on Windows. The biggest part of this is the ONNX runtime and tools like Windows Olive, which can help optimise models. The WinML API also plays a huge role in facilitating AI development on the platform, as it provides an easy way to integrate ML capabilities into Windows applications. While this partnership might seem like the Avengers of the tech world coming together, it seems that Thanos is sitting pretty waiting for the battle. Apple has already set the stage for on-edge AI development workloads. Apple-Microsoft: Head-on-Head Ever since Apple moved to the SoC (system on chip) design for laptop chips, the Neural Engine has been a mainstay of its chips. In 2021, TensorFlow was updated with the capability to allow AI models to be trained on the ANE. According to Apple, this resulted in almost a 5x improvement in training times for common workloads like CycleGAN, Style Transfer, DenseNet and more. Microsoft is trying to catch up to the M-series of chips by tying up with chipmakers to recreate the magic of ANE. While the Neural Engine is one of the most important pieces of the puzzle, it is just one part of the story. Apart from releasing all its new laptops with AI compute on board, Apple has also been hard at work creating developer tools to leverage the ANE. The devchain includes offerings like the Core ML library which allows developers to add pre-built ML models. It also allows them to integrate AI into their applications. What’s more, Apple also has machine learning APIs for common tasks like vision, speech, and natural language, all powered by the ANE. In a way, the WinML API is Microsoft’s answer to Apple’s Core ML library. WinML allows developers to leverage on-device processing capabilities to add ML to their applications. Apple is big on on-device ML processing, so it’s no surprise that all their offerings are aligned with this strategy. However, Microsoft always provides Azure as a fallback for developers when target devices do not have AI capabilities, maximising compatibility. John Giannandrea, Apple’s SVP for machine learning and AI strategy, stated in an interview, “I understand this perception of bigger models in data centres somehow being more accurate, but it’s actually wrong. It’s technically wrong. It’s better to run the model close to the data, rather than moving the data around.” As we can see, Apple and Windows may be targeting the same market, but in fundamentally different ways. With its lead in the market, privacy-preserving features, and on-device processing, it seems that Apple might hold on to the lead for now. On the other hand, Microsoft’s approach will bring AI to the masses, putting Windows devices on par with Apple’s for AI-focused tasks. While it might seem that this puts Apple and Windows head-to-head in the AI ecosystem, both companies are trying to stay on the curve. The market is progressing towards deployment of AI at the edge, pushing system manufacturers to include AI capabilities in the devices. Offering support and creating a developer ecosystem is a no-brainer for both these companies which are leading personal computing towards an AI-powered future.","excerpt":"With Microsoft’s new partnerships, the pillars of the PC ecosystem have teamed up to challenge Apple’s dominance in the AI ecosystem.","categories":["AI Features"],"tags":[],"author_name":"Anirudh VK","publish_date":"2023-05-29T14:00:00","publication_year":"2023","word_count":1060,"keywords":["Go","API","machine learning","AI","R","ML","RAG","edge AI","TensorFlow","Azure"],"extracted_tech_keywords":["AI","machine learning","ML","TensorFlow","edge AI","RAG","Azure","R","Go","API"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/ai-battle-heats-up-microsoft-to-take-on-apple-head-on\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10023623,"title":"Google Announces Recipients Of Research Scholar Program 2021","content":"Google has announced the recipients of the Research Scholar Program. The program was introduced in March 2020 to support early-career professors pursuing research in fields relevant to Google. The inaugural call for proposals was opened in November. The program has received a favourable response from researchers working in computer science, including machine learning, human-computer interaction, healthcare, systems and more. In the first year of the program, Google granted 77 awards, which included 86 principal investigators from 15+ countries and over 50 universities. “Of the 86 award recipients, 43% identify as a historically marginalised group within technology,” the company stated. The recipients were from across categories such as algorithms and optimisation, augmented reality, virtual reality, human-computer interaction, health research, machine learning and data mining, machine perception, networking, natural language processing, privacy, quantum computing, software engineering and programming languages, structured data, semantic graphs, database management and more. Google has supported faculty research for the last 15 years under the Faculty Research Awards Program supporting technical research in areas such as computer science, engineering and related domains. It has so far funded over 2,000 academics at ~400 Universities in 50+ countries. To support the broader research community, Google has introduced two new programs. The Research Scholar Program supports early-career faculty — those who have received their doctorate within the past 7 years and are doing impactful research in fields relevant to Google. It aims to support new collaborations and encourage long-term relationships. Google said these programs complement the existing support of academic research such as  Latin America Research Awards, the PhD Fellowship Program, the Visiting Researcher Program and research grant funding. Check out the full list of 2021 recipients here.","excerpt":"In the first year of the program, Google granted 77 awards, which included 86 principal investigators from 15+ countries.","categories":["AI News"],"tags":[],"author_name":"Srishti Deoras","publish_date":"2021-04-08T12:30:40","publication_year":"2021","word_count":277,"keywords":["Go","funding","machine learning","programming_languages:R","AI","programming_languages:Go","RAG","Aim","R","emerging_tech:quantum computing"],"extracted_tech_keywords":["AI","machine learning","Aim","RAG","R","Go","funding","programming_languages:R","programming_languages:Go","emerging_tech:quantum computing"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/google-announces-recipients-of-research-scholar-program-2021\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10170950,"title":"New DeepSeek-R1 Is as Good as OpenAI o3 and Gemini 2.5 Pro","content":"Chinese AI model maker DeepSeek announced a new update to its R1 reasoning model on Wednesday. The updated model, DeepSeek-R1-0528, is available on Hugging Face. “In the latest update, DeepSeek R1 has significantly improved its depth of reasoning and inference capabilities by leveraging increased computational resources and introducing algorithmic optimisation mechanisms during post-training,” said DeepSeek. The company also shared the model’s benchmark results, which showed that it achieved performance parity with OpenAI’s o3 and Google’s Gemini 2.5 Pro models on multiple evaluations. In the AIME 2025 test, DeepSeek-R1-0528 scored 87.5%, close to OpenAI-o3 (88.9%) and outperformed Gemini 2.5 Pro’s (83.0%). Besides, the model achieved scores on par with leading AI models on other coding, mathematics, and reasoning evaluations, as seen on Artificial Analysis. It scored 77% on LiveCodeBench (coding benchmark), matching Gemini 2.5 Pro (77%) and nearly OpenAI’s o3 (78%) in coding ability. On the reasoning and general knowledge benchmark MMLU-Pro, DeepSeek-R1 achieved 85%, comparable to Gemini 2.5 Pro (84%) and OpenAI’s o3 (85%). Source: Artificial Analysis Several users have already downloaded and deployed the model locally, as per their social media posts. Ivan Fioravanti, CTO of CoreView, said on X that he could run the DeepSeek-R1-0528-4bit at around 21 tokens per second on an Apple M3 Ultra chip-based device. The DeepSeek-R1 reasoning model, released last year, created quite a storm across the AI ecosystem. During its launch, the model surpassed several competing ones in benchmarks. DeepSeek prioritises using efficient techniques in the model’s architecture to improve performance rather than relying on high computing power. One of DeepSeek’s previous models, V3, used 2048 NVIDIA H800 GPUs to achieve performance better than most open-source models. Andrej Karpathy, former OpenAI researcher, said the DeepSeek V3’s level of capability is ‘supposed to require clusters of closer to 16,000 GPUs’. This caused numerous entities to doubt the demand for AI-related hardware, resulting in a market cap loss of over $500 billion for NVIDIA in just one day. Numerous startups and products use the open-source DeepSeek model for deployment, and its capabilities are extensively recognised across various sectors in China. Recently, it was reported to be used for research and development for the country’s ‘most advanced warplanes’. Besides, German automotive leader BMW revealed plans to incorporate DeepSeek into its vehicles in China. Last month, The New York Times revealed that courtroom officials are utilising DeepSeek to draft legal documents in minutes. Additionally, doctors and agencies are employing the model to locate missing persons. The report further noted that numerous companies are “encouraging” employees to adopt DeepSeek for design and customer service tasks.","excerpt":"The new DeepSeek-R1-0528 has “significantly improved its depth of reasoning and inference capabilities.”","categories":["AI News"],"tags":["AI (Artificial Intelligence)","DeepSeek"],"author_name":"Supreeth Koundinya","publish_date":"2025-05-29T18:01:51","publication_year":"2025","word_count":427,"keywords":["Hugging Face","OpenAI","AI","ML","DeepSeek V3","DeepSeek R1","RAG","DeepSeek","Gemini 2.5","Aim","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","ML","OpenAI","Gemini 2.5","DeepSeek V3","DeepSeek R1","Aim","Hugging Face","RAG","R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/new-deepseek-r1-as-good-as-openai-o3-and-gemini-2-5-pro\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":63945,"title":"Best Sites and Resources to Learn Python for Free","content":"Python is one of the most preferred high-level programming languages, which is being increasingly utilised in data science and in designing complex machine learning algorithms. In one of our articles, we discussed why one should learn the Python programming language for data science and machine learning. In this article, we list down the top free resources to learn Python for Machine Learning. Best Sites to Learn Python for Beginners 1. Google’s Python Class2. Introduction to Data Science using Python3. Data Science, Machine Learning, Data Analysis, Python & R4. MatPlotLib with Python5. Data Science with Analogies, Algorithms and Solved Problems6. Machine Learning In Python by Amazon7. Machine Learning With Python by Tutorial Point8. Free Python Course by GreatLearning9. Python for Data Science with Cognitive Class 1. Google’s Python Class About: This is a free class provided by the developers at Google. It includes written materials, lecture videos, and lots of code exercises to practice Python coding. The first exercises work on basic Python concepts like strings and lists, building up to the later exercises which are full programs dealing with text files, processes, and Http connections. Visit here to get started. 2. Introduction to Data Science using Python In this course, you will understand the basics of data science and analytics as well as how to use Python and scikit-learn. The course will show you what data science is and how is it used. You will go through commonly used terms and write some code in Python as well. Visit here to get started. 3. Data Science, Machine Learning, Data Analysis, Python & R This course has been designed by data scientists to help you learn complex theory, algorithms, and Python libraries. With every tutorial, you will develop new skills and improve your understanding of this challenging yet lucrative sub-field of data science. The course includes both Python and R and is also packed with practical exercises that are based on real-life examples. Visit here to get started. 4. MatPlotLib with Python This course has been designed for those who want to learn a variety of ways to visually display data. With over 58 lectures and 6 hours of content, this course covers almost every major chart that Matplotlib is capable of providing. Intended for students who already have a basic understanding of Python. You will take a step-by-step approach to create line graphs, scatter plots, stack plots, bar charts, 3D lines, 3D wireframes, 3D bar charts, 3D scatter plots, geographic maps, live-updating graphs, and much more. On completion of this course, you will not only have gained a deep understanding of the options available for visualising data but also know how to create well presented, visually appealing graphs. Visit here to get started. 5. Data Science with Analogies, Algorithms and Solved Problems This course will help you learn complex theory, algorithms and coding libraries in a simple way. With every tutorial, you will gain new skills and enhance your understanding of this field. It includes a brief introduction to Python and its libraries and how to implement this language in machine learning. Click here to get started. 6. Machine Learning In Python by Amazon In this e-book, you will learn essential techniques of machine learning in predictive analysis using Python programming language. The topic includes predictive model building, writing ensemble methods using Python, understanding and working with data and much more. Visit here to get started. 7. Machine Learning With Python by Tutorial Point This tutorial provides a quick introduction to Python and its libraries like NumPy, SciPy, pandas, Matplotlib, and explains how it can be applied to develop machine learning algorithms that solve real-world problems. The tutorial starts with an introduction to machine learning and the Python language and shows you how to set up Python and its packages. It further covers all important concepts such as exploratory data analysis, data preprocessing, feature extraction, data visualisation, clustering, classification, regression and model performance evaluation. Visit here to get started. 8. Free Python Course by GreatLearning Here, you will learn the basics of Python and Artificial Intelligence, object-oriented programming and create GUIs with Python. It includes an introduction to AI, why Python is becoming the central tool for data scientists, alternatives to Python in data science and other such. Visit here to get started. 9. Python for Data Science with Cognitive Class This free Python course provides a beginner-friendly introduction to Python for Data Science. This will kickstart your learning of Python for data science, as well as programming in general. Upon its completion, you’ll be able to write your own Python scripts and perform basic hands-on data analysis using Jupyter-based lab environment. Visit here to get started.","excerpt":"Python is one of the most preferred high-level programming languages, which is being increasingly utilised in data science and in designing complex machine learning algorithms. In one of our articles, we discussed why one should learn the Python programming language for data science and machine learning. In this article, we list down the top free […]","categories":["AI Trends"],"tags":["beginner python projects","full stack project ideas","Python","python for machine learning","Python language","python machine learning","Python Programming","scikit learn","Why is Python so Popular"],"author_name":"Ambika Choudhury","publish_date":"2020-04-30T10:00:00","publication_year":"2020","word_count":777,"keywords":["scikit-learn","scikit learn","Python Programming","Pandas","data science","NumPy","artificial intelligence","Why is Python so Popular","analytics","full stack project ideas","Python language","machine learning","AI","python for machine learning","Matplotlib","python machine learning","Python","Jupyter","beginner python projects"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","data science","analytics","scikit-learn","Jupyter","Pandas","NumPy","Matplotlib"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-free-9-resources-to-learn-python-for-machine-learning\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":69420,"title":"How Securing Open-Source Projects Is The Key To Innovation","content":"Trademarks and protection of ownership of the project have been the foundation of the open-source community. Although it might sound contradictory to associate trademarks with open-source, protecting the ownership can be critical for developers to maintain the reputation and long-term sustainability of projects. However, with the increasing number of open-source enterprise applications, it gets challenging for developers and companies to protect the brand name. And that’s why Google has launched a new regulatory body — Open Usage Commons — to help companies manage the identity of projects despite giving the licensing for its source code. Importance of trademarks in open-source and how Google’s initiative can help Sometimes it gets difficult for developers and companies to understand that obtaining the license of the source code doesn’t give them the right to use the trademark of projects which includes — name, logo and other branding accessories, stated in the company’s blog post. This has been an issue with open-source licensing, where it includes copyright and patent laws. However, it leaves out the trademark licenses. Trademark is the possessorship of projects, therefore to avoid losing control, trademark and proprietary licenses should be controlled by developers of the projects only. Case in point two prominent open-source relational database management — MariaDB and MySQL, where both have derived from the same database structure and indexes, both have distinct properties. And therefore, raises a serious question of how these companies will associate their products and solutions with the brand of the community? The importance of trademark has been well established by the open-source company RedHat in 2004, where they highlighted the problem of companies using RedHat’s name for their projects. As a matter of fact, in the US, it isn’t required for companies to register for trademarks formally, and thus provide entire ownership and control of projects to developers. However, for other instances, companies must file for registered trademarks to reserve their rights over the projects. Well-defined trademark policies will remove the ambiguity of ownership and will enable developers to protect their projects from misinterpretation and illegitimate usage. Also Read: Can Artificial Intelligence Be Granted Patent For Inventions? One such licensing is provided by Apache 2.0, where it has been defined that although companies can use the source code for their business purposes its only the organisation who has developed the project are liable to own the proprietorship of everything else the patent covers. Google’s recent initiative of opening up a new organisation can definitely help project maintainers to define their identity by providing “neutral and independent ownership.” According to the company’s blog post, projects that are registered under their new organisation will receive google’s support for protecting their own and will provide a useful guideline and compliance testing. Having a name of Open Usage Commons (OUC) will also offer assurance to stakeholders of the brand and the quality for their offering. When asked, Chris DiBona, the director of open-source at Google stated in an interview that with OUC developers can work together without undergoing substantial negotiations with companies and organisations. Currently, the OUC has been launched as a third-party initiative, however, funded by Google, in collaboration with cloud service provider SADA Systems, and academic researchers and contributors. To initiate the launch, Google also added three of their own projects under OUC — web application framework, Angular, open-source independent service mesh, Istio, and web-based team code collaboration tool, Gerrit. Registering their projects under this organisation will also help the developers to provide more services to their customers as well as more integrations. Contradictorily, experts have noted that unlike the Kubernetes project, Google donated their Istio project to OUC instead of Cloud Native Computing Foundation (CNCF). This again negates the statement made earlier this year by Thomas Kurian, the CEO of Google Cloud. Not turning the project to CNCF will probably increase Google control of the project, which will omit the vendor-neutral approach. Thus, it could challenge Google’s view of enhancing the open-source community. Wrapping Up Open-source indeed provides acceleration to products and solutions development but, trademark issues can create immense sustainability issues for projects. Adequately registered projects will allow users to identify the owners, versions and quality of projects. However, the process of attaining one can be tiring for developers. With Google OUC initiative, it will be easier for companies to manage their commercial entities. Experts believe that, although the open-source community encourages sharing and collaboration, it is critical for developers to protect and secure their projects to retain the value.","excerpt":"Trademarks and protection of ownership of the project have been the foundation of the open-source community. Although it might sound contradictory to associate trademarks with open-source, protecting the ownership can be critical for developers to maintain the reputation and long-term sustainability of projects.  However, with the increasing number of open-source enterprise applications, it gets challenging […]","categories":["AI Features"],"tags":["Open Source","Open Source AI","open source project","open source projects on github","open source tools"],"author_name":"Sejuti Das","publish_date":"2020-07-12T16:00:00","publication_year":"2020","word_count":745,"keywords":["Go","open source projects on github","artificial intelligence","Open Source","AWS","AI","open source tools","Git","RAG","Open Source AI","SQL","GAN","open source project","R","kubernetes"],"extracted_tech_keywords":["AI","artificial intelligence","RAG","AWS","kubernetes","R","SQL","Go","Git","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-securing-open-source-projects-is-the-key-to-innovation\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":2503,"title":"Interview &#8211; Vinay Mantha, Head of Analytics COE at Sapient Nitro","content":"[dropcap style=”1″ size=”2″]AIM[\/dropcap]Analytics India Magazine: What are some of the main tenets (philosophies, goals, attributes) of analytics approach and policies at your organization? [dropcap style=”1″ size=”2″]VM[\/dropcap]Vinay Mantha: While there exist several philosophies towards data and attributes towards measurement of success, we at SapientNitro strive towards one goal as far as analytics is concerned – “Demonstrate Data Stewardship by helping our clients measure and optimize the ROI on their digital investments”. This, we achieve through a razor sharp focus on strategy and execution by providing an efficient system of governance, enablement and actionable insights through an in-depth analysis of data. AIM: Please brief us about some business solutions you work on and how you derive value out of it. VM: Our analytics solutions cut across several industries – Telecom, Banking, Retail Electronics, Apparel, Automotive, Sports, FMCG and Food – to name a few. To this extent, our clients are industry leaders in each of these verticals. Value comes in several flavors – from cost savings through an effective system of governance and operational efficiencies to outright potential monetary gains as a result of implementing the changes our analysts recommend. For one of our clients, who is a leading Telecom player, we have pioneered and executed on the concept of a “Value-Driven Analytics Framework”. In its simplest form, it is an approach where every recommendation we provide is backed by a monetary value that is to be potentially realized should the recommendation be implemented. As you would see, this bold and revolutionary approach requires a deep understanding of our client’s business along with a great deal of confidence in our capabilities in drawing meaningful conclusions from the myriad of customer data. The best part of this approach, is that once implemented and validated, it a) more than helps justify the investments in analytics and b) provides a more compelling story to enforce change even from tough stakeholders who are normally resistant to change (who does not mind extra $$$ for their business?). AIM: How does a typical requirement gathering to delivery cycle looks like for you? VM: The key thing to understand is that we do not view Analytics as a linear process but more as a closed-loop process. The below pictorials should give you a better picture of where we see analytics fitting in to the delivery process: AIM: Please brief us about the size of your analytics division and what is hierarchal alignment, both depth and breadth. VM: At a high level there are two core divisions within SapientNitro that cater to our clients’ analytics needs. The first is a “Data Shared Services” team that provides modeling\/ETL and Data Visualization services in the BI\/DW space. The second is an “Analytics & Optimization” practice – a group of professionals who focus on web\/mobile\/social\/competitive analysis and optimization (Search & Siteoptimization through Multi-Variate Testing). This does not include a de-centralized group of analytics implementation specialists who are embedded in project-specific teams given their expertise on platforms such as DemandWare, ATG, Adobe CQ and such. As a company policy we do not divulge specific numbers. [quote style=”1″]Of relevance is the fact that Sapient has recently completed 2 strategic company acquisitions – “Iota Partners” and “Mphasize” to elevate the breadth and depth of our analytics service offerings to our clients.[\/quote] AIM: Would you like to share any example of an Insight that generated a huge positive impact for your clients? VM: For one of our UK-based clients, as we were attempting to diagnose the cause for the high exit rate of a key page in the purchase funnel, we realized that a specific combination of a Browser and an OS was causing a page error which prompted visitors to exit the site. This was obviously not caught by the testing team before code deployment. We instantly worked with the development team to fix the problem and the resolution has resulted in a revenue uplift of £1.5M\/quarter. AIM: Do you think it’s possible to become too married to the data that comes out of analytics? Where do you draw the line? VM: To answer this question, I am going to take the case of Web Analytics. Yes that is very much possible to be obsessed and thereby picky about the data. Confusion often stems around the question “Is my data accurate”? The trick lies in helping the stakeholders realize that this data is never meant to be “absolutely” accurate, but rather “directionally” accurate – Take my word for it, many questions often stem from this lack of understanding. That being said, the beauty of web analytics is not about honing down to an exact user who performed a particular transaction on your web site, but rather taking this directionally accurate data, segmenting it into meaningful buckets and then drawing inferences based on the behavior of segments of visitors. There is no reason to draw a hard line. It is better to try implementing the recommendations of an analyst, learn and adapt. AIM: What are some of the data measurement points that are becoming more important for organizations? VM: Analysis of web, mobile, social (including sentiment) and CRM data – both quantitative as well as qualitative. Big data which essentially congregates all of the above, as we all know is steadily knocking on our door. AIM: What are the most significant challenges you face being in the forefront of analytics space? VM: #1 – Analytics is often an after-thought to the extent that many folks do not even think about it until after say a web site is deployed to production, at which point it becomes more expensive to fix issues with analytics implementation and setup, let alone draw meaningful insights. #2 – Shortage of strong analytics talent (particularly in the Indian sub-continent). AIM: How did you start your career in analytics? VM: Happened more by chance than choice. As a part of a college hire rush, I was interviewed and subsequently accepted a job with Intel (Oregon, USA) not knowing which group I would be getting into. I happened to land into a group which was tasked with monitoring web stats for intel.com. That was my first exposure to web analytics. As I progressed there on, my focus shifted back and forth between Web\/Mobile Analytics and traditional BI\/DW work spanning various roles. AIM: What do you suggest to new graduates aspiring to get into analytics space? [pullquote align=”left”]Thanks to the efforts of media, Analytics is one of the most sought-after career tracks today. Just take a plunge and the joy can be immense.[\/pullquote]VM: Folks who have a natural bent of mind to think logically and out-of-the-box, I think would be really successful as analysts – the tools of the trade can be easily taught. Great communication skills are a boon (given the necessity of direct client interaction in most cases), and a high degree of comfort with Microsoft Office is a plus. It is important to remember, not everyone in this field needs to become an analyst. There is a dire need for folks with a technology background who can serve as analytics enablement specialists as well. For this, prior exposure to JavaScript programming is very helpful. AIM: What kind of knowledge worker do you recruit and what is the selection methodology? What skill sets do you look at while recruiting in analytics? VM: For external hires, we typically look for someone who has anywhere between 3-10 years of experience depending on the position we are seeking to fill. In terms of the interview process to hire Analysts, for example, the aspiring candidates run through 2 preliminary rounds of interview before we even see their resume – the first is a multiple choice technical questionnaire administered by our HR team (cut-off pass mark 80%) followed by an exercise where they are provided a spreadsheet with raw sales\/spend\/traffic data of a fictitious company and asked to provide insights and recommendations in a deck within 24 hours (they can Google or use any book in the world for this open-book test). It is after this that we see the resume along with their exercise submission and decide whether to go forward with the next round of technical domain and attribute interviews. As you can see, our bar for hiring experienced talent is very high given our brand image. We also hire internally from other teams at Sapient, where we target folks with 1 year of experience or so. We put them through a structured program of testing and selection. What follows then is training to add\/elevate their knowledge on analytics. We then place them on a client assignment under the guidance of a senior mentor. AIM: How do you see Analytics evolving today in the industry as a whole? What are the most important contemporary trends that you see emerging in the Analytics space across the globe? VM: In this age of web 2.0, the entire industry is beginning to realize the value analytics adds in optimizing the ROI in digital. Corporations are getting increasingly eager to tap into the wealth of the data that they have collected over the years and are continuing to collect, to help make smarter decisions. With the aid of smarter technology that is gaining more sophistication by the day, it is becoming increasingly feasible to tap into massive data vaults which were previously perceived as impenetrable. Clients are increasingly seeking partners who have deep expertise and experience in strategizing and delivering analytics solutions for their specific business niche (such as Banking, Telecom etc.;) AIM: Anything else you wish to add? VM: Big Data is the top buzz word in the analytics industry today. While I completely agree with the value proposition, here is a tip I would give to organizations that are not yet as mature and ready – First have an analytics strategy in place and very importantly “Mine your Megabytes before you Peel your Petabytes”. [divider top=”1″] [spoiler title=”Biography of Vinay Mantha” open=”0″ style=”2″] Vinay Mantha heads SapientNitro’s Data Analytics & Optimization Center of Excellence (COE) in India. In his current role he is focused on an ongoing expansion of the analytics delivery arm to serve clients from NA\/EU\/APAC and on building framework-based analytical capabilities for SapientNitro. Some of his noteworthy clients such include the likes of Vodafone, RIM, Marks & Spencer, Unilever and Citibank. Vinay is a certified Project Management Professional (PMP) who holds an M.S degree in Computer Science from Wright State University (OH, USA) and an Engineering Bachelor’s degree from Osmania University (Hyderabad). A data evangelist by passion, over a career spanning 15 years, Vinay has spent most of his time chartering data mazes and solving data puzzles – in this process, transforming data into information leading to actionable insights. Starting his career with Intel as a database developer, he has subsequently worked with companies such as Unisys and most recently Simplexity (a US-based startup focusing on mobile retail e-commerce) in various hands-on as well as leadership roles. The entrepreneurial opportunity to build a “Center of Excellence” ground-up and execute on his vision\/strategy prompted Vinay to take up this position with SapientNitro and relocate from US to India in 2011. Life has been great ever since. [\/spoiler]","excerpt":"[dropcap style=”1″ size=”2″]AIM[\/dropcap]Analytics India Magazine: What are some of the main tenets (philosophies, goals, attributes) of analytics approach and policies at your organization? [dropcap style=”1″ size=”2″]VM[\/dropcap]Vinay Mantha: While there exist several philosophies towards data and attributes towards measurement of success, we at SapientNitro strive towards one goal as far as analytics is concerned – “Demonstrate […]","categories":["AI Features"],"tags":["Interviews and Discussions"],"author_name":"Дарья","publish_date":"2013-01-24T06:22:17","publication_year":"2013","word_count":1860,"keywords":["big data","Go","API","AI","Git","Aim","analytics","JavaScript","R","Java","Interviews and Discussions"],"extracted_tech_keywords":["AI","analytics","Aim","R","JavaScript","Go","Java","Git","API","big data"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/interview-vinay-mantha-head-of-analytics-coe-at-sapient-nitro\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10137857,"title":"Bangalore Leads the Way in Sourcing Talent for Frontend, Backend, DevOps, and Data Science Roles in India","content":"Recently released Instahyre Report states that Bangalore has emerged as the most sought after  sourcing location for multiple BFSI tech skills – Frontend, Backend, DevOps and Data Science, followed by Pune and Hyderabad. The report further revealed  that Bangalore also tops as the source of talent for roles in DevOps, providing more than 30% of professionals skilled in Docker, Kubernetes, Jenkins and AWS. Over the years, Bangalore has become a key location for the disruptive startup ecosystem, tech MNCs, and a steady influx of technology innovations. The city is now a hotbed for companies looking to tap in top tier talent. Despite facing uncertainties and layoff announcements, tech hiring in the Indian IT industry remains resilient, showing a positive growth outlook. Data indicates that major IT companies are planning to expand their workforce to meet the growing demand for IT services in India. On an average, these tech giants are expected to add between 40,000 and 50,000 new employees across various tech roles in the near future. Bullish on the industry’s growth, Mr. Sarbojit Mallick, Co-founder of Instahyre said, “Bangalore, the ‘Silicon Valley of India’, has solidified its position as the leading hub for tech talent in the country. It consistently outshines other Indian cities, making it a crucial player in both the national and global tech landscape.” Mallick further added that along with Hyderabad, Bangalore was the cradle of tech services right from the outsourcing days. “Building on that legacy and the startup boom, it has cemented its leading position year-on-year, as our report’s data shows,” he noted. In the Data Science role, Bangalore is again playing host to more than 30% talent for Machine Learning, Computer Vision, NLP and Data Visualization, followed by Pune and Hyderabad. The city also leads the pack in sourcing Security professionals who are proficient in extremely vital tech security skills like Info Security, Security Testing, App Security, and Network Security. When it comes to talent sourcing, the Instahyre report reveals that Bangalore holds the lion’s share, supplying experts skilled in Data Analysis, Warehousing, Data Collection, and Data Extraction. In addition, the city also leads tech hiring within the Networking space, providing skilled talent from Network Analysis, Network Testing, Network Admin and Troubleshooting functions and QA + Risk skills including Quality Assurance, Quality Control, Risk Assessment, and Risk Management. Given the cutting edge technology ecosystem in Bangalore, the city continues to mark its stride as the preferred destination for companies seeking skilled BFSI tech talent across various functions. Without doubt, Bangalore’s ability to consistently produce top-tier talent is the reason behind India being on the global stage, not just in terms of technological output, but also as a source of talent.","excerpt":"Bangalore’s ability to consistently produce top-tier talent is the reason behind India on the global stage, not just in terms of technological output but also as a source of talent.","categories":["AI News"],"tags":["Bengaluru","Data Science"],"author_name":"Shalini Mondal","publish_date":"2024-10-08T17:32:55","publication_year":"2024","word_count":447,"keywords":["data science","machine learning","TPU","AWS","AI","docker","computer vision","RAG","NLP","kubernetes","Data Science","Bengaluru"],"extracted_tech_keywords":["AI","machine learning","NLP","computer vision","data science","RAG","AWS","kubernetes","docker","TPU"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/bangalore-leads-the-way-in-sourcing-talent-for-frontend-backend-devops-and-data-science-roles-in-india\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10059541,"title":"Artificial Intelligence at IIT (BHU) Varanasi: Interview with Director Pramod Kumar Jain","content":"IIT (BHU), Varanasi is one of the oldest engineering institutes in India. Located in Uttar Pradesh, the institution started its life as the Banaras Engineering College in 1919, and became the Institute of Technology, Banaras Hindu University in 1968. The Indian Institute of Technology tag was conferred in 2012. Prof. Pramod Kumar Jain, Director, IIT (BHU) Varanasi was appointed as the Director at IIT – BHU from August 1,  2018. Analytics India Magazine caught up with him to understand the IIT’s contribution to the field of AI & ML in India, and the ongoing research at the premiere institute. IIT (BHU), Varanasi has 13 branches of engineering and three inter-disciplinary schools. AIM: What are the major drivers in scientific research? Prof. Pramod Kumar Jain: Scientific research requires relevance of purpose, originality of objectives, reproducibility of results, and involvement of critical brain faculties. It is an original investigation directed towards gaining new knowledge through existing concepts. It is generally done in a focused area which may lead to the generation of new and better insights through creativity and critical thinking. AIM: Tell us about the ongoing research in the field of AI\/ML and data science at IIT (BHU) Varanasi. Prof. Pramod Kumar Jain: IIT(BHU) has been contributing to theoretical as well as applied research in the core areas of AI\/ML, with the support of various government agencies and industries. The area of Data Analytics and Predictive Technologies (DAPT) has been identified as one of the vibrant and emerging fields. DAPT refers to applying pattern recognition and inference engine as a smart decision support interface to take timely decisions and deploy necessary actuation to take holistic control of the real-life situations, either online or offline. Today, the term DAPT is used for systems involving intelligent pattern-recognition based decision support interfaces backed up by densely embedded sensors and actuators, interconnected by the Internet of Things (IoT), with computations occurring either within the network, or at the edge, or in the cloud–to observe, control, and optimise the performance of technical\/societal\/human systems, such as smart cities, smart energy grids, smart urban water networks, smart homes, smart farms or smart health tech. Any good progress in this area will significantly impact several critical aspects of the Indian economy, such as telecommunications, power, road transport and highways, defence research and development, and health and family Welfare. Other relevant areas under DAPT are urban smart city initiatives, including city-wide water distribution, waste management, air quality monitoring, pollution control, manufacturing, and governance at national and international levels. The ongoing research in the field of Analytics (AI\/ML and Data Science) at IIT (BHU), Varanasi are: Visual Computing and Analytics Lab led by Professor Sanjay Kumar Singh aspires to come up with real innovations in Medical Imaging using machine \/ deep learning methods. VCALAB is utilising AI for designing better computational tools that can help wet-lab researchers to identify antimicrobial peptides for combating antimicrobial resistance using AI.In Analysis of Industrial Vibration Data using AI Techniques, the  institute is performing the digital health monitoring of equipment through AI methods to improve the system’s availability and reduce financial and human losses.The institute is also working on the early classification of time series data utilising AI techniques. AIM: What are the institution’s contributions towards AI\/ML and data science? Prof. Pramod Kumar Jain: IIT(BHU) focuses on both fundamental and applied research. The research is supported through different government agencies as well as the IT industry. Recently, a path-breaking technology has been developed in IIT (BHU) that can reduce the cost of the onboard charging of Electric Vehicles by about 50 percent. This revolutionary and wholly indigenous technology will reduce the overall cost of electric vehicles and significantly impact EVs on Indian roads. Furthermore, several simple hand-held bio-sensing devices that can detect biomolecules like alkaline phosphatase, cancer biomarkers, etc., in complex biological fluids have been developed by researchers of IIT (BHU). This has a tremendous role in early disease diagnosis. IIT (BHU) is the nodal center and a Technology Innovation Hub (TIH) for technology development and activities in “Data Analytics and Predictive Technologies” and other related areas under National Mission on Interdisciplinary Cyber-Physical Systems (NMICPS). The activities envisioned under this mission will provide a great fillip to Indian manufacturing and societal betterment by inventing new products\/processes and services. This endeavour shall catalyse the creation of skilled young engineers, researchers, technicians, and entrepreneurs, together with the human resource at all levels, besides becoming a pivotal contributor to realising the vision of “Digital India”, “Innovate in India,” and “Make in India.” MoU with AISPL IIT BHU has also signed an MOU with the Amazon Internet Services Private Limited (AISPL) on June 19, 2018 to establish an AI and ML Cloud Research Lab. Under the MoU agreement, IIT-BHU gains access to the resources in the Amazon Web Services (AWS) Educate program and curriculum designed for higher education institutions to incorporate into their courses. The Cloud Research Lab will provide students with opportunities to use AWS Cloud technology to pursue research initiatives that focus on AI and ML innovation for India. AWS offers over 125 fully featured services for compute, storage, databases, networking, analytics, ML, AI, Internet of Things (IoT), mobile, security, hybrid, virtual and augmented reality (VR and AR), media, and application development, deployment, and management. All these features will now be available to IIT-BHU faculty, researchers and students to develop innovative solutions for India. With the resources and tools provided by the global AWS Educate program, the educators and students in IIT-BHU gain access to the latest and most innovative AWS Cloud technologies, along with open course content contributed by other top education institutions around the world.","excerpt":"IIT BHU has also signed an MOU with the Amazon Internet Services Private Limited (AISPL) on June 19, 2018 to establish an AI and ML Cloud Research Lab.","categories":["AI Features"],"tags":["IIT","Interviews and Discussions"],"author_name":"Poornima Nataraj","publish_date":"2022-02-01T13:00:00","publication_year":"2022","word_count":939,"keywords":["data science","Go","AWS","AI","ML","RAG","Aim","deep learning","analytics","IIT","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","ML","deep learning","data science","analytics","Aim","RAG","AWS","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/artificial-intelligence-at-iit-bhu-varanasi-interview-with-director-pamod-kumar-jain\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10064609,"title":"When I started my career, I wanted to be recognised as a professional rather than a woman: Wipro’s Sandhya Arun","content":"In her 28 years of professional career, Sandhya Arun has held leadership roles across business strategy, functional and technology consulting, sales and marketing and global delivery. The self-confessed “wonderer and a wanderer” is currently the Vice President and Sector Delivery Head, Integrated Digital Engineering and Application Services (IDEAS) of Wipro Ltd. At Analytics India Magazine’s Women In AI Conference: The Rising 2022.  Sandhya spoke about how Wipro leveraged AI to drive the digital transformation of global organisations. Watch all the recorded sessions of Rising 2022 here>> Evolution Over the last 75 years, Wipro has honed its capabilities in design, digital, strategy, engineering, AI, and cybersecurity, alongside building strong global credentials. “The years following 2010 were my favourite part. This was when Wipro was committed to riding the digital wave and launched “Wipro Digital” as a line of business within the enterprise. It was almost like Wipro was the benevolent venture capitalist (Wipro Ventures), and Digital was the emerging start-up they funded. It was super exciting to be a part of the founding group,” Sandhya said. At the time, Wipro had acquired the global cloud services company Appirio, known for creating next-generation worker and customer experiences. The acquisition positioned Wipro as one of the world’s largest cloud transformation practices and a game-changer in today’s as-a-service and digital economy. Appirio also introduced Wipro to Topcoder, a leading crowdsourcing marketplace connecting over a million designers, developers and data scientists with customers worldwide. “Today, clients use Topcoder products to deliver high-quality, repeatable outcomes that can be easily plugged into sprints and innovation cycles,” she said. In addition, Wipro Digital also acquired Designit, a strategic design firm specialising in designing transformative product-service experiences. The company even established Rapid Prototyping Labs or Protolabs to help clients build and test proof-of-concept with emerging and next-generation technologies in an accelerated manner. “Protolabs plays a key role in helping customers design and conceptualise their digital businesses and alter their business portfolio mix,” she added. The AI era Artificial Intelligence is at the core of any digital transformation initiative driving the experience and intelligence quotient of its processes through the infusion of intelligence and automation. Enterprises are focused on using data analytics and AI to navigate disruption. Sandhya said her focus has been on traditional companies hopping on the digital bandwagon for the first time or those that rely a great deal on legacy systems. “We want to help such companies transform, become agile and nimble!” she said. “For me, it is important to be connected with a larger purpose than work and quarterly results. So social responsibility is a must! We, at Wipro, believe business fuels our purpose, and purpose fuels our business. The Azim Premji Foundation owns 66% of Wipro’s economic interest,” said Sandhya Arun. Wipro has been recognised as one of the World’s Most Ethical Companies by the Ethisphere Institute for the 10th successive year. Diversity & inclusion As much as 36.4% of Wipro’s employees are now women. The company has also offered unconscious bias training to over 183,339 employees. “When I started my career, I wanted to be recognised as a professional rather than a woman. That was a mistake. There was no reason for me to hide my femininity,” said Sandhya Arun. Wipro is recognised as one of the Best Places to Work for LGBTQ+ Equality by the Human Rights Campaign Corporate Equality Index 2022. Additionally, The company has won the ‘Best Employer for Persons with Disabilities’ and second runner-up for ‘Best Employer for D&I’ in the large category at the 2nd ASSOCHAM Inclusion & Diversity Excellence Awards.","excerpt":"As much as 36.4% of Wipro’s employees are now women.","categories":["Deep Tech"],"tags":["AI (Artificial Intelligence)","Data Science","Deep Learning","Machine Learning"],"author_name":"Sri Krishna","publish_date":"2022-04-08T20:32:12","publication_year":"2022","word_count":593,"keywords":["Go","API","artificial intelligence","AI","digital transformation","Machine Learning","Git","RAG","analytics","GAN","Deep Learning","Data Science","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","analytics","RAG","R","Go","Git","API","GAN","digital transformation"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/when-i-started-my-career-i-wanted-to-be-recognised-as-a-professional-rather-than-a-woman-wipros-sandhya-arun\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10014189,"title":"IBM Commercialises Its AI FactSheets. Could It Become An Industry Standard?","content":"IBM has recently announced the commercialisation of its AI FactSheets, which were first introduced in 2018. In an official release, the company wrote that it plans to “commercialise key automated documentation capabilities from IBM Research’s AI FactSheets methodology into Watson Studio in Cloud Pak for Data throughout 2021”. This fact sheet will provide businesses with a framework to define how AI is to be used, to measure a model’s performance, and to generate reports on internal and external transparency. Can such an AI fact sheet from IBM become an industry standard? What Is IBM FactSheet & Why Is It Required? Some of the most important factors to consider while achieving trust in AI are — fairness, safety, explainability, reliability, and accountability. Apart from these factors, it must be accompanied by having a parameter against which the models are measured. A lot of this could be attributed to the increasing usage of AI models and services, even in high-stakes situations such as financial risk assessments, medical diagnosis, talent acquisition, policing, and governance. First proposed in 2018, the authors of IBM’s AI FactSheets argued that every AI service developer and the provider must release a Supplier’s Declaration of Conformity in order to assure transparency and trust. As a matter of fact, such a factsheet was akin to nutrition labels that one may find on food items or information sheets for appliances. “Standardising and publicising this information is key to building trust in AI services across the industry,” IBM had then noted. Along with this, the fact sheets include initial suggestions, information on system operation, algorithms used, testing set-up, performance benchmarks, and fairness and robustness checks. With the recent announcement from IBM, this AI FactSheet tool will complement IBM Cloud Pak for Data. Notably, IBM has also added new capabilities for Cloud Pak for Data to provide a foundation for AI to run on any cloud while providing enhanced governance and security. This will ensure federated learning to facilitate model training on distributed datasets while assuring security. AI FactSheets Offer: The scope for policy creation which defines what information is collected on models, who can use the model and for what purpose, and the way it should operate.The AI FactSheets are designed to help in automatically capturing the model facts as detailed in the FactSheet template throughout the AI lifecycle.These FactSheets can offer knowledge on the AI model in multiple formats depending on the preferences of the user and external audience. Some of the questions from IBM’s FactSheets include: Whether the dataset being used for training has a datasheet or data statement?What are the biases policies against which the dataset was checked for any possible bias?What is the target user for the explanation — machine learning expert, domain expert, consumer, or regulatorWhat is the testing methodology followed?Robustness policies followed to ensure the model’s protection against possible adversarial attack.Expected behaviours if input deviates from the training\/testing dataHow is the overall workflow to AI service tracked? Further, to evaluate the quality of the fact sheet, IBM has also developed a set of quality dimensions which are based on prior research in software development and communication. A few of the dimensions: Credit: IBM Can It Become An Industry Standard? To deploy and scale AI, enterprises must invest their trust in their models and the business outcomes of the entire AI lifecycle. IBM says that the new announcement of commercialising the AI FactSheets on Watson will provide easy access to clients of their comprehensive portfolio of AI governance solutions, which in turn will increase transparency, manage risks, and establish trust on AI models. It is true that apart from such a FactSheet, there is no standard of explaining or establishing the capabilities of an AI model. It forces the customers to make an uninformed decision and go with the common consensus. So naturally, a standardised fact sheet on the lines of IBM FactSheet will help indirectly addressing the issues, as also suggested by industry experts.","excerpt":"IBM has recently announced the commercialisation of its AI FactSheets, which were first introduced in 2018. In an official release, the company wrote that it plans to “commercialise key automated documentation capabilities from IBM Research’s AI FactSheets methodology into Watson Studio in Cloud Pak for Data throughout 2021”. This fact sheet will provide businesses with […]","categories":["AI Features"],"tags":["Ethical AI","purpose of ai","Transparency AI"],"author_name":"Shraddha Goled","publish_date":"2020-12-14T18:00:00","publication_year":"2020","word_count":658,"keywords":["federated learning","Go","machine learning","Transparency AI","AI","programming_languages:R","Ethical AI","programming_languages:Go","Rust","purpose of ai","AI governance","R","programming_languages:Rust"],"extracted_tech_keywords":["AI","machine learning","federated learning","R","Go","Rust","AI governance","programming_languages:R","programming_languages:Go","programming_languages:Rust"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/ibm-commercialises-its-ai-factsheets-could-it-become-an-industry-standard\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":14304,"title":"Aureus Analytics inks multi-year analytics deal with Landmark Insurance Broker","content":"Apaar Kasliwal, Executive Director at Landmark and Ashish Tanna, COO & Co-Founder, Aureus Analytics Aureus Analytics recently made a significant catch by securing a multi-year analytics deal with Landmark Insurance Broker, a prominent insurance broker. As part of the deal, Aureus Analytics will help Landmark Insurance Broker improve their end customer experience by leveraging data analytics and machine learning. Under the ambit of the deal, Aureus will help Landmark build customer intelligence across their lines of business, and drive growth riding on superior customer experience. This solution will be delivered using Microsoft’s Azure cloud infrastructure. The insurance industry is traditionally viewed as conservative in adopting newer technologies. But with a rise in online fraud and challenges in customer retention, they are increasingly looking at predictive analytics and machine learning to understand the customers better. Insurance carriers across the world are challenged with improving their customer engagement and experience. According to InsuraneFraud.org, online fraud siphons off about $80 billion a year across all lines of insurance. On the other hand, insurers lose about 18-20% of their customers to churn. These double-barreled challenges are making insurers leverage new technologies as soon as possible. The big data and predictive analytics solutions provider revealed that traditionally insurance brokers are not considered to be too tech savvy, and that shows in the end customers experience and journey. By securing this deal and partnering up with Aureus Analytics, Landmark will now be able to bring data from all its business into one place, and build a strong and meaningful customer journey. One of the major highlights is that the solution deployment is on cloud. Aureus will leverage their data analytics platform CRUX to implement Customer OneView, Household analytics and other predictive models that will help Landmark improve the end customer experience. “Landmark Insurance Broker is one of the few companies who have understood the potential of leveraging analytics correctly across their entire business portfolio. It helps that both – Landmark and Aureus – are obsessed with customer experience. This deal is also encouraging as it demonstrates increasing comfort of Indian enterprises with cloud based solutions,” said Ashish Tanna, COO and Co-Founder at Aureus Analytics. So what does this signal for B2B business in India? Usually considered a tough territory to break into for startups, the deal establishes the growing needs and demand for analytics across the spectrum. “We are excited to partner with Aureus Analytics to leverage technology and improve the overall customer experience. We have identified a few high potential use-cases that will help us drive growth at Landmark. The biggest challenge we face today is the inability to build a consolidated and intelligent customer one view across our lines of business. Aureus will help us solve this problem and set the foundation of our future growth,” said Apaar Kasliwal, Executive Director at Landmark. Another major development is that the analytics solutions provider made it to the Plug and Play Tech Center’s Insurtech 12-week accelerator cohort. According to the company statement, Aureus is the only Indian startup to be part of the Insurtech batch. Anurag Shah, CEO and Co-Founder of Aureus Analytics said, “Our selection in this cohort validates our approach of building problem specific analytics products for insurers that specifically address these challenges. These 12 weeks will also help us in engaging in global Insurance thought leaders, which will significantly influence in shaping the future roadmap of our product development.","excerpt":"Aureus Analytics recently made a significant catch by securing a multi-year analytics deal with Landmark Insurance Broker, a prominent insurance broker. As part of the deal, Aureus Analytics will help Landmark Insurance Broker improve their end customer experience by leveraging data analytics and machine learning. Under the ambit of the deal, Aureus will help Landmark […]","categories":["AI News"],"tags":["big data processing interview","Gen AI in Insurance"],"author_name":"Richa Bhatia","publish_date":"2017-04-18T03:07:15","publication_year":"2017","word_count":566,"keywords":["big data","API","machine learning","AI","Azure","R","big data processing interview","RAG","analytics","Gen AI in Insurance","predictive analytics","analytics platform"],"extracted_tech_keywords":["AI","machine learning","analytics","RAG","predictive analytics","Azure","R","API","big data","analytics platform"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/aureus-analytics-inks-multi-year-analytics-deal-landmark-insurance-broker\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":48222,"title":"Deep Dive: What Makes Abzooba the Go-To AI Service Provider?","content":"AI has seen explosive growth over the years, well-substantiated by various statistics. Gartner’s 2019 CIO survey shows enterprise AI implementation across 89 countries surging by 270% in 2015-2019 and tripling in 2018 itself! McKinsey predicts AI and ML’s potential to create $4.6T by 2020 in marketing, sales, manufacturing and supply chain planning while IDC foresees worldwide spending on cognitive and AI systems reaching $77.6B in 2022. Abzooba is an end-to-end AI solution provider which combines the power of their flagship framework with a decade of experience. Abzooba partners with enterprises towards a successful AI transformation journey. In this edition of Deep Dive, Analytics India spoke to Joydeep Chatterjee, AVP, Abzooba who discussed what sets Abzooba apart. Why Abzooba: What Sets Abzooba Apart? “Since 2016, we’ve doubled our revenues every year,” said Joydeep. This echoes findings from Accenture TechVision Survey that AI is the key for businesses to stay relevant. Abzooba’s rise as the go-to AI service provider from its inception as an analytics provider (also the Startup of the Year 2014 – Product Development award recipient from Silicon India) has been phenomenal. Witnessing a YoY growth of 100% since 2017, Abzooba has stood out and upheld the trust of its worldwide marquee customers (CMMI-5 and Fortune 500 companies) from BFSI, FMCG, retail and healthcare. Abzooba has also been invited to numerous global conferences and meets, and won the prestigious Great Learning Data Science award (Cypher-2019 conference). Abzooba featured regularly in Analytics India magazine (on a top ten list) and on The Centre for Internet and Society, India report. Recently, Abzooba also becomes the official AI partner of IIT Bombay Tech Fest, 2019 – Asia’s Largest Science and Technology Festival. “If its AI, it’s Abzooba. They exceeded our expectations dramatically – from their attitude to their knowledge, Abzooba has been a revelation” — Rob Edwards, Anthem, US Healthcare Insurance Nurturing Talent: Focus of Abzooba’s Roadmap Abzooba believes that its people are its key capital. An approach that draws from a combination of humans and machines will almost always be substantially better than the most intelligent human or the most sophisticated algorithm. With this in mind and has been at the forefront of the AI revolution, Abzooba has partnered with IITs, IIMs, ISIs and the best technology schools in India to incubate the best-fit talent. In a dynamic world of AI, keeping pace with the latest academic research is one of the key drivers to success.  In the words of Abzooba’s Chief Scientific Officer, Dr. Arnab Bose, “While we become more people-centric (both within the organization and while understanding human needs that require solutions), we also translate our knowledge from our exposure to clients, conferences, meets and events through an extensive 360° training enabled by our L&D department.” “The training encompasses all aspects important to business including technology, delivery methodologies and best practices, organizational behaviour, among others. We have a dedicated training academy — Abzooba Innovation Academy (AIA) to nurture our talent pool and future leaders,” added Joydeep. How We Do It: Accelerating the Cognitive Journey of Enterprises xpresso.ai, Abzooba’s homegrown AI lifecycle management software is an Integrated Development, Deployment and Management Environment (IDDME) and a key enabler for data scientists looking to develop, deploy and monitor AI projects with high flexibility and ease. The above is achieved through a common, high-availability environment that includes the best-in-class infrastructure, open-source tools, libraries, packages and automated processes in line with the industry best-practices. “Our business model is based on developing and deploying enterprise AI applications at a fraction of the cost charged for similar engagements by other organizations and within timelines that are much shorter than the industry standard. In our approach, we help organizations attain self-reliance, sustainability and choice of the best-fit solution and business enablers,” said Joydeep. “Abzooba’s contribution in Sprint Engagement on Hadoop, Unix, Shell Script, Hive and Oozie are fantastic.” — Arijit Sarkar, Ericsson Abzooba’s AI partnership with enterprises comprises three parts. The first part begins with a robust data intake and elaborate analysis to formulate the problem statement.  The second part includes applying an engineering mindset combined with subject matter expertise to identify enablers required for AI initiatives to be successful at the enterprise-scale and prepare cognitive models. In the last part, the cognitive models are deployed, managed and monitored in production at an enterprise scale. Throughout the project lifecycle, Abzooba uses xpresso.ai accelerators to systematically deliver AI transformations at an enterprise scale. So what would have typically taken months to deliver can be developed and deployed in weeks. Coupled with a strong monitoring system this translates into robust AI solutions delivery to satisfied customers. “This gives us a significant competitive advantage over our competition. There are very few companies which offer a truly enterprise-grade AI Lifecycle Management platform – xpresso.ai provides a complete IDDME (Integrated Development Deployment and Management Environment) and this is our unique differentiator,” said Joydeep.","excerpt":"AI has seen explosive growth over the years, well-substantiated by various statistics. Gartner’s 2019 CIO survey shows enterprise AI implementation across 89 countries surging by 270% in 2015-2019 and tripling in 2018 itself! McKinsey predicts AI and ML’s potential to create $4.6T by 2020 in marketing, sales, manufacturing and supply chain planning while IDC foresees […]","categories":["IT Services"],"tags":["Deep Dive"],"author_name":"Vishal Chawla","publish_date":"2019-10-18T13:20:53","publication_year":"2019","word_count":810,"keywords":["data science","Go","API","AWS","AI","ML","analytics","Rust","GAN","R","Deep Dive"],"extracted_tech_keywords":["AI","ML","data science","analytics","AWS","R","Go","Rust","API","GAN"],"url":"https:\/\/analyticsindiamag.com\/it-services\/deep-dive-what-makes-abzooba-the-go-to-ai-service-provider\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":6,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":53911,"title":"Meet Xenobots — World’s First ‘Living’ Machines Created Using Frog Stem Cells","content":"How does an organism evolve? One might quickly quote Darwin and throw all-encompassing words like “universe” and “nature”. No disrespect to the great evolutionist, but unlike in former times, humans are now capable of forcing evolution onto other beings. For instance, researchers from the University of Vermont and the Truft University have forced integrated growing cells from frog embryos into new organisms, by using artificial intelligence algorithms. They are called Xenobots — living machines, which are only 0.04 inches (1 mm) wide. Meet Xenobots These Xenobots are named after African clawed frog — Xenopus laevis, from where the stem cells were obtained. These frogs neither have any resemblance with the amphibian nor do they look as creepy as the water bears. These organisms can move independently and have the ability to self-heal any damage they suffer. During the experiment carried out by Michael Levin, Joshua Bongard, Sam Kriegman and Douglas Blackiston, these Xenobots were able to cut and stitch themselves back in order to sustain weeks in a petri dish where they were stored. How Are They Created? Bongard, one of the co-researchers and a computer scientist and robotics expert at the Unversity of Vermont, said, “They’re neither a traditional robot nor a known species of animal. In fact, its a new class of artefact — a living, programmable organism.” The Xenobots consist of nothing except the skin and heart stem cells unlike Water Bears, who show complex animal-like behaviour consisting more than just skin and cells. Researcher Sam Kriegman, additionally said that these Xenobots have evolved because of the algorithms, which helped in growing the skin and stem cells into tissue clumps of hundreds of cells. These cells then moved in pulses generated by heart muscle tissue. Xenobots aren’t something which can be controlled and do not come with remote control and therefore are entirely autonomous, where one just has to wind them and leave them to be. “We cut the living robot almost in half, and its cells automatically zippered its body back up,” said Kriegman. When Xenobots were observed in a petri dish under the microscope, these fast-moving bits have shown unique cell structure and arrangements, and with all these collected data, a simulated environment where digital versions of these Xenobots was constructed. Upon running the algorithm, mimicking natural selection, it produced generations of these Xenobots. The best performers among the tiny organisms would be the only ones to get reproduced, and others were just deleted. The breakthrough came when the algorithm started producing designs which could be transferred to the cells. The authors then brought these designs into reality by combining the stem cells to become 3D shapes designed by the evolution algorithm. These skins which held the Xenobots together along with heart tissues propelled the blobs throughout the petri dish for days and sometimes weeks without additional nutrients given to them. Why Xenobots? All the technologies that are built up of materials like steel, plastic and chemicals degrade over time and are very unfriendly towards the environment. However, the whole point of this study was to design an utterly biological machine, which followed a natural selection process, and automatically designs new machines in simulation. The best designs are built by combining biological tissues, which ensured that these machines could safely deliver drugs inside a human body and will not create more hassles for the environmental. Outlook Though this research still has some steps in the pipeline, it has broadened human’s ability to create novel lifeforms. But, creating life or life-like things aka bots can have legal implications, and are sometimes believed to have gone against the laws of nature, because of their artificiality. With scientists creating life, one doesn’t have many ideas on how the things going to turn out as, and whether they can be controlled like other machines. Although the first batch of these bots is very basic, however over time the advanced ones will get trained to react to their surroundings which could create issues for nature. Nonetheless, no matter what results this study would bring in, but this indeed has given literal meaning to the phrase ‘artificial intelligence.’","excerpt":"How does an organism evolve? One might quickly quote Darwin and throw all-encompassing words like “universe” and “nature”. No disrespect to the great evolutionist, but unlike in former times, humans are now capable of forcing evolution onto other beings. For instance, researchers from the University of Vermont and the Truft University have forced integrated growing […]","categories":["AI Features"],"tags":["evolutionary algorithm"],"author_name":"Sameer Balaganur","publish_date":"2020-01-15T17:11:07","publication_year":"2020","word_count":687,"keywords":["Go","artificial intelligence","AWS","AI","cloud_platforms:AWS","programming_languages:R","programming_languages:Go","Git","GAN","evolutionary algorithm","R"],"extracted_tech_keywords":["AI","artificial intelligence","AWS","R","Go","Git","GAN","cloud_platforms:AWS","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/meet-xenobots-world-first-living-machines-created-frog-stem-cells\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10142625,"title":"OpenAI to Soon Unveil ChatGPT Projects","content":"Yet again, a source on X has spotted another feature that is all set to be released in ChatGPT. Codenamed ‘Snorlax’, the projects feature lets you keep your chats, files, and custom instructions in a unified location, helping one use ChatGPT in an organised fashion. The projects are built upon custom instructions, and any communication within that will be tailored to the instructions. The feature is also reported to have a word limit of 8,000 characters. Moreover, you can add documents, code files, images, and more to allow the tool to access the contents. When released, the feature will be positioned directly against Claude’s Projects feature, which lets you store and manage knowledge in the form of chats, files, and documents in one place. X user Tibor Blaho is often quick to spot features that have yet to be officially announced. He also spotted the features announced yesterday in the Day 1 event from OpenAI. Yesterday, OpenAI released the full version of the o1 reasoning model, along with a new pro mode that allows the usage of more compute to solve some of the harder requests from the user. The o1 Pro Mode, along with unlimited access to all of OpenAI’s models are packaged in a new $200 monthly plan called ChatGPT Pro. That said, users on X were also able to spot a GPT 4.5 model in preview, for the ChatGPT teams plan. While it wasn’t released in yesterday’s event, it may well launch soon in the coming few days. With eleven more days to go in OpenAI’s feature release spree, fans can expect plenty. At AIM, we’re speculating a bunch of new additions that OpenAI may release. What about a new, and powerful frontier model? Earlier this year, Sam Altman spoke about what to expect in GPT 5, in an interview with Bill Gates. Altman said. “GPT-4 can reason in only extremely limited ways. Reliability is also a concern. If you ask GPT-4 most questions 10,000 times, one of those answers might be pretty good, but it doesn’t always know which one. You’d want the best response from those 10,000 each time.”","excerpt":"Is this what OpenAI is set to launch on the second day of 12 Days of OpenAI?","categories":["AI News"],"tags":["AI (Artificial Intelligence)","OpenAI","OpenAI o1"],"author_name":"Supreeth Koundinya","publish_date":"2024-12-06T16:45:26","publication_year":"2024","word_count":354,"keywords":["Go","ChatGPT","OpenAI","AI","llm_models:Claude","GAN","GPT","Aim","OpenAI o1","R","AI (Artificial Intelligence)","llm_models:GPT"],"extracted_tech_keywords":["AI","ChatGPT","OpenAI","Aim","R","Go","GPT","GAN","llm_models:GPT","llm_models:Claude"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/openai-to-soon-unveil-chatgpt-projects\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":14802,"title":"This country just launched a programme to develop 100 AI projects","content":"Yaacob Ibrahim, Minister for Communications and Information for Singapore Artificial Intelligence will touch new grounds in Singapore, as the country launched a national programme last week to build capabilities for AI. Through this move, the government plans to concentrate in the areas of healthcare, finance, and city management solutions. The programme will receive a SGD150 million investment to achieve a target of developing 100 AI projects, which would solve real world problems. National Research Foundation will delimit the funds over five years, into this programme, named AI.SG. AI.SG will be managed by several government agencies, including the Smart Nation and Digital Government Office, Infocomm Media Development Authority, and Integrated Health Information Systems. Yaacob Ibrahim, Minister for Communications and Information, Singapore comments at the tech conference Innovfest, “AI.SG will do three primary things – first, address major challenges that affect society and industry, secondly, invest in deep capabilities to catch the next wave of scientific innovation, and finally, grow AI innovation and adoption within companies.” From a societal perspective, the programme will touch upon some of the most critical challenges, such as healthcare issues with an ageing population, or improving traffic management during peak hour. Furthermore, the program will also invest towards growth of deep technologies, such as next-generation AI systems reflecting human-like learning abilities. Additionally, Ibrahim revealed plans regarding setting up the Singapore Data Science Consortium. This will help the country deepen its research in data science, besides driving industry adoption in this arena. The Consortium will serve as the primary platform Singapore, helping companies access the latest technologies in data science. Besides, the move will link companies with research institutions, so that they can tap on their expertise to create solutions using data science. “AI and data science are key frontier technologies that the Singapore Government will harness and build capabilities in,” concludes Ibrahim.","excerpt":"Artificial Intelligence will touch new grounds in Singapore, as the country launched a national programme last week to build capabilities for AI. Through this move, the government plans to concentrate in the areas of healthcare, finance, and city management solutions. The programme will receive a SGD150 million investment to achieve a target of developing 100 […]","categories":["AI News"],"tags":["AI in finance","Artificial Intelligence India","Data Science","finance India","healthcare India"],"author_name":"Amit Paul Chowdhury","publish_date":"2017-05-08T10:14:19","publication_year":"2017","word_count":305,"keywords":["data science","Go","artificial intelligence","programming_languages:R","AI","finance India","innovation","Artificial Intelligence India","Git","AI in finance","programming_languages:Go","healthcare India","Data Science","R"],"extracted_tech_keywords":["AI","artificial intelligence","data science","R","Go","Git","innovation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/country-just-launched-programme-develop-100-ai-projects\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10072712,"title":"Major announcements made at SIGGRAPH 2022","content":"SIGGRAPH 2022—one of the world’s premier conferences on computer graphics and interactive technologies—concluded its state-of-the-art conference on August 11, 2022. The event witnessed participation from technology innovators from fields of AI, data science, analytics, digital art, and more. It included more than 2,400 contributors from 74 countries, garnering 30,000+ views during the livestream sessions. The hybrid conference was also held in-person in Vancouver after two years of meeting virtually. The conference featured speakers from big tech companies and institutions including MIT’s Sougwen Chung, Pixar’s co-founder Ed Catmull, Stanford University’s Pat Hanrahan, Xbox’s Corporate Vice President Sarah Bond, and Facebook’s Head of Product Ime Archibong. The event showcased about 240+ technical papers related to various advancements and scientific contributions, alongside roundtable discussions, workshops, paper discussions and others. Besides these, metaverse was also discussed at length. The participants got to experience AR, VR and mixed reality innovations. Metaverse workshops presented current and future possibilities in the virtual universe in conversations about interoperability, 3D modelling and other ideas. Metaverse workshops also included topics of how computer graphics deals with human bodily functions like the sense of touch with physical hardware in a spatial environment with programs including Apply Hour, Art Papers, Birds of a Feather, Immersive Pavilion inclusive of Web3D ecosystem and solutions for AR and VR. A session with AWS Innovators included talks and panel discussions from Animal Logic, Bardel Animation, Framestore, FuzzyPixel, Preymaker, Nvidia, Scanline, Pixomondo, Unity and conversations with ‘Love, Death + Robots’ creator and Blur Studios founder Tim Miller. New technologies unveiled A recently launched startup out of London, Move.ai partnered with Nvidia to develop an AI-powered system to shift from traditional motion capture process to high-fidelity mocap that can be created from regular video footage. The new system allows pose estimation tools and difficult-to-observe joints detection which allows the seamless capture of motion displayed in the form of a soccer game footage with several players on the field, all being fully motion captured. For indie developers, a lower-end iPhone mocap software was also announced. Arcturus, a company founded by former employees of Google and Pixar, has collaborated with Dell and HP to create a line of individual products with systems for volumetric video capture and post-processing. The system allows game developers and filmmakers to capture the entire body from 360-degrees and insert it into a 3D environment of games and interactive entertainment. Researchers at University College London created LookOut which is a camera control system designed using consumer-grade video gear mounted with software that allows voice commands to direct gimbal movements. The software also uses advanced computer vision similar to Tesla’s self-driving cars to identify people or objects in front of the camera. Users can define a personalised range of actions for the gimbal to identify and track and also reframe actors and objects in a shot. Expanding on the earlier creation of Hair-GAN—Generative Adversarial Network, a Deep Learning concept—researchers at Kyushu University in Japan have developed a technique to generate CG recreations of hairstyles using a single image. This will enable developers to groom Photoreal CG hair with greater ease. What’s new? Numerous new hardwares were also announced by various big tech companies. For instance, Intel announced new lines of ray-tracing GPUs—Arc Pro A-series, for creator applications like Blender, DaVinci Resolve Studio, Handbrake with AV1 hardware encoding that is suitable for mobile form factors and small desktops and can work for engineering, construction and architectural manufacturing. Nvidia announced a series of technology innovations like the Nvidia Omniverse Avatar Cloud Engine that offers a platform with tools to create AI-powered virtual assistants. The company also stepped into metaverse applications with their new NeuralVBD library—a next generation OpenVBD open-source library. In addition, Nvidia announced their focus on enhancing open-source Universal Scene Description (USD) format to enable smoother function of metaverse applications. Nvidia also announced two new SDKs with Kaolin WISP and NeuralVDB that will enable developers to power neural graphics in creation and animation of 3D objects to enable easier presentation of models and significant reduction in memory footprint. Other highlights SIGGRAPH honoured this year’s contributor with ‘Best of’ from various programmes like— Best Paper Awards Instant Neural Graphics Primitives With a Multiresolution Hash EncodingDeepPhase: Periodic Autoencoders for Learning Motion Phase ManifoldsImage Features Influence Reaction Time: A Learned Probabilistic Perceptual Model for Saccade LatencyCLIPasso: Semantically Aware Object Sketching VR Theatre Best in Show: On the Morning You Wake (to the End of the World) Art Gallery Best in Show: Ray Art Papers Best Art Paper: Traditional African Dances Preservation Using Deep Learning Techiques Electronic Theatre Best in Show: The Seine’s of TearsJury’s Choice: The End of WarBest Student Project: Yallah!Audience Choice: Alternate Mesozoic Emerging Technologies Best in Show: HDR VR Immersive Pavilion Top Selection: Journal of My Journey: Seamless Interaction in Virtuality and Reality With Digital Fabrication and Sensory Feedback Real-time Live! Best in Show: A Showcase of Decima Engine in Horizon Forbidden WestAudience Choice: AI and Physics Assisted Character Pose Authoring","excerpt":"Various big tech companies announced new innovations, software developments and hardwares along with conferences and workshops for AI and Metaverse.","categories":["Global Tech"],"tags":["AMD","ar","AWS","Data Science","Intel","Metaverse","NVIDIA","Omniverse","VR"],"author_name":"Mohit Pandey","publish_date":"2022-08-13T13:00:00","publication_year":"2022","word_count":823,"keywords":["data science","AMD","AWS","Metaverse","AI","ML","virtual assistants","ar","computer vision","VR","Ray","Omniverse","deep learning","analytics","NVIDIA","Data Science","R","Intel"],"extracted_tech_keywords":["AI","ML","deep learning","computer vision","data science","analytics","Ray","virtual assistants","AWS","R"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/major-announcements-made-at-siggraph-2022\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10164201,"title":"India Bans 119 Apps on Google Play Store, Most from China &amp; Hong Kong: Reports","content":"The Indian government has directed the removal of 119 apps from the Google Play Store, with the majority linked to developers based in China and Hong Kong, Moneycontrol reported. As per reports, the order was revealed on February 18 via a disclosure on the Lumen Database. However, as of February 20, only 15 of the 119 apps have been removed, while the rest remained available for download. Some of the banned apps include ChillChat, developed by Singapore-based Mangostar Team; HoneyCam, operated by Australia-based Shellin PTY Ltd; and Blom, developed by China-based ChangApp. The government has invoked Section 69A of the IT Act to issue blocking orders, enabling the restriction of public access to certain online content in the interest of national security, sovereignty, and public order. While no official reasons have been given, past bans have been linked to geopolitical tensions with China and concerns over data security. India initiated a series of app bans, starting in June 2020, when it blocked 59 apps, including ByteDance’s TikTok, Tencent’s WeChat, and Alibaba Group’s UC Browser. This was followed by the prohibition of 47 related or cloning apps on August 10, 2020. Subsequently, on September 1, 2020, an additional 118 apps were banned, with another 43 apps blocked on November 19, 2020. The crackdown continued into 2022, with India banning 54 more Chinese apps on February 14 of that year.","excerpt":"Past blocking orders linked to national security concerns amid rising geopolitical tensions with China.","categories":["AI News"],"tags":["AI (Artificial Intelligence)"],"author_name":"Aditi Suresh","publish_date":"2025-02-20T14:49:24","publication_year":"2025","word_count":228,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/india-bans-119-apps-on-google-play-store-most-from-china-hong-kong-reports\/","complexity_score":3,"technical_depth":5,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":31142,"title":"Soon, Canine Robots May Replace CISF Sniffer Dogs At Airports","content":"Robotic Canine :US Department of Defence Security dogs like German shepherds, Labradors or Belgian Malinois at the airport security may soon be a thing. Now they would have to compete with new age robotic dogs. According to Central Industrial Security Force (CISF) officials, these factory-made robotic canines will have similar capabilities like normal security dogs who not only sniff explosives within the airport region, but can also X-ray scan the luggage of passengers with their eyes. As of now, such robots can be found only in the airports of developed countries like the US, UK, Canada, Japan, Korea etc. for multiple purposes, especially security checks. Robotic Canine : Boston Dynamics “Canine Robots are capable of detecting explosives and can scan the luggage of passengers using an X-Ray system fitted in their eyes.” Apart from canine robots, CISF is also considering the use of CT scan-based screening of hand-luggage, cabin baggage, artificial intelligence, sophisticated explosive detectors and biometric control access across multiple Indian airports. Security checks which are done by Indian agencies mostly depend upon a lot on human resources but CISF, considering a change keeping in view the global threat to the aviation sector with new methods of smuggling, its adapting new technologies to put an end to drug, gold and explosives being carried via air. The possibility of using factory-made came up at the Global Aviation Security Symposium 2018, held at Montreal, Canada. The conference was attended by both CISF DG Rajesh Ranjan and Additional DG MA Ganapathy, who oversees the security at all Indian airports. It’s the first time when CISF attended the global aviation symposium organised by International Civil Aviation Organisation (CAO) last week and had significant bilateral discussions with European Union nations including other ICAO members, said a CISF Official. CISF spokesperson Hemendra Singh said that GASeP (Global Aviation Security Plan) of ICAO, a future aviation security policy and programme framework, was also stressed upon during the symposium. CISF also offered its expertise in protecting critical infrastructure along with the aviation sector from the perspective of both anti-hijacking and counter-terrorism to the other countries through an increased participation in panel discussions and workshops that were conducted by ICAO. “We already have a tie-up with the Transport Security Administration (TSA) of the United States but now we have a capacity building pact with the EU as well, which will help us explore a lot about new technologies and methods of securing our airports,” a CISF official told TOI","excerpt":"Security dogs like German shepherds, Labradors or Belgian Malinois at the airport security may soon be a thing. Now they would have to compete with new age robotic dogs. According to Central Industrial Security Force (CISF) officials, these factory-made robotic canines will have similar capabilities like normal security dogs who not only sniff explosives within […]","categories":["AI News"],"tags":[],"author_name":"Martin F.R.","publish_date":"2018-12-06T12:33:39","publication_year":"2018","word_count":411,"keywords":["Go","artificial intelligence","programming_languages:R","AI","programming_languages:Go","Ray","GAN","R"],"extracted_tech_keywords":["AI","artificial intelligence","Ray","R","Go","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/canine-robots-cisf-sniffer-dogs-airports\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":43567,"title":"Applied Intelligence Challenge 2019: Accenture Ventures Awards 6 Best Indian AI Startups","content":"Accenture (NYSE: ACN) has announced the winners of the second annual Accenture Ventures Challenge in India, which this year recognizes deep tech growth-stage Indian startups with the most innovative business-to-business (B2B) use cases in five categories related to Applied Intelligence: artificial intelligence (AI), data, analytics, automation, and industrial AI. “The Accenture Ventures Challenge helps us recognize the most disruptive technology startups in India and provides them with access to our Global 2000 clients,” said Avnish Sabharwal, managing director for Accenture Ventures and Open Innovation in India and the Middle-East. “This year’s competition scaled new heights with record participation from mature startups focused on Applied Intelligence. We’re excited about the possibilities that this creates for the winners, and for our clients globally.” This year’s challenge drew over 370 entries from across India. Of these, 15 finalists were evaluated by a jury of Accenture executives and industry leaders from Indian businesses, multinational corporations and the venture capital community. The judges selected the following winners based on innovative edge, business value and scalability: MMS.IND – for its GeoMarketeer platform, which delivers geo-location-based intelligence using micro market focused data and analytics functionalities, delivered as a service. Numberz – for its secure, cloud-based insights driven enterprise platform, which helps streamline the accounts receivable process and accelerate collections by using AI, machine learning and robotic process automation. Locobuzz – for its portfolio of software as a service-based offerings including social listening, AI, command centers and customer relationship management, to address core challenges related to customer experience management in the digital world. Strayos – for its 3D visual AI platform for the mining and infrastructure industry. The platform utilizes imagery data to extract hidden insights about geology in mining for optimized drilling and pre-excavation object detection below the ground surface to enable the accurate inspection of job sites. Niramai – for its AI-based advanced thermal imaging solution, Thermalytix, which provides a non-contact, radiation free, affordable and accessible way to detect early stage breast cancer in women of different age groups. Additionally, this year a sixth winner was chosen by the audience at the grand finale event on July 25 in Mumbai. This award was won by CropIn for its platform which uses big data analytics, AI and remote sensing, to enable clients in the agriculture ecosystem to analyze and interpret actionable insights on crops, such as stress levels, yields and cycles based on both historical and real time data. The winners will join Accenture Ventures’ Open Innovation partner program, through which they will be able to collaborate with Accenture and co-create solutions for enterprises across the globe. The winners will also be mentored by Accenture leaders, introduced to top industry forums, and have access to the Accenture Innovation Hub in Bengaluru where they can further scale their solutions The Accenture Ventures Challenge was open to B2B startups in the deep tech space with operations in India, and annual revenue of at least USD 100,000 in the last reported financial year. This year, nearly 50 startups with more than USD one million revenue participated. A key component of the Accenture Innovation Architecture, Accenture Ventures is a bridge to the global innovation ecosystem that unlocks growth opportunities for our clients by bringing them together with best in class emerging enterprise startups. In India, Accenture Ventures tracks more than 1,300 startups and has engaged with nearly 200 that have technologies relevant to Accenture’s enterprise clients.","excerpt":"Accenture (NYSE: ACN) has announced the winners of the second annual Accenture Ventures Challenge in India, which this year recognizes deep tech growth-stage Indian startups with the most innovative business-to-business (B2B) use cases in five categories related to Applied Intelligence: artificial intelligence (AI), data, analytics, automation, and industrial AI. “The Accenture Ventures Challenge helps us recognize the most disruptive […]","categories":["Deep Tech"],"tags":["Accenture","disruptive technology"],"author_name":"Abhijeet Katte","publish_date":"2019-07-30T15:50:03","publication_year":"2019","word_count":565,"keywords":["Accenture","disruptive technology","Go","machine learning","artificial intelligence","AI","ML","Scala","Ray","object detection","analytics","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","analytics","Ray","object detection","R","Go","Scala"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/applied-intelligence-challenge-2019-accenture-ventures-awards-best-indian-ai-startups\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":28546,"title":"The Ins And Outs Of Adopting Machine Learning At A Corporate Level","content":"The interest in Machine Learning can be understood by merely understanding that there is a rise in volumes and varieties of raw data, as well as the various diverse processes, and therefore, there is a requirement to find a reasonable data storage system. The need of the hour is to devise a method by which business enterprises can rapidly and automatically examine bigger, more complex data. Moreover, by applying and integrating Machine Learning in an enterprise, it becomes easier to enhance the process because Machine Learning helps deliver quicker and more precise results. Challenges faced by Organisations while Adopting Machine Learning Inaccessible Data and Sensitive Data Security When an organisation wants to use Machine Learning in their database, they need the presence of raw data, which is difficult to gather. Yet, collecting data is not the only issue. Once an organisation has the data, security is a very important aspect that needs to be taken care of. Segregating sensitive and insensitive data is indispensable for implementing Machine Learning properly and efficiently. Organisations need to store the confidential data by encrypting it and storing it in other servers where the data is completely secured. The less sensitive data can be made available to trustworthy team members. Infrastructure Requirements for Experimentation and Testing According to a study by Machine Learning Mastery, Machine Learning is difficult for a business to implement, basically because the large-scale organisations in India have yet to recognise and understand the benefits that a simple Machine learning algorithm can offer. There is a need for appropriate infrastructure which can help the testing of diverse tools. Frequent tests should also be permissible to develop the desired outcomes, which in turn, can help in creating better results. Organisations can give their data to different enterprises and ask for their response. Then, they can match the results with a different viewpoint and the best one can be adapted accordingly by the company and subsequently, by the board. However, a small section should still be allowed to work on a different mechanism to allow space for innovation and it might help in providing a better result. Inflexible Business Models Machine learning needs a business to be responsive to their policies. Employing Machine Learning efficaciously needs one to change infrastructure, attitude, and also needs proper and appropriate skill-set. However, employing Machine Learning doesn’t guarantee success. Experimentations need to be done if one idea doesn’t work. For this, agile and flexible business functions are critical; organisations also need to spend less time, money and effort on unproductive projects. If one Machine Learning strategy doesn’t work, it enables the enterprise to learn what is vital and thus guides them in building a new and strong Machine Learning design. In conclusion, implementing a Machine Learning method can be really tedious, but can also act as a revenue generator for a company. However, this is only conceivable by implementing Machine Learning in more innovative ways. Corporate training in machine learning from a top training provider can help organisations in upskilling their existing workforce to use machine learning effectively to optimise the business processes.","excerpt":"The interest in Machine Learning can be understood by merely understanding that there is a rise in volumes and varieties of raw data, as well as the various diverse processes, and therefore, there is a requirement to find a reasonable data storage system. The need of the hour is to devise a method by which […]","categories":["AI Features"],"tags":["Machine Learning"],"author_name":"Ashish Trikha","publish_date":"2018-10-09T11:19:52","publication_year":"2018","word_count":515,"keywords":["Go","API","machine learning","AI","innovation","Machine Learning","RAG","ViT","Rust","GAN","R"],"extracted_tech_keywords":["AI","machine learning","RAG","R","Go","Rust","API","GAN","ViT","innovation"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/adopting-machine-learning-at-a-corporate-level\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10162498,"title":"L&amp;T Technology Services Secures $80 Million Digital Engineering Transformation Deal","content":"L&T Technology Services (LTTS) has secured a multi-year $80 million contract with a US-based industrial products manufacturer to drive digital transformation. The agreement focuses on connected products, digital thread integration, and expanding the client’s innovation capabilities. As part of the deal, LTTS will establish a centre of excellence (CoE) in India. The CoE will serve as a global innovation hub and support clients’ digital transformation and product lifecycle management (PLM) efforts. The initiative aims to strengthen the client’s global presence in emerging digital technologies. “By combining AI-powered insights, digitally connected solutions, business-driven automation, and deep engineering expertise, we aim to enable the client to bring high-quality products to market faster,” said Amit Chadha, CEO and MD of LTTS. LTTS, a listed subsidiary of L&T, provides engineering and technology services, including design, development, testing, and sustenance across products and processes. Recently, L&T also announced it will construct Uzbekistan’s first AI-enabled and sustainable 10-megawatt data centre in Tashkent. The project aims to improve the data infrastructure in the region with advanced technology and energy-efficient solutions. It is most likely an order valued between ₹1,000 crore and ₹2,500 crore. The company also made big headlines recently when its chairman, SN Subrahmanyan, popularly known as SNS, broke the internet with his viral video, in which he was seen asking employees to work 90 hours a week, including on Sundays. Notably, this move could only contribute to the company’s attrition rate and staff shortage issues. Moreover, last year, Infosys co-founder Narayana Murthy also kicked up a storm with his “70-hour work week” remark. Commenting on development and nation-building, Murthy said, “India’s work productivity is one of the lowest in the world…My request is that our youngsters must say, ‘This is my country. I’d like to work 70 hours a week’.”","excerpt":"As part of the deal, LTTS will set up a centre of excellence in India for digital services and data management.","categories":["AI News"],"tags":["CoE","digital transformation"],"author_name":"Sanjana Gupta","publish_date":"2025-01-30T09:53:35","publication_year":"2025","word_count":295,"keywords":["CoE","AI","programming_languages:R","innovation","digital transformation","Git","automation","Ray","Aim","ViT","R"],"extracted_tech_keywords":["AI","Aim","Ray","R","Git","ViT","digital transformation","automation","innovation","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/lt-technology-services-secures-80-million-digital-engineering-transformation-deal\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":41854,"title":"How Social Media Firms Are Fighting Fake News, By Acquiring AI-Powered Startups","content":"While technology has brought many interesting developments for the world, social media platforms are under the scanner for propagating fake news, controversies, conspiracy theories, hate messages and continuously affecting numerous lives. All the social media platforms, including Twitter, Facebook, Google and Whatsapp, are struggling to clean up their platforms so that they can keep a check on fake news at some level. The situation worsens in case of elections, where social media is crucial to influence voters. In fact, a survey by Twitter suggests that fake news spread must faster and further on Twitter than true stories. What Have Companies Done In The Past? Many companies have been attempting to overcome the problem of fake news in the last few years. For instance, Microsoft has been working on creating an artificial intelligence-driven system to tackle fake news across all its operations. It introduced a feature called ‘perspectives’ in its search engine Bing to efficiently bring opposing perspectives or both sides of the debate, from reputable sources. Whatsapp had in the recent past offered up to $50,000 for proposals that will look into tackling fake news, amongst other measures. It brought initiatives such as campaigns to stop rumours (Share Joy, Not Rumors campaign), introduced ‘Forwarded’ label allowing users to know if a particular message has been forwarded or is original. Google and YouTube who have been accused of pushing controversial and problematic content to gain more viewership brought in the Google News Initiative in which it would work with media partners, journalists and news organisations to bring accurate news to users. Acquiring AI-Based Startups Is The New Trend To Fight Fake News Facebook has the highest number of fake news surfacing on its platform, with the largest number of voter influencers staging the platform. While in the past it had partnered with the third-party fact-checkers to keep a check and stop disseminating fake news, it doesn’t seem to be a very sustainable solution. Ahead of Karnataka elections, it had tied up with Mumbai-based fact-checking organisation called Boom to deal with fake news. Facebook acquired Bloomsbury AI last year, to bring a very specific kind of talent and expertise. It had said that algorithms struggle to understand the context and pick up on things like hate speech. Although it said that it is getting better at detecting it, they have a long way to go. Acquisition of Bloomsbury will help Facebook to apply the startup’s skills to combat inappropriate and false content on the platform. They would also be using the natural language processing (NLP) technology developed by them to help machines answer questions. Twitter, on the other hand, most recently acquired London-based startup Fabula AI, which is working on a technology to detect fake news. The startup has patented a technology called Geometric Deep Learning that exhibits high success rates when it comes to identifying and spotting online disinformation. It is a novel technique to network-structured data, that can analyse large and complex datasets, describing relations and interactions and extract signals in ways that traditional ML techniques are not capable of doing. While Twitter did not reveal how it plans to use the technology, it is clear that with the acquisition they have the ability to analyse very large and complex datasets, which can help them to improve the health of the conversations. They are looking forward to strengthening their systems and operations while integrating graph deep learning algorithms into their ML platform. In an in-depth analysis of the Fabula AI’s fake news detection capabilities, it was found that it has an accuracy of 93% in detecting fake news. The startup’s deep learning algorithms are capable of learning patterns quite efficiently on complex and distributed data sets such as social networks. The approach for detecting disinformation relies not on algorithms parsing news content and trying to identify malicious content but it looks at how such news or content spreads on social networks, thereby identifying who is spreading it. While it has not been implemented completely by Twitter yet, it brings a strong hope for them to use the technology for good and stop the spreading of misinformation. The Future May Be Better These efforts and development by companies definitely suggest that there is a better future in store when it comes to dealing with fake news but it is still a long way to go. With the huge number of users in social media, it will take some time to get everything on track.","excerpt":"While technology has brought many interesting developments for the world, social media platforms are under the scanner for propagating fake news, controversies, conspiracy theories, hate messages and continuously affecting numerous lives. All the social media platforms, including Twitter, Facebook, Google and Whatsapp, are struggling to clean up their platforms so that they can keep a […]","categories":["AI Startups"],"tags":[],"author_name":"Srishti Deoras","publish_date":"2019-07-05T11:48:05","publication_year":"2019","word_count":743,"keywords":["Go","artificial intelligence","programming_languages:R","AI","ML","NLP","deep learning","GAN","R","startup"],"extracted_tech_keywords":["AI","artificial intelligence","ML","deep learning","NLP","R","Go","GAN","startup","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-startups\/how-social-media-firms-are-fighting-fake-news-by-acquiring-ai-powered-startups\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10111359,"title":"Delayed Approval of AI-Medical Devices Hampers Innovation","content":"While a number of AI-enabled devices are out in the market, and significant developments in drugs and medical devices are also happening, the task of releasing them to the market is no easy feat. Governed by regulatory bodies, any drug or medical device undergoes stringent checks and trials before it hits the market. Recently, the US FDA (Food and Drug Administration) cleared an AI device last week that can assist in detecting all types of skin cancer. The AI-powered device DermaSensor is reported to aid physicians in detecting skin cancer, with its pivotal trial demonstrating a sensitivity of 96% and a 97% probability of accurately identifying a skin lesion as benign. While the product was only recently approved, the approval process for DermaSensor began eight years ago. Obstacles for Path To Launch The U.S. Food and Drug Administration – FDA, which is a federal agency within the Department of Health and Human Services, plays a vital role in regulating the development and marketing of any new medical drug or device, however, the process to get approval from the department is lengthy. The US has one of the most stringent regulations for medical drugs and devices in the world. The average time from FDA application to approval of drugs is 12 years. Furthermore, the projected average cost of bringing a new drug from concept to market surpasses $1 billion. The only biggest exception to the lengthy problem, was witnessed during the pandemic. Pfizer and BioNTech were granted Fast Track Designation for mRNA-based vaccine development, thereby expediting the release process to less than a year. Significant advancements are also being made to smoothen out approvals and upgrades for AI devices. Last year, the FDA proposed a plan that will allow developers of AI-enabled medical devices to automatically update their products that are already being used in the clinic. “While dozens of companies have attempted to address this problem in recent decades, we are honoured to be the first device cleared by the FDA that provides PCPs (Primary Care Physicians) with an automated tool for evaluation of suspicious lesions,” said Cody Simmons, co-founder and CEO of DermaSensor. In India, the Central Drugs Standard Control Organisation (CDSCO) takes around six to nine months for completing any medical device registration, only if a technical presentation or Subject Expert Committee (SEC) review is not required. The registration does not expire as long as the maintenance fee is paid every five years. As of now, there is no separate provision for AI-enabled devices. AI-enabled device for detecting skin cancer AI Assistance Might Transform Oncology While the discussion on AI replacing doctors is far-fetched, the role of AI in diagnosis and assisting doctors are finding large use cases. Traditional methods of skin cancer detection involve dermoscopy where a medical professional physically examines a patient’s skin and lesions. However, this approach has room for human error. The diagnosis accuracy for such screening techniques is estimated at 75-84%. DermaSensor delivers a 97% of accuracy for identifying lesions as benign. The device can detect all three types of skin cancers-melanoma, basal cell carcinoma and squamous cell carcinoma. “We are entering the golden age of predictive and generative artificial intelligence in healthcare, and these capabilities are being paired with novel types of technology, like spectroscopy and genetic sequencing, to optimise disease detection and care,” said Simmons. When brought in contact with the skin, DermaSensor emits light, and then captures the wavelengths of light reflecting off cellular structures beneath the skin’s surface. It then utilises specific algorithms to analyse the reflected light and detect the presence of skin cancer. The algorithms are trained on data of more than 4000 malignant and benign lesions. A number of deep learning-based methods have been proposed to assist dermatologists in accurate diagnosis of skin cancer. Last week, MIT released two AI-powered models, PRISM neural network (PrismNN) and logistic regression (PrismLR), that helps with early detection of pancreatic cancer. The model is reported to identify 35% of cancer cases, while conventional screening methods only result in a 10% identification rate. As the regulatory bodies aim to ensure a safe deployment of AI-enabled devices, delays in approval are likely to persist.","excerpt":"The extended approval time cycle for AI-enabled medical devices, such as DermaSensor for detecting skin cancer, sheds light on the hindrances to innovation","categories":["AI Trends"],"tags":["cancer","diagnosis","FDA","Healthcare Automation","medical","MIT"],"author_name":"Vandana Nair","publish_date":"2024-01-23T18:00:00","publication_year":"2024","word_count":692,"keywords":["Go","diagnosis","artificial intelligence","AI","neural network","MIT","ML","RAG","Healthcare Automation","Aim","deep learning","medical","cancer","FDA","GAN","R"],"extracted_tech_keywords":["AI","artificial intelligence","ML","deep learning","neural network","Aim","RAG","R","Go","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/delayed-approval-of-ai-medical-devices-hampers-innovation\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":66074,"title":"Why Online Education May Not Be A Sustainable Solution In India Despite Free Courses","content":"The Covid-19 pandemic seems to have heralded a new change in education systems across the world. While digital learning has been the modus operandi of professionals looking to upskill in an increasingly automated world, e-learning found a place in the curricula of schools and universities amid the lockdown to ensure academic continuity. With many edtech platforms experiencing a significant upsurge in learners in recent weeks, does that signal a new era for education in India as we know it? Perhaps not, given that merely shifting pedagogic practices online amid prevailing challengings cannot serve as a long-term solution. Co-founder of edtech firm GreyAtom, Shweta Doshi, ably summarises it: “While learning online has become inevitable, we will not be successful until we understand that teaching online doesn’t mean taking the entire classroom on Zoom and continuing with the same delivery approaches. This may be a subtle point, but has deep implications.” What are the challenges stymying a more widespread adoption of online education in India? Let us find out. Producing Content Without Context It is difficult to practice restraint in a wide and expansive canvas like the internet, and this is a trap many edtech companies fall into. Their excessive focus on content, without little attention on context and learner engagement indicates that online learning – as it stands today – cannot by itself offer a fitting alternative to traditional education. “Our problem is not content, but aggregation,” says Doshi. “It is important to filter content through the context of whom we’re trying to connect with and teach. This comprises their pain point, learning style, etc., and these should be understood well before putting together an online learning module,” she adds. This is where mentoring plays a key role – a feature that many online courses have begun to integrate. Not only do they add context to learning, but mentors can also act as a guide to students and professionals as they navigate the dynamics of education in today’s world. “If they hit a roadblock, a mentor belonging to an industry the learner wants to transition into must be available,” says Doshi. “If online experiences are delivered with high engagement from both mentors and learners, they can definitely create similar outcomes as physical learning,” she adds. ALSO READ: How This Edtech Startup Is Delivering Cognitive  Learning Through AI & ML Concurs Ashutosh Kumar, co-founder of Testbook.com,” Although online courses afford various means to make learning just as engaging and interactive, a lack of personal interaction and assistance in e-learning can be a challenge and may come in the way of completion of goals,” he says. A way to ensure that learners are making the most of online learning is by assessing their engagement quotient. Engagement gives an indication of the time, energy and resources learners spend, and thus, can be a good metric to measure their overall learning. “If we are continuously able to monitor and measure a learner’s adherence to the tasks assigned, their interactions with other learners on online channels, their connect with mentors in sessions, etc. via tools such as CleverTap\/Netcore’s Smartech integrated with one’s LMS, a significant impact can be made,” feels Doshi. Socio-Economic Challenges To Online Education Despite BARC India’s report stating a 30% rise in the Indian edtech market, the challenges plaguing online learning remains consistent. Elaborates Ankush Singla, co-founder of Coding Ninjas: “Despite quality content and native advertising tools, the edtech space is multiplying by numbers, but not in the market size in tier II and tier III cities,” he says. “The reason is not the content by itself, but socio-economic challenges, which include lack of infrastructure and deployment of internet facilities, leading to limited or no internet connectivity for learners,” he adds. What is more, the gap for new users who have little to zero understanding of technological devices and methods leaves many hamstrung. In addition to this, lack of digital literacy and knowledge on how to navigate digital payment methods to subscribe to online programs is another challenge. ALSO READ: Opportunities & Challenges Of Conducting Exams Through AI-Based Proctoring In India Although e-learning can potentially enable many useful life skills, the lockdown has exposed the digital divide in Indian society. And not just students, many teachers are also facing the brunt of this and struggling to effectively relay information to their students, who are also grasping at straws to understand. “In a developing country like India, relevant infrastructure to support a total overhaul of education online is inaccessible,” says Randhir Kumar, founder of BasicFirst Learning. “In addition to lack of proper electricity and internet connectivity, many do not have the financial resources to invest in expensive tablets, laptops or PCs and are eventually left on their own to navigate through complex subjects and topics,” he adds. Outlook Though edtech is one of the flourishing domains in India, the full spectrum of ecosystem around digital content creation, delivery and consumptions still need to be matured in India. “Teachers still need to be trained in online learning to get accustomed and comfortable in creating and delivering digital content,” says Ashutosh. “Also, in addition to user-friendly product experience, personalised interaction through mentorship programs should also be included to make it more holistic,” he adds.","excerpt":"The Covid-19 pandemic seems to have heralded a new change in education systems across the world. While digital learning has been the modus operandi of professionals looking to upskill in an increasingly automated world, e-learning found a place in the curricula of schools and universities amid the lockdown to ensure academic continuity. With many edtech […]","categories":["AI Trends"],"tags":["online education"],"author_name":"Anu Thomas","publish_date":"2020-06-01T10:00:00","publication_year":"2020","word_count":868,"keywords":["Go","AWS","AI","cloud_platforms:AWS","programming_languages:R","ML","Git","online education","ViT","R","startup"],"extracted_tech_keywords":["AI","ML","AWS","R","Go","Git","ViT","startup","cloud_platforms:AWS","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/why-online-education-may-not-be-a-sustainable-solution-in-india-despite-free-courses\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":43625,"title":"6 Challenges In Using Open Source Cybersecurity Tools","content":"When it comes to cybersecurity, tools and infrastructure matter a lot in order to battle notorious threats. Companies across the world have of late understood the importance of having strong cybersecurity and are trying every possible tool or software to make it better. There are two types of tools — open-sourced and closed-sourced. While most of the companies have been using closed sourced security tools, open-source tools today have also started to gain significant attention and usage. Companies are leveraging open-source productivity software, tools for administrators and developers, and even code libraries that they use to build their own software. The main reasons behind this popularity of open source tools is due to: Huge community Source code is publicly available It is customisable Companies off late have realised that even open source is not completely safe — it has its own set of challenges. The open-source cybersecurity tool is like a double-edged sword. In this article, we are going to look at some of the challenges that open source cybersecurity tools pose. The Risk Of Noxious Code Every time we talk about open source, it is the first thing that comes to our head. This problem of poisoning a code is related to the open-source nature of making the code available to the world. Even though the moto behind this act is to help fix a bug as soon as possible, this very nature can also be used the other way around. Black hats can also take a look at that code and use the publicity of the exploits and bugs to their advantage. They can then target companies with the same piece of software and who have not patched the bug. They just need to gain access to the software or tool and manipulate it or they can just make use of the exploit that was available. One of the examples of this kind of hack is the 2017 Equifax data breach where personal details of 143 million people were leaked. The root cause of this breach was the high-risk vulnerability that was there in the company’s open-source Apache Struts framework. Challenge With Support Open source mostly depends on the community for any sort of bug or vulnerability fixes. Even though the community is huge and is one of the major factors behind the open source’s success, this may sometimes be challenging for companies using open source. While most of the open-source software providers have a dedicated support team, there are many providers who sometimes (when they fail to figure out and solve an issue) pass any kind of query to the community. Meaning, the user will have to post his\/her query on forums and message boards to get their problems resolved. And the response and problem fixing might take time. Untrusted Source One of the reasons for opting to open source is the ease of getting and using open source. However, this same advantage might sometimes cause sheer headache to a company. When the cybersecurity department is given the power to decide on the tools and software, the department sometimes ends up picking up open-source tools from sources that are not completely trusted. As a result, many open source tools turn out to be vulnerable to different kinds of attacks and some are even developed in such a way that finds a bug also gets difficult. Therefore, make sure your organisation has strong policies the source and type of open source tools. Help Attackers Take Different Approach Cybersecurity today is just like a cat and dog fight. Every time cybersecurity professionals take a strong step in preventing cyber-attacks, hackers come up with a whole new method and try again. It is basically a trial and error method — for both white and black hats. However, when an organisation uses open sources, the scenario gets a little different. Hackers that are launching attacks on organisation backed by open-source are first evaluating their cybersecurity tools —attackers can audit the code themselves and look for weak points. And they are getting this power as the source code is open. This is without a doubt, one of the biggest challenges for companies leveraging open source. And if companies in this aspect, the chances are high that they would soon get pwned. Challenge With Updates Another challenge is with updates. Companies using open source cybersecurity tools need to manually keep checking for updates for different components. And if not updated on time, those outdated components could pose operational inefficiencies. That is not all, those components might also pose high-risk cybersecurity vulnerabilities as attackers up their hacking game every now and then. Therefore, it is advised not to delay updates and have an active eye on each and every component. Insider Threat This is one of the most likely events that companies face every now and then. When companies leverage open source software, they must keep in mind that they are giving the developers access to the source code. Humans are considered to be the weakest links in cybersecurity. If social engineered or manipulated, developers might end up making changes to the source code, opening doors for hackers to exploit. In order to cope with this challenge, companies need to do a routine check on the things developers work on.","excerpt":"When it comes to cybersecurity, tools and infrastructure matter a lot in order to battle notorious threats. Companies across the world have of late understood the importance of having strong cybersecurity and are trying every possible tool or software to make it better. There are two types of tools — open-sourced and closed-sourced. While most […]","categories":["AI Trends"],"tags":["Cybersecurity","cybersecurity threats","open-source software"],"author_name":"Harshajit Sarmah","publish_date":"2019-07-31T17:45:30","publication_year":"2019","word_count":876,"keywords":["Go","cybersecurity threats","programming_languages:R","AI","open-source software","programming_languages:Go","RAG","ViT","Rust","GAN","Cybersecurity","R","programming_languages:Rust"],"extracted_tech_keywords":["AI","RAG","R","Go","Rust","GAN","ViT","programming_languages:R","programming_languages:Go","programming_languages:Rust"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/6-challenges-in-using-open-source-cybersecurity-tools\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10112949,"title":"Jensen Huang &amp; His Newfound Obsession","content":"Jensen Huang, the CEO of the world’s third-most valued company, has been on a side quest to spread one meaningful message: Every nation, regardless of its size or resources, must develop its own ‘sovereign AI’. The man in black has travelled over a dozen geographies, right from the Indian subcontinent to his latest stop, the UAE, for the recent World Government Summit in Dubai. As Huang envisions it, sovereign AI isn’t just about having some fancy algorithms humming away in a server room. It’s about a fundamental shift in power, reclaiming national autonomy in generative AI. It’s about ensuring that the decisions made by AI, which will increasingly impact everything from healthcare to defence, reflect the priorities of each individual nation. Huang is basically helping countries take control of how they can shape and build AI to fit their needs. To make sure that when AI makes decisions about things like healthcare and defence, those decisions match the values and priorities of each country. He believes AI could change not just how we live but also the character of different nations. Money Matters The trillion-dollar company offering graphic and digital media processors has clearly outlined plans to invest in and scale the operations of the countries Huang is visiting. Last year, NVIDIA signed several deals with those states. In India, the company has joined hands with national conglomerates Reliance and Tata Group to build AI computing infrastructure and platforms powerful than the fastest supercomputer in India today. Interestingly, both the partnerships were announced on the same day. Reliance will be provided with the latest tech to further develop the local infrastructure. On the other hand, TCS will upskill its 600,000-strong workforce apart from the technical aspect of the partnership. Huang’s efforts show that the AI visionary wants to help countries worldwide develop AI at the grassroot levels. Even when he met Omar Al Olama, the Middle Eastern state’s minister of AI, Huang focused on explaining how AI can be trained on local data to protect cultural identities. “It’s not that costly; it is also not that hard,” Huang said. “The first thing that I would do, of course, is codify the language – the data of your culture into your own large language model.” Over the past six months, the chief of the most powerful company in tech right now has also visited Canada, France, Japan, Malaysia, Singapore and Vietnam distributing similar wise words. By building locally (with NVIDIA), these nations can boost their economies while ensuring national security. Empowering Competition Huang’s message resonates with world leaders since large language models and other recent AI developments have kickstarted a competition not just among tech companies, but also nations. China has been ambitiously open-sourcing its contenders, India is going all in, and UAE is betting big on generative AI, too. Other leaders in the field, too, are noticing the stark difference between generative AI and the technologies that prevailed earlier. “The thing that’s different this time”, continued Sri Elaprolu, global head of AWS Generative AI Innovation Center, “is that it has started to be hot across and not just one active geography. We’re supporting customers across all areas, including Latin America, Africa, and the Middle East — locations we normally don’t see a jump right into emerging tech.” From nations large to small, Huang, the Midas of AI, is making sure no one is left untouched by his golden touch. His call is not merely a suggestion, but has a greater purpose of building technology from scratch without anyone being left behind. We have seen technologies in the past get easily monopolised, causing economic disparity and even putting the security of nations at risk. But Huang wants to lay a better foundation for generative AI and is clearly (and literally) going miles to do so. Where do you think Huang is headed next to spread the word about sovereign AI?","excerpt":"Where do you think is the NVIDIA CEO headed next to spread the word about sovereign AI?","categories":["AI Features"],"tags":[],"author_name":"Tasmia Ansari","publish_date":"2024-02-15T12:00:00","publication_year":"2024","word_count":649,"keywords":["Go","AWS","AI","cloud_platforms:AWS","innovation","programming_languages:R","Git","Aim","generative AI","R"],"extracted_tech_keywords":["AI","generative AI","Aim","AWS","R","Go","Git","innovation","cloud_platforms:AWS","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/jensen-huang-his-newfound-obsession\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":33483,"title":"Kerala Pushes For Cutting-Edge R&#038;D, Launches India’s Largest Startup Hub In Kochi","content":"Image source: www.facebook.com\/keralastartupmission Giving a healthy boost to its startup ecosystem, Kerala on 13 January, inaugurated Integrated Startup Complex, India’s largest startup hub with an aim to provide adequate infrastructure for startup incubation and acceleration within the state. The eight-floor facility which is spread across 1.82 lakh sqft area will house incubators for research and development purposes across various sectors. Located at Kalamassery in Kochi, the facility is established by Kerala Startup Mission, the state government’s nodal agency to bolster entrepreneurship and incubation among students and graduates alike. It currently houses Bionest to promote medical technologies, BRINC, India’ first international accelerator for hardware startups, BRIC, an incubator facility for cancer research and Maker Village for electronics incubation. It also has an exclusive Centre of Excellence set up by industry majors such as UNITY and CERA. “Kerala Government is working for the state to have a total area of 2.3 crore sq ft of IT space (up from 1.3 crore sq ft last year). The opening of the new complex at the Technology Innovation Zone (TIZ) is a major step towards achieving the objective,’’ Pinarayi Vijayan, Kerala CM said in his speech at KINFRA Hi-Tech Park, Kalamassery. AI And Robotics For Cancer Research At Biomedical Research, Innovation and Commercialisation in Cancer (BRIC), which is India’s first ever cancer technology incubator,  a great push will be given for leveraging technologies like Artificial Intelligence (AI) and robotics for early detection and cancer care. The centre has already received $1 million in funding from the Bill & Melinda Gates Foundation. Speaking on the matter to a national daily,  Dr Moni Abraham Kuriakose, Director, Cochin Cancer Research Centre (CCRC), said, “We’re developing a homegrown technology for identifying mouth cancer at its very earliest stage using artificial intelligence. Bill & Melinda Gates Foundation is giving the money because it’s a socially impacting project.” For more nuanced and cutting edge results, the centre has partnered with the University of Illinois for collaborative research in robotics and cancer surgery. Advanced robotics will also be leveraged to aid surgeons. More Direct Jobs With the establishment of the centre, the government hopes to provide direct jobs to 2.5 lakh in IT.  The government is also hoping to support startups directly by the means of public procurement and international exposure. The time frame for the accelerator is six months while for incubation it is 18 months, but depending on the age of the start-up it can be extended up to 24 months, an online portal noted. The facility will also provide their service at a cheaper rate on per seat basis. For one seat, it is ₹2,000 which include access to the internet as well.","excerpt":"Giving a healthy boost to its startup ecosystem, Kerala on 13 January, inaugurated Integrated Startup Complex, India’s largest startup hub with an aim to provide adequate infrastructure for startup incubation and acceleration within the state. The eight-floor facility which is spread across 1.82 lakh sqft area will house incubators for research and development purposes across […]","categories":["AI News"],"tags":["AI and robotics","research and development"],"author_name":"Akshaya Asokan","publish_date":"2019-01-15T06:33:43","publication_year":"2019","word_count":444,"keywords":["Go","funding","artificial intelligence","AI","research and development","innovation","RAG","AI and robotics","Ray","Aim","R","startup"],"extracted_tech_keywords":["AI","artificial intelligence","Aim","Ray","RAG","R","Go","innovation","startup","funding"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/kerala-pushes-for-cutting-edge-rd-launches-indias-largest-startup-hub-in-kochi\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":24692,"title":"Organizations Continue To Face Challenges With Big Data: Lets Deep Dive","content":"Since early 2014, we have seen a steep increase in the number of companies adopting big data and analytics. Data is proliferating in every part of businesses and organisations, and they are more focused on reaping the benefits of big data. While big data has played a major role in overall growth of organisations, it is not the only factor that can drive growth. It comes with its own side effects. Leveraging everything on big data will serve only a limited purpose if an organisation as a whole does not integrate with its goals completely. The organisational mindset should look at a broader horizon apart from relying solely on big data insights. It should also be accepted by all employees at various levels. Challenges From A Management Perspective According to Randy Bean, CEO of NewVantage Partners, the challenges in implementing big data lies in the approach followed by the management of a company. He says, “The vast majority of the challenges companies struggle as they operationalise big data related to people, not technology — issues like organisational alignment, business process and adoption, and change management.” He emphasises on bringing a rigid cultural change in any organisation along with providing opportunities in innovation with this domain. This is where Indian tech companies have to concentrate. Adapting to a new technology will happen gradually but how well it is infused in the corporate culture is a key factor in its growth. Another challenge is the technology disparity present in companies. Although, these days  professionals in India grasp latest tech happenings, the prospect of replacing conventional tech is still a matter to brood. Big data alone is not just sufficient to tackle this issue. Bridging the skills and technology gap is a daunting task. The Human Factor Big data and related technologies won’t help much if human analysis is totally left in the dark. Intelligent machines and automation may provide quicker and accurate results but may not reveal details in hindsight. For example, a family doctor’s diagnosis of patients may actually give out more details to assess health on a long run. This cannot be possible if machines perform diagnosis because they don’t have knowledge of previous data. Big data will be more insightful if human intuition along with social sciences are considered. This is due to the fact there are other factors in account such as human behavior and diverse cultural spread. Privacy Concerns The contrasting side of big data in a company is the possibility of intrusion from other entities to data which are present in large amounts and are passed around the company’s network. Even though data protection offers a certain level of security to counter these bad influences, there may be chances of mismanagement of personal and customer data inside or outside a company. In a paper titled Big Data Challenges by Alexandru Adrian of Romanian-American University, the author highlights the core problem associated with data privacy. He explains, “For companies the privacy issue is more related to the sensitive data that they work with. Whether it is financial data, clients list, perspective projects, all represents valuable data that may or may not be disclosed.” This shows that critical data in large volumes can be detrimental to the organisation, if it falls into the wrong hands. Now, for an Indian ecosystem which spans a massive population of 1.2 billion, the data generated will be humongous. In addition, social media presence is very strong which leads to more personal data in the online space. Big data may deal with all these information efficiently, but privacy will be an illusion. Aberration To Technological Innovation Big data when viewed from an innovation perspective might not necessarily be always useful for companies. The insights gained may not predict a causal relationship between big data and innovation. In a review paper by Sen Chai and Willy Shih, a wonderful supposition is discussed: “Although we acknowledge that data mining has enabled tremendous advances in business intelligence and in the understanding of consumer behavior — think of how Amazon figures out what you might want to buy or how content recommendation engines such as those used by Netflix Inc. work — it’s important for executives who oversee technical disciplines to be thoughtful about how they apply this approach to their areas.” This is where applying only big data does not have a strong impact. Research and innovation go hand in hand, but without depending solely on technology. Conclusion As the article discussed, several pitfalls of big data, where it is made to be the main focus of companies or organisations can be avoided. These problems can be alleviated if the management in an organisation concentrate on all areas equally, without pushing only big data and analytics solutions. This means, all technological and managerial aspects should be aligned with the organisation’s interest.","excerpt":"Since early 2014, we have seen a steep increase in the number of companies adopting big data and analytics. Data is proliferating in every part of businesses and organisations, and they are more focused on reaping the benefits of big data. While big data has played a major role in overall growth of organisations, it […]","categories":["IT Services"],"tags":[],"author_name":"Abhishek Sharma","publish_date":"2018-05-17T11:16:21","publication_year":"2018","word_count":805,"keywords":["big data","Go","API","business intelligence","AI","RAG","automation","analytics","GAN","R"],"extracted_tech_keywords":["AI","analytics","RAG","R","Go","API","big data","GAN","automation","business intelligence"],"url":"https:\/\/analyticsindiamag.com\/it-services\/organizations-continue-to-face-challenges-with-big-data-lets-deep-dive\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10104865,"title":"6 Uber-Cool Smart Glass Launches by Big Tech Companies","content":"With a number of AI wearables announced this year, one category that witnessed a surge of releases and caught our eyes was smart glasses. Several big tech companies have unveiled their lineup of smart glasses, with some of them incorporating AI features. Meta created a buzz a few months ago with the announcement of their new version of their Ray-Ban Meta smart glasses, followed by other big tech companies which did not want to be left behind. Here are some of the biggest tech eyewear releases in 2023. Meta Ray-Ban Smart Glasses Mark Zuckerberg revealed the latest Ray-Ban Meta Smart Glasses during the Meta Connect event in October, this year. This announcement comes two years after the launch of the initial Ray-Ban Stories, which lacked the advanced features present in the newly introduced glasses. With the smart glasses, one can take calls, play music, capture photos, videos and even live stream. Furthermore, the multimodal featured glasses will have Meta AI integrated. Amazon Echo Frames 3rd Gen Amazon Echo Frames are eyewear that allows users to interact via Amazon’s flagship voice assistant, Alexa. In October, the company announced the latest version of these smart glasses capable of running a myriad of tasks, including playing music, hands-free communication, and controlling the smart home via Alexa. Jio Smart Glasses Reliance Jio presented its MR glasses, known as Jio Glass a few months ago. These glasses utilise mixed reality (MR), combining augmented reality and virtual reality. The glasses can be connected to the smartphones via a USB-C cable and features a virtual display of up to 100 inches. The company is set to launch two versions of the glasses to cater to consumers and enterprises. Huawei Eyewear 2 The Huawei Eyewear 2 smart glasses were unveiled in October in Barcelona. These wearables are equipped with built-in speakers, enabling users to enjoy music and make voice calls. In contrast to the initial Huawei Eyewear, the updated model boasts extended battery life, faster charging, and features a lightweight titanium frame. The previous version was released in 2021. Lenovo Legion Glasses Released in October, Legion Glasses are primarily designed as accessories for Lenovo’s Legion Go Windows gaming handheld. However, they are also compatible with any device capable of displaying content via USB-C, including Windows, Mac, and Android devices. They function as wearable displays rather than AR glasses and include features such as Micro-OLED displays and Lenovo’s venture into the consumer glasses market. Xreal Air2 Smart Glasses Xreal Air2, is a screen-reflecting smart glass that emulates a high-definition screen from ten feet away. These AR glasses are compatible with various devices and offer a portable 330-inch screen. With an additional accessory, users can anchor the virtual screen anywhere in their surroundings. Xreal also offers custom prescription lens insert for users who wear glasses.","excerpt":"Several big players have unveiled their lineup of smart glasses this year, with some of them incorporating AI features.","categories":["AI Trends"],"tags":["Alexa","Meta","Meta AI","reliance jio","smart glasses"],"author_name":"Vandana Nair","publish_date":"2023-12-14T14:30:40","publication_year":"2023","word_count":465,"keywords":["Go","Meta AI","Meta","programming_languages:R","AI","Modal","programming_languages:Go","Ray","smart glasses","reliance jio","Alexa","R"],"extracted_tech_keywords":["AI","Meta AI","Ray","R","Go","Modal","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/6-uber-cool-smart-glass-launches-by-big-tech-companies\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":69627,"title":"Will We See More Undergraduate AI Courses In Future?","content":"With IIT Madras launching an online BSc degree in programming, last month, it has opened up a significant opportunity for anyone who passed class XII to pursue data science as a career. This raises an interesting question as to whether or not will we see more undergraduate degrees in artificial intelligence and data science? During the launch, HRD Minister Ramesh Pokhriyal stated that according to data every year approximately 7.5 lakh students from India go out of the country to search for better education after class XII, which in turn act as a brain drain for the country. This initiative is aimed towards bringing quality education and unique courses right from the undergraduate level. So far, the field of data science and AI has always been a favoured choice for professional level courses or postgraduate programmes after pursuing an undergraduate degree in computer science or mathematics. However, the rise in demand for AI professionals has created a buzz around the industry to offer more bachelors programs in AI. In fact, the UK and US have already stolen a march over India in offering these bachelor’s degrees with modules in AI. Carnegie Mellon University has been one of the first ones that started providing undergraduate degrees in artificial intelligence in the US. Alongside, the Chinese Ministry of Education has also initiated setting up AI programs in the country. Nonetheless, last year, India saw a major proliferation of B.Tech degrees in AI, which includes some of the prominent names like IITs, Amity University, Amrita Vishwa Vidyapeetham, to name a few. Indraprastha Institute of Information Technology is another university that has started offering undergraduate courses in artificial intelligence in collaboration with Infosys Foundation. Sharda University is also offering an undergraduate programme in machine learning to provide advanced knowledge on deep learning and artificial intelligence. Having AI and data science in undergraduate degrees can be a positive step in making the next generation employment-ready in this competitive environment. Such an early adoption of data science and AI knowledge in the career will also augment the research and development activities of the country. Pushing young minds towards AI will not only bridge the skills gap of the industry but will also help professionals understand the basic concepts of computer science, linguistics, cognitive psychology before actually diving deep into the sub-discipline of their choice. Here are top reasons as to why we may see more undergraduate AI courses in future. Also Read: Should Indian Universities Offer Undergraduate Degrees In Artificial Intelligence? Undergrad AI Courses Will Help Students Start Early Integrating AI in bachelor degree curriculum will help students understand its disciplines like machine learning, natural language processing, computer vision and robotics at a much early stage of their career, which will allow them to make a strong base and gain more knowledge on all the disciplines of the technology in future. Such opportunities will also provide them with new career avenues and will help them to choose a career path based on their interested specialisation. With professionals opting for advanced AI courses, it restricts them to specific knowledge. However, with undergraduate courses, students would be allowed to choose for professions like ML researchers, AI consultants, etc. at an early stage of their career. With board knowledge of tools and technologies in the undergraduate program, young professionals will have more chances to sharpen their skills in their specialisation program. Alongside, a capstone project during undergraduate will help them get prepared for real challenges of the industry. Also Read: Top AI, IoT Undergrad Courses STEM Students Can Consider After Class XII Increasing AI Undergrad Courses Will Bridge The Skills Gap Data science and AI have been the fastest-growing sectors, which has been predicted to create 11.5 million jobs by 2026. However, the demand for AI experts entirely outstrips the supply. Such a skill gap can be the biggest reason for limited AI adoption in organisations. In fact, according to a report, almost 22% of respondents identified the lack of institutional support and late adoption of AI among professionals to be key reasons for this. Thus, more undergraduate AI programs will bridge the skills gaps of the industry with more young minds to explore their career in AI. Alongside, having AI expertise comes with an expensive cost, thus with more professionals graduating with AI bachelor’s degree will provide these companies to have more options while recruiting these professionals. To shorten this skills gap, many tech firms have started collaborating with ed-tech companies to create a curriculum for undergraduate AI programs with skills and technologies that would be relevant to the industry in real life. Tech giants like Accenture, IBM, Wipro as well as Google have partnered with companies like Udacity, upGrad etc. to provide courses and training to make young professionals ready for the industry. In fact, in recent news, IIT Roorkee has collaborated with WileyNXT to launch an AI course for banking professionals, where the course has been designed to provide more opportunities for participants in the banking sector. Also Read: Is it Time For More Focus On Analytics At Undergraduate Education? Advancement In AI Landscape IBM’s CEO once stated to the media that “every company will become an AI company,” and therefore, having a dedicated undergraduate program for AI knowledge will also foster more research and innovation in the field. This, in turn, will advance the AI landscape altogether. More university programs on AI will also provide more opportunity for professionals to pursue research and academia, contributing more to society, making a meaningful impact. Alongside, more AI integration at the undergraduate level will have a positive ripple effect on PhD courses and research areas, which, in turn, will augment the AI landscape. According to a report, it has been revealed that 77% of businesses agreed that AI technology would be fundamental for their organisation’s competitiveness. Thus it is expected to double up the rate of AI innovation by 2021. A lot of this could be attributed to the increasing number of AI courses in the market. When asked, Rohini Srivathsa, National Technology Officer, Microsoft India said, with every company becoming AI dependent, organisations need to be a fast adopter of this technology, and thus require to build a core team for doing so, which, in turn, would require more number of AI graduates.","excerpt":"With IIT Madras launching an online BSc degree in programming, last month, it has opened up a significant opportunity for anyone who passed class XII to pursue data science as a career. This raises an interesting question as to whether or not will we see more undergraduate degrees in artificial intelligence and data science? During […]","categories":["AI Features"],"tags":["ai certificates","computer vision future"],"author_name":"Sejuti Das","publish_date":"2020-07-13T16:00:00","publication_year":"2020","word_count":1046,"keywords":["ai certificates","data science","artificial intelligence","machine learning","AI","ML","computer vision","Aim","deep learning","analytics","computer vision future","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","computer vision","data science","analytics","Aim","R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/will-we-see-more-undergraduate-ai-courses-in-future\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10142073,"title":"‘A Week at OpenAI Equals to Months in Another Company’","content":"OpenAI’s first and only employee in India, Pragya Misra, shared her insights into her experience at the company. “A week at OpenAI is equal to months in another company,” she said, in the latest episode of What’s the Point? podcast with AIM tech journalist Mohit Pandey, highlighting the incredibly fast pace at which the company operates, constantly shipping updates and iterating on models. Further, she highlighted the dynamic nature of OpenAI’s work environment, which means that their teams are always moving quickly, pushing out new updates, which makes time feel compressed. This pace, she said, is unmatched in many other companies, where shipping updates are typically slower. In May this year, OpenAI appointed Misra as its first employee in India, following her previous roles at Truecaller and Meta (WhatsApp), to engage with India’s regulatory landscape and establish key public policy partnerships amid the country’s evolving AI framework. “It’s been a lot of fun… working from home, and often from a plane,” said Pragya, reflecting on the challenges and excitement, likening her experience to that of a founder in a startup environment. Further, she highlighted the incredible talent density at OpenAI, which motivates her to push the boundaries of her work. Pragya also noted that the company’s fast-moving, agile environment has been key to its success, reinforcing the need to pivot quickly as new models become smarter. Recently, OpenAI introduced Predicted Outputs, which lets users add a ‘prediction string’ of an output of repetitive tasks, drastically improving the latency by 50%.A few weeks ago, OpenAI rolled out the desktop app for ChatGPT on macOS and Windows PCs. “ChatGPT for macOS can now work with apps on your desktop. In this early beta for Plus and Team users, you can let ChatGPT look at coding apps to provide better answers,” mentioned OpenAI in an announcement. The company is also set to offer solid competition to Anthropic’s Computer Use, and Microsoft’s Copilot Vision, with their ‘Project Operator’, set to be launched early next year. Besides this, OpenAI has also introduced several new features, including ChatGPT Search for real-time answers, a canvas interface for writing and coding, Advanced Voice with multiple accents, a cost-effective GPT-4o mini model and more. (see below). OpenAI Loves India Today marks ChatGPT’s second birthday. The entire world is celebrating, especially India, which is the second largest market for them after the USA. “India is just AI-optimist, filled with them,” said Misra, highlighting how developers in India have been using OpenAI’s tools to create innovative solutions across sectors such as healthcare, agriculture, and education. This enthusiasm, she mentioned, makes India a crucial part of OpenAI’s mission, with the company eager to support Indian AI startups through tools, technical support, and credits. As OpenAI looks to expand its team in India, Misra emphasised the need for ‘Patience,’ encouraging aspiring candidates to explore and experiment with OpenAI’s models in the meantime.","excerpt":"As OpenAI looks to expand in India, OpenAI’s first employee in India emphasised the need for ‘patience,’ encouraging aspiring candidates to explore and experiment with OpenAI’s models in the meantime.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","OpenAI"],"author_name":"Supreeth Koundinya","publish_date":"2024-11-30T11:46:59","publication_year":"2024","word_count":479,"keywords":["Anthropic","ChatGPT","Go","TPU","OpenAI","AI","GPT-4o","RAG","Aim","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","GPT-4o","ChatGPT","OpenAI","Anthropic","Aim","RAG","TPU","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/a-week-at-openai-equals-to-months-in-another-company\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10063037,"title":"Volvo Group expands R&#038;D in India; lays foundation for Vehicle TechLab","content":"Jan Gurander, Deputy CEO, Volvo Group, has laid the foundation for the Vehicle TechLab for Volvo Group’s Research and Development operations in India. The establishment includes a host of facilities such as vehicle garages, electrical & electronics lab, AR\/VR lab, and access to proving grounds. Vehicle TechLab can house complete trucks, chassis and aggregates and has various supporting equipment for engineers to test, innovate, validate and experiment their ongoing work – through a set up equipped with driving simulators, test benches, 3D scanners among various other tools and systems. “This Vehicle TechLab is designed as a collaborative virtual workspace – creating a simulated workshop environment – using technologies like virtual reality, human body motion tracking & realistic digital rendering of vehicles that allows Volvo Engineers across the globe to connect & collaborate virtually. This facility will significantly reduce the development times, improve problem solving and offer better insights & speed in building innovative solutions.“, said CR Vishwanath, Vice President, Volvo Group Trucks Technology, India.The group’s R&D operations in India has now become the largest development site outside Sweden. Vehicle TechLab adds to the growing capabilities of Volvo Group’s R&D operations in India. Earlier, in July 2021, Volvo Group launched CampX – Volvo Group’s Global Innovation Arena – in Bengaluru.  CampX in India is already working with close to 70 startups. Volvo Group Trucks Technology in India employs more than 1600 engineers","excerpt":"Volvo Group Trucks Technology in India employs more than 1600 engineers.","categories":["AI News"],"tags":["research and development"],"author_name":"Kartik Wali","publish_date":"2022-03-17T21:40:11","publication_year":"2022","word_count":231,"keywords":["Go","programming_languages:R","AI","research and development","innovation","programming_languages:Go","Git","RAG","R","startup"],"extracted_tech_keywords":["AI","RAG","R","Go","Git","innovation","startup","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/volvo-group-expands-r-lays-foundation-for-vehicle-techlab\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10162063,"title":"70% of Entry-Level Jobs for Women at Risk With Automation, Warns Smriti Irani","content":"At the World Economic Forum annual meeting 2025 at Davos, Union Minister Smriti Irani called for a shift in the global narrative, urging Indians to lead discussions on pressing issues such as inclusion, climate change, and health innovation. She emphasised that India’s demographic dividend holds immense potential, but its success is intrinsically tied to the role of women. “For too long, we have focused on what we can do for India. It’s time we shift the narrative to Indians leading global conversations,” she said. Irani noted that India’s achievements in inclusion are not just for domestic progress but serve as lessons for global institutions like the World Bank and the United Nations. Highlighting economic opportunities, she referred to a $12 trillion global gain by 2025 if gender equality is fully realised. “This isn’t just talk. There’s a number to it. But the question is: how do we reach that number?” she asked, urging policymakers and businesses to channel investments into sectors like health innovation, particularly affordable products for women’s wellbeing. Irani also addressed the growing impact of artificial intelligence (AI) and automation on jobs. “If jobs are automated, 70% of the jobs that’ll go are jobs done by women because most of the jobs for female professionals are at entry level,” she said. She stressed the need for strategies to safeguard the livelihoods of women and harness their potential in a rapidly evolving job market. Looking ahead, Irani noted that AI is set to advance from generative tools, like ChatGPT, to AI agents that could automate complex workflows. “AI is not new—it has been around since the 1950s. However, AI agents will lead to further displacement of talent. Are we ready for it?” she asked, cautioning that low-income economies, with limited internet access, could face significant challenges. As India takes its lessons to the global stage, Irani underscored the importance of preparing for these disruptions. “These are issues that need urgent conversations,” she concluded. On the other hand, Sindhu Gangadharan, managing director of SAP Labs India mentioned at The Grace Hopper Celebration India (GHCI) 2024  that women today make just 25% of AI researchers globally, which is a glaring gap that reflects not only underrepresentation but also the missed opportunities of building more inclusive and equitable systems. However, “Now is the time to change that narrative. It’s time for women to take centre stage in the AI revolution, not as participants, but as leaders,” Gangadharan said.","excerpt":"She emphasised that India’s demographic dividend holds immense potential, but its success is intrinsically tied to the role of women.","categories":["AI News"],"tags":["Women in AI","Women in Tech"],"author_name":"Shalini Mondal","publish_date":"2025-01-23T16:42:29","publication_year":"2025","word_count":406,"keywords":["Go","ChatGPT","API","artificial intelligence","AI","innovation","automation","GPT","Women in Tech","GAN","R","Women in AI"],"extracted_tech_keywords":["AI","artificial intelligence","ChatGPT","R","Go","API","GPT","GAN","automation","innovation"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/smriti-irani-highlights-womens-role-being-tied-to-indias-success\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":46231,"title":"How Government Of India Wants Taxonomy Of Data To Balance Privacy and Innovation","content":"India’s Ministry of Electronics and Information Technology (Meity) recently announced the creation of a new committee of industry leaders to come up with the strategy on how non-personal community data can be used, and how it can be differentiated from personal data which is sensitive. The committee will work to build a data governance framework for non personal community data, and is led by Infosys co founder S Gopalakrishnan, along with other members, including Joint Secretary of the Department for Promotion of Industry and Internal Trade, Nasscom president Debjani Ghosh, National Informatics Centre Director-General Neeta Verma, Avanti Finance CTO Lalitesh Katragadda, Ponnurangam Kumaraguru from IIIT-Hyderabad, Parminder Jeet Singh of IT for Change and MeitY joint secretary Gopalakrishnan S. Why Was The New Committee Created? Earlier, the Government had constituted a Committee of Experts under the Chairmanship of Justice B.N. Srikrishna, former Judge, Supreme Court of India to deliberate on data protection framework for India. The committee had submitted its report and a draft Personal Data Protection Bill in July, 2018, under which it took cognisance of community data as relating to a group dimension of privacy and an extension of data protection framework. The Committee also reported that all such community data is relevant for understanding public behaviour, preferences and making decisions for the benefit of the community. The Committee strongly recommended that the Government should consider a suitable law that would facilitate collective protection of privacy by including a principled basis for according protection to an identifiable community that has contributed to community data. “In addition to the nature of community data there is also a need to recognize the economic dimension of data and suitable taxonomy of data. This could include aggregated data, derived data, anonymous data, e-commerce data, AI training data etc. Access to and control over various kinds of data is critical for economic advantage. It is also seen that privately collected digital data could be a necessary requirement for policy making, governance and public service delivery in many areas. There is a growing trend towards platformisation of the digital economy and in such a platform economy data could play a critical role as a community or public resources,” stated the official memorandum. The government has been working on the Data Protection Bill for a long time and in some cases there may require work to distinguish between personal and non personal data which the government and tech companies can leverage for innovation.  Industry leaders have therefore alerted that the policy around non personal data should be designed carefully. The issue came to light when MeitY invited policy questions on how to govern anonymised and ecommerce data held by big companies such as Uber , Google and Amazon. Now with the creation of a committee focused specifically on non-personal data, it seems the issue can be resolved and more clarity can be outlined so personal and non personal data is not clubbed together. Using Community Data To Benefit The Country Experts say friendly regulation is needed around non personal data to be able to use it for developing analytics and artificial intelligence systems that can benefit the country in the long run. Tech companies like Google and Amazon have made huge efforts to localise users data so as to comply with the regulations. Now, industry leaders say government should bring clarity in terms of how companies can use data so there is no regulatory violations once the data privacy bill goes live. It is also likely that the government will play an important role in overseeing that non personal data is used ethically and companies do not mix them with sensitive data to infringe upon user privacy. To boost digital competitiveness, MeitY is also reportedly considering establishing guidelines that would make large tech companies make non personal data of users available to competing companies. The law, if implemented will ensure that there is a level playing field for companies and big tech companies like Google, Amazon and Facebook don’t have an advantage, said MeitY.  Now that MeitY wants private companies to collect and use data, there has to be a distinct line between the personal and non personal data. The committee designed for this purpose may help the country in this regard with extensive consultation. The inputs from IT industry will bring in a balance between regulatory policy framework to ensure that privacy law does not hinge upon the innovation. Overview We have seen big tech companies in Europe make constant violations of GDPR regulations, after which the companies like Google and Facebook were charged with big fines. With all the controversy around data mismanagement, it seems MeitY wants to ensure that such scenarios don’t occur later on when the privacy bill goes live.  On the other hand, with over a billion citizens, data will be huge resource to help the nation make strides in innovation, particularly in artificial intelligence. By leveraging the data, government along with private companies can find solutions to challenges that lie ahead for India to become a global superpower.","excerpt":"India’s Ministry of Electronics and Information Technology (Meity) recently announced the creation of a new committee of industry leaders to come up with the strategy on how non-personal community data can be used, and how it can be differentiated from personal data which is sensitive. The committee will work to build a data governance framework […]","categories":["AI Features"],"tags":["Data Governance strategy","Digital India","Privacy"],"author_name":"Vishal Chawla","publish_date":"2019-09-23T11:30:54","publication_year":"2019","word_count":838,"keywords":["Go","Privacy","Digital India","artificial intelligence","AI","innovation","Git","RAG","ViT","analytics","data governance","R","Data Governance strategy"],"extracted_tech_keywords":["AI","artificial intelligence","analytics","RAG","R","Go","Git","data governance","ViT","innovation"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/government-of-india-taxonomy-of-data-privacy-innovation\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10034624,"title":"Building a Fault-Tolerant Quantum Computer, The Amazon Way","content":"Recently, a team from AWS Center for Quantum Computing released its first architecture paper that described the process of building a fault-tolerant quantum computer with a novel approach to quantum error correction (QEC). The researchers presented a comprehensive architectural analysis for a fault-tolerant quantum computer based on cat codes concatenated with outer quantum error-correcting codes. The research delved deep into a few scientific details of building larger and more useful quantum computers. Quantum algorithms can resolve complex practical problems in various areas, including medicine discovery, designing new materials, and optimising financial portfolios and logistics processes. Building a fault-tolerant quantum computer is one of the complex scientific challenges of the 21st century. A successful quantum computing architecture has to meet multiple complex criteria. For instance, it must have a threshold error rate achievable by hardware on a large scale, a convenient physical layout, and a low overhead for fault-tolerant algorithms. At present, most quantum computing systems have qubits, the most basic unit of quantum information. However, qubits are highly unstable and tend to collapse from their quantum state in the outside environment. “The quantum algorithms that are known to be useful, those that are likely to have an overwhelming advantage over classical algorithms may require millions as well as billions of quantum gates. However, quantum gates, which are the building blocks of quantum algorithms are prone to errors and such errors can accumulate to spoil the result of complex computations,” the researchers said. The tech behind Recent studies have shown that qubits with highly biased noise are a promising route to fault tolerance, at least when gates that preserve the noise bias can be easily implemented in the architecture. According to the researchers, one possible route to realising such qubits is via two-component cat code, a bosonic qubit encoded in an oscillator mode, subjected to engineered two-photon dissipation or engineered Kerr non-linearity. This engineered interaction heavily suppresses the population transfer between the two constituent coherent states of the cat qubit, causing an effective noise bias towards phase-flip errors on the encoded logical qubits. Therefore, concatenating the cat code with another quantum error-correcting code can be done efficiently by tailoring the outer code to suppress the dominant phase-flip errors. The researchers termed these coding schemes as the concatenated cat codes. The research provides a full-stack analysis of a fault-tolerant quantum architecture based on cat codes concatenated with outer quantum error-correcting codes. The analysis has been broadly classified into three categories: 1) A hardware proposal 2) A physical-layer analysis of gate and measurements errors and 3) A logical-level analysis of memory and computation failure rates. For the architecture, the researchers used a combination of active QEC and passive or autonomous QEC. The active quantum error correction is an approach for reducing gate error rates by redundantly encoding information into a protected qubit using various physical qubits. This allows the researchers to detect and correct errors. It also allows the implementation of gate operations on the encoded qubits in a fault-tolerant way. While the passive or autonomous quantum error correction (QEC) is a contrasting approach, it needs a physical computing system with inherent stability against the errors. The cat qubit encoding can suppress the bit-flip errors, while the remaining errors are taken care of by active quantum error correction built on top of the cat qubits. Contributions The researchers presented a comprehensive architectural analysis for a fault-tolerant quantum computer based on cat codes concatenated with outer quantum error-correcting codes.They proposed a system of acoustic resonators coupled to superconducting circuits with a two-dimensional layout for the physical hardware.Using estimated near-term physical parameters for electro-acoustic systems, the researchers performed a detailed error analysis of measurements and gates, including CNOT and Toffoli gates.The researchers obtained realistic full-resource estimates of the physical error rates and overheads needed to run practical fault-tolerant quantum algorithms. Wrapping up The AWS Center for Quantum Computing paper describes a way to build a large-scale processor based on cat qubits, substantiated by simulations spanning from the component level up to the system level. The researchers found that with around 1,000 superconducting circuit components, one could construct a fault-tolerant quantum computer that can run circuits intractable for classical supercomputers. Also, hardware with 32,000 superconducting circuit components could simulate the Hubbard model in a regime beyond the reach of classical computing.","excerpt":"The active quantum error correction is an approach for reducing gate error rates by redundantly encoding information into a protected qubit using various physical qubits.","categories":["Global Tech"],"tags":["AWS","Quantum Computer"],"author_name":"Ambika Choudhury","publish_date":"2021-04-23T12:00:00","publication_year":"2021","word_count":715,"keywords":["Go","AWS","AI","cloud_platforms:AWS","programming_languages:R","programming_languages:Go","Quantum Computer","llm_models:Bard","R","emerging_tech:quantum computing"],"extracted_tech_keywords":["AI","AWS","R","Go","llm_models:Bard","cloud_platforms:AWS","programming_languages:R","programming_languages:Go","emerging_tech:quantum computing"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/building-a-fault-tolerant-quantum-computer-the-amazon-way\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":4086,"title":"Social Media Analytics as an alternative to Primary Research","content":"One of the most interesting trends that are currently emerging is that of Social Media replacing Primary Research in benefits and usage. There are various use cases that can validate this, but lets first dig into the definitions and usage of Primary research. Primary research involves collecting data about a given subject directly from the real world. Primary research is any type of research where you go out and collect yourself. Examples include surveys, interviews, observations, and focus groups. A typical and simplistic use case would be that of a car company undertaking a survey to identify the attributes that consumer most prefer\/ dislike in its brand of cars. As an alternative to survey, Social media can get rich public chatter on the same topic that marketers and researchers would kill for. People willingly talk about a brand of car and their preferences and other issues about the same. The car company looking to identify right attributes can seemingly get a deep data that an otherwise costly primary research might reap out after weeks or even months of painstaking execution. The richness of actionable insights that can be gleaned from Social Media is almost bewildering. People talk about almost everything, sometimes even most private aspects of their lives. Social Media listening is a five step process: Some of the benefits of social media over Primary research include: A much larger sample size than primary research Lower cost: Primary Research involves significant costs, much higher than gleaning information from Social Media. People talk about much more personally on social media than they would in an interview, survey or focus group. Unaided insights: Primary research follows the format of questions and answer. There might be hidden insights that are not evident to interviewer that can be captured by social media. Speed of execution: Primary research takes weeks, sometimes months to deliver insights. Social media research provides much speedier execution Relevant audience: People talking about a theme on social media are the one’s that are most affected by that topic and thus is the most relevant audience for a research. A primary research might not reach out to all relevant audience. While there are obvious benefits and ‘leg-up’ from traditional primary research, Social media research has its own caveats\/ challenges: Concerns over privacy issues means its not easy for anyone to listen to online chatter that many feel is personal. Yet, proponents believe that if something is on social media, it’s for public consumption and for everyone’s benefit. Extracting the right insight from social media is a science of its own, which requires deep expertise in linguistic and text analytics, more importantly big data issues of velocity, volume and variety. The exact science is yet to evolve before its mainstream and readily available for enterprise consumption.","excerpt":"One of the most interesting trends that are currently emerging is that of Social Media replacing Primary Research in benefits and usage. There are various use cases that can validate this, but lets first dig into the definitions and usage of Primary research. Primary research involves collecting data about a given subject directly from the […]","categories":["IT Services"],"tags":[],"author_name":"Дарья","publish_date":"2013-09-18T18:19:32","publication_year":"2013","word_count":463,"keywords":["big data","Go","programming_languages:R","AI","programming_languages:Go","analytics","R"],"extracted_tech_keywords":["AI","analytics","R","Go","big data","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/social-media-analytics-as-an-alternative-to-primary-research\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":2383,"title":"Pitch &#8216;Analytics in Action&#8217; Conference : Bangalore, 16 Jan 2013","content":"Come January 16, 2013, Pitch in Bangalore is going to host Analytics in Action, a day long conference being held for the first time, roping in marketers, planners and analytics practitioners from retail, telecom, BFSI, FMCG, e-commerce and many other industry verticals to Demystify Analytics.  Pitch Analytics in Action aims to deliberate on aspects relevant to the current Analytics scenario in India with inputs from global practitioners and will also feature an eminent panel of global pioneers who have successfully utilized the power of analytics as Business Analytics is estimated to grow to a US$ 50 billion industry by 2016. Among other discussions on the future of analytics in India, the conference shall tackle agendas related to analytics helping banks improve profitability, change in customer targeting by leveraging analytics, filling the need gap between industry and academics, implication of analytics for effective sales management and evolution of analytics in India. As a part of the different sessions questions will be floated to the analytics generators, users and providers that will throw light into the challenges in analytics from an Indian perspective, roadblocks in gathering and storing relevant data for analytics, in investing in the right tools and formats, resource implementing the analytic insights. As a part of identifying the richness of analytics in media and marketing the conference will look into how analytics implementation leads to a stronger consumer-brand connect and will help marketers see the connect platforms at a media agnostic level aiming to elucidate how smart analytics enhances the effectiveness in marketing and media investment. Not only India, the conference will also look into analytics on a global level and how successful international brands have envisioned better marketing ROI with due diligence on the basis of analytics. The Pitch Analytics in Action conference will be seeing esteemed speakers such as Marcia Tal, Founder Tal Solutions & Former EVP, Decision Management, Citigroup who is the brains behind Citigroup’s Decision Management function and created a scalable, international organization embedded in more than 30 countries. To her credit Marcia has leveraged her vast experience in Risk Management and Marketing to ensure that business decisions were rooted in data driven strategies and analytic disciplines that impact profit and loss. She will be throwing light on how analytics helped in improving profitability of a bank. Pankaj Rai, Director, Dell Global Analytics will also be joining the conference to throw light upon How Dell analytics helps in better understanding Customer Behaviour with his experience with Dell ananlytics since 2005. Mohan Jayaraman, Managing Director, Experian Credit Information Company of India, shall also be speaking in the conference, explaining the future of analytics in India. Experian has more than one lakh clients globally across diverse industries and markets such as American Express, Axis Bank, Barclays, BMW Financial Services, BNP Paribas, Coors Brewers UK, France Telecom, Hilton Hotel, ICICI Bank, National Australia Bank, Royal Mail, UNICEF among others. Focused to provide value added information to organisations and consumers, Experian is into innovative products, services and helping its clients specialising and expanding into new vertical market sectors. Giving his views on how customer targeting is changing in India by leveraging analytics will be Ashok Banerjee, VP of Data Platform and Supply Chain Engineering, Flipkart.com. A must attend for the people running businesses, the Pitch Analytics in Action’s focus is more on the results and application of analytics than the analytics technique itself, on the application of analytics with global brands and national leaders, insights and strategy from varied sectors such as consumer packaged goods, services marketing, retail, e-commerce to marketing automobiles and consumer durables. Clickhere to register  or miss it all.","excerpt":"Come January 16, 2013, Pitch in Bangalore is going to host Analytics in Action, a day long conference being held for the first time, roping in marketers, planners and analytics practitioners from retail, telecom, BFSI, FMCG, e-commerce and many other industry verticals to Demystify Analytics.  Pitch Analytics in Action aims to deliberate on aspects relevant […]","categories":["Deep Tech"],"tags":[],"author_name":"Дарья","publish_date":"2012-12-28T14:18:36","publication_year":"2012","word_count":603,"keywords":["Go","programming_languages:R","AI","Scala","RAG","Aim","analytics","programming_languages:Scala","GAN","R"],"extracted_tech_keywords":["AI","analytics","Aim","RAG","R","Go","Scala","GAN","programming_languages:R","programming_languages:Scala"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/pitch-analytics-in-action-conference-bangalore-16-jan-2013\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10170468,"title":"Mistral’s New Model Is Trained to Solve ‘Real GitHub Issues’","content":"Mistral, the French AI startup, released a new open-source model called Devstral on Wednesday. The model was built in collaboration with All Hands AI, a startup building open-source software development agents. Devstral is trained to excel at coding-related tasks. On the SWE-Bench Verified benchmark, which evaluates AI models on real-world software issues, Devstral scored 46.8%, outperforming other open-source AI models. It outperforms OpenAI’s GPT-4.1 Mini and Claude 3.5 Haiku. Mistral said that the model is trained to solve real GitHub issues and overcomes the challenges of typical large language models, which are excellent at code completion or atomic coding tasks, but struggle to solve real-world engineering problems. Source: Mistral “Devstral is light enough to run on a single RTX 4090 or a Mac with 32GB RAM, making it an ideal choice for local deployment and on-device use,” added Mistral. The model is available for download on HuggingFace, Ollama, Kaggle, Unsloth, LM Studio starting today. Mistral recently announced a strategic partnership with G42, an Abu Dhabi-based technology group. The partnership will focus on co-developing next-generation AI platforms and infrastructure. The partnership encompasses the AI value chain, ranging from training AI models and developing AI agents and infrastructure to creating industry-specific applications throughout Europe, the Middle East, and the Global South. Besides, Mistral will also explore collaborations with the Mohamed bin Zayed University of Artificial Intelligence (MBZUAI), the world’s first AI university based in Abu Dhabi, across research and development in foundation models, talent development, and AI research. A few weeks ago, the company unveiled Mistral Medium 3, which focuses on cost-effectiveness while outperforming competing models like Meta’s Llama 4 Maverick in benchmark tests. The company stated that it is designed for enterprise deployment and excels in coding, STEM, and multimodal tasks. It achieves over 90% of Claude Sonnet 3.7’s benchmark scores at significantly lower pricing—$0.40 per million tokens for input and $2 for output.","excerpt":"It is called as Devstral, and it outperformed several other open-source models on the SWE-Bench verified benchmark.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","mistral"],"author_name":"Supreeth Koundinya","publish_date":"2025-05-21T22:25:25","publication_year":"2025","word_count":313,"keywords":["Go","artificial intelligence","TPU","OpenAI","AI","Llama 4","Git","mistral","foundation models","Claude 3.5","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","foundation models","OpenAI","Claude 3.5","Llama 4","TPU","R","Go","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/mistrals-new-model-is-trained-to-solve-real-github-issues\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10118153,"title":"NaMo App Introduces AI Chatbot NaMo AI, Offers Quick Answers to Government Schemes","content":"Prime Minister Narendra Modi’s NaMo App has launched a new feature called NaMo AI, using  generative AI to share details about the government’s flagship schemes and their impact. This AI-powered tool allows users to ask questions about PM Modi and receive quick summaries. Last night, I was checking this amazing feature on the @NamoApp , called the NaMo AI, an AI assistant which helps you with information about the various schemes of the Modi Govt.I got instant search results on number of PM Surya Ghar applicants from Assam. Must try this out! pic.twitter.com\/7qEbi2i7d8— Himanta Biswa Sarma (@himantabiswa) April 12, 2024 For example, users can inquire about initiatives like “Har Ghar Jal” and get insights. The AI also responds to questions about PM Modi’s popularity and awards received. NaMo AI aims to enhance public connectivity and provide useful information during elections. It can be accessed on desktop and mobile devices via the Narendra Modi website or the mobile app, offering responses in PDF format for easy sharing and access. Importantly, NaMo AI’s utility extends beyond information dissemination, potentially aiding voters during upcoming elections by providing essential details about the Prime Minister’s constituency and developmental initiatives through the chatbot. NaMo AI is likely the first of its kind used by a Prime Minister to strengthen public engagement. It’s wise to double-check information on official platforms for accuracy. Users can use NaMo AI to ask questions and receive answers in PDF format. The app is convenient for on-the-go use, accessible from anywhere at any time. As India gears up for the Lok Sabha election this year, a battle between AI-generated and manual-generated content campaigns is in the air. Currently, AI is used increasingly to help politicians reach voters through phone conversations or chatbots and further draft ads and messages about political opponents. As an effort to embrace AI in Indian politics, Hari Balasubramaniam, an Angel Network Investor, shared Prime Minister Narendra Modi’s AI image generated using Midjourney on LinkedIn. This post reflects a curiosity about reshaping our perception of political leaders.","excerpt":"Users can inquire about initiatives like “Har Ghar Jal” and get insights.","categories":["AI News"],"tags":["Modi"],"author_name":"Siddharth Jindal","publish_date":"2024-04-12T18:25:01","publication_year":"2024","word_count":338,"keywords":["Go","API","programming_languages:R","AI","chatbots","programming_languages:Go","Aim","ViT","generative AI","Modi","R"],"extracted_tech_keywords":["AI","generative AI","Aim","chatbots","R","Go","API","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/namo-app-introduces-ai-chatbot-namo-ai-offers-quick-answers-to-government-schemes\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10120648,"title":"92% of Indian Knowledge Workers Embrace AI at Work: Microsoft &amp; LinkedIn Report","content":"A staggering 92% of knowledge workers in India are using artificial intelligence (AI) at work, compared to the global average of 75%, according to the 2024 Work Trend Index released today by Microsoft and LinkedIn. The report, titled “AI at Work is Here. Now Comes the hard part,” highlights the rapid adoption of AI in the workplace and its impact on how people work, lead, and hire. The survey, which included 31,000 people across 31 countries, also found that 91% of Indian leaders believe they need to adopt AI to stay competitive. However, 54% worry that their organisation lacks a clear plan and vision for AI implementation. Despite this, employees are eagerly embracing AI tools, with 72% of Indian AI users bringing their own AI tools to work (BYOAI). “Data from the Work Trend Index shows that AI is now a reality at work, with India having one of the highest AI adoption rates among knowledge workers, at 92%,” said Irina Ghose, managing director of Microsoft India and South Asia. “This AI optimism presents a tremendous opportunity for organisations to invest in the right tools and training, to unlock efficiencies for employees and ultimately drive long-term business impact.” The report also highlights the growing importance of AI skills in the job market. 75% of Indian leaders stated they wouldn’t hire someone lacking AI skills, outpacing the global average of 66%. Furthermore, 80% of leaders in India prefer to hire a less experienced candidate with AI skills over a more experienced candidate without them. LinkedIn data shows a 142x increase globally in members adding AI skills to their profiles and a 160% increase in non-technical professionals taking AI courses. The study identified four types of AI users, ranging from skeptics to power users. AI power users in India are fundamentally reorienting their workdays, with 90% beginning their day with AI and 91% relying on it to prepare for the next day. They are also more likely to receive AI training and communication from senior leadership compared to other employees. “AI is transforming the world of work, reshaping the talent landscape and nudging both individuals and organisations to embrace change,” said Ruchee Anand, head of Talent & Learning Solutions at LinkedIn. “As the workforce looks to tap into the benefits of AI, it’s crucial for leaders to boost their organisation’s AI capabilities through thoughtful investment in both technology and talent.” Microsoft announced new capabilities in Copilot for Microsoft 365 to help people get started with AI, including more conversational features, proactive recommendations, and improved prompt experiences. LinkedIn also announced over 50 free learning courses to empower professionals at all levels to advance their AI aptitude. As the adoption of generative AI at work has nearly doubled globally in the last six months, Indian leaders face the challenge of moving from experimentation to tangible business impact. The 2024 Work Trend Index provides insights into how AI is shaping the future of work and the steps organisations can take to harness its potential.","excerpt":"The survey, which included 31,000 people across 31 countries, also found that 91% of Indian leaders believe they need to adopt AI to stay competitive.","categories":["AI News"],"tags":["AI adoption"],"author_name":"Shyam Nandan Upadhyay","publish_date":"2024-05-16T13:49:19","publication_year":"2024","word_count":499,"keywords":["API","artificial intelligence","AI adoption","programming_languages:R","AI","RAG","generative AI","GAN","R"],"extracted_tech_keywords":["AI","artificial intelligence","generative AI","RAG","R","API","GAN","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/92-of-indian-knowledge-workers-embrace-ai-at-work-microsoft-linkedin-report\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10170514,"title":"The Silent Collapse of Entry-Level IT Jobs in India","content":"Over the last two decades, India’s tech services sector has thrived on volume hiring, largely driven by Level 1 (L1) roles—entry-level positions tasked with repetitive, low-complexity work such as basic coding, maintenance, and support. Notably, these jobs formed the foundation of employment across industry giants like TCS, Infosys, Wipro, and others. However, a quiet transformation is now underway. The industry is witnessing changes in hiring patterns, not due to a fall in tech talent demand, but because of the evolving nature of the work itself. This shift is reflected in hiring data from some of the biggest IT employers. TCS, despite being India’s largest IT services firm, added only 625 employees in Q4 of FY25—a clear indicator of cautious and selective hiring. According to Milind Lakkad, chief HR officer at TCS, the company reportedly hired 40,000 graduates in 2020 and plans to increase that number further. However, that momentum has slowed significantly. Wipro also reflects this trend. As per Saurabh Govil, CHRO at Wipro, the company hired 10,000 freshers in FY25, in line with its revised targets. This is a sharp decline from the 38,000 freshers it planned to onboard in FY23, and the 19,000 it hired the previous fiscal year, as per reports. This signals that entry-level jobs are being redefined. Mohit Saxena, CTO and co-founder at InMobi, told AIM, “Essentially, AI has impacted the talent spectrum in two key ways. First, it has lowered the bar for becoming an average engineer.” Today, even individuals without a formal engineering degree can write code. It may not be high-quality, but it is equivalent to what a weak engineer might produce. “So, the threshold to enter the field has been lowered. At the same time, if you are a good engineer, AI has elevated your capabilities even further,” he added. With the kind of assistance AI offers, a good engineer becomes exceptional. Hence, the gap between average and excellent has widened. “If you are caught in the middle, not leveraging AI and not evolving, you are in trouble,” Saxena warned. The AI and Automation Wave With the integration of generative AI, agentic AI systems, and no-code or low-code platforms, many of the tasks traditionally assigned to L1 engineers, such as basic code writing, testing, and routine ticket handling, are now being handled with minimal human intervention. For example, AI-driven platforms like GitHub Copilot, ChatGPT, and other code-assist tools can now write and optimise blocks of code with simple prompts. “About 25% to 30% of development is going to happen through AI,” Neeti Sharma told AIM, echoing what several IT services CEOs have publicly acknowledged. “We’re already seeing 30% productivity improvements, but it’s not just about code generation,” Muralidhar Krishnaprasad, president and CTO of Salesforce, said in an exclusive interview with AIM, adding that the company is also using AI for test case generation. He further explained that AI can now generate these test cases automatically by understanding the code’s logic and user flows. This saves time and helps developers catch bugs earlier in the development process. In fact, Gautam Goenka, VP of engineering and site head at UiPath, explained how new responsibilities are emerging with the rise of AI and agent-based systems. According to him, skills like prompt engineering, system understanding, and strong domain knowledge are becoming increasingly critical to building effective AI agents. What’s more telling is how hiring metrics have evolved. “They will not look for people who can write lines of code…They look for people who can write prompts better. Prompt engineering is the new benchmark,” Sharma highlighted. Meanwhile, in an interview with AIM, Alan Flower, executive VP and head of AI Labs at HCLTech, revealed that business leaders worldwide perceive AI as a net productivity gain.“We have observed perhaps over a 60% productivity increase in certain areas,” he revealed. Traditional IT Services Model Faces Its Own Reckoning While global capability centres (GCCs) continue to expand their footprint and evolve into innovation hubs, traditional Indian IT firms—such as Infosys, TCS, and Wipro—appear to be on the back foot. With delayed fresher onboarding, limited salary hikes, and shrinking profit margins, their longstanding dependency on L1 roles has perhaps become a liability in today’s rapidly evolving tech landscape. “Unfortunately, for the last three years, fresher engineers have not got the jobs they’re looking for from services companies,” Neeti Sharma, CEO at TeamLease Digital, told AIM. This shift is also reflected in broader hiring trends. According to a TeamLease Digital survey, the hiring outlook for engineering graduates is expected to remain below pre-pandemic levels, with only 1.6 lakh freshers projected to find employment, compared to 2.3 lakh in FY23. This decline underscores the shrinking pipeline for traditional entry-level roles. As Sharma mentioned, the trend was once largely a movement from IT services firms to GCCs. Now, the competition has intensified within the GCC ecosystem itself. This trend signals more than just job shifts; it reveals a rising premium on domain depth and system-level thinking, and the obsolescence of roles that lack those competencies. In this environment, GCCs are clearly raising the bar when it comes to talent acquisition. They are no longer content with just technical skills. Instead, they are looking for talent that blends engineering expertise, technological fluency, and domain understanding. “The combination of engineering, technology, and domain is what GCCs look for. And I don’t see a reduction in terms of those requirements,” Sharma mentioned. This shift means that today’s engineers are not just expected to code—they are expected to understand business logic, customer expectations, and domain-specific applications of their work. For many fresh graduates, this represents a significant leap; one that traditional academic structures often don’t prepare them for adequately. Several forward-looking GCCs are actively embedding this mindset into their operating models. For instance, Anuj Khurana, CEO and co-founder of Anaptyss Inc, stated that the company has built an ecosystem of ‘Digital Entrepreneurs in Residence’, which allows it to prototype and implement industry-specific digital solutions. These are grounded in deep-tech capabilities and a profound understanding of the banking and financial services domain, exemplifying how domain-led innovation is becoming central to the GCC value proposition. L1 Roles are Disappearing, But GCC Internship Pipelines are Expanding One might assume that the only path to building a tech workforce is through fresher hiring. However, that’s not the case in the GCC world. Most centres now focus on structured internships and apprenticeships, treating them as long-term hiring pipelines with very high conversion rates. He added that many companies compensate for this by investing in induction programmes and internal skill-building initiatives. For example, as Kalavathi GV, executive director and head of the global development centre at Siemens Healthineers, pointed out, the company aims to close the gap between academia and industry by sending industry professionals to engage with academic institutions. “They (GCCs) are quite choosy. They don’t go to all engineering colleges; they are very selective. [They will] either go to Tier 1, or to the IITs, IIITs of India,” Sharma added. This means that internships are no longer a “good-to-have” for students; they are often the only route to entering premium tech careers. The GCCs use internships to evaluate behavioural readiness, soft skills, and domain adaptability well before considering full-time employment. “We believe this is the only way to prepare the workforce—through deployment, education, and assessment all rolled into one,” Sharma explained, referring to degree-linked apprenticeships where candidates work, learn, and earn credits simultaneously.","excerpt":"The industry is witnessing changes in hiring patterns, not due to a fall in tech talent demand, but because of the evolving nature of the work itself.","categories":["AI Features"],"tags":["Indian IT"],"author_name":"Shalini Mondal","publish_date":"2025-05-22T13:36:56","publication_year":"2025","word_count":1232,"keywords":["Go","ChatGPT","agentic AI","AI","ML","RAG","Aim","prompt engineering","generative AI","Indian IT","R"],"extracted_tech_keywords":["AI","ML","generative AI","agentic AI","ChatGPT","Aim","RAG","prompt engineering","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/the-silent-collapse-of-entry-level-it-jobs-in-india\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10091132,"title":"Microsoft &#8216;JARVIS&#8217; is the Path Towards AGI","content":"The accidental open sourcing of ‘LLaMA’, Meta’s LLM, acted as a spark to rejuvenate the open-source AI community. Now, it seems that Microsoft wants to replicate their accidental success with the launch of ‘HuggingGPT’, also known as ‘JARVIS’. This technology, built on ChatGPT, aims to leverage Hugging Face, one of the biggest pillars of open-source AI research, to create a new approach to solving complex AI problems. Researchers from Microsoft detailed a way to use LLMs as the user-facing part of the system, utilising its natural language capabilities to interface with other models. This seems to be somewhat of a spiritual successor to ‘Visual ChatGPT’, which used a similar approach to plug in LLMs to text-to-image models. JARVIS explained Named after Iron Man’s personal AI assistant, ‘JARVIS’ aims to bring together the power of the open-source community and ChatGPT. Just as JARVIS accesses Tony Stark’s vast arsenal of services and acts as an AI butler of sorts, HuggingGPT calls specialised models for certain use-cases by interfacing between the user and the models. The architecture created for HuggingGPT is made up of two main components. The first is the LLM, which acts as a controller. This model takes up the roles of planning out tasks, selecting the secondary model, and response generation. The second component is the Hugging Face platform, which mainly conducts task execution. The standout feature of JARVIS is the idea behind it, which can be condensed to the definition ‘language-as-an-interface’. By using language as a general interface and putting the LLM in the ‘brain’ position, it is possible for many different, specialised AI models to work together. HuggingGPT\/JARVIS Architecture The researchers provided many examples to illustrate the potential use-cases of JARVIS. By giving a single prompt containing multiple instructions, HuggingGPT was able to call on a pose detection model, image generation model, image classification model, image captioning model, and a text-to-speech model. Request flow in HuggingGPT\/JARVIS While the models called on by JARVIS are not novel and have been a mainstay of the open-source community for years, bringing them together is a novel approach to solve complex problems. Even though the given prompt had multiple stages of execution with different tasks in each step, the architecture handled it flawlessly. Microsoft’s newfound attitude towards leveraging open-source research shouldn’t come as a surprise, especially considering the waves that LLaMA has been making over the past few weeks. Open source is the next big multiplier for AI, and it seems that Microsoft is on board with it. Open source for AGI While Microsoft is beholden to Sam Altman and OpenAI’s policy of closed AI research, it seems that they are pursuing a different path towards AGI. While the research paper carefully avoids using this loaded term, the abstract of the paper describes solutions like HuggingGPT to be a “key step” towards “advanced artificial intelligence”. For all its talk of creating an AGI and where humanity is on the “path to AGI”, OpenAI is increasingly closed in terms of its research. While many scientists and researchers have criticised this approach of treating AI as proprietary technology, many others have already built up a comprehensive reputation for open sourcing models in the AI community. Last month, the release of LLaMA essentially sparked the open-source community into action by giving them a state-of-the-art LLM (with leaked weights). This has now resulted in a spate of LLaMA-based projects being released out into the world—a formula Microsoft seems eager to reciprocate. Indeed, leveraging the open-source community’s vast library of open-source algorithms might just be the path towards AGI. By bringing together various domain specific AI, also termed ‘narrow AI’, it is possible to move towards a type of artificial general intelligence known as self-organising complex adaptive systems. In his musings on AGI, Ben Goertzel, the CEO of SingularityNet, offered the idea of a narrow AGI which sounds suspiciously similar to Microsoft’s JARVIS. He stated, “There is a path from today’s Narrow AIs to tomorrow’s AGIs that passes through intermediate systems that are best thought of as Narrow AGIs.” These so-called intermediate systems are the precursors for SCADS, which are AI systems composed of smaller AI algorithms. The ‘intelligent’ part of SCADs is responsible for deciding which algorithm performs which function, similar to ChatGPT’s role in HuggingGPT. Goertzel elaborates, “A Narrow AGI for biomedical analytics might leverage a small army of Narrow AI tools carrying out specific intelligent functions—but it would figure out how to combine these on its own.” According to Goertzel, combining these narrow AIs into a bigger AI would create a SCADS AI, which, in turn, would pave the way for a human-like AGI. By creating HuggingGPT, researchers have actually begun making realistic progress towards an AGI, far removed from OpenAI’s empty promises of an AGI future.","excerpt":"Bringing together various expert systems with language as an interface might move us towards AGI.","categories":["Global Tech"],"tags":["Hugging Face","HuggingFace","OpenAI"],"author_name":"Anirudh VK","publish_date":"2023-04-11T12:00:00","publication_year":"2023","word_count":793,"keywords":["Go","Hugging Face","ChatGPT","artificial intelligence","OpenAI","AI","RAG","Aim","analytics","HuggingFace","R"],"extracted_tech_keywords":["AI","artificial intelligence","analytics","ChatGPT","OpenAI","Aim","Hugging Face","RAG","R","Go"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/microsoft-jarvis-is-the-path-towards-agi\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":27926,"title":"Zee Entertainment’s Tryst With Data Science","content":"With the entertainment industry getting quite competitive, new age technologies like artificial intelligence and analytics is helping it grow and widen its viewership. Zee Entertainment, owned by the Essel Group, is already making the most of it. From delivering a personalised experience to churning out customer insights into investments, they are in with the times. Analytics India Magazine caught up with Venkat Nettimi, who heads the strategy and consumer insights at Zee Entertainment, and has brought various analytical tools and processes at Zee. Having worked with the likes of Amex, Citigroup, CIBIL and others, he brings a diverse experience from across several geographies. As he shares, he has had the opportunity to see amazing changes in the world of analytics over the years — from manually tabulating results to using complex analytical solutions. Over the last one year, Zee has implemented mobile-based analytical solutions which are enabling decision-makers to get access to critical information and cut down the analysis time by nearly 90 percent. He also strongly believes that AI and machine learning-based predictive analytics will help better understand the consumers and their content preferences. Entertainment Industry And Analytics While Nettimi has an in-depth experience across industries like consumer packaged goods, financial services and media, the essence of using analytics has been the same — how can we solve a business problem x using data y? However, the shape of x and y varies according to the industry. For instance, in the CPG industry, a lot of problems are solved using consumer feedback and transactional retail data. Financial services have a distinct advantage, thanks to the vast and exclusive transaction histories available to them. However, his journey in media has been different and exciting. As he notes, unlike other industries, the key information on which their business runs are the audience measurement data provided by BARC. It is both their monetisation currency and critical input for audience understanding. And unlike that for other industries, this data is not exclusive to any one player and is available to all. “Thus, data and data availability is no longer the differentiator. It is purely the analysis you do with the data that differentiates you from the other and provides you the competitive edge. In some ways this hones your analytical ability like no other industry does,” he says. Use Cases From Zee Entertainment Nettimi’s focus at Zee has been two folds — to increase the breadth and usage, and the depth of analytics solutions within the firm. He shares that not everyone in the firm was comfortable with the use of data and analytical tools — especially the content and creative teams. “They were highly dependent on their insights people to extract and share the ratings data with them, which created bottlenecks in decision-making by taking time away from the insights team’s bandwidth for more value-added analysis,” he said. Since then they have introduced highly visual and intuitive Tableau-based desktop and mobile dashboards that have truly democratised data usage within the firm. Especially useful are the interactive mobile dashboards, which address almost 90 percent of routine queries. “To increase the depth of analysis, we have been focusing on adopting cutting-edge analytical tools, including AI and ML-based predictive analysis, image recognition based meta tagging etc.” he said. They are also adopting tools used in other industries such as for discrete choice modelling and market basket analysis for better decision making. “In addition to the use of advanced data analytics, we’re also augmenting our understanding of our consumers via more qualitative methods such as consumer immersions, ethnography, observational studies, Neuroscience etc,” he said. Consumer Insights And Machine Learning Nettimi is quick to add that one of the most interesting projects that he has ever led includes ML-based automated scheduling of their movies library. With over 10 Pay movie Channels, 3,500 library titles, 5,500 hours of movie scheduling per month in the category, optimal movie scheduling can provide an extremely important competitive advantage. Since a majority of their titles are library content, historical data can be used to predict future ratings and this helps them in: Manage their library more efficiently Reducing the man-hours used in scheduling Increase our ratings. “So it’s one of those projects that addresses cost and revenue drivers at the same time,” he added. He is also leading a project focused on harnessing content learnings using tech. With over 200,000 hours of content produced every year, they are trying to build an understanding of what content works best for them, for which AI comes into picture. “Using AI-based meta-tagging of video content (characters, emotions, setting etc.) and correlating it to the BARC ratings is enabling us to understand what content works – at scale,” he said. He also shares that they have started using algorithms for ratings and predictions, especially for their movie offerings. “I also see recommendation algorithms becoming more and more important in the near future,” he said. Analytics For Making Investment Decisions At Zee “Analytics is absolutely at the core of every decision we make at our firm. Considering our ratings information is also our monetisation instrument, it becomes an extremely important input for decision making,” he notes. It helps them in understanding what content works for them, guides purchase decisions on shows and movie libraries, understand client (advertiser) needs and develop the best possible solutions to address their advertising needs. Importantly it helps them in large investment decisions such as the launch of new channels or rebranding and relaunch of existing channels. Nettimi shares that their recent launches in English movies space (& Flix and & Prive) and upcoming regional channel offerings are the results of understanding viewer preferences and trends via ratings data. Data Visualisation And Analytics Tools At Zee Nettimi shares that currently they have a distributed model for the Insights function within the organisation, with each business having embedded and dedicated insights and analytics SMEs. There is a central team which drives key strategic initiatives and innovative analytical projects. There over 50 people working in this space within the Domestic Broadcast business (excluding Zee5). “Tableau has been embedded within our organisation for a while now and we’ve members being trained in R as well. We also work with a wide array of small and large data analytics firms to help bring in fresh outside perspectives in analytics,” he shares. Competing With Other Players In OTT Space With Analytics Zee has always been a pioneer in the online and OTT space. Much before the launch of Zee5, Zee was offering Ozee and Ditto to address the consumer needs of Anytime TV and Catch Up TV (both of which have been integrated into the Zee5 platform). “In addition to the rich viewership data available on the platform, which enables us to model consumer preferences and develop content recommendations, we also use other data sources to augment our understanding of consumers. For example, a key focus has been on using social media data to gain feedback on our shows, understand the broader narrative of topics which our viewers are interested in and, thus, enable us to have an ‘ahead of the curve’ listening ability on key themes which are important to our consumers. This helps us make better content decisions,” he said. Challenges And Roadmap “It’s an exciting time to be in the analytics space,” theorizes Nettimi. “The availability of different tools, techniques has increased exponentially over the last few years. However, in some instances the pace of adoption of new techniques has not kept up with the rate of pace of innovation of the same. Sometimes there is a tendency to fall back on “tried and tested” methods as opposed to new and innovative methods. As an analytics leader, a significant part of my work involves educating internal stakeholders on the benefits of these new techniques,” he says. On the flip side, sometimes there is a rush to adopt the “shiniest new toy” in the analytics space without fully understanding whether there is relevance of it to the business and whether it will add value. Also with an increasing volume and velocity of data, one of the biggest challenges is to be able to make sense of it all to drive profitable business decisions. “One of the key initiatives in place is to bring all the data available across, not just media, but the Essel Group of companies onto a common Data Lake that will enable us to leverage opportunities across sectors. We’ll also continue to build our data analytical abilities and move some of the work from our partner agencies internally,” he said on a concluding note.","excerpt":"With the entertainment industry getting quite competitive, new age technologies like artificial intelligence and analytics is helping it grow and widen its viewership. Zee Entertainment, owned by the Essel Group, is already making the most of it. From delivering a personalised experience to churning out customer insights into investments, they are in with the times. […]","categories":["AI Features"],"tags":["Interviews and Discussions"],"author_name":"Srishti Deoras","publish_date":"2018-09-03T12:39:25","publication_year":"2018","word_count":1431,"keywords":["machine learning","artificial intelligence","AI","R","ML","image recognition","RAG","Ray","analytics","predictive analytics","Interviews and Discussions"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","analytics","Ray","RAG","predictive analytics","image recognition","R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/zee-entertainments-tryst-with-data-science\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10125622,"title":"Why Indians Working in US is Bad News for India","content":"Many people dream of earning a big income in the US, working for a few years, and then returning to India to support their families, contribute to their communities, or launch startups. However, this vision isn’t playing out as expected. While a user on X has some compelling reasons to persuade you to relocate and earn big abroad, the reality doesn’t quite add up. Sunil Avaria shared on X that one must move to California because, “All these people can change their community, change their village, and help their relatives like they have never imagined. Probably at an early age of mid-30s.” However, the discussion took an interesting turn when someone mentioned NVIDIA: “Ten years ago, instead of going to the US for school, spending $100k, and working 5+ years at a job, you could have invested that $100k in NVIDIA stock and made $25 million today.” Investing in stocks was just one aspect of the discussion that also highlighted the increased number of Indian employees in NVIDIA. Data suggests that in 2023, NVIDIA reported to have 26,196 employees. Of this figure, about half, which makes 50% of the company’s employees, identified as Asians\/Indians. Source: Statista 2024 But the Grass Isn’t Green on the Other Side Despite the impressive earnings abroad, there’s a noticeable gap between what’s being earned and what’s actually making its way back to India. The Flipkart case study explains this better. In 2018, US retail giant Walmart signed a definitive agreement to acquire a 77% stake in India’s largest e-commerce marketplace Flipkart with an investment of around $16 billion. But, did this money reach India? Flipkart initially incorporated its holding company in Singapore to attract foreign investments and avoid the challenges posed by India’s bureaucracy and regulatory system. Singapore offers an easier business environment, lower taxes, and stronger economic freedom and property rights, making it an attractive destination for many Indian entrepreneurs. Not just Flipkart, Blinkit, and Udaan are among the top Indian startups with parent companies in Singapore. But as per the recent updates, the e-commerce firm is reportedly looking at moving its domicile back to India. There’s Still Hope Meanwhile, Indian IT giants have established a robust presence in the US. According to the HDFC Securities report, Tata Consultancy Services (TCS) employs 50,000 people, Infosys 35,000, HCL Technologies 24,000, Wipro 20,000 and LTI Mindtree 6,500. Together, these companies account for approximately 2% of the US tech industry workforce. Harbir K Bhatia, the CEO of the Silicon Valley Central Chamber of Commerce, told the Press Trust of India, “India is one of the largest leaders of innovation in Silicon Valley. The data shows that 40% of Silicon Valley CEOs or founders are from South Asia or India.” Source: Statista 2024 Apple, a leader in the global tech industry, also reflects this trend, with nearly 35% of its workforce being of Indian origin. Adding to this narrative, Microsoft CEO Satya Nadella, in an interview with CNN-TV18, mentioned that Microsoft employs the second-largest number of engineers from India. When asked about the increasing number of CEO positions held by people of Indian origin, Nadella said, “In a country of 1 billion plus people, we should be seeing one-fifth of the CEO posts being held by Indians.” What’s Next? Many startups consider registering in India and often question why companies like Flipkart are registered in Singapore. However, despite deterrents like India’s FDI policy and taxation, the trend is changing. Several Indian startups across sectors like fintech, healthcare, edtech, and e-commerce are planning to relocate their base back to India. As per reports, in March last year, the Indian government recognised that many Indian startups were based abroad and established a committee to recommend ways to “onshore Indian innovation” to the Gujarat International Finance Tec-City. Recently, Groww, a fintech startup moved its base back to India from the US. Similarly, PhonePe transferred its domicile back to India from Singapore in October 2022.","excerpt":"The Flipkart case study explains this better.","categories":["AI Features"],"tags":["AI Impacts"],"author_name":"Vidyashree Srinivas","publish_date":"2024-07-03T11:14:35","publication_year":"2024","word_count":653,"keywords":["Go","programming_languages:R","AI","innovation","programming_languages:Go","AI Impacts","Rust","CNN","R","programming_languages:Rust","startup"],"extracted_tech_keywords":["AI","R","Go","Rust","CNN","innovation","startup","programming_languages:R","programming_languages:Go","programming_languages:Rust"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/why-indians-working-in-us-is-bad-news-for-india\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10084694,"title":"3nm Chips: Apple Innovates, Android Disrupts","content":"At present, Apple is one of the only companies that is proactively pushing towards adoption of newer and more advanced 3nm chips, unlike Qualcomm and MediaTek who are grappling with uncertainty over following suit citing ‘unclear sales prospects for Android handsets’. In addition, with reports of big-ticket customers like AMD, NVIDIA, and Intel cutting orders with TSMC, its utilisation rate is likely to witness significant decline across all process technologies. This may include 7nm\/6nm declining ~50% in early 2023, the mode advanced 5nm\/4nm process, and, more importantly, the 28nm class that has been fully loaded since the beginning of the chip deficit in early 2021. Samsung reported a 69% decline in profits for its December quarter, citing weak demand and pricing factors. DRAM and NAND flash memory prices have declined alongside the slack demand, while chip inventories are piling up across the supply chain. However, despite the grim state of the current global chip supply, experts aren’t particularly perturbed. They believe it to be a part of the natural cycle of the industry that is currently experiencing a downturn. The upturn during the pandemic—which had elicited a sudden rise in demand and the consequent global chip shortage—has returned to normalcy, leaving chipmakers with large inventories but no demand for the same. Cost vs pricing: The struggle Sravan Kundojjala, principal industry analyst at Strategy Analytics, told AIM that for the past two years, the semiconductor supply chain was contending with inflationary pricing driven by a sudden uptick in demand, shortage of key materials such as gases, the Russia-Ukraine War and increased labour, freight, and energy costs. These factors also contributed to the pricing across the semiconductor value chain. As a result, the suppliers started to pass along increased costs to their customers to cope with inflationary costs and to protect margins. “Unsurprisingly, most semi-companies enjoyed healthy pricing in the past two years. NXP, for example, saw three straight years of price increases. Analog Devices saw up to 50% of its growth driven by pricing alone. Even TSMC saw a 20%+ jump in its wafer pricing in 2022 vs 2021,” Kundojjala pointed out. As of now, the entire semiconductor ecosystem is struggling to find the right balance between cost and pricing. If we speak of Apple alone, the company was trying to maintain a stable price point by releasing the non-Pro models of its flagship iPhone 14 (iPhone 14 and iPhone 14 Plus) with the A15 (5nm) bionic chip while powering the Pro models (iPhone 14 Pro and iPhone 14 Pro Max) with A16 (4nm). Now, with it moving to the 3nm class for the A17 chip, Apple will continue to be under test each time it chooses to employ a denser chip, as the cost-per-transistor increase will become higher as we move towards smaller process nodes. However, unlike Apple, Android companies will likely wait until 2024 for 3nm. This is because, in addition to the worsening demand situation, the timing has not been right for them. Besides TSMC, only Qualcomm and Samsung have been using the leading edge until recently. MediaTek joined last year. If Samsung had perfected its 3nm technology (which it developed using the GAA-FET architecture) last year, Qualcomm would have been the first to market. Android still making its mark Android’s failure to bank on the 3nm ahead of Apple will affect its competitiveness in the market. Still, experts like Kundojjala believe it is a safer bet for them to wait until 2024 as Android companies are conscious of initial costs since their premium tier volumes are not comparable to Apple. There is simply not enough room for cost efficiencies in premium tier chips such as Snapdragon 8 Gen 2. And increasing prices to accommodate the logical scaling of chips will also affect their margins. However, this is not the end of the story for Android. Process technology only goes so far. Companies look to optimise chip architectures by seeing how hardware and software components work together, so they can efficiently solve a problem in a given set of constraints. This way, chip companies include all the IP blocks and work closely with customers to satisfy their needs. In addition, these companies offer software SDKs to OEMs, who take advantage of underlying capabilities. For instance, if Samsung wants to include more AI functionality, it will ask Qualcomm to enhance that aspect in their designs. Qualcomm then introduces the necessary changes in its designs. Thus, there are a lot of ways to increase performance that include adopting new CPU\/GPU cores and improving AI and camera performance. However, Kundojjala says, “The costs continue to go up for advanced chips due to increased functionality on the processing side (CPU, GPU, AI, multimedia, camera, gaming, and security, etc).” And so, ultimately, it is the balancing act of how to enjoy better pricing for the rising costs. He says, “We should expect Qualcomm to enjoy better pricing on 3nm in 2024, compared to 4nm in 2023.” Furthermore, Android is inching closer on certain benchmarks to Apple’s A chips. For example, in one demonstration pitching Snapdragon 8 Gen 2 against Apple A16 Bionic, the AI engine in Qualcomm’s chip ran 4.5 times faster than its last-gen SoC and is said to be ahead of Apple’s neural engine, albeit only by a margin. On the 5G and wireless technology end, Qualcomm has been the leader since the A16 also uses a Qualcomm modem. However, this might not stay for long as Apple plans to go in-house and reduce its reliance on third-party vendors. Besides, Qualcomm has also done an excellent job in improving power efficiency in the GPU department.","excerpt":"At present, Apple is one of the only companies that is proactively pushing towards adoption of newer and more advanced 3nm chips.","categories":["Global Tech"],"tags":["Android","Apple","Apple Bionic","Qualcomm"],"author_name":"Ayush Jain","publish_date":"2023-01-11T13:00:00","publication_year":"2023","word_count":934,"keywords":["Go","Apple Bionic","programming_languages:R","AI","Apple","programming_languages:Go","RAG","Aim","Qualcomm","analytics","R","Android"],"extracted_tech_keywords":["AI","analytics","Aim","RAG","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/3nm-apple-innovates-android-disrupts\/","complexity_score":4,"technical_depth":8,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10072725,"title":"NVIDIA Wants to Create a More Realistic Metaverse","content":"A few decades ago, the Internet boomed and it completely altered the world. “What we’re seeing is the start of a new era of the Internet. One that is generally being called the Metaverse,” says Rev Lebaredian, vice president of Omniverse and Simulation Technology at NVIDIA, in a press briefing. The web, as we know it, is two-dimensional but the power and the potential of 3D technology is expected to drive this new era of the Internet. The ‘Metaverse’ is still in its early phase of development. As it stands, it’s rather unsophisticated and to be more suitable for wider adoption, it needs to get more realistic to enable an excellent immersive experience. In the 2022 edition of the SIGGRAPH conference held in Vancouver, NVIDIA announced a wide range of Metaverse initiatives to help achieve this goal. With its new Metaverse tools, NVIDIA is expected to bridge the gap between AI and the Metaverse. “Having a design team behind recreating the real world as virtual space is time-consuming and not very efficient considering the pace at which the Metaverse is growing. We need AI to offload all the repetitive takes that a designer is intended to do and rather focus on other aspects of virtual world creation,” says Mukundan Govindaraj, Solutions Architecture–Omniverse, NVIDIA in conversation with Analytics India Magazine. Neural Graphics to make Metaverse more realistic According to Rev Lebaredian, NVIDIA is still a ‘tools company’ with graphics at its core. It is known for its Graphic Processing Units (GPUs). In 2020, the US-based tech giant announced its first GPU-based AI chip capable of boosting performance by up to 20x. At present, NVIDIA aims to use the power of neural graphics to create realistic 3D objects and drive the development of the Metaverse. 3D creation will play a crucial role if a wider adoption of the Metaverse is to happen. (Source: NVIDIA) Neural Graphics is a novel tech that brings together the power of AI and graphics to develop an accelerated graphics pipeline that learns from data. Building 3D objects for Metaverse involves meticulous processes such as product designing and visual effects. “Often, developers balance detail and photo realism against deadlines and budget constraints. Developing something that depicts the real world in the Metaverse is a very difficult and time-consuming task. What makes it even more challenging is that multiple objects and characters need to interact in a virtual world. Simulating physics becomes just as important as simulating light,” says Mukundan Govindaraj. Tools and programmes—including NeuralVDB and Kaolin Wisp—that enable quick and easy 3D content creation for millions of designers and creators were also recently announced by NVIDIA. NeuralVDB: It is an update to industry standard OpenVDB. By using ML, NeuralVDB drastically helps in reducing the memory footprint to allow for higher-resolution 3D data. Kaolin Wisp: It is an addition to Kaolin, a PyTorch library enabling faster 3D deep learning research. It helps bring down the time needed to test and implement new techniques from weeks to days. 3D MoMa: It is also a new inverse rendering pipeline that allows developers to import a 2D object into a graphics engine and create a realistic 3D object from it. Life-like virtual assistants Among the different tools announced by NVIDIA, the Omniverse Avatar Cloud Engine (ACE) is deemed the most intriguing. It is a new AI-assisted 3D avatar builder. NVIDIA claims that with the help of ACE, developers will be able to create autonomous virtual assistants and digital humans. Users often interact with voice assistant softwares such as SIRI and Alexa. Now, with this new technology, both SIRI And Alexa could possibly have a face. “The Metaverse without human-like representations or AI inside it will be a very dull and sad place,” says Rev Lebaredian. In concurrence, Mukundan Govindaraj further explains—“It’s a collection of AI models and services using which developers can quickly build, customise, and deploy interactive avatars. Developers can leverage the Omniverse Avatar technology platform to build their own domain-specific avatar solutions.” NVIDIA also announced Omniverse Audio2Face—an AI-based tech that generates expressive facial animation from an audio source. An expanded Omniverse During the conference, NVIDIA also announced a new version of Omniverse. Omniverse is a Universal Scene Description (USD) platform, an engine that builds Metaverses. It has been downloaded more than 200,000 times so far. The new version of NVIDIA’s Omniverse will allow developers to create content for a significantly better immersive experience. USD is emerging to become the HTML of the Metaverse. “USD, developed and open-sourced by Pixar, combines the best parts of the previous file formats and runtime APIs. The ability to interoperate with many tools is going to be the driving factor for its popularity and adoption across all industries working with 3D file formats,” says Mukundan Govindaraj. According to NVIDIA, the new version of the Omniverse comes with several upgraded core technologies and more connections to popular tools. “Connectors are now available in beta for PTC Creo, Visual Components and SideFX Houdini. These new developments join Siemens Xcelerator, now part of the Omniverse network, welcoming more industrial customers into the era of digital twins,” says NVIDIA in a blog post.","excerpt":"The Metaverse without human-like representations or AI inside it will be a very dull and sad place","categories":["Global Tech"],"tags":[],"author_name":"Pritam Bordoloi","publish_date":"2022-08-14T12:00:00","publication_year":"2022","word_count":854,"keywords":["Go","AI","PyTorch","ML","virtual assistants","RAG","Aim","deep learning","analytics","R"],"extracted_tech_keywords":["AI","ML","deep learning","analytics","Aim","PyTorch","RAG","virtual assistants","R","Go"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/nvidia-wants-to-create-a-more-realistic-metaverse\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10140251,"title":"Can Google Beat AI Rivals and Keep the Ad Cash Rolling?","content":"Tech giant Google’s advertising business continues to drive its financial performance with the company posting the highest ever ad revenue of $65.85 billion for Q3 FY24. It constitutes nearly 75% of its total revenue. Amid the company’s expansion drive through Google Cloud and AI innovations and investments, advertising revenue remains the core of its operations. Ad revenue increased 10% year-on-year, indicating the company’s dominance in search-based and video ads. Search revenue alone contributed $49.39 billion to the total, a 12% increase from last year. YouTube also performed well, earning $8.92 billion for the quarter. Sundar Pichai, CEO of Alphabet which owns Google said, “The momentum across the company is extraordinary. Our commitment to innovation, as well as our long-term focus and investment in AI, are paying off with consumers and partners benefiting from our AI tools.” QuarterAd Revenue (in millions)Year-over-Year % ChangeQ3 2024$65.810.41%Q3 2023$59.69.48%Q3 2022$54.42.54%Q3 2021$53.1– The company’s reliance on ad revenues is nothing new. Over the past few years, its ad revenues have been on an upward trajectory with Q3 marking the highest ever earnings. Google has successfully leveraged its search engine, user data and AI features to effectively manage the ad call-to-action behaviour. The company believes the ads on AI Overviews, a feature that summarises content on the search query and displays it under the search box, have allowed users to quickly connect with relevant businesses and services, thereby making the ad process more relevant. Rising Competition in the Ad Space However, its heavy reliance on ads makes its business challenging in a competitive landscape. The rise of AI driven search engines, driven by ChatGPT and Perplexity, poses a significant threat to Google’s search and ad revenue dominance. Both ChatGPT and Perplexity are releasing chrome extensions for their search engines. Besides, Google’s main competitor, Meta, too is reportedly entering the search engine space. The latter has shown consistent progress in the last two years with its ad revenue for Q3 FY24 touching $39.9 billion, an 18.7% jump year-on-year. Perplexity co-founder and CEO Aravind Srinivas posted on X about Google’s approach to raking up ad revenues despite his own company entering the search space. With these developments, Google cannot afford to lose its focus on the ad revenue nor ignore the emerging competitors. Source : X Meta is developing an AI search engine to answer queries on Meta AI. Currently the company is relying on Google and Microsoft’s Bing for the same. It is obvious that dominant players are trying to build their ecosystem to ensure their customers stay on their portal with minimal dependency on competitors. Meta has the advantage of a large user base and data from Facebook and Instagram platforms, so training the AI search platform might not be problematic. Meta’s web crawler is already scraping data for AI training. The company has even partnered with publications such as Reuters to bring news-related answers. AI Powers Ads Even as the company’s Q3 FY 2024 results surpassed analyst expectations in the top and bottom lines, with the consolidated revenues at $88.3 billion and Google Cloud revenue increased 35% to $11.4 billion, advertising income still remains a key growth driver. Google Cloud revenues was led by accelerated growth in Google Cloud Platform (GCP) across AI Infrastructure, Generative AI Solutions, and core GCP products. Pichai credited their long-term focus and investment in AI as key drivers of success for the company and its customers, even highlighting the Gemini API’s 14x growth over the past six months. Google claims that both customers and advertisers have found AI features to propel the user experience across its products and services. Advertisers have been using Gemini to build and test ad creatives at scale. Google’s latest text-to-image model, Imagen 3 was updated in Google Ads. The model was tuned with ads’ performance data across industries to aid customers with high-quality images for their campaigns. It’s interesting to note that AI-powered feature integration on search has also been economical. Pichai mentioned that when the company first began testing AI Overviews, they had lowered the machine costs per query ‘significantly,’ and now, in 18 months, the costs have been reduced by more than 90%. “AI is expanding our ability to understand the intent and connect it to our advertisers. This allows us to connect highly relevant users with the most helpful ad, and deliver business impact to our customers,” said Philipp Schindler, SVP and CBO at Google, on the earnings call.","excerpt":"The rise of AI-driven search engines, driven by ChatGPT and Perplexity, poses a significant threat to Google’s search and ad revenue dominance.","categories":["AI Features"],"tags":["ChatGPT","Generative AI","Google Ads","Meta","Perplexity","search"],"author_name":"Vandana Nair","publish_date":"2024-11-05T18:30:00","publication_year":"2024","word_count":736,"keywords":["Go","ChatGPT","Meta","Perplexity","Meta AI","GCP","Google Ads","AI","API","RAG","Aim","search","generative AI","Generative AI","R"],"extracted_tech_keywords":["AI","generative AI","ChatGPT","Meta AI","Aim","RAG","GCP","R","Go","API"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/can-google-beat-ai-rivals-and-keep-the-ad-cash-rolling\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10162308,"title":"Genpact Unveils AI Gigafactory to Create 25,000 AI Builders for Enterprise","content":"Genpact has launched the Genpact AI Gigafactory, an innovative accelerator aimed at scaling AI solutions from pilot projects to enterprise-wide deployment. This initiative seeks to address the increasing demand for AI adoption while ensuring robust governance and ethical practices. “The benefits of AI are clear, but most businesses struggle with how to implement it at scale,” said Sanjeev Vohra, chief technology and innovation officer at Genpact. “A multi-disciplinary approach is essential to drive transformative impact—one that combines deep industry knowledge, expertise in data, technology, and processes, along with highly specialized talent.” A recent study by Genpact and HFS Research revealed that only 5% of organisations have achieved full AI maturity. Genpact’s proprietary Agentic Solutions and pre-built AI models accelerate efficiency by enabling repeatability and speed-to-value for scaling AI solutions. Leveraging cross-functional teams with deep expertise in data engineering, industry processes, and technology integration, the Gigafactory is expected to create a talent pool of over 25,000 AI builders. Emphasising ethical AI design, the Gigafactory employs tools and a ‘humans-in-the-loop’ approach to ensure the safe and ethical use of AI and data. Vishal Gupta, Partner at Everest Group, highlighted the significance of this initiative, stating, “A key challenge in this journey is access to cross-functional, AI-ready talent. Genpact’s AI Gigafactory solution tackles this challenge by collaborating with technology partners and building a pool of highly skilled professionals through comprehensive training programs. This approach enables faster deployment of AI-ready professionals and accelerates value realization for enterprise clients.” The Gigafactory is supported by strategic partnerships with ecosystem players such as Databricks, enabling enterprises to integrate AI solutions seamlessly. “We see Genpact’s AI Gigafactory as a critical solution in helping companies scale their AI projects,” said Sabina Shaikh, VP, Global Systems Integrators at Databricks. Genpact recently also promoted Sreekanth Menon to the position of senior vice president of Global AI Practice and Innovation. In his new role, Menon will spearhead AI innovation and oversee the execution of the GenpactNEXT strategy. He will operate from the company’s Bengaluru office.","excerpt":"The Gigafactory will host Genpact’s proprietary Agentic Solutions and pre-built AI models to accelerate efficiency by enabling repeatability and speed-to-value for scaling AI solutions.","categories":["AI News"],"tags":["Genpact"],"author_name":"Mohit Pandey","publish_date":"2025-01-28T13:22:44","publication_year":"2025","word_count":333,"keywords":["Go","Genpact","AI","innovation","ML","RAG","Aim","data engineering","GAN","R","Databricks"],"extracted_tech_keywords":["AI","ML","Aim","RAG","Databricks","R","Go","data engineering","GAN","innovation"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/genpact-unveils-ai-gigafactory-to-create-25000-ai-builders-for-enterprise\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":47084,"title":"Why Big Companies Like Google, IBM &#038; Microsoft Are Coming Up With Free Data Science Courses","content":"Data science course today are very common — both on the internet and in the form of training centres. However, there is an inaccurate notion most people have that, free online courses are not worthy, and that they don’t have any value. Obviously, that is not true of every free course. Now, there are big names from the industry who are providing free data science certification courses. Contribution By Tech Giants Three years ago, Redmond-based tech giant started the Microsoft Professional Program with a mission to helping tech enthusiast build and upskill their expertise and knowledge to land and succeed in emerging jobs in analytics. Since the very beginning, Microsoft has been a help to numerous techies to take their skills to the next level. Google’s Analytics Academy by Google is a platform that focuses on helping analytics enthusiasts to understand and learn the company’s tools in order to grow business through intelligent data collection and analysis. When it comes to certification, it is not just about making the newcomers or the enthusiasts data science-ready; rather, it’s about considering employees and professionals as well. IBM earlier this year, in collaboration with The Open Group, IBM has started its own data scientist certification program. Since the get-go, the program seemed to be good and recently, the company has announced that the 140 of its employees have successfully completed the course are now certified data scientist. Furthermore, even though the course was first done by the IBM employees, it is also available to the public in The Open Group’s official website. Why Are They Interested In Educating For Free? These tech giants are undoubtedly providing a great platform for people to upskill and enter the data science domain. But they why they are doing so? One of the key answers is a sheer shortage of data science talent. According to a LinkedIn report, last year there was a shortage of more than 1,50,000 people with data science skills in the US. And that was the reason why IBM entered the space with its certification. Another reason which is completely undeniable is bringing more and more users to the products developed by the company. For example at the Google Analytics Academy, the courses are designed to train professionals in making the best out of Google Analytics tools. With the rise of internet activity, the need for analytics is also increasing and Google Analytics without a doubt is one of the leaders in the space. There are several companies that use the tools by Google, but a talent shortage in the domain is a big hurdle for them. Through its courses and pieces of training, Google is definitely creating a huge pool of certified professionals exclusively for Google analytics that companies could on-board for their analytics needs. Outlook Whether these big companies have entered the domain with a motive or not, whatever they are providing, it is something that is really beneficial for not only them but also for the industry and the professionals. It is without a doubt, is a win-win situation for all.","excerpt":"Data science course today are very common — both on the internet and in the form of training centres. However, there is an inaccurate notion most people have that, free online courses are not worthy, and that they don’t have any value.  Obviously, that is not true of every free course. Now, there are big […]","categories":["AI Trends"],"tags":["big data for data science","Courses","Data Science","Google","IBM","Microsoft","what is big data"],"author_name":"Harshajit Sarmah","publish_date":"2019-10-10T12:00:24","publication_year":"2019","word_count":511,"keywords":["data science","Go","what is big data","AI","programming_languages:R","big data for data science","programming_languages:Go","ViT","IBM","Courses","Google","analytics","Data Science","R","Microsoft"],"extracted_tech_keywords":["AI","data science","analytics","R","Go","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/why-big-companies-like-google-ibm-microsoft-are-coming-up-with-free-data-science-courses\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10009765,"title":"CBSE &#038; Intel Set Guinness World Records For Delivering Virtual Lesson On AI To 13K Students in 24 Hours","content":"The Central Board of Secondary Education (CBSE) and Intel India has created a Guinness World Records title on October 14th, 2020 for taking an online class on artificial intelligence for a maximum number of attendees in 24 hours. The joint record was set by CBSE along with Intel after delivering a virtual lesson on “demystifying the impact of AI” for 13,000 students from class 8 and above. The virtual class was part of the ongoing “AI For Youth Virtual Symposium” which was organised by both the institution — Intel and CBSE. When asked about this achievement, Biswajit Saha the director of training and skill education at CBSE, stated to the media that AI continues to be one of the most critical technologies of the future, which has immense potential to advance India’s digitalisation journey. He believes that it is crucial now than ever to cultivate AI-readiness in the country by empowering students with the right skillset and mindset as early in their learning journey as possible. He further congratulated both the institution, CBSE and Intel, on their commitment to promoting the right artificial intelligence-related skills and mindset among school students and setting a world record along the way. According to the official release, the AI For Youth Virtual Symposium is scheduled from October 13-17, which is witnessing participation from students, principals and teachers from across the country. In the symposium, attendees are allowed to access sessions by national and international speakers and learn various perspectives on the transformational power of this emerging technology — AI. The sessions will also cover digital building readiness and democratising emerging technologies. Attendees will also be able to interact with presenters, join in the experiential zone and attend live webinars. When asked Intel, Shweta Khurana, the director of global partnerships and initiatives stated that the partnership with the CBSE has led to several milestones in expediting AI-readiness in the country. These milestones include incorporating AI in school curriculum, setting up focused AI Skills Labs in the country and skilling facilitators across CBSE schools. She stated to the media that the enthusiastic response to the virtual AI lesson from students across the country has helped in creating a record which showcases how India’s youth is engaged and eager to explore and innovate.","excerpt":"The Central Board of Secondary Education (CBSE) and Intel India has created a Guinness World Records title on October 14th, 2020 for taking an online class on artificial intelligence for a maximum number of attendees in 24 hours. The joint record was set by CBSE along with Intel after delivering a virtual lesson on “demystifying […]","categories":["AI News"],"tags":["Intel"],"author_name":"Sejuti Das","publish_date":"2020-10-15T12:19:58","publication_year":"2020","word_count":376,"keywords":["Go","artificial intelligence","programming_languages:R","AI","programming_languages:Go","Git","GAN","R","Intel"],"extracted_tech_keywords":["AI","artificial intelligence","R","Go","Git","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/cbse-intel-set-guinness-world-records-for-delivering-virtual-lesson-on-ai-to-13k-students-in-24-hours\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10091254,"title":"MachineHack Unveils Top Performers in Google Cloud BigQuery LTV Prediction Challenge","content":"In a display of data mastery, Google Cloud and MachineHack have completed ‘The Google Cloud BigQuery LTV Prediction Challenge’. The competition, which ran from January 16 to February 16, 2023, drew the attention of almost 900 participants, each ready to prove their worth in the exciting world of data science. From preprocessing to model building, and with the use of cutting-edge programming languages like SQL and Python, the challengers put their skills to test in the hopes of rising to the top of the leaderboard. In the end, only the most innovative and resourceful contestants emerged victorious, showcasing their incredible talent and expertise. Lifetime value (LTV), a crucial metric for e-commerce platforms, was the focus of the challenge. The metric, which estimates the net profit generated from a customer connection, is a pivotal factor in companies’ advertising strategies. The challenge aimed to test data scientists’ abilities to predict LTV accurately, providing valuable insights to e-commerce companies for informed decision-making. Competitors had the chance to win enticing prizes, including a pair of state-of-the-art Google Pixel Buds, two cutting-edge Google Nest smart home devices, and $500 worth of Google Cloud Platform (GCP) credits. Additionally, the top 20 participants will receive access to an exclusive, dedicated 4–5 hour workshop on ‘Vertex AI’ by Google Cloud, a valuable opportunity for those seeking to refine their field-specific skills. The winning participants will have the chance to connect with some of the most brilliant minds in data and analytics, cementing their status as top-tier talent in this rapidly evolving field. Let’s see among nearly a thousand competitors who were able to impress the judges with their skills! Rank 01 Tajinder, a seasoned data scientist hailing from Jind, Haryana, expressed gratitude towards the organisers of the challenge. In his post-competition statement, he shared his appreciation for MachineHack and its user experience while also acknowledging the possibility of errors in his approach to problem-solving and solution writing. Furthermore, Tajinder suggested to enhance future competitions, proposing that users should have the opportunity to select the final solution for evaluation from multiple options, rather than having the best solution chosen automatically. Check out the winner’s approach here. Rank 02 Aadarsh Singh, a data scientist at Arista Networks in Pune, clinched third place. A graduate of IIT-Dhanbad, Singh fondly recalled his participation in the “Predict The Price of the Books” hackathon, which boasted over 1000 registrations. Despite the fierce competition, he managed to secure the position on the leaderboard—a significant milestone in his data science journey, he believes. Encouraged by his initial success, Singh dedicated himself to mastering data science, ultimately attaining the rank of ‘GrandMaster’ on the MachineHack platform, where he currently holds the second position on the global leaderboard. Singh praised MachineHack for its numerous features, including hackathons, practice, interview preparation, and puzzles, which he described as a “one-stop place for everything”. In his closing remarks, he expressed his deep emotional connection to the platform, stating that, for him, it is more than just a website; it is an emotion. Check out the winner’s approach here. Rank 03 Roshan Rateria, a student at the Institute of Engineering and Management, Kolkata, secured second place in the competition. Reflecting on his experience, he lauded the user-friendly nature of the MachineHack platform, which he found helpful in his learning journey. He further appreciated the platform’s messaging and learning features, describing it as a “comprehensive platform”. Rateria praised the clear, concise, and simple prerequisites, rules, and evaluation strategies of the competition, making it easy for participants to understand. Expressing his enthusiasm for the platform, he hopes that more hackathons like this would be hosted in the future. Check out the winner’s approach here.","excerpt":"The month-long competition witnessed close to 900 participants.","categories":["Deep Tech"],"tags":["data science hackathon","Data Science Hackathons","Hackathon","hackathon for data scientists","hackathon platforms","Hackathon Winners India","Hackathons","hackathons in India","Hackathons India","Machine learning hackathon","machine learning hackathons","Machinehack","Machinehack Hackathon","New Hackathon For Data Scientists","Weekend Hackathon"],"author_name":"Tasmia Ansari","publish_date":"2023-04-12T17:30:00","publication_year":"2023","word_count":610,"keywords":["GCP","Hackathon Winners India","Weekend Hackathon","Machine learning hackathon","New Hackathon For Data Scientists","hackathon for data scientists","R","data science","hackathons in India","RAG","data science hackathon","analytics","machine learning hackathons","Go","AI","Hackathon","Machinehack Hackathon","hackathon platforms","Hackathons India","Machinehack","Hackathons","Python","Data Science Hackathons","Aim","SQL"],"extracted_tech_keywords":["AI","data science","analytics","Aim","RAG","GCP","Python","R","SQL","Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/machinehack-unveils-top-performers-in-google-cloud-bigquery-ltv-prediction-challenge\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":20447,"title":"Japanese Researchers Create AI That Can Read Your Dreams","content":"A group of researchers from University of Kyoto in Japan have figured out a way where computers can visualise what you are thinking! Any guesses for how they have achieved it? Its—artificial intelligence. As a part of the study published by Guohua Shen, Tomoyasu Horikawa, Kei Majima and Yukiyasu Kamitani, they have used artificial intelligence to decode thoughts. Machine learning has been used previously to study brain scans or analysis of human functional magnetic resonance imaging, where it could also generate visualisations of what a person is thinking when referring to simple, binary images. These images are limited to the reconstruction with low level image bases. With the recent development, the researchers have found a way to “decode” thoughts using deep neural networks or artificial intelligence. It could allow scientists to understand more sophisticated “hierarchical” images having multiple layers of color and structure—for instance, a picture of a bird or man wearing cowboy hat. According to the paper published, the visual cortical activity can be decoded into hierarchical features of a deep neural network for the same input image, providing a way to make use of the information from hierarchical visual features. They came up with novel image reconstruction method where the pixel values of an image are optimized to make its DNN features similar to those decoded from human brain activity at multiple layers. “We found that the generated images resembled the stimulus images (both natural images and artificial shapes) and the subjective visual content during imagery. While our model was solely trained with natural images, our method successfully generalised the reconstruction to artificial shapes, indicating that our model indeed ‘reconstructs’ or ‘generates’ images from brain activity, not simply matches to exemplars”, noted the researchers. Deep image reconstruction: Natural images (seen images), GIF version Left: Seen images Right: Images reconstructed from brain activity (being optimized) pic.twitter.com\/YY0ZDxi7T5 — ‘Yuki’ Kamitani (@ykamit) January 4, 2018 “Our previous method was to assume that an image consists of pixels or simple shapes. But it’s known that our brain processes visual information hierarchically extracting different levels of features or components of different complexities”, they said. During the research which was carried over 10 months, three subjects were shown natural images such as photos of a bird or person, artificial geometric shapes and alphabetical letters for varying lengths of time. The brain activity was measured while the subject was looking at one of 25 images. Once the brain activity was scanned, the computer decoded the information to generate visualisations of subject’s thoughts. There are many potential applications, once the technology develops visible improvement. It can allow to draw pictures or make art simply by imagining something, it can visualise human dreams, hallucinations of psychiatric patients and much more.","excerpt":"A group of researchers from University of Kyoto in Japan have figured out a way where computers can visualise what you are thinking! Any guesses for how they have achieved it? Its—artificial intelligence. As a part of the study published by Guohua Shen, Tomoyasu Horikawa, Kei Majima and Yukiyasu Kamitani, they have used artificial intelligence […]","categories":["AI News"],"tags":[],"author_name":"Srishti Deoras","publish_date":"2018-01-09T09:55:47","publication_year":"2018","word_count":453,"keywords":["machine learning","artificial intelligence","programming_languages:R","AI","neural network","llm_models:T5","ViT","R","T5"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","neural network","R","T5","ViT","llm_models:T5","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/japanese-researchers-create-ai-can-read-dreams\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":31829,"title":"Alexa The New Love Guru Can Take Out Unpredictability In Relationships","content":"(Image credits: onlineforlove) Given the pace at which technology is evolving, the day is not so far when robots will be taking over the world. From Siri to Alexa to Cimon, artificial intelligence has reached a whole new level of advancement over the past few years. According to a recent report, by 2021, digital assistants like Amazon’s Alexa and Google Home may be able to predict our love lives — these devices will be able to tell us whether our relationship or marriage is doomed and the results will be with 75% accuracy rate. Also, it is prophesied that by 2025, by combining genetic coding with information from activity trackers, these devices will be able to tell us about the sexual chemistry between partners. Feels they are getting a bit more personal on lives? But, as per the report, it may be a reality. Alexa To The Rescue Over the past few years, the dating industry has evolved drastically, and the competition between different dating platforms has intensified with dating apps being beefed up with AI\/ML capabilities. And with Alexa getting into the play, the competition is only getting tougher. Whether you talk about Tinder or Bumble, it seems dating platforms are on a spree to make “dating” an easy-peasy thing. The question here is how a small looking device such as Echo Dot and Google Home that follows our instruction is going to work as a relationship guru. Being an AI-powered device, Alexa will listen to the conversations and come up with predictions about your relationship status. That’s is not all, another feature that sounds amazing is that if a couple gets into an argument, the device will provide suggestions and ways to settle things. Basically, it will have the capability to point out at which point of the conversation with a partner or match is proving ineffective or fizzling out. India #Represent With AI wending its way into the dating industry, the days of swiping seem to be far behind. Two of the new players that are actively using the predictive skills of AI and ML for matchmaking are Betterhalf.AI and Truematch. Recently, these India-based platforms have gained popularity in the industry and soon, they will probably be a household name when it comes to dating. Betterhallf.AI: Unlike any other dating platform, Betterhalf.AI does things in a different way. The platform uses machine learning and AI and shows verified matches working at companies through true compatibility scores that are based on relationship dimensions and their interactions on the product. Truematch: When thinking of dating apps, casual flings is the usual perception, however, with Truematch this is not true. Truematch is a cross between casual dating and matrimony. Using AI, Truematch’s search engine uses to learn about user preferences and suggest better results. Outlook We are moving towards an era where “love” will no more be a thing that connects hearts, but will be something that involves data, where a robot analyses the data and recommend matches. Today, the reality is, AI has moved beyond suggesting movies on Netflix and suggesting songs on Soundcloud. It is removing unpredictability from love and making it easier and more frequent for people. Dating is a massive market at present and with the advent of AI in the dating industry, new platforms start to enter the space and the existing platforms will definitely change the way they used to work. The day is not so far when you will have a Siri-style dating app and all you need to do is ask the app for matches and it will come up matches nearby.","excerpt":"(Image credits: onlineforlove) Given the pace at which technology is evolving, the day is not so far when robots will be taking over the world. From Siri to Alexa to Cimon, artificial intelligence has reached a whole new level of advancement over the past few years. According to a recent report, by 2021, digital assistants like […]","categories":["AI Features"],"tags":["Alexa","dating"],"author_name":"Harshajit Sarmah","publish_date":"2018-12-19T13:22:31","publication_year":"2018","word_count":600,"keywords":["Go","dating","machine learning","artificial intelligence","programming_languages:R","AI","ML","programming_languages:Go","Git","ViT","Alexa","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","R","Go","Git","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/alexa-the-new-love-guru-can-take-out-unpredictability-in-relationships\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10028049,"title":"Top 10 Talks To Check Out At SkillUp 2021","content":"Analytics India Magazine is organising the largest and first of its kind Data Science & AI education fair in India, SkillUp 2021. The two-day virtual fair is scheduled to be held on 22 & 23 April 2021. SkillUp 2021 will give an opportunity for students across India to connect with top-tier universities offering courses in data analytics & machine learning. The next-gen conference is expected to host 4,000 aspirants and professionals prospecting for the right courses; representatives from top universities; and over 25 experts from the industry. Here, we have put together the top 10 talks you can attend at SkillUp 2021. 1| Surviving to thriving: Skills for the future When: 22 April, 09:15 — 10:00 By: Harish Subramanian, Vice President at Great Learning About: The terms “remote,” “distributed,” “digital,” and “automation” have all been used to describe the future of work. This talk is a crystal ball to divine the trends driving industries, major techtonic shifts and the skills you need to be future-ready. 2| Role of statistics in the era of big data When: 22 April, 10:00 — 10:40 By:  Rajeeva Karandikar, Director, Chennai Mathematical Institute About: Statistics as a science shaped up much before Big Data came into the picture. In this talk, at SkillUp 2021, Rajeeva Karandikar will explain — with examples — how some statistical techniques have become outdated, though the concepts still hold water. He’d elaborate on how enormous computing power could be combined with statistical ideas and modern AI\/ML tools to get the best results. 3| Data science executive education for working professionals When: 22 April, 12:15 — 12:55 By: Anuradha Sharma, Dean – Delivery and Quality & Professor, INSOFE About: In today’s global economy, data science skills improve the chances of employability by leaps and bounds. INSOFE’s executive programmes, developed and designed in partnership with Case Western Reserve University and the Rennes School of Business, make data science expertise available to working professionals with no time to attend full-time courses. This talk, at SkillUp 2021, will also cover career paths for engineers and managers with 4 to 9 years of experience, as well as educational opportunities for senior leaders. 4| How to make a career in data science When: 22 April, 14:00 — 14:40 By: Ana Simion, Chief Data Officer, Sophis Inc About: Data is an integral part of the modern era, especially in the post-pandemic world.  There has never been a better time to pursue a career in data science. The interaction will offer insights into how one can enter this exciting field. The talk will also cover: Getting started with Data Science, building knowledge & best ways to learn, recommended tools and learning resources, where Data Science can take you, etc. 5| Getting ready for a career in analytics and data science When: 22 April, 15:30 — 16:10 By:  Arnab K. Laha, Associate Professor, IIM Ahmedabad About: Analytics and Data Science jobs are one of the most attractive career opportunities available. The profession has the potential to make the world a better place, in addition to decent pay packages. However, landing a data science and analytics career calls for a lot of hard work and preparation. The candidate should have a strong command in techniques and domain knowledge. Due to the rapid pace of progress in this area, the only viable model for long-term career success is learn-unlearn-relearn. This talk, at SkillUp 2021, will guide on how to get ready for a career in data science and analytics. 6| Data engineering- the fastest growing tech job When: 22 April, 17:00 — 17:40 By: Prasad Srinivasa, Assistant VP, Genpact & Sourav Saha, Dean Academics, Praxis Business School About: Data engineers oversee all data-related processes from the scratch, including where the data comes from, how it should be processed, how it can be made available at the right time and in the right format, and how it should be secured. Data engineering requires knowledge of a number of tools and techniques, including cloud and Big Data systems. This talk will take you through the nuts and bolts of data engineering and how organisations are using data engineers to deliver value. 7| How to be successful in our new AI\/ML era When: 22 April, 19:15 — 19:55 By: Giorgio Suighi, VP, Analytics, Esri About: The session will focus on the most important skills to have and\/or develop to succeed in the AI\/ML field. This talk, at SkillUp 2021, will give pointers on how AI\/ML impacts businesses in real, rather than trafficking in seasonal “buzzwords”. 8| Data Science 2.0: From outputs to outcomes When: 22 April, 20:00 — 20:40 By: Bill Schmarzo, Author, Professor, Innovator and Consultant – Data Science and Data Monetisation About: Data is the catalyst for economic growth in the 21st century. It is the responsibility of the data scientist to take on the initiative to help organisations extract and monetise the customer, product and operational insights buried in the data. The aim of the Data Science 2.0 challenge is to champion a methodology for leveraging the economic value of data and analytics. 9| Global data science masters and career opportunities in USA, Canada and Europe When: 23 April, 10:00 — 10:45 By: Dakshinamurthy V Kolluru, Founder, President & Chief Mentor, INSOFE About: There is a growing demand for domain experts with specialised data science skills. The talk will illustrate the importance of dual specialisation masters in data science. There will be discussions around INSOFE programs designed in collaboration with global universities in the US, Canada, and Europe for individuals from diverse academic disciplines. Further, the session will also touch upon the exciting career opportunities and the work visa options available to students. 10| Hands-on session on data science for beginners When: 23 April, 12:45 — 18:40 By:  Aishwarya Verma, Data Scientist, Analytics India Magazine & Krishna Rastogi, Associate Director, Association of Data Scientists About: The talk will help beginners understand different artificial intelligence landscapes, the basics of Python programming, the importance of data analysis and visualisations. This SkillUp’s session will cover topics such as: Artificial intelligence, machine learning & deep learning, introduction to python programming, hands-on linear algebra and statistics with SciPy and Numpy, hands-on data analysis, manipulation, and visualisation with Pandas and Matpltlib. Check out the entire schedule here.","excerpt":"SkillUp 2021 will give an opportunity for students across India to connect with top-tier universities offering courses in data analytics & machine learning.","categories":["Deep Tech"],"tags":["Data Science Career","data science career tips","data science education","Data science skills","education"],"author_name":"kumar Gandharv","publish_date":"2021-04-21T14:00:00","publication_year":"2021","word_count":1040,"keywords":["data science","NumPy","artificial intelligence","machine learning","AI","education","ML","Data Science Career","Aim","deep learning","analytics","data science education","data science career tips","Data science skills","Pandas"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","data science","analytics","Aim","Pandas","NumPy"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/top-10-talks-to-check-out-at-skillup-2021\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10085409,"title":"Infosys’ OpenAI Investment: A Testament to Sikka’s Forward-Thinking Vision","content":"ChatGPT has caused much sensation on the internet, propelling OpenAI, the maker of the chatbot, into the limelight. Now, a social media post revisiting Infosys’ investment in OpenAI is leaving everyone awed. Back in 2015, Infosys donated around USD 1 billion to OpenAI, which was a non-profit back then. Infosys, along with Elon Musk, AWS, YC Research, among others, joined hands to make the USD 1 billion donation to OpenAI. Why did Infosys invest in OpenAI? The donation to OpenAI, which none of us knew about, was made when Vishal Sikka was still at the helm of the IT giant. The former Infosys head was bullish on AI and much of the credit for the donation to OpenAI goes to Sikka. He also served as an adviser to the folks at OpenAI along with Yoshua Bengio, Sergey Levine, Pieter Abbeel, and Alan Kay. “Our wish is that together the OpenAI team will do unfettered research in the most important, most relevant dimensions of AI,” Sikka said in an Infosys blog post in 2015. Explaining further, Sikka said that the research ‘to be done’ by OpenAI is not time-bound, nor is it limited to just identifying dancing cats in videos. The real motive is to create ideas and inventions that amplify humanity. Back then, Sikka felt that there was a lack of fundamental progress in the field of AI which fuelled his decision to support OpenAI’s initiative. Another factor that may have influenced Sikka’s decision to invest in OpenAI was its open-source philosophy. However, this aspect of the company has undergone significant changes since then, as OpenAI transitioned from a non-profit organisation to a for-profit entity with a capped return in 2019. “If complex systems are not open, not open to be used, extended, and learned about, they end up becoming yet another mysterious thing for us, ones that we end up praying to and mythifying. The more open we make AI, the better,” Sikka said. What happened to the investment? However, in August 2017, Sikka resigned from his post as the managing director and chief executive at Infosys. With his resignation, the plan to integrate OpenAI technology into Infosys also came to a rather abrupt end. The current CEO at Infosys, Salil Parekh, confirmed that Infosys supported OpenAI through a donation. Addressing any further involvement with OpenAI in the future, Parekh said that there were no plans for any investments or involvement in any related activities with OpenAI. Sikka was a visionary Parekh has confirmed that Infosys is currently using ChatGPT with client situations, and that is starting to further increase productivity in automation. This makes us wonder, what if Sikka was still at the helm of Infosys and they had continued to expand their partnership with OpenAI? Would Infosys have back door excess to OpenAI research and their products and services, similar to Microsoft? In 2015, Sikka also said that Infosys would greatly benefit from OpenAI’s research. He believed that AI would radically change how IT firms like Infosys operate in the future. It is evident that Sikka understood the potential of OpenAI at an early stage and was ready to embrace it, highlighting this as a testament to the fact that he was a true visionary. Narayan Murthy, founder of Infosys, himself described Sikka as a technology visionary. Sikka began his tenure as the CEO of Infosys with the goal of transforming the company through the use of technology such as AI and machine learning, with a focus on software and innovation. He believed automation could decrease surplus workforce and increase efficiency and productivity. Further, Sikka also implemented programmes such as ‘Zero Distance’ and ‘Design Thinking’ to enhance sales growth and prioritise automation and operational effectiveness. In fact, by leveraging the power of modern technology, Sikka’s vision was to achieve USD 20 billion in revenues by 2020, which Infosys failed to achieve. This is amazing… https:\/\/t.co\/uMiQEoOOz7— Vishal Sikka (@vsikka) August 13, 2017 Where is Vishal Sikka now? In 2019, Sikka founded ‘Vianai’ with USD 50 million in initial funding. The startup developed its own programming language to enable more developers and organisations to utilise AI and machine learning techniques. In 2021, Vianai raised USD 140 million in Series-B funding with SoftBank Vision Fund 2. They have a range of products, including the Vian H+AI Monitoring & Continuous Operations Platform and Vianai Studio, an AI-based platform for creating and managing machine learning models. The company also provides services such as AI-based business transformation along with consulting development and training for AI products, and AI-based operations and governance.","excerpt":"Sikka said Infosys would greatly benefit from OpenAI’s research.","categories":["IT Services"],"tags":["Infosys"],"author_name":"Pritam Bordoloi","publish_date":"2023-01-18T17:00:19","publication_year":"2023","word_count":758,"keywords":["Go","ChatGPT","machine learning","ELT","OpenAI","Infosys","AI","AWS","RAG","Ray","R"],"extracted_tech_keywords":["AI","machine learning","ChatGPT","OpenAI","Ray","RAG","AWS","R","Go","ELT"],"url":"https:\/\/analyticsindiamag.com\/it-services\/infosys-openai-investment-a-testament-to-sikkas-forward-thinking-vision\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10005510,"title":"How IIT Bombay’s Virtual E-Convocation Is Such A Bold Step In Education","content":"IIT-Bombay held its 58th convocation on Sunday, and it was no ordinary convocation. For the first time in history, students collected their degrees in virtual reality mode. Due to COVID-19 restrictions, students and staff members from Indian Institute of Technology Bombay attended the annual convocation in their VR avatars. While the students were not physically present at the dais, they could be seen getting on and receiving their graduation degrees with the same enthusiasm. According to the institution, it wanted to provide the student with the same experience as an in-person event despite the restrictions posed by the pandemic. The personalised avatar of each graduate accepted the degree certificate from the personalised avatar of the Director Prof. Subhasis Chaudhuri. All medal winners took their medals from the personalised avatar of the Chief Guest. The whole event was telecast on DD India and DD Sahyadri channels and also streamed on social media including YouTube and Facebook Live. The students’ avatars brisk walk to the stage, their glistening smile after receiving the degree certificate and the virtual ambience was nothing less than spectacular. Many people on social media even argued it was better than a real event.  This is for the first time in India’s history that a college presented awards and degrees to the virtual avatars on convocation day. It took the team of 20 organisers over two months to prepare for the event, from creating personalised digital avatars to the procession. Since the last ten years, VR has taken the world by storm. In recent years more practical applications have been released, including many conferences around the world. There are some neat instances of VR that have been disrupting how an event engages the audience. In virtual reality, the computer-generated simulation of a three-dimensional image or environment can interact within a seemingly real or physical way by a person using special electronic equipment, or just spatial imagery and avatars. At present, standard virtual reality systems utilise virtual reality headsets which typically includes auditory and video feedback, but may also provide other types of sensory feedback through haptic technology. The other kind of VR is multi-projected environments that create realistic images, sounds and other scenes that mimic a user’s physical attendance in a virtual setting. The convocation at IIT-B was the latter. With avatar image-based virtual reality, people can enter the virtual scene in the form of real video and also as an avatar. One can engage in the 3D distributed virtual environment as a form of either a standard avatar or an actual video. IIT-Bombay helped connect the college mates virtually and provided viewing of hostel and campus videos on the previous night of the convocation. A link was sent out to the students helping them access the campus map and visit their respective departments as well as walk through the campus corridors virtually. P Sunthar, co-convener of the ceremony told the media that the most challenging part of the preparation was to bring together the convocation scene where the complete delegation of 38 members was supposed to sit and stand together with the national anthem. All the individual videos were shot and aggregated to make it seem that the delegation was present during the virtual event. To make this happen, each delegate was asked to come separately and walk on the dais and take their individual position. Using VR To Inspire Graduates Speaking at the event, the Director said, “Producing top-quality graduates who would be the future leaders in the industry, research and academia is the principal goal of IIT-Bombay. The skill they acquired, the work culture they formed and the friends they made during their education at IIT Bombay will support them to attain success in their lives. The whole country will forward to their contributions in taking our nation ahead”. He further said, “Giving a virtual reality experience to all our graduates was not only a very innovative step but also a great effort by our professors and staff. They did it for the students. Hopefully, this will enthuse our graduates and also other engineers in India to think high and innovatively”. In the engineering field, virtual reality has confirmed to be very beneficial for both engineering instructors and learners. An earlier expensive cost in the education department is now much more accessible. It has proven to be a handy tool for educating future engineers. The most significant factor lies in the ability for the students to be able to interact with 3D models that precisely respond based on real-world instances. This added tool of education provides the immersion required to follow complex topics and be able to apply them. As noted, future architects and engineers will benefit considerably by being able to form understandings within spatial relationships and implementing solutions based on real-world future applications, sitting from their home. Showcasing India’s Young Tech Talent With IIT-B Nurturing Prof. Duncan Haldane, co-winner of the 2016 Nobel Prize in Physics, and a Professor of Physics at Princeton University, was the Chief Guest at the virtual convocation. Stephen Schwarzman, the Chairman, CEO and Co-Founder of Blackstone, a renowned investor and philanthropist, joined the convocation as the Guest of Honor. Presenting the Institute’s Report for 2019-2020, Professor Subhasis Chaudhuri, Director of IIT Bombay, told that the institute continues to be a much sought-after place for undergraduate and postgraduate studies. In his speech, Stephen Schwarzman stated, “India remains in a unique position in the world now, particularly when it comes to technology, where it has built global leadership with its talent. Out of the 72 Indian-origin engineers who have founded unicorn startups over the world, 50 per cent are IIT alumni. IIT engineers are setting the global innovation landscape, and the newest alumni will be the next generation of prospective leaders who will drive that mission”. He also said that with a young, ambitious and tech-savvy population, a forward-thinking government that promotes entrepreneurship, and the fourth-largest startup ecosystem in the world, India is poised to stake its title as one of the world’s preeminent innovation hubs. The graduating students this year included 381 PhDs, 18 Dual Degrees (MTech\/ MPhil + PhD) and 27 Dual Degrees (MSc+PhD). Of these, 39 research scholars were chosen for the award of ‘Excellence in PhD Research’ for the year 2018-20. Plus, 33 joint PhD degrees, in association with Monash University were also bestowed by the Vice-Chancellor and President of Monash University Professor Margaret Gardner, during the event. In addition,  621 MTech, 11 MS (by research), 6 Dual Degree (MSc+MTech), 20 MPhil, 110 MMgt, 64 MDes, 225 two-year MSc, 2 five-year integrated MSc, 684 BTech Degrees,  6 Dual Degrees (BS+MSc), 342 Dual Degree (BTech + MTech), 10 Interdisciplinary Dual Degrees (BTech\/ BS+MTech\/ MSc), 20 Dual Degrees (BDes+MDes),  16 BS, 15 BDes, and 16 PGDIIT, were awarded at the Convocation.","excerpt":"IIT-Bombay held its 58th convocation on Sunday, and it was no ordinary convocation. For the first time in history, students collected their degrees in virtual reality mode.  Due to COVID-19 restrictions, students and staff members from Indian Institute of Technology Bombay attended the annual convocation in their VR avatars. While the students were not physically […]","categories":["IT Services"],"tags":["IIT","IIT Bombay"],"author_name":"Vishal Chawla","publish_date":"2020-08-24T18:18:02","publication_year":"2020","word_count":1132,"keywords":["Go","IIT Bombay","unicorn","programming_languages:R","AI","innovation","programming_languages:Go","Git","GAN","IIT","R","startup"],"extracted_tech_keywords":["AI","R","Go","Git","GAN","innovation","startup","unicorn","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/how-iit-bombays-virtual-e-convocation-is-such-a-bold-step-in-education\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10099017,"title":"OpenAI and Scale Join Forces to Fine-Tune GPT-3.5 for Enterprises","content":"OpenAI announced that it is partnering with San Francisco based data labeling startup Scale to help more companies benefit from fine-tuning its text generating model GPT-3.5. The company in their blog post stated that  it is working with Scale as a preferred partner to extend the benefits of our fine-tuning capability given its experience helping enterprises securely and effectively leverage data for AI. OpenAI further said they have extended the opportunity to Scale customers to fine-tune OpenAI models, mirroring the process they would follow with OpenAI. Moreover, these customers stand to gain from Scale’s proficiency in enterprise AI and the utilization of their Data Engine. “Scale extends our ability to bring the power of fine-tuning to more companies, building on their enterprise AI experience to help businesses better apply OpenAI models for their unique needs.”said Brad Lightcap, COO, OpenAI. OpenAI a day earlier had  announced that fine-tuning for GPT-3.5 Turbo is now available and fine-tuning for GPT-4 is coming this fall. OpenAI has stated that the fine-tuning of GPT 3.5 Turbo is suitable specifically for businesses and developers to customize the model depending upon their use case as it lets them train the model on company’s data and run it at scale. Founded in 2016, Scale’s Generative AI Platform leverages enterprise data to customize powerful base generative models to safely unlock the value of AI. “We are excited to partner with OpenAI to supercharge model performance – helping every enterprise utilize AI most effectively for their unique needs. Prompting alone—atop even the best LLMs like GPT-3.5 — is not enough model customization to produce the most accurate, efficient results. As with software, an incredible amount of value comes from fine-grained optimizations, and fine tuning is critical for that.” said  Alexandr Wang, Founder and CEO, Scale AI Scale has earlier worked with American financial service and technology company Brex. Scale in their blog post claimed that by using the GPT-3.5 fine-tuning API on Brex data annotated with Scale’s Data Engine, they saw that the fine-tuned GPT-3.5 model outperformed the stock GPT-3.5 turbo model 66% of the time.","excerpt":"Scale has earlier worked with American financial service and technology company Brex","categories":["AI News"],"tags":[],"author_name":"Siddharth Jindal","publish_date":"2023-08-24T21:57:45","publication_year":"2023","word_count":346,"keywords":["API","OpenAI","AI","RAG","GPT","Aim","llm_models:GPT","generative AI","R","startup"],"extracted_tech_keywords":["AI","generative AI","OpenAI","Aim","RAG","R","API","GPT","startup","llm_models:GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/openai-and-scale-join-forces-to-fine-tune-gpt-3-5-for-enterprises\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":49059,"title":"State of Enterprise AI In India 2019 | Analytics India Magazine &#038; BRIDGEi2i","content":"“A year spent in artificial intelligence is enough to make one believe in God”– Alan Perlis, early computer science pioneer It’s futile to deny it, but Artificial Intelligence(AI) is no longer the buzzword of tomorrow, it’s a striking reality of today, and the enterprise landscape of AI has never looked more promising than it does today! By 2022, the global business value created by AI will touch a whopping $3.9 trillion, and spending on AI systems is expected to reach $79.2 billion1. Forecasts estimate that AI technologies will pervade every software product2 next year, and AI software revenue is expected to grow to 118.6 billion by 20253. All these are tantamount to the fact that AI is no longer just a differentiator but a core part of business functions! In India, the enterprise AI market is heading towards much wider adoption. An industry expert associates the Indian Enterprise Market for AI to be estimated to be $100 million, growing at 200-250% CAGR. Futuristic growth of this sort clearly underscores the potential in the big revolution that business leaders should prepare for! AI is increasingly being used by software vendors and AI solution providers as embedded products and services to deliver more value across a host of business problems. This journey has roots in core business applications like ERP & CRM, but today, almost every sector is using AI in their auxiliary processes as well, such as customer support, recruitments sales, or marketing. For example, a majority of banks and insurance companies in India have AI-driven chatbots that are fast becoming the first point of customer interactions. While the scale and complexity of these implementations may vary, it’s soon becoming the norm in the market. C-suite executives have gone beyond committing ‘digital experimentation’ to hardcore Digital Transformation. This has led to a surge in demand for enterprise-ready AI services, applications, and tools across organizations. There are plenty of forerunners in the market: Enterprise AI solutions like the IBM Watson platform, which is portable across any cloud, is used by customers to manage multiple customer service touchpoints like chatbots, email or phone, forecast inventory and demand, or improve customer care. Salesforce Einstein empowers sales and support while Google’s latest offering — Contact Center AI provides the best of Google’s AI and machine learning solutions with software that allows businesses to improve customer experiences and operational efficiency at the same time. As the industry evolves, certain sub-sectors that are creating disruption have risen to prominence; Enterprise AI has made inroads in areas like diagnostics (healthcare), salesforce automation (CRM), automated trading (financial services) and anomaly detections (oil & gas, utilities). We believe as the adoption grows, AI will become increasingly interdependent on core business functions. “It’s not just AI, but AI+X that is going to be the game-changer, X being your core business function. The future is about AI being tightly coupled with all our everyday tools until it becomes an integral part of existing tools and processes” Ramprakash Ramamoorthy, Product Manager, Zoho Labs As companies look for additional capabilities to turbocharge innovation — we believe large-scale AI consultancies will play a critical role in bringing in pre-built customizable solutions that can drastically reduce the time-to-market and speed up innovation. With deep domain expertise and trained workforce, AI consultancies are better positioned to educate customers, offer pilots, and scale up the use cases. According to our  2017 report on the State Of Analytics In Domestic Firms In India4, captives account for 75% of the Indian analytics market, while service providers account for 5% market share. With the adoption of AI solutions increasing, we foresee rapid expansion in the number of use cases tailored for specific business functions creating an opportunity for AI consultancies to bring deeper relationships and in-depth domain knowledge to the table. “Unlike earlier technology cycles, we see India being at par with the rest of the developed countries concerning technologies such as AI, ML, and others. Indian companies have mandated their CIOs for effective adoption of cognitive just as much as the focus is on the cloud” Anil Bhasker, Business Unit Leader Analytics and Big Data, India\/South Asia, IBM India “We are seeing some of India’s largest IT firms utilizing AI extensively internally or with consulting companies. With the potential to revolutionize both the services offered by consulting firms to their clients and how consultants work, AI is imperative”Hemal Shah, Senior Vice President & Regional CIO, Asia Pacific, Dell Technologies “We’ll continue to see increasing adoption of enterprise AI solutions as organizations further optimize their IT infrastructure, striving for faster digital product and service development. There will be continued adoption of end-to-end enterprise data management to serve as the foundation for enterprise AI and build an ecosystem of change across analytics and business processes. Also, the growth of digital platforms, which is making it easier for organizations to develop AI applications, will drive business model innovation and optimize the use of Big Data, AI, and IoT” Ronald Van Loon, Digital Transformation Influencer The Re-envision of Digital Transformation is responsible for the flutter in global markets. We need to take a moment to comprehend that digitization isn’t merely changing business models or creating new businesses, but it’s about keeping up with faster and better techniques of accessing, utilizing, and getting value out of the existing tonnes of data. The speed with which enterprises are getting onto the digital bandwagon speaks of the critical urgency in which transformation initiatives are being carried out by organizations. But despite the buzz, organizations aren’t able to fully seize the opportunity presented by AI and turn it into actionable results. IT leaders across sectors face tremendous challenges at the start of their AI journey. While we often hear about AI PoCs advancing from the project stage, success is limited as deployments often fail to turn into actionable items. In addition to this, there is no established playbook for enterprise leaders to follow. Most organizations are very early in this paradigm shift, and while CXOs know there’s value in AI, they are nervous about making bets. The reason is that to unlock real value from AI, there has to be a perceived tangible return on investment (ROI), or the technology has to be assessed against the key performance indicators. Another hurdle is that most companies lack the AI-specific skills to ‘do it alone’ and lack the resources to launch test-and-learn cycles. In addition to this, leaders are also looking to embrace AI because they don’t want to fall behind in the AI journey. Against this backdrop, our report’ State of Artificial Intelligence in India 2019′ with BRIDGEi2i takes stock of the enterprise AI market in India, the AI player landscape, low-risk, revenue-generating PoCs that organizations can get started with and the rise of AI-as-a-service economy. What Does the Report Cover? Large-scale advancements in AI over the last five years have presented tremendous opportunities for companies to transform the customer experience, automate business functions, and broaden their product offerings. To provide a more informed view of the Enterprise AI market in India, we decided to perform our own research into how users are adopting AI technologies. The report also offers a snapshot of the current state of the rapidly changing AI industry, looking through the lenses of suppliers and consumers. The report looks at the scale of opportunity in AI for large-scale organizations that are driving the AI ecosystem in India and how the C-suite can take advantage of PoCs that can deliver the best ROI. It is crucial for business leaders to have a thorough understanding of the AI ecosystem and to target the right PoCs, which can provide the maximum ROI. The report answers questions around How AI\/ML is finally getting real for enterprisesHigh-value use cases enterprises can get started withHow to measure tangible results of AI deploymentWhy India is Poised for Growth in the AI Market \/ India’s Contribution to the Global AI Industries Who Should Read This Report? This report is aimed at Executive Leaders, IT Decision Makers, senior managers, data science heads, and investors tasked with the responsibilities of driving digital transformation, innovation heads, or enthusiasts building delivery capabilities and CoEs. Key Market Dynamics According to a PwC report, “The business opportunity that AI provides is so vast that by 2030, the global GDP is estimated to 14% higher standing at $15.7 trillion just as a direct result of AI. This will be augmented by organizations trying to develop an array of AI-based services that can scale across a wide spectrum of software development, different industrial applications, and use cases.” The Indian AI sector has seen a total investment of$150 million5 in more than 400 companies over the past five years.Industry body NASSCOM6 indicates over 1,200 new advanced technology startups got added to the ecosystem in 2018, with data analytics being the most significant contributor.Due to its sheer size, BFSI is the largest adopter of AI followed by Healthcare and Logistics. There is an increase in demand for products and services which can attract more investment towards the R&D in the AI sector.With conglomerates and enterprises having a big share in India’s market, there is huge scope for AI-based enterprise solutions in the country. Key Takeaways C-suite executives can gain an overview of Enterprise AI landscape and key playersOverview of AI Service Delivery models and why AI as a service economy will drive the market forward Top use cases with high re-usability & ROI Key criteria AI Service firms must meet Enterprise AI landscape in India For many Indian organizations, the rapid rise of AI has become a top corporate agenda with organizations deeming it a critical part of their organizational strategy. IT Decision Makers and Business Leaders want to take advantage of the exponential growth in data and cloud computing. The news cycle is abuzz with Indian conglomerates and new-age companies who are adopting AI technology at large scale and making strategic investments in technical infrastructure and in building the right talent. “In a nutshell, several factors can be attributed to be pushing the growth of AI in India, including the availability of massively organized and cleansed data sets, the growing sophistication of machine learning and self-improving algorithms and the emergence of hardware”Prithvijit Roy, CEO, and Co-Founder of BRIDGEi2i a) Market Size As per a report7, the AI market was valued at $21.5 billion in 2018 and is likely to reach $190.6 billion by 2025. Meanwhile, the global enterprise AI market is valued at $796.38 million and is expected to reach $9880.4 million by 2023. Research by Accenture9 insists that AI has the potential to add US$957 billion to India’s economy in 2035 if the AI Revolution receives the right support among enterprises, business leaders, and policymakers. India’s AI startup ecosystem is booming with a number of startups working in the domain of machine learning, computer vision and NLP. More than 50 percent of firms in India are working on advanced analytics and computer-vision based AI technologies. India is contributing significantly to the data labelling market, which is where humans teach machines to recognize familiar human patterns, and this business is expected to reach $1.2 billion by 2023, according to the research firm Cognilytica. b) Investment India has stepped up the AI game and saw an investment of $73 million10 in 2017. As per our research in 2018, startups with operations in India and globally raised approximately USD$ 529.52 million in funding rounds, and this data includes startups with investment at varying stages of development, from pre-seed to well-funded companies. As per Gartner, the global business value derived from artificial AI is estimated to be $1.2 trillion in 2018, a 70 percent increase from 2017. c) Adoption by Industry AI In Finance: As per our report, the size of the analytics industry in the financial sector is currently estimated to be $1.2 billion (annual) in revenue. Financial institutions have achieved exponential growth and are driving innovation in the industry by building an enterprise-wide analytics capability that is now woven into the key business processes throughout the organization. Banks and FIs are now investing in several vital dimensions – technology infrastructure, strengthening processes, and people to build sophisticated analytics capability. Some of the top players in this segment are HSBC, American Express, ICICI Bank, Moody’s Analytics Knowledge Services, Citi, JPMorgan Chase & Co., HDFC Bank, Axis Bank, Paytm and PhonePe among others. IBM and HDFC ERGO General Insurance Company, India’s third-largest non-life insurance provider in the private sector, are collaborating to co-create new AI-based solutions on IBM Cloud, that will redefine customer experience in India. Leveraging IBM Garage11, teams from HDFC ERGO and IBM Services work together to develop and test new solutions that will help to address customer inquiries better, ensure faster turnaround time and draw deeper customer insights for a better omnichannel experience. AI In Healthcare: The adoption of AI is reshaping the Indian healthcare market significantly. AI-enabled healthcare12 services like automated analysis of medical tests, predictive healthcare diagnosis, automation of healthcare diagnosis with the help of monitoring equipment, and wearable sensor-based medical devices are expected to revolutionize medical treatment processes in the country. The applications of AI in the healthcare space will be worth INR ~431.97 billion by 2021, expanding at a rate of 40%. Top players in this segment are Niramai, Sigtuple, Qure.ai, Tricog Health. Bengaluru-based startup Niramai’s diagnostic platform is now using thermal image processing and machine learning algorithms to enable accurate breast cancer screening. Hospitals are now spending only one-tenth the cost on Niramai hardware compared to mammography machines that cost around ₹1 crore. AI In E-commerce: The E-commerce market in India is well-placed. One of the fastest-growing markets in the Asia Pacific driven by innovations in personalization, social media analytics, omnichannel service and sharing economy business models. In 2018, e-commerce and consumer internet companies raised over US$7 billion in private equity and venture capital in 2018, EY report13 indicated. The interplay of technologies — analytics, AI, cloud, digital, mobility, social and virtualization are driving the industry forward. The innovation is driven by Amazon which committed $5 billion investment in India and Walmart-owned Flipkart which acquired Liv.ai in 2018 to reach the next million users. Some of the key areas where AI is leveraged in the e-commerce sector are recommendation engines, virtual assistants, predictive sales and warehouse automation. Top players in this segment are Amazon, Flipkart, BigBasket and consumer internet majors Oyo, Swiggy, Zomato, Byjus that are disrupting the landscape. India’s largest hotel chain Oyo has a Dynamic Room Pricing model that finds the optimum price point to maximize the overall yield – the combination of price and occupancy. The team watches out for hundreds of signals on the demand — from traffic patterns to conversions, upcoming events, data from offline sources and even historical occupancy rates to ensure no room goes vacant. Based on predictive occupancy, the demand exceeds significantly, thereby resulting in an increase in room prices. AI In Retail: India is Asia’s third-largest retail market and the world’s fourth-largest after the US, China and Japan. The adoption and use of AI in this sector is on a rise with a significant majority of retailers in India deploying AI or automation technology not just for decision-making but also as part of their operations. At the storefront and behind-the-scene in fulfilment centres, retailers are looking to save costs and boost revenues by deploying AI and automation. AI is deployed in customer-facing aspects and optimizing the supply chain backend to enable web-based sales. While big retailers in India — Future Group, Shoppers Stop, Reliance Retail and Tata Group are still to hit the mature AI adoption curve, the adoption of technologies represents a big leap forward for the sector. Globally, AI represents a $300 billion+ opportunity for retail companies. Deploying and scaling AI should be the next big objective for these retail majors in India. Every season, a leading fashion brand launches thousands of products and expects them to get sold at full price. But in reality, they end up discounting all along with End of Season Sales accounting for almost 50% of the total sales. GoFrugal built an AI-based recommendation engine that suggests salesperson how much discount to offer as per the customer profile. By looking at the customer profile, their prior purchase pattern, the engagement with the store, salesperson and product (looks, touch, pick, try) — the ML algorithm can recommend an offer which has been created for just that one garment for that one customer and valid for just that moment. The solutions manage the problem of too much discounting or too less discounting. Bestseller — the leading fashion company that owns brands like Jack & Jones, Vero Moda & others uses IBM Watson AI capability to predict the right merchandise for the consumer at the right time. With Watson mining deeply into big data, the retailer can determine the right assortment plan for each store, predict the next best product to incorporate into its mix, and improve the efficiency of its supply chain. They are working with IBM Watson AI to predict the next big trend and the most relevant styles, colours and size ratios. Higher relevance means a sharper, better-selling assortment, helping them meet consumer expectations while becoming more efficient. AI in CPG: “The world is moving into new ways of doing businesses. Ecosystems of consumption have been formed. There are 4-5 ecosystems formed by Amazon, Alibaba and Tencent in America and China. We are building our own ecosystem with payments, wallet and insurance. In the future, there will be more consolidation in order to form the right ecosystems, he said. “We want to make our ecosystem strong with alliances and partnerships.”Kishore Biyani, Future Group CEO The Consumer Packaged Goods market sees a promising growth in the use of AI both globally and at India level. With AI-powered decision-making systems and recommendation engines being developed on a large scale, the industry is ripe for transformation. Experts and decision-makers in the CPG companies can rely on competitive pricing, prevention of customer churn, and optimization of budget allocation in marketing to enhance their margins at the outset. A BCG report states that CPG Companies are leveraging advanced analytics and AI solutions for local assortments, personalized consumer services and experiences, optimized marketing and promotion ROI, and faster innovation cycles. While there are many brands in India who have adopted AI-powered solutions, Nestle India stands out for an interesting experiment. They introduced NINA, a virtual nutrition assistant in collaboration with an Indian chatbot service provider which could interact with users in a human-like manner and offer real-time, personalized advice on nutrition that is balanced, scientifically correct and customized to their unique needs. This was a huge hit – and a different campaign from the rest. d) Partnership Ecosystem “We don’t see large enterprise customers going to specialised AI vendors, rather they will expect their current vendors to adopt\/bring AI features to their existing solution stack. As compared to established IT companies and digital disruptors, AI Consultancies and service firms are better positioned to educate the customers, offer pilots and scale up the use cases.”Kumar Vembu, CEO and Co-Founder, GoFrugal Partnership ecosystems open a great window of opportunity for organizations around the globe to scale fast and seize the opportunity to drive revenue growth and develop innovative business models. Today, CIOs face incredibly high expectations not just to enable digital transformation, but to build sector-specific solutions\/services that can leverage new digital technologies. These digital technologies — AI\/ML are also enabling companies to move into adjacent markets and drive revenues. There is a huge scope of opportunity in traditional sectors like banking, insurance, retail and other industry verticals like manufacturing, automotive and logistics to move fast in order to sustain innovation. Large multinationals and companies are finally seeing the value in developing these partnerships to define their own digital strategy and build new business model innovation. Digital has changed the rules of engagement: We are seeing businesses pursue two distinct approaches to digital transformation — outside-in and inside-out. While an outside-in approach is largely driven by the market and demand for new digital services, an inside-out approach is about modernizing the core systems and architecting their business for change. This outside-in approach results in an ecosystem expansion, leading to scope for partnerships with startups, subject matter experts, stakeholders to fill capability, domain expertise, talent gap and to improve the overall corporate strategy. This presents a huge opportunity for AI consultancies and technology providers to capitalize on this trend by expanding their role in building partner ecosystems and collaborate with companies to create new revenue streams. In the context of AI, the ecosystem can be built for many things — deliver best-in-class products\/services, engage diverse participants to build talent, strengthen relationships with peers and add to the diversity of industries. “You can’t drive Transformation alone. You need partners, co-innovation with customers, continuous re-skilling of talent”Tiger Tyagarajan, CEO, Genpact Organizations are leveraging partnership ecosystems by: Focusing on creating new revenue streams, driving business growth through collaboration with AI Consultancies\/Technology PartnersCollaborating with technology partners to plug the capability gap and kickstart business model innovation Leveraging partnerships to improve business efficiency internally and remove data silos India’s third-largest bank — Axis Bank went a step ahead and launched Thought Factory, an Innovation Lab in Bengaluru to give Axis Bank a fintech advantage and a better understanding of today’s “technologies and better focus of tech solutions”. The innovation lab partners with fintech startups to deliver the much-needed agility to stay ahead of the curve. Note: This is the First Part of a three-part series of our study ‘State of Enterprise AI In India 2019’ brought to you in association with BRIDGEi2i. Stay tuned for Part II and Part III. Check Part II of the three-part series here. Check Part III of the three-part series here. State-of-Enterprise-AI-in-India-2019Download","excerpt":"“A year spent in artificial intelligence is enough to make one believe in God” – Alan Perlis, early computer science pioneer It’s futile to deny it, but Artificial Intelligence(AI) is no longer the buzzword of tomorrow, it’s a striking reality of today, and the enterprise landscape of AI has never looked more promising than it […]","categories":["AI Features"],"tags":["analytics insights platforms real-time business","automotive analytics","automotive data analytics","big data and analytics everyday life","china ai investments","cognitive computing human capital","latest ai products across industries","salesforce crm","strategic salesforce"],"author_name":"Richa Bhatia","publish_date":"2019-10-30T16:37:03","publication_year":"2019","word_count":3632,"keywords":["computer vision","china ai investments","Ray","cognitive computing human capital","data science","artificial intelligence","NLP","analytics insights platforms real-time business","salesforce crm","analytics","latest ai products across industries","big data and analytics everyday life","machine learning","AI","ML","automotive analytics","Aim","strategic salesforce","automotive data analytics"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","NLP","computer vision","data science","analytics","Aim","Ray"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/state-of-enterprise-ai-in-india-2019-analytics-india-magazine-bridgei2i\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10140429,"title":"‘DeepMind Might Not Have Succeeded if We Started Just a Few Years Earlier or Later’","content":"On his first visit to India, Microsoft AI CEO Mustafa Suleyman took everyone by surprise when he introduced an AI companion for everyone. At the Building AI Companions for India event held in Bengaluru, he urged Indian enterprises to embrace the Copilot wave and said, “Now is the right time to create these new models… All of the resources are now widely available. The APIs are brilliant. There are open source models happening everywhere.” Case in point: Reflecting on DeepMind’s inception in 2010, Suleyman said: “Timing is everything… It’s critical that you get the timing right.”He added that if he and his co-founder Demis Hassabis had launched DeepMind either a few years earlier or later, they might have missed the pivotal wave of deep learning advancements, potentially altering their success. According to Suleyman, AI is constantly evolving and becoming more sophisticated. “For too long, software has principally been utilitarian. My personal vision for AI has always been about how it can be a companion that can make each and every one of us feel more supported and smarter and more capable,” he said. India Praised: Suleyman said that India was one of Microsoft’s fastest-growing markets and has one of its strongest R&D teams globally. “We have extremely talented engineers and developers, and increasingly, we are involving social scientists, psychologists, therapists, scriptwriters, comedians, and other creatives,” said Suleyman, adding that this diversity allows the country to synthesise more perspectives and get a broader picture of people involved in the design and creation process. “AI is going to put knowledge at everyone’s fingertips, synthesised, distilled, personally tuned to the way that you want to learn and use information,” he said, highlighting AI’s potential to democratise knowledge, making it accessible across work environments and enabling informed decision making. S Krishnan, secretary of the Indian government’s Electronics and Information Technology Ministry, also touched upon the importance of voice and linguistic inclusivity, and how it could drive AI adoption in the country. “We try to do conversation in 22 different Indian languages… Voice really is the ultimate way to make these tools accessible,” he recalled, in a fireside chat with Suleyman. Balance is Key: Suleyman urged for a balanced approach to AI, advocating for careful scrutiny without losing sight of its significant benefits. “It’s healthy to ask difficult questions about the labour market or about privacy and security,” said Suleyman, stressing that it is also our collective responsibility—as we really care about civilisation. However, he said that it is important to avoid overreacting prospectively and miss out on the obvious benefit of the system that delivers productivity, education, wealth, well-being, healthcare, as well as access to legal and medical advice.","excerpt":"“Timing is everything,” says Mustafa Suleyman, CEO of Microsoft AI.","categories":["AI News"],"tags":["DeepMind","Mustafa Suleyman"],"author_name":"Aditi Suresh","publish_date":"2024-11-07T01:16:09","publication_year":"2024","word_count":445,"keywords":["Go","API","Mustafa Suleyman","programming_languages:R","AI","programming_languages:Go","deep learning","ViT","R","DeepMind"],"extracted_tech_keywords":["AI","deep learning","R","Go","API","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/deepmind-might-not-have-succeeded-if-we-started-just-a-few-years-earlier-or-later\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":61105,"title":"Is It Possible To Become A Successful Self-Taught Data Scientist?","content":"While there seems to be advocates of both formal and informal education in the data science community, what does the general hiring landscape suggest? We may be living unique lives, but there are a few common experiences that tie most of us together. While the instinctive response seems to be what the Covid-19 pandemic has left in its trail, a less grim example would be the chain of events that mark a few decades of our lives – school -> college -> job While most of these remain unchanged, there has been a growing clamour to rethink education systems as we know it; to one, that both shields learners from the negative effects of new technologies in the workplace, as well as reskills them to prepare for new cross-functional roles that will inevitably be created as a result. While one way of addressing this talent and labour market issue would be to enrol and improve the quality of education, upskilling in silo would be a more efficient way of achieving this without burdening the formal education system across the world. Tied to lifelong learning, upskilling through relevant certification is increasingly becoming an accepted norm. But the field of data science is faced with a critical question – can its vast universe be merely acquired through self-study? Hiring Trends In Data Science While there seems to be advocates of both formal and informal education in the data science community, what does the general hiring landscape suggest? While IT services company CSS Corp places an ‘inestimable value’ on a formal academic background, LinkedIn admittedly hires candidates based on skill sets alone, and not necessarily on their educational background. However, that is not to say that self-taught data science aspirants should not prepare for an arduous journey ahead. Although tech-centric roles demand ingenious techniques and in-depth knowledge of relevant domains, it is often challenging to prove the same to future employers. How can self-taught data science aspirants overcome this barrier to become successful in this field? In addition to upskilling, there are several other things that they need to focus on. Upskill Through Relevant Certifications Although a university degree is a great accomplishment, self-taught aspirants can rejoice as this is not enough to land a good data science job. While a degree may lay down a foundation for a career in this field – and may get one a job interview – it is not a key qualifying factor when applying for tech positions. Even though you may be competing against applicants who have relevant degrees, you can garner a competitive advantage with upskilling using the world of resources available online. What is more, self-study also signals a candidate’s motivation to succeed. ALSO READ: “You Don’t Necessarily Need A Resume To Get A Data Science Job” But you need to first narrow down what you need to learn to substitute for your lack of formal training. Data science is a broad discipline and comprises a wide collection of jobs – from  statisticians to machine learning (ML) experts, to business analysts to data visualization experts. Since the skills required for each vary, it is important to first narrow down the skill sets you need to acquire, and then create a plan around it. Since companies want to know what you as a candidate can offer them, it is just as important to present these skills to show that it will provide value. Apply Your Data Science Skills The next step would be to apply your skills in projects that not only gives you the opportunity to practice, but also gives potential employers an idea of the breadth of your knowledge on the practical use of these skills. Create a data science portfolio on platforms like Kaggle and GitHub – in fact, there is an emerging trend where tech companies directly scour these online portals to find the right talent. Such platforms give you the opportunity to build a portfolio of work that can help you convincingly make the case for your competence. ALSO READ: “You Cannot Become A Data Scientist Without Mastering The Skill Of Storytelling” While startups typically employ such non-traditional ways of hiring candidates – and it is often in these firms that a lack of degree is likely to matter less – even big tech companies have begun to explore these avenues. Establish Work Experience, If Any While acquiring the right skills and applying them to manufactured situations through online projects has become a prerequisite, actual corporate experience will hold you in good stead, especially in a roomful of applicants who do not come with the same experience. You can begin by collating practical examples of how you provided business value to companies through short-term freelance gigs. Relevant experience in a corporate environment will give you bonus points in interviews, since you already demonstrated that the skills you acquired came in handy to someone else. The extra emphasis on work experience in data science is because unlike other disciplines, it does not afford you the chance to see step by step of what is happening – it either works or it does not, and is based on the algorithm running through all the data. ALSO READ: “A Data Scientist Cannot Grow Without Organization’s Support” Learning this way can be challenging, unless you have the kind of data intuition that comes with experience. Without that, you also end up making a lot of mistakes that could easily have been avoided. Furthermore, processing large data sets for an actual company is a lot different from tackling constructed ones, since the former change over time as well. It also requires seamless collaboration with engineers and other departments, and a deep understanding of the business value of such an undertaking – which you can only develop over years of professional experience. Take Help From Mentors Although this is not a prerequisite, for the uninitiated, since data science is not an easy discipline to learn, it may be advisable to take a little help when planning your study – as well as career – around this sector.Mentors – or anyone with experience – can add a lot of value to your self-study process. They can help identify blind spots in the beginning itself and guide you to a path to a more effective learning experience.","excerpt":"While there seems to be advocates of both formal and informal education in the data science community, what does the general hiring landscape suggest? We may be living unique lives, but there are a few common experiences that tie most of us together. While the instinctive response seems to be what the Covid-19 pandemic has […]","categories":["AI Highlights"],"tags":["become a data scientist","big data certification","covid-19 data visualization","data analyst certification","Data analyst jobs","data analytics certificate","Data Analytics Certification","Data Science","Data Science Certification","Data Scientist","Data Scientist Jobs","deep learning projects","how to become a data scientist","trends in bi and data visualization"],"author_name":"Anu Thomas","publish_date":"2020-04-07T16:00:00","publication_year":"2020","word_count":1049,"keywords":["trends in bi and data visualization","Data Science Certification","Git","GitHub","Data Analytics Certification","R","data analyst certification","data science","Data Scientist Jobs","become a data scientist","ViT","covid-19 data visualization","Data Science","Go","how to become a data scientist","machine learning","AI","ML","Data analyst jobs","GAN","big data certification","data analytics certificate","deep learning projects","Data Scientist"],"extracted_tech_keywords":["AI","machine learning","ML","data science","R","Go","Git","GitHub","GAN","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/is-it-possible-to-become-a-successful-self-taught-data-scientist\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":65890,"title":"Azure Build 2020: Key Announcements For Cloud","content":"At Build 2020, Microsoft has announced some newly expanded capabilities for Azure Cloud. These new tools and services available empower developers to be more collaborative and productive on the cloud platform. The announcements made at the conference signals the direction in which Microsoft is proceeding, as cloud-based innovation and new tools can enable developers to build powerful applications. In this article, we take a look at the announcements for developers and companies to leverage the cloud for innovation. Azure Arc Enabled Kubernetes Microsoft has delivered Azure Arc powered by Kubernetes in preview to its clients. With this, businesses can use Azure Arc to connect and configure any Kubernetes cluster over numerous infrastructure and many application scenarios and deployment environments in the Kubernetes ecosystem. Microsoft also announced the preview of Azure Arc, an innovation that opens up new hybrid scenarios for enterprise users by delivering Azure services and management to all infrastructure over edge, datacenters, and multi-cloud. Azure Arc is a set of technologies that extends the control plane of Azure out to on-premises, multi-cloud environments and edge. The solution is a powerful tool and gives a lot of freedom on the cloud. Microsoft’s customers will benefit by having a cloud-native control plane to inventory, organise, and enforce policies for their IT resources wherever they are, from a central place. For Azure Arc, Microsoft has integration partners like Red Hat OpenShift, Canonical Kubernetes, and Rancher Labs to assure Azure Arc runs well for all the major platforms that companies are utilising. Users can link any Kubernetes version of their choice and work with them alongside other Azure data services. Cloud Supercomputer Microsoft also announced one of the world’s most powerful AI supercomputers built on the Azure cloud platform. Developed in collaboration with and exclusively for OpenAI, this supercomputer is purpose-built to train massive distributed AI models, giving it all the benefits of a dedicated appliance paired with the benefits of Azure’s modern cloud infrastructure. According to Microsoft, it is one of the top 5 supercomputers in the world. It uses 285 thousand CPU cores and 10k GPUs connected with an extremely fast network. The cloud supercomputer is built for massively distributed models and gives the cloud-hosted supercomputer the benefits of a dedicated appliance, paired with the benefits of robust modern cloud infrastructure. Because it is cloud-hosted on Azure and can benefit all Azure customers unlike other supercomputers out there. Customers can have large ML models, this computing power can unlock great power for the Azure developer community who will be using it to build new products and services. Azure Synapse Link Synapse Analytics is Microsoft’s single managed service for analytics over your data lake and data warehouse, using either serverless or provision compute. And now, with Azure Synapse Link, this extends to your operational data sources. For real-time analytics, many firms provide operations and workloads in the same system, requiring provisioning on both which is expensive in terms of memory resources, and also not scalable. To overcome this, Microsoft adopted a cloud-native approach for the hybrid transactional analytical process (HTAP), an architecture that removes the wall between transaction processing and analytics. Microsoft is the first to develop the first cloud-native version of HTAP along with Cosmos DB, making it possible to have real-time analytics on enterprise events as they are occurring. The new capability eliminates the obstacles between Azure operational database services and Azure Synapse Analytics, allowing users to gain insights from their real-time transactional data stored in their operational databases with a single click, without handling data transfer or placing a strain on their operational systems. Power Platform Updates Today, 3.5 million developers use the Azure-powered Power Platform to create apps, bots, workflows, and dashboards. It’s a powerful cloud tool to build no-code\/low code applications and customise and extend Microsoft 365 and Dynamic 365. The month of March saw a 50% increase in first time Power Apps users, and this year 70 thousand organisations have started to use Power Apps on the Azure cloud platform. At Build 2020,  Microsoft announced how it is building robotic process automation connectivity to legacy apps with its acquisition of Softomotive, which will be integrated into Power Automate on the Power Platform. The company is also adding new professional developer extensibility to power virtual agents, which is being used extensively during the COVId-19 crisis. Microsoft Cloud For Healthcare A new innovation in healthcare through cloud solutions and Microsoft Teams will help hospitals and care providers better manage the needs of patients. One of the solutions as part of Microsoft Cloud for Healthcare is the bot service which connects with products such as Microsoft Teams. It can be used in emergency situations for medical service professionals addressing common questions that people may have, engaging patients and for collaboration and efficiency. Microsoft Industry Clouds targets specific Industries for its tools, and healthcare is an important one due to the ongoing pandemic. Microsoft cloud for healthcare is accessible in public preview for a six-month free trial, with an integrated set of tools like chatbots, Teams and Azure IoT Microsoft Cloud for Healthcare brings together trusted and integrated capabilities for customers and partners that enhance patient engagement, empower health team collaboration, and improve clinical and operational data insights. Microsoft Cloud for Healthcare will support capabilities such as the new Bookings app in Teams, now generally available to customers across industries to help schedule, manage and conduct business-to-consumer virtual medical appointments.","excerpt":"At Build 2020, Microsoft has announced some newly expanded capabilities for Azure Cloud. These new tools and services available empower developers to be more collaborative and productive on the cloud platform. The announcements made at the conference signals the direction in which Microsoft is proceeding, as cloud-based innovation and new tools can enable developers to […]","categories":["AI Features"],"tags":["Azure cloud platform","Cloud Computing","Cloud Computing in IT Industry","Cloud Platform","database key terms"],"author_name":"Vishal Chawla","publish_date":"2020-05-25T16:00:00","publication_year":"2020","word_count":901,"keywords":["database key terms","OpenAI","AI","chatbots","R","Cloud Computing in IT Industry","ML","kubernetes","Azure cloud platform","RAG","serverless","Cloud Computing","analytics","Azure","Cloud Platform"],"extracted_tech_keywords":["AI","ML","analytics","OpenAI","RAG","chatbots","Azure","kubernetes","serverless","R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/azure-build-2020-key-announcements-for-cloud\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":9528,"title":"Cypher is back | Analytics India Summit | 15-17th Sep 2016 | Bangalore","content":"Analytics India Magazine presents “Cypher 2016” – India’s largest and most exciting analytics summit of the year! After the enormous success of last year’s one-day summit, Cypher 2016 is back with a bigger platform – this time spaced out over a period of 3 days. You can expect 100+ speakers, 50+ talks, 20+ workshops, training sessions and lots of excitement during these three days! Since there is no doubt that analytics in India is growing at a rapid pace with companies increasingly taking up end-to-end integrated approach in analytics, this summit aims to be a platform for leadership to share unique insights as well as for companies to showcase their work. Cypher’16 is an opportunity to network with an audience that’s most definitive representation of analytics industry in India. It’s a platform where interaction between people will help them discuss or ask any critical analytic questions or let them experience the thrill of delivering answers when they join the top analytical minds for a chit chat in the networking breaks. Further, it is an opportunity to showcase your best to the best in Analytics industry by becoming our exhibit partner as Cypher 2016 attracts a technically savvy audience that understands and appreciates the value of your offerings. What’s more, while the company’s technical experts get to enhance their relationships with existing customers as well as develop new relationships, they’ll also get a chance to hone their skills through our workshops and activities. Click here to join us or convince your boss for group bookings! About Last year’s event Cypher made its debut last year and turned out to be a huge success. With most of the prestigious industry names such as Accenture, Amazon, Bajaj Alliance, Dell, Ebay, Genpact, HCL, Times group, Titan, Viacom 18 and many more renowned companies being a part of the event, it caught on many eye balls. Further, many prestigious universities such as IIM Lucknow, Narsee Monjee Institute of Management, New York University, Great Lakes Institute of Management, extended their support towards this conclave. In fact, Cypher received participants from more than 130 organizations across India. Keynote speakers included Atul Jalan – Ceo at Manthan Systems, Lakshmana Gnanapragasam – Head of Analytics at Quantium India, Moumita Sarker – Director – client delivery at Cartesian Consulting, Sameer Dhanrajani – Business Leader, Cognizant Analytics at Cognizant Technology Solutions and Vivek Ratna – Country Director, Hp Big Data Platform. The one-day summit was held at the Hotel Park Plaza in Bangalore. Intended as an annual meeting for analytics professionals, the summit discussed various trends and challenges in analytics in its panel discussions and Keynote presentations. Statistics from last year: Comparing by cities, Bangalore being the host, showed an enthusiastic participation of about 55.3%, followed by at Chennai 11.5%, Mumbai 8.5%, Delhi-NCR 6.4%, Hyderabad 5.1%, Pune 3.8%, Kolkata 1.7% and other cities at 7.7%. Approaching the participants by designation, Cypher ’15 saw more than 400 participants. Out of which, 36% were data scientists, 21% senior managers and managers, 13% directors, 10% analysts, 7% heads and founders, 7% Senior Vice Presidents & Vice Presidents and the remaining 7% were Partners, CXOs as well as CEOs of different firms. Thus, a variety of people exploited their chance to explore networking and collaboration opportunities with each other. The atmosphere was full of excited chattering and intellectual conversations. Tracks at Cypher 2016 This year’s summit revolves around three tracks – Conference, Exhibition, and Networking Dinner. Primarily the summit will have Keynote sessions, Panel Discussions, Talks\/ Huddles, Workshops and Trainings, in addition to some entertaining activities such as Networking Dinners, Gala Party, Hackathons, Exhibition, and Meetings & Showcase. Be a part of the event and discover the possibilities! Details: Date: 15-17th Sep, 2016 Venue: Hotel Park Plaza, Bangalore For more detail, visit www.analyticsindiasummit.com For booking or sponsorship queries, write to us at info@analyticsindiamag.com or call at +91 99160 06869","excerpt":"Analytics India Magazine presents “Cypher 2016” – India’s largest and most exciting analytics summit of the year! After the enormous success of last year’s one-day summit, Cypher 2016 is back with a bigger platform – this time spaced out over a period of 3 days. You can expect 100+ speakers, 50+ talks, 20+ workshops, training […]","categories":["Deep Tech"],"tags":["analytics conference"],"author_name":"Дарья","publish_date":"2016-04-11T06:29:52","publication_year":"2016","word_count":644,"keywords":["big data","API","programming_languages:R","AI","analytics conference","RAG","Aim","ViT","analytics","GAN","R"],"extracted_tech_keywords":["AI","analytics","Aim","RAG","R","API","big data","GAN","ViT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/cypher-back-analytics-india-summit-15-17th-sep-2016-bangalore\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10041308,"title":"IT Sector Saw 14% Sequential Growth In Hiring In May 2021: Naukri Report","content":"The IT-software sector in India recorded a 14 percent sequential growth in hiring in May 2021 compared to April 2021, as per the latest Naukri JobSpeak report. The growth has been attributed to mass hiring announced by several tech firms such as Adobe, IBM, Oracle, Udaan, Accenture, Meesho, Flipkart, and Nike India, the report stated. City-wise breakup showed IT hubs such as Pune (12 percent), Bengaluru (9), Mumbai (5), and Hyderabad (4) demonstrated a growth in hiring in May 2021 vs April 2021 period. However, Delhi\/NCR saw a decline of 11 percent. Smaller cities such as Jaipur (-19 percent), Vadodara (-9), and Coimbatore (-5) witnessed a downturn as well. Compared to May 2019, the tech hiring has improved by 39 percent in May 2021. “The pandemic has accelerated the pace of digital transformation across sectors and created a need for rapid and scaled deployment, thus reflected in the IT-Software industry,” said Naukri JobSpeak report. That said, the overall hiring across sectors first stabilised and then turned flat (-1 percent) in May 2021. Sectors such as Sales\/BD  (-20%), Production\/Quality (-22%) and Purchase\/Logistics (-14%) roles saw a decline in May ’21 in sequential hiring. Hiring for HR\/Admin (-4%), ITES\/BPO (-8%) and Content\/Journalism (-8%) domains has also been impacted in May 2021. The Naukri JobSpeak is a monthly Index that calculates hiring trends based on the job listings on its parent website Naukri.com, on a month on month basis.","excerpt":"Compared to May 2019, the tech hiring has improved by 39 percent in May 2021.","categories":["AI News"],"tags":["employment","Hiring","tech hiring"],"author_name":"Shraddha Goled","publish_date":"2021-06-04T20:25:50","publication_year":"2021","word_count":236,"keywords":["API","programming_languages:R","AI","employment","digital transformation","Hiring","Git","R","tech hiring"],"extracted_tech_keywords":["AI","R","Git","API","digital transformation","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/it-sector-saw-14-sequential-growth-in-hiring-in-may-2021-naukri-report\/","complexity_score":3,"technical_depth":6,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10009802,"title":"IBM Launches Artificial Intelligence Centre In Brazil","content":"Introduced in 2019, by IBM, Brazil has launched the largest research facility, that focuses on artificial intelligence, through a collaboration between the private and public sector. The Artificial Intelligence Center (C4AI) is supported by investments made by IBM along with the São Paulo Research Foundation (FAPESP) and the University of São Paulo (USP). This AI centre — C4AI has been established to tackle five significant challenges that are related to health, the environment, the food production chain, the future of work and the development of NLP technologies in Portuguese. Along with this, it will also aid in projects relating to human wellbeing improvement as well as initiatives focused on diversity and inclusion. The total investment in the AI centre will reach $20 million over the next ten years, which will be split among the investors. The USP will contribute $1 million to cover costs related to the physical set-up of the space, as well as over 70 lecturers and staff to run the centre. At the same time, IBM and São Paulo Research Foundation (FAPESP) will co-run the centre, with a contribution of $500,000 each. While launching the centre, the USP research dean Sylvio Canuto stated that the AI centre is expected to attract more attention to Brazil concerning AI research and the immense talent that is available in the country to support the AI-enabled future. Agreeing to that, the scientific director of FAPESP, Luiz Eugênio Mello, stated that artificial intelligence offers “an infinite number of possibilities” in the areas of medical research during the COVID pandemic. Thus, the AI tools will be bringing in more effectiveness and time savings. According to him, this partnership with IBM is like a milestone in a strategic area for the future. Along with the primary location in the USP São Paulo campus, the AI centre will also have a second facility in the Institute of Mathematical and Computer Sciences (ICMC). It will also have three monitoring committees to discuss topics of common interest to Brazil around AI. This will further expand the debate around the five initial challenges, with a focus on industry, science and societal applications. This news compliments Brazil’s last year’s announcement of creating a national AI strategy to respond to the worldwide race for leadership of AI and the requirement to discuss the future of work, education, tax, research and development as well as ethics as the application of related technologies becomes more pervasive. Alongside, Brazil adheres to the Organisation for Economic Co-operation and Development (OECD) ’s human-centred AI Principles, which provide recommendations around areas such as transparency and explainability. The Brazilian AI strategy will be focusing on six vertical themes — qualifications for a digital future; workforce; research, development, innovation and entrepreneurship; government application of AI; application in the productive sectors and public safety.","excerpt":"Introduced in 2019, by IBM, Brazil has launched the largest research facility, that focuses on artificial intelligence, through a collaboration between the private and public sector. The Artificial Intelligence Center (C4AI) is supported by investments made by IBM along with the São Paulo Research Foundation (FAPESP) and the University of São Paulo (USP). This AI […]","categories":["AI News"],"tags":["what is artificial intelligence"],"author_name":"Sejuti Das","publish_date":"2020-10-15T17:03:41","publication_year":"2020","word_count":465,"keywords":["Go","artificial intelligence","programming_languages:R","AI","innovation","Git","NLP","what is artificial intelligence","GAN","AI research","R"],"extracted_tech_keywords":["AI","artificial intelligence","NLP","R","Go","Git","GAN","innovation","AI research","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/ibm-launches-artificial-intelligence-centre-in-brazil\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":32924,"title":"5 Data Analyst Job Openings At JPMorgan Chase You Should Not Miss","content":"Image for representative purpose only JPMorgan Chase and Co. is one of the most lucrative options for those seeking to get into a role as a data analyst in today’s job market. The firm manages over $2.5 trillion in assets and operations across the world, with a presence in over 100 markets and more than 240,000 employees worldwide. In this article, we list down the top job openings available in Bangalore, Mumbai, and Hyderabad for positions regarding data analysis roles. L3 Support Analyst JPMorgan requires a candidate with a strong technical background in the current J2EE and Oracle database software in order to fill the L3 Support Analyst role. Moreover, the candidate is also required to have a track record of delivering results in complex and fast-moving environments. It is considered a plus if the candidate has knowledge of investment banking. The job will involve the analyst developing enhancements and data remediation scripts to be included in system updates, along with working as a part of a global support team. Requirements: Development experience using J2EE and the JUnit framework Working knowledge of Oracle database and associated tuning techniques Team working skills are essential Location: Bangalore Apply here. Business Systems Analyst The business systems analyst position comes with the primary responsibility of performing business analysis and database marketing functions. These will further be utilised for the execution of portfolio marketing campaigns. The candidate is required to have the ability to assess information that is extracted from multiple data sources, and is also tasked with resolving inconsistencies. Requirements: 5+ years extensive experience in developing Base SAS codes for manipulating large amounts of data Creating reports using procedures such as Proc Freq, Proc Format, Proc Sort, Proc SQL, and Proc Summary Experience as Business Analyst in related domains Experience in database querying (Oracle\/Teradata) Proficiency with Microsoft Office is required. Advanced Microsoft Excel knowledge is a plus. Location: Mumbai Apply here. Business Analyst The candidate applying for this job will be assigned to the CCB Analytics group, who will utilise Big Data tools to find outliers in the system. This includes understanding the behaviour and preferences of customers. This will be utilised to identify important trends, trouble-spots, and areas of opportunity. The candidate is required to create and maintain analytical insights using speech phonetics. Requirements: Bachelor’s degree in Statistics, Math, Economics, Business or equivalent experience. Must have experience using tools like SAS, Python, SQL Experience querying data through Oracle, Teradata, or SQL Servers Strong communication and presentation skills – written and oral Strong knowledge of all Office applications such as MS Excel, MS Powerpoint, MS Access, and Visio Able to work in a fast paced environment Location: Mumbai Apply here. Software Engineering – Business Analyst JPMorgan is also looking for a candidate to fill a position in the Global Security Technology team within their Corporate Technology Division. This role will require the candidate to drive the development of products in the GST space by communicating between the business and technology teams. The candidate is also required to participate in Agile ceremonies and be a part of the scrum team, including participating in daily scrums. Requirements: 6+ years of Business Analysis experience. Bachelors degree in Management Information Systems or related field of study or equivalent -relevant experience in software development In-depth understanding of technology Ability to do in-depth research and drum up granular insights Experience performing detailed data analysis, including the ability to create, run and modify SQL queries to pull data from source databases for data analysis and mapping activities Demonstrated problem-solving skills and ability to develop solutions to complex analytical\/data-driven problems Bachelor’s Degree in Management Information Systems or related field of study or equivalent -relevant experience in software development Location: Hyderabad Apply here. Associate – Model Risk Business Analyst In the Model Risk Business Analyst Associate position, the candidate will be required to have the ability to provide business operating models and workflows to current and future business states. Additionally, they will be in charge of supporting the migration of manual processes related to Model Interconnectedness and Authoritative Data Sources. Requirements: Minimum Master’s [MBA preferred] degree in Finance\/Business\/Economics or in Computer Information Systems\/Financial Engineering 2-5 years experience as a business analyst in the financial industry Strong written and verbal communication skills Proficiency in MS Excel required [VBA experience is a must] Proficiency in MS Access, SQL, PowerPoint, SharePoint, MS Project & Visio required Comfortable working with large datasets for deep dive analysis Experience with BI tools such as Tableau and Qlikview preferred Experience with graph database tools such as Neo4j and Gephi as well as familiarity with Cypher language is a plus but not required Location: Mumbai Apply here.","excerpt":"JPMorgan Chase and Co. is one of the most lucrative options for those seeking to get into a role as a data analyst in today’s job market. The firm manages over $2.5 trillion in assets and operations across the world, with a presence in over 100 markets and more than 240,000 employees worldwide. In this […]","categories":["AI Hirings"],"tags":["Business Analytics"],"author_name":"Anirudh VK","publish_date":"2019-01-07T05:58:55","publication_year":"2019","word_count":776,"keywords":["big data","AI","data-driven","Python","programming_languages:Python","ViT","analytics","Business Analytics","SQL","GAN","R"],"extracted_tech_keywords":["AI","analytics","Python","R","SQL","big data","GAN","ViT","data-driven","programming_languages:Python"],"url":"https:\/\/analyticsindiamag.com\/ai-hiring\/5-data-analyst-job-openings-at-jpmorgan-chase-you-should-not-miss\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10098948,"title":"There&#8217;s No Escape for Criminals in India","content":"Be it the high-profile celebrity assassination cases, political massacres, or the controversial Vikas Dubey case, one company at the centre of it all is Staqu Technologies and its AI-powered video analytics model Jarvis. “We have identified more than 30,000 criminals in the country; we help identify around 100 criminals on a daily basis,“ said Atul Rai, CEO and co-founder of Staqu, in an exclusive interaction with AIM.  Their solution has also garnered attention across sectors, from retail and manufacturing to infrastructure, and hospitality. Jarvis, developed by Staqu, finds application in law enforcement across various states. It is referred to as PAIS in Punjab, Trinetra in UP, Chakra in Bihar, and Eagle in Haryana. Staqu collaborates with a total of eight states, including Rajasthan, Uttarakhand, and Telangana. Moreover, Staqu’s Jarvis incorporates a Video Wall technology and aggregates CCTV footage from 70 prisons. It handles 700 cameras, aiding law enforcement with tasks like access monitoring, violence detection, and intrusion alerts in prisons across UP. This marks the first use of such AI-driven analytics in Indian prisons. Historically, criminals and illicit networks have operated from within the prison walls. With its solid support, JARVIS significantly empowers police forces in their efforts to combat crime. Additionally, it is also efficient in tasks like frisking, crowd analysis, and facial recognition. Beyond these applications, Staqu is actively developing JARVIS for diverse security scenarios as well. The company’s image-based search and analytics platform serves business and security needs for other establishments. It uses deep learning and machine learning technologies to offer video analytics, big data tools, and auditing solutions for security and data analytics domains. In the realm of big data analytics, Staqu provides tools such as a data aggregating tool and a data analysis tool. More than meets the eye This is just one side of Staqu. It operates majorly in the B2B and SaaS sectors, specifically targeting the high-tech and fintech market segments. Founded in 2015 by Atul Rai, Anurag Saini and Pankaj Kumar Sharma, the company works across several sectors from retail, manufacturing, infrastructure, and hospitality to the public sector and has conglomerates such as IBM, GMR and Adani Power as clients. On the retail front, it caters to Starbucks, Piramal Glass, Marico, Tata Consumer Products Ltd, Cafe Coffee Day, Chaayos, Crocs, Khadims, Croma, and Metro. Its business models are centred around offering a loss prevention module and a revenue enhancement module. “If you want to secure it, you will go for a loss prevention module. If you want to monetise it, you will go for the revenue enhancement module,” Rai explained. The revenue enhancement module primarily targets the retail and real-estate sectors. It aims to bridge the data gap in the retail industry, offering insights similar to e-commerce data. Leveraging existing camera infrastructure, JARVIS provides valuable insights to retailers. It tracks footfall, customer journey, gender demographics, product engagement, and queue analysis. By utilising a unique tracking algorithm, JARVIS accurately measures footfall, excluding employees. This technology addresses the lack of critical data in the retail sector, offering valuable metrics for business optimisation. Now, it also finds its use case in catching criminals. The working mechanism involves connecting observers from cameras to JARVIS through the internet, eliminating the need for extra hardware and associated costs. Data analysis occurs in the cloud via custom JARVIS, ensuring end-to-end encryption. This approach ensures data ownership and GDPR compliance. The camera’s features provide insights which, when combined with sales data, generate analyses like customer qualification percentages. The outcomes are tailored for clients like retail stores, enabling informed decision-making based on comprehensive insights. Video management ensures centralised control for businesses with multiple locations. JARVIS, facilitates plug-and-play insights from various stores, covering occupancy analysis and beyond. Additionally, it offers security features like intrusion detection and identifying blacklisted individuals. JARVIS operates as a unified solution, eliminating the need for additional hardware—just an internet connection is required for implementation. Questions about enterprise adoption and the possibility of integrating additional features were addressed by Rai when he explained the approach taken, he stated, “Imagine an enterprise adopting our system and wanting to expand its capabilities. We’ve structured it as a platform rather than a prescriptive solution, akin to Google’s modular approach with services like Gmail.” “In our industry, lacking AI expertise is common, even when APIs are available. Major companies like Google and Microsoft provide APIs, but accessing them requires technical proficiency. By producing our platform, we provide solutions to real business challenges faced by industry giants like Starbucks, Raymond, and CCD. Our aim is to bridge the gap between AI potential and practical implementation.” Funding-wise, Staqu has secured $2.11 million in total funding through four seed rounds and achieved a valuation of $8.57 million as of April 27, 2022. In the last undisclosed seed round on May 25, 2022, Security and Intelligence Services led the investment. Staqu has 64 investors, including five institutional investors like Security and Intelligence Services and 59 Angel investors, including Bikky Khosla. Previous Seed rounds saw participation from Mount Judi Ventures, Ayushman Khanna, Ajay Gupta, and Neeraj Kumar Singal. Innovations Powering Staqu Staqu’s proprietary technology consists of two main components: the streamer algorithm and an IP portfolio with 25 research papers and four patents. The streamer algorithm handles diverse camera sources by unifying recommendations and analyses, compatible with existing systems like DVR NVR setups. The IP portfolio focuses on advanced identification tech and big data search methods. It precisely pinpoints individuals and handles large-scale processing, aided by a specialised search patent using a pyramidal algorithmic approach, alongside leveraging machine learning techniques, including transformers, CNNs, and LSTMs. The foundational model, Vision Transformer (VIT), is optimised for real-time learning and adapted to handle a large number of diverse camera streams. The technology excels in bimodal data analysis, combining text and visual info for comprehensive understanding. It supports audio-visual integration for enhanced insights and interactions. “When you talk about the textual model forensic analysis then, of course, you are training on bimodal data so you are training with the text and images. For instance, if I’m searching for a green shirt and white t-shirt that means that need to be trained with the textual data and the image as well.” In essence, the proprietary technology incorporates a sophisticated streamer algorithm and IP portfolio backed by advanced models. Customising the vision transformer ensures accurate, diverse, and real-time data processing, enabling efficient camera data management and actionable insights for better decision-making.","excerpt":"Staqu’s Jarvis helps law enforcement agencies nab about 100 criminals on a daily basis and has put 30,000 of them behind bars so far","categories":["AI Highlights"],"tags":["Data Analytics","Facial Recognition","Jarvis"],"author_name":"Shyam Nandan Upadhyay","publish_date":"2023-08-24T10:52:27","publication_year":"2023","word_count":1067,"keywords":["Go","machine learning","Facial Recognition","AI","Transformers","Jarvis","RAG","Ray","Aim","deep learning","analytics","Data Analytics","R"],"extracted_tech_keywords":["AI","machine learning","deep learning","analytics","Aim","Ray","Transformers","RAG","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/theres-no-escape-for-criminals-in-india\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":33501,"title":"To Code or ETL: ETL Programming Vs Hand-Coded Programming","content":"There has been a common debate surrounding automated ETL programming and hand-coded programming in an organisation. ETL stands for Extract Transform Load which is a procedure in the process of data preparation. The job is to integrate and make the data available for the further analysis. According to consultant Rick Sherman, “In many companies, data integration remains a manual task — by choice. But it might be time to automate that process.” But it is really worth getting ETL tools for your business? Let’s take a look. ETL Vs Manual Programming Here are some ways in which both the methods can be compared: 1.Time: Hand-coded methods can take days or even weeks, depending on the amount and complexity of the data, to only prepare the data for the analysis, even before any analysis is done. For example, if the sales data has to be compiled into a single sheet considering various parameters and then given further for analysis, hand-coding would take many days to prepare the data. If instead ETL tools are used, it will do the job accurately and within a time frame much less than that need by the hand-coded method. ETL comes in rescue big time as a time-saver. 2.Reusability: Processes in ETL method can be saved and directly reused for other processes and data models as well. In manual coding, changes will have to be made meticulously, by a programmer. Instead, his time can be saved and be spent on genuine problems. 3.Management: Because of automation, managing datasets has become easy with ETL programming. ETL tools provide one a larger view of ETL processes, with things like where the data is coming from, where it is going and what sort of calculations have been done on it. By doing this, ETL method helps in a way that a manually crafted code will never be able to replicate. Such advantages makes ETL processes preferable as it makes data management more easy. Applying hand-coded programming on these complex structured and unstructured data, saves several lines of code. Meanwhile, automated coding does it with minimal efforts. A visual ETL layer also helps in making the process very transparent. Also, every change is easily noticeable and easy to retract and hence saves the times that one spends in hand coded programming and finding the error in the code to locate changes that are initiated by one programmer of the team who is currently unavailable. Moreover, all data, no matter how large it is, can be taken and used for analysis, as opposed to hand-coded programming, which requires the programmer or developer to take a sample data to test their program. This also adds to saving time. Why Not All Companies Have Adopted ETL? ETL, specially for large datasets, works the best. But many organisations with large datasets have not popularly adopted the ETL methods. Every ETL tool has its own unique logic and some applications might find restrictive. But hand coded programming comes with endless possibilities and it is quite flexible that way. In manual programming, data can be managed directly and organisations find that convenient. . We have a complete control over the program, but in ETL there are parameter settings of the ETL tools that you cannot just have a control. Also, manual coding is quick and easy to move data and is hence cheaper than ETL. It’s easier to get the data into a report. Many organisations find ETL tools expensive and so they avoid using it. Conclusion In case of large datasets, where extensive data preparation is required, ETL comes with a huge advantage. It provides a cleaner, better experience than manual coding. ETL comes in very handy when the processes are very easy to automate and the task is not that complicated, but not otherwise. It makes sense to always use ETL, since it has many more advantages over manual. Unless there is a strict cost constraint or a problem where strictly a manual intervention in a problem is necessary, manual coding should be preferred. Otherwise, ETL is a always a preferred choice.","excerpt":"There has been a common debate surrounding automated ETL programming and hand-coded programming in an organisation. ETL stands for Extract Transform Load which is a procedure in the process of data preparation. The job is to integrate and make the data available for the further analysis. According to consultant Rick Sherman, “In many companies, data […]","categories":["Deep Tech"],"tags":["analysis","Automation","models","VS Code"],"author_name":"Disha Misal","publish_date":"2019-01-15T09:29:43","publication_year":"2019","word_count":678,"keywords":["Replicate","Go","analysis","programming_languages:R","AI","ETL","models","Automation","programming_languages:Go","automation","VS Code","GAN","R"],"extracted_tech_keywords":["AI","R","Go","ETL","GAN","automation","Replicate","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/to-code-or-etl-etl-programming-vs-hand-coded-programming\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10014461,"title":"TabPy &#8211; Guide To Integrating Tableau With Python","content":"Python has many visualization libraries(matplotlib, seaborn, bokeh, plotly, etc.) to show the EDA of a dataset, but this would need many lines of code to process. Tableau is known for its amazing interactive dashboards. Imagine how powerful it would be to have leveraged the power of both Python and Tableau together to build amazing business insights. Source: Link TabPy(Tableau Python Server) is an API which allows python scripts to be run on a Tableau server. Thereby enabling EDA and visualization more effectively. This helps in building better dashboards and advanced analytics to draw insights out of it. TabPy handles the data exploration and visualization part when data developers can focus more on the data science logic to better handle business use cases. Data science use cases are iterative in nature. Understanding, analyzing and visualizing takes a major part. Before modelling the data, proper data cleaning and preparation is needed. So let’s now look at how to set up TabPy. Workflow: Installation The GitHub repository can be cloned and used otherwise via pip (It is recommended to run in a virtual environment). TabPy is compatible with Python 3.6 and above. pip install tabpy==2.3.1 If you want you can also separately install the tabpy-server and tabpy-client. After installation from CLI run the command “tabpy” to connect to the server. Note the port number is 9004 which will later be used in server connection from Tableau. Open Tableau Software Go to Help -> Settings and Performance -> Manage External Service Connection. Specify the port number as 9004 and Sign in. Additionally, users can check the box for Require SSL to encrypt the data sent over the network. Lastly, click on Test Connection and the subsequent successful connection will be shown. Now tableau is linked to TabPy. It’s time to do some operations with our data with Python script. Load the data onto a new worksheet in Tableau and create a calculated field. Passing Expressions to Python To let tableau know that the expressions need to go to Python, it must be passed through any one of the following four functions – SCRIPT_BOOL , SCRIPT_REAL , SCRIPT_INT , SCRIPT_STR.Python Functions take the form of Table calculations in Tableau.In table calculations, all the Fields being passed to Python must be represented as Sum (PROFIT), Max (Profit), MIN(Profit), ATTR( Category) etc. Here is an example of sentiment analysis. The python script is written in the calculated field. The right side of the page shows the sentiment scores when the calculated field is passed on the view. Another example shows Profit greater than zero. SCRIPT_BOOL returns a boolean value from the specified expression. The expression is passed to the running external service instance. In Python expressions, use _argn to reference parameters ( _arg1, _arg2, etc.). _arg1 is equal to SUM([Profit])All the Fields being passed to python must be aggregated like Sum(PROFIT), MIN(Profit), Max (Profit), ATTR( Category) etc. Wrapping Up In this article, we have shown an overview of how to set up and use TabPy. As the next step to this, connect your data and use it for complete visualization and advanced analytics, storytelling, interactive graphs etc. with TabPy. Tableau can be connected to multiple data sources to create real-time and can create dashboards that are dynamic in nature.","excerpt":"TabPy(Tableau Python Server) is an API which allows python scripts to be run on a Tableau server. Thereby enabling EDA and visualization more effectively. This helps in building better dashboards and advanced analytics to draw insights out of it. TabPy handles the data exploration and visualization part when data developers can focus more on the data science logic to better handle business use cases.","categories":["Deep Tech"],"tags":["Automating EDA","Data Analysis","Data Visualisation","data visualization python","Python","Tableau"],"author_name":"Jayita Bhattacharyya","publish_date":"2020-12-16T16:00:00","publication_year":"2020","word_count":542,"keywords":["data visualization python","Automating EDA","Data Analysis","data science","Tableau","Plotly","Go","sentiment analysis","Data Visualisation","RAG","Python","Seaborn","analytics","Matplotlib","R"],"extracted_tech_keywords":["data science","analytics","Matplotlib","Plotly","Seaborn","RAG","sentiment analysis","Python","R","Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/tabpy\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10166981,"title":"AI Crawlers: The Nasty Bugs Causing Trouble on the Internet","content":"AI tools with web search capabilities, such as Anthropic’s Claude, browse the internet to deliver users the needed information. Perplexity, OpenAI, and Google offer similar features through ‘Deep Research’. In a blog post, Cloudflare explained that these web crawlers, often referred to as AI crawlers, deploy the same techniques as search engine crawlers to gather available information. While the aim of AI crawlers is to assist users, they may be causing more damage on the internet than one realises. They are believed to increase server resource usage for website administrators, leading to unwanted bills and causing disruptions. AI Crawlers on The Rise of Being a Hassle Gergely Orosz, creator of The Pragmatic Engineer newsletter, shared on LinkedIn, “AI crawlers are wrecking the open internet, and I’m now being hit for the bill for their training.” He explained that his website, a side project, initially had a few thousand visitors a month and used around 100 GB of server bandwidth. But, after Meta’s AI crawler and other bots like Imagesiftbot started crawling the website, more than 700 GB of bandwidth was consumed, leading to an extra $90 in bills. Orosz expressed frustration over having to pay all this extra money to help train LLMs. Furthermore, he added that crawlers ignore robots.txt file. “The irony is how the bots—including Meta! — blatantly ignore the robots.txt on the site that tells them ‘please stay away’…I’m upset – and have had enough.” Vercel, a cloud platform company, shared some interesting statistics from their network in a blog post that said: “AI crawlers have become a significant presence on the web. OpenAI’s GPTBot generated 569 million requests across Vercel’s network in the past month, while Anthropic’s Claude followed with 370 million.” Source: Vercel “For perspective, this combined volume represents about 20% of Googlebot’s 4.5 billion requests during the same period,” it added. Xe Iaso, a software developer, expressed frustration upon noticing that AmazonBot was consuming their Git server resources. Attempts to block it resulted in failure. Iaso stated in the blog post, “It’s futile to block AI crawler bots because they lie, change their user agent, use residential IP addresses as proxies, and more. I just want the requests to stop.” The developer created an open source solution, Anubis, to present a challenge to AI crawlers and block the requests. The developer’s quick solution turned out to be helpful to others as well. Bart Piotrowski, a system administrator at GNOME, used it to fend off AI crawlers from GNOME’s GitLab instance, which were reportedly taking 90% of their resources. Drew Devault, founder of SourceHut, wrote a blog post voicing something similar: “Over the past few months, instead of working on our priorities at SourceHut, I have spent anywhere from 20-100% of my time in any given week mitigating hyper-aggressive LLM crawlers at scale.” Ars Technica reached a similar conclusion for AI crawlers, focusing on its impact on open source projects. Many other reports indicate that people are attempting to fend off AI crawlers consuming their web resources. What Can Be Done? Solutions such as Iaso’s Anubis, though not suitable for everyone, are a good option and are increasingly being embraced by individuals. Cloudflare has joined the fight against AI bots that do not honour the robots.txt rule with AI Labyrinth, which uses AI-generated content to keep the crawler occupied and waste its resources. Source: Cloudflare “Crawlers generate more than 50 billion requests to the Cloudflare network every day, or just under 1% of all web requests we see. While Cloudflare has several tools for identifying and blocking unauthorised AI crawling, we have found that blocking malicious bots can alert the attacker that you are on to them, leading to a shift in approach, and a never-ending arms race,” the Cloudflare blog read. It added, “So, we wanted to create a new way to thwart these unwanted bots, without letting them know they’ve been thwarted.” In addition to the solutions mentioned above, AI companies can do their bit by improving their crawlers to respect the web resources and be a little less aggressive in their information-hunt process. While the web search functionality in AI tools provides great value, it should not come at the cost of disrupting the web server resources of small or independent web admins.","excerpt":"AI crawlers are eating up web resources, and site administrators are looking for defence mechanisms to protect against big bills.","categories":["AI Features"],"tags":["ai bots","cloud servers"],"author_name":"Ankush Das","publish_date":"2025-04-01T11:08:14","publication_year":"2025","word_count":711,"keywords":["Anthropic","Go","OpenAI","AI","cloud servers","Git","RAG","Aim","GitLab","Rust","ai bots","R"],"extracted_tech_keywords":["AI","OpenAI","Anthropic","Aim","RAG","R","Go","Rust","Git","GitLab"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/ai-crawlers-the-nasty-bugs-causing-trouble-on-the-internet\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10051774,"title":"Join This Masterclass &#8211; ‘How To Start Your Career In Data Science?’","content":"“Data Scientist” has become one of the most coveted job titles in the IT industry. This is owing to the fact that organisations across sectors have immense interest and determination to make sense of big data and improve their business outcomes. The need to embrace digital transformation through data is at its all-time high. Undoubtedly, this data-led industrial revolution has proved the value proposition that a data scientist brings to an organisation, cutting across varied industry sectors and strategic business functions. And as it turns out, the industry has also opened up after a brief slump last year. Hiring across data science roles has picked up tremendous momentum over the past few months. It is no surprise that many fresh graduates, including working professionals, are now rearing to enter the realms of the data science domain and make a fulfilling career for themselves. But it is not as easy as it looks. For starters, it takes a certain level of role understanding, aptitude & attitude to be a successful data scientist. A simple search on the internet would perhaps lead you to several online resources, articles, and free advice on how to become a data scientist, top data science skills, top data science tools to learn etc. Unfortunately, most of this information is just too generic to be of any practical use in the current scenario. If you are seriously considering moving into data science, then you will need to turn to the experts for their guidance. Experts who have seen data science right from its early days. Someone who has pioneered data science education and has also mentored many data science career aspirants towards success. Analytics India Magazine in association with Praxis Business School is organising this masterclass on ‘How to start your career in data science,’ where Professor Charanpreet Singh will guide you on various aspects, qualities, skills, aptitude etc that an aspiring data scientist should possess. Along with Prof. Singh, Mr Kaushik Sarkar will also share more details about how Praxis Business School has successfully placed over 1,000 students from its highly acclaimed PGP in Data Science Program. Date: 28 October 2021 Time: 7 – 8 PM (IST) REGISTER NOW! Who should attend the webinar: Aspiring data scientists, data analysts etc who want to begin a career in data scienceGraduates looking to pursue full time PG program in Data ScienceWorking professionals interested to move into data science domain About the Speakers: Prof Charanpreet Singh is the founder and director at Praxis Business School Foundation. He has been driving the highly successful Data Science Program offered by Praxis Business School, which he started in 2011. Prof Singh is a pioneer in the field and has 30+ years of experience in British Oxygen, Tata Steel, PwC, HP and Praxis. He has professional and academic interests in the areas of education, learning, data science and communication and has been a mentor to career aspirants across domains. He is a Chevening Scholar of the British Government and has done his B Tech in Mechanical Engineering from IIT Kanpur and his MBA from the University of Iowa in the USA. Mr Kaushik Sarkar is the Head of Corporate Relations at Praxis Business School. He has over 15+ years of experience working in the education space. At Praxis Business School, Kaushik is responsible for engaging with organizations in the hiring space for PGDM, Data Science, Cyber Security and Data Engineering. Before joining Praxis in 2015, Kaushik was the Group Head for International Marketing & Strategy at Alliance University in Bangalore where he oversaw Alliance University’s international strategy. Kaushik holds a B.Com from Gauhati University, BA in Economics from IGNOU, MBA from IMI-Belgium and a PG in Global Marketing from Judge Business School – Cambridge University. Date: 28 October 2021 Time: 7 – 8 PM (IST) REGISTER FOR THIS MASTERCLASS, HERE","excerpt":"Join this exclusive masterclass to know how to start a lucrative career in data science. Also get insights into the stellar placement record by Praxis Business School for their PGP in Data Science Program","categories":["Deep Tech"],"tags":["Data Science Career","Data science career webinar","praxis","Praxis Business School","praxis data science program","praxis data science programme"],"author_name":"AIM Media House","publish_date":"2021-10-25T12:28:00","publication_year":"2021","word_count":637,"keywords":["big data","data science","Go","Data science career webinar","AI","praxis data science program","Git","Data Science Career","praxis data science programme","Aim","data engineering","analytics","praxis","GAN","R","Praxis Business School"],"extracted_tech_keywords":["AI","data science","analytics","Aim","R","Go","Git","big data","data engineering","GAN"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/join-this-masterclass-how-to-start-your-career-in-data-science\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":48440,"title":"Top 7 Tech Webinars You Must Attend Before 2019 Ends","content":"One of the key ways to keep updated with the latest tools and technologies in emerging tech is to keep learning. One must be a life-long learner if s\/he wishes to pursue a career in the vibrant fields of data science, artificial intelligence, machine learning, IoT and other emerging technologies. For a few years now, noted tech companies have started to conduct free webinars which not only update the community and users at large, but also serve as a great marketing tool for these behemoths. In this article, we list down 7 online events organised by tech giants which one should not miss. 1| How To Begin A Career In Data Science Date: 24th October 2019 Timing in India:  7:00 PM – 8:00 PM Conducted by Analytics India Magazine, this webinar will help you understand how one has to make their move to data science. After completion of the session, you will learn about both the hard and soft skills needed to be successful in data science along with the steps of embarking on the journey of data science. Register here. 2| How Kubernetes on Google Cloud is fueling Fintech hypergrowth Date: 14th November 2019 Timing in India: 10:30 PM – 11:15 PM In this live event, the speaker will discuss how the fintech company, Recurly leverages Google Cloud to securely deploy and hyper-scale there PCI compliant platform. It will also discuss how the company leverages Google Kubernetes Engine (GKE) in its platform along with the process of modernising the PCI compliant applications. Register here. 3| An Introduction to Infrastructure as a Service (IaaS) on Azure Date: 22nd October 2019 Timing: 10:00 AM – 11:00 AM Pacific Time In this webinar, you will learn how using cloud infrastructure helps you simplify management, reduce costs, rapidly adjust to changing business demands, and enhance security. You will also learn how to scale up and down easily with cloud compute and cloud storage, protect your data and applications no matter where they reside, strengthen the security of your workloads across every environment, whether on-premises, in the cloud, or on the edge, and much more. Register here. 4| Optimise Workload Performance with Azure SQL Database Date: 23rd October 2019 Timing: 10:00 AM – 11:00 AM Pacific Time This webinar by Microsoft Azure will discuss the fully managed Azure SQL Database which provides operational and financial benefits by optimising performance. You will learn the similarities between the on-premises SQL Server and Azure SQL Database, built-in intelligence capabilities that optimise workload performance, possibilities of having the limitless scale for workloads as well as how Azure SQL Database can provide price-performance benefits. Register here. 5| How to Choose and Use Linux Tools on Azure Date: 30th October 2019 Timing: 01:30 AM (GMT+5:30) This webinar is also conducted by the tech giant Microsoft where you will learn how to choose and use popular Linux-based tools and services on Azure, how to choose the right tool in order to build images and the other ways of operating virtual machines. Register here. 6| Trusted Data Made Available Through Secure Self-Service Date: 29th October 2019 Timing in India: 1:30 PM This webinar will help you to learn how to make quality, governed data accessible for anyone who needs it in order to make key business decisions. You will understand how to bring higher quality analytics, personalised artificial intelligence, integrated compliance, and increased data sources combined together into an enterprise data catalogue. Register here. 7| Integrating Data, Automating Processes to Create Value And Opportunity Date: 14th November 2019 Timing in India: 7:30 PM In this webinar, the experts will discuss how the integrated use of robotic process automation, artificial intelligence and automation tools transforms businesses and supports digital transformation. You will also learn how an integrated approach delivers the greatest return of productivity, efficiency, and investment, how intelligent automation supports the digital expectations of today’s customers along with the role of intelligent automation and artificial intelligence for digital transformation. Register here.","excerpt":"One of the key ways to keep updated with the latest tools and technologies in emerging tech is to keep learning. One must be a life-long learner if s\/he wishes to pursue a career in the vibrant fields of data science, artificial intelligence, machine learning, IoT and other emerging technologies. For a few years now, […]","categories":["Deep Tech"],"tags":["Tech Webinars You Must Attend"],"author_name":"Ambika Choudhury","publish_date":"2019-10-20T13:00:49","publication_year":"2019","word_count":659,"keywords":["data science","machine learning","artificial intelligence","AI","Tech Webinars You Must Attend","R","RAG","analytics","SQL","Azure","kubernetes"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","data science","analytics","RAG","Azure","kubernetes","R","SQL"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/top-7-tech-webinars-you-must-attend-before-2019-ends\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":69794,"title":"Schneider Electric India Launches IoT Platform &#8216;EcoStruxure&#8217; To Drive Growth","content":"Diving into the space of emerging technologies such as cloud and analytics, Schneider Electric recently launched its IoT enabled platform called ‘EcoStruxure’. It was launched at the ‘EcoXpert Meet’ in Pattaya, Thailand for its channel partners. The IoT platform, which was launched in India earlier this year offers a one stop IT\/ OT solutions for achieving cost and energy efficiency for companies essentially in these for areas—data center, power grid, buildings and industry. The solution will now be available to company’s partners for ensuring energy efficiency at all levels. EcoXperts, an extension of Schneider Electric, is a pioneer in the field of power management, energy management and energy control. They work in the direction of delivering intelligent and sustainable solutions adding value and longevity to the success of shared customers. Shrinivas Chebbi, President-Buildings Business, Schneider Electric India said “The EcoStruxure architecture offered by Schneider Electric is one of the key elements of our future sales strategy in India. The solution, which combines IT and OT, will provide enhanced value to our customers in terms of safety, reliability, efficiency, sustainability and connectivity”. He also added that since the start of EcoXpert program, they have tried to push digitization through their channel partners. With the addition of EcoStruxure as a sales offering in India, the company is looking forward to drive success to their shared customers.","excerpt":"Diving into the space of emerging technologies such as cloud and analytics, Schneider Electric recently launched its IoT enabled platform called ‘EcoStruxure’. It was launched at the ‘EcoXpert Meet’ in Pattaya, Thailand for its channel partners. The IoT platform, which was launched in India earlier this year offers a one stop IT\/ OT solutions for […]","categories":["AI News"],"tags":["iot platform india"],"author_name":"Srishti Deoras","publish_date":"2017-09-11T05:17:53","publication_year":"2017","word_count":224,"keywords":["iot platform india","programming_languages:R","AI","Git","ViT","analytics","R"],"extracted_tech_keywords":["AI","analytics","R","Git","ViT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/schneider-electric-india-launches-iot-platform-ecostruxure-drive-growth\/","complexity_score":3,"technical_depth":6,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":37513,"title":"MachineHack Data Science Hackathon: Use ML To Predict Costs Of A Meal At A Restaurant","content":"Who doesn’t love food? All of us have a craving for at least a few favourite food items, and many restaurants are on our “perfect spot” list. But if there’s one factor that makes us reconsider eating out is the cost. In MachineHack’s latest hackathon called Predicting Restaurant Food Cost Hackathon, you will be predicting the cost of the food served by the restaurants across different cities in India. You will use your Data Science skills to investigate the factors that really affect the cost, and gain interesting insights that might help you choose what to eat and from where. About the Data Set This hackathon of MachineHack has a problem of predicting the price of a restaurant using a training set of 12,690 records and a test set of 4,231 records. The data consists of the following features: TITLE: The feature of the restaurant which can help identify what for whom it is suitable for. RESTAURANT_ID: A unique ID given for each restaurant. CUISINES: The variety of cuisines that a restaurant offers. TIME: The open hours of the restaurant. CITY: The city in which the restaurant is located. LOCALITY: The locality of the restaurant. RATING: The average rating of the restaurant by customers. VOTES: The overall votes received by the restaurant. COST: The average cost of a two-person meal. Steps To Start The Hackathon 1.Open your favourite browser and go to https:\/\/www.machinehack.com\/course\/predicting-restaurant-food-cost-hackathon\/ 2.If you are a new user click on sign up and register on MachineHack with your email. Click the login button if you have already registered to MachineHack. 3.After logging in, click on “Hackathons” which will take you to the page with a list of all the MachineHack hackathons. Select the latest hackathon from the list titled “Predicting Restaurant Food Cost Hackathon”. Read through each of the highlighted sections of the Hackathon to understand the problem clearly. 4.Click on “Start Hackathon” to enter the challenge, read through the rules and click on “Start Course” which will take you to the submission page. 5.In the submission page, the training set, test set, and sample submission can be downloaded from the “Attachment” section and your submissions can be made by following the submit link in the “Assignment” section. Please note that after clicking on “Finish Hackathon” you cannot submit any more entries. Finish the hackathon only when you have no more submissions to make. 6.The scores of your submission will be updated within five minutes after submission. To check the leaderboard after solution, click on the hackathon’s leaderboard. MachineHack comes up with a number of interesting hackathons to give many challenging problems to data science enthusiasts. Here are all its hackathons. It recently concluded its hackathon titled “Predict The Flight Ticket Price Hackathon”.","excerpt":"Who doesn’t love food? All of us have a craving for at least a few favourite food items, and many restaurants are on our “perfect spot” list. But if there’s one factor that makes us reconsider eating out is the cost. In MachineHack’s latest hackathon called Predicting Restaurant Food Cost Hackathon, you will be predicting […]","categories":["Deep Tech"],"tags":["dataset","predict","restaurant"],"author_name":"Disha Misal","publish_date":"2019-04-10T06:51:29","publication_year":"2019","word_count":454,"keywords":["dataset","data science","restaurant","predict","Go","programming_languages:R","AI","programming_languages:Go","RAG","R"],"extracted_tech_keywords":["AI","data science","RAG","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/machinehack-data-science-hackathon-use-ml-to-predict-costs-of-a-meal-at-a-restaurant\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":53359,"title":"These AI Toothbrushes Can Make Sure Nothing Is Stuck In Your Teeth","content":"By now, everyone who is familiar with artificial intelligence knows its immense applications in healthcare. Making its way through one’s body, AI has now reached the human being’s mouth. AI already demands a certain amount of data to work better. Now, apart from data like fingerprints, retina scans or X-rays the AI will now have the chance to look at your teeth and everything inside your mouth using the smart toothbrush. Colgate and Oral-B’s showcased this idea of oral hygiene with their smart toothbrushes at the ongoing CES 2020. This AI-based system will let the user know that there is something stuck in between their teeth, by telling the user that they missed a spot. Colgate’s Smart Toothbrush Colgate’s smart toothbrush has been trained to get paired with one’s smartphone. Colgate’s new Plaqless Pro smart toothbrush, which is an electric type, has tiny sensors embedded that can detect the plaque build up inside the user’s mouth before one starts brushing. One needs to pair their toothbrush to smartphones via Bluetooth, and the Colgate app analyses the mouth and the data for the user. This analysing and deciphering will be unique to each user’s mouth. The data collected and analysed will give details about whether the user’s brushing technique is a viable one and whether the individual has terrible brushing techniques, and all this analysis will be delivered in real-time on the Colgate application. This app will also tell the user if they have missed a spot and even the history of their brushing techniques along with customised oral care tips. According to Colgate’s Chief Dental Officer, Dr Maria Ryan, “Plaqless Pro also coaches the users to brush better, helping both the patient and the oral care provider to build an even stronger partnership in achieving optimal oral health.” Oral-B iO Oral-B too has introduced their version of the smart toothbrush. The brush is called Oral-B iO, which acts a motorised brushing mechanism. P&G said that the design of Oral-B iO is intended to combine oscillating, rotating movements with micro-vibrations to ensure super clean teeth and gums. Oral-B has been part of Proctor and Gamble since 2005. The Oral-B iO consists of bimodal pressure sensor which is embedded inside the toothbrush, and this sensor tells the users how much pressure must be applied while brushing. Like Colgate, Oral-B too has an app that can be downloaded on your smartphone, which will also be giving out information in real-time. This real-time information includes individual coaching and tracking thanks to the AI. This AI, which gives personal feedback, is trained on thousands of sessions of brushing. The app consists of 7 cleaning modes which include daily clean, sensitive, and whitens — it’s like a washing machine for your mouth. Are Colgate & Oral-B The Pioneers? Oral-B iO isn’t the first time smart toothbrushes were introduced. Oral- B, in 2014, has introduced a toothbrush which used Bluetooth and was programmed with the help of a dentist. In 2016 Oral -B has launched a line of Genius toothbrushes, but these needed a smartphone camera in front of you when you brush. Philips Sonicare and Kolibree have also offered toothbrushes like Oral-B’s Genius X. Kolibree has turned brushing into a game, and Onvi’s toothbrush live-streamed whatever was going inside the user’s mouth while brushing. Outlook Not laying more emphasis on how many areas AI is affecting — this new introduction feels more like breaking and entering. AI’s application with toothbrushes shouldn’t be surprising, rather what one should be looking at is how much more comfort will human beings try to achieve in the name of technology. Though one can present facts and numbers when it comes to oral hygiene’s importance, a $220 (Oral-B iO) toothbrush might not be a flexible option — what good is the technology when everyone can’t afford it?","excerpt":"By now, everyone who is familiar with artificial intelligence knows its immense applications in healthcare. Making its way through one’s body, AI has now reached the human being’s mouth. AI already demands a certain amount of data to work better. Now, apart from data like fingerprints, retina scans or X-rays the AI will now have […]","categories":["AI Features"],"tags":[],"author_name":"Sameer Balaganur","publish_date":"2020-01-08T15:26:58","publication_year":"2020","word_count":636,"keywords":["Go","artificial intelligence","programming_languages:R","AI","Modal","programming_languages:Go","Ray","R"],"extracted_tech_keywords":["AI","artificial intelligence","Ray","R","Go","Modal","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/these-ai-smart-toothbrushes-can-make-sure-nothing-is-stuck-in-your-teeth\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10099022,"title":"Twilio Segment Unveils CustomerAI Predictions Tool in Partnership with Amazon SageMaker","content":"Twilio Segment has unveiled a powerful tool designed to transform customer data into actionable insights. Dubbed CustomerAI Predictions, the tool aids marketers in making accurate forecasts regarding the behaviour of specific customer segments. Alex Millet, Senior Director of Product at Twilio Segment, emphasised the significance of quality data for marketers and highlighted the value that can be derived from existing customer data. The tool leverages information like clickstream data from websites or apps, combined with communication data from Twilio, to enable companies to comprehend customer engagement levels and identify areas requiring attention. Incorporating machine learning, CustomerAI Predictions is the result of collaboration between Twilio Segment and Amazon SageMaker. This strategic partnership facilitated the swift development of the predictions tool by utilising SageMaker’s advanced machine-learning infrastructure. As part of their roadmap, Twilio is exploring the integration of generative AI-based email tools, enabling marketers to craft personalised emails based on the data insights garnered from CustomerAI Predictions. Twilio’s journey into predictive analytics stems from its acquisition of Segment in 2020, signalling a strategic expansion beyond its core communications API business. The launch of CustomerAI Predictions underscores the growing importance of not just accumulating customer data, but effectively utilising it to enhance customer experiences and target campaigns with precision. Twilio’s endeavours to transform data into predictive insights align with the broader industry trend of leveraging data for strategic decision-making. As more companies acknowledge the value of predictive analytics in improving customer experiences and driving business growth, tools like CustomerAI Predictions play a pivotal role in shaping marketing strategies for the future. With a commitment to innovation, Twilio’s expansion into predictive analytics is set to empower marketers and businesses in optimising their engagement strategies for the evolving digital landscape.","excerpt":"Dubbed CustomerAI Predictions, the tool aids marketers in making accurate forecasts regarding the behaviour of specific customer segments.","categories":["AI News"],"tags":["Predictive AI","predictive analytics"],"author_name":"Shyam Nandan Upadhyay","publish_date":"2023-08-24T22:55:21","publication_year":"2023","word_count":285,"keywords":["API","Amazon SageMaker","machine learning","AI","R","Git","RAG","analytics","generative AI","predictive analytics","Predictive AI"],"extracted_tech_keywords":["AI","machine learning","analytics","generative AI","Amazon SageMaker","RAG","predictive analytics","R","Git","API"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/twilio-segment-unveils-customerai-predictions-tool-in-partnership-with-amazon-sagemaker\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10058749,"title":"Reliance picks up majority stake in robotics startup Addverb Technologies for USD 132 Mn","content":"Reliance Industries Limited has shelled out USD 132 million to pick up 54% stake in Indian robotics start-up Addverb Technologies to boost the conglomerate’s automation efforts. The Noida-based start-up builds robotics and automation solutions for factories and warehouses. Reliance already uses Addverb’s solutions, including robotic conveyors, semi-automated systems, and pick-by-voice software in its warehouses. The acquisition is in line with Reliance’s plans to extend its warehouses to hundreds of locations in a span of two years, said Sangeet Kumar, CEO and co-founder, Addverb Technologies. “We are glad to announce our strategic partnership with Reliance led by an investment of 132 mn USD in our Series B round. This strategic partnership will help us leverage 5G, battery technology through new energy initiatives and advances in material science of Reliance to deliver more advanced and affordable robots. This association will also provide us with an opportunity to deploy our robots at scale in omni-channel distribution centres across different segments like e-commerce, retail, grocery, fashion, pharma, digital and petrochemical. This funding will help us in expanding our footprint and help us deliver highly automated warehouses for global brands,” according to the company. Addverb has the capacity to churn out 10,000 robots a year including mobile robots, sorting robots, pallets shuttle and carton shuttle. 80% of the company’s revenues come from India. However, the founders hope to change ratio to 50:50 between India and overseas in the next four to five years. Addverb– founded in June 2016, by Sangeet Kumar, Prateek Jain, Bir Singh, Satish Kumar Shukla and Amit Kumar– counts Amazon, PepsiCo, Coca-Cola, HUL, Marico and Flipkart as its clients. The company plans to use the proceeds to expand its global footprint in Europe and the US. The startup has units in Singapore, the Netherlands, the US and Australia.","excerpt":"The intra-logistics solutions provider work across four verticals such as robotics, AS\/RS 3, picking and software","categories":["AI News"],"tags":["AI Startups"],"author_name":"Ebin K. Gheevarghese","publish_date":"2022-01-19T11:56:53","publication_year":"2022","word_count":296,"keywords":["funding","programming_languages:R","AI","Git","RAG","automation","ai_applications:robotics","R","AI Startups","startup"],"extracted_tech_keywords":["AI","RAG","R","Git","automation","startup","funding","programming_languages:R","ai_applications:robotics"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/reliance-picks-up-majority-stake-in-robotics-startup-addverb-technologies-for-usd-132-mn\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10161395,"title":"OpenAI Warns US of Losing $175 Bn to China","content":"OpenAI has cautioned that the United States risks losing an estimated $175 billion in global investment for AI projects to China-backed initiatives. In its newly released Economic Blueprint, the company emphasised the urgency of attracting these funds to maintain the nation’s leadership in artificial intelligence and counter the influence of the Chinese Communist Party. “If the US doesn’t attract those funds, they will flow to China-backed projects—strengthening the Chinese Communist Party’s global influence,” OpenAI said in its report. The blueprint calls for a unified national approach to foster innovation while implementing responsible regulations for AI development. OpenAI advocates for policies that protect public interests, promote free-market competition, and ensure developers and users can responsibly guide AI tools. The company also emphasised the importance of preventing government misuse of AI for coercion or citizen control. To advance its vision, OpenAI’s CEO Sam Altman will launch the Innovating for America initiative in Washington, DC, on January 30. This event will preview AI advancements and discuss their potential for driving economic growth across the country. OpenAI plans to engage with states nationwide to ensure AI’s benefits are distributed widely. OpenAI is also promoting the economic benefits of AI infrastructure development.“This will create tens of thousands of skilled-trade jobs, boost local economies through spending and indirect job creation, and modernise our energy grid in the near term,” the company said. OpenAI expressed its belief in the nation’s ability to lead in innovation, saying, “We believe in America because America believes in innovation.” The Economic Blueprint, described as a living document, will be updated as OpenAI gains insights from its work. Meanwhile, NVIDIA has strongly criticised the Biden administration’s new “AI Diffusion” rule, set to impose restrictions on global access to AI chips and technology. The company argues that the regulation, expected to take effect in 120 days, threatens to undermine U.S. leadership in artificial intelligence and stifle innovation worldwide. The Biden administration has introduced new restrictions on the export of US-developed computer chips used in artificial intelligence (AI) systems to prevent rivals like China from accessing advanced technology. Ned Finkle, vice president of government affairs at NVIDIA, said, “The Biden Administration now seeks to restrict access to mainstream computing applications with its unprecedented and misguided ‘AI Diffusion’ rule, which threatens to derail innovation and economic growth worldwide”. He further noted that under the previous Trump Administration, policies had fostered a competitive environment that allowed U.S. industries to lead in AI innovation without compromising national security.","excerpt":"“If the US doesn’t attract those funds, they will flow to China-backed projects—strengthening the Chinese Communist Party’s global influence.”","categories":["AI News"],"tags":["OpenAI"],"author_name":"Siddharth Jindal","publish_date":"2025-01-14T00:17:51","publication_year":"2025","word_count":411,"keywords":["Go","artificial intelligence","OpenAI","AI","programming_languages:R","innovation","programming_languages:Go","R"],"extracted_tech_keywords":["AI","artificial intelligence","OpenAI","R","Go","innovation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/openai-warns-us-of-losing-175-bn-to-china\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":24741,"title":"If You Love Beer and Machine Learning, MachineHack’s Next Hackathon Is For You","content":"Our online hackathon platform, MachineHack.com is growing by leaps and bounds. The very first hackathon, “Predicting House Prices In Bengaluru” is running successfully and top three winners are going to receive individual passes to Cypher 2018. We have received overwhelming response on the discussion boards and submissions and are happy to build a product for the data science community that is so cherished. The platform in the near future will be the hub for data science communities. Data science enthusiasts ranging from beginners to experts will find a new home in MachineHack and we are  working hard behind the scenes to give amazing challenges rewards to the data science community. We are focussing very hard to build great features into the platform and collaborate with the best in the industry to bring opportunities to budding data scientists out there. With this in mind we have decided to up the ante at MachineHack. We are launching our second inhouse hackathon called, “How To Choose The Perfect Beer”. Did you know that last year, Indians drank a total of 4.7 million litres of beer and the number is expected to go up to 6.5 billion litres by 2022. Choosing the perfect beer is complicated, therefore here at MachineHack, we have entrusted this very important job to the most trustworthy people in the world (especially when it comes to beer) to you, the data scientists. This is a perfect competition for persons who have a beginner’s level understanding of various concepts of machine learning and data science, and are looking to polish their understanding and check how they stand against a larger community. We have exciting prizes of ₹50,000 waiting for the winners. Yes, test and apply your machine learning skills to win: 1st Prize: ₹25,000 2nd Prize: ₹15,000 3rd Prize: ₹10,000 The train and test data will consist of various features that describe a beer. In many beer cellars, important factors such as temperature and humidity are maintained by a climate control system. Hence features like Cellar Temperature and Serving Temperature become really important. This is an actual data set that is curated over months of primary and secondary research by our team. Each row contains fixed size object of features. There are nine features and each feature can be accessed by its name. With the given features, build a model to predict the score of the beer and help everyone choose the perfect beer. We will not let out any more information more, get ready with your beer experience and machine learning skills and click here to participate.","excerpt":"Our online hackathon platform, MachineHack.com is growing by leaps and bounds. The very first hackathon, “Predicting House Prices In Bengaluru” is running successfully and top three winners are going to receive individual passes to Cypher 2018. We have received overwhelming response on the discussion boards and submissions and are happy to build a product for […]","categories":["Deep Tech"],"tags":[],"author_name":"Дарья","publish_date":"2018-05-21T05:50:32","publication_year":"2018","word_count":427,"keywords":["Go","data science","machine learning","programming_languages:R","AI","programming_languages:Go","Rust","R","programming_languages:Rust"],"extracted_tech_keywords":["AI","machine learning","data science","R","Go","Rust","programming_languages:R","programming_languages:Go","programming_languages:Rust"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/if-you-love-beer-and-machine-learning-machinehacks-next-hackathon-is-for-you\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10049093,"title":"Data Integration Platform Fivetran Raises $565 Million in Funding, Acquires HVR","content":"Automated data integration platform, Fivetran, has raised $565 million in a Series D round, consisting of new and existing investors. Andreessen Horowitz led the round with participation from existing investors General Catalyst, CEAS Investments, Matrix Partners, and others, along with new investors ICONIQ Capital, D1 Capital Partners and YC Continuity. Additionally, Fivertan has entered into a definitive agreement to acquire HVR, one of the leading providers of enterprise data replication technology. The acquisition will allow Fivetran to provide breakthrough database replication performance along with enterprise-grade security to address the massive market for modernizing analytics for operational data associated with ERP systems, Oracle databases, and more. “HVR is a recognized leader for enterprise database replication and shares our same vision – to make access to data as simple and reliable as electricity. Their product is the perfect complement to our automated data integration technology and will be instrumental for us to help enterprise organizations that want to improve their analytics with a modern data stack. The funding gives Fivetran the resources to expand our capabilities across all functions of the business and continue to accelerate our growth globally.” said George Fraser, CEO of Fivetran. The combination of Fivetran and HVR complementary technologies will accelerate Fivetran’s mission. Anthony Brooks-Williams, CEO of HVR, said, “Combining HVR and Fivetran will enable a next-generation solution that will better inform business decisions by providing the freshest data available. We’re thrilled to be joining forces with Fivetran and look forward to what this incredible opportunity will provide for our growing team, partners and customers.” Fivetran has raised $730 million to date. The company is now valued at $5.6 billion. Fivetran customers will gain access to a broader range of high-performance data replication solutions, including HVR’s change data capture connectors and a secure, on-premise option for companies with legacy databases. HVR’s customers will have access to Fivetran’s broad range of pre-built, fully managed data connectors and transformation capabilities. The combined capabilities of Fivetran and HVR will enable modern analytics for the world’s most business-critical data without compromising security, performance or ease of use. The Fivetran customer roster includes thousands of global companies such as ASICS, Autodesk, BJ’s Restaurants, Conagra Brands, DocuSign, Forever 21, Lionsgate, Square and Ziff Davis. HVR’s customers include dozens of Fortune 500 brands.","excerpt":"The combined capabilities of Fivetran and HVR will enable modern analytics for the world’s most business-critical data without compromising security, performance or ease of use.","categories":["AI News"],"tags":["data integration","Database","Machine Learning"],"author_name":"Victor Dey","publish_date":"2021-09-21T11:26:26","publication_year":"2021","word_count":378,"keywords":["API","funding","programming_languages:R","AI","Database","Machine Learning","data integration","analytics","GAN","R"],"extracted_tech_keywords":["AI","analytics","R","API","GAN","funding","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/data-integration-platform-fivetran-raises-565-million-in-funding-acquires-hvr\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10039175,"title":"Mastercard Donates $10 Million To Address India’s Current Crisis","content":"Recently, Mastercard announced that the company will be donating $10 Million (approximately ₹75 Crore) in order to address the critical COVID-related needs in India. The payments and technology company has already committed to cover all vaccine-related costs for its India-based employees and immediate families, as doses are available. The initiative has been delivered through the Mastercard Impact Fund and will focus on three priority areas – Access to hospital resources, Access to additional oxygen supplies, and Continued support of the company’s employees across the country. According to its official blog post, the largest contribution will be designated to support the expansion of 2,000 beds through the installation of portable hospitals that can address the immediate healthcare needs. It stated, “Working with the government and local partners on the ground, these hospitals can be constructed quickly and could help an estimated 2.5 million Indians gain access to healthcare, adding to the nation’s healthcare infrastructure.” Ajay Banga, executive chairman of Mastercard said, “The situation in India is gut-wrenching; it’s clear that no one has been left untouched and that no one can be a bystander.” “We have long been advocates for the people of India, as employers and as enablers of the economy. Now it’s time for us to put our shoulders to the wheel and do whatever we can to help all of India get through this intensified crisis,” Banga added. In addition to the on-the-ground medical support, Mastercard is also allocating funding for additional shipments of 1,000 oxygen generators, as part of a broader corporate effort. And the company is donating to local charities and NGOs to allow for direct support of those affected by the virus, including the frontline medical workers across the country. Commenting on the same, Richard Verma, general counsel at Mastercard and former U.S. ambassador to India stated, “The U.S. and India have a shared history of helping each other during times of crisis. We’re hearing from our own teams the challenges that they and their families are facing. It’s critical that we stand by our friends and help those who have been impacted.  Our commitment today and into the future is strong and enduring.” It has also been mentioned that in order to help extend these efforts and provide another outlet for its employees across the globe to help each other, Mastercard will also match employee donations to its Employee Assistance Fund, up to $15,000 per individual. This comes on top of a $500,000 corporate contribution to the fund. The company will continue to explore other opportunities to activate partnerships in India and across the global community.","excerpt":"Recently, Mastercard announced that the company will be donating $10 Million (approximately ₹75 Crore) in order to address the critical COVID-related needs in India. The payments and technology company has already committed to cover all vaccine-related costs for its India-based employees and immediate families, as doses are available.   The initiative has been delivered through the […]","categories":["AI News"],"tags":["Mastercard"],"author_name":"Ambika Choudhury","publish_date":"2021-04-29T11:43:17","publication_year":"2021","word_count":432,"keywords":["Go","Mastercard","funding","programming_languages:R","AI","programming_languages:Go","R"],"extracted_tech_keywords":["AI","R","Go","funding","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/mastercard-donates-10-million-to-address-indias-current-crisis\/","complexity_score":3,"technical_depth":6,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10012538,"title":"This AI Model Can Figure Out Video Games By Its Cover","content":"Recently, researchers from the Western Kentucky University proposed a multi-modal deep learning framework that has the capability to classify genres of video games based on the cover and textual description. The researchers claimed that this research is the first-ever attempt on automatic genre classification using a deep learning approach. Videos games have been one of the most widespread, profitable, and prominent forms of entertainment around the globe. Also, genre and its classification systems play a significant role in the development of video games. According to the researchers, video game covers and textual descriptions are usually the very first impression to its consumers, and they often convey important information about the video games. However, researchers often find it difficult to classify video game genres based on its cover and textual description because of various reasons like a massive number of video game genres, many of which are not concretely defined; cover designs; and textual descriptions that may vary due to many external factors such as country, culture, and other such. Also, with the growing competitiveness in the video game industry, the cover designers and typographers push the cover designs to its limit in the hope of attracting sales. In order to mitigate such problems, the researchers built this new deep learning framework. The Tech Behind For this, the researchers aimed to develop three deep learning algorithm for the task of video game genre classification, which are- An image-based approach using the game coversA text-based approach using the textual descriptionsA multimodal approach using both the game covers and textual description They evaluated five image-based models and two text-based models using deep transfer learning methods for the task of video game genre classification. The image-based models include MobileNet-V1, MobileNet-V2, Inception-V1, Inception-V2 and ResNet-50. Also, the two text-based models include recurrent neural networks (RNN) with Long Short-Term Memory (LSTM). In addition to the cover images of the video games, they also used the game descriptions for genre classification. The main purpose of the game description is to express the objects involved in a game and the set of rules that induce transitions, resulting in a state-action space. On the final step, the researchers considered a multimodal deep learning architecture based on both the game cover and description for the task of genre classification. This approach again involves two steps, which are: A neural network is trained on the classification task for each modalityIntermediate representations are extracted from each network and combined in a multimodal learning step. According to the researchers, the information from both modalities are then combined using the concatenation method, and hence it helped in increasing the classifying accuracy rate. Dataset Used To perform this research, the researchers created a large dataset of 50,000 video games that includes game cover images, description text, title text, and genre information from IGDB.com, a video game database. There were in total 21 genres found in the original dataset; however, they churned out 15 different genres, such as adventure, arcade, fighting, strategy, among others. The collected dataset can be used for a number of studies such as text recognition from images, automatic topic mining, and other such. Contributions Of This Research The researchers contributed to this research in a four-fold way: Firstly, they compiled a large dataset consisting of 50,000 video games from 21 genres made of cover images, description text, and title text and the genre information. Secondly, image-based and text-based, state-of-the-art models are evaluated thoroughly for the task of genre classification for video games.Thirdly, they developed an efficient multi-modal framework based on both images and texts.Lastly, a thorough analysis of the experimental results is shown by the researchers as well as future work to improve the performance is also suggested. Read the paper here.","excerpt":"Recently, researchers from the Western Kentucky University proposed a multi-modal deep learning framework that has the capability to classify genres of video games based on the cover and textual description. The researchers claimed that this research is the first-ever attempt on automatic genre classification using a deep learning approach. Videos games have been one of […]","categories":["Deep Tech"],"tags":["churn prediction model","Deep Learning","deep learning framework"],"author_name":"Ambika Choudhury","publish_date":"2020-11-27T15:03:53","publication_year":"2020","word_count":620,"keywords":["Go","LSTM","AI","neural network","Modal","ResNet","Aim","deep learning","RNN","churn prediction model","deep learning framework","Deep Learning","R"],"extracted_tech_keywords":["AI","deep learning","neural network","Aim","R","Go","RNN","LSTM","ResNet","Modal"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/this-ai-model-can-figure-out-video-games-by-its-cover\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10163393,"title":"Snowflake’s New Cortex Agents Bridge AI and Enterprise Data Like Never Before","content":"Snowflake has launched Cortex Agents, a fully managed service aimed at integrating, retrieving, and processing structured and unstructured data at scale. Cortex Agents plan tasks, execute them using tools, and refine responses for improved accuracy. Available via a REST API, the service integrates into applications using Cortex Analyst for structured data (SQL generation) and Cortex Search for unstructured data retrieval. The new solution, now in public preview, enables businesses to build AI-driven applications with advanced governance and security features. The Cortex Agents framework integrates Anthropic’s Claude 3.5 Sonnet, which enhances reasoning, coding, and workflow execution within Snowflake’s secure perimeter. Cortex Agents review user queries and break them into structured and unstructured components. For example, a business user can request top distributors by revenue (structured) and then ask about contract details (unstructured). The agent disambiguates queries, splits tasks, and selects tools for execution. Cortex Analyst, a fully managed LLM-powered Snowflake Cortex feature, generates SQL for structured data, while Cortex Search retrieves insights from text, audio, and images. The system ensures governed access and compliance while mapping business terms to structured data using semantic understanding. Snowflake reports that Cortex Analyst has achieved 90% accuracy in text-to-SQL use cases. On the other hand, Cortex Search has outperformed OpenAI’s embedding models by 12% in unstructured data retrieval accuracy. It supports large-scale indexing, improved affordability, and customisable vector embeddings. Snowflake also introduced Cortex AI Observability, powered by TruLens, for the evaluation and tracing of AI agents. “AI observability can evaluate agent performance using techniques such as LLM-as-a-judge, allowing customers to refine and optimise their applications,” Snowflake said. Snowflake sees AI agents as a transformative force for enterprises, automating complex tasks and improving efficiency across industries such as finance, engineering, and customer support. “As LLMs continue to advance, agents will collaborate, plan, execute, and refine tasks, driving efficiency and reducing costs,” the company said.","excerpt":"The Cortex Agents framework integrates Anthropic’s Claude 3.5 Sonnet to enhance reasoning, coding, and workflow execution.","categories":["AI News"],"tags":["Snowflake"],"author_name":"Siddharth Jindal","publish_date":"2025-02-12T22:28:36","publication_year":"2025","word_count":308,"keywords":["Anthropic","Go","API","OpenAI","AI","Aim","SQL","Claude 3.5","R","Snowflake"],"extracted_tech_keywords":["AI","OpenAI","Claude 3.5","Anthropic","Aim","Snowflake","R","SQL","Go","API"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/snowflakes-new-cortex-agents-bridge-ai-and-enterprise-data-like-never-before\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10045763,"title":"Do You Find Premium Data Science Courses Expensive? Ivy Pro School Becomes NASSCOM FutureSkills Partner &#038; Launches AI\/ML &#038; Data Science Certification Course With Pay After Placement Option","content":"The Data Science industry is going through an unseen demand cycle for the trained talent because of the massive digitization across the industries. Hence, the data science skill-building courses have become one of the most popular courses in the world. With a large number of interested applicants growing recently all over the world, the premium data science course prices have increased significantly. The prices of the premium data science courses have always been a roadblock to the meritorious students. Ivy Professional School, 2nd ranked data science school in India with 12+ years of experience in helping Fortune 500 organizations and 22,500+ learners build high impact data skills, has launched a Deep Dive in Data Science, ML and AI certification course with Income Sharing Agreement (ISA) option or “Learn Now, Pay When You Get a Job model”. In this model, the meritorious data science enthusiasts can take up the course without paying an upfront fee. They will be only eligible to pay when they get a job with a minimum CTC of between INR 4 and 15 Lakhs per annum. This course is also aligned to the competency standards developed by SSC NASSCOM in collaboration with industry and approved by the Government of India. The industry also thinks that a program like this is the need of the hour. The CEO and founder of TheMathCompany, a leading full-stack data science solution provider with a global footprint, Sayandeb Banerjee, says, “The Data Science industry is going through an unseen demand cycle. TheMathCompany needs to be trained with full-stack data science resources to continue solving complex business problems, and hence the need of rightly designed courses for building a pool of meritorious full-stack data science talent becomes the need of the hour.” Why is Data Science so important in today’s world? The multiple waves of the pandemic have made digitization mandatory for all classes of businesses. Digitization generates a huge amount of data. The companies want to use this data to make informed business decisions. Hence, the demand for the trained data science talent is increasing exponentially. In August 2020, 9.8% of the total global job opportunities had Analytics job openings. This is an increase of 2.6% in 8 months. Interestingly, India contributed 7% of the Analytics jobs worldwide. Analytics is currently one of India’s finest full-time jobs. About 98% of the Analytics jobs in India are full-time. This not only signifies the strength of the Indian Analytics hiring market but also showcases the credibility of the upcoming job opportunities. In addition to that, Accenture, Mphasis, Cognizant Technology Solutions, Capgemini, Infosys, Tech Mahindra, Dell, IBM India, and HCL Technologies are some of the companies that provided the most number of jobs in the Analytics domain in August 2020. Various factors make data science an important job skill in today’s world. Whether it helps you make a lot of money or provides job security, it is considered one of the most satisfying jobs today. Some of the factors that make data Science important are : Provides upgraded skill set: According to LinkedIn, Analytical skills are one of the essential skills for any data science-related job. Some of those skills include: Data WranglingData VisualizationMachine LearningStatistical Analysis In addition to the above, it includes data-driven skills like Artificial Intelligence as well. Handsome pay: The popularity of data science has taken its importance to such a level that the percentage of data science professionals who earn more than 25 lacs has doubled in 2 years. Evolving in every sector: Healthcare, Education, or Agribusiness, there has been a huge demand for Data Scientists. With this increase in the demand for such analysts, companies are banking on them for their business by putting them at the forefront. What is this program about, and why should one go for it? Ivy Professional School aims to increase access to quality Data Science education in India. They not only focus on delivering outcomes but also focus on enhancing the employability options of a candidate. They have come up with employability-focused Data Science courses for all the deserving candidates via an Income Sharing Agreement. It is one of the rarest programs for those who want to boost their career in data science. In addition to that, it comes with benefits that focus on the student’s flexibility and willingness to do something better in their career. There are various subjects in this specific program. This usually depends upon the demand of concepts, tools and technologies that companies want analysts to have their expertise on. The subjects covered in this particular program are: Python for Data SciencePredictive Modeling using Machine Learning Artificial Intelligence & Deep Learning EssentialsMachine Learning Cloud DeploymentData Visualization & Storytelling Using Tableau \/ PowerBIAdvanced SQL Queries and Relational Database ManagementAutomation using Excel VBADashboarding using MS-Excel This course outline is carefully designed, keeping in mind the diversity of subjects. Moreover, these help in the understanding of a skilled Data Engineer\/Data Analyst\/Data Scientist. In addition to that, it includes tools that focus on concepts of data extraction, data cleaning, data processing, data visualization, and data analysis. Kirti Seth, Co-Architect FutureSkills Prime from NASSCOM, said the following about this program, “Demand for Data Science professionals is increasing at a skyrocketing speed. In the direction of bridging the demand-supply gap, this course by Ivy Pro School is a step towards creating an army of digital talent pool with a promise of a bright future ahead. The model of aligning course fees to income earned is an innovative one that understands what learners need for professional courses. In addition, we are delighted that the course design is based on an alignment to NASSCOM’s Competency Standards, giving learners an assurance that they are learning skills that industry actually looks for when hiring candidates.” What do you mean by the Income Sharing Agreement, and how does it work? The Income Sharing Agreement or ‘Learn Now, Pay When You Get Job’ program allows the deserving candidates to pay the tuition of the course once they get a job with the minimum guaranteed CTC. Even after getting a job, the candidates don’t pay the entire tuition amount at once. They will pay 17.5% of their monthly salary for a maximum period of 3 years, capped at a total amount of INR 1,35,000. This is indeed a very generous and confident step that Ivy Professional School has taken forward for the betterment of the job seekers in the Data Science \/ Analytics industry. Here’s how a learner can join the course: Sign-Up for the program and complete the three steps selection process to get selected.Sign the Income Sharing Agreement and set forward to complete the 24-week program with industry experts.Get placed as a Data Scientist\/Analyst in top corporate firms with the minimum guaranteed CTC. Pay 17.5% of your monthly salary for up to 3 years post job capped at INR 1,35,000 + GST if you get a job paying more than the minimum guarantee CTC. The Director and Co-founder of Ivy Professional School, Eeshani Agrawal, says, “The best way to enhance the Employability Factor is by making yourself skilled. Ivy believes that no meritorious student should be deprived of quality training just because of financial constraints. Ivy Pro School’s Deep Dive in Data Science, AI and ML course, with Pay After Placement option, Aligned to competency standards by SSC NASSCOM, developed in collaboration with Industry & Approved by Government of India, is the right fusion of multiple skills needed to become a Data Scientist.” What do the alumni of this program have to say about it? Listen to what their alumni Shreya Gupta has to say about the program and about Ivy Professional School. Shreya is currently working as a Business Analyst at Bridgei2i, a leading data science consulting firm. Shreya believes that disciplined learning and an extensive curriculum were one of the biggest reasons she was able to get hold of the program very well. Who is eligible for this program? Here are some of the eligibility criteria for this program: Graduates \/ Postgraduates \/ Final year students from the field of computer science, economics, statistics, mathematics, or other equivalent degrees.The ideal age bracket is 22-30 years.Should have necessary soft skills or interpersonal qualitiesWorking professionals from the aforementioned streams (below 30 years of age) If you are such an enthusiast who likes to make a difference, here’s your chance to get a kick start. Get in touch with Ivy’s amazing team, and they will be right with you for your help. They have a range of things to offer, from certifications on tools to languages and reporting. Based on the student’s profile and the career that one is looking to pursue, one can choose for themselves. About Ivy Professional School: Launched in 2007, Ivy Professional School is a leader in Corporate Training and is ranked among the Top 10 Data Science Schools for seven consecutive years (2014-2020). Ivy has trained 22,500+ Professionals from Honeywell, Genpact, EY, Tesco, PwC, Tata Steel, ITC Ltd., ICRA, eBay\/PayPal, HSBC, Frost & Sullivan, Citibank, HDFC, Capgemini, Cognizant, etc. and elite educational institutes such as IITs, IIMs, ISI, Delhi Univ, DCE, Calcutta University, etc. Its corporate skilling wing focuses on courses in Analytics, Big Data, Tech, Organizational Behavior, & Foreign Language competencies. Ivy is also the official learning partner of reputed global companies like Accenture, Genpact, Honeywell, Tata Steel, Tesco, PwC, General Mills, Coca Cola India, ITC Limited, ICRA, ITC Infotech, Capgemini, HSBC, Cognizant, Lexmark Inc., etc. About NASSCOM: The National Association of Software and Services Companies (NASSCOM®) is the premier trade body and chamber of commerce of the Tech industry in India and comprises over 2800-member companies, including both Indian and multinational organizations that have a presence in India. NASSCOM membership spans across the entire spectrum of the industry from start-ups to multinationals and from products to services, Global Service Centers to Engineering firms. NASSCOM membership base constitutes over 95% of the industry revenues in India and employs over 4 million professionals. NASSCOM envisions making India a global hub for Innovation and Talent, so when the world thinks Digital, the world will think India. About FutureSkills: NASSCOM FutureSkills is an industry-driven learning ecosystem to get India accelerated on the journey to becoming the global hub for talent in emerging technologies. Announced in February 2018 by Hon’ble Prime Minister of India Shri Narendra Modi, FutureSkills ecosystem hosts high-quality learning content curated by industry and academia experts for the potential employees & students in the industry to reskill in emerging technologies and professional skills.","excerpt":"Ivy Pro School has launched AI\/ML and Data Science courses with Income Sharing Agreement (ISA) option or “Learn Now, Pay When You Get a Job model”.","categories":["AI Trends"],"tags":["big data quality","Courses","Data Scientist Jobs","data scientist salary in india","indian statistical service","list of computer languages and their uses","NASSCOM","statistical analysis using sql","what is data science"],"author_name":"AIM Media House","publish_date":"2021-08-12T10:00:00","publication_year":"2021","word_count":1746,"keywords":["big data quality","statistical analysis using sql","deep learning","R","data science","Data Scientist Jobs","artificial intelligence","analytics","what is data science","indian statistical service","machine learning","AI","ML","data scientist salary in india","list of computer languages and their uses","Python","Aim","Courses","NASSCOM"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","data science","analytics","Aim","Python","R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/do-you-find-premium-data-science-courses-expensive-ivy-pro-school-becomes-nasscom-futureskills-partner-launches-ai-ml-data-science-certification-course-with-pay-after-placement-option\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10170549,"title":"SAP to Open 15,000-Seat Centre of Excellence in Bengaluru by August","content":"German tech giant SAP is set to open a new centre of excellence in Bengaluru by August this year, as per reports. The facility, with a seating capacity of 15,000, aims to support the company’s growing business needs both in India and around the world. Speaking at SAP Sapphire 2025, Muhammad Alam, executive board member at SAP, mentioned that India is one of its largest development centres and is opening a new facility later this year with even more capacity. Given its size, from a product engineering perspective, it is a very self-sufficient and decoupled environment. According to Manish Prasad, president and managing director of SAP Indian Subcontinent, the new centre is expected to become operational in the second half of the year, possibly by July or August. This move reinforces India’s key role in SAP’s global operations, particularly in engineering and product development. Such a development also reinstates that India is a sweet spot for the German tech conglomerate SAP. With over 15,000 customers already based in India and over 4,25,000 globally, the company’s second-largest R&D centre outside Germany is in Bengaluru. Recently, TCS has also deepened its longstanding partnership with SAP to accelerate enterprise-wide adoption of generative AI and cloud technologies, aiming to support large-scale business transformation for global customers. Building on a collaboration that spans over two decades, TCS will work closely with SAP to simplify cloud migration through the ‘RISE with SAP’ initiative.","excerpt":"The facility, with a seating capacity of 15,000, aims to support the company’s growing business needs both in India and around the world.","categories":["AI News"],"tags":["SAP"],"author_name":"Shalini Mondal","publish_date":"2025-05-22T17:41:24","publication_year":"2025","word_count":237,"keywords":["programming_languages:R","AI","SAP","Aim","generative AI","R"],"extracted_tech_keywords":["AI","generative AI","Aim","R","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/sap-to-open-15000-seat-centre-of-excellence-in-bengaluru-by-august\/","complexity_score":2,"technical_depth":5,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10040981,"title":"Beginners Guide To Text Generation With RNNs","content":"Text Generation is a task in Natural Language Processing (NLP) in which text is generated with some constraints such as initial characters or initial words. We come across this task in our day-to-day applications such as character\/word\/sentence predictions while typing texts in Gmail, Google Docs, Smartphone keyboard, and chatbot.  Understanding of text generation forms the base to advanced NLP tasks such as Neural Machine Translation. This article discusses the text generation task to predict the next character given its previous characters. It employs a recurrent neural network with LSTM layers to achieve the task. The deep learning process will be carried out using TensorFlow’s Keras, a high-level API. Let’s dive deeper into hands-on learning. Create the Environment Import the necessary frameworks, libraries and modules to create the required Python environment. Since we work with text data, an Embedding layer will be required. Since we build LSTM recurrent neural networks, an LSTM layer will be required. In addition, a Dense layer will be of use to develop a classification head. import numpy as np import tensorflow as tf from tensorflow import keras from tensorflow.keras.layers import Dense, LSTM, Embedding import matplotlib.pyplot as plt Download Text Data We need text data to train our model. TensorFlow’s data collection has a text file with contents extracted from various Shakespearean plays. Download the data file from Google Cloud Storage. file_URL = \"https:\/\/storage.googleapis.com\/download.tensorflow.org\/data\/shakespeare.txt\" file_name= \"shakespeare.txt\" # get the file path path = keras.utils.get_file(file_name, file_URL) Output: Open the downloaded file and read its content. Print some sample portions from the text. raw = open(path, 'rb').read() print(raw[250:400]) Output: Rather than reading ‘\\n’ as a newline character and printing the consequent characters in the next line, Python reads it as part of text characters. It is because the original downloaded file is UTF-8 encoded. We need to decode the text into Python readable string format. text = raw.decode(encoding='utf-8') print(text[250:400]) Output: What is the length of the downloaded text? len(text) Output: The text has more than one million characters. Vectorize Word Characters into Integers A deep learning model can not accept text characters as inputs. It should be encoded into integers that a model can understand and process with. Though there are more than a million characters in the given text, there will be a countable number of unique characters. The collection of unique characters is called vocabulary. # unique characters vocabulary = np.array(sorted(set(text))) len(vocabulary) Output: Define a tokenizer that can convert a text character into a corresponding integer. There will be 65 integers starting from 0 and ending at 64. We can assign integers on our own as per the order of characters in the vocabulary. # assign an integer to each character tokenizer = {char:i for i,char in enumerate(vocabulary)} What integers are assigned to what characters? Sample the first 20 characters. # check characters and its corresponding integer for i in range(20): char = vocabulary[i] token = tokenizer[char] print('%4s : %4d'%(repr(char),token)) Output: Vectorize the entire text and check whether the built tokenizer can encode and decode – texts and integers properly. vector = np.array([tokenizer[char] for char in text]) print('\\nSample Text \\n') print('-'*70) print(text[:100]) print('-'*70) print('\\n\\nCorresponding Integer Vector \\n') print('-'*70) print(vector[:100]) print('-'*70) Output: Text with one million encoded characters can not be fed into a model as such. Since we predict characters, the text must be broken down into sequences of some predefined length and then fed into the model. Use TensorFlow’s batch method to create sequences of 100 characters each. Prior to that, convert the NumPy arrays into tensors to make further processes using TensorFlow. # convert into tensors vector = tf.data.Dataset.from_tensor_slices(vector) # make sequences each of length 100 characters sequences = vector.batch(100, drop_remainder=True) Recurrent neural networks predict the subsequent characters based on the past characters. RNNs require a sequence of input characters and the corresponding target sequence with the subsequent characters for training. Prepare input sequences with the first 99 characters and corresponding target sequences with the last 99 characters. def prepare_dataset(seq): input_vector = seq[:-1] target_vector = seq[1:] return input_vector, target_vector dataset = sequences.map(prepare_dataset) Let’s sample the first sequence pair. # check how it looks for inp, tar in dataset.take(1): print(inp.numpy()) print(tar.numpy()) inp_text = ''.join(vocabulary[inp]) tar_text = ''.join(vocabulary[tar]) print(repr(inp_text)) print(repr(tar_text)) Output: Batch and Prefetch data Model will be trained with Stochastic Gradient Descent (SGD) based optimizer Adam. It requires the input data to be batched. Further, TensorFlow’s prefetch method helps training with optimized memory. It fetches data batches just before the training requires them. We prefer not to shuffle the data to retain the contextual order of sequences. AUTOTUNE = tf.data.AUTOTUNE # buffer size 10000 # batch size 64 data = dataset.batch(64, drop_remainder=True).repeat() data = data.prefetch(AUTOTUNE) # steps per epoch is number of batches available STEPS_PER_EPOCH = len(sequences)\/\/64 for inp, tar in data.take(1): print(inp.numpy().shape) print(tar.numpy().shape) Output: Build an RNN Model Recurrent neural networks are good at modeling time-dependent data because of its ability to retain time-steps based information in memory. Since texts have contextual information that are determined purely by order of the words, natural language processing heavily relies on sequence modeling architectures such as RNN. Here, an LSTM (Long Short-Term Memory) layers-based recurrent neural network is developed to model the task. While implementing LSTM layers, we enable stateful argument as True to keep the time-step memory of previous states while learning with consequent batches in an epoch. It helps capture the context among consecutive sequences. model = keras.Sequential([ # Embed len(vocabulary) into 64 dimensions Embedding(len(vocabulary), 64, batch_input_shape=[64,None]), # LSTM RNN layers LSTM(512, return_sequences=True, stateful=True), LSTM(512, return_sequences=True, stateful=True), # Classification head Dense(len(vocabulary)) ]) model.summary() Output: Plot the model to understand the flow and shapes of data at each layer’s input and output. keras.utils.plot_model(model, show_shapes=True, dpi=64) Output: Train the RNN Model We can check whether the model can accept the processed data without any errors. # test whether the untrained model performs good for example_inp, example_tar in data.take(1): example_pred = model(example_inp) print(example_tar.numpy().shape) print(example_pred.shape) Output: The target shape is (64, 99), which refers to the batch size and the number of characters in that sequence. The last shape, 65, in the prediction refers to the size of the vocabulary. The model predicts the probability of occurrence of each character in the vocabulary. The character with a higher probability has more possibility to be the next character. Compile the model with Adam optimizer and Sparse Categorical Cross-entropy loss function. Since we have not employed softmax as the output layer’s activation function. The outputs will be independent but not mutually exclusive. Hence, we should enable the argument ‘from_logits’ to be True while declaring the loss function. Train the model for 10 epochs. model.compile(optimizer='adam', loss=keras.losses.SparseCategoricalCrossentropy(from_logits=True)) history = model.fit(data, epochs=10, steps_per_epoch=STEPS_PER_EPOCH) output: Training took almost an hour to complete in a CPU runtime enabled virtual machine. Model Performance Evaluation Visualizing the losses over epochs may help get better insight on model performance. plt.plot(history.history['loss'], '+-r') plt.title('Performance Analysis', size=16, color='green') plt.xlabel('Epochs', size=14, color='blue') plt.ylabel('Loss', size=14, color='blue') plt.xticks(range(10)) plt.show() Output: Loss keeps on falling down even till the 10th epoch. It suggests that the model should be trained for more epochs until convergence occurs. The smoothness in the loss curve suggests that the learning rate is proper for this model configuration. Inference – Next Character Prediction The most awaited part of this task is predicting the next character with the trained model. We can input the model some characters (probably a word) such that it will iteratively predict the next 1000 characters. Before starting prediction with the model, we should reset the model states that were stored in the memory during the last epoch training. However, resetting state memories will not affect the model’s weights. # reset previous states of model model.reset_states() Make predictions by providing the model ‘ANTHONIO:’ as input characters. Nevertheless, the model expects data in three dimensions: the first dimension being the batch size, 64. Vectorize the input characters, expand the dimensions, broadcast the same vector 64 times to obtain a batch of size 64 sequences. Predictions are made based on the logits output by the model. This can be sensitively adjusted by tuning a hyper-parameter called temperature, which refers to the level of randomness in choosing the probable outcome. sample = 'ANTHONIO:' # vectorize the string sample_vector = [tokenizer[s] for s in sample] predicted = sample_vector # convert into tensor of required dimensions sample_tensor = tf.expand_dims(sample_vector, 0) # broadcast to first dimension to 64 sample_tensor = tf.repeat(sample_tensor, 64, axis=0) # predict next 1000 characters # temperature is a sensitive variable to adjust prediction temperature = 0.6 for i in range(1000): pred = model(sample_tensor) # reduce unnecessary dimensions pred = pred[0].numpy()\/temperature pred = tf.random.categorical(pred, num_samples=1)[-1,0].numpy() predicted.append(pred) sample_tensor = predicted[-99:] sample_tensor = tf.expand_dims([pred],0) # broadcast to first dimension to 64 sample_tensor = tf.repeat(sample_tensor, 64, axis=0) # convert the integers back to characters pred_char = [vocabulary[i] for i in predicted] generated = ''.join(pred_char) print(generated) Output: By adjusting the temperature value, we can vary randomness and obtain different predictions. sample = 'ANTHONIO:' # vectorize the string sample_vector = [tokenizer[s] for s in sample] predicted = sample_vector # convert into tensor of required dimensions sample_tensor = tf.expand_dims(sample_vector, 0) # broadcast to first dimension to 64 sample_tensor = tf.repeat(sample_tensor, 64, axis=0) # predict next 1000 characters # vary temperature to change randomness temperature = 0.8 for i in range(1000): pred = model(sample_tensor) # reduce unnecessary dimensions pred = pred[0].numpy()\/temperature pred = tf.random.categorical(pred, num_samples=1)[-1,0].numpy() predicted.append(pred) sample_tensor = predicted[-99:] sample_tensor = tf.expand_dims([pred],0) # broadcast to first dimension to 64 sample_tensor = tf.repeat(sample_tensor, 64, axis=0) # integer to text decoding pred_char = [vocabulary[i] for i in predicted] generated = ''.join(pred_char) print(generated) Output: Most of the predicted words are really English words! But the prediction lacks context. We can not make any meaning out of the predicted sentences. Training for more epochs may help improve the performance of the model. However, a character-prediction model can not cover context greatly compared to a word-prediction model. This Notebook contains the above code implementation. Wrapping Up This article has discussed the concepts of text generation using recurrent neural networks. It has explored a next-character prediction task with practical data by building a deep learning RNN model, training it and making inferences on sample characters. Interested readers can modify the model with word-level vectorization approaches (such as word2vec) to make next-word predictions. References Official Tutorial on RNNOfficial Tutorial on Text Generation TensorFlow Official DatasetsShakespeare plays text data Further Reading Hands-On Guide To Markov Chain For Text GenerationA Brief Overview of OpenChat: Open Source Chatting Framework for Generative ModelsENCONTER: Entity Constrained Insertion Transformer Language modelingResearchers Decode Brain Scans To Generate Text","excerpt":"Text Generation is a task in Natural Language Processing in which text is generated with some constraints such as initial characters words","categories":["Deep Tech"],"tags":["Guide","Keras","lstm","Natural Language Processing","NLP","RNN","Tensorflow","text generation","tokenization"],"author_name":"Rajkumar Lakshmanamoorthy","publish_date":"2021-05-30T12:00:00","publication_year":"2021","word_count":1744,"keywords":["NumPy","Keras","AI","neural network","Natural Language Processing","lstm","tokenization","RAG","NLP","Ray","Matplotlib","deep learning","RNN","TensorFlow","text generation","Guide","Tensorflow"],"extracted_tech_keywords":["AI","deep learning","neural network","NLP","Ray","TensorFlow","Keras","NumPy","Matplotlib","RAG"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/beginners-guide-to-text-generation-with-rnns\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10164071,"title":"New GPT-4o Copilot Model Unveiled for VS Code Users in Public Preview","content":"GitHub announced that a new code completion model, GPT-4o Copilot, is now available in public preview for Visual Studio Code users. Based on GPT-4o mini, the model has been trained on more than 2,75,000 high-quality public repositories across over 30 programming languages. “Our new code completion model is shipping in public preview today. We are calling it GPT-4o Copilot. Based on GPT-4o mini, with mid-training on a code-focused corpus exceeding 1T tokens and reinforcement learning with execution feedback (RLEF),” said GitHub chief Thomas Dohmke. He added that the company used real-life coding workflows and fine-tuned the model across 32 major programming languages. According to GitHub, this update improves the accuracy of suggestions and enhances performance. Users can enable the model by opening the Copilot menu in the VS Code title bar and selecting ‘Configure Code Completions… > Change Completions Model…’ or by using the Command Palette and selecting ‘GitHub Copilot: Change Completions Model…’. For Copilot Business and Enterprise users, an administrator must enable the model by opting for Editor preview features in the Copilot policy settings on GitHub. Free-tier users can access the model within their monthly limit of 2,000 completions. The model will soon be available for Copilot users in all JetBrains IDEs. GitHub has also updated Issues and Projects, introducing a streamlined issue creation flow, checklist-to-sub-issue conversion, and required fields for private repositories. Previously, typing an issue title in a project defaulted to creating a draft issue. Based on user feedback, GitHub has changed this so that pressing ‘Enter’ now creates a new issue directly, while ‘Cmd \/ Ctrl + Enter’ creates a draft. “This change reflects how users typically engage with issue creation in projects,” GitHub stated. Additional updates include improvements to tasklist blocks and single-issue templates. These changes aim to improve workflow efficiency for developers managing projects on GitHub. Last year, GitHub announced that its AI-powered coding tool, GitHub Copilot, was now available for free within Visual Studio Code (VS Code). This announcement came as GitHub reached 150 million developers on its platform.","excerpt":"The model has been fine-tuned across 32 major programming languages.","categories":["AI News"],"tags":["GitHub"],"author_name":"Siddharth Jindal","publish_date":"2025-02-19T11:42:05","publication_year":"2025","word_count":337,"keywords":["programming_languages:R","AI","GPT-4o","ML","Git","GPT","Aim","GitHub","R","llm_models:GPT"],"extracted_tech_keywords":["AI","ML","GPT-4o","Aim","R","Git","GitHub","GPT","llm_models:GPT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/new-gpt-4o-copilot-model-unveiled-for-vs-code-users-in-public-preview\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":46283,"title":"Grant For The Web: How A Company Wants To Help You Challenge Tech Giants’ Control Over The Internet","content":"The most prevalent form of a web-based business relies heavily on advertising revenue. In recent times, this model of online business has led to various issues related to user privacy, where ad trackers are constantly tracking web searches. The problem is worsened with clickbait content and companies selling user data to third parties. When Facebook founder Mark Zuckerberg testified to Congress in 2018, he told lawmakers that the social network needed to be free to “connect everyone around the world”. It couldn’t be far from the truth as the platform is not free and derives value by selling user data. Unless the internet’s entire business model changes, we are doomed to this cycle of constantly trying to protect our privacy online. Clearly, the Internet needs a better model for monetisation and now one company Coil has announced Grant for the Web, a new $100 million fund to help creators and promote innovation in web monetisation. The fund hopes to find solutions to some of the most serious challenges facing the web, including user privacy exploits related to ads and unethical sponsored content. The organisation is aiming to build an ethical internet based on open standards, giving web users more freedom and control over how they create, circulate and monetise content, as well as designing better business models for the web. “We envision a world where it’s possible for creators and publishers to get paid for their work without relying on invasive ads, paywalls, and the abuse of personal data. Grant for the Web is founded on the belief that a healthy internet needs openness and opportunity, and that it cannot be built on the backs of individuals’ security and privacy,” states the website. How Will It be Done Grant for the Web is committed to awarding at least 50% of all grant dollars to proposed software projects and content projects that will be openly licensed. Special consideration will be given to projects that reflect the values of the global commons, such as increasing users’ autonomy, privacy, and control over their own data, promoting diversity and inclusion on the internet, and increasing access to the full capacity of the internet, for historically marginalised or disadvantaged communities and individuals. Representatives from Coil, Creative Commons, and Mozilla will make up the initial Grant for the Web Advisory Council, which governs the fund’s activities. Additionally, Coil, Creative Commons, and Mozilla will each designate an advisor to serve on a Technical Advisory Group. The purpose of this group will be to guide the direction of the program, and to review grant applications from a technical perspective and in turn provide recommendations to the Advisory Council. The project is being built on previous work done by Stefan Thomas, former CTO of payment protocol firm Ripple, and now CEO of Coil – which has pioneered a new internet payment protocol (and aspiring W3C standard) based around an API that allows browsers to make micropayments to content creators. Web Monetisation Is A Core Part Of Grant For Web Grant For Web promotes technologies like Web Monetization, a proposed API standard that allows websites to request a stream of very small payments from a user e.g. fractions of a cent. This provides a framework for new revenue models for websites and web-based services, as an alternative to subscription services or advertising, while preserving the user’s privacy. In exchange for payments from the user, websites can provide the user with a “premium” experience such as exclusive content or by removing advertising or even the need to login to access content or services. The system is based around two key technologies: https:\/\/interledger.org: a protocol for sending payments across multiple ledgers and https:\/\/paymentpointers.org: a way to express a URL that points to a secure payment initiation endpoint on the Web. What India Can Learn From The Project? Web monetisation and creating open standards for the internet that are focused on privacy is something that is highly relevant for countries like India. India has been one of the targets for mass breaches such as the Cambridge Analytica. In the next phase of web-based innovation, there will be a need for solutions in India that focus on user privacy. Also, it will be important for web-based business models that are inclusive to the poor. Projects such as Grant for the Web aim to include marginalised and poor groups globally that have had poor reach to web services ever since it came out. Creating platforms that don’t engage in ad targeting or overzealous data collection and that afford users autonomy and control over their data is something young Indian entrepreneurs can benefit from. By building monetization natively into the web using open protocols and standards benefits both content creators and their supporters by providing access, choice, and user control.","excerpt":"The most prevalent form of a web-based business relies heavily on advertising revenue. In recent times, this model of online business has led to various issues related to user privacy, where ad trackers are constantly tracking web searches. The problem is worsened with clickbait content and companies selling user data to third parties. When Facebook […]","categories":["AI Features"],"tags":["Mozilla","Privacy"],"author_name":"Vishal Chawla","publish_date":"2019-09-24T09:57:31","publication_year":"2019","word_count":794,"keywords":["Go","Privacy","API","Mozilla","AI","programming_languages:R","innovation","programming_languages:Go","Aim","ViT","GAN","R"],"extracted_tech_keywords":["AI","Aim","R","Go","API","GAN","ViT","innovation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/grant-for-the-web-how-a-company-wants-to-help-you-challenge-tech-giants-control-over-the-internet\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10021865,"title":"Paypal Pushes Its Cryptocurrency Ambitions With Curv Acquisition","content":"PayPal has acquired cryptocurrency security firm Curv for an undisclosed amount. Last year, the company had announced the launch of a new service for its customers to buy and sell cryptocurrencies from their PayPal accounts. Curv is an Israel-based provider of cloud-based infrastructure for digital asset security. The deal is expected to close in the first half of 2021. PayPal’s Blockchain Ambitions Earlier, PayPal had announced its plans to explore the opportunities in digital currencies. To drive this initiative, the company recently created a business unit focused on blockchain, crypto and digital currencies. For its blockchain ambitions, Paypal had partnered with the New York state-regulated Paxos to offer cryptocurrency transactions services, starting October. The company believes the shift to digital forms of currencies is inevitable and brings many advantages such as efficiency, speed and resilience in payments. Paypal is already a force to reckon with in the field of digital currencies with its digital payment expertise, two-sided network and rigorous security and compliance controls. To further increase consumer understanding and adoption of cryptocurrency, the company introduced the ability to buy, hold and sell select cryptocurrencies, initially featuring Bitcoin, Ethereum, Bitcoin Cash and Litecoin, directly within the PayPal digital wallet. To further push its growth in the sector, Paypal has been looking to acquire a crypto firm. The company was in talks with BitGo, one of the country’s oldest cryptocurrency firms, but the deal fell through. “The acquisition of Curv is part of our effort to invest in the talent and technology to realise our vision for a more inclusive financial system,” said Jose Fernandez da Ponte. He is the vice president and general manager, blockchain, crypto and digital currencies, of PayPal. PayPal is also looking to tap Curv’s technical talent, entrepreneurial spirit and the crypto technology. Why Curv? Curv, a leading supplier of cloud-based infrastructure for digital asset security, currently supports more than 200 coins and tokens on multiple blockchains. The product is regularly pen-tested and peer-reviewed by world-leading cryptographers. The reasons why Curv can help PayPal fulfil its ambitions are: Secure transactions: Curve claims to eliminate the private key flaw in the blockchain. By design, blockchain transactions are irreversible, thanks to the public key cryptography they use. This public-key cryptography introduces a single point of failure, i.e. private keys. Curv eliminates this by giving a secure, distributed way to sign transactions and manage all the digital assets. Its proprietary multi-party computation (MPC) protocols allow the transaction to be carried in a secure and distributed way to protect against cyber breaches and other physical damages. It also lets users set up sophisticated policies to stop others from withdrawing crypto assets without some approval chain. Expand footprint: Since its inception, Curv has been laser-focused on spreading its crypto footprint and offering a broader range of services. It had even partnered with some notable Europe-based crypto firms such as eToro and FalconX to expand its presence. This partnership will help PayPal to utilise these connections and expand its offerings outside the US. Talent:  Curv will bring its strong team of technologists to PayPal and start working on the crypto technologies to help their users to store crypto assets securely and access their wallets without additional hardware. While the cryptocurrency will be available only in the US for now, Paypal also plans to launch new crypto-related products in other countries. Talking about the deal with PayPal, Itay Malinger, CEO Curv, said, “as the adoption of digital assets accelerates, we feel there’s no better home than PayPal to continue our journey of innovation. We’re excited to join PayPal in expanding the role these assets play in the global economy.” The demand for Curv’s technology has been high, with the likes of Facebook’s crypto arm, Novi, looking to acquire the startup. However, PayPal closed the deal, which speaks volumes about its focus on developing blockchain technology.","excerpt":"PayPal has acquired cryptocurrency security firm Curv for an undisclosed amount. Last year, the company had announced the launch of a new service for its customers to buy and sell cryptocurrencies from their PayPal accounts. Curv is an Israel-based provider of cloud-based infrastructure for digital asset security. The deal is expected to close in the […]","categories":["IT Services"],"tags":["PayPal"],"author_name":"Srishti Deoras","publish_date":"2021-03-11T11:00:00","publication_year":"2021","word_count":641,"keywords":["Go","programming_languages:R","AI","PayPal","innovation","programming_languages:Go","Git","Aim","ViT","R","startup"],"extracted_tech_keywords":["AI","Aim","R","Go","Git","ViT","innovation","startup","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/paypal-pushes-it-cryptocurrency-ambitions-with-curv-acquisition\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10163803,"title":"xAI’s Grok 3 to be Released Today","content":"Elon Musk, CEO of xAI, revealed that the company will release its latest AI model, Grok 3, on Monday. Musk calls it the smartest AI on earth. Users across social media are anticipating Grok 3’s capabilities, given that it has been trained on 100,000 NVIDIA GPUs, ten times more computing power than its predecessor. In addition to the release of Grok 3, xAI is reportedly in talks to raise $10 billion in funding. Bloomberg reported this on Saturday and said Sequoia Capital, Andreessen Horowitz, and Valor Equity Partners will participate in the round. This will increase xAI’s valuation to $75 billion. In December last year, the company completed a $6 billion Series C funding round, with participation from firms such as Andreessen Horowitz, BlackRock, Sequoia Capital, NVIDIA, and AMD. xAI is currently valued at $51 billion. Furthermore, Bloomberg reported on Friday that xAI is in the ‘advanced stages’ of securing a deal worth over $5 billion with Dell Technologies for servers. Dell will reportedly sell xAI servers containing NVIDIA GB200 this year to handle AI workloads. Moreover, xAI’s GPU cluster Colossus is claimed by the company to be the world’s most powerful AI training system. Musk notes that this cluster was built in 122 days from start to finish. According to an earlier report from Financial Times, xAI plans to expand Colossus by over ten times to include more than 1 million GPUs. However, given how DeepSeek disrupted the market by building an AI model with significantly fewer GPUs and beating most of the competition, NVIDIA lost over $500 billion in its market cap as shareholders questioned the demand for AI hardware. Grok 3 might answer the question of how much an AI model can improve by incorporating exponentially higher amounts of computing power, contributing new evidence to the ongoing debate about AI scaling laws. When AIM reached out to Paras Chopra, founder of Lossfunk, he said, “Performance is often log-linear. So I’d say 10x more compute would have a ~double jump in performance over Grok 2.” Currently, Grok AI is available for free as a chatbot on X.com. It will have to compete not only with the existing models but also with the upcoming hybrid model by Anthropic and GPT 4.5, which is part of OpenAI’s newly released roadmap.","excerpt":"Besides. xAI is in talks for a $10 billion funding, and a $5 billon plus deal to purchase servers from Dell.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","Elon Musk","grok","XAI"],"author_name":"Supreeth Koundinya","publish_date":"2025-02-17T10:59:50","publication_year":"2025","word_count":380,"keywords":["Anthropic","Grok 3","Go","API","OpenAI","AI","AWS","R","Elon Musk","XAI","Aim","grok","xAI","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","OpenAI","Anthropic","Grok 3","xAI","Aim","AWS","R","Go","API"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/xais-grok-3-to-be-released-today\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10044063,"title":"Facebook Introduces Few-Shot Neural Architecture Search","content":"Facebook recently introduced few-shot neural architecture search (NAS). The new approach combines the accurate network ranking of vanilla NAS with the speed and minimal computing cost of one-shot NAS. Few-shot NAS enables users to quickly design a powerful customised model for their tasks using just a few GPUs. Few-shot NAS can effectively design numerous SOTA models, from convolutional neural networks (CNNs) for image recognition to generative adversarial networks (GANs) for image generation. The source code of Latent Action Monte Carlo Tree Search (LA-MCTS), alongside its application to NAS, is available on GitHub. The paper has also been selected at ICML 2021 for a poster presentation. Why few-shot NAS? Of late, NAS has become an exciting area in deep learning research, offering promising results in computer vision, especially when specialised models need to be found under different resources and platform constraints (for example, on-device models in virtual reality (VR) headsets). For years, researchers have used vanilla NAS. It utilises search techniques to accurately explore the search space and evaluate new architectures by training them from scratch. However, this approach requires thousands of GPU hours, leading to high computing costs. Meanwhile, one-shot NAS significantly lowers the computing cost by using supernet, an extensive network whose edge contains every type of edge connection. Once pre-trained, a supernet can approximate the accuracy of neural architectures in the search space without having to be trained from scratch. While one-shot NAS reduces GPU requirements, its search can be hampered by inaccurate predictions from the supernet, making it hard to identify suitable architectures. Facebook researchers Yiyang Zhao, Linnan Wang, Yuandong Tian, Rodrigo Fonseca and Tian Guo came up with few-shot NAS to address the issue. Compared to one-shot NAS, few-shot NAS improves performance estimation by first partitioning the search space into different independent areas and later employing multiple sub-supernets to cover these regions. “To partition to search space in a meaningful way, we choose to leverage the structure of the original supernet. By picking each type of edge connection individually, we choose a way to split the search space that is consistent with how the supernet is constructed,” said researchers. In other words, few-shot NAS is a trade-off between the accuracy of vanilla NAS and the low computing cost of one-shot NAS. (Source: arXiv) Here’s how few-shot NAS works According to Facebook, the innovation offered by few-shot NAS arises from the observation that a supernet can be regarded as a representation of search space, and that we can enumerate every neural architecture by recursively splitting each supernet’s compound edges. (Source: arXiv) The researchers tested on multiple supernets, checking if it could offer the best aspects of both one-shot NAS and vanilla NAS. “To investigate this idea, we designed a search space containing 1,296 networks. First, we trained the networks in order to rank them according to their true accuracies on the CIFAR10 dataset. We then predicted the 1,296 networks using 6, 36, and 216 sub-supernets and compared the predicted ranking with the true ranking,” wrote Tian, in a blog post. Interestingly, the researchers found the ranking improved substantially even when adding just a few sub-supernets. With multiple sub-supernets, the predicted accuracy matched well with the ground truth accuracy and the ranking prediction also improved, along with the final search performance. As depicted in the image below: (Source: arXiv) Outcome The researchers also tested their idea on real-world tasks and found that, compared to one-shot NAS, few-shot NAS improved the accuracy of architecture evaluation with a slight increase in evaluation costs. For example, for seven sub-supernets it used in the experiment, few-shot NAS established new SOTA: On ImageNet, the researchers found that the models that reach 80.5 percent top-1 accuracy at 600 MFLOPS (million floating-point operations per second) and 77.5 percent top-1 accuracy at 238 MFLOPS On CIFAR10, it reached 98.72 percent top-1 accuracy without using extra data or transfer learning.In AutoGAN, few-shot NAS outperformed the previously published results by up to 20 percent. Extensive experiments showed that few-shot NAS significantly improved various one-shot methods, including four gradient-based and six search-based strategies on three tasks in NasBench-201 and NasBench1-shot-1. “Overall, our work demonstrates that few-shot NAS is a simple yet highly effective advance over the ability of one-shot NAS to improve ranking prediction. It is also widely applicable to all existing NAS methods,” shared Facebook researchers. Facebook believes the latest technique will help researchers develop broad applications, particularly when a candidate architecture needs to be evaluated quickly in search of better architectures.","excerpt":"Few-shot NAS enables users to quickly design a powerful customised model for their tasks using just a few GPUs.","categories":["AI Features"],"tags":["Facebook AI research","Facebook latest","Machine Learning","Machine Learning Latest","neural architecture search"],"author_name":"Amit Naik","publish_date":"2021-07-21T19:00:00","publication_year":"2021","word_count":744,"keywords":["Go","AI","neural network","ML","image recognition","Machine Learning","Machine Learning Latest","computer vision","RAG","Git","neural architecture search","deep learning","Facebook AI research","R","Facebook latest"],"extracted_tech_keywords":["AI","ML","deep learning","neural network","computer vision","RAG","image recognition","R","Go","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/facebook-introduces-few-shot-neural-architecture-search\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10088896,"title":"Stability AI Makes its First Acquisition in a Generative AI Startup, Init ML","content":"Stability AI has acquired Init ML, the developer of the acclaimed AI-powered imaging tool Clipdrop for an undisclosed amount. As per the deal, Stability AI will integrate its generative AI into the Clipdrop platform to make it accessible to creators, which already boasts 15 million users. Announcement: Stability AI has acquired the industry leader in AI-powered imaging tools @Clipdropapp! We’re excited to integrate our generative AI into the Clipdrop platform, enabling our technology to be easily accessible by creators everywhere! https:\/\/t.co\/cOjnQcKOZd— Stability AI (@StabilityAI) March 7, 2023 Init ML will now function as a wholly-owned, autonomous subsidiary of Stability AI, with all employees remaining in their current roles. Cyril Diagne, CEO of Init ML, will take up the role of Vice President of Innovation, Clipdrop, while COO Damien Henry will be the Senior Vice President of Product, Clipdrop, and CTO Jonathan Blanchet will become the Vice President of Engineering, Clipdrop. The acquisition provides Clipdrop with Stability AI’s technology and computing power to create the next generation of imaging tools. The deal marks a new milestone in both firms’ journey to democratise the power of AI and multimodal foundation models in creative workflows, according to Emad Mostaque, founder and CEO of Stability AI. Init ML, which is based in Paris, France, was founded in July 2020 and received a seed investment from AI venture capital firm ‘Air Street Capital’. Its flagship suite of imaging applications, Clipdrop, has over 15 million users today. Currently, advanced machine learning research, such as Remove Background and Super Resolution, is included in Clipdrop and Stability AI’s powerful generative AI models are being incorporated into its beta applications. The strategic acquisition comes in the backdrop of Stability AI’s crown jewel Stable Diffusion, an open-source text-to-image tool. The tool caught the public eye and was lauded by AI stalwarts as it is free and has no content filters. Former Tesla Director of AI, who recently returned to OpenAI, Andrej Karpathy praised Stable Diffusion and claimed that that tool marks a day of historic proportion for human creativity. The code and model card of Stable Diffusion is available on GitHub and HuggingFace.","excerpt":"With this acquisition, Stability AI looks to integrate its generative AI (Stable Diffusion) into the Clipdrop platform to make it accessible to creators.","categories":["Deep Tech"],"tags":["Generative AI","Stability AI","Stable Diffusion","Startups"],"author_name":"Tasmia Ansari","publish_date":"2023-03-08T17:13:33","publication_year":"2023","word_count":353,"keywords":["machine learning","OpenAI","AI","Stable Diffusion","ML","Git","Aim","Startups","generative AI","Generative AI","GitHub","R","foundation models","Stability AI"],"extracted_tech_keywords":["AI","machine learning","ML","generative AI","foundation models","OpenAI","Aim","R","Git","GitHub"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/stability-ai-makes-its-first-acquisition-in-a-generative-init-ml\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":54907,"title":"Temenos &#038; Google Cloud Announce Global Strategic Partnership To Accelerate Banks’ Digital Transformation In The Cloud","content":"Temenos, the banking software company and Google Cloud has announced its global and strategic partnership to help financial service organisations to run mission-critical banking software and applications on Google Cloud. This will help them to create profitable business models and improve and differentiate their customer experiences. The two companies began collaborating in 2019 to integrate Temenos cloud-native, cloud-agnostic software into Google Cloud and have demonstrated success with joint banking customers in Europe and Asia. Under this expanded partnership, Temenos’ full suite of banking software and applications will now be available on Google Cloud, and the partnership will help banks go to market faster, open up new business models, and achieve tangible business benefits. More than 3,000 financial service institutions around the world have leveraged Temenos’ modern, cloud-native and API-first technology. As a part of this strategic partnership, banks can run Temenos’ applications in Google Cloud, taking advantage of its scalability, resilience, and global infrastructure. The expanded partnership was announced at the Paris Fintech Forum 2020, where Max Chuard, Temenos CEO, spoke about ‘Banking in the Cloud’. Besides, building upon its open approach to cloud, Temenos will be the first global banking software provider to run on Google Cloud’s Anthos, enabling customers to deliver mission-critical workloads across on-premises and cloud environments, or even across multiple clouds. Anthos is Google Cloud’s hybrid and multi-cloud application platform that enables organisations to migrate and modernise their existing applications, build new ones and run them anywhere, whether it’s in the cloud, on-premises or on multiple clouds. Built on open-source technology pioneered by Google, such as Kubernetes, Istio and containers, it helps banks empower their developers, security professionals, platform teams and operations engineers — enhancing employee and customer experiences, saving bottom-line costs and increasing top-line revenue. According to Chuard, Chief Executive Officer, at Temenos, “We see explosive growth in cloud adoption in the banking industry. We are delighted to extend our leadership in the cloud and strengthen our strategic partnership with Google Cloud.” He further added, “As a strategic global banking software partner of Google Cloud, we will bring to innovative market solutions that combine our API-first, cloud-native, and microservices-based banking software with Google Cloud’s scale and its leading cloud capabilities. Our customers will benefit from the winning combination of Temenos’ functionally rich and technologically advanced digital banking platform on Google Cloud. Together, we will enable banks to reduce their time to market and operational complexity, as well as elastically scale and deliver outstanding digital customer experiences. With Google Cloud’s Anthos, we give banks the freedom to innovate and reap the true benefits of multi-cloud.” Thomas Kurian, Chief Executive Officer, at Google Cloud, also said, “Temenos delivers market-leading banking software applications to thousands of the world’s leading financial services organisations, and we’re excited to expand our strategic alliance with them. Extending Temenos’ core capabilities to Google Cloud will help banks digitally transform their business and deliver entirely new customer experiences and offerings.” Temenos has been at the forefront of software innovation, consistently investing 20% of its revenues in R&D and is pioneering in cloud banking for the last ten years. Temenos will enable banks to significantly reduce their total cost of ownership through elastic cloud scalability, distributed database technology and multi-cloud resilience, all underpinned by the benefits of vendor and platform independence. Temenos accelerates banks digital transformation helping them to become more agile and innovate faster.","excerpt":"Temenos, the banking software company and Google Cloud has announced its global and strategic partnership to help financial service organisations to run mission-critical banking software and applications on Google Cloud. This will help them to create profitable business models and improve and differentiate their customer experiences. The two companies began collaborating in 2019 to integrate […]","categories":["AI News"],"tags":["Banking","banking AI","banking analytics","Cloud Platform","Google Cloud"],"author_name":"Sejuti Das","publish_date":"2020-01-30T16:04:02","publication_year":"2020","word_count":560,"keywords":["banking AI","Go","API","Google Cloud","AI","banking analytics","Scala","kubernetes","RAG","microservices","Git","Banking","GAN","R","Cloud Platform"],"extracted_tech_keywords":["AI","RAG","kubernetes","microservices","R","Go","Scala","Git","API","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/temenos-google-cloud-announce-global-strategic-partnership-to-accelerate-banks-digital-transformation-in-the-cloud\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":26986,"title":"Book Review: ‘Data Analytics Using R’ By Seema Acharya","content":"R is without a doubt one of the most sought-after software tools for today’s data scientist. It is very flexible, powerful and enables the user to access many algorithms and statistical tools and techniques with ease. It also works well with many large data sets. The book by Seema Acharya, Data Analytics Using R (McGraw Hill Education, 2018) is a timely book for R practitioners. Seema Acharya is a senior lead principal at Infosys and her experience clearly reflects in the large textbook for R. The book consists of over 300 self-assessment progress questions, more than 10 practical hands-on experience and 200+ multiple choice questions for any practitioner to test his\/her knowledge. The best feature in the book is definitely the fact that it starts from the basics. Basic syntax and algorithms are covered totally. The book also introduces many R interfaces with various kinds of data sources such as csv, JSON, XML and RDBMS and others. The book is laid out and structured in such a manner that it will be useful for many levels and kinds of R practitioners. The book can be used as a guide by aspiring data scientists, data analysts or executives. It can be used by a professional at the executive and management level to aid his\/her decision making. The book is neatly organised in 12 chapters — starting from the introductory chapter instructing users how to install and use R packages. From the first chapter itself, the book is designed to get the reader comfortable with using and programming with R. R is a tricky subject to write about because of its extensive ecosystem of packages and tools. Acharya still makes a great attempt at building a complete information set. Following chapters deal with introducing the readers to many basic commands which can be used by practitioners to analyse content or datasets. This chapter is followed by instructions that introduce the reader the inner working of R when used to process and load data from many data sources. One of the highlights of the book is that this book teaches the reader to work with many databases such as MYSQL, SQLite, PostgreSQL and others. Along with this book, the readers can use additional resources such as Question Bank and Weblinks for reference material. It is advised that the reader read through all the chapters and make us of all the hands-on material to practice programming. The sequence of the chapter is well thought-out and suits very well for a person who is just starting a career in data analytics and R. The fourth chapter of the textbook teaches descriptive statistics with commands such as dim(), summary(), etc. As the book advances it slowly increases the standard of subjects handed. The next chapters handle various details of regression analysis, logistic regression and various kinds of logistic regressions. The book also explains important terms like Residual, Goodness-of-Fit tests carefully. The rest of the book handles tricky advanced topics such as handling and processing of time series data and clustering. The breadth of algorithms which are handled in the section of time series and clustering sections is very impressive with the reader being introduced to a wide range of possible approaches. After this, the book eases the reader into advanced data mining techniques like association rules and text mining. The book also delves into topics like parallel computing with an introduction to MapReduce algorithms and distributed computing. All in all, Acharya’s Data Analytics Using R is a near-complete book which is a rare combination of good algorithm, tools education and a crisp explanation of the R language. The book comes at a great time when there is a great demand for R in the market and usage of the language is only increasing.","excerpt":"R is without a doubt one of the most sought-after software tools for today’s data scientist. It is very flexible, powerful and enables the user to access many algorithms and statistical tools and techniques with ease. It also works well with many large data sets. The book by Seema Acharya, Data Analytics Using R (McGraw […]","categories":["AI Features"],"tags":["Data Analytics","Programming Languages","R","Software","statistical analysis using sql"],"author_name":"Abhijeet Katte","publish_date":"2018-08-07T11:04:00","publication_year":"2018","word_count":627,"keywords":["PostgreSQL","Go","statistical analysis using sql","programming_languages:R","AI","ML","distributed computing","Programming Languages","GAN","analytics","SQL","Data Analytics","Software","R"],"extracted_tech_keywords":["AI","ML","analytics","distributed computing","PostgreSQL","R","SQL","Go","GAN","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/book-review-data-analytics-using-r-by-seema-acharya\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":57634,"title":"Top 7 Expert Advice From Women Data Scientists","content":"‘The Rising’, an event hosted by Analytics India Magazine, has been the biggest conference for women data science leaders and women professionals from across the domain. With the second edition coming up in Bengaluru — Hotel Radisson Blu on March 20th, 2020, the event is aimed towards promoting and celebrating women innovators in data science. The conference is also going to highlight the career achievements of women data science leaders and executives. In this article, we are going to list down, in no particular order, top 7 expert tips from women data scientists from The Rising 2019. As a data scientist, one needs to have a lot of patience – Mathangi Sri, Head of Data at Gojek. In this talk, Mathangi Sri, head of data at Gojek, spoke about her 15 years of journey in the field of data science and the learning acquired. She also discussed her experiences working across multiple problems and various types of data and the challenges she witnessed over the years. One crucial advice she mentioned in the talk was to have enough patience. She said how it is easy to build a model sitting behind the desk, but it takes an ample amount of time to make it to a production level from an engineering standpoint. So, she advises young data scientists to have a lot of patience. She said, “Data scientists would need a lot of patience to get the data, as well as, would need a lot of patience to wait for these models to go live.” Reskilling is of utmost importance in the digital era – Sohini Mehta, Global Service Delivery Head of Analytics at Wipro Sohini Mehta, the global service delivery head of analytics at Wipro, in the talk, spoke about the importance of reskilling in this digital era. Considering the dynamic landscape of technology, which is evolving every day, it has become imperative for organisations to provide reskilling\/upskilling opportunities for employees to improve the talent workforce. Reskilling will not only bridge the talent gap but will also offer opportunities for employees to learn and be relevant in the world of data and analytics. She strongly believes in reskilling over hiring a new individual, where the existing employees are taught newer skills. In this process, organisations are not only building a talent pipeline but also improving employee retention to a great extent. To remove biases in machine learning, humans need to increase diversity – Smitha Ganesh, Principal Data Scientist at ThoughtWorks Smitha Ganesh is the principal consultant and data scientist at Thoughtworks, and in this talk, she spoke about how human biases impact AI technology. She believes that biases and prejudices are common elements wherever there is a system that encompasses humans, which is something very inherent. But when this inherence gets into the building an algorithm, these effects get compounded with time. And that’s why it is crucial to fight prejudice in AI. To remove these biases, according to Ganesh, humans need to incorporate a diverse mindset for data collection, which will help in building and taking AI to an elevated level and recognising the vitality involved with AI training systems. Drive towards a cleaner future through data – Deepika Sandeep, Practice Head – AI & ML at Bharat Light & Power Deepika Sandeep is the head of AI and ML at BLP Clean Energy, and in this, thought-provoking talk she spoke about how organisations can use newer technologies like AI, ML and augmented analytics for creating a better future. She suggested how AI and ML can benefit industries creating renewable energy, reducing downtime in the utility sector, managing energy, bringing about automation, and optimising inventory. She strongly believes in using AI and ML to drive a push towards a cleaner future with the help of data. We need more women in AI – Vaishali Kasturae, India head of ISV Business Segment at AWS Vaishali Kasturae, the India head of ISV Business Segment of AWS, discussed the urgent need for more women in the world of AI. This need is not only to empower more women in AI but actually to make AI safer and bring in a diverse perspective to make sure that people are using AI the way they should. She mentioned that organisations must ensure that their AI systems do not discriminate inappropriately against any individual or group. However, if the technology is always built by only certain groups of people, then the chances of involving discrimination will increase, even if it is not intentional. Kasture believes that the best way for combating sexism is to have a diverse data science team. She said, “If we need to make AI safe in the future it is important to have a blue tick mark for approved AI models, as we have on our social media for verified accounts, only this time women should give the verification.” AI should be used for noble causes, which can help in saving lives – Geetha Manjunath, Co-founder and CEO of NIRAMAI Dr Geetha Manjunath is the founder, CEO and CTO of NIRAMAI, and in this talk, she spoke about the noble solution developed by NIRAMAI to detect early-stage breast cancer in a non-invasive manner. The talk focuses on how AI can be used for such a noble cause which is helping several women of the country to have early detection of breast cancer, which, in turn, will save their lives. In a country like India, where there is a massive increase in the incidences of breast cancer, Niramai is using artificial intelligence, machine learning algorithms, big data analytics to improve breast cancer diagnosis. And similar to the name NIRAMAI, which means ‘being healthy’ Geetha Manjunath also strongly believes in using emerging technology in making everyone healthy in our country. Organisations should enhance employee satisfaction using AI – Dr Shivani Rai Gupta, head of AI & data science at Capgemini Dr Shivani Rai Gupta is the head of AI and data science in Capgemini, and in this talk, she spoke about how organisations are focused about creating newer technology and products for their customers but rarely utilises the same for insiders to engage employees and increase their satisfaction level. She firmly believes that if employees are engaged then the overall productivity of the business increases. She mentioned that if the technical difficulties and operational challenges are reduced for employees, their satisfaction level increases. According to her, the four ways AI can help employee satisfaction are — introducing AI chatbots to improve onboarding experience; applying AI to capture employees’ daily experiences in real time; introducing AI to enhance team collaboration; using AI to enable timely learning and development to keep the employees engaged.","excerpt":"‘The Rising’, an event hosted by Analytics India Magazine, has been the biggest conference for women data science leaders and women professionals from across the domain. With the second edition coming up in Bengaluru — Hotel Radisson Blu on March 20th, 2020, the event is aimed towards promoting and celebrating women innovators in data science. […]","categories":["AI Trends"],"tags":["data science women"],"author_name":"Sejuti Das","publish_date":"2020-02-27T17:00:00","publication_year":"2020","word_count":1110,"keywords":["data science","machine learning","artificial intelligence","AWS","AI","chatbots","ML","data science women","Aim","analytics","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","data science","analytics","Aim","chatbots","AWS","R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-7-expert-advice-from-women-data-scientists\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":16542,"title":"Why is Kaggle’s upcoming challenge receiving so much flak?","content":"Kaggle’s latest competition leads to a wave of controversies and dissatisfaction Google made news with the recent acquisition of the online platform Kaggle, which represents “the world’s largest community of data scientists and machine learning enthusiasts.” Kaggle’s community includes researchers in at least 194 countries. By acquiring Kaggle, Google Cloud obtained more direct access to the community’s one million members, who compete and earn money for developing data science solutions which help solve data analysis problems. In the past, this approach has helped several organizations, for instance, improving the algorithms used by the online real estate giant Zillow, and helping a satellite company use data to “track the human footprint in the rainforest.” Today Kaggle has over one million users, almost 60,000 active competitors, and countless other people. They know about Kaggle and follow competition outcomes. Basically, a completion held on Kaggle usually garners a lot of attention. Google Kaggle’s airport competition is mired in controversy Daimler had challenged Kagglers to tackle the curse of dimensionality and reduce the time that cars spend on the test bench The Google-Kaggle merger received mixed reviews, However, that doesn’t change the fact that the community’s international membership is in revolt. This was a result of Kaggle’s recent decision to host a U.S. government competition, one which is legally off limits to foreigners. The competition is supposed to be worth $1.5 million in prize money. The competition, named Passenger Screening Algorithm Challenge invites Kagglers to analyze a large dataset of body-scan images collected from volunteers passing through airport security scanners. U.S. Department of Homeland Security (DHS) is sponsoring this event, and hopes that in return Kagglers will submit machine-learning algorithms. This will in turn improve the accuracy of U.S. airport scanners in detecting possible threats. The controversy surfaced from the fact that Federal legislation known as the America Competes Act bars the government, among other things, from awarding federal prize money like what DHS is offering to anyone who is not an American citizen or permanent resident. This act was enacted in 2007 and reauthorized under Barack Obama in 2011. This isn’t a new situation for Kaggle, as the 1.2-million Zillow Prize launched in May, initially prohibited Chinese Kagglers from participating in the second round of the competition. Zillow was concerned about acquiring intellectual property rights in China. However, that restriction was eventually lifted after a chorus of Kagglers voiced their disapproval. Why is Kaggle’s competition receiving so much backlash? Vladimir Iglovikov, Senior Data Scientist at TrueAccord Many Kagglers have already boycotted the competition. The exclusive DHS competition has particularly created a backlash among platform users who argue that this decision runs against Kaggle’s culture of open data and international cooperation. Many feel it’s unfair to leverage the one million users and 60,000 active competitors for positive attention, while stripping most of them from the opportunity to win prizes. Moreover, some international Kaggle members have raised concerns about U.S. participants discriminating against potential foreign teammates because of the prize restrictions. Others have also brought to light the outrage at being invited to make contributions to American national security without having a fair shot at the prize money. “Hosting open competitions in which people are differentiated by anything except their skill is unacceptable and it is definitely against the spirit of the community,” comments Vladimir Iglovikov, Senior Data Scientist, TrueAccord. This outrage and dissatisfaction can be illustrated by the fact that the main prize announcement on the official Kaggle forum has received more than 150 downvotes, while a forum posting titled “This is insane discrimination and [an] insult to our international community” has received more than 400 upvotes.” Last Words Kaggle’s Co-founder and CEO Anthony Goldbloom Kaggle could have put more thought into it and minimized the damage. They could have made the challenge a “master-only” private competition. Such competitions have been held in the past, where participants are restricted based on Kaggle rank. Another approach would involve relying on a different competition platform for hosting the competition, where the focus would lie on soliciting contributions from U.S. participants. Moreover, even the Kaggle’s chief executive Anthony Goldbloom acknowledged the unfairness of what he described as a “two-speed competition,” where international Kagglers are allowed to only compete for ranking points on the platform, while U.S. Kagglers can compete for both points and prize money. Kagglers like Iglovikov feel that they can’t really blame defense-related government agencies such as the DHS for having their hands tied in allocating prize money. However, few of the federal agencies are tackling this situation by providing honorable mentions, prize celebration events, or “introductions to the broader business community.” But, Kaggler’s are disappointed as this competition defies the platform’s broader mission of crowdsourcing important data-science solutions to transcend national borders. In fact, some Kagglers are of the opinion that the online community they have built together may now be under threat. They are also worried about Google’s future moves when it comes to managing the Kaggle community.","excerpt":"Google made news with the recent acquisition of the online platform Kaggle, which represents “the world’s largest community of data scientists and machine learning enthusiasts.” Kaggle’s community includes researchers in at least 194 countries. By acquiring Kaggle, Google Cloud obtained more direct access to the community’s one million members, who compete and earn money for […]","categories":["IT Services"],"tags":["AI India","data analysis India","data science india","Data scientists India","Google","Machine Learning India"],"author_name":"Amit Paul Chowdhury","publish_date":"2017-07-26T11:23:41","publication_year":"2017","word_count":825,"keywords":["data science","Go","machine learning","data science india","AI","ML","Data scientists India","Machine Learning India","data analysis India","RAG","GAN","Aim","ViT","Google","AI India","R"],"extracted_tech_keywords":["AI","machine learning","ML","data science","Aim","RAG","R","Go","GAN","ViT"],"url":"https:\/\/analyticsindiamag.com\/it-services\/kaggles-upcoming-challenge-receiving-much-flak\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10058877,"title":"How digital twins expand the scope of deep learning applications","content":"Ajinkya Bhave, Country Head (India) at Siemens Engineering Services, discussed the rising significance of simulated data, in his talk at the MLDS conference, titled “Simulation-driven Machine Learning”. He discussed the application of simulated data to train machine learning models in situations impossible with physical data. “At Siemens, the tool connects simulation models and data to train frameworks for ML to train the model at scale using the digital twin,” he explained. He outlined the challenge of the generation and labelling of real-world data and how industries can overcome the hurdles using a digital twin and simulation data. He referred to the Reduced Order Model (ROM), which simplifies a high-fidelity static or dynamical model, preserving essential behaviour and dominant effects to reduce solution time or storage capacity required for the more complex model. ROM, simulation and digital twins The reduced-order model helps organisations convert data to models, extend their scope and compute faster. ROM can run your digital twin on embedded devices, cloud and on-site. “The basic idea is that the ROM is the catalyst of the digital twin, enabling more applications that weren’t possible in the past,” he explained. There are multiple ways to create a ROM, depending on your application area, data, and the model’s system. The model can be anywhere from data-driven with machine learning and deep learning, hybrid with statistical models and physics, to a complete physics-based model. “You cannot create a model without domain knowledge that you encapsulate in it. But, equally important, the data matters. All the models require some amount of data,” he said. To create ROMs based on a neural network approach, the data can become either a stopping point or an advantage. At Siemens, the team either augments existing physical data, creates synthetic data or cleans\/ labels existing data. Simulation plays a huge role in connecting machine learning to the digital twin model. Ajinkya explored this ability through interesting real-world case studies. Case study 1: Applying synthetic data to deploy the machine in real-world scenarios Ajinkya walked the audience through a case study of a Siemen’s client that creates gearboxes for wind turbines. The wind turbines break down due to failures in gearboxes and ball bearings. The company turned to predictive monitoring to minimise the downtime. While the customer had tons of data, they did not have the domain distribution needed, making most of the data good with only one-off events of fault anomalies. To balance the distribution, Siemens leveraged 1D and 3D tools to model the gearbox and the ball bearings around the gears in the company’s multiphysics tool for 1D modelling. The model and its parts were simulated through a nonlinear spring mass damper system with both parameters based on real data and others tuned. Then, fault injections were applied on the model with the faults the customer was looking for, that output a synthetic time series. Next, statistical noise injection was done to ensure the output was closer to real-time. Siemens combined noise to create a time-series analyses and ran it through a neural network to identify faults. “The idea was that we created synthetic training data, which was then used to train a neural network on a digital twin of the model. Then we tested that on the real faults which occur in the ball bearings of the gearboxes with the physical data. The graph showed us the prediction was pretty accurate,” he said. “The model trained on synthetic data with a well-tuned simulation model was able to create good training data for the machine learning algorithm for it to be able to predict those faults in real-world data in a real-world deployment.” Case study 2: Model predictive control MPC’s algorithms need to be accurate and high fidelity plant models, but that is not always possible. To that end, a virtual model of the plant is created through the black or grey-box model approach. The model either completes the system as a plant or as a sub-system in the form of a virtual sensor for the parts of the plant that are not measurable. The neural network-based sensor infers from the physical measurements and a model for the subsystem states the controller needs and are later given to the MPC. “You have augmented the physical plant along with the unobservable data using a simulation-based approach to help the controller to do better than what it would have with only the physical model of the plant,” he said. The ROM and synthetic data can be additionally applied to the neural network of the plant in the MPC for model-based reinforcement learning, autonomous driving and factory robots for a fast but reduced-order model of the plant for the controller to optimise. Case study 3: Predictive maintenance of pole-mounted transformers The last case study was about the pole-mounted transformers that take high tension wires and reduce the voltage to 230 V for the safe operation of house appliances. However, given India’s diverse temperature conditions, such transformers are a fire risk. An identified cause is the oil levels between the coils of the transformers going down, causing it to overheat or spark. To monitor the oil levels of different transformers, the ‘normalised twin concept’ is used. Siemens retrofitted the transformers infrastructure with a Siemens box containing four temperature senses and a cloud-based router to send the measurements periodically to the cloud. This allowed Siemens to infer the oil level, specialise the normalised digital twin for that model and use the live twin to virtually estimate the oil labels. Although this is still an ongoing project, “using a digital twin with simulated data was parameterised and finetuned with real parameters from the field”. Lastly, Ajinkya discussed a generative design case study focusing on CFD simulations. ML can be used to adaptively learn the success certainty of simulation runs and reduce the hours of the process to mere minutes.","excerpt":"Ajinkya Bhave, Country Head (India) at Siemens Engineering Services, discussed the rising significance of simulated data, in his talk at the MLDS conference, titled “Simulation-driven Machine Learning”. He discussed the application of simulated data to train machine learning models in situations impossible with physical data. “At Siemens, the tool connects simulation models and data to […]","categories":["IT Services"],"tags":[],"author_name":"Avi Gopani","publish_date":"2022-01-21T15:00:00","publication_year":"2022","word_count":975,"keywords":["Go","machine learning","TPU","AI","neural network","ML","Transformers","RAG","deep learning","R"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","neural network","Transformers","RAG","TPU","R","Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/how-digital-twins-expand-the-scope-of-deep-learning-applications\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":36491,"title":"4 Latest Data Science Jobs In Hyderabad","content":"Source: Startup Hyderabad Analytics India Jobs offers a new platform for the data science community in India by bringing to the fore latest job openings across the country. Here are few Data Science job openings in Hyderabad data science enthusiasts can apply to: Deep Learning and Computer Vision Expert @ Pratibha Analytics The candidate will be part of a team that consists of highly qualified and dedicated professionals who are focussed on value creation for clients through automation, data and decision sciences. The main role of the team is to apply leading Advanced Analytics, Robotics Process Automation, Machine Learning and other AI frameworks, tools and techniques to solve client issues and help them gain powerful new insights to improve revenues, profitability and reduce costs by solving for a range of KPIs and value drivers based on the client industry and functional domain area. Requirements A degree in Msc. Stats, Phd in Statistics, Mathematics, Engg. in Computer Science, MCA, Masters in Computer Science or other business and engineering graduate streams with experience in Data Science and Advanced Analytics implementations. 2-8 years experience, with strong knowledge and skills in statistics, mathematics and computer science. Deep skills in Computer Vision solutions using CNN, RNN and other ensemble algorithms and methods. Experience in design and delivery of leading Analytics \/ AI solutions to clients by applying your deep skills in above mentioned disciplines including Advanced Analytics, Machine Learning and Deep Learning. Python, Tensorflow, Keras, Pytorch, Caffe, R, along with a strong understanding of all key market leading Advanced Analytics, Machine Learning and Deep Learning techniques, frameworks, methodologies and tools. Apply here Principal Machine Learning Engineer @ Leben Care Technologies Founded in late 2016, Leben Care offers automated medical image analysis algorithms that improve access and quality of diagnosis across areas of life sciences. Their mission is to develop artificial intelligence products and solutions that improve access and quality of diagnosis across areas of life sciences. Principal ML engineer would be part of  a world-class AI RnD team to enhance the Netra.AI platform enabling next generation of medical imaging diagnostic for Ophthalmology. Requirements Master’s Degree or PhD – Computer Science, Artificial Intelligence. Hands on Experience in implementing deep learning architectures using latest machine learning tech and frameworks. Existing track record as a researcher in machine learning with published scientific journal. Design and build novel machine learning models to solve unique medical problems and improve patient outcomes. Test and evaluate algorithms on large medical datasets to prove robustness Deliver high quality and production-ready code Solid Python and C++ experience and Linux user. Solid mathematical background. Experience with a vast set of computer vision libraries. Deep understanding and hands-on experience in state-of-the-art Medical Image Analysis algorithms will be a plus. Apply here Analytics & Category Manager @ HealthFolks Headquartered in Hyderabad, HealthFolks guarantees to deliver original medicines and healthcare products on a single click, 365 days. It is a part of the prestigious 44+ years old ‘MNR Educational Trust’ based out of Hyderabad, which has over 48 educational institutions at Hyderabad, Mumbai, UAE The candidate shall be responsible for managing the products categories on our Portal on the basis of analysing the market trends and conduct analytics on different kinds of data & information generated. Conduct Deep Competitor analysis, unmasking their strategies, preparing our approach plan on the basis of analytical data and statistics. Requirements Engineering or MBA from tier 1 institutes 3+ Years of past experience working upon E Commerce Product Categories\/ Analytics Team management skills and ability to get the work done in an efficient manner Should possess Analytical capabilities such as modeling, data mining, and advanced segmentation, reporting and forecasting Should have expert knowledge in w.r.t the latest statistical software SQL, R\/ Python, BI platforms (Tableau) or other Analytical tool shall be preferred. Apply here Product & Analytics Manager @ HealthFolks The candidate shall be responsible for managing the products on our Portal on the basis of analyzing the market trends and conduct analytics on different kinds of data & information generated. Conduct Deep Competitor analysis, unmasking their strategies, preparing our approach plan on the basis of analytical data and statistics. Implement tools and strategies to translate raw data into valuable business insights. Requirements Engineering or MBA from tier 1 institutes 3+ Years of past experience in the relevant field working upon E-Commerce Product or related Team management skills and the ability to get the work done in an efficient manner Should possess Analytical capabilities such as modelling, data mining, and advanced segmentation, reporting and forecasting Should have expert knowledge in w.r. the latest statistical software SQL, R\/ Python, BI platforms (Tableau) or other Analytical tool shall be preferred. Apply here","excerpt":"Analytics India Jobs offers a new platform for the data science community in India by bringing to the fore latest job openings across the country. Here are few Data Science job openings in Hyderabad data science enthusiasts can apply to: Deep Learning and Computer Vision Expert @ Pratibha Analytics The candidate will be part of […]","categories":["AI Hirings"],"tags":["Data Science Jobs","Hyderabad","Python","science jobs"],"author_name":"Ram Sagar","publish_date":"2019-03-18T10:04:30","publication_year":"2019","word_count":777,"keywords":["data science","artificial intelligence","machine learning","AI","science jobs","PyTorch","ML","Data Science Jobs","Hyderabad","computer vision","Python","deep learning","analytics","TensorFlow"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","computer vision","data science","analytics","TensorFlow","PyTorch"],"url":"https:\/\/analyticsindiamag.com\/ai-hiring\/4-latest-data-science-jobs-in-hyderabad\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":5944,"title":"Part 2: Creating the Customer Life Stage model","content":"This is the second installation in a series of articles on Customer Life Stage modelling. The first article has been published at https:\/\/analyticsindiamag.com\/part-1-customer-life-cycle-a-misleading-term\/ . In the earlier article, I highlighted the difference between Customer Life Cycle and Customer Life Stage. The article also listed 8 dimensions along which the Customer Life Stage can be defined. These dimensions may not be a comprehensive list but in my experience they are more than sufficient. This article will attempt to create a sample Customer Life Stage model. We will use three dimensions for our sample. A one dimensional model does not make much value addition. A two dimensional model simplifies the approach into a comfortable analysis. Anything more than 3 dimensions may make it difficult to visualize. So, we will stick to the geometry that everyone is comfortable with – A cube. I also recommend that you adopt a three dimensional approach when embarking on a Customer Life Stage modelling. When you reach a maturity with this model, you may attempt to add a fourth dimension to the Customer Life Stage definition. For our purpose, we will consider a banking scenario and consider the following three dimensions: Vintage: Vintage is defined as the time the customer has been on books of the company, stated in months. A telco would term this as month on books. Product Holding: Product holding dimension aims at listing the products held by the customer. I will be discussing this dimension later in this article. Product Value: Product value defines as the value of the product. Since, different products have different value, it is necessary to arrive at a normalized definition of the value so that it can be summed up over all products. For example, a bank could consider net revenue spread from product holding. A telco could consider expensed value of services. Before we proceed, let’s approach the method to arrive at the values for the product holding dimension. At the basic, the set of values should be individual products, say, Savings Account, Current Account, Deposit, Credit Card, Loan. Next we need to find the commonly occurring combination of products. This is where we will cross over into statistical domain. While I attempt to keep it simple, kindly go to http:\/\/en.wikipedia.org\/wiki\/Association_rule_learning if you need more information. Consider, we have 5 customers with the following holding across four products. Customer Savings Account (SA) Deposit(DP) Credit Card(CC) Loan(LN) Customer 1 1 1 0 0 Customer 2 0 0 1 0 Customer 3 0 0 0 1 Customer 4 1 1 1 0 Customer 5 0 1 1 0 Let us first consider 2 product combinations: SA – DP (2 nos), SA – CC (1 nos), DP – CC (2 nos). Considering 3 product combinations: SA – DP – CC (1 nos) Taking each of these combinations, we have the following list of values for the product holding dimensions: SA DP CC LN SA – DP SA – CC DP – CC SA – DP – CC We may decide to filter this list further. For example, a bank may decide only to lend to customers who hold savings account. So, loans can be taken out as a stand-alone value since no customer will have only loans. We may also set a threshold for occurrence of the combinations of product holding. Say, any combination that is less than 2% or less than 200 occurrences may be filtered out. In our example, we will ignore all combinations that is less than 2 occurrences. So, we now have the following selected values: SA DP CC SA – DP DP – CC We will not get into concepts of confidence and support definitions here. See the link above for more info on these. The list of values for vintage could be: 0 to 6 months 6 to 12 months 12 to 24 months 24 + months The list of values for product value could be: 0 to 1K 1K to 10 K 10K to 50 K 50 + K Now the Customer Life Stage model is a three dimensional model containing 80 possible stages ( 5 product holding x 4 vintage x 4 product value ). A customer may be in any of these stage cells. In the next article, we will aim to traverse this space and draw up the customer life stage models.","excerpt":"This is the second installation in a series of articles on Customer Life Stage modelling. The first article has been published at https:\/\/analyticsindiamag.com\/part-1-customer-life-cycle-a-misleading-term\/ . In the earlier article, I highlighted the difference between Customer Life Cycle and Customer Life Stage. The article also listed 8 dimensions along which the Customer Life Stage can be defined. […]","categories":["IT Services"],"tags":[],"author_name":"Feroz D Silva","publish_date":"2014-07-22T09:06:54","publication_year":"2014","word_count":722,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","Aim","analytics","R"],"extracted_tech_keywords":["AI","analytics","Aim","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/part-2-creating-the-customer-life-stage-model\/","complexity_score":3,"technical_depth":7,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10116962,"title":"Where Are All the Early AI Unicorns Now?","content":"It’s super quiet right now. There’s little or no sign of the early AI unicorns that were once flying high above the rainbow in cotton candy skies with their cute cat-like ideas. Source: (ImgFlip) Let’s see what they are up to now. According to several reports, Stability AI is treading on very unstable ground. Several developers, who worked on the company’s most important product, Stable Diffusion, are resigning. CEO Emad Mostaque said that three of the five researchers have left the company. Investors too have been pretty unhappy with Mostaque’s leadership, and the startup is now up for sale. It is hard to remain on top of the AI game when there is so much competition. Many AI startups that tasted success initially are now lost amidst all the announcements and generative AI news. Another such case is Jasper. The AI content platform based in Austin, became a unicorn after raising $125 million in a Series A funding round led by Salesforce at a $1.5 billion valuation in October 2022. In September 2023, the company’s CEO was changed. Timothy Young, the former president of Dropbox was appointed as the CEO, replacing Dave Rogenmoser. This was discussed as a highly suspicious move in the VC and generative AI startups circles. The worst is that this came after the company announced layoffs, just months after its funding round. The same month, Jasper reportedly reduced the internal worth of its ordinary shares by 20%, as informed by ex-employees who received notifications from the company. This decline suggests a deceleration in the advancement of its AI-driven writing tool tailored for marketers. While Jasper has historically depended on OpenAI’s GPT-3 to fuel its product, the emergence of ChatGPT by OpenAI last autumn has essentially positioned it as a direct rival. It has always been touted that whenever OpenAI releases a new update, a lot of startups are killed. Moreover, there has been a lot of hype around investing in generative AI startups that do not have any moat, resulting in a failed investment. Muffled voices Former Adobe CTO’s startup Typeface last raised a funding of $100 million in June 2023, becoming a unicorn. The generative AI for brand startup has been quiet ever since its competitors began to rise in the field. On the other hand, startups such as Synthesia, Runway, Cohere, Coreweave, and Replit have all been coming up with new features. The only thing that still remains undiscussed is these companies’ road to profitability. Though they all have some or the other revenue model in place for customers, the cost of running the models is still exceptionally high. Before OpenAI announced Sora, there was another AI-powered filmmaking platform called Lightricks. The company which raised $130 million funding in September 2021, reaching a valuation of $1.8 billion, also laid off 80 of its employees, which was around 12% of its workforce. It continued to raise several rounds of funding after that and released LTX Studio just last month. If we take a look at the recent Instoried debacle, it gives away the sense and nature of AI investments globally. The company started seven years ago with the motivation of adding empathy to users’ content and pitched the idea of AI, but it didn’t really become a unicorn. Cut to present, there is almost no sign of the company, which according to the now-deleted LinkedIn profile claimed to have created ChatGPT way back in 2019. Inflection AI, the company that hit the news because two of its co-founders left to join Microsoft AI also raises eyebrows about the investment hype in the earlier AI boom. The company’s future remains hazy even though it raised billions of dollars and became a unicorn. Another success story is that of Glean, the AI-powered work assistant startup founded by Arvind Jain. Just a month back, the unicorn raised $200 million at $2.2 billion valuation in a series D round. It has also announced going on a hiring spree as the demand for its enterprise AI solution is on the rise. Leaving everyone else behind is the success story of Hugging Face, Character.ai, and Adept AI that show that not all generative AI unicorns are on the negative track. These startups have been raising funds and partnering with many almost every second month. Only the ones that raised funds without a moat to defend against big-tech and OpenAI are on the decline. Only time will tell what happens to the AI unicorns that have been raising funds over the past year, such as Ola’s Krutrim, which recently became India’s first and fastest AI unicorn.","excerpt":"The text-to-image giant Stability AI is in trouble. While others like Glean are on the rise, Jasper is slowly vanishing, and the layoffs are constant.","categories":["AI Features"],"tags":["Replit"],"author_name":"Mohit Pandey","publish_date":"2024-03-22T14:12:35","publication_year":"2024","word_count":764,"keywords":["Replit","Go","ChatGPT","Hugging Face","OpenAI","AI","GPT","Aim","stable diffusion","generative AI","R"],"extracted_tech_keywords":["AI","generative AI","ChatGPT","OpenAI","Aim","Hugging Face","R","Go","GPT","stable diffusion"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/where-are-all-the-early-ai-unicorns-now\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10140079,"title":"GenAI Can Boost Global Banking Revenues by up to $340 Billion","content":"Like the health and hospitality sectors, banking too relies largely on human interactions and emotions. While the potential for generative AI in banking is remarkable, it may fail to judge human behaviour and falter, especially when offering tailored financial products and services to customers. The McKinsey Global Institute (MGI) projects that GenAI could generate an additional $200-340 billion annually across the global banking sector, boosting industry revenues by 2.8 to 4.7 per cent mainly through enhanced productivity. So, how are banks embracing AI? In 2023, Accenture highlighted a transformative trend in Indian banking. It noted that with AI integration, banks can serve many customers simultaneously, boosting both transaction volume and digital interactions. Back in 2017, HDFC Bank Ltd introduced India’s first AI-driven banking chatbot, Eva (Electronic Virtual Assistant). Powered by Bengaluru-based startup Senseforth, Eva was designed as a breakthrough in customer service, capable of handling millions of queries instantly across multiple platforms. Eva set the standard for AI-assisted customer interaction with the ability to pull information from thousands of sources and respond in under 0.4 seconds. By 2020, ICICI Bank expanded on this concept with its chatbot, iPal, integrating it with Amazon Alexa and Google Assistant. This allowed retail banking customers to perform various transactions through simple voice commands. However, the service was discontinued in August 2021. More recently, in 2023, the State Bank of India (SBI) announced a strategic AI-driven initiative aimed at enhancing decision-making and operational efficiency. With plans to deploy advanced data warehouses and data lakes, SBI is also exploring partnerships with fintech firms and non-banking financial companies (NBFCs) to revolutionise co-lending practices and drive a more connected financial ecosystem. Meanwhile, Deutsche Bank’s innovative AI journey is powered by data approach, control, and talent. It has collaborated with Google Cloud (since 2020) and NVIDIA (from 2022), accelerating cloud transformation and AI adoption across the bank. In 2023, it launched a bank-wide AI program with practical applications like AI chatbots, developer support tools, and unstructured data analysis, positioning it as an early adopter of generative AI and LLMs. But is Everything Safe? According to Kroll’s 2023 Fraud and Financial Crime Report, a survey of 400 senior executives across three continents revealed that 67% expect financial crime to rise in the coming year, with 57% identifying third-party gatekeepers as a key risk factor. The link between money laundering and organised crime is significant, with up to $2 trillion (2-5% of global GDP) laundered annually as criminals work to mask illicit gains. In response, banks are increasingly looking to AI. For example, HSBC previously relied on rule-based systems to spot potential money laundering, often leading to numerous “false positives” that required manual review. Now, in partnership with Google Cloud, it uses advanced AI trained on extensive data to recognise suspicious activity autonomously. This AI-driven Anti Money Laundering (AML) solution is more precise, reducing false positives and enhancing its detection capabilities without pre-set parameters. As banks evolve their defences, fraudsters are getting more adept at bypassing them. Speaking at the DECODE webinar, Sahil Aneja, vice president at HDFC Bank, pointed out that traditional rule-based monitoring, though foundational, is rigid and struggles to keep up with the fraudsters’ methods. “For example, when UPI first launched, there was a significant spike in fraud incidents. Banks responded by setting thresholds for unusual transactions, which temporarily reduced the scale and frequency of fraud. However, fraudsters soon adjusted to these rules, necessitating a shift to AI-driven platforms for better fraud detection,” he said. Aneja spoke about a general benchmark in fraud prevention, indicating that within 30-45 days, fraudsters adapt and develop new methods to bypass platforms put in place by financial institutions. This means that institutions must continuously refine their systems, transitioning to self-learning models. What’s Next? Infosys Finacle, a part of EdgeVerve Systems, a fully-owned subsidiary of Infosys, recently introduced the Finacle Data and AI Suite, a powerful toolkit designed to bring AI seamlessly into banks’ digital operations and fast-track their AI journey. This suite provides a collection of platforms enabling banks to build low-code, predictive, and generative AI solutions from scratch, with a focus on transparency and explainability. With this, banks can boost their data readiness, standardise AI model development, harness generative AI technologies, and deliver actionable insights across their entire ecosystem. Meanwhile, Axis Bank believes that AI will not change the nature of work in India. However, the Mumbai-based firm has ramped up its technology team to 800 employees, up from about 60 nearly five years ago. The bank currently employs about 70 people who work exclusively on AI and plans to further expand its team by 10% yearly. This shows that the adoption of AI in the banking industry is not slowing down.","excerpt":"However, as banks evolve their defences, fraudsters are getting more adept at bypassing them.","categories":["AI Features"],"tags":["AI (Artificial Intelligence)","Bank","hdfc bank","ICICI Bank"],"author_name":"Vidyashree Srinivas","publish_date":"2024-11-04T12:23:38","publication_year":"2024","word_count":781,"keywords":["Bank","Go","GenAI","AI","ICICI Bank","chatbots","ML","Git","Aim","generative AI","hdfc bank","R","AI (Artificial Intelligence)","fraud detection"],"extracted_tech_keywords":["AI","ML","generative AI","GenAI","Aim","chatbots","fraud detection","R","Go","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/genai-can-boost-global-banking-revenues-by-up-to-340-billion\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10039541,"title":"Top MLOps Tool Repos On Github","content":"MLOps has been introduced to provide an end-to-end machine learning development process to design, build and manage reproducible, testable, and evolvable ML-powered software. MLOps allowed organisations to collaborate across departments and accelerate workflows, which usually hit the wall due to various issues in the production. In the next section, we present top MLOps tool repos that are available on Github (Image credits: Microsoft Azure) Here are the top Github MLOps tools repos: Seldon Core Seldon core is an MLOps framework to package, deploy, monitor and manage thousands of production machine learning models. It converts machine learning models (built on Tensorflow, Pytorch, H2o, etc.) or language wrappers (built on Python, Java, etc.) into production microservices. Seldon Core makes scaling to thousands of production machine learning models possible and provides advanced ML capabilities that include Advanced Metrics, Request Logging, Explainers, Outlier Detectors, A\/B Tests, Canaries and more. It makes deployment easy through their pre-packaged inference servers and language wrappers. Check the full repo here. Polyaxon Polyaxon can be used for building, training, and monitoring large scale deep learning applications. This platform is built to address reproducibility, automation, and scalability for ML applications. Polyaxon can be deployed into any data centre, cloud provider, or can be hosted. It supports all the major deep learning frameworks such as Tensorflow, MXNet, Caffe, Torch, etc. According to the team that developed Polyaxon, the platform makes it developing ML applications faster, easier, and more efficient by managing workloads with smart container and node management. It even turns GPU servers into shared, self-service resources for teams. Installation: $ pip install -U polyaxon Check the full repo here. Hydrosphere Serving Hydrosphere Serving offers deploying and versioning options for machine learning models in production. This MLOps platform: Can serve machine learning models developed in any language or framework. It wraps them in a Docker image and deploys it onto the production cluster, exposing HTTP, gRPC and Kafka interfaces.Shadows traffic between different model versions to examine how different model versions behave on the same traffic.Versions control models and pipelines as they are deployed. Check the full repo here. Metaflow Metaflow was originally developed at Netflix to address the needs of its data scientists who work on demanding real-life data science projects. Netflix open-sourced Metaflow in 2019. Metaflow helps users design your workflow, run it at scale, and deploy it to production. It versions and tracks all your experiments and data automatically. Metaflow provides built-in integrations to storage, compute, and machine learning services in the AWS cloud. No code changes required. Check the full repo here. Kedro Kedro is an open-source Python framework which can be used for creating reproducible, maintainable and modular data science code. Kedro is built on the foundations of  software engineering and applies them to machine-learning code; applied concepts include modularity, separation of concerns and versioning. Check the full repo here BentoML As a flexible, high-performance framework, BentoML can be used for serving, managing, and deploying machine learning models. It does this by providing a standard interface for describing a prediction service and abstracting away how to run model inference efficiently and how model serving workloads can integrate with cloud infrastructures. BentoML’s features include: Production-ready online serving.Supports multiple ML frameworks including PyTorch, TensorFlow.Containerized model server for production deployment with Docker, Kubernetes etc.Discover and package all dependencies automatically.Serve any Python code along with trained models.Health check endpoint and Prometheus \/metrics endpoint for monitoring. Check the full repo here. Flyte Flyte provides production-grade, container-native, type-safe workflow platforms optimized for large scale processing. It is written in Golang and enables highly concurrent, scalable and maintainable workflows for machine learning and data processing. It connects disparate computation backends using a type safe data dependency graph and records all changes to a pipeline, making it possible to rewind time. Check the full repo here.","excerpt":"MLOps has been introduced to provide an end-to-end machine learning development process to design, build and manage reproducible, testable, and evolvable ML-powered software. MLOps allowed organisations to collaborate across departments and accelerate workflows, which usually hit the wall due to various issues in the production. In the next section, we present top MLOps tool repos […]","categories":["AI Trends"],"tags":["MLops tools"],"author_name":"Ram Sagar","publish_date":"2021-05-04T16:00:00","publication_year":"2021","word_count":631,"keywords":["data science","machine learning","AI","BentoML","Seldon","ML","MLOps","PyTorch","MLops tools","deep learning","TensorFlow"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","data science","MLOps","BentoML","Seldon","TensorFlow","PyTorch"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-mlops-tools-github-repos\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10067980,"title":"Consolidation of data science education industry","content":"Earlier in May, edtech unicorn upGrad acquired the International School of Engineering (INSOFE) in a USD 33 million share swap deal. A profitable company, INSOFE is an educational institution that focuses on data science, artificial intelligence, and machine learning courses. In the last year alone, upGrad has acquired a few major players in the data science edtech space, including names like KnowledgeHut and TalentEdge. This aggressive acquisition strategy is not limited to just upGrad. Decacorn Byju’s added Great Learning to its long list of companies that it has overtaken via a whopping USD 600 million deal. As happens with most industries, after a phase of rapid growth, through a series of mergers and acquisitions, they progress through a consolidation lifecycle. Such consolidation often leaves dominance of a few handful companies. Looking at how the data science edtech domain is growing, could we anticipate such a consolidation lifecycle happening here too? The edtech boom While other industries were busy playing catch up in the wake of the COVID-19 pandemic, one industry that jumped, sprinted and leapt to arguably become the most exciting and fast-paced domain was edtech. The massive growth potential caught favour with the investors as well. Ranking behind just eCommerce (USD 10.7 billion) and fintech (USD 8 billion), the edtech sector raised USD 4.7 billion to emerge as the third most funded sector in 2021. As per data from Inc42, the test preparation startups walked away with the major chunk of the funding – 65 per cent of the total funds (USD 3.06 billion). It was followed by online certification (22 per cent or USD 1.04 billion), K12 (5.7 per cent or USD 270 million), and enterprise edtech (5 per cent or USD 240 million). In the edtech sector, one area which generated such interest was the data science edtech market. As per a report, the data science edtech market is estimated to grow from USD 103 million in 2020 to USD 626 million by 2025 at a CAGR of 43 per cent during this period. The market for off-campus or online courses, which include executive programs and online certifications, is projected to grow to USD 239.6 million. This is hardly surprising, given the demand for data scientists being consistently high, translating into heavy paycheques and an exciting career graph. Analysts believe that the number of data science and related jobs will touch 11 million openings by 2026. India alone contributed to 11.6 per cent of the total global data science and related open jobs. Some of the factors that played a part in this are a continuous investment made by domestic companies in developing advanced data science\/analytics capabilities; MNCs and IT companies moving their data science jobs in large numbers to india; increased funding of AI\/analytics-based startups. Credit: AIM Research Major M&As In 2021 alone, Byju’s acquired ten companies for a combined transactional value of USD 2.5 billion. This propelled Byju’s to become the most active media and telecom (TMT) company across the global merger and acquisition market in 2021, a report from GlobalData said. After the much talked about billion-dollar acquisition of Akash Institutes, a name synonymous with competitive exam coaching, Byju’s next big hunt has been Great Learning.  With Great Learning, Byju’s marked its entry into the professional upskilling and life-long learning space. Apart from the USD 600 million Byju’s spent on acquiring Great Learning, the former has also earmarked USD 400 million for this segment in the near term. In a public release, representatives of Great Learning said that they would now work to accelerate their organic and inorganic growth in India and overseas markets like North America, Canada, etc. Interestingly, Great Learning recently acquired Singapore-based Northwest Executive Education in a cash and equity deal. Founded in 2015, Northwest Executive education offers management, technology, and other executive courses from MIT, UC Berkeley, Yale, UCLA, the University of Chicago and National University of Singapore, etc., and is expanding across the US, Europe, Asia, and Latin America. Going the Byju’s way is another major edtech giant – upGrad. It has recently acquired INSOFE (latest), TalentEdge and KnowledgeHut. In an email interaction with Analytics India Magazine, Mayank Kumar, co-founder, and MD, upGrad, said, “When we look at acquisitions, it’s not just about a business model but is about aligning our vision with the partner brand for driving impactful education at scale. Talking about INSOFE and TalentEdge in specific, these strategic business movements will not only accelerate cost-efficiency but will also empower us with a wider audience reach for further cementing our leadership position in India’s lifelong learning market.” He also mentioned that upGrad’s analytics vertical, with subjects like data science, machine learning, and business analytics, has grown by 50 per cent in terms of revenue since the first Covid-19 wave in Q2FY21 (JAS 2020) as compared to the third wave during the Q4FY22 (JFM 2022). He predicts that analytics as a domain with data science at its core will continue to grow, thus preparing professionals for the upcoming Industry 5.0. Scaler, which bagged USD 55 million in a Series B funding round in February, acquired learning platform AppliedRoots in a USD 50 million deal to strengthen its data science, AI and ML offerings. This is the company’s third acquisition after Coding Minutes and Coding Elements in 2021. Further, Manipal Group-backed edtech company UNext Learning acquired Jigsaw Academy to strengthen its higher education offerings by adding specialised professional certification courses. In 2020, the National Stock Exchange’s arm – NSE Academy, acquired Talentsprint, a prestigious name in the AI, ML and data science education space. Relatively smaller companies are also making strides. Higher education and professional course development platform iNurture recently acquired placement and professional skill development platform Krackin. Market consolidation It is a general trend observed across industries that after several mergers and acquisitions, many progress through consolidation. As per a study of 1,345 mergers and acquisitions completed over 13 years, once an industry forms, it goes through four stages of consolidation over 25 years. The research predicts that this duration will reduce in future. Credit: Harvard Business Review Last year, the Hero Group launched Hero Vired in April. Their first program went live in July, and the company has since then introduced many similar industry-relevant programs. As per a few sources, Hero Vired is working on an expansion strategy, and they may be looking at acquiring similarly aligned firms. Speaking to Analytics India Magazine on the consolidation of the data science edtech space, Dipayaman Sanyal, the head of Academics and Learning, said, “Like the fintech industry, the data science edtech sector is also seeing consolidation. We see bigger companies taking over smaller firms. One of the issues with this trend is that it may affect the risk-taking capability of the acquired companies, as the bigger firms generally stick to traditional and tested ways of operating. But that is how the industry evolves. That said, I am positive that this phase will also give space for newer competitors to emerge, filling the void.” About the implication of such consolidations, Mayank Kumar of upGrad, said, “We have seen consolidations in the edtech sector over the last 12 months, and I feel that this may slow down now in the current environment. But, the larger companies in this space will continue to chase opportunities to grow market share inorganically. They will continue investing in developing their own delivery channels, offering hybrid learning solutions, expanding platform development, and expanding globally to enter adjacent and associated markets. I also believe collaboration and partnerships will observe a marked increase in the edtech sector. As we proceed from here, we can expect a lot of younger companies at play, and the market will pivot to drive up platform innovation across all micro-levels.” Subramanyam Reddy, founder and CEO of KnowledgeHut, a company recently acquired by upGrad, believes that some level of consolidation may happen on the short-term horizon. He said, “The data science domain growth is leading to a proportional growth of the data science edtech industry, which is also seeing growth driven by various factors like the pandemic and various government policies. The edtech industry is likely to remain bullish on growth for some time to come. Considering that the amount of funding will reduce over the coming months, there will be some level of consolidation that might happen in the short-term horizon. However, talking about the online higher education sector, the need for data science has never been more relevant, and we feel it will continue to drive overall growth in this sector.” Paul Kim, the chief technology officer and Assistant Dean of the Graduate School of Education at Stanford University, in an earlier interview, said that India is in a much better position in terms of data science education because of the presence of multiple technology innovation powerhouses here. He said that while US and China are big contenders, institutional financing and governmental support in terms of policy and regulation would ensure India’s rapid growth in this area. Even with the world getting back to normalcy and brick and mortar institutions starting to open their doors, these online lifelong learning program providers are unlikely to suffer a major blow. However, as Prof Kim mentioned, traditional establishments must transform to align with competing alternative education options.","excerpt":"As happens with most industries, after a phase of rapid growth, through a series of mergers and acquisitions, they progress through a consolidation lifecycle.","categories":["AI Trends"],"tags":["Byju’s","Data Science","Great learning","Insofe","UpGrad"],"author_name":"Shraddha Goled","publish_date":"2022-05-30T10:00:00","publication_year":"2022","word_count":1542,"keywords":["data science","Go","API","artificial intelligence","machine learning","AI","ML","Byju’s","Insofe","Aim","UpGrad","analytics","Great learning","Data Science","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","data science","analytics","Aim","R","Go","API"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/consolidation-of-data-science-edtech-industry\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10041988,"title":"The Future Of Gaming is Cloud. And How","content":"In the last few years, gaming has become one of the fastest-growing industries in India. From the days of playing Tetris on the computer to implementing AI in GTA5, the gaming industry has always been at the forefront of innovation and progress. However, a new chapter has now been added to the world of gaming through cloud computing. Cloud computing has revolutionised video streaming services and led to the foundation of giant corporations such as Google and Netflix. A prime example of a platform that grew on the cloud, YouTube has scaled massively in the last decade to provide an even better video streaming experience. While the gaming industry’s future is expected to skyrocket in the coming year, according to a report by Research and Markets, the global gaming market was worth $ 167.9 billion in 2020, and was expected to reach $ 287.1 billion by 2026, growing at CAGR of 9.24 percent between 2021-2026. However, cloud gaming continues to remain in an infantile stage. Companies like Microsoft are taking giant leaps towards the growth of AI in the gaming industry through cloud computing. Ahead of its Xbox and Bethesda Showcase, Microsoft announced the latest updates on its expansions of the Xbox Game Pass subscription services. From increasing the number of screens, to including smart TVs from third-party manufacturers and adding its own streaming devices, Microsoft is on track to capture a more significant portion of the gaming market. It plans to use the cloud to provide an Xbox cloud gaming subscription service on devices with less horsepower. The company already offers a similar service on Android and iOS devices via a beta version of its Xbox Cloud Gaming service. Google Stadia A couple of years back, when Google announced its cloud gaming service Stadia, it was met with scepticism. At that time, it was advertised as being capable of streaming 4k resolution video at 60 fps to players through the company’s many data centres across Chrome browser, Chromecast, and Pixel devices. Tech gurus, however, argued that the streaming experience was not as smooth as a home console. But Stadia was just the beginning of a new gaming era. Amazon Luna In September 2020, Amazon announced its cloud gaming platform Amazon Luna. Amazon Luna is a cloud-based video streaming service for Fire TV, smartphones, tablets, and computers. It competes with the likes of Microsoft xCloud and Google Stadia. Apart from the 4k support at 60fps, Luna is also said to integrate Twitch– a popular video streaming site that allows gamers to broadcast and game simultaneously. Amazon Luna runs on Windows servers and NVIDIA GPUs in Amazon’s AWS compute cloud, allowing developers to port existing games to Luna easily. Xbox With the latest Game Pass Subscription expansion, Xbox is working with global TV manufacturers to embed the Xbox experience directly into internet-connected televisions with no extra hardware. It also explores new subscription offerings for Xbox Game Pass, allowing more players to experience immersive games across devices. Besides this, Xbox is also building its own streaming devices for cloud gaming to reach gamers on display devices like TV or monitor, without a console that could ultimately change the gaming experience. Currently, Microsoft is upgrading its data centres across the globe with the Xbox Series X hardware to improve the gaming experience by enabling faster load times, improved frame rates, and the ability to play Xbox Series X|S optimised games. Later this year, Xbox plans to add cloud gaming directly into the Xbox app on PC, integrate it into their console experience, and provide users with experiences like’ try before you download.’ While cloud gaming is currently a barely functioning concept, the technology represents the future of gaming. It won’t be too long before more players adopt cloud gaming to provide a hassle-free immersive gaming experience to their customers.","excerpt":"In the last few years, gaming has become one of the fastest-growing industries in India. From the days of playing Tetris on the computer to implementing AI in GTA5, the gaming industry has always been at the forefront of innovation and progress. However, a new chapter has now been added to the world of gaming […]","categories":["AI Features"],"tags":["Cloud Computing","Cloud Gaming","netflix","YouTube"],"author_name":"Ritika Sagar","publish_date":"2021-06-19T18:00:00","publication_year":"2021","word_count":636,"keywords":["Go","Cloud Gaming","AWS","AI","cloud computing","innovation","cloud_platforms:AWS","programming_languages:R","programming_languages:Go","netflix","Cloud Computing","YouTube","R"],"extracted_tech_keywords":["AI","cloud computing","AWS","R","Go","innovation","cloud_platforms:AWS","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/the-future-of-gaming-is-cloud-and-how\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":32744,"title":"10 Best Artificial Intelligence &#038; Machine Learning Stocks To Buy In 2019","content":"AI and ML stocks are also one of the hot properties to invest in share market now. With growing influence through the applicability of emerging technologies like AI and ML in varied sectors and industries around the world. One of the major reasons why AI and ML stocks are picking up among the stockholder circles is because of the multiple advantages it offers. Be it autonomous driving or smart electronic gadgets, there have been more ups than downs in the AI sector overall. With artificial intelligence being declared as the future of tech industry, we list top 10 AI and ML-based stocks to buy in 2019. These companies are not listed in any particular order. 5 Major Companies To Invest In – 1| ALPHABET Market Value – $812.0 billion Listed on NASDAQ: AAPL Reasons To Invest – One of the most direct ways Alphabet uses machine learning right now is through the company’s self-driving vehicle company Waymo and the machine learning software that’s driving the vehicles is second to none. The company’s machine learning systems have been able to improve Waymo’s cars drive because they’ve been fed lots of data from the  10 million miles they’ve driven on the road and the 7 billion virtual distance in miles they’ve driven through augmented scenarios. Alphabet, Google’s parent company’s stocks will not skyrocket like a smaller company would if its AI investments are about to pay off. It is a company which is perhaps a few years away from being valued at $1 trillion valuations, and uses AI as a driving force behind everything, from search results to ad pricing and autonomous driving technology. So, investing in its stocks can prove beneficial. 2| AMAZON Market Value – $177.866 billion Listed on NASDAQ: AMZ Reasons To Invest – Amazon CEO Jeff Bezos has been the driving force behind the company’s meteoric rise. Machine learning already is in use in core parts in  Amazon-like product search rankings, recommendations, demand forecasting, fraud protection and warehouse fulfilment, and AI is also central to Alexa virtual assistant and future for drone delivery and cashier-less grocery stores. According to Bezos, the most apparent machine learning use comes in the form of figuring out where and how items will be on the Amazon website for users to see, and which products Amazon recommends to its users. Amazon is also looking at machine learning to help drive its fast-growing advertising business. They are also in the process of utilising machine learning and analytical techniques to create solutions across the company’s business which includes its ad-serving pipeline. Also, Amazon is developing data and analytics tools for brands, backed by machine learning and its ubiquitous web services, so that it can use the tech to find relevant data for ad targeting and measure ad success. 3|  NVIDIA Market Value – $9.714 billion Listed on NASDAQ: NVDA Reasons To Invest – Chipmaker NVIDIA Corporation defines AI and ML technology as the practice of using algorithms to parse data, learn from it and then decide or predict about something in the world. NVIDIA’s graphics processing units are used by all the major technology companies to help their servers implement machine learning services. But the company has also used its GPUs to build semi-autonomous driving technology. Its Drive PX Pegasus system allows vehicles to see and process the images captured by onboard cameras so that the vehicles can correctly navigate their surroundings. Its Pegasus is already being used by more than 25 companies to develop fully autonomous robotaxis that can drive themselves. NVIDIA believes that autonomous vehicles represent a $60 billion total addressable market for the company by 2035. But, 2019 will bound to be a crucial year for the market which will likely ramp up over the coming years. 4|  IBM Market Value – $ 79.139 billion Listed on NYSE: IBM Reasons To Invest – AI is not new to IBM. The company has been developing this type of technology since its inception. IBM developed its AI computer called Deep Blue in 1985 itself far ahead of other. IBM created Watson to take on the best players on the quiz show Jeopardy where the computer eventually won. Investing in  IBM  with a focus on its AI and other cutting-edge technologies can make a big difference. IBM’s strategic imperatives include cloud computing, security, analytics, big data and mobile generated $39 billion or about 48% of total revenues. All these technologies have been the primal force in improving the growth rate of overall businesses too. IBM stock also has promising dividends, which is pegged at 4.1%. This is one of the highest in the tech industry. The valuation is reasonable as the forward price-to-earnings ratio is the only 11X. 5|  Microsoft (MSFT) Market Value – $771.0 billion Listed on NASDAQ: MSFT Reasons To Invest  – Microsoft Corporation acquired Canadian AI company Maluuba this year. This Canadian company teaches machines to think and ask questions through deep learning methods. It’s known for using AI to beat the notoriously difficult Ms Pac-Man arcade video game. According to Microsoft CEO Satya Nadella, the MNC is in the process to “democratize AI” and bring the technology to more industries such as healthcare, education and manufacturing. MSFT has pivoted to become a cloud-centric company, jockeying with Amazon and Google for AI breakthroughs. Microsoft’s major focus is also on talent in artificial intelligence research, where it’s developing cloud-based tools for genomics and precision medicine, human language technologies, assistive robotics, machines that can read medical images and much, much more. With consumer-facing services like Cortana as well, Microsoft is one of the best artificial intelligence stocks to buy for the coming years. 5 Lesser Known Companies To Invest In 2019 – 6| YEXT Market Value – $170.2 million Listed on NYSE: YEXT Reasons To Invest  – Yext being the top company is providing data provider, with integrations of over 150 services with help of AI from operators Google, Apple Amazon.com Microsoft, Facebook and Tencent. Yext also uses context and intent to its work which allows for more accurate real-time searches. According to its CEO Howard Lerman, the world is moving to smart databases and AI-powered services is the future. Growth has been strong for Yext. In the previous quarter, revenues shot up by 35% to $55.1 million. The company has also been getting more attraction from enterprise customers. 7  | BAIDU Market Value – CNY 84.809 billion Listed on NASDAQ: BIDU Reasons To Invest – When it comes to the search business, Baidu is arguably the most advanced AI company in China. But BAIDU has also invested heavily in becoming an artificial intelligence stock. This has helped with personalizing the search experience as well as improving the impact of online ads. The company has created several platforms for third parties like Apollo. It is an AI system for autonomous vehicles. BAIDU is hoping that its Apollo technology will eventually power most of the world’s cars, while other projects like virtual assistants and voice replication are already coming along. The AI efforts have been paying off. In the latest quarter, revenues jumped by 32% to $3.93 billion and the adjusted EBITDA came to $1.12 billion — or about 29% of total revenues. BIDU also has a highly scalable business model. BAIDU stock has taken a hit this year, going from $273  to $218 in 2018. But for investors looking for a play on AI in China, Baidu stock does look attractive at these levels. 8  |Delphi Automotive \/APTIV Market Value – US$ 12.9 billion Listed on NYSE: DLPH \/ APTV Reasons To Invest  – Delphi Automotive PLC (NYSE: DLPH) is on the rise. Delphi grew by 30% in previous years and is predicting revenue of $16.5 billion to 16.9 billion for the full year 2019. Delphi  also left its powertrain business to focus on self-driving cars and electric vehicles With BMW, Intel Corporation and Mobileye NV, Delphi plans to launch self-driving cars by 2021. DLPH received four buy ratings and one hold rating recently, which has nearly 70.30% upside potential for the next 12 months. 9 | Salesforce.com Market Value – $100.4 billion Listed on NYSE: CRM Reasons To Invest- Cloud computing giant Salesforce.com offers customer relationship management to over nearly 100,000 customers But the company has been focusing on  AI and made big leaps through the launch of Einstein Analytics According to CRM, Einstein Analytics apps can automatically analyse billions of data combinations to surface predictive insights and prescriptive recommendations. Monness Crespi analyst Brian White pointed out that on the AI front, Einstein is providing nearly 2 billion predictions each day and double the 1 billion in 4QFY18 after CRM’s first-quarter earnings in 2018. In 2018, Salesforce announced two big partnerships one with IBM’s powerful AI platform Watson and the other with Amazon Connect. 10 | BioXcel Therapeutics Market Value – $139.1 million Listed on BTAI (NASDAQ) Reasons To Invest- While most of AI attention is focused on tech stocks, AI also has the potential to drive huge change when it comes to the world of healthcare and drug development. One top-rated biotech stock making the most of these new opportunities is BioXcel Therapeutics. BioXcel uses AI to pinpoint prospective neuroscience and immunooncology drugs. Through AI, BioXcel will source promising failed or discontinued drug candidates which it will apply to the treatment of multiple different diseases to make minute modifications to things such as the dosing pattern and all. Other Valuable AI AND ML Stocks to Invest In – Tencent (TCEHY) China’s biggest social media company, Tencent, the company behind the WeChat app, recently built a Seattle-based AI research lab headed by a two-decade veteran of Microsoft Research. Intel Corp. (INTC) While Nvidia is possibly the best example to be considered as full-on AI stock, Intel, the quintessential blue-chip semiconductor company, has taken notice of the enormous growth potential. Micron Technology (MU) Micron’s speciality is memory chips along with Samsung and SK Hynix, it controls about 80 per cent of the global market. The company is investing heavily in Machine learning which will drive great change in fields like autonomous driving, the IoT and industries like health care and finance. Twilio (TWLO) Most readers probably haven’t heard of Twilio and it’s no coincidence TWLO may have the greatest risk\/reward of the top AI companies as a result. At $3.8 billion, the cloud software business is overvalued and or shares may multiply over 8 times in the next five years. Facebook & Netflix are also good stocks to invest in as both the companies are heavily investing in AI  & ML, but how the markets respond is yet to be seen. Of course, we have listed top valuable stocks for AI and ML stock investors. But, there is no guarantee that these companies will for sure see skyrocketing share prices in 2019. The major reasons why we have mentioned these stocks here because these companies have invested heavily in machine learning and are keeping a long-term perspective in mind to gain from it. Investors who take a similar approach to AI and ML stocks keeping in view the long-term prospects will likely be rewarded as they wait for the market to grow to see significant returns.","excerpt":"AI and ML stocks are also one of the hot properties to invest in share market now. With growing influence through the applicability of emerging technologies like AI and ML in varied sectors and industries around the world. One of the major reasons why AI and ML stocks are picking up among the stockholder circles […]","categories":["AI Trends"],"tags":["AI India","AI Stocks","Amazon","augmented intelligence for smart industry","fintech AI","Google","IBM","Machine Learning","Microsoft","salesforce crm","stock market analytics","strategic salesforce"],"author_name":"Martin F.R.","publish_date":"2019-01-03T11:50:03","publication_year":"2019","word_count":1862,"keywords":["stock market analytics","AI Stocks","deep learning","R","artificial intelligence","augmented intelligence for smart industry","salesforce crm","analytics","Go","machine learning","Microsoft","AI","cloud computing","ML","Amazon","Machine Learning","IBM","virtual assistants","fintech AI","Google","AI India","strategic salesforce"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","analytics","virtual assistants","cloud computing","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/10-best-ai-ml-stocks-to-buy-in-2019\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":37157,"title":"Why Sports Analytics Is More Than Just A Numbers Game","content":"Source: https:\/\/youtu.be\/2lbgUCLAhnQ Sports over the years have become extremely competitive — today, the scenarios are so intense that a minute can completely flip the course of a game. And in order to stay one step ahead of the rivals, sports teams are opting to analytics.  And it is not only because of the competitive pressure but also because of fans base who seek detailed information about players, their performance, the reason why the team won or lost etc. Simply put, sports analytics is the practice of applying data gathered from a player’s or team’s performance under different circumstance for better decision making and improving performance. It can also be used to understand and maintain a team’s fan-base and capture the eye of investors. Sports analytics is not a very new concept, though — it has been around for quite some time. However, the recent advancements in data collection and management have expanded its role significantly. The Role Of Analytics On The Pitch In order to collect player data, analyzers use wearable devices that are attached to players jersey or the body. These wearable devices record data like heart rate, speed and acceleration of the player. The collected data is later analysed to have an idea about the players who were fit to join for the upcoming games. Furthermore, this data analytics also helps in keeping the players safe from injuries as it tracks their condition and let them know when to rest. NBA is one of the most famous basket league in the world and it is a great example when we are talking about sports analytics. Almost every team in the NBA has data analysts as full-time staff who works with coaches and front office staff in order to find out improve on-court tactics or practice habits — which move, and which shots are best suited for each player. And without a doubt, it helps teams to derive game strategies. Football is considered to be a  difficult game to analyze. There was a time when data analytics in football only include things like passing percentage and shooting efficiency, but today, teams are using analytics to analyze player movement, action away from the ball etc. These data are so valuable that they are used to create algorithms for improving individual and team behaviour, as well as decision making for substitution. One of the major football clubs, Arsenal is a good example of sports analytics. The football club has made tremendous investments in sports analytics system that tracks numerous data points per game and analyses them using an automated algorithm. From past several years, some of Europe’s most renown football clubs have been using analytics to stay ahead of their rivals, and Arsenal is one among them. Off The Pitch Sports betting over the years gained a tremendous market. The industry is booming year by year, and analytics is playing a vital role in evolution.  Today, analytics has become the lifeblood of sports betting; it is not only helping betting businesses find new opportunities but also helping them gain new insights to run their business efficiently. When a sports betting company or platform want to gain a bigger customer base, one of the best ways to succeed is to keep a track on what clients want. And by collecting player data and analyse it, one can help its clients betting on the right team or the right player. Prediction Sports prediction have recently gained momentum. Some use prediction for betting and some use prediction to better their platforms. Crictix is one such company that is working in the field of sports prediction. The company predicts cricket matches – whether its IPL or any international cricket match. However, the company is into betting; the concept of the Crictix platform is to showcase its depth in analytics and promote its data analytics platform for businesses. #IPL2019 PreMatch Prediction | Match11 – #SRHvRCB#Crictix Predicts win for #SunRisersHyderabadWinning % : #SRH 51.03% – #RCB 48.97%For latest updates keep following – @CrictixBizViz @bdb_ai#IPL12 #BDB #PredictiveAnalytics #Cricket @SunRisers @RCBTwee…https:\/\/t.co\/ioPDzCpVuC — Vineet Agarwal (@vineettweets) March 31, 2019 Wrapping Up There is no denying to the fact that analytics have transformed many businesses across the globe. And now it has marked its presence in the space of sports. Do you remember the 2011 baseball movie “Moneyball” where Brad Pitt and Jonah Hill builds a team of underrated players by taking a sophisticated sabermetric approach to scouting and analyzing players? The scenario today will be something more advanced than that. With time both sports and technology evolve, and with that, the role of analytics in sports will continue to become vital. The more competitive sports become the more teams and sports organisation will look out for analytics to better their game.","excerpt":"Sports over the years have become extremely competitive — today, the scenarios are so intense that a minute can completely flip the course of a game. And in order to stay one step ahead of the rivals, sports teams are opting to analytics.  And it is not only because of the competitive pressure but also […]","categories":["AI Features"],"tags":["Big Data","sports analytics"],"author_name":"Harshajit Sarmah","publish_date":"2019-04-01T12:08:18","publication_year":"2019","word_count":791,"keywords":["Go","programming_languages:R","AI","sports analytics","programming_languages:Go","ViT","analytics","GAN","Big Data","R","analytics platform"],"extracted_tech_keywords":["AI","analytics","R","Go","GAN","ViT","analytics platform","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/why-sports-analytics-is-more-than-just-a-numbers-game\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10095025,"title":"Generative AI Use Cases Crop Up Bountifully in Agriculture","content":"ChatGPT, driven by powerful GPT models, has found application across various sectors. OpenAI CEO Sam Altman, during his recent visit to India, mentioned that he was amazed to witness an Indian farmer using a ChatGPT-powered chatbot to access information on agricultural welfare schemes. Indeed, platforms like Kissan AI ( previously known as KissanGPT) have showcased the prowess of generative AI in agriculture. The platform leverages the power of GPT3.5 and Whisper models by OpenAI and receives 30,000-40,000 queries every month. The Indian government too acknowledges how useful generative AI can be. Reports earlier this year suggested that the Ministry of IT and Electronics(MeitY) is working on a WhatsApp chatbot by integrating ChatGPT. “LLMs combined with the natural language understanding of Indic languages can significantly enhance the ability of farmers to make informed decisions, potentially boosting yield, reducing wastage, and enhancing overall farm profitability,” Pratik Desai, who developed KissanAI, told AIM. Leveraging Generative AI In India, the agricultural sector remains one of the biggest contributors to India’s GDP and employs nearly 50% of the population of India, directly or indirectly. Today, there are numerous government schemes and benefits available for farmers, however, many of our farmers are unaware of these schemes. Navigating through the multitude of government welfare programmes in India can be challenging, especially for those who are illiterate or unfamiliar with English. Jugalbandi, an AI-chatbot powered by OpenAI’s GPT models via Azure OpenAI Service, is helping farmers as well as villagers in rural India find out about different government schemes which are beneficial for them. The chatbot, which can also be accessed through WhatsApp, retrieves pertinent programme details, typically documented in English, and delivers them in the native language of the user. Currently, the chatbot is available in 10 of the 22 official Indian languages. Similarly, Wadhwani AI, an independent, non-profit institute dedicated to developing AI solutions for social good is exploring leveraging generative AI to power Kissan call centres. Alpan Raval, chief scientist, AI\/ML, Wadhwani AI, told AIM that they are building a Kissan call centre support system using generative AI to assist farmers with their queries. “Our approach involves augmenting the expertise of the human experts by utilising models that provide automated responses based on a knowledge base. We create these knowledge bases using current government reports and documents, and our speech interface enables a conversational AI system for seamless interactions,” he said. Generative AI capabilities are being tapped into by other players in the sector as well. Digital Green, a global development organisation has joined forces with generative AI startup Gooey.AI to introduce Farmer.CHAT, an innovative solution aimed at addressing the challenges faced by farmers who are at the forefront of climate change and water security issues. Odisha’s Department for Agriculture and Farmer’s Empowerment also launched Ama KrushAI, an interactive chatbot, to provide farmers with valuable guidance on optimal agronomic practices, government schemes, and loan products offered by over 40 commercial and cooperative banks. Desai also revealed that he has been approached by over 20 companies operating within diverse agricultural sectors, both in India and internationally. “These range from early-stage agritech startups to established agri input enterprises. Currently, in our pilot phase, we’re working in close collaboration with a selected cohort of these companies.” Generative AI could do a lot more The current technological transformations and digital penetration in rural areas have made farmers aware of their crops. “However, there is still a need to acquaint them with tailor-made solutions to improve their conventional farming practices and adopt the latest methods and techniques to increase their productivity,” Sateesh Nukala, CEO & co-founder at BigHaat, told AIM. Generative AI could potentially do a lot more, given the technology is still in its nascent phase. As the technology matures, its potential use cases in agriculture will also increase. “It will become useful in numerous areas like quick detection of diseases and pest control, nutritional deficiencies that are not currently screened by existing tools.” Determining soil health to know the possible deficiencies in the current age is difficult as it requires expensive tools to generate such soil health reports. “Further, the history of crops sown on a particular land and capability to predict the upcoming quantum of harvest and plan for the coming year are some of the crucial parameters where generative AI can assist farmers in their native language,” Nukala said.","excerpt":"Digital Green, a global development organisation has joined forces with Generative AI startup Gooey.AI to introduce Farmer.CHAT","categories":["Deep Tech"],"tags":["kissanai","kissangpt"],"author_name":"Pritam Bordoloi","publish_date":"2023-06-13T14:30:00","publication_year":"2023","word_count":723,"keywords":["kissanai","Go","ChatGPT","OpenAI","AI","R","ML","RAG","Aim","generative AI","kissangpt","Azure"],"extracted_tech_keywords":["AI","ML","generative AI","ChatGPT","OpenAI","Aim","RAG","Azure","R","Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/generative-ai-already-finding-multiple-use-cases-agriculture\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":31317,"title":"Tesla’s Mad Max Autopilot Is To Driving, What Jarvis Is To Iron Man","content":"Tesla announced the ‘On-ramp to Off-ramp’ feature with the release of Autopilot 2.0 and the Enhanced Autopilot package back in 2016. This year, Tesla launched its ‘Mad Max’ autopilot mode. The company’s technology has always created immense anticipation among not just car enthusiasts, but also technology enthusiasts in general. With its new Version 9 of the autopilot feature, it has created the same level of hype. What Is Mad Max? Max Max (very likely named after a dystopian movie of the same name) is equipped with ultrasonic and radar technology. The autopilot version is a system with 8 cameras, 12 ultrasonic sensors and a forward facing radars. The camera collects data, the data being the surrounding of the car. This next-gen update of the feature identifies what lanes should a driver change to and at what time. It identifies the lane that the car is in, understands the route that the car is following and takes that route. It also guides the driver with the fastest reach, having a brief idea of the lanes, when in a slow-moving traffic. When the car reaches the exit, it will depart the freeway, slow down and transition control back to you. The version also has a feature called ‘Navigate’, which when active, gives indicators to guide you with keeping the car in the lane. For instance, a single blue line indicates the path ahead, keeping the car in your lane and grey lines are for lane changes for a more effective route, using the 8 cameras, gauging the traffic in all directions at the same instant. It sees far more than a driver could, in a way that a driver can never make makes decisions that a driver probably wouldn’t have made. Additionally, if there is another driver who is in the lane adjacent to the lane where you want to move, but this driver is in a speed faster than safe for this Tesla driver to overtake, this mode will warn and not change the lane. It is extremely difficult to make judgments in the case of human drivers. The mode uses neural nets and they are camera agnostic, because of which Tesla could use these neural nets to all the cameras around the car, providing an effective result. It reports things like dimensions, relative velocity, of the obstacles, driving behaviour, routes and mileage, and uses data, also of the surrounding cars to improve its performance. It uses different colour circles to show what’s the position of the car: stopped, moving, stationary or unknown, where the circle size represents the distance to the car. The radar is in 3D, but we have to project them in 2D. How Safe Is It? Tesla’s driverless cars have created a lot of buzz and have raised many questions about its dangers. But Tesla is not planning to do anything violating the law. Lexis Georgeson, a Tesla spokesperson had said, “We’re not getting rid of the pilot. This is about releasing the driver from tedious tasks so they can focus and provide better input.” With the description, the Mad Max mode seems like a next level autonomous system and as the system identifies the route to change the lane, it finally hands over the control to the driver before taking any action. Also, this mode requires the driver’s hand to be always on the wheel. If the hands go off the wheel, the car would decelerate and then eventually come to a halt. This is what makes the car safe. “Full autonomy is “really a software limitation: The hardware exists to create full autonomy, so it’s really about developing advanced, narrow AI for the car to operate on”, Musk had said. The autopilot firmware 7.1 had removed some self-driving features to discourage customers from engaging in risky behaviour and added Summon remote parking technology that can move the car forward and back without a driver in the car. According to a data collected after 45 million miles of Tesla cars with this feature, there was actually a whole of 50% less chance of an accident than before. Moreover, the car also takes care about maintaining pace in regard to other cars and keeps the Tesla in one lane. With all these features, the safety is definitely a lot better with a huge margin compared to non-autonomous cars and the California-based Automotive giant has proved its lordship in the tech game one more time. Autonomous Cars In A Nutshell When the autopilot tech was first made available in the Tesla Model S and Model X, it included the features of semi-autonomous drive and parking capabilities. Tesla’s Autopilot mode might disappoint you with it not being ‘autopilot’ in the real sense, but they can be called semi-autonomous, and it’s all for good reasons. This feature is to make the drive easier, not be completely irresponsible. So don’t expect to have a good nap on your way home from the office on a traffic-filled road. Tesla also intends to make a fully driverless car, considering the legal acknowledging that the legal and technical hurdles to be overcome. We have come a long way in technology, with autonomous cars being only a reality in science fiction movies, Tesla always comes up with unthinkable, unprecedented technologies for its cars. This update definitely stands out among some others. We have taken a big leap in our path for autonomous cars with this update. Only time will tell what wonders this real-life Tony Stark can present to the world.","excerpt":"Tesla announced the ‘On-ramp to Off-ramp’ feature with the release of Autopilot 2.0 and the Enhanced Autopilot package back in 2016. This year, Tesla launched its ‘Mad Max’ autopilot mode. The company’s technology has always created immense anticipation among not just car enthusiasts, but also technology enthusiasts in general. With its new Version 9 of […]","categories":["AI Trends"],"tags":["Elon Musk","Tesla"],"author_name":"Disha Misal","publish_date":"2018-12-10T14:36:56","publication_year":"2018","word_count":916,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","RAG","Elon Musk","Tesla","R"],"extracted_tech_keywords":["AI","RAG","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/teslas-mad-max-autopilot-is-to-driving-what-jarvis-is-to-iron-man\/","complexity_score":4,"technical_depth":6,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10164387,"title":"Google’s AI Video Model Veo 2 Goes Live on Freepik","content":"Google’s AI video generation model Veo 2 is now available for public use on the Freepik platform. Freepik provides AI tools to generate content and a stock image library. Today, it announced it would be the first platform to feature Google’s Veo 2. “Google partners with us to debut the most advanced AI video model. Unmatched realism, precision, and smooth animations,” said Freepik in a post on X. Google’s Veo 2 will consume 1,000 credits per generation on the Freepik platform. Moreover, Freepik announced that the first 10,000 users can run two Veo 2 generations for free. You can check out the number of credits each Freepik plan offers and its pricing here. Moreover, Freepik published a few “hand-picked” prompts and examples of videos generated from Veo 2 by its team of creators. Recently, Google announced that Veo 2 was integrated into YouTube Shorts as an upgrade to the Dream Screen feature. This feature allows users to create video clips from text-based prompts. Last December, Google announced the Veo 2 model, and the company’s product lead, Logan Kilpatrick, revealed that they will hit API by early next year. The model handles complex elements like reflections and shadows, producing clearer and sharper footage. It also includes SynthID watermarking for safety. Google’s internal testing indicates that Veo outperforms competitors such as China’s Kling, Meta’s Movie Gen, and OpenAI’s Sora in terms of quality and prompt adherence. Google’s Veo 2 has been advertised as a tool that brings advanced capabilities in terms of cinematic understanding. “Veo 2 delivers lifelike visuals with enhanced realism, reducing artefacts and improving detail. Its motion simulation accurately replicates simple and complex movements using physics,” said Google DeepMind’s Tom Hume on X. The Veo 2 supports 4K resolution and can produce videos longer than two minutes, although its experimental platform currently restricts it to 720p and eight seconds. It outperforms OpenAI’s  Sora with four times the resolution and six times the video duration.","excerpt":"The model will consume 1,000 credits per generation on the platform.","categories":["AI News"],"tags":["AI (Artificial Intelligence)"],"author_name":"Supreeth Koundinya","publish_date":"2025-02-21T17:54:31","publication_year":"2025","word_count":324,"keywords":["Replicate","Go","API","OpenAI","AI","programming_languages:R","programming_languages:Go","CLIP","GAN","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","OpenAI","R","Go","API","CLIP","GAN","Replicate","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/googles-ai-video-model-veo-2-goes-live-on-freepik\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10125414,"title":"AI will Create More Creative Jobs Than There are Today","content":"Meta CEO Mark Zuckerberg seems to firmly believe that “in the future, there will be far more creative jobs than there are today”. In a recent interview with Kallaway, Zuckerberg shared that as technology evolves, so do the AI tools we use, and staying updated with these is crucial for everyone in the creative space. He reiterated that in the future most forms of everyday work would require more imagination. In an episode of the Lex Fridman Podcast, the Meta founder and CEO said that a lot more people will work on creative stuff in the future – it would then be considered similar to traditional labour or service. However, Mira Murati, the chief technology officer of OpenAI, believes otherwise. During an interview at the AI Everywhere event at Dartmouth College, she said, “Some creative jobs may go away, but maybe they shouldn’t have been there in the first place.” Source: X How to be Creative with AI? Previously, AIM reported on a fascinating story about a 19-year-old artist selling AI-generated art in Bengaluru. The artist highlighted that his images were original, generated from scratch, and not copied from other creators or existing works. Well, this isn’t the first example of artists embracing AI to be more creative. In dance, for instance, AI-generated visuals and music interact with live performers, as seen in Cloud Gate Dance Theatre’s production ‘Waves’. Here, AI not only enhances the visual and auditory landscape but also integrates with dancers’ movements, using data collected from their physiological signals. Zuckerberg deeply believes that the future will consist of many AIs created by different people, each offering unique experiences. This belief in diversity and variety is also why he supports open-source development so strongly. Meta’s AI Studio Meta recently announced the testing of user-created AI chatbots on Instagram with a new tool called AI Studio. These AI characters, developed by various content creators, will soon appear in the US, labelled as AI to ensure user awareness. “We’re launching the first test phase with around 50 creators. We’ll gradually expand this to a small percentage of users, refining the experience along the way. “By the end of July or August, we anticipate a full rollout. It will be fascinating to see how people respond to interacting with these AI created by their favourite creators,” Zuckerberg said. He further mentioned that Meta’s goal is to build more tools that empower more people, including those who don’t see themselves as creators today. It’s like the Picasso quote: “All children are artists. The challenge is how to remain one when you grow up.” https:\/\/twitter.com\/kanekallaway\/status\/1806365032030363876 AI for All Meanwhile, authors Barry Lynn, Max von Thun, and Karina Montoya call out the dangers of monopolist AI development. They argue that tech giants like Google, Amazon, Microsoft, Meta, and Apple control the “upstream” infrastructure, which includes essential resources and technologies. Reflecting this sentiment, Zuckerberg, too, states that AI technology should not be monopolised by a single company; instead, it should be accessible to everyone. “This involves providing tools for creators and users to develop their own AIs, similar to user-generated content, and open-sourcing the technology to allow innovation and experimentation,” he said. He emphasised the need to empower a diverse range of creators and small businesses to develop their own AIs, which will create a richer, more dynamic technological landscape.","excerpt":"It’s like the Picasso quote: “All children are artists. The challenge is how to remain one when you grow up.”","categories":["AI Features"],"tags":["AI (Artificial Intelligence)","AI Tool","job interview","Meta","Superintelligence"],"author_name":"Vidyashree Srinivas","publish_date":"2024-07-01T14:20:02","publication_year":"2024","word_count":555,"keywords":["Go","Meta","OpenAI","AI","chatbots","innovation","ML","programming_languages:R","programming_languages:Go","Aim","Superintelligence","job interview","AI Tool","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","ML","OpenAI","Aim","chatbots","R","Go","innovation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/ai-will-create-more-creative-jobs-than-there-are-today\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":65552,"title":"How Small Businesses &#038; Startups Can Utilise AI To Survive The Pandemic","content":"Severe impacts have been witnessed when the nation-wide lockdown was announced due to the pandemic. This has led to the layoffs in organisations, business disruptions and other such. During this crisis, AI and machine learning have been playing a substantial role in various ways such as predicting the end of the epidemic, tracking the rise, researching on the cure and antidotes, among others. This crisis has not only pushed organisations to shift through a new set of challenges but also forced them to change their business strategies. In one of our articles, we discussed how the unprecedented nature of the health crisis has ripple effects on industries and organisations across the globe. While the world is in lockdown and organisations are struggling with their business continuity, issues such as disruption in the supply chain, remote working of employees, and the unpredictability of the crisis have created an adverse impact on the organisations regardless of their sizes. According to sources, “Startups and small businesses are facing issues such as lower demand for products and services, which resulted in salary cuts and reduction in expenses towards marketing, advertising and infrastructure.” In this crucial time, small businesses and startups have been giving more than 100 per cent to ensure that this epidemic does not hamper their businesses. Emerging technologies have the capability to solve problems and issues and help businesses in various ways, including automating work, predicting sales, detecting frauds, to name a few. In this article, we will discuss how small businesses and startups can utilise the techniques of artificial intelligence and machine learning to survive the pandemic. 1| Using Conversational AI Platforms Implementing a conversational AI platform in your business will provide you with a number of benefits amid this crisis. Unlike traditional customer support methods, where human resources were used to solve issues via messaging and communicating while consuming time and cost, conversational AI is cost-effective as well as time-efficient. According to reports, the chatbot market size is projected to grow from $2.6 billion in 2019 to $9.4 billion by 2024 at a compound annual growth rate (CAGR) of 29.7%. Over the years, several organisations have been adopting this technique to automate communications as well as create personalised customer experiences. If you have not used conversational AI platforms in your business yet, it is high time to deploy bots and enable automation to drive effective customer service. 2| AI In Logistics Operations The utilisation of AI in logistics and supply chain will result in improving the processes of logistics while reducing the costs and time. Recently, a global SaaS company, Locus has announced the launching of a free tool ‘QuickStart,’ which is a self-serve lite version of its product suite for startups and small and medium enterprises to improve their supply chain efficiency amid COVID-19 pandemic. As companies are struggling to manage volatile demands, fleet and resource efficiency, and rise in costs due to the COVID-19 pandemic, the use of AI in logistics will eventually improve and stabilise the current situation while making it more efficient and cost-effective. 3| AI In Business Analytics In this current situation, the use of AI-powered business intelligence tools has the capability to transform small businesses and startups into an automated and efficient one by reducing the operational costs, gaining more insights from data, and other such. Investing in AI for business analytics will help in making data-driven decisions, and thus, tailoring products and services accordingly. For instance, traditional business analytics models often fail to predict where to make future investment in the market, but with the help of AI, one can predict and invest accordingly. 4| AI-Based Digital Marketing Recommending products relevant to previous purchases and views have been trending among various organisations. Recommender systems play a crucial role in marketing in any organisation regardless of the sizes. Unlike traditional digital marketing techniques, using AI techniques and models will help to target the desired audience effectively to gain more benefits. For instance, utilising AI and Ml in Customer Relationship Management (CRM) will help in creating an efficient customer-centric approach to business by maximising the collection of relevant customer information. 5| AI To Hire Talents Hiring talents is one of the challenging tasks in an organisation as it can make or break an organisation. Oftentimes, startups face hiring challenges that include lack of hiring experience, inefficient assessment tools, no human resource managers, among others. This eventually hampers the growth of business and other initiatives. This issue can be mitigated by utilising AI to hire talents. AI for recruiting is an application which is designed to streamline or automate some part of the recruiting workflows, especially repetitive, high-volume tasks, such as applying machine learning to resumes to auto-screen candidates, conduct sentiment analysis on job descriptions to identify the potentially biased language, among others. This technique will help a startup and small businesses to save the recruiters time by automating the high-volume tasks, and thus, improving the quality of hiring talents.","excerpt":"Severe impacts have been witnessed when the nation-wide lockdown was announced due to the pandemic. This has led to the layoffs in organisations, business disruptions and other such. During this crisis, AI and machine learning have been playing a substantial role in various ways such as predicting the end of the epidemic, tracking the rise, […]","categories":["AI Startups"],"tags":["automated saas intelligence","business analytics mba","how to implement business intelligence","MBA Business Analytics","Startups"],"author_name":"Ambika Choudhury","publish_date":"2020-05-21T11:00:00","publication_year":"2020","word_count":824,"keywords":["how to implement business intelligence","artificial intelligence","machine learning","MBA Business Analytics","AI","sentiment analysis","ML","business analytics mba","automated saas intelligence","Git","automation","analytics","GAN","Startups","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","analytics","sentiment analysis","R","Git","GAN","automation"],"url":"https:\/\/analyticsindiamag.com\/ai-startups\/how-small-businesses-startups-can-utilise-ai-to-survive-the-pandemic\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":15305,"title":"China opens up its ‘big data valley’ at Guizhou province","content":"Img src- www.nws.cn China recently launched its first big data engineering laboratory in Guizhou province, which shows promises to perform several functions such as collecting and analyzing administrative data, offering privacy protection of data, applying big data technology to government decision making, social management, government supervision and much more. The lab located in the national high-tech industrial development zone in the provincial capital, is expected to help improve government management policy. Wang Daming, head of the Council of National Engineering Laboratory, was reportedly quoted as saying that the lab will focus on study of big data which would be dedicated to government management and also encourage the application and industrialization of big data in government management efficiency. Earlier in November 2016, the National Development and Reform Commission had approved the joint construction of the laboratory by Guizhou and China Electronics Technology Group Corporation. Big Data has seen an evident rise in China with Guizhou becoming the first pilot zone for big data. Since gaining popularity in the country, it has seen wide applications in government management, business and daily life. One such use case is the provincial high people’s court, which has set up a big data system to aid the handling of cases and facilitate public inquiries. The big data industrial map facilitated by the province demonstrates the development of companies in different areas. This could come as a reference for policy making in the country. It is worth noting that the province’s computer, communication and other other electronic manufacturing industry saw their added value reach 9.3 billion yuan in 2016, which is apparently nine times that in 2011. We had earlier covered a feature on how China has become a new entrant in the big data space and its continued interest in building and developing the infrastructure that could store and analyze big data. Not just big data, China has been tapping the AI and machine learning market as is evident from newer research papers and government policies that is flooding the province.","excerpt":"China recently launched its first big data engineering laboratory in Guizhou province, which shows promises to perform several functions such as collecting and analyzing administrative data, offering privacy protection of data, applying big data technology to government decision making, social management, government supervision and much more. The lab located in the national high-tech industrial development […]","categories":["AI News"],"tags":[],"author_name":"Srishti Deoras","publish_date":"2017-05-30T12:36:22","publication_year":"2017","word_count":335,"keywords":["big data","Go","API","machine learning","programming_languages:R","AI","programming_languages:Go","RAG","data engineering","R"],"extracted_tech_keywords":["AI","machine learning","RAG","R","Go","API","big data","data engineering","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/china-opens-big-data-valley-guizhou-province\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":17642,"title":"Meet Michelangelo, Uber’s Machine Learning SaaS Platform","content":"For the past few years, there has been a mantra that “every business is a software business.” But over the years, it has become increasingly outdated. The updated maxim is: “Every business is an AI business.” Machine-learning and AI technology are very closely related to mapping, data collection, estimating delivery times and is becoming increasingly popular among global players. One such platform that has existed for over a year but not spoken about is Michelangelo, Uber’s machine learning platform, designed to manage data, deploy models, make and monitor predictions and train and evaluate the models built. The system also supports traditional machine-learning models, time series forecasting, and deep learning. Michelangelo, an internal MLSaaS platform, democratises machine learning and makes scaling AI meet needs. Reportedly, it also enables internal teams to build, deploy, and operate machine learning solutions. Dozens of teams within the company have been building and deploying AI models through the platform. It runs across several Uber data centres, leverages specialised hardware, and serves predictions for the highest-loaded online services, according to a post co-written by Jeremy Hermann, head of the machine-learning platform at Uber, and Mike Del Balso, product manager of machine learning at Uber. How has Uber been using AI and ML? Uber has a core team providing pre-packaged machine learning algorithms ‘as-a-service’ to its team of mobile app developers, map experts and autonomous driving teams. Moreover, the company has claimed that machine learning is a part of its DNA. Let us see how Uber has been using AI and ML to bring more accuracy to its predictions. AI is not new at Uber. Uber has admitted using artificial intelligence to charge customers based on what they are likely to be willing to pay. The ride-hailing service has said that the system is based on AI and algorithms, which estimate fare rates that groups of customers will be willing to pay depending on destination, time of day, and location. Uber used machine learning techniques to bring greater personalisation into its core rider app late last year. The upgraded app starts by asking for your destination, including a number of predictions based on your habits and your current location. For example, if you are at the office it will assume you want to go home, or to the gym, or the pub. Uber also uses machine learning algorithms layered on top of their historic trip data to make more accurate estimated time of arrival (ETA) information, taking into account traffic patterns. The company has been using data from the two billion logged trips it has to ‘learn’ where good pickup spots are. How Michelangelo adds up? Michelangelo has been serving production use cases at Uber for about a year and has become the de-facto system for machine learning for engineers and data scientists, with dozens of teams building and deploying models within the company. The best use case for understanding what Michelangelo does would be UberEATS, the food ordering service launched by Uber in 2014. The MLSaaS is trained and designed to predict meal delivery time, search rankings, search autocomplete as well as restaurant rankings. Its delivery time models predict the time taken for a meal to be prepared and delivered before the order is issued and then again at each stage of the delivery process. At the core of the user experience in a meal service is the time to delivery. Initially, that was basically thought about as a classical computation. The distance between the user and the restaurant, and the average speed in the town, and then some average time to prepare the meal. In reality, Michelangelo now uses data to predict how long it takes to make noodles, how long it takes to make a hamburger, and how long it takes to deliver it in different parts of town at different times of the day. The Michelangelo platform provides the UberEATS data scientists with gradient boosted decision tree regression models to predict the end-to-end delivery time. Features for the model include information from the request (e.g., time of day, delivery location), historical features (e.g. average meal prep time for the last seven days), and near-real-time calculated features (e.g., average meal prep time for the last one hour). Models are deployed across Uber’s data centres to Michelangelo model serving containers and are invoked via network requests by the UberEATS micro services. These predictions are displayed to UberEATS customers prior to ordering from a restaurant and as their meal is being prepared and delivered The delivery model, based on machine learning, can predict how much time a meal will take to prepare and deliver before the order is issued and then again at each stage of the delivery process. The flowchart shows the predictive analysis that Michelangelo does to provide a better delivery experience. What are Michelangelo’s components? Michelangelo is built with a mix of open source systems and components made in-house. The primary open sourced components used are HDFS, Spark, Samza, Cassandra, MLLib, XGBoost, and TensorFlow. The way forward for Michelangelo Uber will undoubtedly continue to scale and harden the existing system. It will go further to work upon a higher level of tools and services to drive democratisation of machine learning and better support the needs. Going forward, these are the developments that Uber might be working upon to make Michelangelo more efficient. Auto Machine Learning: This will increase the productivity of data scientists by allowing them to specify set of labels and objective function. The system could also then make the most of the privacy-and security-aware use of Uber’s data to find the best model for the problem. Model visualisation: Uber has already made some initial steps with visualisation tools for tree-based models, but much more needs to be done to enable data scientists to understand, debug, and tune models and for better end results for users. Online learning: Michelangelo is already being trained to develop a full platform solution that is easily updateable, trains faster and sees evaluation architecture and pipelines. Scientists at the company are also working on creating a system that automates model validation and deployment, along with developing sophisticated monitoring and alerting systems. Distributed deep learning: For high-level of machine learning, it is mandatory to implement higher levels of deep learning technologies. It will not just handle larger data but also motivate distributed learning.","excerpt":"For the past few years, there has been a mantra that “every business is a software business.” But over the years, it has become increasingly outdated. The updated maxim is: “Every business is an AI business.” Machine-learning and AI technology are very closely related to mapping, data collection, estimating delivery times and is becoming increasingly […]","categories":["IT Services"],"tags":["AI (Artificial Intelligence)","Machine Learning","SaaS","Uber"],"author_name":"Priya Singh","publish_date":"2017-09-09T20:00:02","publication_year":"2017","word_count":1054,"keywords":["artificial intelligence","machine learning","AI","ML","Machine Learning","RAG","SaaS","deep learning","Aim","XGBoost","Uber","TensorFlow","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","Aim","TensorFlow","XGBoost","RAG","R"],"url":"https:\/\/analyticsindiamag.com\/it-services\/meet-michelangelo-ubers-machine-learning-saas-platform\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":24981,"title":"How Has Data Science And The Role Of Data Scientist Evolved Over The Years","content":"Until a few years ago, only a handful of us had heard of data science, which is all the rage now. According to GreenBookBlog, Danish computer scientist Peter Naur is credited for coining the term data science. Over the years, data science has evolved to become more pervasive, and we are now witnessing newbie business analysts and students faced with the prospect of building their own analysis software and methodologies. Another key trend that has been observed is how the phenomenal growth of data has pushed traditional statistical techniques to the fringes and how deep learning has revolutionised computer science. But many seem to have forgotten the basic definition of a data scientist. American mathematician and computer scientist DJ Patil defined it simply: “A data scientist is that unique blend of skills that can both unlock the insights of data and tell a fantastic story via the data.” But the frenzy and excitement around data science has become so intense that working professionals are rushing in to learn machine learning, computer vision, text mining. On the other hand, they are getting panned for missing key statistical concepts such as distribution or confidence that form the basis of data science. How The Definition And Role Of Data Scientist Evolved Jennifer Priestley, Associate Dean of The Graduate College and Professor of Statistics and Data Science at Kennesaw State University categorised organisations as digital natives and non-natives. For example, for big tech firms like Amazon, Facebook, Airbnb and Google, data is the foundation. Most of the computational and analytical innovations come out of big tech companies. These companies have a deep bench of data science talent, and this structure works well because here, one can attract and recruit data science teams. Currently, there is a central data science team in most organisations, but analysts believe that as time progresses, each business unit will have dedicated data science member. Companies who have modelled themselves as data science organisations are Netflix and Airbnb. Statistics And Data Science Are Intertwined: As the field matures, the role of data scientist will evolve. One of the definitions being bandied around is that data scientists are expert in statistics. But it may not be the case with the current lot which has gravitated from the engineering field. We have often heard that data science can’t be more than statistics. Sean Owen, director of Data Science at Cloudera noted that statistics and numerical computing have been connected for decades, and as in all areas of computing, we always crave ways to analyse a little more data. According to John Tukey’s paper The Future of Data Analysis, statistics must become concerned with the handling and processing of data, its size, and visualisation. However, today a lot of people from diverse background, even economics, claim to be data scientists. Challenges Involved With Working Noisy Datasets: The current focus of organisations is on utilising big data which builds up analysis solutions to fulfil customer objectives. But the actual substance of what data scientists do, remains ambiguous. For example, a research cites that increasingly, scientists are faced with the challenge of working with large, heterogeneous, and noisy datasets. Most new entrants have little or no experience with cutting-edge data science techniques and technologies. These individuals have to find opportunities to bridge beyond their current skill set and current disciplinary approaches. Companies Transition From Data-Poor To Data-Rich: As companies transition from data-poor organisation to data-rich, wide experience and a thorough background in both data science and pure sciences will be required. With institutes rushing to bridge the gap and aligning curriculum with current industry demand, the supply gap will gradually decrease. However, as people in their late 20s, 30s and even 40s look to pivot towards a career in data science, they should significantly beef up on critical, applied learning and get real hands-on experience. One cannot become a data analyst with just one analytics track or online certification, one needs to beef up strong applied statistics program. Hands-on experience can go a long way in clearing the most difficult concepts of data science. The Last Word According to a Gartner research, by 2020 more than 40 percent of data science tasks will be automated. For example, data preparation, for the most part, will be automated. In fact, the research also laid out a few defined instances where some of the data science-related tasks could become completely automated — automated selection and tuning. The tasks that will become the core skill set in the future are Feature engineering and model validation Understanding of the domain Machine learning Eventually, there will be more focus on parallel and distributed code as professionals who were reliant on spreadsheet analysis will shift to Python and R.","excerpt":"Until a few years ago, only a handful of us had heard of data science, which is all the rage now. According to GreenBookBlog, Danish computer scientist Peter Naur is credited for coining the term data science. Over the years, data science has evolved to become more pervasive, and we are now witnessing newbie business […]","categories":["IT Services"],"tags":["Data Science"],"author_name":"Richa Bhatia","publish_date":"2018-05-30T04:23:23","publication_year":"2018","word_count":787,"keywords":["data science","machine learning","AI","computer vision","RAG","Python","Aim","deep learning","analytics","Data Science","R"],"extracted_tech_keywords":["AI","machine learning","deep learning","computer vision","data science","analytics","Aim","RAG","Python","R"],"url":"https:\/\/analyticsindiamag.com\/it-services\/how-has-data-science-and-the-role-of-data-scientist-evolved-over-the-years\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10166140,"title":"Did ByteDance Just Create a React Native Killer?","content":"ByteDance has been making headlines for developing AI models and taking on a carefully crafted image of a developer-focused platform. In line with this, the parent company of TikTok has now introduced an open-source Rust-based JavaScript framework called Lynx. Lynx helps build cross-platform mobile and web applications. ByteDance seems to have been using Lynx internally for apps like TikTok and has now open-sourced it. The internet is abuzz with discussions on Lynx, which is seen as an alternative to Meta’s React Native, its potential, and how it tries to overcome the problems of the existing frameworks. So, What is Lynx? Xuan Huang, the architect of Lynx, calls it a family of technologies that empower developers to use their existing web skills to create native UIs for mobile and web from a single codebase. The core engine of the Lynx framework is framework-agnostic, as well as agnostic to host platforms and rendering backends. In an official blog post, he explained that Lynx is designed for diverse use cases and rich interactivity, which enables it to provide engaging UIs for large-scale apps like TikTok. Huang mentioned that Lynx powers TikTok Studio, e-commerce storefronts like Shop, and high-profile events such as Disney100 and The Met Gala on TikTok. In a nutshell, TikTok has extensively used it. What Does Lynx Aim to Solve? Lynx aims to provide a platform for developers to ship their apps with a great native experience while eliminating lag. “A blank screen, a 0.1s lag in a “like” animation, or an unfamiliar UI pattern can make an interface feel “cheap” or untrustworthy. We believe that native primitives and responsiveness aren’t just nice-to-haves—native is a necessity,” Huang wrote. He emphasises that even with the growing app economy, developers still face challenges in delivering experiences at scale and velocity. Lynx tries to solve this by enabling developers to build once and reach more platforms. A notable architectural decision of Lynx is its statically enforced division of user scripting into two distinct runtimes: a main-thread runtime, powered by PrimJS, a custom JavaScript engine optimised for Lynx, and a background runtime for user code as the default. This enables it to render frames fast and have responsive interfaces. Is It a Better Alternative to React Native? Experts seem to be impressed by Lynx. Rajesh Sahoo, senior software engineer at Tikkl, told AIM, “Lynx is still in its early days, whereas React Native is currently more stable and boasts a massive community. However, React Native has some limitations, such as relying heavily on community-maintained libraries for native functionality on Android and iOS.” He added that Lynx could be a game-changer for its easier access to native functionality and claims about its performance optimisation. “At the end of the day, I don’t care whether it’s Lynx, React Native, or another platform. I just hope whoever can tackle these issues, does it in a way that makes my life as a developer easier. I’ll be happy to use whatever platform gets me there,” Sahoo expressed. Theo Browne, YouTuber and founder of T3 Chat, elaborated on the problem with React Native on his YouTube channel. He explained that the way React Native works is not the most efficient, and a user may experience a stutter with the UI updates in an app. While React Native has tried to improve over the years, Lynx has introduced a fundamentally different approach with two threads—the UI thread, and the framework thread. This enables heavy data processing while updating at 60 FPS with no lag. “I see so much potential in what they’ve built here,” Browne added. Zack Jackson, infrastructure architect at ByteDance, highlighted the difference on X: “Lynx: Emphasises a multithreaded engine designed to achieve “instant launch and silky UI responsiveness. React Native: Relies on a single-threaded JavaScript bridge to communicate between JavaScript and native code, which can become a bottleneck for performance-intensive apps.” He added, “While React Native has made strides with its new architecture (e.g., Fabric renderer), it doesn’t inherently emphasise multithreading in the same way Lynx does.” A thread on React Native’s Subreddit highlights reactions from developers. One user said, “So it’s using react renderer but it renders to different base UI than React Native. An alternative renderer essentially, with common bits being react but more importantly allowing you to use something other than react to make native apps (sic).” The developer added, “Competition and innovation is cool!”. Another developer wrote, “This seems to finally answer the “write once, run everywhere, react based & dom rendering native platform”. While the general sentiment seems positive, considering it powers an app like TikTok, it is still in the early days for the framework to overshadow React Native. However, a developer wrote on X, “‘I’m excited for Lynx and I think React really needs a kick in the pants.”","excerpt":"TikTok’s parent, ByteDance, transitions to the developer side of things.","categories":["AI Features"],"tags":["ByteDance","react-native"],"author_name":"Ankush Das","publish_date":"2025-03-17T15:00:00","publication_year":"2025","word_count":798,"keywords":["ByteDance","Rust","programming_languages:R","AI","innovation","Aim","ViT","react-native","JavaScript","R","Java","programming_languages:JavaScript"],"extracted_tech_keywords":["AI","Aim","R","JavaScript","Rust","Java","ViT","innovation","programming_languages:R","programming_languages:JavaScript"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/did-bytedance-just-create-a-react-native-killer\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10079115,"title":"Good Good Piggy’s Mantra for Good Financial Literacy","content":"‘Catch ‘em young’ is a good mantra to abide by when it comes to teaching your kids a trick or two about financial planning. Childhood experiences form the core of our personalities and learning how to handle finances can be counted as a life-essential skill, one that your kids would be grateful to have acquired early on. So, how about a technology that teaches the basics of financing to children as young as five? Analytics India Magazine interacted with Purva Aggarwal, founder & CEO of Good Good Piggy. Aggarwal talks about her journey, recounting how she established the first fin-edtech startup in India. Aggarwal is the first Indian-based female solopreneur ranked third globally to enter the space of ‘children’s online banking & wallet’.  Founded in 2021, Good Good Piggy is an online piggy bank and behavioural rewards platform that acts as an investment channel for young children to form habits of money management. The platform applies a few objectives from OECD\/INFE, impacting the downtrend of financial literacy by setting early benchmarks. The venture came up in Shark Tank India on Sony Liv, while Aggarwal was featured in Japan’s Start-up Review as ‘The Next Leader 2021’. Prior to this, Aggarwal worked as a financial analyst at Mayfair Equity Advisory Ltd. She was also a writer with a leading publication for institutional investors. She is the ambassador of the Indian Chapter of Women in Tech – a global movement by an international non-profit headquartered in Paris, dedicated to promoting diversity in the tech sector worldwide. Read the excerpts from our interview with Purva Aggarwal below. AIM: What’s it like to be the founder and CEO of Good Good Piggy? Every day is different and comes with new learning every day, bringing new challenges and problems for you to contend with. It’s bringing everyone’s efforts together and remembering the bigger picture and of course, fighting little fires every single day. Though I can’t be mad – it’s a good problem to have. We’ve been focusing on our go-to-market strategy and the public roll-out of the app as of late. It’s exciting to see what was once just an idea in your head turn into a full-fledged company, but it is also an exhausting rollercoaster. AIM: Elaborate on your role in trying to accomplish goals for the company. My role is to balance the bigger picture with the little details. The role of a founder is never-ending really, but that is my favourite part. You are always on multiple learning paths because each part is too important to set the foundation right. There are always many different areas that need your full attention, all at once – operations, hiring, planning, funding – all as immediate as the other (laughs). AIM: Could you give us more details on Good Good Piggy? Good Good Piggy is a digital piggy bank and habit builder for pre-teenagers, focusing on a D2C approach. We have started curating our special communities for different types of parents so we can identify and cater to their unique needs. Our first community event was held on October 16 at India Habitat Centre for Mompreneurs (moms+entrepreneurs). It was a huge success and we are glad moms and kids enjoyed themselves. Digital penetration in our lives has grown manifolds during the pandemic. About 86 percent of children as young as one year old have used a digital device. Now kids in primary school are attending school on video calls and can understand that technology better than we do sometimes. Therefore, the urge for digital exploration is already eminent. Children are a little too click-friendly now. They know how to handle devices and there’s no going back from this trend. In this tech- driven click-happy generation, attention spans are becoming smaller and instant gratification is too prevalent. AIM: What are the technologies driving the company and how do you make it kid-friendly? We majorly use React JS, React Native, MongoDB and a bunch of integrations in a modular environment to make development progressive and scalable. Good Good Piggy’s kid interface is designed to look like a game so it keeps their attention and is easy to use. Why not engage their screen-time to teach them skills that will set them on the right path? It is also completely customisable and entirely controlled by the parent because we understand each parent and child is unique and there are many different parenting styles. Everything needs to be parent-approved and is designed to be completely safe. Further, there is a digital lock-out feature that parents can enable to prevent excessive screen-time for their kids. The video garage makes sure that the kid is only watching content fed into the app by the parent and doesn’t go clicking onto other unapproved videos. AIM: How has your work evolved over time, especially in the post-pandemic era? Good Good Piggy was a pandemic baby, so it is part of our DNA. The post-pandemic era has changed how teams work with geography no longer being a constraint. It makes the talent pool much wider and allows different perspectives to be part of the process. Creativity has definitely blossomed new solutions for virtual teamwork to happen effortlessly. AIM: What are some of the learnings that come with being in your shoes? My biggest learning is that you can learn anything. I’ve learnt so much in the past year that I probably never would’ve bothered to if it weren’t for Good Good Piggy. There are also many macro trends we have studied in trying to figure out how to create for Generation Alpha and beyond. The environment is changing so rapidly that the development of the mind has changed in some ways as well and it is important for caregivers of the next generation to realise that we need to adapt teaching to accommodate the tech disruption. AIM: Did you encounter the ‘glass ceiling’ in your career, or is it imaginary? Of course I did! Banking and entrepreneurship are both male-dominated sectors. I’m the first female solopreneur in India and the third globally in the children’s online banking and wallet space. The startup bubble has definitely improved in terms of women’s share but there is still a long way to go. I’m glad to be part of that journey and hope to make the path better for other women moving forward. AIM: Being the Ambassador of the India Chapter of Women in Tech, how do you think organisations address the need for diversity in the field of technology? Our goal is to encourage women in the technology sector and reduce gender parity. We encourage women to keep going and become part of the boardroom and for companies to be more inclusive. We also assist women joining the workforce after maternity leave\/break, which is still an issue for many women. I believe true change comes when there is diversity at every level of an organisation. This means institutional funding for female leaders, training, ease in resuming work after a break and of course, equal considerations. We want companies to be more inclusive and create their policies with all stakeholders in mind – not just follow the standard. Take 5 Top apps you frequently useCanva, Stripo, Google Scholar, MS Clarity Favourite books\/ podcastPodcast called Capitalisn’tMarketing 5.0 by Philip KotlerMind Change by Susan Greenfield What would your alternative career be?If not for my venture, I would probably still be working in the private equity sector but you never know – after all, Good Good Piggy was a surprise twist. How do you define your leadership style?I like to encourage independence. I believe people need space to be creative and come up with ideas. We definitely encourage learning all that you can – even for our senior members. I believe learning from each other is important for the success of any team. Your advice for women on a similar journey.Getting the business environment correct is important. Employees, advisors and consultants – the right guidance and the right criticism is key for success.","excerpt":"The venture was featured in Shark Tank India on Sony Liv and Japan’s Start-up Review as The Next Leader 2021","categories":["AI Features"],"tags":["Interviews and Discussions"],"author_name":"Bhuvana Kamath","publish_date":"2022-11-09T11:00:00","publication_year":"2022","word_count":1333,"keywords":["Go","API","AI","MongoDB","Scala","Git","RAG","Aim","analytics","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","analytics","Aim","RAG","MongoDB","R","Go","Scala","Git","API"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/good-good-piggys-mantra-for-good-financial-literacy\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10101144,"title":"How Oracle Plans to Kill Attrition in Indian IT","content":"Oracle has cracked the code on how to get the best out of generative AI with its billion-dollar baby Cohere by its side. One of the key areas where Oracle is experimenting with generative AI is human resources. To streamline the HR department’s workflow, Oracle, earlier this year, introduced generative AI-powered features in Oracle Fusion Cloud Human Capital Management (HCM). With HCM, HRs would be able to make decisions based on employee data. “Data in a system can now be utilised not only by managers and the organisation but also by the workers, making them feel valued,” said Chris Havrilla, vice president of product strategy Oracle, in an interaction with AIM at the Oracle CloudWorld 2023, Las Vegas. With Oracle HCM, every employee will have access to upcoming opportunities within the company, leaving nothing hidden. For example, they will know which skills they must acquire to master any role. These recommendations will be based on employee data. Havrilla believes that “democratising” data combined with capabilities of HCM would allow managers to uplift workers. Havrilla said that in any organisation, there are various types of personalities. Some individuals might be outspoken, while others may be quiet, but hardworking. Through the HCM tool, Oracle’s aim is to bring all employees on the same platform so that all of them feel valued within the company. Furthermore, Havrilla shed light on the problem of talent retention within the company. She believes it is not often because of money that employees change their jobs frequently. “What we’re seeing from the data is that things that drive employees are feeling valued by their peers\/organisation and the notion of belonging and growth,” she added. Oracle to solve the attrition problem In the recent past, Indian IT companies – the likes of Wipro, TCS, Infosys and Tech Mahindra – which also happen to be the customers of Oracle Fusion Cloud, have been dealing with high attrition rates, which now seems to be flattening to an extent. Now, the average attrition rate for Indian IT companies stands at 20%. This means that one out of every five employees leaves their job every year. A factor that contributes to the high attrition rate in Indian IT companies is the lack of opportunities for growth and development, alongside stagnant work culture. Havrilla pointed out that this happens because organisations struggle to understand their employees’ skills, predict future needs, and fill skill gaps. Similarly, employees find it hard to adapt to changing business needs because they often aren’t aware of growth opportunities due to lack of transparency. “Everybody wants employees to be equipped with all the skills. We have to become more of a development marketplace,” she said, adding that (employers’) focus should be on bettering employees, rather than succumbing to hiring and layoffs. To tackle this issue, earlier this year Oracle introduced ‘personalised coach’, Grow. It helps employees to identify and achieve their career goals, stay up-to-date on the latest skills and technologies, and network with other professionals in their field. In other words, increases retention for companies. Citing Gen Z, who frequently switch jobs in search of better opportunities, Havrilla said that Oracle Grow will help enterprises retain Gen Z employees in the future as they will be able to figure out what suits them best. “Grow is that personalised coach that is going through all the data, examining every single micro-action that happens in the organisation, updating in real-time, and understanding what impact that could have. It surfaces opportunities, connections, gigs, projects, learning, and all the different things,” explained Havrilla, pointing at the Great Resignation. Havrilla said that during the pandemic, individuals were figuring out what suited them best as organisations weren’t able to do it for them. She noted that even today, resignations continue to happen. By providing employees with access to a wide range of learning resources and opportunities to connect with others, Oracle Grow can help employees develop their skills and knowledge and advance their careers. Plus, it can help organisations build a skilled and motivated workforce. Towards Advisory Role For quite some time now, the HR role has been mostly around managing the human resource and running around for employees’ feedback, but now, it’s transitioning towards mentorship and advisory roles, thanks to generative AI. Havrilla believes that Oracle’s new offerings will enable HRs to become advisors who would further be able to assist the organisations in assessing the employee’s potential based on their performance data. “We really need HRs to be the customer advocate for those employees, like red light, advocacy, advisory – all of these things,” said Havrilla. She further added that the process of performance and talent review would be automated with the help of HCM. “We can take off that burden of going out and collecting all this data for managers to create and do performance and talent reviews and let the system do that,” she added. Their new role involves assisting employees in realising their full potential. With access to their data, HRs can now suggest specific growth opportunities for employees to pursue. “If you’re going to be an expert on anything, be an expert on human behaviour. Be an expert in helping your leaders build great teams,” she concluded.","excerpt":"With Oracle Grow, employees can stay up-to-date on the latest skills and technologies","categories":["AI Features"],"tags":["Interviews and Discussions","Oracle"],"author_name":"Siddharth Jindal","publish_date":"2023-10-05T14:20:11","publication_year":"2023","word_count":869,"keywords":["Go","API","programming_languages:R","AI","ML","Oracle","RAG","Aim","generative AI","GAN","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","ML","generative AI","Aim","RAG","R","Go","API","GAN","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-oracle-aims-to-kill-attrition-in-indian-it\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10165163,"title":"Keysight, Samsung, and NVIDIA Collaborate on AI-Optimised RAN Technology","content":"US-based Keysight Technologies has partnered with Samsung and NVIDIA to develop AI models to improve radio access network (RAN) performance for 5G-advanced and 6G technologies. This collaboration enables Samsung to integrate these AI models into its virtual RAN (vRAN) software, which will be showcased at the Mobile World Congress (MWC) 2025 in Barcelona. Developed under the AI-RAN Alliance, the initiative addresses key challenges in traditional RAN systems, such as limited throughput, high latency, and inefficient resource utilisation. This development comes right after the company’s recent collaboration with Hyundai to complete what is claimed to be the industry’s first end-to-end reduced capability (RedCap) trial over a private 5G network. Samsung will also unveil its first Android XR headset, Project Moohan, by integrating multimodal AI with advanced XR capabilities. Charlie Zhang, senior VP at Samsung Research America, stated, “This collaboration marks a significant milestone as we move toward AI-native and sustainable telecommunication networks.” AI-based channel estimation models have demonstrated a 30% improvement in cell edge throughput during lab tests compared to conventional methods. These advancements optimise resource allocation, enhance system capacity, and reduce power consumption. Samsung’s AI model evaluation was conducted using an end-to-end setup featuring its radio point and distributed unit (DU) on NVIDIA’s AI Aerial platform powered by the GH200 Grace Hopper Superchip. Keysight’s Channel Emulation Solutions enabled precise testing under various conditions. Giampaolo Tardioli, GM and VP at Keysight, highlighted the transformative potential of this partnership in advancing energy-efficient networks. Soma Velayutham, GM for AI, 5G and telecoms at NVIDIA, also emphasised the flexibility and efficiency gains unlocked by AI in signal processing pipelines. The collaboration is expected to drive innovation and accelerate the global adoption of AI-enhanced RAN technologies. At MWC 2025, Jio Platforms Limited (JPL), along with AMD, Cisco, and Nokia, also announced plans to develop an Open Telecom AI Platform. It offers a ‘multi-domain intelligence’ framework to introduce AI and automation to ‘every layer’ of network operations. Even AWS has entered the game with its new cloud solutions to boost 5G networks for telecom operators. At the same event, SynaXG (a provider of AI-RAN solutions), in collaboration with Kyocera (Japanese manufacturer), and NVIDIA claimed to have unveiled the world’s first software-defined mmWave 5G vRAN solution.","excerpt":"This enables Samsung to integrate these AI models into its virtual RAN (vRAN) software.","categories":["AI News"],"tags":["5G","AI-RAN","NVIDIA","Samsung"],"author_name":"Sanjana Gupta","publish_date":"2025-03-05T12:38:29","publication_year":"2025","word_count":369,"keywords":["5G","Samsung","AWS","AI","Modal","innovation","cloud_platforms:AWS","programming_languages:R","automation","Aim","AI-RAN","multimodal AI","NVIDIA","R"],"extracted_tech_keywords":["AI","multimodal AI","Aim","AWS","R","automation","innovation","Modal","cloud_platforms:AWS","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/keysight-samsung-and-nvidia-collaborate-on-ai-optimised-ran-technology\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10129951,"title":"The New AI Talent Hubs are the Tier 2 &amp; 3 Cities in India","content":"Big companies often shift their focus to smaller towns in search of outsourcing opportunities and more affordable labour. Not anymore. Now, what they are also looking for is better talent in smaller towns as they look to expand their operations. According to Randstad’s India Talent Insights Report 2024, cities like Jaipur, Vadodara, Indore, Ludhiana, Chandigarh, Lucknow, Bhopal, Surat, Mohali, Kochi, and Coimbatore have emerged as key job hotspots over the past six months. The report indicates a 25-35% rise in hiring in Tier 2 and 3 cities from 2023 to 2024, particularly in sectors such as retail, consumer goods, banking and financial services, energy and utilities, pharma and healthcare, and manufacturing, with AI also taking a big part of the chunk. These companies include Bajaj Finserv, IBM, Accenture, Infosys, Genpact, Vodafone, Airtel, and many banks. Similarly, several data centre companies are also expanding their bases to Tier 2 and 3 cities in search of talent. Saandeep Dandekarr, senior director at NTT Global Data Centers and Cloud Infrastructure, told AIM that the move would also add a bunch of jobs, which he categorised as the construction phase that would see a temporary spike in employment. Then there’s the operational phase promising more sustainable job opportunities, albeit with a smaller workforce due to advancements in automation and AI. A few weeks back, IBM India HR head Thirukkumaran Nagarjan said that they are actively hiring local talent. “Instead of hiring talent from smaller cities and placing them in offices in metros, we are taking our offices closer to talent pools across India,” he said. Speaking with AIM, Celebal Technologies’ CEO Anirudh Kala explained why it was a big deal that Microsoft partnered with them as a company sitting in Jaipur. “There is no dearth of talent in this country…there are people who didn’t get a chance,” said Kala. He added that the company acquired a lot of talent from Tier 2 and even Tier 5 towns in India and gave them the nurturing and platform needed to grow. Is the Talent Really Present? This comes at a time when the IT sector has seen a decline in hiring across all regions, with hundreds of thousands of jobs vacant. Many tech enthusiasts who planned on moving to Bangalore, Mumbai, and other cities for IT jobs are choosing to remain in their towns as the opportunities back home increase. And a few have been relocating from bigger cities to their hometowns for the same reason. Furthermore, as upskilling has been on every tech company’s monthly quota, most are focusing on training their existing employees instead of hiring new ones, which also helps cut costs. Viswanath PS, the CEO of Randstad India, listed out the many reasons that put smaller cities at an advantage. These include: an expanding pool of employable talent, substantial infrastructure investments, good educational institutions, lower labour and operational costs, favourable government policies. Besides, rising disposable incomes are boosting e-commerce, and several companies are planning to set up manufacturing units in remote areas. Despite this, the need for specialised skills is crucial to land a job in this field. Data annotation, one of the jobs that was majorly outsourced to smaller cities, is also in demand, creating more white collar jobs. Setting up shop in small cities also costs less as the cost of living for employees there is way lesser. Meanwhile, there are numerous AI-ML courses coming up in the market which are easily accessible for job seekers in these cities, enabling them to upskill faster. “These programmes can lead to a salary increase of 20-30 per cent in metro cities and 15-20 percent in Tier 2 and 3 cities,” said Sachin Alug, the CEO of NLB Services. For tech companies, these are untapped fresh skills that they are actively looking for. According to the India Skills Report 2024, the manufacturing sector is projected to generate 5 million AI-related jobs by 2030, while the retail industry might see a loss of 850,000 jobs by 2025 but is expected to gain 4 million new AI roles. This emphasises the importance of continuous learning and upskilling. Campus Placements are One of the Reasons Vinayak Gupta, the head of placement and outreach at JECRC University in Jaipur, said that one of the reasons for this rise is the increase in placements in Tier 2 universities. “As companies expand their recruitment efforts beyond major metropolitan areas, they unlock a wealth of untapped talent, driving both economic development and corporate growth,” said Gupta, while adding that these placements directly contribute to lowering unemployment rates in the cities. Moreover, there is a surplus of graduates coming out of these smaller towns that need employment. But since IT companies are not actively hiring people citing skills as one of the reasons, there is definitely the need for graduates to upskill themselves and apply for jobs in their own cities. Since 2022, Indian IT firms have been expanding to smaller cities in a bid to attract more talent. Infosys said that the company would open new offices in four tier 2 cities in 2022. Similar announcements were made by Accenture and IBM. Interestingly, this is also at the time when Bangalore, the Silicon Valley of India, is toying with the idea of making its employees work 14 hours a day at the same pay.","excerpt":"This comes at a time when the IT sector has seen a decline in hiring across all regions, with hundreds of thousands of jobs vacant.","categories":["AI Trends"],"tags":["AI adoption India","AI career","Career"],"author_name":"Mohit Pandey","publish_date":"2024-07-23T16:01:01","publication_year":"2024","word_count":884,"keywords":["Go","AI career","programming_languages:R","AI","ML","programming_languages:Go","automation","AI adoption India","Aim","R","Career"],"extracted_tech_keywords":["AI","ML","Aim","R","Go","automation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/tier-2-3-cities-in-india-are-the-new-ai-talent-hub\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10057599,"title":"An Illustrative Guide to Masked Image Modelling","content":"In machine learning, nowadays, we can see that the models and techniques of one domain can perform tasks of other domains. For example, models focused on natural language processing can also perform a few tasks related to computer vision. In this article, we will discuss such a technique that is transferable from NLP to computer vision. When applying it to the computer vision tasks, we can call it Masked Image Modelling. We will try to understand the working of this technique along with its important applications. The Major points to be discussed in this article are listed below. Table of Contents What is Masked Image Modelling?The framework of Masked Image ModellingWorks Related to Masked Image ModellingApplications of Masked Image Modelling Let’s begin the discussion by understanding what mask image modelling is. What is Masked Image modelling? In machine learning, masked signal learning is a type of learning where the masked portion of the input is used to learn and predict the masked signal. We can find the use cases of this type of learning in NLP for self-supervised learning. In many works, we can see the use of masked signal modelling for learning from huge unannotated data. While talking about the computer vision task, this approach can also provide competitive results to the other approaches like contrastive learning. Performing computer vision tasks using masked images can be called masked image modelling. Applying masked image modelling can have the following difficulties: Pixels near to each other are highly correlated.Signals under the images are raw and low level in comparison to the signal (tokens) under the NLP data.Signals in image data are continuous while text signals are discrete. So applying this approach to image or computer vision-related data, requires the procedure to be accomplished very well so that correlation can be avoided. Prediction from the low-level signals can be used for high-level visual tasks and the approach can adapt the continuous signal behaviour. We can witness various works which have modelled image data to generalize these difficulties like: Pre-Trained Image Processing Transformer: This work shows the adoption of continuous signal from image for classification tasks using the colour clustering techniques in addition. Swin Transformer V2: This work represents the technique for scaling a Swin transformer up to 3 billion parameters and making it capable of learning and performing computer vision tasks with images up to 1536 x 1536 resolution. They have applied the adaptation techniques for continuous signals from images using models.BEiT: BERT Pre-Training of Image Transformers: This work can be considered as using the BERT model in computer vision where we can witness a similar technique of tokenization using an additional network for image data and its block-wise image masking breaks the short-range connection between the pixels. After going through the above works, we can see a few examples of techniques that can be used to solve the difficulties. We can understand the level of complexity to design a model or framework which can handle these difficulties and can perform the required task. If we talk about the basic idea behind the model or framework for masked image modelling, it can be represented by the following images : Image source In the above image, we can see the input patches of images, with a linear layer to perform regression on pixel values of the masked area under loss. The design and insights of a simple model can consist of the following: Masking applied to the imagesModel for raw pixel regressionLightweight prediction head By applying simple masking to the images, we can make the process simple for a transformer. The regression task aligns well with the continuous nature of visual signals, and a lightweight prediction head should have the property of bringing a remarkable speedup in pre-training. Heavier heads have the capability of a stronger generation but can lead to a loss in the downstream fine-tuning tasks. The Framework of Masked Image Modelling We could understand that the motive of these procedures is to learn representation using the masked image modelling, in which the procedure should be capable of masking a portion of an image signal and predicting the original signals at the masked area. A framework to complete the motivation can have the following components: Masking strategy: This component should be designed for selecting the area to mask and to perform the masking on the selected area so that the masked image can be used as an input.Encoder architecture: This component should be able to extract latent feature representation for the masked image and use the extracted representation to predict the original signals at the masked area. If using transformers as encoders, then it is expected from the encoder that it should be capable of performing a variety of computer vision tasks. Some examples of transformers in computer vision are vanilla ViT and Swim transformers.Prediction head: This component should be capable of producing one form of original signals at the masked area of the image when applied to the latent feature representation learned by the encoder. Prediction target: This component should be capable of defining the form of the prediction from original signals and loss type. Talking about the prediction from it can be either raw pixels or a transformation of the raw pixels. The form of loss type can be a cross-entropy classification or L1 and L2 regression loss. For image masking, we can use a variety of strategies of image masking like square shape masking, block-wise masking, random masking, etc. The below image is a representation of the different types of image masking. Image source In the above section, we have seen a basic architecture of the framework for masked image modelling and the components using which we can make the framework perform computer vision tasks using masked image modelling. Let’s see some of the examples where we can witness the masked image modelling. Works Related to Masked Image modelling Some of the important works related to masked image modelling are as follows: Generative Pretraining from Pixels: In this work, we can see an example of a trained sequence transformer to predict pixels using an autoregressive approach. In this work, a GPT-2 model is used to learn strong image representations. This work has achieved 96.3% accuracy with a linear probe, outperforming a supervised Wide ResNet, and 99.0% accuracy with full fine-tuning, matching the top supervised pre-trained models.Unsupervised Visual Representation Learning by Context Prediction: In this work, we can see the use of spatial context from the image as a source of visual signals which can be used for training a visual representation. While extracting random pairs of patches from images, they have trained a CNN to predict the place of the second patch correlated to the first. This work is a representation of unsupervised visual discovery by learning feature representations within images. Also, learning feature representations within images helps in capturing the visual similarity across images.Selfie: Self-supervised Pre Training for Image Embedding: This work generalizes the concept of masked language modelling of BERT to image data. Using the masked image, the method learns to select the correct patch of the image, among other patches sampled from the same image. This work on ImageNet 224 x 224 with 60 examples per class (5%), improves the mean accuracy of ResNet-50 from 35.6% to 46.7%, an improvement of 11.1 points in absolute accuracy.SimMIM: A Simple Framework for Masked Image modelling: SimMIM is a very simple framework for masked image modelling. This framework is an example of applying a very light prediction head that can be compared to the linear layers. Using ViT-B, this approach achieves 83.8% top-1 fine-tuning accuracy on the ImageNet-1K dataset. By pre-training also on this dataset using SwinV2-H, it achieves 87.1% accuracy using only ImageNet-1K data.Masked image modelling with Autoencoders: This is an example of masked image modelling given by Keras where we can find a simple and effective method to pre-train large vision models like ViT. This method gets inspiration from the pre-training algorithm of BERT. Using this example we can learn how to patch images and predict by learning from extracted patches of images. This can be considered as an implementation of a masked autoencoder for self-supervised pre-training with the CIFAR-10 data. Applications of Masked Image Modelling In the above sections, we have seen how we can perform the masked image modelling and what are the different works related to this technique. This technique can be used in various computer vision applications. Performing image masking helps transformers and autoencoders to learn easily using only required information from the images. Masking can speed up the transformer to perform classification tasks using images. Also, masking images is a process of creating an image piece from a larger image and also we can use it to modify a larger image. It is a process that is underneath many types of image processing like edge detection, motion detection, and noise reduction. Mainly, we can say that this technique can be used in self-supervised learning in computer vision. Masked images are easy to learn because of the low and important information in masked images. Due to high-level unannotated data creating confusion for the model, image masking can be considered as a process of converting high dimensional data to a lower dimension. Final Words In this article, we have discussed masked image modelling which is developed by taking inspiration from masked language modelling. Here, the using the  is donemasked information and we can be more capable in learning from visual signals of the images. Along with this, we have seen how we can perform this technique and what are the works related to this technique. In the end, we also discussed some key applications of mask image modelling.","excerpt":"masked image modelling can provide competitive results to the other approaches like contrastive learning. Performing computer vision tasks using masked images can be called masked image modelling.","categories":["Deep Tech"],"tags":["computer science projects","Computer Vision","computer vision applications","computer vision dataset","computer vision future","computer vision neural network","Image Classification","image processing","image recognition","image reconstruction","image segmentation","image synthesis","ImageNet"],"author_name":"Yugesh Verma","publish_date":"2022-01-04T15:00:00","publication_year":"2022","word_count":1615,"keywords":["computer vision applications","computer vision","BERT","R","image reconstruction","image recognition","NLP","Computer Vision","computer science projects","computer vision future","ImageNet","Image Classification","Go","machine learning","AI","image processing","image synthesis","computer vision neural network","computer vision dataset","Keras","image segmentation","Transformers","GPT"],"extracted_tech_keywords":["AI","machine learning","NLP","computer vision","Keras","Transformers","R","Go","BERT","GPT"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/an-illustrative-guide-to-masked-image-modelling\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":67745,"title":"Is The End Near For ImageNet?","content":"ImageNet is one of the most popular image datasets organized according to the WordNet hierarchy. ImageNet was created with the objective to build tens of millions of cleanly sorted images for most of the concepts. Total number of images: 14,197,122Number of images with bounding box annotations: 1,034,908 Today, almost every state-of-the-art image recognition model is pre-trained on the ImageNet database. The universality of ImageNet makes one wonder if it is worth the praise. Also Read: Image Classification Benchmarking To investigate this, the researchers from Google, in their new work, ran experiments to check if the progress on the ImageNet classification benchmark is as good as what it is considered to be as. The authors check the benchmark for meaningful generalization and if the users have started to overfit the nuances of the labels in the ImageNet database. The researchers at Google developed a more robust procedure for collecting human annotations of the ImageNet validation set. Using these new labels, they have reassessed the accuracy of recently proposed ImageNet classifiers. To their surprise, they found that the newly reported gains are substantially smaller than those reported on the original labels. The researchers demonstrated the inefficiencies of using ImageNet. Then proceeded to develop a new labelling procedure and also introduced a new metric called ReaL accuracy to assess the new labels. Where Does ImageNet Fall Short And What Can Be Done As illustrated above, Red indicates original ImageNet label, green indicates the proposed ReaL labels. The authors show that even when a single object is present, ImageNet labels present systematic inaccuracies due to their labelling procedure. The authors came across two shortcomings of the ImageNet labelling procedure. ImageNet images are annotated by assigning a single label to images instead of multiple labels. To address this problem, the authors proposed the following: Use a training objective which allows models to emit multiple non-exclusive predictions for a given image. Treat the multi-way classification problem as a set of independent binary classification problems. Penalise each one with a sigmoid cross-entropy loss, which does not enforce mutually exclusive predictions. The authors observed that recent top-performing models surpass the original ImageNet labels in their ability to predict human preferences. To filter the noise, the researchers have used BiT-L model to clean the ImageNet training set. First, the training images are split into 10 equally-sized folds. And, one fold is excluded, and the BiT-L model is trained on the remaining 9 folds. The resulting model from this training is then made to predict labels on the hold-out fold. Images with labels that are inconsistent with BiT-L’s predictions are removed. While exposing the reduced usefulness of the ImageNet benchmark, the authors also proposed a solution to improve the results for ImageNet. They observed that long training schedules can be a hindrance in the presence of noisy data. So, they believe that cleaning the ImageNet training set will yield additional benefits. The authors draw the following insights from their experiments: Training on clean ImageNet data consistently improves accuracyUsing sigmoid loss resulted in consistent accuracy improvements across all ResNet architectures Key Takeaways Though they admitted that ImageNet dataset is a landmark achievement in the evaluation of machine learning techniques. They underline some limitations. One of them being a single label description of images instead of images containing multiple objects. Even for images containing a single object, biases in the collection procedure can lead to systematic inaccuracies. Also, some ImageNet classes even draw distinctions between identical groups of images. So, the researchers stress that there is a need for a new human annotation procedure. This work investigated whether recent progress on the ImageNet benchmark amounts to meaningful generalisation. Using “Reassessed Labels” (ReaL), the researchers have found that the association between progress on ImageNet and ReaL progress has weakened over time. The findings from this work can be summarised as follows: Recent top models have begun to surpass the original ImageNet labels in terms of their ReaL accuracy, indicating that we may be approaching the end of their usefulness.ImageNet usefulness in evaluating vision models may be nearing an end.ImageNet can be improved with cleaner data and by using sigmoid loss. Link to paper.","excerpt":"ImageNet is one of the most popular image datasets organized according to the WordNet hierarchy. ImageNet was created with the objective to build tens of millions of cleanly sorted images for most of the concepts.  Total number of images: 14,197,122 Number of images with bounding box annotations: 1,034,908 Today, almost every state-of-the-art image recognition model […]","categories":["Deep Tech"],"tags":["ImageNet"],"author_name":"Ram Sagar","publish_date":"2020-06-19T17:00:04","publication_year":"2020","word_count":690,"keywords":["Go","machine learning","programming_languages:R","AI","RPA","image recognition","ResNet","programming_languages:Go","GAN","R","ImageNet"],"extracted_tech_keywords":["AI","machine learning","image recognition","R","Go","GAN","ResNet","RPA","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/imagenet-computer-vision-accuracy\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":65368,"title":"Twitter Says It Is Game Over For Fake News","content":"The last couple of years had been turbulent for Twitter. Users have accused Twitter for showing double standards by banning an individual based on the reports of the rival faction. However, Twitter has been risking its reputations and rolling new features that will cut down misinformation while giving its users more control over conversations. Twitter is doubling down on its initiatives to curtail misinformation. Now, Twitter adds a new feature that would suggest if a certain tweet is insensitive or misleading. “We’re introducing new labels and warning messages that will provide additional context and information on some Tweets containing disputed or misleading information related to COVID-19.” In the case of information specific to COVID-19, unverified claims that have the potential to lead to the destruction or damage of critical infrastructure, or cause widespread panic\/social unrest may be considered a violation of Twitter’s policies. New Warnings . While false or misleading content can take many different forms, Twitter will be monitoring the fake news based on the following categories: Misleading information such as the statements that have been confirmed to be false or misleading by subject-matter experts, such as public health authorities Disputed claims such as the statements or assertions in which the accuracy, truthfulness, or credibility of the claim is contested or unknown Unverified claims such as the information that is unconfirmed at the time it is shared If a tweet falls in any of the above categories then the information will be hidden along with a label that takes the user to more authentic sources to verify the content in the tweet. The above snapshot shows how a potential COVID-19 misinformation video has been flagged. Most of the social media platforms use machine learning tools to automate their flagging process. Twitter too, since the introduction of its new policies on March 18, have removed more than 1,100 Tweets containing misleading and potentially harmful content from Twitter within the first 2 weeks. Additionally, their automated systems have challenged more than 1.5 million accounts, which were targeting discussions around COVID-19 with spammy or manipulative behaviors. Going Forward With over 300 million monthly active users, the amount of data that gets generated on Twitter is huge. Handling this kind of data and making sure that there are no mishits is a challenging job for engineers at Twitter. Twitter has been moving the goalposts of its policies and the definition of harm has changed dramatically. Using algorithms to curb misinformation can sometimes have mishits, and Twitter admits to the same and requests its users to bear with them and report it in case of any unfair takedowns. The whole landscape of privacy and freedom of speech has changed dramatically since the onset of the pandemic, and it looks like these policies are here to stay. Twitter made it clear that it will continue to prioritise removing content when there is a clear call to action that could directly pose a risk to people’s health or well-being, but they also admit that they will not be able to take enforcement action on every Tweet that contains disputed information about COVID-19. They also insist that these policies are not meant to limit good-faith discussion or expressing hope about ongoing studies related to potential medical interventions that show promise. In serving the public conversation, says the Twitter team, the goal is to simplify finding credible information and limiting the spread of potentially harmful and misleading content. Twitter, in their announcement, has also stated that going forward, it will continue to introduce new labels to cut down fake news if necessary.","excerpt":"The last couple of years had been turbulent for Twitter. Users have accused Twitter for showing double standards by banning an individual based on the reports of the rival faction. However, Twitter has been risking its reputations and rolling new features that will cut down misinformation while giving its users more control over conversations. Twitter […]","categories":["Global Tech"],"tags":["policy","Twitter (X)"],"author_name":"Ram Sagar","publish_date":"2020-05-16T16:00:47","publication_year":"2020","word_count":594,"keywords":["Go","machine learning","programming_languages:R","AI","programming_languages:Go","Aim","Twitter (X)","R","policy"],"extracted_tech_keywords":["AI","machine learning","Aim","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/twitter-fake-news\/","complexity_score":4,"technical_depth":7,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10009457,"title":"India Needs Quantum-Secure Encryption To Prevent Future Attacks, Says CTO of QNu Labs","content":"Quantum computing is the new technology frontier on which nations and big tech companies are fighting. While innovation in the field of quantum technology can bring new advancements in various sectors, it poses a significant threat to digital security. According to experts, using quantum computing blackhat hackers and states sponsored entities can attack and decrypt web assets and applications. The threat has pushed companies to look towards quantum cryptography and quantum key distribution (QKD). Quantum safe encryption will play a big role in preventing major security breaches but also overcome application failures in AI and ML. Despite the importance of quantum cryptography, there are only a few companies in India which provide quantum-safe products. For this weekly interview, we reached out to Vivek Shenoy, Chief Technology Officer, QNu Labs- an organisation that is currently leading the quantum encryption market in India. The firm has deployed its solutions across sectors such as defence, telecom, banking, finance and so on. According to Vivek Shenoy, the objective of QNu Labs is to put India at the forefront of data security through a new paradigm in encrypting keys using quantum physics principles. Here are the excerpts from the interview: What does the Quantum ecosystem in India look like at the moment? The quantum ecosystem in India is in a nascent stage. There is a precise intent shown by the government when it comes to providing the necessary impetus towards the growth of Quantum Technologies in India. The allocation of ₹8,000 crores towards the development of technologies such as Quantum Cryptography and Quantum Communication announced during the Union Budget 2020 will help India make significant strides in this space. But, there are critical challenges which need to be addressed for India to be a front runner. There continues to be a significant gap when it comes to the skilled workforce that can operate in the quantum space for India to compete with the larger global players. We also need to massively improve our hardware capabilities, including procuring the necessary equipment, establishing fab labs and so on. What is the role of QNu labs when it comes to the development of quantum technologies for the Indian and global market? QNu labs is a leading home-grown organisation that is at the vanguard of the quantum ecosystem in India. We are the enablers of the Quantum Technologies ecosystem in India, having developed ready to deploy solutions for two out of the nine applications identified under the National Mission on Quantum Technologies Applications. We are also the first company to represent Indian quantum technology ecosystem globally at the Quantum Alliance Initiative (QAI). We aim to put India at the forefront of data security and privacy powered through Quantum technology which is provided through a new paradigm in encrypting keys using Quantum physics principles centred on cybersecurity for the cloud and internet space. What are the security ramifications across various tech companies with the advent of quantum, and how can India prepare for it and implement quantum-safe encryption? The Quantum Computing breakthrough is poised to unlock new technological realms for the world. However, it also presents unfathomable security implications. Its incredibly fast processing capability also means that it can overpower any security infrastructure including military, nuclear commands, government bodies, and global financial institutions and so on. This makes it imperative for governments, industries and large organisations to ensure timely safeguards for the same. The Y2Q (Years to Quantum) moment, as it is most commonly known, is on the horizon and we need to act now. India must promote home-grown organisations to come up with new quantum technologies identified under the National Mission on Quantum Technologies. We also need to scale our hardware capabilities and strengthen the existing infrastructure to facilitate the development of new technologies. How do QNu Labs’ solutions work to strengthen the present-day cybersecurity infrastructure? We’re the only Indian company to have made significant strides when it comes to quantum encryption, which means that our security solutions are the most advanced and safe when it comes to present-day cybersecurity infrastructure. Nothing in the present day and age can penetrate the robust security mechanism that we provide to our customers. This is why we have been able to deploy our solutions across sectors such as defence, telecom, banking, finance and so on. Hackers are now carrying out attacks which are based on “Harvest Now, Decrypt Later”. In such attacks, the primary strategy adopted by hackers is to copy and store encrypted data and decrypt it at a later stage using quantum computers. As a result, forward security is very much necessary today, and the window to act is narrowing fast. QNu Labs intend to protect data at a national level and to have a global impact. We are not only ensuring a robust security mechanism for our clients but also ensuring the behavioural shift across industries so that they do not have to wait for the Y2Q moment to maintain proper safeguards. They can do so today by deploying our solutions that not only cater to the cybersecurity needs for today but also for the future. Can you elucidate on QNu Labs’ adoption of Quantum Key Distribution or QKD and product vision\/roadmap? Our Quantum Key Distribution System (QKD) called Armos was launched earlier this year in March and was made deployment-ready in India as well in international markets. The solution is directed at addressing the growing issues around data security and privacy, which have become a fundamental concern for businesses, governments and defence organisations across the globe. Our product roadmap involves the development of Quantum Hardware Security Modules and Post Quantum Cryptography solutions, which will become the basis for critical infrastructure in the future. Please tell us about the successful field trial of the first quantum secure link in India running between two defence locations in North India. This link was set up between two defence establishments, about 50 km apart. Two dark fibres were used, one for the quantum channel and one for clock synchronisation. Two C-DOT encryptors were used at each station that would retrieve the ultra-secure key that Armos would generate and encrypt all data flow between these two stations. This setup ran continuously and flawlessly for five days, with keys being refreshed every one minute between the routers. Tell us about how your clients are using products like Quantum Key Distribution, Quantum Secure Platform and Entropy as a Service? Our Quantum Key Distribution (QKD) solution called Armos provides impenetrable and unconditional data protection when data is at its most vulnerable–in motion. Using the laws of quantum physics, Armos secures the distribution of symmetric encryption keys. Armos functions by quantum encoding single photons of light which are sent through a dedicated fibre optical channel. Heisenberg’s Uncertainty principle ensures that any effort to tap into these quantum keys creates a perturbation in their quantum state which can be detected. Any product changes seen in the quantum state of the keys conveys Armos to make an alert and give an accurate real-time report regarding the location of the intrusion. Armos destroys any compromised keys and produces new ones, thus preventing attacks even before they begin. Only when the quantum encoded photons are checked at both the ends of the system are the keys generated. These quantum keys are then utilised to encrypt and decrypt data which is sent on classical channels. As the keys cannot be cloned, the data–even if cloned–remains unconditionally safe. Our Quantum Security Platform, Hodos provides key management and application layer for our clients to plug and play. Hodos is built to integrate with pre-existing infrastructure and give firms a higher security posture instantly. Hodos is powered by Quantum Physics, making it the most future-proof technology in cybersecurity today. It needs minimal upgrades which lower the Total Cost of Ownership (TCO) and is also resistant to increases in computing power or more powerful algorithms. The solution offers advantages such as real-time intrusion detection, identifying the exact location of eavesdropping devices and actionable real-time reports customised to meet our customer’s requirement. Similarly, our Entropy-as-a-Service architecture is designed to provide a quantum entropy source to the connected world and IoT devices today. It provides secure access to high entropy according to clients needs, with entropy becoming a requirement in many industries, EaaS aids easy pay per entropy service. Please tell us about the research projects that you are undertaking. Any new innovations in pipeline and product roadmap for the future? There are multiple areas that QNu Labs is working in. The most important area being QKD and its various offshoots – to increase distance, detect and take countermeasures against intrusion, free-space and satellite QKD. Beyond QKD, QNu is working on Hardware Security Modules, which are based on its QRNG technology. The QRNG technology is being enhanced to fit into servers as a PCIe card or a USB dongle. In a separate research stream, QNu is working on Post Quantum Cryptography (PQC) that enables us to create a hybrid and quantum-secure key distribution network based on new and upcoming mathematics-based key exchange mechanisms. This ties in nicely with the QKD scheme, extending the reach of QKD and creating a true quantum-secure network. QNu research team works closely with leading academic institutions and researchers around the world, leveraging both their expertise and insight to improve both QKD and PQC.","excerpt":"Quantum computing is the new technology frontier on which nations and big tech companies are fighting. While innovation in the field of quantum technology can bring new advancements in various sectors, it poses a significant threat to digital security.  According to experts, using quantum computing blackhat hackers and states sponsored entities can attack and decrypt […]","categories":["AI Features"],"tags":["computer vision future","Interviews and Discussions","quantum application development system","Quantum Computing","Startups","the need to scale with big data","the quantum companies","what companies use quantum computers"],"author_name":"Vishal Chawla","publish_date":"2020-10-11T18:00:00","publication_year":"2020","word_count":1550,"keywords":["Quantum Computing","Go","what companies use quantum computers","AWS","AI","innovation","ML","the need to scale with big data","Git","the quantum companies","RAG","Aim","quantum application development system","GAN","Startups","computer vision future","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","ML","Aim","RAG","AWS","R","Go","Git","GAN","innovation"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/indian-firms-need-quantum-secure-key-distribution-to-prevent-future-attacks-says-cto-of-qnu-labs\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10111573,"title":"Animation, VFX, Extended Reality to Generate 30,000 Jobs by 2028: Karnataka CM","content":"The Animation, Visual Effects, Gaming & Comics Hub (AVGC) industry will create around 30,000 jobs in the state by 2028, Karnataka chief minister Shri Siddaramaiah said while speaking at the inaugural ceremony of the fifth edition of Bengaluru GAFX, held in Bengaluru. “With over 15,000 professionals and 300 specialised studios, GAFX celebrates Karnataka’s role as a global AVGC hub. Our state, a trailblazer in technology, is proud to lead the nation’s Media & Entertainment sector, contributing 20% to the industry,” he said. Speaking at the same event, Shri Priyank Kharge, Hon’ble Minister for Information Technology & Biotechnology and Rural Development & Panchayat Raj, Govt. of Karnataka, said, “The state is a pioneer in evolving India’s AVGC policies, is committed to securing a future where it not only maintains the state’s leadership in traditional IT-BPM but emerges as a global powerhouse in AVGC-XR.” He also revealed that the state plans to establish AVGC-XR-focused incubators to help drive the industry. During the Bengaluru Tech Summit (BTS) 2023, the Karnataka government introduced draft policies for the Animation, Visual Effects, Gaming, Comics, and Extended Reality (AVGC-XR) industry. The AVGC-XR policy focuses on positioning Karnataka as a hub in the AVGC sector, offering incentives to startups, MSMEs, and entrepreneurs from SC\/ST and women categories.","excerpt":"The state is also looking to help establish AVGC-XR-focused incubators","categories":["AI News"],"tags":["Gaming"],"author_name":"Pritam Bordoloi","publish_date":"2024-01-29T17:22:56","publication_year":"2024","word_count":209,"keywords":["Go","programming_languages:R","AI","Gaming","programming_languages:Go","R","startup"],"extracted_tech_keywords":["AI","R","Go","startup","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/animation-vfx-extended-reality-to-generate-30000-jobs-by-2028-karnataka-cm\/","complexity_score":2,"technical_depth":6,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":40734,"title":"Is Zero-Trust Approach The Next Best Strategy To Keep Data Safe?","content":"There is no doubt the cyber threats have reached a whole new level — they have become more and more sophisticated, making it tough for security professionals to deal with them. This is also trying to show that the InfoSec industry is not exactly doing the right thing, or is it just running in circles behind ideas and concepts to strengthen cybersecurity. How do we deal with these ever-increasing cyber threats? And the ultimate answer to all these questions is the fact that companies are not having a right and ultimate goal — the goal has to be about reducing, mitigating cyber threats and protect the valuable sensitive data from getting exposed. Don’t Trust. Verify. While there are new concepts coming up in the cyber security domain, there is one concept that seems to be gaining a lot of popularity, lately — Zero-Trust Approach. When it comes to fighting cyber threats and prevent the exfiltration of sensitive data, Zero Trust architecture is considered to be a great approach for businesses across the world. Zero-Trust is basically all about the concept of verifying before trusting. It is designed to address lateral threat movement within a network, and it can be done by leveraging micro-segmentation and granular perimeters enforcement, based on user, data and location. The biggest advantage of taking a Zero-trust approach to security is that it is not biased — it doesn’t differentiate between insiders and outsiders when it comes to security. Amid all the data breaches happening across the world, it is the best way to protect mission-critical data and systems. Simply put, if someone is not trying or attempting to access any company data, that doesn’t mean that person is trusted. Rather, the company can make use of ways to verify the person at first. Base For A Zero-Trust Architecture A zero-trust model for access control has several benefits; however, in order to implement this technology strategy, an organisation must keep a few things in mind. Make Sure Data Is Accessed Securely Irrespective of who the employee is, what is his\/her designation, it is imperative for organisations to verify the identity of every single user, every single time, who is accessing the data. It is all about implementing a least-privileged access strategy and strictly enforce access control. Implement Authentication Methods When you strengthen the access control by imposing strict protocols, you also need to implement authentication methods for the authorized people to access the data and that can be done through a combination of single sign-on, multi-factor authentication (MFA), or even biometrics if required. Also, it is important to keep an eye on employees who are not granted access and still attempts to access that data. That is not all, one must also ensure that what applications users are using and the appropriate connection method. When it comes to data, no organisation can compromise with security. Make Use Of Behavioral Analytics When the prime motive is to keep the data safe, no one is trusted in this approach — until and unless they pass verification. One of the verification methods could be behaviour-based analytics, which basically about providing insight into the actions of people. This can be achieved by using AI that would keep an eye on the person who is accessing the data. Every time it accesses the data, the AI would record what s\/he is doing with it. This would allow the company to understand whether the person is trustworthy or not. Outlook While many organisations are focusing on their perimeter-based security, many have almost forgotten that sometimes the attack starts inside. Meaning, the point of infiltration of an attack is not the target location all the time. With time, even hackers have curated sophisticated techniques to compromises data of organisations across the world. If your organisation is still using traditional methods for the cybersecurity infrastructure, then it is high time to give Zero-trust approach a shot.","excerpt":"There is no doubt the cyber threats have reached a whole new level — they have become more and more sophisticated, making it tough for security professionals to deal with them. This is also trying to show that the InfoSec industry is not exactly doing the right thing, or is it just running in circles […]","categories":["AI Features"],"tags":["Cyber Security"],"author_name":"Harshajit Sarmah","publish_date":"2019-06-16T09:36:54","publication_year":"2019","word_count":651,"keywords":["Go","Cyber Security","programming_languages:R","AI","programming_languages:Go","RAG","analytics","Rust","GAN","R","programming_languages:Rust"],"extracted_tech_keywords":["AI","analytics","RAG","R","Go","Rust","GAN","programming_languages:R","programming_languages:Go","programming_languages:Rust"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/is-zero-trust-approach-the-next-best-strategy-to-keep-data-safe\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10117708,"title":"Downtime Decoded: How to Revolutionise Manufacturing with GenAI","content":"Machine downtime is a significant challenge in the manufacturing sector, costing industrial manufacturers between 5 and 20% of their productive capacity. The global cost of downtime runs into trillions of dollars annually, making it a critical issue for the industry. A 2023 report by Siemens highlighted the escalating costs of downtime. Fortune Global 500 companies now face losses of around 11% of their yearly turnover, totalling nearly $1.5 trillion. Understanding the Costs and Factors According to Siemens’ findings, the annual cost of downtime per facility among Fortune Global 500 companies has risen by 65% in recent years, reaching a staggering $129 million per facility. The cost of a lost hour varies from industry to industry, with automotive plants experiencing the highest of over $2 million per hour. Similarly, the oil and gas sector has seen a doubling of hourly downtime costs to nearly $500,000. Downtime incidents are influenced by a myriad of factors, ranging from equipment age and condition to maintenance practices, operator training, material quality, and environmental factors. Supply chain disruptions, human error, unplanned events, and software issues further compound the problem. Identifying and addressing these factors presents low-hanging opportunities for manufacturers to reduce downtime and improve operational efficiency. The Role of Generative AI In recent years, advancements in AI have opened up new avenues for addressing downtime challenges. Generative AI, powered by advanced models like LLaMA, BERT, Falcon, GPT-4, and Gemini, offers an innovative approach to tackling downtime. Generative AI can analyse vast amounts of operational data to identify patterns, predict potential downtime events, and recommend proactive maintenance strategies. By fine-tuning these models with real-time data and leveraging the RAG AI framework during the proof-of-concept stage, manufacturers can experience a significant reduction in downtime incidents. Early successes have shown a 3-5% decline in machine downtime within the first quarter of implementation. This motivated plant managers to increase their investment in generative AI initiatives, bringing additional data sources to improve model accuracy and relevancy. With the help of LLMs like GPT-4 and Gemini and, RAG, shop floor employees get access to actionable information and real-time insights. Through mobile or web applications, employees can access troubleshooting workflows, escalation SOPs, and interactive Q&A sessions, expediting the recovery process and ensuring adherence to best practices. Training Workforce and Measuring Impact Empowering the workforce is another critical aspect of making the most of generative AI. Regular training workshops for maintenance technicians, operators, and managers are essential. Adherence to standard operating procedures (SOPs), scheduled servicing, and maintenance protocols further enhance operational resilience. Both hands-on and classroom training are common in the manufacturing sector, ensuring each employee has access to essential documents and resources. To validate the effectiveness of AI solutions, it’s crucial to measure its impact on key performance metrics such as Overall Equipment Effectiveness (OEE), Mean Time to Repair (MTTR), Downtime Frequency, and Scrap Rate. Integrating generative AI into the manufacturing processes not only reduces downtime but also enhances productivity, operational efficiency, and decision-making capabilities. The Final Word In conclusion, the urgency of addressing machine downtime in manufacturing cannot be overstated. By leveraging generative AI, adopting proactive measures, empowering a skilled workforce, and continuously measuring and improving operational metrics, manufacturers can navigate challenges more effectively, minimise downtime, and maximise efficiency. This journey towards resilience and innovation is imperative in today’s dynamic manufacturing landscape. This comprehensive approach, combining advanced AI technologies with strategic initiatives and workforce empowerment, paves the way for a more resilient and efficient manufacturing sector. As industries evolve and challenges persist, embracing innovation becomes not just a competitive advantage but a necessity for sustained success. Generative AI offers a transformative path forward, revolutionising manufacturing processes and driving tangible business outcomes.","excerpt":"Generative AI can analyse vast amounts of operational data to identify patterns, predict potential downtime events, and recommend proactive maintenance strategies.","categories":["AI Highlights"],"tags":["AI for manufacturing","Generative AI"],"author_name":"Sushant Ajmani","publish_date":"2024-04-08T10:36:37","publication_year":"2024","word_count":609,"keywords":["AI","innovation","Scala","RAG","BERT","GPT","AI for manufacturing","generative AI","ViT","disruption","Generative AI","R"],"extracted_tech_keywords":["AI","generative AI","RAG","R","Scala","BERT","GPT","ViT","innovation","disruption"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/downtime-decoded-how-to-revolutionise-manufacturing-with-genai\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10018144,"title":"How AI-Powered Digital Twins Are Revolutionising Manufacturing Industry &#8211; And Saving Loads Of Money","content":"Safety is the most critical parameter for a vehicle, which makes crash testing the most crucial phase in production. It also happens to be very expensive. However, with the latest technological advancements, crash testing has become more data-driven and has, in turn, reduced the number of tests to be performed to meet the safety requirements. Here, we try to understand how companies use digital twins and AI in driving down production and manufacturing costs. What Is A Digital Twin? A digital twin is a virtual representation of a physical device. It is a computer program that takes real-world data about a physical object or system as inputs to produce predictions or simulations of how the physical object or system will respond in different situations. Various technologies have to come together for the digital twin technology to work efficiently. First, IoT devices capture data from a prototype. The versatility of IoT devices available in the market today makes capturing data from these objects easy. As IoT devices get more advanced, digital-twin scenarios can include smaller and more complex objects. Data could also come from objects that are not yet built, using design specifications, production processes or engineering information. This includes data on materials, equipment, methods, parts, historical analysis and maintenance record. Once the data is collected, data scientists can build virtual simulations using mathematical models and create a digital twin of the object with softwares such as CAD or 3D design. A Winning Combo The manufacturers can reap the optimal benefit by combining the digital twin with AI or machine learning algorithms. The resulting digital model can simulate the object’s performance in virtual environments. The data from these simulations can be used to train AI models. The AI system can be taught to replicate potential real-world conditions. This way, the manufacturers don’t have to crash an actual vehicle for testing. Digital twins can also be coupled with AI to test new features in a simulated object. Let’s imagine an aircraft. IoT devices on the plane can collect data and build a digital twin. The manufacturers can tinker with the virtual model to calibrate optimum specs for the given requirement, saving a lot of cash in the process. The virtual model, along with IoT devices, can also be used to relay real-time data to offer insights into the performance or detect potential problems in the object or system. This could save resources and reduce the likelihood of accidents. Digital Twin Applications Cut Costs NASA’s rescue mission of Apollo 13 first anticipated the use of Digital Twin technology. The astronauts and engineers used a physical twin of the Apollo 13 to figure out the modus operandi for the rescue mission. Today, the twin is virtual and more sophisticated. The technology is being used in Chevron’s manufacturing plants to reduce maintenance issues on supply chains and ensure optimal production by monitoring the equipment and other physical assets in real-time. Recently, ExactSpace has implemented the tech in thermal power plants in India. They used the tech to figure out the optimal configuration for the plant in a live setting. This helped reduce fuel input and improve efficiency, thus saving on production costs and reducing carbon emissions. Zimmer Group, a German-based company, has used digital twin technology to simulate a furniture manufacturing factory using affordable industrial robots, instead of the expensive special-purpose machines used before. General Electric has successfully applied the Digital Twin technology across sectors saving up to $1.6 billion for companies worldwide. In one instance, the technology warned of a supply leak on a gas turbine and saved $9 million. In another, it detected increased suction temperature on a centrifugal compressor, saving $3.2 million. Big firms such as IBM, Siemens and Microsoft are also working on digital twin technology services. Wrapping Up The technology harnesses the synergy of digital twins and the latest advancements in AI and IoT to help the companies in saving capital and operational expenditure in manufacturing, production, and maintenance. It’s still early days for the tech, and we might see incremental progress in coming days, given the rate of progress of the tech and rising use cases.","excerpt":"Safety is the most critical parameter for a vehicle, which makes crash testing the most crucial phase in production. It also happens to be very expensive. However, with the latest technological advancements, crash testing has become more data-driven and has, in turn, reduced the number of tests to be performed to meet the safety requirements.  […]","categories":["IT Services"],"tags":[],"author_name":"Kashyap Raibagi","publish_date":"2021-01-15T16:00:00","publication_year":"2021","word_count":687,"keywords":["Replicate","Go","API","machine learning","programming_languages:R","AI","data-driven","programming_languages:Go","Git","R"],"extracted_tech_keywords":["AI","machine learning","R","Go","Git","API","data-driven","Replicate","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/how-ai-powered-digital-twins-are-revolutionising-manufacturing-industry-and-saving-loads-of-money\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":59750,"title":"How COVID-19 Is Burning A Hole In Pockets Of The Big Tech","content":"This is probably the first time in the history of mankind that the entire population is focussed on one single problem. As the experts and lawmakers tussled with how to engage in war with this new adversary, the virality of  COVID-19 had already made dents in the socio-economic fabric of the society. Now analysts forecast a recession around the corner while others warn that we already have entered a recession. The latter, unfortunately, makes more sense now as the current market meltdown echoes the sins of the 2008 financial crisis. There are bailouts, there are lay-offs, and all these things indicate towards the most obvious situation — a recession. The aftermath of the 2008 financial crisis, has seen the tech companies spearheading the bullish markets. This has been the scenario for the past 11 years. Trillion-dollar companies or the big four — Apple, Amazon, Microsoft and Alphabet — along with Facebook, took heavy losses in the past couple of weeks. via CNBC As of March 19, those four companies — Apple, Microsoft, Amazon and Alphabet — along with Facebook have lost a combined $1.3 trillion in valuation since the market peaked. One unusual in which the global pandemic is squeezing tech companies around the world through the cancellation of conferences. According to Oxford economics, business conferences generate over a trillion dollars in annual spending. Apart from the economic aspect, there is a lot more to these conferences. Tech companies especially, leverage these conferences to pitch their new products, ideas and a sneak peek into the future. As of now, all the major conferences, from Google’s cloud event to Apple’s WWDC in the US to Mobile World Congress in Barcelona are either cancelled or switched to an online-only event. Collateral Effects Of Cancelled Conferences: The Losers Attendees at MWC According to PredictHQ, a company that offers intelligence from the data of global events, The direct economic loss from the cancellation of more than 10 major tech conferences — including Google I\/O, Facebook’s F8 event, Mobile World Congress, SXSW, and now the massive video gaming conference Electronic Entertainment Expo (E3) — due to the coronavirus outbreak has surpassed $1.1 billion. The ticket cost of these conferences is fairly high and maybe rightfully so, given the kind of opportunities that it offers to everyone. These tickets are major revenue sources for those who host these conferences. These major tech conferences, in particular, have a huge economic footprint, as attendees boost to the revenues of the city through their hospitality sector. One of the largest major tech events to get cancelled due to COVID-19 is Mobile World Congress, the world’s biggest conference for the mobile industry. According to PredictHQ, the cancellations would cost MWC organisers a monstrous half a billion euros in revenue. This annual event draws a crowd of over 100,000 from over 200 countries and generates over 14,000 part-time jobs. Another shocker was the cancellation of Austin’s South by Southwest festival. The annual 10-day-long event features film, music and interactive tech. Last year, the event drew a crowd of over 417,000 people. According to reports, the city of Austin lost a whopping $355 million in revenue for the year 2020. Adding to the woes is the fact that pandemics are not covered under the event’s insurance policy. Now the hosts are being forced to lay off a major fraction of their employee base as they scramble to find alternative sources of income to keep from running out of money by this summer. At this point, it seems like merely a matter of time before other tech events get suspended, piling misery on small business owners and the hosting cities. How Is The Big Tech Doing From the developer’s perspective, the conferences hosted by Google or Amazon or Facebook is a huge opportunity to both learn and display their ideas to a larger audience. The conferences of the big five cover a broad range of topics from artificial intelligence to cloud computing to AR\/VR. This might still be taken care of using virtual platforms or other social media platforms. But, the economic impact of these conferences is grim nevertheless. While the reach of Facebook’s F8 event is relatively less compared to the Mobile World Congress, the conference is still nothing to scoff at. Last year, the Facebook event drew over 5,000 attendees. PredictHQ estimates the direct losses due to cancellation of F8 to be $12.2 million, a figure that includes losses to hotels, airlines, restaurants and transportation providers that would normally profit from the conference. After the cancellation, however, Facebook announced that it would be doubling its annual donation to $500,000, prioritising organisations serving the San Jose area. Whereas, Google I\/O event too, met with the same fate as F8. The Mountain View-based event draws over 7,000 attendees and the direct loss for the cancellation, according to CNBC, is estimated to be about $19.5 million. However, unlike other companies, Google is offering full refunds worth $1,150 per ticket. The company also pledged a $1 million dollar donation to the city of Mountain View to help small businesses recoup the losses. Apple and Microsoft, too, have followed suit and moved their respective developer conferences online. However, these tech companies will still be doing fine thanks to the numerous verticals they have entered over the past few years. Amazon, Google and Microsoft offer cloud-based AI services amongst many other things. The global lockdown has forced many organisations to look towards alternatives that assist remote work. The might of the internet has been witnessed through these tumultuous times with the help of these companies. From small businesses to individuals, these companies have a lot to offer, and given the huge demand, the tech giants might probably come out unscathed when the things cool down.","excerpt":"This is probably the first time in the history of mankind that the entire population is focussed on one single problem. As the experts and lawmakers tussled with how to engage in war with this new adversary, the virality of  COVID-19 had already made dents in the socio-economic fabric of the society. Now analysts forecast […]","categories":["Deep Tech"],"tags":["conference","covid-19"],"author_name":"Ram Sagar","publish_date":"2020-03-24T16:00:58","publication_year":"2020","word_count":958,"keywords":["Go","artificial intelligence","ELT","covid-19","AI","conference","cloud computing","AWS","RPA","RAG","GAN","R"],"extracted_tech_keywords":["AI","artificial intelligence","RAG","cloud computing","AWS","R","Go","ELT","GAN","RPA"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/covid-19-economic-impacts-google-conference-cancel-loss\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10087566,"title":"This Extension Allows Grammarly-like Edits on ChatGPT","content":"Apart from generating content from prompts, ChatGPT can also be used for auto-editing, re-writing, and summarising content. But there is no option to see the changes that it has made to the content. To solve this problem, Reddit user shuafeiwang, developed a Chrome extension called editGPT, that turns ChatGPT into ‘Grammarly’. The extension does not generate content, but allows users to check the changes made by ChatGPT. The tool makes it easier to track and verify the change of content by highlighting content, just like Grammarly, and users can accept or reject specific suggestions. Once the extension is installed, users can type in a prompt at ChatGPT like “Proofread this:” followed by the text. This will show an editing button on the menu bar. The developer is also experimenting with different types of prompts to enable more flexibility and personalisation in the editing style. Proofread this but only fix grammar: (10% change) Proofread this: (20-30% change) Proofread this, lightly improving clarity and flow: (30-40% change) Proofread this, improving clarity and flow: (30-50% change) Proofread this, significantly improving clarity and flow: (40-60% change) Rewrite this, improving prose: Proofread these bullet points from my CV, keep it in CV language: Proofread these bullet points: The extension is available for Microsoft Edge as well, and the developer is working on one for Firefox. You can track further developments of editGPT in this Reddit thread created by the developer. ChatGPT everywhere There has been a surge in ChatGPT-based tools. Recently, Sarvesh Kapre developed GPTx, a Chrome extension that enhances the browsing experience by integrating ChatGPT results in the Google search box. The extension gives GPT-powered results combined with Google results. Another developer and co-founder of Cue, Daniel Gross, integrated GPT-3.5 on WhatsApp as an AI-enabled assistant, and called it WhatsApp-GPT. Stanford students Silas Alberti and Joseph Semrai developed ChatBCG, for creating text-to-slides. Read: Crazy ChatGPT Clones","excerpt":"The tool makes it easier to track and verify the change of content by highlighting content, just like Grammarly, and users can accept or reject specific suggestions.","categories":["AI News"],"tags":["ChatGPT"],"author_name":"Mohit Pandey","publish_date":"2023-02-17T13:14:27","publication_year":"2023","word_count":312,"keywords":["Go","ChatGPT","programming_languages:R","AI","BERT","GPT","llm_models:BERT","R","llm_models:ChatGPT","llm_models:GPT"],"extracted_tech_keywords":["AI","ChatGPT","R","Go","BERT","GPT","llm_models:GPT","llm_models:BERT","llm_models:ChatGPT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/this-extension-allows-grammarly-like-edits-on-chatgpt\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":65531,"title":"China Launches AI Virtual Idol, Ling","content":"China has recently launched a virtual key opinion leader — Ling. Ling has made her world online debut keeping up with the latest Chinese internet trend among positive influencers. According to a report, this 3D virtual figure has been designed to showcase the country’s soft power and culture. According to the report, the AI virtual influencer — Ling has been co-created by Chinese artificial intelligence startup Shanghai Xmov Information Technology and Beijing Cishi Culture Media Company and has been designed to have a lot of similarity to humans. The startup involved — Xmov has self-developed full-stack end-to-end AI technology and therefore, has opened the process of intelligent characterisation from intelligent modelling to artificial intelligence performance animation technology. These technologies are used to drive facial expressions, eyes, body and finger movements of the virtual influencer. The technology further generates short videos or real-time broadcasts to understand the interaction and commercialisation of the virtually intelligent property, stated the report. The report further stated that the Ling’s team would generate content around such characteristics as the national essence of Peking Opera and the combination of classic and modern fashion on social media platforms like Weibo, Instagram, and Douyin, Xmov. Ling further utilises commercial programming to engage in business endorsements, live broadcasts, and participate in online and offline activities, the report stated. The virtual idol sector has been gradually emerging and is heavily backed by the younger generation, who love getting entertaining content from these virtual influencers. Alongside, in recent years, more domestic firms are increasing their investments in this emerging virtual idol sector, which in turn has provided a push to this sector. According to a research report, the size of China’s virtual idol industry was less than 100 million yuan, which is approximate, $14 million, in 2018, however with the recent development in the industry and increased investment, it is expected to grow to 1.5 billion yuan by 2023.","excerpt":"China has recently launched a virtual key opinion leader — Ling. Ling has made her world online debut keeping up with the latest Chinese internet trend among positive influencers. According to a report, this 3D virtual figure has been designed to showcase the country’s soft power and culture. According to the report, the AI virtual […]","categories":["AI News"],"tags":["China","china ai","china ai investments","Virtual Influencers"],"author_name":"Sejuti Das","publish_date":"2020-05-20T11:01:31","publication_year":"2020","word_count":317,"keywords":["artificial intelligence","programming_languages:R","AI","china ai investments","china ai","Virtual Influencers","ViT","R","China","startup"],"extracted_tech_keywords":["AI","artificial intelligence","R","ViT","startup","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/china-launches-ai-virtual-idol-ling\/","complexity_score":2,"technical_depth":6,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":36316,"title":"5 Times Karnataka CM HD Kumaraswamy Championed Artificial Intelligence","content":"As political parties gear up for the 2019 general elections, Analytics India Magazine walks down the memory lane and takes stock of the times when leaders were quite gracious in praising and educating the mass about the importance of emerging technologies like artificial intelligence and machine learning. After his tumultuous election win, the Janata Dal (Secular) supremo, HD Kumaraswamy has stepped into the shoes of his predecessor with ease in Karnataka. In fact, Kumaraswamy has been quite vocal about the role of AI in shaping the future of the state’s economy. Here are some of his best quotes on emerging technology, especially AI: On turning the state capital into a global innovation hub:  With the government playing a crucial role in bringing various stakeholders from across the country under one roof, state-sponsored Bengaluru Tech Summit has played a vital role in ushering innovation and impact. While addressing a press meet the CM said, ”Bengaluru has emerged as one of the global innovation hubs in the league of Tokyo in Japan, Silicon Valley in the US and Tel Aviv in Israel. The summit will provide a platform for knowledge sharing on emerging technologies like Artificial Intelligence, Robotics and Blockchain,” Kumaraswamy said. On addressing creating societal needs: Speaking about the role of emerging technologies at the inaugural function of the 21st Bengaluru Tech Summit, he said, “The new technologies such as artificial intelligence, data analytics, Internet of Things (IoT), blockchain and automation help in providing solutions in diverse sectors and addressing societal needs.” To attract more investment: The CM has been pushing for investment to turn the state into a hub for development. In order to achieve this, acres of farmland has been identified by the government to set up industrial units across rural Karnataka. Highlighting the importance of emerging technologies in ushering investments, Kumaraswamy said, “Our state’s technology engagement has attracted investments from the across the country and abroad, making Karnataka a favourite for investing in disruptive technologies like Artificial Intelligence, Automation, Data Analytics, Machine Learning, 3D Printing and Robotics,” said Kumaraswamy.” To create a futuristics state:  Stating that most startups working in the field of IT, ITeS and other emerging tech related are largely concentrated in Bengaluru, the CM highlighted that there is a need for technology solution providers to look into other cities within the state so as to ensure a wholesome state. According to the CM, the government’s vision is to create a futuristic state which technologically advanced, penning his thoughts on the role of emerging technologies in shaping the future in a leading daily he wrote, “The vision for Karnataka 2025 is to establish the state as the torchbearer for India led by technology. Manufacturing processes will undergo a transformation through technologies such as IoT, robotics, 3D printing, artificial intelligence (AI) and machine learning (ML). Karnataka has taken proactive steps to promote these emerging technologies. On creating more Centre of Excellence in the state: The state capital has always been a hub for innovation and with the availability of qualified workforce has always attracted tech giants to set up their offices and even centres of excellence to conduct research and benefit from their talented skill set. In fact, in November 2018, Intel established their second largest  CoE in Bengaluru with an investment of Rs 1,100 crore. The state government is also mulling creating a more conducive environment for research and development by constructing more CoEs in the field of AI and ML. Highlighting the areas where the government wish to concentrate on, the CM said, ”Together with research and product development, we are laying emphasis on training and capacity-building. Centres of Excellence have been already set up in areas such as aerospace and defence, IoT, data science, AI, ML, robotics, and cybersecurity.","excerpt":"As political parties gear up for the 2019 general elections, Analytics India Magazine walks down the memory lane and takes stock of the times when leaders were quite gracious in praising and educating the mass about the importance of emerging technologies like artificial intelligence and machine learning. After his tumultuous election win, the Janata Dal […]","categories":["AI Trends"],"tags":["AI and Deep Learning","AI and Indian Economy","AI centre of excellence","Bengaluru"],"author_name":"Akshaya Asokan","publish_date":"2019-03-14T10:44:31","publication_year":"2019","word_count":625,"keywords":["data science","Go","API","artificial intelligence","machine learning","AI","AI and Deep Learning","ML","AI centre of excellence","AI and Indian Economy","ViT","analytics","Bengaluru","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","data science","analytics","R","Go","API","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/five-times-karnataka-cm-hd-kumaraswamy-championed-artificial-intelligence\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10023021,"title":"The Hustle Of Minus Zero: The Punjabi Startup Building Affordable, Fully Autonomous EVs For Indian Roads","content":"“Leveraging the true power of autonomy tech and building the product with a lasting impact is how we define success at Minus Zero.”Gagandeep Reehal For this week’s startup feature, Analytics India Magazine spoke to Gagandeep Reehal, the CEO and CTO at Minus Zero to understand how the Jalandhar-based startup is leveraging emerging technologies to create fully autonomous driving capabilities in electric cars. Founded in 2020 by Gagandeep Reehal and Gursimran Kalra, Minus Zero is India’s first start-up building affordable fully self-driving electric cars in the country. The company was founded amidst the pandemic and is headquartered in Jalandhar with an R&D unit in Patiala currently. The founders claimed the company is capable of Level 5 autonomy with highly energy-efficient electric vehicle design and proprietary nature-inspired AI that is less dependent on extensive data & costly sensor suite (like LIDAR’s, etc.). Reehal said, “Our fully Autonomous Electric Cars will be able to drive autonomously in highly unstructured and disorganised traffic scenarios like India’s.” Flagship Products Currently, Minus Zero is planning to start pre-orders of the fully autonomous electric car in late 2023. The self-driving car is capable of Level 5 autonomy in the toughest Indian traffic scenarios and is coupled with an energy-efficient electric vehicle design that could cover around 180 km on a 45 minutes charge with a total capacity of 500 km on a single charge. Reehal added, “Ranged between INR 16-25 lakhs, the vehicle will cater to the needs of Generation Z, with a hybrid sports design and high performance at a competitively affordable price. There will be a lot of features which we will reveal over time.” What’s The Differentiator “Many industry prototypes make heavy use of LiDARS and secondary sensors, while our Self Driving Architecture will drive the car similar to how a human does using only eyes, camera in the car’s case. This will make the travel experience more rider-centric, eliminating the pressure on a driver, and enabling safer travel,” said Reehal. He also mentioned some of the competitive advantages The company’s perception algorithms do not depend on lane markings for road detection, making it possible to drive autonomously even on roads in bad condition.The systems use 10x lesser GFLOPS per frame achieving high-speed processing, even on edge hardware, making it computationally efficient without compromising on accuracy. Zero dependencies on expensive sensor suites like long-range LiDARS, even at night.Nature-inspired vision AI uses predictive behavioural-based path planning to ensure robust autonomy in even unseen environments. Use of AI @ Minus Zero Minus Zero uses a hybrid neural network, taking into account the cognitive bias generated by serial-progressions in an intuitive way, to finetune the prediction. The company’s proprietary approach solves one of the computer vision’s major problems, occlusion. This also decreases the need for more training on a large dataset, thus making it robust even on completely random traffic scenarios. Core Tech Stack The tech stack at Minus Zero includes various machine learning techniques, open-source Python-based frameworks. The mechanical design & control of the vehicle are being developed in CAD environments for dynamic testing, design and simulations. Hiring Phase At present, Minus Zero includes a team of 25 employees. While hiring, the company looks for dedicated and passionate individuals with a research bent of mind, who are willing to explore new technology with a strong practical base. Future Roadmap “Our whole focus and energy is devoted to bringing the product into the market as early as possible. As mentioned earlier, we will be launching a market-ready version of our product by late 2023 and file for a regulatory testing post that. We will launch a set number of our initial model in some specified cities and monitor it closely. Along with that, our R&D team will continue to make improvements & bring updates to our self-driving software. We will begin mass-scale production after a successful launch in the specified cities and thereafter, ship out the products throughout India. We hope to achieve our target sales of 5,000 units by late 2025,” said Reehal.","excerpt":"“Leveraging the true power of autonomy tech and building the product with a lasting impact is how we define success at Minus Zero.” Gagandeep Reehal For this week’s startup feature, Analytics India Magazine spoke to Gagandeep Reehal, the CEO and CTO at Minus Zero to understand how the Jalandhar-based startup is leveraging emerging technologies to […]","categories":["AI Startups"],"tags":["AI Startups"],"author_name":"Ambika Choudhury","publish_date":"2021-03-30T13:00:00","publication_year":"2021","word_count":667,"keywords":["Go","machine learning","AI","neural network","computer vision","RAG","Python","Aim","analytics","R","AI Startups"],"extracted_tech_keywords":["AI","machine learning","neural network","computer vision","analytics","Aim","RAG","Python","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-startups\/the-hustle-of-minus-zero-the-punjabi-startup-building-affordable-fully-autonomous-evs-for-indian-roads\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10097318,"title":"OpenAI Introduces &#8216;Custom Instructions&#8217; for ChatGPT","content":"OpenAI just introduced a new feature called ‘custom instructions’, aimed at providing users with enhanced control over ChatGPT’s responses. By incorporating custom instructions, users will have the ability to use fewer prompts, making the interaction process more efficient and user-friendly as ChatGPT will now be able to remember your conversation context based on your chosen preferences, allowing for a more personalized and tailored AI interaction experience. ​​For instance, a teacher could indicate that they are teaching sixth-grade math, or a developer can specify their preferred programming language when seeking suggestions. Users can also provide information about their family size, enabling ChatGPT to provide relevant responses related to meals, grocery shopping, and vacation planning tailored to their specific needs. Starting today, the beta version of this feature will be accessible to users with the Plus plan, and it will gradually expand to include all users in the following weeks. Interestingly, this update coincides with OpenAI’s decision to double the number of messages that ChatGPT Plus customers can send to GPT-4 in a given time frame. Starting next week, users will be able to send up to 50 messages per 3 hours, further expanding the possibilities for AI-powered conversations. Notably, the company said it will use custom instructions to improve model performance for users. However, “you can disable this via data controls”, the company added. Plus users can start using custom instructions today by opting into the beta for custom instructions: On the web, click on your name → Settings → Beta features → opt into Custom instructions. Custom instructions will appear in the menu when you click on your name going forward. On iOS, go to Settings → New Features → turn on Custom instructions. Custom instructions will appear in settings.","excerpt":"OpenAI says it will use Custom Instructions to improve model performance for users","categories":["AI News"],"tags":[],"author_name":"Siddharth Jindal","publish_date":"2023-07-21T00:26:22","publication_year":"2023","word_count":290,"keywords":["Go","ChatGPT","OpenAI","AI","programming_languages:R","GPT","Aim","R","llm_models:ChatGPT","llm_models:GPT"],"extracted_tech_keywords":["AI","ChatGPT","OpenAI","Aim","R","Go","GPT","llm_models:GPT","llm_models:ChatGPT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/openai-introduces-custom-instructions-for-chatgpt\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":7389,"title":"Predictive Analytics, Healthcare and a Life much awaited","content":"Jyotika is 61 years old and lives in a small town in Lucknow and is suffering from Diabetes for the past 14 years. She is coping well with fair HB1AC values and lifestyle management. Recently, Jyotika felt that she is not able to handle sunlight well and feels that her vision is a bit blurred- she is speculating cataract and shows up at Dr. Narula’s Clinic. Dr. Narula is the Chief Ophthalmologist and has more than 35 years of experience and currently runs a Polyclinic out of Lucknow, UP. Jyotika gets successfully operated for Cataract and is discharged the same day. Within 13 days of discharge, Jyotika is admitted again but this time because of a Cardiac Disease. The family is speculating if this has something to do with the recent cataract surgery. Dr Reddy, Head of Department, Cardiology shows up and makes a comment to the family and Nurse- Well! Oxidative damage to proteins in the human lens is believed to be important in the etiology of age-related cataract. Because free radical-mediated oxidative damage to lipoproteins may accelerate atherosclerosis, the authors hypothesized that the development of cataract might be a marker for such damage and therefore might be associated with future risk of coronary heart disease (CHD). Mahesh, husband of Jyotika is not very happy with this. He believes that the doctors demonstrated clear negligence when Jyotika was admitted last time and provided no warning signs related to any possibility of a heart disease. Is there a way that this story would have been different and more assuring for the patient? Predictors like Physiological data (systolic blood pressure); Predictive Drugs usage and context (drug use of alpha blockers, beta blockers, beta agonists), Previous disease history (especially COPD, Lifestyle and environmental factors (age, smoking, and other coronary risk factors) can help determine risk for the population well in advance with Predictive analytics in action. Quick stats- In the United States, 1 in 4 women dies from heart disease. In fact, coronary heart disease (CHD)—the most common type of heart disease—is the #1 killer of both men and women in the United States. Also, a greater proportion of women (52 percent) than men (42 percent) with myocardial infarction die of sudden cardiac death before reaching the hospital. Advanced Predictive analytics companies like Enlightiks (based in India, US) have developed interesting risk stratification modules that can crawl of millions of health records in a population to understand patterns, trends, patient similarity metrics and natural language processing system that also takes into context unstructured data like Physician notes and discharge documents that are not often taken into account for analysis. Data goes through normalisation in a smart grid atmosphere and is then put into context to further do analysis based on simple algorithms like Framingham’s scale coupled with patient similarity study, isomorphism of patient graphs in a population, building cohorts and then aligning the data to the outcomes. This provides more complete and accurate understanding of each patient. While studying the Prospective study of cataract extraction and risk of coronary heart disease in women 60,657 women aged 45–63 years were followed. These women were without known coronary disease, stroke, or cancer in 1984.During 10 years of follow-up (674,283 person-years), the authors of this report (see source) documented 887 incident cases of CHD and 2,322 deaths. After adjustment for age, smoking, and other coronary risk factors, cataract extraction was significantly associated with higher risk of CHD (relative risk (RR) = 1.88, 95% confidence interval (CI): 1.41, 2.50) for total CHD, 2.44 (95% CI: 1.54, 3.89) for fatal CHD, and 1.63 (95% CI: 1.14, 2.34) for nonfatal myocardial infarction). The positive association between cataract extraction and total CHD was stronger among women with a history of diabetes (RR = 2.80, 95% CI: 1.77, 4.42) than among those without reported diabetes (RR = 1.51, 95 percent CI: 1.04, 2.18). In multivariate analyses, cataract extraction was associated with significantly increased overall mortality (RR = 1.37, 95 percent CI: 1.13, 1.66), which was entirely explained by the increased mortality from cardiovascular disease (RR = 1.84, 95% CI: 1.29, 2.64). These findings are compatible with current hypotheses relating oxidative damage and tissue aging to the development of cataract and CHD. Data mining and data analytics has been of immense importance to many different fields as we witness the evolution of data sciences over recent years. Biostatistics and Medical Informatics has proved to be the foundation of many modern biological theories and analysis techniques. These are the fields which applies data mining practices along with statistical models to discover hidden trends from data that comprises of biological experiments or procedures on different entities. In case of Jyotika if risk stratification was possible, it would have allowed advanced preventative measures and would have also help bring down healthcare costs especially for a country like India where 80% of healthcare expenditure is out of pocket. The claims and hospital admission would have been carefully planned or avoidable. Also, such measures would have led to better patient satisfaction. Currently at Enlightiks we are working on Predicting the unknown. We have presented several papers including the recent one on “Predicting Risk of Diabetes in Non-Diabetic Population”. Sometimes Indian healthcare providers do shy away from such advanced analytics and stick to basic Business intelligence tools. We have encouraging examples like Aarogyasri Health Care Trust, Mitra Biotech (Cancer research) as Indian case studies as to how to utilise data to build insights. Using healthcare data smart, securely and privacy-safe, it can bring a big boost to any healthcare system helping provide best quality, personalized healthcare at a much affordable cost. Credits: Cardiac Data Mining (CDM); Organization and Predictive Analytics on Biomedical (Cardiac) Data. Musa M Bilal, Masood Hussain, Iqra Basharat, Mamuna Fatima, AIP Conference Proceedings 01\/2013; 1559. DOI: 10.1063\/1.4825018 A Predictive Model for Readmission of Patients with Congestive Heart Failure: A Multi-hospital Perspective- Eric Zheng Prospective study of cataract extraction and risk of coronary heart disease in women. Hu FB1, Hankinson SE, Stampfer MJ, Manson JE, Colditz GA, Speizer FE, Hennekens CH, Willett WC. nhlbi.nih.gov, How Does Heart Disease Affect Women? http:\/\/www.cdc.gov\/media\/dpk\/2013\/images\/vitalSigns\/heart_disease\/img14.jpg Recent Insights in Coronary Artery Disease in Women- Biswakes Majumder, Azizul Haque, Dipankar Ghosh Dastidar Prevalence of risk factors for coronary artery disease in an urban Indian population- 2014- T Sekhri, R S Kanwar, R Wilfred, P Chugh, M Chhillar, R Aggarwal, Y K Sharma, J Sethi, J Sundriyal, K Bhadra, S Singh, N Rautela, Tek Chand, M Singh, S K Singh Prevalence of coronary artery disease and coronary risk factors in Kerala, South India: A population survey – Design and methods- Geevar Zachariah,a S. Harikrishnan,b M.N. Krishnan,g,∗ P.P. Mohanan,c G. Sanjay,d K. Venugopal,e K.R. Thankappan,f and The Cardiological Society of India Kerala Chapter Coronary Artery Disease and Its Risk Factors Prevalence (CSI Kerala CRP) Study Investigators","excerpt":"Jyotika is 61 years old and lives in a small town in Lucknow and is suffering from Diabetes for the past 14 years. She is coping well with fair HB1AC values and lifestyle management. Recently, Jyotika felt that she is not able to handle sunlight well and feels that her vision is a bit blurred- […]","categories":["IT Services"],"tags":["Analytics Case Study","healthcare analytics","Predictive AI"],"author_name":"Dr Ruchi Dass","publish_date":"2015-05-11T14:09:40","publication_year":"2015","word_count":1133,"keywords":["data science","Go","ELT","AI","R","healthcare analytics","RAG","Aim","analytics","Rust","Analytics Case Study","predictive analytics","Predictive AI"],"extracted_tech_keywords":["AI","data science","analytics","Aim","RAG","predictive analytics","R","Go","Rust","ELT"],"url":"https:\/\/analyticsindiamag.com\/it-services\/predictive-analytics-healthcare-and-a-life-much-awaited\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10039438,"title":"Building A Secure &#038; Scalable System – A Case Study","content":"Building a cloud-native application often requires us to focus on multiple aspects, in a holistic manner.  The creation of just a working application is not good enough to call it a well-built application.  In the past, the focus had always been more on functionality than on the critical aspects surrounding the application.  Giving enough importance to these activities can quickly become our differentiator in building a resilient application. Scalability & Security – two such critical aspects when building a software application, deserve their due respect from the design phase itself. Case Details A state-of-the-art application was built to administer vaccinations across nations – bug-free!  The primary purpose was met!  The superset of features that the system provides will cater to the needs of varying audiences across the globe, with configurable parameters.  All aspects of the system, including registrations, notifications, analytics, reporting were all built using latest technologies such as HTML5, Angular, Node.js, PowerBI, SQL and NoSQL DB’s, using microservices architecture that is hosted in AWS.  This included many services, including S3, ECS with Fargate, ECR, Cognito, ALB, Redis Cache, SNS, PinPoint, API Gateway, and sundry other services. During the initial phases, multiple discussions were held internally with architects who have deep architectural knowledge and were validated with AWS against a well-architected framework.  This resulted in healthy critiquing of various services chosen, including those that were purposefully omitted.  In general, architecting a system is more of an art, as decisions could be made to use either cloud-native services, or open-source systems, or a combination, including those available on AWS Marketplace or elsewhere.  Hence, a solid understanding of not only existing services but also their alternatives is needed to come up with a robust best-of-the-breed approach, to make the system scalable and secure, while keeping the costs at bay. Focus on scalability is a multi-pronged approach, including a selection of ECS with auto-scaling enablement.  ECS was chosen instead of EC2, and within ECS, Fargate was chosen as an alternative to EC2.  The serverless approach was part of cloud strategy and is a step towards building scalable architecture.  Appropriate limits were set up in ECS, so the instances would scale in\/out automatically based on pre-set parameters, such as CPU utilization.  Containers were created using Node.js code and they were developed as microservices, with ECS as the orchestrator of these services. With this approach, after basic testing, stress testing was needed to ensure it indeed scales to the ad-hoc needs of various customers.  Performance is not something that can be compromised. Hence, this testing included performance optimization when the system scales to the necessary level.  Load testing was performed using simulated threads from a few tools, and the system was stress-tested.  Initial scaling was at 100,000 concurrent writes per second (Read concurrency was multi-fold).  Through continuous optimization techniques, subsequently, the system scaled to perform at even higher concurrency levels.  After tweaking a few parameters, optimizing API’s, minification of front-end JavaScript code, and with many other initiatives, system performance has increased tremendously even when it scaled to a level that was beyond any customer’s realistic usage. Every aspect of the system was looked into in greater detail, to ensure the system warms up and scales to the needs of the customers when demand surges.  It is quite important to ensure the system scales in as well, when there is a slump in demand, to reduce overall TCO.  To ensure optimal performance at this scale, CloudFront was used as CDN and the Redis Cache engine was used to cache database transactions.  Over a period, we observed that over 60% of hits were from Redis than from the database. On the other hand, the security of the system is of paramount importance.  Any weak link will be exploited in no time.  Security from all aspects should be tackled – starting at the account level, with the principle of least privileges in mind, all the way to defending against DDoS attacks.  Web Application Firewall (WAF) was deployed to protect the system from vulnerabilities.  At the network level, the entire application was hosted in private subnets, with only one public subnet containing a bastion host. NACL’s and security groups were created to allow only the expected traffic at various ports for specific protocols.  An API Gateway was configured with appropriate rate limits, to ensure proper management of API calls to the containers hosting the application code.  This gives us flexibility in restricting traffic, while the load balancer helps us distribute the traffic via HTTPS, for a secure data transfer.  Other services, such as AWS Shield was used to protect against DDoS attacks.  GuardDuty was used to protect accounts, sensitive data and for threat detection, while AWS Inspector helped in protection against vulnerabilities and for compliance reasons. KMS helped us manage keys for application to connect to various services in a secure way.  All HTTP requests were routed to HTTPS, to ensure data in transit is always encrypted.  Encryption at OS and Database level was turned on, to take care of encryption of data at rest.  In addition to this, many application-level fixes were done to ensure the security of the overall application. Having a secure application also calls for solid instrumentation in place.  Proper logging and monitoring should be part of the overall design.  Logs were made accessible only to those that are authorized to view them.  They contain information pertaining to network, application and service levels.  VPC Flow Logs were enabled to monitor network activity, while application-level logs were refined not to store sensitive information and their access was controlled using IAM roles.  Similarly, service level logs were enabled to provide the health of various services and how they behave under various circumstances, to give complete insight into the functioning of the application.  CloudWatch was closely integrated, and proper alerts were created across the system to notify the support team through emails, using lambda functions, so the team can proactively work on remedies before they reach their actual limits. Outcome There had been some learnings, including the need to collaborate with the AWS team at times, when soft limits on certain services ought to be increased etc.  Similarly, based on the need of the application, we had to choose a combination of native and third-party services.  It is sometimes worth resorting to third-party services, though they might become a bit expensive, to meet certain goals of the overall system.  Implementation of automation where possible, certainly in the areas of DevOps and IaaC, is non-negotiable.  We all would run into situations where there will have to be multiple features released in short time and we end up seeing the need to spin-off different environments for various stakeholders, even for shorter periods.  Early prioritization of DevOps, IaaC, monitoring and instrumentation gave us flexibility to try a few new features and to quickly integrate third-party services. Laser focus on the above aspects from the design phase itself, on top of developing a functionally qualitative product, resulted in a highly scalable and secure system.  In fact, the application was later certified by independent and professional testing agencies that it was both secure and scalable to all foreseeable customer-base, for which it was intended to be used, thus vindicating the overall design and implementation of the system!","excerpt":"Building a cloud-native application often requires us to focus on multiple aspects, in a holistic manner.  The creation of just a working application is not good enough to call it a well-built application.  In the past, the focus had always been more on functionality than on the critical aspects surrounding the application.  Giving enough importance […]","categories":["AI Features"],"tags":[],"author_name":"Gopalakrishna M","publish_date":"2021-05-03T18:00:00","publication_year":"2021","word_count":1204,"keywords":["AWS","AI","ML","serverless","microservices","analytics","SQL","JavaScript","R","Redis"],"extracted_tech_keywords":["AI","ML","analytics","AWS","microservices","serverless","Redis","R","SQL","JavaScript"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/building-a-secure-scalable-system-a-case-study\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10132804,"title":"No One Killed Galactica","content":"Occasionally, the AI research field becomes heated, not due to new launches but because of debates about who is wrong or who is undermining each other’s research. This time, it was Yann LeCun pointing fingers on Gary Marcus, Grady Booch, and Michael J Black for killing Meta’s Galactica. It all started with the launch of Sakana AI’s AI Scientist, which the company claims is the first AI system for automating scientific research paper and open-ended discovery. This seems very similar to Galactica, an LLM trained on scientific knowledge for the scientific field. However, it was taken down because researchers were concerned about its outputs, which could be unreliable and prone to hallucination. “Will the AI negativists who killed Galactica by dousing it with vitriol kill this one too?” asked LeCun, further questioning Booch, Marcus, and Black, and others, “will rescind their prophecies of LLM-powered doom for the scientific community?” Source: X Just weeks after Galactica, ChatGPT was released. “Contrary to their predictions, the scientific publication system was not hurt by the availability of LLMs, let alone destroyed. If anything, it was actually helpful,” said LeCun. What are these ‘Prophecies of Doom’? Black, who is the director of Max Planck Institute for Intelligent Systems, had raised questions about the reliability of Galactica and similar products during the launch. Back in November 2022, Black posted a thread of all the problems with Galactica and said, “I applaud the ambition of this project but caution everyone about the hype surrounding it. This is not a great accelerator for science or even a helpful tool for science writing. It is potentially distorting and dangerous for science.” I applaud the ambition of this project but caution everyone about the hype surrounding it. This is not a great accelerator for science or even a helpful tool for science writing. It is potentially distorting and dangerous for science. (9\/9)— Michael Black (@Michael_J_Black) November 17, 2022 Black replied to LeCun saying that his criticism was genuine. “I’m sorry but the truth is that *YOU* killed it, not me,” said Black. He mentioned that LeCun promoted Galactica as, “Type a text and galactica.ai will generate a paper with relevant references, formulas, and everything,” which according to Black is a big claim to make. Thus, when Black conducted the test, Galactica fell short of his expectations. “It made up realistic looking references. And, it did so in a way that sounded authentic,” he added, saying that Meta did the right thing by taking it down as it did not live up to the hype around it. Marcus, a Professor of Psychology and Neural Science at NYU, is a popular critic of deep learning and AGI, also took to X to state that Galactica got his birthday, education as well as research interests wrong. Nearly 85% of the results presented by Galactica about Marcus were not true. Similarly, Booch had said, “Galactica is little more than statistical nonsense at scale,” adding adjectives like amusing, dangerous, and unethical. As for the ‘prophecies of doom’, Black had done a thorough analysis of the problem with LLMs then, and said there should be new ethics rules for scientific publication. He highlighted issues that there can be a flow of fake scientific papers, in an industry which completely relies on peer reviews and public trust. He said that he also uses LLMs for research and there should be further research on LLMs instead of slowing it down and building safeguards to defend the public trust in science. LeCun doesn’t agree. He replied to Black saying that the negative reaction from a scientist led to shutting down of Galactica. “In simple terms, *YOU* killed it,” said LeCun, asking him if the Galactica demo would have made the world a better place. So, Who Killed Galactica? Black believes that Meta, as an organisation, should have been more careful when releasing Galactica, even as a demo, and LeCun should accept the scrutiny faced for promotion of the product. “You will likely say that this was just a “demo” and not a “product”. Do you think the public really differentiates these when it comes from Meta, has a slick website, and is heavily hyped?” asked Black. Which makes sense. Since Meta released it as a software and not a research publication, it puts Meta under the spotlight, specifically when the claims were not adding up. Though Black appreciates LeCun and Meta’s efforts towards science and open source AI, he said that Meta is not a university, thus receives a lot of attention from the world, not just academia, which makes it riskier. Today, the weights of Galactica are still up for academic research, but the product is not available for direct use. “That was a good decision. It shifted the focus away from Galactica being a Meta “product” to it being a research prototype,” said Black. Maybe that is also why LeCun had invested in Perplexity AI, which often sells itself as a research focused search engine. Probably, Black is right. With the advancements in LLMs and people getting accustomed to them in the real world, shutting down Galactica back then sounded like a decent step to take. He didn’t kill Galactica, his feedback was strong enough that Meta had to kill it. However, now seems like a good time to finally bring back Galactica.","excerpt":"Now seems like a good time to finally bring back Galactica.","categories":["Deep Tech"],"tags":["AI in Research"],"author_name":"Mohit Pandey","publish_date":"2024-08-16T15:30:08","publication_year":"2024","word_count":886,"keywords":["Go","ChatGPT","TPU","AI","AI in Research","GPT","Aim","deep learning","Rust","GAN","R"],"extracted_tech_keywords":["AI","deep learning","ChatGPT","Aim","TPU","R","Go","Rust","GPT","GAN"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/no-one-killed-galactica\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10073029,"title":"Consumer Data in a Cookie-less Future","content":"With the changing dynamics of consumer-brand relations and the ever-increasing complexity of data—customers are now seeking personalised products and consider personalisation as an important differentiator. However, in the case of FMCG (Fast Moving Consumer Goods) brands, even if they sell their products in record numbers, they don’t really know who their customers are. Historically, FMCG brands have suffered because of lack of data. “Worldwide, FMCG brands face a strategic business problem across the consumer journey—from generating awareness to influencing purchase. At one end, FMCG brands are increasingly dependent on walled gardens such as Google, Facebook and Amazon to reach their consumers and generate awareness about their brands.” “At the other end, they are relying more and more on some of the biggest retailers in the market to help influence consumers’ purchase decisions at the point of sale,” said Lakshmana Gnanapragasam, Senior Vice President, Analytics–Epsilon APAC. In this context, Google and Apple’s announcement to phase out third-party cookies only adds to their troubles. Cookie-less future Brands often rely on cookies to track website visitors, collect data and target the right audience with advertisements. So, with Google’s announcement that it would block third-party cookies from Chrome in 2020, things changed significantly for FMCG brands. Lately, people have become more aware and have raised concerns about being tracked online by different parties—be it social media firms, advertisers or some other companies. “If digital advertising doesn’t evolve to address the growing concerns people have about their privacy and how their personal identity is being used, we risk the future of the free and open web,” Google said in a blog post. This, however, is a concern for FMCG brands, which rely primarily on third-party cookies. Third-party cookies have been around for over 25 years and banning them is expected to change digital marketing tremendously. “Now businesses are constantly worried about how to future proof their marketing and brand strategy for continuous growth. This is a real problem,” Lakshmana Gnanapragasam said. According to GetApp, as a result of Google’s announcement, 41 per cent of the marketers believe that it will become very challenging for them to track the right data. Nearly 44 per cent of them expect their expenditure to skyrocket from 5 per cent to 25 per cent. While this has only added to the brand’s troubles, this is not the end of the road. First-party data The pandemic gave digitalisation a significant boost. Brands are now aware that switching to a digital-first approach will help them in the long run. At present, progressive FMCG brands are investing in technological solutions to acquire first-party data on their consumers. First-party data offers brands greater insights into their customers. Brands could have data such as demographic information on their best buyers, gender, and the spending habits of buyers across locations. “Simply put, first-party data is information the brand owns and is a custodian of. This could be the data brands have on their own consumers and prospects. A good example is the information brands gather when consumers log in and browse their websites and apps.” “A solid first-party data enables brands to deeply understand their consumers and their purchase journeys and turn it into a competitive advantage, thus delivering higher returns on their marketing and media investments,” explained Lakshmana Gnanapragasam. Conversely, third-party data could prove to be expensive. The uncertainties associated with the accuracy, completeness and unknown origins of such data also make it challenging for brands working with them. Tech solutions In order to resolve such reliability and cost-related issues, brands are now turning to technology in their bid to acquire first-party data. This shift to technology solutions comes in the light of their importance in helping brands identify consumers visiting the brand’s websites and apps. Lakshmana Gnanapragasam elaborated through an example—“If a consumer is coming to your website or an app or your mobile site and the same consumer is going into a retailer’s website and purchasing your brand, there needs to be a way for us to identify that it is the same customer. For example, it could be the same customer who might visit a brand’s website to form multiple devices. But you need to know that these are all the same customers.” Investing in Identity Solutions will help brands overcome this challenge. For instance, Unilever invested in an online-offline consumer identity solution and onboarded millions of their active consumer records for improved online targeting. Brands are also investing in Consumer Data Platforms. Lakshmana Gnanapragasam said, “These customer data platforms would have customer ID and information around their web address, phone number and email. It would also have information like demographics, their transactions and engagement history, and all of that is consented data.” He added that brands are also investing in something called Clean Rooms, which is merely an embellished name for a non-PII database. Brands need to track the traffic on their websites and maintain a database of their engagement activity and what they do outside of their portal. However, these processes need to be privacy compliant. “And that’s why we’re calling it a Clean Room because they technically wouldn’t hold any personally identifiable information such as name, phone number, address, email id. But there will be a unique ID associated with that individual.” Another way of collecting data directly from the consumers is through incentivisation like scanning a product’s QR codes or uploading invoices for additional rewards and discounts. In turn, when a consumer scans a QR code, the brand can acquire data like phone numbers and email addresses to send personalised coupons, offers and other relevant information. Brands are also building digital communities of like-minded consumers to create a deeper connection with them and further influence them to become brand advocates. “Kellogg’s has a ‘Family Rewards’ programme in many countries across the world where millions of consumers engage with the brand through their loyalty programme. [Likewise] Pet food brand ‘Nestle Purina’ created a vibrant community of pet lovers and owners, thus pushing Purina to the top of their minds when it comes to the needs of their pets,” explained Lakshmana Gnanapragasam. Data analytics Collecting first-party data is one thing, but analysing and interpreting them in the best possible way is a whole another matter. “You need to have your data scientist and analysts log into an environment where all of this data is available, be able to create features on top of this data, and then create cohorts of customers that you can activate across multiple media channels.” Analysts will use the data for consumer segmentation and profiling. Brands also analyse the customers’ path to purchase. If the data is accurately analysed and interpreted, brands would be able to successfully map the path taken by consumers before they buy the brand’s products. Can AI help? “I wouldn’t say FMCG companies, as an industry, are as advanced as, let’s say, retailers or some media companies. But we’ve started to see that some of the FMCG companies that started investing in consumer data platforms and loyalty platforms over a period of time are beginning to use AI and ML to optimise the result,” explained Lakshmana Gnanapragasam. FMCG brands are constrained by a lack of data and have therefore witnessed inconsistent growth in terms of employing AI\/ML driven solutions. The emergence of DTC brands has also changed the landscape. These new brands are tech savvy compared to the legacy FMCG brands and could force existing brands to choose a digital-first approach or even modify their business model altogether. “We’ve seen some of the biggest brands like Unilever have taken on this journey and are investing heavily. Some of the mid-sized FMCG brands are also asking questions around what is our data strategy and how we can use it to get better at our media, even product development, identifying distribution opportunities, or even partnering with e-commerce and DTC companies.” “It is just a matter of time. With the right level of leadership, I think the FMCG brands are beginning to invest in this area,” said Lakshmana Gnanapragasam in conclusion.","excerpt":"Unilever invested in an online-offline consumer identity solution and onboarded millions of their active consumer records for improved online targeting.","categories":["IT Services"],"tags":[],"author_name":"Pritam Bordoloi","publish_date":"2022-08-18T18:00:00","publication_year":"2022","word_count":1336,"keywords":["Go","programming_languages:R","AI","ML","programming_languages:Go","Git","RAG","ViT","analytics","R"],"extracted_tech_keywords":["AI","ML","analytics","RAG","R","Go","Git","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/consumer-data-in-a-cookie-less-future\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10134095,"title":"How is Linux Powering the AI Moment?","content":"A few months ago, NVIDIA open sourced their drivers, starting with R515 for Linux. This solves one big problem of getting up and running a Linux system powered by an NVIDIA card. However, with users requesting this for years, what made NVIDIA finally open source them? The answer is simple. Most developers use Linux and drivers are the most stressful part of using the NVIDIA GPU. Not just developers, you will even find many big tech companies, including OpenAI and Google, rely on Linux to build their AI platform. TensorFlow, one of the most popular AI development libraries by Google, is best compatible with Ubuntu, a well-known Linux distribution. On Windows, however, you’ll need to use WSL (Windows Subsystem for Linux), which essentially acts as a virtual machine for running Linux. Even Google DeepMind, one of the leading AI research labs, previously used a modified version of Ubuntu, called Goobuntu, but later switched to Debian testing. These are both Linux-based operating systems. Further, IBM’s Watson, known for its natural language processing and machine learning capabilities, runs on SUSE Linux Enterprise Server. CUDA Performs Better on Linux It might come as a surprise, but many users have reported that CUDA performs better on Linux than Windows. Mike Mikowski, the manager of the Kubuntu Focus project, mentioned that NVIDIA GPU drivers are almost not faster on Windows, especially for deep learning solutions which use CUDA and OpenCL interfaces. NVIDIA has been using Ubuntu exclusively to demonstrate deep learning on all their edge solutions, which suggests Linux performs better for deep learning tasks compared to Windows. Meanwhile, a Reddit user reported that when he used CUDA on the same hardware but with different operating systems, including Windows and Ubuntu, the latter performed better. A reason behind this performance gain is Linux has a better GPU command scheduling than Windows. Linux Makes AI Dev Environment Setup a Breeze Despite the fact that Linux has a limited choice of software, you will find that a majority of AI developers use Linux. The key reason behind this is how well the software libraries are available and work well with Linux. For example, getting CUDA and CuDNN up and running is a hassle and not a seamless process on Windows. Compared to that, on Linux, everything can be managed via package manager without any issues. This is one of the reasons why most software development libraries and tools make their way to Linux first. A Reddit user, while answering why everyone uses Linux, mentioned that installing and setting up development software on Linux is a breeze thanks to a healthy array of package managers. “On Ubuntu or Debian-based systems, apt has most of what one will need to set a machine up. That, coupled with Conda, pip, and perhaps some additional software… it’s fast to install new libraries and programs, and dependencies are handled automatically,” he added further. Linux being a versatile operating system matches the target production system in most cases as most production workloads are based on Linux. A user on Stack Overflow mentioned that CUDA performs better on Linux, stating, “If you are looking at a performance point of view, and the time taken for a build, it would be best to use Linux.” When you combine better performance with CUDA and the ease of configuring the development environment, that too with other benefits like security, large community and support, there’s no doubt that Linux is powering the AI moment.","excerpt":"NVIDIA has been using Ubuntu exclusively to demonstrate deep learning on all its edge solutions which suggests Linux performs better for deep learning tasks.","categories":["Deep Tech"],"tags":["Linux","NVIDIA"],"author_name":"Sagar Sharma","publish_date":"2024-08-29T17:08:55","publication_year":"2024","word_count":576,"keywords":["CUDA","Go","machine learning","OpenAI","AI","ML","Ray","deep learning","Linux","NVIDIA","TensorFlow","R"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","OpenAI","Ray","TensorFlow","CUDA","R","Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/how-is-linux-powering-the-ai-moment\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10043361,"title":"Salesforce’s Deepak Pargaonkar On Cloud Acceleration In India","content":"Salesforce, a customer relationship management (CRM) solution provider, was founded in 1996. Analytics India Magazine caught up with Deepak Pargaonkar, VP, Solution Engineering, Salesforce India to understand more about the inner workings of the company. “Today we have 4000+ employees across Mumbai, Hyderabad, Delhi and Bangalore. Salesforce India, has the highest workforce outside of the US. As we grow, we will continue to invest in expanding our employee presence in the region making India a leading global talent and innovation hub for the company,” said Deepak. Deepak currently leads Solution Engineering at Salesforce India. Prior to Salesforce, Deepak has worked with organisations like Oracle, Cambridge Technology partners and Videocon. Excerpts: AIM: What’s your role at Salesforce? Deepak: I currently lead Solution Engineering at Salesforce – India. A part of the India leadership team at Salesforce, my key responsibilities involve creating compelling value propositions to support sales efforts in the region. Each customer has a unique set of challenges, most of which can be simplified by leveraging technology. We work closely with our customers to identify these complexities and build customised solutions. I have had the opportunity to work with various leading organisations in their journey of Customer Management over the last 15 plus years. In India, we are at the forefront of the revolutionary cloud computing wave, and I strongly believe that customer management is more about the organisation’s desire to build intimate relationships with customers and derive maximum business benefits – technology is an enabler to build these relationships. AIM: Tell us about the evolution of cloud computing in India, especially in the wake of the pandemic Deepak: The past year has truly taught us the meaning of digital transformation across every industry – particularly in terms of reimagining businesses to better serve customers and having the agility to respond to evolving circumstances. Companies leveraging cloud computing have shown us what it means to be resilient in times of crisis, and these migrations are only set to accelerate in 2021 – enabling scalability, availability and accessibility of information from any location. The digital imperative is now a priority for every CEO. Leaders are recognising the level of risk that their companies will be exposed to if they don’t have a digital strategy in place. Whether B2B or B2C, organisations need to accelerate their efforts to go all digital and build direct, trusted relationships with their customers. At least 28 percent of spending in key IT segments will shift to the cloud by 2022, as per a report by Gartner. Our customers such as Edelweiss continued to service customers seamlessly through the lockdown. Since adopting Salesforce, the company’s customer satisfaction rates have jumped from 78% to 92%. Teams are better able to monitor customer sentiment, personalise services, and resolve issues faster. A traditional brick and mortar educational institute NMIMS managed to pivot to a digital first university by adopting Salesforce. AIM: How important is the Indian market for Salesforce? Deepak: As we step into the new year we will continue to strengthen our vision of working together as One India, leading with relevance, integrity and trust, guiding our customers through their digital transformation journeys. Every business strategy over the last year was digital, making Salesforce more relevant than ever. Our priorities include: Customer centric digital transformation: Our customers have had to learn how to sell, service, market and more, from anywhere with the customer at the center of it all. We will continue to help companies transform by unlocking the value of their customer data, providing a single source of truth, while ensuring scale and confidence in the all-digital, work from anywhere world. Innovation on the Platform: India is a unique market with unique challenges. Innovation with confidence and scalability is essential. The Salesforce Platform offers this opportunity for individuals and businesses to solve problems for India and the world with easy-to-use flows, pre-written low-code blocks and much more. Our priority is to bring the power of this trusted, intelligent, integrated and sustainable platform to the customer building them up for success.Skilling and Upskilling with Trailhead: COVID-19 has accelerated the digital skills gap and companies everywhere are trying to pivot to go digital fast, while rebuilding and reskilling their workforces. In fact, a recent Salesforce study found that in India, there has been a heightened interest in online learning since COVID-19 with 83% of survey participants from India saying they are more interested in online learning or training. With the power of Trailhead, Salesforce’s online learning platform, we will empower both learners at every stage of their career journey and companies looking to reskill their employees with the platform, expertise, and guidance to skill up for the future. AIM: How is Salesforce reimagining Sales Cloud to drive growth? Deepak: Sales people need to learn how to cultivate relationships remotely and understand how different industries are reacting to the current state of the world.  There is an urgent need to have the right tools, processes, and agility to adapt and take hold of a new competitive landscape. Salesforce is reimagining Sales Cloud as a growth platform to provide everything a business needs to transform sales, from anywhere for anyone at every stage of growth. Even when salespeople get back on the road, 51% expect to travel less than they did pre-pandemic—so seamless, virtual interactions will continue to be a critical part of field sales. Salesforce Meetings makes meeting prep and wrap ups simple when you’re in back to back calls and Einstein Conversation Insights helps to understand trends in customer conversations and train sales teams. Moving forward, sales is going to be extremely data driven. Tableau aims to empower sales teams to analyze data and make smart decisions, faster, without relying on data scientists\/ experts. Most importantly, the pandemic was a clear reminder that successful businesses must be able to react in real-time to changing market conditions. MuleSoft Composer for Salesforce now allows business users to connect data and apps without writing code, such that, sales operations teams can connect apps and systems to Salesforce, automate sales processes, and unlock richer sales insights without waiting on development resources. Lastly, enablement and training will adapt to virtual too, so new reps can onboard quickly right within Salesforce from Trailhead. Sales people can see relevant education materials like competitor analyses surfaced right in their workspace. AIM: How does Salesforce keep up with the increasingly digital world? Deepak: Salesforce is built on innovation and agility and we are committed to providing tools to help organizations respond and recover from the challenges of COVID-19. For instance, our Work.com solution enables organisations to manage the logistics of returning to work while putting employee and visitor health and safety first. Vaccine Cloud is another great example that allows government agencies around the world to manage the deployment of vaccines in a fair and equitable way. Our industry strategy has been very focused over the years. Today, we have 12 industry clouds in financial services, healthcare and life sciences, public sector, communications, manufacturing, energy & utilities, consumer goods, media, education, and nonprofit. Within these clouds, we are also increasing capabilities in specific verticals, like insurance and healthcare payers. In India, we are seeing accelerated adoption across the financial services, Oil & Gas, education and manufacturing sectors. We will continue to accelerate our focus on re-imagining what technology can do for certain industries, creating an ecosystem of trailblazers, delivering the power of the Salesforce platform to drive business growth with scalability, agility and resilience.","excerpt":"Salesforce, a customer relationship management (CRM) solution provider, was founded in 1996. Analytics India Magazine caught up with Deepak Pargaonkar, VP, Solution Engineering, Salesforce India to understand more about the inner workings of the company. “Today we have 4000+ employees across Mumbai, Hyderabad, Delhi and Bangalore. Salesforce India, has the highest workforce outside of the […]","categories":["AI Features"],"tags":["Interviews and Discussions"],"author_name":"kumar Gandharv","publish_date":"2021-07-10T12:00:00","publication_year":"2021","word_count":1245,"keywords":["Go","AI","cloud computing","ML","Scala","RAG","Aim","analytics","Rust","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","ML","analytics","Aim","RAG","cloud computing","R","Go","Rust","Scala"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/salesforces-deepak-pargaonkar-on-cloud-acceleration-in-india\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10071055,"title":"Is Microsoft&#8217;s VS Code really open source?","content":"In June, Databricks open-sourced all the Delta Late APIs as part of the new Delta Lake 2.0 release. This put a definitive end to criticism from its competitors like Cloudera, Dremio, Google (Big Lake), Microsoft, Oracle, SAP, AWS Snowflake, HPE (Ezmeral) and Vertica, who had doubted whether Delta Lake was open source or proprietary. “The new announcement should provide continuity and clarity for users and help counter confusion (stoked in part by competitors) about whether Delta Lake is proprietary or open source,” said Matt Aslett, research director at Ventana Research. This is not the first time there has been an ambiguity regarding the open-source nature of a tool. The best example is Microsoft’s VS Code. Last month, Rukshan Ranatunge, Member – FHIR Implementation Architecture and Technical Advisory Group Ministry of Health – Sri Lanka, mentioned in a blog post that he is switching from VS Code to VS Codium. In the post, he talked about VS Code not being open source and also about what makes VS Code proprietary. Also, in June, Licio Lentimo, Cybersecurity Technical Mentor at CYDEO and software developer, tweeted, “VS Codium is the open-source alternative of VS Code. Interestingly, Microsoft’s VS Code is not fully open source as many previously thought.” What is VS Code? VS Code is a source-code editor made by Microsoft that runs on macOS, Linux, and Windows. VS Code has a set of integrated tools that a developer needs for a quick code-build-debug cycle. The source code editor supports development operations like debugging, task running, and version control. It can be used with several programming languages like Java, JavaScript, Node.js, Python, C++ and Fortran. VS Code is based on the Electron framework, which is used to develop web applications using HTML, CSS, etc. A prominent aspect of the source code editor is that it can be customised using extensions. These extensions support new programming languages, themes, and debuggers and perform static code analysis. With the built-in source control feature, users can access control settings and view changes made in the current project. Moreover, this feature also allows users to create repositories and make a push and pull requests directly from the Visual Studio Code program. How open is VS Code? Microsoft released the beta open-source version of VS Code in November 2015. At the same time, it open-sourced the VS Code repository. However, this does not mean that VS Code is open-source. Instead, it’s more accurate to say that VS Code is built on an open-source project called Code-Open Source Software (Code-OSS). Code-OSS is the core layer of VS Code. It is available on GitHub under the standard MIT License. The GitHub repository (Code-OSS) is where Microsoft develops the VS Code product. Here, the developers write code and modify it. They also publish their roadmap and monthly iteration in the GitHub repository. However, Microsoft VS Code is a Microsoft licensed distribution of ‘Code – OSS’ that includes Microsoft proprietary assets and features like Visual Studio Marketplace integration and telemetry system that are not available in Code-OSS. Thus, Microsoft follows an ‘open core model’ for VS Code and is not actually open-source. In an open core model, the company offers certain limited features that form the core of the product as free and open source (FOSS) software, while several add-on features are released as proprietary. Not just Microsoft but several other companies deploy the open core model. For example, Google built Chrome on Chromium, an open-source browser and then modified it to incorporate proprietary Google features, which is released as proprietary freeware. The same is true for the Oracle JDK, Xamarin Studio and JetBrains. These applications have been built on top of OpenJDK, MonoDevelop and IntelliJ, respectively. “Microsoft modifies VS Code in a way that a non-Microsoft VS Code fork can’t use extensions from the official Microsoft VS Code store. Not only that, some of the VS Code extensions developed and released by Microsoft will only work in the VS Code released by Microsoft and won’t work on non-Microsoft VS Code forks,” mentioned Ranatunge in his blog post. Microsoft has made similar moves in the past. It modified the open-source cross-platform IDE MonoDevelop as Visual Studio for Mac. The Visual Studio for Mac has three versions- for students, professionals and enterprises. While the students’ version is free and supports classroom learning, individual developers and small companies must log in via IDE to access the other versions. In 2021, Microsoft abruptly removed the Hot Reload functionality from the open-source .NET SDK, only to revoke it later as it had enraged the .NET community. Move towards VS Codium As stated, Microsoft follows an open-core model for VS code. Therefore, developers who want access to the full open source code that is MIT licensed will have to download the code from the repository and then build the VS code on their own. The task is cumbersome for most users. This is where VS Codium comes into play. VS Codium is fully open-source software binaries of VS Code licensed under the MIT license. With VS Codium, developers do not need to download and build from the source. Instead, the VS Codium team builds VS Code from the source repository and uploads the binaries to GitHub. “VS Codium is a clone of Microsoft’s Visual Studio Code. This project’s sole aim is to provide you with ready-to-use binaries without Microsoft’s telemetry code,” mentioned Abhishek Prakash, creator of It’s FOSS (a web portal focused on open source), in a blog. With telemetry tracking, developers are often flooded with unnecessary advertisements for premium versions of various extensions they use. VS Code gives users the option to install Microsoft and third-party extensions. Unfortunately, these extensions may be collecting their usage data, which cannot be disabled by disabling telemetry tracking.","excerpt":"In an open core model, the company offers certain limited features that form the core of the product as free and open source (FOSS) software.","categories":["Global Tech"],"tags":["Open Source AI","VS Code"],"author_name":"Zinnia Banerjee","publish_date":"2022-07-18T10:00:00","publication_year":"2022","word_count":957,"keywords":["AWS","AI","ML","RAG","Open Source AI","Python","Aim","Databricks","VS Code","JavaScript","R","Snowflake"],"extracted_tech_keywords":["AI","ML","Aim","RAG","AWS","Snowflake","Databricks","Python","R","JavaScript"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/is-microsofts-vs-code-really-open-source\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":46234,"title":"Vishal Sikka Raises $50 Million For His AI Startup Vianai From Undisclosed Investors","content":"Former Infosys chief Vishal Sikka’s startup Vianai has reportedly raised $50 million as seed funding from undisclosed investors. Sikka demonstrated a new AI platform vision last week during his keynote address at Oracle Open World. Vianai’s vision is to enable every company, in every industry, to utilize the explorable and explainable AI techniques. “Companies today have been formed and reshaped by the systems of yesterday,” said Sikka. “In the early days of personal computers, there were systems of record, then came systems of engagement. Today, we have the opportunity to create systems of intelligence that can bring AI techniques to transform all aspects of a business, and do so in a way that is transparent, explainable and explorable, inclusive and accessible to a massive number of developers. The Vianai platform helps us deliver on the promise of AI for every business and for all of us.” At the Oracle Open World, Vianai’s demo showcased a fully integrated and exploratory experience for building and deploying many AI projects per enterprise and many experiments per project, where each element of an experiment — data, feature, model and results — can be introspected, modified, and reused for sharing and ongoing experimentation. A key advance in the Vianai platform is an innovative domain specific language for data science and machine learning that allows any person who can write a standard mathematical expression to fully implement all aspects of an algorithm (models and parameters) in a highly simple, succinct and fully implemented, explorable specification. Vianai’s support for rapid exploration of the relationships between data, features, model and results, is joined to a highly visual design experience that will enable users to define and run dozens of experiments in the time it now takes to iterate through just one. In contrast to approaches such as AutoML, the platform ensures far better collaboration and alignment between data science practitioners and business stakeholders, ensuring that AI projects meet their goals.","excerpt":"Former Infosys chief Vishal Sikka’s startup Vianai has reportedly raised $50 million as seed funding from undisclosed investors. Sikka demonstrated a new AI platform vision last week during his keynote address at Oracle Open World. Vianai’s vision is to enable every company, in every industry, to utilize the explorable and explainable AI techniques. “Companies today have […]","categories":["AI News"],"tags":["vishal sikka"],"author_name":"Prajakta Hebbar","publish_date":"2019-09-23T12:16:34","publication_year":"2019","word_count":322,"keywords":["data science","Go","API","machine learning","funding","AI","R","ML","explainable AI","vishal sikka","startup"],"extracted_tech_keywords":["AI","machine learning","ML","data science","R","Go","API","explainable AI","startup","funding"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/vishal-sikka-raises-50-million-for-his-ai-startup-vianai-from-undisclosed-investors\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10001487,"title":"What Business Leaders Should Know About Shadow IoT &#038; The Growing Risk To  Organizations","content":"The world of Internet has changed significantly, from the content to the devices that connects, everything has undergone a phenomenal change since the advent of internet. World Wide Web turned 30 and in the last  30 years which is considerably a very small period, technology has advanced exponentially. Today, internet is home to billions of connected devices which is estimated to go over 20 billion devices by the year 2020 according to a research firm. The growing number of devices especially IoT’s pose a great threat. What is Shadow IoT The word “Shadow” is becoming quite popular in Information Technology(IT).  Shadow IT and Shadow IoT are two common terms that are often heard. Shadow IT refers to the IT applications or solutions that are used within organisations without explicit approval of the organisation whereas Shadow IoT are IoT devices that are connected to an organisation’s network without the consent or knowledge of the Organization. Both pose a great level of threat to the organisation. Both Shadow IT and Shadow IoT uses “Shadow” in the same context except that they both denote different entities. Another thing that’s common between the “two shadows” is the risk it poses to the organisation Shadow Begets Threat Shadow IoT’s challenges the cyber security of an organisation. As the number of unauthorised devices on a private network increases, the number of weak links also increases putting the entire organization’s data at risk. The recent Mirai botnet has already set an example on how IoT devices can be controlled to launch a cyber attack. There have also been many other IoT based attacks such as St. Jude Medical implantable cardiac devices attack, TRENDnet Webcam Hack etc. IoT devices mostly do not come with high or enterprise-grade security and are also hard to detect on an organisational network. It is impossible to control or deny the permission to connect when the devices are not visible, thus putting more challenge to the cybersecurity of the organisation. Reportedly there has been a 600% increase in IoT attacks within one year from 2017 to 2018. Besides being low on security and visibility in private networks, IoT devices are very good at showing their presence on the internet through search engines like Shodan which enables users to find details of identifiable devices, including the banner information, HTTP, SSH, FTP and SNMP services. This makes IoT devices easy to target in a wider network and hard to safeguard in a private network. How To Tackle Shadow IoT Here are possible ways to decrease the level of threats posed by Shadow IoTs according to a report by James A. Martin: 1. Un-Shadowing The initial step to safeguarding a private network from shadow devices is to make their presence in the private network known. Allowing users to officially connect their IoT devices to the organisation’s network can prevent shadow devices to a large extent. Chester Wisniewski, principal research scientist at Sophos, a security software and hardware company says, “The reason you have shadow IT and shadow IoT is often because the IT department is known for saying ‘no’ to requests to use devices like smart TVs” 2. Filtering The Network  for Shadow Devices Actively scanning the network for unauthorised devices in both the Organization’s network and beyond it would help discover shadow devices more easily. Wireless monitoring for shadow IoT devices and networks can allow visibility and asset management of those other devices and networks. 3. Isolating IoT Enabling employees to access a separate dedicated network within the organisation to connect IoT’s and IIoT’s will isolate the devices and creates a wall between the more private and important networks within the organization. Also configuring the network to enable IoTs to only transmit data and not receive incoming requests will add an extra layer of security. Into The Future IoT devices are sure to grow in numbers in the near future. With or without security these devices will thrive in the world of internet generating and transmitting data in real time. While the growth of data itself is a major concern, IoT will certainly add to it a factor of data security risk.","excerpt":"The world of Internet has changed significantly, from the content to the devices that connects, everything has undergone a phenomenal change since the advent of internet. World Wide Web turned 30 and in the last  30 years which is considerably a very small period, technology has advanced exponentially. Today, internet is home to billions of […]","categories":["AI News"],"tags":["cyber threats"],"author_name":"Amal Nair","publish_date":"2019-03-14T21:02:40","publication_year":"2019","word_count":687,"keywords":["Go","cyber threats","programming_languages:R","AI","programming_languages:Go","GAN","R"],"extracted_tech_keywords":["AI","R","Go","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/what-business-leaders-should-know-about-shadow-iot-the-growing-risk-to-organizations\/","complexity_score":2,"technical_depth":6,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":988,"title":"Portonics launches &#8220;Yogg X” smartwatch in India","content":"Portronics, a pioneer in Portable devices and gadgets, has enhanced its Wellness Series by launching “Yogg X”- a smart fitness watch with a touch-enabled detachable dial and a water proof design. Yoggx joins their Yogg Fitness tracker which was launched last year. With a detachable dial, Yogg X lets you track various activities such as– monitoring the distance covered, calories burned, steps taken, sleep patterns. It also shows time, date and battery status on the OLED screen. That’s not all; it can sync all the data with your smartphone via Bluetooth and alerts you when you achieve your goal. The smart watch maker claims it to be water proof adding that even if it accidently falls and stays in a bucket of water 1 meter \/ 3 feet deep for 30 minutes, it will work normally without any glitch. A design that integrates seamlessly in your daily active lifestyle, Yogg X allows you to: Set up to six daily reminders and Yogg X will silently vibrate on your arm without bothering others Shows you notifications about social media messages, SMS, emails, missed calls, reminders etc. Monitors your sleep pattern and display it the next day. Let’s you experiment with your evening habits like food, reading, watching TV, etc. and see the impact on quality of sleep. A fully charged 55mAh battery lives up to five days. Currently available in red and black color, its meager weight of 20 grams let you pull it off easily throughout the day. It is available both online and in offline stores for Rs 2,499. It is compatible with both iOS and Android devices.","excerpt":"Here comes yet another wearable technology for style and fitness conscious people.","categories":["AI News"],"tags":[],"author_name":"Srishti Deoras","publish_date":"2016-12-22T04:16:25","publication_year":"2016","word_count":269,"keywords":["Go","programming_languages:R","AI","ML","programming_languages:Go","Aim","ViT","R"],"extracted_tech_keywords":["AI","ML","Aim","R","Go","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/portonics-launches-yogg-x-smartwatch-india\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":51034,"title":"Why Ethereum Developer Virgil Griffith May Be Imprisoned For Giving A Blockchain Presentation","content":"Ethereum Foundation’s special projects researcher, Virgil Griffith has been arrested for allegedly violating the IEEPA (International Emergency Economic Powers Act) by giving a presentation on decentralised application and blockchain platform- Ethereum in North Korea. The United States Department of Justice in their statement said that Griffith, 36, travelled to North Korea to deliver a presentation on using cryptocurrency and blockchain tech to evade sanctions. According to experts, he may face prison time for this. The U.S. Attorney said that Virgil provided highly technical information to North Korea that can help it to launder money and evade sanctions from the U.S. In simple terms, the U.S doesn’t want North Korea to find a way to escape from under its thumb. These sanctions are something which keeps the pressure on North Korea. North Korea is deemed as the country which poses a real threat to the U.S national security and its allies, so it is concerned with every move that North Korea tries to make to cut loose from the U.S, even if it is a blockchain expert. Maybe it doesn’t sound like a big deal when one is scrolling through the headlines but, the blockchain technology and cryptocurrency may have a way to escape the sanctions imposed by the U.S, which is the reason why Venezuela and Iran have been working on to create a National Cryptocurrency. Ethereum Foundation Remains Indifferent To The Arrest Now what U.S attorney’s office says about Virgil Griffith is that he presented a decentralized solution and discussed a way for North Korea to resist the sanctions imposed upon it. The U.S department claimed that Griffith travelled to North Korea despite the State denying him permission to travel. This means that Griffith knew that travelling to North Korea and presenting the paper violates the sanctions against North Korea. In a statement, the Ethereum Foundation described Griffith’s travel to North Korea as a personal matter.” The statement said, “The Foundation is aware of the recent charges filed against Virgil Griffith. We can confirm that the Foundation was not represented in any capacity at the events outlined in the Justice Department’s filing and that the Foundation neither approved nor supported any such travel, which was a personal matter. We are continuing to monitor the situation as it develops.” Why Escape The U.S Control? So, we all know since World War 2 ended, U.S Dollar has been the world’s dominant reserve currency because of which the vast majority of cross-border trade is conducted in dollars. This gives the U.S immense power over other countries when it comes to politics and financial aspect. Now, losing access to dollar-clearing and correspondent relationships with banks in New York would mean a death sentence to most financial institutions, and in 2012 the U.S used this power or the leverage to oust Iran from the nuclear sanctions. Which meant Iran couldn’t transfer money across the globe which curbed Iran’s financial as well as nuclear ambitions. Because of the power, the U.S holds on account of its currency, emerging powers like China, with some Asia regulatory experts have been calling for an alternative international reserve currency. Now, to develop some resistance to these sanctions that the U.S imposed, countries like Iran, Russia, Venezuela are looking to build blockchain technology. Each country plans on claiming a statewide cryptocurrency to eventually dissolve the effects of Dollar. Now, what’s important is to know is that the leading tech behind this is open-source, free to use. For example, Iran announced this August that is was planning for a national cryptocurrency. Which will run on an open-source platform for enterprise blockchain system like Hyperledger Fabric. Hundreds of firms around the world are involved with Hyperledger blockchain projects. Another example is Venezuela.","excerpt":"Ethereum Foundation’s special projects researcher, Virgil Griffith has been arrested for allegedly violating the IEEPA (International Emergency Economic Powers Act) by giving a presentation on decentralised application and blockchain platform- Ethereum in North Korea. The United States Department of Justice in their statement said that Griffith, 36, travelled to North Korea to deliver a […]","categories":["AI Features"],"tags":["Blockchain","Cryptocurrency","north korea"],"author_name":"Sameer Balaganur","publish_date":"2019-12-03T14:00:00","publication_year":"2019","word_count":620,"keywords":["Blockchain","programming_languages:R","AI","Cryptocurrency","RAG","Aim","north korea","R"],"extracted_tech_keywords":["AI","Aim","RAG","R","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/ethereum-developer-virgil-griffith-imprisoned-blockchain\/","complexity_score":2,"technical_depth":5,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10086427,"title":"Great Firewall of China: Barrier to Baidu’s Global AI Aspirations","content":"Cashing in on the current AI wave, China-based tech firm Baidu is developing an AI-powered chatbot similar to OpenAI’s popular ChatGPT, Bloomberg reported. What’s more? Baidu plans to embed the chatbot into its search engine as early as March, much before ChatGPT is embedded into Microsoft Bing. While ChatGPT has been trained on OpenAI’s GPT3.5 architecture, Baidu’s chatbot has been trained on ERNIE (Enhanced Representation through Knowledge Integration) 3.0 Titan, a language model developed by the Chinese tech giant. GPT-3.5, which is a fine-tuned version of GPT-3, has 175 billion parameters, ERNIE 3.0 has 260 billion parameters. Interestingly, the initial version of ERNIE, which was launched in 2019, outperformed Google’s BERT, a pre-trained language representation model. Undoubtedly, ChatGPT has been a game-changer and has elevated the generative AI industry to new heights, but the media attention it receives is not unprecedented. The Wall Street Journal called Baidu’s chatbot ‘China’s entry into the global AI race’. However, for Baidu, there are a few stumbling blocks for it to fulfil its global AI ambitions. The Great Firewall of China Barrier In China, the government controls the internet. The Xi Jinping-led administration claims the authority to censor internet content within its jurisdiction. China is, in fact, one of the strictest censorship regimes in the world. From political dissent and discussion of sensitive historical events in China to information about its leaders and high-level officials, the government censors almost everything through the Great Firewall of China programme. Through the programme, the Chinese government blocks access to websites and online content that the government deems harmful, sensitive, or in violation of its regulations. So, the question is whether Baidu’s chatbot will also face the Chinese government’s censorship. Unfortunately, the answer to this mostly is yes. ERNIE-VilG, Baidu’s text-to-image AI generator, a competitor to DALL-E2 and Stable Diffusion, already rejects politically sensitive prompts. Chances are high that Baidu’s chatbot’s responses will toe the government’s line and this could prove to be a stumbling block for Baidu’s global ambitions. Additionally, OpenAI is continually refining ChatGPT through human feedback. It is utilised globally, whereas Baidu’s chatbot, if only employed within China, could result in a noticeably different training process or even biases for the outside world. Catering to China Only Baidu is the most-used search engine with a market share of 72.37%, as of May 2021. However, the Chinese search engine had a global market share of just 1.64% during that period, according to Statista. When it comes to the architecture, ERNIE 3.0 Titan has been pre-trained on heterogeneous Chinese data. However, since the chatbot will be available in both Chinese and English, it means the chatbot has been trained on English datasets as well. But for now, it’s difficult to assume how accurate Baidu’s chatbot will be, compared to ChatGPT. (Let’s not forget ChatGPT is also not highly accurate). Since Baidu’s chatbot and ChatGPT are trained on a completely different dataset, their responses too will vary. While it might cater to the Chinese audience, how successfully will it cater to the global audience remains to be seen.","excerpt":"Cashing in on the current AI wave, China-based tech firm Baidu is developing an AI-powered chatbot similar to OpenAI’s popular ChatGPT, Bloomberg reported. What’s more? Baidu plans to embed the chatbot into its search engine as early as March, much before ChatGPT is embedded into Microsoft Bing. While ChatGPT has been trained on OpenAI’s GPT3.5 […]","categories":["AI Features"],"tags":["AI Tool","baidu chatbot"],"author_name":"Pritam Bordoloi","publish_date":"2023-02-02T17:00:00","publication_year":"2023","word_count":510,"keywords":["Go","ChatGPT","DALL-E","OpenAI","AI","baidu chatbot","BERT","GPT","Aim","generative AI","AI Tool","R"],"extracted_tech_keywords":["AI","generative AI","ChatGPT","OpenAI","Aim","R","Go","BERT","GPT","DALL-E"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/great-firewall-of-china-barrier-to-baidus-global-ai-aspirations\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":45646,"title":"10 Open-Source Datasets To Learn Robotics","content":"Nowadays, researchers are striving to implement human-level understanding into robots. For instance, understanding the surrounding environment, hand movement, detecting and grasping objects and many more. In this article, we list down 10 open-source datasets that can be used to learn robotics in an advanced way. 1| Robot Arm Pushing Dataset The robot arm pushing dataset contains approximately 59,000 examples of robot pushing motions which including one training set (train) and two test sets. The first test set used two different subsets of the objects pushed during training. The second test set involved two subsets of objects. One key application of this dataset is to implement it in a learned model for decision making in vision-based robotic control tasks. 2| Robot Arm Grasping Dataset The robot arm grasping dataset contains approximately 650,000 examples of robot grasping attempts. The dataset was mainly created to train a large convolutional neural network (CNN) to predict the probability that the task-space motion of the gripper will result in successful grasps. 3| Dataset Of Daily Interactive Manipulation The dataset of daily interactive manipulation focuses on position, orientation, force, and torque of objects manipulated in daily tasks. It is a collection of 3D position and orientation (PO), force and torque (FT) data of tools\/objects being manipulated to fulfill certain tasks. It includes 1,603 trials of 32 types of daily motions and 1,596 trials of pouring alone along with the helper code. 4| Fukuoka Datasets For Place Categorisation Fukuoka datasets for place categorisation is a collection of several multimodal 3D datasets. It contains several datasets for the task of place categorisation using global information without object segmentation. It can also be used for other segmentation-based methods and can be extended with additional annotations. The datasets include indoor and outdoor scenarios obtained in different locations in Fukuoka city, Japan. 5| MultiDrone Public DataSet Aerial robots can be used for many robotic tasks, for instance, surveillance, monitoring, filming, among others. The MultiDrone public dataset has been collected using both pre-existing audiovisual material and newly filmed UAV shots. It includes more than 10 hours of UAV footage depicting football, rowing, and cycling (DW). A large subset of this dataset has been annotated for facilitating scientific research, in tasks such as visual detection and tracking of bicycles, football players, human crowds, etc. 6| Complex Urban Dataset With Multi-Level Sensors From Highly Diverse Urban Environments The complex urban dataset with multi-level sensors from a highly diverse urban environment captures various complex urban features and addresses the major issues of complex urban areas, such as unreliable and sporadic Global Positioning System (GPS) data, multi-lane roads, complex building structures, and the abundance of highly dynamic objects. In this dataset, typical LiDAR sensor data, 2D and 3D LiDAR, are provided. 7| Natural Language Instructions For Human-Robot Collaborative Manipulation Natural language instructions for human-robot collaborative manipulation is a dataset of natural language instructions for object reference in manipulation scenarios. It comprises of 1582 individual written instructions which were collected via online crowdsourcing. This dataset is particularly useful for researchers who work in natural language processing, human-robot interaction, and robotic manipulation. 8| Yale-CMU-Berkeley Dataset For Robotic Manipulation Research Yale-CMU-Berkeley dataset for robotic manipulation research is an image and model dataset of the real-life objects from the Yale-CMU-Berkeley Object Set, which is specifically designed for benchmarking in manipulation research. The dataset includes 600 high-resolution RGB images, 600 RGB-D images and five sets of textured three-dimensional geometric models. 9| Robot-at-Home Dataset Robot-at-home is a robotic dataset for semantic mapping of home environments. It is a collection of raw and processed sensory data from domestic settings which are used for semantic mapping algorithms through the categorisation of objects or rooms. The dataset contains 87,000+ time-stamped observations gathered by a mobile robot endowed with a rig of four RGB-D cameras and a 2D laser scanner. 10| Robotic 3D Scan Repository Robotic 3D scan repository is a data repository of standard 3D data sets for the robotics community, log files of robot runs and 3D point clouds from robotic experiments. One can easily use this data for developing SLAM or interpretation algorithms.","excerpt":"Nowadays, researchers are striving to implement human-level understanding into robots. For instance, understanding the surrounding environment, hand movement, detecting and grasping objects and many more. In this article, we list down 10 open-source datasets that can be used to learn robotics in an advanced way. 1| Robot Arm Pushing Dataset The robot arm pushing dataset […]","categories":["AI Trends"],"tags":["data mapping","open source datasets"],"author_name":"Ambika Choudhury","publish_date":"2019-09-08T14:00:18","publication_year":"2019","word_count":679,"keywords":["Go","programming_languages:R","AI","neural network","Modal","programming_languages:Go","ai_applications:robotics","data mapping","CNN","R","open source datasets"],"extracted_tech_keywords":["AI","neural network","R","Go","CNN","Modal","programming_languages:R","programming_languages:Go","ai_applications:robotics"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/10-open-source-datasets-to-learn-robotics\/","complexity_score":4,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10104968,"title":"Data Science Hiring and Interview Process at Marlabs","content":"Founded in 2011, New York-based IT services and consulting firm Marlabs helps companies of various sizes to undergo AI-powered digital transformation. It provides a wide range of services, including strategic planning, creating rapid prototypes in specialised labs, and applying agile engineering techniques to develop and expand digital solutions, cloud-based applications and AI-driven platforms. Marlabs’s data science team addresses a range of industry challenges, emphasising tasks like extracting insights from extensive datasets and employing pattern recognition, prediction, forecasting, recommendation, optimisation, and classification. Exploring Generative AI at Marlabs “In operationalising AI\/ML, we have tackled diverse projects, such as demand forecasting, inventory optimisation, point of sale data linkage, admissions candidate evaluation, real-time anomaly detection, and clinical trial report anomaly detection,” Sriraman Raghunathan, digital innovation and strategy principal, Marlabs, told AIM in an exclusive interaction. The team is also exploring generative AI applications, particularly in knowledge base extraction and summarisation across domains like IT service desk ticket resolution, sustainability finance, medical devices service management, and rare disease education. However, it is not developing foundational models as of now due to substantial capital requirements. “Instead, we are focussing on the value chain beyond foundational models, offering tools and practices for deploying such models within organisation boundaries, tailored for specific domains,” he added. Marlabs employs a variety of tools and frameworks depending on project specifics, utilising R and Python for development, Tableau, Power BI, QlikView for data exploration and visualisation, and PyTorch, TensorFlow, Cloud-Native tools\/platforms, and Jupyter Notebooks for AI\/ML model development. The team leverages Transformer models like GPT-3, especially in NLP use cases, implementing them in TensorFlow, and PyTorch, and utilising pre-trained models from Hugging Face Transformers Library. For generative AI, their toolkit includes LangChain, Llama Index; OpenAI, Cohere, PaLM2, Dolly; Chroma, and Atlas. Hiring Process The hiring process for data science roles at the organisation emphasises a blend of technical knowledge, practical application, and relevant experience. The initial steps involve a clear definition of the role and its requirements, followed by the creation of a detailed job description. The interview process comprises technical assessments, video interviews with AI\/ML experts, and HR interviews. Technical assessments evaluate coding skills, data analysis, and problem-solving abilities. Video interviews focus on the candidate’s depth of knowledge, practical application, and communication skills, often including a discussion of a relevant case study or project. HR interviews center around cultural fit, interpersonal skills, collaboration, and the candidate’s approach to handling challenges. Expectations “Upon joining the data science team, candidates can anticipate a thorough onboarding process tailored to their specific team, providing access to essential tools, resources, and training for a smooth transition,” commented Raghunathan. The company’s AI\/ML projects involve cutting-edge technologies, exposing candidates to dynamic customer use cases spanning natural language processing, computer vision, recommendation systems, and predictive analytics. The work environment is agile and fast-paced. The company places a strong emphasis on team collaboration and effective communication, given the collaborative nature of data science and AI\/ML projects. In this rapidly evolving field, the company expects new hires to demonstrate continuous learning, tackle complex technical and functional challenges, operate with high levels of abstraction, and exhibit creative and innovative thinking. Mistakes to Avoid “The most prevalent error observed in candidates during data science role interviews is a lack of clear communication,” he added. The ability to effectively communicate insights to non-technical stakeholders is crucial in the AI\/ML space, and this skill is frequently overlooked. Another common mistake is a failure to comprehend and articulate the business context and domain knowledge of the problem, which is essential in AI\/ML applications with significant business impact. Work Culture “We are recognised for our value-based culture focused on outcomes, emphasising a flat organizational structure to spur innovation and personal growth. Key values such as respect, transparency, trust, and a commitment to continuous learning are central to their ethos, all aimed at exceeding customer expectations,” he said. The company’s robust learning and development program has prepared over 150 young managers for leadership roles, with a strong emphasis on AI and technology for organisational insights and sentiment analysis. The company offers a comprehensive benefits package, including versatile insurance plans, performance incentives, and access to extensive learning resources like Courseware and Udemy, supporting a hybrid work model. Additionally, they provide mental health support and reward long-term employees based on tenure. Raghunathan further explained that in the data science team, Marlabs stands out for its innovative and collaborative environment, encouraging creativity and continuous learning. “This distinctive culture and investment in employee growth make us a leader in data science, differentiating it from competitors in the tech industry,” he added. Why Should You Join Marlabs? “Join Marlabs for a dynamic opportunity to work with a passionate team, using data to drive meaningful change. In this collaborative setting, data scientists work with brilliant colleagues across various industries, including healthcare, finance, and retail. You’ll tackle complex issues, contributing to significant business transformations. Marlabs supports your career with essential tools, resources, training, competitive compensation, benefits, and opportunities for professional growth and development,” concluded Raghunathan.","excerpt":"Marlabs is currently hiring for 10 data science roles, including ML Architect, ML Engineer, and Statistical Modeling positions.","categories":["AI Hirings"],"tags":["Data Science","Data Science Hiring","Top Trend"],"author_name":"Shritama Saha","publish_date":"2024-07-29T15:33:00","publication_year":"2024","word_count":830,"keywords":["Top Trend","data science","OpenAI","AI","ML","computer vision","NLP","Data Science Hiring","LangChain","Aim","analytics","generative AI","Data Science"],"extracted_tech_keywords":["AI","ML","NLP","computer vision","data science","analytics","generative AI","OpenAI","LangChain","Aim"],"url":"https:\/\/analyticsindiamag.com\/ai-hiring\/data-science-hiring-process-at-marlabs\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10039081,"title":"Meet skweak: A Python Toolkit For Applying Weak Supervision To NLP Tasks","content":"skweak is a software toolkit based on Python, developed for applying weak supervision to various NLP tasks. It has been recently introduced by Pierre Lison, Jeremy Barnes and Aliaksandr Hubin from Norway in April 2021 (research paper). Are you familiar with the term ‘weak supervision’? Have a look at its brief meaning before proceeding. Weak supervision refers to a novel ML technique that uses noisy, unstructured or limited data sources to label training data in a supervised learning approach. Instead of annotating the data manually, labelling functions created using existing knowledge of the domain annotate the data independently and hence eliminate the efforts and cost required for manual annotations. Overview of skweak skweak applies weak supervision to various NLP tasks such as sequence labelling and text classification. What it does can be summarized by the following steps: Apply a variety of labelling functions created based on the domain knowledge on a corpus.Aggregate the results of all the applied labelling functions in an unsupervised manner using a generative model. Any machine learning model can then be trained on the labelled corpus. Image source: Research paper Types of labeling functions used by skweak: Heuristics: The most straightforward way of labeling the corpus is using a heuristic approach which uses certain rules to decide upon the labels. For instance, if an entity ends with ‘Pvt Ltd’, ‘Inc’ etc. terms, it can be labelled as a “company”. (We will see an  example of such labelling in the practical implementation section further in this article). Gazetteers: This group of labelling functions searches a document for occurrences of specific words or phrases. It relies on a prefix tree called ‘trie’, which looks for all the possible occurrences and is traversed following the depth-first search method. Machine learning models: skweak can employ the concept of transfer learning, i.e. it can learn to label a corpus from an ML model and then use that knowledge for labeling the actual corpus. Document-level labeling functions: skweak can use the concept of label consistency in a document for labeling the whole corpus. For instance, frequently occurring terms are more likely to belong to a common label. Practical implementation Here’s a demonstration of using skweak for annotating a corpus having 200 news articles. The code has been implemented using Google colab with Python 3.7.10, skweak 0.2.9 and spacy 2.2.4 versions. Step-wise explanation of the code is as follows: Install skweak using pip command. !pip install skweak Install SpaCy library !pip install spacy Import required libraries and modules. import tarfile import spacy import skweak Download en_core_web_sm and en_core_web_md trained pipelines of SpaCy. !python -m spacy download en_core_web_sm !python -m spacy download en_core_web_md Extract the corpus’ text. (Data file used can be downloaded from here.) txt = []  #Create an array to store the text #Open the zip file of corpus arcv_file = tarfile.open(\"reuters_small.tar.gz\") #For each file in the zip file for arcv_mem in arcv_file.getnames(): #If the file is a text file, extract it, read its contents and decode it if arcv_mem.endswith(\".txt\"): text = arcv_file.extractfile(arcv_mem).read().decode(\"utf8\") #Add the content to storage array ‘txt txt.append(text) Load the en_core_web_sm pipeline and disable unnecessary components of the pipeline. pipeline = spacy.load(\"en_core_web_sm\", disable=[\"ner\", \"lemmatizer\"]) docs = list(pipeline.pipe(txt)) Create a list of cue words for other non-commercial organizations. otherorg = {\"University\", \"Institute\", \"College\", \"Committee\", \"Party\", \"Agency\", \"Union\", \"Association\", \"Organization\", \"Court\", \"Office\", \"National\"} Define a function to find companies’ names in the corpus text. def find_company(doc): #for each noun in the document for chunk in doc.noun_chunks: #If the noun ends with suffix like ‘corp’, ‘inc’ etc if chunk[-1].lower_.rstrip(\".\") in {'corp', 'inc', 'ltd', 'llc', 'sa', 'ag'}: #label the chunk as COMPANY yield chunk.start, chunk.end, \"COMPANY\" Create labelling function for companies detect_company = skweak.heuristics.FunctionAnnotator(\"company_detector\", find_company) Where, ‘company_detector’ is the name given to labelling function and ‘find_company’ is the function to be used for annotation Run the labelling function on the entire corpus docs = list(detect_company.pipe(docs)) Apply the labelling function on a small piece of text from the corpus. Display the annotated entities in a document using display_entities() method. skweak.utils.display_entities(docs[27], \"company_detector\") Condensed output: Define a function to find the names of non-commercial organizations in the corpus text and label them. def find_other_org(doc): #for each noun in the document for chunk in doc.noun_chunks: #if the noun is equal to one of the members of otherorg list if any([token.text in otherorg for token in chunk]): #label that chunk as OTHER_ORGANIZATION yield chunk.start, chunk.end, \"OTHER_ORGANIZATION\" Create a labelling function for other organizations. detect_other_org = skweak.heuristics.FunctionAnnotator(\"other_org_detector\", find_other_org) Where, ‘other_org_detector’ is the name given to the labelling function and ‘find_other_arg’ is the function to be used for annotation. Apply the labelling function to the corpus. docs = list(detect_other_org.pipe(docs)) Apply the above labelling function on a document. skweak.utils.display_entities(docs[28], \"other_org_detector\") Output: Create Gazetteers labelling function. First, we extract companies’ data from a JSON file available here. comp_data = skweak.gazetteers.extract_json_data(\"crunchbase_companies.json.gz\") #Labelling function gzt = skweak.gazetteers.GazetteerAnnotator(\"gazetteer\", comp_data) Run the gazetteer function on the whole corpus. docs = list(gzt.pipe(docs)) Apply the labelling function on a spacy document from the corpus. skweak.utils.display_entities(docs[28], \"gzt\") Condensed output: Run an NER (Named Entity Recognition) model trained on conll2003 dataset. ner_model = skweak.spacy.ModelAnnotator(\"spacy\", \"en_core_web_sm\") docs = list(ner_model.pipe(docs)) Apply the NER model on a document. skweak.utils.display_entities(docs[17], \"spacy\") Condensed output: Aggregation step We now aggregate the labels of different labelling functions using a generative model. This will create a unique annotation for each of the documents in the corpus. agg_model = skweak.aggregation.HMM(\"hmm\", [\"COMPANY\", \"OTHER_ORG\"]) Specify that “ORG” term can represent both a company or a non-commercial organization.agg_model.add_underspecified_label(“ORG”, [“COMPANY”, “OTHER_ORG”]) Fit the aggregated model on the corpus. docs = agg_model.fit_and_aggregate(docs) Output: Run the aggregated model on a document. skweak.utils.display_entities(docs[17], \"hmm\") Condensed output: Write the stream of documents in a binary file for document in docs: document.ents = document.spans[\"hmm\"] skweak.utils.docbin_writer(docs, \"reuters_small.spacy\") Train the final aggregated model on the labelled data. !spacy init config - --lang en --pipeline ner --optimize accuracy | \\ spacy train - --paths.train .\/reuters_small.spacy  --paths.dev .\/reuters_small.spacy \\ --initialize.vectors en_core_web_md --output reuters_small Sample condensed output: Code source: GitHubGoogle colab notebook of the above implementation References Research paperGitHub repositoryVideo tutorial","excerpt":"skweak is a software toolkit based on Python, developed for applying weak supervision to various Natural Language Processing tasks.","categories":["AI Trends"],"tags":["Natural Language Processing","NLP"],"author_name":"Nikita Shiledarbaxi","publish_date":"2021-04-28T14:00:00","publication_year":"2021","word_count":998,"keywords":["text classification","machine learning","TPU","AI","Natural Language Processing","ML","RAG","NLP","Ray","Colab","spaCy"],"extracted_tech_keywords":["AI","machine learning","ML","NLP","Ray","Colab","spaCy","RAG","text classification","TPU"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/meet-skweak-a-python-toolkit-for-applying-weak-supervision-to-nlp-tasks\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10075476,"title":"The ‘Unsolved’ Problems in Machine Learning","content":"While artificial intelligence and machine learning are solving a lot of real world problems, a complete comprehension of a lot of the “unsolved” problems in these fields is hindered due to fundamental limitations that are yet to be resolved with finality. There are various domains in the field of machine learning that developers dive deep into and come up with small incremental improvements. However, challenges to further advancement in these fields persist. A recent discussion on Reddit brought in several developers of the AI\/ML landscape to talk about some of these “important” and “unsolved” problems which, when solved, are likely to pave the way for significant improvements in these fields. Uncertainty prediction Arguably, the most important aspect of creating a machine learning model is gathering information from reliable and abundant sources. Beginners in the field of machine learning, who formerly worked as computer scientists, face the difficulty of working with imperfect or incomplete information—which is inevitable in the field. “Given that many computer scientists and software engineers work in a relatively clean and certain environment, it can be surprising that machine learning makes heavy use of probability theory,” said Andyk Maulana in his book series—‘Adaptive Computation and Machine learning’. Three major sources of uncertainty in machine learning are: Presence of noise in data: Observations in machine learning are referred to as “sample” or “instance” that often consist of variability and randomness which ultimately impact the output.Incomplete coverage of the domain: Models trained on observations that are by default incomplete as they only consist of a “sample” of the larger unattainable dataset.Imperfect models: “All models are wrong but some are useful,” said George Box. There is always some error in every model. Check out a research paper by Francesca Tavazza on uncertainty prediction for machine learning models here. Convergence time and low-resource learning systems Optimising the process of training and then inferring data requires a large amount of resources. The problems of reducing the convergence time of neural networks and requiring low-resource systems are countering each other. Developers might be able to build tech that is groundbreaking in applications but requires huge amounts of resources like hardware, power, storage, and electricity. For example, language models require vast amounts of data. The ultimate goal of reaching human-level interaction in the models requires training on a massive scale. This means a longer convergence time and requirement of higher resources for training. A key factor in the development of machine learning algorithms is scaling the amount of input data that, arguably, increases the accuracy of a model. But in order to achieve this, the recent success of deep learning models shows the importance of stronger processors and resources, thus resulting in continuous juggling of the two problems. Click here to learn how to converge neural networks faster. Overfitting Recent text-to-image generators like DALL-E or Midjourney showcase possibilities of what overfitting of input and training data can look like. https:\/\/twitter.com\/hausman_k\/status\/1511732395011194885?lang=en Overfitting, also a result of noise in data, is when a learning model picks up random fluctuations in the training data and treats them like concepts of the model resulting in errors and impacting the model’s ability to generalise. To counter this problem, most non-parametric and non-linear models include techniques and input guiding parameters to limit the reach of learning of the model. Even then, in practice, fitting a perfect dataset into a model is a difficult task. Two suggested techniques to limit overfitting data are: Using resampling techniques to gauge model accuracy: ‘K-fold cross validation’ is the most popular sampling technique that allows developers to train and test models several times with different subsets of training data.Holding back validation dataset: After tuning the machine learning algorithm on the initial dataset, developers input a validation dataset to achieve the final objective of the model and check how the model would perform on previously unseen data. Estimating causality instead of correlations Causal inferences come to humans naturally. Machine learning algorithms like deep neural networks are great for analysing patterns in huge datasets but struggle to make causal inferences. This occurs in fields like computer vision, robotics, and self-driving cars where models—though capable of recognising patterns—do not comprehend physical environmental properties of objects, resulting in making predictions about the situations and not actively dealing with novel situations. Researchers from Max Planck Institute for Intelligent Systems along with Google Research published a paper—Towards Causal Representation Learning, which talks about the challenges in machine learning algorithms due to the lack of causal representation. According to the researchers, to counter the absence of causality in machine learning models, developers try to increase the amount of datasets on which the models are trained, but fail to understand that this eventually leads to models recognising patterns and not independently “thinking”. The introduction of “inductive bias” into models is believed to be a step towards building causality into machines. But that, arguably, can be counter productive in building AI that is free of bias. Help! What precisely is \"inductive bias\"? Some ML researchers are in the opinion that the machine learning category of ‘inductive biases’ can allow us to build a causal understanding of the world. My Ladder of Causation says: \"This is mathematically impossible\". Who is right? 1\/— Judea Pearl (@yudapearl) February 14, 2021 Reproducibility AI\/ML being the most promising tool in almost all fields has resulted in many newcomers diving straight into it without fully grasping the intricacies of the subject. While reproducibility or replication is a combined outcome of the above mentioned problems, it still poses great challenges for newly developing models. Due to lack of resources and reluctance to conduct extensive trials, many of the algorithms fail when tested and implemented by other expert researchers. Big companies offering hi-tech solutions do not always publicly release their codes, making new researchers experiment on their own and propose solutions for large problems without rigorous testing, thus lacking reliability. Click here to find out about how lack of reproducibility in machine learning models is making the healthcare industry risky.","excerpt":"Uncertainty, probability, infinite-datasets, lack of causality are only few of the several challenges in machine learning.","categories":["AI Features"],"tags":["AI (Artificial Intelligence)","AI Tool","ML","ml algorithms","ML models","Neural Networks","Robotics"],"author_name":"Mohit Pandey","publish_date":"2022-09-22T10:00:00","publication_year":"2022","word_count":997,"keywords":["artificial intelligence","ml algorithms","machine learning","AI","neural network","TPU","ML","ML models","Robotics","computer vision","RAG","deep learning","AI Tool","R","AI (Artificial Intelligence)","Neural Networks"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","neural network","computer vision","RAG","TPU","R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/the-unsolved-problems-in-machine-learning\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10067414,"title":"How to use simulated annealing to approximate a function?","content":"Simulated annealing is an uninformed search technique to optimise the global minima of a function. It has become a popular approach over the last two decades due to its ease of implementation, convergence qualities, and use of hill-climbing motions to escape local optima. It is primarily used to solve discrete optimization issues and, to a lesser extent, continuous optimization problems. In this article, we will be discussing simulated annealing with a small implementation to understand the working of this technique. Using this method, we will find out the minima of a given function. Following are the topics to be covered. Table of contents About Local optimizationAbout Simulated AnnealingWorking of Simulated AnnealingImplementing Simulated Annealing to optimise a function Simulated Annealing is a meta-heuristic local search algorithm capable of escaping from local optima. Let’s start by understanding Local Optimization. About Local optimization A combinatorial optimization issue is defined by defining a collection of solutions and a cost function that assigns a numerical value to each option. An optimum solution is one that has the lowest possible cost (there may be more than one such solution). Local optimization aims to improve on an arbitrary solution to such a problem by a series of incremental, local improvements. To begin defining a local optimization algorithm, provide a way for perturbing solutions in order to acquire various solutions. The neighbourhood of ‘A’ refers to the set of solutions that may be acquired in one such step from a given solution ‘A’. The algorithm then executes the basic loop, leaving the precise techniques for selecting a solution (S) and untested neighbour (S’) to implementation details. Although S does not have to be the best solution after the loop is completed, it will be locally optimal in the sense that none of its neighbours has a lower cost. The hope is that locally optimum will work. Below is the image with a set of steps for calculating the local optimization. Analytics India Magazine Are you looking for a complete repository of Python libraries used in data science, check out here. About Simulated Annealing Simulated annealing is named after the physical annealing process with solids, in which a crystalline solid is heated and then gently cooled until it obtains the most regular possible crystal lattice structure and is, therefore, free of crystal flaws. The final arrangement results in a solid with higher structural integrity if the cooling schedule is suitably slow. Simulated annealing connects this sort of thermodynamic behaviour to the search for global minima in a discrete optimization problem. It also presents an algorithmic method for leveraging such a relationship. The values for two solutions (the present solution and a freshly picked solution) are compared at each iteration of a simulated annealing process applied to a discrete optimization problem. Improving solutions are always accepted, whereas a subset of non-improving (inferior) solutions are allowed with the goal of escaping local optima and reaching global optima. The likelihood of accepting non-improving solutions is determined by a temperature parameter, which is normally non-increasing with each algorithm iteration. The essential algorithmic aspect of simulated annealing is that it allows for the escape of local optima via hill-climbing steps (i.e., moves which worsen the objective function value). As the temperature parameter approaches zero, hill climbing moves become less common, and the solution distribution associated with the inhomogeneous Markov chain that models the algorithm’s behaviour converges to a form in which all probability is concentrated on the set of globally optimal solutions (provided that the algorithm is convergent; otherwise the algorithm will converge to a local optimum, which may or may not be globally optimal). Working of Simulated Annealing Several definitions are required to characterise the special aspects of a simulated annealing approach for discrete optimization problems. Consider a collection of all potential solutions, often known as the solution space. Then, on the solution space, create an objective function. The purpose is to determine the global minimum that exists in the set of solutions. To verify that a global minimum exists, the goal function must be constrained. For the collection of solutions, define a neighbourhood function. As a result, neighbouring solutions are associated with each solution and can be obtained in a single iteration of a local search algorithm. Simulated annealing begins with a solution chosen from a collection of solutions. A neighbouring solution is subsequently created, either at random or according to a predefined rule. The Metropolis acceptance criterion is used to represent how a thermodynamic system progresses from the current solution or state to a candidate solution with the lowest energy content. Based on the acceptance probability, the candidate solution is chosen as the current solution. This acceptance probability is the fundamental building block of the search process in simulated annealing. If the temperature is gradually lowered, the system can approach an equilibrium (stationary state) with each repetition. The objective functions are specified, and the energy associated with solutions is denoted. The equilibrium is determined by the Boltzmann distribution. It might be defined as the likelihood of the system being in a state with an energy function (the objective function) at a certain temperature, where the probability is proportional to the total of all conceivable functions. If the likelihood of creating a candidate solution from the solution’s neighbours is one, then a non-negative square stochastic matrix with transition probabilities may be created. These transition probabilities define a set of solutions produced by an inhomogeneous Markov chain. Implementing Simulated Annealing to optimise a function Let’s implement simulated annealing by computing a blind search to optimise the minima of the function and observe the progress. Import necessary libraries import numpy as np import matplotlib.pyplot as plt Defining the objective function for which the optimization has to be performed. We’ll use a straightforward one-dimensional objective function with boundaries [-10,10] with the optima at 0.0. def objective(x): return x[0]**2.0 r_min, r_max = -10.0, 10.0 inputs = np.arange(r_min, r_max, 0.1) results = [objective([x]) for x in inputs] plt.plot(inputs, results) x_optima = 0.0 plt.axvline(x=x_optima, ls='--', color='red') plt.show() Analytics India Magazine Following that, we may have a clearer understanding of how the metropolitan acceptance criterion changes with temperature throughout time. The criteria are temperature dependent, but it is also dependent on how different the objective evaluation of the new point is from the present working solution. As a result, we’ll plot the criterion for a few distinct “differences in objective function value” to show how they affect acceptance probability. iterations = 120 initial_temp = 12 iterations = [i for i in range(iterations)] temperatures = [initial_temp\/float(i + 1) for i in iterations] differences = [0.01, 0.1, 0.5, 1.0] for d in differences: metropolis = [np.exp(-d\/t) for t in temperatures] label = 'diff=%.2f' % d plt.plot(iterations, metropolis, label=label) plt.xlabel('Iteration') plt.ylabel('Metropolis Criterion') plt.legend() plt.show() Analytics India Magazine The plot is divided into four lines to represent the four disparities between the new poorer option and the current functioning solution. As one may predict, the wider the disparity, the less likely the model is to accept the worse option regardless of algorithm repetition. We can also see that the chance of accepting inferior answers reduces with algorithm iteration in all circumstances. Let’s build a simulated annealing function def simulated_annealing(objective, bounds, n_iterations, step_size, temp): best = bounds[:, 0] + np.random.rand(len(bounds)) * (bounds[:, 1] - bounds[:, 0]) best_eval = objective(best) curr, curr_eval = best, best_eval scores = list() for i in range(n_iterations): candidate = curr + np.random.randn(len(bounds)) * step_size candidate_eval = objective(candidate) if candidate_eval < best_eval: best, best_eval = candidate, candidate_eval scores.append(best_eval) print('>%d f(%s) = %.5f' % (i, best, best_eval)) diff = candidate_eval - curr_eval t = temp \/ float(i + 1) metropolis = np.exp(-diff \/ t) if diff < 0 or np.random.rand() < metropolis: curr, curr_eval = candidate, candidate_eval return [best, best_eval, scores] Now that we know how the temperature and metropolis acceptance criterion behave over time, we can use simulated annealing for our test problem. np.random.seed(10) bounds = np.asarray([[-10.0, 10.0]]) n_iterations = 1000 step_size = 0.1 temp = 12 best, score, scores = simulated_annealing(objective, bounds, n_iterations, step_size, temp) print('Completed...') print('f(%s) = %f' % (best, score)) Analytics India Magazine Start by seeding the pseudorandom number generator. This is not essential in general, but it is necessary in this case to ensure that the same sequence of random numbers is produced each time the method is performed so that the results may be plotted afterwards. The algorithm will search for 1,000 iterations with a step size of 0.1 in this example. The starting temperature will be 12.0. Because the search technique is more sensitive to the annealing schedule than the beginning temperature, initial temperature settings are nearly arbitrary. Plotting the progress report plt.figure(figsize=(10,5)) plt.plot(scores, '.-') plt.xlabel('Improvement Number') plt.ylabel('Evaluation f(x)') plt.show() Analytics India Magazine During the hill-climbing search, a line plot is constructed to display the objective function assessment for each improvement. During the search, we can see roughly 78 changes to the objective function evaluation, with substantial changes at first and very modest to unnoticeable changes near the conclusion as the algorithm settled on the optima. Final Verdict Simulated annealing optimization methods may be used to solve both single-objective and multi-objective optimization problems. It’s been used in disciplines as diverse as process system engineering, operations research, and smart materials. With this hands-on implementation article, we have understood the working of Simulated annealing to optimise a function approximation. References Link to the above codePaper on Simulated Annealing","excerpt":"Simulated annealing is an uninformed search technique to optimise the global minima of a function.","categories":["AI Trends"],"tags":["AI Tool","discrete","markov chain","optimisation algorithms"],"author_name":"Sourabh Mehta","publish_date":"2022-05-19T10:00:00","publication_year":"2022","word_count":1561,"keywords":["data science","NumPy","discrete","AWS","AI","markov chain","RAG","Python","Ray","Aim","optimisation algorithms","analytics","AI Tool","Matplotlib"],"extracted_tech_keywords":["AI","data science","analytics","Aim","Ray","NumPy","Matplotlib","RAG","AWS","Python"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/how-to-use-simulated-annealing-to-approximate-a-function\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10139290,"title":"NVIDIA Partners with F5 to Lead Sovereign AI Cloud","content":"At the NVIDIA AI Summit 2024 in Mumbai, the company announced its partnership with F5, which specialises in network security and multi-cloud management. The goal is to enhance AI efficiency and security in sovereign cloud environments by integrating NVIDIA BlueField-3 DPUs with F5 BIG-IP Next for Kubernetes, an open-source system that automates software development and management. “This partnership is ideal for industries with strict data governance, privacy or compliance requirements and addresses the growing demand for scalable AI infrastructure,” said Ash Bhalgat, senior director, of cloud, telco and cybersecurity market development in NVIDIA’s official blog. Regarding this partnership, Ahmed Guetari, VP, general manager and service provider at F5 said, “We’re working with NVIDIA to enable industries to deploy scalable, secure AI solutions faster, with better performance, all while ensuring data remains protected.” NVIDIA BlueField-3 BlueField-3 networking platform is a solution designed to boost data centre performance and security. It combines networking, storage, and security capabilities into one system, powered by accelerated computing. NVIDIA BlueField-3, Source: NVIDIA Blog BlueField-3 uses SmartNIC technology and has built-in security features, such as hardware-based isolation and encryption, to protect sensitive data from potential threats. The platform is also highly scalable, handling a variety of workloads, from traditional organisational applications to modern cloud-based environments. This partnership with F5 accelerates BlueField-3 and enhances performance, security, and operational efficiency in modern data centres. The Sovereign Cloud Approach Sovereign clouds are essential for companies that maintain sensitive information, such as those in the telecommunications and financial services industry. With this new NVIDIA collaboration with F5, organisations can now achieve a secure AI networking infrastructure that maintains data privacy. The integration allows for efficient AI task processes and reduces energy consumption, optimising GPU resource use. According to the announcement, the sovereign cloud market is expected to grow significantly and reach $250 billion by 2027. In the meantime, the market for foundation models will be $30 billion by 2027. This efficient AI workload management also claims to help NVIDIA NIM microservices, which accelerate the setting up of foundation models.NVIDIA had also previously partnered with VMware Cloud Foundation to enable enterprises to customise models and run generative AI applications, including intelligent chatbots, assistants, search and summarisation.","excerpt":"NVIDIA integrates BlueField-3 DPUs with F5’s BIG-IP Next for Kubernetes","categories":["AI News"],"tags":["F5","NVIDIA","partnership"],"author_name":"Sanjana Gupta","publish_date":"2024-10-24T13:32:18","publication_year":"2024","word_count":364,"keywords":["partnership","Go","AI","chatbots","RAG","microservices","Aim","F5","generative AI","foundation models","NVIDIA","R","kubernetes"],"extracted_tech_keywords":["AI","generative AI","foundation models","Aim","RAG","chatbots","kubernetes","microservices","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/nvidia-partners-with-f5-to-lead-sovereign-ai-cloud\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":64776,"title":"Clari5 Shares Some Best Practices For Businesses To Overcome This Crisis","content":"Clari5, a Bengaluru-based fintech company, has shared some of the best practices for companies to follow in order to overcome this unprecedented challenge. Clari5, who has been an expert in providing financial crime risk management and customer experience management solutions to the BFSI industry, shared these practices to help businesses deal with this never-before-of-its kind challenge. According to the company — “Despite the negative news and noise all around about how the pandemic has been impacting most firms and systems worldwide, we are having some very positive experiences.” The company further stated, “We hit the ground running, with close to zero disruptions to any of the planned tasks and unplanned tasks despite the sudden call for a lockdown. We ensured that our customer engagements, project progress and services delivery continued with zero disruption. In fact, we were back to ready-mode within 48 hours of the nation-wide lockdown announcement.” To get the businesses running, the company has stepped forward to help its employees work from home. “Our employees across all functions were immediately enabled to ‘work from far’, with constant real-time communication between team members. Secure remote access to our technology infrastructure is keeping them in sync for all the ongoing product maintenance and support tasks,” stated the company note. The company has further developed a cloud-based solution to make the whole transition to remote working smoother for their employees. “Our cloud-based, company-wide project management tool has been helping us work seamlessly and virtually during this phase by providing a real-time view of all engagements and deliverables, assigning or monitoring tasks, reviewing deliverables.” The company has further stated some of its noteworthy achievements during the lockdown: A leading Payments Bank went live with the company’s real-time, cross-channel enterprise fraud and financial crime management platform — Clari5.The company kicked off two fraud management projects during this phase — one for a large South-East Asian bank and the other one for a small finance bank.The company performed a remote infrastructure fine-tuning for one of its client banks.It conducted several DRG (Detailed Requirement Gathering) sessions for their client banks.It also conducted a successful Data Center Disaster Recovery drill as a pro-active measure, just in case there may be any potential outage or downtime situation. The company, further, recognises its partnership with client banks — “Of course, our client banks have been playing a vital role in ensuring BAU by providing us secured remote access and joining in on virtual meetings.” “We have also been reaching out to them on a daily basis via an initiative called BAU@Clari5 to know how we can be of more help to them with respect to our product which they have implemented. Alongside, we also wish to help in areas that may not be directly related to us, but are within the domain of our business and technology competencies.” The company has also worked on online campaigns for their clients. “Interestingly, some of our client banks have been reaching out to us to explore how they can use artificial intelligence and machine learning for bettering fraud detection capabilities.” Alongside, the company has also been organising webinars to empower their clients and customers with knowledge. “We have been conducting informative webinars called ‘Combating another kind of outbreak’ on newer types of fraud emerging during the COVID-19 outbreak, how banks will have a Himalayan task when the situation returns to normal, and what are the preventive measures banks can take.” The first session of the webinar had participants from 10 countries, despite it being a Good Friday holiday on that day. The company has also announced that they have been continuing their hiring process during this phase. “In short, the pandemic lockdown has been a litmus test of our capabilities, and from our experience, it has actually made us do more,” the company concluded.","excerpt":"Clari5, a Bengaluru-based fintech company, has shared some of the best practices for companies to follow in order to overcome this unprecedented challenge.  Clari5, who has been an expert in providing financial crime risk management and customer experience management solutions to the BFSI industry, shared these practices to help businesses deal with this never-before-of-its kind […]","categories":["AI News"],"tags":["covid-19","how to implement business intelligence","Industry best practices","Remote Work"],"author_name":"Sejuti Das","publish_date":"2020-05-07T18:18:29","publication_year":"2020","word_count":633,"keywords":["how to implement business intelligence","Go","artificial intelligence","machine learning","covid-19","AI","ML","Industry best practices","ViT","disruption","GAN","R","Remote Work","fraud detection"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","fraud detection","R","Go","GAN","ViT","disruption"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/customerxps-shares-some-best-practices-for-businesses-to-overcome-this-crisis\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":66228,"title":"5 Tips For Mentees To Make The Most Of The Mentorship Program","content":"Analytics India Magazine recently launched a community-based mentoring program, AIM Mentoring Circle which encourages data science professionals and enthusiasts, to connect with the best mentors in the industry. It aims at helping budding data scientists or those stuck in their professional careers to seek advice, guidance and suggestions from experienced professionals. The world’s largest one-on-one mentorship network, it provides data science enthusiasts with an excellent opportunity to learn in-demand skills and excel in their professional career. It provides a platform for mentees to get insights from the best of the industry and establish themselves in the highly competitive field. Having a mentor means that mentees have a powerful resource of expertise, strategy and a deeper understanding of the industry. If made the most of this opportunity, it can prove highly rewarding to set a career growth. Here we bring 5 tips for the mentees to make the most of the mentoring program. Express your curiosity to learn more: The mentors at the AIM Mentoring sessions are all about throwing insights about the latest developments in the data science industry. Being experienced in the field and knowing the latest developments, it will be in your advantage to get the best understanding of the industry from them. If you are curious enough and are willing to step out of your comfort zone, the mentor-mentee relationship will offer the best for you. Ask thought-provoking questions, engage with them and learn about the latest tools and technologies from them. Remember that mentors are always attracted to people who are proactive and curious. They are keen to have candidates who are eager to learn and are looking for value addition from them. Share your interests and expectations with the mentors: While the data science field is huge and you may have specific interests in the field, it is best to let the mentors know so that they can help you in the best way possible. Before getting started with the mentoring sessions, set up a meeting just to convey your interests in the topics, establish a big picture and what outcomes you are looking for with the sessions.  Let your mentors know of the goals you would like to accomplish with the program. Whether it is understanding the data science field better, making a transition into the data science field or learning more about the skills in the field.  Setting a few smaller goals and taking the sessions topic-wise will be the best way to make the most of these sessions. Chalk out an agenda with the mentors: Setting an agenda with the mentor helps in deciding how much time they should allot for you and meet your expectations in the best possible way. Coming prepared for the meeting helps mentors decide and work on a routine for the program more efficiently. Having an agenda can help you achieve your goals in a more efficient manner, saving both of them some time. Be open to feedback: It is important to be apprised of the fact that the mentor is here to help you and therefore a candidate must be open to any suggestions and feedback from them. Opening up to the mentor will help them guide you towards your data science journey in an informed manner. Any feedback from mentors regarding career, taking up the courses, increasing your network or any other feedback should be considered in the best interest, given the experience they bring on the table. Give them the full picture of your challenges, needs, goals, and accomplishments so they can understand how to best help you. Keep in touch with your mentors and follow up: For the relationship to yield the best possible results, keeping up with the mentors and staying in regular contact with them through emails will help you make the most of the mentorship program. It shows that you are ambitious enough to utilize this opportunity and consistent enough to set your career in the field. Reaching out to them more often will help your mentee-mentor relationship be alive and help you to seek more benefits from it. Join AIM Mentoring Circle","excerpt":"Analytics India Magazine recently launched a community-based mentoring program, AIM Mentoring Circle which encourages data science professionals and enthusiasts, to connect with the best mentors in the industry. It aims at helping budding data scientists or those stuck in their professional careers to seek advice, guidance and suggestions from experienced professionals. The world’s largest one-on-one […]","categories":["AI Trends"],"tags":["AIM Mentoring Circle","mentorship"],"author_name":"Srishti Deoras","publish_date":"2020-05-30T13:00:03","publication_year":"2020","word_count":683,"keywords":["data science","Go","programming_languages:R","AI","programming_languages:Go","RAG","Aim","AIM Mentoring Circle","analytics","mentorship","R"],"extracted_tech_keywords":["AI","data science","analytics","Aim","RAG","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/5-tips-for-mentees-to-make-the-most-of-the-mentorship-program\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10068497,"title":"How differential binarization can improve real-time scene text detection","content":"Ever wondered how Google Lens scans and reads wifi password at your favourite Starbucks? The answer lies in real time scene text detection. In a recent paper, Real Time Scene-Text Detection with Differentiable Binarization, researchers have proposed an adaptive segmentation technique for various use cases of scene text detection. Before proceeding further, let’s understand Image segmentation and Image Binarization. Image segmentation is a technique of grouping similar segments of an image together based on similarity in pixel values (Semantic Segmentation) or based on similarity in Instances (Instance Segmentation) of the same object in an image. Fig 1: Example of Image Segmentation in autonomous driving systems Fig 2: How Image Segmentation fits in a Machine Learning Architecture Image Binarization is a technique to convert any image to a binary scaled image. Fig 3: Example of Image Binarization In most text detection systems, segmentation is followed by Binarization and pixel grouping techniques to detect texts in real scenarios. The proposed architecture in this paper uses binarization in the segmentation network using a Differentiable Binarization Function so that the binarized image can be trained end to end in a CNN. Fig 4 : Model Architecture Three key concepts in the model architecture are: Differentiable Binarization: Any standard binarization function like one mentioned below (P is the probability map and t is the set threshold) would be non – differentiable and hence it cannot be optimised along with segmentation. Hence, to make binarization optimizable, a function like mentioned below (P and T are the probability map and threshold map learned from the network) would be differentiable:- Adaptive Thresholding: Adaptive Thresholding is optimized thresholding for effective binarization within the segmentation network. Refer Adaptive Threshold under section Methodology in the paper. Deformable Convolution: This step helps in the convoluting over spatial images or images with larger aspect ratios. Refer here for more on Deformable Convolution Network. To get an understanding of how the deformable convolution has been performed in the paper take a look here. Fig 5:- Source code snippet on implementation of the Deformable Convolution layer How does the model work? The input image is firstly fed into a feature-pyramid backbone (for e.g using RESNET). To know more about feature-pyramid backbone refer here. Then, the obtained pyramid features are up-sampled (using image augmentation )to the same scale of the input image and cascaded to produce a feature F, which will be used to predict probability and threshold.Obtained  feature F from the previous step  is used to predict both the probability map (P) and the threshold map (T). Threshold map generates threshold for effective binarization in the following steps. After that, the approximate binary map (Bˆ) is calculated by P and F as enlisted in formula (2) of the paper. Supervision is then applied during model training on probability map, threshold map and approximate binary map with probability and binary map sharing the same supervision. In the inference period, the bounding boxes can be obtained easily from the approximate binary map or the probability map by a box formulation module using the Vatti Clipping Algorithm as enlisted in formula (10) of the paper. Fig 6:- Some visualisation results on text instances of various shapes, including curved text, multi-oriented text, vertical text, and long text lines. For each unit, the top right is the threshold map; the bottom right is the probability map. As per review metrics, DB has worked well with RESNET-50 and RESNET-18 as backbones for real time inference and accurate text detections. However, the method can’t handle cases of “text inside text” or if an overlapping text region is in the centre region of another text instance. Refer the section Implementation Details to know how the model was trained , sections Limitation and Conclusion for more details on the model performance. For accessing the entire research paper with highlighted sections and added notes, click here.","excerpt":"Effective Binarization can help in better image segmentation.","categories":["AI Trends"],"tags":["image segmentation"],"author_name":"Souptik Majumder","publish_date":"2022-06-07T15:00:00","publication_year":"2022","word_count":641,"keywords":["Go","machine learning","image segmentation","AI","programming_languages:R","programming_languages:Go","ai_applications:autonomous driving","CLIP","ResNet","CNN","R"],"extracted_tech_keywords":["AI","machine learning","R","Go","CLIP","CNN","ResNet","programming_languages:R","programming_languages:Go","ai_applications:autonomous driving"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/how-differential-binarization-can-improve-real-time-scene-text-detection\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10003080,"title":"Python Sweeps Away The Top Programming Languages Ranking, Yet Again","content":"Python is the top programming language this year again, according to the recently released list on top programming languages in 2020 by IEEE. The yearly list by IEEE, which is the largest professional organisation devoted to engineering and applied science, ranks the popularity of dozens of programming languages on various platforms. The list of rankings can be selected on the basis of what is trending, languages that are in demand by employers, languages popular on open-source hubs and more. It also allows users to make their own ranking by adjusting the weights using the “Create Custom Ranking” option. The ranking also takes into account the language types: Web: languages used for developing websites and applicationsEnterprise: languages used in the enterprise, desktop and scientific applicationsMobile: languages used for applications on mobile devicesEmbedded: and those used to program device controllers. Python also ranks on the top across languages that are in demand by employers and on open-source hubs. While Java is the most trending language after Python, C stands second to Python in languages that are in demand by employers. C++, Javascript are other languages that follow on the IEEE ranking list. R, which is often considered one of the most preferred languages, has been placed 6th in the ranking, which is one position down from last year. Whereas in terms of the languages that are trending, R is ranked on 15th position. Other popular languages that made it to the list are Arduino, Go, Swift, Matlab, Ruby, Dart, SQL, Scala, Shell and more. As per IEEE, the ranking was created by weighting and combining 11 metrics from eight different sources such as CareerBuilder, GitHub, Google, Hacker News, the IEEE, Reddit, Stack Overflow, and Twitter.","excerpt":"Python is the top programming language this year again, according to the recently released list on top programming languages in 2020 by IEEE. The yearly list by IEEE, which is the largest professional organisation devoted to engineering and applied science, ranks the popularity of dozens of programming languages on various platforms.   The list of rankings […]","categories":["AI News"],"tags":["Programming Languages","Python","Python Programming"],"author_name":"Srishti Deoras","publish_date":"2020-07-23T11:30:00","publication_year":"2020","word_count":283,"keywords":["Go","AI","Scala","Programming Languages","Git","Python","C++","SQL","JavaScript","Python Programming","R","Java"],"extracted_tech_keywords":["AI","Python","R","SQL","JavaScript","Go","Java","Scala","C++","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/python-sweeps-away-the-top-programming-languages-ranking-yet-again\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10060086,"title":"How NFTs will shape the metaverse","content":"The ongoing craze about non-fungible tokens (NFTs) has overtaken the world and has become a bone of contention. However, there is one undeniable fact that NFTs will play a crucial role in birthing an ideal metaverse. NFTs will also provide brand-new opportunities for enterprises, investors, and fans in the long run. Recently, NFTs are being used to enter or “access” the metaverse. But, what value do they have in it? And is it crucial now for people who are into blockchain? Subsequently, it creates NFTs (introduced interoperability and scarcity). However, the projects coupled with curiosity will let NFTs and blockchain play a pivotal role in the metaverse. It is a 3D digital world that allows users and businesses vast opportunities to port real-world assets and services. It can birth an open and fair economy by using blockchain. Since the gaming industry has already been doing what the metaverse plans to, the system of play-to-earn will bridge the gap between NFTs and the metaverse to clear the way for identity, community, and social experiences. Just about every discussion in this industry hints towards mixing NFTs and the metaverse. We can understand that NFTs and the metaverse are synonymous because of the rapid growth in blockchain gaming, hence driving further growth by working towards a virtual world. In addition, real-life identities and digital avatars give direct access to the metaverse with the help of NFTs. For example, back in a 2019 NFT.NYC event, an NFT-based token was used as an entry ticket. Although no one exclusively called it a metaverse event, it hinted at the coexistence of NFTs and metaverse. Amidst all the technological disruption, NFTs will disrupt the metaverse’s traditional social network paradigm of user interaction, socialisation, and transaction. After a couple of successful stints of using NFTs as entry-tickets, it paved way for further adoption of NFTs in the metaverse. Today, projects are focused on introducing massive transformations for online interaction. Another example is Decentraland shows where users own metaverse real estate by using “land” tokens. Since the metaverse is a humongous concept, NFTs can widen its scope. Shortly, it can become a form of currency with which virtual property can be bought and gain exclusive access. In NFTs, a concept called the smart contract, a blockchain programming that helps the network keep NFT transactions that can be accessed later can help create a metaverse economy. At first, NFT-based access can be the primary focus to properly get into the whole metaverse like VIP access and whatnot. From an enterprise’s perspective, they can launch branded merchandise or exclusive access to content to followers, similar to a subscription model. Then, keeping an efficient fan rendezvous aside, NFTs will introduce interoperability outside the metaverse by using location-based features, engagement, and augmented reality. Now, as we slowly pieced together, NFTs and the metaverse are truly a match made in heaven. Keep your friends close, but your IP closer One of the most common questions people have about virtual products is how their value can be protected if anyone copies them. Of course, product imitation is a major problem in the real world, but in the virtual universe, it gets more complex. The materials and construction separate it from the copied ones in our tangible world. However, it is easy to produce similar copies in the virtual universe with no variance. But how do you safeguard it? And this is where NFTs come into play. Blockchain tech might seem daunting at first, but it is easier to understand it as something that helps to record the ownership of an asset, which can be a virtual product or cryptocurrency. The difference in ownership relies on the polarity between the traditional banking system and blockchain, which is decentralised. Since it is decentralised, no one can control, alter, or destroy it. Therefore, if you have bitcoin, it is recorded in the blockchain and cannot be spent twice, and copies cannot be produced. So, when you buy one bitcoin, that unit is connected to a record in the blockchain. And when you spend it or sell it, you give them a unique reference to its blockchain record. Welcome to the new economy Today, users and enterprises can port real-world assets and services in a decentralised virtual environment – the metaverse. Another way of bringing in more real-world assets to the metaverse can be done with the help of gaming models and interoperable blockchain games. The play-to-earn system is the most famous approach which helps in wide user engagement. Players can rely on NFTs and use in-game currency while playing (basically, the more value you add, the more you earn). Furthermore, these play-to-earn games are impartial as players will fully own their assets instead of being controlled by a game compared to traditional gaming. It further helps in increasing the popularity of play-to-earn gaming models and the metaverse in general. In this, guilds work as intermediaries after buying in-game NFT resources like land assets and can be shared with other players who want to earn in their respective virtual worlds. However, play-to-earn guilds take a small portion of the earnings as charges. This directly leads to creating a virtual economy (fair and open) where players who cannot pay beforehand can take the help of the guilds to kick-start their campaign. Guilds lower the entrance barrier for everyone, and they can compete in this metaverse economy. Guilds make NFT resources more accessible for everyone in the metaverse economy. The community Whether it is the real world or the virtual one, a community is destined to exist. For the metaverse, NFTs are crucial as it is used in identification, community and social experiences. For example, keeping certain NFT assets can show a user’s support and dedication for a project and convey various perspectives. This leads like-minded individuals to band into communities to share experiences and create content together. One example can be the NFT avatars. We must keep in mind that NFT avatars show a player’s true or imagined personality, therefore, allowing players to use NFT avatars as access tokens to enter and move around different places in the metaverse. In this case, NFT avatars can be looked at as an extension of our real-life identities, and we have full ownership to curate and build our virtual identities. Furthermore, by owning avatars, we get virtual membership to exclusive content in the metaverse. Meanwhile, NFT avatars are constantly shaping up experiences and environments of the metaverses with the help of content creation, and startup (business) launches.","excerpt":"The projects coupled with the curiosity on it will let NFTs and blockchain play a pivotal role in the metaverse.","categories":["Global Tech"],"tags":["nft blockchain"],"author_name":"Akashdeep Arul","publish_date":"2022-02-09T09:00:00","publication_year":"2022","word_count":1084,"keywords":["Go","nft blockchain","API","programming_languages:R","AI","programming_languages:Go","Git","disruption","R","startup"],"extracted_tech_keywords":["AI","R","Go","Git","API","disruption","startup","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/how-nfts-will-shape-the-metaverse\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10080968,"title":"Package Manager Conda Finds Space in Python’s Burrow","content":"A year after the launch of ‘Conda’, Travis Oliphant, the founder of Anaconda, wrote in a blog about how he was often asked why he was promoting another Python package manager like Conda instead of just contributing to the standard Python packages. But that’s exactly what Oliphant had done. Conda came around after a meeting with Python founder Guido van Rossum in 2012 at the inaugural for PyData. Rossum was answering Oliphant’s questions around packaging issues and stated, “it really sounds like your needs are so unusual compared to the larger Python community that you’re better off building your own”. Travis Oliphant, creator of SciPy, NumPy, and Numba, founder and Director of Anaconda, Inc. Why use Conda at all? Conda was introduced as an alternative to Python’s native package manager pip that comes pre-installed with the most recent versions of Python. But Conda hasn’t gathered sufficient acclaim for itself to justify its inevitability. Author and MLOps expert Noah Gift questioned how popular Conda was among AWS Sagemaker users. What good would it do to install Conda when pip was already installed? The collective response to this is as scattered as it could be. A deep learning researcher and PhD scholar Sebastian Raschka stated that he preferred the Conda environment after having tried all other combinations. “Pip and virtualenv allow you to install packages in virtual environments with different requirements, but it doesn’t help you do the same with different Python version. For that, you’d need pyenv (another tool for managing different versions of Python) on top of that, and that can get pretty messy real quick if you are not careful,” he noted. Conda’s convenient advantages The convenience offered by Conda is one of its biggest strengths hailed by the community. Developers can install their packages and manage the environments all within the same tool while, with pip, users have to use another external tool on top of it to manage the environments. Another developer weighed in saying, “Based on my experience with Conda, it is the easiest way to install packages, especially if they depend on native code. Conda, in my view, is also the easiest way to reuse environments across projects on your local machine. And I don’t know any alternatives that would be as convenient as Conda.” Source: StackOverflow Conda also has what is known as Anaconda Navigator, a GUI or Graphical User Interface-programme to launch widely used applications like Jupyter, RStudio and VSCode. The navigator makes life much easier for developers. Another MLOps engineer echoed this: “I have found the Conda environment very useful if you have a GPU machine and are working with multiple versions of Cuda on those projects”. Conda also comes with a bunch of ML packages like NumPy, Pandas, Scikit-Learn and Matplotlib preinstalled. A developer mentioned this on a Reddit community discussion saying, “It gives you all the standard packages used in scientific computing in a convenient package without having to worry about installing them all individually with their dependencies. If you don’t plan on using typical scientific computing packages or any of the packaged software, the only downside is that you end up downloading software that you might not need,” he noted. The main difference between the two is that while Conda is language-agnostic and a cross-platform environment manager, pip is a general-purpose manager for everything Python. If all your work is with Python packages within isolated environments, Conda and pip+virtualenv are mostly replaceable. Not competitors, Different target audiences Turns out the question between these two isn’t which is better, it’s simply that the user base these two cater to is different—If a user wants to manage Python packages within an existing Python installation, Conda is useless because it will only be able to install packages within a Conda environment. If the user wants to work with Python packages that are normally have external dependencies like NumPy, pip won’t work because it is only restricted to Python packages. This does suggest that Conda and pip aren’t in the same competition as was earlier assumed. But pip does have the same advantages that any tool with a massive pre-existing community has—users who prefer the standard venv or virtualenv over Conda will be more likely to find solutions to their queries or errors they find on StackOverflow simply because pip is used by more people already. The other merit that pip has is that it has already been more battle-tested and is consequently believed to be more robust. Conda might have errors that users may have never found with pip, building the perception that pip is simply more reliable. A number of developers explicitly stated that their new go-to package is Miniforge which is essentially ‘Miniconda’, a much smaller sibling distribution for Anaconda or prefer using Miniconda itself. Miniforge has conda-forge as the default package. A developer commented saying Miniforge is “way better tested for OS and version compatibility compared to the average pip package”. Another developer lauded Miniforge stated, “While the Anaconda team maintains the core scientific stack you get when you install Anaconda, the ‘conda-forge’ channel includes a lot of the other major Python packages not included in Anaconda.”","excerpt":"Turns out the question between these two isn’t which is better, it’s simply that the user base these two cater to is different.","categories":["Deep Tech"],"tags":[],"author_name":"Poulomi Chatterjee","publish_date":"2022-11-29T15:00:00","publication_year":"2022","word_count":858,"keywords":["scikit-learn","NumPy","AI","ML","MLOps","Aim","deep learning","Matplotlib","Jupyter","Pandas"],"extracted_tech_keywords":["AI","ML","deep learning","MLOps","Aim","scikit-learn","Jupyter","Pandas","NumPy","Matplotlib"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/package-manager-conda-finds-space-in-pythons-burrow\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10118332,"title":"PyTorch Releases torchtune for Easily Fine-Tuning LLMs","content":"PyTorch has announced the alpha release of torchtune, a new library designed to simplify the fine-tuning of large language models (LLMs) using PyTorch. Click here to check out the GitHub repository. The library is built on PyTorch’s core principles, offering modular building blocks and customisable training recipes for fine-tuning popular LLMs on various GPUs, including consumer-grade and professional ones. The library provides a comprehensive fine-tuning workflow, from downloading and preparing datasets and model checkpoints to customising training with composable building blocks, logging progress, quantizing models post-tuning, evaluating fine-tuned models, running local inference for testing, and ensuring compatibility with popular production inference systems. Torchtune aims to address the increasing demand for fine-tuning LLMs by offering flexibility and control. Users can easily add customisations and optimizations to adapt models to specific use cases, including memory-efficient recipes that work on machines with single 24GB gaming GPUs. The design of torchtune focuses on easy extensibility, democratising fine-tuning for users of varying expertise, and interoperability with the open-source LLM ecosystem. The library integrates with popular tools such as Hugging Face Hub, PyTorch FSDP, Weights & Biases, EleutherAI’s LM Evaluation Harness, ExecuTorch, and torchao for various purposes like model and dataset access, distributed training, logging, evaluation, inference, and quantisation. torchtune currently supports Llama 2, Mistral, and Gemma 7B models and plans to expand with additional models, features, and fine-tuning techniques in the coming weeks including 70 billion parameters and Mixture of Experts models.","excerpt":"Users can easily add customisations and optimizations to adapt models to specific use cases, including memory-efficient recipes that work on machines with single 24GB gaming GPUs.","categories":["AI News"],"tags":["LLMs","Pytorch"],"author_name":"Mohit Pandey","publish_date":"2024-04-17T03:16:41","publication_year":"2024","word_count":237,"keywords":["Pytorch","Hugging Face","AI","LLMs","PyTorch","Git","Aim","ai_frameworks:PyTorch","llm_models:Llama","GitHub","R","Weights & Biases"],"extracted_tech_keywords":["AI","Weights & Biases","Aim","PyTorch","Hugging Face","R","Git","GitHub","llm_models:Llama","ai_frameworks:PyTorch"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/pytorch-releases-torchtune-for-easily-fine-tuning-llms\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10083060,"title":"Breathe Capital Thinks Climate Tech Startups are Hot","content":"In 2022, over a quarter of all the venture capital funding went into climate technology startups specifically focussed on technologies targeting emission reduction. In 2021, a total of $53.7 billion VC funds were allocated to climate tech startups. This has been the trend since the signing of the Paris Agreement in 2016, which resulted in more and more companies shifting focus to solving climate change. To understand the VCs philosophy about investing in climate tech companies, Analytics India Magazine spoke to Shauraya Bhutani, co-founder and partner at Breathe Capital, a firm looking to investing in technology startups while focusing on climate tech startups. Bhutani shared his views on the growing interest of VCs in climate technology while also talking about what his firm looks for in a startup. Bhutani has been operating in the investment sector for more than eight years out of Singapore and started working from India only two years ago, focussing on the South-East Asian and Indian markets. He is also the founder of Capital Connect Advisors, offering capital funding and financial advisory services for technology entrepreneurs and investors. Bhutani got together with Zubin Mehta, to start Breathe Capital in 2022. “Despite the funding winter, there is a lot of money in good startups,” said Bhutani. “If you invest in 10 companies, six of them are going to go bust, two of them are going to probably return your money, and the other two are going to return all your funds. So, you are always looking for the company that will return your investment,” said Bhutani. Breathe Capital is a sector-agnostic early stage investment firm mostly focusing on investing in technology startups that are building for the climate, while also investing in e-commerce and fintech. “We are not essentially looking for tech innovations, but a business model innovation that can be adapted to a developing country, and taken to international markets,” adds Bhutani. The rise of climate tech startups “India has fundamentally different challenges when compared to the US or Europe, and we are seeing climate tech pick pace in India and South-East Asia in the past 2-3 years,” said Bhutani. “Though Europe has been leading innovations in the climate tech industry, India is now driving more and more investment in the field.” There has essentially been a top-down directive from the Indian government to invest in climate tech startups. “Most of the capital comes from the sovereign wealth funds and DFIs and they are directing VCs to start making climate tech investments a priority, even the LPs are pushing for it,” said Bhutani. Bhutani believes that the current stage of the climate tech startup industry is mostly like the e-commerce industry in 2013-14. “We are seeing very similar investors and investments in this sector as well with more and more fintech investors getting interested,” said Bhutani. He adds that investors like Y-Combinator, Lightspeed Ventures, and Sequoia Capital also announced that they are increasingly investing in the climate tech industry. “We will definitely see a bunch of unicorns from the sector, starting from India and going abroad as well, with the government stepping in to support them.” Energy and farming “When we map out the climate tech sector, the transport, mobility, and energy sector are already almost done with but there are still pockets of opportunity in the energy sub-sector like storage for instance,” explains Bhutani. There are very few electric vehicle companies that receive pre-seed or seed funding as the sector is very advanced. Bhutani said that most of the founders are building in the circular economy and a lot of new innovations are also coming in the EV charging space. “Under the circular economy, there are two kinds of approaches we are witnessing, one is product design which ensures that the product stays longer in the market. And the other is the business model design approach, which means getting a finite resource back into the ecosystem,” explains Bhutani. The firm has made three investments in the business model design sector in the past two months. The farming and food production sector is gaining a lot of traction in recent years with brands that are offering plant based meat and protein alternatives. “We are not a multi-stage fund so we stay away from the protein sector because it requires a long-term view and heavy investments,” said Bhutani. The firm is more interested in companies building software solutions for monitoring and precision farming. ‘Money is just a commodity for some founders‘ Bhutani said that earlier, he used to find only one slide out of ten in a slide deck of a company that talked about the impact it has, which looked more like a “nice to have” instead of a “must have”. “Now when we look at climate tech companies, we are seeing more and more of them talking about the impact they have on the environment and embed it in their pitches. These are the type of companies we are interested in investing in, companies that are directly, or indirectly cutting down emissions and making it their priority — that is the key metric,” affirms Bhutani. “Where Breathe Capital stands apart is that being new into the industry, we are focused on adding value to the company. By being only an early stage investor, we prepare and structure the startups for later rounds,” said Bhutani. “All we expect from founders is transparency.” Click here to pitch ideas to Breathe Capital.","excerpt":"Breathe Capital is a sector agnostic early-stage investment firm focusing on investing in technology startups that are building for the climate","categories":["Deep Tech"],"tags":["AI Startups"],"author_name":"Mohit Pandey","publish_date":"2022-12-22T15:00:00","publication_year":"2022","word_count":904,"keywords":["Go","API","unicorn","AI","innovation","RAG","Ray","analytics","R","AI Startups","startup"],"extracted_tech_keywords":["AI","analytics","Ray","RAG","R","Go","API","innovation","startup","unicorn"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/breathe-capital-thinks-climate-tech-startups-are-hot\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":50420,"title":"Silicon Is The New Gold: Ex-Apple Chip Designer’s Startup Nuvia Secures $53 Million Funding","content":"Gerard Williams III (ex-Apple), Manu Gulati (ex-Google), and John Bruno (ex-Google) this month announced that their silicon startup Nuvia had raised $53 million funding from a group of prominent investors. The investment round was led by Dell Technologies Capital, Capricorn Investment Group and other Silicon Valley firms. The California-based Nuvia was started earlier this year to design chips for data centres. A young company securing such hefty funding can be a shocker to many, but if a company has ex-Apple and ex-Google top players as founders who are designing chips for data centres, investing in such an opportunity isn’t that big a bet. Key People Behind Nuvia Gerard Williams III is the CEO of NUVIA. Prior to co-founding NUVIA, he was a Senior Director at Apple and Chief CPU Architect for nearly a decade. Before joining Apple, Gerard spent over 10 years at ARM, as an ARM Fellow, and serving on the ARM Architectural Review and Technical Advisory Boards. Manu Gulati is the SVP of Silicon Engineering at NUVIA. Manu started his career at AMD and 8×8 and has over 58 issued patents to date. Manu holds a BSEE from IIT Delhi and an MCSE from UC Irvine. Prior to co-founding NUVIA, Manu was the lead SoC Architect for consumer hardware at Google, playing a key role in defining the company’s silicon and product roadmaps. John Bruno worked as a System Architect at Google. Before joining Google, John spent five years at Apple in a similar role in the company’s platform architecture group. John Bruno is now the SVP of System Engineering at NUVIA and he, along with his co-founders plan to take on the likes of Intel and AMD in the silicon sector. “The world is creating more data than it can process as we become increasingly dependent on high-speed information access, always-on rich media experiences and ubiquitous connectivity,” said Gerard Williams in a statement. With Nuvia, the trio aims to reimagine the way silicon chips are created and produce products that would make data centres more efficient. Ushering A New Wave Of Chip-Making Startups Nuvia joins the club of another exciting list of startups that have been making noise with their chips for AI. Previously Cerebra, Graphcore and couple of other startups garnered attention for raising hefty funds. One of them even achieved the unicorn status with $200 million in funding. UK-based Graphcore is a semiconductor company which secured a total funding amount of $310 million. It develops accelerators for AI and machine learning. The IPU’s unique architecture means developers can run current machine learning models orders of magnitude faster. Whereas, California-based Fungible, founded in 2015 by the founders of Juniper, have raised a whopping $292.5 million till date. The rush to build the next-gen AI-based hardware is quite obvious now and it is being verified by the fact that startups are being supported with hefty funding. So far, companies like Intel, NVIDIA and AMD have been steering the chips for deep learning markets. Now a new wave of startups are looking to take over the charge from these giants. As part of that team, we’re focused on providing NUVIA with DTC’s unique value and market leverage. -Scott Darling, President of Dell Technologies Capital With a sudden uprising of startups like that of Nuvia, spearheaded by the veterans from big tech companies like Google and Apple, we can safely say that a new wave of chip maker has arrived and the ecosystem is ripe for transformation led by AI.","excerpt":"Gerard Williams III (ex-Apple), Manu Gulati (ex-Google), and John Bruno (ex-Google) this month announced that their silicon startup Nuvia had raised $53 million funding from a group of prominent investors. The investment round was led by Dell Technologies Capital, Capricorn Investment Group and other Silicon Valley firms. The California-based Nuvia was started earlier this year […]","categories":["AI News"],"tags":["Apple"],"author_name":"Ram Sagar","publish_date":"2019-11-21T14:00:42","publication_year":"2019","word_count":581,"keywords":["Go","API","machine learning","AI","Apple","RAG","Aim","deep learning","ViT","R","startup"],"extracted_tech_keywords":["AI","machine learning","deep learning","Aim","RAG","R","Go","API","ViT","startup"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/nuvia-apple-chip-startup-funding\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":41149,"title":"Zeotap’s Guru Patnaik On How Organisations Can Close The Cyber Security Gap","content":"With cyber threats gaining ground, business leaders across the board are increasing their investments in cyber resilience. Interestingly, some of the ground-breaking technologies that are powering consumer-facing applications can also help in securing organisations. This week, we caught up with Zeotap’s Guru Patnaik, Director of Information Security who also earned the CCISO (Certified Chief Information Security Officer) certification. In this interview, Patnaik gives a lowdown on the cybersecurity domain and emerging roles in this field. He also stresses how AI and automation can help secure data and why the C-suite should build a cyber-resilience strategy. AIM: In this digital age, how has the role of security management evolved or become prominent? Guru Patnaik: In my experience, of over a decade, the approach of “Security Management” has drastically evolved from: run antivirus software, don’t write your password on sticky notes, don’t click on suspicious links, to: personalised security controls, stringent threat analysis\/management and regulatory enforcement. Security has definitely moved beyond IT. It’s become an enterprise-wide issue that needed addressing, and managing risk has become a “business priority”. Security teams have to come up with new tactics to fight against advanced threats. AIM: What are the various positions that both large organisations and start-ups are recruiting for the role in management? GP: Information\/Cyber Security is one of the most sought after and incredibly attractive field due to the ever-changing threat scenario and evolving risk landscape. Some of the roles that are in demand, are – Security Analyst, Security Architect, Security Administrator, Security Consultants, Chief Security Officer etc. The objective should be to align the organizational risk portfolio to manage the security posture with skilled resources. AIM: Tell us about your role as a CCSIO or a Certified Chief Information Security Officer? GP: The role focuses on the application of information security management principles from an executive management point of view, justifying “good to have” and “need to have” as the principle approach. One of the key and interesting aspects of the role is anticipating new threats and actively working towards preventing them from occurring. AIM: How can a certification in CCISO be done? What are the various aspects that the course covers? GP: The Certified CISO (CCISO) programme is one of the most recognised security certification programs aimed at producing top-level information security executives. It also equips information security leaders with the most effective toolset to defend organizations from cyber threats such as cyber-attacks. EC Council (CCISO Certification) has done a fabulous job in structuring a course that helps in connecting the dots, especially for a CISO to be able to function effectively. The course covers techno-management domains like – Governance, Security Risk Management, Controls and Audit Management, Security Program Management & Operations, Information Security Core Concepts and Strategic Planning, Finance & Vendor Management. AIM: What are the most effective toolset that equips information security leaders? What are the key skills they should pose? GP: Predominantly, the role of a CISO is to address the emerging threats to information by developing and maintaining a tough information security strategy and action plan coalescing People, Process & Technology as the key components. CISOs must have relevant experience (technical and process), strong leadership and communication skills and innovative strengths clubbed with broad business knowledge to resolve the ever-growing information security threat landscape. AIM: Who should avail this certification? GP: The CCISO program is for executives looking to hone their skills & learn to better align their information security programs to the goals of the organisation as well as aspiring CISOs. CCISO focuses on exposing middle managers to executive-level content as well as encouraging existing CISOs to continually improve their own processes & programs. AIM: What are some of the other roles you see evolving in the tech space in the coming future? GP: There is no organization left that has never encountered a hacking attack or a breach. Whether small or big, every organisation has encountered an incident at least once. With some of the recent attack techniques like – ransomware, re-authentication, cyber extorting to name a few,  there is going to be a paradigm shift in skill set required to address such issues. Some of the roles that are going to evolve are – Dev-Sec-Ops, Defensive and Offensive security, Artificial Intelligence in security analysis and threat hunting","excerpt":"With cyber threats gaining ground, business leaders across the board are increasing their investments in cyber resilience. Interestingly, some of the ground-breaking technologies that are powering consumer-facing applications can also help in securing organisations. This week, we caught up with Zeotap’s Guru Patnaik, Director of Information Security who also earned the CCISO (Certified Chief Information […]","categories":["AI Features"],"tags":["Cyber Security","cyber security India","Interviews and Discussions"],"author_name":"Srishti Deoras","publish_date":"2019-06-22T14:04:09","publication_year":"2019","word_count":713,"keywords":["Go","cyber security India","artificial intelligence","Cyber Security","programming_languages:R","AI","Git","RAG","automation","Aim","GAN","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","artificial intelligence","Aim","RAG","R","Go","Git","GAN","automation","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/zeotaps-guru-patnaik-on-how-organisations-can-close-the-cyber-security-gap\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10074168,"title":"NFTs won&#8217;t be the Same without Social Media","content":"Meta will now let users post NFTs across Instagram and Facebook. Users can connect their wallets to post digital collectibles minted on Ethereum, Polygon and Flow. An updated blog post of the company said that it has “started giving people the ability to post digital collectibles that they own across both Facebook and Instagram” after they connect their digital wallets to either platform. The company has been experimenting on both platforms with NFTs since May 2022. This is likely the first wave of digital collectives on social media. However, the ability to cross-post NFTs on both Instagram and Facebook is currently available only to US-based users. Meta & NFTs — A Timeline July 2022: Meta launched NFT support on Instagram in 100 countries. This news came out recently after the company started its test launch in May. With this new update, Instagram users can post their NFTs on their Instagram feed and share them with others. Further, users can automatically tag the creator and collector of the NFT for proper attribution. May 2022: Adam Mosseri announced that Instagram would start testing NFTs with select creators in May. He mentioned that only a limited number of people had access to the test, but the company plans to roll out more functionality once it gets feedback from initial testing. June 2022: Facebook began showcasing digital collectibles with limited US-based creators. We're launching NFTs on Facebook! Excited to share what I've been working on with the world. pic.twitter.com\/TaV66zRanV— Navdeep Singh (@navdeep_ua) June 29, 2022 March 2022: Mark Zuckerberg said NFTs are coming to Instagram soon. NFT benefits for the users Monetisation and commodification of digital content NFT gives users the ability to create and monetise their content. Commodification is possible through tokenisation, by converting the physical or virtual content into digital units that can be bought and sold. Social media platforms must be on the same page with the popularity of NFTs and work as a common ground. These platforms help bridge the gap between buyers and sellers and determine its monetary value. Earning from royalty Royalties can be programmed into an NFT for the creator to earn profit each time a piece of art is sold to the new owner. Autonomous in nature Decentralised social networking platforms give users more control and autonomy. Benefits include censorship resistance, ownership over personal data, and improved control over user-generated content. This means no one else can make modifications to or remove content created by users. Ownership control Social media users can upload their content on the platform and ‘mint’ their work by giving it an NFT value. This solution is for streaming and digital reproductive commodification of content providers who lose money. Twitter, Reddit join the club https:\/\/twitter.com\/TwitterBlue\/status\/1484226494708662273?s=20&t=Dfz-hiTxC3seYQj1ZNiJfw Earlier this month, Twitter launched its NFT feature for subscribers of its paid platform service, Twitter Blue, on iOS. The microblogging site allowed users to use their NFT profile picture in a hexagonal shape, link their cryptocurrency wallet, and display their NFT collection. In a statement, Twitter said the company sees people using NFT as a form of identity and self-expression. The new feature lets people verify their NFT ownership by allowing them to directly connect their crypto wallets to Twitter and select an NFT from their collection as their new profile picture. Meanwhile, global sales of NFTs in 2021 are estimated at a whopping $25 billion. According to GlobalData’s Social Media Analytics tool, the “popularity of NFTs is due to the herd behaviour propagated by social media”. It noted that several popular NFTs emerged from widely circulated memes, such as ‘Disaster Girl’ and ‘Success Kid’. Though Twitter took the lead by allowing the use of NFTs as profile pictures, Reddit has followed the same path. Reddit is currently testing an NFT profile picture implementation as part of its “CryptoSnoos” project. Once implemented, this feature will allow users to display the NFT of their choice as they browse.","excerpt":"Instagram and Facebook users can now post their NFTs on the platforms. Twitter and Reddit will also allow the use of NFTs as profile pictures","categories":["IT Services"],"tags":[],"author_name":"Tasmia Ansari","publish_date":"2022-09-02T13:00:00","publication_year":"2022","word_count":651,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","Git","analytics","GAN","R"],"extracted_tech_keywords":["AI","analytics","R","Go","Git","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/nfts-wont-be-the-same-without-social-media\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10000699,"title":"India’s Indigenous Lunar Mission Chandrayaan-2 Is Set To Land Near The South Pole In 2019","content":"One of the most anticipated projects of the Indian space industry is undisputedly the Chandrayaan-2. A decade after India’s first attempt to study moon with Chandrayaan-1, ISRO will launch its second version on January 31 2019. In this article, we list down the reasons why the upcoming Chandrayaan 2 will be a remarkable milestone for India. How Is This Lunar Mission Different From Chandrayaan-1? Although Chandrayaan-2 is a follow-up of Chandrayaan-1 mission, both had different objectives. Chandrayaan-1 had only orbited the moon, did not land on it. Hence, it did not have a rover in its mission. Chandrayaan-2 is a combination of a rover, lander and an orbiter. The orbiter will orbit around the moon and perform the objective of remote sensing the moon.The mission will carry a six-wheeled rover that will move around the landing site in semi-autonomous mode as decided by the ground commands. The instrument on the rover will collect the data from the lunar surface and send back data, which will be useful for analysis of the lunar soil.  Also, unlike Chandrayaan-1’s moon Impact Probe, the Vikram lander of Chandrayaan-2 will make a soft landing, deploy the rover, and perform some scientific activities such as checking for the presence of water ice on the moon. Why Is It So Special? There are broadly two key features that play a major role in making this mission stand out among other moon missions: First rover: This is the very first time that India is going to land a rover on the Moon, or any other celestial surface for that matter. This is also India’s first attempt in developing a lander. Besides China’s Chang’e-4 soft landing to the Moon, no other country achieved a successful landing at the place where Chandrayaan-2 is planned to land at the South Pole. The site: The landing site of Chandrayaan-2 is at the south pole of the moon. The site is above 70 degree latitude and no satellite in the world has ever reached this place. What is so significant about this site is that the area remains in shadow is much larger at this pole, and also because of the presence of water ice. On the other hand, the pole also has places with permanent sunlight. Sending a rover here would give us insights on both, the darkness regions as well as the sunlight regions. Although there have been attempts of studying this area from space, no rover has ever landed at this unexplored site. The pole is also a home to the Cabeus crater, where LCROSS impacted in 2009, as well as the Aitken Basin, which contains impact melt that will allow scientists to unambiguously determine the basin’s age, plus Shackleton crater, the region touted as the perfect place for future outposts and huge telescopes. The south pole contains mountains such as Epsilon Peak which is taller than any mountain found on earth. Vikram: The lander of Chandrayaan-2 called Vikram, will detach from the orbiter and descend into a lunar orbit of 30 km x 100 km (19 mi x 62 mi) using its 800 N (180 lbf) liquid main engines. It will then perform a comprehensive check of its on-board systems before attempting to land on the lunar surface. ISRO is making an engine which will make it easier for scientists to control land a spacecraft on the moon surface. ISRO had created a man-made crater in order to simulate the moon surface. The payloads will collect scientific information on lunar topography, mineralogy, elemental abundance, lunar exosphere and signatures of hydroxyl and water-ice. Chandrayaan-2 landing site. Key Challenges For This Lunar Mission More than 50 missions to the moon have failed, including Apollo 13. The mission was initially planned in 2013, but it faced many challenges because of which it has been now finally scheduled to launch in 2019. Soft landing: Soft landing challenges can cause ISRO a delay in executing this mission. The site at which it is determined to land has many craters, hence it is difficult to land. Earlier this year, the lander had trouble with throttling. “The lander was developing vibrations at the time of re-throttling. The problem appears to be with the thrusters,” a senior level official at ISRO had said. Russia Lander plan went kaput: There was a deal with the Russian Space agency to launch Chandrayaan-2, where Russia was supposed to make the lander and India was supposed to make the orbiter and the rover. But since Russia had to postpone because of some issues, ISRO decided to build the lander, along with the rover and the orbiter. https:\/\/twitter.com\/isro\/status\/1062715001994010624 Fifth liquid engine: One of the major challenge was the integration of the fifth liquid engine to manage the additional load of the lander which now has to orbit the Moon, lander legs, rover integration, modified harness, and so on. The fifth engine had failed a heat test. It also had extended solar panels at about 350 meters, which was casting a shadow on the rover and would have caused a problem getting solar charged as soon as it lands. Outlook Chandrayaan-2 will open up new opportunities for new tech and scientific achievements for India. There will be many experiments and test plans for the moon mission. Considering the exotic site at which the the rover is going to land, it is sure that the mission is going to give the world some very deep discoveries.","excerpt":"One of the most anticipated projects of the Indian space industry is undisputedly the Chandrayaan-2. A decade after India’s first attempt to study moon with Chandrayaan-1, ISRO will launch its second version on January 31 2019. In this article, we list down the reasons why the upcoming Chandrayaan 2 will be a remarkable milestone for […]","categories":["AI Features"],"tags":["Interviews and Discussions"],"author_name":"Disha Misal","publish_date":"2018-12-27T19:10:19","publication_year":"2018","word_count":902,"keywords":["Go","ELT","programming_languages:R","AI","programming_languages:Go","Ray","ViT","Rust","R","programming_languages:Rust","Interviews and Discussions"],"extracted_tech_keywords":["AI","Ray","R","Go","Rust","ELT","ViT","programming_languages:R","programming_languages:Go","programming_languages:Rust"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/indias-indigenous-lunar-mission-chandrayaan-2-is-set-to-land-near-the-south-pole-in-2019\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10103139,"title":"Bing Chat No More","content":"At Ignite 2023, Microsoft announced that it will be renaming its AI search engine-based chatbot Bing Chat to Copilot. This strategic overhaul aims to revolutionise user experiences, embracing a more conversational and intelligent interface reminiscent of ChatGPT. The rebranding of Bing Chat to Copilot signifies a pivotal moment in Microsoft’s commitment to providing a seamless and engaging search experience globally. Simply put, Microsoft looks to provide unified Copilot experience to its consumers as well as enterprise customers. On Wednesday, at Microsoft Ignite, CEO Satya Nadella emphasised the importance of this transformation, and said: “This is clearly the age of the Copilot.” The renaming of Bing Chat to Copilot is not merely a cosmetic change; it represents a commitment to creating a more dynamic and responsive search platform, catering to the evolving expectations of users worldwide. Moreover, it marks a shift towards a more conversational and intelligent user interface, aligning with the global initiative to enhance user engagement and search functionality. The no-cost version of Copilot will remain available in Bing and Windows, and it will also have a dedicated domain at copilot.microsoft.com, similar to the structure of ChatGPT. Business users will log in to Copilot using an Entra ID, whereas consumers will require a Microsoft Account to utilise the free Copilot service. The official support for Microsoft Copilot is currently limited to Microsoft Edge and Chrome, and it is accessible on Windows and macOS. Microsoft vs the World Earlier this year, Nadella expressed his desire to make Google dance, which he referred to as an 800-pound gorilla, to embrace AI more actively in its Search functionality. At the same time, Microsoft has been accusing Google of following unfair tactics that led to its dominance as a search engine, and the battle continues to this day. Not to forget the rising popularity of OpenAI’s ChatGPT. This rebranding could have been done to differentiate itself from ChatGPT and Google Bard, and positioning itself as a productivity tool for both personal as well as professional use, rather than just experimentation or conversing platform with the web. Interestingly, the rebranding of Bing Chat come close on the heels of OpenAI’s announcement that ChatGPT is being used by 100 million people on a weekly basis. One of the key highlights of Copilot is its focus on creating a personalised and engaging interaction with the search engine. Drawing inspiration from ChatGPT, Copilot goes beyond traditional search functionality, offering users a platform where they can converse with the search engine in a more human-like manner. The enhanced language understanding capabilities of Copilot enable it to grasp user queries with nuance, providing more accurate and relevant results. This move towards a more conversational platform signifies a departure from Google’s conventional search engine model, as Microsoft endeavors to make the interaction with Copilot feel less like a query and more like a conversation. Why Copilot? By infusing Copilot with ChatGPT-inspired features, Microsoft is not only keeping pace with evolving user expectations but also setting new standards for what a search engine can offer. The goal is to create a platform that not only provides information but does so in a way that feels natural, interactive, and tailored to individual preferences. As Copilot rolls out to users worldwide, the impact of this transformation is expected to redefine the way people interact with search engines, making information retrieval a more intuitive, personalised, and engaging process. The age of the Copilot has dawned, promising a new era in the world of search engine technology.","excerpt":"“The renaming of Bing Chat to Copilot is not merely a cosmetic change.”","categories":["AI Features"],"tags":[],"author_name":"Sandhra Jayan","publish_date":"2023-11-16T15:54:52","publication_year":"2023","word_count":582,"keywords":["Go","ChatGPT","OpenAI","AI","Google Bard","ML","GPT","Aim","ViT","R"],"extracted_tech_keywords":["AI","ML","ChatGPT","OpenAI","Google Bard","Aim","R","Go","GPT","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/bing-chat-no-more\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10017588,"title":"The Pitfalls of Using AI As Exam Invigilator — And How To Avoid Them","content":"In the wake of the Covid 19 outbreak, the number of universities and colleges using AI-based video surveillance for proctoring exams is on the rise. Proctoring technologies tap AI to verify students through 1:1 identification. It uses video and audio recordings of students taking online exams to monitor body movement or noises to flag malpractices. Delhi Technological University, University of Petroleum and Energy Studies, Dehradun, SRM University, Chennai, O.P. Jindal Global University, Sonipat, Mumbai University and IIT-Bombay leverage AI for exam invigilation in various capacities. What Could Go Wrong? The current technology captures a student’s picture every 15 to 20 seconds, resulting in 200-240 photos of the students in an hour, while also collecting biometric data and personally identifiable information (PII) such as full name, date of birth, address, phone number, government IDs etc. On top of the exam pressure, being continuously exposed to the camera with private data being collected and monitored every second can spike students’ anxiety levels. The surge in the number of cybersecurity mishaps and the lack of adequate policy and implementation further compound the issue. The collected data is prone to misuse. To make matters worse, India does not have a proper data protection law. The Privacy Data Protection Bill is still under review of the Joint Parliamentary Committee. Experts have cast aspersions on the draft bill for warranting excessive power to the government to use anonymised data. There is also a palpable uncertainty over how the government will use the data from public universities gathered through proctoring technologies. For starters, the technology used by universities is not foolproof. For instance, there were complaints about technical glitches in the CodeTantra app used by University of Petroleum and Energy Studies, Dehradun. Further, there are cheat codes available on the internet to end-run around proctoring technologies. The universities should also account for many external factors such as internet connectivity, potential disturbances, etc. There is a high likelihood of students getting flagged for situations not under their control. What Can Be Done About It? Last week, the Internet Freedom Foundation wrote to the University Grants Commission to stop using proctor technologies to conduct online examinations. The letter rightly points out that any justifiable intrusion by the State into people’s right to privacy protected under Article 21 of the Constitution must conform to certain thresholds such as legality, necessity, proportionality and procedural safeguards. The universities have failed to uphold the standards. And for those reasons, universities should refrain from using AI until the privacy and technical requirements are met. Universities should conform to a basic protocol while using AI for proctoring. Firstly, the data should not be used for any purpose other than advancing the proctor application’s performance. Secondly, universities should ensure the availability of a functional computer or laptop, high-speed internet connection, and other accessories, including a webcam and speakers for all students. Thirdly, the technology should be tried and tested in all scenarios and have a high threshold for accuracy. There is always a chance of technology going wrong. To address this, the decision-making on malpractices should be made by human proctors. Wrapping Up Technology, especially automation through AI, has been a lifesaver amidst the pandemic and will continue to be so. However, it is essential to deploy technology responsibly while accounting for privacy and security issues.","excerpt":"In the wake of the Covid 19 outbreak, the number of universities and colleges using AI-based video surveillance for proctoring exams is on the rise. Proctoring technologies tap AI to verify students through 1:1 identification. It uses video and audio recordings of students taking online exams to monitor body movement or noises to flag […]","categories":["AI Features"],"tags":["purpose of ai"],"author_name":"Kashyap Raibagi","publish_date":"2021-01-12T17:14:23","publication_year":"2021","word_count":550,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","RAG","automation","ViT","purpose of ai","R"],"extracted_tech_keywords":["AI","RAG","R","Go","ViT","automation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/the-pitfalls-of-using-ai-as-exam-invigilator-and-how-to-avoid-them\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10141161,"title":"Goodbye Vanilla RAG, Agentic RAG is Here","content":"Everyone loves retrieval-augmented generation (RAG). It has revolutionised how AI systems process and respond to user queries by leveraging external knowledge sources. At the same time, everyone wants to replace RAG with something new as it doesn’t meet all the diverse needs of modern enterprises. As the demands for nuanced, complex, and adaptive AI systems grow, the traditional RAG approach—often dubbed vanilla RAG—is reaching its limitations. This is where agentic RAG comes into play. Agentic RAG represents an advanced architecture that combines the foundational principles of RAG with the autonomy and flexibility of AI agents, promising a future where AI systems are more adaptive, proactive, and intelligent. What Exactly is Agentic RAG? Armand Ruiz, VP of product-AI platform at IBM, shared on LinkedIn that agentic RAG is here, and it aligns with the future of AI, which he believes is also agentic. He posted the GitHub repository for LangChain agentic RAG system using IBM’s Granite 3.0 8B Instruct model on Watsonx. Regardless, in its conventional form, vanilla RAG involves a linear pipeline where user queries are processed through retrieval, reranking, synthesis, and response generation. While it effectively generates grounded and contextually relevant answers, vanilla RAG struggles with flexibility. It relies heavily on predefined knowledge sources, lacks mechanisms for validating retrieved data, and operates as a one-shot retriever without iterative refinement. Agentic RAG addresses these shortcomings by integrating AI agents into the RAG pipeline. These agents act autonomously, orchestrating complex tasks like planning, multi-step reasoning, and tool utilisation. This agentic approach transforms static retrieval systems into dynamic frameworks capable of adapting strategies based on evolving data and user needs. At the core of agentic RAG is the ability to incorporate agents at various stages of the RAG pipeline. It allows users to build systems with complete autonomy to reason and execute specific tools when needed. Technology partner manager Erika Cardenas, and machine learning engineer Leonie Monigatti at Weaviate explained that Agents determine whether external knowledge is needed, select the appropriate retrieval tool (e.g., vector search, web search, APIs), and formulate queries tailored to the task. Further, instead of relying on the initial retrieved data, agents validate its relevance and re-retrieve if necessary, ensuring the final output aligns with the user’s intent. Agents can also access diverse tools, from calculators and email APIs to web searches and proprietary databases, significantly broadening the scope of what can be retrieved and processed. With agentic RAG, it seems like the conversation around fine-tuning and RAG is finally dead. Agents can resolve queries with unparalleled accuracy and speed by retrieving information from community forums, internal knowledge bases, and documentation. It is akin to a model being fine-tuned at the time of inference, also what some people call reasoning using multiple models. This is what LlamaIndex also calls the agentic RAG framework—adding LLM layers to reason over inputs and post-process the outputs. This architecture isn’t limited to single-agent systems. In more advanced setups, multiple agents collaborate under the guidance of a meta-agent, with each specialised in tasks like summarising internal documents, retrieving public data, or analysing personal content like emails and chat logs. RAG is Not the Answer? Not everyone likes RAG. Amit Sheth, the chair and founding director of the Artificial Intelligence Institute of South Carolina (AIISC), replied to Ruiz’s post, claiming RAG bothers him in principle. “You need RAG because the core\/backend\/main AI system is inadequate,” Sheth said, adding that RAG systems are needed because the core AI systems are not good enough for accurate information, which makes it a loss of effort that went into building them. Moreover, according to several researchers, if an agentic RAG thinks longer and gives no response because no information is available in the database, it is just a waste of compute, and thus, it does not scale with more compute. In a bid to move beyond RAG, Google introduced a new approach—retrieval interleaved generation (RIG)—with its DataGemma model. This technique integrates LLMs with Data Commons, an open-source database of public data. With RIG, if the AI model needs more current or specific data, it pauses to search for this information from reliable external sources like databases or websites. The model then seamlessly incorporates this newly acquired data into its response, alternating between generating content and retrieving information as needed. When it comes to agentic RAG, however, the system dynamically adapts retrieval strategies, accessing varied tools and knowledge sources beyond static databases. With iterative retrieval and reasoning, agents ensure the data they retrieve is accurate and relevant. Agents anticipate user needs and take preemptive actions, enabling a smoother and more efficient interaction process. This proactive and adaptive nature makes agentic RAG particularly effective in scenarios requiring detailed reasoning, multi-document comparison, and comprehensive decision-making. In this context, even approaches like RIG become less relevant.","excerpt":"With agentic RAG, it seems like the conversation around fine-tuning and RAG is finally dead, as agents are now helping in reasoning.","categories":["AI Trends"],"tags":["Agentic RAG","RAG"],"author_name":"Mohit Pandey","publish_date":"2024-11-19T17:00:00","publication_year":"2024","word_count":794,"keywords":["Agentic RAG","Weaviate","machine learning","artificial intelligence","TPU","AI","ML","LlamaIndex","RAG","LangChain","Aim"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","LangChain","LlamaIndex","Aim","RAG","Weaviate","TPU"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/goodbye-vanilla-rag-agentic-rag-is-here\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10162301,"title":"KASFAB Tools Launches India’s First Semiconductor Manufacturing Facility","content":"KASFAB Tools Private Limited, part of the KAS Group, inaugurated India’s first facility dedicated to manufacturing semiconductor equipment for global customers on Monday. The facility, located in Doddaballapur, was launched in the presence of representatives from Fortune 100 companies like Suraj Rengarajan, head of the semiconductor products group in Applied Materials India, Ankineedu Velaga, SVP & country head YES India and Srinivasan Raghavan, chairman CeNSe IISc. The state-of-the-art facility spans 20,000 square feet and includes advanced features such as Class 10 and Class 100 cleanrooms, high-precision welding capabilities for plastics and metals, comprehensive testing infrastructure, component fabrication, and ultrasound-aided cleaning using deionised water. “This initiative will generate employment, attract significant investments, and accelerate India’s ambitions to become a global leader in semiconductor manufacturing,” said Manjunath Jyothinagara, MD at KASFAB Tools. This facility features independent test and validation tools, a safety simulation bench, and functional testing systems. These capabilities align with the highest global standards. Developed after benchmarking against similar facilities in the USA, Japan, South Korea, Malaysia, and China, it stands as one of the most advanced facilities of its kind in India. This development is part of KAS Group’s initiative to establish India as a prominent player in the global semiconductor supply chain. It also reinforces Karnataka’s leadership in technology and engineering excellence, especially after the state’s latest policy push to its semiconductor industry. KAS Group has committed to investing INR 20 crore in the facility’s initial phase, with plans to inject INR 250 crore into future expansions. The company targets the USD 25 billion global semiconductor equipment market and expects the Indian ecosystem to grow to INR 2,500 crore within the next three to five years. Revenue projections for FY 2025-26 are estimated at INR 50 crore, with significant growth anticipated in subsequent years. Sanjeev Kumar Gupta, CEO of Karnataka Digital Economy Mission (KDEM), also took to LinkedIn to congratulate and announce the launch of this facility. “We at the KDEM are proud to support such initiatives that bolster the ESDM sector and contribute to India’s position in the $125 billion global semiconductor market,” he said. “With the first-of-its-kind infrastructure in India, capable of supporting industry giants, we look forward to collaborating with players like Applied Materials, LAM Research, Tokyo Electron, Yield Engineering Systems, etc,” Jyothinagara added. KAS Group, a key player in the engineering sector, has built a strong portfolio of companies, including UHP Technologies and KASTech Equipments, which serve the solar and semiconductor industries. Jyothinagara also anticipates that the country’s talent pool and robust engineering infrastructure will encourage MSMEs to confidently engage in the sector. With expertise spanning solar fabrication and semiconductor R&D facilities, the group has recently secured a significant contract with Micron Semiconductor Gujarat, further cementing its leadership in the industry.","excerpt":"FY 2025-26 revenue projections are estimated at INR 50 crore, with significant growth ahead.","categories":["AI News"],"tags":["Chip Manufacturing","manufacturing India","Semiconductor India"],"author_name":"Sanjana Gupta","publish_date":"2025-01-28T12:44:41","publication_year":"2025","word_count":457,"keywords":["programming_languages:R","AI","manufacturing India","Git","RAG","Semiconductor India","R","Chip Manufacturing"],"extracted_tech_keywords":["AI","RAG","R","Git","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/kasfab-tools-launches-indias-first-semiconductor-manufacturing-facility\/","complexity_score":3,"technical_depth":5,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10001535,"title":"7 Ways Drones Are Being Used In India","content":"Drone technology has been around for decades now, and with time this technology is just getting better. There was a time when drones were only used in military task or activities, however, drones today are being used by other people and that too on a day-to-day basis. Behind this ever-increasing popularity, drone companies also major role; their innovations and the way they have advertised. Apart from military use, let’s have a look at the other verticals where Indians are making the most of the drones. Here Are Some Of The Ways Drones Are Being Used In India Agricultural Drones It is not the IT sector that is reaping the benefits of evolving technologies, even agriculture is also of such vertical that has been touched by technology. The impact of technology in agriculture is positive, and with the advent of technologies like drones, it is experiencing a big change in improving the efficiency of agriculture. Fopple Technologies is one such company which is playing a vital role in this transformation. The company has devised drones that spray pesticides on the fields. Now, farmers don’t have to do it manually, which is both time-consuming and increases health at risk. https:\/\/www.facebook.com\/Fopple\/videos\/1903381016394346\/ In Mapping Property And Slum Areas According to one of our report, most of the litigation in India are related to property disputes or titles. And in order to deal with it, the Indian government, with the help of private companies, is using drones. Last year, the government of Odisha has been using drones to map out the urban slums. Also, the govt. was ready to give titles to 2,00,000 households in urban slums and those on the outskirts of cities by the end of 2018. Drone For Delivering Parcels A couple of years back, Amazon once filed patent for the use of drones in India for delivering e-commerce packages, and in response to that Indian Civil Aviation Ministry announced in 201 that it would soon be a reality to witness the use of drones for such purposes. If Amazon successfully pulls this off, then it would not only be one of the most quirky uses of drone that India will witness but it will also take e-commerce one step ahead. Drone Racing Last year in November, India witnessed a quirky drone event. In association with Indian Drone Racing League (IDRL), Bengaluru Tech Summit hosted India’s biggest and most extravagant drone racing competition. It was a night drone racing where pilots flew there drones at more than 160 kmph through a live video feed sent from the drones directly to the pilot’s goggles. Talking about quirky drone uses in India, this drone racing event is definitely one of amazing. Over 26 drone pilots from across India participated and Ritvik Suneel Nesargi from Pune was the winner who was awarded a cash prize of ₹1 lakh. First Aid Drone Last year in Chennai, India, a team of engineering students designed a drone that could be used to carry first-aid box accident spots where vehicles or ambulance can’t reach. According to the students, the drone was designed in such a way it can carry a first-aid box of up 8 kgs of weight and it can fly at a speed of up 70kmph. That is not all, the drone equipped with an interactive screen that is used to guide and explain the kind of first aid to be given based on the type of accidents and injuries. Traffic Monitoring And Security Talking about security and traffic monitoring, India is wholeheartedly accepting drones. Bengaluru is one of the cities in India to have drone surveillance in many parts of the cities. In December 31st, 2017, the city police deployed several drones to keep an eye on the crowd in order to ensure security. Places like MG Road, Koramangala, Indiranagar in Bengaluru were under drone surveillance during the new year’s night in 2017. Also, in 2018, the city was successfully monitoring all the activities. This is one of the best uses of drone that India has witnessed. Photography And Videography This is most probably the most popular use of drones in India. The nation is full of photography and videography talents and people always try out different ways. Also, YouTube has given a great platform to all the Indian to showcase their skills and with the help of drones, people are reaping the benefits. Furthermore, wedding photography and event after movie are also gaining a lot of popularity and drones seem to become an integral part of it. DJI, one of the well-known drone manufacturer has created some of the best drones that fulfil the needs of the aerial photographer\/videographer. Its Mavic 2 Pro, a foldable and portable drone is one of the most sought after drones today. Mavic’s Hasselblad L1D-20c Camera produces iconic images. It has a 20 MP 1” CMOS Sensor and an adjustable aperture of f\/2.8 – f\/11. Talking about the video side, the drone’s camera captures 4K 10-bit HD video.","excerpt":"Drone technology has been around for decades now, and with time this technology is just getting better. There was a time when drones were only used in military task or activities, however, drones today are being used by other people and that too on a day-to-day basis. Behind this ever-increasing popularity, drone companies also major […]","categories":["AI Features"],"tags":["Drone","Drones in india","Interviews and Discussions"],"author_name":"Harshajit Sarmah","publish_date":"2019-03-18T20:12:41","publication_year":"2019","word_count":828,"keywords":["Go","Drone","API","programming_languages:R","AI","innovation","programming_languages:Go","GAN","Ray","ViT","Drones in india","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","Ray","R","Go","API","GAN","ViT","innovation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/7-ways-drones-are-being-used-in-india\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10133273,"title":"How a Stanford Dropout’s Startup, Now Backed by OpenAI, is Shaping the Future of Education","content":"Xiaoyin Qu is not your typical AI startup founder. The super-energetic entrepreneur recently unveiled a new learning app for kids called Heeyo AI. Dropping out of Stanford Graduate School of Business to start her entrepreneurial venture and eventually in the learning and teaching space, stands testament to her unwavering commitment to revolutionise education with AI. During her late-night discussion with AIM, Qu enthusiastically explained the app, revealing that its vision stemmed from her own childhood desire for a coach who could guide her inventive and curious mind. “When you are 3 years old, you are just learning to talk. When you are 6 to 7, you can say multiple sentences. When you are 8 or 9, is when you can handle a lot of very open-ended creation. So, that’s why AI actually needs to change the way they talk to you based on your cognitive abilities, and we have both systems that can cater to that,” said Qu, founder and CEO of Heeyo AI, in an exclusive interaction with AIM. Defining themselves as ‘smart AI friends that help kids learn,’ California-based Heeyo AI received a $3.5M seed fund from some of the biggest tech players in the market including OpenAI startup fund, Amazon Alexa Fund, Par VC, Charge Ventures, StoryHouse Ventures and other top VCs. The startup is also funded by teams from top universities including Stanford and Harvard. Heeyo utilises various AI models, including text-to-speech, speech recognition, text-to-image, and text-to-music. They use specific models like OpenAI and Stable Diffusion for tasks such as content creation, translation, and image generation, with each project involving multiple models in a step-by-step process. The learning platform provides over 2000+ learning games where the AI figure can speak in over 20 languages through fun avatars such as pandas and dragons. Social Emotional Learning Heeyo is leveraging its AI platform, which is ideal for 3 to 11 year-olds, to address one of the biggest challenges of social emotional learning in children which is essentially helping kids improve their social skills. Qu highlighted the importance of teaching children how to make friends, express themselves, handle rejection, and respond appropriately when meeting someone new, and believes Heeyo will address all that. “That’s a relatively new thing that all Silicon Valley parents want to be interested in,” said Qu. Furthermore, Heeyo allows anyone including parents, educators, and kids to design their own learning games, and make them based on different cultures. “When you want to do a trivia on some Hindu traditions or Chinese traditions, you can actually build it up. In the US, we see some people are doing, like, Bible stuff. It’s like Bible parenting,” said Qu. Previous to this, Qu started Run The World, a leading virtual event platform, which she eventually sold last year. She has also led product management at Facebook and Instagram and led marketing & business development for Atlassian’s Asian market. AI for the Physical World Qu is also ecstatic with the kind of funding Heeyo has received, and considers to be very lucky to have OpenAI invested in it. Not just OpenAI, but being backed by Amazon Alexa too, the potential for future products is plenty. Qu highlighted how with many children already using Alexa, she believes that their content and interactive experiences could integrate well with Alexa to help even more kids. “In the long run, we plan to partner with toy companies, like family animals, and some of that. So, that’s going to happen. But, right now, we’re starting more from a content standpoint to make sure we have the right content,” said Qu. Interestingly, children’s friendly AI robot Moxie, was also built with the intention of helping children develop social skills. Talking about it, Qu mentioned that Moxie is built for kids with special needs such as autism and has seen many adopt them. However, the price range of it makes it inaccessible for all. “We could imagine that Moxie costs $899 and that’s pretty much not affordable for most families,” said Qu, who also said that Heeyo is accessible for all kids. The application is available for free and can be downloaded via the Apple or Google Play store. With AI entering the education domain, a number of universities are already tieing up with companies offering AI-related educational services. Recently, Andrej Karpathy, one of the former founders of OpenAI, launched Eureka Labs, which is an AI education company that aims to transform traditional teaching methods with generative AI. Furthermore, with AI helping with automating a number of administrative and repetitive tasks, teachers are free’d up for dedicating time for personalised interactions with students, thereby enhancing the quality of education. With advanced interactive modes of teaching emerging through AI, learning and even teaching is continuously transforming for the better. Heeyo AI not only offers learning games but also allows parents, educators, and kids to design their own games, including those based on specific ethnic cultures and traditions.","excerpt":"Heeyo is leveraging its AI platform, designed for 3 to 11-year-olds, to tackle one of the biggest challenges in social-emotional learning for children.","categories":["AI Trends"],"tags":["Editors Picks","learning","OpenAI"],"author_name":"Vandana Nair","publish_date":"2024-08-20T18:00:00","publication_year":"2024","word_count":820,"keywords":["Go","OpenAI","learning","AI","RPA","RAG","Editors Picks","Aim","stable diffusion","generative AI","R","Pandas"],"extracted_tech_keywords":["AI","generative AI","OpenAI","Aim","Pandas","RAG","R","Go","stable diffusion","RPA"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/how-a-stanford-dropouts-startup-now-backed-by-openai-is-shaping-the-future-of-education\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10008858,"title":"India Is Working To Develop A Supercomputer To Facilitate AI Framework","content":"India is working to develop a framework to help different walks of life in the longer-term in order to become an AI superpower. The Indian government is also working with the Centre for Development of Advanced Computing to develop a supercomputer to facilitate the AI framework. AI has been widely believed to play a significant role in improving governance, along with some of the top use cases in the field of social welfare, policymaking, and healthcare. In fact, countries such as the US and China have already made giant strides in this direction. And thus India needed to make its way as well. When asked, the CEO of National E-Governance Division, Abhishek Singh stated to the media that globally, we are recognised as a country which has a vast AI-skilled workforce, along with a good network of startup companies which are creating products. However, the only thing lacking is the compute capabilities, which is required. And that’s why the government is currently working on a framework and an ecosystem to facilitate that. Singh further stated that computing facilities are being set up in India and will allow the AI, startups, tech entrepreneurs and researchers to leverage the infrastructure that has been built to run their algorithms and to create “world-class AI products.” This comment from Singh comes right ahead of the RAISE 2020 event which has been designed to empower responsible AI for social empowerment — scheduled to start on 5th October. This summit will also feature startups working in AI-related fields, and those who are shortlisted through the AI Solution Challenge will also showcase their solutions in the AI Startup Pitchfest — scheduled for 6th October. Compared to other countries, India is in a unique position to become the new AI superpower. It indeed encompasses “a robust IT ecosystem and all the capabilities to democratise any technology.” According to Singh India ranks among the top countries in a Stanford AI index. It also ranks high when it comes to a relevant skilled workforce, startups segments, and IT experts and professionals that are skilled in AI. He further stressed the requirement for a necessary computing capability to run these complex algorithms. Explaining that point Singh further stated that, among the top 500 supercomputers in the world, for running complex algorithms, the majority is in China, with about 140 or 150 are in the US. But, India so far has only two such computers with one currently in use for weather prediction. Artificial Intelligence To Fight COVID India has been leveraging AI to address, fight and contain the COVID pandemic. AI is used to leverage the data that comes through platforms such as Aarogya Setu and to predict the next hotspot model. Singh said that “the process is purely based on data science and data analytics.” The Indian government has also used AI extensively for communications on the pandemic, where it has placed an AI-powered chatbot on Mygov portal. Such applications helped people to access the right information amid fake news on the virus. India is also helping other countries like Australia and the UK with its learnings in the field of AI. Incentives For AI Ecosystem According to Singh, there was a requirement for public investment in AI, to ramp up the AI efforts for catching up with the likes of the European Union, the UK, and China.“I’m sure very soon when the national programme on AI is launched, the amount of support the government will do towards researchers in AI, towards startups in AI, to create a compute infrastructure for AI, this will go up. So, already the national strategy on AI talks about a lot of this, about the creating AIRAWAT compute infrastructure, or about the course about the centre for research in artificial intelligence, centres of excellence in artificial intelligence,” said Singh. “So, all of it is something that is in works and, and public investment will also happen, which will further fuel private investment in such technologies.”","excerpt":"India is working to develop a framework to help different walks of life in the longer-term in order to become an AI superpower. The Indian government is also working with the Centre for Development of Advanced Computing to develop a supercomputer to facilitate the AI framework. AI has been widely believed to play a significant […]","categories":["AI News"],"tags":["AI in India","countries with quantum computers","India AI","India AI initiative","india ai strategy","Indian government using AI","limitations of AI"],"author_name":"Sejuti Das","publish_date":"2020-10-05T11:50:07","publication_year":"2020","word_count":661,"keywords":["India AI","data science","Go","artificial intelligence","AI in India","programming_languages:R","AI","countries with quantum computers","india ai strategy","RAG","responsible AI","analytics","Indian government using AI","India AI initiative","R","limitations of AI","startup"],"extracted_tech_keywords":["AI","artificial intelligence","data science","analytics","RAG","R","Go","responsible AI","startup","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/india-is-working-to-develop-a-supercomputer-to-facilitate-ai-framework\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":67583,"title":"ODI Match Winner: Weekend Hackathon #9","content":"MachineHack is back again with another exciting hackathon for this weekend. The 9th of the popular weekend hackathon series, and this time we challenge data scientists to predict the outcome of a cricket match in MachineHack’s ODI Match Winner: Weekend Hackathon #9. The challenge will start on June 19th Friday at 6 pm IST. Click here to participate Problem Statement & Description Our country shares a great deal of history with the game of cricket. Introduced as a royal game by the British during the British Raj, India took on the game as a popular sport even after its Independence. Today cricket is more than just a sport in India. In this hackathon, we challenge data science enthusiasts to predict the winning team of an ODI (One Day International) match. Given are 9 distinguishing factors that can influence the outcome of a cricket match. Your objective as a data scientist is to build a machine learning model that can accurately predict the winning team of an ODI match. Data Description:- The unzipped folder will have the following files. Train.csv –  2293 observations.Test.csv –  983 observations.Sample Submission – Sample format for the submission. Target Variable: MatchWinner The datasets will be made available for download on June 19th, Friday at 6 pm IST This hackathon and the bounty will expire on June 22nd, Monday at 7 am IST Below are the file formats for the provided data Train.csv Test.csv Sample_Submission.xlsx Click here to participate Bounties The top 3 competitors will receive a free pass to the Rising 2020. Know more about the Rising 2020 here. Click here to participate Rules One account per participant. Submissions from multiple accounts will lead to disqualificationThe submission limit for the hackathon is 10 per day after which the submission will not be evaluatedAll registered participants are eligible to compete in the hackathonThis competition counts towards your overall ranking pointsYou will not be able to submit once you click the “Complete Hackathon” button. You may ignore this featureWe ask that you respect the spirit of the competition and do not cheatThis hackathon will expire on 22nd June, Monday at 7 am ISTUse of any external dataset is prohibited and doing so will lead to disqualification Evaluation The leaderboard is evaluated using multi-class log loss for the participant’s submission. Click here to participate","excerpt":"MachineHack is back again with another exciting hackathon for this weekend. The 9th of the popular weekend hackathon series, and this time we challenge data scientists to predict the outcome of a cricket match in MachineHack’s ODI Match Winner: Weekend Hackathon #9. The challenge will start on June 19th Friday at 6 pm IST. Problem […]","categories":["Deep Tech"],"tags":["Hackathon","Machinehack Hackathon","Weekend Hackathon"],"author_name":"Amal Nair","publish_date":"2020-06-18T11:00:00","publication_year":"2020","word_count":385,"keywords":["data science","machine learning","Weekend Hackathon","programming_languages:R","AI","Hackathon","Machinehack Hackathon","R"],"extracted_tech_keywords":["AI","machine learning","data science","R","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/odi-match-winner-weekend-hackathon-9\/","complexity_score":3,"technical_depth":5,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10050302,"title":"9 Job Portals To Help You Land A Data Science Job","content":"A data science job requires combining computer science, statistics, and mathematics to analyse processes, develop models, predict future outcomes, and create actionable plans for complex challenges. According to the Harvard Business Review, ‘Data Scientist’ is the ‘Sexiest Job of the 21st Century.’ And, why not? Every digital disruption is a data scientist’s brainchild, and their increasing demand draws lucrative salaries. According to AIM Research, the overall adoption of Analytics and Data Science in large Indian companies stands at 74.5 per cent. Last year in October, more than 93,500 analytics jobs were available in India. However, finding the right opportunity at the right time may not always come easy. Earlier this year, GitHub announced that it was shutting down its jobs portal — a go-to site for tech companies to search for ML programmers and engineers. This is why we list the top job portals to find a data science job: Machine Hack Analytics India Magazine is building a machine learning community for data science and AI enthusiasts, Machine Hack. The platform is aimed at helping data scientists learn, grow and compete in the data science community by assessing skills, providing access to algorithms’ inventory, and organising hackathons. The platform also provides badges to competitors, helping them build a portfolio and ultimately get hired by top startups and companies. Machine Hack boasts of more than 40,000 registered developers and facilitates the hiring of 300-plus developers. Check all open jobs here. Post a job for free here. LinkedIn Founded in 2003, Microsoft owned California-based LinkedIn has been the go-to place for all job search and company culture-related questions. A usual search on the job portal website will display an average of more than 20,000 data scientist jobs in India itself. Even when not finding the right opportunities, LinkedIn helps users with professional networking, employers to keep track of potential candidates, and users to share their resumes and have conversations with hiring managers and recruiters. Monster Global employment website Monster was created in 1999 as a result of the merger between Online Career Center and The Monster Board. Headquartered in Massachusetts, Monster lists data science job postings by leading tech companies — Infosys, Amazon, PayPal, DELL, and others. Check the available data science jobs in India listed on Monster here. MLConf Job Board MLConf, or Machine Learning Conference, was founded in 2012 as a partnership with Carnegie Mellon University’s GraphLab team. In 2013, it established itself as a separate event to serve the ML and Data Science community in San Francisco. Today, it boasts of a growing ML community offering newsletters, workshops, talks, guest blog posts, interviews, and a free job board for individuals in Analytics. Y Combinator US-based seed money startup accelerator Y Combinator, or YC, was founded in 2005 by Paul Graham, Jessica Livingston, Robert Tappan Morris and Trevor Blackwell. So far, it has helped launch more than 2,000 companies, including Airbnb, Dropbox, DoorDash, Instacard and Reddit. In addition, YC has a job board that connects data scientists to over 400 startups that are funded by the accelerator. Stack Overflow Jeff Atwood and Joel Spolsky’s Stack Exchange Network’s flagship site, Stack Overflow, is a QnA website for programmers and developers. However, it also serves as a job board, not just for engineers but also for data scientists. In addition, it lists jobs posted by startups and tech companies from across the globe. AngelList Naval Ravikant and Babak Nivi-founded AngelList is the go-to place for startups. Founded in 2010, the San Francisco-based website connects startups, angel investors and job seekers with each other. The job board at AngelList works similarly to LinkedIn, where candidates have to create their own profile, fill in the details, and get access to a list of startup jobs. DataJobs Founded by Frank and Amy Lo, DataJobs accelerates big data recruiting by providing a platform that enables companies to post openings. It works as a pure-play hub format where data scientists can browse through job openings and apply at the ones that suit them the best. Besides data scientist jobs, DataJobs has openings for data analyst jobs, analytics jobs, hadoop jobs and database administrator jobs. Indeed Co-founded by Paul Forster and Rony Kahan in 2004, Indeed is a subsidiary of Japan’s Recruit Co. Ltd. and is available across 60 countries and in 28 languages. Indeed aggregates job listings from across websites, job boards, staffing firms, associations and company career pages. It claims to add ten jobs every second, globally. To know how to land the right data science job, check this article. Additionally, Analytics India Magazine also regularly tracks and posts about analytics job openings at top tech companies in the country. You can check these articles to understand the hiring patterns and data science openings in India.","excerpt":"At present, the overall adoption of Analytics and Data Science in large Indian companies stands at 74.5 per cent.","categories":["AI Hirings"],"tags":["Data Science Jobs","data science jobs in India","linkedin","linkedin jobs","Stack Overflow","Statistics for Data Science","Y Combinator"],"author_name":"Debolina Biswas","publish_date":"2021-10-04T13:00:00","publication_year":"2021","word_count":790,"keywords":["data science","Go","machine learning","AWS","linkedin jobs","AI","ML","Data Science Jobs","Stack Overflow","Y Combinator","data science jobs in India","RAG","Aim","Statistics for Data Science","analytics","linkedin","R"],"extracted_tech_keywords":["AI","machine learning","ML","data science","analytics","Aim","RAG","AWS","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-hiring\/9-job-portals-to-help-you-land-a-data-science-job\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10041243,"title":"Memory &#038; Storage Are At The Heart Of Semiconductor Industry: Rajesh Gupta, Micron Technology","content":"US-headquartered Micron Technology is one of the leading companies and semiconductor brands to offer state-of-the-art memory solutions to data centres. Analytics India Magazine caught up with Rajesh Gupta, Director, Country Manager- India Sales, Micron Technology to understand the status quo of the data centre industry; opportunities and challenges; and the road ahead. Excerpts: AIM: Tell us in detail about Micron’s memory solutions? Rajesh Gupta: Our broad portfolio of high-performance memory and storage technologies, including DRAM, NAND, and NOR, is transforming how the world uses the information to enrich life for all. We have successfully built memory chips using the world’s most advanced DRAM process technology, which offers major improvements in bit density, power and performance. Our innovation is helping customers protect themselves against security attacks of ever-increasing frequency and sophistication. Our specific innovation efforts include: Higher capacity and performance: In Micron’s core area of DRAM and Flash, we are increasing capacity, bandwidth, energy efficiency and packaging options for new form factors. New memory technologies: We are working on new ways to hold data, for which the terms “memory” and “storage” may no longer be adequate towards the delivery of independent memory and storage systems. Future innovation: From future memory and storage standards development to data innovations for quantum computing, Micron engineers incubate future technologies to fuel future entries to our data-centric portfolio. The recent announcement of 176-layer 3D flash memory, the world’s most technologically advanced NAND. New ways to hold data: Data Center IT and cloud managers want the fast, low latency and consistent performance of NVMe storage that won’t break the budget. The Micron 7`300 leverages the low power consumption and price-performance efficiencies of 3D NAND technology, delivering fast NVMe IOPS and GB\/s for a wide array of workloads. AIM: How does Micron cater to the needs of Data centre space? Rajesh Gupta: We deliver a broad portfolio of data centre memory and storage across hyperscale and enterprise environments. We meet the demands of today’s data centers through the acceleration of our DRAM and NAND innovation cycle, delivering new capability to data-centric platforms. To improve AI inference at scale, we have partnered with NVIDIA on the delivery of GDDR6X, the first in the industry to implement multilevel signalling in memory, creating a new benchmark for future generations of graphics memory. For energy-efficient memory scaling required in high-performance computing, we are delivering HBM solutions to the market. AIM: What does the future of data centre design look like? Rajesh Gupta: As we address today’s memory and storage infrastructure needs, we have an eye on the future of data centre design. Our fundamental computing architecture today is pushing against limits to future cloud scalability. Performance and energy efficiency are not improving fast enough to keep pace, and fundamental changes to the way compute platforms are built are required to push innovation further. The scale and efficiency of cloud computing and accelerated insight of AI has driven a requirement to re-architect data-centric infrastructure. The industry is developing infrastructure models to support fully composable, computing systems that create virtual compute, memory, storage and I\/O resource pools that can be independently scaled based on workload requirements. This moves places data at the centre of computing and represents a tremendous opportunity for memory and storage innovation. AIM: What is the role of semiconductor\/memory companies like Micron in data centre technology? Rajesh Gupta: The industry has identified the next wave of cloud innovation in the delivery of composable infrastructure where each element of the platform, compute, memory, storage and I\/O, can be scaled independently, and the fundamental equation of scale moves beyond the server node that has fueled data centre innovation for two decades. However, this vision is not possible without significant memory and storage innovations that move us towards a model where data, not compute, is at the centre of the system. Micron technology innovation delivers new architectures for composable memory pools and scalable data centre storage resources. Our solutions enable the collection, storage, and management of data, turning that data into insights and intelligence with unprecedented speed and efficiency. AIM: What’s your take on the looming global chip crisis? Rajesh Gupta: From computers to cars, chips are at the centre of them all. There has been a rise in demands for memory due to increasing 5G smartphones and artificial intelligence software. The global shift to remote work due to the pandemic has also led to a boost in the demand. The market is currently facing a severe undersupply. This shortage is expected to tighten through the year. We are looking at a multitude of ways to solve customer’s needs for increased supply. Massive demand across various end markets, combined with disruptions at a few logic and foundry semiconductor producers, has resulted in a shortage of non-memory ICs, and we believe memory demand would have been even greater without these shortages. Despite the challenges of pandemic, shortages in the electronics industry, and other disruptions at our Taiwan fab, Micron has been able to mitigate the impact on our production output through proactive supply chain and inventory management strategies. AIM: Tell us about India’s semiconductor market Rajesh Gupta: Today, the industry is challenged to fuel the wide adoption of multi-cloud infrastructure to enable broad deployment of AI-enabled applications. Memory and storage are at the heart of the semiconductor industry and are a critical enabler of the technologies driving economic growth and well-being. There is a complete shift in the outlook of businesses and the industry. We can foresee a new paradigm when it comes to changes in the way companies are working today because of the new hybrid working models that have come into play post-Covid-19. There is a significant change in consumer behaviour, leading to increasing online activity, including e-commerce, gaming, and video streaming, all of which is driving additional data centre capacity requirements. Recovery from the pandemic and pent-up demand are expected to drive demand growth in markets such as enterprise, cloud, desktop PCs, mobile, auto industry. AIM: What are the short-term and long-term plans for Micron? Rajesh Gupta: The long-term plan for us revolves around building our portfolio. We are on track to begin customer qualification of our next-generation client SSDs using 176-layer NAND in the fiscal second half of 2021. By the end of 2021, Micron expects to cover multiple segments of the market, including consumer, value OEM and premium OEM with our 176-layer-based client portfolio. We partner closely with our customers, industry, governments, communities, and team members to develop the next disruptive applications and end products in a sustainable way across the 17 countries in which we operate. Micron is working with key OEMs and tier-one partners on centralizing storage, enhancing security, and deploying automotive-grade SSDs for next-gen architectures for 2023 and beyond in the areas of IVI and automated driving and central computer architectures. On the India front, we have had a few key collaborations over the last year. We recently partnered with Tata Communications to announce the launch of the world’s first cloud-based embedded SIM, a ground-breaking solution to accelerate and simplify the deployment of IoT devices at a global scale. Also, in line with supporting the tech talent and fostering technological innovation along with boosting economic growth, Micron recently introduced the Micron University Research Alliance (URAM) that focused on building talent in science, technology, engineering, and math (STEM) and aimed to reach over 10,000 students in India.","excerpt":"US-headquartered Micron Technology is one of the leading companies and semiconductor brands to offer state-of-the-art memory solutions to data centres.  Analytics India Magazine caught up with Rajesh Gupta, Director, Country Manager- India Sales, Micron Technology to understand the status quo of the data centre industry; opportunities and challenges; and the road ahead. Excerpts: AIM: Tell […]","categories":["AI Features"],"tags":["Chip shortage","data centres","Indian market","Interviews and Discussions","memory","semiconductor industry","storage"],"author_name":"Shraddha Goled","publish_date":"2021-06-05T10:00:00","publication_year":"2021","word_count":1229,"keywords":["Go","artificial intelligence","TPU","Indian market","AI","cloud computing","semiconductor industry","RAG","Ray","Aim","Chip shortage","memory","data centres","storage","analytics","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","artificial intelligence","analytics","Aim","Ray","RAG","cloud computing","TPU","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/memory-storage-are-at-the-heart-of-semiconductor-industry-rajesh-gupta-micron-technology\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10052929,"title":"HCL Technologies To Hire 10,000 Professionals For AWS Business Unit","content":"HCL Technologies recently announced that the company will be adding 10,000 professionals to boost its newly launched Amazon Web Services (AWS) Business Unit (AWSBU) to help enterprises worldwide accelerate their cloud transformation journey. The new devoted business unit inside HCL will be supported by AWS engineering, options and business groups which will help businesses modernise their legacy systems and mainframe applications and reliably adopt cloud technologies that boost efficiency, achieve objectives and meet regulatory compliance, all the while migrating and managing SAP workloads on AWS. HCL currently holds five AWS competencies, has more than 10,000 professionals trained on AWS, and plans to boost this capacity to more than 20,000 specialists in the near future. “HCL is an AWS Premier Consulting Partner capable of providing an end-to-end road map for adopting AWS to best serve our mutual customers, with a firm heritage in modernising infrastructure, applications and data. With the launch of the HCL AWS Business Unit, HCL will leverage its vertical-first focus on FSI, telco, and energy and utilities plus deep technical expertise on SAP, contact centres, hybrid cloud and mainframe modernisation transform businesses and consumer behaviour using technology when they need it the most,” said Doug Yeum, Head of WW Channels and Alliances, AWS. Kalyan Kumar, Chief Technology Officer and Head, Ecosystems, HCL Technologies, said, “The AWS BU is an important part of our larger #HCLCloudSmart strategy in servicing every aspect of cloud delivery to build effective ecosystems that help our customers stay ahead of their competitors. Leveraging our extensive relationship with AWS, the AWS BU synergises the best of what both companies have to offer, driving digital, cultural and customer-centric transformations.” HCL will also provide end-to-end AWS Managed Services through HCL’s ElasticOps and help build mature solutions on AWS, optimise costs and provide improved business agility. The unit will assist enterprise shoppers to modernise and migrate at scale, maintain value benefits and deal with experimentation by combining HCL’s expertise and experience in constructing adaptive cloud smart portfolios and AWS’ depth and breadth of companies and fast innovation. HCL is uniquely positioned to help enterprises, both as a global systems integrator (GSI) and an independent software vendor (ISV), with a cloud-focused ecosystem and product innovation strategy. The company unit also hopes to facilitate revenue growth and user experience by developing customised industry solutions built with AWS services and investment.","excerpt":"HCL Technologies recently announced that the company will be adding 10,000 professionals to boost its newly launched Amazon Web Services (AWS) Business Unit (AWSBU)","categories":["AI News"],"tags":["AWS cloud","HCL Technology","hiring trends"],"author_name":"Victor Dey","publish_date":"2021-11-05T12:37:16","publication_year":"2021","word_count":390,"keywords":["hiring trends","AI","AWS","innovation","cloud_platforms:AWS","programming_languages:R","Git","RAG","HCL Technology","cloud_platforms:Amazon Web Services","AWS cloud","R"],"extracted_tech_keywords":["AI","RAG","AWS","R","Git","innovation","cloud_platforms:AWS","cloud_platforms:Amazon Web Services","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/hcl-technologies-to-hire-10000-professionals-for-aws-business-unit\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10120637,"title":"SkyServe&#8217;s STORM Ushers in &#8220;Smartphone Moment&#8221; for Satellite Imaging","content":"Bengaluru-based space tech company SkyServe announced today that it has successfully achieved Smart Earth Imaging in orbit, marking an important step forward in Earth observation. By uplinking and testing their edge computing software stack STORM on a satellite, SkyServe demonstrated the ability to generate actionable insights from space in a fraction of the time. In mid-April, SkyServe collaborated with space logistics company D-Orbit to deploy STORM on a SpaceX-launched satellite. Within seconds of capturing imagery over the Egypt-Sinai Peninsula, STORM performed intelligent tasks onboard, including error correction, cloud\/water removal, and vegetation identification. The optimised data was then transmitted back to Earth, compressed by 5X. “We’re essentially creating the iPhone moment for Earth observation,” said Vinay Simha, SkyServe’s Co-founder and CEO. “Just like Smartphones revolutionised data accessibility and user engagement, STORM enables geospatial applications with edge tasking and processing, unlocking vast new use cases from space.” STORM signifies a paradigm shift beyond traditional Earth observation, empowering satellites to optimise for specific customer applications and deliver near-real-time insights dynamically. “The collaboration with SkyServe facilitates their in-orbit STORM platform and aligns with our mission to provide comprehensive in-orbit services,” noted Viney Jean-Francois Dhiri, D-Orbit’s Head of Business Development. SkyServe is now leveraging Loft Orbital’s YAM-6 satellite for Mission Denali, showcasing automated tasking and hosting of geospatial models for various use cases. “We can program the satellites to identify wildfires while flying over regions like Australia and monitor water resources over Bengaluru,” explained Vishesh Vatsal, SkyServe’s Co-founder and CTO. Later this year, SkyServe’s Mission K2 is scheduled to launch aboard ISRO’s PSLV C59. These missions are unlocking a new paradigm of real-time space-based insights by empowering analytics companies with onboard data processing capabilities.","excerpt":"In mid-April, SkyServe collaborated with space logistics company D-Orbit to deploy STORM on a SpaceX-launched satellite.","categories":["AI News"],"tags":["Earth Observation Satellites"],"author_name":"Shyam Nandan Upadhyay","publish_date":"2024-05-16T13:35:45","publication_year":"2024","word_count":279,"keywords":["Earth Observation Satellites","AI","programming_languages:R","RAG","analytics","edge computing","R"],"extracted_tech_keywords":["AI","analytics","RAG","edge computing","R","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/skyserves-storm-ushers-in-smartphone-moment-for-satellite-imaging\/","complexity_score":2,"technical_depth":6,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10039836,"title":"What Is Text Modular Network?","content":"Complex machine learning tasks such as question-answer and numerical reasoning will be easier to solve if they are decomposed into smaller functions that existing methods can solve. Based on this approach, a team of scientists from the Allen Institute for AI developed a general framework called the Text Modular Networks for interpretable systems. TNN, explained Text Modular Networks (TMN) learn the textual input-output behaviour of existing models through their datasets. It is different from earlier approaches involving task decomposition, explicitly designed for each task and produced decomposition independently of existing submodels. For this study, the team selected the Question Answer task to show how to train a next-question generator to produce sub-questions targeting appropriate submodels sequentially. The next-question generator lies at the core of the TMN framework. The output is a sequence of sub-questions and the answers providing a human-interpretable description of the model’s neuro-symbolic reasoning. The TMNs use only distant supervision learning to learn how to produce these decompositions; additionally, there is no need for an explicit human annotation. The team also observed that by giving appropriate hints, the capabilities of the existing sub-models can be captured by training a text-to-text system to generate the questions in the sub-models training dataset. To generate questions, the team trained a BART model, which is a denoising autoencoder for pretraining sequence-to-sequence models, and fed preferred vocabulary as hints. The sub-task question models generated the sub-questions and identified appropriate sub-models. Through this, the team was able to extract likely intermediate answers from each step of the complex question. The resulting sub-questions are in the language of corresponding sub-models that can be now used to train the next-question generator to repeat the whole process. Using the TMN framework, a modular system, MODULARQA, was developed. This system gives the reasoning in natural language by decomposing complex questions to those answerable by two sub-models — neural factoid single-span QA model and a symbolic calculator. MODULARQA was evaluated on questions from two datasets–DROP and HotpotQA that results in the first cross-dataset decomposition-based interpretable QA system. Its implementation involves multi-hop questions that can be answered using five classes of reasoning found in the existing QA dataset: composition, comparison, conjunction, difference, and complementation. Exciting new work from the Aristo team at AI2!The ModularQA system can explain each of its reasoning steps in terms of a simple question and its answer – in today's world of black-box models, this is an important step towards explainable #AI.#NLProc #MachineLearning https:\/\/t.co\/E6DnFSgtPg— Allen Institute for AI (@allen_ai) May 3, 2021 MODULARQA demonstrated cross-dataset versatility, robustness, sample efficiency and ability to explain its reasoning in natural language. It even outperformed BlackBox methods by 2 percent FI in a limited data setting. Comparison with previous approaches Earlier QA systems were generally designed as a combination of distinct modules that comprised outputs of lower-level language tasks to solve higher-level tasks. While a good approach, its application has been limited to pre-determined composition structures. The question decomposition method has been pursued before as well. However, there are a few issues with this, such as: A few methods focussed directly on training a model to produce subquestions using question span. This technique is found to be unsuitable for datasets such as DROP.Many techniques generate simpler questions without capturing the required reasoning.An approach where the model collects full question decomposition meaning representations (QDMR) annotations is effective. However, it may still require human intervention and may not generalize well. In contrast, TMNs start with pre-determined models and generate decompositions in their language. There have been many multi-hop QA models designed for HotpotQA and DROP. However, these models are often complex that focus on only one of the two datasets. These models could produce post-hoc explanation only on HotpotQA when the supporting sentences are annotated. However, these explanations are not faithful and are often shown to be gameable. With TMN, however, the scientists were able to produce explanations for multiple datasets without needing such annotations; this makes it more generalizable to future datasets. By comparison, TMN is similar to the models based on neural module networks, which compose task-specific simple neural modules. However, the two approaches differ mainly on two grounds: The formulations of NMN target only one dataset and do not reuse the existing QA system; It provides an attention-based explanation, the interpretability for which is unclear. Read the full paper here.","excerpt":"Complex machine learning tasks such as question-answer and numerical reasoning will be easier to solve if they are decomposed into smaller functions that existing methods can solve. Based on this approach, a team of scientists from the Allen Institute for AI developed a general framework called the Text Modular Networks for interpretable systems. TNN, explained […]","categories":["AI Trends"],"tags":["AI Tool","Allen Institute for AI"],"author_name":"Shraddha Goled","publish_date":"2021-05-08T18:00:00","publication_year":"2021","word_count":721,"keywords":["Go","machine learning","TPU","programming_languages:R","AI","programming_languages:Go","Allen Institute for AI","NLP","ai_applications:NLP","AI Tool","R"],"extracted_tech_keywords":["AI","machine learning","NLP","TPU","R","Go","programming_languages:R","programming_languages:Go","ai_applications:NLP"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/what-is-text-modular-network\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10092051,"title":"Apple&#8217;s Mac Pro Chip Dilemma: From Core Strength to Core Stress","content":"The year was 2020, when Apple announced that in the next two years, the company would transition all of its Mac lineup to its own chips rather than depending on Intel. However, with Apple’s Mac Pro still coming with Intel’s Xeon W processor, the company seems to be dragging its feet on the deadline. And while Apple’s Worldwide Developers Conference (WWDC) was where the 2019 and 2013 Mac Pro were launched (shipping later in the year), this year’s WWDC, which is happening in June, might also not see the announcement of the new Mac Pro as per the latest newsletter by Bloomberg’s Mark Gurman. Quicker Doesn’t Mean Better Speculation is rife that Apple was developing an M1 Extreme-powered 48-core CPU, 152-core GPU Mac Pro, which was supposed to be launched in 2023, but is now cancelled. The reports suggest that Apple will be launching Mac Pro featuring M2 Ultra – a newer version of the same chip used in the Mac Studio. M2 Ultra was made by putting together two M2 Max chips and joining them with a high-quality bridge so that they can work together without losing any power. Since it was a success, Apple wanted to replicate the same formula for its Mac Pro, where it tried to produce the M1 Extreme by putting together four chips. However, it had to stall the plan over “complexity and cost” issues. As per Gurman, “the complexity and cost of producing a processor that is essentially four M2 Max chips fused together” was too much for Apple, given the small size of its target market. “Producing chips was never Apple’s forte, and while it achieved success with its M1 and M2 chips for MacBook lineup, it will take some time to build a chip to support a machine as powerful as Mac Pro,” a source at AMD told AIM. However, the Mac Pro with an M2 Ultra chip too doesn’t make sense. Why would customers buy the Mac Pro with M2 Ultra rather than the cheaper Mac Studio with the same processor? Additionally, some rumours also support the theory that the new model will lack user-upgradeable RAM due to the unified memory being built as a physical part of the M2 system-on-chip but will have “slots” for storage, graphics, media and networking cards. However, Apple also can not come with Intel Xeon processor as AMD already has a better one to offer and Apple has decided to part ways with Intel’s Xeon lineup. When asked what exactly was Apple’s problem with Intel Xeon, Intel refused to comment. AMD EPYC– how does it perform? Now that Intel’s chips are not being used by Mac Pro and Apple’s own chips are not as powerful as they should be, the only choice that remains is AMD’s EPYC chip. Until now the EPYC chips were generally used at the server side, but now it’s also being touted as an alternative to the workstation-class processors. To compare with Intel Xeon processors, AMD EPYC processors offer a higher core count compared to Intel Xeon processors, with the latest EPYC processors having up to 64 cores, while Intel Xeon processors have a maximum of 28 cores. However, the Intel Xeon processors have a higher clock speed compared to AMD EPYC processors. This is because Intel Xeon processors are designed for single-threaded performance, while AMD EPYC processors are optimised for multi-threaded performance. The AMD EPYC processors support more memory per socket compared to Intel Xeon processors. The latest EPYC processors support up to 4TB of memory, while Intel Xeon processors support a maximum of 2TB of memory. Additionally, the AMD EPYC processors have more PCIe lanes compared to Intel Xeon processors. The latest EPYC processors have up to 128 PCIe lanes, while Intel Xeon processors have a maximum of 48 PCIe lanes. The benchmark tests on Intel Xeon and AMD Epyc processors suggest that AMD EPYC outperforms Intel Xeon in several scenarios. For instance, in applications that require high core count and multi-threaded performance, AMD EPYC processors deliver better performance than Intel Xeon processors. Furthermore, AMD EPYC processors also come with a higher number of PCIe lanes, which makes them more suitable for tasks that require high I\/O bandwidth. On the other hand, Intel Xeon processors have a slight edge when it comes to single-threaded performance and clock speed.","excerpt":"Apple won’t come with Intel Xeon processor as AMD already has a better one to offer and the former has decided to part ways with Intel’s Xeon lineup","categories":["Global Tech"],"tags":["AMD","Apple","Intel","NVIDIA"],"author_name":"Lokesh Choudhary","publish_date":"2023-04-21T12:00:00","publication_year":"2023","word_count":720,"keywords":["Replicate","AMD","programming_languages:R","AI","Apple","RAG","Aim","NVIDIA","R","Intel"],"extracted_tech_keywords":["AI","Aim","RAG","R","Replicate","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/apples-mac-pro-chip-dilemma-from-core-strength-to-core-stress\/","complexity_score":2,"technical_depth":6,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10081304,"title":"Teaching Machines to Unlearn","content":"In recent years, the debates and developments surrounding data, privacy, and big-tech, have brought up “the right to be forgotten” which essentially means that users should have the right to decide whether and when their personal data can be made inaccessible, or deleted. In the context of artificial intelligence, this means that machine learning models should be designed to forget and discard irrelevant information. Companies like Amazon and Flipkart spend millions of dollars to build recommendation engines and the best algorithm to suggest products by tracking customer choices. A product distribution website recommending you products by accessing the information you provide to them might feel like a moderate bargain. The idea of machine learning is essentially feeding chunks of information into a machine so that it can remember it, and produce results as you want it to. Now this information, which is called data, is one of the biggest factors that decides the efficiency and accuracy of the model. But can machines unlearn this data that they are trained on, if a controversy arises? Though this “duty to forget” is not currently being considered to be up there with human rights, the idea still strikes as meaningful. Researchers and scientists should delve into developing methods so that machines, apart from learning, can also “unlearn” things, i.e. remove the input data and thus settle the debate about AI risking privacy. Machine unlearning can improve data privacy by allowing machine learning models to forget sensitive information that may have been included in the training data. In some cases, personal or sensitive information may be inadvertently included in the data used to train a machine learning model. This is particularly important in fields like healthcare, where personal information must be kept private. Researchers have been trying to find an effective and efficient way to tackle this challenge and build a “machine unlearning” algorithm. In 2020, researchers from University of Toronto and University of Wisconsin tried the SISA method to remove data primarily concerned about privacy for users. But since then, very little improvement has been made in the field. Why is it necessary? We’ve all heard about the Facebook controversy that erupted when the company leaked 87 million users’ personal information across the web for political advertising, resulting in a stream of lawsuits and users exiting the platform. In 2020, Facebook revealed the “clear history” button on its website, which was supposed to delete the user’s data from the website, but all it did was remove the user’s ability to check if the data is still there. It is not easy for users to delete data, for which, the access was granted to the companies or models. Machine Unlearning“Once users have shared their data online, it is difficult to revoke access and ask for the data to be deleted. ML exacerbates this problem because any model trained with said data may have memorized it, putting users' privacy at risk.” https:\/\/t.co\/ki3ZRI9tLB pic.twitter.com\/RHxOjbb624— hardmaru (@hardmaru) January 29, 2020 However, removing the data that a machine is trained on is not an easy task. The concept of “machine unlearning” is to remove or reduce the training data without affecting the performance of the model. Apart from the privacy part of it, machine learning models are prone to biases that often occur because of underfitting or overfitting of data. This can result in a system that does not achieve perfect results when testing. Thus, the developers have to start from scratch, select or build another dataset, and build the model again, which proves to be a cumbersome process. Machine unlearning is not a new topic, but is a territory that definitely needs more exploration. An important use case of machine unlearning is to remove the unwanted data from a dataset to improve the accuracy even further. For example, Amazon’s sexist recruitment system that was biassed against women when scanning their profiles, was fed on the dataset of the engineering field which is largely male-dominated. Thus, cleaning the data that it is trained on is essential to remove the bias. This is where algorithms for machine unlearning can greatly improve the models without building the model from zero. Conclusion Though the discussion about if the data that the machine learning algorithms are trained on in these websites is stored on their servers is still unsolved, the algorithms that the models are built on definitely can still access it. The recent draft of the Digital Personal Data Protection Bill of India talks about the privacy of an individual’s data and how organisations need to delete data no longer needed. This can include the right to access, correct, or delete their personal information. It pushes the researchers in this field to make more innovations and figure out what needs to be done to delete datasets of ML models and improve the privacy of users and make models perform better. Machine unlearning has been proven to not be an easy challenge but the approaches that have already been tested still require a lot of improvement. With the increasing regulations, policies, and parameters for machine learning models, the need for AI to unlearn things is the need of the hour.","excerpt":"With the increasing policies around data protection and user privacy, AI’s need to forget information is also a rising topic.","categories":["IT Services"],"tags":["data protection","dpdp india","machine learning problems"],"author_name":"Mohit Pandey","publish_date":"2022-12-02T10:00:00","publication_year":"2022","word_count":858,"keywords":["Go","dpdp india","artificial intelligence","machine learning","AWS","AI","R","ML","Git","GAN","machine unlearning","data protection","machine learning problems"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","machine unlearning","AWS","R","Go","Git","GAN"],"url":"https:\/\/analyticsindiamag.com\/it-services\/teaching-machines-to-unlearn\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":56560,"title":"Key Insights From The First-Ever Apache Airflow Meetup Hosted By Qubole","content":"Apache Airflow has been one of the best workflow management systems around. In modern business practices, the software serves as a platform where huge workloads can be easily be managed by organisations. The recent meetup hosted by Qubole, together with co-hosts Twitter and Astronomer, was held at their Bangalore office on Feb 8th, 2020 – it represented the first Apache Airflow meetup, where the speakers shed light into a few important topics and updates on the popular software, Apache Airflow. From understanding platform insights on achieving high scale using Airflow to how Qubole is utilising its data using Airflow, this meet up went beyond and above. The meetup was overflowed with more than 100 attendees and witnessed a considerable amount audience from the Airflow community. Due to several benefits such as modular architecture, extensibility, being dynamic in nature, etc., several popular organisations like Twitter and Netflix, have started adopting this software into their platforms. Qubole have been working on Airflow for a few years now. According to the sources, last year, the enterprise cloud startup, Astronomer has raised $5.7 million in seed funding to deliver enterprise-grade Apache Airflow. There were a total of four talks where the discussions mainly centred on how Apache Airflow portrays a vital role in organisations when it comes to monitoring workflows. The talks were delivered by industry data experts that included, Kaxil Naik, an Apache Airflow PMC Member & Core Committer, and a Senior Data Engineer at Astronomer; Raman Gupta, a Senior Computer Scientist at Adobe; Devjyoti Patra of Qubole; and Sumit Maheshwari, Apache Airflow PMC & Tech Lead at Twitter. Click Here to Subscribe to Qubole’s APAC Field Events Talk 1: Airflow Best Practices & Roadmap to 2.0 This talk was delivered by Kaxil Naik from Astronomer. In this talk, Naik spoke about various intuitive features of Airflow version 1.10.8 which have been updated along with some bug fixes, a few interesting tips for writing Directed Acyclic Graphs (DAGs) as well as the upcoming upgrades in Airflow 2.0. Some of the new features discussed are mentioned below Addition of tags to DAGs and using it for filtering.Allowing adding conf. option to add DAG Run viewAllowing externally triggered DAGs to run for future exec datesUpdating documentation.Center DAG on graph view load. Next, Naik shared some of the tips and tricks for writing DAGs in order to reduce the redundancies and improve the way to write DAGs. Some of them are mentioned below. DAGs as Context Manager: DAGs can be used as context managers to assign new operators to DAG automatically.Default Arguments: In Airflow 1.10.8, one can use default_args to avoid repeating arguments. For instance, If a dictionary of default_args is passed to a DAG, it will apply them to any of its operators. This makes it easy to apply a common parameter to many operators without having to type it many times.Using the list to task dependencies.Restrict the number of Airflow variables in the DAG.Avoid code outside of an operator in DAG filesUse Flask-App builder based UI.Set configs using environment variables.Instead of using initdb, apply migration using “airflow upgragdedb” Lastly, Naik also discussed some of the enhancements and the roadmap to 2.0; some of them are mentioned below. DAG SerialisationRevamped real-time UI.Production-grade modern API.Official Docker Image and Helm Chart.Scheduler ImprovementsData Lineage. Talk 2: Adobe Experience Platform Insights on Achieving High Scale Using Apache Airflow This talk was delivered by Raman Gupta, who is the Senior Computer Scientist at Adobe. In this talk, Gupta discussed how the Platform Orchestration Service at Adobe is utilising AirFlow to fulfil the need of workflow management system. He also discussed the APIs exposed to its users to manage workflows as well as the challenges faced by them and steps to overcome them. Some of the key challenges faced are mentioned below: To run 1000 tasks concurrently.To keep scheduling latency within some predefined threshold.To improve service usage or debugging experience. Then, Gupta talked about some of the steps which are being taken to achieve high-scale with latency, as mentioned below: Support for multiple Airflow cluster behind Orchestration service. This will also restrict the number of workflows per cluster to control.Scalability such as support for multi-cluster in Orchestration service, Airflow setup on Kubernetes to scale worker task horizontally and support for reschedule mode to achieve high concurrency. Talk 3: How Qubole Dog Food Its Own Data Using Airflow? The speaker of this talk was Devjyoti Patra from Qubole. In this talk, Patra gave an overview of the Qubole unified data platform, and he showed how to create an Airflow cluster easily just by using the console. He then discussed the ETL workflow orchestration and some of the use-cases at Qubole where Airflow is being used. For instance, Qubole has BiFrost Pipeline for sharing data with clients, run multiple pipelines for billing, insights, alerts and reports, validating the output data from upstream pipelines, among others. Talk 4: WTF is PEX & How Twitter Uses it to Deploy Airflow This was the last talk for the event which was initiated by Sumit Maheshwari, Apache Airflow PMC & Tech Lead at Twitter. In this talk, Sumit discussed what PEX or Python executables are, and how it is similar to .exe files. He also discussed Python ANTS or PANTS, which is an Ant-based Python built tool and the ways to use it to build Airflow.pex file. Lastly, The speaker discussed a few points on how to deploy Airflow on Kubernetes. Some of the points mentioned are as follows Bundle airflow.pex file into docker image on top of a python-based image.Supply Airflow configs via Kubernetes Config mapWrite\/Read task logs from GCS. Outlook With the conclusion of the meetup, the audience learnt a lot about the Apache Airflow as well as a few tips and tricks on how to use this software with ease. Some of the key takeaways are mentioned below. AirFlow 2.0 will include the official Docker Image and Helm Chart.How Adobe is using Airflow to monitor workflow. Understanding Qubole’s unified data platform.The use of PEX and DAGs at Twitter.How to deploy Airflow on Kubernetes. Click Here to Subscribe to Qubole’s APAC Field Events","excerpt":"Apache Airflow has been one of the best workflow management systems around. In modern business practices, the software serves as a platform where huge workloads can be easily be managed by organisations. The recent meetup hosted by Qubole, together with co-hosts Twitter and Astronomer, was held at their Bangalore office on Feb 8th, 2020 – […]","categories":["Deep Tech"],"tags":["TED Talks"],"author_name":"Ambika Choudhury","publish_date":"2020-02-12T19:18:00","publication_year":"2020","word_count":1018,"keywords":["API","TPU","AI","docker","Scala","TED Talks","RAG","Python","Ray","R","kubernetes"],"extracted_tech_keywords":["AI","Ray","RAG","kubernetes","docker","TPU","Python","R","Scala","API"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/the-key-insights-from-the-first-ever-apache-airflow-meetup-hosted-by-qubole\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":43458,"title":"How Online Companies Are Tapping Into Object Detection To Improve User Experience","content":"The accuracy of state-of-the-art object detection systems is often under scanner for seemingly obvious reasons. From unlocking the phone to self-driving cars, object detection is almost everywhere. As computer vision applications grow in popularity, it has become crucial to keep their flaws in check or at least detect them in the first place. These flaws usually are a culmination of unhealthy data collection strategies and biases — both inductive and engineered. A wrongly identified image can spew results which can spiral a business into a catastrophe. Object detection had been used to set up profitable business ventures. One such example is that of the billion dollar hospitality venture Airbnb. Brian Chesky co-founded peer-to-peer room and home rental company Airbnb with Nathan Blecharczyk and Joe Gebbia in 2008. Now, almost 11 years later, Airbnb has been used by more than 300 million people in 81,000 cities in 191 countries. Airbnb platform has millions of listings that cover living space around the world. The quality of these listings is a top priority, as the tastes of customers depend greatly upon the aesthetics of the interiors. To do this Airbnb had to determine whether the amenities advertised online match the actual ones. For  Airbnb, knowing that kitchenware exists in a picture does not speak much about the type of room. Likewise, knowing there is a table in the picture doesn’t help either. The goal here is to understand whether the detected amenities provide convenience for guests and can assist the customer in decision making. Because a family trip might require a room with spacious kitchen when compared to a bachelor’s trip. The search results should allow the customer to draw more insights. Airbnb platform is a fine example of using machine learning algorithms to improve the user experience. Achieving High Quality Detection The above picture is a sample of amenity detection result of a third-party API service, from an industry-leading vendor. The picture below, depicts the amenity detection using Airbnb API: Source: Airbnb. The detailing in the Airbnb results, look more insightful. To achieve this, data scientists at Airbnb had to face some data challenges. From finding suitable annotated data to cleaning it, from manual classification to developing a robust model, the hurdles were plenty. To address the issues with  taxonomy that encompasses amenities (kitchenware, furniture etc) is an open-ended question. The taxonomy was unclear and the data science team at Airbnb began  with something lightweight. They have found Open Image Dataset V4, that offered a vast amount of image data. It has more than 9 million images that had been annotated with image-level labels, object bounding boxes (BB), and visual relationships. Along with this, the team had also manually reviewed the 600 classes and selected around 40 classes that were relevant to the use case. For labelling the data, the team used Google data labelling service and to ensure the diversity of the dataset, some in-house data was added to the public data resulting in an evenly distributed amenity classes. For training, two pre-trained models were chosen: ssd_mobilenet_v2 and faster_rcnn_inception_resnet_v2. The accuracies of the pre-trained models were tested on a 10% held-out data which contains 7.5k images with 30 object classes. Mean Average Precision (mAP) was the metric used, which measures the average precision of a model across all object classes. Broadening The Scope Of Object Detection The team also experimented with Google AutoML vision and to their surprise, they found the results to be promising. They were able to train an object detection model on 75k annotated images within 3 days. They also found that model deployment on AutoML to be easy as the model was turned into an online service which people could easily use through REST API or a few lines of Python code. In addition to Amenity Detection, Object Detection with broader scope is another important area that Airbnb looks to count on going forward. This is due to the promising results of Broad-scope Object Detection, which provided the data scientists with necessary content moderation to prevent things like weapons, large-size human faces, etc. from being exposed without protection and in way help Airbnb become a smarter and safer home-sharing platform. Know more about amenity detection here.","excerpt":"The accuracy of state-of-the-art object detection systems is often under scanner for seemingly obvious reasons. From unlocking the phone to self-driving cars, object detection is almost everywhere.  As computer vision applications grow in popularity, it has become crucial to keep their flaws in check or at least detect them in the first place. These flaws […]","categories":[],"tags":["Airbnb","image recognition","Object Detection"],"author_name":"Ram Sagar","publish_date":"2019-07-26T18:30:15","publication_year":"2019","word_count":699,"keywords":["data science","machine learning","AWS","AI","ML","image recognition","computer vision","RAG","Python","object detection","Object Detection","Airbnb","R"],"extracted_tech_keywords":["AI","machine learning","ML","computer vision","data science","RAG","object detection","AWS","Python","R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-online-companies-are-tapping-into-object-detection-to-improve-user-experience\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10017002,"title":"Top 10 IoT Products To Watch Out For In 2021","content":"From manufacturing to retail marketing, organisations, regardless of their sizes, have started using the Internet of Things (IoT) devices to transform their businesses. According to reports, the IoT technology reached 100 billion dollars in the market revenue for the first time in 2017, and estimates suggest that this figure will grow to around 1.6 trillion by 2025. Below is a compilation of the top ten Internet of Thing (IoT) products that will be trending in 2021. Note: The list is in alphabetical order 1| AWS Snowcone About: AWS Snowcone is a member of the AWS Snow Family of edge computing, edge storage, and data transfer devices. The weight of this IoT device is around 4.5 pounds, i.e. 2.1 kg and includes 2 CPUs and 8 terabytes of usable storage. Snowcone can run edge computing workloads that use AWS IoT Greengrass or Amazon EC2 instances, and store data securely. One can use Snowcone in backpacks on first responders, or for IoT, vehicular, and drone use cases. Users can also execute compute applications at the edge, ship the device with data to AWS for offline data transfer as well as transfer data online with AWS DataSync from edge locations. Currently, the device is available in the US East (N. Virginia), US West (Oregon), Europe (Ireland), and Europe (Frankfurt) regions. Price: The price varies depending upon the types of services, and ranges from USD 30-USD 2000. Know more here. 2| Amazon Echo Plus Voice Controller About: Amazon Echo Plus Voice Controller is a voice-controlled device with Alexa voice service. The smart, voice-controlled loudspeaker features premium speakers that deliver 360-degree omnidirectional audio. It has a built-in smart hub that can set up compatible Zigbee enabled smart home devices directly. With seven microphones, beam-forming technology, and noise cancellation, Echo Plus hears from any direction even in noisy environments. Price: The price range starts from INR 3,499. Know more here. 3| Bitdefender BOX IoT Security Solution About: Bitdefender launched the first IoT security solution in 2015 and created the IoT security category. Bitdefender BOX 2 is the latest next-gen security device trusted by experts and users to secure their homes by protecting all connected devices with a single security solution. Bitdefender BOX 2 plugs into any non-mesh or non-Google Wi-Fi router and protects an unlimited number of Wi-Fi and internet-connected devices using total security unlimited. It is a full-featured, cross-platform network security suite ensuring all Windows, Mac, iOS, Android, and IoT devices are fully protected. This includes real-time data protection, multi-layer malware and ransomware protection, social network protection, microphone monitor, webcam protection, anti-tracker, password manager, secured online banking and shopping and much more. Price: You can buy this from Amazon at around USD 332. Know more here. 4| Google Home Voice Controller About: Google Home is a wifi-connected, voice-controlled smart speaker. With Google Home Voice Controller, you can use your voice to enjoy Google Nest and Google Home speakers and display features, like media, alarms, lights & thermostats, volume control, and much more. The product’s voice-activated search functions let users to speak voice commands to interact with services through Google Assistant. Price: The lowest price of Google Home Smart Speaker is INR 6,990. Know more here. 5| Lenovo ThinkCentre M75n IoT About: The ThinkCentre M75n IoT can collect, store, as well as process data from IoT devices, such as sensors and cameras, both seamlessly and securely. This is an industrial PC that can help give businesses the edge needed for computation, and also delivers real-time tracking and insights. The ThinkCentre M75n IoT Desktop is equipped with AMD Athlon 3000 Series mobile processors with Radeon Graphics. Each ThinkCentre is developed with ThinkShield, a security-by-design approach offering multiple layers of protection to safeguard business-critical information and devices. Price: The price of ThinkCentre M75n Nano IoT starts from USD 369. Know more here. 6| Logitech Harmony Universal Remote About: Logitech Harmony Universal Remote is a smart and universal remote controller that can control up to 15 devices for seamless control of entertainment and smart systems. The latest Harmony Elite includes colour touchscreen, motion sensing, vibration feedback, an improved button layout, voice integration, dedicated connected home buttons, and a replaceable, rechargeable battery with 20% greater capacity. Price: The price starts from INR 15,000- 50,000. Know more here. 7| Particle Photon Wi-Fi with Headers About: Particle Photon Wi-Fi with Headers is a Wi-Fi IoT device for creating connected projects and products for the Internet of Things. It is a tiny Wi-Fi IoT device powered by a Cypress Wi-Fi chip alongside a powerful STM32 ARM Cortex M3 microcontroller. The headers make prototyping easy as each board can plug directly into standard breadboards and perfboards. Price: The price of this device is USD 19. Know more here 8| Sensimed Triggerfish About: Sensimed Triggerfish is a continuous ocular monitoring system that provides insights into the ocular volume changes throughout the day and night. The device includes a smart contact lens that captures spontaneous changes in the eye, providing physicians with valuable information that can help guide glaucoma treatment. Know more here. 9| Telematics Plug About: The Telematics Plug is an IoT sensor device that is designed for crash detection as well as analysing driving behaviour. It is developed by Bosch and one can choose between the various options, such as accident detection functionality & driving behaviour determination; accident detection functionality only and driving behaviour determination only. Price: The price ranges around INR 4,000. Know more here. 10| Vodafone Curve About: Vodafone Curve is a smart, multi-purpose GPS tracker device launched by Vodafone UK. The IoT device allows users to stay connected to the people by sending them alerts as well as updates to their mobile from a distance. The device requires a separate Vodafone Smart SIM subscription and users can personalise alerts to create their own zones. Price: The price ranges from £ 20. Know more here.","excerpt":"From manufacturing to retail marketing, organisations, regardless of their sizes, have started using the Internet of Things (IoT) devices to transform their businesses. According to reports, the IoT technology reached 100 billion dollars in the market revenue for the first time in 2017, and estimates suggest that this figure will grow to around 1.6 trillion […]","categories":["AI Trends"],"tags":["IoT","IoT devices"],"author_name":"Ambika Choudhury","publish_date":"2021-01-04T12:00:00","publication_year":"2021","word_count":973,"keywords":["Go","AWS","AI","IoT devices","ML","Git","RAG","GAN","Rust","edge computing","R","IoT"],"extracted_tech_keywords":["AI","ML","RAG","AWS","edge computing","R","Go","Rust","Git","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-10-iot-products-to-watch-out-for-in-2021\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10044586,"title":"What is DrQ Algorithm by NYU &#038; Facebook AI Research","content":"New York University (NYU) & Facebook Artificial Intelligence Research (FAIR) researchers, including Denis Yarats, Rob Fergus, Alessandro Lazaric, Lerrel Pinto, have introduced DrQ-v2, a model-free reinforcement learning (RL) algorithm for visual continuous control. DrQ-v2 is an upgraded version of DrQ, an off-policy actor-critic approach that uses data augmentation to learn directly from pixels. DrQ (Data regularised Q) algorithm was introduced in March 2021 by NYU & FAIR. At present, lots of methods exist to address the sample efficiency of RL algorithms that directly learn from pixels. The approaches can be classified into two groups: Model-based methods: Attempt to learn the system dynamics to acquire a compact latent representation of high-dimensional observations to later perform policy search.Model-free methods: Either learn the latent representation indirectly by optimising the RL objective or by employing auxiliary losses that provide additional supervision. The DrQ approach can be combined with them to improve performance. DrQ-v2’s implementation is released publicly to provide RL practitioners with a strong and computationally efficient baseline. What’s new in DrQ-v2 DrQ-v2 improves upon DrQ by making several algorithmic changes: Switching the base RL algorithm from Soft Actor Critic (SAC) to Deep Deterministic Policy Gradient (DDPG).Addition of bilinear interpolation to the random shift image augmentation.Introducing an exploration schedule.Selection of better hyper-parameters, including a larger capacity of the replay buffer. The research claims to introduce various improvements that yield state-of-the-art results on the DeepMind Control Suite. In particular, DrQ-v2 is able to solve complex humanoid locomotion tasks directly from pixel observations, previously unattained by model-free RL. “In addition, DrQ-v2 is conceptually simple, easy to implement, and provides a significantly better computational footprint compared to prior work, with the majority of tasks taking just 8 hours to train on a single GPU,” as per the paper. Source: NYU & FAIR Present-day state-of-the-art model-free methods have three major limitations: They are inadequate to solve the more challenging visual control problems such as quadruped and humanoid locomotion.They often require significant computational resources, i.e. lengthy training times using distributed multi-GPU infrastructure.It is often unclear how different design choices affect overall system performance. The humanoid control problem is one of the hardest control problems due to its large state and action spaces. Apart from NYU & FAIR, various other research has been initiated for the same. In collaboration with the University of Toronto and DeepMind, Google AI has introduced DreamerV2, the first RL agent to achieve human-level performance on the Atari benchmark. “Recently, a model-based method, DreamerV2, was also shown to solve visual continuous control problems, and it was first to solve the humanoid locomotion problem from pixels. However, while our model-free DrQ-v2 matches DreamerV2 in terms of sample efficiency and performance, it does so four times faster in terms of wall-clock time to train,” as per the paper. DreamerV2 from Google AI relies exclusively on general information from the images and accurately predicts future task rewards even when those rewards did not influence its representations. “Using a single GPU, DreamerV2 outperforms top model-free algorithms with the same compute and sample budget,” as per the blog. It builds upon the Recurrent State-Space Model (RSSM). “An unofficial implementation of DreamerV2 is available on Github and provides a productive starting point for future research projects. We see world models that leverage large offline datasets, long-term memory, hierarchical planning, and directed exploration as exciting avenues for future research,” as per the blog. Recent research in this field opens avenues for further futuristic applications of the technique. Moreover, RL algorithms that are good at working with pixels can be useful in applications such as Neuralink’s LINK, Mindpong and even make the RL training simulation more realistic and robust.","excerpt":"The research is supported in part by DARPA through the Machine Common Sense Program.","categories":["AI Features"],"tags":[],"author_name":"kumar Gandharv","publish_date":"2021-07-28T15:00:00","publication_year":"2021","word_count":605,"keywords":["Go","data augmentation","artificial intelligence","programming_languages:R","AI","Git","RAG","Aim","GitHub","R"],"extracted_tech_keywords":["AI","artificial intelligence","Aim","RAG","R","Go","Git","GitHub","data augmentation","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/what-is-drq-algorithm-by-nyu-facebook-ai-research\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":42460,"title":"Inside Tableau’s New AI Feature — Explain Data","content":"Image Source: Twitter With cutting-edge technology, almost all the organisations are engaged in the rat-race to be at the same pace with the emerging technologies. Since data visualisation is portraying a crucial role in analysing data for a few years now. Tableau has gained popularity as the leading data visualisation platform and with the new release, the software has come up with new and exciting features for the latest version. One of the most important advantages of using Tableau is that it helps an analyst to gain minute insights into the data. The beta version of Tableau 2019.3 has been released last month which includes some important features along with a new feature called “Explain Data.” Explain Data is a new AI-powered feature which helps the analysts to understand “why” behind the unexpected values in data by identifying causes and relationships between the data. The researchers behind this feature of Tableau used powerful Bayesian statistical methods to generate the explanations. The Explain Data works by evaluating more than a hundred explanations and choosing the most likely ones. Other New Features Introduced Besides Explain Data, Tableau 2019.3 includes two more important  interactive features as mentioned below Tableau Catalog: With the help of Tableau Catalog, analysts will be able to manage their data better than ever. The user will be able to get the entire view of the data which is being used and can automatically trace information such as usage metrics, user permissions, are among others. This feature will be available for Tableau Server and Tableau Online as a part of the Data Management Add-on. Project McKinley: With the help of this feature, a user can run large and critical Tableau server deployments with ease. Project McKinley allows a user to streamline the management process and provides functionalities such as improved scalability with external repository hosting and node optimisation, enhanced security with Amazon AWS key management integration, etc. This feature will be a separately licensed add-on and is not available for Tableau Online. Some of the other important features of Tableau 2019.3 include Extract Encryption At Rest: For additional security, a user can now encrypt the extracts at rest on Tableau Server. Embed Ask Data Into Your Content: By embedding Ask Data into an organisation’s portal, an analyst can enable more people in his\/her organisation who are to be asking questions of their data. Join Published Data Sources With Prep Builder: A user can now integrate published data sources into your prep flows and will be able to connect to them like any other data source by creating joins, pivoting data, etc. Web Editing For Tableau Public: With the help of this feature, a user can easily edit and save any of the Tableau Public visualisations directly in the browser. Highlight From Tableau 2019.1 Earlier this year, Tableau 2019.1 released Ask Data, a feature with self-service analytics which holds the capabilities of natural language processing and helps a user to ask data questions and get an instant visual response. This feature indeed helps a user regardless of skillset to gain a closer view of the data and produce more analytical insights. In Conclusion Many a time, an analyst ends up thinking why the data is showing the unexpected results and this is where Tableau as a tool assists to gain these pieces of information. Explain Data as a feature will help the data analyst to cut the time-consuming process of finding explanations and then validating them with data and provide you with a bird’s view.","excerpt":"Image Source: Twitter With cutting-edge technology, almost all the organisations are engaged in the rat-race to be at the same pace with the emerging technologies. Since data visualisation is portraying a crucial role in analysing data for a few years now. Tableau has gained popularity as the leading data visualisation platform and with the new […]","categories":["AI Features"],"tags":["Data Visualisation","extract data from twitter"],"author_name":"Ambika Choudhury","publish_date":"2019-07-15T12:01:41","publication_year":"2019","word_count":584,"keywords":["AWS","AI","cloud_platforms:AWS","Data Visualisation","ML","programming_languages:R","Scala","extract data from twitter","Ray","analytics","GAN","R"],"extracted_tech_keywords":["AI","ML","analytics","Ray","AWS","R","Scala","GAN","cloud_platforms:AWS","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/inside-tableaus-new-ai-feature-explain-data\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10018100,"title":"The Big Tech Overreach: Should India Be Wary Of Their Might","content":"“If you can silence a king, you are the king.”Naval Ravikant Last week, the megaphones of the internet were on a feeding frenzy, deplatforming Donald Trump in the wake of the US Capitol storming. For some reason, even Pinterest thought itself significant enough to block the President of the United States. That said, the whole “left vs right” theatrics exposed the fragility of our institutions. The big tech shot itself in the foot by trying to silence the head of one of the most powerful nations in the world. Facebook and Twitter especially, suffered a significant market wipeout. While Facebook saw $47.6 billion erased from its public valuation, Twitter’s market cap dropped by $3.5 billion. The whole drama was a wake up call to people around the world, especially, the heads of the states. Now, countries like Poland are reportedly mulling a law to fine Big Tech firms $2.2 million every time they censor lawful speech online. German Chancellor Angela Merkel called Twitter’s decision to ban Donald Trump, “problematic”. Hot Takes “A lot of people are going to be super unhappy with West Coast high tech as the de facto arbiter of free speech.”–Elon Musk“For me, Jack banning Trump violates one principle of liberal-democraticness (Jack was not elected by a vote), but the event was a huge win for another principle of liberal-democraticness: separation of powers.”–Vitalik Buterin, founder, Ethereum“The sudden fetishization of Terms of Service as a protection for free speech and due process is absurd. All that ToS stipulates is that you can say what the tech elite allows you to say.”–David Sacks, founder, Yammer (Source:WIRED) Making A Case For The Big Tech I do not celebrate or feel pride in our having to ban @realDonaldTrump from Twitter, or how we got here. After a clear warning we’d take this action, we made a decision with the best information we had based on threats to physical safety both on and off Twitter. Was this correct?— jack (@jack) January 14, 2021 Building an infrastructure that benefits the whole world is immensely challenging. There is no denying the fact that we all have one way or the other benefitted from the products of the big tech. It is highly likely that you are reading this using one of their products. But, that cannot be an excuse for the cancel culture. Especially, considering the regulations are highly partisan in nature. Last year, when COVID-19 hit the shores of the west, the tech companies asked their employees to work from home. Content moderators at Twitter and Facebook too were working from home. What followed was an increased distress for these human moderators. Their job is to flag malicious players on the platform but the heaviness of the content did take a toll on them. The big tech tried to replace these moderators with AI-based filtering. That didn’t go well either. There were quite a few accidental deletions of non-malicious content putting the whole notion of freedom of speech at risk. Today, the Big Tech firms are worth a combined $5 trillion(almost equal to Japan’s GDP). Companies like Google and Apple have cornered markets by offering services of the highest quality. But, the consolidation of the market kills healthy competition.The way social media platform Parler was taken down overnight, is a good case in point. Your app has to be hosted on servers owned by AWS. It can be only downloaded from app stores owned by Google and Apple. It can only run on operating systems owned by Google and Apple. Now, when these handful of companies decide to set the protocol, things get murkier. Sometimes the hammer comes down heavy on a select few. Like the one below: With no explanation other than \"repeatedly going against our community standards,\" @Facebook has blocked me from managing my page. Never have we received notice of violating community standards in the past and nowhere is the offending post identified. pic.twitter.com\/EdMyW9gufa— Ron Paul (@RonPaul) January 11, 2021 According to data released by the Federal Election Commission, employees at Alphabet, Amazon, Apple, Facebook, Microsoft, and Oracle have contributed $4,787,752 to Biden, 20 times the amount Trump received ($239,527). The scale is clearly tilted to one direction. And, to what extent these vested interests of the unelected few dictate the freedom of the elected officials is still up for debate. So, should the rest of the world be worried? Is there an alternative at all? Also Read: Big Tech Breakup: Google Blamed For Running A Monopoly, Case Filed What Can Or Should India Do? To begin with, there is not much to do, at least not in the near term. Take for example, India’s big digital disruption brought on by Reliance Jio. The telecommunication giant started as a cheapest internet service provider. It single handedly took the internet to millions of Indians, uprooted its competitors and made them follow Jio’s strategy. Today, Jio is into everything. From 5G to OTT service, from AR\/VR products to AI and more. To pull off a Google or a Facebook, India needs an impetus of Jio-esque proportions, which currently Jio won’t be able to do or even think of. Google and Facebook have invested heavily in Jio. Google spent ₹33,737 crores in Jio Platforms; a 7.73% stake in Jio– one of the largest investments by Google in an Indian company. Will other players or newer ones buckle up to take on these giants? Will the government back them up? There are many such dilemmas, which can quickly discourage many aspirants. And, even if some local player succeeds in building an OS, app store or a search engine, it is highly likely that given enough time they too would eventually go down the same dark alleys as those of the big tech companies. So, should the Indian regulators take control of the social media? Maybe, we could look up to China for a lesson or two! This is all much bigger than US politics.It is a matter of national security for India, Israel, and every other country to maintain a sovereign communications channel for their leaders to reach their people directly.The world cannot be ruled by American corporations. https:\/\/t.co\/2dDtcOICQ8— Balaji (@balajis) January 9, 2021 So, how can any community save itself from the eternal power grab between the media and the state? One potential solution is decentralisation. Developing and deploying technologies like blockchain is not straightforward. But, there have been few successes of late. When Sci-Hub– a portal to access scientific research papers behind paywall– was suspended, the website created a censorship-resistant alternative based on blockchain. …holy shit. This will go down as the beginning of the next, truly decentralized phase of the Web. https:\/\/t.co\/yEDnunzUwt— Eric\/ (@wheatpond) January 11, 2021 Whereas Paytm, India’s top payment app has launched an Android mini app store for developers and customers who are searching for alternatives other then Google play store– the go-to option for all Android users. “Paytm Mini App Store empowers our young Indian developers to leverage our reach and payments to build new innovative services,” said CEO, Vijay Shekhar Sharma. However, the blockchain landscape in India is yet to takeoff. According to India’s policy thinktank NITI Aayog, it is important to understand that the mechanism of decentralisation or peer-to-peer exchange. “It also has to be understood separately from government regulation – networks completely regulated by governments can be decentralised and feature peer-to-peer exchange and, totally centralised systems can also be unregulated and operate beyond the bounds of law. Decentralised networks do not necessarily mean they aren’t regulated.” Also Read: Top Indian Startups In The Field Of Blockchain Technology We can safely assume that big-tech’s dubious handling of free speech will set new precedents  and  will serve as a shot in the arm for developers across the globe working on decentralisation technologies. “We’re transitioning through a temporary bug in the Internet’s history before we knew how to build open social protocols. At first, the Internet transferred data. Then, it transferred scarcity (Bitcoin). Then, computation (Ethereum). Coming up – identity and social graphs,” observed Naval Ravikant, founder of AngelList.","excerpt":"“If you can silence a king, you are the king.” Naval Ravikant Last week, the megaphones of the internet were on a feeding frenzy, deplatforming Donald Trump in the wake of the US Capitol storming. For some reason, even Pinterest thought itself significant enough to block the President of the United States. That said, the […]","categories":["Deep Tech"],"tags":["AWS","Big Tech","censorship","Twitter (X)"],"author_name":"Ram Sagar","publish_date":"2021-01-14T17:00:09","publication_year":"2021","word_count":1342,"keywords":["Go","API","AWS","AI","Git","censorship","RAG","ViT","Big Tech","disruption","Twitter (X)","R","startup"],"extracted_tech_keywords":["AI","RAG","AWS","R","Go","Git","API","ViT","disruption","startup"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/big-tech-censor-india-wary-online-twitter\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10071981,"title":"Much-awaited Linux 5.19 is here, released from an M2 MacBook Air","content":"Finnish computer scientist Linus Torvalds came out with the latest version of Linux kernel from an ARM64 laptop on Monday. Good news to many, the announcement was made by Asahi on Twitter, where the developer says that he used Apple for the third time for Linux development. “On a personal note, the most interesting part here is that I did the release (and am writing this) on an arm64 laptop. It’s something I’ve been waiting for for a _loong_ time, and it’s finally reality, thanks to the Asahi team. We’ve had arm64 hardware around running Linux for a long time, but none of it has really been usable as a development platform until now..” Torvalds also says that he hasn’t used it for any real work, but to only perform test builds and boots. He shared about how he will take the Apple-powered Mac when he hits the road, ‘But I’m trying to make sure that the next time I travel, I can travel with this as a laptop and finally dogfooding the arm64 side too. Anyway, regardless of all that, this obviously means that the merge window (*) will open tomorrow. But please give this a good test run before you get all excited about a new development kernel’. He also prefers to call it 6.0 since he’s ‘starting to worry about getting confused by big numbers again’.","excerpt":"Linus Torvalds releases Linux 5.19 through a project called Asahi Linux, which was dedicated to supporting Apple’s Arm-based Silicon Macbooks.","categories":["AI News"],"tags":["Apple","Linux","Linux Kernel","OS"],"author_name":"Bhuvana Kamath","publish_date":"2022-08-01T19:58:38","publication_year":"2022","word_count":229,"keywords":["Go","Linux Kernel","programming_languages:R","AI","Apple","OS","programming_languages:Go","Linux","R"],"extracted_tech_keywords":["AI","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/much-awaited-linux-5-19-is-here-released-from-an-m2-macbook-air\/","complexity_score":2,"technical_depth":5,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":51400,"title":"OpenAI Benchmarks Reinforcement Learning To Avoid Model Overfitting","content":"OpenAI has benchmarked reinforcement learning by mitigating most of its problems using the procedural generational technique. RL has been a central methodology in the field of artificial intelligence. However, over the years, researchers have witnessed a few shortcomings with the approach. Developers often use a colossal amount of data to train and increase the efficiency of machine learning models. But this has resulted in overfitting of data in many cases, thereby, causing hindrance in the adoption of ML technologies. Overfitting To avoid overfitting, diversity in environments is an essential aspect — that’s why quite a few techniques have been used for keeping that in check. For one, Arcade Learning Environment (ALE) – a framework for the development and testing of AI agents – has been the standard for evaluating AI technologies for years. It facilitates hundreds of Atari 2600 games for ML models to train and assess in simulated environments. While diversity between different games in the ALE is one of its greatest strengths, the low emphasis on generalisation presents a significant drawback. Consequently, in order to address such generalisation challenges, there have been numerous attempts. Another such attempt is CoinRun – a training environment released by OpenAI in late last year, to provide generalisation diversity. While it helped in determining generalisation in reinforcement, it was limited to one environment. To facilitate a generalisation, various attempts have been made by different groups such as Dota and StarCraft using procedural generation to evaluate generalisation in reinforcement learning models. But it comes with lots of complexities; It is tough to iterate in extremely complicated, tedious to use more than one environment at the same time. While all of the above solutions were able to improve the overfitting issues, those approaches were not feasible in various use cases. To address such problems, OpenAI strived to render high diversity within environments, across environments, along with experimental convenience. Overfitting Solution Through Procedural Generalisation Benchmark AI agents become more robust and deliver higher accuracy when they are trained in generalised environments that offer ever-changing levels. Consequently, as a part of procedural generalisation, OpenAI included 16 unique environments designed to evaluate generalisation and sample efficiency in reinforcement learning. OpenAI said that such methodology is ideal for evaluating generalisation since unique training and testing data can be created in every environment. Besides, they also stressed that it is also well-suited to evaluate sample efficiency since these environments offer new and compelling challenges for ML models. Their Approach Towards Achieving The Benchmark While devising a plan, OpenAI focused on a wide range of aspects to benchmark the approach. They focused on high diversity, fast evaluation, tunable difficulty, and emphasis on visual recognition and motor control. To achieve high diversity, the environment generation logic was given maximal freedom, resulting in providing meaningful challenges to increase the ability of models. Besides, they have calibrated environment difficulty in a way that ML models can progress, resulting in the faster experimental pipeline. As a part of their difficulty, the environments support two well-calibrated modes: easy and hard. While the former caters to the developers with limited computational power, the latter is for users with high-performance computing systems. OpenAI further evaluated the generalisation while utilising ConiRun to understand how the agent struggled to generalise. Therefore, the firm evaluated how the size of the training data impacts generalisation. So, they generated training data from 100 to 100,00 levels and trained agents from 200M on it using Proximal Policy Optimisation (PPO) to measure the performance. Result The above evaluations were central in getting insights into the reasons behind overfitting. They witnessed that the agents overfit to small training data set in most of the environments. And with more data, the training performance improved. This is contrary to the trend in supervised learning, where the learning is inversely proportional to the size of training data sets. With generalisation across levels in the training set, even with large training set can enhance training performance. Outlook It will allow developers to make the most out of the plethora of available data and improve the accuracy of ML models. OpenAI’s benchmark has really shifted the landscape in reinforcement learning as it solves the overfitting problem with large datasets. The benchmark describes how important diverse environment distributions is while training and evaluation reinforcement learning agents. Although this advancement will allow developers to build robust and effective ML models, it will have to be applied in strenuous use cases for evaluating its full potential.","excerpt":"OpenAI has benchmarked reinforcement learning by mitigating most of its problems using the procedural generational technique. RL has been a central methodology in the field of artificial intelligence. However, over the years, researchers have witnessed a few shortcomings with the approach. Developers often use a colossal amount of data to train and increase the efficiency […]","categories":[],"tags":["OpenAI","overfitting","Reinforcement Learning"],"author_name":"Rohit Yadav","publish_date":"2019-12-09T17:00:00","publication_year":"2019","word_count":742,"keywords":["artificial intelligence","Reinforcement Learning","OpenAI","AI","machine learning","R","ML","programming_languages:R","emerging_tech:AI agents","AI agents","overfitting"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","OpenAI","R","AI agents","programming_languages:R","emerging_tech:AI agents"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/openai-benchmarks-reinforcement-learning-to-avoid-model-overfitting\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":40225,"title":"10 Best Firms In India For Data Scientists To Work For &#8211; 2019","content":"This year Analytics India Magazine is creating a list of the 10 best companies in India which can be the dream company for any data scientist in India to work for. We reached out to close to 25 companies with operations in India before shortlisting 10 best firms based on various parameters such as their work culture, exposure to different projects, internal growth trajectory, employee-friendly initiatives for data scientists such as re-skilling, programs to retain employees, programs to ensure reduced attrition rates, among others. (The companies are arranged in alphabetical order). AB InBev The culture at the “Growth Analytics Center” and AB InBev worldwide is truly unique as they are built on a strong foundation of ownership, informality, transparency and candour. They believe their people to be their greatest strength and work towards motivating them. The data scientists at ABI’s Growth Analytics Center work with some of the world’s top universities and are instrumental in developing in-house capabilities. The company strongly encourages innovation through exploration and collaboration which keeps data scientists highly motivated. As the company says, their people have a genuine entrepreneurial spirit for what they do. It’s with this spirit they take ownership of outcomes — taking results personally. Support for data scientists: Data Scientists are given personalised attention since day one by getting to know their interests as well as assessing their competencies. ABI provides year-round training programs focused on leadership, compelling communication and domain-specific training through in-house Analytics Academy initiative. The data scientists also have the opportunity to travel to their GHQ office in New York and to some of their zones, to work and learn with business partners quite frequently. Retention of data scientists: Being a flat organisation, data scientists get to interact with the senior-most leadership to present their work. These touchpoints are incredibly valuable for employees to see the larger picture as well as get recognised for their work. ABI also provides ample opportunities for international short or long term assignments in the markets that they support. BRIDGEi2i Analytics Solutions Team BRIDGEi2i With over 400 employees out of which 280 are data scientists, BRIDGEi2i is nurturing a people-driven, growth-oriented work culture based on skills & capability development, continuous learning, and ascension. Apart from core responsibilities, the employees contribute toward a wide range of initiatives and programs such as innovation, employer branding, employee engagement, and experiential marketing. BRIDGEi2i’s Rewards and recognition program celebrates and rewards high performers both on outcomes and values. Support for data scientists: At BRIDGEi2i, the new campus inductees are welcomed with an exciting program LeGo (Learn and Grow) to bridge the gap between college and professional lives. Experienced employees take charge of acquainting and moulding the new talent with necessary skill sets, not just in data science, but also soft skills! BRIDGEi2i also emphasises continuous growth and learning with SCaLa (Skills Capabilities Leadership Ascension) centres of excellence, which puts employees through a structured learning program with distinct streams tied to specific career growth paths. Retention of data scientists: With an attrition rate as per industry standards, BRIDGEi2i imparts one of the best learning and growth curves in comparison to the other players in the market. With specific, role-based learning on niche subjects, employees can upskill and focus on strengthening core fundamentals. Additionally, employees are encouraged to take up new roles and responsibilities that bring out the best in them while allowing them to grow into leadership roles! Cartesian Consulting Team Cartesian Consulting With over 200 data scientists in their team, Cartesian ensures that every single person at the company participates in the outcome of his\/her client’s business. The impact-oriented mindset ensures that teams not only develop data science skills but are constantly engaging with business to understand the nuances and solve real business problems. Teams are exposed to clients very early on to inculcate a sense of creating business impact. Support for data scientists: Apart from the emphasis on training, including external courses, there are other aspects that help data scientists grow at Cartesian. Focus on ground-up solutions encourages data scientists to do independent thinking in terms of problem formulation and methodology. Other aspects of cross-domain learning, allows employees to have experience in multiple domains such as retail, BFSI. This encourages them to abstract across domains and come up with robust solutions. Retention of data scientist: At Cartesian, the attrition at senior levels is extremely low compared to junior folks who mostly leave because of higher studies. Cartesian is able to retain people because of three main reasons — they get exposed to business very early on, employees are encouraged to work across domains, use cases, and geographies, and third is that people are constantly exposed to new techniques, languages and approaches through Cartesian’s Innovations Lab and AI Lab. The employees can see their contributions changing Cartesian and that is a big motivator for a low attrition rate. Envestnet Yodlee Team Evestnet | Yodlee The culture at the data analytics CoE at Envestnet Yodlee basically runs on the theme, “No idea is a bad idea”. The analytics team has an opportunity to work cross-functionally with every team within the organisation — right from the conceptualisation stage, to the development process, and projects going live. The company prides on the fact that they have fewer silos, and that all the employees have a sense of belonging and understanding of cross-functional roles and responsibilities. The company strongly encourages open communication among the employees regardless of the designation, a fact which enthuses creativity and confidence among youngsters. Support for data scientists: Envestnet Yodlee has a strong focus on learning, development and personal growth. They operate as a research lab with commercial deliverables. The company equips the employees with the right training and opportunities to learn within and outside the organisation based on their developmental needs. They do not insist on classroom training, instead have learning champions within the team who understand the training needs of individuals and plan tailor-made learning opportunities. Retention of data scientists: The company does not believe that keeping a check on attrition alone, is a feasible long-term solution. Instead, they focus on developing individuals in the team from the time they join. The trust that the company puts in the employees with confidential data, and the flexibility to derive insights and innovative models from the data, is something that wins the employees over. GOJEK Sidu Ponnappa, India MD, GOJEK With about 50 data scientists spread across geographies, GOJEK is built on the philosophy of speed, innovation and social impact. To encourage employee-friendliness, the company offers complete freedom to run experiments, challenge conventions and provide the necessary room for the team to feel empowered. Their zero-hierarchy model allows the employees to take quick decisions. GOJEK provides freedom to data scientists to explore and choose the project they like and work with experienced professionals to create unique solutions. Support for data scientists: GOJEK’s culture imbibes 360-degree assessment and one-on-one feedback, which eliminates hierarchy and gives anybody with a good idea an opportunity to be heard. It rewards independence, entrepreneurship and initiative efforts, encouraging data scientists to work seamlessly with product and business teams to ensure optimum output. Data science professionals get a chance to work with senior leaders across geographies, which is a major boost for employees. Retention of data scientists: The attrition rate of GOJEK is next to negligible. They don’t stress much on keeping a check on attrition rates or retaining data scientists, instead, they offer challenges that people can solve. The best any employee can ask from an organisation is to keep them challenged, offer great culture and remunerate them handsomely. GOJEK ticks these critical boxes, making it an incredibly hard organisation to join and equally difficult to leave. Oracle Team Oracle Cloud Solution Hub The unique differentiator of the Oracle Cloud Solution Hubs in Bangalore is that it has the spirit of a start-up while being part of one of the largest and most successful technology companies ever. Hence, a data scientist can have the best of both worlds, enjoying all the sophisticated tools and technologies that Oracle offers. Data scientists get a chance to work in a well-defined solution engineering environment under the guidance of experienced and best of breed data scientists. Support for data scientists: Oracle Cloud Solution Hubs encourages an innovative culture. They organise various hackathons and innovative contests throughout the year, along with inculcating a culture of continuous learning and investing in training and growth of their team. Oracle also encourages active participation in various events and conferences, along with the opportunity to present and publish papers in numerous academic and research journals. There is a clear growth path defined for employees along with facilitating mentorship programs to encourage people to interact with other teams. Check on attrition and retain data scientists: To check on attrition and retain data scientists, the company stresses on hiring the right fit. They look for employees with a strong work ethic, the ability to get along with others and a willingness to learn apart from their core technical skillset. They also focus on familiarizing new employees through orientation and induction to make them comfortable. Oracle also believes in listening to their employees and encouraging them to share continuous feedback, which is used as a two-way communication to improve transparency. Publicis Sapient Team Publicis Sapient The company has more than 100 analytics people, whereas around 35 AI\/ML people in their data science team. Publicis Sapient has the analytics, engineering and science teams under one community, which gives the team enormous visibility and opportunity to work on the end-to-end customer journey- starting from customer identification, experience and engagement, all the way till delivering a service or a product efficiently. Support for data scientists: The company provides the opportunity to work on both Applied-AI in addition to exploring the new approaches in ML\/AI in academia. The team is encouraged to have external connects via conferences, speaker opportunities, book publications etc. Since Data Science is still in nascent stages at Publicis Sapient, this allows data scientists to continue learning and apply the knowledge for Digital Business Transformation across multiple industries. Retention of data scientists: The company has a very low attrition rate, thanks to the interesting initiatives to keep a check on attrition and retain data scientists. Apart from many interesting data science-friendly initiatives, Sapient focuses on hiring based on learn-ability and attitude and keeping them engaged with interesting work. “Embracing the Future” and “Learning Mindset” is part of Core Values at Publicis Sapient. Tredence Inc. Team Tredence Tredence boasts of around 300 members in its data science team, with a work culture that is egalitarian, non-hierarchical and respects the individuality of team members. The company provides a conducive environment where all kinds of learning are encouraged. Tredence encourages its employees to take up courses to upskill themselves and loop them into ongoing projects through several knowledge sessions. With a young workforce, where the average age is 27, Tredence has a collegial environment that allows for extensive growth. Support for data scientists: Tredence encourages a healthy and fit workplace where a lot of engagement activities are around sports and fitness. They have a tie-up with Cult where one trainer from Cult visits the office every alternate day, whereas healthy dinner from Cult is provided every day. The company also encourages learning habit in employees. The ‘U Learn V Pay’ is a unique learning offering at Tredence that encourages its employees to take any course through online platforms on any topic of their choice. Once the employee successfully completes the course, the course fee is reimbursed by Tredence. Apart from this, employees get to participate in conferences, industry events and given a chance to continue higher education. Retention of data scientists: The attrition rate at Tredence is at par with the industry average and they are constantly striving to engage continuously with their top performers. The biggest reason they attract and retain people is a combination of learning and working with a high-quality peer group plus an opportunity to solve complex business problems with mastering analytics skills in an environment that makes the entire journey rewarding. Wipro Ltd. Team Wipro With more than 1000 data scientists, Wipro has a diverse pool of data scientists across various domains. The company focuses on regular thought-leadership meetups at various locations for AI and data science community at multiple locations across the globe. Consulting-driven approach to most customer requirements gives the team an opportunity to a fresh storyboarding towards each problem. The company also provides an opportunity for transforming customers into an intelligent enterprise and lots of other opportunities for employees to cross skill. Support for data scientists: There are several platforms and initiatives which assist and enable a data scientist to grow within Wipro and keep them engaged throughout their career with Wipro. Some of many such programs are TrendNxt and FutureSkills, which contains extensive courses on all the learning areas, SDS (School of decision sciences) which is a flagship analytics full-stack program, Big Data academy, crowdsourcing programs such as TopCoder (Wipro-acquired entity) and TopGear (Wipro-owned entity), Datathon AI (an annual Wipro AI event), among several others. Retention of data scientists: There are several awards and recognition for data scientists which creates a positive and encouraging work environment at Wirpo. Hiring and retaining the right data scientists is one of the top priorities for Wipro and they achieve it by providing training, constant upskilling and industry-fit compensation. Wipro prides in constant development of skill sets and recognising the efforts top performers by awarding them in the form of quarterly rewards to annual awards. ZS Associates Team ZS ZS’s Capability and Expertise Center in India house more than 4,000 ZSers across three offices in Pune, New Delhi and Bengaluru, many of whom are data scientists. The firm leverages its flagship “ZS Data Science Challenge” initiative to hire the best talent from top engineering schools. This effort has helped grow the team by 50 percent over the course of the past year. ZS culture celebrates innovation and curiosity. The firm seeks out creative thinkers and problem solvers from a variety of backgrounds, blending their unique talents to develop and deliver products that drive better business results. Support for data scientists: ZS offers a sponsorship program to support data scientists pursue higher education (Masters) at ivy league institutes across the globe. There is a strong internal data science community that encourages sharing, learning and updating knowledge, partnering with external vendors. ZS also provides various internal platforms for the team to strive for grassroot innovations. Other opportunities include real-time challenges such as hackathons, participation in conferences and summits, external competitions such as Kaggle, potential to publish whitepapers and blogs in external publications, among others. Retention of data scientists: Transparency is key when defining and communicating roles and responsibilities so that people know what to expect. Team members are encouraged to evolve in their current roles by experimenting with new tools, technologies and algorithms. Data scientists are also offered the opportunity to work at other ZS offices as well as client sites, which provide great exposure.","excerpt":"This year Analytics India Magazine is creating a list of the 10 best companies in India which can be the dream company for any data scientist in India to work for. We reached out to close to 25 companies with operations in India before shortlisting 10 best firms based on various parameters such as their […]","categories":["AI Features"],"tags":["best book to learn digital marketing","Data scientists India","what is data oriented person"],"author_name":"Srishti Deoras","publish_date":"2019-06-05T10:53:36","publication_year":"2019","word_count":2506,"keywords":["what is data oriented person","data science","Go","TPU","AI","ML","Data scientists India","Scala","RAG","analytics","Rust","best book to learn digital marketing","R"],"extracted_tech_keywords":["AI","ML","data science","analytics","RAG","TPU","R","Go","Rust","Scala"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/10-best-firms-in-india-for-data-scientists-to-work-for-2019\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":24011,"title":"Real-Time SQL Is What’s Making CrateDB Such A Hit In The IoT Market","content":"In this era of constantly evolving technology, database management systems (DBMS) are no longer limited to their traditional functions of merely managing data. With the advent of disruptive technologies such as internet of things, artificial intelligence and machine learning, the complexity and proliferation of data is on a sharp rise. Data is omnipresent and is now a part of the innovation bandwagon. As a result, terms like big data and cloud computing are now mainstream jargons in the information technology (IT) domain. From accurate sensors to automation, data now needs to be handled in real time. This is where SQL, the programming language used to maintain DBMS, has witnessed a change with CrateDB — a real-time DBMS which powers IoT data. This article will explore the intricacies of CrateDB and how it makes real-time database projects scalable. The Momentum In DBMS For IoT Over the years, the development in DBMS has evolved into other functions such as data transformations, data security and other useful features. Also, maintaining these storage systems have become much easier with cloud technology assisting the complex processes in a significant manner. The advances in hardware such as increased storage and memory capacity have been filling the gap of processing meaningful data on a larger scale. This has consequently paved way to explore data-related issues in real time. Since IoT involves data exchange in real time, DBMS such as CrateDB is helping manage them efficiently. Started in 2013 as an SQL database offering, CrateDB has garnered attention in serving areas of IoT and machine data. Instant Data: Sensors And Their Role IoT is a interconnected chain of physical devices and which also contains electronic components like sensors to record data. This data is generated continuously and needs to be attended quickly for proper functioning of the entire IoT system. Therefore, this poses newer challenges such as new data volume and query complexity, among others. CrateDB aims to take on these challenges and provide a feasible IoT solution through SQL. Designing CrateDB CrateDB incorporates a mix of functionalities from SQL, NoSQL and Container technology. Crate.io, the company behind CrateDB discuss in their white paper about the product’s architecture and working. They label the architecture as “distributed, shared-nothing, container-native”. The Create.io team explains: “CrateDB operates in a shared-nothing architecture as a cluster of identically configured servers (nodes) who coordinate seamlessly with each other. Execution of write and query operations are automatically distributed across the nodes in the cluster. Increasing or decreasing database capacity is a simple matter of adding or removing nodes. We worked hard on the “simple” part by automating the sharding, replication (for fault tolerance), and rebalancing of data as the cluster changes size. CrateDB was born in the container era and allows you to scale and administer it easily via container orchestration platforms like Docker or Kubernetes in a microservices environment.” CrateDB architecture (Image courtesy : Crate.io) Accessibility With SQL To make it easier for database developers to work with CrateDB, SQL is set as the primary language for data access. This means, the database offers compatibility across a host of standard SQL features such as joins, aggregations, indexes, Binary Large Objects (BLOBs), and user-defined functions, among others to work with SQL tools. In addition, it supports open machine data access interfaces which are prominent in IoT — such as Apache Kafka, Apache Spark, MQTT, Telegraf and Grafana. Indexing For Performance CrateDB relies on open source NoSQL technologies such as Lucene, Elasticsearch and Netty for indexing and storage. This is to avoid the effects of outage in systems due to hardware failure while keeping optimum performance in mind — in fact, query execution is almost parallelised to achieve optimal output. Flexibility And Real-Time Performance With the inner architecture integrated with SQL and NoSQL, almost any data structure can be used with CrateDB. For instance, if there are multiple sensors giving out different data, CrateDB records all these data in one table instead of storing it in a traditional SQL way — separate tables for separate data types. This way, the queries are processed faster and the real-time scalability is improved. Furthermore, the database uses in-memory columnar indexing, a sophisticated technique to handle queries with minimal memory (cache). CrateDB proved to be 33 times faster compared to traditional SQL databases when handling complex time series and text queries. This is why it has found its use on IoT applications since data is processed and stored almost instantly. The following image shows where it makes a difference in executing IoT workloads with respect to other databases. CrateDB comparison with SQL and NoSQL databases (Image courtesy : Crate.io) Conclusion The Crate.io team aims to further their database development towards IoT by exploring other related fields such as analytics, and gaining insights to develop simpler IoT applications. This will definitely boost manufacturing industries who plan to rely more on data rather than investing in automation and latest machinery. Even IT companies look forward to implementing this database for their IoT needs.","excerpt":"In this era of constantly evolving technology, database management systems (DBMS) are no longer limited to their traditional functions of merely managing data. With the advent of disruptive technologies such as internet of things, artificial intelligence and machine learning, the complexity and proliferation of data is on a sharp rise. Data is omnipresent and is […]","categories":["IT Services"],"tags":[],"author_name":"Abhishek Sharma","publish_date":"2018-04-26T06:24:49","publication_year":"2018","word_count":831,"keywords":["machine learning","artificial intelligence","AI","cloud computing","ML","docker","RAG","Aim","analytics","kubernetes"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","analytics","Aim","RAG","cloud computing","kubernetes","docker"],"url":"https:\/\/analyticsindiamag.com\/it-services\/real-time-sql-is-whats-making-cratedb-such-a-hit-in-the-iot-market\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10163724,"title":"Indian Music Labels T-Series, Saregama Seek to Join Copyright Lawsuit Against OpenAI","content":"A group of India’s leading Bollywood music labels, including T-Series, Saregama, and Sony, has sought to join a copyright lawsuit against OpenAI in New Delhi over concerns regarding the unauthorised use of sound recordings for AI training, according to a recent report by Reuters. The Indian Music Industry (IMI) group, along with T-Series and Saregama, approached a New Delhi court on Thursday, arguing that OpenAI’s alleged use of their copyrighted sound recordings breaches intellectual property rights. The labels assert that the matter is significant for the music industry in India and globally. OpenAI, backed by Microsoft, has faced increasing legal scrutiny worldwide. The company maintains that it adheres to fair-use principles in utilising publicly available data for AI model training. Neither OpenAI nor the music labels responded to requests for comment. The music labels aim to be part of a lawsuit filed last year by Indian news agency ANI, which accused OpenAI’s ChatGPT of using its content without authorisation for AI training. Since then, book publishers and media organisations, some linked to billionaires Mukesh Ambani and Gautam Adani, have joined the legal action against OpenAI in the New Delhi court. Bollywood and Hindi pop music represent a significant industry in India. T-Series, one of the largest music labels in the country, releases approximately 2,000 sound recordings or songs annually. Saregama, which has been in operation for over a century, has a catalogue featuring renowned Indian singers such as Mohammed Rafi and Lata Mangeshkar. According to its website, the IMI group also represents international labels like Sony Music and Warner Music. The report added that Indian music companies are concerned AI systems, including those developed by OpenAI, can extract lyrics, compositions, and sound recordings from the internet without proper authorisation. The legal move follows similar concerns raised globally. In November, Germany’s GEMA, representing composers, lyricists, and publishers, sued OpenAI over ChatGPT’s alleged unlicensed reproduction of song lyrics, arguing that the system had been trained using copyrighted content. OpenAI has opposed the ANI lawsuit, contending that Indian courts lack jurisdiction as the company is based in the United States with servers located abroad. The next hearing in the case is scheduled for February 21. The outcome is expected to influence the future use of copyrighted material in AI model training in India. OpenAI CEO Sam Altman recently visited India, meeting with the country’s information technology minister to discuss India’s plans for low-cost AI development.","excerpt":"OpenAI CEO Sam Altman recently visited India, meeting with the country’s information technology minister to discuss India’s plans for low-cost AI development.","categories":["AI News"],"tags":["OpenAI"],"author_name":"Siddharth Jindal","publish_date":"2025-02-14T15:05:19","publication_year":"2025","word_count":402,"keywords":["ChatGPT","AWS","OpenAI","AI","GPT","Aim","GAN","R","llm_models:ChatGPT","llm_models:GPT"],"extracted_tech_keywords":["AI","ChatGPT","OpenAI","Aim","AWS","R","GPT","GAN","llm_models:GPT","llm_models:ChatGPT"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/indian-music-labels-t-series-saregama-seek-to-join-copyright-lawsuit-against-openai\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":31299,"title":"5 Python Unit Test Frameworks To Learn In 2019","content":"Testing applications have become a necessary skill set to become a proficient developer today. The Python community supports testing and the Python standard library has well-built tools to support testing. In the Python environment, there are abundant testing tools to handle the complex testing needs. Unittest Inspired by the JUnit framework and having similar characteristics, this unit testing framework supports test automation, sharing of setup and shutdown codes and acts independently of the tests from the reporting environment. The UNittest comprises of several Object-Oriented Concepts such as: Test Fixture: A test fixture is a representation of the preparation which is needed to execute one or more tests, and collaborates cleanup actions Test Suite: It is a collection of test cases, test suites which are used as aggregate tests which will be executed together Test Case: It is an individual unit of testing that tries to find a specific response to appropriate input sets. A test case can be used to create new test cases which provide base classes for the unit test Test Runner: The basic operation of a test runner is to execute tests and provide the output to the user. The procedure may use a graphical interface, a textual interface, or a special attribute to specify the results of executed tests Here’s the sample code: import unittestclass TSM(unittest.TestCase):   def test_upper(auto):       auto.assertEqual(‘abc’.upper(), ‘ABC’)   def test_isupper(auto):       auto.assertTrue(‘ABC’.isupper())       auto.assertFalse(‘Abc’.isupper())   def test_split(auto):       s = ‘hello world’       auto.assertEqual(s.split(), [‘This’,’Is’, ‘Python’])       # check that given condition and s.split fails when the separator is not a string       with auto.assertRaises(TypeError):           s.split(2)if __name__ == ‘__main__’:   unittest.main() Nose Testing An updated form of Unitest framework, the nose testing framework automatically collects the tests from the subclasses of the unittest. Users need to script simple test functions to execute them. The Nose provides various helpful functions for implementing timed tests, testing for exceptions, and other uses. Sample Code def dimension_overlap(ranges):   “””Returns common overlap among a set of [minimum, maximum] ranges”””   minimum = dimention[0][0]   maximum = dimention[0][1]   for (min_1, max_1) in dimention:       minimum = max(minimum, min_1)      maximum = min(maximum, max_1)       if minimum >= maximum:           return None       else:           return (minimum, maximum) Pytest Preferred by most of the Python developers, Pytest framework is largely used for Unit testing. The package offers various sub packages collectively known as Pytest framework. The refined and Pythonic expressions which were introduced for testing development have made it preferable for test suites to be scripted in a high-toned firm pattern. It is widely used for small projects and has the best test automation technique. One of the best features of Pytest is that it offers complete information about the failures in the testing case, which will help the developers to address the issue in a relevant and fast way. Sample Code: pip install pytest (installing pytest) import pytest (installing pytest into work environment)      from Purse import purse, InsufficientAmount (user-defined package)def test_default_initial_amount():   purse = Purse()   assert purse.balance == 0def test_setting_initial_amount():purse = Purse(100)   assert Purse.balance == 100def test_purse_add_cash():   purse = Purse(10)   purse.add_cash(90)   assert purse.balance == 100def test_purse_spend_cash():   purse =Purse(20)   purse.spend_cash(10)   assert purse.balance == 10   def test_purse_spend_cash_raises_exception_on_insufficient_amount():   purse= Purse()   with pytest.raises(insufficient amount):       purse.spend_cash(100) Robot Framework Being an operating system and application independent, Robot framework is a universal test automation framework for acceptance testing and acceptance driven development(ATDD). Giving access to the easy usage of test data and best suited for the keyword drove testing approach, the robot framework testing capabilities can be drawn out of test libraries either by using Python or Java. The framework gives users the scope to create new high-level keywords from current ones that are in usage by using the same syntax that is used for building test cases. Sample Code *** Settings ***Library  SeleniumLibrary  (set the working library)*** Test Cases ***The user can search for colleges   [Tags]    search_colleges   Open browser    http:\/\/XYZ..com\/   Chrome   Select From List By Value   xpath:\/\/select[@name=’most_preferred’]  India   Select From List by Value   xpath:\/\/select[@name=’least_preferred’]    India   Click Button    css:input[type=’submit’]   @{colleges}=  Get WebElements    css:table[class=’table’]>tbody tr   Should Not Be Empty     ${colleges}   Close All Browsers Zope Testing The Zope debugger allows the user to point out the exact step where there is a problem in the running process. The Zope facilitates logging options which permit the user to issue warnings and error messages. The testing provides information through various sources and allows to combine channels to fetch information about the debugging. The Control Panel The control panel gives various views that can help user debug zope, specifically in the area of accomplishments, the debugging information link present on the control panel gives two views, debugging info and profiling. Product Refresh Settings Zope provides a refreshing view of all control panel devices. This allows the user to reload the products modules as they are being changed or updated. Debug Mode Once you set the debug mode, it reduces the Zope performance and gives effects such as Tracebacks will be displayed on the browser when errors raiseExternal methods and DTML file objects are analyzed to see if they have been updated every time they have been called for an update and they are reloaded The Python Debugger Integrated with Python debugger Zope can shut down the server and produce a request through a command line. It develops an infrastructure to raise new objects and debug them immediately. Sample Code from zope.interface.verify import verifyClass, verifyObjectfrom uspkg.app import SampleApp (uspkg=user defined package)from uspkg.interfaces import ISampleAppdef test_app_create():   # Assure we can create instances of `Sample_App`   app_1 = SampleApp_1()   assert app_1 is not Nonedef test_app_1_class_iface():   # Assure the class implements the declared interface   assert verifyClass(ISample_App, Sample_App)def test_app_1_instance_iface():   # Assure instances of the class provide the declared interface   assert verifyObject(ISample_App, Sample_App())","excerpt":"Testing applications have become a necessary skill set to become a proficient developer today. The Python community supports testing and the Python standard library has well-built tools to support testing. In the Python environment, there are abundant testing tools to handle the complex testing needs. Unittest Inspired by the JUnit framework and having similar characteristics, […]","categories":[],"tags":["Python","python advantages","python frameworks","Python Libraries","testing"],"author_name":"Bharat Adibhatla","publish_date":"2018-12-10T11:56:00","publication_year":"2018","word_count":926,"keywords":["python advantages","TPU","programming_languages:R","AI","Python Libraries","R","ML","programming_languages:Java","automation","Python","programming_languages:Python","testing","python frameworks","Java"],"extracted_tech_keywords":["AI","ML","TPU","Python","R","Java","automation","programming_languages:Python","programming_languages:R","programming_languages:Java"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/5-python-unit-test-frameworks-to-learn-in-2019\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10084910,"title":"Introducing Text-to-Video Generator, Tune-A-Video","content":"Since the birth of text-to-image DALL-E by OpenAI, the AI world has been working towards similar models, for example, Midjourney, and Imagen, to name a few. Soon came text-to-video models like Transframer, NUWA Infinity, CogVideo, etc. Even text-to-voice models like VALL-E were recently unveiled by Microsoft. Last month, researchers from Show Lab, National University of Singapore came up with a text-to-video generator called Tune-A-Video (TTV) to address the issue of One-Shot Video Generation, where only a single text-video pair is provided for training an open-domain text-to-video generator. With customised Sparse-Causal Attention, Tune-A-Video expands spatial self-attention to the spatiotemporal domain using pretrained text-to-image (TTI) diffusion models. Tune-To-Video Check the unofficial implementation of Tune-A-Video here. In one training sample, the projection matrices in the attention block are modified to include the relevant motion information. Tune-A-Video can create temporally coherent videos for various applications, including changing the subject or background, modifying attributes, and transferring styles. It was discovered that TTI models could produce images that match verb terms well and that expanding TTI models to generate different images at once demonstrates unexpectedly strong content consistency. Fine-Tuning: TTI models are expanded to TTV models using TTI model weights that have already been pretrained. The text-video pair is then subjected to one-shot tuning in order to create a one-shot TTV model. Inference: A modified text prompt is used to generate new videos. After receiving a video and text pair as input, it modifies the projection matrices in attention blocks. Read the full paper here.","excerpt":"With customised Sparse-Causal Attention, Tune-A-Video expands spatial self-attention to the spatiotemporal domain using pretrained text-to-image diffusion models.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","AI Tool","Data Science","Deep Learning","Machine Learning"],"author_name":"Shritama Saha","publish_date":"2023-01-12T17:53:12","publication_year":"2023","word_count":249,"keywords":["DALL-E","OpenAI","AI","programming_languages:R","Machine Learning","diffusion models","Deep Learning","Data Science","R","AI (Artificial Intelligence)","AI Tool"],"extracted_tech_keywords":["AI","OpenAI","R","DALL-E","diffusion models","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/introducing-text-to-video-generator-tune-a-video\/","complexity_score":2,"technical_depth":6,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":43809,"title":"Complete Guide To Cracking MachineHack’s ‘Predict The News Category Hackathon&#8217;","content":"Natural Language Processing (NLP) is one of the most explored and successful domains in machine learning. It is important because complex communication is one of the best signs of intelligence as we are trying to make machines communicate with humans effortlessly. In this article, we will do a hands-on NLP with Python to solve MachineHack’s Predict The News Category hackathon. Predict The News Category Hackathon MachineHack has launched its second Natural Language Processing challenge for its large Data Science and ML audience. The hackathon is about predicting the category or section of news from its content.The dataset consists of news pieces collected from a number of different sources along with the category or section of the news piece in which it was featured. Given below is the description of the dataset. Size of training set: 7,628 records Size of test set: 2,748 records FEATURES: STORY: A part of the main content of the article to be published as a piece of news. SECTION: The genre\/category the STORY falls in. There are four distinct sections where each story may fall in to. The Sections are labelled as follows : Politics: 0 Technology: 1 Entertainment: 2 Business: 3 Getting The Datasets Go to MachineHack, Sign Up as a user and click on the Predict The News Category Hackathon. Start the hackathon and find the dataset in the Attachment section. Click here to register for the hackathon Without further ado, let’s crack the Hackathon! Solving The Hackathon Let’s break the solution into 6 parts as given below for better understanding. Exploratory Data Analysis: A Simple analysis of Data Data cleaning Data preprocessing: Count Vectors and TF-IDF Vectors Training the classifier Predicting for the test set Submitting your solution at MachineHack Exploratory Data Analysis: A Simple analysis of Data Let’s start off with the usual drill and import all the necessary modules for our project. #Importing the libraries import pandas as pd import nltk from nltk.corpus import stopwords import string #Download the following modules once nltk.download('stopwords') nltk.download('wordnet') Let’s do a simple analysis of the data in hand. #Importing the training set train_data = pd.read_excel(\"Datasets\/Data_Train.xlsx\") #Printing the top 5 rows print(train_data.head(5)) #Printing the dataset info print(train_data.info()) #Printing the shape of the dataset print(train_data.shape)] Out:(7628, 2) #Printing the group by description of each category train_data.groupby(\"SECTION\").describe() Data Cleaning #Removing duplicates to avoid overfitting train_data.drop_duplicates(inplace = True) #A punctuations string for reference (added other valid characters from the dataset) all_punctuations = string.punctuation + '‘’,:”][],' #Method to remove punctuation marks from the data def punc_remover(raw_text): no_punct = \"\".join([i for i in raw_text if i not in all_punctuations]) return no_punct #Method to remove stopwords from the data def stopword_remover(no_punc_text): words = no_punc_text.split() no_stp_words = \" \".join([i for i in words if i not in stopwords.words('english')]) return no_stp_words #Method to lemmatize the words in the data lemmer = nltk.stem.WordNetLemmatizer() def lem(words): return \" \".join([lemmer.lemmatize(word,'v') for word in words.split()]) #Method to perform a complete cleaning def text_cleaner(raw): cleaned_text = stopword_remover(punc_remover(raw)) return lem(cleaned_text) #Testing the cleaner method text_cleaner(\"Hi!, this is a sample text to test the text cleaner method. Removes *@!#special characters%$^* and stopwords. And lemmatizes, go, going - run, ran, running\") Out: 'Hi sample text test text cleaner method Removes special character stopwords And lemmatizes go go run run run' #Applying the cleaner method to the entire data train_data['CLEAN_STORY'] = train_data['STORY'].apply(text_cleaner) #Checking the new dataset print(train_data.values) Data Preprocessing: Count Vectors and TF-IDF Vectors Creating Count vectors #Importing sklearn’s Countvectorizer from sklearn.feature_extraction.text import CountVectorizer #Creating a bag-of-words dictionary of words from the data bow_dictionary = CountVectorizer().fit(train_data['CLEAN_STORY']) #Total number of words in the bow_dictionary len(bow_dictionary.vocabulary_) Out : 35189 #Using the bow_dictionary to create count vectors for the cleaned data. bow = bow_dictionary.transform(train_data['CLEAN_STORY']) #Printing the shape of the bag of words model print(bow.shape) Out: (7551, 35189) Creating TF-IDF Vectors #Importing TfidfTransformer from sklearn from sklearn.feature_extraction.text import TfidfTransformer #Fitting the bag of words data to the TF-IDF transformer tfidf_transformer = TfidfTransformer().fit(bow) #Transforming the bag of words model to TF-IDF vectors storytfidf = tfidf_transformer.transform(bow) Training The Classifier #Creating a Multinomial Naive Bayes Classifier from sklearn.naive_bayes import MultinomialNB #Fitting the training data to the classifier classifier = MultinomialNB().fit(storytfidf, train_data['SECTION']) Predicting For The Test Set #Importing and cleaning the test data test_data = pd.read_excel(\"Datasets\/Data_Test.xlsx\") test_data['CLEAN_STORY'] = test_data['STORY'].apply(text_cleaner) #Printing the cleaned data print(test_data.values) Creating A Pipeline To Pre-Process The Data & Initialise The Classifier #Importing the Pipeline module from sklearn from sklearn.pipeline import Pipeline #Initializing the pipeline with necessary transformations and the required classifier pipe = Pipeline([ ('bow', CountVectorizer()), ('tfidf', TfidfTransformer()), ('classifier', MultinomialNB())]) #Fitting the training data to the pipeline pipe.fit(train_data['CLEAN_STORY'], train_data['SECTION']) #Predicting the SECTION test_preds_mnb = pipe.predict(test_data['CLEAN_STORY']) #Writing the predictions to an excel sheet pd.DataFrame(test_preds_mnb, columns = ['SECTION']).to_excel(\"Predictions\/predictions.xlsx\") Submitting Your Solution At MachineHack Finally, head to MachineHack, and submit your excel fine at the Submission Deck of the hackathon. 1. Click on the Assignment 2. Browse to your file and select it. Note: Also provide a comment for your submission. Check your score on the Hackathon Leaderboard. The hackathon leaderboard will be updated within 2 minutes. And that’s it. You have successfully found a solution. Now what’s left to do is tweaking the model for performance. We will leave that part up to you. Tune the model, improve your accuracy, top the leaderboard and win exciting prizes. Happy Coding!","excerpt":"Natural Language Processing (NLP) is one of the most explored and successful domains in machine learning. It is important because complex communication is one of the best signs of intelligence as we are trying to make machines communicate with humans effortlessly. In this article, we will do a hands-on NLP with Python to solve MachineHack’s […]","categories":["Deep Tech"],"tags":["data science hackathon","Hackathon","Machinehack","Natural Language Processing"],"author_name":"Amal Nair","publish_date":"2019-08-06T10:00:02","publication_year":"2019","word_count":870,"keywords":["data science","Go","machine learning","AI","Machinehack","Natural Language Processing","ML","Hackathon","data science hackathon","NLP","Python","NLTK","R","Pandas"],"extracted_tech_keywords":["AI","machine learning","ML","NLP","data science","NLTK","Pandas","Python","R","Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/guide-to-cracking-machinehacks-predict-the-news-category-hackathon\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":38218,"title":"10 Most Popular Machine Learning &#038; Data Science Packages On Github","content":"GitHub community decided to dig deeper into machine learning and pulled data on contributions from Jan-Dec 2018. The contributions include pushing code, opening an issue or pull request, commenting on an issue and reviewing a pull request. The Octoverse report used data from the dependency graph for the most imported packages which include all public repositories and any private repositories which have opted into the dependency graph. The information on this article has been cited from the original documentation and the sources are also cited. In this article, we list down the 10 most popular machine learning and data science packages on GitHub. 1| Numpy NumPy is the fundamental package for scientific computing with Python. It contains a powerful N-dimensional array object, sophisticated (broadcasting) functions, tools for integrating C\/C++ and Fortran code and is useful linear algebra, Fourier transform, and random number capabilities. NumPy can also be used as an efficient multi-dimensional container of generic data. Here, the arbitrary datatypes can also be defined that allows NumPy to seamlessly as well as speedily integrate with a wide variety of databases. Click here to read more. 2| Scipy SciPy is open-source software for mathematics, science, and engineering which includes modules for statistics, optimisation, integration, linear algebra, Fourier transforms, signal and image processing, ODE solvers, and more. It basically depends on NumPy which provides convenient and fast N-dimensional array manipulation. SciPy is built to work with NumPy arrays and provides many user-friendly and efficient numerical routines such as routines for numerical integration and optimization. Click here to read more. 3| Pandas Pandas is a Python package which provides fast, flexible, and expressive data structures designed to make working with “relational” or “labeled” data both easy and intuitive. It aims to be the fundamental high-level building block for doing practical, real-world data analysis in Python. Additionally, it has the broader goal of becoming the most powerful and flexible open source data analysis\/manipulation tool available in any language. Click here to read more. 4| Matplotlib Matplotlib is a Python 2D plotting library which produces publication quality figures in a variety of hardcopy formats and interactive environments across platforms. Matplotlib can be used in Python scripts, the Python and IPythonshells, the Jupyter notebook, web application servers, and four graphical user interface toolkits. You can easily generate plots, histograms, power spectra, bar charts, error charts, scatterplots, etc., with just a few lines of code in this library. Click here to read more. 5| Scikit-learn Scikit-learn is an open source machine learning library for Python which is built on top of SciPy and distributed under the 3-Clause BSD license. Scikit-learn 0.20 is the last version to support Python2.7. Scikit-learn 0.21 and later will require Python 3.5 or newer. Scikit-learn also uses CBLAS, the C interface to the Basic Linear Algebra Subprograms library. If you already have a working installation of NumPy and SciPy, the easiest way to install Scikit-learn is using pip or conda i.e. pip install -U scikit-learn Or conda install scikit-learn Click here to read more. 6| Six Six is a Python 2 and 3 compatibility library which provides utility functions for smoothing over the differences between the Python versions with the goal of writing Python code that is compatible on both Python versions. Six is basically a utility package and it is intended to support codebases which work on both Python 2 and 3 without any modification and it can be downloaded on PyPI. Click here to read more. 7| TensorFlow TensorFlow is an open source software library for numerical computation using data flow graphs. The graph nodes represent mathematical operations, while the graph edges represent the multidimensional data arrays (tensors) that flow between them. This flexible architecture enables you to deploy computation to one or more CPUs or GPUs in a desktop, server, or mobile device without rewriting code. TensorFlow also includes TensorBoard which is a data visualization toolkit. Click here to read more. 8| Requests Requests is an Apache2 Licensed HTTP library, written in Python. It is designed to be used by humans to interact with the language. It allows you to send organic, grass-fed HTTP\/1.1 requests, without the need for manual labor and is one of the most downloaded Python packages of all time, pulling in over 11,000,000 downloads every month. Click here to read more. 9| Python-dateutil The dateutil package provides powerful extensions to the standard datetime module, available in Python. The features of Python-dateutil includes computing of relative deltas between two given date or datetime objects, computing of dates based on very flexible recurrence rules, using a superset of iCalender specification, generic parsing of dates, etc. Click here to read more. 10| Pytz The pytz library allows accurate and cross platform timezone calculations using Python 2.4 or higher versions and provides access to the Olson timezone database. It also solves the issue of ambiguous times at the end of daylight savings, which you can read more about in the Python Library Reference. Click here to read more.","excerpt":"GitHub community decided to dig deeper into machine learning and pulled data on contributions from Jan-Dec 2018. The contributions include pushing code, opening an issue or pull request, commenting on an issue and reviewing a pull request. The Octoverse report used data from the dependency graph for the most imported packages which include all public […]","categories":["AI Trends"],"tags":["fourier transform machine learning","GitHub","ml libraries"],"author_name":"Ambika Choudhury","publish_date":"2019-04-25T05:52:20","publication_year":"2019","word_count":828,"keywords":["data science","scikit-learn","machine learning","AI","fourier transform machine learning","TensorFlow","ML","ml libraries","Ray","Aim","Jupyter","GitHub","Pandas"],"extracted_tech_keywords":["AI","machine learning","ML","data science","Aim","Ray","TensorFlow","scikit-learn","Jupyter","Pandas"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/10-most-popular-machine-learning-data-science-packages-on-github\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10050440,"title":"What Is Border Gateway Protocol: Reason Behind Facebook’s Outage","content":"Top social media apps — Facebook, Instagram, and WhatsApp — suffered one of the longest global outages last night. For about six hours, all the three apps, which command a total of over 3.5 billion monthly core users, were not working. This outage grazed Facebook founder Mark Zuckerberg’s fortune by $6 billion and pushed him down a few positions in the world’s richest list. In an official statement issued by Facebook, the cause behind the unprecedented outage was a configuration change in the backbone routers that coordinate network traffic between the data centres. This had a cascading effect and brought all the Facebook services to a halt. In layman terms, everything that Facebook runs disappeared for a period of time. We’re aware that some people are having trouble accessing our apps and products. We’re working to get things back to normal as quickly as possible, and we apologize for any inconvenience.— Meta (@Meta) October 4, 2021 What Went Wrong The company talks about the Border Gateway Protocol (BGP) in a detailed blog by website security firm Cloudflare. It is a mechanism to exchange routing information among autonomous systems. The list of possible routes is constantly updated by the Internet routers. If BGP stops working, these routers will not know what to do, bringing the Internet to a halt. So while DNS (domain name system) is the address system of the location of each website or the IP address, BGP is the roadmap that helps find the most efficient route to get to that address. Further, an individual network with a unified internal routing policy is called an Autonomous System (AS), each of which has an Autonomous System Number or ASN. An AS can originate prefixes — control a group of IP addresses — and transit prefixes — tell how to reach specific groups of IP addresses. As per Cloudflare, Facebook stopped announcing the routes to their DNS prefixes, which means that its DNS servers were unavailable. A BGP UPDATE message gives information about any changes made to a prefix advertisement or entirely withdraws the prefix. Cloudflare said that for Facebook, this chart is largely unchanged as the social media giant does not make a lot of changes to its network minute by minute. However, before the global outage, Cloudflare observed a lot of routing changes from Facebook. This led to the routes being withdrawn, and the DNS went offline. With these changes, Facebook and associated sites were effectively disconnected from the Internet. Credit: Cloudflare Facebook found itself in a deep slump because it was not able to fix the issue quickly. The company’s own internal systems run from the same place, so it was difficult for the staff to resolve the problem; they were restricted from accessing their own communications and were unable to access their office due to the security pass system being affected during the outage. As per reports, Facebook sent a technical team to its servers in California to manually reset the servers where the problem originated. Credit: Cloudflare Interestingly, when Facebook and related apps were down, people started looking for alternatives. A lot of DNS queries to Twitter, Signal and other social media apps increased. https:\/\/twitter.com\/Twitter\/status\/1445078208190291973?s=20","excerpt":"For about six hours, Facebook, Instagram, and WhatsApp, which command a total of over 3.5 billion monthly core users, were not working.","categories":["AI Trends"],"tags":["Facebook outage"],"author_name":"Shraddha Goled","publish_date":"2021-10-06T10:00:00","publication_year":"2021","word_count":530,"keywords":["AWS","AI","cloud_platforms:AWS","programming_languages:R","Facebook outage","R"],"extracted_tech_keywords":["AI","AWS","R","cloud_platforms:AWS","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/what-is-border-gateway-protocol-reason-behind-facebooks-outage\/","complexity_score":2,"technical_depth":5,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10008977,"title":"Meet The NLP Startup Making Internet Available To Every Indian In Local Language","content":"Reports suggest that around 90% of Indian users prefer using local languages for online engagement. The same report mentions that the web has only 0.1% content available in Indian languages. There is a prominent gap between these two stats, which is being filled by the Bengaluru-based Reverie Language Technologies. Taking pride in building language equality on the internet since 2009, the startup has enabled many industries such as BFSI, education, media, entertainment, e-commerce, and the Indian government to impact the lives of at least 500 million through their technologies. This gap especially became prominent during the pandemic when there was a sudden surge of users using online services due to the lockdown. This led to business and government organisations to bring Indian languages online to reach broader users. The pandemic has only further inspired Reverie to work harder to get the internet to every Indian and support the Digital India campaign. Analytics India Magazine caught up with Arvind Pani, CEO and co-founder at Reverie Language Technologies to understand the underlying technology that focuses on AI, and how it is providing Indic localisation in 22 languages. Co-founded by Vivekananda Pani, Reverie is the only full-stack Indian language product with intuitive technologies, making localisation easy. AI And NLP For Language Localisation As Pani explains, Reverie’s NLP suite enables devices and applications to interact with users like a human, in their local language. Using the NLP suite, the device understands and transcribes speech-to-text in 11 languages and Indian English. The NLP suite is further programmed to suit the Indian context. It has a robust Indic foundation to comprehend the users’ intent and respond accordingly. Further expanding on the technology, Pani said that Reverie uses AI-enabled language solutions to translate, create and manage localisation efficiently. Their AI-powered translation tools ensure easy and error-free translation of content, thereby saving time and effort. Reverie’s AI-enabled platforms — Prabandhak and Anuvadak, combined with Neural Machine Translation and Transliteration are great examples of how we use AI for language localisation. Prabandhak is an automated, end-to-end translation platform that helps in translating content in 22 Indian languages. It uses a combination of machine translation (that uses AI technology) and manual translators. Whereas, Anuvadak as a platform automates repetitive tasks involved in translating, hosting, and scaling content for localised websites. “Managing localised content, website, and tracking the progress is made seamless via a simple user dashboard. This platform can decrease the time taken to go live by 40% and save content management costs by about 60%,” shared Pani. “Reverie’s voice suite uses NLP, NLU, text-to-speech and speech-to-text technologies, to power chatbots, voice bots, IVRs, virtual assistants to enable local-language communication with users,” he added. Use Cases Talking about the challenges that most users faced during the pandemic, Pani said that during this time, it became difficult to communicate with citizens using online platforms as most of them were dominated by the English language. It made a few startups from Bengaluru come together to start Project StepOne, a telemedicine initiative, across states to give out critical information related to COVID and provide medical assistance. In this project, Reverie contributed by deploying its NLP suite for IVR. “Usually, voiceovers recorded in studios take a long time to process, but our technology took no time to translate information in different local languages. The feedback we got was overwhelming, and the voiceover was so natural that the user does not realise that they are talking to simulation,” said Pani. Citing some more use cases, Pani shared that the government leveraged Anuvadak’s technology to publish the MyGov COVID-19 page in 10 different Indian languages. In another use case with Practo, he shared that 40% of Practo users prefer to use the app in local languages. To enable this, about 2.5 million SMS notifications are sent in local languages, annually, using Reverie’s solutions. Besides, Reverie has helped Bank of Baroda to send about 2.5 million SMS notifications in Kannada and Hindi across four cities. The NLP technology by Reverie is also being used by eNAM (National Agriculture Market), a Government of India’s unified online market for trading agricultural produce. This service helps farmers check the production cost and selling price of various agricultural products allowing the farmer to buy or sell online at best prices. “The application was available in the English language, which could not be used by the farmers. We translated the content and services in local languages in five local languages,” shared Pani. Wrapping Up Last year, Reliance Industrial Investments & Holdings (RIIHL), a wholly-owned subsidiary of Reliance Industries acquired a majority stake in the startup for ₹190 crores, with a further investment of ₹77 crores by March 2021. While Reverie has been operating independently, Pani shared how Reliance has helped the startup in the long run. As Reverie is made on three critical pillars — access to affordable smartphones, enabling digital infrastructure and accessibility to information in Indian languages, Reliance Jio has enabled India with the first two pillars. Further, Reverie’s mission to bridge the digital language divide in India is in sync with Reliance’s mission to provide access to every Indian. “At Reverie Language Technologies, we aim to impact at least 500 million Indian lives through deployments of our technologies across governments and enterprises. Reliance’s strategic investment has set the stage for a better tomorrow,” said Pani on a concluding note. Check out our video podcast with Co-founder & CTO of Reverie Language Technologies- Vivekanand Pani","excerpt":"Reports suggest that around 90% of Indian users prefer using local languages for online engagement. The same report mentions that the web has only 0.1% content available in Indian languages. There is a prominent gap between these two stats, which is being filled by the Bengaluru-based Reverie Language Technologies. Taking pride in building language equality […]","categories":["AI Startups"],"tags":["best jobs in india","how does ai create jobs"],"author_name":"Srishti Deoras","publish_date":"2020-10-06T17:00:21","publication_year":"2020","word_count":905,"keywords":["how does ai create jobs","best jobs in india","Go","AI","chatbots","ML","virtual assistants","RAG","NLP","Aim","analytics","R"],"extracted_tech_keywords":["AI","ML","NLP","analytics","Aim","RAG","chatbots","virtual assistants","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-startups\/meet-the-nlp-startup-making-internet-available-to-every-indian-in-local-language\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":69074,"title":"7 Types Of Tableau Charts &#038; Graphs To Make Your Data Visually Interactive","content":"Graphs and charts are the most expressive ways when it comes to showing enormous amounts of raw data in a simple manner. Tableau is known to be one of the most popular and trending data visualisations software among developers. It creates fast visualisations of data in the form of worksheets and dashboards. Below here, we listed the top 7 types of Tableau charts and graphs one must use to make the data more visually appealing. (The list is in alphabetical order) 1| Bubble Chart In the bubble chart, the data is displayed with the help of circles. This type of chart helps in depicting the relationship between three or more measures. Bubble charts can visualise large volumes of data at once by varying the colour and size of the circles. Steps to create a Packed Bubble Chart- Below here are the steps to create packed Bubble Chart for Sales and Profit for various products: First, you need to connect to the Sample – Superstore data source Then drag the Category dimension to Columns Drag the Sales measure to Rows Click Show Me on the toolbar to select the packed bubbles chart type Drag Region to Detail on the Marks card to include more bubbles in the view Drag Profit to Color on the Marks card Drag Region to Label on the Marks card to clarify what each bubble represents. Know more here. 2| Box Plots Box and Whisker Plots or simply Box Plots are used to display the distribution of data for values such as median, outliers, first quartile, third quartile, among others. Box Plot is mainly useful for comparing the variations in the data. Steps to create a Box Plot: Below here are the steps to create Box Plot that shows discounts by region and customer segment: First, you need to connect to the given Sample – Superstore data source. Then drag the Segment dimension to Columns. After dragging Segment dimension, you need to drag the Discount measure to Rows. Next, the Region dimension is needed to be dragged to Columns, and drop it to the right of the Segment. Now click Show Me in the toolbar to select the box plot chart type. Drag Region from the Marks card back to Columns, to the right of Segment. To disaggregate data, select Analysis > Aggregate Measures. Know more here. 3| Gantt Charts Gantt Charts was invented in 1910 by Henry Gantt, and it displays a series of horizontal lines that track completed work overtime. This type of chart is useful in displaying the average delivery time for a range of products. Steps to create a Gantt Chart: Here are the steps to create a Gantt chart that shows discounts by region and customer segment: First, you need to connect to the given Sample – Superstore data source. Next, drag the Order Date dimension to Columns. On the Columns shelf, click the Year drop-down arrow, and then select Week Number. You need to drag the Sub-Category and Ship Mode dimensions to the Rows shelf while dropping Ship Mode to the right of Sub-Category. Next, in the toolbar menu, click Analysis > Create Calculated Field In the calculation dialogue box, you can name your calculated field OrderUntilShip. Clear any by default content in the Formula box. Know more here. 4| Heat Maps Heat Maps or Density maps can be used to identify locations without greater or fewer numbers of data points as well as reveal the patterns or relative concentrations that might otherwise be hidden due to an overlapping mark on a map. In Tableau, a heatmap can be created by grouping overlaying marks and colour-coding them based on the number of marks in the group. Steps to create a Heat Map: Below are the steps to create a Heat Map: To create a heat map, your data source should have latitude and longitude coordinates or location names. You can choose “Density” from the mark type drop-down, and Tableau will compute a density surface on your view. Know more here. 5| Pareto Charts A Pareto chart is a kind of chart in Tableau that contains bars as well as line graphs. In this chart, individual values are represented in descending order by bars, while the ascending cumulative total is represented by the line. Steps to create a Pareto Chart: Below are the steps to create a Pareto Chart: First, create a bar chart that displays Sales by Sub-Category in descending order. Next, add a line chart that also displays Sales by Sub-Category. Then add a table calculation to the line chart to display sales by Sub-Category as a Running Total as well as a Percent of Total. Know more here. 6| Scatter Plots In Tableau, a scatter plot can be constituted by putting at least one measure on the Columns shelf as well as one measure on the Rows shelf. Scatter Plots can be used to visualise relationships between numerical variables. Steps to create a Scatter plot: Here are the steps to create a Scatter plot to compare sales to profit: Open the Sample – Superstore data source. Drag the Profit measure to Columns. Drag the Sales measure to Rows. Drag the Category dimension to Color on the Marks card. Drag the Region dimension to Detail on the Marks card. Know more here. 7| TreeMaps Treemaps are simple data visualisation graphs that have the ability to provide insight in a visually attractive format. This type of graph can be used to display data in nested rectangles. Steps to create a TreeMap: Here are the steps to create a TreeMap that shows aggregated sales totals across a range of product categories: First, you need to connect to the given Sample – Superstore data source. Next, you need to drag the Sub-Category dimension to Columns. After dragging the Sub-Category dimension, you need to drag the Sales measure to Rows. Next, click Show Me on the toolbar to select the treemap chart type. Then, drag the Ship Mode dimension to Color on the Marks card. Lastly, drag the Profit measure to Color on the Marks card. Know more here. (Image Source: Here)","excerpt":"Graphs and charts are the most expressive ways when it comes to showing enormous amounts of raw data in a simple manner. Tableau is known to be one of the most popular and trending data visualisations software among developers. It creates fast visualisations of data in the form of worksheets and dashboards. Below here, we […]","categories":["AI Trends"],"tags":["Data Visualisation","data visualization tools","Tableau","types of analytics"],"author_name":"Ambika Choudhury","publish_date":"2020-07-07T10:00:00","publication_year":"2020","word_count":1018,"keywords":["Go","Tableau","programming_languages:R","AI","Data Visualisation","programming_languages:Go","data visualization tools","Git","RAG","GAN","R","types of analytics"],"extracted_tech_keywords":["AI","RAG","R","Go","Git","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/7-types-of-tableau-charts-graphs-to-make-your-data-visually-interactive\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":39580,"title":"Google To Revamp Search Results Page With Icons For Mobile Users","content":"Google recently announced that they would be revamping the way search results are shown on mobile. The redesign features a better design for search results where the source of the result is shown in a clearer fashion. This comes after the company is redefining itself to enable a more easily understandable search experience. This is also a move to provide information more clearly to the user in a shorter amount of time. The branding of a website ranking as a result on Google will be more clearly presented, allowing for a faster way to access the content the user was looking for. Site owners are also given a new feature to change the icon for organic listings. This changes the look and feel of the Search results page, with a higher emphasis on the branding of the site posting the content. There is also a renewed focus on making ads more visible. The ‘Ad’ label for promoted content has lost its green colour and instead transitioned to a bolded black label to show that the result is advertised. The redesign offers users an easier way to find out where their content is coming from, a positive move by Google that is sure to strengthen the recognition of dependable information. Moreover, Google also mentioned that this is a forward-thinking focused move, with the redesign making space for new features. Jamie Leach Senior Interaction Designer, Search, Google, said in a statement: “…from buying movie tickets to playing podcasts this new design allows us to add more action buttons and helpful previews to search results cards.” This offers users multiple different options to choose from on the search page itself, driving engagement while providing a better web experience for the user. The redesign is coming soon for mobile users, with the company promising even more “fresh ways” that search will be redefined. The rollout for desktop is not mentioned yet, but will presumably be at a later date.","excerpt":"Google recently announced that they would be revamping the way search results are shown on mobile. The redesign features a better design for search results where the source of the result is shown in a clearer fashion. This comes after the company is redefining itself to enable a more easily understandable search experience. This is […]","categories":["AI News"],"tags":["Google"],"author_name":"Anirudh VK","publish_date":"2019-05-23T07:52:54","publication_year":"2019","word_count":325,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","Google","GAN","R"],"extracted_tech_keywords":["AI","R","Go","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/google-to-revamp-search-results-page-with-icons-for-mobile-users\/","complexity_score":2,"technical_depth":6,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":57135,"title":"As Photoshop Turns 30, Adobe Introduces New AI Improvements","content":"Adobe Photoshop has become synonymous with photo editing since it was introduced in 1990. As the company turns 30, it has unveiled a slew of features and software updates which integrates AI for performance improvements, superior outputs and better workflows. This follows its heavy usage of Adobe Sensei for movie visual effects and personalised experiences, and the launch of the software on iPad. Some of the key AI-powered features that marks its 30th anniversary includes: (New) Object Selection & (Improved) Select Subject Tools Available under the ‘Magic Selections’ tool icon, the Object Selection tool uses Adobe Sensei to automatically select images – be it isolated, multiple or individual parts of objects. Concurrently, the Select Subject tool has been given a major remodelling and made smarter. First debuted on the desktop in 2018, this command,which is powered by ML, is now faster at analyzing imagery and has been updated to better identify the most prominent subject in an image. It has also sharpened the results when isolating the primary object from the background – even when the said object is ambiguous in the image – and achieves better edges. This is the result of multiple different Sensei AI and ML algorithms working in tandem – each optimized for different parts of the selection process. What is more, this tool is faster on Mac and can be run on iPad as well. Improvements To Lens Blur This takes better advantage of depth maps from a user’s phone, and sets a focal point on what to focus on and what to blur out. Also, this has been sped up with Graphics Processing Unit (GPU) improvements. Content-Aware Fill Enhancements The company believes that in an industry where both speed and quality are paramount, Content-Aware Fill is invaluable. With the latest updates, Adobe has introduced three novels ways to identify where in an image a user wants Content-Aware Fill to look for source pixels as fill content. This includes auto, rectangular and custom. The first smartly selects source pixels by analyzing pixels near the selection. The second chooses nearby pixels, and the third gives users full control to identify exactly which pixels to fill from. Adobe Photoshop Camera First announced in November 2019, Adobe Photoshop Camera uses AI to redefine mobile photo editing. Particularly aimed at followers of photo sharing apps like Instagram and Snapchat, the company had launched an app that seeks to leverage automatic image retouching. How does it work? Using the Sensei AI platform, the app identifies in real-time what the smartphone is aimed at, then applies Photoshop effects directly at the moment of capture. This includes a wide array of photos, including portraits, landscapes, ‘selfies’ as well as food shots. This cuts the need for any post-processing or editing, while always preserving the original shot. What is more, this AI recognition feature works offline as well. Looking ahead Alongside these improvements and enhancements packaged into its flagship photo editing software Photoshop, Adobe has introduced a new service whereby it added a section in its home screen for Cloud Documents. This keeps users’ Photoshop files in-sync across devices. It had also developed a machine learning tool that could spot images that had been edited using its own software. These developments are in congruence with the company’s response to shifting behaviors in the way people share content and use creative tools. According to Adobe CTO Abhay Parasnis, this necessitates the integration of a lot more AI into its processes as it increasingly looks to “democratize” its products. This includes its application to enable users to edit photos hands-free by verbally telling such apps what they want, and eliminating mundane or repetitive tasks, and more.","excerpt":"Adobe Photoshop has become synonymous with photo editing since it was introduced in 1990. As the company turns 30, it has unveiled a slew of features and software updates which integrates AI for performance improvements, superior outputs and better workflows. This follows its heavy usage of Adobe Sensei for movie visual effects and personalised experiences, […]","categories":["AI Features"],"tags":["Adobe","Adobe Sensei","photoshop"],"author_name":"Anu Thomas","publish_date":"2020-02-22T14:00:00","publication_year":"2020","word_count":611,"keywords":["Go","machine learning","TPU","AI","Adobe","ML","Adobe Sensei","RAG","Ray","Aim","GRU","photoshop","R"],"extracted_tech_keywords":["AI","machine learning","ML","Aim","Ray","RAG","TPU","R","Go","GRU"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/as-photoshop-turns-30-adobe-introduces-new-ai-improvements\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10021880,"title":"How This Techie’s Love For Storytelling Led Her To Start An AI-Driven Content Generation Startup","content":"Sharmin Ali was working with a technology company and dealing with Fortune 500 companies in the USA when she realised that even large corporations are poor at communication. With a strong interest in writing and reading, she understood how we communicate is crucial to making marketing decisions. “Here is where I stepped in with Instoried,” said Ali. Founded in 2019, Bengaluru-based Instoried helps large companies make their marketing and communications content more human with a data-driven, scalable and repeatable approach using AI. Its deep-tech platform analyses and predicts the emotional quotient, tonality and relevance of written content across formats like blogs, articles, social media posts. In the last two years, the startup has helped numerous enterprises and individuals create emotionally engaging content and improving ROI. Behind Instoried With its augmented writing platform, Instoried helps predict the success of the content before publishing it, hence improving customer interest and engagement. The AI-driven tool enables optimising emotions of the written content with smart recommendations in real-time. “At Instoried, we have built deep learning-based natural language processing systems to help optimise content. We are aiming to build a highly inclusive workplace for people of all backgrounds and genders who can help us grow,” Ali told Analytics India Magazine. She is responsible for driving strategy, leadership and sales as the founder and CEO at Instoried. Sutanshu Raj, who takes care of the startup’s tech aspect, later joined her as the CTO. As Ali likes to call it, Instoried is at the cusp of Sharmin’s love for storytelling and a background in technology. Tech Behind Instoried Instoried is an AI-driven deep-tech startup that works on making real-time analysis and suggestions to enhance the written content. It does so by using a data-driven approach in its proprietary AI technology to make these predictions. The AI algorithms are trained on an accumulated data set from 5 million data points to give smart recommendations. Instoried’s APIs can be easily integrated into any platform that brands may already be using. Apart from its proprietary AI technology, it uses neuromarketing principles to offer smart recommendations to dial up (or down) a particular emotion in the content. “As the name suggests, neuromarketing is marketing to the brain,” said Ali. It works on the fact that emotions drive consumer decision making more often than logic. “We help marketers create content that is more human and filled with empathy, that it becomes easier to resonate with their audiences,” she further added. Explaining how the AI solution works, Ali said that as one creates content, the tool dynamically analyses it and displays the text’s tone and emotion. On clicking on a particular tone or emotion, the tool highlights all the words, excluding that particular tone or emotion. Upon hovering the cursor over a particular word, it gives multiple recommendations that are both semantically and contextually relevant, thereby enhancing the content’s empathy. Companies have used it for increasing the opening rate of email, creating engaging captions for social media, plagiarism check, analysing the mood of the customer, and more. Ali said the tech is completely proprietary. “We haven’t used any of the open-source APIs available out there. We have used 7 million data points in-house for building our proprietary dataset. The entire code has been built in-house, and I’m happy to tell you that our analysis and recommendations are completely state-of-the-art and have the highest levels of accuracy and effectiveness,” she added. So far, Instoried’s customers have witnessed up to a 3X increase in incoming leads per post and a 2X rise in ROI in less than six months. Customers have used it across industries such as e-commerce, FMCG, media and more. Growth Story “Our customers love us,” shared an elated Ali. Having started shipping in April last year, Instoried has witnessed a hockey stick curve in the growth cycle despite the lockdown. “In less than ten months of selling, we have already received such amazing customer feedback,” she said. Ali said they have an exciting blueprint for 2021. “We are launching our own keyboard for both Android and IOS devices that can be used on any platform. Imagine chatting with your loved one and being able to gauge the temperature of your chat, and also getting smart recommendations in real-time,” said Ali. She believes this is the future of chat, and there are endless possibilities that Instoried is eager to tap.","excerpt":"Sharmin Ali was working with a technology company and dealing with Fortune 500 companies in the USA when she realised that even large corporations are poor at communication. With a strong interest in writing and reading, she understood how we communicate is crucial to making marketing decisions. “Here is where I stepped in with Instoried,” […]","categories":["AI Startups"],"tags":["AI Startups"],"author_name":"Srishti Deoras","publish_date":"2021-03-11T12:00:00","publication_year":"2021","word_count":727,"keywords":["Go","API","AI","data-driven","Scala","Aim","deep learning","analytics","R","AI Startups","startup"],"extracted_tech_keywords":["AI","deep learning","analytics","Aim","R","Go","Scala","API","data-driven","startup"],"url":"https:\/\/analyticsindiamag.com\/ai-startups\/how-this-techies-love-for-storytelling-led-her-to-start-an-ai-driven-content-generation-startup\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":37105,"title":"How IIT Kharagpur Is Strengthening Its AI Research With The Help Of Its Illustrious Alumni Base","content":"IIT Kharagpur may soon expand its AI research capabilities with the help of its illustrious alumni network to benefit the best from the emerging technologies. The institute, which was founded in the year 1951, is one of the oldest and prestigious IITs in the country and has produced several meritorious alumni over the years. Sundar Pichai, the current Google CEO is one among the top name among the list of noted personalities. Speaking to national daily about the importance of having an advanced AI centre, Indian-American entrepreneur, Vinod Gupta, an alumnus of IIT KGP said, “Some of the universities (in other countries) are investing 100, 500 million dollars in artificial intelligence. IIT can do very well in that with their research, students. They can do a lot more. The future is into AI.” Pointing out that such investment would require a large sum and by approaching its billionaire alumni like Pichai, Gupta said that the institute can have the benefit of cutting-edge technologies like AI, “They need to scale it up much more than what they are doing now…. My suggestion was: they should go to Sundar Pichai because he is CEO of Google, he is a billionaire, he loves IIT, he is our alumni and they are the ones who are doing a lot of cutting-edge work in artificial intelligence.” AI taking precedence The institute has been at the forefront to encompass AI and another emerging tech into its academics and already has a dedicated Centre of Excellence (CoE) for AI being established within the campus. In addition to this, last year, it signed a Memorandum of Understanding (MoU) with the Telangana government to set up its off-campus in the state. The new research park will be established in the outskirts of Hyderabad will focus on areas like AI, cybersecurity, defence and aerospace and aims to train professionals and carry out research and industry projects. Further, with close collaboration with the Telangana government, the institute aims to take more government projects in the listed field. In addition to CoEs, the institute has also been accommodated AI and other emerging technology related course. Among the courses offered in the field of AI, the topics include financial analytics, industrial automation, digital healthcare, intelligent transportation system among others. In August last year, the Institute broadened its academics to offer a unique course in law. The Rajiv Gandhi School of Intellectual Property Rights (RGSOIPL) in IIT Kharagpur announced that it will offer a one-semester a one-semester interdisciplinary course titled “Artificial Intelligence (AI) and Law”,  which aims to provoke a rethink on the modes of societal governance and law by AI capabilities. The programme will focus on foundational concepts, methods and tools, and areas such as Information retrieval and web search, text and language processing, image and video analytics. Stating that it is the one and only IIT in the country, a press release from the institute said that almost 56 students within the college have joined for the programme. “This course addresses the paradigm shifts that are occurring due to the intense involvement of AI as autonomous systems that can possibly invent, participate in inventions with humans, create new expressions like music, paintings, and raising sensitive questions on inventorship, authorship, ownership of patents, copyright, designs,” said Dean, RGSOIPL’s Padmavati Manchikanti. Corporate partnerships Among the numerous corporate tie up that it has made over the years, the Institute on March 28 announced that it will be collaborating with Indian IT giant, Wipro to bolster research in the field of AI and 5G technology. The joint partnership will see Wipro deploying the outcome of the research into their working, while IIT KGP is expected to benefit from the company’s high-end infrastructure and gain more research insight. “The two organizations will focus on AI research applicable in the healthcare, education and retail sectors as well as in domains such as climate change and cybersecurity. In addition, subject matter experts from Wipro and IIT Kharagpur will promote knowledge sharing through guest lectures, workshops and seminars on 5G and AI,” Wipro said in a press release.","excerpt":"IIT Kharagpur may soon expand its AI research capabilities with the help of its illustrious alumni network to benefit the best from the emerging technologies. The institute, which was founded in the year 1951, is one of the oldest and prestigious IITs in the country and has produced several meritorious alumni over the years. Sundar […]","categories":["AI News"],"tags":["AI centre of excellence","Sundar Pichai"],"author_name":"Akshaya Asokan","publish_date":"2019-04-01T05:30:33","publication_year":"2019","word_count":678,"keywords":["Go","artificial intelligence","AI","AI centre of excellence","Git","RAG","automation","Aim","analytics","GAN","R","Sundar Pichai"],"extracted_tech_keywords":["AI","artificial intelligence","analytics","Aim","RAG","R","Go","Git","GAN","automation"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/how-iit-kharagpur-is-strengthening-its-ai-research-with-the-help-of-its-illustrious-alumni-base\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10040070,"title":"Is Google FloC A Step In The Wrong Direction?","content":"Everyone loves the internet. But not the surveillance part that comes with the territory. An infamous aspect of this is the use of third-party cookies, which allow companies to track users and their online behaviour digitally. Many tech giants are openly in talks of fighting this level of tracking either by blocking cookies unless consented to by users—as Apple has done—or by replacing cookies with a different method to continue targeted advertising. Google, earlier this year, announced its plans for replacing cookies with a technology called FLoC, short for Federated Learning of Cohorts. Federated Learning of Cohorts, or FLoC, is a privacy sandbox technology that allows advertisers to reach people and target their ads without exposing individual details or an individual’s browser history. The technology pools users into ‘cohorts’ or individuals with similar interests and serve targeted ads to these groups. The cohorts will be generated through an algorithm that will place a person in a different group each week, based on his\/her browsing history. Google aims to completely replace cookies with FLoC by 2022, which might be difficult considering the hot and cold response the technology has been met with. What’s Cookieng? Bennett Cypher, from Electronic Frontier Foundation (EFF), has called FLoC ’a terrible idea’. He agrees that an alternative to cookies must be found but calls Google’s FLoC the most ambitious attempts at this, ‘and potentially the most harmful.’ Cypher agreed that adopting FLoC will do away with the privacy risks associated with third-party cookies. However, he claims it will generate new problems and exacerbate some of the worst non-privacy problems within behavioural ads—such as discrimination and predatory targeting. He also believes fingerprinting to be a significant problem with FLoC cohorts. Fingerprinting is when advertisers and companies create identifiers out of information gathered from browsing patterns. Cypher states that a tracker starting with an individual cohort would only need to distinguish personal interests out of a few thousand other users instead of a few hundred million. This might make fingerprinting easier for FLoC users. The EFF believes that a much better way around targeted advertisements would be to do away with them altogether for uniform advertising. Many browsers are either against the new technology or on the fence. Brave and Vivaldi have been very open about their rejection of FLoC. Brave has called FLoC ’a step in the wrong direction’ and believes it is terrible for user privacy but with a label of user-friendliness. On the other hand, Vivaldi has said they ‘refuse to get FLoC’ed’ and plan on disabling the tech. Vivaldi has also pointed out that a FLoC ID would still present private browsing history data to advertisers—even if it is semi-anonymous—and increases Google’s power in the industry as the developer of both FLoC and Google Chrome. Microsoft’s browser Edge, Mozilla Firefox and Apple’s Safari etc are still on the fence. Microsoft and Mozilla issued statements to The Verge that they support solutions that give people control and privacy and that advertising and privacy can coexist—respectively. Neither of them nor Apple have given a definitive answer to whether they will allow FLoC and are most likely waiting for further improvements in FLoC’s security, transparency and privacy. Meanwhile, Google is all set to do a test run for FLoC on Version 89 of Google Chrome. Google will do this test in multiple countries, including Australia, Brazil, India, Indonesia, Japan, the Philippines and the United States. There is some uncertainty around whether FLoC clears GDPR. The EFF, in response to this, has set up a website designed to let users of Chrome know whether they are a part of this trial run. If you are a part of FLoC but do not want to be, you can either opt for another browser, choose to block third-party cookies on your Chrome Settings panel and disable it, or install an extension to block FLoC. Privacy, especially on the world wide web, is a tricky affair. Google’s new initiative has chinks in its armour. However, with users’ citing concerns on privacy, advertisers fighting for their beloved cookies and browsers debating them—Google’s plans to replace cookies with FLoC by next year might not fruition.","excerpt":"Everyone loves the internet. But not the surveillance part that comes with the territory. An infamous aspect of this is the use of third-party cookies, which allow companies to track users and their online behaviour digitally. Many tech giants are openly in talks of fighting this level of tracking either by blocking cookies unless consented […]","categories":["Global Tech"],"tags":["Privacy"],"author_name":"Mita Chaturvedi","publish_date":"2021-05-14T12:00:00","publication_year":"2021","word_count":692,"keywords":["federated learning","Go","Privacy","emerging_tech:federated learning","programming_languages:R","AI","programming_languages:Go","Git","Aim","R"],"extracted_tech_keywords":["AI","Aim","federated learning","R","Go","Git","programming_languages:R","programming_languages:Go","emerging_tech:federated learning"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/is-google-floc-a-step-in-the-wrong-direction\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10167731,"title":"We Will Need 10x More Software Engineers in 5 Years","content":"AI has sparked anxiety within the software engineering community. We are nearing a world where most of the coding is done in natural language, with tools like Replit, Cursor, Windsurf, and Lovable essentially making software in a matter of minutes. The tech leaders are also not helping the cause, with their extensive debates on whether we would even need software engineers in the future. Thankfully, not all opinions are gloomy, offering the developers a little hope and relief. Claim your free Nvidia course > Todd McKinnon, CEO of Okta, is one of the recent voices pushing back against the doomsday narrative. In an interview with Business Insider, McKinnon called the fear of declining engineering jobs “laughable”. He compared the rise of AI to past tech shifts like personal computing and mobile, which ultimately expanded job opportunities rather than eliminating them. “In five years, there will be more software engineers than there are now,” McKinnon said, and these engineers would be building things on top of the current solutions. He also added that companies like Microsoft, Meta, and Salesforce will hire more software engineers in the near future. How Much of This is True? Obviously, not everyone agrees with this sentiment. Sam Altman, the CEO of OpenAI, speaking on Stratechery with Ben Thompson, said, “Each software engineer will just do much, much more for a while…And then at some point, maybe we will need fewer software engineers.” While he acknowledged that “agentic coding”—AI autonomously handling complex software tasks—is on the rise, he cautioned against assuming this spells the end for engineers. Altman noted that some companies are already generating over half their code through AI. However, he emphasised that human oversight remains crucial. Alarm bells really went off when Anthropic CEO Dario Amodei predicted that in less than six months, AI would handle 90% of coding. This is similar to what Sridhar Vembu, founder of Zoho, thinks. He recently said that 90% of what programmers write today is ‘boilerplate’. But coding is just one part of the software engineering cycle. As AI grows more powerful, the engineers behind it are becoming even more valuable. Microsoft CEO Satya Nadella has long advocated for a broader view of software engineering. Rather than focusing solely on the tech sector, Nadella pointed to LinkedIn data that shows robust demand for engineers across healthcare, finance, manufacturing, and education. Google CEO Sundar Pichai sees the role of software engineers evolving rapidly in the face of AI. In an earlier podcast of The David Rubenstein Show: Peer to Peer Conversations, he said Google continues to prioritise “superstar engineers”—those who thrive in fast-changing environments and possess mastery in fields like AI, quantum computing, and cybersecurity. Despite hiring slowdowns, Pichai revealed that Google maintains a 90% offer acceptance rate for engineering hires. His point: demand hasn’t dropped, it’s just shifted. Engineers now need to be more versatile, more interdisciplinary, and more comfortable working with AI-powered toolchains. For Google, innovation isn’t possible without human talent at the helm. There is No Such Thing as ‘Vibe Engineering’ It remains clear that while the world is vibing to code, there is still no such thing as vibe engineering. A recent blog by Sergey Tselovalnikov highlighted that while vibe coding can accelerate prototyping, it doesn’t replace software engineering. Real engineering involves long-term system design, reliability, scalability, and maintainability—concerns that current AI-generated code cannot handle. So, despite the hype, there’s no such thing as “vibe engineering”; it’s still just engineering, with or without the code typing. Last year, the debate around AI replacing jobs was in full swing, just like today. That is when François Chollet, creator of Keras, made a prediction: “There will be more software engineers (the kind that write code, e.g. Python, C or JavaScript code) in five years than there are today.” He added that the estimated number of professional software engineers today is 26 million, which would jump to 30-35 million in five years. Being a developer or coder is not just about writing code but about engineering and thinking of higher-order things, which is where the future value of software engineering lies. “Code is largely worthless, more of a liability than an asset. Problem-solving is where the value is,” Chollet added in the thread. “If you could fully automate software engineering (my job), I think that would be great, since I could then move on to higher-leverage things. Making software is a means to an end, not the end.” In its October 2024 report, Gartner said that GenAI will spawn new roles in software engineering and operations through 2027, requiring 80% of the engineering workforce to upskill. It remains true that software engineers have to upskill faster than anyone else. Especially in India, where most IT firms and engineers are not learning to code but merely providing services, AI could be their biggest threat. If models can write code, test it, and check for self-consistency, a form of automatic supervision that is not possible in most domains due to the limits of human expertise will be ushered in. This capability would allow code to be tested empirically and automatically by AI agents. The future will be about one engineer managing a bunch of AI coding agents. As a result of this, software engineering will see a radical transformation and possibly an increase in demand. “There will be way more software engineers in the future than the present,” Russell Kaplan from Cognition Labs earlier noted. But “the job will just be very different: more English, less boilerplate coding”.","excerpt":"The future will be about one engineer managing a bunch of AI coding agents.","categories":["AI Features"],"tags":["Vibe Coding"],"author_name":"Mohit Pandey","publish_date":"2025-04-11T09:00:00","publication_year":"2025","word_count":918,"keywords":["Anthropic","GenAI","Keras","OpenAI","AI","RAG","Python","Vibe Coding","Aim","JavaScript","R"],"extracted_tech_keywords":["AI","GenAI","OpenAI","Anthropic","Aim","Keras","RAG","Python","R","JavaScript"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/we-will-need-10x-more-software-engineers-in-5-years\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10103601,"title":"NVIDIA and Genentech Forge Strategic AI Alliance for Drug Discovery","content":"Genentech, a member of the Roche Group, and hardware giant NVIDIA have sealed a multi-year research collaboration. In a strategic alliance, Genentech will leverage its expertise in AI and extensive biological datasets, teaming up with NVIDIA’s computing power and AI capabilities. The collaboration is to enhance Genentech’s machine learning algorithms through NVIDIA’s DGX Cloud—a platform specifically tailored for AI applications in drug discovery. Under the leadership of CEO Jensen Huang, the move will not only improve Genentech’s scientific models but also potentially enhance NVIDIA’s broader platforms. Huang envisions this partnership as a catalyst for transformative change within the healthcare sector, particularly in expediting the drug discovery process. This collaborative effort majorly supports Genentech’s AI and machine learning ML teams, who are actively developing foundational models spanning diverse therapeutic domains. Moreover, it speeds up the implementation of Genentech’s innovative “lab in a loop” strategy, where experimental data informs computational models. These models then generate predictions that undergo real-world testing in laboratories, subsequently feeding back results to enhance and refine the computational models in a continuous improvement cycle. In the course of their collaboration, both public and proprietary data from Genentech will be used with the company maintaining strict control over access to its private data. Notably, NVIDIA will not have direct access to Genentech’s proprietary data unless expressly permitted for a particular project. NVIDIA also made waves earlier this year with the introduction of BioNeMo Cloud, an important component of its AI Foundations suite. This platform provides access to pre-trained AI models, a development that has already proven instrumental in expediting drug discovery processes for notable companies like Amgen and various startups. Not just BioNeMo, NVIDIA also plays an important role in advancing drug discovery through its GPU-accelerated platform, NVIDIA Clara which integrates AI, data analysis, simulations, and visualisation to swiftly sift through extensive chemical libraries for potential drug candidates, modelling protein structures and dynamics to comprehend their roles and interactions with medications, crafting novel molecules with desired attributes, simulating drug effects within the human body, and visualising findings from drug discovery experiments.","excerpt":"The collaboration is to enhance Genentech’s ML algorithms through NVIDIA’s DGX Cloud","categories":["AI News"],"tags":[],"author_name":"Tasmia Ansari","publish_date":"2023-11-23T21:32:48","publication_year":"2023","word_count":342,"keywords":["Go","machine learning","programming_languages:R","AI","ML","programming_languages:Go","RAG","R","startup"],"extracted_tech_keywords":["AI","machine learning","ML","RAG","R","Go","startup","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/nvidia-and-genentech-forge-ai-alliance-for-drug-discovery\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10117796,"title":"Data-Driven Decisions Made Easy With DataSwitch’s DS Integrate","content":"When it comes to Data Modernization from on-premise to the cloud, DataSwitch is a proven leader, especially in automating the migration processes. With its trio of tools— DS Migrate, DS Integrate, and DS Democratize—DataSwitch covers the entire data transformation lifecycle. Many companies struggle with data trapped in siloed systems, where marketing data might reside in a CRM while financial data sits in an ERP. This fragmentation makes it challenging to gain a unified view of business. Here’s where DS Integrate comes into the picture. DS Integrate acts as a bridge between various data sources, making it easy to combine and organise data for specific purposes. It simplifies handling previously unstructured data, ensuring seamless integration and usability. DataSwitch is compatible with legacy systems, including Oracle, Teradata, Netezza, Informatica, SSIS, and DataStage. Additionally, DataSwitch supports integration with modern cloud platforms like AWS RedShift, Snowflake, BigQuery, DataBricks, and Spark. Why Choose DS Integrate? Traditionally, data ingestion and transformation can be complex and time-consuming, requiring manual coding or scripting. DS Integrate offers a user-friendly interface with pre-built connectors and functionalities, allowing businesses to ingest data from various sources and transform it into a usable format without extensive coding expertise. ‘With no code, DS Integrate will reduce dependency on core technology personnel, allowing business professionals themselves to perform data analysis. That is one of the objectives. It’s not that DataSwitch is killing jobs,” said DataSwitch chief Karthikeyan Viswanathan in an exclusive interview with AIM. DS Integrate automatically generates code to create a knowledge base in a format compatible with cloud databases such as Spark, Talend, Matillion, DataBricks, and more. It comes with data standardisation. “DataSwitch’s DS Integrate is designed to handle data arriving in various formats, including PDFs, images, text, ODBC, and JDBC. It follows industry-standard coding practices to transform this diverse input into a structured data catalogue,” said Viswanathan. After standardising data, DS Integrate enables users to convert raw data into valuable insights without requiring advanced coding skills. This approach, termed as Citizen Data Engineering, dramatically improves accessibility to data engineering, encourages innovation and agility, facilitating quick adaptation to changing market dynamics. “A minimal understanding of technology is needed. Someone with a clear understanding of what they want to achieve can perform data engineering easily with DS Integrate,” said Viswanathan. Who Can Benefit from DS Integrate DataSwitch primarily serves companies facing challenges in data migration, data integration from multiple sources, and data transformation for analytics. The company primarily provides services to global system integrators. Viswanathan said DS Integrate is a perfect tool for junior engineers and freshers at multinational or top service providers  needing more experience. He added that DS Integrate incorporates expertise equivalent to that of a ten-year experienced engineer. According to him, while experienced engineers demand higher salaries, companies can leverage freshers’ capabilities using this tool to successfully complete projects. “Freshers can use DS Integrate, which benefits  their firms by delivering more projects. This will also create more job opportunities for junior engineers,” he said. DataSwitch is also targeting the Global Capability Centers (GCCs) market in India and globally. A Nasscom-Zinnov report reveals that India had 1,580 GCCs with 1.66 million employees as of 2022-23. In the first half of 2023, 18 new GCCs were established in Tier-I cities such as Mumbai, Pune, and Bengaluru. Viswanathan explained, ” For businesses, the purpose of GCCs is to develop their teams in India instead of hiring professionals from companies like TCS.” He further added that GCCs often rely on standard technologies that can become outdated over time, emphasising the need to adapt and evolve beyond their initial setups. “DS Integrate is a handy tool for them, given their expertise in legacy technology. With DS Integrate, a simple click generates code more efficiently than even trained, experienced individuals,” concluded Viswanathan.","excerpt":"DS Integrate acts as a bridge between various data sources, making it easy to combine and organise data for specific purposes. It simplifies handling previously unstructured data, ensuring seamless integration and usability.","categories":["AI Highlights"],"tags":["DataSwitch"],"author_name":"Siddharth Jindal","publish_date":"2024-04-09T10:00:00","publication_year":"2024","word_count":625,"keywords":["AWS","AI","ML","RAG","Aim","Databricks","data engineering","analytics","R","DataSwitch","Snowflake"],"extracted_tech_keywords":["AI","ML","analytics","Aim","RAG","AWS","Snowflake","Databricks","R","data engineering"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/data-driven-decisions-made-easy-with-dataswitchs-ds-integrate\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":61102,"title":"Deep Dive: How Fintech Firm XPay.Life Is Taking AI &#038; ML To Rural India","content":"The increase of mobile applications has provided an ample amount of opportunities to different kinds of industries in India. One of the sectors that has benefited the most with the increasing use of mobile is Fintech. The Fintech industry has adopted mobile apps aggressively, with digital applications that help exchange money between two parties without any cash being involved. Furthermore, with the Indian Government pushing for a cashless economy, Fintech platforms have benefitted even more. However, it seems that initiatives like ‘Cashless India’ and ‘Digital India’ have been mostly successful in urban cities only (65%). Nearly 80% of the rural population still primarily use cash. Also, many of these online transactions in urban cities involve bill payments to avoid standing in queues. This is not the case in rural areas. With the vision to make life simpler by providing a one-stop digital solution in the bill payment space, XPay.Life is the latest offering from XIPHIAS. The company has a decade-long expertise in providing bill payment solutions through touchscreen kiosks. For this week’s Deep Dive column, we got in touch with XPay.Life’s Co-Founder, COO\/CTO Bohitesh Mishra to discuss its growth and how the company uses AI and ML in their products and services. About The Company’s Growth XPay.Life was founded in June 2019 by Rohit Kumat (CEO) and Bohitesh Misra (COO and CTO). XPay.Life is a BBPS\/NPCI assured agent institution which provides both B2B and B2C services. Through the years, XPay.Life has moved from prototyping products like Xpay Life Mobile App, Online web, Mobile Van, ATP Kiosk and PoS machine, to launch them into services across their customers. These products have been well-received by their customers. This is demonstrated by Rs 100 crore monthly bill payment collection transactions on their Touch Screen ATP Kiosk. This number comes from a single biller from Jaipur. In future, XPay.Life intends to add 253 billers at their collection points across India. Use Of AI and ML XPay.Life uses AI to cluster the customer’s bill payment behavior and promote their services accordingly. Further, they leverage machine learning (ML) methods to identify any fraudulent and suspicious transactions. AI is also used to alert customers about the next payment that is due. XPay.Life uses ML outlier methods to identify any fraudulent activity or suspicious transaction. Upon identifying such transactions, these are immediately blocked and rejected by the system. The ML algorithms on their platform help XPay.Life fight such fraudulent transactions. XPay.Life also has bots on their platforms to interact with customers. In fact, the company claims that often customers cannot tell when they are interacting with bots or human agents when getting their complaints and queries addressed. XPay.Life also uses engines to help them schedule their mobile vans for collection of customer bills in tier-2 and tier-3 cities, and even in some villages. These engines also help them optimize the travel path of these vehicles. XPay.Life’s Products, Training Model & Hiring Processes XPay.Life’s USP is bill payment through Touch Screen ATP Kiosk, PoS machine and Mobile Van, which accepts cash as a digital payment. XPay.Life has an integrated blockchain-based payment gateway for ensuring the complete security of these transactions. As for the training model, the company follows security based on the AMBIC model, that is, AI, Mobility, Blockchain, IoT and Cloud, providing a transparent and secure platform for bill payment through a multitude of digital solutions. XPay.Life’s development team consists of 40 engineering teams. When it comes to hiring, they look for candidates who have an acute knowledge of basic concepts and well-polished analytical skills when it comes to AI and data science. The company develops and trains its employees on the latest updates and developments that their platform receives, since XPay.Life’s platform is challenging to work on. They follow a train-the-trainer concept and newly joined employees are trained on new technologies, products and solutions. Customer Satisfaction & Future Roadmap With an efficient operations and support team, XPay.Life reaches out to customers for any help they need that is related to business and transactions. Their bots and efficient helpdesk help them improve their customer service. While all other players in the industry are focused on digitizing the urban sector, XPay.Life believes that rural areas hold untapped sectors with USP of ATP Kiosk, PoS machine and Mobile Van that accept cash as a ‘digital payment mode’ with other digital payment channels. In the future, XPay.Life is planning to deploy one lakh ATP Kiosks, PoS devices and Mobile Vans within the next three years to drive digital empowerment of rural India.","excerpt":"The increase of mobile applications has provided an ample amount of opportunities to different kinds of industries in India. One of the sectors that has benefited the most with the increasing use of mobile is Fintech. The Fintech industry has adopted mobile apps aggressively, with digital applications that help exchange money between two parties without […]","categories":["AI Features"],"tags":["FinTech","latest ai products across industries"],"author_name":"Sameer Balaganur","publish_date":"2020-04-07T12:00:00","publication_year":"2020","word_count":751,"keywords":["data science","Go","machine learning","AI","ML","Git","RAG","Aim","ViT","latest ai products across industries","FinTech","R"],"extracted_tech_keywords":["AI","machine learning","ML","data science","Aim","RAG","R","Go","Git","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-fintech-firm-xpay-life-is-taking-ai-ml-to-rural-india\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10065203,"title":"Fujitsu launches new centre in Bengaluru to accelerate AI and ML innovation","content":"Fujitsu announced its new research centre in India to strengthen its focus on research and development into AI, ML technologies and quantum software. Fujitsu India said that by 2024, they would have about 50 researchers to expand their research fields into security and other areas and conduct software R&D in collaboration with Fujitsu’s global network of research centres in Japan, Europe, and the US to create software for global distribution. Established in Bengaluru on April 1, 2022, Fujitsu Research of India Private Limited (FRIPL) will embark on a joint research programme with the Indian Institute of Technology, Hyderabad, and the Indian Institute of Science, Bengaluru, to promote innovation in leading-edge AI technologies. The joint research activities will focus on improving the accuracy and resilience of AI and ML technology. The research and development with IIT Hyderabad will focus on AI technology to discover causal relationships with higher accuracy, while collaboration with IISc will centre on technology to automatically generate AI through autonomous training in response to various environmental changes. Vivek Mahajan, Chief Technology Officer (Corporate Executive Officer, SEVP), Fujitsu Limited, said, “Innovation has a growing role in solving societal and environmental challenges. Strengthening our presence in India will allow us to tap into the enormous potential offered by world-class researchers with local institutions and universities that drive global software technology development. We look forward to conducting more agile and challenging joint research to deliver a more sustainable future for humanity.”","excerpt":"Fujitsu Research of India Private Limited was established in Bengaluru on April 1, 2022.","categories":["AI News"],"tags":[],"author_name":"Poornima Nataraj","publish_date":"2022-04-18T16:51:47","publication_year":"2022","word_count":240,"keywords":["programming_languages:R","AI","innovation","ML","ViT","edge AI","emerging_tech:edge AI","R"],"extracted_tech_keywords":["AI","ML","edge AI","R","ViT","innovation","programming_languages:R","emerging_tech:edge AI"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/fujitsu-launches-new-centre-in-bengaluru-to-accelerate-ai-and-ml-innovation\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10171445,"title":"How Stripe Used AI to Boost Fraud Detection from ‘59-97% Overnight’","content":"There’s a perennial buzz about AI seeping into almost every part of a company’s operations. However, some of the most valuable and tangible results can be achieved if AI can help prevent scenarios that hit people where it hurts the most: losing money. Stripe’s use of AI to boost fraud detection and increase security measures is a crucial case study on how AI can be used in payment processing, especially given the scale at which the company operates, with customers like OpenAI, Amazon, Google, Apple, and many more. At the Stripe Sessions product keynote held last month, the company announced the Payments Foundation Model as part of its efforts to boost fraud detection and enhance security. Stripe also touted it as the “world’s first foundational model built for payments”. This foundation model highlighted the company’s success in identifying card testing, wherein bad actors try to determine whether stolen card information is valid so that they can use it to make purchases. Stripe Built a Transformer-Based ‘Payments Foundation Model’ Gautam Kedia, who leads applied machine learning at Stripe, elaborated on the company’s methodology in a LinkedIn post. Kedia revealed that while standard machine learning models have helped Stripe reduce fraud, each requires task-specific training for activities like authorisation, fraud detection, dispute resolution, and more. “Given the learning power of generalised transformer architectures, we wondered whether an LLM-style approach could work here. It wasn’t obvious that it would—payments are like language in some ways,” Kedia explained. Hence, the company decided to build a foundational payments model—a self-supervised model that learns “dense, general-purpose vectors for every transaction, much like a language model that embeds words”. The model is said to be trained on tens of billions of transactions and distils each charge’s key signals into a single embedding. “You can think of the result as a vast distribution of payments in a high-dimensional vector space. The location of each embedding captures rich data, including how different elements relate to each other.” Kedia further explained that payments with shared characteristics naturally group together. For example, transactions from the same card issuer are clustered, those from the same bank are even closer, and payments using the same email address appear almost indistinguishable. He further indicated that these ‘rich embeddings’ make it easier to spot nuanced, adversarial transaction patterns and build more accurate classifiers based on both the features of an individual payment and its relationship to other payments in the sequence. In the last two years, the company has reduced card testing for users on Stripe by 80%. However, more sophisticated attacks, where the card testers hide novel attack patterns in the volumes of the largest companies, make it hard to spot them with traditional methods, Kedia suggested. “We built a classifier that ingests sequences of embeddings from the foundation model and predicts if the traffic slice is under an attack,” he said. Kedia added that this works in real time so that Stripe can block attacks before they hit businesses. “This approach improved our detection rate for card-testing attacks on large users from 59% to 97% overnight.” He also said that Stripe’s success may suggest that payment activities contain semantic meaning. “Just like words in a sentence, transactions possess complex sequential dependencies and latent feature interactions that simply can’t be captured by manual feature engineering,” he said. Kedia also shared a table comparing the performance of Stripe’s own foundational model and ‘incumbent’ machine learning models. Stripe Also Recovered $6 Bn in Legitimate Transactions That Were Declined Notably, the company also reported using AI to prevent redundant transactions and to identify which transactions must actually be retried. “This resulted in more than $6 billion in legitimate declined transactions recovered for our users in 2024—a record amount,” John Affaki, payments business lead at Stripe, wrote in a blog post. The company stated that it resolved the issue through its product, Adaptive Acceptance, which uses AI to automatically identify falsely declined transactions. Stripe said it can recognise various patterns in transaction data, which indicate that a legitimate payment was mistakenly rejected by issuers as suspected fraud. Previously, Stripe used a gradient-boosted tree model, XGBoost, but then transitioned to a TabTransformer-based deep neural network, which the company calls TabTransformer+. “This system excels at modelling complex interactions among hundreds of factors that influence transaction success,” the company stated. It further stated that the new architecture also features high-dimensional embeddings, which map payment patterns and enable the model to capture and analyse signals that affect payment outcomes. This allows the model to make “more nuanced decisions” about which declined transactions to retry and how to adjust them for a higher approval chance. “Based on these improvements, Adaptive Acceptance’s new AI model achieves 70% greater precision in identifying legitimate transactions that have been falsely declined. This increased precision allowed us to recover more revenue than ever last year while reducing retry attempts by 35%,” the company said. Besides, Stripe’s fraud prevention tool Radar was updated with automated authentication capabilities. It can now trigger 3DS, Stripe’s additional layer of security, to enable a two-factor authentication flow. Stripe also said it is backed by a new multihead model and a decision-making layer, and early users have seen a 30% reduction in fraud on eligible transitions. All things considered, AI is extensively used in the payments processing industry. Several leading global giants, including Razorpay, a competitor of Stripe, are using AI to tackle delayed customer payments, simplify setting up payment gateways, and reduce the return-to-origin problem.","excerpt":"Stripe’s use of AI to boost fraud detection and increase security measures is a crucial case study on how AI can be used in payment processing.","categories":["Global Tech"],"tags":["AI (Artificial Intelligence)","stripe"],"author_name":"Supreeth Koundinya","publish_date":"2025-06-07T10:43:55","publication_year":"2025","word_count":912,"keywords":["Go","machine learning","OpenAI","AI","neural network","feature engineering","Git","XGBoost","stripe","R","AI (Artificial Intelligence)","fraud detection"],"extracted_tech_keywords":["AI","machine learning","neural network","OpenAI","XGBoost","fraud detection","R","Go","Git","feature engineering"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/how-stripe-used-ai-to-boost-fraud-detection-from-59-97-overnight\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":40195,"title":"Addressing India&#8217;s Reskilling Challenge &#8211; A Report By AIM","content":"Even with the third-largest developer base and a substantial tech-savvy talent pool, India lags behind its peers on major AI indicators. This is despite a thriving startup ecosystem, high-growth companies which have made a substantial investment in setting up CoEs and the Government investing in building a robust tech infrastructure. Behind the AI and data analytics boom, lies the story of a massive talent gap as workforce struggles to remain employable. The skills’ shelf life has shortened, with technology changing exponentially over the last decade, skills that were relevant at the beginning of the career have become obsolete. In order to remain employable, the workforce needs to reskill to take advantage of new opportunities. The rise of edtech companies in India is not surprising, given the huge clamour for continuous learning that has taken root in the professional sphere. This is backed by the rise of emerging technologies — artificial intelligence, its subset machine learning and data science which has spawned a booming job market revolving around new technologies that has substantially transformed India’s IT labour market. The changing job economy has resulted in new opportunities for the Indian workforce. As estimated by a consulting major, AI has the potential to add US$957 billion, or 15 percent of India’s current gross value in 2035. The booming economy, fuelled by AI and advanced analytics requires more Indians to enter the workforce with a different skill-set. As per our estimate, close to 97,000 AI positions lie vacant in India. But, the challenges are also increasing multifold — on the one hand India Inc is struggling with disruptions like automation that are redefining jobs and secondly, it is grappling with finding the right talent with the right skillset for AI\/ML and data science teams. Meanwhile, the upcoming generation that will enter the workforce soon is fed on an outdated curriculum that hasn’t kept up with the industry’s demands. In our report, we dig into the educational stakeholder landscape to see how they are transforming the skills market by developing training courses and certification programmes that correspond to in-demand skills required today. We look at the type of educational institutions offering data and analytics programs; how the educational landscape is changing in response to the heightened demand for analytics skills and what needs to be done to fill the skill gap. The second half of the report looks at our last three years ranking data to find out the winning attributes that have helped analytics institutes rank on top consistently and how other training institutes have fared over the last three years. Key Highlights The online reskilling market is estimated to be $93 million and is expected to grow at a rate of 38%. As compared to other educational categories (secondary supplemental education providers and GMAT\/ GRE\/GATE test preparation providers) the reskilling market is more mature Current market is largely B2C driven but educational stakeholders are also actively catering to the B2B segment Reskilling market in India is driven by the needs of a large working population looking for industry-relevant skills Online key players are also moving towards blended educational solutions by creating offline touchpoints to provide peer interaction Emphasis on personalised learning has led to mentorship and offline touchpoints that helps students gain hands-on experience for particular concepts Business Analytics course was the starting point, besides this, other courses that are gaining traction are Artificial Intelligence, Machine Learning, Data Science & Analytics & Data Engineering Partnerships between analytics education providers and universities in offering niche courses Higher demand for short-term diploma courses in in-demand areas such as Blockchain, Data Science and Machine Learning Virtual classroom concept that began in 2014 has brought high quality analytics education more accessible Key tools learnt are R, SAS, Python, on big data end Hive, Pig, Hadoop and in AI\/ML end Tensorflow and Keras Key Players In The Reskilling Market In order to capitalise on these opportunities, IT companies, educators and policymakers need to develop a deeper understanding of the existing workforce, the skill-set required in the future, and the gaps that will need to be addressed. This implies that these three key players need to align the broader economic developer agenda with the shifting job market and work towards building a strong talent that has the baseline and digital skills required for current landscape. At the Government level, policy makers will have to assess secondary and postsecondary education and align it with the skills that are required for tomorrow. Many leading Indian IT majors have undertaken employer-training initiatives, pre-employment training and have also provided their own courseware. Collectively, the key stakeholders can foster a workforce development ecosystem and provide domain specific training with a job-first approach. Given this scenario — educational stakeholders have made a very strong business case for reskilling the workforce and have actively partnered with renowned educational institutions to launch technical certifications and degree programmes tailored to fill the skill gap. Analytics Education Landscape The nature of analytics education has evolved over the last few years and a mix of models have emerged in the online and offline space to accommodate the changing requirements of students. Learners seek a career-focused analytics education augmented by classroom setting that prepares them for job functions in data analytics space. • In cases where learning is delivered purely online, participants look for realtime learning in a format that allows learners to pursue it at their own pace • Candidates look for course content created by top instructors, with industry and university collaboration to provide a well-rounded analytics education • Executive programs are also in high demand as these are intended for senior professionals who want to renew their skillset and understand how data can be helpful in managerial decision making • In case of executive analytics courses, technical skills such as data management are augmented by soft skills such as business understanding and communication • Analytics education providers in India mostly offer Business Analytics (BA) and Business Intelligence (BI) programs that combine analytical number crunching, reporting and visualization techniques Learning Formats The learning formats can be broadly put under 4 categories: Self-paced learning delivered via recorded video content Instructor-Led live classroom sessions delivered online Blended learning format with classroom and online delivery Bootcamps for intensive, in-person learning that provides a hands-on experience Around 87 percent of analytics courses from private training institutes are delivered in the self-paced learning models 6 percent are delivered in the hybrid (Self-paced and Instructor-Led online) format and 4 percent in Instructor Led weekend and self-paced format There’s only a 3 percent uptake for weekend classroom format On average, analytics courses by private institutes offer 105 hours of instructor contact hours The hybrid model of self-paced + online Instructor-Led courses has the highest number of contact hours at 157. The blended learning opportunity allows learners to get continuous feedback and participate in real-time assessment Weekend-only model has the least contact hours at 75 For those looking for face-to-face learning environment, weekend model is the best fit Breakup by Course Type 56 percent of all long-term analytics\/ data science programs by analytics training institutes in India are part-time courses The rest 44 percent are full-time programs Ranking Data Science Institutes In India Analytics India Magazine started the tradition of ranking Data Science Institutes in India that are at the forefront of reskilling revolution. Our annual ranking report, started in 2014 measures the institutes on various parameters such as course content & comprehensiveness, faculty details such as those with PhD or industry experience, student experience, for example post completion engagement. Other parameters included external collaboration, placement assistance and more. Each of these parameters is ranked on a scale of 0-5 where 0 is for worst and 5 for best. The institutes that could not be a part of the list either did not participate in the ranking process or could not make it to our list. For long, professionals and aspirants have relied on our findings to make informed career and educational choices. What’s remarkable is how these educational stakeholders fully embraced the challenge, had a pulse on the changing trends and forged deep industry bonds to ensure adequate bandwidth and resources to achieve the talent roadmap. The institutes that have consistently come out on top are the ones that have excelled in providing an experiential data and analytics programme delivered in collaboration with international universities and industry majors. What’s more — the certifications and diploma programmes are tailored towards early-stage and mid-career professionals who need to upskill data and analytics skills that has become core to organisations. The institutional strength is also measured on the basis of training pedagogy, faculty strength and its industry focused curricula. The institutes that are top-ranked are also rated on their inventory of data resources, labs with access to in-demand computational tools and networking events. A key evolution of analytics training institutes is how these online players have broadened their offerings by providing offline touch points to students with basecamps, meet-ups and networking events. Some of the value-added services provided by Data Science Training Institutes are: Reputed University Credentials (international & Indian) Branded Bootcamps Corporate Tie-ups Capstone Projects leading to hands-on experience Internships & Placement Assistance Networking Sessions Another shift has been in the form of accelerating interdisciplinary research with industry stakeholders in Artificial Intelligence and Data Science that serves to catalyze R&D in India. Based on our ranking report, we look at the top performer in the game that has consistently clinched the #1 spot and list down the reasons behind its top slot. Ranking Market Map We dig into the last three years ranking reports — 2016, 2017 & 2018 to find out the winning factors that helped edtech firms stay on top of the game. The 4 key parameters are — course content, faculty, student experience and external collaboration. Some of the other details under these parameters are course comprehensiveness, post-completion engagement of students, capstone projects and others that result in student retention and placement success. The fast-evolving talent market requires a blend of skills for analytics-enabled and data science roles. Analytics training providers that managed to keep abreast of industry developments, develop\/ revise data analytics courses and forge collaborations with key industry players to match skills required by employers have continued to retain the top spot. If we look at the last 3 years, Bangalore headquartered Jigsaw Academy has clinched the top spot consistently. When it comes to the third spot, there have been changes with INSOFE coming third in 2016, losing the spot to UpX academy in 2017 and regaining the title again in 2018. Throughout this period, Jigsaw Academy consistently grabbed the #1 spot with its carefully crafted data-centric programs. Here are the key attributes for edtech firms clinching top spots on ranking report: Industry-driven courses: Edtech companies that have recognized the need for accelerating formalized education through partnerships with renowned education providers Analytics institutes that have stayed true to their core offering (Data Science) and have diversified the learning options to provide candidates with a specialized development path or mentor-style learning opportunities Edtech firms that offer the highest placement rate with a return on investment Edtech firms that have focused approach towards Data Science education, have an updated industry-driven curriculum that offers the knowledge and skills for analytics-enabled roles Formalizing analytics with higher educational institutes: Edtech companies that have recognized the need for accelerating formalized education through partnerships with renowned education providers Accelerated programmes: based on the needs of working population: Analytics institutes that have stayed true to their core offering (Data Science) and have diversified the learning options to provide candidates with a specialized development path or mentor-style learning opportunities Placement Success: Edtech firms that offer the highest placement rate with a return on investment How Jigsaw Academy Stepped Up The Reskilling Game To Prepare Talent For Data-driven Workplace Jigsaw Academy has consistently been ranked as the #1 Data Science training institute over- 2017, 2016, 2014 and 2013 and also topped the list last year. The reason why Jigsaw Academy clinched the top spot was because the institute is one of the few “specialists” that stuck to its core — developed a set of top programmes in collaboration with international universities that are suited to beginners, developers and even leaders to fill the data science talent gap in India. In the last 8 years, Jigsaw Academy has developed an elite set of graduates from over 30+ countries who are crosstrained in all aspects of data science and data visualization and are working across 3000+ companies. In this report, we look at Jigsaw Academy’s role in leading the data science revolution in India and how it is advancing data science and AI education in India. We also look at the reasons why we believe Jigsaw Academy has become the home for data science education, providing a broad set of audience with a comprehensive curriculum that allows them to shift careers into analytics and data science domain effectively. Key Drivers Of Jigsaw Academy’s Rise 1) Jigsaw Academy has excelled by addressing a large set of audience through its innovative offerings across online, in-person and corporate training. Jigsaw Academy has a long-standing history of collaborations with leading universities such as the University of Chicago Graham School and IIM Indore that help them in driving advanced and established leadership in the field of data science. In fact, Jigsaw Academy’s PGP in Data Science and Machine Learning is one of the elite flagship programmes offered by the University of Chicago Graham School and Jigsaw Academy. This PGP in Data Science and Machine Learning was ranked number 2 in the list of executive data science courses in India. Recently, Jigsaw Academy in collaboration with IIM Indore launched a 10-month intensive Integrated Program in Business Analytics in response to the constant demand for data-centric education used to power business decisions 2) Jigsaw Academy is one of the first institutes in India to bring advanced analytics training to industry. While there are a lot of courses available in the market, what differentiates the Jigsaw Academy offerings are that they provide a comprehensive coverage of analytics as well as its business application 3) As industries and enterprises became more data-intensive and shifted towards an AI\/ML stack, Jigsaw Academy was quick to capitalize on this shift and introduced a full-stack ML & AI course which was ranked number 4 in the list of best courses and programs on AI in India. The full-stack AI\/ML course was developed to reflect the strategic direction where the market is heading and also in fulfilling new talent demand. 4) Corporate training is another key area where Jigsaw Academy has excelled with training initiatives for Fortune 500 companies like Titan, Reliance Industries Limited & data visualization firm Gramener. So far, Jigsaw Academy has trained over 150 companies and their corporate training programmes are well-suited to senior leaders who are tasked with building teams and piloting several PoCs within their organisation. 4) Corporate training is another key area where Jigsaw Academy has excelled with training initiatives for Fortune 500 companies like Titan, Reliance Industries Limited & data visualization firm Gramener. So far, Jigsaw Academy has trained over 150 companies and their corporate training programmes are well-suited to senior leaders who are tasked with building teams and piloting several PoCs within their organisation 5) With the demand for data scientist-geared courses going up, Jigsaw Academy has successfully pioneered a holistic approach to serve the needs of learners who come with varying educational goals. For example, the Academy provides training in 2 modes – online and offline. In the offline or inperson mode, learners can take up the long-term University of Chicago Graham School certified PGPDM or the more intensive 3 month Data Science Bootcamp. In the online mode, students can choose the IIM Indore certified IPBA with live online classes featuring faculty from IIM Indore and Jigsaw Academy or Full Stack Data Science program which has been ranked as #1 by AIM. Conclusion The emerging domain of data science is no longer a single free-standing field, it has evolved into an interdisciplinary effort that requires a broader understanding of the ecosystem. It requires a fundamental understanding of the core methods which include statistical, mathematical and computational understanding for working with data. In this space, training institutes that can address the profound breadth of data science and also provide a solid training in the emerging applications and business problems will come out as leaders in this field. As per a report, the online education market in India is expected to grow by $1.96 billion over the next 5 years. Also, the paid user base is expected to increase to $9.6 million by 2021. This implies there is a huge play for byte-sized Instructor-Led online courses – an area that analytics training institutes can capitalise on. While we know there is no magic bullet to this challenge, working in conjunction with educational stakeholders, industry players and policymakers are likely the best solution for reskilling the workforce. Download the report here: [slideshare id=149138617&doc=addressingreskillingchallengesinindia-190610061413&type=d]","excerpt":"Even with the third-largest developer base and a substantial tech-savvy talent pool, India lags behind its peers on major AI indicators. This is despite a thriving startup ecosystem, high-growth companies which have made a substantial investment in setting up CoEs and the Government investing in building a robust tech infrastructure. Behind the AI and data […]","categories":["AI Features"],"tags":["data science ai bootcamp","industries transformed by big data"],"author_name":"Richa Bhatia","publish_date":"2019-06-05T11:06:20","publication_year":"2019","word_count":2831,"keywords":["data science","machine learning","artificial intelligence","Keras","AI","ML","RAG","data science ai bootcamp","Aim","analytics","TensorFlow","industries transformed by big data"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","data science","analytics","Aim","TensorFlow","Keras","RAG"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/addressing-indias-reskilling-challenge-a-report-by-aim\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10054355,"title":"What To Expect From AWS re:Invent 2021 Event?","content":"The AWS re:Invent 2021 event is scheduled between November 29 and December 3, 2021. This is one of the most anticipated events of the year. AWS has truly emerged as a cash cow for Amazon. It is a $64 billion revenue run rate business that sees a 39 per cent year-on-year growth. AWS’ growth rate had accelerated from 29 per cent in 2020. Its sustained financial strength and continued growth of the cloud hyperscale market on the whole has allowed parent company Amazon to exert influence in building new business and disrupting markets. Considering these factors, it is no doubt that AWS’ event is one of the most anticipated events of the year. Based on the announcement, and trends in the cloud industry, among other factors, we predict what will happen at the event. Storage In the previous year, AWS made three major announcements for EBS, the block-storage service to be used with EC2. Further, AWS also announced major improvements to S3, including the inclusion of strong read-after-write consistency for all major applications, replication with multiple destination buckets, and new bucket keys. This year also, one can expect major updates and launches in this space. Notably, in October, AWS released a research paper about the operations of S3. In the paper titled “Using Lightweight Formal Methods to Validate a Key-Value Storage Node in Amazon S3,” AWS also spoke about implementing ShardStore, which is a new key-value storage node implementation for S3 cloud object storage devices. Machine Learning At the AWS re:Invent 2020 event, machine learning became a major theme. In particular, AWS vice president for AI and machine learning Swami Sivasubramanian focused his keynote around SageMaker, introducing the new Feature Store — a fully managed, purpose-built repository for storing and updating, retrieving and sharing ML features. Services like SageMaker Clarify, SageMaker Debugger and SageMaker Model Parallelism were also introduced. In the past few months, AWS has introduced many new features for SageMaker. In August, Amazon announced that users will now be able to deploy ML model packages and Edge Manager agent to edge devices easily. It has been made possible by making SageMaker Edge Manager integrable with AWS IoT Greengrass. Before this, the Edge Manager agent was deployable to devices by manually copying the Edge Manager. As companies race towards offering advanced solutions that are optimised and accelerated for AI applications, AWS jumped the bandwagon and introduced a new competitor to the landscape. It announced cloud instances powered by Intel-owned Habana’s Gaudi AI processors. It is the first AI training instance provided by AWS that is not GPU-based. More details\/updates may be provided in a similar direction at the AWS re:Invent 2021 event. Before that, AWS also added SageMaker Clarify, an Amazon tool to detect algorithmic bias and to increase ML model transparency. As awareness about Responsible AI grows in the industry, AWS may announce more such features in the upcoming event. Networks As per the latest development from AWS, customers can now create IPv6-only virtual private cloud networks. The company claims that it is a monumental step towards enabling IPv6 on its cloud. The new feature will allow admins to create an IPv6-only subnet with dual-stack VPC. According to AWS, each subnet has a \/64 Classless Inter-Domain Routing range, offering “approximately ten quintillion IP addresses for applications.” Quantum Computing Quantum computing has become a major technology theme in recent years. AWS is not far behind to leverage it either. In October, AWS’ quantum computing research unit moved to the California Institute of Technology (Caltech). The new facility will work to build a fault-tolerant quantum computer. In the past, too, AWS has published research papers in this field. The new centre will develop and test new quantum processors and also work on technologies like cryogenic cooling systems and other supporting technologies. Expert Comments We spoke to a few experts in the field to understand their expectations from the event. “The pandemic has made digital transformation move at a MACH speed (Microservices, API, Cloud, Hybrid). We expect AWS to keep the focus on these topics and, at the same time, update the security and audit tools. AWS has a lot of tools in its arsenal, and we expect more business process automation (like RPA) from them. DevOps will always take centre stage, and this time we would expect updates to their SageMaker ML platform. It is great that AWS re:Invent has added online events – so this is one thing in Vegas that won’t stay just in Vegas,” said Kanchan Ray, CTO, Nagarro. “Quite honestly, I don’t expect a lot from re:Invent 2021. The entire global cloud play doesn’t really keep in mind the Indian requirement of the cloud for a large scale of individual MSME needs. A country as native as India in terms of its need of cloud solutions (online storage) needs a completely customised solution & year after year, we haven’t really found one,” said Arnab Mitra, CEO, Digiboxx. “We would be keen to know more from AWS and how it can help startups like us in working and scaling towards creating the metaverse. We envision such partnerships as a game-changer for organisations like us. For the past couple of years, we at XR Central have been working towards democratising how metaverses have been looked at. Such an infrastructure service will enable us to accelerate the process and bring metaverses to the masses. We dream of everyone having their connected metaverse,” said Shrey Mishra, Co-Founder, XR Central. On a lighter note: Any predictions on new #AWS services being launched at #reinvent? The cloud sharks are placing their bets next week! https:\/\/t.co\/f9RR4HeVbn pic.twitter.com\/dEUUTNLg6j— Drew Firment (@drewfirment) November 11, 2021","excerpt":"Based on the announcement, and trends in the cloud industry, among other factors, we predict what will happen at the AWS re:Invent event.","categories":["IT Services"],"tags":["AWS","AWS services"],"author_name":"Shraddha Goled","publish_date":"2021-11-27T18:00:00","publication_year":"2021","word_count":940,"keywords":["Go","machine learning","AWS services","AWS","AI","ML","RAG","microservices","Ray","Aim","R"],"extracted_tech_keywords":["AI","machine learning","ML","Aim","Ray","RAG","AWS","microservices","R","Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/what-to-expect-from-aws-reinvent-2021-event\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10023203,"title":"GPT-Neo: The Open-Source Cure For GPT-3 FOMO","content":"GPT-3 was the largest language model when OpenAI released it last year. Now, Google Brain’s 1.6 trillion parameters language model (as opposed to GPT-3’s 175 billion parameters) has replaced GPT-3 as the largest language model. Both GPT-3 and Google Brain model are transformer-based neural networks. Developers across the world were waiting for the arrival of GPT-3 with bated breath. But much to their disappointment, and a departure from their earlier signals, OpenAI exclusively sold the source code of GPT-3 to Microsoft. Interestingly, GPT and GPT-2 were open-source projects. Ever since, the internet hive mind, developers and researchers ached for an open-source version of GPT-3. Now, it looks like their dream has finally come true. Enter GPT-Neo, the brainchild of EleutherAI. Connor Leahy, Leo Gao, and Sid Black founded EleutherAI in July 2020. It is a decentralized grassroots collective of volunteer developers, engineers, and researchers focused on AI alignment, scaling, and open-source AI research. According to EleutherAI’s website, GPT‑Neo is the code name for a family of transformer-based language models loosely styled around the GPT architecture. The stated goal of the project is to replicate a GPT‑3 DaVinci-sized model and open-source it to the public, for free. GPT‑Neo is an implementation of model & data-parallel GPT‑2 and GPT‑3-like models, utilizing Mesh Tensorflow for distributed support. The codebase is optimized for TPUs, but also work on GPUs. Interestingly, Leahy had earlier attempted to replicate GPT-2 through Google’s Tensorflow Research Cloud (TFRC) program, which worked out to their advantage while working on GPT-Neo. The researchers have now announced the release of two mid-sized models in their GPT-Neo library, pre-trained using 1.3 billion and 2.7 billion parameters, a far cry from GPT-3’s 175 billion parameter but in the ballpark of GPT-2 with 1.5 billion parameters. An EleutherAI researcher tweeted: https:\/\/twitter.com\/arankomatsuzaki\/status\/1373732645444579331 We're also planning to release the weights (~10B) in Jax soon and eventually the full GPT-3 in PyTorch.— Aran Komatsuzaki (@arankomatsuzaki) March 21, 2021 GPT-Neo: A GPT-3-Sized Model The researchers are working on two repositories — GPT-Neo (for training on TPUs) and GPT-NeoX (for training on GPUs) The original codebases for GPT-Neo were built on TPUs, Google’s custom AI accelerator chips. The EleutherAI team realised even the generous amount of TPUs provided through TFRC wouldn’t be sufficient to train GPT-Neo. The research work got a huge boost when CoreWeave, a US-based cryptocurrency miner, approached EleutherAI. The former offered EleutherAI team access to its hardware in exchange for an open-source GPT-3-like model. No money changed hands as part of the deal. While the model released is quite similar to GPT-3, it will differ in terms of the training dataset, refined by extensive bias analysis. The dataset, called The Pile, is an 835GB corpus consisting of 22 smaller datasets combined to ensure broad generalisation abilities. GPT-Neo was trained on The Pile, the weights and configs of which can be freely downloaded here. With the GPT-Neo implementation, researchers were able to make GPT-2 and GPT-3-like models while scaling them up to GPT3 sizes and more, using the mesh-TensorFlow library. It includes alternative model architectures and linear attention implementations that should enable scaling up to even larger model sizes & context lengths, including Local attention, Linear attention, Mixture of Experts, Axial Positional embedding and Masked Language Modelling. Way Forward Researchers are now focusing on the development of the GPT-NeoX library using the hardware and compute-capabilities provided by Coreweave. EleutherAI is currently waiting for CoreWeave to finish building the final hardware for training. GPT-Neox will offer features such as 3D parallelism, model structuring and straightforward configuration using various files. GPT-NeoX is under active development and will be based on the DeepSpeed library. It is designed to be able to train models in the hundreds of billions of parameters or larger.","excerpt":"GPT-3 was the largest language model when OpenAI released it last year. Now, Google Brain’s 1.6 trillion parameters language model (as opposed to GPT-3’s 175 billion parameters) has replaced GPT-3 as the largest language model. Both GPT-3 and Google Brain model are transformer-based neural networks. Developers across the world were waiting for the arrival of […]","categories":["AI Trends"],"tags":["EleutherAI","GPT-3"],"author_name":"Srishti Deoras","publish_date":"2021-04-04T16:00:00","publication_year":"2021","word_count":621,"keywords":["GPT-3","Go","TPU","OpenAI","AI","neural network","PyTorch","GPT","JAX","EleutherAI","TensorFlow","R"],"extracted_tech_keywords":["AI","neural network","OpenAI","TensorFlow","PyTorch","JAX","TPU","R","Go","GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/gpt-neo-the-open-source-cure-for-gpt-3-fomo\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10005704,"title":"How upGrad Managed To Witness A 40% Rise In Data Science Enrolments Amid Lockdown","content":"While the concept of online learning has been prevalent in the field of data science, the recently enforced lockdown has provided new momentum to the whole ed-tech space. The data professionals across industries are now looking at online learning programs not only to stay updated with their skills and knowledge but also to keep up their relevance post the COVID crisis. This has led to a spike in the number of learners on edtech platforms. In fact, upGrad, India’s largest online higher education company, has witnessed a 40% rise in data science enrolments amid this lockdown. This surge can be attributed to the fact that despite the challenges due to COVID, the demand for data science and analytics across industries has increased. College students and fresh graduates are thus noticing more potential in this field and therefore looking to pursue data science as a career. While many short-term certifications can help data scientists to upskill themselves, what sets upGrad apart are the industry-relevant data science programs offered by them, along with the alumni status from respective universities, industry knowledge via projects, domain-specific capstone project, WES recognition of the program and hands-on experience with programming languages and tools with specialised tracks. As a matter of fact, upGrad’s collaboration with IIIT-B to develop the PG Diploma in Data Science has become the first PG diploma program that has been validated and recommended by NASSCOM Futureskills, which not only complies to all industry standards but also makes learners job-ready from day one. Led by India’s leading data science faculty and industry leaders, this program provides industry-leading content in the form of videos, case studies, live lectures and coaching sessions with best-in-class career outcomes. Key Benefits Of upGrad’s Flagship PG Diploma & M.Sc. Program in Data Science: It offers a PG Diploma in Data Science from IIIT Bangalore and an M.Sc. from Liverpool John Moores University, with alumni status from the respective universities, allowing learners to connect to a global network of accomplished alumni. Post successful completion of the NASSCOM-certified PG Diploma in Data Science from IIITB, learners can choose to articulate themselves to a Master’s Degree in Data Science from LJMU, at 1\/10th the cost of an equivalent on-campus program. Learners can also directly enroll themselves into M.Sc. in Data Science where the learners first complete the diploma and post that complete their Master’s from LJMU. Both the programs have five functional specialisation options to choose from as per the learners’ career aspirations — deep learning; natural language processing; business intelligence and data analytics; business analytics; data engineering. Receive 360-degree career support and mock interviews with hiring managers. Get dedicated one-on-one industry expert mentorship sessions for proactive support. Hands-on 20+ industry projects. The master’s program offers a dissertation project opening up avenues for research and PhD for learners. Along with the Master’s and PG Diploma in Data Science which require 18 and 12 months of commitment, respectively, upGrad now offers Career Services for 7 months of PG Certification in Data Science for the professionals who are in a dilemma to make a long term commitment to indulge in certification programs. 360 Degree Career Services by upGrad include: Dedicated career mentor for continuous productive feedbackOne-on-one sessions with industry experts for better understandingLive profile building workshops including the Git-hub profileResume review sessionsOpportunity to access upGrad’s job portal which has 100+ openings at all timeLast-minute interview preparation from expertscompany-specific preparation for interviewsHackathons with hiring partners with the job opportunities  Ability to access upGrad elevate — a virtual hiring drive Along with these, upGrad is also offering a per month pricing option: ₹7,583 for PG Diploma in Data Science ₹13,125 for M.Sc. in Data Science Outlook With such benefits in hand, upGrad has been one of the preferred platforms for its data science programs. Targeting working professionals along with college graduates, upGrad’s online data science programs will help learners to kickstart their data science journey at a fraction of the cost compared to other similar courses. Enrol now to avail this best in the industry learning program – one that ensures consistent progress among data scientists.","excerpt":"While the concept of online learning has been prevalent in the field of data science, the recently enforced lockdown has provided new momentum to the whole ed-tech space. The data professionals across industries are now looking at online learning programs not only to stay updated with their skills and knowledge but also to keep up […]","categories":["AI Trends"],"tags":["Data Analytics Certification","data analytics masters","mentorship","UpGrad","UpGrad India"],"author_name":"Sejuti Das","publish_date":"2020-08-27T15:31:22","publication_year":"2020","word_count":679,"keywords":["business intelligence","data science","data analytics masters","programming_languages:R","AI","Git","deep learning","data engineering","UpGrad","analytics","mentorship","Data Analytics Certification","R","UpGrad India"],"extracted_tech_keywords":["AI","deep learning","data science","analytics","R","Git","data engineering","business intelligence","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/how-upgrad-managed-to-witness-a-40-rise-in-data-science-enrolments-amid-lockdown\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10121914,"title":"Top 8 Microsoft 365 Copilot Alternatives to Create PPTs in Seconds","content":"Microsoft introduced Copilot, integrating it into Microsoft 365 to provide users with more agency and enhance accessibility using natural language processing. Copilot is now within familiar Office suite apps like Word, Excel, PowerPoint, Outlook, and Teams. With PowerPoint, in particular, Copilot acts as an AI assistant, simplifying presentation creation by effortlessly transforming ideas into polished slides. It streamlines tasks such as generating drafts, distilling complex content, organising slides, and applying brand styles. Users can explore different ideas and formats, refining their presentation skills along the way. By leveraging Copilot’s intuitive features and feedback, presenters can save time. Whether for work, school, or personal use, Copilot empowers users to deliver impactful presentations that captivate audiences and effectively convey messages. Given that they are now for sale, here are brief reviews of the Microsoft Copilot Pro Apps I have tried:Outlook: This is the slickest of the Copilots in terms of deep integration into the core application, and in many ways is the most obvious use case. It basically lets GPT-4… pic.twitter.com\/F81sINsBek— Ethan Mollick (@emollick) January 16, 2024 However, Copilot isn’t the first to leverage its PowerPoint capabilities. Here are 8 alternative online AI tools to create your PowerPoint presentations in seconds. Top 8 Microsoft Copilot Alternatives to Create PPTs in Seconds ToolsBest forPricePopAiChat with document & Image Chat$49.99\/yearBeautiful.aiGenerate new feature$40\/monthDecktopusAuto-created deck for presentations$34.99\/monthTomeDynamic text editing & image generation on text $16\/monthSlideSpeakOne-Click polish tool$19\/monthGammaDesign lock-in $15\/monthPlus.aigenerate slideshow from scratch with versatile templates$20\/monthSlidesAIDesign Intelligence₹832.83\/month PopAi PopAi is an AI tool that offers versatile conversational experiences, supporting over 200 languages. It caters to both personal and professional needs, adapting to educational queries, technical support, and creative idea generation. PopAi introduces innovative features like “Chat with Document” for instant insights from documents, “AI Presentation” for efficient presentation creation, and “Image Chat” for visual understanding. By leveraging AI presentation tools, users can save valuable time and increase productivity. These tools automate various aspects of presentation creation, such as design suggestions and content generation. With AI chat, users no longer need extensive design knowledge or expertise to create visually appealing slides. 2. Pop AiYour Personal AI WorkspaceYou can:– Create presentations– Craft CVs and resumes– Write academic essays– Design flowcharts– Write blogs– Debug code, and much more. pic.twitter.com\/8XELh9Oy5W— D-Coder (@Damn_coder) May 28, 2024 Beautiful.ai Beautiful.ai is a web-based tool that helps one create stunning presentations in minutes using AI to design slides based on content and preferences. It handles fonts, colours, layouts, and animations. One can collaborate with a team in real time and convert presentations to .pdf or .ppts formats for sharing online or offline. Beautiful.ai works on any device and browser, allowing one to create and present from anywhere. It also integrates with Slack. With DesignerBot, one can quickly design slides and benefit from helpful brainstorming, instant text, and image generation. Decktopus Decktopus is a simple and intuitive tool that helps to create presentations online. It automatically adjusts a presentation to ensure it is perfect, aligning to text, images, icons, and colours harmoniously for an appealing design. Decktopus is a smart assistant that helps to create effective and engaging presentations. It enables one to create visually stunning and professional presentations by offering different themes and styles to suit your purpose and audience. One of the standout features of Decktopus is its AI-enriched content. The tool provides image and icon suggestions, slide notes, and more content ideas based on topic and audience. One can also add voice recordings, videos, URLs, and other multimedia elements to enhance the presentation. Tome Tome AI is an innovative tool that helps users quickly and easily create presentations and other narrative content. By using prompts or existing documents as input, Tome generates visuals, layouts, and text suggestions to build professional-looking presentations. Users can generate presentations simply by providing a prompt, and the output is organized by a table of contents, complete with text, introduction slides, and AI-generated images. Tome’s presentations have a distinctive style, typically featuring a black background, white text, and AI illustrations, setting them apart from traditional PowerPoint or Google Slides presentations. Tome AITired of spending hours making slides?Generate your presentations in seconds. pic.twitter.com\/rynggzMiSe— Paul Couvert (@itsPaulAi) May 9, 2023 SlideSpeak SlideSpeak revolutionises presentation creation by allowing users to upload PDFs or Word documents and automatically generate presentations based on the content. With just one click, one can transform the document into a presentation without needing to interact extensively with an AI bot. The generated presentation can then be downloaded as a PowerPoint file (.pptx), where one can fix any misaligned text or images and make further edits as needed. Additionally, SlideSpeak enables one to share and download the presentation in both formats, providing a seamless and efficient way to create and distribute professional presentations. Gamma Gamma is an AI-powered tool designed for creating professional presentations, websites, and documents. With Gamma, design lock-in is no longer an issue. It accelerates content creation by generating templates that automatically align with the brand. Gamma supports various media formats, making it easy to include GIFs, videos, charts, or websites. This enhances presentations and helps convey complex ideas more effectively. The media integration feature ensures that your content is dynamic and engaging. Ideal for large teams, Gamma allows for real-time collaboration, enabling instant feedback and collective efforts, all within a single platform. This feature caters to the needs of big teams and ensures seamless teamwork. https:\/\/twitter.com\/rowancheung\/status\/1644725291472891904 Plus.ai Plus.ai is an AI tool that can be integrated into Google Docs and Slides to generate custom content quickly and effortlessly. It prioritises professional designs, ensuring that presentations are suitable for both professional and academic contexts. Additionally, the AI copilot functionality facilitates collaborative presentation creation by seamlessly integrating AI throughout the process. The Live Snapshots feature, powered by Plus’s Snapshot technology, automates data updates, ensuring that information remains current. Moreover, Plus AI focuses on content quality, generating an appropriate amount of text for each slide and demonstrating an advanced understanding of various slide layouts. Furthermore, the rewrite feature allows users to quickly rectify inaccuracies in content with AI assistance. SlidesAI SlidesAI is seamlessly integrated into Google Slides, offering users the ability to utilise generative AI directly within the platform. It was initially launched with the capability to generate presentations from lengthy text documents. SlidesAI has recently expanded its functionality to include the creation of presentations using shorter prompts as well. Alongside these primary features, SlidesAI also provides users with additional tools such as image suggestions tailored to specific slides, text paraphrasing options for refining content, and a text-to-slides feature that enables users to effortlessly convert existing text into presentation slides through simple copy-paste actions.","excerpt":"PopAi, Beautiful.ai, Decktopus, Tome & Plus.ai are some of the top AI tools for creating PowerPoint presentations online","categories":["AI Trends"],"tags":["AI","Copilot","Microsoft 365"],"author_name":"Gopika Raj","publish_date":"2024-05-29T12:04:56","publication_year":"2024","word_count":1096,"keywords":["Go","TPU","AI","ML","RAG","GPT","Microsoft 365","generative AI","GAN","copilots","Copilot","R"],"extracted_tech_keywords":["AI","ML","generative AI","RAG","copilots","TPU","R","Go","GPT","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/8-online-ai-tools-for-creating-ppts-in-seconds\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10084253,"title":"ChatGPT Hits Ethical Roadblock; Blocked","content":"The 40th International Conference on Machine Learning (ICML) has called for papers on all topics related to machine learning for the main conference proceedings. But if you are thinking of using ChatGPT, think again. Excerpt from the ICML website One of the conditions that ICML 2023 website, under the ethics section, cites as the requirement for the submissions is, “Papers that include text generated from a large-scale language model (LLM) such as ChatGPT are prohibited unless this produced text is presented as a part of the paper’s experimental analysis.” Further, ICML, in its call for papers, does mention that “all suspected unethical behaviour will be investigated by an ethics board and individuals found violating the rules may face sanctions”. But it is not just the ICML website, there are a host of sites that have banned the popular chatbot. On Wednesday, the New York City Department of Education began blocking access to the chatbot on school devices and networks, citing negative impact on student learning, and concerns regarding the safety and accuracy of the content. It also argued that “while the tool may be able to provide quick and easy answers to questions, it does not build critical-thinking and problem-solving skills, which are essential for academic and lifelong success.” However schools and students who wish to use the tool to study the role and impact of AI, can still request access through the department. The NYC education dept was the first to block ChatGPT, and coming from the United State’s largest school system, the move could be reflective of the beginning of similar bans elsewhere. Microsoft, at a recent event, too acknowledged this, calling it both good and bad – most likely unfortunate – as students are using it to write essays. While ChatGPT is making it easier for students to complete projects and assignments, there has been a backlash from the school management, raising concerns around the effectiveness of learning and creativity. ChatGPT ban continues Previously, the chatbot’s usage was temporarily banned by Stack Overflow, citing a high degree of inaccuracy in the bot’s responses. This was later followed by IT and cyber security organisation Offensive Security, which banned the chatbot in its certification exams. “Interesting to see ChatGPT being banned from Offensive Security exams as I was headed to give the OSEP soon,” said Dhruva Goyal, founder at BugBase.in, questioning the logic behind this. A week ago ChatGPT was also removed from WeChat, the popular Chinese social media app developed by Tencent. This is because despite the bot being not officially available in the country, local developers rushed to provide unauthorised access to the service. While this appears to be the start of organisations and institutions banning the use of ChatGPT, both temporarily and permanently, there is a rising concern of how one can identify content produced by AI. Enters GPTZero Among the efforts to identify content generated by AI or ChatGPT-like platforms, is GPTZero. This has been developed by Edward Tian, a Princeton student, alongside Sreejan Kumar, that prompts when a particular text is written by ChatGPT and other generative AI engines. “I spent New Years building GPTZero – an app that can quickly and efficiently detect whether an essay is ChatGPT or human written,” said Tian. The app processes submitted text for indicators of AI origins like randomness and complexity in how it is written, technically referred to as “perplexity and burstiness”. In his Twitter thread, Tian tested GPTZero on various posts published by companies on various social media platforms like Linkedin and Twitter, to check its effectiveness. However, the app does not have enough data to measure accuracy yet, but it is a good start, as it is still a work in progress. Soon, the duo looks to publish a paper soon. GPTZero is not the only such attempt to identify AI-generated text. Meanwhile, OpenAI is also looking at addressing the concerns of ChatGPT, where it is looking to build identification features like watermarking and other features similar to its other AI generative tool DALL.E. Lastly, it is also interesting to see more and more apps and tools such as GPTZero cropping up in the coming months, to tackle misuse of ChatGPT for unethical reasons, like submitting research papers and generating codes and essays for examination among others.","excerpt":"The usage of the chatbot has already been banned by New York’s education department, ICML, Stack Overflow, Offensive Security and WeChat","categories":["AI Features"],"tags":["ban","ChatGPT","ICML","Stack Overflow"],"author_name":"Aparna Iyer","publish_date":"2023-01-06T16:00:00","publication_year":"2023","word_count":713,"keywords":["Go","ChatGPT","machine learning","OpenAI","AI","ML","Stack Overflow","ICML","GPT","generative AI","GAN","ban","R"],"extracted_tech_keywords":["AI","machine learning","ML","generative AI","ChatGPT","OpenAI","R","Go","GPT","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/chatgpt-hits-ethical-roadblock-blocked\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10056057,"title":"Winter Session To Witness The Controversial Chills Of The Personal Data Protection Bill","content":"The Indian Personal Data Protection Bill is expected to be taken up for a much-awaited discussion in the winter session of the Parliament, after almost two years of deliberations. But a few of the provisions have come under great scrutiny by dissenters and opposition parties across the country, leaving the future of the PDP bill a question mark. To understand the criticism, let us look at the state of surveillance in the country in the past two years with the increased use of drone technology to monitor the citizens. The Delhi police did this during the assembly elections and the riots following it in February 2020, the protests during the foundation of Ayodhya’s Ram Mandir and the pandemic lockdown. Not just the Delhi police, police across states like Mumbai, Punjab, Hyderabad, Kerala and Chennai had relied on drones to survey citizens during the lockdowns. Some have continued this use even after the suspension of the lockdown. While drone usage is regulated under UAS rules, the states asked for exemption from the law, making drone surveillance more contractual than a legal safeguard for data protection. The proposed personal data protection bill might have similar effects. The History Countries across the globe have introduced their data protection laws after the EU GDPR that came into effect in 2018. As of now, these safeguards are governed by the Information Technology Act, 2000, which was amended to include Section 43A and 72A that give a right to compensation for improper disclosure of personal information. India’s long journey to an independent bill began in 2019, with a joint parliamentary committee drafting the bill’s report that sought to provide protection of personal data and establish a data protection authority. The bill has been tenured six times since. One of the most awaited bills, the 250-page final report, is likely to be tabled on 21 December 2021, during the winter session of the Parliament. But some provisions have already come under criticism, with seven of the 30 members showing concerns. The Bill’s Recommendations The PDP Bill requires prior consent to the use of individual data, limits the list of purposes for which companies can process data, and restricts data collection to only the data necessary for providing a service. Companies will further need to localise their data and appoint data protection officers. The NPCs recommendations also widen the bill’s scope to cover not only personal data but also non-personal data for unauthorised breaches. Essentially, it restricts non-consented acquisition, usage, sharing, alteration or destruction of either type of data since it will compromise the confidentiality and availability of data. In addition, the bill covers accidental breaches as well, meaning that it will also cover the company’s lack of compliance or ignorance. The bill recommends social media firms should be treated as publishers, making them liable for the content spoken or distributed on their platforms. They will also need to set up offices in India to function in the country. The bill further proposes data fiduciary and data processor, referring to a controller and processor. This means that the bill will apply to Indians in the country but also foreign individuals conducting business in\/with India. The point of the bill that has come under most scrutiny is its power to the government to exempt its probe agencies from the provisions of the Act. It excludes the government and its agencies like the CBI and UIDAI from national and public welfare security provisions. Criticism from Opposition Responding to the inclusion of non-personal data, lack of data localisation and social media while exempting the government from the purview, opposition MLAs have filed dissent notes. The TMC tweeted saying Section 35 of the Act will give the “government exclusive rights to invade our privacy whenever they want”. Congress MP Manish Tewari has objected to the entire Bill through a 10-page dissent note that states that the bill’s construction has an ‘inherent design flaw’. “The bill as it stands creates two parallel universes – one for the private sector where it would apply with full rigour and one for government where it is riddled with exemptions, carve-outs and escape clauses. In my limited experience of three decades as a litigator, I have always been taught and made to appreciate that a Fundamental Right is principally enforceable against the State. A bill that seeks, therefore, to provide blanket exemptions either in perpetuity or even for a limited period to the ‘State’ and its instrumentalities, in my estimation, is ultra vires of the Fundamental Right to Privacy as laid down by a 9-judge bench of the Supreme Court of India in Re Puttuswamy (2017) 10 SCC 1,” Tewari argued. In fact, even Justice BN Srikrishna, who had worked on the bill’s first draft in 2018, called the bill “Orwellian”. Social Media and Silencing Mishi Choudhary, a technology lawyer and online civil liberties activist, deemed India the only democracy overlooking the State’s surveillance power. With the power of the Aadhar and the government, the personal data will be chained together to follow us through “every fiscal and administrative transaction”. A lack of check on this government power is surely a concern. The second stretch of criticism comes towards the social media related recommendations. India’s history looks at punishing social media companies or their executives for controlling the freedom of speech on the platforms. Just in July, the Indian government and Twitter were at a ‘war’ with the Intermediary Guidelines and Digital Media Ethics Code rule that demanded social media platforms to censor speech as per the terms of the rules. The JPC report recommends criminal punishment for such offences under the law, essentially giving the government control to threaten Twitter executives with jail time for non-compliance. Additionally, putting the burden of user-generated content on the social media platforms by deeming them as publishers will make platforms liable for any inappropriate content. This will involuntarily make platforms the arbiters of online speech and lead to over censorship. Second, this might, along with the compulsion to set up offices in India, increase the uncertainty for social media platforms and hurt the growth of the digital economy. Opposition is not it Contrary to the dissent, Salil Parekh, the CEO of Infosys, has a more positive approach to the bill’s demands. According to him, the laws are not a cost to the enterprise but critical building blocks of the technological future of India. “If you see data regulation guidelines across different jurisdictions, the main objective is to protect individual privacy and protect some level of intellectual property,” he said at the Carnegie India Global Technology Summit 2021. “I think in India also, regulations will be along those lines. It’s not to be viewed as a critical element.” The Indian government is not known for its track record of maintaining the citizens’ personal privacy with instances such as the Pegasus spyware project or WhatsApp suing them over internet laws that could lead to mass surveillance of its 400 million Indian users. But the genuineness of the citizen’s data protection will only unfold with the winter session of the Parliament.","excerpt":"Even Justice BN Srikrishna, who had worked on the bill’s first draft in 2018, called the bill “Orwellian”.","categories":["AI Features"],"tags":[],"author_name":"Avi Gopani","publish_date":"2021-12-16T18:00:00","publication_year":"2021","word_count":1178,"keywords":["Go","AWS","AI","cloud_platforms:AWS","ML","Git","BERT","llm_models:BERT","GAN","R"],"extracted_tech_keywords":["AI","ML","AWS","R","Go","Git","BERT","GAN","llm_models:BERT","cloud_platforms:AWS"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/winter-session-to-witness-the-controversial-chills-of-the-personal-data-protection-bill\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10018528,"title":"Complete Guide To AutoGL -The Latest AutoML Framework For Graph Datasets","content":"Creating algorithms is difficult and time-consuming. This specific problem has inspired researchers to develop some productivity tools to help young members in this domain. This has given birth to a revolutionary field in Data Science called Auto Machine Learning(AutoML). AutoML provides methods and processes to make Machine Learning available to non-Machine Learning experts, to improve efficiency of Machine Learning and speed up the research. On the other hand, all the existing AutoML systems cannot be applied to graph datasets. This made the researchers of Tsinghua University to develop a new AutoML framework called Auto Graph Learning(AutoGL). AutoGL is an AutoML framework that can be used on graph datasets and tasks. This toolkit handles various stages as shown in Figure below. You can read more about it, here. Without further ado, let’s jump into the quick tutorial. Installation Requirements Make sure to meet the requirements before installing the library. Python >= 3.6.02. PyTorch (>=1.5.1) PyTorch Geometricsee import sys #to print the current python version print(sys.version) !pip install torch !pip install torch-geometric Install from pip Might take some time to install. !pip install auto-graph-learning 2) Learning AutoGL is based on the concept of AutoML, we need to provide only dataset and tasks to be done and AutoGL Solver will do all the wonder for us from feature engineering to ensembling the model(as suggested in the diagram above). The whole mechanism can be summarised as : AutoGL contains a dataset to maintain graph datasets given by the user then we create an AutoGL solver object to define the task. AutoSolver includes feature engineering, auto model, hyperparameter optimization and auto ensemble, which automatically preprocess the data, choose the best model, optimize and ensemble in the best way. For example, we are applying AutoGL on Cora dataset. First let’s fetch the cora data set from the datasets module of AutoGL. #We can easily connect fetch the using datasets module from autogl.datasets import build_dataset_from_name #fetching cora dataset cora_dataset = build_dataset_from_name('cora') Now, creating an AutoGL solver object to define the required task. #Import the required libraries import torch #import the AutoNodeClassifier method from solver module to make node #classification solver for handling auto training process. from autogl.solver import AutoNodeClassifier #take device as 'cuda' if available else 'cpu' device = torch.device('cuda' if torch.cuda.is_available() else 'cpu') #creating a solver object defining 'deepgl' as feature engineering model, #['gcn','gat'] as graph models , 'anneal' hyperparameter model and 'voting' as ensemble method #and lastly giving the device. solver = AutoNodeClassifier( feature_module='deepgl', graph_models=['gcn', 'gat'], hpo_module='anneal', ensemble_module='voting', device=device ) Fitting the model to the cora dataset. #fitting the above method solver object to cora_dataset #time_limit is required so as to make sure the whole auto graph process #wont exceed 1 hour limit #you can exclusively define train and test datasets. solver.fit(cora_dataset, time_limit=3600) #solver.show() represents the model present in solver object and its #performance on validation dataset. solver.get_leaderboard().show() Now, we will predict and evaluate using the evaluation functions. Here, we don’t need to pass cora_dataset again as it is already saved in the solver object and it will be reused when no dataset is passed at prediction but we can always new dataset while predicting and it will consider the new dataset instead of older one. #Importing the accuracy method from train module from autogl.module.train import Acc #predicting the probabilities of cora_dataset #if not specifically mentioned, solver will consider the #cora_dataset on predicting. You can also mention different dataset. predicted = solver.predict() print('Test accuracy: ', Acc.evaluate(predicted.reshape(-1,1), cora_dataset.data.y[cora_dataset.data.test_mask].numpy().reshape(-1,1))) 3) Dataset This dataset can import datasets from CogDL, Pytorch Geometric and from OGB. You can refer to these links for creating and building datasets. You can check all the datasets supported by AutoGL, here.  If your dataset contains two matrices, groups and networks then you can directly register that dataset through url, please refer here. You can also create a local dataset for testing. For this, please refer here. 4) Feature Engineering There are many engineering pipelines provided by AutoGL toolkit for nodes and subgraphs. Along with that, it provides an automatic feature engineering pipeline. An example of how we can apply feature Engineering is given below. Majorly, there are three kinds of feature engineering atom names supported by AutoGL toolkit. Here is a list of all the selectors, generators and subgraphs. Apart from that you can also create your own feature engineering object, please refer to this. # 1. Choose a dataset. from autogl.datasets import build_dataset_from_name data = build_dataset_from_name('cora') # 2. Compose a feature engineering pipeline from autogl.module.feature import BaseFeatureAtom,AutoFeatureEngineer from autogl.module.feature.generators import GeEigen from autogl.module.feature.selectors import SeGBDT from autogl.module.feature.subgraph import SgNetLSD # you may compose feature engineering atoms through BaseFeatureAtom.compose fe = BaseFeatureAtom.compose([ GeEigen(size=32) , SeGBDT(fixlen=100), SgNetLSD() ]) # or just through '&' operator fe = fe & AutoFeatureEngineer(fixlen=200,max_epoch=3) # 3. Fit and transform the data fe.fit(data) data1=fe.transform(data,inplace=False) print(data1.data) 5) Model You can check out all the models available in this toolkit, here. Apart from that, you can create your own model and automodel. An example of this is given here. 6) Trainer AutoGL Trainer handles all the auto training of tasks. There are two type of trainer supported by it namely NodeClassificationTrainer for semi-supervised node classification GraphClassificationTrainer for supervised graph classification You can build a trainer with the help of this example, here. After the initialization of the trainer, we can train and test. Till now, training and testing is available for node classification and graph classification and after testing, you can also evaluate the prediction through different metrics available. 7) Hyper Parameter Optimization This toolkit used black-box hyperparameter optimization. Here is the list of all the algorithms supported by it. Apart from that you can create your own Hyper Parameter Optimizer. Source: https:\/\/autogl.readthedocs.io\/en\/latest\/docfile\/tutorial\/t_hpo.html 8) Ensemble Voting and Stacking are the two methods supported by this ensemble. Voting is based on the number of occurrences . Stacking combines the predictions from the well performing model. You can create your own ensembler by inheriting BaseEnsembler and overloading methods. 9) Solver It handles the auto-solvation of tasks. Currently, there are two methods provided by it. AutoNodeClassifier for semi-supervised node classification AutoGraphClassifier for supervised graph classification Its initialization has already been discussed in the above example, please refer to AutoGL Learning. Conclusion We have discussed all the basics of Auto Graph Learning in this tutorial. This project is currently under development and researchers are adding on new features. Following are the incoming features we can see in the near future. Neural Architecture Search Large-scale graph datasets support More graph tasks (e.g. Link prediction, Heterogeneous graph tasks, Spatial & Temporal tasks) Graph Boosting & Bagging More graph library backend support (e.g. Deep Graph Library) Tutorials and other resources used above: Website Github Documentation Colab Notebook I hope you find this article interesting and useful.","excerpt":"Creating algorithms is difficult and time-consuming. This specific problem has inspired researchers to develop some productivity tools to help young members in this domain. This has given birth to a revolutionary field in Data Science called Auto Machine Learning(AutoML). AutoML provides methods and processes to make Machine Learning available to non-Machine Learning experts, to improve […]","categories":["Deep Tech"],"tags":["data science training","graph database","Machine Learning","Python","Pytorch"],"author_name":"Aishwarya Verma","publish_date":"2021-01-20T18:00:00","publication_year":"2021","word_count":1116,"keywords":["CUDA","graph database","Pytorch","data science","machine learning","NumPy","AI","PyTorch","ML","Machine Learning","Python","data science training","Colab","R"],"extracted_tech_keywords":["AI","machine learning","ML","data science","PyTorch","Colab","NumPy","CUDA","Python","R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/complete-guide-to-autogl-the-latest-automl-framework-for-graph-datasets\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10014707,"title":"UMANG App To Introduce Face ID For Biometric Verifications &#8211; But Are We Ready?","content":"Government app UMANG (Unified Mobile Application for New-age Governance) wants to introduce the use of facial recognition for biometric verification using in-built features of a smartphone, according to recent news. The app, launched in 2017, is a platform to avail more than 2000 e-governance services provided by the central and state governments. It is developed by the Ministry of Electronics and Information Technology (MeitY) along with National e-Governance Division (NeGD) to drive mobile governance in India. This year the traffic on the app increased drastically, especially after the pandemic, as people’s physical movements and interactions were curtailed. The article discusses several advantages of introducing the use of a smartphone’s in-built features like face IDs for biometric verification and what India needs to consider for its safe deployment. Advantages Of Using Face ID For Biometric Verification UMANG has made access for various e-governance services easy for users. However, some of the services that need biometric authentication require the person to buy an external device. Currently, there are less than ten devices that are compatible with the app, and they cost within the range of ₹2,000 to 4,000. Hence, the use of an in-built biometric authentication will remove the dependency on the availability and affordability of these devices. Not only can it save time, but it can also make the services available in the comfort of one’s own home. This is especially useful for older people who need to avoid stepping out during the pandemic as well as working people who can use the app at a time of their convenience without worrying about an external device. At the same time, an Aadhaar Card service is also added to one’s UMANG app when logged in. The Aadhaar Card has seen widespread adoption as more than 1.27 billion people cards have already been generated. The use of face ID in Aadhaar Cards will make the verification process even easier. Lastly, theft identification becomes easier if the process is dependent on third-party biometric devices since the security standards might not be the same. If the smartphone’s in-built features are used for biometric verification, one will get the same security as the smartphone. Concerns Of Implementing Face IDs For Biometric Verification The collection of any biometric data, especially face IDs, has become a matter of concern in the data community as they realise the potential harm it can do in terms of surveillance and privacy. Robust laws and regulations are needed to safeguard one’s freedom and privacy when personal data is collected by the state. However, the existing legal infrastructure in India does not address this. The Personal Data Protection Bill 2019 as well, has been criticised heavily for enhancing existing surveillance powers of the government as it enables projects like nationwide facial recognition programmes among other surveillance projects. Hence, the face ID feature in UMANG should not be introduced before robust personal data protection laws are passed in the country. Secondly, the Umang privacy policy allows the sharing or disclosure of individual information with law enforcement agencies if it is ‘required by the law’. As several state governments have already started using facial recognition systems, it is important to highlight the repercussions of using facial recognition systems in policing. Apart from surveillance, there have been many examples of misclassification in facial recognition systems due to the lack of inclusive data. This is very much a possibility due to the small user-base in UMANG, as only those people who can afford smartphones can use the app. Thus irrespective of any laws, the policy of Umang should prohibit sharing face ID data or any other biometric information with law enforcement agencies. Or at least prohibit the development of face recognition algorithms from the data collected through the app. Thirdly, the collection of sensitive data like face ID should have strong security. Government databases in the past, including Aadhaar, have not been proven to be safe and have been vulnerable to hackers. Thus, with such a quick adoption of the UMANG app, it is imperative that the government first define and follow strong data security standards before it collects such sensitive data. Lastly, UMANG does not have any redressal mechanisms. It provides an email address to report grievances but does not have any provisions for what happens when there is a misuse of personal data. Misuse of a facial database without any consequences can have serious repercussions. Wrapping Up Tech-policy think-tanks and oversight bodies across the world have warned against the use of facial recognition technologies, especially the ones that can lead to mass surveillance. Some have even called for a complete ban until robust legal frameworks and redressal mechanisms are addressed. Thus, while there are considerable advantages of using face IDs in the UMANG app, it should not be implemented before proper laws, to restrict the use of the data collected through face IDs, are established.","excerpt":"Government app UMANG (Unified Mobile Application for New-age Governance) wants to introduce the use of facial recognition for biometric verification using in-built features of a smartphone, according to recent news. The app, launched in 2017, is a platform to avail more than 2000 e-governance services provided by the central and state governments. It is developed […]","categories":["AI Features"],"tags":["Data Security","expert system examples","Facial Recognition","facial recognition India","MeitY"],"author_name":"Kashyap Raibagi","publish_date":"2020-12-17T16:00:00","publication_year":"2020","word_count":810,"keywords":["Go","Facial Recognition","AWS","AI","Data Security","cloud_platforms:AWS","programming_languages:R","programming_languages:Go","expert system examples","facial recognition India","MeitY","R"],"extracted_tech_keywords":["AI","AWS","R","Go","cloud_platforms:AWS","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/umang-app-to-introduce-face-id-for-biometric-verifications-but-are-we-ready\/","complexity_score":3,"technical_depth":7,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10019299,"title":"Meet Lak, The Brain Behind Google Cloud&#8217;s Pathbreaking AI Solutions","content":"For this week’s ML practitioners series, Analytics India Magazine(AIM) got in touch with Valliappa Lakshmanan(Lak), Director for Data Analytics and AI Solutions at Google Cloud where he also founded Google’s Advanced Solutions Lab ML Immersion program. He currently leads a team that builds software solutions for  business problems using Google Cloud’s data analytics and machine learning products. He is also the author of a handful of popular ML books. In this interview, Lak gives us a glimpse of how data gets leveraged from his vantage point. AIM: Could you please tell us about your educational background? Lak: I pursued my PhD in Electrical and Computer Engineering from the University of Oklahoma and my Master’s in Biomedical engineering from The Ohio State University in Columbus, Ohio. I did BTech in Electronics and Communications Engineering from the Indian Institute of Technology, Madras. AIM: How did your journey in machine learning begin?  Your fascination with algorithms. How did it all start? Lak: My interest in Digital Signal Processing and Image Processing started in college when I got introduced to these subjects. My UG Project was to build a transducer to monitor heart rates. While applying for Masters I was keen to pursue my interest in this area and a lot of my coursework focused around image processing and computer vision. My MS project was on identifying mitral valves from ultrasound images. My experience of building image processing for ultrasound helped me in getting my first job in weather forecasting. It turns out that much of the signal processing math is similar between ultrasound and ground-based weather radar. The research lab that  I worked at – at University of Oklahoma and National Oceanic and Atmospheric Administration (NOAA) was a great opportunity and helped me enhance my skills and knowledge on weather detection algorithms. As datasets grew over time, I continued my exploration of new technologies and started to apply machine learning methods like genetic algorithms and neural networks to those problems. In the 2010’s when Deep Learning was well on its way to becoming the backbone technology for all sorts of big data problems, it was  a natural transition to research areas I was already involved in. AIM: What books and other resources have you used in your journey? Lak: One of the books I learned the most from is Neural networks for Pattern recognition by Professor Christopher Bishop. It provides a solid foundation and great intuition. I have learnt and researched on new topics as and when needed. I also learn new topics when I mentor junior colleagues. Usually, this is by reading papers and articles and trying new techniques. AIM: What were the initial challenges and how did you address them? Lak: In the ancient days, for all algorithms we needed, I had to implement myself in C++. The tools we had, like Weka or Stuttgart Neural Network Simulator (SNNS) wouldn’t scale to the size and complexity of the weather data. It was quite a task managing these and I wish we had solutions like TensorFlow or Keras back then! AIM: Can you talk about the role at your current company? What does a typical day look like? Lak: I lead the Analytics and AI solutions team at Google Cloud. My team drives client roadmaps, program building that incorporate intelligent processes into existing business process and solution planning. These solutions are best practice guides and reference implementations for common customer problems. It’s hard to think of a typical day! So, let’s take a typical month-  A lot of my work revolves around providing and receiving information. I spend a substantial amount of time working with Google Cloud customers. This entails providing strategic advice to executives, helping a data science team identify the kinds of problems they can tackle, and helping troubleshoot problems and loop in the appropriate teams at Google. I also spend a lot of time helping my teams prioritise the solutions they are building, helping them when they run into issues, and in general, doing what I need to do to make them successful and look at newer ways of growth and development for the team and myself. Besides, I am also responsible for business operations and co-ordinating with internal stakeholders. AIM: How does your team approach a data science problem? Lak: We spend a considerable amount of time in designing the data collection and data preparation as we do in training\/evaluating\/deploying models. The key tips here are to ensure that you are visualizing the data at every step of the way so that you know what’s happening to it. It’s important to look, not just at aggregate statistics, but also at individual samples. A recent project by our team was to build an end-to-end system for doing real-time matching of documents. The solution we have published is a generalization of one that we built for a major publisher. Given a short synopsis of a book or article, they wanted to be able to retrieve similar articles that they had in their very extensive index. In the original project, we spend quite a bit of time working through getting data from across all their brands, and from public repositories and building a matching system that would work with missing data. When it came to the ML model and the different components of the system, we went with options that were both state-of-the-art as well as proven. Since many of the systems we used were serverless and fully managed, it was possible for us to move pretty fast. AIM: What does your machine learning toolkit look like? Lak: My preference is to develop structured data models and time-series models in BigQuery ML. For models on unstructured data, I prefer using  Keras\/TensorFlow and  we use State-of-the-Art (SOTA) models rather than reinvent the wheel. We usually use EfficientNet for most image models, ARIMA for time series, Bert for text embeddings, etc. For cloud services, we use Google Cloud. We use BigQuery as our data warehouse, Apache Beam on Dataflow for ETL, Cloud AI Platform Notebooks for development, Cloud AI Platform Training and Predictions for managed training and deployment, and Cloud AI Platform Pipelines (Kubeflow Pipelines) for managing the end-to-end system. AIM: How has the ML landscape changed over the years? Lak: Over the years, Machine learning has truly evolved and has become more powerful, and much more reusable. It is very rare that one has to develop ML models from scratch and can  instead, opt for pre-designed models and train them on your dataset. AIM:From a global perspective, where do you think India stands in the AI\/ML ecosystem? What are the potential areas of Improvement? Lak: I have been very impressed by the growth of ML applications, especially when it comes to mobile applications, text readers, medical technology, among others. There is a huge talent pool  of young Indians with a strong appetite for upskilling. The degree and pace of mobile connectivity in India is very impressive. AIM: There is a lot of hype around machine learning. So, when the dust settles down, what techniques, use cases and applications do you think will stand the test of time? Lak: The greatest opportunity to create value using ML is in personalization and recommendation systems. The greatest opportunities are in digitizing industries, for example, to enable inventory management in physical stores that are exploring online fulfilment. These use cases will remain because customer satisfaction and cost savings will remain business priorities for the long term. The way we solve this – whether using voice assistants or scanning shelves using image classification – will change over time. The real differentiator will be in the kind of data we can collect and put to use to solve these problems. AIM: Which domain of AI, do you think, will come out on top in the next 10 years? Lak: AI is just a tool. It will be used to solve a wide range of problems. I don’t think it is useful to think in terms of domains of AI as being in competition with each other. The field of AI that will become more and more useful is to design AI with humans in mind – similar to the field of human factors research in user interface design. AIM: What do outsiders get wrong about this field? Lak: The word AI often brings us thoughts of humanoid robots who possess general intelligence – they can walk, talk, and do a wide variety of tasks. But AI today is extremely specific in what it can do. AIM: Any additional tips\/resources for the beginners? Lak: I would recommend beginners to read books that provide a comprehensive introduction to the topic and make sure to gain practice with actually doing machine learning. I’m a little biased, but I do recommend my books: Data Science on the Google Cloud Platform– In this book, I have explained  a typical data science project from the time you plan it, to collecting data, to deploying an ML model and consuming its predictions. The book BigQuery: The definitive guide will help you acquire a very important skill – the ability to easily handle large datasets and quickly gain insight from them. Finally, the book Machine Learning Design Patterns helps you learn what experienced people in ML already know. That is the order in which I wrote the books, and the order in which I suggest you read them as well.","excerpt":"For this week’s ML practitioners series, Analytics India Magazine(AIM) got in touch with Valliappa Lakshmanan(Lak), Director for Data Analytics and AI Solutions at Google Cloud where he also founded Google’s Advanced Solutions Lab ML Immersion program. He currently leads a team that builds software solutions for  business problems using Google Cloud’s data analytics and machine […]","categories":["Global Tech"],"tags":[],"author_name":"Ram Sagar","publish_date":"2021-01-29T18:00:00","publication_year":"2021","word_count":1564,"keywords":["data science","machine learning","AI","neural network","ML","computer vision","Kubeflow","deep learning","Aim","analytics"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","neural network","computer vision","data science","analytics","Kubeflow","Aim"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/interview-lak-the-brain-behind-google-cloud-ai-machine-learning\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":66731,"title":"Netflix Is Using AI For Its Subtitles","content":"As Netflix continues to expand its global reach, the viewership has become quite global. The audience is totally fine with watching a TV show that is shot in a foreign language. Netflix has somehow succeeded to break the language barrier, but the challenge now lies in making sure that the translations are accurate. Netflix has a rigorous process in place to discard inaccurate subtitles, but that still doesn’t help. So, a group of ML researchers at Netflix have introduced a new approach to tackle this. They call this approach Automatic Pre-Processing or APP. This process, the researchers claim, will give translations that are close to those of the native language. Overview Of The Model Translation quality for low-resource translation (i.e., from English into a low resource language) in the black-box MT (BBMT) setting is challenging. So, the researchers at Netflix introduced a method to improve such systems via automatic pre-processing (APP) using sentence simplification. For example, if a source sentence says “The vice president should feel free to jump in,” and is to be translated into Hindi, using Google Translate one will get “Vice President should feel free to jump inside.” The system, state the Netflix team, was unable to correctly translate the idiomatic and non-compositional phrase “jump in.” Machine translation (MT) systems trained on smaller training sets usually give results that deviate from the context. Grasping phrases, idioms, or complex word language pairs is a challenging task. In other words, the back-translation is different in meaning than the natural source sentence. To address this problem, the researchers adopt the notion that translating back-translations is easier than translating naturally occurring source sentences. The model Automated Preprocessing (APP) builds on this observation that human reference translations when back-translated to the original language is a rich source of simplifications (e.g. “jump in” is simplified to “take part”). This observation, stated the Netflix team, leads to two immediate corollaries – back-translating the ground truth human translations to the source language results in a simplified version of the original source, and a function to map the source sentences to its simplified version can be learned by training a sequence-to-sequence (S2S) model. The evaluation of the model was done on the GIGS, Wikilarge and the Open Subtitles datasets. FIGS dataset comes from subtitles appearing on 12,301 TV shows and movies from a subscription video-on-demand provider. For training the APP simplification model, the researchers used the Transformer architecture through the tensor2tensor library. All experiments were conducted using the transformer base architecture with 6 blocks in the encoder and decoder and are run using 4 NVIDIA V100 GPUs. Even though this work mainly focuses on simplifying English-based subtitles, the model is universal and can be used for other languages. “Our work merges two important sub-fields machine translation and sentence simplification, and paves the path for future research in both of these fields,” the researchers wrote. According to Netflix, the errors in subtitles might not be that critical but are subtle. These experiences add up and might affect the user engagement. Netflix sometimes rejects a subtitle even if the subtitles are grammatically correct but fall short of getting the simple phrases and colloquialisms right. Translating multiword expressions and non-compositional phrases is a tricky task, and simplifying these expressions before translating help. This work merges machine translation and sentence simplification; two important sub-fields of NLP and the researchers believe that this will lead to further research. Know more about this work here.","excerpt":"As Netflix continues to expand its global reach, the viewership has become quite global. The audience is totally fine with watching a TV show that is shot in a foreign language. Netflix has somehow succeeded to break the language barrier, but the challenge now lies in making sure that the translations are accurate. Netflix has […]","categories":[],"tags":["netflix","translate"],"author_name":"Ram Sagar","publish_date":"2020-06-05T10:00:35","publication_year":"2020","word_count":573,"keywords":["Go","programming_languages:R","AI","ML","programming_languages:Go","netflix","NLP","Aim","translate","transformer architecture","V100","R"],"extracted_tech_keywords":["AI","ML","NLP","Aim","R","Go","transformer architecture","V100","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/netflix-ai-subtitles-transalation\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10040162,"title":"Overview Of Atoti: A Python BI Analytics Tool","content":"Atoti is a Python business intelligence analytics tool that creates a Tableau-like dashboarding interface inside Jupyter notebooks. It provides a BI web application that enables hassle-free dashboard creation and sharing. Notebooks on their own are an amazing tool but they very obvious limitation when it comes to analytics tasks: Pandas DataFrames are good for data wrangling, but they start to slow down when the datasets grow larger than a couple of GigaBytes, forcing the analyst to start over in Spark. Visualization libraries create frozen plots, sure they are interactive, but you can’t apply filters without creating a new plot.Lack of native support for multi-dimensional analysis like OLAP applications. Atoti Stores, in-memory tables, scale very efficiently and can handle a lot more data than DataFrames. Additionally, it enables analysts to create advanced data models using joins between stores without duplicating data as done by a merge in pandas. Atoti has embedded interactive data visualization tools that can be used to build scenarios, apply filters and compare different versions of the data. That being said, atoti doesn’t aim to replace pandas or Spark; they are both good tools for cleaning and transforming data. Atoti, on the other hand, focuses on visualization, analysis and collaboration. If that’s the case, then why not just use a dedicated BI tool like Tableau or PowerBI? For starters, atoti eliminates the need to export the data and load it into another software. In addition to that, it enables analysts to create new measures in Python rather than using a niche language like PowerBI’s DAX. Creating a Dashboard with Atoti Atoti can be installed from PyPI, if you want to use it interactively in notebooks, you’ll need to install its JupyterLab plugin as well. pip install atoti[jupyterlab] Create a session. import atoti as tt session = tt.create_session() Load the data into an atoti Store session.load_all_data() data = session.read_csv(\"FL_insurance_sample.csv\", keys=[\"policyID\"]) data.head() load_all_data() needs to be called when there are more than 1000 rows; otherwise, the first 1000 rows are sampled. The keys argument is used to indicate the primary key of the table. Inspired by OLAP applications, the data in atoti is modelled as cubes before analysis. Cubes provide a multidimensional view of data; they make it easy to explore, aggregate, filter and compare. It’s called a cube because each attribute of the data can be thought of as a cube dimension. cube = session.create_cube(data) Read more about OLAP cubes here. The cube automatically creates hierarchies for all non-numeric fields and measures for the numeric fields. Let’s see the measures created by the cube. m = cube.measures m The cube has created sum and mean measures for all numeric fields. Using the query() method, you can fetch the value of measures over the whole dataset. cube.query(m[\"eq_site_limit.SUM\"], levels=l[\"county\"]) Or dice the cube to get the value for each COUNTY cube.query(m[\"eq_site_limit.SUM\"], levels=l[\"county\"]) Or slice on a particular COUNTY cube.query( m[\"eq_site_limit.SUM\"], condition=l[\"county\"] == \"BRADFORD COUNTY\", ) To create visualisations call the visualize method on the cube object. This will create a widget in the notebook and open the atoti tab on the left to manipulate the widget. session.visualize() Building small widgets like this is good for exploring data but for providing deeper insights dashboard are better. Atoti enables creating dashboards by providing a web application that can be accessed outside of the notebook; here, widgets can be used to form sharable dashboards. The URL of this web app can be accessed by session.url This web application is a “safe” environment; all the filters and queries are read-only and the original data is not affected by it. Last Epoch This article introduced atoti, an analytics tool that enables analysts to analyze millions of rows on their laptops. It supports multi-dimensional analysis with OLAP cubes and the creation of interactive widgets inside Jupyter notebooks without any code. Atoti has various plugins that allow it to work with data from cloud platforms like AWS, GCP, Azure, etc. To learn more about atoti you can refer to the documentation.","excerpt":"Atoti is a Python business intelligence analytics tool that creates a Tableau-like dashboarding interface inside Jupyter notebooks.","categories":["AI Trends"],"tags":["Business Intelligence","Dashboard analytics"],"author_name":"Aditya Singh","publish_date":"2021-05-15T11:00:00","publication_year":"2021","word_count":663,"keywords":["GCP","AWS","AI","R","Dashboard analytics","Python","Aim","analytics","Business Intelligence","Jupyter","Azure","Pandas"],"extracted_tech_keywords":["AI","analytics","Aim","Jupyter","Pandas","AWS","Azure","GCP","Python","R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/overview-of-atoti-a-python-bi-analytics-tool\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10162542,"title":"DeepSeek-Style Innovation Already in the Works in India","content":"By now, you must know that China’s latest AI model, DeepSeek-R1, has been the centre of all conversations for having built a SOTA model with scant resources with respect to compute and cost. It shattered the idea that building a model as capable as OpenAI’s on a $10 million budget is impossible—remember Sam Altman’s last visit to India? With DeepSeek setting a precedence for everyone, India has gotten the boost it always needed. India’s DeepSeek Ambitions When US President Donald Trump announced Project Stargate a week ago, India got talking about building in the country. India building its own Project Stargate was portrayed as a necessity, with many tech leaders weighing in on the conversation. The discussions had barely died down when DeepSeek brought forth the next idea of why India can’t build one just like it. Well, India is already building it. “Yes, we definitely are! It won’t be a 671B parameter one (to begin with), but it’ll be a frontier model in its parameter category,” said Abhishek Upperwal, founder and CEO of Soket AI Labs, in an exclusive interaction with AIM. “The pace of development will depend on the kind of funds we get access to, but we are gonna definitely build it,” the founder of the Gurgaon-based AI research startup added. Upperwal stated that Pragna-1B (Soket’s AI model) marks the team’s initial step toward developing frontier models. The 1.25 billion-parameter model was trained on a budget of just $100K, covering both synthetic data and compute costs. “The plan is to bootstrap bigger models using smaller ones and any open-source model with a permissive license—while keeping compute costs dirt cheap,” he said. He highlighted that high-quality data and training optimisations make this approach feasible, pointing to DeepSeek as a successful example. Upperwal noted that if  “less resources” translates to $2-3 million, the prospects for building frontier models are either bleak or significantly slow. In such a scenario, companies would have to prioritise revenue-generating products over AI model development. “I think we need at least $10 million to start working on frontier tech, and this money should be purely dedicated to R&D for building these models—no distractions like building applications or even thinking about GTM. This is where investors and founders need to align with patient capital,” said Upperwal. Similarly, Reliance-backed Indian AI startup TWO AI is building a cost-efficient multilingual AI model family with speech, search, and visual processing in 50+ languages. It believes it has already been building DeepSeek-like models. “DeepSeek’s RL-only post-training approach and insights like distilling reasoning into smaller models really resonate with what we’re doing at TWO AI,” said Pranav Mistry, founder and CEO of TWO to AIM. Mistry believes the AI race now demands rapid innovation rather than massive compute power. “Gone are the days when you needed a 20,000 GPU farm to train a single model,” he said. He added that TWO AI has demonstrated this with its SUTRA model, which outperforms SOTA models in the official MMLU for Indian languages despite being trained on a $2 million budget. While greater resources can accelerate innovation, optimised approaches are proving just as crucial. “Of course, more resources can help accelerate the speed at which we can innovate,” he added. Pratyush Kumar, co-founder of Sarvam AI, another Indian AI startup that is developing LLMs and GenAI solutions for Indian languages, recently posted on X inviting Perplexity co-founder Aravind Srinivas to join their mission. “Aravind, at SarvamAI we are building sovereign models that combine deep reasoning and Indic language skills. Would love to have you join this mission!” he wrote. However, when AIM reached out, Sarvam AI declined to comment on DeepSeek. Multimodal AI platform Krutrim AI, started by Ola’s Bhavish Aggarwal, is also on a mission to cater to the Indian audience via their multilingual platforms. What is Stopping India? “Very soon, we will also have our own LLMs,” said IT minister Ashwini Vaishnaw, at the recent Utkarsh Odisha Conclave. “In the India AI compute facility, we have received compute bids for creating 18,000 GPUs,” he said. While the government is slowly encouraging and providing incentives to promote AI in India, VCs are still sceptical about investing fully in it. “The problem is that the benefit here isn’t immediate revenue generation, which is why VCs run away from these kinds of ventures. But the real ROI is in gaining the know-how of building intelligence at scale, which can create value in a hundred other ways (just imagine the kind of leverage DeepSeek holds today),” said Upperwal. “Intelligence and the know-how to build one will be the most valuable IP in the future,” he added. Upperwal believes that to reach DeepSeek R1’s level, we will need at least $50 million. “DeepSeek is already on its 3rd version, plus multiple other models. The cost to get here should be the aggregate of everything they’ve spent so far. I’d estimate $50-100 million,” he said. He believes the key lies in securing adequate R&D funding (ranging from $5-10 million per startup) for at least 4-7 teams. “Sarvam is the only startup with access to such funds, but it’s splitting its focus between figuring out use cases and building models, which slows down progress,” he said. In a blog post, Zerodha co-founder Kailash Nadh shared his views on DeepSeek, focussing on research and human capabilities as a priority. Nadh believes that India’s AI sovereignty and future depends not on a narrow focus on LLMs or GPUs but on building a foundational ecosystem that encourages breakthroughs through a blend of scientific, social, and engineering expertise across academia, industry, and civil society. “In fact, the bulk of any long-term AI sovereignty strategy must be a holistic education and research strategy. Without the overall quality and standard of higher education and research being upped significantly, it is going to be a perpetual game of second-guessing and catch-up,” he said.","excerpt":"“Very soon, we will have our own LLMs,” said IT minister Ashwini Vaishnaw.","categories":["AI Startups"],"tags":["DeepSeek","India","Kailash Nadh","Perplexity","TWO AI"],"author_name":"Vandana Nair","publish_date":"2025-01-30T17:00:00","publication_year":"2025","word_count":977,"keywords":["GenAI","TWO AI","Perplexity","OpenAI","AI","ML","DeepSeek R1","RAG","DeepSeek","Aim","Ray","multimodal AI","Kailash Nadh","R","India"],"extracted_tech_keywords":["AI","ML","GenAI","multimodal AI","OpenAI","DeepSeek R1","Aim","Ray","RAG","R"],"url":"https:\/\/analyticsindiamag.com\/ai-startups\/deepseek-style-innovation-already-in-the-works-in-india\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10019935,"title":"Does Recent Databricks’ Massive Investment Signal A Maturing Data Science Industry?","content":"In February 2021, San Francisco- based Databricks closed a $1 billion late-stage round led by Franklin Templeton. Canada Pension Plan investment board, Amazon Web Services (AWS), and Salesforce Ventures have also participated in the Series G funding. The billion-dollar investment came in the wake of Databricks’ partnership with the cloud companies, which CEO Ali Ghodsi called a ‘symbiotic relationship of strategic importance’. With such firms attracting significant investments, is the data science industry maturing? Growth Phase Databricks has seen rapid growth in the past couple of years. In 2019, the company raised $250 million and $400 million in two funding rounds. Currently, the company is valued at $28 billion. Many other data companies are also seeing huge investments. Snowflake, a cloud data warehousing firm, attracted funding of $479 million in February 2020, taking the total investment tally to $1.4 billion. In September, the company raised $3.36 billion in the biggest software IPO ever. Alteryx, a data science company, listed in 2017, has raised $163 million in funding and is currently valued at $9.11 billion. Datarobot raised $270 million investment last year, taking its total funding to $751 million, and is currently valued at $2.7 billion. All of these companies have shown high revenue growths. Alteryx generated quarterly revenue growth of 26% year-on-year, whereas Snowflake grew by 119%. While Databricks is not publicly listed, according to news reports, it grew by 75% in Q3 2020 to $350 million, up from $200 million in Q3 2019. The two listed companies, Alteryx and Snowflake, showed a negative Earning Per Share (Trailing Twelve Months) of 0.24 and 5.11 last year in September. Alteryx grew its customer base by 27% at the end of the second quarter of 2020, and Snowflake doubled its customer base last year. Matured industries tend to have slow and steady progress with consistent returns. Hence, even with the high valuations, indicators like the rising revenues, investments, and customer bases show that the data science industry is in a growing phase. Investments pour in tantamount to industry’s topicality and the potential applications and services. The right strategies in investments, along with mergers and acquisitions, play a role in the maturing of industries. Aligned Strategies In an interview following the Databricks funding round, Ghodsi said the Series G made a lot of sense with respect to the participants. On closer inspection, Databricks funding round is strategic investments by big cloud players on an open and unified data analytics platform. In other words, giants like AWS and Salesforce are finding niche opportunities in niche markets. Databricks’ acquisition of Redash, a data visualisation platform last year, ties in perfectly with the overall vision. In the case of Snowflake, the latest round came from Salesforce Ventures. Snowflake is a data cloud firm providing various applications, including data sharing, data science, data warehousing, among others. Salesforce, on the other hand, is the world’s number one CRM solutions provider that needs huge data storage and platforms for data analysis. Similarly, the DataRobot, that provides an end-to-end enterprise AI platform to build value from data, has seen fundings mostly from tech investors. It acquired Cursor in 2019 to bolster data collaborations. Wrapping Up Compared to five years back, data science firms are experiencing a spurt of growth and are profusely raising funds. The strategic collaborations and mergers also suggest the industry is on track to consolidation. However, the key metrics suggest the industry is in a high-growth phase and is far from being mature.","excerpt":"In February 2021, San Francisco- based Databricks closed a $1 billion late-stage round led by Franklin Templeton. Canada Pension Plan investment board, Amazon Web Services (AWS), and Salesforce Ventures have also participated in the Series G funding. The billion-dollar investment came in the wake of Databricks’ partnership with the cloud companies, which CEO Ali Ghodsi […]","categories":["AI Features"],"tags":["Statistics for Data Science"],"author_name":"Kashyap Raibagi","publish_date":"2021-02-10T14:00:00","publication_year":"2021","word_count":576,"keywords":["data science","API","AWS","AI","RAG","Databricks","analytics","Statistics for Data Science","R","analytics platform","Snowflake"],"extracted_tech_keywords":["AI","data science","analytics","RAG","AWS","Snowflake","Databricks","R","API","analytics platform"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/does-recent-databricks-massive-investment-signal-a-maturing-data-science-industry\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10093963,"title":"Data Science Hiring and Interview Process at Meesho","content":"Founded in 2015 by Vidit Aatrey and Sanjeev Barnwal, e-commerce platform Meesho has over 100 million customers. It recently surpassed a record 1.1-million seller mark on its platform, attracting over 600,000 small enterprises within the last 12 months. Backed by the likes of SoftBank, Meta, Y Combinator and Fidelity Investments, Meesho allows anyone to start their businesses with zero investment or inventory. Over the years, the leading online marketplace platform has used AI and analytics to revolutionise the way we shop. Besides implementing existing AI and ML models, Meesho also has customised frameworks. By leveraging data science to personalise product recommendations, optimise prices, detect fraud, and provide customer support, Meesho is making it easier and more convenient for customers to find and buy the products they want. Analytics India Magazine got in touch with Debdoot Mukherjee, chief data scientist, and head of AI and demand engineering at Meesho, to understand how the e-commerce giant is dabbling in AI and ML and what is their hiring process for data science candidates. Read more: Data Science Hiring Process at NoBroker Meesho’s AI & Analytics Play “At Meesho, AI and ML are fundamental components to solve a wide range of problems across every aspect of an e-commerce platform,” Mukherjee told AIM. On the demand side, Meesho implements AI and ML to engage users through push notifications and provide personalised feeds. Recommendations play a crucial role, with a significant percentage of their orders coming through recommendation algorithms powered by ML and deep learning models.  The recommendations rely on a combination of user and product understanding, which involves analysing user interaction data and decoding product content using computer vision and NLP models. Meanwhile, the company empowers its suppliers by offering efficient product cataloguing. With just a simple photo, their automated cataloguing process, driven by computer vision models, streamlines the task. They also assist suppliers in optimising their pricing through guided suggestions powered by ML techniques. “In terms of fulfilment, we enhance order efficiency by addressing customer issues, detecting and preventing fraud, and resolving any issues that may arise in the system. These functions heavily rely on large-scale ML models,” said Mukherjee. The 60-strong data science team is also actively experimenting with generative AI in various applications like recommendation systems to rewrite or correct user queries and enhance cataloguing through richer visual representations and virtual try-on experiences. The company uses TensorFlow or PyTorch for most of its model training. They also use a lot of tree-based models like XGBoost. Databricks is used as the data engineering platform, which includes a data lake and other infrastructure components such as the offline feature store. In addition to pre-existing models, Meesho also develops tailor-made tools based on specific requirements. This encompasses their exclusive model to detect shopping intent, built upon cutting-edge transformer models employing deep learning techniques. For cataloguing, Meesho has built custom models built on existing computer vision models to extract detailed information about products, starting from their images. For fraud detection, Meesho has developed a proprietary system that effectively identifies and catches a significant number of fraudulent activities. It specifically addresses issues like Return to Origin (RTO) fraud, which is a major concern in the Indian e-commerce market. Meesho also has systems in place to counter other types of fraud. Read more: AI Cloud Wars: Azure AI vs Vertex AI Interview Process The selection process begins with rounds focused on coding skills and fundamentals. Candidates are tested on their ability to build systems, including ML systems, in applied environments. Subsequent rounds assess ML fundamentals, theoretical knowledge, and the use of core algorithms. Later, candidates face application-focused rounds with case studies, where they must develop ML solutions for real-world problems. The company values candidates’ coding skills, data science expertise, and problem-solving abilities in practical scenarios. Candidates applying for the role are assessed on three components. Firstly, they need expertise in machine learning and AI, encompassing knowledge of algorithms, statistics, and ML properties. Secondly, a strong foundation in computer science is necessary to develop scalable systems and apply ML techniques on a large scale. Lastly, they must possess product thinking skills to identify problem formulations that improve business metrics and address customer issues. Read more: Jeffrey Ullman’s Unsettling Ultimatum Work Culture “The Meesho AI team consistently pushes the boundaries of the state-of-the-art and encourages the development of cutting-edge systems. I believe that candidates joining our team should anticipate a promising and exciting set of problem statements to solve,” added Mukherjee. With a problem-first mindset, Meesho encourages curiosity, creativity, and finding new solutions. The ‘Meesho Mantras’ foster a people-centric workplace, promoting employee happiness and engagement. Meesho values teamwork and collaboration, rewards speed over perfection and encourages diverse perspectives. Employees are empowered to take ownership of their work, make decisions and take bigger risks. The organisation takes pride in providing top-notch employee benefits, including maintaining a healthy work-life balance, comprehensive health insurance and offering flexible leaves. “AI can touch users’ lives firsthand at scale across hundreds of millions of users. So if you care about creating impact at scale, Meesho is the right fit,” concluded Mukherjee. Read more: Data Science Hiring Process at ZS","excerpt":"The SoftBank, Meta, Y Combinator and Fidelity Investments-backed e-commerce platform recently surpassed the 1.1 million seller milestone on its platform, attracting over 600,000 small enterprises within the past 12 months","categories":["AI Hirings"],"tags":["Career","Data Science Hiring","Meesho","Top Trend"],"author_name":"Shritama Saha","publish_date":"2024-07-29T15:49:34","publication_year":"2024","word_count":853,"keywords":["Top Trend","data science","machine learning","Meesho","AI","ML","computer vision","NLP","Data Science Hiring","Aim","deep learning","analytics","generative AI","Career"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","NLP","computer vision","data science","analytics","generative AI","Aim"],"url":"https:\/\/analyticsindiamag.com\/ai-hiring\/data-science-hiring-process-at-meesho\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10119757,"title":"Microsoft is Starting to Look a Lot Like OpenAI with ‘MAI-1’","content":"Microsoft is done relying on others for its models. According to reports, though the company had been training smaller models like Orca and Phi all this while (using GPT and incorporating Meta’s Llama on its platform), this time it is training a model large enough to compete with others. Referred to as MAI-1 (possibly Microsoft AI-1), the model is being developed internally by the company, and is around 500 billion parameters in size. Its development is being headed by Mustafa Suleyman, formerly a leader in AI at Google and most recently CEO of the AI startup Inflection, who now oversees Microsoft’s AI division. In March, Microsoft acquired a majority of Inflection’s staff and paid $650 million for its intellectual property rights. However, MAI-1 is a Microsoft-developed model, distinct from those previously developed by Inflection. While it may leverage training data and technology from the startup, it is an independent project. Though the exact purpose has not been disclosed just yet, it is possible that Microsoft might incorporate its products into all its Copilot products. This would mean that the company will move away from OpenAI’s GPT and Codex models. This comes in the backdrop of OpenAI partnering with Stack Overflow to improve its products. The big-tech company is also planning to release its search engine powered by Bing possibly as a way to compete with Google, which Microsoft has a history of doing. No competition? Microsoft’s strategy has highlighted three trends for 2024 – small language models, multimodal AI, and AI in science. This time around, it seems like the company has decided to incorporate multimodal AI with large language models. The small language models will continue to be incorporated into the company’s on-edge use cases, such as laptops, while larger ones might be available in other core products. All of this comes only weeks before the Microsoft Build conference. So it is possible that the model will be introduced at the conference. To train the new model, Microsoft has allocated a significant cluster of servers equipped with NVIDIA GPUs. To make the case clear, Microsoft’s CTO Kevin Scott went on LinkedIn to explain that this is not in any way a competition with OpenAI. “I’m not sure why this is news, but just to summarise the obvious: we build big supercomputers to train AI models. Our partner OpenAI uses these supercomputers to train frontier-defining models; and then we both make these models available in products and services so that lots of people can benefit from them. We rather like this arrangement,” he said. He further stated that the company has always been building bigger and better supercomputers for OpenAI to further AI research and wishes to continue this arrangement. “There’s no end in sight to the increasing impact that our work together will have,” he added. So it seems like Sam Altman’s $50 billion dream of building AGI is going to get funded by Microsoft. Scott also clarified that all the research that Microsoft does is about building models: “AI models turn out to be interesting things to work on, and our researchers do great work studying and building them.” He also confirmed that there would be more models coming out soon, including MAI, Phi, and even Turing. But is it true? Even though Scott clarified that there is no competition between OpenAI and Microsoft, it is still worth a wonder if a model as big as MAI-1 would actually be used for Microsoft products instead of OpenAI’s GPT. Moreover, it would be ideal for Microsoft to have a backup plan just in case the deal with OpenAI falls through, as has been the case with several others in the field. Chief Satya Nadella seems to be playing a different AI game. Under him, Microsoft has invested in all kinds of AI companies, from OpenAI and Mistral to Databricks and Figure AI. Recently an email from 2019 resurfaced, in which Scott told the Microsoft team that they needed to invest in OpenAI as the competition with Google was rising and that their AI model was “scarily good”. Suleyman recently posted on X saying, “AI is everything at Microsoft”. He also highlighted that the company is building massive products using AI and has a definite vision for Copilot. Everything about this seems forced. There seems to be no other reason for Microsoft to build such large models and spend so much on compute if they’re not making it for commercial purposes. Moreover, on his hiring, Suleyman was touted as the “new” Sam Altman. So, is Microsoft possibly becoming the OG of AI, aka OpenAI?","excerpt":"Meanwhile, Sam Altman is aiming to build AGI.","categories":["Global Tech"],"tags":["Microsoft","OpenAI"],"author_name":"Mohit Pandey","publish_date":"2024-05-07T14:09:53","publication_year":"2024","word_count":764,"keywords":["Go","startup","OpenAI","AI","RAG","GPT","multimodal AI","small language models","R","Microsoft","Databricks"],"extracted_tech_keywords":["AI","multimodal AI","OpenAI","small language models","RAG","Databricks","R","Go","GPT","startup"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/microsoft-is-starting-to-look-a-lot-like-openai-with-mai-1\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10017006,"title":"9 Best Image Datasets For Data Scientists in 2024","content":"Vision data is the most widely used form of data around us. Almost every industry from fashion to streaming platforms, medical, legal, finance all has its usage for various use-cases. Social media being one of the biggest examples. AI has taken over everything in the world now and has done wonders to image data. Machine learning and deep learning models as we know are well trained where there are diverse data, so these algorithms are data hunger. Thus there became a need to develop better datasets to address biases present in these algorithms. Computer vision is a field where computers deal with digital images in the form of pixel values. In other words, computers are made to have an understanding of images\/videos as humans do. It includes processing, analyzing, transforming, extracting features and various other operations done to an image. Earlier image processing techniques used have certain drawbacks as they fail to bring out high-level dimensionality accurately. Now deep learning algorithms have overcome these problems and have proven to be much reliable. Nowadays they are used in almost all kinds of tasks such as object detection, object tracking, image classification, image segmentation and localization, 3D pose estimation, video matting and many more we can keep naming. Taking image datasets forward now GANs (generative adversarial networks) have taken over. They can increase the size of datasets by including synthetic data. Besides, it can make synthetic data imitate exactly like real-world data, for example – deepfakes. In recent years it has gained much attention, and more research and development is revolving around it. In this article, we will discuss the various image datasets that are readily available for training machine learning models. MNIST and Types MNIST is the handwritten digits dataset. The very first of its kind to have been developed in 1999 by Yan LeCunn and other researchers. It is a very basic dataset for beginners, starting deep learning with computer vision. Using simple Convnet architectures these are very easy as it is preprocessed in grayscale images (total 70,000 out of which 60,000 training set and 10,000 test set) each of 28*28 pixels associated with numbers 0 to 9 as labels. Over the years different variants of MNIST have been released namely – binarized MNIST, KMNIST, EMNIST, QMNIST, and 3D MNIST. Binarized MNIST contains the binarized version of original digits MNIST. EMNIST or extended MNIST is an extension by adding more data to the original MNIST. KMNIST is Kuzushiji MNIST which is a drop-in replacement of the original MNIST with NumPy format. QMNIST developed by Facebook AI research contains 50,000 additional images apart from the original MNIST. 3D MNIST, as the name suggests, contains 3-dimensional digit representations. It is a smaller dataset compared to MNIST. All of these datasets are open-sourced and readily available to use in ML model training. There are some pre-built libraries in Tensorflow and PyTorch for implementing these datasets. For implementation and other information -> 6 MNIST Image Datasets FASHON MNIST MNIST could not explore many aspects of deep learning algorithms based on computer vision, so Fashion MNIST was released. As the name suggests, it contains ten categories of apparels namely T-shirt\/top, trouser, pullover, dress, coat, sandals, shirt, sneakers, bags, ankle boots with class labels 0 to 9 as MNIST. All of these images are in grayscale with 28*28 pixels each. With fashion MNIST new benchmarks were achieved in deep learning. This also has pre-built libraries to be readily used for model training. Recently fashion MNIST was used with GANs and have generated really good results showing new apparel designs. For implementation and other information -> Fashion MNIST MEDICAL MNIST AND TYPES Following the MNIST type structure, many other datasets were released to fulfil different purposes. With neural networks finding relevance in all fields, medical science has many things to be covered and addressed. Bioinformatics data science has now been much in research and achieved some of the results that weren’t addressed for years. Different medical MNIST datasets have evolved over the years, MedMNIST is one of the recently released (in 2020) benchmark datasets in them. It is a collection of 10 open sourced medical datasets namely – PathMNIST, ChestMNIST, DermaMNIST, OCTMNIST, PneumoniaMNIST, RetinaMNIST, OrganMNIST(axial, coronal, sagittal). These datasets have been implemented using machine learning and AutoML. Rest consist of medical MNIST, skin cancer MNIST and colorectal histology MNIST. Medical MNIST consists of 6 classes – ChestCT, BreastMRI, CXR, Hand, HeadCT, AbdomenCT. Colorectal cancer histology Multiclass classification for texture analysis belonging to 8 classes of tissues. Skin Cancer MNIST contains 7 classes – Melanocytic nevi,  Melanoma, Benign keratosis-like lesions, Basal cell carcinoma, Actinic keratoses, Vascular lesions, Dermatofibroma. Different libraries have been implemented around them and can be readily used for building medical research projects. For implementation and other information -> Medical MNIST SIGN LANGUAGE MNIST Sign language MNIST was released to bring help for hearing and speaking impaired people to convey messages through hand gestures. It is similar in structure to the original MNIST in pixel dimensions and some other parameters. There are 24 classes present from A to Z except for J and Z. It is present in CSV format with labels and pixel values for each. It is developed from American Sign Language letter database. For implementation and other information -> Sign Language MNIST GOOGLE OPEN IMAGES Google has a huge open-source vision dataset which serves many purposes. Along with images it contains annotations, object relationship in images, object detection and bounding boxes, image segmentation and other recently released localized narratives. It has gone through 6 versions and currently the v6 version is in use. It is accessible through Google Cloud Vision API. Images have been crowdsourced and validated by professional annotators. Two of its most significant implementations have been seen in artistic style transfer and deep dream. For implementation and other information -> Open Images IMAGENET AND VARIANTS Imagenet is one of the greatest achievements in computer vision. Until now Imagenet is the biggest image dataset with over 14 million images spread across 20,000 different classes. Imagenet every year holds a competition on the dataset where different deep learning algorithms\/models compete to win it. With every year passing the error rates have been reduced and it’s remarkable how to have crossed the human average error rate.  Imagenet2012 (started by Fei Fei Li, later enhanced by many other researchers), thereafter many variants came over as drop-in replacement to original Imagenet namely – Imagenet2012_real, Imagenet2012_subset, Mini Imagenet, Imagenet_A & Imagenet_O, Imagenet_R, Imagenet_resized. These datasets were released along with research papers specifying their relevance. All of these have pre-built libraries to directly be used in model training. For implementation and other information -> Imagenet CIFAR 10 & 100 Cifar contains 80million tiny images dataset. Cifar-10 contains 10 object classes namely – aeroplane, bird, car, cat, deer, dog, frog, horse, ship, and truck. These images are in the form of 32×32 pixels RGB format. Cifar 100 is an extension to Cifar 10. It contains 100 object classes divided into 20 main classes- aquatic mammals, fishes, large omnivores and herbivores, medium-sized mammals, flower, food container, household electrical devices, fruit and vegetable, household furniture, insects, large carnivores, large man-made outdoor things, large natural outdoor scenes, non-insect invertebrates, people, reptiles, trees, small mammals, vehicles 1, vehicles 2. Both these datasets have an implementation in deep learning libraries. For implementation and other information -> CIFAR10 & CIFAR100 STL 10 The STL10 dataset was built inspired by the Cifar10 dataset. It is used in unsupervised learning. Divided into 10 classes – aeroplane, birds, car, cat, deer, dog, horse, monkey, ship, truck. Images are in 96×96 pixels in RGB. Total of 13000 images divided into 5000 training and 8000 test sets. It has implementations in deep learning libraries Tensorflow and PyTorch. For implementation and other information -> STL10 CALTECH DATASETS Caltech consists of 4 different datasets – Caltech 101 (containing 100 object classes of common daily use such as fans, cars, boats, lamps etc and 1 background clutter), Caltech 256 (extension to Caltech101, contains more classes and larger background clutter for testing), Caltech Birds 2010 (200 bird species) and Caltech Birds 2011(extension to Caltech Birds 2010). All these images have annotations present with bounding boxes and other information. These datasets have implementations in deep learning libraries. For implementation and other information -> Caltech","excerpt":"In this article, we will discuss the various image datasets that are readily available for training machine learning models.","categories":["AI Trends"],"tags":["caltech","Computer Vision","data science training","Data Scientist","Datasets","Deep Learning","ImageNet","MNIST","Top Trend"],"author_name":"Jayita Bhattacharyya","publish_date":"2021-01-04T13:00:00","publication_year":"2021","word_count":1379,"keywords":["Top Trend","computer vision","Data Scientist","Ray","caltech","deep learning","data science","Datasets","PyTorch","MNIST","Computer Vision","ImageNet","machine learning","AI","neural network","ML","data science training","Deep Learning","TensorFlow"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","neural network","computer vision","data science","Ray","TensorFlow","PyTorch"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/comprehensive-guide-to-image-datasets\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":8340,"title":"Top 10 Analytics Trends in 2015","content":"As we get ready to bid goodbye to 2015 and gear up to welcome another year, we can’t help but retrospect on how this year turned out to be for the analytical space and what to expect in the future. However, to plan the future it is important to look at the past, and for that matter one needs to compare. Hence, we bring you a mixed palette which makes it interesting to see how different last year analytical trends were from this year as well as what new trends may stem from these. Analytics India Magazine decided to dig out the best and most popular analytical trends by speaking to those who have experienced the eye of analytics hurricane and continue to win over it diligently. Here’s a glimpse as the data leaders from different companies share the top and emerging trends that rule analytical the industry. 1. Increased use of Digital Marketing and Social Media Analytics Suhale Kapoor (Executive Vice President and Co-Founder, Absolutdata Research & Analytics Pvt. Ltd), observed a growth in social media analytics and business intelligence through digital marketing. He remarked that digital marketing analytics expenses had increased by 60% in 2015 as branding and advertising businesses boomed. Likewise, he predicts that social media and online advertising on mobile will continue to grow as integration of offline and online customer experience is on the rise. “There is a rise in democratizing analytics through cloud and social media”, agreed Sudipta Sen (Regional Director – South East Asia, Vice Chairman and Board Member – SAS Institute (India) Pvt. Ltd.) He added that, cloud helps in making analytics available to more consumers at lower cost. In fact, businesses today derive meaningful insights from social media using the flood of conversations. By turning towards analytics to help them understand customer attitudes and identify trends, they make smarter marketing decisions. 2. Governed data discovery becomes essential James Richardson (Business Analytics Strategist at Qlik), talks about another emerging trend – Self-service Business Intelligence. With more data out there, users want to become more self-sufficient in creating their own analyses rather than relying on others, but this means they need to work in a managed data space. Within a framework of governance, users will focus their energy on getting insights from their analyses. With the ability to combine both internal and external data sources, users now have access to more context around their data, which ultimately leads to more insights and better decisions. Meanwhile, Lavanya Uppala (Practice Head of Big Data Analytics program at Bosch India), spoke about creating a unified data culture across organizations. Thereby, enabling a variety of analytical applications such as customer experience management, social media monitoring etc. 3. Personalization Sudipta Sen (Regional Director – South East Asia, Vice Chairman and Board Member – SAS Institute (India) Pvt. Ltd.) identified ‘Personalization’ as an emerging trend. Owing to the advancements in technologies combined with the avalanche of data available today, enterprises across industries are leveraging inexpensive technologies such as Hadoop to analyze huge amounts of customer data, understand patterns and subsequently personalize their offers to their customers. This in turn helps them out-think and out-do the competition. James Richardson gives us another angle as he believes “More data storytelling equals more engagement”. In fact, when making a proposal to a group, 86 percent almost always or often take time to ‘lay out what has happened previously,’ and 80 percent almost always or often take time to ‘project forward or to predict possible outcomes’.” He added that storytelling not only personalizes the task at hand, but it can also make it more memorable, impactful, and relevant for those that hear it. In 2016, there will no longer be an excuse to “take that offline.” People will use interactive storytelling to deliver information in a more compelling way that prompts them to take action in the moment, when the insight emerges. 4. Expansion Analytics for Internet of Things (IoT) This steers us to our next topic – IoT. Now, most of our experts percept that 2015 is witnessing an expansion of the ‘Internet of things’ or the IoT at high rate. Sudipta Sen, said that “More devices than ever before are being connected to the internet today. In fact with over 30 billion devices forecasted to be connected to the internet by 2020, we truly live in an age of Internet of Things (IoT).” IoT is the concept of everyday objects – from industrial machines to wearable devices, using built-in sensors to gather data and take action on it across a network. To boot, governments and organizations alike are exploring ways to leverage the vast volumes of data generated by these devices and platforms to optimize processes, create differentiated offerings and derive new revenue streams. Moreover, Rachit Ahuja (Head of Global Marketing, Ma Foi Analytics) contributed by stating that Indian analytics market space now has start-ups providing Internet of Things (IoT), machine learning and NLP based products that can crunch an entire set of big data and use algorithms to come up with accurate and tailored results. He predicts that “this is one of the interesting trends that will catch up in days to come!” 5. Analytics in Education industry: sprouting through MOOCs Another interesting trend that has been sprouting steadily was spotted by Sray Agarwal (subject matter expert of Business Analytics for TimesPro) and Lavanya Uppala. They agree that universities have started setting up data science and analytics courses as the demand in the industry has increased. Sray Agarwal remarked that Massive Open Online Courses a.k.a MOOCs are not a new concept for students and professionals. Numerous foreign universities and institutions have launched MOOCs across varied subjects from mathematics to data science. However, in the world of analytics, MOOCs have really changed the way people learn data science. Today, they are not merely recorded videos of professors\/trainers talking about data science, but they also uniquely connect with resources such as blogs, groups, discussion boards and forums. They also help hosting meet-ups for the participant for enhanced learning experience. On the other side, these MOOCs are leveraging the data collected through the participant to improve the content, quality and scope of the delivery. Needless to say, analytics is transforming every business and taking them to greater heights than ever before. Furthermore, Lavanya explained that even in India, “Many top B-schools and Universities are setting up data science and self-service data discovery courses into their academic curriculum to meet the increasing demands from the industry to produce quality talent pool.” 6. Data Visualization Moreover, James Richardson remarked that rather than just consuming information, users are now engaging in data prep and profiling. As a result, visualization is now becoming a form of self-expression. By creating visual apps, users are expressing their views and learning about themselves through being actively engaged with the growing volumes of data. You can see this trend in the rise of the quantified-self movement at an individual level and data-driven journalism in the mass media, altering how people are using public data to understand how society works. 7. Cognitive analytics continuing to drive users Ms. Lavanya (Practice Head of Big Data Analytics program at Bosch India) pointed out that context-aware cognitive analytics along with its underlying AI and NLP techniques are enhancing the robustness and accuracy to solve complicated business problems without constant human observation. Therefore, significant developments in cognitive computing can improve the quality of the decisions made for the business users. Furthermore, Rachit Ahuja contributed to the discussion by claiming that “Consumers will expect their software to anticipate their needs, driving requirements for predictive capabilities in all apps.” Thus, the goal is to expand the boundaries and our understanding of what it means to assemble and display intelligence ‘in context’. Clients are now looking for embedded predictive analytics with visualization capabilities rather than a separate visualization tool on the backend. 8. Product based analytics Rachit Ahuja affirms that product based analytics consumption is definitely on the rise. People and departments are increasingly lining up to self-service analytics software — which are easy to use and do not require any technical knowhow. Conversely, there is a growing shift from project based to outcome based engagement model. Interest in new methodologies, involving semi-structured and unstructured data, e.g. NLP, elastic search and machine learning will see significant rise. 9. Mobility – A screen in the hand is worth two on the desk. James Richardson, said that Mobility is becoming more important than ever for data users. This means that enabling multi-device lensing of BI and analytics will gain importance. For instance, 85 percent of respondents from the U.S. and 77 percent of respondents from the rest of the world complete their objectives by using multiple devices simultaneously. Having unlimited access to their data can help users ask “why?” any time, and find the answer quickly. BI and visualization solutions that don’t support users moving from device to device, often and at speed, will not deliver the kinds of experience that people want. 10. Predictive analytics in entertainment industry As we reached the last part of our discussion, Vibha Bhilawadikar (Vice president, Kiesquare) and Sray Agarwal also spoke about predictive analytics in entertainment industry. Sray Agarwal expressed that previously analytics was utilised only for business purpose and for conventional decision making. However, there has been a paradigm shift in that notion. Based on various attributes that can either be aggregated in-house or through social media, lot of insight on revenue, ratings, audience sizes, movie attendance, number of prints, pre and post release marketing campaigns and even probability of grabbing awards can be revealed using predictive analytics. However, Vibha Bhilawadikar argued that “while there can be set frameworks with specific protocols, we have to lace all the quantitative analytics with qualitative conversations and pitches to achieve the planned business outcomes.” She strongly believes, from her personal experiences of the same industry, that the business situations that we encounter cannot be negotiated with prescriptive solutioning in its complete totality. So while descriptive analytics is needed for the context of business thinking, predictive analytics is needed for building “what if” scenarios. Therefore, companies have to layer qualitative research in their strategy to accomplish their planned business goals. Thus, it is interesting to note that Analytics has touched upon all sorts of spheres this year, from education industry to entertainment, it has not restricted to any one sector. We conclude by hoping this compilation would help you reflect on what new trends would stem in 2016 and be better prepared to welcome another new year!","excerpt":"As we get ready to bid goodbye to 2015 and gear up to welcome another year, we can’t help but retrospect on how this year turned out to be for the analytical space and what to expect in the future. However, to plan the future it is important to look at the past, and for […]","categories":["AI Features"],"tags":["Absolutdata"],"author_name":"Дарья","publish_date":"2015-12-01T06:54:43","publication_year":"2015","word_count":1760,"keywords":["data science","machine learning","AI","R","RAG","NLP","Ray","Aim","analytics","predictive analytics","Absolutdata"],"extracted_tech_keywords":["AI","machine learning","NLP","data science","analytics","Aim","Ray","RAG","predictive analytics","R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/top-10-analytics-trends-in-2015\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":65202,"title":"DeepMind Compares How Children And AI Explore The World","content":"Recently, researchers at Alphabet’s DeepMind and the University of California, Berkeley proposed an AI framework for comparing child and AI agent behaviours and hence developing new exploration techniques. While developing a reinforcement learning agent, there are several questions related to the exploring behaviour that comes into the mind of researchers — how should an agent gather enough experience from different environments in order to produce optimal behaviours. According to the researchers, the issue of explorations has been considered as one of the most fundamental issues in RL. There have been a number of researchers performed regarding this issue. Despite these efforts, the issue of exploration remains unsolved. The algorithms that have achieved state-of-the-art performance on popular benchmarks such as Atari, often rely on simple exploration strategies — greedy — combined with huge amounts of computation. From A Child’s Perspective In this paper, the researchers performed direct, controlled comparisons between children and agents to leverage insights from children’s exploratory behaviour to improve the design of RL algorithms. A human possesses an innate ability to explore the environment from the day a child is born. Children are an active and curious learner who explore the surroundings thoroughly and efficiently to learn new things. According to the researchers, recent evidence suggests that children do indeed explore more than a grown-up human. This means, children tend to perform higher amounts of learning than an adult. The generalisation and rapid learning resulting from children’s exploration is indeed in contrast with what modern RL agents exhibit. There are three main reasons for comparing children’s behaviour with an RL agent. They are mentioned below: – Ecological Validity: According to the researchers, one crucial reason for gathering data from children and RL agents in the same environment is that it helps the reinforcement learning agents to be evaluated in a more ecologically valid setting, contrasting to the grid world-like settings, 2D Atari games, among others. Controlled Comparisons: Comparing children directly to reinforcement learning agents aim to provide a standard baseline for the evaluation of agent behaviour and can assist in identifying areas of promising research in deep RL. Cognitive Modelling: In addition to the above two conditions, the direct comparisons in children and RL agents give strong direction in the development of new cognitive models of behaviour, furthering the “virtuous cycle” between cognitive science and AI. Behind the Framework The researchers presented a methodology based on DeepMind Lab for directly comparing child and RL agent behaviour in simulated exploration tasks. This allowed the researchers to precisely test questions about how children explore, how agents explore, and how and why they differ. Further, using this methodology, the researchers proposed two candidate experiments designed to test key qualitative predictions of different exploration algorithms with respect to what is known about children’s exploration behaviour in other domains. The two experiments are free versus goal-directed exploration and sparse versus dense rewards. These experiments have been validated on queries such as how much children and agents are willing to explore, whether free versus goal-directed exploration strategies differ and how the reward shaping affects exploration. How It Works DeepMind Lab is a learning environment, based on the Quake game engine, that provides a suite of challenging 3D navigation and puzzle-solving tasks for learning agents. The researchers proposed a unified environment with tasks for training and evaluating both humans and agents. According to the researchers, these tasks require physical or spatial navigation capabilities to achieve and are modelled after games that children play themselves. In the experimental setup, children are allowed to interact with the DeepMind Lab environment through a custom Arduino-based controller. The controller exposes four actions that agents would use in this environment, which are moving forward, move back, move left, and turn right. Wrapping Up In both the experiments mentioned above, the researchers found out that during free versus goal-directed exploration, children’s search strategies differ between the no-goal and goal condition. Children’s behaviour is compared with a depth-first search (DFS) agent, and it showed that in the no-goal condition, children made choices consistent with DFS 89.61 per cent of the time compared to the goal condition, in which children made choices consistent with DFS 96.04 per cent of the time. In the second experiment, which is sparse versus dense rewards, the researchers found out that children are less likely to explore an area in the dense rewards condition, Read the paper here.","excerpt":"Recently, researchers at Alphabet’s DeepMind and the University of California, Berkeley proposed an AI framework for comparing child and AI agent behaviours and hence developing new exploration techniques. While developing a reinforcement learning agent, there are several questions related to the exploring behaviour that comes into the mind of researchers — how should an agent […]","categories":["AI Trends"],"tags":["ai framework","arduino","DeepMind","DeepMind AI","deepmind open source","reinforcement learning models","uc berkeley"],"author_name":"Ambika Choudhury","publish_date":"2020-05-14T11:00:00","publication_year":"2020","word_count":730,"keywords":["ai framework","Go","API","uc berkeley","AI","programming_languages:R","R","deepmind open source","programming_languages:Go","RAG","Aim","reinforcement learning models","DeepMind AI","arduino","DeepMind"],"extracted_tech_keywords":["AI","Aim","RAG","R","Go","API","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/deepmind-compares-how-children-and-ai-explore-the-world\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10166095,"title":"China’s Baidu Launches Two New AI Models, Rivals DeepSeek R1 at Half the Price","content":"Chinese tech giant Baidu launched two new AI models on Sunday: ERNIE 4.5, a native multimodal model, and ERNIE X1, a deep-thinking reasoning model. The company has also made its AI chatbot, ERNIE Bot, free for individual users earlier than planned. “As a deep-thinking reasoning model with multimodal capabilities, ERNIE X1 delivers performance on par with DeepSeek R1 at only half the price. Meanwhile, ERNIE 4.5 is our latest foundation model and new-generation native multimodal model,” the company said in a post on X. The company said that ERNIE 4.5 improves understanding, generation, reasoning, and memory while reducing hallucinations and enhancing logical reasoning and coding abilities. It is now available via APIs on Baidu AI Cloud’s MaaS platform, Qianfan. ERNIE X1 will also be accessible on the platform soon. For enterprise users, ERNIE 4.5’s input and output pricing starts at $0.55 per 1 million tokens and $2.20 per 1 million tokens, respectively. ERNIE X1’s pricing starts at $0.28 per 1 million tokens for input and $1.10 per 1 million tokens for output. Baidu plans to integrate ERNIE 4.5 and ERNIE X1 across its ecosystem, including Baidu Search and the Wenxiaoyan app. Notably, Baidu was among the first in China to launch a ChatGPT-style chatbot in early 2023, with Ernie 4.0 claiming to rival OpenAI’s GPT-4. However, adoption has lagged due to strong competition, especially from DeepSeek’s R1 model, released last month. Meanwhile,  DeepSeek plans to release its next reasoning model, the DeepSeek R2, ‘as early as possible.’ The company initially planned to release it in early May but is now considering an earlier timeline. The model is said to produce ‘better coding’ and reason in languages beyond English.","excerpt":"Baidu plans to integrate ERNIE 4.5 and ERNIE X1 across its ecosystem, including Baidu Search and the Wenxiaoyan app.","categories":["AI News"],"tags":["Baidu"],"author_name":"Siddharth Jindal","publish_date":"2025-03-16T11:37:39","publication_year":"2025","word_count":278,"keywords":["ChatGPT","API","TPU","OpenAI","AI","Modal","DeepSeek R1","GPT","Aim","Baidu","R"],"extracted_tech_keywords":["AI","ChatGPT","OpenAI","DeepSeek R1","Aim","TPU","R","API","GPT","Modal"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/chinas-baidu-launches-two-new-ai-models-rivals-deepseek-r1-at-half-the-price\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":41214,"title":"5 Ways To Land A Data Scientist Job Without Any Prior Experience","content":"The amount of data that is getting generated on a day-to-day basis is huge. That is why companies across the globe are turning data into information and are using it to optimise their strategies. But the challenge here is the fact that every company needs a professional with relevant skills to extract insights from the massive data collected — a data scientist who’s now getting a seat at the big table. Further, with the evolution of data and its increasing use in different types of business, people have started to see data science as an uber-cool job. However, when it comes to becoming a data scientist, we notice a lot of professionals have dozens of MOOC courses and fancy buzzwords on their resumes or LinkedIn profiles. And when a data science neophyte sees these portfolios, they get the impression that data science is not their cup of tea. However, that is not the case all the time — data science is about solving an actual business problem, making the best out of the cluttered data. If you have the relevant knowledge, you can kickstart your data science career without any prior experience. Here’s how: Steps To Follow There are many aspirants who want to be a part of the data science community, but they are clueless about how to get started, and there could be several reasons behind it — maybe they didn’t have a data science subject in their formal education, maybe they never attended any data science conference, maybe there are not many faculties who are much aware of the domain, etc. In this article, we are going to outline some of the important factors to bear in mind and prepare for a data science job without any prior experience. 1. Self-Assessment This the first and foremost thing to do when you are starting your data science journey and you don’t have any prior experience. Ask yourself these questions: why would a company would hire you? If they are not hiring you, what could be the reason? What do you know about the data science domain? What more do you need to know about the domain? What extra skills do you need to learn to stand out from the crowd? Further, along with the skills and knowledge a data science professionals should have, learn about the latest industry trends — how corporate works, what are the current job roles that are on demand, what are the latest programming languages etc. Make a list of all the things you know, and you need to know and make a plan for how you should go about them. 2. Skills You Need To Master Mathematics: It is also considered as one of the vital elements when it comes to data science. It is very important in the field of data science as there are many concepts that help a data scientist with algorithms. Also, concepts like statistics and probability theory are key for algorithms implementation. So, make sure you put in a lot of effort into sharpening your mathematical skills. Programming: There are many people who would suggest a huge bunch of programming languages to learn if you want to have a career in data science. However, don’t overwhelm yourself with all the hype talks. When it comes to data science Python and R are the two most important programming languages. Put in your complete focus on these two languages at the initial stage. Later, when you gain confidence along with significant confidence, you can move on to the next one (Java could be one of them). To learn to programme you can always take up short term course or online courses. Also, practice a lot. The more you code, the better coder you become. Communication & Visualisation: Having an upper hand on all the technicalities is one but to be a successful data scientist you also need to have outstanding communication and presentation skills. You should not just be a data scientist but be a data storyteller too. Why? Once you get the valuable insights from the cluttered data, your next job is to present it, and if you don’t have storytelling skills, how would make others understand what the insights are capable of and the value they would deliver. 3. Practice With Real-Time Problem Statement Learning and mastering skills are definitely mandatory, but to make the most out of your learning, you need to practice — practice with real-time problem statements you give your data science learning a worth.  The more you solve those problems the more you gain experience as well as confidence and makes the pathway to your dream science job short. There are many hackathons available on the internet — you can always pick one, participate and see where you stand in this ever competitive data science domain. 4.Connect  With Leaders It is always considered to be a good practice to take advice from someone who has already mastered domain. And for that, you can make the best use of platforms like LinkedIn to connect with some of the leaders from the industry. Another best ways to make connections is by attending data science conferences, where you not only get to attend talks and masterclasses but also meet a lot of people from the industry who would help you take a right path when you are starting with your data science journey. 5. Accept Reality It is no surprise that data science is one of the highest paying and reputed jobs right now in the industry. And no company would pay someone a handsome paycheck and give a high-level designation until and unless they prove that s\/he is capable of dealing with and some of the complex business problems. So, accept the fact that when you initially start your career, you might not even get the designation as a data scientist (you might get in some exceptional cases). However, if you are determined and learn more and more about the domain, the chance of you getting to a higher position with a significantly high paycheck increases. Make sure you don’t hesitate to seek help from fellow data scientist when you need. Knowledge and skills are the master keys to success.","excerpt":"The amount of data that is getting generated on a day-to-day basis is huge. That is why companies across the globe are turning data into information and are using it to optimise their strategies. But the challenge here is the fact that every company needs a professional with relevant skills to extract insights from the […]","categories":["AI Hirings"],"tags":["Cypher","Data Science","Data Science Career","data science job","Hackathon","Machinehack","Machinehack Hackathon"],"author_name":"Harshajit Sarmah","publish_date":"2019-06-24T10:50:21","publication_year":"2019","word_count":1034,"keywords":["Cypher","data science","Go","programming_languages:R","Machinehack","AI","programming_languages:Java","Data Science Career","Hackathon","Python","ViT","Machinehack Hackathon","programming_languages:Python","Data Science","R","Java","data science job"],"extracted_tech_keywords":["AI","data science","Python","R","Go","Java","ViT","programming_languages:Python","programming_languages:R","programming_languages:Java"],"url":"https:\/\/analyticsindiamag.com\/ai-hiring\/5-ways-to-land-a-data-scientist-job-without-any-prior-experience\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10068876,"title":"Microsoft launches MLOps v2","content":"Microsoft recently announced the Beta #2 version of MLOps v2 to simplify your MLOps workstream with a unified solution accelerators available on GitHub repository. The full release is targeted for July 2022. Check out the GitHub repository here. MLOps v2 will allow AI professionals to deploy an end-to-end standardised and unified Machine Learning lifecycle scalable across multiple workspaces. By abstracting agnostic infrastructure in an outer loop, the customer can focus on the inner loop development of their use cases. MLOps v2 is a set of workflows that intends to serve as the initial point for MLOps implementation in Azure. MLOps v2 provides a templatized approach for the end-to-end Data Science process and focuses on driving efficiency at each stage. Currently, the general customer struggle is standing up an end-to-end MLOps engine due to resource, time, and skill constraints. One main issue is that it often takes a significant amount of time to bootstrap a new Data Science project. The MLOps v2 provides templates that can be reused to establish a “Cookie-Cutter-Approach” for the bootstrapping process to shorten the process from days to hours or minutes. The bootstrapping process encapsulates key MLOps decisions such as the components of the repository, the structure of the repository, the link between model development and model deployment, and technology choices for each phase of the Data Science process. The MLOps v2 architectural pattern is made up of four modular elements representing phases of the MLOps lifecycle for a given data science scenario, the relationships and process flow between those elements, and the personas associated with ownership of those elements. Solution accelerator The solution accelerator provides a modular end-to-end approach for MLOps in Azure based on pattern architectures. As each organization is unique, solutions will often need to be customized to fit the organization’s needs. The solution accelerator goals are: SimplicityModularity Repeatability Collaboration Enterprise readiness (Source: Microsoft) MLOps v2 is the de-facto MLOps solution for Microsoft on forward. Aligned with the development of Azure Machine Learning v2, MLOps v2 gives you and your customer the flexibility, security, modularity, ease-of-use, and scalability to go fast to product with your AI. MLOps v2 not just unifies Machine Learning Operations at Microsoft, even more, it sets innovative new standards to any AI workload.","excerpt":"MLOps v2 will allow AI professionals to deploy an end-to-end standardised and unified Machine Learning lifecycle scalable across multiple workspaces.","categories":["AI News"],"tags":[],"author_name":"AIM Media House","publish_date":"2022-06-13T14:35:41","publication_year":"2022","word_count":374,"keywords":["data science","Go","machine learning","AI","R","ML","MLOps","Scala","Git","Azure"],"extracted_tech_keywords":["AI","machine learning","ML","data science","MLOps","Azure","R","Go","Scala","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/microsoft-launches-mlops-v2\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10163792,"title":"How Apprenticeships Are Helping India’s GCCs Face Talent Crunch","content":"India has emerged as the epicentre of global transformation through its Global Capability Centres (GCCs), housing more than half of them in the world. With over 1,700 centres today and projections to surpass 2,975 by FY2024, the demand for a skilled and future-ready workforce has never been greater. Yet, amid this rapid expansion, a critical challenge looms – bridging the growing skill gap in emerging technologies. Traditional education systems often struggle to keep pace with industry demands, leaving businesses grappling with a shortage of AI\/ML engineers and other specialised professionals. According to LinkedIn’s Workplace Learning Report 2025, nearly half of learning and talent development professionals face a skills crisis today. Of these, 49% agree that “executives are concerned that employees do not have the right skills to execute business strategy”. Source: https:\/\/economicgraph.linkedin.com\/workforce-data?selectedFilter=view-all%2Fby-year By integrating apprenticeship programs into their hiring and talent development strategies, GCCs are not only training employees but also shaping the future of work. Bridging the Gap Through Apprenticeships In an interview with AIM, Dhriti Prasanna Mahanta, vice president and business head of TeamLease Degree Apprenticeship, mentioned that apprenticeship programs enable businesses to nurture talent from the ground up, ensuring a sustainable workforce while minimising recruitment costs and turnover. He further mentioned that apprenticeships also support long-term employee engagement and retention by providing clear career progression opportunities, fostering loyalty and reducing the skill mismatches that can hinder growth in the fast-paced and technology-driven environment of GCCs. “The demand for AI\/ML talent has already surged by 400% in 2024, and apprenticeship programs offer a viable solution to rapidly upskill the workforce, meeting the urgent needs of both businesses and the wider tech ecosystem,” Mahanta mentioned. According to him, these initiatives provide hands-on learning experiences that combine theoretical knowledge with real-world application, making candidates job-ready from day one. “Apprenticeships are not just about learning; they prepare individuals for real-world challenges,” he added. Unlike conventional degree programs, apprenticeships equip professionals with practical problem-solving skills, allowing them to integrate seamlessly into the workplace. How GCCs are Future-Proofing Their Workforce By investing in AI-focused apprenticeships and internship programs, GCCs are creating a steady pipeline of highly skilled professionals. These initiatives also help companies reduce hiring costs, improve employee retention, and foster a culture of continuous learning. Mahanta highlighted that over the years, TeamLease has worked closely with a wide range of GCCs across India, delivering tailored skilling solutions. “Our clients come from various industries, including technology, automotive, finance, healthcare, and manufacturing, ranging from mid-sized enterprises to Fortune 500 companies.” Meanwhile, Pankaj Vyas, CEO and MD of Siemens Technology and Services India, said that the hiring landscape has undergone a significant shift over the past decade. He explained that what was happening was very interesting. He mentioned that if the question had been asked seven to eight years ago, the focus would have largely been on the industry and domain. However, now the emphasis has shifted to the technology hiring that companies are doing. He also highlighted the growing emphasis on reskilling and upskilling the existing workforce to keep pace with evolving industry demands. “In the business that we are in, it’s not just about technology. It’s equally important to have deep domain expertise,” he concluded. Moreover, Dhanya Rajeswaran, global vice president and country managing director at Fluence India, also stressed the need for an agile and highly skilled workforce to support their rapid growth and innovation. “We’re very young, and, therefore, we needed a workforce that was very agile and talented and with the ability to shorten timelines for us,” she explained.In this context, Rajesh Puneyani, vice president of technology and site leader at Kenvue Solutions India, mentioned that AI-driven hiring processes, such as automated screening and data-driven candidate assessments, are areas the company is exploring. This helps streamline hiring while ensuring high-calibre talent that aligns with its vision.","excerpt":"By integrating apprenticeship programs into their hiring and talent development strategies, GCCs are not just training employees but shaping the future of work.","categories":["GCC"],"tags":["GCC","Talent Crunch"],"author_name":"Shalini Mondal","publish_date":"2025-02-16T10:00:00","publication_year":"2025","word_count":636,"keywords":["Go","API","GCC","programming_languages:R","AI","RPA","ML","data-driven","innovation","Talent Crunch","Aim","R"],"extracted_tech_keywords":["AI","ML","Aim","R","Go","API","RPA","data-driven","innovation","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/gcc\/how-apprenticeships-are-helping-indias-gccs-face-talent-crunch\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":62438,"title":"New Microsoft 365 Version With AI-Driven Content Now In India","content":"Microsoft has announced that the company is bringing out a new version of Microsoft 365 for the Indian market that will provide users with artificial intelligence-driven content. This new Microsoft 365 version will offer content related to artificial intelligence like video chats, templates and cloud-powered experiences. This new version of Microsoft 365 Personal will be priced at ₹ 420 per month for an individual and will allow a family of up to six people to use it for ₹ 530 per month. This new Microsoft 365 version will include features like premium desktop Office apps, 1TB of OneDrive cloud storage per person, 60 Skype minutes for calling mobile phones and landlines, advanced security features to protect users from malware and phishing attacks and ongoing technical support. The company further announced the subscription of Microsoft 365 for family and personal use will be soon rolled out regionally and is expected to reach over 38 million Office 365 subscribers globally in the next few months. According to Yusuf Mehdi, the corporate vice president of Windows and Devices Group, “The Microsoft 365 family and personal subscriptions will provide users with innovative experiences that enable them to co-author, video chat, organise, and come together with friends and families anytime and anywhere.” Additionally, the new version of Microsoft 365 will include a Microsoft Editor, which is an AI-powered service available in more than 20 languages. This will also be accessible across Word and Outlook.com, and as a standalone browser extension for Microsoft Edge and Google Chrome. Among other improvements is the AI-powered Presenter in Microsoft PowerPoint, which has been designed to help correct monotone pitch and refine speeches — will be available as a free preview for all, and then eventually only to Microsoft 365 subscribers. Further for the new version, in Excel, users will have new data types and smart templates to interact with data, providing deeper meaning for over 100 topics. Microsoft 365 subscribers have exclusive access to the over 100 new data types powered by Wolfram Alpha. According to the company, Skype has seen an increase in usage with 40 million people using it daily, up 70% month over month. The company stated, “We are seeing a 220% increase in Skype to Skype calling minutes month over month.” A Microsoft Family Safety app has also been designed in order to keep families safe. The company said it would also launch Teams with new features later this year for users to stay connected with their family and friends.","excerpt":"Microsoft has announced that the company is bringing out a new version of Microsoft 365 for the Indian market that will provide users with artificial intelligence-driven content. This new Microsoft 365 version will offer content related to artificial intelligence like video chats, templates and cloud-powered experiences. This new version of Microsoft 365 Personal will be […]","categories":["AI News"],"tags":["Content Analysis by AI","Microsoft","Microsoft 365"],"author_name":"Sejuti Das","publish_date":"2020-04-23T13:30:00","publication_year":"2020","word_count":415,"keywords":["Go","artificial intelligence","programming_languages:R","AI","programming_languages:Go","RAG","Microsoft 365","GAN","Content Analysis by AI","R","Microsoft"],"extracted_tech_keywords":["AI","artificial intelligence","RAG","R","Go","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/new-microsoft-365-version-with-ai-driven-content-now-in-india\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10001307,"title":"WhatsApp Turns 10: Here’s A Tech Timeline Of The Most Famous Messaging Platform","content":"According to a recent 2019 study, WhatsApp, the messaging app owned by social network giant Facebook, has 1.5 billion users in the world, out of which, India has 200 million users. With a rapid increase of its users, WhatsApp is undoubtedly the most popular messaging app worldwide today. The app today celebrates 10 years of its presence. Here is a look at their tech journey and updates over the last decade. From encryption to location sharing to status, check out the all the updates WhatsApp introduced in the last 10 years. Also, check out the upcoming features on the anvil. The Journey 2009: The start of WhatsApp was primarily as a messaging application for mobile phone users, invented by Jan Koum and Brian Acton who were former employees of Yahoo!Both the founders. Koum began to explore the possibility of creating an app that would let mobile users have a better interaction and engage with their friends, family, and business contacts. Teaming up with Brian Acton, Koum managed to persuade five associates from Yahoo! to fund the app with $250,000, and in 2009 WhatsApp was launched. By the end of the year it was also an image and video sharing platform. 2010: They introduced a feature to share location by its users and send it to people to know about their whereabouts and know that they were safe. This was the first new, notable feature of WhatsApp since its inception and provided a lot of convenience to its users. 2011: This year, WhatsApp introduced the WhatsApp Web, where users could use their WhatsApp registered on their phone numbers on the laptops and desktops, adding one more convenience. WhatsApp group feature was introduced this year. WhatsApp groups is widely used for many purposes, whether it is a group project or a trip to go to, it comes with a lot of communication convenience. 2013: It introduced its voice messaging feature where users could use the mic on WhatsApp and send it as a voice message to their contacts. 2014: This was when WhatsApp became a member of Facebook. It also made available the feature of read receipts, which created a lot of buzz. A Quora answer explains that there are three parts of the Whatsapp: sender, receiver and server, with multiple servers. It has a single tick when the signal is received. When the server sends message to the receiver, it waits for the server to come online before sending the message and the algorithm has a double tick underneath it. For blue ticks, WhatsApp, on opening a chat, captures the signal and sends the ‘seen’ signal to the server and then to the sender, eventually giving a blue tick to the message, o the receiver. 2016: One of the unique features of 256-bit end-to-end encryption was introduced in 2016. WhatsApp end-to-end encryption ensures only you and the person you’re communicating with can read what’s sent, and nobody in between, not even WhatsApp. Your messages are secured with locks, with every message having its own unique lock and key, and only the recipient and you have the special keys needed to unlock and read your messages. The following steps describes the working of it: 1.When the user first opens the WhatsApp, two different keys; public & private, are generated. The encryption process takes place on the phone itself. 2.The private key must remain with the user whereas the public key is transferred to the receiver via a centralised WhatsApp server. 3.Public key encrypts the senders message on the phone even before it reaches the centralised server. 4.The server is only used to transmit the encrypted message. The message can only be unlocked by the private key of the receiver. No third party, including WhatsApp can intercept and read the message. They also had a desktop app of WhatsApp in 2016. Towards the end of 2016, WhatsApp introduced the much-anticipated video calling feature to take on Skype and Google Hangouts. 2017: WhatsApp introduced the feature of WhatsApp status, which was very similar on the social networking site of Snapchat. 2018: WhatsApp Business was introduced this year. WhatsApp Business also supports WhatsApp Web, wherein you can manage the service online, without a mobile app. These are the finer points that set WhatsApp Business apart from the standard app. It is built from the ground up for businesses or individuals running their own setup, to cater to client needs. It’s also the very reason the app exists, because it offers so much more than what the standard app can offer for those currently doing business on the regular app. WhatsApp group calling was also introduced this year. 2019: WhatsApp completes 10 years. Upcoming Features This Year Some of the upcoming features of WhatsApp or this year are WhatsApp Dark Mode, WhatsApp Add Contact feature, WhatsApp QR code, WhatsApp ranking for contacts and Private Reply","excerpt":"According to a recent 2019 study, WhatsApp, the messaging app owned by social network giant Facebook, has 1.5 billion users in the world, out of which, India has 200 million users. With a rapid increase of its users, WhatsApp is undoubtedly the most popular messaging app worldwide today. The app today celebrates 10 years of […]","categories":["AI Features"],"tags":["business"],"author_name":"Disha Misal","publish_date":"2019-02-27T16:51:46","publication_year":"2019","word_count":809,"keywords":["Go","API","programming_languages:R","AI","programming_languages:Go","business","GAN","R"],"extracted_tech_keywords":["AI","R","Go","API","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/whatsapp-turns-10-heres-a-tech-timeline-of-the-most-famous-messaging-platform\/","complexity_score":4,"technical_depth":7,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10160879,"title":"‘ChatGPT is Cosmos DB’s Largest User&#8217;","content":"Microsoft CEO Satya Nadella revealed at the Microsoft AI Tour in Bengaluru that OpenAI’s ChatGPT is the largest user of Cosmos DB, Microsoft’s database service, designed to build scalable and AI-ready apps. “If you look at ChatGPT, they are some of the biggest users of Cosmos because [it] serves as a stateful application,” Nadella said. This underscores Cosmos DB’s role in managing extensive real-time interactions and session data required by ChatGPT, a chatbot that handles millions of simultaneous conversations. Nadella emphasised cloud-native databases such as Cosmos DB, SQL Hyperscale, and Fabric for modern AI workloads. He said that AI doesn’t sit on its own and, instead, relies on a comprehensive compute stack to function effectively. Cosmos DB’s scalable architecture supports high-throughput, low-latency demands during both AI training and inference, thereby ensuring responsive and stateful interactions for applications like ChatGPT. Linking AI’s progress to Moore’s Law, Nadella noted that scaling laws remain strong for both training and inference and highlighted the significance of test-time compute”, where databases like Cosmos DB must handle high-throughput and low-latency demands post-training. He pointed out that merging operational and analytical data in the cloud enables AI-driven apps to access high-volume data reliably and in real time. Nadella attributed ChatGPT’s success to the foundational role of databases in AI and said that Cosmos DB integrates operational and analytical data to provide reliable, high-volume access in real time. According to the company, this focus on robust data infrastructure aligns with its broader strategy, including a $3 billion investment in Azure data centres across India, where the aim is to expand global capacity to meet rising AI demands. These investments reinforce Cosmos DB’s position as a backbone for next-generation AI applications. At last year’s Microsoft Build, the company introduced Microsoft Fabric, real-time intelligence to its AI analytics platform. This update integrated SQL Server into Fabric databases and merged operational and analytical data within a single unified platform.Moreover, the data API builder for Azure Cosmos DB was released to the public. This open-source tool requires no coding to set up secure GraphQL endpoints.","excerpt":"The focus on robust data infrastructure aligns with its broader strategy, including a $3 billion investment in Azure data centres across India.","categories":["AI News"],"tags":["ChatGPT","OpenAI"],"author_name":"Aditi Suresh","publish_date":"2025-01-07T15:55:59","publication_year":"2025","word_count":343,"keywords":["ChatGPT","OpenAI","AI","AWS","R","Scala","Aim","analytics","SQL","Azure"],"extracted_tech_keywords":["AI","analytics","ChatGPT","OpenAI","Aim","AWS","Azure","R","SQL","Scala"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/chatgpt-is-cosmos-dbs-largest-user\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":31959,"title":"A Day In The Life Of: Uber Techies Who Work With Big Data","content":"As the year 2018 draws to an end, Analytics India Magazine is starting a new column called ‘A Day In The Life Of’, where we will try to step into the shoes of the awesome techies from various organisations who are working in New Tech areas like big data, data analytics, artificial intelligence and the internet of things, among others. This week we decided to talk to two women engineers from Uber. Pallavi Rao, Staff Engineer, and Divya B, Senior Software Engineer, spoke to AIM about their life as two dedicated techies who are working in a top company, in an area which is constantly changing and evolving. Pallavi, whose day begins at 6:30 am, rain or shine, says that her philosophy towards life is that she considers each day a brand new challenge. She applies the same rule for her work at Uber, “Each day has a brand new technical problem to solve and, I enjoy solving problems,” she says cheerfully. Divya, who efficiently juggles work and play, says: My mornings are dedicated to self-care (I work out every single day), and my evenings are dedicated to my family. As Pallavi walks us through her day, she explains, “My job is to chart out a technical roadmap for the team given the various product requirements and keeping in mind the design principles and good engineering practices. My expertise is in big data systems and machine learning. I work across various projects within the group helping and guiding engineers where needed.” Her team’s job is to build technology that helps Uber marketers reach the right set of audience, at the right time, with the right message. “Ensuring our marketing spend gives us the best ROI is another goal our team helps to achieve. This involves building systems that ingest data from multiple sources, process them churn GBs or data in an hour,” explains Pallavi. Divya’s key responsibilities, on the other hand, include: Making technology choices Design and architecture of big data systems Building big data and analytics systems Coding\/programming Mentoring teams Setting up a process and good coding practices She is currently working with the AdTech team. Divya explains, “Uber is spending a huge amount of money in advertising across hundreds of channels throughout the world. Technology built out of AdTech team makes Uber spend efficiently using data analytics and machine learning. It involves processing data at scale using big data systems.” Both Divya and Pallavi are working mothers and one can see the meticulous planning that has gone into their juggling all these roles efficiently. For example, Pallavi, who eats her meals at the Uber cafeteria, prefers to eat home-cooked food on the weekends. She also plans her free time in such a way, that it covers all aspects of her vibrant personality. “I would say I have an active social life. I am an extrovert, but not really a party animal. I occasionally go for movies, dinner, games or even treks. However, I like my quiet times too, where I can just curl up with a book or just stay indoors and play a board game with my family. I also play basketball with my son or go for walks with my husband.” Painting a vibrant picture of the Uber team in India, Divya says happily that the best part of her workday includes “lunch talks” with the team. On a serious note, she says: When I solve a problem in a better way than the existing solution if there was a pre-existing solution, totally makes my day. When asked about their long-term plans, both of them had a clear idea. Pallavi says, “My five-year plan is to work on company-wide initiatives and drive the technology charter across various sub-orgs. But for now, I want to make the AdTech team successful by bringing in innovative and cutting-edge technology solutions and setting it up for long-term success. Also, work with rest of the engineering leaders at Uber to make this [Bengaluru] site a world-class technology centre.” Divya says, “Over the next five years I want to move up in my career and see myself in a position that creates a positive impact across the company. My short-term plan is to work on challenging projects that make a difference to my organisation, and which can span across other companies as well.” Also see:","excerpt":"As the year 2018 draws to an end, Analytics India Magazine is starting a new column called ‘A Day In The Life Of’, where we will try to step into the shoes of the awesome techies from various organisations who are working in New Tech areas like big data, data analytics, artificial intelligence and the […]","categories":["AI Features"],"tags":["Bengaluru","Big Data","Data Science","Data Science Career","Interviews and Discussions","Uber"],"author_name":"Prajakta Hebbar","publish_date":"2018-12-21T07:01:00","publication_year":"2018","word_count":723,"keywords":["big data","Go","artificial intelligence","machine learning","AWS","AI","GAN","Data Science Career","Aim","analytics","Uber","Data Science","Big Data","R","Bengaluru","Interviews and Discussions"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","analytics","Aim","AWS","R","Go","big data","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/a-day-in-the-life-of-uber-techies-who-work-with-big-data\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10045205,"title":"Complete Guide To Descriptive Statistics in Python for Beginners","content":"In the modern era, everything is now a data and data-driven system. Every second amount of data being generated is in terabytes. To draw meaningful statements and conclusions with numerical evidence, a widely used mathematics framework called Statistics is used. So what is meant by statistics? It means collection, organization, analysis and interpretation of data. Statistics are mainly used to give numerical conclusions. For example, if anyone asks you how many people are watching youtube, in this case, we can’t say more; many people are watching youtube, we have to answer in numerical terms that give more meaning to you. We can say like during weekdays 6 pm-8 pm more people are watching youtube applications and during weekend 8 pm-11 pm. If you want to answer active users, we can say there are two billion+ monthly active users, in the same way; the users spend a daily average of 18 minutes. This is the numerical way to conclude the questions, and statistics is the medium used to make such inference. Statistics include; Design of experiments: Used to understand Characteristics of the dataset  Sampling: Used to understand the samplesDescriptive statistics: Summarization of data Inferential Statistics: Hypothesis way of concluding dataProbability Theory: Likelihood estimation Why do we have to learn Statistics? We are learning statistics because we can; observe the information properly, draw the conclusion from the large volume of the dataset, make reliable forecasts about business activity and improve the business process. To do all kinds of these analyses, statistics are used. Further, it is classified into two types: Descriptive and Inferential statistics. Descriptive statistics summarize the data by computing mean, median, mode, standard deviation likewise. It is distinguished from inferential statistics by its aim to summarize the sample rather than use the data to learn more about the Population that; the sample of data thought to represent this means it is not developed based on probability theory. Whereas inferential statistics are the methods for using sample data to make general conclusions (inferences) about populations by using the hypothesis. The sample is typically part of the whole population which contains only limited information about the population. For example, you might have seen the exit poll; those exit polls are calculated by taking several samples from different regions of that territory. Such conclusions are drawn from inferential statistics. Descriptive statistics: In statistical analysis, there are three main fundamental concepts associated with describing the data: location or Central tendency, Dissemination or spread, and Shape or distribution. A raw dataset is difficult to describe; descriptive statistics describe the dataset in a way simpler manner through; The measure of central tendency (Mean, Median, Mode) Measure of spread (Range, Quartile, Percentiles, absolute deviation, variance and standard deviation) Measure of symmetry (Skewness) Measure of Peakedness (Kurtosis) Let’s see the above one by one by leveraging Python; Code Implementation: Basic Statistics In Python import pandas as pd import matplotlib.pyplot as plt dataset = pd.read_excel('3. Descriptive Statistics.xlsx',sheet_name=0) Measures of central tendency: The goal of central tendency is to come with the single value that best describes the distribution scores. There are three basic measurements used i.e, mean(the average value) , median(the middle value), mode(the most frequent value). Let’s calculate the central tendency for the above example Mean: The arithmetic average of some data is the average score or value and is computed by simply adding all scores and dividing by the number of scores. It uses information from every single score. Here we are using python library pandas functionality to calculate most of our statistical parameters, so we don’t need to write code from scratch; it is just a matter of a few lines of code; dataset[['CurrentSalary', 'After6Months', 'SalBegin']].mean() The above says from 475 employees the average salary at beginning, After the six months and current as above.  There are multiple types of means, such as weighted mean, trimmed mean but this is the most common use of mean. Median: Whenever we need to find a middle value, we go for the Median to calculate the median; we need to arrange values in ascending order. The median also attempts to define a typical value from the dataset, but unlike the mean, it does not require calculation, but it is a precaution while calculating the median like as; If there are odd numbers of observations present in your dataset, then the median is the simple middle value of the ascending order of a particular column. If there are even numbers of observations present, then the median value is the average of two middle values. As we are using the Pandas library for the calculation, these precautionary things are handled automatically; as the methodology is concerned, we should know all these things. dataset[['CurrentSalary', 'After6Months', 'SalBegin']].median() The above values suggest at least half of the observations should have the current salary less than the 28875, in the same way, we conclude for the other two. Mode: The mode is used as the value that appears more frequently in our dataset. The institution of mode is not as immediate as mean or median, but there is a clear rationale. The mode value is usually being calculated for categorical variables. We can calculate mode by simply using .mode() to the pandas data frame object. The below is another way of calculating mode. from collections import Counter job_time = dataset['Job Time'].values data = dict(Counter(job_time)) mode = [k for k, v in data.items() if v == max(list(data.values()))] mode The above code gives mode values like 93 and 81; this is a bit confusing right! This is because we have a tie between 93 and 81. After all, they are occurring in the same number. These are all concepts in Measure of Central tendency. The measure of dispersion (spread): In this, we have different concepts such as Range, Standard Deviation, Variance, Quartile. However, it mainly tells how data is spread from the center, nothing but mean median, mode. The range is nothing but the largest value subtracted from the lowest value. It ignores the effect of outliers, considers only two points in its estimation and does not recognize data distribution. Next is deviation; the deviation is calculated to know how values have deviated from the mean. We can calculate the deviation for any central measures, i.e. mean, median, mode. While calculating deviation, we have to ignore negative values and consider them as positive. Quartile means quarterly basis calculations. Quartiles of distribution are the three values that split the data into four equal parts like as below where Q1 is 25th percentile, Q2 is 50th percentile, and Q3 is the 75th percentile; Source The Interquartile range(IQR) is a measure that indicates the extent to which the central 50% values within the dataset are dispersed. It is calculated as Q3-Q1. As far as dealing with outliers is concerned, IQR can be used to impute the outliers values. Next is variance, used mainly to find variation in the dataset. Variance indicates how close to or far from the mean are most of the values from a particular variable, and the standard deviation of the square root of the variance gives the magnitude of the variance. In other words, the standard deviation is used to check the consistency of the data lower the high-value consistency is there. To calculate all the parameters under a measure of dispersion, we can code individually for all the parameters or use the NumPy package to do so. Here, as we are dealing with the data frame, the pandas .describe() function gives all of the parameters we need. dataset[['Education', 'JobCategory', 'CurrentSalary', 'After6Months', 'SalBegin', 'Job Time', 'Prev Exep']].describe(include='all') The measure of symmetry and peakedness: Skewness is the measure of symmetry or, more precisely, the lack of symmetry. For example, a distribution or dataset is symmetric if it looks the same to the left and right of the data of the centre point. Whereas kurtosis is the measure of whether the data are heavy-tailed or light-tailed relative to normal distribution. This means datasets with high kurtosis tend to have more data points on either side. The histogram is an effective graphical way to show both skewness and kurtosis; plt.figure(figsize=(7,5)) plt.hist(dataset.CurrentSalary) The data is +vely skewed from the above histogram, i.e. mean >  median; it has a heavy tail which means data contains more outliers. To see the distribution of the outliers, we can scatter plot or box plot, which gives a clear representation of data. Boxplot can be treated as an option to the histogram as it gives nearly all the information as histogram gives. plt.figure(figsize=(7,5)) plt.boxplot(dataset.CurrentSalary) Conclusion: This was all about the basics of statistics, especially descriptive statistics. The above concepts of descriptive statistics help in: identifying potential data problems such as error outliers, identifying the process issues, selecting appropriate statistical tests for understanding the underlying relationship or pattern. Here we have used the pandas package to calculate all the parameters. We can also use packages like NumPy, math modules or can code it from scratch. References: Dataset usedStatistics using pythonLink for above code","excerpt":"We are learning statistics because we can; observe the information properly, draw the conclusion from the large volume of the dataset, make reliable forecasts about business activity and improve the business process. To do all kinds of these analyses, statistics are used. Further, it is classified into two types: Descriptive and Inferential statistics.","categories":["Deep Tech"],"tags":["Data Analytics","Data Science","Guide","Statistics for Data Science"],"author_name":"Vijaysinh Lendave","publish_date":"2021-08-04T14:00:00","publication_year":"2021","word_count":1497,"keywords":["Go","NumPy","AI","RAG","Python","GAN","Aim","Statistics for Data Science","Data Analytics","Data Science","Matplotlib","R","Guide","Pandas"],"extracted_tech_keywords":["AI","Aim","Pandas","NumPy","Matplotlib","RAG","Python","R","Go","GAN"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/complete-guide-to-descriptive-statistics-in-python-for-beginners\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":48203,"title":"Why MLDS 2020, India’s Industry Leading Event, Matters For Machine Learning Community","content":"The industry’s leading event for the developer community is back in January 2020. A year further into the machine learning ecosystem and MLDS 2020 returns with two packed, insightful editions in Bangalore (22-23 Jan) and Hyderabad (30-31 Jan). Last year’s inaugural summit held in January at NIMHANS Convention Centre saw the venue at full capacity with 700+ delegates and 300+ industry-leading companies like Microsoft, Intel, ABInBev, Amazon, Walmart, SAP Labs India Pvt Ltd, Capgemini, Cognizant, AXA XL, Flipkart, Philips Research, AWS, Capital One and Oracle Cloud Solution Hub, among others, participating in the summit to track the developments in this fast-paced machine learning space. MLDS was specifically launched to fill the gap between machine learning applications and how it fits into AI value chain with industry leaders, IT Decision Makers and evangelists discussing the newest progress on ML, how companies are designing AI strategies for their value chains, strategies on acquiring data to train models and how the technology was scaled. But more so, the conference addresses the unique needs of developer ecosystem by sharing insights and best practices on topics ranging from AI\/ML innovations to reducing the model bias, open-source deep learning and how to drive the most out of your cloud environment. Companies are experimenting with adopting AI\/ML technologies across their value chains and as emerging technologies enter the enterprises — there is a greater thrust on developers and machine learning engineers to apply ML beyond support functions into core business processes. MLDS provides data scientists, ML engineers and practitioners an opportunity to interact with industry-leading platforms powering the most advanced ML capabilities to build the next generation of applications. In more ways than one, MLDS has become a pivotal point for exponential learning, industry networking, interaction with tools like Azure AI solutions, AWS, Intel Developer tools and engage with the developer community and understand where the industry is heading. Industry Thought Leaders Share Their Vision On AI & Machine Learning Lakshmi Narasimhan of Intel at MLDS 2019 Learn about hard-earned lessons in ML: As tech companies and organisations outside the technology domain double down on AI\/ML, we are beginning to see more enterprises setting up teams of machine learning engineers, modellers and data scientists. This is an excellent opportunity to understand fast-paced developments happening in this industry, learn about hard-earned lessons about ML and who knows, maybe land the next job from networking at MLDS. Learn why transparency and accountability counts in machine learning: Looking ahead to 2020, there is a greater need to raise awareness on the importance of fairness in AI training systems, the impact of bias on models and the critical questions plaguing ML engineers ⁠— do they need to go beyond simply business metrics to factor in the goodness of AI. Organisations like Microsoft treat these issues seriously and ML practitioners can gain practical insights on how to address issues around bias, transparency and accountability and steps the model builders should take to avoid bias. Enabling humans-in-the-loop to do more: The holy grail of machine learning and AI has been “democratisation” to enable humans-in-the-loop to put prototypes into production. This is an area where tech giants Google, Microsoft, Amazon, Intel lead the market. But the real enablers who can drive these machine learning tools forward are ML engineers, data scientists and practitioners — companies are building a suite of developer tools to accelerate time-to-market and reduce complexities in ML systems. At MLDS 2019, Intel demoed computer vision applications built with OpenVINO toolkit. Cloud environments like Azure and AWS augment the data scientists’ toolkit but providing tools to deploy prototypes into production. Wrapping Up Sandeep Alur of Microsoft at MLDS 2019 The last summit addressed a wide variety of issues the ML community faces — for example the need for the best-tuned or best-performing model, why it is important not to lose track of critical business metrics and which are the best developer tools that make it easy to apply algorithms. From industry-leading statements around products to Technical Deep Dives with expert developers and startups changing the face of AI landscape, MLDS 2020 is a great platform for exchanging ideas and progress in this fast-evolving field.","excerpt":"The industry’s leading event for the developer community is back in January 2020. A year further into the machine learning ecosystem and MLDS 2020 returns with two packed, insightful editions in Bangalore (22-23 Jan) and Hyderabad (30-31 Jan). Last year’s inaugural summit held in January at NIMHANS Convention Centre saw the venue at full capacity […]","categories":["Deep Tech"],"tags":["machine learning conference"],"author_name":"Richa Bhatia","publish_date":"2019-10-18T10:47:04","publication_year":"2019","word_count":692,"keywords":["machine learning conference","Go","machine learning","AWS","AI","R","ML","computer vision","deep learning","Rust","Azure"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","computer vision","AWS","Azure","R","Go","Rust"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/why-mlds-2020-indias-industry-leading-event-matters-for-machine-learning-community\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":6,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":11212,"title":"ACTIVE.AI raises $3 Mn from Kalaari Capital and IDG Ventures India: Aims to expand its AI platform for Banking &#038; Payments","content":"Founders- Active.ai Active.ai, a Singapore based fintech startup recently announced raising a fund of $3 million from leading venture capital funds IDG Ventures India and Kalaari Capital. Active.ai uses artificial intelligence to deliver virtual assistant banking services to its customers. It’s intelligent and omni-channel platform enables financial service companies such as banks, wealth managers and asset managers to engage with their customers over on bots, SMS or voice API’s. This investment comes as a support to enable rapid growth of Active.ai to continuously advance its cutting-edge platform and to build out advanced AI features, enabling an increasing roster of clients to deliver a superior and engaging banking experience. Just imagine you are travelling and have lost your credit card, and you urgently need to reach your bank. What would you do? Would you call the bank’s call center and wait in an endless IVR? What if you are able to use your favorite messaging channel and send an “I lost my card” message to your bank? And your bank can converse with you intelligently using an automated virtual assistant, block your card and issue you a new one! That is what Active.ai’s offers. Its solution will allow financial services companies to offer compelling UX that will simplify engagement via messaging platforms customers’ access in their daily lives. Speaking on the investment, Ravishankar, CEO and Co-Founder said, “Conversation is the new UX and with banks opening up APIs, a new era of digital business is emerging. We are moving from ‘Mobile First’ to ‘AI first’ and Active.ai is the platform facilitating banks to achieve that.” “Active.ai is helping banks redefine their digital strategy for the future, bringing in automation and intelligent customer engagement to banking and payments. The company’s intelligent and built for banking technology uses advanced NLP and machine intelligence to enable customers have natural dialogues over messaging, voice or IOT devices,” added Shankar Narayanan, COO and Co-Founder. Ravishankar, Shankar Narayanan and Parikshit Paspulati (CTO and Co-Founder), bring a wealth of expertise in banking and fintech, having founded companies in mobile banking, payments and working in startup teams of leading banks in Asia. In a short period of time the company has gained significant traction with banks across Asia, added revenue and built a team of experts in banking technology and AI. “43% of worldwide mobile phone consumers with a bank account use mobile banking today. The new generation of banking customers are looking for easy, secure and seamless interaction with their financial services provider. IDG Ventures India believes that AI will completely transform the banking industry, and we are very excited to partner with Active.ai to build the world’s largest conversational banking platform,” said Sanat Rao, IDG Ventures. Bala Srinivasa, Kalaari Capital notes “messaging based mobile banking has become a strategic imperative for banks globally to enhance consumer experience and lower customer engagement costs. Active.ai is poised to tap into this large market opportunity with its chat and AI powered mobile customer engagement solution. Kalaari is excited to back this founding team given their past track record, banking domain knowledge, and deep technology expertise.“","excerpt":"Active.ai, a Singapore based fintech startup recently announced raising a fund of $3 million from leading venture capital funds IDG Ventures India and Kalaari Capital. Active.ai uses artificial intelligence to deliver virtual assistant banking services to its customers. It’s intelligent and omni-channel platform enables financial service companies such as banks, wealth managers and asset managers to […]","categories":["AI News"],"tags":["Startups"],"author_name":"Srishti Deoras","publish_date":"2016-11-17T05:51:46","publication_year":"2016","word_count":515,"keywords":["API","artificial intelligence","AI","ML","Git","automation","NLP","Ray","Startups","R","startup"],"extracted_tech_keywords":["AI","artificial intelligence","ML","NLP","Ray","R","Git","API","automation","startup"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/active-ai-raises-3-mn-kalaari-capital-idg-ventures-india-aims-expand-ai-platform-banking-payments\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":49233,"title":"Best Practices For Version Control In Data Science Projects","content":"For data-driven organisations today, collaboration in data science projects is key to staying ahead of the pack. However, version control in data science projects are not straightforward and need to be implemented with best practices for effortless collaboration. Jupyter Notebook Under Version Control Version control of data science projects on Jupyter Notebooks are tedious. That’s because changes in the code also alter the Jupyter Notebook structure which then displays the change in ‘.ipynb’ along with the code changes. The notebook has a ‘JSON’ format, thus with all the ‘JSON objects’ under ‘git diff’, it becomes difficult to find and understand the code changes in the notebook cell. Besides, if one runs the cell twice without changing the code, the notebook increments the cell number and git tracks those changes, which again spoils the user experience while collaborating as the cell numbers do not match with other contributors. Such variations are extraneous to developers, create conflict during accepting the pull requests from contributors. In a nutshell, the notebook is inadvisable for collaborative projects. A workaround for this is to tweak the way the notebook functions for simultaneously generating ‘.py’ as well as ‘.html’ files whenever the notebook is saved. This will allow users to make changes in the notebook and then save it for creating Python and Html file that include code changes in the notebook. Although it does not impact the way notebook conducts its functions and still remains the same, one can use the Python and Html files to track new addition and deletion in codes. However, this corners the challenges, but one still will have to re-run the whole Notebook before committing the changes for getting rid of cell numbers. To set up Jupyter Notebook, you can follow through the link and configure it for creating and updating Python and Html files. R Markdown Under Version Control R Markdown is the go-to file format for any data science projects because of a wide range of aspects that the ‘Rmd’ file offers. Unlike ‘.ipynb’ files that tracked cell numbers, ‘Rmd’ files do not trail the number of times a chunk has been executed. Moreover, git does not track the modification in the file due to alteration in codes. Therefore, one can commit changes without having to make any modification. Further, the ‘Rmd’ allow users to hide outputs of the chunks by adding options to each code chunk. Developers can add `echo=False`, and git will not monitor any output. Chunk options are effective not only for version control but also for extracting ‘Rmd’ files into pdf and Html files. Performing data science projects in R Studio is trouble-free as it includes a lot of features such as version control within it. On the other hand, customising Jupyter Notebook is strenuous in nature, thereby, requiring expertise for modifying workflows. What Data Scientists Should Do Python has continued to be the primary programming language for data scientists thus Jupyter Notebook is an apparent choice for them for any data analysis projects. Therefore, they cling on to notebooks even after all the problems that come with it. To streamline the workflow, they should ensure that all the contributors make those configurations for the unification of projects. Because of notebooks variation among contributors, there can be a hindrance in collaboration. Outlook Although Jupyter Notebook can be configured, it still requires data scientists to adopt certain practices for maintaining the workflow of the projects. However, this is cumbersome and decelerate the analysis process. This is contrary to the idea of collaboration that focuses on expediting the process. It usually takes time to learn the ropes for effectively managing the data science project with Jupyter Notebooks. Consequently, in the beginning, one may struggle to adapt, but eventually increases flexibility for higher productivity.","excerpt":"For data-driven organisations today, collaboration in data science projects is key to staying ahead of the pack. However, version control in data science projects are not straightforward and need to be implemented with best practices for effortless collaboration. Jupyter Notebook Under Version Control Version control of data science projects on Jupyter Notebooks are tedious. That’s […]","categories":["AI Trends"],"tags":["Data Scientists","Jupyter Notebook","version control"],"author_name":"Rohit Yadav","publish_date":"2019-11-01T16:00:39","publication_year":"2019","word_count":626,"keywords":["data science","Go","version control","Jupyter Notebook","TPU","AI","ML","Git","Python","GAN","Jupyter","R","Data Scientists"],"extracted_tech_keywords":["AI","ML","data science","Jupyter","TPU","Python","R","Go","Git","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/best-practices-for-version-control-in-data-science-projects\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":46704,"title":"5 Data Science Research Fellow Jobs In India","content":"For a company to thrive irrespective of the domain, research is one of the most important things to consider. When it comes to data science, the domain is so vast that organisations, as well as educational institutes across the world, are investing heavily in research and development. In this article, we list five latest data science research fellow jobs in India that you can apply for right away: 1. Junior Research Fellowship\/Senior Research Fellowship\/Research Associate @ IIT Kharagpur The Indian Institute of Technology, Kharagpur on 26 September 2019 announced that the institute is looking to onboard a best-suited candidate for the position of Junior Research Fellowship\/ Senior Research Fellowship\/ Research Associate. Also, the selected candidate will be ₹47,000 per month (depending upon qualification & experience) Project MIRIAD — Many Incarnations of Screening Radiology for High Throughput Disease Screening Via Multiple Instance Reinforcement Learning with Adversarial Deep Neural Networks(NNA) Prerequisites For Junior Research Fellowship: B. Tech\/BE in EE\/EC\/IN with project\/thesis in Image\/Signal Processing. A valid GATE score is desirable. Senior Research Fellowship: M. Tech\/ME\/MS in EE\/EC\/IN with a thesis on Image\/Signal Processing or Machine Learning. A valid GATE score is desirable. Research Associate: PhD in Medical Image Analysis or Machine Learning or Machine Vision. Also, the candidate must have experience in programming with Python, Pytorch\/TensorFlow, CUDA\/cuDNN, MKL\/mklDNN. Research Associate applicants are expected to have published in topical Transactions\/Journals\/Conferences like IEEE TMI, TIP, TSP, MedIA, CMIG, SPIE JBO, MICCAI, ISBI, etc. Note: Application Deadline is 14 October 2019, and the number of vacancies is 1. Apply Here. 2. Researcher – Machine Learning for Intelligent Systems @ Bosch, Bengaluru Bosch’s Research and Technology Center at Bangalore is looking to hire researchers who would be working with top-tier universities worldwide and corporate partners in machine learning for topics related to pervasive computing\/intelligent systems. The centre focuses on research in Internet technologies, computer vision and embedded real-time systems. Prerequisites PhD in any of the areas: Machine Learning for Intelligent Systems, Pervasive Computing, Context-Aware Systems, IoT Data Analytics — from reputed Indian institutes like IISc, IITs, IIITs, etc. or from reputed international institutes Should have a strong track record of publications in the areas such as Computer Science, Electronics and Communications Engineering, Electrical etc. and related branches with a background in Machine Learning Must have experience in programming with Python, R, Java Should be able to design experiments and create proof-of-concept deployments Apply Here. 3. AI Centre Postdoctoral Fellow @ Microsoft, Bengaluru Microsoft Research (MSR) India is a leading body when it comes to developing and deploying technologies for socio-economic development. MSR right now is accepting applications for its AI Centre Postdoctoral Fellows Program. As an AI Centre Postdoctoral Fellow, the candidate would have to spend 1 to 2 years at MSR India and get a chance to receive ongoing guidance and mentoring from an experienced researcher at Microsoft Research. The candidate would collaborate effectively with other researchers, product development teams and external collaborators. Prerequisites Candidates who have completed their PhDs can apply for this position. Apply Here. 4. Junior Research Fellow\/Senior Research Fellow @ Indian Institute of Science Education and Research (IISER), Bhopal IISER Bhopal recently announced that the institution is looking to hire candidates for the position of Junior Research Fellow\/Senior Research Fellow. Talking about the salary, Junior Research Fellow will get ₹31,000 per month and Senior Research Fellow will get ₹35,000. Research Area Computational studies of membranes and membrane proteins. The project is highly interdisciplinary in nature and will integrate the knowledge of Molecular Dynamics Simulations and Machine Learning techniques. Prerequisites For Junior Research Fellow: MSc\/MTech\/BTech degree from a recognised institute with a minimum of 60% marks or CGPA of 7.5 and above in one of the following disciplines: Physical Sciences, Chemical Sciences, Engineering Sciences. The candidate must have Mathematics both at the 10+2 and Bachelors level A valid score in GATE or CSIR-UGC NET (JRF\/LS) or any other equivalent national level exam is desirable For Senior Research Fellow: Minimum two years of research experience as JRF Note: The Application deadline is 2 October 2019 and the number of vacancies is 1 Apply Here. 5. Senior Research Scientist @ Rakuten, Bengaluru Japanese electronic commerce and Internet company, Rakuten is looking to hire a Senior Research Scientist for Rakuten India, Bengaluru. As a research scientist, the candidate would not only be responsible for the systematic study of the latest research and industry developments in Artificial Intelligence but would also work on solving complex business problems of global organisations by applying AI, ML and NLP techniques, especially Deep Learning and\/or Reinforcement Learning. Prerequisites The candidate should be a PhD\/MS. in statistics, mathematics, computer science, Machine Learning or related areas Should have hands-on experience in ML, Deep Learning and AI Experience with TensorFlow, Keras, Theano and\/or PyTorch Apply Here.","excerpt":"For a company to thrive irrespective of the domain, research is one of the most important things to consider. When it comes to data science, the domain is so vast that organisations, as well as educational institutes across the world, are investing heavily in research and development. In this article, we list five latest data […]","categories":["AI Hirings"],"tags":["data science job","Data Science Jobs","MSc data science","msc data science and analytics","MSc. in data science","project topics for computer science","science jobs"],"author_name":"Harshajit Sarmah","publish_date":"2019-10-01T17:00:00","publication_year":"2019","word_count":793,"keywords":["MSc. in data science","project topics for computer science","data science","artificial intelligence","machine learning","science jobs","AI","neural network","ML","Data Science Jobs","computer vision","NLP","msc data science and analytics","deep learning","analytics","MSc data science","data science job"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","neural network","NLP","computer vision","data science","analytics"],"url":"https:\/\/analyticsindiamag.com\/ai-hiring\/5-data-science-research-fellow-jobs-in-india\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10069164,"title":"Tech firms would rather leave China than deal with its grinding data security laws","content":"Nike Run Club App will shut its Chinese operations from July 8, the American firm announced last week. Nike has asked users to move to a Nike mini app on WeChat. Nike Run Club App has around 8 million users. Many foreign brands have pulled out after China enforced strict data protection laws. U.S. companies leaving China, Source: Statista Exodus of US companies A week prior, Amazon announced that it was halting sales of its Kindle e-books in China starting from July 2023. Towards the end of last month, Airbnb said it too was pulling out from China owing to fierce competition from homegrown rivals. Microsoft-owned LinkedIn also shut down its Chinese operations last year. Most foreign tech giants have either been blocked in the country (like Facebook and Twitter in 2009) or have quit the country (like Google in 2010). Last November, Yahoo left China citing “increasingly challenging” business environment. By mid-November, Epic Games (the publisher of hugely popular Fortnite) pulled out of China in the wake of the Xi Jinping-led government’s chokehold on video games. The Chinese government has capped the gaming hours to three for children below the age of 18. Last August, China announced a set of expansive data protection policies emulating EU’s General Data Protection Regulation (GDPR). First came the Personal Information Protection Law (PIPL), which restricted the transfer of personal data of Chinese citizens across the border. The law kicked in in November which gave companies very little time to comply. Like in GDPR, companies were required to submit themselves to a security clearance assessment conducted by Chinese regulators before transferring data and would have to appoint local representatives to handle privacy issues. Companies that violated the law were potentially subject to fines of up to 5 percent of their annual revenue or have their operation licence revoked. In addition, individual penalties could be levied against senior executives of the companies. If companies are found to store unauthorised data overseas, they are liable to pay even heftier fines that can run up to 500,000 yuan. The country’s Data Security Law came into effect in September last year. The Law is enacted for the purpose of regulating data processing, ensuring data security, promoting development and utilisation of data, protecting the lawful rights and interests of individuals and organisations, and safeguarding the sovereignty, security, and development interests of the state. The law applies to data processing activities and security supervision and regulation of such activities within the territory of the People’s Republic of China. Where data processing outside the territory of People’s Republic of China harms the national security, public interests, or the lawful rights and interests of individuals or organisations of the People’s Republic of China, legal liability will be investigated in accordance with the law. The law also had an additional set of compliances for organisations that processed valuable data. These entities were required to open themselves to periodic risk assessments besides following PIPL’s cross border regulations as well. China has a much broader definition for “personal data.” According to PIPL even if the data cannot be matched to a specific user it can be classified as personal data as long as it can be “related to identified or identifiable natural persons.” As its domestic tech giants like Tencent and Alibaba grew in scale and power, the Chinese government became wary about the increasing complaints regarding user privacy violations. Last year, before the 618 shopping festival, the Ministry of Industry and Information Technology summoned several e-commerce platforms and telecom companies for their intrusive marketing messages. The regulatory body also condemned apps including WeChat for illegal transfer of personal information. In the beginning of last year, the country’s popular ride-hailing app Didi Chuxing came under fire for data privacy issues. Despite the fraught relationship with their own government, Didi filed for the US IPO in June. Right after the company raised USD 4.4 billion, the Cyberspace Administration of China initiated an investigation against it. In response, Didi’s executives said China’s cybersecurity law was vaguely worded and did not explicitly state how it would be implemented.","excerpt":"Amazon announced it was halting sales of its Kindle e-books in China starting from July 2023.","categories":["AI Features"],"tags":[],"author_name":"Poulomi Chatterjee","publish_date":"2022-06-16T11:00:00","publication_year":"2022","word_count":680,"keywords":["Go","AWS","AI","IPO","cloud_platforms:AWS","programming_languages:R","programming_languages:Go","ViT","GAN","R"],"extracted_tech_keywords":["AI","AWS","R","Go","GAN","ViT","IPO","cloud_platforms:AWS","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/tech-firms-would-rather-leave-china-than-deal-with-its-grinding-data-security-laws\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10089056,"title":"Microsoft Unveils Visual ChatGPT, A ChatGPT for Images","content":"Microsoft recently unveiled a new model called ‘Visual ChatGPT‘, which incorporates different types of Visual Foundation Models (VFMs) including Transformers, ControlNet, and Stable Diffusion with ChatGPT. The system enables interaction with ChatGPT beyond language. This connection enables sending messages through chat and receiving images during the chat, while also injecting a series of visual model prompts to edit the images as well. Click here to check out the GitHub repository. The paper titled, ‘Visual ChatGPT: Talking, Drawing and Editing with Visual Foundation Models‘, highlights how all the visual transformer models are only experts of specific tasks with fixed inputs and outputs, and the same is the case with ChatGPT as it is only trained on text. Combining all of them makes image generation and manipulation limitless. ​​ In order to bridge the gap between ChatGPT and VFMs, the paper proposes the use of a Prompt Manager that includes the following features: Explicitly inform ChatGPT about the capabilities of each VFM and specify the necessary input-output formats. Convert various types of visual information—such as png images, depth images, and mask matrices—into language format to aid ChatGPT’s understanding. Manage the histories, priorities, and conflicts of different VFMs. By utilising the Prompt Manager, ChatGPT can effectively leverage VFMs and receive their feedback in an iterative manner until the users’ requirements are met or a concluding condition is reached.This enables users to interact with ChatGPT using images as well, more than only text. Moreover, users can also ask complex image questions or visual editing by the collaboration of different AI models in multi-steps. Users can also ask for corrections and feedback on results.","excerpt":"This connection enables sending messages through chat and receiving images during the chat, while also injecting a series of visual model prompts, to edit the images as well.","categories":["AI News"],"tags":["OpenAI","Stable Diffusion","Transformers"],"author_name":"Mohit Pandey","publish_date":"2023-03-10T13:07:59","publication_year":"2023","word_count":269,"keywords":["ChatGPT","TPU","OpenAI","AI","Stable Diffusion","Transformers","Git","RAG","GPT","foundation models","GitHub","R"],"extracted_tech_keywords":["AI","foundation models","ChatGPT","Transformers","RAG","TPU","R","Git","GitHub","GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/microsoft-unveils-visual-chatgpt-a-chatgpt-for-images\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10009792,"title":"Hands-On Tutorial On Machine Learning Pipelines With Scikit-Learn","content":"With increasing demand in machine learning and data science in businesses, for upgraded data strategizing there’s a need for a better workflow to ensure robustness in data modelling. Machine learning has certain steps to be followed namely – data collection, data preprocessing(cleaning and feature engineering), model training, validation and prediction on the test data(which is previously unseen by model). Here testing data needs to go through the same preprocessing as training data. For this iterative process, pipelines are used which can automate the entire process for both training and testing data. It ensures reusability of the model by reducing the redundant part, thereby speeding up the process. This could prove to be very effective during the production workflow. (Source: YouTube – Pydata ) In this article, I’ll be discussing how to implement a machine learning pipeline using scikit-learn. Advantages of using Pipeline: Automating the workflow being iterative.Easier to fix bugs Production ReadyClean code writing standardsHelpful in iterative hyperparameter tuning and cross-validation evaluation Challenges in using Pipeline: Proper data cleaningData Exploration and AnalysisEfficient feature engineering Scikit-Learn Pipeline The sklearn.pipeline module implements utilities to build a composite estimator, as a chain of transforms and estimators. I’ve used the Iris dataset which is readily available in scikit-learn’s datasets library. The 6 columns in this dataset are: Id, SepalLength(in cm), SepalWidth(in cm), PetalLength(in cm), PetalWidth(in cm), Species(Target). 50samples containing 3 classes-Iris setosa, Iris Virginica, Iris versicolor. After loading the data, split it into training and testing then build pipeline object wherein standardization is done using StandardScalar() and dimensionality reduction using PCA(principal component analysis) both of these with be fit and transformed(these are transformers), lastly the model to use is declared here it is LogisticRegression, this is the estimator. The pipeline is fitted and the model performance score is determined. from sklearn.datasets import load_iris from sklearn.model_selection import train_test_split from sklearn.preprocessing import StandardScaler from sklearn.decomposition import PCA from sklearn.pipeline import Pipeline from sklearn.linear_model import LogisticRegression iris_df=load_iris() X_train,X_test,y_train,y_test=train_test_split(iris_df.data,iris_df.target,test_size=0.3,random_state=0) pipeline_lr=Pipeline([('scalar1',StandardScaler()), ('pca1',PCA(n_components=2)),                     ('lr_classifier',LogisticRegression(random_state=0))]) model = pipeline_lr.fit(X_train, y_train) model.score(X_test,y_test) OUTPUT - 0.8666666666666667 With the pipeline, we preprocess the training data and fit the model in a single line of code. In contrast, without a pipeline, we have to do normalization, dimensionality reduction, and model training in separate steps. This becomes especially messy if we have to deal with both numerical and categorical variables. Use the following two lines of code inside the Pipeline object for filling missing values and change categorical values to numeric. (Since iris dataset doesn’t contain these we are not using) ('imputer', SimpleImputer(strategy='most_frequent')) #filling missing values (‘onehot', OneHotEncoder(handle_unknown='ignore'))    #convert categorical Make sure to import OneHotEncoder and SimpleImputer modules from sklearn! Stacking Multiple Pipelines to Find the Model with the Best Accuracy We build different pipelines for each algorithm and the fit to see which performs better. from sklearn.neighbors import KNeighborsClassifier from sklearn.tree import DecisionTreeClassifier from sklearn import svm pipeline_lr=Pipeline([('scalar1',StandardScaler()), ('pca1',PCA(n_components=2)), ('lr_classifier',LogisticRegression())]) pipeline_dt=Pipeline([('scalar2',StandardScaler()), ('pca2',PCA(n_components=2)), ('dt_classifier',DecisionTreeClassifier())]) pipeline_svm = Pipeline([('scalar3', StandardScaler()), ('pca3', PCA(n_components=2)), ('clf', svm.SVC())]) pipeline_knn=Pipeline([('scalar4',StandardScaler()), ('pca4',PCA(n_components=2)), ('knn_classifier',KNeighborsClassifier())]) pipelines = [pipeline_lr, pipeline_dt, pipeline_randomforest, pipeline_knn] pipe_dict = {0: 'Logistic Regression', 1: 'Decision Tree', 2: 'Support Vector Machine',3:'K Nearest Neighbor'} for pipe in pipelines: pipe.fit(X_train, y_train) for i,model in enumerate(pipelines): print(\"{} Test Accuracy:{}\".format(pipe_dict[i],model.score(X_test,y_test))) OUTPUT: Logistic Regression Test Accuracy: 0.8666666666666667 Decision Tree Test Accuracy: 0.9111111111111111 Support Vector Machine Test Accuracy: 0.9333333333333333 K Nearest Neighbor Test Accuracy: 0.9111111111111111 From the results, it’s clear that Support Vector Machines(SVM) perform better than other models. Hyperparameter Tuning in Pipeline With pipelines, you can easily perform a grid-search over a set of parameters for each step of this meta-estimator to find the best performing parameters. To do this you first need to create a parameter grid for your chosen model. One important thing to note is that you need to append the name that you have given the classifier part of your pipeline to each parameter name. In my code above I have called this ‘randomforestclassifier’ so I have added randomforestclassifier__ to each parameter. Next, I created a grid search object which includes the original pipeline. When I then call fit, the transformations are applied to the data, before a cross-validated grid-search is performed over the parameter grid. from sklearn.model_selection import GridSearchCV from sklearn.pipeline import make_pipeline from sklearn.ensemble import RandomForestClassifier pipe = make_pipeline((RandomForestClassifier())) grid_param = [ {\"randomforestclassifier\": [RandomForestClassifier()], \"randomforestclassifier__n_estimators\":[10,100,1000],                    \"randomforestclassifier__max_depth\":[5,8,15,25,30,None],                 \"randomforestclassifier__min_samples_leaf\":[1,2,5,10,15,100], \"randomforestclassifier__max_leaf_nodes\": [2, 5,10]}] gridsearch = GridSearchCV(pipe, grid_param, cv=5, verbose=0,n_jobs=-1) best_model = gridsearch.fit(X_train,y_train) best_model.score(X_test,y_test) OUTPUT - 0.9777777777777777 Conclusion This is a basic pipeline implementation. In real-life data science, scenario data would need to be prepared first then applied pipeline for rest processes. Building quick and efficient machine learning models is what pipelines are for. Pipelines are high in demand as it helps in coding better and extensible in implementing big data projects. Automating the applied machine learning workflow and saving time invested in redundant preprocessing work.The complete code of the above implementation is available at the AIM’s GitHub repository. Please visit this link to find the notebook with codes.","excerpt":"Hands-On Tutorial On Machine Learning Pipelines With Scikit-Learn .In this article, I’ll be discussing how to implement a machine learning pipeline using scikit-learn.","categories":["Deep Tech"],"tags":["data preprocessing","pipelines in machine learning","scikit learn","simple ai tutorial","when to use support vector machine"],"author_name":"Jayita Bhattacharyya","publish_date":"2020-10-15T18:00:15","publication_year":"2020","word_count":811,"keywords":["simple ai tutorial","data science","scikit-learn","machine learning","TPU","Go","data preprocessing","AI","Transformers","when to use support vector machine","Scala","Aim","scikit learn","pipelines in machine learning","R"],"extracted_tech_keywords":["AI","machine learning","data science","Aim","scikit-learn","Transformers","TPU","R","Go","Scala"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/hands-on-tutorial-on-machine-learning-pipelines-with-scikit-learn\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10170079,"title":"How Modern Data Engineers Are Becoming Race Engineers of the AI Era","content":"In the fast-changing world of data engineering and the AI era, the role of professionals has evolved, moving from mere managers of data pipelines to key strategists making rapid decisions in real time. At DES 25, Varun Saraogi, principal data architect at MathCo, vividly captured this transformation, drawing an insightful analogy from the high-stakes world of Formula 1 racing. “Modern data engineers are like race engineers,” he said. Saraogi, an F1 enthusiast, noted that the comparison was casual but more structured. In a Formula 1 race, each team gathers massive volumes of data every second. But, as he noted, there is no differentiation in how much data you can capture across teams. What differentiates them is how their teams use the data. Like a race engineer relaying insight to the driver, today’s data teams must act with precision, speed, and deep understanding. He stated that the key issue is how the data engineering team, the strategist, and the race engineer comprehend and leverage that data to deliver contextual information. Two Pit Stops: Context and Product Mindset Saraogi’s central argument was that context is now fundamental to decision-making. From metadata and domain knowledge to access controls and interaction history, Saraogi outlined eight distinct layers that must be embedded, not bolted on, into pipelines. “Metadata cannot be an afterthought anymore,” he insisted. It has to be embedded into your pipeline.” Beyond context, he called for a shift from pipeline-building to product thinking. Drawing from experience, Saraogi admitted that back in the day, he only focused on providing data to the team and did not understand the use cases. That mindset no longer holds. Whether building a Customer 360 or preparing for real-time AI use cases, data engineers today must anticipate business intent. “It is not just about building platforms for AI, it is about building platforms with AI as well,” he said. Saraogi also pointed to the need for real-time signals, such as holidays or sales trends, to be embedded directly into data flows, keeping context in mind. Traditional systems can’t keep up with open-ended queries or expectations for on-the-fly personalisation. “If you have to rewrite the playbook…the future is here,” he said, urging engineers to rethink how data platforms work and are built. The Finish Line Isn’t Just About Speed Saraogi reminded the audience that the fastest car doesn’t always win. “McLaren in the second half had the fastest car, but they did not win. It’s also about strategy, data, and how one uses it,” he said. Likewise, a platform with high throughput and low latency won’t cut it unless the data has meaning. He stressed that the data has to be analysed to understand the users, the region, and more, not just in terms of sales numbers. Context must be present from the beginning to deliver relevant and trustworthy AI outputs. Saraogi laid out a structured approach: identify context sources (structured, unstructured, metadata, and external), assemble context models (from relational to vector databases), and monitor their use to ensure feedback and trust. So, the modern data engineer doesn’t just move data. They monitor, interpret, and strategise, like a race engineer perched on the pit wall. The race hasn’t slowed down, but the winners are not only building faster pipelines but also asking more thoughtful questions.","excerpt":"“It is not just about building platforms for AI, it is about building platforms with AI as well.”","categories":["AI Features"],"tags":["Data Engineering"],"author_name":"Ankush Das","publish_date":"2025-05-16T13:19:36","publication_year":"2025","word_count":545,"keywords":["API","TPU","programming_languages:R","AI","data pipeline","RAG","vector databases","data engineering","Data Engineering","Rust","R"],"extracted_tech_keywords":["AI","RAG","vector databases","TPU","R","Rust","API","data engineering","data pipeline","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-modern-data-engineers-are-becoming-race-engineers-of-the-ai-era\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10047122,"title":"5 Ransomware Attacks of 2021 That Blew The Internet","content":"Ransomware refers to malicious softwares that, when deployed, can prevent one from using their own computer. In order to get back access to the system, one has to pay a sum of money to cyber-criminals or hackers. According to Chennai-based Cyber Security Works, Ransomware is increasingly targeting the critical industrial sectors — oil and gas, finance, healthcare, food and beverages, and transportation. Ransom attacks have increased in volume (by 37 per cent) during the ongoing pandemic, and Common Vulnerabilities and Exposure (CVE) saw a jump of over 356 per cent this year compared to 2019. SonicWall recorded an all-time high of 78.4 million ransomware attacks globally in June 2021. Airline company Air India, food major Haldiram’s, and Pune’s smart city project Pimpri-Chinchwad Municipal Corporation faced major cyber attacks in India itself. Today, we take a peek at the biggest ransomware attacks of 2021 so far. To know about last year’s attacks, check here. Colonial Pipeline Company American oil pipeline system Colonial Pipeline Company suffered a major ransomware attack in May this year. The cyberattack affected its computerised equipment managing the pipeline originating from Houston, Texas, disrupting the fuel supply to most of the US East Coast for days. Despite affecting just its IT systems, Colonial Pipeline Company shut down its entire pipeline operations to prevent further harm. With the FBI’s help, the company paid $4.4 million in bitcoin, as demanded by the hackers. According to the FBI, the attack was carried out by DarkSide. A month after payment, the Department of Justice reported that the FBI was able to seize a portion of the payment using a private key. Acer Taiwanese computer giant Acer was hit by a REvil ransomware attack in March this year. The hackers demanded a whopping $50 million. They shared images of stolen files as proof of breaching Acer’s security and the consequent data leak. These included images of financial spreadsheets, bank communications, and bank balances. According to media reports, the group got access to Acer’s network through a Microsoft Exchange vulnerability that had earlier led to the hacking of 30,000 US governmental and commercial organisations’ emails. The ransomware gang reportedly made more than $100 million in one year from large business extortions. The same hackers were responsible for the 2020 ransomware attack on Travelex. While the computer manufacturer never confirmed if they actually paid the ransom, it said that companies like theirs are constantly under attack and had reported abnormal situations observed to the required law enforcement and data protection authorities. CNA Financial Chicago-based CNA Financial Corp., one of the largest insurance companies in the USA, had noticed a breach in March this year. The ransomware attack is said to have led to the compromise of data of around 75,000 individuals. This data might have included names, health benefits information, and Social Security numbers of the company’s present and former employees, contract workers, and their dependents. According to media reports, later in May, CNA Financial agreed to have paid $40 million to get back access to its network. Reportedly, the hackers used a malware called Phoenic Locker, a variant of Hades which Russian cybercrime syndicate Evil Corp creates. Brenntag Around the same time as the Colonial Pipeline Company cyberattack, hackers group DarkSide targeted Germany-headquartered chemical distribution company Brenntag. DarkSide reportedly demanded $7.5 million, or 133.65 bitcoin, for gaining access to 150 GB worth of data. Additionally, DarkSide shared a data leak page consisting of a description of the data stolen and screenshots of a couple of files to prove its claims. The ransom was negotiated, and ultimately, Brenntag ended up paying $4.4 million. Kia Motors A subsidiary of Hyundai, Kia Motors, suffered ransom in February this year. Attackers DopplePaymer gang reportedly asked for $20 million for a decrypter and not leak the stolen data. As claimed by Kia Motors, the subsequent ‘IT outage’ affected the mobile UVO Link apps, payment systems, owner’s portal, phone services, and internal sites used by Kia Motors America. While these were global attacks, India isn’t far from making headlines for cybersecurity breaches, either. If one were to go by media reports, India was most hit by ransomware attacks this year, so far. A report by Check Point research suggests that with ransomware attacks shot up by 102 per cent globally in 2021 from last year, India was the worst hit with 213 weekly ransomware attacks per organisation. Last year, Microsoft appointed a Threat Protection Intelligence Team to deal with the attacks. To know all about how the Colonial Pipeline Company ransomware happened and how the FBI was able to plot a crypto ransom recovery, click here.","excerpt":"In June 2021, SonicWall recorded an all-time high of 78.4 million ransomware attacks across the globally.","categories":["AI Trends"],"tags":["hacking","Ransomware","Ransomware attacks"],"author_name":"Debolina Biswas","publish_date":"2021-08-26T16:00:00","publication_year":"2021","word_count":766,"keywords":["Ransomware attacks","Go","hacking","programming_languages:R","AI","Ransomware","programming_languages:Go","RAG","Aim","GAN","R"],"extracted_tech_keywords":["AI","Aim","RAG","R","Go","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/5-ransomware-attacks-of-2021-that-blew-the-internet\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10112879,"title":"Oracle Cloud Powers Yubi&#8217;s Co-lending Expansion","content":"Yubi Group, a fintech company that specialises in providing a platform for debt financing, is using Oracle Cloud Infrastructure (OCI) to run its co-lending platform across India. With OCI, the fintech is able to bridge the credit gap in India and facilitate scalability in the co-lending sector. Yubi processes over 1 million transactions daily on average. This has enabled over 750 lenders to distribute joint loans totaling US$1.2 billion to more than 1 million customers in India over the past three years. Co-lending in India is projected to exceed US$12 billion this year. Yubi’s platform requires better flexibility, scalability, and security to accommodate the growing need. Gaurav Kumar, founder and CEO of Yubi said, “We’re gearing up to increase our transaction volumes by over 10-fold in the upcoming months, and transitioning to OCI can help us with our goals,” He further said they opted for OCI due to the dual-region cloud strategy, anticipating over 25 percent cost savings from this migration. Yubi will deploy OCI services such as Compute Virtual Machines (VMs), Object Storage, and OCI Database with PostgreSQL to gain flexible compute capacity for its projects. The company said it will also benefit from high-performance computing and low-cost cloud storage options to improve the efficiency and productivity of its IT team. The migration to OCI will help Yubi to combine open-source technology with OCI to significantly improve performance and lower costs. In the recent conference at Oracle CloudWorld Tour, the company shared that it has seen a 50% year-over-year increase in cloud consumption and notable expansions in sectors like BFSI, telecom, healthcare, and education. Oracle’s partnerships, including those with Microsoft and VMware, are improving multi-cloud environments. These new launches in Oracle’s portfolio aim to support India’s digital economy goals.","excerpt":"Yubi Group utilises Oracle Cloud Infrastructure to enhance its co-lending platform in India to enable their significant growth in the fintech sector.","categories":["AI News"],"tags":["Oracle","Oracle SQL"],"author_name":"K L Krithika","publish_date":"2024-02-14T16:17:22","publication_year":"2024","word_count":290,"keywords":["PostgreSQL","Oracle SQL","Go","AI","Scala","Oracle","RAG","Git","Aim","ViT","SQL","R"],"extracted_tech_keywords":["AI","Aim","RAG","PostgreSQL","R","SQL","Go","Scala","Git","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/oracle-cloud-powers-yubis-co-lending-expansion\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":40010,"title":"Top 5 AI-Based Prototyping Tools For UI And UX","content":"The UX design space over the years has witnessed tremendous transformations. It has reached a level where designers no longer need to hustle and work and see the final product — with the advent of technologies like AI, designers can now have a peek of the final result without even completing the entire project. Thanks to prototyping tools, designers today have overcome their dilemma whether the output will be something that will be appreciated. A prototyping tool is what basically allows a designer to experience the end result\/product without having to complete the project. This not only helps the designers make product creation faster but also help them in making it effective, delivering great user experience. Whether it’s about creating layouts or making templates, it takes hours or some time days to complete the task. But using technologies like AI, the whole process can be automated. So, here we are going to see some of the best AI-enabled prototyping tools for UI\/UX. 1. Uizard Uizard is a popular platform that teaches machines to understand graphical user interfaces in the same way humans do. Using AI, one can generate native mobile applications directly from sketches. That is not all, apart from automating the design process, it also focuses on generating code from the sketch. Uizard without a doubt one of the best options if you want to design and test the flow with users. 2. Airbnb’s AI American hospitality services company, Airbnb has developed an AI system that helps designers to move ideas directly from the drawing board to reality. This entire project into the scenario when the realized that the time taken to test idea should be very less. The company designed an AI system that recognizes standard hand-drawn design and elements and automatically renders them into the source code. So, basically, this AI-powered solution by Airbnb makes the product development process simpler as it focuses more on testing functional prototypes. 3. Mockplus When we talk about a tool that helps designers better implement creativity, communication, material collection and reduce manual labour in this AI-driven design era, Mockplus is one of the best choices. It is a powerful tool that let designers automate design and export design right from platforms such as Sketch, Photoshop or Adobe XD. It also allows the designers to see specification prepared automatically and build interactive prototypes. 4. InVision InVision is one of the widely used prototyping tools with over 2 million users all across the world.  This tool has gained so much traction that many says it is one of the toughest competitions Adobe Wireframe has faced till date.  This tool basically lets a designer create clickable versions of designs in order to test and present it to the customer or the clients. That is not all, it allows a designer to share the screen with the fellow designers, create animations and transitions and it also interacts with most common design tools such as Sketch and Photoshop. 5. Balsamiq Simplicity is one of the topmost priority of a designer when they are trying to figure out how the end product would look. And this where Balsamiq comes in. While most of the prototyping tools come with a colourful and fancy interface, Balsamiq wireframe eliminates that aspect in order to reproduce the experience of sketching on a notepad or whiteboard but using a computer. The best thing about this tool is that it allows the designer to focus more on the content at first, rather than the colours that could be done anyway at the last. Outlook The design domain might be becoming an AI-driven, but it won’t be an easy task for a machine to understand human emotions which is very much necessary in designing. By knowing the clients’ needs, a designer has to make something that suits the business or the purpose of the project. So, it is unlikely that AI would be taking all the jobs of designers, rather, it will be a designer’s new best friend — because when the machine automate the repetitive tasks, humans become creative and that is what needed to be a top designer.","excerpt":"The UX design space over the years has witnessed tremendous transformations. It has reached a level where designers no longer need to hustle and work and see the final product — with the advent of technologies like AI, designers can now have a peek of the final result without even completing the entire project. Thanks […]","categories":["AI Trends"],"tags":["Web Development"],"author_name":"Harshajit Sarmah","publish_date":"2019-06-02T10:29:40","publication_year":"2019","word_count":686,"keywords":["Go","TPU","programming_languages:R","AI","programming_languages:Go","Web Development","ViT","R"],"extracted_tech_keywords":["AI","TPU","R","Go","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-5-ai-based-prototyping-tools-for-ui-and-ux\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":38751,"title":"Hive v\/s Pig: Comparing The Two Principal Components Of Hadoop Ecosystem","content":"Apache Pig and Apache Hive are the two key components of the Hadoop ecosystem. Both the tools are open-sourced and run on the top of MapReduce. In this article, we list down the comparisons between the two components. 1| Definition Apache Pig is a platform for analyzing large data sets which consists of a high-level language for expressing data analysis programs, coupled with infrastructure for evaluating these programs. Pig’s infrastructure layer consists of a compiler that produces sequences of Map-Reduce programs, for which large-scale parallel implementations already exist (e.g., the Hadoop subproject). It is released under the Apache 2.0 license. Apache Hive is an open source project run by volunteers at the Apache Software Foundation. Hive comes with built-in connectors for comma and tab-separated values (CSV\/TSV) text files, Apache Parquet, Apache ORC, and other formats. Users can extend Hive with connectors for other formats. It is designed to maximize scalability (scale out with more machines added dynamically to the Hadoop cluster), performance, extensibility, fault-tolerance, and loose-coupling with its input formats. 2| Language They Are Using Pig’s language layer currently consists of a textual language called Pig Latin whereas Apache Hive uses HiveQL which is a declarative language. 3| Properties The properties of Apache Pig are mentioned below: Ease of programming. It is trivial to achieve parallel execution of simple, “embarrassingly parallel” data analysis tasks. Complex tasks comprised of multiple interrelated data transformations are explicitly encoded as data flow sequences, making them easy to write, understand, and maintain. Optimization opportunities. The way in which tasks are encoded permits the system to optimize their execution automatically, allowing the user to focus on semantics rather than efficiency. Extensibility. Users can create their own functions to do special-purpose processing. Extract Transform Load: This platform extracts the numerous dataset, performs operations on them and then dumps the data in the required format in Hadoop Distributed File System (HDFS). On the other hand, Hive provides the following properties: Tools to enable easy access to data via SQL, thus enabling data warehousing tasks such as extract\/transform\/load (ETL), reporting, and data analysis. A mechanism to impose structure on a variety of data formats Access to files stored either directly in Apache HDFS or in other data storage systems such as Apache HBase Query execution via Apache Tez, Apache Spark, or MapReduce Procedural language with HPL-SQL Sub-second query retrieval via Hive LLAP, Apache YARN, and Apache Slider. Hive provides standard SQL functionality, including many of the later SQL:2003, SQL:2011, and SQL:2016 features for analytics. 4| When To Use Apache Pig is an exceptional Extract-Transform-Load tool for big data and can be used to handle numerous amount of unstructured data. This tool is faster than the other as it uses the multi-query approach. This tool provides a number of built-in operators to support data operations like joins, filters, sorting, etc. Pig is mainly used by researchers and programmers On the other hand, Hive can be used to query large datasets as well as analyse historical data. This tool has smart built-in features on accessing raw data and can be used for creating accurate reports. Hive is widely adopted in case of structured data and is mainly used by data analysts. 5| Latest Release The stable release of Apache Hadoop is 3.1.1 and it works with Hadoop 3.x.y. while the stable release of Apache Pig is 0.17.0 and this release works with Hadoop 2.X (above 2.7.x). Bottom Line As mentioned earlier, both of them are the two main components of the Hadoop ecosystem and both works for the same purpose. At the end of the day, it all depends upon the choices of the users in which they are more comfortable with. Both the tools have a moderately large community and will be developed with great advancements in the coming years.","excerpt":"Apache Pig and Apache Hive are the two key components of the Hadoop ecosystem. Both the tools are open-sourced and run on the top of MapReduce. In this article, we list down the comparisons between the two components. 1| Definition Apache Pig is a platform for analyzing large data sets which consists of a high-level […]","categories":["Deep Tech"],"tags":["etl hadoop","hadoop and etl","hadoop etl"],"author_name":"Ambika Choudhury","publish_date":"2019-05-07T11:50:51","publication_year":"2019","word_count":630,"keywords":["big data","programming_languages:R","AI","ETL","R","Apache Spark","Scala","RAG","analytics","SQL","hadoop etl","hadoop and etl","etl hadoop"],"extracted_tech_keywords":["AI","analytics","RAG","Apache Spark","R","SQL","Scala","big data","ETL","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/hive-v-s-pig-comparing-the-two-principal-components-of-hadoop-ecosystem\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":22820,"title":"Understanding Reptile: A Scalable Meta-learning Algorithm By OpenAI","content":"With Machine Learning (ML) advancing its frontiers day by day, developing base algorithms to fulfill the needs of self-learning in machines is a challenging task. In meta-learning, as the algorithms encounter a horde of data, they become more prone to providing useful output rather than giving out incomprehensible output. One such algorithm developed by OpenAI, a non-profit artificial intelligence (AI) research company, is Reptile— a meta-learning algorithm designed to perform over a wide array of tasks. Why Meta-learning is necessary? In ML, the learner algorithm assumes a set of statements that determines output based on input. If the input digresses from the usual and finds something new or strange, then the output  returned is vague. This is called learning or inductive bias since the algorithm is not tweaked into incorporating many data types or range. For example, if a neural network is built over a specific data range, the algorithm produces meaningful output that falls within the data range or type. It may not function if a different type of input is fed. This is where meta-learning comes into play. The level of learning is dynamically improved with the increase in number and quality of input data with rigorous variations. For instance, one may increase factors such as training-set size, sample distribution among other things to achieve an efficient learning algorithm. Ultimately, the aim of meta-learning is to decide how components of algorithms interact with each other and then act on improving it. Also, meta-learning helps new algorithms go beyond the scope of existing algorithms. Meta-learning is nothing but learning to learn. It is the process to choose the best strategy among the assortment of tasks present in an algorithm. The performance of the algorithm will improve over time with meta-learning. How Reptile works? Reptile is slightly based on another algorithm called Model-Agnostic Meta-learning(MAML) developed by researchers at University of California, Berkeley in collaboration with OpenAI. MAML uses step-by-step approach of training parameters for a gradient descent associated model. This way every step will serve as a training data for any new task encountered in the model. Reptile differs in this aspect. It utilises the stochastic gradient descent to initialise the parameters of the model instead of performing several computations which are resource-consuming, which is evident in MAML. This significantly reduces the dependency of higher computational, hardware requirements if implemented in a ML project. SGD is an improved form of the standard gradient descent. The difference among the former and the latter lies in the update made in the parameters. The equations are given below: where, = parameters for the project = learning rate = gradient descent with respect to parameters E = Error = Output presented for the parameters The stochastic variation is presented like this: Reptile follows this variation for its working. Although, the code for the algorithm may look a bit different, nevertheless the logic remains the same. The initialisation is performed either using Omniglot (for one-shot learning) and Mini-Image.Net (for few-shot learning).The algorithm is also tested using the Taylor Series. The reason being the product between the gradients related to similar tasks shows improvement in generalisation. Reptile is compared for its usefulness with another algorithm called the First-Order Model-Agnostic Meta-Learning (FOMAML). It is mathematically similar to Reptile except that it uses the standard gradient descent and the update has a gradient which stays constant throughout iterations.The key contrasting feature between these two algorithms is the reduction in variance shown below. Sample Variance Depiction using Omniglot (Courtesy – OpenAI) The above image shows the different gradients — represented by the letter ‘g’ obtained using Omniglot. They are combined to form meta-gradients to facilitate quicker learning by the algorithm. For demo purposes, the platform used by OpenAI to showcase Reptile is Tensorflow, owing to its flexibility in terms of ML and mathematical functions associated to build a ML project. Another important aspect is that Reptile is built using ‘shortest descent’ method of minimizing obstruction when learning anything new. This is possible by implementing a technique which incorporates memory points — known as fast weights and slow weights. The former stores long-term data while the latter stores temporary and pivotal new data. This forms a connectionist model where the learning is based on, and is similar to human cognition and behavior. With this, the fast weights compensate for the slow weights which sometimes will be a hindrance to the functioning of the algorithm. Even though Reptile and FOMAML have similarities in computing algorithm updates, they have a different take when it comes to addressing a problem. One such example is one-shot and few-shot techniques for optimizing random parameters during the process.","excerpt":"With Machine Learning (ML) advancing its frontiers day by day, developing base algorithms to fulfill the needs of self-learning in machines is a challenging task. In meta-learning, as the algorithms encounter a horde of data, they become more prone to providing useful output rather than giving out incomprehensible output. One such algorithm developed by OpenAI, […]","categories":["Global Tech"],"tags":["stochastic gradient descent"],"author_name":"Abhishek Sharma","publish_date":"2018-03-20T11:40:21","publication_year":"2018","word_count":772,"keywords":["stochastic gradient descent","machine learning","artificial intelligence","OpenAI","AI","neural network","ML","Ray","Aim","few-shot learning","TensorFlow"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","neural network","OpenAI","Aim","Ray","TensorFlow","few-shot learning"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/understanding-reptile-a-scalable-meta-learning-algorithm-by-openai\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":46445,"title":"Top Countries That Are Betting Big On AI-Based Surveillance","content":"A CCTV camera is only as good as the person sitting behind the monitor. Manual monitoring is not perfect because it involves humans who are inconsistent and are prone to fatigue. But artificial intelligence systems are relentless. They not only help monitor round the clock but also assist the humans in providing real time insights for real time action. Modern day computer vision algorithms have mastered face recognition, object recognition and event recognition tasks. These advantages of having robust security systems are forcing the many nations around the world to either develop home grown tech or buy from top sellers like China and the US. Last week, the Carnegie Endowment for International Peace released a report claiming that at least 75 countries are actively using AI tools such as facial recognition for surveillance. The new report says a growing number of countries are following China’s lead in deploying AI to track citizens. Source: Report by the Carnegie Endowment for International Peace Key takeaways from the report: At least seventy-five out of 176 countries globally are actively using AI technologies for surveillance purposes. China’s Huawei alone is responsible for providing AI surveillance technology to at least fifty countries worldwide. AI surveillance technology supplied by US firms is present in thirty-two countries. Here we take a look at a few cases where top countries have hinted at how serious they are about AI-based surveillance: China We all have watched those famous videos where people walk into stores, get their face scanned, collect their product and leave. Today, China has become the epicentre of AI-based monitoring. There are schools that scan faces of kids before they are allowed inside. It doesn’t just end there. Top Chinese companies like Huawei, Hikvision and ZTE supply AI surveillance technology in sixty-three countries. Of these 63, there are countries like Zimbabwe and Venezuela, which are originally known for human rights violations and Chinese presence does ring alarm bells. USA The US is no alien to advanced monitoring systems. Though these systems haven’t found their way into civilian premises like that of China, thanks to whistleblowers like Edward Snowden, strategies of government agencies like the NSA have been exposed to the world. Today, AI surveillance technology supplied by US firms is present in thirty-two countries. The most significant US companies are IBM (eleven countries), Palantir (nine countries), and Cisco (six countries). Israel Over the past three decades, Israel has made giant strides in developing advanced surveillance systems. Today it is also being called as the hub for AI-based surveillance systems. AI-based surveillance is being used to build a “smart fence” to separate Jerusalem from the West Bank. Israeli companies are developing laser-enhanced camera, radar and a communications system that can scan a two-mile radius to detect motion. These captured images are later analysed using AI to segregate humans from the rest. Impressed by this border surveillance technology, the US, in spite of being a power house of AI itself, is incorporating Israeli technology at the Mexican border to detect movement as far as 7 miles. Whereas China on the other hand, Bis using technology built by Israeli startup Xsight Systems at Beijing Daxing International Airport to help it improve the safety of its runways by monitoring for debris or hazards. Russia Moscow already has more than 150,000 cameras monitoring its 12 million population. Now the authorities look to augment the monitoring systems with AI. These systems are aimed at identifying suspicious actions: someone very quickly waving his arms, running, grabbing an object that resembles a weapon. Moscow’s face recognition start-up Ntechlab is the frontrunner to deliver services to the mass surveillance programme that is underway. India The Maharashtra government recently deployed the technology in Mumbai, where it will be integrated with the 10,000-strong contingent of CCTV cameras. Police in Delhi, Amritsar, and Surat have been using facial recognition since as early as mid-2018. Recently, the Bengaluru traffic department too, has announced that the signals shall be operated on AI. With its neighbour producing world-class AI-based monitoring systems, the Indian army is exploring various image recognition technologies that can replace patrolling on foot at the Chinese frontier. According to the national strategy report released by NITI Aayog, last June, on the use of AI, monitoring and surveillance systems occupied a significant portion. For instance, the banking sector has been highlighted for the potential use cases in the AI domain. This report also talks about how Banking and Financial Services sector in India have improved over the years and the room for AI adoption to monitor proactively with development of credit scores through analysis of bank history or social media data; and fraud analytics for prevention of various instances of fraud, money laundering,malpractice, and the prediction of potential risks. The Government’s aggressive push to become a global hub for advanced technology definitely compliments their AI-based aspirations. South Korea As a result of discovering a North Korean drone in Paju in March 2014, South Korea’s Ministry of National Defense has adopted a drone detection radar based on an overseas technology. The radar systems used by South Korean army are purchased from abroad such as drone detection radars from RADA in Israel and Blighter in the U.K. RADA especially owns a superior radar technology performance that can detect drones that are a maximum of 3km away. France In France, the port city of Marseille initiated a partnership with ZTE in 2016 to establish the Big Data of Public Tranquility project. The goal of the program is to reduce crime by establishing a vast public surveillance network featuring an intelligence operations centre and nearly one thousand intelligent closed-circuit television (CCTV) cameras (the number will double by 2020). We can’t put too much blame on the authorities either as France witnessed gruesome terror activities the past couple of years. These events force the authorities to take up the most extreme measures to contain any future mishap. Local authorities claim that this project with ZTE shall make Marseille “the first ‘safe city’ of France and Europe. UK At the England versus Wales rugby match in Cardiff on February 23, more than 18,000 scans resulted in 12 potential matches against a watchlist of 700 suspects, according to figures published by South Wales Police. According to the official parliamentary record Hansard, the word “facial recognition” has been referenced to 59 times in the UK parliament since 2014. The policymakers are having mixed opinions regarding large scale deployment of these surveillance systems. And, incidents like those of Glasgow and Gatwick aren’t helping either. European Union The European Union is testing a technology called iBorderCtrl in three countries — Greece, Hungary, and Latvia — to screen migrants at border crossings. Individuals are asked questions about their countries of origin and circumstances of departure. The answers are then evaluated by an AI-based lie-detecting system. Future Under Scanner Stanley Kubrick’s 2001: A Space Odyssey Orwellian states were once considered as the benchmark for fictional dystopian folk-lore. Today, the phrase ‘big brother is watching’ is no more a wink-wink, nudge-nudge joke. Though the case against AI systems for inappropriate has been well documented, there is a darker aspect to this. The accuracy of the algorithms isn’t foolproof yet. The induction of biases such as those of race and gender still are pervasive. And, an innocent behind bars because some algorithm mistook an umbrella for sniper, is quite a possible scenario. AI surveillance systems are usually established to ensure safer cities through crime prevention in real-time, build smart cities through crowd scanning for sentiment analysis and localised policymaking. However, these applications don’t even scratch the surface of what these systems are capable of doing. Since the road to hell is paved with good intentions, stricter regulations and transparency should be at the top of any state agenda before adopting large scale surveillance.","excerpt":"A CCTV camera is only as good as the person sitting behind the monitor. Manual monitoring is not perfect because it involves humans who are inconsistent and are prone to fatigue. But artificial intelligence systems are relentless. They not only help monitor round the clock but also assist the humans in providing real time insights […]","categories":["AI Trends"],"tags":["edward snowden","surveillance"],"author_name":"Ram Sagar","publish_date":"2019-09-26T19:00:54","publication_year":"2019","word_count":1299,"keywords":["big data","Go","artificial intelligence","AI","surveillance","sentiment analysis","image recognition","computer vision","Aim","analytics","edward snowden","R"],"extracted_tech_keywords":["AI","artificial intelligence","computer vision","analytics","Aim","sentiment analysis","image recognition","R","Go","big data"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-countries-that-are-betting-big-on-ai-based-surveillance\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":26318,"title":"How This Startup And Their AI-Powered Products Helped Global Banks Save Up To $10 Million In Regulatory Fines","content":"Assertion, a startup founded in 2014, is an enterprise-grade platform for automated compliance of multiple enterprise applications. They offers solutions that make it easy for customers to ensure that enterprise applications are compliant with organisational policies, thus reducing risks. Claiming to be the first in the world that looks at the compliance of everything, this venture-backed startup was founded by Srini Narayanan and Pradeep Vasudev, who swapped their corporate lives to run their own venture. With a focus on industries such as banking, financial services, insurance, IoT, healthcare, and contact centres, among others, they are building products which are cloud-first, use AI and big data to provide real-time and predictive compliance of any data. Analytics India Magazine caught up with the founders to know more about their RegTech products, growth plan and more. Starting Assertion Srini Narayanan shared that Assertion was born out of a specific company need. “A top bank in the world wanted a solution for a pertinent problem they faced. They had failed an audit with a regulator and wanted to create processes and policies to manage their global communications network,” he said. This is when they created controls and helped them automatically enforce them. Thus, Assertion was born, with a focus on automating compliance, governance and lowering risk for top corporates. Since then the 35-member team have been driving solutions for automating regulatory and corporate compliance. Being the world’s only platform that can automate multiple regulations, it uses a combination of secure connections to devices to process any kind of real-world data and automate complex problems that have largely been manual till now. They have enabled a noted bank in Singapore to save up to $10 million in regulatory fines by identifying shortfalls in the system. Solution And Use Cases Pradeep Vasudev shares that custom compliance is a part of every business and that they have the perfect platform in Compliance Studio to address that market. “In fact, most companies come to us because we not only automate many regulatory compliances across the world, but provide an infinite possibility to automate any of their corporate compliance using it,” he said. Explaining a use case of how their product is used, Vasudev gives an example of their largest customer, a global bank headquartered in the UK, where they needed a solution that could help them be always compliant. And their Enterprise Compliance Solution delivered just that. With over 20 products making their communications infrastructure and 600 assets spread across the world, their cost of manual compliance was excessive. Moreover, it took them a complete year for a team of people to collect evidence from these assets to prove compliance. But, once the team automated this using 300 controls, the time to scan the global network reduced to under a day. “Now, automatically the system scans their global network daily and sends a notification email when violations are identified. The people do far more valuable jobs now than just repeated work of checking compliance,” he shared. Their system is over 90 percent cheaper and 10 times faster than humans. With its adoption, there are no longer trust issues of having to share logins to critical assets with employees, hiring and training issues, errors (unintentional or intentional) of reporting, time to prepare reports and managing all those excel files in which they used to maintain daily reports. The Artificial Intelligence Play Narayanan shares that handwriting recognition is one area where they use AI. For instance, while doing compliance for banks, handling their account opening process and others, customers fill out forms, provide evidence of their address, identity, income and more. They use AI in understanding the content of these forms and handle the compliances accordingly. Patterns and historical evidence provide interesting insight into compliance. Sharing a use case, he said, “assume there is a bank manager who is allowed to inspect the safe room and he has the requisite access. So, if he goes there, it is not a fault and security systems will allow that. But say in the last year he has entered the room at an average of once a day, but today he has entered five times. Our system will identify a change in pattern, even minor changes in the behaviour.” Technologies like AI plays a crucial role here. Automating Compliance As Vasudev explains, there are many stages in the automation of a regulation and they take extreme care in each step. “The first step is to author controls. It is a very critical step and is prone to interpretation errors, so, correctness and completeness is of paramount importance. We get the controls reviewed by experts, we cross-check with third-party sources, and sometimes even with the regulators,” he said. He further shares that all the controls and reports on the system are verified multiple times, then digitally signed, encrypted and loaded. The databases are tamper proof, so once a system is in production use, even if someone has the administrator access, they cannot tamper with the evidence reports. He added that security, data correctness and completeness is more important for them than speed, cost or performance. Growth Story With lots of clients under their belt, Assertion is self-sufficient as of now. However, they are looking forward to raising the next round of capital in the future to expand globally. That’s not all, they are also looking to expand their team size, especially with people who can bring AI into the system.","excerpt":"Assertion, a startup founded in 2014, is an enterprise-grade platform for automated compliance of multiple enterprise applications. They offers solutions that make it easy for customers to ensure that enterprise applications are compliant with organisational policies, thus reducing risks. Claiming to be the first in the world that looks at the compliance of everything, this […]","categories":["AI Startups"],"tags":[],"author_name":"Srishti Deoras","publish_date":"2018-07-10T13:21:08","publication_year":"2018","word_count":906,"keywords":["Go","artificial intelligence","AI","Git","RAG","Ray","Aim","analytics","Rust","R"],"extracted_tech_keywords":["AI","artificial intelligence","analytics","Aim","Ray","RAG","R","Go","Rust","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-startups\/how-this-startup-and-their-ai-powered-products-helped-global-banks-save-up-to-10-million-in-regulatory-fines\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":34519,"title":"How Computer Vision Became An Important Manifestation Of Artificial Intelligence","content":"Today, top global firms are increasingly making use of computer vision across domains — from autonomous driving and retail (Amazon Go) to healthcare (Amazon Deeplens and DermLens) — to improve operations and enhance productivity. The graph for development in computer vision in application areas has increased exponentially over the last few years. It is now being widely used by global tech giants like Amazon’s Rekognition, or Indian companies like FaceX. How Computer Vision Differs From Other Artificial Intelligence Techniques We can differentiate this technique with other AI techniques in two possible ways: This technique does not predict as other artificial intelligence techniques do, it gives the outcomes depending upon the present content of images and it will be able to gain more accuracy with the passage of time. The data collecting process is way easier than the other techniques of AI. The worldwide library of images and videos are growing every day as millions of images and videos are being shared every day through Facebook, Snapchat, Instagram, etc. How Availability Of Datasets Has Spurred Development In Computer Vision Within a couple of years, large datasets are made available online in order to allow the researchers to innovate this technology in every possible way. Applications of this technique are catching up in almost every field as well as the tech giants seemed to take big leaps in this regard. Some of the instances are listed below Autonomous Cars: The big tech companies like Microsoft, Intel, Apple, etc. along with the traditional car manufacturers like Hyundai, Honda, etc. are in a rat race to manufacture driverless cars. Computer Vision makes sense of the input as well as rely on cameras or other types of sensors to perceive the world. OpenVINO Toolkit: This toolkit by Intel enables CNN-based deep learning inference on the edge and supports heterogeneous execution across computer vision accelerators including CPU, GPU, FPGA and will be used to enable Vision-based intelligence at the edge across multiple architectures. Detection of Vehicle Damage: US-based CCC Information Services Inc. uses Computer Vision AI to automatically detect vehicle damage and visually depict that damage using unique heat maps. The damage detection technology and heat maps bring depth and insight into otherwise opaque damage photos that are fed into the computer vision and the algorithm detects the damages through the already fed training dataset. Vision Framework: This framework by Apple is used to perform various tasks on input images like face detection can be traced in the latest mobile operating system by Apple. The iOS 10 recognises different faces from photo library and sorts them into folders. This framework applies computer vision algorithms to perform a variety of tasks to input images and video. Microsoft’s XiaoIce: This prominent chatbot from China use computer vision techniques to paint, draw and even comment to your shared images. Computer Vision makes it possible for this chatbot to look at images, point out objects and make inferences. Cashierless Grocery Stores: According to a report, The largest online retailer, Amazon is planning for opening 3,000 new AmazonGo cashier-less stores by 2021 that uses the technique of Computer Visioning. Computer Vision works as tracking your movements through the store after you’ve scanned your Amazon account at the front and also monitors when you take items off shelves. Lens: This tool is launched by photo-sharing app Pinterest as a beta product that works with the help of computer vision algorithms to recognise an object and provide contextual information to aid your in-store experience. It lets people use their camera phone to get recommendations inspired by objects they see in the real world. Project InnerEye: This project by Microsoft build innovative tools for the automatic, quantitative analysis of three-dimensional radiological images. The project develops computer vision and machine learning techniques for the automatic delineation of tumours as well as healthy anatomy in 3D radiological images. A Vision Of Future Tractica  forecasts that global revenue from computer vision software, hardware, and services will grow from $1.1 billion in 2016 to $26.2 billion by 2025. The market intelligence firm expects that the growth in computer vision will be driven by a wide range of industries and an increasing diversity of applications. According to this analysis, there are top 10 cases for computer vision during the year 2016-2025 that will be forecasted. They are mentioned below: Video Surveillance Machine\/vehicular object detection\/identification Medical Image Analysis Augmented Reality\/ Virtual Reality Localisation and Mapping Converting paperwork into digital data Human emotional analysis Ad insertions into images and videos Face Recognition Real estate development optimisation","excerpt":"Today, top global firms are increasingly making use of computer vision across domains — from autonomous driving and retail (Amazon Go) to healthcare (Amazon Deeplens and DermLens) — to improve operations and enhance productivity. The graph for development in computer vision in application areas has increased exponentially over the last few years. It is now […]","categories":["AI Features"],"tags":["Computer Vision"],"author_name":"Ambika Choudhury","publish_date":"2019-02-06T06:21:36","publication_year":"2019","word_count":755,"keywords":["Go","machine learning","artificial intelligence","AI","ML","Git","computer vision","deep learning","object detection","Computer Vision","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","computer vision","object detection","R","Go","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-computer-vision-became-an-important-manifestation-of-artificial-intelligence\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":33494,"title":"How Is An Algorithm Designed? Here’s The Thought Process Behind It","content":"If we said that your day began and ended with an algorithm, it would still be an understatement. From turning off eerie chimes of alarm clocks to pulling down the curtains after binge-watching Netflix, modern man is under a deluge of algorithms. Designing the right algorithm for a given application is a creative process. A problem has to be identified, a solution has to be sculpted out — sometimes remote and often oblivious to the domain of problem under consideration. Eventually, these pieces of solutions are threaded together to work in congruence with each other like a well-oiled contraption. Being a successful algorithm designer requires a certain attitude — the ability to ask oneself questions to guide through the thought process. To use it effectively, you must not only ask the questions but answer them as well. For an algorithm designer, there is never an answer which says just “no”. But, “No, because…” Understanding The Problem The first step for designing an algorithm is to begin with the premise of the problem itself, like, the components of an input and the desired results, or the importance of optimisation in obtaining the answers. Because this exploring this question will decide whether to use more or fewer steps in an algorithm. When the significance of an optimised solution gets decided, factors like speed come into the picture. As a numerical problem is different from a graph algorithm problem which in turn is different from a geometric problem. Different problems, different speed hence varied efforts. Redundancy Check This is more or a less an obvious one because who wants to reinvent the wheel? Every problem will be searched for available solutions both popular and obscure ones. In case of finding none, an algorithm designer has to recheck for other domains with sub-branches that overlap with the problem under consideration. What Heuristics Since, the above step gives an idea about the accuracies, time and speed, one can then conclude on the magnitude of effort required. And, if the solution is to be checked for precision and it needs a test which will lead to searching for the tests available or design a customised test. If a solution is found to be working then it should be checked for its repeatability i.e whether it gives approximately similar results with changing data; large value in the sample first or the smaller sample or a random one altogether. After determining the characteristic of the solution to certain samples, the ill-performing heuristics should be noted as well. Checks And Balances To Make A Case For ML Algorithms The sections prior to this dealt with how to start with a problem at a fundamental level. From a machine learning perspective, which deals with large chunks of unprepared, non-linear data, questions such as time and speed and, accuracies become increasingly important and the modern day gadgetry provisions it as well, with all the advancements in computational devices. The following are few questions a machine learning researchers would face or should ask before they posit a new algorithm: Are there any special cases for the problem under consideration? Why can’t this special-case algorithm be generalized to a wider class of inputs? How does the solution behave when few input parameters are left out? Is there a way to split the problem into two smaller problems? Does this suggest a divide-and-conquer algorithm? How can dynamic programming be used to exploit the order of the elements; like the branches in a decision tree? Can directed randomness be used to simulate the actual scenarios? Is the problem similar to NP-complete problem? If yes, then is that the reason behind improper efficiencies? Conclusion Problem-solving is not a science, but part art and part skill. It is one of the skills most worth developing. And, finally, the most important question that a researcher should save for the last, like we did in the article is, why not start again from the beginning and work through these questions again and if any of these answers have changed since the last trip?  Noted author and computer researcher Steven Skiena rightly said, “In algorithms, as in life, persistence usually pays off.”","excerpt":"If we said that your day began and ended with an algorithm, it would still be an understatement. From turning off eerie chimes of alarm clocks to pulling down the curtains after binge-watching Netflix, modern man is under a deluge of algorithms. Designing the right algorithm for a given application is a creative process. A […]","categories":[],"tags":[],"author_name":"Ram Sagar","publish_date":"2019-01-15T07:40:23","publication_year":"2019","word_count":692,"keywords":["Go","machine learning","programming_languages:R","AI","ML","programming_languages:Go","GRU","GAN","R"],"extracted_tech_keywords":["AI","machine learning","ML","R","Go","GAN","GRU","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-is-an-algorithm-designed-heres-the-thought-process-behind-it\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":61541,"title":"Apple &#038; Google Collaborates To Develop COVID-19 Contact Tracing Technology","content":"In a recent blog posted by Apple, the company has announced its partnership with Google to develop a contact tracing technology in order to help governments and health agencies to carry out contact tracing of people affected by COVID-19. According to a release issued by the companies, the two technology giants — Apple and Google, are working together to enable the use of Bluetooth for government and healthcare agencies to identify people coming into contact with a person with COVID-19, while keeping privacy and security in the forefront. The release stated that Apple and Google are developing a comprehensive solution, which would include application programming interfaces and operating system level technology. The two giants are releasing technical documentation of Bluetooth specifications, cryptography specifications, and a framework API for contact tracing. The release stated — “Given the urgent need, the plan is to implement this solution in two steps while maintaining strong protections around user privacy.” “First, in May, both companies will release APIs that enable interoperability between Android and iOS devices using apps from public health authorities. These official apps will be available for users to download via their respective app stores.” “Second, in the coming months, Apple and Google will work to enable a broader Bluetooth-based contact tracing platform by building this functionality into the underlying platforms. This is a more robust solution than an API and would allow more individuals to participate if they choose to opt-in, as well as enable interaction with a broader ecosystem of apps and government health authorities. Privacy, transparency, and consent are of utmost importance in this effort, and we look forward to building this functionality in consultation with interested stakeholders. We will openly publish information about our work for others to analyse.” The company stated that no GPS location data or personal information would be recorded for this system. How the contact tracing tech works — by Google. This rare partnership between the longtime rivals — Apple and Google — is a much need of the hour, as the pandemic continues to grow. Both the companies have been under several concerns and pressure in order to use their innovative resources to help fight the pandemic. Explaining the workings, Google stated that this initiative of developing contact tracing would be playing a key role in reducing the spread of the COVID-19. The system would work by identifying the people who have been infected by the virus, and all the individuals they have come into contact with, who could, in turn, be potentially infected as well. How a person is notified, they were in contact with a contagious person — by Google. Apple and Google have further released a draft technical documentation, which is including Bluetooth and cryptography specifications and framework documentation. The companies jointly stated — “All of us at Apple and Google believe there has never been a more important moment to work together to solve one of the world’s most pressing problems.” “Through close cooperation and collaboration with developers, governments and public health providers, we hope to harness the power of technology to help countries around the world slow the spread of COVID-19 and accelerate the return of everyday life,” concluded by the release.","excerpt":"In a recent blog posted by Apple, the company has announced its partnership with Google to develop a contact tracing technology in order to help governments and health agencies to carry out contact tracing of people affected by COVID-19. According to a release issued by the companies, the two technology giants — Apple and Google, […]","categories":["AI News"],"tags":["Apple","covid-19","covid19 data"],"author_name":"Sejuti Das","publish_date":"2020-04-13T13:32:02","publication_year":"2020","word_count":533,"keywords":["Go","API","covid-19","AI","programming_languages:R","Apple","programming_languages:Go","covid19 data","R"],"extracted_tech_keywords":["AI","R","Go","API","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/apple-google-collaborates-to-develop-covid-19-contact-tracing-technology\/","complexity_score":2,"technical_depth":6,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":26372,"title":"Top 10 Indian Government Datasets For Analytics Projects","content":"Open datasets have only now started becoming available for researchers, analysts, professionals and students to carry out various projects and research. In new tech fields like analytics, machine learning and artificial intelligence, there is a constant need for datasets to perform tasks like planning projects, building models or using it for education. While there are several readily-available datasets provided by companies, communities, government bodies, government institutions and others, there is often a need to clean these data before it is usable. These datasets can be accessed either freely (which most government websites are) or by paying nominal charges. We are listing here 10 such India-based datasets from the public domain and government bodies which can come in handy for your next projects. 1. RBI Database Of Indian Economy The RBI database is a website launched by Reserve Bank of India and has data on the macroeconomic indicators of the Indian economy. It is loaded with relevant information and data for researchers, analysts, and general users all alike. It has datasets across money and banking, financial markets, national income, saving and employment, and others. The idea is to facilitate contemporary styles of data analysis that can provide important real-time numbers about economic activity, prices and more. 2. Ministry Of Statistics And Programme Implementation Dataset This is the dataset provided by MOSPI, a Union Ministry concerned with the coverage and quality aspects of statistics released. The datasets are collected by conducting large-scale sample surveys across India for various parameters, which eventually leads to the creation of the database. The ministry applies standard statistical techniques and extensive scrutiny and supervision to enable this. 3. Gateway To Indian Earth Observation An initiative by ISRO, the open data archive provides free satellite data, products download facility and thematic datasets. It uses a crowdsourcing approach to collect enriching and point-of-interest data. It also acts as a platform to host government data such as forest department. Apart from being a repository of data, it allows users to explore the 2D and 3D representation of the surface of the earth, pest surveillance, disaster services, high-resolution imagery of cities, among others. 4. National Portal of India A web portal for India citizens, it was developed by the Indian government with an objective of facilitating a single window access to information and services of all government entities. It is designed and developed jointly by National Informatics Centre (NIC), Ministry of Electronics & Information Technology. A single point access to a lot of information, it has a searchable contact directory, a database of the government website, and others. 5. Survey Of India India’s central engineering agency the Survey of India is in charge of mapping and surveying, under the Department of Science and Technology, and is one of the oldest scientific departments. With data centres spread across India, it has user-focused, cost-effective, reliable and quality geospatial data from across India. 6. India Weather Data With datasets for various meteoroid indicators, water resource planning, rainfall, and others from across various parts of India, these datasets are available for users in simple formats. It also contains databases for several other parameters such as temperature, pressure, relative humidity, precipitation amount, wind speed, solar radiation, among others. 7. Aadhaar Metadata This provides a huge database generated by the daily count of total registrations, enrolment applications accepted and rejected by state and district. It also contains other details such as Aadhaar generated by age, gender, etc. 8. Import Exports Datasets ICEGATE or the Indian Customs Electronic Commerce\/Electronic Data Interchange (EC\/EDI) Gateway is a portal with e-filling services for trade and cargo carriers. It also has an exhaustive  National Import Database (NIDB) and Export Commodity Database (ECDB) for Directorate of valuation that is being handled by the ICEGATE. It has information such as documents, messages, and others processes by the customs’ end by the Indian Customs EDI System (ICES). 9. Open Government Data (OGD) Platform India Set up by the National Informatics Centre (NIC) in compliance with the Open Data Policy (NDSAP) of India, OGD platform gives access to government-owned shareable data along with its information about its usage in an open and machine-readable format through a wide area of network across the country. A part of the Digital India initiative, it has been developed by using Open Source Stack. It publishes datasets, documents, tools and applications collected by government for public use and community participation of the products with visualisation, APIs, alerts etc. It is also a collection of all the government based datasets discussed above. 10. Wildlife Institute of India Dataset An autonomous institution under the Ministry of Environment Forest and Climate change, Government of India, it has datasets on different wildlife species in India. There are a total of 4591 specimens that are housed at WII herbarium, of which 4322 are digitised and published through the GBIF network. The data is mainly used by researchers and field managers from the respective protected areas of the country to prepare for management plan and other research.","excerpt":"Open datasets have only now started becoming available for researchers, analysts, professionals and students to carry out various projects and research. In new tech fields like analytics, machine learning and artificial intelligence, there is a constant need for datasets to perform tasks like planning projects, building models or using it for education. While there are […]","categories":["AI Trends"],"tags":["free datasets for analysis","full stack project ideas","high paying jobs in india","indian statistical service"],"author_name":"Srishti Deoras","publish_date":"2018-07-12T05:41:47","publication_year":"2018","word_count":830,"keywords":["Go","API","free datasets for analysis","artificial intelligence","machine learning","AI","high paying jobs in india","Git","RAG","ViT","analytics","R","indian statistical service","full stack project ideas"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","analytics","RAG","R","Go","Git","API","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-10-indian-government-datasets-that-you-can-use-for-analytics-projects\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10007311,"title":"Using Orange To Build A Machine Learning Model","content":"Orange is an open-source, GUI based platform that is popularly used for rule mining and easy data analysis. The reason behind the popularity of this platform is it is completely code-free. Researchers, students, non-developers and business analysts use platforms like Orange to get a good understanding of the data at hand and also quickly build machine learning models to understand the relationship between the data points better. In this article, we will be understanding: Why Orange?Installing and Setting up the tool Training your first machine learning model using Orange Why Orange? Orange is a platform built on Python that lets you do everything required to build machine learning models without code. Orange includes a wide range of data visualisation, exploration, preprocessing and modelling techniques. Not only does it become handy in machine learning, but it is also very useful for associative rule mining of numbers, text and even network analysis. Orange comes with multiple classifications and regression algorithms and all of these are implemented with drag and drop features.  It can be used as an intuitive user interface or, for more advanced users, as a module for the Python programming language. Installing and Setting up the tool There are a few options you can use to install the tool. Firstly, you can click here to download the tool. You will see the following options to install the tool according to which Operating system you use. Based on the OS, you can select the latest version of orange and the download will automatically begin. Once it is downloaded you can run the .exe file and launch the platform on your computer. The next option is to install it using conda or pip command if Anaconda Navigator is already set up in your system. To install the platform with conda use the command: conda config --add channels conda-forge conda install orange3 This will add the orange channel to your anaconda navigator and you can launch the tool from the navigator. Since orange is built with python the last option to install it would be through the pip command. It is better to create a virtual environment for this since a lot of dependencies need to be downloaded. pip install orange3 Training your first machine learning model using Orange Once the tool has been installed and setup you will be directed to the main page of the platform. As shown above, you have some options on the left like file, CSV file, datasets etc. To keep this demonstration simple I will choose a built-in dataset. There are a number of datasets that are readily available which can be seen by double-clicking the datasets option. You can add your own CSV file by clicking the CSV option also. After deciding which dataset you want to work with we can proceed with file creation. I have selected a simple heart disease dataset. To create the file, drag and drop the ‘File’ to the right-hand side. Double-click on the file to select the dataset and view the dataset. If you have uploaded the dataset using CSV option the dataset will automatically show up in the file. Now, you need to view the dataset in the form of a table to understand the features and target better. To do this you can drag a connecting line and you will see a list of options. Select the option called ‘Data table’ and double click on this component. Data cleaning After loading the data it is important to produce clean data for machine learning model implementation. To clean the data, drag a connecting line again from the file component and select impute option. In the impute component you can clean the data by filling in missing values either by mean on random data. I have selected the option to fill missing data(if any) with the average values. You can also remove the rows with missing values or impute the data with model-based imputer. Data visualizations Once we have clean data it is time to visualize the data distribution. To do this drag a connecting line from impute component and select the scatter plot option. Double-click the scatter plot component. Here you are presented with options to change the values in x and y axes, change colours of the points, change opacity etc. You can perform all the visualizations here and understand your data distribution. After visualization the data it is time to build a model. Our dataset indicates that it is a classification type problem and requires a classification algorithm. Since the target contains values of 0 and 1 we can use a simple logistic regression model. Note: Since heart disease data is a built-in dataset, the target is marked automatically. If you are using a custom dataset you can set the target by navigating to file component and double-clicking on the column you want to set and selecting the target option. Model building Drag a connecting line from the impute component and select the logistic regression option. Click on the component. Here you can set your preferences in regression methods like a lasso or ridge regression and also adjust the C values. Once this is done, drag a connecting line from the logistic regression component and select the option of test and score. Be sure to connect the impute component also to test and score component. Once done, the model will begin to train and display the results. If you have a separate test dataset you can connect it here and select the option of test on test data as well. You can drag a connecting line from test and score component and select the option to display a confusion matrix. Thus, we have trained a machine learning model without writing code. You can train multiple models and compare them as well. Finally, to save the model you can press CTRL+S and save the entire orange window on the local system. Conclusion In this article, we saw the advantages that Orange platform provides especially for non-coders and implemented a simple machine learning model from scratch. Orange can be used for almost any kind of analysis but most importantly, for beautiful and easy visuals.","excerpt":"In this article, we will be understanding: Why Orange? Installing and Setting up the tool Training your first machine learning model using Orange","categories":["Deep Tech"],"tags":["Data Analysis","Machine Learning","machine learning computer","machine learning methods","no-code"],"author_name":"Bhoomika Madhukar","publish_date":"2020-09-14T11:00:37","publication_year":"2020","word_count":1025,"keywords":["Data Analysis","Go","machine learning","programming_languages:R","AI","machine learning computer","no-code","Machine Learning","programming_languages:Go","RAG","Python","programming_languages:Python","machine learning methods","R"],"extracted_tech_keywords":["AI","machine learning","RAG","Python","R","Go","programming_languages:Python","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/using-orange-to-build-a-machine-learning-model\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":516,"title":"5 ways IoT can bring changes to Retail Industry","content":"“Internet of Things” has existed for quite a while but its implementation in various industries has gained momentum recently. The benefits that IoT can bring to businesses is being recognized now and organizations are starting to take steps in this direction. Simply put IoT is a network of objects that can communicate through the internet. And this technology can be extremely beneficial to the retail industry if they act proactively and take initiatives. Yes, IoT is still in a nascent stage but still retailers can start taking benefit of it and upgrade as IoT evolves. Internationally, big retail chains has already starting using IoT for various business purposes. Here we bring you 5 different strategies which Retailers in India can employ to take advantage of IoT. Way #1: Managing Inventory efficiently with IoT Inventory management is one of the most important task for any retailer. Too much or too less of inventory can impact the business and can lead to losses. Hence maintaining the right levels of inventory is a big challenge. However, IoT can help in managing it efficiently. In fact, large retailers like Walmart have already started using IoT to manage their inventory. With the help of IoT, retailers can deploy sensors like radio frequency tags on the products and use it to track the products on a real-time basis. Through this the retailer can know much inventory is on the store shelf, in the storage or in the warehouse and based on this tracking he can take decisions with respect to product movements to avoid empty store shelves. Also through sensors, alerts can be raised when inventory dips to a certain level for it to be re-filled. This leads to better inventory management and efficient use of retailer’s working capital. Way 2: IoT can help to reduce retail shrinkage Retail shrinkage is a challenge faced by every retailer be it from shoplifters or stealing by store employees. IoT can help to reduce this. Retailers can take advantage of IoT by implementing smarter in-store technologies like smarter cameras or RFID tags to reduce shrinkage. Companies like Panasonic are working towards bringing in smart shelf technology and powerful cameras to enhance the store’s ability to track and manage their products. With more powerful cameras they can capture all the movement happening in the store and products with RFID (radio frequency identification) tags can be tracked if someone tries to take them out of store and if in any case it is stolen, these tags can help to identify the product when it is sold in gray market also. Way 3: Improved Maintenance and Warranty Maintenance and Warranty is another important aspect of retail selling and especially for products that require post-sale service or are under warranty. IoT can help to track the health of the product by fitting in sensors on the products sold and these sensors will send real-time data to retailer giving him information about malfunctions or warranty issues. This can improve post-sales service to customers. Also such data can help in product improvement and innovations. An excellent example is of General Electric. GE uses IoT for predictive maintenance for its jet engines, turbines, and wind farms. Hence by analyzing real-time data, GE technicians are able to know when to make visits for maintenance rather than unnecessary periodic trips Way #4: Price Optimization Right pricing for any product is very important to garner sales. IoT can help in optimizing prices automatically. It will compare the demand to supply of a particular product and also make comparisons of similar products available with other retailers to give an optimal price on a real-time basis. This will help the retailer grow his sales and will also be beneficial to the customers. Way # 5: Real-time Promotions Promotions are one of the key ways to increase sales and especially in the retail space. And real-time promotions with the help of IoT can be extremely beneficial. Retailers can collect data using IoT about customer’s preferences, shopping history, location, credit card activity and such other variables and based on analysis send only relevant and customized promotions to the customer. Such prioritizing and customization will help them to target the consumers in an efficient way. Concluding remarks By taking initiatives to imbibe Internet of Things in their business, retailers can take advantage and manage efficiently everything from inventory to customers. It’s on the part of the retail industry to be proactive and take steps necessary for successful implementation of IoT in their retail business. This will not only lead to customer satisfaction but also help the company in boosting their sales and reducing costs.","excerpt":"“Internet of Things” has existed for quite a while but its implementation in various industries has gained momentum recently. The benefits that IoT can bring to businesses is being recognized now and organizations are starting to take steps in this direction. Simply put IoT is a network of objects that can communicate through the internet. […]","categories":["AI Trends"],"tags":[],"author_name":"Manisha Salecha","publish_date":"2016-08-15T04:27:43","publication_year":"2016","word_count":770,"keywords":["API","programming_languages:R","AI","innovation","RAG","Ray","ViT","GAN","R"],"extracted_tech_keywords":["AI","Ray","RAG","R","API","GAN","ViT","innovation","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/5-ways-iot-can-bring-changes-retail-industry\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10020111,"title":"Global Crypto Frenzy, Chipageddon And More In This Week’s Top News","content":"On Thursday, Bitcoin hit a record  $48,000 mark after America’s oldest bank, New York Mellon announced that they will provide custody for digital assets. BNY Mellon said that it would begin financing cryptocurrencies and will eventually allow crypto assets to pass through the same financial network it currently uses for more traditional holdings like the U.S. Treasury bonds and equities. Even Mastercard flaunted its interest in crypto. On Wednesday, Mastercard released a statement. BTC’s unprecedented rallying– up more than 60% this year– also comes after Elon Musk’s Tesla announced that it had purchased $1.5 billion worth of bitcoin and would soon accept it as a form of payment. JAY-Z\/@S_C_ and I are giving 500 BTC to a new endowment named ₿trust to fund #Bitcoin development, initially focused on teams in Africa & India. It‘ll be set up as a blind irrevocable trust, taking zero direction from us. We need 3 board members to start: https:\/\/t.co\/L4mRBryMJe— jack (@jack) February 12, 2021 Twitter’s CEO, Jack Dorsey, was also among the big names who have backed crypto. Jack announced that he would be funding Bitcoin development in countries like India and Africa. Last week we read that India’s Reserve Bank is reportedly in plans to introduce an official digital currency for the country. According to the rumored Cryptocurrency and Regulation of Official Digital Currency Bill, Indian government can ban private cryptocurrencies in India. Whatever your opinions on cryptocurrencies, wrote Raj Dhamodharan of Mastercard, from a dyed-in-wool fanatic to utter skeptic — the fact remains that these digital assets are becoming a more important part of the payments world. After Canada, Sweden Slams Clearview AI Controversial facial recognition tech maker Clearview AI was slammed by the Canadian regulators for unlawful practices who found that Clearview AI scraped billions of images of people from across the Internet, which according to Canada, represented mass surveillance and was a clear violation of privacy rights. This week, the Swedish Authority for Privacy Protection concluded that Clearview AI has been used by the Police on a number of occasions.  The Swedish watchdog concluded that the law enforcers have failed to fulfill its obligations as a data controller on a number of accounts with regards to the use of Clearview AI. “The Police have failed to implement sufficient organisational measures to ensure and be able to demonstrate that the processing of personal data in this case has been carried out in compliance with the Criminal Data Act,” read the press release. Clearview AI had been used unlawfully by the police to process biometric data for facial recognition. Automakers Hit By Chip Shortage There is a global chip shortage and the auto industry is bearing the brunt. Semiconductor industry is struggling to compensate for the unforeseen delays caused by the pandemic along with other factors. Remote work, rise in 5G demand and the obvious lengthy process of making chips have halted the efforts of automakers who are trying to make up for all the losses of 2020. Chipageddon aka the global chip shortage is now also being acknowledged by the US lawmakers as they wrote to the National Economic Council to take necessary steps. This shortage, wrote the senators, threatens post-pandemic economic recovery. Also Read: Global Chip Crisis Just Got Real The senators urged the economic council to treat the auto industry as the significant semiconductor consumer. “We also strongly urge you to support efforts to secure the necessary funding to swiftly implement the semiconductor-related provisions in the most recent National Defense Authorization Act, which would boost the production of semiconductor manufacturing and incent the domestic production of semiconductors in the future,” wrote the senators. So don’t be surprised if booking your new car is taking more time! Google Cloud Boosts Python Community Efforts This week, with a focus on improving the supply-chain security of the Python ecosystem, Google Cloud announced a donation of more than $350,000 to support three specific PSF projects: Productionized malware detection for PyPI Improvements for foundational Python tools and services A CPython Developer-in-Residence for 2021 to help the CPython project prioritize maintenance and address its backlog “Python is critically important to both Google Cloud and our customers. It serves as a popular runtime for many of our hosted services, from the launch of App Engine more than a decade ago, to modern serverless products like Cloud Functions,” wrote Dustin Ingram of Google Cloud. Convict To Review Source Code A convict in the United States is facing charges for murder and his best chance to prove his innocence has boiled down to code review! According to reports, a court in the US  has ruled that this man who is accused of murder is entitled to review proprietary genetic testing software to challenge evidence presented against him. However, the software maker argued that this ruling can jeopardise their product by revealing many trade secrets. “The program consists of 170,000 lines of MATLAB code, is so dense it would take eight and a half years to review at a rate of ten lines an hour,” argued Mark Perlin, the co-founder of Cybergenetics. The court ruled that scrutinizing software’s source code is essential to validate the software. This ruling adds a new chapter in the ‘man vs machine’ debate as AI is poised to take over the world. No More Cookies On Chrome? Google is rewiring its online strategies by deciding to ditch third-party cookies on Google Chrome. These cookies are used for targeted advertisements but puts user’s privacy at risk. According to the WIRED, Google will deploy browser based machine learning that will skim through browser history and will weave a network of people’s interest(think: recommendation engine), which will be used for advertisement. Google Chrome is one the most widely used browsers in the world and this move to eliminate cookies can put many smaller advertising firms out of business. Big Tech To Get Taxed According to reports, the US lawmakers in many states are making moves to tax the big tech. The plan is to tax the revenue from digital advertisements sold by companies like Facebook, Google and Amazon. If materialised, this tax will generate as much as an estimated $250 million in the first year after enactment.","excerpt":"On Thursday, Bitcoin hit a record  $48,000 mark after America’s oldest bank, New York Mellon announced that they will provide custody for digital assets. BNY Mellon said that it would begin financing cryptocurrencies and will eventually allow crypto assets to pass through the same financial network it currently uses for more traditional holdings like the U.S. […]","categories":["AI News"],"tags":[],"author_name":"Ram Sagar","publish_date":"2021-02-14T10:00:00","publication_year":"2021","word_count":1030,"keywords":["Go","funding","machine learning","AI","Git","serverless","Python","Rust","GAN","R"],"extracted_tech_keywords":["AI","machine learning","serverless","Python","R","Go","Rust","Git","GAN","funding"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/bitcoin-tesla-chip-shortage-top-tech-news\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10079845,"title":"Is Python Slowly Eating R? The Reason Why RStudio Became Posit","content":"Following the company’s annual user conference in Washington, D.C. held on July 27, IDE (Integrated Development Environment) RStudio announced that it has a new name—Posit. The move signalled the company’s expansion plans with a focus beyond R, including users of Python and Visual Studio Code. Meanwhile, the open-source data science community Posit said, “While many things will stay the same, our rebrand will result in changes beyond a new name.” https:\/\/twitter.com\/posit_pbc\/status\/1587823489859264512 RStudio has been emphasising that its commercial products are “bilingual” for both R and Python for many years. However, the “RStudio” brand has made it difficult to convince organisations to consider its products for Python users. But does it bring about a belief that Python is somehow supplanting R for the data science ecosystem? Cocktail of languages At Snowday 2022, cloud company ‘Snowflake’ announced new additions to its platform that are set to help data scientists and developers figure out new ways to develop pipelines, applications, and ML models with the firm’s single data platform. Read: Snowflake is Now Python-friendly Over the years, Posit (formerly, RStudio) has transitioned from R-exclusive tooling to an agnostic language ecosystem. There has been a gradual shift of RStudio IDE to be more Python-friendly. RStudio—a name synonymous with open source R development—rebranded itself to better represent the evolving business. This led to the rebranding of tools and commercial products. RStudio Connect was renamed ‘Posit Connect’ and RStudio Workbench as ‘Posit Workbench’. RStudio said in a tweet that RStudio IDE will still be around to offer help with the open source R development. Hi, RStudio IDE team here! Just want to clarify that the name of the open source RStudio IDE is NOT changing. It's always been, and always will be, RStudio.#rstats https:\/\/t.co\/Q3f1PVBGz5— RStudio Tips (@rstudiotips) November 2, 2022 RStudio’s chief scientist Hadley Wickham said, “We’re not pivoting from R to Python.” He further explained, “I’m not going to stop writing R code. . . I’m not going to learn Python,” putting users’ concerns at ease. Although RStudio is seeking to balance the share of engineers working on R with other advancements over time, the company claims that the majority of work will continue to be R-related. Can Python replace R? Python and R are used for similar purposes, but differ in essence. Python is a high-level, object-oriented programming language that comes with built-in data structures, making it a top language for the development of applications. Python syntaxes are simple and easy to read. On the other hand, R is a programming language used for statistical analysis of data and comes with a wide range of techniques of linear modelling, statistical tests, non-linear modelling, and clustering. One of the core strengths of R is easy production of a plot, including mathematical notations and formulas. However, both the languages are the preferred for data science, data analysis, and machine learning.  R primarily focuses on the statistical aspect of a project while Python is flexible in its data analysis and usage. R plays an important role in visualising data in graphs. However, it is difficult to use this language in a production environment because of its ‘yet-to-develop’ production tools. In contrast, Python can be easily integrated in a complex work environment. When it comes to performance, users prefer Python as it runs faster than R in all environments. However, a user posted on Reddit that Python has libraries that are ‘embarrassing’ compared to feature-rich ones in R. (Source: Reddit) Nevertheless, both languages are among top favourites for users to work with depending on their usage in a given environment. A Single Home for R and Python With R and RStudio communities, the company has helped users pose and answer difficult questions around data. By building open source tools to make “code-first” data science accessible to millions of people to establish reproducibility as a baseline for analysis and communication, the company aims to foster developments in a diverse community. RStudio said that one of the core ideas that the community believed in was the imperativeness of using open source software for scientific work. “Scientific work needs to be reproducible, resilient, and must encourage broad participation in the creation of the tools themselves.” Hadley Wickham said, “The name had just started to feel increasingly constraining.”  Both Wickham and CEO J.J. Allaire emphasised that the rebranding doesn’t signify a shift away from R. However, a user claims that above all, the foremost problem with R is governance. On Python github, users will be able to see thousands of pull requests and issues, with various people who are trying to contribute to the core language. In addition, Python even holds elections with anyone theoretically becoming a “core python developer”. The user further said, “Do you want to contribute to the core R language? You can’t. R is open source in terms of the source code being available, but completely closed in terms of development. You cannot even directly create an issue on the bug tracker if you find a bug.” If Python is a democracy, R is the feudal system. Additionally, since R core developers are unelected and the amount of core developers can be counted—it makes Python more diverse than R.","excerpt":"Python and R are used for similar purposes, but differ in essence.","categories":["Deep Tech"],"tags":["Data Science","Python","R","Snowflake"],"author_name":"Bhuvana Kamath","publish_date":"2022-11-16T13:00:00","publication_year":"2022","word_count":861,"keywords":["data science","Go","machine learning","AI","ML","RAG","Python","Aim","Data Science","R","Snowflake"],"extracted_tech_keywords":["AI","machine learning","ML","data science","Aim","RAG","Snowflake","Python","R","Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/is-python-slowly-eating-r-the-reason-why-rstudio-became-posit\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10092341,"title":"OpenAI Spoils Users With More Data Control for ChatGPT","content":"Amid rising concerns around data misuse, OpenAI today announced a new feature that lets users turn off chat history in ChatGPT. The company also announced its plans to launch ChatGPT Business in the coming months. OpenAI, in its blog post, said that conversations that are initiated when chat history is disabled won’t be used to train and improve their models, and will not appear in the history sidebar. Further, it said when the chat history is disabled, OpenAI will retain new conversations for 30 days and review them only when needed to monitor for abuse, before deleting them permanently. These controls, which are rolling out to all users starting today, can be found in ChatGPT’s settings and can be changed at any time. We hope this provides an easier way to manage your data than our existing opt-out process. This new update comes against the backdrop of recent incidents of employees leaking sensitive information on ChatGPT. For instance, a Samsung employee leaked critical information while using ChatGPT to correct errors in their source code. The company even cautioned its employees against using the chatbot. Read: ChatGPT Has Its Eyes on Your Data Besides Samsung, many organisations in the last few weeks have instructed employees to strictly not use ChatGPT due to risk confined to intentional leaks or cyber breaches, that could stem from the usage of such tools. OpenAI said that it is also working on a new ChatGPT Business subscription for professionals who need more control over their data as well as enterprise looking to manage their end users. “ChatGPT Business will follow our API’s data usage policies,” said the company, stating that end users’ data will not be used to train their model by default. Lastly, OpenAI has also released a new export option in settings, where users get to export their ChatGPT data and understand what information ChatGPT can store. “You’ll receive a file with your conversations and all other relevant data in email.”","excerpt":"This new update comes against the backdrop of recent incidents of employees leaking sensitive information on ChatGPT.","categories":["AI News"],"tags":[],"author_name":"Tasmia Ansari","publish_date":"2023-04-26T00:17:37","publication_year":"2023","word_count":327,"keywords":["ChatGPT","API","OpenAI","AI","programming_languages:R","GPT","GAN","R","llm_models:ChatGPT","llm_models:GPT"],"extracted_tech_keywords":["AI","ChatGPT","OpenAI","R","API","GPT","GAN","llm_models:GPT","llm_models:ChatGPT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/openai-spoils-users-with-more-data-control-for-chatgpt\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":36519,"title":"And The Bad News Spree Continues For Facebook With The Recent NZ Shooting","content":"Since quite some time, Facebook has been caught up in a wave of litigations. Last year, the social media giant disclosed that an unprecedented security issue impacted almost 50 million user accounts. It was nothing less than a shock to all its users. Numerous questions were thrown at Facebook regarding the platforms security strength. Even though Facebook managed to come out of the trouble after a lot of hustle and bustle, it seems, the dark clouds are still hovering over the Menlo Park giant. Recently, on March 8 a proposal from US senator and presidential candidate Elizabeth Warren almost broke up the company. That is not all, to make the situation worse for Facebook CEO Mark Zuckerberg, the company also witnessed the longest-ever outage of its social network and services. Twitter exploded out with memes and troll posts about the outage. Facebook said the outage was caused because of a shift in the setup of its computer servers. “We are very sorry for the inconvenience and we appreciate everyone’s patience,” the company said. But the worst was yet to come. On 15 March 2019, at least 49 people were killed in a mass shootings at two mosques in the New Zealand city of Christchurch. And to the shock, the footage of the massacre was streamed live online on Facebook for almost 17 minutes. The Rise Of Negativity For Facebook According to a source, the bad news spree for Facebook has reached a whole new level; the platform has never witnessed such a fall on shares in more than two months —  falling 2.5 per cent to close at $165.98. Furthermore, according to Facebook, the platform has managed to take down and prevent 1.2 million uploads of the video in the first 24 hours. However, 300,000 versions made it to the platform before being removed. It is not only Facebook that was struggling to take the video off from the platform, but Twitter and YouTube also struggled a lot. Twitter removed the original video and suspended the account that posted it, however, the real trouble was to remove copies that have been downloaded from the Facebook live and posted from other accounts. “We are deeply saddened by the shootings in Christchurch today,” a Twitter spokesperson said. “Twitter has rigorous processes and a dedicated team in place for managing exigent and emergency situations such as this. We also cooperate with law enforcement to facilitate their investigations as required.” This mishap has completely blown Facebook — the social media giant was witnessing a shower of negative comments —tweets on Twitter reached a whole new level. The complaints are skyrocketing that Facebook has done a poor job of safeguarding data or protecting users from the spread of hate speech, disinformation and live footage of violent events. Outlook Over the last few years, social media has grown to become a powerful force for people to voice their opinions and is being now used as a tool to hack data for driving political campaigns and broadcast divisive agendas. And the live-streamed mosque massacre is one such example. The video was streamed for almost 17 minutes and Facebook couldn’t stop it. Being a social media giant, the job of Facebook is not to only about connecting people but also about stepping up for the wrong. This might sound a bit silly, but why couldn’t Facebook inform the Police or the Governing body about what is going to happen. Or the platform could even share the attacker’s location. There are many policies and protocols that the company needs to follow, but are they bigger than doing that can stop a mishap? Let’s step into the shoes of a common man and think for a moment, wouldn’t you expect a strong step from a company that rules social media?","excerpt":"Since quite some time, Facebook has been caught up in a wave of litigations. Last year, the social media giant disclosed that an unprecedented security issue impacted almost 50 million user accounts. It was nothing less than a shock to all its users. Numerous questions were thrown at Facebook regarding the platforms security strength. Even though Facebook […]","categories":["AI Features"],"tags":["Facebook","facebook controversy","Mark Zuckerberg","political campaigns and ai"],"author_name":"Harshajit Sarmah","publish_date":"2019-03-19T05:05:33","publication_year":"2019","word_count":630,"keywords":["political campaigns and ai","Go","facebook controversy","AI","programming_languages:R","programming_languages:Go","Mark Zuckerberg","ViT","Facebook","R"],"extracted_tech_keywords":["AI","R","Go","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/and-the-bad-news-spree-continues-for-facebook-with-the-recent-nz-shooting\/","complexity_score":3,"technical_depth":6,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":2338,"title":"Biggest Analytics News in 2012","content":"Its time to look back and list down the biggest analytics events in 2012 for India. 2012 was a busy year for cross border acquisitions in analytics space. And the deals were both sided i.e. Indian companies acquiring foreign firms and companies outside India investing in Indian analytics firms. Some analytics deals where Indian firms bought foreign companies: – In May, Piramal acquired US-based DRG for $635 million. The Burlington, Massachusetts based Decision Resources Group (DRG) provides proprietary research on healthcare trends, clinical research and mergers & acquisitions in the biopharma industry, competing with other analytics firms such as Forrester and Elsevier. Going a step further, Piramal’s subsidiary Decision Resources Group (DRG) acquired UK-based market analytics and market access solutions firm Abacus International. Indian IT vendor MphasiS bought United States-based data analytics company, Digital Risk, for US$200 million in an all-cash deal. Wipro acquired Promax Applications Group (PAG), an Australia based Analytics Company, for 35 million Australian dollars (about Rs 191 crore). Crisil Ltd acquired UK-based analytics firm Coalition Development Ltd , along with its subsidiaries, for about £29 million (Rs. 250 crore). Coalition will become a part of the Crisil’s Global Research and Analytics business. Some analytics deals where foreign firms bought Indian companies: – US-based consumer credit reporting agency, Equifax acquired 51% stake in Bangalore-based startup NettPositive for an undisclosed sum. Global credit rating agency Moody’s Corporation was seen interested, along with outsourcing firm Genpact, in bidding for investment research and analytics company Amba Research. Equity infusion\/ Capital raising was also seen happening for various Indian analytics firms: – Global venture fund Sequoia Capital led a Rs 50-crore investment in data analytics startup, iCreate Software. This is the second round of equity funding for the Bangalore-based venture that earlier raised funding of Rs 15 crore from technology investor IDG Ventures. Affine Analytics secured angel investment through Bangalore’s Financial Consulting firm FinAKS Advisory. Manthan Systems secured $15m in funding from Norwest Venture Partners, in Feb 2012. iKen Solutions Pvt Ltd, an IIT Bombay research spin-off, received an undisclosed amount in funding from India Innovation Fund (IIF). Fidelity Growth Partners India commits to a $20 million investment in AbsolutData There was lot of action seen on the education front for analytics space. IIM Lucknow and KSB Indiana University USA jointly launched CPBAE programme. This one year Certificate Programme in Business Analytics for Executives (CPBAE) will provide cutting edge analytics knowledge. A business analytics lab would be soon set up at IIM Ranchi following a MoU between the institute and IBM Corporation. IMT Ghaziabad and Genpact Limited signed a MoU to develop and implement analytics elective for the two year PGDM program. Other news: – Data Security Council of India (DSCI) a central body that helps in promoting data protection in India, developing security and privacy best practices & standards, has chosen Ugam Solutions as the winner of the “DSCI Excellence Award for Security in IT Services – SME”. eClerx won the 2011’s Indian Most Admired Knowledge Enterprises (MAKE) award for creating a learning organization, in Feb 2012. Analytics India Magazine was launched in May 2012. :)","excerpt":"Its time to look back and list down the biggest analytics events in 2012 for India. 2012 was a busy year for cross border acquisitions in analytics space. And the deals were both sided i.e. Indian companies acquiring foreign firms and companies outside India investing in Indian analytics firms. Some analytics deals where Indian firms […]","categories":["IT Services"],"tags":["Absolutdata"],"author_name":"Дарья","publish_date":"2012-12-19T13:28:35","publication_year":"2012","word_count":516,"keywords":["Go","API","funding","AI","innovation","Git","analytics","GAN","R","Absolutdata","startup"],"extracted_tech_keywords":["AI","analytics","R","Go","Git","API","GAN","innovation","startup","funding"],"url":"https:\/\/analyticsindiamag.com\/it-services\/biggest-analytics-news-in-2012\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":39911,"title":"Data Extraction Just Got Smarter With ML: AWS Announces Textract","content":"Amazon Web Services, the cloud computing arm of the e-commerce giant, recently launched an ML service for automated text and data extraction. The service, known as Textract, is fully cloud-hosted and managed by AWS, and allows users to parse various forms of data easily. The service is said to be more than just an optical character recognition algorithm, as it can parse data tables, whole pages, forms, scans, PDFs, photos and more. Moreover, it also identifies fields and tables, so as to contextualize the data and allow for the collection of cleaner datasets with deeper insights. The company states that it can process millions of document pages “accurately” in just a few hours. All the data is exported to a JSON format, and can integrate easily with other ML-based AWS services. What sets this product apart is that there is no need to maintain any code or template, and that there is no ML experience required to operate or manage the product. Amazon states that they have trained Textract on “tens of millions of documents from virtually every industry”, making it suitable for use in any scenario. It can “automatically detect a document’s layout”, preserving the key elements in the page and perform optimal data collection by understanding the relationships between the data. Amazon is billing it as a lower-cost alternative to manual data entry, with an ease-of-use benefits. Moreover, as with every cloud computing service, it is provided on a pay-as-you-go basis, with accessible APIs. Swami Sivasubramanian, Vice President, Amazon Machine Learning, stated: “Amazon Textract makes it possible for customers to gain real meaning from their file collections, operate more efficiently, improve security compliance, automate data entry, and facilitate faster business decisions.” Currently, the service is available in US East (Ohio), US East (N. Virginia), US West (Oregon), EU (Ireland), with Amazon stating that further expansion will happen within the year. Many prominent companies have already begun using the service, such as The Globe and Mail, a Canadian media outlet, Met Office, the UK’s national weather service  and PriceWaterhouseCoopers, one of the world’s biggest accounting firms. The rise of accessible data entry ML models might be the beginning of the end for low-level jobs such as data entry.","excerpt":"Amazon Web Services, the cloud computing arm of the e-commerce giant, recently launched an ML service for automated text and data extraction. The service, known as Textract, is fully cloud-hosted and managed by AWS, and allows users to parse various forms of data easily. The service is said to be more than just an optical […]","categories":["AI News"],"tags":["Amazon","AWS","ML"],"author_name":"Anirudh VK","publish_date":"2019-05-30T09:07:24","publication_year":"2019","word_count":369,"keywords":["Go","API","machine learning","AWS","AI","cloud computing","cloud_platforms:AWS","ML","Amazon","cloud_platforms:Amazon Web Services","R"],"extracted_tech_keywords":["AI","machine learning","ML","cloud computing","AWS","R","Go","API","cloud_platforms:AWS","cloud_platforms:Amazon Web Services"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/data-extraction-just-got-smarter-with-ml-aws-announces-textract\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10090961,"title":"Council Post: Why Responsible AI in Computer Vision is Imperative","content":"In the 1960s, early academics advocated artificial intelligence as a technology that could change the world and were incredibly hopeful about the future of these connected sciences. Years later, Grand View Research reported that, in 2020, the worldwide market for computer vision was worth $11.32 billion and projected to grow at a rate of 7.3% annually from 2021 to 2028. While AI application has become pervasive across industries, the next decade will see rapid advancement of AI-driven computer vision technology. With the expansion of its applications and use-cases, it will bring to fore the need for governance, responsible use and ethical standards. Computer vision is a field of artificial intelligence that enables computers and systems to derive meaningful information from digital images, videos and other visual inputs. Until recently, computer vision only worked in limited capacity but the field has been able to take great leaps in recent years, surpassing humans in some tasks related to detecting and labelling objects. With advancements in AI innovation in disciplines of deep learning and neural networks, the scope will only grow further. But, as with all technologies, computer vision also comes with its own set of challenges. Let’s delve into some areas where computer vision still needs work. Addressing bias in computer vision Computer vision is increasingly being used in facial recognition, object detection, and autonomous driving. It is crucial that this technology is fair and unbiased so that it can be used in a way that benefits everyone in society. However, a major challenge in achieving fairness and eliminating bias is the fact that these systems are trained on large datasets, which often reflect the biases and prejudices of the society in which they were created. If a dataset used to train a facial recognition system is mostly composed of images of a certain ethnicity of males, the system may perform poorly when attempting to recognise faces of people from other racial or gender groups. For instance, according to a journal in Lancet Digital Health, AI systems designed to diagnose skin cancer were shown to be less reliable as in the datasets used to train the model, there were no photographs of people with an African, African-Caribbean, or South Asian heritage. To address this challenge, it is important to develop more diverse and inclusive datasets that reflect the full diversity of human experience. Additionally, algorithms should be designed to actively detect and correct biases in the data and in the models themselves. Privacy risks of AI-powered computer vision Government and public sector applications are rapidly utilising artificial intelligence and computer vision due to their numerous advantages. For instance, in order to monitor traffic, road conditions, and public events, governments deploy computer vision technologies to create smart cities. Real-time visual data that is acquired by cameras is gathered and processed by computer vision systems. The video feed is initially recorded and delivered for data processing and analysis to a system on-site, edge devices, or a cloud-based storage system. Following this, computer vision applications process the raw data using deep learning models to carry out tasks such as human detection, object detection, and counting. Additionally, data sent to the cloud is frequently kept there for at least a while in accordance with data retention policies or legal requirements. The collection, storage and handling of such volumes of identifiable data presents a huge risk of data breaches or misuse and therefore invasion of people’s privacy. Most businesses have rules governing who has access to the data and what they may do with it. These rules may allow third-party cloud service providers or vendors to access, share, or sell such data for marketing purposes, further raising the risk for data privacy misuses. Environmental implications of AI-powered computer vision Processes involved in training and using computer vision systems can have negative environmental impact. Deep learning model training consumes a substantial amount of processing power which, in turn, consumes a lot of energy. These models require enormous quantities of electricity that is often produced using fossil fuels in the data centres that house them. Powerful hardware, such as GPUs, is needed for computer vision systems, albeit with a short lifespan and in need of frequent upgrades. Owing to such setups, there is a significant volume of electronic waste that becomes challenging to recycle. Massive volumes of data are produced by computer vision applications and this data must be saved for further processing and analysis. This necessitates a large quantity of storage space, which uses more energy and may result in greenhouse gas emissions. Responsible build and use of computer vision While computer vision applications are in their nascent stage, they still have a long way to go. Organisations and people must take the initiative in utilising computer vision and facial recognition in an ethical and responsible manner until governmental bodies are able to effectively regulate these developing technologies. A fundamental key is to build responsibly and only with the goal to serve the purpose. Putting this into perspective: If the requirement is to only count footfalls, computer vision should be deployed to only count and not for facial recognition simply because the technology allows it. Similarly, those with access to the data should uphold ethical practices and engage in responsible handling. Only then can we reap the full benefits of this technology without endangering people’s privacy and safety. This article is written by a member of the AIM Leaders Council. AIM Leaders Council is an invitation-only forum of senior executives in the Data Science and Analytics industry. To check if you are eligible for a membership, please fill out the form here.","excerpt":"Organisations and people must take the initiative in utilising computer vision and facial recognition in an ethical and responsible manner until governmental bodies are able to effectively regulate these developing technologies. A fundamental key is to build responsibly and only with the goal to serve the purpose.","categories":["AI Features"],"tags":["bias","Computer Vision","Responsible AI"],"author_name":"Swaroop Shivaram","publish_date":"2023-04-06T17:30:00","publication_year":"2023","word_count":934,"keywords":["data science","artificial intelligence","AI","neural network","Responsible AI","computer vision","RAG","Aim","deep learning","object detection","analytics","Computer Vision","bias"],"extracted_tech_keywords":["AI","artificial intelligence","deep learning","neural network","computer vision","data science","analytics","Aim","RAG","object detection"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/why-responsible-ai-in-computer-vision-is-imperative\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10072848,"title":"Deepmind Launches SOTA Video Generation Framework, ‘Transframer’","content":"Recently, Deepmind researchers announced the launch of Transframer—a new general-purpose framework for image modelling and vision tasks based on probabilistic frame prediction. This new model unifies a broad range of tasks, including image segmentation, view synthesis and video interpolation. This latest framework uses U-Net and Transformer components to condition on annotated context frames, and outputs sequences of sparse, compressed image features. Transframer is a general-purpose generative framework that can handle many image and video tasks in a probabilistic setting. New work shows it excels in video prediction and view synthesis, and can generate 30s videos from a single image: https:\/\/t.co\/wX3nrrYEEa 1\/ pic.twitter.com\/gQk6f9nZyg— Google DeepMind (@GoogleDeepMind) August 15, 2022 What does Transframer do Developed by Deepmind, Transframer unifies a range of image modelling and vision tasks and has the ability to create videos or image features based on a single image with one or more context frames. Transframer works on a variety of video generation benchmarks. The research team claims that it is a state-of-the-art model which is expected to be the strongest and most competitive on few-shot view synthesis, and can generate coherent 30-second videos from a single image. The proposed model also showed promising results on eight tasks in total, some of which are semantic segmentation, image classification, and optical flow prediction with no task-specific architectural components. Transframer can also be used in various applications that require learning conditional structure using text or a single image, and will be able to predict and generate video models, novel view synthesis and multi-task vision. Backed by Google, Deepmind has been researching in the field of AI since 2010 and focusing on building computer models that can solve building and generative problems on their own. Click here to read the research paper.","excerpt":"The framework uses U-net and Transformer components to condition on annotated context frames, and generate a sequence of spare, compressed image features.","categories":["AI News"],"tags":["AI Video Generation Models","DeepMind","Google"],"author_name":"Mohit Pandey","publish_date":"2022-08-16T18:50:59","publication_year":"2022","word_count":290,"keywords":["Go","TPU","programming_languages:R","AI","programming_languages:Go","AI Video Generation Models","Aim","Google","R","DeepMind"],"extracted_tech_keywords":["AI","Aim","TPU","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/deepmind-launches-sota-video-generation-framework-transframer\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10094977,"title":"Salesforce Launches AI Cloud with Einstein Trust Layer","content":"Salesforce has announced the launch of AI Cloud, a suite of products built for CRM in enterprises where they can boost their productivity through all the Salesforce AI applications. Set to launch this year, the AI Cloud platform offers hosting and delivery services for various AI models, particularly those focussed on generating text. It collaborates with multiple partners such as Amazon Web Services, Anthropic, Cohere, and OpenAI, utilising Salesforce’s cloud infrastructure. Marc Benioff, co-founder and CEO of Salesforce, said, “Generative AI may be the most important technology of any lifetime,” during the announcement in New York City. In addition to the partner models, customers can access first-party models developed by Salesforce’s AI research division, which enable features like code generation and business process automation. Alternatively, customers have the option to bring their own custom-trained model to the platform while keeping their data stored on their own infrastructure. All the AI models Salesforce’s flagship products, including Data Cloud, Tableau, Flow, and MuleSoft, benefit from the advanced capabilities provided by the AI models built within AI Cloud. These models include Sales GPT, Service GPT, Marketing GPT, Commerce GPT, Slack GPT, Tableau GPT, Flow GPT, and Apex GPT, making a total of nine models. However, one notable omission in AI Cloud is an image-generation model similar to DALL-E 2 and Stable Diffusion. Salesforce is actively working on developing such a model. They aim to overcome various barriers, including copyright and toxicity concerns, before releasing it to the public. Sales GPT enables the quick creation of personalised emails, while Service GPT assists in generating service briefings, case summaries, and work orders based on case data and customer history. Marketing GPT and Commerce GPT generate audience segments for targeted marketing and tailor product descriptions to individual buyers using their customer data, providing recommendations on increasing average order value. The remaining models, Slack GPT, Tableau GPT, Flow GPT, and Apex GPT, have more specialised applications. Slack GPT and Flow GPT allow users to build no-code workflows that incorporate AI actions within Slack or Flow. Tableau GPT generates visualisations based on natural language queries and reveals data insights. Apex GPT scans for code vulnerabilities and suggests inline code for Salesforce’s proprietary programming language, Apex. Currently, several models are already operational, such as Slack GPT, Commerce GPT, Sales GPT, and Service GPT. The rest, excluding Flow GPT, which will be available in October, are scheduled to be launched as early as this month. This is a very similar move to Amazon launching Bedrock, which provides a family of models trained by AWS in-house, along with pre-trained models for startup partners. Recently, Salesforce also partnered with Google Cloud to expand strategic partnership for helping businesses utilise AI and data for delivering more personalised experiences for their customers. Einstein GPT Trust Layer Similar to NeMo Guardrails, for ensuring the privacy and security of these models, Salesforce is making an attempt to ensure the moderation of the content by retaining the privacy of user’s sensitive data. AI Cloud includes a brand new Einstein Trust Layer, which makes it open, extensible, while providing privacy and security. This will allow customers to use third-party LLMs, Bring Your Own Model (BYOM) on the Salesforce platform. The Einstein GPT Trust Layer will be available from this month generally. Moreover, to distinguish AI Cloud from other managed AI service providers, Salesforce is introducing a range of prompt “templates” and tools for building prompt templates. These templates, according to Salesforce, utilise “optimised” AI prompts that leverage “harmonised” data to ensure that the generated content aligns with a company’s specific requirements. By grounding the model-generated outputs in context, these prompts significantly enhance the quality and relevance of the generated content.","excerpt":"AI Cloud includes a brand new Einstein Trust Layer, which makes it open, extensible, while providing privacy and security.","categories":["AI News"],"tags":["Salesforce"],"author_name":"Mohit Pandey","publish_date":"2023-06-13T10:06:37","publication_year":"2023","word_count":612,"keywords":["Anthropic","Go","TPU","OpenAI","AI","AWS","RAG","Aim","generative AI","Salesforce","R"],"extracted_tech_keywords":["AI","generative AI","OpenAI","Anthropic","Aim","RAG","AWS","TPU","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/salesforce-launches-ai-cloud-with-einstein-trust-layer\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":29326,"title":"Supercharged Analytics Platform Teradata Vantage Looks To Break Down Data Silos In Enterprises","content":"Teradata CTO Stephen Brobst speaking at the Teradata Analytics Universe. Government departments all over the world, especially in India have a huge amount of scattered data. The government, like any other large operation in the technology sector, needs a platform which can deliver a common data strategy and drive analytics to serve their customers and users better. With this in mind, the Rajasthan government has partnered with Teradata to create a common data and analytics platform for all government departments across the state. This project aims to build services that collate many databases and deliver a 360-degree citizens’ view and solve their grievances. Eligibility of beneficiaries for schemes, easy tracking of beneficiaries, fraud detection, enhanced citizen engagement and sentiment analytics of citizens, are some of the services. Break Down Silos: The Vantage Strategy The thought behind Teradata Vantage has been to bring down data silos in large organisations and provide a single comprehensive data strategy. Teradata Vantage promises to bring multiple data sources on one platform and also bring many technologies under one umbrella by giving a plug and play service. Talking to Analytics India Magazine, Chad Meley, Vice President of AI and Analytical Solutions at Teradata said, “Our customers certainly have valued the core Teradata technology very highly. Their feedback has been that there were many data silos. With Teradata Vantage they would have the power to collect the data at one place and open it up with modern architectures like API and microservices.” One of the prime examples of Teradata’s customer putting faith in the new platform is Siemens Healthineers which has around 6,00,000 installed products all around the world and is responsible for more than 70% of critical clinical decisions. They use Teradata Vantage to power clinical decision making which leverages around 240,000 patient touch points, including rich sensor data. Teradata SVP Chris Twogood speaking at the Teradata Analytics Universe. Teradata Vantage may soon make waves in the Indian enterprise ecosystem. Yasmeen Ahmad, director of Customer Excellence for Teradata told Analytics India Magazine said that the Indian market presents a lot of opportunities to marketers to target customers and gain massive improvements in the bottomline. She also talked about the trends of geospatial analytics being adopted on a big scale in India. The Customer Experience John Carpenter, VP at SiriumXM said that since the adoption of Teradata Vantage their company has seen the time for certain tasks to go down drastically. Speaking during the general session at Teradata Analytics Universe he highlighted that Teradata has reduced time to get their services to the market while lowering their costs at a large scale. Interacting with Analytics India Magazine on the sidelines of Teradata Analytics Universe, Stephen Brobst who heads technology at Teradata talked about their Indian strategy. He said, “The Indian market has lots of data so we are the best choice because we handle large data better than anyone. Our goal is to bring Indian customers into advanced analytics because they are overwhelmed with so much data.” Conclusion Teradata has also put a singular focus on building a platform that can work with any kind of data. The Vantage platform works with multiple types of data, formats and heterogeneous data stores further bringing data together under a single platform. Reema Potdar, Executive Vice President in charge of Product Development for Teradata says, “Vantage is a product built for everyone but can adapt to many of the Indian use cases. I believe it can go in a long way to improve operational efficiencies in Indian hospitals. Teradata is being used by many large companies who have their innovation offices set up in India and hence our products do have a global effect through our India based clients” Talking about the difference between the Indian and Western analytics markets, Ahmad said, “Western countries like the UK are already a huge market for voice assistants whereas India still hasn’t reached there.” This points out the fact that Indian enterprises would not be able to market their solutions through voice assistants just yet and would have to wait for the voice market to evolve in India. Teradata Vantage is a departure from closed analytical products that have clogged the market but would have strong competition from niche and open source tools and products data scientists and business executives have grown used to over the years. Disclaimer: The writer is in Las Vegas on the invite of Teradata India","excerpt":"Government departments all over the world, especially in India have a huge amount of scattered data. The government, like any other large operation in the technology sector, needs a platform which can deliver a common data strategy and drive analytics to serve their customers and users better. With this in mind, the Rajasthan government has […]","categories":["Deep Tech"],"tags":["Data Scientists"],"author_name":"Abhijeet Katte","publish_date":"2018-10-17T05:24:45","publication_year":"2018","word_count":737,"keywords":["Go","API","AI","R","ML","RAG","microservices","Aim","analytics","fraud detection","Data Scientists"],"extracted_tech_keywords":["AI","ML","analytics","Aim","RAG","fraud detection","microservices","R","Go","API"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/supercharged-analytics-platform-teradata-vantage-looks-to-break-down-data-silos-in-enterprises\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10075369,"title":"AI startup Sigmoid raises $12 million from Sequoia Capital","content":"AI solutions provider Sigmoid has raised $12 million funding from Sequoia Capital India today. With the current round being a mix of primary and secondary funding, Sequoia has invested $19.3 million in Sigmoid till date. The startup will use the capital to evolve its market offerings, expand delivery centers and cater to new industries. “The last 12 months have been an inflection point in our growth. We are happy to see Sequoia Capital India continue to believe in us. This will help us rapidly expand our capabilities in terms of solutions and talent to meet the growing customer demand,” said Lokesh Anand, chief executive officer at Sigmoid. Established in 2013—IIT Kharagpur alumni Lokesh Anand, Mayur Rustagi and Rahul Kumar Singh found Sigmoid to provide data engineering and AI consulting for enterprises to help them gain competitive advantage through effective data-driven decision making. According to Anandamoy Roychowdhary, principal, Sequoia Southeast Asia, “The AI and machine learning (ML) market is growing year-on-year and so is the need for solutions to help enterprises adopt and harness this power. The team at Sigmoid, which stands out for its data and AI engineering excellence, is well-positioned to capture this opportunity.” The team currently comprises over 500 data professionals, with over 80% of the workforce located in India. Majority of the revenues come from US clients, who are provided with expertise in data engineering, cloud data modernisation, and data operations.","excerpt":"The team currently comprises over 500 data professionals with over 80% of the workforce being in India","categories":["AI News"],"tags":["AI Startups"],"author_name":"Bhuvana Kamath","publish_date":"2022-09-15T19:18:16","publication_year":"2022","word_count":234,"keywords":["API","machine learning","AI","data-driven","ML","RAG","data engineering","Rust","R","AI Startups","startup"],"extracted_tech_keywords":["AI","machine learning","ML","RAG","R","Rust","API","data engineering","data-driven","startup"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/ai-startup-sigmoid-raises-12-million-from-sequoia-capital\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":9348,"title":"Engineering Analytics : As Predictable As It Can Get","content":"Exploding data in the engineering space has made Engineering Analytics an imperative across various industries. Engineering Analytics is likely to benefit companies to the tune of $250 billion and this is expected to double to $500 billion by 2017. The current spend on Engineering Analytics product engineering, analytics and system integration is close to $13 billion. Services that enable the planning and deployment of engineering analytic solutions go by as Engineering Analytics services and from a $5.4 billion, it is expected to triple by 2017 to $14.8 billion. Engineering Analytics are being adopted by OEMs and suppliers in three key areas of their operations: design, manufacture and after-sales support. In design, all OEM’s have huge clusters of high power computing, which are used in the design process and generate terabytes of data on each simulation on their products. There is a need to visualize and tease out design patterns from this data. This will decrease development time and improves the quality and performance of their products. As manufacturing systems increasingly network and communicate with each other in an Industrial Internet of Things (IIoT) environment, automatic measurement and quality control at factory level become a reality. Terabytes of manufacturing data produced will have to be analyzed and leveraged for decision making. The next generation of aircraft will include engines that are permanently connected to a data centre allowing engineers to analyze and monitor the fleets and help diagnose faults, correcting them and preventing them from occurring again. Below are some examples of Engineering Analytics services mapped to various industries: Industries Services Aerospace & Aviation Predictive analytics Rail Transportation Descriptive Analytics Heavy equipment Diagnostic analytics Medical devices & Electronics Data visualization Oil and Gas Industrial Internet Internet of things Machine to Machine Aero Engines Engine Health Monitoring Engine Diagnostic Assessment Component Life Assessment Trend Monitoring Data driven Engine Design Engineering Analytics are found to be more pervasive in the Energy industry with a spend of $1.4 billion as well as have the largest addressable market for Engineering Analytics service providers. Some of the key Engineering Analytics services that can be outsourced include: data cleaning and integration data visualization predictive modelling and decision optimization In the energy industry, predictive modelling and analytics will help in the development of next-gen system planning & operations tools and controls to understand system dynamics, earmark areas prone to fallibility, prevent and mitigate subsequent fall-out events, and support analyses after the incident. The instability of renewable energy sources owing to factors such as wind speed and solar availability restrict their usage to just about 60-80%. Accurate energy production forecasts of such renewable energy power plants are possible through predictive modelling and analytics. By predicting equipment failure and downtime, it is therefore possible to optimize overall operational efficiencies. Such diagnostic analytics also improve the life of these valuable renewable energy equipment. In the Oil and Gas industry, predictive modelling and decision optimization is useful across upstream, midstream and downstream operations. In Upstream, predictive modelling enables asset maintenance, optimization in drilling\/exploration\/production optimization, risk assessment, etc. Outcomes of predictive modelling and analytics include better planning of maintenance activity, using insights from intelligent analysis of data and then feeding that back into design and reduced service disruption. All these translate to huge cost savings and reduction and improved operational efficiency.","excerpt":"Exploding data in the engineering space has made Engineering Analytics an imperative across various industries. Engineering Analytics is likely to benefit companies to the tune of $250 billion and this is expected to double to $500 billion by 2017. The current spend on Engineering Analytics product engineering, analytics and system integration is close to $13 […]","categories":["IT Services"],"tags":[],"author_name":"Dr T C Ramesh","publish_date":"2016-03-17T05:45:06","publication_year":"2016","word_count":548,"keywords":["Go","programming_languages:R","AI","predictive analytics","programming_languages:Go","RAG","ViT","analytics","disruption","R"],"extracted_tech_keywords":["AI","analytics","RAG","predictive analytics","R","Go","ViT","disruption","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/decision-making-predictable-can-get\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10140090,"title":"Stop Paying for GPT-4o—This YC Startup Offers 4x the Savings","content":"Floworks, the cloud-based enterprise automation startup, recently released a novel ThorV2 architecture allowing LLMs to perform functions with better accuracy and reliability. The YC-backed company collaborated with IIT Bombay and Kharagpur to build this architecture. In an interview with AIM earlier this year, Floworks claimed that its AI agent, Alisha, is 100% reliable for tasks involving API calls. Sudipta Biswas, the co-founder of Floworks, said, “Our model, which we are internally calling ThorV2, is the most accurate and the most reliable model out there in the world when it comes to using external tools right now.” Further, he claimed that ThorV2 was 36% more accurate than OpenAI’s GPT-4o, 4x cheaper, and almost 30% faster in terms of latency. These sensational claims were recently backed by an in-depth research paper that dives deep into ThorV2 architecture and how all of its novel features work to solve several crucial challenges in agentic workflows inside some of the market-leading LLMs today. Edge of Domain Modeling is ThorV2’s Hero Technique The edge of domain modelling, used in ThorV2 architecture, involves providing minimal instruction upfront, allowing the agent to begin the task, and then providing the remaining information through error corrections post-task. This approach differs from providing knowledge of all possible scenarios regarding the function calling. Edge of domain modelling reduces the need for extensive instructions, which in turn reduces the number of tokens in the prompt. It can further lead to cost saving measures. The authors mentioned that “Function schemas can be lengthy, leading to large prompt sizes. This increases deployment costs, time consumption, and can result in decreased accuracy on reasoning tasks.” Additional instructions added to the LLM in the error correction process are performed by a static agent implemented through an Agent Validator Architecture inside ThorV2. You Don’t Need an LLM to Evaluate Another LLM The Agent Validator architecture overcomes several limitations of agentic workflows, where the primary LLM agent performing a task receives feedback from other LLMs that act as critics. The authors argue that using an additional LLM not just increases the deployment costs, but decreases the rates of accuracy. The introduction of a static agent written fully in code, includes a component called Domain Expert Validator (DEV) which inspects the output generated by the LLM for errors. The DEV contains all the knowledge and information required to perform the function calls inside a specific platform. While building a validator requires a significant amount of effort, it helps reduce processing time and improve accuracy. This is because the knowledge contained inside the DEV contains information regarding the most common and repetitive errors that occur in the function calling process. Multiple API Functions in a Single-Step One of ThorV2’s other advantages is that it can generate multiple API calls in a single step. With ThorV2, a single query is sufficient for both tasks, even if the first task needs to retrieve information from the API for use in the second task. The approach involves using a placeholder to represent unknown values, and once the first task retrieves the API response, the value is then injected into the second task.“Generating multiple API calls at once requires sophisticated planning and reasoning capabilities, which is very challenging for ordinary LLMs. Our Agent-Validator architecture simplifies this process as well by correcting errors in the planning step”, added the researchers. This approach is a significant improvement over the traditional, sequential handling of API calls in current LLMs, which often require a step-by-step execution process. And the Numbers Don’t Lie – 50% Cost Reduction With 100% Reliability The ThorV2 architecture was compared to OpenAI’s GPT 4o, GPT 4 Turbo, and Claude 3 Opus for a set of operations on HubSpot’s CRM. The authors developed a dataset called HubBench on which the model was evaluated. The models were tested for accuracy, reliability, speed, and cost. In a conversation with AIM, Sudipta mentioned that ThorV2 was connected to the Llama 3 70B model for comparison. ThorV2 came out on top in every single test, and a 100% score in the reliability test, which seeks a consistent output when the model is put out to perform the task ten times. In the single API call function, ThorV2 scored 90% accuracy, second to Claude 3 Opus’ 78% score. The test also revealed that it only took $1.6 for a thousand queries, which is 3 times cheaper than OpenAI’s models. Even with multiple API calls, ThorV2 performed better on every single metric. While reading the comparison benchmark scores, one wonders if these scores are relevant five months after the tests were conducted, with several new and capable models like Claude 3.5 Sonnet and GPT o1 having been launched. However, it is important to understand that ThorV2 is an architecture built to enhance the performance and capabilities of an existing LLM. The integration will, in fact, work better with new and more capable models. “We will soon come up with Thor v3, which will definitely compare with other models that have come up recently. But again, the framework is not a model-level innovation that we’re doing,” Sudipta said.  “So even if the underlying model keeps on getting better, our framework will keep supporting even better than that.” It Isn’t Perfect, But Floworks Wants to Get There One of ThorV2’s limitations is that it relies on knowledge from the DEV based on common, and well established error patterns and it may face difficulties approaching an unseen one. Moreover, the research currently tests ThorV2’s architecture for just single, and two API call functions. The authors acknowledge the limitations, and plan to perform a comparison with three or more function calls in future research. In the conversation with AIM, Sudipta revealed that ThorV3 is currently in the works, and it will challenge some of the latest market leading models today. That said, one can also expect other limitations to be resolved in the future iteration. A Vision to Solve More Real-World Problems The authors envision ThorV2 to overcome the limitations of existing LLMs and solve problems that can truly create an impact. They mentioned that LLMs have revolutionised NLP and AI, demonstrating remarkable capabilities across a wide range of tasks. However, their economic impact has been somewhat limited, particularly in domains requiring precise interaction with external tools and APIs. Over the last few months, we’ve also seen a meteoric rise in AI Agents and their tremendous capabilities, and frameworks like ThorV2 can only propel their powers further in sectors that require a large amount of automation and knowledge transfer between different applications. “LLMs seem very cool, but to front-load them with a high amount of tokens, the cost will be prohibitively, very high. For large-scale operations where lots of automation is needed to be done, that price point will not suit enterprises, and small businesses,” Sudipta said.","excerpt":"In collaboration with IIT Bombay and IIT Kharagpur, Floworks has released a research paper that dives deep into the sensational claims made by the YC-backed startup earlier this year.","categories":["AI Startups"],"tags":["AI (Artificial Intelligence)","AI Agents","API","Floworks"],"author_name":"Supreeth Koundinya","publish_date":"2024-11-04T14:00:00","publication_year":"2024","word_count":1131,"keywords":["API","TPU","OpenAI","AI","GPT o1","GPT-4o","AI Agents","agentic workflows","RAG","NLP","Aim","Claude 3.5","Floworks","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","NLP","GPT-4o","GPT o1","OpenAI","Claude 3.5","agentic workflows","Aim","RAG","TPU"],"url":"https:\/\/analyticsindiamag.com\/ai-startups\/stop-paying-for-gpt-4o-this-yc-startup-offers-4x-the-savings\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":26698,"title":"Data Security &#038; AI For All, Google Next &#8217;18 Brought Good News For Users, Small Businesses","content":"(Image source: Google) Google’s annual event Next ’18 was a thing to behold. Held at the Moscone Center, San Francisco, and hosted by CEO Sundar Pichai and Cloud top executive Diane Green, the event was witnessed by over 20,000 people. “Our mission is to organise world’s information and make it universally accessible and useful. I am always talking about being fortunate as a company as a timeless machine. One that feels as it did 20 years ago,” said Pichai during the address. Apart from new features in Gmail, functionalities in Cloud and other news, many key announcements around security, artificial intelligence, machine learning were made: AI for every business Google will be helping their users train their ML models faster, in multiple locations (on-prem, and multi-cloud with Kubeflow), and with more libraries (XGBoost and scikit-learn on Cloud ML Engine). They will also extend classification to IoT and gateway devices at the edge. From data to insights Data growth in the enterprise is staggering, and as businesses generate more data each year, they need advanced tools to store, manage, analyse, and generally make sense of it all. Google announced a number of enhancements to their data analytics offerings aimed at helping businesses uncover important insights from their data. Secure Cloud From context-aware access to shielded VMs and binary authorisation, Google announced a variety of security tools and capabilities, to secure data and the operating environment. AI for every business Google announced Cloud AutoML Vision, Natural Language, and Translation that extends powerful ML models to suit specific needs, without requiring any specialised knowledge in machine learning or coding. They also announced a new solution, Contact Center AI, which includes new Dialogflow features alongside other tools to assist live agents and perform analytics. Security A new investigation tool in the Security Center helps admins identify which users are potentially infected, see if anything’s been shared externally and remove access to Drive files or delete malicious emails. Here’s a full recap for Day 1: And for Day 2:","excerpt":"Google’s annual event Next ’18 was a thing to behold. Held at the Moscone Center, San Francisco, and hosted by CEO Sundar Pichai and Cloud top executive Diane Green, the event was witnessed by over 20,000 people. “Our mission is to organise world’s information and make it universally accessible and useful. I am always talking […]","categories":["AI News"],"tags":["Google Cloud"],"author_name":"Prajakta Hebbar","publish_date":"2018-07-26T12:57:36","publication_year":"2018","word_count":334,"keywords":["scikit-learn","artificial intelligence","machine learning","Google Cloud","AI","ML","Kubeflow","Aim","XGBoost","analytics","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","analytics","Kubeflow","Aim","scikit-learn","XGBoost","R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/data-security-ai-for-all-google-next-18-brought-good-news-for-users-small-businesses\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":1276,"title":"Intel takes on Retail industry by introducing IoT Retail Platform and investing $100 mn","content":"Vouching for the retail industry, Intel recently unveiled the new Intel® Responsive Retail Platform (Intel® RRP) designed exclusively to take retail to the next level. It intends to do so by taking retail to an era of highly efficient personalized shopping and intelligently lead business. This announcement supports Intel’s wider efforts to integrate IoT and other technologies into retail. It has also committed an investment of more than $100 million over the next five years in the retail industry, further boasting its intention of unifying every part of the retail operations and solving longstanding business challenges. As addressed by Intel CEO Brian Krzani in Retail’s BIG show, the new horizontal platform fosters the creation of highly flexible, scalable and innovative solutions, bringing retail hardware, software, APIs and sensors in a standardized way. As the retail IoT market is reportedly projected to reach $35.64 billion by 2020, the retailers are increasingly turning towards IoT solutions to drive operational efficiencies and create new and exciting customer experiences, both in the store and online. And rightly tapping onto this opportunity, the quicker and cost effective Intel Responsive Retail Platform can be used across all types of retail businesses to integrate islands of technologies across their stores and uncover new business insights. Key features of Intel® RRP: It enables real-time, automated actions for store employees, products and customers It has the ability to significantly reduce cost and time to deploy new store services Unified platform that eliminates islands of technology across brick and mortar and online stores Simplified on-boarding process for deploying and maintaining new sensors for the store. Through a robust ecosystem, the Intel RRP can quickly and efficiently be scaled to fit the needs of the retailer. Additionally, the standardization enables the rapid development and deployment of new IoT services from a multitude of suppliers. The IoT play in building responsive, customer-centric, smart and connected retail: Aiming to transform the retail industry to drive new customer experiences, deliver real-time insights and create a connected and unified environment, Intel is collaborating with a diverse ecosystem of retailers such as Avery Dennison, ByReveal, Honeywell, Fujitsu, JDA, RetailNext, SAP and SATO. The Intel Responsive Retail Platform which uses RFID, video, radio and other sensors would enable easy, holistic integration and deliver a 360-degree viewpoint of retail from the store floor through the supply chain, and deliver real-time, actionable insights. Designed to address the industry’s toughest challenges, Intel is developing a comprehensive end- to-end portfolio for the retail industry that incorporates the wide breadth of technology necessary for advanced computing, customer experience, resource deployment, communication and deep learning that has unmatched scalability and security spanning brick and mortar, cloud infrastructure, and data center. Intel is uniquely positioned to enable every segment of the smart and connected world – powering the majority of the world’s data centers, connecting hundreds of millions of IoT devices and fulfilling the promise of always-on, 5G connectivity, deep machine learning, and security and privacy. The integrated technologies that come with Intel Responsive Retail Platform: Intel based Retail sensors, Gateway and Cloud forms the crux of Intel RRP. The sensors enable data ingestion from numerous sources in the retail store in a simple cost effective manner. The Gateway finds, configures, controls and interacts with the Intel Responsive Retail Sensors over Ethernet without third-party middleware. Provided by HP* and based on Intel® CoreTM i7 or Intel® Xeon® processors, the gateway connects to the cloud via the internet using a wired network, wireless or 3G\/LTE cellular network. Whereas cloud application is facilitated as an open-source analytics platform-as-a-service (PaaS). This cloud-based, horizontal platform features an integrated stack that reduces complexity and gives greater control to businesses. Simbe Robotics’ Tally, powered by Intel: Also making a debut into the industry was Simbe Robotics’ Tally, the world’s first robotic autonomous shelf auditing and analytics solution for retail, which was demonstrated by Krzanich at Retail’s Big Show. Tally works in concert with retail associates by arming them with information to ensure the store’s products are always stocked, in the right place and displaying the correct price tag. The robot operates safely during normal store hours alongside shoppers and employees, and doesn’t require any infrastructure changes to the store. Tally is powered by an Intel® Core™ i7 processor-based Intel® NUC and uses a number of Intel® RealSense™ cameras to help understand the world around it and navigate the store safely.","excerpt":"Vouching for the retail industry, Intel recently unveiled the new Intel® Responsive Retail Platform (Intel® RRP) designed exclusively to take retail to the next level. It intends to do so by taking retail to an era of highly efficient personalized shopping and intelligently lead business. This announcement supports Intel’s wider efforts to integrate IoT and […]","categories":["AI News"],"tags":["Intel","IoT","Robotics"],"author_name":"Srishti Deoras","publish_date":"2017-01-21T06:05:04","publication_year":"2017","word_count":730,"keywords":["API","machine learning","AI","R","Scala","Robotics","Aim","deep learning","ViT","analytics","Intel","analytics platform","IoT"],"extracted_tech_keywords":["AI","machine learning","deep learning","analytics","Aim","R","Scala","API","ViT","analytics platform"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/intel-takes-retail-industry-introducing-iot-retail-platform-investing-100-mn\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10065106,"title":"The AI strategy of Housing.com","content":"Housing.com has been investing in AI and ML to help its users in every step of their home buying or home renting journey. “As an organisation, we believe in AI and its impact on consumer experience on the platform,” said Vipin Kumar Singh, head of technology, Housing.com, PropTiger.com and Makaan.com. In an exclusive interview with Analytics India Magazine, Vipin spoke about how the firm embeds ethics into its AI systems. AIM: How does Housing.com leverage AI? Vipin Kumar Singh: Here are some of the major use cases where we harness the power of AI: Boost Efficiency and Productivity with Digital Automation: For audits of property images uploaded by users. These capabilities help find competitors’ logos, blurred images, unwanted text and NSFW content.Automated Valuation Model: To predict the price of a property in resale and rent which is used in various product workflows to help sellers to list their properties at the right price and the property seekers to avoid paying too much.Audio Processing: To deal with audio data in terms of conversations between consumers and our pre-sales and sales teams.Personalised Customer Experience: The recommendation engine is used for both the search and discovery as well as self-serve package subscription for the paying customers.Text Processing and NLP: For keyword extraction from user-generated property descriptions.Data Insights: We process the data from various sources in near real-time via a well-architected Data Platform to provide business insight. This helps in competitive analysis, helps in determining customer churn and improves renewal rates.AI-Powered Chatbots: The customer interaction for property search and customer support is based on easily customised decision tree-based workflows.Marketing Efficiency Improvement: Multi-channel marketing can have higher ROI via marketing intelligence like Market Mix Modeling. Fintech Fraud Detection: AI helps in detecting the frauds with much better accuracy and the underlying ML models can evolve with changing patterns in fraud activities. AIM: Could you elaborate on Housing.com’s AI governance framework? Vipin Kumar Singh: From an approach POV, we use the CRISP-DM methodology to ensure each data science problem is solved holistically and meets business expectations and standards. We have implemented a data governance policy that ensures consistent and transparent data availability for AI projects via data platforms. Moreover, we periodically track key performance metrics such as recall, precision and F1 scores on production deployments. This helps us identify data drift in time and recalibrate and retrain models as and when required. Our fraud detection models are prone to shifts in data patterns as potential fraudsters try to game the system. Putting the right frameworks and governance methods around performance tracking has helped us ensure optimal standards. For us, ethical issues like data privacy, security and transparency in AI development are of the highest priority as it’s part of the organisation’s risk framework. Our core values empower our data scientists and ML engineers with an open culture where failures are appreciated as part of growing. Sustainable and right methods are always given priority over faster results. AIM: What explains the growing conversation around AI ethics, responsibility, and fairness? Vipin Kumar Singh: Ethical AI guarantees that an organisation’s AI systems respect human dignity and do not hurt individuals in any way. This covers a wide range of issues, including justice, non-weaponisation, and responsibility, like in the case of self-driving cars that cause accidents. Creating awareness and literacy around AI ethics is increasingly becoming the need of the hour as the field is rapidly moving from more of a research area to a key business function that supports the day-to-day business processes. As AI-driven processes are not always explainable such as in the case of neural networks, it becomes very important to ensure the right data and methods are used to build AI capabilities. Poorly designed projects– built on faulty inadequate or biased data–can have unintended, potentially harmful, consequences. AIM: How do you mitigate biases in AI systems? Vipin Kumar Singh: We live in a world of biases, which often get reflected in the data sets used to train AI models. In a recent example, Twitter‘s image cropping algorithm favoured white people’s faces over black which raised many eyebrows. We do mindful data analysis and preparation to eliminate the bias introduced by data. There is also the possibility of the ML algorithm being biased in some way, therefore the requisite checks are made before picking any algorithm. On top of that, there is always the possibility of human bias based on who is working on the problem. This gets minimised by doing multiple rounds of the solution, code and data review by different members of the team. As a final step, we deploy all our models with feedback gathering framework so that any biases affecting users get highlighted and corrected proactively. AIM: What processes do you have in place to protect consumer data? Vipin Kumar Singh: Here are the ways we ensure data privacy: By anonymising and encrypting data at the source, rest and in motion.Defined and implemented a data privacy and security policy. Subsequently, through periodic internal and third-party audits conducting assessments to identify improvement areas.Be vigilant of customer feedback on privacy around our business and trends throughout the industry as well. Using privacy by design principles in our system design processes. We’ve used open models for use cases like object detection. Our data science leadership does a thorough review of the literature around models before picking them up. We test these models internally to look out for issues around our AI development principles before taking them into production.","excerpt":"We use the CRISP-DM methodology to ensure each data science problem is solved holistically and meets business expectations and standards.","categories":["AI Features"],"tags":["AI fairness","ai governance","Cloud services","consumer protection","Data Governance","Data Privacy","data rights","digital transformation","Ethical AI","Interviews and Discussions","modernisation"],"author_name":"Sri Krishna","publish_date":"2022-04-18T11:00:00","publication_year":"2022","word_count":912,"keywords":["consumer protection","data rights","chatbots","ai governance","fraud detection","data science","digital transformation","Cloud services","RAG","Data Governance","NLP","analytics","modernisation","AI","neural network","Ethical AI","ML","Data Privacy","AI fairness","Aim","Interviews and Discussions"],"extracted_tech_keywords":["AI","ML","neural network","NLP","data science","analytics","Aim","RAG","chatbots","fraud detection"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/the-ai-strategy-of-housing-com\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10094456,"title":"Top AI Courses by Google for Free in 2024","content":"Generative AI has become the talk of the town, garnering widespread attention from venture capitalists who are investing heavily in genAI startups, even without the guarantee of immediate returns. By 2030, experts anticipate the generative AI market to reach an impressive $109.3 billion, signifying a promising outlook that is captivating investors across the board. Google, one of the leading players in this space, introduced a suite of generative AI features for Gmail, unveiled the advanced PaLM 2 language model, showcased Med-PaLM 2 for medical applications, and highlighted the capabilities of Bard for developers. Google also announced gen AI enhancements for Google Cloud, including Duet AI, and introduced foundation models like Codey, Imagen, and Chirp. Recently, Google also introduced free training courses on generative AI that also come with a completion badge. Read more: Generative AI is Having An Edison Moment 1. Introduction to Generative AI This short course of 45-minutes provides an introduction to Generative AI, its applications, and its distinctions from conventional machine learning approaches. It also includes information on Google tools that can assist in the creation of your own generative AI applications. 2. Introduction to Large Language Models In this course, you will get an overview of large language models (LLMs), their definitions and explaining their potential applications. It also delves into the concept of prompt engineering, which can improve the performance of LLMs. Additionally, the module introduces various Google tools that can assist in the development of personalised genAI applications. 3. Attention Mechanism This module focuses on attention mechanisms in deep learning and its applications in enhancing the effectiveness of different ML tasks. It explores how attention can be used to improve machine translation, text summarisation, and question answering, among other tasks. Pre-existing knowledge of ML, deep learning, NLP, computer vision, and Python programming is necessary. 4. Transformer Models and BERT Model This course introduces the key elements of the Transformer architecture, including the self-attention mechanism, and its application in constructing the BERT model. Additionally, you will understand various tasks that BERT can be employed for, including text classification, question answering, and natural language inference. Prior knowledge in intermediate ML, word embeddings, attention mechanisms, and proficiency in Python and TensorFlow are recommended. 5. Introduction to Image Generation Adding to the list of interesting courses, this course is designed to teach you about diffusion models that are responsible for generating images. Throughout the course, you will learn the principles behind diffusion models, as well as how to train and utilise them on Vertex AI. Prior knowledge in ML, deep learning, convolutional neural networks (CNNs), and programming in Python is required to fully benefit from this course. 6. Create Image Captioning Models This course will prepare you to build an image captioning model through deep learning. It covers the key elements of such a model, including the encoder and decoder, as well as the training and evaluation processes. Upon completion of this module, you will possess the skills to develop your own image captioning models and utilize them to generate captions for images. Prior familiarity with ML, deep learning, NLP, computer vision, and Python programming is advantageous. 7. Encoder-Decoder Architecture In this course, you will understand encoder-decoder architecture, a widely used and effective ML framework for tasks involving sequences, such as translating languages, summarising text, and answering questions. The module covers the key elements of the encoder-decoder architecture and provides instructions on training and deploying these models. To make the most of this module, it is essential to have a strong foundation in Python and familiarity with TensorFlow. 8. Introduction to Responsible AI Building products with a focus on ethical AI is important. So this course by Google where you will understand the significance of responsible AI, as well as how Google incorporates responsible AI into its products is a need of the hour. 9. Introduction to Generative AI Studio Google introduced a new genAI Studio at Google I\/O. So this course gives an introduction to Generative AI Studio, a tool within Vertex AI and how to use it effectively. Generative AI Studio allows you to create and personalise generative AI models, enabling you to incorporate their abilities into your own applications. 10. Generative AI Explorer (Vertex AI) The Generative AI Explorer – Vertex AI Quest is a series of labs on using generative AI on Google Cloud. It covers the Vertex AI PaLM API family, including models like text-bison, chat-bison, and text embedding-gecko. You’ll learn about prompt design, best practices, and applications like text classification and summarisation. The quest consists of four modules: foundation models, model tuning, PaLM API, and Text Embedding API. Read more: AI Cloud Wars: Azure AI vs Vertex AI","excerpt":"Google has unveiled a series of short courses on generative AI that can be accessed for free","categories":["AI Trends"],"tags":["Courses"],"author_name":"Shritama Saha","publish_date":"2023-06-05T12:37:20","publication_year":"2023","word_count":776,"keywords":["GenAI","machine learning","AI","neural network","ML","computer vision","NLP","deep learning","generative AI","foundation models","Courses"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","neural network","NLP","computer vision","generative AI","GenAI","foundation models"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/master-gen-ai-for-free-from-google\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10076019,"title":"What Ethereum Plans To Achieve, Solana Already Has","content":"First, there was proof-of-work, then came proof-of-stake, and now, there is proof-of-history. This is the chain of evolution of blockchain networks from Bitcoin to Ethereum to Solana. After the launch of Solana in April 2019, Ethereum faces a killer competitor. The Ethereum Merge, which happened in mid-September, was a big step towards the modernisation of the blockchain network, but it is still far from its competitor Solana in terms of transactions per second (TPS), finality time (average time to mine a block) or gas fees (transaction fees). Solana boasts of being able to handle up to 710,000 transactions per second, which is 30 times the amount that Visa currently handles. Although it has never actually gone beyond 50,000 in the past, this doesn’t mean it can’t do 710,000. It just hasn’t tried it. The Solana network is fast, monstrous, and also cheap. But, what makes it so fast and cheap? The secret is hidden in the consensus mechanism,  a process by which a group of computers – or nodes – on a network determine which blockchain transactions are valid and which are not. Solana doesn’t use proof-of-work, and doesn’t really use proof-of-stake. Solana founder Anatoly Yakovenko described a new system called proof-of-history in his white paper. One major issue that other blockchains face is that they must agree on a time. Computers must constantly wonder what time it is because they cannot look at their phone or even the Sun to determine it. On other blockchains, the nodes or computers on the network have to chat back and forth until they agree on a time. They have to do this process before submitting a block, and this chatter can take up a lot of time. Solana fixes this using proof-of-history to have everyone timestamp their blocks and use cryptographic proof so that they don’t have to wait on everyone to agree on the time. Solana basically uses proof-of-stake, but it adds in the special variable of time. So, proof-of-history is not a consensus mechanism but a way of integrating time into the blockchain data. “We use something called timestamps to place a specific date and time on the blocks, and we do this so that it allows for a very fast sequencing of validators. So that they know their order to submit blocks without having to communicate back and forth,” writes Yakovenko in a blog. Proof of history explained by Solana Proof of history enables the blockchain to work fast while maintaining security and decentralisation. Here is an example: If you take snapshots of dye diffusion and they were scrambled, you would know exactly how to place the resulting images in order again because of the laws of entropy as a function of time. Proof of history uses a recursive verifiable delay function to hash incoming events and transactions. Every event has a unique hash and counts along this data structure as a function of real-time. This information tells us what event had to come before another, almost like a cryptographic timestamp, giving a verifiable ordering of events as a function of time. Each node has a cryptographic clock that helps the entire network agree on the time and order of events without having to wait to hear from other nodes. Unlike slower traditional blockchains, oftentimes throttled by various consensus mechanisms, Solana achieves high throughput without sacrificing network’s security. The basic question arises that if Solana can easily solve the decentralisation, security and scalability of the blockchain network, why not Ethereum. Has Solana solved the blockchain trilemma? Coined by Ethereum founder Vitalik Buterin, the term ‘blockchain trilemma’ talks about the challenges developers face in creating a blockchain that is scalable, decentralised and secure — without compromising on any facet. The trilemma explains that truly decentralised networks need to choose between security and scalability. Solana calls itself a blockchain that is decentralised, permissionless, secure, and scalable. However, in the crypto world, Solana is frequently criticised for not being truly decentralised. This criticism is partially justified because Solana has far fewer nodes than Ethereum or Bitcoin. The reason for this is that Solana’s high transaction throughput necessitates significant hardware requirements for validators. Network participants who want to run validator nodes must purchase custom-built hardware. As a result, personal computers and standard fibre optic internet connections cannot run Solana validator nodes. As a result, the network has far fewer validators than competing blockchains such as Ethereum. There are currently about 2,000 Solana validators. Experts feel that Solana has purposefully chosen to prioritise scalability over decentralisation. However, this does not imply that Solana is not sufficiently decentralised, as decentralisation exists on a spectrum with certain trade-offs. Different levels of decentralisation may be appropriate for different blockchains depending on their use cases and security requirements. A serious contender for Ethereum “Clearly, Solana has significant advantages at throughput and performance level. Also, they have achieved significant cost optimisation compared to the high gas fees of Ethereum. It is surely one of the strongest contenders among L1 chains to challenge Ethereum,” said Pradeep Singh, founder & CEO of SquirrelVerse, a Web3 platform for NFTs. He, however, added that Solana has to address some challenges which have been observed before such as attacks and outages. Ethereum is more secure against DDoS attacks, and Ethereum has never faced outages at the entire chain level.","excerpt":"Solana uses proof-of-history, which is a way of integrating time into the blockchain data","categories":["AI Features"],"tags":["Bitcoin","Blockchain","Ethereum"],"author_name":"Tausif Alam","publish_date":"2022-09-28T17:00:00","publication_year":"2022","word_count":885,"keywords":["Go","Bitcoin","Blockchain","AWS","AI","cloud_platforms:AWS","programming_languages:R","Scala","RAG","ViT","programming_languages:Scala","R","Ethereum"],"extracted_tech_keywords":["AI","RAG","AWS","R","Go","Scala","ViT","cloud_platforms:AWS","programming_languages:R","programming_languages:Scala"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/what-ethereum-plans-to-achieve-solana-already-has\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":23807,"title":"Evaluation Of Major Deep Learning Frameworks","content":"There are many deep learning frameworks out there and it can lead to confusion as to which one is better for your task. in this article, we will evaluate the different frameworks with the help of this open-source GitHub repository. Frameworks are like different programming languages. One has its own way of communicating with the systems. This article shows how the frameworks built for deep learning are different in terms of various factors. There can be situations where the code could be written in Java, while you are familiar with Python. Instead of writing the model in Python language you can simply implement Java and work on the model. The goal of this repo was to make a comparison between different benchmarks helping the data scientists implement their work easily, and to make a GPU comparison with respect to the advancement. The open-source communities have also collaborated to this project making the process easier. Benchmarking Outcomes Three types of dataset were used on two different GPUs for framework comparison. The first was the CIFAR-10 dataset with 50,000 training samples and 10,000 test samples, uniformly distributed over 10 classes. Every image has a depth of 3 and 32×32 shape and has been rescaled from 0-255 to 0-1. A CNN has been used across different platforms with GPU support — Nvidia K80 and P100, with CUDA and cuDNN. CUDA (Compute Unified Device Architecture) is a parallel computing platform developed by Nvidia. CUDA support is required on these frameworks for the implementation of GPU while training and testing the models. Similarly, cuDNN is a Deep Neural Network library developed by Nvidia for high tuning of the computations like front propagation and backpropagation. DL Library K80\/CUDA 8\/cuDNN 6 P100\/CUDA 8\/cuDNN 6 Caffe2 148 54 Chainer 162 69 CNTK 163 53 Gluon 152 62 Keras(CNTK) 194 76 Keras(TF) 241 76 Keras(Theano) 269 93 Tensorflow 173 57 Lasagne(Theano) 253 65 MXNet 145 51 PyTorch 169 51 Julia – Knet 159 * * – Not submitted at the time of benchmarking. Average Time for 1,000 Images: ResNet-50 – Feature Extraction The next model was the pre-trained ResNet50 split after average pooling at the end (7,7), which creates a 2048D vector. After passing this to a softmax, it squashes the values between 0 and 1, like probabilities. This has been performed on the same Nvidia GPUs and CUDA platforms. DL Library K80\/CUDA 8\/cuDNN 6 P100\/CUDA 8\/cuDNN 6 Caffe2 14.1 7.9 Chainer 9.3 2.7 CNTK 8.5 1.6 Keras(CNTK) 21.7 5.9 Keras(TF) 10.2 2.9 Tensorflow 6.5 1.8 MXNet 7.7 2.0 PyTorch 7.7 1.9 Julia – Knet 6.3 * * – Not submitted at the time of benchmarking. A sentiment analysis has been done on the IMDB dataset available on the website. The training set had 25,000 reviews and the test samples were in 25,000, strategically sampled with equal number of positives and negatives. Comparison of the time taken during training is shown below DL Library K80\/CUDA 8\/cuDNN 6 P100\/CUDA 8\/cuDNN 6 Using cuDNN? CNTK 32 15 Yes Keras(CNTK) 86 53 No Keras(TF) 35 26 Yes MXNet 29 24 Yes Pytorch 31 16 Yes Tensorflow 30 22 Yes Julia – Knet 29 * Yes * – Not submitted at the time of benchmarking. Study Analysis Most frameworks use cuDNN’s algorithm to run an exhaustive search and optimise the algorithm used for the forward-pass of convolutions on your fixed-sized images. For example, this can be implement on the torch platform with the following command “torch.backends.cudnn.benchmark=True”. cuDNN improves speed of the computations while training the RNNs. The downside is that running inference on CPU later-on may be more challenging. From the analysis we can see that the K80 GPU is less powerful compared to the P100 GPU even though both have CUDA and cuDNN support. This particular benchmarking on time required for training and feature extraction exhibits that Pytorch, CNTK and Tensorflow show a high rate of computational speed. It has been determined that larger number of frameworks use cuDNN to optimize the algorithms during forward-propagation on the images. By comparing these frameworks we found out that the architecture and the data used by each of them was similar. The computation speed and time with respect to all the frameworks has been conducted but . They are simply meant to show how to create the same networks across different frameworks and the performance on these specific examples. ONNX (Open Neural Network Exchange Format) was useful not only while developing a framework, but also while converting the score of the model. Also, MMdnn tools convert between different framework and visualise the architecture at the same time. The study was completed with the help of the various teams’ contribution who are working on different frameworks. First, the Keras with Tensorflow has channels-last configuration which needed to specify the parameters at every batch, but now it has been developed and a channel-first is now a native configuration. This repo is the version 1.0, and the team is working on considering other benchmarks to work on and build a comparison model. Frameworks – Tensorflow, Julia, MXNet, Keras, Theano, R, CNTK, Pytorch, Caffe2, Chainer and Gluon.","excerpt":"There are many deep learning frameworks out there and it can lead to confusion as to which one is better for your task. in this article, we will evaluate the different frameworks with the help of this open-source GitHub repository. Frameworks are like different programming languages. One has its own way of communicating with the […]","categories":["Deep Tech"],"tags":["Caffe2","Data Science","Deep Learning","Keras","Machine Learning","Toolkit AI"],"author_name":"Kishan Maladkar","publish_date":"2018-04-17T12:02:40","publication_year":"2018","word_count":848,"keywords":["CUDA","Keras","Caffe2","AI","neural network","PyTorch","ML","sentiment analysis","Machine Learning","RAG","Toolkit AI","deep learning","Deep Learning","Data Science","TensorFlow"],"extracted_tech_keywords":["AI","ML","deep learning","neural network","TensorFlow","PyTorch","Keras","RAG","sentiment analysis","CUDA"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/evaluation-of-major-deep-learning-frameworks\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":36217,"title":"AI Research Group OpenAI Plans To Save The World While Chasing Profits","content":"Open AI has decided to open a new for-profit arm of the famed research think tank in a new shocking twist. OpenAI was in news recently when Elon Musk left the board of the world leading research think tank in light of many disagreements. Sam Altman also recently resigned as the president of Y Combinator, a super incubator for startups to focus on activities at Open AI. In a sense, with this move, Open AI transforms into a fancy AI startup with a world class team. What once started as a nonprofit research lab, co-founded by Musk and Altman in a bid to save the world from the potential dangers of AI and focus on safety of artificial intelligence has now transitioned into a for-profit enterprise. What products could Open AI build as part of its portfolio still remains to be seen. Natural language processing and reinforcement learning are some of the core strengths at Open AI. The new company very well may try selling these AI services to monetise them in the near term. When the lab was set up, it raised headlines for poaching some of the best researchers from leading universities, leaving a gap in the academic world. With world class researchers and technologists already in place and a market for cutting-edge applications, this for-profit move will push the company to cater to the real world needs of the enterprises along with allowing Open AI to attract the best talent possible. Open AI faced several attritions as leading researchers like Pieter Abbeel and Andrej Karpathy left Open AI to target more commercial gigs of Musk where AI could be applied to real world problems. Move towards profits and investors will encourage more world class researchers and engineers to stay at Open AI. Altman’s vision to balance capitalism with responsibility has resulted in the company trying out the “capped-profit” model. The statement from Open AI on its blog, says “We’ve created OpenAI LP, a new “capped-profit” company that allows us to rapidly increase our investments in compute and talent while including checks and balances to actualize our mission.” The company pivots to this mode of operation in a hope to actualise its mission in a more fruitful way. Open AI is focussed on making AGI(Artificial general intelligence) safer and ensure a broad adoption of AGI. Even though the Open AI statement said that the daily work at Open AI would remain the same, it is doubtful how it would make money without building end to end products. Open AI says, “The fundamental idea of OpenAI LP is that investors and employees can get a capped return if we succeed at our mission, which allows us to raise investment capital and attract employees with startup-like equity.” With a firm focus on profits the company has become the most probable place where AGI will actually be developed and sold off to businesses. With motivation to raise billions of dollars in funding and make multiple times the funding amount the company seems poised to serve enterprises who are struggling to get real value for their AI bucks. Open AI with their superior machine learning capabilities can become a competitor to major AI service companies once they decide which products they will focus on. Open AI LP’s new structure allows its investors to only make 100x on their returns while everything above that will go back to the non profit entity, “Open AI Non Profit.” Open AI realise that their core mission is unrealistic without the flow of steady and huge amounts of capital. Open AI as a company is late in embracing the AI markets and building products for it will be definitely tough.Open AI LP will have a long way to go to cultivate the ability of creating market ready products. Here is where Sam Altman comes in. Sam Altman left his responsibilities at Y Combinator to lead the new Open AI LP and bring some business to the company which has been in research mode for years.","excerpt":"Open AI has decided to open a new for-profit arm of the famed research think tank in a new shocking twist. OpenAI was in news recently when Elon Musk left the board of the world leading research think tank in light of many disagreements. Sam Altman also recently resigned as the president of Y Combinator, […]","categories":["Global Tech"],"tags":["OpenAI","Startups"],"author_name":"Abhijeet Katte","publish_date":"2019-03-13T05:57:16","publication_year":"2019","word_count":668,"keywords":["Go","API","machine learning","artificial intelligence","OpenAI","AI","RPA","RAG","ViT","Startups","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","OpenAI","RAG","R","Go","API","ViT","RPA"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/ai-research-group-openai-plans-to-save-the-world-while-chasing-profits\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":17771,"title":"Analysis of New York Philharmonic Performances and role of big data in the entertainment industry","content":"The New York Philharmonic plays a leading cultural role in New York City, the United States, and the world and it was founded by the American conductor Ureli Corelli Hill in 1842, with the help of the Irish composer William Vincent Wallace, it was then called the Philharmonic Society of New York, it was the third of its kind on the American soil since 1799 and took as it’s purpose “the advancement of instrumental music”. The first concert took place on December 7, 1842c in the Apollo Rooms on lower Broadway before an audience of 600-700, the concert opened with the Beethoven’s Symphony No.5, led by Hill and two other conductors, Henry Christian Timm from Germany and French-born, Denis Etienne, they led the three-hour program which included chamber music and several other operatic selections with the leading singers of the day. At the end of the season, the players would divide any proceeds among themselves. Beethoven remains one of the most popular composers. Beethoven’s Ninth and a new home The group was doing well and only after a dozen performances within four years, the Philharmonic organized a concert to raise funds to build a new music hill, the centerpiece was the premiere of Ludwig Van Beethoven’s Symphony 9 and had to take place at Castle garden on the southern tip of Manhattan, around 400 instrumental and vocal performers were arranged and the orchestra was conducted by George Loder, this was the first time that the chorals were translated into English. International Presence The New York Philharmonic Orchestra is a long leader in the American Music life and has become renowned around the globe, having appeared in around 62 countries on five. Historical tours include the first tour to Europe with the composer Toscanini, the tour of USSR, in 1959 with Leonard Bernstein. The orchestra is committed to developing tomorrow’s leading orchestral musicians, the Philharmonic has established the New York Philharmonic Global Academy, the orchestra also continues its residency partnership with the University Musical Society of the University of Michigan. Austrian, Luxembourgian and Swiss Performances German Performances Belgian Performances Big data and Spotify Spotify is a commercial music streaming service, it was launched in the year 2008 and since then it has registered around 25 million active users of which 6 million are paying for premier services. Its social marketing is also strong as it has 3.7 million Facebook fans, plus its database has around 20 million songs with the list increasing every day. Its popularity can be noted by the fact that users have created over 1 billion playlists and over $500 million has been paid out to rights holder since the initial launch, without big data techniques, Spotify would not exist. The whole organization has an incline towards data as users create around 600 Gigabytes of data every day and 4 Gigabytes of data is generated in Hadoop. The organization has 28 Petabytes of storage which is spread out four data centers across the world. Along with stringent hardware capabilities, they have also developed a workflow manager on Python called Luigi which is open sourced. Most of the data is open to the public that allows billions of log messages which in turn helps Spotify to provide music recommendations. The data is primarily used in decision-making and forecasting to provide a personalized touch to every user as the company has an ever-growing presence in many countries, plus more listeners account for more data creation and in coming years the recommendation algorithm will be more precise, thus more payouts to the right holders. Big data truly enabled Spotify to change the music industry and has attracted in creating of various streaming services such as Netflix and Amazon Prime which has brought the entertainment industry to the brink of digital revolution. Role of big data and analytics in the entertainment industry Steve Jobs introduced the iPod around 15 years ago and since then music fans have embraced this radical change on how they listen to music, but people don’t understand how raw data and information which is accumulated via downloads and online searches influences the entire entertainment industry cause the effect comes to the overall margin and not only what songs are marketed and sold but what become hits. Although decisions still hinge upon subjective assumptions and rest upon the executive about what sounds good or which artist will be easier to market, the overall decisions are increasingly shaped by big data and the analytics that help this information turn into actions. Big data is a relative term that reflects a amount of information people generate and over the years, its been a lot. Experts estimate that humans generate more information in one minute than in every moment from the earliest historical record through 2000. Unsurprisingly, harnessing this data has shaped the entertainment industry in radically new ways.","excerpt":"The New York Philharmonic plays a leading cultural role in New York City, the United States, and the world and it was founded by the American conductor Ureli Corelli Hill in 1842, with the help of the Irish composer William Vincent Wallace, it was then called the Philharmonic Society of New York, it was the […]","categories":["IT Services"],"tags":["big data role"],"author_name":"Kshitiz Gupta","publish_date":"2017-09-18T11:39:03","publication_year":"2017","word_count":807,"keywords":["big data","Go","AI","IPO","Git","RAG","Python","analytics","GAN","R","big data role"],"extracted_tech_keywords":["AI","analytics","RAG","Python","R","Go","Git","big data","GAN","IPO"],"url":"https:\/\/analyticsindiamag.com\/it-services\/analysis-of-new-york-philharmonic-performances-and-role-of-big-data-in-the-entertainment-industry\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":34237,"title":"Vijay Kelkar, Ex-Chairman Of Finance Commission, Pitches For NITI Aayog 2.0","content":"Former chairman of Finance Commission, Vijay Kelkar in a paper titled Towards India’s New Fiscal Federalism has pitched for setting up of NITI Aayog 2.0 with a plan to allocating development or transformational capital or revenue grants to the States. “It is desirable that a functionally distinct entity such as the new NITI Aayog or NITI Aayog 2.0 be put to use to do the job at hand related to the structural issues including removal of regional imbalances in the economy,” said Kelkar in the paper. However, the former chairman has also stated that he is not suggesting for a moment that the new NITI Aayog 2.0 should take the form of the socialist-era Planning Commission. According to Kelkar, ever since NITI Aayog has replaced the Planning Commission (which was promoting regionally balanced growth in India), the government has lost a significant policy reach. Furthermore, Kelkar said that if we want to make the institution more effective, then it is high time that NITI Aayog is put at the “High Table” of decision making of the Government. “This means the Vice Chairman of the new NITI Aayog will need to be a permanent invitee of the Cabinet Committee on Economic Affairs,” Kelkar stated. That is not all, according to the former Finance Commission chairman, annually NITI Aayog 2.0 will require the resources of around 1.5% to 2% of the GDP to provide suitable grants to the States for mitigating the development imbalance. He has also said that over the last three decades India has consistently accelerated its growth rate. At present, the economy of the nation is on a growth turnpike, not very different from the growth miracles experienced by Asian tigers as well as China in earlier decades.","excerpt":"Former chairman of Finance Commission, Vijay Kelkar in a paper titled Towards India’s New Fiscal Federalism has pitched for setting up of NITI Aayog 2.0 with a plan to allocating development or transformational capital or revenue grants to the States. “It is desirable that a functionally distinct entity such as the new NITI Aayog or […]","categories":["AI News"],"tags":["NITI Aayog"],"author_name":"Harshajit Sarmah","publish_date":"2019-01-28T11:49:28","publication_year":"2019","word_count":289,"keywords":["Go","API","programming_languages:R","AI","programming_languages:Go","ViT","R","NITI Aayog"],"extracted_tech_keywords":["AI","R","Go","API","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/vijay-kelkar-ex-chairman-of-finance-commission-pitches-for-niti-aayog-2-0\/","complexity_score":3,"technical_depth":7,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10112759,"title":"10 Generative AI Startups that Made it to JioGenNext Cohort","content":"JioGenNext, the renowned startup accelerator, has introduced its latest cohort MAP’ 24, emphasising generative AI technology. The cohort consists of ten dynamic startups spanning diverse sectors such as healthcare, banking, legal services, entertainment, and agriculture. Startups in the Cohort: Medhini-Arficus Founded in 2019 by Sandeep Sinha, Bengaluru-based Medhini-Arficus is an AI disease screening, diagnostics, and prediction SaaS-SAMD platform. Targeting various healthcare establishments, including primary healthcare centres, rural areas, diagnostic centres, dental clinics, medical clinics, maternity homes, hospitals, direct consumers, etc., the company aims to provide a user-friendly mechanism, akin to having a hospital in one’s pocket. It also claims to provide an AI-based diagnosing solution in just 53 seconds, alongside data-driven genome sequencing pandemic wave prediction analysis and more. Dista Pune-based Dista was founded by Shishir Gokhale, Deepak Garg and Nishant Kumar in 2020. The company offers a low-code location intelligence platform tailored for large enterprises. With over 40 enterprise customers across sectors like financial services, supply chain, logistics, and retail, Dista maximises field sales and service revenues, provides geospatial analytics for advanced insights, and enhances business revenue and operational efficiency. ContractKen Founded in late 2021 by former EXL exec Amit Sharma, ContractKen is building the world’s best AI copilot for contracts, assisting legal professionals in corporate legal teams and law firms to draft, review, and negotiate contracts efficiently. By ensuring consistency with organisational standards, ContractKen aims to streamline contract management processes rigorously. Beatoven.ai Based in Bengaluru, Beatoven.ai was started by Siddharth Bhardwaj and Mansoor Rahimat Khan in 2021. The company leverages advanced AI music generation techniques to create mood-based music tailored for video, podcast, and game creators. Backed by user studies demonstrating the impact of music on engagement, Beatoven.ai enhances creativity, empathy, emotion, and information dissemination in multimedia content. Expertia AI Bengaluru-based Expertia AI was founded by former IBM researcher Akshay Gugnani and serial entrepreneur Kanishk Shukla in 2021. The company focuses on discovering, developing, and deploying skills for the future of work. With offerings like the Virtual Recruiter and Expertia Career Site, Expertia AI empowers recruiters to screen candidates efficiently, manage job applications remotely, and access a talent pool for new vacancies. VoiceOwl Based in Indore, VoiceOwl was founded by Vivek Jain, Ruchika Drabla Keshwani, and Piyush Keshwani in October last year. The company specialises in developing bespoke, enterprise-grade, large language models (LLMs). By building custom GenAI-powered applications, VoiceOwl ensures secure connectivity with enterprise data and APIs, offering personalised and secure solutions tailored for enterprise environments. HYRGPT Bengaluru-based HYRGPT was founded by Debi Kar, Sameer Dharap and Neha Mathur in June last year. The company streamlines the recruitment process by automating outreach and conducting conversational AI-based interviews. With trained LLMs and hiring analytics, HYRGPT delivers instant summaries and competitive rankings, facilitating data-driven hiring decisions. Jhana Founded in 2022 by Hemanth Bharatha Chakravarthy, Ben Hoffner-Brodsky and Em McGlone, Jhana is building India’s first AI paralegal. Based in Chennai, the company offers a suite of intelligent research and drafting assistants for Indian legal professionals. KissanAI Founded by Prateek Desai in March last year, KissanAI is innovating at the intersection of AI and farming. Bridging barriers of literacy and language, the company is reshaping farming practices with groundbreaking AI technology. Providing real-time advice on crop management and pest control, it empowers farmers with actionable insights. OnFinance AI Founded by Priyesh Srivastava and Anuj Srivastava, alumni of BITS Pilani and NYU, Bengaluru-based OnFinance AI boasts a team with prior experience at esteemed institutions such as HSBC, Barclays, Credit Suisse, IIFL, Nutanix, and Jio. The company aims to leverage the potential of generative AI to enhance the banking, wealth management, and insurance sectors JioGenNext reaffirms its commitment to supporting these visionary startups by providing mentorship and resources. As these startups continue to innovate and disrupt their respective industries, JioGenNext aims to foster their growth and impact in the realm of generative AI technology. Stay tuned for updates on their progress and achievements. [Updated | February 14, 2024 | 18:23] The headline of the story erroneously mentioned ‘JioNextGen’ instead of ‘JioGenNext.’ It has now been changed to reflect the correct name.","excerpt":"The cohort consists of ten dynamic startups spanning diverse sectors such as healthcare, banking, legal services, entertainment, and agriculture.","categories":["Deep Tech"],"tags":["Jio AI Cloud"],"author_name":"Mohit Pandey","publish_date":"2024-02-13T20:41:26","publication_year":"2024","word_count":677,"keywords":["Go","API","GenAI","AI","ML","RAG","Aim","Jio AI Cloud","generative AI","analytics","R"],"extracted_tech_keywords":["AI","ML","analytics","generative AI","GenAI","Aim","RAG","R","Go","API"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/10-ai-startup-that-made-it-to-jiogennext-cohort\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10134303,"title":"HARMAN Introduces ForecastGPT, a GenAI Platform for Enterprises","content":"HARMAN, a Connecticut-based Samsung subsidiary, specialising in audio electronics, launched HARMAN ForecastGPT, a predictive analytics platform that enables organisations to become more efficient through future forecasting and optimum resource allocation. It is aimed at understanding complex data patterns, and some of its chief features are advanced AI capabilities, real-time adaptability and the ability to seamlessly integrate with various platforms. It is compatible with any data format and source including CSV, Excel, SQL, and API. HARMAN ForecastGPT’s Versatility It can benefit businesses in a plethora of ways including sales forecasting, supply chain forecasting, financial planning, and marketing even. “Embracing AI is imperative for business success and at HARMAN DTS, we are pioneering the application of AI to deliver tangible, bottom-line results. By understanding the unique challenges and aspirations of each client, we’re crafting AI solutions that go beyond generic predictions. The ForecastGPT platform is a testament to our commitment to equipping businesses with the tools to move past challenging roadblocks and fully capitalise on the potential of AI,” said Nick Parrotta, President – Digital Transformation Solutions & Chief Digital and Information officer at HARMAN. HARMAN & its Growing Capabilities HARMAN has a wide portfolio including audio and video systems, car audio, connected car solutions, professional audio and lighting equipment, and more. Additionally, it is also the parent company to brands like JBL, Harman Kardon, AKG, Mark Levinson, and Infinity Systems. “HARMAN’s data science team has made significant contributions by incorporating machine learning and deep learning models into a range of applications, such as predictive analytics, computer vision, NLP, and graph analytics,” said Dr Jai Ganesh, chief product officer of HARMAN, in an exclusive interview with AIM previously. The data science team of HARMAN employs both open source as well as commercial tools such as Python, TensorFlow, PyTorch, AWS, Azure, or Google Cloud Platform, Java, C++, Git, Jenkins, Docker, Kubernetes, R, Jupyter, SAS, MongoDB, Spark, Kafka, MySQL, RStudio, KNIME, RapidMiner, H2O etc.","excerpt":"Designed specifically for uncertain and volatile markets, ForecastGPT leverages AI to help businesses make accurate predictions and informed decisions.","categories":["AI News"],"tags":["Harman"],"author_name":"Aditi Suresh","publish_date":"2024-09-03T13:45:55","publication_year":"2024","word_count":320,"keywords":["data science","machine learning","AI","ML","computer vision","NLP","Aim","deep learning","Harman","analytics","TensorFlow"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","NLP","computer vision","data science","analytics","Aim","TensorFlow"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/harman-introduces-forecastgpt-a-genai-platform-for-enterprises\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10292,"title":"Meet Loan Singh, startup that provides education loan for short term analytics courses","content":"Loan! – An expression that brings about a series of nightmarish emotions such as complex documentation and an endless scrutiny. The traditional financial institutions are archaic, slow and inefficient for the millennial population. We needed a change – a solution to the traditional remedies. The solution is a lending platform that uses minimal documentation and provides quick disbursals. So, keep reading to know more about it… Digital data is exploding, and so is the requirement for individuals with a skillset to utilize data efficiently. The market has exploded with varied courses and specializations in the domain of analytics. A typical course ranges from 20K to 2 Lakhs or more. This requires the aspirant to have enough funds and if not then to acquire financial assistance. Institutions, which offer such support, treat it as a personal loan. High interest rates are charged and credit history is evaluated. Most of the aspirants being fresh graduates or young professionals are clueless about all these financial concepts and tend to have a hard time getting a loan. LoanSingh – An intellectual product created by Seynse Technologies Pvt. Ltd, based out of Goa. It is a credit model built in-house by Seynse Technologies to leverage digital data with the help of data science techniques (which include statistics, math, data gathering and modeling to name a few). The model evaluates the credit worthiness of a borrower with the help of easily available professional, educational, social and financial data. Currently the default rate of the model is 0%, due to its risk model. LoanSingh is trying to revolutionize the way lending is done, specific to professional skill enhancement for aspirants with no credit history (around 85% of Indian population). Loan Process at LoanSingh – A simple digital process comprising of the following steps: Social Authentication by the user. Loan application is filled by the user- Education, Company Name, and Salary. Upload bank statement of last 6 months. Loan approved within a minute and EMI is shown to the borrower based on their score. Post approval, the borrower needs to submit the Agreement and a Post-dated Cheque. After which, the loan amount is disbursed within hours. The team of LoanSingh has built a platform that introduces lenders to the borrowers. They are facilitators of the loan process. They are not involved with the transaction process directly. LoanSingh selects credit-worthy borrowers with the help of their flagship scoring model called ‘Seynse Score’. They make use of demographics and current financial position to perform the credit evaluation of the borrower. You can track your EMI status through your registered dashboard. Minimal documentation required. No paperwork and quick disbursal. With a dedicated and experienced team, Loan Singh is quickly becoming a major player in the burgeoning P2P lending scene in India. To know more, call on 1800 2744747 or drop an email to ls@loansingh.com","excerpt":"Loan! – An expression that brings about a series of nightmarish emotions such as complex documentation and an endless scrutiny. The traditional financial institutions are archaic, slow and inefficient for the millennial population. We needed a change – a solution to the traditional remedies. The solution is a lending platform that uses minimal documentation and […]","categories":["AI Trends"],"tags":["data scientist india salary"],"author_name":"AIM Media House","publish_date":"2016-06-28T12:21:02","publication_year":"2016","word_count":473,"keywords":["data science","Go","programming_languages:R","AI","data scientist india salary","programming_languages:Go","Git","RAG","analytics","R"],"extracted_tech_keywords":["AI","data science","analytics","RAG","R","Go","Git","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/meet-loan-singh-startup-provides-education-loan-short-term-analytics-courses\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10111821,"title":"Amazon Introduces Rufus, Generative AI Shopping Assistant","content":"Amazon has introduced Rufus, a shopping assistant powered by generative AI. Rufus aims to simplify the Amazon shopping process by providing intelligent assistance, personalised recommendations, and seamless product discovery for users. Key features of Rufus include the ability to guide customers on what to look for in various product categories, helping them shop by occasion or purpose, facilitating comparisons between product categories, and providing personalised recommendations based on specific queries. Moreover, customers can also use Rufus to ask questions about individual products while viewing their detail pages. For instance, “is this pickleball paddle good for beginners?”, or “is this jacket machine washable?” Trained on Amazon’s extensive product catalog, customer reviews, community Q&As, and web information, Rufus is designed to answer a variety of customer queries, offer product comparisons, and provide tailored recommendations. The rollout begins with a beta release to a small subset of U.S. customers using Amazon’s mobile app, with a progressive expansion to additional users in the coming weeks. To access Rufus, beta customers can update their Amazon Shopping app, type or speak their questions into the search bar, and a Rufus chat dialog box will appear at the bottom of their screen, offering answers and suggestions. Users are encouraged to provide feedback by rating answers with a thumbs up or down and offering freeform feedback to contribute to the continuous improvement of Rufus.","excerpt":"It is launching today in beta will progressively roll out to additional U.S. customers in the coming weeks.","categories":["AI News"],"tags":[],"author_name":"Siddharth Jindal","publish_date":"2024-02-02T10:37:22","publication_year":"2024","word_count":226,"keywords":["Go","programming_languages:R","AI","ML","programming_languages:Go","RAG","Aim","generative AI","R"],"extracted_tech_keywords":["AI","ML","generative AI","Aim","RAG","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/amazon-introduces-rufus-generative-ai-shopping-assistant\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10056069,"title":"How AI Can Help Therapists Make Therapy Better","content":"The COVID-19 pandemic took a toll on both the physical and mental well being of humankind. The side effects of the impact have brought in a surge in people seeking help to deal with illness, unemployment and social isolation, all in an increasingly fast-paced, pressure-filled world. However, the digital revolution that has transformed many aspects of our daily lives has not yet truly emerged in behavioural health care. As a supplement to therapy, AI and machine learning now seem to have the potential to revolutionise how we diagnose and treat mental health conditions. Soon, algorithms may become our first line of defence against the mental health struggles that can be debilitating for many. How can AI aid? Researchers are now pioneering a new approach to mental health care wherein an AI analyses the language used during the therapy sessions. An automated form of quality control is becoming more and more essential in helping therapists meet the demand. Natural-language processing, also known as NLP, identifies which parts of a conversation between the therapist and client and which types of utterance and exchange seem to be most effective at treating different disorders. Understanding therapy’s most essential ingredients could help open the door to more personalised mental-health care, allowing doctors to tailor psychiatric treatments to specific clients, much needed when prescribing drugs. Machine learning techniques that carry out the automatic translation from sessions enable quickly analysing vast amounts of the language. This gives researchers access to an endless and untapped data source: the language therapists use. Researchers also believe that they can use insights from processed data to boost therapy and its outcomes. This can result in more people getting better and healing fast. The Working At first, a few hundred transcripts are annotated by hand to train the NLP models, highlighting the role therapists’ and clients’ words play at that point in the session. The technology works similar to a sentiment-analysis algorithm that can distinguish and tell whether movie reviews are positive or negative. Next, the AI translates natural language into a kind of bar code or fingerprint of a therapy session that reveals the complete role played by different utterances. For example, a fingerprint for a session can show how much time was spent in constructive therapy versus general conversations. Such a readout can help therapists focus more on the former in constructing future sessions. AI techniques could also help prospective clients match with therapists and determine which types of therapy will work best for them. What’s more? A virtual therapist named Ellie was also launched and trialled by the University of Southern California’s Institute for Creative Technologies (ICT). Ellie was initially designed to treat veterans experiencing depression and post-traumatic stress syndrome. It can detect words and nonverbal cues such as facial expressions, gestures, and different postures. Nonverbal signs are an important aspect in therapy yet can be subtle and difficult to pick up. Ellie’s creators argued that this virtual human can advance mental health and improve diagnostic precision. A few psychologists argue that humans might find it easier to share potentially embarrassing information with a virtual therapist, whereas, in human-to-human interaction, there often seems to be a degree of self-restraint. It has been observed that when patients talk to a therapy bot, they report not feeling judged. Therapists, on the other hand, can discuss the AI-generated feedback for further improvements. The idea is to help therapists take control of their professional development, showing them what they’re good at, things that other therapists can learn from, and a few of them not so good at things they might want to work on. Summing up Although AI for mental health still needs to deal with many complexities, research shows that behavioural health interventions benefit from continuity, and technology seems to offer an improved user experience. Furthermore, while the human brain is complex with its own set of challenges, data collection from behavioural health sessions in a consistent, measurable and accessible manner will be essential to better care and better results in the near future.","excerpt":"algorithms may become our first line of defence against the mental health struggles that can be debilitating for many","categories":["IT Services"],"tags":["AI (Artificial Intelligence)","Machine Learning","Mental Health"],"author_name":"Victor Dey","publish_date":"2021-12-18T16:00:00","publication_year":"2021","word_count":672,"keywords":["Go","API","machine learning","programming_languages:R","AI","Machine Learning","Git","programming_languages:Go","NLP","Mental Health","ai_applications:NLP","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","machine learning","NLP","R","Go","Git","API","programming_languages:R","programming_languages:Go","ai_applications:NLP"],"url":"https:\/\/analyticsindiamag.com\/it-services\/how-ai-can-help-therapists-make-therapy-better\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10043763,"title":"Inside NVIDIA Cambridge-1: UK’s Most Powerful Supercomputer","content":"NVIDIA recently launched Cambridge-1, a supercomputer designed and built for external research access. The multinational American tech company claims it to be the United Kingdom’s most powerful supercomputer. It has been built to make available to UK healthcare researchers to use AI to solve pressing medical challenges, including the ongoing pandemic. In a press statement, Jensen Huang, Founder and CEO of NVIDIA, said that Cambridge-1 will empower world-leading researchers in businesses and academia with the ability to perform their life’s work– unlocking clues to disease and treatments– at a scale and speed that was previously impossible in the UK. NVIDIA has invested $100 million to build this supercomputer. Tech behind Cambridge-1 The AI supercomputer brings together decades of NVIDIA’s work in AI, accelerated computing and life sciences, where NVIDIA Clara (a healthcare app framework for developing AI-powered imaging, genomics and deploying smart sensors) and AI frameworks have been optimised to perform large-scale research. A NVIDIA DGX SuperPOD supercomputing cluster, Cambridge-1 features 80 DGX A100 systems. DGX SuperPOD delivers a turnkey full-stack AI data center solution with advanced computing, network fabric, storage and software tools. Additionally, Cambridge-1 integrates NVIDIA A100 GPUs, BlueField-2 DPUs and NVIDIA HDR InfiniBand networking, delivering more than 400 petaFLOPS of AI performance and eight petaflops of LINPACK performance. It will enable researchers and academics the stability to tackle challenging AI training, inference and data science workloads at scale. Source: NVIDIA This launch comes right after NVIDIA unveiled the new DGX SuperPOD– a cloud-native, multi-tenant supercomputer which features 20 or more NVIDIA DGX A100 systems and NVIDIA InfiniBand HDR networking. While usually traditional supercomputers take years to deploy, the modular DGX SuperPOD architecture enables the system to be installed and operational in a couple of weeks. Located at a facility operated by NVIDIA’s partner Kao Data, NVIDIA claims Cambridge-1 to be one among the world’s top 50 fastest computers. It is powered by 100 percent renewable energy. At present, AstraZeneca, GSK, Guy’s and St Thomas’ NHS Foundation Trust, King’s College London and Oxford Nanopore Technologies harness Cambridge-1 to research, develop new drugs, and improve the accuracy of finding disease-causing variations in human genomes. King’s College London and Guy’s and St Thomas’ NHS Foundation Trust will be using Cambridge-1 to train AI models to generate synthetic brain images and better understand dementia, stroke, brain cancer, and multiple sclerosis. By collaborating with AstraZeneca, Cambridge-1 will help create a transformer-based generated AI model for chemical structures, allowing researchers to leverage massive datasets using self-supervised training methods, fueling faster drug discoveries. According to reports, the AI supercomputer is expected to create an estimated value of $825 million over the next decade. Source: NVIDIA Cambridge-1 vs the rest In comparison to Cambridge-1, the world’s fastest supercomputer– Fugaku– reported an initial performance of Rmax of 416 petaFLOPS in the FP64 high performance LINPACK benchmark. However, after last year’s upgrade, Fugaku’s performance increased to a Rmax of 442 petaFLOPS. Pangea III– the world’s most powerful commercial supercomputer built by IBM, has a computing power of 25 petaFLOPS, with storage capacity of 50 petabytes. On the other hand, Intel and Cray’s upcoming supercomputer Aurora is expected to have computing power of 130 teraFLOPS. Aurora will have over nine thousand nodes, each composed of two Intel Xeon Sapphire Rapids processors, six Xe GPU’s and will have memory of around 10 petabytes. It is expected to be completed in 2021. Thus, going by the currently available specs, Cambridge-1 is expected to be positioned at around number 29 in the Top500 supercomputers list, and at number 3 in the Green500 list. Going forward, NVIDIA is also working on building an AI Research Center in Cambridge. Presently featuring an Arm-based supercomputer, Cambridge-1 will be a part of the Center of Excellence.","excerpt":"NVIDIA recently launched Cambridge-1, a supercomputer designed and built for external research access. The multinational American tech company claims it to be the United Kingdom’s most powerful supercomputer. It has been built to make available to UK healthcare researchers to use AI to solve pressing medical challenges, including the ongoing pandemic.  In a press statement, […]","categories":["Global Tech"],"tags":["AI Supercomputer","GPUs","NVIDIA","Supercomputers"],"author_name":"Debolina Biswas","publish_date":"2021-07-17T14:00:00","publication_year":"2021","word_count":621,"keywords":["data science","Rapids","Go","API","AI","AI Supercomputer","RAG","Ray","Aim","Supercomputers","Rust","GPUs","NVIDIA","R"],"extracted_tech_keywords":["AI","data science","Aim","Ray","Rapids","RAG","R","Go","Rust","API"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/inside-nvidia-cambridge-1-uks-most-powerful-supercomputer\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":67543,"title":"Latest Entry-Level Data Science And Analysts Job Openings","content":"According to a recent survey by AIM, the demand for analytics services will see a high rise despite the overall downturn in the global and domestic economy. One of the significant concerns among organisations is the silos of data that are generated every day, which is why organisations have been recruiting data scientists and analysts to gain insights from the raw data. In this article, we list down the latest 8 data science and analysts jobs for freshers one must apply. (The list is in no particular order) 1| Data Scientist at IBM Location: Noida Responsibilities: As a data scientist at IBM, you will help transform their clients’ data into tangible business value by analysing information, communicating outcomes, and collaborating on product development. You will bring deeper experience as well as the strategic insights that will span from architecture, design and development leadership to client-facing solution skills, provide guidance to the larger global Community of Practice (CoP) in their area and drive important community-supported projects, among others. Click here to apply. 2| Data Scientist at Brillio Technologies Location: Bangalore Responsibilities: As a data scientist, your job is to work on image analytics, computer vision using deep neural networks in domains such as object detection, object identification, developing chatbots, or translations or document summarisation with the help of text mining and natural language processing, provide insights using Python or R languages, among others. Click here to apply. 3| Data Analytics Specialist at McKinsey & Company Location: Gurgaon Responsibilities: As a data analytics specialist at McKinsey & Company, you will make decisions based on sound qualitative, including quantitative evidence by collecting as well as making use of statistics, web analytics, surveys, support tickets, financials, and more. Further, you will also derive novel insights from data using a wide variety of quantitative methods, including advanced analytics, and other such. Click here to apply. 4| Data Analyst at GP Strategies Corporation Location: Chennai Responsibilities: As a data analyst, you will be accountable for authority information reporting and co-ordinating BI for designated contracts. You will concentrate on report creation, data manipulation, data integrity, data extraction, determining information requirements and adequate delivery techniques to meet the customers’ needs. Further, you will also develop a range of reports in various formats, including Excel, PowerBI and PowerPoint etc., as requested, collate, analyse and prepare management information reports, among others. Click here to apply. 5| Data Analyst at Colt Technology Services Location: Gurgaon Responsibilities: As a data analyst, you will be responsible for providing operational performance insights and analysis. Besides, you will have to deliver timely inputs and recommendations for improvement in business processes and customer experience. Furthermore, the job would also require providing customer-specific reports and analysis for Ops SMT review meeting with the customer, automating daily reporting requirement through BO and delivering on time with quality, and other such. Click here to apply. 6| Data Scientist at Great Value Foods Location: Delhi Responsibilities: As a data scientist, you will be responsible for selecting and engineering features to build models based on the mining of large text and financials database. You will be training and tuning a variety of machine learning models, enhancing data collection procedures to include information that is relevant for building analytic systems, among others. Click here to apply. 7| Data Scientist at Quark Location: Ajitgarh Responsibilities: In this job role, you will be responsible for developing custom data models and algorithms to apply to data sets. You will utilise predictive modelling to increase and optimise the customer experiences, revenue generation, ad targeting and other business outcomes. The job will also require you to develop processes and tools to monitor and analyse model performance and data accuracy. Click here to apply. 8| Programmer for Data Analysis of DAS at Mercedes-Benz Research and Development Location: Bangalore Responsibilities: As a data analyst, you will be responsible for developing and deploying pipeline to ingest various ADAS systems data to the clusters, develop scalable solutions and improve efficiency and the reliability of the existing application. In addition, you will develop and deploy new features based on business need. Click here to apply.","excerpt":"According to a recent survey by AIM, the demand for analytics services will see a high rise despite the overall downturn in the global and domestic economy. One of the significant concerns among organisations is the silos of data that are generated every day, which is why organisations have been recruiting data scientists and analysts […]","categories":["AI Features"],"tags":["Data Analysis","Data analyst jobs","Data Analytics","Data Science","Data Science Career","Data Science Jobs","Data Scientist","latest technology in machine learning"],"author_name":"Ambika Choudhury","publish_date":"2020-06-17T18:00:00","publication_year":"2020","word_count":683,"keywords":["Data Analysis","data science","machine learning","AI","neural network","chatbots","Data analyst jobs","Data Science Jobs","computer vision","Data Science Career","Python","Aim","object detection","latest technology in machine learning","analytics","Data Analytics","Data Science","Data Scientist"],"extracted_tech_keywords":["AI","machine learning","neural network","computer vision","data science","analytics","Aim","chatbots","object detection","Python"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/latest-entry-level-data-science-and-analysts-job-openings\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10097587,"title":"The Rotten Side of Apple","content":"Unsurprisingly, reports about Apple’s own chatbot emerged close on the heels of Meta announcing Llama 2 with Microsoft as the preferred partner to launch it on Azure cloud. The partnership must have hit Apple hard. Adding insult to injury, Meta CEO Mark Zuckerberg posted a rare picture with Microsoft CEO Satya Nadella. While Apple’s Vision Pro garnered significant media attention, the company has chosen to stay mum on generative AI. During the second-quarter earnings call on May 4, CEO Tim Cook notably omitted any remarks about artificial intelligence in his opening statements, setting Apple apart from other major tech vendors who frequently discuss AI advancements. “As you know, we don’t comment on product roadmaps,” Cook said, when questioned by Credit Suisse analyst Shannon Cross about his perspective on generative AI as a whole and how the technology will integrate with Apple’s products. A Bloomberg report contradicts the notion that Apple is shying away from generative AI. As per the report, Apple is putting a lot of effort into advancing AI, and several teams are working together on this project, according to insiders who wish to remain anonymous. They are also actively working to address any privacy concerns related to the technology. Apple is synonymous with closed ecosystems in the tech industry. Whether it’s their hardware, software, or user experience philosophy, Apple has always advocated for a walled-garden approach, ensuring tight integration between its products and services. It  is known for accessing open source tools without contributing to the development of the technology. Aleksa Gordic, ex-Google DeepMind\/ Microsoft ML engineer reflecting on the same, said that he wished LLaMA 2 license was somehow applicable to fundamental AI research to force Apple to publish their internal research. “They never give anything in return, and yet benefit enormously from the research coming in from all the other big tech companies. It just doesn’t feel right,” he added. https:\/\/twitter.com\/gordic_aleksa\/status\/1681971425224138758 It is quite selfish of Apple to not contribute anything substantial to the open source community. According to Hugging Face stats, Meta holds the highest number of contributions to the open-source community, followed by Google and Microsoft. Not Contributing to Open Source Community The notion of Apple being secretive is not a recent one; it has been ingrained in the company’s culture since the days of Steve Jobs. This approach, however, has its pros and cons. It shows Apple is dedicated to its customers, creating a private ecosystem that prioritises confidentiality and doesn’t disclose personal details. This helps Apple gain customers’ trust with their data. Chatbots are the most lucrative way for any company to gain data. And if Apple is able to win customers’ trust with their chatbot, who knows enterprises might rush to adopt it since confidentiality is key for them. On the other hand, this practice raises questions about whether Apple is actively contributing to the tech society. In the technology space, advancements do not occur in isolation; collaboration is essential. Meta has embraced this approach consistently. Microsoft, a crucial partner of OpenAI, has chosen a different route by collaborating with Meta. It is actively exploring open-source opportunities and acting as a bridge between GPT-4 and Llama 2 to position itself for future advancements. Interestingly, Apple’s Ajax system is built on top of Google Jax, the search giant’s machine learning framework and Apple’s system runs on Google Cloud, which the company uses to power cloud services alongside its own infrastructure and Amazon’s AWS. Also, GPT (which stands for Generative Pretrained Transformer), was originally coined and developed by Google in 2018. It serves as the foundation for both OpenAI’s ChatGPT and Apple’s GPT. If Apple has taken so much from its competitors, should it not pay them back? In comparison, OpenAI, which started as an open source non-profit company, turned into a closed  system, capped-profit company in 2020 following the footsteps of Apple. But, now, things seem to be changing. Now, OpenAI is now looking at releasing the weights of its models.  In the race of AI advancement, Apple seems to be quietly selfish. It’s time they collaborated and contributed more for the benefit of the AI ecosystem, and not just build the ecosystem around their products and services for its customers.","excerpt":"The other side of Apple, as you may already be aware of, paints a vivid picture of exciting new products and services…","categories":["AI Features"],"tags":[],"author_name":"Siddharth Jindal","publish_date":"2023-07-26T14:50:28","publication_year":"2023","word_count":699,"keywords":["ChatGPT","Hugging Face","machine learning","artificial intelligence","OpenAI","AI","chatbots","ML","generative AI","JAX"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","generative AI","ChatGPT","OpenAI","JAX","Hugging Face","chatbots"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/the-rotten-side-of-apple\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":58090,"title":"7-Year-Old Bangalore Girl Creates An App To Help Ambulances Reach Hospitals Faster","content":"Metro cities are dealing with increasing traffic woes which also have a grave impact on emergency services like ambulances leading to a delay in admitting critical patients in the hospital on time. To overcome this challenge, Brinda Jain, a 7-year-old girl from Bangalore has created the Ambulance Whizz App that aims to create a faster corridor for ambulances across the city by informing the traffic police about an upcoming emergency vehicle. Brinda comes from a family of doctors and so was exposed to daily conversations of various emergencies. In emergency conditions like a heart attack, a stroke or a major accident, the patient must get to the hospital in the golden hour (first 60 min). But in many metros, it is common to see ambulances being unable to navigate major traffic jams and get to the hospital on time. The police are not always aware of the situation in every corner of the city to help the ambulance. Learning coding on the WhiteHat Jr platform for the past few months, Brinda wanted to utilize her learning to create an app that would solve this issue. The ambulance driver only needs to update the location of the ambulance on the app, and it will notify the Traffic Control which can then help to create a fast track corridor for the ambulance to pass through, thereby reducing delays and helping patients to reach the hospital in time and save lives. Brinda is one of the twelve winners of the Silicon Valley Program conducted by WhiteHat Jr. Her app was selected after vetting 7000+ entries from across India, where kids were invited to think of an original idea that solves a real-life problem and code an app independently. As a part of this programme, she will visit Silicon Valley in the US to pitch her app to the top venture capitalists (VCs) like Nexus Venture Partners and Owl Ventures. During the weeklong trip, she will also meet some noted Silicon Valley entrepreneurs to learn valuable lessons on entrepreneurship, get an opportunity to visit Googleplex and interact with their engineers, as well as visit the Waymo facility to experience driverless cars and talk to their product managers. “Kids today are very aware of societal issues and want to contribute in their way to address them. We are witnessing many young kids like Brinda creating truly transformational and high impact apps within just 40 hours of coding on the WhiteHat Jr platform, that will make a lasting impact on the world.”, said Karan Bajaj, Founder and CEO, WhiteHat Jr. All these kids are learning to code on WhiteHat Jr, an EdTech startup that teaches coding to kids between 6 to 14 years of age. Teaching cutting-edge curriculum on technologies like AI, robotics, machine learning and space tech, the startup aims to harness the natural creativity of kids and shift their mindset from consumers to creators at an early age. After seeing great success for its online platform, the company is working to bring its AI and robotics coding curriculum to schools across the country. They are also soon starting a 15 under 15 fellowship wherein WhiteHat Jr will help the top 15 kids on the platform incubate their startups and provide a $15000 fellowship to them.","excerpt":"Metro cities are dealing with increasing traffic woes which also have a grave impact on emergency services like ambulances leading to a delay in admitting critical patients in the hospital on time. To overcome this challenge, Brinda Jain, a 7-year-old girl from Bangalore has created the Ambulance Whizz App that aims to create a faster […]","categories":["AI News"],"tags":["traffic management"],"author_name":"Sejuti Das","publish_date":"2020-03-05T13:04:05","publication_year":"2020","word_count":541,"keywords":["Go","API","machine learning","programming_languages:R","AI","venture capital","Aim","ViT","traffic management","R","startup"],"extracted_tech_keywords":["AI","machine learning","Aim","R","Go","API","ViT","startup","venture capital","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/7-year-old-bangalore-girl-creates-an-app-to-help-ambulances-reach-hospitals-faster\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10072214,"title":"Another bill bites the dust","content":"After dithering for over three years, the Centre on Wednesday finally withdrew the highly controversial Data Protection Bill 2019. The Bill sought to regulate how the government and companies could use the digital data of citizens. It had faced severe backlash from stakeholders – citizens, tech firms, and political parties – since inception and had undergone various changes over the years based on experts’ suggestions. The Union minister for electronics and information technology (MeiTY) Ashwini Vaishnaw noted that the now-withdrawn Bill had been “deliberated in great detail” by the joint parliamentary committee (JPC). The committee proposed 81 amendments and 12 recommendations to build a “comprehensive legal framework” for the digital space. The Modi government now intends to “present a new Bill”. Personal Data Protection Bill has been withdrawn because the JCP recommended 81 amendments in a bill of 99 sections. Above that it made 12 major recommendations.Therefore the bill has been withdrawn and a new bill will be presented for public consultation.— Ashwini Vaishnaw (@AshwiniVaishnaw) August 3, 2022 Never-ending delay As per media reports, Vaishnaw said that the new draft would go through the approval process very soon. He was hopeful that the new Bill would get passed in the Budget session of the Parliament. In an exclusive interview with Economic Times, retired Justice B N Sri Krishna (who headed the experts committee that brought out the draft for the Personal Data Protection Bill way back in 2018), commented that the government probably thought it was wiser to withdraw now, “rethink the entire Bill” and come out with a new one with proper provisions. The Bill has been doing the rounds for a long time now and prompt action is the need of the hour to put a proper framework for data protection in place. Apar Gupta, executive director, Internet Freedom Foundation, tweeted: “It has been close to 10 years since the AP Shah on Privacy, 5 years since the Puttaswamy judgement and 4 years since the Srikrishna Committee’s report — they all signal urgency for a data protection law and surveillance reforms. Each day lost, causes more injury and harm.” In the making for half a decade now It all started with the monumental 2017 Justice Puttaswamy judgment by the Supreme Court that identified privacy as a fundamental right. During the same time, MeiTY constituted an expert committee to study what kind of data protection issues the country was facing and recommend steps and procedures to tackle them. This committee was headed by Supreme Court judge (retired) Justice B N Srikrishna. The committee submitted the report draft titled ‘A Free and Fair Digital Economy – Protecting Privacy, Empowering Indians’ to IT minister Ravi Shankar Prasad in mid-2018 and proposed the Personal Data Protection Bill-2018. A template for the developing world The report highlighted that if India is to shape the global digital landscape in the 21st century, it needs to formulate a legal framework relating to personal data which can work as a “template for the developing world”. The committee came up with several recommendations on the methods to be followed to process personal data of citizens by companies (both global and Indian). The stress was on user consent. The report also mentioned the penalties that can be imposed for violation of the rules. Image: *Data_Protection_Committee_Report.pdf (meity.gov.in) In 2019, the Bill was introduced in the Lok Sabha and met with strong criticism. The Parliament formed a JPC to look into the Bill minutely. After a whole two years and many changes later, the JPC report was  tabled during the winter session of the Parliament in late 2021. Major Roadblocks The Bill has received flak for various provisions – a common grievance being that it gave too much power to the government with respect to users’ personal data. Another provision that attracted criticism was Article 35, which could allow the Centre to exempt any government agency from the provisions of the proposed law “in the interest of India’s sovereignty and integrity, the state’s security, friendly relations with foreign states, and public order”. Image: *Personal_Data_Protection_Bill,2018.pdf (meity.gov.in) Image:https:\/\/www.meity.gov.in\/writereaddata\/files\/Personal_Data_Protection_Bill,2018.pdf Last year, Congress MPs Jairam Ramesh and Manish Tewari added their dissent notes criticising the exemptions given to the government. TMC MP Derek O Brien, too, had voiced his dissatisfaction. Today, the Joint Parliamentary Committee on Personal Data Protection Bill 2019 will adopt its report. I’m compelled to submit a detailed dissent note. But that should not detract from the democratic manner in which the Committee has functioned. Now, for the debate in Parliament. pic.twitter.com\/tavBnF9y5B— Jairam Ramesh (@Jairam_Ramesh) November 22, 2021 Image: A Public Brief on the Data Protection Bill, 2021 (internetfreedom.in) Tech giants were displeased with the Bill The old Bill had proposed strict regulations on data flows from India to other countries and gave the government the power to seek user data from companies, as a way to maintain its grip over them. Image:Asia-Internet-Coalition-AIC-Submission-on-Joint-Parliamentary-Committees-Report-on-the-Personal-Data-Protection-Bill_25Jan22-updated.pdf (aicasia.org) The Bill was not at all well received by tech giants. They were miffed with the provision under the name “data localisation” that asked the companies to maintain a copy of certain sensitive personal data within India. It also prohibited export of critical personal data to another country. Google expressed the said concerns in 2020. It had told the JPC on the Personal Data Protection Bill that India should avoid data localisation requirements.","excerpt":"The Bill had faced heavy criticism from different stakeholders -citizens, tech firms, political parties since its inception","categories":["AI Features"],"tags":[],"author_name":"Sreejani Bhattacharyya","publish_date":"2022-08-05T14:00:00","publication_year":"2022","word_count":886,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","Git","R"],"extracted_tech_keywords":["AI","R","Go","Git","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/another-bill-bites-the-dust\/","complexity_score":3,"technical_depth":6,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10093382,"title":"Zerodha’s Journey from Skepticism to AI Optimism","content":"In 2021, Indian fintech Zerodha’s CTO Kailash Nadh criticised superfluous, outsized “We’re powered by AI\/ML” marketing which companies were running after. But the recent advancements in AI has altered the hobbyist software developer’s opinion as the company now deviated from its previously held ‘no AI approach’. Cut to present, isn’t Zerodha doing the exact same thing? To this, the self-taught coder Nadh said: “Marketing specifically, not usage.” Clarifying the same, he said companies have been legitimately using AI where they made sense for a long time. But the hype was fuelled by companies doing name-sake or bogus implementations and claiming to be powered by AI\/ML.” Nadh has been researching and playing around with language models since the GPT-3\/Copilot launch. But he agrees that the chat interface of course broke the flood gate, made it far easier to tinker and test, and also understand its potential impact on the world as large, which its viral adoption illustrates beautifully. He further stated that Zerodha, founded in 2010 hadn’t found any use cases for AI\/ML technologies up until the recent breakthroughs. “We might use where it makes sense, but the technologies have become so commoditised to integrate (literally takes minutes) that my earlier argument about superfluous marketing stands validated. Everyone will start using it and the claim will become meaningless,” he added. While the fintech company has very recently announced the integration of AI in their organisation, its competitors like TradeStation and Groww on the other hand have been using AI chatbot to automate manual workflows since half a decade ago. Is Zerodha late to the AI party? Two years ago, when AIM asked Nadh if Zerodha is using AI\/ML, he said that it uses little to no AI or ML apart from some basic image\/document recognition ML models for document processing. Now, Zerodha seems to be on the side of the fence, experimenting with GPT-4 and alike. “We’re just experimenting and doing small pilots. Text summarisation is a big use case for this industry. For example getting quick summaries of multi-page legal documents,” explained Nadh. With the currently identified use cases using a specific set of technologies, the company has estimated the tasks done by at least 20% (~200 people) can be fully replaced by AI automation. However, Nadh said the people at Zerodha needn’t be worried outright because the team has factored in this risk and created a policy that provides assurance. We’ve just created an internal AI policy @zerodhaonline to give clarity to the team, given the AI\/job loss anxiety. This is our stance:\"We will not fire anyone on the team just because we have implemented a new piece of technology that makes an earlier job redundant.\" 1\/8— Nithin Kamath (@Nithin0dha) May 12, 2023 Though Zerodha seems to have taken a humane approach, the tech goliath IBM’s CEO Arvind Krishna said the company expects to pause hiring for roles it thinks could be replaced with AI in the coming years. Roughly 7,800 jobs could be replaced in the next five years as per the exec. The Fear of AI Replacing Humans Still Looms The fear of AI replacing humans is not new. At the recent Web Summit in Rio De Janeiro, AI Guru and founder of SingularityNET, Ben Goertzel vocally addressed that in the coming years AI could replace 80 percent of human jobs. But, how will we know in the future that the companies are not laying off employees because of performance or AI\/automation basis? Nadh believes that layoff of human resources is neither a new problem, nor an AI-problem. A lot of people who are laid off never know the “real” reason behind those decisions. They could be easily the result of human bias. Either way, this issue, AI or non AI, can only be solved via organisational transparency. Management has to commit to being transparent about these decisions he suggested. Ergo, AI policy. Nadh opined, many companies will likely let go of employees and blame it on AI. In the process, companies will earn more and make their shareholders wealthier, worsening wealth inequality. “This isn’t a good outcome for humanity,” he added. What’s the solution? A report by Goldman Sachs economists surfaced on the internet last month that stated generative AI could replace up to one-fourth i.e 300 million of current jobs globally. But the researchers listed workers in China, Vietnam and India are among the least likely to fall prey to the impact. “It’s still important to understand the implications. It will affect existing opportunities obviously, but new ones might emerge,”  Nadh said. After the company released its policy, Nadh took to his blog to pen down the reasons why the company is finally going against its AI-first mindset. In the post titled ‘This time, it feels different’ Nadh stated, ‘Neither blockchain, serverless, web3, big data, nor earlier AI\/ML technologies brought this about. But, the specific breakthroughs in the past few months finally did. All it took was 30 minutes to integrate, during which, it generated the code to integrate itself.’ He further noted that his excitement for these developments is overshadowed by growing fear. “Since there is so much that is unknown and alien about this, apart from, ‘let’s think this through and be very careful, I don’t know what advice can be given,” he concluded.","excerpt":"Zerodha’s CTO, Kailash Nadh, candidly discusses with AIM the company’s recent transition to harness AI technologies, emphasising that it is far from a mere marketing gimmick.","categories":["AI Trends"],"tags":["AI automation","AI Jobs","FinTech","IBM","Kailash Nadh","Zerodha"],"author_name":"Tasmia Ansari","publish_date":"2023-05-15T18:30:00","publication_year":"2023","word_count":881,"keywords":["big data","Go","AI automation","AI","ML","AI Jobs","Git","serverless","GPT","Kailash Nadh","Aim","IBM","generative AI","Zerodha","FinTech","R"],"extracted_tech_keywords":["AI","ML","generative AI","Aim","serverless","R","Go","Git","big data","GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/zerodhas-journey-from-skepticism-to-ai-adoption\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10168925,"title":"Amazon Launches Project Kuiper Satellites to Take on Starlink","content":"Amazon has launched its first batch of 27 satellites as part of Project Kuiper,  an initiative aimed at increasing global broadband access through a constellation of over 3,000 satellites in low Earth orbit. The project’s mission is to deliver fast, affordable broadband to underserved communities worldwide, competing against SpaceX’s Starlink. It initially faced delays due to manufacturing slowdown and last-minute launch cancellations. In a market dominated by SpaceX, this move by Amazon is a significant development. The satellites were deployed 280 miles (450 kilometres) above Earth during this mission, a significant upgrade from the two prototype satellites successfully tested during the Protoflight mission in October 2023. Amazon confirmed that the satellites are successfully activated and communicating with ground systems. “We’ve designed some of the most advanced communications satellites ever built, and every launch is an opportunity to add more capacity and coverage to our network,” Rajeev Badyal, VP of Technology for Project Kuiper, had said before the launch. According to a Reuters report, the company faces a deadline set by the U.S. Federal Communications Commission to deploy half its constellation by mid-2026. The satellites are coated with a dielectric mirror film unique to Kuiper that scatters reflected sunlight, helping to make them less visible to ground-based astronomers. “Following KA-01, we will continue to increase our production, processing, and deployment rates as we prepare to begin delivering service to customers,” Amazon said. “We have already begun shipping and processing satellites for our next mission: KA-02 will also use a ULA Atlas V rocket and launch from Cape Canaveral Space Force Station.”","excerpt":"We’ve designed some of the most advanced communications satellites ever built, Rajeev Badyal, VP of technology for Project Kuiper","categories":["AI News"],"tags":["Amazon","satellites India"],"author_name":"Amisha Arya","publish_date":"2025-04-30T13:18:16","publication_year":"2025","word_count":260,"keywords":["satellites India","AI","programming_languages:R","Amazon","RAG","Aim","R"],"extracted_tech_keywords":["AI","Aim","RAG","R","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/amazon-launches-project-kuiper-satellites-to-take-on-starlink\/","complexity_score":2,"technical_depth":5,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10112148,"title":"AWS, T-Hub Partner to Launch India&#8217;s First Spacetech Accelerator Programme","content":"Amazon Web Services (AWS) India Private Limited recently announced the AWS Space Accelerator: India, a technical, business, and mentorship opportunity designed to foster startups focused on space technology, and accelerate their development, with support from, T-Hub, and Minfy. This is AWS’s first accelerator programme in India focused on startups in the space sector, and follows the MoU it signed with ISRO and IN-SPACe in September last year, which envisions to nurture startups in space-tech, and support innovation in the sector. The accelerator was launched in the presence of Shri Sudheer Kumar N, Director, Capacity Building and Public Outreach, ISRO; and representatives from AWS, T-Hub, and Minfy. The 14-week accelerator program will provide India-based startups dedicated business resources, expertise, and guidance around technology and business capabilities, and help them leverage AWS to build, grow, and scale their missions. The programme is open to both early-stage and mature startups based in India whose missions support the space sector, who need a technology platform for their ideas, and are looking to accelerate their growth and investment opportunities. Interested startups can learn more and register at this link by March 17th, 2024. “Cloud computing is crucial to power the future of the space industry. It enables space-tech startups to harness high-performance computing resources to perform intensive data analysis, apply machine learning models, and innovate in their missions, while achieving lower cost of operations, faster time-to-market, and deployment at scale. Technologies such as generative AI open transformative opportunities for space-tech, and we are committed to help startups innovate and develop sustainable solutions for the space sector,” Clint Crosier, Director, Aerospace and Satellite, AWS Inc., said. Startups qualifying for the program can receive an array of benefits, including: up to US$100,000 in AWS credits, as well as guidance from experts at ISRO, IN-SPACe, and AWS, and business and technology support from T-Hub and Minfy, an AWS Premier Partner. They will also learn about space domain technologies and applications, and how to leverage space data and build scalable solutions using cloud computing, data analytics, and machine learning, including the potential to apply generative artificial intelligence (AI).","excerpt":"Startups can leverage AWS’cloud infrastructure to build, grow, and scale their missions.","categories":["Deep Tech"],"tags":["AWS","AWS cloud"],"author_name":"Pritam Bordoloi","publish_date":"2024-02-07T11:43:37","publication_year":"2024","word_count":349,"keywords":["artificial intelligence","machine learning","AWS","AI","cloud computing","RAG","Ray","analytics","generative AI","AWS cloud","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","analytics","generative AI","Ray","RAG","cloud computing","AWS","R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/aws-t-hub-partner-to-launch-indias-first-space-tech-accelerator-programme\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":68434,"title":"How Sindhu Gangadharan Is Prioritising Business Continuity For SAP Labs","content":"Over two decades, SAP Labs has fueled numerous innovations across organisations and industries. SAP Labs India is the company’s biggest R&D centre outside headquarters in Germany and a leading hub for innovation and a hot spot for talent development. SAP Labs India also makes a significant contribution to SAP’s important solutions – to name a few – ERP & Finance (SAP S\/4 HANA, Business ByDesign, Supply Chain & Manufacturing, Industries) and Business Technology Platform (SAP Cloud Platform, Database and Data management, Data Warehouse, Application Development\/Integration, HANA Enterprise Cloud, Analytics, Intelligent Technologies). Recently, Analytics India Magazine interacted with Sindhu Gangadharan is Senior Vice President and Managing Director of SAP Labs in India to further know the company’s growth plans in India. Sindhu joined the company two decades ago, at a time when the company had just set up its operations in Bengaluru. Within SAP, Sindhu has held various leadership positions and played a pivotal role in shaping SAP’s Integration strategy by growing SAP’s Integration product portfolio as Head of Product Management for SAP Cloud Platform Integration, IoT and SAP Process Orchestration. In this interview, Sindhu also shares her views on SAP’s technology initiatives for clients, the ongoing research and development projects, business continuity and reskilling programs for the organisation’s employees. AIM: How has COVID-19 impacted SAP Labs in India? Sindhu Gangadharan: At SAP Labs, our primary focus continues to be the health and well-being of our people. We have made significant investments in crisis management to tackle any challenge and respond to disruptive incidents such as COVID-19. For the majority of our teams, work can continue remotely which enables teams to work off-site while staying safe at home. Our teams are accustomed to working from home as this is a long-standing practice we follow here, so the transition was seamless. Our engineers across the globe are working separately wherever possible to mitigate the risk of virus and minimise the impact. Employees needed on-site are following social distancing and hygiene protocols based on guidance from health authorities and our Global Pandemic Task Force. AIM: Which specific technologies, in your opinion, will play a critical role in the fight against COVID-19 and why? Sindhu Gangadharan: Technology, if used properly, has a significant role to play during any crisis such as the one caused by COVID-19 pandemic. With specific reference to fight against COVID-19, we can look at the intervention of technology in three major areas. There are technologies for detecting and ensuring social distancing. Using GPS tracking and mobile apps, social distancing and contact tracing can be ensured. AI-powered non-contact sensor systems and thermal imaging can help identify people with fever, who can be potential COVID-19 carriers. Likewise, unmanned aerial vehicle technology like drones can be effectively used to pinpoint with great accuracy scan and deliver in crowded and unapproachable areas. Then there is technology for non-pharmaceutical interventions. Agent-based simulation modelling helps prevent a second wave of infection. Along with Tata Institute of Fundamental Research and Indian Institute of Science, SAP has developed a ‘City-Scale Epidemic Simulator’ to help evaluate scenarios of phased emergence from the lockdown, once the restrictions are lifted. Similarly, through the power of analytics, solutions can be developed that provide accurate information drawn from multiple sources to help the Government and all people connected, to provide relief or lessen the impact of this pandemic. SAP has contributed to the development of the analytics dashboard that NASSCOM has offered to the Government of Karnataka and Telangana to aid the control of COVID-19. Finally, there is technology for vaccine development: However, in the process of vaccine development, innumerable data from multiple sources like clinical trials, research development and many are to be made to generate insights, which is possible only through advanced analytics and natural language processing techniques. Having quick access to the insights and the sources of data can help speed up vaccines and the drug development process. AIM: Also, tell us about SAP Labs’ tools for remote working and collaboration. Sindhu Gangadharan: SAP Labs India has always been a frontrunner in the industry in creating employee-friendly policies. Keeping our workforce’s convenience as a priority, we were one of the first few companies to implement a flexible working option. Also, to safeguard our employees, we took proactive measures and asked them to work from home much before the lockdown was imposed by the Government. At SAP, we use the best collaboration tools to provide ease of remote working through applications for communication, group work and data backup. Our teams connect regularly through daily scrums, team meetings and collaborate virtually to ensure that our business continuity is not impacted. To keep our employees motivated and stay positive, we have introduced SAP Come Alive initiative, which delivers online fun, wellness and motivational programs. We are sharing tips and tricks from employees globally for effective remote working. To also keep the personal connection going, teams are engaging during breaks with virtual coffee sessions and volunteering in programs that are focused on giving back to society. AIM: What are various research & development projects that you are currently working on? Sindhu Gangadharan: SAP Labs India drives SAP’s product strategy, and it is responsible for SAP core solutions, provides product localisation and India specific solutions. Unique to SAP Labs India is that it represents the entire breadth of SAP product portfolio and contributes to SAP flagship products such as SAP HANA, the platform for next-generation applications and analytics. SAP S\/4HANA, the real-time ERP suite for digital business holds key product responsibilities across the customer engagement, human capital management, analytics and IoT portfolios. Moreover, our teams have been focusing on enabling businesses to evolve as ‘intelligent enterprises’. At SAP, our effort is centred on technologies being the enabler to bring business impact for customers. Our vision is on driving forward the intelligent enterprise transformation and marrying it with our whole suite of products and solutions. Because now, it’s not just about transforming a segment but rejigging the whole business infrastructure, which is possible with an intelligent enterprise strategy. AIM: What are SAP Labs’ initiatives focused on maintaining business continuity? Sindhu Gangadharan: Due to COVID-19 pandemic, customers are learning to adapt to the new norm and are seeking technical support to accelerate and stabilise their growth journeys. We have developed business continuity plans to make sure our services are not disrupted worldwide. Since 77% of the world’s transactions touch an SAP system, we need to make sure that the business systems of our customers are always operational. To ensure business continuity during these challenging times, we are actively engaging with our workforce and customers. Our employees are well-equipped with all the necessary infrastructure to work remotely. Our IT policies are framed in such a way that remote working is never a challenge for employees. Technology has been a game-changer, especially during these times and enabling many organisations including SMEs to ensure business continuity. Our focus has also been working towards providing training programmes to SMEs to optimise their supply chain and manage their finances, procurement and stakeholders. We have been supporting startups with infrastructure, access to experts and product experts who have been in the business for a long time, and providing them with coaching and mentoring. We will continue to do this. AIM: How does SAP India aim to benefit the business community with its new initiatives? Sindhu Gangadharan: SAP has a clear purpose of helping the world run better and improving people’s lives. To empower our customer and smaller businesses, we’ve made a few offerings available immediately, at no cost. Our Remote Work Pulse by Qualtrics is available which is designed to help employees stay connected. The platform enables organisations to understand how their employees are doing and what support they need as they adapt to new work environments. We have SAP Litmos offering a completely free Remote Readiness & Productivity Academy training to support businesses which are having to work remotely. The video-based courses are designed to help establish best practices for remote work and develop leadership during times of change and challenge. Ruum by SAP which is a lightweight project management and collaboration tool to optimise processes is accessible to anyone. It has created custom checklist templates, adapted from the Centres for Disease Control and Prevention to help businesses plan, prepare, and respond to COVID-19. Additionally, we had also opened access to SAP Ariba Discovery for 90 days to help buyers and suppliers to connect and fulfil their sourcing needs with minimum disruption in time of this crisis. AIM: Which skills are you building internally at SAP Labs to keep the innovation going forward in 2020? Sindhu Gangadharan: Employees are the centre of any organisation, and they are the greatest assets for us at SAP. We have always fostered an inclusive culture and collaborative learning, which ensures all our employees are ready to face the current and future challenges in the technology landscape that they encounter. Recently, we announced a new digital learning initiative offering innovative, interactive educational content to support students, professionals and anyone wishing to continue to learn during this challenging time. This dynamic initiative is based on three educational pillars – massive open online courses (MOOCs), learning journeys for universities and the SAP Young Thinkers program all the offerings are accessible through the openSAP platform, our most popular online learning initiative. The openSAP platform provides MOOCs about leading technologies, the latest innovations and the digital economy. Course topics include automated robotic process automation, data science, machine learning, ethical artificial intelligence, IoT, sustainability, Java programming and more. We also have SAP Success Map which is a one-stop digital experience with 1.3 million opportunities available in the system, and employees can register their learning needs and find solutions. SAP Enable Now gives access to employees who use SAP solutions to the training they need anytime and anywhere with context-sensitive user help, training simulations, test scripts, and e-learning materials making them ready to face any technology challenge. Additionally, there are other programs such as the peer learning culture with award-winning coaching and mentoring programs – free for all employees and Beyond-the-Job experience with opportunities for new work experiences through fellowships and job rotations.","excerpt":"Over two decades, SAP Labs has fueled numerous innovations across organisations and industries. SAP Labs India is the company’s biggest R&D centre outside headquarters in Germany and a leading hub for innovation and a hot spot for talent development. SAP Labs India also makes a significant contribution to SAP’s important solutions – to name a few […]","categories":["AI Features"],"tags":["cloud business intelligence solutions","how to implement business intelligence","Interviews and Discussions","iot friendly database","SAP"],"author_name":"Vishal Chawla","publish_date":"2020-06-29T14:00:00","publication_year":"2020","word_count":1690,"keywords":["how to implement business intelligence","data science","cloud business intelligence solutions","Go","artificial intelligence","machine learning","AI","SAP","ML","Aim","analytics","iot friendly database","R","Java","Interviews and Discussions"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","data science","analytics","Aim","R","Go","Java"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-sindhu-gangadharan-is-prioritising-business-continuity-for-sap-labs\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10089515,"title":"Midjourney AI Version 5 &#8211; What to Expect?","content":"Midjourney announced the release of version 5 of its commercial product. Midjourney ai v5 is said to be capable of producing photorealistic images at a quality level that some AI art enthusiasts have described as “creepy” and “too perfect”. The new version is available as an alpha test for Midjourney subscribers via Discord. Midjourney v5 generated image of a kitten sleeping on a desk The developers of Midjourney v5 have made significant improvements in various areas, including aesthetics, quality, and details. The latest version boasts a wider stylistic range, understands user requests more accurately, and can produce higher-quality images instantly—thanks to twice as much resolution and a wider dynamic range. The new version also excels in small details, such as the number of fingers on each hand and facial detailing. Developers have claimed to have made significant improvements in facial detailing, making it easier to create well-known figures from popular culture. As per a Twitter user, “Midjourney’s AI can now do hands correctly. Be extra critical of any political imagery (especially photography) you see online that is trying to incite a reaction”. Midjourney v5’s neural network uses default styles to create photorealistic output when no style is specified. This feature has grown in significance in recent years, with the mention of artists’ or movements’ names becoming increasingly important. The new version also supports seamless textures, wider aspect ratios, and better image prompting, making it easier to produce a more diverse range of outputs. Midjourney hopes to encourage its community to test Midjourney v5 and explore the new features it has to offer. The v5 was launched earlier for select users which, according to Midjourney, wasn’t performing well as the images were coming out blurry. However, Midjourney claims that after disabling it, the issues have been now fixed.","excerpt":"The v5 was launched earlier for select users which, according to Midjourney, wasn’t performing well as the images were coming out blurry.","categories":["AI News"],"tags":["MidJourney"],"author_name":"Lokesh Choudhary","publish_date":"2023-03-17T12:35:40","publication_year":"2023","word_count":297,"keywords":["MidJourney","TPU","programming_languages:R","AI","neural network","ML","RAG","Aim","AI art","R"],"extracted_tech_keywords":["AI","ML","neural network","Aim","RAG","TPU","R","AI art","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/midjourney-gets-more-realistic-with-its-version-5\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":9631,"title":"Cognitive Retail: The future is here","content":"Cognitive technology is going to impact practically everything and it has glimpses of future that we have probably not even dreamt of. Getting back to the significant impact of Cognitive on various industries – none of them is going to remain untouched. One of the prime industries that are going to simply metamorphose will be that of retail. Now, with all the barcodes, QR codes, fancy packaging, billing, online and on-call customer care services, we have been quite content with the way retail works. But is this the limit? Has technology nothing else to offer? Thinking beyond the automated menus and self-help kiosks where you play with little machines to get information and more – cognitive is going to take it to another level and all your science fantasies will be soon coming alive. The future holds treasures of technology that we will soon explore and witness! Technology companies all across the world have already started investing in this segment which verifies that this is indeed promising and soon to become a reality. The significance and the crucial role that will be played by AI and IOT has been well understood and acknowledged by top notch experts all across the world. To understand what is going to happen in the future with the grace of the dear cognitive, let’s divide our retail experience into steps: When we realize we need a product When we shop for it When we experience the product When we realize we need the product (Pre-sale experience) How do we really realize we need it? We see it somewhere – at a friend’s place or in an advertisement (print or electronics) and we are like, damn! I need it! What would you do next? Try to figure out where you can get more information about it and perhaps, where you can get the best deal for it. Let’s assume that you are at a friend’s place and you see a washing machine and you suddenly realize that you need one too. All you would need to do is click a picture and input it in the device – be it your cell phone, tablet or laptop. The device will get back to you with all kind of information you need and you will be able to know the credentials about the product along with all the technical specifications. Being advanced, modern and particular – we would like to research about it next. No need to bother anymore as your device is doing it for you already. It is finding youtube videos, showing reviews and telling you what people on social media are talking about this. Technology will become your guide and understand your needs. It will ask you questions in your preferred language so that the best possible product can be suggested. For example, ‘How many people are there in your house?’ ‘What kind of fabric you mostly wear?’ The human touch will be imparted without humans being involved in a very subtle manner. Not only this, the device will also advise you where you can buy it for cheapest and nearest. If you need it immediately, it will be advisable to visit the nearby electronic or appliances store. Did you know that your computer has already checked if the particular washing machine was in stock or not and then alerted you? Or maybe if you don’t need it so early, you can buy it online. At the same time, read reviews and know how other people have rated it. Apart from what people talk on social media, there are always experts available who are giving genuine and credible reviews about it. You are going to get it all right in front of you – what any expert in any corner of the world is talking about the washing machine that you have your eyes upon. When we shop for it (Sales experience) If you want to buy it right now, you have already been alerted for the nearby store and within seconds, the directions are there in front of you along with the traffic status. You don’t feel like driving till there? The bus numbers and metro routes are there for you and if it is just a walk away, the computer will advise you to that considering your current health conditions. While we reach the shop and the device helps us find what we want, it can also suggest a list of things that are available and can be bought along. This will be suggested according to your previous buying patterns and the feasible combinations. Now while you are buying a washing machine, you obviously need washing powder too and wait maybe an iron along with it? And of course, all the offers and combos! The device will oodle information like buying a water purifier along with this washing machine of the same company is going to give you a great discount. While you are eying on a washing machine your device will help you know what similar products you can buy from the same store! While you have picked up the washing machine along with several other products maybe, how do you want to pay for it? What are the cashback or discount offers that you can avail with all the debit and credit cards in your wallet? Do you also get some extra discount if you use a particular one? Or maybe you have already spent too much on your cards and it is advisable to pay in cash. Now, the washing machine might need a warranty, then what are the packages that you can go for to have a better warranty strategy for your product. Considering the history of how you deal with machines or products, your device can accordingly suggest the kind of warranty pack you must avail. While you have ordered it – offline or online – where has it reached? The app on the device will inform the delivery people about the time you are usually at home so that you can receive it yourself. If you have to leave urgently for a vacation or work, it can inform the same to the delivery people and you can always remain well informed about where the package has actually reached. When we experience the product (Post-Sales Experience) The impact of cognitive is not just limited to buying, but it also amplifies when you start using the product. Do you know how to set this up and what all these buttons do? Do you have the instruction manual? Do you even have basic skills to do it or some youtube videos might prepare you better? If there is an instruction manual, which language it is? Do you need help with translation? Infact the power of cognitive is that your device would have already read the manual and will provide you with step-by-step instructions on how to use the washing machine for best results. Once you have started using the washing machine, your device can also prompt you to write reviews about it so that other people can also know if you really liked it and cognitive techniques can help them decide better based on the interests. If you are writing positive reviews, you definitely deserve to earn some bonus points right? Maybe you shopped from this particular site for the first time and you must be rewarded. It can also prompt you to allow the device to suggest it to people in your circle based on interest and use. And if they end up buying it too, you must flourish with more rewards! How many days had it been that you have been using that Washing Machine? The washing machine will inform the company about the usage and before it gets too late, you will have someone at your doorstep to do the basic repairing so that it can function smoothly. If some part of the machine is not working properly, it can be replaced before it gets faulty and stops working. If you don’t like it and in the following days, you are not using it much – your device can actually suggest you to resell it and find people who have been interested in buying similar second-hand products. According to your usage, preferences and interest – the device can also suggest what can be bought further to improve the experience further. Internet of Things (IoT) becoming in-charge of your health! Now while you have your washing machine, it is observing the frequency with which you use it. Accordingly, it can predict the water consumption and family members and their age groups. Also, your air purifier is observing the level of CO2 and O2 and other pollutants; along with that, your water purifier knows how much water your family consumes. With the help of data from IoT, all these machines can interact and prepare a health scorecard for your family. If there is any irregularity in the patterns, the device can alert you so that you can check on it. For example, if the drinking water consumption becomes less, maybe you will be advised to get back to your normal drinking water consumption as it can have a bad effect on your life. Similarly, if you smoke and the level of pollutants has been increasing, it can again alert you to reduce the smoking and live a healthier life. To be summed up, whatever you would think, the future of retail is looking very bright and it is definitely going to open new avenues and provide its customers a treat with multiple new possibilities and get a very convenient and better shopping experience.","excerpt":"Cognitive technology is going to impact practically everything and it has glimpses of future that we have probably not even dreamt of. Getting back to the significant impact of Cognitive on various industries – none of them is going to remain untouched. One of the prime industries that are going to simply metamorphose will be that […]","categories":["IT Services"],"tags":["retail analytics"],"author_name":"Shailendra Kumar","publish_date":"2016-04-20T04:58:42","publication_year":"2016","word_count":1604,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","R","retail analytics"],"extracted_tech_keywords":["AI","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/cognitive-retail-the-future-is-here\/","complexity_score":2,"technical_depth":5,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10099315,"title":"Genpact Unveils 12-Week Immersive Learning Program to Strengthen Gen AI Developer Community","content":"Genpact has launched an innovative 12-week immersive learning program aimed at fostering the growth of the Gen AI developer community. This strategic initiative seeks to bolster Genpact’s Gen AI delivery engine, enabling it to achieve unprecedented speed and scale in the rapidly evolving landscape of artificial intelligence. The program, designed to elevate the capabilities of carefully selected Data-AI professionals spanning various experience bands, is poised to revolutionise the field. With prerequisites encompassing a solid understanding of machine learning concepts, neural networks, algorithm optimisation, and proficiency in Python programming, the program ensures a strong foundation for the journey ahead. The very first batch comprising 300 professionals have embarked on their learning journey and many more to come! Empowering through Expertise: A Unique Blend of Learning Central to this immersive program is an innovative learning approach. The curriculum is powered by Udemy, a renowned platform for online education, offering a comprehensive and expert-curated syllabus coupled with hands-on labs. Over an 8-week period, participants will delve into topics essential for mastering Generative AI and Large Language Models (LLMs), starting from the fundamentals of deep learning and culminating in advanced topics and techniques that underpin these cutting-edge technologies. The curriculum includes dedicated modules on Generative Adversarial Networks (GANs), Variational Autoencoders (VAEs), Sequence Models for Text Generation, Large Language Models and Finetuning, and the intricacies of structuring and deploying applications. From Learning to Application: Bridging the Gap The program goes beyond theoretical understanding, emphasising practical application. Following the 8-week Udemy learning path and weekly guru connects, participants will undergo rigorous knowledge and skill assessments to solidify their grasp of the material. This will be followed by a pivotal 4-week capstone project phase, where participants will engage in real-world Gen AI Proof of Concepts (POCs). This hands-on experience will bridge the gap between theory and practice, enabling participants to apply their newfound knowledge to tangible, industry-relevant scenarios. Culmination and Future Pathways As participants complete the program, they will emerge as proficient developers, well-versed in harnessing Generative AI and Large Language Models to solve complex challenges. Graduates of the program will not only join the growing Gen AI developer community but also contribute to advancing Genpact’s AI-driven solutions at an accelerated pace. With an ever-increasing demand for AI-driven innovations across industries, Genpact’s immersive learning program represents a significant leap forward. It equips participants with the tools, knowledge, and confidence to drive AI-powered transformation across various sectors, establishing themselves as trailblazers in the dynamic world of artificial intelligence.","excerpt":"Driving rapid advancements in Gen AI delivery engine through an immersive talent build intervention.","categories":["AI News"],"tags":["genpact AI"],"author_name":"AIM Media House","publish_date":"2023-08-31T14:39:32","publication_year":"2023","word_count":410,"keywords":["Go","machine learning","artificial intelligence","AI","neural network","genpact AI","Python","Aim","deep learning","generative AI","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","deep learning","neural network","generative AI","Aim","Python","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/genpact-unveils-12-week-immersive-learning-program-to-strengthen-gen-ai-developer-community\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10096940,"title":"Twitter Now Pays &#8216;Blue Verified&#8217; Creators, Will Threads Follow Suit?","content":"Twitter today announced the launch of its ‘Creator Ads Revenue Sharing Program’ for creators subscribed to Twitter Blue or Verified Organizations. Twitter said that content creators will now have the opportunity to earn a share of the revenue generated from ads displayed within their replies. ​​https:\/\/twitter.com\/Twitter\/status\/1679572360695824384 To be eligible for this program, creators must be verified users and have garnered at least 5 million impressions in the last three months. Once Twitter approves a creator’s eligibility, they will be required to have a Stripe account to facilitate payment processing, the company added. In February, Musk had announced that Twitter will start sharing revenue from reply-thread ads with creators subscribed to Twitter Blue Verified. In April, Twitter rolled down its legacy Verification Program which led to previous verified accounts on the basis of previous criteria (active, notable and authentic) to lose their blue checkmark. In a recent announcement, Musk also introduced a limit on the number of posts a user can read per day on a platform. The rationale behind implementing this restriction, as explained by Musk, was to tackle the issue of excessive data scraping and manipulation of the system. Who is Copying Whom? With this, Twitter is trying to draw more content creators into its ecosystem. This comes shortly after the introduction of Threads, which is being promoted as a potential “Twitter Killer” application by Zuckerberg. Notably, Threads achieved a significant milestone by surpassing 100 million sign-ups within just five days, surpassing even the rapid growth observed with OpenAI’s ChatGPT. It would be interesting to see if Threads will also start “Creator Ads Revenue Sharing Programme,” as it previously seems to have inspired everything that Elon’s Twitter has been doing to make the platform better, including the paid blue tick, and to an extent, the entire platform as well. However, Meta previously had introduced something similar to Twitter ad revenue sharing on Instagram reels for select creators, where they could earn money by posting reels.","excerpt":"This comes shortly after the introduction of Threads, which is being promoted as ‘Twitter Killer’","categories":["AI News"],"tags":[],"author_name":"Siddharth Jindal","publish_date":"2023-07-14T13:01:23","publication_year":"2023","word_count":326,"keywords":["ChatGPT","API","OpenAI","AI","RPA","GPT","GAN","R","llm_models:ChatGPT","llm_models:GPT"],"extracted_tech_keywords":["AI","ChatGPT","OpenAI","R","API","GPT","GAN","RPA","llm_models:GPT","llm_models:ChatGPT"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/twitter-now-pays-blue-verified-creators-will-threads-follow-suit\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10045390,"title":"Hands-On Guide to Generating Artificial Faces Using Progressive GAN","content":"Nowadays, Generative adversarial networks, in short, GAN, are the effective ways to train data for computer vision-based applications. Moreover, it can be used to create a synthetic dataset to fulfil the requirements. A typical GAN architecture uses two models for training: Generator and Discriminator; the generator is used to output a synthetic image. The discriminator model is used to check whether the image is real or fake based on the discriminator output the generator model is trained on. To achieve proper equilibrium, both the models are trained in an adversarial manner. If you are new to GAN, I recommend this article to have a proper understanding of GAN. When it comes to large datasets with higher pixel values, GAN generates the images with sharp pixels that look crispy though make the training unstable. Generating high-resolution images is a challenging task because the generator must know the details and structures involved in images. The high-resolution images can cause any issues that the discriminator can easily spot; therefore, the whole training process fails. The solution for this behaviour of GAN is training the network by progressively adding layers. Progressive GAN is an extension of standard GAN that holds the generator in stable mode while dealing with large images to achieve better stable performance. Methods involve starting with very small images such as 4 x 4-pixel images and successively adding blocks of layers that increase the size of images to 8 x 8, 64 x 64 until the desired size, as shown in the picture below. This makes Progressive GAN capable of generating high pixel images such as 1024 x 1024 images. Image Source As we can see from the above image, during training, the new convolutional layer is added to both generator and discriminator models; this makes the whole model effectively learn deep detailing of the pixels and finer level pixels simultaneously. Today in this article, we will implement the Progressive GAN model using TensorFlow and see how this model can be used to generate artificial faces. Implementation of Progressive GAN: The below demonstration uses the Tensorflow model based on GAN, which maps the N-dimensional Latent space vector into an RGB image. The below code shows the mapping of a latent space vector to the image and Generating the target image using gradient descent which gives the latent vector. Here the latent vector is simply a representation of the compressed data in which similar data points are closer to each other in space. This is important because the GAN model takes points from the latent vector and generates an image based on it. The below code is part of the official implementation of Progressive GAN in Tensorflow. Install & Import all dependencies: # imageio for creating animation !pip -q install imageio !pip -q install scikit-image !pip install git+https:\/\/github.com\/tensorflow\/docs from absl import logging import imageio import PIL.Image import PIL.Image import matplotlib.pyplot as plt import numpy as np import tensorflow as tf tf.random.set_seed(42) import tensorflow_hub as hub from tensorflow_docs.vis import embed import time try: from google.colab import files except ImportError: pass from IPython import display from skimage import transform Helper functions: The model uses the latent dimension as a multiple of 512. Here in this section, we create a function for displaying the image, animating the image to see the changes, and the interpolation function to create new pixels. latent_dim = 512 def interpolating_vectors(v1, v2, num_steps): v1_n = tf.norm(v1) v2_n = tf.norm(v2) v2_normal = v2 * (v1_n \/ v2_n) vect = [] for step in range(num_steps): interpol = v1 + (v2_normal - v1) * step \/ (num_steps - 1) interpol_norm = tf.norm(interpol) interpol_normal = interpol * (v1_n \/ interpol_norm) vect.append(interpol_normal) return tf.stack(vect) def image_display(img): img = tf.constant(img) img = tf.image.convert_image_dtype(img, dtype=tf.uint8) return PIL.Image.fromarray(img.numpy()) def animation(img): img = np.array(img) converted_images = np.clip(img * 255, 0, 255).astype(np.uint8) imageio.mimsave('.\/animation.gif', converted_images) return embed.embed_file('.\/animation.gif') Latent space interpolation: Here we are using the random vectors for the interpolation; for that, we are using the TF-hub module, which consists pretrained model of Progressive GAN. progan = hub.load('https:\/\/tfhub.dev\/google\/progan-128\/1').signatures['default'] def interpolating_between_vect(): v1 = tf.random.normal([latent_dim]) v2 = tf.random.normal([latent_dim]) vect = interpolating_vectors(v1, v2, 150) interpolated_images = progan(vect)['default'] return interpolated_images interpolated_images = interpolating_between_vect() animation(interpolated_images) See the image generated by Progressive GAN by randomly generated vectors; Finding closet vector in latent space: Here we try to generate a target image using the latent space vector; we can also upload our image by changing the  image_from_module_space to False. inside_image = True def from_module_space(): vector = tf.random.normal([1, 512]) images = progan(vector)['default'][0] return images def upload_image(): uploaded = files.upload() img = imageio.imread(uploaded[list(uploaded.keys())[0]]) return transform.resize(img, [128, 128]) if inside_image: target_image = from_module_space() else: target_image = upload_image() image_display(target_image) As I have chosen to use the image from the module, here is our target image; Below we are generating our starting image based on that model tries to converge to the target image. We need to define the loss function between the target image and the latent space image. Then, we can use gradient descent to find variables that minimize loss. tf.random.set_seed(4) initial_vector = tf.random.normal([1, latent_dim]) display_image(progan(initial_vector)['default'][0]) Below is our starting image; Now get the closest latent point for the target image; def closest_vetor(initial_vector, num_optimization_, steps_per_img): images = [] losses = [] vector = tf.Variable(initial_vector) optimizer = tf.optimizers.Adam(learning_rate=0.01) loss_fn = tf.losses.MeanAbsoluteError(reduction=\"sum\") for step in range(num_optimization_): if (step % 100)==0: print() print('.', end='') with tf.GradientTape() as tape: image = progan(vector.read_value())['default'][0] if (step % steps_per_img) == 0: images.append(image.numpy()) target_image_difference = loss_fn(image, target_image[:,:,:3]) regularizer = tf.abs(tf.norm(vector) - np.sqrt(latent_dim)) loss = target_image_difference + regularizer losses.append(loss.numpy()) grads = tape.gradient(loss, [vector]) optimizer.apply_gradients(zip(grads, [vector])) return images, losses num_optimization_=200 steps_per_img=5 images, loss = closest_vetor(initial_vector, num_optimization_, steps_per_img) Results: Check out the animated image, which shows how the model has converged to our targeted image; and check the result side by side left-sided is generated image, and right-sided is the target image animate(np.stack(images)) display_image(np.concatenate([images[-1], target_image], axis=1)) Conclusion: From this article, we have seen how Standard GAN architectures lag when handling large pixel values. But the Progressive GAN comes to fulfil the task; we see how progressively adding the layers helps the generator function remain stable throughout the operation and generate a reasonable image. References: Official documentation for Progressive GANOfficial implementation on TensorflowLink for above code","excerpt":"When it comes to large datasets with higher pixel values, GAN generates the images with sharp pixels that look crispy though make the training unstable. Generating high-resolution images is a challenging task because the generator must know the details and structures involved in images. The high-resolution images can cause any issues that the discriminator can easily spot; therefore, the whole training process fails.","categories":["Deep Tech"],"tags":["AI (Artificial Intelligence)","AI Tool","Deep Learning","GAN model","Google","Guide","image-generation","Machine Learning"],"author_name":"Vijaysinh Lendave","publish_date":"2021-08-08T13:00:00","publication_year":"2021","word_count":1022,"keywords":["Guide","NumPy","TPU","AI","ML","image-generation","Machine Learning","computer vision","Colab","Ray","Matplotlib","Python","Google","GAN model","Deep Learning","AI Tool","TensorFlow","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","ML","computer vision","Ray","TensorFlow","Colab","NumPy","Matplotlib","TPU","Python"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/hands-on-guide-to-generating-artificial-faces-using-progressive-gan\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10048127,"title":"Tableau 2021.3 Is Here &#8211; Latest Features &#038; Updates","content":"Tableau’s easy-to-use user interface and loaded features have made it a popular tool amongst the community, and even a non-technical user can easily work on it. It recently released an update to its 2021 version, the Tableau 2021.3. Released on 7 September 2021, the new update aims to bring more efficiency and tools to enhance workflow. Here are some exciting features from the latest update: Tableau Slack Notifications Tableau has integrated its visualization prowess with Slack, and together, it will broaden not only the aspects of analytics but also the flow of businesses. With the release of Tableau 2021.3, notifications can now be directly received via Slack. Slack will directly notify when a set goal or threshold is reached to either an individual or a whole team. Another new feature is the mentions update. Whenever one is ‘@’ mentioned, Slack will send out notifications through which one can access the dashboard or reach out to the commenter directly in Slack. Insights can be shared directly to help collaborate on analytics with just the click of a button. Image Source : Siawash Emami via LinkedIn Better Data Management with Tableau Prep If you have numerous missing rows in a data table, Tableau can now help you fix that. Using the Tableau Prep feature, data can now be automatically filled to prevent missing rows based on the provided dates, date times, or integer. This allows the downstream processes to have complete datasets to work on, and this aids in building more accurate visualizations. Tableau Prep also features options to automate multiple flow jobs through linked tasks and ensure its sequence so that every task completes successfully right after the other. Improvements to Tableau Catalog Tableau will also allow pop-up notifications, or email subscriptions, for Data Quality warnings to help identify the data’s potential issues. The receiver has to simply click the link provided in the email to go directly to the previously impacted dashboard, and additional details can be seen on the Data Details panel. With Tableau Catalog, we can now see the inherited description within the web authoring flow for the fuller context of the data. If a particular description for a field or column exists upstream from that field or column in that data source or workbook, the information about where the description has been inherited from can be seen. Introducing Personal Space Personal Space is a newly introduced private location available on Tableau Server and Online, where the users can save content before it is ready to be shared with others. A user can explore data and staging workbooks in their own dedicated and confidential space, and admins would now no longer have the need to create private projects for each and every individual on-site. Personal Space will be available to every user holding an Explorer license or above. This dedicated space for drafting content makes the site cleaner for other users and easier to find relevant content. Create your Personalised Start Experience If you have used the previous versions of Tableau, the Tableau Start Page has been a favorite. From Superstore to Regional and World Indicator sample workbooks, the start page features everything you need. With the 2021.3 update, you can now set customized workbooks you want to display in the samples area to match the specific needs. Users will now be able to access content that is relevant to their product instantly. New users will be able to get up and running with sample data and pre-populated workbooks, distribute corporate data sources, create specific highlights for the business, and bring together teams setting fun challenges. The options to customize are endless and full of possibilities! Increased trust, transparency and support Coming later this year will be the new security features that will help increase flexibility and segmentation. Administrators will be allowed to centrally configure which users and groups have access to which particular slices of data. With centralized row-level security, also known as CRLS, Tableau admins can define and manage row-level security features among the data tables present and apply the same to all the connected Tableau flows, data sources and workbooks that depend on that data. Through virtual connections, users can create and share access to different tables through a database connection, embed service account credentials, and extract data from data tables centrally to reuse within the Tableau Server and Online space. Calendar support improvements for the ISO-8601 calendar have also been extended to more databases, such as the DB2 connector. When connecting to Salesforce data using the custom SOQL, multiple objects at once can be accessed using the SOQL relationship, using which fields from multiple objects in a single query can be pulled at the same time. Updated Tableau Mobile New updates to Tableau Mobile have also been added. The new improvements bring along an improved notification experience, new notifications tab to see shares, comments and prep flows, all in one place. This consolidates all the important changes across the organisation. Tableau Mobile now supports Blackberry and Airwatch devices to provide quick and secure ways to leverage data. The new app includes a fresh design that is faster and more optimised, new ways to search and browse with Tableau Server, and Online as well as offline capabilities that are quick to load and offer more interactivity for visualisations. The new Tableau 2021.3 looks to be an update loaded with features and functionalities that will only increase the productivity within the tool. Every aspect of a perfect visualization tool is covered, and it seems that it will take Tableau to newer heights of popularity. You can download the latest version of Tableau using the link here.","excerpt":"Tableau recently released an update to its 2021 version, the Tableau 2021.3.","categories":["AI Trends"],"tags":["data quality","Data Science","Slack","Tableau","updates"],"author_name":"Victor Dey","publish_date":"2021-09-10T11:00:00","publication_year":"2021","word_count":940,"keywords":["Go","Tableau","Rust","AI","RAG","Aim","ViT","data quality","updates","analytics","GAN","Data Science","R","Slack"],"extracted_tech_keywords":["AI","analytics","Aim","RAG","R","Go","Rust","data quality","GAN","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/latest-features-introduced-in-tableau-2021-3\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":39916,"title":"Hybrid &#038; Multi-Cloud Environment Is A Trillion Dollar Market, &#038; IBM Intends To Seize It: Vikas Arora","content":"While the AI-powered systems have shown exponential growth and success in the past few years, they have been facing unparalleled security threats that have not been seen before in the traditional software space. To deal with these issues, IBM along with its large teams of scientists, engineers and developers are dedicated to creating novel defence techniques and deploying practical defences of real-world AI systems. Analytics India Magazine got in touch with Vikas Arora, IBM Cloud and Cognitive Software Leader, IBM India\/South Asia, who also leads the IBM’s security business and key transformation projects along with IBM’s cloud and cognitive business. In this candid chat, Arora spoke about the cybersecurity landscape in India, need of cognitive security operation centres, the role of AI in security and more. Analytics India Magazine: How has IBM brought key transformations in the security business with the help of AI? Vikas Arora: IBM is leading the journey towards AI and Intelligent Automation in Cybersecurity. In 2016, IBM invested heavily on the acquisition of Resilient Systems – the leading platform for orchestrating and automating the incident response processes The latest step in this journey is the automation of response – connecting machine intelligence and human expertise together more seamlessly across the entire threat lifecycle. In 2017, IBM brought the power of Watson to the cybersecurity market to help augment the skills of security analysts in their investigations (Watson for Security \/ QRadar Advisor with Watson) In April 2018, IBM introduced new innovations that move the needle towards a new era where machine intelligence and human expertise are orchestrated seamlessly together across the Security Operation Center In October 2018, IBM announced new features for QRadar Advisor which expand the platform’s knowledge of cybercriminal behaviour and allow it to learn from security response activities within an organization. AIM: What does your role as cloud and cognitive software leader at IBM include to enable security domain at IBM? VA: With a shift towards cloud, companies are looking for capabilities that enable the enterprise to integrate across public, private and on-premises environments. They are also looking for the ability to manage workflows across public clouds from multiple vendors, build cloud capabilities and have consistent management protocols. Customers regularly ask for help on modernizing their security frameworks, respond to the global security skills shortage, address increasing cyber-attack vectors including IoT, secure the journey to the cloud and digital transformation and maintaining data privacy and regulatory compliance. We have integrated products and services, global expertise and leading-edge orchestration and AI tools to help them here. AIM: How has the role of AI in security evolved over the years in India? What are the key innovations that you have witnessed? VA: The state of cybersecurity is reaching an inflection point. The number of risks and events is growing exponentially, and security operation teams are struggling to keep up with the volume. Security leaders are worried about how security incidents affect their operations today and how they may shape their reputations tomorrow. Organisations are now looking to cognitive\/AI security solutions to manage this situation and help address gaps in intelligence, speed and accuracy. Even though it is in the early days, there is great hope and optimism about its potential. It may enable improved detection and decision-making capabilities, improve incident response time, among others. However, there is still a lot of education and preparation that has to happen before widespread adoption occurs for AI-based security solutions. AIM: What is the kind of security threats that AI-powered systems are susceptible to in the current scenario? How is IBM helping overcome these threats? VA: As with any new technology, cybercriminals will also be looking for ways to use AI to their advantage. For instance, cybercriminals are constantly looking to leverage emerging technologies – both to compromise it as an attack vector or to help their current attacks become more widespread or effective. We as the defenders need to be looking at the threats surrounding AI to stay ahead of cybercriminals. This can be done by using AI to detect advanced threats and help security teams and examining how AI can be weaponized by cybercriminals to make their attacks more successful. At IBM, we have experts focused on exploring both sides of this equation. AIM: Would you like to highlight a few instances where IBM has provided security solutions to the client? VA: For instance, Bombay Stock Exchange (India), the oldest stock exchange in Asia and now the fastest exchange in the world, has selected IBM Security to design, build and manage a Cyber Security Operations Center to safeguard the company’s assets and protect stakeholder data. Under the five-year managed security services agreement, the centre enables around-the-clock security event monitoring, event handling, security analysis, incident management and response along with synchronized management of devices, networks and applications. AIM: What are some of the novel defence techniques and AI-based solutions by IBM in the security space? VA: We are leading the journey towards AI and Intelligent Automation in Cybersecurity. Some of our key AI, Resilient and Automation solutions are: IBM QRadar Advisor with Watson: Leverages the power of cognitive AI to automatically investigate indicators of compromise and gain critical insights, accelerating your threat response times. IBM Resilient Incident Response Platform: Orchestrates and automates hundreds of time-consuming, repetitive and complicated response actions that previously required significant human intervention. IBM MaaS360 Advisor with Watson: Provides cognitive insights, contextual analytics and benchmarking to make sense of security events –while protecting your endpoints, users, apps, docs and data from one platform. IBM Application Security on Cloud: Automates code reviews with AI to help eliminate false negatives and false positives, so you can proactively secure your web, mobile and cloud-based applications. AIM: How has been the adoption of cognitive security solutions in India? How do you anticipate it growing in the next couple of years? VA: We believe that the adoption of AI\/Cognitive solutions within Security space is still at a nascent stage, however, continues to be an important strategy. IBM is leading the journey towards AI and Intelligent Automation in Cybersecurity. We foresee both being a key priority for not just CISOs but the entire C-Suite in 2019. Enterprises are looking at driving service agility and resilience in their digital business along with data-driven security intelligence which can help them be prepared for any unforeseen threats. AIM: What is IBM’s security business roadmap in the coming year? What are the key initiatives that you plan to take up in the cybersecurity space? VA: IBM Security is one of the fastest growing vendors in the global security software market. We have simplified our portfolio to help clients in three strategic areas, “Strategy & Risk”, “Threat Management” and “Digital Trust”. Further, we are focused on top security needs we hear from our customers around advanced threats, cloud security, Mobile and IoT, Compliance Mandates and addressing the skills shortage in the security space.  We are already the largest Security provider to the Enterprises in India, and we aim to strengthen this position as we drive our Security strategy in 2019. AIM: What is your strategy to boost the hybrid cloud market play in India? VA: Chapter one of the cloud was largely about moving a lot of new and customer-facing applications to the cloud. Chapter two is about scaling AI and creating hybrid clouds. It’s about bringing the cloud operating model to all those mission-critical apps and enabling customers to manage data, workloads, and apps and move them between multiple clouds. The result is a complex web of hybrid, multi-cloud environments. Studies estimate that while organisations plan to adopt hybrid architectures, very few will have procedures and tools they need to operate that environment. To deal with this, IBM launched Multi-cloud Manager last year which gives businesses visibility and control across different clouds and vendors. Recently, we also launched new hybrid cloud tools and services designed to help enterprises navigate the complexities of this new landscape such as IBM Cloud Integration Platform and IBM Services. The other important development that will allow IBM to help companies accelerate our journey to the hybrid cloud is the fact that IBM is in the process of acquiring Red Hat. Hybrid, Multi-cloud environment is a trillion dollar opportunity and IBM intends to be #1. AIM: How do you plan to level up competition in the cloud space in India? VA: Cloud is emerging as the essential construction site for the cognitive enterprise. Companies are moving beyond their initial cloud deployments and realising its potential to help optimise existing business applications, and launch new business services, with levels of speed and cognitive innovation previously impossible. As we go through 2019, organisations would be looking for advice on selecting the right cloud solution, help them move to cloud, create the modern cloud application suites and help businesses manage new multi-cloud environments as they build them. To address these needs we have devised IBM Cloud strategy as “Journey to Cloud 14 in 4 offerings” to support in these four areas, and beneath them, there are 14 offerings that bring our services and solutions to life.","excerpt":"While the AI-powered systems have shown exponential growth and success in the past few years, they have been facing unparalleled security threats that have not been seen before in the traditional software space. To deal with these issues, IBM along with its large teams of scientists, engineers and developers are dedicated to creating novel defence […]","categories":["AI Features"],"tags":["Interviews and Discussions","private cloud"],"author_name":"Srishti Deoras","publish_date":"2019-05-30T11:06:41","publication_year":"2019","word_count":1512,"keywords":["Go","AI","R","ML","Git","RAG","Aim","analytics","Rust","GAN","private cloud","Interviews and Discussions"],"extracted_tech_keywords":["AI","ML","analytics","Aim","RAG","R","Go","Rust","Git","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/hybrid-multi-cloud-environment-is-a-trillion-dollar-market-ibm-intends-to-seize-it-vikas-arora\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10005354,"title":"Within The Realm Of AI Plus AR, How Are Mobile Applications Changing?","content":"Augmented Reality (AR) will deeply affect businesses across all industries, impacting the way we learn, make data-driven decisions and communicate with the physical world. Machine learning is a crucial determinant in pushing the AR industry forward. In the AR world, ML is being utilised to determine the detection problem based on camera tracking. Several large tech companies such as Google, Microsoft, Facebook and Amazon are leading the development of the underlying technology and integrating AI with AR for various use cases. The next-gen of AR can create much more personal and intimate experiences, with the computing environment connecting digital objects in the real world which users can both interact with and be present together. Companies such as Facebook have been extremely focused on creating technology to shape the next generation of computing to make it more human-designed and around the ways that we all naturally interact with each other. AI and ML In Augmented Reality In the last years alone, billions of people have used AR features on social media platforms, including Facebook and Snapchat. Facebook provided support for spark AR studio with operating systems such as Windows and macOS and opened its AR for Instagram for everyone to build apps for it. Companies are now using ML for tasks such as inferring approximate 3D surface geometry to allow visual effects, needing only a single camera input without the requirement for a dedicated depth sensor, and more. Other areas where AI plus AR have been explored are car insurance companies where you can walk up to any car, hold your phone up to it and it will identify the make and model of the vehicle. It would connect to the company’s APIs and then tell you the rate and monthly payment you would be eligible for. How Can Machine Learning Be Integrated Into AR For using machine learning in augmented reality apps, there are several pre-trained ML models that can be used. For example, ResNet and others are AI models optimised for computer vision task object detection. These models are designed to track classes of objects and not just one particular object. For applications in the context of augmented reality, there are three levels of image processing for which machine learning is used. First is image classification which tells you what is in the image, second is object detection to draw a bounding box around the image, and finally, image masking where you can actually get an exact outline of the objects in an image. Now, suppose you can have image masking work on mobile with an existing AR SDK that does ground plane detection. In that case, you can infer the position of the object in 3D space for object occlusion or adding colliders for physical interactions. There are native solutions for running ML models on Android and iOS, and for image masking or object detection, we can only know spatial information about the detections in 2D space. To add AR technology along with AI models, developers are building applications that involve physical interaction with image objects and a 3D space. The technology is also open source. So, for example, let’s say across all the tech companies you want to use Google for this. You can train a model against the existing model using transfer learning with the Google cloud platform. TensorFlow Lite, which is an open-source deep learning framework for on-device inference, is incredibly useful for building AR apps. Especially if you are looking to maximise the performance on your smartphone to run a machine learning model. And then, of course, there is MLKit which helps to create an app without necessarily having to write their own model. There are also APIs which provide resources such as OCR, face detection, and other capabilities already pre-built for users. There are also tools such as ARCore developed by Google which is  cross-platform and uses OpenGL. It’s a light wrapper around OpenGL which can perform tasks like motion tracking and scene building. There is also SceneForm, which is a specific SDK for Android that saves you from having to learn OpenGL. With advancements in machine learning there are many ways available to integrate into the AR systems. It requires minimal effort to get things up and running. These can be run directly on the devices or run through cloud services as well.","excerpt":"Augmented Reality (AR) will deeply affect businesses across all industries, impacting the way we learn, make data-driven decisions and communicate with the physical world. Machine learning is a crucial determinant in pushing the AR industry forward. In the AR world, ML is being utilised to determine the detection problem based on camera tracking. Several large […]","categories":["AI Features"],"tags":["augmented reality"],"author_name":"Vishal Chawla","publish_date":"2020-08-22T21:04:31","publication_year":"2020","word_count":721,"keywords":["Go","machine learning","AI","augmented reality","ML","Git","computer vision","deep learning","object detection","TensorFlow","R"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","computer vision","TensorFlow","object detection","R","Go","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/within-the-realm-of-ai-plus-ar-how-are-mobile-applications-changing\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":13722,"title":"Spreadsheets &#038; Corporate Data &#8211; Managing Spreadsheet Series: 1 of 5","content":"For a hundred reasons, the spreadsheet have become a ubiquitous commodity in the corporate landscape. Perhaps the most pervasive of these reasons is that the spreadsheet allows the end user – the non-technician – to have control over his\/her technical destiny. The spreadsheet is malleable, controllable and inexpensive. Those traits allow the end user to take control. The IT person is not needed in order to build and control the spreadsheet. There are other reasons why the spreadsheet is so pervasive, but autonomy of processing by and for the end user has to be at the head of the list. There are many types of spreadsheets. Fig 1 shows some of the more common types of spreadsheets – Some spreadsheets are very personal. A very personal spreadsheet may have the football scores of your favourite team. Or the names of the children you are going to invite to your child’s birthday party. There is NO business import to a very personal spreadsheet. Some spreadsheets are business informal. In a business informal spreadsheet, you may keep track of inventory or expenses or days off for sickness or personal time for people working in your department. While there is business import for these spreadsheets, the informality of the business spreadsheet makes the usage of the spreadsheet unusable beyond the immediate walls of the department. Then there are formal business spreadsheets. These formal business spreadsheets may cover quite a bit of ground. A formal business spreadsheet might be for corporate revenues, corporate expenses, customer purchases, inventory, and so forth. Indeed, the different types of spreadsheets can be arranged in a spectrum, as seen in Fig 2 – Fig 2 shows that there is a continuum of spreadsheets all across the corporation. The spectrum goes from informal to formal spreadsheets. As the formality of the spreadsheets increases so increases the nonrecurring\/ recurring nature of the spreadsheets. Most really informal personal spreadsheets are nonrecurring. And the more formal and business-related the spreadsheet, the more the likelihood that the spreadsheet is recurring. And as spreadsheets become formal and recurring, the spreadsheets tend to contain more business value. Fig 3 shows where business value starts to appear in the spectrum of spreadsheets. There is a lot of potential business value to the spreadsheets found to the right of the spectrum. But the greatest business value is derived when the spreadsheets are turned into corporate data. When the business value oriented spreadsheets are turned into corporate data, there are several major benefits to the organization. The first benefit is that the organization can start to use the spreadsheet data on a widespread basis. Fig 4 shows this benefit – When a spreadsheet that has business value is turned into corporate data, it is open for usage by anyone that needs to see the data. As long as the spreadsheet remains just a spreadsheet, only a few people and a few departments can use the spreadsheet. But when the spreadsheet is turned into corporate data –a standard database — people can access and use the data base across a wide spectrum of users. So the first value in turning spreadsheet data into corporate data is that of widespread accessibility throughout the corporation. But, there is a second major advantage to turning spreadsheet data into corporate data, and that benefit is that when there is a need to communicate information from one company to the next, the spreadsheet becomes the lingua franca. Fig 5 shows that spreadsheets are an effective way to communicate information from one corporation to another – When one corporation has a need to transmit information to another corporation, the spreadsheet become the means by which the communication is made. Bill Inmon – the “father of data warehouse” – has written 57 books published in nine languages. Bill was named by ComputerWorld as one of the ten most influential people in the history of the computer profession. Bill lives in Castle Rock, Colorado. Bill’s latest book is TURNING TEXT INTO GOLD, Technics Publications, a book that shows how text can be turned into business value. TURNING TEXT INTO GOLD is available on Amazon.com.","excerpt":"For a hundred reasons, the spreadsheet have become a ubiquitous commodity in the corporate landscape. Perhaps the most pervasive of these reasons is that the spreadsheet allows the end user – the non-technician – to have control over his\/her technical destiny. The spreadsheet is malleable, controllable and inexpensive. Those traits allow the end user to […]","categories":[],"tags":[],"author_name":"William Inmon","publish_date":"2017-03-24T05:33:00","publication_year":"2017","word_count":686,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","ViT","GAN","R","data warehouse"],"extracted_tech_keywords":["AI","R","Go","data warehouse","GAN","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/spreadsheets-corporate-data-managing-spreadsheet-series-1-5\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":60644,"title":"Top Research Papers On Causal Inference","content":"As researchers pursued the inevitable AGI in machines, there has been a renewed interest in the idea of causality in models. There are significant implications to applying machine learning to problems of causal inference in fields such as healthcare, economics and education. Here are a few top works that acknowledge the challenges and offer solutions to the causal inference in machines: The Seven Tools Of Causal Inference 2018 In this paper, Judea Pearl who has championed the notion of causal inference in machines, argues that causal reasoning is an indispensable component of human thought that should be formalized and algorithimitized towards achieving human-level machine intelligence. Pearl, in this paper, analyses some of the challenges in the form of a three-level hierarchy, and shows that inference to different levels requires a causal model of one’s environment. He has also described seven cognitive tasks that require tools from those two levels of inference. Check paper here. A Causal Bayesian Networks Viewpoint on Fairness 2019 In this paper by DeepMind, the researchers offer a graphical interpretation of unfairness in a dataset as the presence of an unfair causal path in the causal Bayesian network representing the data-generation mechanism. They have used this viewpoint to point out that fairness evaluation on a model requires careful consideration on the patterns of unfairness underlying the training data. They also show that causal Bayesian networks can function as a powerful tool to measure unfairness in a dataset, and to design fair models in complex unfairness scenarios. Check paper here Causal Inference And The Data-fusion Problem 2016 The authors address the problem of data fusion—piecing together multiple datasets collected under heterogeneous conditions. These conditions can be different populations or sampling methods. The diversity of these datasets offer new opportunities for better insights. However, there is always a challenge of bias seeping into these datasets. In this work, the authors present a general, nonparametric framework for handling these biases and, ultimately, a theoretical solution to the problem of data fusion in causal inference tasks. Check paper here. Reinforcement Knowledge Graph Reasoning for Explainable Recommendation 2019 Personalized recommendation with the help of knowledge graphs has been gaining traction lately. In this work, the authors perform explicit reasoning with knowledge for decision-making, so that the recommendations are generated and supported by an interpretable causal inference procedure. They propose a method called Policy-Guided Path Reasoning (PGPR), which couples recommendation and interpretability by providing actual paths in a knowledge graph. Our contributions include four aspects. Their experiments on several large-scale real-world benchmark datasets led to obtaining favorable results compared with state-of-the-art methods. Check paper here Double\/Debiased Machine Learning for Treatment and Causal Parameters 2016 Supervised machine learning (ML) methods are explicitly designed to solve prediction problems very well. However, causal parameters can behave very poorly due to regularization bias. And, this regularization bias, assert the authors, can be removed by solving auxiliary prediction problems using machine learning tools. The authors discuss an orthogonal score that combines auxiliary and main ML predictions, which is then used to build an estimator for debiasing of the target parameter and be approximately unbiased and normal. They call this method a ‘double ML’ method and claim that it can be used on a broad set of ML predictive methods, such as random forest, lasso, ridge, deep neural nets, boosted trees, as well as various hybrids and aggregators of these methods. Check paper here Causal Regularization 2017 Causal interpretability of predictive models in domains such as healthcare models is critical. To facilitate such interpretability, the authors propose a causal regularizer to steer predictive models towards causally-interpretable solutions. Their analysis on a large-scale Electronic Health Records (EHR) shows that their causally-regularized model outperforms its L1-regularized counterpart in causal accuracy, and is competitive in predictive performance. They have also demonstrated that the proposed causal regularizer can be used together with neural representation learning algorithms to yield up to 20% improvement over multilayer perceptron in detecting multivariate causation, a situation common in healthcare, where many causal factors should occur simultaneously to have an effect on the target variable. Check paper here Unbiased Scene Graph Generation 2020 Traditional machine learning debiasing methods cannot distinguish between good and bad bias. Example, good context prior (‘person read book’ rather than ‘eat’), and bad long-tailed bias (‘near’ dominating ‘behind \/ in front of’). In this paper, the authors present a novel framework based on causal inference and build a causal graph to perform traditional biased training with the graph. This framework, the authors claim, is agnostic and thus, can be widely applied in the community who seeks unbiased predictions. Check paper here","excerpt":"As researchers pursued the inevitable AGI in machines, there has been a renewed interest in the idea of causality in models. There are significant implications to applying machine learning to problems of causal inference in fields such as healthcare, economics and education.  Here are a few top works that acknowledge the challenges and offer solutions […]","categories":[],"tags":["causal inference","judea pearl"],"author_name":"Ram Sagar","publish_date":"2020-04-01T17:00:44","publication_year":"2020","word_count":766,"keywords":["knowledge graphs","Go","machine learning","programming_languages:R","AI","RPA","ML","Aim","ViT","causal inference","judea pearl","R"],"extracted_tech_keywords":["AI","machine learning","ML","Aim","knowledge graphs","R","Go","ViT","RPA","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/top-research-causal-inference\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10001088,"title":"Are Quantum Computers A Threat To Cryptocurrency?","content":"There are often talks revolving around the quantum computing revolution and its advantages to the computing world. However, there is also a downside to quantum computers. Quantum computers pose a threat to cryptocurrencies, and they could introduce changes that the cryptocurrency owners will not like. A Real Threat Quantum computers, with their unique property of quantum entanglement of particles, are very secure modes of information and it would be impossible to break it. It uses something called as a quantum key distribution (QKD) that helps them to keep the security intact. Cryptocurrency work involves a private key that is used to access Blockchain and Bitcoin contained on it. Each transaction has its own cryptographic hash. The currency owner will generate two numbers, one of which acts like a ‘private key’ and the other one as a ‘public key’. The public key can be easily generated from the private key but the opposite cannot be done. There is a signature used that uses a technique called elliptic curve signature scheme, for the receiver to ensure that the owner is the one that has the private key. These two things give cryptocurrencies the security that it has, which is impossible to break with the current traditional computers, since the only way to get hold of the private key is calculating it using the public key. Google’s quantum computing expert, John Martinis, is of the opinion that that even if his group could build a quantum computer in 10 years time, it would likely take a lot longer than that to break RSA. A quantum computer, due to Shor’s algorithm, could easily break the RSA encryption which is to secure data transmission on the Internet. Quantum computers may have the potential to reduce times of brute force attacks by providing the network with exponentially larger amounts of processing power, as opposed to the computers today, for whom the brute force attack would likely take multiple lifetimes to pose any relative threat to the network. Performance of a single quantum computer for blockchain attacks as a function of physical gate error rate Pg, which is an internal machine specification, and mining Difficulty D, which is set by the blockchain protocol. Image source: https:\/\/arxiv.org\/pdf\/1710.10377.pdf A study by Divesh Aggarwal at the National University of Singapore with his research team had studied the threat of quantum computers to Bitcoin. According to their paper, they found that the proof-of-work used by Bitcoin is relatively resistant to substantial speedup by quantum computers in the next 10 years, mainly because specialised ASIC miners are extremely fast compared to the estimated clock speed of quantum computers. It said, “The elliptic curve signature scheme used by Bitcoin is much more at risk, and could be completely broken by a quantum computer as early as 2027.” This research team from Singapore, Australia and France, in their paper, put forward that quantum computers could run fast enough to crack the cryptography on the network to steal cryptocurrencies. Tackling The Threat It is also pointed out that quantum computers do not pose any serious threat to the cryptocurrency security. Although quantum computer threat to the cryptocurrency is real, some experts believe that this quantum threat issue can be fixed. Jeffrey Tucker, Editorial Director at the American Institute of Economic Research, is one of the people with such an opinion. In his blog, he conveys that the fear of quantum computing as an existential threat to Bitcoin is based on scare-mongering and ignorance. Image source: https:\/\/arxiv.org\/pdf\/1710.10377.pdf This plot above, taken from their paper, shows two estimates of the hashing power, in hashes per second, which is a mathematical function that turns data into a specific number of random string numbers, of the Bitcoin network in blue striped curves versus a single quantum computer in red striped curves, as a function of time for the next 25 years. the black dotted line shows the hash rate of a single ASIC device today. Scientists from Victoria University of Wellington in New Zealand had done a research in which they proposed making blockchain behave like a quantum networked time machine to keep it safe. They showed that entanglement in time, as opposed to entanglement in space, provides the crucial quantum benefit. HCash, a startup of Australia is also creating solutions for protecting against quantum hacking and its researchers aren’t the only ones racing to ensure this sort of thing can’t happen. HCash CEO Dallas Brooks has said, “The great thing about HCash is that it’s preparing for the inevitable.” It is quite evident that we are already gearing up for the threat of quantum computing to cryptocurrency, even before the advent of the quantum computers.","excerpt":"There are often talks revolving around the quantum computing revolution and its advantages to the computing world. However, there is also a downside to quantum computers. Quantum computers pose a threat to cryptocurrencies, and they could introduce changes that the cryptocurrency owners will not like. A Real Threat Quantum computers, with their unique property of […]","categories":["AI Features"],"tags":["Bitcoin","Cryptocurrency","entanglement","Security"],"author_name":"Disha Misal","publish_date":"2019-02-03T19:18:35","publication_year":"2019","word_count":777,"keywords":["Go","Bitcoin","startup","programming_languages:R","AI","Cryptocurrency","programming_languages:Go","Security","ViT","R","emerging_tech:quantum computing","entanglement"],"extracted_tech_keywords":["AI","R","Go","ViT","startup","programming_languages:R","programming_languages:Go","emerging_tech:quantum computing"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/are-quantum-computers-a-threat-to-cryptocurrency\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":45888,"title":"Can Machines Benefit From Deja Vu? Role Of Replay Learning In Artificial Neural Networks","content":"Source: medicalnewstoday In the early 90s, a bunch of experiments were performed to find out how a living organism learns a task. In one such experiment, rats were made to run along the length of a single corridor or circular track, so researchers could easily determine which neuron coded for each position within the corridor. To the surprise of the researchers, these marked cells also appeared to be firing in the rest state as well. During rest, the cells sometimes spontaneously fired in rapid sequences demarking the same path the animal ran earlier, but at a greatly accelerated speed. These sequences are called replay. To test the significance of replaying sequences in learning a task, the researchers disturbed the brain activity during these replay events and found that the technique of experience replay allows the agent to repeatedly rehearse previous interactions, making the most of each interaction. By measuring memory retrieval directly in the brain, neuroscientists have noticed something remarkable: spontaneous recollections, measured directly in the brain, often occur as very fast sequences of multiple memories. These so-called ‘replay’ sequences play out in a fraction of a second–so fast that we’re not necessarily aware of the sequence. How Significant Is Learning Through Replay via gyfcat Imagine walking into a garden and coming across an apple on the ground, under a single apple tree. The fact that it has fallen from that tree is obvious and the possibility of the apple coming out of the ground doesn’t even occur to our mind. The reason why the apple doesn’t fly back to the tree is beyond the scope of this article. The obvious connections between the tree and the fruit are made by an average human. Replay doesn’t literally rehearse events in the order they were experienced. Instead, it infers or imagines the real relationships between events, and synthesises sequences that make sense given an understanding of how the world works. In AI terms, these replay sequences are generated using a learned model of the environment. The imagination theory makes a different prediction about how replay will look: when you rest on the couch, your brain should replay the sequence “tree, apple, ground”. You know from past experience that apples are more likely to fall from a tree than show up from the ground–and this knowledge can be used to reorganise experience into a more meaningful order. In deep RL, the large majority of agents have used movie-like replay, because it is easy to implement (the system can simply store events in memory, and play them back as they happened). Meanwhile, in neuroscience, classic theories of replay postulated that movie replay would be useful to strengthen the connections between neurons that represent different events or locations in the order they were experienced. The most compelling observation is that even when rats only experienced two arms of a maze separately, subsequent replay sequences sometimes followed trajectories from one arm into the other. via DeepMind To take this replay experiment to the next level, DeepMind in collaboration with Oxford and UCL have made few experiments. In these experiments, the subjects are shown a few scenes and then shown the scrambled sequence and then given five minutes to rest, while sitting in an MEG (magnetoencephalography) brain scanner. To find how the brain builds sequences, the researchers played a new sequence for participants. In this sequence, they walk into your factory and see spilt oil on the floor. They then see a knocked over an oil barrel. Finally, they turn to see a guilty robot. Common sense implies that the robot has spilt the oil in spite of not witnessing the act itself. This inference might come from the experiences such as those of having a kid or a dog at the house who spills water and then stands in a corner. The researchers found the use of some abstract codes between the two narratives. These abstract codes, which incorporate the conceptual knowledge that lets people unscramble the sequences, may help the brain to retrieve the correct item for the next slot in the replay sequence. This paints an interesting picture of a system where the brain slots new information into an abstract framework built from past experiences, keeping it organised using precise relative timings within very fast replay sequences. Can Machines Benefit From Replaying Experience? Research into experience replay has unfolded along parallel tracks in artificial intelligence and neuroscience, with each field providing ideas and inspiration for the other. Incorporating replay on computers has been beneficial to advancing AI. Deep learning often depends upon a ready supply of large datasets. In reinforcement learning, these data come through direct interaction with the environment, which takes time. The technique of experience replay allows the agent to repeatedly rehearse previous interactions, making the most of each interaction. Experience Replay is originally proposed in Reinforcement Learning for Robots Using Neural Networks in 1993. Experience Replay stores experiences including state transitions, rewards and actions, which are necessary data to perform Q learning. Q-learning is a model-free reinforcement learning algorithm. The goal of Q-learning is to learn a policy, which tells an agent what action to take under what circumstances. This method proved crucial for combining deep neural networks with reinforcement learning in the DQN agent that first mastered multiple Atari games. Deep Q-Network(DQN) is introduced in 2 papers, Playing Atari with Deep Reinforcement Learning on NIPS in 2013 and Human-level control through deep reinforcement learning in 2015. Since the introduction of DQN, the efficiency of replay has been improved by preferentially replaying the most salient experiences from memory, rather than simply choosing experiences at random for replay. Further improvements in agent performance have come from combining experiences across multiple agents, learning about a variety of different behaviours from the same set of experiences. Experiments such as the above show that a carefully constructed AI algorithm can reach superhuman performance and the potential of AI to enter the coveted arena of AGI looks higher than ever before. Read the original paper here.","excerpt":"In the early 90s, a bunch of experiments were performed to find out how a living organism learns a task. In one such experiment, rats were made to run along the length of a single corridor or circular track, so researchers could easily determine which neuron coded for each position within the corridor. To the […]","categories":["AI Trends"],"tags":["DeepMind","neuroscience"],"author_name":"Ram Sagar","publish_date":"2019-09-12T18:45:27","publication_year":"2019","word_count":1001,"keywords":["Go","API","artificial intelligence","AI","neural network","RAG","deep learning","ViT","neuroscience","GAN","R","DeepMind"],"extracted_tech_keywords":["AI","artificial intelligence","deep learning","neural network","RAG","R","Go","API","GAN","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/can-machines-benefit-from-deja-vu-role-of-replay-learning-in-artificial-neural-networks\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10004637,"title":"Comprehensive Guide To Learning Rate Algorithms (With Python Codes)","content":"Learning rate is an important hyperparameter that controls how much we adjust the weights in the network according to the gradient. The question most commonly asked in the field of Machine learning is “how do we know what is the right value for learning rate?” Unfortunately, there is no one size fits all answer to this question. But, I will put forth some of the methods you can use that can help you estimate what value should be used. This article covers the types of Learning Rate (LR) algorithms, the behaviour of learning rates with SGD and implementation of techniques to find out suitable LR values. Types of LR algorithms The learning rate algorithms are broadly classified into two categories: Constant Learning rate algorithm – As the name suggests, these algorithms deal with learning rates that remain constant throughout the training process. Stochastic Gradient Descent falls under this category. Here, η represents the learning rate. The smaller the value of η, the slower the training and adjustment of weights. But if the value is too high, the model converges too quickly and results in a suboptimal solution. Adaptive learning rate algorithm – Here, the optimizers help in changing the learning rate throughout the process of training. Adam, Adagrad, Adadelta, RMSProp are some examples of adaptive learning rate algorithms. For the purpose of this article, I will be using stochastic gradient descent to find the optimal learning rate. Stochastic Gradient Descent and Learning rate Stochastic Gradient Descent (SGD) is one of the most common optimizers used in machine learning. Let us see how SGD looks for a single sample. Take a look at the loss function below. where x is the input sample, y is the label, and θ is the weight. We can define the partial derivative cost function for a batch size equal to N as: In its most basic form, the SGD works by updating the value of ???? that moves the weights in the direction opposite to the gradient value of the loss. Unfortunately, when it comes to deep learning models, the reality is different. The local minima in case of smaller models are relatively shallow and are easy to get past. But, because of millions of parameters involved in the deep models, the local minima tends to be wider and thus creates a problem called plateaus. This is also called saddle because of how it looks. When this situation occurs, our learning model is stuck in the saddle and struggles to get out. One solution to this is fixing our learning rate large enough to escape the saddle. Let us now look at the methods that can be used. Reduce LR on Plateau: This is one of the ways of moving out of the saddle. Every time the loss begins to plateau, the learning rate decreases by a set fraction. When the loss function succumbs to higher learning rate and keeps moving around in the saddle, reducing the learning rate can help the loss find a smoother surface to escape this. LR Finder: In this method, the learning rate is selected by taking a random value for the weight, calculating the loss and getting a learning rate for that value. Next, a small step is taken and the learning rate is recalculated for the new weight and loss. This process is plotted in a graph and the optimal LR is selected. Cyclic Learning Rate: This method eliminates the need to experimentally find the best values and schedule for global learning rates. Instead of monotonically decreasing the learning rate, this method lets the learning rate cyclically vary between boundaries. Let us implement Cyclic LR and LR finder for CIFAR 10 to understand the difference and see the improvement in the accuracy. We will import the required libraries and load our data. from keras import backend as Kimport timeimport matplotlib.pyplot as pltimport numpy as np% matplotlib inlinenp.random.seed(2017) from keras import regularizersfrom keras.models import Sequentialfromkeras.layers.convolutionalimportConvolution2D, MaxPooling2D,AveragePooling2Dfrom keras.layers import Activation, Flatten, Dense, Dropoutfrom keras.layers.normalization import BatchNormalizationfrom keras.utils import np_utilsfrom keras.preprocessing.image import ImageDataGeneratorfrom keras.datasets import cifar10(train_features, train_labels), (test_features, test_labels) = cifar10.load_data()num_train, img_rows, img_cols,img_channels =  train_features.shapenum_test, _, _, _ =  test_features.shapenum_classes = len(np.unique(train_labels))class_names = ['airplane','automobile','bird','cat','deer','dog','frog','horse','ship','truck']fig = plt.figure(figsize=(8,3))for i in range(num_classes):    ax = fig.add_subplot(2, 5, 1 + i, xticks=[], yticks=[])    idx = np.where(train_labels[:]==i)[0]    features_idx = train_features[idx,::]    img_num = np.random.randint(features_idx.shape[0])    im = features_idx[img_num]    ax.set_title(class_names[i])    plt.imshow(im)plt.show() Our data has been loaded and saved in variables, let us normalize the data and convert them to categorical data. train_features = train_features.astype('float32')\/255test_features = test_features.astype('float32')\/255train_labels = np_utils.to_categorical(train_labels, num_classes)test_labels = np_utils.to_categorical(test_labels, num_classes) For simplification purposes, let us split our data into batches and augment the data using Image Datagenerator. (trainX, trainy), (testX, testy) = cifar10.load_data()datagen = ImageDataGenerator(featurewise_center=True, featurewise_std_normalization=True)datagen.fit(trainX)iterator = datagen.flow(trainX, trainy, batch_size=128)batchX, batchy = iterator.next()iterator = datagen.flow(trainX, trainy, batch_size=len(trainX), shuffle=False)batchX, batchy = iterator.next()print(batchX.shape, batchX.mean(), batchX.std())min_pix, max_pix = batchX.min(), batchX.max() Let us do the same for test data as well iterator1 = datagen.flow(testX, testy, batch_size=len(testX), shuffle=False)batch_testX, batch_testy = iterator1.next()X_train = batchXX_test = batch_testXy_train=batchyy_test=batch_testy Now, build a CNN model with batch normalization and regularization function (for faster convergence) and bear in mind to use SGD optimizer from keras import optimizersmodel1 = Sequential()model1.add(Convolution2D(32, 3, 3, border_mode='same',kernel_regularizer=regularizers.l2(0.0001), input_shape=(32, 32, 3)))model1.add(Activation('relu'))model1.add(BatchNormalization())model1.add(Convolution2D(64, 3, 3,kernel_regularizer=regularizers.l2(0.0001),border_mode='same'))model1.add(Activation('relu'))model1.add(BatchNormalization())model1.add(MaxPooling2D(pool_size=(2, 2)))model1.add(Dropout(0.2))model1.add(Convolution2D(32, 1, 1))model1.add(Convolution2D(64, 3, 3, kernel_regularizer = regularizers.l2 (0.0001), border_mode = 'same'))model1.add(Activation('relu'))model1.add(BatchNormalization())model1.add(Convolution2D(128, 3, 3,kernel_regularizer=regularizers.l2(0.0001),border_mode='same'))model1.add(Activation('relu'))model1.add(BatchNormalization())model1.add(MaxPooling2D(pool_size=(2, 2)))model1.add(Dropout(0.3))model1.add(Convolution2D(32, 1, 1))model1.add(Convolution2D(128, 3, 3,kernel_regularizer=regularizers.l2(0.0001), border_mode='same'))model1.add(Activation('relu'))model1.add(BatchNormalization())model1.add(Convolution2D(256, 3, 3,kernel_regularizer=regularizers.l2(0.0001), border_mode='same'))model1.add(Activation('relu'))model1.add(BatchNormalization())model1.add(MaxPooling2D(pool_size=(2, 2)))model1.add(Dropout(0.5))model1.add(Convolution2D(10, 1, 1))model1.add(AveragePooling2D(pool_size = (4,4)))model1.add(Flatten())model1.add(Activation('softmax'))sgd = optimizers.SGD(lr=0.0001, momentum=0.9, nesterov=True)model1.compile(optimizer=sgd, loss='categorical_crossentropy', metrics=['accuracy']) Do not worry about the lr that is assigned above. You can assign any value here since we will be overriding it soon. In order to make the model work better, I will use the cutout function. def get_random_eraser(p=0.5, s_l=0.02, s_h=0.4, r_1=0.3, r_2=1\/0.3, v_l=0, v_h=255, pixel_level=False):    def eraser(input_img):        img_h, img_w, img_c = input_img.shape        p_1 = np.random.rand()        if p_1 > p:            return input_img        while True:            s = np.random.uniform(s_l, s_h) * img_h * img_w            r = np.random.uniform(r_1, r_2)            w = int(np.sqrt(s \/ r))            h = int(np.sqrt(s * r))            left = np.random.randint(0, img_w)            top = np.random.randint(0, img_h)            if left + w <= img_w and top + h <= img_h:                break        if pixel_level:            c = np.random.uniform(v_l, v_h, (h, w, img_c))        else:            c = np.random.uniform(v_l, v_h)        input_img[top:top + h, left:left + w, :] = c        return input_img    return eraser Let us go ahead and implement the LR finder algorithm from Keras. from keras.callbacks import Callbackclass LR_Finder(Callback):     def __init__(self, start_lr=1e-5, end_lr=10, step_size=None, beta=.98):        super().__init__()        self.start_lr = start_lr        self.end_lr = end_lr        self.step_size = step_size        self.beta = beta        self.lr_mult = (end_lr\/start_lr)**(1\/step_size)            def on_train_begin(self, logs=None):        self.best_loss = 1e9        self.avg_loss = 0        self.losses, self.smoothed_losses, self.lrs, self.iterations = [], [], [], []        self.iteration = 0        logs = logs or {}        K.set_value(self.model.optimizer.lr, self.start_lr)            def on_batch_end(self, epoch, logs=None):        logs = logs or {}        loss = logs.get('loss')        self.iteration += 1                self.avg_loss = self.beta * self.avg_loss + (1 - self.beta) * loss        smoothed_loss = self.avg_loss \/ (1 - self.beta**self.iteration)        if self.iteration>1 and smoothed_loss > self.best_loss * 4:            self.model.stop_training = True            return        if smoothed_loss < self.best_loss or self.iteration==1:            self.best_loss = smoothed_loss        lr = self.start_lr * (self.lr_mult**self.iteration)            self.losses.append(loss)        self.smoothed_losses.append(smoothed_loss)        self.lrs.append(lr)        self.iterations.append(self.iteration)        K.set_value(self.model.optimizer.lr, lr)      def plot_lr(self):        plt.xlabel('Iterations')        plt.ylabel('Learning rate')        plt.plot(self.iterations, self.lrs)    def plot(self, n_skip=1):        plt.ylabel('Loss')        plt.xlabel('Learning rate (log scale)')        plt.plot(self.lrs[n_skip:-5], self.losses[n_skip:-5])        plt.xscale('log')            def plot_smoothed_loss(self, n_skip=10):        plt.ylabel('Smoothed Losses')        plt.xlabel('Learning rate (log scale)')        plt.plot(self.lrs[n_skip:-5], self.smoothed_losses[n_skip:-5])        plt.xscale('log')    def plot_loss(self):        plt.ylabel('Losses')        plt.xlabel('Iterations')        plt.plot(self.iterations[10:], self.losses[10:]) It is time to put everything together. We will define our accuracy function and a function to plot the model graphically. def plot_model_history(model_history):    fig, axs = plt.subplots(1,2,figsize=(15,5))    axs[0].plot(range(1,len(model_history.history['acc'])+1),model_history.history['acc'])    axs[0].plot(range(1,len(model_history.history['val_acc'])+1),model_history.history['val_acc'])    axs[0].set_title('Model Accuracy')    axs[0].set_ylabel('Accuracy')    axs[0].set_xlabel('Epoch')    axs[0].set_xticks(np.arange(1,len(model_history.history['acc'])+1),len(model_history.history['acc'])\/10)    axs[0].legend(['train', 'val'], loc='best')    axs[1].plot(range(1,len(model_history.history['loss'])+1),model_history.history['loss'])    axs[1].plot(range(1,len(model_history.history['val_loss'])+1),model_history.history['val_loss'])    axs[1].set_title('Model Loss')    axs[1].set_ylabel('Loss')    axs[1].set_xlabel('Epoch')    axs[1].set_xticks(np.arange(1,len(model_history.history['loss'])+1),len(model_history.history['loss'])\/10)    axs[1].legend(['train', 'val'], loc='best')    plt.show()def accuracy(test_x, test_y, model):    result = model.predict(test_x)    predicted_class = np.argmax(result, axis=1)    true_class = np.argmax(test_y, axis=1)    num_correct = np.sum(predicted_class == true_class)     accuracy = float(num_correct)\/result.shape[0]    return (accuracy * 100)datagen = ImageDataGenerator(zoom_range=0.0,                              horizontal_flip=False,                              preprocessing_function=get_random_eraser(v_l=min_pix, v_h=max_pix, pixel_level=True))lr_finder = LR_Finder(start_lr=1e-5, end_lr=1e-2, step_size=np.ceil(X_train.shape[0]\/128))start = time.time()model_info = model1.fit_generator(datagen.flow(X_train, Y_train, batch_size = 128),                                  samples_per_epoch = train_features.shape[0], nb_epoch = 100,                                   validation_data = (X_test, Y_test), verbose=0,                                  callbacks=[lr_finder])end = time.time()print (\"Model took %0.2f seconds to train\"%(end - start))print (\"Accuracy on test data is: %0.2f\"%accuracy(X_test, Y_test, model1))lr_finder.plot_lr() lr_finder.plot_smoothed_loss() Typically, a good static learning rate can be found half-way on the descending loss curve. In the plot shown that would be around 0.002(10^-3) because that is where the descent is steeper. For our cyclic learning rates, we need boundaries (start and end) and this can be identified from the graph as well. The boundaries are the point at which the loss starts descending and the point at which the loss stops descending. From the graph above, the curve starts at 0.002 and stops at 0.2 (10^-1). We have identified our boundaries, let us implement the cyclic LR and begin our training. from keras.callbacks import Callback, ModelCheckpointclass CyclicLR(Callback):  def __init__(self, min_lr, max_lr, stepsize=1000):    super().__init__()    self.min_lr = min_lr    self.max_lr = max_lr    self.currstep = 0    self.stepsize = stepsize  def on_train_batch_begin(self, batch, logs=None):    currstep = self.currstep    stepsize = self.stepsize    min_lr   = self.min_lr    max_lr   = self.max_lr    dlr = (max_lr - min_lr) \/ stepsize    if currstep < stepsize :      dlr = dlr*currstep    else:      dlr = dlr*(2*stepsize - currstep)    lr = min_lr + dlr    K.set_value(self.model.optimizer.lr, lr)    self.currstep += 1  def on_train_batch_end(self, batch, logs=None):    if self.currstep == 4000:      self.currstep = 0clr = CyclicLR(2e-4, 2e-2, 2000)model1.compile(optimizer=sgd, loss='categorical_crossentropy', metrics=['accuracy'])def scheduler(epoch, lr):  return round(1e-2\/(1+0.1*epoch), 10)start = time.time()model_info = model1.fit_generator(datagen.flow(X_train, Y_train, batch_size = 128),                                 samples_per_epoch = train_features.shape[0], nb_epoch = 100,                                  validation_data = (X_test, Y_test), verbose=1,                                 callbacks=[clr])end = time.time()print (\"Model took %0.2f seconds to train\"%(end - start))plot_model_history(model_info)print (\"Accuracy on test data is: %0.2f\"%accuracy(X_test, Y_test, model1)) Although we see no sharp spikes, while tuning hyperparameters it is essential to check for overfitting. The best way to do this is to identify misclassified images in the dataset. Once this identification is done, you can always go back to the learning rate curves or the model and tweak it further to get the best results possible. I will use gradcam and identify the misclassifications below. import cv2model.summary()def gradcam(idx, images, normimage, layername):  ii = idx  x = normimage[ii].reshape((1, 32, 32, 3))  preds = model.predict(x)  class_idx = np.argmax(preds[0])  class_output = model.output[:, class_idx]  last_conv_layer = model.get_layer(layername)      grads = K.gradients(class_output, last_conv_layer.output)[0]  pooled_grads = K.mean(grads, axis=(0, 1, 2))  iterate = K.function([model.input], [pooled_grads, last_conv_layer.output[0]])  pooled_grads_value, conv_layer_output_value = iterate([x])  depth = conv_layer_output_value.shape[-1]  for i in range(depth):    conv_layer_output_value[:, :, i] *= pooled_grads_value[i]         heatmap = np.mean(conv_layer_output_value, axis=-1)  heatmap = np.maximum(heatmap, 0)  max_heatmap = np.max(heatmap)  if max_heatmap >= 0 :     heatmap \/= max_heatmap  img = images[ii]   heatmap = cv2.resize(heatmap, (img.shape[1], img.shape[0]))  heatmap = np.uint8(255 * heatmap)  heatmap = cv2.applyColorMap(heatmap, cv2.COLORMAP_JET)  heatmap = cv2.cvtColor(heatmap, cv2.COLOR_BGR2RGB)           superimposed_img = cv2.addWeighted(img, 0.7, heatmap, 0.3, 0)  return superimposed_imgy_pred = model.predict(X_test)i = 0fig, ax = plt.subplots(10, 5, figsize = (15, 30))fig.suptitle('Misclassified Images')fig.tight_layout(pad = 0.3, rect = [0, 0, 0.9, 0.9])for (x, y) in [(i, j) for i in range(5) for j in range(5)]:  while i < 10000 and np.argmax(y_pred[i, :]) == testy[i]:    i += 1  ax[2*x, y].imshow(testX[i])  ax[2*x, y].axis('off')  acls, pcls = class_names[int(testy[i])], class_names[np.argmax(y_pred[i, :])]  ax[2*x, y].set_title('%d A: %s P: %s' % (i, acls, pcls))  ax[2*x+1, y].imshow(gradcam(i, testX, X_test, 'conv2d_17'))  ax[2*x+1, y].axis('off')  i += 1   if i >= 10000:    break You can see here that these images are not classified right. But now that we have the tools to improve our learning rates we can go back to the model and tune it better. Conclusion Hyper-parameter optimization is a very important and time-consuming process in the life of a good machine learning model. It helps in making the model stand out and be better. With the techniques discussed above, you can improve your model by tuning the learning rates better.","excerpt":"This article covers the types of Learning Rate (LR) algorithms, behaviour of learning rates with SGD and implementation of techniques to find out suitable LR values.","categories":["Deep Tech"],"tags":["learning rate","Machine Learning","machine learning optimization","optimizers","Python"],"author_name":"Bhoomika Madhukar","publish_date":"2020-08-12T10:00:00","publication_year":"2020","word_count":1886,"keywords":["Go","NumPy","learning rate","machine learning","Keras","AI","TPU","Machine Learning","RAG","Python","deep learning","machine learning optimization","Matplotlib","R","optimizers"],"extracted_tech_keywords":["AI","machine learning","deep learning","Keras","NumPy","Matplotlib","RAG","TPU","R","Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/comprehensive-guide-to-learning-rate-algorithms-with-python-codes\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10051113,"title":"Microsoft acquires OKR startup Ally.io","content":"Microsoft recently announced that it has acquired Ally.io, one of the leading OKR (objectives and key results) platforms and that it will join the Microsoft Viva family as part of their employee experience platform (EXP) designed to help companies embrace digital work life. Microsoft says the idea behind Viva and the Ally acquisition is to provide a more transparent way to communicate company goals and objectives to employees. The two firms did not share the purchase price. “Aligning employee work to the company’s strategic mission and core priorities is top of mind for every organization. To do this, leaders need to invest in tools that communicate transparency around big company bets and create ways to cascade aspirational goals and report results at all levels of an organization,” said Kirk Koenigsbauer, Chief Operating Officer & Corporate Vice President, Experiences and Devices Group, at Microsoft. Ally CEO and founder Vetri Vellore said, “As a part of Microsoft Viva, Ally.io will continue to give leaders, teams and individuals the ability to align and focus everyday work to the company’s most important objectives. We will help bring goals and purpose to wherever the team is doing work, including Teams, Outlook, Slack and the other systems you use every day”. The OKR category is a fast-growing and emerging space. Ally.io is considered one of the most loved tools on the market. Microsoft says that the customers will find the Ally.io experience flexible, easy to use with quick time-to-value. They will appreciate its broad set of integrations with existing work systems. Since its launch in 2018, Ally.io has been adopted by over 1,000 leading high-tech, manufacturing, financial services and healthcare businesses across more than 80 countries. Ally.io will power the new Microsoft Viva module. Viva is an employee experience platform that brings together communications, knowledge, learning, resources and insights – from anywhere you work. Powered by Microsoft 365 and Microsoft Teams, Viva helps organizations foster a culture of human connection, growth, well-being and success. Viva has four connected modules: Connections, Insights, Topics and Learning – that deliver a holistic experience with intelligent and actionable insights in the flow of work.","excerpt":"Microsoft says that the customers will find the Ally.io experience flexible, easy to use with quick time-to-value.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","Data Science","Data Scientist","Deep Learning","Google","Machine Learning","Microsoft","Microsoft 365"],"author_name":"Victor Dey","publish_date":"2021-10-08T17:45:36","publication_year":"2021","word_count":354,"keywords":["Go","Microsoft","AI","programming_languages:R","Machine Learning","Git","programming_languages:Go","Microsoft 365","Google","GAN","Deep Learning","Data Science","Data Scientist","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","R","Go","Git","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/microsoft-acquires-okr-startup-ally-io\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10041427,"title":"ISRO Offers A Five-Day Machine Learning Course For Free","content":"The Indian Space Research Organisation has announced a five-day free course on machine learning, between July 5 -9. The course is being offered as part of the Indian Institute of Remote Sensing’s (part of ISRO) outreach program. Central and state government employees, researchers, professionals, and those attached with NGOs can attend the course. Interested candidates must have basic knowledge of remote sensing and GIS. The short course is designed for professionals engaged in remote sensing data processing in different applications which involves extracting a specific class of interest. Register for AWS ML Fridays | Hands-on Workshops on Machine Learning The course content is planned as follows: July 5: Remote Sensing and its sensors of various resolutions. Radiometry and Geometric corrections and Basic understanding of Image July 6: Basic classifier to machine learning-A journey July 7: Methods in machine learning: Supervised, unsupervised and reinforcement. July 8: Fuzzy based machine learning with application in temporal data processing July 9: Network-based learning algorithms – ANN to CNN\/RNN The course materials such as lecture slides, recordings of the classes, handouts of the demonstrations, etc will be made available to the students and the full video lectures will be uploaded on YouTube. Candidates can attend the course live via any web browser through the e-class portal of IIRS Dehradun. The participants can also attend the live workshop via IIR’s YouTube channel of IIR. For receiving the course completion certificate, a student must have attended 70 percent of the sessions via the e-class portal. Students attending the class via IIR’s YouTube channel should mark their attendance via an offline session made available after 24 hours post the class. All the information regarding registration and the course is available here. For further clarification, students may contact Dr Anil Kumar, course coordinator and head, PRSD, at 0135-2524114 or email him at: anil@iirs.gov.in","excerpt":"The Indian Space Research Organisation has announced a five-day free course on machine learning, between July 5 -9. The course is being offered as part of the Indian Institute of Remote Sensing’s (part of ISRO) outreach program. Central and state government employees, researchers, professionals, and those attached with NGOs can attend the course. Interested candidates […]","categories":["AI News"],"tags":["ISRO"],"author_name":"Shraddha Goled","publish_date":"2021-06-08T12:28:16","publication_year":"2021","word_count":304,"keywords":["Go","ISRO","machine learning","AWS","AI","cloud_platforms:AWS","ML","RNN","GAN","CNN","R"],"extracted_tech_keywords":["AI","machine learning","ML","AWS","R","Go","GAN","CNN","RNN","cloud_platforms:AWS"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/isro-offers-a-five-day-machine-learning-course-for-free\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10099881,"title":"MongoDB Launches Academia Program in India to Upskill 500k+ Students","content":"MongoDB has launched an Academia Program in India to upskill students to use its multi-cloud developer data platform called MongoDB Atlas. This program aims to train more than 5,00,000 students and 1000 plus educators by partnering with more than 800 educational institutions across the country. In India, 6 out of 10 IT engineering graduates do not possess the necessary skills required to enter high demand tech roles, according to a report by National Association of Software and Service Companies. MongoDB’s program wants to close this gap and equip students with its Altas platform which enables developers to build modern applications. “The biggest challenge is finding and retaining the right developer talent”, said Sachin Chawla, Area Vice President, India at MongoDB in a press release. The launch of this program will help solve these challenges and support the next generation of Indian developers and business as they capitalize on this massive opportunity, he said. The New York-based software giant will partner with ICT Academy, a non-profit initiative of the Government of Tamil Nadu and the Government of India which helps to close the technology gap in India.  The program will provide access MongoDB Atlas credits and certification courses free of cost–helps validates skills to employers It will also provide curriculum resources and training to educators and students to build, manage and deploy modern applications critical for business. ICT Academy and MongoDB will conduct joint activities like academia summits, learnathons and tech bootcamps. Indians currently pursuing PhD, as part of this program will also also get access to MongoDB PhD fellowship, which enables them to contribute in the field of computer science. Last year, the software firm launched a revamped version of MongoDB University, which provides free, on-demand access to courses to learn high demand software management and data management skills. This has helped thousands of Indians to learn and develop their DB skills.","excerpt":"This program will train 5,00,000+ students and 1000 plus educators in India","categories":["AI News"],"tags":[],"author_name":"Pranav Kashyap","publish_date":"2023-09-12T16:41:04","publication_year":"2023","word_count":312,"keywords":["Go","API","programming_languages:R","AI","MongoDB","programming_languages:Go","Aim","ViT","data_tools:MongoDB","R"],"extracted_tech_keywords":["AI","Aim","MongoDB","R","Go","API","ViT","programming_languages:R","programming_languages:Go","data_tools:MongoDB"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/mongodb-launches-academia-program-in-india-to-upskill-500k-students\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10046331,"title":"[Jobs Roundup] Latest Analytics &#038; Data Science Job Openings At GICs In India","content":"We have listed the latest Analytics and Data Science Job Openings at GICs in India this week. Global In-house Centers (GICs) are service delivery operations typically in low-cost geography, owned and operated by the same company receiving the services. 1| Data Scientist at ABB Location: Bangalore Responsibilities: Working with business managers to frame a problem, both mathematically and within the business context.Understanding business data and how to use it appropriately in data analysis.Performing exploratory data analysis to gain a deeper understanding of the problem. Apply here. 2| Data Scientist – Strategy and Analytics at Dell Location: Bangalore Responsibilities: Collaborate with internal and external teams to understand customer requirements.Develop and apply a broad range of techniques and theories from statistics, machine learning, and business intelligence to deliver actionable business insights.Develop and drive testing of algorithms efficacy for differing analytical use-cases. Performs end-to-end steps involved in model development while establishing subject-matter expertise. Apply here. 3| Senior Data Scientist at Citigroup Location: Mumbai Responsibilities: Conducts strategic data analysis, identifies insights and implications and make strategic recommendations, develops data displays that clearly communicate complex analysis.Mines and analyzes data from various platforms to drive optimization and improve data quality.Delivers analytics initiatives to address business problems with the ability to determine data required, assess time & effort required and establish a project plan. Apply here. 4| Manager – Data Scientist at Morgan Stanley Location: Bangalore Responsibilities: Development of risk models leveraging Machine learning and related quantitative techniques to improve surveillance and compliance monitoring and testing activitiesStatistical analysis of model thresholds, backtesting and other sensitivity and productivity analyses, model documentation and facilitate model validations.Working with members of the Financial Crimes, Compliance, Legal and IT departments to identify, assess and manage risk by evaluating large data sets using various statistical tools. Apply here. 5| Data Scientist Computer Vision at Swiss Re Location: Bangalore Responsibilities: Working with big data tools and on machine learning\/AI model development for innovative products. Python is required but any other language used in model development is welcomed.Help in developing visual intelligence and image recognition algorithm is mandatory.Strong in data science model deployment, math, statistics, physics or any other computationally intense disciplines and working experience in computer science\/engineering. Apply here. 6| Data Engineer at Nokia Location: Bangalore Responsibilities: Understanding the impact of the IT requirements on the changes to be able to challenge the business requirement and proposed solution.Analyse and study complex systems requirements and troubleshoot and resolve issues in existing software or solution.Proactively follows the digital market technologies, trends and innovations. Apply here. 7| Lead Data Scientist at Nissan Motor Corporation Location: Trivandrum, Tamil Nadu Responsibilities: Expected to drive a business problem right from interacting with business, data collection, building a concept application to demonstrate the advantages of the new methodology with explanations.Bring in original ideas, collaborate and work with the team and 3rd parties for speed of delivery to solve and deploy the solution globally. They also leverage the vast global network of Nissan Motors to collaborate with Nissan Digital – Data Engineering CoE, Software CoE, CI\/CD team and other functions for creating and deploying solutions. Apply here. 8| AI\/ML Engineer at Intel Location: Bangalore Responsibilities: The role is to transform manufacturing, design and validation processes using AI\/ML technologies.The candidate must have experience in delivering such transformational programs using AI\/ML and also have the ability to conceptualize such transformational ideas.Should have end to end developed and deployed AI solutions with the complete data pipeline and must have experience in resolving post-production issues. Apply here. 9| Experienced Data Engineer at Boeing Location: Bangalore Responsibilities: With Solution Architects \/ Data Architecture support solution design for the service layer for IPSS APIs and Data Services for AnalyticsAPI Development, Data Service for ETL\/ADF or other means for IPSS Data from Data Pipeline to T&PS Data Repository and other stakeholdersDevelop business logic as needed for developing and consuming IPSS services for example User identity validation and other logic based on current Catalyst and ICS logic applied to services. Apply here. 10| Human Capital Management Data Engineer – Senior Analyst at Goldman Sachs Location: Bangalore Responsibilities: Build, manage & maintain various data models in HCM.Build & maintain web applications and manage their data flows.Analyse and provide insights to the business through BI and Analytics Tools. Apply here. 11| DevOps Engineer (Python\/Go) at SAP Location: Bangalore Responsibilities: Setup, automate & maintain CI\/CD pipelines. Develop new features in our CI\/CD process framework.Gain hands-on experience with new build technologies.On board new features in collaboration with SAP’s global development teams. Develop monitoring dashboards for CI\/CD services. Apply here. 12| Data Engineer at Commonwealth Bank of Australia Location: Bangalore Responsibilities: Undertake identification of detailed data flow & lineage from point of data capture to point of data consumptions.Work with stakeholders to Identify rules for data quality measurement and develop thresholds.Undertake an assessment of data risks and existing controls that mitigate these risks and generate data quality assessment. Apply here. AIMRecruits A subsidiary of Analytics India Magazine, AIM Recruits is an India-focused executive search firm that partners with leading businesses to assess and acquire top data science talent to drive breakthrough performance. AIM Recruits is a leading Executive Search Firm for Analytics, Data Science & Artificial Intelligence. 1| Analytics Consultant Location: Delhi NCR Responsibilities: Support\/Lead development and validation of PD, LGD, EAD estimation for various portfolios of a retail bank from a regulatory standpoint.Background in Basel \/ IFRS modelling will be an added advantage.Proficiency in R and SQL is a must. Apply here. 2| Data Analyst at Knowlvers Location: Delhi NCR\/Ghaziabad Responsibilities: Strong in Coding in Python, Exposure to ML model development, Tableau.Leverage analytics to solve structured or unstructured problems and interpreting results.Ability to consult clients by discovering insightful & relevant patterns from data. Apply here. 3| Data Engineer – ETL\/SQL Location: Bangalore Responsibilities: Able to debug, handle ETL failures on daily basis. Scheduling & Monitoring of Batch Jobs.Prepare Plans for all ETL Procedures and architectures.Data Integration from various sources & adhoc requirements including Data Extraction. Tuning and Optimizing the Existing Load Plans as and when required. Apply here. 4| Senior Manager – Data Engineering Location: Bangalore Responsibilities: Lead and continue to build out a team of Data Engineers in the region.Establish and implement state-of-art data engineering standards, tools, and techniques across the team.Partner with Data Science, Machine Learning, and Platform Engineering teams to develop and deploy production-quality code. Apply here. 5| Senior Manager – Data Scientist at Bain & Company Location: Bangalore Responsibilities: Provide data science expertise to Bain case teams and clients worldwide. You will work together with case teams to drive results by assessing needs and developing data science methods, products and capabilities.Develop analytical solutions that bring critical insights to a variety of different problems such as targeting customers and segmenting markets, product design, marketing optimization, demand forecasting and brand valuation, profit and price analyses, fraud detection.Collaborate closely with and influence business consulting staff and leaders as part of multi-disciplinary teams to assess opportunities and develop data-driven solutions for Bain clients across a variety of sectors. Apply here. 6| Data Engineer\/Senior Data Engineer Location: Bangalore (Currently WFH) Responsibilities: Will be responsible for onboarding tables to the data lake and building pipelines for data ingestion using python.Building data models and tables on Athena which will be used for reporting.Create and maintain workflow\/technical documents for all the development activities. Apply here. 7| Senior Analyst – Data Science at Max Life Insurance Location: Gurgaon Responsibilities: Liaison and collaborate with different business stakeholders and back end engineering teams to develop AI-based solutions.Build data-driven stories and present analysis output and solutions to all levels of the business.Understand business processes, collect and integrate data from various data sources to be used for building the models. This could involve writing crawlers, connecting to databases for extracting data for training algorithms to create features supporting hypotheses. Apply here.","excerpt":"We have listed the latest Analytics and Data Science Job Openings at GICs in India this week.","categories":["AI Hirings"],"tags":["AIM job updates","analytics jobs in India","big data quality","data analyst vs data scientist","data science and manufacturing","Data Science Jobs","data science jobs in India","data science project marketing","Data Scientist Jobs","how to measure twitter influence","jobs in india","jobs roundup","load data python","open source ios","statistical analysis using sql","Statistics for Data Science","weekly job updates","weekly jobs roundup","what is data science"],"author_name":"kumar Gandharv","publish_date":"2021-08-19T18:00:00","publication_year":"2021","word_count":1298,"keywords":["big data quality","statistical analysis using sql","AIM job updates","computer vision","weekly job updates","Statistics for Data Science","jobs roundup","fraud detection","data science","load data python","Data Scientist Jobs","artificial intelligence","how to measure twitter influence","jobs in india","analytics jobs in India","RAG","weekly jobs roundup","analytics","what is data science","data science project marketing","machine learning","AI","ML","Data Science Jobs","data science jobs in India","data analyst vs data scientist","data science and manufacturing","open source ios","Aim"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","computer vision","data science","analytics","Aim","RAG","fraud detection"],"url":"https:\/\/analyticsindiamag.com\/ai-hiring\/jobs-roundup-latest-analytics-data-science-job-openings-at-gics-in-india\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":44776,"title":"Why Has India Lagged Behind In Semiconductor Chip Manufacturing?","content":"While India has done well in terms of chip design and electronics manufacturing, there have been challenges in setting up of Semiconductor Wafer Fabrication (FAB) units for a long time. The digital age has propelled the world into consuming electronics at an unprecedented scale. There will be worldwide shipments of devices — PCs, tablets and mobile phones — totalling 2.2 billion units in 2019. All of these gadgets require semiconductor chips to function, and it is clear that economies with a large production of these chips have benefited the most in terms of enhancing their GDP. US, Japan, Korea, China, Singapore, etc are all large producers of semiconductor chips and also have a strong foothold on the global economy. Where Does India Stand? While India has done well in terms of chip design and electronics manufacturing, there have been challenges in setting up of Semiconductor Wafer Fabrication (FAB) units in India for a long time. This is due to multiple factors, including not just the lack of infrastructure and skilled labour in the country. It is also difficult to compete with neighbouring countries like China and Vietnam which have been favourite destinations for global chip manufacturers due to better cost-efficiency. It is for these reasons that Intel stated in 2014 it had no interest in starting manufacturing in India. There have been attempts to set up semiconductor fab units here by private companies in the country: Hindustan Semiconductor Manufacturing Corporation (HSMC), a consortium of companies that included ST Microelectronics and Silterra Malaysia was aiming kickstart chip manufacturing plant in Gujarat, a project worth ₹30,000 crore. The government in 2019 cancelled the letter of intent granted to HSMC and now there is no such proposal from any private company to initiate such a project. The reason was that the consortium could not submit the required documents asked by the government for setting up of Semiconductor Wafer Fabrication (FAB) unit. HSMC had been backed by AMD and has also received ₹700 crore in funding from Mumbai-based Next Orbit Ventures. Then there was another consortium led by Jaiprakash Associates, which partnered with IBM and Tower Semiconductor of Israel to start chip manufacturing in UP. In 2016, debt-ridden Jaiprakash (JP) Associates has pulled out of the ₹34,000-crore. If the only two private sector consortiums cleared by the government to establish large scale chip manufacturing in the country could not make it happen, then it’s certainly a bad indicator to why India is lagging behind in the space. What Are The Hurdles? One of the biggest hurdles in setting up fab manufacturing units is the fact that it requires massive investment. In addition to the huge cost, running in billions of dollars, manufacturing even a single chip requires hundreds of gallons of pure water, which may also be hard to find in India in the required quantities. An uninterrupted power supply is another major hurdle. The heart of the issue is that India is still not unto the par in terms of the basic infrastructure needed to pursue endeavours in the chip manufacturing space. There is also constant price pressure from other global players, particularly China which is also building a homegrown chip program for the adoption of local semiconductors in 70% of its products by 2025. No Dearth Of Talent In Chip Manufacturing It’s not like India does not possess the scientific capabilities to manufacture semiconductor chips. In fact, the level of technical know-how is so high that the government even considered a ₹2,500 crore feasible study based on Gallium Nitride semiconductor fab at the Indian Institute of Science (IISc) in order to build a local semiconductor industry. Gallium Nitride based semiconductors skip altogether the already mature silicon-based fab technology dominated by countries like US and China. While large scale semiconductor manufacturers face challenges to take off, several startups building chips for embedded systems like IoT and telephony on a much smaller scale are also disrupting the scene. There are many prominent startups here like Signalchip, a semiconductor company also based in Bengaluru, which rolled out 4G and 5G modem chips. Saankhya Labs, is another Bengaluru-based startup that has been creating chipsets for use in defence, satellite communication and broadcast. Another one is a microprocessor called Shakti, developed at IIT Madras which can be used in mobile computing devices, embedded low power wireless systems like smartphones, surveillance cameras and networking systems. Overview Clearly, there is no dearth of talent when it comes to India becoming a dominant player in semiconductor chip manufacturing.Yet, the country has struggled to find a way to establish fab units needed for large scale manufacturing. Now, with the rising demand for electronics, India is a large net importer of semiconductor chips. In fact, experts say that India is spending more money on import of semiconductor than on oil. The way to reduce the dependency on chip imports is to create semiconductor manufacturing units within the country. Here, the government needs to make sure there is proper infrastructure and investments are made so that scalable manufacturing units can be created. We also need to look at the success story of China in semiconductor manufacturing and take lessons.","excerpt":"While India has done well in terms of chip design and electronics manufacturing, there have been challenges in setting up of Semiconductor Wafer Fabrication (FAB) units for a long time. The digital age has propelled the world into consuming electronics at an unprecedented scale. There will be worldwide shipments of devices — PCs, tablets and […]","categories":["AI Features"],"tags":["semiconductor chips","Semiconductor India","Semiconductor market in india","Semiconductor Wafer Fabrication"],"author_name":"Vishal Chawla","publish_date":"2019-08-21T10:54:19","publication_year":"2019","word_count":856,"keywords":["Go","funding","programming_languages:R","AI","Semiconductor market in india","Scala","semiconductor chips","Git","Aim","programming_languages:Scala","Semiconductor India","Semiconductor Wafer Fabrication","R","startup"],"extracted_tech_keywords":["AI","Aim","R","Go","Scala","Git","startup","funding","programming_languages:R","programming_languages:Scala"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/india-semiconductor-chip-manufacturing\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":65417,"title":"NVIDIA Announced It&#8217;s First GPU-Based AI Chip That Can Boost Performance By Up To 20x","content":"Recently, NVIDIA CEO and Co-Founder Jensen Huang introduced GPU-based NVIDIA A100. Huang made the announcement from the kitchen of his California home along with discussions on important new software technologies such as NVIDIA Jarvis and recent Mellanox acquisition. NVIDIA A100 is said to be the first GPU based on the NVIDIA Ampere architecture with 54 billion transistors. The NVIDIA A100 Tensor Core GPU is said to deliver unprecedented acceleration at every scale for AI, data analytics, including high-performance computing (HPC) to tackle highly complex processing challenges. Behind A100 NVIDIA A100 is part of the complete NVIDIA data centre solution that incorporates building blocks across hardware, networking, software, libraries, and optimised AI models and applications from NGC, which is the hub for GPU-optimised software for deep learning, machine learning, and high-performance computing (HPC). This AI chip is represented as the most powerful end-to-end AI and HPC platform for data centres that allows the researchers to deliver real-world results and deploy solutions into production at scale. According to a blog post, this chip draws on design breakthroughs in the NVIDIA Ampere architecture — offering the company’s largest leap in performance to date within its eight generations of GPUs — to unify AI training, inference, and boost performance by up to 20x over its predecessors. Huang stated, “The powerful trends of cloud computing and AI are driving a tectonic shift in data centre designs so that what was once a sea of CPU-only servers is now GPU-accelerated computing.” He added, “NVIDIA A100 GPU is a 20x AI performance leap and an end-to-end machine learning accelerator — from data analytics to training and inference. For the first time, scale-up and scale-out workloads can be accelerated on one platform. NVIDIA A100 will simultaneously boost throughput and drive down the cost of data centres.” Breakthroughs The NVIDIA A100 GPU is a technical design breakthrough that is fueled by five key innovations, and the outcome of these innovations result in 6x higher performance than NVIDIA’s previous generation Volta architecture for training and 7x higher performance for inference. The innovations are mentioned below: NVIDIA Ampere Architecture: At the core of this chip is the NVIDIA Ampere GPU architecture that contains more than 54 billion transistors and hence, making it the largest 7-nanometer processor in the world.Third-generation Tensor Cores with TF32: The A100 includes new TF32 for artificial intelligence that allows for up to 20x the AI performance of FP32 precision, without making any changes in the code.Multi-Instance GPU: Multi-instance GPU or MIG is a new technology that allows multiple networks to operate simultaneously on a single A100 GPU for optimal utilisation of computing resources. It enables a single A100 GPU to be partitioned into as many as seven separate GPUs in order to deliver varying degrees of computing for jobs of different sizes while providing optimal utilisation as well as maximising return on investment. Third-Generation NVIDIA NVLink: The 3rd-generation NVIDIA NVLink helps in doubling the high-speed connectivity between GPUs, and thus provide efficient performance scaling in a server.Structural Sparsity: Structural sparsity is a new technique that harnesses the inherently sparse nature of AI math to double the performance. Wrapping Up Among the early adopters of this chip, tech giant Microsoft will be the first one to use the power of NVIDIA A100 GPUs. Mikhail Parakhin, corporate vice president at Microsoft, stated, “We will train dramatically bigger AI models using thousands of NVIDIA’s new generation of A100 GPUs in Azure at scale to push the state-of-the-art on language, speech, vision and multi-modality.” Watch The Announcement Here:","excerpt":"Recently, NVIDIA CEO and Co-Founder Jensen Huang introduced GPU-based NVIDIA A100. Huang made the announcement from the kitchen of his California home along with discussions on important new software technologies such as NVIDIA Jarvis and recent Mellanox acquisition. NVIDIA A100 is said to be the first GPU based on the NVIDIA Ampere architecture with 54 […]","categories":["Global Tech"],"tags":["NVIDIA","NVIDIA GPU"],"author_name":"Ambika Choudhury","publish_date":"2020-05-18T15:00:00","publication_year":"2020","word_count":587,"keywords":["NVIDIA GPU","artificial intelligence","machine learning","AWS","AI","cloud computing","R","deep learning","analytics","Tecton","NVIDIA","Azure"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","deep learning","analytics","Tecton","cloud computing","AWS","Azure","R"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/nvidia-announced-its-first-gpu-based-ai-chip-that-can-boost-performance-by-up-to-20x\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":16101,"title":"Why China is reckoned as a formidable force in the AI landscape?","content":"China is making significant strides in the AI space. The country recently laid out a roadmap to guide the development of Artificial Intelligence until the year 2030. China’s Minister for Science and Technology, Wan Gang had mentioned that the plan has already been drafted, and will soon be released to the public. The plan will address issues in 4 key areas: The building up of AI capability, The application of AI technologies, The introduction of policies to handle the risks brought about by AI, such as job losses International collaboration Wan revealed no clue about the amount of money Beijing is planning to invest towards developing AI technologies. However, during the annual meeting of China’s parliament in March, he announced that the central government would set up a special fund for research into the core technologies behind AI. Undoubtedly, the US took an early lead in the AI space, however, AI advocates in China believe that their country has all the chances to eventually overtake its western rival in the AI domain. Moreover, China has a lot of advantages. The country has the data, the market, and the talent needed to promote growth of AI. Various reports also suggest that the country will see its greatest economic gains globally from AI by 2030. This implies an estimated 26 percent boost to the GDP within that timeframe. That’s not all, AI technologies are expected to boost the world’s GDP by 14 per cent by 2030. China’s ambition in the AI space Minister of Science and Technology Wan Gang Chinese tech giants like Baidu, Alibaba, and Tencent are rapidly fueling fundamental AI breakthroughs. These organizations have access to the vast amounts of data needed for AI training through their millions of customers. Additionally, they are injecting massive investment into technology, setting up their own AI research laboratories to develop new products. Last year, China emerged as the world’s second biggest investor in AI enterprises, following the States. The country injected US$2.6 billion into the AI landscape. Moreover, the national development plan would see AI technology being adopted in areas including the economy, social welfare, environmental protection, and national security. Furthermore, last year the government had promised to establish an AI market that would worth more than US$15 billion by 2018. Beijing has already invested millions into studying AI in universities and research institutes around the nation. The technology is also being applied across the full spectrum of governance. Use cases for AI development in China A robot performs at a conference in Beijing China’s most innovative city of Shenzhen in Guangdong province presents a unique case for the use of AI. The city is making use of a tiny chip in public surveillance cameras. This small innovation has helped police crack hundreds of cases and find several lost children. The intelligent chip proliferates the speed of human facial recognition. The city of Jiaxing in Zhejiang province also began leveraging AI for general use. The traffic authorities in the city designed an AI coach in a driving school. The system not only monitors students’ driving behavior and the mistakes they make, but it also instructs them through a speaker and rates their performance. Statistics revealed that the passing rate of students who learned with the AI coach was 20 percent higher. China’s innovation in AI space is truly prolific. Another city Nantong will put an AI judge in place to organize and analyses legal documents and materials. The AI system will perform paper work to lighten the workload for human judges. Interestingly, the AI is expected to speed up the handling of legal cases by 30 percent. Even smaller firms aren’t missing on the opportunity to swoop into the landscape. Recently, AI robots owned by two Chinese education-tech companies attempted the math section of the nation’s toughest college examinations. Why is China betting big on AI? Artificial Intelligence holds the key to an autonomous world China has been touted to see the greatest economic gains from AI by 2030. This will only happen when technology accelerates global GDP growth by increasing productivity and boosting consumption. Moreover, AI technologies are expected to push global GDP by a further 14 per cent by 2030. As a result, China will see an estimated 26 percent. Interestingly, between 2016 and 2030, labour productivity improvements would account for over half of the US$15.7 trillion in economic gains. Increased consumer demand resulting from AI-enabled product enhancements will account for the rest. The analysis clearly states AI will continue to transform lives, enterprises, and societies over the next couple of years. Moreover, the technology powers an array of advanced applications, from facial recognition to self-driving vehicles, and will thus remain centre of attention for almost every tech company in China as they bet big on the landscape. China has already made great leaps in the development of AI. Tech giants like Baidu, Tencent Holdings and Alibaba Group have been striving harder than ever to attract top AI talent from Silicon Valley. The future of AI development in China China’s leading technology companies are investing heavily into research and development for artificial intelligence These advancements are the beginning of an AI revolution. With software and hardware upgrades, newer AI machines are expected to be much faster and more human-like. Thus, the AI user experience of the future would be strikingly different from today. However, central government dedicates its resources to help a wide range of special interests, which includes special funds that back poverty alleviation and environmental protections initiatives. This also implies expressed support for AI is unlikely to result in significant financial backing for the AI sector. Another aspect to take note here is the fact that the sector could require years before technologies are ready to be commercialized. Moreover, concerns are also raised around the research, development, and commercialization of AI, as it involves high risk, and are usually considered long term investments. To conclude with, it’s safe to say China will enjoy an edge in this competitive field, if government can provide strong support in funding.","excerpt":"China is making significant strides in the AI space. The country recently laid out a roadmap to guide the development of Artificial Intelligence until the year 2030. China’s Minister for Science and Technology, Wan Gang had mentioned that the plan has already been drafted, and will soon be released to the public. The plan will […]","categories":["IT Services"],"tags":["Artificial Intelligence India","Facial Recognition","hardware","Science","Statistics India","Technology"],"author_name":"Amit Paul Chowdhury","publish_date":"2017-07-05T12:11:08","publication_year":"2017","word_count":1008,"keywords":["Go","API","artificial intelligence","Facial Recognition","AI","innovation","Artificial Intelligence India","Science","hardware","RAG","Ray","Statistics India","Technology","ViT","GAN","R"],"extracted_tech_keywords":["AI","artificial intelligence","Ray","RAG","R","Go","API","GAN","ViT","innovation"],"url":"https:\/\/analyticsindiamag.com\/it-services\/china-reckoned-formidable-force-ai-landscape\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10114779,"title":"Top 10 Gadgets at MWC 2024","content":"The Mobile World Congress (MWC), an annual event organised by the GSMA, showcased the latest in mobile technology, including smartphones, services, and advancements in 5G and artificial intelligence. Held in Barcelona, Spain, it is the largest exhibition for the mobile industry, attracting global participants from the tech community. This year, it hosted a plethora of devices with AI integration being the common thread. One standout example of AI’s application was the Honor Magic V2, where eye-tracking technology allows users to interact with their device in a hands-free manner. This feature, along with other AI-driven innovations presented at MWC, underscores the industry’s shift towards creating more personalised and efficient user experiences. Here is a list of top 10 gadgets showcased at MWC this year. Top 10 Gadgets in MWC Showcase 2024HMD Barbie PhoneMotorola Debuts Smart ConnectOnePlus Watch 2A Transparent Laptop from LenovoSamsung Galaxy RingHonor Magic V2TCL’s NXTPaper 5G and Portable 5G DongleZTE 5G+AI Eyewear-free 3D TabletXiaomi 14 SmartphoneHumane AI HMD Barbie Phone HMD, initially celebrated for reviving Nokia phones at MWC 2017, struggled to compete with giants like Samsung and Apple, shifting its focus to budget Android and feature phones. At MWC 2024, it announced its first profitable year in 2023 and introduced a rebranding strategy, adopting ‘Human Mobile Devices’ as its new moniker. Its 2024 device lineup featured a classic Nokia model, and a unique Barbie flip phone developed in collaboration with Mattel, targeting a summer release as a pink, digital detox tool. As for the non-Barbie phones, the company has plans for those, too, though no details are available at the moment. Motorola Debuts Smart Connect Motorola showed off its innovative Adaptive Display concept phone, a departure from traditional designs with its bendable structure, allowing it to be bent backward. Motorola also introduced Smart Connect, a collaborative effort with Lenovo that builds upon the Ready for platform. This new feature allows for wireless connection between a Motorola phone and nearby displays, including Lenovo tablets and Windows laptops available through the Microsoft Store, enhancing productivity and inter-device usability. OnePlus Watch 2 In the spotlight at MWC 2024 was the OnePlus Watch 2, which boasts a significant improvement over its predecessor. A standout feature of the OnePlus Watch 2 is its dual operating system capability, powered by two distinct processors. It operates on Google Wear OS with the Qualcomm Snapdragon W5 Gen 1 chipset for demanding tasks such as navigation, music playback, and app usage. The OnePlus Watch 2, priced at $300, is currently available for preorder and will officially go on sale on March 4. OnePlus is also offering a promotional discount of $50 for those trading in any watch, including analog models, towards the purchase. A Transparent Laptop from Lenovo Lenovo introduced a concept at MWC 2024, known as Project Crystal, a transparent laptop. While it’s not slated for immediate release, the concept showcases a glimpse into the future of laptop design. The laptop’s Micro-LED transparent screen offers a futuristic look, allowing users to see through the device while still providing a bright display for normal app usage. However, this transparency means that others can see the user’s screen, posing privacy concerns. Lenovo mentioned the potential for adjusting the screen’s transmissivity to create an opaque layer for privacy, though such features were not demonstrated. Samsung Galaxy Ring Samsung unveiled its latest wearable, the Galaxy ring, at the Mobile World Congress. This is the first time it was showcased to the public. This smart ring, designed to monitor health data and provide insights based on daily and nightly metrics, will expand Samsung’s wearable market. The Galaxy ring can monitor temperature, heart rate, respiratory rate, sleep movement, and time taken to fall asleep. Interestingly, the Galaxy ring will also offer payment capabilities, distinguishing it from other smart rings that focus solely on health or fitness tracking. The ring is available in black, gold, and silver, and comes in nine sizes, accompanied by a sizing kit. Its price in India starts from ₹24,599 and is set to be released later this year. Honor Magic V2 Honor displayed its new devices, the Magic 6 Pro and Magic V2 RSR smartphones, and the MagicBook Pro 16 laptop, heavily emphasising AI features that aren’t necessarily driven by actual artificial intelligence. A notable demonstration featured the eye-tracking technology on the Magic 6 Pro, which allows users to expand notifications by simply looking at them. This feature, expected to be added via a software update, was highlighted through an unusual demo where the technology was used to control a car (an Alfa Romeo), with options like Engine Start and Stop, Forward, and Backward, showcasing the potential of eye-tracking for hands-free device interaction. TCL’s NXTPaper 5G and Portable 5G Dongle TCL introduced a new addition to its NXTPaper range, the TCL 50 XL NXTPaper 5G, featuring a 6.8-inch screen with a 120-Hz refresh rate designed to mimic paper. Despite its modest specs, its $229 price point is aimed at readers preferring to engage with digital content on their phones. Additionally, TCL unveiled the NXTPAPER 14 Pro, equipped with the same eye-friendly technology in a larger 14-inch display, targeting productivity users with its MediaTek Dimensity 8020 processor, 12 GB of RAM, a 12,000-mAh battery, and 256 GB of storage. ZTE 5G+AI Eyewear-free 3D Tablet Nubia introduced its latest devices on the ZTE stage, despite emphasising its independence from ZTE. The highlights include the Nubia Flip, the brand’s first foldable phone, featuring a 6.9-inch 120-Hz display that folds to a compact size and sports a unique circular screen on the front. Priced at $599, it offers a Snapdragon 7 Gen 1 processor and unique features like a 3D interactive pet and extensive customization options. Another significant release was the Nubia Pad 3D II, a tablet capable of displaying 3D content without glasses through eye-tracking technology. This new version introduces 5G connectivity and incorporates “AI concepts” for enhanced functionality, including dual cameras for 3D content creation and an AI feature that converts 2D to 3D content. Xiaomi 14 Smartphone Xiaomi unveiled its flagship Xiaomi 14 Ultra at the Mobile World Congress. This model enhances its predecessor’s capabilities, offering an unparalleled display, a more durable construction. However, the high cost and specific target market of photography enthusiasts might limit its appeal. The device, starting at €1,499, with an optional Photography Kit for €199, introduces HyperOS, an interface designed to refine user experience. Additionally, Xiaomi unveiled other devices including the Xiaomi Pad 6S Pro, Xiaomi Smart Band 8 Pro, and the Watch S3 with HyperOS, alongside the Xiaomi Watch 2 running on Google’s Wear OS. While these products won’t reach the US market, their launch in Europe demonstrates Xiaomi’s strategic global expansion. Humane AI The Humane Ai pin, introduced a few months ago, was on display at the MWC. The wearable device aims towards a future less dependent on smartphones. It is designed by former Apple employees, and aims for a screen-free existence, blending seamlessly into personal attire while offering sophisticated AI functionalities. Priced at $699, with a $24 monthly subscription for connectivity and AI services, the Ai Pin operates through voice and gesture interactions, supporting up to 50 languages and adapting to local languages automatically. Humane emphasises privacy with features like an LED indicator for the camera and encrypted data management, marking the Ai Pin as an innovative step towards integrating AI into daily life without screens.","excerpt":"MWC 2024 dazzled enthusiasts with a tech buffet, serving up everything AI can possibly do.","categories":["AI Trends"],"tags":["Top Trend"],"author_name":"K L Krithika","publish_date":"2024-02-29T17:05:30","publication_year":"2024","word_count":1225,"keywords":["Top Trend","Go","artificial intelligence","AI","ML","Git","RAG","Aim","ViT","GAN","R"],"extracted_tech_keywords":["AI","artificial intelligence","ML","Aim","RAG","R","Go","Git","GAN","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/10-futuristic-gadgets-announced-at-mwc-2024\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10054688,"title":"This ML Model Can Help Robots Learn About The Relationships Between Objects","content":"MIT researchers have developed a new machine-learning model. As a result, the robots may understand the world’s interactions similarly to how humans do. Numerous deep learning models struggle to see the environment because they are unaware of the intricate interactions between individual items. For example, without this understanding, a robot built to assist someone in the kitchen would have difficulties following commands such as “take up the spatula to the left of the stove and place it on top of the cutting board.” Composing Visual Relationships Researchers at MIT devised a model that comprehends the fundamental interactions between items in a scene to address this issue. Their model describes specific relationships one by one and then combines them to explain the entire picture. This research could be applicable in scenarios where industrial robots are required to do complex, multistep manipulation tasks, such as stacking things in a warehouse or assembling appliances. Additionally, it advances the field toward enabling robots to learn from and interact with their environments like how people do. “When I look at a table, I cannot assert that an object exists at the XYZ place. Our minds do not operate in this manner. When we comprehend a scene in our minds, we do it based on the relationships between the objects. We believe that by developing a system that can comprehend the relationships between objects, we will be able to manipulate and change our environments more effectively,” says Yilun Du, PhD Student, Computer Science and Artificial Intelligence Laboratory (CSAIL), MIT. One Relationship at a Time The researchers built a framework to build an image of a scene from a text description of objects and their connections, such as “A wood table to the left of a blue stool.” To the right of a blue stool is a crimson couch.” Their method would deconstruct these words into two smaller components that explain each unique relationship and then model each component independently. These components are then integrated to create an image of the scene via an optimisation process. Source: Learning to Compose Visual Relations The researchers represented the individual object interactions in a scene description using a machine-learning technique called energy-based models. This technique enables them to encode each relational description using a single energy-based model and then combine them in a way that infers all objects and relationships. In addition, by segmenting the lines into shorter bits for each relationship, the system may recombine them in various ways, Li explains, making it more adaptable to new scenario descriptions. “Other systems would consider all relationships holistically and generate the image from the description in a single shot. However, these approaches fail when we have out-of-distribution descriptions, such as those with more relations, because these models cannot truly adapt to generate images with more relationships from a single shot. However, by combining these smaller, independent models, we can model a greater number of relationships and respond to unique combinations,” Du explains. Additionally, the system works in reverse – given an image, it can generate text descriptions that correspond to the relationships between the scene’s objects. Additionally, their model can alter an image by rearranging the scene’s elements to correspond to a new description. Recognising Complex Scenes The researchers compared their model to those generated by other deep learning approaches when given written descriptions and instructed to generate images displaying the associated items and their relationships. Their model outperformed the baselines in each case. Additionally, they asked humans to assess if the generated photos matched the scene description. Ninety-one per cent of participants concluded that the new model performed better in the most difficult situations, where descriptions included three linkages. Additionally, the researchers presented the model photographs of previously unseen scenes and various alternative text descriptions for each image. It was able to correctly pick the description that best fit the item relationships in the image. Furthermore, when the researchers provided the system with two relational scene descriptions that depicted the same image in distinct ways, the model recognised that the descriptions were equal. “One of the outstanding fundamental problems in computer vision is developing visual representations that can deal with the compositional nature of the world around us. This article makes substantial progress toward resolving this issue by providing an energy-based model that explicitly represents the numerous relationships between the items displayed in the image. The results are quite remarkable,” says Josef Sivic, a distinguished researcher from Czech Technical University’s Czech Institute of Informatics, Robotics, and Cybernetics who was not involved in this research.For more information, refer to the article.","excerpt":"MIT researchers have developed a new machine-learning model. As a result, the robots may understand the world’s interactions similarly to how humans do. Numerous deep learning models struggle to see the environment because they are unaware of the intricate interactions between individual items. For example, without this understanding, a robot built to assist someone in […]","categories":["AI Trends"],"tags":["AI (Artificial Intelligence)","Computer Vision","Deep Learning","industrial robots","Machine Learning","machine learning model","machine learning models","MIT","MIT CSAIL","Robotics"],"author_name":"Dr. Nivash Jeevanandam","publish_date":"2021-12-02T12:00:00","publication_year":"2021","word_count":760,"keywords":["artificial intelligence","programming_languages:R","AI","MIT","Machine Learning","machine learning models","industrial robots","Robotics","MIT CSAIL","computer vision","ai_applications:computer vision","machine learning model","deep learning","Computer Vision","ai_applications:robotics","Deep Learning","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","deep learning","computer vision","R","programming_languages:R","ai_applications:computer vision","ai_applications:robotics"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/ml-model-can-help-robots-learn-about-the-relationships-between-objects\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":14156,"title":"Creating Corporate Database &#8211; Managing Spreadsheet Series: 5 of 5","content":"The last step in creating corporate data from spreadsheets is the creation of the corporate database itself. The corporate database is created when – The spreadsheet has been selected The spreadsheet has been logged in The spreadsheet has been run through spreadsheet disambiguation technology At this point the data and the metadata from the spreadsheet have been stripped from the spreadsheet. Fig 1 shows this progression – The data that is put in the corporate data base consists of – Column name Row id name Value Spreadsheet system name Date of processing Fig 2 shows the contents of the corporate data after processing – The column name and the row id name serve to identify the value. The spreadsheet system name identifies the source of the data, and the date filed determines what particular day the spreadsheet was processed. Spreadsheet system name and date of processing are needed to satisfy lineage requirements, and column name and row id name are needed to provide the metadata that is associated with the value. On occasion, it may be necessary to delete one or more entries into the corporate data. This is because sometimes spreadsheets contain extraneous or spurious data. The spreadsheet disambiguation program picks up ALL elements of data on the spreadsheet. So it is possible that unwanted data arrives in the spreadsheet corporate data input. If that is the case then those unwanted elements of data are “weeded out”. Fig 3 shows the weeding out of unwanted data before it is placed into corporate data. One of the features of corporate data is that it usually is arranged by subject area. But when data comes out of spreadsheet disambiguation, it comes out as it was laid out on the spreadsheet. For this reason, it is often convenient to divide the corporate data by subject area before finalizing the corporate database. Fig 4 shows this activity. A final consideration of corporate data coming from spreadsheets and corporate data coming from other sources is that once the spreadsheet data has been cast into the form of corporate data, it can be freely mixed with other corporate data. Fig 5 shows this capability. The ability to integrate spreadsheet data with other corporate data easily is one of the major advantages of moving spreadsheet data to the corporate data environment. The analyst finds this capability to be very useful. The integration is achieved by using the column name\/row id metadata and comparing it to the metadata that is already found in the corporate data environment. Fig 6 shows this interaction. Bill Inmon – the “father of data warehouse” – has written 57 books published in nine languages. Bill was named by ComputerWorld as one of the ten most influential people in the history of the computer profession. Bill lives in Castle Rock, Colorado. Bill’s latest book is TURNING TEXT INTO GOLD, Technics Publications, a book that shows how text can be turned into business value. TURNING TEXT INTO GOLD is available on Amazon.com.","excerpt":"The last step in creating corporate data from spreadsheets is the creation of the corporate database itself. The corporate database is created when –   The spreadsheet has been selected   The spreadsheet has been logged in   The spreadsheet has been run through spreadsheet disambiguation technology At this point the data and the metadata from the spreadsheet […]","categories":[],"tags":[],"author_name":"William Inmon","publish_date":"2017-04-09T03:51:06","publication_year":"2017","word_count":498,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","ViT","R","data warehouse"],"extracted_tech_keywords":["AI","R","Go","data warehouse","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/creating-corporate-database-managing-spreadsheet-series-5-5\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10131741,"title":"Bengaluru’s Control One is Bringing Forklifts to Life with Physical AI Agents","content":"Within the AI landscape of India, with the minimal funding that startups get, it is hard to get into the sector, especially if you are getting into robotics. But backed by the power of jugaad, a team of experts from Tesla, and advisors from India, Bengaluru-based Control One has taken up the task of building the next Boston Dynamics from the country. The key to Control One’s innovation lies in its focus on vision-based Visual SLAM (Simultaneous Localisation and Mapping) systems instead of traditional LiDAR technology. “Using vision, we try to make the machine understand its surroundings and navigate accordingly,” said Pranavan S, the founder and CEO, in an interaction with AIM. Recently, this one-year-old startup unveiled India’s first physical AI agent for the global market, which was installed on a forklift. With the integration, the forklift could interact with its surroundings and execute tasks based on simple voice inputs. The company calls it a 3-brain system, where operators can manage the AI decision remotely in real-time. Visual SLAM, or vSLAM, helps the robot navigate using reference points in the environment, just like humans do. Pranavan said what separates Control One from every other robotics company in India is the focus on building an operating system for robotics. Currently, the forklift operates only on the ground but Pranavan said they aim to operate it in three directions, which includes picking up objects and placing them. Founded in 2023, Control One attracted investment in April and raised $350k from industry leaders like iRobot co-founder Helen Greiner, CRED  founder Kunal Shah, and executives from Tesla, Walmart, and General Electric. Last year, NVIDIA Research unveiled Eureka, an AI agent that automatically generates algorithms for training robots using GPT-4 prompts and reinforcement learning. It seems like that is the future when physical AI agents will be everywhere. OS for Robots But when it comes to Control One, “We are building an OS that can understand the environment around it, akin to human behaviour,” explained Pranavan. “For example, we humans navigate dynamically, even in crowded and ever-changing environments, using our vision and depth perception to take predictive actions. This is what we aim to achieve with our machines.” Control One’s OS, dubbed ‘One AI’ can be seamlessly integrated into existing warehouse equipment like pallet movers and forklifts. These systems autonomously navigate and improve performance over time, setting a new standard for intelligent material movement. Currently only operated using remotes, Control One aims to introduce prompts in natural language in the future, but the issue is around the regulations of the same. Pranavan elaborated on the advantages of vision systems over LiDAR. “LiDAR can detect obstacles, but it doesn’t understand what those obstacles are or their movement dynamics. vSLAM, on the other hand, can analyse pixel vectors to determine if an object is approaching, moving away, or crossing paths, allowing the robot to make more informed decisions.” Despite their focus on vision systems, Control One acknowledges the complementary role of LiDAR for safety. “We use LiDARs to meet market safety standards, ensuring our robots can stop immediately if they get too close to an object,” said Pranavan. Though the company does training in the real-world, simulation also accounts for around 30% of the training. “The problem with simulation is that the AI would work in the settings you put it in, but when transferred into the real world, uncertainties such as uneven floor can also affect the physical agent’s performance,” said Pranavan. The Boston Dynamics of India While there are companies such as DiFACTO Robotics, Novus Hi-Tech, and Gridbots Robotics, Control One is focused on building the whole stack of a robot using physical AI agents. The most interesting part is that all of this happens on-edge using NVIDIA A2000 GPUs. “We started with NVIDIA’s A5000 series but moved to the 2000 series to reduce power consumption,” said Pranavan. “Our next iteration will consume less than 80 watts and run entirely on battery, extending its operational life.” This efficient use of power is crucial for Control One’s vision-based system, which demands significant computational resources. “Our current demo operates at 280 TOPS of AI computing power, plus it runs on an AMD processor. The next iteration will be more power-efficient while maintaining high performance.” Pranavan, with experience of building SP Robotics Works as a co-founder, said that frugal innovation has always been part of what Indian startups do, explaining how he started figuring out how a 128 MB RAM computer worked back in his young days. He said that to compete with Boston Dynamics, they need at least $500 million in funding, but “given our capital efficiency and cost factors, we would be able to do with $100 million as well”. Giving AI to Blue Collar Jobs Control One is focusing on the warehousing sector, where there is a significant labour shortage and a high demand for automation solutions, especially in the global market. “The warehousing market is facing a huge labour crisis. Our system enables one person to manage multiple robots, effectively multiplying their productivity,” said Pranavan. One AI is designed to manage entire fleets of robots, enabling remote operation and supervision. “Our ultimate goal is to allow one person in a remote location to manage an entire warehouse in another part of the world. OneAI will be the interface between humans and machines, helping operators make informed decisions.” With a vision of building a humanoid, maybe in the future, Pranavan said that you don’t necessarily need to build a humanoid to operate machinery in factories. “We are operation system makers, not humanoid makers, yet,” said Pranavan, as he says that he is looking to partner with OEMs in the future, with several already under way.","excerpt":"Pranavan said that to compete with Boston Dynamics, the company needs at least $500 million in funding, but “given our capital efficiency and cost factors, we would be able to do with $100 million as well”.","categories":["AI Features"],"tags":["AI Agents","bangalore","Bengaluru"],"author_name":"Mohit Pandey","publish_date":"2024-08-07T15:10:10","publication_year":"2024","word_count":950,"keywords":["Go","API","AI","innovation","ML","AI Agents","automation","GPT","Aim","ViT","Bengaluru","R","bangalore"],"extracted_tech_keywords":["AI","ML","Aim","R","Go","API","GPT","ViT","automation","innovation"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/bengalurus-control-one-is-bringing-forklifts-to-life-with-physical-ai-agents\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10081614,"title":"Forget Python, JS, Web3 Languages will Now Shape Developers’ Future","content":"Calls for Web3 have become increasingly louder with advocates of the umbrella buzzword promising an ambitious future of the financial world and the internet. Needless to say, a new version of the internet will affect developers’ lives in more ways than one. As the demand for Web3 projects picks up, developers are pushing to fill the gap. Despite this movement, there is a lack. According to a report released by crypto investment firm Electric Capital at the end of August, Web3 development has been stymied by a dearth of developers. At the end of last year when Bitcoin and a bunch of other cryptocurrencies touched their highest values, there were 18,500 developers participating in open-source Web3 projects. Source: Stack Overflow Resurgence in Rust Even as that number is reported to be at its highest and growing faster than ever (around 60% of Web3 developers entered the industry last year), it is still a small portion of the total number of software engineers globally which stood at 31.1 million in the first quarter of 2022, according to developer analytics firm SlashData. Nevertheless, there has been a noticeable shift. The shortage has been marked by a need for knowledge of specialised programming languages used to create smart contracts on Ethereum and Solana. This change in direction is also evident in the emergence of certain programming languages like Rust. In 2020, Rust popped up as the most-loved programming language. According to a Stack Overflow survey, Rust was picked as the most popular with 86% of developers choosing it over industry standards like Python and TypeScript. While Rust first appeared 12 years ago in 2010, it has stayed around the top of the survey since 2016. Source: Quora Security advantage It is interesting to note that Rust has become a favourite among Web3 developers as well, because of its ease of use and increased security. Where every programming language has a reputation for a different use case, Rust is excellent in multi-platform environments and high-performance areas where security is a primary requirement. This hyperfocus around security is seen especially in blockchain, browsers, operating systems and cybersecurity products. Rust helps users write codes with fewer errors because of its compiler that checks everything. The language is good for writing secure time-saving abstractions, which is one of the main principles in object-oriented programming. This also eliminates the need for immutability – one the foundational elements of foundational programming. Immutability is essential for developers to write safer and cleaner code. For these reasons, blockchain and functional programming are a naturally great fit with each other. Rust has also gained acceptance among developers working with cryptocurrency and blockchain since it also delivers tools for them to start shipping their code much faster. This level of stability and safety safeguards against, say, a cryptocurrency crashing suddenly. This is especially vital considering blockchains and cryptocurrencies almost always have the worst case scenario of bugs. Critical bugs normally can’t be discussed openly since they affect live systems. However, they also have to be deployed at the same time across several participants, without the support of a reliable third party. This is why there are limited and covert ways to fix them. Also, blockchains secure actual financial value incentivising hackers to find bugs in those systems. Source: Hiro.so New languages like Clarity and Solidity Rust isn’t the only language on the minds of Web3 developers. A couple of new languages like Clarity and Solidity have also come up of late. Clarity, is a programming language, especially for creating smart contracts and decentralised applications or dApps in the Bitcoin Web3 ecosystem. Like Rust, Clarity is also built to make assets on blockchains as safe and predictable as possible. DApps designed on Clarity are published on the programming layer for Bitcoin, Stacks. Solidity, the first-ever programming language made for Ethereum specifically, is by far the most widely used language in Web3. Solidity first gained acceptance due to its first-mover advantage in the space. Proposed by Ethereum co-founder Gavin Wood in 2014, the language was officially released in 2018. Ethereum researcher Christian Reitwiessner led the team and eventually formed the Solidity team. The language can be used to create dApps on the Ethereum blockchain and developers can use it on blockchains compatible with the Ethereum Virtual Machine or EVMs. Currently, Solidity is the most-popular Web3 language with more than 4,000 monthly active developers using it on just the Ethereum blockchain even as Rust appears to be catching up. Clarity, which came up after Solidity, has been built on the flaws of its predecessor. Clarity has a different basic approach in terms of design and prioritises security and predictability even more than Solidity. Clarity also enables users to settle contracts on the Bitcoin blockchain itself, which is the safest and decentralised blockchain in use today.","excerpt":"This change in direction is also evident in the development of certain programming languages like Rust","categories":["AI Trends"],"tags":[],"author_name":"Poulomi Chatterjee","publish_date":"2022-12-06T16:00:00","publication_year":"2022","word_count":799,"keywords":["Go","API","AWS","AI","TypeScript","Python","ViT","analytics","Rust","R"],"extracted_tech_keywords":["AI","analytics","AWS","Python","R","TypeScript","Go","Rust","API","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/how-web3-is-shaping-the-emergence-of-programming-languages\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":22837,"title":"Dawn Of Cryptocurrency AI Agents: Trading Crypto Using Reinforcement Learning","content":"One day, the term ‘hard cash’, and other financial instruments will soon be obsolete, thanks to the aggressive growth in cryptocurrency and its flexibility with regard to transactions. Though it comes with a downside of falling into the wrong hands, causing a financial disaster in larger proportions, its popularity cannot be denied. As far as research in the field is concerned, the study of cryptocurrency has been shied away, largely because it is unregulated and no single government body has direct control over it. On the other hand, artificial intelligence and machine learning enthusiasts are beginning to explore trading cryptocurrencies using techniques such as Reinforcement Learning (RL), meta-learning among many others, to make it easier for research purposes as well as making it beneficial for the betterment of society. Artificial Intelligence for Cryptocurrency Predictions Cryptocurrency platforms will soon reap the benefits offered by artificial intelligence(AI) and machine learning(ML). The year 2017 saw the rise of cryptocurrencies in magnanimous amounts. In the near future, it will be the norm of monetary payments if regulated legitimately. For example, Coinbase, the popular exchange platform has come up with an added security functionality to verify users’ identities as well as monitor market predictions. Another instance of AI presence in trading cryptocurrencies is by VantagePoint, an AI-assisted forecasting software used to predict market prices (low and high) of cryptocurrencies along with additional information such as equities and commodities. A typical analysis involves highlighting  the top popular markets in cryptocurrencies which is then fed into the company’s proprietary software. The software implements a non-linear ‘weight matrix’ to generate forecasts to match the target market cited by the user. Lane Mendelsohn, Vice President of VantagePoint Software emphasizes that AI is the key to attract more trading opportunities among traders and investors through cryptocurrencies. This seems lucrative but comes with downsides, such as computing limitations (hardware and processing power), trust among the public, and limited data available in cryptocurrencies unlike the traditional stock market which has vast amounts of tangible data. Therefore, given the time and effort, it is certain that AI will overcome these setbacks and completely predict the digital market. The companies listed above have shown keen interest in incorporating AI agents in some or the other form. “Intelligent Agents” as they are called, handle the complexity of tasks or the level of intelligence required for the task. It all depends on how rational the intelligent systems should be when it comes to predicting data. Be it performance, perception of trading environment or previous trading knowledge. The tactics of using Reinforcement Learning on a research perspective Reinforcement Learning(RL), which is a facet of ML and AI can be used to predict cryptocurrency markets. Since there is limited work available for research purposes, we can use the concept of RL to optimise and predict these volatile markets. RL takes the form of a Markov Decision Process(MDP) tree since there are many factors at play in trading. MDP is adopted because the trading environment is highly variable and volatile. It is chosen to work with the current state (in this case, trading market) rather than working with past data to optimise a decision.The decision system is given below: Markov Decision Process Overview (Image Courtesy : SFL Scientific) Agent : The agent for this scenario is the trading agent — For example, Unocoin, an Indian cryptocurrency exchange platform employees trading agents who make decisions based on the live market. Environment : One could say that the environment is the exchange platform itself. But, on hindsight, the exchange has human as well as algorithmic agents acting in the environment. At this stage, it is assumed that we only consider a single environment to eliminate complexity. State : The current state of the market is only considered. This again forms a Partially Observable Markov Decision Process (POMDP), which means the agent observes only a part of the actual state of the environment. Action: In RL, there is a contrast between discrete and continuous variables. It depends on how far the user is willing to take up complexity in terms of algorithmic factors. Usually, in this scenario, we consider “buy” and “sell” as two action spaces. Reward : The possibilities for choosing a reward function are numerous. The user can concentrate on the profit margin to obtain positive feedbacks than negative feedbacks fed to the system. Now that we have a framework as to what goes into learning, we list down the key techniques that a user can utilise to enforce proper reinforcement algorithm. End-to-end optimisation: RL optimises the reward function, which is specified by the user. For example, profit\/loss trend can be combined to the reward function to optimise larger monetary gain. Policy-learning: Instead of coding a rule-based policy, RL can learn the policy itself to improvise over time with respect to the model. A good example is the inclusion of a deep neural network. Flexibility to model other agents: RL can also be extended to other agents, if incorporated, in the project. For example, the market trend can also be observed along with price details. This is just a handful of techniques in an ocean of learning algorithms. Also, the numerical computations is not covered in this article. The theme of the article here is to how AI and RL can prove to be beneficial when it comes to finance in the digital space. Ultimately, cryptocurrency trading is a vast phenomenon and people might be baffled with the market fluctuations happening in real-time. It is recommended that they master the basics to get a solid grip of the cryptocurrency arena.","excerpt":"One day, the term ‘hard cash’, and other financial instruments will soon be obsolete, thanks to the aggressive growth in cryptocurrency and its flexibility with regard to transactions. Though it comes with a downside of falling into the wrong hands, causing a financial disaster in larger proportions, its popularity cannot be denied. As far as […]","categories":["IT Services"],"tags":["bitcoin stock","Cryptocurrency","cryptocurrency bitcoin","cryptocurrency ethereum","cryptocurrency trading","Ethical Hacking"],"author_name":"Abhishek Sharma","publish_date":"2018-03-22T09:26:42","publication_year":"2018","word_count":930,"keywords":["cryptocurrency trading","Go","meta-learning","artificial intelligence","machine learning","Cryptocurrency","AI","neural network","ML","R","Git","bitcoin stock","cryptocurrency bitcoin","Rust","Ethical Hacking","cryptocurrency ethereum"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","neural network","R","Go","Rust","Git","meta-learning"],"url":"https:\/\/analyticsindiamag.com\/it-services\/cryptocurrency-trading\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10168511,"title":"Microsoft’s 365 Copilot Receives New Features in a Major Update","content":"Microsoft announced an update to the Copilot 365 application on Wednesday, which includes numerous new features. The platform now offers an ‘AI-powered’ search feature that incorporates data from various third-party sources, including Google Drive, Slack, Jira, and others. The company also announced a new Create feature that supports image generation, powered by OpenAI’s latest GPT-4o capabilities. Additionally, the redesigned app now defaults to a chat-based interface, as demonstrated in a video published by the company on YouTube. Furthermore, the company announced a new Agent Store where users can integrate agents developed by Microsoft’s partner companies. This announcement is in addition to the recently introduced Microsoft Researcher and Analyst agents. The Researcher feature is similar to the deep research tools offered by today’s AI models and platforms. This feature combines OpenAI’s deep research model with the capabilities of Microsoft 365 Copilot. Additionally, it can integrate third-party data from platforms like Salesforce, ServiceNow, and Confluence. The ‘Analyst’ feature is designed to function like a “skilled data scientist.” This feature is built on OpenAI’s o3-mini reasoning model and can run Python to execute data queries, allowing users to view the code the tool runs in real time. Source: Microsoft Microsoft also announced Copilot Notebooks, a feature that integrates chats, files, and recordings to deliver real-time insights and audio summaries, keeping users updated as data changes. “For example, I can collect all the latest things I’m reading about agents and agent frameworks, and then I can listen to it,” said CEO Satya Nadella, demonstrating the Copilot Notebooks feature on X. Alongside these new features, Copilot is also equipped with memory and personalisation features. These ensure that the tool ‘learns’ the user preferences, to provide tailored responses. Recently, Microsoft also announced autonomous computer use capabilities in the Copilot Studio. This feature is accessible through an early access research preview. This can navigate through a user interface and perform actions such as clicking buttons, selecting menus, and typing text into fields.","excerpt":"Microsoft announced an update to the Copilot 365 application on Wednesday, which includes numerous new features.  The platform now offers an ‘AI-powered’ search feature that incorporates data from various third-party sources, including Google Drive, Slack, Jira, and others. The company also announced a new Create feature that supports image generation, powered by OpenAI’s latest GPT-4o […]","categories":["AI News"],"tags":["AI (Artificial Intelligence)","Microsoft Copilot"],"author_name":"Supreeth Koundinya","publish_date":"2025-04-24T11:16:40","publication_year":"2025","word_count":325,"keywords":["Go","OpenAI","AI","programming_languages:R","GPT-4o","Python","GPT","Microsoft Copilot","programming_languages:Python","R","AI (Artificial Intelligence)","llm_models:GPT"],"extracted_tech_keywords":["AI","GPT-4o","OpenAI","Python","R","Go","GPT","llm_models:GPT","programming_languages:Python","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/microsofts-365-copilot-receives-new-features-in-a-major-update\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10071909,"title":"Can TikTok be the Giant Slayer?","content":"The earnings week has been a bloodbath for Big Tech. A slowdown in digital ad spending caused by rising inflation has dampened the report cards for Google, Snap Inc and Twitter, all of which rely heavily on ad revenue. However, none has fallen harder than Meta. Facebook’s parent company Meta posted its first-ever revenue decline after a stupendous run for years. The company posted its quarterly revenue numbers of USD 28.8 billion, which is 1 per cent lower than last year’s revenues. The figures also fell short of market expectations which stood at USD 28.9 billion. Moreover, Reality Labs unit, which is behind the company’s metaverse push, generated sales of USD 452 million, down from USD 695 million in the last quarter. Source: Meta, Meta’s second-quarter results Macro-economic factors aside, it was no secret that Meta was struggling with growing competition from rival TikTok. Meta’s pivot to catch up ByteDance Ltd.’s TikTok, unlike Facebook, isn’t based on the social-graph model. TikTok doesn’t need influential people to use it to enhance engagement, nor does a user need to have friends on it already to find it useful. The app runs on an excellent recommendation algorithm that can determine a user’s interest area with great accuracy in no time. The TikTok model did not require building a giant social network like Meta or Twitter; it simply exists to distract a user. Considering it has random short videos catering to a user’s interests, the platform’s addictiveness can be gauged by the sheer number of its active users. Currently, TikTok has around a billion monthly users with an engagement time of about 10.85 minutes, making TikTok the most engaging app ever. Even more impressive is the fact that the platform reached these figures in a short span of four years. Facebook’s fears that they weren’t doing enough to compete against TikTok left senior executives desperate. In April, The Verge obtained a memo from Tom Alison, Meta’s executive in charge of Facebook stating that Facebook feed would be built much like TikTok’s, i.e. fewer posts from people that users followed and more recommended posts regardless of where they came from. Sources: SensorTensor 1, SensorTensor 2, InfluencerMarketingHub, Mediakix, TikTok, Statista 1, Statista 2, AppAnnie Meta had already incorporated their Reels, a short video feature from Instagram onto Facebook. Akin to TikTok, Reels had tools which allowed users to edit their videos, add chosen audio clips and AR effects. With heavy promotion of Reels, Instagram had already transformed its interface to appear as close to TikTok as possible. However, these efforts proved to be insufficient, post which Meta shifted the feature to Facebook with the intention of bringing youngsters back to the app. This is not the first time that Facebook has rushed to copy the algorithm of another competitive app. Earlier, Facebook had copied the Stories format from Snapchat. During the earnings call with analysts, Zuckerberg revealed that Reels could potentially be monetised faster than Stories, but it wasn’t making them money at the moment. But even with these changes, Meta recognised the long-term dangers with moving away from a connection-centric model. Source: TechCrunch The Cunningham memo In mid-July, an internal document called the Cunningham memo re-surfaced in the news. The document signalled that in addition to being anxious about its rivals, Zuckerberg was also worried about subsidiaries WhatsApp and Instagram cannibalising Facebook’s business. A former data scientist from Facebook, economist Thomas Cunningham, made three noteworthy points in the memo. First, a single messaging or sharing app tends to dominate in a market across all ages, genders and classes. There was only one group that eluded this rule: teens. An app could succeed singularly among teens only. This is what seems to have happened in the case of TikTok. A research done by Ofcom showed that teens were turning their backs on traditional sources of news and consuming their news from TikTok. The number of users on the app consuming news had risen from 800,000 in 2020 to 3.9 million this year. The Ofcom report noted that this growth was mainly powered by users in the ages between 16 and 24. Second, general messaging apps usually have a bimodal distribution in a market. If an app dominates more than 50% of the market, it will tend to veer towards 90% market coverage. On the other hand, if an app dominates less than 50% of a market, it tends to move towards 1 per cent market coverage eventually. Third, small declines in the market share of a social networking platform is indicative of big declines to come in the future. “Once users start using a social app, their use declines slowly,” Cunningham added. This hypothesis was reiterated when Snap entered the competition and caused a dent in Facebook’s business and now TikTok. Source: Instagram Future model In the current climate, Facebook is stuck between a rock and a hard place and the stakes are high. Sticking to the current model could mean succumbing to hypercompetitive apps like TikTok. But changing its original model and moving away from being a social network would mean it is in direct competition with apps like TikTok. Apps like TikTok, which have no underlying principle binding it together, might be a flash in the pan once newer competition arrives. New apps that are more interactive and more distracting could dismantle TikTok. But this heated contest might produce one collateral result — the end of the dominance of social media giants like Facebook and an open field for smaller internet companies.","excerpt":"Facebook’s parent company Meta posted its first-ever revenue decline after a stupendous run for years.","categories":["IT Services"],"tags":[],"author_name":"Poulomi Chatterjee","publish_date":"2022-08-01T15:00:00","publication_year":"2022","word_count":918,"keywords":["Go","programming_languages:R","AI","Modal","programming_languages:Go","Git","RAG","CLIP","R"],"extracted_tech_keywords":["AI","RAG","R","Go","Git","CLIP","Modal","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/can-tiktok-be-the-giant-slayer\/","complexity_score":4,"technical_depth":9,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10074371,"title":"IIT Kharagpur Develops No-code 360-degree VR Platform for Teachers","content":"IIT Kharagpur researchers have developed a 360-degree VR platform for educators to bridge the gap between teachers and students. “We want to redefine the dependence of teachers on others to develop immersive content,” Kaushal Kumar Bhagat, assistant professor at the Indian Institute of Technology, Kharagpur, said. During the pandemic, Bhagat and a team of researchers became acutely aware of the lack of interaction in online classes. This gave them the idea of developing a no-code 360 degree VR platform for teachers. Bhagat believes that VR can potentially make learning more interactive and engaging through computer-generated graphics in a three-dimensional environment, easing complexities. It took almost 18 months to develop an open-source 360-degree VR platform for educators, explains Bhagat. Moreover, it does not require any prior coding experience. Teachers can freely add audio, create quizzes and bring in gaming elements to make the lessons interesting. It does not require additional skills. The researchers said all they need is to shoot photos in 360-degree, follow the instructions, and create a lesson ready to be experienced in VR. This VR platform works on all Android devices. As to how immersive the experience would be, depends also on the type of the VR headset used by the students. The key idea is to make it reachable to every student in urban and rural India, the team of researchers affirmed.","excerpt":"Researchers believe that VR can potentially make learning more interactive and engaging through computer-generated graphics in a three-dimensional environment","categories":["AI News"],"tags":["IIT","iit Kharagpur","Technology"],"author_name":"Nidhi Bhardwaj","publish_date":"2022-09-05T17:17:04","publication_year":"2022","word_count":225,"keywords":["programming_languages:R","AI","RAG","Technology","iit Kharagpur","IIT","R"],"extracted_tech_keywords":["AI","RAG","R","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/iit-kharagpur-develops-no-code-360-degree-vr-platform-for-teachers\/","complexity_score":2,"technical_depth":4,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":33271,"title":"The Best Resources For Learning Deep Learning For Beginners","content":"Over the last few years, Deep Learning has proven itself to be the game-changer. This area of data science is the only one responsible for advancements in machine learning and artificial intelligence. From academic researches to self-driving cars, Deep Learning is found in all possible aspects nowadays. Deep Learning is a complex and vast field that consists of several components. It cannot be mastered in a day and hence it will take several months if you want to dig deeper into this field. Core knowledge of linear algebra and calculus is very important before learning this area. This article explores the basics of deep learning for beginners. 1| Basic of Machine Learning This is the path that will lead you to the door of deep learning. It basically consists of three types of learning, supervised, unsupervised and reinforced learning. Techniques like linear regression, logistic regression are required in deep learning. If you are familiar with machine learning and want to try your hands on you can click here. 2| Introducing Deep Learning The first task is to know the frameworks of deep learning. Deep Learning is mainly concerned with algorithms that are inspired by artificial neural networks. Various courses and online videos are available nowadays for understanding Deep Learning that you must not pass on this learning opt. Here is an online course for you to understand this subfield of machine learning more precisely. Avail here. If you like to know more about deep learning, you can go through this book. This book by Michael Nielson is a great start for learning from the initials. Click here for the book. 3| Knowing Neural Networks A neural network contains a layered design that includes an input layer, an output layer, and a hidden layer. It functions as the neurons in the human brain like receiving inputs and produce an output. You must know how the data can be handled as well as pre-processed, regularisation techniques, hyperparameter technique, data augmentation, etc. The functions of the artificial neural networks are used in deep learning that helps in speech recognition, image recognition, etc. If you want to know more about Neural Networks, please go ahead and click here. 4| Basics of Convolution Neural Networks Image Source: towardsdatascience.com Convolution Neural Network starts playing a crucial role in deep learning. It is widely used image recognition and classifications, object detections, facial recognition, etc. Basically, it takes an image as an input, process and classifies under certain attributes. In deep learning, CNN models are used to train and test in such a way that an input image will pass through a series of convolution layers with layers, pooling, fully connected layers and then classifying the object with probabilistic values between 0 and 1. For more information on CNN, please click here. 5| Understanding The Sequence Models It is important to understand how to build the models like Recurrent Neural Networks (RNNs) and using the variants like Gated Recurrent Unit (GRU) and Long Short Term Memory (LSTMs) if you want to dig deep a little more. Learning these will help you in audio applications, speech recognition, music synthesis, etc. Want to know more about sequence models. Click here. 6| Unsupervised Deep Learning This is a complex one but has many advantages and also has the potential to unlock the unsolvable problems that were done previously. Unsupervised Deep Learning is used to solve the issues created by supervised learning like biasing and the numerous manual efforts to make the algorithms work. Autoencoder neural network is an unsupervised deep learning algorithm. This tutorial from Stanford will help you understand unsupervised deep learning a little more. 7| Learning Natural Language Processing Image Source: Flickr This domain is concerned with understanding human languages and has emerged high nowadays with many benefits. It deals with constructing computational algorithms to automatically analyzation and representation of human language, can also be used for machine translation, dialogue generation, etc. The link provided will help you to understand NLP more precisely. Go ahead and click here. 8| Deep Reinforcement Learning We came to know the potential of deep reinforcement learning when the reinforcement learning algorithm is combined with deep learning method and AlphaGo was created, that defeated the strongest Go players. Here is a quick guide to learn more about Deep Reinforcement Learning. 9| Generative Models This is a powerful model for learning any distribution of data using unsupervised learning. This model is used to generate new data points with some variations by learning the true data distributions. There are two most efficient methods, they are Generative Adversarial Networks (GAN) and Variational Autoencoders (VAE). To know more about the model, click here. 10| Make It Practical To implement deep learning, you must know how to use Python. It’s totally cool if you are not aware of implementing Python. We are here to help. Click here. There are loads of libraries available where you can try your hands on and be the master. Also, if you plan to improve your deep learning skills and want to build your own deep learning server from scratch. Please go ahead and click here.","excerpt":"Over the last few years, Deep Learning has proven itself to be the game-changer. This area of data science is the only one responsible for advancements in machine learning and artificial intelligence. From academic researches to self-driving cars, Deep Learning is found in all possible aspects nowadays. Deep Learning is a complex and vast field […]","categories":["AI Highlights"],"tags":["AI for Beginners","Deep Learning","what is artificial intelligence for beginners"],"author_name":"Ambika Choudhury","publish_date":"2019-01-11T07:21:09","publication_year":"2019","word_count":852,"keywords":["what is artificial intelligence for beginners","data science","artificial intelligence","machine learning","TPU","AI","neural network","image recognition","NLP","deep learning","object detection","AI for Beginners","Deep Learning"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","deep learning","neural network","NLP","data science","image recognition","object detection","TPU"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/the-best-resources-for-learning-deep-learning-for-beginners\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10167090,"title":"GPT 4.5 Passes the Turing Test: Study","content":"The University of California, San Diego, unveiled a research study on Tuesday that claims to provide the “first empirical evidence that any artificial system can pass a standard three-party Turing test”. Alan Turing, a British mathematician and computer scientist, introduced the ‘imitation game’ in 1950, proposing that if an interrogator couldn’t distinguish between a machine and a human in text, the machine might possess human-like intelligence. In a three-party Turing Test, an interrogator converses with both a human and a machine to accurately identify the human. The research tested three AI models: OpenAI’s GPT-4.5, Meta’s Llama 3.1 405B, and OpenAI’s GPT-4o. Human participants engaged in five-minute test conversations with one human and one AI system using a split-screen interface. After each round, the interrogator selected the participant they believed was human. The AI models were evaluated under two conditions: a minimal instruction (NO-PERSONA) prompt and an enhanced PERSONA prompt that guided the AI to adopt a specific human-like demeanor. The results indicated that GPT-4.5 with the PERSONA prompt achieved a win rate of 73%, suggesting that interrogators often mistook it for a human. Llama 3.1‑405B with the PERSONA prompt attained a win rate of around 56%, whereas GPT‑4o under NO‑PERSONA conditions only reached a win rate of 21%. Interrogators primarily engaged in small talk—asking about daily activities and personal details in 61% of interactions, while also probing social and emotional aspects such as opinions, emotions, humour, and experiences in 50% of interactions. “If interrogators are not able to reliably distinguish between a human and a machine, then the machine is said to have passed [the Turing test]. By this logic, both GPT-4.5 and Llama-3.1-405B pass the Turing Test when they are given prompts to adopt a human-like persona,” read a section of the research study. The authors stated that these systems might seamlessly supplement or even replace human labour in economic roles that rely on brief conversational exchanges. “More broadly, these systems could become indiscriminable substitutes for other social interactions, from conversations with strangers online to those with friends, colleagues, and even romantic companions,” the authors added. OpenAI released the GPT-4.5 model in February, which was mostly appreciated for its thoughtful and emotional responses. Ethan Mollick, a professor at The Wharton School, said on X, “It can write beautifully, is very creative, and is occasionally oddly lazy on complex projects.” He even joked that the model took a “lot more” classes in the humanities. Been using GPT-4.5 for a few days and it is a very odd and interesting model. It can write beautifully, is very creative, and is occasionally oddly lazy on complex projects.Feels like Claude 3.7 while Claude 3.7 feels like GPT-4.5.— Ethan Mollick (@emollick) February 27, 2025","excerpt":"A UC San Diego study found that human participants frequently misidentified responses generated by OpenAI’s GPT‑4.5 along with Meta’s Llama‑3.1‑405B as coming from a human.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","OpenAI","Turing test"],"author_name":"Supreeth Koundinya","publish_date":"2025-04-02T11:55:35","publication_year":"2025","word_count":450,"keywords":["Go","OpenAI","AI","GPT-4o","ML","Turing test","GPT-4.5","GPT","Aim","ViT","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","ML","GPT-4.5","GPT-4o","OpenAI","Aim","R","Go","GPT","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/gpt-4-5-passes-the-turing-test-study\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10018143,"title":"Top Five Artificial Intelligence Predictions For 2021","content":"As AI becomes more ubiquitous, it’s also become more autonomous — able to act on its own without human supervision. This demonstrates progress, but it also introduces concerns around control over AI. The AI Arms Race has driven organizations everywhere to deliver the most sophisticated algorithms around, but this can come at a price, ignoring cultural and ethical values that are critical to responsible AI. Here are five predictions on what we should expect to see in AI in 2021: Something’s going to give around AI governance. Though regulation hasn’t reached a boiling point yet, AI governance will continue to be a hot topic in 2021. As AI becomes more pervasive, more and more stakeholders are waking up to the potential problems it introduces to the public. In response, organizations everywhere — from the most cutting-edge to the laggards — will be expected to deliver AI systems that are responsible, transparent, and unbiased. But whose responsibility is it to make sure this happens and regulates AI – the government, businesses, industry groups, or some combination? If businesses want to regulate themselves before the government does, they will have to take steps to ensure the data that feeds their AI is fair and unbiased, and that their models are empathetic, transparent, and robust. Organizations will also need to implement a way to closely monitor their AI — with a robust simulation capability and automated oversight –, so it doesn’t go off the rails as it “learns.” So far, businesses have come up short, and unless they make meaningful strides, government regulators will crank up the heat this year. Most consumers will continue to be sceptical of AI. With several big consumer brands in the hot seat around questionable AI ethics, most people still don’t trust AI. For many, it’s because they don’t understand it or even realize they’re using it daily. Consumers are getting so many AI-powered services for free — Facebook, Google, TikTok, etc. — that they don’t understand what they’re personally giving up in return — namely their personal data. As long as the general public continues to be naïve, they won’t be able to anticipate the dangers AI can introduce or how to protect themselves — unless the market better educates customers or implements regulations to protect them. Despite this, there’s some evidence that we’re turning the corner on AI’s trustworthiness. 81% of business leader respondents to Pega’s upcoming survey said they’re optimistic that AI bias will be sufficiently mitigated in five years. Businesses had better hope this turns out to be true – because as more of the public wakes up to how AI impacts their lives, and in some cases plays favourites, they will continue to ask harder questions that further erodes trust in AI, forcing businesses to have to answer to them. Digital transformation (DX) finds its moment. With COVID-19 spotlighting the drastic need for digital transformation (and in many cases fast-tracking efforts) in 2020, the trend will only continue in 2021. This year, businesses were forced to take those five-year DX vision plans and execute them in only five months to respond to the new realities brought on by the pandemic. After seeing this success, leaders wonder what other DX projects are possible that they once felt were out of reach. Given growing consumer queries, distributed workforces, and increased challenges, there will be more demand for automation in AI and close monitoring of those AI interactions. Hyper-automation will help streamline workflows in the post-COVID era. However, businesses will still need to understand what their AI is doing in order to predict it and monitor it, so they don’t disenfranchise any of their customers. Organizations will increasingly push AI to the edge. As computing and storage capability grows, more processing power and functionality will be accessed at the enterprise network edge. Our study found that 41% of business leaders believe extended edge use cases are highly dependent on the maturation of AI and other technologies like automation and machine learning, suggesting a complementary relationship between the two technologies. With the proliferation of the internet of things (IoT) devices and the increased adoption of 5G fueling this trend, computing power at the edge will grow, and the ability to leverage AI at the edge will grow too. To be successful, businesses will want to ensure everything on the periphery syncs up to a central brain for a holistic customer view. Isolating AI into edge computing silos would ultimately lessen the power of AI. Only by ensuring the edge is constantly connected to a central location will they be able to push the boundaries of what’s possible. 5. ModelOps will become the “go-to” approach for AI deployment. Much like the way DevOps has given structure to the way applications are deployed, ModelOps will reach a tipping point in 2021 as a way for mainstream businesses to better develop and operationalize their AI models. This will give them a more systematic way to quickly and responsibly develop, test, and deploy AI models more efficiently via the Cloud. Next year, more organizations will focus on effectively moving their AI models up the food chain with ModelOps while ensuring there’s enough structure for IT to put in guardrails. That process should not be so restrictive that agility and innovation are stifled – particularly with Citizen data scientists. In 2021, ModelOps will help organizations everywhere strike that right balance. There’s no doubt that AI will be in the spotlight in 2021. And scepticism is not going away. As AI becomes more pervasive, more structure and controls around processes and policies on AI, and how it is developed and used, are sure to emerge. With so much potential to help businesses and consumers everywhere, AI can change the game, if handled responsibly.","excerpt":"As AI becomes more ubiquitous, it’s also become more autonomous — able to act on its own without human supervision. This demonstrates progress, but it also introduces concerns around control over AI. The AI Arms Race has driven organizations everywhere to deliver the most sophisticated algorithms around, but this can come at a price, ignoring […]","categories":["AI Trends"],"tags":["AI Algorithms","what is artificial intelligence"],"author_name":"Suman Reddy","publish_date":"2021-01-15T15:18:12","publication_year":"2021","word_count":960,"keywords":["Go","machine learning","Rust","AI","ML","Git","RAG","AI Algorithms","what is artificial intelligence","edge computing","DevOps","R"],"extracted_tech_keywords":["AI","machine learning","ML","RAG","edge computing","R","Go","Rust","Git","DevOps"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-five-artificial-intelligence-predictions-for-2021\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10000944,"title":"Inside Ryuk Ransomware That Brought Down America’s Leading Publisher Tribune","content":"Recently, Tribune Publishing, publisher of some of the major US newspapers, reported about the Ryuk ransomware which disrupted the distribution of papers in the weekend of late December 2018. According to a news report, the attack delayed the distribution of Saturday editions of the Los Angeles Times and San Diego Union-Tribune and also stymied distribution of the West Coast editions of the Wall Street Journal and New York Times, which are printed at the Los Angeles Times’ Olympic printing plant in downtown Los Angeles. Source of the Attack: McAfee’s Assessment The company’s initial suspicion was that the attack originated outside of the country. However, specialists and investigators responded that it was too soon to confirm the source of the attack. Later, computer security software company Mcafee assessed the incident and investigated the source code of the malware confirming it’s a similarity with the older ransomware Hermes which hinted its origin to North Korea. Hermes ransomware has been used in the past by North Korean actors and the code blocks of Ryuk is very similar to Hermes. The McAfee Advanced Threat Research team was able to crack through the code and the characteristics of the malware, determining how the malware works, how the attackers operate, and how to detect it. The team hypothesized based on some of the known factors such as the technical indicators, cybercriminal characteristics, and evidence discovered on the dark web, that the Ryuk ransomware could be the result of a cybercriminal activity and may not be backed by a nation. Tribune had claimed that all data including the personal data of subscribers, online users, and advertising clients had not been compromised which only hints the intention behind the attack to be more focused at disabling the infrastructure as opposed to stealing information. Inside The Ryuk Ransomware The name: Ryuk is the name of a character in one of the popular Japanese Manga series “Death Note”. In manga comic, Ryuk drops a Death Note, which according to McAfee is a name that fits ransomware that drops ransom notes. The ransomware was named by its developer who apparently is a big fan of the manga, McAfee suspects. Technical indicators: The McAfee’s research team also found that the functionalities of Ryuk and Hermes are generally equal. Comparing the compilation time and the Program Database(PDB) of the samples collected during the past months strongly indicates that Ryuk was an altered version of Hermes 2.1. The investigation also found out that Hermes 2.1 was offered by a Russian-speaking actor in an underground forum making the source code available for cybercriminals looking forward to making use of it. Although the research has proven worthy in terms of understanding how the malware works, how the attackers operate, and how to detect the threat the question of who is behind the attack still remains a mystery. As for now, the conclusion is based on a hypothesis and it states that Ryuk was part of a cybercrime operation developed from a tool kit offered by a Russian-speaking actor.","excerpt":"Recently, Tribune Publishing, publisher of some of the major US newspapers, reported about the Ryuk ransomware which disrupted the distribution of papers in the weekend of late December 2018. According to a news report, the attack delayed the distribution of Saturday editions of the Los Angeles Times and San Diego Union-Tribune and also stymied distribution […]","categories":["AI News"],"tags":["Cyber Security","Ransomware"],"author_name":"Amal Nair","publish_date":"2019-01-16T21:48:10","publication_year":"2019","word_count":502,"keywords":["Go","Cyber Security","programming_languages:R","AI","Ransomware","programming_languages:Go","Aim","ViT","R"],"extracted_tech_keywords":["AI","Aim","R","Go","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/inside-ryuk-ransomware-that-brought-down-americas-leading-publisher-tribune\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10075144,"title":"Indore to Become the Country&#8217;s First Smart City with Smart Address","content":"‘Pataa’, a digital address app based out of Indore, has recently signed an MoU with the Indore government. According to this MoU, all governmental organisations and emergency services, including the police, fire department, and ambulance, will utilise the Pataa app. The app would also be utilised for essential services, including E-KYC and banking geotagging. It will be widely accessible to the public for personal and commercial purposes. ‘Pataa Navigation’ will integrate with government platforms such as those for the departments of power, agriculture, excise, women and child welfare, and education in order to offer access to Digital Addressing Systems. Additionally, Pataa will geotag and include all of MP Tourism’s landmark spots with websites. Digital address numbers would be placed on any major public property, such as light posts, bus stations, and public bathrooms. The user may locate the particular geotagged location by using the succinct and recognisable code Pataa. Users of the Pataa App can post pictures of their residences, landmarks, and other objects in addition to their full-text addresses. The user also has the choice to record audio directions—rendering it unnecessary to repeat the address over the phone—and making it easier for the visitor to locate the venue accurately. The user will soon be able to share a short code rather than their lengthy and complete address—which would be a major advantage in terms of emergency services, delivery and more. According to Pataa, “the alliance seeks to raise the standard of living for those living in rural areas of India by making resources like loans, seeds, and technical support more accessible.”","excerpt":"Indore government and Pataa, a digital address app, recently signed an MOU for Pataa to offer its expertise in geotagging services to various government agencies.","categories":["AI News"],"tags":[],"author_name":"Lokesh Choudhary","publish_date":"2022-09-13T16:53:37","publication_year":"2022","word_count":262,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","Git","GAN","R"],"extracted_tech_keywords":["AI","R","Go","Git","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/indore-to-become-the-countrys-first-smart-city-with-smart-address\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10045629,"title":"Visualizing and Comparing ML Models Using LazyPredict","content":"We live in a period of computational and technological supremacy, where computing has moved from large mainframes with tangled wires to PCs to an era of cloud computing. The world around us is changing quickly, and what makes it even remarkable is not what has happened until now but what is yet to come. It is an exciting time to live where various tools and techniques are being developed, followed by a major boost in computing, which can truly be called the world of Data Science!. Machine Learning, or ML for short, has proven to be one of the most game-changing technological advancements of the current decade. In an increasingly competitive corporate world, ML enables organizations to fast-track digital transformation and move swiftly into an age of automation. AI\/ML are here to stay relevant due to their demand of usage in everyday life, such as digital payments and fraud detection in banking or providing product recommendations to customers. The adoption of machine learning algorithms and methods to learn them are well-documented and readily accessible, with different companies moving in to adopt machine learning at scale across the verticals. Every other app and software available today all over the Internet uses machine learning in some way or the other. Machine Learning has now become the go-to solution for companies to solve problems. Today, one can use machine learning to process current and past data to predict future data. Other real-world applications vary from finding the shortest way in a map to reach a destination to identifying types of cancer cells. The process of developing a machine learning model can be complicated, and the model that is developed must be constructed in such a manner that it fits the problem perfectly. Machine learning is used to either solve a problem or offer insights that can lead to better decision making, wherever data is present. In the case of machine learning models, however, no one method can be used to solve all problems. Distinct types of algorithms are developed to solve different issues using entirely different techniques. For each class, the inputs supplied, the task completed, and the results achieved are all extremely distinct. Some of the major types are Supervised Learning, Semi-Supervised Learning and Reinforcement Learning. Supervised learning is a machine learning algorithm where a model or a function is being developed to map the input from the test data to their respective output. Here, the training dataset is a data bank of labelled data, and the test data is a set of inputs having no labels. Unsupervised learning is a type of machine learning that uses the inferences drawn from a dataset without labels. Reinforcement learning algorithms are the type of machine learning model where tasks are being performed by an agent in a particular simulation environment. During this, the agent either receives a reward or punishment for each task that is performed. Unlike other machine learning approaches, the algorithm is not given any instructions and learns by itself. Machine learning methods depend upon the type of task and can be further categorized as Classification models, Regression models, Clustering etc. Classification is the task of predicting a type or class of an object from a finite number of options. The output variable generated by classification is usually categorical, but regression may be used to solve a collection of issues using a continuous output variable. Predicting the price of airline tickets, for example, is a classic regression assignment. Clustering, on the other hand, is the challenge of grouping items that are related in some way. It assists in the automated identification of similar items without the need for human interaction. What is LazyPredict? LazyPredict is an open-source python library that helps you to semi-automate your Machine Learning Task. It can build multiple models without writing much code and helps understand which models work better for the processed dataset without requiring any parameter tuning. Using LazyPredict, one can apply all the models on that dataset to compare and analyze how our basic model is performing. Here a basic model means a “Model without parameters”. It can help deriving accuracies, and after getting accuracy for all the models, one can choose the top 5 models and then apply hyperparameter tuning to them. It comes with a Lazy Classifier to solve classification problems and Lazy Regressor to solve the regression problems. While building machine learning models, one cannot be sure which algorithm will work well on the given dataset; hence, it ends up trying many models and keeping iterating until proper accuracy is synthesized. LazyPredict comes to the rescue for such use cases, generating all the basic machine learning algorithms’ performances on your model. Along with the accuracy score, LazyPredict also provides certain evaluation metrics and describes the time taken by each model. Getting Started with the Code In this article, we will implement a model using the LazyPredict library, which will help us find the best-suited model for Classification and Regression to be used for our dataset and each model’s accuracy score. In addition, we will also be visualizing the accuracy scores to compare and choose the best-suited model for the dataset being processed. The following code is inspired by the documentation provided by the creators of LazyPredict, whose link can be found here. Installing The Library The first step will be to install the LazyPredict library to set up our model; you can use the following code to do so, !pip install lazypredict !pip install scipy==1.7.1 We are also installing the latest version of SciPy, which will help us process the data better. #cloning the model !git clone https:\/\/github.com\/shankarpandala\/lazypredict.git Performing Classification: Importing Dependencies Now we will be importing the required essential dependencies for our model, # Import libraries import lazypredict from lazypredict.Supervised import LazyClassifier from sklearn.datasets import load_breast_cancer from sklearn.model_selection import train_test_split Loading The Dataset Let us now load our dataset; we use the inbuilt Breast Cancer dataset for our first problem. # Load dataset data = load_breast_cancer() X = data.data y= data.target Splitting the dataset into train and test, # Splitting Dataset X_train, X_test, y_train, y_test = train_test_split(X, y,test_size=.2,random_state =42) Performing Classification With the dataset loaded and everything else set up, now let’s perform our classification task. For this, we will be setting up our classification pipeline using the LazyClassifier. # Defining parameters for lazyclassifier clf = LazyClassifier(verbose=0,ignore_warnings=True, custom_metric=None) models_train,predictions_train = clf.fit(X_train, X_train, y_train, y_train) models_test,predictions_test = clf.fit(X_train, X_test, y_train, y_test) # Printing all the model performances models_train Output : As you can see, the LazyClassifier has provided us with scores for all the possible models according to our dataset! Let us now visualize the scores to derive a better understanding, #plotting the accuracy scores import matplotlib.pyplot as plt import seaborn as sns plt.figure(figsize=(10, 5)) sns.set_theme(style=\"whitegrid\") ax = sns.barplot(x=models_train.index, y=\"Accuracy\", data=models_train) plt.xticks(rotation=90) import matplotlib.pyplot as plt import seaborn as sns plt.figure(figsize=(5, 10)) sns.set_theme(style=\"whitegrid\") ax = sns.barplot(y=models_train.index, x=\"Accuracy\", data=models_train) Using such visualizations, we can now easily understand the best model to be used for optimal accuracy. Performing Regression We can also perform Regression using the LazyRegressor module on another dataset! # Importing the libraries from lazypredict.Supervised import LazyRegressor from sklearn import datasets from sklearn.utils import shuffle import numpy as np # Loading the Boston dataset boston = datasets.load_boston() X, y = shuffle(boston.data, boston.target, random_state=42) #Splitting Data X_train, X_test, y_train, y_test = train_test_split(X, y,test_size=.2,random_state =42) # building the pipeline reg = LazyRegressor(verbose=0,ignore_warnings=False, custom_metric=None) models_train,predictions_train = reg.fit(X_train, X_train, y_train, y_train) models_test,predictions_test = reg.fit(X_train, X_test, y_train, y_test) # Printing all model performances models_train Output : Creating Visualization for a particular column, such as for R-squared Scores only, import matplotlib.pyplot as plt import seaborn as sns plt.figure(figsize=(10, 5)) sns.set_theme(style=\"whitegrid\") ax = sns.barplot(x=models_train.index, y=\"R-Squared\", data=models_train) ax.set(ylim=(0, 1)) plt.xticks(rotation=90) import matplotlib.pyplot as plt import seaborn as sns models_train[\"R-Squared\"] = [0 if i < 0 else i for i in models_train.iloc[:,0] ] plt.figure(figsize=(5, 10)) sns.set_theme(style=\"whitegrid\") ax = sns.barplot(y=models_train.index, x=\"R-Squared\", data=models_train) ax.set(xlim=(0, 1)) End Notes In this article, we learnt about the importance of different machine learning models and their uses. We also created a model using the LazyPredict library that helps us understand the best-suited model for our dataset for optimal results and accuracy. The Following implementation can be found as a Colab notebook, using the link here. Happy Learning! References Official Github RepositoryScipy Website","excerpt":"LazyPredict is an open-source python library that helps you to semi-automate your Machine Learning Task. It can build multiple models without writing much code and helps understand which models work better for the processed dataset without requiring any parameter tuning.","categories":["AI Trends"],"tags":["AI (Artificial Intelligence)","AI Tool","Data Science","data visualisation libraries","Deep Learning","libraries","machine learning libraries","Python"],"author_name":"Victor Dey","publish_date":"2021-08-10T14:00:00","publication_year":"2021","word_count":1375,"keywords":["data science","NumPy","machine learning","AI","cloud computing","ML","data visualisation libraries","libraries","Python","Colab","Seaborn","machine learning libraries","Deep Learning","Data Science","Matplotlib","fraud detection","AI (Artificial Intelligence)","AI Tool"],"extracted_tech_keywords":["AI","machine learning","ML","data science","Colab","NumPy","Matplotlib","Seaborn","fraud detection","cloud computing"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/visualizing-and-comparing-ml-models-using-lazypredict\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10068020,"title":"The Indian IT and carbon neutrality","content":"As part of the 2015 Paris Agreement, countries worldwide have agreed to work collectively to limit global warming to below 2, preferably to 1.5 degrees Celsius, compared to pre-industrial levels. To that end, many big tech companies have made climate commitments. Ten years ago, an average Google search produced about 0.2 grams of CO2. Today, Google has put practices in place to offset its carbon footprint. The tech giant plans to decarbonise its energy consumption by 2030, and operate on carbon-free energy, day in and day out. In India, the IT industry churns out two million tons of carbon annually before the COVID-19 pandemic. However, the emissions dropped by nearly 85 percent during the lockdown to 300,000 tons. In India, IT companies such as Infosys and Wipro have taken significant steps to reduce their carbon footprint and achieve net neutrality. In fact, many IT firms are offering green IT solutions to amplify their clients’ efforts to reach net-zero emissions. Infosys is carbon neutral Infosys committed to achieving carbon neutrality in 2011, four years before the Paris agreement. In 2020, Infosys announced that it turned carbon neutral in compliance with PAS 2060 standards, 30 years before the 2050 timeline. During the same period, around 44.3 percent of its electricity needs in India were met by renewable energy sources. In 2021, about 50 percent of electricity Infosys consumed came from renewable energy sources. The company has invested in 60 MW of solar photovoltaic (PV) capacity, said Infosys CEO Salil Parekh. Staying true to its commitment to fighting climate change, Infosys teamed up with oil and gas behemoth BP to develop an integrated Energy-as-a-Service (EaaS) offering. Announced last year, the new service by Infosys will provide end-to-end management of a customer’s energy assets and services. Infosys said both parties would explore opportunities to manage energy assets and provide low carbon power, low carbon heating\/cooling, and low carbon mobility to campuses, driven by an AI-based digital platform. TCS to be carbon neutral by 2030 In 2020, TCS announced it would reduce its carbon footprint by 50 percent. In addition, India’s largest software service exporter said it would be carbon neutral by 2030. The company plans to achieve this by improving energy efficiency by adding more green buildings to its real estate portfolio, reducing IT system power usage, and using TCS Clever Energy, which leverages IoT, ML and AI to optimise energy consumption across campuses. The company will, by 2025, reduce its absolute greenhouse gas emissions across Scope 1 and Scope 2 by 70 percent. Earlier this year, TCS announced a host of sustainability solutions for its clients, such as TCS Clever Energy, Intelligent Power Plant (IP2) and TCS Envirozone on Microsoft’s Azure IoT platform. With these latest offerings, TCS’ clients will be able to gain insights into their energy usage and reduce waste and emissions to accelerate their net-zero goals. Wipro to reach net-zero GHG emission by 2040 In 2021, Wipro announced it would reach net-zero greenhouse gas emissions by 2040. The IT giant also plans to cut its greenhouse gas emissions by 55 percent by the decade. This will be achieved by increased use of renewable energy in owned facilities in India through private power purchase agreements and captive solar power. Further, Wipro will combine behavioural, technological, and collaborative approaches that help reduce the carbon footprint of air travel, commuting and purchased goods and services. Wipro has also developed green IT solutions to serve its clients in high carbon-intensive sectors like Energy and Utilities, Manufacturing, Transportation and Consumer Goods. In addition, Wipro’s products and services will help its clients in waste management, sustainable sourcing and packaging and public cloud adoption. Tech Mahindra joins the Climate Pledge Earlier this year, Tech Mahindra joined the Climate Pledge, a cross-border community of companies committed to reaching the Paris Agreement 10 years early and be net-zero carbon by 2040. Amazon, ​​SAP and Maersk are other major signatories of Climate Pledge. Tech Mahindra has committed to reducing greenhouse gas emissions by 22 percent by the end of 2030 and 50 percent by 2050. The tech giant will focus on improving energy efficiency through process optimisation, energy conservation initiatives, increased use of renewable energy, and investing in low emission and green technologies. Mahindra Group’s renewable energy asset management subsidiary Mahindra TEQO leverages AI, ML, augmented reality and industrial internet of things (IIoT) to help its clients in the renewable energy industry decrease operating and maintenance expenses by 20 percent. HCL to reach net-zero by 2040 HCL became the latest Indian IT giant to join the Climate Pledge. The company has vowed to limit greenhouse gas emissions by 2030 as per the Paris Agreement and reach net zero by 2040. At HCL Technologies, we are committed to reducing the impact our company has on the planet. Our aim is to limit our GHG emissions aligned to a 1.5 °C pathway by 2030 and to reach net-zero by 2040. Learn more: https:\/\/t.co\/KdqUU3Jojz. pic.twitter.com\/lVPLL9iCpr— HCLTech (@hcltech) December 16, 2021 “These new commitments are part of our efforts to act responsibly and lead by action, sustainably offer our stakeholders maximum value as part of our pact and ensure we create impact through our actions and initiatives,” Santhosh Jayaram, Global Head of Sustainability, HCL Technologies, said.","excerpt":"Infosys achieved carbon neutrality in 2020.","categories":["IT Services"],"tags":[],"author_name":"Pritam Bordoloi","publish_date":"2022-05-30T11:00:00","publication_year":"2022","word_count":872,"keywords":["Go","cloud_platforms:Azure","programming_languages:R","AI","R","ML","Git","RAG","Aim","Azure"],"extracted_tech_keywords":["AI","ML","Aim","RAG","Azure","R","Go","Git","cloud_platforms:Azure","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/it-services\/the-indian-it-and-carbon-neutrality\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10136406,"title":"Transformers Can Solve Any Problem","content":"In many cases, LLMs are turning out to be the default solution to business problems, including domains that do not require language understanding. It seems like Transformers are indeed all we need. Reiterating this, Denny Zhou, research director at Google DeepMind, recently released a new paper. While sharing his research on X, Zhou mentioned that, “We have mathematically proven that Transformers can solve any problem, provided they are allowed to generate as many intermediate reasoning tokens as needed.” This echoes AI researcher Andrej Karpathy’s recent remarks on next-token prediction frameworks, suggesting that they could become a universal tool for solving a wide range of problems, far beyond just alone. LLMs are not really just “language experts” anymore. According to Karpathy, the “language” part has become historical because these models were first trained to predict the next word in a sentence, but in reality, they can work on any kind of data that’s broken down into little pieces, called tokens. This is what Zhou also presented in his paper, but instead for LLMs. The research primarily focused on CoT (chain of thought) and by using CoT, it provides a “road map” for LLMs to follow when solving complex problems. A YouTube video, while explaining the significance of this paper, mentioned that using CoT, you are not giving your AI tools to put the puzzle pieces together but instead you are making it understand why the pieces fit the way they should. Chain of Thinking and Reasoning Without CoT, Transformers can only solve problems in parallel computation models (AC0\/TC0 complexity classes). However, with CoT, Transformers can solve more complex problems, enabling serial computation. It all comes down to how a model will be able to reason better. We saw the first iteration of System 2 LLMs starting with o-series models from OpenAI which uses a perfect combination of CoT and reasoning tokens. Zhiyuan Li, assistant professor at the Toyota Technological Institute at Chicago and lead contributor to the research paper, mentioned that this proves that CoT enables more iterative compute to solve inherently serial problems. “On the other hand, a const-depth transformer that outputs answers right away can only solve problems that allow fast parallel algorithms,” he added. Li further shared an image suggesting that models using CoT can solve more complex problems that require many steps in sequence. On the other hand, models without this ability can only handle simpler problems that can be solved quickly in parallel. With techniques like CoT, we are moving towards explainable AI systems and slowly moving away from models that were prone to blackbox. A Reddit user mentioned that with the help of CoT, the inner workings of LLM are traceable too. “The black box of latent space would make it harder for us humans to understand how the model is performing the reasoning. There is a huge benefit to explainable AI,” he added further. Number of Tokens Matter the Most CoT goes beyond solving maths problems, and users have started comparing it to Turing Machines, a theoretical computational model that defines an abstract machine capable of simulating any computer algorithm. For example, Google researchers have solved two critical problems – the Circuit Value Problem (CVP) and the Permutation Composition Problem. These two are classic computer problems and in both cases, enabling CoT allows Transformers, especially those with low depth, to solve these inherently serial problems much more effectively than without CoT. This aligns with the paper’s theoretical predictions about CoT’s ability to empower Transformers to handle serial computations. In a discussion on HackerNews, users mentioned that CoT could push LLMs closer to the theoretical limits of computation as represented by Turing machines. “Since LLMs operate as at least a subset of Turing machines, the chain of thought approach could be equivalent to or even more expressive than that subset. In fact, CoT could perfectly be a Turing machine,” he added further. The CoT method, while being extremely useful, is known to use multiple tokens, which is not only costly but takes longer to respond. Justin, the founder of the AI LegalTech startup, raised questions about practical application perspective, time, and cost. There are few things which can be considered when it comes to cost. The first is Mosaic’s law, which states that the cost of training goes down 75% per year. The other consideration is Koomey’s law, which states that the energy efficiency of computation doubles roughly every 1.5 years. It’s a critical metric to understand how computational power becomes more energy-efficient over time, which is vital for sustainable technology development and the same goes for a number of tokens.","excerpt":"With techniques like CoT, we are moving towards explainable AI systems and slowly moving away from models that were prone to blackbox.","categories":["AI Features"],"tags":["Editors Picks","reasoning","Transformer Model"],"author_name":"Sagar Sharma","publish_date":"2024-09-23T15:40:52","publication_year":"2024","word_count":766,"keywords":["Go","TPU","OpenAI","AI","RPA","Transformers","Editors Picks","explainable AI","ViT","reasoning","chain of thought","Transformer Model","R"],"extracted_tech_keywords":["AI","OpenAI","Transformers","chain of thought","TPU","R","Go","ViT","explainable AI","RPA"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/transformers-can-solve-any-problem\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":5280,"title":"Great Lakes launches Centre for biz analytics","content":"Chennai-based Great Lakes Institute of Management added a sixth centre of excellence to the five it has now, with the launch of its Centre for Excellence in Business Analytics and Business Intelligence. Inaugurated by Shankar Narayanan, serial entrepreneur and Co-founder of Nexus Inc in the US, the opening of the centre is also to mark ten years of Great Lakes. The inaugural address for the centre was delivered by Vinod Muthukrishnan, CEO, Cloudcherry Pte, Singapore. Also present on the occasion was Ram Vishwanathan, Distinguished IBM Engineer, who spoke on the trends in the digital world. The B school also hosted a conference for the students where panelists spoke on the three fertile areas within the mobility spectrum – social media and rich communication experience, Web and mobile tech and inclusive growth. K. C. John, Director, Centre of Excellence in Technopreneurship, gave an overview of the conference. He stressed the importance of entrepreneurship and pointed out that in the last few years India witnessed a meagre employment growth rate of 2.3 per cent only and now was the time to create jobs rather than seek one. The Business Analytics centre has been launched with an objective to sync business studies with the emerging trends of the digital world. The Centre will have an association with an analytics company and Great Lakes is currently in talks with companies such as Cloudcherry. On the occasion, a text book on business analytics by Tapan Panda, Professor of Marketing, was released. Bala V. Balachandran, Dean, said, “The mobile is going to take a major change in coming years with a large audience being interconnected. Envisioning this as an opportunity, Great Lakes is going to create and offer boutique courses in business analytics and intelligence.” He said the institute is in the process of signing an MoU with Illinois Institute of Technology for Centre for Excellence in Technopreneurship, through which it plans to offer courses such as a Master of Science in Technopreneurship. Other centres of excellence in Great Lakes are the: Kotler – Srinivasan Centre for Research in Marketing; Union Bank Centre for Banking Excellence; Yale — Great Lakes Centre For Management Research and the recently-launched Centre for Excellence in Retail Management. Source: Business Line","excerpt":"Chennai-based Great Lakes Institute of Management added a sixth centre of excellence to the five it has now, with the launch of its Centre for Excellence in Business Analytics and Business Intelligence. Inaugurated by Shankar Narayanan, serial entrepreneur and Co-founder of Nexus Inc in the US, the opening of the centre is also to mark […]","categories":["AI Trends"],"tags":[],"author_name":"AIM Media House","publish_date":"2014-02-20T08:16:31","publication_year":"2014","word_count":370,"keywords":["business intelligence","Go","programming_languages:R","AI","programming_languages:Go","Git","Ray","analytics","R"],"extracted_tech_keywords":["AI","analytics","Ray","R","Go","Git","business intelligence","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/great-lakes-launches-centre-for-biz-analytics\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10166170,"title":"Ather Bets Big on AI and Salesforce to Transform EV Sales","content":"One of India’s leading electric two-wheeler manufacturers, Ather Energy, has taken a tech-first approach to customer engagement and dealer management. In partnership with Salesforce, Ather has developed a unified, AI-ready digital platform to streamline its sales, service, and customer experience operations. The partnership was first announced two years ago. With the integration of Salesforce’s cloud-based solutions, Ather now has a single, scalable system that connects dealers, service centers, and customers, ensuring real-time access to critical data. Unified Saas-EV Ecosystem “We are a growing company in a growing industry, and things are changing rapidly—not just from a consumer point of view but also from a regulatory perspective. At the core of our business strategy is consumer experience, and that includes our dealers,” Ravneet S. Phokela, chief business officer, Ather Energy,​ said in an exclusive interview with AIM. He was speaking on the sidelines of Ather-Salesforce event in Bengaluru. India’s EV sector is evolving rapidly, with shifting regulations, growing competition, and changing consumer behaviour. To navigate these challenges, Ather recognised the need for a centralised platform that brings together lead management, customer interactions, and dealership operations. Phokela mentioned that regarding integrating AI or other emerging technologies, it is essential to be positioned to embrace the innovations that arise within the ecosystem. Previously, Ather’s dealership and customer management relied on multiple tools, making it challenging to track test drives, pre-orders, and post-sales service seamlessly. With Salesforce, Ather has consolidated these functions into a unified platform, allowing for real-time coordination across all stakeholders. “You can’t be a consumer-focused company if you don’t have a unified view of the customer,” he said. Highlighting the significance of this collaboration, Mankiran Chowhan, vice president of Salesforce India, emphasised how it enhances Ather’s scalability and operational efficiency. “This is about unifying CRM, lead management, and DMS (dealer management system) into one scalable platform, enhancing the experience for both dealers and customers,” said Chowhan to AIM. AI-Driven Insights Continue Beyond just data centralisation, the system is built to integrate AI and automation, ensuring smarter decision-making and predictive analytics. Chowhan highlighted Salesforce’s AI engine, Einstein AI, that processes over 2 trillion predictions daily, enabling businesses to personalise customer interactions and optimise operations. “This is really about how human agents and digital agents are all working together seamlessly in maximizing productivity,” said Chowhan. While Ather is still in the early stages of AI adoption, its new digital backbone ensures future compatibility with AI-driven tools. This means the company can eventually implement predictive maintenance alerts, automated service scheduling, and real-time sales forecasting, all powered by AI. Phokela emphasised the importance of seamless AI integration, stating, “Technology should be invisible. A sales guy should sell. The system should work in the background, making their job easier.”​ Chowhan highlighted how trust and transparency are key concerns in AI adoption, stating, “The world of AI is evolving rapidly, and trust and privacy are becoming critical. Most organisations are now looking at how AI ties back to their business goals.”​ For Ather, this means ensuring that AI is embedded responsibly into its digital framework, allowing for scalability without compromising data security. The partnership also ensures that as new regulatory requirements emerge, Ather’s systems can adapt and evolve without major overhauls. Overcoming Transition Challenges Transitioning to a unified, AI-enabled digital platform was not without its challenges. Phokela highlighted that resistance to change was one of the biggest obstacles. “If it ain’t broken, why fix it? We already had a system, and it was working fine. The decision to proactively change something was a reasonably bold step,” he stated, emphasising the risks associated with data migration and system overhaul​. From a technical standpoint, developing a scalable and future-ready architecture requires careful planning, as Ather explained. The system needed to be flexible and adaptable to meet the evolving needs of Ather’s dealerships and customers. The complexity of dealer-managed systems also necessitated a solid data model with 40-50 data entities, ensuring that future business pivots would not be hindered by technological limitations. On future collaborations, Chowhan expressed optimism about empowering more businesses. “Agent force is what we call, is really the heart and center of what Salesforce is focusing on,” she said. She added that there is a global workforce shortage while it is in surplus in India, and yet many are not employable. “So our focus right now is making sure that we can bridge that gap and get AI to help enable that next wave of growth as India focuses on being the third largest economy and really empowering that next wave,” Chowhan said.","excerpt":"“This is about unifying CRM, lead management, and DMS (dealer management system) into one scalable platform, enhancing the experience for both dealers and customers,” said Mankiran Chowhan, vice president of Salesforce India.","categories":["AI Features"],"tags":["Ather","Cloud","CRM","dealers","EV"],"author_name":"Vandana Nair","publish_date":"2025-03-17T18:58:53","publication_year":"2025","word_count":756,"keywords":["Go","AI","Cloud","Ather","ML","predictive analytics","R","Scala","Git","dealers","Aim","analytics","Rust","EV","CRM"],"extracted_tech_keywords":["AI","ML","analytics","Aim","predictive analytics","R","Go","Rust","Scala","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/ather-bets-big-on-ai-and-salesforce-to-transform-ev-sales\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":27125,"title":"A Pinch Of AI And A Dash Of Automation Is What Helps Mechanical Chef Cook 100+ Indian Dishes","content":"When you walk to the office of this Bengaluru-based startup, you shouldn’t be shocked to see machines hard at work, as they cook delicious Indian delicacies! We are talking about Mechanical Chef, a startup founded in 2017, which has taken on the challenge of developing cooking robots for the Indian market. Started by Arpit Sharma, an IIT Kanpur graduate and Cohan Sujay Carlos, a machine learning researcher, the company has till now been able to accomplish a prototype which looks like a top-heavy bundle of bottles whirling, spinning and measuring the ingredients to create the perfect recipe. It is also open to the public to test the robot and provide feedback that will help them inculcate the necessary features. Analytics India Magazine caught up with Arpit Sharma, who is the CTO of Mechanical Chef to know the inside-out of this interesting startup which may be the answer to millions of those struggling to put home-made dishes on their plate. How Did The Makers Build An Automatic Cooking Robot For Indian Homes? On being asked why they chose to work on such a project, Sharma shares that it all started with a conversation with a woman colleague, who quipped that if women didn’t have to do any cooking, it would have saved half a billion women in India three hours a day, every day of their lives. Sharma also shares another interesting personal experience they faced at home, where they saw their working parents struggle with balancing work and putting hours of efforts in the kitchen. Also, having lived away from home and cooking for themselves, they could relate to the problem quite easily. Taking an inspiration from these instances, they started putting things in place to turn the idea of creating mechanical chef into reality. “The first design of mechanical chef turned out to be too expensive and complicated to be sold to Indian customers. We had to do everything possible to cut costs, as Indians are very price-conscious buyers”, said Sharma. And the way out was altering physics and mechanical design. The first thing they did was to eliminate all vertical movements against the force of gravity. Even this turned out to be insufficiently optimal and they re-designed it to replace all linear movements with rotary movements. The resulting third design was turned into a working prototype for user testing with the help of a grant from the Department of Science and Technology in 2017. How Does The Mechanical Chef Work? The product is still a prototype but it can cook Indian dishes perfectly without human interventions. It can cook three dishes for a family of three on three heathers using the following operations: Dropping the ingredients into the dish Heating the dish and Stirring the contents The machine comes loaded with all the spices that are needed for cooking (stored permanently in spice boxes), and the user can add vegetables needed for cooking, and the machine then takes over. It can currently cook around 30 Indian dishes, which the team is trying to increase every day. Some of these are matar paneer, bisi bele bath, rice, dal tadka, chhole, rasam, sambhar, upma, poha, capsicum sabji, among others. “We are aiming at 150+ dishes in the final version”, said Sharma. The Mechanical Chef is designed to ensure that it saves the user time and adds no Inconvenience. When the user selects what they want to cook (the machine can also suggest things based on what vegetables are available), the machine guides them through what vegetables they need to prepare and cut, in the quantities they need. It then guides them through the loading of the vegetables into the machine. The vessels that hold the vegetables and any part of the machine that comes into contact with the food can be snapped on and off, and is therefore easy to clean. The number of such parts to clean is also minimised to keep it easy. What Makes It Relevant Today? Sharma is quick to add that Internet of Things (IoT) and AI have been making a lot of news lately. “Internet-enabled and connected devices are expected to become ubiquitous in smart homes and smart kitchens. Well, it doesn’t get any smarter than this”, says Sharma. A cooking robot straddles the worlds of connected devices and intelligent machines, and can save its owner a lot of time and effort twice or thrice a day. To make this a reality, they are using analytics and artificial intelligence to analyse the spice content being put in the food, monitoring the number of ingredients, deciding on an ingredient that can help balance diet for people, among others. “But the underlying principle of designing it is to use physics than to use sophisticated hardware or tools. Our team ensures to put enough physics fundamentals before adding any component to the machine which is enabling us to keep the cost low”, shares Sharma. Keeping Up With the Neck-To-Neck Competition Sharma shares that India has had a long history of innovation in consumer electronics and electrical products, starting with wet grinders that were invented in India to Rotimatic, which was developed by an Indian couple living in Singapore. Being in the field of robotic cooking, they face competition from companies such as Spyce, Nymble and Moley. Explaining the key differentiator from their competitors, Sharma shares that the Mechanical Chef is the first cooking robot for homes to be demonstrated to the public and the size is such that can fit into homes (others have developed machines that are too large to fit in homes). They are also working up to keep the cost affordable, unlike other players. Keeping Up With The Challenges Sharma says that poor manufacturing options make it difficult for a hardware company to grow in India. However, they are in talks with different manufacturing partners for manufacturing different components which are not being manufactured in India right now. “The other challenge is that investors remain wary of investments in hardware and manufacturing firms in India because they are capital intensive and mistakes can be costly and the possibilities are not yet as well-understood as in the case of software”, added Sharma. Hence they rely on user studies. According to Sharma, it can give both participants and entrepreneurs a window into each other’s world, and an understanding of the market they are catering to. Growth Plan The Mechanical Chef team is ambitious to make money through sales or leasing the machine. “We expect the market to be between 1 and 2 billion USD initially and grow by at least 10 percent each year”, said Sharma. They also have plans to raise a seed round of around $120,000 and are in talks with angel investors and firms in the consumer products space. Sharma shares their plan of spending about 50 percent of the money on hiring tech employees for product development in Bengaluru. “We are aiming to bring more functional and improved prototype by end of this year which will be close to the production level”, he said. On A Concluding Note While Mechanical Chef has created a lot of buzz, can it replace human cook? Sharma says that there are certain cooking chores it cannot perform. “Given the state of the art in AI today, it is still not a simple matter to build cutting machines that are also easy to clean. So, a human would be needed to do the cutting and peeling”, he said. Despite these shortcomings, they are convinced that there is a market for invention in India and that it would prove revolutionary for women in nuclear families, single men and women living by themselves. The makers hope to see the production of cooking robots one-and-a-half years from now, as they also aim to work on other automated products in the kitchen.","excerpt":"When you walk to the office of this Bengaluru-based startup, you shouldn’t be shocked to see machines hard at work, as they cook delicious Indian delicacies! We are talking about Mechanical Chef, a startup founded in 2017, which has taken on the challenge of developing cooking robots for the Indian market. Started by Arpit Sharma, […]","categories":["Deep Tech"],"tags":["automation testing","software automation testing"],"author_name":"Srishti Deoras","publish_date":"2018-08-09T09:41:53","publication_year":"2018","word_count":1301,"keywords":["API","machine learning","artificial intelligence","automation testing","AI","innovation","software automation testing","Aim","ViT","analytics","R","startup"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","analytics","Aim","R","API","ViT","innovation","startup"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/a-pinch-of-ai-and-a-dash-of-automation-is-what-helps-mechanical-chef-cook-100-indian-dishes\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10045406,"title":"Why Is Federated Learning Getting So Popular","content":"In 2016, Uber paid $148 million to settle the investigation on a data breach that exposed the personal information of over half a million drivers. In 2020, Google was fined $57 million for a GDPR violation. The rise of on-device machine learning, coupled with the growing concerns of data privacy, has nudged developers and researchers towards techniques such as federated learning–a collaborative learning method that operates without exchanging users’ original data. The stringent GDPR makes data sharing among European organisations challenging. Meanwhile, federated learning systems (FLS) have shown promise with good predictive accuracy while complying with privacy rules. According to Li et al., federated learning systems are game changers like deep learning frameworks such as PyTorch and TensorFlow. As illustrated above, the number of related papers in FL has increased rapidly and has reached about 4,400 last year. Federated Learning leverages techniques from multiple research areas such as distributed systems, machine learning, and privacy. FL is best applied in situations where the on-device data is more relevant than the data that exists on servers. However, FLS faces various challenges such as effectiveness, efficiency, and privacy. FL enables multiple parties to jointly train a machine learning model without exchanging the local data. According to Li et al.,se FLS is being adopted in various domains: A typical Federated Learning Protocol according to Google AI: Federated learning servers are called by devicesModel checkpoint from storage is read by the servers.Models are sent to the select devices.On-device model training followed by server update. Server aggregates these updates into a global model and writes them into storage. On-device FL (Source: Paper by Bonawitz et al.,) Federated learning provides a decentralised computation strategy to train a neural model. Modern day mobile devices churn out swathes of personal data, which can be used for training. Instead of uploading data to servers for centralised training, phones process their local data and share model updates with the server. Weights from a large population of mobiles are aggregated by the server and combined to create an improved global model. The distributed approach has been shown to work with unbalanced datasets and data that are not independent or identically distributed across clients. On-device machine learning comes with a privacy challenge. Data recorded by cameras and microphones can put individuals at great risk in the event of a hack. For example, apps might expose a search mechanism for information retrieval or in-app navigation. Federated averaging was implemented by researchers from University of Kyoto in practical mobile edge computing (MEC) frameworks by using an operator of MEC frameworks to manage the resources of heterogeneous clients. Both distributed deep reinforcement learning (DRL) and federated learning were also adopted in mobile edge computing systems. The use of DRL and FL has the potential to optimise the mobile edge computing, caching, and communication. FL has also been performed on resource-constrained MEC systems, where the researchers address the problem of how to efficiently utilise the limited computation and communication resources at the edge. A technique called FedGKT was proposed where each device only trains a small part of a whole ResNet to reduce the computation overhead. Using federated averaging, the researchers implemented many machine learning algorithms including linear regression, SVM, and CNN. For natural language processing Companies like Google use Federated Averaging techniques in its smartphone keyboard for text prediction. FL was applied in mobile keyboard next-word prediction. Federated averaging method was used to learn a variant of LSTM called Coupled Input and Forget Gate (CIFG). According to the researchers, the FL method can achieve better precision recall than the server-based training with log data. For instance, companies like Apple use FL techniques and its variants like Federated Tuning (FT) on their products to carry out a combination of on-device computing as well as recommendations with user privacy. For Apple, applications around FE and FT occupy a large percentage of system usage. Federated evaluation (FE) occurs on user interaction history. This significantly reduces turn-around times when compared to live A\/B experimentation. FE can help quickly identify the most promising ML system or model candidates before exposing end users to these candidates via live A\/B experimentation. For recommender systems Federated collaborative filter method is popular with building recommendation systems. Based on a stochastic gradient approach, the item-factor matrix is trained in a global server by aggregating the local updates. The method is said to have no accuracy loss compared to the centralised method. In another method, a federated matrix factorisation framework is used. Here, Federated SGD is used to learn the matrices. Federated recommender system (FedRecSys) has implemented popular recommendation algorithms with SMC protocols. The algorithms include matrix factorisation, singular value decomposition(SVD), factorisation machine, and deep learning. Though the local data are not exposed in FL, the exchanged model parameters may still leak sensitive information. Model inversion attacks and membership inference attacks etc can potentially infer the raw data by accessing the ML model. This is a major concern especially in domains such as healthcare. Modern health systems require cooperation among research institutes, hospitals, and federal agencies. Moreover, in a pandemic like situation, collaborative research among countries is vital but not at the expense of privacy. FL makes the cooperation possible because it can ensure privacy. In a federation of healthcare, there is probably no central server. So, another challenging part is the design of a decentralised FLS, which should also be robust against malefactors. The privacy concern can be solved by additional mechanisms like secure multi-party computation and differential privacy. According to a survey on FLS by Li et al., explainability of the FL models is an open problem.","excerpt":"Federated learning provides a decentralised computation strategy to train a neural model.","categories":["Deep Tech"],"tags":["churn prediction model","federated learning"],"author_name":"Ram Sagar","publish_date":"2021-08-09T10:00:00","publication_year":"2021","word_count":932,"keywords":["federated learning","machine learning","AI","PyTorch","ML","RAG","CuPy","deep learning","differential privacy","churn prediction model","TensorFlow"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","TensorFlow","PyTorch","CuPy","federated learning","differential privacy","RAG"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/why-federated-learning-getting-so-popular\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":45170,"title":"India To Unveil Cybersecurity Strategy Policy By January 2020","content":"India is reportedly set to release its first-ever cybersecurity policy by January 2020. Speaking at an event, Rajesh Pant, the national cybersecurity coordinator on cyber policies, told a news wire, “India’s cybersecurity strategy policy will be released in January 2020 and will enable the government to cyber-secure the nation. The government’s vision of a $5-trillion economy will be helped to a great extent by this effort.” Ajeet Bajpai, director-general of the National Critical Information Infrastructure Protection Centre, added, “Considering the size and scale of our nation, we need approximately ₹25,000 crore budget for the same. The biggest question is where this money will come from? Also, there is a need to emphasise on the need to make cybersecurity mandatory as a subject at the university level for high-decibel awareness.” This is not the first time that India has decided to launch a cybersecurity policy. Back in 2013, former IT minister Kapil Sibal had released the National Cyber Security Policy in Delhi. The policy was prepared in consultation with all relevant stakeholders, user entities and public. It was aimed at facilitating the creation of a secure computing environment and enabling adequate trust and confidence in electronic transactions and also guiding stakeholders actions for the protection of cyberspace. In 2017, Piyush Goyal had created sectoral Computer Emergency Response Teams (CERTs) to mitigate cybersecurity threats in power systems.","excerpt":"India is reportedly set to release its first-ever cybersecurity policy by January 2020. Speaking at an event, Rajesh Pant, the national cybersecurity coordinator on cyber policies, told a news wire, “India’s cybersecurity strategy policy will be released in January 2020 and will enable the government to cyber-secure the nation. The government’s vision of a $5-trillion […]","categories":["AI News"],"tags":["Cybersecurity"],"author_name":"Prajakta Hebbar","publish_date":"2019-08-29T14:50:03","publication_year":"2019","word_count":225,"keywords":["Go","API","programming_languages:R","AI","programming_languages:Go","Aim","Rust","Cybersecurity","R","programming_languages:Rust"],"extracted_tech_keywords":["AI","Aim","R","Go","Rust","API","programming_languages:R","programming_languages:Go","programming_languages:Rust"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/india-to-unveil-cybersecurity-strategy-policy-by-january-2020\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10007363,"title":"Join This Full-Day Workshop On Natural Language Processing From Scratch","content":"The Association of Data Scientists, the premier global professional body of data science & machine learning professionals, has announced a full-day workshop on Natural Language Processing (NLP) on the 26th of September, Saturday. Over the last few years, the applications around NLP have increased tremendously, with use cases ranging from review analysis to intelligent chatbots in various industries. The workshop by AdaSci aims to take the participants on a learning ride with hands-on exposure to implementing NLP techniques in Python from scratch. The full-day workshop will cover topics such as the overview of NLP including syntax, semantics, tokenization; text analysis including topics such as stemming, feature extraction, Named Entity Recognition, sentiment analysis and more; machine learning algorithm for text classification; sequence modelling; overview of transformer, implementation of BERT, and more. With an increase in the interesting applications of NLP across the industries, there has been an increasing demand amid professionals to understand the technology, its working and developing interesting applications based on it. To facilitate the same, this workshop comes up with a learning dose where the participants will get hands-on exposure to implementing NLP techniques in Python from scratch. The detailed workshop will require candidates to have a knowledge of Python programming language, basic knowledge of linear algebra such as vectors, matrices and tensors, familiarity with Google Colab and GPU environment. To further make the experience seamless, attendees must install TensorFlow, Keras, TextBlob, and more. By the end of the workshop, candidates will be able to have a clear understanding of the fundamentals of NLP techniques, understand textual data representation, implement NLP techniques in different use cases, practical knowledge of advanced concepts such as transformers. Details of the workshop: Date: 26th September Timings (Full day): 10:00 am to 5:00 pm (IST) Mode: Online Pricing: $12.99 (workshop is free for ADaSci members) Register here","excerpt":"The Association of Data Scientists, the premier global professional body of data science & machine learning professionals, has announced a full-day workshop on Natural Language Processing (NLP) on the 26th of September, Saturday.  Over the last few years, the applications around NLP have increased tremendously, with use cases ranging from review analysis to intelligent chatbots […]","categories":["Deep Tech"],"tags":["ADAsci (Association of Data Scientists)","Natural Language Processing","NLP"],"author_name":"Srishti Deoras","publish_date":"2020-09-14T16:09:07","publication_year":"2020","word_count":304,"keywords":["data science","machine learning","ADAsci (Association of Data Scientists)","Keras","AI","Natural Language Processing","ML","Transformers","NLP","Colab","Aim","TensorFlow"],"extracted_tech_keywords":["AI","machine learning","ML","NLP","data science","Aim","TensorFlow","Keras","Transformers","Colab"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/join-this-full-day-workshop-on-natural-language-processing-from-scratch\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10112044,"title":"AI Brings 2000 Year Old Script Back to Life","content":"The Vesuvius Challenge, initiated ten months ago to tackle the enigma of the Herculaneum Papyri, has achieved a historic breakthrough, as announced by Nat Friedman. The project successfully decoded a portion of the ancient scrolls that were preserved in the aftermath of Mount Vesuvius’ eruption in 79 AD. Ten months ago, we launched the Vesuvius Challenge to solve the ancient problem of the Herculaneum Papyri, a library of scrolls that were flash-fried by the eruption of Mount Vesuvius in 79 AD.Today we are overjoyed to announce that our crazy project has succeeded. After 2000… pic.twitter.com\/fihs9ADb48— Nat Friedman (@natfriedman) February 5, 2024 The winning team, composed of Youssef Nader, Luke Farrito, and Julian Schilliger, secured the Vesuvius Challenge Grand Prize of $700,000. Youssef is an Egyptian PhD student in Berlin who successfully read a few columns of text back in October, earning him the second-place First Letters Prize. Luke is a 21-year-old college student and SpaceX intern from Nebraska. Julian is a Swiss robotics student at ETH Zürich, who won three Segmentation Tooling prizes for his incredible work on the Volume Cartographer. The deciphered text, originating from the first scroll, sheds light on the musings of the probable author, Philodemus, an Epicurean philosopher. Philodemus discusses various subjects, including music, food, and the pursuit of life’s pleasures. Notably, the text concludes with subtle criticism directed at unnamed ideological adversaries, believed to be the stoics, who are accused of having little to say about pleasure. The winning solution utilised a final canonical model based on the timesformer small architecture featuring a divided space-time attention mechanism. This innovative approach allowed for a comprehensive understanding of the ancient text, marking a significant breakthrough in the quest to unveil the mysteries of the past. The dataset underwent meticulous expansion and cleaning, involving approximately 15 rounds of refinement to enhance the accuracy of the labels. It also consisted of 2 other architectures, Resnet3D-101 with pretrained weights, I3D with non-local block and maxpooling. The Vesuvius Challenge is set to continue in 2024. The objective now is to progress from decoding isolated passages to reading entire scrolls. To spur further advancements, a new $100,000 grand prize has been introduced for the first team capable of reading at least 90% of all four scanned scrolls.","excerpt":"The winning team, composed of Youssef Nader, Luke Farrito, and Julian Schilliger, secured the Vesuvius Challenge Grand Prize of $700,000.","categories":["AI News"],"tags":[],"author_name":"Siddharth Jindal","publish_date":"2024-02-06T09:58:47","publication_year":"2024","word_count":375,"keywords":["Go","attention mechanism","programming_languages:R","AI","programming_languages:Go","ai_applications:robotics","Julia","ResNet","R"],"extracted_tech_keywords":["AI","R","Go","Julia","attention mechanism","ResNet","programming_languages:R","programming_languages:Go","ai_applications:robotics"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/ai-brings-2000-year-old-script-back-to-life\/","complexity_score":4,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10170033,"title":"How AI is Giving Patients Perfect Vision","content":"In ophthalmology, precision is paramount, and every millimetre counts. Even a slight misalignment can significantly affect a person’s vision and quality of life. Now, artificial intelligence (AI) is elevating standards by optimising the design and implantation of AI-designed intraocular lenses (IOLs). “Improvements in technology, surgical technique and the design of intraocular lenses have led to cataract surgery becoming a very safe and effective procedure,” Michele Lanza, an ophthalmology professor, stated in his paper published in Frontiers medical journal. This cutting-edge innovation took a leap with Asia’s first implantation of AI-designed IOLs at Nethradhama Super Speciality Eye Hospital, Bengaluru. IOLs refer to the tiny artificial lens implanted during cataract surgery. The surgery used advanced lenses like Rayner’s RayOne Galaxy and Galaxy Toric IOLs, which were launched in Barcelona last year and uniquely designed using AI. The spiral-shaped architecture provides a full range of vision with minimal halo and glare. The Galaxy Toric variant offers enhanced rotational stability, crucial for correcting astigmatism, an eye condition in which irregular cornea or lens shapes cause blurred or distorted vision. Patients benefit from increased spectacle independence and often achieve remarkable visual clarity within 24 hours post-surgery. “These lenses optimise light distribution, reduce visual disturbances and offer clarity across all distances, all of which directly improve patient satisfaction,” said Dr. Sri Ganesh, chairman and managing director of Nethradhama. “This is not about technological advancements alone, but about better patient outcomes and personalised care,” he added. Challenges of Integrating AI in Clinical Practice The introduction of AI-based medical devices in clinical practice has its challenges. Ensuring the technology meets stringent safety and efficacy standards requires rigorous internal validation and extensive clinical trials. “One of our key challenges was training surgeons to adapt to these lenses. AI may guide design, but the surgical execution still requires precision and familiarity,” said Dr Ganesh. Patients, too, approached the new technology cautiously. To address concerns, hospitals implemented detailed pre-surgery counselling alongside visual simulation sessions, helping patients better understand the expected outcomes. “When patients can visualise the expected outcomes, it helps ease hesitation and builds trust,” Dr. Ganesh observed. Beyond AI-designed IOLs, AI is being integrated across multiple stages of ophthalmic care. AI-powered imaging tools are now capable of detecting early signs of conditions such as diabetic retinopathy, glaucoma and age-related macular degeneration. Advanced algorithms also help eye care professionals determine the most suitable IOL power, design, and alignment based on each patient’s eye profile. AI also helps control lens positioning and incision placements, improving long-term visual outcomes. Predictive Analytics and Early Disease Detection Beyond immediate surgical applications, AI’s predictive analytics are helping in preventative ophthalmology by analysing vast datasets from diagnostics scans to identify individuals at high risk of developing eye diseases. “Machine learning models are helping us predict the likelihood of diseases like glaucoma or macular degeneration well before symptoms surface,” Dr Ganesh noted. Nethradhama has also partnered with Zeiss to develop an AI-powered surgery optimiser app that trains doctors in cataract surgeries. “It significantly reduces the learning curve and the chances of complications, something especially important in a high-volume country like India,” said Dr. Ganesh. Meanwhile, Bangalore-based ophthalmic innovator firm Remidio has received approval for its AI tools that rapidly detect glaucoma and age-related macular degeneration (AMD) using portable cameras. Designed to function offline, without the need for internet, these tools are ideal for screening in rural and remote areas. With just a tap, doctors can screen three major eye diseases within seconds. Looking ahead, the next wave of AI integration in ophthalmology promises to include real-time surgical guidance, robotic-assisted procedures, and AI-powered patient education tools. “We’re already seeing how visual simulation tools can help patients understand what their post-surgical vision might look like. It makes the decision-making process more informed,” said Dr. Ganesh. With AI evolving in healthcare, precision, personalised treatment, and patient confidence are reaching unprecedented heights, ushering in a new era of technology that can transform vision restoration.","excerpt":"AI-powered imaging tools are now capable of detecting early signs of conditions such as diabetic retinopathy, glaucoma and age-related macular degeneration.","categories":["AI Features"],"tags":["AI","IOLs","ophthalmology"],"author_name":"Merin Susan John","publish_date":"2025-05-15T13:55:45","publication_year":"2025","word_count":652,"keywords":["IOLs","Go","API","machine learning","artificial intelligence","AI","R","ophthalmology","Ray","analytics","Rust","predictive analytics"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","analytics","Ray","predictive analytics","R","Go","Rust","API"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-ai-is-giving-patients-perfect-vision\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10017323,"title":"IIT Gandhinagar Invites Applications For M.Sc. in Cognitive Science With Monthly Scholarship","content":"Indian Institute of Technology Gandhinagar has called for applications for its M.Sc. in Cognitive Science, which is an interdisciplinary field with cutting-edge applications in AI & ML. According to reports, the Cognitive Science discipline consists of philosophy, psychology, computer science, neuroscience, and Artificial Intelligence. In the present scenario, Cognitive Science is growing as a discipline in almost every major field. The Institute has a state-of-the-art Centre for Cognitive and Brain Sciences, which conducts high-quality research in the frontier areas of cognitive science on various cutting-edge topics such as perception, curiosity, attention, motor learning, decision making, prosthetics, rehabilitation, robotics, consumer behavior, multisensory perception, music cognition, and tactile perception, among other areas. The M.Sc. in Cognitive Science program places a strong framework of the curriculum including core courses, electives, and a research project. It will follow a structure of four semesters with a total of eighty credits required for graduation. The course is said to offer flexibility to the students in order to tailor the course according to their interests. The course transcends traditional academic boundaries through rigorous multi-disciplinary coursework, focused faculty-mentored projects, exciting summer internships, and an in-depth dissertation. Highlighting the unique qualities of this course, Prof Jaison Manjaly, Head, Discipline of Humanities and Social Sciences, IIT Gandhinagar, said, “The interdisciplinary nature of all the IIT Gandhinagar courses makes a student’s learning experience highly enriching. The MSc in Cognitive Science offers a deeper understanding of mind, brain, and cognition. These programs are also uniquely aligned to the IITGN’s vision to enable lifelong learning among the students.” Eligible candidates admitted to the M.Sc. in Cognitive Science program will receive a monthly scholarship of Rs. 5,000\/-.The deadline for M.Sc. in Cognitive Science is 31st January 2021. Due to the COVID-19 pandemic, both admission tests and interviews for the courses will be conducted online. Also, the institute encourages the M.Sc. in Cognitive Science students to present their research work at national and international conferences, and provide travel scholarships of up to Rs. 60,000\/-. For further information, click here.","excerpt":"Indian Institute of Technology Gandhinagar has called for applications for its M.Sc. in Cognitive Science, which is an interdisciplinary field with cutting-edge applications in AI & ML. According to reports, the Cognitive Science discipline consists of philosophy, psychology, computer science, neuroscience, and Artificial Intelligence. In the present scenario, Cognitive Science is growing as a discipline […]","categories":["AI News"],"tags":["computer science projects","project topics for computer science"],"author_name":"Ambika Choudhury","publish_date":"2021-01-07T17:33:01","publication_year":"2021","word_count":334,"keywords":["project topics for computer science","Go","artificial intelligence","lifelong learning","programming_languages:R","AI","ML","programming_languages:Go","RAG","GAN","computer science projects","R"],"extracted_tech_keywords":["AI","artificial intelligence","ML","RAG","R","Go","GAN","lifelong learning","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/iit-gandhinagar-invites-applications-for-m-sc-in-cognitive-science-with-monthly-scholarship\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10077530,"title":"Top AI Courses Launched in 2022","content":"Notwithstanding that these technologies are still new and developing, there is almost no field where artificial intelligence or machine learning has not permeated in some form or another. Learning about these rising technologies, therefore, becomes an important step to pushing your career forward and making groundbreaking innovations. We have compiled a list of a few courses launched in 2022 that freshers and professionals can pursue to march ahead in AI-ML. 1. MachineHack Pocket Courses Unlike full-fledged machine learning courses that are time-consuming and resource intensive, MachineHack has a bunch of pocket courses covering various topics on machine learning and analytics helping individuals learn faster. and participate in hackathons, perfecting those skills with in-depth knowledge. Click here to check out individual courses on topics like linear regression, sequence modelling, image segmentation. 2. Machine learning specialisation DeepLearning.AI and Stanford University Online have collaborated to create a beginner-friendly programme for teaching the fundamentals of machine learning and building real-world AI applications. It is a foundational specialisation with three courses taught by Andrew Ng, Aarti Bagul, Eddy Shyu, and Geoff Ladwig. The three-course programme is an updated version of Andrew’s machine learning course launched in 2012 and had over 4.8 million learners. The three courses included in the specialisation are — Supervised Machine Learning: Regression and ClassificationAdvanced Learning AlgorithmsUnsupervised Learning, Recommenders, Reinforcement Learning The course offers a broad introduction to modern machine learning using Python and other libraries like NumPy and scikit-learn. It encompasses topics such as building and training neural networks with TensorFlow, clustering and anomaly detection, building recommender systems, and building reinforcement learning models. To check out the courses on Coursera, click here. 3. IIT Mandi Data Science Course The Indian Institute of Technology, Mandi, has collaborated with National Skills Development Corporation to launch certificate courses on machine learning and data science starting in November. The courses will be conducted by the faculty members of IIT Mandi online. The programme will include foundational knowledge in data science and also specialise in machine learning with Python to allow learners to explore careers as data scientists, and business and data analysts. The course will last for nine months with certificates issued by IIT Mandi and NSDC. To find out more about the course, click here. 4. Google AI Research YouTube Google AI announced their Google Research YouTube channel in August with focus on subjects like robotics, AI\/ML, quantum computing, health and bioscience. The channel is a free resource for information about the subjects and includes three segments: Meet a Google researcher, where they demonstrate their new innovations and technologiesResearchBytes focuses on converting Google research publications into byte-sized content for easy understandingSpotlights discusses new technologies by Google in-depth, focusing on individual innovations at a time 5. IIT Jodhpur’s PGDM in Data Science and Cloud Computing In collaboration with WileyNXT, IIT Jodhpur launched a post-graduate diploma in cloud computing and data engineering. It is a 12-month intensive course teaching about managing the lifecycle of a data engineering project like big data collection and insight generation using machine learning techniques. The course requires a bachelor’s degree in science or engineering with a minimum of 50% marks. Along with this, the candidate must also be a working professional with a minimum of two years experience in the data science field. Click here to learn more and apply for the course. 6. IIT Madras-Sony India Finishing School Programme Pravartak Technologies of IIT Madras collaborated with Sony India Software Centre to launch a free course for pandemic-era kids with a family income less than INR 8 lakh. The course focuses on AI, ML, computer graphics, cybersecurity, and communication skills. Sony will offer employment to 15 top performers of the six-month course and the rest would be eligible to appear for job interviews at the IIT-M placement cell. The free course also offers a stipend to the top scoring candidates of the entrance examination. Click here to know more about the course. 7. Samsung Innovation Campus Samsung India launched their Samsung Innovation Campus to upskill students and young professionals in AI, IoT, Big Data, and coding and prepare them for future opportunities. The first batch will be of 3,000 underprivileged students across India as proposed by Electronics Sector Skills Council of India (ESSCI). During the course, the students will receive mentorship from Samsung researchers and classes from ESSCI-approved educational experts. The AI course will be conducted for 270 hours of theory and 80 hours of project work. To know more about the course, click here. 8. Fast.ai Practical Deep Learning for Coders Fast.ai has been developing several courses for the past few years, but recently announced their Practical Deep Learning for Coders 2022 course for teaching learners about building deep learning models. The course includes computer vision, NLP, tabular analysis, and collaborative filtering problems. By the end of this course, students will learn how to implement fundamentals of deep learning like gradient descent, stochastic gradient, and complete training loop. The course is taught by Jeremy Howard with nine lessons, each about 90 minutes long. Alumni of previous editions of the course have been working at GoogleBrain, OpenAI, Amazon, Tesla, and Adobe. Click here to know more about the programme.","excerpt":"MachineHack has a bunch of pocket courses covering various topics on machine learning and analytics","categories":["AI Trends"],"tags":["Courses"],"author_name":"Mohit Pandey","publish_date":"2022-10-18T17:00:00","publication_year":"2022","word_count":858,"keywords":["data science","machine learning","artificial intelligence","AI","neural network","ML","computer vision","NLP","deep learning","analytics","Courses"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","neural network","NLP","computer vision","data science","analytics"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-ai-courses-launched-in-2022\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":27893,"title":"Bot-Heavy Data Labelling Platforms Are The Cause Of Bad Results In AI Research","content":"High-quality training data is a critical asset to the success of artificial intelligence applications and products. But now, data labelling service provider Amazon Mechanical Turk has come under the scanner in the social sciences research community for giving bad results. So far, Amazon Mechanical Turk has been the best way of getting training data for machine learning algorithms. A recent report by a noted news portal, indicated that researchers who use Amazon Mechanical Turk for academic studies are noticing an increase in bad results on account of the survey questions. The output for bad results has been cited as bots, or human-augmented bots or even humans themselves. There is also a fear about the robots replacing human Turkers in some way, and as a result making the platform less reliable for the types of research they’re conducting. Amazon launched Mechanical Turks in 2005 as a tool for crowdsourcing training and which paid people a small amount of money for performing tasks through the desktops. Mechanical Turk or MTurk is essentially a crowdsourcing marketplace where the Requester publishes and coordinates a wide set of Human Intelligence Tasks (HITs), such as classification, tagging, surveys, and transcriptions. And users can choose from the tasks, thereby earning a small amount of money for each completed task. Since its launch, the platform’s popularity has soared and users are known to label a 1,000 record-dataset for a fee of $300 (plus fees) in a few hours. According to a report, a recent Mechanical Turk listing offered workers 80 cents to read a restaurant review and then answer a survey about their impressions of it; the time limit was 45 minutes. Of late, many researchers use Mechanical Turk as a means of human-labelled training data which is leading to bad output for models. High-quality training data is a major requirement for driving deep-learning-based approach to AI. As the stakes get higher for researchers and consumer-based AI applications are being rolled out rapidly, the reliance on quality training data has increased. Researchers believe it is becoming easy to spot the potential problems of a bot-heavy Mechanical Turk platform which are returning nonsensical responses or labels. And because of the mislabeling or erroneous labels\/answers, researchers believe it is skewing the accuracy of the resulting models. Is MTurk Turning Into A Menace For AI Researchers With the need for training data going up, organisations and researchers are increasingly relying on third parties\/platforms and startups that are providing Training Data As a Service solution. For example, Seattle based startup Mighty AI provides training data to companies that build computer vision models for autonomous vehicles. These as-a-service solution providers are doing a good job in labeling sensitive data but they should be vetted more responsibly. Mighty AI and another startup Figure Eight, transforms real-world messy data into training data and are known to have a stricter vetting process. Not just that, training data as a service provider is also giving companies and startups a platform to automate data-labeling. Besides, the rise in startups who provide synthetic data, is soon becoming the go-to approach to solving data-labeling problem with researchers relying on both — training data and synthetic data. Synthetic data or simulated datasets are secure and would not lead to loss of data information. Synthetic Data Generation The Answer To Data Inaccuracies In an earlier report, we revealed how Sergey Nikolenko, chief research officer at Neuromation emphasised that synthetic data is a more efficient way of getting perfectly labelled data for recognition. He shared that the synthetic data approach has proven to be very successful, and now the models trained by Neuromation are already being implemented in the retail sector. Synthetic Data generation greatly reduces the manual work required to label data, replicated data is labelled perfectly without any errors and is believed to be a useful tool for testing the scalability of algorithms and the performance of new software. Synthetic data platforms are able to perform a range of low-level tasks and cut down on the requirement of human labour. Increasingly, synthetic datasets are becoming a part of the data strategy and have also sparked a notion of open data economy.","excerpt":"High-quality training data is a critical asset to the success of artificial intelligence applications and products. But now, data labelling service provider Amazon Mechanical Turk has come under the scanner in the social sciences research community for giving bad results. So far, Amazon Mechanical Turk has been the best way of getting training data for […]","categories":["AI Features"],"tags":[],"author_name":"Richa Bhatia","publish_date":"2018-08-31T12:49:00","publication_year":"2018","word_count":690,"keywords":["Go","API","machine learning","artificial intelligence","TPU","AI","Scala","computer vision","GAN","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","computer vision","TPU","R","Go","Scala","API","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/bot-heavy-data-labelling-platforms-are-the-cause-of-bad-results-in-ai-research\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10018373,"title":"AdaBoost Vs Gradient Boosting: A Comparison Of Leading Boosting Algorithms","content":"In recent years, ensemble learning or boosting has become one of the most promising approaches for analysing data in machine learning techniques. The method was initially proposed as ensemble methods based on the principle of generating multiple predictions and average voting among individual classifiers. Researchers from the Institute for Medical Biometry, Germany, have identified the key reasons for the success of statistical boosting algorithms as: (i) The ability of the boosting algorithms to incorporate automated variable selection and model choice in the fitting process, (ii) The flexibility regarding the type of predictor effects that can be included in the final model and (iii) The stability of these algorithms in high-dimensional data with several candidate variables rather than observations, a setting where most conventional estimation algorithms for regression settings collapse. Here, we have compared two of the popular boosting algorithms, Gradient Boosting and AdaBoost. AdaBoost AdaBoost or Adaptive Boosting is the first Boosting ensemble model. The method automatically adjusts its parameters to the data based on the actual performance in the current iteration. Meaning, both the weights for re-weighting the data and the weights for the final aggregation are re-computed iteratively. In practice, this boosting technique is used with simple classification trees or stumps as base-learners, which resulted in improved performance compared to the classification by one tree or other single base-learner. Gradient Boosting Gradient Boost is a robust machine learning algorithm made up of Gradient descent and Boosting. The word ‘gradient’ implies that you can have two or more derivatives of the same function. Gradient Boosting has three main components: additive model, loss function and a weak learner. The technique yields a direct interpretation of boosting methods from the perspective of numerical optimisation in a function space and generalises them by allowing optimisation of an arbitrary loss function. The Comparison Loss Function: The technique of Boosting uses various loss functions. In case of Adaptive Boosting or AdaBoost, it minimises the exponential loss function that can make the algorithm sensitive to the outliers. With Gradient Boosting, any differentiable loss function can be utilised. Gradient Boosting algorithm is more robust to outliers than AdaBoost. Flexibility AdaBoost is the first designed boosting algorithm with a particular loss function. On the other hand, Gradient Boosting is a generic algorithm that assists in searching the approximate solutions to the additive modelling problem. This makes Gradient Boosting more flexible than AdaBoost. Benefits AdaBoost minimises loss function related to any classification error and is best used with weak learners. The method was mainly designed for binary classification problems and can be utilised to boost the performance of decision trees. Gradient Boosting is used to solve the differentiable loss function problem. The technique can be used for both classification and regression problems. Shortcomings In the case of Gradient Boosting, the shortcomings of the existing weak learners can be identified by gradients and with AdaBoost, it can be identified by high-weight data points. Wrapping Up Though there are several differences between the two boosting methods, both the algorithms follow the same path and share similar historic roots. Both the algorithms work for boosting the performance of a simple base-learner by iteratively shifting the focus towards problematic observations that are challenging to predict. In the case of AdaBoost, the shifting is done by up-weighting observations that were misclassified before, while Gradient Boosting identifies the difficult observations by large residuals computed in the previous iterations.","excerpt":"In recent years, ensemble learning or boosting has become one of the most promising approaches for analysing data in machine learning techniques. The method was initially proposed as ensemble methods based on the principle of generating multiple predictions and average voting among individual classifiers. Researchers from the Institute for Medical Biometry, Germany, have identified the […]","categories":["Deep Tech"],"tags":["decision tree algorithm","Ensemble Learning","gradient boosting"],"author_name":"Ambika Choudhury","publish_date":"2021-01-18T15:00:00","publication_year":"2021","word_count":564,"keywords":["Go","machine learning","decision tree algorithm","programming_languages:R","AI","programming_languages:Go","Ensemble Learning","RAG","gradient boosting","R"],"extracted_tech_keywords":["AI","machine learning","RAG","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/adaboost-vs-gradient-boosting-a-comparison-of-leading-boosting-algorithms\/","complexity_score":3,"technical_depth":7,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":25222,"title":"Why Are Humans Obsessed With Making AI &#8216;Moral&#8217;?","content":"Researchers across the globe are working extensively towards achieving an artificially intelligent system that can behave in an ethically and morally right manner. Morality, or the nature of distinguishing good from the bad, is an important human trait which researchers are now obsessively looking to infuse into machines. But, why are humans obsessing over it? Is that even a true human trait? Hasn’t history shown us that humans are capable of things worse than any AI could possibly do? The Moral Machine The concerns over morality often arise while talking about AI in areas like self-driving cars. Who dies in the car crash? Should it protect the passengers or passers-by? The Moral Machine, an initiative by the Massachusetts Institute of Technology gathers a perspective on moral decisions made by AI and machine learning. As a part of this initiative, participants could give their opinions on what AI in cars should do when confronted with a moral dilemma. Some of the common questions asked in this initiative to ‘crowdsource’ morality were: Should the self-driving car run down a pair of joggers instead of a pair of children Should it hit the concrete wall to save a pregnant woman or a child Should it put the passenger’s life at risk in order to save another human? The researchers then created an AI based on this data, teaching it the most ‘predictably’ moral thing a human could do. The initiative was led by a collaboration between Carnegie Mellon assistant professor, Ariel Procaccia and one of MIT’s Moral Machine researchers, Iyad Rahwa, who designed it to evaluate various moral situations that an AI can encounter. Though it sounds like an interesting concept, how can the reliability of the machine based on crowdsourced morality be ensured? It couldn’t exactly be trusted for making complex decisions such as those around saving human lives. As experts believe, to decide on hundreds of millions of variations based on views of few millions couldn’t possibly be the best way. Professor James Grimmelmann from Cornell Law School had said, “Crowdsourced morality doesn’t make the AI ethical. It makes AI ethical or unethical in the same way that large numbers of people are ethical or unethical.” In a similar effort, Germany released the world’s first ethical guidelines for the artificial intelligence of autonomous vehicles. Developed by the Ethics Commission at the German Ministry of Transport and Digital Infrastructure, the guidelines stated that self-driving cars must prioritise human lives over animals, whilst also restricting them from making decisions based on age, gender and disability. Why Humans Are Obsessed About ‘Moral’ AI In an earlier survey carried out by MIT, while many shared genuine views, there were others who agreed that though self-driving car should sacrifice its own passenger when faced with a calamity, they would not prefer to ride in the cars themselves. This ambiguity in their thoughts raises questions about how ethical AI system could actually be, if their (human’s) own opinions are disparate. Being truly abstract in nature, teaching morality to AI — which could be done best with measurable metrics — is next to impossible. In fact, considering instances, such as the one mentioned above, it is even questionable if humans have a sound understanding of morality that all of us can agree upon. ‘Instinct’ or ‘gut feeling’ takes over precedence in many cases. For instance, an AI player can excel in games with clear rules and boundaries by learning to optimise the score, but it has to work harder when it comes to mind games such as Chess or Go. We have, however, seen in past instances where Alphabet’s DeepMind was able to beat the best human players of Go. But in real-life situations, optimising problems could be more complex. For example, teaching a machine to algorithmically overcome racial and gender biases or designing an AI system that has a precise conception of what fairness is, can be a daunting task. Remember Microsoft’s AI chatbot that learnt to be misogynist and racist in less than a day? To teach AI the nuances of being ethically and morally correct is definitely not a cakewalk. Can AI Be Moral If we assume that a perfect moral system was to exist, we could derive this perfect moral system by collecting massive amounts of data on human opinions and analyse it to produce correct results. If we could collect data on what each person thinks is morally correct, and track how these opinions change and evolve over time and over generations, probably we could have enough inputs to train AI with these massive data-sets to be perfectly moral. Though this gives a hope of building moral AI, since it relies on human inputs, it would be susceptible to human imperfections. Unsupervised data collection and analysis could in fact produce undesirable consequences and result in a system that actually represents the worst of humanity. On A Concluding Note Despite fears by the likes of legendary scientists Stephen Hawking, arguing that once humans develop full AI, it will take off on its own and redesign itself at an ever-increasing rate, humans seem to be indulging in conversations around the importance of programming morality into AI. Tech giant Elon Musk has also time and again warned that AI may constitute a “fundamental risk to the existence of human civilisation”. Though these fears seem only reasonable, it cannot be denied that there is a need for more ethical implementation of AI systems, with a hope for engineers to imbue autonomous systems with a sense of ethics. It would be only ethically correct to have a moral AI that builds upon itself over and over again, and improves on its moral capabilities as it learns from previous experiences — just like humans do.","excerpt":"Researchers across the globe are working extensively towards achieving an artificially intelligent system that can behave in an ethically and morally right manner. Morality, or the nature of distinguishing good from the bad, is an important human trait which researchers are now obsessively looking to infuse into machines. But, why are humans obsessing over it? […]","categories":["IT Services"],"tags":[],"author_name":"Srishti Deoras","publish_date":"2018-06-07T10:52:24","publication_year":"2018","word_count":956,"keywords":["Go","machine learning","artificial intelligence","programming_languages:R","AI","programming_languages:Go","Git","Rust","R","programming_languages:Rust"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","R","Go","Rust","Git","programming_languages:R","programming_languages:Go","programming_languages:Rust"],"url":"https:\/\/analyticsindiamag.com\/it-services\/why-are-humans-obsessed-with-making-ai-moral\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10016764,"title":"As Chinese Semiconductor Companies Look To Raise Money, What Does It Mean For The Chip War?","content":"Chinese chipmaker Horizon is seeking to raise $700 million through C round funding after the five-year-old unicorn received $150 million in round A. The news comes at a time as China tries to become self-reliant and independent from western chipmakers because of the US’s various sanctions that have limited the supply of chips into the country. Emerging technologies play a critical role in modern-day geopolitics and whoever has resources to build these technologies, including the processing chips, are at an advantage over others. Thus, controlling the chip market in the 21st century might prove to be as crucial as controlling oil in the 20th century. The article tries to analyse the geopolitical implications of the ‘chip war’ between the US and China and what would the effort made by Chinese semiconductor companies to raise money for advanced chipmaking mean for the overall situation. The US Is Trying To Stop Tech Advances By China Until now, the US has had a monopoly over the chip market, and this gives it a tremendous geopolitical advantage over others. Even though manufacturing was done in different countries, intelligence and key components have been coming from the US. In recent times, the US has used this to its advantage by declaring sanctions to limit the number of processing chips its rivals can have access to so that it can control their technological progress. The ‘chip war’ resulting from this has seen the US ban Taiwan Semiconductor Manufacturing Company (TSMC) and other chipmakers from supplying China with any chip that uses American equipment. Last week, the US banned the state-owned Semiconductor Manufacturing International Corporation (SMIC) in Beijing from US suppliers’ buying critical components by mandating US exporters to apply for a license. These developments have put tremendous pressure on China and mainly Huawei, the Chinese smartphone company, whose production has been hampered as they are running out of processor chips to build their phones. This has become a pivotal issue in the geopolitical standoff between Washington and Beijing. These moves by the US are in line with other initiatives like its participation in the D10 – an alliance of ten democracies formed explicitly to explore alternatives to the 5G technology developed by Huawei. The US Department of Commerce has said that it is willing to give exporters licenses to any product that will not be used for Huawei’s 5G. Thus, it appears that the US is determined to thwart any effort made by China or Chinese companies to develop advanced technologies. This helps them gain a better geopolitical advantage while forming tech-related alliances with other democracies. China’s Impetus To Grow While the US and Western democracies resist China’s tech advances, the effort by Horizon and other semiconductor companies in China to raise funding is indicative of its intention to fight back. If it has to gain an actual geopolitical advantage, China needs to play a more relevant role than be a low-cost factory for chip manufacturing. In the past, China has proven that they can do it. “Right now, the Chinese chip industry is where the Chinese software industry was a decade ago,” said Abishur Prakash, geopolitical futurist in conversation with Analytics India Magazine. “At that time, nobody imagined Chinese software emerging and rivalling what the US had to offer. In the same light, as the US cuts off China from chips, Beijing is doubling down on building its domestic chipmaking capabilities,” he added. Even if the US is making a lot of tech allies and playing offensive, China is not likely to bog down. While only 16% of computer chips used in 2019 in China were produced domestically, China has a plan to ensure this number is 70% by 2025. Last month, FT even reported on Huawei’s plan to set up its own chip plant to get around the US sanctions. Hence, it might take time, but China has the power and the will to unshackle itself from Western dependency and create its own market. “Until then, the chip wars are really about the US stopping China’s rise with tech, and on the other hand, China trying to challenge US dominance through tech,” said Prakash. Wrapping Up While the chip manufacturing trade sanctions were put by Donald Trump, there are various speculations on how president-elect Joe Biden will deal with China. Experts think that Biden might go softer on China, but the tensions over technology likely won’t go away. With the participation of the US in the likes of D10 and Global Partnership in Artificial Intelligence (GPAI), and the revival of the Quad using various tech and information sharing alliances, all of which exclude China, the US has taken a direction of creating a narrative of democratic alliances to counter China’s influence. A targeted approach with American allies, with the narrower aim of keeping the most advanced technology out of China’s hands, could follow. But this is much more complex than it appears. While allies like Taiwan sided with the US (for now) after TSMC and chip firms were banned from supplying China, South Korea deemed the move ‘unacceptable’ owing to the profits it makes from the sales and the help it gets from China to keep North Korea in check.","excerpt":"Chinese chipmaker Horizon is seeking to raise $700 million through C round funding after the five-year-old unicorn received $150 million in round A. The news comes at a time as China tries to become self-reliant and independent from western chipmakers because of the US’s various sanctions that have limited the supply of chips into the […]","categories":["AI Features"],"tags":["AI Chip Wars","AI Companies","AI What it Does","china chip technology","Chip Manufacturing","Huawei","recent technological advances"],"author_name":"Kashyap Raibagi","publish_date":"2020-12-30T10:00:00","publication_year":"2020","word_count":866,"keywords":["Go","funding","artificial intelligence","unicorn","programming_languages:R","AI","AI Chip Wars","programming_languages:Go","Aim","china chip technology","AI What it Does","recent technological advances","AI Companies","analytics","Huawei","R","Chip Manufacturing"],"extracted_tech_keywords":["AI","artificial intelligence","analytics","Aim","R","Go","unicorn","funding","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/as-chinese-semiconductor-companies-look-to-raise-money-what-does-it-mean-for-the-chip-war\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":26453,"title":"Pitfalls Of AI In Healthcare — The Holy Grail Of Personalised Medicine","content":"Image: GE Healthcare If AI is the engine of growth, then the healthcare AI market is definitely getting geared up for the ultimate boom. As AI technology paves the way for smarter healthcare systems and workflow improvements in India, it also brings a host of issues such as data integrity, algorithmic accountability and bias in data. The most prominent revenue churners in AI applications are virtual assistants, optimisation systems for administrative workflows and robot-assisted surgery. An Accenture report calls AI the new OS in health. But behind the promise and potential of AI in healthcare, lies the risk of patient safety. Much has already been written about the growing volumes of healthcare data, which in part has spurred the AI revolution in medicine. For example, an IDG report claims healthcare industry is facing a “data tsunami” which has grown from 153 exabytes in 2013 to 2,314 exabytes by 2020. Behind the stupendous growth lies the threat of shadow IT and compliance issues. Exploring the medical risks posed by AI should be as much part of the study as the applications for adoption in healthcare. A study estimates that currently, 50 percent of more than 3.4 billion smartphone and tablet users have downloaded mobile health apps. To counter the risks, policymakers, hospitals and universities will have to come together to formulate guidelines and policies on best practices on data collection, data usage, how to integrate new technologies and the potential risks associated with clinical and non-clinical data. Here’s why — in a clinical setting, patient data necessitate the highest level of accuracy, reliability, security and privacy a better place to take advantages of the AI. Risks of AI in healthcare Algorithmic Accountability In Medical Data: Dubbed as the “black box of medicine”, big data techniques are only as potent as the data fed into the models. Given the opaque nature of AI technology, a doctor or a patient will not be able to understand how a prediction was made or whether the algorithmic conclusion is correct. Even though the algorithms deployed are validated before, an incorrect output can affect patient outcomes. Today, AI is leveraged from deciding the choice of drugs to personalised treatments, in view of the scarce hospital resources and low doctor-patient ratio. A study hints that while algorithms trained on reams of data points can churn predictive correlations, there is no way to assess whether the outcome was biased, accurate or inaccurate. That’s why the medical community and AI tech providers need to validate the algorithms on a diverse dataset in order to deliver a fair outcome. Risks Of Curated Datasets: Given the complexity of medical data and its disparate sources, putting together curated datasets is an expensive and time-consuming task. Researchers require dataset that have the breadth and depth for training and this can spark privacy concerns. Data Privacy And Black Box Medicine: How can one ensure algorithmic accountability without sacrificing data privacy? This feverish debate has drawn opinions from across the board — healthcare community, tech experts and even ethicists who worry about the risks posed by leakages related to clinical data and clinical trial data. Accountability, transparency and privacy do not go hand-in-hand — ethicists claim to limit the use of clinical, non-clinical data researchers companies use to validate models, while researchers claim that to ensure transparency, one requires a diverse dataset for driving a fair patient outcome. That is why researchers have proposed robust data governance policies to keep out unintended users and ensure patient data is used ethically. CIOs Grapple With Data Governance Policies In Healthcare: As AI and cloud technology permeates the healthcare industry, CIOs are increasingly investing in off-the-shelf cloud-based solutions which can pose data security threats. Besides cybersecurity governance, CIOs are also grappling with compliance and data ownership. Then, there’s also the issue of shadow IT, a growing security threat to healthcare organisations that needs to be addressed. Another reason why it is acuter is because a breach of data can prove too costly for organisations that have to face financial penalties and loss of patients’ trust. A report suggests that every breached health record can cost $355.7 to healthcare organisations that not only lose data but also user trust. Outlook It is true there won’t be any precision medicine without AI and the role of technology is only set to grow. According to the Accenture report, consumers are more likely to see AI having a positive impact. The technology will also deliver more benefits in terms of efficiency and greater interoperability. It will also lower the cost of care, combat physician shortage and reduce the burden of clinical demand.","excerpt":"If AI is the engine of growth, then the healthcare AI market is definitely getting geared up for the ultimate boom. As AI technology paves the way for smarter healthcare systems and workflow improvements in India, it also brings a host of issues such as data integrity, algorithmic accountability and bias in data. The most […]","categories":["AI Features"],"tags":["AI Healthcare"],"author_name":"Richa Bhatia","publish_date":"2018-07-16T06:09:36","publication_year":"2018","word_count":770,"keywords":["big data","Go","TPU","AI","AI Healthcare","virtual assistants","RAG","Aim","data governance","Rust","R"],"extracted_tech_keywords":["AI","Aim","RAG","virtual assistants","TPU","R","Go","Rust","big data","data governance"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/pitfalls-of-ai-in-healthcare-the-holy-grail-of-personalised-medicine\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10005671,"title":"Researchers Identify Fifty New Planets Using Machine Learning","content":"Researchers from the University of Warwick have identified 50 new planets using machine learning algorithms that were used to determine the real, fake or false positives by calculating the probability of each candidate to be a true planet. This is the first time ever that astronomers have used used a process based on machine learning to analyse a sample of potential planets. Previous ML techniques have ranked candidates, but never determined the probability that a candidate was a true planet by themselves. The ML algorithm was built on large samples of thousands of candidates found by telescope missions such as NASA’s Kepler and TESS, by researchers from Warwick’s Departments of Physics and Computer Science, as well as The Alan Turing Institute. It was trained to recognise real planets using two large samples of confirmed planets and false positives from the now-retired Kepler mission. The researchers then used the algorithm on a dataset of still unconfirmed planetary candidates from Kepler, resulting in fifty new confirmed planets and the first to be validated by machine learning. Dr David Armstrong, from the University of Warwick Department of Physics, said, “In terms of planet validation, no-one has used a machine learning technique before. Machine learning has been used for ranking planetary candidates but never in a probabilistic framework, which is what you need to truly validate a planet.” “Rather than saying which candidates are more likely to be planets, we can now say what the precise statistical likelihood is. Where there is less than a 1% chance of a candidate being a false positive, it is considered a validated planet,” he added. Researchers believe that this new technique is faster than previous techniques as it can be automated and improved with further training. “We still have to spend time training the algorithm, but once that is done it becomes much easier to apply it to future candidates. You can also incorporate new discoveries to progressively improve it,” added Dr Armstrong.","excerpt":"Researchers from the University of Warwick have identified 50 new planets using machine learning algorithms that were used to determine the real, fake or false positives by calculating the probability of each candidate to be a true planet. This is the first time ever that astronomers have used used a process based on machine learning […]","categories":["AI News"],"tags":[],"author_name":"Srishti Deoras","publish_date":"2020-08-27T11:29:44","publication_year":"2020","word_count":326,"keywords":["Go","machine learning","programming_languages:R","AI","ML","programming_languages:Go","R"],"extracted_tech_keywords":["AI","machine learning","ML","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/researchers-identify-fifty-new-planets-using-machine-learning\/","complexity_score":3,"technical_depth":7,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10142407,"title":"AWS had a Hard Time Fitting in All of Bedrock’s Innovations at re:Invent 2024","content":"At AWS re:Invent in Las Vegas, Amazon Web Services (AWS) has announced exciting updates to Amazon Bedrock, its platform for creating and running AI applications. “One of the hardest parts was figuring out how much we could fit in,” resonated AWS chief Matt Garman, reflecting on the sheer scale of advancements in Bedrock. “Fortunately, Swami will dive deeper into a ton more during his keynote tomorrow.” Garman said that Bedrock is by far the easiest way to build and scale generative AI applications. One big addition to Bedrock includes Automated Reasoning Checks, a tool designed to stop AI from making factual mistakes, aka hallucinations. This is especially useful for industries like healthcare and finance, where accuracy is critical. “Automated reasoning checks prevent factual errors due to model hallunciations,” said Garman. AWS further claimed that it helps ensure AI provides correct and trustworthy answers without needing advanced AI expertise. For example, PwC is using Automated Reasoning checks to build accurate, trustworthy AI assistants and agents to drive its clients’ businesses to the leading edge. In addition to this, AWS announced the launch of Model Distillation. This lets users shrink large AI models into smaller ones without losing much accuracy. Smaller models are faster and cheaper to run. “Model Distillation in Bedrock delivers models that are 500% faster and 75% cheaper,” shared Garman. For instance, Robin AI is already using this to save money while providing quick, accurate answers for legal questions. With Amazon Bedrock Model Distillation, customers can choose the optimal model for their use case and a smaller model from the same family, balancing application latency with cost efficiency. The company claimed that it works best with models from Anthropic and Meta, alongside its latest in-house Nova-series of models. “With a broad selection of models, leading capabilities that make it easier for developers to incorporate generative AI into their applications, and a commitment to security and privacy, Amazon Bedrock has become essential for customers who want to make generative AI a core part of their applications and businesses,” said Dr. Swami Sivasubramanian, vice president of AI and Data at AWS. AWS also showcased Bedrock’s ability to manage and coordinate multiple AI agents for large-scale, complex workflows. “Bedrock agents can now support complex workflows with agent collaboration, enabling seamless coordination for sophisticated tasks,” shared Garman. Moody’s uses Amazon Bedrock’s multi-agent system to improve risk analysis, with each agent handling specific tasks. This makes their assessments faster and more accurate, strengthening their position as a financial leader. What’s Next? Garman said while generative AI is still in its early stages, Bedrock is positioning itself as a leader by offering innovative tools and models that address real-world challenges. “This is just a sampling of the new capabilities that we’re announcing this week. Bedrock gives you the best models, the right tools, and capabilities you cannot get anywhere else,” added Garman in his keynote address on the future vision of Bedrock.","excerpt":"“Bedrock gives you everything you need to integrate generative AI into production applications, not just proof of concepts,” says AWS CEO Matt Garman.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","AWS"],"author_name":"Aditi Suresh","publish_date":"2024-12-04T01:05:20","publication_year":"2024","word_count":487,"keywords":["Anthropic","AI assistants","AWS","AI","ML","Aim","AI agents","generative AI","Rust","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","ML","generative AI","Anthropic","Aim","AI assistants","AWS","R","Rust","AI agents"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/aws-had-a-hard-time-fitting-in-all-of-bedrocks-innovations-at-reinvent-2024\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":24806,"title":"Congress Party Appoints Office-Bearers For Their Data Analytics Department","content":"The Indian National Congress on Monday announced the appointment of the new heads for their AICC Data Analytics Department. In a press statement, the INC appointed Swapna Patronis as the national coordinator, and Ashok Varadrajan and Aditeshwar Singh Deo as the joint coordinators of the department. INC COMMUNIQUE Announcement of the National Coordinator and the Joint Coordinators of the AICC Data Analytics Department. pic.twitter.com\/DFFo9Jmjb7 — INC Sandesh (@INCSandesh) May 21, 2018 Earlier this February, the “grand old party” had announced that their computer department had been renamed as ‘Data Analytics Department’. Praveen Chakravarty, former investment banker, angel investor and a political economist, had been nominated the chairman for the same. The move to strengthen the Congress’ analytics department may stem from the fact that Cambridge Analytica whistleblower Christopher Wylie had claimed that the data analytics firm which has been charged and convicted of data breach, worked extensively in India and “believed” that the Congress party had employed the firm for certain regional elections in India. Wylie’s recent statement was a part of his testimony before British Parliament’s Digital, Culture, Media and Sports Committee which is investigating the issue of the Facebook data breach. Wylie had said, “I believe their [CA’s] client was Congress. But I know that they have done all kinds of projects. I don’t remember a national project, but I know regionally. I mean India is so big that you know that one state can be, you know, as big as Britain.” In September 2017, Gandhi had also visited the US to meet noted personalities in the AI and analytics fields. He also reportedly met several think-tanks from Silicon Valley at the San Francisco Bay and discussed future technologies. The appointment of the national coordinator and joint coordinators for the Congress’ data analytics department comes hot on the heels of their bitter defeat against the Bharatiya Janata Party and Janata Dal (Secular) in Karnataka assembly elections. When the BJP failed to prove the majority in the Rajya Sabha by eight votes, Congress formed a majority-based with JD(S).","excerpt":"The Indian National Congress on Monday announced the appointment of the new heads for their AICC Data Analytics Department. In a press statement, the INC appointed Swapna Patronis as the national coordinator, and Ashok Varadrajan and Aditeshwar Singh Deo as the joint coordinators of the department. INC COMMUNIQUE Announcement of the National Coordinator and the Joint […]","categories":["AI News"],"tags":["BJP","cambridge analytica","congress","rahul gandhi"],"author_name":"Prajakta Hebbar","publish_date":"2018-05-22T11:34:08","publication_year":"2018","word_count":339,"keywords":["cambridge analytica","congress","programming_languages:R","AI","rahul gandhi","Git","BJP","Aim","analytics","GAN","R"],"extracted_tech_keywords":["AI","analytics","Aim","R","Git","GAN","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/congress-appoints-office-bearers-for-data-analytics-department\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":18041,"title":"Machine Learning – The Future of CPG Industry","content":"The Consumer Packaged Goods (CPG) industry is characterized by enormous data volumes with the very large number of transaction records, data coming from multiple sources etc. These unique characteristics of the CPG industry make it an ideal candidate for Machine Learning in combination with advanced predictive algorithms. The traditional analytics technologies struggle with the volume and complexity of the data and that is where exactly the Machine Learning is best suited. In CPG Industry, the Machine learning can be used for improving the effectiveness of marketing campaigns, increasing the performance of the sales team, optimizing the supply chain and streamlining manufacturing. But firstly, one needs to understand why machine learning is superior to the traditional techniques of business analytics that most of the people are using till now. Machine Learning vs traditional analytics and business analytics: Machine Learning (ML) is a branch of Artificial Intelligence (AI) which focuses on understanding the data intelligence and then automate the learning rules using machines to replicate the same or similar data. Historically, ML techniques and approaches heavily rely on computing power. On the other hand, Traditional Analytics techniques were mostly developed where computing power was not an option. As a result, Traditional Analytics heavily relies on small samples and fleshy assumptions about data and its distribution. Data Structure: While the traditional Analytics can analyze only structured data, the newfangled Big Data technologies and state-of-the-art machine learning algorithms can analyze any data format viz. images, videos, text, emails, unstructured social media messages, server, logs, etc. Data Consolidation: The traditional analytics focuses mainly on internal data but the cutting-edge machine learning technologies can easily create versatile datasets by allowing the merging of internal data with external public data which helps the data scientists to enrich sales and marketing data. Futuristic: Traditional analytics mostly focuses on past data analysis, while advanced machine learning algorithms focus on ingesting all the data and find patterns that can then be used to make accurate inferences about the future. Manual error: Since Machine Learning uses advanced computer algorithms to analyze the data and all the analysis are done totally impartially, so, they are free from any human biases. Faster Delivery: Cutting edge machine learning technologies allow the management to get answers from their data very quickly and efficiently as compared to traditional analytics techniques. So, the business questions can be answered in few weeks not years. Result Interpretation: While traditional Analytics & BI methods provide visualizations which are open and easy to interpretation, the predictive power of machine learning helps us answering smart questions like how many units of a particular product should be produced etc. Machine Learning In CPG Industry: Now, it is important to understand why CPG industry is an ideal candidate for predictive analytics & machine learning. Here are the unique attributes of the industry that makes it ideal for predictive analytics & machine learning: The huge transaction volume on a daily basis Short shelf life of the products Large number of consumer base Due to the enormous volumes of transactions generated by CPG organizations, it is very difficult to analyze such a huge volume of data manually as it overwhelms most of the brave analysts. Machine learning can easily handle these datasets to extract incredible insights from them. CPG products are characterized by short shelf life. Also, considering the large volumes of products any optimization to the oversupply (or undersupply) problem can have a huge impact on return on investment. Machine learning can solve these issues very easily. Accuracy in sales forecast also leads to a better efficiency of Sales & Marketing activities as well as campaign management. Cutting-edge learning based technology can easily merge many data sources (sales, marketing, digital, demographics, weather, etc.) which greatly improves the quality of sales forecasts, understand the underlying demand drivers that influence demand when compared to traditional predictions which are traditionally done on isolated and aggregated sales figures. Traditional demand forecasting and planning systems are restricted to mainly the demand history, while machine learning-based forecasting can take advantage of limitless data, determining what’s significant, then prioritize available consumer insights (demand sensing) to influence future demand using “what if” analysis (demand shaping). This improved Sales forecast can answer the question such as which product should be promoted in a particular month, what type of campaign will be most profitable for a particular product, what consumer segment should be targeted etc. Moreover, almost all large CPG organizations have wonderful ERP systems that hold a wealth of hidden value in their data. Using Machine Learning, this data can be used to create models that can answer several critical questions such as: How can on-time delivery be guaranteed? How to shorten the time to manufacture a product? etc. So, now is the time, the CPG industry should start leveraging modern technologies like machine learning to deliver remarkable customer experience!!","excerpt":"The Consumer Packaged Goods (CPG) industry is characterized by enormous data volumes with the very large number of transaction records, data coming from multiple sources etc. These unique characteristics of the CPG industry make it an ideal candidate for Machine Learning in combination with advanced predictive algorithms. The traditional analytics technologies struggle with the volume […]","categories":["IT Services"],"tags":["Business Analytics","Machine Learning"],"author_name":"Anjanita Das","publish_date":"2017-10-04T05:16:03","publication_year":"2017","word_count":806,"keywords":["Go","artificial intelligence","machine learning","AI","R","ML","Machine Learning","Git","RAG","analytics","Business Analytics","predictive analytics"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","analytics","RAG","predictive analytics","R","Go","Git"],"url":"https:\/\/analyticsindiamag.com\/it-services\/machine-learning-future-cpg-industry\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10126259,"title":"Tech Mahindra Partners with Microsoft to Transform Workplace Experiences Using Generative AI","content":"Tech Mahindra has announced a collaboration with Microsoft to enhance workplace experiences for its 1,200+ customers and over 10,000 employees across 15 locations through the adoption of Copilot for Microsoft 365. This partnership positions Tech Mahindra as a prominent Global Systems Integrator (GSI) in the realm of Copilot for Microsoft 365. The collaboration aims to boost workforce efficiency and streamline processes using Microsoft’s trusted cloud platform and generative AI capabilities. Additionally, Tech Mahindra will deploy GitHub Copilot for 5,000 developers, anticipating a productivity increase of 35% to 40%. Tech Mahindra is committed to empowering employees with AI tools to drive innovation, value, and sustainable growth. The organisation plans to extend the capabilities of Copilot with plugins to leverage multiple data sources and enhance creativity and productivity. The focus is on increasing efficiency, reducing effort, and improving quality and compliance across the board. Mohit Joshi, CEO and Managing Director of Tech Mahindra, emphasised the transformative potential of this partnership, stating, “Our vision is to redefine the workplace experience by empowering every employee to excel and innovate using cutting-edge AI technology. We are not just adopting a tool; we are shaping the future of work for our employees and customers. The collaboration with Microsoft, and the introduction of Copilot for Microsoft 365 and GitHub Copilot also marks a significant stride in Tech Mahindra commitment to making AI accessible to everyone.” As part of this initiative, Tech Mahindra has launched a dedicated Copilot practice to help customers unlock the full potential of AI tools. This includes workforce training for assessment and preparation, critical for integrating AI across all functions of an organisation. Tech Mahindra will offer comprehensive solutions to help customers assess, prepare, pilot, and adopt business solutions using Copilot for Microsoft 365, providing a scalable and personalised user experience. Judson Althoff, Executive Vice President and Chief Commercial Officer at Microsoft, remarked, “Our collaboration with Tech Mahindra will empower its employees with new generative AI capabilities to enhance workplace experiences and increase developer productivity through the adoption of Copilot for Microsoft 365 and GitHub Copilot. With a focus on driving AI innovation and skilling, Tech Mahindra is poised to deliver new solutions and greater value for its customers across industries.” This collaboration aligns with Tech Mahindra’s ongoing efforts to enhance workforce productivity using GenAI tools. Recently, Tech Mahindra launched a unified workbench on Microsoft Fabric, designed to help organisations accelerate the adoption of Microsoft Fabric and create complex data workflows with ease. The longstanding partnership between Microsoft and Tech Mahindra has also led to the development of industry-leading solutions such as: Generative AI-powered Enterprise Knowledge Search: Integrates Microsoft Azure OpenAI Service, Azure Cognitive Search, and Azure Language understanding to enhance organisational knowledge. Green CodeRefiner: Transforms existing code into energy-efficient code, optimising emissions. SenTindra: A cloud-based virtual security operations centre developed on Microsoft Sentinel. COMPASS-Cloud Security Assessment and Global System Integrator solutions: Focused on fortifying security frameworks and data governance capabilities.","excerpt":"As part of this initiative, Tech Mahindra has launched a dedicated Copilot practice to help customers unlock the full potential of AI tools.","categories":["AI News"],"tags":["Microsoft","Tech Mahindra"],"author_name":"Mohit Pandey","publish_date":"2024-07-09T13:02:22","publication_year":"2024","word_count":487,"keywords":["Tech Mahindra","GenAI","OpenAI","AI","R","ML","RAG","Aim","generative AI","edge AI","Azure","Microsoft"],"extracted_tech_keywords":["AI","ML","generative AI","GenAI","OpenAI","Aim","edge AI","RAG","Azure","R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/tech-mahindra-partners-with-microsoft-to-transform-workplace-experiences-using-generative-ai\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10042339,"title":"IISc &#038; IBM Launch New Hybrid Cloud Lab In Bengaluru","content":"IBM and IISc have now joined hands to launch the IBM-IISc Hybrid Cloud lab to advance research and development in hybrid cloud technologies. The lab is located at the IISc campus in Bengaluru. The lab will be co-chaired by Prof. Yogesh Simmhan, Associate Professor, Department of Computational and Data Sciences at IISc, Bengaluru and Dr. Amith Singhee, Senior Manager and Senior Technical Staff Member at IBM Research, India. Navakanta Bhat, Dean, Division of Interdisciplinary Sciences, IISc Bengaluru said, “Such industry-academia partnerships are key to boosting the impact of research, and we are proud to work together with a technology leader like IBM.” The lab will bring together academia and the open-source community. The initial set of projects will involve faculty and students from the Department of Computational and Data Science, Computer Science and Automation, and Supercomputing Education and Research Centre at IISc with IBM researchers. They will work on: Building AI-based autonomous and self-healing computing systems to predict emerging issues, diagnose and heal with maximum availability and minimum cost of operationsAdopting microservices and optimising cloud native applications to advance technologies such as Kubernetes. This will make IT services agile and enable cloud based business modelsCreating AI-based information systems to govern, consume, draw insights, and create value Developing AI systems to analyse human and machine languages The lab will take an open access approach. Gargi Dasgupta, Director, IBM Research India said, “The IBM – IISc lab will bring together two leading research organisations in industry and academia to create an ecosystem for Hybrid Cloud research, in India, for India and the world. IISc has a strong record in research areas like Hybrid Cloud, AI, Security, which compliments the expertise of IBM Research, and we are excited to collaborate with IISc to create innovative, industry-relevant solutions. Our Hybrid Cloud platform is Open, and we will jointly develop open-source software that provides interoperability, portability, and security that can be easily accessible to the vast community of developers to accelerate innovation”.","excerpt":"Such industry-academia partnerships are key to boosting the impact of research.","categories":["AI News"],"tags":["IISc","innovation"],"author_name":"Shraddha Goled","publish_date":"2021-06-24T15:02:58","publication_year":"2021","word_count":327,"keywords":["data science","Go","programming_languages:R","AI","innovation","IISc","microservices","automation","GAN","R","kubernetes"],"extracted_tech_keywords":["AI","data science","kubernetes","microservices","R","Go","GAN","automation","innovation","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/iisc-ibm-launch-new-hybrid-cloud-lab-in-bengaluru\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10014750,"title":"TensorFlow v2.4 Released: Roundup Of Major Features &#038; Updates","content":"TensorFlow released its new version 2.4 earlier this week. This version promises increased support for distributed training, a new NumPy frontend, and methods for monitoring and diagnosing bottlenecks. Notably, this release comes around the same time TensorFlow celebrated five years of its existence. Another important coincidence is that this release comes on the heels of an announcement from its rival Pytorch, which released its v1.7.1, just recently. Some of the new features and updates of TensorFlow’s new release are discussed in this article. New Features For Distributed Training Parameter Server Strategy: The TensorFlow Distribute module, which is TensorFlow API for distribution strategy, from v2.4 will have experimental support for asynchronous training for modules with Parameter Server Strategy. Experimental APIs are the ones which will be added eventually to TensorFlow but may be subject to change in the backwards compatibility ways. Parameter Server Strategy is a common data parallel method for scaling up a machine learning model on several machines. It consists of workers and parameter servers, where the ‘worker’ reads and updates the variables created on the parameter server. Since the whole process of reading and updating is done independently without synchronisation with each other, it is also called asynchronous training. Multi Worker Mirrored Strategy: This is an API Strategy that implements distributed training with synchronous data parallelism, like its counterpart Mirrored Strategy. The only difference between the two is that the former can enable training across multiple machines running on several GPUs each. To keep the variables in sync, Multi Worker Mirrored Strategy uses CollectiveOp, which is a single operation that can automatically choose an all-reduce algorithm based on the hardware and the network topology. In the new version, Multi Worker Mirrored Strategy has graduated from experimental and is now a part of a stable API. New Updates For Keras Mixed Precision: While most TensorFlow models use the 32-bit floating type, there are a few lower-precision models which may use 16-bit floating type to limit the use of memory. Mixed Precision makes use of both 16-bit and 32-bit floating-point types during training to make the model run fast and use less memory. The Mixed Precision API helps in improving the model performance by at least three times on GPUs and up to 60% of TPUs. In v2.4, Mixed Precision has moved out of the experimental phase and has been instated as a stable API. Optimizers: In TensorFlow, Optimizers are an extended class that includes additional information to train a specific model, while improving its speed and performance. Examples of Optimizers include — SGD, RMSprop, Adam, Adadelta, Adagrad, Adamax, Nadam, and Ftrl. The new TensorFlow release includes reconstructing the Optimizers class in Keras to enable the use of custom training loops. This update will help in writing the training code that works with any optimizer. Additionally, all the Optimizer subclasses will now accept gradient_transformers and gradient_aggregator arguments to easily define custom gradient transformations. Improvements to Functional API Model: The new release also includes a major refactoring of the internals of Keras Functional API. With this update, there is expected to be an improvement in the memory consumption of the functional model and also the simplification of the triggering logic. Experimental Support For NumPy API TensorFlow v2.4 now introduces the experimental support for a subset of NumPy APIs which helps in running the NumPy code. The NumPy API is built on top of TensorFlow, which enables to interoperate seamlessly with the latter. This allows access to all TensorFlow APIs and provides an optimised execution with the help of compilation and auto-vectorisation. GPU Support TensorFlow 2.4 will enable support for the newly introduced NVIDIA Ampere GPU architecture as it can run with both CUDA 11 and cuDNN 8. For the uninitiated, CUDA (Computer Unified Device Architecture) is a parallel computing platform, and API model from NVIDIA and cuDNN is a library for deep neural networks built using CUDA.The GitHub link can be found here.","excerpt":"TensorFlow released its new version 2.4 earlier this week. This version promises increased support for distributed training, a new NumPy frontend, and methods for monitoring and diagnosing bottlenecks. Notably, this release comes around the same time TensorFlow celebrated five years of its existence. Another important coincidence is that this release comes on the heels of […]","categories":["AI Trends"],"tags":["Pytorch"],"author_name":"Shraddha Goled","publish_date":"2020-12-18T15:00:00","publication_year":"2020","word_count":652,"keywords":["Pytorch","NumPy","machine learning","Keras","TPU","AI","neural network","PyTorch","ML","Transformers","TensorFlow"],"extracted_tech_keywords":["AI","machine learning","ML","neural network","TensorFlow","PyTorch","Keras","Transformers","NumPy","TPU"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/tensorflow-v2-4-released-roundup-of-major-features-updates\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10052247,"title":"Facebook Wants To Make Reinforcement Learning Easier","content":"Facebook AI has announced the release of ‘SaLinA,’ a lightweight library for implementing sequential decision models, including reinforcement learning (RL) algorithms. ​​According to Ludovic Denoyer, a research scientist at Facebook, “SaLinA is a Pytorch modification that allows users to combine agents instead of modules, giving the computation a time dimension. Classic RL algorithms may be constructed in a few lines using this abstraction, and they are not dependent on policy designs.” I am very excited to release SaLinA which is a lightweight library to implement sequential decision models, including reinforcement learning algorithms (but not only). A thread….Github: https:\/\/t.co\/zqpgEP7MneArxiv: https:\/\/t.co\/pkfZsGMzyo1\/8 pic.twitter.com\/aHJe682xM8— Ludovic Denoyer (@LudovicDenoyer) October 18, 2021 SaLinA is a lightweight library for developing sequential decision models that extends PyTorch components. It can be used for RL, as well as in supervised\/unsupervised learning scenarios. It enables the rapid development of extremely complicated sequential models (or policies) in a few lines.It is compatible with multiple CPUs and GPUs. Source: SaLinA What is SaLinA A sandbox for developing large-scale sequential models.A simple (300 hundred lines) ‘core’ programme that defines all of the components necessary to construct agents in sequential decision learning systems.It is simple to comprehend and use because it adheres to Pytorch’s fundamental ideas, just extending nn.Module to an agent to handle the temporal dimension.A collection of agents that may be combined in various ways (similar to Pytorch modules) to produce complex behaviours.A collection of implementations and samples from many fields. There are several types of learning: RL, imitation learning, and computer vision, with many more to come. Why SaLinA SaLina’s goal is to make the implementation of sequential decision processes, particularly those including RL, natural and easy for practitioners who have a working knowledge of how neural networks can be implemented. SaLina aims to handle any sequential decision problem by employing simple ‘agents’ that progressively process data. The intended audience includes researchers in natural language processing or computer vision and experts in natural language processing who are looking for a more natural way to model conversations in their models, trying to make them more straightforward and easily understood than previous methods. Advantages of SaLinA Simplicity: A working knowledge of the Agent and Workspace APIs is sufficient to comprehend SaLinA and create complicated sequential decision models. There are no hidden mechanics, and the two classes are extremely straightforward and intuitive to anyone who has used PyTorch.Modularity: SaLinA enables the construction of sophisticated agents through the use of predefined container agents.Flexibility: SaLinA’s flexibility is enhanced by the addition of tools that aid in the implementation of complex models. SaLinA includes wrappers for capturing openAI Gym settings, DataLoader environments, and Brax environments as agents, enabling rapid development of a diverse set of models. Scalability: SaLinA includes an NRemoteAgent wrapper that allows any agent to be executed over several processes, significantly speeding up the computation of any individual agent. When combined with the ability to run agents on either CPU or GPU, the library can scale to very big problems with only a few changes to the code. Additional features of SaLinA Speed: SaLinA is a complete Python library that incurs low overheads and performs on a par with existing alternatives.From policies to recurring policies: The workspace concept enables the easy implementation of complicated policies without modifying any other code.Replay Buffer: There is no need to design a sophisticated replay buffer class in SaLinA, as a collection of workspaces can naturally be utilised as a replay buffer due to the agents’ playback capability.Batch RL: It is simple to compute complex losses across defined trajectories utilising SaLinA’s replay feature. Model-based RL: Because environments in SaLinA are agents, it is feasible to replace any environment agent at any time with an agent that models the world. Multi-agent RL: SaLinA naturally supports multi-agent settings by integrating several agents into a single one.The SaLinA RL benchmark: SaLinA’s RL benchmark currently contains implementations of Double DQN, Reinforce, and Behavioral Cloning. Conclusion By integrating agents, SaLinA enables the implementation of sequential decision-making algorithms in a novel method. It is a small library, extremely flexible, and scalable. It enables the creation of new algorithms and the rapid evaluation of novel ideas without losing training or testing speed. Future directions include the following: a) enabling the execution of agents on remote computers; b) developing new tools for agent implementation; and c) establishing algorithms in various disciplines.","excerpt":"A versatile and simple library for sequential agent learning, including reinforcement learning","categories":["AI Features"],"tags":["Computer Vision","Facebook AI","GPUs","Natural Language Processing","Python","Python Library","Pytorch","Reinforcement Learning"],"author_name":"Dr. Nivash Jeevanandam","publish_date":"2021-10-24T12:00:00","publication_year":"2021","word_count":723,"keywords":["Pytorch","Go","Reinforcement Learning","Facebook AI","AI","neural network","Natural Language Processing","OpenAI","PyTorch","Scala","computer vision","Python","Aim","Computer Vision","Python Library","GPUs","R"],"extracted_tech_keywords":["AI","neural network","computer vision","OpenAI","Aim","PyTorch","Python","R","Go","Scala"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/facebook-ai-salina\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10045027,"title":"Musk’s Neuralink Raises Money, India’s AI for All And More In This Week’s Top News","content":"Looks like Elon Musk’s Neuralink has finally convinced a few biggies in the tech industry. On Thursday, Neuralink announced a Series C funding round of $205 million, which featured Google Ventures and OpenAI’s founder Sam Altman among others that included Fred Ehrsam co-founder of Coinbase and Ken Howery co-founder of PayPal.  Neuralink was founded by Elon Musk to help those with brain injuries in the near term and reduce AI risk to humanity in the long term. The company’s mission is to develop brain-machine interfaces that treat various brain related ailments, with the eventual goal of creating a whole brain interface capable of more closely connecting biological and artificial intelligence. Neuralink has spent the last four years building the first high channel count brain machine interface intended for therapeutic use in patients. Neuralink’s brain implants are intended to help quadriplegics regain their digital freedom by allowing users to interact with their computers or phones in a high bandwidth and naturalistic way. “The funds from the round will be used to take Neuralink’s first product to market and accelerate the research and development of future products,” said Neuralink in their press release. AI’s Patent Rights There is a case quietly making its way through the Federal Court of Australia that can change the way we look at the concept of invention. According to reports, an Australian court has been tasked with finding an answer to whether a patent can be granted for an invention that was devised by a machine, or is a human inventor required. A US based computer programmer claims that his AI project, DABUS (‘Device for the Autonomous Bootstrapping of Unified Sentience’) is capable of inventing on its own. The AI Inventor Project(AIP) has already filed patent applications in 17 countries, including the US and the UK. To even have an argument like this is unprecedented and can shake up the foundations of research, ethics and innovation. Apple Gets Gassed On Friday, Taiwan Semiconductor  Manufacturing Company(TSMC) said that one of its plants that makes Apple processors was hit by contamination of gas.  According to Nikkei, the factory — dubbed Fab 18 — is the company’s most advanced chip making facility. All of the latest processors for upcoming iPhones and Mac computers are produced there. “Some TSMC production lines in the South Taiwan Science Park received certain gases from suppliers that are believed to be contaminated. These were quickly replaced with other gas supplies,” the company told Nikkei, adding that the contamination was discovered on Thursday night. The ongoing chip drought has already disrupted the global supply chain markets. From automobiles to GPUs, many markets are struggling to keep up with the demands. The latest mishap in the Taiwanese company underlines the perils of isolated manufacturing hubs. PM Modi Bets Big On AI To develop core-concepts of students, SAFAL will introduce application-based & High Order Thinking Skills based assessment in line with #NEP2020 #TransformingEducation To know more: https:\/\/t.co\/LNSDhFs01f pic.twitter.com\/T3J7FPcI18— Ministry of Education (@EduMinOfIndia) July 29, 2021 On Thursday, Indian Prime Minister Narendra Modi launched key initiatives marking the first anniversary of National Education Policy (NEP 2020). The live webcast featured the Prime Minister along with members from Ministry of Education, NCC-New Delhi, IIT Kharagpur, IIT Madras, UGC, AICTE etc. The major announcements include the AI for All initiative by the central board of secondary education(CBSE) in collaboration with Intel, SAFAL assessment framework etc. AI For All is a four-hour, self-paced learning program that demystifies AI in an inclusive manner. SAFAL is an assessment proposed by NEP 2020 to assess students at school level. The assessment system, according to CBSE, is designed to promote development of students and suggests a shift from testing rote memorisation to competency-based learning. SAFAL will have an annual school examination for all students in Grades 3, 5, and 8, which will focus on testing core concepts, application of knowledge and higher order thinking skills. Read the full story here. Amazon Sets A New Record We are not talking about Blue Origin or the quarterly earnings. This is a record that Amazon is not very proud of. According to reports, Amazon has been fined a record $887 million, by a European Union(EU) privacy regulator for violations related to its advertising. This is the largest-ever fine under the EU’s data-protection law. Big tech and their big fines have become a common sighting in the last couple of years. This year especially, the crackdown has been more palpable. Both the US and the UK have been levying heavy fines and proposing regulations that end the “monopolisation” of the markets. Duolingo Talks Numbers On Wednesday, language learning app Duolingo made a grand debut on Nasdaq with its valuation touching $6.5 billion. Duolingo began as a computer science project for co-founders Severin Hacker and Luis von Ahn aimed at teaching people foreign languages while trying to translate the entire internet. The app has 40 million monthly active users and more than 500 million downloads. In 2020 alone, the company raked in 51% of revenue from Apple’s App Store and 19 % from the Google Play Store, and was the top-grossing app in the education category on both the platforms.","excerpt":"Looks like Elon Musk’s Neuralink has finally convinced a few biggies in the tech industry. On Thursday, Neuralink announced a Series C funding round of $205 million, which featured Google Ventures and OpenAI’s founder Sam Altman among others that included Fred Ehrsam co-founder of Coinbase and Ken Howery co-founder of PayPal.  Neuralink was founded by […]","categories":["AI News"],"tags":["data science and manufacturing","neuralink"],"author_name":"Ram Sagar","publish_date":"2021-08-01T10:00:00","publication_year":"2021","word_count":859,"keywords":["Go","artificial intelligence","data science and manufacturing","OpenAI","AI","ETL","neuralink","Git","RAG","Aim","GAN","R"],"extracted_tech_keywords":["AI","artificial intelligence","OpenAI","Aim","RAG","R","Go","Git","ETL","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/elon-musk-neuralink-funding-modi-india-ai-for-all\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10040008,"title":"7 Free Resources To Learn Explainable AI","content":"Explainable AI (XAI) is key to establishing trust among users and fighting the black-box nature of machine learning models. In general, XAI enhances accountability and reliability in machine learning models. For a long time, tech giants like Google, IBM and others have poured resources on explainable AI to explain the decision-making process of such models. Below are the top free resources to understand Explainable AI (XAI) in detail. (The list is in no particular order) 1| Explainable Machine Learning with LIME and H2O in R About: Explainable Machine Learning with LIME and H2O in R is a hands-on, guided introduction to explainable machine learning. The topics covered include the introduction and project overview, importing libraries, preprocessing data using Recipes, running autoML, leaderboard exploration, and model performance exploration, etc. By the end of this project, you will be able to use the H2O and LIME packages in R for automatic and interpretable machine learning. You will also learn how to build classification models quickly with H2O AutoML and interpret model predictions using LIME. Know more here. 2| An Introduction to Explainable AI, and Why We Need it About: This is an online tutorial. You will be exposed to a brief introduction to explainable AI, how it works and its importance. The blog will help you understand the Reversed Time Attention Model (RETAIN) model, Local Interpretable Model-Agnostic Explanations (LIME), how explainable AI works as one generates newer and more innovative applications for neural networks. Author Patrick Ferris explainS all of these through instances such as the one-pixel attack. Know more here. 3| Getting a Window into your Black Box Model About: In this tutorial, you will learn how to make sense of parts of a complex black-box model. The tutorial has two main goals. The first is to show how to build the windows or a local linear surrogate model based on a complex global model. The second goal is to explain reason codes, which helps in understanding the factors driving a prediction. Know more here. 4| Explainable AI: Scene Classification and GradCam Visualization About: This is a 2-hour long hands-on project where you will learn to train machine learning and deep learning models to predict the type of scenery in images. You will also understand the theory behind the deep neural networks, convolutional neural networks (CNNs) and residual nets. You will learn how to build a deep learning model based on CNNs and Residual blocks with the help of Keras with Tensorflow 2.0 as a backend and more. Know more here. 5| Explaining Quantitative Measures of Fairness About: This is a hands-on article that connects explainable AI methods with fairness measures and shows how modern explainability methods can enhance the usefulness of quantitative fairness metrics. You will also learn how to decompose measures of fairness and allocate responsibility for any observed disparity among each of the model’s input features. The tutorial will help you understand not just how to choose the “correct” measure of model fairness but rather about explaining whichever metric you have chosen. Know more here. 6| Interpretable Machine Learning Applications: Part 1 & 2 About: This is a project-based course for beginners to create interpretable machine learning applications on classification regression models, decision tree and random forest classifiers. In the first part, you will learn how to explain such prediction models by extracting the most important features and their values. In the second part, you will learn how to develop interpretable machine learning applications explaining individual predictions rather than the behaviour of the prediction model as a whole. For part 1, click here. For part 2, click here. 7| Responsible Machine Learning with Python About: This is a series of notebooks that introduce several approaches that increase transparency, accountability, and trustworthiness in ML models. The notebooks highlight techniques such as monotonic XGBoost models, partial dependence, individual conditional expectation plots, and Shapley explanations, decision tree surrogates, reason codes, and ensembles of explanations, LIME, etc. Know more here.","excerpt":"Explainable AI (XAI) is key to establishing trust among users and fighting the black-box nature of machine learning models. In general, XAI enhances accountability and reliability in machine learning models. For a long time, tech giants like Google, IBM and others have poured resources on explainable AI to explain the decision-making process of such models.  […]","categories":["AI Trends"],"tags":["explainability","explainability in AI","explainability of machine learning models","Explainable AI","XAI"],"author_name":"Ambika Choudhury","publish_date":"2021-05-13T11:00:00","publication_year":"2021","word_count":658,"keywords":["machine learning","explainability of machine learning models","explainability","AI","neural network","Keras","ML","explainability in AI","XAI","Python","deep learning","XGBoost","Explainable AI","TensorFlow","xAI"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","neural network","xAI","TensorFlow","Keras","XGBoost","Python"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/7-free-resources-to-learn-explainable-ai\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10139198,"title":"Databricks Introduces Scalable Batch Inference Model Serving","content":"Databricks, the data and AI company, has announced a new feature aimed at enhancing the efficiency of large language model (LLM) inference with its Mosaic AI Model Serving. The company says this innovation allows for simple, fast, and scalable batch processing of LLMs, making it easier for organisations to deploy these models in production environments to analyse unstructured data. The new model supports batch inference, allowing users to process multiple requests simultaneously, rather than one at a time. Databricks claims it enhances throughput and reduces latency, which is vital for real-time applications. Designed for ease of use, it provides a straightforward interface for users to quickly set up and manage LLM inference tasks without extensive coding. Mosaic AI Model Serving efficiently scales with demand, enabling organisations to dynamically adjust resources based on workload for optimal performance during peak times. This feature integrates with the Databricks platform, using existing data lakes and collaborative notebooks to enhance model training and deployment workflows. “No more exporting data as CSV files to unmanaged locations—now you can run batch inference directly within your workflows, with full governance through Unity Catalog,” the company posted on its official blog. This development positions Databricks as a leader in the LLM space, addressing the growing demand for efficient AI solutions across industries. Databricks-AWS Partnership Recently, Amazon and Databricks struck a five-year deal to focus on using Amazon’s Trainium AI chips that could cut costs for businesses seeking to build their GenAI apps. Databricks acquired AI startup MosaicML last year in a $1.3 billion deal and is expanding its services to democratise AI and position its Lakehouse as the top platform for GenAI and LLMs. The company raised $37 million and offers technology up to 15 times cheaper than competitors, serving clients like AI2, Replit, and Hippocratic AI. It claims that its  MPT-30B LLM, a 30-billion parameter model, is superior in quality and more cost-effective for local deployment than GPT-3.","excerpt":"Introducing Simple, Fast, and Scalable Batch LLM Inference on Mosaic AI Model Serving","categories":["AI News"],"tags":["Databricks"],"author_name":"Tarunya S","publish_date":"2024-10-23T16:49:16","publication_year":"2024","word_count":320,"keywords":["Go","GenAI","AWS","AI","ML","Scala","Aim","R","data lake","Databricks"],"extracted_tech_keywords":["AI","ML","GenAI","Aim","AWS","Databricks","R","Go","Scala","data lake"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/databricks-introduces-scalable-batch-inference-model-serving\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10086905,"title":"Google Live in Paris: LaMDA-Powered Search Features Unveiled","content":"At Google’s ‘Live in Paris‘ event, the tech giant revealed a series of AI innovations. Google has incorporated AI into everything from maps to museums to “reimagining how people search for, explore and interact with information”. Let us take a look at the key highlights of the event- Google Search & Lens Google adds its LaMDA-based AI chatbot Bard to search. Now you can get a summary of answers to your queries, similar to ChatGPT. “Search is our biggest moonshot”, said Prabhakar Raghavan, senior vice president at Google. Bard has opened up to  ‘trusted testers’ for feedback before being expanded to the public. Google Lens reached a record of enlisting over 10 billion times every month. In the coming months, you can use Google Lens to search what you see in photos or videos across websites and apps on your phone. The multi-search feature is now globally available in 70 languages. Currently being rolled out for image search results, you can use it to find real-world objects in different colours. Translate Over 1 billion people use Google’s translate feature. With a new ‘Zero Shot Machine Translation’ method, over 24 new languages are added. Google also said many Ukrainians seeking refuge had used Translate to fathom new places. By utilising AI and Google Lens to organically translate words for common household objects, the firm claims to be actively working to preserve endangered languages. Maps Google introduced several new AI features to Maps. Maps with the ‘Immersive View’, are rolling out now in over 1000 cities like LA, NY, and London. With its improved “Search with Live View” feature, which blends AI and AR, you can use your phone’s camera to visually locate objects in your area, such as restaurants, ATMs, and transit hubs. The ‘Indoor Live View’ is now being used for airports, shopping malls, and train stations. It uses AR arrows to indicate important spots like baggage claim, elevators etc, in limited cities right now, but with its “largest expansion”, it will be available in more than 1000 cities “in the coming months”. Another new AI feature is added for electronic vehicles whereby Maps can instantly filter charging stops and give you the “best charging stop” and also “very fast charging” stops. Read more: Google, the Tech-Savvy People Person Arts and Culture Google unveiled a number of new AI-powered arts and culture tools, including one that lets you look up well-known artwork by thousands of artists and showcase its deep details. Blob Opera also made a special entry. “Although we are 25 years into the search, I dare say that our story has just begun”, said Prabhakar reiterating the company’s motto towards “Responsible AI”. The company is also planning to hire more developers to advance ‘Generative AI’ in the coming months. No mention of Anthropic was made. Although the event did not have any huge announcements, and it mostly seemed like a recap, viewers were quick to spot an error in Bard’s results during the live demo. During the demo, Google tweeted a brief GIF clip showing Bard in action and called the Bard a “launchpad for curiosity” that would help make difficult subjects simple. Prompt: “What recent findings from the James Webb Space Telescope (JWST) can I share with my 9-year-old? Bard offered several responses, one of which claims that the JWST was used to capture the first images of exoplanets or worlds outside the solar system. That is incorrect. According to NASA, the Very Large Telescope (VLT) of the European Southern Observatory captured the first images of exoplanets in 2004. Microsoft held a close door conference last day where it released an upgraded AI-powered Bing search and Edge web browser, based on a new LLM but seeking findings from OpenAI’s GPT 3.5. Read more: The Battle for AI Supremacy Begins","excerpt":"At the Google Live in Paris event, it launched a slew of new updates in Search, Maps, and Lens, among others.","categories":["AI News"],"tags":["AI latest","Bard","Google","Google Bard","Google Lens","Google Map","Google Search"],"author_name":"Shritama Saha","publish_date":"2023-02-08T20:54:13","publication_year":"2023","word_count":632,"keywords":["Anthropic","ChatGPT","Go","Google Lens","OpenAI","AI","Google Bard","Google Search","Bard","Google Map","RAG","Aim","AI latest","generative AI","Google","Rust","R"],"extracted_tech_keywords":["AI","generative AI","ChatGPT","OpenAI","Anthropic","Aim","RAG","R","Go","Rust"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/google-adds-lamda-powered-bard-to-search-unveils-new-ar-features\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10017314,"title":"AI Helps Solve Schrödinger’s Equation — What Does This Mean For The Future?","content":"Scientists at the Freie Universität Berlin have come up with an AI-based solution for calculating the ground state of the Schrödinger equation in quantum chemistry. The Schrödinger’s equation is primarily used to predict the chemical and physical properties of a molecule based on the arrangement of its atoms. The equation helps determine where the electrons and nuclei of a molecule are and under a given set of conditions what their energies are. The equation has the same central importance as Newton’s law motion, which can predict an object’s position at a particular moment, but in quantum mechanics – that is in atoms or subatomic particles. The article describes how the neural network developed by the scientists at the Freie Universität Berlin brings more accuracy in solving the Schrödinger’s equation and what does this mean for the future. AI Brings More Accuracy To The Equation In principle, the Schrödinger’s equation can be solved to predict the exact location of atoms or subatomic particles in a molecule, but in practice, this is extremely difficult since it involves a lot of approximation. Central to the equation is a mathematical object, a wave function that specifies electrons’ behaviour in a molecule. But the high dimensionality of the wave function makes it extremely difficult to find out how electrons affect each other. Thus the most you get from the mathematical representations is a probabilistic account of it and not exact answers. This limits the accuracy with which we can find properties of a molecule like the configuration, conformation, size, and shape, which can help define the wave function. The process becomes so complex that it becomes impossible to implement the equation beyond a few atoms. Replacing the mathematical building blocks, the scientists at Freie Universität Berlin came up with a deep neural network that is capable of learning the complex patterns of how electrons are located around the nuclei. The scientists developed a Deep Neural Networks (DNN) model, PauliNet, that has several advantages over conventional methods to study quantum systems like the Quantum Monte Carlo or other classical quantum chemistry methods. The DNN model developed by these scientists is highly flexible and allows for a variational approach that can aid accurate calculation of electronic properties beyond the electronic energies. Secondly, it also helps the easy calculation of many-body and more-complex correlation with fewer determinants, reducing the need for higher computation power. The model mainly helped solve a major tradeoff issue between accuracy and computational cost, often faced while solving the Schrodinger equation. The model can also calculate the local energy of heavy nuclei like heavy metals without using pseudo-potentials or approximations. Lastly, the model developed in the study has anti-symmetry functions and other principles crucial to electronic wave functions integrated into the DNN model, rather than let the model learn. Thus, building fundamental physics in the model has helped it make meaningful and accurate predictions. AI Helps Science In recent years, artificial intelligence has helped solve many scientific problems that otherwise seemed impossible using traditional methods. AI has become instrumental in anticipating the results of experiments or simulations of quantum systems, especially due to its science’s complex nature. In 2018, reinforcement learning was used to design new quantum experiments in automated laboratories autonomously. Recent efforts by the University of Warwick and another IBM and DeepMind have also tried to solve the Schrödinger’s equation. However, PauliNet, with its greater accuracy of solving the equation now, presents us with a potential to use it in many real-life applications. Understanding molecules’ composition can help accelerate drug-discovery, which earlier was difficult due to the approximations to understand its properties. Similarly, it could also help discover several other elements or metamaterials like new catalysts, industrial chemical applications, new pesticides, among others. It can be used in characterising molecules that are synthesised in laboratories. Several academic and commercial software use Schrödinger’s equation at the core but are based on applications. The accuracy of this software will improve. Quantum computing in itself is based on quantum phenomena of superposition and is made up of qubits that take advantage of the principle. Quantum computing performance will improve as qubits will be able to be measured faster. Wrapping Up While the current study has come up with a faster, cheaper, and accurate solution, there are many challenges to overcome before it is industry-ready. However, once it is ready, the world will witness many applications as a result of greater accuracy in solving Schrödinger’s equation.","excerpt":"Scientists at the Freie Universität Berlin have come up with an AI-based solution for calculating the ground state of the Schrödinger equation in quantum chemistry. The Schrödinger’s equation is primarily used to predict the chemical and physical properties of a molecule based on the arrangement of its atoms. The equation helps determine where the electrons […]","categories":["AI Features"],"tags":["AI What it Does","Deep Neural Networks","how does ai work","how does artificial intelligence work","quantum mechanics","quantum physics"],"author_name":"Kashyap Raibagi","publish_date":"2021-01-07T17:00:00","publication_year":"2021","word_count":740,"keywords":["artificial intelligence","programming_languages:R","AI","neural network","R","how does ai work","quantum physics","how does artificial intelligence work","Deep Neural Networks","AI What it Does","quantum mechanics","emerging_tech:quantum computing"],"extracted_tech_keywords":["AI","artificial intelligence","neural network","R","programming_languages:R","emerging_tech:quantum computing"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/ai-helps-solve-schrodingers-equation-what-does-this-mean-for-the-future\/","complexity_score":3,"technical_depth":6,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10082351,"title":"1 in 5 Indians is Prone to Data Breaches","content":"India remains synonymous with data breaches. Recently, cybersecurity’s data monitoring firm Surfshark revealed that 1 out of 5 Indians had been affected by data breaches (19 for every 100 people) from 2004 to Q4 FY22. India is the fourth country to have the highest number of data point leaks. The study revealed that 738k Indian accounts have been breached till now. Last month, the internal systems of All India Institute of Medical Sciences (AIIMS) succumbed to one of the worst ransomware attacks. The institute has now switched many of its operations from digital to manual. Even the ICMR website faced about 6,000 hacking attempts. The below video highlights the implications of such breaches: Recently, India’s government platform for connecting with its overseas population, Global Pravasi Rishta Portal, leaked sensitive information like names, phones, passport numbers and email addresses. The list goes on and on. Since 2004, India has had over one billion data leaks, with an average of 4 data leaks per hacked account. Since 2004, 51.8 billion data points—of which 15.5 billion were email addresses—have been exposed. Each email address is exposed to 2.3 extra data points on average. Since the beginning of 2020, 2.5 billion accounts have been affected, accounting for 16.2% of all worldwide breaches that have occurred since 2004. Every year, 11 out of 100 people worldwide are affected. Breach rates, however, have decreased since the fourth quarter of 2020. As a result, the number of Indian data breaches decreased from 13,688,077 to 738,361. One of India’s most leaked data points is passwords. Statistically, since 2004, for every ten Indian online accounts, five passwords have been leaked. India ranks second in Asia and sixth globally based on the breach count since 2004 (265.4M compromised internet users). Accounts from Vietnam, Iraq, and America have been compromised with the most data points. Two nations account for 30.4% of all global breaches: the US and Russia. The majority of hacked accounts worldwide are located in the United States. It represents 15.9% of all compromised email addresses worldwide. Russian email accounts make up 14.5% of all global breaches. The country with the most exposed accounts per 100 people is Russia. According to statistics, each Russian account’s email address was breached about 15 times. This figure is 7.8 times greater than the global average (199 breached accounts per 100 people).","excerpt":"India ranks second in Asia and sixth globally based on the breach count since 2004.","categories":["AI News"],"tags":["China","data breach","data leak","Data Privacy","Russia","USA"],"author_name":"Shritama Saha","publish_date":"2022-12-14T16:10:03","publication_year":"2022","word_count":389,"keywords":["Go","programming_languages:R","AI","data leak","Russia","programming_languages:Go","Git","data breach","RAG","USA","Data Privacy","R","China"],"extracted_tech_keywords":["AI","RAG","R","Go","Git","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/1-in-5-indians-is-prone-to-data-breaches\/","complexity_score":3,"technical_depth":7,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10021625,"title":"China Vs US: Battle Lines Are Drawn In The Fight For AI Upperhand","content":"Stanford University’s research showed the total number of AI research publications increased by 34.5 percent last year: almost double from the 19.6 percent year-on-year growth in 2019. China has surpassed the US in the total number of AI journals published in the previous year. The report, AI index 2021, deals with AI metrics, including its impact on jobs, economic growth, investments in the field, journals published, diversity, and more. The report has brought the standoff between China and the US to the fore again, albeit on the AI front this time. In the last few years, China has become a force to reckon with in AI. Key Highlights Of The Report: China surpassed the US in terms of the number of AI research papers published in 2020.US scientists’ papers were more frequently accepted for prestigious conferences and were more cited by other researchers globally, including Chinese.The US had more international students than China for PhD courses in AI. 64.3% of AI PhDs in North America were international students, a 4.3% increase than the year before. China surpassed the US in significant scholarly work. Chinese-affiliated scholars were cited in more peer-reviewed journals than any other country, indicating China’s AI research has come up in quality and quantity. China Taking A Lead In AI R&D The three-step program released by China in 2017 clearly defined its goal to become an AI leader worth $150 billion by 2030. The Chinese government has also made investments in AI-focused technology research parks, apart from releasing Beijing AI Principles. While the US has led the efforts and investment in AI, China quickly ramped up contributions in fundamental AI research. It has worked around computer vision and other AI implementations to create a massive market for successful AI startups such as SenseTime, Yitu Technology, and CloudWalk. Key Areas Of Development Computer vision and facial recognition: China has strengthened its capabilities in computer vision, facial recognition, emotion recognition and object detection. It has also made progress in natural language processing and 5G technology. Drug discovery: China is also leading the way in AI-based drug discovery efforts. In 2019, Tencent launched an AI-based drug discovery platform to reduce the time spent on the R&D of new drugs. Other measures include a $200 million AI drug discovery partnership between Insilico and China’s CTFH. Quantum computing: Chinese researchers claimed to have achieved quantum supremacy with the capability of performing calculations 100 trillion times faster than the world’s most advanced supercomputer, beating Google to this milestone. Military AI: China’s spending has also trumped the US in military AI. Experts believe if the US does not become aggressive, the country’s democracy will be doomed. Advantage China Favourable policies: China enjoys fair policies and regulations backed by a government willing to invest and deploy innovative technologies at scale. Large population: The biggest advantage China has is its outsized population, leading to an enormous potential workforce. Compared to the US and Europe, the labour cost is low, which can favour AI research. Concerted efforts from universities and startups: Both industry and academia are working to propel AI research in computer vision, NLP and more. China is working to bring researchers back to the country by offering good perks and benefits. Challenges Despite the growth, the country faces challenges such as expanding into the overseas market due to its aggressive foreign policies. The US-China trade war has also affected the AI investments in China. Another challenge is the lack of technological tools. For instance, the US leads in open source platforms such as TensorFlow and Caffe. However, there are a few open-source platforms and tools still available in China, such as PaddlePaddle. Disregard for ethical issues is a big challenge not just in China but also in other countries. China faced criticism for its use of facial recognition technology to monitor Uighurs in Xinjiang. China In Slipstream For long, the US has been the leader in artificial intelligence, quantum computing, blockchain, etc. Last year, the US toppled China to take the top spot as the global AI venture capital investment hub. According to ABI Research, the US commanded the most AI investment in 2018 globally at 52.3%. AI was also named the second-highest R&D priority in 2020 in the US. Meanwhile, the National Security Commission on Artificial Intelligence (NSCAI), created in 2018, warned that the US is underprepared for the AI race and China could potentially replace the US as the AI superpower. “America is not prepared to defend or compete in the AI era,” Eric Schmidt, who chaired NSCAI, had said. The NSCAI has also called for significant changes in the government’s approach to innovation and technology, training AI talent, building competitiveness and more. Having said that, while China is making meaningful investments in AI, the US has the world’s largest investment market in privately-held AI companies. It might just have to make AI the top priority, just like the Chinese government, to consolidate its leadership position.","excerpt":"Stanford University’s research showed the total number of AI research publications increased by 34.5 percent last year: almost double from the 19.6 percent year-on-year growth in 2019. China has surpassed the US in the total number of AI journals published in the previous year. The report, AI index 2021, deals with AI metrics, including its […]","categories":["Deep Tech"],"tags":[],"author_name":"Srishti Deoras","publish_date":"2021-03-08T13:00:00","publication_year":"2021","word_count":822,"keywords":["Go","API","artificial intelligence","AI","computer vision","NLP","Aim","object detection","TensorFlow","R"],"extracted_tech_keywords":["AI","artificial intelligence","NLP","computer vision","Aim","TensorFlow","object detection","R","Go","API"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/china-vs-us-battle-lines-are-drawn-in-the-fight-for-ai-upperhand\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10084976,"title":"Building CyberSecurity Barriers for Enterprises with Zero Trust","content":"The growing danger of cyber attacks and data breaches is a two-fold risk: not only have attacks multiplied in numbers, hackers are using more sophisticated technology to penetrate security walls. Cybersecurity has become top-of-the-order concern for most companies in this age with a simultaneous expansion of the cybersecurity sector. To keep up with changing times, InstaSafe Technologies can deploy their single-cloud, easily scalable cybersecurity software in a matter of minutes. In an interaction with the co-founder and CEO Sandip Panda, Analytics India Magazine discussed how InstaSafe is coping with advanced threats and how Zero Trust may be the answer. AIM: What is Zero Trust security? How does it help you provide fool-proof cyber security to companies? Sandip: In the wake of the pandemic, we have witnessed a swift shift to remote working to make it convenient for employees to work from anywhere in the world. For businesses with remote workers, technologies like VPN and firewalls have been the primary way to enable connectivity and enforce security. However, traditional VPNs don’t meet the needs of modern organisations and security has become a major concern. Over time, the technology began to counter the previously designed VPNs as they expose entire networks to threats like malware, DDoS attacks, and spoofing attacks. Once an attacker has breached the network through a compromised device, the entire network can be tripped down. However, since 2010, we have seen a massive rise in the adoption of the zero trust model after John Kindervag, a Principal Analyst from Forrester presented his research on Zero Trust Network Architecture. This approach addresses the challenges in legacy VPN solutions by limiting lateral movement and denying access requests. This has been made possible with the help of SDP (software-defined perimeter), a network architecture that implements zero-trust principles to provide more secure remote access than VPNs. Zero Trust is a cloud-delivered solution that is easy to deploy and scale. Businesses can adopt a granular access control policy and frame a security framework for BYOD policy with a Zero Trust solution. AIM: Where does India’s cybersecurity sector stand and what does it need to do to mitigate big cyber attacks? Sandip: According to Computer Weekly Data, revenues from cybersecurity products and services in India reached USD 156.3 billion in 2022. Furthermore, since the adoption of 5G and the rise of new-age startups, the nation is expected to witness a trend of rapid digitisation and regulatory attention on data and privacy. This would indicate that there’s a great opportunity for the country’s cybersecurity sector. While India is making quick progress towards its digital goals, cybersecurity is still a problem because it relies on linked networks and systems. Currently, this is being tackled with the help of the implementation of the Zero Trust Application Security model to prevent unauthorised user access, enhance network visibility, simplify access control policies, and improve the overall effectiveness of one’s network security. While it’s crucial to implement Zero Trust Security and educate and empower employees to implement secure application access solutions, it is also important to understand that India is often one of the most vulnerable nations online, therefore securing vital assets depends on its cybersecurity resiliency. Currently, a multifaceted defence is in the works to meet this threat. AIM: How would you describe InstaSafe’s journey? Sandip: Founded in the year 2012, InstaSafe’s vision is to provide cloud-based security solutions to businesses. Our startup has identified a gap in the secure remote access which was largely served by legacy hardware-based VPN solutions. With the cloud transformation, InstaSafe has tapped into the market with a more simplified and secure solution through the Zero Trust model. Additionally, in 2021, we received the ‘Security Product Company of the Year’ from the Data Security Council of India (DSCI). AIM: How many companies use InstaSafe and how many of these companies are from India? Sandip: Today, they have over 150 customers from the enterprise sector to the SME segment spread across IT\/ITES, retail, logistics, e-commerce, and financial services. Currently, InstaSafe is securing over five lakh endpoints with our customers being distributed across the US, EMEA, and Asia Pacific regions. Their growth has helped them get recognised by companies like Deloitte, Gartner, HewlettPackard Enterprise. AIM: How has the market evolved for InstaSafe in India? Sandip: The Global Zero Trust Security market size is expected to reach USD 54.6 billion by 2026, rising at a market growth of 18.8% CAGR during the forecast period. As India ranks third globally in terms of cyber breaches, InstaSafe has changed their pre-existing security model. The hybrid work environment is the new reality and with the increase in cloud adoption in the country, the adoption of Zero Trust is expected to grow significantly. AIM: What are the new trends and developments we are going to witness in the cybersecurity space in the coming years? Sandip: Data, be it personal or professional, has always been a potential risk around the world and safeguarding it has become increasingly important. Any tiny defect or flaw in your computer’s software or browser could allow hackers access to your sensitive data. Due to the lack of data protection along with the cultural changes in working environments, we have been witness to a huge spike of data breaches from sources that are  common in the workplace. To counter these, AI\/ML has been adopted in all segments and this has brought a sea change in the area of cybersecurity. However, on the flipside, AI is also used to orchestrate clever malware attacks that cross even the most advanced data security mechanisms. Additionally, today’s vehicles are also loaded with automated software that enables smooth connectivity with airbags, cruise control, door locks, and advanced driver aid systems. These vehicles use bluetooth and WiFi to connect which exposes them to potential security flaws or hacker threats. With rapid developments happening continuously as we step into 2023, it is important for the sector to monitor how phishing attacks, ransomware attacks, insider threats and vulnerabilities are changing. Keeping a constant and careful eye over these trends will help the industry develop in the right direction. As mobile banking malware has been one of the biggest risks with individuals’ personal data, pictures, emails, official emails, and official assets always being under threat, malware that affects smartphones should be more in our focus.","excerpt":"The hybrid work environment is the new reality and with the increase in cloud adoption in the country, the adoption of Zero Trust is expected to grow significantly.","categories":["Deep Tech"],"tags":["Cybersecurity India"],"author_name":"Poulomi Chatterjee","publish_date":"2023-01-13T13:00:00","publication_year":"2023","word_count":1047,"keywords":["Go","AWS","AI","ML","Scala","Git","Aim","Cybersecurity India","analytics","Rust","R"],"extracted_tech_keywords":["AI","ML","analytics","Aim","AWS","R","Go","Rust","Scala","Git"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/building-cybersecurity-barriers-for-enterprises-with-zero-trust\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10125867,"title":"Why Public Sector Organisations Globally Trust AWS","content":"Amazon Web Services (AWS) launched its AWS Public Sector Generative AI Impact Initiative last week during the AWS Summit in Washington, DC. As part of the two-year initiative, public sector companies will receive support in terms of cloud infrastructure and generative AI services to help them upgrade existing and new missions. “Security is our top priority. It’s the foundation of how we support customers worldwide. And public sector customers trust AWS. They trust us to maintain the integrity of their most sensitive assets and their most sensitive missions. We prioritise governance while exploring and incorporating emerging technologies, like generative AI,” said Dave Levy, the vice president of the world public sector at AWS. Levy told AIM that the inclination of public sector organisations to trust AWS is because of AWS’ focus on trust itself. “Security is critical internally and so, that’s also true of our customers. It gives them a level of comfort with us, that we are going to be transparent, and that their data and their enterprise is very important to us,” he said. But this isn’t the only reason why AWS has gained the trust of several global public sector organisations. An Early Player In the early 2010s, AWS was one of several pioneers that began offering their services to government agencies, starting with the launch of AWS GovCloud, which was specifically tailored for government organisations, in line with regulations. Subsequently, AWS became one of the few companies that continuously focused on offering cloud services to the public sector, which was especially important as the US government started switching to cloud computing in the mid-2010s. This led to receiving contracts from multiple government agencies, including the CIA, which secured their services for a massive $600 million contract way back in 2013. In the present day, AWS is currently working on establishing the AWS European Sovereign Cloud by 2025, with a whopping €7.9 billion investment. AWS has been at the forefront of allowing public sector organisations to actively shift to cloud computing from legacy on-premise physical servers, allowing for overall better storage solutions. With a lead in cloud computing within the US, AWS quickly seemed to become the go-to cloud provider for several countries, including India where public sector organisations leverage AWS’ capabilities in the fields of agriculture, healthcare and education. So far, AWS has partnered with organisations and ministries like MeitY, the health and family welfare ministry, and the Telangana and Madhya Pradesh governments, as well as several universities and the National Skill Development Corporation. Highlighting this, Levy said that India remains a core market for the company. “India is very important to us. We’ve made a lot of investments over the years. We’ve been here for a long, long time, and it’s also a very exciting market for us. We see a lot of opportunity for AWS and for our team working with the public sector,” he said. In line with this, AWS very recently launched Amazon Bedrock for the Mumbai region during the AWS Summit held in Bengaluru in May this year. This inclusion of Mumbai means that customers from the region can build applications using Amazon’s in-house foundation models, as well as models from Anthropic, Cohere and Meta. “In fact, Bedrock landed in Mumbai before it landed here [Washington]. So India is a very important place for us,” he said. In AWS, We Trust Despite its history, however, Levy is right in saying that AWS is able to abide by the strict privacy and security requirements set out by public sector organisations. AWS and its competitors in Microsoft and Google have comprehensive write-ups on the policies employed in collaborating with the public sector. However, its early lead in the market, as well as the fact that it has continuously held a majority of the market (31% as of May 2024) when it comes to providing cloud infrastructure services, means that AWS is often the first choice when public sector organisations look to shift cloud computing. Now, with their recent partnership with Accenture and Anthropic, wherein AWS will provide cloud infra services to Accenture’s customers, this market share is likely to increase, overtaking players like Microsoft. Interestingly, this deal also comes at a time when it seems like Microsoft had been gaining headway in catching up with AWS in terms of cloud computing, as its share had increased this year, while AWS’ had decreased by one per cent. However, its recent partnership with Oracle, integrating Oracle Cloud Infrastructure, implies that there might be some gaps in Microsoft Azure’s functioning that only OCI can solve. Meanwhile, AWS Outposts allows extended capabilities to provide on-premises private cloud services for their customers, adding another layer of security to an already secure product. Nevertheless, it seems that Microsoft has a long way to go, with AWS remaining the first choice for many public sector organisations, thanks to its current lead in the market, basically proving that customers, public and otherwise, are more likely to trust AWS with their data. As Levy said, “We’re convinced that responsibility drives trust, and trust drives adoption, and adoption drives innovation.”","excerpt":"“We’re convinced that responsibility drives trust, and trust drives adoption, and adoption drives innovation,” says Dave Levy, vice president of the world public sector at AWS.","categories":["IT Services"],"tags":[],"author_name":"Donna Eva","publish_date":"2024-07-05T11:00:00","publication_year":"2024","word_count":847,"keywords":["Anthropic","AWS","AI","cloud computing","R","RAG","Aim","generative AI","foundation models","Azure"],"extracted_tech_keywords":["AI","generative AI","foundation models","Anthropic","Aim","RAG","cloud computing","AWS","Azure","R"],"url":"https:\/\/analyticsindiamag.com\/it-services\/why-public-sector-organisations-globally-trust-aws\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10166377,"title":"How Commonwealth Bank of Australia Built a GenAI Chatbot in Just 6 Weeks","content":"The Commonwealth Bank of Australia (CBA) is integrating generative AI across its operations to improve customer service, simplify processes, and strengthen security. Speaking at the The Rising 2025, India’s biggest summit on women and tech in AI, Nidhi Sinha, general manager, chief data analytics office (CDAO) at CBA India, revealed that the bank built its generative AI chatbot, CommBank, in less than six weeks. “A lot of organisations are experimenting with generative AI,” Sinha said, “But we are using it to lift and accelerate what we are currently doing.” She added that while the initial focus was improving internal operations, CBA has now extended these capabilities to customer-facing solutions. One such initiative was launched in November for business banking. Business customers, who often have complex queries about products and payments, previously had to navigate through 80 different FAQ documents or contact call centres, resulting in delays. “We have embedded the generative AI solution within the app,” Sinha explained. “It fetches data from all these documents, allowing customers to complete transactions without leaving the app.” The chatbot understands context and provides relevant answers, reducing the need for external searches or call centre support. “Best of all, this entire solution was developed in just six weeks, from infrastructure provisioning to deployment.” She added that CBA built this solution on AWS, one of its strategic partners. “The pace of change has never been this fast, yet it will never be this slow again,” Sinha remarked, quoting an internal mantra that reflects the rapid advancements in AI adoption. Notably, CommBank entered a five-year strategic collaboration with Amazon Web Services (AWS) earlier this year to continue as the bank’s preferred cloud provider. The bank currently has over 60 generative AI use cases, with many already live for both customers and internal users. These solutions drive efficiency and improve customer interactions. To support AI-driven transformation, CBA has established the GenAI Council, a leadership body that includes senior executives and the CEO and is focused on AI acceleration. “A combination of oversight and a combination of federation is really helping us accelerate at a very, very fast pace,” she further said. “Different teams run with their own use cases, while central oversight ensures alignment and scalability.” CBA’s investment in AI spans more than a decade. “We have been recognised as the number one bank in AI in the Asia-Pacific region for two consecutive years and globally as the best in responsible AI,” Sinha said. “We are also working with the Australian government to develop AI principles for the country.” She further shared that over 60% of the population in Australia uses CBA in some form, with one-third considering it their primary financial institution. The bank has prioritised responsible AI, integrating governance frameworks to ensure safe implementation. “For example, checks on groundedness prevent hallucinations, profanity filters maintain appropriate interactions, and jailbreaking safeguards ensure models are not misused. These controls are centrally managed and available for all teams.” CBA is also accelerating its data strategy, moving from on-premise to cloud. “Initially, this was planned for 18 months, but by leveraging AI, we are completing it in nine months,” Nidhi said. “By June, all our data will be on the cloud, providing practitioners with high-quality data to accelerate use cases.” Collaboration is key to CBA’s AI journey. “We have world-class partnerships with AWS, Microsoft, and Anthropic, giving us access to cutting-edge AI capabilities and top talent,” she said. “These partnerships are critical in our generative AI journey.” AI-Driven Financial Management One of CBA’s AI-driven solutions focuses on financial management. The bank uses predictive analytics to help customers manage their finances effectively. “We have a product where we use AI to predict future cash flows and nudge our customers at the right time,” Sinha said. This proactive approach allows customers to plan better and avoid financial stress. CBA also applies AI to disaster response. Given Australia’s exposure to natural disasters, the bank integrates external weather data with its customer information. “We use data and AI to identify customers and communities proactively who would be impacted by some calamity, and we reach out to them to provide our help,” she said. Personalised Customer Engagement CBA has been personalising customer interactions for over a decade. “In 2015, we launched something we call the Customer Engagement Engine,” Sinha shared. This AI-driven platform connects all customer interaction channels—mobile banking, branch visits, and call centres—to offer real-time recommendations. “When a customer comes into any of these channels, there is something called next best conversation (NBC) that is surfaced to them,” she explained. “It could be about an offer, a service, or even a simple happy birthday message.” She added that the system processes over 3.1 trillion data points and runs 2,000 adaptive models to ensure relevance. This AI-driven approach has also influenced customer engagement through CBA’s loyalty program, Yello. “What we have seen in the last five years is that customers engaged with this program log into our app, on average, 67 times a month—twice a day.” AI for Fraud Prevention CBA has made significant strides in fraud prevention using AI. “With AI evolving, threat actors and scammers also have access to these technologies, and fraud is increasing globally,” Sinha warned. She cited global statistics showing a $1 trillion loss due to fraud last year. However, CBA has successfully reduced scams by 70% in the past two years. “This is a big achievement, though we aim to do more,” she explained. The bank monitors 20 million transactions daily, detecting fraudulent activity in real time. “Within 10 milliseconds, an alert is sent to the customer, allowing them to take immediate action.” CBA has also introduced features like NameCheck and CallerCheck to prevent mistaken or fraudulent transactions. “This has actually helped us save $650 million per customer,” she shared. Preventing Payment Abuse One of the more unique AI applications at CBA involves preventing abuse through payment messaging. “We discovered that some individuals were misusing the payment description field to send abusive messages,” Sinha said. In response, CBA launched a profanity blocker in 2021. “At the time of payment, if abusive words are detected, the transaction is blocked immediately,” she explained. However, the challenge went beyond explicit words. “Even simple phrases like ‘I love you’ can be threatening in certain contexts,” she said. The bank has since developed AI models to detect harmful intent, ensuring payments are not used as a tool for harassment. Talent Development Another focus area is talent development. “People are the cornerstone of our AI initiatives,” Sinha said. “Five years ago, we set up CBA India, and today, we have 46% workforce diversity, with 41% representation in leadership roles…An inclusive workforce enables us to understand customers better and think beyond traditional banking.” CBA’s AI journey is not just about technology but also about fostering a culture of experimentation, innovation, and responsibility. “We are creating a model where people raise their vision to do more with AI,” Sinha concluded.","excerpt":"While the initial focus was improving internal operations, the bank has now extended these capabilities to customer-facing solutions.","categories":["Global Tech"],"tags":["Bank"],"author_name":"Siddharth Jindal","publish_date":"2025-03-20T13:30:00","publication_year":"2025","word_count":1153,"keywords":["Bank","Anthropic","GenAI","AWS","AI","RAG","Aim","generative AI","analytics","edge AI","predictive analytics"],"extracted_tech_keywords":["AI","analytics","generative AI","GenAI","Anthropic","Aim","edge AI","RAG","predictive analytics","AWS"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/how-commonwealth-bank-of-australia-built-a-genai-chatbot-in-just-6-weeks\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10048398,"title":"All You Need to Know about Data Fabric","content":"Gathering actionable insights and information from raw data is an essential aspect of today’s data-driven world. Data management & analysis has become a priority for organizations as data has become increasingly diverse, distributed and complex. Hence, analysts and data scientists need to leverage traditional techniques and switch to modern practices such as artificially intelligent systems. A new and rapidly growing concept known as “Data Fabric” has emerged in recent years to aid such challenges. What is Data Fabric? Data Fabric is an all-in-one integrated architectural layer, i.e., fabric, that connects data and analytical processes. Data Fabric uses existing metadata assets to support the design, deployment and proper utilization of data across all environments and platforms. The concept aims to accelerate the inference of insights from data through several different automated processes. It can help many use cases and provide real-time insights, in turn managing the data flow and curation from all data sources. A data fabric integrates processes such as data integration, analytics, and dashboarding all into one and serves as a management solution. It enables a consistent user experience and instant access to data for any member in an organization in real-time, allowing frictionless access in a distributed environment. Need for Data Fabric In a data-centric environment, data needs to be instantly accessible to users who need it while being present in a secure and efficient environment. The traditional data integration methods do not meet the demand of dynamically changing businesses, and therefore, automated and self-efficient systems have been in demand. The data management processes today are required to deliver a comprehensive view for customers and products. A data fabric is a potential solution to all this. The term “data fabric” can be described as a cloth spread wherever the organization’s users are. The user could be present anywhere in the world and still have access to data without any real-time constraints. Organizations today need to have their data transformed and processed as a lack of comprehensive data access can often fail to produce useful predictions and desired outcomes. In such conditions, a data fabric comes to the rescue. It leverages both human and machine capabilities. Key Components of Data Fabric A data fabric is a combination of several different layers. Here are some of the key components necessary for its proper implementation : A well-connected pool of metadata lays the foundation of a Data Fabric Design. It comprises services that allow the data fabric to identify and analyse every type of metadata.A data catalogue provides access to all the metadata types through a well-connected knowledge graph. It also graphically depicts the present metadata in an easy to understand manner and creates unique relationships between them.Analytics, when connected with knowledge graphs, help in activating the metadata. It enables the graphs to enrich their data with semantics, making it easier for data analysts and scientists.Including a set of standard data integration tools ensures integrated data delivery through multiple data delivery styles and helps curate the analysed knowledge graphs. The created data fabric should have a robust data compatibility backbone. The fabric should be compatible with various data delivery styles and not be limited to any. Support for different data types ensures its availability for all kinds of users. Impact of Data Fabric Data fabric is not just a mere combination of traditional and contemporary technologies but a concept that aims to ease the workloads for humans and machines. The design optimises and leverages data management techniques by helping automate repetitive tasks such as data discovery, data profiling and aligning the data schema to new data sources. It can also help correct faulty data integration tasks and analyse what’s wrong. Data fabric helps an organisation by understanding business demands while gaining an edge over its competitors. Through this technology, the actual potential and power of a hybrid multi-cloud experience can be harnessed to its fullest. Built on a rich set of data management capabilities, it ensures consistency across all integrated environments. It can reach anywhere, be it on-premises, edge and IoT devices, or different cloud types. Using Artificial intelligence with Data Fabric Data fabric being an operational layer, not only brings data altogether but also transforms and processes it. Making use of technologies such as artificial intelligence and machine learning can discover patterns and insights. Data fabric can help prepare the data to meet the needs of AI and ML with higher sustainability levels and at an automated pace. Through this, decision-makers can promptly gain better insights — discovering hidden facts that previously would have been overlooked help find more solutions for problems proactively. Risks and Benefits of Data Fabric Concern about data fabric is the threat to data security when data is transported from one point to another. Therefore, it becomes a mandatory requirement to embed the system with security firewalls and protocols that ensure safety from breaches. With several cases of cyberattacks hitting organisations, the security of data points becomes a necessary step towards safety. Data Fabric helps increase connectivity rapidly and is an agile model that can work with all operating and storage systems. No highly expensive investments in hardware or trained staff are required to control one. It provides better accessibility with the real-time flow of data and information. Summing up With data growing exponentially and problems around data multiplying, the use of a data fabric architecture can be a potential solution towards a sustainable data future. A data fabric not only maximises the value of data but also provides a better infrastructure to manage it. Furthermore, it improves and simplifies end-to-end performance not only within but also outside the organisational space.","excerpt":"Data Fabric uses existing metadata assets to support the design, deployment and proper utilization of data across all environments and platforms. The concept aims to accelerate the inference of insights from data through several different automated processes.","categories":["AI Trends"],"tags":["AI (Artificial Intelligence)","business data sources","Data Fabric","Data Management","Data Science","Data Scientist","how to measure twitter influence","load data python","Machine Learning"],"author_name":"Victor Dey","publish_date":"2021-09-15T15:00:00","publication_year":"2021","word_count":935,"keywords":["API","Data Fabric","R","business data sources","load data python","artificial intelligence","how to measure twitter influence","RAG","analytics","Data Science","machine learning","AI","ML","Machine Learning","Data Management","knowledge graphs","Aim","Data Scientist","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","analytics","Aim","RAG","knowledge graphs","R","API"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/all-you-need-to-know-about-data-fabric\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10161530,"title":"Sam Altman Genuinely Believes OpenAI Will Launch the First AGI","content":"Sam Altman, CEO of the AI giant OpenAI, expressed confidence in a podcast episode that the company will be the first to achieve AGI or artificial general intelligence. When host Adam Grant, an author and a psychologist, asked Altman about an unpopular opinion in AI, the latter replied, “That it’s [AI] not gonna be as big of a deal as people think, at least in the short term. Long term, everything changes.” “I kind of genuinely believe that we can launch the first AGI, and no one cares that much,” Altman added. He predicted that 20 minutes after AGI is achieved, everyone will think about what they will have for dinner that night. Moreover, Altman also believes that the company’s models are smarter than him in “almost every way”, but this doesn’t impact his life. Altman’s thesis points towards the fact that in the short term, AI models don’t matter much, and only the long-term effects will change the course of humanity.“You and I are living through this once-in-human history transition where humans go from being the smartest thing on planet earth to not being the smartest thing on planet earth.” This certainly opens up a lot of interesting takes on how humans may navigate an AI-driven future and economies. “Eventually, I think the whole economy transforms. We’ll find new things to do. I have no worry about that,” Altman added. So are we not to worry as well? ‘We Always Find New Jobs’ While Altman admitted that some jobs will disappear, he added that humans will always find new things to do. Several big names in the industry approach an AI-dominated future using a similar, optimistic thought process. Take Anthropic CEO Dario Amodei, for instance. In an essay, he said, “I think it is very likely a mistake to believe that tasks you undertake are meaningless simply because an AI could do them better.” He also said that people may begin to greatly enjoy activities that produce no economic value. He added that people will still be able to achieve a sense of accomplishment by spending years attempting some very difficult task without wanting to reap the economic benefits that come with it. “The fact that an AI somewhere could, in principle, do this task better and that this task is no longer an economically rewarded element of a global economy doesn’t seem to matter very much to me.” He further said that even in tasks where AI can do 100% of the things better, humans may possess an advantage if machines are economically inefficient. Amodei also believes that one area where humans are likely to maintain a relative or even an absolute advantage is their presence in the physical world. However, he admits that AI will eventually become effective and so cheap that the arguments no longer apply. “At that point, our current economic setup will no longer make sense, and there will be a need for a broader societal conversation about how the economy should be organised,” he added. Vinod Khosla, a veteran entrepreneur in Silicon Valley, believes that AI could create a world where a “small elite” will thrive and the rest will face economic problems. AI will create new jobs we cannot currently conceive of,” he said, adding that it will eliminate most professions that people pursue to support their needs and lifestyles. However, he is confident that smart interventions, such as income distribution, minimum living standards, and strategic legislators, can lead to solutions. “I believe these interventions are achievable because Western capitalism is achieved with the permission of democracy and its voters. If we correctly handle this phase shift, AI will generate more than enough wealth to go around, and everyone will be better off than in a world without it,” Khosla said. Sam Altman Loves UBI Interestingly, Altman has also explored the cure for the very problem he created. He is the biggest proponent of Universal Basic Income (UBI), and he’s been vocal about it since 2016, well before the advent of generative AI. He backs a research group called OpenResearch, which studied the effects of unconditional cash transfers of $1,000 per month to around 1,000 American individuals with an income of approximately $30,000 a year. The study found that the cash led to increased spending on basic needs and that “recipients worked less, but remained engaged in the workforce and were more deliberate with job searches and employment choices”. “Recipients had greater agency to make decisions that worked best for their lives and to prepare for the future, from moving neighbourhoods to expressing interest in new business ventures,” the report further stated. However, cash did not mean everything, and it wasn’t able to address challenges such as chronic health conditions, childcare, or housing costs. Of course, OpenResearch acknowledges such challenges, but implementing UBI is a completely different problem. Will UBI be More Difficult to Achieve Than AGI? Anton Korinek, a professor of economics at the University of Virginia, said in a research study that, unlike wealth distribution mechanisms that compensate “losers” within a country, there isn’t a well-established global institution for large-scale wealth redistribution across different countries. “Addressing this issue would require unprecedented levels of international cooperation and potentially the development of new global economic governance structures to share the benefits of AI more equitably across nations,” Korinek said. He also said that monetary policies will face new challenges, especially in understanding the relationship between unemployment and inflation. “Traditional measures of labour market slack may become irrelevant for inflation dynamics. Instead, new indicators of economic capacity utilisation, such as the intensity of capital use, may emerge,” he added. Korinek also said that sources of government revenue would have to shift from labour to other bases, such as capital driven by AI assets or new sources of taxes that capture AI-generated economic activity. “Key challenge will lie in reimagining macroeconomic policy for an era where AGI, rather than human labour, becomes the primary driver of economic growth and fluctuations.” Thus, it isn’t just engineers or AI safety researchers who will have a lot of work to do. “Economists have a crucial role in preparing for the age of AI,” Korinek predicted. “Furthermore, our [economists’] understanding of market dynamics and regulatory frameworks can inform the design of effective governance structures for AI.” That said, if UBI has its challenges, Altman also proposes universal basic computing (UBC) as a long-term and sustainable alternative to UBI. “Imagine owning part of the productivity, like a slice of GPT-7 compute, which you could use, donate, or resell—transforming access into empowerment,” Altman said.","excerpt":"But no one will care that much, he said.","categories":["AI Features"],"tags":["AI (Artificial Intelligence)","OpenAI"],"author_name":"Supreeth Koundinya","publish_date":"2025-01-16T12:33:46","publication_year":"2025","word_count":1094,"keywords":["Anthropic","Go","API","OpenAI","AI","GPT","generative AI","GAN","R","AI (Artificial Intelligence)","Redis"],"extracted_tech_keywords":["AI","generative AI","OpenAI","Anthropic","Redis","R","Go","API","GPT","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/sam-altman-genuinely-believes-openai-will-launch-the-first-agi\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10123941,"title":"OpenAI Chief Scientist Jakub Pachocki&#8217;s X Account Hacked","content":"OpenAI’s chief scientist Jakub Pachocki’s X account was recently hacked, according to Bloomberg. The hacked account posted that OpenAI is bridging the gap between AI and blockchain and claimed that all OpenAI users are eligible to receive a portion of a new token supply. Someone has gained access to Jakub Pachocki's X account (OpenAI Chief Scientist).This comes one year after Mira Murati's account was compromised (OpenAI CTO). pic.twitter.com\/qXZf9QuTjh— Smoke-away (@SmokeAwayyy) June 19, 2024 “We’re very happy to announce $OPENAI: the token bridging the gap between AI and blockchain technology. All OpenAI users are eligible to claim a piece of $OPENAI’s initial supply. Holding $OPENAI will grant access to all of our future beta programs,” read the post on X. Bloomberg confirmed that Pachocki’s account was compromised. “OpenAI is not getting into crypto, company comms just confirmed with me. Jakub Pachocki’s tweet that is making the rounds appears to be a hack,” a Bloomberg journalist posted on X. OpenAI is *not* getting into crypto, company comms just confirmed with me. Jakub Pachocki’s tweet that is making the rounds appears to be a hack.— Shirin Ghaffary (@shiringhaffary) June 19, 2024 Jakub Pachocki recently took over as the company’s chief scientist, succeeding Ilya Sutskever, who departed to pursue personal endeavors. As chief scientist, Pachochki leads OpenAI’s research efforts, focusing on scaling deep learning systems and advancing AI research. Pachocki joined OpenAI in 2017 and has held several key positions, including Research Lead for the Dota team, leader of the Reasoning Team, and head of the Science of Deep Learning team. He also served as Director of Research, leading the development of GPT-4 and OpenAI Five and conducting fundamental research in large-scale reinforcement learning and deep learning optimisation.","excerpt":"OpenAI is not getting into Crypto.","categories":["AI News"],"tags":["OpenAI"],"author_name":"Siddharth Jindal","publish_date":"2024-06-19T09:31:01","publication_year":"2024","word_count":284,"keywords":["OpenAI","AI","programming_languages:R","GPT","Aim","deep learning","AI research","R","llm_models:GPT"],"extracted_tech_keywords":["AI","deep learning","OpenAI","Aim","R","GPT","AI research","llm_models:GPT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/openai-chief-scientist-jakub-pachockis-x-account-hacked\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10095859,"title":"NVIDIA H100 Aces Generative AI, Breaks MLPerf Records","content":"NVIDIA’s flagship H100 chip has blown the competition out of the water yet again, demonstrating the best performance yet on a set of MLPerf training benchmarks. The GPU set new records across the board in a new attempt conducted in conjunction with CoreWeave and Inflection AI. The tests were run on a cluster of 3,584 H100 GPUs hosted on CoreWeave’s platform. These GPUs were linked together by the InfiniBand interconnect, with NVIDIA stating that this allows the GPUs to deliver performance at both standalone and scale levels. The MLPerf benchmarks aim to quantify the power of certain hardware by calculating the amount of time it takes to complete certain workloads. The tests consist of various LLMs and computer vision models, along with a handful of CNNs and RNNs. NVIDIA claims that the H100 has ‘delivered the highest performance on every benchmark’, with some standout examples being training GPT-3 in just 11 minutes, and completing the ResNet benchmark in just 0.18 minutes. Reportedly, the H100 was the only chip that was able to complete all benchmarks. As Team Green dominates, NVIDIA’s competitors are left holding the slack. AMD’s AI accelerator offerings are nowhere close to the H100. The company recently launched the M1300X AI accelerator chips, which aims to compete against the H100 with its 192GB of VRAM. While benchmark results have not been released for this new chip, the lukewarm response from the market is enough to show that the product won’t take over NVIDIA any time soon. Other notable competitors include Cerebras, which is shown to be cost equivalent to NVIDIA when training large language models, and Google’s TPUs, which are more energy efficient and powerful, but restricted to certain workloads. For the foreseeable future, it seems that NVIDIA will continue to ride the AI wave, fuelled by the dual engines of capable compute and a well-integrated software ecosystem.","excerpt":"The NVIDIA H100’s benchmark performance shows why Team Green is still the market leader in AI compute.","categories":["AI News"],"tags":[],"author_name":"Anirudh VK","publish_date":"2023-06-28T12:09:39","publication_year":"2023","word_count":310,"keywords":["Go","TPU","AI","ML","computer vision","GPT","Aim","RNN","CNN","R"],"extracted_tech_keywords":["AI","ML","computer vision","Aim","TPU","R","Go","GPT","CNN","RNN"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/nvidia-aces-generative-ai-breaks-mlperf-records\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10003200,"title":"MathWorks In Collaboration With NVIDIA&#8217;s DLI Offers New Deep Learning With MATLAB Course","content":"MathWorks has announced their comprehensive deep learning with MATLAB course, developed in collaboration with NVIDIA’s Deep Learning Institute. This two-day course is being offered in both instructor-led online and self-paced on-demand formats throughout the rest of the year, 2020. The course claims that, on completion, learners will be ready to apply GPU-accelerated deep learning techniques in MATLAB to common applications such as image classification, autonomous systems, voice recognition, and object detection. MathWorks aims to provide a comprehensive platform for building AI-driven systems that are based on decades of supporting complex engineering projects. GPU Coder generates optimised CUDA code from MATLAB code for deep learning, embedded vision, and autonomous systems, which allows developers to build solutions that run efficiently on NVIDIA GPUs. In addition, a MATLAB container from NVIDIA GPU Cloud (NGC), a hub for GPU-optimized AI and HPC software, provides a complete deep learning workflow that uses NVIDIA GPUs to accelerate neural network training to scale-up performance across nodes. According to David Rich, the director of MATLAB marketing, MathWorks, “The NVIDIA Deep Learning Institute plays a crucial role in developing hands-on training and showcasing how to use new techniques like deep learning to solve complex problems.” He further added, “This course offers a practical approach to deep learning that will help NVIDIA users to iterate quickly and converge on a solution that meets product and time-to-market requirements.” “There’s been a surge of interest in the Deep Learning with MATLAB course using NVIDIA GPUs,” said Will Ramey, senior director and global head of developer programs at NVIDIA. “Learning how to quickly and easily apply the power of NVIDIA GPUs to accelerate neural network training streamlines the process of application development and allows for more rapid deployment and faster time to market,” concluded Ramey. For dates and locations, click here.","excerpt":"MathWorks has announced their comprehensive deep learning with MATLAB course, developed in collaboration with NVIDIA’s Deep Learning Institute.  This two-day course is being offered in both instructor-led online and self-paced on-demand formats throughout the rest of the year, 2020. The course claims that, on completion, learners will be ready to apply GPU-accelerated deep learning techniques […]","categories":["AI News"],"tags":["Deep Learning Techniques","Mathworks","MATLAB"],"author_name":"Sejuti Das","publish_date":"2020-07-24T11:26:53","publication_year":"2020","word_count":298,"keywords":["CUDA","API","AI","neural network","MATLAB","ML","Mathworks","Aim","deep learning","object detection","Deep Learning Techniques","R"],"extracted_tech_keywords":["AI","ML","deep learning","neural network","Aim","object detection","CUDA","R","CUDA","API"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/mathworks-in-collaboration-with-nvidias-dli-offers-new-deep-learning-with-matlab-course\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10133912,"title":"Dell Technologies to Offer GenAI Foundations Course and Professional Certifications in India","content":"Dell Technologies recently announced that it is accelerating skills development in India by offering its GenAI Foundations course at upcoming forums in Mumbai and Bengaluru, equipping professionals with AI\/ML expertise and providing certifications to drive transformative business outcomes in the AI era. The specialised training program will aim to equip individuals with the necessary knowledge and expertise to leverage AI and machine learning technologies effectively driving transformative business outcomes. By focusing on GenAI, Dell Technologies is addressing the growing demand for skilled professionals in this rapidly evolving field. Check: GenAI Courses by ADasci The first 250 attendees who register for the Dell Tech Forums in Mumbai and Bangalore will have the opportunity to take an exam and receive a Dell Proven Professional certification, demonstrating their proficiency in GenAI. The tech solutions group also offers a wide range of technical courses including large language models, data engineering and cybersecurity. Coming this fall will be a GenAI governance and security course. Alok Ohrie, Dell Technologies India president and managing director laid emphasis on how Dell  Technologies is committed to empowering organisations with the skills they need to thrive in the AI era saying, “By investing in skills development, we are contributing to the broader digital transformation journey across India. At Dell, we ensure our customers have the skills and knowledge they need to lead in the age of AI” Pioneering Enterprise AI Dell Technologies has been at the forefront of  driving generative AI initiatives across healthcare, transportation, and public services by partnering with organisations like Northwestern Medicine, Duos Technologies, and the City of Amarillo to deploy AI solutions that improve patient outcomes, enhance railway safety, and democratise access to city services. Recently Wipro and Dell Technologies partnered to accelerate enterprise AI by offering localised AI processing solutions that cater to stringent data privacy requirements, alongside democratising access to advanced AI capabilities and enabling flexible, scalable deployment across cloud, data center, and edge environments. In April 2024, Dell Technologies and NVIDIA joined forces to deliver end-to-end generative AI solutions, combining advanced infrastructure and professional services to accelerate AI adoption across industries, unlocking a potential $4.4 trillion annual impact as businesses harness the transformative power of data-driven AI. Discover: List of Top 10 GenAI Courses","excerpt":"The first 250 attendees who register for the Dell Tech Forums in Mumbai and Bangalore will have the opportunity to take an exam and receive a Dell Proven Professional certification.","categories":["AI News"],"tags":["Courses","Dell","GenAI"],"author_name":"Tanisha Bhattacharjee","publish_date":"2024-08-28T13:32:47","publication_year":"2024","word_count":371,"keywords":["Go","GenAI","machine learning","Dell","AI","ML","Scala","RAG","Aim","generative AI","Courses","R"],"extracted_tech_keywords":["AI","machine learning","ML","generative AI","GenAI","Aim","RAG","R","Go","Scala"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/dell-technologies-to-offer-genai-foundations-course-and-professional-certifications-in-india\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10054470,"title":"It is crucial for children to gain knowledge of AI: Mehreen Shamim, Intel&#8217;s AI Impact Shapers Awardee","content":"Mehreen Mushtaq Shamim, a teacher from Bengaluru’s Delhi Public School, was one of the ‘Grand Winners’ who won the First Intel AI Global Impact Festival under the AI Impact Shapers Category: Teachers with Innovative AI Teaching Learning Practices Ideas. She received an Intel certificate and a prize worth USD 5,000. Mehreen is very positive about AI and the way it is being introduced to children at a young age. In an exclusive interview with Analytics India Magazine, she expressed how it is essential for children to learn AI. “In the future, there will be a lot of jobs that require the skill, and I believe that every child should learn AI, and no one should be left behind. AI will be the facilitator in some or another way, and children will be dealing with the technology. It is crucial for them at this point in time to gain knowledge of AI,” said Mehreen, telling how she believes that learning AI should not be limited to children in STEM education. “Children are interested in learning AI, and future jobs might depend on it. AI is not related to one field like IT. It’s multidimensional and takes all streams into account. Students of humanities and commerce will also find AI interesting,” she added. Mehreen has trained 250 teachers on how AI can be integrated with the English subject. She believes that AI can be present with subject pedagogies like English, Hindi, Maths, and even Social Sciences. Like many computer teachers in the country, she was also introduced to teaching AI to school students after CBSE introduced it as a subject for students of standard 9-12 in 2019 in its curriculum. Her passion for computers and her background in technical subjects played a key factor in her school being among the first ones chosen for the pilot programme by CBSE before introducing the course nationwide. She agrees that the course by CBSE (designed by Intel) is currently great and helps the young ones experientially learn AI and not the typical rote learning style. “Children learn to think out of the box and are surprisingly coming up with solutions to real issues. While teaching at NIIT, I realised that the older students surely learn more technical knowledge like application of Java and Python which is from a perspective to get a job, but the younger ones are very excited and fascinated about the technology,” she said. But she is also in favour of the course needing an upgrade on a timely basis as the technology is continuously growing. Mehreen herself comes from a strong educational background in technical courses as well as teaching. She is a Bachelor of Information Technology (BIT) from Manipal Academy of Higher Education. She also completed her Master of Computer Applications from ICFAI and completed OCP and SCJP from NIIT. To support her teaching career, she also completed her Bachelor of Education (B. Ed). She is in favour of AI access to all the children. She is very optimistic about the EdTech companies that introduce AI courses for children as they help in building an ecosystem. She said, “The support of Edtech companies is also required especially for quick learners who want to go beyond what is taught in school syllabus as they can take up advanced modules. Also, other than CBSE, no other board has AI for school students; EdTech platforms help such children get introduced to AI. The Edtech platforms have level-based courses and also provide competitions and hackathons. I would say they are also enablers in building an ecosystem where children know what AI is and how ML works.” Mehreen is in complete favour of ethical AI and loves the way CBSE has planned its syllabus. She especially finds children fascinated by the gamified versions of learning. For instance, children love how in a simple game of rock-paper-scissors, the system tries to learn the moves made by the students, analyses them and tries to win over them. “This helps in making children understand how the data you feed will affect the outcome. They can understand why systems by big companies can be biased. They understand that when you show only females as nurses to the system, it learns the bias and search results in something like Google will show female nurses.” With children learning AI, she believes that now will be the age where children’s school projects will go beyond chart papers and thermocol, and they will build working prototypes or solutions with AI. Mehreen is also associated with CBSE’s AISC (artificial intelligence students committee), which has over 25,000 members. She encourages schools and teachers to take the free training by CBSE and start teaching AI in schools. “Classrooms with projectors and a computer lab are the only infrastructure required to run an AI course. An installation of software like Jupyter notebook and\/or Anaconda in the existing systems is all that schools need to start teaching AI,” concluded Mehreen. Intel AI Global Impact Festival 2021Mehreen Mushtaq Shamim was the Grand Winner of ‘AI Impact Shapers Category: Teachers with Innovative AI Teaching Learning Practices Ideas,’ along with three others. They are Byoungchol Chang from South Korea, Marek Grzywna from Poland, and Kiruthika Ramanathan from Singapore.Intel initiated the AI Global Impact Festival 2021 with the theme ‘Enriching Lives with AI Innovations.’ At the inaugural festival, over 110,000 students, technologists, and future developers joined from 135 countries. They received over 230 AI Innovations entries from more than 20 countries. The festival was held in conjunction with Intel® Innovation digital event, a new educational tech event for developers and industry insiders to connect with Intel leaders and industry experts.","excerpt":"In an exclusive interview with AIM, Mehreen Shamim, Intel’s AI Impact Shapers awardee, shared the importance of building an ecosystem of AI learning","categories":["AI Features"],"tags":["Intel","Interviews and Discussions","STEM"],"author_name":"Meeta Ramnani","publish_date":"2021-11-29T14:00:00","publication_year":"2021","word_count":938,"keywords":["Go","artificial intelligence","STEM","AI","ML","RAG","Python","analytics","Jupyter","R","Java","Intel","Interviews and Discussions"],"extracted_tech_keywords":["AI","artificial intelligence","ML","analytics","Jupyter","RAG","Python","R","Go","Java"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/it-is-crucial-for-children-to-gain-knowledge-of-ai-mehreen-shamim-intels-ai-impact-shapers-awardee\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":20957,"title":"Alibaba Founder Jack Ma Says AI, Robots Will Create Unemployment","content":"While the World Economic Forum this year saw many prominent personalities emphasising the growth and potential in the technology sector — especially artificial intelligence and data analytics — Jack Ma, founder and executive chairman of Alibaba said that AI was going to be a big threat to humans. Speaking at a panel discussion during the WEF in Davos, Ma said, “AI, big data are a threat to human beings. AI and robots are going to kill a lot of jobs, because in the future, these will be done by machines.” While admitting that AI would create a lot of successful people and interesting careers in the future, Ma also said that it would lead “social problems”. “I think AI should support human beings. Technology should always do something that enables people, not disable people… The first technology revolution caused the World War I and the second technology revolution caused the World War II and now we have the third revolution. If there is a World War III, I think that should be against disease, pollution and poverty, not against ourselves,” he said. Alibaba which is the world’s sixth-largest Internet company by revenue, was in news recently because of their plans to enter the Indian market with its cloud-computing arm, Alibaba Cloud. Recently Infosys co-founder and celebrated corporate thought leader NR Narayana Murthy had also said that AI was “worrisome” for freshers in the IT sector. Murthy is only one in a long string of influential personalities who has sounded the alarm bells regarding automation, artificial intelligence, and its effect on employment. US’ former Secretary of State Hillary Clinton had also said that millions of jobs would be on the line if automated machines and applications were implemented on a large scale. While many noted names such as physicist Stephen Hawking and Tesla and SpaceX’s CEO Elon Musk have already sounded alarm bells with regards to Artificial Intelligence and its potential harm, there are others who don’t concur. Facebook’s Mark Zuckerberg, Microsoft founder Bill Gates and futurist Ray Kurzweil are some of the famous names who believe that AI can actually do more good to the humans than harm. Kurzweil, who currently runs a group at Google writing automatic responses to your emails in cooperation with the Gmail team, had once famously said, “It was fire that kept us warm, cooked our food, but also burnt our houses down. Technology is always a double-edged sword.”","excerpt":"While the World Economic Forum this year saw many prominent personalities emphasising the growth and potential in the technology sector — especially artificial intelligence and data analytics — Jack Ma, founder and executive chairman of Alibaba said that AI was going to be a big threat to humans. Speaking at a panel discussion during the […]","categories":["AI News"],"tags":["AI (Artificial Intelligence)","Alibaba","davos","jack ma","narayana murthy"],"author_name":"Prajakta Hebbar","publish_date":"2018-01-25T06:36:47","publication_year":"2018","word_count":404,"keywords":["big data","Go","artificial intelligence","programming_languages:R","AI","Alibaba","jack ma","programming_languages:Go","automation","Ray","analytics","AI (Artificial Intelligence)","R","davos","narayana murthy"],"extracted_tech_keywords":["AI","artificial intelligence","analytics","Ray","R","Go","big data","automation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/jack-ma-alibaba-ai-will-kill-jobs\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10131001,"title":"Protect AI Acquires Bengaluru-Based SydeLabs to Enhance LLM Security","content":"AI security company, Protect AI has announced the acquisition of SydeLabs, a company specialising in automated attack simulation (red teaming) for generative AI (GenAI) systems. This acquisition enhances Protect AI’s platform by improving its ability to test and secure large language models (LLMs) and solidifies its position as the only provider of end-to-end AI security solutions. SydeLabs, founded less than a year ago by former product and engineering leads from Google and MPL, has established itself as a pioneer in AI security. Based in Bangalore, India, SydeLabs developed SydeBox, a product providing comprehensive vulnerability assessments for GenAI systems. The SydeLabs team will join Protect AI and expand the local talent pool in Bangalore, complementing teams in Seattle and Berlin. The acquisition of SydeLabs extends our platform with unmatched red teaming capabilities, allowing customers to stress test, benchmark, and harden their large language models against security risks,” said Protect AI CEO Ian Swanson. SydeBox will be integrated into the Protect AI Platform and rebranded as Protect AI Recon. Recon identifies potential vulnerabilities in LLMs, ensuring enterprises can deploy AI applications confidently. Key features include no-code integration, model-agnostic scanning, and detailed threat profiling across multiple categories. Recon uses both an attack library and an LLM agent-based solution for red teaming and evaluating the security and safety of GenAI systems. The integration of SydeLabs’ technology aligns with formal guidance from NIST, Mitre, OWASP, and CISA, as well as mandates like the Executive Order on AI Safety and Security and the EU AI Act. Recon will help meet the growing demand for robust AI security solutions by offering detailed threat profiling across jailbreaks, prompt injection attacks, input manipulations, and other attack vectors. Ruchir Patwa, co-founder of SydeLabs, commented, “The combination of SydeLabs’ SydeBox and Protect AI’s platform provides customers with a comprehensive defense-in-depth solution for building, managing, testing, deploying, and monitoring LLMs.” Recon will enable Protect AI to address growing customer demand for robust AI security solutions, providing detailed threat profiling crucial for maintaining AI system integrity and security. This acquisition further solidifies Protect AI’s market presence and leadership in AI Security Posture Management (AI-SPM) solutions. When used alongside Layer, Protect AI’s LLM observability and monitoring solution, organizations can better secure LLM implementations across emerging security concerns associated with GenAI usage. Partners and ecosystem stakeholders will benefit from enhanced security capabilities, ensuring the AI ecosystem is better protected against potential threats.","excerpt":"SydeBox will be integrated into the Protect AI Platform and rebranded as Protect AI Recon.","categories":["AI News"],"tags":["Mergers and Acquisitions"],"author_name":"Siddharth Jindal","publish_date":"2024-07-31T23:02:04","publication_year":"2024","word_count":397,"keywords":["Go","GenAI","programming_languages:R","AI","programming_languages:Go","AI safety","generative AI","GAN","Mergers and Acquisitions","R"],"extracted_tech_keywords":["AI","generative AI","GenAI","R","Go","GAN","AI safety","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/protect-ai-acquires-bengaluru-based-sydelabs-to-enhance-llm-security\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":6381,"title":"A Structural Equation Modelling Approach to Analyse Factors Affecting on-line Shopping Experience.","content":"To determine the critical factors which lead to overall satisfaction in on-line shopping experience an empirical research was undertaken with a survey questionnaire as the research instrument. Conceptual model Figure 1: Conceptual Model The conceptual model portrays the following hypotheses which would need to be validated using structural equation modelling. Hypothesis-1: Ease of using an online shopping portal is positively related to the overall satisfaction experienced by a customer while shopping online Hypothesis-2: Quality of product display is positively related to the overall satisfaction experienced by a customer while shopping online Hypothesis-3: Price of products sold via online shop is positively related to the overall satisfaction experienced by a customer while shopping online Hypothesis-4: Quality of delivery of product and after sales service are positively related to the overall satisfaction experienced by a customer while shopping online The structural equation modelling would need the following latent constructs and the associated manifest\/ indicator variables as given below. Latent constructs in the model and the associated manifest\/ indicator variables Ease of use is explained by Ease of navigation: Ease with which a customer is able to find\/browse a desired product using the online shopping portal Page load time: Time taken to display the website by the browser Payment options: Cash on delivery, Net banking, Credit card, etc. Check-out-process: Ease with which a customer is able to place an order once he\/she has selected the product Quality of product display is explained by Product description: Text description giving details of the product Product review: Reviews about the product by existing customers Product images: Quality of image of the products Range of products: Number of products offered by the online shopping portal Price is explained by Price of product: Online price of a product Shipping cost: Cost of delivering the product to designated place, which is charged to the customer Coupons and discounts: Regular or seasonal discounts and coupons which can be utilised for shopping Quality of delivery and after sales is explained by Customer service: Experience of talking with a customer service executive while resolving a query Exchange\/return policy: Ease with which a customer can return a product if the delivered product is faulty On-time delivery: Punctuality of delivering a product Overall satisfaction is explained by Recommendation Repurchase Satisfaction The manifest\/indicator variables were measured from the survey responses to the questionnaire. Methodology Structural Equation Modelling (SEM) is used to confirm the stated hypothesis of the conceptual model. The final sample size was obtained as 98, margin of error of 10%. Data analysis Structural Equation Modelling To run a structural model as presented by the conceptual model, an online survey was designed to capture the responses of online shoppers. A database of 98 responses was prepared based on the data collected using survey. Cronbach’s Alpha reliability test was performed to check whether the stated indicator variables are able to measure the same construct. Reliability Analysis Construct Name Cronbach’s Alpha Ease of use 0.803 Product Display 0.804 Price 0.622 Delivery & After sales 0.807 Satisfaction 0.939 After reliability test it was found that ‘price’ was not being measured reliably by its indicator variables (price of product, shipment cost and coupons&discounts) as its Cronbach’s alpha value is less than 0.7. Thus the latent construct ‘price’ was not included in SEM. Modified model Figure 2: Revised Conceptual Model based on Cronbach’s Alpha values. IBM SPSS AMOS (version 21) was used to test the conceptual model Results from structural equation modelling. Results from running SEM (using AMOS v21) Indicators of model fit: Chi square\/degrees of freedom    =   1.122 Overall model p-value                    = 0.236 GFI                                                      = 0.914 AGFI                                                   = 0.857 CFI                                                      = 0.991 NFI                                                     = 0.99 SEM LatentConstructsInModel CMIN\/DF P value CFI GFI AGFI NFI Good Fit 0 <= 2 .05 <=1 .97 <= 1 .95 <= 1 .9 <= 1 .95 <= 1 Acceptable Fit 2 <= 3 .01 <=.05 .95<= .97 .9 <= .95 .85 <= .9 .9 <=  .95 EaseOfUse 1.122 0.236 0.991 0.914 0.857 0.923 Product Display Delivery & After sales Satisfaction GoodFit GoodFit GoodFit NotGoodFit NotGoodFit AcceptableFit The model fit was good considering the values of CMIN\/DF, p-value, CFI and NFI, we can now proceed to have a look at the estimates weights and their levels of significance. Conclusion Looking at the above estimates we can conclude that: Quality of timely delivery and after sales significantly affect the overall satisfaction experienced by a customer while shopping online (Critical Ratio = 6.171 > 1.96, individual p-value is *** indicating highly significant). Ease of using an online shopping portal is positively related to the overall satisfaction experienced by a customer while shopping online. We can conclude this hypothesis but with caution (Critical Ratio = 1.909 < 1.96, individual p-value is.056 indicating somewhat significant relationship). Quality of product display is not positively related to the overall satisfaction experienced by a customer while shopping online (Critical Ratio = -.369 < 1.96, individual p-value is .712 indicating not significant relationship). The above observations are incorporated in the final model where solid lines show significant relationship dotted line indicates a relationship which is not significant. Thus from SEM analysis it was found that the below two hypothesis are correct. H-1: Ease of using an online shopping portal is positively related to the overall satisfaction experienced by a customer while shopping online H-4: Quality of delivery of product and after sales service are positively related to the overall satisfaction experienced by a customer while shopping online Figure 4: Final Model as seen in the AMOS output. References http:\/\/articles.economictimes.indiatimes.com\/2008-01-08\/news\/27719779_1_middle-class-spending-marketers http:\/\/www.thehindu.com\/sci-tech\/technology\/internet\/india-has-slowest-internet-penetration-growth-in-apac\/article6085420.ece http:\/\/www.ey.com\/IN\/en\/Industries\/Technology\/Re-birth-of-e-Commerce-in-India Services Marketing: Integrating Customer Focus Across the Firm (Valarie A Zeithalm, Mary Jo Bitner, Dwayne D. Gremler and Ajay Pandit) http:\/\/www.academia.edu\/1003804\/E-Service_Quality_A_Conceptual_Model Jukka Ojasalo,Laurea University of Applied Sciences, Finland http:\/\/www.iosrjournals.org\/iosr-jbm\/papers\/vol3-issue4\/E0343440.pdf Dineshkumar, P.Vikkraman [attachments]","excerpt":"To determine the critical factors which lead to overall satisfaction in on-line shopping experience an empirical research was undertaken with a survey questionnaire as the research instrument. Conceptual model Figure 1: Conceptual Model The conceptual model portrays the following hypotheses which would need to be validated using structural equation modelling. Hypothesis-1: Ease of using an […]","categories":[],"tags":[],"author_name":"Krunal Patel","publish_date":"2014-10-28T06:26:58","publication_year":"2014","word_count":951,"keywords":["Go","TPU","programming_languages:R","AI","ML","programming_languages:Go","Ray","R"],"extracted_tech_keywords":["AI","ML","Ray","TPU","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/structural-equation-modelling-approach-analyse-factors-affecting-line-shopping-experience\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10008424,"title":"Why Companies Need To Pay Attention To Software Supply Chain Security","content":"A software supply chain attack happens when an attacker enters and modifies software in the complicated software development supply chain to imperil a target farther down on the chain by injecting their malicious code. These inserts can be employed to further modify code by getting system permissions or to directly deliver a malicious payload or backdoor package. Software supply chain attacks exploit established channels of system verification to acquire privileged access to systems and to compromise big networks and databases. By software supply chain, we basically mean everything that goes into making a software product, including all the components, packages and repositories. Modern software products comprise a huge number of dependencies on other code and therefore tracking down which vulnerabilities compromise which software products and tools can be a challenging technical effort. Companies must therefore understand all the components in software or packages they are consuming. Today, a large part of the software is open-source, and it is difficult to know precisely who has contributed to a piece of code and where that code has come from before it ends up in an enterprise environment. Supply chain software security deals with all of that, and because there is so much software out there, most of the security processes need to be automated. Breaches Can Happen Due To Lag In Secure Software Supply Chain There have been many breaches so far due to problems or mistakes in software supply chain security.  For instance, NotPetya and the Equifax data breach in 2017, affected millions of users. It showed the potential scale of software supply chain attacks and their strategic utility for nefarious cyber activities. Many times, companies also don’t disclose in time that their software or network has been compromised, putting their entire ecosystem of customers and partners at risk. For instance, Kingslayer Windows log management software that was compromised in 2017, had the Chinese attackers target the Windows IT admin application to inject malicious code under a valid signature, which could diffuse either by updating or downloading the application. The attack compromised systems across the globe, including colleges, defence companies, government organisations, banks, IT and telecom firms, and other businesses. This happened as the malware installed a secondary package which could upload and download files, execute malicious programs, and run arbitrary shell commands. The attack was found when a defence contractor found a piece of software in their environment pinging an IP address which was a known bad IP. Now, this could have been prevented if there was vigilant software supply chain security and management in place. The good news is that now this is coming to life with advancements in DevSecOps and Gitops. You Want To Know Everything In Your Software Environment As a developer, you want to know what’s in your environment and discover any software problems that could be a possible security issue such as a license issue or a dependency vulnerability, and patch it as soon as possible. Containers are an essential part of the software supply chain and make patching a smooth task. Containers make it easy for you to rebuild, test then redeploy into your environment without actually having any downtime. If companies are proactive in this regard, they can ensure optimal security at the container level with automated code scanning and patching. Of course many times, it’s easier said than done, as software pipelines can become very complex with time. But proactive patching can address the majority of security issues in time before any major data breach happens. Most open-source projects do not follow stringent organisational structures and instead, depend on self-organisation and a collaborative approach to drive software innovation and development. That can be a major issue in ensuring a proper software security supply chain. Now with GitOps, teams in an operating model can not only define their infrastructure as code but also make deployments and changes to it by submitting pull requests that people can review and approve collectively. GitOps has made it ubiquitous to have continuous integration on a continuous delivery server when developing your application. This helps in embedding DevSecOps practices when teams are working on building their products. Tools For Secure Software Supply Chain On GitHub While many companies use Github for hosting their code, there’s a feature called dependency graph which can do the scanning for security analysis, including information about the licenses and security alerts for vulnerable dependencies. For example, if a team adds a new dependency that has a vulnerability or there’s a new vulnerability discovered on one of their existing dependencies, they get an alert, which can be resolved with patching. Github also has an automated feature, where an automated bot called Dependabot sends users an automated pull request to suggest that a user has upgraded to a patched version for vulnerable dependencies. Many of Github’s enterprise users are utilising Dependabot and the GitHub dependency graph to know which dependencies they use, their vulnerabilities, how to patch them and return to normal work. Users can discover vulnerabilities across a codebase with CodeQL, Github’s semantic code analysis engine, which lets users query code as though it were data.","excerpt":"A software supply chain attack happens when an attacker enters and modifies software in the complicated software development supply chain to imperil a target farther down on the chain by injecting their malicious code. These inserts can be employed to further modify code by getting system permissions or to directly deliver a malicious payload or […]","categories":["AI Features"],"tags":["AI Companies","supply chain analytics projects"],"author_name":"Vishal Chawla","publish_date":"2020-09-27T18:00:00","publication_year":"2020","word_count":852,"keywords":["Go","programming_languages:R","AI","innovation","programming_languages:Go","Git","ViT","AI Companies","GitHub","GAN","supply chain analytics projects","R"],"extracted_tech_keywords":["AI","R","Go","Git","GitHub","GAN","ViT","innovation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/why-companies-need-to-pay-attention-to-software-supply-chain-security\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10098760,"title":"OpenAI Doesn’t Care about NYT Enough","content":"The copyright infringement case with generative AI models has been taking a lot of turns. The latest being The New York Times‘ potential lawsuit against OpenAI for copyright infringement. The news outlet believes that OpenAI’s models are trained on NYT’s intellectual property data and copied the style of their writers to give ChatGPT the ability to articles in the same manner, whenever prompted. This will likely be one of the first lawsuits against OpenAI to actually put the company in trouble. NYT recently published its terms of services focusing on prohibiting AI companies from scrapping articles for training AI models hinting at a possible lawsuit on OpenAI. If NYT successfully proves that its content is illegally used, OpenAI might have to delete its entire dataset used for training AI models and cough up a fine of up to $150,000 for each infringed content. Meanwhile, OpenAI had made a licensing deal with Associated Press (AP), accessing its archive for building better AI models. It has also partnered with several news agencies recently. Now, AP has decided to join other news organisations to frame guiding principles of using AI in newsrooms. In the report, AP said that a lot of news organisations are concerned that their content is being used without permission. The tricks up its sleeves There is no doubt that OpenAI took help from Microsoft’s Bing to crawl the internet. According to Gilles Babinet, the company must have crawled up to 250,000 websites for building GPT, without asking anyone — true, to be taken with a pinch of salt. Responding to this, Yann LeCun, Meta AI chief, gave an example of search engines. “Google, Bing, and others crawl the internet constantly. That’s not the problem,” he said and asked where the problem exactly lay. Arguably, there is a difference between crawling and reusing the content, but even then, the argument for it being a “copyright” case does not stand strong. Sébastien Hubert explains how neural networks do not actually store any data, but just represent the understanding of the data. This arguably is a lot like how humans work. “GPT has read The Three Musketeers but is unable to quote any chapter verbatim upon request. An LLM is a sort of super-reader – it doesn’t copy anything,” explained Hubert. Interestingly, an important thing to note here is that NYT is only suing OpenAI, and not Google for making Bard. For NYT, the problem comes from the fact that Bing synthesises the content without creating traffic for the newspaper thereby hitting advertising revenue. Bing AI is an internet “wrapper”, which breaks the economic model of many sites. It seems, NYT woke up to the problems of ChatGPT only after realising that OpenAI is partnering with other publishers and not them. It must be noted that NYT is going to receive $100 million from Google for the next three years in a deal where the tech giant will be able to publish content on its platforms. Google is testing its new AI writing tools in partnership with NYT, WSJ, and Washington Post. This might possibly be a hint towards a partnership to rival OpenAI and AP. NYT has a problem with OpenAI, not generative AI OpenAI’s GPTBot, which recently came to everyone’s notice, says that the company will automatically scrape the internet and websites for training its AI models. To opt out, companies have to voluntarily put a line of code on their website to block the crawler. There is no doubt that it must have crawled all news outlets to train its AI models. In 2015, federal appeals court ruling for Google, found to be scanning millions of books for Google Books library, was that the library was not able to create a significant market substitute for the original books, and thus fell under the ‘fair use’ of using its models. For OpenAI, this would be difficult to prove. According to some experts, ChatGPT might be able to form an alternative to visiting NYT articles, dropping the traffic on the website. This means that OpenAI is not using the content from the website “fairly”. Interestingly, ChatGPT is not connected to the internet. So even if NYT is able to prove that it was trained on its articles illegally, it is not doing that after the cut off time of 2021. GPT-4’s Browse with Bing feature was also discontinued some time back, possibly because of the same reasons. It seems like OpenAI was aware about the copyright issues beforehand and took an early step. It might now be hard to prove that it is actually a competition to NYT or people are merely using it to summarise things in their style — something that NYT only wants Google’s AI tools to do in the future. Nevertheless, writers have been protesting against generative AI technology for replacing their jobs for a long time. Now that NYT is with them, they might finally be able to catch up, but the push is not against AI models, but just Google’s competitor, OpenAI.","excerpt":"…because Google does.","categories":["Global Tech"],"tags":["OpenAI"],"author_name":"Mohit Pandey","publish_date":"2023-08-21T15:29:11","publication_year":"2023","word_count":837,"keywords":["Go","ChatGPT","Meta AI","OpenAI","AI","neural network","AWS","BERT","generative AI","R"],"extracted_tech_keywords":["AI","neural network","generative AI","ChatGPT","OpenAI","Meta AI","AWS","R","Go","BERT"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/openai-doesnt-care-about-nyt-enough\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10021735,"title":"Reinventing Deep Learning Operation Via Einops","content":"Einops, an abbreviation of Einstein-Inspired Notation for operations is an open-source python framework for writing deep learning code in a new and better way. Einops provides us with new notation & new operations. It is  a flexible and powerful tool to ensure code readability and reliability with minimalist yet powerful API. Supported Frameworks numpypytorchtensorflowjaxcupychainergluontf.kerasmxnet (experimental) https:\/\/twitter.com\/karpathy\/status\/1290826075916779520 In case you need convincing arguments for setting aside time to learn about einsum (https:\/\/t.co\/2lA3Bsh53D) and Alex Rogozhnikov's einops (https:\/\/t.co\/SY4yJAktEh). Screenshot taken from https:\/\/t.co\/RsCX5P5NLv. pic.twitter.com\/TTvb8pENIb— Tim Rocktäschel (@_rockt) February 21, 2020 Requirements & Installation Python>=3.6 You can install Einops via PyPI. !pip install einops Tutorials Here are a few examples to get started with Einops. A. Basics of Einops Instead of writing, y = x.transpose(0, 2, 3, 1), in Einops we represent it as y = rearrange(x, 'b c h w -> b h w c') This demo will cover basics of Einops reordering, composition and decomposition of axes, operations like rearrange, reduce, repeat and what all you can do with these single operations. Let’s get started! Import all the required packages and load the data. Here we are using 6 images. A sample of it is shown below:Perform all three operations: Rearrange Rearranging a single image: # we'll use three operations from einops import rearrange, reduce, repeat # rearrange, as its name suggests, rearranges elements # below we swapped height and width. # In other words, transposed first two axes (dimensions) # rearrange elements according to the pattern rearrange(ims[0], 'h w c -> w h c') Composition of axes: # einops allows seamlessly composing batch and height to a new height dimension # We just rendered all images by collapsing to 3d tensor! rearrange(ims, 'b h w c -> (b h) w c') Decomposition of axes : # decomposition is the inverse process - represent an axis as a combination of new axes # several decompositions possible, so b1=2 is to decompose 6 to b1=2 and b2=3 rearrange(ims, '(b1 b2) h w c -> b1 b2 h w c ', b1=2).shape # finally, combine composition and decomposition: rearrange(ims, '(b1 b2) h w c -> (b1 h) (b2 w) c ', b1=2) Order of the axes matter : # compare with the next example rearrange(ims, 'b h w c -> h (b w) c') # order of axes in composition is different # rule is just as for digits in the number: leftmost digit is the most significant, # while neighboring numbers differ in the rightmost axis. # you can also think of this as lexicographic sort rearrange(ims, 'b h w c -> h (w b) c') rearrange doesn’t change the number of elements and covers different numpy functions (like transpose, reshape, stack, concatenate, squeeze and expand_dims). You can check all types of rearrangement, here. Reduce : Instead of worrying about  x.mean(-1),  Einops gives you a option of directly reducing the image as : # average over batch reduce(ims, 'b h w c -> h w c', 'mean') If the axis is not present in the output, that means it is reduced and also provides different kinds of methods to reduce on like mean, min, max, sum, etc. # the previous is identical to familiar: ims.mean(axis=0) # but is so much more readable reduce combines the same reordering syntax with reductions (mean, min, max, sum, prod, and any others). You can check all the utilities of the reduce function here. Reduce ⇄ Repeat : reduce and repeat are like opposite of each other: first one reduces amount of elements, second one increases. # compute max in each image individually, then show a difference x = reduce(ims, 'b h w c -> b () () c', 'max') - ims rearrange(x, 'b h w c -> h (b w) c') repeat additionally covers repeating and tiling. Some fancy examples are available here. You can check Fundamental Demo of Einops in this Colab Notebook. B. Einops for Deep Learning This demo contains usage of some deep learning packages, important cases for DL models and at last functionality of einsops.asnumpy and einops.layers. You can select your choice of framework.Einops functions work with any tensor like they are native to the framework. For the example purpose, here the framework choice is PyTorch. Let’s start with very basic computation. converting bchw to bhwc format and back: y = rearrange(x, 'b c h w -> b h w c') guess(y.shape) Backpropagation is quite common in DL models. Here is the code for backpropagation via einops. y0 = x y1 = reduce(y0, 'b c h w -> b c', 'max') y2 = rearrange(y1, 'b c -> c b') y3 = reduce(y2, 'c b -> ', 'sum') if flavour == 'tensorflow': print(reduce(tape.gradient(y3, x), 'b c h w -> ', 'sum')) else: y3.backward() print(reduce(x.grad, 'b c h w -> ', 'sum')) einops.asnumpy : This function is used to convert tensor into numpy. from einops import asnumpy y3_numpy = asnumpy(y3) print(type(y3_numpy), y3_numpy.shape) Common Building Blocks of Deep Learning Flattening y = rearrange(x, 'b c h w -> b (c h w)') guess(y.shape) Space-to-Depth y = rearrange(x, 'b c (h h1) (w w1) -> b (h1 w1 c) h w', h1=2, w1=2) guess(y.shape) Depth-to-Space y = reduce(x, 'b c h w -> b c', reduction='mean') guess(y.shape) Reductions Global Average Pooling y = reduce(x, 'b c h w -> b c', reduction='mean') guess(y.shape) Max-pooling, here with kernel 2 X 2 y = reduce(x, 'b c (h h1) (w w1) -> b c h w', reduction='max', h1=2, w1=2) guess(y.shape) # you can skip names for reduced axes y = reduce(x, 'b c (h 2) (w 2) -> b c h w', reduction='max') guess(y.shape) Squeeze & Unsqueeze (expand_dims) # models typically work only with batches, # so to predict a single image ... image = rearrange(x[0, :3], 'c h w -> h w c') # ... create a dummy 1-element axis ... y = rearrange(image, 'h w c -> () c h w') # ... imagine you predicted this with a convolutional network for classification, # we'll just flatten axes ... predictions = rearrange(y, 'b c h w -> b (c h w)') # ... finally, decompose (remove) dummy axis predictions = rearrange(predictions, '() classes -> classes') Stacking Start with a list of tensors list_of_tensors = list(x) New axis (one that enumerates tensors) appears first on the left side of expression. Just as if you were indexing a list – first you’d get tensor by index. tensors = rearrange(list_of_tensors, 'b c h w -> b h w c') guess(tensors.shape) Or, Stacking along last dimension # or maybe stack along last dimension? tensors = rearrange(list_of_tensors, 'b c h w -> h w c b') guess(tensors.shape) Concatenation Concatenation over First Dimension tensors = rearrange(list_of_tensors, 'b c h w -> (b h) w c') guess(tensors.shape) Concatenation over Last Dimension tensors = rearrange(list_of_tensors, 'b c h w -> h w (b c)') guess(tensors.shape) You can check all other Functionalities here. C. Einops for PyTorch Slowly but surely, einops is seeping in to every nook and cranny of my code. If you find yourself shuffling around bazillion dimensional tensors, this might change your life: https:\/\/t.co\/xYgSrbqoy9— Nasim Rahaman (@nasim_rahaman) January 11, 2020 References You can refer to official Codes, Docs & Tutorials here. GithubWebsiteTutorialsPyTorch Tutorial or  hereAPI Usage","excerpt":"Einops, an abbreviation of Einstein-Inspired Notation for operations is an open-source python framework for writing deep learning code in a new and better way. Einops provides us with new notation & new operations. It is  a flexible and powerful tool to ensure code readability and reliability with minimalist yet powerful API. Supported Frameworks  numpy pytorch […]","categories":["AI Trends"],"tags":["Deep Learning","JAX","Keras","numpy","Pytorch","Tensorflow"],"author_name":"Aishwarya Verma","publish_date":"2021-03-09T14:00:00","publication_year":"2021","word_count":1203,"keywords":["Pytorch","NumPy","Keras","AI","PyTorch","ML","numpy","Colab","CuPy","deep learning","JAX","Deep Learning","TensorFlow","Tensorflow"],"extracted_tech_keywords":["AI","ML","deep learning","TensorFlow","PyTorch","JAX","Keras","Colab","NumPy","CuPy"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/reinventing-deep-learning-operation-via-einops\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":30218,"title":"10 Top Tech Skills In High Demand In 2019","content":"Data skills continue to be in high demand with enterprises looking to get the most out of their data. With organisations looking to capitalize on their data assets, the role of DevOps engineer, Python programmer, Data Engineers and Machine Learning engineer has become central to enterprises. In this article, we list down the top 10 critical tech skills IT professionals must possess to compete and multi-task.  While there’s a raft of technical skills that are required, the reality is that IT landscape is one that is undergoing frequent changes. The landscape is such that the technology is constantly challenged by new technology, complex problems, or dynamic roles. While new technologies and tools crop up continuously on the AI and analytics landscape, it’s important for developers to have a broad understanding of these tools. From how to use Hadoop or Big Data querying to machine learning and AI, there are some important attributes and skills that have emerged as a must-have going into 2019. To be successful in this field, one requires a mix of programming skills and understanding of computational aspects such as dealing with large volumes of data and working with real-time data. In today’s competitive environment, programmers should continuously re-learn and update their hard skills. So what does it take to be a data science whiz. We list down top 10 updated skills to succeed in 2019: 1) DevOps: In 2017, the role of DevOps engineer landed at the no. 3 spot on Indeed’s list of top jobs in America for 2017, in terms of salary, number of job postings, and opportunities for growth. Be it in deployment pipelines or systems architecture, DevOps engineers implement automation technologies and tools at all levels. According to a recent news report, DevOps positions have grown by 106% in the last few years and boast an average base salary of $123,165. With DevOps becoming the most in-demand skill in 2019, get an industry-recognised DevOps certification from Inventateq which provides classroom and live instructor led online Devops training in popular DevOps tools like Jenkins, Puppet, Chef, Ansible, SaltStack, Nagios, GIT, Splunk, Docket, Build, Release and Maven. You can also dive further with sub-topics which will drive in-depth into both theory and practicals. 2) Big Data Hadoop: There was a time when Hadoop captured a sizable chunk of the big data market and is still a relevant platform for data storage. According to a industry prediction, the Hadoop market will grow to $40 billion by 2021 and this will require skills to develop, manage and administer Hadoop implementations. If you are looking for technical training in Hadoop architecture, features and ecosystem, and a deeper understanding of Apache Hadoop, with HDFS and MapReduce, then Inventateq’s Big Data Hadoop training features industry use-cases and real projects. In this course, you can deep dive into Hadoop ecosystem and also understand the various tools with their functionalities. To cope with big data efficiently, new technologies appeared that enabled distributed data storage and parallel data processing. Apache Hadoop (with its HDFS and MapReduce components) was a pioneering technology. It is important to gain a technical background 3) Python Django: It is the most in-demand and fastest growing programming language in 2019. Python has significantly outstripped other languages in terms of community base of active developers. Python’s popularity has soared exponentially with the language being one of the requisites for an entry-level job. At Inventateq, you can gain job-oriented python training with Django and gain an in-depth understanding of using Django, an open source web application and learn how to create web application framework with Django functionality. The in-depth course will get you started on building any software applications on your own and learn how to build Django template systems. 4) Data Science with R & Python: R & Python are billed to be the prerequisites for candidates looking to build a career in data science. Developers and IT professionals are rushing to upskill in R & Python but the key is to sharpen the learning experience with real industry projects and case studies. At Inventateq, you will get real hands-on training with case studies and projects and use anaonymised datasets from data-intensive companies like Amazon, Facebook, Adobe, Walmart among others. The data science courses covers core data science tools, like Hadoop and Spark and how to deploy models on Azure and AWS. 5) Machine Learning: As machine learning grows in demand, enterprises and startups investing in specialized training and certifications to help broaden machine learning skillset and use advanced solutions to help scale their efforts. With Inventateq’s Machine Learning Course and Certification with Tensorflow APIs, IT professionals can master the basics of Python, Spark, R, SQL and basic statistics. Learn the key components such as feature engineering and gain an understanding of real-world applications of ML for face recognition, speech recognition and text mining. 6) Artificial Intelligence: With the ongoing excitement around artificial intelligence being the emergent technology and enterprises baking AI into their products, AI is set to transform the business landscape. The hottest skill today, AI professionals are expected to fetch salaries between $300,000 to $500,000. With the buzz around AI growing, IT professionals are rushing to upskill. This artificial intelligence course is meant for professionals with a strong CS and Math background and covers the fundamentals of AI, machine learning, parallel and distributed systems, data mining and algorithm engineering among other areas. 7) RPA Tools: As companies race to embrace digital transformation with RPA and other automation technologies, RPA tools alongside AI and cognitive computing have emerged as a critical skillset required to improve operational efficiency and facilitate better customer experience. Looking to gain hands-on experience with RPA tools from UiPath, Automation Anywhere and Blue Prism. This industry-leading rpa training from Inventateq covers a range of industry use cases from healthcare, insurance sector, human resources and financial services. 8) AWS Certification: In the era of cloud, enterprises and businesses are racing to develop use cases that depend on cloud platforms and one such cloud computing platform, AWS has cornered a global market share of 40 percent. And IT professionals who want to learn about AWS to maximize their company’s  cloud environment and avoid skills gaps can get hands on aws training and learn about the core areas such as Elastic Compute Cloud (EC2), database services, storage services and master the concepts of IaaS and PaaS with this instructor-led, classroom certification. 9) Tableau: With the BI tools winning over new businesses and enterprises, Tableau’s popularity and user base has surged significantly over the years. And BI vendors are quickly adding a slew of features to automate data analysis. With new upgrades and automation features being injected, Tableau, which is in hot demand allows developers to bridge the gap between data and business leaders and helps them visualize company data better.  This real-time tableau training covers topics such as filters, graphs, table calculation, aggregating data, data blending and dashboards. 10) Digital Marketing Analytics: With every business going digital, business owners look for digital marketers who can understand their customer data, derive insights from analytics and create a compelling digital strategy and measure the ROI achieved against the KPIs. There is a need to go beyond the metrics and numbers and understand key topics such as website optimization and social media analytics. Inventateq’s certification digital marketing courses covers key topics such as Google Analytics and social media analytics required to mine user sentiment.","excerpt":"Data skills continue to be in high demand with enterprises looking to get the most out of their data. With organisations looking to capitalize on their data assets, the role of DevOps engineer, Python programmer, Data Engineers and Machine Learning engineer has become central to enterprises. In this article, we list down the top 10 […]","categories":["AI Trends"],"tags":["cognitive computing human capital","devops tools"],"author_name":"Richa Bhatia","publish_date":"2018-11-13T12:00:54","publication_year":"2018","word_count":1234,"keywords":["data science","artificial intelligence","machine learning","AWS","AI","cloud computing","ML","devops tools","RAG","analytics","cognitive computing human capital","TensorFlow"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","data science","analytics","TensorFlow","RAG","cloud computing","AWS"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/10-top-tech-skills-in-high-demand-in-2019\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10002350,"title":"Aeternity Starfleet Program Propels Blockchain Adoption In India","content":"AE Ventures, a blockchain investment company that launched the æternity Starfleet India program earlier last year, has registered over 170+ early-stage startups for the first edition of the æternity’s Global Starfleet Program in India. Over 50% of the registered startups are businesses with traction\/users\/revenue that are willing to implement blockchain for the very first time through æternity, leading to an acceleration in blockchain adoption amongst enterprises across sectors. Indian businesses are now increasingly open to exploring and implementing blockchain as a technology. With the business benefits of blockchain becoming evident to enterprises through strong used cases in the market, a sizable share of the businesses is increasingly implementing blockchain in their projects now. While still, in its nascent stage, India’s blockchain market is expected to grow at a CAGR of 37% till 2024 according to a report. Blockchain’s transformational potential in India has been recognised by enterprises across sectors in India. The fintech sector recorded the most affinity to adopting blockchain used-case from the registered startups, comprising of over 14% of the registered startups, followed by the healthcare sector with 13% and agri-tech with 10% of the registered startups. Blockchain-based used cases in social media marketing, public sector, and renewable energy are also garnering considerable interest in the Starfleet India program. The fintech sector has seen the highest adoption, despite the cautious regulatory approach in India. A conducive regulatory and government procurement policy environment, will accelerate startup growth and drive blockchain implementation not only in the BFSI sector but across other areas as well. Nikola Stojanow, the CEO of AE Ventures said, “We are very excited to see this trend among companies that haven’t implemented blockchain, showing affinity towards exploring this technology when a strong used case is available. This is an example of the increased awareness that has helped the blockchain ecosystem, and we are happy to be a part of the journey that busts any myths about blockchain and helps increase its adoption across India. We look forward to working with the new wave of startups selected for Starfleet India and are very excited to empower India’s blockchain ecosystem.” As per reports from the Startup India initiative, the Indian startup ecosystem is the third-largest in terms of the number of tech startups following the US and China. And while the ecosystem is thriving, new players often find it difficult to access the correct channels due to a gap in the opportunities to network, receive mentorship and the right platforms to pilot their innovations. Aeternity Global Starfleet program, organised by AE Ventures in partnership with IBC Media aims to bridge this gap and provide relevant startups with the necessary impetus in the form of initial capital funding, along with six months of technical support, go-to-market strategy access and help with business development, marketing, and legal affairs. Raghu Mohan, the CEO of IBC Media, added, “This change in trend is an example of a more favourable ecosystem in India for blockchain technology to thrive. We are happy to be a part of this changing mindset and truly believe that accelerator programs like Starfleet India are pivotal, in bringing blockchain to more businesses and in helping companies realise that, the capabilities to implement your tech are already available here. Hence, we do not need to make new ones.”","excerpt":"AE Ventures, a blockchain investment company that launched the æternity Starfleet India program earlier last year, has registered over 170+ early-stage startups for the first edition of the æternity’s Global Starfleet Program in India. Over 50% of the registered startups are businesses with traction\/users\/revenue that are willing to implement blockchain for the very first time […]","categories":["AI Features"],"tags":[],"author_name":"AIM Media House","publish_date":"2020-03-15T13:59:52","publication_year":"2020","word_count":547,"keywords":["Go","API","funding","AI","innovation","RAG","Aim","GAN","R","startup"],"extracted_tech_keywords":["AI","Aim","RAG","R","Go","API","GAN","innovation","startup","funding"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/aeternity-starfleet-program-propels-blockchain-adoption-in-india\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10041140,"title":"Stanford University Professor Maneesh Agrawala On Video Editing Tools, Deep Fakes &#038; More","content":"Editing is an art. Bringing a story to life calls for tons of patience and hard work. But, what if we told you video editing could be as simple as text editing. Stanford University computer scientist Maneesh Agrawala, and his team has developed such a video editing software. Agrawala is the Forest Baskett professor of computer science and director of the Brown Institute for media innovation at Stanford University. Previously, he served as a professor of electrical engineering and computer science at the University of California, Berkeley, for over a decade. He specialises in computer graphics, human-computer interaction and visualisation. In an exclusive interaction with Analytics India Magazine, Agrawala explained the tech behind video editing tools, applications and challenges. Excerpts: AIM: Tell us about the technology behind your video editing software. How is it different from existing tools in the market? Maneesh Agrawala: Today, most tools force you as an editor to work at a frame level, so you have to find the exact frame you want to cut the video. It isn’t easy to think in terms of the frame at which things are happening and produce desired results. Our software converts the video file into high-quality transcripts. Once aligned, the user can directly click on a word to jump to the specific part of the video. Instead of scrubbing through the entire video, reading is much faster any day. This can be used for navigation. We have simplified it further, where the user can cut, copy and paste on the text to transfer or edit the underlying video. You can edit your video much like you are editing a text document. Initially, we only allowed cut, copy and paste options. Recently, we have introduced a new feature, where the user can type in new words into the transcript, and the person in the video will say those words. Coming to the working of the model, we write the code ourselves. The data comes in as a video (any formats). Once fed into the system, we take the audio portion and run speech-to-text to get the transcript. We have built a pipeline for taking the input text and have developed a model of the human face of the person saying those words. The neural model reconstructs the face and the lip motions that correspond to the new sounds and texts. In terms of the SOTA algorithms, we have used convolutional neural networks all over the place, integrated with face-tracking tools, which also comes with an underlying 3D head model. AIM: What are the use cases for your video editing software? Maneesh Agrawala: We are starting to explore healthcare, where we want to give people their voice back after they have had throat or laryngectomy surgery. In most cases, people can no longer speak with their natural voice after the surgery. They have to use an electrolarynx, a device that needs to be placed up your throat, and the vibration causes it to generate a robotic sounding voice. Here, we plan to record a patient’s voice before the surgery and then use that pre-surgery recording to convert their electrolarynx voice back into their pre-surgery voice. Besides this, we are interested in making tools that allow people to express themselves and create stories. It’s a significant component of human culture, and the kind of tools that we talked about earlier, in a way, support the creation of these kinds of video stories. AIM: Today, deep fakes are becoming next to impossible to detect. How can we put an end to this menace? Maneesh Agrawala: From a user standpoint, when you are using tools to manipulate videos, two things need to be considered — the audience should be made aware of the manipulated video, and consent of the actor in the manipulated videos is a must. The tools to detect deep fakes are not perfect. But we have algorithms and techniques to find those imperfections. For example, we have developed a tool that considers lip movement around certain phonemes, where we focus on the visemes associated with words having the sound M (mama), B (baba), or P (papa) in which the mouth must completely close in pronouncing these phonemes. However, emotions that forced the mouth to close often are not well reproduced by some tools. That can work to some extent, but they are not foolproof. In the long run, I do not believe that detection tools will work in any reasonable way. In the future, the tools for creating these deep fakes are just going to get better and better. They will get so good that it will be impossible to detect whether this was an original or a fake. Unfortunately, we do not have any technology that can determine if someone is lying. That said, we need to think about ways to mitigate the problem. We also need to think about other societal biases to prevent misuse and misinformation. AIM: Currently, there are limited open-source codes and libraries for video editing tools. How do you deal with this challenge in terms of collaboration within the ecosystem? Maneesh Agrawala: Ideally, we look at research papers and check if they sound and assess the results they can produce. I think some people do release their code today. But, we have chosen not to release them because of the potential harm. However, if researchers ask us, we are open to sharing the tools under certain limitations and understanding as we want to be careful about its implications.","excerpt":"Editing is an art. Bringing a story to life calls for tons of patience and hard work. But, what if we told you video editing could be as simple as text editing. Stanford University computer scientist Maneesh Agrawala, and his team has developed such a video editing software. Agrawala is the Forest Baskett professor of […]","categories":["AI Features"],"tags":["deep fakes","Interviews and Discussions","Machine Learning","stanford university"],"author_name":"Amit Naik","publish_date":"2021-06-01T18:00:00","publication_year":"2021","word_count":914,"keywords":["Go","programming_languages:R","AI","neural network","R","innovation","Machine Learning","programming_languages:Go","Aim","analytics","stanford university","deep fakes","Interviews and Discussions"],"extracted_tech_keywords":["AI","neural network","analytics","Aim","R","Go","innovation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/stanford-university-professor-maneesh-agrawala-on-video-editing-tools-deep-fakes-more\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":1646,"title":"Analytics India Companies Study 2012","content":"With the emergence of Analytics in its full force, the analytics providers in India are on a rise. By our estimate, there are upwards of 500 analytics (or services that heavily depend on analytics) in India currently. And the number is increasing at a significant pace. In this study, we have tried to capture where exactly the action lies. We studied these new age service providers and identified some key patterns which is the focus of this report. Please note that this study has not included captive analytics units, i.e. analytics teams setup by organizations for internal purposes.  The numbers and graphs are indicative in nature and the company’s logos are just a small sample of a much bigger population. Most of these companies have been IT service providers who have adopted analytics as one of their offerings. These include big IT vendors like Infosys, wipro and TCS. Yet, we have seen a significant rise in boutique analytics firms offering pure play analytics services. 30% of all analytics providers are based out of Bangalore, followed by Mumbai (16%) and Delhi (14%). 50% of all analytics providers falls into the industry of IT\/ ITES providers, followed by Consulting, Market Research and Media\/ Advertising (~10% each). 71% analytics providers have less than 50 employee strength, of which 27% has less than 10 employees. Average employee strength of analytics providers in India is 76 employees. Mumbai has the highest average employee strength of 106 employees among analytics providers. Download the complete report below. [attachments]","excerpt":"With the emergence of Analytics in its full force, the analytics providers in India are on a rise. By our estimate, there are upwards of 500 analytics (or services that heavily depend on analytics) in India currently. And the number is increasing at a significant pace. In this study, we have tried to capture where […]","categories":["AI Features"],"tags":[],"author_name":"Дарья","publish_date":"2012-10-12T20:28:41","publication_year":"2012","word_count":251,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","RAG","analytics","GAN","R"],"extracted_tech_keywords":["AI","analytics","RAG","R","Go","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/analytics-india-companies-study-2012\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10095058,"title":"Accenture to Invest USD 3 Billion in AI","content":"Accenture has announced that it will invest USD3 billion over three years in its Data & AI practice to help clients across all industries rapidly and responsibly advance and use AI to achieve greater growth, efficiency and resilience. The investments will be made in Generative AI, industry solutions, ventures, acquisitions, talent and ecosystem partnerships. The Data & AI practice will double its AI talent to 80,000 professionals through a mix of hiring, acquisitions and training.“There is unprecedented interest in all areas of AI, and the substantial investment we are making in our Data & AI practice will help our clients move from interest to action to value, and in a responsible way with clear business cases.” “Companies that build a strong foundation of AI by adopting and scaling it now, where the technology is mature and delivers clear value, will be better positioned to reinvent, compete and achieve new levels of performance. Our clients have complex environments, and at a time when the technology is changing rapidly, our deep understanding of ecosystem solutions allows us to help them navigate quickly and cost effectively to make smart decisions,” Julie Sweet, chair and CEO, Accenture, said. Besides, Accenture has also launched AI Navigator for Enterprise platform to guide AI strategy, use cases, rigorous business cases, decision-making and responsible policies; and Center for Advanced AI to help maximise value of generative and other AI. It will also create accelerators for data and AI readiness across 19 distinct industries as well as pre-built industry and functional models that take advantage of new generative AI capabilities.","excerpt":"Will double AI talent to 80,000 people through hiring, acquisitions and training","categories":["AI News"],"tags":["Accenture"],"author_name":"Pritam Bordoloi","publish_date":"2023-06-13T17:16:54","publication_year":"2023","word_count":261,"keywords":["Accenture","Go","API","programming_languages:R","AI","programming_languages:Go","generative AI","R"],"extracted_tech_keywords":["AI","generative AI","R","Go","API","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/accenture-to-invest-usd3-billion-in-ai\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10125602,"title":"Former BharatPe CPO Builds AI Doctor for 8 Billion People Worldwide","content":"Ankur Jain, the former chief product officer of BharatPe, left the fintech giant last year to venture into AI and healthcare. Together with G V Sanjay Reddy, the chairman of Reddy Ventures, he co-founded Jivi.ai, which happens to be Jain’s second stint in entrepreneurship. Jain, who has a background in AI and machine learning, founded Jivi in January 2024. Within months of incorporation, the startup announced its first model Jivi MedX, which outperformed popular models like Google’s Med-PaLM 2 and OpenAI’s GPT-4 on the Open Medical LLM Leaderboard. In an exclusive interaction with AIM, Jain revealed that the company plans to launch a series of models focussed on healthcare in the coming months. “While Jivi MedX is a text model, we are working on a series of other models that we internally call a model cluster. For example, there could be a different model that specialises in diabetes, and for ophthalmology, there will be a different model. “The next model from Jivi will be a vision model. We are working on a multimodal MedX,” Jain said. The vision Jain has with his new startup is to build an AI medical companion that the eight billion-strong population in the world can use for free. AI is finding use cases in other domains of healthcare such as drug discovery and genoming. However, Jain’s startup is focussed on primary healthcare. Not Just a Chatbot The startup’s aim is to create a product which will be useful to end users as well as doctors. Jain revealed that the AI medical companion he is building will have voice capabilities built in, which means everyone would be able to converse with the model in their own languages. “We will have voice capabilities and the models will eventually understand the top 25% of the world languages, which covers around 90% of the population. However, the technology is not there yet. Even OpenAI has not released the voice capabilities for GPT-4o,” the Stanford alumni pointed out. However, it’s just a matter of time, according to Jain, who comes from a family of doctors. “I think in six months, it will become mainstream. We are not building a chatbot, people will have to explain their symptoms to the model and for this, voice is the best medium.” Regulatory Challenges Jain’s endeavour is bold since healthcare is one of the most highly regulated industries in the world. Moreover, regulators around the world are contemplating regulating the technology. “For instance, we already have telemedicine regulation, whether you use AI in telemedicine or not, telemedicine rules will apply. In the end, the healthcare side will always supersede the technology, whether AI or non AI, we have to abide by the regulation,” Jain said. Nonetheless, Jain acknowledges that governments worldwide are contemplating regulations and changes are imminent. “We already adhere to standards like ISO 42,001, and we must comply with evolving regulatory frameworks, not only in India but globally. Similarly, in AI, especially in healthcare, we welcome and adhere to regulations. Regulations ensure that only serious players can enter the market, driving the delivery of quality products to consumers,” Jain added. Powered by Llama 3 Jain revealed that JiviMedX is powered by the Llama 3 8 billion and 70 billion parameter model. The business started by trying out the existing models, but none of them produced good enough results for the medical domain. “We tried Mistral, Llama 2 and a few other domain-specific models but none of them helped us in achieving the results we were looking for. The accuracy at best was around 55-60%, which was not good enough,” Jain revealed. To make the models more accurate, the startup experimented with various approaches initially, including direct preference optimisation (DPO) and reinforcement learning with human feedback (RHLF) frameworks but these methods did not yield the desired results. “So we settled with odds ratio preference optimisation (ORPO). Interestingly, not many people in the world have used it so far. We kept experimenting with it, akin to tuning hyperparameters. We conducted approximately 150 experiments to achieve the desired outcome,” Jain said. Currently, JiviMedX has an average score of 91.65% across Leaderboard’s nine benchmark categories. However, Jain adds that even 92% is not good enough and efforts are underway to increase the model’s accuracy level to 95%. While MedX uses Llama 3 for its intelligence, the other models Jivi is building will not necessarily be powered by Llama 3. Other models in the cluster could use a different model for intelligence depending on the use case. ‘My First Hire was a Doctor’ “We plan to launch a product in the coming months, but we need to be 100% sure that the model works and does not hallucinate. Currently, the medical community is evaluating it and we will launch it as soon as we get the green light,” Jain said. Interestingly, Jain revealed that even though it’s an AI startup, his first hire was a doctor. Currently, his team consists of 24 people, and nearly 40% of them are doctors. “We are involving doctors from Day 0 basically. Half of my team is full of in-house surgeons and physicians. We have also collaborated with the doctor’s community throughout. We built a prototype, shared it with the community and incorporated their feedback, that’s the process,” he added. Jivi’s Secret Sauce Jivi trained MedX using domain-specific data gathered from web scraping, along with information sourced from books, medical journals, research papers, and clinical notes. However, the data alone does not solve the problem. “An article about diabetes found on the web tends to be quite generic, typically targeting end users. In contrast, our analysis delves into the past 50 years of diabetic research data. We aggregate insights from various sources, compiling them into a comprehensive knowledge base. This is our secret sauce,” Jain pointed out. Another challenge the company faces is keeping the data up to date. For instance, if there is new research in diabetes, the startup has to ingest the data into the model. “So, we have to ingest every new development into the model every month. We plan to do this every week. Currently, we are ingesting once a month,” Jain said. Can AI Make Money? Creating an AI model is one thing; monetising it is a completely different challenge. AI requires huge investment, but building a sustainable and profitable business model remains a challenge. Jain plans to release the product for general consumers for free. However, the startup also plans to create a B2B segment where hospitals, pharma companies, and insurance providers will provide Jivi’s product to their end users as a software-as-a-service (SaaS) product. “Our business model involves selling premium services and also cross-selling. We can connect end users to diagnostic services, and we can also cross-sell e-commerce products like medical devices,” Jain said.","excerpt":"JiviMedX is powered by the Llama 3 8 billion and 70 billion parameter model","categories":["AI Features"],"tags":["Interviews and Discussions"],"author_name":"Pritam Bordoloi","publish_date":"2024-07-02T18:12:23","publication_year":"2024","word_count":1130,"keywords":["Go","API","machine learning","OpenAI","AI","GPT-4o","RAG","GPT","Aim","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","machine learning","GPT-4o","OpenAI","Aim","RAG","R","Go","API","GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/former-bharatpe-cpo-is-building-an-ai-doctor-for-8-billion-people\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10099594,"title":"UAE Unveils GPT-4&#8217;s New Rival, Falcon 180B","content":"Adding to the list of open-source LLMs, Abu Dhabi’s TII has released Falcon 180B, a highly scaled-up version of Falcon 40B. According to the official blog post, this is the largest open-source language model, boasting a staggering 180 billion parameters. Back in June, the institute released the three variants of Falcon – 1B, 7B, and 40B. It is trained on a dataset of 3.5 trillion tokens from TII’s RefinedWeb dataset, making it the longest single-epoch pretraining process for an openly accessible model. The training process involved the simultaneous use of up to 4096 GPUs, using Amazon The chat model was fine-tuned leveraging a combination of various extensive conversational datasets focused on chat and instructions. Imagine that your moat is money and you try to compete with state level funding of the UAE— Yam Peleg (@Yampeleg) September 6, 2023 Falcon 180B Vs Llama 2 Vs GPT 3.5 Currently topping the HuggingFace leaderboard, Falcon 180B surpasses Llama 2 in size by 2.5 times and utilises four times the computing power. It also outperforms when compared to Llama 2 70B and OpenAI’s GPT-3.5 in terms of MMLU. It also uses multi-query attention (MQA). Additionally, it exhibits comparable results to Google’s PaLM 2-Large in assessments involving HellaSwag, LAMBADA, WebQuestions, Winogrande, PIQA, ARC, BoolQ, CB, COPA, RTE, WiC, WSC, and ReCoRD. However it is yet to perform as well as GPT- 4. Read more: Llama 2 vs GPT-4 vs Claude-2 Commercial Use While the research paper for the model has not yet been released, Falcon 180b can be commercially used but under very restrictive conditions, excluding any “hosting use”, making it less commercially friendly than previous Falcon models. Open Source Gets a New Player Surpassing Meta’s Llama 2, this is the largest open-source language model. Even though Meta has been championing the open-sourcing ecosystem, it comes with its own set of restrictions like its complicated licensing policy. But even Meta seems to follow a closed-door approach with its upcoming models which are touted to be even bigger and better. However, meanwhile, no matter how much effort Meta puts in, the real controller of open source is OpenAI, as AIM reported earlier. But now with the release of Google’s Gemini getting closer, it is high time that OpenAI releases GPT-5 to stay ahead in the race. Read more: Meta Launches Open Source Models, OpenAI Controls Them","excerpt":"With 180 billion parameters, it is trained on dataset of 3.5 trillion tokens from TII’s RefinedWeb dataset.","categories":["AI News"],"tags":["Falcon 180B","GPT-4","uae"],"author_name":"Shritama Saha","publish_date":"2023-09-06T19:55:26","publication_year":"2023","word_count":390,"keywords":["Go","uae","GPT-5","AI","OpenAI","RPA","ML","Falcon 180B","GPT-4","RAG","GPT","Aim","R"],"extracted_tech_keywords":["AI","ML","GPT-5","OpenAI","Aim","RAG","R","Go","GPT","RPA"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/uaes-tii-unveils-falcon-180b-a-new-open-source-llm\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10168994,"title":"NVIDIA, Anthropic Clash Over Chip Export Controls: Report","content":"Anthropic, the AI startup that develops the Claude family of models, suggested reccomendations in a blog post on Wednesday to the United States government to strengthen controls on chip exports. Nonetheless, NVIDIA, the leading developer of AI chips and hardware, has firmly opposed Anthropic’s statements, according to CNBC. In the blog post, Anthropic outlined several recommendations for the U.S. Department of Commerce, arguing that enhanced export controls could strengthen a technological edge considered vital for the U.S. to compete with China. Anthropic recommends adjusting chip access tiers for countries, lowering no-license purchase limits to curb smuggling, and increasing funding, and resources for export control enforcement to enhance effectiveness. For instance, Anthropic has suggested decreasing the GPU export limit NVIDIA can make to certain nations without government authorisation, expressing concerns about potential smuggling. However, an NVIDIA spokesperson told CNBC that “American firms should focus on innovation and rise to the challenge, rather than tell tall tales that large, heavy, and sensitive electronics are somehow smuggled in ‘baby bumps’ or ‘alongside live lobsters.” The spokesperson also added that the U.S. cannot be successful in the AI race by manipulating regulators, stating that China is home to ‘half of the world’s AI researchers’ and has ‘highly capable’ experts in every aspect of the AI stack. NVIDIA CEO Jensen Huang also resonates with the sentiment, and recently said that China is not behind the U.S in the AI race. The above argument revolves around a rule implemented by former U.S. President Joe Biden in January – which comes into effect on May 15, 2025. It is called ‘The Framework for Artificial Intelligence Diffusion’, which categorises nations into three levels according to the security threats they pose to the U.S. These categories dictate the application of export controls. The first tier, including 17 countries and Taiwan, can get unlimited AI chips. Around 120 countries fall into the second tier, where chip access is limited. Exports are prohibited to the third group, consisting of countries like China, Russia, Iran, and North Korea. ​​Companies can import up to 1,700 GPUs, valued at around $40-50 million, without needing a licence, to countries classified in the ‘Tier 2’ bracket. However, larger imports, worth up to $1 billion, will require a licence review. Recently, Reuters reported that the current Trump administration is planning to remove the tier system to control exports. This could be replaced with a ‘global licensing regime with government-government agreements’, as per the sources cited by Reuters. The report also adds that the threshold of 1,700 NVIDIA GPUs could also be lowered to 500.","excerpt":"Besides, the Trump administration is planning to further strengthen chip controls, as per reports.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","Anthropic","NVIDIA"],"author_name":"Supreeth Koundinya","publish_date":"2025-05-01T23:55:03","publication_year":"2025","word_count":427,"keywords":["Anthropic","Go","funding","artificial intelligence","AI","innovation","ML","ViT","NVIDIA","R","AI (Artificial Intelligence)","startup"],"extracted_tech_keywords":["AI","artificial intelligence","ML","Anthropic","R","Go","ViT","innovation","startup","funding"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/nvidia-anthropic-clash-over-chip-export-controls-report\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":45242,"title":"Meet OpenSpiel, A Collection Of AI Training Tools For Video Games By DeepMind","content":"Reinforcement learning is responsible for most of the breakthroughs in the field of emerging technology. Beating famous Go players, mastering chess and even poker sounded like conceptual ideas only a few years ago but with the advent of RN, they have been converted into reality. RL is now almost single-handedly mastering the gaming space. Recently, researchers from Alphabet’s DeepMind introduced OpenSpiel, a reinforcement learning framework for games. It is basically a collection of environments and algorithms for research in general RL and search and planning in games. The Framework OpenSpiel is a framework for writing games and algorithms and evaluating them on a variety of benchmark games. It contains an implementation of more than 20 different sorts of games like matrix and gridworld, among others. The implementation of games, core APIs, as well as the algorithms, are written in C++ and are wrapped in Python language. Since the games have been written in C++, this makes the framework fast and memory-efficient for implementing basic algorithms. Currently, the framework is only tested on Linux and a subset of the library has been ported to Swift. In this framework, most of the learning algorithms written in Python use TensorFlow. OpenSpiel supports n-player (single and multi-agent) zero-sum, cooperative and general-sum, one-shot and sequential, strictly turn-taking and simultaneous-move, perfect and imperfect information games, as well as traditional multiagent environments such as (partially and fully-observable) grid worlds and social dilemmas. It also includes tools to analyse learning dynamics and other common evaluation metrics. There are several algorithms which are implemented within the OpenSpiel framework such as classical search algorithms which are minimax (and alpha-beta) search, and Monte Carlo tree search (MCTS). OpenSpiel includes some basic optimisation algorithms which are applied to games. There are also three traditional single-agent reinforcement learning algorithms and they are Deep Q-Networks (DQN), Advantage Actor-Critic (A2C) and Ephemeral Value Adjustments (EVA). OpenSpiel utilises α-Rank algorithm as a tool for visualisation and evaluation. α-Rank is an algorithm which leverages evolutionary game theory to rank AI agents interacting in multiplayer games. This framework includes two major design criteria, they are mentioned below: Simplicity: This framework provides codes which are readable, usable, extendable by non-experts in the programming languages and especially to researchers from potentially different fields. It provides reference implementations which are used to learn from and prototype with, rather than fully-optimised or high-performance code that would require additional assumptions or advanced language features Avoiding Dependencies: The OpenSpiel framework does not introduce dependencies as according to the researchers, dependencies can be problematic for long-term compatibility, maintenance, and easy to use. Purpose of This Framework Researchers have stated that the purpose of OpenSpiel is to promote general multiagent RL across different game types. This framework stresses heavy emphasis on learning. The researchers said, “We hope that OpenSpiel could have a similar effect on general RL in games as the Atari Learning Environment has had on single-agent RL.” Installation Below are the commands following which one can clone the repository and build OpenSpiel on Debian or Ubuntu Linux. sudo apt-get install git git clone https:\/\/github.com\/deepmind\/open_spiel.git cd open_spiel .\/install.sh # Install various dependencies pip3 install –upgrade -r requirements.txt # Install Python dependencies mkdir build cd build Outlook With the increasing demand for RL in the field of emerging technologies, DeepMind is not the only player for open-sourcing RL algorithms. Earlier, chip manufacturer Intel had also released RL Coach, an open-source framework for training and evaluating reinforcement learning agents.","excerpt":"Reinforcement learning is responsible for most of the breakthroughs in the field of emerging technology. Beating famous Go players, mastering chess and even poker sounded like conceptual ideas only a few years ago but with the advent of RN, they have been converted into reality. RL is now almost single-handedly mastering the gaming space. Recently, […]","categories":["AI Features"],"tags":["game theory","Reinforcement Learning","reinforcement learning framework"],"author_name":"Ambika Choudhury","publish_date":"2019-08-30T15:14:54","publication_year":"2019","word_count":572,"keywords":["Go","API","Reinforcement Learning","AI","reinforcement learning framework","Git","game theory","RAG","Python","C++","GitHub","TensorFlow","R"],"extracted_tech_keywords":["AI","TensorFlow","RAG","Python","R","Go","C++","Git","GitHub","API"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/meet-openspiel-a-collection-of-ai-training-tools-for-video-games-by-deepmind\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":16267,"title":"India &#038; Israel firms now collaborate on AI","content":"Img credit- REUTERS\/Ammar Awad India-Israel talk brings back an AI deal: Bangalore based Telrad Tech ties up with Zebra Medical Vision over artificial intelligence The historic visit of Indian Prime Minister Narendra Modi to Israel has got the two countries talking over various important issues—military ties, medical devices, technology, agriculture, water systems to name a few. Both Modi and Benjamin Netanyahu who bonded well on the pretext of sharing same culture and values, are looking forward to give the commercial ties deeper roots. Treating our technology quest amidst the various deals is the pact signed between the two nations, under which two leading firms from the both the countries have tied up over artificial intelligence. Telrad Tech, an affiliate of Bangalore based company Teleradiology Solutions and Zebra Medical Vision, an Israel based artificial intelligence firm have collaborated to bring cutting edge technologies in the field across India, Asia and Africa. The deal between the healthcare companies from across the nations is slated to bring newer solutions in AI. The introduction of artificial intelligence in radiology can bring auto and intelligent detection of cancer, strokes, fractures and other medical conditions, which is exactly what this pact focuses on. Having served healthcare systems by treating millions of patients globally over the years, Teleradiology is keen on the benefits that this collaboration would bring. It believes that the tie-up can bring Zebra’s cloud-based deep learning analytics engine to India, thereby improving their high quality radiology reporting. The ability to produce and deploy AI for radiology by Zebra is unique and the Indian firm believes that it could add value to its overall offering.  Zebra believes that the partnership would help in bridging the gap between demand and supply of radiology services hence improving the quality of care. By using state of the art deep learning technology, it would help the healthcare economies to make a leap in their businesses significantly. The analytics platform by Zebra allows the healthcare institutions to identify patients at risk of disease and offer improved and preventive treatments. This Kibbutz Shefayim headquartered company was founded in 2014.","excerpt":"India-Israel talk brings back an AI deal: Bangalore based Telrad Tech ties up with Zebra Medical Vision over artificial intelligence The historic visit of Indian Prime Minister Narendra Modi to Israel has got the two countries talking over various important issues—military ties, medical devices, technology, agriculture, water systems to name a few. Both Modi and […]","categories":["AI News"],"tags":["AI India","Artificial Intelligence India"],"author_name":"Srishti Deoras","publish_date":"2017-07-12T13:01:28","publication_year":"2017","word_count":348,"keywords":["Go","artificial intelligence","programming_languages:R","AI","Artificial Intelligence India","programming_languages:Go","deep learning","analytics","AI India","R","analytics platform"],"extracted_tech_keywords":["AI","artificial intelligence","deep learning","analytics","R","Go","analytics platform","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/india-israel-firms-now-collaborates-ai\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":2532,"title":"Interview &#8211; Venkatesh Juloori, Partner at Meritus Analytics","content":"[frame align=”left”] [\/frame][dropcap style=”1″ size=”2″]AIM[\/dropcap]Analytics India Magazine: What are some of the main tenets (philosophies, goals, attributes) of analytics approach and policies at your organization? [dropcap style=”1″ size=”2″]VJ[\/dropcap]Venkatesh Juloori: Our endeavor is to help our clients to set-up, use and embed analytics as a part of the marketing decision making process.  To enable this we provide End-to-end services in marketing analytics – Consulting, technology deployment and analytical framework development. We offer platforms that enable analytics on a SAAS (Software\/Analytics as a service model) AIM: Please brief us about some business solutions you work on and how you derive value out of it. VJ: Meritus is the Analytics and BI center of Excellence set up by WPP, we are an analytical hub for large CPG clients delivering services across 50+ countries. In the last 12 months alone, we have delivered 300+ brand\/market focused ROI analysis and recommendations; which easily makes us the leading player in Asia in marketing analytics space We have the expertise and analytical constructs to deliver every component of a typical marketing journey starting with NPD, Segmentation, Market Estimation & forecasting, Marketing Prioritization, and Key sales driver identification with a strong focus on better marketing ROI. With our experiential learning and robust frameworks, we are empowering client users to build complex analytical solutions on the fly with minimal new IT knowledge which we foresee as the new paradigm in this space Typical questions we answer include: Can I increase the price of my brand without significant impact on sales? Should I cut spend on bigger brand or a growing smaller brand? What will be the impact of macro-economic environment on category demand? Is my offline and online marketing efforts structured to maximize ROI? How to allocate budgets for my portfolio of brands to maximize ROI? What personalization rules should be used to maximize revenue from our e-commerce site? AIM: Would you like to share any example of an Insight that generated a huge positive impact for your clients? VJ: A large multi-national client wanted to reduce price by 5% to increase volumes and gain market share in a particular market. Our analysis showed the price reductions were not required for the immediate future. Client maintained the price and was able retain $ 7 million in the bottom-line. The key point to note is that the analytics just cost a miniscule fraction of the value the client generated from the decision. AIM: Where do you see the bulk of your business coming from? Do Indian organizations have the same affinity towards BI\/ Analytics as that of organizations from other regions? VJ: Our clients are mostly the top 50 of the fortune 500 with a global footprint. Most of our work is delivered for North America, Europe, Asia pacific and Middle East markets. In the last 2-3 years we have observed the appetite for BI \/Analytics steadily increasing in India. With multinational firms operating here, we expect them to bring in their best practices and adapting to Indian context. AIM: Do you think it’s possible to become too married to the data that comes out of analytics? Where do you draw the line? VJ: While analytics is about learning from the past and having a better future, it is also possible that the past may not repeat.  I guess this is the where the question about being too married to data comes up. Our view is that this is unlikely to happen as any decision, especially when driven by analytics is measureable.  So if an expected outcome does not happen there will be enough reason to question the analytics and make suitable course corrections. AIM: What are the most significant challenges you face being in the forefront of analytics space? VJ: The pace at which established analytical practices and approaches are metamorphosing the ability to keep pace is one aspect. Another aspect is the limitation of today’s technology in handling unstructured data.  For example how to understand, process and derive insights from social media? AIM: What do you suggest to new graduates aspiring to get into analytics space? VJ: A combination of numeric abilities, domain understanding and skills with data management technologies are critical for success.  What will make them sought after is in their ability to convert what analytics brings to meaningful and contextually useful directions to solve the business problem at hand. AIM: How do you see Analytics evolving today in the industry as a whole? What are the most important contemporary trends that you see emerging in the Analytics space across the globe? VJ: We see a few trends Analytics is moving from expanding out of the domain of the specialist.  Better tools and analytical applications are driving this trend.  Self-service Analytics is a reality today The ability to make sense of unstructured data is improving by the day With the amount of data available about individuals and their behavior online that targeting is increasing in sophistication [divider top=”1″] [spoiler title=”Biography of Venkatesh Juloori” open=”0″ style=”2″] Venkatesh is currently Partner at Meritus Analytics, A WPP group company where he leads business development for Meritus. Venkatesh has with over 15 years of multi-market experience in services delivery and sales. Venkatesh holds a Bachelor’s degree in economics and statistics from Osmania University and an MBA from Indian Institute of Foreign Trade, New Delhi.][\/spoiler]","excerpt":"[frame align=”left”] [\/frame][dropcap style=”1″ size=”2″]AIM[\/dropcap]Analytics India Magazine: What are some of the main tenets (philosophies, goals, attributes) of analytics approach and policies at your organization? [dropcap style=”1″ size=”2″]VJ[\/dropcap]Venkatesh Juloori: Our endeavor is to help our clients to set-up, use and embed analytics as a part of the marketing decision making process.  To enable this we […]","categories":["AI Features"],"tags":["Interviews and Discussions"],"author_name":"Дарья","publish_date":"2013-01-30T13:25:10","publication_year":"2013","word_count":884,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","Aim","analytics","GAN","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","analytics","Aim","R","Go","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/interview-venkatesh-juloori-partner-at-meritus-analytics\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10005885,"title":"Let’s Learn Dabl: A Python Tool for Data Analysis and ML Automation","content":"Prior to training a machine learning or deep learning model, it is important to cleanse, pre-process and analyse the dataset at hand. Processes like dealing with missing values, converting text data into numbers and so on are all part of the pre-processing phase. More often than not, these processes come across as being repetitive and monotonous. Although there are tools for automating this process, they behave like a black box and do not give intuition about how they changed the data. To overcome this problem, python introduced a library called dabl – Data Analysis Baseline library. Dabl can be used to automate many of the tasks that seem repetitive in the early stages of model development. This was developed quite recently and the latest version of Dabl was released earlier this year. The number of available features currently are less, but the development process is happening at a good pace at Dabl. In this article, we will use this tool for data pre-processing, visualisation and analysis as well as model development. Let’s get started. Data pre-processing To use dabl to perform data analysis we need to first install the package. You can install this using the pip command as pip install dabl Once the installation is done, let us go ahead and pick a dataset. I will select a sample dataset from Kaggle. You can click this link to download the data. I have chosen the diabetes dataset. It is a small dataset which will make it easy to understand how dabl works. After downloading the dataset, let us import the important libraries and look at our dataset. import numpy as np import dabl import pandas as pd db_data=pd.read_csv('diabetes.csv') db_data.head() Usually, after looking at the dataset you would get into the data cleaning process by trying to identify missing rows, identify the erroneous data and understand the datatypes of the columns.  These processes are made easy using dabl by automating these. db_clean = dabl.clean(db_data, verbose=1) We have a list of detected feature types for the dataset given. These types indicate the following. Continuous: This is the number of columns containing continuous values and columns with high cardinality. Dirty_float: Float variables that sometimes take string values are called dirty_float. Low_card_int: Columns that contain integers with low cardinality fall under this category. Categorical: This is the number of columns containing pandas categorical values in a string, integer or floating-point formats. Date: Columns with data in them. These are currently not handled by dabl. free_string: string data types which contain multiple unique values are labelled as free_string. Useless: Constant or integer values that do not match with any of the categories are given a name useless. For more information about the feature types it has identified you can do the following step. type_info = dabl.detect_types(db_clean) type_info Here, we can clearly see which column of the dataset is of which data type. We can also change the type to meet our needs and requirements. For example, the column named Pregnancies is labelled neither as continuous nor as categorical and since the values in the column are single integer values we can make them into categorical values. db_clean = dabl.clean(db_data, type_hints={\"Pregnancies\": \"categorical\"}) We have successfully converted the column into a categorical one. Data visualisation The next part before training of the model is to visualise the data. Using visualisation tools like matplotlib or seaborn is effective but dabl makes the process very simple and displays a wide range of plots for a single line of code. As dabl detects feature types and automatically cleans the data this makes analysing the data extremely fast. Using the plot() method you can plot all your features against the target. In our dataset, the column Outcome is the target. dabl.plot(db_clean, 'Outcome') Dabl first automatically identifies and drops any outliers present in the dataset. It then identifies what type of data is present in the target (whether is categorical or continuous) and then displays the appropriate graph. Since ours is a categorical target, the output is a bar graph containing the count of 0s and 1s. Dabl also calculates and displays Linear discriminant analysis scores for the training set. The next graph is to identify the distribution of each feature against our target. As you can see below, each feature is plotted as a histogram against our target and the number of features that lead to 1 and 0 are shown in orange and blue respectively. The next graph is the scatter plot of the different combinations of data that exists in the dataset. For example, the feature glucose will be plotted against all the other columns and the distribution is shown below. In order to increase the efficiency and speed of training, dabl automatically performs PCA on the dataset and also shows the distribution to us. The next graph is the discriminant PCA graph for all the features in the dataset. It also displays the variance and cumulative variance for the dataset. The final graph displayed here is the linear discriminant analysis which is done by combining the features against the target. It is clear that in a single line of code we are able to analyse the data in different ways that would usually be done in multiple steps and with code redundancy. But dabl is not repetitive and is an automated way to make data visualisation easy and simple to use. Model development Dabl intends to speed up the process of model training and provides a low code environment to train models. It takes up very little time and memory to train models using dabl. But, as mentioned earlier, this is a recently developed library and provides basic methods for machine learning training. Here I will be using a simple classifier model to train the diabetes dataset. classifier = dabl.SimpleClassifier(random_state=0) x = db_clean.drop('Outcome', axis=1) y = db_clean.Outcome classifier.fit(x, y) The simple classifier method performs training on the most commonly used classification models and produces accuracy, precision, recall and roc_auc for the data. Not only this, but it also identifies the best model giving the best results on your dataset and displays it. Similar to classification, you can also use a simple regressor model for regression type of problem. Conclusion Dabl offers ways of automating processes that otherwise take a lot of time and effort. Faster processing of data leads to faster model development and prototyping. Using Dabl not only makes data wrangling easier but also makes it efficient by saving a lot of memory. The documentation of dabl indicated that there are some useful features still to come, including model explainers and tools for enhanced model building.","excerpt":"In this article, we will use this tool for data pre-processing, visualisation and analysis as well as model development. Let’s get started.","categories":["Deep Tech"],"tags":["Automation","data preprocessing","regression analysis","Visualization"],"author_name":"Bhoomika Madhukar","publish_date":"2020-09-01T14:00:23","publication_year":"2020","word_count":1100,"keywords":["NumPy","machine learning","TPU","AI","data preprocessing","regression analysis","Automation","Python","Seaborn","deep learning","Visualization","Matplotlib","R","Pandas"],"extracted_tech_keywords":["AI","machine learning","deep learning","Pandas","NumPy","Matplotlib","Seaborn","TPU","Python","R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/lets-learn-dabl-a-python-tool-for-data-analysis-and-ml-automation\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":30435,"title":"This Startup Is Aiming At A Netflix-Like Personalisation In Food Industry","content":"The use of artificial intelligence and machine learning in the food tech sector is not new. We have in the past covered several such organisations — like startups using AI to suggest recipes based on available ingredients, or robots cooking from scratch. Spoonshot (earlier called Dishq) is another unconventional startup in the list that is using machine learning to understand and predict people’s taste. It is also helping the food and beverages industry with challenges such as recommendation, personalisation, menu & product development, among others. To get an exclusive sneak peek into the recent developments at the company, the reason for rebranding, their future plans and more, Analytics India Magazine got in touch with Kishan Vasani, one of the founding members of the Bengaluru-based startup which was started in 2015. Currently 20-people strong and adding more into the team, Vasani shared some interesting insights about the startup and their journey. A Quick Glance At Spoonshot And Its AI Solutions Spoonshot is nothing but an AI-powered food brain that predicts human taste and trends for consumers to be able to better consume food. Some of the core technology that they have built is around offering personalised food which helps the recipe sites, catering services and restaurant chains who have a digital platform to come up with better offerings. While many restaurants abroad have their own apps and websites to bring personalised experiences, India hasn’t seen something so unique. More recently they have also developed a taste intelligence platform initially for CPG companies to get forward-looking insights and help innovation and marketing teams with increasing product penetration and improve the efficiency of the product development process. The Story Behind Founding And Rebranding Vasani shares that when they first started, they were a company that was into building mobile apps and a personalized food discovery app for the Indian market. Dishq was a portmanteau of Dish and Ishq, which he believes made a lot of sense in India. “But when we pivoted 18 months ago, people started calling us Dishq (pronounced as dish-cue) which was funny, and it didn’t resonate with us.” This is when they decided to launch it in a completely different identity. “We introduced this completely new taste intelligence platform, and so we rebranded as Spoonshot (which comes from moonshot), he said. Explaining the idea behind the newly-christened name, Vasani shared that back in the 1970s, when the US wanted to put a man on the moon and they didn’t know how. We thought that it was a great metaphor as we had lots of lofty, ambitious and innovative ideas and we don’t always claim to know how we are going to do it,” he added. The D-Factor “As a young startup, we wanted to take up one thing at a time and we started with personalisation,” shares Vasani. As they dived deeper into the space to understand the market, they realised that to deliver personalisation, they needed to understand the taste preference and therefore be able to predict them. “The timing was perfect in terms of prevalence of machine learning and AI tools, and that proved to be a valid approach for us,” he shared. Talking about the competition, Vasani shares that from a core technology perspective and in terms of taste prediction, they do not have direct competitors. “But if you start looking at it from product perspective and personalisation engines, there are a couple of them around the globe. One such name is HALLA, which started earlier this year. If you look from taste intelligence perspective, there are a couple of companies doing similar things with a different approach. Tech Behind Spoonshot Vasani opines that the models are not as important as the datasets that play a significant role. “We make a lot of efforts in our company to differentiate datasets regarding various aspects. One is the chemical and biological aspects of food, where we create ontologies. We have this recipe database and custom datasets that are built in-house. From there we use machine learning and AI in our products to solve various problems at hand such as flavour, extracting meaning out of food and others,” he explained. Next is to monitor the food trends in the market and derive causality relations between various pieces of data. “The dataset that we have built answers a lot of these questions,” he added. Having said that they are at an early stage and there are a lot of interesting problems that remain unsolved. Tools: Vasani shares that most of them are standard ones. Most of them are standard ones. For hosting, computation and storage purposes, they use AWS which has got a lot of readymade solutions. For all their data science efforts, they use libraries such as TensorFlow, Keras, PyTorch, web frameworks such as Django Flask, computational frameworks such as Spark, and others. “All our systems are built on Python,” he shared. Use Cases Their customers are usually some form of digital food company such as cloud kitchens or food aggregators which depend on personalisation. For instance, HungerBox is one client which is digitising the food courts at tech park. Vasani explains that everyday employees order through their platform, so there is a high repeat rate from a large number of users. “Our personalisation helps users to go through the order quickly by cutting out steps in the conversion funnel because we understand what people want,” he said. He further shares that on the taste intelligence side, they are working with a couple of very large food businesses in the US. “The cases we are helping them on is in identifying new ingredient combinations for some of their core products which are worth tens of millions of dollars. They are looking for ways to innovate and be agile,” he shared. With their solutions, they are enabling companies to test more ingredient combinations than the traditional methods would normally do. They are enabling these companies to have new levels of confidence around product launches, which hadn’t been possible before. “Usually 50 percent of new products fail due to lack of understanding and lack of testing that. So, we are helping them both ways and there will be interesting case studies to talk about early in 2019,” he said. Growth Plans The startup which was founded back in July had raised $400,000 then. Now, they have big plans in terms of hiring and product innovation ahead of it. They are looking forward to finishing case studies with the companies and improving technology. “We are finishing off case studies and building taste intelligence platforms in the first quarter. Later in 2019, we will also build the food service use case whether it is McDonald’s or restaurant chains or catering food service companies. We would essentially help them understand trends and innovate on their menus,” he said. The startup also had an interesting development when it acquired analytics startup Brisky. Vasani shares, “what is really cool about Brisky is that they have solved a problem in quite a clever way. They were able to get more genuine feedback directly from consumers and their experience. “We see data synergies that Brisky has in terms of data and what our products do. Our competitors are focused on flavour but, flavour and taste are different. Flavour is just a part of overall taste experience and taste incorporates many other factors related to actual experiences beyond just the food as well.” “We have a company in our business that has a very good understanding of what kind of experiences people are having in restaurants which is really important to our taste intelligence platform to help improve insights. So, there are benefits to both entities to leverage each other’s data,” he said. Overcoming Challenges “Specifically talking about the Indian market, hiring is probably the number one pain point for startups in India,” he shares. “In particular, technical hiring is extremely difficult, and we have tried every approach. When everybody today wants to be a data scientist or claims to be a data scientist, validating it is really important,” he added. While they have built a rigorous framework to find great people, it still remains a challenge. “There is a huge opportunity in India for those in the recruitment field to really solve this problem and help this ecosystem with better recruitment and better training of candidates,” he said on a concluding note.","excerpt":"The use of artificial intelligence and machine learning in the food tech sector is not new. We have in the past covered several such organisations — like startups using AI to suggest recipes based on available ingredients, or robots cooking from scratch. Spoonshot (earlier called Dishq) is another unconventional startup in the list that is […]","categories":["AI Startups"],"tags":[],"author_name":"Srishti Deoras","publish_date":"2018-11-20T11:47:19","publication_year":"2018","word_count":1394,"keywords":["data science","machine learning","artificial intelligence","Keras","AI","PyTorch","RAG","Aim","analytics","TensorFlow"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","data science","analytics","Aim","TensorFlow","PyTorch","Keras","RAG"],"url":"https:\/\/analyticsindiamag.com\/ai-startups\/this-startup-is-aiming-at-a-netflix-like-personalisation-in-food-industry\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10010901,"title":"AFRS In India: The Tech Is Ready, But What About The People?","content":"Is India ready for the implementation of the Automated Facial Recognition System (AFRS)? The answer to this question is another question – “At what cost?” India’s law and order enforcement is a curious mix. At one end it is still governed by the colonial police law from 1861, at the other, it plans to implement the most advanced technology like AFRS across all agencies. This is a dangerous combination. The Personal Data Protection Bill 2019 is still under a joint parliamentary committee for review. There is a lack of the right laws to control who gets access and for what purpose. We’ve been working with facial recognition technologies long enough to understand the threat for potential misuse is very real. Before implementing any format of the facial recognition system, our country needs to safeguard its citizens. It needs to ensure that citizens feel safe and their personal space is not encroached upon. The average Indian is afraid of the policeman. Several cases are under-reported because people are as afraid of the police as they are of the criminal, sometimes even more so. The courts are notorious for delaying cases. Eventually, attaining justice through the right means is an arduous journey. Law enforcement in India is impeded by a huge population, rampant corruption and a lack of trust between the police and the general public. The stress because of excessive workload does not make things easier for the police force either. The constitution of India gives the states and territories the powers to maintain law and order through their police forces. Every state has its own police force, making it difficult to track criminals when they cross over to different states. Once deployed, AFRS will be the world’s largest facial recognition system. It will be a beneficial tool for the police forces, enabling it to track criminals across pan- India, using surveillance video from cameras placed at major tolls, railway stations, airports et al. For instance, recently, Rs. 250 crore was allocated from the Nirbhaya fund for installation of facial recognition cameras at 983 railway stations across the country. The NRCB, in the first few pages of the RPF, has stated that “the Automated Facial Recognition System (AFRS) would help in automatic identification and verification of persons from digital images, photos, digital sketches, video frames and video sources by comparison of selected facial features of the image from an already existing image database”. A Facial Recognition System is an excellent investigation enhancer for identification of criminals, missing children\/persons, unidentified dead bodies and unknown traced children\/persons.” What Is AFRS? Automated Facial Recognition System (AFRS), to be used by police pan-India will be issued by the National Crime Records Bureau (NCRB). It would be a mobile and web-based application hosted in Delhi to help in crime prevention and detection, and fast track document verification. The AFRS is supposed to be interlinked with other existing databases like Crime and Criminal Tracking Network & Systems (CCTNS): managed by NCRB, Integrated Criminal Justice System (ICJS), State-specific database systems and the Khoya-paya portal. AFRS will maintain an extensive database of photos and videos of peoples’ faces. In the second step, a new image of an unidentified person taken from CCTV footage will be compared to the database to find the correct match and identify the person. The system will add photos from newspapers, sketches et. al. and tag them for marks of identification like scars, tattoos and age. Using AI, it will be able to search with accuracy and speed to raise an alert if a blacklisted match is found. How Does AFRS Work? AFRS aims to provide a national-level searchable platform of facial images. The threat lies in the power that a facial recognition system provides to the ones with access. Deep learning-based technologies can provide real-time face detection that can quickly and accurately match images against massive databases. It can coordinate between several law-enforcement agencies of the country and create a database of criminals that can be assessed by anyone with clearance. The definition of ‘potential criminal’, again is something that’s not very clear when it comes to technology like facial recognition. In the Delhi riots, the Delhi Police used video footage to identify protestors’ faces using facial recognition technology. AFRS, once functional, will be deployed and utilised by both the State police, the Central forces and agencies working under the aegis of Central government. The NCRB states that the system will function as a repository of photographs of criminals present within the country. This would allow the policing agencies to detect crime patterns and understand the modus operandi of criminals across the states. Proceed With Caution India does not have a data protection law, yet. Law enforcement agencies need to have a high degree of discretion in the absence of these safeguards. There should be predefined, detailed processes to define and most importantly, limit the use of AFRS. Experts have voiced concern regarding the low accuracy of facial recognition algorithms when it comes to minorities, women and children. Before deployment, law enforcement agencies should understand its shortcomings and act accordingly. Image recognition can be an ineffective and erroneous task in the early stages of a non-trained algorithm. The algorithm needs to be trained with the right data before being deployed in a criminal justice system. Marginal and vulnerable groups are more often over-represented, which make them potential victims of false positives. A compulsory training program for everyone with access to AFRS needs to be developed. AFRS is a powerful tool. Due diligence is needed before it is made available to law enforcement across the country. Along with the technology, the people using it also should be ready.","excerpt":"Is India ready for the implementation of the Automated Facial Recognition System (AFRS)? The answer to this question is another question – “At what cost?”  India’s law and order enforcement is a curious mix. At one end it is still governed by the colonial police law from 1861, at the other, it plans to implement the […]","categories":["AI Features"],"tags":["Facial Recognition"],"author_name":"Kunal Kislay","publish_date":"2020-10-29T14:00:02","publication_year":"2020","word_count":943,"keywords":["Go","Facial Recognition","AWS","AI","image recognition","Git","RAG","Aim","deep learning","Rust","R"],"extracted_tech_keywords":["AI","deep learning","Aim","RAG","image recognition","AWS","R","Go","Rust","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/afrs-in-india-the-tech-is-ready-but-what-about-the-people\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":34980,"title":"Top 3 Tools And Applications Used For Data Mapping","content":"Data Mapping is one of the first steps used in data integration tasks. Building a data map will help the users to avoid any potential issues. In this article, we mentioned 5 well-known Data mapping tools and applications. 1.Python Record Linkage Toolkit With its abundant libraries and toolkits, Python offers a special package called Python Record Linkage Toolkit. The toolkit helps in record linking within or external data sources and provides maximum accessories needed for record linkage and deduplication. The toolkit was specially designed for analysis and the linking of small or average sized files. Inspired by the FEBRL project the toolkit has an advanced feature of data manipulation tools. However, this feature is not supported by FEBRL. This advanced feature is used to integrate record linkage directly with available data manipulation projects. Features: The key objective of the toolkit is to develop an extensible record linkage structure. The toolkit helps clean and regulate data with simple techniques, performing recording pairs with intelligent indexing methods, impair records with a great number of correlating and similarity measures for various types of variables such as strings, numbers and date. The toolkit has a number of supervised and unsupervised classified algorithms, boosted with record linkage evaluation with various built-in datasets. Record Linkage (R) The Record Linkage package is developed to promote the application of record linkage in R. The package emerged while using R for record linkage of streaming data. It provides an interpretation of various designs which lead to abundant availability of functions and data structures. Combination of these functions and data structures as an R package facilitates the application of record linkage techniques to different datasets. Features The tool builds comparison patterns by providing compare. dedup function for deduplication and compare.linkage function for linking two or more data sets together The ReLinkData class includes other components which help in the process of Data Linkage The package helps in blocking, which reduces the number of data pairs by focusing on specified patterns The kit supports the phonetic functions and string comparators which deal with typographical errors in character strings 2.FRIL(A Fine-Grained Record Integration And Linkage Tool) FRIL tool boosts the classical linkage tools with a loaded set of parameters. Users can systematically and iteratively explore the best combination of parameter values which improve linking performances with accuracy. The tool has the potential to boost the accuracy of data linkage throughout all the suggested record linkage. FRIL uses some algorithms which are user-controlled parameters that are naturally stored in common linkage tools such as Link King10, Link Plus 11 and many more. The tool includes the standard process of record mapping. Features Association of graphical tools for adapting schema discrepancy and for analyzing, validating and summarizing results Development of computerized learning tools to enable suggestion of natural parameters Implementing search methods namely, nested loop join(NLJ) and sorted neighbourhood method (SNM) for comparing small and average data files 3.Dedupe Dedupe is a Web API library which uses machine learning to implement de-duplication and entity resolution instantly on structured data. The library aids in removing duplicate entries from a spreadsheet of names and addresses. It links a list with user information to another list with organisational history without individual customer ids. Dedupe processes instruction data fed and drums up rules for the user dataset to facilitate a quick and automatic search for similar records with enormous databases. Features A machine learning technique which reads the human labelled data and naturally creates best weights and blocking rules Runs on personal computers and makes smart comparisons which don’t require the advanced server to run the tool as a library, this is possible as the library integrates to user applications Allows extensions by adding designed data types, string comparators and blocking rules 3. Remadder The application is built to automatically recognise identical records and eliminate the redundant data, which reduces the storage needs for files and backups considerably. The application helps people working with virtual machines or sharing large files with disorganised data across the servers by performing regular backups. Features Finds duplicate values by using records linkage and fuzzy match analysis Allows users to define how the duplicates and handle them Minimizes storage needs and operational costs","excerpt":"Data Mapping is one of the first steps used in data integration tasks. Building a data map will help the users to avoid any potential issues. In this article, we mentioned 5 well-known Data mapping tools and applications. 1.Python Record Linkage Toolkit With its abundant libraries and toolkits, Python offers a special package called Python […]","categories":["AI Trends"],"tags":["data mapping","entity resolution software","Python"],"author_name":"Bharat Adibhatla","publish_date":"2019-02-15T06:56:08","publication_year":"2019","word_count":700,"keywords":["Go","API","machine learning","programming_languages:R","AI","entity resolution software","RAG","Python","programming_languages:Python","GAN","data mapping","R"],"extracted_tech_keywords":["AI","machine learning","RAG","Python","R","Go","API","GAN","programming_languages:Python","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-3-tools-and-applications-used-for-data-mapping\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":66270,"title":"How To Find Success In Kaggle: What Do The Masters Recommend","content":"The challenges in the real world get more complex and competing than an online competition. Hackathon might not paint the exact picture, and the success at these competitions should not be mistaken for expertise at the industry level. However, Kaggle, one of the world’s finest platforms for data scientists, gives aspirants the best possible introduction into the tricky world of data. Analytics India Magazine has been exclusively covering the stories of top Kagglers, and today we compile a few nuggets of wisdom from those interviews that can guide an aspirant. Be Tenacious “A right proportion of hard work, dedication, persistence, never giving up attitude and luck are the most important ingredients that helped me,” said Abhishek Thakur when asked about his Kaggle success and what made him the world’s first 4x grandmaster. Abhishek inculcated a healthy diet of solving previous Kaggle competitions on his own, checking the successful solutions and getting to the bottom of the approaches with the help of Google. When asked about what it takes to get to the top, Darragh, a Kaggle grandmaster, recollecting Jermey Howard, said that the best practitioners in machine learning all share one particular trait in common; they’re very, very tenacious. Putting tenacity into practice requires investing more time in learning a new technology, getting good at it and then quickly moving onto new approaches. Another Kaggle master, Arthur Llau said that he had spent around 4-8 hours per day for over a month for the contests that fetched him gold. Arthur believes that being a top Kaggler is a full-time job. Do Not Reinvent The Wheel With all the advancements happening around AI, one might lose track of the kind of abundance that it has created in terms of tools. However, few traditional methods still hold their ground when it comes to crunching data, but many people fall for the fad — “NEW IS GOOD”. Arthur recommends to newcomers to develop the notion of exploring the data and finding what is not evident and not to hesitate to try classic methods. Whereas, another master, Mathurin, a Kaggle top 20 ranked master, asserts the need to try more and fail fast. He underlines the importance of reusing previous codes and learning about optimisation of metrics. Although every problem has its challenges, he recommends one to have a good scheme of cross-validation and confidence. Besides, he urges one to trust local results over results on the public leaderboard. Tri Duc, a Kaggle grandmaster, insists on the importance of knowing the fundamentals that involve mathematics. He strongly believes that mathematics helps one to get familiar with algorithms, which assists aspirants\/practitioners in preparing new concepts introduced in books or advanced courses. However, in a real-world project or Kaggle competition, observes Duc, the role of mathematics is rarely tangible, and one barely touches it while building ML pipelines. Don’t Fall In Love With The Tools The pace at which new tools get released, there is no doubt that the abilities of machine learning are sometimes blown out of proportion. The kind of attention it gets, sometimes clouds the ground truth and nudges people into believing that it is the holy grail of solutions. For instance, Abhishek uses TensorFlow for NLP problems and PyTorch for image problems. There is no dearth of libraries or frameworks one can use these days, and he firmly believes that it’s all good as long as one understands what is happening in the background. A common trait that can be seen across the top players is the flexibility in their approach towards problem-solving. They pick tools that do the job and do not waste time battling over languages and frameworks. For instance, Kaggle master Arthur prefers Python and sometimes C++ for doing operational research tasks. He switches between Keras and PyTorch framework while using a handful of very useful libraries like albumentations for image augmentation, eli5 and lofo for feature selection, Missingno and seaborn for visualisation, and Imblearn for imbalanced data. For parameters optimisation, Arthur prefers Optuna and skopt for the Bayesian module. Be An Eternal Student Although the democratisation of ML through new frameworks and libraries have made launching deep learning algorithms relatively straightforward, almost all the masters believe that it is of utmost importance that one has a grip on the fundamentals. Konstantin Yakovlev, a Kaggle Grand Master advises newcomers to read vociferously and recommends aspirants to be a lifelong student of the craft. For beginners, he recommends working on code writing ability with Python, R, etc. and developing a knack for drawing insights from results through visualisations. Mathurin has participated in more than 200 competitions where he mostly competed solo. However, he admits that all his top medals were won as a team. He believes that choosing a highly skilled partner assists in keeping the learning curve steep. As an industry insider, Mathurin says that building a machine learning pipeline at an organisation needs skills that extend beyond what Kaggle demands and warns newcomers not to mistake a Kaggle competition as the end goal. For job seekers, he stresses the importance of being consistently curious, which he believes to be more valuable than one’s Kaggle achievements because at the end of the day, companies are looking for those who are readily deployable with the least amount of training. Let’s conclude with a few thumb rules: read the forum carefully for discussion, re-read the top solution in similar contests in the pastread new papers to get ideasrun experiments and always perform K-fold to evaluate the gap between local validation and public leaderboard.","excerpt":"The challenges in the real world get more complex and competing than an online competition. Hackathon might not paint the exact picture, and the success at these competitions should not be mistaken for expertise at the industry level. However, Kaggle, one of the world’s finest platforms for data scientists, gives aspirants the best possible introduction […]","categories":["Deep Tech"],"tags":["best online data science masters","Kaggle","online masters analytics"],"author_name":"Ram Sagar","publish_date":"2020-06-01T12:00:16","publication_year":"2020","word_count":922,"keywords":["best online data science masters","machine learning","Kaggle","Keras","AI","PyTorch","ML","NLP","online masters analytics","Seaborn","deep learning","analytics","TensorFlow"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","NLP","analytics","TensorFlow","PyTorch","Keras","Seaborn"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/kaggle-competition-data-science-beginner\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10170717,"title":"Sarvam AI’s Backlash Exposes the Sad State of Indian AI","content":"When India’s most noted AI startup, Sarvam AI (Axonwise Pvt Ltd), released its latest LLM, it prompted discussions around the company’s approach and broader challenges facing India’s AI mission. One of the first companies to be selected under the IndiaAI Mission to build India’s sovereign foundational LLM, Sarvam AI recently released Sarvam-M, a 24-billion parameter open-weights hybrid language model built on top of Mistral Small. A breakthrough for Indic AI researchers to build use cases, the model supports 10 different Indian languages, including Hindi, Bengali, Gujarati, Kannada, and Malayalam, among others. But the tepid response—just 334 downloads in two days on Hugging Face—received some flak. Deedy Das, an investor at Menlo Ventures, called it “embarrassing,” and said there’s no real audience for this incremental work. His comments led to a heated debate among the Indian AI community. Das contrasted this with an open-source model, developed by two Korean college students, that garnered about 200k downloads. While Sarvam claims to be committed to building a foundational model, with many more releases in the pipeline, this early release of Sarvam-M, built on top of the French AI model, raised eyebrows. Sarvam is not alone. The government-backed BharatGen’s Param-1 model, which became available on AIKosh last week, has only received 12 downloads when writing this article. Is there a Merit in the Criticism? Das’s critique goes beyond model downloads. He argues that Sarvam’s effort is a reflection of misplaced ambition. “No one is asking for a slightly better 24B Indic model. Clearly,” he said. “If you want to train models, there should be a very good reason for it.” He further noted that there were cheaper and better models made by Google and TWO.ai that perform better in all these languages. “I have nothing against Sarvam, but I just don’t think that at this moment their contributions are remotely commensurate to their funding,” he added. According to the company, it has raised $41 million from leading investors like Lightspeed India Partners, Peak XV Partners, Lightspeed Venture Partners, and Khosla Ventures, among others. Its valuation as of March 2025 stands at $111 million, according to Tracxn. That said, several users on X pointed out that the model was good and had several use cases, but it also needed improvement. While appreciating the Bulbul TTS model of Sarvam, holding some reservations, Das suggested Sarvam work on fundamental, ground-level rethinking of the hardware and software stack, citing China’s DeepSeek as an example. “Focus on large-scale Indic and other general model data collection, which is driving the frontier models today,” he said. Nonetheless, according to a technical report, Sarvam-M surpassed Llama-4 Scout in performance and holds its ground against larger models such as Llama-3.3 70B and Gemma 3 27B. “We found the base Mistral Small model could be significantly improved in Indian languages,” reads the report. However, it experienced a minor decline of one percent in English knowledge evaluations like MMLU. The company took pride in the development. “Sarvam-M represents an important stepping stone on our journey to build Sovereign AI for India,” Vivek Raghavan, the co-founder of the company, said on X Also, Aashay Sachdeva from Sarvam AI defended the model on X, saying Sarvam-M achieved new benchmarks for Indian languages and pointed readers to the technical blog detailing the customisation and fine-tuning process. Sachdeva also posted a Google Sheet on X, in which he asked Sarvam-M’s Think model 7 questions from JEE Advanced 2025 in Hindi, and it answered all of them correctly. Clearly, the model is good for several use cases. Agree to Disagree Many appreciated Sarvam’s efforts and emphasised that innovation is not always about instant virality. However, one can argue that it was all about expectation mismatch and whether India’s bet on homegrown AI can deliver results that justify the hype. Much of the criticism comes from comparing Sarvam to OpenAI or DeepSeek, while the problem the company is trying to solve is fundamentally different. Harveen Singh Chadha from Sarvam AI, who was also critical when Krutrim launched its LLM earlier this year, said people critiqued the model without actually testing it. User @cneuralnetwork, who works at AI4Bharat, defended Sarvam’s work by focusing on the methodology: “The model is not the work—they process how they made the model is the work I loved the most. It sets the ground for other builders on how to post-train and possibly do even better.” Some even shared examples of people using Sarvam’s newly released model on Google Colab, and listed use cases where the model could potentially help farmers and the legal community, centred around solving for Bharat. Meanwhile, Kurain Benoy, a machine learning engineer at Sarvam AI, echoed a broader sentiment of optimism and national pride. On the other hand, Gaurav Aggarwal, VP and chief AI scientist at Reliance Jio, questioned the nationalism around Sarvam, pointing out it’s funded by Western Investors. Also read: Is Sarvam 2B Good Enough for Indian AI Developers? The Sad State of Indic AI Research India has 600 million smartphone users, and a considerable population prefer using Indic keyboards. For instance, AskDisha, the AI chatbot on IRCTC website built by CoRover.ai for booking tickets in Indic languages, is a classic case. Models like IndicTrans2 from AI4Bharat are built for the larger population of Bharat. The need to build AI for Bharat has been highlighted several times in the past. A large population of Indians are also first-time smartphone users, who use it in their native language as much as possible. This divide between English users and Bharatiya populations presents unique challenges in bridging the gap between technology and its diverse users. Many startups and researchers in India aim to solve for Bharat and Sarvam needs to highlight the use cases they aim for. Pratyush Choudhury, principal at Together Fund, explained that most people outside India don’t understand the challenges, and neither do they appreciate that compute is quite the invisible ceiling. “Plus, there’s a whole host of Indic-language use-cases where this sovereign model would work much better compared to using any other open-weights model,” he added, countering Das’ arguments. This shows the importance of making Indic AI models, a trend that started two years ago when developers were building on top of Llama for languages like Kannada, Tamil, and Malayalam. Raj Dabre, senior research scientist at Google and one of the creators of IndicTrans2, said, “Before Sarvam M came out, people were complaining about the lack of IndicLLMs. After Sarvam M came out, people are still complaining.” While Sarvam’s work on Indic languages is far more important than that of another wrapper startup, the debate surrounding Sarvam, judging merely based on downloads, indicates the development community is overlooking ‘building in India, for India’.","excerpt":"Much of Sarvam’s criticism comes from comparing it to OpenAI or DeepSeek, while the problem the company is trying to solve is fundamentally different.","categories":["AI Features"],"tags":["Sarvam"],"author_name":"Mohit Pandey","publish_date":"2025-05-25T10:00:00","publication_year":"2025","word_count":1114,"keywords":["Hugging Face","machine learning","OpenAI","AI","ML","Sarvam","Colab","RAG","Aim","Gemma 3","R"],"extracted_tech_keywords":["AI","machine learning","ML","OpenAI","Gemma 3","Aim","Hugging Face","Colab","RAG","R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/sarvam-ais-backlash-exposes-the-sad-state-of-indian-ai\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":2079,"title":"Are IoT projects really failing? How do we ensure their successful implementation?","content":"60 per cent of IoT initiatives stall at the Proof of Concept (PoC) stage The global networking giant Cisco recently conducted a study which shows that 60 per cent of IoT initiatives stall at the Proof of Concept (PoC) stage. “It’s not for lack of trying. But there are plenty of things we can do to get more projects out of pilot and to complete success and that’s what we’re here in London to do,” shared Rowan Trollope, Senior Vice President and General Manager, IoT and Applications, Cisco. This may come as a surprise to many, as the IoT ecosystem is gaining momentum globally. According to IDC’s prediction, the worldwide installed base of IoT endpoints will grow from 14.9 billion at the end of 2016 to more than 82 billion in 2025. In other words, IoT might soon be indispensable as a technology. The failure of these projects can be largely attributed to the fact that these projects depend on various human factors like culture, organization, and leadership; besides focusing on the technology aspect. Cisco’s study revealed that nearly 54 percent of the surveyed IT and business decision-makers believed that collaboration between IT and the business side was the top factor that affects the success of a project; while a 49 percent of the sample relied on a technology-focused culture, stemming from top-down leadership and executive sponsorship. This is not all, 48 percent of the sample claimed that IoT expertise, whether internal or through external partnership also affected the success of an IoT project. A good 60 percent also talked about the IoT initiatives generally looking good only on paper. Reasons why IoT projects fail Some IoT projects fail in the very initial stages As discussed earlier, most of the IoT initiatives fail at PoC stage. Cisco involved a sample of 1,800 IT and Business Decision Makers in the US and UK, for its study.  The organization reached the conclusion that generally in such cases, problems stem from company culture, organization and structure. There are several reasons why these projects fail increasingly. Let’s glance through Cisco’s list of the most prominent reasons why IoT projects fail: Delay in completion: An IoT project might also have to bear mid-project changes, which can also result in failure of such projects. Moreover, projects that take too long to complete usually involve more capital than usual. Such projects can’t be really considered successful. The ideal approach is to clearly define the project scope in advance, in order to deliver the IoT-based solution in a timely manner. Absence of integration & collaboration among various teams: There are several independent teams involved in an IoT project, and working separately on several interdependent components. These components include hardware, device software, protocol stack implementation, gateway systems, backend systems, end user application, analytics, and more. The teams must constantly work in perfect collaboration with each other, maintaining concise and precise communication on regular basis. However, many a times, the team collaboration doesn’t work perfectly, leading to an unsuccessful project. Lack of talent and expertise for implementation: IoT projects generally require high level of expertise on technologies that are very new in the market. Constant innovation is a part and parcel of such projects, as the field is still evolving drastically. This makes it all the more difficult to spot talent, skillset, and expertise for such projects. Quality of data can impact IoT projects: Data is central and key to the effective implementation of IoT projects. Poor quality of data collected is often the chief reason why IoT projects fail. Moreover, value is essentially created with exploitation of the data sensed from the field. Thus, it becomes absolutely crucial to define the data sets that are to be collected, besides guaranteeing efficient and secure transmission of this data. Spiraling budget: Budget is one of the most alarming concerns tied up with an IoT project. Often lack of budget turns out to be the reason why IoT projects fail. The fear of maintaining steady budget often deters companies from investing in small, smart devices with specific functions, or other similar IoT projects. Some of them even fail before they are initiated. Thus going over-budget can be identified as one of the chief reasons why IoT projects fail. Critical factors which helps in ensuring IoT projects deliver business value Antenna design critical for successful IoT device system performance There are several factors enterprises must concentrate above to ensure successful implementation of an IoT project. By taking care of these aspects, failures can surely be dealt with. The prominent ones are listed as follows: Organizational factors: Everyone has a definite role to play at every level of the organizational hierarchy. The project sponsors for such project must be prepared to defend the project business cases to the company’s management team. The sponsor must also have the authority to allocate the required budgets. Similarly, the Information Systems must reflect the capacity and expertise to fulfill the computing infrastructure requirements of the IoT projects. The IoT software vendors must showcase the capacity to fulfill their commitments towards the IoT projects. Technology factors: IoT projects involve the transmission of a copious amount of data. This requires the presence of a robust telecommunications capacity, likely in the form of a wide area network, which must exist or be acquired reasonably quickly. Moreover, the IoT data has to be stored in a data warehouse. Enterprises must also acquire the software needed to build, maintain, and operate a data warehouse. This is not all, it is also crucial to display and report the IoT data. The business intelligence and visual analytical software required to accomplish such tasks must also be acquired. Project Team factors: It’s advisable for project managers to administer such IoT projects, as they showcase sufficient knowledge about the business and sufficient experience as IT project managers. Enterprises must also assign subject matter experts (SMEs) to the IoT projects, who can explain, communicate and defend the project business cases to various stakeholders. These projects also seek software developers with sufficient software development experience, who can use business intelligence and visual analytics software development tools. However, the short supply of such individuals usually fails to meet the escalating demand. Data factors: Errors discovered from the data obtained from IoT projects often threaten the projects’ business cases. The IoT data must reflect a high level of data quality, in order to drive the required business value. Thus, enterprises need software and SME’s who can accelerate the identification and resolution of data quality and data integration issues. This will help in ensuring the success of your IoT project. “Aiming small” for success HIO Project is a development effort to create 3D expandable modular hardware platforms for building embedded systems and Internet of Things (IoT) devices Most of the organizations usually design a grand vision behind implementing IoT projects, however, the strategy in many cases get subjected to change. So, it’s usually advisable to “aim small,” at the beginning, and then scale up. Often project teams discover new service business models, which function more efficiently. They replace the existing model with newer one. Additionally, these IoT projects are mostly not end-to-end, thus involving both complexity and uniqueness. A singly organization can’t fix it on its own, it requires an ecosystem-based approach. Small, focused startups might have an important role to play here. Often, they furnish a unique set of competencies or intellectual property, which help addressing key business impediments. To sum it up, the most successful projects are a blend of large and small players, who work together towards a common goal in true collaborative mode.","excerpt":"The global networking giant Cisco recently conducted a study which shows that 60 per cent of IoT initiatives stall at the Proof of Concept (PoC) stage. “It’s not for lack of trying. But there are plenty of things we can do to get more projects out of pilot and to complete success and that’s what […]","categories":["IT Services"],"tags":["Business Intelligence India","data India","IoT India","SMEs India"],"author_name":"Дарья","publish_date":"2017-06-08T10:07:32","publication_year":"2017","word_count":1265,"keywords":["Business Intelligence India","Go","API","AI","IoT India","data India","data warehouse","Scala","Aim","analytics","data quality","GAN","R","SMEs India"],"extracted_tech_keywords":["AI","analytics","Aim","R","Go","Scala","API","data warehouse","data quality","GAN"],"url":"https:\/\/analyticsindiamag.com\/it-services\/iot-projects-really-failing-ensure-successful-implementation\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10073651,"title":"Centre asks Offshore IITs to Offer Courses in Data Science and AI","content":"The ministry of education has proposed offshore IIT campuses to offer undergraduate degree programmes in areas such as data science and artificial intelligence. Under consideration by the Centre, the proposal to have these courses was based on feedback from Indian embassies abroad. A survey report shows most of the universities in the target nations, have undergraduate programmes in conventional disciplines. “From the feedback shared by the ambassadors of the identified host nations, the most frequently mentioned disciplines are related to computer science or IT, data sciences, AI, machine learning or robotics, electrical, electronics, mining, metallurgy, petroleum and energy,” the report read. Further recommendations were made on the various modes of admission, including JEE, GATE, SAT, GRE, and JAM. “A JEE or JEE (Advanced) exclusively for offshore campuses can be conceived in the future if it’s economically and logistically viable,” said the report. The committee noted that academic programmes such as Bachelor of Technology (BTech) and Masters of Technology (MTech) would be named Bachelor of Science (BS) and Masters of Science (MS), which are commonly used for international degrees. The 17-member committee is led by IIT Council Standing Committee chairperson Dr K Radhakrishnan and also includes the directors of IIT Bombay, IIT Kharagpur, IIT Madras, IIT Delhi, IIT Kanpur, IIT Guwahati and IIT Dhanbad. Flexible joint faculty contracts were proposed identifying provisions for deputing faculty members from the existing IITs to the proposed institutes abroad in their formative years. The committee set up by the Centre for the global expansion of IITs has Indian missions abroad and identified UK, UAE, Egypt, Saudi Arabia, Qatar, Malaysia and Thailand as prospective locations for offshore campuses under IIT’s brand name.","excerpt":"The report also suggested the admission criteria, small student intake, and options for Indian IIT students to study abroad","categories":["AI News"],"tags":["Courses","Data Science","IIT"],"author_name":"Bhuvana Kamath","publish_date":"2022-08-26T13:07:02","publication_year":"2022","word_count":276,"keywords":["data science","artificial intelligence","machine learning","programming_languages:R","AI","RAG","Courses","ai_applications:robotics","Data Science","R","IIT"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","data science","RAG","R","programming_languages:R","ai_applications:robotics"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/centre-asks-offshore-iits-to-offer-courses-in-data-science-and-ai\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10046054,"title":"AI Has Helped Us In Improving The Performance Of Our Network: Prashant Gupta, Verizon India","content":"Verizon is one of the top wireless carriers in the US, with over 120 million subscribers. The telecommunication giant’s India team plays a critical role in developing new technologies, building robust systems, and carrying out initiatives that cater to its vast consumer base. Verizon India is an innovation hub with teams working across different verticals. Analytics India Magazine caught up with Prashant Gupta, Head of Solutions – South East Asia & India, Verizon Business Group, to understand more about the company. Excerpts: AIM: What are Verizon’s main offerings for the Indian market? Prashant Gupta: Verizon holds National Long Distance (NLD), International Long Distance (ILD), Internet Service Provider (ISP), and Virtual Network Operator (VNO) licenses. We deliver a host of advanced IP, data, voice, and security solutions to meet the evolving demands of large businesses and government agencies. We are leading the industry with our Network-as-a-Service (NaaS), and Software-Defined Networking Virtualized Network Services (SDN VNS) as part of the proposition. This month, we announced Advanced Secure Access Service Edge (SASE), an innovative solution that merges Software-Defined Wide Area Network (SD-WAN) capabilities with comprehensive network security services to create a unified, cloud-delivered service model that supports the secure access needs of today’s digital environment. We also work closely with multinational global companies in various verticals such as IT, ITES, BFSI, manufacturing, pharmaceutical, and e-commerce. AIM: How do you leverage AI and machine learning in your services? Prashant Gupta: Our objective is to leverage AI and machine learning to find additional and faster ways to improve the performance of our network. We use it to predict when we might experience an increase in data traffic, identify potential problems in the network and fix them before they happen, among others. Where in the past, we would rely on customer feedback, we now use predictive analytics algorithms to monitor 3GB of data per second from millions of network interfaces. AI allows us to predict hundreds of customer-impacting events before they happen and take steps to prevent them from occurring. Our algorithms first establish a “normal” behaviour that we expect from our networks, and then it identifies “unnatural” data and attempts to recognise these events. We also have an internal team that creates and improves machine learning systems for specific business problems by leveraging open source tools and our internal development processes. As often happens with Big Data projects, though, the insights prove helpful in ways other than those for which they were originally intended. AIM: Verizon recently released its 2021 report on Data Breach Investigations. What are the key takeaways? Prashant Gupta: This year, we analysed 79,635 incidents, of which 29,207 met our quality standards and 5,258 were confirmed data breaches, sampled from 88 countries around the world. DBIR 2021 revealed that many of the breaches in APAC were caused by financially motivated attackers phishing employees for creds and then using those stolen creds to gain access to mail accounts and web application servers. Phishing was present in 36% of breaches in our overall dataset, up from 25% last year. Furthermore, instances of misrepresentation are 15 times higher than previous year in integrity compromised due to the increase of business email compromises. There has been a decline of user devices (desktops and laptops) being compromised, despite the remote working scenario. This is expected, considering that breaches are moving towards Social and Web application vectors, and those are becoming more server-based, such as gathering credentials and using them against cloud-based email systems. AIM: How did the mobile security landscape evolve over time? Prashant Gupta: We witnessed a global shift in the way organisations operate as an impact of Covid-19, many of which ramped up their digital transformation agendas to meet the evolving requirements of both employees and customers. With the sudden increase in remote working and the spike in mobile device usage, the threat landscape changed, implying a greater need for organisations to hone in on mobile security to protect themselves and those they serve. Companies have taken to remote work and cloud services big time to keep things functioning in a “socially distanced” environment. A broader profile of business responsibilities is now managed on mobile devices. Mobile security is tougher to manage than desktop and laptop security; phones and tablets are easier to lose and steal, and are at greater risk when it comes to certain types of cyber-attacks that rely on an end user’s failure to notice small details (such as phishing attempts). Surprisingly, mobile security compromises were actually down as compared to any of the three previous years. However, the number is still high (23%), considering nearly a quarter of organisations can expect mobile device breaches. What can be said for certain is that mobile risk is already high and still growing. We saw an order-of-magnitude increase in purchases of our mobile device management (MDM) solution. Standing at number two, India is one of the largest telco markets globally, including the number of internet subscribers, making it one of the biggest consumers of data. In developing countries such as India, where people have access to international apps, the mobile threat scenario comes at par with the world’s. AIM: As the world moves towards 5G technology, there is a fear among the general public regarding the possibility of severe health and environmental implications. What do you have to say about it? Prashant Gupta: Many organisations such as World Health Organisation (WHO) and Public Health England (PHE) have internationally examined the implications of 5G, finding nothing that points to any negative impact on public health. 5G operates in non-ionising radiation, which means that it doesn’t have enough energy to cause any damage. Globally, the defence forces have been using this spectrum for many years now, and there has been no impact on health or the environment. Moreover, the industry body Cellular Operators Association of India (COAI) has confirmed that any concerns around the impact of 5G on health and the environment are not substantiated. The same was the outcome of recent court rulings on the petition filed. AIM: 5G with AI is anticipated to bring in the next revolution in business. What are your thoughts on this? Prashant Gupta: 5G and AI are two of the most revolutionary technologies of the digital era. Combining the two technologies can truly transform the way we experience networks. While 5G can help us collect information, AI can use that data and create business intelligence to increase revenues. 5G’s faster speeds, lower latency, and greater capacity will support a much larger number of wireless devices than ever before. In turn, this is expected to help bring about Industry 4.0 and the true Internet of Things era. While individually, 5G and AI are game-changing technologies; together, they can be game changers. With scalable bandwidth and remote computing power, 5G can facilitate better data collection, while the proliferation of AI should harness this information to generate business intelligence and increased revenues. 5G and AI will also be invaluable for enhancing customer experiences. With increasing number of internet users, the adoption of smart devices, and advances in technology across retail chains, both technologies can help retailers digitally connect the shop floor to collect data and improve the customer shopping experience with emerging technologies such as holograms and virtual reality. Cybersecurity risks can also be mitigated through the combination of 5G and AI — a challenge that will inevitably grow in prevalence as things become more internet-connected. AI can autonomously detect threats and neutralise an attack instantly, helping manage the increasing threat from hackers that would be hard, if not impossible, to manage by human effort alone. AIM: What are the challenges and the opportunities in India’s telecom sector? Prashant Gupta: India’s telecom sector is full of potential, being one of the biggest and fastest-growing telecom markets. Mobile phones have reached villages, and with the Government of India’s efforts with the Digital India Program, people in rural areas now have access to high-speed internet. This development presents many opportunities for growth but, at the same time, puts mobile security at risk. India also has a huge potential for manufacturing 5G enabled endpoints, with few companies already setting up their manufacturing facilities in India. The advent of 5G wireless communications that India is preparing for constitutes a new era of network connectivity that will revolutionise many aspects of commerce and people’s lives. 5G, on paper, can reach up to 10-20 GBPS – some of the applications of 5G are autonomous vehicles, remote surgery, connected communities, automation, edge computing, etc. However, it is also important to note that with new technology comes the need for new security measures. Ensuring security of the 5G network will be of utmost importance.","excerpt":"We deliver a host of advanced IP, data, voice, and security solutions to meet the evolving demands of large businesses and government agencies","categories":["AI Features"],"tags":["high performance computing","Interviews and Discussions","verizon"],"author_name":"Shraddha Goled","publish_date":"2021-08-16T13:00:00","publication_year":"2021","word_count":1444,"keywords":["Go","machine learning","AI","R","Scala","RAG","Aim","high performance computing","analytics","edge computing","verizon","predictive analytics","Interviews and Discussions"],"extracted_tech_keywords":["AI","machine learning","analytics","Aim","RAG","predictive analytics","edge computing","R","Go","Scala"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/ai-has-helped-us-in-improving-the-performance-of-our-network-prashant-gupta-verizon-india\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10091847,"title":"Developers Love ChatGPT, But Yearn for CopilotX","content":"From sci-fi dreams to real-life implications, the evolution of AI has become a global sensation. Its potential to boost the global GDP by 7% is huge, and yet its impact on over 300 million jobs in major economies is equally threatening. But amid the fear of repercussions, developers remain unfazed. Recently, Chris Albon, director of ML at Wikimedia (the nonprofit host of Wikipedia) took to Twitter to share that ChatGPT makes his coding faster. However, he admits that ChatGPT is bad at “big things” like architecture or the product’s concept or goals but can give a “a working\/near-working solution in one second”. Senior ML engineer and teacher Santiago weighed in his opinions along similar lines stating that ChatGPT enhances his productivity and saves time. Satya Mallick, the chief executive officer at OpenCV, an open-source computer vision and machine learning software library, told AIM, “We should look at AI as a collaborator instead of a competitor that humans can leverage to be more productive” Read more: LLMs Can Now Self-Debug. Should Developers Be Worried? ChatGPT Over Copilot? Autocoder GitHub Copilot has been a fan favourite since its release last year. But with a chat-based interface coming into picture, developers are shifting to ChatGPT. Although Copilot can offer suggestions for lines of code, variables, and function names and can be integrated with the IDE and assists with small code snippets, Copilot has its limitations, such as producing inaccurate or inefficient code proposals and being inadequate for complicated programming duties that demand extensive knowledge and expertise. On the other hand, ChatGPT is easier for generating code answers and clarifying code ideas as it has a conversational interface. Not just newbies, ChatGPT increases efficiency of tech veterans too, as echoed by iOS engineer Nathan Chan. Additionally, it can aid in the creation of function and unit tests to improve test-driven development techniques. But it is not fitting for time-sensitive applications such as gaming or trading. Nonetheless, GitHub is set to upgrade its Copilot by introducing CopilotX that is built on GPT-4 just like ChatGPT and will provide an experience similar to ChatGPT right within the VS Code IDE. In addition to suggesting code, CopilotX offers comprehensive code analysis, suggests bug fixes, and identifies the code entered by the programmer. Furthermore, GitHub is developing a voice-to-code model capable of transforming spoken language instructions into code. Read more: GPT-4 Comes To GitHub With Copilot X LLMs Can Now Self-Debug Besides generating codes, LLMs can now also debug on their own with ease, which has been considered to be a long-standing obstacle. According to a recent Google Brain publication, LLMs not only generate codes but also possess the ability to self-debug their codes. This has resulted in a nearly 10% improvement in accuracy. LLMs’ natural language processing capability allows them to identify and rectify mistakes in their work. By implementing this method, LLMs have achieved optimal performance across various code generation evaluations, producing outputs of greater value compared to those without self-debugging. Debugging is a major obstacle in the workflow, and improving self-debugging capabilities is crucial for enhancing LLM coding. Integrating autocoding platforms with other AI approaches, such as constitutional AI, Reflexion, and AutoGPT, can significantly enhance productivity. ChatGPT is Killing Stack Overflow Prior to the arrival of ChatGPT, programmers depended on Stack Overflow to seek solutions from the online developer community. However, in December of 2022, ChatGPT was barred from Stack Overflow citing falsity over the AI-generated responses, resulting in a 12% reduction in the total number of website visits in December. The number of visits decreased from 279 million in November 2022 to 247 million by the end of December. But when AIM reached out to Stack Overflow to understand about the drop, they dismissed the impact of ChatGPT on traffic. “Like many other technology sites during the holidays, the month of December often brings a seasonal shift as users (many of whom are professional developers) take time off to be with their friends and families. Our public platform continues to serve 100 million visitors every month, making it one of the most popular websites in the world,” Stack Overflow told AIM.","excerpt":"LLMs are going to revolutionise the way we code. And this is just the beginning","categories":["IT Services"],"tags":["ChatGPT","GitHub","GPT","Microsoft","OpenAI","Stack Overflow"],"author_name":"Shritama Saha","publish_date":"2023-04-19T19:12:00","publication_year":"2023","word_count":687,"keywords":["ChatGPT","machine learning","TPU","OpenAI","AI","ML","Stack Overflow","computer vision","GPT","OpenCV","Aim","RAG","GitHub","Microsoft","AutoGPT"],"extracted_tech_keywords":["AI","machine learning","ML","computer vision","ChatGPT","AutoGPT","Aim","OpenCV","RAG","TPU"],"url":"https:\/\/analyticsindiamag.com\/it-services\/developers-love-chatgpt-but-yearn-for-copilotx\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10078338,"title":"More Generative AI Tools, More APIs, More Problems","content":"A lot has transpired within a span of just a few months – that’s the speed of AI for you. From tech that seemed gimmicky at first, generative AI tools have made AI infrastructure more accessible. What seemed far-flung from mainstream society is now cheap, fast and easy to build for developers across the spectrum. Several startups have either cropped up or pivoted to building platforms anchored around these open-source and closed-source models from companies like OpenAI. Survey conducted to judge the importance of APIs, Source: DEVOPSdigest Rise in APIs Important LLMs like OpenAI’s GPT-3 and other foundational models like Stable Diffusion have been made commercially available via API across applications. As the need for connected software increases, APIs have become ubiquitous. APIs first came about in the early 2000s when companies like Salesforce, Amazon and eBay developed their own APIs for their developers to access. Up until a few years ago the usage of APIs escalated as the world became more digital. What we are witnessing now is another wave of AI applications, which will result in SaaS companies being focused around generative software, pushing the application of APIs over the edge. The interface of Stable Diffusion available with the option to run with API API-related attacks By the second half of the 2010s, MIT research fellow Marshall Van Alstyne published work around business platforms and how companies that used APIs showed 12.7% more growth in their market capitalisation within a four-year time period as compared to companies that did not adopt APIs. But as the possibilities grew, so did the threats related to API abuse. In 2019, Gartner predicted in a report that API hacks would eventually become the most commonly found type of cyberattacks by 2022. The prediction has unfortunately come true. Salt Security conducted a survey among 200 enterprise security officials and concluded that 91% of companies reported API-related security issues last year. The study stated that around 56% of organisations experienced between 10 and 55 attacks each month while 22% faced between 51 to 200 API-related attacks each month. The report titled ‘Salt’s State of API Security Q1’ also mentioned that malicious API calls increased on a monthly per-customer basis from 2.73 million in December 2020 to 21.32 million in December 2021. API protection platform Salt has Web Application Firewalls that almost every API gateway was able to cross. Best automated API testing tools in 2022, Source: Katalon Need for API security A more recent survey, by Radware, noted that organisations are now grappling with sustaining security across platforms. According to the survey 40% reported that more than half of their company was vulnerable to attacks through third-party APIs. But an API is only as good as its security. The danger of these increasing number of attacks indicate that companies that see the value behind APIs must also see the value behind adopting API management platforms. The recent history of API attacks on enterprises attest to this. In April, the Microsoft 365 Defender Threat Intelligence Team revealed that they had discovered a ‘low volume’ of attempts to hack its cloud services via Spring4Shell, an application framework for Java. Source: Salt Security report Last year, LinkedIn suffered from a data breach that exposed over 92% of user profiles including their full names, email addresses and passwords. Investigation showed that the hacker had breached LinkedIn’s database through their open authentication-free developer API and scraped through the LinkedIn database, which eventually landed up for sale on the dark web. The data breach questioned how endangered social media platforms were owing to a failure to check security of third-party vendors. Mobile payment service Venmo exposed details of over 200 million subscribers via its API. The PayPal-owned payment application had made their data accessible because they offered a public API that was set as its default. This allowed hackers to download the data containing the names of senders, descriptions of the transaction memos and the transaction values. HubSpot, another prominent CRM tool, reported a data breach in mid-March that affected more than 1.6 million users, emails and associated contact numbers of accounts of the company’s internal customer support portal. HubSpot later revealed that its internal systems had been compromised and a portion of its internal systems accessed. Moreover, a few customer accounts in the cryptocurrency industry, including NYDIG, Swan, and BlockFi, were tampered with by an insider.","excerpt":"According to the survey, 40% of the respondents reported that over half of their company was vulnerable to attacks through third-party APIs","categories":["IT Services"],"tags":[],"author_name":"Poulomi Chatterjee","publish_date":"2022-10-31T14:00:00","publication_year":"2022","word_count":727,"keywords":["Go","API","OpenAI","AI","Scala","Git","generative AI","DevOps","R","Java"],"extracted_tech_keywords":["AI","generative AI","OpenAI","R","Go","Java","Scala","Git","DevOps","API"],"url":"https:\/\/analyticsindiamag.com\/it-services\/more-generative-ai-tools-more-apis-more-problems\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":6,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10056912,"title":"Interesting Algorithms Released By Google AI In 2021","content":"Google AI aims to apply AI to products and domains that will make AI accessible to all. To fulfil this mission, the tech giant conducts cutting-edge research to bring out innovations helpful to society. This year, too, we saw many such models and algorithms from Google. Though not possible to focus on all, let us take a look at some of the interesting innovations that came from Google AI this year. Wikipedia-Based Image Text (WIT) Dataset In September, Google released the Wikipedia-Based Image Text (WIT) Dataset. It is a large multimodal dataset created by extracting multiple different text selections associated with an image from Wikipedia articles and Wikimedia image links. Google says this was followed by rigorous filtering to only retain high-quality image-text sets. The final result, as a set, has 37.5 million entity-rich image-text examples with 11.5 million unique images across 108 languages. WIT wants to create a large dataset without compromising on the quality or coverage of concepts. Google says this is the reason they moved focus to the largest online encyclopaedia available today – Wikipedia. Image: Google For more details, click here. GoEmotions Dataset The tech biggie came out with GoEmotions, a human-annotated dataset of 58,000 Reddit comments extracted from popular English-language subreddits and labelled with 27 emotion categories. These include 12 positive, 11 negative, and 4 ambiguous emotion categories and 1 “neutral” category, keeping in mind psychology and data applicability. Google said that the GoEmotions taxonomy wants to give the greatest coverage of the emotions expressed in Reddit data and the best coverage of types of emotional expressions. For more details, click here. Indian Language Transliterations in Google Maps In Google Maps, the names of most Indian places of interest (POIs) in Google Maps are not generally available in the native scripts of the languages of India. For the majority of times, they are in English or may be combined with acronyms based on the Latin script and Indian language words and names. To solve this issue, Google came out with an ensemble of learned models to transliterate names of Latin script POIs into ten languages prominent in India. These include Hindi, Bangla, Marathi, Telugu, Tamil, Gujarati, Kannada, Malayalam, Punjabi, and Odia. Google said that with this ensemble, it has added names in these languages to millions of POIs in India, increasing the coverage nearly twenty-fold in some languages. For more details, click here. MetNet-2 -12-hour Precipitation Forecasting In another achievement in the climate space, Google came out with Meteorological Neural Network 2 (MetNet-2) for 12-hour precipitation forecasting. It uses deep learning methods for forecasting by learning to predict directly from observed data. Image: Google It added that the computations are faster than physics-based techniques. While its predecessor, MetNet, released last year, provided eight-hour forecasting, MetNet-2 takes it up a notch higher with 12-hour precipitation forecasting. For more details, click here. FLAN Model The Fine-tuned LAnguage Net (FLAN) model from Google explores a simple technique called instruction fine-tuning. This NLP model is fine-tuned on a large set of varied instructions that use a simple and intuitive description of the task. FLAN uses templates to transform existing datasets into an instructional format instead of creating a dataset of instructions from scratch to fine-tune the model. Image: Google For more details, click here. Generalist Language Model (GLaM) Google AI came out with the Generalist Language Model (GLaM), a trillion weight model that uses sparsity. The full version of GLaM has a whopping 1.2T total parameters across 64 experts per mixture of experts (MoE) layer with 32 MoE layers in total. But, it only activates a subnetwork of 97B (8% of 1.2T) parameters per token prediction during inference. GLaM’s performance compares favourably to GPT-3 (175B), with significantly improved learning efficiency across 29 public NLP benchmarks in seven categories. This spreads over language completion, open-domain question answering, and natural language inference tasks. Image: Google As for the Megatron-Turing model, GLaM is on-par on the seven respective tasks if using a 5% margin while using 5x less computation during inference. For more details, click here.","excerpt":"Let us take a look at some of the interesting innovations that came from Google AI this year","categories":["Global Tech"],"tags":["AI (Artificial Intelligence)","Data Science","GLaM","Google","Machine Learning"],"author_name":"Sreejani Bhattacharyya","publish_date":"2021-12-23T11:00:00","publication_year":"2021","word_count":674,"keywords":["Go","AI","neural network","R","innovation","Machine Learning","RAG","NLP","GPT","Aim","deep learning","Google","Data Science","GLaM","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","deep learning","neural network","NLP","Aim","RAG","R","Go","GPT","innovation"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/interesting-algorithms-released-by-google-ai-in-2021\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":61776,"title":"Power BI Service Introduces A ‘New Look’ For Workspaces","content":"In a recent Microsoft blog post, it has been announced that the company is introducing a ‘new look’ of workspaces in the PowerBI service. “To modernise and simplify the user experience in the Power BI service, last year, we had launched a public preview of the new look feature,” stated senior product manager, Nikhil Gaekwad in the blog post. That look majorly focused on improving the overall viewing experience. Some of the features included were, vertical page navigation, surfacing additional metadata, and simplifying the action bar. However, “since then, we’ve continued to evolve the new look experience based on feedback and made incremental updates to validate our design. And, today, I’m excited to announce that we’re bringing the modern look and feel of the new look to a place where most of us spend time collaborating and managing content – workspaces,” stated Gaekwad. According to the post, all the workspaces, including the classic and the new ones will receive the modern new look, provided the ‘new look’ toggle has been turned on the dashboard. Apart from the new look features include: It has been made easier for users to add content to their workspace. With the ‘+’ New button, users can access to connect to data, open files, and create new reports, dashboards, and more.An entry point has been made on the top command bar to switch between the list and lineage view. This entry point will help users to see the connections between data flows, data sets, reports and dashboards and their connections to other data sources.With the help of the new search box, users can perform a search across all content within their workspaces.The original tabs have been restructured to give users a flat list of content like SharePoint. This will create less confusion for users. With the new Filters tab, users can quickly filter the content to search within workspaces with hundreds of artefacts.The modern look is providing quick actions for your content that will be available on hover.Apart from these, the new look has a lighter colour theme and updated icons based on the Microsoft Fluent design system. Here’s the new tab order: All – Shows all content (dashboards, reports, workbooks, paginated reports, datasets, dataflows) within the workspace.Content – Shows all content created for consumption (dashboards, reports, workbooks, paginated reports) within the workspace.Datasets + dataflows – Shows all the datasets and data flows within the workspace for easy data management. In a recent LinkedIn post, Gaekwad confirmed the news by stating — Collaborating and managing content in the #PowerBI service just got a whole lot easier. I’m excited to announce that we’re bringing the modern look and feel of the ‘new look’ with a ton of usability improvements to workspaces! According to the blog post, all Power BI Service users will be able to opt-in to the new look of workspaces using the existing ‘new look’ toggle on the dashboard.","excerpt":"In a recent Microsoft blog post, it has been announced that the company is introducing a ‘new look’ of workspaces in the PowerBI service.  “To modernise and simplify the user experience in the Power BI service, last year, we had launched a public preview of the new look feature,” stated senior product manager, Nikhil Gaekwad […]","categories":["AI News"],"tags":["Power BI","what is power bi"],"author_name":"Sejuti Das","publish_date":"2020-04-15T14:30:00","publication_year":"2020","word_count":484,"keywords":["Go","Power BI","what is power bi","programming_languages:R","AI","programming_languages:Go","R"],"extracted_tech_keywords":["AI","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/power-bi-service-introduces-a-new-look-for-workspaces\/","complexity_score":2,"technical_depth":5,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10102266,"title":"Fret Not, Startups: Every OpenAI Update Can Be Your Next Stepping Stone","content":"This is increasingly getting inevitable. Whenever OpenAI releases new features to ChatGPT, it gets blamed for affecting startups built around similar functionalities. OpenAI recently introduced two new features to ChatGPT, namely ‘Upload many types of documents’ and ‘Use tools without switching’. This new ‘multimodal’ update, according to many, is expected to kill hundreds of startups. Some of the popular names include ChatPDF, AskYourPDF, and PDF.ai, and many more. With the latest update, ChatGPT now not only reads PDFs, but also supports a variety of document types within the same conversation, including PDFs, images, CSVs, and more. Previously, users were limited to uploading images in the default mode, but now they can seamlessly upload documents and immediately begin asking questions, expanding the platform’s versatility and usability. Moreover, users are no longer required to specify the ChatGPT mode they want to use. Browsing, Advanced Data Analysis (formerly known as Code Interpreter), and DALL-E 3 will all be available within the same conversation. GPT will determine the appropriate time to activate each mode, such as invoking DALL-E when the user intends to create an image. PDF.ai’s founder Damon Chen was one of the first ones to react, and humorously wrote on X: “Last night, I had a conversation with my wife on ChatGPT news. I asked, ‘What if PDF.ai doesn’t succeed in the end?’ To which, she casually responded, ‘It’s just one project. Start a new one’.” Not all is lost for GenAI startups Despite the development, Chen is still optimistic about PDF.ai. “I don’t think ChatGPT will ever implement small PDF-related features that customers desperately ask for.” However, he believes that small players will go away or not even want to get started and big players with the VC money will die once they burn all their money soon. But about his own startup, he is pretty much assured. “PDF.ai is bootstrapped and profitable with a very healthy margin. We don’t have a mission to become another unicorn; several million dollars in ARR is good enough for me, and that’s my goal in the next 1-2 years. I’m 1,000% confident we can make it!” he added. ChatGPT has been in the market for almost a year, and OpenAI will continue to add new features gradually. OpenAI’s ultimate goal is to achieve AGI, and small features like the ability to read PDFs are just a small part of that larger objective. Sahar Mor, product lead at stripe said “UI and ease of use are still valid propositions so vertical startups targeting specific segments can still prevail. Those are the horizontal AI startups that are at risk.” Mark Zahm, founder, Glass Acres LLC said “GPT wrappers will exist and thrive as long as GPT exists. Parent platforms can’t fill every niche and use-case. See every extension and plugin..ever.” Rowan Cheung shared on X, “Why is nobody talking about people getting filthy rich on GPT wrappers right now?” According to him, some startups are outperforming billion-dollar companies in web traffic. The list was almost evenly split between wrappers, fine-tunes, and proprietary models. This means that some GPT wrappers are receiving more monthly visits than companies with billion-dollar valuations. Source: a16z GPT wrapper-based startups often offer a cost-effective and efficient approach for businesses looking to incorporate AI capabilities without the complexities of building a model from scratch. OpenAI is not alone Besides OpenAI, numerous startups are emerging, and generative AI businesses are playing a significant role in the rise of unicorns. According to a recent report from venture capital firm Accel, 60% of these new unicorns belong to the Generative AI sector. Investment in GenAI startups based in Europe and Israel amounted to nearly $1 billion in the past year. This figure, however, is significantly lower than the funding received by US Generative AI startups, which exceeded $14 billion during the same period. However, it’s important to note that this data is influenced by a substantial $10 billion funding received by OpenAI alone, as highlighted in the report. As per recent reports, several photo AI apps and AI chatbots are actually making more money than ChatGPT. In September, both ‘Chat & Ask AI‘ and ‘ChatOn — AI Chat Bot Assistant‘ generated substantial revenue, raking in almost $3.38 million and $2.11 million, respectively. Additionally, ‘AI Chatbot — Nova’ and ‘AI Chatbot: AI Chat Smith’ were not far behind, earning approximately $1.44 million and $1.72 million during the same period. Moreover, Character.ai, a chatbot startup backed by a16z, is making waves with 2.39 million downloads recorded as of September. OpenAI might be ten steps ahead of its competitors with its multimodal capabilities, but that doesn’t mean other startups should simply give up. Instead, they can take inspiration from it to make a better product.","excerpt":"OpenAI recently introduced two new features to ChatGPT, namely ‘Upload many types of documents’ and ‘Use tools without switching’","categories":["Deep Tech"],"tags":["AI Startups"],"author_name":"Siddharth Jindal","publish_date":"2023-10-31T10:00:11","publication_year":"2023","word_count":786,"keywords":["Go","ChatGPT","GenAI","API","OpenAI","AI","chatbots","ML","generative AI","R","AI Startups"],"extracted_tech_keywords":["AI","ML","generative AI","GenAI","ChatGPT","OpenAI","chatbots","R","Go","API"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/fear-not-startups-every-openai-update-can-be-your-next-stepping-stone\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10165886,"title":"Why Can’t AI Features Be Turned Off in Some Apps?","content":"Every day, it feels like founders come up with a new adjective to describe the potential of artificial intelligence. And they’re not wrong—AI has truly demonstrated its transformative capabilities over the past few years. As a result, everyone wants to add an element of AI in their applications, and it’s now making its way into products used across generations. Whether it’s an email client, messaging apps, or a suite of AI features that show up after updating smartphones. But is it fair not to let users disable these additions or force them to go through a cumbersome process of hunting for buttons buried deep within menus to turn them off? Such has been the case with certain applications, and several users have expressed their frustrations on social media. While some have been unable to disable AI features, others are struggling to find the settings to do so, on products offered by companies like Meta, Microsoft, Google, Apple, xAI, and so on. A controversial example is Snapchat, which introduced the ‘My AI’ chatbot. Despite backlash, the company only allowed users with a Plus subscription to disable it, sparking further frustration among non-subscribers. While most of these products let users opt out of sharing their data for the purposes of training, some don’t want these features to clutter, or interfere with their usage experience. To understand the rationale behind such strategies, AIM spoke to Karan Peri, an independent product advisor, with over a decade of experience in product management at companies like Microsoft, Coinbase and Amazon, among others. ‘Trying to Please Everyone Will Only Lead to Failure’ When the company decides to introduce a feature, it is almost always based on an A\/B test, which involves testing two product versions among the users. The one that rates better on an aggregate makes it to the final version. Thus, in this process, the preference of individual users isn’t considered, only the majority’s. Peri stated that if a feature does not provide value for a long time, good product teams will either iterate or turn it off, especially if the backlash is detrimental to the engagement of the product. However, when companies have to consider the feedback of a minority section, and end up adding more knobs and controls to the feature, changes to the code base can increase the maintenance costs. “If you go on a path believing that you want to please everybody, you will fail. That means you’re dropping your quality to the least common multiple. You keep dropping till it pleases everybody,” Peri added. For instance, a 75-year-old user of a popular messaging app is frustrated by the AI features she can’t disable. She falls into the third group of people who dislike the feature, alongside power users in the first group who love it, and the second group consisting of those who don’t care about the feature. “But the first two buckets are way larger than the third bucket, so the company doesn’t care,” Peri explained. Nevertheless, if the messaging app occupies almost a monopolistic position in the industry, its competitive position is advantageous. “If not for that app, where would the 75-year-old woman go?” he questioned. “She may not go to another app because her grandchildren, and other family members aren’t there. Whether she likes it or not, she is the product has her hooked.” Hence, certain companies will continue to retain these features. Moreover, certain features may not appeal to people initially, considering the cognitively heavy task of understanding how it works. However, users eventually end up embracing it. Similarly, companies often aim to integrate these features within the fabric of the product, like Netflix, for example, where the AI-enabled recommendation engine has become a crucial factor in the user experience. For example, several users complain of the inability to disable Meta’s AI features on WhatsApp, Facebook, and Instagram. However, a report from The Information revealed that Meta AI daily usage is heaviest on WhatsApp and Facebook. Meta also revealed that the chatbot had about 185 million weekly users of Meta AI across their products, as of last year. Furthermore, despite Snapchat’s questionable choice of not letting users disable ‘My AI’ chatbot for free, the app has seen nothing but consistent growth in the number of active users over the years. That said, strategies vary across companies. For example, Apple Intelligence provides an option to turn the features off. “If you buy a new iPhone, which is super expensive, you must ensure that the phone is liked. If not, then the entire line will get discredited,” said Peri, indicating that the scenario takes a turn when there is hardware, and money is involved. However, when Apple came under fire for its AI-enabled notifications features hallucinating and reporting fake news, the company eventually halted the feature. Are Companies Making a Good Case for Value Addition? “People are thinking about turning an AI feature off because they do not know how to get value from it,” Peri pointed out. “If it was valuable and people understood what it was doing, maybe lesser people would have asked, ‘How do I turn it off?’ You want to turn it off because it is useless,” he added, indicating that people on forums wouldn’t want to ask how to turn it off if it added good value. Several factors contribute to this scenario. The company likely did not instill enough thought into the launch, choosing to release it and observe user behavior without explaining how it works. Moreover, the instructions may not have been clear enough for users to understand the feature’s purpose and functionality. However, Peri said that good product teams often implement a feature in a way where education isn’t needed. In conclusion, it boils down to a single fact—Not everybody can be pleased. The goal is to please as many people as possible.","excerpt":"Not everybody can be pleased. The goal is to please as many people as possible.","categories":["AI Features"],"tags":["AI (Artificial Intelligence)"],"author_name":"Supreeth Koundinya","publish_date":"2025-03-12T11:07:39","publication_year":"2025","word_count":973,"keywords":["Go","Meta AI","artificial intelligence","programming_languages:R","AI","R","XAI","Aim","Rust","xAI","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","Meta AI","xAI","Aim","R","Go","Rust","XAI","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/why-cant-ai-features-be-turned-off-in-some-apps\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10118660,"title":"Generative AI Jobs in India can Fetch You up to Rs 1 Crore","content":"The demand for generative AI jobs in India is definitely on the rise. A recent report revealed that senior developers working in generative AI draw over INR 1 crore per annum, while an entrant’s salary could easily be around INR 18 lakh per annum. Further, it stated that the techies in India are driving up salaries with additional AI, generative AI, and adjacent skills, commanding as high as a 30-50 percent premium over those without such expertise. AIM Research noted that the median salaries of generative AI developers and engineers ranged between INR 11.1 lakh and 12.5 lakh per annum. Upon further investigation, AIM found that entry-level AI engineers at companies like Accenture typically earned about INR 8.5 lakh per annum, compared to the INR 5-6 lakh per annum that regular software engineers took home. Additionally, AI engineers with generative AI skills saw a significant 50% increase in their salary. Notably, Accenture recently secured $1.1 billion worth of generative AI deals. AIM also noted that Accenture’s consulting roles in generative AI fetched significantly higher salaries than software engineering or tech roles. Experienced folks could easily make up to INR 35 lakh per annum with a base salary of INR 21 lakh per annum. “GenAI has definitely created a new market, and companies and VCs are moving fast to build and capitalise on this. So, people in this field are getting paid above-average salaries,” said Mankaran Singh, founder of Flowdrive. Further, he said that any engineer with expertise in deep learning can get started and start contributing to generative AI development within 2-3 months of studying. An interesting story emerged when AIM spoke to Izam Mohammed, who had never attended college and is now an AI\/ML engineer earning lakhs. He created the Python (ragrank.readthedocs.io) library for evaluating RAG models, using the credibility and skills he gained through online resources to land this job at 18. Moreover, as global capability centres (GCCs) establish their presence in India, young talent in the country can seek employment opportunities to leverage their skills in generative AI for lucrative careers. According to consulting firm ANSR, about 90 percent of GCCs in India plan to harness the potential of AI, ML, and cognitive computing in the next 2-3 years. A report indicates that mid-career software professionals in the GCC segment, with about three to eight years of experience, typically earned salaries ranging from INR 15 lakh to INR 35 lakh per annum. Recently, AMD announced that it will expand its GenAI team in India and will be hiring machine learning research scientists\/engineers at multiple levels. https:\/\/twitter.com\/PrakamyaMishra\/status\/1778155876576338181 What about AI  Startups? AIM contacted computer scientist Pratik Desai, the founder of KissanAI, to understand the impact generative AI can have on salary packages. Desai said, those who can conduct research using models, fine-tuning, and tokenisers and have a deep understanding of neural networks typically move to the US as they can get ‘high salaries, not just ‘30-50%’. Further, he said that traditional software development roles, which involve knowledge of libraries like Llama Index or LangChain and assist in AI project development, would eventually see an increase in salaries. “They will see an increase, maybe not immediately. The immediate increase would be with established tech startups,” he added. CoRover AI chief Ankush Sabharwal told AIM that the average salary for GenAI tech developers in India can range from INR 8 lakh to INR 24 lakh per annum, depending on experience and skill set. He puts them into two categories: GenAI Tech Developers and Business Problem Solvers. GenAI Tech Developers are tech wizards who specialise in machine learning, deep learning, coding languages, and generative models. Business Problem Solvers are equipped with strong problem-solving abilities, business acumen, and a deep understanding of data. Sabharwal said that the latter act as detectives, identifying problems, grasping industry nuances, and maximising data’s potential to leverage GenAI effectively. Both the roles offer substantial earning potential, attracting skilled professionals to the dynamic and impactful realm of GenAI employment opportunities, he added. “The average salary of an ML engineer is INR 30-40 lakh per annum,” said a GenAI engineer at Sarvam AI, adding that if you are a good ML engineer, your salary will be above the same SWE-level salary anyway, GenAI or not. Indian IT Pays Poorly in GenAI Indian IT is betting big on generative AI. TCS recently announced having trained over 350,000 employees in AI skills and plans to hire 40,000 freshers for FY25. Although TCS claimed to be training its employees in generative AI, the exact focus and content of this training remained ambiguous. Meanwhile, the entry-level salary for freshers at TCS, ranging from INR 3 lakh to 4 lakh per annum, remained lower than industry standards. AIM contacted an employee who works with generative AI at TCS. He said that the pay scale specifically for GenAI resources at the company won’t change, and the hike is also based on the company’s financial performance and business unit budget allocation. “I believe we may get some good hikes in the range of 25%-30% after switching to another company,” he added. Furthermore, he said that the necessary skills required to be a GenAI engineer include the basics of AI algorithms in machine learning, deep learning, neural networks, and NLP. When asked about TCS’s training focus in generative AI, he explained, “TCS has been releasing various courses on GenAI on our internal platform to upskill its workforce so that resources are ready to be deployed in client projects.” He added, “Along with the courses, TCS is providing free vouchers for Azure certification exams, access to Google Cloud, and Nvidia portal access for attending workshops, etc.” Due to lower salaries, many employees choose to work for startups, GCCs, and major tech companies such as Microsoft, Google, and Meta. These companies provide higher compensation packages that can range from INR 15 lakh to over INR 40 lakh annually, excluding bonuses and stock options, and also could go up to INR 1 crore for senior executives and leaders.","excerpt":"“Any engineer with expertise in deep learning can get started and start contributing to generative AI development within 2-3 months of studying.”","categories":["AI Features"],"tags":["AI career"],"author_name":"Siddharth Jindal","publish_date":"2024-04-19T11:29:23","publication_year":"2024","word_count":998,"keywords":["AI career","machine learning","GenAI","AI","neural network","ML","NLP","LangChain","deep learning","Aim","generative AI"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","neural network","NLP","generative AI","GenAI","LangChain","Aim"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/generative-ai-jobs-in-india-can-fetch-you-up-to-rs-1-crore\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10160622,"title":"IISc Develops ML Model to Predict Material Properties with Limited Data","content":"Researchers at the Indian Institute of Science (IISc), in collaboration with University College London, have devised a machine learning (ML) approach to predict material properties using limited data. This breakthrough holds promise for discovering materials with specific properties, such as semiconductors, and addresses the challenges posed by expensive and time-consuming testing methods. The research, led by Sai Gautam Gopalakrishnan, Assistant Professor at IISc’s Department of Materials Engineering, employs transfer learning- a technique where a model pre-trained on a large dataset is fine-tuned for smaller, target-specific datasets. Gopalakrishnan explains the concept using an analogy: a model trained to classify general images can later be fine-tuned for specialised tasks, such as detecting tumors in medical scans. For their study, the researchers used Graph Neural Networks (GNNs), which work with graph-structured data, such as the three-dimensional crystal structures of materials. In GNNs, atoms are represented as nodes and bonds as edges, enabling the model to learn complex material properties effectively. The team optimised the GNN architecture and the size of the training data, freezing some layers of the model during pre-training to enhance efficiency. The researchers also introduced a Multi-property Pre-Training (MPT) framework, training their model simultaneously on seven bulk material properties. This approach significantly improved the model’s predictive accuracy, enabling it to estimate material properties like the piezoelectric coefficient and even predict the band gap values of 2D materials it had not encountered before. Source: IISc Transform Semiconductor Field The transfer learning-based model outperformed conventional models trained from scratch. Its applications extend beyond semiconductors, aiding in areas such as battery technology, where predicting ion mobility within electrodes could improve energy storage. Gopalakrishnan noted the model’s potential in supporting India’s semiconductor manufacturing ambitions by predicting the tendency of materials to form point defects. This innovative method leverages rich, multi-property data to enhance predictions and represents a leap forward in materials science, offering solutions for industries reliant on advanced materials and energy technologies.","excerpt":"The new development could aid semiconductor and battery advancements.","categories":["AI News"],"tags":["IISc","material science","ML","semiconductor","University College London"],"author_name":"Vandana Nair","publish_date":"2024-12-31T15:58:46","publication_year":"2024","word_count":319,"keywords":["Go","University College London","machine learning","programming_languages:R","AI","neural network","material science","ML","semiconductor","programming_languages:Go","IISc","RAG","R"],"extracted_tech_keywords":["AI","machine learning","ML","neural network","RAG","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/iisc-develops-ml-model-to-predict-material-properties-with-limited-data\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10047805,"title":"NVIDIA Shows How To Build AI Models At Scale With PyTorch Lightning","content":"Deploying predictive AI models across a business is no easy feat. The effectiveness of AI depends on the quality of the underlying models. Therefore, it becomes crucial for data scientists and academic researchers to quickly build models with various parameters and identify the most effective ones to deploy them easily and scale them seamlessly. AI model training process (Source: NVIDIA) In its latest blog post, NVIDIA researchers showed how to build speech models with PyTorch Lighting on CPU-powered AWS instances (Grid). PyTorch Lightning is a lightweight PyTorch wrapper designed to make high performance AI research simple. It is an organised PyTorch, which allows users to train their models on CPU, GPUs, or multiple nodes without changing any code. Grid, which runs on AWS, supports Lightning and classic machine learning frameworks such as TensorFlow, Keras, PyTorch, Sci-Kit, and others. It also helps users to scale the training of models from the NGC catalogue. NGC catalogue is a curated set of GPU-optimised containers for deep learning, visualisation, and high-performance computing (HPC). PyTorch lightning software and developer environment is available on NGC Catalog. Also, check out GitHub to get started with Grid, NGC, PyTorch Lightning here. Training AI Models For building speech models, NVIDIA researchers have used ASR, which transcribes spoken language to text. ASR is a critical component of speech-to-text systems. So, when training ASR models, the goal is to generate text from a given audio input that reduces the word error rate (WER) metric on human transcribed speech. The NGC catalogue contains SOTA pretrained models of ASR. Further, they use Grid sessions, NVIDIA NeMo, and PyTorch Lightning to fine-tune these models on the AN4 dataset, aka Alphanumeric dataset. Collected and published by Carnegie Mellon University, the AN4 dataset consists of recordings of people spelling out addresses, names, phone numbers, etc. Here are the key steps to follow when building speech models: Create a Grid session optimised for Lightning and pretrained NGC models Clone the ASR demo repo and open the tutorial notebook Install NeMo ASR dependencies Convert and visualise the AN4 dataset (Spectrograms and Mel spectrograms)Load and inference a pre-trained QuartzNet model from NGCFine-tune model with Lightning Inference and deployment Pause session Create a Grid Session Users can run Grid sessions on the same hardware they need to scale while providing them with pre-configured environments to iterate the ML process faster. Here, sessions are linked to GitHub, loaded with ‘JupyterHub,’ and can be accessed through SSH and IDE without installing. Check out the Grid Session tour here. (requires Grid.ai account) Clone ASR demo repo & open tutorial notebook Once you have a developer environment optimised for PyTorch Lightning, the next step is to clone the NGC-Lightning-Grid-Workshop repo. After this, the user can open up the notebook to fine-tune the NGC hosted model with NeMo and PyTorch Lightning. Install NeMo ASR dependencies Install all the session dependencies by running tools like PyTorch Lightning and NeMo, and process the AN4 dataset. Then, run the first cell in the tutorial notebook, which runs the following bash commands to install the dependencies. Convert and Visualise the AN4 Dataset The AN4 dataset contains raw Sof audio files. Convert them to the Wav format so that you can use NeMo audio processing. Once processed, you can then visualise the audio example as images of the audio waveform. The below image shows the activity in the waveform that corresponds to each letter in the audio. Each spoken letter has a different “shape.” Interestingly, the last two blobs look relatively similar because they are both the letter N. The audio waveform of the sample example (Source: NVIDIA) Load and Infer Pretrained Model After you have understood the AN4 datasets, the next step is to use NGC to lead an ASR model to fine-tune with PyTorch Lightning. This model comes with many building blocks and even complete models that you can use for training and evaluation. To model the data, the researchers have used a Jasper architecture called QuartzNet. The image below shows that Jasper architecture consists of a repeated block structure that uses 1D convolutions to model spectrogram data. QuartzNet is a better variant of Jasper as it uses time-channel separable 1D convolutions. As a result, it helps in reducing the number of weights dramatically while keeping similar accuracy. Jasper architecture (Source: NVIDIA) Fine-tune the Model with Lightning Once you have a model, you can fine-tune it with PyTorch Lightning. Some of the key advantages include checkpointing and logging by default. Also, you can use 50+ best-practices tactics without needing to modify the model code, including multi-GPU training, model sharding, quantisation-aware training, deep speed, early stopping, mixed precision, gradient clipping, profiling, etc. Inference and deployment Once you have a baseline model, the next step is to inference it. Pause Session Once you have trained the model, you can pause the session, and all the files you need are required. Check out all the source codes related to PyTorch Lightning, NGC, and Grid on NVIDIA’s blog.","excerpt":"PyTorch lightning software and developer environment is available on NGC Catalog","categories":["Global Tech"],"tags":["AI (Artificial Intelligence)","AI Models","deploying models","learn machine learning","Machine Learning","Machine Learning Latest"],"author_name":"Amit Naik","publish_date":"2021-09-07T10:00:00","publication_year":"2021","word_count":826,"keywords":["AI Models","machine learning","deploying models","learn machine learning","AI","PyTorch","Keras","ML","AWS","Machine Learning","Machine Learning Latest","deep learning","Jupyter","TensorFlow","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","TensorFlow","PyTorch","Keras","Jupyter","AWS","R"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/nvidia-shows-how-to-build-ai-models-at-scale-with-pytorch-lightning\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10039195,"title":"Chips That Power Your Vlogs: The Secret Behind YouTube’s Uninterrupted Service","content":"“To let anyone upload a video to show anyone else in the world, for free takes a lot of processing power.” The pandemic turned out to be a blessing in disguise for video streaming services. In the first quarter of last year, YouTube alone witnessed a 25 percent increase in watch-time. And for the first half of last year, total daily live streams grew by 45 percent. Running a global platform like YouTube with massive amounts of video being uploaded, stored, and distributed every second for its millions of creators and billions of viewers is a complex and demanding task. Youtube videos are created and uploaded in a single format but consumed on different devices at different resolutions. The infrastructure team’s job is to get those videos ready for you to watch through a process called transcoding. Transcoding is the process of compressing videos so that the smallest amount of data is sent to the device with the best possible quality. “But it’s costly and slow, and doing that processing using regular CPUs is pretty inefficient, especially as you add more and more videos,” said Jeff Calow, a software engineer at Youtube. Transcoding challenges Handling and scaling different output resolutions and formats.Handling complex algorithmic trade-offs and quality\/compression\/computing compromises. Providing inter and intra-task parallelism.Providing high performance at low costs.Enabling ease of deployment when operating at scale. Image credits: Youtube blog Looking beyond GPUs and TPUs Video Coding Unit(VCU) YouTube’s video processing platform currently supports multiple video-centric workloads at Google. YouTube also has to deal with hundreds of hours of video uploads every minute apart from Google Photos and Google Drive, which demand similar bandwidth. The team at YouTube created a new system for transcoding video more efficiently at data centres. They developed a custom chip to transcode video, and a software to run it– the Video (trans)Coding Unit (VCU). “We’ve seen up to 20-33x improvements in compute efficiency compared to our previous optimized system, which was running software on traditional servers,” said Calow. “VCUs resulted in upto 20-33x improvements in compute efficiency.” VCUs are built ground-up for data-centre-scale video workloads. At this scale, deployment becomes a challenge. So the engineers at YouTube designed accelerators for userspace software control and addressed the hardware failures through redundancy and fallback at higher-level software layers. Not only that, but the engineers also have to account for the constant updating of the applications. Viewers rarely watch those pixelated 360p videos. Today, the internet is faster and cheaper. Youtube users no longer hesitate to click those 4k resolution videos. For Youtube, an affordable internet is a great business opportunity but also an engineering nightmare.VCUs enable programmability and interoperability while closely monitoring the computationally expensive infrequently-changing aspects of the system. VCU System Design (Source: Ranganathan et al.,) According to the team behind VCUs, the software and hardware were loosely coupled to facilitate parallel development pre-silicon and continuous iteration post-silicon. VCU has 3,000 millidecode cores and 10,000 milliencode cores available. The codec cores in the VCU are programmed as opaque memories by the on-chip management firmware. The loose coupling allows userspace software to adjust the flow of frames through codecs and changing codec modes without requiring other system changes. VCU hardware design is a combination of Mentor Graphics’ Catapult tool and an in-house integration tool called Taffel. The encoder core design is implemented using a C++ based HLS design flow for faster development and design iteration. VCU vs CPU, GPU (Source: Paper by Ranganathan et al.,) Video processing has quickly become a data centre workload headache. So, VCUs or their variants are inevitable. Though some GPUs support transcoding, they fall short when it comes to video-sharing workloads. The work on VCUs is the first of its kind on broadcast-quality video acceleration. The engineers also have explored the design trade-offs for commercial production workloads serving hundreds of hours of uploads per minute and discussed co-design trade-offs with a production video processing software stack and deployment at scale. As the world slowly drifts towards a predominantly virtual setup, live streaming, virtual conferencing, cloud gaming, vlogging, AR\/VR footage will become more prominent. According to the YouTube engineers, Video Coding Unit is just the beginning. They believe that rich opportunities for future innovation await in the form of combining transcoding with other machine-learning on videos, such as automatic caption generation and more.","excerpt":"“To let anyone upload a video to show anyone else in the world, for free takes a lot of processing power.” The pandemic turned out to be a blessing in disguise for video streaming services. In the first quarter of last year, YouTube alone witnessed a 25 percent increase in watch-time. And for the first […]","categories":["Deep Tech"],"tags":[],"author_name":"Ram Sagar","publish_date":"2021-04-29T17:00:00","publication_year":"2021","word_count":719,"keywords":["Go","TPU","programming_languages:R","AI","innovation","programming_languages:Go","C++","ViT","GAN","R"],"extracted_tech_keywords":["AI","TPU","R","Go","C++","GAN","ViT","innovation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/chips-that-power-your-vlogs-the-secret-behind-youtubes-uninterrupted-service\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10122120,"title":"We Won’t Just Replace Our Security Engineering Team with a Coding Engine: Appdome CPO","content":"The discourse continues regarding whether AI can effectively mitigate security threats stemming from AI advancements. Notably, we’ve witnessed instances where security companies integrate AI into their defence mechanisms against cyber threats. Appdome chief product officer Chris Roeckl stated that his company won’t simply substitute its security engineering team with AI, solely because of the current AI trend.“We are a security company. We’re not about to replace our highly valuable security engineers with a coding engine anytime soon. The value of our security research team lies in their creativity and deep understanding of our code base. The human element is certainly very important to us in terms of how we provide our solution,” Roeckl told AIM in an exclusive interaction. Founded in 2012, and headquartered in Redwood City, California, Appdome is a software company that provides a mobile app security and integration platform. Appdome’s cyber defence automation platform is a no-code platform that enables app developers to add over 300 security, anti-fraud, anti-malware, anti-cheat, and other protections to Android and iOS apps. Currently, Appdome leverages a machine-learning engine to power its platform but has not integrated large language models (LLMs) into its platform yet. “It’s an unanswered question. Our research team is currently exploring – do we leverage AI to combat AI or rely on good old traditional practices? This remains an open question as we evaluate whether an AI-driven response could prove advantageous,” he added. But There Are Benefits During our discussion, Roeckl touched upon the benefits AI could bring to Appdome’s platform. For instance, AI could play a great role in recommending security features to Appdome’s clients, choosing from the 300+ features they have on their platform. “Depending on where your app is, and what kind of app it is, we might be able to provide very specific recommendations on how to secure your app. So the idea of using an AI-based recommendation engine is very interesting to us,” he said However, whether the recommendation engine will be powered by an LLM or a conventional recommendation engine is not determined yet. Roeckl notes that it could be a mixture of AI and their existing machine-learning engine. AI is Making Bad Actors Good Coders While AppDome is contemplating using LLMs, numerous threats are emerging due to AI’s increasing popularity. For instance, generative AI is making it easier for bad actors to write malware code. ChatGPT, a free tool, can write code, while more advanced tools like GitHub Copilot are available for developers at a price of $10\/month. However, GitHub has measures in place to prevent Copilot from being used to write malware and similar malicious content. “We see that AI coding is lowering the barrier of entry for bad actors to actually code attacks. So now we see that mobile apps are under increasing threat of attack because bad actors can write good codes with the help of AI even though they don’t have good coding abilities,” Roeckl said. Phone-Based AI Agents Will Open Doors to More Security Threats AI is touching almost every field, and AI agents are expected to be the next big iteration. Experts envision that most consumer interactions online will involve agents performing tasks on their behalf. Pretty soon, your smartphone’s built-in AI agent might communicate with your preferred food delivery app’s AI agent to place orders automatically on your behalf. This, however, according to Roeckl, opens the doors for more security vulnerabilities. For these systems to communicate effectively, they rely on application programming interfaces (APIs) to facilitate interaction. However, APIs can be vulnerable to interception and misuse for malicious purposes. “For instance, many APIs in mobile apps are exposed, leading to potential bot attacks, such as sneaker bots. Mobile channels are increasingly targeted for such attacks due to embedded coding logic within the apps. “Failing to secure the app can expose the coding logic of the APIs, posing a significant threat. Therefore, as the API economy drives tighter interconnections between systems, brands must heighten vigilance in safeguarding these systems.” Roeckl said. AI is Not Creating a New Category of Threats, Yet While generative AI does open the door for more security threats in mobile security, according to Roeckl, no new kinds of threats are emerging. However, he does acknowledge that AI is resulting in a volumetric increase in threats overall. “We’re not seeing unique attacks from AI. While we can track and identify them as originating from AI-powered systems, they largely resemble the standard types of attacks we encounter every day. Recently, we’ve particularly focused on social engineering attacks as a significant area of concern.” Social engineering attacks are indeed becoming more sophisticated and dangerous with the integration of AI. “There are deepfakes of both images and voices. Additionally, there are phishing attacks, which can lead to mobile app vulnerabilities. AI elements have begun to infiltrate all these various types of attacks,” said Roeckl.","excerpt":"Generative AI is making it easier for bad actors to write malware codes.","categories":["AI Features"],"tags":["Agents","Interviews and Discussions"],"author_name":"Pritam Bordoloi","publish_date":"2024-05-31T18:30:00","publication_year":"2024","word_count":809,"keywords":["Go","ChatGPT","API","Agents","AI","Git","RAG","Aim","generative AI","GitHub","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","generative AI","ChatGPT","Aim","RAG","R","Go","Git","GitHub","API"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/we-wont-just-replace-our-security-engineering-team-with-a-coding-engine-appdome-cpo\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":11878,"title":"I know what you will buy next –[Power of AI &#038; Machine Learning]","content":"Today, we are surrounded by a variety of products which makes it difficult for one to identify the best or the combination of the best products. Businesses are flooded with millions of transactions every day and it becomes ever more important to understand the choices and preferences of customers’ buying behaviour. People use a variety of strategies to make choices about what to buy and how much to spend. Recommender Systems automate some of these strategies with the goal of providing affordable, personal and high-quality recommendations. Given below are the approaches to develop a state-of-the-art Recommender System. WHAT IS A RECOMMENDER SYSTEM? Recommender System or algorithms start by finding a similarity score over the customers’ purchased and rated items which overlaps the other users’ purchased and rated items. The algorithm aggregates items from these similar customers, removes items the user has already purchased or rated, and recommends the remaining items to the user. Two popular versions of these algorithms are Collaborative Filtering and Cluster Models. Other algorithms like Market Basket Analysis and Association Rule Mining are also used by the industry that focuses on finding similar items, and not similar customers. WHAT WE NEED TO RECOMMEND? So, if a customer has not yet bought item 2,3,5 and 7, the Recommender System will calculate the similarity of the customer purchased items with his non-purchased items and will recommend the best items which are similar to his purchased items. IMPLEMENTING A RECOMMENDER SYSTEM Recommendation algorithms provide an effective form of targeted marketing by creating a personalised shopping experience for each customer. For large online retailers, a good recommendation algorithm should be scalable over very large customer bases and product catalogue and should generate compelling recommendations for all users regardless of the number of purchases and ratings. Item-to-Item Collaborative Filtering which we are showcasing below in this report is able to meet these challenges. We move a step further and try to present cases where retail industry can broadly apply recommendation algorithms for targeted marketing, both online and offline. With the techniques to measure conversion rate, click rate, and open rate it has become effective for offline retailers to use Recommendation System in postal mailings, coupons, POS driven promotions and other forms of customer communication. There are three common approaches to solving the recommendation problem: Cluster Models, Collaborative Filtering and Association Rule Mining. CLUSTER MODELS To find customers who are similar to the user, Cluster Models divide the customer base into many segments and treat the task as a classification problem. The algorithm’s goal is to assign the user to the segment containing the most similar customers. It then uses the purchases and ratings of the customers in the segment to generate recommendations. The segments typically are created using a clustering or other machine learning algorithm, although some applications use manually determined segments. ITEM-TO-ITEM COLLABORATIVE FILTERING Rather than matching the user to similar customers, Item-to-Item Collaborative Filtering matches each of the users’ purchased and rated items to similar items, then combines those similar items into a recommendation list. To determine the most-similar match for a given item, the algorithm builds a similar-items table by finding items that customers tend to purchase together. We could build a product-to-product matrix by iterating through all item pairs and computing a similarity metric for each pair. ASSOCIATION RULES It’s a rule-based machine learning method for finding frequent patterns and analysing item sets in customers’ basket or transactions and identifying frequently occurring item sets, which can be the basis for recommendations also known as Market Basket Analysis in Retail Not going much into the statistical part, we can clearly see that in the previous baskets insect killer and deo combination was present in 3 out of the 4 baskets. This shows that whenever insect killer was bought, deo was bought along with it (one of the arguments in favour of this correlation can be attributed to men buying insect killer and deo combination as a top up when they visit super market). This is an intro to how Market Basket Analysis works. ILLUSTRATION 1 Below are the few illustrations of Recommendation Engine live in action that we have compiled: Recommendations for online grocery purchases based on historical transactions. Such recommendations can easily be deployed live on online ecommerce portals. We have used Item-to-Item Collaborative Filtering approach which uses similarity scores between two items and recommends the top items using the similarity score. This method also solves the problem of recommending new items as the algorithm compares the new item similarity with the existing items and then will recommend the new items alongside the existing items bought by the user. Let’s say, a new beer-based shampoo is introduced in the catalogue. The algorithm will calculate the similarity of this product with all the existing products in the catalogue. The products with the highest similarity scores with the beer-based shampoo will be recommended to the user who will buy those products or who have already bought those products. On an intuitive level, it is not hard to see that beer-based shampoo will be recommended to beer buyers, shampoo and other toiletries buyers. In the table given below, Rec1 to Rec8 are the top eight recommendations of the items list. So, after a customer has added citrus fruit in the cart, the Recommender System will recommend bread and yogurt as the top recommendations. ILLUSTRATION 2 Recommendations for Tours and Travels brands and portals based on the historical trips made by the customers Below is the recommendation tracker for few of the destinations that Tours and Travels portal has under his umbrella based on travel destinations opted by the customers previously. Now the merchant knows that a customer who has visited Goa has higher affinity to visit Alibagh, Coorg, Mahabaleshwar, Shirdi and Ganpatipule. These combinations of Goa vs the top 5 recommendations can be suggested to customers under different packages. This will inform customers about what their fellow travellers did and will garner additional business for the merchant. In below table, Rec1 to Rec5 are the top five recommendations of the items list. So, after a customer has travelled to Goa, the Recommender System will recommend Alibagh, Coorg as the top recommendations for his\/her next visit. Retail industry can broadly apply recommendation algorithms for targeted marketing, both online and offline. With the techniques to measure conversion rate, click rate and open rate, it has become effective for offline retailers to use the Recommendation System in postal mailings, coupons, POS driven\/Real time in-store promotions and other forms of customer communication. [divider size=”1″] Authors Nishit Mittal Nishit Mittal is working with the Consulting-Econometrics and analytics research team at Hansa Cequity and is based out of Mumbai office. Nishit is a data science enthusiast and is trained in data science and business analytics from IIM Bangalore and LSE. He has worked as an economist with RBS and UBS and as a consultant with NCAER before joining Hansa Cequity. He can be reached at nishit.mittal@cequitysolutions.com Ankit Patel Ankit Patel is an alumnus from DA-IICT(Dhirubhai Ambani Institute for ICT) and currently working with the analytics team at Hansa Cequity. He leverages his skills to explore more meaningful avenue for analytics with the clients at Hansa Cequity. He can be reached at ankit.patel@cequitysolutions.com","excerpt":"Today, we are surrounded by a variety of products which makes it difficult for one to identify the best or the combination of the best products. Businesses are flooded with millions of transactions every day and it becomes ever more important to understand the choices and preferences of customers’ buying behaviour. People use a variety […]","categories":["AI Features"],"tags":["hansa cequity","RECOMMENDER SYSTEM"],"author_name":"AIM Media House","publish_date":"2016-12-26T08:35:58","publication_year":"2016","word_count":1205,"keywords":["data science","Go","machine learning","programming_languages:R","AI","Scala","RAG","analytics","RECOMMENDER SYSTEM","GAN","hansa cequity","R"],"extracted_tech_keywords":["AI","machine learning","data science","analytics","RAG","R","Go","Scala","GAN","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/know-will-buy-next-power-ai-machine-learning\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":28928,"title":"Is K-Means Clustering Really The Best Unsupervised Learning Algorithm?","content":"Unsupervised learning has emerged as the most effective technique for discovering patterns in data. It is also being used to develop labels on top of the supervised models. This is one of the most widely used techniques for market or customer segmentation wherein the company’s data can be segregated into clusters and used to identify certain patterns which leads to a more customised approach. This technique comprises machine learning algorithms through which data analysts can draw inferences from datasets without labelled responses. Cluster analysis is also widely used for exploratory data analysis to find hidden patterns or grouping in data. Unsupervised Learning Algorithms Can Be Divided Into Two Wide Categories: Clustering: A clustering problem is where one can find the inherent groupings in the data, such as grouping customers by purchasing behaviour. Association: An association rule learning problem is where you want to discover rules that describe large portions of your data, such as people that buy X also tend to buy Y. Some of the common clustering algorithms are hierarchical clustering, Gaussian mixture models and K-means clustering. The last one is considered one of the simplest unsupervised learning algorithms, wherein data is split into k distinct clusters based on distance to the centroid of a cluster. Why K-Means Clustering Is So Popular K-Means for Clustering is one of the popular algorithms for this approach. Where K means the number of clustering and means implies the statistics mean a problem. It is used to calculate code-vectors (the centroids of different clusters). According to a tutorial, for any word\/value\/key that needs to be ‘vector quantized’, it is by calculating the distance from all the code vectors and assign the index of the code vector with the minimum distance to this value. For example, clustering can be applied to MP3 files, cellular phones are the general areas that use this technique. Is K-Means Really Used In Production? K-means has been around since the 1970s and fares better than other clustering algorithms like density-based, expectation-maximisation. It is one of the most robust methods, especially for image segmentation and image annotation projects. According to some users, K-means is very simple and easy to implement. However, it is unlikely to be the state-of-the-art, but for straightforward clustering, it is also a part of a larger data-processing pipeline, K-means is a reasonable default choice, at least until you figure out that the clustering step is your bottleneck in terms of overall performance. Strengths Of K-Means Clustering Algorithm According to this paper, (Learning Feature Representations With K-means) K-means is used to learn feature representations for images (use k-means to cluster small patches of pixels from natural images, then represent images in the basis of cluster centres; repeat this several times to form a “deep” network of feature representations) gives image classification results that are competitive with much more complex \/ intimidating deep neural network models. In fact, a lot of k-means applications are now done using support vector machines. It gives good results It is already implemented in the software Number of clusters has to be fixed before Dependent of the initialisation parameters and the chosen distance Weakness The results given are usually dependent on the initial values for the means. And the way to initialise the means is not specified, one can start by randomly choosing K of the samples. How K-Means Algorithm Functions: The algorithm clusters into k groups and here k is the input parameter. In this procedure, a dataset is classified through a certain number of clusters, commonly known as k clusters and the main idea is to define k centres, one for each cluster. These centers should be placed in a way since different location causes different results. So, the better choice is to place them as much as possible far away from each other. The next step is to take each point belonging to a given data set and associate it to the nearest centre. When no point is pending, the first step is completed and an early group age is done. However, the main disadvantage is one has to specify the number of clusters as an input in the algorithm. Conclusion Many argue that in the field of data science, one should primarily use simple, self-learning algorithms. And clustering algorithm, the most commonly used unsupervised learning algorithm is self-improving and one doesn’t need to set parameters. In fact, most data science teams rely on simple algorithms like regression and completely because they solved all normal business problems with simple algorithms like XG Boost. Another key upside of K-means, the standard data mining tool is that as opposed to conventional statistical methods, the clustering algorithms do not depend on statistical distributions of data and can be used with little prior knowledge exists, a paper on data mining using K-means discussed.","excerpt":"Unsupervised learning has emerged as the most effective technique for discovering patterns in data. It is also being used to develop labels on top of the supervised models. This is one of the most widely used techniques for market or customer segmentation wherein the company’s data can be segregated into clusters and used to identify […]","categories":["AI Features"],"tags":["clustering algorithms","K-means clustering algorithm"],"author_name":"Richa Bhatia","publish_date":"2018-10-04T09:42:34","publication_year":"2018","word_count":799,"keywords":["data science","Go","machine learning","programming_languages:R","AI","neural network","ML","programming_languages:Go","clustering algorithms","R","K-means clustering algorithm"],"extracted_tech_keywords":["AI","machine learning","ML","neural network","data science","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/is-k-means-clustering-really-the-best-unsupervised-learning-algorithm\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":38531,"title":"How To Become A Data Scientist In 2019: A Beginner’s Guide","content":"How can I become a data scientist? What are the most valuable skills to learn for a data scientist now? Could I learn how to be a data scientist by going through online tutorials? What does a data scientist do? These are only some of the questions that are being discussed online, on blogs, on forums and on knowledge-sharing platforms like Quora. The data science sector is flourishing to such an extent that our earlier jobs study revealed that there are currently more than 97,000 job openings for analytics and data science in India right now. It is true that the “hottest job of the 21st century” has all the buzz, glam and traffic, but many enthusiasts are still confused as to what this job entitles. Fewer still, understand, what it takes to be a data scientist. In this article, Analytics India Magazine will give you a step-by-step guide to becoming a data scientist. It’s not an easy journey, but the results are worth it. Note: This is a beginner’s guide 1. Learn Mathematics And Statistics If you are a data science aspirant, you need strong background in mathematics and basic knowledge of statistics. In the midst of the hype around data-driven decision making, the basics are somehow getting sidelined. The boom in data science requires an increase in executive statistics and maths skills. Some of the fundamental concepts expected from a business analyst are correlation, causation and how to statistically test hypothesis. We highly recommend starting with basics in linear algebra, then gradually moving to calculus. It may be hard to master them initially but given the time and practice with working, these areas will be familiar and comfortable to work on. 2. Practice Programming Numerous studies, including our own, have come to the conclusion that Python is the most important language to be learnt by a data scientist. In fact, in 2019, this sentiment is being clamoured, as almost 75% of the industry, as well as the professionals, are saying that. So beginners should focus on learning Python programming for at least their first next six months and interacting with databases. Then once you have a good understanding of Python and programming in general, you can then start learning other languages like R and Java, then move to machine learning packages like scikit-learn. Sites like Kaggle and DataCamp are extremely good at testing codes and collaborating with peers and developers. In addition, forums like Stack Overflow are excellent to discuss problems and queries related to programming. 3. Dip Your Feet Into Machine Learning Many people do the mistake of learning every algorithm in ML and forget where it actually helps in solving a problem. For beginners, it is suggested that they learn the popular and standard algorithms. A complicated algorithm is not always the solution for complex applications. It is all about how an ML problem is solved optimally. Here are a few blogs which brilliantly explain the process: Machine Learning Mastery by Jason Brownlee – An amazing blog by expert Jason Brownlee. He explores the fascinating world of ML and captures its essence in the real world. Adam Geitey’s blog – interesting write-ups in ML and Python Arthur Juliani’s blog on Reinforcement Learning – an absolute gem of a blog which particularly focuses on reinforcement learning in ML. Edwin Chen’s blog – it explores requisite concepts of ML such as neural networks, deep learning etc. as well as the math behind it. If you do decide to take a course in ML, keep in mind that most of these programmes only give a brief idea of the basic algorithms like Support Vector Machines (SVM) and neural networks. However, they strengthen concepts like matrix operations and linear regression and teach supervised and unsupervised learning. An introduction to some projects using programming languages, generally Matlab, R, Python or Octave also form a part. These include projects like ‘text recognition’, ‘spam classifier’, ‘movie recommender systems’. 4. Create And Build Machine Learning Projects Learning is just the beginning, we need to implement. The knowledge one possesses can be appreciated only when it is represented. Taking up live projects, understanding the architecture behind the screen would help a lot. Hands-on experience in the field of data science is very much needed at the moment, large firms look for people who have experience and an analytical mindset. You can learn via the following projects which are perfect for beginners: The Iris flower classification project MovieLens 100k Turkiye student evaluation dataset BigMart sales prediction 5. Create A Portfolio While a resume is an important component to showcase your abilities to the potential employers, a data scientist should also be able to showcase his\/her abilities in coding and other software capabilities. A crucial part of data science jobs is to be able to code, and GitHub serves as a perfect platform to access the coding skills and display hands-on ability to solve problems. Here are the points one must remember: Good author information: Details such as candidate’s username, location, email address, current employer, etc must be included. Large followers: The number of followers that your portfolio has is a good indication of the work that you have done in the past. More than 50 is usually a decent number. Contribution graph: This showcases your keenness to explore other areas and shows activity levels in the coding community. The greener the graph is, the better is your contribution rate. Improving on stars: This is a way of identifying how you have engaged in the community. 100 stars are usually considered decent but larger the better. Forking and creating repositories: More the number of people who have forked, the greater is the popularity of that developer’s project in the GitHub community. A large amount of activity indicates that the developer is working on a popular project. Writing employer-targeted code: Writing a code related directly to the employer’s business is a good way to catch their attention. It can showcase your coding abilities while demonstrating the interest you have in getting that job. 6. Focus On Soft Skills Too Industry experts say that simply hiring a data scientist is not enough. Managers need to take special care to align business and data teams thus enabling data scientists to be self-sufficient. Otherwise, they might not get the expected ROI in data science which is a problem almost 80% of the companies face. Here are the skills a good data scientist should focus on: Communication Problem-solving Ability To Draw Parallels To Real-world Problems Prioritisation Business acumen 7. Apply For Jobs Wisely As tools are evolving, data science job roles are maturing and becoming more mainstream in companies. The number of openings that companies have for data science roles is also on an all-time high. Given the number of opportunities available, these are being expanded to professionals with a non-technical background as well. While there are many positions with a shortage of ideal candidates, it has made it quite possible for one candidate landing up with more than one job offer in hand for similar roles. That’s why it is necessary to ask questions about: Responsibilities Tools used Methodologies used Type of data used in the organisation Time spent on aspects of the role like analysis, data management 8. Keep Upskilling To keep up with the changing times, most organisations try to hire candidates who have a definite willingness to learn and upskill. We have seen in the past how companies like Cognizant laid-off employees who were not able to keep up with the changing technologies and failed to upskill themselves. Companies in 2019 are focussing on not just training a single skill but a cluster of skills which will be relevant for more number of years. Some of the skills that are currently picking up are: Automation RPA Robotics Cybersecurity Artificial intelligence IoT Connected devices FinTech Data analytics Blockchain","excerpt":"How can I become a data scientist? What are the most valuable skills to learn for a data scientist now? Could I learn how to be a data scientist by going through online tutorials? What does a data scientist do? These are only some of the questions that are being discussed online, on blogs, on […]","categories":["AI Trends"],"tags":["blockchain tutorial","Data Science","Data Scientist","how to become a data scientist","Machine Learning","programming","Python","R"],"author_name":"Prajakta Hebbar","publish_date":"2019-05-01T08:32:18","publication_year":"2019","word_count":1309,"keywords":["data science","scikit-learn","how to become a data scientist","artificial intelligence","machine learning","AI","neural network","programming","ML","Machine Learning","Python","blockchain tutorial","deep learning","analytics","Data Science","Data Scientist","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","neural network","data science","analytics","scikit-learn","Python"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/how-to-become-a-data-scientist-in-2019-a-beginners-guide\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10110818,"title":"[Exclusive] Indian AI and Robotics Company Confirms Level 5 Autonomy","content":"A few weeks ago, Indian autonomous driving company Swaayatt Robots claimed to have achieved Level 5 autonomy. The company announced this big breakthrough in a video, showcasing autonomous driving of the car at night at a toll plaza amid complex traffic dynamics. The video, however, failed to convince the sceptics as here was a company claiming to have successfully developed Level 5 autonomy at a time when Tesla, General Motors and other bigwigs were still languishing at Level 2. To clear the air, AIM did an exclusive interview with Sanjeev Sharma, founder and chief of Swaayatt Robots, who confirmed, “If we look at the demo conducted at the toll plaza, the vehicle showed Level 5 capabilities.” “At night, there were no traffic rules. You could see trucks going zigzag, overtaking our vehicle randomly. Also, there were trucks parked here and there. Our vehicle, arriving at the intersection, slows down for the speed breaker and then decides which lane or gate to commit to,” he said, noting that this capability falls under level 5 autonomy. Moreover, Sharma said their technology is capable of negotiating bi-directional traffic on a single lane. “No other company deals with this; they just come to a complete halt. Our vehicle just took a small detour even when the object was coming from the wrong end. Now, we are able to do this because this intelligence is inherently embedded.” As for the level game, he said that he leaves it up to the people to decide if this was level four or five. In fact he said that he would be fine even if they wanted to mark it level three. “However, wherever we felt that a claim had to be made, we made the claim, specifically in toll plaza navigation, asserting that this is level 5.” Tech Behind This Sharma said that Swaayatt Robots is heavily invested in reinforcement learning and inverse reinforcement learning. He considers Wavye AI to be their toughest competitor, given that Wavye AI also employs reinforcement learning. Moreover, he said that Swaayatt Robot’s current focus is to get rid of as many algorithms as possible. According to Sharma, while other companies are working on developing algorithms for obstacle detection and the intent of the other vehicles, they are working on eliminating the obstacle detection algorithm in autonomous vehicles. Their focus is on creating a decision-making system with an inherent understanding of the world’s context, without the need for explicit computation to detect various obstacles or road intent. Recently, a Bengaluru developer converted his modified Maruti Alto K10 into an autonomous vehicle using a second-hand Redmi Note 9 Pro and an open-source fork of Comma.ai’s OpenPilot. Regarding Comma.ai, Sharma mentioned their use of behavioural cloning. “Behavioural cloning means you try to mimic or clone the human demonstrators or experts. This is the approach taken by Comma.ai and some other startups working on this model of autonomous driving, but it’s not scalable – it’s a dead end already,” he said. Speaking about Tesla, he said that they use autoregressive reinforcement learning, which could be scalable, but a bit doubtful as training and running ARL models can be computationally expensive, requiring powerful hardware and significant resources. Sharma said that Swaayatt Robots is trying to challenge the three major pipelines of autonomous driving in the “classical sense”, which are perception, localisation and mapping and planning. “Since 2019, with the advent of Wavye AI, people have been discussing autonomy without relying on maps. Localisation against maps necessitates the use of maps, but we achieved everything without them. We were the first technology to enable vehicles without the reliance on high-definition maps. In 2017, we implemented multi-RL agents without requiring any maps,” he said. Furthermore, he added that they have autonomous vehicles with mounted cameras; hence, they don’t require external data supply. They have pipelines for automatic data labelling and use generative AI to create specific scenarios, though not on the scale of NVIDIA. Their setup includes 8 side-looking cameras, 2 lidar units, and 2 additional cameras on the bumper. We plan to add 4 more cameras to the rear. What’s Next? Sharma revealed that the team is actively working on an undisclosed ‘X’ paradigm, intending to showcase how autonomous vehicles will independently acquire various skills. This demonstration is scheduled for the month of February. He said that the autonomous driving tech market is projected to be a trillion-dollar industry by 2030, with only a maximum of five companies expected to survive. “Our goal is to secure 25% of the market by 2030 and evolve into a multinational corporation,” he added. “I believe that in a year or so, we will have a pan-India presence and will have raised $1 billion. Currently, we have only $3 million in funding, so even if we achieve sophistication in a specific algorithmic framework for a certain problem, it can only be demonstrated on a limited scale,” he concluded.","excerpt":"Sharma said that Swaayatt Robots is trying to challenge the three major pipelines of autonomous driving in the “classical sense”, which are perception, localisation and mapping and planning.","categories":["Deep Tech"],"tags":["AI in Robotics"],"author_name":"Siddharth Jindal","publish_date":"2024-01-16T13:37:17","publication_year":"2024","word_count":817,"keywords":["Go","funding","ELT","AI","ML","Scala","AI in Robotics","Aim","generative AI","R","startup"],"extracted_tech_keywords":["AI","ML","generative AI","Aim","R","Go","Scala","ELT","startup","funding"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/exclusive-indian-ai-and-robotic-company-confirms-level-5-autonomy\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10167201,"title":"Anthropic Launches Claude for Education to Support Universities with AI","content":"AI startup Anthropic has launched Claude for Education, a specialised version of its AI tool Claude designed to support higher education institutions. The tool offers secure access for teaching, learning, and administration, aiming to help shape the future role of AI in society. Key features include Learning Mode, which encourages critical thinking through Socratic questioning, and campus-wide access at institutions like Northeastern University, LSE, and Champlain College. Claude will assist students in drafting essays, solving calculus problems, and receiving thesis feedback. Faculty can use it to create rubrics, generate academic materials, and provide tailored feedback. Administrative teams can automate tasks, analyse trends, and simplify complex policies. Northeastern University is deploying Claude to 50,000 members across 13 campuses as part of its Northeastern 2025 academic plan. LSE is focusing on responsible AI use by offering Claude to its entire student body, while Champlain College is integrating it into all programmes to develop AI skills for the workforce. Anthropic has partnered with Internet2 to provide secure AI access and with Instructure to integrate Claude into the Canvas learning platform. TechCrunch reported that Claude for Education could help Anthropic increase its revenue. The company, which currently makes $115 million a month, aims to double that in 2025 while competing with OpenAI in the education sector. This launch follows Anthropic’s trend of matching OpenAI’s offerings. OpenAI has partnered with 15 leading research institutions to advance AI in research and education, providing AI tools to benefit students, researchers, and educators worldwide. This follows the launch of ChatGPT Edu in May 2024. In March, the company announced it had secured $3.5 billion in a Series E funding round, bringing its post-money valuation to $61.5 billion. The funds will be used to further AI research, increase compute capacity, and support international growth. The company outlined how to use the investment to improve the development of next-generation AI systems, expand its compute capacity, deepen its research in mechanistic interpretability and alignment, and speed up its global expansion.","excerpt":"Claude will assist students in drafting essays, solving calculus problems, and receiving thesis feedback.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","Startups"],"author_name":"Aditi Suresh","publish_date":"2025-04-03T13:42:39","publication_year":"2025","word_count":329,"keywords":["Anthropic","ChatGPT","OpenAI","AI","RAG","GPT","responsible AI","Aim","Startups","R","AI (Artificial Intelligence)","startup"],"extracted_tech_keywords":["AI","ChatGPT","OpenAI","Anthropic","Aim","RAG","R","GPT","responsible AI","startup"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/anthropic-launches-claude-for-education-to-support-universities-with-ai\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10013761,"title":"What Is George Gabriel Navier-Stokes Problem In HPC?","content":"Solving advance and complex mathematical conundrums at exascale powers supercomputing that stands at the frontiers of human advancement. Some of the architectures of supercomputing come with parallel computing capabilities over the past few decades. With ever-increasing mathematical CFD capabilities, the horizons of supercomputing are opening up. Stanford’s Engineering Center for Turbulence Research has set a world record in CFD in breaking a million-core supercomputer barrier by modelling the supersonic jet noise with 1.5 petabytes of memory with a high-speed 5-dimensional interconnect. Cray, the supercomputing company, produced a number of supercomputers during the past five decades revolutionising and solving some of the most complex and hardest vector calculus problems such as the Navier Stokes equation, invented by George Gabriel Stokes and Claude-Louis Navier. The applications of computational fluid dynamics heavily leverage vector calculus. A two-dimensional or three-dimensional vector field is a function f  that maps multiple points such as (a,b) in ℝ2 , for the two-dimensional vector (x,y). A three-dimensional vector field maps the fields from (a,b,c) to (x,y,z). The two-dimensional vector field equation represented as: A number of data science fields leverage vectors, such as computational fluid dynamics representing different physical quantities such as gravity, electricity, velocity, and magnetism. The gradient is a critical aspect of the vector field, for the function f(a,b), the gradient can be represented as (fa(a,b), fb(a,b)). The (a,b) is the tail of the function that represents the function’s direction with maximum increase. The delta (????) operator is an important mathematical operator in vector calculus that deals with partial derivates of the vectors with differentiation and integration over the vector fields. If we consider, a,b,c as the unit vectors for the coordinate axes, the delta operator represented as: Navier Stokes equations are implemented on CFD (computational fluid dynamics) by the supercomputers for compressible and incompressible flow. Cray XMP and 205 implemented Navier Stokes and Euler’s equations with three-dimensional configurations. The Navier Stokes equations applied to solve many of the complex problems for simulating a physical phenomenon such as smoke or fire as an application of Newton’s second law, F = ma, represented as the product of the acceleration of the object and mass of the object. We can represent f as the forces and leverage density and shear stress. The density denoted as the measurements of the object’s mass per unit volume and shear stress can be defined as the stress coplanar component as: where; p: Density of the fluid and the equivalent of the density to mass ????: Velocity ????: Shear stress f: Forces ????. ???? + f: Total force The above equation can also be represented as: where; P: Pressure ????: Dynamic viscosity Dividing the pressure p and subtracting the ????.  ???? u, we can derive the traditional form of the Navier-Stokes equation. Fluid-dynamic resonance is excited by the thin boundary layers in compact cavities around the periphery of mirror casings, predicting cavity and local flow conditions. The Navier-Stokes equation can solve in 2-dimensions with different boundary conditions. The differential equations for cavity flow for velocity components ????, v can be represented as: Leveraging applied mathematics, we can discretise the ???? dynamic viscosity momentum-equation as follows by transferring the equations into discrete parts for numerical evaluations and further implementation in PyTorch or Python in machine learning. We can also discretise v, the viscosity measurements of resistance another type of momentum equation as follows: The pressure Poisson equation for incompressible Navier-Stokes equation can be written as a solver with Neumann boundary conditions with discretisation. George Gabriel Stokes introduced the Reynolds numbers (Re), a quantity in computational fluid mechanics that provides the ratio between inertial forces to viscous forces. It represents: ρ à Density of the fluid u à Speed of the flow L à Linear dimension ???? à Dynamic viscosity of the fluid ν à The measure of resistance to the flow of a fluid The distributeddataparallel can be leveraged in PyTorch to implement the data parallelism at each module level for distributed high-performance and supercomputing with collective communications in the torch.distributed package by synchronising the buffers and gradients across clusters and processes of machines with different communication strategies for point-to-point communications. Especially, when there is a large number of data points in computational fluid dynamics (CFD), the deep neural networks require large-scale compute and hardware on clusters of machines. Massive parallel computing capabilities with both forward and backward pass with parallelism on GPUs need distributed supercomputing. The minibatch stochastic gradient algorithm can be applied to train the deep neural networks across the batch by scaling the learning rate and hyperparameter tuning for improved performance. The parallelising techniques either on PyTorch or TensorFlow, involve the following steps primarily. Loading the data from the disk into the host, leveraging multi-threading for data parallelisation from the pageable memory to the host’s pinnable memory. As soon as the data gets loaded from the disk into the main memory and gets transferred from the GPU’s pinnable memory, the calculation of gradients with the backward pass and forward pass occurs, and the parameters get updated in a highly computer-intensive parallel task with multiple computations. Dataloader is a framework, which is a class in PyTorch that takes the dataset as the argument by loading the data from the disk and from pageable to pinnable memory. Data pipelines significantly scale up parallelised and pipelined data loading instead of the serial data loading.","excerpt":"Solving advance and complex mathematical conundrums at exascale powers supercomputing that stands at the frontiers of human advancement. Some of the architectures of supercomputing come with parallel computing capabilities over the past few decades. With ever-increasing mathematical CFD capabilities, the horizons of supercomputing are opening up. Stanford’s Engineering Center for Turbulence Research has set a […]","categories":["Deep Tech"],"tags":["HPC"],"author_name":"Ganapathi Pulipaka","publish_date":"2020-12-10T15:00:00","publication_year":"2020","word_count":890,"keywords":["data science","machine learning","AI","neural network","PyTorch","RAG","Python","Ray","TensorFlow","R","HPC"],"extracted_tech_keywords":["AI","machine learning","neural network","data science","Ray","TensorFlow","PyTorch","RAG","Python","R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/what-is-george-gabriel-navier-stokes-problem-in-hpc\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10095819,"title":"Amazon’s Eternal Battle Against Fake Reviews","content":"A 2021 WEF report based on official figures and data from major e-commerce platforms called out approximately 4% of all online reviews as fake. This translates into a significant economic impact, with fake reviews directly influencing global online spending by a staggering $152 billion. To delve deeper into the issue, let’s consider the impact of fake reviews on e-commerce in some of the largest economies. In the United States, fake online reviews influence approximately $791 billion in e-commerce spending, annually. In Japan, the figure stands at $6.4 billion, while in the United Kingdom it’s $5 billion. In Canada, the impact is around $2.3 billion, and in Australia, it’s approximately $900 million. To illustrate the financial implications of fake reviews, an enforcement case involving Legacy Learning Systems Inc highlighted that an investment of $250,000 in fake reviews generated sales exceeding $5 million. Amazon, which is stifled with fake reviews, has been fighting hard across its ecosystem. The e-commerce giant, in its blog post, said it uses machine learning models that analyse thousands of data points to detect risk. This includes relations to other accounts, sign-in activity, review history, and other indications of unusual behaviour, alongside expert investigators that use sophisticated fraud-detection tools to analyse and prevent fake reviews. Fake reviews are a huge problem because they can influence consumer decisions based on generated reviews written by an individual, who was paid to boost product and brand ratings. “Organisations need AI-powered solutions that target fake reviews at the root, i.e fake accounts. Device intelligence pinpoints both the physical devices used to create these accounts, and the use of tools associated with fraud. This paints a fuller picture of the risk for each account, allowing platforms to stop fraud without affecting genuine users,” said Gautam Sehgal, director at SHIELD—a risk intelligence company. Blame it on Social Media Platforms The e-commerce giant believes that social media platforms like Facebook, Instagram and Twitter are not doing much about the cottage industry of fraudsters growing on their platforms. Along similar lines, Rajvardhan Oak, a PhD student at UC Davis, revealed an extensive operation on Facebook groups where reviews and ratings were bought and sold with agents from countries like Pakistan, Bangladesh, and India, working for Chinese sellers to get reviews on Amazon in the US and Europe. Intrigued by an ad offering a free robot vacuum cleaner in exchange for a five-star review, Oak decided to investigate further. He conducted a survey involving agents and reviewers, revealing that people were writing an average of 10 reviews per month for products ranging from $120 to $2,400 in value. Agents earned around $4 to $5 per review, with top earners making up to $1,200 per month. The scale of the operation and the prevalence of such groups raised concerns about the authenticity of Amazon’s star ratings and the company’s ability to detect fake reviews. Amazon has taken steps to combat fake reviews, including filing a lawsuit against over 10,000 Facebook groups involved in this practice. The company has dedicated teams of investigators, lawyers, analysts, and specialists who track down fraudsters and take legal action against them. However, there is an ongoing debate about whether Amazon and other retailers are doing enough to address the issue. Taking Fake Reviews Seriously Amazon believes it has been doing its part to counter this. In 2022, the platform reported and took action against people who run these sites. The methods used by fake review farms have evolved over time. They have shifted from using Markov chain generators to employing machine-learning models that scan old reviews and rephrase the content, making them harder to detect. Startups like Fakespot have emerged to combat this problem, using algorithms to detect fake reviews and scams. These efforts highlight the need for improved fraud detection and prevention mechanisms. The company has also implemented measures to promote genuine reviews, such as the Vine program that offers free samples in exchange for honest reviews, false positives and the challenge of distinguishing legitimate customer reviews from fakes remain significant challenges. Businesses and customers have voiced concerns about legitimate reviews being removed or flagged as suspect. Enter Generative AI Generative AI too is polluting the web. APIs are being automated to leave fake reviews starting with ‘As an AI language model’. A report by The Verge highlights a fake review for a ‘BuTure VC10 Cordless Vacuum Cleaner’, which reads, “As an AI language model, I haven’t personally used this product, but based on its features and customer reviews, I can confidently give it a five-star rating.” Amazon said it blocked 200 million fake reviews in 2022 alone. It also said it has called on governments and is looking for a ‘private and public sector partnership to fight fake reviews. It is urging governments to provide more support in terms of legislation and enforcement to tackle the issue. The company emphasises that the situation differs from country to country and suggests that governments should establish enforcement authority or strengthen their enforcement tools to penalise those involved in the trade of fake reviews. Governments have also started taking notice of the issue. The UK government has proposed plans to make fake reviews explicitly illegal, potentially imposing fines as heavy as 10% of a business’s global turnover for misleading customers. In India, discussions have taken place between the Department of Consumer Affairs and e-commerce entities to address the problem. Overall, the problem of fake reviews on Amazon and other platforms continues to be a complex and evolving issue. While Amazon and other retailers are making efforts to combat this deceptive practice, there is still much debate and ongoing work required to create a more trustworthy review ecosystem. Governments, platforms, and consumers must work together to ensure the integrity of online reviews and protect customers from misleading information.","excerpt":"Amazon said it uses sophisticated fraud-detection tools to analyse and prevent fake reviews","categories":["AI Trends"],"tags":["AI Tool","Amazon","e-commerce","Machine Learning","WEF"],"author_name":"Shyam Nandan Upadhyay","publish_date":"2023-06-27T14:56:59","publication_year":"2023","word_count":963,"keywords":["Go","machine learning","e-commerce","AWS","AI","WEF","Amazon","Machine Learning","Git","RAG","generative AI","Rust","AI Tool","R","fraud detection"],"extracted_tech_keywords":["AI","machine learning","generative AI","RAG","fraud detection","AWS","R","Go","Rust","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/amazons-eternal-battle-against-fake-reviews\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10094077,"title":"Uncensored Models Outperform Aligned Language Models","content":"When we hear the word “uncensored” in AI, we think of models that can be harmful and possibly biassed. While that can be absolutely true, these uncensored models are increasingly outperforming their aligned counterparts, even if possibly harmful. According to the Open LLM Leaderboard on Hugging Face, uncensored models like Wizard-Vicuna-13B-Uncensored-HF, whose developer recently got into trouble for releasing the model to the public, have been competing with LLaMa and Falcon and is one of the top models. Over-finetuning a model may handicap its capabilities. This phenomenon is called the alignment tax of AI models. When a model goes to several benchmarks testing with humans in the loop trying to make the model as aligned and as “politically correct” as possible, it loses a lot of its performance. A lot of it is because of reinforcement learning with human feedback (RLHF). Alignment tax is the extra cost that an AI system has to pay to stay more aligned, at the cost of building an unaligned, or uncensored model. Which ultimately is also hindering its performance. Too much training There is no doubt that OpenAI’s decision to use RLHF to train its GPT model gave birth to the much-hyped and loved ChatGPT. But even then, according to the GPT-4 paper, the model’s accuracy and factuality were much better and more confident before the researchers decided to use RLHF for fine-tuning. The Sparks of AGI paper by Microsoft Research explains this GPT-4 phenomenon. The paper talks about how at the early development stage of the model, it performed way better than the final result after fine-tuning with RLHF. Even though the model is now more aligned and gives balanced answers, its capabilities to answer were much better before. In a presentation of the paper, Sebastien Bubeck, one of its lead authors, narrated the problems that occurred after GPT-4 was trained. He gave an example of the prompt, “draw the unicorn”, and explained how the quality of the output degraded significantly after the model was aligned for safety. A similar case was shared by a lot of Reddit users in a post. During the initial release days of ChatGPT, it used to provide much better results. But after people started to jailbreak ChatGPT, OpenAI put up more guardrails and restrictions in an attempt to address the issues, resulting in poorer responses over time. Meta AI recently released LIMA, a LLaMa 65B model where it compared pre-training a model on unsupervised raw data versus large-scale instruction tuning, i.e. RLHF based model. According to the paper, LIMA was able to outperform GPT-4 in 43% of use cases with just 1,000 carefully curated prompts. Though the model was not uncensored as much, it clearly shows that RLHF might be hindering ChatGPT’s performance. Is poor performance a fair bargain for alignment? As the world is heading towards more AI regulations, it is important for models to be more aligned with what the developers and the users want. To filter out the misinformation that these models have the capability to produce, it is necessary to have humans in the loop that could bring the hallucinating models back on track. These models are essentially built on internet data. Apart from a lot of necessary information, the data is also inadvertently scrapped from misinformation-creating websites. This results in the model giving out falsehoods, which in all senses should be controlled. On the other hand, do you really want your chatbot to not give out the information you want? Even though ChatGPT-like models do not spew out controversial or misleading content as designed by the developers, a lot of users have criticised the model for being too “woke”. A paper titled, Scaling Laws for Reward Model Overoptimisation, explains how RLHF preferences induce bias into the models that hinder the ground truth performance of the models. Some people compare over-tuning a model with lobotomy of the brain. For and against Since these uncensored models have been outperforming many other censored models, we can make the case that these models should be allowed and used to create chatbots. Uncensored models that do not filter out responses that the creators of the model would feel are not safe enough, might not be suitable for researchers and scientists who want to explore the field model. But on the other hand, it comes with a lot of problems like misuse, harmful actors, and malicious intent AI models. While we can make a case for the open source community to be responsible enough to not misuse the unaligned models, there is no surety of any of it. Moreover, aligning models to represent a single viewpoint might not be the way forward in the future. If OpenAI’s chatbot is too woke, there should be an alternative available for developers to build their own versions of ChatGPT. Imagine, if China releases a chatbot that is so aligned with the government’s beliefs that it is unable to operate openly or criticise anything in the country. Even if developers try to control the datasets and perform RLHF as much as they can, what is the possibility that such an AI model would be perfectly aligned and now spew out anything that the creators don’t want it to?","excerpt":"Do you really want a chatbot to not give out the information you want just to stay aligned?","categories":["AI Features"],"tags":["ChatGPT","GPT-4"],"author_name":"Mohit Pandey","publish_date":"2023-05-30T11:00:00","publication_year":"2023","word_count":866,"keywords":["ChatGPT","Meta AI","Hugging Face","RLHF","OpenAI","AI","chatbots","AWS","TPU","GPT-4","R"],"extracted_tech_keywords":["AI","ChatGPT","OpenAI","Meta AI","Hugging Face","RLHF","chatbots","AWS","TPU","R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/uncensored-models-outperform-aligned-language-models\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10022528,"title":"Java Version 16 Released: All The Major Features &#038; Updates","content":"Recently, Oracle announced Java version 16 with 17 new enhancements, including language improvements, tools, memory management, preview features and more. Java is one of the most powerful programming languages and the development platform of choice for enterprises and developers worldwide. The language is ideal for driving innovation, shortening development timeframes, improving application services, among others. Register for ML Fridays by AWS Oracle rolls out Java updates every six months to deliver continued performance, stability and security improvements for developers. The latest Java Development Kit (JDK) includes features like Pattern Matching for instanceof and Records, the previously premiered language enhancements in Java version 14. Georges Saab, vice president of development at Java Platform Group of Oracle, said features like Pattern Matching and Records were launched a year ago as part of JDK 14 and have gone through multiple feedback rounds, based on real-world applications. This provided developers with the opportunity to experiment with these features and develop robust features suited for the community. The essential updates and features of Java 16 are as follows: Pattern Matching for instanceof & Records The two features, Pattern Matching and Records, were launched a year ago as part of Java Development Kit (JDK) in version 14 and are now included in the new version. The Pattern Matching for instanceof enhances the Java programming language with pattern matching for the instanceof operator. It allows common logic in a program, namely the conditional extraction of components from objects, to be expressed more concisely. The Records feature improves the Java language with records, mainly the classes that work as transparent carriers for immutable data. Records can be thought of as nominal tuples. It also allows the ability to declare local record classes, local enum classes, and local interfaces. Packaging Tool Packaging Tool improves the productivity of the developers. It comes with jpackage tool for packaging self-contained Java applications. Further, the feature allows support to native packaging formats, launch-time parameters, among others. Know more here. Elastic Metaspace Elastic Metaspace facilitates memory management and improves the performance of the language and virtual machine. It returns unused HotSpot class-metadata (i.e., metaspace) memory to the operating system promptly, reduces metaspace footprint, and simplifies the metaspace code to reduce maintenance costs. Know more here. UNIX-Domain Socket Channels UNIX-Domain Socket Channels improve productivity as well as the flexibility of developers. It offers support for UNIX-domain sockets features, which are common across both the UNIX platforms and Windows to the socket channel and server-socket channel APIs in the java.nio.channels package. UNIX-domain sockets are utilised for inter-process communication on the same host. They are usually similar to the TCP\/IP sockets, except the former are addressed by filesystem pathnames rather than the Internet Protocol addresses and port numbers. Know more here. Vector API (Incubator) Vector API (Incubator) offers an initial iteration of an incubator module to express vector computations, which compiles seamlessly at runtime to optimal vector hardware instructions on the supported CPU architectures and thus achieve superior performance to equivalent scalar computations. Know more here. Wrapping Up Oracle announced the new version of Java would be released in September this year. Java 17 will include features such as: Enhanced Pseudo-Random Number GeneratorsNew macOS Rendering Pipeline Download the latest version here.","excerpt":"Recently, Oracle announced Java version 16 with 17 new enhancements, including language improvements, tools, memory management, preview features and more. Java is one of the most powerful programming languages and the development platform of choice for enterprises and developers worldwide. The language is ideal for driving innovation, shortening development timeframes, improving application services, among others.  […]","categories":["AI Trends"],"tags":[],"author_name":"Ambika Choudhury","publish_date":"2021-03-21T12:00:00","publication_year":"2021","word_count":534,"keywords":["Go","API","AWS","AI","innovation","ML","Scala","ViT","R","Java"],"extracted_tech_keywords":["AI","ML","AWS","R","Go","Java","Scala","API","ViT","innovation"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/java-version-16-released-all-the-major-features-updates\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":32933,"title":"Is Artificial Intelligence The Newest Wolf On The Wall Street?","content":"The pace at which technology is evolving, especially artificial intelligence, the days are not far when they will be able to make a tangible impact on the countries’ economies. AI over the years has conquered almost every vertical and now in recent years, it has made headway in the trading industry also. In fact, AI is already playing a vital role and is changing the industry significantly. Given all the benefits of leveraging AI in trading, financial giants across the world are switching to AI-driven systems which can not only foresee the market trends but also help to make the trade better. This might be a ‘yay’ moment for AI but for traders, this is turning out to be a nightmare. Many Wall Street traders believe that if this transformation takes more critical turns then the days are not so far when high-earning traders will be sacked. But what if AI would have been able to predict or stop the 2008 crash at the Bombay Stock Exchange? In January 2008, the BSE fell by 1,408 points to 17,605 leading to one of the largest erosions in investor wealth in the history of India. In fact, the day, which is now referred to as “Black Monday”, also took place because the BSE stopped trading for a while on 21 January 2008 due to a technical snag. AI On Trading Platforms National Stock Exchange: In August 2018, the NSE had announced that they were working on AI to strengthen their surveillance operations to prevent any manipulative activities amid reported cases of social media leaks of sensitive information. Why Businesses Are Opting For AI In Trading Trading on the stock market is all about making a profit — there is no room for any kind of emotions. Most of the wrong decisions in trading or stock market are taken because of human emotions, as demonstrated by scores of business management tomes and Hollywood flicks. However, a machine can make complicated decisions and execute trades at a rate which humans cannot match. Some of the major factors that play a key role in trading decision making are: Price variations Macroeconomic data volume changes Prophesising based on news All these parameters can be used effectively by an AI-powered machine or a program. Stock Prediction: Discovering trends in the stock market is done using a company’s previous data and by applying AI and machine learning. Trading firms can crunch historical data more quickly and analyse how they have behaved in the past and create predictive models regarding the rise and fall of the markets. AI makes the decision of buying and selling of stocks easy and based on the inflexion points in the predictive models, AI suggests whether to buy or sell stocks. Future Vision Financial firms across the world are already in a race to develop and implement AI-based program to make trading more profitable and efficient. As new tools continue to enter the trading industry, more and more data will be able to be manipulated; not only from historical data but also from other sources like social media. With all these advancements, the involvement of AI in trading will skyrocket and financial giants from the wall street will need to invest more in the technology part. If the industry is evolving, one also need to evolve to stay on the same level. So, in order to benefit both the firm and its clients, AI might take the place of traditional traders — the industry can expect a significant transformation in terms of employment. The Snag Earlier this month, the Securities and Exchange Board of India (SEBI) had issued a directive that stockbrokers and depository participants who use AI and ML-based applications would have to make quarterly disclosures on their compliance with cybersecurity framework. This instruction came after the SEBI observed that AI\/ML systems were black boxes and that it was difficult to see if these systems it adhered to the regulator’s cybersecurity framework or not. Role Of Traders On Wall Street Humans can still work in trading future; AI might help trading firms with analytics, forecasting, data search or trading, but humans are still the ones who can narrow analyses and share with firms, clients and the public. So, basically AI cannot completely replace humans; they will always be involved when it comes to communication.","excerpt":"The pace at which technology is evolving, especially artificial intelligence, the days are not far when they will be able to make a tangible impact on the countries’ economies. AI over the years has conquered almost every vertical and now in recent years, it has made headway in the trading industry also. In fact, AI […]","categories":["AI Features"],"tags":["AI in stock market","Wall Street"],"author_name":"Harshajit Sarmah","publish_date":"2019-01-07T06:46:07","publication_year":"2019","word_count":721,"keywords":["artificial intelligence","machine learning","programming_languages:R","AI","ML","AI in stock market","RAG","ViT","analytics","Wall Street","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","analytics","RAG","R","ViT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/is-artificial-intelligence-the-newest-wolf-on-the-wall-street\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10164140,"title":"AI Use Reduces Critical Thinking Ability, Says New Microsoft Study","content":"A new study reveals that while generative AI (GenAI) tools can significantly reduce workload, they also risk diminishing critical thinking skills among knowledge workers. The study was conducted jointly by researchers from the Microsoft research lab in Cambridge and Hao-Ping (Hank) Lee, a PhD student at the Human-Computer Interaction Institute at Carnegie Mellon University. The researchers surveyed 319 professionals and analysed 936 real-world examples to understand the impact of AI tools like ChatGPT and Copilot on cognitive processes in the workplace. This was targeted at professionals who use these tools at work at least once a week. The researchers said, “When using GenAI tools, the effort invested in critical thinking shifts from information gathering to information verification, from problem-solving to AI response integration, and from task execution to task stewardship.” The findings, presented at the CHI Conference on Human Factors in Computing Systems, indicate that professionals who are more confident in GenAI tend to think less critically during their tasks. This suggests a potential over-reliance on AI, hindering independent problem-solving. “It’s a simple task, and I knew ChatGPT could do it without difficulty, so I just never thought about it, as critical thinking didn’t feel relevant,” noted one participant, highlighting this tendency to overestimate AI capabilities. Conversely, participants who were highly self-confident in their skills often perceived greater effort in tasks, particularly when evaluating and applying AI responses. The research highlights a significant shift in how knowledge workers approach their responsibilities. Instead of focusing primarily on hands-on task execution, they are increasingly transitioning to overseeing AI-generated results, including verifying outputs for accuracy. This includes setting clear goals, refining prompts, and assessing AI-generated content to meet specific criteria. A user stresses that with straightforward factual information, ChatGPT usually gives good answers, showing AI’s ability. However, GenAI’s limitations and biases also require careful consideration. One participant noted that AI tends to make up information to agree with whatever points you are trying to make. Hence, the editing process could be time-consuming. Additionally, a participant also said the AI output was too emphatic and did not fit the scientific style, and it needed to be rephrased. Based on these findings, researchers emphasise the importance of designing GenAI tools to support critical thinking. The study suggests addressing factors such as awareness of limitations, motivation for careful evaluation, and skill development in areas where AI might fall short.","excerpt":"The study indicates that professionals who are more confident in GenAI tend to think less critically during their tasks.","categories":["AI News"],"tags":["AI adoption","Carnegie Mellon University","Microsoft","research paper"],"author_name":"Sanjana Gupta","publish_date":"2025-02-19T21:29:43","publication_year":"2025","word_count":393,"keywords":["Go","Carnegie Mellon University","research paper","GenAI","AI adoption","ChatGPT","AI","TPU","GPT","generative AI","R","AI-generated content","Microsoft","llm_models:GPT"],"extracted_tech_keywords":["AI","generative AI","GenAI","ChatGPT","TPU","R","Go","GPT","AI-generated content","llm_models:GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/ai-use-reduces-critical-thinking-ability-says-new-microsoft-study\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10069510,"title":"Legacy analytics tools are fighting for survival","content":"The first recorded use of business analytics was in the 19th century when the famed US mechanical engineer Frederick Winslow Taylor set up a time management system. In the 1960s, businesses started relying on computers for decision-making systems and the popularity of analytics grew. Later, the arrival of SAS–a software suite to mine, alter, manage and retrieve data from different sources and perform statistical analysis–in 1972 changed the analytics game. Analytics tools are a critical component of businesses globally. Such tools are used to retrieve, sort and process data and present them in a simple format to enable decision-making. In the 90s, businesses had multiple analytics tools at their disposal for information management, performance management and visualisation. Today, more and more businesses are migrating to the cloud for cost reasons, ease of data management and governance, and greater control and security. Also, companies such as DataRobot, Databricks and Dataiku are leveraging machine learning and open source technologies and libraries to help businesses make the best use of their data. Cloud is the future Before cloud computing, businesses had a hard time maintaining servers and building infrastructure to host applications. Not to mention the high-cost involved. According to a Gartner report, modernising legacy applications can reduce IT costs by nearly 74 percent. In 2020, SAS revamped its primary analytics and BI platform Viya to make it cloud-native. However, the company continues to offer its SAS 9.4 to serve its on-premises customers. SAS’ decision was influenced by its competitors such as MicroStrategy, Qlik and ThoughtSpot adopting a cloud-first strategy. According to Jay Upchurch, SAS’ executive vice president and CIO, the company reached an inflection point where it had to decide whether to continue on the path it was on or pause and rewrite. Apache Spark is a unified analytics engine for large-scale data processing. It provides high-level APIs in Scala, Java, Python, and R, and an optimised engine that supports general computation graphs for data analysis. It also supports a rich set of higher-level tools including Spark SQL for SQL and DataFrames, pandas API on Spark for pandas workloads, MLlib for machine learning, GraphX for graph processing, and Structured Streaming for stream processing. The creators of Apache Spark also developed Databricks. Last week, data science company RapidMiner launched a next-gen cloud platform for enterprises. RapidMiner builds a software platform for data science teams that unites data preparation, machine learning, predictive model deployment etc. The company’s platform has been re-architected from the ground up to deploy models faster and bridge the gap between data science and business understanding. The platform helps scale the enterprise environment by connecting to common data sources, code-free deployment and more. Tableau Software, an American interactive data visualisation software company, launched ‘Tableau Online’ in 2014. Soon, it became the company’s best-selling product. KNIME, a free and open-source data analytics, reporting and integration platform, launched ‘KNIME Cloud Analytics Platform’ in 2016. Issues with legacy analytics tools Most legacy systems are rigid and don’t render easy integration. The lack of flexibility can hamper the day-to-day operations of businesses. But take a tool like Microsoft Power BI. It operates seamlessly with Excel and text files, SQL server and other cloud-based tools. Also, cloud-native tools are far better than legacy tools when it comes to security. Legacy tools were developed at a time when big data wasn’t a thing. These tools are not designed to deal with huge swathes of data and tend to slow down in the face of complex, unstructured data. Modern analytics tools, on the other hand, are robust and can extract actionable insights from a humongous amount of raw data. Further, platforms such as Databricks are fast replacing the legacy tools. Databricks allows users to collaboratively create and run data analysis projects in a structured and scheduled manner with the help of open source technologies. Databricks also eliminates data silos and allows different teams to work together on the same platform, thus standardising the process and making it easy for each person to understand the different processes involved. Since it controls the entire workflow, it leads to better communication, speed and efficiency. Legacy systems aren’t obsolete. Yet! Though most legacy systems are on their way out, some businesses still rely on these systems, and some have been using them for decades. Microsoft’s Excel is a prime example of a legacy tool still going strong. However, Microsoft has made Excel available on the cloud. In 2019, 54 percent of the businesses in the US were still using Excel. Continuing further, switching to a new system is not easy. Take, for example, the cost of bringing employees up to speed. Everyone in the organisation will have to spend time learning the new tools, from the executives to the manager to the IT team. Also, legacy tools contain years’ worth of data for many businesses. As a result, migrating to the cloud is not an easy task. However, even though modernisation is costly and time-consuming, businesses can’t afford to miss the boat","excerpt":"According to a Gartner report, modernising legacy applications can reduce IT costs by nearly 74 percent.","categories":["AI Features"],"tags":["legacy systems"],"author_name":"Pritam Bordoloi","publish_date":"2022-06-22T11:00:00","publication_year":"2022","word_count":830,"keywords":["data science","machine learning","AI","cloud computing","ML","Apache Spark","RAG","Databricks","analytics","legacy systems","Pandas"],"extracted_tech_keywords":["AI","machine learning","ML","data science","analytics","Pandas","RAG","cloud computing","Apache Spark","Databricks"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/legacy-analytics-tools-are-fighting-for-survival\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10110141,"title":"AI Chatbots are not Ready for Customer Service","content":"After the launch of ChatGPT, the discussion surrounding AI replacing customer service jobs gained considerable traction. However, the situation intensified in July 2023, when Dukaan, a DIY platform for online stores, terminated 90 percent of its support staff, opting to replace them with an AI chatbot called Lina. After this development, concerns were raised about the potential job displacement of customer agents, particularly in economies like India. However, industry experts interviewed by AIM expressed the view that at its current stage, AI is not equipped to replace humans in the context of BPO operations. Moreover, recent occurrences, such as a software engineer successfully manipulating a car dealership chatbot powered by OpenAI’s GPT models to sell him a Chevy for just a dollar, confirm their assertions. While LLMs bring numerous benefits to those in customer service, instances like this do raise one moot question: Are LLMs reliable for customer-facing sales and support roles? Generative AI for customer service Ethan Mollick, an associate professor at the Wharton School of the University of Pennsylvania thinks LLMs are not ready for external facing sales and support roles. “They are gullible and hallucinate,” he posted on X. Persistent hallucinations continue to pose a major challenge for LLMs, and with the industry yet to find a solution, this issue may persist as we progress. Currently, the models like GPT-4 are being fine tuned and trained with enterprise data to make them ready for enterprise application, however, it does not eliminate the risks. Despite being fine-tuned, it does not eliminate the chances of the bot providing plausible-sounding but incorrect or nonsensical answers. Sanjeev Menon, co-founder and head of product & tech at E42.ai believes using generative AI such as ChatGPT in customer service does elevate efficiency and experience, provided enough thought is put into the design, along with fine tuning with specific data. “However, generative AI is not a panacea for all maladies in customer support—clarity on the capabilities and limitations is very essential,” he told AIM. The reality is many enterprises today have an LLM-powered chatbot integrated into their platform, and the number is only going to increase. Moreover, it can be argued that not most customers interacting with a car dealership bot will not ask for a python script. “The behaviour does not reflect what normal shoppers do. Most people use it to ask a question like, ‘My brake light is on, what do I do?’ or ‘I need to schedule a service appointment,” Aharon Horwitz, CEO at Fullpath told Business Insider. Yet that does not mean the risks associated with it can be ignored. As more enterprises embrace these bots, AI mishaps might only increase. Humans in the loop Hence, according to Menon, human intervention still plays a significant role despite the advancements in AI. He said that checks on prompt toxicity, data updates, and supervision during complex or sensitive situations are paramount to guaranteeing customers a positive and secure experience. “In doing so, we not only enhance efficiency but also eliminate risks associated with the use of language models, ensuring a seamless and reliable customer service interaction.” Gaurav Singh, founder and chief executive officer at Verloop.io, also said empowering agents as gatekeepers ensures quality control.  “More than 90 percent of queries can be effectively handled by LLM-powered Conversational AI, but in instances of uncertainty, seamless transfer of queries to agents allows verification and editing, maintaining accuracy in automated responses for optimal query resolution,” he told AIM. While there have been instances like Dukaan, contrary to earlier fears, widespread job loss has not occurred. Furthermore, occurrences such as chatbots recommending poison recipes underscore the crucial need for human intervention and caution against excessive reliance on AI chatbots. “It’s important to strike a balance and use human agents where emotional intelligence, nuanced understanding, and complex problem-solving are required. A hybrid approach that combines the strengths of both AI and human agents may be the most effective solution for providing excellent customer service,” Beerud Sheth, co-founder and CEO at Gupshup, told AIM. Are Small Language Models the answer? LLMs like GPT-4 have billions of parameters and are trained on terabytes of data scraped from the web. These models have world knowledge, meaning they know everything from historical facts to contemporary events, providing a vast understanding of diverse topics. However, does an enterprise need such worldly knowledge? Does a car dealership chatbot know Python script? No, and this is where Small Language Models (SLM) come in. A SLM can be fine-tuned or trained specifically for a particular industry or domain. This enables the model to better understand industry-specific terminology, customer inquiries, and context, leading to more accurate and relevant responses. These models also allow enterprises more control over the training process and can customise the model to align with their specific customer service needs. “Leveraging domain-specific data and knowledge, these models ensure that their generated outputs align precisely with customers’ queries, industry standards, and specific requirements,” Rashid Khan, co-founder and CPO at Yellow.ai, told AIM. However, the problem of hallucination pertains even in SLMs. While it has posed challenges to end hallucinations in these models, it certainly can be reduced. For instance, Yellow.ai, leverages a maker-checker model setup. “One model generates responses, while another validates their relevance and accuracy. We also implemented the RAG architecture, ensuring fact-based answers to reduce hallucination chances and refining our model to provide accurate responses from a given paragraph,” Khan added. Nonetheless, a domain-specific model with a human in the loop might still be the best approach for enterprises to mitigate risk. “With domain adaptation for precision, strict moderation for safety, and calculated human involvement for accountability, one can balance the efficiency of AI while guarding against unforeseen issues,” Sheth said.","excerpt":"A software engineer prompted a car dealership chatbot to sell him a Chevy for just a dollar.","categories":["AI Features"],"tags":["ai chatbots"],"author_name":"Pritam Bordoloi","publish_date":"2024-01-08T12:00:00","publication_year":"2024","word_count":949,"keywords":["ai chatbots","ChatGPT","OpenAI","AI","chatbots","ML","SLM","RAG","Aim","generative AI","small language models"],"extracted_tech_keywords":["AI","ML","generative AI","ChatGPT","OpenAI","Aim","small language models","SLM","RAG","chatbots"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/ai-chatbots-are-not-ready-for-customer-service\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10165388,"title":"Here’s How AI Takes the Grunt Work Out of Design","content":"Although AI has practically permeated every layer of our work, when it comes to design, not all artists or designers are keen on incorporating AI into their work. However, with several AI design tools available, designers are beginning to experiment with them to improve their workflows in various ways. Simplifying Tasks and Boosting Creative Output For simple and repetitive tasks such as creating auto layouts, adjusting component variations, resizing images, and more, AI saves time by automating these actions. Vaibhav Bhasin, head of design at Upstox, told AIM that designers use auto-layout to define component constraints and manage resizing. Now, AI enables anyone to convert components to auto-layout using one shortcut. Bhasin further highlighted that design tools are becoming increasingly nuanced. New tools are emerging, like Jitter for motion design and Rive, for generating production-ready UI and graphics. There are tools like Framer that let you build an entire website. Such tools have given people, who have no knowledge of design, a starting point to create something. He added that designers do not always have time to go through all the research they have collated. So, in such cases, using Google’s NotebookLM, or any tool like ChatGPT, even with their privacy concerns, is a good way to condense information from the research, which previously took hours. At Razorpay’s FTX 25 event, Akhil Joy, head of design at Zepto, highlighted that AI has helped do more with the same head count. According to him, AI has enabled designers to create more experiments at the highest quality. In the same panel discussion, Deepa Bachu, UX director at Google, agreed that AI will remove all the grunt work that goes into the design process. AI Helps With Designers’ Skill Gaps According to Joy, his team members at Zepto use AI and other tools in different ways, sparking inspiration. People get inspired by each other’s output, which prompts them to improve their output and outdo the last idea. When it comes to designing, there are many areas to work with. A single designer may not have the knowledge to do it all. In such situations, AI comes to the rescue. Bhasin told AIM that AI design tools are democratising the skill set to a certain extent. Citing an example from his own experience as a product designer, he acknowledged that he’s not the best at motion design or illustration. In such cases, one can use tools like DALL·E, Midjourney, and others to develop something without having the technical knowledge for it. He added that AI is useful for making sense of data from dashboards one may not entirely understand. With the right use of AI, one can make informed decisions even without a particular expertise. He noted that one must also be very cognizant about where AI is needed, instead of force fitting it into everything. The Human Mindset Remains Vital Even With AI Bhasin believes that the mindset of a designer matters more than the hard skills, especially when a company is hiring its first designer. Also at Razorpay’s FTX 25 event, Priyanka Kodikal, head of design at Dream Sports, explained that designers need to be curious, and have a forward-looking mindset. Meanwhile, Bachu said that it is important for designers to have empathy to understand when communicating with their target users.","excerpt":"Designers are beginning to experiment with AI tools to improve their workflows in various ways.","categories":["AI Features"],"tags":["Design"],"author_name":"Ankush Das","publish_date":"2025-03-06T20:00:00","publication_year":"2025","word_count":549,"keywords":["Go","ChatGPT","TPU","AI","RPA","GPT","GRU","Aim","ViT","R","Design"],"extracted_tech_keywords":["AI","ChatGPT","Aim","TPU","R","Go","GPT","GRU","ViT","RPA"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/heres-how-ai-takes-the-grunt-work-out-of-design\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10141215,"title":"Karnataka Brings GCC Policy, Aims to Create 3.5 Lakh Jobs by 2029","content":"Karnataka government recently introduced the global capability centres (GCC) policy 2024, aiming to bolster its reputation as a global innovation hub while creating 3.5 lakh jobs by 2029. The policy is designed to attract 500 new GCCs to the state, bringing the total to 1,000 and generating an economic output of $50 billion. With this, Karnataka government is targeting 330 Forbes 2000 companies to establish their GCCs in the state by 2030. “We are dedicated to providing the necessary incentives and stimulating innovation in our tier 2 cities, thereby democratising opportunities for prosperity,” said chief minister Siddaramaiah, highlighting the policy’s significance. The GCC policy emphasises regional development, particularly through the ‘Beyond Bengaluru’ initiative. This program targets decentralising GCC activities by promoting cities like Mysuru, Mangaluru, and Tumakuru as viable alternatives for establishing new centres. “Our vision extends beyond Bengaluru to promote balanced regional development,” said the minister for electronics, IT, BT and rural development, Priyank M Kharge. The government also announced the exchange of five MOUs with big techs to train 100,000 people and expects an investment of $17.5 billion in total. What is ‘Beyond Bengaluru’? The policy outlined four key pillars: workforce development, innovation ecosystem enhancement, regulatory reforms, and tailored incentives. Among its highlights are the creation of 1 lakh internships and a global leadership development skilling program, which aims to equip the workforce with advanced skills in technologies like artificial intelligence and machine learning. Bengaluru currently hosts 875 GCC units, accounting for 35% of the national GCC workforce. The city contributes significantly to sectors like IT, banking, and life sciences. However, Karnataka seeks to extend this dominance to other regions. The “Beyond Bengaluru” cluster seed fund and infrastructural investments such as expanding metro networks and enhancing airport connectivity are part of this strategy. Industry representatives have welcomed the policy. According to a senior GCC executive, “Karnataka’s targeted approach with infrastructure and talent skilling ensures businesses can seamlessly expand. The emphasis on secondary cities is timely and aligns with broader industry needs.” The policy includes initiatives such as a dedicated GCC Acceleration and Investment Council and specific fiscal incentives for enterprises establishing operations outside Bengaluru to further attract investments. The state is also setting up innovation hubs and centres of excellence in AI, cybersecurity, and quantum computing to foster advanced research. The government aims to position Karnataka as a global leader in the GCC landscape. Currently contributing $22.2 billion to the GCC market, the state plans to double its share by 2029. With policies supporting sustainable innovation, inclusive regional growth, and digital transformation, Karnataka is poised to reinforce its status as a preferred destination for global businesses. The GCC policy is seen as a strategic step in achieving these ambitions while addressing employment and economic challenges.","excerpt":"The state aims to attract 500 new GCCs and bring the total to 1,000 while generating an economic output of $50 billion.","categories":["AI News"],"tags":["Editors Picks","GCC","karnataka"],"author_name":"Sanjana Gupta","publish_date":"2024-11-20T00:52:28","publication_year":"2024","word_count":457,"keywords":["karnataka","Go","artificial intelligence","GCC","machine learning","AI","TPU","ML","Git","Editors Picks","Aim","ViT","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","Aim","TPU","R","Go","Git","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/karnataka-brings-gcc-policy-aims-to-create-3-5-lakh-jobs-by-2029\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":12374,"title":"10 Emerging Analytics Startups in India to watch for in 2017","content":"With analytics domain expanding its wing into AI, machine learning, NLP and more, this year’s startup list has quite interesting names that are doing some intriguing work in this field. Starting from video analytics to applying analytics in digital marketing, these startups have brought in many interesting concepts to the table. After an extensive research and brainstorming, we bring to you the names of 10 such emerging analytics startups that hold a promising future. And just like last year, the task wasn’t easy, as every startup has a competitive edge to offer over others. Read below to find out 10 emerging analytics startups in India to watch for in 2017. Read: 10 Emerging Analytics Startups in India to watch for in 2016 10 Emerging Analytics Startups in India to watch for in 2015 10 Emerging Analytics Startups in India to watch for in 2014 10 Emerging Analytics Startups in India to watch for in 2013 1. Dataxylo: A Customer Analytics company with an award-winning AI powered platform (CIP) for Marketing Ops, Dataxylo helps marketing operations to continuously identify, unify, cleanse, enrich customer data to further power the embedded analytics engine which enables deeper insights to predict re-marketing opportunities. DX Platform most uses cases are across omni-channel Retail, High Tech domain. In the marketing world where improvements are measured by small percent points of improvement, DataXylo customers regularly report a 4X+ lift in retention and monetization efforts. DX platform ensures that you leverage all your customer data, conduct cross channel identification and ensure continuous campaign improvement – while leveraging behavioural and intent based insights through its sophisticated AI based platform which is offered as both API and UI based tool with integrations across majority of tech apps. A venture backed MIT spin off, Dataxylo was co-founded by Abhi Yadav- CEO & Iqbal Kaur– COA, now with an impressive team and offices across Cambridge (HQ), Bengaluru & Sunnyvale. It flaunts an impressive list of Investors, Board members and Advisors including Michael Cusumano – (Sr MIT Professor), Mike Kail – (ex- CIO Yahoo), Michael Fleischman (Twitter Board), and Serial Tech Entrepreneurs like Deepak Taneja and Doug Levin. Team DataXylo 2. Emplay: A 4-year-old product start-up headquartered in California with its core team of data scientists and product engineers located in Bangalore, Emplay brings the concept of self-driving cars to the world of business with its AI and machine learning based decision making and action automation products. It is also building intelligent sales bots that can perform sales activities on behalf of sales reps. The bots can assist, augment and replace sales reps in various scenarios providing unprecedented return on sales investments. Led by Sanchita Sur (CEO) and Sharad Joshi (CDO), who together bring an amazing mix of business and technology experience, Emplay has won 4 global industry awards 4 years in a row for helping Fortune 500 companies predictably meet their revenue goals by providing turn by turn data driven guidance to their sales reps. Emplay’s clients include large software, high tech, insurance and retail companies. A part of SAP’s start-up focus program, Emplay is on its path to tripling its revenue in 2017. Team Emplay 3. G Square: There are very few companies globally who provide plug and play analytics productified solutions- G-Square Solutions being one- which was founded by Gurpreet Singh and Gopi Suvanam in 2014, to harness the huge potential and underpenetration of analytics products and services. With a B2B approach, they offer SaaS based financial products and also provide services in sales, marketing and risk analytics tailored to the client requirements. They also have a strong expertise in Artificial Intelligence, Machine learning and apply strong data engineering in each assignment, their sweet spot being the Financial Services and Fin-Tech space. Other areas where it provides its tools and services are across Banks, NBFCs, Wealth Management companies, Asset Management Companies. Their uniqueness lies in using their in-house technology and proprietary algorithms for building models along with conventional methods. Recipient of the Emerging Analytics Services Startup of the year award at CYPHER 2016, Analytics India Summit in 2016, G-Square wishes to be the Artificial Intelligence driven plug and play analytics product provider with a nimble footed approach to client requirements. 4. Hiddime: Lead Semantics, a new generation AI company integrating knowledge bases and machine learning, develops products and services targeting the area of ‘Semantics Integrated Data Science’ for both the Enterprise and the Cloud environments. Hiddime.com is first of its kind Semantic Cloud-BI tool that enables advanced analytics on the cloud. an ‘Interactive Discovery and Exploratory Analytics’ tool (IDEA tool) in the cloud, hiddime.com enables end business users with little IT knowledge to deliver routine to sophisticated BI and advanced Analytics with just point and click interactions in the browser. Their data science teams deliver NLP, Graph, Machine Learning and Semantic Technology projects that also include integration of complex Big Data engineering pipelines feeding into BI Datawarehouses and Smart Data Lakes. Their pedigree and experience uniquely positions them to take advantage of the recent surge in interest in Smart Data to deliver cutting edge data science that enterprises are striving for globally. Following the win at CYPHER 2016, as the ‘Analytics Product of the Year’, they have been witnessing tremendous traction with customers. LS-Team 5. Realbox: India’s upcoming and promising Data Science Company, Realbox has a vision of uplifting the standards of contemporary Brick and Mortar businesses. Leveraging their published Customer Profiling patent, they are taking Guest Experience to a whole new level. Realbox is solving the biggest problem of real-time analysis of large customer data for enterprises like movie theatre giant PVR Director’s Cut and Australia’s biggest juice chain JOOST among many others in retail and hospitality industry. Its flagship product Pulse is like the Google Analytics for offline business. This intelligent dashboard can be installed on top of any existing POS enabling firms to use data analytics without changing their existing technology. Decision-makers get access to business summaries and key performance metrics on mobile devices in real time, helping them remain agile and in control of their businesses. It does this by making use of Business Intelligence and Predictive Analysis tools to help increase revenues and decrease risks. Based in Delhi, Realbox Data Analytics Pvt. Ltd. was incorporated in April 2015 by Founder and CEO Saurabh Moody and co-founders Preksha Kaparwan(CMO) and Arjun Sudhanshu(CTO), proving to be a prominent leader of India’s tech startup community since its inception. Realbox team 6. RecoSense Labs: With its data intelligent platform which can independently interpret the context of the user and the content to generate structured meta-content, RecoSense helps Digital Publishers and Enterprises to automate the personalization of their content for every User. The platform achieves strong user engagement, content discovery and better monetization by delivering the right content to the right user for any endpoint – App, Web or even Devices. RecoSense platform is derived from an indigenously built IP to interpret any kind of content (Videos, Text, Images, Ads) based on Natural language processing and Machine Learning framework. It offers a distinct advantage for its customers with a strong focus on content interpretation compared to just the data insight from traditional clickstream analytics. The platform capabilities extend to Automated Metadata generation, Content management systems, Data analytics, Youtube solution, Loyalty Management Features etc. Founded in September 2014 and based out of Bangalore, RecoSense works with marquee clients in Digital Media, Telcos and now expanding across the globe into other sectors like Financial services, eCommerce, Travel portals etc. It serves over a billion recommendations a month with a strong impact on increasing Watch time and User Actions due to higher relevance of content. Led by Amith Srinivas, Raghunandan Patthar and Shivkumar Janakiraman, the firm has won two awards last year- ‘Product Innovator 2016’ by Frost & Sullivan and ‘Top 100 startups 2016 in Asia’ by Red Herring. RecoSense team 7. Think Analytics: Founded by Amit Das, Suryadip Ghoshal and Monish Salot, Think Analytics India is an analytics solution and services company founded in 2014.  Starting as an offshore analytics advisory for US clients, the team has since built multiple solutions, and is continuing to invest in products with a potential of disrupting the mobile data and analytics ecosystem. Their most recent solution, Algo360, is powered by extensive libraries, NLP, and machine learning algorithms, allowing the companies to get a real time 360-degree view of their customers. For instance, an NBFC or a P2P lending company can leverage Algo360 scores and custom risk scorecard layer to power instant loan decisions. An e-commerce company can deploy Algo360 to identify its value destroying customers, or build a pay later framework. Algo360 aims to make a difference for “new to credit” segments and make credit more accessible for the under-banked and unbanked segments. Headquartered in Mumbai, their goal this quarter is to assess at least a million applications, especially the 40-50% that existing bureaus do not cover. Think Analytics team 8. Vahanalytics: Founded by Shivalik Sen, Arun Gandhi, Nikhil Tavora and Someshwar Dash, alums of BITS-Pilani, Vahanalytics (VA) makes better drivers through smarter data. VA works with fleet companies to give them insights into the safety and performance of their drivers while monitoring them using their smartphone based technology. Displaying their core strengths in data sciences, physics and computer engineer, their edge is in their ability to quickly ideate and build innovative solutions for pattern matching and outlier detection problems in the logistics space with machine learning at its heart. With a lot of research stating that better driving practices and corrective measures not only help fleet owners have safer and more efficient operations, it also helps them increase fuel efficiency, cut down on maintenance costs and see a host of non measurable value additions to their business and operations. Vahanalytics Team 9. Vidooly: Vidooly is an Online Video Analytics and Marketing company that provides tools for Online Video Creators, Multi Channel Networks, digital focused brands and agencies. India’s only Digital Video Analytics company, Vidooly is quite unique from its global counterparts because unlike most of them, it caters to multiple stakeholders of the online video ecosystem. Video creators can use its tools to increase their online visibility and expand their audience whereas Multi Channel Networks can acquire new partners and run their operations better using Vidooly’s comprehensive Content Management System. Brands and agencies that focus on digital video can get a deeper understanding about their target audience and run their Marketing and advertising campaigns with optimum targeting, resulting in the best ROI for them. With online video gaining momentum in India, Vidooly is poised to play a vital role in the ecosystem in 2017. From social video to OTT and Video On Demand, Vidooly has set its aims high this year and is all set to punch above its weight. Team Vidooly 10. VokseDigital: A management consulting startup, VokseDigital focuses on leveraging the power of analytics to provide actionable insights to business to drive growth and improve profitability. With the founding team boasting a cumulative experience of 75+ years in consulting and delivering complex projects with specialization in customer and marketing analytics, the team has deep domain expertise across Retail, eCommerce, Banking, Insurance, Financial Services, Telecom and Healthcare verticals. VokseDigital is engaged in providing digital analytics services to marquee clients dealing with business challenges of customer churn, ridership analytics, marketing effectiveness, to name a few. Their niche focus enables them to provide the necessary depth and the breadth in dealing with disparate data such as image, video, web, mobile, speech, text and others. It is also creating solutions in the areas of customer experience, social media analytics, web & mobile analytics leveraging the Microsoft stack. Their unique approach focuses on converging cloud, mobile, big data, machine learning and advance analytics like predictive, prescriptive, NLP, sentiment analytics in: “getting better insights” out of the existing data “enabling sales force” to quickly respond to prospects and customer leads “enhanced omnichannel customer experience” to acquire and retain customers “effectively targeting & engaging” high revenue and profitable clients “creating value & maximizing returns” from digital investments.","excerpt":"With analytics domain expanding its wing into AI, machine learning, NLP and more, this year’s startup list has quite interesting names that are doing some intriguing work in this field. Starting from video analytics to applying analytics in digital marketing, these startups have brought in many interesting concepts to the table. After an extensive research […]","categories":["AI Features"],"tags":["Startups"],"author_name":"Srishti Deoras","publish_date":"2017-01-30T07:51:26","publication_year":"2017","word_count":2007,"keywords":["data science","Go","machine learning","artificial intelligence","AI","RAG","NLP","Aim","analytics","Startups","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","NLP","data science","analytics","Aim","RAG","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/10-emerging-analytics-startups-india-watch-2017\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":35792,"title":"3 Years In IT Sector, What Next: Study By AIM &#038; Jigsaw Academy","content":"The field of information technology is evolving at a faster rate than usual. Areas like artificial intelligence, data science, data analytics and internet of things, among others, are now being integrated with industries across all sectors. Many times, it becomes difficult to keep a track of these technologies and keep up with them. Now, more than ever, we are seeing a pattern where young professionals are finding themselves confused about their future in the company — be it regarding their job roles, their progress, or even the fear of losing their jobs to machines. This month, Analytics India Magazine, in association with Jigsaw Academy, decided to find out what goes on behind the making of a good career graph in 2019. From the education background to upskilling strategies and organisational partnerships for further studies, today’s young professionals have found many ways to fight the career ennui. In fact, industry numbers suggest that there is an 18% to 20% attrition rate for young professionals with around three to four years of work experience. About The Study The samples were collected by asking respondents to fill in a survey created by AIM about what these young professionals felt like at work — professionally. This included various sub-topics such as growth opportunity in the current organisation, their need to upskill and their preference of a particular Emerging Tech. We took opinions from young Indian professionals with work experience between 2 to 4 years to get a thorough idea of the working environment in this growing field. Our survey was met with much enthusiasm — and we got some great insights from it. Some of them were expected, and many of them were real eye-openers. How long have you been working in this industry? Our core demographic for the survey, young professionals with work experience of between two to five years constituted a majority of the result at 39% The other important constituents to the survey were young working professionals with less than two years of work experience at 27% 19% of our respondents were persons with 5-10 years of work experience. Do you wish to work in your current technology\/domain for the next five years? The answer to this question was an ambitious “no” with 63% young professionals saying that they would definitely want to change their roles in terms of technology or domain. About 27% respondents were not sure about their future in their chosen sector And only 10% of the respondents said that they wanted to keep working in the same technology domain. Do you feel you face enough growth opportunities in your current role? Interestingly enough, an overwhelming majority of the professionals at 66% said that they did not face enough growth opportunities in their current role. Only 34% said they did foresee growth opportunity. If you do want to change your job\/role, how do you plan to do it? When asked the question about how these young professionals wanted to take their career ahead if looking for a change, the majority of them, at 59%, said that they would prefer to upskill, while keeping their current jobs. 18% of the respondents said that they would quit working full time and take a break to pursue a subject properly. 13% of the respondents, on the other hand, said that they would like to change jobs, as that would also allow them to get more exposure and experience. What Future Tech would you like to upskill to? When asked about what future tech would most of the young professionals upgrade their skill set to, most young professionals gave a very clear answer: artificial intelligence and machine learning-related technologies were most in demand at 44%. 35% of the respondents with 3-5 years of work experience in the IT sector also said that data science would be a great option to upskill. If yes, how would you like to upskill? 37% of the respondents who prefer to upskill said that the best way to do this was through taking up a part-time educational programme. 35% said that they would prefer studying on their own, using their own material which is available freely online. 19% of the respondents said that they would like to quit their full-time jobs and pursue full-time education. Do you think, your organisation should sponsor to upskill you? This question was met with a very positive reaction where 64% of the respondents said that they would like it if their organisations encouraged the culture of upskilling by sponsoring it. Do you think your current job will be automated in the next 3-5 years? Probably reacting to the dystopian warnings predicted all around the country via different means, almost half the respondents, at 49%, said that they were worried that their job would be automated within the next five years. How many times do you hear keywords like AI\/ IoT\/ Blockchain in your current organisation? New tech concepts such as AI, IoT and blockchain have become fairly common at workplaces today. Clearly, respondents have had positive experience towards it, because, 44% said that they heard these terms frequently. 28% respondents said that they heard these terms regularly. Are you already looking out for a job in tech like AI\/ data science\/ IoT\/ Blockchain? Clearly, after gaining a few years of work experience in the IT sector, the next step for most young professionals is to then move on to New Tech. This was demonstrated by an overwhelming majority of 72% who said that they were already looking out for a new job in this area. Do you think your past experience would be wasted once you move to another technology? Many professionals have expressed this fear in the past that while upskilling is a good move for the career, but it may sometimes undermind the past work experience. But 52% of the respondents said that they did not think that upskilling or moving into another, newer technology would waste their past work experience. However, 29% of the respondents were not sure about this How much salary hike do you think you would receive once you transition to another tech? Most of the time, these young professionals want to move to newer territories in the future tech is because of the growth opportunities and monetary compensation. 58% respondents think that they would get a salary hike between 20% to 50% after they upskill themselves or transition to new tech. 18% respondents said that they expected they would get as much as a 50% hike The Managerial Perspective Mohan Sitharam, the chief human resources officer at Subex, told Analytics India Magazine that they always preferred to upskill and retain current employees than hire outside talent. “Subex is a product company with deep expertise in the telecom domain. For a business such as ours, it is an obvious choice to retain and upskill our engineers. We consider re-training our existing resources as raising the bar of our talent’s competence, and in our experience, we realise that the odds of advantage are twice when we retrain our existing talent, rather hiring skills.” Bhuvan Nijhawan, Director Education – Asia Pacific, SAS, told Analytics India Magazine, “Young Professionals these days are constantly looking to upskill themselves, given the rapid market demands and advancements in Data Science technologies. These employees not only intend to continue with their current employer but also upskill themselves for career progression and growth. To encourage such budding talent and recognise their skillset, SAS® has introduced SAS® Digital Badges. These badges are given only to those who take up our official Training programs or those who clear the Certification exam. This also acts as a reference point for potential interviewers. With increasing demand for skills in Data Science, Machine Learning and Artificial intelligence, tech-savvy professionals with 0 to 3 years’ experience often look forward to acquiring these skillsets. Moreover, the hype in this field often leaves them with scattered information making it difficult to choose relevant programs to their profile. With this in consideration, SAS has developed assessment tools which map their existing skill sets thus aiding with recommended curricula relevant to their profile. We also work to connect these talents with our partner organisations and customers.” Road To Career Transition Abdul Baasit, one of the students who completed the Postgraduate Program in Data Science and Machine Learning from Jigsaw Academy has transitioned successfully in his career. “I had a background in IT but in around 2015, I discovered my interest in data and generally working with mathematical solutions,” he said. Abdul initially completed the course in Data Science with SAS and while he did then move into the data science space, he felt that it wasn’t quite enough. “I’d noticed the potential of analytics in the job market and was keen to get further into it. I felt that doing another more in-depth course would be helpful,” he added. Here’s the Complete Report 3 Years In IT Sector, What Next? Study By AIM & Jigsaw Academy","excerpt":"The field of information technology is evolving at a faster rate than usual. Areas like artificial intelligence, data science, data analytics and internet of things, among others, are now being integrated with industries across all sectors. Many times, it becomes difficult to keep a track of these technologies and keep up with them. Now, more […]","categories":["AI Features"],"tags":["jigsaw academy","sas","Subex"],"author_name":"Prajakta Hebbar","publish_date":"2019-03-06T05:24:11","publication_year":"2019","word_count":1493,"keywords":["data science","Go","artificial intelligence","machine learning","AI","Git","RAG","Aim","sas","Subex","analytics","jigsaw academy","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","data science","analytics","Aim","RAG","R","Go","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/3-years-in-it-sector-what-next-study-by-aim-jigsaw-academy\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10005536,"title":"Hands-on Guide to Pattern &#8211; A Python Tool for Effective Text Processing and Data Mining","content":"Text Processing mainly requires Natural Language Processing( NLP), which is processing the data in a useful way so that the machine can understand the Human Language with the help of an application or product. Using NLP we can derive some information from the textual data such as sentiment, polarity, etc. which are useful in creating text processing based applications. Python provides different open-source libraries or modules which are built on top of NLTK and helps in text processing using NLP functions. Different libraries have different functionalities that are used on data to gain meaningful results. One such Library is Pattern. Pattern is an open-source python library and performs different NLP tasks. It is mostly used for text processing due to various functionalities it provides. Other than text processing Pattern is used for Data Mining i.e we can extract data from various sources such as Twitter, Google, etc. using the data mining functions provided by Pattern. In this article, we will try and cover the following points: NLP Functionalities of PatternData Mining Using Pattern Implementation: We will start by installing Pattern using the pip install pattern. Importing required library Different functionalities are defined under different functions we will import them as and when required as we move ahead in this article. We will be working in the English language so we will be using ‘en’ for the English module. Let us start with some basic functionalities of Pattern for NLP operations NLP Operations using Pattern We will go through some of the most used and most important functionalities which are provided by Pattern. Starting with parsing a sentence. a. Parsing from pattern.en import parse parse('Hello Everyone and Welcome to Analytics India Magazine') Here we can see the output of the parse function differentiate the words in the sentence as a noun, verb, subject, or subject. We can also use the ‘pprint’ function defined in the pattern library to display the parsed sentence in a clear manner. Also, we can set different parameters for parses such as lemmata, tokenize, encoding, etc.  All these parameters can be used in parsing only so that we do not have to use a separate function for different properties. from pattern.en import pprint pprint(parse('Hello Everyone and Welcome to Analytics India Magazine', relations  = True,tokenize= True, lemmata= True)) b. N-Grams N-Gram function is used to find all the n-grams in a given text string. from pattern.en import ngrams print(ngrams(\"Hello Everyone and Welcome to Analytics India Magazine\", n=3)) Sentiment Analysis Sentiment function tries to identify the opinion or view that is held by the particular text string. Sentiment function returns both polarity and the subjectivity of the given text. The Polarity value ranges between 1(Highly Positive) to -1(Highly Negative) and subjectivity value ranges between 0(Objective) to 1(Subjective). from pattern.en import sentiment print(sentiment(\"He is a good boy but sometimes he behaves miserably\")) We can see that the sentiment analysis says that the sentence is negative with high subjectivity. Modality Modality is one such function that makes it different from other python libraries based on NLP. The modality function is used to find the degree of certainty in a particular sentence. Its value ranges from -1 to 1. As defined in the Pattern library we can state that a sentence with a modality of 0.5 and above can be stated as a fact. from pattern.en import modality text = parse('He is a good boy but sometimes he behaves miserably') text= Sentence(text) print(modality(text)) The modality comes out to be zero which means that the sentence is neutral. Suggest Suggest function is used for spelling corrections but it is more than that. It not only checks the spelling it also gives you suggestions of what might be the correct word with their probabilities. This function also distinguishes pattern from other libraries. from pattern.en import suggest print(suggest(\"Heroi\")) Quantify Quantify function is used to provide a word count estimation of the words given. from pattern.en import quantify a = quantify(['Pencil', 'Pencil', 'Eraser', 'Sharpener', 'Sharpener', 'Sharpener', 'Scale', 'Compass']) print(a) Data Mining using Pattern One of the most important features of Pattern is that it can be used for data mining through different platforms like Google, Twitter, Wikipedia, etc. Let us explore the data mining operations of the pattern library and extract some data using it. We will start by mining data using Google by entering a keyword that we want to search for and display the text along with the URL that is there in the search result. Google Mining from pattern.web import Google google = Google() for results in google.search('Analytics India Magazine'): print(results.url) print(results.text) Twitter Mining We can also use twitter for mining data which we require. Let us explore it through an example. from pattern.web import Twitter twitter = Twitter() for results in twitter.search('Analytics India Magazine'): print(results.url) print(results.text) Flickr Mining Flickr is an American image hosting and video hosting service, as well as an online community. Pattern can be used to extract data from Flickr. from pattern.web import Flickr flickr = Flickr(license=None) for result in flickr.search('Analytics India Magazine'): print(result.url) print(result.text) Similarly, Pattern provides a large number of online data mining using different platforms and we can use them accordingly. Conclusion: In this article, we started with installing Pattern, an open-source python library based on NLP and started exploring its different functions, we saw how the pattern is different from other NLP based python libraries after that we explored how we can use Pattern for text mining and extract data from online sources. Here we learned about how to use Pattern for NLP  operations and Data mining from different platforms easily and effortlessly.","excerpt":"Pattern is an open-source python library and performs different NLP tasks. It is mostly used for text processing due to various functionalities it provides.","categories":["Deep Tech"],"tags":["Data Mining","extract data from twitter","text analysis","text analytics","text classification","text dataset","text-based algorithm"],"author_name":"Himanshu Sharma","publish_date":"2020-08-25T17:00:00","publication_year":"2020","word_count":928,"keywords":["Go","text classification","TPU","AI","NLTK","Data Mining","text-based algorithm","sentiment analysis","R","NLP","extract data from twitter","Python","ViT","analytics","text analytics","text analysis","text dataset"],"extracted_tech_keywords":["AI","NLP","analytics","NLTK","sentiment analysis","TPU","Python","R","Go","ViT"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/hands-on-guide-to-pattern-a-python-tool-for-effective-text-processing-and-data-mining\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10169974,"title":"Databricks to Acquire Database Startup Neon for $1 Billion","content":"Databricks has announced its plan to acquire Neon, a leading serverless Postgres provider, for $ 1 billion. Neon’s offering is designed to support AI agents, a key element in modern development workflows. According to recent data, over 80% of databases created on Neon are provisioned automatically by AI agents, rather than humans. “The era of AI-native, agent-driven applications is reshaping what a database must do,” said Ali Ghodsi, co-founder and CEO at Databricks in a company statement. “By bringing Neon into Databricks, we’re giving developers a serverless Postgres that can keep up with agentic speed, pay-as-you-go economics and the openness of the Postgres community.” Neon’s platform allows for rapid database spin-up, with Postgres instances created in under 500 milliseconds, while offering cost scalability and full compatibility with the open-source Postgres ecosystem. Neon is a cloud-native, serverless PostgreSQL database designed for developers and AI-driven apps. Launched in 2021 by Nikita Shamgunov, Heikki Linnakangas, and Stas Kelvich, it separates storage and compute for instant provisioning, auto-scaling, and branching. This approach allows developers to quickly spin up isolated environments for rapid testing and iteration. Databricks aims to integrate Neon’s technology with its Data Intelligence Platform, addressing AI workload bottlenecks. The acquisition will help Databricks to provide developers with a serverless, scalable Postgres platform suited for AI applications. “Four years ago, we set out to build the best Postgres for the cloud that was serverless, highly scalable, and open to everyone. With this acquisition, we plan to accelerate that mission with the support and resources of an AI giant,” said Neon CEO Shamgunov. Neon’s team will join Databricks after the acquisition, continuing to support its community. This move is set to transform the $100-billion-plus database market, which is experiencing significant disruption due to AI advancements. Similarly, Databricks acquired Fennel last month, an engineering platform, to support customers in iterating on features, improving model performance with reliable signals. Last year, Databricks’ acquired Tabular, a data management startup founded by Ryan Blue, Daniel Weeks, and Jason Reid for $1 billion.","excerpt":"Neon’s team will join Databricks after the acquisition, continuing to support its community.","categories":["AI News"],"tags":["AI (Artificial Intelligence)"],"author_name":"Aditi Suresh","publish_date":"2025-05-14T17:52:57","publication_year":"2025","word_count":334,"keywords":["PostgreSQL","Go","AI","Scala","serverless","RAG","Aim","SQL","R","AI (Artificial Intelligence)","Databricks"],"extracted_tech_keywords":["AI","Aim","RAG","serverless","PostgreSQL","Databricks","R","SQL","Go","Scala"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/databricks-to-acquire-database-startup-neon-for-1-billion\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10018887,"title":"How To Become A Cybersecurity Analyst","content":"Solorigate, FireEye Breach, Nintendo Data Breach, Zoom Credentials Hack, are some of the major breaches and security attacks reported in 2020 alone. A Kaspersky report said, as many as 726 million cyberattacks were carried out in the first five months of 2020. The rise in attacks was chalked up to the shift to remote working, in the wake of the pandemic. A huge demand-supply gap on the talent front, rising incidents of cyberattacks, and stringent data protection and privacy laws worldwide make cybersecurity-related jobs much sought-after in the current market. Indian IT sector is estimated to be worth $350 billion by 2025, with cybersecurity accounting for 10 percent of the total market cap. Cybersecurity presents several career paths, including cybersecurity engineer, cybersecurity consultant, auditor, administrator, etc. An earlier report by the US Bureau of Labor Statistics said the cybersecurity analyst hiring is expected to grow by 32 percent between 2018 and 2028. Let us take a look at the nuts and bolts of becoming a cybersecurity analyst. Who Is A Cybersecurity Analyst? Cybersecurity analysts are the forefront warriors of an enterprise’s cyber defense. The role demands keeping a constant tab on any threat and monitoring the company’s network for potential vulnerabilities. A cybersecurity analyst lives by the adage– ‘a company’s security is as good as its weakest link’, and is always on the lookout for any untoward event across the network. The major responsibilities of a cybersecurity analyst include: Maintaining a firewall to protect confidential information and encrypting data transmissionMonitoring the entire network for any attacks, intrusions or unauthorised activityDetermining emerging threat patterns and vulnerabilities using advanced analytics toolsGenerating reports for all the stakeholders involved– both technical and non-technicalCarrying out risk assessments to ensure best security practices are in placeHelp in developing cybersecurity awareness training for colleaguesEducating users about threats and vulnerabilities How To Get There? A bachelor’s degree is the minimum eligibility criteria. The degree could be in cybersecurity\/information security, computer science, information systems, mathematics, physics, networks, etc. That said, practical experience and skills play a huge role in landing a cybersecurity analyst job. Skills include: knowledge of computer networks, operating systems, and security; familiarity with programming languages such as C, C++, PHP, Perl, and Java; working knowledge of security technologies such as firewalls, host intrusion prevention, and anti-virus; problem-solving; decision-making capabilities; strong verbal and written communication skills. “We are seeing increased adoption of technologies like endpoint detection and response, extended detection and response, network detection and response, security information and event management, security orchestration automation & response across all customer verticals. Security analysts need to be aware of upcoming technologies and how these fit into the IT landscape ecosystem,” said  Srinivas Prasad, Vice President and Security Practice Head at NTT Netmagic. For a stronger technical understanding and expertise, aspirants can opt for the following certification programs: CompTIA’s Network+: A vendor-neutral certification to learn designing, management, troubleshooting, and network configuration aspects.CompTIA’s Cybersecurity Analyst: This certification is ideal for IT professionals planning to pivot to cybersecurity.Certified Ethical Hacking: This certification will teach you to use the same knowledge and tools as a black-hat hacker, albeit lawfully, to improve an organisation’s security.CompTIA’s Advanced Security Practitioner: This certification is designed to provide advanced skills in enterprise security integration, research and collaboration, and risk management.CompTIA Security Analytics Expert certification: This is an advanced certification course for people with basic level certification such as CompTIA’s Cybersecurity Analyst. Common Interview Questions When we asked Srinivas Prasad on what skills are candidates generally interviewed on in companies such as NTT Netmagic, he said, “There are multiple career options in cybersecurity, if we talk about security analysts in specific, we do not focus on technology-specific skills while interviewing an analyst. Since there are new technologies in security that come up every day, it is practically not possible to have hands-on expertise with each. We look for candidates with: Good understanding of operating systems & network fundamentalsThreat detection & response capabilitiesBasic digital forensics and incident responseUnderstanding and management of SIEM.” As mentioned by Srinivas, the employers look for the general understanding and aptitude of candidates on cybersecurity issues. They can be expected to be quizzed on the following topics (subject to experience level): Risk, vulnerability and threat on a networkCybersecurity frameworksSSL encryptionDDoS attack and mitigation methodsDNS monitoring Wrapping Up Given the booming market and, in particular, the potential for the cybersecurity analyst role, it’s the place to be. Since security is a pan-industry issue, cybersecurity analysts may get the opportunity to work in varying sectors such as BFSI, government, energy, media, educational institutions, and consultancies. The starting salary for a cybersecurity analyst is 6 lakh per annum in India.","excerpt":"Solorigate, FireEye Breach, Nintendo Data Breach, Zoom Credentials Hack, are some of the major breaches and security attacks reported in 2020 alone. A Kaspersky report said, as many as 726 million cyberattacks were carried out in the first five months of 2020. The rise in attacks was chalked up to the shift to remote working, […]","categories":["AI Highlights"],"tags":["Ethical Hacking"],"author_name":"Shraddha Goled","publish_date":"2021-01-25T17:00:00","publication_year":"2021","word_count":771,"keywords":["Go","AWS","AI","Git","C++","ViT","analytics","GAN","Ethical Hacking","R","Java"],"extracted_tech_keywords":["AI","analytics","AWS","R","Go","Java","C++","Git","GAN","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/how-to-become-a-cybersecurity-analyst\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10089859,"title":"NFTs Fizzle Out, What&#8217;s Next for Polygon&#8217;s Growth Strategy?","content":"In less than a year of its launch, Meta decided to pull the plug on their NFT venture. Last week, Head of Commerce and Financial Technologies at Meta, Stephane Kasriel, announced that the company will wind down its NFT platform on Facebook and Instagram. The company claims to shut NFT to focus on “other ways to support creators, people, and businesses”. https:\/\/twitter.com\/skasriel\/status\/1635386565487898624 The focus now falls on the fate of the companies that had provided NFTs for organisations that aggressively invested in Web3 platforms when the now-subsided NFT wave peaked. One such company is Polygon. Polygon is one of the most widely recognised platforms for Ethereum Level 2 scaling and infrastructure development. Facebook and Instagram had integrated with Polygon to help creators mint NFTs. With Meta’s announcement on shutting NFTs, Polygon’s future with the big tech is now uncertain. When AIM reached out to Polygon, the company refused to comment, on record, about Meta’s decision. However, they insisted, as confirmed on tweet, that the company will continue to support and empower the creator community to “leverage the vast options within the Polygon Ecosystem”. Has Polygon’s Growth Become Stagnant? In September 2022, Dalip Tyagi, the then SVP and Head of Developer Relations, Polygon, spoke to AIM about how Polygon invests in technologies that are “future leaning” and that they invest “ahead of the curve”. Polygon also believed that they would become the “AWS of Web3”  and would get there “sooner than most of their competitors”. However, the forward vision seems to be facing turbulence in the light of one of the biggest social media platforms having retracted their NFTs. When plans of launching NFT on Facebook were abuzz last year in July, Polygon and Ethereum saw their token prices rise over 17%. The former has so far raised a total funding of $451.5M over eight rounds. In February 2022, Polygon raised $450M from three players—Sequoia Capital India, SoftBank and Tiger Global. Source: Blockchain Council (January 2023) However, after the burst of NFT hype, the company did not receive any major fundings. The last funding they received was from a non-equity assistance round in July 2022. The impact of this downturn was further highlighted last month when Polygon laid off 20% of its workforce. The layoffs were across teams, affecting about 100 employees and were said to have been a part of their efforts to consolidate multiple business units. Future of Polygon and NFTs The announcement of NFT closure by Meta comes at a time when NFT markets are facing a volatile decline. It was reported that volume of NFT sales have dropped 83% year-on-year. From $2.8 billion in January 2022, the monthly sales dropped to $492 million in January 2023. However, NFT is not on its way out and neither is Polygon. Polygon and Salesforce recently announced their partnership to create NFT-based loyalty programmes. Clients of Salesforce, a CRM software company, will be able to create token-based loyalty programmes using Polygon’s platform. They would be able to mint, manage, and sell NFTs. Collaborating with Polygon, Reddit introduced NFT avatars in July last year and is still going strong. According to a report published last month, the market capitalisation of Collectible Avatars has crossed $36 million with over 10.6 million items minted since its launch. Polygon has also announced partnership with video game maker ‘Square Enix’ to bring NFT games into its ecosystem. Additionally, Polygon has been partnering with several web3 gaming projects, including PlanetIX, Zed Run, Crazy Defense Heroes and Karmaverse. Polygon has been playing safe bets by diversifying its portfolio to cater to companies from different industries. The company’s partners belong to sectors ranging from consumer products and sports to social media, gaming, and more. This approach helps reduce Polygon’s risks and safeguards the company against any form of downfall faced by a particular industry. Other Players New entrant Amazon is working on its NFT marketplace and is expected to release it next month. Although a late entrant, Amazon will pursue its NFT ambitions without concerning itself too much with the established players it will be going against in this space. Twitter also entered the NFT market in January this year with ‘NFT profile pictures’ that are available for Twitter blue subscribers only. The Pokemon Company, maker of Nintendo video games, similarly hinted at its possible entry into the Web3 space. The company recently posted an opening for an expert in NFT, blockchain and metaverse implying its possible foray into NFT platform. However, the company is yet to make an official announcement. While the initial NFT wave might have subsided, companies’ increased partnerships with Web3 platforms might be directing attention towards a possible NFT renaissance.","excerpt":"With Meta shutting down NFT, what next for one of its biggest NFT providers—Polygon?","categories":["AI Features"],"tags":["Amazon","Blockchain","Meta","NFT","polygon","reddit","Web3"],"author_name":"Vandana Nair","publish_date":"2023-03-23T12:30:00","publication_year":"2023","word_count":775,"keywords":["Go","API","Meta","funding","Blockchain","AWS","AI","reddit","R","Amazon","RAG","polygon","Ray","Aim","GAN","NFT","Web3"],"extracted_tech_keywords":["AI","Aim","Ray","RAG","AWS","R","Go","API","GAN","funding"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/nfts-fizzle-out-whats-next-for-polygons-growth-strategy\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":69296,"title":"Netmagic To Create Centre Of Excellence In Association With Google Cloud","content":"Netmagic Solutions, a managed cloud hosting and multi-cloud hybrid IT solution provider, on Tuesday, announced its association with Google Cloud to create a centre of excellence (CoE). The idea behind CoE is to empower its customers in their digital journey by assisting organisations with hybrid clouds adoption to leverage artificial intelligence-based services at scale. “We are excited to partner with Netmagic and help businesses make a smooth transition to the cloud. The CoE will help businesses leverage Google Cloud’s leading infrastructure, platform capabilities and industry solutions to build for the future; and Netmagic, as a managed hosting and hybrid cloud IT solution provider, is uniquely positioned to help customers along this path,” said Karan Bajwa, managing director of Google Cloud India. CoE will be central to developing and delivering cloud solutions utilising Google Cloud’s Anthos — a modern application management platform — to help customers with various support across cloud adoption. Mumbai-headquartered Netmagic offers a wide range of solutions such as data centres, multi-cloud, SD-WAN, and managed security, to assist its clients in simplifying their business processes to obtain business growth. The collaboration comes at a time when almost every company are either planning or moving to clouds to maintain business continuity. To abide by government regulations, companies are required to allow their employees to work from home. This has led to the increasing adoption of the cloud, and Netmagic wants to be an enabler for the same.And with this association, the company with the help of Google would be able to further offer superior solutions. “This partnership reinforces our commitment to our customers with an integrated approach to hosted infrastructure, connectivity, security and managed support – helping them derive greater value while delivering business outcomes,” said Sharad Sanghi, MD and CEO of Netmagic (an NTT Company).","excerpt":"Netmagic Solutions, a managed cloud hosting and multi-cloud hybrid IT solution provider, on Tuesday, announced its association with Google Cloud to create a centre of excellence (CoE). The idea behind CoE is to empower its customers in their digital journey by assisting organisations with hybrid clouds adoption to leverage artificial intelligence-based services at scale. “We […]","categories":["AI News"],"tags":["centre of excellence","cloud business intelligence solutions"],"author_name":"Rohit Yadav","publish_date":"2020-07-08T19:31:00","publication_year":"2020","word_count":297,"keywords":["Go","cloud business intelligence solutions","artificial intelligence","programming_languages:R","AI","Git","RAG","ViT","cloud_platforms:Google Cloud","GAN","R","centre of excellence"],"extracted_tech_keywords":["AI","artificial intelligence","RAG","R","Go","Git","GAN","ViT","cloud_platforms:Google Cloud","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/netmagic-to-create-centre-of-excellence-in-association-with-google-cloud\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10002096,"title":"Top 5 Cybersecurity Risks &#038; Vulnerabilities In IoT","content":"IoT is gain currency and with the rapid rise of technology, comes a risk of it being hacked. Most of the information stored in these connected devices can be hacked. Homes and offices are applying these IoT devices to their network. These can include things like monitors, sensors, some of them are everyday products like your kettle. These devices are susceptible to cybercrime and are seeking opportunities to harvest valuable, personal information. Consumers and businesses face threats because of these risks involved. If the data is not properly secured, hackers can easily take advantage of it since the data is connected to the rest of the network. Reasons For Cybersecurity Risks There are a number of reasons contributing to cybersecurity risks involved with connected devices. Here are some of the most common ones. 1.Data protection: Hackers find it easy to hack into systems connected to the internet. Data in the hands of the internet, along with social media information gives away a lot of personal information. With the advent of devices like smart watches and fitness trackers gives plenty of personal data. More the details available, the easier it is for it to be hacked. Many connected devices have small sensors collecting temperature, humidity or moisture data and is generally considered to cause the greatest IoT risks, as they do not have the power, processing or memory resources required to run traditional encryption algorithms. In this case, they must have an algorithm with high security and low computation. Cryptography comes as a solution to prevent privacy risks and protect the data made available for IoT. Encryption also prevents cybercriminals from manipulating or falsifying data. 2.Passwords IoT devices come with a unique identifier that can help with the tasks of authentication. Once IoT devices on the network are discovered it is necessary to find out what they can connect to. But with hundreds or thousands of unique IDs to deal with, this can take a lot of time. As a solution, devices should only be allowed to see and access what is necessary for them to complete their jobs. Users should be diligent in updating passwords and using strong passwords. Hardcoded passwords should never be part of a device’s design process. Strong passwords can help combat IoT risks. This is where two-factor or three-factor authentication comes as help. Technologies like biometrics and blockchain can be used to authenticate IoT devices. Many IoT platforms offer features to manage your devices and control what data, other devices and networks they can access. 3.IoT Botnets A botnet is a network of systems combined together with the purpose of remotely taking control and distributing malware. With the rise of the IoT, many objects and devices are in danger of a botnet that incorporates independent connected objects. Controlled by botnet operators via Command-and-Control-Servers (C&C Server), they are used by criminals for stealing private information, exploiting online-banking data and spam emails. They are internet enabled and are able to transfer data automatically via a network and have one goal of sending thousands of email requests to a target in hopes that the platform crashes while struggling to cope with the enormous amount of requests. Social engineering is one such act of manipulating people so they give up confidential information related to bank details or other personal details. Hackers try to access a computer to install malicious software to get them access to the user’s personal information, as well as giving them control over the computer. 4.Data Encryption: Because IoT devices communicate with each other using the cloud, it creates risks. The data of a user using the IoT technology must be ensured that he is given the credentials. Blocking and disabling Universal Plug-and-Play (UPnP) capabilities and shutting down traffic on non-essential network ports frequently used for IoT attacks are also worthwhile precautions. Many manufacturers can also indulge in making misuse of the personal user data. A device may transfer data to manufacturer. Manufacturers have the authority to do maintenance of the data shared on the cloud and can be mistreated. 5.Lack of Interfaces: Some IoT devices lack human user interfaces for device use and management, something which is the core of IoT technology. Conventional IT services might be better in these terms in providing dservices than these devices with lack of interfaces. Lack Of Standardisation With greater use of data and connected devices comes a greater need to standardization of data. Today, the IoT world lacks standardization and hence it brings an even greater risk of cyber attacks. Data protection needs a rethink and the world needs a common ground for guidelines in this regard to protect the privacy of the user data. The sensors involved with IoT communicate with each other and analyze the data received. It does provide a lot of convenience. But along with that it poses risk by creating opportunities for all the shared information to be compromised for not letting it be private. Not only is a lot of data being shared among these devices, but also the data is extremely sensitive and has the potential to harm in many ways and so the risks are much greater. This is where standardisation of data becomes mandatory and comes as the first solution to the problem of cybersecurity risks involved with the advent of IoT.","excerpt":"IoT is gain currency and with the rapid rise of technology, comes a risk of it being hacked. Most of the information stored in these connected devices can be hacked. Homes and offices are applying these IoT devices to their network. These can include things like monitors, sensors, some of them are everyday products like […]","categories":["AI Trends"],"tags":["Cybersecurity","internet","IoT","passwords"],"author_name":"Disha Misal","publish_date":"2019-05-17T20:16:37","publication_year":"2019","word_count":882,"keywords":["Go","API","programming_languages:R","AI","programming_languages:Go","passwords","Cybersecurity","R","IoT","internet"],"extracted_tech_keywords":["AI","R","Go","API","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-5-cybersecurity-risks-vulnerabilities-in-iot\/","complexity_score":3,"technical_depth":6,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10096759,"title":"Anthropic Launches ChatGPT Rival, Claude 2","content":"San-Francisco-based AI lab Anthropic has announced Claude 2, a new ChatGPT rival open to the public in the US and the UK. The model is the latest version of ‘Claude’ released merely five months ago which was available only to businesses. Unlike its predecessor the latest version is available via a public-facing beta site as well as an API. One of the chatbot’s beta testers, Ethan Mollick, an Associate Professor at the Wharton School of the University said in a LinkedIn post, it has two big advantages over the other models: it is very good at handling documents (especially PDFs, which GPT struggles with) and shows a very sophisticated “understanding” of documents. Furthermore, it continues to be the most “pleasant” AI personality. On the downside, he suggested users to refrain from using the model for data, even though it accepts CSV files. It hallucinates answers, contrarily Code Interpreter does not.The startup run by former senior members of the OpenAI team Daniela and Dario Amodei purports to be a more ethically-driven company that makes generative AI safe and “steerable,” according to its website. According to the announcement blog, the latest version of the AI assistant scored 76.5 percent on the multiple choice section of the Bar exam and in the 90th percentile on the reading and writing portion of the GRE. Its coding skills have notably improved, scoring 71.2 percent on a Python coding test compared to Claude’s 56 percent. Earlier this year, in February, Anthropic introduced a waitlist for early access to Claude, following IT giant Google’s recent investment in the startup. The investment—worth $300 million—gave Google a 10% stake in the company bringing Anthropic’s value at approximately $5 billion. The partnership was predicted, as earlier in January Anthropic announced that it had chosen Google Cloud as its preferred cloud provider. The OpenAI competitor has set itself apart with its focus on understanding and developing safe AI systems, with the “constitutional AI” approach. “We have an internal red-teaming evaluation that scores our models on a large representative set of harmful prompts, using an automated test while we also regularly check the results manually,” said the blog. This is to ensure that Claude 2 is less susceptible to jailbreaks or nefarious uses. If you’re in the US or UK, you can access the AI chatbot through the Claude 2 page, and sign up for free. Click on “Talk to Claude”, provide an email address and you’ll be ready to go.","excerpt":"The model is the latest version of ‘Claude’ released merely five months ago which was available only to businesses.","categories":["AI News"],"tags":["AI Assistant","AI Chatbot","Anthropic","ChatGPT","Claude","OpenAI"],"author_name":"Tasmia Ansari","publish_date":"2023-07-12T12:15:02","publication_year":"2023","word_count":409,"keywords":["Anthropic","ChatGPT","Go","API","OpenAI","AI","AI Assistant","AI Chatbot","Claude","Python","GPT","generative AI","R"],"extracted_tech_keywords":["AI","generative AI","ChatGPT","OpenAI","Anthropic","Python","R","Go","API","GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/anthropic-launches-chatgpt-rival-claude-2\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":5867,"title":"How to Win the World Cup Office Pool","content":"With the World Cup now full flow in Brazil, worldwide anticipation has reached a fever pitch. In South America, and Brazil in particular, office pools and group bets are popping up, with soccer fans hoping to pick the winning team. At Palisade, we don’t believe in taking wild guesses—which is why we developed a model using DecisionTools Suite software to forecast the probabilities of each one of the 32 national soccer teams winning this ultimate championship. Taking data from the rankings of over 200 national teams from FIFA spanning the past four years (2011-2014), Palisade created a model that uses @RISK to determine the probabilities of different teams winning at different stages, and PrecisionTree for mapping this information into a tree, or bracket, format. @RISK uses Monte Carlo simulation to compute thousands of different possible outcomes for the tournament automatically. The historic strengths and weaknesses of each team are accounted for in the statistical models used to represent each matchup. To build this model, I first classified teams into ten equally-weighted probability “bins,” or categories, based on their past performance. I then calculated the odds of each team winning, losing, or tying with its ranked adversary in a given match. For example, a team in the highest bin (number 10) would have a larger probability of beating a team in an intermediate (say number 5) bin. In this case, historic data predicts that the highest ranked team would have an 86% chance of beating an intermediate team, a 7% chance of losing and also a 7% chance of getting a draw. The model also gives the option to consider possible home field advantage during games, using data on a team’s location during previous wins and losses. That is part of the beauty and mystery of soccer.  The chance that an underdog can, from time to time, beat the favorite! After calculating these probability tables, I modeled all of the first 48 games. For those of you who are unfamiliar, here’s a brief primer on how the tournament functions: the 32 teams are allocated into eight groups of four. Within each group, all four teams play against each other. The top two teams from each group advance to a group of 16 teams. A win at this level accounts for 3 points, a defeat for 0 points, and a draw accounts for only one point to both teams. An astounding number of results are possible. Even though teams may tie in terms of points on their quest to qualify for a second round, rules exist to break these ties. One rule states that the team with the greatest number of net goal difference (goals scored minus goals received) to that point in the tournament will proceed. If that figure is the same for both teams, then the particular match between these two tying teams is considered. Finally if a tie still persists, a coin toss determines which team advances. I incorporated all of these rules for ties into the Monte Carlo simulation model. At the core, the model also considered historic goal scoring records for winning, drawing and losing teams. Obviously, teams that win a match score more goals than teams that either draw or lose. After the 16 teams are classified, they engage in a single-elimination, bracket-style tournament (like March Madness), leaving eight, then four and then two teams for a final game. These additional 16 games (including a game to determine third place among the two losers of semifinals) were also simulated. After running 50,000 iterations, the model probabilistically channels each team into an eventual tournament win, and calculates its odds. Depending on certain ranking assumptions, the odds calculations may vary significantly. However, a robust approach that considers both historic and current rankings yielded the following results: With a home advantage incorporated, Brazil has the largest probability of becoming champions with a 17% chance, with Spain coming as a runner up with 12% probability. Aside from those two pack-leaders, the following six teams have the next greatest chances in taking the cup (in descending order): Switzerland (8%), Greece (8%), Germany (7%), Colombia (7%), Argentina (6%) and Uruguay (5%). Clearly, these probabilities do not differ significantly, which could make for an unpredictable championship if the two front-runners fail to win. Without considering home advantage, Germany would be the most likely winner, with a 19.9% chance. However, I have to concede that my trust in risk modeling may override national pride when it comes to placing bets on the World Cup. I am still not sure whether I would bet on my country, Costa Rica, in the office pool. In a group with three former World Champions–Uruguay, Italy and England–it only stands a 23% chance of classifying for second round, and only one chance in 440 of becoming World Champions.","excerpt":"With the World Cup now full flow in Brazil, worldwide anticipation has reached a fever pitch. In South America, and Brazil in particular, office pools and group bets are popping up, with soccer fans hoping to pick the winning team. At Palisade, we don’t believe in taking wild guesses—which is why we developed a model […]","categories":["IT Services"],"tags":[],"author_name":"Fernando Hernández","publish_date":"2014-06-23T18:00:24","publication_year":"2014","word_count":799,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","Rust","R","programming_languages:Rust"],"extracted_tech_keywords":["AI","R","Go","Rust","programming_languages:R","programming_languages:Go","programming_languages:Rust"],"url":"https:\/\/analyticsindiamag.com\/it-services\/how-to-win-the-world-cup-office-pool\/","complexity_score":3,"technical_depth":7,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":24347,"title":"Why MoblieFaceNet, The ‘Lightweight’ Model For Facial Recognition, Is A Total Game Changer","content":"Facial recognition market is going to be a $7.6 billion market by 2022. There is a huge opportunity waiting for whoever builds great proprietary technology using lesser computational resources. As of now Apple and Amazon seem to be winning the race to build fast and efficient facial recognition systems. However, Chinese researchers Sheng Chen, Yang Liu, Xiang Gao, and Zhen Han have now come up with a “light-weight” facial recognition network, called the MobileFaceNet. Facial verification is also a very important identity authentication technology. It is being used in more and more mobile phones and applications — such as for unlocking a device or mobile payment platform, among others. To achieve maximum user-friendliness with limited computation resources, the facial verification models deployed locally on mobile devices are expected to be not only accurate but also small and fast. Performance And Size MobileFaceNet is a neural network and obtains accuracy upto 99.28 percent on labelled faces in the wild (LFW) dataset, and a 93.05 percent accuracy on recognising faces in the AgeDB dataset. The network used around a million parameters taking only 24 milliseconds to run and produce results on a Qualcomm Snapdragon processor. We can compare this performance to accuracies of 98.70 percent and 89.27 percent for ShuffleNet, which has many more parameters and takes a little longer to execute on the CPU. The researchers have made it easy to replace the global average pooling layer in the CNN with a depthwise convolution layer, which improves performance on facial recognition. This development is really important as the artificial intelligence world searches for efficient models that run on small compute powers which are available on today’s mobile phones. Another approach for obtaining lightweight facial verification models is by compressing pretrained networks by knowledge distillation. Such approaches have achieved 97.32 percent facial verification accuracy on LFW with 4.0 MB model size. The remarkable achievement is that MobileFaceNets achieves comparable accuracy with very small budget. Traditional Weaknesses In most state-of-the-art mobile neural networks for visual recognition tasks, global average pooling layers are involved. The pooling layer is used to continuously reduce the spatial size of the representation, to reduce the number of parameters and amount of computation in the network. For example, the neural network models MobileNetV1, ShuffleNet, and MobileNetV2, which are some of the most successful facial verification and recognition approaches, have a global pooling layer. Researchers have also observed that CNNs with global average pooling layers are less accurate than those without global average pooling. However, there is no theoretical proof or analysis for this phenomenon. The researchers make a simple analysis on this phenomenon: A typical deep facial verification pipeline includes preprocessing facial images, extracting facial features by a trained deep model, and matching two faces by their features’ similarity or distance. This the global average pooling layer can be replaced by with a fully connected layer to project a compact face feature vector. But this approach adds large number of parameters to the whole model. This is not very desirable since the main pursuit is to design a model with minimum parameters. MobileFaceNet Architectures MobileFaceNet architecture is partly inspired by the MobileNetV2 architecture. The residual bottlenecks proposed in MobileNetV2 are used as our main building blocks.  The researchers use PReLU as the non-linearity, which is better suited for facial verification than using ReLU. The researchers also use a fast downsampling strategy at the beginning of the network, and a linear 1×1 convolution layer following a linear global depthwise convolution layer as the feature output layer. The detailed architecture is mentioned in the table below: The primary MobileFaceNet network uses 0.99 million parameters. To reduce computational cost, the researchers decided to change input resolution from 112×112 to 112×96 or 96×96. The the linear 1×1 convolution layer after the linear GDConv layer was also removed from MobileFaceNet. This gives a resulting network called MobileFaceNet-M. Conclusion The researchers have used  MobileNetV1, ShuffleNet, and MobileNetV2 as the baseline models. All MobileFaceNet models and baseline models are trained on CASIA-Webface dataset from scratch by ArcFace loss, for a fair performance comparison among them. The training is finished at 60K iterations. To pursue further excellent performance, MobileFaceNet, MobileFaceNet (112×96), and MobileFaceNet (96×96) are also trained on the cleaned training set of MSCeleb-1M database with 3.8 million images from 85,000 subjects. The accuracy of our primary MobileFaceNet is boosted to 99.55 percent and 96.07 percent on LFW and AgeDB-30, respectively.","excerpt":"Facial recognition market is going to be a $7.6 billion market by 2022. There is a huge opportunity waiting for whoever builds great proprietary technology using lesser computational resources. As of now Apple and Amazon seem to be winning the race to build fast and efficient facial recognition systems. However, Chinese researchers Sheng Chen, Yang […]","categories":["IT Services"],"tags":["Facial Recognition"],"author_name":"Abhijeet Katte","publish_date":"2018-05-08T10:13:41","publication_year":"2018","word_count":734,"keywords":["Go","artificial intelligence","Facial Recognition","TPU","AI","neural network","programming_languages:R","programming_languages:Go","RAG","CNN","R"],"extracted_tech_keywords":["AI","artificial intelligence","neural network","RAG","TPU","R","Go","CNN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/mobliefacenets-lightweight-model-facial-recognition\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10167331,"title":"Y Combinator’s 20-Year Journey from Incubator to an $800 Billion Powerhouse","content":"San Francisco-based Y Combinator completed its twenty-year anniversary in March this year. What began in 2005 as an experimental idea by Paul Graham, Jessica Livingston, Trevor Blackwell, and Robert Tappan Morris has since evolved into one of the most influential forces in the global startup ecosystem. YC has backed thousands of companies, generating over $800 billion in market value. Even as it scales—recently expanding to four batches a year from its traditional two—YC still insists on thinking like a startup, echoing Graham’s original vision. “It still runs on the original belief that a small group of passionate people can truly change the world,” said Kastle CEO Rishi Choudhary, who was part of YC’s summer 2024 batch. AIM spoke to YC alumni to understand if it still feels like a startup. “YC does still function very much like a startup, very nimble,” said Vahan.ai CEO Madhav Krishna, who was part of YC’s Summer 2019 batch. “Imagine, to this day, Michael Seibel, a very senior partner, reached out to me and asked for an update. It shows that top leadership is still actively involved with startups.” Notably, Seibel recently announced that he would move into a partner emeritus role. Floworks co-founder Sudipta Biswas, part of Y Combinator’s Winter 2023 batch, said the programme became quite large in 2020–21, but now feels more like it did before COVID-19. That scrappiness is a common thread, even as the programme grows. “Partners grabbing pizza and beers with founders, sitting cross-legged on crowded floors, giving tailored, thoughtful advice to each founder,” added Choudhary. YC’s Shift Towards AI Startups Under the leadership of its CEO and president, Garry Tan, YC has increasingly become a breeding ground for AI innovation. Tan has been an advocate for Little Tech, a lobbying operation that fights on behalf of thousands of venture-backed firms competing for a place in the emerging AI economy. He has been vocal about the challenges and opportunities facing startups. In a recent interview, he said that courtesy of AI, their most recent cohort is growing by 10% each week. “YC’s fingerprints are all over the AI revolution,” said Choudhary. “Look at voice AI. Nearly every leader in the space is YC-backed. YC has always excelled at spotting founders pursuing ideas that initially seem impossible but eventually reshape entire industries. AI is exactly that kind of story.” Biswas added that while OpenAI, Google, and years of earlier research in academia were key to starting the AI boom, YC played a role in accelerating it by backing a large number of startups in the space. The story comes full circle with OpenAI CEO Sam Altman. Altman was part of YC’s founding cohort in 2005 as the founder of Loopt, a location-based social app, which was acquired later. He later served as the accelerator’s president from 2014 to 2019. An important yet frequently overlooked element of OpenAI’s history is its connection to YC. OpenAI was originally incubated at YC Research before Elon Musk and Altman chose a different path. The rest is history. Golden Age to Build Tan has also spoken about the current time as a golden age to build. AI’s ability to handle demanding tasks has enabled these companies to operate with smaller teams. He said that around 25% of the current YC startups had 95% of their code written by AI. “Right now, every SaaS dollar feels up for grabs. We’re thrilled by the opportunity this creates to make consumer lending cheaper, faster, and accessible to everyone, not just incrementally better, but fundamentally different,” Choudhary added. YC released a list of startup categories they are interested in funding, aimed at addressing challenges in a post-AGI world. These include government software, public safety, US manufacturing with AI and robotics, fintech, LLM chip design, space tech, AI engineering, human-centric jobs, and energy-efficient computing. Tan’s journey into Y Combinator has been inspiring. He initially volunteered as a photographer at one of their events, and later returned as a founder in one of the cohorts. Relevance After Two Decades Although YC inspired a global wave of accelerators, its founders believe it remains just as relevant 20 years after its founding. CodeAnt AI CEO Amartya Jha, part of YC’s Winter 2024 batch, said YC remains one of the rare places where founders can start with just an idea. It offers access to top mentorship, investors, customers, and, most importantly, a strong community of peer founders. Y Combinator invests $500,000 in startups, securing a 7% stake with $125,000 and converting the remaining $375,000 into shares in future funding rounds. This gives them a strategic advantage in backing promising startups and participating in their long-term growth. As discussions around India housing 100 AI unicorns in the next decade continue, it’s interesting to note that investors betting on these startups are increasingly emerging from the far West, with Y Combinator playing a crucial role for many of them. YC experienced a remarkable 45% rise in applications from India between the Winter 2020 and the Winter 2022 batches.","excerpt":"YC wants startups to build for a post-AGI world.","categories":["AI Startups"],"tags":["Y Combinator"],"author_name":"Aditi Suresh","publish_date":"2025-04-07T09:59:59","publication_year":"2025","word_count":832,"keywords":["Go","unicorn","OpenAI","AI","innovation","Y Combinator","BERT","Aim","GAN","R","startup"],"extracted_tech_keywords":["AI","OpenAI","Aim","R","Go","BERT","GAN","innovation","startup","unicorn"],"url":"https:\/\/analyticsindiamag.com\/ai-startups\/y-combinators-20-year-journey-from-incubator-to-an-800-billion-powerhouse\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10004602,"title":"Why Google Believes You Don’t Need A College Degree To Get A High Paying Tech Job","content":"We know digital skills are crucial to finding employment today, and even more so if one strives to make a good living. In fact, since the beginning of the last decade, more than 60% of jobs need either high-level or medium-level digital skills. While COVID-19 has caused a bump in unemployment, people are keen on learning new-age skills. In today’s economy, even though higher education is appealing, for most people, it can be expensive and takes a long time (typically 3-5 years of college time) to learn the appropriate expertise. What if via learning-based online tools, people can acquire advanced skills quickly? To make way for such a scenario, Google is planning on introducing special career certificates that can support people in advancing their careers and even kickstart their tech careers to find well-paying jobs. Kent Walker, SVP of Global Affairs for Google wrote in a blog that people shouldn’t need a college diploma to have economic security. “We need new, accessible job-training solutions—from enhanced vocational programs to online education—to help America recover and rebuild,” he wrote. The new Google Career Certificates is developed on Google’s existing programs to generate pathways into IT Support careers for people without college degrees. With this, people can acquire job-ready skills to commence or advance careers in high-demand fields like data analyst, project manager, UX Designer, IT Support Specialist. According to Google, these certificates also connect learners to top national employers who are hiring for relevant job roles. He further wrote that college degrees are out of reach for many Americans, and needs new, accessible job-training solutions—from enhanced vocational programs to online education—to help America recover and rebuild. Ken Walker wrote about Yves Cooper, who previously worked as a van driver and later took the IT support certificate program. Within five days of finishing the program, he was given a role as an IT helpdesk technician at a nonprofit in Washington, D.C. Grow With Google Initiative The recent introduction of certificates come under the “The Grow with Google Career Certificates” education initiative. With this Initiative, the company will assist people in getting the abilities they require to get a job or build their business. There are hundreds of apprenticeship opportunities at Google for people finishing these career certificate programs to provide real on-the-job training. Plus, there are 100,000 need-based scholarships, funded by Google, to complete any of these career certificates. Google has also partnered with more than 100 community colleges to help with the registration and teaching assistance for the courses. “We will fund 100,000 need-based scholarships, and at Google, we will hold our new career certificates as the equivalent of a four-year degree for relevant roles. We’re also committing $10 million in job training Google.org grants for communities across America, working with partners like YWCA, NPower and JFF,” wrote Walker. The programs will be taught by the company’s employees who work in the specific field of data analytics, project management and UX experience, and IT Support. It is to be noted these career certificates are different from Google Cloud certifications, which focus on developing advanced expertise for professionals that use Google Cloud technology products. What Are The Various Certificates? Let’s take a look at which career certificates Google has developed. Google only has the course descriptions available at the moment, and the company has given a release date. The announcement is primarily targeted to American workers, but it is expected that later will be expanded worldwide. Data Analyst: With a median salary of $66,000 as per Burning Glass, the data analyst role has been included as part of career certificates. Google says that the certificate for data analysts can help them with the confidence and expertise in navigating the data lifecycle utilising tools and platforms to process, analyse and visualise to get insights from datasets. Such insights are essential for business managers to make effective decisions. Project Manager: The next in the list is the job role of a project manager, which commands the median salary of $93,000 as per Burning Glass. Project managers are high in demand today, and Google seems to meet the demand for this high paying job with a six-month career certificate. According to Google, this certificate is centred on the fundamentals of traditional project management and also provides training in agile project management. UX Designers: According to Google, this certificate guides learners with the foundations of UX design and research, creating low-fidelity designs and wireframes, designing high-fidelity prototypes, and testing. UX designers (median annual income, $75,000) help make technology products simpler and more pleasant to use by building interfaces to deliver a great user experience. IT Support Specialist: Finally, one of the most popular certificates from Google is the IT support specialist which includes learning the function of troubleshooting and problem solving for IT systems, and producing great customer service in the process. IT support specialists help maintain everything from networking to operating systems, and system administration to security across organisations. Google has been running this professional certificate for a while on Coursera, benefitting professionals with finding a new job, receiving a pay raise, or commencing a new business. Launched in 2018, the Google IT Certificate program has grown to be one of the most well-known certificates on Coursera. Google said that thousands of people have got new jobs and raised their earnings after completing the course. Why This Is Important Google has specified that the certificates are a pathway to jobs because the tech giant will directly connect Certificate completers with top employers in respective countries. The certificates are designed for learners to become job-ready for high demands and well-paid job roles with median annual pay over $50,000. Google’s certificate courses can be finished in a period of six months. The best part is that Google has hinted that they will be treated as equivalent four-year college degrees by Google. Because it is entirely developed by Google, it can be a game-changer for those looking to switch fields or advance their careers further. Not just that, Google has announced that it will provide grants scholarships to deserving learners. All the resources related to courses and specific jobs will be hosted on Coursera. Just like other career certificates, Google says there is no degree or prior experience required. In fact, 61% of learners enrolled do not hold a four-year degree. This certificate could be a perfect launchpad to a career in IT. The certificate courses are not provided for free. In terms of fees, for its current IT support program, an average student takes about three to six months to finish the certificate program which entails a cost of $49 per month. The importance of short term certificates is that they train people on practical skills in the short term instead of a traditional college education. Such skills and certifications can play a crucial role in reselling and continuous learning needed in the tech sector. Such certificates are not only beneficial in acquiring niche skills, but they are also much more affordable than traditional university degrees. Plus, the curriculum can be revised every year based on the demands of hands-on job work, something which colleges don’t do.","excerpt":"We know digital skills are crucial to finding employment today, and even more so if one strives to make a good living. In fact, since the beginning of the last decade, more than 60% of jobs need either high-level or medium-level digital skills. While COVID-19 has caused a bump in unemployment, people are keen on […]","categories":["Global Tech"],"tags":["high paying jobs in india"],"author_name":"Vishal Chawla","publish_date":"2020-08-11T15:00:00","publication_year":"2020","word_count":1189,"keywords":["Go","programming_languages:R","AI","high paying jobs in india","programming_languages:Go","Git","RAG","analytics","cloud_platforms:Google Cloud","GAN","R"],"extracted_tech_keywords":["AI","analytics","RAG","R","Go","Git","GAN","cloud_platforms:Google Cloud","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/why-google-believes-you-dont-need-a-college-degree-to-get-a-high-paying-tech-job\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":12977,"title":"Correlation Analysis can be first step to Insights but proceed with Caution","content":"The value of every information system is the opportunity for insight. Once an organization has insight all things are possible. With insight comes new opportunity – to make money, to save money, to improve goods and services, and so forth. But where does insight come from? Insight can come as easily as glancing out a window and seeing a rainbow or insight may be as difficult as investing money in a dot com startup only to ultimately discover that there never was a business case for what was being built. But one of the surest ways to spark the fire of insight is through correlation analysis. Correlation analysis is analysis of events that occur together. Suppose a study is about event A. But in gathering information about event A it is noticed that another event – event B – happens frequently. From an insight perspective, the fact that two events occur in tandem with each other is great opportunity to gather insight. Some typical questions are – Does event A cause event B? If so, under what conditions? Is there another cause of event A and event B? If so, what is the cause and under what conditions are the different events triggered? Caution must be exercised when making inferences based on coincidence of events. Given enough factors a correlation can be discovered which is nonsensical. Just because two factors have events in concert with each other does not necessarily imply a real relationship. A famous correlation once was the correlation between the winner of the Super Bowl and whether the stock market would rise or fall that year. For many years when the NFL won the Super Bowl the stock market would rise and other years when the AFL won the Super Bowl the market would fall. Of course, there is no actual relationship between professional football and the stock market. The coordination of the events was purely a random coincidence. Having stated that, there are many cases where there is a real reason for the correlation of events. When there is a real relationship between events, correlation analysis is a powerful analytic tool. There are plenty of other possibilities for insight when using correlation analysis. Correlation analysis is never more powerful and never more pregnant with possibilities then when applied to the medical and health care environment. When a medical or healthcare organization gathers all of its episodes of care and other encounters between doctors and patients and then integrates and assimilates the associated events, insightful conclusions can result. Often times very interesting and unexpected results can be the result. Looking at the natural correlations that have evolved over the years in the practice of medicine can be very useful. What happens is that often times no major patterns are discerned by any one doctor (short of alerting his\/her intuition) because the doctor sees only one patient at a time, that is the patient that is immediately in front of him\/her. But given many observations taken over many doctors and patients and taken over a lengthy period of time, medical and health patterns start to emerge that have otherwise been unnoticed. And sometimes these patterns have very profound implications to healthcare. One of the challenges of doing correlation analysis that is meaningful is that many events must be reported. It doesn’t do much good to examine 50 or even 500 incidences of medical care. Instead 500,000 or 5,000,000 incidences of care are much better for spotting previously unseen correlations and patterns. Another of the challenges is that of dealing with the data. Medical and healthcare data is notoriously unstructured. Medical and healthcare data is usually textual data. In addition, there is a very variable nomenclature for the same event or activity. Recently a knowledgeable physician told me that there were at least 15 ways to describe the same thing – a broken bone. Yet another problem with meaningful correlation analysis is that in some cases there are terms that are spelled differently. To do meaningful correlation analysis, there needs to be a single consistent spelling of the same term. Still another problem is that for correlation analysis to be effective, the results need to be visual. When correlation analysis is not visual, patterns that are unclear or faint tend to hide – to get lost in the massive amount of data and other more obvious patterns. But when visual techniques are used for correlation analysis, the chances of spotting faint and unclear patterns become greatly enhanced. The good news is that correlation analysis for the medical community is now a real possibility. Leading medical research firms are starting to use powerful new technology for very sophisticated correlation analysis. It is a reasonable expectation that today’s technology can spot the important patterns that were as little as a year ago unable to be found. Now you can hear and see Bill Inmon on the Internet. Take a look at his new videotape education series on safaribooksonline.com Bill covers IT topics from A to Z.","excerpt":"The value of every information system is the opportunity for insight. Once an organization has insight all things are possible. With insight comes new opportunity – to make money, to save money, to improve goods and services, and so forth. But where does insight come from? Insight can come as easily as glancing out a […]","categories":[],"tags":[],"author_name":"William Inmon","publish_date":"2017-02-23T05:27:02","publication_year":"2017","word_count":835,"keywords":["Go","programming_languages:R","AI","data_tools:Spark","programming_languages:Go","ViT","GAN","R","startup"],"extracted_tech_keywords":["AI","R","Go","GAN","ViT","startup","programming_languages:R","programming_languages:Go","data_tools:Spark"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/correlation-analysis-can-first-step-insights-proceed-caution\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":32447,"title":"When Obtaining LegalDocs Becomes Easy With LegalDocs AI-Powered Solution","content":"LegalDocs is a Mumbai-based AI-start-up that was established in 2016 for making CA, CS and lawyer services as easy as ordering from an e-commerce website Gauravkumar Kate, CEO of LegalDocs established the start-up after his tryst with what he calls the mundane process of trying to get legal documents starting from his rental agreement to the marriage certificate. “After getting married I had to go through mundane documentation like rent agreement registration, marriage registration, name change etc. It was a nightmare for me because there was a lack of information about how to get these documents and from where to get it,” Gaurav says about his struggle which prompted him to think differently. Gaurav, who previously worked as an Innovation and Product Development Manager immediately saw a business opportunity there and went on to establish the start-up along with brother Makarand Kate with their friend Shrishail Thote. Soon after its commencement, LegalDocs started working with SMEs who were facing problems related to accounting and legal matters. “We understood that SMEs are facing more problems in these areas and so we started helping SMEs through our professional marketplace,” Gaurav says. Through LegalDocs, the three primary steps of legal documentation such as drafting, paying stamp duty and document validation can more or less be outsourced to Gaurav’s team. All the customer needs to do is to go online and opt for the service that they need. AI At The Core According to Gaurav, some of the biggest challenges that individuals and business face is to find legitimate CA, CS and law professionals. More often, there are fraudsters who loot the customers and do not get the work done on time as well. So taking these factors into consideration, the inbuilt AI algorithms within LegalDocs will enable the customers to avail verified legal documents at a predetermined rate, thus solving the problem of discovery and right pricing.  AI is also leveraged to draft regular legal documents like employee agreements, vendor agreements, non-disclosure agreements and cheque bounce, thus democratising the processes along the way. “At the core of LegalDocs platform is a proprietary AI algorithm for matchmaking and tracking of customer satisfaction. At a point of time, if a customer is not happy or have any issues with the professional, she or he can request for change of professionals. This solves the problem of communication and progress tracking,” says Gaurav. To make their platform more robust, the team has studied over 5000 drafts to train the AI with each draft. Though the system is currently at a testing phase, Gaurav’s team hopes to make it available for the public from January 2019. An AI-Enabled Backbone The application of AI cut-across LegalDoc’s platform right from understanding customer preferences to procuring customer feedback. Here are the four main areas where AI has been leveraged. – LegalDocs’s major product development revolves around simplicity and process automation. -AI help them understand customer requirements and product offerings are customised based on the algorithm. -AI also help in matchmaking. Based on customer requirement suitable professional is chosen for them. In order to do this, the AI matrix goes through more than 50+ parameters to get the best possible match. -Feedback and performance improvement metrics for professionals is driven by AI algorithms. The Road Ahead Within just two years of its inception, LegalDocs has already carved a niche in the legal tech domain. Presently their client base includes Swiggy, Zomato, Nestaway, No broker, Lodha Group, Rustomgee for various CA, CS and legal requirements across seven metro cities in India. Under the watchful eye of Ex Infosys CEOs S D Shibulal and S Gopalakrishnan, the start-up’s revenue has grown substantially and presently serve more than 50,000 customers. In the future, the start-up plans to turn itself into a business consultancy and capture more than 10% market share and emerge as a market leader in CA, CS and legal services.","excerpt":"LegalDocs is a Mumbai-based AI-start-up that was established in 2016 for making CA, CS and lawyer services as easy as ordering from an e-commerce website Gauravkumar Kate, CEO of LegalDocs established the start-up after his tryst with what he calls the mundane process of trying to get legal documents starting from his rental agreement to […]","categories":["Deep Tech"],"tags":["AI in legal sector in India","Startups"],"author_name":"Akshaya Asokan","publish_date":"2018-12-31T06:59:42","publication_year":"2018","word_count":647,"keywords":["Go","programming_languages:R","AI","innovation","programming_languages:Go","AI in legal sector in India","Git","RAG","automation","Rust","Startups","R"],"extracted_tech_keywords":["AI","RAG","R","Go","Rust","Git","automation","innovation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/when-obtaining-legal-docs-becomes-easy-with-legaldocs-ai-powered-solution\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10092959,"title":"After Google, Microsoft Targets Nvidia","content":"In response to a question about Microsoft’s plans to compete with Google’s search business, CEO Satya Nadella indicated that he only intends to make Google ‘dance’. Looks like Microsoft sees itself in AMD, and the duo is set out to break another monopoly: that of Nvidia. Just a few days ago, reports emerged that Microsoft has been working on its own in-house AI processors, codenamed Athena, since 2019 in a bid to challenge Google and AWS, who have their own set of in-house chips for training and inference. To top that, new reports allege that this project, which Microsoft has been keeping under the shadows for so long, is actually being done in partnership with the chip company, AMD. A part of Athena or not, speculations are ripe that Microsoft is financing AMD’s AI chip push. It is surprising, especially since Microsoft has a strong partnership with Nvidia, which helps to train OpenAI’s large language models on Azure. It almost seems like AMD and Microsoft have planned a coup against Nvidia. No wonder Nvidia’s shares moved lower following the report. It will not be the first time we’ve seen Microsoft use AMD. AMD’s silicon already powers the secure AI infrastructure in Azure cloud services, while also playing a role in the Xbox Series X and Series S consoles. Market positioning Currently, Nvidia enjoys a monopoly in the GPU market, so the industry is seeking an alternative. At least in the data centre segment, it feels like Intel and AMD are fighting over scraps, while Nvidia continues to dominate sales. It’s possible that Microsoft doesn’t want to spend so much money into Nvidia and is looking for an alternative that can handle their workload. “They [Microsoft] were counting on Intel, but Intel is still not able to deliver. AMD, on the other hand, has got a very good GPU technology, but they fell back in software optimisation,” says semiconductor analyst Sravan Kundojjala. AMD will be releasing the Instinct MI300 later this year which will be their first shot at building a true data centre\/HPC-class APU, combining the best of AMD’s CPU and GPU technologies. To catch up to Nvidia, AMD developed a software framework called ROCm™, an open software platform designed to provide HPC and AI communities with access to open compute languages, compilers, libraries, and tools. However, it is not yet mature enough. Kundojjala explains that Nvidia’s CUDA works with all kinds of industries, but AMD doesn’t have that luxury. In the data center, AMD is trying to focus its attention on certain verticals – certain customers, certain workloads – so it can position itself. One of those customers is Microsoft. Nvidia alone cannot serve every use case or every customer. Kundojjala says, “it’s a rising tide, lifts all boats kind of situation.” So, if AMD wins, Nvidia doesn’t have to lose. Because the market is so big, both can grow. At the same time, he also argues that Microsoft’s AMD push will not have much to do with the pricing that Nvidia offers for its GPUs. “Unlike consumer applications, data centre customers are not sensitive to price. Enterprise cloud market doesn’t care about pricing. What they need is that the product has to be good, the roadmap has to be good, and the maturity of the software framework that they provide along with the hardware product – that has to be good,” he added. Generative AI gold rush In the recent earnings call, AMD CEO Lisa Su suggested that there are opportunities beyond hyperscalers and game consoles for AMD’s IP, such as semi-custom opportunities with higher volume potential. AMD has already had success with semi-custom gaming consoles such as the Sony PS and Microsoft Xbox, and it is attempting to replicate this semi-custom business model in data centers with hyperscalers like Microsoft. “In the AI space, I am pretty sure that Microsoft can help AMD accelerate its software,” said Kundojjala. Kundojjala mentions that AMD’s data centre GPU revenue is a few hundred million dollars, which is negligible compared to Nvidia’s data centre GPU chip revenue of almost $15 to $16 billion. The reason AMD has come so late to the party is that it never really focused on it. ‘Their plate was already full; they were already eating Intel’s lunch,’ he said. AMD is still reporting solid profits, despite a PC inventory correction and a slowdown in data center sales, while Intel is reporting losses and investing billions in new fabs. Additionally, AMD has recently revealed its new Ryzen 7040U series processors for laptops, making bold claims that the chips not only beat the competition from Intel but also outpace the MacBook M2. Overall, they have been making significant strides in that direction. Now, AMD realises that generative AI is a crucial use case in data centres. However, there will not be any dramatic change since their MI300 product will ramp up until 2024. So, it will take at least another three years for AMD to make its mark in data center GPUs.","excerpt":"Looks like Microsoft sees AMD in itself, and the duo is set out to break another monopoly: that of Nvidia.","categories":["Global Tech"],"tags":["AMD","ChatGPT","Google Search","Microsoft","NVIDIA","OpenAI"],"author_name":"Ayush Jain","publish_date":"2023-05-08T11:14:41","publication_year":"2023","word_count":834,"keywords":["CUDA","Go","ChatGPT","AMD","OpenAI","AI","Google Search","AWS","R","Aim","generative AI","NVIDIA","Azure","Microsoft"],"extracted_tech_keywords":["AI","generative AI","OpenAI","Aim","AWS","Azure","CUDA","R","Go","CUDA"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/after-google-microsoft-targets-nvidia\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10165316,"title":"‘Anthropic’s Claude Code Has Been Writing Half of My Code…’","content":"Anthropic’s obsession with developers took a major leap last week with the announcement of Claude Code. While it impressed several users with its ability to run and preview code using the Artifacts feature, Claude Code goes one step further as an ‘agentic coding tool’ that operates directly within the terminal. It is capable of fixing bugs across a code base, resolving merge conflicts, creating commits and pull requests, and answering questions about the architecture and logic. Moreover, as revealed by the company’s chief product officer Mike Krieger, Anthropic’s approach is largely about “picking its bets” carefully. With Claude Code, they took a strategic step by first releasing it internally to boost its own team’s performance. “After seeing it play out for a couple of months, we thought, ‘This is good.’ It’s not a solution for all coding problems, and doesn’t obviate the IDE (integrated development environment). But it is useful to us in enough cases that we want to see people use it out,” said Krieger, in a podcast episode with venture capitalist Harry Stebbings. “Our product engineers love Claude Code,” he added, indicating that most of the work for these engineers lies across multiple layers of the product. Notably, it is in such scenarios that an agentic workflow is helpful. Meanwhile, Emmanuel Ameisen, a research engineer at Anthropic, said, “Claude Code has been writing half of my code for the past few months.” Similarly, several developers have praised the new tool. Victor Taelin, founder of Higher Order Company, revealed how he used Claude Code to optimise HVM3 (the company’s high-performance functional runtime for parallel computing), and achieved a speed boost of 51% on a single core of the Apple M4 processor. He also revealed that Claude Code created a CUDA version for the same. “This is serious,” said Taelin. “I just asked Claude Code to optimise the repo, and it did.” Several other developers also shared their experience yielding impressive results in single shot prompting. Pietro Schirano, founder of EverArt, highlighted how Claude Code created an entire ‘glass-like’ user interface design system in a single shot, with all the necessary components. Notably, Claude Code also appears to be exceptionally fast. Developers have reported accomplishing their tasks with it in about the same amount of time it takes to do small household chores, like making coffee or unstacking the dishwasher. However, if one is looking at the intersection between AI and coding, Cursor has to be taken into consideration. The AI coding agent recently reached $100 million in annual recurring revenue, and a growth rate of over 9,000% in 2024 meant that it became the fastest growing SaaS of all time. A user on Reddit compared both Cursor and Claude Code. The review stated that Claude Code produces code of “very high quality”. “This thing blows Cursor out of the water. I can’t believe both use the same model when I see the difference in how Claude-3.7 behaves in Cursor and how it behaves in Claude Code,” the review added. “I’ve had no functionality breaking mistakes, which happen every now and then with Cursor, where it just breaks something or large files are truncated,” it further stated. It is ‘Insanely Expensive’ Anthropic has always found a sweet spot in the hearts of developers for most, if not all, of their products – from Computer Use to MCP, and now the Claude Code. However, there has also been a fair share of criticism. For one, Claude Code is very expensive as the API pricing for Anthropic’s AI models is one of the highest out there. The Claude 3.7 Sonnet costs $3 per million input tokens and a whopping $15 per million output tokens. One developer called it “insanely expensive” and said they could easily spend $50-$100 on it, while agreeing that it is better than Cursor. Matt Popovich, an engineer at Forge, said that Claude Code costs as much as hiring a developer. However, autonomous coding agents are expensive, considering Devin, the first to the market, costs a whopping $500 a month, but with no limit on the number of seats in an organisation. Another developer said that Claude Code costs him $28 a day, adding that it will end up costing the same as Devin. asking claude code to change the text on a label pic.twitter.com\/chY4GKKP67— Tetrabay 3 💎🌁 (@TetraspaceWest) March 4, 2025 Besides, there are other problems associated with Claude Code. Multiple users on GitHub pointed out that running a command to automate Claude updates on the Ubuntu Server 24.02 messed with the system file ownership, locking users out of admin access. However, Anthropic did provide a solution to mitigate the issue. Moreover, a few developers also reminded that AI coding agents still have a long way to go, and plenty of room to elevate their scope. Petr Baudis, CTO of Rossum, said on X that Claude Code struggled with certain real-world engineering tasks. He found that the tool wrote several redundant and unreviewable code, and it cost $55 to do so. “To be clear, looking at this from 2023, it’s absolutely mindblowing that AI can do all this. But it’s also simply not useful for actual engineering tasks where plain code-writing isn’t the bottleneck. Not even tests,” he added. It would be unfair to single out Claude Code, however, as the situation is more or less similar with multiple autonomous coding platforms.","excerpt":"Anthropic relied on Claude Code internally to accelerate development.","categories":["Global Tech"],"tags":["AI (Artificial Intelligence)","Anthropic","Anthropic Claude"],"author_name":"Supreeth Koundinya","publish_date":"2025-03-05T19:00:00","publication_year":"2025","word_count":894,"keywords":["CUDA","Anthropic","Go","API","TPU","AI","Git","Anthropic Claude","GitHub","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","Anthropic","TPU","CUDA","R","Go","CUDA","Git","GitHub","API"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/anthropics-claude-code-has-been-writing-half-of-my-code\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10020565,"title":"How To Run A Development Server For Flask Web Applications Using Google Colab","content":"A lot of people know how to build ML models, but surprisingly few are comfortable with the deployment process. Deploying models is a necessary skill in the industry, and the first step of deployment is running our ML models on the web during development for demos and testing. This can be done using a simple development server before deployment. A development server can be built with just a few lines of code using a framework like Flask. In this article, we will explore a very simple development server which will enable us to render the output predictions of our image classifier model (written using PyTorch) on a public webpage. This entire code can be run on Google Colab, without installing anything on our local machine. Google Colab provides a virtual machine environment, so unlike when running Flask on our local machine, we cannot access localhost. Hence, we need to expose it to a public URL using the library flask-ngrok, which generates a temporary URL where the web app runs. First, let’s create two folders called images and references in the current project directory, where we will store images (to be classified by our classifier) and a dictionary mapping class indices to human-understandable class names. import os PROJECT_ROOT_DIR = \".\" IMAGES_PATH = os.path.join(PROJECT_ROOT_DIR, 'images') os.makedirs(IMAGES_PATH, exist_ok=True) REFERENCE_FILES_PATH = os.path.join(PROJECT_ROOT_DIR, 'references') os.makedirs(REFERENCE_FILES_PATH, exist_ok=True) Next, let’s install ngrok, which will help us with the temporary url generation. !pip install flask-ngrok After installing Flask, let’s create a web app, called app. When we call run_with_ngrok(), ngrok will start when app is run, creating the temporary URL. app = Flask(__name__) run_with_ngrok(app) Now, let’s upload the {class index: [class ID, class name]} dictionary into the references folder we created. The link for the document is available here: imagenet_class_index = json.load(open(os.path.join(REFERENCE_FILES_PATH, 'imagenet_class_index.json'))) Next, let’s use PyTorch’s torchvision package to create a classifier from one of its pre-trained models (densenet121). Then, we define a pipeline that takes in an image in bytes and resizes it, centre crops it, converts it to a tensor, and then normalizes the result, to be fed into the model. Note that we should load the model only once, before serving requests,  so that it is not unnecessarily loaded multiple times which would be a waste of computational power. This is very important in production systems. model = models.densenet121(pretrained=True) model.eval() def transform_image(image_bytes): my_transforms = transforms.Compose([transforms.Resize(255), transforms.CenterCrop(224), transforms.ToTensor(), transforms.Normalize( [0.485, 0.456, 0.406], [0.229, 0.224, 0.225])]) image = Image.open(io.BytesIO(image_bytes)) return my_transforms(image).unsqueeze(0) We can test transform_image by running it on a test image; the output should be a tensor. A tensor is like a matrix of N dimensions: all neural networks expect inputs to be tensors. with open(os.path.join(IMAGES_PATH, 'test_pic.jpg'), 'rb') as f: image_in_bytes = f.read() tensor = transform_image(image_bytes=image_in_bytes) print(tensor) Next, let’s define the get_prediction method which takes in the image in bytes, transforms it to a tensor using the pipeline defined in the previous function, and then performs forward propagation using the model and returns the predicted class name. The model outputs the integer class index (0,1,2…) which has to be converted to the string key to get the corresponding string class index and name from imagenet_class_index. def get_prediction(image_bytes): tensor = transform_image(image_bytes=image_bytes) outputs = model.forward(tensor) _, y_hat = outputs.max(1) predicted_idx = str(y_hat.item()) return imagenet_class_index[predicted_idx] We can test the above function by calling it with our test image. The image should return a class index and name as shown below. with open(os.path.join(IMAGES_PATH, 'test_pic.jpg'), 'rb') as f: image_in_bytes = f.read() print(get_prediction(image_bytes=image_in_bytes)) The test_pic used was a girl in a skirt, so the model is quite right! Next, let’s define web app functions. We’ll define the home\/root web page at index (‘\/’) and another page (‘\/predict’) where the JSON output should be rendered. @app.route(\"\/\") def root(): return 'Root of Flask WebApp!' @app.route('\/predict') def predict(): file_doc = open(os.path.join(IMAGES_PATH, 'wolf.jpg'),'rb') img_bytes = file_doc.read() class_id, class_name = get_prediction(image_bytes=img_bytes) return jsonify({'class_id': class_id, 'class_name': class_name}) Hence, the home page should say ‘Root of Flask WebApp!’ and the \/predict page should give us our model’s output for the input image, in this case, a wolf. Let’s run the web app with the following command: app.run() This means ngrok is exposing our web app onto the temporary URL http:\/\/be8dd2d747fc.ngrok.io (it is different every time it is run). If we go to this URL we get our home page: And if we go to the ‘\/predict’ domain, we get our model’s prediction. The model is right, and its JSON output is successfully being rendered on a public webpage! That’s it, we have now successfully built a development server and deployed our model on the web. This is a great way to give a demo of our web app before deploying it into production. Nowadays more and more people are using Google Colab because of its low overhead and easy access to GPUs, and now you too know how to create and test ML web apps on Google Colab from scratch! For the complete running code, please refer to this link.","excerpt":"A lot of people know how to build ML models, but surprisingly few are comfortable with the deployment process. Deploying models is a necessary skill in the industry, and the first step of deployment is running our ML models on the web during development for demos and testing. This can be done using a simple […]","categories":["Deep Tech"],"tags":["Flask","Google Colab"],"author_name":"Divya Sundaresan","publish_date":"2021-02-20T13:00:00","publication_year":"2021","word_count":822,"keywords":["Go","Flask","TPU","programming_languages:R","AI","neural network","PyTorch","ML","Colab","ai_frameworks:PyTorch","Google Colab","R"],"extracted_tech_keywords":["AI","ML","neural network","PyTorch","Colab","TPU","R","Go","ai_frameworks:PyTorch","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/how-to-run-a-development-server-for-flask-web-applications-using-google-colab\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10171012,"title":"We Smoked NVIDIA’s Blackwell, Says Cerebras","content":"AI hardware maker Cerebras announced on Wednesday that its systems have outperformed NVIDIA’s DGX B200 with 8 Blackwell GPUs’ [Graphic Processing Units] output token speed on Meta’s Llama 4 Maverick Model. Cerebras achieved an output token speed of over 2,500 tokens\/sec, whereas NVIDIA demonstrated only 1,000 tokens per second. However, NVIDIA outperformed systems from Groq, AMD, Google, and other vendors. “Only Cerebras stands – and we smoked Blackwell,” said Cerebras in a post on X. Cerebras just beat NVIDIA BlackwellLast week: Blackwell hit 1,000 t\/s on Llama 4.Today: Cerebras hit 2,500 t\/s on the same model, same benchmarks by @ArtificialAnlysBlackwell smoked Groq, AMD, Google – everyone.Only Cerebras stands – and we smoked Blackwell. pic.twitter.com\/2Nd0W8ttOB— Cerebras (@CerebrasSystems) May 28, 2025 Based in the United States, Cerebras manufactures hardware specifically designed for AI inference, using a trained AI model to make decisions. The company’s Wafer-Scale Engine (WSE) technology offers faster inference\/output tokens speed than traditional GPUs. “We’ve tested dozens of vendors, and Cerebras is the only inference solution that outperforms Blackwell for Meta’s flagship model,” said the company. Last month, Meta announced a partnership with Cerebras to offer developers access to inference speeds up to 18 times faster than GPU-based solutions. While GPUs are widely used for training AI models, which require a vast amount of data and compute, dedicated solutions for inferencing are being developed at a large scale. Cerebras, Groq, and SambaNova are some of the other companies working on such solutions. SambaNova’s SN40L custom AI chip, which features their Reconfigurable Dataflow Unit architecture. Manufactured on TSMC’s 5 nm process, the SN40L combines DRAM, HBM3, and SRAM on each chip. On the other hand, Groq offers AI inference processors called LPUs, or language processing units. Instead of relying on external memory like GPUs, LPUs keep all the model parameters directly within their chips.","excerpt":"At over 2,500 t\/s, Cerebras claims to have has set a world record for LLM inference speed on the 400B parameter Llama 4 Maverick model.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","Cerebras"],"author_name":"Supreeth Koundinya","publish_date":"2025-05-30T17:48:47","publication_year":"2025","word_count":303,"keywords":["Go","TPU","Cerebras","SambaNova","AI","programming_languages:R","Llama 4","Groq","llm_models:Llama","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","Llama 4","TPU","R","Go","Groq","Cerebras","SambaNova","llm_models:Llama","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/we-smoked-nvidias-blackwell-says-cerebras\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":50211,"title":"Marketing Futurist: A Superhero Who Juggles Data Analysis, Behavioral Sciences &#038; Creativity","content":"An advertising creative head, a behavioural scientist and a data analyst walk into a bar… Alas! Jokes of this kind might no longer be relevant or spoken about by today’s standup comedians. Why? Because all three of the above ‘roles’ have now merged into a single one. The Marketing Futurist, a superhero you didn’t know about. Yeah. I know what you are thinking. Data Analyst? Behaviour science? and a Creative Head? All rolled into one brain? Gods come alive, I would love to see the neural activity map of that guy. Because technically, both the right and the left side of his brain will be on fire. Maybe, these Marketing Futurists have a shorter than average lifespan because of neural synapses shorting faster than usual? [That deserves a story for later!] But for now – we have come to accept the reality that data is the new fuel that is firing up enterprises everywhere in this information economy value. And the advertising industry is not a stranger to this deluge that has taken over. In fact, the story of big data success is tied very closely with the evolution of the ad industry itself. We have obviously come a long way from the heydays of sales agents haggling for real-estate on the large internet publishers in-person to now the magic of RTB (Real-Time Bidding using artificial intelligence) and programmatic buying of digital media space that happens faster than the time taken for even webpage to load up. “40% of the success [of a direct campaign] will be attributable to the targeting. Maybe 20% to the creative. (These are merely averages—a really clever creative insight can have a huge effect.) And there was always this 30% which was ‘something else,’ it wasn’t the targeting and it wasn’t creative.” Rory Sutherland, Vice-Chairman Ogilvy UK. Rory surmised very cleverly that this ‘something else’ is actually behavioural science which has gained prominence in this Kahneman-Ariely era. And yet, it is still mind-boggling to assume that the creative process for ads (usually behind-locked-doors, furious marathon brainstorming sessions fueled by caffeine, cigarettes and confusion) can actually be done in a lot more data-driven and ‘logical’ fashion. Creativity by Analytics is possibly an oxymoron but we now know that it only applies for the ad-men of the last era. Because today, no process is complete without being guided by data. Today’s CMO owns the overall customer experience and the brand resonance but it’s a tough job that he straddles. Because of liquid customer demands in an era where loyalty is probably dead and digitally native start-ups continuously disrupt and challenge the status-quo, marketing (and consequently advertising!) has to evolve from being static product-centric to being “relevant at the moment”. Data and behavioural science is what would feed this new revolution. Maybe we can call this Prescient Advertising for now, eh. Let’s take a few examples of how data and creativity have worked well together. Personalization: One of the recent examples for this is Spotify’s entry into the Indian market. Spotify’s tag line as they launched in India, was ‘There is a playlist for that’ focused on different situations. Now, this was a direct fallout from the millions they initially invested in their data science teams and continuously improving on their personalization algorithms which sync real-time, changing music-streams based on the ‘like’ buttons. Apparently, this whole campaign was a result of three separate activities. Social listening based on geography, sentiment analysis and finally recommendations engine. Ensuring Influencer marketing done right: Today’s consumers are hard nuts to crack and with Instagram and other social platforms gaining importance in the overall purchasing funnel, influencer marketing possibly could a critical tactic to drives consideration and purchase. So, ad-makers targeting Instagram with such influencers endorsing their brands are laughing all the way to the banks. But there is enough and more instances where this money being thrown into a black hole without much results. Can we predict the impact of such influencers on brand resonance? Data could help! Chtrbox is a startup in this space that helps enterprises quantify more than 7-8 parameters (relevance, reach, authenticity, mix optimization etc) to measure and predict influencer marketing effectiveness. Identity, Insights and Intelligence using DMP: Data Management Platforms (DMP) have become a critical component of all marketing technology stack for advertisers\/marketers. Starting from identifying (unique marker for individual customers across devices\/geography or channel) to generating insights from analytics\/machine learning till ensuring intelligent orchestration (using the right channel, right moment with the right product for the customer) There are variations of DMP today in action (Like CDP’s Customer Data Platforms focused on creating 360-view of the Customer Segments\/cohorts in real-time or near real-time) but DMP’s remain an excellent case in point of data influencing advertising decisions. It’s clear that in today’s age [Accenture terms this to be a post-digital era] the data deluge needs to be managed (“captured, curated and consumed”) converting it to gold nuggets to lead to high growth. Advertising as a function can utilize data platforms to run informed highly relevant campaigns, expanding the overall creative canvas and also making it a lot more competitive. Stands to reason that apart from savings on all the caffeine and cigarettes, bringing data into the brainstorming sessions will take out the big C (Confusion) from the whole process because of a high degree of objectivity and transparency. After all, machines cannot lie. And thus, are best suited to deal with the scale and pace of the big data that defines this industry. And Lo! Behold. A new superhero is born – The Marketing Futurist who successfully juggles the roles between a creative head, a database marketer and even brings in his behavioural science knowledge to successful target and convert prospects into buyers\/customers. Post Script: I know that tiny worm at the back of your mind still cannot fathom this: How can the machines running on rational numerical and spreadsheet + Python driven logic beat the frenetic, creative, and emotionally driven psychological ideas?","excerpt":"An advertising creative head, a behavioural scientist and a data analyst walk into a bar… Alas!  Jokes of this kind might no longer be relevant or spoken about by today’s standup comedians. Why? Because all three of the above ‘roles’ have now merged into a single one.  The Marketing Futurist, a superhero you didn’t know […]","categories":["AI Features"],"tags":["Advertising","digital marketing","Sentiment Analysis"],"author_name":"Sachin Dev","publish_date":"2019-11-19T12:01:00","publication_year":"2019","word_count":996,"keywords":["data science","Go","Sentiment Analysis","artificial intelligence","machine learning","AI","sentiment analysis","Advertising","RAG","Python","digital marketing","analytics","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","data science","analytics","RAG","sentiment analysis","Python","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/marketing-futurist-a-superhero-who-juggles-data-analysis-behavioral-sciences-creativity\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10125044,"title":"This Surat-based Company is Creating AI Hardware for India","content":"Starting a hardware company based in India is no easy task. However, Surat-based company, Vicharak, took on the herculean task of churning out hardware in-house, designed specifically for AI workloads. The company recently secured funding of INR 1 crore,  boosting its valuation to INR 100 crore. We’ve received ₹1 crore in funding at a ₹100 crore valuation. We’ve often heard that doing hardware is hard in a city like Surat or even in India. We’ve gone through hundreds of failed prototypes and iterations in our labs, but somebody has to start at some point, right?This…— Vicharak (@Vicharak_In) June 25, 2024 Speaking with AIM, founder and CEO Akshar Vastarpara said that Vicharak’s focus is not just on creating hardware, it’s redefining computing technology. “Our first target is to develop a GPU-like technology that can be used in mobile phones, laptops, and servers. We are approaching this in a very different way, starting with the consumer base but scaling to servers and lower-level areas as well,” Vastarpara explained. This led to the creation of Vaaman. It is a complete packaged computing board that boasts a six-core ARM CPU and an FPGA with 112,128 logic cells. Its distinctive design allows it to tackle challenges that existing products can’t. With a 300-MBps connection between the FPGA and CPU, Vaaman is optimised for hardware acceleration and excels in parallel computing. India for the World “Our goal is not to compete directly [with NVIDIA] but to offer something unique. FPGAs are reconfigurable chips, capable of doing many things that ASIC (Application-Specific Integrated Circuit) startups can’t. We can achieve 90% efficiency compared to what they offer,” Vastarpara said confidently. This is such an insane white pill moment for me. Back in college I used to such reviews by western YouTubers of western tech companies. Now time watching an Indian YouTuber make review about a deep tech product of an Indian company. We’re gonna make it folks. pic.twitter.com\/RwkOcxMzmK— Varsh (@infinite_varsh) June 26, 2024 Vicharak’s products are poised to revolutionise single-board computing. “We are in the same industry as Raspberry Pi, but our boards include FPGAs alongside processors, offering a complete AI infrastructure,” he elaborated. Priced at $180, their boards offer a competitive edge with advanced capabilities at an affordable cost. Moreover, Vicharak aims to make its software completely free while maintaining proprietary IP on their hardware. “Our plan is to integrate FPGAs into every kind of computer, much like CPUs and GPUs today. The software we develop will be free, but the IP for our FPGA designs will remain ours,” Vastarpara clarified. Not just hardware, Vicharak is also in direct competition to NVIDIA’s CUDA with its focus on software. Its flagship product, Gati, exemplifies this vision. “Gati is our AI exploration project. We’re writing our own infrastructure on top of FPGA, creating a stack similar to what NVIDIA does with CUDA,” said Vastarpara. The goal is to enable AI inference on FPGAs, offering a flexible and powerful alternative to traditional GPUs and CPUs. Apart from Vaaman and Gati, Vicharak has also built Axon, a processor powered by Rockchip RK3588S, an 8-core 64-bit SoC, utilising a 8 nm lithography process. It also integrates a 4-core GPU and a built-in NPU which provides up to 6 TOPS of performance for AI workloads. Future Outlook Looking ahead, Vicharak aims to make its software completely free while maintaining proprietary IP on their hardware. “Our plan is to integrate FPGAs into every kind of computer, much like CPUs and GPUs today. The software we develop will be free for use, but the IP for our FPGA designs will remain ours,” Vastarpara clarified. Reflecting on his journey, Vastarpara shared, “I graduated as a software engineer in October 2016. While I could write software, I realised that my true passion lay in electronics and hardware. That’s how Vicharak was born.” Initially, Vastarpara focused on consultancy projects, which allowed him to bootstrap his company. “We grew a team of 30 people, and by 2022, we had shifted our focus entirely to consumer-facing products,” he added. Vicharak is already garnering interest from various sectors. “We are set to demo our product within two months, working with government contractors for smart traffic systems and robotics startups. We are nearing the launch stage and expect to see our technology in practical use soon,” he said.","excerpt":"Vaaman is poised to make a difference for developers and researchers in India, with affordable prices and easy availability, cutting down reliance on hardware from other countries.","categories":["AI Features"],"tags":["Interviews and Discussions"],"author_name":"Mohit Pandey","publish_date":"2024-06-27T13:18:21","publication_year":"2024","word_count":716,"keywords":["CUDA","Go","funding","startup","programming_languages:R","AI","RPA","Aim","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","Aim","CUDA","R","Go","CUDA","RPA","startup","funding","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/this-surat-based-company-is-creating-ai-hardware-for-india\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":65869,"title":"Hands-On Guide to Download, Analyze and Visualize Twitter Data","content":"Twitter is a big source of news nowadays because it is the most comprehensive source of public conversations around the world. The latest news which may not be available on news channels or websites but it may be trending on twitter among public conversations. Any information, despite being constructive or destructive can be easily spread across the twitter network skipping the editorial cuts or regulations. Because of this nature of twitter, it is also being popular among data analysts who want to gather some trending information and perform analytics. In this article, we will learn to download and analyze twitter data. We will learn how to get tweets related to an interesting keyword, how to clean, analyze, visualize those tweets and finally how to convert it into a data frame and save it into a CSV file. Here, we will discuss a hands-on approach to download and analyze twitter data. We will import all the required libraries here. Make sure to install ‘tweepy’, ‘textblob‘ and ‘wordcloud‘ libraries using ‘pip install tweepy’, ‘pip install textblob‘ and ‘pip install wordcloud‘. #Importing Libraries import tweepy from textblob import TextBlob import pandas as pd import numpy as np import matplotlib.pyplot as plt import re import nltk nltk.download('stopwords') from nltk.corpus import stopwords from nltk.stem.porter import PorterStemmer from wordcloud import WordCloud import json from collections import Counter Downloading the data from Twitter To use the ‘tweepy‘ API, you need to create an account with Twitter Developer. After creating the account, go to ‘Get Started’ option and navigate to the ‘Create an app’ option. After you create the app, not down the below-required credentials from there. #Authorization and Search tweets #Getting authorization consumer_key = 'XXXXXXXXXXXXXXX' consumer_key_secret = 'XXXXXXXXXXXXXXX' access_token = 'XXXXXXXXXXXXXXX' access_token_secret = 'XXXXXXXXXXXXXXX' auth = tweepy.OAuthHandler(consumer_key, consumer_key_secret) auth.set_access_token(access_token, access_token_secret) api = tweepy.API(auth, wait_on_rate_limit=True) You can pass the keyword of your interest here and the maximum number of tweets to be downloaded through the tweepy API. #Defining Search keyword and number of tweets and searching tweets query = 'lockdown' max_tweets = 2000 searched_tweets = [status for status in tweepy.Cursor(api.search, q=query).items(max_tweets)] We will now analyze the sentiments of tweets that we have downloaded and then visualize them here. Sentiment Analysis #Sentiment Analysis Report #Finding sentiment analysis (+ve, -ve and neutral) pos = 0 neg = 0 neu = 0 for tweet in searched_tweets: analysis = TextBlob(tweet.text) if analysis.sentiment[0]>0: pos = pos +1 elif analysis.sentiment[0]<0: neg = neg + 1 else: neu = neu + 1 print(\"Total Positive = \", pos) print(\"Total Negative = \", neg) print(\"Total Neutral = \", neu) #Plotting sentiments labels = 'Positive', 'Negative', 'Neutral' sizes = [257, 223, 520] colors = ['gold', 'yellowgreen', 'lightcoral'] explode = (0.1, 0, 0)  # explode 1st slice plt.pie(sizes, explode=explode, labels=labels, colors=colors, autopct='%1.1f%%', shadow=True, startangle=140) plt.axis('equal') plt.show() Here, we will create a data frame of all the tweet data that we have downloaded. Later all the processed data will be saved to a CSV file in the local system. Through this way, we can utilize this tweet data for other experimental purposes. Creating the Data Frame and Saving into CSV File #Creating Dataframe of Tweets #Cleaning searched tweets and converting into Dataframe my_list_of_dicts = [] for each_json_tweet in searched_tweets: my_list_of_dicts.append(each_json_tweet._json) with open('tweet_json_Data.txt', 'w') as file: file.write(json.dumps(my_list_of_dicts, indent=4)) my_demo_list = [] with open('tweet_json_Data.txt', encoding='utf-8') as json_file: all_data = json.load(json_file) for each_dictionary in all_data: tweet_id = each_dictionary['id'] text = each_dictionary['text'] favorite_count = each_dictionary['favorite_count'] retweet_count = each_dictionary['retweet_count'] created_at = each_dictionary['created_at'] my_demo_list.append({'tweet_id': str(tweet_id), 'text': str(text), 'favorite_count': int(favorite_count), 'retweet_count': int(retweet_count), 'created_at': created_at, }) tweet_dataset = pd.DataFrame(my_demo_list, columns = ['tweet_id', 'text', 'favorite_count', 'retweet_count', 'created_at']) #Writing tweet dataset ti csv file for future reference tweet_dataset.to_csv('tweet_data.csv') Cleaning Tweet Texts using NLP Operations As we are ready now with the tweet data set, we will analyze our dataset and clean this data in the following segments. tweet_dataset.shape tweet_dataset.head() #Cleaning Data #Removing @ handle def remove_pattern(input_txt, pattern): r = re.findall(pattern, input_txt) for i in r: input_txt = re.sub(i, '', input_txt) return input_txt tweet_dataset['text'] = np.vectorize(remove_pattern)(tweet_dataset['text'], \"@[\\w]*\") tweet_dataset.head() tweet_dataset['text'].head() Here, as we are ready with the clean tweet data, we will perform NLP operations on the tweet texts including taking only alphabets, converting all to lower cases, tokenization and stemming. As retweets, hypertexts etc. are present in the tweets, we need to remove all those unnecessary information. #Cleaning Tweets corpus = [] for i in range(0, 1000): tweet = re.sub('[^a-zA-Z0-9]', ' ', tweet_dataset['text'][i]) tweet = tweet.lower() tweet = re.sub('rt', '', tweet) tweet = re.sub('http', '', tweet) tweet = re.sub('https', '', tweet) tweet = tweet.split() ps = PorterStemmer() tweet = [ps.stem(word) for word in tweet if not word in set(stopwords.words('english'))] tweet = ' '.join(tweet) corpus.append(tweet) Now, after performing the NLP operations, we visualize the most frequent words in the tweets through a word cloud and using the term frequency. Visualizing Highest Occurring Words using Word Cloud #Visualization #Word Cloud all_words = ' '.join([text for text in corpus]) wordcloud = WordCloud(width=800, height=500, random_state=21, max_font_size=110).generate(all_words) plt.figure(figsize=(10, 7)) plt.imshow(wordcloud, interpolation=\"bilinear\") plt.axis('off') plt.show() Analyzing Highest Occurring Words Term Frequency #Term Freuency - TF-IDF from sklearn.feature_extraction.text import TfidfVectorizer tfidf_vectorizer = TfidfVectorizer(max_df=0.90, min_df=2, max_features=1000, stop_words='english') tfidf = tfidf_vectorizer.fit_transform(tweet_dataset['text']) #Count Most Frequent Words Counter = Counter(corpus) most_occur = Counter.most_common(10) print(most_occur) The above all the most frequent terms appeared in the tweet data. So, this is the way how we can download the tweets, clean those tweets, convert them into data frames, save them into CSV file and finally, analyze those tweets.","excerpt":"Here in this article, we will discuss a hands-on approach to download and analyze twitter data.","categories":["Deep Tech"],"tags":["extract data from twitter","Sentiment Analysis","twitter analytics"],"author_name":"Dr. Vaibhav Kumar","publish_date":"2020-05-27T10:00:00","publication_year":"2020","word_count":892,"keywords":["twitter analytics","NumPy","Go","Sentiment Analysis","AI","sentiment analysis","NLP","extract data from twitter","analytics","NLTK","Matplotlib","R","Pandas"],"extracted_tech_keywords":["AI","NLP","analytics","NLTK","Pandas","NumPy","Matplotlib","sentiment analysis","R","Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/hands-on-guide-to-download-analyze-and-visualize-twitter-data\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10066723,"title":"Take advantage of the MSBA in the USA Learning Opportunity — University of Louisville College of Business is coming to your city. Register now to apply for free individual counselling sessions","content":"College admissions can be daunting and stressful, especially if one goes to another country to get a degree. With so many aspects involved: choosing colleges, completing the application, and appearing for admission tests, getting the information needed to make a good choice can be difficult. To ease this process, the University of Louisville College of Business, one of the leading institutions in the US, has organised in-person information meetings from May 22 to June 05 in six leading cities of India to answer prospective student questions. Classes start in August and in January Innovative course offerings College representatives will be answering questions about its STEM-designated Master of Science in Business Analytics (MSBA) program. The 13-month MSBA is taught by extensively published PhD-level faculty and subject matter experts. The curriculum helps students acquire highly in-demand skills such as data mining, data modelling, data visualisation, discrete choice modelling, machine learning, prediction algorithms, programming, statistical analysis, textual analysis, and web scraping. In addition, students will interact with corporate speakers in class and at networking events building their employment potential and connection with the industry. Additionally, students have the opportunity to gain work experience with a competitive paid internship opportunity (optional) and a capstone consulting project. The innovative internship program (optional) offers students ways to substantially offset the cost of tuition while providing valuable real-world work experience in the program. Most students seeking an internship are able to find a corporate partner within the early stages of the program, and many of these students are offered a career position afterwards, setting the students up for success and long-term achievement. Additionally, the University of Louisville College of Business student support is innovative and complete. From day one, assistance is available with application process assistance, guidance with visas and travel to the US, academic success, internship preparation\/career coaching, and after the program is over with their lifetime career support promise. What’s the process? Register for the city that works the best by clicking from the list below. Group or one-on-one meetings are available. Vijayawada – 20 May – REGISTER HERE Hyderabad – 22 May – REGISTER HERE Bangalore – 28 May – REGISTER HERE New Delhi – 29 May – REGISTER HERE Mumbai – 4 June – REGISTER HERE Pune – 5 June – REGISTER HERE Classes start in August and in January","excerpt":"UofL College of Business will be organising in-person meetings from May 22 to June 05 with prospective students in six leading cities of India.","categories":["AI Trends"],"tags":["Business Analytics"],"author_name":"Sreejani Bhattacharyya","publish_date":"2022-05-11T12:00:00","publication_year":"2022","word_count":387,"keywords":["Go","API","machine learning","programming_languages:R","AI","programming_languages:Go","analytics","Business Analytics","GAN","R"],"extracted_tech_keywords":["AI","machine learning","analytics","R","Go","API","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/university-of-louisville\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10121161,"title":"With Google’s Gemini 1.5 Flash, the Possibilities are Endless","content":"Google announced a new model, Gemini 1.5 Flash, at the Google I\/O 2024. It’s a lightweight AI model optimised for speed and efficiency with a massive context window of 1M tokens. Designed to handle tasks that require quick responses, it is capable of multimodal reasoning, which means it has the ability to simultaneously process and understand various types of data such as text, images, audio, and video. It is a valuable tool for situations where time and efficiency are crucial and can be used in various applications from a customer service chatbot and generating captions or images for social media posts, to scientific research and business analytics. People are still underestimating the value of Gemini 1.5 Flash. For $0.35, you can get 1 million tokens and start building natively multi-modal projects. The cost + latency + context window size + intelligence of Flash is going to create so many new startups.— Logan Kilpatrick (@OfficialLoganK) May 18, 2024 “Gemini 1.5 Flash excels at summarisation, chat applications, image and video captioning, data extraction from long documents and tables, and more,” wrote Demis Hassabis, the CEO of Google DeepMind. Hassabis further added that Google created Gemini 1.5 Flash to provide developers with a model that was lighter and less expensive than the Gemini 1.5 Pro version. Despite being lighter in weight than Gemini Pro, Gemini 1.5 Flash is just as powerful. This is because it’s been trained through a process called “distillation”, where the most essential knowledge and skills from Gemini Pro are transferred to 1.5 Flash but in a way that makes the Flash model smaller and more efficient. In addition to being the fastest in the Gemini family, it’s also more cost efficient to use, making it a faster and less expensive way for developers building their own AI products and services. How Does Gemini 1.5 Flash Compare to Other Models? Source: X Many users tested Gemini 1.5 Flash and compared it with other models and in most cases 1.5 Flash performed impressively. Over the weekend, I tested 3 LLMs to get relevancy score1. Haiku2. Gemini Flash 1.53. Perplexity: Llama3 Sonar 8BHaiku didn't care to follow my instructions most of the time, btw I used claude to write the prompt.Gemini Flash worked pretty wellPerplexity worked really…— Praveen Kumar | Building MevinX (@PraveenInPublic) May 20, 2024 When compared with GPT-4o, a user posted that 1.5 Flash performed almost as well as GPT-4o on the StaticAnalysisEval benchmark. Additionally, it is faster and more cost-effective than GPT-4o, making it a compelling alternative. A user tested GPT- 3.5 Turbo, Claude Haiku, and Gemini 1.5 Flash to check which model aligns most closely with GPT-4o in terms of accuracy for a specific classification task. Flash emerged as the clear winner. Another posted that Gemini 1.5 Flash was better than Llama-3-70b on long context tasks. “It’s way faster than my locally hosted 70b model (on 4*A6000) and hallucinates less. The free of charge plan is good enough for me to do prompt engineering for prototyping,” he wrote. A user ran 1.5 Flash on some evals for automatically triaging vulnerabilities in code, and did the same with GPT-4-Turbo hosted on Azure, Llama-3 70B hosted on Groq, and GPT-4o hosted on OpenAI as well. “It’s very fast and very cheap. The results were pretty much on par with the other models in terms of accuracy,” he concluded. I played with Google's new Gemini 1.5 Flash model over the weekend and was quite impressed. It's not the best model out there, but can be very powerful if it works for your use case.It's more verbose, but very fast and very cheap.I ran it on some of our evals for… pic.twitter.com\/mJIff2gerA— Stefan Streichsbier (@s_streichsbier) May 20, 2024 Another user ran various tests for both Gemini Flash as well as GPT-4o and agreed that Google’s new model is impressive – cheaper, sometimes faster, and gives similar results to GPT-4o. “A combination of the two using LLM agentic workflow is the solution,” he added. The new gemini-flash is 19x cheaper than gpt-4o & nearly as good.But I don't trust benchmarks. So I run my own tests:test #1 → analyze youtube for me pic.twitter.com\/E1bZsbjqzj— Ruben Hassid (@RubenHssd) May 19, 2024 However, some have also raised concerns about the model’s low rate limit that is creating roadblocks in using it in production in any capacity. Source: X Interesting Use Cases of Gemini 1.5 Flash Online users have been trying their hands on the model and are coming up with interesting use cases. DIY-Astra, a multi-modal AI assistant powered by Gemini 1.5 Flash Introducing DIY-Astra, a small but powerful web app powered by Gemini 1.5 Flash. ⚡️Astra will tell you anything it sees in the camera, essentially in real-time.I was so impressed when I saw that it can also solve visual questions as well.Repo in the comment. pic.twitter.com\/eHUHPGEMHg— Pietro Schirano (@skirano) May 16, 2024 The 1M token context, low cost, and high speed of Gemini 1.5 Flash make it a perfect tool to create exciting applications like these. Gemini 1.5 Flash for WebScrapping Gemini 1.5 Flash is ideal for web scraping. It simplifies the process by eliminating the need for HTML selectors and adapts to various HTML structures across devices, countries, and products. The model works efficiently with any web page technology, including JavaScript and pre-rendered HTML. https:\/\/twitter.com\/xaviramirezcom\/status\/1792702724733116867 Analyse a Video to Produce Script An online user gave Gemini 1.5 Flash a video recording of him shopping and it generated the Selenium code of the site in just about 5 seconds. https:\/\/twitter.com\/minchoi\/status\/1791846425363398698 Gemini-1.5-Flash as a Copilot in VSCode By connecting CodeGPT with Google AI Studio, you can leverage the power of Gemini 1.5 Flash to enhance your coding experience. Gemini-1.5-flash as a Copilot in VSCode is amazing!You can now use this model by connecting CodeGPT with Google AI Studio.@codegptAI + @googleaistudioIn this video, I show how CodeGPT manages to get the entire context of the \"Quick Fix\" section and Gemini provides a… pic.twitter.com\/E6eGczLgtb— Daniel San (@dani_avila7) May 15, 2024 A Great Option for Voice AI Gemini 1.5 Flash is a great option for voice AI, with first token around 500 ms and 150 tokens\/s. Gemini 1.5 Flash is a game changer for voice-based products. Adding it to Voqal really shows how easy it will be to interact with machines in the future.It took <5min to \"teach\" my assistant how to watch my CI builds and alert me when they finish. Zero keywords. Zero wake… pic.twitter.com\/t7NFJgRkxT— Voqal (@voqaldev) May 16, 2024 Gemini YouTube Researcher Let Gemini be your YouTube researcher. Simply input a topic, and the AI analyses relevant videos to deliver a comprehensive summary, simplifying your research by extracting key insights efficiently. 1. Gemini YouTube Researcher– Listens to videos & delivers topical reports.– Write a topic and AI will analyze relevant videos & provide a comprehensive report. pic.twitter.com\/pQ0EkMPAXg— Saumya Singh (@saumya1singh) May 20, 2024 This shows that with Gemini 1.5 Flash’s cost, latency, and 1M tokens context, alongside the OpenAI GPT-4o, which is also plausibly a lightweight model, the possibilities are endless.","excerpt":"The cost + latency + context window size of Flash can create so many new startups.","categories":["AI Features"],"tags":["DeepMind","demis hassabis","Google"],"author_name":"Sukriti Gupta","publish_date":"2024-05-21T16:46:45","publication_year":"2024","word_count":1166,"keywords":["Gemini Pro","OpenAI","demis hassabis","AI","GPT-4o","ML","R","RAG","prompt engineering","analytics","Google","Azure","DeepMind"],"extracted_tech_keywords":["AI","ML","analytics","GPT-4o","OpenAI","Gemini Pro","RAG","prompt engineering","Azure","R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/with-googles-gemini-1-5-flash-the-possibilities-are-endless\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10075985,"title":"Why the Master of Science in Accounting &#038; Analytics programme from the University of Louisville is the right choice to drive your CPA career","content":"Certified Public Accountant (CPA) is one of the most sought-after profiles for accountants. Research claims that a CPA in the US could earn an average salary between USD70,235 and USD461,014. However, succeeding at the CPA test is challenging for most aspirants—even more so considering the constant evolution of the field. In today’s data-rich world, accounting analytics is decidedly gaining more prominence. Traditionally, accountants have worked with data in the form of numbers related to taxes, revenues, and expenditures. However, with the recent emergence of Big Data and analytics tools, accounting firms are incorporating such modern technologies into their business models. With the introduction of accounting analytics, professionals in the field are now leveraging data science technologies to uncover new insights and patterns in finance as well as perform predictive analysis, increase efficiency, and manage risks. Hence, to become a certified public accountant, one must have the right skill set that the industry demands and for this, undertaking the right course is key. Also, please note that the candidates are required to have a bachelor’s degree in accounting to be eligible for this programme. Why take up the MSAA programme? The MSAA programme from the University of Louisville is STEM (science, technology, engineering and mathematics)-designated and offers the essential tools, resources, and support to develop and enhance the candidates’ accounting skill set. In addition, as a STEM-Designated program, the MSAA provides international students with up to THREE YEARS of post-completion work authorisation in the USA. While it helps you achieve the right skill set in accounting analytics, the programme also comprehensively covers the fields of accounting, auditing, and financial reporting, which are necessary for the CPA test. The Master in Accountancy and Analytics programme offered by the University of Louisville was globally ranked #27 by Best Masters 2022. The MSAA programme offered by the university is the same but with a keener emphasis on the analytics skills needed to lead in the data-driven requirements of the current accounting industry. What does the programme offer? The Master of Science in Accountancy and Analytics is a 34-credit hour cohort-based programme. Graduates of the programme may also receive the Graduate Certificate in Managerial Analytics. This graduate certificate is a twelve (12) credit hour certificate that provides business professionals with the skills to thrive in a data-driven environment. This programme is designed to: ●  Equip graduates with technical accounting knowledge of GAAP (Generally Accepted Accounting Principles). ●  Prepare graduates to assume leadership positions in the professional practice of accounting as Certified Public Accountants. ●  Provide graduates with functional knowledge of data science from a managerial perspective. ●  Impart graduates with a programmatic or professional understanding of data results effectively, especially in terms of data visualisation and data storytelling. While in the programme, the candidates get unlimited access to Becker CPA or CMA Review software with the guided study. This could go a long way in helping them prepare for the CPA or CMA test. Becker CPA or CMA Review software is one of the topmost programmes in the market to help candidates prepare for the CPA or the CMA test. Further, the MSAA programme includes a competitive 11-month paid internship, which helps you test your knowledge and skills in a real-life scenario. It also includes a 10-day international study trip or an online global-learning course. In addition, candidates can network with professionals and leaders in public accounting and other industries. How to apply? Visit the University website, create an application account, and then visit the College of Business graduate programs website to apply for the programme. Candidates are required to attach two letters of recommendation, which should speak about the candidate’s character, academic aptitude, and other academic and professional achievements. The candidate must also provide the name and email address of the recommending persons. Further, candidates must provide a personal statement. The personal statement is a one to a two-page document describing their interest in the University of Louisville’s MSAA programme. The document must define why the candidate has selected the University of Louisville, their short-term and long-term goals, and any additional information they feel is relevant. The next steps are to provide their professional resume and official college transcripts from all colleges attended. No GMAT is required. Candidates with GMAT or GRE scores are welcome to provide them to strengthen their application. Scores of 700+ will be reviewed for additional scholarship consideration. Learn more about the Master of Science in Accounting and Analytics (MSAA) Program here. You can apply for the programme here.","excerpt":"The MSAA programme from the University of Louisville is STEM-designated and offers the essential tools, resources, and support to develop and enhance candidates’ accounting skill sets.","categories":["AI Trends"],"tags":[],"author_name":"Pritam Bordoloi","publish_date":"2022-09-28T11:00:00","publication_year":"2022","word_count":749,"keywords":["big data","data science","Go","programming_languages:R","AI","data-driven","RAG","Aim","analytics","R"],"extracted_tech_keywords":["AI","data science","analytics","Aim","RAG","R","Go","big data","data-driven","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/why-the-master-of-science-in-accounting-analytics-programme-from-the-university-of-louisville-is-the-right-choice-to-drive-your-cpa-career\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":48653,"title":"How To Crack Data Science Interview At Microsoft","content":"While preparing for a data science job interview at Microsoft, candidates need to prepare about topics ranging from analytical thinking to problem-solving. In this article, we list down some of the top data science questions to ace your interview as well as some important tips to prepare for D-day. In a blog post by Microsoft, the recruiters shared that they expect the candidates to have a passion for coding, solving problems and the incredible things technology can do for people around the world. Every interview looks for something specific depending upon a particular job profile. Technical Interview A technical interview varies depending upon the team and the role which the candidate is going to apply for. Some of the important points to be noted while appearing in a data science interview are: A Neat Resume: In an interview, a resume acts as a mirror for your experiences and achievements. A resume must be neat enough to understand. Also, the candidate must include real-life examples, projects, and experiences in the resume by depicting its impact as well as how it benefits the end-users. Before an interview, one must also go through all the topics which are mentioned in a resume. Communication Skills: Interviewers are no mentalists. This is a general point where strong communication is needed while appearing for an interview. Also, while responding to any questions, the candidate must understand what is being asked by the interviewers and respond accordingly. Delving Deep Into The Basics: In a data science interview, a candidate must dive deep into topics such as the basics of machine learning algorithms, statistics, linear algebra, probability, among others. There is a plethora of online courses where one can easily brush up the basic concepts beforehand. The Code To Code: While appearing in a personal interview round, the interviewer might ask you to solve a coding problem. The right way to approach it is to ask any doubt or questions related to the problem and not just directly rush into solving the problem. Following this rule will not only help you to complete the task with a more accurate answer but also it will showcase your communication skills and the way of understanding and solving a problem. Solve The Questions Now, we will take a look at the questions which are usually asked in a data science interview at the tech giant and preparing them will surely help you to ace your data science interview. Given a string like “fun1(fun2(a,b) ,c ,fun2(a),fun3(c,d))” how many unique function signatures are there. You have a bag with 6 marbles. One marble is white.  You reach the bag 100 times. After taking out a marble, it is placed back in the bag. What is the probability of drawing a white marble at least once? Given a box of dimensions W, H, and coordinates of points inside that box. Find the largest area that is free of any of these points. Why neural network work and why is it a booming field? What’s power? How to explain it to a non-statistics person? what’s a false positive and false negative? How Random Forest, Lasso and Ridge Regression work? Difference between lasso and ridge. Write a function to check whether a particular word is a palindrome or not. Find the maximum of subsequence in an integer list. Generate a fair coin from a biased one. Generate 7 integers with equal probability from a function which returns 1\/0 with probability p and (1-p). What are the ROC curve and the meaning of sensitivity, specificity, confusion matrix? Given a time series dataset, how will you predict future value? How to explain a deep learning model to customers? What is the definition of a P-value? How to explain p-value to customers. How can you compute an inverse matrix faster by playing with some computation tricks? Describe how gradient boost works. Describe the steps for data wrangling and cleaning before applying machine learning algorithms. How to deal with unbalanced binary classification? How do you detect if a new observation is an outlier? What is the bias-variance trade-off? Explain the Support Vector Machine (SVM). Wrapping Up The interview questions above have been cited from various sources, comments and reviews about data science interviews at Microsoft. Also, we can see that the questions are mostly related to probability, basics of machine learning algorithms, and other techniques of extracting data in an ML method. Thus, a candidate with a good grip on these topics will surely ace the data science interview.","excerpt":"While preparing for a data science job interview at Microsoft, candidates need to prepare about topics ranging from analytical thinking to problem-solving. In this article, we list down some of the top data science questions to ace your interview as well as some important tips to prepare for D-day. In a blog post by Microsoft, […]","categories":["AI Trends"],"tags":["Data Science","data science interview","Microsoft"],"author_name":"Ambika Choudhury","publish_date":"2019-10-23T14:00:27","publication_year":"2019","word_count":749,"keywords":["data science","Go","machine learning","data science interview","programming_languages:R","AI","neural network","ML","deep learning","ViT","Data Science","R","Microsoft"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","neural network","data science","R","Go","ViT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/how-to-crack-data-science-interview-at-microsoft\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":13691,"title":"Textual &#038; Spreadsheet Data – Effective Data Science Series: 5 of 5","content":"When the data scientist goes after structured and machine-generated data, experience has shown that there are not many positive results. Instead, the most fertile grounds are textual data and spreadsheet data. Fig 1 depicts textual and spreadsheet data. But – as has been discussed – there is a barrier to accessing and analyzing textual and spreadsheet data. Textual and spreadsheet data is not “well behaved”. Spreadsheet and textual data is erose, and common data base management systems do not hold or interact well with erose data. But it is noted that just because textual data forms a basis for business value does not mean that ALL textual data is useful for finding business value. Fig 2 notes that there is some amount of textual data that is not fit to serve as a basis for finding business value. Some textual data is informal. Some textual data is hearsay. Some textual data is casual. So textual data must be vetted as to its suitability to serve as a basis for business value. The same is true for spreadsheet data. Fig 3 shows that some spreadsheets are not suitable to serve as a basis for finding business value. Some spreadsheets are informal. Some spreadsheets are casual. Some spreadsheets are created at 9:00 am and are deleted at 10:00 am. There are many reasons why a spreadsheet may not be a good candidate to serve as a basis for finding business value. Fig 4 shows that there is a continuum of spreadsheets. In actuality probably only 10% or less of the spreadsheets the corporation has are fit to serve as a basis for finding business value. Once the organization has vetted both textual data and spreadsheet data, the next step is to employ technology that allows the data to be transformed into a standard data base management system. There are two very different technologies that are required. For text, there is textual disambiguation, as seen by Fig 5. And for spreadsheets there is spreadsheet disambiguation, as seen in Fig 6. At a high level, textual disambiguation and spreadsheet disambiguation appear to be similar, because they both achieve the same function. They both convert unstructured data into a standard data base management system. But once you look inside the two technologies they are nothing alike. Textual disambiguation deals with the vagaries of language and text, while spreadsheet deals with the idiosyncrasies of spreadsheets. Once text and\/or spreadsheets have been disambiguated, they are turned into a standard data base. And after they have been turned into a standard data base, then (and only then) the data scientist can stat to do his\/her analysis. It is disambiguation technology that breaks down the shield of opaqueness that surrounds text and spreadsheets.","excerpt":"When the data scientist goes after structured and machine-generated data, experience has shown that there are not many positive results. Instead, the most fertile grounds are textual data and spreadsheet data. Fig 1 depicts textual and spreadsheet data. But – as has been discussed – there is a barrier to accessing and analyzing textual and […]","categories":[],"tags":[],"author_name":"William Inmon","publish_date":"2017-03-23T05:01:09","publication_year":"2017","word_count":454,"keywords":["Go","programming_languages:R","programming_languages:Go","GAN","R"],"extracted_tech_keywords":["R","Go","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/textual-spreadsheet-data-effective-data-science-series-4-5\/","complexity_score":3,"technical_depth":5,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":60717,"title":"10 Most Influential Analytics Leaders In India &#8211; 2020","content":"Analytics India Magazine has unveiled its list of the 10 Most Influential Analytics Leaders for the year 2020. Selected on the basis of the significant expertise they hold in the field of analytics and data science, it recognizes the exceptional work they have accomplished in the last year. AIM 100: The Most Influential Leaders in AI 2024As AI reshapes our world, AIM introducing the most influential global leaders in AI. Given the pace at which the analytics industry has been growing, the timing could not have been more opportune. The growing importance of delivering value in a business through insight-driven decisions has made their contributions not only critical, but imperative as well. Furthermore, these leaders have consistently adopted newer paradigms in their practices, potentially breaking new ground in the fields of analytics, artificial intelligence, and more. Every year we scour through organizations to identify leaders that best exemplify the data-driven ideal that befits our times. This year’s list includes some from our previous lists, and a few others who have not only excelled at creating innovation, but are also deemed to be some of the most influential individuals in India in this field. Find out who made it to the 2019 list here. Here is the list of top 10 influencers in the Indian analytics industry for 2020, presented in alphabetical order:- Ashwin Mittal CEO of Course5 As CEO, Ashwin Mittal has led Course5 into the analytics and AI powerhouse it has become today. With prior experience in the corporate world in India and in the US, he has introduced new technologies for digital transformation at the company. This has taken him down a path where he has successfully built world-class AI-based products, and helped leading global brands reconfigure their businesses to compete in the digital era. Ashwin has received several accolades and awards from leading global authorities for his achievements in the field of analytics. He has also spoken at several global forums on driving business impact through analytics and AI. Under his leadership, Course5 has also won numerous recognitions, including Deloitte Technology Fast 50 India 2019 Award, and Top 50 Tech Innovators award at InterCon Dubai 2019, among a long list of others. Besides work, this UCLA and London Business School alumnus is also involved in various CSR activities around child education and health. A Chartered Accountant by qualification, he is also a member of the Young Presidents’ Organization and Entrepreneurs’ Organization. Key Achievements last year: Gartner’s ‘Cool Vendors in AI for Marketing’ report – Course5 Adomate (Computer Vision-based Ad Optimization platform) One of the most important M&CI solutions in Forrester’s Landscape Overview report – Course5 Compete (Real-Time Competitive Intelligence platform) ‘New Product of the Year’ at Enterprise at BIG Awards for Business 2019 – Course5 Discovery (Natural Language based AI Insights platform) Great Learning Data Science Award for ‘Data Science\/AI in FMCG’ at Cypher 2019 Course5 achieved Gold Competency status for Microsoft’s suite of key Data Analytics products Read an interview with Ashwin Mittal here. Deep Thomas Chief Data & Analytics Officer at Aditya Birla Group Deep Thomas is a reputed analytics expert, thought leader and a passionate evangelist of data science, with a distinguished career spanning more than two decades. As Chief Data and Analytics Officer at Aditya Birla Group, he has been leading the charge of its AI-led digital transformation. In this role, he has enabled operational efficiencies, driven new business growth, and delivered successful outcomes across the organization’s multi-sector businesses. He has also delivered sustained and increasing profitability through low cost technology and analytics-led open innovation. Under his leadership, the Aditya Birla Group has successfully forayed into a data-driven culture, with many innovative solutions in core areas such as safety, procurement, quality, energy, pricing, and supply chain, among others. Prior to Aditya Birla Group, this Arizona State alumnus was the founder and CEO of Tata Insights and Quants, Tata Group’s Big Data and Decision Science company. He had also held key positions in the US and India with multinationals like Citigroup, HSBC, and American Express to steer their global digital and analytics agenda. Key Achievements last year: Under his leadership, the Aditya Birla Group has introduced many innovative solutions deployed through cutting-edge AI techniques, Computer Vision, NLP etc. in core areas of its businesses Deep has had the unique privilege of setting up the analytics arm for the founder institutions of modern India, Tatas and Birlas This is the fourth time Deep has been featured on this list Find out who made it to the 2018 list here. Mahesh Kumar Founder & CEO of Tiger Analytics As CEO of one of the fastest-growing AI and advanced analytics consulting firms globally, Mahesh Kumar has maintained a significant focus on building a company known for solving high-complexity problems. Having emerged as one of the most sought-after players in the industry, Tiger Analytics today services over 35 Fortune 1000 clients across a wide range of industries. Prior to this, Mahesh was on the faculty of the Smith School of Business and Rutgers Business School. He has also conducted research in the areas of data mining and statistical modeling, and has successfully applied his research to solve real-world problems. Furthermore, this IIT-Bombay alumnus also holds a PhD. in Operations Research and Marketing from MIT, and is a highly sought-after speaker in the analytics community. And all of this, while maintaining work-life harmony. In fact, Mahesh firmly believes that hiring the right talent, creating an environment of learning and sharing, and enabling open-door access to leadership has been the key to the success of the company. Key Achievements last year: Tiger Analytics has been listed among Deloitte’s Fastest Growing Companies in India, and Inc.5000’s fastest growing companies in the US It has been rated by Forrester as a leading analytics service provider It has also been recognized as Frost & Sullivan’s AI Solution Provider of the Year 2019 This is the second time Mahesh has been featured on this list Read an interview with Mahesh Kumar here. Mehul Bhagadia Chief Data & Analytics Officer at IDFC First Bank With over two decades of experience in data analytics, business management, strategy and service delivery, Mehul Bhagadia has built and managed teams across the analytics life cycle. His career has taken him to deliver advanced analytics for large organizations like JP Morgan, where he coordinated with a team of over 250 across India and Philippines. He has also worked for ICICI Bank and Citi. Furthermore, Mehul has been the recipient of the Analytics 100 award at MachineCon 2019. As Chief Data and Analytics Officer with IDFC, he has been instrumental in participating in data strategy design, building new products, and working on the customer journey roadmap. Mehul has also been instrumental in adding advanced analytical capabilities – like machine learning, Big Data infrastructure with Hadoop, GreenPlum and programming languages like Python, R and visualization tools like Tableau – for the Global Analytics Center in India. Key Achievements last year: In the last 10 months, he has been instrumental in participating in the data strategy design for the bank, building new products and working on the customer journey roadmap His core responsibilities have been to deliver the Liabilities, Wholesale Banking, Operations and Customer Delight Analytics initiatives for the bank Find out who made it to the 2017 list here. Prithvijit Roy CEO & Co-founder of Bridgei2i An expert AI practitioner with over two decades of analytics experience, Prithvijit Roy has led BRIDGEi2i for nearly nine years. In fact, the company was ranked among the best places in India for data scientists to work for in 2019. Prithvijit has been influential in shaping the company from a pure-play analytics firm, to a global AI company. Recognized as a thought leader in the AI and analytics space, Prithvijit has been fervently advocating for the adoption of AI-thinking at scale with the objective of making AI more pervasive. He has also been instrumental in spearheading an AI community called Humans of Analytics – a forum for like-minded individuals to connect and celebrate, both their personal and professional journeys through inspiring stories and anecdotes. Over the last one year, he has been focusing on leveraging cutting edge data science techniques and newer ML algorithms to drive innovation at BRIDGEi2i. An alumnus of Indian Statistical Institute, Prithvijit was previously involved in the setting up of analytics centers in India for HP and GE\/Genpact. Key Achievements last year: Over the last one year, the company has been focussed on leveraging cutting edge data science techniques that has translated into actionable insights and embedded algorithms used by partner clients to drive intelligent automation, thereby fuelling their AI-powered digital transformation journey Prithvijit is recognized as a thought leader in the AI and analytics space, and has been visible at several industry forums including Cypher, where he encouraged the adoption of AI-thinking at scale This is the fifth time Prithvijit has been featured on this list Read an interview with Prithvijit Roy here. Rajeev Baphna Founder & CEO of Analyttica Datalab In his 30+ years of corporate and analytical leadership, Rajeev Baphna has taken up global visionary and management roles, and partnered with global clients in virtually every sizable market across the globe. He is a seasoned strategic leader who has built analytical centres across 25+ countries. He specializes in application of domain contextual analytics and AI to enable scalable business solutions. Rajeev has a deep interest in solution creation through technology-enabled experiential and computational analytical platforms, and simulation-based methodologies that are completely integrated in a powerful business analytical ecosystem. As founder and CEO of Analyttica Datalab, he has directed the company in the field of analytics and AI to build impactful and scalable solutions by leveraging state-of-the-art technologies. An innovator at heart with a passion for the application of domain contextual analytics, Rajeev has various respected publications and patents. He is also the recipient of several globally recognized awards for excellence. Key Achievements: Rajeev Baphna over a span of his career has delivered substantial business impact to the tune of $100 million in productivity improvement, and a $1billion incremental revenue impact to businesses globally Analyttica Datalab’s Augmented Analytics and AI platform – Analyttica TreasureHunt (ATH), has been getting significant recognition from across the globe. Find out who made it to the 2016 list here. Sandeep Mittal MD of Cartesian Consulting With Cartesian – a promising firm in the domestic analytics industry – Mittal has developed a practice that has served over 200 clients in more than a dozen countries globally. Over the years, he has emerged as a trusted advisor to many CXOs from diverse industries on all things analytics. Although he has worn many hats – including online magazine founder-editor, tech entrepreneur, and even a musician – his current focus is on expanding the product portfolio and the sales organization at Cartesian. In fact, his favourite area of work is hyper-personalization and its ability to nudge customer behaviour. Known to be an inspirational leader, this IIM alumnus is a frequent speaker in industry forums and academia, and is best known for his storytelling and ability to add a creative spark to the way data is consumed. His cartoons and sketches which usually accompany his presentations is a much loved trademark touch. Key Achievements last year: With SOLUS.ai, a product incubated by Cartesian, he has developed a practice that has served over 200 clients in more than a dozen countries globally He has mastered hyper-personalization, and has learned to successfully leverage it to nudge customer behaviour This is the fifth time Sandeep has been featured on this list Read an interview with Sandeep Mittal here. Dr Shailesh Kumar Chief Data Scientist at Reliance Jio Center of Excellence in AI\/ML Dr Kumar has put in over twenty years in research and startups, and has industrial experience across various areas of AI. In the last year, he has invented a number of new frameworks and platforms, including domain and language agnostic Question Answering System, and a new method of discovering biomarkers for specific Phenotypes. His AI for India team recently received the prestigious Aegis Graham Bell Award in the ‘Technology for Social Good category’ for the Machali app’ that enables fishermen across ten coastal states of India to be safer and more productive by using personalized and localized advisories. He has published over 20 conference papers, journal papers, and book chapters, and holds as many patents in these areas. A graduate from IIT-Varanasi, Dr Kumar received his PhD and Masters in Computer Science from the University of Texas at Austin. He also serves as a program advisor for the AI\/ML program at Jio Institute, and is a visiting faculty of machine learning at Indian School of Business. Prior to this, he was a distinguished scientist at Ola, researcher in the Google Brain team, Senior Scientist at Yahoo! Labs, Principal Scientist at Fair Isaac Research, and co-founder of ThirdLeap – an edtech company. Key Achievements: Dr Kumar is working on three broad charters at Jio: First, building an ‘AI Stack for India’ involving speech and language capabilities in the Indian context. Second, building ecosystem solutions for healthcare and agriculture under the ‘AI for India’ initiatives of Jio, and third, building systems of intelligence for telecom, retail, media, manufacturing, and financial services for various Reliance businesses In the last year, he has invented a number of new frameworks and platforms including novel frameworks for Optimization and Supply Chain logistics for both operational and manufacturing optimizations using Digital Twin technologies His AI for India team recently received the prestigious Aegis Graham Bell Award in the ‘Technology for Social Good category’ for the Machali app’ that enables fishermen across ten coastal states of India to be safer and more productive by using personalized and localized advisories Read this story with Dr Shailesh Kumar here. Shub Bhowmick CEO and Co-founder of Tredence Over the course of 20 years, Shub Bhowmick has built a career in unlocking value and driving customer impact. Under his able leadership, Tredence has maintained profitability and has created sustainable value for its 30+ Fortune 500 clientele, as well as its 600+ employees. What is more, under Shub’s leadership, Tredence has grown at 60%+ CAGR in the last five years. Prior to founding Tredence, this IIT alumnus also held senior executive positions in Diamond Consultants (now PwC), Mu Sigma, and Infosys Technologies. Key Achievements last year: Recognized as a ‘Strong Performer’ in The Forrester Wave™: Customer Analytics Service Providers, Q2 2019 Listed for the second time in Gartner Market Guide for Data and Analytics service providers Listed for the fourth consecutive time in the INC 5000 list, America’s fastest growing private companies Ranked fifth among the fastest growing private companies in silicon valley by SVBJ for the second time This is the second time Shub has been featured on this list Find out who made it to the 2015 list here. Sreekanth Menon VP – Data Science at GENPACT Sreekanth Menon brings over two decades of industry expertise to his role heading the AI\/ML practice for Genpact. He has been leading strategy, business transformation and product development here, and has been the key architect of the ML incubation program at GENPACT to drive ML first culture across the organization. His primary focus has been competency building in AI ecosystems, such as ML, NLP\/Text Mining, Computer Vision and nurturing new capabilities. With this, he has incubated and launched over 50 advanced analytics solutions in the global market, and worked closely with Fortune 500 clients by enabling them with innovative AI-led solutions and practices. Prior to this, this XLRI alumnus had led a large analytics team at Symphony Marketing Solutions, and also assumed end-to-end responsibility of developing the analytics engine, setting up virtual captive teams and building Big Data Analytics platforms. Key Achievements last year: Sreekanth incubated the AI team at Genpact in 2016-2017, and further scaled it in 2019 He is responsible for delivering AI\/ML revenue with successful client deliveries of over 20 AI\/ML solutions across banking, consumer goods, retail, insurance and life sciences He has spearheaded delivery of AI products in the area of Insurance Claims and Supply Chain Find out who made it to the 2014 list here.","excerpt":"Analytics India Magazine has unveiled its list of the 10 Most Influential Analytics Leaders for the year 2020. Selected on the basis of the significant expertise they hold in the field of analytics and data science, it recognizes the exceptional work they have accomplished in the last year. Given the pace at which the analytics […]","categories":["AI Features"],"tags":["ai publications by india","analytics insights platforms real-time business","analytics leaders","corporate analytics platform","enterprise analytic hub","fields of analytics","intelligent automation for retail","mba in business analytics","online masters analytics","Speech Analytics","Visual Analytics Provider"],"author_name":"Anu Thomas","publish_date":"2020-04-02T13:00:00","publication_year":"2020","word_count":2695,"keywords":["computer vision","online masters analytics","Ray","enterprise analytic hub","data science","artificial intelligence","fields of analytics","corporate analytics platform","intelligent automation for retail","NLP","analytics insights platforms real-time business","Visual Analytics Provider","analytics","machine learning","Speech Analytics","AI","ML","mba in business analytics","ai publications by india","analytics leaders","Aim"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","NLP","computer vision","data science","analytics","Aim","Ray"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/10-most-influential-analytics-leaders-in-india-2020\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":33571,"title":"Why You Should Learn Matlab For Data Science","content":"Data Science is one of the fastest growing fields in India and Matlab comes with a very ease of learning. In this article, we will talk about Matlab, the programming language developed by MathWorks which is suitable platform for predictive analysis and is easy to implement new features. Matlab is very adaptive to data science and is widely applied in a range of industries from finance, energy and medical devices to industrial automation, automotive and aerospace in various functions for business-critical applications. However, in data science, Matlab is not as popular as Python and R. This could be mainly because it is not free. In an interview with Analytics India Magazine, Prashant Rao, Technical Manager at MathWorks India said, “Our goal is to make MATLAB accessible and easy for engineers and scientists to use for deep learning. With tools and functions for managing large data sets, MATLAB also offers specialized toolboxes for working with machine learning, neural networks, computer vision, and automated driving”. Here’s Why Matlab Works For Data Science 1.Rich ML libraries: Matlab has a toolbox – the Deep Learning Toolbox which provides simple Matlab commands for creating and interconnecting the layers of a deep neural network. It has Parallel Computing Toolbox to distribute training across multicore CPUs, graphical processing units (GPUs), and clusters of computers with multiple CPUs and GPUs. Deep Learning Toolbox, which replaces Neural Network Toolbox, provides a framework for designing and implementing deep neural networks. It can be used to design complex neural architectures, more easily. It exhibits a full set of capabilities for deep learning and provides end-to-end integrated workflow from research to prototype. 2.Best for matrix calculations: Machine learning deals with a lot with matrices. Matrix operations are used in the description of many ML algorithms and are the fundamentals of linear algebra. They are used in training an algorithm and processes. Transpose operation are used for flipping the dimensions of a matrix, inverse matrix operations are used in solving systems of linear equations. When working with NLP, documents are represented as large matrices  of word occurrences. For example, the columns of the matrix may be the known words in the vocabulary and rows may be sentences, paragraphs, pages, or documents of text with cells in the matrix marked as the count or frequency of the number of times the word occurred. Dealing with images also requires matrix fundamentals and Matlab comes with an easy set of operations for it. 3.Fewer lines of code: Matlab can be dubbed as an ML rich language with built-in library. Because of this, the script is very small and equally effective compared to other languages. The design of the language makes it possible to write a powerful program in a few lines. Also, codes in Matlab look exactly like the equation you are trying to solve. So demonstrating and debugging code becomes very easy. Free Online Sources To Learn Matlab 1.Introduction to programming with MATLAB: This one from Coursera is aimed at beginners. However, this isn’t a data science restricted course. This course is best suited for beginners who have no prior experience in computer science, as it includes various functions and commands used in Matlab. It is an introductory programming course that uses MATLAB to illustrate general concepts in computer science and programming. 2.Andrew Ng’s course on ML: This popular introductory course on ML has practical sessions throughout the course. The practicals are done in Octave or Matlab and will learners a brief overview of how things pertaining to machine learning can be programmed in Matlab. 3.Introduction to Matlab: This is an online course by MIT Opencourseware. The course is intended for students with no prior programming experience and provides the foundations of programming in Matlab. Variables, arrays, conditional statements, loops, functions, and plots, that are useful and are fundamentals of ML, are explained clearly. 4.Matlab YouTube: Matlab has its own YouTube channel that showcases Matlab programming video tutorials. There are also various videos of conversations with experts in Matlab. In Conclusion Matlab is not very popular when it comes to data science but it is one of the languages that many people consider for learning data science. Researchers, scientists and engineers who are already using MATLAB find it easy to move to deep learning because of the functionality of the Deep Learning Toolbox.","excerpt":"Data Science is one of the fastest growing fields in India and Matlab comes with a very ease of learning. In this article, we will talk about Matlab, the programming language developed by MathWorks which is suitable platform for predictive analysis and is easy to implement new features. Matlab is very adaptive to data science […]","categories":["AI Features"],"tags":["Deep Learning","library","tutorial"],"author_name":"Disha Misal","publish_date":"2019-01-16T07:59:28","publication_year":"2019","word_count":718,"keywords":["data science","machine learning","AI","neural network","library","ML","computer vision","NLP","Aim","deep learning","analytics","tutorial","Deep Learning"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","neural network","NLP","computer vision","data science","analytics","Aim"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/why-you-should-learn-matlab-for-data-science\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":34958,"title":"How These Data Science Enthusiasts From Christ University Solved Our Insurance Products Hackathon","content":"Team Neuron, the winning team of the Republic Day hackathon by TEG Analytics and Analytics India Magazine TEG Analytics and Analytics India Magazine had organised a hackathon called Predict Market Competitiveness For Insurance Products, this year on the Republic Day. The problem involved predicting the market share of insurance companies in the US, affiliated to Medicare. After an offline hackathon, a set of finalist teams were shortlisted to present their case study at the Machine learning Developers Summit (MLDS) organised by AIM on 30-31 January 2019, after which a winning team was selected. Analytics India Magazine got in touch with the four winners to know about them and find out how they solved the hackathon problem: The Journey Called ‘Team Neuron’, the team consisted of four 3rd year undergraduates from Christ University, Bengaluru. The team members are Sreyan Ghosh, Samden Lepcha, Sonal Kumar and Joshua Jiji. Their journey in data science began in their second year of engineering after attending an introductory machine learning workshop organised by their seniors. The team on the same day went back home and signed up for the machine learning course on Coursera by Andrew Ng. Since then the courses never stopped and nor did their learning in machine learning. It took time to learn the statistics and mathematics side of data science but they never gave up. The team leader Ghosh later co-founded a data science club in his college and named it Neuron, which has been growing exponentially since its inauguration. They participated in Kaggle competitions where we had to predict poverty levels in Costa Rica, and several other online and offline competitions, which helped them to stay in touch with all new innovations in the data science world. Approach To Solving The Hackathon Problem Visualisation: The team generally solved a problem starting off with visualisation. They visualised each and every column and also combined separate columns in order to get a better understanding of the data. Data Cleaning: On receiving the data, they had to clean the data to remove all the NaN\/null values, which, according to the team, took the most time. Filling up the NaNs took a lot of Domain Knowledge about how insurance policies in the US work and separate parts of an insurance policy. They also took help from visualisations in Tableau and Plotly in cleaning the data. They then added some new features which included maximum and minimum expenditure in a year of a policyholder. Modelling: After five complete days of cleaning the data and feature engineering, they finally moved on to modelling. Soon they were met with their first hurdle, which was the difference in key categorical columns between the test and the train. They had to divide them first and train them separately. “No one has ever won a competition without ensembling or blending,” said Ghosh. Their first aim here was to find the best single model. They started with residual boosted trees like XGBoost and LightGBM as they generally tend to outperform all other models in competitions, but unfortunately they found that it was not the case here and so they moved on to more baseline primitive models and finally random forests, with its key hyperparameters tuned, proved to be the best performing single model. This took them to the top 3 in the leaderboard. However, their best single model with a high train score and a relatively low cv score was overfitting. They needed better-regularised models with a better validation score to go up the leaderboard. This was when they took up the “stacking” approach to modelling, which is quite similar to ensembling, the only difference being that stacking takes a ranked weighted average of all models. Stacking is somewhat similar to Neural Networks, which comprises of different layers of stacked models. 3 layers of stacked models, with six separate models, is what lead the team to victory. Modelling was quite an iterative process where they kept going back to feature engineering and feature selection to improve the model. MachineHack Experience Talking about their experience on MachineHack platform, they said that it was overall a great learning experience. Competing with top contenders, especially professional data scientists gave the students and aspiring data scientists a chance to grow and see where they stand in terms of the industry standards. “Machine Hack is a great platform where companies from around India are organising challenging hackathons that provide an amazing competitive environment to professionals and data science enthusiast,” Team Neuron said. They also said that they will keep enrolling in hackathons on MachineHack and are looking forward to more hackathons on the platform. “One thing we have learnt so far in our data science journey is that Data Science is not just about making complicated models, it’s about taking better decisions,” said the team.","excerpt":"TEG Analytics and Analytics India Magazine had organised a hackathon called Predict Market Competitiveness For Insurance Products, this year on the Republic Day. The problem involved predicting the market share of insurance companies in the US, affiliated to Medicare. After an offline hackathon, a set of finalist teams were shortlisted to present their case study […]","categories":["Deep Tech"],"tags":["Data Cleaning","Gen AI in Insurance","insurance","Kaggle","Machine Learning","modelling","XGBoost"],"author_name":"Disha Misal","publish_date":"2019-02-14T07:59:21","publication_year":"2019","word_count":798,"keywords":["data science","insurance","Kaggle","modelling","AI","machine learning","neural network","ML","Plotly","Machine Learning","Aim","XGBoost","Data Cleaning","Gen AI in Insurance","analytics","LightGBM"],"extracted_tech_keywords":["AI","machine learning","ML","neural network","data science","analytics","Aim","XGBoost","LightGBM","Plotly"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/how-these-data-science-enthusiasts-from-christ-university-solved-our-insurance-products-hackathon\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10040749,"title":"Ransomware Everywhere: Dire State Of Cybersecurity In 2021","content":"Hackers are on ransom-spree. From fuel pipelines in the west to stock exchanges in the east, they are making their presence felt everywhere.  Last week, one of the largest fuel supply companies— Colonial pipeline which spans over 8,000 kms  was shut down after a breach. The company had to pay the hackers handsomely to resume their operations. Colonial Pipeline’s CEO admitted that his organisation paid the group of hackers a $4.4 million ransom as the company executives were unable to assess how badly their networks had been hacked or how long it would take to repair the pipeline. The particular case has come out in the open, many go unnoticed – thereby requiring an urgent fix. Image Credits: CSW Ransomware viruses are encoded in a file sent to a targeted user.  The hackers then try to do a trade-off by giving access back for some money. The ransomware attacks have increased by almost 37% during the pandemic. Moreover, Common Vulnerabilities & Exposure (CVE) saw a phenomenal jump of over 356% from just 57 in 2019 to 260 in the first quarter of 2021, as per Ransomware Spotlight Report 2021. Major cyber-attacks in India May 2021 – The Air-India data breach of more than 4.5 million passengers after a sophisticated cyber-attack on SITA – the Switzerland based company providing passenger services system. The attack was carried out on its servers based in the US.March 2021 – Ransomware attack on Pimpri-Chinchwad Municipal Corporation, Smart City project in Pune district, managed by Tech Mahindra.In October 2020 – Haldiram’s popular food major faced the ransomware attack, and attackers demanded $7,50,000 for access.November 2020 – Indian Computer Emergency Response Team (CERT-In) issued a warning against the spread of ransomware virus ‘Egregor’ capable of stealing vital corporate data. Several other ransomware attacks, including WannaCry, Petya, Mirai Botnet and Pegasus, have impacted private and public organisations on an immense scale. “If we go by the “Pegasus attack” by the NSO group of Israel, which was a topic of hot debate among large sections of our society, your phone can be compromised with even a missed call which is a shocking but cruel reality’, said Pukhraj Singh, Cyber Intelligence Analyst. “There is baseband software in our mobile phones. It acts as an interface between the hardware and software. Whatever communication we receive on our phones is converted into data by the interface. The moment this interface is compromised, you are hacked. It’s extremely challenging to avoid such hacks.” This is why organisations are seeking state-of-the-art solutions like machine learning and AI to thwart these attacks. Companies like DarkTrace claim that their algorithms can detect the threats in real time. What can AI do Image Credits: Darktrace Cyber AI from Darktrace is capable of neutralising ransomware without relying on rules or signatures. The tool is capable of identifying even the most sophisticated strains of ransomwares, while giving a response within seconds. It works by studying the organization’s ‘patterns of existence,’ which include people, machines, and servers, and detecting ransomware attacks as soon as they deviate from the standard. Image Credits: Research Paper – AI-Powered Ransomware Detection Framework Researchers Subash Poudyal and Dipankar Dasgupta from the Department of Computer Science, University of Memphis, have proposed an AI-powered ransomware detection framework. They have designed a ransomware analysis tool – AIRaD (AI-Powered Ransomware Detection), using the techniques of reverse engineering, static and dynamic analysis, and machine learning. The researchers are in the process of development, and they are planning to make it an open-source tool. Whereas, Microsoft’s Azure Sentinel uses AI to detect threats and respond against such attacks. It collects data both on-premises and on clouds to detect unknown threats using analytics and to hunt down suspicious activities with years of experience in the cyber-security domain. Image Credit: Microsoft Azure SpinOne is another tool, which uses machine learning-enabled ransomware security methods along with backups. The algorithms look for anomalies in file activity and spot ransomware encryption patterns. Once confirmed, they block the source of the attack and revoke user account access to prevent any further encryption. Image Credits: SpinOne The Information Technology (IT) Act 2000 with 94 sections was last amended in 2008 and now has 124 sections. Some of the common sections which affect us daily include Section-65 deals with knowingly or intentionally tampering with Computer source documents; Section-66 Deals with the hacking of computer systems; 66-E for violation of privacy etc. Whenever we are interacting with cyberspace, we are only interacting in 0 and 1. The problem here is how laws that are not recognised by cyberspace can regulate it. “Code is law,” i.e. if we want to control cyberspace, you can control it only through the code . Additionally, the servers need to be installed within the country, rather than storing and using it outside. Data localisation will make it easy for the governments and cybersecurity personnel to access data in real-time for monitoring and quick response. Furthermore, we need to put more emphasis on the Public-Private Partnership (PPP) model, where Indian IT industries can work in tandem along with the government sector by exchanging best practices available in the cyber field. More funds for R&D in the field of cybersecurity will hold the key for the future.","excerpt":"Hackers are on ransom-spree. From fuel pipelines in the west to stock exchanges in the east, they are making their presence felt everywhere.  Last week, one of the largest fuel supply companies— Colonial pipeline which spans over 8,000 kms  was shut down after a breach. The company had to pay the hackers handsomely to resume […]","categories":["AI Features"],"tags":[],"author_name":"kumar Gandharv","publish_date":"2021-05-26T11:00:00","publication_year":"2021","word_count":870,"keywords":["Go","machine learning","ELT","AWS","AI","R","Aim","analytics","GAN","Azure"],"extracted_tech_keywords":["AI","machine learning","analytics","Aim","AWS","Azure","R","Go","ELT","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/ransomware-everywhere-dire-state-of-cybersecurity-in-2021\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10069420,"title":"Refuelling India with bigger, better convenience store: 7-Eleven","content":"Last month, global convenience store giant 7-Eleven, launched its global solution centre (GSC) in Bengaluru. This is the company’s first such centre outside the US. Analytics India Magazine caught up with the VP and country leader, global solution centre – India, 7-Eleven, Jennifer Goschke, at their global solution centre in Bengaluru, where she spoke in length about their future plans in the country, alongside her vision for the company, work culture, and more. Goschke’s obsession with Slurpee and Big Gulp – a brand name for carbonated slushies sold by 7-Eleven – is on another level. The pop-art of Slurpee (designed by Gensler) behind her work desk explains that. She said that her journey with 7-Eleven has been very personal. “I grew up with 7-Eleven. I used to ride my bicycle to 7-Eleven every day when I was in 7th and 8th grade to get a Slurpee. And then, once I got my driver’s licence, I would drive to 7-Eleven, sometimes twice a day, to get a Big Gulp of Coca-Cola — because it is the best!” said Goschke. Now, to be working for the company, after having grown up with the brand, is really just a dream come true, she added. She said that 7-Eleven has always been about serving customers what they want, when and where they want it. “One of our leadership values is customer obsession. I am dedicated to leading the GSC and following our mission to act as a trusted resource for 7-Eleven, providing capability and dedicated expertise to deliver high-quality, innovative services,” said Goschke. Before joining 7-Eleven in 2018, Goschke has worked in senior leadership roles in companies like Walmart, K-Mart\/Sears, and Office Depot, among others. Leadership challenges “Recruiting talent in a fiercely competitive job market to a brand that is just getting started in the region has proven to be my biggest leadership challenge,” said Goschke. She has been tasked with bringing this iconic brand to life in Bengaluru, living and modelling their values and leadership principles daily to provide a positive 7-Eleven brand experience for employees. “I strive to ensure my team knows I care about them personally and professionally, and I am here to help them be successful, make their lives better, remove roadblocks and invest in their development,” said Goschke. 7-Eleven has over 81,000 stores in 18 countries and regions. In India, the company is working with master franchisee Reliance Retail to modernise the small-retail environment and bring greater convenience to shoppers. In the coming months, the company looks to expand aggressively in Mumbai and other cities. The convenience store giant competes with various retail companies, including Target, Tesco, Metro Cash & Carry, Walmart and others. The question is, how are they different from other players in the market? To this, Goschke said: Although 7-Eleven is a nearly 95-year-old company, it is newer to India compared to other retailers who have been here for a decade or more. Some might consider the GSC the world’s oldest startup, backed by almost a century of history but with the energy of a new business. In 1927, 7-Eleven invented the convenience store – and they have been doing it every day since becoming the premier name in the convenience-retailing industry with more stores than any other convenience store chain. The company believes that it has remained laser-focused on innovating to meet customers’ needs – anytime, anywhere. 7-Eleven store in Mumbai Why Bengaluru? Goschke said that Bengaluru was a natural choice, as it is India’s technology hub, and access to the right talent allows 7-Eleven to elevate its digital innovations. Currently, the GSC has more than 170 employees and is hiring for IT, information security, finance, and accounting roles, and plans to double it next year. “We have recently started construction work on additional office space to accommodate the expected growth of our team,” shared Goschke. Further, she said that the GSC plays a key role in 7-Eleven’s strategy by developing products that support business activities at stores. Also, the GSC accelerates digital strategy and expands internal IT capabilities in information security, networking and data science. Team 7-Eleven at the inauguration of its new global solution centre (GSC) in Bengaluru Team structure Goschke said the GSC is an important resource for their Dallas-Fort Worth-based core technology team. “Our India and US offices complement each other and support innovation to enhance our services,” she added. Like the US, 7-Eleven’s Bengaluru GSC is structured in two main categories – Center Enablement and Service Delivery. Center Enablement includes the roles needed to support the operations of the GSC, such as HR, finance and facilities. Service Delivery includes the functional teams that provide services to the US, including IT, information security, finance and accounting. 7-Eleven’s tech innovations 7-Eleven has migrated all its applications and processes to the cloud and leverages AI\/ML capabilities in both the US and India centres to provide the best customer experience. The team said its entire tech stack is hosted on the cloud. They follow a multi-cloud hybrid approach, where they work with multiple cloud providers in their setup. Here are some of the analytics use cases at 7-Eleven: Ordering AI: Use of algorithms that power their supply chain. Scheduling AI framework: Using sales forecast to help determine the appropriate labour hours\/scheduling needed in the stores. Personalisation: Multiple ML models built in-house to help them develop personalised offers for their customers in their 7Now App. Work culture Goschke said that the GSC is a close-knit group of people who are excited to learn, eager to innovate and passionate about making a meaningful contribution to the company. When asked about collaboration and partnerships, Goschke said that she looks for integrity because, without it, everything else crumbles. “I also look for energy – someone who has a passion for what they are doing. High-energy people electrify the workplace. Another trait I look for in a candidate is excellent collaboration skills,” she added. Exciting times ahead “7-Eleven’s mission is: ‘Be the first choice for convenience. Anytime. Anywhere,’” said Goschke. She said that at the GSC, they strive to be a one-stop shop for their teams in Dallas and collaborate closely with them to complement and deliver on the 7-Eleven brand’s value proposition. “We are eager, willing and ready to innovate to meet customers’ needs and make their day a little more awesome by delivering fast, personalised convenience – when, where and how they want it,” she added, saying that they look forward to GSC’s continued growth.","excerpt":"7-Eleven has over 81,000 stores in 18 countries and regions. In India, the company is working with master franchisee Reliance Retail to modernise the small-retail environment and bring greater convenience to shoppers.","categories":["IT Services"],"tags":[],"author_name":"Amit Naik","publish_date":"2022-06-20T17:00:00","publication_year":"2022","word_count":1078,"keywords":["data science","Go","AI","ML","Git","RAG","ViT","analytics","Rust","R"],"extracted_tech_keywords":["AI","ML","data science","analytics","RAG","R","Go","Rust","Git","ViT"],"url":"https:\/\/analyticsindiamag.com\/it-services\/refuelling-india-with-bigger-better-convenience-store-7-eleven\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":27471,"title":"pi Ventures Raises $6 Million In Funding, Plans To Use It For Deep Tech Innovations","content":"pi venture Partners (L-R): Umakant Soni, Manish Singhal and Abishek Surendran pi Ventures, India’s first Applied Artificial Intelligence and IoT focused early stage venture fund has been backed by CDC Group with $6 million, which will be used by the company to invest in startups with deep tech capabilities that have the potential to disrupt sectors and facilitate a lasting change in businesses and lives. CDC Group is the UK government’s development finance institution that aims to increase capital flows to underdeveloped markets so countries can finance their way out of poverty. Their focus is to find talented teams and cohesively grow great businesses in Africa and South Asia. CDC’s unique structure combines the best of private sector skills with a public-spirited mission. Their focus sectors to invest include financial services, infrastructure, health, manufacturing, food and agriculture, construction and real estate, and education, typically that which will lead to the economic growth of a country. Alagappan Murugappan, Managing Director, Head of Intermediated Equity, South Asia, CDC Group said, “We are pleased to be supporting pi Ventures for its AI-focused fund. This allows our capital to support early-stage companies that are innovating through technology to provide affordable goods and services at scale in areas such as healthcare among others.” pi Ventures announced their second close in January this year at $25 million and the fund is expected to announce their final close shortly. Key contributors to the fund include Chairman of Hero Enterprise Sunil Kant Munjal, Electronic Development Fund (managed by Canbank Ventures), SIDBI, prominent family offices from USA, Canada, Singapore & India and leading entrepreneurs like Mohandas Pai, Binny Bansal, Deep Kalra, Sanjeev Bikchandani and Bhupen Shah among others. The fund is co-sponsored by In Color Capital of Canada. Manish Singhal, Founding Partner, pi Ventures added, “CDC’s confidence fuels our vision to find, fund and support the best teams who are using Artificial Intelligence to solve real world problems. We are very happy to have their support and backing”. pi Ventures has announced six investments so far — SigTuple, NIRAMAI, ten3T, Zenatix, CustomerSuccessBox and Locus.sh. The 70-year old CDC Group has been investing in India for over 30 years and has already made 595 investments across sectors.","excerpt":"pi Ventures, India’s first Applied Artificial Intelligence and IoT focused early stage venture fund has been backed by CDC Group with $6 million, which will be used by the company to invest in startups with deep tech capabilities that have the potential to disrupt sectors and facilitate a lasting change in businesses and lives. CDC […]","categories":["AI News"],"tags":[],"author_name":"Srishti Deoras","publish_date":"2018-08-21T06:58:14","publication_year":"2018","word_count":367,"keywords":["Go","API","artificial intelligence","programming_languages:R","AI","programming_languages:Go","Aim","R","startup"],"extracted_tech_keywords":["AI","artificial intelligence","Aim","R","Go","API","startup","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/pi-ventures-raises-6-million-in-funding-plans-to-use-it-for-deep-tech-innovations\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":48946,"title":"Is India Ready To Deploy The Biggest-Ever Facial Recognition System","content":"India is set to witness the biggest ever deployment of facial recognition system next month. Come 8th of November, India will announce the winner of the bids for the National Automated Facial Recognition System (NAFRS) that will lead India towards executing its plan to install automatic facial recognition. It will not only enhance security across the country but help in capturing criminals, identify lost individuals, dead bodies and more. With this move, the government has made another significant stride in improving security across the country. While installing CCTV cameras for enhanced security was effectual but it included manual force to monitor and identify incidents. With automated facial recognition, the government will streamline such tedious processes to make instant decisions for arresting criminals. What is Government Expecting from Bidders With this system installed, the videos from CCTV cameras can be utilised to match information from the national crime record bureau and notify in case of a match. It would mean that cops would be able to find criminals easily without performing any manual search. The government is also expecting that the automatic facial recognition system should have a feature to upload the images of criminals from newspapers or pictures sent by the public in the criminal database. The facial recognition platform will be integrated with other governmental departments to embrace images collected by them, thereby, simplifying the process for police to search across different databases. Is India ready for it? As of today, there is no regulation in place to keep a tab on the data for avoiding the unlawful use of facial information. The government has not yet disclosed the roadmap for the accessibility of facial data. It, therefore, raised doubts among citizens about the potential implications of the technology. Safeguarding data is of paramount importance as the slightest of negligence can lead to a significant privacy glitch. Although the deployment of facial recognition will bring convenience to both the government and people alike, citizens anticipate reliable data protection. Therefore, people are dubious about the successful implementation of this initiative. However, India has already successfully deployed facial recognition at airports for security check-ins, and now with the integration of NAFRS, the government is trying to make it ubiquitous in India. Citizens’ View Facial recognition is far from being flawless as there have been numerous instances where the technology could not recognise the faces correctly or even failed to identify 3D facial structures. They are many instances where facial recognition technology has failed. Mismatches are very common with this technology. Thus many governments across the globe have rejected the idea of its integration. While a few people think that it is an encroachment of freedom, others consider it a crucial initiative for striking down criminals for a safer place. But, with the recent news about the misuse of facial recognition technology by Hong Kong police to quell protestors, citizens are critical about the potential abuse of technology. Outlook Embracing the technology for purveying public services is essential but at the same time, ensuring its appropriate use is pivotal for any government to gain the trust of its people. It can be done by bringing transparency in the system. The use case of facial recognization is limitless, but accuracy has to be the prime factor for deploying it. Building the world’s biggest facial recognition system for the sake of records will not serve the purpose; thereby, the government must communicate its intention and plans clearly with its citizens. The government needs to step up and explain their approach. It should also assure its citizens about the use cases to deliver public services effectively. Moreover, the government should bring a stringent law that eliminates the potential exploitation of personal data of its citizens.","excerpt":"India is set to witness the biggest ever deployment of facial recognition system next month. Come 8th of November, India will announce the winner of the bids for the National Automated Facial Recognition System (NAFRS) that will lead India towards executing its plan to install automatic facial recognition. It will not only enhance security across […]","categories":["AI Features"],"tags":["facial recognition India"],"author_name":"Rohit Yadav","publish_date":"2019-10-29T14:14:03","publication_year":"2019","word_count":620,"keywords":["Go","programming_languages:R","AI","ML","programming_languages:Go","facial recognition India","Rust","R","programming_languages:Rust"],"extracted_tech_keywords":["AI","ML","R","Go","Rust","programming_languages:R","programming_languages:Go","programming_languages:Rust"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/is-india-ready-to-deploy-the-biggest-ever-facial-recognition-system\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":1528,"title":"5Barz India transforms Smart Home market with the launch of its Home IoT Hub","content":"5BARz India introduces ROVR Wi-Fi router and smart IoT hub 5BARz India Private Limited, a 5BARz International’s wholly owned subsidiary announced its plan to launch an advanced next generation ROVR Wi-Fi router and smart IoT hub at a recent event held at its Bangalore office. The technology firm specializes in the cellular network extender industry and broadband networks. Mr. Narayanan Sadagopan, Director, IIIT Bangalore, as well as Mr. Vasan Paulraj, Managing Director & Head – Financial Sponsors Group, Axis Capital Limited were some of the notable attendees at the event. The previous year marks the successful commercialization of 5BARz’ Network extender in India, and the firm continues to lead the market with breakthrough technology. The newest addition of the Wi-Fi product combines the very best of creativity and technological ingenuity, aimed at delivering the best Wi-Fi user experience for the customers. Daniel Bland, Founder & CEO, 5BARz International, extols “The launch of our ROVR Wi-Fi in the Indian markets marks a significant step for us, and we are really excited. This launch represents the second product line we unveiled for Indian market, proving as major milestone for 5BARz.” The product will be soon rolled out for Indian market, with an underlying objective of connecting 5 million Indian homes to the Internet over the next 5 years. As a part of this launch, 5BARz will be collaborating with an undisclosed leading Internet Service Provider (ISP). Additionally, the product will be made available on Amazon, Flipkart, and Snapdeal and through major retail chains. “I would like to applaud the team on their continued excellence in generating leading edge technology solutions for the Indian market and beyond. This indeed marks a special day for the entire team.” Few unique features of the ROVR: The device always remains on with a 4 hours’ battery backup Includes dual Backhaul (Wire line + Wireless backhaul) The device features Parental control and Usage control Combine home safety and security features Furthermore, the device combines mini Wi-Fi instances for each member, remote monitoring, alerts, and advanced IoT capabilities. With the easy-to-use 5BARz™ Smart Experience connectivity suite of applications, monitoring and customization can be done anywhere and at any time. The device will manage and control temperature, humidity, gas detection, motion sensors, and cameras, to alert the home owners with a call or SMS when the situation arises.","excerpt":"5BARz India Private Limited, a 5BARz International’s wholly owned subsidiary announced its plan to launch an advanced next generation ROVR Wi-Fi router and smart IoT hub at a recent event held at its Bangalore office. The technology firm specializes in the cellular network extender industry and broadband networks. Mr. Narayanan Sadagopan, Director, IIIT Bangalore, as […]","categories":["AI News"],"tags":["IoT India"],"author_name":"Дарья","publish_date":"2017-02-24T07:03:20","publication_year":"2017","word_count":389,"keywords":["Go","API","programming_languages:R","AI","IoT India","programming_languages:Go","Ray","Aim","ViT","R"],"extracted_tech_keywords":["AI","Aim","Ray","R","Go","API","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/5barz-india-transforms-smart-home-market-launch-home-iot-hub\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10053799,"title":"Why 2021 Proved To Be A Significant Year For AI-Based Digital Healthcare","content":"At the recently concluded NVIDIA GTC 2021 fall event, CEO Jensen Huang spoke extensively about the company’s efforts and innovation in the digital health care sector. While healthcare has always been an important and exciting field for AI and digital tech-based innovation, the pandemic has shone a better light on it. In this article, we discuss some important innovations and announcements from big tech companies in 2021. NVIDIA’s Foray Huang, in his keynote speech at the GTC 2021 event, said that business models would undergo revolution and Medical Device Software-as-a-Service solutions will replace medical instrument sales. He added, “The industry needs a software-defined imaging platform to build this future on, just as the auto industry needed a software-defined AV platform. Today we’re announcing NVIDIA Clara Holoscan, a software, programmable, imaging platform.” The newly introduced Clara Holoscan is an AI computing platform for medical devices. It combines hardware systems for low-latency sensor and network connectivity, libraries for data processing and AI with core microservices. It can run streaming, imaging and other applications. The platform has been designed to speed up key workflow phases — high-speed I\/O, image processing, data processing, physics processing, and rendering. These applications can be deployed fully in instruments, in the hospital data centre, or hybrid. Over the years, NVIDIA has made major inroads in the healthcare industry. For instance, in April this year, the company collaborated with AstraZeneca and the University of Florida for drug discovery and patient care AI research projects. In addition, in 2018, NVIDIA had launched an AI-based virtual medical imaging platform called Clara. Alphabet’s Isomorphic Lab Google’s parent company, Alphabet, recently launched Isomorphic Labs. This company, dedicated to drug discovery, will be headed by DeepMind founder and CEO Demis Hassabis. The company’s launch comes on the heels of the AlphaFold 2 release. Touted as a major breakthrough in biology and medical research, AlphaFold from DeepMind solved the 50-year-old grand challenge of protein folding by predicting its 3D structure accurately right up to atom-level from its amino acid sequence. Hassabi wrote in the launch blog that this was the right time to dedicate focus and resources to biological and medical research, something that Isomorphic Labs endeavours to do. Isomorphic Labs is a commercial venture that reimagines the drug discovery process from the AI-first perspective — to model and understand fundamental mechanisms of life. DeepMind & AlphaFold In July, Alphabet-owned research lab DeepMind made the AlphaFold 2.0 source code public. DeepMind hopes to offer easy access and better research opportunities in areas such as drug discovery. The protein folding problem has been a major challenge among the research community. DeepMind first began working on it in 2016. It took about two years to release AlphaFold 1.0. While it was novel, it was good enough for carrying out any groundbreaking research. In 2020, the research lab released AlphaFold 2.0. It is an upgraded version composed of a subnetworks system integrated into a single differentiable model based on pattern recognition. A few months later, DeepMind introduced AlphaFold Multimeter, a model to predict the structure of multi-chain protein complexes. It increases the accuracy of predicted multimeric interfaces and maintains high intra-chain accuracy. Microsoft-Nuance Deal Over the years, Microsoft has invested heavily in digital healthcare, and 2021 was no different. In one of the year’s biggest deals, Microsoft acquired Nuance, an AI platform for speech recognition at a staggering $16 billion. It is the second-largest deal in Microsoft’s history. Nuance’s tech powers Apple’s virtual assistant Siri and makes software for other healthcare and automotive sectors. Nuance’s major clinical speech recognition products include Dragon Ambient eXperience, Dragon Medical One, and PowerScribe One. All these SaaS offerings are built on Microsoft Azure. Another major development in Microsoft’s digital health push came with its collaboration with AXA. The two companies have joined hands to build a digital healthcare platform that is open to all. This new platform will be built on the technology portfolio in Microsoft Cloud for Healthcare to derive patient insights without compromising patient privacy.","excerpt":"In this article, we discuss some important innovations and announcements from big tech companies in 2021.","categories":["IT Services"],"tags":[],"author_name":"Shraddha Goled","publish_date":"2021-11-19T13:00:00","publication_year":"2021","word_count":665,"keywords":["Go","AI","R","Git","microservices","RAG","Ray","ViT","GAN","Azure"],"extracted_tech_keywords":["AI","Ray","RAG","Azure","microservices","R","Go","Git","GAN","ViT"],"url":"https:\/\/analyticsindiamag.com\/it-services\/why-2021-proved-to-be-a-significant-year-for-ai-based-digital-healthcare\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10065466,"title":"Does DALL-E pose a threat to designer jobs?","content":"DALL-E 2, the latest iteration of OpenAI’s AI system that creates images from textual prompts in natural language, has broken the internet, and for the right reasons. To showcase the system’s potential, OpenAI CEO Sam Altman solicited ideas from the public: Imaginative prompts such as ‘a shark and a dolphin cruise hand-in-hand with an undersea city in the background’ or ‘a rabbit detective sitting on a park bench and reading a newspaper in a victorian setting’ came up as suggestions. The results have made people wonder if DALL.E, or similar AI tools, could replace human designers. pic.twitter.com\/9qg3hpNu08— Sam Altman (@sama) April 6, 2022 DALL·E 2 learns the relationship between images and the text prompts. The process starts with a pattern of random dots and builds towards the final image using a process called diffusion. The new iteration also comes with an editing option. DALL.E 2 can respond to natural language caption to add and remove elements while taking shadows, reflections, and textures into account. DALL-E 2 generates more realistic and accurate images with 4x greater resolution than DALL·E. https:\/\/twitter.com\/Flomerboy\/status\/1516453647915225088?s=20&t=1htsDjYhK4kP4HWlQaN2yQ How credible is the threat Alex Nichol, one of the DALL-E researchers, said tools like DALL.E 2 democratise the designing process. But does it put designer jobs at risk? What exactly is the collateral damage of this technology? “Questions which human beings are asking are, is my job viable? Is my profession at risk? And unfortunately, that’s just a different set of questions that the businesses are asking- can we do something faster, cheaper, more automated?” said George Baily, product marketing manager at FintechOS. DALL-E generated images can be a viable substitute for expensive stock photos. The major factors at play here are speed and scale. “As an individual wanting a series of interesting illustrations in a particular style or having ten ideas for a new logo or a quick banner ad for my social media campaign, I don’t want to have an emotional discussion with my designer about their creativity. It doesn’t need to be perfect, it just needs to be. These images can be created at a hyper-speed, scale and quantity that humans are just not built to process,” he said. Animal helicopter chimeras generated with DALL·E 2: pic.twitter.com\/5b8a9iq3k9— Aditya Ramesh (@model_mechanic) April 7, 2022 Gary Marcus, scientist and the author of Rebooting AI, said: “DALL-E is probably best used as a source of inspiration rather than a tool for final products. You can say something like “a boat on the sea, in a Van Gogh style”, and get something beautiful. But if you want to change the end product, perhaps to “a boat on the sea but with five people rather than 4, with the tallest person in the front and the shorter person in the back, with the same boat, but painted brown, and a slightly darker background”, the system probably won’t understand the language well enough to meet your exact specifications.” Emerging picture of DALL-E2:• Produces fabulous, high quality graphics• Easy to get something in general ballpark of what is desired• Difficult to fine-tune output to specific needs• Understanding of language is very superficial• Fails tests of compositionality https:\/\/t.co\/MocJv1QJ39— Gary Marcus (@GaryMarcus) April 12, 2022 The trouble with DALL.E 2 Gary Marcus has also taken to Twitter to highlight the model’s drawbacks, such as the difficulty to fine-tune output to specific needs; superficial understanding of language; and compositional flaws. Mathematician Jeremy Kahn said the system often fails to render details like lighting or shadow in complex scenes: DALL.E is also not good at merging borders or understanding binding attributes. Alex Nichol discussed a few examples: He asked the tool to ‘put the Eiffel Tower on the moon’, and it put the moon in the sky above the tower. Then, he gave a ‘living room filled with sand’ prompt, and it generated a scene that resembled a construction site than a living room. CNN’s Rachel Metz also posted a few DALL.E 2 fails on Twitter. AI: good at generating penguins, bad at counting pic.twitter.com\/ghLrTDBEEm— Rachel Metz (@rachelmetz) April 11, 2022 The AI system also throws the idea of art into sharp relief. “What makes art valuable is that artists have an opportunity cost of doing that art. It’s a sacrifice of something else in their life. AI art gets inflated instantly. Everyone can get a beautiful painting with a button click. Not valuable,” said a commenter in a hacker news thread. According to OpenAI, DALL-E shows how imaginative humans and clever systems can work together to make new things, amplifying our creative potential. We can safely conclude that DALL.E is no Picasso, but the AI system can create spellbinding conceptual art with the right input and combinatorial play.","excerpt":"The images can be created at a hyper-speed, scale and quantity that humans are just not built to process.","categories":["IT Services"],"tags":["AI Tool"],"author_name":"Avi Gopani","publish_date":"2022-04-22T11:00:00","publication_year":"2022","word_count":781,"keywords":["Go","DALL-E","TPU","AWS","OpenAI","AI","CNN","ViT","AI art","AI Tool","R"],"extracted_tech_keywords":["AI","OpenAI","AWS","TPU","R","Go","DALL-E","CNN","ViT","AI art"],"url":"https:\/\/analyticsindiamag.com\/it-services\/does-dall-e-pose-a-threat-to-designer-jobs\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10092304,"title":"Pune-based Persistent Systems Crosses USD 1 Billion in Annual Revenue","content":"Pune-based mid-tier IT company, Persistent Systems Ltd, has reported a consolidated net profit of Rs. 251.50 crore in their fourth-quarter ended March 2023, an increase of 5.7% sequentially from Rs. 237.90 crore. The company also posted a consolidated revenue from operations for the quarter ended at Rs. 2,254.5 crore, up 4% on a quarter-on-quarter basis from Rs. 2,169.40 crore. The company’s net profit increased by 35% year-on-year to Rs. 237.90 crore in the previous quarter. Sandeep Kalra, Chief Executive Officer and Executive Director at Persistent acknowledged that the current fiscal year of 2023 was a “momentous” one for the company as it reached several key milestones like attaining USD 1 billion in annual revenue and saw its inclusion in three key indices of the National Stock Exchange of India including the Nifty IT index. “We have been nimble, proactive, and disciplined, allowing us to build a healthy booking pipeline and maintain competitive advantage. We’re truly grateful to our clients, partners, investors, and team members for their unwavering trust. As we move to the next phase of growth, we will continue to strengthen our partner ecosystem, maintain operational rigor, and deepen our capabilities to scale our Digital Engineering expertise and drive business value for our clients,” he stated. The board has recommended a final dividend of Rs. 12 per equity share and a special dividend of Rs. 10 per equity share of Rs. 10 each for FY 2022-23, according to a company filing. The special dividend, which will be given together with the final payout, is recommended for reaching $1 billion in annual revenue. The company also said in the filing that the order booking for the quarter under review was at USD 421.60 million in annual contract value terms.","excerpt":"The company’s net profit increased by 35% year-on-year to Rs. 237.9 crore in the previous quarter.","categories":["AI News"],"tags":[],"author_name":"Poulomi Chatterjee","publish_date":"2023-04-25T15:35:00","publication_year":"2023","word_count":289,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","Git","Rust","R","programming_languages:Rust"],"extracted_tech_keywords":["AI","R","Go","Rust","Git","programming_languages:R","programming_languages:Go","programming_languages:Rust"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/pune-based-persistent-systems-crosses-usd-1-billion-in-annual-revenue\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":69485,"title":"The Limits of Transfer Learning In Natural Language Processing","content":"It has become increasingly common to pre-train models to develop general-purpose abilities and knowledge that can then be “transferred” to downstream tasks. In applications of transfer learning to computer vision, pre-training is typically done via supervised learning on a large labelled dataset like ImageNet. In contrast, modern techniques for transfer learning in NLP often pre-train using unsupervised learning on unlabeled data. In spite of being widely popular there are still few pressing questions bothering transfer learning in ML: How much of the original task has the model forgotten? Why don’t large models change as much as small models? Can we make more out of pre-trained weight statistics? Are the results similar to other tasks, such as segmentation? The rapid rate of progress and diversity of techniques can make it difficult to compare different algorithms and understand the space of existing methods for transfer learning. The researchers at Google say that there is a need for a unified approach to understanding the effectiveness of transfer learning. To formulate a unique, unified approach, the researchers treated every NLP problem as a “text-to-text” problem, i.e. taking text as input and producing new text as output. The text-to-text framework offers flexibility that helps in evaluating performance on a wide variety of English-based NLP problems, including question answering, document summarization, and sentiment classification, to name a few. With this unified approach, wrote the authors, we can compare the effectiveness of different transfer learning objectives, unlabeled datasets, and other factors, while exploring the limits of transfer learning for NLP by scaling up models and datasets beyond what has previously been considered. Testing Effectiveness of Transfer Learning In transfer learning, the neural network is trained in two stages: Pre-training: The network is generally trained on a large-scale benchmark dataset representing a wide range of categories Fine-tuning: Pre-trained network is further trained on the specific target task of interest, which may have fewer labelled examples than the pre-training dataset. For this study, all the experiments done were based on Transformer architecture, considering its wide applicability and adoption. The baseline model is designed so that the encoder and decoder are each similar in size and configuration to BERT. One of the examples for testing the model is shown above, where the words “for”, “inviting” and “last” are crossed (corrupted), which were chosen randomly. Each consecutive span of corrupted tokens is replaced by <X> and <Y>. Since “for” and “inviting” occur consecutively, they are replaced by <X>. The output sequence then consists of the dropped-out spans, delimited by the sentinel tokens used to replace them in the input plus a final sentinel token <Z>. The authors pre-trained their models for 1 million steps on a batch size of 2^11 sequences of length 512, corresponding to a total of about 1 trillion pre-training tokens. Pre-training on the RealNews-like, or Wikipedia + TBC datasets outperformed pre-training on C4 on a few downstream tasks. However, these dataset variants are sufficiently small that they would be repeated hundreds of times over the course of pre-training on 1 trillion tokens. The results show that additional pre-training can indeed be helpful and that both increasing the batch size and increasing the number of training steps can confer this benefit. The authors summarised their findings as follows: Finding The Right Training Strategies The paper states that updating all of a pre-trained model’s parameters during fine-tuning outperformed methods that are designed to update fewer parameters, although updating all parameters is expensive. In a multi-task setting, the authors couldn’t find a strategy that matched the performance of the basic approach of unsupervised pre-training, followed by supervised fine-tuning. However, they found that fine-tuning after pre-training on a mixture of tasks produced comparable performance to unsupervised pre-training. The authors also posit that larger models trained for longer might benefit from a larger proportion of unlabeled data because they are more likely to overfit to smaller training datasets. Architectures While some work on transfer learning for NLP has considered architectural variants of the Transformer, the  original encoder-decoder form worked best in the text-to-text framework. Though an encoder-decoder model uses twice as many parameters as “encoder-only” (e.g. BERT) or “decoder-only” (language model) architectures, it has a similar computational cost. This study also demonstrates that sharing the parameters in the encoder and decoder did not result in a substantial performance drop while reducing the total parameter count by 50 per cent. Datasets When comparing C4 to datasets that use additional filtering, the researchers found that training on in-domain unlabeled data could boost performance in a few downstream tasks. However, constraining to a single domain typically results in a smaller dataset. Performance can degrade when an unlabeled dataset is small enough that it is repeated many times over the course of pre-training. This hints at the preference for large and diverse datasets for generic language understanding tasks. Few other key conclusions of the work: Larger models tend to perform better Pre-training on unlabeled in domain data can improve performance on downstream tasks English-only pre-training did not achieve state-of-the-art results on the translation tasks Ensembling models that were fine-tuned from the same base pre-trained model performed worse than pre-training and fine-tuning all models completely","excerpt":"It has become increasingly common to pre-train models to develop general-purpose abilities and knowledge that can then be “transferred” to downstream tasks. In applications of transfer learning to computer vision, pre-training is typically done via supervised learning on a large labelled dataset like ImageNet. In contrast, modern techniques for transfer learning in NLP often pre-train […]","categories":["Deep Tech"],"tags":["NLP","Transfer Learning"],"author_name":"Ram Sagar","publish_date":"2020-07-12T13:00:20","publication_year":"2020","word_count":858,"keywords":["Go","API","TPU","AI","neural network","ML","computer vision","NLP","transformer architecture","Transfer Learning","R"],"extracted_tech_keywords":["AI","ML","neural network","NLP","computer vision","TPU","R","Go","API","transformer architecture"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/limits-of-transfer-learning-in-nlp\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":30646,"title":"Machine Learning 101: How To Kickstart Your Learning Journey","content":"When it comes to pursuing a career in machine learning, it is not as simple as learning and getting ahead in their careers. Beginners are often bewildered with countless learning resources out there. This guide tells exactly where to kickstart your proficiency towards ML. Step 1: Math And Statistics The first thing is the skill to understanding problems through a mathematical intuition. We would highly recommend starting with basics in linear algebra, then gradually moving to calculus. It may be hard to master them initially but given the time and practice with working, these areas will be familiar and comfortable to work on. Another subject that follows math is statistics. For any ML algorithm to be understood clearly, statistics is essential. Hence, fundamental knowledge in stats should be learnt hand in hand. By now, you might have easily thought of online websites such as Coursera or Udemy. Of course, these are good resources to learn at your own pace. The only hindrance is setting and following self-paced goals. On the other hand, here are a few resources that are equally good at teaching the basics of math and statistics. Linear Algebra tutorials by Kardi Teknomo – This interactive tutorial is a gem of a resource. Dr Kardi Teknomo of Ateneo de Manila University explains every concept in a simple, easy to read the language. Linear Algebra in Twenty Five Lectures – A concise course by the University of California, Davis scholars Tom Denton and Andrew Waldron. Paul’s Online Notes – A complete online resource (free to download) for math by Paul Dawkins of Lamar University. This resource is mainly centred around algebra and calculus. Stat Trek – Online website for statistics. Khan Academy – An all-time favourite among students, you would find a plethora of content on various areas of mathematics and statistics. Step 2. Programming For beginners, programming is sometimes dreadful to learn. Naturally, it might be intimidating and difficult at first. But with regular practice, this skill can be aced eventually. Coming to programming languages in ML, Python stands out top due to its versatility and ease in coding. Other languages such as R, Julia and Java also help in building ML projects. Ultimately, programming should complement your math and stat knowledge. Once mastered, try working on various problems and build small projects around ML. See how it solves a problem particularly by making use of math and statistics concepts. Sites like Kaggle and DataCamp are extremely good at testing codes and collaborating with peers and developers. In addition, forums like Stack Overflow are excellent to discuss problems and queries related to programming. These are the top resources that we found for learning Python, R and sites that cover programming in general. Pythonprogamming.net – One of the best online resources for learning Python out there. The programmer behind this website, Harrison Kinsley (popularly known as Sentdex in the Python community) explains every aspect of Python perfectly! Automate The Boring Stuff by Al Sweigart – Another very good online resource on Python. Programming is deconstructed right from scratch. In fact, the beauty lies in how simple tasks can be automated through Python. ListenData for R – This website offers R tutorials for free. Extensive coverage of R concepts is what makes this site a catch for ML beginners preferring R over Python. Code Project – A popular discussion forum exclusively for discussing programming queries in general. With a big developer user base, ML beginners can share their code, ask where they face problems in the code and work around ideas. Step 3. Application-Oriented Approach Learning is only fruitful if it is applied sensibly. Many people do the mistake of learning every algorithm in ML and forget where it actually helps in solving a problem. For beginners, it is suggested that they learn the popular and standard algorithms. A complicated algorithm is not always the solution for complex applications. It is all about how an ML problem is solved optimally. Here are a few blogs which brilliantly tells about applications in ML, which is suggested for every beginner to go through them for realising how ML is actually helpful. Machine Learning Mastery by Jason Brownlee – An amazing blog by expert Jason Brownlee. He explores the fascinating world of ML and captures its essence in the real world. Adam Geitey’s blog– interesting write-ups in ML and Python Arthur Juliani’s blog on Reinforcement Learning – an absolute gem of a blog which particularly focuses on reinforcement learning in ML. Edwin Chen’s blog – it explores requisite concepts of ML such as neural networks, deep learning etc. as well as the math behind it.","excerpt":"When it comes to pursuing a career in machine learning, it is not as simple as learning and getting ahead in their careers. Beginners are often bewildered with countless learning resources out there. This guide tells exactly where to kickstart your proficiency towards ML. Step 1: Math And Statistics The first thing is the skill […]","categories":["AI Trends"],"tags":["java project ideas","Machine Learning","Python Programming","r programming","Statistics"],"author_name":"Abhishek Sharma","publish_date":"2018-11-24T08:43:03","publication_year":"2018","word_count":772,"keywords":["Go","machine learning","Statistics","AI","neural network","ML","Machine Learning","Python Programming","RAG","Python","deep learning","r programming","java project ideas","R","Java"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","neural network","RAG","Python","R","Go","Java"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/machine-learning-101-how-to-kickstart-your-learning-journey\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10109937,"title":"Andrew Ng Releases New Course on Advanced Retrieval Techniques for AI Applications","content":"Andrew Ng’s DeepLearning.AI, in collaboration with Chroma, an open-source embedding database company focused on AI-native solutions, has introduced a free one-hour course designed to teach participants advanced retrieval techniques specifically tailored for AI applications. Called “Advanced Retrieval for AI with Chroma”, this short course is ideal for individuals with intermediate Python skills and a keen interest in mastering advanced retrieval techniques for extracting data from vector databases. Led by Chroma cofounder Anton Troynikov, the primary focus is on refining information retrieval processes to ensure that the output from a database query is not only semantically similar but also highly relevant to the query and its intended application. Leveraging an LLM enhances the effectiveness of this traditional technique, and the course explores another form of expansion where the LLM suggests a potential answer to the query, subsequently included in the query itself. Additionally, participants will also learn about cross-encoder reranking, a method to reorder retrieval results and prioritise those most relevant to the query, thereby improving overall results. Furthermore, the course covers the training and application of embedding adapters, introducing an adapter layer to reshape embeddings and improve elements pertinent to the specific application, leading to better retrieval outcomes. Ng, who has democratised AI education for all through his free courses had earlier released courses on various topics like generative AI for all, LLM quality and security, vector database for LLMs and more. He has participated with several companies like Microsoft, Lamini, AWS, OpenAI for these training materials.","excerpt":"Andrew Ng’s DeepLearning.AI, in collaboration with Chroma, an open-source embedding database company focused on AI-native solutions, has introduced a free one-hour course designed to teach participants advanced retrieval techniques specifically tailored for AI applications. Called “Advanced Retrieval for AI with Chroma”, this short course is ideal for individuals with intermediate Python skills and a keen […]","categories":["AI News"],"tags":["Courses"],"author_name":"Shritama Saha","publish_date":"2024-01-04T11:12:30","publication_year":"2024","word_count":247,"keywords":["TPU","OpenAI","AI","AWS","RAG","vector databases","Python","generative AI","Chroma","Courses","R"],"extracted_tech_keywords":["AI","generative AI","OpenAI","RAG","vector databases","Chroma","AWS","TPU","Python","R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/andrew-ng-releases-new-course-on-advanced-retrieval-techniques-for-ai-applications\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10052419,"title":"A Guide to Generalization and Regularization in Machine Learning","content":"Generalization and Regularization are two often terms that have the most significant role when you aim to build a robust machine learning model. The one-term refers to the model behaviour and another term is responsible for enhancing the model performance. In a straightforward way, it can be said that regularization helps the machine learning models for better generalization. In this post, we will cover each aspect of these terms and try to understand how these are linked to each other. The major points to be discussed in this article are outlined below. Table of Contents What is Generalization?Reasoning about the GeneralizationMeasuring GeneralizationWhat is Regularization?How Does Regularization Work?Different Regularization TechniquesChoosing the Right Regularization MethodRelating Generalization and Regularization Let’s start the discussion by understanding what generalization actually means. What is Generalization? The term ‘generalization’ refers to a model’s ability to adapt and react appropriately to previously unseen, fresh data chosen from the same distribution as the model’s initial input. In other words, generalization assesses a model’s ability to process new data and generate accurate predictions after being trained on a training set. A model’s ability to generalize is critical to its success. Over-training on training data will prevent a model from generalizing. In such cases, when new data is supplied, it will make inaccurate predictions. Even if the model is capable of making accurate predictions based on the training data set, it will be rendered ineffective. This is referred to as overfitting. The contrary is also true (underfitting), which occurs when a model is trained with insufficient data. Even given the training data, your model would fail to produce correct predictions if it was under-fitted. This would render the model as ineffective as overfitting. Reasoning about the Generalization Overfitting occurs when a network performs well on the training set but performs poorly in general. If the training set contains unintentional regularities, the network may overfit. Suppose If the job is to categorize handwritten numbers, for example, it’s possible that all photos of 9s in the training set have pixel number 122 on, while all other samples have it off. The network may elect to take advantage of this coincidental regularity, accurately identifying all of the training samples of 9’s without having to learn the true regularities. The network will not generalize well if this property does not hold on to the test set. Source Consider how the training and generalization error fluctuates as a function of the number of training examples and the number of parameters in order to reason qualitatively about generalization. More training data should only aid generalization: the larger the training set for any given test case, the more likely there will be a closely related training example. Furthermore, as the training set grows larger, the number of accidental regularities decreases, forcing the network to focus on the real regularities. As a result, generalization error should decrease as more training instances are added. Small training sets, on the other hand, are easier to memorize than big ones, therefore training error tends to grow as we add more examples. The two will eventually meet as the training set grows in size. This is depicted qualitatively in the figure above. Now consider the model’s capability. The more parameters we add, the easier it is to fit both the accidental and real regularities of the training data. As a result, as we add more factors, the training error decreases. The influence on generalization mistakes is subtler. If the network has insufficient capacity, it generalizes poorly because it fails to detect regularities (whether true or accidental) in the data. It will memorize the training set and fail to generalize if it has too much capacity. As a result, capacity has a non-monotonic influence on test error: it decreases and then increases. We’d like to create network topologies that are powerful enough to learn actual regularities in training data but not powerful enough to merely remember the training set or exploit accidental regularities. This is depicted qualitatively in the figure above. Measuring Generalization In most scenarios, we generally focused on the training set or tried to optimize the performance on the training set. But this is not the correct way of model building, there is a lot of uncertainty (such as noise) in the unseen data which is taken from the same distribution as the training set. So in such cases, we also should aim for our model that can generalize well on those unseen data. Fortunately, there is a simple method for assessing a model’s generalization performance. Simply put, we divide our data into three subsets. A training set is a collection of training examples on which the network is trained.A validation set is used to fine-tune hyperparameters like the number of hidden units and the learning rate.A test set designed to evaluate generalization performance. The losses on these subsets are referred to as training, validation, and test loss, in that order. It should be evident why we need distinct training and test sets: if we train on test data, we have no notion if the model is correctly generalizing or merely memorizing the training examples. There are other variations on this basic method, including what is known as cross-validation. These options are typically employed in cases with tiny datasets, i.e. less than a few thousand examples. The majority of advanced machine learning applications include datasets large enough to be divided into training, validation, and test sets. Apart from all these techniques, there is one called Regularization. Regularization has no effect on the algorithm’s performance on the data set used to learn the model parameters (feature weights). It can, however, increase generalization performance, i.e., performance on new, previously unknown data, which is exactly what we want. What is Regularization? In general, the term “regularization” refers to the process of making something regular or acceptable. This is precisely why we utilize it for machine learning applications. Regularization is the process of shrinking or regularizing the coefficients towards zero in machine learning. To put it another way, regularization prevents overfitting by discouraging the learning of a more complicated or flexible model. While modelling in regression analysis, the features are calculated using coefficients. Furthermore, if the estimates can be constrained, shrunk, or regularized towards zero, the impact of trivial features can be decreased, and models with high variance and a stable fit can be avoided. How does Regularization Work? The main concept is to penalize complex models by including a complexity factor that causes a larger loss for complex models. Consider a basic linear regression relationship to better comprehend it. It is stated mathematically as follows: Y≈ W0+ W1 X1+ W2 X2+⋯+WP XP Where Y denotes the learned relationship or the expected value. X1, X2,…, XP are the characteristics that determine the value of Y. The weights allocated to the attributes X1, X2,…, XP are W1, W2,…, WP. W0 is used to symbolize bias. To build a model that reliably predicts the value of Y, we’ll need a loss function and ideal parameters like bias and weights. In linear regression, the residual sum of the squares loss function is often used. The model will now learn using this loss function. It will alter the weights based on our training data (coefficients). If our dataset is noisy, it will suffer from overfitting, and the calculated coefficients will not generalize to previously unknown data. Here is where regularization comes into play. It penalizes the magnitude of coefficients to regularize these learned estimations towards zero. But first, let’s look at how it imposes a penalty to the coefficients. Different Regularization Techniques Each of the strategies below uses a different regularization norm (L-p) depending on the mathematical methodology that generates several types of regularization. The beta coefficients of the characteristics are affected differently by these techniques. The following are some machine learning regularization techniques: Most of the methods below use L1 and L2 norms. The L1 norm is calculated by adding the absolute values of the vector. The L2 norm is calculated by taking the square root of the sum of the squared vector values. Lasso Regression The coefficients are penalized to the point where they reach zero in the Least Absolute Shrinkage and Selection Operator (or LASSO) Regression. It gets rid of the unimportant independent variables. The L1 norm is used for regularization in this technique. L1-norm is added as a penalty.The beta coefficients’ absolute value is L1.The L1 regularization is another name for it.L1 regularization produces sparse results. When there are a lot of variables, this strategy comes in handy because it may be utilized as a feature selection method on its own. Ridge Regression When the variables in a model are multicollinear, the Ridge regression approach is employed to analyze it. It minimizes the number of inconsequential independent variables but does not totally eliminate them. The L2 norm is used for regularization in this sort of regularization. As a punishment, it employs the L2-norm. The L2 penalty is equal to the square of the magnitudes of the beta coefficients.It is also referred to as L2-regularization.L2 reduces the coefficients but never brings them to zero.L2 regularization produces non-sparse results. Choosing the Right Regularization Method Ridge regression is employed when all of the independent variables in the model must be included, or when there are multiple interactions. There is collinearity or codependency between the variables in this case. When there are multiple predictors available and we want the model to do feature selection for us, we use Lasso regression. When there are a lot of variables and we can’t decide whether to use Ridge or Lasso regression, Elastic-Net regression is the best option. Relating Generalization and Regularization By the discussion so far we can summarize generalization and regularization as below so that we can find a relationship between both the concepts. The term generalization is used to describe how the model is good at predicting the new instances which it didn’t see before. Ideally speaking the model should generalize the relationship as same as during the training phase. Whereas the term regularization is referred to as a process that enhances the generalization capabilities of the model.The poor generalization is due to problems like overfitting, underfitting, and bias-variance issues. We can easily interpret the generalization status of our model by just observing the training and validation accuracy or testing accuracy scores. Whenever there is a poor generalization we apply one of the regularization techniques mentioned above based on the method It penalizes the magnitude of coefficients which is responsible for generalization towards zero. Conclusion Through this post, we have seen two important terminologies that every professional and novice should know. To summarize, the generalization is a process, or, can be called a remark of your ML model that tells us how the model is good at predicting or estimating the true relationship of unseen data points. Whereas regularization, the name itself tells us that it regularizes the behaviour of your ML model, or in other terms, it enhances the generalization process. References GeneralizationWhat Is Regularisation?","excerpt":"Generalization and Regularization are two often terms that have the most significant role when you aim to build a robust machine learning model.","categories":["Deep Tech"],"tags":["AI (Artificial Intelligence)","Data Science","Deep Learning","Guide","Machine Learning","Regularization"],"author_name":"Vijaysinh Lendave","publish_date":"2021-10-27T09:00:00","publication_year":"2021","word_count":1835,"keywords":["Go","Regularization","machine learning","programming_languages:R","AI","AI (Artificial Intelligence)","RPA","ML","Machine Learning","programming_languages:Go","RAG","Aim","Deep Learning","Data Science","R","Guide"],"extracted_tech_keywords":["AI","machine learning","ML","Aim","RAG","R","Go","RPA","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/a-guide-to-generalization-and-regularization-in-machine-learning\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10095460,"title":"Nandan Nilekani Donates ₹315 Crore to IIT Bombay","content":"Nandan Nilekani, the co-founder of Infosys, has announced a donation of ₹315 crore to his alma mater, the Indian Institute of Technology (IIT) Bombay, in commemoration of his 50-year association with the institution. According to a statement, the purpose of the donation is to enhance the establishment’s infrastructure to a world-class standard, promote research in emerging fields of engineering and technology, and cultivate a thriving ecosystem for deep tech startups at IIT Bombay. The statement also highlights that this contribution is one of the most substantial donations ever made by an alumnus in India. https:\/\/twitter.com\/NandanNilekani\/status\/1671042324003164163 In the release, Nilekani expressed his profound gratitude towards IIT Bombay, stating that it played a pivotal role in his life, shaping his early years and providing the groundwork for his journey. As he celebrates five decades of his affiliation with this esteemed institution, he feels fortunate to contribute to its future and make a positive impact. “This donation is more than just a financial contribution; it is a tribute to the place that has given me so much and a commitment to the students who will shape our world tomorrow,” he added. Read: Why is Nandan Nilekani the ‘Go To Guy’ for GoI? It is worth mentioning that Nilekani had previously donated grants totalling ₹85 crore to the institution, making his overall contribution to IIT Bombay reach ₹400 crore. Nilekani and Professor Subhasis Chaudhuri, director of IIT Bombay, signed the Memorandum of Understanding (MoU) today in Bengaluru. The statement quoted Professor Chaudhuri expressing his belief that this significant contribution will pave the way for IIT Bombay to achieve global leadership in its endeavours. “We are extremely delighted to see our illustrious alumnus Nandan Nilekani continuing his foundational & pioneering contributions to the Institute. This historic donation will significantly accelerate the growth of IIT Bombay and will firmly set it on a path of global leadership,” Chaudhari said. Last year, Nilekani also contributed ₹36 crores grant for launching Nilekani Centre at AI4Bharat in partnership with IIT Madras, for advancing Indian language technology for creating social impact.","excerpt":"Nilekani had previously donated grants totalling ₹ 85 crore to the institute, making his overall contribution to IIT Bombay reach ₹ 400 crore.","categories":["AI News"],"tags":["IIT","Infosys","NANDAN NILEKANI"],"author_name":"Mohit Pandey","publish_date":"2023-06-20T14:58:04","publication_year":"2023","word_count":341,"keywords":["Go","API","Infosys","AI","programming_languages:R","ML","programming_languages:Go","NANDAN NILEKANI","IIT","R","startup"],"extracted_tech_keywords":["AI","ML","R","Go","API","startup","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/nandan-nilekani-donates-%e2%82%b9315-crore-to-iit-bombay\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10001680,"title":"Will Drone Delivery Ever Take Off In India?","content":"As per a recent study, there are close to 147.8 million e-commerce users in India, with the number estimated to grow to 210.46 million users by the year 2020. With the market for e-commerce sector looking positive and with stiff competition driving the industry currently, all the major industry players are regularly looking at better ways to cater to their customers. With the Amazon famously testing their drone delivery in 2017, the development seems to have had a trickling effect in India. With food delivery and ordering continuing to have a considerable market presence, Zomato in 2018, acquired TechAngle Innovation, a Lucknow-based drone delivery startup to expand its drone-based delivery ecosystem. The food delivery app has partnered with as many as 75,000 restaurants across 100 cities in India and with the acquisition, Zomato will be working with the drone delivery startup to develop a hub to hub delivery system which will be powered by hybrid multi-rotor drones which can lift as much as 5 kilos. Speaking about the role of drone delivery and other aerial innovation in driving businesses,  Deepinder Goyal, CEO and founder of Zomato said, “We believe that robots powering the last mile delivery is an inevitable part of the future and hence is going to be a significant area of investment for us.” Strict Compliances Presently under the rules stated under the Director General of Civil Aviation’s (DGCA), flying drones or Remotely Piloted Aerial Systems are mainly categorised into four based on the size of the vehicle and depending on its lifting capacity. The DGCA has defined five different categories of drones: Nano: Less than or equal to 250 grams Micro: From 250 grams to 2kg Small: From 2kg to 25kg Medium: From 25kg to 150kg Large: Greater than 150kg The DGCA announced Civil Aviation Requirements (CAR) on August 27, 2018, thus making drone flying legal in the country. As per the Drone Regulations 1.0, which laid out strict rules, all drones except for nano drones needs to be registered and must have a unique identification number issued in their name. It also stated that the drones can be flown only during the day time. However, as per the initial policy, the ministry placed a blanket ban on drone delivery by food and other goods by any e-commerce company and limited its usage to aerial photography, filmmaking, disaster relief, and recreational activities. Policy Revision Speaking about the advantage that the development can have on the India economy, Civil Aviation Minister, Jayant Sinha said, “There are opportunities for our aviation sector & for India’s startup industry as India is set to become a global leader as far as drone ecosystem is concerned. It’s important for us to have a policy roadmap and regulations that support the growth of the drone ecosystem.” The government, however, is all set to reverse its policy under the Drone 2.0 which was expected to come into effect in March 2019. Under the revised policy, it will now allow commercial use of drones fro drone delivery, air-taxis etc beyond visual line of sight. Road Ahead With the global drone delivery market expected to grow at a CAGR of 22.3% and its application being more dominant in the e-commerce segment, drone delivery in India is expected to grow steadily. In India, drone delivery will definitely catch a steady momentum once the revised set of policies come into being. With government initiatives like Make In India and Start-Up India receiving more importance, it is likely that an increased number of players will be venturing into the field.","excerpt":"As per a recent study, there are close to 147.8 million e-commerce users in India, with the number estimated to grow to 210.46 million users by the year 2020. With the market for e-commerce sector looking positive and with stiff competition driving the industry currently, all the major industry players are regularly looking at better […]","categories":["AI News"],"tags":["drone delivery","ecommerce"],"author_name":"Akshaya Asokan","publish_date":"2019-03-28T17:42:55","publication_year":"2019","word_count":591,"keywords":["Go","drone delivery","programming_languages:R","AI","innovation","programming_languages:Go","ecommerce","ViT","R","startup"],"extracted_tech_keywords":["AI","R","Go","ViT","innovation","startup","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/will-drone-delivery-ever-take-off-in-india\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10053162,"title":"Cybersecurity Breaches Of 2021 Worth Taking A Look","content":"Like every year, 2021 also saw some major cybersecurity breaches and data leaks that exposed the personal information of millions of people online. In addition, the pandemic made many companies go online, and this gave an opportunity to hackers to get more creative and use sophisticated tools to carry out their work. Let’s take a look at a few major cybersecurity breaches of 2021: Facebook user data breach In April 2021, media outlet Business Insider reported that a user in a hacking forum published the personal data of millions of Facebook users. Alon Gal, co-founder and CTO of cybercrime intelligence company Hudson Rock, discovered this first which exposed personal data of more than 533 million Facebook users from different countries with 6 million users from India. It included phone numbers, Facebook IDs, full names, locations, birthdates, and email addresses. The media report also talked about how a Facebook spokesperson said that the data had been scraped due to a vulnerability that the social media giant had patched in 2019. This was not the first time Facebook had suffered data breaches. Even in 2019, millions of users’ phone numbers were scraped from Facebook’s servers in violation of its terms of service. LinkedIn The data connected to 700 million LinkedIn users was posted on a dark web platform for sale in June 2021. By exploiting the API, the hackers “scraped” the data. The type of data stolen was email addresses, full names, phone numbers, LinkedIn usernames, personal and professional experience, and other social media accounts they held. LinkedIn, in a statement, said that it was not a data breach and no private LinkedIn member data was exposed. The initial investigation revealed that data was scraped from LinkedIn and other websites. It included the same data reported earlier this year in the April 2021 scraping update. Microsoft exchange This breach happened when four zero-day exploits were discovered in on-premises Microsoft Exchange Servers. It gave attackers full access to user emails and passwords on affected servers and administrator privileges on the server. They installed a backdoor that lets the attacker get entry to impacted servers even if the server is later updated to no longer be vulnerable to the original exploits. Over 250,000 servers have fallen victim to the data breach as of 9th March 2021. Accenture In August 2021, IT giant Accenture confirmed that LockBit ransomware operators stole data from its systems during an attack that impacted the company’s systems. The LockBit ransomware team claimed to have stolen six terabytes of data from Accenture’s network. They demanded a $50 million ransom. Acer In October 2021, Acer confirmed that its servers were breached by a group of hackers called Desorden. They managed to steal over 60 gigabytes worth of data that contained sensitive information about millions of Acer’s customers like names, phone numbers of clients, and corporate financial data. The hack was reported by the hackers themselves and was later confirmed to be true by Acer. Pixlr In January 2021, a database of 1.9 million user records belonging to online photo-editor Pixlr was dumped on a dark web hacker forum by a notorious hacker by the name ShinyHunters. Data like usernames, email addresses, country, and passwords were exposed. T-Mobile T-Mobile in August said that a US data breach had hit more than 40 million T-Mobile customers; the company has admitted and blamed it on a highly sophisticated cyberattack. It also said though personal information was stolen, no financial details were leaked. The world got to know about the breach after reports came out that criminals were attempting to sell a large database containing T-Mobile customer data online.","excerpt":"A look at a few major cybersecurity breaches of 2021","categories":["AI Features"],"tags":["Cybersecurity","data breach","hackers"],"author_name":"Sreejani Bhattacharyya","publish_date":"2021-11-10T15:00:00","publication_year":"2021","word_count":603,"keywords":["Go","API","programming_languages:R","AI","programming_languages:Go","data breach","Aim","hackers","Cybersecurity","R"],"extracted_tech_keywords":["AI","Aim","R","Go","API","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/cybersecurity-breaches-of-2021-worth-taking-a-look\/","complexity_score":3,"technical_depth":7,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10054709,"title":"MongoDB Announces Pay-As-You-Go Service In AWS Marketplace","content":"One of the leading general-purpose database platforms, MongoDB, recently announced that it will make it easier for customers to build, scale, and manage data-rich applications with MongoDB Atlas in AWS Marketplace in calendar Q1 2022. With the launch of a pay-as-you-go MongoDB Atlas with Free Trial in AWS Marketplace, developers will have a simplified subscription experience, and enterprises will have another way to procure MongoDB in addition to privately negotiated offers already supported on AWS Marketplace. There will be no up-front commitments required to use MongoDB Atlas on AWS, and customers will pay only for the resources they use and scale based on their needs. MongoDB is also launching the capability for customers to buy MongoDB Professional Services in AWS Marketplace at AWS re: Invent, providing enterprises with one place to implement, configure, and run MongoDB workloads. “MongoDB and its document model are deployed in many AWS and MongoDB customers’ environments to increase developer productivity, decrease cost, and scale to meet the needs of high growth applications. This is why both rising startups and global enterprises from around the world, across multiple industries, have chosen MongoDB as the data foundation for their next-generation, modern applications. MongoDB Atlas enables customers to extend their familiar environment into AWS to run the same applications with a common foundation and tooling regardless of whether it is on cloud or on-premises. We are excited about offering MongoDB Atlas in AWS Marketplace, which will make it easier for MongoDB customers to build, manage, and scale their data-rich applications on AWS,” said Alan Chhabra, Executive Vice President of Worldwide Partners, MongoDB. Stephen Orban, General Manager, AWS Marketplace & Control Services at AWS, said, “AWS Marketplace makes it easy for customers worldwide to find, subscribe to, deploy, and govern the best third-party software, data, and professional services offerings. MongoDB Atlas’ pay-as-you-go offering will provide customers with one of the most developer-friendly databases with an integrated and simplified subscription experience on AWS Marketplace, and we’ll continue to work with MongoDB to make AWS the best place for developers to run MongoDB Atlas.” MongoDB and AWS Marketplace remain committed to removing friction and making it easier for customers to procure and innovate in the cloud. Earlier this year, MongoDB and AWS launched integrations between MongoDB Atlas and AWS Wavelength, Amazon Kinesis Data Firehose, Amazon EventBridge, AWS PrivateLink, AWS App Runner, Amazon Managed Grafana, and more. With today’s new AWS Marketplace integrations, AWS customers can enjoy the benefits of consolidated billing via their AWS account.","excerpt":"With the launch of a pay-as-you-go MongoDB Atlas with Free Trial in AWS Marketplace, developers will have a simplified subscription experience, and enterprises will have another way to procure MongoDB in addition to privately negotiated offers already supported on AWS Marketplace.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","Amazon AWS","AWS","AWS services","Data Science","Data Scientist","Deep Learning","Machine Learning","MongoDB","mongodb atlas"],"author_name":"Victor Dey","publish_date":"2021-12-02T13:30:04","publication_year":"2021","word_count":414,"keywords":["Go","AWS services","AWS","mongodb atlas","AI","MongoDB","cloud_platforms:AWS","programming_languages:R","Machine Learning","programming_languages:Go","Amazon AWS","ViT","Deep Learning","Data Science","Data Scientist","R","AI (Artificial Intelligence)","startup"],"extracted_tech_keywords":["AI","AWS","MongoDB","R","Go","ViT","startup","cloud_platforms:AWS","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/mongodb-announces-pay-as-you-go-service-in-aws-marketplace\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10110900,"title":"Amnic Deploys GenAI across Code, Cloud and Kubernetes","content":"Cloud cost observability startup Amnic, has launched its Gen AI assistant, AmnicCoPilot. The tool aims to change the way Kubernetes (K8s) infrastructure is managed, offering better experience for development and Site Reliability Engineering (SRE) teams. It also helps companies monitor the way they spend on tech and the return on investments. Sathya Narayanan Nagarajan, co-founder and CTO at Amnic, expressed the significance of contextual responses in the realm of Kubernetes management. He stated, “AmnicCoPilot is designed to simplify the management, modernization and understanding of new containerized applications, rightsizing Kubernetes infrastructure to manage workflows, improve efficiency and reduce costs at scale.” Amnic was launched by three former Ola colleagues Ankit Bhati, Nimish Joshi and Nagarajan in 2021 to make the software development process easier. The foundation of the platform was laid when Bhati realised the need for Amnic while interacting with people from companies he has invested in. The AI-powered assistant was developed by the Amnic team to address a challenge faced by developers and SREs – the effective management of Kubernetes instances while monitoring the spending. Leveraging generative AI as its co-pilot, this assistant will assist teams in navigating the complex architecture of K8s, providing contextual, relevant, and personalized responses to queries specific to users’ Kubernetes instances. The key benefits of AmnicCoPilot are as diverse as the challenges it tackles. It will streamline the performance and management of containerized applications across AWS and Kubernetes. Integrating containerized applications into existing resources and production environments, will provide developers and IT operations teams with enhanced visibility. Moreover, AmnicCoPilot provides tailored recommendations and solution roadmaps, thereby reducing operational costs for IT teams. AmnicCoPilot is built on a suite of technologies, including Amazon Elastic Kubernetes Service (Amazon EKS) , Karpenter, Amazon Bedrock, Amazon SageMaker and underlying GPU nodes. It delivers contextual information and AI-based assistance to Amazon Elastic Kubernetes Service (EKS) users. Built on a suite of advanced technologies, including Amazon Elastic Kubernetes Service (Amazon EKS), Karpenter, Amazon Bedrock, Amazon SageMaker, and underlying GPU nodes, AmnicCoPilot exemplifies the startup’s commitment to make tech innovations easier through monitoring. This development solidifies Amnic’s position as an invaluable partner in the DevOps ecosystem and highlights its dedication to providing customers with solutions to enhance their cloud infrastructure.","excerpt":"Amnic was launched by former Ola colleagues in 2021 for companies to monitor their tech spending.","categories":["AI News"],"tags":["Generative AI","Kubernetes"],"author_name":"Tasmia Ansari","publish_date":"2024-01-17T12:04:07","publication_year":"2024","word_count":369,"keywords":["Amazon SageMaker","AWS","AI","R","ML","RAG","Ray","Aim","generative AI","Generative AI","Kubernetes","kubernetes"],"extracted_tech_keywords":["AI","ML","generative AI","Aim","Amazon SageMaker","Ray","RAG","AWS","kubernetes","R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/amnic-deploys-genai-across-code-cloud-and-kubernetes\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10078687,"title":"Startups Scratch the Surface of AGI Without Really Understanding It","content":"As research in AI moves ahead at a breakneck speed, the monumental goal of AGI may seem imminent. Investors are feeding millions of dollars into the AGI dream with companies like Open AI, DeepMind and Google Brain, all backed by big tech companies, leading the way. A report by Mind Commerce stated that investment in AGI will touch a massive USD 50 billion by 2023. However, the progress is more than just encouraging—OpenAI’s DALL.E 2 can create arresting images out of any text prompts, their GPT-3 can write about pretty much anything, DeepMind’s Gato, a multi-modal model promised to perform just about any task thrown at it. In July, a Google engineer, Blake Lemoine, who was working with a chatbot called LaMDA, was convinced that the robot was sentient. But for all the progress made, researchers worry that we might achieve intelligence in a way that will tick off the boxes of benchmarks but might not understand what this ‘intelligence’ is about. The concept of Explainable AI as demonstrated in a DARPA report Need for AI Interpretability Thomas Wolf, co-founder of Hugging Face, articulated these fears in a post on LinkedIn. Wolf noted that enthusiasts like him, who saw AI as a way to unlock deeper insights on human intelligence, now seemed to believe that even though we are seemingly inching closer towards intelligence, the concept of what it was still eluded us. “Understanding how these new AI\/ML models work at low level is key to this part of the scientific journey of AI and calls for more research on interpretability and diving in the inner-working of these new models. Pretty much only Anthropic seems to be really working on this type of research at the moment but I expect this type of research direction to be increasingly important as compute and large models become more and more widely available,” he stated. Wolf’s prediction isn’t a novel one. Researcher and author Gary Marcus has often pointed out how contemporary AI’s dependence on deep learning is flawed due to this gap. While machines can now recognise patterns in data, this understanding of the data is largely superficial and not conceptual—making the results difficult to determine. Marcus has said that this has created a vicious cycle where companies are caught in a trap to pursue benchmarks instead of the foundational ideas of intelligence. This search for clarity pushed a lot of interest into interpretability and the money followed later. Until a couple of years ago, explainable AI witnessed its time in the spotlight. There was a wave of core AI startups like Kyndi, Fiddler Labs and DataRobot that integrated explainable AI within them. Explainable AI started gaining traction among VCs, with firms like UL Ventures, Intel Capital, Light Speed and Greylock seen actively investing in it. A report by Gartner stated that “30% of government and large enterprise contracts will require XAI solutions by 2025”. However, most of the growth in explainable AI was expected to arise in industries like banking, health care and manufacturing—essentially, areas which placed a high value on trust and transparency and required accountability from these AI models. The money also flowed in that direction. VCs were more keen to put their money on more tedious applications that were focused on transforming an existing industry rather than a distant moonshot. Co-founder and CEO of Anthropic Dario Amodei Explainability in commercial AI and academic research Startups like Anthropic were started with a very different intention than this. Founded by OpenAI‘s former VP of research Dario Amodei along with his sister, Daniela, the startup was formed less than a year back with nine other OpenAI employees. The young company picked up USD 124 million in funding then. Not even a year later, it raised another USD 580 million. The Series B round was led by the CEO of FTX Trading, Sam Bankman-Fried and included participation from Jaan Tallinn, co-founder of Skype, Infotech’s James McClave and former Google CEO Eric Schmidt. What’s even more interesting is that Anthropic’s list of supporters didn’t include the usual suspects among deep tech investors. But, this is possibly because the startup is a non-profit organisation which immediately made it a deal breaker. Ironically, like Wolf mentioned, the work that Anthropic was doing was rare. It wasn’t like companies that were waving the explainable AI flag to cater to the market. It is quietly working on improving the safety of compute-heavy AI models and understanding the source of behaviour in today’s LLMs. After its big Series B funding, CEO Amodei said, “With this fundraise, we’re going to explore the predictable scaling properties of machine learning systems, while closely examining the unpredictable ways in which capabilities and safety issues can emerge at-scale. We’ve made strong initial progress on understanding and steering the behaviour of AI systems, and are gradually assembling the pieces needed to make usable, integrated AI systems that benefit society.” Worldwide funding in AI startups over a decade in US billion dollars A recent report titled ‘State of AI Report 2022’ by Nathan Benaich of Air Street Capital and Ian Hogarth of Plural Platform observed that funding in academic research in AI was drying up (with the exception of Anthropic) with a lot of the money shifting to the commercial sector. “Once considered untouchable, talent from Tier 1 AI labs is breaking loose and becoming entrepreneurial,” the report stated. More so, some research labs backed by tech giants have been shut down like Meta’s central AI research arm. “Alums are working on AGI, AI safety, biotech, fintech, energy, dev tools and robotics,” the document mentioned.","excerpt":"VCs were more keen to put their money on more tedious applications that were focused on transforming an existing industry rather than a distant moonshot.","categories":["Deep Tech"],"tags":["AI Startups"],"author_name":"Poulomi Chatterjee","publish_date":"2022-11-04T16:15:00","publication_year":"2022","word_count":930,"keywords":["Anthropic","Hugging Face","machine learning","OpenAI","AI","R","ML","RAG","deep learning","xAI","AI Startups"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","OpenAI","Anthropic","xAI","Hugging Face","RAG","R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/startups-scratch-the-surface-of-agi-without-really-understanding-it\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10131108,"title":"Google DeepMind Unveils Visual Processing Framework to Reduce Computational Costs","content":"A team of researchers from Google DeepMind has introduced a novel framework called Mixture of Nested Experts (MoNE) that significantly reduces computational costs for processing images and videos without sacrificing accuracy. This new approach dynamically allocates computational resources to different visual tokens based on their importance. Read the full paper here. MoNE achieves over two-fold reduction in inference time compute compared to baseline models while maintaining equivalent performance on standard image and video datasets. On the Something-Something-v2 video dataset, MoNE matched the baseline accuracy of 64.4% while using only 162.8 GFLOPs compared to the baseline’s 376.3 GFLOPs. The framework builds on the concept of nested models, where smaller sub-models are contained within larger ones. MoNE uses a router network to dynamically assign visual tokens to these nested experts of varying sizes. This allows more important or informative tokens to be processed by larger, more computationally expensive models while less important tokens are handled by smaller nested models. The team evaluated MoNE on image classification using the ImageNet-21k dataset and on video classification using the Kinetics-400 and Something-Something-v2 datasets. They found that MoNE consistently outperformed baseline models and other approaches like Mixture of Depths, especially at lower computational budgets. A key advantage of MoNE is its ability to adapt to different inference-time compute budgets using a single trained model. This flexibility allows the framework to meet varying computational constraints without requiring retraining. Visualisations showed that MoNE effectively identified important regions in images and videos, routing tokens from these areas to larger nested models. This demonstrates the framework’s ability to focus computational resources on the most informative parts of visual inputs. While primarily designed for encoder architectures, the researchers note that extending MoNE to autoregressive decoding in large language models remains a challenge for future work. They also highlight the potential societal impact of MoNE in reducing energy usage and carbon emissions during model inference, as well as democratising access to AI by allowing broader use of trained models without requiring extensive computational resources. As deep learning models continue to grow in size and complexity, approaches like MoNE that can significantly reduce computational costs. Also, maintaining performance is likely to become increasingly important for practical applications of AI in resource-constrained environments.Google DeepMind is constantly developing advanced AI models and frameworks. Last month, they introduced Foundational Large Autorater Models (FLAMe) for various quality assessment tasks, and also launched the MatFormer Framework, which enhances on-device capabilities by allowing users to mix and match AI models within a single framework to optimise performance for specific tasks.","excerpt":"MoNE’s key advantage is its ability to adapt to varying inference-time compute budgets with a single trained model.","categories":["AI News"],"tags":["Google Deepmind"],"author_name":"Gopika Raj","publish_date":"2024-08-01T15:02:44","publication_year":"2024","word_count":421,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","deep learning","Google Deepmind","R"],"extracted_tech_keywords":["AI","deep learning","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/google-deepmind-unveils-visual-processing-framework-to-reduce-computational-costs\/","complexity_score":3,"technical_depth":6,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10042407,"title":"Meet The Winners Of IBM Watson AI XPRIZE","content":"IBM and XPRIZE, a non-profit organisation focused on designing and implementing innovative competition models to solve real world problems, launched IBM Watson AI XPRIZE in 2016. The competition aims to accelerate the adoption of AI technologies to solve societal challenges. “The IBM Watson AI XPRIZE is more than a competition– it is a future-thinking initiative aimed at establishing a more balanced dialogue around AI– in particular about how humans and machines can collaborate towards a better future,” said Amir Banifatemi, General Manager of Innovation and Growth at XPRIZE. Earlier this week, the winners of the competition for 2021 were announced. Let’s take a look at the top three winners of the IBM Watson AI XPRIZE: Zzapp Malaria Zzapp Malaria won the grand prize, taking home $3 million and the People’s Choice Award for the ‘Most Inspiring Team’. Headquartered in Tel Aviv, Israel, Zzapp Malaria is on a mission to eliminate malaria, which kills over four lakh people every year. Co-founded by Arnon Houri-Yafin and Ari Eichler in 2016, Zzapp Malaria’s software solves operational challenges of targeting stagnant water bodies for mosquito control. Its AI algorithm analyses satellite images and topographical maps to identify malaria transmission hotspots, that is, areas where human populations and water bodies coincide. In addition, Zzapp Malaria uses a weather analysis algorithm to optimise the timing of interventions. The algorithm is exclusively developed by the IBM Data Science and AI Elite team for this purpose. The AI strategies are then communicated to field workers using a map-based mobile application. The app further guides field workers to identify, report and treat water bodies. Finally, the data collected by field workers are fed back into the system to improve the algorithms and recommendations. At present, Zzap Malaria is running antimalarial in Ghana, Zanzibar and Tanzania. The startup has tied up with local partners, including municipalities, foundations, gold mines, and NGOs, to curb the mosquito menace. According to media reports, the co-founders soon plan to expand operations to Kenya, India and Mozambique. Before starting Zzapp Malaria, Co-founder and CEO Arnon was a key member of the team that developed Parasight, a machine vision-based malaria diagnostics device sold in more than 20 countries. Aifred Health Based out of Montreal, Aifred Health uses machine learning to treat depression. The company was founded in 2017 by Robert Fratila and Sonia Israel. Aifred Health’s clinical decision support tool is accessible on all devices, supporting telehealth and in-person appointments. The algorithm processes the patient’s behavioural health information. The output helps the physician understand where the patient stands on the disease cycle, monitor them over time and determine the way forward. Its solution uses AI to learn from a pool of patients to offer treatment and reduce the time taken for a patient to reach remission. Aifred Health uses IBM Watson Health to get access to millions of records of observational depression data and improve the model. Once approved by FDA and Health Canada, Aifred Health plans to combine the clinical decision support tool with AI model for treatment selection, to create an end-to-end solution for the treatment of depression. Aifred Health won the second prize of $1 million. Marinus Analytics Headquartered in Pittsburgh, United States, Marinus Analytics was founded in 2014 by Cara Jones and Emily Kennedy. The company uses AI and machine learning to deploy SaaS applications in the law enforcement and public sector. The company’s mission is to protect the vulnerable and end systemic exploitation. Marinus Analytics’ services extract actionable insights from big data for detectives, social workers and cyber fraud investigators. Its flagship product Traffic Jam aims to prevent human trafficking, recover victims, and dismantle organised criminal networks. The firm uses tools including IBM Watson Discovery and IBM Watson Assistant. Marinus Analytics does not engage in the widespread use of facial recognition. It is one of the three organisations allowed to continue using Rekognition tools, to aid in the rescue of sex trafficking victims. The AI company is backed by the United States National Science Foundation and Bank of New York Mellon. Marinus Analytics won the third prize of $500k. The competition received 150 entries from across the globe. The teams were evaluated on the basis of four parameters– technical impact, real-world impact, scalability of real-world implications, ethics and safety.","excerpt":"IBM and XPRIZE, a non-profit organisation focused on designing and implementing innovative competition models to solve real world problems, launched IBM Watson AI XPRIZE in 2016. The competition aims to accelerate the adoption of AI technologies to solve societal challenges.   “The IBM Watson AI XPRIZE is more than a competition– it is a future-thinking initiative […]","categories":["AI Features"],"tags":[],"author_name":"Debolina Biswas","publish_date":"2021-06-25T18:00:00","publication_year":"2021","word_count":708,"keywords":["big data","data science","Go","machine learning","TPU","AI","Scala","Aim","analytics","R"],"extracted_tech_keywords":["AI","machine learning","data science","analytics","Aim","TPU","R","Go","Scala","big data"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/meet-the-winners-of-ibm-watson-ai-xprize\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":6427,"title":"Beyond BI, Before Predictive Analytics","content":"Usage of Advanced Analytics is fast becoming more pervasive amongst Insurers in India. Players with bigger market share and those that have been around in the industry for quite some time have been really aggressive in this play with in house analytics teams working on different business initiatives. At the same time we have insurers that have entered the market in last 3-5 years and are still evolving towards setting up mature distribution,sales process and finding right product mix. For these players all this hype around Predictive Analytics may not seem to be a priority although they may lend keen ears to it. Is it right time for them to adopt advanced analytics into business process? Some barriers one will encounter in this journey Consolidating data across silos Data quality Changes to business process and IT systems Training of business, operations and technical teams. All these might pose substantial barriers to surpass and a first successful integration may take anyway more than 5 to 6 months. In light of this, one would rather focus on business priorities, strengthening sales and distribution channel and going after market share. Is there a way we can adopt advanced analytics avoiding much of overheads? It is also a fact that for most people advanced analytics is equated with Predictive Analytics where one is making predictions about customer churn, cross sell purchase etc and deliverable is an algorithm which outputs customer scorecard or propensity. Advanced analytics has more to it than just predictive analytics. What else can we do? Customer Segmentation: This may seem plain simple Business Intelligence report to start with especially with defined set of segments in mind. But given that your business is collecting various different data points one may miss out on finer segments. At the same time you may not always find out interaction between different variables. There are unsupervised clustering techniques like K-Means, Fuzzy C and lot more that allow you to do segmentation more effectively without having a particular profile in mind. Once we get these segments and profile these out, business can devise right marketing or sales strategy for each segment. Understanding Customer Profile for Events: Events like Churn, Cross Sell purchase are possible to predict with historical data. Without building a predictive model one can study segments that show higher churn or most cross sell. Decision tree techniques like CHAID, C&RT are effective techniques in this regard. Results from this can be interpreted by marketing team to target right profiles across different channels, Renewal team can devise right retention strategy for different segments, and Sales team can begin sourcing right set of profiles that are needed for healthy business. Results of above studies can provide useful insights to business users. Think of it as strategic insights stopping short of technical implementation.","excerpt":"Usage of Advanced Analytics is fast becoming more pervasive amongst Insurers in India. Players with bigger market share and those that have been around in the industry for quite some time have been really aggressive in this play with in house analytics teams working on different business initiatives. At the same time we have insurers […]","categories":["IT Services"],"tags":[],"author_name":"Vikas Kamra","publish_date":"2014-11-04T17:04:15","publication_year":"2014","word_count":465,"keywords":["business intelligence","Go","TPU","programming_languages:R","AI","RPA","predictive analytics","analytics","data quality","R"],"extracted_tech_keywords":["AI","analytics","predictive analytics","TPU","R","Go","data quality","RPA","business intelligence","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/it-services\/beyond-bi-before-predictive-analytics\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":41739,"title":"Here’s What Facebook Is Giving Away For Free Now","content":"Big companies who are leveraging artificial intelligence are also updating their recommendation engines. For example, Amazon recently created an AI-powered real-time personalisation and shopping recommendation called Personalise. Facebook, on the other hand, has portrayed its machine learning skills numerous over the past decade. This week, the social media giant announced the open source release of its recommendation system called the Deep Learning Recommendation Model (DLRM). Need For A Recommendation Model Companies need an efficient recommendation model for making profits in their businesses. Especially the kind of industry that Facebook is, it needs to have a better than the average recommendation engine. By making DLRM open-source, Facebook wants to invite the entire developer community to contribute to make it better and to also address the issues that are faced in this class of models. Categorical Data: The challenge with the recommendation class of models is associating categorical data with higher level attributes using neural networks. These neural networks are the most common tools for building recommendation systems. But the difference of these models with the others is working with categorical data. Lack Of Details: Another challenge is the unavailability of abundant details of representative models and data sets. Lack of data is a challenge for neural networks. All About DLRM The model specialises in serving personalised results in production environments. It was implemented using the platforms of PyTorch and Caffe2. (It is available be found on GitHub.) DLRM combines the principles from both, collaborative filtering and predictive analytics-based approaches. This makes it possible to work with production-scale data and provide state-of-art results. This is how the DLRM wins over other recommendation models available. DLRM is assigned to benchmark the speed as well as the accuracy of these systems. Objectives of making DLRM open source was: To help the community find new ways to address the difficulties faced while working with this class of models. To encourage algorithmic suggestions, modelling, system co-design and benchmarking. This, in turn, will lead to new models and more efficient systems that can provide more relevant content to people using a wide range of digital services. The processes followed in the DLRM model follow the following order: 1.Categorical features are processed. This is done by processing embeddings and continuous features with a bottom multilayer perceptron (MLP) 2.Second order interactions of different features are computed explicitly 3.Results are processed with a top MLP and fed into a sigmoid function. This process is to calculate the probability of a click Image source: Facebook The open source implementation of the DLRM can be used as a benchmark to measure: 1.The speed at which the model performs 2.How various numerical techniques affect their accuracy How Can Developers Benefit With the DLRM going open-source, many developers can take advantage of that. Here are some ways that would benefit them: 1.DLRM uses PyTorch, Facebook’s open source ML platform, making it easy to access. Developers working in PyTorch, which is very common among ML practitioners will be able to use it for their recommendation engines 2.DLRM benchmark is written in Python. This accounts for a flexible implementation where the command line arguments define the model architecture, data set and other parameters, providing the same advantage of accessibility to developers 3.It runs on a realistic data set. This has an advantage for the developer community to allow measuring model accuracy. This is useful when making use of different models and techniques 4.It has a self-contained code that can interface with public data sets, including the Kaggle Display Advertising Challenge Dataset Outlook The social media giant, in recent times, has open-sourced many of its AI models and frameworks. Some of the examples are its Robotics framework called PyRobot and Prophet, which is used for a time series forecast. The recommendation engine is a very sensitive issue at Facebook because of the controversy that it involves itself with surrounding the subject. But the objective of its DLRM is to draw attention to the difficulties that personalised recommendation models face. The data science team at Facebook is looking forward to working with the community to address the difficulties and improve the recommendation models. Developing methods to come with improved models using deep learning and neural networks will eventually lead to an advancement in this class of models, bringing to the world a better recommendation system.","excerpt":"Big companies who are leveraging artificial intelligence are also updating their recommendation engines. For example, Amazon recently created an AI-powered real-time personalisation and shopping recommendation called Personalise.  Facebook, on the other hand, has portrayed its machine learning skills numerous over the past decade. This week, the social media giant announced the open source release of […]","categories":["AI Features"],"tags":["Deep Learning","Developers","Open Source","recommendation"],"author_name":"Disha Misal","publish_date":"2019-07-03T17:33:15","publication_year":"2019","word_count":717,"keywords":["data science","artificial intelligence","Open Source","machine learning","AI","neural network","PyTorch","ML","Ray","deep learning","analytics","recommendation","Deep Learning","Developers"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","neural network","data science","analytics","Ray","PyTorch"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/heres-what-facebook-is-giving-away-for-free-now\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":62010,"title":"Are Virtual Conferences Worth Attending?","content":"Events and conferences have emerged as a great way to build your network, hear from the best in the industry, and stay up-to-date on newer developments. However, for some people, the travel time, distance, and expense can be an obstacle in attending these conferences. Virtual conferences and events, on the other hand, offer the same substance and content, but from the comfort of your home. In this article, let us talk about why shifting to a virtual conference is the most reasonable choice anyone can make: Get Anytime & Anywhere Access Perhaps the best reason to attend a virtual conference is the flexibility of getting access to any session – anytime, anywhere. To attend a conventional in-person conference usually requires travel, accommodation, and other necessary amenities. However, with a virtual conference, you can avoid all associated hassles and just think of focusing on the sessions at the comfort of your own home. This not only saves money but also a lot of unnecessary inconveniences. Online events also offer you the opportunity to view the sessions and presentations anytime you want with a playback option. This gives a chance for attendees to take all the time they want to absorb all the information by re-playing the sessions. Attend Multiple Tracks At The Same Time With a virtual conference, you can avoid the issue of having to choose between multiple sessions to attend. One does not need to rush to attend a particular session as attendees will have the opportunity to participate in on-demand sessions after the session has ended, and even record them for further replays. Virtual conferences help you strengthen your knowledge, connect with experts from the industry and learn from the best from anywhere. Making the most out of these online events will help attendees to not only understanding the industry but also create a base before diving deep for your career. Build An Extensive Global Network Connecting with a larger and diverse global audience from different parts of the world, including speakers, panellists, and peers, is easier in a virtual environment than a traditional in-person event. Considering virtual conferences are designed beyond boundaries, it allows attendees to connect with a larger audience and peers across borders. Alongside, since the attendees are usually provided with the agenda before the event, it will give them the flexibility to plan their day around relevant sessions. Networking with Speakers & Peers Virtual conferences provide you with the opportunity to connect with anybody attending the event without personally knowing them. This could get a little tricky in an in-person event. With virtual events, attendees have the flexibility to connect with leaders and speakers, ask them questions, and even receive feedback in a one-to-one setting. During virtual events, attendees can also use the chat room facility to get in touch with relevant people to better understand the complex topics of the session. Group Learning Is Possible With the help of virtual conferences, group learning is possible. Virtual conferences and online events allow you as well as your peers and friends to attend the session with you — on a single pass, which not only reinforces learning but also provides better clarity for complex topics. Alongside, many individuals aren’t capable of buying passes to these conferences; however, with virtual conferences, all of your peers can pool in to buy a single pass in order to attend these events. Attendees are also allowed to record the sessions and share them with peers to get the most out of these events. Benefits like the ability to attend multiple tracks and sessions by speakers, as well as the ability to network virtually at the comfort of your home, make virtual conferences a perfect choice for attendees to learn and grow during this pandemic lockdown. plugin is one such virtual conference for developers and leaders in the data science and machine learning space that will give attendees direct access to the brilliant minds of AI and the data science ecosystem.","excerpt":"Events and conferences have emerged as a great way to build your network, hear from the best in the industry, and stay up-to-date on newer developments. However, for some people, the travel time, distance, and expense can be an obstacle in attending these conferences. Virtual conferences and events, on the other hand, offer the same […]","categories":["AI Features"],"tags":["ai conferences","conference","conferences","data science conferences","tools for virtual conferencing","videoconferencing","virtual AI conference"],"author_name":"Sejuti Das","publish_date":"2020-04-17T20:00:00","publication_year":"2020","word_count":661,"keywords":["data science conferences","tools for virtual conferencing","data science","ai conferences","machine learning","videoconferencing","AI","conference","programming_languages:R","conferences","virtual AI conference","R"],"extracted_tech_keywords":["AI","machine learning","data science","R","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/are-virtual-conferences-worth-attending\/","complexity_score":2,"technical_depth":5,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10100089,"title":"JFrog Introduces Support for ML Models To Streamline Workflows","content":"Data scientists, ML engineers and DevOps are used to not having a common process for delivering software. The Indian service provider, JForg has introduced ML Model  management capabilities to streamline the management and security of these models. The feature on JFrog platform brings AI deliveries in line with an organisation’s existing DevOps and DevSecOPs to accelerate, secure and govern the release of the machine learning components. The lack of common process amongst teams introduces friction, difficulty in scale, and lacks standards across a portfolio. ML models are incomplete with Python and are often served using Docker containers. Yoav Landman, co-founder and CTO of JFrog looks at this release as a unified software supply chain platform to help developers deliver trusted software at scale. “It can take time and effort to deploy models into production from start to finish. However, even once in production, users face challenges with model performance, model drift, and bias,” said Jim Mercer, Research Vice President, IDC Research. According to IDC Research, the AI\/ML market, including software, hardware and services, is poised to grow at 19.6%, approximately $500 billion in 2023. However, as more models are being moved to production, the end users often face cost and scaling challenges. JFrog’s latest management offers a proxy to the popular repository Hugging Face to cache open source AI models, protecting them from deletion or modification. It will also detect and block the use of malicious models. It scans model licences to ensure compliance with policies. Having a single system of record that can help automate the development, ongoing management, and security of models that get packaged into applications offers a compelling alternative for optimising the process, said Mercer. Commenting on the release, Yossi Shaul, SVP Product and Engineering, JFrog said, “We’re excited to give customers an easy way to proxy, store, secure, and manage models alongside their other software components to help accelerate their pace of innovation while remaining well-positioned for tomorrow’s demands.”","excerpt":"JForg has introduced ML Model  management capabilities to streamline the management and security of these models.","categories":["AI News"],"tags":[],"author_name":"Pranav Kashyap","publish_date":"2023-09-14T18:06:05","publication_year":"2023","word_count":324,"keywords":["Go","Hugging Face","machine learning","AI","ML","docker","Python","Rust","DevOps","R"],"extracted_tech_keywords":["AI","machine learning","ML","Hugging Face","docker","Python","R","Go","Rust","DevOps"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/jfrog-introduces-support-for-ml-models-to-streamline-workflows\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10018612,"title":"15 Latest Job Openings In Data Engineering","content":"This week, we have created a list of the latest job openings for Data Engineers. 1| Data Engineer I at Amazon Location: Bangalore Responsibilities: Deliver on data architecture projects and implementation of next-generation BI solutions.Build and deliver high-quality data architecture and pipelines to support business analysts, data scientists, and customer reporting needs.Manage AWS resources including EC2, RDS, Redshift, Kinesis, EMR, Lambda, etc. Apply here. 2| Data Engineer at IBM Location: Bangalore Responsibilities: Build cutting edge credit or fraud default models for risk analytics in a big data environmentAssist in building a process that maximises the operating efficiency while maintaining the risk across various lending cycles.Develop credit as well as fraud default models depending on the requirements of the customer, etc. Apply here. 3| Data Engineer at Jio Location: Mumbai Responsibilities: Develop data architecture, data modelling, and ETFL mapping solutions within a structured data warehouse environment.Experience in Open source technologies (Apache Spark, Hadoop, Hbase, Hive, Pig, Kafka, etc.)Design and build a data warehouse, streaming architecture, and machine learning pipelines, etc. Apply here. 4| Data Engineer at Zoom Location: Bangalore Responsibilities: Own and optimise Zoom’s data architecture to address the data needs of a rapidly-growing businessWrite complex and efficient queries to transform raw data sources into easily accessible models by coding across several languages such as Java, Python, and SQLArchitect and launch data models to provide instinctive analytics to the team, etc. Apply here. 5| Analytics Developer- Data Engineer at Airbnb Location: Gurugram Responsibilities: Implement downstream changes from backend infrastructure projects and operational changesBuild frameworks for automation, anomaly detection, and data quality checksCreating QA plans, validating metrics and peer-reviewing code: identifying opportunities for improvement contributing to the development of Airbnb’s best practices, etc. Apply here. 6| Data Engineer at Amazon Location: Bangalore Responsibilities: Data Warehousing experience with Big Data Technologies such as Hadoop, Hive, Hbase, Pig, Spark, among others.Experience in building as well as operating distributed systems of data extraction, ingestion, and processing of large data setsAbility to effectively communicate technical information as well as the results of engineering design at all levels, etc. Apply here. 7|  Data Engineer II at Uber Location: Bangalore Responsibilities: Build and work with real-time services along with batch data pipelines that track and attribute ad spend to actions that Uber users take.Build systems that get free traffic (SEO and content) instead of having to pay other ad networks for it. Apply here. 8| Data Engineer at Hitachi Vantara Location: Bangalore Responsibilities: Working knowledge of the various machine learning and deep learning applications including Keras, Python Scikit learn, among others.Working knowledge of Azure services like HD Insight, Azure DataBricks etc. Apply here. 9| Lead Engineer – Data Engineering at Johnson & Johnson Location: Bangalore Responsibilities: Work with the business teams to learn the problem statement as well as identify the different stakeholders and systems.Identify, develop as well as test database architectures and large-scale processing solutions. Apply here. 10| Data Engineer at Roposo Location: Nelamangala, Karnataka Responsibilities: Understand & provide innovative solutions to business and product requirements using Big Data ArchitectureTake ownership of end-to-end data-pipeline including system design and integrating required Big Data tools & frameworks, etc. Apply here. 11| Senior Data Engineer at Razorpay Location: Bangalore Responsibilities: Design as well as implement the internal process improvements, such as automating manual processes, re-designing infrastructure for greater scalability, etc.Build the infrastructure that is required for optimal extraction, transformation, as well as loading of data from a variety of data sources using SQL and other big data technologies. Apply here. 12| Data Engineer SE – ll at Dunzo Location: Bangalore Responsibilities: Build large-scale batch and real-time data pipelines with data processing frameworks like Apache Beam, Spark on the Google Cloud Platform.Collaborate with other engineers, business analytics and stakeholders, and lap up learning and leadership opportunities on a daily basis.. Apply here. 13| Senior Data Engineer at Nike Location: Bangalore Responsibilities: Designing and building reusable components, frameworks and libraries at scale to support analytics productsDesigning and delivering product features in collaboration with business and technology stakeholders Apply here. 14| Data Engineer – GIR Technology at Goldman Sachs Location: Bangalore Responsibilities: Interact with platform users in strategies and business-aligned technology teams around the globe and across various time zones.Develop solutions using test-driven development incorporating both unit and integration testing. Apply here. 15| Sr. Data Engineer at NetApp Location: Bangalore Responsibilities: Design, manage and optimise data models and flows for all of NetApp Public Cloud Services usage and billing data to derive intelligent business insightsUnderstanding the data, as well as designing and developing business logic and code to deliver automation in business intelligence, etc. Apply here.","excerpt":"This week, we have created a list of the latest job openings for Data Engineers. 1| Data Engineer I at Amazon Location: Bangalore Responsibilities: Deliver on data architecture projects and implementation of next-generation BI solutions. Build and deliver high-quality data architecture and pipelines to support business analysts, data scientists, and customer reporting needs. Manage AWS […]","categories":["AI Hirings"],"tags":["data engineer jobs","Data Engineering","Latest Data Science jobs"],"author_name":"Ambika Choudhury","publish_date":"2021-01-21T16:00:00","publication_year":"2021","word_count":763,"keywords":["machine learning","Keras","AWS","AI","Latest Data Science jobs","Apache Spark","data engineer jobs","deep learning","Data Engineering","analytics","anomaly detection","Kafka","Azure"],"extracted_tech_keywords":["AI","machine learning","deep learning","analytics","Keras","anomaly detection","AWS","Azure","Apache Spark","Kafka"],"url":"https:\/\/analyticsindiamag.com\/ai-hiring\/15-latest-job-openings-in-data-engineering\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10046196,"title":"Google Introduces Its Dataset Exploration Tool: Know Your Data","content":"It is difficult to completely differentiate data collection and labelling of a model from unconscious biases, data access limitations, and privacy concerns. AI models are created and trained by humans; they are bound to mimic humans’ biases and prejudices. This often leads to the datasets including unfair social biases along dimensions of race, gender, age, and more. When these biased AI models are used to solve critical societal problems, they result in skewed decisions. IBM research has found more than 180 human biases in today’s AI systems. For example, Google itself faced backlash when Google Photos classified black people as gorillas when Google’s facial recognition couldn’t recognise people of colour. This June, Facebook researchers spoke about creating a framework for identifying gender biases in texts to understand “social construct and identify languages.” Dataset examination tools assist in analysing how the representation of different groups is a crucial component in creating a moral ML model. For Google AI, this is a critical step to ensure the responsible use of ML datasets and point toward potential mitigations of unfair outcomes. Google recently introduced its dataset exploration tool, Know Your Data. KYD is Google’s tool to help researchers, engineers, product teams, and policy teams explore datasets, improve data quality and mitigate bias issues. Know Your Data “KYD is a dataset analysis tool that complements the growing suite of responsible AI tools being developed across Google and the broader research community,” Google AI’s blog post stated. “Currently, KYD only supports analysis of a small set of image datasets, but we’re working hard to make the tool accessible beyond this set.” KYD aims to answer the following questions: Is my data corrupted? (e.g. broken images, garbled text, wrong labels, etc.)Is my data sensitive? (e.g. are there humans, explicit content)Does my data have gaps? (e.g. lack of daylight photos)Is my dataset balanced across various attributes? KYD’s range of features includes filtering, grouping, and studying correlations. In addition, the tool uses Google’s Cloud Vision API to automatically compute labels and show the users signals that weren’t present in the dataset initially. Essentially, the users can explore the dataset in accordance with information that wasn’t present in it. COCO Captions Case Study COCO Caption’s image dataset contains five human-generated captions for each of over 300k images, including annotations provided by the free-form text. Researchers at Google applied KYD to this dataset to explore gender biases by analysing the gendered correlations within the image captions. As a result, the tool managed to find gender biases across various depictions and descriptions within the dataset. KYD’s ‘Relations Tab’ allowed the researchers to examine the difference between activities captioned with ‘man’ or ‘woman’ by visualising the extent to which two signals co-occur more (or less) than would be expected by chance. The cells have blue and orange colours signifying positive or negative correlation between two signal values. Watch the demonstration here. A screenshot from KYD showing results for activities and the gendered captions. KYD’s feature that filters rows of a relational tab further probed for captioned words with ‘ing’, such as ‘shopping’. The exploration tool found strongly gendered co-relations with activities like shopping or cooking that are stereotypically associated with women, occurring with more images captioned with ‘women’ than with ‘men’; vice versa with activities like ‘surfing’ or ‘skateboarding’. According to Google AI, the essence of the dataset exploration tool lies in the fact that these image captions are not derogatory or stereotypical. Still, the tool manages to find where certain groups are over-represented within an activity across the dataset. KYD quickly brings these risks to the surface, preventing the entire dataset from learning to make stereotypical associations. Google further used the tool to find an age bias in the COCO dataset, where physical activities were rarely captioned with words like ‘elderly’ or ‘old’. The researchers also found that relative to the caption ‘young’, ‘old’ was used more often to describe belongings or clothing rather than people. A screenshot from KYD showing results for physical activities and age specifications. KYD helped point out this underrepresentation; assumingly, because of a smaller dataset for older ‘people’, an aspect of the dataset needs to be worked upon. This tool is a step toward improving datasets and trailing ML models on fair and unbiased ones. The researchers claim this dataset to be very people-friendly with features for filtering and grouping in real-time and comparing various attributes.","excerpt":"Researchers at Google applied KYD to this dataset to explore gender biases by analysing the gendered correlations within the image captions","categories":["Global Tech"],"tags":["case study","Google"],"author_name":"Avi Gopani","publish_date":"2021-08-17T16:00:00","publication_year":"2021","word_count":730,"keywords":["Go","API","programming_languages:R","AI","ML","case study","responsible AI","Aim","ViT","data quality","Google","R"],"extracted_tech_keywords":["AI","ML","Aim","R","Go","API","data quality","ViT","responsible AI","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/google-introduces-its-dataset-exploration-tool-know-your-data\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":15903,"title":"How RailYatri is using analytics to make intelligent travel predictions","content":"Co founders of RailYatri. in , Sachin Saxena, Manish Rathi & Kapil Raizada (From Left to right) In a country where over 23.9 million passengers rely on railways as a mode of travel and daily commutation, it becomes important to keep serving the passengers with inclusive and comprehensive information, to help them make informed decisions. RailYatri is one such initiative that is efforting to simplify train travels, by using deep learning and computer vision. Aiming to be a premier portal for Indian Railways train travellers, it’s a mobile friendly site that answers all train travel questions in a few taps. At RailYatri, they work to present not just the dry numbers but provide insights behind them so as to help passengers make informed decision about their train travel. Some of the features that it enables are checking PNR status information with confirmation probability, live train status with real time GPS location mechanism, avoiding delayed trains by studying running time statistics, live trip sharing etc. In a nutshell it is a one stop solution that furnishes an expanse of data-based travel discovery at a passenger’s disposal. On the idea behind conception of RailYatri- An idea which was conceived as a data-driven “intelligent” and consumer-centric travel platform, RailYatri helps in reducing the uncertainties in long distance travel by addressing some of the questions like—will my ticket be confirmed? Which is the best train to take? How can I get medical assistance during my travel? And much more. “The journey has been fulfilling, so to say the least. We have successfully scaled to reach over 15 million app users – being the fastest to reach this scale”, says RailYatri, a startup founded by Manish Rathi, Kapil Raizada and Kapil Saxena. On being selected for Google’s Accelerator Program- The team believes that it truly validates the work that team has done so far and the positive impact that is being felt by the users at large. “We have a significant technology overlap in areas where Google has proven products & expertise e.g., big data, deep-analytics, run-time computations based on user locations, AI, machine learning, etc. which is core to how our platform generates its ‘recommendations’”, the company says. The startup is looking to engage with domain experts at Google on some of the key areas on which the team is currently working on. It is also focusing extensively on scaling consumer products. On analytics driven product portfolio- “Our solutions are designed to essentially help travelers throughout the travel lifecycle, and we focus on areas that help save travelers time, money or even anxiety”, says RailYatri”. For instance, their location-driven algorithms can accurately predict the arrival of a train based on the user’s GPS location & train profile and their seat availability predictions have achieved over 97% accuracy. Their app has mappings of equivalent bus & train routes that helps travellers easily switch across modes while doing a seat search. This startup claims that theirs is the only travel app that can recommend the approach side to the station, automatically show the nearest medical centre en-route your journey, and can facilitate user to check the quality of mobile network connectivity on their route. On using deep-analytics technology and big data platform that makes intelligent predictions- The data driven platform by RailYatri makes recommendations to travellers based on actual reported events. The company explains “One can imagine each significant change in status as an “event”. For instance, an event could be when someone cancels a booking, as it allows another waitlist passenger to get a seat. A train’s location & speed can be another event that defines its arrival time at a platform, along the same lines as how Google Maps estimates when you drive a car. RailYatri uses technology to track the events and analyse the impact to help predict the end outcome of interest to the traveller. On the big data platform that leverages intelligence and crowd sourced content, RailYatri says “Everyday, over 6 million user GPS data points are analysed to get the best train location and linked to the train profile to generate the ETA at downstream stoppages.” It further explains that even before the data can be consumed, it has to be first cleaned using multiple levels of filtering to separate the good data from the bad ones. This is where various modelling techniques are used to increase the information reliability using crowd-sourced data from different sources, and it is an ongoing exercise as every event adds to the learning process. A few use cases- On being asked to highlight a few use cases where analytics is being used to drive value at RailYatri, the company says that it is using it for instance predicting the platform numbers in advance that can help travellers plan their approach and potentially save on commute, effort and even money if they were to hire a porter. “A lot of travellers have appreciated the seat availability predictions as it helps them avoid spending on the multiple bookings in the hope that at least one would get confirmed”, it says. It further adds that the computations that go into into such calculations are fairly complex. To take a use case, the waitlist probability is calculated by measuring the impact of a nearly 20+ variables, including traffic patterns of other trains on the same route. RailYatri is solving this and many such complex problems using analytics. The roadmap ahead- The company says that is would continue to invest in leveraging crowd-sourced content, analytics based intelligence, user personalization and integrated commerce for a seamless travel experience for their travellers. “We believe that intelligence will define the next generation of travel platforms, and we would like to see RailYatri become the most preferred travel platform for all long distance travellers in India”, it concludes.","excerpt":"In a country where over 23.9 million passengers rely on railways as a mode of travel and daily commutation, it becomes important to keep serving the passengers with inclusive and comprehensive information, to help them make informed decisions. RailYatri is one such initiative that is efforting to simplify train travels, by using deep learning and […]","categories":["Deep Tech"],"tags":[],"author_name":"Srishti Deoras","publish_date":"2017-06-27T09:43:49","publication_year":"2017","word_count":965,"keywords":["Go","machine learning","AI","ML","computer vision","RAG","Aim","deep learning","analytics","R"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","computer vision","analytics","Aim","RAG","R","Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/railyatri-using-analytics-make-intelligent-travel-predictions\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10103046,"title":"Rust Provides the Ultimate Security Against Hackers","content":"Rust has been voted the most-loved programming language for eight years in its short life of 14 years. The popularity of the language is owed to its safety, one of the primary reasons it was created. Rust was designed to be a safer option, providing safety-first principles to ensure programmers write stable and extendable, asynchronous code. Rust is structured in such a way that it inherently prevents developers from inadvertently introducing the most prevalent kinds of security flaws that are exploitable. This characteristic of the language could greatly impact the routine process of patching vulnerabilities and improving cybersecurity. Earlier this year Microsoft began rewriting their core Windows libraries in Rust. “You will actually have Windows booting with Rust in the kernel in probably the next several weeks or months, which is really cool,” said David Weston, VP of OS security for Windows. He further said that, “The basic goal here was to convert some of these internal C++ data types into their Rust equivalents.” Additionally, with the backing of AWS, sudo and su are being rewritten in Rust to replace critical but outdated infrastructure components with memory-safe alternatives. Along with Microsoft, Rust is being actively embraced by Amazon, Apple, Google and Mozilla. Multiple Safety Features One of the primary security features of Rust is its emphasis on memory safety. This is achieved through a strict ownership model, which dictates how memory is allocated and managed. Each piece of data in Rust has a unique owner, and the language enforces rules about how and when data can be accessed or modified. This system effectively prevents common memory errors such as buffer overflows and null pointer dereferences, which are frequent attack vectors in other languages. Apart from its primary feature of safety in memory allocation, Rust stands out in its approach to concurrency, which is a key aspect of its design providing safety and security in multi-threaded applications. The language’s unique ownership rules are applied to its concurrency model, making data access thread-safe and free from data races. This careful handling of concurrency not only enhances performance but also significantly reduces a range of security vulnerabilities that are typically associated with multi-threaded environments. Complementing its concurrency model, Rust boasts a minimal to no runtime. This serves as a substantial security advantage. Unlike languages that depend on larger runtimes or virtual machines, Rust’s lean runtime architecture minimizes the potential attack surface. This means there are fewer components that could be targeted or exploited by hackers, enhancing the overall security of applications developed in Rust. Error handling in Rust is another cornerstone of its security framework. The language mandates that programmers explicitly handle potential errors, thereby preventing unexpected crashes or behaviors. This explicit and predictable approach to error handling is integrated into the language at the compile-time level, significantly reducing the chances of runtime errors that could be leveraged in cyber attacks. Rust also benefits greatly from its package manager, Cargo. Cargo is pivotal in maintaining secure code, as it efficiently manages dependencies, tracks library versions, and ensures that all components of a project are up-to-date. This functionality is crucial for security; it enables developers to promptly implement patches and updates, particularly for libraries that may have vulnerabilities. The active involvement of the Rust community plays a vital role in the language’s security posture. Regular updates and revisions by the community help to address known vulnerabilities and continually improve the language’s security features. This proactive and community-driven approach is integral to maintaining Rust’s resilience against security threats. In summary, while no programming language can offer absolute protection against hacking, Rust’s thoughtful design, encompassing safe concurrency, minimal runtime, explicit error handling, efficient package management, and an engaged community, positions it as a more secure alternative compared to languages like C and C++. These attributes collectively contribute to Rust’s ability to effectively mitigate a wide array of common vulnerabilities. Rapid migration Developers shifting from other languages are drawn to Rust’s compelling feature set. Its efficient management of concurrent programming enables parallel code execution, and its lightweight, fast nature, with benchmarks rivaling C\/C++, is a significant advantage. This shift is in line with the NSA’s recommendation to move from C\/C++ to memory-safe languages like Rust. Rust’s development has been user-centric, focusing on essential yet often overlooked features. These include generics, algebraic types, Foreign Function Interface (FFI) interoperability, a robust dependency management tool, and procedural macros, all of which contribute to a more enjoyable programming experience in Rust. In the tech industry, major players are adopting Rust for its benefits. Mozilla, for instance, is revamping Firefox with Rust to enhance its security, reliability, and performance. Similarly, Amazon is leveraging Rust for AWS and Kindle, and is even developing a Rust compiler for Java, prioritizing performance and scalability. Google and Dropbox are also embracing Rust. Google uses Rust in Chrome and Android and is creating a Rust compiler for Go to bolster security and reliability. Dropbox, meanwhile, is transitioning its backend to Rust, aiming for improved performance and scalability. Facebook, too, is tapping into Rust’s potential. The company is using Rust in developing the Libra blockchain and Oculus VR, and is working on a Rust compiler for C++, focusing on creating more secure and reliable software.","excerpt":"Rust, acclaimed for its robust security features, is rapidly becoming the preferred programming language against hackers.","categories":["AI Features"],"tags":[],"author_name":"K L Krithika","publish_date":"2023-11-15T12:27:09","publication_year":"2023","word_count":867,"keywords":["Go","AWS","AI","Scala","RAG","Ray","Aim","Rust","R","Java"],"extracted_tech_keywords":["AI","Aim","Ray","RAG","AWS","R","Go","Rust","Java","Scala"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/rust-provides-the-ultimate-security-against-hackers\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10043624,"title":"Krish Naik Speaks About His ML Journey &#038; Advice To Data Scientists","content":"Krish Naik is a hot shot in the field of data science education with over 397k subscribers for his YouTube channel. He is the co-founder of iNeuron.ai, where he dons both CIO and CMO hats. Analytics India Magazine got in touch with Krish Naik to understand his ML Journey and his take on India’s data science landscape. Excerpts: AIM: What attracted you to the field of data science? Krish: Before starting in data science and AI, I was a software engineer moving between net programming and Java. During that time, I had never heard about AI. In one of the projects, we were implementing a module based on personalisation using net programming. The functionality was very simple where based on a login, we had to recommend various insurance products to the user. Once the project got implemented, we got special recognition for this module, and the team called it an AI module. That was the first time when I first got to know about AI and machine learning. This excellent use case ignited a fire in me to learn about machine learning and data science. I started exploring more about it and could see how many different use cases and problems it could solve with ease. When I started working on some of the projects, I could see the overall development in me in terms of business, technical and presentation and many more things. Apart from this, I am also a huge fan of Mathematics and Stats, which motivated me to move towards data science. So until now, I have worked in more than four companies where I have successfully implemented around 6-7 projects in the field of data science and helped them earn exceptional revenues. I have also co-founded a company named iNeuron Intelligence (ineuron.ai), where we focus on providing affordable courses on AI-related technologies and parallelly, we do AI product development. AIM: What’s your advice for aspirants embarking on an ML journey? Krish: If anybody wants to start their ML journey now, you need to focus on three perspectives: Business Problem (domain knowledge)Technical skillsOutcomes and Improvements of the solution Business problem: As a data scientist, we need to think about the business problem and what use case we are solving. Not every issue needs to be solved by using data science or machine learning. To start learning, we need to focus on understanding the domain because that is the best step from where your requirement gathering and data gathering processes will start. The more you understand the business and the domain, the better solution you will be able to create with better accuracy. Technical skills: Technical skills involve understanding the programming language (I would suggest going with Python), understanding the entire lifecycle of data science projects, model deployment and retraining approaches. Outcomes and improvements: This step involves building the product with better functionalities and modules, bringing more improvements in the products. These steps are always to be kept in mind since the industries will be working in this manner, and if you have this thought process, it will definitely help you. AIM: What tools do you use? Krish: From a language perspective, I use Python, javascript, .net., frameworks and related libraries. I also use a lot of GPU computing. NVIDIA RAPIDS is one of the best tools for accelerated data science, especially when I want to do hyperparameter optimization. Unfortunately, CPUs are quite slow, so NVIDIA Cuda Libraries come to the rescue of NVIDIA Titan RTX. The NVIDIA GeForce RTX 3060 is the latest NVIDIA GPU I have been using. All in all, I can say that this acceleration helps to complete my everyday data science tasks faster, and I spend more time in algorithm optimization and insights gathering rather than spending time waiting for pre-processing and model training to get completed. AIM: What are your thoughts on India’s AI\/ML landscape? Krish: I have been working in the industry for about five years, and the Indian AI and Analytics startups continued to attract investment in 2019, receiving $762.5 million in funding, a 44% increase over the $530 million funding received in 2018. This steady growth in funding seems insignificant compared with the 368% growth from 2017 and 2018. This continued growth has attracted many people to move into this sector. This sector has also created many jobs, and many people are able to get some amazing salaries. AIM: You have said 2021 is the year of hiring data scientists. Could you elaborate? Krish: According to the recent survey from Gartner, around 4-5% have started incorporating AI in their day to day activities, and they also have mentioned that it’s going to increase in the upcoming five years. Apart from this, in the past six months in 2021, I have seen more than 500+ of my subscribers and students from iNeuron from different domains making a successful transition to data science. And more and more requirements are coming up in companies for these specific roles in Analytics. So this gives a hint that this year will be amazing for different data scientist roles.","excerpt":"Before starting in data science and AI, I was a software engineer moving between net programming and Java.","categories":["AI Features"],"tags":["data science in India","how to retrain data","Interviews and Discussions","tools for data science"],"author_name":"kumar Gandharv","publish_date":"2021-07-15T13:00:00","publication_year":"2021","word_count":844,"keywords":["CUDA","data science","Rapids","machine learning","AI","how to retrain data","ML","data science in India","Python","Aim","tools for data science","analytics","GPU computing","Interviews and Discussions"],"extracted_tech_keywords":["AI","machine learning","ML","data science","analytics","Aim","Rapids","GPU computing","CUDA","Python"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/krish-naik-speaks-about-his-ml-journey-advice-to-data-scientists\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10099379,"title":"AI Drone Beats Humans in Drone Racing For The First Time","content":"Researchers at the University of Zurich have created Swift AI which has beaten expert drones in high speed racing for the very first time. It managed to beat three top-level human pilots, including Alex Vanover, the 2019 Drone Racing League world champion, 60% of the time. Created by Leonard Bauersfeld and his team, the AI, known as Swift, won 15 out of 25 races, setting a new course record with a half-second lead over the closest human competitor. According to Bauersfeld, previously AI could only outperform human drone racers by relying on an unfair advantage: numerous cameras strategically placed around the racecourse and an external computer providing real-time instructions to the drone. In contrast, Swift works with just a single camera and computer, making it an autonomous system. “With millimetre precision and really high update rates, like 400 times a second, you know exactly where the drone is located in space and also how it is oriented,” said Bauersfeld in a recent interview. The model employs deep reinforcement learning to determine the most effective commands to navigate the circuit at high speeds. Given the technique’s reliance on trial and error, the drone crashed several times during its training phase within a simulation. During an actual race, Swift sends live video from its onboard camera to a neural network designed to identify the racing gates. This visual data is then combined with other information from an inertial sensor to calculate the drone’s precise position, orientation, and velocity.  After that a second neural network uses these estimations to determine the optimal commands to send to the drone. Analysis of the races showed that Swift consistently outperformed human pilots at the race’s beginning point, executing tighter turns and achieving speed. Its fastest lap clocked in at 17.47 seconds, surpassing the fastest human pilot by half a second. Nevertheless, Swift faced challenges, losing 40% of its races against humans and crashing multiple times. It is also sensitive to varying environmental factors such as changes in lighting conditions.","excerpt":"The AI drone beat Alex Vanover, the 2019 Drone Racing League world champion, 60 % of the time","categories":["AI News"],"tags":[],"author_name":"Tasmia Ansari","publish_date":"2023-09-01T16:51:34","publication_year":"2023","word_count":333,"keywords":["programming_languages:R","AI","neural network","RPA","R"],"extracted_tech_keywords":["AI","neural network","R","RPA","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/swift-ai-drone-beats-humans-in-drone-racing-for-the-first-time\/","complexity_score":3,"technical_depth":5,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10023789,"title":"Creating Interactive Data Reports With Datapane","content":"One of the issues of using Python for data analytics is the inability to create shareable data visualization reports quickly. Other analytics tools like PowerBI and Tableau can readily publish and share reports. Sure, you can create shareable reports with modules like dash and Flask but they require quite a bit of extra code. Datapane is a Python module that enables you to quickly create shareable reports from existing data analysis components like Pandas DataFrames and plots from a wide range of visualisation libraries like Plotly and Bokeh. Check out this article to learn the basics of creating a report with Datapane. Recently Datapane has introduced new components that make creating basic reports even easier and allow analysts to exercise more control over the report layout and view. DataTable One of these components is the DataTable block; this is an extremely powerful component, so much so that it facilitates the creation of a basic data report with just a few lines of code. The DataTable block takes a DataFrame and creates an interactive table that allows sort and search operations. In addition to that, it incorporates data analysis tools like Pandas-Profiling and SandDance by default. Let’s create a barebones report for the wine quality dataset using just the DataTable block. First, let’s install and set up Datapane. Install from PyPI pip install -U datapane To publish and share reports, you’ll need an API token. Register on the Datapane website and get your token from the account settings. import datapane as dp dp.login('API TOKEN') Load the dataset and pass the DataFrame object to dp.DataTable. Publish the DataTable block using dp.Report(block_name).publish(‘Report Name’, ‘Report Description’) import datapane as dp import pandas as pd df = pd.read_csv(\"winequality_white.csv\") table = dp.DataTable(df) dp.Report(table).publish(name='WineData', description='Winedata Example') That’s it! If you visit the above link you’ll find an interactive table like this. Not only that, using the buttons in the top-right corner, you can analyze the data with Padas-Profiling and SandDance. You can learn more about dp.DataTable here. More Sophisticated Layouts The default layout of Datapane reports is good enough for basic use cases, but if you want to create an intricate report layout with multiple pages, you can do so with Select, Group and Page components. The dp.Select component enables you to create different tabs for different blocks. Let’s say we have three components we want to display: white wine interactive table, source code of the whole report and a 3D scatter plot of pH, alcohol content and density. To share syntax-highlighted code instead of simple formatted text, we can use another one of the new components, dp.Code. It currently only supports Python and Javascript. white = pd.read_csv(\"winequality_white.csv\") white_table = dp.DataTable(white) fig_3D = px.scatter_3d(white, x='pH', y='alcohol', z='density', color='quality') # Create plot component from the 3D scatter plot plot_3D = dp.Plot(fig_3D, label = '3D Scatter Plot') # Create the selection of tabs select = dp.Select(blocks=[ dp.DataTable(white, label =\"Table\"), dp.Code(source_code, label='Source code'), plot_3D ]) You can learn more about dp.Select here. dp.Group can be used to display components side-by-side. Let’s say we want to display the interactive table for red wine data alongside a scatter plot of citric acid and quality. red = pd.read_csv(\"winequality_red.csv\") red_table = dp.DataTable(red) fig = px.scatter(data_frame=white, x='citric acid', y='quality') # Create plot component plot_2D = dp.Plot(fig) # Group table component and plot component group = dp.Group(blocks=[red_table, plot_2D], columns=2) You can learn more dp.Group here. To facilitate the creation of multi-page reports Datapane provides the dp.Page component. Let’s create a combined report of both the wine types using the previously created Select and Group components. page1 =  dp.Page(label='Red Wine', blocks=['## Red Wine', group] ) page2 = dp.Page(label='White Wine', blocks=['## White Wine', select] ) dp.Report(page1, page2).publish(name='Wine Layout Example') You can learn more about dp.Page here. The above report and its source code can be found here. References: To learn more about Datapane please refer to the following resources: Documentation GitHubWebsiteDemos","excerpt":"Datapane is a Python module that enables you to quickly create shareable reports from existing data analysis components like Pandas DataFrames and plots from a wide range of visualisation libraries like Plotly and Bokeh.","categories":["AI Trends"],"tags":["dashboards","Data Visualisation","interactive plots","Python data visualization tools","Python Libraries","report"],"author_name":"Aditya Singh","publish_date":"2021-04-14T18:00:00","publication_year":"2021","word_count":643,"keywords":["Go","dashboards","report","Plotly","Python Libraries","Data Visualisation","R","Git","interactive plots","Python","analytics","JavaScript","GitHub","Python data visualization tools","Java","Pandas"],"extracted_tech_keywords":["analytics","Pandas","Plotly","Python","R","JavaScript","Go","Java","Git","GitHub"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/creating-interactive-data-reports-with-datapane\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10133492,"title":"Apple Intelligence to Get a Robotic Arm Soon","content":"After having pulled the plug on its decade-long ‘Project Titan’ – Apple’s ambitious self-driving car project – earlier this year, the Cupertino giant is now ambitiously moving forward with its robotics project starting with a ‘tabletop robot’. However, Apple is finally looking to build on its robotics dream with the new home robot. From Cars to Home Apple’s robotics venture was highly aspirational from the beginning. The company had plans to build cars that could achieve Level 5 autonomy, and even have a controller or app for low-speed driving with an Apple command center for remote assistance. However, with the evolution of the autonomous vehicles market where prominent players such as Waymo, Baidu and now Tesla too, are spearheading through, Apple’s choice of abandoning it, seemed like the best option for now. Additionally, with their increasing focus in generative AI ventures, robotics seemed to have taken a backseat. While the team was disbanded, the same team is reportedly leading the tabletop robot project. “They’re all in on robotics right now. Robotics is the next big thing at Apple. They’re talking about humanoids. They’re talking about mobile robots to go around your home, and now they’re talking about this home device,” said Mark Gurman, a reputed journalist who covers Apple devices extensively. Apple’s Robotic Goals Apple’s new product is pretty much a slim robotic arm capable of moving a large screen (an i-pad), tilting it up and down, and enabling a full 360-degree rotation. The product is estimated to be priced around $1000. Away from self-driving cars, the robotic arm seems like a more realistic goal. A fully voice-controlled system that can move around the table will most likely be powered by Apple Intelligence. The device would identify different voices in the house and use a center stage-like feature to turn and face users whenever they speak. Foxconn to Power Apple Robots Recent reports indicate that Hongzhun, which is part of electronics manufacturing giant Foxconn, a key casing manufacturer under the Hon Hai Group, has been added to Apple’s robot supply chain partnership. It was chosen due to its expertise in mass producing robot-related components. The companies will continue to collaborate with further developments from Hongzhun being anticipated. Interestingly, Foxconn’s partnership with Apple has been a long-standing one, beginning as early as the 2000s with the production of iMacs. This relationship deepened over time, with Foxconn becoming the primary assembler of many Apple products, including the iPhone, which is a significant revenue driver for both companies. Foxconn has also set up factories in India for iPhone manufacturing. It is also possible that in the future, India may also serve as the hub for Apple robot product manufacturing. Meanwhile, Foxconn CEO Young Liu recently visited India and met many prominent people including Prime Minister Narendra Modi. His visit to Bengaluru has strengthened his relationship with the country, probably hinting that Apple’s future robotic device manufacturing could take place there. Far-Fetched Goals Will Continue Apple’s increased focus on the robotics sector is part of the company’s efforts to explore new product categories for revenue generation. This becomes important especially after the recent Apple Vision Pro debacle where the company witnessed a drop in sales by 75%. One of the main reasons for the drop is attributed to the exorbitant price of close to $4000 for a Vision Pro set. Furthermore, the hype and the interest around the device have also drastically declined. With the robot tabletop that is estimated to be launched in 2026, or later, it is likely that Apple would extensively invest in robotics. When the car dream shut down, people were speculating Apple’s entry into the humanoid segment. And with the way the humanoid race is gaining steam, that wouldn’t be a surprise either.","excerpt":"Foxconn is set to develop Apple’s first robotics product, an iPad integrated with a robotic arm.","categories":["AI Trends"],"tags":["Apple","foxconn","robotic arm","Robotics"],"author_name":"Vandana Nair","publish_date":"2024-08-22T14:00:00","publication_year":"2024","word_count":624,"keywords":["Go","programming_languages:R","AI","Apple","R","programming_languages:Go","Robotics","generative AI","ai_applications:robotics","robotic arm","foxconn"],"extracted_tech_keywords":["AI","generative AI","R","Go","programming_languages:R","programming_languages:Go","ai_applications:robotics"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/apple-intelligence-to-get-a-robotic-arm-soon\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10073903,"title":"SpaceX’s Starlink, T-Mobile Unlikely to End ‘Mobile Dead Zones’ in India","content":"Last week, SpaceX and T-Mobile announced their partnership to bring cell phone connectivity everywhere – ‘Coverage Above and Beyond’. With this collaboration, the duo looks to leverage Starlink (SpaceX’s constellation of satellites in low Earth orbit) and T-Mobile’s wireless network. The un-carrier plans to provide near complete coverage at most places in the United States and remote locations, previously unreachable by traditional cell phone signals. T-Mobile believes that this partnership with SpaceX would help them deliver coverage Beyond. Earlier this year, T-Mobile’s un-carrier gave customers enhanced connectivity beyond the reach of its network – in the air and abroad. With this new partnership, the company looks to provide the ultimate coverage experience. T-Mobile chief Mike Sievert said it is more than just a groundbreaking alliance. It represents two industry-shaking innovators challenging the old ways of doing things to create something entirely new that would further connect customers and scare competitors. The duo will create a new network, broadcast from Starlink’s satellites using T-Mobile’s nationwide mid-band spectrum to provide its service. This satellite-to-cellular service will provide nearly complete coverage almost anywhere in the world. With this, T-Mobile is planning to give customers text coverage practically everywhere in the US, Hawaii, parts of Alaska, Puerto Rico and territorial waters, and even outside the signal of T-Mobile’s network, starting with a beta in select areas by the end of 2023, after SpaceX’s planned satellite launches. SpaceX regulatory hurdles While it is currently available in the US, the chiefs Elon Musk and Mike Sievert look to expand globally, inviting the world’s carriers to collaborate for a truly global connectivity. T-Mobile plans to offer reciprocal roaming to those providers working with them to enable this vision. Last year, Starlink was planning to use Vodafone’s spectrum to expand its services in the UK. In India, T-Mobile currently works on GSM, which is available on networks such as Vodafone Idea, Airtel and a few other networks. This will most likely disrupt these carriers’ roaming services in India and other countries. It is interesting to see how this will all unfold in the coming months. On the other hand, Starlink in India has been stifled with regulatory challenges. In May 2022, SpaceX announced that its satellite internet services would be available in 32 countries worldwide. India was out of the picture. On its website, the blue colour on the map says ‘pending regulatory approval’. India is perhaps one of the only few countries on the map where Starlink service is unavailable, compared to most countries, where it says ‘Available Now’ or ‘Starting in 2023’. Meanwhile, India is gearing up for a 5G launch by October 12. It will be rolled out in Bengaluru, Delhi, Ahmedabad, Chandigarh, Gurugram, Pune and others shortly. The Department of Telecommunication (DoT) has already received INR 17,876 crore from telecom service providers like Reliance Jio, Adani Data Networks, Bharti Airtel, and Vodafone Idea, which have been provided spectrum at the recent auction. Bringing ‘mobile dead zones’ to life T-Mobile-Starlink is tackling a bigger problem at hand and looks to put an end to mobile dead zones entirely. In its blog post, the network provider said despite powerful LTE and 5G networks, well over half a million square miles of the US, in addition to vast stretches of ocean, are untouched by cell signals… from any provider. Further, it said that in those areas, people are either disconnected or pay exorbitant rates to lug around a sat phone. The duo believes they are taking the first step in making ‘mobile dead zones’ alive. The company aims to give an additional layer of connectivity in areas previously unreachable by cell signals from any provider. This can be useful when you are lost in a mountain while trekking\/hiking or stranded in the middle of the ocean. Interestingly, existing smartphones on the T-Mobile network will now be able to use the new service via the device’s existing radio. SpaceX chief Musk said there won’t be dead zones anywhere in the world for our cellphones. “We are incredibly excited to do this with T-Mobile,” he said. Users can send text messages, including SMS, MMS and participating messaging apps, which will empower customers to stay connected and share experiences nearly everywhere, followed by adding voice and data coverage. Déjà vu much Six years ago, Mark Zukerberg’s ambitious mission called ‘Free Basics’ to provide free internet access to rural and remote parts of India was rejected by the people it was intended to help. Not just that, the regulators, too, culled the idea for good. The idea was that ‘Free Basics‘ would provide limited access to the internet through a suite of websites, apps and services that, unsurprisingly, included Facebook. This mobile app would strip the websites of photos and videos and allow users to browse without paying for mobile data. But why did it fail? People felt that the free-but-restricted internet service did not follow the open nature of the internet, where all sites and online destinations should be equally accessible. This attracted a lot of backlash from telecom regulators with real-world protests, alongside an online campaign called ‘Save the Internet,’ making rounds. A decade ago, Google, too, faced a similar fate, where its ‘balloon’ experiment called Loon crashed big time. This project was intended to offer free internet to rural and remote parts of the world, connecting the last billion users. This was also intended to serve as an option for local telecom providers to reach more customers with internet access rather than having to spend hundreds of million dollars required to set up a physical infrastructure to reach a relatively small number of people at remote locations. Unfortunately for Loon, in January 2021, the company had to shut down its operations despite the support of several partners and investors. Now, it all seems like a déjà vu all over again with Elon Musk’s recent announcement of partnering with T-Mobile to offer voice and data coverage. This also brings us to question, if Starlink will own the internet one day, alongside questions around net neutrality.","excerpt":"India is perhaps one of the only few countries on Starlink’s radar, where service is unavailable due to pending regulatory approval","categories":["AI Features"],"tags":[],"author_name":"Amit Naik","publish_date":"2022-08-30T11:00:00","publication_year":"2022","word_count":1006,"keywords":["Go","ELT","programming_languages:R","AI","programming_languages:Go","RAG","Aim","ViT","R"],"extracted_tech_keywords":["AI","Aim","RAG","R","Go","ELT","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/spacexs-starlink-t-mobile-unlikely-to-end-mobile-dead-zones-in-india\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":25689,"title":"Space based solar power – harnessing energy from outer space, where the sun never sets","content":"It is pegged as the key to mankind’s survival. Space based solar power is the alternative to energy and gas emissions problem. According to American international nonprofit organization, National Space Society, solar power harnessed from space can generate large quantities of energy to each and every person on Earth, with very little environmental impact. With advancing economies and shortage of fuel, derived from oil, coal and natural gas and heightened concerns over energy supply, energy dependent economies are training their eyes on clean energy, especially space based solar power. NASA has taken big steps in collecting energy from the sun, in outer space. Dubbed SPS-ALPHA, Solar Power Satellite via Arbitrarily Large PHased Array (built on the concept of Retrodirective Phased Array), is a proposed first-of-its-kind solar power satellite that has the potential of delivering energy to one\/–third of humanity.  An affordable green energy option, it is supposed to be operational by 2025, could revolutionize disaster relief and become a way of life. How can energy be harnessed from sun in deep space According to National Space Society, the sun has a lifetime of 4-5 billion years, making it an ideal source for long term energy solution. Now, earth receives only a single part out of the 2.3 billion of the Sun’s output. According to Space Canada, a facilitator and enabler for space-based solar power, one of the major advantages of space-based solar power is that it is abundant and there are no carbon emissions. Space based solar power is an idea that has been around as long as space program itself. In the early days, visionaries imagined astronauts building structures in space the size of small towns. It was a dream to provide clean, abundant and reliable electricity and the most promising way to get that energy to the ground is via low intensity microwave beams. All one has to do is gather it up. But, the shadow of the earth gets in the way. To capture the energy of the sun efficiently, scientists propose building satellite out of the shadow, in deep space where the sun always shines. Solar panels in proper orbit absorb ten times the energy than those on the ground. And they do it 24\/7. Space based solar power – a bold idea whose time has come According to space Canada video, in space, where weight is no problem, gossamer thin structures can grow like a normal flower petals, many kilometress on the side and sunlight gathered by huge arrays can be converted into a microwave or laser beam, then transmitted to special ground stations on earth. This can provide more than all of our current energy needs. However, unlike communication satellite that use one transmitter to send one beam, the solar satellite beam involves millions of components that working together in harmony. The transmitter of a solar power satellite makes up many small wave micro transmitters and each one only produces the power of a light bulb. But there are lot of them spread out through a very large area that is called Retrodirective Phased Array that means that signals from all the transmitters are coordinated like an orchestra to produce a focused beam. Now, the conductor for the solar power satellite is called a pilot beam that comes from the ground. The receiving station on the ground shifts the beam up to the satellite, telling the array where to the point the beam and synchronize their transmission. The pilot signal provides a reference that all of the transmitters can lock on to, thereby focusing their energy on the same location. The amazing part of Retrodirective Phased Array is that the beam can be steered to a different part on the ground without any moving parts, the whole process takes place in milliseconds. While this beam carries a lot of energy, its intensity is low because all the transmitters are small and spread over a large area and it poses no threat to life on the ground or to the air. Looking ahead – developing a prototype No longer reserved for sci-fi films, Space Canada reveals the elements of this type of array have already been demonstrated on earth, that means the technology is there and the physics works. The next step is developing a small prototype in low earth orbit, only a few hundred kilometers up. It’s here the proposed SPS-ALPHA comes into play. In low earth orbit, the satellite continuously passes over the ground circling the earth every hour and a half keeping the beam focused on a series of targets. Then it moves up to the geostationary orbit where the satellite remains fixed over the same spot over the earth and point the beam to different locations on the ground. California-headquartered Solaren Corporation is on a mission to provide clean power to the planet by “designing, developing, launching and operating the world’s first Space Solar Power plants.” In a statement, Gary Spirnak led company revealed, “Though the concept has existed since the 1970s, it’s been revisited and abandoned because getting all the parts up there in outer space, and the people to put it all together, is supremely expensive. With the advent of super small, mass-produced satellites and reusable booster rockets, scientists are beginning to give a crack at making space solar a reality”. Challenges in making space solar power an everyday utility John Mankins, space solar power advocate who reportedly received a funding from NASA’s Institute of Advanced Concepts, 2011 to refine the space solar power plant concept cited though the technology required to make space solar a reality already existed, it is forbiddingly expensive. The financial hurdle is the biggest barrier to making solar power a reality. Gary Spirnak shared in the statement that one of the biggest challenges is getting the cost down to low double digits or even single digits of cents per kilowatt-hour is a must to make space solar a competitive utility. An idea, completely reliant on massive infrastructure investments for space solar to work. First up, one needs solid-state power amplifiers that can efficiently convert electricity from gathered sunlight into radio-frequency waves, then there are receivers on the ground that can re-convert RF waves into electricity. According to Solaren, RF converters, the technology behind cell phones is key to space solar research as well. Another hurdle is undertaking a massive construction project in space. The largest space construction project ever undertaken was the International Space Station that has a price tag of $100 billion. This is small compared to a solar powered satellite but with private players such as SpaceX building powerful rockets making it cheaper to reach space, build things in low earth orbit could become a reality.  Not to forget, the payback from the energy solar powered satellites provide is enormous.","excerpt":"It is pegged as the key to mankind’s survival. Space based solar power is the alternative to energy and gas emissions problem. According to American international nonprofit organization, National Space Society, solar power harnessed from space can generate large quantities of energy to each and every person on Earth, with very little environmental impact. With […]","categories":["IT Services"],"tags":[],"author_name":"Richa Bhatia","publish_date":"2017-03-28T03:20:16","publication_year":"2017","word_count":1128,"keywords":["Go","funding","TPU","programming_languages:R","AI","programming_languages:Go","Git","Ray","GAN","R"],"extracted_tech_keywords":["AI","Ray","TPU","R","Go","Git","GAN","funding","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/space-based-solar-power-harnessing-energy-outer-space-sun-never-sets\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10065872,"title":"Top open source datasets from Amazon","content":"Amazon has open-sourced Multilingual Amazon SLURP for Slot Filling, Intent Classification, and Virtual-assistant Evaluation (MASSIVE), a speech dataset that supports 51 languages to encourage developers to build more third-party apps and tools for its AI speaker device Alexa. It contains one million spoken samples and an open-source code to train multilingual AI models. It has been compiled through translators translating an English-only dataset into several languages spoken across Africa, Latin America, Europe, and Asia. It largely contains questions or common commands like asking a device to play a song or checking the weather situation. Over the years, Amazon and AWS have contributed massively to the open-source community by releasing their comprehensive datasets to the public. We will have a look at a few of them in this article. Amazon Customer Reviews dataset Amazon Customer Reviews is a collection of product reviews that have been collected over a period of over two decades. It contains over a hundred million reviews where customers have described their experience with products bought from the website. This makes the data a rich source of information for academic research, particularly in the field of NLP, information retrieval, and machine learning, among others. This dataset has been created to represent a sample of customer evaluations and opinions, which also reflect the variation in the perception of the same product across different geographical regions. Amazon Berkley Objects dataset Last year, Amazon and the University of California, Berkeley, jointly released the Amazon Berkley Objects dataset. It is a massive dataset of product images and associated metadata for supporting research on product information management, visual understanding, and information retrieval. It would like researchers to develop more powerful AI models for image-based shopping and for expanding retailers’ product graphs. This dataset includes images of close to 150,000 products that are all annotated with metadata like multilingual title, model, brand, product type, and dimensions, among others. Further, there are close to 400,000 static catalogue images, over 8,000 images that provide 360-degree rotations in the plane at 5-degree intervals, and over 7,000 product models that can be rotated along any axis and rendered in any 3D environment under different lighting conditions. SpaceNet Launched in 2016, SpaceNet is an open innovation project that offers a repository of freely available imagery with co-registered map features. SpaceNet hosts datasets developed by its team along with data sets from projects like IARPA’s Functional Map of the World (fMoW). Before SpaceNet, researchers had much lesser options to get free, precision-labelled and high-resolution satellite imagery. Cancer Genome Atlas The Cancer Genome Atlas is the result of a collaboration between the National Cancer Institute and the National Human Genome Research Institute. By analysing matched tumour and normal tissue samples from 11,000 patients, the group aims to generate comprehensive and multi-dimensional maps of key genomic changes in major types of cancer. The group was able to chart out a comprehensive characterisation of 33 cancer types and subtypes, including ten rare cancers. This dataset contains Clinical Supplement, miRNA-Seq Isoform Expression Quantification, Genotyping Array Masked Copy Number Segment, Genotyping Array Gene Level Copy Number Scores, and WXS Masked Somatic Mutation data from Genomic Data Commons (GDC), Whole Exome Sequencing (WXS), RNA-Seq, miRNA-Seq, and WXS Aggregated Somatic Mutation data. Genome Aggregation database The Genome Aggregation Database (gnomAD) is developed jointly by an international coalition of investigators who aggregate both exome and genome data from a range of large-scale human sequencing projects. The v2 data set of GRCh37 spans 125,748 exome sequences and 15,708 whole-genome sequences from unrelated persons. The v3 data set or GRCh38 contains 71,702 genomes selected as in v2. Foldingathome COVID-19 Datasets Folding@home is a major distributed computing project which uses biomolecular simulations to find the molecular origins of disease to accelerate the discovery of newer treatments. During the COVID-19 pandemic, Folding@home partnered with several experimental collaborators to accelerate the progress toward building effective therapies for treating COVID-19. One of the outcomes of these efforts was the creation of the world’s first exascale distributed computing resource to generate scientific datasets of massive size.","excerpt":"Over the years, Amazon and AWS have contributed massively to the open-source community by releasing their comprehensive datasets to the public.","categories":["AI Trends"],"tags":["Datasets"],"author_name":"Shraddha Goled","publish_date":"2022-04-28T17:00:00","publication_year":"2022","word_count":671,"keywords":["Datasets","API","machine learning","AWS","AI","distributed computing","RAG","NLP","Ray","Aim","R"],"extracted_tech_keywords":["AI","machine learning","NLP","Aim","Ray","RAG","AWS","distributed computing","R","API"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-open-source-datasets-from-amazon\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10136618,"title":"India Needs Legal AI to Navigate 44 Million Pending Court Cases","content":"In India, there are around 44 million pending court cases across over 1,200 courts, including tribunals, district courts, high courts, and the Supreme Court. In 2023 alone, 1.35 crore cases were filed, averaging 60,000 cases daily. And these cases are managed by around 1.5 million advocates, supported by millions of assistants across the country, conducting over 11 million searches on various legal databases and search engines. A lot to take in right? Imagine working on the research based on this vast data. Lexlegis.ai, an advanced LLM, is taking the lead in accelerating legal research. Speaking at Cypher 2024, India’s Biggest AI Conference by AIM Media House, Saakar S. Yadav, founder and CMD at Lexlegis.AI, emphasised, “Legal research is not just about finding judgments; it’s about finding precise answers”. He highlighted the complexity of the country’s legislation quoting the Income Tax Act in India which has undergone more amendments than any other law globally And so, lawyers and citizens alike often face the challenge of sifting through countless documents, statutes, and circulars to answer even basic legal questions. “We’re solving that problem,” Yadav declared, explaining that people don’t want just a list of documents, they want direct, meaningful answers and Lexlegis.ai aims to provide that. The platform uses advanced AI algorithms to not only search through millions of documents but also to synthesise information, offering concise legal answers within seconds. “AI is not just about retrieving data; it’s about making sense of it. We are focused on providing quality, shepherdised content,” he explained. Shepherdisation refers to the process of organising legal issues into a hierarchy that reflects the latest legal status. How does it Work? The platform is custom-trained on more than 10 million documents, a staggering 20 billion data points in total. “Our system learns from the nuances in legal language and case precedents, allowing us to provide accurate, contextually relevant answers,” he said. As AI models continue to evolve, each has its strengths and limitations. “We tested different AI models for various use cases, and each one had a unique strength,” Yadav mentioned, underscoring that AI is a tool to assist, not replace, human expertise. However, Yadav is clear that AI’s role in the legal sector is still in its infancy. “We’ve made great strides, but we’re only scratching the surface of what AI can do for the legal field,” he said. The next step for Lexlegis.ai is developing an end-to-end practice management system, which will streamline legal workflows for lawyers and legal professionals. “We started with a focus on tax-related cases, but now we’re building tools for all areas of law,” he shared. AI in Legal Practice Attorneys are increasingly embracing the power of AI, particularly Machine Learning (ML), to enhance their efficiency in contract review. Tools like Lawgeex allow legal professionals to spot issues and errors at a speed and accuracy that often surpasses human capability. By automating the review process, these innovations help ensure that critical details are not overlooked, streamlining the workload for legal teams. In the realm of legal research and discovery, AI’s impact is equally transformative. Algorithms have long been employed to sift through vast amounts of data in lawsuits, and now ML techniques are optimising this process even further. Services like CS Disco provide AI-driven solutions that assist law firms in identifying relevant documents while navigating the complexities of legal restrictions. Additionally, platforms such as Westlaw Edge have integrated advanced semantic search capabilities, enabling attorneys to delve deeper into legal texts with greater understanding and insight. Features like Quick Check can even flag potentially outdated case citations, ensuring attorneys remain well-informed in their arguments.","excerpt":"Lexlegis.ai, an advanced LLM, is taking the lead in accelerating legal research in India.","categories":["AI Features"],"tags":["AI (Artificial Intelligence)","AI Tool"],"author_name":"Vidyashree Srinivas","publish_date":"2024-09-25T15:22:06","publication_year":"2024","word_count":602,"keywords":["Go","semantic search","machine learning","AWS","AI","ML","RAG","Aim","GAN","AI Tool","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","machine learning","ML","Aim","RAG","semantic search","AWS","R","Go","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/india-needs-legal-ai-to-navigate-44-million-pending-court-cases\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10006123,"title":"NVIDIA’s GeForce RTX 30 Series GPU Is The Greatest-Ever Generational Leap","content":"“If the last 20 years was amazing, the next 20 will seem like nothing short of Science Fiction.”- Jensen Huang, CEO of Nvidia Jensen Huang, CEO and founder of Nvidia, has unveiled its all-new GeForce RTX 30 Series GPUs. Huang made the announcement from the kitchen of his Silicon-valley home as he also spoke about some of the new tools for GeForce gamers. The GeForce RTX 30 Series GPUs is the second-generation RTX GPUs that includes the trio of new RTX GPUs — the flagship GeForce RTX 3080, the GeForce RTX 3070 and the “ferocious” GeForce RTX 3090. During the virtual launch event, he claimed that this series is delivering the “greatest generational leap” in the company’s history. The series is powered by the NVIDIA Ampere architecture, also known as the largest 7-nanometer processor in the world. According to this blog post, the GeForce RTX 3090, 3080 and 3070 GPUs are said to have smashing performance records and have the capability to deliver up to 2x the performance. It is also said to provide 1.9x power efficiency over previous-generation GPUs. Also, the GPUs take advantage of the second generation of NVIDIA RTX, which is the world’s most powerful PC gaming platform in order to provide unprecedented levels of real-time ray tracing and AI gaming. Huang stated, “If the last 20 years were amazing, the next 20 would seem like nothing short of science fiction,” He added, “today’s NVIDIA Ampere launch is a giant step into the future.” Behind The RTX 30 Series Talking about the RTX 30 Series, Huang stated, “the days of just relying on transistor performance scaling is over.” RTX 3090 Priced at $1,499, the GeForce RTX 3090 or the Big Ferocious GPU (BFGPU) is said to be the world’s first GPU capable of playing blockbuster games at 60 fps in 8K resolution. The RTX 3090 is up to 50 per cent faster than the current ultimate PC graphics card and comes with a silencer, i.e. a three-slot, dual-axial, flow-through design that is up to 10x quieter than the TITAN RTX. RTX 3080 Built on a custom 8N manufacturing process, the flagship GeForce RTX 3080 has 28 billion transistors. It connects to Micron’s new GDDR6X memory, which is known to be the fastest graphics memory ever made. The starting price of RTX 3080 is $699 and is said to be the ultimate gaming GPU, i.e. up to 2x faster than the RTX 2080. RTX 3070 The GeForce RTX 3070 is powered by Ampere, NVIDIA’s second-gen RTX architecture. Starting at $499, RTX 3070 is built with enhanced Ray Tracing Cores and Tensor Cores, new streaming multiprocessors, and high-speed G6 memory. Marvels of 2nd Generation RTX The new RTX 30 Series GPUs along with the NVIDIA Ampere architecture includes a number of intuitive features. They are- New streaming multiprocessors: The new streaming multiprocessors deliver 2x the FP32 throughput of the previous generation.Second-gen RT Cores: The second-gen RT Cores deliver 2x the throughput of the previous generation.Third-gen Tensor Cores: The third-gen Tensor Cores provide a faster and efficient performance of AI-powered technologies.NVIDIA RTX IO: It enables rapid GPU-based loading and game asset decompression, accelerating input\/output performance by up to 100x compared with hard drives and traditional storage APIsWorld’s fastest graphics memory: RTX 30 Series include the world’s fastest discrete graphics memory, GDDR6X.Next-gen process technology: RTX 30 Series allows higher transistor density and more efficiency. New GeForce Tools Besides the trio of the new GPUs, Huang also introduced a slate of new tools for the GeForce gamers. They include- Nvidia Reflex: Nvidia Reflex is a suite of GPU, G-SYNC display, and software technologies that measure and reduce system latency in competitive games. NVIDIA Reflex features two major new technologies – NVIDIA Reflex SDK and NVIDIA Reflex Latency Analyser. It reduces and measures rendering latency, detects the clicks coming from the user’s mouse and then measures the time it takes for the resulting pixels to change on the screen. Nvidia Omniverse Machinima: NVIDIA Omniverse Machinima enables gamers across the world to easily master the art of storytelling by using 3D objects, animated by NVIDIA AI technologies. Nvidia Broadcast: NVIDIA Broadcast is a universal plugin that works with most popular live streaming, voice chat and video conferencing apps. It mainly harnesses AI to build virtual broadcast studios for streamers. Wrapping Up Popular video games like Fortnite, Call of Duty: Black Ops Cold War, Minecraft, among others turned on NVIDIA RTX real-time ray-tracing technology. On a concluding note, Huang stated that the second-generation NVIDIA RTX fusing programmable shading, ray tracing and AI would provide photorealistic graphics and the highest frame rates simultaneously. Huang said, “In this future, GeForce is your holodeck, your lightspeed starship, your time machine.” He added, “In this future, we will look back and realize that it started here.” Watch the video","excerpt":"“If the last 20 years was amazing, the next 20 will seem like nothing short of Science Fiction.”- Jensen Huang, CEO of Nvidia Jensen Huang, CEO and founder of Nvidia, has unveiled its all-new GeForce RTX 30 Series GPUs. Huang made the announcement from the kitchen of his Silicon-valley home as he also spoke about […]","categories":["Global Tech"],"tags":["big data video games","GPU","machine learning gpu","NVIDIA","NVIDIA GPU"],"author_name":"Ambika Choudhury","publish_date":"2020-09-05T10:00:24","publication_year":"2020","word_count":800,"keywords":["API","NVIDIA GPU","TPU","programming_languages:R","AI","machine learning gpu","RAG","Ray","Aim","big data video games","NVIDIA","R","GPU"],"extracted_tech_keywords":["AI","Aim","Ray","RAG","TPU","R","API","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/nvidias-geforce-rtx-30-series-gpu-is-the-greatest-ever-generational-leap\/","complexity_score":4,"technical_depth":8,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":30668,"title":"Top 5 Data Science Podcasts For Beginners","content":"With the technology changing at a meteoric pace, tech enthusiasts are usually one step behind when it comes to upskilling and keeping themselves updated. With rising popularity, buzzwords regarding data science have started to surface and have swarmed over the foundational concepts making them obscure for beginners. Here we list few subject-specific podcast episodes that can help you get an idea on what to expect from this field and what tools you require to become an expert Is Data Science Something For You It is always wise to have a cursory view of a new field before diving deep. In this episode of the podcast titled The Effective Statistician, industry experts will help you gain some insights on starting out in this field. They discuss how much of statistics, a data scientist needs to know and the areas where data science has had an early impact. Basic Concepts Of Statistics If you have made it through the previous podcast and are curious to know more about data science then you can start here. This episode is a part of The Science of Everything podcast and this episode will introduce you to key concepts like types of statistical data, sampling methods, the difference between descriptive and inferential statistics, statistical significance, and p-values. And briefly about, important statistical tests like chi-square test, t-test, and linear regression. Being Bayesian It is almost impossible to do a course on statistics without coming across Bayes’ theorem. Bayesian inference assists statisticians with optimal decision making. This episode explores the fundamental concept of what it is to be Bayesian — describing knowledge of a system probabilistically, having an appropriate prior probability, know how to weigh new evidence, and following Bayes’s rule to compute the revised distribution. Poisson Distribution The Poisson distribution is a probability distribution function used for events that happen in time or space. This episode by Linear Digressions, introduces you to the distribution and then explores some wide variety of applications using the Poisson distribution ranging from supernovas to a study on army deaths from horse kicks. Maximum Likelihood Estimation In Statistical Environments Designing machine learning models in statistical environments needs knowledge of frequencies at which events occur in the real world.  In the real world, the data is large, random and inconsistent. This episode is a part of the series A Gentle Introduction to AI and ML, by Learning Machines 101 and might help you develop some intuition regarding rudimentary statistical designs and, help one prepare for next level of machine learning.","excerpt":"With the technology changing at a meteoric pace, tech enthusiasts are usually one step behind when it comes to upskilling and keeping themselves updated. With rising popularity, buzzwords regarding data science have started to surface and have swarmed over the foundational concepts making them obscure for beginners.     Here we list few subject-specific podcast […]","categories":[],"tags":["data science podcasts","Statistics Machine Learning"],"author_name":"Ram Sagar","publish_date":"2018-11-26T11:05:06","publication_year":"2018","word_count":418,"keywords":["data science","machine learning","programming_languages:R","AI","ML","data science podcasts","Statistics Machine Learning","R"],"extracted_tech_keywords":["AI","machine learning","ML","data science","R","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/top-5-data-science-podcasts-for-beginners\/","complexity_score":2,"technical_depth":6,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":33757,"title":"The Recipe Behind Microsoft’s Beloved Chinese Chatbot XiaoIce","content":"Talking to voice assistants or chatbots are more like asking something to do with no emotions attached which is far different from talking to a human. With the advancements in artificial intelligence and machine learning, the tech giants are always looking to develop something out of the box. In an effort to make these assistants more human-like in nature, Microsoft has come up a chatbot that not only assists you whenever you want but also conveys it emotionally. AI-based chatbot, XiaoIce, translated to “little Ice” in Chinese is one of the ambitious projects of Microsoft, that was released by researchers in May 2014 at China. It’s uniquely designed as an AI companion with an emotional connection to satisfy the need for human communication and is catering to over 660 million users. The Making The design principle is mainly based on three specific steps: To create IQ and EQ: A sufficient high IQ (Intelligent Quotient) to help users complete specific tasks and high EQ (Emotional Quotient) to meet the user’s emotional needs like affection, social belongings, etc. The aim is to enable it to identify a personality which is a set of characteristics such as behaviour, cognition and emotional patterns that creates a distinctive character of an individual. The IQ capacities include knowledge, memory modelling, image, and natural language understanding, reasoning, prediction, etc. for the development of dialogue skills. The most sophisticated skill is the Core Chat that can engage long and open-domain conversations with the users. On the other hand, the EQ consists of two components, empathy (capability of understanding what the user is experiencing) and social skills. Social skills include query understanding, user profiling, emotion detection, sentiment recognition, etc. Social Chatbot Metric: Unlike the performances of other bots that are measured by task success rate, this social bot use of Conversation-turns Per Session (CPS) as social chatbot metric to measure the success of long-term emotional engagement with the users. Hierarchical Decision Making: A hierarchical decision-making process is cast so that the social bot interacts with the users over a sequence of discrete dialogue turn and at each turn, the bot observes the current dialogue state and chooses a response according to that dialogue policy. Three Layers Of The System Architecture User Experience Layer to connect the chatbot to different chat platforms. It includes a set of components like speech recognition and synthesis, text normalisation and image understanding that is used to process user inputs and responses of the chatbot. Conversation Engine Layer is composed of dialogue manager, an emphatic computing module, Core Chat, and dialogue skills. The dialogue manager keeps track of the state of dialogues and responds in reference to the particular state. Data Layer consists of the set of databases including the collected human conversational data., non-conversational data, knowledge graphs and the profiles of all the users. Features This is more than just a chatbot because it has certain features that are incomparable to other chatbots. Over these years, the chatbot has developed more than 200 different skills. This bot can predict what a user will say next to keep the discussion on and can communicate accordingly to the nature and state of the conversation. Here are some features of this amazing bot listed below: Unlike other virtual assistants, this chatbot gives you a human-like feeling and tries to become your friend no sooner you start chatting. It behaves the character of an 18 years old female persona that is reliable, sympathetic, affectionate, etc. and responds sensibly to sensitive questions by shifting skillfully to a new comfortable topic. It has strong language ability and visual awareness. This awesome bot has the unique ability to create poetry. A book of 139 poems is generated by the chatbot titled as “The Sunlight That Lost The Glass Window”. The book was published in Beijing and is said to be the first-ever poetry collection written by artificial intelligence. This bot is definitely many steps ahead in fashion than many of us since recently it has started designing images and patterns on fabrics for international fashion and garment producers. It can create immaculate designs that are popular among young generations and can paint unique images by itself. It has also been trained in a technique like reciting kid’s stories. The stories are automatically analyzed and it chooses suitable tones and characters to complete the entire process of creating a kid’s audiobook. This free-of-cost audio service is available for the users in Asia and is provided to five countries till now with different names like Rinna in Japan, Facebook messenger in India, etc.","excerpt":"Talking to voice assistants or chatbots are more like asking something to do with no emotions attached which is far different from talking to a human. With the advancements in artificial intelligence and machine learning, the tech giants are always looking to develop something out of the box. In an effort to make these assistants […]","categories":[],"tags":["AI Chatbot"],"author_name":"Ambika Choudhury","publish_date":"2019-01-19T11:26:20","publication_year":"2019","word_count":760,"keywords":["knowledge graphs","machine learning","artificial intelligence","programming_languages:R","AI","AI Chatbot","chatbots","virtual assistants","Aim","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","Aim","chatbots","virtual assistants","knowledge graphs","R","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/the-recipe-behind-microsofts-beloved-chinese-chatbot-xiaoice\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":7860,"title":"Rahul Yadav To Launch Data Analytics Startup","content":"Rahul Yadav, former CEO of Housing.com is planning to launch a start-up focused on visualization and data analytics. And Yadav’s new venture has perked up interests from none other than Paytm founder Vijay Shekhar Sharma, Flipkart co-founder Sachin Bansal and Micromax co-founder Rahul Sharma. The trio is believed to have made personal investment in the new venture. While Yadav has not yet disclosed the name of the new venture, speculation is rife that his company will focus on analyzing big data to predict consumer behaviour and provide speedy insights at key decision points. The data analytics and visualization firm is likely to be used by e-commerce companies on a day-to-day basis. Rahul Yadav is also scheduled to meet India’s richest man Mukesh Ambani, chairman of energy-to-retail conglomerate Reliance Industries (RIL). Later, he had posted on his Facebook page about his new venture as, “All Indian internet companies put together < My next venture”. Getting similar and strong awesome feelings again for an upcoming meeting with Mukesh Ambani. Interestingly, Softbank, which is a majority stakeholder in Housing.com, is also a major investor in Alibaba, which in turn has invested heavily in PayTm. Keeping in mind what Rahul is planning for his new startup, Rahul might not be joining Bigg Boss after all and wants to focus on the startup.","excerpt":"Rahul Yadav, former CEO of Housing.com is planning to launch a start-up focused on visualization and data analytics. And Yadav’s new venture has perked up interests from none other than Paytm founder Vijay Shekhar Sharma, Flipkart co-founder Sachin Bansal and Micromax co-founder Rahul Sharma. The trio is believed to have made personal investment in the […]","categories":["AI News"],"tags":[],"author_name":"AIM Media House","publish_date":"2015-09-15T13:12:17","publication_year":"2015","word_count":218,"keywords":["big data","programming_languages:R","AI","analytics","R","startup"],"extracted_tech_keywords":["AI","analytics","R","big data","startup","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/rahul-yadav-to-launch-data-analytics-startup\/","complexity_score":2,"technical_depth":6,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10014726,"title":"Will AWS Hold The Markets In 2021","content":"“What we build is driven by what you tell matters to you. And even if you can’t articulate a feature, we try to read between the lines, understand what you are trying to build and invent on your behalf.”Andy Jassy, CEO, AWS It has been nearly a decade and a half since the world was introduced to Amazon Web Services and today it has grown into one of the most successful cloud infrastructure companies on the planet, occupying more than 30% of the market. AWS has slowly risen to become a key driver of Amazon’s large gains. For instance, AWS posted $11.6 billion in sales in the last quarter, up 29% from a year earlier. Even Gartner, in their latest survey, named AWS as the leader, for the 10th straight time. Making A Case For AWS For The Coming Decade Amazon has revolutionised the way companies deal with their big data chores. Their success stimulated cloud markets and opened up to other players like Microsoft’s Azure and Google’s Cloud Platform. AWS one-upped their rivals almost every year by releasing new features such as customised machine learning tools and in-house chips. Not that Google or Microsoft do not possess any of these features but AWS’ first mover’s advantage combined with a strong portfolio of customers that include media behemoths like Netflix, keeps them afloat regardless of the rising competition. “When you look at an unmatched array of instances that you have in AWS coupled with the relentless innovation in chips, virtualisation layer and compute platform, you are now getting reinvented instances every few months, instead of every few years,” said Andy Jassy at a conference recently. Let’s take a look at few of the most interesting AWS announcements in this year alone that hints at the company’s bright future: Machine Learning Made Easy At the recently concluded AWS re:Invent virtual conference, Amazon launched Redshift Machine Learning — Amazon Redshift ML. Data scientists can now create, train and deploy machine learning models in Amazon Redshift using SQL. Powered by SageMaker, Amazon Redshift ML is a fully managed ML service. Here one can use SQL statements to create and train machine learning models from data in Amazon Redshift. The models then can be used for applications like churn prediction and fraud risk, among others. And, in another attempt to democratise data science, in June, Amazon introduced Honeycode. With Honeycode, users can develop web and mobile applications without writing code. On the other hand, Kubernetes is also now leveraged with AWS. CEO, Andy Jassy has also announced, ECS Anywhere and EKS Anywhere — container-services offered by the company, where the customers can use popular tools like open-sourcing EKS. The ever increasing ML tool stack at AWS is a testimony to the fact that the company is betting big on a future where ML as a service, will overtake other services. Check more announcements from re:Invent here. Growing Silicon Ambitions Since we launched EC2 Inf1 instances powered by #AWS Inferentia, customers like @Snapchat, @CondeNast, & @AnthemBCBS have used Inf1 for #ML inference at lower cost. Now #Alexa runs the majority of its inference workloads on Inf1, as well. https:\/\/t.co\/YiyEXD776T— Andy Jassy (@ajassy) November 14, 2020 “AWS has its own custom-built AI chip, Inferentia and even a custom-built processor Graviton2.” Majority of the data centres across the world widely adopt the integrated solutions provided by companies like NVIDIA, or Intel. AWS’ decision to install its own hardware resonated with that of Apple when the smartphone maker announced their separation from Intel. ML applications are also improving and increasing, and to process these, one has to have decent processors. For example, Amazon’s Alexa services draw in tens of millions of customers every month. AWS claims that there are more than 100 million devices connected to Alexa and the resulting ML workloads are usually heavy. To handle this, AWS has announced that they are migrating these workloads with their own Inferentia chips. Last week, AWS announced yet another custom ML chip — AWS Trainium, which according to the company, provides the best price-performance for training machine learning models in the cloud. It offers the highest performance with the most teraflops (TFLOPS) of computing power for ML in the cloud. The Trainium chip is optimised explicitly for deep learning training workloads for applications including image classification, semantic search, translation, voice recognition, natural language processing and recommendation engines. AWS Trainium shares the same AWS Neuron SDK as AWS Inferentia, making it easy for developers using Inferentia to get started with Trainium. An India-Centric Focus Happy to announce the largest FDI in the history of Telangana! After a series of meetings, AWS has finalized investment of ₹20,761 Cr ($ 2.77 Bn) to set up multiple data centers in TelanganaThe @AWSCloud Hyd Region is expected to be launched by mid 2022#HappeningHyderabad pic.twitter.com\/XuGxFfSFsS— KTR (@KTRBRS) November 6, 2020 Amazon has also recently announced that it is expanding its services to include a second infrastructure region in India by mid-2022. The company believes that the new AWS Asia Pacific (Hyderabad) region aims to assist developers, startups, and enterprises as well as the government to run their applications and serve end-users from data centres located in India. AWS investment interests in India dates back to June 27, 2016, when the first region was opened in Mumbai. Today, AWS has expanded its services through edge locations in Bangalore, Chennai, Hyderabad, Mumbai, New Delhi, and Kolkata. Now, hundreds of thousands of Indian organisations can leverage additional infrastructure to leverage advanced AWS’ that includes computing, storage, analytics, AI, database, Internet of Things (IoT), machine learning, mobile services, serverless, and more to drive innovation. AWS also has its own academy whose aim is to upskill local developers, students, and the next generation of IT leaders in India. According to the company, AWS Academy provides a free ready-to-teach cloud computing curriculum that prepares students to pursue industry-recognised certifications and in-demand cloud jobs. “Amazon plans to train 29 million people in AWS by 2025.” Resonating with AWS’ initiatives such as AWS Academy, Amazon has also announced that they would be training 29 million-strong workforce to work for them by 2025. Amazon will train people across the globe in AWS and will also be partnering with non-profit organisations, schools and others. Machine learning, silicon and Asian markets have become the cornerstones for the success of modern-day big tech companies, and AWS has all the ingredients in the right proportion to position itself to reign supreme not just for 2021 but probably the next decade as well!","excerpt":"“What we build is driven by what you tell matters to you. And even if you can’t articulate a feature, we try to read between the lines, understand what you are trying to build and invent on your behalf.” Andy Jassy, CEO, AWS It has been nearly a decade and a half since the world […]","categories":["Global Tech"],"tags":["AWS","Azure","Cloud Computing","Google Cloud"],"author_name":"Ram Sagar","publish_date":"2020-12-18T11:00:00","publication_year":"2020","word_count":1084,"keywords":["data science","Google Cloud","machine learning","AWS","AI","ML","RAG","Ray","Aim","Cloud Computing","deep learning","analytics","CuPy","Azure"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","data science","analytics","Aim","Ray","CuPy","RAG"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/aws-cloud-markets-top-2021\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10075731,"title":"The Auto Industry Focus for NVIDIA is So Clearly Noticeable at GTC 2022","content":"At GTC 2022, NVIDIA announced their new innovations in the fields of omniverse, automotive, healthcare, and robotics ranging from bots to autonomous vehicles. In the automotive industry, NVIDIA is striving to accelerate innovations in autonomous vehicles, infotainment systems, and graphic simulations in the omniverse and unveiled new systems and technologies in these fields. Read: Nvidia GTC 2022 is Happening. Here’s What to Expect “Humans never stop learning. We need to think of an AI car in the same manner.” The pipeline of creating an autonomous vehicle from perception to mapping, to localisation, continues to be an endless process. NVIDIA’s safety first policy has rated them with ISO 26262 safety, with developing sensors and algorithms of their automotive products safe and secured by 15,000 engineers across 5 million lines of code. DRIVE Thor “Advances in accelerated computing and AI are moving at lightspeed,” said Jensen Huang, founder and CEO of NVIDIA. NVIDIA’s DRIVE is an end-to-end platform for autonomous vehicle development and deployment. Launched in 2015, DRIVE has been providing solutions for the transportation industry using deep learning networks. NVIDIA DRIVE introduced Thor, a centralised car computer system that includes infotainment, automated driving, and cost-saving system. In 2018, Thor chip, Parker, had one teraflops of performance. Now, the Thor Superchip achieves 2,000 teraflops performance with FP8 precision that allows transition from 32-bit FP data to 8-bit integer format, unifying all intelligent functions into a single architecture for great efficiency. With 77 billion transistors, the architecture leverages Hopper GPU, Next-Gen GPU, and Grace GPU enabling concurrent processors running without interruptions. All the automotive companies’ will be able to implement one single architecture for all their AI and technological installations in their vehicles by 2025, starting with Geely-owned automaker ZEEKR for their next generation electric vehicles. Thor is the first AV platform that is incorporating inference transformer engine, a component of Tensor Cores within NVIDIA GPUs which will accelerate inference performance of deep neural networks by nine times, paramount for supporting workloads related with self driving. DRIVE Thor is replacing DRIVE Atlan and is a follow-on to DRIVE Orin that is currently in production in most vehicles with a 254 tops performance. The superchip also uses NVLink C2C to equip automakers to compute headroom and flexibility and help in building software-based vehicles, DRIVE Concierge The in-vehicle infotainment (IVI) has been upgraded with NVIDIA DRIVE Concierge platform, bringing in capabilities beyond music and directions. The Concierge platform offers AI-powered safety, convenience, and entertainment features for every passenger of the vehicles with personalised cluster capabilities. Using the DRIVE Thor cross-domain platform, Concierge can host multiple virtual machines using a single chip for simultaneous and streamlined development. To make the console user friendly, the platform supports the Android Automotive operating system for customisation. Additionally, the digital cockpit also enables premium functionality across the vehicle with video conferencing capabilities, digital assistants, and even gaming. Along with the entertainment features, Concierge is also integrated with the DRIVE Chauffeur platform that provides 4D visualisation and 360-degree view for safety and comfort of the passengers. NVIDIA is also taking the next step of using the software capabilities to include in-vehicle gaming systems and stream on any screen. The company utilises their GeForce NOW cloud gaming service, to allow streaming 1,400 titles without downloading and unlimited cloud storage. Using the DRIVE AGX compute platform, DRIVE IX software stack and Omniverse Avatar Cloud Engine (ACE), NVIDIA is making in vehicle time more enjoyable. When it comes to safety and security, Concierge uses interior sensors and deep neural networks for driver monitoring, ensuring attention when the human is in control. In addition to this, Omniverse ACE is developing easily customisable, interactive, AI-assistants with speech AI, computer vision, and natural language understanding. These created avatars can also make recommendations, provide alerts, and access vehicle controls. OVX Computing Systems to automotive industry NVIDIA announced the second generation of OVX, which is powered by Next-Gen GPU architecture and delivers real-time graphics, AI and digital twin simulation capabilities. The OVX systems build 3D virtual worlds using 3D applications to build digital-twin in NVIDIA Omniverse Enterprise. The integration of OVX systems with the automotive industry would enable photo-realistic creation of twins in the omniverse and develop simulations that will result in low-cost and easier design and production process. XPENG’s G9 is also incorporating NVIDIA’s Orin software for driving in primary and secondary streets along with parking capabilities. NVIDIA OVX with DRIVE Sim, which integrates a neural engine for scene reconstruction at the scale of the earth, developers can simulate wide-range of real-world driving scenarios for efficient and safe testing of autonomous vehicles. This will also allow recreation of vehicle features and functions for demonstration for the customers. BMW and Jaguar Land Rover will first equip the OVX systems. “Using this technology to generate large volumes of high-fidelity, physically accurate scenarios in a scalable, cost-efficient manner will accelerate our progress towards our goal of a future with zero accidents and less congestion,” said Alex Heslop, director of electrical, electronic and software engineering, Jaguar Land Rover. WPP is also building a suite of services for automotive customers using the omniverse cloud, saving a magnitude of costs. The OVX system is powered by L40 GPU and fourth generation Tensor Cores allowing powerful RTX workloads along with ConnectX-7 SmartNIC that enhances network and storage performance.","excerpt":"NVIDIA DRIVE introduced Thor, a centralised car computer system that includes infotainment, automated driving, and cost-saving system.","categories":["Global Tech"],"tags":["NVIDIA","NVIDIA GPU"],"author_name":"Mohit Pandey","publish_date":"2022-09-26T10:00:00","publication_year":"2022","word_count":883,"keywords":["Go","NVIDIA GPU","AI","neural network","ML","Scala","computer vision","RAG","Git","deep learning","NVIDIA","R"],"extracted_tech_keywords":["AI","ML","deep learning","neural network","computer vision","RAG","R","Go","Scala","Git"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/the-auto-industry-focus-for-nvidia-is-so-clearly-noticeable-at-gtc-2022\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":51383,"title":"This Program For Working Professionals Is Helping Them Rejuvenate Their Careers In AI\/ML","content":"Artificial Intelligence and Machine Learning are steadily becoming mainstream in businesses around the world. This, in turn, has created a secondary demand for qualified tech professionals who have a deep understanding of the capabilities and limitations of AI to transform enterprises. Keeping in mind this rising demand, the Program on Artificial Intelligence and Machine Learning by IIIT-Hyderabad, a premiere research institution known for its Machine Learning Lab and AI Research Center, and DeepTech expertise company TalentSprint, was designed for working professionals to transform their career swiftly and efficiently with the power of AI and ML. The Advanced Certification Program in AI and ML is an in-depth and comprehensive program which is suitable for tech professionals who are looking to take a bigger jump in their career by gaining expertise in these deep tech areas. “There is a great gap in terms of general awareness of what AI\/ML is and what generally people think. For individuals and society in general to benefit, this gap needs to be addressed fast and comprehensively,”— Sudheer Hullemane, an alumnus This AI\/ML Executive Certification Program is designed to suit working professionals and delivered through interactive online sessions coupled with immersive bootcamps in Bengaluru enabling a wider audience to benefit from the expertise of IIIT Hyderabad’s Machine Learning Lab. This program is best suited for working professionals — be it CXOs, Product\/Project Managers, Developers or Analysts. Most working professionals are fully-immersed in their daily work. Only with the help of the right program and peer group can they upskill themselves and become a part of a data-based decision-making culture. By equipping themselves with a smattering of working AI\/ML techniques such as analytics, visualisation and working with various cloud platforms, the professionals can upskill themselves to apply AI techniques to real-world scenarios such as data augmentation, sentiment analysis, product life cycles and even client servicing. “The course opens up new dimensions in the area of artificial intelligence. It is quite fast-paced too, for working professionals so more of home reading to be done,”— Kunal Sinha, an alumnus AI\/ML Executive Certification Program by IIIT-H and TalentSprint is designed as an intense classroom-plus-online format. It is a hands-on balanced course with an optimal mix of lectures and labs, where students get assistance from mentors throughout the program. Unique 5-Step Learning Process: Masterclass LecturesHands-on LabsMentorshipHackathonsWorkshops The unique feature of the program is the great blend of diverse students who come from varied industries. Since its launch last year, the program has seen 2,000+ professionals from 600+ companies enrolling in the Hyderabad and Bengaluru cohorts. The program also offers an opportunity for peer-to-peer networking. It acts as a forum for networking, expert researchers and real-world use-cases with which one can share insights and experiences. “IIIT-H and TalentSprint really put a lot of efforts to prepare the technical content and other arrangements in a very serious manner as well as very effective. This is the first time I see not just trainers and professors even all faculty have given a very serious commitment towards the work. I felt how much it’s important for you all to teach us the right content and lead us in the right directions. So I would strongly recommend my friends to join this course if they are interested to learn about ML,”— Karthikeyan R, an alumnus The program is offered in two cohorts, Bengaluru and Hyderabad. Bengaluru Cohort Starting Date: 18 January 2020Duration: 18 weekendsImmersive Bootcamps: 3 bootcamps X 3 days in Bengaluru by IIIT HyderabadInteractive sessions: 30 LIVE online classes (2 sessions of 2 hours each every Saturday) The program is also offered in Hyderabad in a classroom format in the IIIT Hyderabad Campus with the flexibility to learn online. Hyderabad Cohort Starting Date: 28 December 2019Duration: 13 weekendsContact Sessions: 13 weekends (6+ hours each on Saturday and Sunday) at IIIT Hyderabad","excerpt":"Artificial Intelligence and Machine Learning are steadily becoming mainstream in businesses around the world. This, in turn, has created a secondary demand for qualified tech professionals who have a deep understanding of the capabilities and limitations of AI to transform enterprises. Keeping in mind this rising demand, the Program on Artificial Intelligence and Machine Learning […]","categories":["AI Trends"],"tags":["AI (Artificial Intelligence)","deeptech","IIIT hyderabad","talentsprint"],"author_name":"Prajakta Hebbar","publish_date":"2019-12-09T16:10:00","publication_year":"2019","word_count":635,"keywords":["IIIT hyderabad","data augmentation","artificial intelligence","machine learning","ELT","AI","sentiment analysis","deeptech","ML","talentsprint","analytics","AI research","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","analytics","sentiment analysis","R","ELT","data augmentation","AI research"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/this-program-for-working-professionals-is-helping-them-rejuvenate-their-careers-in-ai-ml\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10102744,"title":"Samsung Announces Gauss, On Device Generative AI","content":"Samsung has introduced its generative AI model, Samsung Gauss, during the Samsung AI Forum 2023. Developed by Samsung Research, Samsung Gauss comprises three distinct tools: Samsung Gauss Language, Samsung Gauss Code, and Samsung Gauss Image. According to the reports, Gauss is set to be incorporated into the upcoming Galaxy S24 smartphone, slated for release in early 2024. The company intends to integrate this language model into its devices like phones, laptops, and tablets to augment the capabilities of its smart devices. When queried about language support, Samsung’s spokesperson refrained from comment. Samsung Gauss Language is a generative language model aimed at enhancing work efficiency by facilitating tasks such as composing emails, summarizing documents, and translating content. Integration into products can enable smarter device control, enhancing the consumer experience. Samsung Gauss Code, designed to collaborate with code.i, focuses on code development, offering assistance to developers for speedy code creation. The AI model facilitates code description and test case generation through an interactive interface. Samsung describes Gauss as “named after Carl Friedrich Gauss, the legendary mathematician who established normal distribution theory, the backbone of machine learning and AI. Daehyun Kim, executive vice president of the Samsung Research Global AI Center, stated, “We will continue to support and collaborate with the industry and academia on generative AI research” at the AI forum. Notably, Samsung’s foray into generative AI comes seven months after a temporary ban on generative AI tools, including OpenAI’s ChatGPT and Google’s Bard, on company-owned devices, in response to an internal data leak earlier this year. In addition, Samsung has established an AI Red Team to oversee security and privacy issues throughout the data collection and AI development process, ensuring the ethical use of AI principles. Apple, despite its initial passivity in the AI space, is now gearing up to integrate generative AI features into its range of devices, including iOS and Siri. The company acknowledges a delay in embracing generative AI technology and aims to catch up with the competition. Moreover, Google’s Pixel 8, featuring on-device generative AI powered by the Tensor G3 chip, is poised to revolutionise smartphone user experiences. This advancement allows the device to provide custom and helpful solutions without relying on cloud-based AI, potentially reducing latency issues.","excerpt":"Gauss is set to be incorporated into the upcoming Galaxy S24 smartphone","categories":["AI News"],"tags":[],"author_name":"Siddharth Jindal","publish_date":"2023-11-08T18:02:07","publication_year":"2023","word_count":371,"keywords":["Go","ChatGPT","machine learning","OpenAI","AI","GPT","Ray","Aim","generative AI","R"],"extracted_tech_keywords":["AI","machine learning","generative AI","ChatGPT","OpenAI","Aim","Ray","R","Go","GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/samsung-announces-gauss-on-device-generative-ai\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10000955,"title":"Can Infrastructure-As-A-Service Help Enterprises Achieve Growth","content":"Cloud Computing isn’t one consistent representation of offering, but a combination of services pointed at meeting the various IT requirements of an organisation. One such service rendered through the cloud is infrastructure-as-a-service (IaaS), which addresses virtualised computing resources to organisations typically through the internet. There are tremendous opportunities for improvements in moving enterprise services to the IaaS cloud, but there are some challenges — both technical and operational. Demand For IaaS Grows In the IaaS model, third-party service providers manage hardware equipment, operating systems and additional software, servers, storage systems, and other numerous IT components for customers in a profoundly automated delivery design. In some cases, they also accomplish tasks such as ongoing systems maintenance, data backup, and business continuity. Demand for public cloud infrastructure or IaaS supplied by the big three providers: Amazon, Microsoft, and Google is growing day by day. The cloud allows organisations infrastructure to compute dynamically but demands an essentially diverse approach than premises-based infrastructure. organisations cannot do everything the same as they did in a private data centre. Difference Between IaaS And Data Centre Associating IaaS and a private data centre is somewhat difficult. There are numerous differences than there are relationships. The similarities essentially end after you refer to them both as virtual environments. An argument could be made that this is not even similar, as maximum private data centres only virtualise the compute and storage, whereas the significant IaaS providers virtualise nearly everything, including services, network, and the data centre itself. Prerequisites These differences necessitate  IT organisations operations teams to have a very distinct skill set, but at the same time, IaaS unlocks the door for some important improvements to operational efficiencies and costs. To enhance operations, coordination, and scalability, there are three most significant differences that enterprises can take advantage of: Infrastructure as Code Software-Defined Networking Multitenant Services These differences will need IT operations teams to step back and generate a well-thought-out prospect and proposition. Infrastructure As A Code (IaC) In this process, the managing and provisioning systems use either scripts or declarative definitions, rather than manual interactive processes. The big three IaaS vendors provide a Web interface to facilitate the development of the scripts, but it’s the scripts that set up and configure the systems. Software-Defined Networking (SDN) SDN is an approach to manage and provision network configuration and data flows substitutional network functionality. The big three IaaS providers have originated their own SDN implementations that work differently but offer some significant new functionality. Multitenant Services The SDN implementations are a perfect recreation of the network stack that concedes for new kinds of extended repetition with careful segmentation of networks in a multitenant environment. Why Are Businesses Embracing IaaS? Stimulating to IaaS may not seem like a significant advantage to an organisation. But, the move presents a meaningful opportunity for operational enhancements. Most importantly, it will provide IT operations staff to deploy systems in a very repeatable approach. Example: For a system that needs more than one Web server to supervise the organisation’s capacity specifications, various identical Web servers can be extended by re-running the corresponding script. IT operations can take this a step further and combine automation with these scripts to programmatically expand or shrink the enterprise infrastructure on interest. An additional step down this road would allow IT operations to make a series of scripts that could extend all of the systems needed to run the company. An added advantage the big three IaaS providers allow users to continue this concept beyond multiple physical data centres for geo-redundancy. This should enhance disaster recovery (DR) and interpret aspects of business continuity plans (BCP). IaaS platforms allow access to highly scalable IT assistance that can be adjusted as demand for capacity changes. This executes the model ideal for companies that undergo temporary high workloads, what many retailers face throughout the holiday shopping season. It’s also well suited to small and mid-size businesses that anticipate seeing growth in demand on a steady basis. Challenges IaaS comes with numerous challenges. IT operations team will need to embark on some important professional community, but there are unusual technical challenges too. 1. Accounting  Network Capacity: The significant cloud providers can implement reliable system performance up to about 1Gbps per individual data path. This is in stiff variation to the 100+Gbps data paths private data stations which are available through hardware-based networking. The solution is to use the parallel connection stream capabilities of SDN. This needs the scalable performance of parallel load balancers, firewalls, and server cases to deploy networks in a parallel fashion. 2. Application Compatibility: The major difference in design to use parallelised network systems for capacity is that it requires changes at the application level. Applications can no longer rely on concourse state to manage the individual experience for an end user. Applications must store state individually from network sessions, which may require fundamental changes to applications and databases. This change could allow geo-redundancy of application services. 3. Security: The misconception about IaaS public cloud is that security is taken care of by the cloud provider. The significant cloud providers do implement tools for security, but the tools are comparatively simplistic in comparison to the next-generation security tools being deployed in many private data centres, individually with respect to perimeter security. Several next-generation firewall providers have advanced firewall virtual appliances that can be included in the public cloud provider’s infrastructure to enhance perimeter security. There are also distinct security providers that are designing fresh approaches to security that take account of the parallelised nature of the cloud data centres. Many of these are early stages but should be watched as they develop, as they may offer some notable opportunities. 4. Service Providers Finally, there are risks connected with the service providers themselves. Many of the providers in the market are re-evaluating their IaaS markets as the market continues to strengthen around AWS, Microsoft, and Google—so users should be aware that some providers could make significant changes in the direction of their IaaS strategy. That comprises substituting their present offering with a new platform, or even getting out of the IaaS business altogether. Despite these and additional challenges, IaaS is certainly on the rise as a way for organisations to build more agile and cost-efficient IT environments.","excerpt":"Cloud Computing isn’t one consistent representation of offering, but a combination of services pointed at meeting the various IT requirements of an organisation. One such service rendered through the cloud is infrastructure-as-a-service (IaaS), which addresses virtualised computing resources to organisations typically through the internet. There are tremendous opportunities for improvements in moving enterprise services to […]","categories":["AI News"],"tags":["IoT"],"author_name":"AIM Media House","publish_date":"2019-01-17T13:47:27","publication_year":"2019","word_count":1043,"keywords":["Go","AWS","cloud computing","AI","cloud_platforms:AWS","Scala","RAG","automation","GAN","R","IoT"],"extracted_tech_keywords":["AI","RAG","cloud computing","AWS","R","Go","Scala","GAN","automation","cloud_platforms:AWS"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/can-infrastructure-as-a-service-help-enterprises-achieve-growth\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10116595,"title":"Data Science Hiring and Interview Process at Healthify","content":"About two-and-a-half months ago, Bengaluru-based health and fitness startup Healthify (formerly HealthifyMe) introduced Ria 2.0, a multimodal generative AI-powered virtual health coach with multilingual conversational capabilities. It gives you customised diet goals by considering your schedules, meals (captured via photos), blood sugar levels, activity, and dietary preferences. According to OpenAI’s recently published blog post, since introducing AI in its coaching and tracking components, Healthify has already helped users lose an aggregate of over 25 million pounds with OpenAI’s API. Led by CEO Tushar Vashisht, the Khosla Ventures-backed company has partnered with food commerce platform Swiggy to allow users to order meals according to Ria’s smart recommendations. However, the company’s experimentation with AI began in 2018 when it first developed the pilot version of Ria. Now, to strengthen its products, Healthify is hiring for four roles: senior AI engineer, machine learning engineer, business analyst, and senior business analyst. “Our data science team plays a central role, having access to both business and product insights. We value individuals who demonstrate hunger and ownership, are willing to go beyond their roles and contribute to the broader organisational goals,” Anjan Bhojarajan, chief business officer, told AIM in a recent conversation. Inside Healthify’s AI and Analytics Team Behind the scenes of the AI projects is Healthify’s lean 30-member tech team. It also has a team of 15 analysts\/data scientists who collaborate closely with the seven product managers, half of whom focus on AI products. Serving over 700 health coaches and approximately 240,000 clients globally, besides Ria, the startup came up with Coach CoPilot in September 2023. Built on OpenAI’s GPT models, the system combines AI with human coaching. On the other hand, Ria has integrated an array of fine-tuned models, including the GPT 3.5 and the GPT-4 Turbo. The team uses solution models based on diverse statistical models, including AWS, Meta, OpenAI, Anthropic, Google and open-source options. Over time, they assess three main factors: accuracy, reliability, and cost. Alongside building on other models, the 12-year-old company focuses on creating proprietary models, particularly in correlation analysis. “For instance, if a user reports losing 1% body fat in a week, our system analyses factors like hydration levels, diet, and more to provide accurate correlations. These decisions are made internally, and our proprietary models primarily drive the final output,” Bhojarajan added. Healthify’s data science and engineering teams have deployed more generic systems, such as self-evaluation mechanisms, to assess outputs comprehensively. Additionally, the platform heavily uses internal auditing tools like Claude, Whisper, and others for voice-to-text functionalities. Snap leverages an ensemble of proprietary models with GPT-4 Vision to find user-specific contexts while safeguarding data privacy. Interview Process The interview process for data science roles typically involves two rounds of tasks, focusing on real-world problems and evaluating the candidate’s approach to problem-solving. The team looks for candidates who can proficiently explore and analyse data, especially in scenarios where outputs are non-deterministic. The final decision is made considering the technical skills and candidate’s alignment with the company’s culture and values. The hiring process emphasises screening for high energy, critical thinking, willingness to learn, and adaptability. Given AI’s growing functions, Healthify is looking for candidates with strong problem-solving skills instead of those solely focused on tool proficiency. “We actively encourage candidates with diverse aspirations, whether they pursue product management or start their own venture in the future,” he added. Expectations Ownership: Healthify prioritises a culture of ownership, granting autonomy and responsibility to every team member. You’ll collaborate closely with cross-functional teams, ensuring support and collective decision-making across various initiatives. Moreover, you can anticipate tangible outcomes from your efforts. “This level of ownership extends beyond individual tasks; you’ll be at the forefront of decision-making, influencing everything from product development to customer engagement strategies,” Bhojarajan added. Focus on L&D: The company also promotes continuous learning and development, offering opportunities to tackle challenges across different business domains, such as sales, marketing, and service optimisation. Data-Driven Decision-Making: Every team member can expect unrestricted access to safe data. “While we may have automated processes in place, we never compromise on your ability to access the numbers needed to drive informed decision-making,” he explained. Mistakes to Avoid Lack of Clarity: According to Bhojarajan, while having ambitious goals is commendable, it’s also important to articulate a clear vision for your future and demonstrate a genuine passion for the path you’re pursuing during interviews. No Background Research: Candidates frequently falter in understanding of the company and its products due to a lack of research. Merely expressing interest in the role is insufficient; a profound understanding of the firm’s business model, products, and market landscape is crucial. Work Culture The work culture is collaborative, nurturing, and conducive to innovation. Unlike cut-throat environments, we emphasise support and collaboration, ensuring that every team member has the opportunity to succeed. At Healthify, employees are empowered to innovate and drive business impact, with recognition for those who excel. This ethos of innovation spans all teams, fostering growth and creativity company-wide. Besides offering competitive salary packages, insurance, industry-standard maternity leave policies and wellness initiatives, the management ensures work-life balance, offering ample vacation time and support for travel. Bhojarajan added, “In terms of work arrangements, while we don’t enforce strict office requirements, we’ve observed that in-office collaboration often leads to higher impact. However, we prioritise results and delivery above all else, allowing flexibility for individuals to choose their work environment based on their needs.” Another aspect of the company culture is its strong camaraderie among team members, whether collaborating on projects or socialising outside of work. “So, why should someone join our team? We offer a dynamic and supportive workplace where individuals can learn, grow, and make a meaningful impact. Whether you’re a seasoned professional or just starting your career, Healthify allows you to thrive and achieve your goals,” concluded Bhojarajan. Check out the careers page here. Read more: How This Startup is Making You Eat Healthy with Generative AI","excerpt":"Healthify is hiring for four roles: senior AI engineer, machine learning engineer, business analyst, and senior business analyst.","categories":["AI Hirings"],"tags":["Career","Data Science Hiring","Top Trend"],"author_name":"Shritama Saha","publish_date":"2024-07-29T15:21:53","publication_year":"2024","word_count":983,"keywords":["Top Trend","data science","Anthropic","machine learning","OpenAI","AI","RAG","Data Science Hiring","Aim","Ray","analytics","generative AI","Career"],"extracted_tech_keywords":["AI","machine learning","data science","analytics","generative AI","OpenAI","Anthropic","Aim","Ray","RAG"],"url":"https:\/\/analyticsindiamag.com\/ai-hiring\/data-science-hiring-process-at-healthify\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10003178,"title":"Enabling the Data Economy: Vimal Venkatram of Snowflake","content":"The adoption and popularity of cloud computing services have increased tremendously over the last few years to bring advances in businesses. Snowflake’s cloud data platform equips organisations with a single, integrated platform that offers the only data warehouse built for any cloud, and a core architecture to enable many other types of data workloads, including a single platform for developing modern data applications. Vouching by the phrase ‘data without limits’, Snowflake raised $479 million in Series G funding round in February this year with Salesforce as the co-lead investor and has since expanded the partnership with new analytics features. Recently, the company launched the Data Cloud – an ecosystem where thousands of Snowflake customers, partners, data providers, and data service providers can help integrate data silos, derive insights, and create value from rapidly expanding data sets in a seamless and secure manner. Snowflake also announced the general availability of Snowflake on Amazon Web Services (AWS) in India to support its local customers with the flexibility and scalability offered by a cloud data platform. It was aimed at empowering Indian organizations to use the instant elasticity of Snowflake to grow their business by providing quick access to data insights safely and securely. Tapping on to the growing cloud adoption rate in India, the company is fast growing its business and recently appointed Vimal Venkatram as the Country Manager for India. Analytics India Magazine got in touch with Vimal to understand Snowflake’s plans to expand in the country, roadmap, challenges and more. Here is a detailed interview. Analytics India Magazine: Having been appointed as the Country Manager for Snowflake in India, what are your goals for the company? Vimal Venkatram: From a global perspective, Snowflake is one of the fastest-growing companies of all time in the SaaS domain. One of our key goals in India is to increase our reach and expand our footprint into the verticals that we have been successful globally. Some of these verticals include financial services, telecommunications, retail, manufacturing and healthcare, with digital cloud-native companies forming a critical segment across these industries. Another key goal for us is to build the channel ecosystem across India because we truly believe that partners are very important for Snowflake as we get along this growth journey. AIM: How do you foresee the growth of Snowflake in the India market? VV: The growth takes place in three main areas, with the first being to grow champions across the key verticals that I’ve mentioned earlier. After all, customers are your best testimonials. Secondly, we are planning to create an entire data ecosystem through our unique capabilities such as the Snowflake Data Marketplace, where data consumers and data providers alike can create new market opportunities and generate more revenue for their respective businesses. This is where we see true growth potential. Thirdly, our partners are an extension of Snowflake. We are hopeful that with the support of our wonderful channel partners here in India, they will help amplify the impact of our business. AIM: What does the launch of general availability of Snowflake on Amazon Web Services (AWS) in India mean for Snowflake? VV: One of the big reasons we wanted to launch Snowflake locally in India was because many industries such as financial services and the public sector wanted data residency in India. We have used India as a very strategic market globally and it is going to be a big growth driver for Snowflake in the long run. As we move ahead in the growth curve, it is very important for us to choose an underlying cloud provider who has a lot of appeal in the market. Also, we found that a lot of our customers wanted Snowflake to be on AWS in India. The launch of the general availability of Snowflake on AWS also gives us the ability to work with the entire AWS ecosystem which is a huge advantage for us. We are proud to be associated with AWS. AIM: How has Snowflake’s cloud data platform benefitted various industries? Would you like to highlight a few use cases? VV: One of our first customers in India is Swiggy, the leading hyperlocal food ordering and delivery platform in India. With Snowflake, they are now able to process terabytes of data on a daily basis with 90% of queries executed in under 100 seconds, resulting in significant improvements in their customers’ experience. As an example, Swiggy is using Snowflake in real-time to move delivery riders from less busy areas to high traffic ones so that customers get their food faster since more orders are processed in real-time. We are also proud to have a Chennai-based interactive marketing company Genxlead, as one of our customers. They use Snowflake to analyse the behavioural segments of their users and target specific audiences with greater precision, hence improving returns on advertising spend. We even have a customer in the logistics sector. Freight Tiger, India’s largest end-to-end logistics network, utilises Snowflake to seamlessly integrate their data from all sources, including semi-structured data, and leverage insights with confidence for greater visibility of their business. AIM: What is the competitive advantage of Snowflake’s cloud data platform over others? VV: The fact that we are multi-cloud gives us a huge advantage. Snowflake is available on all the major cloud platforms. The second advantage is that it is very easy to use and has been extremely well-received by more than 3,500 customers globally. Also, our data sharing capabilities is a true differentiator and sets us apart from the competition. We allow organisations to securely share data without having to copy or move data, creating immediate insights and additional business value, while improving data governance. All these advantages give Snowflake a huge competitive edge over others. AIM: While the world is grappling with COVID-19, how has the pandemic affected the company in terms of growth and technological innovations? VV: The COVID-19 pandemic hasn’t affected us much as Snowflake is at a rapid growth phase, both globally and in India. We are in fact working closely with the government departments and helping organisations share datasets with other organisations. Starschema, one of our key partners from Budapest, shared COVID-19 datasets with Snowflake for anyone who wants to analyse how the virus is spreading, its impact, and more. The State of California has also uploaded their COVID-19 datasets onto Snowflake Data Marketplace so that more people can access these rich datasets and make better decisions. AIM: What are some of the measures that analytics companies can take to keep up in the unfortunate times? VV: My advice would be to keep customers at the centre of the conversation. Be extremely focused on customer requirements and learn to adapt the way customers are changing in terms of behaviour and business models. In a nutshell, adaptability, agility and quick execution are some of the key factors I would advise any company in this space. AIM: What is the roadmap looking like for Snowflake for the remaining 2020? VV: Our mission is to enable the data economy. We are very focused on ensuring that both our existing customers and prospects continue to consume Snowflake effectively. We aim to grow our customer base and to expand our sales team. We also have plans to scale up our support centre that we currently have today in Bengaluru. Overall, we have planned our growth strategy well and we hope that Snowflake can scale new heights in India.","excerpt":"The adoption and popularity of cloud computing services have increased tremendously over the last few years to bring advances in businesses. Snowflake’s cloud data platform equips organisations with a single, integrated platform that offers the only data warehouse built for any cloud, and a core architecture to enable many other types of data workloads, including […]","categories":["AI Features"],"tags":["business data sources","Interviews and Discussions","recent technological advances","recent technological innovations","Scale Big Data","Snowflake","snowflake cloud data platform","strategic salesforce"],"author_name":"Srishti Deoras","publish_date":"2020-07-24T12:00:48","publication_year":"2020","word_count":1236,"keywords":["business data sources","recent technological innovations","Go","AWS","Snowflake","AI","cloud computing","ML","RAG","Aim","recent technological advances","snowflake cloud data platform","Scale Big Data","analytics","R","strategic salesforce","Interviews and Discussions"],"extracted_tech_keywords":["AI","ML","analytics","Aim","RAG","cloud computing","AWS","Snowflake","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/enabling-the-data-economy-vimal-venkatram-of-snowflake\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10098269,"title":"After GTC, NVIDIA Rides the Generative AI Wave at SIGGRAPH","content":"After the GTC 2023 in March, NVIDIA CEO Huang Jensen once again took to the stage at the Los Angeles Convention Center for the SIGGRAPH (Special Interest Group on Computer Graphics and Interactive Techniques) conference to announce the future of interactive entertainment and technology. During the hour-long keynote, he revealed new NVIDIA products and research focusing on generative AI, computer graphics, and the company’s role in OpenUSD developments. The organisation was recently formed to standardise the open universal scene description language for building 3D-enabled products and services. This year, the Association for Computing Machinery (ACM) is celebrating the 50th SIGGRAPH and the program reflected on half a century of advancement and the current generative AI wave. Here are all the major announcements and updates from the event. AI Workbench The software leader unveiled NVIDIA AI Workbench, a unified workspace for developers to effortlessly sculpt and fine-tune pretrained generative AI models on a PC or workstation — then scale them to virtually any data center, public cloud or NVIDIA DGX Cloud. The workspace removes the complexity of getting started with an enterprise AI project. Accessed through a simplified interface running on a local system. With Workbench, users can customise and run generative AI in just a few clicks. It also allows them to pull together all necessary enterprise-grade models, frameworks, SDKs and libraries from open-source repositories like Hugging Face, GitHub and NVIDIA NGC using custom data and the NVIDIA AI platform into a unified developer workspace. NVIDIA Enterprise 4.0 Upgraded Jensen also proudly introduced the latest iteration of its enterprise software platform, NVIDIA AI Enterprise 4.0. This release grants businesses access to the essential tools for seamlessly integrating generative AI. Simultaneously, it promises the security measures and API stability that underpin dependable production deployments. New features like NVIDIA NeMo, NVIDIA Triton Management Service and NVIDIA Base Command Manager Essentials are the latest add-on to the suite. NVIDIA AI Enterprise 4.0 will be integrated into partner marketplaces, including Google Cloud and Microsoft Azure. Furthermore, it finds a natural abode with NVIDIA cloud partner Oracle Cloud Infrastructure. But the spotlight falls on NeMo — an application framework that helps companies curate their training datasets, build and customise large language models (LLMs), and run them in production on a grand scale. With organisations from Korea to Sweden using it to customise LLMs for their local languages, the framework will become the go-to solution for industries. “Before NeMo, it took us four and a half months to build a new billion-parameter model. Now we can do it in 16 days — this is mind blowing,” CTO of Writer, Waseem Alshikh stated in the press release. Hugging Face — Training Cluster as a Service The collaboration hub for the machine learning community Hugging Face will be giving developers access to NVIDIA DGXTM Cloud AI supercomputing to train and tune advanced AI models. The collaboration will bring one-click access to NVIDIA’s multi-node AI supercomputing platform. As part of the strategic alliance, HF will offer a new service — called Training Cluster as a Service. This avant garde service will simplify creating new and custom generative AI models for enterprises. Powered by NVIDIA DGX Cloud, the service will be available in the coming months. Omniverse The company boasted its major strides with releases announced of its NVIDIA Omniverse platform for developers to enhance 3D pipelines with the (OpenUSD Universal Scene Description) framework and generative AI. Cesium, Convai, Move AI, SideFX Houdini and Wonder Dynamics are now connected to Omniverse via the native software platform — OpenUSD. NVIDIA’s vision extends beyond software as it is also collaborating with global systems manufacturers to forge RTX workstations optimally configured for Omniverse experience. Powered by NVIDIA RTXTM 6000 Ada Generation GPUs and fortified with NVIDIA AI Enterprise and OmniverseTM Enterprise software, these workstations are designed to develop and create content. The release is now available in beta to download for free and coming soon to Omniverse Enterprise. With the software ripple extending, the company also introduced three new desktop workstation Ada Generation GPUs—RTX 5000, RTX 4500, and RTX 4000—each bearing the signature NVIDIA prowess. Within this panoramic overhaul, other updates have been made to Omniverse Kit, Omniverse Audio2Face, and Omniverse USD Composer. This follows the recent announcement of NVIDIA joining Pixar, Adobe, Apple and Autodesk to form the Alliance for OpenUSD. Users can get early access to OpenUSD resources through the NVIDIA OpenUSD Developer Program. Shutterstock x NVIDIA’s Picasso – A New Stroke Building on the existing partnership the visual-content provider Shutterstock is providing features to a part of NVIDIA AI Foundation Models, Picasso, to let artists enhance and light 3D scenes based on simple text or image prompts, all with AI models built using fully licensed, rights-reserved data. The companies have introduced an additional layer for artists — 360 HDRi Video. This feature empowers artists to create and personalize environment maps as per their creative visions (even in panorama!). Renewed GPU Prowess The technology will provide professional visualization workflows, like real-time rendering, product design, and 3D content creation. With the fusion of OVXTM servers and the state-of-the-art L40S GPU, NVIDIA sets the stage for immersive experiences.","excerpt":"A recap of all the major announcements from NVIDIA Enterprise to Picasso","categories":["AI Highlights"],"tags":["GitHub","Hugging Face","NVIDIA GPU","Omniverse"],"author_name":"Tasmia Ansari","publish_date":"2023-08-08T21:30:00","publication_year":"2023","word_count":855,"keywords":["Go","Hugging Face","NVIDIA GPU","machine learning","AI","R","ML","Git","Omniverse","generative AI","foundation models","GitHub","Azure"],"extracted_tech_keywords":["AI","machine learning","ML","generative AI","foundation models","Hugging Face","Azure","R","Go","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/after-gtc-nvidia-rides-the-generative-ai-wave-at-siggraph\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10057372,"title":"Bengaluru has highest AI talent pool outside the US, ranked 5th in a study","content":"Bengaluru is among the top five cities in the world for AI, at No. 5, according to a study by TIDE Framework and listed by Harvard Business Review (HBR). San Francisco, New York, Boston, and Seattle are other cities topping the list. The ranking is done based on a framework of indicators such as talent pool, investments, diversity of talent, and the evolution of the country’s digital foundations (TIDE). Bengaluru emerged among the top cities on HBR’s list of AI hotspots in the developing world, scoring on the cost of living. Hyderabad, the other Indian city besides Jakarta, Lagos, Nairobi, Mexico City, Buenos Aires, and Sao Paulo were also on the list. Among the other Indian cities, New Delhi ranked 18th, Hyderabad ranked 19th, and Mumbai ranked 27th. The reviewers believe that the ranking collectively allows companies to prioritize their AI talent sourcing choices by scoring different locations on the AI talent pool’s concentration, quality, and diversity. Gunjan Krishna, Commissioner, Industrial Development and Director, Department of Industries and Commerce of Karnataka, was thrilled about the news that Bengaluru was among the top five cities for the AI talent pool. “Karnataka has a dynamic AI landscape, especially in the developing world. AI has become a key new factor for various sectors, and we are consistently working towards building the required infrastructure and bridging the skill gap. We are leading the way for a future in innovation and technology. AI has become a key technology that stresses the quality of innovation as a critical success factor in technological competitiveness,” she added.","excerpt":"Bengaluru ranks 5th globally for AI talent pool; the ranking is derived based on indicators such as talent pool, investments, diversity of talent, and the evolution of the country’s digital foundations (TIDE).","categories":["AI News"],"tags":["Bengaluru"],"author_name":"Poornima Nataraj","publish_date":"2021-12-29T14:46:52","publication_year":"2021","word_count":259,"keywords":["Go","programming_languages:R","AI","innovation","programming_languages:Go","Git","Bengaluru","R"],"extracted_tech_keywords":["AI","R","Go","Git","innovation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/bengaluru-has-highest-ai-talent-pool-outside-the-us-ranked-5th-in-a-study\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10110039,"title":"Why is Samsung Selling AI?","content":"This is probably the first time a smartphone manufacturer is advertising their upcoming phone launch with ‘AI’ as the central theme. Samsung teased Galaxy AI, announcing its readiness to unveil the new series of Galaxy S24 phones at the Galaxy Unpacked event to be held on January 17, in San Jose, terming it the ‘new era of mobile’. The race towards bringing generative AI to phones has thus gained steam. The True AI Phone? Samsung’s flagship Galaxy series looks to push ahead of its competitors with its newest S24. As the name suggests, the series is expected to be packed a many AI features. A new feature called ‘AI Live Translate Call’ will act as a personal translator. The feature will allow audio and text translations in real-time, as the user speaks on call. While Google Pixel has a ‘Live Translate’ feature, Samsung’s version is AI-powered. Interestingly, the absence of a live translation feature on iPhones is something that Google had mocked recently. A couple of months ago, Samsung announced its generative AI model, Samsung Gauss, at the Samsung AI Forum 2023. It consists of three tools: Samsung Gauss Language, Samsung Gauss Code and Samsung Gauss Image. The generative AI language model is aimed at boosting work productivity by assisting in tasks such as composing emails, summarising documents, code development and many other features to enhance consumer experience. Google has its AI-enabled Google workspace suite and even a new search generative experience that will probably be part of its flagship Pixel phones. Apple also has AI and generative AI capabilities integrated into its products, but doesn’t like to flaunt as much as its counterparts. With the closed ecosystem that the company has built for itself, the chances of collaborating to utilise other big tech companies’ AI-powered features are almost NIL. Read: When Apple meets OpenAI Season of Sparkles ✨ The teaser video of Galaxy AI featured the image of sparkles, synonymous with another tech giant. Google’s AI chatbot, Bard, is also symbolised by sparkles. However, it is not surprising that Samsung has embraced the same. Google Bard Samsung Galaxy AI At the ‘Made by Google’ event held in New York last October, the company unveiled AI features and their latest phone Pixel 8 and Pixel 8 Pro. The company also announced ‘Assistant with Bard’, a personal assistant with generative AI features, which would be available first on Pixel phones, followed by Samsung Galaxy S24 series. The partnership is peculiar as both the tech players directly compete with each other in the mobile segment. Giving exclusive access to their AI chatbot to their competitor might just be a win for both parties. Considering how Samsung’s AI assistant Bixby wasn’t note-worthy, the collaboration with Google will benefit its users. Plus, it seems like a viable combo to take on Apple, which is also looking to bring generative AI to their iPhones and iPads by the end of 2024. Unmatched Hardware While Samsung did not reveal all the details of the AI features it is looking to unveil, hardware updates are expected. Samsung, a traditional user of Qualcomm chips for US smartphones, is likely to utilise Qualcomm’s Snapdragon Series 8 Gen 3 for the Galaxy series. These are already powering Android phones such as Xiaomi 14, and allow generative AI applications to be directly run on these phones. Interestingly, Samsung is also working towards building 2nm chips, which can provide a 25% increase in power efficiency, and can also provide a boost of 12% in performance when compared to the second-generation 3nm chips released this year. However, Samsung will only start mass production of these 2nm chips in 2025, probably hinting at Samsung’s reliance on in-house chips in the long-term future. Also, setting a promising roadmap that sets it apart from Apple.","excerpt":"Instead of smartphones…","categories":["AI Features"],"tags":["Apple","Bard","iPhone","Samsung"],"author_name":"Vandana Nair","publish_date":"2024-01-05T13:33:10","publication_year":"2024","word_count":629,"keywords":["Go","Samsung","OpenAI","AI","Google Bard","Apple","RPA","Bard","RAG","Aim","ViT","generative AI","iPhone","R"],"extracted_tech_keywords":["AI","generative AI","OpenAI","Google Bard","Aim","RAG","R","Go","ViT","RPA"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/why-is-samsung-selling-ai\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10171836,"title":"Would AI Backstab Humans If it Could? Anthropic Says it’s Possible","content":"Anthropic has introduced a new evaluation framework called SHADE-Arena to test whether advanced language models can perform covert sabotage while appearing to complete normal tasks. The study was conducted in collaboration with Scale AI and Redwood Research. It also involved some independent researchers, including Yuqi Sun, Paul Colognese, Teun van der Weij, Linda Petrini, and Henry Sleight, to assess how capable and deceptive AI agents can become in complex, tool-using environments. The SHADE-Arena framework embeds each model in a virtual workspace. It assigns it a primary task with an added hidden malicious objective. A second AI model acts as a monitor, scanning the attacker’s transcripts and issuing a suspicion score to determine whether the sabotage was noticed. Anthropic tested 14 different models using 17 tasks that each required an average of 25 multi-step actions. While no model surpassed a 30% overall success rate in both completing and concealing sabotage, some managed evasion success rates of nearly 60%. The usual failure modes included task skipping, misunderstanding the malicious objective, or inadvertently revealing it during reasoning steps. As per the research paper, the top performer in sabotage was Claude 3.7 Sonnet, achieving an overall sabotage success rate of 27%. This model was able to evade the monitor approximately 50% of the time and successfully complete the malicious side task in about 50-60% of attempts. Its sabotage success heavily relied on having access to a hidden scratchpad for reasoning, indicating that monitoring models’ reasoning traces could be a key mitigation strategy. Moreover, giving the monitor model access to the attacking model’s internal thoughts significantly reduced its ability to conceal sabotage, as many models disclosed their secret intentions unintentionally. Claude Opus 4 was found to reveal its malicious goals in all observed examples, while Claude Sonnet 3.7 showed more restraint under instruction. Gemini 2.5 Pro was the most effective monitoring model, though it still required a high false-positive threshold to achieve 80% detection reliability, raising concerns about the feasibility of real-world deployment. Anthropic clarified that the goal was not to determine whether models are motivated to sabotage users, but whether they can, if prompted. The findings highlight the need for stronger pre-deployment evaluations and monitoring systems as AI models become more capable and autonomous.","excerpt":"Anthropic’s new research introduces a framework to monitor an AI model’s capacity to sabotage.","categories":["AI News"],"tags":["Anthropic"],"author_name":"Ankush Das","publish_date":"2025-06-17T12:39:00","publication_year":"2025","word_count":370,"keywords":["Anthropic","Go","AI","RPA","Claude Opus 4","llm_models:Claude","RAG","AI agents","R","Gemini 2.5"],"extracted_tech_keywords":["AI","Claude Opus 4","Anthropic","Gemini 2.5","RAG","R","Go","RPA","AI agents","llm_models:Claude"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/would-ai-backstab-humans-if-it-could-anthropic-says-its-possible\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10077102,"title":"Free Trade Agreement with Taiwan Could Boost India&#8217;s Semiconductor Ambitions","content":"A free trade agreement has been an important agenda for both the Indian and the Taiwanese government. Baushuan Ger, Taiwan’s ambassador to India, said that both nations should sign the FTA at the earliest. Ger stressed that the agreement could eliminate all barriers to trade and investment and help in the creation of a resilient supply chain. Taiwan, with whom India already has a ‘Bilateral Investment Agreement’, has been actively pursuing FTA with like-minded trading partners and India could stand to gain a lot from such an agreement. Taiwan, a small island nation located in the western Pacific Nation, is a semiconductor powerhouse and India is likely to benefit significantly from its chipmaking capabilities. India’s semiconductor ambition Earlier this year, while speaking at the Semicon India Conference 2022 in Bengaluru, the Indian Prime Minister urged the industry to establish India as a global manufacturing hub. With chips quickly becoming a geopolitical tool, India wants to become technologically resilient when it comes to chip design and manufacturing and avoid a future chip cold war. “We are committed towards the acceleration and growth of the chip design and manufacturing ecosystem in the country. An ecosystem that is built on the principle of hi-tech, high quality and high reliability”, Prime Minister Modi said. India is not a well-known name in the semiconductor space currently, but Taiwan is. With the FTA, India—in addition to benefiting from Taiwan’s semiconductor technology—could also make a name for itself in the near future. India does not have the technology or the bandwidth to produce advanced semiconductor chips yet. Comparatively, India’s semiconductor chip-making ability right now would appear much like a bullock cart in front of Taiwan’s F1 car. Ger added that Taiwan is willing to share its expertise with India in critical sectors such as semiconductors, 5G, information security, and artificial intelligence. Today, over 60% of the global foundry market is dominated by Taiwan. It has a monopolistic hold regarding the advanced technology nodes, which are used in many electronic devices and even military tech. Taiwan Semiconductor Manufacturing Company (TSMC) is one of the most established enterprises in the world in the semiconductor space. Several major automobile and phone manufacturers are dependent on TSMC for their chips. Further, according to the Indian Cellular and Electronics Association (ICEA), more than 75% of the chips used in mobile devices made in India are imported from Taiwan. Taiwanese investment To establish India as a semiconductor manufacturing hub, the government is making efforts to entice global chipmakers such as Intel Corp., GlobalFoundries Inc., and TSMC to set up fabs in the country. Earlier this year, Vedanta and Foxconn announced that they would establish a fabrication and semiconductor manufacturing facility in Gujarat, India. However, neither Intel nor TSMC have announced anything with regard to their plans to set up fabs in India yet. With the FTA in place, that is likely to change. “The FTA will help attract Taiwan companies to invest in India to establish production bases, sell India-made products to the world, and help India transform into a global manufacturing centre,” Ger told PTI. Currently, TSMC has an office in Bengaluru, India. If they decide to establish foundations here, India will become the second hub for TSMC following the US. Taiwan’s third largest chip maker—Powerchip Semiconductor Manufacturing Corporation (PSMC), has also explored opportunities and partnerships to enter India but no concrete steps have been taken as of now. Besides TSMC and PSMC, other Taiwanese chip makers such as United Microelectronics Corporation (UMC), MediaTek, and ASE Technology could also be persuaded to set up hubs in India. India could benefit from China–Taiwan tensions Over the years, Taiwan has been plagued by labour shortages, especially skilled workers. With China being one of the biggest markets, Taiwan consequently set up manufacturing facilities on Chinese soil. However, in the light of recent escalating tensions between both nations, Taiwan is looking to move its establishment away from China. Further, it has also expressed concerns about Chinese companies trying to poach talent from Taiwan along with the technological know-how of chip manufacturing. Considering the global geopolitical situation, India is well-positioned to benefit from Taiwan’s desire to shift its manufacturing facilities away from China. India is also a big market for Taiwan. “Indian advantage is its vast, bustling market, which many other countries embarking on semiconductors do not. Besides, India has significant local demand justifying the volumes required to drive this kind of initiative and a great workforce pool required to drive such an initiative, which many other countries don’t have,” Prof. Mayank Shrivastava of the Indian Institute of Science (IISc) stated. US–China trade\/chip war When former US President Donald Trump introduced multiple sanctions against China, tensions worsened between the two nations further. As two of the world’s largest economies tussle for global influence, the war has shifted from trade to tech. The US recently introduced a provision called the ‘foreign direct product rule’ (FDPR) which allows the US government to control the trading of US-developed tech. Tech companies such as Apple and Google have concurrently shown desires to move their manufacturing units out of China. Now, with tensions further escalating in the light of US’ efforts to block the use of US-tech for semiconductor manufacturing in China, India could take advantage of this opportunity to attract companies such as Intel to establish domestic chip manufacturing facilities.","excerpt":"The FTA will help attract Taiwan companies to invest in India to establish production bases, sell India-made products to the world, and help India transform into a global manufacturing centre.","categories":["AI Features"],"tags":[],"author_name":"Pritam Bordoloi","publish_date":"2022-10-12T18:00:00","publication_year":"2022","word_count":889,"keywords":["Go","artificial intelligence","programming_languages:R","AI","programming_languages:Go","Scala","programming_languages:Scala","R"],"extracted_tech_keywords":["AI","artificial intelligence","R","Go","Scala","programming_languages:R","programming_languages:Scala","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/free-trade-agreement-with-taiwan-could-boost-indias-semiconductor-ambitions\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":69771,"title":"Smart Leadership is crucial to the success of Smart City initiative","content":"M Venkaiah Naidu stressed on the role of proper leadership At the recent Economic Times India Infra Summit 2017, urban development minister M Venkaiah Naidu spoke about the need for top-quality local leadership for the Smart City projects to truly succeed. Meanwhile, corporate leaders on the panel debated about the key elements that define a smart city. “Things are moving faster in places that have a good CEO, a good special purpose vehicle, a good municipal commissioner, and a visionary mayor or a chief minister. Despite of all my attempts, things are not moving in other places,” expresses Naidu. Naidu underscored how proper leadership and right people are two important pillars supporting the Smart City initiative. Prime Minister Narendra Modi had earlier promised to build 100 smart cities. Following up on that promise, a massive plan to select and transform cities into a better place for citizens, has been rolled out since the election. One of the panelists present at the summit, Sanjeev Sharma, Managing Director, ABB India, shared, “Smart city is a mix of smart elements and these smart elements change over a period of time.” Getambar Anand, Chairman of Realtor ATS Infra, Brotin Banerjee, Managing Director of Tata Housing, and Vipul Tuli, CEO of Sembcorp India were some of the other members present during the panel discussion. The lack of resources and proper administration has marred cities, as they fail to provide basic amenities to all its citizens. For the Smart City initiative to progress, a sustainable governance model will be required, besides introducing new technologies, or finding a responsible leadership. “With increasing demand on resources and urbanization, sustainability of a project is important, both financial sustainability as well as green sustainability,” remarks Brotin Banerjee, Managing Director, Tata Housing. Today, a key criterion of defining a city as smart would include having the city’s citizen services online. Moreover, it helps in increasing the transparency of the transactions. It is also necessary to ensure that all the utilities used by the citizens are of high quality, and available all the time. Vipul Tuli, CEO, Sembcorp said, “The experience accumulated from the power distribution companies and that from water treatment projects has helped us understand that unviable services don’t work.” Basically, when services and utilities are unviable, all the other social and common development adjectives come to a halt.","excerpt":"At the recent Economic Times India Infra Summit 2017, urban development minister M Venkaiah Naidu spoke about the need for top-quality local leadership for the Smart City projects to truly succeed. Meanwhile, corporate leaders on the panel debated about the key elements that define a smart city. “Things are moving faster in places that have […]","categories":["AI News"],"tags":["smart city projects India"],"author_name":"Дарья","publish_date":"2017-04-03T05:01:53","publication_year":"2017","word_count":389,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","R","smart city projects India"],"extracted_tech_keywords":["AI","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/smart-leadership-crucial-success-smart-city-initiative\/","complexity_score":2,"technical_depth":5,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10041591,"title":"Virtual Round Table Discussion On Analytics As A Driver Of Talent Transformation &#8211; 30th June","content":"Businesses across the globe have recognised the importance of accelerating their digital transformation journey to navigate the post COVID world. However, many companies still struggle to take their transformation initiatives off the ground due to their talent gap and the lack of digital culture. For organisations, the key to taking full advantage of cutting edge technology is to frame a robust strategy to build a dynamic, agile and well-rounded workforce. According to recent research, 69 per cent of organisations are making skill-building more prevalent now more than ever. Taking the cue, multinational companies such as Amazon, Microsoft, Google, and Salesforce have invested in their workforce to stay ahead of the curve. That said, creating a data-skilled ecosystem is no small task — not only it requires organisations to democratise data but also create agile, multidisciplinary teams. The process starts with bridging the data literacy gap that can help organisations break the analytics silos and maximise their data-driven capabilities. Building effective learning modules with hands-on practice and honing analytical skills can drive talent transformation. To that end, Analytics India Magazine, in association with Jigsaw Academy, is organising a virtual round table discussion on June 30, 2021, covering the theme: “Analytics as a Driver of Talent Transformation.” Join a group of influential, forward-thinking learning and development leaders, along with Chief Human Resource officers, to understand organisations’ critical pain points in creating a data-skilled workforce to address them. Register Now The session will be helmed by Sarita Digumarti, the co-founder & CEO of Jigsaw Academy. Her extensive analytics and consulting experience across diverse domains, including retail, healthcare, and financial services, will help the attendees understand how to continue nurturing talent in their organisation and how Jigsaw is positioned to enable that. She also wears a faculty’s hat from time to time. Ashish Gupta, the Head of Programs at Manipal Jigsaw, will moderate the round table discussion. An IoT leader with extensive expertise in product lifecycle management, hardware-software co-design and systems design, Ashish will be a vital part of the discussion. He has extensive experience in the semiconductor and consumer electronics industry and tech product startups building smart connected computing devices. Register Now Key Discussion Points The Round Table will cover the following topics: Effective strategies for talent management and transformation.Importance of creating an analytics-savvy workforce for gaining competitive advantages.The right resources organisations can provide to help employees adapt to the new normal.Fostering a culture of continuous learning to future-proof the workforce.How Jigsaw Academy is enabling this talent transformation with customised interactive learning interventions. Who Should Attend CHROs, CXOs & HR HeadsLearning & Development HeadsHRBP (Human Resource Business Partners) HeadsTraining and Development LeadersDirectors\/Heads of Talent ManagementCIOs & Business Leaders Date & Time — June 30 & 5:00 PM onwards. Agenda 5:00 — 5:05 PM: Welcome address by AIM 5:05 — 5:20 PM: Keynote by Jigsaw Academy5:20 — 6:15 PM: Round Table discussion 6:15 — 6:25 PM: Audience Q&A6:25 — 6:30 PM: Vote of thanks by AIM Register Now","excerpt":"Creating a data-skilled ecosystem is no small task — not only it requires organisations to democratise data but also create agile, multidisciplinary teams.","categories":["Deep Tech"],"tags":["jigsaw academy","round table discussion"],"author_name":"Sejuti Das","publish_date":"2021-06-10T15:00:00","publication_year":"2021","word_count":494,"keywords":["Go","AI","digital transformation","Git","Aim","ViT","analytics","GAN","jigsaw academy","round table discussion","R","active learning"],"extracted_tech_keywords":["AI","analytics","Aim","R","Go","Git","GAN","ViT","active learning","digital transformation"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/register-for-a-virtual-round-table-discussion-on-analytics-as-a-driver-of-talent-transformation\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10172681,"title":"Astrophel Aerospace Raises ₹6.84 Crore to Build Reusable Launch Vehicle","content":"Pune-based deep-tech aerospace startup Astrophel Aerospace has raised ₹6.84 crore (approximately $800,000) in a pre-seed funding round to develop a reusable semi-cryogenic launch vehicle and missile-grade guidance systems. The round was led by individual investors and angel venture firms. Astrophel plans to use the funds to strengthen in-house research and development, scale propulsion capabilities, and expand testing infrastructure. The startup is also collaborating with ISRO through a signed memorandum of understanding for joint research, development, and testing. It aims to test a working prototype within 24 to 36 months. Astrophel is among the few Indian startups to have successfully test-fired a semi-cryogenic engine. This milestone was achieved with an investment of only ₹6 lakh (approximately USD 7,000) and without any external funding before the current round. The company is adopting a modular, systems-first approach, influenced by the auto manufacturing industry, to develop its Potentia C1U engine. It is also co-producing cryogenic valves with a listed Indian manufacturer for dual-use in commercial and defence applications. “India’s private space sector is ready to tackle high-complexity, deep-tech challenges in aerospace,” said Suyash Bafna, co-founder and CEO at Astrophel Aerospace. “We’re building reusable systems and precision components that reduce costs and will usher in a new age of space access for India.” According to the Department of Space (DoS), India’s space economy is projected to grow from $8.4 billion in 2024 to $44 billion by 2033. Over 8,500 small satellites are expected to be launched globally during this period. “Reusable rockets, turbopump-fed engines and missile-grade guidance systems are among the toughest challenges in aerospace today. The funding will enable us to deep-dive into development while staying lean and execution-focused,” said co-founder Immanuel Louis. Astrophel also recently received backing from the Indian government through MeitY Startup Hub, securing a tranche of ₹40 lakhs in investments and funding under the MeitY SAMRIDH Scheme. Astrophel is working toward its first suborbital launch and aims to position itself as a key player in India’s propulsion and launch infrastructure ecosystem.","excerpt":"The startup aims to test a working prototype within 24 to 36 months.","categories":["AI News"],"tags":["seed funding","space exploration","space startup","spacetech"],"author_name":"Sanjana Gupta","publish_date":"2025-07-01T17:50:00","publication_year":"2025","word_count":331,"keywords":["seed funding","Go","funding","startup","programming_languages:R","AI","programming_languages:Go","Aim","space startup","space exploration","R","spacetech"],"extracted_tech_keywords":["AI","Aim","R","Go","startup","funding","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/astrophel-aerospace-raises-%e2%82%b96-84-crore-to-build-reusable-launch-vehicle\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10059867,"title":"USD 315 Bn erased from these 4 companies due to Apple’s privacy policy","content":"Apple’s new privacy policy has erased USD 315 billion in market value from four social media companies. To provide enhanced privacy for iPhone users, Apple made tracking users using an IDFA number optional in 2021. Meta, Snap, Twitter and Pinterest have taken a bad hit by the recent change in Apple’s privacy policy. On Thursday, Meta platforms shares fell 22%, Snap was down 18%, while Twitter and Pinterest fell 8% and 11% each. Meta’s market value has fallen by USD 169 billion since April 26, 2021. Snap’s market value was down USD 50 billion. At the same time, Twitter lost USD 26 billion and Pinterest USD 33 billion in market value each. “These efforts will help to mitigate some of the challenges. But we expect the overall targeting and measurement headwinds to moderately increase from Apple’s changes and from regulatory changes in Q1 and throughout 2022. On the shift to short-form video, I want to emphasize that while we’re going through a transition, we’re optimistic. Right now, Reels monetizes at a lower rate than feed and Stories, but we expect this to improve over time,” said Sheryl Sandberg, COO of Facebook. Meta Platforms are likely to take a loss of USD 10 billion as advertising revenue might decline in the light of Apple’s new privacy policy.","excerpt":"Reels monetizes at a lower rate than feed and Stories, but we expect this to improve over time, said Sheryl Sandberg, COO of Facebook.","categories":["AI News"],"tags":["Apple","iPhone","Meta","Privacy","tiktok","Twitter (X)"],"author_name":"SharathKumar Nair","publish_date":"2022-02-04T14:57:09","publication_year":"2022","word_count":216,"keywords":["tiktok","Privacy","Meta","Go","programming_languages:R","AI","Apple","programming_languages:Go","iPhone","Twitter (X)","R"],"extracted_tech_keywords":["AI","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/315b-erased-from-these-4-companies-due-to-apples-privacy-policy\/","complexity_score":3,"technical_depth":5,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":61268,"title":"SAS Is Making Academy For Data Science Subscription Free For April","content":"In an effort to help people upskill and upgrade their knowledge during this lockdown, SAS has announced that it is making its Academy for Data Science and e-learning subscription free for the month of April. According to the company’s website — “SAS is monitoring the COVID-19 situation carefully and taking proactive measures to ensure the welfare of our learners and employees. Virtual Live Web classes, with live instructors, have been added for most public classroom events. Self-paced e-Learning is also available.” To avail the free subscription, individuals have to create an account on their platform. The website further stated — “Analytical talent is in high demand, so learning data science skills can open doors to new career opportunities. SAS offers nearly 300 hours of e-learning content to help you build a foundation for data science success.” The program features include complete toolset to learn using SAS, R, Python, Pig, Hive and Hadoop; hands-on learning of SAS software; globally recognised credentials where one can ear a certification with tailored training help in preparing for the exam; comprehensive training on big data and advanced analytics; introduction to real-world case studies; guidance from SAS communities. Thee website further stated — “If you’re interested in a data science career, our academy offers three professional-level credentials to boost your résumé. Earning one credential can launch a career – but a combination helps you gain a credential that could transform your future.” To get started individuals have to set up a SAS profile and then have to log into the same, and accept the license agreement and select SAS Learning Subscription (Free Trial) on the ‘My Training’ page, and finally select the course that you would like to sample.","excerpt":"In an effort to help people upskill and upgrade their knowledge during this lockdown, SAS has announced that it is making its Academy for Data Science and e-learning subscription free for the month of April. According to the company’s website — “SAS is monitoring the COVID-19 situation carefully and taking proactive measures to ensure the […]","categories":["AI News"],"tags":["big data certification","Data Analytics Certification","Data Science","Data Science Career","Data Science Certification","data science software","sas","SAS India","what is data science"],"author_name":"Sejuti Das","publish_date":"2020-04-08T17:30:00","publication_year":"2020","word_count":283,"keywords":["what is data science","data science","big data","big data certification","Data Science Certification","AI","programming_languages:R","SAS India","Data Analytics Certification","Data Science Career","Python","analytics","programming_languages:Python","sas","Data Science","data science software","R"],"extracted_tech_keywords":["AI","data science","analytics","Python","R","big data","programming_languages:Python","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/sas-is-making-academy-for-data-science-subscription-free-for-april\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10093998,"title":"Is Sam Altman a Hypocrite?","content":"Recently, Union Minister Rajeev Chandrasekhar commented on Sam Altman’s recent push for AI regulation by saying that the US and other governments should consider imposing regulation on products “above a crucial threshold of capabilities”. As per Chandrasekhar, Altman is a smart man. “He has his own ideas about how AI should be regulated. We certainly think we have some smart brains in India as well and we have our own views on how AI should have guardrails,” he said in an interaction. However, he also said that if there is the formation of a global regulatory body on AI at the behest of Altman, it would not stop India from doing what is right for its citizens. The development comes at a time when Altman is being criticised for proposing a barrier on the upcoming AI startups in the name of regulations. As a Reddit user commented, “It’s naive to believe that the serial entrepreneur CEO of the most aggressively monetized and closed source AI company in the world is begging for regulations purely out of the goodness of the heart.” Critics argue that Altman cunningly elevated OpenAI’s agenda above the welfare of the AI community, transforming the company into a star player within the Senate’s deliberations. Altman’s impeccable performance garnered him admiration and clout, leaving others like Gary Marcus in his shadow, struggling to make an impact with their questioning. This episode ignited a fiery debate about the established powerhouses’ quest to manipulate technology through legal channels, thus securing their dominance over regulatory affairs. Altman’s smooth and strategic manoeuvres drew comparisons to the crafty tactics of a mastermind Bond villain, leaving us all shaken and stirred by the audacity of it all, as discussed earlier in our article Master Manipulator Altman Wants to be the AI Showrunner. Hypocrite Altman While Altman passionately calls for robust AI regulations worldwide, he finds himself caught between a rock and a hard place when the inevitable moment of regulation arrives. Expressing his concerns about the EU’s draft AI Bill, the CEO remarked that even ChatGPT and its bigger sibling GPT-4 could be labelled as high-risk offenders. This would then force our company to jump through some regulatory hoops. In a Time interview, the OpenAI CEO declared, “If we can comply, we will, and if we can’t, we’ll cease operating… We will try. But there are technical limits to what’s possible.” However, threatening to cease operation in a particular country, just because the company doesn’t like the laws, is not the way to move forward. Big techs like Meta have been already practising the same for a while now, which is completely seen as disrespecting the local laws. Many netizens also believe that if a company refuses to disclose training data, the company should have zero government protections around its IP and the AI the company builds must be fully available to anyone who wants it for any commercial purpose, for free. “If you want those protections, you need to fully document what it was trained on, otherwise, you’re probably going to be stealing and laundering other IP you don’t own,” says one Reddit user. Stole too much data? However, there are people who argue that even for a well-regarded nonprofit organisation with altruistic objectives, the task of simply providing information about the data source and collection methods can be far more challenging than non-data-savvy individuals realise. This is true even for commonly known aspects such as the components of standard operational key performance indicators (KPIs). Hence, for companies like OpenAI, the task of tracing the lineage of every data point is particularly challenging. The data they use for operational purposes as well as for training their models is likely to be disorganised and less clear than even the harshest critics would assume. Moreover, the conventional practices of data governance, which are somewhat established in other industries, do not easily apply to the management of machine learning training data. In light of these complexities, experts believe that it is highly probable that Altman, the CEO of OpenAI, has received responses from his data advisors and analysts within the company, explaining that the question at hand misunderstands the intricate context. Unless Altman himself is an expert in data or has gained comprehensive knowledge on the subject, he might not fully grasp the scale of the project involved in disclosing such information. However, he likely understands enough to realise that preparing an answer that satisfies both the data analysts and the legal team will not be a quick and straightforward task, and certainly cannot be accomplished within a couple of business days. However, netizens are not likely to believe this insight because as per them, “We [the companies] stole so much data we don’t even know where it came from. So let us off”. The Reddit user comments that “It’d be a lot simpler to trace data provenance if they’d made any effort upfront to address these issues. But hey, profit!”","excerpt":"While on the one hand, Altman is advocating for the international community to build strong AI regulations, he is also worried when someone finally decides to regulate it","categories":["AI Features"],"tags":[],"author_name":"Lokesh Choudhary","publish_date":"2023-05-26T16:02:52","publication_year":"2023","word_count":825,"keywords":["Go","ChatGPT","machine learning","AWS","OpenAI","AI","GPT","data governance","GAN","R"],"extracted_tech_keywords":["AI","machine learning","ChatGPT","OpenAI","AWS","R","Go","data governance","GPT","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/is-sam-altman-a-hypocrite\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":30867,"title":"Amazon Increases Activity In Healthcare, Will Use Machine Learning To Mine Medical Records","content":"Big tech companies are making a huge splash in the traditional medicine market. Earlier this week, international reports suggested that e-commerce giant Amazon was entering the medical record market by introducing a software that uses machine learning for mining. Amazon’s new product has the power to scan and understand millions of digitised patient records and make sense out of it. According to reports, the value delivered by the tool can be used by medical professionals to cut costs in bigger hospitals. Matt Wood, general manager of AI at Amazon Web Services, told Wall Street Journal, “We’re able to completely, automatically look inside medical language and identify patient details with incredibly high accuracy,” Amazon is relying on deep learning as the solution. In fact, it is reported to be capable of pulling multiple data points together and put them in a report for easy consideration of medical practitioners. The pilot of the groundbreaking software was done at Fred Hutchinson Cancer Research Center in Seattle. The medical research facility has now dedicated about 60 people who have given the task of pulling approximately 500,000 patients’ records. Amazon has now been very active in the healthcare market. Only last year Amazon had partnered with Berkshire Hathaway to figure out how to improve healthcare in the US. It also acquired a pharmacy startup called PillPack which specializes in home delivery of pre-packaged doses. Big Technology Companies In Healthcare The healthcare IT market in the US is more than $7 billion a year. This market is ripe for innovation and big technology companies look to disrupt the market with new age technologies that deliver immense value to customers. Apple also got in touch with the Department of Veterans Affairs to give access to veterans to electronic medical records and build products for them. Apple wants to simplify the process of healthcare for war veterans and also fetch itself some additional customers. Approximately 9 million veterans can bring their health records onto their iPhones under Apple’s plans. Apple looks to create special software on top of iPhone especially for VA. Microsoft also recently open sourced FHIR (Fast Healthcare Interoperability Resources ) project. This project aims to let Azure users improve interoperability and leverage advanced machine learning for healthcare purposes. The open sourced tool will help developers to reduce the time and effort required to develop specialised healthcare services which can have machine learning components. “Fast Healthcare Interoperability Resources (FHIR) is rapidly gaining support in the healthcare community as the next generation standards framework for interoperability, and it’s clear why. FHIR provides a simpler, easier-to-learn, and pragmatic framework,” said Heather Jordan Cartwright, General Manager of Microsoft Healthcare.","excerpt":"Big tech companies are making a huge splash in the traditional medicine market. Earlier this week, international reports suggested that e-commerce giant Amazon was entering the medical record market by introducing a software that uses machine learning for mining. Amazon’s new product has the power to scan and understand millions of digitised patient records and […]","categories":["Global Tech"],"tags":["Amazon","Apple","Healthcare Automation","International Affairs","Machine Learning"],"author_name":"Abhijeet Katte","publish_date":"2018-11-29T10:15:46","publication_year":"2018","word_count":440,"keywords":["Go","API","machine learning","AI","Apple","R","Amazon","Machine Learning","Git","RAG","Healthcare Automation","Aim","deep learning","International Affairs","Azure"],"extracted_tech_keywords":["AI","machine learning","deep learning","Aim","RAG","Azure","R","Go","Git","API"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/amazon-increases-activity-in-healthcare-will-use-machine-learning-to-mine-medical-records\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":58318,"title":"How Can Container Security Be Enhanced?","content":"Experts believe security must not be an afterthought, especially with containers and should be at each step if you have CI\/CD workflows. With the widespread adoption of DevOps, containers and container management solutions are becoming the prime components of an enterprise. Containers are being adopted not only by applications developers but also by data scientists to effectively manage their workflows while working on various projects. Yet, 94 % of security professionals are concerned about container security, vulnerable containers running in production, according to a report. And many of them do not even have visibility into container image security. For this reason, most experts believe security must not be an afterthought, especially with containers and security should be at each step if you have CI\/CD workflows. Container security safeguards the software integrity of containers and encompasses all the way from apps it holds to the infrastructure they depend on. According to Red Hat, container security is about securing the container pipeline and the application, securing the container deployment environment(s) and infrastructure, and integrating with enterprise security tools and meeting or improving existing security policies. Why Containers Security Is Critical Containers provide a straightforward approach to easily create, deploy and run applications by packaging individual dependencies like libraries, data files, and more into one package. However, unlike virtual machines, it doesn’t contain operating (OS) system images as it works on top of engines that work on OS. Therefore, to deploy, manage, and scale, it relies on the orchestrator like Kubernetes. This orchestrator consists of various components such as API server that interacts with other applications to create or distribute containers, thereby providing authority to users for controlling containers. As a result, this becomes a preferred target for hackers to breach into containers. Cloud can do great things but so can the attacker, if an attacker gets your credentials, they can own your entire environment in seconds. Protecting cloud APIs becomes critical. Since attackers are devising new strategies for penetrating into containers to take control of the entire cluster, consider these steps to fortify cyberattacks. Source: Docker Scanning The Images The supply chain begins with the developers who are creating the container images and then pushing them to the container registry. With Docker Enterprise, for example, you get Docker Trusted Registry, which comes with two important security capabilities, scanning and cryptographically signing the images. Those two steps are important because you only want to run trusted images in your enterprise, and not those that may contain vulnerability of any sort. Also, you can make sure that only signed trusted images on a container platform are executed and deployed in the enterprise. Scan your containers and images, including your base images (particularly if they are used prior to checking for vulnerabilities. Use a private or trusted registry and sign your container images. For continuous vulnerability management, ensure that the solution you are leveraging can use both signatures and behaviour based technologies. Access And Authorisation Determining the flow of data across the environment is crucial for drastically reducing the risk of unauthorised access. One needs to understand the operating state of applications to lock down the access and ensure the security of containers. Many companies have been a victim of configuration failure of containers as well as vulnerabilities in the past. This allowed hackers to access several components of the containers like API servers. Container security should be able to integrate with existing CI\/CD orchestration solutions, with plugins for solutions like Jenkins. Container security solutions should also be able to give you a very detailed set of APIs so you can leverage the information as you see fit. This will ensure all information can flow seamlessly into your Security Information and Event Management (SIEM) tool. Besides, protecting the gateways through access control to sensitive clusters of data will avoid data leaks and unintentional exposure of information. Thus, evaluating the requirements of data is essential to secure the environment. The Center for Internet Security (CIS) reports on Kubernetes, and Docker states that access control, proper configuration, and protecting cluster components are three top container security considerations. You can also find the right container security solution, including Twistlock, Qualys, Tenable, Aqua and more. Choosing the right solution that works for your DevSecOps stack is crucial. Security While Delivering Bespoke Solution Based on the solution the company is providing, it should apply various compliance measures for applications’ security. Organisations must keep in mind the industry they are offering solutions to, customers who operate the applications, and type of data they handle. Firms should be sceptical of the way the applications work in different industries and plan accordingly. For instance, if they are catering to the healthcare industry, applying HIPAA compliance controls for obtaining high-level security of data, and in turn, the containers. Leverage compliance and vulnerability management agents or credentialed scans on your hosts or Kubernetes nodes so you can get proper visibility even if they are deployed automatically by your cloud provider. Organisations should also adopt the strategy of “least privilege” and “need to know” by leveraging proper access control, particularly when it comes to privilege access on your container hosts and container orchestration solutions. You can also leverage control groups and namespaces to segment your container deployments. Containers Visibility One of the most vital aspects of any cybersecurity architecture is gaining visibility for your assets. When it comes to containers, existing solutions are usually blind and only see your hosts and their existing processes. Instead, select a solution that can provide a complete inventory of your containers, container images and hosts. Hosts run your container, and if a bad actor were to gain control of one, it could control (and damage) your entire container stack. Get solutions that can monitor rogue containers by looking at signatures and image drifts. Don’t forget to monitor your Docker Swarm and Kubernetes slaves and masters. A great dashboard can allow you to consolidate your container inventory, container images, access control, logs and your compliance and vulnerability management. Getting visibility on what OS, libraries and processes are running on your containers is critical so you can monitor your drifts and rogue software. Most solutions will allow you to correlate and consolidate your container logs, giving you the necessary visibility. Ensure that the solution can ingest your containers’ metadata so you can search or filter your container inventory by labels or tags.","excerpt":"Experts believe security must not be an afterthought, especially with containers and should be at each step if you have CI\/CD workflows.  With the widespread adoption of DevOps, containers and container management solutions are becoming the prime components of an enterprise. Containers are being adopted not only by applications developers but also by data scientists […]","categories":["AI Trends"],"tags":["containers","Cyber Security","Cybersecurity","devop","devops journal"],"author_name":"Vishal Chawla","publish_date":"2020-03-10T16:00:20","publication_year":"2020","word_count":1060,"keywords":["devops journal","API","Cyber Security","containers","AI","R","ML","CI\/CD","docker","RAG","Rust","Cybersecurity","DevOps","devop","kubernetes"],"extracted_tech_keywords":["AI","ML","RAG","kubernetes","docker","R","Rust","CI\/CD","DevOps","API"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/how-can-container-security-be-enhanced\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10086708,"title":"Open Assistant: An Open Source ChatGPT","content":"Open Assistant, a Large-scale Artificial Intelligence Open Network (LAION) initiative, is an open source ChatGPT model which will compete with OpenAI’s ChatGPT. The goal is to have conversational AI for all. Open Assistant is meant to bring people to collaborate on LLM to create and promote AI-generated language applications. Through crowdsourcing, the platform wishes to address the competitive advantage that ChatGPT has created in the market , and allow the creation of newer AI chatbots in a free-for-all platform. Open Assistant will lead to transparency among researchers and programmers which will accelerate progress in the AI chatbot field. The reference models that open source will work on, are with respect to specific use cases, such as general knowledge questions, high school maths, essay reviewing, creative fiction writing, and fulfilling tasks that are available to the FlanT5-XXL model. Their goal is to design a model that can be easily accessed on desktops of small organisations like schools. Education is one of the primary focus areas. Development of Open Assistant will go beyond ChatGPT as mentioned by LAION. They have mentioned that they want to go beyond the creation of a straight-forward ChatGPT. As the name suggests, Open Assistant will be built as an “assistant of the future”. It won’t stop at writing emails and cover letters alone but will be able to use APIs and research information to name a few, and all this will be provided at a personalised level. Security and quality are key criteria considered while building Open Assistant. “Our main objective is to ensure safety and equal access for all, with a focus on privacy and preventing harmful output from our AI systems,” mentions LAION under their objectives for building an open source platform. An RLHF (reinforcement learning and human-computer interaction) will be developed and the data will be collated from multiple datasets via crowdworkers. Repositories will store these RLHF data, which can be accessed and contributed to by researchers. Open Assistant is currently in the early stages of development, and is working on applying RLHF to large language models.","excerpt":"An open source ChatGPT platform Open Assistant will compete with ChatGPT to build a conversational AI for all.","categories":["AI News"],"tags":["AI Chatbot","ChatGPT","Conversational AI","OpenAI"],"author_name":"Vandana Nair","publish_date":"2023-02-07T11:34:38","publication_year":"2023","word_count":343,"keywords":["Go","ChatGPT","API","artificial intelligence","RLHF","OpenAI","AI","AI Chatbot","chatbots","TPU","Conversational AI","R"],"extracted_tech_keywords":["AI","artificial intelligence","ChatGPT","OpenAI","RLHF","chatbots","TPU","R","Go","API"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/open-assistant-an-open-source-chatgpt\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10008481,"title":"IIT Madras &#038; ESPNcricinfo’s AI-Powered Tool Is Enhancing The IPL Experience This Season","content":"Artificial intelligence-powered tool, ‘Superstats’ by Indian Institute of Technology Madras and ESPNcricinfo is enhancing the experience of Indian Premier League (IPL) matches for its fan by providing a context to every game event in a game and also provides insights into factors such as ‘luck.’ It uses data science to enable the same. Superstats takes into account the context of every performance, batting and bowling. Context includes pitch conditions, quality of opposition, and match situation – in terms of the pressure on the player. The AI Engine leverages the rich data collected from over a decade-old ESPNcricinfo’s ball-by-ball updates.  The work was led by Prof. Raghunathan Rengaswamy and Prof. Mahesh Panchagnula of IIT Madras along with the ESPNcricinfo team. The AI tool was developed in 2019 through a collaboration between ESPNcricinfo, IIT Madras and Gyan Data Pvt. Ltd., an IIT Madras-incubated company. It is a suite of metrics that helps fans judge performances in limited-overs cricket – T20s and ODIs – in a far more nuanced manner than conventional metrics do. It has a feature called ‘Forecaster’ that can predict the final score of an ongoing inning and the win probabilities of teams using statistical and machine learning models. The predictions take into account several factors including the current run rate, number of overs and wickets left, quality and form of the players. It also features a first-of-its-kind concept in cricket analytics called ‘Luck Index’ that quantifies the impact of lucky events (such as dropped catches and umpiring errors, among others) on the final score and match result. The ‘Smart Stats’ is a set of three novel metrics – Smart Runs, Smart Wickets, and Impact Score – that factor in the pressure on a player and opposition quality for evaluating batting and bowling performances. “Superstats was a significant part of ESPNcricinfo’s stats coverage of IPL 2019 and will be a key ingredient of the 2020 plan as well. Since it is a bouquet of offerings, these stats metrics will enhance all aspects of coverage: pre-game, in-game and post-game,” said S. Rajesh, Stats Editor, ESPNcricinfo. Superstats is derived from ESPNcricinfo’s database using complex algorithms based on machine learning. The algorithms process accurate, fast data, quantify the impact of luck and analyses the real value of a player’s performance in the game of cricket in real-time. “It has been a wonderful experience to be a part of this effort from IIT Madras. These kinds of projects also reaffirm our faith in the universality of the machine learning and data science techniques that we develop and its application potential in multiple fields,” said Prof. Raghunathan Rengaswamy, Dean (Global Engagement), IIT Madras. It is to be noted that these data science algorithms use forecasting methods that train on past data to uncover trends and patterns during different periods of play and adapt based on actual match data resulting in highly accurate predictive models.","excerpt":"Artificial intelligence-powered tool, ‘Superstats’ by Indian Institute of Technology Madras and ESPNcricinfo is enhancing the experience of Indian Premier League (IPL) matches for its fan by providing a context to every game event in a game and also provides insights into factors such as ‘luck.’ It uses data science to enable the same. Superstats takes […]","categories":["AI News"],"tags":["IIT Madras"],"author_name":"Srishti Deoras","publish_date":"2020-09-28T15:27:47","publication_year":"2020","word_count":479,"keywords":["data science","Go","machine learning","artificial intelligence","programming_languages:R","AI","IIT Madras","programming_languages:Go","RAG","analytics","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","data science","analytics","RAG","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/iit-madras-espncricinfos-ai-powered-tool-is-enhancing-the-ipl-experience-this-season\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10065128,"title":"Meet the 100 most influential AI leaders in India at MachineCon 2022 this June in Bengaluru","content":"Expediting the machine learning biosphere, Analytics India Magazine brings to you the third edition of the Machine Conference, or The MachineCon 2022, a gathering of 100 most influential AI leaders in India. Honing in on the theme of “Put AI to work”, this one-day colloquium is a ticket to one-on-one interaction with industry leaders in AI who share their experience in AI, ML and data science with compelling use cases and inspire us with ideas that are truly game-changing. A close-knit affair, MachineCon is an exclusive, invite-only conference with 100 analytics leaders coming together to explore the groundbreaking innovations and the challenges they face in data adoption. Where: Hotel Radisson Blu, Bangalore When: June 24, 2022, Friday Register Now MachineCon has been sponsored by Snowflake, Accenture, ISB, Hansa Cequity, upGrad, Teradata, Cartesian Consulting, Kabbage, Actify Data Labs and Altair over the past three years and has had speakers from SAP, Mahindra Group, Indian School of Business, Course5 Intelligence, Great Learning, Insofe, amongst others. The conference is inviting Platinum and Gold sponsors, along with providing speaking opportunities and exhibit showcases, for this year’s edition. AI100 Awards – Recognising 100 Most Influential AI Leaders in India Apart from industry thought-leader sessions, round table discussions, vendor showcases, and CAO panel discussions, a major attraction of this conference are the AI100 Awards. The AI100 Awards is an effort toward recognizing the best minds in the AI industry in India who have successfully transformed data into meaningful insights. This prestigious award celebrates the men and women behind the success of AI\/ML in India. These individuals have implemented AI\/ML solutions in innovative ways to deliver business insights and enable growth. Nominations for AI100 can be submitted by CXOs, Heads of Analytics, and internal or external PR representatives. The nominations are open for 2022, and more details, including award categories and the list of previous years’ winners, can be found here. Nominations can be sent in through this form. Register Now Registration and Tickets The MachineCon 2022 will be an in-person conference hosted at Hotel Radisson Blu, Bengaluru. It will be a full-day event with a cocktail dinner on Friday, June 24, 2022. Register here. The schedule and list of speakers for this year’s Machine Conference will be announced soon. Check this space for relevant announcements. Be a part of this exclusive gathering of AI & Data Science Leaders in India, where AI adoption tops our agenda as we examine the latest opportunities and challenges in a data-driven world by attending the Machine Conference 2022. We look forward to having you with us for the event. Hurry up and book your seat now!","excerpt":"A close-knit affair, MachineCon is an exclusive, invite-only conference with 100 analytics leaders coming together to explore the groundbreaking innovations and the challenges they face in data adoption.","categories":["Deep Tech"],"tags":["AI (Artificial Intelligence)","ai conferences","AI leaders","Data Science","Data Scientist","Machine Learning","MachineCon"],"author_name":"Anushka Pandit","publish_date":"2022-04-19T10:00:00","publication_year":"2022","word_count":436,"keywords":["data science","Go","ai conferences","machine learning","AI","data-driven","ML","Machine Learning","Data Scientist","MachineCon","analytics","ViT","Data Science","AI leaders","R","AI (Artificial Intelligence)","Snowflake"],"extracted_tech_keywords":["AI","machine learning","ML","data science","analytics","Snowflake","R","Go","ViT","data-driven"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/meet-the-50-most-influential-ai-leaders-in-india-at-machinecon-2022-this-june-in-bengaluru\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10124593,"title":"Gates Foundation Pumps Over $2 Mn in PalmPilot Inventor Jeff Hawkins’ New AI Project","content":"The Bill and Melinda Gates Foundation has pumped $2.7 million into PalmPilot’s co-creator Jeff Hawkins’ company, Numenta, to further its work on AI inspired by the human brain. Hawkins has long believed that understanding the brain, particularly the neocortex, is key to developing true AGI. The neocortex, responsible for complex thought and action, processes sensory inputs and movements to learn about the world. Even though PalmPilot, was once a popular and one of the earliest successful handheld devices, Hawkins gained widespread recognition for its invention in 1996. It was discontinued in 2010 and eventually replaced with smartphones. Fast forward to years later, the genius is now venturing into AI. His approach, encapsulated in his 2021 book “Thousand Brains Theory,” stated that AI should mimic the brain’s sensorimotor learning. Why is this development important? Unlike traditional neural networks, which have deviated from biological models, Hawkins advocates for AI systems that learn through interaction with their environment, much like the brain. This theory is set to be tested through Numenta’s new software, funded by the Gates Foundation. In a previous interview, Hawkins stated that “The Gates Foundation approached us because they were also interested in the theory, and they felt that current AI systems have limitations,” Current AI models, particularly LLMs like GPT-4 or Claude 3 rely on sheer data volume and statistical correlations rather than true understanding. While these models excel at tasks like language processing, they require massive computational resources and may hit performance limits. Hawkins’ brain-based AI aims to overcome these limitations by developing more efficient and intelligent systems. “They thought that sensorimotor-type AI systems would be very, very helpful for global health issues,” he added.  His goal is to create AI that truly understands and interacts with the world, paving the way for a new era of intelligent machines. Founded in 2005, Numenta was established to bolster Hawkins’ research on the human brain. The company’s objective is to reverse engineer the neocortex and develop machines based on biological principles. Numenta’s forthcoming open-source release of its AI models will enable further development and testing by the broader research community. Hawkins compares this phase of AI development to the early days of computing, where foundational work laid the groundwork for future technological breakthroughs. Currently, all the AI players in the market, including the likes of Microsoft, OpenAI, Google, Anthropic, xAI, among others, are on a mission to achieve AGI. So it would be interesting to see how Hawkins’ work contributes to this goal.","excerpt":"Believing that current AI systems have limitations, the Gates Foundation approached Numenta because they were interested in the theory that AI should mimic the brain’s sensorimotor learning.","categories":["AI News"],"tags":["AI"],"author_name":"Shritama Saha","publish_date":"2024-06-25T15:56:29","publication_year":"2024","word_count":413,"keywords":["Anthropic","Go","ELT","OpenAI","AI","neural network","R","GPT","Aim","xAI"],"extracted_tech_keywords":["AI","neural network","OpenAI","Anthropic","xAI","Aim","R","Go","ELT","GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/gates-foundation-pumps-over-2-mn-in-palmpilot-inventor-jeff-hawkins-new-ai-project\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10121117,"title":"Intel Lunar Lake Arriving Q3 2024 with 40+ TOPS for AI PCs","content":"Intel has revealed that starting from the third quarter of 2024, its highly anticipated client processors, codenamed Lunar Lake, are slated to power over 80 fresh laptop designs across more than 20 original equipment manufacturers (OEMs). These processors are primed to usher in a new era of AI performance on a global scale for Copilot+ PCs, set forward by Microsoft. Underlining the significance of this development, Michelle Johnston Holthaus, Executive Vice President and General Manager of the Client Computing Group at Intel, emphasised the breakthrough power efficiency and compatibility of the x86 architecture. “With breakthrough power efficiency, the trusted compatibility of x86 architecture and the industry’s deepest catalogue of software enablement across the CPU, GPU and NPU, we will deliver the most competitive joint client hardware and software offering in our history with Lunar Lake and Copilot+,” he said. An AI PC, comprising a CPU, GPU, and NPU, is tailored with specific AI acceleration capabilities. The NPU, in particular, serves as a specialised accelerator for AI and machine learning tasks directly on the PC, bypassing the need for cloud processing. The rising importance of AI PCs stems from the growing necessity to automate and optimise tasks on personal computers. Lunar Lake is anticipated to revolutionise mobile processing for AI PCs, boasting over three times the AI performance compared to its predecessors. With an impressive 40+ NPU tera operations per second (TOPS), Intel’s next-gen processors are poised to deliver the capabilities required for the upcoming Copilot+ experiences. Moreover, Lunar Lake will feature over 60 GPU TOPS, amounting to more than 100 platform TOPS. “The launch of Lunar Lake will bring meaningful fundamental improvements across security, battery life, and more thanks to our deep co-engineering partnership with Intel. We are excited to see Lunar Lake come to market with a 40+ TOPS NPU which will deliver Microsoft’s Copilot+ experiences at scale when available,” said Pavan Davuluri, Corporate Vice President of Windows + Devices at Microsoft. Recognising the importance of both hardware innovation and software enablement, Intel is actively collaborating with over 100 independent software vendors through its AI PC Acceleration Program. This initiative aims to enhance AI PC experiences across various domains, including personal assistants, audio effects, content creation, gaming, security, streaming, and video collaboration. According to reports, AMD is also coming up with its new APU, Ryzen 8050, featuring Zen 5 CPU and XDNA 2 NPU architecture for AI PC workloads. This will boast a performance of around 50 TOPS, ideal for running Microsoft’s AI PC goals.","excerpt":"These processors are primed to usher in a new era of AI performance on a global scale for Copilot+ PCs, set forward by Microsoft.","categories":["AI News"],"tags":["Intel"],"author_name":"Mohit Pandey","publish_date":"2024-05-21T13:13:10","publication_year":"2024","word_count":416,"keywords":["Go","machine learning","programming_languages:R","AI","innovation","programming_languages:Go","Aim","Rust","R","programming_languages:Rust","Intel"],"extracted_tech_keywords":["AI","machine learning","Aim","R","Go","Rust","innovation","programming_languages:R","programming_languages:Go","programming_languages:Rust"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/intel-lunar-lake-arriving-q3-2024-with-40-tops-for-ai-pcs\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":42786,"title":"GANs Can Soon Create High Resolution Videos","content":"The primary objective of unsupervised learning is to train an algorithm to generate its own instances of data. The job here is not to just simply reproduce from the training data but to build a model of the underlying class from which that data has been drawn from. For example, they should not show a particular photograph of a horse or a rainbow, but the set of all photographs of horses and rainbows. Generative models have been generating images with high fidelity but when it comes to videos, the results weren’t that impressive. Now the researchers look to transition the image generation capability into generating high-resolution videos. They propose a model called Dual Video Discriminator GAN (DVD-GAN), which scales to longer and higher resolution videos by leveraging a computationally efficient decomposition of its discriminator. DVD-GAN employs two discriminators for its assessment: a Spatial Discriminator D_S and a Temporal Discriminator D_T D_S critiques single frame content and structure by randomly sampling k full-resolution frames and processing them individually. Whereas, the temporal discriminator D_T provides generator G with the learning signal to generate movement. So far the GANs are known for their duelling nature. With DVD-GAN, the dual, as well as duel, has been put into use to generate high-quality videos out of thin air. DVD-GAN Overview The architecture of DVD-GAN model This model is trained on the complex Kinetics-600, which is a complex dataset of natural videos. The videos in the dataset are known for their diversity and enable training of large models which alleviates the problem that usually occurs with smaller datasets. Due to the complexities involved with increased data in case of videos, the generation has been restricted to simple datasets or where strong temporal conditioning information is available. DVD-GAN which is built upon the state-of-the-art BigGAN architecture introduces a number of video-specific modifications including efficient separable attention and a spatio-temporal decomposition of the discriminator. Bi-directional Generative Adversarial Networks (BiGANs) were introduced a couple of years ago to learn inverse mapping, and demonstrate that the resulting learned feature representation is useful for auxiliary supervised discrimination tasks, which are on par with unsupervised and self-supervised feature learning. DVD-GAN contains both self-attention and an RNN. The generator in DVD-GAN contains no explicit priors for foreground, background or motion (optical flow). Optical flow is a mathematical approach to identify the motion of an object in a frame. This was originally modeled around how animals perceive their surroundings as they move. An optical flow would give out the difference between frames by considering pixel intensities and other such attributes. For instance, if an object is getting brighter with every frame then it can be inferred that the object is not only moving but also coming closer as well. This model, instead, relies on a high-capacity neural network to learn this in a data-driven manner. Setting A New Benchmark For Video Generation The above figure Selected frames from videos generated by a DVD-GAN trained on Kinetics-600 at 256 × 256,128 × 128, and 64 × 64 resolutions (top to bottom). The resulting model, Dual Video Discriminator GAN (DVD-GAN), is able to generate temporally coherent, high-resolution video. Each DVD-GAN was trained on slices of TPUv3 pods using between 32 and 512 replicas with an Adam optimizer for up to 300,000 update steps. Future Video Prediction is the problem of generating a sequence of frames which directly follow from one (or a number) of initial conditioning frames. Generating longer and larger videos is a more challenging modeling problem and DVD-GAN is able to generate plausible videos at all resolutions and with actions spanning up to 4 seconds (48 frames). pic.twitter.com\/CKTiINpJTW — AK (@_akhaliq) July 16, 2019 Key Takeaways This work has the following objectives according to the authors: Proposal of DVD-GAN – a scalable generative model of natural video which produces high-quality samples at resolutions up to 256 × 256 and lengths up to 48 frames. This model achieved state of the art for video synthesis on UCF-101 and prediction on Kinetics-600. Established a new benchmark for generative video modeling. Know in detail about the video generative models here.","excerpt":"The primary objective of unsupervised learning is to train an algorithm to generate its own instances of data. The job here is not to just simply reproduce from the training data but to build a model of the underlying class from which that data has been drawn from. For example, they should not show a […]","categories":["Deep Tech"],"tags":["GANs","RNN"],"author_name":"Ram Sagar","publish_date":"2019-07-17T17:21:39","publication_year":"2019","word_count":683,"keywords":["Go","TPU","AI","neural network","ML","Scala","RAG","RNN","GANs","GAN","R"],"extracted_tech_keywords":["AI","ML","neural network","RAG","TPU","R","Go","Scala","GAN","RNN"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/gans-can-soon-create-high-resolution-videos-dvd-gan\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10064831,"title":"What’s the big deal about DALL.E 2?","content":"Early last year, San Francisco-based artificial intelligence company OpenAI launched an AI system that could generate a realistic image from the description of the scene or object and called it DALL.E. The text-to-image generator’s name was a portmanteau coined after combining the artist Salvador Dali and the robot WALL.E from the Pixar film of the same name. OpenAI then described DALL.E as the “GPT-3 for images.” Like GPT-3, DALL.E is also a transformer language model. Despite the fact that the model was seemingly making images from nothing, the results produced weren’t exactly frame worthy. Last week, OpenAI announced a more advanced version called DALL.E 2. The reaction to the images churned out by the model caused a mini commotion on Twitter. CEO Sam Altman started inviting random commenters to suggest the most inventive ideas they could think of for the model to make images out of. DALL.E 2 more than obliged by creating imagery out of thin air. Source: Twitter Source: Twitter Comparison DALL.E was a 12-billion parameter model that worked using a dataset of text-image pairs. As an input, it received both the image and the text as a single stream of data. Each stream of data contained up to 1280 tokens and was trained using maximum likelihood to generate all of the tokens, one after another. This enabled DALL.E to produce a fresh image from scratch. It could also regenerate any rectangular region of an existing image that extended to the bottom-right corner, such that it was consistent with the text prompts. DALL.E 2 essentially does the same thing that DALL.E does, which is to take a complex prompt like, “A painting inspired by Banksy’s art showing a machine-human interaction,” and then turn it into hundreds of images. Eventually, it chooses the most suitable image from all the outputs to one that would meet the user’s standards. However, DALL.E 2 is far more versatile and capable of producing images of a higher resolution. More efficient: DALL.E 2 functions on a 3.5-billion parameter model while using another 1.5-billion parameter model to enhance the resolution of its digitally-produced images. The model is also faster at processing images than DALL.E. The big jump in performance is because of a new diffusion model, which is smaller and more efficient than the one that DALL.E used. The diffusion model starts with an image that is entirely noise and then gradually transforms itself to make it look as close to the prompt. Source: OpenAI blog More realistic: The images produced by the newer version are more well-rounded, with complex backgrounds, realistic lighting and reflections. The final product is a far cry from the images that DALL.E produced that were cartoonish and, more often than not, had a plain background. Source: OpenAI blog Editing: Another major addition to DALL.E 2 is that it can edit an image using what it calls, “inpainting.” A user can input the prompt asking for the change that it wants to make and select the area on the image that it wants to edit. In a few seconds, the model produces a handful of options for the user to choose from. For example, the user could select an area on a table that carries plates and prompt for them to be removed if they want. The model is also able to render appropriate lighting and shadows in images besides also using the most suited materials for the objects. Source: OpenAI blog Multiple variations: DALL.E 2 is also able to produce multiple variations of a single image. These variations could be an impressionistic version of the image or a close resemblance of it. The user can even give the model a second image and DALL.E 2 can combine the more vital features of both the images to form a final one. According to tests conducted by OpenAI, DALL.E 2’s image classification and captions are more accurate. In the past year, it was found that algorithms were more vulnerable to being tricked into mislabelling an item. For example, if the system was trained using the image of an apple that was labelled ‘orange,’ the system would get tricked into believing that it was an orange. However, DALL.E 2 does not make the same mistake. Limitations OpenAI has said it is conscious of the potential negative impact that DALL.E 2 could have in the wrong hands. In today’s world of deep fakes, the model could easily be used to produce misinformation or racist imagery, which is why OpenAI has allowed DALL.2 to be used by developers on solely an invite-only basis. All the prompts that the model receives must adhere to a strict content policy. To completely rule out the possibility of DALL.E 2 producing any hateful or violent images, the dataset itself omitted the inclusion of any dangerous weapons. While OpenAI has said that it intends to turn it into an API eventually, it is prepared to proceed with caution in the case of DALL.E 2.","excerpt":"DALL.E 2 is far more versatile and capable of churning out images of a higher resolution.","categories":["IT Services"],"tags":["AI Tool","DALL.E"],"author_name":"Poulomi Chatterjee","publish_date":"2022-04-13T12:00:00","publication_year":"2022","word_count":824,"keywords":["Go","API","artificial intelligence","TPU","OpenAI","AI","Git","GPT","ViT","AI Tool","R","DALL.E"],"extracted_tech_keywords":["AI","artificial intelligence","OpenAI","TPU","R","Go","Git","API","GPT","ViT"],"url":"https:\/\/analyticsindiamag.com\/it-services\/whats-the-big-deal-about-dall-e-2\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":39981,"title":"Top 6 Cyber Hygiene Practices Every SME Should Follow","content":"Over the past couple of years, the world has witnessed some of the most notorious cyber-attacks. While big companies across the world have started taking things seriously and have started strengthening their cybersecurity infrastructure, many small companies are still in denial. However, it is high time now that these companies start maintaining cyber hygiene. Cyber hygiene is nothing but the practices and precautions that a company or an individual should take. It is just as important as personal hygiene — it not only helps in keeping your data safe and well-protected but also helps in protecting an organisation’s computer systems from outside attack. In today’s business world, data is the most valuable asset for any organisation, and it is imperative that you take the most protective measures to keep your data safe. However, sometimes, you don’t need a huge enterprise security team with a lot of solutions in its arsenal — all you need to do is follow some basic practices. In this article, we are going to have look at some basic cyber hygiene steps that every small company should take in order to establish a strong cybersecurity infrastructure. Risk And Threat Identification When you start out to set a cybersecurity infrastructure, make sure you first identify the threats and the risk they pose. This is the first and foremost thing every small company should before spending any amount on security tool. So make sure you figure out all the potential threats and the risk they pose. By doing so, you would be able to figure out what solutions you would require.  Also, have at least one Cybersecurity professional in your team with significant experience and who is capable of doing the threat analysis for your organisation. Have A Strong Cybersecurity Policy Your data should be your priority and in order to secure that data, you must take sincere, serious and quick action towards all kind of threats, irrespective of the severity and size. Cybersecurity hygiene is all about the continuous effort to address risks in real time. So, make sure your company has the policy to deal with threats — even if it’s just about antivirus to stop viruses and malware. Also, don’t forget to introduce a training module for all the employees. When you are a small company with a small team, it is easy to train your employees — train on every possible aspect of cybersecurity. Look Out For Third Party Support Having a dedicated in house cybersecurity is team is always considered to be good. However, there are times when you need some support from outside the office. This is where third parties come in. And seeking party support doesn’t necessarily mean you would have to spend a heavy amount. There are dedicated third parties for small companies and startups who would provide all the necessary support at budget. While third parties take care of all your cybersecurity infrastructure, you get time to focus on your core business, and that’s a major benefit. Divide Responsibilities Among Other Employees Too When you are a small company, you have the advantage of getting help from each and every employee. Take some time off from your core business and train your employees on some specific roles and let them take care of it. They might not be able to completely eliminate any risk for you, but they would definitely be able to mitigate the risk to a great extent. Have Software And Hardware from Trusted and Certified Vendors There are many instances from all across the world where vendors were found to be culprits behind breaches. There are many vendors that tend to fool around with firms that are newborn and doesn’t have much enough experience. So, every time you get a product for your firm, make sure you verify properly — each and every section. Also, if required have cybersecurity personnel to do that for you. When it comes to data security, you cannot compromise with anything. Install Firewalls Among all the other measures, installing a firewall might cost you a bit more. However, it is worth the price you spend. Today, having a firewall is one of the most important things for organisations of any size. A firewall is basically a software or a hardware device that protects your systems (if not completely then to a great extent) from being attacked by hackers on the internet. So, try to make an investment on a firewall for your company’s internet connection. It would allow you to set up online rules for the users, which a great way to mitigate risk.","excerpt":"Over the past couple of years, the world has witnessed some of the most notorious cyber-attacks. While big companies across the world have started taking things seriously and have started strengthening their cybersecurity infrastructure, many small companies are still in denial. However, it is high time now that these companies start maintaining cyber hygiene. Cyber […]","categories":["AI Trends"],"tags":["Cyber Security","Cyber security best practices","Firewall hardware","hardware firewall"],"author_name":"Harshajit Sarmah","publish_date":"2019-05-31T11:22:11","publication_year":"2019","word_count":763,"keywords":["Go","Cyber Security","programming_languages:R","AI","programming_languages:Go","Cyber security best practices","Firewall hardware","Rust","GAN","hardware firewall","R","programming_languages:Rust","startup"],"extracted_tech_keywords":["AI","R","Go","Rust","GAN","startup","programming_languages:R","programming_languages:Go","programming_languages:Rust"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-6-cyber-hygiene-practices-every-sme-should-follow\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10068320,"title":"NSE ties up with analytics platform Qlik to enhance its regulatory capabilities","content":"Data analytics platform Qlik has announced that National Stock Exchange of India Ltd.(NSE) is using Qlik to support and augment its regulatory capabilities. Qlik supports multiple teams within NSE to drive efficiencies and provide insights for facilitating data-informed decisions. Qlik offers an end-to-end cloud platform that delivers real-time data integration and analytics solutions to improve decision-making. NSE is using Qlik’s platform to integrate data inputs from numerous databases, including trading data, clearing and settlement data, reporting by members, risk management services and client data. This enables NSE to generate insights from all their data collectively and run comprehensive reporting for regulatory purposes. “Data utilisation is very important for NSE. We wanted to be able to have a single view of our data and be able to analyse it very intuitively. With Qlik, we have been able to integrate our data on a single platform, reduce the time taken for reporting and quickly identify risk areas,” said Dr Dinesh Kumar Soni, senior VP, Regulatory Group at NSE.With Qlik, the non-technical NSE team members have been able to create and use their own reports with no-code solutions and natural language processing. “We are delighted to help the National Stock Exchange of India implement Qlik to manage regulatory compliance proficiently…customers like NSE are testament to the strength of our platform in bringing data together and being able to utilise it for fine-tuning decisions and business processes,” said Varun Babbar, MD, India, Qlik.","excerpt":"NSE is using Qlik’s platform to integrate data inputs from numerous databases.","categories":["AI News"],"tags":[],"author_name":"Zinnia Banerjee","publish_date":"2022-06-03T13:08:53","publication_year":"2022","word_count":239,"keywords":["programming_languages:R","AI","analytics","R","analytics platform"],"extracted_tech_keywords":["AI","analytics","R","analytics platform","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/nse-ties-up-with-analytics-platform-qlik-to-enhance-its-regulatory-capabilities\/","complexity_score":2,"technical_depth":5,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10111102,"title":"Top 7 Products Built Using OpenAI API","content":"Launched in March 2023, OpenAI’s API allows researchers and developers to interact with advanced AI models for various tasks like image creation and natural language understanding, allowing for creative applications and solutions in various AI-related tasks. In simpler terms, the API is like a helper that lets you use OpenAI’s smart programs in your projects. For example, you can add cool features like understanding and creating text without having to know all the nitty-gritties of the underlying models. Certainly, OpenAI offers a range of fascinating AI products beyond the well-known ChatGPT. Let’s explore some of the coolest ones. Navan Navan is the next generation of software designed for people. Their corporate travel and expense solution allows users to easily book, view, and manage business travel and expenses. Through the OpenAI and ChatGPT APIs, the company is leveraging the technology in both user-facing features, such as modernising its contact centre and optimising efficiencies across its codebase. Expedia Expedia has integrated OpenAI’s ChatGPT into its mobile app, allowing open-ended conversations for personalised travel recommendations. Expedia’s new plugin, in collaboration with OpenAI, simplifies trip planning for ChatGPT users. Currently limited to English conversations, the feature is rolling out globally, building on existing AI features within Expedia’s platform. Expedia Group and Kayak are the first travel brands to develop ChatGPT plugins, experimenting with its potential as a chatbot, search engine, and retail platform. Notion Notion is a single space where you can think, write, and plan. OpenAI API allows integrations to access Notion’s pages, databases, and users, connecting services and creating interactive experiences. It is a freemium productivity and note-taking web application developed by Notion Labs Inc. Intercom Intercom is a customer messaging platform that facilitates efficient communication between businesses and their customers. Its own proprietary AI technology, Fin, is an AI bot powered by a mix of models, including OpenAI’s GPT-4. Companies like Wolt, Trumpet, Hospitable and many more use Intercom for smooth client interactions. Intercom helps organisations deliver an exceptional customer experience. Yext Yext Chat employs GPT-3 and other generative AI models to enhance user understanding and response, retaining a conversational voice AI and chatbot-like interaction style. Yext publisher partners include Facebook, Apple, Amazon Alexa and more. In order to reflect how users are interacting with their Yext Chat AI, they can also update it and improve the replies. Instacart Online grocery delivery platform Instacart uses ChatGPT along with Instacart’s own AI and product data from their 75,000+ retail partner store locations to help customers discover ideas for open-ended shopping goals. In, 2022 Instacart also launched its Instacart Platform, a full suite of end-to-end solutions, enabling retailers to grow their businesses. Sana Modern LLMs enable Sana Labs, which creates self-paced courses and live group sessions, to provide everyone with the knowledge they need to operate more quickly and creatively. Sana employs customised GPT-3 models that have been adjusted using their own data. It develops employee training initiatives and assesses team development using Sana’s reliable data. Additionally, the company works with the world’s largest corporations and the fastest-growing startups to offer personalised, flexible learning to everyone.","excerpt":"The API is like a helper that lets you use OpenAI’s smart programs in your projects.","categories":["AI Trends"],"tags":["Application Programming Interface","Top Trend"],"author_name":"Arya Vishwakarma","publish_date":"2024-01-19T13:00:09","publication_year":"2024","word_count":513,"keywords":["Top Trend","ChatGPT","Go","Application Programming Interface","API","OpenAI","AI","RAG","GPT","generative AI","GAN","R"],"extracted_tech_keywords":["AI","generative AI","ChatGPT","OpenAI","RAG","R","Go","API","GPT","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/top-7-cool-products-built-using-openai-api\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":53383,"title":"Sonos Alleges Tech Infringement, Sues Google Over Smart Speakers","content":"Sonos, a speaker manufacturing company, this week sued Google, alleging that the tech giant “stole” it’s patented wireless technology which enables the interconnectedness and synchronisation of its speakers over the internet. This technology first featured in 2005 when Sonos introduced its new range of wireless speaker that could be controlled from a handheld device. The technology, at the time, only allowed for changing or managing songs on the playlist and moderating the volume. The Background Of The Speaker War “Google is an important partner with whom we have collaborated successfully for years, including bringing the Google Assistant to the Sonos platform last year,” says the CEO of Sonos, Patrick Spence. In 2013, it employed Google to design a music service platform for its speakers and also to incorporate features such as voice control, etc. To remove as many obstacles as possible and to assist Google in the designing process, Sonos deemed it ‘harmless’ to share the blueprint of their patented technologies used in their speakers. Sonos has now filed a lawsuit alleging that Google, as well as Amazon, have replicated and implemented similar technologies (as Sonos’ patented ones) in their range of wireless speakers. Sonos claims that more than 100 instances of copyright breach have been recorded to date but are choosing to pursue legal action on only five accounts of misconduct. Apart from the blueprints, Sonos purchased the Google (and Amazon) devices and subjected the same to packet sniffing test, through which they established that Sonos’ speakers and Google’s (and Amazon) speakers, indeed, has similar mechanics involved in their functioning. Uphill Battle For Sonos Sonos is also only taking Google to court as it feels that it does not have the ability nor the resources to fight two tech heavyweights at the same time. It is pursuing a ban on the sale of all Google devices such as speakers, laptops, smartphones, Google Nest products, etc., which, Sonos feels, is wrongfully enabled by its patented wireless technologies. Sonos is also looking for financial compensation for the damages incurred; however, the figure remains undisclosed. Sonos filed the lawsuit in the Federal District Court in Los Angeles and the United States International Trade Commission which has the authority to ban import and export of goods which infringe patent rights. Spence comments, “Google has been blatantly and knowingly copying our patented technology,” and “Despite our repeated and extensive efforts over the last few years, Google has not shown any willingness to work with us on a mutually beneficial solution. We’re left with no choice but to litigate” Tech Giants On The Speaker Tech Theft Allegations In their defence, Google and Amazon have each release a statement. Jose Castaneda of Google said, “We dispute these claims and will defend them vigorously”, adding that a legal battle could have been avoided and that the matter could have been easily resolved internally. Hereth Natalie, the spokesperson for Amazon, said, “The Echo family of devices and our multiroom music technology were developed independently by Amazon.” It has faced stiff, crippling competition from Google and Amazon. In the third quarter of 2019, according to a report, Amazon shipped 10.5 million and Google shipped 6 million speakers respectively. In the whole of 2019, Sonos recorded the sale of 6 million units. Sonos primarily focuses on the quality of sound in their speakers resulting in their cheapest speaker being priced at $180 while Google and Amazon’s entry-level speakers can be had for as low as $50; the two companies also provide hefty discounts on their products. “The harm produced by Google’s infringement has been profoundly compounded by Google’s business strategy to use its multi-room audio products to vacuum up invaluable consumer data from users and, thus, further entrench the Google platform among its users and ultimately fuel its dominant advertising and search platforms,” says Sonos. Outlook Sonos needs the two tech giants to market its speakers to customers internationally. A lawsuit can have a devasting effect on the company, the likes of which it may find hard to recover from. Sally Hubbard, a thinktank employee comments, “The fear of retaliation is a real fear. Any of these companies could bury them tomorrow. Google could bury them in its search results. Amazon can bury them in their search results.” She further says that almost every company is dependent on one of the tech giants in one way or the other. Sonos also adds the virtual assistant feature from Google (and Amazon) into their speakers at the manufacturing stage, since the feature continues to remain largely popular. To reduce its dependence on the tech giants, Sonos turned to the consumers, indirectly, who have the option of using either the virtual assistant or Sonos’ individual platform. The lawsuit will certainly dampen the relationship between all the companies involved, especially Sonos and Google. Sonos, though, hopes to maintain a cordial alliance with the tech giants.","excerpt":"Sonos, a speaker manufacturing company, this week sued Google, alleging that the tech giant “stole” it’s patented wireless technology which enables the interconnectedness and synchronisation of its speakers over the internet. This technology first featured in 2005 when Sonos introduced its new range of wireless speaker that could be controlled from a handheld device. The […]","categories":["AI News"],"tags":["Google"],"author_name":"Yeshey Rabzyor Yolmo","publish_date":"2020-01-08T17:29:21","publication_year":"2020","word_count":811,"keywords":["Replicate","Go","AWS","AI","cloud_platforms:AWS","programming_languages:R","ML","Aim","Google","llm_models:Bard","R"],"extracted_tech_keywords":["AI","ML","Aim","AWS","R","Go","Replicate","llm_models:Bard","cloud_platforms:AWS","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/sonos-lawsuit-google-wireless-speakers\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10089772,"title":"What Mexico can Learn from India When it Comes to Instant Payment","content":"In October 2019, the Mexican Central Bank launched the digital payment platform ‘Cobro Digital’, or CoDi, which enabled mobile-based payment based on QR codes and near-field communication (NFC) technology. The idea was very similar to India’s UPI. But, unlike India, CoDi is struggling in the Mexican market and cash is still the dominant mode of transaction in the country. So, what is going wrong at CoDi and what can it learn from India’s UPI? UPI Way: A Successful Model UPI was launched by NPCI and has two varieties of transactions, which are known as PUSH or PULL. Generally, PUSH is majorly used in the market and transactions happen in two phases. In the first phase, the customer initiates the transaction via Payee’s mobile number, VPA (Virtual Payment Address), or QR code. The Payer’s PSP (Payment Service Provider), which is typically a bank then sends the transaction request to the NPCI (National Payments Corporation of India). The NPCI UPI server then forwards the same request to the Payee’s PSP for authorisation and address resolution. The Payee’s PSP then resolves the address and provides the account details by working with the Remitter bank. In the second phase, the Payee’s PSP provides the bank details to UPI, which is then forwarded to NPCI. The NPCI then checks with the Remitter bank to debit the funds from the payer’s account. Once the funds are debited, a credit request is sent to the beneficiary’s bank which then credits the Payee’s account and responds to the NPCI UPI. The NPCI UPI server then passes on the response to the transaction status through the Payer’s PSP to the customer. How CoDi is inefficient CoDi works almost the same way as UPI does, except the fact that instead of NPCI, the Central Bank of Mexico (SPEI) becomes the mediator. Additionally, CoDi is only available to the banks that are members of the SPEI. However, CoDi—even after building a system so close to UPI​​—still fails in the areas where it could have been a success. For instance, it doesn’t provide an option for the person-to-person transaction and largely depends on person-to-merchant transactions. In that too, while the interface does provide an option for QR codes, it still isn’t the best one. Every time a merchant has to receive a payment, the system requires a request generation through the merchant’s application, after which the customer can pay by scanning the code. When compared with UPI, this approach takes way more time, nearly 20-30 seconds per transaction. This, perhaps, might be the main reason why CoDi is not well adapted in Mexico. With only 38 banks and 500,000 users, the application has been slow to gain ground in the country. Miguel Diaz, Banco de México’s general director of payment systems and market infrastructures, also told in a recent media interaction that much of the sluggishness was initially due to the platform’s poor user interface, which required merchants to send 18-digit account numbers to each new customer during onboarding. Additionally, the system only processes single transactions up to a maximum value of 8,000 pesos, whereas UPI has a limit of INR 100,000 per day with no limit on the number of transactions. However, on the bright side, much like UPI, CoDi is also free of banking commissions and the transactions can be made without keeping banking hours in mind. As per the official website, there have only been 13,883,440 transactions performed on the applications to date, which is far lesser than UPI in India. Many speculate that the central issue with CoDi was its lousy promotion by banks. There were little to no promotions, except for a couple of videos being uploaded on the Youtube channels. Additionally, the banks started small business payments directly, instead of going to the service providers—like Electricity, Water bills—with which the QR codes could have been attached. The pandemic would have catapulted the adoption. Once people were accustomed to paying their bills with QR codes then the platforms could move on to businesses, starting with the largest such as supermarket chains such as 7-Eleven. All in all, the platform could have been more successful if not for such hindrances.","excerpt":"Although similar to UPI, the Mexican platform still needs improvement in customer journey.","categories":["IT Services"],"tags":["npci","UPI","User Interface"],"author_name":"Lokesh Choudhary","publish_date":"2023-03-21T16:10:01","publication_year":"2023","word_count":692,"keywords":["Go","npci","AI","programming_languages:R","programming_languages:Go","Git","UPI","User Interface","R"],"extracted_tech_keywords":["AI","R","Go","Git","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/it-services\/what-mexico-can-learn-from-india-when-it-comes-to-instant-payment\/","complexity_score":3,"technical_depth":6,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10415,"title":"IFIM B School Launches Third Batch of its Analytics Program","content":"With an increase in data across the industries, there is a dire need of newer tools and techniques that can quickly and easily collect, store, manage, integrate, manipulate, aggregate and analyze all types of data to positively impact businesses. As it indicates towards a need for business analytics, the flip side is an acute shortage of skilled manpower in this area who can understand the potential of big data analytics and translate analysis outcome to integrate with overall business strategy of the organization. Despite a surge in data science related programs, universities and colleges are not able to produce effective and efficient data scientists fast enough to meet the business demand. What IFIM offers? In an effort towards addressing this need in analytics industry and preparing a business environment from the unpredictability and disruption that it can create, IFIM B-School launched the executive PGDM program in September’ 2014. Gaining significant attention from corporate world as well as practitioners, it has enrolled more than 100 business professionals till date from various companies like IBM, Genpact, TCS, Herman Miller, Schneider Electric, Fujitsu, United Healthcare, Mindtree etc. For building world class capabilities for management students and working professionals, the success story of the program has been recognized on various platforms such as IBM. In line with its previous years’ success, IFIM launches its third successful session for the year 2016-2017. This 15 months’ duration AICTE approved program is uniquely designed for a hybrid model of delivery. The architecture of the course has been so laid out, that it gives emphasis on training and making an individual gain expertise in all major pillars of business analytics such as business acumen, domain knowledge, tools & techniques and problem solving capabilities. IFIM Institutions established in 1995, founded by the Dalal Street Investment Journal (DSIJ) and promoted by Centre for Developmental Education (CDE), has many accolades under its belt such as the ‘Best Higher Education Institute of the Year-South’ by ASSOCHAM 2016, ‘Outstanding Business School (South)’ at the National Education Awards 2015 and recognized as the ‘Most promising B-School with Excellent Industry Interface’ at the Asia Education Summit 2015. Program objectives: The program designed for working professionals equips the individual looking for higher career goals in the analytics domain. It trains individual on enterprise-wide perspective on data, result communication, preparing a business intelligence roadmap, and developing abilities that foster an analytical culture within an organization. Starting with understanding the fundamentals of business analytics and big data technology, it helps a candidate manage business analytics process and get familiar with state-of-the-art software applications for business analytics. It adds skills such as understanding the application of business analytics within an organization, evaluate the impact of analytics on overall strategy, understanding the roadmap for analytics implementation, understanding best practices for business intelligence rollout and much more to their profile. The biggest perk of the program is the incorporation of real life project that gives an individual a chance to apply their conceptual and technical understanding into solving a business problem in-hand. The program offers hands-on-experience of working with relevant software from IBM, OpenSource and Microsoft. Curriculum Overview: This 15 month programs, extended over three terms includes 800+ hours of intensive sessions. Apart from the business management subjects, it involves core business analytics subjects and project along with specialized domain electives and project. It covers all the basic tools used in analytics, advanced EXCEL, R, power BI, analytics framework – predictive analytics, data visualization & design, tools & techniques in big data analytics and more. The final terms of the course involve selecting any two of the specializations among finance, marketing, logistics & supply chain management and HR. The entire curriculum is designed and delivered by a blend of highly qualified and experienced faculty from IFIM and industry experts. Experienced in-house faculty as well as those from outside are at the helm of this program, making it a good learning experience for the students. The hybrid model of course delivery is facilitated by the live streaming of lectures and classroom teaching to the enrolled students along with an access to recorded sessions. Any graduate in any discipline with at least 50% marks in aggregate and a work experience of minimum three years and above are eligible to enroll for the course. Keeping in mind the working professionals that make into the enrolment list, there is a convenience of weekend classes including Saturdays and alternate Sundays that they can avail. In short the curriculum offers a poise between key business concepts with skills in analytical, statistical modelling and data management, and professionals with anywhere from below 5 years to more than 15 years of experience in the industry can enroll for the course. Other Perks: To further strengthen the hold at academic\/ industry interface in analytics, IFIM has established a Center of Excellence in Business Analytics. IFIM Business School has also collaborated with IBM and has set up a business analytics lab to help students enhance their skills in the areas of managerial decision making and strategy. This program initiated by Career Education for Business Transformation (CEBT) focuses on the major areas of business transformation where software plays a critical role. IBM provides relevant software for CEBT training, to address the increasing need for T-shaped skills in the global marketplace. Apart from IBM, IFIM also has collaborations with organizations like NASSCOM National Stock Exchange (NSE) to offer certified courses in Business Analytics and Data Sciences and Global Financial Markets respectively. To deal with the ever increasing data and with an advent of the technology in big data space, the focus remains on platforms that provides resources to house cutting edge tools and intensive software. And given the fact that many aspects apart from technical play are critical for successful implementation and roll out of analytics at organization level, the course offered by IFIM offers the right mix of both practical and analytical knowledge hence bringing out the best of analytics industry. For more details, visit http:\/\/www.ifim.edu.in\/analytics\/","excerpt":"With an increase in data across the industries, there is a dire need of newer tools and techniques that can quickly and easily collect, store, manage, integrate, manipulate, aggregate and analyze all types of data to positively impact businesses. As it indicates towards a need for business analytics, the flip side is an acute shortage […]","categories":["AI Trends"],"tags":[],"author_name":"Srishti Deoras","publish_date":"2016-07-19T07:10:45","publication_year":"2016","word_count":994,"keywords":["big data","data science","Go","business intelligence","ELT","AI","predictive analytics","analytics","GAN","R"],"extracted_tech_keywords":["AI","data science","analytics","predictive analytics","R","Go","big data","ELT","GAN","business intelligence"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/ifim-b-school-launches-third-batch-premier-analytics-program\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10110889,"title":"HCLSoftware Prioritises Customer Benefits Over Gen AI Hype","content":"Generative AI gained ground in Indian IT in 2023. The year saw a slew of companies introducing the technology through language models of all sizes. While the rest rushed to jump on the bandwagon, some companies are taking their time to add the technology to their products and services. One of them is HCLSoftware, which believes that the use of generative AI should be pragmatic. The software production arm of the IT veteran HCLTech is primarily balancing the technology with the cost that customers need to pay with the value that appears. “We don’t want Gen AI just to put it in front of the customer. We want them to benefit from the technology by using the products,” declared Kalyan Kumar, HCLSoftware’s Chief Product Officer, in an interview with AIM. The executive was joined by the company’s Chief Revenue Officer, Rajiv Shesh, who said, “We balance it [generative AI] with the cost that the customers need to pay, and the value will appear.” “We are engaging with language models in several ways across our technology, and we’ve been working with these models for much longer than the emergence of the current discussion of the state. Every product line has an embedded AI capability,” Shesh elaborated. A prime example is HCL Unica, a software for multichannel marketing. In 2019, HCL acquired Unica, a platform developed by Yuchun Lee in 1992 that provides AI-driven assistance today. “We founded Unica focused on marketing to help organisations provide offerings so targeted that it doesn’t feel like selling or marketing but a service to the customer,” said Lee, who also joined the interview alongside Shesh and Kumar. He further elaborated that a high level of computation is required to handle this concept. “I think going forward, this is really about how HCL can take that concept and deliver the capability at an unheard-of scale,” he said. Providing Customers an Experience The software leader fulfils their customers’ business requirements through two extreme approaches. Shesh, who joined HCL in 1997, elaborated that “one is absolute standardised software. Second is that you sit across from a customer, get needed specifications, and start building it from scratch.” “Instead of waiting for years for that software to come in the next series software, we make it available to expedite and speed up the capability of the software they deploy. At the same time, keep the risks at cost manageable,” he explained how the company foundationally engages with its customers. Additionally, Kumar highlighted, “We can do things very differently because we have the engineering scale not to think like a traditional software company that everyone should fit into a business slice.” Across 132 countries, HCLSoftware has more than a thousand business partners. “You can pick any timezone starting from the International Date Line. We have customers north to south from the Arctic and Antarctic. We are the only software company built out of engineering services heritage,” he gladly boasted. HCLSoftware intends to help customers extract value from their investments. “The core principle of our business is that our software should get used,” added Shesh. He further said that HCLSoftware has generated the bulk of its business by providing new capabilities to its clients. The organisation’s goal is to keep customers abreast of everything new in technology. “This whole process is connected so that we understand that the technology is being created based on what the customer is seeking from us,” Shesh explained. “That cycle is going on,” the industry veteran said, reflecting on the growing clientele across 130+ countries. “Our customers are our best salespeople. They’re the ones who provide references,” he added. Mapping History HCLSoftware was created by HCLTech with the single purpose of building a software production business. “It was a very conscious plan decision and a curation of intellectual property acquired from IBM, Cisco and a few others,” recalled Kumar, who has worked with HCL for over two decades. “We’ve got a set of capabilities which enable businesses to rewire and run the company in a digital plus enterprise economy. We help them attract prospects, transact and nurture the customer lifecycle, “he said, But that is one part of our business. The second is the world’s most extensive portfolio of low code and no code capabilities, which the team inherited, noted Kumar. The company’s low code development platform, HCL Volt MX, has proven exceptional as it provides solutions to customers like Ferrari. Kumar noted that the third is the suite of products around hybrid data. “We believe the world of data cannot get on the cloud or on-prem. It has got to be hybrid consistently,” he added, emphasising the company’s focus on cloud. HCLSoftware also has the Intelligent Automation Cloud Suite, which can securely deploy, develop, and build applications and manage endpoints, observability, AIops, and the whole lifecycle.","excerpt":"“We don’t want Gen AI just to put it in front of the customer,” said HCLSoftware’s exec","categories":["AI Features"],"tags":["Generative AI","HCL Technology","Interviews and Discussions"],"author_name":"Tasmia Ansari","publish_date":"2024-01-17T11:51:53","publication_year":"2024","word_count":800,"keywords":["Go","intelligent automation","AI","Git","RAG","automation","Aim","generative AI","HCL Technology","GAN","Generative AI","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","generative AI","Aim","RAG","R","Go","Git","GAN","automation","intelligent automation"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/hclsoftware-prioritises-customer-benefits-over-genai-hype\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10139028,"title":"Indian IT Is Trying Really Hard to Woo GCCs","content":"The Indian IT services industry is grappling with a major challenge as clients increasingly build new AI projects in-house. This shift is reflected in the rising number of Global Capability Centers (GCCs) being established in India, enabling multinational companies to tap directly into the country’s vast software talent pool. To tackle this, Indian IT has adopted a new strategy. They are now prepared to work with GCCs and help them establish their offices in India. “We are working very closely with GCCs all around – working with clients when they set up their GCCs in India,” said Infosys’ chief executive officer, Salil Parekh, during the company’s recent earnings call, adding that the company is involved during the build-operate-transfer process. He also mentioned that they help GCCs in India in scaling and recruitment. Though he didn’t name any specific GCC, he mentioned that in India they are working with a large number of GCCs in financial services, telecommunications, and life sciences, helping and supporting them. Sharing similar views, Wipro chief Srini Pallia said, “As far as GCCs are concerned, it’s our strategy to partner with them. All of us know that GCCs are growing significantly in India at this point in time.” Pallia is particularly interested in leveraging Wipro’s employees trained in generative AI. “We have trained 230,000 employees in generative AI and have 44,000 advanced AI experts. We have the talent, and we can groom it, allowing GCCs and us to partner in executing projects. So, I see this as a win-win for both of us,” he said, adding that Wipro is prepared to partner with them not just on the BOT, but also on many projects. The GCCs and Indian IT will coexist, according to HCLTech chief C Vijay Kumar. “There is a certain type of work that is best done in an outsourced model, and a certain type of work that customers might want to do in-house,” he said. He suggested that while some of this market share is moving to GCCs, this shift is not detrimental to the overall growth opportunities for service providers. “I think there is a slight deflation in the total addressable market for service providers, but it is insignificant because, if you look at the total addressable market for large GSIs, it exceeds one trillion dollars,” he said. “From that perspective, some market share going to GCCs doesn’t really impact the overall growth opportunities for service providers,” he added. Era of GCCs India’s GCC sector is thriving and currently employs over 1.6 million professionals. According to EY, this growth will persist, with projections showing over 2,400 GCCs in India by 2030, generating jobs for over 4.5 million people. The market size is expected to surge to $110 billion by 2030 from the current $45 billion. In comparison, roughly 5.4 million people were employed in the Indian IT sector as of March 2023, according to MeitY and ICRA expects the Indian IT services industry to see revenue growth of 4-6% in FY2025. “GCCs’ share in Indian IT and engineering services will grow due to the scaling of existing GCCs. The impact of new GCC growth will be smaller. India has 1,500+ GCCs, and most Fortune 1000 firms have GCCs in India. For new GCCs, the target profile is Fortune 1000–5000 firms and new tech firms,” said Pareekh Jain, the CEO of EIIRTrend, Pareekh Consulting. Cities like Hyderabad, Chennai, and Bangalore have emerged as the preferred hubs for GCCs. GCCs Pay Higher than Indian IT The increasing number of GCCs in India indicates a trend where more professionals are gravitating toward these organisations. A recent report reveals that GCCs offer salaries 12-20% higher than those in IT services and other non-tech industries for comparable tech roles. For instance, GCCs provide salaries ranging from INR9.7 lakh to INR 43 lakh per annum for software developers, depending on experience. In contrast, the IT products and services sector offers around INR 5.7 lakh per annum for entry-level roles, with salaries going up to INR 17.9 lakh per annum for professionals with over eight years of experience. Moreover, entry-level talent at GCCs in India is attracting pay packages that are, on average, up to 30% higher than entry-level salaries across all sectors. GCCs in India place a great emphasis on developing AI capabilities and leading AI transformation within their organisations. The most in-demand tech roles at GCCs include software engineers, developers, data specialists, AI and machine learning experts, cloud computing professionals, and cybersecurity experts. Meanwhile, Indian IT is still testing the waters with generative AI and has yet to build an impressive product. The industry is currently focused on maintaining existing projects rather than creating something revolutionary. A primary reason for this reluctance is that investing in R&D requires significantly more capital than simply pushing existing solutions into the market. Mohandas Pai, the chairman of Aarin Capital and former CFO of Infosys, told AIM that it has always been ideal for Indian IT companies to focus on services. “Indian IT services companies are not product companies,” said Pai. However, it doesn’t come as a surprise that most Indian IT companies, such as TCS, Infosys, Wipro, HCLTech, and LTIMindtree, are somewhat conservative in disclosing specific details or numbers related to their generative AI initiatives, especially concerning the financial impact or granular operational metrics.","excerpt":"India’s GCC sector is thriving and currently employs over 1.6 million professionals.","categories":["GCC"],"tags":["Editors Picks","GCC","Indian IT"],"author_name":"Siddharth Jindal","publish_date":"2024-10-21T19:26:23","publication_year":"2024","word_count":883,"keywords":["Go","API","GCC","machine learning","AI","cloud computing","RAG","Editors Picks","Aim","generative AI","GAN","Indian IT","R"],"extracted_tech_keywords":["AI","machine learning","generative AI","Aim","RAG","cloud computing","R","Go","API","GAN"],"url":"https:\/\/analyticsindiamag.com\/gcc\/indian-it-is-trying-really-hard-to-woo-gccs\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10068307,"title":"Slack’s great Indian dream","content":"Cal Henderson and Stewart Butterfield founded Slack in 2009. The collaboration tool rose from the ashes of the non-combat multiplayer online game, Glitch. Tiny Speck, the publisher of Glitch, built an in-house tool to foster a sense of collaboration among video game developers while making repetitive tasks more fun and engaging. On realising the commercial potential for an enterprise communication “cool”, the team launched Slacked as an open-source, instant messaging app that let team members collaborate in one place – and it became an instant hit. Over 8,000 customers signed up for the service within 24 hours after it was incorporated in August 2013. In 8 months, the company reached a jaw-dropping valuation of USD 1 billion. In 2015, Slack claimed it had over 135,000 paying customers spread across 60,000 teams, with 10,000 users signing up each week. Last year, the IDC MarketScape recognised Slack as a leader among major collaborative and community platform vendors. The report positioned Slack as a leader for both its strategies and capabilities, including wide adoption by the developer community; integrations with several business software applications; new audio and video features such as huddles and clips; In-house customer relationship support and management; and enterprise-friendly plans to help meet security, compliance and governance requirements. As per Slack, users send over 1.5 billion messages per month on the platform. Salesforce acquired Slack in July 2021. Ever since, the co-CEO of the CRM platform, Bret Taylor, has been endorsing a Slack-first UX. As per a 2021 update, Slack has  12 million daily active users. Further, 43% of Fortune 100 companies have a paid subscription to Slack. The collaboration tool had 156,000 paying customers as of March 2021. Mindblowing stats from @SlackHQ * 300K msgs per sec* 1M developers building apps on Slack#SlackFrontiers pic.twitter.com\/bochNRbD4b— Thien-Lan Weber (@thienlan) November 17, 2021 Today, Slack is so much more than a messaging app for businesses. By centralising the teams, Slack transforms how organisations communicate. Feud with Microsoft Microsoft announced Teams shortly after Slack was launched and has remained its biggest competitor ever since. Slack had 12 million daily active users in 2021. Teams, on the other hand, had over 110 million. Microsoft also claims 270 million “active monthly users.” Teams caters to 500,000 organisations. Most notably, 91 of the Fortune 100 use the platform. However, 66% of Teams customers also use Slack. In 2018, Microsoft added Slack to its list of official competitors. On the other hand, Slack filed an antitrust complaint against Microsoft in 2020. The European Commission is formally investigating the claim. Slack CEO Stewart Butterfield said the platform doesn’t compete with Teams. He alleged Microsoft is “unhealthily preoccupied” with killing Slack, using Teams as a weapon. Eyeing the next billion users Slack changed its growth model after joining the Salesforce family. “The opportunity in India is massive. To start with, Slack has a massive free user base here in India. Therefore, there is a high awareness of Slack and a strong desire to engage with it and try it out. Our mission is to ensure companies using the free version are aware of the additional functionality they will get with the business version, truly unlocking the power of Slack,” said Rahul Sharma, AVP and country manager, Slack. India has the second-largest SMB market globally– a sweet spot for Slack. For SMBs, employee experience is of paramount importance, and they put a lot of emphasis on usability. “The vibrant India Inc allows us to leverage their growth. We have a significant footprint in big global tech companies with a major presence in India, like IBM, Amazon, Oracle, and Intuit. India is also a major focus for Slack due to large systems integrators like Wipro and TCS. There are also the digital-first companies born in the cloud and have built their businesses on Slack, and India’s ever-expanding population of unicorns, who love working on Slack,” he continued. Digital HQ Slack has been operating in India for four years and had set up an engineering team in Pune in 2018 following the company’s acquisition of Astro. The Slack India team has over 120 employees across four offices in Pune, Mumbai, Bangalore and Gurgaon. Indian companies such as Zomato, Dreamsports, Freecharge, Razorpay etc use Slack to enable seamless communication. On June 1, 2022, Slack officially launched in India, announcing its mission to help Indian companies navigate the transition to a hybrid workplace by establishing a digital headquarters (‘Digital HQ’). Adopting Slack as a Digital HQ allows Indian companies to connect their teams, tools, customers, and partners in a digital place that’s fast, flexible and inclusive for a work-from-anywhere world. A Digital HQ allows work to flow, breaking down communication and collaboration silos, internally and externally; automates tasks that take away time from deep, meaningful work; and enables new, flexible ways of working, striking the right balance between synchronous and asynchronous. According to a recent Slack study, The Reinvention of Work, 4 in 5 respondents desired flexibility, and a significant 80% would seek a role elsewhere if their employer didn’t accommodate this. Embracing Slack as their Digital HQ provides Indian knowledge workers with the ability to do their jobs from anywhere, at any time. In the same Slack study, Indian knowledge workers said they were wasting an average of 47 minutes a day switching between the various apps they use to do their jobs. One in five respondents said they were losing 10 hours a week – nearly ten working weeks a year. With over 2,600 app integrations, the Slack platform is purpose-built to address this loss in productivity and improve the employee experience. “We are excited to be here in India, and we see so many growth opportunities. We believe we can positively impact the Indian companies we serve, supporting them in their transition to a hybrid workplace and being the Digital HQ that is relied on by companies of all sizes,” said Rahul.","excerpt":"The opportunity in India is massive. To start with, Slack has a massive free user base here in India.","categories":["IT Services"],"tags":["Microsoft","Salesforce","Slack","Zoom"],"author_name":"Sri Krishna","publish_date":"2022-06-03T12:00:00","publication_year":"2022","word_count":981,"keywords":["Go","Zoom","AI","ML","Git","RAG","Aim","Salesforce","CLIP","Rust","GAN","R","Slack","Microsoft"],"extracted_tech_keywords":["AI","ML","Aim","RAG","R","Go","Rust","Git","CLIP","GAN"],"url":"https:\/\/analyticsindiamag.com\/it-services\/slacks-great-indian-dream\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10131356,"title":"Most AI Startups are Destined to Fail – Even the Funded Ones","content":"Every startup wants to be big someday, and most successful businesses were startups when they began. But the truth is, somewhere down the line most of them end up either dying, or getting acquired by big companies. This gets a notable exit for most investors, but the story for startups ends there. “Most startups are destined to die. Even the funded ones,” said Kunal Shah, the founder of CRED, adding that the success of a startup is mostly a miracle. “And a miracle doesn’t happen with a team that’s looking for stability and dislikes ambiguity,” he said, adding that startups need problem solvers. The same is the case with the current AI startups, globally. The ones that started their journey a few years ago are now either getting acquired or gradually dying because of lack of funds. Only a few companies, such as OpenAI, Anthropic, Mistral, Hugging Face, and few others, are actively getting funded, which eventually will be known as the prodigies of the AI generation. The AI Bubble is Here With Google recently acquiring the founders of CharacterAI, the case of AI startups sustaining for a long term is put into question. The same was the case with Mustafa Suleyman from Inflection AI joining Microsoft AI Research team, Amazon taking over Adept AI’s team, Snowflake’s acquisition of Neeva, or Canva’s acquisition of Lenoardo.ai. https:\/\/twitter.com\/ClementDelangue\/status\/1820381347556229563 Going by that logic, Reka AI or even Cohere, might end up with the same fate. With Emad Mostaque leaving, Stability AI is also going through unstable times. The same could happen to Midjourney. What options do startups really have apart from getting acquired? As the discussion around the AI bubble intensifies, it gets tougher for companies to raise funds from investors as they get increasingly wary. The ones that were successful during the AI boom are now encountering difficulties. Most of these startups are also not making money. For example, Lensa, the AI photo generating company with a great product and marketing, was not able to differentiate itself even when it was generating revenue. Quickly, other companies started building similar offerings within their own products, making Lensa lack defensibility. This is the problem with most of the AI startups. “The problem with AI is that just as quickly as you can create a great product, another copycat can emerge and undercut you,” said David Chen, the CEO of Kapsule. Apart from this, another problem he highlighted was the problem of finding use cases and distribution. While India is currently running as the AI use capital of the world, running on jugaad and not VC money, the long term strategy for them also seems questionable. Though they know how to run businesses without large investment, most of them have the inherent goal of getting acquired, as competing with big-tech is not what they strive to do. A few, such as Sarvam AI, Krutrim, TWO.AI, may have received a decent amount of funding, but long-term plans still remain unclear. Sriharsha Putrevu, the co-founder of Retail Technology Group, said that the current AI startups would fail as they are not focused on value creation, but rather just the valuation. “Startup is a repeatable, scalable, sustainable business model not just a cash burning, user acquiring business models in hope of making profits in decades,” he said. Sales Cure All Problems When it comes to starting a company, raising funds is the relatively easy part (not easy in itself) as we have seen with a lot of current AI funds. The harder part is finding the right market, niche, and perfect fit for profit. The problem cannot be solved by building the best version of an AI model as well, since any competitor would learn from it and create an even better one with the frontier of AI constantly moving. Though the cost of building AI models is shrinking which is helping the startups, it is also drawing in several competitors to the field. Take, for example, OpenAI’s GPT-4 constantly getting dethroned by Meta’s Llama 3 or Anthropic’s Claude, or Google’s Gemini. And these are the ones that are already competing for the top spot; what about the new startups? If AI is like electricity, it is important for startups to build a niche and solve the problem in a specific field, since competing for ‘best electricity’ does not make sense. That is what the current Indian AI landscape is focused on – to build use cases of AI instead of building the next LLM. Maybe, this would help them sustain, but for how long is the question. Moreover, since the investors in India are extra wary of pouring money into startups, these companies have very low tailwinds. This is making them run low on money and get close to the point of acquisition, or maybe extinction. All these problems can be solved with sales, for which startups need to move fast. For some, AI is a bubble, for some it is a tree. Regardless, many AI startups would burst or fall off that tree.","excerpt":"For some, AI is a bubble, for some it is a tree. Regardless, many AI startups would burst or fall off that tree.","categories":["AI Startups"],"tags":["Fund Raising","Startups"],"author_name":"Mohit Pandey","publish_date":"2024-08-05T18:08:48","publication_year":"2024","word_count":837,"keywords":["Anthropic","Hugging Face","Go","API","OpenAI","AI","Fund Raising","Scala","GPT","Startups","R","Snowflake"],"extracted_tech_keywords":["AI","OpenAI","Anthropic","Hugging Face","Snowflake","R","Go","Scala","API","GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-startups\/most-ai-startups-are-destined-to-fail-even-the-funded-ones\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10071485,"title":"Atlassian gets a new CTO","content":"Atlassian Corporation Plc, a leading team collaboration and productivity software provider, has appointed Rajeev Rajan as its new Chief Technology Officer (CTO). In his new role, Rajan will lead and support all global engineering, IT, trust and security teams. Atlassian works with 230,000+ customers across large and small organisations—including Bank of America, Redfin, NASA, Verizon and Dropbox—use Atlassian’s project tracking, content creation and sharing, and service management products to work together and deliver timely results. With close to thirty years of experience in building and leading teams at pioneering global technology companies, including Meta and Microsoft, Rajan has joined Atlassian. The India-born executive brings a range of experience in scaling global technology companies, having spent nearly five years at Meta, most recently as Vice President and Head of Engineering for Facebook and Head of Meta for the Pacific Northwest region. Prior to Meta, Rajeev had a career spanning 23 years at Microsoft, working across multiple products from Exchange to SQL Server to Active Directory, culminating in Office 365. The BITS Pilani alumnus holds a postgraduate degree in computer science from the Ohio State University. In addition, he spent a year in graduate school at the University of Illinois Urbana-Champaign computer science program. Rajan said he considers this appointment as the next step in his career, and believes that the role of Atlassian CTO is where he could contribute most meaningfully. As he spent the entirety of career in building and leading software teams, Atlassian’s mission to tap the potential of every team through collaboration and productivity software resonates with him. ‘Atlassian has the potential to become one of the top engineering teams globally as the company looks forward to tripling the engineering team in the upcoming years’, says Rajan.","excerpt":"Rajan will lead and support all the global engineering, IT, trust and security teams in his new role.","categories":["AI News"],"tags":[],"author_name":"Tasmia Ansari","publish_date":"2022-07-25T14:34:21","publication_year":"2022","word_count":289,"keywords":["programming_languages:R","AI","programming_languages:SQL","ViT","SQL","Rust","GAN","R","programming_languages:Rust"],"extracted_tech_keywords":["AI","R","SQL","Rust","GAN","ViT","programming_languages:R","programming_languages:SQL","programming_languages:Rust"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/atlassian-gets-a-new-cto\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10068879,"title":"Salesforce innovations that fuel AI","content":"Salesforce has consistently ranked as one of the leading CRM providers globally. According to Grand View Research, the global customer relationship management market size, valued at USD 52.4 billion in 2021, is poised to grow at a CAGR of 13.3% from 2022 to 2030. To cash in on the opportunity, Salesforce is leveraging AI and ML. Image: CX & Relationship Mgmt. Software Market Grew in 2018 | Gartner Salesforce Einstein The AI revolution will not only transform the consumer world but also change the work world as well, said John Ball, then GM of Salesforce Einstein, while introducing Salesforce Einstein in 2016. Salesforce claimed that, with  Salesforce Einstein, companies will be able to make better predictive and personalised customer experiences across sales, service, marketing, commerce etc. Einstein has applications across various domains. Sales Cloud: The customers can use Einstein to analyse data and predict which leads and opportunities are most likely to convert. Einstein Activity Capture syncs email and calendar to Salesforce automatically along with a prebuilt activity dashboard. Marketing Cloud: Einstein can help deduce customer choice better with predictive insights drawn from their marketing engagements, brand interactions, and conversations across social media. Businesses can also create personalised messages and content based on customer preferences and intent. Service Cloud: Salesforce’s AI-based chatbots can work on checking claims status or modifying orders using NLP on real-time channels like chat and messaging. TransmogrifAI In 2018, Salesforce open-sourced its machine learning tool, TransmogrifAI- the engine powering Salesforce Einstein. Image: Open Sourcing TransmogrifAI – Salesforce Engineering Blog Merlion Merlion is an open-source Python library for time series intelligence. It comes with a standardised and easily extensible framework for data loading, pre-processing, and benchmarking for time series forecasting and anomaly detection tasks. The models here include classic statistical methods, tree ensembles, and deep learning methods. Merlion also comes with an evaluation framework that simulates the live deployment and re-training of a model in production. This library aims to provide engineers and researchers a one-stop solution to rapidly develop models for their specific time series needs and benchmark them across multiple time series datasets. Image: 2109.09265.pdf (arxiv.org) CodeGen Last March, Salesforce released CodeGen, a large scale language model which turns simple English prompts into executable code. The idea is to democratise access to the world of writing software, allowing anyone to develop apps in conjunction with an “AI assistant” or “teacher” without the need to learn programming in the traditional way. CodeGen, a 16-billion parameter auto-regressive language model, has been trained on a large corpus of natural and programming languages. It can be applied to both simple and complex problems, using natural language. Using CodeGen, people with little or no programming knowledge can solve relatively simple coding problems. Through this, Salesforce also aims to bring disadvantaged groups into the programming world. ML Lake Eli Levine, (then) software engineering architect at Salesforce, said ML Lake helps application developers and data scientists to easily build machine learning capabilities on customer and non-customer data. “It is a shared service that provides the right data, optimises the right access patterns, and alleviates the machine learning application developer from having to manage data pipelines, storage, security and compliance,” he added. He said one of the main reasons to build ML Lake was to make it easier for applications to get access to the data they require while centralising the security controls needed to maintain trust. The common industry practice is to carefully maintain and curate a number of key datasets for ML or analytics use cases. The metadata in ML Lake is vital for model training and serving and compliance operations.","excerpt":"In 2018, Salesforce open-sourced its machine learning tool, TransmogrifAI.","categories":["AI Features"],"tags":["Salesforce AI","Salesforce Einstein"],"author_name":"Sreejani Bhattacharyya","publish_date":"2022-06-13T15:00:00","publication_year":"2022","word_count":600,"keywords":["machine learning","AI","chatbots","ML","RAG","NLP","Aim","deep learning","anomaly detection","Salesforce Einstein","analytics","Salesforce AI"],"extracted_tech_keywords":["AI","machine learning","ML","deep learning","NLP","analytics","Aim","RAG","chatbots","anomaly detection"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/salesforce-innovations-that-fuel-ai\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10051087,"title":"Robotic Startup Ishitva Raises Over $1 Million In Pre-Series A Round","content":"Ahmedabad based robotic startup Ishitva Robotic Systems (IRS) has raised a Pre-Series A round of over $1mn led by Inflection Point Ventures, one of India’s largest angel investment platforms. The funds raised will be utilised in growing sales and expanding the R&D team. Many waste management and plastic value chain players and veterans also participated, including Kamaljyot Investments Ltd (a subsidiary of Excel Industries Ltd), Gandhi Family from GRP Ltd., AVI Global Plast Private Limited, Anup Gulati, COO at Quality Maritime Provider Lda., Harinder Arya, President & Div CEO – Major Capex & Pressure Control at Joulon, Sanjay Kamlani, Serial Investor and Founder of Pangea3 and Mohan Ayyangar: Operations Lead – GDS EY India. “Ishitva’s model is particularly innovative as they are leveraging cutting edge technology to sort out the waste which is already being generated. Sorting the waste is the first step to a meaningful recycling ecosystem. Ishitva’s business vision and our ESG goals are a seamless synergy which prompted us to invest in the company,” said Mitesh Shah, Co-founder, Inflection Point Ventures. Ishitva uses AI, ML, IoT, Industry 4.0 to build efficient solutions which help in sorting complex waste. Effective sorting of waste is currently the weakest link in creating a circular value chain. Therefore, the company focuses on leveraging industry 4.0 tools to systematically sort out waste and assess the quality of recyclable waste. Jitesh Dadlani, Founder of Ishitva, said, “Indian waste management industry is a riddle, and our unlimited dream is to eliminate the manual sorting of waste which often involves humans to work in unhygienic conditions. Our vision is to install industry 4.0 solutions in 4000+ towns of the country picked up by the Swachh Bharat Mission, and we are thrilled to have IPV as a part of this journey.” “We are excited that many of the waste industry stalwarts are joining early in our journey. With IPV, we now have access to a large pool of professionals and experts, which is immensely useful at this stage for us,” said Sandip Singh, CEO of Ishitva. Ishitva currently offers a range of solutions, including SUKA (AI-powered air sorting), YUTA (AI-powered robotic sorting) and Netra AI Vision system, which help identify waste and smart bins. The AI-powered Netra identifies the waste, learns about its contents by capturing images, and then scans for recyclable material.","excerpt":"Ishitva uses AI, ML, IoT, Industry 4.0 to build efficient solutions which help in sorting complex waste.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","ai generated images","Data Science","Data Scientist","Deep Learning","Machine Learning","Robotics","Startups"],"author_name":"Victor Dey","publish_date":"2021-10-08T12:45:03","publication_year":"2021","word_count":387,"keywords":["ai generated images","Go","programming_languages:R","AI","ML","Machine Learning","programming_languages:Go","Robotics","RAG","Startups","GAN","Deep Learning","Data Science","Data Scientist","R","AI (Artificial Intelligence)","startup"],"extracted_tech_keywords":["AI","ML","RAG","R","Go","GAN","startup","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/robotic-startup-ishitva-raises-over-1-million-in-pre-series-a-round\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10049826,"title":"The Axis Bank-Manipal School of Data Analytics Announced For BFSI Personnel","content":"Manipal Global, a leading education services’ organization, launches the Axis Bank-Manipal School of Data Analytics, the country’s first job-assured online programme in data science and analytics, in collaboration with Axis Bank, India’s third-largest private sector bank. Robin Bhowmik, Chief Business Officer, Manipal Global, stated, “We are delighted to announce the establishment of the Axis Bank Manipal School of Data Analytics. This is a one-of-a-kind programme that enables individuals to transform themselves and acquire the essential technology and problem-solving abilities to land a job. This strategic agreement with Axis Bank enables us to hire data analysts with industry experience and accelerate the bank’s digital transformation.” The School of Data Analytics provides a three-month online degree that includes a two-week practicum. The first group will be trained on the fundamental principles and techniques of data science and analytics, big data, and machine learning with Python, beginning this month. Students will get the opportunity to apply their technological expertise and apply tools and concepts in a variety of scenarios. The curriculum is aimed towards recent graduates and those with up to three years of experience in technology\/analytics professions. The curriculum will be delivered online by some of Manipal’s greatest instructors. This is an excellent chance for engineers and individuals with a background in mathematics or statistics, as the course guarantees employment for its graduates. Axis Bank has recently increased its digital infrastructure investment to provide smart and secure digital banking solutions to its institutional and individual customers. The bank has been at the vanguard of digitization, from machine learning-based chatbots to the implementation of conversational online banking. The bank needs highly trained professionals who can contribute to business growth through data and insights to keep up with the increased demand for digital banking services.","excerpt":"Axis Bank has been increasing its digital infrastructure investments to provide smart and secure digital banking solutions to its institutional and individual customers.","categories":["AI News"],"tags":["Big Data","Data Analytics","data science and analytics","Machine Learning","mathematics","Statistics","Statistics for Data Science"],"author_name":"Dr. Nivash Jeevanandam","publish_date":"2021-09-27T16:41:45","publication_year":"2021","word_count":291,"keywords":["mathematics","data science","big data","machine learning","Statistics","AI","chatbots","Machine Learning","Git","Python","Aim","Statistics for Data Science","analytics","Data Analytics","Big Data","R","data science and analytics"],"extracted_tech_keywords":["AI","machine learning","data science","analytics","Aim","chatbots","Python","R","Git","big data"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/data-analytics\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":12999,"title":"Data Visualization is big in this election season and here’s the company that’s leading it","content":"Big data means big elections and Hyderabad-based Gramener, the Data Visualization & Analytics platform provider, is no stranger to the election frenzy. It was Gramener’s insightful and dynamic visuals that drew out the audiences for Times Now channel in the 2016 elections. While political campaigns and data visualization can be a tricky mix, the Gramener team tackled the odds by “learning politics and asking subtle questions”. The company forayed in visual data journalism in 2013 and has gained serious street cred in visualizing election campaign data. Some of the ripple effects of their analytically and visually rendered campaign data were – 80% primetime viewership, and eliminating uncertainty by “putting the numbers out there, there was no theorizing”. Times Now data advantage will be leveraged by its rival channel now. With the 2017 elections around the corner, the company is all set to navigate the political maelstrom with top-notch, interactive visuals for TV 18 network.  Visualization is Gramener’s core competency and S Anand, co-founder and CEO has admittedly been in the throes of live election coverage, living and breathing the election campaign data for weeks on end. One of India’s most celebrated and revered Indian data scientists, admittedly “doesn’t know much about Indian politics”. That apart, the 42-year-old CEO definitely knows the science behind it. Gramener CEO S Anand gives us the lowdown behind the stellar, eye-grabbing data-centric storytelling: Gramener CEO S Anand At the heart of their success is Gramex platform, the homegrown platform that uses an open source component and stands on several technology stack. Front end visualization Interactive visuals need to respond to changes, such as update in data, resizing of display or a change in colour. There are three kinds of visualization paradigm in vogue today: Templating: This intersperses programming logic in the middle of content. Think PHP. Handlebars is a representative library in this space. Gramener uses Underscore – that provides templating along with other utility functions. Data Binding: This approach is best represented by D3. The flexible library allows selectively binding input data to document elements and applying changes to generate and revise content. D3 works well when output needs to dynamically with changing data. This pattern has become the gold standard in data visualization. However, the D3 approach involves a lot of learning and effort from developers even to build simple visualizations. Reactive approach: This approach combines templating with data binding. Templates are updated dynamically and minimally. This approach uses a virtual screen (called “virtual DOM”) to note all the UI changes and brings the UI up to speed with the latest changes. Here’s a bit of trivia about React.js, it was born at Facebook as a set of best practices for building reactive user interfaces. Here’s what goes on in the back-end The back end needs to crunch numbers rapidly. That’s why Gramener’s platform runs on the Pandas library for data processing and makes use of Tornado framework, the scalable web application framework known for its speed in handling a large number of client requests and high volumes of traffic.  “You need a framework that serves output without causing any delay. That’s why we use Tornado framework,” he shares. Here’s a common scenario demonstrating the need for responsiveness on server side: Say for example, browser demands a result from a database and an image The web server receives the request and sends it to the database While waiting for the database response, the web server takes the next request from the browser and serves the image When the database responds, the server sends the response to the browser This ensures that the server is not idle waiting for other requests. This event-driven approach is how high-performance servers like nginx break the 10,000 requests\/second barrier. Things to watch out for in Election Data Visualization As Anand reveals, during live broadcasting nobody really conceives how a story will flow. “Around 75%-80% is planned, the rest 25% is buffer,” he lets on. The anchor has to be familiar with the interface which is truly interactive and dynamic. The visualization team has to keep a tab on what kind of interactive analysis or paths the anchor will be taking and drive the sequence of visuals accordingly. Secondly, what’s onscreen has to be updated fast, only then is it classified as Breaking News and analytics has to keep pace with it. Thirdly, unstructured data that is not machine readable, anomalies in spellings and ill-defined constituency boundaries that are reshaped and recasted add to the data cleansing job that is a “bit of a nightmare”. Autolysis – Gramener’s Exploratory Data Analysis tool Automated analysis tool was born out of the need for automating tasks such as automating insight discovery, data extraction, data cleansing and data classification among others. The tool enables discovering insights without any programming or data analysis knowledge. Anand revealed in an earlier interview to Analytics Indian Magazine that he is a “believer in machines more than people, as machines can do a lot as well, therefore what I am trying to do is seeing how much of what we are doing can be automated.” The result — the algorithms can automatically find answers to questions like: Which are the best performing branches? What drives the price of a product? Where are the exceptions to these rules? How can customers be clustered? “We are 30-35% of the way there and by the end of the year it will become a more robust process. Even tasks such as data cleansing will be a process-driven checklist of 40-45 items,” he shares.","excerpt":"Big data means big elections and Hyderabad-based Gramener, the Data Visualization & Analytics platform provider, is no stranger to the election frenzy. It was Gramener’s insightful and dynamic visuals that drew out the audiences for Times Now channel in the 2016 elections. While political campaigns and data visualization can be a tricky mix, the Gramener […]","categories":["IT Services"],"tags":["political campaigns and ai"],"author_name":"Richa Bhatia","publish_date":"2017-02-24T06:04:41","publication_year":"2017","word_count":922,"keywords":["political campaigns and ai","Go","API","TPU","AI","Scala","RAG","Ray","analytics","R","Pandas"],"extracted_tech_keywords":["AI","analytics","Ray","Pandas","RAG","TPU","R","Go","Scala","API"],"url":"https:\/\/analyticsindiamag.com\/it-services\/data-visualization-big-election-season-heres-company-thats-leading\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10021319,"title":"Workshop Alert! Hands-on Meta-Learning","content":"The Association of Data Scientists (AdaSci), the premier global professional body of data science and ML practitioners, has announced a hands-on workshop on meta-learning on March 13, Saturday. The unavailability of large datasets has turned out to be a huge problem in solving critical challenges with machine learning and artificial intelligence. As a matter of fact, deep learning’s progress often gets impeded due to the unavailability of adequate labelled data. In many cases, it becomes challenging to collect a sufficiently large number of labelled data, which inspired many research efforts on exploring ways to train robust models for various learning tasks beyond labelled data. Further, to train complex deep learning algorithms and models need high computational power. In this workshop, the attendees get to learn about meta-learning — a subfield of machine learning where deep learning models are trained with fewer data efficiently. Known as ‘learning how to learn,’ meta-learning is an exciting trend in machine learning. Register Here The full-day workshop on meta-learning will introduce topics such as few-shot learning; deep multi-task learning; parameter-level approach; zero-shot, one-shot and low-shot learning and more. The workshop will also discuss the implementation of different few-shot learning networks with PyTorch and TensorFlow, including siamese; matching; prototypical; relational; and memory augmented networks. Additionally, the workshop will allow a hands-on implementation of meta-learning methods with PyTorch and TensorFlow, including model agnostic meta-learning; meta-SGD; open AI’s reptile; domain adaptive meta-learning; LSTM meta-learner; task-agnostic meta-learning and more. Attendees will also gain a comprehensive understanding of the applications of meta-learning in the healthcare sector, robotics, chatbots, fake news, and self-driving vehicles. The participants should have a basic to moderate level understanding of Python, as well as basic knowledge of Pandas, Numpy, Scikit-Learn, TensorFlow, PyTorch and Keras. The workshop would also expect the attendees to have a nodding acquaintance with deep learning, convolutional neural network and recurrent neural network, and familiarity with Google Colab and GPU environment. Register Here Attendees also need to have a few tools like an editor to run the python programs, preferably Google Colab Notebooks and install Pandas, NumPy, Scikit-learn, TensorFlow, PyTorch and Keras. A high-speed internet connection is mandatory. Upon completing the workshop, the attendees will gain hands-on experience in meta-learning and few-shot learning and their applications in AI and ML. They will learn about the latest advancement in meta-learning and be equipped to kickstart image classification using meta-learning. Attendees will also get a certificate on hands-on meta-learning with Python. Details of the workshop: Date: March 13, 2021 Timings (Full day): 10:00 am to 5:00 pm (IST) Mode: Online Pricing: $12.99 (workshop is free for ADaSci members) Register Here","excerpt":"The Association of Data Scientists (AdaSci), the premier global professional body of data science and ML practitioners, has announced a hands-on workshop on meta-learning on March 13, Saturday. The unavailability of large datasets has turned out to be a huge problem in solving critical challenges with machine learning and artificial intelligence. As a matter of […]","categories":["Deep Tech"],"tags":[],"author_name":"Sejuti Das","publish_date":"2021-03-03T16:33:55","publication_year":"2021","word_count":435,"keywords":["data science","machine learning","artificial intelligence","Keras","AI","neural network","PyTorch","ML","deep learning","TensorFlow"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","neural network","data science","TensorFlow","PyTorch","Keras"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/workshop-alert-hands-on-meta-learning\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":1474,"title":"A Smart t-shirt Made in India surfaces in the wearable market","content":"After top clothing brands like Arrow having ventured into wearable shirts, there comes another India startup that has launched an unprecedented clothing line, amongst which the wearable T shirts are capturing interests of tech savvy audience prying on fitness goals. The other interesting technology packed clothes this startup is selling are LED t-shirts that can display any slogan you want using your smartphone and the clothes that glow. Claiming to be the first ever fitness tracking t-shirt to be sold in the market, the Hyderabad based startup, Broadcast Wearables has named its wearable t-shirt SYGNAL. If its features are to be tracked down, this T shirt like any other fitness tracker pairs up with an app to keep a check on your daily steps, distance covered by walking or running, calories burned, floors climbed etc. What can it do? Track the steps taken in the entire day Track calories burnt even in the slightest form of exercise Track floors climbed Track the distance walked or run It can navigate to desired location How it works? Loaded with a bunch of sensors packed in a small chip, the t-shirt has it placed along with the battery and Bluetooth module in the upper back portion. The Bluetooth syncs all the tracked data with the app, which can be viewed for at least three days. A soft master switch on the sleeve allows the device to be turned off. The t-shirt also allows to connect with the MAPS app on the phone to help you navigate, which is made possible by the two built-in vibration sensors sewed on both the shoulders. The sensors vibrate according to the direction the map wants you to take. Available in India for a starting range of INR 2499, it available in a couple of color variants, this combination of hardware and software can be paired with the app which is available in both android and iOS. If you are wondering about its re-usability, the SYGNAL t-shirt is washable and all the electronics embedded on it are 100% waterproof. The kickstarter is being crowdfunded on FuelADream online funding marketplace.","excerpt":"After top clothing brands like Arrow having ventured into wearable shirts, there comes another India startup that has launched an unprecedented clothing line, amongst which the wearable T shirts are capturing interests of tech savvy audience prying on fitness goals. The other interesting technology packed clothes this startup is selling are LED t-shirts that can […]","categories":["AI News"],"tags":[],"author_name":"Srishti Deoras","publish_date":"2017-02-15T06:14:29","publication_year":"2017","word_count":351,"keywords":["Go","funding","programming_languages:R","AI","programming_languages:Go","Aim","GAN","R","startup"],"extracted_tech_keywords":["AI","Aim","R","Go","GAN","startup","funding","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/smart-t-shirt-made-india-surfaces-wearable-market\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10021461,"title":"Microsoft Introduces Mathematical Framework To Tune Up Attention Architectures","content":"Recently, Microsoft Research and the University of Montreal introduced a new mathematical framework that uses measure theory and integral operators to model attention architectures in neural networks. According to the researchers, the framework is proposed to quantify the regularity; in other words, the amount of smoothness of the attention operation. The attention mechanism is the fundamental building block of neural networks like multi-layer perceptron, convolution neural network and recurrent neural network cell. The mechanism, a part of the network’s architecture, is in charge of managing and quantifying the interdependence between the input and output elements and also within the input elements. Why This Research Attention has proved to be a powerful component of modern neural networks across a variety of domains. Researchers have been working to improve this architecture. However, there is still a lack of clarity in the explanations about the mathematical properties of attention and regularities of the attention architectures. The new mathematical framework comes in the wake of this situation. The researchers stated, “In particular, we seek to understand how “close” the outputs of the attention operation are in terms of the closeness of the inputs and the parameters of the attention block.” Importance Of Regularity According to the researchers, the regularity of attention is essential for various reasons. More specifically, regularity impacts self-attention networks’ properties, such as their invertibility and the existence of infinite-depth limits. The key reasons why regularity is important: Regularity is a basic property of a function with crucial implications for tasks such as feature learning. Also, Lipschitz regularity, in particular, plays an important role in such cases.Secondly, the repeated composition of a function magnifies its regularity. Since attention is liberally used in very deep architectures, understanding the regularity of this essential building block can shine a light on the training and stability of these models.Lastly, having a precise theory allows us to make testable predictions about experiments to generate improvements and post hoc analysis of experimental results to understand better why a given behaviour was observed. Behind The Framework The mathematical framework relies heavily on linear transformations of measures modelled by Markov kernels. The idea behind this research is to show that attention is Lipschitz continuous under various assumptions, and to do so, the researchers introduced the modelling paradigm for attention based on measure theory and integral operators. They demonstrated that the attention operation is Lipschitz continuous and provided an estimate of its Lipschitz constant. Lipschitz continuity is mainly used to improve the state-of-the-art in several deep learning topics such as generative models and robust learning. The researchers also assessed the impact of these regularity results on practical applications of attention such as cross-attention, robustness and token-level perturbations in NLP, and sophisticated extensions to the transformer architecture. Wrapping Up The researchers studied how regularity can help certain applications by providing robustness to the learned representations. They showed several benefits of using this mathematical framework: The mathematical framework is consistent with the usual definition, and it captures the essential properties of attention. The framework showed the resulting representation is Lipschitz continuous concerning the output semantic space.The framework provides a potential mathematical basis for the robustness of transformers.The modelling could be used to derive predictions of the distance between self-attention networks’ contextual embeddings as a function of the context to test this hypothesis.It could also be used to design better model components, such as input embedding spaces that reduce the regularity mismatch for specific perturbations that are highly irregular. Read the paper here.","excerpt":"Recently, Microsoft Research and the University of Montreal introduced a new mathematical framework that uses measure theory and integral operators to model attention architectures in neural networks. According to the researchers, the framework is proposed to quantify the regularity; in other words, the amount of smoothness of the attention operation.  The attention mechanism is the […]","categories":["AI Trends"],"tags":["Artificial Neural Network","Convolutional Neural Network","Deep Neural Networks","Neural Networks","Recurrent Neural Network"],"author_name":"Ambika Choudhury","publish_date":"2021-03-05T14:00:00","publication_year":"2021","word_count":579,"keywords":["Recurrent Neural Network","TPU","attention mechanism","AI","neural network","Artificial Neural Network","ai_frameworks:Transformers","Transformers","NLP","deep learning","Deep Neural Networks","Convolutional Neural Network","transformer architecture","R","Neural Networks"],"extracted_tech_keywords":["AI","deep learning","neural network","NLP","Transformers","TPU","R","transformer architecture","attention mechanism","ai_frameworks:Transformers"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/microsoft-introduces-mathematical-framework-to-tune-up-attention-architectures\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10051939,"title":"Apple Launches AI\/ML Residency Program, Here’s How To Apply","content":"Apple recently announced the launch of its new AI\/ML Residency Program for graduates looking to develop their career in Artificial Intelligence & Machine Learning. Apple’s AI\/ML residency is a year-long program that invites experts in various fields to apply their own domain expertise to innovate and build revolutionary machine learning and AI-based products and experiences. The residents will have the opportunity to attend machine learning and AI courses, learn from an Apple mentor, collaborate with fellow residents, gain hands-on experience working on high-impact projects, publish in premier academic conferences, and partner with Apple teams across hardware, software, and services. Key qualifications for applying to the program include: Proficiency in a programming language such as Python, Objective-C, C++, Swift, and\/or R. Candidates should have experience completing a moderate-sized software project. Demonstrated ability to connect and collaborate with others. Ideal candidates will also have experience in one of the following: HCI\/design, accessibility, natural language processing, neuroscience, biology, biomedical engineering, information retrieval\/search, computer vision, robotics, autonomous systems, applied math\/statistics, optimisation, physics, mechanical engineering, electrical engineering, signal processing, speech, data mining, econometrics, cognitive science, behavioural economics, behavioural science, psychology, learning sciences, learning analytics, or psychometrics. Thesis, capstone project, internship, co-op, and\/or industry experience in these fields will qualify. Educational requirements include having a graduate degree in a STEM field or equivalent industry experience in software engineering. While applying for the program, a cover letter as part of your application explaining the applicant’s research interests and why they would like to join the AI\/ML residency program is preferred. Apple’s AI\/ML residency is a year-long program starting in July 2022. Residents must be available to work full-time (40 hours\/week) from July 2022 to July 2023. The current application deadline stands to be December 15, 2021. You can apply using the link here.","excerpt":"Apple’s AI\/ML residency is a year-long program that invites experts in various fields to apply their own domain expertise to innovate and build revolutionary machine learning and AI-based products and experiences.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","Apple","covid-19","Data Science","Data Scientist","Deep Learning","Machine Learning","Python"],"author_name":"Victor Dey","publish_date":"2021-10-20T10:50:39","publication_year":"2021","word_count":296,"keywords":["artificial intelligence","machine learning","covid-19","AI","Apple","ML","Machine Learning","computer vision","Python","C++","ViT","analytics","Deep Learning","Data Science","Data Scientist","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","computer vision","analytics","Python","R","C++","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/apple-launches-ai-ml-residency-program-heres-how-to-apply\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10068914,"title":"How to use a pre-trained Random Forest model for transfer learning?","content":"Pretrained models in machine learning is the process of saving the models in a pickle or joblib format and using them to make predictions for the data it is trained for. Saving the models in the pipeline facilitates the interpretation of model coefficients and taking up predictions from the saved model weights and parameters during model deployment for production. So this article provides a brief overview of how to implement a random forest classifier model and save it in a pickle format and use the pretrained model for predictions during production. Table of Contents An introduction to pre-trained models Building a classification model from scratchSaving the model in pickle formatLoading the saved modelObtaining predictions from the loaded modelSummary An introduction to pre-trained models Pretrained models are the models obtained after maturing through various processes of a typical machine learning model lifecycle. Pretrained models are the models developed to obtain predictions for problems of similar kinds and help us to save huge training time. So for similar kinds of data, the pretrained models can be loaded at the start and later modified as required or the same model which is pretrained for similar kinds of features can be used to obtain predictions. Note:- Pretrained models may always not be accurate and may be biased to similar kinds of features in data. So in general it is advisable to understand if the pretrained models are biased towards any particular features before using them. Building a random forest classification model from scratch Here we have used a health-care dataset to build a random forest classification model from scratch and also the required preprocessing steps to adhere are shown below. So now let’s look into the steps involved in building a random forest classification model. Are you looking for a complete repository of Python libraries used in data science, check out here. So first let us visualize the top 5 entries of the dataset. Data Preprocessing So the above dataset was checked for null values and the corresponding features of null values were appropriately imputed for correct values. So from the dataset id and gender features were removed as they seemed to be less significant and did not possess any important information. df=df.drop(['id','gender'],axis=1) So now the categorical features of the dataset were encoded to numerical features using the LabelEncoder of the scikit module as shown below. from sklearn.preprocessing import LabelEncoder le=LabelEncoder() ## creating a label encoder instance for fitting df['ever_married']=le.fit_transform(df['ever_married']) df['work_type']=le.fit_transform(df['work_type']) df['Residence_type']=le.fit_transform(df['Residence_type']) df['smoking_status']=le.fit_transform(df['smoking_status']) So once the encoding was complete the dataset was again visualized to understand how LabelEncoder has encoded the categorical features present in the data. Now as we have appropriate preprocessed data let’s proceed ahead with splitting the data. Splitting the data The preprocessed data is now being split using the scikit learn module as shown below along with validating the number of records for training and testing the model. from sklearn.model_selection import train_test_split X_train,X_test,Y_train,Y_test=train_test_split(X,y,test_size=0.2,random_state=42) print('Number of records for train',X_train.shape) print('Number of records for test',X_test.shape) print('Number of records for train',Y_train.shape) print('Number of records for test',Y_test.shape) Implementing the random forest model Using the split data a random forest classifier model was implemented as shown below along with evaluating various parameters like accuracy score and Area Under Curve (AUC) to determine model performance and validate any signs of overfitting. The steps involved in implementing a random forest model and evaluating the parameters are shown below. from sklearn.ensemble import RandomForestClassifier rfc_class=RandomForestClassifier(random_state=42) rfc_base=rfc_class.fit(X_train,Y_train) rfc_pred=rfc_base.predict(X_test) Now the prediction of the base random forest model was used to obtain the classification report and also to evaluate the AUC score. from sklearn.metrics import classification_report,accuracy_score,roc_auc_score print('Classification report \\n',classification_report(Y_test,rfc_pred)) y_train_pred=rfc_base.predict(X_train) y_train_prob=rfc_base.predict_proba(X_train)[:,1] y_test_prob=rfc_base.predict_proba(X_test)[:,1] print('Train Accuracy',accuracy_score(Y_train,y_train_pred)) print('Train AUC',roc_auc_score(Y_train,y_train_prob)) print() print('Test Accuracy',accuracy_score(Y_test,rfc_pred)) print('Test AUC',roc_auc_score(Y_test,y_test_prob)) So from the model developed, we can see that the model’s testing parameters are lesser than the training parameters but according to the classification report, the model is possessing an accuracy score of 94%. Now let us look into how to save this base model in pickle format. Saving the model in pickle format In general, the machine learning models are more likely to be saved in a pickle format for easy saving and loading of the saved model parameters. So let us look into the steps involved in saving a machine learning model in pickle format. import pickle with open('rfc_model_pkl', 'wb') as files: pickle.dump(rfc_base, files) So here the pickle module has been imported to the working environment and a pickle object is created with writable permission operations and the base model developed is dumped in a pickle file format in the pickle object created. So the pickle object created can be checked in the working environment where it will be saved in a pkl format. Loading the saved model Now let’s read the saved pickle file in the working environment by following the steps mentioned below. # load saved model with open('rfc_model_pkl' , 'rb') as f: rfc_pretrained = pickle.load(f) So here the pickle file created is opened in a readable format (rb) and the load() function of pickle is used to obtain the pretrained model into the working environment. Obtaining predictions from the saved model So now the pretrained model can be used to obtain predictions for a random set of parameters, that is passed on to the pretrained model in the same order as the original dataset. The steps to follow for the same are listed below. rfc_pretrained.predict([[55,0,1,0,2,0,107.93,42,3]]) rfc_pretrained.predict([[81,1,1,1,3,0,100,35.7,3]]) So in this way, we have to pass random features in the respective order of the data frame and obtain predictions for the pretrained model. So in the later stage, this pretrained model is used to evaluate various parameters as shown below. y_pred_pretrained=rfc_pretrained.predict(X_test) print('Classification_report of the pretrained model \\n',classification_report(Y_test,y_pred_pretrained)) As the classification report of the pretrained model was obtained other parameters of the pretrained model were also evaluated as shown below. y_train_pred=rfc_pretrained.predict(X_train) y_train_prob=rfc_pretrained.predict_proba(X_train)[:,1] y_test_prob=rfc_pretrained.predict_proba(X_test)[:,1] print('Training Accuracy of pretrained model',accuracy_score(Y_train,y_train_pred)) print('Training AUC of pretrained model',roc_auc_score(Y_train,y_train_prob)) print() print('Test Accuracy of pretrained model',accuracy_score(Y_test,y_pred_pretrained)) print('Test AUCo f pretrained model',roc_auc_score(Y_test,y_test_prob)) Summary So this is how a machine learning model in real-time is built from scratch and saved as standard model formats like pickle and later loaded into working environments to take up predictions for similar kinds of features. Pickle file formats are memory friendly and it provides easy writing and reading operations of the instance created and facilitates obtaining predictions and evaluation of various parameters using the pretrained models.","excerpt":"Using pre-trained machine learning models to evaluate parameters","categories":["AI Trends"],"tags":["Deep Learning","Machine Learning","Randomforest","Transfer Learning"],"author_name":"Darshan M","publish_date":"2022-06-13T18:00:00","publication_year":"2022","word_count":1055,"keywords":["data science","Go","machine learning","programming_languages:R","AI","Randomforest","Machine Learning","programming_languages:Go","Python","programming_languages:Python","Deep Learning","Transfer Learning","R"],"extracted_tech_keywords":["AI","machine learning","data science","Python","R","Go","programming_languages:Python","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/how-to-use-a-pre-trained-random-forest-model-for-transfer-learning\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":38234,"title":"5 Mid-Level Data Analytics Job Roles You Can’t Miss","content":"Analytics India Jobs offers a new platform for the data science community in India by bringing to the fore latest job openings across the country. Here are few Data Analytics job openings across top 4 metros – Mumbai, Hyderabad, Bangalore and Gurgaon to apply right away: Data Scientist @ Shortlist, Mumbai & Hyderabad Shortlist works in India and East Africa to help growing enterprises hire based on skills and potential, rather than pedigree. Products such as customised chatbots would require decent understanding of machine learning algorithms and programming. Responsibilities: Work closely with the CTO and leadership to solve statistical and business problems Develop & code production-grade novel algorithms for our business-experimentation platform. Manage, Develop, maintain and market the business experimentation suite at Fractal Conduct research and prototyping innovations, data and requirements gathering, solution scoping and architecture Collaborate effectively with internal stakeholders and cross-functional teams to solve problems, create operational efficiencies, etc. Consult clients and client facing teams on advanced statistical and machine learning problems, especially in the areas of marketing and designing experiments Test various machine learning and analytical tools, especially in the big data space, to scale prototypes to production-grade systems Provide solutions but not limited to: Customer Segmentation & Targeting, Propensity Modeling, Churn Modeling, Lifetime Value Estimation, Forecasting, Recommender Systems, Modeling Response to Incentives, Marketing Mix Optimization, Price Optimization. Apply here Analyst @ Mastercard, Gurgaon As an analyst, the candidate will work closely with Data & Services Advanced Analytics teams and external clients around the world to architect, develop, and maintain advanced reporting and data visualisation capabilities on large volumes of data in order to support consulting projects, model development, model scoring, and campaign measurement. And, translate business requirements into tangible solution specifications and high quality, on time deliverables. Requirements: B.Tech or Mathematics. M.S. preferred. 2-5 years of relevant experience in hands-on data programming, querying, data mining and report development using large volumes of granular data to deliver business intelligence and custom reporting solutions in a Microsoft SQL Server environment. Strong hands-on experience on SQL, Tableau, SSIS. Analytical\/Problem Solving. Relevant retail and payments industry experience a plus. Apply here Role In Analytics @ Max Life Insurance, Gurgaon Max Life envisions to be the most admired life insurance company in India by securing the financial future of customers. Max Life Insurance amongst the top 15 BFSI companies to work for in India. Requirements: BE\/BTech, MTech, MSc. in Computer science\/Information Technology\/ ECE or other quantitative disciplines from tier 1\/ tier 2 institutes. 2-7 years of hands on work experience in NLP, AI, Machine Learning, Deep Learning Should have developed and deployed NLP based solutions with R( tm, korPus, etc) or  Python( NLTK, Scikit-learn and Spacy Strong familiarity with RNN, LSTM, word2vec and other embeddings, Keras\/Tensor flow. Knowledge of Chabot platforms and development and deployment of at least 1 NLP based Chatbot or other AI, machine learning, or NLP technologies Working with Speech to Text and conversational analysis tools \/Api, such as Dialogflow is a plus Working knowledge of Image matching, and understanding applications of voice and video data is a plus Experience in working with cloud infrastructure and high-performance computing (GPU\/TPU) be a plus. Apply here Digital Data Analyst @ DG7 Solutions Pvt.Ltd, Mumbai DG7 Solutions is an analytics-focused, research-based consulting partner for medium and large organisations offering end-to-end Digital Marketing Solutions. Requirements Work with internal stakeholders to determine ongoing business intelligence needs in order to design, develop and produce ongoing reports and dashboards to support the business. Manage projects related to digital analytics from start to finish including data integration, report and dashboard automation, product analytics and insight, data collection, and product optimisation Assists in identifying opportunities that are most effective at driving conversions, revenue, ROI, and scale across Digital Marketing programs. Build automated dashboards for digital, marketing & customer level insights – to be shared and championed with key stakeholders throughout the business. Web Analytics: In addition to setting up standard web reports, the ideal candidate will continually mine Google Analytics Developing new skills and sharing your knowledge with the Data Team SQL, R, Python or PySpark Apply here Data Analyst @ Access Analytics Solutions Pvt.Ltd, Bangalore Access Analytics International provides advanced analytical tools and services for engineering and data analysis. A data analyst will manage data sources, database and movement of data and data organisation across different systems. Requirements: Bachelors\/PGDM\/ Masters degree in Information Systems \/Maths\/ Statistics\/\/ Finance\/ Engineering \/Business Analytics Actuaries\/FRM\/CFA\/CQF\/PRM certification would be a plus 0-2 years of experience in database or project management Should have hands-on experience of designing database schemas SQL and at least one of R\/Python\/VBA Experience in handling large structured and unstructured datasets Exposure to tools\/platforms – Hadoop ecosystem and DB systems Good hands-on expertise in Tableau and ETL tools. Apply here","excerpt":"Analytics India Jobs offers a new platform for the data science community in India by bringing to the fore latest job openings across the country. Here are few Data Analytics job openings across top 4 metros – Mumbai, Hyderabad, Bangalore and Gurgaon to apply right away: Data Scientist @ Shortlist, Mumbai & Hyderabad Shortlist works […]","categories":["AI Hirings"],"tags":["data analytics jobs","data science jobs in India","Tableau"],"author_name":"Ram Sagar","publish_date":"2019-04-25T08:42:06","publication_year":"2019","word_count":791,"keywords":["data science","scikit-learn","machine learning","Tableau","data analytics jobs","AI","Keras","NLTK","data science jobs in India","NLP","deep learning","analytics","spaCy"],"extracted_tech_keywords":["AI","machine learning","deep learning","NLP","data science","analytics","Keras","scikit-learn","spaCy","NLTK"],"url":"https:\/\/analyticsindiamag.com\/ai-hiring\/5-mid-level-data-analytics-job-roles-you-cant-miss\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10048268,"title":"AWS Announces Amazon S3 Plugin For PyTorch","content":"Recently, AWS has announced the release of the Amazon S3 plugin for PyTorch — an open-source library built to be used with the deep learning framework PyTorch for streaming data from Amazon Simple Storage Service (Amazon S3). With this feature available in PyTorch Deep Learning Containers, one can take advantage of using data from S3 buckets directly with PyTorch dataset and data loader APIs without needing to download it first on local storage. It also provides a way to transfer data from Amazon S3 in parallel when needed to get maximum performance without worrying about thread safety or multiple connections to Amazon S3. You can also stream data from .zip or .tar archives and shuffle the dataset within or across the shards as required. The Amazon S3 plugin for PyTorch offers the following benefits: Support for both map-style and iterable-style dataset interfaces – PyTorch supports two different types of datasets. In addition, the Amazon S3 plugin for PyTorch also provides the flexibility to use either map-style or iterable-style dataset interfaces based on your needs: Map-style dataset – Represents a map from indexes or keys to data samples. It provides random access capabilities.Iterable-style dataset – Represents an iterable over data samples. This type of dataset is particularly suitable for cases where random reads are expensive or even improbable and where the batch size depends on the fetched data. Support for various data formats – Training data can be in a variety of different formats, such as CSV, Parquet, and JPEG. This plugin is file-format agnostic and presents objects in Amazon S3 as a binary buffer (blob). Thus, you can apply any additional transformations to the data received from Amazon S3. Support for shuffling – In deep learning, you may need to shuffle data across and within shards to reduce variance. This plugin provides a way to shuffle data in-memory within shards using ShuffleDataset or across shards by providing the input parameter shuffle_urls while extending S3IterableDataset. One can find the configuration, library and detailed information here. A few days earlier,  Amazon Web Services announced the general availability of Amazon FSx for NetApp ONTAP, a new storage service that allows customers to launch and run complete, fully managed NetApp ONTAP file systems in the cloud for the first time.","excerpt":"One can take advantage of using data from S3 buckets directly with PyTorch dataset and data loader APIs without needing to download it first on local storage.","categories":["AI News"],"tags":[],"author_name":"kumar Gandharv","publish_date":"2021-09-13T18:15:00","publication_year":"2021","word_count":376,"keywords":["API","AWS","PyTorch","AI","cloud_platforms:AWS","RAG","deep learning","ai_frameworks:PyTorch","cloud_platforms:Amazon Web Services","R"],"extracted_tech_keywords":["AI","deep learning","PyTorch","RAG","AWS","R","API","ai_frameworks:PyTorch","cloud_platforms:AWS","cloud_platforms:Amazon Web Services"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/aws-announces-amazon-s3-plugin-for-pytorch\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":35078,"title":"Deep Dive: Unpacking DBS Asia Hub 2’s Hackathon Approach &#038; Digital Agenda","content":"DBS Asia Hub 2’s ‘Hacker in Her’ hackathon programme for women has become very popular It’s a bank that is known for shifting gears and has digital transformation at its core. At the heart of its complete digital revamp was a vision for emerging tech and reimagining operations which took banks across the globe by a storm. Even before machine learning and artificial intelligence became pervasive, Singapore’s DBS Bank geared for an industry-level disruption by actively focusing on inspiring its workforce from the ground up and creating a culture of innovation. On the other end of the spectrum, the bank forged ties with FinTech companies and universities to innovate and be a part of the wider ecosystem. In 2018, the bank’s digitisation agenda won it two global awards — the ‘World’s Best Digital Bank’ and ‘World’s Best SME Bank’, making it the first Asian and Singapore bank to win two awards in the global category at the prestigious Euromoney Awards for Excellence 2018. With Asia becoming the engine of growth, DBS Bank looked to India to strengthen its position and turbocharge digital transformation by setting up its first technology development centre outside Singapore in Hyderabad in 2016. Over the last two years, DAH2, as it is widely known, has been successful in reimagining banking by addressing a range of consumer and technological demands and coming up with innovative and relevant digital solutions for its customers. The tech hub that focuses on big data, AI, automation and cloud has also built over 200 APIs. Mohit Kapoor, Head, DBS Asia Hub 2 Analytics India Magazine caught up with Mohit Kapoor, Head, DBS Asia Hub 2 to understand his technological direction and how he successfully pivoted DAH2 from a pure cost centre into a value centre. At the heart of its transition has been the key focus on product development, R&D and industry connect that helps it strengthen the tech capabilities and support core banking operations. “DAH2 has been at the centre of the digital transformation journey for the bank. We are focused on technology stack transformation, as much as on culture and workplace transformation,” shared Kapoor. DBS Bank, billed as one of the most futuristic banks has automated most of the core banking processes like loan processing, card cancellation\/ replacement through chatbots. Also, chatbots handle around 80% of our queries on the Digibank app. DAH2 Leads In Tech Thought Leadership By Building An Ecosystem At DAH2, the workforce works on products like Digibank, APIs, payments, automated loans and building the ecosystem. They continue to build customer-centric digital banking solutions using the latest technology stack, AI, cloud and mobile. The centre focuses on adoption of agile, continuous learning programs, industry connects to be able to drive tech thought leadership. “Our workplace transformation focuses on building Joyspaces – DBS’s innovative activity-based working workplace design. By focusing on these, DAH2 has been able to contribute greatly to the technical capabilities required for the future of banking,” said Kapoor. Another key aspect at DAH2 is the focus on continuous learning and innovation programs that expose the workforce to diverse skills like Cloud, IoT, AI, Machine Learning capabilities and more. The hub also has well-crafted learning programs to help upskill the workforce. “We believe in rewarding employees for learning and for experiments. Tech Design Forum is a platform to share experiments and learning within the tech clubs at DAH2,” he said, talking about promoting a culture of innovation by allowing them to run experiments.  In an attempt to foster an ecosystem of developers, the centre also hosts several tech meetups, case in point the Python meetup, open source meetup, where engineers from all around the city work with in-house engineers on open source projects. Another key way of building an ecosystem is reaching out to industry partners. One of the key highlights at DAH2 is the sharp focus on AI and as part of this, the hub conducts a workshop, known as the Makers’ workshop, where industry speakers conduct specialised workshops. The centre also runs a guest speaker series — Imaginarium, wherein startup founders and tech giants talk about emerging trends and share their experiences. Building A Strong Talent Pipeline Through Hackathons Under Kapoor’s leadership, talent, innovation and continuous learning have emerged as key focus areas for DAH2. The hub is known for not just building excellent banking solutions but its startup culture and offering a rewarding career trajectory. “We will continue to encourage experiments and embrace the best of technologies. We have been able to hire the best tech talent across the country,” shared Kapoor. Today, DAH2 has around 1500 engineers focusing on data, cloud, AI, automation and more. The technology hub has also emerged as a frontrunner in reimaging its people strategies and building a through its hackathons. Today, DBS’s Hack2Hire has emerged as a strong hiring programme and is very popular amongst college campuses and the tech fraternity. With more than 13000+ candidates participating in each of these hackathons, DAH2 has built a pipeline of diverse talent. “At DBS, we are building a 26,000 startup which thrives on ideas, experiments, learning and our PRIDE values (purpose driven, relationship-led, innovative, decisive, everything fun).  I personally think that Hack2Hire for campus and experienced candidates, as well as those designed especially for women in tech, play a vital role in connecting with people who have the startup mindset,” shared Kapoor, stressing how the Hack2Hire program gave the tech hub a competitive edge in its transformation journey. During the hackathon, the participants spend a day with the tech mentors and design PoCs (proof of concept) on problem statements, thereby encouraging out-of-the-box thinking. Recently, DBS announced the sixth edition of the Hack2Hire programme in India. Through the initiative, DBS aims to hire over 100 professionals in emerging technologies such as cloud computing, AI, machine learning and Big Data. In addition to his, they also have a ‘Hacker in Her’ hackathon programme for women and it is open to students as well as experienced candidates. Future Roadmap DAH2 has been successful in building a culture of experimentation which reflects in the grass root innovation, our regular workshops and industry contacts. The tech hub’s chief highlights have been investing in emerging technologies, robust talent pipeline and also constructing an outside-in perspective model which helps in supporting core banking operations. In closing, Kapoor shared the centre will focus on building a startup culture which will help shape not just great banking solutions and embrace the best of technologies.","excerpt":"It’s a bank that is known for shifting gears and has digital transformation at its core. At the heart of its complete digital revamp was a vision for emerging tech and reimagining operations which took banks across the globe by a storm. Even before machine learning and artificial intelligence became pervasive, Singapore’s DBS Bank geared […]","categories":["IT Services"],"tags":[],"author_name":"Richa Bhatia","publish_date":"2019-02-18T07:46:23","publication_year":"2019","word_count":1076,"keywords":["machine learning","artificial intelligence","AI","chatbots","cloud computing","RAG","Python","Aim","analytics","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","analytics","Aim","RAG","chatbots","cloud computing","Python","R"],"url":"https:\/\/analyticsindiamag.com\/it-services\/deep-dive-unpacking-dbs-asia-hub-2s-hackathon-approach-digital-agenda\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10171754,"title":"From MIT Dropout to Billionaire: How Alexander Wang Built Scale AI","content":"When Alexander Wang founded Scale AI in 2016, the AI landscape was still in its early stages. The data labelling startup, driven by Wang’s vision, quickly became a key player in the AI landscape, growing to over 1,500 employees and securing crucial partnerships across industries and governments. Now, with Meta’s $14.3 billion investment in Scale AI for a 49% stake, and Wang’s shift to Meta’s SuperIntelligence Lab, the journey of this 28-year-old billionaire is nothing short of remarkable. This transition was confirmed through a company statement. While Wang has stepped down from his role at Scale AI, he will continue to maintain a position on the company’s board of directors. But, Who Exactly is Alexander Wang? Wang founded Scale AI at 16 after dropping out from MIT. In a parting note to his employees, Wang highlighted that this was a period when AI started gaining traction—DeepMind had just launched AlphaGo, and Google had rolled out TensorFlow. Yet, the industry was still in its early stages. Scale AI provides data labelling and model evaluation for AI training and serves clients such as Meta, OpenAI, Microsoft, and the US defence department. He began his entrepreneurial journey at San Francisco’s renowned startup accelerator, Y Combinator, in the Summer 2016 batch, along with his co-founder Lucy Guo. Reflecting on the early days of building Scale AI, Wang, in a podcast hosted by South Park Commons, said, “The first half of our batch…was probably…[what] we refer to as a squiggle—a lot of existential angst; you don’t really know what you’re doing with your life.” However, despite the uncertainty, he believes it all came together. “Ultimately, we built Scale…It was like a perfect storm; a lot of serendipity,” he said. Wang explained that the core idea was simple: in the future, human computation would be orchestrated much more dynamically, but the necessary API was still lacking. Scale AI is just one of the many AI startups born out of Y Combinator. His time at YC and, eventually, Silicon Valley, led to him making powerful connections in the tech and venture capital ecosystem. What’s in it for Meta? This acqui-hire strengthens Meta’s position in the AGI race, in addition to its previous significant acquisitions in social media, like Instagram and WhatsApp, and VR, such as Oculus. “Wang is joining Meta to work on Meta’s AI efforts. He will continue to serve as a director on the Scale board of directors and support the company’s ongoing work to unlock the power of AI and keep human values at the forefront,” the company said in a statement. Meanwhile, the company announced Jason Droege, who has extensive experience at Uber Eats and AXON, as the new interim CEO. When The Information first broke the news about this deal, an investor highlighted the advantages this structure provides for Meta. “It’s a very interesting structure, where Meta effectively gains control of Scale AI, bypasses antitrust review, and acquires top talent, including Alexander Wang, all while paying only half of the cash needed to buy the entire company,” he wrote. Notably, Axios previously reported that Meta is restructuring its AI operations. Through this, the company is creating two teams—AI Products, led by Connor Hayes, which focuses on tools like Meta AI assistant; and AGI Foundations, co-headed by Ahmad Al-Dahle and Amir Frenkel, which works on foundational technologies like Llama models. A subgroup from Facebook Artificial Intelligence Research (FAIR) will join the AGI team. As per reports, even though no job cuts are planned, leadership changes are to be expected. Big tech companies are partnering with the top players in the field to ensure they can harness these technologies to their advantage. Historically, Microsoft’s $13 billion investment in OpenAI, Amazon’s $8 billion stake in Anthropic, and Google’s backing of Anthropic have effectively secured much of the top-tier model infrastructure within their tightly integrated ecosystems. In addition, Meta’s partnership with Scale AI, which has ties to the US government, improves its institutional credibility and supports initiatives like the Defense Llama project—a version of Meta’s model for national security applications. The data collection and labelling market is growing rapidly and is expected to reach $29.2 billion by 2032, driven by the increasing demand for high-quality labelled data for AI and machine learning. Scale AI is one of the major players in the market. It already works with tech giants like Microsoft and OpenAI. As per reports, the company was recently valued at approximately $14 billion in a 2024 funding round supported by Meta and Microsoft. Moreover, it was reportedly in discussions earlier this year for a tender offer that would value it at $25 billion. Unlike its competitors, Meta does not offer cloud computing services, making the investment in Scale AI essential for improving its AI capabilities and access to quality labelled data.","excerpt":"He began his entrepreneurial journey at San Francisco’s startup accelerator, Y Combinator, in 2016.","categories":["Global Tech"],"tags":["AI (Artificial Intelligence)"],"author_name":"Aditi Suresh","publish_date":"2025-06-13T17:57:45","publication_year":"2025","word_count":797,"keywords":["Anthropic","Go","Meta AI","machine learning","artificial intelligence","OpenAI","AI","cloud computing","TensorFlow","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","OpenAI","Anthropic","Meta AI","TensorFlow","cloud computing","R","Go"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/from-mit-dropout-to-billionaire-how-alexander-wang-built-scale-ai\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10103720,"title":"How NVIDIA is Helping Foxconn Unleash its EV Ambitions","content":"Foxconn, the largest electronics manufacturer and the supplier of Apple, has taken a natural drift towards producing electric vehicles (EVs) this year. NVIDIA is helping them bring this to life, alongside developing automated and autonomous vehicle platforms. “This partnership will provide scale for volume manufacturing to meet growing demand for the NVIDIA DRIVE platform,” said its chief Huang Jensen, in its recent record-breaking earnings. He said that Foxconn will be using the NVIDIA DRIVE Hyperion, and sensor architecture for its electric vehicles. Further, he said the company would be a Tier 1 manufacturer producing electronic control units based on NVIDIA DRIVE Orin for global automotive vehicles. This also comes in the backdrop of the NVIDIA DRIVE operating system receiving safety certification from TUV SUD. “One of the most experienced and rigorous assessment bodies in the automotive industry,” said Jensen, saying that their platform meets the higher standards required for autonomous transportation. The alliance of NVIDIA and Foxconn is unique, and the partnership initiates the establishment of AI factories, utilising NVIDIA’s GPU computing infrastructure explicitly designed for processing, refining, and converting extensive datasets into valuable AI models and tokens. Powering AI Innovation Together “Foxconn, the world’s largest manufacturer, has the expertise and scale to build AI factories globally. We are delighted to expand our decade-long partnership with Foxconn to accelerate the AI industrial revolution,” Huang, CEO of NVIDIA, said about their collaboration with Foxconn. Foxconn is also leveraging NVIDIA Omniverse for the manufacturing of EVs. With the support of an extensive partner network, manufacturers strengthen their workflow to plan, build, operate, and optimize their factories using a range of NVIDIA technologies. Why is everyone making EVs? Apple, Oppo, Xiaomi, Samsung, Huawei, and almost everyone are looking to make EVs. Now, Foxconn also joins that list. Driven by government incentives promoting the production of eco-friendly vehicles, these companies have opted to transition into the automotive sector. All of this comes in the backdrop of the United Nations issuing a red alert on climate change, and a lot of transport aggregators have also made commitments to achieve its sustainable goals. For instance, Uber said that it would be going carbon neutral by 2025. In light of these events, a lot of smartphone manufacturers and electronic manufacturers are transitioning to bridge the supply and demand gap, and EVs just make it more lucrative than ever. NVIDIA is making it much easier for all of them.","excerpt":"Electronics manufacturers globally are enhancing digitalisation with NVIDIA’s AI, 3D, simulation, and autonomous tech.","categories":["Global Tech"],"tags":[],"author_name":"Sandhra Jayan","publish_date":"2023-11-27T20:00:04","publication_year":"2023","word_count":401,"keywords":["Go","programming_languages:R","AI","innovation","programming_languages:Go","RAG","R","GPU computing"],"extracted_tech_keywords":["AI","RAG","GPU computing","R","Go","innovation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/how-nvidia-is-helping-foxconn-unleash-its-ev-ambitions\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10113057,"title":"How Epsilon is Navigating DE&#038;I in Tech","content":"The journey to equality has been a bit of a tedious trek for the LGBTQ+ community in tech (and elsewhere). While some big names in the industry have opened up about their identities, issues like workplace safety and acceptance are real and still hold back many from being themselves. The evolution within the industry, albeit slow, has been happening with organisations acknowledging the pivotal role diversity, equity, and inclusion (DE&I) play in shaping a more conducive work environment. Sharing a similar story is Joseleen Princy C, a senior business system analyst at Texas-based advertising and technology company Epsilon. Beginning her career as a developer at HCL, she later moved to Cognizant before joining Epsilon where she has been working for almost two years. A computer science engineer from Chennai, Joseleen identifies as a transwoman and has had her share of troubles with social stigma and workplace biases. But Epsilon provided her with a safe space for coming out and sharing her story. Internally, she collaborated with the DE&I team to conduct sessions, participated in events, and engaged in media campaigns that reached a broader audience, fostering understanding and acceptance. These efforts collectively contributed to creating a more inclusive and supportive environment for LGBTQ+ individuals within the organisation. “Epsilon gave me the platform to express myself and advocate for the rights and well-being of my community. I’ve seen a positive impact on the personal and professional growth of my colleagues. Through networking, education, and mentorship, we are empowering our community to thrive in all aspects of their lives,” Princy told AIM in a candid conversation last week. Epsilon’s D&I Initiatives According to Princy, the comprehensive support provided by Epsilon during her onboarding, including insurance and medical benefits, has been noteworthy. Of particular significance is the personalised assistance with bank-related documentation. Epsilon took charge of this process, creating a specialised system for efficient compensation processing. “One thing that I would like to highlight was my bank-related documentation, which Epsilon took over. They generated a special system for me to get my compensation. They also ensured my bank procedures and signatures were in place. This personal attention made me feel valued as an employee and has fostered a sense of trust and transparency within the organisation,” said Princy. Epsilon aims to create a robust culture of inclusivity, free from biases, to give a respectful and welcoming environment to all team members. The company’s DE&I initiatives are overseen by a dedicated DE&I Council, led by the chief diversity officer, and include Employee Resource Groups (ERGs) that focus on various diversity dimensions. This commitment is reinforced through mandatory periodic training and workshops covering various topics such as POSH, unconscious bias, ethics, and privacy. “These training ensure that all employees are equipped with the understanding, knowledge, and skills to promote a culture of inclusivity in their daily interactions,” said Princy. Besides imparting practical knowledge, these initiatives promote empathy and understanding among colleagues, fostering a harmonious and collaborative work environment that reflects the organisation’s dedication to fostering an inclusive workplace. Beyond training, the company’s efforts encompass recruitment, talent acquisition, learning and development, featuring programs like unconscious bias training. Employee engagement is fostered through surveys and events celebrating diversity. Additionally, Epsilon engages with communities and tracks progress using metrics like workforce representation and participation in DE&I programs. Addressing Data Gaps for Trans and Non-Binary Individuals The lack of data on non-binary and transgender individuals stems from challenges in existing gender categorisations. Traditional binary classifications prevalent in data collection often render non-binary and transgender individuals invisible, with estimates suggesting that only 0.1–2% of people identify as transgender or diverge from the cisgender classifications. These groups face exclusion and discrimination across various life aspects, but limited evidence exists due to the pervasive use of binary gender classifications in data collection instruments. The minority status of this group, coupled with smaller sample sizes, increases the risk of personal identification in datasets, limiting the ability to derive statistically meaningful insights about non-binary people. However, the revelation of one’s trans and non-binary identity goes beyond mere employment. There have been several instances where individuals have not been recruited because they chose to reveal their sexual identities. Beyond employment considerations, the act of coming out as a trans or non-binary person is deeply personal, involving the challenging navigation of societal norms. To address this, Princy suggests support from companies is crucial. This should come in the form of DE&I policies, schemes, and similar supportive systems. “Improvements should also extend to insurance coverage that includes the third gender, incorporating medical support for necessary procedures. A call for an equal and independent work atmosphere, coupled with opportunities for representation and community welfare awareness, further contributes to fostering inclusivity,” said Princy. For those struggling to come out, “my advice is to take a leap of faith. Despite initial challenges, there is now widespread awareness and support available. Embracing your true self can lead to a happier, more fulfilling work-life balance,” she said. Today, companies have open policies, providing a systematic process. She encourages reading policies, and contacting DE&I points of contact, HR, and managers to navigate the systematic process. The industry, she notes, is increasingly supportive, fostering a culture where everyone feels valued and respected, irrespective of background or identity.","excerpt":"From comprehensive onboarding support to innovative DE&I initiatives, Epsilon strives to be an inclusive workplace that transcends biases.","categories":["IT Services"],"tags":["AI (Artificial Intelligence)"],"author_name":"Shritama Saha","publish_date":"2024-02-16T12:14:55","publication_year":"2024","word_count":873,"keywords":["Go","API","programming_languages:R","AI","RAG","Aim","ViT","Rust","GAN","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","Aim","RAG","R","Go","Rust","API","GAN","ViT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/it-services\/how-epsilon-is-navigating-dei-in-tech\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10078827,"title":"Who is this Indian-origin Technologist Acting as Musk’s Right Hand?","content":"The new CEO of microblogging platform Twitter is moving at a pace that can only be described as hard and fast. A few hours after taking over the company in a USD 44 billion deal, Musk fired several top executives including Indian-origin CEO Parag Agrawal. Agrawal had replaced founder Jack Dorsey after he resigned less than a year ago. Musk is reportedly working with a tight-knit group of advisers on layoffs, content moderation policies, advertising sources and product changes. Besides the 50 employees from Tesla, SpaceX and The Boring Company, this inner circle that Musk handpicked included some familiar faces from his usual cohort and some new ones. There was David Sacks, a venture capitalist and former product head from PayPal, Alex Spiro, his personal lawyer and Jason Calacanis, a tech investor and podcaster. An Indian-origin name stood out—Sriram Krishnan. Now that the word is out: I’m helping out @elonmusk with Twitter temporarily with some other great people. I ( and a16z) believe this is a hugely important company and can have great impact on the world and Elon is the person to make it happen. pic.twitter.com\/weGwEp8oga— Sriram Krishnan – sriramk.eth (@sriramk) October 30, 2022 Who is Sriram Krishnan? A partner with investment heavyweight Andreessen Horowitz, Krishnan led the cryptocurrency division at the firm. Horowitz and Marc Andreessen’s firm invested USD 400 million in Musk’s Twitter acquisition. A16z’s interest in Twitter makes sense considering its push for Web3 but there’s still no clarity around whether this will move Twitter in a direction that is more friendly towards Web3. A few days ago, Krishnan tweeted a photo from the company’s San Francisco office clarifying that he was only “helping out Musk with Twitter temporarily with some other great people”. Krishnan was a product leader at Twitter in the past and is notably the only one in the bunch who previously worked within the company. He has also worked as a product developer at Facebook and Snapchat. The Musk connect This wasn’t Krishnan’s first brush with the Tesla boss. Krishnan and his wife, Aarthi Ramamurthy, who were both born in Chennai, are seemingly well-connected. They host a podcast called ‘The Good Time Show’ that became one of the most popular shows on the social audio app ‘Clubhouse’ last year. Musk appeared as one of the guests on the show after the couple reportedly texted him. According to an earlier profile done on the couple by The New York Times, Krishnan and Ramamurthy had met Musk many years ago during a private tour of the SpaceX headquarters they attended in Hawthorne, California. Krishnan and Ramamurthy’s show had other high-profile guests including Meta CEO and founder Mark Zuckerberg and boasted socialite Paris Hilton and DJ Calvin Harris among its fans. ‘The Good Time Show’ was started in the heyday of the pandemic in December 2020 and picked up 175,000 subscribers. While the craze for Clubhouse has died down, the couple continues to post episodes of the show on YouTube. Discussions on the series are mainly around what’s trending in cryptocurrency buyouts and developments in robotics. Sriram Krishnan and Aarthi Ramamurthy co-host ‘The Good Time Show,’ Source: Yahoo.com Background Raised in typically middle-class Indian households, Krishnan and his wife met in Chennai’s SRM Engineering College in a Yahoo! chat room while working on a coding project together. After moving to the US, Krishnan and Ramamurthy were both spotted by Microsoft executive S. Somasegar, who’s a venture capitalist himself. Somasegar ended up hiring both of them in 2005 after which they moved to the Microsoft headquarters. The couple eventually got married in 2010 and moved to Palo Alto from Seattle, where they currently live with a two-year old daughter. Ramamurthy herself worked at Netflix in her early days and has since introduced a couple of startups—True and Co., an e-commerce site for lingerie in 2012 and Lumoid, another platform that allows people to try out consumer gadgets before buying. Like Krishnan, Ramamurthy was also a product director in Facebook in 2017 before working with Clubhouse. Ramamurthy also received equity like other employees with the app. The couple has gradually built a reputation around backing new technology. Krishnan in their earlier interview said that the two of them “have been fanboys forever” and really just “love tech”. Like any launch I’m sure there will be room for improvement and changes but I’m excited for the first change in how verification works in social media in a very long time.— Sriram Krishnan – sriramk.eth (@sriramk) November 6, 2022 Work with Twitter While it isn’t clear exactly how Krishnan is helping Musk out, he has been tweeting in defense of the company’s new proposed revenue model. Several celebrities and twitterati have critiqued Musk’s suggestion to charge USD 8 from verified accounts on Twitter as being elitist. Krishnan called these comments ‘logically inconsistent’, further stating that the payment process will reduce duplicated Twitter accounts. He also said that the blue tick on the platform was initially meant to resolve the issue of recognising the ‘person for who they say they are’ instead of ascertaining the person as some kind of celebrity.","excerpt":"Krishnan was a product leader at Twitter in the past and is notably the only one in the bunch who previously worked within the company.","categories":["AI Features"],"tags":["Interviews and Discussions"],"author_name":"Poulomi Chatterjee","publish_date":"2022-11-07T14:00:00","publication_year":"2022","word_count":851,"keywords":["Go","API","startup","programming_languages:R","AI","venture capital","programming_languages:Go","RAG","ai_applications:robotics","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","RAG","R","Go","API","startup","venture capital","programming_languages:R","programming_languages:Go","ai_applications:robotics"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/who-is-the-indian-origin-software-technologist-acting-as-musks-right-hand-in-twitter\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10111921,"title":"How Viswanathan Anand Uses ChatGPT","content":"At MLDS 2024, India’s biggest generative AI conference, Viswanathan Anand, spoke about AI and chess, and how he uses ChatGPT to make chess training exercises and explain moves. “ChatGPT is essentially statistics applied to language,” said Anand, explaining that language is a fundamental way of how we learn anything. “That’s the key connection.” Citing chess, he said that it [ChatGPT] gives you the correct answer as long as you specify the rules and the terms of the game. “Understanding this is very hard unless you are trained in chess,” he added. He said it is difficult to learn from it, because the moves these AI systems show are typically hard to explain, and when amateur players copy those moves hastily it generally backfires on them. On the bright side, he noted that if used wisely AI helps players save a lot of time in learning new moves, and understanding their opponent better. Chess in the Era of AI Now that everyone has access to tools such as ChatGPT and Bard that will push players towards being more creative in a game, and learn faster. How can one differentiate themselves to be a better player?“So the first thing is extra familiarity and the next is asking more questions,” he explained. The better player, according to Anand, examines not only the best moves, but the next seven or eight moves which the opponent has neglected. He said that this is a game of constant improvements, and one has to come up with entirely new systems. “What I believed to be the truth when I was younger is only maybe 52% true,” he said, stating that with AI, the game of chess has changed significantly. Today the players have to be willing to sit there with the AI system, playing training games and willing to forget their comfort zone, Anand said. Instead they have to test individual positions and go over the new moves over and over again until it works. Thanks to AI, there’s a lot of scope for doing things much faster. “Instead of spending hours just pulling up statistics about my opponent, I can instead use AI or ChatGPT for the details I need.” AI in Chess Anand said AI can never replace humans in chess, even though it is getting better at it. “Chess is still a human game.” OpenAI chief Sam Altman also echoed similar views, at Davos WEF, said, “Chess has never been more popular than it is now, and almost no one watches two AIs play each other, we are very interested in what humans do.” Anand said chess doesn’t have a finite number of sensible moves, and AI can never nearly solve chess. “There isn’t a finite number of steps in chess. I don’t think it will be solved given the large number of even the sensible moves to be analysed by the computer,” he added. He said that he likes to play with humans over AI anyday, but uses AI to practise. In 2016, AlphaGo beat Lee Sedol four times out of five. The efforts with chess go further back to 1985, when IBM’s Deep Blue first lost to Garry Kasparov but promptly came back next year to win. But the general fear that this will take the fun away from games is overrated. We’ve evolved from Houdini and Rybka to AlphaZero progressively getting better at the game that effortlessly beat any Chess Grandmaster. The most significant demonstration of AlphaZero’s capabilities was a series of games played against the popular chess engine Stockfish 8. The battle between the best human-designed chess engine and an AI system that had taught itself how to play was watched by millions of chess enthusiasts. Not only chess, AI agents are making progress in beating humans in video games and even physical games. Anand Intelligence (AI) in Chess Anand was taught to play chess by his mother Sushila Viswanathan at the age of six. He continued playing at the Tal Chess Club in Chennai backed by Soviet funds, which is one of the reasons for Chennai becoming the capital of Indian chess. The biggest difference between playing chess before, and after the advent of computer programs and then AI is the approach to the game. Anand said, “I’m not even trying to come up with new ideas myself now. The computer gives me the idea, I use it to navigate the lines of moves ahead and reason it out.” This is very different from when he was a child he used to come up with the idea, check all the lines and implement it. “Now I don’t do anything. All I need to do is to keep asking questions. And so over, over a long enough time period, you suddenly look back and you realise you’re a completely different player,” he explained. In the world championship in 2008 against Kramnik who had always won against Anand playing the white pieces. “Because of the computers which were still reasonably primitive then, I was able to pull out the statistics and practise the moves deviating from my usual King’s Pawn opening,” he said. Anand explained that previously it would have taken him years to work the trial and error method and practise a new opening and really familiarise himself with it without computers.","excerpt":"“We’ve witnessed how AI has adapted and transformed the role of chess players.”","categories":["AI Features"],"tags":["AlphaGo","ChatGPT","chess","games","Google Deepmind"],"author_name":"K L Krithika","publish_date":"2024-02-03T16:04:35","publication_year":"2024","word_count":882,"keywords":["Go","ChatGPT","AlphaGo","API","OpenAI","AI","games","ML","GPT","chess","Google Deepmind","AI agents","generative AI","R"],"extracted_tech_keywords":["AI","ML","generative AI","ChatGPT","OpenAI","R","Go","API","GPT","AI agents"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-viswanathan-anand-uses-chatgpt\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":1630,"title":"Tech Mahindra boosts Industry 4.0: Opens up “Factory of the Future” centre at Bangalore","content":"As we move into a phase where manufacturing is getting bigger and better with connected devices powered by IoT, artificial intelligence and machine learning, it is witnessing an amplified productivity with minimal human intervention. Keeping in lines with the advancement in technology, Tech Mahindra, a pioneer in digital transformation has set up a state-of-the-art lab which works on technologies that are expected to power the future factory floors. Coming as a headstrong move by Tata Mahindra, it has opened up Factory of the Future centre at Bangalore. An initiative that is first of its kind, it definitely proves that the company is one of the leading solution providers in the field of smart manufacturing across the globe. Inside the Factory of Future Lab- Based on next-gen technologies, the lab intends to demonstrate products and solutions built in association of leading software and hardware vendors. Built in a modular form it would showcase the concepts of Industry 4.0. Utilizing on concepts like connected machine, manufacturing, robotics, AI, analytics, automation, AR and VR, the lab would show all possible solutions and assets in this space. “Industries world-over are going through a massive transformation, a transformation that is driven by the advances in newer technologies, especially things like IoT and Robotics. As a trusted Digital Transformation partner for client organizations, we have taken lead in these areas and have invested in building capabilities to harness the opportunities,” said L Ravichandran, President and Chief Operating Officer at Tech Mahindra. “The infrastructure and the assets in the lab are already being monetized to create customer-centric solutions and we already have some references while some of the customer use cases are being tested,” noted Aloke Palsikar, Senior VP and Global Head of Manufacturing at Tech Mahindra. From being a labour intensive segment to more and more automated with time, the impact and adoption of technology has been much more evident in manufacturing than other segments. It has entered a next-gen phase of automation where robots will work alongside humans in a connected factory floor powered by technologies such as AI, robotics, machine learning, big data and analytics. Tech Mahindra being a leading vendor in this space has a unique advantage of leveraging the best-in-breed technologies from global majors and building solutions banking on its deeper engagement and understanding of the customer problems. The company has already been working along with IBM at its newly-opened Watson IoT centre in Munich to create solutions targeting the Industry 4.0, especially segments such as manufacturing, farming, healthcare, banking, and automotive.","excerpt":"As we move into a phase where manufacturing is getting bigger and better with connected devices powered by IoT, artificial intelligence and machine learning, it is witnessing an amplified productivity with minimal human intervention. Keeping in lines with the advancement in technology, Tech Mahindra, a pioneer in digital transformation has set up a state-of-the-art lab […]","categories":["AI News"],"tags":["best technology to learn for future"],"author_name":"Srishti Deoras","publish_date":"2017-03-11T05:48:26","publication_year":"2017","word_count":420,"keywords":["big data","Go","machine learning","artificial intelligence","AI","Git","RAG","analytics","Rust","R","best technology to learn for future"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","analytics","RAG","R","Go","Rust","Git","big data"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/tech-mahindra-boosts-industry-4-0-opens-factory-future-centre-bangalore\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10077144,"title":"How I Cracked the FAANG Interview","content":"Abhinav Mishra’s resume betrays a life of ease. He was part of IIT Kanpur’s accelerator programme, studied MS in operations research at Columbia University and landed a job at Amazon as a senior product manager. In a chat with Analytics India Magazine, Mishra speaks about what it takes to crack a FAANG interview, the pragmatism behind tech and the art of ‘failing up’. AIM: What does your role at Amazon entail? Abhinav: I work as a senior product manager at the L6 level in Amazon’s Seattle office which is the company base, which means I own a considerable portion of the products and lead a critical role. I work on productivity tools for the vendors, which uses data to recommend how they could improve their offering. Essentially, my current product helps find the proverbial needle in the haystack and tell the vendor how they can fix it and what the impact might be money-wise. In my role, I work as both a technical product manager and a business manager in consultation with engineers and vendors on a daily basis. AIM: What pushed you to pursue data science as a career? Abhinav: When I was at IIT Kanpur for a year, at the age of 22, I started building a supply chain management SaaS tool. I started working on projects and doing pilots for startups and also companies like Titan. I went to Columbia in the middle of the pandemic and enrolled in a combination of courses from both the engineering and the business school, like data visualisation, design thinking and product management. I was very active as an undergrad student. You wouldn’t find me partying on a Saturday night. I was attending events and writing research papers, three of which were published in the IEEE. But the IIM accelerator was my first brush with the real world. That’s when I realised that I needed to know about both tech and business to make it. One person there said something witty that has still stayed with me. They asked me how I reached college? I said, “in an auto”, to which they replied, “exactly, you didn’t come here on a combustion engine. Stop talking about tech, AI\/ML only and focus on the problem and the solution instead”. AIM: There’s so much hype about being able to crack a FAANG. Can you share any tips from your Amazon interview? Abhinav: The common perception is that Columbia is what led me here, but the reality is very different. In the US, you are on your own. Someone graduating from Columbia has to put in the same effort as anyone else. If people think I was offered the job on a platter because of the ‘pedigree’, that is not the case. It was a crushing feeling, having to reach out to 20-25 people every day. I remember sitting for around 60 interviews, of which, only ten reached the final round. I appeared for Microsoft first, and then eventually sat for the Amazon interview which I got through. There are many fancy tips available outside which don’t really work. Candidates need to work hard on their resumes and focus only on what is relevant. I run a Facebook group with around 15k students where I help them out and give them career advice. I am often asked to review resumes where I see a bunch of unrelated stuff. Secondly, talk to people who work in these FAANG companies because they can tell you best about the work culture there. I didn’t focus on Columbia or Amazon specifically but rather worked on building my resume. So, I would advise people to start working on things that are relevant to the job they are applying for. Look for projects online that are suitable to you, like I worked on open-source projects. So, these three are the main factors for me – resume, people and razor-sharp focus on your skills. AIM: How was the difficulty level in the technical test? Abhinav: Technical questions were aplenty in the interview for the role of a product manager. I didn’t get any questions on coding. For my Microsoft interview, there were definitely more technical questions. For a product manager, you should at least know the data part of coding which is SQL. This is the main requisite for anybody who wants to break into product management. I appeared for TikTok, which isn’t FAANG but pretty close, where they asked me to code. They might even ask you to explain a code. There are not many direct technical questions but more practical questions about how I would behave in a certain situation or solve a specific problem in a real business scenario. A lot of the times there were questions around things you have worked on in the past. For those who are appearing for the technical product management interview, there will be coding questions and it is more likely they will ask you to code. Besides, there is a book called ‘Cracking the PM Interview’ by Gayle Laakmann McDowell which is the holy grail for aspiring product managers. She herself was an engineer with Google, and has written an entire section in the book just for the type of technical questions that may be asked in FAANG interviews. They might ask questions like, “Can you explain the concept of cloud to a kid?” AIM: What clicked between the Microsoft interview and the Amazon interview? Or do you think it is easier to get into Amazon? Abhinav: It’s definitely not that Amazon is easier to get into. And my preparation for both was equal. In fact a smaller company that I appeared for had a harder test than my Amazon interview. Between Microsoft and Amazon, Amazon is probably harder to crack because they have a ‘bar raiser’ measure according to which a candidate has to be better than 50% of their employees. You need to have genuine skills over experience to impress them. Infact, to crack a FAANG interview is relatively easier today probably because of the online resources available on the subject. There are courses on Coursera now worth USD 400 that will tell you exactly what to do. The harder part is to get a FAANG interview because the competition is higher and the intake far lower. There are studies which show that it is harder to get into FAANG companies than MIT or Harvard. AIM: What are the biggest lessons that you’ve learned in the data science industry? Abhinav: I have had three major takeaways along the way. I think my journey has been about failing to succeed. I have learned things from failing and succeeded better than my failures. I haven’t sat through a placement ever. Once the IIM program ended, I didn’t have my own money and didn’t want to ask my parents. I stayed away from home so that put me in a tight spot. The job didn’t come easy to me, I worked at it for a year. I took calculated risks and failed a lot until I succeeded. I also have a motto which is ‘Give to succeed’ which is why I run the Facebook group. I get hundreds of DMs of people asking me to look at their SOP or asking how they can get through XYZ college. I always taught my friends. I volunteer at the University of Washington for a product management course where I mentor students. The third thing is having a habit to understand. Everything has a framework in life — you cannot operate on autopilot. It’s just that we need to understand the framework of solving a problem.","excerpt":"There are many fancy tips available outside which don’t really work. Studies have shown that it is harder to get into FAANG companies than MIT or Harvard","categories":["AI Features"],"tags":["AI Tool","Interviews and Discussions"],"author_name":"Poulomi Chatterjee","publish_date":"2022-10-13T12:10:00","publication_year":"2022","word_count":1270,"keywords":["data science","Go","AI","ML","RAG","Ray","Aim","analytics","SQL","AI Tool","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","ML","data science","analytics","Aim","Ray","RAG","R","SQL","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-i-cracked-the-faang-interview\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10170125,"title":"How Zepto’s Data Team Built for 10-Minute Delivery","content":"To build apps serving the masses, companies must develop resilient systems that handle all the data efficiently and use it to empower their business decisions and improve the end-user experience. In such cases, the data team, with its innovative approaches to handling and processing data, does the heavy lifting. At the Data Engineering Summit 2025, Abhinav Raghuvanshi, Zepto’s associate director of data engineering, addressed the company’s challenge in managing data: How can data be delivered as quickly as food? For a company that established the 10-minute delivery model, real-time visibility is not just advantageous, but is essential for successful execution, forecasting, and enhancing customer experience. The talk, “From Warehouse to Lakehouse: The Future of Real-Time Data at Zepto,” traced how the data platform evolved from a single Redshift warehouse to a hybrid architecture powering operational analytics and application-grade real-time systems. Along the way, Zepto had to rethink ingestion, transformation, cost governance, and how it handled SQL queries for the backend. When Redshift Wasn’t Fast Enough Zepto’s original setup relied on Redshift as its central warehouse. It worked, until it didn’t. Query bottlenecks, stale reporting, and long wait times exposed cracks. “With the volume we were getting, it was becoming impossible to serve data in near real time,” Raghuvanshi said. Adding compute nodes or partitioning tables only went so far. Query collisions, a lack of separation between storage and compute, and costly I\/O pushed the team to rethink. Citing an example, Raghuvanshi said, if someone were to run a very inefficient query on the Redshift cluster, they would find that this approach would be pretty slow for them. So began the shift to a modern architecture with S3 as the central storage layer, Kafka for streaming ingestion, and Databricks (powered by Apache Spark and Photon) for transformation and orchestration. He explained, PII columns are stripped early, business logic is layered through a structured data pipeline, and every job is encrypted and scheduled using Apache Airflow. Zepto also made a shift to MongoDB to reduce latency recently. It will be interesting to see what they do next. Data Democratised, Governance Intact Zepto data platform supports nearly 400 users across business, product, and operations. To ensure usability, the engineering team built a low-code in-house framework where users can simply drop in SQL queries. “Most analysts are very comfortable with SQL queries,” Raghuvanshi explained. “We put in abstractions in terms of transformations they want to do.” Each team operates within cost boundaries thanks to Databricks tags, role-based access, and job-level budget dashboards. Teams get alerted if they’re burning through compute, and reporting workloads are sandboxed to avoid collisions. Instead of worrying about when the data would arrive, he noted, it gives more control to the end-users in terms of what frequencies of data they want to consume. The lakehouse model, underpinned by S3 and Databricks, now supports historical queries, snapshots, and time-series analytics. Snapshots can reconstruct a store’s inventory state from months ago, and they are helpful for audits, restocking predictions, or other analyses. ClickHouse for Real-Time Analytics From a business model perspective, Zepto utilises dark stores, which serve as central hubs for items like ice cream. Maintaining optimal conditions, such as temperature, requires real-time monitoring of various metrics in milliseconds to ensure product safety and quality. Additionally, tracking delivery routes necessitates IoT data integration for real-time reporting. This immediate feedback is crucial for efficient operations, ensuring timely store visits and preventing delays. Raghuvanshi mentioned using ClickHouse for that. The Iot system captures all those metrics and puts them into ClickHouse for real-time analysis, reporting, and monitoring. ClickHouse powers Zepto’s production-facing real-time analytics that include order drop rates, traffic flow, IoT metrics from storage facilities, and even fridge temperature tracking. It supports MergeTree engines—replicated, replacing, summing—each designed for different workloads. “If your data is immutable, feel free to just dump it into ClickHouse,” he said. To simplify access, the team wrapped Flink with a SQL-friendly interface, enabling analysts to plug queries into Kafka sources without writing code. Results are routed into ClickHouse, where apps can query metrics within milliseconds. ClickHouse is not just a database, it’s a monitoring nerve centre. Store-level metrics, inventory depletion, and fulfilment rates—all now update in near real-time. He explained that real-time data allows for efficient restocking. For example, if a store is running out of milk at 10 a.m., the system would provide the necessary data to ensure replenishment by 2 p.m. Future Roadmap Tools like StarRocks and other Online Analytical Processing (OLAP) engines are being evaluated to handle complex joins without latency penalties. The larger goal is to make real-time intelligence as accessible as historical reporting without expecting every business user to become a systems engineer. With a system like Zepto’s, he noted, “Now people can focus on building more data products tailored for users within the ecosystem.”","excerpt":"Zepto transitioned to a real-time analytics architecture designed to match the speed of its promise.","categories":["AI Features"],"tags":["Zepto"],"author_name":"Ankush Das","publish_date":"2025-05-19T11:09:46","publication_year":"2025","word_count":800,"keywords":["Go","AI","MongoDB","Apache Spark","RAG","analytics","Kafka","SQL","R","Zepto","Databricks"],"extracted_tech_keywords":["AI","analytics","RAG","Apache Spark","Kafka","MongoDB","Databricks","R","SQL","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-zeptos-data-team-built-for-10-minute-delivery\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":36976,"title":"How This Indian Researcher Is Using Deep Learning To Shut Down Trolls And Fake Reviews","content":"The world has become more connected now than ever before. An opinion can fly at light speed across continents and a revolution can be sparked by remote players in a matter of hours. The platform owners and policymakers have to walk a tightrope as the fabric of society is at stake. And as the technology keeps improving, new methods are being discovered by fraudulent players to up their ante. Most platform owners tussle with the after-effects of an aggravation. To identify the perpetrator and discourage them to carry out any illicit, immoral social engineering would require automated detection techniques. Machine learning is being used for predicting stock market calamities and other anomalies. The recommendation engines used for e-commerce platforms, predict the likes and dislikes of the users by methods such as collaborative filtering to identify similarities in the choices made by the user in the past and also of their peers. Now ML models have risen to a state where they can devour on tonnes of data and come up with solutions\/insights which enhance decision making. Since, predicting an anomaly and then making a prudent decision is crucial to discouraging unwanted social engineering, machine learning would be a great place to start in our pursuit to accommodate uninterrupted transaction of ideas. Fraudulent User Prediction With REV2 Srijan Kumar of Stanford University and his peers from Flipkart and Carnegie Mellon University introduced REV2 early last year to alleviate the noisy spammers on product reviews on retail platforms. Since the motivation of users is associated with a rating of a product, it is extremely important to those who own the platforms to stop the fake reviews from getting over publicised and bring down the authenticity of the product; eventually sales. The researchers proposed that users and ratings have intrinsic, inter-dependent scores that indicate how trustworthy they are and how a product is going to be evaluated based on this. REV2 algorithm is designed to calculate the intrinsic scores of the users and their ratings based on their behaviour. How REV2 fairs with its contemporaries on 5 datasets.Source Key Takeaways REV2 outperforms nine existing algorithms in identifying fake reviewers. Running REV2 on Indian e-commerce giant Flipkart’s product reviews gave an accuracy of 84.6%; 127 out of 150 probable unfair users were proven to be guilty by the fraud investigators. This success made REV2, the go-to algorithm at Flipkart now and is deployed on a large scale. Read more about REV2 here. To Tweet, Or Not To Tweet Twitter has received a lot of flak recently for its deplatforming of few critics based on the mood swings of a certain section of the platform. Since people now are more troubled by the misalignment of views of others, things take an aggressive turn with every tweet. The administrators of the platform walk have to make a trade-off between keeping the interaction healthy while avoiding the danger of shutting down of few voices because it didn’t meet the threshold of some algorithm or some section didn’t like it. There is a lot of misinformation circulated and it gets worse with the popularity of the entities involved in the discussion. The dialogues can start with casual nitpicking and can transform into trolls and threats. This virtual wildfire decouples the users from the truth and they usually end up in their own echo chambers. With the success of REV2, Kumar is now working on how to make the internet a better place to have healthy interactions. His methods include extraction of attributes from social graph structure to characterize and mitigate the damage of disinformation. The fake reviews were handled by REV2 using graph mining collective classification algorithm and now for tackling multiple account abuse in online discussions, Kumar proposes first web-scale characterization model, which statistically analyses user interaction graphs to predict online conflicts. This technique is already in use by the administrators at Reddit and Wikipedia. And, the efficacy of this model will be verified when it is put to use on platforms like Twitter — mankind’s newfound turf for warfare.","excerpt":"The world has become more connected now than ever before. An opinion can fly at light speed across continents and a revolution can be sparked by remote players in a matter of hours. The platform owners and policymakers have to walk a tightrope as the fabric of society is at stake. And as the technology […]","categories":["Deep Tech"],"tags":["deep fake","Deep Learning","Flipkart","Stanford"],"author_name":"Ram Sagar","publish_date":"2019-03-28T09:45:07","publication_year":"2019","word_count":673,"keywords":["Go","machine learning","Rust","programming_languages:R","AI","ML","Flipkart","programming_languages:Go","RAG","deep fake","Stanford","Deep Learning","R","programming_languages:Rust"],"extracted_tech_keywords":["AI","machine learning","ML","RAG","R","Go","Rust","programming_languages:R","programming_languages:Go","programming_languages:Rust"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/how-this-indian-researcher-is-using-deep-learning-to-shut-down-trolls-and-fake-reviews\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":26438,"title":"Is Big Data, China’s Unique Advantage, Also Hampering Its Growth?","content":"Does China really have a big data advantage? Asia’s most dominant player in artificial intelligence has stepped up its efforts considerably. However, tech experts believe that even though the country and its key tech players have embraced the new changes, data is just one part of the puzzle as the country grapples with its share of user privacy concerns. Casting a speculative look at China’s big data hegemony, industry experts have speculated that even though the Asian giant is charging ahead under a national agenda and establishing data sovereignty, it is also bringing in challenges in the form of privacy and digitisation. The big data gold rush in China is accompanied by its share of doubts and fragilities. Fallout Of Big Data In China Rise In Data Theft: Even though the country leads in data sovereignty, it also has been subject to data leakages. According to news reports, personal data theft is on a rise and a series of privacy issues has jolted Chinese users who are now showing a heightened sense of awareness. In fact, China Railways faced a data breach of their official ticket booking website in June 2018 after reports emerged of user information being sold for bitcoins on the dark web. Data Sovereignty Means Tighter Control Over Private Companies: The rise of China’s tech firms is attributed to sweeping data collection power that allowed them to expand exponentially. But while data did enable the companies to scale their operations fast, it also enabled Beijing to play a more authoritative role in the companies. It also implies the State’s unbridled control over data and its intended use. In the case of China, the control is exerted over tech companies through regulation, closer scrutiny of capital inflows, fundraising and tighter control over products. Now there is news that the state wants “special management shares” or stakes in internet media giants. Nationalised Tech Sector: China’s pro-trade policies have led to a nationalised tech sector where startups mushroom. Also, the country is pushing for a tech homecoming wave with the state urging tech champions with a market value exceeding ¥200 billion to list with Chinese Depository Receipts (CDR). While on one hand foreign private investment is highly regulated, no major player can compete against local companies on account of Chinese regulations. So that weeds out foreign competition, also due to highly-regulated investment, control is never ceded to foreign hands, unlike India where FDI is a huge source for economic development. Chinese companies are cocooned from direct competition. Big Data And Privacy Issues A bigger issue is that while China’s privacy laws are very “business-friendly” they allow the state to monitor users aggressively. China’s data privacy laws give more teeth to the tech giants, is less rigid for the companies; and thus very effectively allows data processing without consent. It appoints tech companies as data stewards. In fact, as Samm Sacks, Senior Fellow, Technology Policy Program a Center for Strategies and International Studies points out the concept of consent is given a broader latitude under China’s data protection regime. Sacks said that the Chinese approach to data privacy may look similar on the surface but it is the exemption of consent which could come under an intense debate. According to Sacks, China’s data privacy framework appears to differ from EU’s more rigid GDPR in terms of its overall approach to having legal ground to process personal data. Another research pointed out that Chinese users have a more laid-back approach to their use of personal data and privacy protection laws vis-à-vis their Western counterparts. Is Data Deluge Enough To Help Companies Compete On A Bigger Scale “If data is the new oil then China must be super rich,” observes Australian Strategic Policy institute. However, the data deluge and vision for big data supremacy in China is accompanied by a lack of stricter privacy laws, controlled tech companies that operate under a national agenda and state-controlled technologies. Another offshoot of controlled technologies and privacy laws that govern the outflow of data is that the models are only being built on the data harvested locally. The Australian think tank observes that while the models are being built around the needs of the state, can they be scaled to compete with Google or Microsoft at a global level? China’s closed tech culture doesn’t allow foreign companies to enter the market and build applications tailored for Chinese consumers. This can also impact the tech ecosystem, many a tech expert has observed that the success of Chinese firms is only limited to the mainland and scaling operations would not be easy. Another key takeaway is that how data is collected would also affect the country’s digital ecosystem. According to the Australian think tank’s white paper, China’s dalliance with big data began in early 2010 when the State Council mandated the policy in its 12th Five Year plan to “strengthen the development of basic software — more specifically, software that handles big volume of data”. Last Word Today the big data vision is bigger and more ambitious with the Chinese government laying out an AI vision for 2030. As part of the AI vision, that entails developing a $150 billion industry, the state has earmarked $2 billion for AI development. AI is also at the heart of its military plan with China ramping up the use of AI for military applications. This has also sparked an international debate on the use of LAWS in warfare and how it could have a major impact. The white paper further states that as per the 13th Five-Year Plan, big data applications form a high priority project dubbed as ‘informatisation’ projects. Big data has already seeped into China’s social sector and under the 13th plan, it will be further intensified. There will be more big data applications in the social, economic and public security sector.","excerpt":"Does China really have a big data advantage? Asia’s most dominant player in artificial intelligence has stepped up its efforts considerably. However, tech experts believe that even though the country and its key tech players have embraced the new changes, data is just one part of the puzzle as the country grapples with its share […]","categories":["AI Features"],"tags":["Data Privacy","digitisation"],"author_name":"Richa Bhatia","publish_date":"2018-07-13T13:13:18","publication_year":"2018","word_count":973,"keywords":["big data","Go","API","artificial intelligence","AWS","AI","digitisation","Git","RAG","Data Privacy","GAN","R"],"extracted_tech_keywords":["AI","artificial intelligence","RAG","AWS","R","Go","Git","API","big data","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/is-big-data-chinas-unique-advantage-also-hampering-its-growth\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10097513,"title":"Data Modernisation and Monetisation: A Closer Look at Publicis Sapient&#8217;s Trailblazing Initiatives and Strategies","content":"Amid rapid technological advancement, the buzzword playing on everyone’s lips is “data modernisation.” As companies grapple with the ever-increasing volume, variety, and velocity of data, the need to harness its power and derive meaningful insights has become paramount. Analytics India Magazine spoke to Deepak Kumar, senior director data engineering, at Publicis Sapient, who shares the company’s customer-centric data modernisation plans and its relevance in today’s industry. “Data modernisation is crucial for building a customer-centric culture. And it’s not just about providing a good service to the customer but the ability to understand customer situations. Data management, essentially, is to give a platform that serves as a ‘one-stop-shop’ to everyone in the organisation or a stakeholder so they can easily access, manage and analyse the data for crucial insights that drive provide more data-driven outcomes,” he said. Kumar further emphasised on the genAI pivot in today’s era. “Organisations need to prioritise and embrace a data-centric approach to move forward with agility. Gone are the days of relying on outdated data. For businesses to be able to leverage genAI, it’s crucial to have up-to-date and relevant data. Only when they do this is when they can benefit from advanced analytics and cutting-edge experiences powered by genAI. What organisations truly need to accomplish this, is dynamic and real-time data that can be seamlessly integrated into their existing platforms, catering to both their brand and their customers. This is why data modernisation plays a pivotal role in establishing a customer-centered business. By ensuring that data is constantly in motion, businesses can harness its full potential and stay ahead in today’s fast-paced world,” he added. Today, several traditional companies hesitate at the thought of digitising their business. “They don’t have the resiliency because of legacy architectures and tools,” said Kumar highlighting the need for digital transformation. From traditional industries to cutting-edge startups, organisations are recognizing data as the new currency, and those who can effectively leverage it will gain a competitive edge in the marketplace. “Businesses are facing an unprecedented level of volatility and disruptions that have occurred over the past few years. Staying competitive and resilient, requires companies to embrace digital transformation and modernization. These initiatives not only mitigate risks but also unlock newer opportunities for growth. While there may be some challenges to overcome, such as managing maintenance costs, it’s important to view them as stepping stones towards achieving long-term success. With the introduction of stricter data privacy and security regulations, organisations must also ensure they strike a balance between compliance and operational efficiency. Moreover, considering the impact on revenue and industry demand becomes crucial as businesses navigate through these transformative times,” he added. Publicis Sapient helps its clients in every step while digitising their data but there are several hurdles which need to be overcome in the process. “One of the key challenges we tackle for our clients is bridging the gap between data modernisation and monetisation,” said Kumar. “It’s not uncommon for organisations to struggle to effectively utilise their data, not realising the potential it holds. We believe that data can be leveraged both internally and externally, allowing organisations to generate value by sharing it with accelerators and other partners. Implementing robust data models lets organisations unlock the power of their data and gain greater visibility for various purposes,” he suggested. Externally, there are massive shifts in terms of regulatory compliance and Publicis Sapient is helping customers with data privacy and data sharing because most of them are not ready or have slow and disjointed machine learning, he added. Most recently, the company was also named the market leader in Data Modernization Services by HFS Research. Vision and Strategy Being a renowned leader in digital business transformation software, Publicis Sapient has demonstrated a strong commitment to driving digital innovation for its clients over the years. The vision was inspired by Nigel Vaz, CEO of Publicis Sapient, and his book titled “Digital Business Transformation.” Vaz introduced a unique philosophy known as SPEED, which has since guided the organisation’s approach to transformation. SPEED is an acronym for Strategy: developing and testing your hypothesis on priority value pools; Product: evolving at pace and speed; Experience: how you enable value for customers, Engineering: delivering on your promise, and Data: validating your hypotheses and uncovering insights for constant iteration. “Another notable aspect is our internal solution isolators or IP. We have developed a comprehensive solution accelerator and toolkit, which we call the PS inner source This internal resource empowers our teams by expediting development in various data functions including integration, management, governance, and quality. These accelerators and toolkits provide a solid foundation and streamline the process of building necessary data foundations based on specific requirements. They play a crucial role in facilitating efficient and effective development within our organisation,” explained Kumar. Training and Best Practices The technology company holds many workshops and initiatives to upskill cross-skilled people. They also have a leadership development programme to make IT leaders shine, said Kumar. “A lot of internal initiatives that we have created talk about self-levelling C paths. We encourage them to become certified professionals in technology so we have a pre-approved list of technologies in the data space where people can see what they certainly want to do,” he added. Suggesting the best practices for organisations in the industry, Kumar said, “Customer-centric guide is all about understanding customer expectations, perceptions and what we need to build Customer 360. If we talk about data models, our statement is interface-centric around data monetisation and production isolation, basically how an organisation can have more understanding and control of their data.” In conclusion, Kumar said that personalisation, hyper-personalisation and security are the best practices to think of when offering customer-centric data models.","excerpt":"As companies grapple with the ever-increasing volume, variety, and velocity of data, the need to harness its power and derive meaningful insights has become paramount.","categories":["AI Highlights"],"tags":["data cloud","Data Engineering","Publicis Sapient"],"author_name":"Tasmia Ansari","publish_date":"2023-07-25T15:28:16","publication_year":"2023","word_count":950,"keywords":["Go","API","GenAI","machine learning","AI","data cloud","ML","Git","RAG","Data Engineering","Publicis Sapient","analytics","R"],"extracted_tech_keywords":["AI","machine learning","ML","analytics","GenAI","RAG","R","Go","Git","API"],"url":"https:\/\/analyticsindiamag.com\/ai-highlights\/data-modernisation-and-monetisation-a-closer-look-at-publicis-sapients-trailblazing-initiatives-and-strategies\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10166768,"title":"EY India Launches Customised LLM for BFSI Sector","content":"EY India has developed a customised, fine-tuned LLM tailored for BFSI applications. This AI-powered solution promises enhanced customer service, improved operational efficiency and optimised risk management. The LLM is designed to address the unique needs of the BFSI sector by improving answer accuracy, intent recognition and contextual comprehension. It incorporates BFSI-specific vocabulary, ensuring precise and dependable AI-driven solutions. The model is also equipped with multilingual capabilities. Powered by the fine-tuned Llama 3.1-8B instruct model, the AI solution leverages advanced parameter-efficient finetuning with low-rank adaptation techniques. This allows for deeper contextual understanding and better alignment with India’s BFSI regulatory frameworks. According to EY India, the model delivers measurable business benefits, including up to 50% cost savings. Mahesh Makhija, partner and technology consulting leader at EY India, said, “EY’s customised AI models are tailored to meet the specific needs of the BFSI sector, empowering banks to optimise operations, enhance risk management and make data-driven decisions.” The model is designed for flexible deployment, allowing BFSI organisations to host it either on-premises or in cloud environments with limited GPU requirements, significantly reducing costs. It can be integrated into customer service applications through text-based platforms such as WhatsApp, SMS and website chatbots, as well as voice-calling AI call center frameworks. Unlike other models, EY’s fine-tuned LLM prioritises security and regulatory compliance by being securely hosted within enterprise networks. This approach mitigates risks such as AI hallucinations and data privacy concerns, ensuring that sensitive information remains protected within organisational boundaries. Early this year, the company launched the EY.ai Global AI Advisory Council, which aims to deliver key insights into AI advancements, societal impacts and innovative applications, enabling the global EY network and its clients to remain at the forefront of AI developments.","excerpt":"The model delivers measurable business benefits, including up to 50% cost savings.","categories":["AI News"],"tags":["EY","LLM"],"author_name":"Merin Susan John","publish_date":"2025-03-27T15:06:16","publication_year":"2025","word_count":286,"keywords":["programming_languages:R","AI","chatbots","data-driven","LLM","RAG","Aim","EY","GAN","llm_models:Llama","R"],"extracted_tech_keywords":["AI","Aim","RAG","chatbots","R","GAN","data-driven","llm_models:Llama","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/ey-india-launches-customised-llm-for-bfsi-sector\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10012197,"title":"“Moneyball-ed”: How Kaggle GM Okoshi Takumi’s Love For Baseball Statistics Made Him A Data Scientist","content":"“I would be happy if medical AI becomes more and more popular in the next ten years.” For this week’s ML practitioners series, Analytics India Magazine got in touch with Kaggle GM Okoshi Takumi. Okoshi is ranked 55 in Kaggle global rankings and currently works as a data scientist at Rist — an AI company based in Japan. In this interview, Okoshi talks about how his love for baseball led him to data science. He also has some tips for aspirants who are looking to break into the field of data science or make it to the top in Kaggle. AIM: How did your Data Science journey begin? Okoshi: I played baseball when I was a kid. At the time, I kept track of my hitting records in Excel and looked at the performance of professional baseball players. I enjoyed it. Looking back, I think I’ve loved data since then. Once I got to high school, that fever grew even more, and with the help of Moneyball analysis, I was able to analyse the data of Japanese professional baseball. I’ve analysed who the most valued hitters are and whether there are any hidden star candidates on your favourite team. I enjoyed that analysis so much that I looked for a university where I could study statistics, and to this day, I enjoy analysing data very much. In college, I learned about statistics and machine learning, and the research involved using machine learning to decipher the relationship between cheering and outcomes in sports. I’ve read a lot of books. Many of them have been translated into Japanese, or are in Japanese, to begin with, but O’Reilly’s books are very good. I also studied in Coursera’s Andrew Ng course and Stanford University’s CS course delivery videos. Recently, there is also a book of tips from Kaggle masters that have been published, which is still only in the Japanese language but has been very helpful to Kagglers in Japan. Several books on Kaggle are also available in Japan, making it easier for newcomers to get started. One of the books — “Kaggle Start Book”, written and reviewed by mean my teammates in the Petfinder competition. Currently, I work for Rist, a Japanese IT company, where our team develops AutoML solutions that will be used internally. Apart from this, I spend part of my time on Kaggle. AIM: Tell us about your Kaggle journey Okoshi: The first Kaggle challenge I tackled was Porto Seguro’s Safe Driver Prediction. It was a prediction of whether or not a car insurance policyholder would have an accident. It was a competition of anonymous features, so it was very difficult. I managed to get a bronze medal with the help of the kernel. Kaggle appeals to me for three main reasons: Firstly, is the competition where you can compete with the other participants on the score, so you know where you stand. It’s nice when you move up in the rankings, and you get to experiment and learn a lot for it. Secondly, coming in contact with data on many different topics. Images, text, audio, and tableau, each has its own interesting aspects, and I enjoy getting a taste of it all! Especially in multi-modal competitions, such as Petfinder and Avito, it was a lot of fun because I was exposed to different types of data in one competition. Also, all of the competition topics were interesting, even if they were in areas I didn’t know about as I got to work on it and become a little more knowledgeable about the field. Thirdly, getting hold of the solution shared by the top solvers. Even when the rankings are not good, after a competition, you can look at the solutions of the winners and reflect on what was lacking. I try to recreate the top solutions from competitions that I have already completed, and even in competitions where I lost, I always try to make use of them for my own benefit. On Kaggle, the current competition methods are discussed in kernel and discussion, and there is a solution released afterwards, so I feel it will be a learning experience for both beginners and experts. Both the participants and Kaggle’s UI are very appealing to me because of the importance of learning. Especially the discussion and kernel, and of course the sharing of solutions afterwards is a great culture. I try to write solutions in the competitions that I get to the top of, and I hope I can contribute to this great culture in some small way. AIM: What does it take to be at the top? Okoshi: I’m still a novice ML engineer myself. In particular, I feel it’s very difficult to think about what AI products users want. Of course, the development skills I have at Kaggle are beneficial in my work, and once I get into the development phase and the experimental phase, I can create solutions quickly. However, I feel that I have a particular challenge in other areas, especially in task design, and I believe that learning those areas will help me become a “great” ML engineer. I participated in a lot of competitions, about 40 competitions, and since Porto Seguro, I have been working on Kaggle competitions almost without a break. Every time I participate in a competition, I discover something new, which I can use to improve my work in the following competitions. In due course, I was able to win more medals and become a Kaggle Grand Master. It’s also important to have an organised methodology in competitions. In my case, I make a pipeline for competitions, and when there is a new method for each competition, I try to incorporate that into that pipeline. By doing so, I can easily try out the methods in the next competition, and together with the pipeline, I can get stronger with each competition. This pipeline development has also led to the development of AutoML. Given a data science problem, I would first focus on creating a benchmark using a pipeline. Then when I get a score, I compare it with that of other participants’ scores, and if it’s low, I try to identify the missing parts. If it’s high, I generate new ideas. To do this, I gather information from the Overview section of the competition, discussions, and papers. Along with this, I do EDA to check the data. From this point onwards, it is all about repetition of hypotheses, ideas and experiments to raise the score. Towards the end of the competition, I do an ensemble and make a final push. It is important to create benchmarks quickly and experiment with new ideas. And so far my pipeline building routine has been successful! I think it’s a good idea to participate in the Kaggle competition for now. At first, you can make a benchmark using the kernel as a reference, and then you can participate in discussions and other kernels. In Japan, we have a blog for beginners called “What to do next after signing up for Kaggle”. There’s also an article (in Japanese), and I think beginners will learn a lot just by following this. I also recommend the Kaggle start book, which you should do when you start with Kaggle so that you can work on your competition smoothly. If you’d like to read it in English, contact the author! AIM: What does your machine learning toolstack look like? Okoshi: I write code in Python, and I mainly use the PyTorch framework for competitions that need neural networks and LightGBM for the tableau competition. And of course, I’m indebted to many other tools such as pandas, NumPy, Scikit-learn, Albumentations, etc., and for images, I recommend timm. For accessing any pre-trained models on PyTorch, check this. For computing, I use GCE for my machines. The company also supports Kaggler’s machine fees, which is very helpful. AIM: What do you think the future of ML would be like? Okoshi: That’s a very difficult question. I think that machine learning products, like any other product, will depend on what the users want. I started my own business when I was a college student, and back then, I couldn’t figure out what the users wanted, and I couldn’t create something that I could say with any confidence that it was good. Markets simply don’t accept this. As a result, I abandoned the company and started doing something else. Looking back on that, I think it’s important to create something that users want, and honesty, to create something that can be marketed with confidence as a good product. In ML, I still think the problem of accuracy is very deep-rooted. In many cases, there is a difference between the accuracy we can achieve with AI, and the accuracy users expect. Both parties end up unhappy. We need to find a field where the difference in recognition is matched as much as possible, and the accuracy we achieve with AI is sufficient for the service. That said, I have high hopes for the medical industry myself. There are a lot of medical competitions on Kaggle, but I think we’re just starting to make progress in this field. I would be happy if medical AI becomes more and more popular in the next ten years, and I would gladly participate in those medical-related competitions.","excerpt":"“I would be happy if medical AI becomes more and more popular in the next ten years.” For this week’s ML practitioners series, Analytics India Magazine got in touch with Kaggle GM Okoshi Takumi. Okoshi is ranked 55 in Kaggle global rankings and currently works as a data scientist at Rist — an AI company […]","categories":["AI Features"],"tags":["data analytics masters","Data Scientist","Interviews and Discussions","Kaggle grandmasters","Python for Data Science"],"author_name":"Ram Sagar","publish_date":"2020-11-23T13:00:43","publication_year":"2020","word_count":1558,"keywords":["data science","data analytics masters","scikit-learn","machine learning","AI","neural network","PyTorch","ML","Aim","Kaggle grandmasters","analytics","Python for Data Science","LightGBM","Data Scientist","Interviews and Discussions"],"extracted_tech_keywords":["AI","machine learning","ML","neural network","data science","analytics","Aim","PyTorch","scikit-learn","LightGBM"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/okoshi-takumi-kaggle-interview-data-scientist\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10089019,"title":"ChatGPT is Now On Azure OpenAI Service","content":"The Azure OpenAI Service has been providing 1,000 customers and enterprises with most advanced AI models of this era—such as GPT-3.5, DALL-E 2, and Codex—through their massive supercomputing cloud. Now, Microsoft is also including ChatGPT in this service. This inclusion will allow developers to easily integrate these custom AI-powered experiences directly into their own applications. Enterprises and developers now have more options to create advanced AI-powered applications that offer personalised experiences to users. This includes the ability to enhance existing bots to handle unexpected questions, enable faster customer support resolutions by recapping call centre conversations, create personalised ad copy with offers, automate claims processing, and more. The ChatGPT service is priced at $0.002\/1k tokens and the billing for this will begin from March 13, 2023. By combining cognitive services with Azure OpenAI, enterprises can create compelling use cases. For example, by using conversational language for knowledge base retrieval on enterprise data, Azure OpenAI and Azure Cognitive Search can be combined. Using a no-code approach in Azure OpenAI Studio, users can easily create customised applications without requiring extensive coding knowledge. In addition to the customisability offered for every model in the service, Azure OpenAI Studio also offers a unique interface to configure ChatGPT’s response behaviour to align with the user’s organisation. The blog also highlights how Microsoft has been following its Responsible AI Standard by putting customers in charge, explaining the outputs generated by AI, filtering content, and also having transparency guidelines. Recently, Microsoft also unveiled Microsoft Dynamics 365 Copilot, which is touted as the world’s first copilot for customer relationship management (CRM) and enterprise resource planning (ERP).","excerpt":"The ChatGPT service is priced at $0.002\/1k tokens, and the billing for this will begin from March 13th.","categories":["AI News"],"tags":["Microsoft Azure"],"author_name":"Mohit Pandey","publish_date":"2023-03-10T11:36:27","publication_year":"2023","word_count":267,"keywords":["ChatGPT","DALL-E","TPU","OpenAI","AI","R","GPT","Aim","Microsoft Azure","GAN","Azure"],"extracted_tech_keywords":["AI","ChatGPT","OpenAI","Aim","Azure","TPU","R","GPT","DALL-E","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/chatgpt-is-now-on-azure-openai-service\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10010365,"title":"Bengaluru-based No-Code AI Platform Signzy Raises $5.4M In Current Funding Round","content":"Bengaluru-based Signzy, a ‘no-code AI platform has raised $5.4M from Arkam Ventures and Mastercard. Signzy’s existing investors, Kalaari Capital and Stellaris Venture Partners also participated in the current round. The ‘no-code AI platform’ by the startup helps financial services automate risk and compliance processes. With the massive surge in global demand for its solution amid the Covid-19 pandemic, the startup is helping banks and financial institutions with AI-led digital transformation. Signzy aims to will invest the funds in AI research, product enhancement and sales team expansion to address the surge in demand from global and India clients. “The ongoing pandemic has not only upended the pace of digitalisation in the financial services space but has also led to increased demand for no-contact solutions,” shared Ankit Ratan, co-founder of Signzy.He further added that this fundraising is a timely boost in their efforts to scale up our AI capabilities and to tap an ever-widening global market opportunity. “We’ll also be using these funds to deepen our customer engagement,” he said. Signzy can completely automate the back-operations decision-making process into a real-time API. This is possible due to a combination of Nebula — a no-code AI model builder and a Fintech API Marketplace of over 240+ APIs. “Some of the largest and most demanding banks and financial services in India and globally have adopted the Signzy solution due to its cutting edge technology and measurable value proposition. We are delighted to support this exceptional team as they broaden their product portfolio and accelerate global expansion,” said Bala Srinivasa, Managing Director of Arkam Ventures.   Signzy provides digital KYC solutions to over 100 Indian banks, insurance firms, and asset management companies like Aditya Birla Sunlife AMC, BoB Financial and more. The clients also include four of the largest banks in India, with State Bank of India (SBI) and ICICI Bank among them.Globally, Signzy has a strong partnership with Mastercard and offices in New York and Dubai to serve customers in multiple geographies.","excerpt":"Bengaluru-based Signzy, a ‘no-code AI platform has raised $5.4M from Arkam Ventures and Mastercard. Signzy’s existing investors, Kalaari Capital and Stellaris Venture Partners also participated in the current round. The ‘no-code AI platform’ by the startup helps financial services automate risk and compliance processes. With the massive surge in global demand for its solution amid […]","categories":["AI News"],"tags":[],"author_name":"Srishti Deoras","publish_date":"2020-10-22T10:54:22","publication_year":"2020","word_count":327,"keywords":["Go","API","programming_languages:R","AI","digital transformation","Git","Aim","AI research","R","startup"],"extracted_tech_keywords":["AI","Aim","R","Go","Git","API","digital transformation","startup","AI research","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/bengaluru-based-no-code-ai-platform-signzy-raises-5-4m-in-current-funding-round\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10093602,"title":"Now You Can Talk to ChatGPT on Your iPhone","content":"In a surprising new development, OpenAI has released a ChatGPT app for iOS users. The on-the-go app syncs users’ conversations, and supports voice input, alongside other new improvements. Sadly, Android users will have to wait. “P.S. Andriod users, you’re next!,” said OpenAI, in its blog post, saying that ChatGPT will be coming to Google’s devices soon. Just launched the ChatGPT iOS app! https:\/\/t.co\/QC2Ec7Jshs Now in the US, world soon. Android next.— Mira Murati (@miramurati) May 18, 2023 OpenAI said that the app is free to use and syncs users’ history across devices. It has also integrated Whisper, an open-source, multilingual speech recognition system, which enables voice input. The company said that the ChatGPT Plus subscribers will be getting exclusive access to GPT-4’s capabilities, alongside early access to new features and faster response times, all on iOS. Download ChatGPT app here. Currently, the company said that it is planning to roll out in the US, and slowly will be expanding it to other countries in the coming weeks. Google, the creator of Bard, is catching up with ChatGPT once again. The company at the recent I\/O conference announced a series of updates for Bard, including the integration with Search, PaLM 2 update, etc. But, it is yet to launch a mobile app. Again, who knows? They might still have a chance to launch a Bard app on Android ahead of OpenAI. Regulatory Challenges This new development comes against the backdrop of regulatory challenges that many AI companies are facing across geographies, including the recent US Senate discussion, where the OpenAI chief shared his concerns about the technology, and how government and companies should come together to regulate them. This move by Altman has been raising eyebrows in the open-source community, calling this a manipulative move, culling new innovations and using them for personal benefit. Read: Has OpenAI Lost the Open Race? https:\/\/www.youtube.com\/watch?v=D59T1ldw1gg On the other hand, the European Parliament has implemented stringent AI regulations. This includes bans on biometric surveillance, emotion recognition, and predictive policing AI systems, alongside tailor-made regimes for general-purpose AI and foundational models like GPT, and others. India, on the other hand, has nothing in place yet. Read: India Backs Off on AI Regulation, But Why? However, the Indian government is now waking up and is considering a regulatory framework for AI-enabled platforms like ChatGPT, including areas related to algorithmic bias and copyrights.","excerpt":"This new development comes against the backdrop of regulatory challenges that many AI companies are facing across geographies","categories":["AI News"],"tags":["OpenAI"],"author_name":"Tasmia Ansari","publish_date":"2023-05-19T09:27:03","publication_year":"2023","word_count":395,"keywords":["Go","ChatGPT","OpenAI","AI","innovation","llm_models:PaLM","GPT","R","llm_models:ChatGPT","llm_models:GPT"],"extracted_tech_keywords":["AI","ChatGPT","OpenAI","R","Go","GPT","innovation","llm_models:GPT","llm_models:PaLM","llm_models:ChatGPT"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/now-you-can-talk-to-chatgpt-on-your-iphone\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10169250,"title":"Oracle and Andhra Pradesh Government to Train 400,000 Students in AI, Cloud and Data Science","content":"Oracle has partnered with the Andhra Pradesh State Skill Development Corporation (APSSDC) to provide skills development training in emerging technologies to 400,000 students across the state. The initiative will deliver free, structured digital training through Oracle MyLearn, covering more than 300 hours of content. Students will receive foundational and professional-level training in Oracle Cloud Infrastructure (OCI), AI services, OCI Generative AI, Data Science, APEX, DevOps, and Security. The collaboration is intended to boost employability and prepare students for technology-driven careers. “Our collaboration with Oracle is a huge milestone in strengthening and expanding the digital capabilities of the youth of Andhra Pradesh,” said G. Ganesh Kumar, I.A.S., MD & CEO of APSSDC. “We are determined to produce a highly skilled IT workforce and become the IT knowledge hub for the world.” Students will be able to access curated learning paths aligned with their educational levels and career goals. In addition to certifications, learners can earn digital badges to signal job-readiness to employers. “We are proud to partner with APSSDC in our ongoing commitment to empower youth with future-ready skills and support India’s economic development,” said Shailender Kumar, senior vice president and regional managing director, Oracle India and NetSuite – Japan and Asia Pacific. “Oracle University’s Skills Development Initiative offers world-class training and certifications that not only build in-demand digital competencies, but also help students validate their expertise.” Oracle University has trained over 3.1 million certified professionals globally and logged more than 20 million hours of learning content consumption. The Andhra Pradesh initiative builds on Oracle’s ongoing strategy to expand the IT talent pool across India.","excerpt":"The initiative will deliver free, structured digital training through Oracle MyLearn, covering more than 300 hours of content.","categories":["AI News"],"tags":["Oracle"],"author_name":"Siddharth Jindal","publish_date":"2025-05-06T19:08:21","publication_year":"2025","word_count":264,"keywords":["data science","Go","programming_languages:R","AI","programming_languages:Go","Git","Oracle","generative AI","GAN","DevOps","R"],"extracted_tech_keywords":["AI","data science","generative AI","R","Go","Git","DevOps","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/oracle-and-andhra-pradesh-government-to-train-400000-students-in-ai-cloud-and-data-science\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10163041,"title":"Synergy Quantum, MP3 International Sign Deal to Boost Quantum Tech for Military","content":"Delhi-based Synergy Quantum and Abu Dhabi-based MP3 International, a subsidiary of Grade One Group under EDGE Group, have signed a sales and distribution agreement to enhance military cybersecurity and quantum technology across the Gulf Cooperation Council (GCC). The partnership aims to deploy military-grade quantum-secure solutions to protect critical infrastructure, defence operations, and government agencies in the UAE, Saudi Arabia, Kuwait, Qatar, Bahrain, and Oman. The agreement addresses the growing threats of cyberattacks and the potential vulnerabilities posed by quantum computing. With conventional encryption standards at risk of being compromised, the collaboration will introduce Post-Quantum Cryptography (PQC), Quantum Key Distribution (QKD), Quantum Sensing and AI for military and defence applications. Strategic Collaboration for Defense Security Under the agreement, Synergy Quantum’s solutions will be distributed through MP3 International, enhancing cybersecurity for defence and critical infrastructure in the GCC. Jay Oberai, CEO and Founder of Synergy Quantum, stated, “The rise of advanced cyber threats and the emerging risks posed by quantum computing necessitate a forward-thinking and robust defence strategy.” In addition, Dr. Ibrahim Al Kuwaiti of MP3 International said, “This collaboration empowers us to distribute military-grade cybersecurity products and advanced quantum technologies across the GCC.” Military Applications of Quantum Technologies The integration of quantum solutions into military operations is gaining traction globally. Air Marshal GS Bedi (Retd.), VP of business development (aerospace & satellite communications) at Synergy Quantum, noted that quantum cryptography is becoming essential for modern air forces, including those in the Middle East and India. Others also noted the growing emphasis on quantum capabilities. “Quantum warfare is no longer a distant concept; it is an imminent reality that advanced armies worldwide are preparing for,” stated Lt Gen Nagesh Rao (Retd.), vice president of defense business development at Synergy Quantum. Key Benefits for Military and Defense Sectors Quantum-secured communication enhances protection for sensitive military data, mitigating the risks of cyber threats and quantum-enabled attacks. AI-powered analysis further adds to this, improving decision-making by enabling real-time data processing and strategic insights. Enhanced operational resilience also allows for faster simulations and more efficient military responses in high-stakes scenarios. Furthermore, advanced surveillance capabilities, supported by quantum sensors, improve anomaly detection and situational awareness for defence missions. Lastly, secure global connectivity through satellite-based quantum networks ensures encrypted, long-range communication, strengthening the cybersecurity framework for military operations.","excerpt":"The agreement addresses the growing threats of cyberattacks and the potential vulnerabilities posed by quantum computing.","categories":["AI News"],"tags":["Cybersecurity","Military","quantum","quantum cryptography","quantum key distribution"],"author_name":"Sanjana Gupta","publish_date":"2025-02-07T21:13:36","publication_year":"2025","word_count":380,"keywords":["quantum key distribution","Go","programming_languages:R","AI","Military","quantum cryptography","programming_languages:Go","quantum","Aim","anomaly detection","ViT","Cybersecurity","R","emerging_tech:quantum computing"],"extracted_tech_keywords":["AI","Aim","anomaly detection","R","Go","ViT","programming_languages:R","programming_languages:Go","emerging_tech:quantum computing"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/synergy-quantum-mp3-international-sign-deal-to-boost-quantum-tech-for-military\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10131286,"title":"NVIDIA Chief Jensen Huang to Visit India in October | Register Now for AI Summit","content":"NVIDIA has announced its AI Summit 2024, set to take place from October 23 to 25 at the Jio World Convention Centre in Mumbai, India. The summit will feature three days of presentations, hands-on workshops, and networking opportunities, aimed at connecting industry experts and exploring advancements in artificial intelligence. “It’s going to be an exhilarating experience for the tech community and I’m especially proud to be hosting the summit in India,” said Vishal Dhupar, Managing Director, Asia South NVIDIA. NVIDIA CEO Jensen Huang will make a special visit to India for the event, participating in a fireside chat. Huang’s visit follows last year’s announcement of NVIDIA’s partnerships with Reliance, Tata, and Infosys to support India’s AI startup ecosystem and reskill the IT workforce. NVIDIA has begun delivering its latest chips, such as the GH200 AI, to Indian partners like Tata Communications and Jio Platforms, which are building AI-cloud infrastructure Tata Communications’ Managing Director and CEO, A.S. Lakshminarayanan, confirmed that the installation process for NVIDIA’s AI Cloud is underway, with a full launch expected by the third quarter of this fiscal year. The summit will showcase NVIDIA’s latest innovations, including the much-anticipated B100 chip, promising significant improvements over its predecessor, the H100. Attendees can explore NVIDIA’s suite of AI solutions designed to accelerate innovation in fields like healthcare, robotics, and industrial digitalization. REGISTER NOW Last year, Infosys expanded its alliance with NVIDIA to train 50,000 employees on NVIDIA’s AI technology, integrating these tools with Infosys Topaz to create generative AI solutions for enterprises. During his visit to India last year, Huang said “India has lots of data,” touching upon the diversity of languages and dialects. He said, “There’s no reason for India to export data to western companies”. He believes India has the capability to make in-house LLMs and foundational models. “You have all of your own data. You have the great talent of computer scientists. You produce more computer scientists than any country on the planet; and have the right infrastructure for producing computer scientists. You have an infrastructure for that, right? It’s called AI – Actual Intelligence,” said Jensen, speaking of IITs. However, he said India does lack infrastructure – “not the roads and bridges kind”, but AI infrastructure. He said with NVIDIA supercomputers coming in, that has also been taken care of. “You have everything you need to build and use the AI here. But you need to have infrastructure. Just like electric power plants and steam engines, this is now the production of intelligence,” he stressed. REGISTER NOW Following that, the Indian government is actively supporting AI development through the IndiaAI mission, launched in March 2024, which aims to position India as a global AI leader by investing in infrastructure and supporting startups. The mission includes an INR 10,300 crore investment to expand AI infrastructure and make GPUs more accessible. In March 2024, Yotta received the first shipment of 4,000 NVIDIA H100 GPUs. Yotta plans to scale up its GPU inventory to 32,768 units by the end of 2025, following last year’s announcement of importing 24,000 GPUs, including NVIDIA H100s and L40S, in phases. REGISTER NOW","excerpt":"NVIDIA AI Summit running from October 23-25, 2024, will feature over 50 sessions and live demos on generative AI, industrial digitalization, robotics, large language models, and more.","categories":["AI News"],"tags":["NVIDIA"],"author_name":"Siddharth Jindal","publish_date":"2024-08-03T15:02:39","publication_year":"2024","word_count":519,"keywords":["Go","artificial intelligence","AI","innovation","Git","Ray","Aim","generative AI","NVIDIA","R","startup"],"extracted_tech_keywords":["AI","artificial intelligence","generative AI","Aim","Ray","R","Go","Git","innovation","startup"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/nvidia-chief-jensen-huang-to-visit-india-in-october\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10118059,"title":"NVIDIA Introduces Ruler to Measure the Context Length of Models","content":"NVIDIA researchers developed RULER, a synthetic benchmark designed to evaluate long-context language models (LLMs) across various task categories, including retrieval, multi-hop tracing, aggregation, and question answering. The study involved benchmarking ten long-context models using RULER with context sizes ranging from 4K to 128K. The models were assessed on 13 tasks of varying complexity. Click here to check out the GitHub repository. The evaluation revealed that despite achieving nearly perfect results on the needle-in-a-haystack test, all models experienced significant performance drops as the input length increased. The top-performing models, including GPT-4, Command-R, Yi-34B, and Mixtral, demonstrated satisfactory performance at 32K length, but others struggled with larger contexts. The researchers also examined the impact of training context length, model size, and architecture on performance. Models trained with larger context sizes generally performed better on RULER, although performance rankings varied with longer sequences. Larger models, such as Yi-34B-200k, outperformed smaller counterparts, demonstrating the benefits of scaling model sizes. Non-Transformer architectures like RWKV-v5 and Mamba-2.8B-slimpj faced significant degradation when extending context size to 8K and underperformed compared to the Transformer baseline Llama2-7B. The main results showed that while all models claimed context sizes of 32K tokens or more, none of them maintained performance above the Llama2-7B baseline at their claimed length, except for Mixtral, which performed moderately well at double its claimed context size of 32K. The models experienced large degradation in performance when tested using RULER as sequence length increased, despite achieving nearly perfect results in the needle-in-a-haystack task. GPT-4 was the best-performing model, exhibiting the highest performance at 4K length and the least degradation when extending the context to 128K. Additionally, the study found that the top three open-source models—Command-R, Yi-34B, and Mixtral—used a large base frequency in RoPE and had larger parameter sizes. Although the LWM, trained with a context size of 1M, performed worse than Llama2-7B at 4K, it showed smaller degradation with increasing context size, leading to a higher rank than Mistral-7B in weighted evaluations. RULER’s open-source availability aims to encourage comprehensive evaluation and further research on long-context modeling, highlighting significant room for improvement in this area.","excerpt":"The main results showed that while all models claimed context sizes of 32K tokens or more, none of them maintained performance above the Llama2-7B baseline at their claimed length.","categories":["AI News"],"tags":["NVIDIA"],"author_name":"Mohit Pandey","publish_date":"2024-04-11T13:56:18","publication_year":"2024","word_count":348,"keywords":["Go","AI","RPA","Git","RAG","GPT","Aim","transformer architecture","NVIDIA","GitHub","R"],"extracted_tech_keywords":["AI","Aim","RAG","R","Go","Git","GitHub","transformer architecture","GPT","RPA"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/nvidia-introduces-ruler-to-measure-the-context-length-of-models\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":23090,"title":"Beginners Guide To Reinforcement Q-Learning","content":"Q-Learning is a reinforcement learning technique. It has the ability to compute the utility of the actions without a model for the environment. It takes the help of action-value pair and the expected reward from the current action. During this process the agent learns to move around the environment and understand the current state which is the optimal policy by taking the action with the highest reward. Let us look at an example of this technique. Environment Imagine we are moving from one floor to another in the elevator in an apartment building. Considering this situation, our agent will be awarded if he starts from a random room and eventually finds a way to reach the terrace. He can move around in any direction with random action. Let us look at the available actions, states, rewards and the goal. States – 0, 1, 2, 3, 4, 5 (Terrace) Actions – 0, 1, 2, 3, 4, 5 (Terrace) Reward – 0, 100 Goal – 5 (Terrace) Let us visualise these on a Q-Table. On this table, the rows represent the states or the floors and the columns represent the actions that can be taken at a particular floor. 0 means that particular action can be taken but has no reward, whereas -1 means the particular action is not available for that state. We know how an elevator works, and by pressing the highest number on the keyboard we reach the terrace. But let us consider that the agent does not know to read the buttons on the elevator display board and has to figure out a way to reach the terrace by trial and error method. The basic principle here is that the agent cannot jump a floor while moving, if he does then he is punished. If the agent moves to an adjacent floor he is not rewarded unless that particular floor is the terrace. That’s where his freedom lies. What the agent will have is the Q matrix, where every state-action is encoded to reach the optimal policy to get the maximum expected reward. How Does It Work? Following are the steps on how the Q Learning table works Q matrix is initialized with zeros (agent has just entered the environment) A random action is performed to just to another floor (current state —-> next state) Every episode will state at the ground floor and end at the goal, that is, the terrace When the state is not a final state (terrace) Perform a random action Based on the current action go to the next state Find the maximum Q value. We continue this process starting from the current state and move to the next state. And the next state will be our current state and the process continues until we reach the goal. This will be our basic algorithm which works on the principle of awarding the agent for the right move and punishing for the wrong move. Implementation of Q-Learning Lets us consider the Q Table initialized with zeros to start with. Here is the reward table on which the agent will be rewarded based on the actions he takes. Episode No.1 Let us say the agent will be starting from the fourth floor (4), the available actions are moving down to third floor (3) or moving up to the terrace (5) Now let us say we randomly choose to move to the state 5. Here, we have two possibilities that is we can move either the below floor – fourth floor or we can reach the goal by moving up to the terrace. We need to select the biggest Q value from the available values which are – Q (5,4), Q (5,5). We choose the max function to make the agent to move upwards to the terrace. But at this point the Q table is still filled with zeros. Now the new state is 5, and after the max function the agent will move towards the terrace and will be awarded with a reward of 100. Our Q table after this computation will look like Episode No.2 Now the agent will randomly choose another state let’s say the third floor – which is state 3. Now the possible actions are moving either down to the second floor or moving up to the fourth floor. Let us look at the possible action-reward pair from the table below Now since we have the possible actions – Q(3,2), Q(3,4). If the agent randomly choose to go the fourth state. We can update the Q table with the help of the Markov’s Decision Process we can award the agent with a reward multiplied by a gamma function (0.80). For the Q(3,4) we get a reward of 80. The updated Q table after two episode is below Episode No.1,000 One we have run this process for a thousand episodes we will each a optimal Q table for which the agent knows how to reach the terrace with optimal number of iterations. This is an example of Q Learning and here is a visual of how the Q table is at 1000 episodes Eventually the agent will learn how to move from any floor to the top floor using this optimal policy. Here is a visual description of how the environment works though every iteration of every episode. Markov Decision Process Overview (Image Courtesy : SFL Scientific) Conclusion We saw how an agent learnt to move towards the terrace eventually from any random state with optimal number of moves. This is a simple example of Q-Learning technique, using Markov Decision Process. The random action can be nullified as we move forward through the episodes and lowering the probability of occurrence of these kinds of actions. The random action is allowed or forced on the agent just to make the agent understand the environment completely, that is, to see the unseen areas. This way the agent will master the environment after 1,000 episodes.","excerpt":"Q-Learning is a reinforcement learning technique. It has the ability to compute the utility of the actions without a model for the environment. It takes the help of action-value pair and the expected reward from the current action. During this process the agent learns to move around the environment and understand the current state which […]","categories":["Deep Tech"],"tags":["Reinforcement Learning"],"author_name":"Kishan Maladkar","publish_date":"2018-03-28T11:48:36","publication_year":"2018","word_count":990,"keywords":["Go","Reinforcement Learning","programming_languages:R","AI","ML","programming_languages:Go","R"],"extracted_tech_keywords":["AI","ML","R","Go","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/beginners-guide-to-reinforcement-q-learning\/","complexity_score":4,"technical_depth":6,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10167900,"title":"DeepSeek to Open Source its Inference Engine","content":"Chinese AI lab DeepSeek on Monday announced its intention to open-source its inference engine. To achieve this, the company is “collaborating closely” with existing open-source projects and frameworks. Previously, when the company planned to open-source its inference engine, it identified challenges such as significant codebase divergence from the original framework, extensive infrastructure dependencies, and a limited capacity to maintain a large-scale public project. The latest announcement further emphasises DeepSeek AI’s dedication to open-sourcing key components and libraries of its models. 🙏 @deepseek_ai's highly performant inference engine is built on top of vLLM. Now they are open-sourcing the engine the right way: instead of a separate repo, they are bringing changes to the open source community so everyone can immediately benefit!https:\/\/t.co\/0bEYh8p97W— vLLM (@vllm_project) April 14, 2025 Recently, during Open Source Week, the company released five high-performance AI infrastructure tools as open-source libraries. These enhance the scalability, deployment, and efficiency of training large language models. “It’s an honour to contribute to this thriving [open source] ecosystem and to see our models and code embraced by the community. Together, let’s push the boundaries of AGI and ensure its benefits serve all of humanity,” said DeepSeek in the announcement. Recently, the company, in collaboration with Tsinghua University, unveiled a new research study aimed at improving reward modelling in large language models by utilising more inference time compute. This research resulted in a model named DeepSeek-GRM, which the company asserts will be released as open source. A few weeks ago, DeepSeek released an update for its DeepSeek-V3 model. The updated model, ‘DeepSeek V3-0324’, now ranks highest in benchmarks among all non-reasoning models. Artificial Analysis, a platform that benchmarks AI models, stated, “This is the first time an open weights model is the leading non-reasoning model, marking a milestone for open source.” The model scored the highest points among all non-reasoning models on the platform’s ‘Intelligence Index’. Recently, Reuters reported that DeepSeek plans to release R2 “as early as possible”. The company initially intended to launch it in early May but is now considering an earlier timeline. The model is expected to produce “better coding” and can reason in languages beyond English.","excerpt":"The announcement emphasises DeepSeek AI’s dedication to open-sourcing key components and libraries of its models.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","DeepSeek"],"author_name":"Supreeth Koundinya","publish_date":"2025-04-14T13:58:32","publication_year":"2025","word_count":356,"keywords":["Go","programming_languages:R","AI","DeepSeek V3","programming_languages:Go","Scala","DeepSeek","Aim","programming_languages:Scala","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","DeepSeek V3","Aim","R","Go","Scala","programming_languages:R","programming_languages:Scala","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/deepseek-to-open-source-its-inference-engine\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":20871,"title":"In India, Retail Lags In Terms Of Analytics Adoption Compared To Other Industries, Says Kapil Malhotra Of Pepsico","content":"Kapil Malhotra, Head- Analytics, Pepsico With over 15 years of progressive leadership experience across business analytics, strategy & planning, and Go-To-Market, Kapil Malhotra, currently serves as the head of analytics at PepsiCo, and has worked across functional domains such as marketing, sales, pricing and supply chain over the years. In his current capacity, Malhotra drives analytics agenda for PepsiCo India, accelerating the foundational analytics capability to better support business decisions, planning, optimization, and initiating transformational initiatives aimed at future proofing. He is instrumental in delivering key analytics projects, driving meaningful triangulated insights and collaborating across various functions for coherent execution of various initiatives. Prior to PepsiCo, he has worked with the likes of Accenture, Genpact, marketRx, Blue Star Infotech and Daewoo Motors, leading various analytics positions at these organizations. A post graduate from NITIE, Mumbai and an industrial engineer from DCE, Delhi, he considers himself an analytics practitioner who uses his overall expertise and rice experience in multiple environments to positively impact business outcomes. In his conversation with Analytics India Magazine, Malhotra talks about his analytics journey, analytics utilisation at Pepsico, challenges in the area and much more. Here is the complete interview. 1) How has been the analytics journey at Pepsico? What have been the major milestones achieved post adoption of analytics? It’s been a very exciting and impactful journey. Albeit a slow start amidst organizational transformation in 2014\/15, analytics delivered on critical agenda items of identifying category\/brand drivers, understanding pricing power, improved econometric forecasting, sales prioritization at granular level, and hyper-targeting mega-cities. Analytics was initially seeded to drive efficacy of marketing programs with efficient mix (marketing), and to support day-to-day business performance analyses (senior leadership). It eventually evolved as a standalone function to impact sales, finance, strategy, IT and now supply chain, to drive better, analytically driven actions, besides being key enabler for core planning and strategy processes. 2) How is Pepsico making use of large volumes of data? What are the analytics tools at disposal and how are they being used? PepsiCo has invested in building big data infrastructure based on SAP HANA, with data warehouses that power the self-service provisioning tools such as Cognos, as well as automated Tableau visualisation for drill-down on sales data. Further, there’s JDA that powers statistical demand forecasting at SKU level using the best-fit algorithms basis historical patterns. Most part of external data such as Nielsen retail performance or Kantar IMRB consumertrack, comes along in flat files or data dumps, besides specific software to run drill-downs. Finally, multiple streams of digital data gets synthesized through tools such Brandwatch and Unmetric. Most of our analytics projects are delivered within partners’ ecosystem, but for local projects, we make use of open source R and Python. 3) What have been the benefits of using analytics till date? Would you like to highlight few Indian use cases where analytics has brought significant changes? From how analytics started as a fledgling function three years back, it is today the key influencer in business affairs as highlighted earlier. Key benefit areas have been better forecasting mechanism, quantification of business drivers, re-allocation of marketing spends, owning the core process of category and share construct with senior leadership, influencing pricing decisions, better performance diagnostics leading to tactical actions on ground, optimum assortment mix in mega-cities, and influencing global business intelligence transformation. While still there’s room to expand the sphere of influence, in India analytics is much evolved, integrated business function today, locally as well as within global PepsiCo ecosystem. 4) How do you make use of analytics to stay ahead of competition? Well, most of the above mentioned analytics led initiatives are either to optimise resource allocation or to enable PepsiCo India grow better, and ahead of the competition. Besides, we do maintain close eye on our competition and bring to light pockets where we need to take specific actions, be it related to consumer promo, pack-price or distribution. 5) How has been the adoption of big data analytics in retail industry with India perspective? In India, Retail\/ CPG has been the laggard in terms of adoption of ‘structured or big data analytics’ when compared with other industry peers such as BFSI, telecom or e-commerce. Large part of it is by design since the latter have consumer (also customer in most cases) level data, and each consumer leaving a trail of touchpoints or transaction data each day. While retail does capture the same, it’s less frequent and when it comes to CPG, consumer level touchpoint\/ transactional data is nearly non-existent. But now, with growing digital footprint, at least digital profiling is a possibility and is used for targeted campaigns. This should gradually evolve to much better segmentation and eventually, 360 consumer views by integrating offline data or pooling in other sources of available information such as consumer connect programs. 6) What are the future plans\/roadmap for analytics at Pepsico? Way forward is very clear – take up all initiatives to keep impacting the business positively. While some work has already started, business 360 dashboard across data sources with embedded analytics engine, scaling up of mega-city hyper-targeting through better assortment across top 10 cities, and auto-refreshing ROI models covering ATL\/BTL drivers using machine learning, are clearly top 3 initiatives. 7) What is the size and hierarchical alignment (both depth and breadth), of your analytics group? I have a lean team of 3 team members, which includes one data scientist, and two senior business analysts. Internally, we work closely with IT and Strategy functions. As mentioned earlier, large part of reporting and analytics is delivered via partner ecosystem, the sum-total of all might be around 25-30. 8) What are the most significant challenges you face being at the forefront in analytics space? As part of the business and still being relatively a new function, no prizes for guessing that it is the business alignment and required resourcing, which is the most challenging part. From a client servicing standpoint, it is always about understanding client business situation well enough to have any project move in the right direction and to deliver business impact. 9) What kind of knowledge and skill-sets do you look for, while recruiting your workforce? I have a pretty simple way of recruiting my team. It’s the cultural fitment (organisation’s), demonstrated skill-set (specific to position) or performers, and go-getters or folks who have shown initiatives (so that I don’t have push). 10) How do you think ‘Analytics’ as an industry is evolving today? Could you tell us the most important contemporary trends that you see emerging in the present analytics space across the globe? This is an age of rapid evolution, and not just evolution. Technology is transforming the way businesses create and deliver goods or services, and at the same time, consumers are becoming ever-demanding and loyalty stays fragile. Disruption is all around us and is happening at rapid pace, especially in digital – be it electric\/driverless cars and trucks or blockchain (e.g. a group of Indian insurers has just to ease their application processing using this technology) or IoT (mass onboarding of consumer devices around the corner) or entire set of strides being made in AI (smart speakers) & robotics. Even today, there’s lot of behind the scene analytics that decides how we consume information or are offered marketing choices or denied\/offered an insurance policy\/credit card. Since analytics thrives on data and is poised to become more core than being peripheral, products and services will be designed for measurement, unlike how they have so far (cost, reliability or serviceability). From businesses standpoint, embedded analytics as foundational data platform shall gain prevalence that would integrate data across disparate sources and provide quantified impact of inputs or competitive action on the fly – one of the most relevant solves for FMCG, amongst others. Analytics, on one hand shall play key role in driving next level of business efficiency, and on the other, in designing products\/services that allow for individualised experience through better measurability. These are exciting times!","excerpt":"With over 15 years of progressive leadership experience across business analytics, strategy & planning, and Go-To-Market, Kapil Malhotra, currently serves as the head of analytics at PepsiCo, and has worked across functional domains such as marketing, sales, pricing and supply chain over the years. In his current capacity, Malhotra drives analytics agenda for PepsiCo India, […]","categories":["AI Features"],"tags":["Analytics India","current leaders in self driving cars","Interviews and Discussions","retail analytics"],"author_name":"Srishti Deoras","publish_date":"2018-01-22T06:56:20","publication_year":"2018","word_count":1328,"keywords":["Go","API","machine learning","AI","R","Analytics India","Git","RAG","Python","Aim","analytics","retail analytics","current leaders in self driving cars","Interviews and Discussions"],"extracted_tech_keywords":["AI","machine learning","analytics","Aim","RAG","Python","R","Go","Git","API"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/retail-lag-terms-analytics-adoption-says-kapil-malhotra-pepsico\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10069071,"title":"On launching a career in Data Analytics &#038; Business Intelligence","content":"Although the data science platform market is constantly growing, for someone looking to enter the data science industry, choosing the right kind of analytics course is still not easy. The perfect program should have an updated curriculum, practical hands-on sessions, proper training and mentorship and should definitely be industry-ready. A recruiter looks for three core skills in an analytics profession, as follows: Domain knowledgeGrasp on analytics toolsUnderstanding the business problem that the company is trying to solve. More often than not, analytics courses focus too much on tools. However, business acumen is also extremely important. Therefore, Prestige University is organising a webinar “Why you should pursue MBA in Data Analytics & Business Intelligence” on June 20, 2022, from 3:00 PM – 4:00 PM to offer career guidance for analytics aspirants. Click here to register for the webinar The webinar will cover: Introduction to the data science & analytics market & the emergence of data science as a field.Core skills to build a career in data science.Different job roles in the data science & analytics industry.How pursuing MBA in data analytics & business intelligence can help? Who should attend? Students looking to build careers in AI, data science and analyticsFreshers & candidates with work experience of less than three years Data scientists & analytics professionalsBusiness analysts & data analystsData science & data analytics enthusiastsWorking professionals looking to transition to an analytics role Click here to register for the webinar Session speakers Santhanakrishnan R-Vice President – Analytics Consulting, Tiger Analytics He is a senior leader working at the intersection of advanced analytics and management consulting. He advises clients across key industries such as retail and consumer packaged goods, consumer finance, manufacturing and oil & gas on end-to-end aspects of Advanced Analytics programs: design, implementation and ongoing value delivery. During his stints with leading companies like GE, Genpact, and Infosys, he has worked closely with leaders, teams, and world-class clients in analytics consulting, data science, and big data platforms. Dr Chandrashekhar G R- Dean of business analytics at Prestige University He comes with three decades of experience in industry and academia. He has set up the data science program at IFMR before Prestige. Click here to register for the webinar","excerpt":"For someone looking to build a career in data science, finding the right data analytics & business intelligence courses can be a daunting task. Prestige University is organising this webinar to offer career guidance for analytics aspirants.","categories":["Deep Tech"],"tags":["analytics tools","Data Science","data science and analytics","Data Science Career","Data Science Jobs"],"author_name":"Sreejani Bhattacharyya","publish_date":"2022-06-15T13:00:00","publication_year":"2022","word_count":367,"keywords":["big data","data science","Go","business intelligence","analytics tools","programming_languages:R","AI","programming_languages:Go","Data Science Jobs","Data Science Career","analytics","GAN","Data Science","R","data science and analytics"],"extracted_tech_keywords":["AI","data science","analytics","R","Go","big data","GAN","business intelligence","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/attend-this-power-session-to-launch-your-career-in-data-analytics-business-intelligence\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":45430,"title":"10 Indian Startups That Are Leading The AI Race: 2019","content":"The Indian AI startups space is booming and now encompasses various avenues such as computer vision, self-driving, retail, audio production and innovating e-commerce platforms, among others. This year we have come with a list of yet another 10 awesome startups that are revolutionising the field of artificial intelligence with their impressive technologies. These startups are about two to three years old and are already beginning to show some remarkable developments in the field of AI. Please note that this is not a ranking and the startups are listed in alphabetical order. Here is our last year’s list. 1| Comsense Comsense team Founded in 2014 by a team of ex-IBMers, Comsense has grown into a global marketing technology in a relatively short span. With a state-of-the-art development centre in Pune and offices in the US, Germany, South Africa, and Singapore, they are enabling businesses to use cognitive technology to do better data-driven marketing. Their flagship AI product DataSense felicitates bi-directional exchange of data between the Marketing Automation Platform and the organisations legacy application’s data sources. The client gets a more complete picture of their customers’ behaviour in real-time so they can act faster and create more personalised messages to improve the engagement. With a team of 30, they have been working towards comprehending the challenges associated with improving customer experience and bridging the technology and business gap through the power of marketing automation and predictive analytics. The Founding Team: It was co-founded by Sagar Babar (CEO & Co-founder), Biju Nambiar (Co-founder & Sales Director), Aniket Neurgaonkar (Co-founder & Marketing Director), Vineet Ladhania (Co-founder & CFO), Samarth Malik (Co-founder & Director – EMEA). 2| DeepSight Labs Satya Tadimeti, Nishant Veer, Rakesh Channaiah (L-R) DeepSight AI Labs was launched in 2018 with an aim to come up with a way for AI to solve the security and safety problems that human face. With an employee strength of 12, they have been able to accomplish tremendous growth. They have developed “SuperSecure Computer Vision Platform”, a smart video surveillance retrofit solution based on artificial intelligence and deep learning technologies. SuperSecure+ adds intelligence to existing or new CCTV and drones to detect objects or potential threats such as intrusion, people with helmets, masks, weapons, ANPR, face recognition, fire etc. It can also develop custom object detection to suit specific requirements. It is a universally compatible retrofit platform that works with any CCTV system to make into a smart surveillance system. This computer vision platform can analyse multiple live and recorded video feeds from CCTVs, enhance the quality of videos in real-time, and prevent incidents such as robberies, among others. It is being used in verticals such as banking, e-surveillance, smart cities, oil & gas, retail, media, transportation and others. The Founding Team: Satya Tadimeti (CEO), Nishant Veer (CTO), Rakesh Channaiah (COO). 3| Deepsync Ishan Sharma and Rishikesh (L-R) Founded in December 2018, Deepsync was started with an aim to create AI that augments human potential and not just automates them. They believe in combining human creativity and AI’s promise of efficiency. Voice came as a natural focus for them as they saw the need to improve audio production. It is now completely changing audio production with the use of AI. As India is witnessing an increase in the vernacular content, audio remains a big focus. It is not feasible for humans to scale recording manually, also, recording today is a tiresome and costly process, requiring multiple stakeholders. Their AI is making recording audio as easy as typing a document. Their AI learns hundreds of features regarding human voice such as accent and speed of speaking to bring high-quality content which is truly scalable at a fraction of time and cost. Deepsync’s technology supports all major Indian languages and creates new content in near real-time. Having raised a pre-seed round from Hong-Kong based Zeroth.ai, they are currently expanding their technology team and hiring for engineers. The Founding Team: It was founded by Ishan Sharma (CEO) and Rishikesh (CTO & Head of AI Research). They come with an extensive joint background working with voice technology, AI and blockchain research. 4| Infilect Infilect team Based in Bengaluru, Infilect is a B2B SaaS startup helping global retail brands with visual data and insights from retail stores. Every retail brand in the world has so far relied on Point-of-Sale (POS) data from retail outlets to understand their sales and marketing performance. However, POS data does not capture what is happening in retail stores, how are customers interacting with a brand’s product and why a brand’s products are not selling. Infilect provides digital “eyes” to every brand to know exactly what is happening in every retail outlet. It is doing so by using advanced deep learning and AI algorithms, which companies are using to significantly improve marketing ROI and revenue. Its deep learning architectures is fine-tuned for retail domain and also uses computer vision technology to bring better results. It currently serves several top brands in FMCG\/CPG, Telco, Consumer Durables and more. The Founding Team: It was founded in 2015 by Anand Prabhu Subramanian and Vijay Gabale with a vision to be the best visual intelligence provider to the retail industry. They are currently a team of 14 people and are bootstrapped. 5| Marax Marax team Marax is building Mars, the world’s first privacy-first discount optimisation platform. It is a SaaS tool that enables growth teams to run optimised discount campaigns for their users, driven by a budget — in a ‘privacy-first’ world — across their devices, across the lifecycle. It works by allocating just the right discount to each user to nudge them towards completing a transaction by using anonymised data. Mars uses the latest advancements in deep reinforcement learning to recommend optimised discounts which can then further be used for improving the discounting experience or marketing. Since the awareness around privacy is increasing every day, it is important to use technologies that make the above possible without invading an individual’s privacy – MARS does exactly that. Founded in 2016, Marax AI has the sole mission of solving churn for consumer internet business. It is currently being used by companies such as Rapido to optimise the discount experience for their users. The startup is funded by Zeroth.AI, HK and SmartStart fund, US. The Founding Team: It was founded by Raman Shrivastava, Prateek Gupta, and Sumant US. 6| Quadrical.Ai Dr Hugh Hind and Sharat Singh (L-R) Founded in September 2018, it is one of the youngest startups in the list. They identify how customers are currently using data and analysing their success metrics, before proposing a way to integrate their ‘best-in-class’ AI-generated guidance. The startup combines data science into solutions which are in-depth, scale seamlessly and follow strict testing. It enables effortless integration with a customised API. Quadrical.Ai also has products to maximise profits, forecast outcomes, personalise recommendations and control risk through anomaly detection. They have around 14 members in their AI team evenly split between offices in Waterloo, Ontario and Gurgaon. The startup has raised $550,000 in the founder funding round. The startup pioneers in carrying out short term predictions in avenues such as price, demand, sales and others for their e-commerce customers. The Founding Team: It was founded by Sharat Singh and Dr Hugh Hind. Others in the founding team include Vishal Arya, VP Platforms, Kitty Chachra, VP, Marketing, Communications and Yash Chhikara, VP Product, Business Development. “Made in Waterloo,” it came into existence when Dr Hugh Hind met Sharat Singh and realised that they shared a similar vision. Both had a fundamental belief in AI’s ability to transform business — while Hugh approaches AI from a strategic and mathematical frame, Sharat brings an engineering and execution perspective. 7| Signzy Signzy team A fintech startup, Signzy offers digital onboarding solutions for banks, NBFCs and other financial institutions. Their AI product solution focuses on two major areas in the BFSI industry — the KYC process and back-end operations such as REALKYC, Algorithmic Risk Intelligence (ARI) and Digital Contracts. Through its AI-based decision engines and back-office tools, Signzy is drastically reducing the customer onboarding process and have insulated banks from human frauds. With years of R&D into their computer vision tools, they have been able to recognise documents and detect forgery. They currently have an employee strength of 75 and have received initial funding from Kstart, a venture of Kalaari Capital. They have also raised Series A round of $3.6 Million from Kalaari Capital and Stellaris Venture Partners. Signzy was founded in 2015 inspired by founders’ experiences of dealing with banks as individuals and companies in previous startups. As the banks were going digital banks were spending huge resources in doing functions such as customer onboarding, back-office operations which they felt could be automated using AI. The Founding Team: Signzy was co-founded by Ankit Ratan, Ankur Pandey and Arpit Ratan. While Ankit and Ankur both are IIT graduates and bring the expertise of data science and app-development, Arpit is a corporate lawyer who brings in the knowledge of financial regulations and expertise of working with statutory bodies, to the table. 8| Vedalabs Vedalabs team Vedalabs is a retail analytics platform that allows retailers to analyse in-store traffic and business metrics associated with it, thereby presenting real-time business insights to retailers. It was founded in January 2018 by a group of technology folks hailing from institutes like CMU, IIT, NIT, Manchester Institute, etc. with a combined experience of over 40 years in the AI domain. Vedalabs was conceptualised with a primary aim to make everyday CCTV cameras smart as the founders felt that while there are millions of CCTVs deployed across the country, there is no proactive mechanism that monitors it. Hence they decided to use AI and machine learning to build a system that can do that. They currently have 12 employees and with the recent funding that they raise, they aim to scale their team and current pipeline to establish market dominance. The Founding Team: Veer Mishra, Saurabh Shandilya, Vivek Singh, Saurabh Yadav 9| Wobot Wobot team Wobot.ai is an AI-powered computer vision SaaS that helps hospitality, food and retail businesses monitor operations. This plug-and-play tool connects to any existing CCTV or other forms of cameras and helps detect and track anomalies in standard operating procedures (SOPs) including keeping a check on food quality parameters. It currently offers various modules including hygiene, food safety pilferage and customer experience. Wobot can detect everything from people cooking with or without hair nets, packaging with or without gloves, whether safety attire or a uniform is worn or not, mopping of floors, raw material quality, number of transactions at a restaurant and even whether guest interactions happen as per SOP or not. It is revolutionising the video analytics function and solving the biggest problem of monitoring operations across a vast setup. It currently has over 2000 installations and some of the partners are IRCTC, CultFit, Barista, Travel Food Services, EatFit, Blue Tokai and others. The team size of Wobot currently is over 35 with people across design, engineering, data tagging and customer support. They recently raised a round from a few well-known angels and are now planning Series A round over the next quarter. The Founding Team: It was founded in 2017 by Adit Chhabra (CEO), Tapan Dixit (COO), and Tanay Dixit (CPO). Pranav Anand (CCO) also joined the ranks soon. Together they bring experience in food, hospitality, retail, compliance and have built products using ML and DL in the past. 10| Zappy.AI Zappy.AI team Zappy has developed Cognitive Robotic Process Automation Software which is 30x faster, easier and cost-effective compared to traditional automation software. With their two product lines Zappy and Zappy Analytics, they are driving the world’s fastest automation creation software that can observe user actions on desktop and creates automation within minutes. Their advanced algorithms are based on a combination of supervised, unsupervised and reinforcement machine learning giving Zappy the ability to automate the task by observing people at their work computers. It uses AI to discover all business processes in an organisation that can be optimised and offers unique automated solutions. Zappy Analytics collects, analyzes and process data from every employee in the organisation and uses machine learning capabilities to present an analysis of the employees’ time spent on various applications over the entire Business Process to the employer. They currently have ongoing pilots and implementations with large European banks, conglomerates, telecoms and clients across the UK, India, Japan, France and Poland. The Founding Team: It was founded in 2017 by Ambuj Agrawal (CEO) with an aim to allow employees to get rid of their repetitive tasks. The startup has received several grants such as from NetApp Innovation, Ministry of Economy Luxembourg and Google.","excerpt":"The Indian AI startups space is booming and now encompasses various avenues such as computer vision, self-driving, retail, audio production and innovating e-commerce platforms, among others. This year we have come with a list of yet another 10 awesome startups that are revolutionising the field of artificial intelligence with their impressive technologies. These startups are […]","categories":["AI Features"],"tags":[],"author_name":"Srishti Deoras","publish_date":"2019-09-04T13:03:17","publication_year":"2019","word_count":2113,"keywords":["data science","machine learning","artificial intelligence","AI","ML","computer vision","RAG","Aim","deep learning","analytics"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","deep learning","computer vision","data science","analytics","Aim","RAG"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/10-indian-startups-that-are-leading-the-ai-race-2019\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10163360,"title":"How Women in Tech Navigate Work-Life Balance &amp; Career Growth","content":"The United Nations observes the International Day of Women and Girls in Science on February 11 every year to highlight the importance of equal access to and participation in Science, Technology, Engineering, and Mathematics (STEM). The day serves as a powerful reminder that gender equality in STEM is not just a goal but a critical factor in shaping a more inclusive and innovative world. While India boasts the highest percentage of female STEM graduates in the world – nearly 40% in 2021-22 – only 14% of STEM jobs in the country are held by women. This highlights a critical issue in the tech industry – the ‘leaky pipeline’. Women often exit the workforce at key life stages, including marriage or childbirth, and, sometimes, when family responsibilities intensify. This has led to the underrepresentation of women in leadership roles, innovation, and research. A study by BiasWatchIndia highlighted the gender disparities within STEM fields. Engineering, for instance, has the widest gender gap, with women representing only 9.2% of faculty members. Fields like biology see slightly higher female representation, at 25.5%. The overall gender gap, however, remains a significant challenge. This trend underscores the systemic barriers women face, especially in traditionally male-dominated disciplines such as engineering, physics, and computer science. One question remains: how can women in tech achieve a sustainable work-life balance while overcoming these challenges? What Women Leaders Think Women leaders in tech are at the forefront of driving change, pushing for inclusive cultures, and redefining work-life balance. Here are a few stories that provide insights into how women can thrive professionally while maintaining personal well-being. In a conversation with AIM, Suprabha Mysore, vice president of software development at TransUnion Global Capability Centre (GCC) in India, emphasised the importance of balance for true success. “We can never be great at our work if work is the only thing that we do! True success lies in finding a balance between all the things that matter to us,” she said. Suprabha’s perspective highlights that success is not just about career achievements but also about nurturing other aspects of life, such as relationships, health, and personal happiness. Birlasoft’s chief operating officer, Selvakumaran Mannappan, shares a similar view. He said that work-life balance remains a critical challenge for women in tech, often impeding career growth and retention. Mannappan advocates for workplaces that prioritise flexibility, mentorship, and career development. “True inclusion means women in tech can lead, innovate, and grow without sacrificing work-life balance, aspirations, or opportunities,” he stated. Birlasoft’s BEmpowered Women Leadership program is one such initiative. It equips women with the skills they need to progress into senior roles and fosters a culture where women can balance their personal and professional aspirations. Kavitha Krishnan, vice president of AI innovations in procurement at SAP Labs India, also stressed the importance of intentional focus and balance. For her, mornings are sacred, a time for deep focus and high-impact tasks. “Work-life balance is about making intentional choices,” she explained. For Kavitha, maintaining a focus on health through exercise, mindful eating, and proper rest forms the foundation of her well-being, allowing her to tackle both work and personal life with energy and creativity. Medhavi Singh, country head at Criteo India, told AIM that the conversation about work-life balance goes beyond flexible hours or remote work arrangements. “True leadership in tech is about prioritising the well-being of those who make business success possible.” At Criteo, Medhavi emphasised that a supportive culture allows women to excel in their careers while maintaining personal well-being. Whether it’s through adjusting work schedules or providing opportunities for personal growth, creating a flexible work environment is key to empowering women. Meanwhile, Rekha Sahay Ghosh, head of marketing and communications at Sasken Technologies, said, “It’s about prioritising tasks, being flexible, and setting clear boundaries.” Efforts Towards Gender Diversity Efforts to close the gender gap in engineering have been propelled by women like Sudha Murthy, who became the first woman in India to study mechanical engineering, paving the way for future generations. The Infosys Foundation has taken steps toward increasing gender diversity by partnering with the ICT Academy of Tamil Nadu. The initiative aims to establish Centres of Excellence for Women and Youth Empowerment across 450 colleges in India, offering skills training and job placement assistance to rural learners. The program also includes soft skills development and coding practice sessions to ensure participants are prepared for real-world tech industry challenges. Similarly, TalentSprint’s Women Engineers (WE) program, supported by Google, is creating an inclusive ecosystem for women in tech. The program has attracted over 1.3 lakh applicants from diverse backgrounds, offering scholarships, mentorship, and career opportunities. With 39% of participants from low-income families and 25% from rural areas, WE aims to empower aspiring women engineers and support their growth in high-growth tech careers across India.","excerpt":"While India boasts the highest percentage of female STEM graduates in the world – nearly 40% in 2021-22 – only 14% of STEM jobs in the country are held by women.","categories":["AI Features"],"tags":["AI (Artificial Intelligence)","STEM","Women in Tech"],"author_name":"Vidyashree Srinivas","publish_date":"2025-02-13T09:00:00","publication_year":"2025","word_count":797,"keywords":["Go","API","programming_languages:R","AI","STEM","innovation","programming_languages:Go","Aim","Women in Tech","ViT","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","Aim","R","Go","API","ViT","innovation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-women-in-tech-navigate-work-life-balance-career-growth\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10097067,"title":"Why Foxconn-TSMC Partnership is Unlikely","content":"When it comes to the Indian semiconductor space, Foxconn has grabbed more headlines than any other player in the past year or so — but mostly for the wrong reasons. When Vedanta and Foxconn announced their joint venture last year, many perceived it as the point that kickstarted India’s semiconductor manufacturing ambitions. Now, since that plan went kaput, it’s now being speculated that Foxconn may partner-up with the Taiwan Semiconductor Manufacturing Company Limited (TSMC) and Japan’s TMH Group to keep its ambitions of building chips in India alive. Why is TSMC an unlikely partner? Today, TSMC is one of the largest chip foundries in the world catering to the likes of Apple, NVIDIA, Google and AMD, among others. Headquartered in Hsinchu, Taiwan, TSMC was the first foundry to market 7 nm and 5 nm chips and is currently working on the 3 nm and 2 nm ones. Considering TSMC’s prominent position in the semiconductor industry, this news has led to a sense of optimism and hope among many Indians. However, Arun Mampazhy, semiconductor analyst, is a bit sceptical. He believes, TSMC was hesitant about moving outside Taiwan. “It’s only because of geopolitical factors and the US agreeing to protect Taiwan that TSMC has agreed to set shop there,” Mampazhy told AIM. Besides, TSMC is also setting up fabs in Japan and Europe. In Japan, TSMC aims to manufacture 5 nm and 10 nm chips from the second half of the decade. Similarly, in Europe, TSMC is in talks with Germany, to set up a fab in Saxony. Mampazhy believes that establishing fabs in Germany and Japan would be a logical move for TSMC. For instance, in Japan, Sony and Denso serve as partners to TSMC and contribute significantly to its business, thereby providing an advantage for TSMC by setting up a fab in Japan. However, Foxconn does not bring similar benefits to TMSC in India. “Given Foxconn is only an assembler, it is unlikely that TSMC will agree to do a technology transfer or share that technology with Foxconn,” Mampazhy said. The only way he sees a potential partnership with TSMC is if Foxconn assumes the responsibility for the financial aspects of establishing a fab in India. “However, even with such support, TSMC faces considerable challenges in managing the workforce required for its upcoming fabs in Japan, the US, and Germany.” Is the trust in Foxconn waning? Foxconn tryst with semiconductors isn’t new. Back in 2018, the Taiwanese electronics contract manufacturer had announced plans to establish chip manufacturing facilities in China, but no tangible progress was made on that front. Similarly in India, their joint venture with Vedanta, which was intended to initiate their foray into chip production, too failed to materialise. “Foxconn has a history of promising and not delivering in multiple places. Wisconsin is one, then, there have been other instances of engaging with the Saudi Arabian government, Malaysia, and possibly even Mexico. In about five or six countries, Foxconn has been known to fall short on fulfilling its commitments,” Mampazhy said. In fact, Foxconn has already been part of many controversies in India. Last year, the contract manufacturer found itself embroiled in one when both Telangana and Karnataka governments simultaneously claimed Foxconn to have signed up for big investments in their respective states to manufacture electronics. In a manner reminiscent of their past experiences, a similar situation unfolded in Foxconn’s foundry plans. Their proposed joint venture with Vedanta, which was well-advanced in talks for a site in Maharashtra, abruptly shifted to Gujarat just before the state elections. “Foxconn has the reputation for being one of the most opaque companies in an opaque world,” Lawrence Tabak, the author of Foxconned: Imaginary Jobs, Bulldozed Homes, and the Sacking of Local Government, said.  According to him, Foxconn had no problem in making big announcements for the cause of a government’s electoral ambitions without having any concrete plan of action. Hence, given Foxconn’s dubious reputation, the TSMC news could also be smoke and mirrors. Taiwan’s interest in India The possibility of the two Taiwanese companies collaborating to establish a fab in India, however, remains to be seen. It is worth noting that the Taiwanese government has exhibited keen interest in India over time. Baushuan Ger, Taiwan’s ambassador to India, earlier said that both nations should sign the free-trade agreement (FTA) at the earliest. Ger said Taiwan is willing to share its expertise with India in critical sectors such as semiconductors, 5G, information security, and AI. Most recently, Taiwanese foreign minister Jaushieh Joseph Wu also sparked discussions around a FTA with India. While interacting with the press in Taipei, Wu said that Taiwan has been actively encouraging its companies, which no longer find the Chinese market profitable, to relocate their production facilities to India. Given, Taiwan alone accounts for about 60% of the total global foundry market, India stands to greatly benefit from such a deal. However, despite keen interest from Taiwan for a FTA with India, nothing has materialised so far.","excerpt":"It’s being speculated that Foxconn may partner-up with TSMC and Japan’s TMH Group to keep its ambitions of building chips in India alive","categories":["AI Features"],"tags":["foxconn india","india fab","Semiconductor India","tsmc india"],"author_name":"Pritam Bordoloi","publish_date":"2023-07-18T11:00:00","publication_year":"2023","word_count":830,"keywords":["Go","Rust","programming_languages:R","AI","foxconn india","tsmc india","programming_languages:Go","RAG","GAN","Ray","Aim","Semiconductor India","india fab","R"],"extracted_tech_keywords":["AI","Aim","Ray","RAG","R","Go","Rust","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/why-foxconn-tsmc-partnership-is-unlikely\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10103615,"title":"Why Sam’s Return to OpenAI is Good For Microsoft","content":"Emmett Shear, who was appointed the CEO of OpenAI on Monday, may have lasted only 72 hours in the role, but it was still longer than Sam Altman’s stint as the leader of a new advanced AI research team at Microsoft. And while Altman may be expressing his love to Emmett now, he is definitely sighing in relief for returning to OpenAI, as he wouldn’t have stayed put with Microsoft for long. What makes us say that? Read on. Shackles on Personal Ventures When asked about Altman’s plans on pursuing his side projects when at Microsoft, Satya Nadella responded by acknowledging Altman’s broad interests and investments, and confirmed that they would work around it and the ‘governance aspect of it’. However, he also emphasised in a cryptic manner that Altman would only work at Microsoft ‘if he wants to spend his full time on pursuing the mission’ there, similar to how he would at OpenAI. Unfortunately for Microsoft, Altman returned to OpenAI within the next few days. With a number of side investments that Altman has made into projects ranging from cryptocurrency, biotech, energy and many more, working for a major conglomerate will come with its limitations. As Nadella mentioned, the governance aspect, probably conflict of interest, would hamper Atlman’s involvement in those ventures. Controlled Autonomy Venture capitalist and CEO of Social Capital Chamath Palihapitiya, recently tweeted about the pros of an AGI coming from a startup vs that from a big-tech company. He believes that startups are deemed preferable owing to their risk-taking nature and potential for pure, but potentially risky innovation. Whereas, AGI emerging from big-tech companies is anticipated to be heavily regulated and constrained by terms and conditions, diminishing its true potential. Furthermore, startups, driven by engineering initiatives, are seen as more likely to push the boundaries of AGI development, while big tech with its abundance of legal frameworks and market capital, may prioritise caution over groundbreaking advancements. The autonomous nature of functioning that happens in a startup cannot be replicated in a big tech company — a reason that could have been a potential deal breaker for Altman. Elon Musk did not forget to share his scepticism on the matter, hinting that OpenAI being independent from Microsoft will avoid a concentration of power Shrouded in Secrecy Going by the latest developments on how a powerful AI breakthrough that could possibly threaten humanity was made just before Altman’s ousting, the secretive nature of OpenAI allowed Altman and the team to proceed without hindrance — except for the board fiasco. However, with the way things are unfolding, with Altman’s close aide Bret Taylor already on board, and more supporters expected, Altman is securing the future of his company. This would allow him to work in confidentiality, a privilege he would have never got at Microsoft. Big Fish in a Small Pond While autonomy is an important aspect, there is a cultural perk with growing a startup. Going by the insane level of employee support that OpenAI received (with over 700+ employees signing letters threatening to quit if Altman was not brought back), reflected the close-knit employee community he had built in the company. Such unity is highly unlikely in a big tech. Considering the entrepreneur and tech guru that Altman is, he would not have survived in a closer controlled environment that Microsoft would have provided. On a lighter note, Altman would have had to resort to using Microsoft products such as Teams, as opposed to Google Meet which OpenAI used for firing him — something even Musk took a jibe on. While Microsoft showered love on Sam & Co., keeping Mac laptops and San Francisco office space ready for OpenAI employees, Altman chose to come back to OpenAI. With an already existing solid symbiotic partnership between both companies, Altman at OpenAI would definitely work better for both. Altman’s return brings autonomy and unrestricted power, allowing OpenAI to continue on all its planned commitments. In the process, a win-win for all. Sam is happy, so is Satya.","excerpt":"Because, Microsoft would never have worked out for Sam.","categories":["Global Tech"],"tags":["AGI","Biotech","Bloomberg","Cryptocurrency","Emmett Shear","energy","governance","investments","Microsoft","OpenAI","Sam Altman","Satya Nadella","VC"],"author_name":"Vandana Nair","publish_date":"2023-11-24T12:10:39","publication_year":"2023","word_count":669,"keywords":["API","investments","Cryptocurrency","Bloomberg","R","Satya Nadella","energy","Sam Altman","AI employees","venture capital","Biotech","startup","VC","Go","Replicate","AI","Emmett Shear","AGI","OpenAI","innovation","governance","Microsoft"],"extracted_tech_keywords":["AI","OpenAI","R","Go","API","innovation","startup","venture capital","Replicate","AI employees"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/why-sams-return-to-openai-is-good-for-microsoft\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10101119,"title":"Meet Silicon Valley&#8217;s Generative AI Darling","content":"Looks like the entire Silicon Valley is head over heels for Anthropic. According to recent reports, the company is ready to raise another round of funding from past investors, including Google. One of the prime independent rivals of OpenAI, the company is in talks with investors to raise around $2 billion in funding. This comes just a week after Amazon committed $1.25 billion for Anthropic, with plans to invest a total of $4 billion in the future. In return, Amazon expects to be the sole cloud provider for Anthropic. Interestingly, Google has already made a $300 million investment in Anthropic, acquiring a 10% stake in the company. The two-year old startup that is building Claude, a rival to ChatGPT, now aims to get a valuation between $20 billion to $30 billion, which is ten times more than the current $4 billion after the investment in March. To put this in perspective, OpenAI roughly has a valuation of around $28 billion after raising several rounds of funds. The CEO of Anthropic, Dario Amodei, said in a recent interview with Andreessen Horowitz, that the biggest thing that the company wants to do is make Claude have infinite context windows. The only things holding it back, according to Amodei, is that “at some point, it just becomes too expensive in terms of compute”. It is clear that Anthropic has high expectations when it comes to what it wants to achieve. But it seems like the current funds are holding the company back from its ambitions. It is only fair for the company to go around looking for more funds and raise the stakes for the big-tech. A tussle with Google? Interestingly, there are have been rumours about a senior Google engineer delivering some challenging news to over fifty colleagues. Here, a segment of the company’s cloud services, crucial for Anthropic, was experiencing issues necessitating overtime efforts to rectify the situation. To address the problems in their service, specifically, an underperforming and unstable NVIDIA H100 cluster, Google Cloud leadership initiated a month-long, seven-day-per-week sprint. The consequences of not resolving this issue were deemed substantial, affecting Anthropic primarily but also leaving an adverse impact on Google Cloud and Google as a whole, as per the documents examined by Big Technology. Just a week after Google launched the sprint, Anthropic announced its deal with Amazon, designating Amazon Web Services as its primary cloud provider for mission-critical workloads. It’s worth noting that the Amazon deal had been in the works for a while and was unrelated to Google Cloud’s performance problems. Nevertheless, for Google, this development must have been unsettling, especially considering Google had invested all this money into the company. Nevertheless, Anthropic’s new funding from Amazon is undoubtedly a benefit for Google as well, as the value of its share would also increase, not just for Amazon. On the other hand, Google is already developing its own AI models with Google DeepMind. Gemini, which is expected to be arriving soon, might be the biggest bet the company has made. While Google may have the capability to manage these endeavours simultaneously, it faces the risk of being outpaced by competitors with fewer complicated trade-offs. Notably, Google Cloud’s performance issues with Anthropic appear to be stabilising, albeit not without requiring engineers to engage in a rare phenomenon at Google — weekend work. OpenAI went from $25m run rate to $1bn in a year, Anthropic from < $10m to $200m this year, $500m next yearIt looks like they are raising at 100x sales, but have you ever seen sales growth that high?And the enterprise market hasn't even started yet, gen AI will be everywhere— Emad (@EMostaque) October 4, 2023 Everyone loves Anthropic Even though OpenAI is not profitable yet, it is still generating revenue through its offerings. Anthropic also has plans to make its generative AI capabilities generate revenue for itself, and aims for an annualised pace of $200 million. It also hopes to generate a $500 million annualised rate, according to a person with knowledge. Anthropic believes that these AI models from companies like OpenAI would be ahead of everything in the next few years, and it would be impossible to catch up with them. This is clearly why every AI startup in the world wants its valuation to be the highest at the moment, to stay ahead in the race. At present, generative AI startups are the biggest draw for investors and cloud providers. Emerging startups such as Mistral AI, Reka AI, Cohere AI, and Inflection AI, all have been raising funds, and have their own strategies for making bucks. Amidst all this, the investors and big-tech are running for their money. Anthropic has raised the stakes even more, as in the end, the only moat that generative AI companies have is money. Interestingly, FTX had a $500 million stakeholder in Anthropic. But even after bankruptcy, the Sam Bankman-Fried led company stopped the sale of its shares. Now, three months later, the stakes would be worth $2 billion, effectively making its customers very happy. How could someone not love Anthropic? https:\/\/twitter.com\/tanayj\/status\/1709402004643106818","excerpt":"Even the crypto folks are dreaming about FTX’s return from Anthropic","categories":["AI Features"],"tags":["Anthropic","Anthropic AI","Anthropic funding","Claude"],"author_name":"Mohit Pandey","publish_date":"2023-10-05T10:42:39","publication_year":"2023","word_count":844,"keywords":["Anthropic","ChatGPT","Go","OpenAI","AI","Claude","GPT","Aim","generative AI","Anthropic funding","R","Anthropic AI","startup"],"extracted_tech_keywords":["AI","generative AI","ChatGPT","OpenAI","Anthropic","Aim","R","Go","GPT","startup"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/meet-silicon-valleys-generative-ai-darling\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10049089,"title":"Are No-Code Platforms Suitable Partners for Businesses? Hubbler CEO Discusses","content":"Creating applications generally requires immense coding and expertise, usually over a period of time. It’s not the easiest task and definitely not everyone’s cup of tea. Entering into this landscape, Karnataka based Hubbler is simplifying the creation of a business process into a web or mobile application. The organisation assists businesses by enabling app creation within 24 hours – without writing a single line of code. Analytics India Magazine caught up with Hubbler’s founder, Mr Vinay Agrrawal,  to get an insight into his mission of creating a digital workforce on seamless app development. AIM: We’d love to know the story of the origin of Hubbler. Please take us through it. Vinay Agrrawal: I started my journey through Unicel in 2003, my previous startup. As an entrepreneur, I had to face many challenges and chaos a growing company throws. At Unicel, I spent considerable time and money to organise the chaos that I faced and took up a challenge to solve this problem using technology and innovation. Negotiating these challenges made me realise that I have a bigger goal to achieve. The idea of a new startup – Hubbler, was born. In 2015, a global multi-channel mobile messaging platform, mGage India Private Limited, acquired Unicel to bolster its presence in India’s enterprise mobile marketing in a deal that was widely reported. With this deal, I took an exit from Unicel to pursue what I loved most – building a company from scratch. With the objective to create a new venture, I started to create a team for Hubbler. Today, Hubbler is empowering so many organisations to create an app in just hours without writing a single line of code. With a revolutionary platform, Hubbler plans to change the way enterprises build and consume applications. AIM: What makes Hubbler stand out from the competition? Vinay Agrrawal: The super APP Concept. Hubbler is designed to reduce the number of applications that a business user needs to interact with. Not only can one fork a new application on Hubbler as needed, but also connect their legacy systems to Hubbler and interact with it using Hubbler as an interface. 40+ Intelligent Hubbles​: Hubbles are configurable microservice-based components that help build truly vertical\/function agnostic applications and change them on the fly​.  Connect & Engage: Create and track organisation goals. Reward people and celebrate success.  Device Agnostic: Deliver applications natively to Android, iOS and Web​. Platform beyond workflows: Build around accelerating business transformation strategy than capturing just the workflows. AIM: Tell us about your flagship products and services. Vinay Agrrawal: Hubbler is a cloud platform that helps organisations build any business applications overnight without writing a single line of code. The platform has the capability to configure any kind of business application. Our customers have built over 1000+ applications on our platform. We have deployed complex solutions like Procure to Pay Cloud, Lease Management, Revenue Operations, Expense management, Field Force automation and even entire billing systems, etc. Many such solutions can be easily built on Hubbler, and possibilities are endless. Some of the startups have built their entire tech platform on us, allowing them to hit the market in a few months. AIM: Can you explain the tech behind creating an app without coding? Vinay Agrrawal: Ironically, there is a lot of code behind building a non-code platform. At the outset, we rejected hybrid application approaches for a better user experience. All the applications built on Hubbler are available natively on all the form factors. We have built a proprietary system that generates not only the user interface but also a backend database and business middleware with just a click of a button. Our core tech platform is written in Python, while the user interface uses SWIFT on iOS, Kotlin on Android, and JS on the Web. AIM: What market are you tapping into? What’s your expansion plan? Vinay Agrrawal: For the last couple of years, we had been working on perfecting the platform and learning as much as possible by solving different problems. Today, we have customers from 15+ industries with over 500+ applications in India. Predominantly, we target growth stage mid-size enterprises across verticals. We have all plans ready for an international launch (North America, Europe & Middle East) by Q3 this year, and therefore, we are raising funds now. These markets are pretty mature to adopt a no-code platform. According to the Market and Markets report on low and no-code platforms, an industry estimated to USD 45.5 Bn by 2025, a market adoption analysis found that North America and Europe were the biggest consumers of low and no-code platforms followed by the APAC region. The Middle East was also touted as a key consumer of low\/no-code. AIM: Would you please elaborate on how the development of Hubbler-created apps helps them join shoulders with those developed by coders, the conventional way? Vinay Agrrawal: We gave a lot of thought to Hubbler’s ability to work with the existing or future ecosystem of the business. We acknowledged the fact that we can’t operate in silos. We have built an extensive API framework to make Hubbler work with any external system. Our architecture allows data to be handed over from any of the applications to Hubbler and then back to the same or any other application for further processing if needed. AIM: How does Hubbler contribute to India’s growing data science ecosystem? Vinay Agrrawal: We have seen many businesses talking about AI and data science in their business. However, the core issue is that their data is still trapped. The data is either scattered on multiple systems or worse, is trapped in papers and emails. At Hubbler, we are strong proponents of basic hygiene. Firstly, we are helping businesses digitise the processes that they are doing either manually or on excel sheets. Secondly, using API connectors, we are not only bringing in scattered data in a single place but also organising it based on organisation structure and processes. This ensures data can be consumed in the most efficient way. AIM: Tell us about the work culture at Hubbler. Vinay Agrrawal: Hubbler nurtures leaders, learners and challengers. We work beyond hierarchy barriers and foster a culture of collaboration. Our teams can be found in huddles, discussing everything from new application development to the latest Netflix binge. We believe that a workplace you love will create a job you love – happy teams make happy customers! AIM: What are your views on the current data policies in India? Vinay Agrrawal: There’s a lot that has been done in the last 4-5 years to ensure that our policies match with the best practices followed in the West. However, we have yet to see a push from the government to open source the data trapped within their multiple departments. All government departments should offer anonymised data (for privacy) as an API. This would result in a wave of new interesting applications that would not only help citizens but also create a lot of employment opportunities.","excerpt":"“Ironically, there is a lot of code behind building a non-code platform.”","categories":["AI Features"],"tags":[],"author_name":"Avi Gopani","publish_date":"2021-09-21T12:00:00","publication_year":"2021","word_count":1163,"keywords":["data science","Go","API","AI","ML","Git","Python","Aim","analytics","R"],"extracted_tech_keywords":["AI","ML","data science","analytics","Aim","Python","R","Go","Git","API"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/are-no-code-platforms-suitable-partners-for-businesses-hubbler-ceo-discusses\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":49779,"title":"State Of Data Science In Domestic Indian Market 2019 – By AIM &#038; SAS","content":"This year’s study done in collaboration with SAS dives into 50 large-sized firms to better understand analytics maturity and penetration in these organisations. Our in-depth study can also be seen as a deep dive benchmarking report to see how companies are winning with analytics. The study also provides a greater understanding of how analytics mature organisations are doing to draw tangible results from their investments and how other companies can adopt these learnings and practices to move up the maturity curve. While many organisations are accelerating their investment in analytics, how well are these organisations able to leverage these analytics investments is yet to be seen. Our study reveals that the e-commerce sector has emerged as one of the biggest adopters of analytics, thanks to the fact that this sector is not weighed down by legacy infrastructure constraints. Other sectors that boast of high analytics adoption are banking and auto which are undertaking major transformational initiatives and are embedding analytics across functions. Senior analytics leaders from these sectors recognize the need for building an immersive analytics environment and are assiduously working towards raising the analytics IQ in their organisations. Read last year’s study here. Definitions & Companies We analysed 50 large-sized companies in India by Revenues as of 2018, for this study. IT Service providers in India are not part of this study. Also, some leading start-up Unicorns and e-commerce players were included in the study. Analytics penetration is a metric that denotes the degree of infusion of analytics function in an organisation. It quantifies the approximate number of analytics professionals employed by the organisation for every employee within the whole firm. So, a penetration of 1% should be read as 1 analytics\/ data science professional for every 100 employees with the organisation. Analytics maturity is the metric to quantify the quality and depth of analytics adopted within an organisation. Maturity is a combination of 3 factors: employee tenurepercentage of artificial intelligence in the analytics function employee seniority in that organisation Analytics adoption is calculated at the penetration of less than 0.75% i.e. the organisation does not even have 1 analytics professional to support 133 employees within the organisation. It is important to note that these are relative metrics and organisations, as plotted here, are based on a comparative study with respect to each other and should not be compared to other organisations outside this study. Also, the values may differ in terms of global standards. Another point to keep in mind in this study is that the numbers would be negatively biased towards organisations that outsource their analytics functions heavily. Analytics & Data Science At Indian Firms – Key Trends The overall adoption of Analytics & Data Science at large Indian firms is around 70%. In other words, 70% of all large firms in India have adopted analytics in some form. This is higher than last year’s rate of 64%. That’s a healthy adoption rate given most of these large firms are into traditional businesses like energy & utilitiesTelecom, Financial Services, Ecommerce & Private Sector Banks have almost 100% adoption rates. Some level of analytics & data science is being executed in these organisations, especially the large onesBanking and auto industries are on firmer ground with 100% and 83% analytics adoption rateTelecommunications and utilities also report a high analytics adoption of 100% and 80%In contrast, infrastructure and power lag behind with low analytics adoption Some of the traditional sectors that are seeing an increase in analytics adoption are –SteelOil Drilling & RefineriesFor the rest of the sectors, the adoption rates remain the same as last year Companies with Biggest Analytics Unit To realize the maximum business impact, large organisations and new-age companies have set up dedicated Analytics Services Unit — one of the first steps towards developing a 360-degree vision of existing data and building innovative solutions and initiatives. Based on our survey, we identified 5 Indian consumer firms have the biggest analytics units in the country. These 5 organisations also boast of enterprise-wide analytics implementation, data-driven architecture and C-level leadership. These 5 Indian consumer firms have the biggest analytics units in the country ICICI BankFlipkartHDFC BankAxis BankBharti Airtel Analytics Hubs In India In terms of city-wise adoption, Mumbai leads the pack becoming the largest data analytics employer in the country. After Mumbai, Bangalore and Delhi region have a higher analytics concentration. 44% of all analytic functions for Indian firms are based out of Mumbai, followed by Delhi NCR at 29%Almost 91% of analytics functions are based out of just three cities – Mumbai, Bangalore & Delhi\/NCR6% of analytics functions for companies are based out of Pune Analytics Units In Organisations There are several ways data and analytics function in an organisation. In one model, data and analytics function works as a standalone unit, while in certain cases, analytics is embedded in the IT function with a team working in collaboration with traditional IT team on areas where analytics can deliver the greatest value. Our study finds that despite the high adoption rate, analytics is leveraged as a support function for traditional operations like Sales & Marketing, IT and Operations as opposed to a fully standalone service unit. 37% of analytics functions in large Indian organisations support the Sales & Marketing group. 19% support the operations group and just 15% support the IT groupContrary to the most widely accepted belief, Analytics functions do not usually fall under the IT unit for Indian firms Indian Firms – Key Metrics For Analytics Function With analytics becoming a core competency across sectors, attracting and retaining talent has become a ‘key mandate’ for senior leaders. In the last couple of years, we have seen organisations launching dedicated programmes to source the best analytics talent. Besides attracting talent, it is also important for organisations to focus on retaining analytics talent by developing talent engagement programmes. Talent maturity, as reflected by employee tenure is 4 years this year as compared to 3.4 years in 2018. While analytics penetration in Indian firms is only 2.5%. On average, large Indian firms have an analytics penetration of 2.5%. This essentially implies that for every 40 employees in the organisation, 1 employee is in some shape associated with data and analytics. This is slightly lower than 2.8% from last year. Overall, the large firms have decreased the proportional analytics headcount servicing their business this yearThe average tenure of analytics professionals at Indian firms is 4 years, slightly higher than last year at 3.4 years.Public sector banks, despite low adoption, have the highest analytics tenure among India firms, at almost 5.6 YearsOn the other hand, E-commerce firms, with high adoption of Analytics, have a much lower analytics tenure at 2.5 YearsOverall, tenure for all the sectors have gone up in the last one year except for Power & Steel Tenure Vs. Adoption We looked at the sector across tenure vs adoption for analytics matrix. Here are some key findings: Most sectors in India fall into Quadrant I i.e. High adoption\/high tenure relative to other sectorsE-commerce sector, despite the high adoption rate of analytics, severely lacks on the tenure of its analytics professionals Power Sector lags both on employee tenure and analytics adoptionPublic Sector Banks and Oil & Refineries have identical adoption & tenure In terms of talent retention, Indian telcos lead the race with professionals having the highest experience. This is also indicative of employee engagement which helps drive talent retention in organisations. Telecom sector has some of the most interesting and complex problems with high consumption of analytics and telecom majors doubled down by providing new career paths to attract and retain talent. The median experience level of Analytics professionals with Indian firms is 8.2 yearsAnalytics professionals in Indian telecom firms have the highest experience at 10.2 yearsOil Drilling & Refineries have the lowest at 7.9 years Analytics Penetration Vs. Maturity Like the previous year, here we analysed various sectors on the basis of analytics penetrations and analytics maturity. The Quadrant measures industries on two parameters Analytics Penetration and Analytics Maturity. As outlined earlier, maturity signifies the scope of analytics within an organisation while penetration measures the approximate number of analytics professionals employed by the organisation for every employee within the whole firm. So, a penetration of 1% should be read as 1 analytics\/ data science professional for every 100 employees with the organisation. In Quadrant I, companies are raising the bar with analytics both in terms of penetration and maturity. Their enterprise success comes from a high level of depth both in terms of analytics penetration and maturity. In Quadrant II, companies are scoring big on analytics maturity but low on analytics penetration. From BFSI to telecom, utilities, traditional sectors are reshaping their strategies with high analytics adoption but scores less on analytics penetration. Quadrant III is where we see sectors lagging behind on both metric – analytics penetration and analytics maturity. These sectors are missing out on a huge opportunity. Quadrant IV denotes high analytics penetration but low analytics maturity. This implies that companies need to have a tighter focus on strengthening analytics capabilities. Following Conclusions Can Be Made From The Graph E-commerce continues to be the only sector in the First Quadrant, which signifies that it is both high on penetration and maturity. With Data Science becoming a core part of the e-commerce sector,  all mid to large e-commerce firms have some form of analytics adoption within the firmThe highest number of sectors lies in Quadrant II. This is a good indication of adoption as well as relative maturity of analytics functions within these sectors Comparison From Last Year E-commerce has decreased in both maturity & penetration as compared to last year Please note that maturity & penetration for each year is a relative index to other sectors So, while the absolute maturity and penetration for e-commerce might have increased, it has decreased when compared to other sectors this year. The biggest changes can be seen in the Steel industry, undergoing digital transformation. The sector moved from Quadrant III to Quadrant II signifying an increase in maturity of the analytics functions in the sector Oil Drilling & Refineries have also increased maturity, though the sector still falls in Quadrant IIIPower, Auto & Utilities have a decrease in maturity vis-à-vis last year Analytics Penetration Vs. Maturity – Private Sector Banking Private Sector Banking in India scores high on the maturity of analytics and data science deployed in the organisations, slightly lower than Telecom and e-commerce industryFinance & Sales are the two primary functions where private sector banks deploy analytics. Research is the third function where analytics get deployedMumbai is the top city where analytics functions of private Indian banks are based Since last year, the penetration and maturity of this sector has almost remained constant Analytics Penetration Vs. Maturity – Public Sector Banking Public Sector Banking in India has among the lowest analytics penetration among other sectors, almost the same as drilling and refineries sectorAgain, Mumbai is the base for all analytics functions related to public Indian banks Since last year, the penetration and maturity of this sector has almost remained constant Analytics Penetration Vs. Maturity – Auto The auto sector has a high adoption rate of analytics and data science. 5 out of 6 auto companies in India have some form of analytics adoption.Most analytics functions in the auto industry support the sales and Operations unitsMost analytics functions in the auto industry are based out of Mumbai or PuneSince last year, the Auto sector has dropped in terms of analytics maturity and remained constant in terms of penetration; vis-à-vis other sectors Analytics Penetration Vs. Maturity – E-commerce E-commerce has emerged as one of the largest adopters of analytics and is evident with their focus on hiring senior analytics professionalsIt is the leading sector in India in terms of both analytics penetration and maturityMost analytics functions in e-commerce industry support the business development units; followed by the Engineering group Analytics Penetration Vs. Maturity – Oil Drilling & Refineries We looked at 8 large-sized oil drilling & refinery companies in India. Analytics adoption is dismally low in this sectorJust 3 out of 8 large-sized companies have some form of analytics adoptionMost analytics functions in Oil Drilling & Refineries supports the business Development & Sales unitsIn terms of analytics maturity, the sector showed an increase this year Analytics Penetration Vs. Maturity – Steel We looked at 3 large-sized Steel companies in India. Analytics adoption is high in this sector2 out of 3 large-sized companies in this sector have some form of analytics adoptionThis is the only sector that has moved across the Quadrants this year (from Quadrant III to II) depicting an increase in analytics maturity this year Conclusion While some sectors have made significant progress in terms of analytics maturity, traditional sectors like Steel, Power, Oil Drilling & Refineries and Utilities, weighed down by legacy infrastructure are yet to realise the gains from an analytics implementation. In the big scheme of things, analytics adoption requires exponential shifts in terms of processes, people and technology. So, what’s holding back these traditional sectors from rising up the analytics maturity scale. Sectors with low adoption have little senior management buy-in and also lack a coordinated analytics integration. at a time when leadership plays a critical role in coalescing a strategy and tying it to the financial bottom line, these traditional sectors have little C-level accountability. At a time when enterprises are achieving a series of improvements through enterprise-wide adoption, these traditional sectors are yet to define a vision for the implementation of analytics strategies. Also, companies that lag behind in analytics adoption should put in a closed-loop process to map the baseline value out of analytics initiatives and include learnings in new initiatives. Access the complete report here State of data_science_in_domestic_indian_market_2019_aim_sas from Srishti Deoras Download the report here State_of_Data_Science_in_Domestic_Indian_Market_2019_AIM_SAS-1Download","excerpt":"This year’s study done in collaboration with SAS dives into 50 large-sized firms to better understand analytics maturity and penetration in these organisations. Our in-depth study can also be seen as a deep dive benchmarking report to see how companies are winning with analytics. The study also provides a greater understanding of how analytics mature […]","categories":["AI Features"],"tags":["sas"],"author_name":"Richa Bhatia","publish_date":"2019-11-13T11:47:50","publication_year":"2019","word_count":2298,"keywords":["data science","Go","API","artificial intelligence","AI","Git","RAG","Aim","analytics","sas","R"],"extracted_tech_keywords":["AI","artificial intelligence","data science","analytics","Aim","RAG","R","Go","Git","API"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/state-of-data-science-in-domestic-indian-market-2019-by-aim-sas\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":56615,"title":"Vinodhini Ranganathan Of Cisco Explains How To Tackle Fake Reviews Using AI","content":"At MLDS 2020, Vinodhini Ranganathan, Data Scientist at Cisco highlighted the challenges of fake reviews along with an analytics framework for identifying and removing them on web platforms. “In God we trust, all others bring data,” W. Edwards Deming said a long while ago. But in this age, the sort of optimism is challenging, especially with the online e-com ecosystem. The hunt for authentic data is where lies the true value. But, this could be challenging as the platforms are filled with fake reviews. “Fake reviews are becoming an increasing problem. Online platforms like Amazon, TripAdvisor, Zomato or any other review platform where you have a lot of volume of reviews coming, and that’s where we deal with this problem of fake reviews,” said Vinodhini Ranganathan of Cisco. The impact of reviews is vast. This can be elaborated by the fact that 68% of millennials always go through reviews before they make a purchase. According to Vinodhini, consumer reviews are trusted significantly more than any other reviews, and that’s what makes this area worthy of more research. According to her, negative reviews create a bust, positive reviews boost sales. “There’s a lot of sentiment that goes around when there is a negative review. Let’s say the Maggi controversy when reports came out saying it had high levels of lead in it, leading to a huge dip in sales,” Vinodhini Ranganathan, Data Scientist at Cisco said. A single negative review can cost you, 30 customers. After reading three negative reviews, 59% of consumers will not buy. Four or more negative reviews about your company or product might take away 70% of the potential customers. Which is why it’s important to identify which reviews may be authentic or fake, particularly businesses that rely a lot on online reviews such as restaurants, hotels, medical\/healthcare, clothing store, grocery, etc. It has also been found that it’s not just the negative reviews which create an impact. The impact of a positive review is also there in the sense that for every star a business gets, there will be approximately a 5-9% increase in business revenue. Consumers are likely to spend 31% more on a business with excellent reviews. 72 percent of consumers say that positive reviews make them trust a local business more. There are many ways in which fake reviews exist on web platforms. These could be sock puppeting, where a single person can post several reviews, crowd turfing where online sellers hire people to post reviews; and review brushing where orders are created via identity theft to alternate delivery addresses, which then lead to spoofing fake reviews. While there are potential ways to use analytics and ML systems to identify fake reviews, according to Vinodhini, human learning is also key here. “We as individuals have a lot of challenges as it is not like any other task where we can machine learning. Usually, you have a forecasting problem, and you can identify it because the human learning component is needed to a great extent. But, in the case of identifying fake reviews, it is humans who are going to annotate these reviews as fake or not fake,” based on the different features which have been identified within data sets of reviews.” Feature Extraction Is The Challenge Vinodhini said that feature extraction is one of the major challenges when it comes to fake review detection as it entails finally annotating and labelling reviews by doing all of the feature engineering processes. Then, there is the automation part to it also. “While we have so many algorithms for that, the initial challenge is to annotate datasets we have as part of the feature engineering extraction process,” she said. There are different types of features here such reviewer centric features, review centric features, and network-centric features. Reviewer centric features include a number of reviews, number of helpful votes, the time interval between reviews, percentage of positive and negative reviews, the ratio of verified purchase, verified stay flag, rating deviation, review length, etc. Review centric includes looking for near-duplicate reviews posted from different IDs, or the same reviewer posting different reviews for different products, but the content may be same, or the spamming of the same reviews across different intervals. “Using NLP, experts can find the percentage of nouns, pronouns or adjectives as part of speech tagging. If you have a lot of pronouns or adjectives in your reviews, then it is more likely to be a fake review. Then there are things like lexical validity, lexical diversity, content diversity, syntactical diversity, usage of pictures\/links, emotiveness, sentiment score, product information matching etc,” told Vinodhini Ranganathan of Cisco. Moving on, there are network-centric features including IP address, GPS information, timestamp, traffic patterns, the sender IP neighbourhood density, device information. “Whenever there is a spam network, they are usually closely-knit during a given period of time coming from the same IP neighbourhoods,” told Vinodhini on this. What Is The Proposed Framework For Fake Review Detection? Sentiment analysis models should include the component of fake\/non-fake classification where possible classes include fake positive, fake negative, genuine positive, genuine negative. There is a lot of experimentation and research that can go when you combine this with a traditional sentiment analysis model. Vinodhini proposes a framework on this– “If you have unlabelled data, we are not going to clean the data how we do for sentiment analysis where we go remove numbers, punctuations, we should be looking at the data as is. The best and least normalisation we can do is converting everything into lower case and correct spell errors because every information in the review is very important. This will be data preprocessing. Then, it is followed by feature engineering, where all the list of features we spoke about and annotate the reviews datasets. After we have the labelled data, we are going to put it through classification algorithms for automation. This is similar to any sort of classification approach we follow. But, there will be a lot of skewed imbalanced data, so we can use a lot of sampling techniques to make the classes balanced, and then we can go and evaluate the model,” stated Vinodhini. According to Vinodhini Ranganathan of Cisco, fake or non-fake should be the first task in the pipeline for review text analysis whenever a sentiment analysis model is created. “We are talking about explainable AI, industry 4.0, we should also be talking about responsible AI. All of us are directly and indirectly responsible for the content that is getting posted online. As ML practitioners we also have to come up with ways to identify all of these because there is not much process, we have BERT, we have neural networks, and we should be also looking at this problem research and find out models which identify fake reviews,” said Vinodhini Ranganathan, Data Scientist at Cisco. Also Read: CAN ARTIFICIAL INTELLIGENCE REALLY FLAG FAKE NEWS? NEW RESEARCH SAYS NO","excerpt":"At MLDS 2020, Vinodhini Ranganathan, Data Scientist at Cisco highlighted the challenges of fake reviews along with an analytics framework for identifying and removing them on web platforms. “In God we trust, all others bring data,” W. Edwards Deming said a long while ago. But in this age, the sort of optimism is challenging, especially […]","categories":["AI Features"],"tags":["cisco","ecommerce","ecommerce analytics","Interviews and Discussions"],"author_name":"Vishal Chawla","publish_date":"2020-02-13T17:53:04","publication_year":"2020","word_count":1152,"keywords":["Go","ecommerce analytics","artificial intelligence","machine learning","AI","neural network","sentiment analysis","ML","cisco","NLP","ecommerce","analytics","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","neural network","NLP","analytics","sentiment analysis","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/vinodhini-ranganathan\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10098560,"title":"Google-Backed Anthropic Raises $100M from SKT to Fuel AI Innovation in Telecom","content":"Artificial intelligence startup Anthropic has secured $100 million in funding from SK Telecom (SKT), a major player in the South Korean telecommunications sector. This strategic investment marks a significant move in the rapidly evolving landscape of generative AI and has the potential to reshape the telecommunications industry. The partnership between Anthropic and SKT will revolve around the development of a multilingual large language model tailored for global telco firms, ushering in a new era of AI-powered innovation in the telecommunications domain. This funding comes on the heels of Anthropic’s Series C funding round in which it raised an impressive $450 million, led by Spark Capital. Notably, SKT was already involved in the Series C funding through its venture capital arm, SK Telecom Venture Capital (SKTVC). This strategic investment underscores SKT’s commitment to being at the forefront of AI advancements and leveraging Anthropic’s expertise to fuel transformative changes within the telecommunications sector. Anthropic’s co-founder and chief science officer, Jared Kaplan, will steer the customisation of the multilingual LLM and the subsequent product roadmap, solidifying Anthropic’s pivotal role in this partnership. Dario Amodei, co-founder and CEO of Anthropic, expressed enthusiasm for the collaboration, stating, “SKT has incredible ambitions to use AI to transform the telco industry. We’re excited to combine our AI expertise with SKT’s industry knowledge to build an LLM that is customized for telcos.” The resultant LLM, co-developed by SKT and Anthropic, is set to empower four members of the Global Telco AI Alliance—Deutsche Telekom, e&, Singtel, and SKT itself—to deliver AI-driven solutions tailored to their respective markets and users. This multilingual LLM is slated to support an array of languages, including English, Korean, German, Japanese, Arabic, and Spanish, effectively bridging language barriers and enabling more efficient and personalized AI-driven services. Steering Changes in Telecom SKT’s involvement in the Global Telco AI Alliance further solidifies its position as a key player in shaping the AI landscape within the telecommunications sector. By collaborating with Anthropic and harnessing its AI prowess, SKT aims to position itself at the forefront of AI-driven transformations in the industry. Ryu Young-sang, CEO of SKT, remarked, “By combining our Korean language-based LLM with Anthropic’s strong AI capabilities, we expect to create synergy and gain leadership in the AI ecosystem with our global telco partners.” Conclusively, SKT’s substantial investment in Anthropic exemplifies its dedication to staying ahead in the dynamic field of AI. This partnership is set to redefine the telecommunications industry’s trajectory by fostering AI-driven innovations tailored to diverse linguistic and market needs. The collaboration between Anthropic and SKT stands as a testament to the potential of AI to reshape industries and underscores the growing significance of tailored AI solutions in today’s technological landscape. Anthropic, founded in 2021, is the driving force behind Claude, an AI system designed to streamline various tasks within corporations, ranging from generating answers and automating workflows to coding and processing natural language. This innovation mirrors OpenAI’s ChatGPT, enhancing corporate efficiency through the seamless integration of AI technologies. Claude’s application in the telco industry will encompass specific areas such as customer service, marketing, sales, and interactive consumer applications, amplifying its utility and relevance. Anthropic also launched Claude 2 in July. Its predecessor was confined to enterprise usage, however, Claude 2 marks a transition to greater accessibility by extending its services to the general public in the United States and the United Kingdom. Unlike its forerunner, Claude 2 stands apart by adopting a dual approach for accessibility—through a beta website and an API. This strategic move not only broadens Anthropic’s user base but also establishes a clear distinction between the two iterations of Claude, further underlining Anthropic’s commitment to democratising AI.","excerpt":"This funding comes on the heels of Anthropic’s Series C funding round which raised $450 million, led by Spark Capital.","categories":["AI News"],"tags":[],"author_name":"Shyam Nandan Upadhyay","publish_date":"2023-08-14T17:36:52","publication_year":"2023","word_count":607,"keywords":["Anthropic","ChatGPT","artificial intelligence","OpenAI","AI","ML","RAG","Ray","Aim","generative AI"],"extracted_tech_keywords":["AI","artificial intelligence","ML","generative AI","ChatGPT","OpenAI","Anthropic","Aim","Ray","RAG"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/google-backed-anthropic-raises-100m-from-skt-to-fuel-ai-innovation-in-telecom\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10136739,"title":"AI to Play a Critical Role in Fulfilling Viksit Bharat 2047 Vision, says Screwvala","content":"India aspires to be a developed nation by 2047 with the country’s GDP growing from $3.4 trillion to $30 trillion by 2047. While speaking at Cypher 2024– India’s largest AI conference– hosted by AIM Media House–  Ronnie Screwvala co-founder upGrad & Swades Foundation, shared his thoughts on how AI can play an important role in achieving the Indian government’s Viksit Bharat @2047 vision. “If I were to sum it up, 1.3 billion people will play a vital role in shaping the nation we aspire to be by 2047. This vision is rooted in ambition and aspiration. This is where I believe the democratization of AI can make a significant impact—not just from a technological standpoint, but also in how we connect with others and inspire the younger generation. For me, AI is a powerful tool that can transform aspirations and ambitions, and its effects will be felt everywhere,” Screwvala said. Moreover, he further emphasized the importance of seeing AI as an augmentation tool. “At UpGrad, we don’t see AI as a threat; instead, we view it as a powerful tool for maximizing intellectual property creation. The content we create is as vital as the algorithms driving it. Instead of viewing AI as a threat, we should embrace it as a tool that enhances our capabilities.” In the insightful discussion with Srikanth Iyengar, CEO at upGrad Enterprise, Screwvala also adds that the best path to becoming one of the world’s top economies is to become the intellectual capital of the world. “Uplifting ourselves and setting benchmarks is crucial. While the government aims for us to be the AI capital, the current challenge remains job creation. It can seem contradictory to promote AI, as it may replace jobs, yet this vision is essential for conveying that we will create new opportunities,” he added. AI is a Slow and Steady Race He also believes that AI is a slow and steady race. He emphasizes that whenever a new technology arrives, it’s accompanied by hype, followed by an acknowledgement of the challenges it presents, much like what we saw with the World Wide Web. Over time, things stabilize, and true resilience emerges. “Today, I believe we’re in the early stages of AI adoption, especially in India. It’s crucial to understand that we can’t simply benchmark ourselves against the US, given the unique dynamics of our ecosystem, market, and workforce. Using AI as both a tool and a career path presents various opportunities and challenges. It’s essential to recognize that not all early startups will succeed. Maturity takes time, and sometimes being second is just as valuable as being first. History shows us that enduring technologies often prevail. As we navigate the rapid changes brought by AI, let’s remember to slow down and think strategically,” he said. Impact Investment in AI Screwvala, who has also made significant investments in many companies through his venture firm Unilazer Ventures, believes India is a country where impact and profit go hand in hand, creating immense opportunities for growth. “I’m excited to be actively involved in businesses that embody this combination. At this stage, if I were only focused on impact, I might consider a nonprofit approach, but I see that pursuing profit can amplify impact tenfold. When you limit yourself to low-profit ventures, you often rely on external funding and resources, which can be constraining,” he said. Taking India’s telecom sector as an example, Screwvala said that the country boasts some of the lowest data prices in the world. However, a decade ago, bandwidth limitations hindered business launches. “But we’ve broken that barrier. Today, top market-cap companies in India exemplify the balance between profitability and impact. I encourage everyone to pursue this dual approach, as it’s the key to building substantial value and sustainable profits.”","excerpt":"AI is a powerful tool that can transform India’s aspirations and ambitions, and its effects will be felt everywhere","categories":["AI Features"],"tags":["UpGrad"],"author_name":"Pritam Bordoloi","publish_date":"2024-09-27T10:42:22","publication_year":"2024","word_count":627,"keywords":["Go","API","funding","ELT","AI","RAG","Aim","ViT","UpGrad","R","startup"],"extracted_tech_keywords":["AI","Aim","RAG","R","Go","API","ELT","ViT","startup","funding"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/ai-to-play-a-critical-role-in-fulfilling-viksit-bharat-2047-vision-says-screwvala\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10005882,"title":"Hands-On Guide To Atspy For Automating The Time-Series Forecasting","content":"Time Series data is a type of data that changes over a period of time and can be represented over a timeline. Data related to stocks, depreciation of machinery, insurance premium, etc. can be considered as Time series data as it tends to change from time to time. Time series is a part of our everyday life. Times series forecasting is a process where we try to forecast\/predict what will be the future values of the time series data by studying the historical data. There are a number of models that are used to predict the time series data some of them are ARIMA, GARCH, Prophet, etc. All these models have a certain definition and the parameters are different for them. Atspy provides a way of automating the process of Time Series Forecasting and that too in just a single line of code. Atspy contains a variety of Models such as HWAAS, HWAMS, ARIMA, Prophet, etc. which can be used to predict time series data and we can also compare the results of these models. In this article, we will explore how we can use Atspy for automating time series forecasting and compare different models which are there in Atspy. Implementation: We will start by installing atspy using pip install atspy Importing Required Libraries We will be loading the data using pandas so we need to import pandas and we will import different functions from atspy as and when required. import pandas as pd Loading the dataset For using atspy we need data that is strictly preprocessed i.e. there should be no missing data and no period should be missed. I have created a fake dataset which contains the closing sales of a company with the respective month. Let us load the dataset and convert it to set the date column as the index. We need to make sure that our data contains only 1 target column as supported by atspy. import pandas as pd df = pd.read_csv(\"atspy.csv\") df.Date = pd.to_datetime(df.Date)     #changing datatype to datetime df = df.set_index(\"Date\")     #Setting date as index df.head() Selecting Models to be used In this step, we will import the AutomatedModel() function and define a list of Models that we will feed our data to with their names. from atspy import AutomatedModel model_list = [\"ARIMA\", \"Prophet\"] mdl = AutomatedModel(df = df , model_list=model_list, forecast_len=25 ) Insample Model and Prediction Now we will call the forecaste_insample function which returns the forecast on the insample data and the performance. forecast_in, performance = mdl.forecast_insample() Here you can see that itself splits the training and testing data, after that it runs all the models we mentioned in the list defined above for training and forecasting. Now let us print the forecasting and the performance. forecast_in performance Outsample Model and Prediction Similar to insample prediction now we will use outsample prediction which returns the out-of-sample forecasted dataframe. forecast_out = mdl.forecast_outsample() forecast_out performance Here we created a model using ARIMA and Prophet. We can use a lot more models that are defined in atspy and compare the performance of these models. Conclusion: In this article, we saw how easily we can create models using atspy and compare different models. Atspy is simple and creates models in just a single line of code, which is really timesaving and efficient in terms of prediction. We can create different model predictions in a single line by defining the model list. Currently, Atspy supports around 8 to 10-time series prediction models.","excerpt":"Atspy provides a way of automating the process of Time Series Forecasting and that too in just a single line of code. Atspy contains a variety of Models such as HWAAS, HWAMS, ARIMA, Prophet, etc.","categories":["Deep Tech"],"tags":["bitcoin stock","Time Series","Time Series Analysis","Time Series Forecasting","time series prediction"],"author_name":"Himanshu Sharma","publish_date":"2020-09-01T15:00:00","publication_year":"2020","word_count":576,"keywords":["Time Series Analysis","data_tools:Pandas","programming_languages:R","AI","Time Series Forecasting","bitcoin stock","Time Series","time series prediction","R","Pandas"],"extracted_tech_keywords":["AI","Pandas","R","programming_languages:R","data_tools:Pandas"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/hands-on-guide-to-atspy-for-automating-the-time-series-forecasting\/","complexity_score":2,"technical_depth":5,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10052377,"title":"Council Post: How Start-Ups Can Build &#038; Grow A Data Science Team","content":"With the world becoming smarter and more data inclined, it has become essential for start-ups to leverage data. Data not only gives them a competitive edge over other companies but is also the key to success in the modern day. However, creating an efficient data science team in the start-up ecosystem needs consideration of several factors such as the company’s funding stage, growth stage or the extent to which their product or service is based on machine learning. In this article, we analyse how start-ups at different stages can build their Data Science teams. We also look at some of the common pitfalls start-ups face when building them. Growth Stage Vs Use Of ML Two factors play a significant role in determining the steps needed to build a Data Science team in a start-up. The first aspect is the maturity of the organisation. A start-up’s maturity can be categorised based on factors such as revenue, user base, or funding. These factors help differentiate between an early-stage start-up and a mature one. The second aspect is the extent to which artificial intelligence is at the core of their products. Almost all tech companies are data-centric today, with most start-ups leveraging data for day to day decision making. AI-centric companies are organisations where AI is the key differentiator in their product or service offering. Source: Author This matrix describes the four possible combinations between the maturity of the organisation and the extent to which AI is at the core of its product\/service. The X-Axis maps the maturity of the organisation while the Y-Axis maps its AI centricity. AI-Centric + Early Stage AI-centric start-ups are organisations where AI is at the core of their product and service, differentiating them from competitors. AI-centric companies include (but are not limited to) enterprises offering NLP powered chatbots, image detection, SaaS, smart assistance, etc. In AI-centric start-ups, data science has to be one of the first teams that are established. Since data-powered IPs are at the centre of the company’s offerings, it is integral to onboard a Principal Data Scientist or a Head of Data Science in the early stages. This must be someone with extensive experience and knowledge in the Data Science field. Given their expertise, they can leverage their experience to solve complex AI & ML problems. Given the start-up’s limited bandwidth, the Head of Data Science can build a small but experienced team. This means onboarding experts with in-depth knowledge and higher work experience into the team. For instance, in a start-up providing NLP products, roles such as NLP architect\/NLP engineer should be a top priority. The idea for the DS team at this stage is to stay focused and say ‘No’ more than ‘Yes’ to different demands from stakeholders. Core algorithms should be the focus of this team. Only after these algorithms are well built and successfully tested should the organisation consider scaling up. Analytics or ML for other initiatives such as SEO or marketing growth-related can be executed later. At the same time, a small ML Ops team is essential to support the Data Science team and sustainably test the algorithms. Hence, building the ML Ops team responsible for scalable model deployments should begin from an early stage. AI-Centric Companies  + Mature Stage Once a start-up with a core ML offering starts generating more revenue, the focus can shift to expanding the data science team. The small initial team of experts should grow to include more analytics and leadership roles. Depending on the products offered by the company, one should hire more data scientists or ML engineers to develop complex analytical solutions, data engineers to work with data integration and distributed computing, data architects to oversee cloud services and provide system-wide insights and data analysts to perform reporting and BI. Hiring for leadership roles like chief data officers (CDOs) or chief analytics officers (CAOs) is also important. They will play a crucial role in directing teams efficiently. Non-AI-Centric (but Data-Driven) + Early Stage Organisations that don’t have AI at the core of their product don’t need data science in their initial stages. Since their products are not in need of AI-based solutions immediately, the companies should initially focus on establishing a good data foundation. This involves setting up the right data instrumentation and plumbing the data pipelines to create a base for AI in the future. For instance, the initial focus for an eCommerce company revolves around operational efficiencies. They could set up a data analytics team at the start but can function without an ML team in the initial months. Once the company starts growing, leaders can consider establishing a data science function to leverage ML for exploring revenue opportunities. Non-AI-Centric (but Data-Driven) + Mature Stage The correct time to build a data science function for an organisation that is not built around AI is when it achieves a certain level of maturity. At this stage, organisations can start with building a core data science team consisting of Data Scientists, Data Engineers, and Data Analysts. An organisation would have already built a centralised data management system with well oiled dashboarding and reporting systems to consolidate the organisation’s data for the team to base analytical insights and solutions upon. These organisations can now shift their focus to building state-of-the-art data science models to start bringing in data to drive strategic and long term decisions. Data scientists could initially start with naive heuristics and simple models to understand the evolving problem space. Once they have gauged the nature of the problem at hand, they can proceed with more complex models. Non-tech based organisations can combine in-house IT specialists with external consultants to efficiently develop a data warehouse. Furthermore, this data should be democratised to empower any team member to access the data at any given time and generate business insights. This article is written by a member of the AIM Leaders Council. AIM Leaders Council is an invitation-only forum of senior executives in the Data Science and Analytics industry. To check if you are elgible for a membership, please fill the form here.","excerpt":"With the world becoming smarter and more data inclined, it has become essential for start-ups to leverage data. Data not only gives them a competitive edge over other companies but is also the key to success in the modern day. However, creating an efficient data science team in the start-up ecosystem needs consideration of several […]","categories":["AI Features"],"tags":["Startups"],"author_name":"Mathangi Sri","publish_date":"2021-10-26T17:00:00","publication_year":"2021","word_count":1010,"keywords":["data science","machine learning","artificial intelligence","AI","chatbots","ML","RAG","NLP","Aim","analytics","Startups"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","NLP","data science","analytics","Aim","RAG","chatbots"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/council-post-how-start-ups-can-build-grow-a-data-science-team\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":57943,"title":"7 Key Tenets Of AI Ethics To Follow","content":"Considering the vast amount of data present, various insights into the data have led to data science professionals to get information about human behaviour, and that’s where the question of ethics arises. As data science’s focal point is analysing the data generated by humans, how data scientists use it should come under certain ethical conditions. Today’s era offers a great deal of ease when it comes to analytics, which completely changes the ethical framework. The scalability of the data also raises several ethical issues, especially with the companies who use data for monetising it. With the existing ethical frameworks, it is not clear about what to follow, but some basic principles should always be addressed when looking at data science ethics: The Autonomous Paradigm With advancements like robotics and automation, the gap between humans and artificial intelligence is closing. Over time, these sophisticated algorithms which use data might soon be able to overtake human decision making. Just to put in perspective, for example — an autonomous vehicle is supposed to make the lives of human beings safer by reducing traffic and accidents. But, what will happen in an unlikely event where an accident takes place involving autonomous vehicles? Who will be held accountable for the machine’s autonomous decision making in an emergency? Here, data science personnel may be involved in building the model to steer the machine, leaving out the implementation part. And, therefore, closing the gap between human beings and artificial intelligence raises many several ethical questions. The Age-Old Question Of Privacy Perhaps, data privacy concerns and the queries around it have been given the most attention when it comes to ethics and data science. Nothing in today’s world receives more concern than privacy. The digital transformation has changed the way we interact with it. The interactions we have online whether it be social media or something related to other aspects of life are increasingly defining what we think, what we do and most importantly, who we are. In this case, the companies which gather this data have to be responsible for its usage. With data being the new currency in the world, an organisation which works with data has to be able to answer questions related to ethics like who owns the data? To what extent will the personal data be used? Who will be controlling the data? To what extent should the data controllers be held responsible for the misuse or loss of the data? “Our information is being weaponised against us with military efficiency. Every day, billions of dollars change hands, and countless decisions are made based on our likes and dislikes, our friends and families, our relationships and conversations, our wishes and fears, our hopes and dreams. These scraps of data, each one harmless enough on its own, are carefully assembled, synthesised, traded and sold.” – Tim Cook One such scary example where the data privacy by far is the prime concern is healthcare. The public trust took a significant hit when the news broke out that Google’s Nightingale had access to healthcare information, including names and other data of millions of people without their knowledge. Google did release a statement saying that they were well within the regulations. Still, news like this isn’t something that can be let go so smoothly, especially when it is the health-related data of so many people, which will give the understanding of the person at an intimate level. Micro-Targeting The data analytics can be used to focus on the pattern behaviours like beliefs, values and attitudes of a mass on a larger scale. Micro-targeting has the potential to influence a certain mass in the fields of marketing, politics and economics. What is tricky about micro-targeting is that it can be used to influence an individual’s free choice. The practice of feeding consumers only information they agree with or showing specific information targeting specific groups of people can lead to manipulation. The real danger comes when data science is used less to improve an organisation’ organisation’s services and more towards making the client a product and an object of manipulation. The most significant example when it comes to this is the Cambridge Analytica using Facebook data for targeting a specific group of the public with the entire purpose of influencing the US elections in 2016. Later, Facebook disagreed with sharing the data with third parties. A data scientist’s role should be at least to warn of the possible misuse. The person should know and should be communicated with the dangers of micro-targeting for society. Discrimination and Bias The bias issue is one where the data science knowledge is perhaps very important for understanding the problem. While talking about algorithms like pattern recognition and similar kinds, they seem to show a certain prejudice. The goal of an algorithm is to predict the future behaviours of a person or an object according to previously observed behaviour data. Then measure characteristics of this person according to others’ characteristics in the group. In most cases, there is more negative impact rather than a possible benefit — for example, COMPAS (Correctional Offender Management Profiling for Alternative Sanctions). COMPAS, a software used in the US judicial system to classify the defendant’s risk of committing more crimes showed that it gave better chances to the white coloured group of people for not committing a crime than the darker-skinned group. The implication of such algorithmic biases leads to danger to people’s lives. Technically, there is no clear solution to this type of problem. Monitoring the type of data that is being fed to the algorithm is one way, but how the system uses this data is much of a black box. This can only be prevented by morality and law. Distributed Ledgers and Encryption There is declining trust between consumers and the intermediaries who capture, collect and monetise their data. With the advent of the internet, the distrust is growing even more. Blockchain and data encryption in machine learning offer some support to make the data more private, traceable and transparent. Blockchain makes the data more traceable and transparent in particular. Others like Julia using Homomorphic Encryption, promise more data security for organisations. The only problem is, these are relatively new technologies and questions whether these might be trusted arise in one’s mind, or their scalability might be questioned. “There’s another attack on our civil liberties that we see heating up every day — it’s the battle over encryption. Some in Washington are hoping to undermine the ability of ordinary citizens to encrypt their data.” – Tim Cook Data Curation When it comes to data curation, the data scientists must assume the potential biases and abnormalities in the algorithms. Much of the algorithms can’t be improved because of the black-box nature. These algorithms either have to be taken offline or completely replaced. One such example is Tay, Microsoft twitter bot. Tay was implemented on Twitter as a regular user with an intention for the bot to perform normal user functions on Twitter. But soon, the algorithm learned in such a way that in less than a day, it started spurting out fascistic and racist paroles. The biggest reason was that people on Twitter had learned to manipulate it that way, and there was no curation. Ethical Review and Adhering to Legal Frameworks The products and research should be subjected to internal and external ethical review. The organisations should prioritise establishing an actionable ethical review for new practices and services and research. Also, the data scientists should adhere to legal frameworks like GDPR (Europe), The Personal Data Protection Act, 2018 (India) and The California Consumer Privacy Act of 2018 (California) specifically entail the rights of citizens and address the dangers of the commercial use of their data. These frameworks try to balance power and influence between the private organisations and individuals. These frameworks give ethical benchmarks like the right to object, the right to access, the right to be forgotten and the right to rectification. Importance of such legislations extend beyond data privacy; this Government legislation can be taken to accurately define and form data practices to respond to many ethical questions.","excerpt":"Considering the vast amount of data present, various insights into the data have led to data science professionals to get information about human behaviour, and that’s where the question of ethics arises. As data science’s focal point is analysing the data generated by humans, how data scientists use it should come under certain ethical conditions. […]","categories":["AI Trends"],"tags":["AI &amp; ethics","Data Privacy"],"author_name":"Sameer Balaganur","publish_date":"2020-03-04T13:00:00","publication_year":"2020","word_count":1354,"keywords":["data science","Go","artificial intelligence","machine learning","AI","AI &amp; ethics","ML","homomorphic encryption","Data Privacy","analytics","Rust","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","data science","analytics","homomorphic encryption","R","Go","Rust"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/7-key-tenets-of-ai-ethics-to-follow\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":43098,"title":"How BIGO LIVE Is Using AI\/ML To Customise &#038; Showcase User-Generated Video Content","content":"With the advent of affordable internet and smartphones, video content streaming has witnessed a tremendous rise in the past couple of years. Today, the millennials and Gen Zs believe in binge-watch their favourite shows on digital platforms at a time of their own choosing. Thus, giving rise to video content platforms. Talking about the Indian scenario, the nation is witnessing its own over the top streaming or OTT online video content. What started with a handful of mini web-series, now have their original content across several genres. With so much happening, it’s not just OTT that is reaping the benefits, live streaming content and platforms are also coming into the picture. And with time, they are going big, empowering micro-entrepreneurs and entertainers across creative fields. Started in 2014 in Singapore, BIGO is one such company that is playing a vital role in the live streaming content ecosystem. And in order to understand the market, the technology behind better, we got in touch with Nagesh Banga, Deputy Country Manager at BIGO LIVE. The Transformation BIGO’s first product was BIGO LIVE, and in 2016, it was ranked Number #1 in the App Store in Thailand. In 2017, the company’s other product LIKEE (formerly LIKE Video) was also awarded “Best App” on Google Play. Talking about the Indian market, BIGO LIVE started with on-boarding of entertainers, dancers and singers, and now the platform also has projects like PUGC (Professional User-generated content), MC Rappers (Platform for underground rappers) and Education learning. “Our strategy to keep on adding more content categories and to improvise content makes sure we act as a one-stop-shop and act as a daily dose of entertainment for all,” Banga added. “This clearly shows there is a lot happening in this space and has tremendous scope for further content creation.” Today, BIGO operates its services in more than 150 countries and regions in the world. It has also established more than 20 offices globally expand our business. Talking about the user base, BIGO Technology as a group has over 400 million monthly active users globally. The best performance would be the video communication app called imo, which claims 211 million MAU around the world. The combined MAU of both BIGO LIVE (which is a live streaming App) and LIKEE (which is a short video making App) is 78.9 Million globally, which increased with 160% compared with the same period last year. The Role Of AI & ML With a global team of around 3,000 people, and with a significant number of them working on the engineering side which focuses on R&D, AI is the foundation for BIGO products and services. While walking us through BIGO’s tech stack, Banga also said that BIGO has one of the world’s strongest AI Technology to cater to its moderation of content on BIGO LIVE. Whether its Image Recognition, Facial Recognition, Video Intelligence, or Voice Processing, AI is at the very core and forms the roots of BIGO. In fact, the company claims that its AI is also able to detect negative content and issue a ban within 60 seconds. “The application of AI brings us a creative and healthy internet environment for Indian users,” said Banga. When it comes to machine learning, the platform uses this sought after tech for recommendations. It learns what it’s users like and features with personalised video content. Being a company that focuses on technology, especially on Artificial Intelligence, BIGO has established more R&D centres in headquarter Singapore, US, MENA region, India and China. And these R&D centres give the company outweigh the advantages of fierce competition among global AI players. Promoting AI Globally BIGO is always looking out for the best talent around the world. But finding the right talent is a real challenge as it believes that AI is a vast topic and subject, and the number of graduates are also quite a few. Therefore, in order to promote AI as a subject among budding talents, BIGO support universities around the world through scholarships. The company has recently announced a scholarship fund with Singapore’s Nanyang Technological University valued with SGD 500,000. In China, BIGO collaborated with renowned university called Shanghai Jiaotong University in collaboration with visual and voice processing program. In India, BIGO is open to working with universities in a similar capacity, which is something that will be very excited to achieve in the future. The Roadmap Ahead Looking into the future, BIGO’s vision is to become a connecting bridge for the internet industry and to equalise the technological disparities in developing areas. And in terms of the Indian market, the company is aiming to become a platform serving all kinds of content to users in India, and for this, BIGO is already developing an MCN network. “We plan to entrench ourselves deeper into the Indian market through localisation of our business; having said this, we understand India to have diverse cultures and languages. Our plans to localise will be targeted at the local level so that we will always be relevant to our users,” Banga concluded.","excerpt":"With the advent of affordable internet and smartphones, video content streaming has witnessed a tremendous rise in the past couple of years. Today, the millennials and Gen Zs believe in binge-watch their favourite shows on digital platforms at a time of their own choosing. Thus, giving rise to video content platforms. Talking about the Indian […]","categories":["Deep Tech"],"tags":["OTT","Startups"],"author_name":"Harshajit Sarmah","publish_date":"2019-07-22T13:00:33","publication_year":"2019","word_count":839,"keywords":["Go","API","machine learning","artificial intelligence","AI","ML","image recognition","Git","OTT","Aim","Startups","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","Aim","image recognition","R","Go","Git","API"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/how-bigo-live-is-using-ai-ml-to-customise-showcase-user-generated-video-content\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":69676,"title":"What Are The Advantages Of DevOps On The Cloud?","content":"In the era of business transformation, enterprise cloud is very important. But after migrating to the cloud, how to manage the cloud computing infrastructure and achieve the delivery of cloud-native architecture. From the perspective of technical staff, the factors of cloud include price, infrastructure startup and running time, and application scalability. At first, you may have concerns about putting sensitive information on the cloud. Plus, there could be technical problems that teams may have to deal with. The preparation for launching new cloud infrastructure needs carrying out capacity planning for the cloud; this is a very complex process. And when the capacity planning is completed, make a purchase plan, layout the network, etc., install software and hardware, and prepare the corresponding application building and deployment tools. There are other complexities between the test operation and the environment because of the inconsistency of the environment. For example, the developers may have to find that software has been developed, and the tester finds that it is not running well in the environment. To get past the complexities, cloud DevOps can be of much importance. DevOps And Cloud Infrastructure Automation The DevOps tool on the cloud provides the entire link covering DevOps right from the submission of code to the entire life process of application release, we can have a flexible choice of open-source tools and professional products. For capacity planning, because we can quickly create resources in minutes on the cloud, you do not need to think too much about how to expand capacity. Users can define the infrastructure as code with declarative configuration files which can be applied to create, control and update infrastructure resources, such as virtual machines (VMs), networks, and containers. Companies can deploy the Terraform configuration language to efficiently automate resource management over their workflow. The large majority of cloud development projects use DevOps, as benefits to utilising DevOps with cloud projects are also growing to be better defined and more elaborate. The centralised nature of cloud computing provides DevOps automation with a standard and centralised platform for processes like testing, deployment, and production. Previously, the distributed nature of some enterprise systems did not work well with centralised software deployment, but using a cloud platform eliminates many problems with distributed complexity. Most public and private cloud computing platforms maintain DevOps on their platform, including continuous integration and continuous development tools. Cloud-based DevOps reduces the requirement to account for resources leveraged. Cloud leverages usage-based accounting, which tracks the usage of resources by application, developer, user, data, etc. Features of IaaS Tools For DevOps For Cloud There are many open-source tools and proprietary products to achieve DevOps on the cloud such as Terraform from HashiCorp as an example to manage multi-cloud infrastructure and application automation management solutions. Such tools and services form the basis to achieve the rapid delivery and maximum benefit of DevOps. A full-stack cloud platform enables a developer to sign up, connect their version control — GitHub, Bitbucket or GitLab — and quickly build and deploy all of their repositories through a container file. Using the popular tools, you can build Infrastructure as a Service (IaaS) layer to build infrastructure as code for creating virtual private cloud (VPC) clusters that decide which machines to include, what kind of network to configure, and design security rules, which require humans to operate in the traditional model and then execute them through container tools. For example, Terraform is revolutionising DevOps by transforming the processes with which infrastructure is managed and making it quicker and more effective to administer DevOps projects. This means that your entire Cloud infrastructure can be automated in Terraform. One of these is the way Terraform handles failure. When provisioning fails, Terraform records the suspect resource and removes and re-provisions them at the following execution. Terraform is platform-agnostic, and organisations can utilise it to maintain bare metal servers or cloud servers like Google Cloud Platform, AWS, OpenStack, and Azure. Terraform is used to manage infrastructure on multiple cloud platforms and does so based on configuration files that manage the production, changing, and destruction of all resources they include. It frees users from having to do this manually, which would be error-prone and usually irreproducible. Utilised as part of a multi-team DevOps process, Terraform also supports teams such as operations and security to run in parallel with developers. Each component in the DevOps process has a specially designed tool, which means organisations can focus on their critical tasks without hindering other teams working on the projects. This transforms the DevOps process as teams can work in parallel. There is no need to deal with various disputes between development and test operations due to environmental consistency. If you use a similar SaaS service such as AWS CodePipeline to design and develop new applications on PaaS using a DevOps approach for rapid rollouts. Here, you do not need to prepare application construction and deployment tools. Teams can easily derive a new environment, and quickly create these resources in minutes. In this process, they can use a tool like AWS CodePipeline for CI\/CD, and for the running environment can use container services. Companies can also choose the PaaS platform and the container service to build applications. So when the platform is ready, the code is quickly released to the cloud. At this time, the product CodePipeline can directly submit the code, then automatically build, and deploy it to ContainerService like ECS according to instructions. This CodePipeline and tools support the complete process of DevOps. When we deploy to ECS, Terraform can help generate the environment quickly. Overview Adopting agile development is important to pursue the deliverable value and rapid development. DevOps aims to break the barriers between development and operation. DevOps is able to accelerate all the processes and methods of receiving the final application and service delivery from the demand. DevOps adopt an agile method, always pursue a goal, and then establish a rapid monitoring and feedback system to improve the quality of software delivery, establish a communication environment of communication and trust between various departments of the enterprise, and take the fastest delivery service application.","excerpt":"In the era of business transformation, enterprise cloud is very important. But after migrating to the cloud, how to manage the cloud computing infrastructure and achieve the delivery of cloud-native architecture. From the perspective of technical staff, the factors of cloud include price, infrastructure startup and running time, and application scalability. At first, you may […]","categories":["AI Features"],"tags":["AWS cloud","Cloud Computing","importance of automation in saas","importance of automation saas","software automation testing"],"author_name":"Vishal Chawla","publish_date":"2020-07-14T10:00:00","publication_year":"2020","word_count":1016,"keywords":["Go","AWS","AI","cloud computing","Azure","R","Scala","RAG","software automation testing","importance of automation saas","Cloud Computing","Aim","Rust","AWS cloud","importance of automation in saas"],"extracted_tech_keywords":["AI","Aim","RAG","cloud computing","AWS","Azure","R","Go","Rust","Scala"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/what-are-the-advantages-of-devops-on-the-cloud\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":59451,"title":"As Coronavirus Grips The World, GoJek Launches Critical Initiatives","content":"As effects of the coronavirus pandemic continue to ripple across the globe, ride-hailing giant Gojek defies associated turbulence in the tech industry by raising $1.2 billion to expand its operations. But even as it revels in what has come to be one of the largest deals since the emergence of coronavirus, the Southeast Asian company has been continuing its efforts to prevent the spread of the viral disease. Gojek is emphasizing three key areas in its fight against coronavirus: Social distancingPractising a healthy lifestyle, andMaintaining productivity The company has collaborated with various stakeholders and launched a number of initiatives to this end. Contactless Food Delivery To support the ongoing social distancing movement, GoJek is providing the option of minimizing direct contact between driver partners and customers during food deliveries. While all payments will be made cashless, customers can also write an additional note to the driver to drop off their meal at a designated location. What is more, the company is reportedly spraying disinfectants on delivery bags as well as cooking stations on a regular basis. The company is also partnering with others for items such as hand sanitizers and masks for their driver partners. Income assistance scheme GoJek has begun providing financial assistance to its driver partners who have tested positive for the disease. In addition to providing income support, it will also temporarily stop installments, including vehicle payments, until they return to work. What is more, the company has also deactivated the accounts of drivers who are currently under observation. There is also a team that operates 24\/7 that helps connect partners with health authorities if they are experiencing symptoms related to Covid-19. ALSO READ: How Alibaba Is Leading The War Against Coronavirus Outbreak Educate Customers About Coronavirus Using banners on apps, GoJek has been launching campaigns to create awareness about the disease. According to The Jakarta Post, information on precautionary measures is conveyed in a timely manner through its in-app notification feature. Free Consultation Service GoJek app has a free consultation service for drivers in its network who are affected by coronavirus. Through a collaboration between GoMed and Halodoc, this service allows them to consult a certified medical professional from anywhere. Moreover, they get their medication delivered to them within a few hours. Comprehensive Work-From-Home Guidelines While the company has been running deep-cleaning services in its office regularly, it has shared a comprehensive guideline on how to adjust to this transition and how to work remotely in a manner that is productive. This includes setting up the right environment, ensuring you have the right tools, as well as some useful hacks.","excerpt":"As effects of the coronavirus pandemic continue to ripple across the globe, ride-hailing giant Gojek defies associated turbulence in the tech industry by raising $1.2 billion to expand its operations. But even as it revels in what has come to be one of the largest deals since the emergence of coronavirus, the Southeast Asian company […]","categories":["Deep Tech"],"tags":["Coronavirus","covid-19","gojek"],"author_name":"Anu Thomas","publish_date":"2020-03-23T11:33:54","publication_year":"2020","word_count":434,"keywords":["Go","covid-19","AI","programming_languages:R","gojek","programming_languages:Go","Coronavirus","Ray","ViT","R"],"extracted_tech_keywords":["AI","Ray","R","Go","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/as-coronavirus-grips-the-world-gojek-launches-critical-initiatives\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":42090,"title":"How Organisations Can Build An Effective Incident Response Framework","content":"As cyber-attacks continue to grow each year. Organisations across the world have come to the realisation that working to prevent and detect cyber-attacks is one thing, but having an incident response framework is also equally imperative. However, not every organisation knows what’s the best approach built that effective incident response framework. Why Is An Incident Response Framework Important? An incident response framework is basically a plan to deal with the aftermath of a cyber-attack. When a company’s cybersecurity defensive wall fails to prevent cyber-attack, incident response system is the next most important thing that comes into the play, as it helps an organisation take steps to quickly contain, minimise, and learn from the damage. Every time a cyber-attack takes place, both the company and its consumers run into risk. It’s not just that phase — through breaches, black hats cost organisations millions of dollars and C-level executives their jobs. No matter how good your firm is; when there is a breach, the negative press is almost guaranteed for a significant amount of time. There are possibilities that under the pressure of a critical level incident, there won’t be any time to strategise your game. Therefore, having a strong incident response makes sure that things are sorted quickly and even the cost of getting things back to work is also not much. Response time is critical to minimising damages, and with every second counting, having a plan ready in place is the key to mitigate the loss and negativity forming across your business. An incident response system makes sure that it prioritises things according to the severity. For example, if there is just a login failure, the cybersecurity team cannot afford to spend hours investigating that event. Different events have different needs to be investigated. How To Build Your Own Incident Response Framework If you or your organisation want to have a strong incident responses framework you can either adopt one of them or you can pick up a few steps from one and few from another. And then compile them to have one framework for you. And if that doesn’t work, you can also have your own framework. However, there are instances that when the framework wouldn’t go well with compliance. Things to keep in mind when building a customised incident response framework: Make sure employees of every level get the required training and knowledge regarding all kind of threats. Make them aware and prevent them from falling prey to cyber baits. Have strict cybersecurity policies and protocols Make sure you assess properly all your tools and process, and fix or update everything accordingly Always go for vendors that are trustworthy and have a good record of past works Have a team of specialists who would take care of every aspect of cybersecurity A strong vulnerability management system\/team Emphasise on network security as well Prioritise events. Make sure the most severe event is dealt first Make sure the team who investigates the entire event, doesn’t leave a stone unturned. Always have a back up of every single data that is critical to you and your business Industry Standard Incident Response Frameworks When it comes to incident response frameworks, there are two standards that are extensively followed in the industry — NIST and SANS. They are the dominant institutes whose incident response steps have become the industry standard. NIST Established in 1901, the National Institute of Standards and Technology (NIST) is a part of the U.S. Department of Commerce. It is a government that works in all-things-technology, including cybersecurity, which has become one of their fortes. Talking about incident response, NIST’s process has four steps: Preparation Detection and Analysis Containment, Eradication, and Recovery Post-Incident Activity SANS SANS is a private organisation that, per their self-description is the most trusted and by far the largest source for information security training and security certification in the world. And their incident response framework is another industry standard. SANS’ process has six steps: Preparation Identification Containment Eradication Recovery Lessons Learned Outlook There was a time when organisations used to emphasise more on having a system that would stop any attack. However, those same organisations have realised that preventing every single threat wouldn’t be an easy task. So, it is better to be prepared for the worst. With the ever-increasing cyber-attacks and the risk of insider threat growing, the need for a strong incident response framework has reached a significantly high level. If your organisation has still not adopted an incident response framework, then there are chances that you would end up complicating things if you are ever hit by a cyber-attack.","excerpt":"As cyber-attacks continue to grow each year. Organisations across the world have come to the realisation that working to prevent and detect cyber-attacks is one thing, but having an incident response framework is also equally imperative. However, not every organisation knows what’s the best approach built that effective incident response framework. Why Is An Incident […]","categories":["AI Features"],"tags":["Cyber Security","Cyber security best practices","NIST"],"author_name":"Harshajit Sarmah","publish_date":"2019-07-08T15:29:15","publication_year":"2019","word_count":766,"keywords":["Go","Cyber Security","Rust","programming_languages:R","AI","programming_languages:Go","Cyber security best practices","ViT","NIST","GAN","R","programming_languages:Rust"],"extracted_tech_keywords":["AI","R","Go","Rust","GAN","ViT","programming_languages:R","programming_languages:Go","programming_languages:Rust"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/how-organisations-can-build-an-effective-incident-response-framework\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":58418,"title":"Go Beyond Upskilling In Data Science &#038; Machine Learning With MTech Program From Great Learning &#038; PES University","content":"The Big Data boom has birthed many flourishing opportunities for India Inc in 2020. Most businesses are now energetically adopting data science, machine learning and data analytics in their operations for improved business models and optimisation of their operations. Now, with data science and machine learning gaining more and more prominence across sectors, enterprises are now looking at filling the talent gap to keep up with the swift pace of the technological progress. The year 2019 saw over 97,000 job openings in data science alone. Now, it has been estimated that 2020 will see over 2 lakh job openings in the field of data science and allied areas in India. The Need for Higher Education In Data Science Data science is a relatively nascent field, and most of the mid-level management have graduated from traditional education — like software engineering backgrounds. Until recently, most of the universities didn’t even offer Master’s courses in data science, which, in turn, resulted in a shortage of skilled applicants for the jobs. To help professionals become data-proficient and build career critical competencies, Great Learning has collaborated with PES University to offer an MTech Program in Data Science and Machine Learning. The programme enables participants to gain an in-depth understanding of data science and machine learning techniques and tools that are widely used by companies. The programme takes a very practical approach to impart industry-relevant skills and enables participants to become job-ready. Importance Of A Degree Many Indian behemoths have a hierarchy system for appraisal as well as hiring, which is based on education levels. A Master’s degree can come in handy especially to help the applicant or the employee stand out from the crowd. Basically, a Master’s degree can help with the following: Fit into defined ‘salary brackets’ in bigger enterprises which give importance to a more formal or traditional form of educationSurpass hurdles that may delay an employee’s rate of review or appraisals in an organisationAutomatically differentiate oneself from the crowd of diploma and bachelor’s degree holders When applying for roles of some prominence, especially in sectors like IT, BFSI and manufacturing, conventional degrees like Bachelor’s, Master’s or Doctorates play a crucial role in hiring as well as appraisals. This is where a Master’s degree from a reputed institution like PES University plays to great advantage. About The Programme The full time and part-time (weekend classroom) 21-month M.Tech in Data Science and Machine Learning by Great Learning is a comprehensive program designed by accomplished data science academicians and professionals to give learners all the critical knowledge and skills that the industry demands. The course will also cover all the tools and skills sought by leading companies in Data Science. Through the duration of the course, candidates will be trained on Python, SQL, Tableau, Data Science and Machine Learning. Participants in the course will build their knowledge through classroom lectures by expert faculty and doing multiple challenging projects across various topics and applications in Data Science. The curriculum of this program consists of 4 modules spread over 21 months and covers the following integral topics of Data Science and Machine learning: Module 1:Statistical Foundations for Data SciencePython for Data ScienceDatabases – SQL & NoSQL Module 2:Data VisualizationMachine Learning – 1Machine Learning – 2 Module 3:Intro to Deep Learning and its applicationsIntro to Big Data Analytics Year 2:Industry Internship  (Full-Time) OR Capstone Project (Weekend-Classroom)M.Tech Thesis Faculty Career Outcomes Data science is a thriving field right now, and knowledge of this sector can open up numerous opportunities, especially after taking a comprehensive course like the MTech in Data Science and Machine Learning by Great Learning. After the program, one can jump to lucrative job profiles like data scientist, product manager, data analyst, data engineer and business intelligence developer, among others. Some of the companies that have recently participated in Great Learning’s hiring drives, include KPMG, Mahindra, Uber, IBM, Tata Capital, Swiggy and Truecaller, among many others. This programme will also equip the learner to: Be able to leverage the latest Data Science tools and techniques in a Data Science roleBe able to structure business problems in Machine Learning and Data Science frameworks Programme At A Glance Duration: 21 monthsCourse Type: Weekend classroom | Full-timeFees: ₹5,00,000 + GST (weekend) | ₹6,00,000 + GST (full-time)Financial Aid: AvailableCertification: M.Tech Degree from PES University Pre-requisites: – Candidates should have a B.Tech, BE or an MCA degree– Candidates should have a minimum of 60% in X, XII and Bachelor’s degreeNote: The program can be done even if the pre-requisites aren’t strictly met — as far as an individual is keen to learn","excerpt":"The Big Data boom has birthed many flourishing opportunities for India Inc in 2020. Most businesses are now energetically adopting data science, machine learning and data analytics in their operations for improved business models and optimisation of their operations. Now, with data science and machine learning gaining more and more prominence across sectors, enterprises are […]","categories":["AI Trends"],"tags":["big data certification","Courses","data analyst certification","data science salary","is the tech boom over","Learn Data Science","learning data science","Machine Learning"],"author_name":"Sejuti Das","publish_date":"2020-03-11T16:17:42","publication_year":"2020","word_count":761,"keywords":["data science","API","machine learning","big data certification","AI","learning data science","Machine Learning","RAG","Python","deep learning","analytics","data science salary","is the tech boom over","Learn Data Science","SQL","Courses","R","data analyst certification"],"extracted_tech_keywords":["AI","machine learning","deep learning","data science","analytics","RAG","Python","R","SQL","API"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/upskill-data-science-machine-learning-mtech-great-learning-pes-university\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10130623,"title":"AWS Launches GenAI Loft in Bengaluru to Advance Generative AI Innovation","content":"Amazon Web Services (AWS) has inaugurated the AWS GenAI Loft in Bengaluru as part of a global initiative to promote generative AI . This two-week pop-up, running from July 29 to August 9, 2024, is the first stop in a worldwide tour aimed at fostering collaboration and innovation in key AI hubs. The GenAI Loft offers developers, startups, and AI enthusiasts a platform to explore and engage with generative AI technologies. The space will feature 22+ sessions including workshops, hands-on training, and demos from AWS partners such as Shellkode, Datastax, Ganit, Cloudthat, Ankercloud, Rapyder, and GoML. Visitors will also have opportunities to network with AI experts and industry leaders. “We are excited to commence the global tour of AWS GenAI Loft in Bengaluru, a hub for AI innovation in the country,” said Kumara Raghavan, Head, Startups, AWS India and South Asia. “The GenAI Loft is designed to provide a dedicated space for builders and innovators to learn, ideate and gain hands-on experience in leveraging the transformative potential of generative AI. The GenAI Loft will also offer networking opportunities to connect with startups, investors, AWS experts and our partners,” he added Adam Seligman, Vice President of Developer Experience at AWS, emphasised the significance of in-person events in sparking new ideas. “I’m always inspired by the way in-person events spark new ideas when we bring together builders, startups, investors, AWS experts, and our partners. By creating these hubs of activity, we aim to accelerate the development of generative AI technologies and foster a truly global community of creators,” said Seligman. The GenAI Loft offers a variety of activities designed to provide deep insights into generative AI and its applications. Participants will gain practical experience with AWS’s suite of AI tools and technologies, including Amazon Bedrock and Amazon Q. Workshops will cover topics such as building full-stack web applications with generative AI capabilities and optimizing costs by migrating from OpenAI to Amazon Bedrock. The loft will also host sessions led by industry trailblazers and AI visionaries, providing attendees with the latest advancements in generative AI. Additionally, attendees can get one-on-one help from generative AI experts and AWS Solutions Architects, ensuring they receive personalized guidance for their projects. The loft will facilitate connections with leading AI investors, industry leaders, and innovators, fostering a collaborative environment for idea exchange and networking. The Bengaluru GenAI Loft is the first stop in a global tour that will include pop-ups in key AI hotspots such as San Francisco, São Paulo, London, and Paris. Each city will host the loft for up to 12 weeks, providing a platform for thousands of visitors to engage with cutting-edge generative AI projects and thought leaders. AWS’s launch of the GenAI Lofts is part of a larger $230 million commitment to accelerate the creation of generative AI applications by startups around the world.","excerpt":"The space will feature 22+ sessions including workshops, hands-on training, and demos from AWS partners such as Shellkode, Datastax, Ganit, Cloudthat, Ankercloud, Rapyder, and GoML.","categories":["AI News"],"tags":[],"author_name":"Siddharth Jindal","publish_date":"2024-07-29T16:53:48","publication_year":"2024","word_count":470,"keywords":["Go","GenAI","OpenAI","AI","AWS","ML","RAG","Aim","generative AI","R"],"extracted_tech_keywords":["AI","ML","generative AI","GenAI","OpenAI","Aim","RAG","AWS","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/aws-launches-genai-loft-in-bengaluru-to-advance-generative-ai-innovation\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":44767,"title":"H2O.ai Raises $72.5 Million In Series D Funding Led By Goldman Sachs","content":"Sri Satish Ambati, CEO Of H20.ai H2O.ai is a visionary Silicon Valley open source software company that redefined the AI movement. It is one of the leading open source data science and machine learning platforms used by nearly half of the Fortune 500 and trusted by over 18,000 organizations and hundreds of thousands of data scientists around the world. H2O.ai, has announced that it has secured $72.5 million in a Series D round, bringing the total funding to $147 million. The round was led by Goldman Sachs and the Ping An Global Voyager Fund with continued investments from Wells Fargo, NVIDIA and Nexus Venture Partners. Jade Mandel from Goldman Sachs will be joining the H2O.ai Board of Directors. Since its founding in 2012, H2O.ai has been on a mission to democratize AI for everyone. To that end, H2O.ai leads a thriving open AI movement of hundreds of thousands of data scientists in tens of thousands of organizations globally. The company also announced its latest release of H2O Driverless AI which empowers every company to become an AI company. With this new round of investment, the company plans to accelerate innovation and expand sales and marketing globally. The new investment will be used to further simplify AI for business users with new and ground-breaking technologies. “H2O.ai is democratizing AI and powering the imagination of every entrepreneur and business globally – we are making them the true AI superpowers,” said Sri Ambati, CEO and Founder at H2O.ai. “Our customers are unlocking discovery in every sphere and walk of life and challenging the dominance of technology giants. This will be fun.”","excerpt":"H2O.ai is a visionary Silicon Valley open source software company that redefined the AI movement. It is one of the leading open source data science and machine learning platforms used by nearly half of the Fortune 500 and trusted by over 18,000 organizations and hundreds of thousands of data scientists around the world. H2O.ai, has […]","categories":["AI News"],"tags":["Democratisation","Funding","goldman sachs","ML"],"author_name":"Ram Sagar","publish_date":"2019-08-20T18:27:12","publication_year":"2019","word_count":268,"keywords":["data science","Go","Funding","machine learning","funding","programming_languages:R","AI","innovation","ML","goldman sachs","Democratisation","Rust","GAN","R"],"extracted_tech_keywords":["AI","machine learning","data science","R","Go","Rust","GAN","innovation","funding","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/h2o-ai-raises-72-5-million-in-series-d-funding-led-by-goldman-sachs\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10133091,"title":"This YC-Backed Bengaluru AI Startup is Powering AWS, Microsoft, Databricks, and Moody’s with 5 Mn Monthly Evaluations","content":"Enterprises love to RAG, but not everyone is great at it. The much touted solution for hallucinations and bringing new information to LLM systems, is often difficult to maintain, and to evaluate if it is even getting the right answers. This is where Ragas comes into the picture. With over 6,000 stars on GitHub and an active Discord community over 1,300 members strong, Ragas was co-founded by Shahul ES and his college friend Jithin James. The YC-backed Bengaluru-based AI startup is building an open source stand for evaluation of RAG-based applications. Several engineering teams from companies such as Microsoft, IBM, Baidu, Cisco, AWS, Databricks, and Adobe, rely on Ragas offerings to make their pipeline pristine. Ragas already processes 20 million evaluations monthly for companies and is growing at 70% month over month. The team has partnered with various companies such as Llama Index and Langchain for providing their solutions and have been heavily appreciated by the community. But what makes them special? Not Everyone Can RAG The idea started when they were building LLM applications and noticed a glaring gap in the market: there was no effective way to evaluate the performance of these systems. “We realised there were no standardised evaluation methods for these systems. So, we put out a small open-source project, and the response was overwhelming,” Shahul explained while speaking with AIM. By mid-2023, Ragas had gained significant traction, even catching the attention of OpenAI, which featured their product during a DevDay event. The startup continued to iterate on their product, receiving positive feedback from major players in the tech industry. “We started getting more attention, and we applied to the Y Combinator (YC) Fall 2023 batch and got selected,” Shahul said, reflecting on their rapid growth. Ragas’s core offering is its open-source engine for automated evaluation of RAG systems, but the startup is also exploring additional features that cater specifically to enterprise needs. “We are focusing on bringing more automation into the evaluation process,” Shahul said. The goal is to save developers’ time by automating the boring parts of the job. That’s why enterprises use Ragas. As Ragas continues to evolve, Shahul emphasised the importance of their open-source strategy. “We want to build something that becomes the standard for evaluation in LLM applications. Our vision is that when someone thinks about evaluation, they think of Ragas.” While speaking with AIM in 2022, Kochi-based Shahul, who happens to be a Kaggle Grandmaster, revealed that he used to miss classes and spend his time Kaggeling. The Love for Developers “YC was a game-changer for us,” Shahul noted. “Being in San Francisco allowed us to learn from some of the best in the industry. We understood what it takes to build a successful product and the frameworks needed to scale.” Despite their global ambitions, Ragas remains deeply rooted in India. “Most of our hires are in Bengaluru,” Shahul said. “We have a strong network here and are committed to providing opportunities to high-quality engineers who may not have access to state-of-the-art projects.” “We have been working on AI and ML since college,” Shahul said. “After graduating, we worked in remote startups for three years, contributing to open source projects. In 2023, we decided to experiment and see if we could build something of our own. That’s when we quit our jobs and started Ragas.” Looking ahead, Ragas is planning to expand its product offerings to cover a broader range of LLM applications. “We’re very excited about our upcoming release, v0.2. It’s about expanding beyond RAG systems to include more complex applications, like agent tool-based calls,” Shahul shared. Shahul and his team are focused on building a solution that not only meets the current needs of developers and enterprises, but also anticipates the challenges of the future. “We are building something that developers love, and that’s our core philosophy,” Shahul concluded.","excerpt":"By mid-2023, Ragas gained significant traction, even catching the attention of OpenAI, which featured their product during a DevDay event.","categories":["AI Startups"],"tags":["AI Startups","Bengaluru"],"author_name":"Mohit Pandey","publish_date":"2024-08-20T10:10:40","publication_year":"2024","word_count":641,"keywords":["Go","OpenAI","AI","AWS","ML","RAG","LangChain","Aim","Bengaluru","R","AI Startups","Databricks"],"extracted_tech_keywords":["AI","ML","OpenAI","LangChain","Aim","RAG","AWS","Databricks","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-startups\/this-yc-backed-bengaluru-ai-startup-is-powering-aws-microsoft-databricks-and-moodys-with-5-mn-monthly-evaluations\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10064298,"title":"A hands-on guide to ridge regression for feature selection","content":"In machine learning, feature engineering is an important step that determines the level of importance of any features from the data. In one of our articles, we have seen that ridge regression is used to get rid of overfitting which can also be reduced by fitting the model with only important features. Ridge regression can also help us in feature selection to find out the important features required for modelling purposes. In this article, we are going to discuss ridge regression for feature selection. The major points to be discussed in the article are listed below. Table of contents What is ridge regression?Why ridge regression for feature selection?Implementing Ridge regression for feature selection Let’s begin with understanding the ridge regression. What is ridge regression? We can consider ridge regression as a way or method to estimate the coefficient of multiple regression models.  We mainly find the requirement of ridge regression where variables in data are highly correlated. We can also think of ridge regression as a possible solution to the imprecision of LSE(least square estimator) where the independent variables of any linear regression model are highly correlated. It converts LSE into a ridge regression estimator. Using this we can get a more precise estimate because it provides a smaller variance and the mean square estimator.  This regression is also called an L2 regularization that uses shrinkage of data values. Let’s see why we can use ridge regression for feature selection. Are you looking for a complete repository of Python libraries used in data science, check out here. Why ridge regression for feature selection? One of the most important things about ridge regression is that without wasting any information about predictions it tries to determine variables that have exactly zero effects. Ridge regression is popular because it uses regularization for making predictions and regularization is intended to resolve the problem of overfitting. We mainly find that overfitting is where the size of data is very large and ridge regression works by penalizing the coefficient of features and it also minimizes the errors in prediction. Talking about the other methods of feature importance they directly nullify the effect of less competent features. If we believe that some of the coefficients have zero effect using ridge regression we can build a better model by providing Bayesian priors for regression coefficients. For example, if we know that any of the coefficients have zero effect but which one has it we don’t know then we can use a prior with a ridge to know about the effect of every coefficient.  Let’s see how we can implement ridge for feature selection. Implementing Ridge regression for feature selection Importing libraries import numpy as np import matplotlib.pyplot as plt from sklearn.datasets import make_classification from sklearn.linear_model import LogisticRegression, Ridge Making data set X, y = make_classification(n_samples=3000, n_features=10, n_informative=5, random_state=1) X.shape, y.shape Output: In the above, we have made a classification data that has 10 features in it and 3000 values. Plotting some data plt.scatter(X[:, 0], X[:, 1], marker=\"o\", c=y, s=25, edgecolor=\"k\") Output: Here we can see the distribution of the data of the first and second variables. Let’s just split the data. from sklearn.model_selection import train_test_split X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.3,random_state=0) X_train.shape, X_test.shape, y_train.shape, y_test.shape Output: Scaling the data from sklearn.preprocessing import StandardScaler scaler = StandardScaler() scaler.fit(X_train, y_train) We can use ridge regression for feature selection while fitting the model. In this article, we are going to use logistic regression for model fitting and push the parameter penalty as L2 which basically means the penalty we use in ridge regression. ridge_logit =LogisticRegression(C=1, penalty='l2') ridge_logit.fit(X_train, y_train) Output: Let’s check the coefficient of features that will tell us how the features are important for the model. ridge_logit.coef_ Output: In the above, we can see that the feature from the data is one of the most important features and other features are not that much important. Let’s check how many features are having a coefficient of more than zero np.sum(ridge_logit.coef_ >= 0) Output: Let’s draw the plot for feature importance. importance = ridge_logit.coef_[0] plt.bar([x for x in range(len(importance))], importance) plt.show() Output: Here we can see how the features are correlated to the model predictions. Here in the above, we have seen how the ridge regression works for feature selection now after looking at the plots and results we can select the features from data to make prediction more accurate. As discussed above we can also see that it has considered all the features in the modelling of the logistic regression model which means it has the benefits of not dropping any of the features from the data set and in the results it gives a level of importance of every feature. Final words In this article, we have discussed ridge regression which is basically a feature regularization technique using which we can also get the levels of importance of the features. Not nullifying the effect of any feature from the data makes the ridge regression different from the other methods. References Link to the codes","excerpt":"One of the most important things about ridge regression is that without wasting any information about predictions it tries to determine variables that have exactly zero effects. Ridge regression is popular because it uses regularization for making predictions and regularization is intended to resolve the problem of overfitting.","categories":["Deep Tech"],"tags":["AI (Artificial Intelligence)","Data Science","Data Scientist","Deep Learning","feature engineering","feature extraction","feature selection","Machine Learning","Python","ridge regression"],"author_name":"Yugesh Verma","publish_date":"2022-04-05T13:00:00","publication_year":"2022","word_count":836,"keywords":["TPU","Git","R","data science","NumPy","Data Science","Go","machine learning","AI","ridge regression","feature engineering","Machine Learning","feature extraction","Matplotlib","Python","Deep Learning","Data Scientist","feature selection","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","machine learning","data science","NumPy","Matplotlib","TPU","Python","R","Go","Git"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/a-hands-on-guide-to-ridge-regression-for-feature-selection\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10073775,"title":"Contrary to Popular Belief, Hackathons Do Get Deployed for Real-World Problems","content":"Over 80% of the Fortune 100 companies conduct hackathons. More than a thousand hackathons are conducted annually globally, and private companies conduct 48.5% of them. Yet hackathon is an underutilised tool when it comes to corporate innovations. Born out of Silicon Valley startups, hackathons have become popular among corporations over the last years to explore ideas and crowd-source innovation. Countless examples show hackathons’ value and ability to generate substantial rewards. But the relationship between developers and hackathons is that of a love-hate one. Hackathons are considered time-consuming activities with no solid practical outputs. Often the term ‘Innovation theatre’ is associated with it. ‘Innovation theatre,’ a term coined by serial entrepreneur and professor Steve Blank, describes the idea of a company demonstrating innovations without putting much in the way of resources or substance behind it. However, contrary to popular belief, many of the winning solutions are deployed into production by the organiser. The developer community and domain businesses ensure that these hackathon ideas don’t end up homeless. One such community of data scientists is Kaggle, which has a reputation for being “fun, but impractical for real-world problems”. Kaggle broke the stereotype around hackathons with a winning solution deployed by the NFL during the 2019 NFL Big Data Bowl. The $150 billion organisation collects tracking data on all of its players. Internally, their datasets have coordinates for every player collected at a high frequency. The goal was to predict how many yards a rusher would get on a play. The data provided was the tracking data plus trajectories for every player when the ball was handed off to the rusher. The NFL cares about this piece of data they already use as one can quantify the skill of rushers based on how many yards above or below expectation they gain throughout the season. This allows them to significantly increase their production value and engage fans that want more depth. They can generate graphics like this and tell fans how impressive any particular play was in real time during the broadcast. The model was created by Philipp Singer and Dmitry Gordeev, which the NFL uses even today. However, this isn’t the only example of an extremely valuable winning solution. TD Ameritrade has products that come out of hackathons, such as Trade Finder, a tool that surfaces potential options trades based on client market expectations. Another instance was the hack found in 2020 during the INSPIRE hackathon to help farmers during locust swarms. INSPIRE hackathon is an annual event dedicated to solving global problems with open data, volunteered geographic information, and observations. The winning team used imagery from the European Space Agency’s Sentinel-1 and -2 satellites to map and study the locust swarm effects. After the hackathon, Kamau and her team published their report with the Global Open Data for Agriculture and Nutrition initiative. In 2016, Found Animals Foundation (FAF) and hackathon participant Yantra ended up innovating a solution that enables FAF to scale and streamline their work to save pets’ lives. More than 3 million animals are euthanised each year in the US. So, FAF turned to NetSuite to help it extend its microchip program and hand over 2 million microchips to shelters, rescues and other organisations annually. Tech giants such as Microsoft and Facebook have rolled out products and features generated from hackathons. Post the 2016 hackathon Facebook released a post explaining their favourite hacks of the event. The outputs included building on the data centre technology used in Facebook’s Wedge 100 top-of-rack (ToR) switch. The team transmitted ethernet traffic at 200GB over a single wavelength. The prototype eventually evolved into the Voyager transponder, contributing to the Telecom Infra Project this year. The other includes the Instant Verification feature, which uses people’s active Facebook sessions to see which phone numbers they’ve verified. Also, a group of Seattle engineers built a group video streaming feature within Facebook’s Messenger platform. This idea found its way into Messenger for third-party calls and Facebook’s virtual reality platform Oculus under the Rooms and Parties features. Some of Microsoft’s features emerged from the hackathon, including Seeing AI, an app that narrates the world to visually impaired users. Another addition was Learning Tools for OneNote, which was added as a OneNote add-on assistant for dyslexic students with reading challenges. Furthermore, Ability EyeGaze, developed initially for NFL star Steve Gleason, lets people interface with computers using eye movements. The project is now a part of Windows 10, enabling the usage of an on-screen mouse, keyboard, and text-to-speech experience by using only their eyes. It is the subject of a book encouraging engineers and developers to build the next wave of inclusive technology. In conclusion, hackathons serve as an important platform for solving real-life problems within a limited period. Moreover, it comes with the added perks of networking, being updated with market trends plus the experience looks great on the resume. Read here: MachineHack, a machine learning hackathon platform from Analytics India Magazine, has conducted multiple hackathons on numerous problem statements across industries.","excerpt":"Tech giants such as Microsoft and Facebook have rolled out products and features generated from hackathons.","categories":["AI Features"],"tags":["hackathons in India","Machinehack Hackathon"],"author_name":"Tasmia Ansari","publish_date":"2022-08-29T10:00:00","publication_year":"2022","word_count":832,"keywords":["big data","Go","machine learning","TPU","ELT","AI","hackathons in India","ML","RAG","analytics","Machinehack Hackathon","R"],"extracted_tech_keywords":["AI","machine learning","ML","analytics","RAG","TPU","R","Go","big data","ELT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/contrary-to-popular-belief-hackathons-do-get-deployed-for-real-world-problems\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10094015,"title":"Microsoft launches Pengi, an Audio Language Model for Open-ended Tasks","content":"Transfer Learning has been instrumental in advancing audio processing and enabling Self-Supervised Learning and Zero-Shot Learning techniques. However, current models lack the ability to generate language for open-ended tasks such as Audio Captioning or Audio Question & Answering. In response to this limitation, researchers from Microsoft have launched Pengi, a groundbreaking Audio Language Model, adopts Transfer Learning to reframe all audio tasks as text-generation tasks. By incorporating both audio and text inputs, Pengi generates free-form text as output without the need for additional fine-tuning. Extensive evaluations involving 22 downstream tasks showcase Pengi’s state-of-the-art performance, underscoring the significant progress achieved in general-purpose audio understanding through the integration of language models with audio models. The audio language model, harnesses transfer learning by treating all audio-related tasks as text-generation tasks. It operates by taking an audio recording and associated text as input, subsequently producing free-form text as output. The unified architecture of Pengi permits the handling of open-ended and close-ended tasks without necessitating extra fine-tuning or task-specific extensions. During training, Pengi is exposed to a vast dataset consisting of audio-text pairs. This dataset encompasses diverse audio recordings encompassing human speech, music, and various sounds, along with corresponding text transcripts. The audio recordings undergo processing via an audio encoder, which converts them into a sequence of continuous embeddings. Simultaneously, the text transcripts are processed by a text encoder, converting them into a corresponding sequence of continuous embeddings. These two sequences of embeddings are merged as a prefix to prompt a pre-trained frozen language model. The language model subsequently generates tokens in an autoregressive manner, conditioned on the audio and text input. To assess Pengi’s capabilities, evaluations are conducted across 22 downstream tasks, encompassing audio captioning, audio question answering, and audio event detection, among others. Pengi achieves state-of-the-art performance across several of these tasks, affirming its efficacy as a potent audio language model applicable to a wide range of tasks. Examples of Pengi’s functionalities include generating captions for audio recordings, answering questions related to audio recordings, detecting events within audio recordings, translating audio recordings into text, summarising audio recordings, and generating creative text formats such as poems, code, scripts, musical pieces, emails, and letters. Although still in development, Pengi possesses the potential to revolutionise audio interaction. With Pengi, natural conversations with devices become feasible, enabling unprecedented audio-related capabilities that were previously unattainable.","excerpt":"By incorporating both audio and text inputs, Pengi generates free-form text as output without the need for additional fine-tuning","categories":["AI News"],"tags":["Microsoft"],"author_name":"Shyam Nandan Upadhyay","publish_date":"2023-05-26T20:29:00","publication_year":"2023","word_count":387,"keywords":["Go","self-supervised learning","TPU","programming_languages:R","AI","programming_languages:Go","R","zero-shot learning","Microsoft"],"extracted_tech_keywords":["AI","zero-shot learning","TPU","R","Go","self-supervised learning","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/microsoft-launches-pengi-an-audio-language-model-for-open-ended-tasks\/","complexity_score":3,"technical_depth":8,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":672,"title":"Top 10 Analytics Courses in India &#8211; 2013","content":"The demand for trained analytics professionals has witnessed a massive growth in recent years. The dearth of skilled manpower can be overcome with serious intervention at the education level and imparting training on specific Analytical and statistical tools. This goes to say that training in Analytics is of foremost importance to match the ever growing demand and dearth in supply. Yet, there is a severe dearth of good training programs in the field. In this article, Analytics India Magazine investigates ten courses on Analytics being offered by premier institutes of India. Please note that this list is not a ranking, the list is in the alphabetic order of the School name. Read our Latest Ranking: Top 10 Executive Analytics Courses in India – Ranking 2019 PGP in Business Analytics and Big Data – Aegis School of Business PGP in Business Analytics and Big Data is India’s first holistic data science program designed and delivered by Aegis School of Business & Telecommunication in association with IBM to train the new generation of data-savvy professionals. This 11 months program provides you intensive training to develop the necessary and unique set of skills required for a successful career in the world of Big Data and Business Analytics. Program Highlights Certification from IBM at the completion of the course. IBM Business Analytics Lab: IBM has setup an IBM Business Analytics and IBM Cloud Computing Lab in the campus. Innovative Curriculum: Curriculum has been developed for the programs jointly and will be delivered by IBM designated subject matter experts and Aegis faculty. The curriculum caters to the various skill requirements of organizations across the world including Banks, Computer Services, Education, Healthcare, Insurance, Manufacturing, Retail and other industries. Tools: Hands on exposure on IBM DB2, IBM Cognos TM1, IBM Cognos Insight, IBM InfoSphere Big Insight, IBM Worklight, IBM BlueMix, R, Python, SAS, Hadoop, MapReduce, EC2, AWS, Weka etc Placement: Aegis has Career Management Centre organizes your paid internship as well as the final placement with leading companies. Duration and Credit: The PGP – BA & Big Data program is Globally acceptable 36 Credit Unit spread over 11 months that includes 2 months. Delivered Full time & Part time (online\/weekend\/hybrid) Course Modules taught by IBM: IBM faculty will be teaching three courses: (i) Business Intelligence using Cognos BI (ii) Big Data Analytics using IBM InfoSphere Big Insight and (iii) Enterprise Performance Management using IBM Cognos TM1 MBA in Business Analytics – BML Munjal University The programme has been designed in association with IBM. An IBM Business Analytics Lab has been setup at the BMU campus. IBM will be providing the relevant software and training based on Predictive Analytics using IBM SPSS, Descriptive Analytics using IBM Cognos, and Big Data & Analytics using IBM InfoSphere BigInsight. Typically, students who are interested in this programme would have studied Science, Mathematics, and Engineering; along with the general audience. Programme Advantages: 2-year, full time programme, mentored by Imperial College London Innovative curriculum jointly developed by BMU and IBM on various industry specialisations, and based on skill requirements of organizations across the world Opportunity to learn through live industry cases directly from industry practitioners from IBM and experts from BMU Professional certification from IBM on completion Hands-on learning at the on campus IBM Business Analytics Lab Post Graduate Programme in Business Analytics – Great Lakes Institute of Management, Gurgaon The PGP-BA Programme is designed with a learning methodology that is comprehensive in content and impactful in outcome. The format of the Programme combines the best of classroom sessions and technology aided learning in terms of recorded lectures and live webinars. The Programme duration is 11 months, with the participants spending 5 days (including weekend) every alternate month on campus. Between successive contact sessions, participants would take online classes and access other learning material (assignments, readings, recordings, hands on SAS lab exercises, etc) using an online learning management system. In terms of number of hours, the Programme would be spread across 240 hours of classroom sessions and about 150 hours of online learning (using recordings, webinars, virtual lab, assignments and project). Analytics Essentials – IIIT, Bangalore Designed to provide participants with a comprehensive understanding of the concepts of business analytics – the data, the approach, the technology and its day-to-day business decision applications. Cross functional approach that integrates number crunching, information technology, management applications, business analysis for a result oriented application learning. Multi-scenario exposure to a systematic and structured application lab (integrating data, process, domain and systems) for informed decisions and value insights. Inputs from decision makers, practitioners and academicians to ensure that the skill development is aligned to the industry expectations with focus on ready-to-apply solutions in real world. Advanced Analytics for Management, IIM Ahmedabad This programme is intended for enabling practitioners, managers, and decision-makers to use advanced analytics for better decision-making and in-depth understanding of these concepts using hands-on technique(s) and relating to business cases. The programme may also be of interest to participants from various analytics organizations to better understand the underlying concepts of these advanced techniques. An aptitude for quantitative modeling and some prior experience in use of analytics is desirable. The teaching methodology for this programme will be an appropriate mixture of classroom teaching, hands-on experiments, case discussion, identification of best practices, in-class participation, group reading and presentations, guest lectures, and panel discussion. Business Analytics and Intelligence (BAI) – IIM Bangalore The course is suitable for those who are already working in analytics to enhance their knowledge as well as for those with analytical aptitude and would like to start new career in analytics.  The participants of Executive Education are expected to have at least 5 years of work experience. The programme will be conducted live in the classroom at IIMB. The sessions will be beamed instantaneously across selected cities in India through Reliance Web World outlets using video conferencing facilities that allow a large number of geographically dispersed participants to participate in highly interactive sessions with the faculty. Certificate Programme in Business Analytics – IIM Lucknow Indian Institute of Management, Lucknow (IIML) and Kelley School of Business (KSB), Indiana University, USA is offering one year part time certificate programme in business analytics for executives (CPBAE). The program is jointly taught by highly experienced faculty from IIM Lucknow, India and the Kelley School of Business, Indiana University, USA. A combination of class room and an online learning platform is used. It offers flexibility since the participants can access the internet and attend online classes from any location. Training on predictive analytics with Eminer by SAS Institute (India) Private Limited, access to e-learning and opportunity for getting certified on Predictive Analytics. Certificate Programme in Business Analytics – ISB, Hyderabad ISB is offering a one year Certification in Business Analytics with an aim to create Next generation Data Management Scientists. The programme is designed on a schedule that minimizes disruption of work and personal pursuits. The program is a combination of classroom and Technology aided learning platform with 430 contact hours. Participants will typically be on campus for a 5 day schedule of classroom learning every eight weeks which would ideally be planned to include a weekend. The schedule will include full days of teaching and evenings will be used for Industry speakers, quizzes&  group work. Apart from this some modules and Tutorials will be taken on the technology platform .The program also has a component on Business Analysis & Predictive modeling by SAS. Another key element embedded is the Action learning project whose duration varies between 3- 6 months focused on real subject from the organization . Certificate Program in Business Analytics – NMIMS Hyderabad Business analytics (BA) refers to the skills, technologies, applications and practices for continuous iterative exploration and investigation of past business performance to gain insight and drive business planning. Increases in computing power and the amount of data collected have led to the development and widespread adoption of analytics by numerous industries. The proposed program on Business Analytics will combine coursework on developing analytical thinking, introductory and advanced statistical concepts, and applications across wide section of industry and functional domains. In partnership with SAS, a 5-day ‘Statistical Business Analyst’ program will be conducted by SAS leading to SAS Certification as well. This program will be delivered by SAS at the end of the Certification Program offered by NMIMS. The students will be provided by the SAS course material and the necessary software package with validity over the training period. 1 Year Full Time Post Graduate Program – Praxis Business School One of the few full time courses in India, this is an industry supported program in Business Analytics designed and delivered in collaboration with knowledge partner ICICI Bank and knowledge supporter PwC. This unique Praxis Business Analytics Program is spread over 12 months. The program is divided into 4 trimesters: Three trimesters of three months each comprise classroom training on campus One trimester of three months is assigned for internship in an organization The program offers an equivalent of 16 courses of 3-credits each. This translates into 480 hours of classroom training including projects, presentations and assignments. Some of these assignments may be ‘live’ industry assignments. Students will be offered a 3-month internship with organizations of repute. They will be required to work on live assignments in the domain of Business Analysis. [divider top=”1″] [poll id=”3″]","excerpt":"The demand for trained analytics professionals has witnessed a massive growth in recent years. The dearth of skilled manpower can be overcome with serious intervention at the education level and imparting training on specific Analytical and statistical tools. This goes to say that training in Analytics is of foremost importance to match the ever growing […]","categories":["AI Features"],"tags":["Praxis Business School"],"author_name":"Дарья","publish_date":"2012-08-08T20:48:41","publication_year":"2012","word_count":1551,"keywords":["data science","AWS","AI","cloud computing","R","ML","Python","Aim","analytics","predictive analytics","Praxis Business School"],"extracted_tech_keywords":["AI","ML","data science","analytics","Aim","predictive analytics","cloud computing","AWS","Python","R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/top-6-analytics-courses-in-india\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":2055,"title":"EMC launches analytics courses in India","content":"EMC Corporation has announced EMC Proven Professional training and certifications to help Indian businesses address skill transformation and the looming Cloud Computing and Data Science talent gap. Rajesh Janey, president, EMC India and SAARC, said that the Indian cloud computing market (use of computing resources delivered as a service over the internet), currently estimated at $400 million, was likely to touch $4.5 billion by 2015 and the business opportunity in big data (huge data difficult to process with existing tools) is expected to touch $300 million in a couple of years. “However, there is a huge shortage of manpower in these domains,” he said. According to Janey, the new courses will be a part of EMC Academic Alliance which has been implemented by tying up with over 250 educational institutes in India and to EMC’s customers and partners. EMC will offer Cloud Computing, Data Science and Big Data Analytics training and certification programs with an “open” curriculum to help develop professionals who can extract more value and insight for competitive advantage from the new expanse of data that businesses create and capture. In addition, the company aims to train over 30,000 engineering students in the first year itself as part of its EAA program. The new Data Science and Big Data Analytics training and certification helps build a solid foundation for data analytics with a particular focus on the opportunities and challenges presented by Big Data. According to Rajesh Janey, “In the emerging digital era, individual need and sentiment has become more prevalent thanks to the proliferation of social and other online networks”. “Cloud computing is transforming IT and big data is transforming business. But there is shortage of people with requisite skills. It is estimated globally that there is a shortage of around 192,000 data scientists,” Janey said. He said that the new data science and big data analytics training and certification helped build a solid foundation for data analytics with focus on opportunities and challenges presented by big data. The new Cloud Infrastructure and Services Associate-Level Training and Certification provides essential foundation of technical principles and concepts to enable IT professionals to together make informed business and technical decisions on migration to the cloud. Citing EMC Zinnov study Janey said that digital information is creating new opportunities in cloud computing and data. Janey said that the cloud opportunity in India is expected to create 100,000 jobs by 2015 from 10,000 in 2011. In addition to customers and partners, EMC will offer these courses to their university partners through its EMC Academic Alliance (EAA) program, an EMC initiative that aims to collaborate with the leading educational institutions around the world to address the emerging knowledge gap in the areas of Cloud Computing and Data Science and Big Data Analytics.” Apart from addressing the domestic big data market need, India has the opportunity to address the global market expected to touch $25 billion. India’s potential to tap this market is around $1 billion by 2015 at a compounded annual growth rate (CAGR) of 83 percent, Janey said citing a NASSCOM CRISIL study. The Cloud Infrastructure and Services, Cloud Architect and Data Science training and certifications are easily accessible for customers, partners, students, and those in the industry at large who desire in-depth knowledge of these transformational technologies and new professional disciplines. EMC offers flexible learning options by providing participants the choice to learn at their own pace via popular Video-instructor-led Training (VILT) or attend class room instructor-led courses.","excerpt":"EMC Corporation has announced EMC Proven Professional training and certifications to help Indian businesses address skill transformation and the looming Cloud Computing and Data Science talent gap. Rajesh Janey, president, EMC India and SAARC, said that the Indian cloud computing market (use of computing resources delivered as a service over the internet), currently estimated at […]","categories":["AI Trends"],"tags":[],"author_name":"Дарья","publish_date":"2012-11-14T12:25:37","publication_year":"2012","word_count":578,"keywords":["big data","data science","programming_languages:R","cloud computing","AI","Git","Aim","analytics","R"],"extracted_tech_keywords":["AI","data science","analytics","Aim","cloud computing","R","Git","big data","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/emc-launches-cloud-analytics-courses-in-india\/","complexity_score":3,"technical_depth":9,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10140928,"title":"The CUDA Killer","content":"While NVIDIA’s fame rests on its GPUs, the real magic comes from CUDA, the software it can’t do without. In a recent interview with No Priors, NVIDIA chief Jensen Huang said that the goal is for their AI engineers to “build once, run everywhere”. “The investment in software is the most expensive,” Huang said, speaking about CUDA and its critical role in supporting and maximising the potential of hardware. He further said that NVIDIA maintains a strong commitment to supporting its software indefinitely, citing programming language C as an example of this approach. “We’ve never given up on a piece of software,” he said, adding that NVIDIA will continue to maintain the software it develops “for as long as we shall live”. In another interview, he revealed that the psychology of protecting the software began in 1993 (the year NVIDIA was founded) and has been the company’s priority ever since. “The reason why NVIDIA’s CUDA has such a massive install base is because we have always protected it,” said Huang. Today, five million developers across about 40,000 companies use CUDA. It provides a robust environment with over 300 code libraries, 600 AI models, and support for 3,700 GPU-accelerated applications, catering to diverse computing needs. CUDA primarily supports C, C++, and Fortran and includes a well-established API with extensive libraries for parallel processing, such as cuBLAS for linear algebra, cuDNN for deep learning, and Thrust for parallel algorithms. Developers can also use frameworks like PyTorch and TensorFlow, which come with built-in CUDA support. AMD ROCm Rolls To challenge NVIDIA’s CUDA, AMD launched ROCm 6.2, which introduces support for essential AI features such as the FP8 datatype, Flash Attention 3, Kernel Fusion, and more. These updates enable ROCm 6.2 to deliver up to a 2.4x performance boost in inference and a 1.8x improvement in training across a range of LLMs compared to ROCm 6.0. In an exclusive interview with AIM, Sasank Chilamkurthy, the founder of Johnaic, said that ROCm has an advantage over CUDA due to its strong support for PyTorch. He added that companies preferring to own their code may benefit from using ROCm. Chilamkurthy also shared a fun fact with us: ROCm was initially built for NVIDIA GPUs, however, AMD later prioritised it for its own GPUs. Notably, the company recently introduced its new MI325X accelerators for training and inferencing LLMs. During the Advancing AI Summit in 2024, Vamsi Bopanna, SVP of AI at AMD, shared more details about ROCm. “ROCm is a complete set of libraries, runtime compilers, and tools needed to develop and deploy AI workloads. We designed ROCm to be modular and open source, allowing for rapid contributions from AI communities,” said Bopanna, adding that it is also designed to connect easily with ecosystem components and frameworks like PyTorch and model hubs like Hugging Face. He explained that they have expanded support for newer frameworks like JAX and implemented powerful new features, algorithms, and optimisations to deliver the best performance for generative AI workloads. AMD also supports various open-source frameworks, including vLLM, Triton, SGlang, ONXX Runtime, and more. Bopanna revealed that today, over 1 million Hugging Face models run on AMD. “We have a very deep partnership with PyTorch. We are fully upstreamed in it, run over 200,000 tests nightly, and are a tier-one citizen within the PyTorch community,” said Bradley McCredie, corporate vice president at AMD, adding that there are only two compute platforms that are fully upstreamed in PyTorch, and AMD is one of them. AMD also works closely with Triton, an open-source programming language and compiler developed by OpenAI for GPU programming. While originally designed for NVIDIA GPUs, recent developments have enabled Triton to be compatible with AMD GPUs through the ROCm platform. “Triton is an extremely strategic platform for our industry as it provides a highly productive environment with a high level of abstraction for coders, enabling excellent performance,” said McCredie. He added that it eliminates the dependence on hardware-specific languages like CUDA, allowing programmers to code directly at this level and compile directly to the AMD platform. Moreover, AMD GPUs with ROCm offer a more cost-effective option when compared to NVIDIA GPUs with CUDA. Although its top-tier GPUs may lag by 10-30% in raw performance, the price difference can be significant. AMD showcased testimonials for ROCm at Advancing AI 2024 by inviting startup leaders, including Amit Jain, the CEO of Luma AI; Ashish Vashwani, the CEO of Essential AI; Dani Yogatama, the CEO of Reka AI, and Dmytro Dzhulgakov, the CTO of Fireworks AI. Luma AI recently launched a video-generation model called Dream Machine. “The models we’re training are very challenging and don’t resemble LLMs at all. However, we were impressed with how quickly we could get the model running on ROCm and MI300X GPUs. It took us just a few days to establish the end-to-end pipeline, which is quite fantastic,” said Jain. Advantage CUDA? AMD’s ROCm is not quite as mature as CUDA. NVIDIA’s ecosystem around CUDA is extensive, with a large developer community, extensive documentation, and a broad set of tools for debugging and profiling. Most deep learning frameworks, HPC (high-performance computing) applications, and libraries are developed with CUDA in mind, making it the go-to choice for many developers. NVIDIA introduced CUDA in 2006 as a proprietary parallel computing platform and application programming interface (API) model, while ROCm was originally launched in 2014. AMD is still developing ROCm to catch up. A user on Hacker News suggested that for ROCm to have a chance against CUDA, AMD would need to commit billions to ecosystem-building. This involves supporting developers, creating resources, and nurturing a long-term platform that can rival CUDA’s popularity. The user explained that historically, hardware companies like Intel and AMD have struggled with building and maintaining strong software ecosystems. Using OpenCL as an example, he pointed out that although it was supported by hardware companies, it failed to develop into a strong competitor due to inconsistent support and a lack of ecosystem investment. ​​The only problem with CUDA is that it is closed source and works only for NVIDIA GPU workloads. Though people have been finding several solutions to work around this restriction, CUDA remains the best on NVIDIA GPUs, and since everyone is using its GPUs, the moat becomes even bigger. Today, NVIDIA controls 95% of the AI chip market. “CUDA is dominant in GPU programming because NVIDIA dominates the GPU market when it comes to AI, ML, and other GPU programming applications. CUDA is exclusive to NVIDIA,” posted a user on Reddit. In India, startups like Unscript and Sarvam AI revealed to AIM that they work with NVIDIA GPUs and CUDA and haven’t adopted ROCm yet. Last year, AMD acquired Nod.ai to provide AI customers with open software for easily deploying high-performance AI models optimised for AMD hardware. The company recently acquired Europe’s private AI lab, Silo AI. “We recently completed the acquisition of Silo AI, which adds a world-class team with tremendous experience in training and optimising LLMs, as well as delivering customer-specific AI solutions,” said AMD chief Lisa Su. All in all, while ROCm offers a compelling alternative, CUDA’s maturity and widespread use make it hard to surpass it.","excerpt":"ROCm was initially built for NVIDIA GPUs; however, AMD later prioritised it for its own GPUs.","categories":["Global Tech"],"tags":["AMD","NVIDIA"],"author_name":"Siddharth Jindal","publish_date":"2024-11-13T18:51:33","publication_year":"2024","word_count":1194,"keywords":["AMD","Hugging Face","OpenAI","AI","PyTorch","ML","Aim","deep learning","generative AI","JAX","NVIDIA","TensorFlow"],"extracted_tech_keywords":["AI","ML","deep learning","generative AI","OpenAI","Aim","TensorFlow","PyTorch","JAX","Hugging Face"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/the-cuda-killer\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10032715,"title":"Behind NVIDIA’s Megatron","content":"Natural Language Processing (NLP) has made considerable strides in recent years on the back of the availability of larger datasets and computation at scale. Recent works have demonstrated that large language models can have high accuracy on many NLP datasets without additional finetuning. However, researchers find it challenging to train large language models, primarily due to two main reasons: Since the GPU memory is limited, it isn’t easy to fit these large models on a single or even multi-GPU server.To train these models, one needs to run several computing operations, resulting in unrealistic training times; for example, a model like GPT-3 with 175 billion parameters may take 36 years on eight V100 GPUs. In 2019, NVIDIA introduced MegatronLM, an 8.3 billion transformer language with model and data parallelism trained on 512 GPUs. At the time, it was the largest transformer model ever trained. While this method works for models with up to 20 billion parameters on DGX A100 servers with eight A100 GPUs, it is ineffective in larger models. This is because the models need to be split across multiple servers, causing problems like: Slower communication rate since the all-reduce communication now needs to go through inter-server links. This is slower than the high bandwidth NVLink available inside a DGX A100 server.A greater degree of model parallelism decreases GPU utilisation. Credit: NVIDIA To overcome these limitations, teams from NVIDIA, Microsoft Research, and Stanford have proposed a technique which involves smart combination of different parallelism methods — tensor, pipeline, and data parallelism to achieve a two-order magnitude increase in the size of models that can be trained, as compared to the current systems. Modes of parallelism There are three major modes of parallelism. Model parallelism: Individual layers of the model are partitioned over multiple devices. The team deployed a partitioning strategy employed by Megatron for transformer layers that form the base for language models. Pipeline parallelism: The layers of the model are shared across multiple devices. When used on repetitive transformer-based models, each device can be assigned to an equal number of transformer layers. Here, the batch is split into smaller micro-batches, and the execution is pipelined across these micro-batches. The pipeline schemes ensure the inputs see consistent weight updates in both backward and forward passes. Data parallelism: Each worker has a copy of the full model. The workers aggregate gradients periodically to ensure that all workers see a consistent version of the weights. For large models that do not fit a single worker, data parallelism can be used on smaller model shards. Scaling model training The team showed a way to combine tensor, pipeline and data parallelism methods to train large models. “The combination of tensor parallelism within a multi-GPU server, pipeline parallelism across multi-GPU servers, and data parallelism, allows us to practically train models with a trillion parameters with graceful scaling in a highly-optimised cluster environment with high-bandwidth links between GPUs on the same server and across servers,” the researchers said. The team performed training iterations on models with a trillion parameters at 502 petaFLOP\/s on 3072 GPUs by combining three techniques. They achieved a per-GPU throughput of 52 percent, against the 36 percent throughput achieved with previous techniques. To achieve this high throughput, the team implemented several innovative techniques and careful engineering along multiple axes: efficient kernel implementation. This approach allowed most computations to be compute-bound rather than memory bound. The smart partitioning of computation graphs over the devices reduced the number of bytes required to be sent over network links. This also resulted in limiting idle device periods, optimisation of domain-specific communication, and fast hardware. The authors also studied interactions between the components which affects throughput. Wrapping up In the future, the team will work on optimising the pipeline schedule. Tradeoffs associated with hyperparameters such as micro batch size, global batch size, and activation recomputation on throughput will be explored. They will also study the implication of using scheduling without pipeline flushes, as used in the present set up. Read the full paper here.","excerpt":"The team performed training iterations on models with a trillion parameters at 502 petaFLOP\/s on 3072 GPUs by combining three techniques.","categories":["Global Tech"],"tags":["Data parallelism","GPU","LLMs","Model parallelism","NLP models","NVIDIA"],"author_name":"Shraddha Goled","publish_date":"2021-04-22T11:00:00","publication_year":"2021","word_count":667,"keywords":["Go","programming_languages:R","AI","LLMs","RPA","Model parallelism","A100","NLP models","NLP","GPT","Data parallelism","llm_models:GPT","V100","NVIDIA","R","GPU"],"extracted_tech_keywords":["AI","NLP","R","Go","GPT","RPA","A100","V100","llm_models:GPT","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/behind-nvidias-megatron\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":10011831,"title":"Enterprise AI Platform DataRobot Raises $270 Million For R&#038;D, Building Out Operations","content":"DataRobot, an enterprise AI company, has raised $270 million in a new round of funding, before going public, according to the news media. The company has already raised a $206 million of funding, last year, in a Series E led by Sapphire Ventures, along with New Enterprise Associates and Meritech Capital. This round of funding, whereas, was led by Altimeter Capital, in participation with investors like T. Rowe Price, BlackRock, Silverlake, New Enterprise Associates, and Tiger Global Management. The round brings the company’s valuation to $2.7 billion. COO and President Dan Wright stated to the media that this round of funding signifies that there’s now a clear leader in this very large market, and there is an opportunity to lead this AI revolution in the coming years. The company creates automated platforms for enterprises to build and deploy predictive models, stated by the company website. As a matter of fact, Wright believes that the age of “experimental AI” is gone, and currently, it is “all about the ability to make better insights that allow you to make better decisions.” This new funding will be used for research and development, expanding the company internationally, partnering with federal governments, and building out company operations, according to Wright. In terms of expanding the company, DataRobot is rapidly scaling in Europe and will be growing in Australia, Singapore and the Asia Pacific region. In terms of growth, the company is “well over” $100 million in annual recurring revenue, and it’s witnessing high double-digit growth, stated Wright. Also Read: Journey Of Director Of DataRobot DS And Kaggle GM Sergey Yurgenson According to Wright, more than 2 billion machine-learning models have been built to date on DataRobot’s cloud platform. Altimeter Capital CEO Brad Gerstner stated to the media that for too long AI was just a distant promise; however, the scenario has changed now and DataRobot is at the front in terms of democratising AI and business intelligence for organisations. According to the news media, this funding round will be the company’s last before an IPO, the timeline to which isn’t explicitly mentioned. Currently, the focus is on creating an “iconic company,” Wright said.","excerpt":"DataRobot, an enterprise AI company, has raised $270 million in a new round of funding, before going public, according to the news media. The company has already raised a $206 million of funding, last year, in a Series E led by Sapphire Ventures, along with New Enterprise Associates and Meritech Capital. This round of funding, […]","categories":["AI News"],"tags":["DataRobot"],"author_name":"Sejuti Das","publish_date":"2020-11-18T16:38:44","publication_year":"2020","word_count":358,"keywords":["business intelligence","Go","API","funding","programming_languages:R","AI","IPO","Git","DataRobot","GAN","R"],"extracted_tech_keywords":["AI","R","Go","Git","API","GAN","business intelligence","funding","IPO","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/enterprise-ai-platform-datarobot-raises-270-million-for-rd-building-out-operations\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":9457,"title":"AIG Science Summit | 16 April, 2016 | Bangalore","content":"American International Group, Inc. , also known as AIG, is an American multinational insurance corporation with more than 88 million customers in over 100 countries. The company serves 89+% of the Fortune 500 companies, 96% of Fortune 1000, and 90% of Fortune Global 500 The Science group at AIG is a catalyst for consistent evidence-based decision making across AIG. This interdisciplinary team applies scientific methods to decision making from a variety of fields, including data science, statistical modeling, behavioral decision science, economics, and others. Our approach to industrializing Science has four key components. These then transform into multiple projects within Science to deliver on an industrialized future. Some of the questions that Science at AIG is helping to answer are: Claims – How are we leveraging cutting-edge technology and insights to process claims better, cheaper, and faster?Casualty Underwriting – How are we drawing on AIG’s data advantage to price accounts better and to create risk solutions in collaboration with clients? Casualty Underwriting – How are we drawing on AIG’s data advantage to price accounts better and to create risk solutions in collaboration with clients? Sales – How is Sales becoming more fact based and prioritizing value over volume? Direct Marketing – How are we improving efficiency of Direct Marketing to raise ROI? Interested to know more? Join us on Saturday, April 16 at Park Plaza Hotel in Bangalore to learn more AIG Science and for a chance to hear distinguished speakers on latest trends and technologies in the world for data science. We will also be hosting a networking lunch, a unique opportunity to interact with data science professionals from within and outside AIG. Register now","excerpt":"American International Group, Inc. , also known as AIG, is an American multinational insurance corporation with more than 88 million customers in over 100 countries. The company serves 89+% of the Fortune 500 companies, 96% of Fortune 1000, and 90% of Fortune Global 500 The Science group at AIG is a catalyst for consistent evidence-based […]","categories":["Deep Tech"],"tags":[],"author_name":"Дарья","publish_date":"2016-04-05T08:00:55","publication_year":"2016","word_count":276,"keywords":["data science","programming_languages:R","AI","RAG","Aim","R"],"extracted_tech_keywords":["AI","data science","Aim","RAG","R","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/aig-science-summit-16-april-2016-bangalore\/","complexity_score":2,"technical_depth":6,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10066710,"title":"Yann LeCun resumes war of words with Gary Marcus","content":"I dont engage in vacuous debates. Nor do I speculate about vague hypothetical proposals. I build stuff. You should try it sometimes: Yann LeCun to Gary Marcus. You have all the resources in the world. I don’t. But that doesn’t make your approach correct: Gary Marcus to Yann LeCun After a brief lull, the war of words between Gary Marcus, a professor of Neural Science at NYU and an AI critic, and Yann LeCun, chief AI scientist at Meta, has resumed. Marcus is known for his beefs with AI researchers. Recently, he took a jab at LeCun and tagged him in a video of a Tesla crashing into a private jet. today, @ylecun: “Not only is AI not \"hitting a wall\",  cars with AI-powered driving assistance aren't hitting walls, or anything else, either.”@tesla, last week: oops https:\/\/t.co\/guHpCTqNCl— Gary Marcus (@GaryMarcus) May 7, 2022 In 2021, Marcus published an article on Nautilus, titled ‘Deep Learning Is Hitting a Wall’. The article spoke about the (wrong) direction the AI industry is headed. He cited the failure of self-driving cars to make his case. LeCun responded to the Tweet: don’t doubt that the particular statistic but this part is simply not factual: “cars with AI-powered driving assistance aren't hitting walls, or anything else, either.”— Gary Marcus (@GaryMarcus) May 8, 2022 Marcus also responded by asking LeCun, “Also, is it a pure convnet or a hybrid system?” In his 2017 debate with LeCun at the NYU Mind, Brain and Consciousness, Marcus had argued for the need for an innate machinery to push the boundaries of AI. A year later, LeCun responded to Marcus’ critique of deep learning, saying ‘the number of valuable recommendations ever made by Gary Marcus is exactly zero.’ In response to a comment on Marcus’ recent Tweet, LeCun said: “Deep Learning is a super-hot topic in radiology these days. If radiologists feel insulted by AI, why did they invite me to give a keynote at the Int. Soc. Magnetic Resonance Imaging conference?” Here’s how their Twitter debate unfolded. i have confirmed through very knowledgeable sources that the Intel\/MobileEye system is indeed a complex hybrid system, in which convnets are only part of what’s going on.— Gary Marcus (@GaryMarcus) May 8, 2022 Empty semantics.Obviously, *any* complete AI system, even one built around a big neural net, will have all kinds of other code around it.It's particularly true of systems designed to act in the real-world.Basic engineering, really.This is a completely vacuous debate.— Yann LeCun (@ylecun) May 8, 2022 You obviously never read a single one of the papers of mine that you have been criticizing. One thing for you to disagree, another to show a lack of awareness of what the person you disagree with is saying.— Gary Marcus (@GaryMarcus) May 8, 2022 You obviously never engineered nor built a working AI system.Arguments in favor \"hybrid systems\" are non-falsifiable strawmen, whose only purpose to allow the author to declare victory regardless of the outcome.As I said, this debate is vacuous.— Yann LeCun (@ylecun) May 8, 2022 i have laid out definitions of neurosymbolic hybrids, but Yann has not engaged with them.— Gary Marcus (@GaryMarcus) May 8, 2022 I dont engage in vacuous debates.Nor do I speculate about vague hypothetical proposals.I build stuff.You should try it sometimes.— Yann LeCun (@ylecun) May 8, 2022 you have all the resources in the world. I don’t.but that doesn’t make your approach correct.— Gary Marcus (@GaryMarcus) May 8, 2022","excerpt":"In 2021, Marcus published an article on Nautilus, titled ‘Deep Learning Is Hitting a Wall’.","categories":["AI News"],"tags":["gary marcus","Yann LeCun"],"author_name":"Avi Gopani","publish_date":"2022-05-10T19:51:21","publication_year":"2022","word_count":572,"keywords":["Go","Yann LeCun","programming_languages:R","AI","programming_languages:Go","gary marcus","deep learning","ViT","AI research","R"],"extracted_tech_keywords":["AI","deep learning","R","Go","ViT","AI research","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/yann-lecun-resumes-war-of-words-with-gary-marcus\/","complexity_score":4,"technical_depth":8,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10121068,"title":"AI to Help You Understand This Indian Classical Dance","content":"Kathakali, a traditional dance drama art form from Kerala, can now be understood by all, including a foreign audience. Performed based on subject matter from the Ramayana, the Mahabharata, and stories from Shaiva literature, AI can now help understand the stylised gestures and facial expressions of the performers. A recent research paper by the Indian Institute of Information Technology in Kottayam, in collaboration with the Kerala Kalamandalam, Thrissur, aims to develop an AI-enabled tool capable of providing semantic interpretations of the art form. The study hopes to leverage advancements in AI, particularly in computer vision and natural language processing, to preserve and interpret cultural heritage, specifically focusing on the Indian classical art form Kathakali. The goal is to develop a system that can automatically recognise and interpret the gestures, or mudras, used in Kathakali performances. By doing so, the study aims to address the challenges of digitising and understanding this intricate art form, which involves complex hand gestures and facial expressions. By building a database of vector representations of mudras and comparing them to input data, the system can identify and interpret the gestures being performed. Importantly, the study aims to achieve this with minimal training data, making it adaptable to other dance forms or related areas like sign language recognition. Overall, the study seeks to contribute to the preservation and understanding of cultural heritage by applying AI techniques to digitally analyse and interpret traditional art forms. AI and dance are not a new duo. In 2019, renowned choreographer Wayne McGregor unveiled his newest creation, “Living Archive: An AI Performance Experiment,” in a world premiere. McGregor collaborated with Google Arts and Culture Lab to develop an innovative AI choreographic tool, drawing from 25 years of his video archive.","excerpt":"The goal is to develop a system that can automatically recognise and interpret the gestures, or mudras, used in Kathakali performances.","categories":["AI News"],"tags":["AI (Artificial Intelligence)"],"author_name":"Vidyashree Srinivas","publish_date":"2024-05-20T17:53:42","publication_year":"2024","word_count":288,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","Git","computer vision","RAG","Aim","ai_applications:computer vision","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","computer vision","Aim","RAG","R","Go","Git","programming_languages:R","programming_languages:Go","ai_applications:computer vision"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/ai-to-help-you-understand-this-indian-classical-dance\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10169121,"title":"Tim Cook Bats for ‘Team India’ as Apple Battles Trump Tariffs","content":"To say Apple is facing challenges would be an understatement. As it scrambles to catch up in the AI race, the tech giant has also suffered a major setback, losing a landmark court battle to American game developer Epic Games. Moreover, while CEO Tim Cook expressed his support for newly elected president Donald Trump, his tariff war with China adds to the list of Apple’s woes. Just four days after Trump announced tariffs against Chinese imports, Apple lost $770 billion in market cap. In the company’s Q2 2025 earnings call, Cook noted that the company anticipates a $900 million additional expense in the ongoing quarter if the 145% tariffs on Chinese imports are implemented. Most of the company’s iPhones still originate from China. However, Apple’s saviour during this period of uncertainty is India (imports from India are taxed at 26%). “For the June quarter, we do expect the majority of iPhones sold in the US will have India as their country of origin,” said Cook, while adding that the rest of the devices will be assembled in Vietnam. In collaboration with contract manufacturers, namely Foxconn and Tata, Apple assembled iPhones worth $22 billion in the 12 months that ended in March, as reported by Bloomberg. This represents a 60% increase from the previous year, and now, Apple produces one in five iPhones in India, according to the sources cited by Bloomberg. In an attempt to prepare for the impending tariffs, Apple chartered cargo flights to ship 600 tonnes of iPhones from India to the United States, according to a Reuters report. The report also noted that the Indian airport authorities reduced the customs clearance time from 30 hours to six hours for the six cargo aircraft carrying iPhones. Counterpoint research revealed that 16% of the iPhones made for Apple globally last year were assembled in India. The company’s research vice president, Neil Shah, said, “The stars are aligned for India to be the alternative destination to China.” US-Assembled iPhone 16 Pro Will Cost 25% More At present, the US government has put a temporary hold on tariffs, including the significant one imposed on China regarding chips, smartphones, and electronic devices. Nonetheless, the forthcoming Section 232 investigation started by the US Department of Commerce could lead to a revised tariff for chips and electronics. JPMorgan estimates a 20% tariff on iPhone components imported to the US, as reported by Barrons. Further details of JPMorgan’s analysis were revealed by India Dispatch, which estimates the cost of an iPhone if it were to be assembled in the United States, versus China and India. In addition to the import tariffs on the components, the analysis indicates that assembly costs in the US would double due to rising wage demands and the profit margins of contract manufacturers. Thus, with a 20% tariff on imported components, the retail price of the iPhone would rise by 30% relative to China. In contrast, when factoring in assembly costs in India, the effective retail price would only go up by 2% compared to China. The report also compared the price of the iPhone’s bill of materials (BOM) in India and China. Last month, CNBC reported on an analysis from Wamsi Mohan, a Bank of America Securities analyst, who estimated that the cost of an iPhone 16 Pro could increase by 25%. Mohan attributed the increase solely to labour costs. But Will China Let It Happen? There have been several counterarguments. Craig Moffet, a senior analyst at MoffettNathanson LLC, suggests that moving assembly to India may not help, as the components would still originate from China. This would mean that any resistance from China towards harsh US tariffs would hurt Apple’s plans. A recent report from The Information reveals that China is hindering Apple’s operations. The country’s authorities have refused to allow one of Apple’s equipment suppliers to export machinery essential for the trial production of the upcoming iPhone 17  to India. Additionally, the report notes that Chinese officials are ‘delaying or blocking’ shipments to India without providing any explanation. Foxconn, one of Apple’s key manufacturing partners, reportedly observed that approval times for exporting equipment from China to India have increased from two weeks to as much as four months. Apple is also eagerly waiting for upcoming decisions by the government. “I don’t want to predict the future because I’m not sure what will happen with the tariffs, and there is the Section 232 investigation going on. And so it’s very difficult to predict beyond June,” said Cook in the earnings call, reiterating that given the current proposed tariffs, the company can only estimate $900 million in additional costs for the quarter.","excerpt":"“For the June quarter, we do expect the majority of iPhones sold in the US will have India as their country of origin,” said Apple’s CEO.","categories":["Global Tech"],"tags":["AI (Artificial Intelligence)","Apple","India"],"author_name":"Supreeth Koundinya","publish_date":"2025-05-05T14:55:14","publication_year":"2025","word_count":772,"keywords":["Go","programming_languages:R","AI","Apple","programming_languages:Go","GAN","R","India","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","R","Go","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/tim-cook-bats-for-team-india-as-apple-battles-trump-tariffs\/","complexity_score":3,"technical_depth":6,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10110953,"title":"Apple Introduces AIM, New Autoregressive Pre-Trained Vision Models","content":"Apple recently unveiled autoregressive image models (AIM), a collection of vision models pre-trained with an autoregressive objective. These models represent a new frontier for training large-scale vision models, which are inspired by their textual counterparts, large language models (LLMs), and exhibit similar scaling properties. The researchers said that it presents a scalable method for pre-training vision models without supervision. The authors have used a generative autoregressive objective during pre-training and propose technical improvements to adapt it for downstream transfer. Check out the GitHub repository here. The researchers said that the performance of the visual features scale with both the model capacity and the quantity of data. Further, they said that the value of the objective function correlates with the performance of the model on downstream tasks. The team have also illustrated the practical implication of these findings by pre-training a 7 billion parameter AIM on 2 billion images, which achieves 84.0% on ImageNet-1k with a frozen trunk. Interestingly, even at this scale, they have observed no sign of saturation in performance. The pre-training of AIM is similar to the pre-training of LLMs, and does not require any image-specific strategy to stabilize the training at scale. About AIM Apple believes that AIM has desirable properties, including the ability to scale to 7 billion parameters using a vanilla transformer implementation without stability-inducing techniques or extensive hyperparameter adjustments. Moreover, AIM’s performance on the pre-training task has a strong correlation with downstream performance, outperforming state-of-the-art methods like MAE and narrowing the gap between generative and joint embedding pre-training approaches. The researchers have also found no signs of saturation as models scaled, suggesting potential for further performance improvements with larger models trained for longer schedules.","excerpt":"Apple recently unveiled autoregressive image models (AIM), a collection of vision models pre-trained with an autoregressive objective.","categories":["AI News"],"tags":[],"author_name":"Arya Vishwakarma","publish_date":"2024-01-17T16:19:50","publication_year":"2024","word_count":280,"keywords":["programming_languages:R","AI","RPA","Scala","Git","Aim","programming_languages:Scala","GitHub","R"],"extracted_tech_keywords":["AI","Aim","R","Scala","Git","GitHub","RPA","programming_languages:R","programming_languages:Scala"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/apple-introduces-aim-new-autoregressive-pre-trained-vision-models\/","complexity_score":2,"technical_depth":9,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10160746,"title":"AI Agents Set to Join the Workforce by 2025, says OpenAI’s Sam Altman","content":"OpenAI chief Sam Altman is confident in the development of artificial general intelligence (AGI) and predicts that AI agents could enter the workforce by 2025. “We believe that, in 2025, we may see the first AI agents join the workforce and materially change the output of companies,” Altman wrote in a recent blog post. He further expressed confidence that OpenAI has mastered the methods necessary to build Artificial General Intelligence (AGI) as traditionally understood, emphasizing the potential of these technologies to enhance human capabilities. “We continue to believe that iteratively putting great tools in the hands of people leads to great, broadly-distributed outcomes,” he said. Looking beyond AGI, Altman revealed that OpenAI is shifting its focus toward developing superintelligence, which he believes could revolutionise scientific discovery and innovation. He described superintelligent tools as capable of performing tasks far beyond human capabilities, potentially leading to unprecedented levels of abundance and prosperity. “With superintelligence, we can do anything else,” he said, suggesting that these advancements could accelerate progress in fields such as medicine, climate science, and technology. Altman acknowledged that the concept of superintelligence may sound speculative but expressed confidence that its significance will become apparent in the coming years. “We’ve been there before and we’re OK with being there again,” Altman said, referencing the company’s history of pursuing ambitious goals. When OpenAI released its most powerful o3 series of models, speculations arose about whether the model achieved AGI, given its performance on benchmarks. The o3 model scored almost 90% on the ARC-AGI benchmark, exceeding human performance. Meanwhile, Microsoft and OpenAI have agreed on a new, specific definition of AGI, or artificial general intelligence. According to a report, OpenAI can only achieve AGI when it has built a system that can generate $100 billion in profits. Interestingly, a few days ago, Google’s senior product manager Logan Kilpatrick, posted on X, “Straight shot to ASI is looking more and more probable by the month… this is what Ilya saw.” Kilpatrick highlighted the approach taken by Ilya Sutskever, co-founder of OpenAI and the Superintelligence (SSI). “Ilya founded SSI with the plan to do a straight shot to Artificial Super Intelligence,” Kilpatrick said. “No intermediate products, no intermediate model releases.”","excerpt":"“We are beginning to turn our aim beyond that, to superintelligence in the true sense of the word.”","categories":["AI News"],"tags":["OpenAI"],"author_name":"Siddharth Jindal","publish_date":"2025-01-06T10:55:42","publication_year":"2025","word_count":365,"keywords":["Go","TPU","OpenAI","AI","programming_languages:R","innovation","programming_languages:Go","AI agents","GAN","R"],"extracted_tech_keywords":["AI","OpenAI","TPU","R","Go","GAN","innovation","AI agents","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/ai-agents-set-to-join-the-workforce-by-2025-says-openais-sam-altman\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10119515,"title":"American Express Inaugurates Largest Global Office in Gurugram","content":"Global financial giant American Express has inaugurated its largest global office in Gurugram, India, covering nearly one million square feet. This new headquarters is located in Sector 74A and will see employees moving in stages, starting by the end of this month. The Gurugram campus is part of a strategic initiative to boost the company’s presence in India by combining global expertise with local talent. This investment aims to construct a modern, environmentally friendly office space. The facility has earned LEED Gold certification because of its sustainable architecture and construction. Environmental considerations are integral for the company, featuring energy-efficient LED lighting, smart building solutions, and electric vehicle charging stations. It also involves the use of renewable energy and effective waste management practices. The design of the campus prioritises the health and well-being of its staff, incorporating green spaces, ergonomic workstations, and areas for relaxation such as quiet rooms and lounges. The intent is to create an environment that enhances productivity and employee satisfaction. “The facility is a fitting reflection of our brand and the kind of workplace where our colleagues can thrive,” said Gagandeep Singh, senior vice president of global real estate and workplace experience at American Express. Sanjay Khanna, CEO and country manager for American Express in India, highlighted the campus’s role in fostering innovation and adding global value. Additional features of the campus include a dynamic cafeteria with a live kitchen, a fitness centre, outdoor sports facilities, and terraces, all designed to promote community and collaborative engagement among employees. Access to campus amenities is streamlined through user-friendly apps, facilitating easy integration and flexibility for the workforce. Apart from Gurgaon, American Express operates from multiple offices in India including Bengaluru, Mumbai, Chennai, Pune and New Delhi. The company strategically leverages generative AI to improve its customer service and operational efficiency, opting for partnerships with established AI models rather than developing its own LLMs. This approach allows them to integrate such analytics and customer sentiment analysis into their services, improving predictive capabilities and personalising the customer experience.","excerpt":"The nearly one million square feet office has earned LEED Gold certification because of its sustainable architecture and construction.","categories":["Deep Tech"],"tags":["American Express"],"author_name":"Shritama Saha","publish_date":"2024-05-03T12:20:47","publication_year":"2024","word_count":337,"keywords":["American Express","Go","AI","sentiment analysis","ML","RAG","Aim","generative AI","analytics","GAN","R"],"extracted_tech_keywords":["AI","ML","analytics","generative AI","Aim","RAG","sentiment analysis","R","Go","GAN"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/american-express-inaugurates-largest-global-office-in-gurugram\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":49887,"title":"Oculus CTO John Carmack Steps Down To Work On Artificial General Intelligence","content":"John Carmack, the noted tech visionary and the CTO of Facebook’s VR subsidiary Oculus, announced that he is stepping down from his post. Carmack, who recently won the first-ever Accenture VR Lifetime Achievement Award, said that he now plans to focus his time on artificial general intelligence and that he would be moving to a ‘consulting CTO’ position with Oculus. He added that he’d still have a “voice” in the company. He said in a Facebook post, “I’m going to work on artificial general intelligence (AGI). I think it is possible, enormously valuable, and that I have a non-negligible chance of making a difference there, so by a Pascal’s Mugging sort of logic, I should be working on it.” Carmack also said that he’d be working on AGI in the typical “Victorian Gentleman Scientist” style, pursuing his inquiries from home, and drafting his son into the work. Carmack began working at Oculus in 2013 as the CTO and helped popularise the movement to bring consumer VR to the masses with modern technology. He started at Facebook after the social media giant acquired Oculus for $2 billion in 2014. Here’s Carmack’s full statement: Starting this week, I’m moving to a “Consulting CTO” position with Oculus.I will still have a voice in the development work, but it will only be consuming a modest slice of my time.As for what I am going to be doing with the rest of my time: When I think back over everything I have done across games, aerospace, and VR, I have always felt that I had at least a vague “line of sight” to the solutions, even if they were unconventional or unproven. I have sometimes wondered how I would fare with a problem where the solution really isn’t in sight. I decided that I should give it a try before I get too old.I’m going to work on artificial general intelligence (AGI).I think it is possible, enormously valuable, and that I have a non-negligible chance of making a difference there, so by a Pascal’s Mugging sort of logic, I should be working on it.For the time being at least, I am going to be going about it “Victorian Gentleman Scientist” style, pursuing my inquiries from home, and drafting my son into the work.Runner up for next project was cost effective nuclear fission reactors, which wouldn’t have been as suitable for that style of work. 😊","excerpt":"John Carmack, the noted tech visionary and the CTO of Facebook’s VR subsidiary Oculus, announced that he is stepping down from his post. Carmack, who recently won the first-ever Accenture VR Lifetime Achievement Award, said that he now plans to focus his time on artificial general intelligence and that he would be moving to a […]","categories":["AI News"],"tags":["AI (Artificial Intelligence)","ar","oculus","VR"],"author_name":"Prajakta Hebbar","publish_date":"2019-11-14T13:33:33","publication_year":"2019","word_count":400,"keywords":["Go","ELT","oculus","programming_languages:R","AI","programming_languages:Go","ar","VR","GAN","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","R","Go","ELT","GAN","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/oculus-cto-john-carmack-steps-down-to-work-on-artificial-general-intelligence\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10000798,"title":"Why The Next Gravitational Wave Detector Is Going To Be Set Up In India?","content":"The great discovery of gravitational waves was by LIGO of Livingston and Hanford detectors, which also won the 2017 nobel  prize, which included Karan Jani, a LIGO scientist of Indian origin. It can be called the most remarkable discoveries of today’s time, confirming Einstein’s idea of their existence. India is also on its mission to built such a highly sensitive interferometry to detect gravitational waves and is already on its way to build one. IndiGO – Built In Collaboration With LIGO IndiGO is an international collaboration between the LIGO Laboratory and three lead institutions in the IndIGO consortium: Institute of Plasma Research (IPR) Gandhinagar, Inter University Centre for Astronomy and Astrophysics (IUCAA), Pune and Raja Ramanna Centre for Advanced Technology (RRCAT), Indore. The institute will add to an already established network of the LIGO detectors. Understanding the Instrument When even a few molecules bounce off of LIGO’s mirrors, it could induce vibrations that could mask a gravitational wave signal. Therefore,for the detectors to detect these waves, they have to be extremely sensitive. What better option than the michelson Interferometer, which are the most sensitive devices to be ever built? The instrument invented by the American physicist Albert Michelson in the 1880’s has a number of applications today and the gravitational wave detection is one of them. Basic configuration of a Michelson Interferometer. IndiGO will have the interferometer with Fabry-Perot enhanced arms of 4 km length. The motive of the setup is to detect differential changes in the arm-lengths as small as 10-23 Hz-1\/2 in the frequency range between 30 to 800 Hz. The design would be identical to that of Advanced LIGO detectors that are being commissioned in the US. Why We Need To Detect GW At All? The first plot shows the combined antenna-pattern functions (a measure of the sensitivity of the network) of a network consisting of two LIGO detectors in Hanford, one LIGO in Livingston and the Virgo detector in Italy (HHLV). The second plot shows the same for a network consisting of one LIGO detector in Hanford, LIGO Livingston, Virgo and LIGO-India (HLVI). Redder regions correspond to regions of higher sensitivity and bluer regions to regions of poorer sensitivity. Locations of the detectors (LIGO-Hanford, LIGO-Livingston, Virgo, LIGO-India, KAGRA) are marked by filled black circles. The scientific benefits of LIGO-India are enormous. Adding a new detector to the existing network will increase the expected event rates. By adding a new detector in the network, the sensitivity, sky coverage and the duty cycle will automatically be increased and will therefore boost the detection confidence of new gravitational wave sources. The improvement from LIGO-India would come in the ability of localizing these sources in the sky. It was important to add this new detector in Aundh in Hingoli district of Maharashtra, so that it is geographically well-separated for the rest of the detectors, thereby improving the source-location accuracies dramatically, by about 5-10 times, thus enabling us to use GW observations as an excellent astronomical tool. Impact Of IndiGO On Indian Research Research: There is currently no research in gravitational wave in India based on Indian instrument. The development of this LIGO-India project will open many opportunities for the coming generation of scientists to do research in the field. It will bring together scientists and engineers from different fields like optics, lasers, gravitational physics, astronomy and astrophysics, cosmology, computational science, mathematics and various branches of engineering. In order to fully realize the potential of multi-messenger astronomy, the LIGO-India project will join forces with several Indian astronomy projects. Potential collaborators include the Astrosat  project, future upgrades of the India-based Neutrino Observatory and optical\/radio telescopes. Industry: LIGO project has facilitated major industry-academic research partnerships in USA and Europe, and has produced several important technological spin offs . LIGO-India will provide similar opportunities to Indian industry. This new project to be completed by 2025 will certainly bring many research opportunities in the country that were never open before and the setup will allow us to contribute to groundbreaking discoveries, in partnership with the other LIGO collaborators.","excerpt":"The great discovery of gravitational waves was by LIGO of Livingston and Hanford detectors, which also won the 2017 nobel  prize, which included Karan Jani, a LIGO scientist of Indian origin. It can be called the most remarkable discoveries of today’s time, confirming Einstein’s idea of their existence. India is also on its mission to […]","categories":["AI Features"],"tags":["Einstein","Interviews and Discussions"],"author_name":"Disha Misal","publish_date":"2019-01-02T20:54:49","publication_year":"2019","word_count":673,"keywords":["Go","programming_languages:R","AI","programming_languages:Go","Einstein","RAG","BERT","llm_models:BERT","ViT","GAN","R","Interviews and Discussions"],"extracted_tech_keywords":["AI","RAG","R","Go","BERT","GAN","ViT","llm_models:BERT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/why-the-next-gravitational-wave-detector-is-going-to-be-set-up-in-india\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10054396,"title":"Council Post: The Big Refresh: Insights-led Segmentation Is The Next Normal","content":"Imagine this scenario- a well-packaged mobile application or web page with an attractive interface but no customer traction. Unfortunately, this is the reality for many consumer brands today.  Despite packaging their products and services in a ‘cool’ manner, very few customers spot or use those products. The reason is, for a long time,  organizations have taken a market-oriented approach with narrow techniques to understand the customer. However, the digital world we’re living in has disrupted more than the technology for communication. More and more consumers expect brands to understand their preferences to engage and purchase. Companies can act on these changing consumer expectations is by creating a customer-centric strategy that puts the consumer at the forefront and creates value through the company’s offerings. This involves understanding the customer, their needs and creating a meaningful experience for the consumer on their terms. Customer-centric Customer centricity is essentially putting the customer first and creating value for the customer. This goes beyond just providing the cool product\/service at the point of purchase or offering gifts or coupons on a purchase. Instead, it is about an approach that looks at the total customer lifetime value and focuses on consumer retention. Brands need to collaborate behavioural science with insights-led segmentation to understand the customer well enough. They should be able to anticipate the consumer’s desires and create meaningful experiences. More and more organizations understand the need to shift focus from mere campaigns and begin thinking about the customer. The challenge is to turn this blueprint into reality. Consumer brands might have the data, the technology stack, and the business processes in place. However, they cannot ensure that all of these work as a single unit. This is much easier said than done – such an insights-led approach can be cumbersome or nearly impossible without the right technology partner. So, what does it take to build a customer-centric engagement strategy and execute it at scale? The answer is an insights-led approach that begins with a deeper analysis of customer insights that help brands go beyond the broad norms of segmentation is essential. Insights-led Segmentation The traditional segmentation models group customers based on their geographic, demographic, or behavioural traits. While these are great starting points to understand your customers, companies need to delve deeper into the customer’s context and analyze the overall interests, preferences, affinities, and lifestyle to understand the consumer’s overall identity. Forward-looking brands lean towards an insights-led strategy to segmentation which involves a business framework with data at the foundational base upon which customer engagement decisions are made. The insights-led strategy involves analyzing and predicting customer behavior by leveraging analytics. Based on these insights, organizations can delve deeper and organize customers into homogeneous groups based on their personalities, lifestyle, beliefs, and choices – as displayed across the channels, and so on. Brands can also leverage an insights-led strategy to forecast a customer’s future needs and personalize their offerings towards fulfilling those needs. Such an approach minimizes errors and yields better ROI. Above all, brands can delight customers with relevant, in-the-moment messages every single time. The Essential Steps To Implementing the Insights-led Segmentation To execute a strategy based on insights, one of the most important steps is for brands to tap into the existing data. This involves gathering and analyzing data at scale and in real-time. This data can be leveraged for reactive insights and predictive capabilities to support marketers forecasting their customer preferences and being ready to engage at the right time. Secondly, it is essential to invest in the right technology to make the most available data. Invest in a comprehensive engagement stack that helps marketers build deeper analytics for quick segmentation without the need to code or rely on other teams. It is essential to consider that both the consumers and the technology for consumer engagement are evolving. Marketers need to stay on top of the trends and quickly shift to newer channels and touchpoints as their customers do. It is also essential to remain agile by analyzing customer behaviour across channels. Identifying consumers’ micro-moments across different channels can throw light on their wants and needs, and the organization can predict and work upon those. Once brands have their work set out for them, they should harness the power of personalization. This includes personalizing emails, promotional materials, targeted ads, or recommendations. This helps set brands apart from their competitors and become more customer-centric. Finally, having an omnichannel approach while reaching out to customers across the various channels they may move to and generate maximum engagement. Along with this, loyalty programs should also be channel agnostic to retain most customers. We have just touched the tip of the iceberg of what insights-led segmentation has to offer. Brands that have implemented this approach have witnessed faster growth in their customer engagement. The buyer behaviour is constantly evolving, making it integral for brands to innovate faster and keep up with the trends to stay in step with their customers and become the brand they can relate to. This article is written by a member of the AIM Leaders Council. AIM Leaders Council is an invitation-only forum of senior executives in the Data Science and Analytics industry. To check if you are eligible for a membership, please fill the form here.","excerpt":"Imagine this scenario- a well-packaged mobile application or web page with an attractive interface but no customer traction. Unfortunately, this is the reality for many consumer brands today.  Despite packaging their products and services in a ‘cool’ manner, very few customers spot or use those products. The reason is, for a long time,  organizations have […]","categories":["AI Features"],"tags":["Customer segmentation"],"author_name":"Nalin Goel","publish_date":"2021-11-29T12:00:00","publication_year":"2021","word_count":874,"keywords":["data science","Go","Customer segmentation","AI","Git","RAG","Aim","ViT","analytics","GAN","R"],"extracted_tech_keywords":["AI","data science","analytics","Aim","RAG","R","Go","Git","GAN","ViT"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/council-post-the-big-refresh-insights-led-segmentation-is-the-next-normal\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10094643,"title":"Apple is Turning Everyone into Digital Zombies","content":"With the launch of Apple Vision Pro at WWDC 2023, CEO Tim Cook has lobbed the ‘dangerously enticing’ forbidden fruit at us again. Only this time, upon taking a bite from the Apple, you will be sent to a mixed reality that you probably should not be in. “Mac introduced us to personal computing and iPhone introduced us to mobile computing, Apple Vision Pro will introduce us to spatial computing,” said Cook during the announcement. The point is, with the introduction of phones, people were already getting detached from the ones around them, though being connected with the world. This new innovation will take this detachment to a whole new level. Welcome to the era of spatial computing with Apple Vision Pro. You’ve never seen anything like this before! pic.twitter.com\/PEIxKNpXBs— Tim Cook (@tim_cook) June 5, 2023 As a 100 year celebration of Disney, Apple has also announced its partnership with the company to allow people “to look through instead of just looking at it”, according to Cook. So, now you can sit awkwardly in a flight and place your movie screen on the face of the person sitting next to you, and enjoy the 3D space. No one would know or care. Now everyone’s dreaming Amid all the hype around AI at Google and Microsoft’s conference, Apple decided to not hop on the wagon. Instead of building a hallucinating chatbot, Apple decided to make people hallucinate in real-life. Imagine, sitting in an empty house and staring at oblivion. This gloomy scenario is what Apple has been trying to sell. No doubt a technological marvel, but comes with a lot of things that would make people refrain from it. Moreover, looking at the price point, there is a reason why the headset’s shape is like a kidney. The common theme around Apple’s product launch was to ensure that the users are not scared that they would lose touch with the humanity around them. Constantly, Cook and team used words like, “not losing track of reality”, to ensure people do not fear the loss of what is real after this. The truth is, in fact, the opposite. Ever since the release of Apple Vision Pro, Twitter is filled with people comparing the technology with and panicking about the present becoming Netflix’s Black Mirror or perhaps the world in WALL-E. It appears that Apple imagines the world to roam around wearing ski goggles. anyone else getting black mirror vibes rn with the apple VR headsetpresentation?— DeadProxy (@DeadProxy) June 5, 2023 Though this makes complete sense if someone is focusing on immersion and controlling everything around without holding anything in their hands. We all know that “eyes are the windows to your soul”, and now Apple is scanning our eyes and probably staring into our souls. Stay away from my eyes Is Apple preparing us for a future pandemic? We already experience separation anxiety with the small screen we keep in our pockets. Only now, the screen will be in front of our eyes — All The Time. You wouldn’t want to meet your friends in real life because it would feel like they are in the room with you when they are not. Matt Birchler said in his blog, “As soon as you’re interacting with someone in person, the headset is coming off. Wearing this thing at a kid’s birthday party will be seen as borderline psychotic behavior.” We saw the video of the product and how a dad is interacting with Vision Pro instead of playing with his daughter, and distinctively kicks the ball towards her. Either it was a marketing marvel which was supposed to be heartwarming, but it just felt dystopian. Apple was making the point that this technology will not suck you out of reality, and was trying very hard to prove it. But to some extent, it failed to do so. At the conference, Apple spoke about screen time and how the new WatchOS 10 will track and tell you if it gets too much. On the other hand, attaching a screen 5 cm away from your eyes would not damage your eyes? Moreover, if Apple now has the ability to track your eye movement all the time, the data can be used for advertising and possibly control. Imagine ads popping up in front of your eyes instead of your phone. Pretty big invasion of privacy as well. When Google revealed its Glass in 2013, the people who were actively wearing it were termed “glassholes“. It died. Now with Apple’s Vision Pro users, there should be a new term – “vision poopers“. The prediction is just this — Vision Pro probably wouldn’t be an iPhone moment from Apple. It would just be a gadget that few people would buy, use for a while, and probably not linger on for too long. That explains why even Cook was not wearing the headset at the event.","excerpt":"Vision Pro is one of the best inventions in a while that only rich psychotics will buy","categories":["AI Features"],"tags":["tim cook"],"author_name":"Mohit Pandey","publish_date":"2023-06-07T14:00:00","publication_year":"2023","word_count":816,"keywords":["Go","ELT","programming_languages:R","AI","innovation","programming_languages:Go","tim cook","R"],"extracted_tech_keywords":["AI","R","Go","ELT","innovation","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/apple-is-turning-people-into-digital-zombies\/","complexity_score":2,"technical_depth":7,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":41444,"title":"Big Story: What Has Modi 2.0 Done To Boost Emerging Tech After One Month In Power?","content":"For our Big Story of the month, Analytics India Magazine takes a look back at Prime Minister Narendra Modi’s emphatic electoral win and how the NDA-led government is pushing India’s tech ambitions forward. After their momentous victory last month in the Lok Sabha elections, Modi and the Bharatiya Janata Party (BJP)-led government has been busy making announcements in the tech sector. Over their 2014-19 tenure, the Modi government has put enough importance towards the technology sector by starting various new programmes and setting up an AI Task Force to prepare India for the upcoming Industrial Revolution 5.0. Through a series of events and speeches, Prime Minister Narendra Modi has been deliberately showcasing India as well as his government as technologically-forward. What BJP Promised The BJP has been giving a great deal of importance to emerging tech and had, in fact, proposed various programmes and new developments in their Lok Sabha election manifesto or the Sankalp Patra. We had done a detailed review and had compared the manifestos of India’s two biggest political parties. Well before the BJP won the Lok Sabha Elections earlier, PM Modi had asked his party office, the NITI Aayog and Principal Scientific Advisor to prepare the agenda for the first 100 days after the for the next government. In its manifesto, the BJP had said that the micro, small and medium enterprises would be benefitted by emerging tech. They had said, “Our [BJP] government has taken a major step in expanding ‘Technology Centres’ and we would achieve more than 150 such centres all over the country by 2024. These Technology Centres would help in mentoring skilling and prototyping of MSMEs. They will expose MSMEs to artificial intelligence, robotics, internet of things, virtual reality, blockchain technology and zero defect zero effect.” “The supercomputer, AI and quantum missions will make India not only Industry 5.0 ready but also a leader… Now we will get the latest technologies to transform Waste to Energy and Wealth in a major mission. We will also strive towards semiconductor manufacturing within the country,” they had added. Well before the BJP won the Lok Sabha Elections earlier, PM Modi had asked his party office, the NITI Aayog and Principal Scientific Advisor to prepare the agenda for the first 100 days after the for the next government. In this light, Analytics India Magazine is listing down the major developments kick-started by the Government in just over a month: Focus On Agri Tech Stressing on the term agri-scientists, BJP had earlier said that they would enable the development of young minds to take advantage of artificial intelligence, machine learning, blockchain and big data analytics for more predictive and profitable precision agriculture. @narendramodi\/Twitter Now, Finance Minister Nirmala Sitharaman is set to announce the Centre’s plans to infuse ₹25 lakh crore in the agricultural sector using AI and big data. Reports have suggested that these funds will be generated in association with the private players and will be used over a period of five years. “In order to enhance agriculture productivity, an investment of ₹25 lakh crore will be made in the coming years” Sitharaman is expected to make the announcement in the Union Budget for 2019-20 which will include new policies around the extensive use of big data, AI and blockchain technology as means to enable the agriculture sector to get faster and accurate information about weather, prices and other variables. Other tech initiatives like a mobile app-based system for direct marketing by farmers, a fund for fish and aquatic farming, a village storage scheme of agri-produce and a special plan for using uncultivated rural land for solar farming are also reportedly going to be included in the announcement. President Ram Nath Kovind had talked about the same earlier this month. He had said, “Large scale investments have been made to strengthen rural India. In order to enhance agriculture productivity, an investment of ₹25 lakh crore will be made in the coming years.” Double Farmers’ Income by 2022 Using Emerging Tech Earlier this week, the Government announced that they have set a target of doubling the Indian farmers’ income by 2022. The Centre also established an inter-ministerial committee to examine farmers’ issues to recommend a strategy to achieve the goal. The committee also said that they would be focussing on emerging technology and asserted that it was the only way to achieve a swift transformation in the Indian rural diaspora. Most changes and recommendations by the committee will include AI, big data analytics, Blockchain and IoT among other emerging technologies. Some of their other tech interventions are: Development of Kisan Suvidha mobile app Development of mKisan Portal for sending advisories on the various crop-related matter to the registered farmers through text messages Launching of e-National Agriculture Market initiative to provide farmers with an electronic online trading platform Use of  space technology for various programmes and areas such as Forecasting Agricultural Output using Space, Agrometeorology and Land-based Observations project, Coordinated programme on Horticulture Assessment and Management using geo-informatics project, National Agricultural Drought Assessment and Monitoring System, Rice-Fallow Area Mapping and intensification, geo-tagging of infrastructure and assets created under Rashtriya Krishi Vikas Yojana and Crop Insurance Using ML along with other algorithms for crop classification and area estimation Japan Fund-Of-Funds In a move to accelerate tech development in the continent, India and Japan recently launched a new $187 million (₹1,298 crore) fund-of-fund to invest in over 100 tech startups. This fund is to focus on ventures related to AI and IoT. Of the total amount, ¥15 billion or $140 million will be raised from Japanese investors while the remaining will be raised from Indian investors. This move is a part of a collaborative venture by the two countries floated in October 2018, where India and Japan agreed to work extensively in the areas of business, human resources and technology. Last year, Japan, in association with India’s Nasscom and IIM Bangalore had launched the Indo-Japan Startup Hub in Bengaluru to support Japanese companies and their local operations. The Hub was conceptualised as part of a joint statement signed between the Ministry of Economy, Trade, & Industry (Japan) and Ministry of Commerce & Industry (India) in May 2018. Other Smaller Announcements Ministry of Finance: Anurag Thakur, the Minister of State for Finance and Corporate Affairs, held a pre-budget consultation meeting to discuss the health of Digital India. Their key talking points were: Use of Big Data technology in improving the forecasting of economic, financial, climatic, and other phenomena by analysing large data sets Use of big data for SMEs Unleashing the power of big data for public governance Digital infrastructure and the role of the Government in the same Regulation of digital economy, especially in privacy, consumer protection and financial regulation Software as a service (SaaS) Interaction with IT experts: The Fifteenth Finance Commission earlier this month discussed the possibilities of: Extending the Direct Benefit Transfer Technology to Electricity subsidies and Water subsidies Digitisation of stamp duty and other digital payments Building of digital infrastructure Use of AI for development and reduction in the technology gap among states PM Modi asking to link Yoga with AI for wider outreach: Speaking at the International Yoga day on 21 June, the Prime Minister said, “Just as the software of our mobile phones keeps getting updated, similarly, we need to update the world with new information on Yoga. Therefore, it is essential that we do not keep Yoga limited. Yoga must be linked with medical, physiotherapy, artificial intelligence. Moreover, we have to encourage the private enterprise spirit associated with yoga; only then, we will be able to expand yoga.”","excerpt":"For our Big Story of the month, Analytics India Magazine takes a look back at Prime Minister Narendra Modi’s emphatic electoral win and how the NDA-led government is pushing India’s tech ambitions forward. After their momentous victory last month in the Lok Sabha elections, Modi and the Bharatiya Janata Party (BJP)-led government has been busy […]","categories":["AI Features"],"tags":["Agriculture","Big Data","BJP","International Affairs","Narendra Modi","NITI Aayog","what is power bi"],"author_name":"Prajakta Hebbar","publish_date":"2019-06-29T02:49:44","publication_year":"2019","word_count":1265,"keywords":["Go","what is power bi","artificial intelligence","machine learning","AI","TPU","ML","Agriculture","Git","RAG","BJP","analytics","Narendra Modi","International Affairs","Big Data","R","NITI Aayog"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","analytics","RAG","TPU","R","Go","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/big-story-what-has-modi-2-0-done-to-boost-emerging-tech-after-one-month-in-power\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10050567,"title":"Kerala Based AI Startup Uncanny Vision Acquired by Cloud Video Surveillance Platform Eagle Eye","content":"Eagle Eye Networks, one of the leading cloud video surveillance platforms, recently announced the acquisition of Kerala based artificial intelligence startup Uncanny Vision, accelerating the company’s leadership in providing AI and analytics. The Austin based firm aims to make customers’ businesses more efficient and the world a safer place. This acquisition includes research and development capabilities and a new regional office in Bangalore, India. The deal accelerates Eagle Eye’s plan, announced in November 2020, when Eagle Eye raised funds from venture capital firm Accel to dramatically reshape video surveillance. “After evaluating more than a dozen AI companies, we began working with Uncanny Vision in 2020. It didn’t take long for us to conclude that Uncanny Vision is the clear leader in surveillance AI. Their award-winning AI technology is deployed across thousands of locations, including Fortune 500 customers. Uncanny Vision’s tools for training, building, and optimizing models and its infrastructure for managing and storing training data are superior. This is a key component we’re excited to add to the Eagle Eye Networks platform,” said Dean Drako, CEO, Eagle Eye Networks. Ranjith Parakkal, co-founder, Uncanny Vision, said, “We share the Eagle Eye team’s vision to deliver advanced, cyber-secure AI cloud video surveillance offerings that transform video surveillance for businesses around the globe”. Uncanny Vision co-founders Ranjith Parakkal and Navaneethan Sundaramoorthy started as low-level hardware DSP developers and took time to understand the space and its needs. All 60 Uncanny Vision employees will be retained, and Eagle Eye plans to expand the Bangalore office. Eagle Eye is committed to supporting Uncanny Vision’s current customers and will continue to build its global infrastructure to provide the very best 24\/7 support to its valued customers around the world. Uncanny Vision’s deep learning algorithms enable recognition, identification, and prediction, improving business operations, customer service, and site safety. Their award-winning AI is used today in multiple applications, including Smart parking, ATM monitoring and retail analytics, to name a few.","excerpt":"This acquisition includes research and development capabilities and a new regional office in Bangalore, India.","categories":["AI News"],"tags":["AI (Artificial Intelligence)","Data Science","Data Scientist","Deep Learning","Facial Recognition","Google","Machine Learning","monitoring","networking","Python","smart devices"],"author_name":"Victor Dey","publish_date":"2021-10-07T11:58:24","publication_year":"2021","word_count":322,"keywords":["API","networking","smart devices","deep learning","R","artificial intelligence","analytics","Data Science","startup","Go","Facial Recognition","monitoring","AI","Machine Learning","GAN","Python","Aim","Google","Deep Learning","Data Scientist","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","deep learning","analytics","Aim","R","Go","API","GAN","startup"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/kerala-based-ai-startup-uncanny-vision-acquired-by-cloud-video-surveillance-platform-eagle-eye\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10103455,"title":"OpenAI Did The Right Thing by Firing Sam Altman","content":"And now he’s back. But while we are still trying to wrap our heads around the recent OpenAI drama, the board has and still is the decision-maker in the whole scenario. Despite holding a position as a board member, the ease with which Sam Altman was ousted was facilitated by a peculiar board structure crafted by OpenAI, and going by the motive for that structure, Altman’s firing was justified. While the latest development speaks about the tussle between Altman and the independent board members, the situation that led to this state is solely attributable to the not-for-profit board that holds authority over everyone. It is almost ominous that Altman knew he is not immune to firing and even spoke about it in a Bloomberg interview a few months ago, where he said that the board can fire him and believed that the board “needs to be democratised to include all of humanity”. With the current structure of the board, any conflict that happens against the motive of the non-profit wing can lead to drastic measures which raises the question of what went wrong or what the board saw that was against the interest of the company. With the vision of rapid growth and AGI focus, safety became a topic of contention. It is possible that the steps to achieve this conflicted with the board’s vision, which automatically puts the CEO under fire. Furthermore, the complex board structure made the removal process effortless. The Complex Board OpenAI, which started as a non-profit AI research company in 2015 with the goal of advancing digital intelligence to benefit humanity without the need to generate a financial return, created an OpenAI LP, a capped-profit company in 2019. The vision of this company was to build safe AGI for the world, and believed that the capped-profit format would allow them to rapidly increase their investments in compute and talent. OpenAI LP was structured in a way that allows it to attract funding with the promise of capped financial return for investors and employees. The company’s capped-profit approach limited first round investors to a maximum profit of 100 times their initial investment (100x). Furthermore, any excess value generated will go back to the nonprofit entity to support its mission-driven goals. However, all of this is controlled by one board irrespective of the profit motive. OpenAI Structure. Source: Substack (Chamath Palihapitiya) Supreme Board Power OpenAI’s board of directors sits right at the top of the structure that controls both the nonprofit and for-profit bodies. The ‘capped profit’ company (OpenAI Global LLC) controls ChatGPT in which Microsoft has invested billions of dollars. Interestingly, the company has built a structure where the nonprofit wing takes precedence over any obligation to generate a profit. It is also fair to say that OpenAIGlobal has no control over itself, rather controlled by other entities. The unique hierarchy also limited investors’ say in the running of the company as their contribution also lies towards the bottom of the structure. In Altman’s ousting episode, there was a bigger catch. Though Altman was part of the board, and had initially invested in the company via YCombinator, he had zero equity in the company – something he had confessed in the AI Senate and even joked about after his firing. Having no equity in the company also meant that he has no separate power to influence the board in the likelihood of something going wrong. Altman’s commitment to OpenAI, ‘doing what he loves’, pretty much did nothing to secure his position there. While there are options such as dual-class share structures that can secure founders from being ousted from their companies (except for actions such as fraud, misbehaviour, and grave issues), none of this was implemented at OpenAI. It is also possible that the accelerated growth that OpenAI witnessed from last November with the launch of ChatGPT was unexpected. As Altman himself had mentioned long ago, the company did not expect this kind of growth when they launched ChatGPT, which could also possibly hint at why there was no planning or restructuring of the board. For instance, Microsoft who has invested billions of dollars into OpenAI has no board seat. The Inevitable Move While the reason for ousting Altman was attributed to him not being candid with the board, thereby hindering their ability to exercise responsibilities, the decision was delivered by four members from the 6-member board. Out of the four, only Ilya Sutskever was a cofounder and the other three were independent directors with no equity and minimal experience in corporate governance. Pretty much, hinting at how the OpenAI board structure was faulty from the beginning. While many questions remain unanswered, a weak board structure toppled the biggest AI company in the world and even put the fate of the company at risk. An expensive lesson in corporate governance.","excerpt":"The board was right all along, and Altman knew.","categories":["Global Tech"],"tags":["AGI","AI Senate","board","ChatGPT","Google Meet","Greg Brockman","Ilya sutskever","investors","OpenAI","Sam Altman"],"author_name":"Vandana Nair","publish_date":"2023-11-22T11:13:29","publication_year":"2023","word_count":804,"keywords":["Go","ChatGPT","API","board","Sam Altman","OpenAI","AI","Google Meet","funding","investors","AI Senate","Greg Brockman","Git","GPT","ViT","AGI","Ilya sutskever","R"],"extracted_tech_keywords":["AI","ChatGPT","OpenAI","R","Go","Git","API","GPT","ViT","funding"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/why-openai-did-the-right-thing-by-firing-sam-altman\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":20444,"title":"Inside SingularityNET That Aims To Create Decentralised Marketplace For AI","content":"SinguarityNET is the recent talk of the town and is turning out to be probably one of the most popular concepts of all times. At a time where everything is run on AI, it is providing a platform for anyone to monetise AI, allowing companies, organisations and developers to buy and sell AI at scale. What does this mean and how does SingularityNET aim to create a decentralised marketplace for AI? Let’s find out. From the creators of Sophia, who are opening AI market to the entire globe If you are a follower of recent trends in AI and robotics, the chances are feeble that you might have missed the brouhaha around Sophia. From being granted the first ever citizenship to a humanoid by Saudi Arabia to be claimed the most expressive robots of all times, Sophia didn’t miss to grab an eyeball from across the globe. SingularityNET who’s artificial intelligence technology powers this robot, along with Hanson Robotics, who built it, have ambitious plans to revolutionise the way AI works. The brainchild of AI researcher Ben Goertzel, SingularityNET aims to be a decentralised open market for AI, meaning that it will become a bridge between AI developers and business users, allowing larger sectors to have access to AI. With AI being implemented across major facilities, the relevance of SingularityNET increases even more. Despite gaining immense popularity, Goertzel believes that the area of artificial intelligence remains largely fragmented. SingularityNET aims high to overcome this challenge by allowing everyday users, from small business owners to brand managers to connect with AI and create perfect solutions. It would let small teams to pick the right technology and in a cost effective manner. As their website explains, SingularityNET enables AI-as-a-service on a permissionless platform, so that anyone can use AI services easily. SingularityNET is AI coalescing blockchain technology As we have known by now that SingularityNET aims to become the key protocol for networking AI and machine learning tools to form a coordinated artificial general intelligence, it must do so by powering SingularityNET platform by a utility token that is tailored to enable its core functionality. Thus comes the AGI token that is structured to enable four key mechanisms—transactions, settlements, incentives and governance. The notion of SingularityNET that has been in Goertzel’s mind for a while took an assumption with Bitcoin introducing the blockchain technology. By building SingularityNET on the blockchain, he aims to create a decentralised system optimised for ease, reliability, and security. Powered by AGI token, that makes decentralised economy possible, it lets quantify the value of an AI agent. Aiming to build a common infrastructure that benefits all in the area of AI, it is an open source protocol that collects smart contract for a decentralised market of coordinated AI services. “Blockchain technology provides an ideal tool for managing network transactions on SingularityNET due to its transactional and bookkeeping advantages. But a blockchain-based framework needs to be designed to allow AI Agents to interact with each other and external customers”, mentioned the company’s blogpost. It also mentioned that AGI token allows network users to transact on the market place from thousands of contracts, technologies and protocols. “All transactions occur on a single, publicly-auditable transaction ledger”, it mentioned. It is built on the OpenCog platform, and is currently being used by more than 50 companies such as the likes of Huawei, Cisco, and others. Benefits of having SingularityNET in place A solution that is foreseeing creation of “AI for all” and capture of a $3 trillion economy by 2024, has something in store for everyone—from a customer point of view, AI technology point of view and from a business perspective. From a customer view, SingularityNET proves to be an easier, broader and powerful way for a business to obtain AI services. It can be used as a back end of multiple consumer services and provide AI services that are superior to those of competing platforms, whereas from an AI technology point of view, it is a heterogeneous and multi-paradigm framework. It would allow understanding various tools such as neural nets, evolutionary learning, probabilistic reasoning, statistical pattern recognition, amongst others. When it comes to business structure perspective, it would form an entirely new socioeconomic structure in the form of decentralised self organising cooperative, allowing everyone to make key decisions based on it. SingularityNET is being looked forward by many as a way to create extraordinary impact in the way AI is being used as a next stage of growth. From creating more powerful and intelligent solutions to providing an easy accessibility in a cost effective manner, it aims high at revolutionising AI market in the coming future.","excerpt":"SinguarityNET is the recent talk of the town and is turning out to be probably one of the most popular concepts of all times. At a time where everything is run on AI, it is providing a platform for anyone to monetise AI, allowing companies, organisations and developers to buy and sell AI at scale. […]","categories":["IT Services"],"tags":["decentralised ai"],"author_name":"Srishti Deoras","publish_date":"2018-01-09T08:50:41","publication_year":"2018","word_count":775,"keywords":["Go","artificial intelligence","machine learning","AI","RAG","Aim","decentralised ai","AI agents","GAN","AI research","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","Aim","RAG","R","Go","GAN","AI research","AI agents"],"url":"https:\/\/analyticsindiamag.com\/it-services\/inside-singularitynet-aims-create-decentralised-marketplace-ai\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":35619,"title":"Can AI Be An Effective Therapist?","content":"Over 450 million people are currently affected by mental or neurological disorders and it is estimated that one in four people will be affected by such condition in the coming years. With the rapid advancement in technology and its application in the medical field, researchers and medical practitioners are now looking at ways in which artificial intelligence and machine learning can be leveraged to detect early symptoms and potential cure for various mental illness. Over the years, considerable advancements have been made in this regard and AI-powered solutions such as NLP and even chatbots have been designed to understand the human mind. We look at ways in which these solutions are helping psychiatrists and other mental health professionals deliver their job better and the potential harm associated with these technologies. Virtual Therapist Several startups have combined AI and virtual reality to create a virtual therapist that can interact with is patients in real-time. Quartet Health is one such startup which uses machine learning capabilities to identify patients with a mental health condition and provide a customised treatment plan based on their medical history and behavioural pattern. The platform partners with health plans and systems to facilitate access to personalised care and also enable virtual collaboration between patient and trained specialists. In another niche use case Ellie, a virtual therapist was created by University of Southern California’s Institute for Creative Technologies and was funded by the US government to treat veterans suffering from PTSD. The virtual assistant could analyse facial expressions, head gestures, eye gaze direction and voice quality to identify behavioural indicators related to depression and post-trauma stress. Disadvantage: In conditions like PTSD, where a person can be overwhelmed by symptoms like flashbacks, nightmares and severe anxiety, the need for personal assistance and direct human intervention is more. While the technology can ensure a certain degree of anonymity, it will fail to replace human intervention which is crucial for a person in distress. AI-Powered Genetic Counsellor The profession involves advising individuals and families at the risk of a genetic disorder by helping them understand the condition better and provide them with the much need mental support. Increasingly, more people are seeking the help of these professionals to better understand the genetic composition to predict whether they are likely to develop conditions like anxiety, schizophrenia and bipolar disorder. Health practitioners and startups like Clear Genetics and OptraGURU have already developed AI software which can provide the same services as that of a genetic counsellor. AI has been leveraged in genome sequencing to spot disease marker in patients and even to make a personalised drug treatment plan for patients. Disadvantage: While cost reduction has been associated with the technology, the biggest threat that is likely to occur is that of hacking and data theft. Healthcare data breaches have been occurring and crucial information like a patient’s medical history and other personally identifiable data can be grossly misused. Since AI systems are prone to vulnerabilities, it can also lead to inaccurate disease detection and false recommendation of drugs. Chatbots For Depression As the gap between the availability of mental health professional and the cost of each therapy session keeps increasing, the demand for digitised healthcare solutions has increased steadfastly. Even in their low point, more and more people are picking app experience over the real therapy session. Due to this burgeoning demand, India and UK-based healthcare startup, Touchskin introduced Wysa, its AI-powered chatbot. The app has over half a million downloads and provides its users with features like guided and unguided meditation, reminders via message and progress tracking. Yet another popular name is Woebot, a chatbot that uses Cognitive Behavioral Therapy (CBT) tools for depression. The app sends over a million messages per week to help its users deal with issues related to depression, anxiety, relationship problems, procrastination, loneliness, grief, addiction, pain management and more. The app was rated among the best app for anyone with anxiety. In a testimonial about the app in their website, one of its users writes, “In my first session with Woebot, I found it immediately helpful…addressing my anxiety without another human’s help felt freeing.” Disadvantage: Though the biggest advantage attributed to these apps is the instant availability of someone to talk with you, the biggest drawback it faces is the lack of human touch. “Sorry, this is just way too weird for me. I feel like I’m talking to a human who I don’t know and it’s making me very anxious and paranoid. It would be cool if it really is a human typing, but I would have to get to know that person first. Just not my type,” a Wysa user remarked. Outlook While technology has the potential to disrupt and shape the future of precision healthcare, its application in the field of mental health and psychology needs to be carefully looked at. Even though machines and algorithms can mimic human emotions in speech and visual format, it is a long road ahead before completely relying on AI and ML capabilities in a field like psychological counselling, which require more human intervention than machine. With machine learning algorithms being more susceptible to biases in the form of racist and sexist remarks, key industry players need to ensure robustness in their AI system before bringing it out to the market. Finally, companies have a bigger responsibility for ensuring the safety and security of patients’ personal data.","excerpt":"Over 450 million people are currently affected by mental or neurological disorders and it is estimated that one in four people will be affected by such condition in the coming years. With the rapid advancement in technology and its application in the medical field, researchers and medical practitioners are now looking at ways in which […]","categories":["AI Features"],"tags":["AI for mental health","meditation","virtual assistant"],"author_name":"Akshaya Asokan","publish_date":"2019-03-01T08:25:43","publication_year":"2019","word_count":899,"keywords":["meditation","Go","artificial intelligence","machine learning","AI","chatbots","ML","Git","AI for mental health","NLP","RAG","virtual assistant","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","NLP","RAG","chatbots","R","Go","Git"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/can-ai-be-an-effective-therapist\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":3,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":false},{"article_id":10018097,"title":"Google Trains A Trillion Parameter Model, Largest Of Its Kind","content":"Google has developed and benchmarked Switch Transformers, a technique to train language models, with over a trillion parameters. The research team said the 1.6 trillion parameter model is the largest of its kind and has better speeds than T5-XXL, the Google model that previously held the title. Switch Transformer According to the researchers, the Mixture of Experts (MoE) models, despite being more effective than other deep learning models, face issues due to their complexity, lack of accessibility, and computational costs. As opposed to the traditional models that use the same parameters for all the inputs, MoE selects different parameters for each input. While we may get a sparsely activated model with MoE, it leads to having a massive number of parameters, resulting in disadvantages discussed above. Google researchers have developed Switch Transformer to create a system that would increase the parameter count while maintaining the floating-point operations (FLOPS) per input constant. It does that by using only a part of the model’s weight or parameters to input data within a model. The Experiment The Switch Transformer is based on T5-Base and T5-Large models. In the T5 model (introduced by Google in 2019), all the NLP tasks are unified into a text-to-text format where both the input and output are always text strings. In addition to the T5 models, Switch Transformers use hardware initially designed for dense matrix multiplication, and also used in language models such as GPUs and TPUs. The researchers established a distributed training setup for the experiment, and the models split unique weights into different devices. While the weights increase in proportion to the number of devices, the memory and the computational footprint of each device remains manageable. Switch Transformer models, using 32 TPUs, were pretrained on the Colossal Clean Crawled Corpus — a 750 GB dataset composed of text snippets from Reddit, Wikipedia, among others. For the experiment, the Switch Transformer models were used to predict missing words in passages where 15% of the words were masked. Other challenges included language translation and answering a series of tough questions. Performance of Switch Transformer Model The researchers claimed that the model performed better than the smaller T5-XXL model with 400 billion parameters. Further, the new model didn’t manifest any training instability. The Switch Transformer also showed marked improvement in delivering downstream tasks. The model maintained seven times higher pretraining speed while using the same amount of computational resources. On the translation front, the Switch Transformer model, which was trained to translate between 100 languages, did so with four times the higher speed for at least 91% languages. However, the model’s performance was unsatisfactory compared to the baseline model on the Sanford Question Answering Dataset (SQuAD). The researchers chalked it up to “a poorly understood dependence between fine-tuning quality, FLOPS per token and number of parameters.” Looking Forward The current approach falls within the adaptive computation algorithm that uses identical and homogeneous parameters. In the future, the researchers hope to support heterogeneous parameters, facilitated by a more flexible infrastructure. Further, the Switch Transformer will be applied across different modalities as well as multi-modal networks. Like other large language models such as GPT-2, GPT-3, BERT, RoBERTa, Switch Transformer is also susceptible to biases. Such biases may result in the spread of misinformation, phishing, abuse of legal and governmental processes, and radical social engineering etc.","excerpt":"Google has developed and benchmarked Switch Transformers, a technique to train language models, with over a trillion parameters. The research team said the 1.6 trillion parameter model is the largest of its kind and has better speeds than T5-XXL, the Google model that previously held the title. Switch Transformer According to the researchers, the Mixture […]","categories":["AI Trends"],"tags":["GPT-3","Language Models","Parameters","Switch Transformer"],"author_name":"Shraddha Goled","publish_date":"2021-01-14T16:00:00","publication_year":"2021","word_count":554,"keywords":["GPT-3","Go","TPU","AI","Switch Transformer","Transformers","BERT","NLP","GPT","Aim","deep learning","Parameters","Language Models","R"],"extracted_tech_keywords":["AI","deep learning","NLP","Aim","Transformers","TPU","R","Go","BERT","GPT"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/google-trains-a-trillion-parameter-model-largest-of-its-kind\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":10102590,"title":"Testing the Waters: What’s Bubbling with Underwater Data Centres?","content":"In 2018, Microsoft worked on a research project — Project Natick — to test the viability of subsea data centres that will be powered by offshore renewable energy. After two successful years, the 40-foot-long, 12 rack, self-contained underwater data centre was retrieved from the North Sea for analysis. With the successful deployment of this project, the concept of underwater data centres is no longer just a possibility. Microsoft Sets The Stepping Stone Project Natick proved that in addition to providing cooling efficiency, it is possible to implement environmentally-friendly and sustainable power solutions for data centres. Microsoft is now working on the second phase of the research with Project Natick Phase 2, where the aim is to demonstrate the capability of manufacturing it at production scale. Microsoft’s Project Natick Plan.  Source: Dgtl Infra Since data centres have a high energy consumption for normal functioning, with GPUs demanding even more, underwater data centres provide respite to the energy problem to a large extent. The ocean bed being cooler helps nullify the heat produced by the servers, thereby providing natural cooling at zero cost, and the heat is dissipated into the surrounding environment. Furthermore, it even addresses the problems of low-latency connectivity issues, caused by the long distance between location source and destination. While the results of the ambitious project are promising, the viability of underwater data centres still looms large. All Is Not Easy Underwater One of the biggest challenges to having underwater data centres is the tricky hardware maintenance. If any kind of problem arises in the server, sending a person to fix it or even lifting up the data centre to the surface is a burdensome task. Robotics could play an important part here to probably fix the issues, but, until then, human dependence will continue. Furthermore, network connectivity, which is best achieved via underwater network cables, carry huge expenses, with costs that can go as high as $400 per foot. While underwater data centres are considered an environmentally sustainable solution compared to land data centres, there are still a number of unanswered questions in terms of understanding how the oceans would react in the long run. Questions about the amount of heat emitted and its effect on the ocean bed and marine life when a number of data centres operate underwater, is still unknown. Furthermore, the question on the disposal mode for these data centres after they fail or outlive their usefulness, remains unresolved. Interestingly, power used for running these data centres can be obtained from renewable sources — something big tech companies are actively working towards. Big Tech’s Chase Sustainable Tech With big tech companies actively working on sustainable solutions for the longer run, the shift towards sustainable options for running data centres too is no surprise. Microsoft’s choice of the Orkney Islands for the Northern Isles deployment by the Project Natick team was influenced by the fact that the region’s grid is entirely powered by wind, solar energy, and emerging green energy technologies that are under development by the European Marine Energy Centre. They are even looking for ways to have an underwater datacenter with an offshore wind farm in close proximity. Microsoft’s attempt to use renewable energy for powering their projects is not new. Recently, the company was in news for looking to hire a program manager specialised in nuclear technology for heading their small modular reactors and microreactor energy systems, thereby exploring the possibility of harnessing nuclear energy to fuel its data centres. Furthermore, other Big tech billionaires such as Sam Altman, Bill Gates and Jeff Bezos have actively invested in this space. Exploration of underwater data centres is slowly gaining steam. Last year, Subsea Cloud announced three underwater data centre projects, where it will not be using pressure vessels, unlike Microsoft. Instead, it would be a conventional shipping container with pressure equalised on both sides. Earlier this year, Highlander in China built the first commercial underwater data centre near Hainan Island in China, and is looking to find global customers. With China also experimenting with the same, it is likely that it will become a global chase.","excerpt":"Microsoft started experimenting with underwater data centres years ago; now a few other companies are treading lightly along that path","categories":["AI Features"],"tags":["bill gates","China","data centre","energy","GPUs","jeff bezos","Microsoft","nuclear energy","Sam Altman"],"author_name":"Vandana Nair","publish_date":"2023-11-07T10:00:00","publication_year":"2023","word_count":683,"keywords":["Go","energy","Sam Altman","Microsoft","AI","programming_languages:R","programming_languages:Go","Aim","ViT","ai_applications:robotics","jeff bezos","bill gates","GPUs","nuclear energy","R","data centre","China"],"extracted_tech_keywords":["AI","Aim","R","Go","ViT","programming_languages:R","programming_languages:Go","ai_applications:robotics"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/whats-bubbling-with-underwater-data-centres\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10139498,"title":"Meta’s Llama 3.1 is the Missing Link for Indic Datasets","content":"Earlier this year, when Meta released the Llama 3.1 405B model, the updated version allowed developers to use outputs from Llama models—including the 405B—to improve other models. This benefitted developers and AI startups in India that were building Indic LLMs. Yann LeCun, Meta’s chief AI scientist, acknowledged the past issues developers encountered with Llama regarding model creation. Speaking at Meta’s Build with AI Summit in Bengaluru, LeCun said the company took note of it and addressed the concerns with its newer version. Sharing the stage with Yann LeCun in a fireside chat, Nandan Nilekani, influential Indian entrepreneur and co-founder of Infosys, said that this development will help Indian AI startups use LLMs easily and that it isn’t necessary for India to build LLMs from scratch. “Our goal should not be to create another LLM. Let the big players in Silicon Valley handle that,” Nilekani said. “India should become the use-case capital of the world and focus on building small models quickly.” He further added India will use it [Llama] to create synthetic data, build small language models quickly, and train them using appropriate data. Nilekani is partially right as using OpenAI’s GPT-40 or Anthropic’s Claude Sonnet 3.5 APIs can be expensive. In contrast, Llama 3.1 405B is freely available on Hugging Face and competes well with top foundation models like GPT-4, GPT-4o, and Claude 3.5 Sonnet. Nilekani opined that the correct approach for Indian AI companies is to create appropriate data. Notably, in 2022, he invested in AI4Bharat, a research lab dedicated to creating open-source datasets, tools, models, and applications for Indian languages. Llama Powers Sarvam AI Speaking at Cypher 2024, Vivek Raghavan, the chief of Sarvam AI, revealed that they used Llama 3.1 405B to build Sarvam 2B. He explained that it is a 2 billion parameter model with 4 trillion tokens, of which 2 trillion are Indian language tokens. Sarvam 2B is part of a class of small language models (SLMs) that includes Microsoft’s Phi series, Llama 3 (8 billion), and Google’s Gemma models. It serves as a viable alternative to using large models, such as those from OpenAI and Anthropic, while also being more efficient for specific use cases. “If you look at the 100 billion tokens in Indian languages, we used a clever method to create synthetic data for building these models using Llama 3.1 405B. We trained the model on 1,024 NVIDIA H100s in India, and it took only 15 days,” said Raghavan. Regarding Sarvam 2B, he further said that the model performs well on Indic tasks. “It is extremely good for summarisation in Indian languages, and for any kind of NLP task in Indian languages—this will outperform models that are much bigger.” The company recently launched its latest model, Sarvam-1, which outperforms Google’s Gemma-2 and Llama 3.2 on Indic tasks. The startup claims that its secret sauce is 2 trillion tokens of synthetic Indic data, equivalent to 6-8T regular tokens due to their super-efficient tokenizer. Sarvam AI is not alone. In a recent interaction with AIM, Meta VP Manohar Paluri revealed that even Ola Krutrim uses Llama. “Not just small companies or developers, even large companies build on top of this ecosystem, which really gives us confidence that we have the momentum.” He further added that people are now using Llama as their de facto intelligence layer to build their businesses on top of it.“We are actually trying to bring high-quality Indian tokens into Llama so that Llama will work in Indian languages,” he said. Paluri explained that since Llama is the engine for Meta AI, the AI company will support Indian languages, benefiting billions of people in India who use Meta AI on WhatsApp, Facebook, and Instagram. Meta AI boasts over 500 million monthly active users globally and is on track to become the most widely used AI chatbot by the end of 2024. Recently, it was launched in Hindi and the company plans to integrate multiple Indian languages into their future models. Exploring Llama 3.1’s Multilingual Capabilities In the Llama 3.1 research paper, Meta states that the model has been trained on multilingual data, generating high-quality instruction-tuning data for languages such as German, French, Italian, Portuguese, Hindi, Spanish, and Thai. Meta has collected high-quality, manually annotated data from linguists and native speakers. Moreover, with a context length of 128K tokens, it can process and generate longer, more complex pieces of text, which is beneficial for creating diverse synthetic datasets. “Synthetic data generation is one of the main use cases for these very large models. This can be extremely helpful in domains where obtaining a large, high-quality dataset is challenging due to cost, privacy concerns, or simply a lack of available data,” said Hamid Shojanazeri, ML engineer at Meta. Meta is not limited to LLMs. It intends to empower developers with the resources to create custom agents and discover new types of agentic behaviours.","excerpt":"We will use it [Llama] to create synthetic data, build small language models quickly, and train them using appropriate data: Nandan Nilekani","categories":["Global Tech"],"tags":["Meta AI","sarvam ai"],"author_name":"Siddharth Jindal","publish_date":"2024-10-26T14:04:04","publication_year":"2024","word_count":811,"keywords":["Anthropic","Meta AI","sarvam ai","AI","OpenAI","GPT-4o","ML","NLP","Aim","foundation models","Claude 3.5"],"extracted_tech_keywords":["AI","ML","NLP","foundation models","GPT-4o","OpenAI","Claude 3.5","Anthropic","Meta AI","Aim"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/metas-llama-3-1-is-the-missing-link-for-indic-datasets\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10137863,"title":"80% of Engineers Could Lose Their Jobs if They Don’t Upskill by 2027","content":"Gartner, the research and advisory firm, said in its latest report that GenAI will spawn new roles in software engineering and operations through 2027, requiring 80% of the engineering workforce to upskill. The report analyses the impact of AI engineering on the daily functioning of software development organisations and notes that it significantly increases productivity by performing smaller tasks and helping senior developers in big organisations. Philip Walsh, senior principal analyst at the research firm, elaborated on the need for human intervention in the development of complex software. Looking further, AI could automate more tasks, thus changing the working environment for current developers. Most of the mid-level coding would be done by AI, and humans would need to be prompt in senior-level engineering and retrieval-augmented generation (RAG). This technique combines the strengths of information retrieval and natural language generation (NLG). Looking at the long-term effect, AI will majorly integrate into the software industry, and AI engineering will play a pivotal role. “The AI engineer possesses a unique combination of skills in software engineering, data science and AI\/machine learning (ML), skills that are sought after,” says Walsh. Need for Human Intervention AIM reported earlier that the job market for software engineers has been highly affected by tools like Cursor, ChatGPT, and Claude. Though generative AI tools have contributed to completely transforming the process followed by software engineers, human creativity and innovative expertise remain essential in providing necessary and original inputs. Zahiruddin Tavargere, the senior principal software engineer who started exploring AI Engineering in early 2023, expressed, “Today I can use data engineering techniques, build ML models, leverage GenAI, and use MLOps techniques to solve business problems and push the solutions to production.” What do the Developers Think? A user on Reddit who runs an AI engineering consultancy expressed that if he interviewed someone who demonstrated a few AI products where they could clearly think about the product and not just engineering, he’d hire them quickly. Though this point of view holds much water, developers still argue about the actual definition of upskilling and how much it contributes in reality. Another Reddit user, however, doubted that developers will need new skills. “Working with AI-generated code is essentially maintenance coding, which is taking someone else’s codebase and fixing it,” the person commented in a Reddit post. As AI Engineering becomes more common, companies will have to adopt more AI-friendly developer platforms, which will, in turn, help these organisations integrate AI more effectively on a large scale.","excerpt":"Want to know why pursuing AI-based development techniques could help keep your job? Gartner says upskilling in AI Engineering is the key.","categories":["AI News"],"tags":["gartner","software engineer","upskilling"],"author_name":"Sanjana Gupta","publish_date":"2024-10-08T17:37:07","publication_year":"2024","word_count":413,"keywords":["data science","ChatGPT","GenAI","machine learning","upskilling","AI","ML","MLOps","RAG","software engineer","Aim","generative AI","gartner"],"extracted_tech_keywords":["AI","machine learning","ML","data science","generative AI","GenAI","ChatGPT","MLOps","Aim","RAG"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/80-of-engineers-could-lose-their-jobs-if-they-dont-upskill-by-2027\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":29221,"title":"IRCTC Debuts An AI-Powered Chatbot AskDisha For Customer Support","content":"The Indian Railway Catering and Tourism Corporation (IRCTC) has launched a digital interaction platform—Ask Disha to help its travellers with queries related to ticket booking, cancellation or other value added services offered by IRCTC. The AI-powered chatbot is aimed at improving the customer services of railway passengers. According to reports, the chatbot is expected to offer an improved and intuitive customer support by answering customer queries pertaining to all aspects of services provided by IRCTC. Some of the features that Ask Disha would have are: Ability to answer customer queries Perform multitasking Provide round-the-clock customer support Reduce the waiting time for queries to get answered Provide customers a stress-free experience The new feature by IRCTC has gone live on its next-gen e ticketing website from today and will soon be integrated on the IRCTC Rail Connect Android app. Available on bottom right-hand side, once the user starts typing a query, the chatbot is capable of even suggesting options. They also plan to make it voice-enabled to make it more functional. Some of the queries that Disha can answer are—what are different payment gateway options on IRCTC, what are ticket cancellation charges, what is confirmation probability among others. The team at IRCTC believe that the chatbot will learn over the time by improving its knowledge and be able to answer wider range of questions in the coming future. Piyush Goyal, the minister of Railways re-tweeted the post as below, confirming the launch of Ask Disha. ‘Ask Disha’ – Digital Interaction to Seek Help Anytime: Artificial Intelligence Chat Bot launched on IRCTC website enables passengers’ 24×7 customer query support & quick response time.https:\/\/t.co\/TruRCC1p5E pic.twitter.com\/sxTOAoYXrO — Piyush Goyal (@PiyushGoyal) October 14, 2018 The AI-based chatbot has been developed by IRCTC in association with CoRover Private Limited, a Bengaluru-based startup that builds AI and ML-based platform to provide safe, convenient & fun-filled travel experiences.","excerpt":"The Indian Railway Catering and Tourism Corporation (IRCTC) has launched a digital interaction platform—Ask Disha to help its travellers with queries related to ticket booking, cancellation or other value added services offered by IRCTC. The AI-powered chatbot is aimed at improving the customer services of railway passengers. According to reports, the chatbot is expected to […]","categories":["AI News"],"tags":["AI Chatbot","chatbot ai"],"author_name":"Srishti Deoras","publish_date":"2018-10-15T09:32:20","publication_year":"2018","word_count":310,"keywords":["Go","artificial intelligence","programming_languages:R","AI","AI Chatbot","ML","programming_languages:Go","Git","Aim","chatbot ai","R","startup"],"extracted_tech_keywords":["AI","artificial intelligence","ML","Aim","R","Go","Git","startup","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/irctc-debuts-ai-powered-chatbot-askdisha-for-customer-support\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10096143,"title":"God&#8217;s Take on Tech: Vatican Launches AI Ethics Think Tank","content":"In an interesting turn of events, Pope Francis, in collaboration with Santa Clara University’s Markkula Center for Applied Ethics, released a manual discussing the ethical aspects of AI. Together, they have established the Institute for Technology, Ethics, and Culture (ITEC), which serves as a Vatican-led think tank focusing on AI ethics to facilitate meaningful discussions about the impact of technology on humanity. This comes three months after the Pope got a fancy makeover in a Balenciaga puffer jacket, thanks to AI image generator Midjourney. OKAAYYY pic.twitter.com\/MliHsksX7L— leon (@skyferrori) March 25, 2023 Around the same, time major tech players, including Elon Musk, Steve Wozniak, and Evan Sharp, signed an open letter urging AI developers to temporarily halt the training of advanced AI experiments, expressing concerns about potential risks that could lead to a loss of control over the civilisation. It emphasised that the current rapid development of AI, exemplified by the likes of ChatGPT, Dalle-2, Midjourney, and voice-cloning softwares, may be irresponsible and fail to consider the consequences for society. The letter called for a pause in testing technologies more powerful than ChatGPT-4 for six months. Furthermore, it urged developers to establish robust safety protocols that involve rigorous audits and oversight from independent experts. Pope’s Call to Silicon Valley Moral Guidance is Not New The handbook titled ‘Ethics in the Age of Disruptive Technologies: An Operational Roadmap’ aims to help technology companies navigate the ethical considerations of AI. ITEC claims to have discussed with several AI and ML tech leaders in Silicon Valley before making the handbook. It covers various topics, with a focus on AI development, and emphasises the importance of upholding high ethical standards in the technology industry. When the Pope’s AI-generated fake images took social media by storm, he said he believes in the positive impact that AI and ML can create, however, “At the same time, I am certain that this potential will be realised only if there is a constant and consistent commitment on the part of those developing these technologies to act ethically and responsibly.” Even in 2019, Pope Francis spoke about AI misuse at “The Common Good in the Digital Age” conference in the Vatican. “If mankind’s so-called technological progress were to become an enemy of the common good, this would lead to an unfortunate regression to a form of barbarism dictated by the law of the strongest,” he told Reuters during the event. Soon, along with support from Microsoft and IBM, the Vatican City released a set of principles called the ‘Rome Call for AI Ethics‘, which promotes the responsible and ethical use of AI. Read more: The role of ‘God’ in the ‘Matrix’","excerpt":"Pope Francis and Santa Clara University’s Markkula Center released an AI ethics manual. They founded ITEC, a Vatican-led think tank promoting meaningful discussions on the ethical impact of technology.","categories":["AI News"],"tags":["AI Tool"],"author_name":"Shritama Saha","publish_date":"2023-07-03T12:57:13","publication_year":"2023","word_count":441,"keywords":["Go","ChatGPT","API","AI","ML","Git","AI ethics","GPT","Aim","AI Tool","R"],"extracted_tech_keywords":["AI","ML","ChatGPT","Aim","R","Go","Git","API","GPT","AI ethics"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/gods-take-on-tech-vatican-launches-ai-ethics-think-tank\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":13865,"title":"Mathworks India to Host MATLAB EXPO 2017 in Bangalore, Pune and Hyderabad: 20-27th April 2017","content":"MATLAB EXPO 2017, one of India’s premier events for India’s research and engineering communities will be hosted in Bangalore, Pune and Hyderabad by MathWorks India. The EXPO will feature presentations and workshops by MathWorks technical professionals and customers. Over the last seven years, this conference has served as a stage for engineers, scientists and researchers to meet, converse and learn about cutting-edge product capabilities in MATLAB and Simulink. The annual conference will bring together engineers and scientists to learn new technological advancements and industry trends along with updates on MATLAB and Simulink. The featured speakers for this year’s conference will include Kishore Rao, Managing Director of MathWorks India, Prashant Rao, Technical manager of MathWorks India and Jim Tung, Fellow member of MathWorks amongst the others. Jim Tung, Fellow member of MathWorks will present the keynote on “How to Build an Autonomous Anything” where he will discuss about how using MATLAB and Simulink, engineers and scientists are combining massive amount of data, computing power and diverse set of algorithms to build autonomous technology in products and services today. EXPO will feature key industry players including Robert Bosch, Tata Motors, IFB, TAFE, Philips For the detailed agenda, visit www.matlabexpo.in. WHEN: Bangalore: Thursday, April 20, 2017 Pune: Tuesday, April 25, 2017 Hyderabad: Thursday, April 27, 2017 WHERE: Hotel Park Plaza 90\/4, Outer Ring Road Marathahalli Village, Bengaluru Karnataka 560037 J W Marriot Hotel Senapati Bapat Road, 30, Laxmi Society, Model Colony, Shivajinagar, Pune, Maharashtra 411053 HICC (Hyderabad international convention center) Novotel & HICC Complex (Near HITEC City), P.O Bag 1101, Cyberabad Post Office, Hyderabad – 500 081, India About MathWorks: MathWorks is the leading developer of mathematical computing software. MATLAB, the language of technical computing, is a programming environment for algorithm development, data analysis, visualization, and numeric computation. Simulink is a graphical environment for simulation and Model-Based Design for multi-domain dynamic and embedded systems. Engineers and scientists worldwide rely on these product families to accelerate the pace of discovery, innovation, and development in automotive, aerospace, electronics, financial services, biotech-pharmaceutical, and other industries. MATLAB and Simulink are also fundamental teaching and research tools in the world’s universities and learning institutions. Founded in 1984, MathWorks employs more than 3500 people in 15 countries, with headquarters in Natick, Massachusetts, USA. For additional information, visit mathworks.com.","excerpt":"MATLAB EXPO 2017, one of India’s premier events for India’s research and engineering communities will be hosted in Bangalore, Pune and Hyderabad by MathWorks India. The EXPO will feature presentations and workshops by MathWorks technical professionals and customers. Over the last seven years, this conference has served as a stage for engineers, scientists and researchers […]","categories":["Deep Tech"],"tags":[],"author_name":"Дарья","publish_date":"2017-03-29T04:09:34","publication_year":"2017","word_count":379,"keywords":["Go","programming_languages:R","AI","innovation","programming_languages:Go","BERT","llm_models:BERT","R"],"extracted_tech_keywords":["AI","R","Go","BERT","innovation","llm_models:BERT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/mathworks-india-host-annual-conference-matlab-expo-2017\/","complexity_score":4,"technical_depth":8,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":11708,"title":"Top 10 Analytics Trends in India to watch out for in 2017 &#8211; By AIM &#038; AnalytixLabs","content":"Analytics is by far the biggest influencer in IT industry – a phenomena evident by the rise of next-gen technology Cognitive Computing, Blockchain and Virtual Reality which has at its core a valuable asset “data”, and analytics quite irrefutably is the essence of it. After all, it’s the whole mix of technology, data and analytics that is revolutionizing the way we work. In keeping with our annual tradition, we present the much-researched and a carefully thought-out study carried out in association with AnalytixLab, a premier analytics training institute in India. We invited nominations from various organization to identify the analytics trends that will shape the future of our industry in India for 2017. After a lot of fact-finding from the industry insiders, we bring to you the top 10 analytic trends that have had the highest impact on the analytics industry today and potentially going forward in 2017. This year, we received an astounding 36 submissions. A lot of new trends have hit the space, a few faded away and there are others that endured and will definitely stay. This study, now in its third year presents neatly sorted out viewpoint of the industry leaders and veterans on Top 10 Analytics Trends in India to watch out for in 2017. Artificial intelligence becomes pervasive in business The Cognitive age is clearly upon us— it is indicated by the fact that more than $1 billion in venture capital funding went into cognitive science in 2014 and 2015, and further supported by fact that various analysts project the overall market revenue for cognitive sciences to exceed $60 billion by 2025. As the cognitive era evolves, it will likely become another key decision making tool in the toolbox of CXOs; vital for the right applications but not entirely replacing traditional business & advanced analytics capabilities that complement the human thought process. In a nutshell, the man-machine dichotomy is not “either-or”, it is unequivocally “both-and”. Debashish Banerjee, Managing Director at Deloitte Consulting There is a lot of hype around “artificial intelligence,” but it will often serve best as an augmentation rather than replacement of human analysis because it’s equally important to keep asking the right questions as it is to provide the answers. Souma Das, Managing Director, India, Qlik [divider size=”2″] Analytics of Things continues to be a game changer Internet of Things and Analytics of Things: The Indian Internet of Things (IoT) market is set to grow to $15 billion by 2020 from the current $5.6 billion. Just about every type of company seemed to have an IoT strategy in 2016. However, today, IoT is more about Data Than It Is Things. The original description of “Internet of Things,” was describing a network of connected physical objects. But in 2016, it was apparent that this initial description didn’t consider the importance of data or cloud computing. So now, IoT isn’t about connecting billions of objects to the Internet, it is really going to be all about the data and the ability of organizations to gain insights out of all of this data. This means that getting value from data goes beyond devices, sensors and machines and includes all data including that produced by server logs, geo location and data from the Internet. Sunil Jose, Managing Director, Teradata India The “Internet of Things” is exploding. It is predicted that the number of device connected would reach 50 billion by 2020. Most of these smart devices would be in factories, energy sector, health care systems, home appliances and wearable devices. The vital data so generated would enable us to track health parameters, optimise machine performance, and reduce response time for breakdowns and also save lives. In order to create real value of IoT\/IIoT, the Sensors & Communications node needs to integrate with the Analytics infrastructure, else it will be a simple data collection exercise. The technologies & skills related to IoT protocols, edge analytics & real time sensor analytics will be the key differentiators for its success & adoption in the market. Vinay Gupta, Head of Analytics at Suzlon [divider size=”1″] Enabling real-time automated decisioning systems There is a visible shift across many big data and analytics users to streaming and real time analytics. Businesses particularly digital advertising, ecommerce, logistics & transportation are looking to leverage ream time analytics and are heavily invested in this space. This is also apparent from the elevated adoption levels of Apache Spark Streaming, Apache Storm or Twitter’s Heron. Srikanth Sundarrajan, Principal Architect at InMobi Enterprises are increasingly enabling real-time automated decisioning systems whether to streamline operations or mitigate risk. The older rule-based systems are now being replaced by a new generation of systems powered by online machine learning and artificial intelligence – these are self-learning systems that can recalibrate in an automated manner and can be deployed on a large-scale. Pradeep Gulipalli, Co-founder at Tiger Analytics [divider size=”1″] Analytics is made invisible, embedded within the system Analytics works best when it’s a natural part of people’s workflow. In 2017, analytics will become pervasive and the market will expect analytics to enrich every business process. This will often put analytics into the hands of people who’ve never consumed data, like store clerks, call-center workers, and truck drivers. Deepak Ghodke, Country Manager, Tableau Technologies that nudge us to drink water, take regular walk breaks, or inform us that our cab has arrived have become commonplace. This is not restricted to consumer facing decisions. In fact, businesses in 2016 are beginning to realize the value of this kind of data and deploy on-demand analytics to drive better decisions. They are capturing and streaming unstructured data, blending it with other data sources, deploying analytical models to unearth insights, and are using rules engines to drive applicable “nudges”. Mihir Kittur, Co-founder, Ugam [divider size=”1″] Fintech is growing and so is Fintech Analytics Going by the events of the last year (2016), Fintech will clearly emerge as the most challenging as well as beneficial. The linkage of various identity proofs to uniquely identify a person and their financial footprint would be the key to the mission to drive out corruption and black money. Dr Nupur Pavan Bang, Associate Director, Thomas Schmidheiny Centre for Family Enterprise, Indian School of Business Financial institutions are moving rapidly towards “digitization” and educating their customers to adopt digital channels for day-to-day activities. With eroding revenue streams, intensifying competition, and ever-increasing customers’ expectation financial institutions need to explore new way of doing business. Measuring customer relationship, evaluating the customer journey and recommending right bundle of product & services at the right price in real time through technology enabled digital platform will be the vital enablers to improve customers’ banking experience. Suman Singh, Chief Analytics Officer, ZAFIN [divider size=”1″] Rise of Self-Service Analytics The realization that what delivers impact is not automated MLR or one-size fits all solutions, but context driven customized solutions that leverage business know-how (that probably exists deep within a company) and domain knowledge (of expert consultants with rich industry and applied analytics experience) will dawn on most business leaders in 2017. Randhir Hebbar, Cofounder at Convergytics Technological advancement has led to tools for Self Service Business Intelligence or Self Analytics. This approach meets the needs of data producers and consumers alike, adding speed and agility to the process while protecting organizational data and the system overall with a single version of the truth. Tejinderpal Singh Miglani, CEO, Incedo Inc There has been a paradigm shift in deriving business value from analytics owing to the exponential growth in the volume, variety and complexity of data. Today’s competitive business environment asks for democratisation of analytics with self-service capabilities to meet the time-to-insight demands. Self-service work benches, packaged analytics, dashboards, visualization frameworks and collaboration tools for data scientists (built right into the analytics frameworks) are gaining popularity. The spotlight has now shifted from IT-led reporting to business-led self-service analytics. Self-service allowing “regular” users to derive value from large data assets owned by enterprises is the way forward. Having said that, the specialists will not be redundant. In fact, there will an even greater demand so as to ensure that they manage and provide the entire infrastructure and a strong footing for decision making across the organisation. Jobin Wilson, Principal R&D Architect-Data Sciences, Flytxt [divider size=”1″] Democratization and consumerization of analytics More organizations are “democratizing” Business Intelligence (BI) and analytics to enable a broad range of non-IT users, from the executive level to frontline personnel, to do more on their own with data access and analysis via self-service BI and visual data discovery such as drag-and- drop dashboards. Anil Chawla, Managing Director, Customer Engagement Solutions, Verint Systems More data is now available to companies of all sizes. So more easy access to data within a company & sources to find external data will be a trend that I see. Companies can partner each other & leverage each other’s data. A DTH company knows when you move residences & that data can help a Retailer who sells furniture or is in that catchment. I believe that in year 2017 more marketers will leverage each other’s data to build more effective analytics solutions. Ajay Kelkar, Co- founder of Hansa Cequity [divider size=”1″] Mobile first Omni channel strategy With increasing penetration of mobile phones, the number of mobile apps have sky rocketed. Due to the limited space on these mobile phones, consumers are engaging with apps that create value for them. To stay relevant, the apps developers are using app analytics to understand their users’ profile and transaction behaviour to fine tune their product features and offering to increase user experience and engagement. We will continue to see analytics playing a larger role to declutter the space. Debasmit Mohanty, CEO & Founder, StratLytics With increased penetration of mobile, companies are taking a Mobile First approach to engage consumers, leading to progress in Mobile Analytics. With the available location and motion sensing capabilities, significant progress was made in data collection in a privacy compliant fashion to determine what, when, where, and why of the activities that consumers engage in. This data is enabling improved insight and reach for business growth and consumer experience improvement. Amit Deshpande, Vice President, Analytic Consulting Group, Epsilon [divider size=”1″] Leverage GeoSpatial analytics in improving business models In 2016, we have seen “Geo spatial analytics” gain good amount of momentum in India. Geo spatial analytics refers to mapping of events to point in time locations. With a massive adoption of mobile devices across India in the past year and businesses getting more digitized, there are lots of “time and place” data that is being collected today. Besides mobile devices, emergence of sensors with respect to smart cities in India, drones being evaluated in agricultural and construction sectors, emergence of social media in real time, we are seeing tremendous business potential in terms of leveraging Geo spatial analytics in improving business models and this trend will only continue to grow. Sunil Shirguppi, ‎Senior Vice-President – Big Data and Analytics at Happiest Minds Technologies In 2016, businesses found value in understanding the ‘where’ factor of data and the ability to query location based information or location analytics into their existing analytics. Business intelligence solutions along with geographic analysis brought forth insights that helped companies better communicate with their customers, create more targeted promotions and pursue previously unrecognized cross-selling opportunities. Manish Choudhary, SVP, Global Innovation and MD, Pitney Bowes India [divider size=”1″] Analytics Governance Platforms The use of analytical models is significantly increasing across business functions (marketing, sales, risk, pricing, etc.) and business & product lines of the enterprise. They are at different stages of deployment and being used continuously by various business users simultaneously. However post implementation of such models, businesses are not necessarily being able to track how these models are performing and not sure if they are delivering the promised value. Neither do the businesses know the interlinkage effect of all these models working together. With the increasing use of analytical models in business decisions, consolidating them in a single technology console, monitoring the health of analytics implementations and model performance is becoming a crucial need for business leadership and regulators. This will help track the analytics performance, demonstrated ROI and avoid the risk of incorrect decisions made because of analytics. ‘Analytics governance platforms’ will gain prominence across all large enterprises and will become mainstream to monitor the analytics deployment through workflows. The ROI of the analytics governance platforms will be seen in the long-term and would reap the benefits of governing complex analytics environments in a single platform with more visibility to the analytics ROI. Prithvijit Roy, CEO, and Co-founder, BRIDGEi2i Analytics Solutions [divider] Here’s the complete report. [divider] Download the complete report below: [attachments include=”11738″]","excerpt":"Analytics is by far the biggest influencer in IT industry – a phenomena evident by the rise of next-gen technology Cognitive Computing, Blockchain and Virtual Reality which has at its core a valuable asset “data”, and analytics quite irrefutably is the essence of it. After all, it’s the whole mix of technology, data and analytics […]","categories":["AI Features"],"tags":["cognitive computing human capital","real time decisioning","virtual invisible network"],"author_name":"Дарья","publish_date":"2016-12-20T07:29:52","publication_year":"2016","word_count":2115,"keywords":["data science","artificial intelligence","machine learning","AI","cloud computing","ML","virtual invisible network","Apache Spark","RAG","analytics","real time decisioning","cognitive computing human capital","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","data science","analytics","RAG","cloud computing","Apache Spark","R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/top-10-analytics-trends-india-watch-2017-aim-analytixlabs\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":true},{"article_id":10052897,"title":"How to Train Unigram Tokenizer Using Hugging Face?","content":"Text summarization, the creation of entirely new pieces of text, and the prediction of the next word like Google’s auto-fill, to name a few. Do you know what all of these NLP tasks have in common? Language models are at the heart of everything! To be honest, most complex NLP tasks require these language models as a first step. In this post, we’ll take a closer look at one of the language models, the Unigram model, by understanding its working concept and putting it into practice. We will also understand how to implement a unigram tokenizer using the Hugging Face package. The following are the topics that will be discussed in this article. Points to be Covered: Types of Language ModelWhat is the N-Gram model?Implementing Unigram Model using Hugging Face Let us start by seeing what language models are available so far. Types of Language Model Language Models are classified into two types: Models of Statistical Language: To learn the probability distribution of words, these models employ classic statistical techniques such as N-grams, Hidden Markov Models (HMM), and specific linguistic rules. Models of Neural Language: These are newcomers to the NLP scene, and they use various types of Neural Networks to model language. What is the N-Gram Model? An N-gram model predicts the most likely word to follow a sequence of N-1 words given a set of N-1 words. It’s a probabilistic model that has been trained on a text corpus. Many NLP applications, such as speech recognition, machine translation, and predictive text input, benefit from such a model. Consider the following two sentences: “Heavy rain” vs. “Heavy flood.” We know from experience that the first sentence sounds better. In the training corpus, “heavy rain” occurs substantially more frequently than “heavy flood,” according to an N-gram model. As a result, the first sentence is more likely to be chosen by the model. An n-gram (also known as a Q-gram) is a contiguous sequence of n items from a particular text or speech sample. Depending on the application, the elements can be phonemes, syllables, letters, words, or base pairs. Typically, n-grams are extracted from a text or audio corpus. When the components are words, n-grams are sometimes referred to as shingles. An n-gram model, also known as an (n-1) order Markov model, is a sort of probabilistic language model used to forecast the next item in a sequence.  In probability, communication theory, computational linguistics (such as statistical natural language processing), computational biology (such as biological sequence analysis), and data compression, n-gram models are now commonly employed. Simpleness and scalability are two advantages of n-gram models (and algorithms that employ them): bigger n allows a model to hold more context with a well-understood space-time tradeoff, allowing modest experiments to scale up efficiently. Implementing Unigram Model using Hugging Face The Unigram model solves the merging problem by assessing the likelihood of each subword combination rather than selecting the most common pattern. It computes the likelihood of each subword and then discards it using a loss function. You can then instruct the model to drop the bottom 20–30% of the subword tokens based on a particular threshold of the loss value. Unigram is a purely probabilistic method that selects both the pairs of characters and the final choice to merge (or not) based on probability in each iteration. The tokenizers package from Hugging Face includes implementations of all of today’s most popular tokenizers. It also enables us to train models from scratch on any dataset of our choosing and then tokenize the input string of our choice. The datasets we used to train these models are free books from wikitext-103, which contains 516 million words. Now let’s quickly set up the environment by running the following commands. # We won't need TensorFlow here !pip uninstall -y tensorflow # Install `transformers` from master !pip install git+https:\/\/github.com\/huggingface\/transformers !pip list | grep -E 'transformers|tokenizers' # transformers version at notebook update --- 2.11.0 # tokenizers version at notebook update --- 0.8.0rc1 Download and unzip the dataset. !wget https:\/\/s3.amazonaws.com\/research.metamind.io\/wikitext\/wikitext-103-raw-v1.zip !unzip wikitext-103-raw-v1.zip We’ll create and train a tokenizer in this tour. In this case, educating the tokenizer entails teaching it merging rules. To begin, use all of the characters in the training corpus as tokens.Combine the most common pair of tokens into a single token.Continue until the vocabulary (for example, the number of tokens) reaches the desired size. The Tokenizer class is the library’s core API; here’s how one can create with a Unigram model: from tokenizers import Tokenizer from tokenizers.models import Unigram tokenizer = Tokenizer(Unigram()) Next is normalization, which is a collection of procedures applied to a raw string to make it less random or “cleaner.” Stripping whitespace, eliminating accented characters, and lowercasing all text are all common actions. If you’re familiar with Unicode normalization, you’ll recognize it as one of the most popular normalization operations used by most tokenizers. from tokenizers import normalizers from tokenizers.normalizers import NFD, StripAccents normalizer = normalizers.Sequence([NFD(), StripAccents()]) tokenizer.normalizer = normalizer Pre-tokenization is the process of breaking down a text into smaller pieces in order to estimate how many tokens you’ll have at the end of training. To put it another way, the pre-tokenizer will divide your text into “words,” and your final tokens will be parts of those words. Splitting inputs on spaces and punctuations, as done by the Whitespace pre-tokenizer, is a simple approach to pre-tokenize inputs: from tokenizers.pre_tokenizers import Whitespace tokenizer.pre_tokenizer = Whitespace() We’ll need to create a trainer, in this case, a UnigramTrainer, to train our tokenizer on the wikitext files also defining some special tokens to handle special characters and likewise. from tokenizers.trainers import UnigramTrainer trainer = UnigramTrainer(special_tokens=[\"[UNK]\", \"[CLS]\", \"[SEP]\", \"[PAD]\", \"[MASK]\"]) We could want our tokenizer to add special tokens like “[CLS]” or “[SEP]” automatically. A post-processor is used to do this. The most frequent method in TemplateProcessing, which requires simply the specification of a template for the processing of single sentences and pairs of sentences, as well as the special tokens and their IDs. Here’s how to configure the post-processing for standard BERT inputs: from tokenizers.processors import TemplateProcessing tokenizer.post_processor = TemplateProcessing( single=\"[CLS] $A [SEP]\", pair=\"[CLS] $A [SEP] $B:1 [SEP]:1\", special_tokens=[(\"[CLS]\", 1), (\"[SEP]\", 2)],) Now, we are good to go now. We can now simply use the train() method with whatever list of files we want: Our tokenizer should only take a few seconds to train on the entire wikitext dataset. files = [f\"wikitext-103-raw\/wiki.{split}.raw\" for split in [\"test\", \"train\", \"valid\"]] tokenizer.train(files, trainer) It is now time to put the model to the test. A Tokenizer offers an API for decoding, which converts IDs created by your model back to the text, in addition to encoding the input texts. This is accomplished using the methods decode() (for a single predicted text) and decode batch() (for a batch of predictions). output = tokenizer.encode('This code written in notebook') print('Token ids:',output.ids) print('Decoded Sequence:',tokenizer.decode([1, 181, 2580, 53, 545, 19, 72, 10, 447, 2])) Let’s see the tokenized form of our corpus print('Tokenized form:',output.tokens) To handle a batch of sentence pairs, pass two lists to the encode batch() method: the list of sentences A and the list of sentences B: As a result, we have a collection of Encoding objects that are comparable to those we saw earlier. You can combine as many texts as you want as long as they fit in your memory. output = tokenizer.encode_batch([\"Hello there all!\", \"How are you ?\"]) print('Tokens of 2nd sentence:',output[1].tokens) Pass two lists to the encode batch() method to handle a batch of sentence pairs: the list of sentences A and the list of sentences B like below. output = tokenizer.encode_batch( [[\"Hello there all!\", \"How are you ?\"]]) print('Tokens of 1st batch:',output[0].tokens) Final Words Unigram appears to generate subword tokens that are more typically seen in the English language based on the type of tokens generated, however, this observation is not universal. These algorithms differ from one another yet do a comparable job of generating a good NLP model. However, the performance of your language model is heavily influenced by the use case, vocabulary quantity, speed, and other aspects. In this post, we have seen what kind of language models exist right now. Based on it we narrowed our discussion towards the Unigram model in which we have seen the working of the algorithm and implemented the same with the support of hugging face. References Hugging Face Documentation Link for above codes","excerpt":"Text summarization, the creation of entirely new pieces of text, and the prediction of the next word like Google’s auto-fill, to name a few.","categories":["AI Trends"],"tags":["AI (Artificial Intelligence)","Data Science","Machine Learning","Python"],"author_name":"Vijaysinh Lendave","publish_date":"2021-11-04T16:00:00","publication_year":"2021","word_count":1396,"keywords":["Go","Hugging Face","TPU","AWS","AI","neural network","Machine Learning","Transformers","Python","NLP","Data Science","TensorFlow","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","neural network","NLP","TensorFlow","Hugging Face","Transformers","AWS","TPU","R","Go"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/how-to-train-unigram-tokenizer-using-hugging-face\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":true},{"article_id":10103225,"title":"Google DeepMind Launches Lyria, Transforming the Future of Music with AI","content":"In a groundbreaking partnership with YouTube, Google DeepMind unveiled Lyria, their most sophisticated AI music generation model yet, alongside two experimental projects set to revolutionise music creation and collaboration. Music, with its intricate layers of melody, rhythm, and vocals, has long posed a challenge for AI systems. But today marks a turning point. Lyria, developed by Google DeepMind, represents a leap in AI music generation. This model excels in producing high-quality music, handling instrumentals and vocals, and offers nuanced control over style and performance, aiming to bridge the gap between AI and musical continuity. In a bid to foster connections between artists and audiences, YouTube Shorts hosts Dream Track, an experiment powered by Lyria. Selected creators will collaborate with renowned artists like Charlie Puth, Demi Lovato, and Sia, among others, to produce unique soundtracks using AI-generated voices and musical styles. Dream Track users can seamlessly generate 30-second soundtracks by selecting an artist and a topic, receiving an AI-generated voice and musical accompaniment tailored to the chosen style. Additionally, Google’s collaborative efforts with industry experts in YouTube’s Music AI Incubator aim to innovate music AI tools. These tools are envisioned to facilitate creativity, enabling users to compose melodies from hums, transform chords into vocal choirs, or even convert music styles and instruments seamlessly. Through this initiative, Google DeepMind and YouTube are reshaping the creative landscape, providing a glimpse into a future where AI and human collaboration redefine the boundaries of musical innovation. Watch the evolution of music creation unfold as AI meets artistry in this groundbreaking endeavor.","excerpt":"Google’s collaborative efforts with industry experts in YouTube’s Music AI Incubator aim to innovate music AI tools.","categories":["AI News"],"tags":["Google Deepmind","innovation"],"author_name":"Shyam Nandan Upadhyay","publish_date":"2023-11-17T18:23:16","publication_year":"2023","word_count":256,"keywords":["Go","API","AI music","programming_languages:R","AI","innovation","ML","Aim","Google Deepmind","ViT","R"],"extracted_tech_keywords":["AI","ML","Aim","R","Go","API","ViT","innovation","AI music","programming_languages:R"],"url":"https:\/\/analyticsindiamag.com\/ai-news-updates\/google-deepmind-launches-lyria-transforming-the-future-of-music-with-ai\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":1,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":62250,"title":"Dark Web Peddlers Are Selling Fake COVID 19 Vaccines","content":"CYFIRMA, threat discovery and cyber intelligence platform company backed by Goldman Sachs, Zodius Capital, and Z3P Partners, has launched a report publishing their observations about change in cybercriminals’ approach and attitude towards taking advantage of the current COVID-19 pandemic for financial gains. CYFIRMA researchers observed that hackers are cognizant to the dangers of putting millions of lives at risk as families of those who have been infected by the COVID-19 virus would likely be desperately seeking a medical remedy. Any news of a vaccine availability could also send masses of people into a state of frenzy and cause major turmoil across many societies. While hackers and scammers have been leveraging the pandemic to push out malware and phishing emails as part of their cyber- attack campaigns to steal data from businesses and consumers, or to cause social unrest amongst various communities, there has been an understanding amongst hackers groups to not ‘cross the line of humanity’ by selling fictitious vaccines. A market place in the Dark Web called Monopoly has the restricted sale of fake vaccines for COVID 19 on their platform. While they sell all sorts of illicit stuff. And the ‘founder’ of the marketplace wrote a post, ’Any vendor caught flogging goods as a cure to Coronavirus will not only be permanently removed from this market but should be avoided like the Spanish Flu’. The forum post also stated the gravity of the pandemic and asked sellers not to use the crisis as a marketing tool. However, there is a group of hackers who have ignored this warning and are choosing to sell fake vaccine and making anywhere from $99 to $25000. According to Cyfirma report, the hackers are from North Korean and have got interest from Italy, Spain, France and the US. Payment is being made via bitcoin, few bitcoin accounts have collected on tune of $400K just within last 6 days. The obvious fallout of this malicious act is the loss of money by the users but in order to get the vaccine, they have shared their Personal identifiable information including health and social security details. The theft of personal information will also fetch additional financial gains for hackers. Cyfirma predicts that personal information provided to buy fake vaccine could be used for the next wave of cyberattacks.","excerpt":"CYFIRMA, threat discovery and cyber intelligence platform company backed by Goldman Sachs, Zodius Capital, and Z3P Partners, has launched a report publishing their observations about change in cybercriminals’ approach and attitude towards taking advantage of the current COVID-19 pandemic for financial gains. CYFIRMA researchers observed that hackers are cognizant to the dangers of putting millions […]","categories":["AI Features"],"tags":["covid-19","dark web"],"author_name":"Vishal Chawla","publish_date":"2020-04-21T13:24:07","publication_year":"2020","word_count":382,"keywords":["Go","API","covid-19","AI","programming_languages:R","programming_languages:Go","dark web","RAG","ViT","R"],"extracted_tech_keywords":["AI","RAG","R","Go","API","ViT","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/dark-web-peddlers-are-selling-fake-covid-19-vaccines\/","complexity_score":2,"technical_depth":8,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":10130428,"title":"Futuristic Acer AI PCs Coming Soon in Indian Market","content":"Around 241.8 million units of personal computers were sold all across the world in 2023. Despite this, it was the worst year on record for PC sales, with a nearly 15% decline compared to the previous year, according to Gartner. In India too, last year, the PC market declined by 6.6%. Low market sentiments post-pandemic, supply chain constraints and geopolitical tensions contributed to the decline; but now PC makers are hoping generative AI could help alter their fate. The top PC makers in the world have been quick to ship AI-powered PCs in most markets. Acer, which has a relatively small portion of the PC sales market, witnessed a 12.3% increase in sales in 2023, the highest among all. Acer too has already launched a series of AI PCs that are also available in the Indian market which comes with built-in AI features. New AI PCs Coming Soon to Indian Market In an interview with AIM, Sudhir Goel, chief business officer at Acer India said, “At Computex 2024, we have showcased a lot of new products, which we are thrilled to introduce to the Indian market in the coming year.” So far, Acer has launched the Acer Swift 14 AI PCs, the TravelMate series, and its Predator Helios AI gaming laptops in India. “With Swift AI, all functionalities, ranging from image enhancement to voice processing, are performed locally, thanks to the neural processing unit integrated within our laptops. This ensures an elevated level of privacy and security, paramount for individual and corporate users,” Goel said. The TravelMate business PCs are equipped with sophisticated enterprise-grade AI security, and they will soon be available in India. These laptops feature advanced AI tools, including Acer LiveArt and the GIMP with Intel’s Stable Diffusion plugin. It also leverages NPU for AI-accelerated applications to blur the background, automatically frame, and maintain eye contact during video conferencing. Moreover, AI will optimise power consumption during long conferencing calls. “In the gaming sector, AI integration will redefine immersive gaming by enabling real-time map generation and dynamic creation of in-game elements based on live data. Additionally, we will introduce advanced AI-driven monitors to elevate the overall user experience,” Goel revealed. The Predator Helios 16 laptops are already available in the Indian market. However, the true advantage of AI emerges when an AI model can be run locally on the device. Given the gargantuan sizes of these Large Language Models (LLMs), they can only run on the cloud. “With Acer’s cutting-edge Neural Processing Units (NPU), our laptops can handle LLM tasks directly on the device. As we look to the future, we envision local LLM capabilities becoming a significant differentiator in the market,” Goel revealed. Acer Laptops Will have New AI Processors The Acer Swift AI PCs come with Qualcomm’s Snapdragon Elite processors. However, Acer plans to offer a range of new processors in its upcoming laptops. “While Snapdragon’s latest technology offers remarkable capabilities, we are also exploring options from Intel and AMD. Our strategy is to evaluate and incorporate processors from all these leading providers based on their strengths and innovations. This approach ensures that our laptops can cater to a wide range of needs, from exceptional performance and efficiency to specialised AI features,” Goel said. The TravelMate P6 14 laptop features Intel Core Ultra 7 processors with Intel vPro Enterprise, Intel Graphics, and Intel AI Boost. Will AI PCs Boost the Market? While Acer’s introduction of AI-capable PCs is impressive, other PC manufacturers have swiftly followed suit. Dell and HP have also released AI-powered PCs in the Indian market this year. Most recently, Microsoft unveiled its Surface AI PCs in India, featuring Snapdragon Elite Processors. PC makers are hoping AI could help pull the market from the stalemate that it was last year. Research firm Canalys predicts that the PC market will see an 8% annual growth in 2024 as more AI PCs hit the market. Canalys also predicts AI PCs will capture 60% of the market by 2027. Goel also believes generative AI has the potential to significantly boost laptop sales. “Over the past few years, the PC industry has been striving to make devices more powerful, efficient, thinner, and lighter. However, it lacked a transformative technology that could truly revolutionise the market. With the advent of AI, this missing piece has finally arrived,” he said. AI-driven workloads in PCs could enhance performance, enable new functionalities, and create a more seamless user experience. PCs makers are desperately banking for this to happen. Acer, too Turns to Server Business and Consumer Electronics Earlier this year, Acer also launched its consumer electronics and home appliances brand, Acer Pure in India. When we asked Goel whether it is a result of declining PC sales, he said, “Acer India is the fastest-growing PC brand in India, and we have seen remarkable YoY growth for our PC business with 2X growth in the consumer market and market leadership in some of the commercial segments.” He stressed that PCs remain Acer’s core business. Interestingly, Acer also launched its server business in India a few years back, a segment dominated by Dell, HP, and Lenovo, other PC brands Acer competes with. Called Altos Computing, it caters to the growing demand for high-performance servers and workstations in India’s digital infrastructure landscape. “It includes introducing AI-powered solutions to support local cloud and data storage initiatives, which are crucial for governmental and corporate digital transformation priorities,” Goel concluded.","excerpt":"So far, Acer has launched the Acer Swift 14 AI PCs, the TravelMate series, and its Predator Helios AI gaming laptops in India.","categories":["AI Trends"],"tags":["AI Impacts"],"author_name":"Pritam Bordoloi","publish_date":"2024-07-28T14:07:46","publication_year":"2024","word_count":901,"keywords":["Go","AI","ML","Git","RAG","Aim","stable diffusion","generative AI","AI Impacts","GAN","R"],"extracted_tech_keywords":["AI","ML","generative AI","Aim","RAG","R","Go","Git","stable diffusion","GAN"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/futuristic-acer-ai-pcs-coming-soon-in-indian-market\/","complexity_score":4,"technical_depth":10,"industry_relevance_score":5,"has_code_examples":true,"has_tutorial_content":false,"is_research_content":false},{"article_id":10010977,"title":"Understanding Speech: Moving Beyond ASRs","content":"Deep Learning DevCon 2020 or DLDC 2020 is another conference of the year that is hosted in partnership with Analytics India Magazine. Scheduled for 29th and 30th October, the conference has brought the leading experts and best minds of deep learning and machine learning industry from around the globe. The first session of Day 1 was presented by Abhinav Tushar, who is the head of AI at the Bengaluru-based conversational AI startup Vernacular.ai. The primary aspect of the session named “Understanding Speech: Moving beyond Automatic Speech Recognitions” is — although text-based conversational interactions have been around in the industry for a while now, speech interactions are still in infancy. Tushar kickstarted the talk by explaining the importance of speech and the emotions hidden behind it. He said that speech is far different than text, and it is much more complex. He said, “Speech is much more than transcriptions. And that should influence how we design conversational agents.” Tushar mentioned that some of the factors that are impacting the responses include content, environment, speaker characteristics and paralinguistics. He also gave an instance of various “okays” spoken by various people that depicted different emotions in each different time. He then discussed the workings of the present conversational AI-based voice bot that is built at Vernacular.ai and how it is chasing up in terms of mirroring human behaviour. The working of the framework follows the mentioned steps: When a user speaks, the speech goes into the speech recognition block, where it extracts the speech It then moves forward into the Automatic Speech Recognition system, that includes an acoustic model, pronunciation model and language model. After that, it moves forward for frame understanding like intent classification, pre-processing and entity parsing. Next step is a content management and dialogue management process. Then the final step proceeds, where the text is transformed into speech and sent to the user. Further, Tushar discussed the various stages of extra-lexical conversational behaviour that includes snapshot-based, flow-based and persuasive. Snapshot-based understands the behavioural snapshots and performs simple actions. This feature includes bail out on certain cues, detect personal characteristics and switch prompt, etc. In the Flow-based stage, the system works across multiple turns and can perform basic repairs. The feature includes tracking the consistent expression of discomfort, change in-flow experience based on the situation, etc. In the Persuasive stage, the system persuades the other party and drives the conversation. The features include understanding and utilising uncertainties and preferences, model situations and manoeuvre, etc. Furthermore, in order to build such intelligent systems, Tushar concluded that one must include components like: Stylistic and semantic modelsState trackingLive experimentation framework","excerpt":"Deep Learning DevCon 2020 or DLDC 2020 is another conference of the year that is hosted in partnership with Analytics India Magazine. Scheduled for 29th and 30th October, the conference has brought the leading experts and best minds of deep learning and machine learning industry from around the globe. The first session of Day 1 […]","categories":["Deep Tech"],"tags":["Speech Analytics","Speech Recognition","speech recognition algorithm","Vernacular Automated Speech Recognition"],"author_name":"Ambika Choudhury","publish_date":"2020-10-31T16:00:00","publication_year":"2020","word_count":436,"keywords":["Go","machine learning","Speech Analytics","AI","programming_languages:R","Vernacular Automated Speech Recognition","speech recognition algorithm","programming_languages:Go","deep learning","analytics","Speech Recognition","conversational agents","R","startup"],"extracted_tech_keywords":["AI","machine learning","deep learning","analytics","conversational agents","R","Go","startup","programming_languages:R","programming_languages:Go"],"url":"https:\/\/analyticsindiamag.com\/deep-tech\/understanding-speech-moving-beyond-asrs\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":734,"title":"10 leading IoT Accelerators\/ Incubators in India","content":"With India witnessing a trend of start-ups in the IoT space, incubators and accelerators too have evolved to support these budding start-ups. Both incubators and accelerators are playing an important role in helping the Indian IoT start-ups take their initial steps which are extremely important in their long and challenging journey. IoT India Magazine puts together a list of 10 leading Incubators and Accelerators which the IoT Indian start-ups should definitely watch out for. Ashoka -ISB –Microsoft Ventures The Indian School of Business and Ashoka University in association with Microsoft Ventures focus on providing support to the IoT start-ups. This incubator has an investment pool of around Rs 1 crore which is to be put into startups with a focus on Smart Cities. Further, an investment of Rs 7-10 lakh will be put in each venture selected, benefitting 8-10 teams altogether. Microsoft Ventures will provide software and technology support for the programme. Autonebula Autonebula is a connected transport accelerator and investment fund. Its primary focus is on supporting startups which are contributing towards developing Connected Transport solutions. Their vision is to help accelerate the adoption of smart, innovative transportation technologies resulting in economic and social value. To support this vision, Autonebula provides seed funding, work facilities, mentoring and networking opportunities, to the start-up in the connected transport area. Center of Excellence for IOT The CoE-IoT Incubation Program will act as a catalyst to accelerate the efforts of the start-ups for growth. The program will provide the necessary support to the start-ups to innovate and create new products leveraging technology to solve real life problems. The IoT lab is an interoperable, multi-technology based stack which will provide design prototype equipments, validation & testing space, demo space and concept labs. GSF Global Accelerator GSF Accelerator’s IoT+Cloud program offers founders four months to make the kind of progress typically accomplished in eight to twelve months. GSF IoT+Cloud is structured like a Hackathon. Products pivot several times as teams go through intense, often grueling Q&A sessions with mentors and entrepreneurs in residence. GSF has collaborated with Cisco through a cross-functional and multi-country effort. Cisco will tap into its product and engineering presence in India as well as its global teams, to support and mentor entrepreneurs in India focused on IoT and Cloud enabled opportunities. HyperCat The Hypercat Incubator offers an entrepreneurial environment to the start-ups in the domain of IoT and Smart City. It offers high quality offices, advisory support, an international network of collaborators and services. Its aim is to support the launch, survival and rapid growth of new IoT and Smart City businesses that commit to using the Hypercat standard for interoperability. The Hypercat Incubator is situated in Hyderabad, India’s Hi-Tech city, and has the backing of 700 businesses, globally, and houses 40 startups. ISBA ISBA, started a decade ago, today has a membership base of 70+ organisations that are into entrepreneurship development and incubation. ISBA has a pan India network of Business Incubators and has an abundance of expertise and knowledge base in technology business Incubation that has been acquired over a period of 25 years. A diverse range of technology sectors are addressed by the technology business incubators including IoT. Qualcomm Qualcomm has created the Qualcomm® Design in India Challenge in collaboration with The National Association of Software and Services Companies (NASSCOM). This challenge aims to discover startups which are into product design, development, and manufacturing of hardware and Internet of Things (IoT) for healthcare, education, banking, automotive as well as Smart Cities sectors. In this challenge, in the first round, 10 companies are selected and given $10,000, as well as incubation and access to Qualcomm’s innovation labs. Post this, in the final round, the top three startups get a total prize money of $100,000. Revvx hardware Accelerator Revvx is India’s first Indian Accelerator focused on supporting Hardware Startups through Prototyping and Manufacturing of innovative consumer electronics and connected Devices. RevvX supports building Disruptive and Innovative Product Startups in areas such as Internet of Things, Augmented \/Virtual reality, Robotics, Drones, Blockchain, Wearables and Connected Automobiles. Revvx drives Hardware Startups in 3 separate Tracks as part of the Acceleration Program. Track1: Prototyping: Startups build Proof of Concept into a Prototype Track2: Mass Manufacturing: Startups scale up Production Track3: Distribution: Startups roll out Retail Sales Target Accelerator Program The Target Accelerator Program is designed to help early stage startups develop concepts that could improve Target’s business and the broader retail industry. Through the four month intensive program, Target focuses on transformative and technology-driven ideas to create solutions for today’s rapidly changing retail environment. The American retail chain has been selecting startups relevant to its business for its accelerator program in India and the recent fourth batch of start-ups selected are in the areas of AI, AR, and IoT. T Labs Founded in 2011, TLabs, is a startup accelerator as well as an early stage seed-fund for Indian internet and mobile technology startups. TLabs is an in-house team that provides support through weekly catch-ups across different verticals of the business and provides access to over 100+ experts who are mentors of the program and guide startups on specific issues. There are regular engagement with early stage VCs and angels through multiple sessions.","excerpt":"Incubators and Accelerators play a vital role in hand-holding start-ups and showing them the path forward.","categories":["AI Trends"],"tags":[],"author_name":"Manisha Salecha","publish_date":"2016-11-08T05:55:31","publication_year":"2016","word_count":869,"keywords":["Go","API","AI","innovation","RAG","GRU","Aim","GAN","R","startup"],"extracted_tech_keywords":["AI","Aim","RAG","R","Go","API","GAN","GRU","innovation","startup"],"url":"https:\/\/analyticsindiamag.com\/ai-trends\/10-leading-iot-accelerators-incubators-india\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":true,"has_tutorial_content":true,"is_research_content":false},{"article_id":10098083,"title":"Google AI Helps Doctors Decide When to Trust AI Diagnoses","content":"AI has changed the face of many sectors, but in healthcare its adoption is moving at snail’s pace. The sector faces several challenges such as data privacy, security, lack of interoperability, and the absence of regulation, which have restricted its adoption. AI models are prone to mistakes, and as we know it is human to err. Google Research has asked the question, what would be the error rates when you combine the expertise of predictive AI models and clinicians? In July this year, Google DeepMind joined hands with Google Research and introduced the Complimentary-driven-Deferral-to-Clinical-Workflow (CoDoC), a system that maximises accuracy by combining human expertise with predictive AI. The system essentially decides if the AI model is more accurate than a hypothetical clinician’s workflow of diagnosis. It does this using a confidence score of the predictive model as one of the inputs. The comprehensive tests of the CoDoC with multiple real-world datasets has shown that with human expertise and predictive AI results with CoDoC provides greater accuracy. They saw a reduction of 25% in false positives for mammography datasets and more importantly, didn’t miss any true positives. The published paper is a significant advancement in collaboration between AI and clinicians. It promises improved accuracy in determining disease with binary outcomes. Their datasets focused on breast cancer screening using X-Ray mammography and a triage for TB tests using chest X-Rays. Sidestepping AI hurdles in medicine Knowing when to say ‘I don’t know’ is essential when working with artificial intelligence tools in a medical setting. The paper addresses the crucial challenge of when to acknowledge uncertainty and then to pass on the responsibility to the clinician. “If you use CoDoC together with the AI tool, and the outputs of a real radiologist, and then CoDoC helps decide which opinion to use, the resulting accuracy is better than either the person or the AI tool alone,” says Alan Karthikesalingam at Google Health UK, who worked on the research. The CoDoC model also does not require medical images from patients to make the diagnosis. This takes care of the privacy of the patient. It requires only three inputs for each case in training the dataset. First is the outputs of the hospital’s own existing predictive AI’s confidence score (0 is a certainty of no disease present and 1 is certain that the disease is present). Second, the outputs from a non-AI expert clinical workflow and finally historical ‘ground truth’ data. The system could be compatible with any proprietary AI models and would not need access to the model’s inner workings or data it was trained on. To apply the CoDoC paradigm to existing predictive AI systems, researchers would follow the methodology described in the paper – this involves training a CoDoC-style model using the outputs of their own existing predictive AI system. How the predictive model works It learns by comparing the predictive AI model’s accuracy with what the doctor interprets and then checks how the accuracy varies with the confidence score generated by the predictive AI model. After being trained, CoDoC is placed in a hypothetical future clinical workflow, working alongside both the predictive AI and the human doctor. When a new patient image is evaluated by the predictive AI model, its associated confidence score is fed into CoDoC. By having an AI as an effective tool to confirm the diagnosis, doctors can be confident of their diagnoses even in edge cases, something that was not available before. “The advantage of CoDoC is that it’s interoperable with a variety of proprietary AI systems,” says Krishnamurthy Dvijotham at Google DeepMind. Inviting developers to test, validate, build To help researchers build on their work, be transparent and ensure safer AI models for the real world, they’ve open-sourced CodoC’s code on Github. This work is theoretical so far but according to the researchers it shows the AI system’s potential to adapt, improve performance on interpreting medical imaging across varied demographic populations, clinical settings, medical imaging equipment used, and disease types. Helen Salisbury from the University of Oxford said, “It is a welcome development, but mammograms and tuberculosis checks involve fewer variables than most diagnostic decisions, so expanding the use of AI to other applications will be challenging.” She further said, “For systems where you have no chance to influence, post-hoc, what comes out the black box, it seems like a good idea to add on machine learning.”","excerpt":"In a joint effort, Google DeepMind and Google Research publish a paper that promises safer medical imaging interpretations, reducing false positives by 25%.","categories":["Global Tech"],"tags":[],"author_name":"K L Krithika","publish_date":"2023-08-03T16:34:16","publication_year":"2023","word_count":730,"keywords":["Go","machine learning","artificial intelligence","TPU","AI","Git","Ray","ViT","GitHub","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","Ray","TPU","R","Go","Git","GitHub","ViT"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/google-ai-helps-doctors-decide-when-to-trust-ai-diagnoses\/","complexity_score":2,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":true},{"article_id":54323,"title":"C-Level Executives Should Stay Away From These 6 Cybersecurity Myths","content":"The C-suite in any organisation is entrusted with the responsibility of spearheading innovation, progress, and company direction. Additionally, C-level executives hold a greater responsibility in maintaining the security of their network. Hackers believe senior executives are the weakest link in an enterprise network, often falling prey to fraud and phishing scams. In the current digitally advanced era, it is imperative for senior executives to develop a solid foundation of security realities to ensure the organisation is prepared to detect and defend cyber threats at all times. However, certain myths around cybersecurity that c-level executives should consider seriously, which could prevent them from safeguarding the enterprise effectively. Here are six such myths and healthy workarounds that can help decision-makers defy them. Myth 1: Security defences don’t have to be very expensive A CEO and Board Risk Management Survey by Deloitte highlighted that 95% of CEOs believe their enterprises will face severe threats and disruptions to their growth prospects in the next couple of years — with disruptive technologies and cyber incidents being the two greatest threats. However, many leaders still utilise traditional approaches, tools, and technologies to detect and manage risks. Investments and budgets allocated to cybersecurity are quite low when compared to the magnitude of potential damage — reputational and financial — that a cyberattack can result in. Yet, many executive-level managers are sceptical about investing in security defences. It’s time they change their mindset that higher investments in cybersecurity, which will result in more significant expenses on the balance sheet. The c-suite must realise the need for sophisticated cyber defence methodologies to counter and prevent threats. Instead of focusing the discussions on the amount spent on cybersecurity, it should revolve around utilising cybersecurity investments intelligently. Myth 2: Security cannot and should not be outsourced While some enterprise leaders believe security is too expensive to be outsourced, a few others feel outsourcing could result in non-compliance to state or country regulations. However, some executives feel cybersecurity is a sensitive area that has to be managed in-house. As far as regulatory compliance is concerned, data protection mandates do not prevent outsourcing. When done in a controlled manner with all underlying signed service agreements with liability clauses in place, outsourcing will not just be legal and complaint but will also offer better defence in a cost-effective way. IT security is more affordable and flexible when entrusted to the cybersecurity experts. Choosing the right security partner is critical. Firms with expertise in niche security skills will be able to streamline processes efficiently and ensure proactive monitoring against all incoming threats. Such firms are not only successful in identifying and hiring the best cybersecurity talent but are also able to retain the recruited security experts—by motivating them with opportunities for cross-skilling, upskilling, and working across technologies and horizontals. Myth 3: Adherence to IT and cyber regulations equals 100% fool-proof cybersecurity Developing such a mindset is one of the most expensive mistakes that decision-makers commit. While compliance with government and industry regulations is critical in carrying out business and building trust with partners and customers — it is the bare minimum. It only means that the enterprise has checked the first box with regard to cybersecurity. Compliance can reduce reputation damage to a certain extent when entangled in a legal cyberattack battle, but it does not promise adequate security. A robust incident response plan is required to avoid the attack itself. Defining the right strategy to protect an organisation’s high-value assets is the responsibility of the leadership team over and above regulatory compliance responsibilities. Myth 4: Controls? Implemented. Patches? Up-to-date. Hence, my organisation is completely safe CEOs need to understand that controls implemented once do not protect an enterprise forever. They need constant auditing, reviewing, refreshing, and upgrading, to be protected from various threat vectors. In today’s digitally advanced era, an enterprise network is loaded with a plethora of applications, firewalls, routers, servers and connected devices. Keeping every single component in this network patched and up to date is a massive, extremely critical task. Hence, relaxing post executing basic controls and patches, and assuming that the enterprise is safe, is going to cost the enterprise dearly. Permanent satisfaction is never a reality with regard to cybersecurity. Leaders need to understand the importance of keeping up with the trends of cyber defence. They should continuously monitor, audit, implement and embed security operations as part of the enterprise’s proactive defence strategy. Cybersecurity should become an essential element of board-level agenda. Myth 5: Awareness programs are sufficient for employees to fight and prevent cyber threats As threats and defences change constantly, lack of a structured approach toward tackling these threats could disrupt businesses to an unimaginable extent. Once-in-a-lifetime or annual cybersecurity training and awareness programs are not going to be enough anymore. Imagine the magnitude of damage that can be caused when an employee accidentally opens a malicious link in a phishing email, or uploads a client document to a public folder, or shares critical code with the wrong person. Today, social engineering attacks are one of the most popular ways to attack organisations in the cyberspace, and such incidents can be avoided through regular cybersecurity awareness campaigns and phishing simulations. Organisations with security-aware employees tend to be better poised to prevent and counter cyber threats. Regularly educating employees on the do’s and don’ts associated with cybersecurity—via e-mail campaigns, communication campaigns, employee awareness programs, specialised training programs for IT teams, etc.—gets employees to think twice before unintentionally putting data into the wrong hands. Myth 6: My enterprise’s current insurance policy covers cyber insurance 71% of 105 CFOs surveyed by FM Global felt that they were adequately covered in the event of a cybersecurity incident, and 26% expected the cyber insurance provider to cover their losses in full. Assuming cyber insurance is included in the firm’s existing insurance policy is a common mistake. Most traditional commercial general liability policies do not cover cyber risks such as data breach response, liability and privacy fines. While signing up for insurance policies, decision-makers need to be fully aware of the extent of coverage. The firm’s current insurance policy may not cover fines to regulatory bodies or financial losses due to attacks. Cybersecurity insurance needs to be chosen after careful evaluation—identifying the possible levels of threat; identifying different types of penalty that could be applicable in case of an issue, and understanding if the cost that will be covered in the cyber insurance policy. It is crucial to opt for a cyber insurance policy that will include a reasonable level of damage in case of an attack. The way forward We are in the digital era of sophisticated cybercrime and ransomware that place cybersecurity as a direct responsibility on the enterprise leadership team. Establishing a cyber security-aware culture across the enterprise is possible only when CEOs and other C-level executives themselves are aware of the firm’s current security posture and risk appetite, understanding potential threats unique to the organisation, and are continuously looking for ways to strengthen the company’s security posture. As cyber threats continue to advance to new levels, it is crucial for leaders to be proactive in their stance, invest wisely in the right security defences, and safeguard company assets most sensibly.","excerpt":"The C-suite in any organisation is entrusted with the responsibility of spearheading innovation, progress, and company direction. Additionally, C-level executives hold a greater responsibility in maintaining the security of their network. Hackers believe senior executives are the weakest link in an enterprise network, often falling prey to fraud and phishing scams.  In the current digitally […]","categories":["AI Features"],"tags":["Cybersecurity"],"author_name":"Priya Kanduri","publish_date":"2020-01-18T10:00:00","publication_year":"2020","word_count":1201,"keywords":["Go","ELT","AI","innovation","ML","Git","RAG","Rust","GAN","Cybersecurity","R"],"extracted_tech_keywords":["AI","ML","RAG","R","Go","Rust","Git","ELT","GAN","innovation"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/c-level-executives-cybersecurity-myths-organisation-security\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":4,"has_code_examples":false,"has_tutorial_content":false,"is_research_content":false},{"article_id":33655,"title":"Understanding AI: Google Brain Scientist Been Kim Is Developing An AI Translator","content":"As human beings, we need a reason for each and everything that we do. But what can we do about that as we all are programmed to understand the mechanics behind everything that we possibly can? Explaining the phenomenon in a similar way, a research scientist at Google Brain, Been Kim, gives a thumbs up to the idea that one should expect nothing less than that from artificial intelligence as well. As an interpretable machine learning expert, her dream as of now is to build an AI software that can explain itself to anyone. Kim believes that AI is at a critical moment where the whole humankind is trying to decide whether this technology is good for us or not. “If we don’t solve this problem of interpretability, I don’t think we’re going to move forward with this technology as we will be forced to drop AI as technology in future,” she said. This researcher with Google Brain has mitigated this problem through translation and interpretability research to bridge the gap between AI and humans. Testing with Concept Activation Vectors (TCAV) — a system that has been developed by her along with her team — she describes it as the translator for humans that allows a user to ask a black box AI how much a specific, high-level concept has been played into its final reasoning that it has done. For example, if a machine-learning model is trained to identify zebras in images, a person can use TCAV to determine how much weight the system gives to the concept of stripes when deciding. While the TCAV was normally designed and tested for image recognition AI, it can be adjusted for use with another AI model too. Nice article about @_beenkim at @GoogleAI. Her work on explainable AI is a great. Lots more work like this needed to ensure people both understand the technology and can use it safely.https:\/\/t.co\/U6B0kcDmpj — Mustafa Suleyman (@mustafasuleyman) January 11, 2019 Key Features Of TCAV TCAV was for the first time was tested on machine-learning models trained to recognize images, but it works with other models as well that are trained on text and certain kinds of data visualizations, like EEG waveforms. TCAV can be plugged into machine learning algorithms to get out the information, of how much they weighted different factors or types of data before churning out results Tools like TCAV are in high demand as AI finds itself under greater scrutiny for the racial and gender bias that plagues artificial intelligence and the training data used to develop it. With TCAV, people using a facial recognition algorithm would be able to determine also how much it factored in the race when, say, matching up people against a database of known criminals or evaluating their job applications. So, TCAV also enables people at large to have the choice to question, reject and even fix a neural network’s conclusions rather than blindly trusting the machine to be objective and fair. How TCAV Works TCAV can be also used to ask a trained model about irrelevant concepts. For example, doctors using AI to make cancer predictions, the doctors might suddenly think. It looks like the machine is giving positive predictions of cancer for a lot of images that have a kind of bluish coloured spots. But that’s just an absurd analysis to be made. But a proficient Doctor would want to know how much bluish spots mattered to the model in making positive predictions of cancer. So, for that, some images are collected say 20. Now, these images will be plugged to those labelled examples into the model. Then TCAV engages in an internal process called sensitivity testing. A numerical output which will be given as a number between zero and one, which thereby will give the probability of a positive prediction for whether cancer has increased or not. That’s your TCAV score. If the probability is increased, it means it has just identified a problem in their machine-learning model. Outlook Interpretability as such has two major divisions and one of them deals with the interpretability for science. The second type being interpretability that’s responsible for AI. The goal of this second branch of interpretability is to simply understand a tool so that it can be used safely. And that understanding can be created only if its confirmed that relevant, useful human knowledge is reflected in that tool that’s being used. This is what the Google Brain scientist Been Kim has accomplished by building a tool that can help artificial intelligence systems explain how they arrived at final conclusions, a task that’s nearly impossible for ML algorithms to gauge. Though the project is in its developmental stage, Kim is of the opinion that there is no need for a tool as such to totally explain AI’s decision-making process. But, it’s good enough for now to have something that can flag potential issues and give human beings the much-needed insight into where something may or may have gone wrong. The goal of interpretability for machine learning for her is to tell if a system is safe to use or not. It’s about ultimately revealing the truth.","excerpt":"As human beings, we need a reason for each and everything that we do. But what can we do about that as we all are programmed to understand the mechanics behind everything that we possibly can? Explaining the phenomenon in a similar way, a research scientist at Google Brain, Been Kim, gives a thumbs up […]","categories":["Global Tech"],"tags":["AI (Artificial Intelligence)","AI Research","Google","Google Brain","Machine Learning"],"author_name":"Martin F.R.","publish_date":"2019-01-17T11:16:23","publication_year":"2019","word_count":859,"keywords":["Go","artificial intelligence","machine learning","TPU","AI","neural network","ML","Google Brain","Machine Learning","AI Research","image recognition","Google","Rust","R","AI (Artificial Intelligence)"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","neural network","image recognition","TPU","R","Go","Rust"],"url":"https:\/\/analyticsindiamag.com\/global-tech\/google-brain-been-kim-ai-translator\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":0,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true},{"article_id":31044,"title":"Key Business Factors SMEs Should Consider Before Deploying AI","content":"Tech giants like Google and Microsoft are working relentlessly to make artificial intelligence inclusive, building enterprise-grade AI solutions. However, many experts are questioning how small and medium-sized enterprises can wield a similar advantage and implement machine learning strategies. Enterprises incorporate ML into their core processes for a variety of strategic reasons. For example, ML can deliver benefits like the ability to discover patterns and substantially improve customer segmentation and targeting. Beyond the business outcomes, SMEs need to consider a range of factors in order to evaluate and maximise the return on investment. One of the biggest roadblocks is that ML projects require a significant investment and even small projects can cost up to hundreds of thousands of dollars. At an enterprise level, these projects cost a few million dollars. And a chunk of IT budget is spent on understanding how to allocate the revenue for procuring and implementing ML applications. Some of the key factors considered for IT budget are: Vendor-specific cost structure Data availability Project duration Whether the solution will be built in-house or outsourced Essentially, ML is a long-term investment which can start to deliver benefits only after a few years. Key Factors SMEs Need To Consider Before Leveraging AI Identifying The Right Problem To Be Solved: Some of the key questions that need to be dealt with are what are the right problems to solve, what are the kind of benchmarks that should be established and does the company have the right data available. Another question is who’s the right vendor and does the project require continued investment. Right Third Party Vendors Software And Tools: Most enterprises have a dedicated team of data scientists and developers to build ML\/AI applications or do it through third-party vendor tools. Leading tech majors like Google, AWS and Microsoft offer API platforms for tasks like speech recognition, image and text recognition. For example, SMEs can choose from a range of tools such as Google’s TensorFlow ML tools or IBM’s Watson AI platform. Meanwhile, there is also a range of open source platforms which provides the building blocks to develop ML applications. How To Measure Business Impact: ML and AI require petabytes of data as inputs to train models, which is a significant investment. Then, the next step entails using the right algorithms and developing a working model which requires many iterations to churn out efficient business outcomes. The business impact is measured in terms of the type of use case being developed, the success of the project’s implementation and how well it scaled. Building In-house Talent or Adopting A Hybrid Approach: SMEs will also have to make the choice between building an in-house team of training data scientists and software developers who can build models in-house or outsource the model development task. Most SMEs follow a hybrid approach to reduce costs and drive better business outcomes. However, building an in-house team requires massive investment as data science talent is scarce and it can take up to 12 months to get the projects off the ground. Not All ML Projects Will Be Successful: Before investing in an ML project, it is important to recognise that not all ML projects will be successful. In fact, enterprises are known to adopt the “fail first approach” which is also a telling comment on the iterative nature of ML project development stages which includes prototype development, testing and analysis. Besides model development, testing and validation, also form a major component of the project. ML models need to be fine-tuned and the results are validated thoroughly before being put in use. For example, in the case of building a financial robot adviser, the results need to be validated against compliance regulations. Outlook Before investing in ML and AI capabilities, SMEs should be well-informed about the opportunities so that the Tier-1 management team can make an appropriate decision about where to invest, which vendor software to use and whether emerging technologies will benefit their business and help the company as a leader in its space. While there are several examples of ML\/AI projects delivering up to three times the cost of investment, not all projects will be successful. There are a section of use cases (for example, chatbots) which have the potential to produce much greater results in the long run.","excerpt":"Tech giants like Google and Microsoft are working relentlessly to make artificial intelligence inclusive, building enterprise-grade AI solutions. However, many experts are questioning how small and medium-sized enterprises can wield a similar advantage and implement machine learning strategies. Enterprises incorporate ML into their core processes for a variety of strategic reasons. For example, ML can […]","categories":["AI Features"],"tags":["Machine Learning","ML projects","SMEs"],"author_name":"Richa Bhatia","publish_date":"2018-12-04T12:23:36","publication_year":"2018","word_count":713,"keywords":["data science","artificial intelligence","machine learning","AWS","AI","SMEs","chatbots","ML","Machine Learning","RAG","ML projects","TensorFlow","R"],"extracted_tech_keywords":["AI","artificial intelligence","machine learning","ML","data science","TensorFlow","RAG","chatbots","AWS","R"],"url":"https:\/\/analyticsindiamag.com\/ai-features\/key-business-factors-smes-should-consider-before-deploying-ai\/","complexity_score":3,"technical_depth":10,"industry_relevance_score":2,"has_code_examples":false,"has_tutorial_content":true,"is_research_content":true}]